[kernel] r12535 - in dists/trunk/linux-2.6: . debian debian/bin debian/config debian/config/alpha/vserver debian/config/amd64 debian/config/amd64/vserver debian/config/amd64/xen debian/config/arm debian/config/featureset-openvz debian/config/featureset-vserver debian/config/featureset-xen debian/config/hppa debian/config/i386 debian/config/i386/openvz debian/config/i386/vserver debian/config/i386/xen debian/config/ia64 debian/config/ia64/vserver debian/config/powerpc debian/config/powerpc/vserver debian/config/s390 debian/config/sparc debian/lib/python/debian_linux debian/patches/bugfix/all debian/patches/bugfix/all/stable debian/patches/bugfix/alpha debian/patches/bugfix/arm debian/patches/bugfix/parisc debian/patches/bugfix/powerpc debian/patches/bugfix/s390 debian/patches/bugfix/x86 debian/patches/debian debian/patches/features/all/vserver debian/patches/features/all/xen debian/patches/features/arm debian/patches/features/powerpc debian/patches/features/sparc debian/patches/features/x86 debian/patches/series debian/templates
Bastian Blank
waldi at alioth.debian.org
Sun Dec 28 16:55:53 UTC 2008
Author: waldi
Date: Sun Dec 28 16:55:48 2008
New Revision: 12535
Log:
Merge /dists/sid/linux-2.6.
Added:
dists/trunk/linux-2.6/debian/config/featureset-vserver/defines
dists/trunk/linux-2.6/debian/config/featureset-xen/defines
dists/trunk/linux-2.6/debian/config/i386/xen/config
dists/trunk/linux-2.6/debian/patches/bugfix/all/ext3-add-support-for-non-native-signed-unsigned-htr.patch
dists/trunk/linux-2.6/debian/patches/bugfix/all/ext4-add-support-for-non-native-signed-unsigned-htr.patch
dists/trunk/linux-2.6/debian/patches/bugfix/all/sata_mv-clear_irq.patch
dists/trunk/linux-2.6/debian/patches/bugfix/all/wireless-regulatory-default-EU.patch
dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_ignore_pci.patch
dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_powerdown.patch
dists/trunk/linux-2.6/debian/patches/bugfix/parisc/
dists/trunk/linux-2.6/debian/patches/bugfix/parisc/add-lockless-fw-funcs.patch
dists/trunk/linux-2.6/debian/patches/bugfix/parisc/hijack-jump-to-start_kernel.patch
dists/trunk/linux-2.6/debian/patches/bugfix/s390/
dists/trunk/linux-2.6/debian/patches/bugfix/s390/prevent-ptrace-padding-area-read-write-in-31-bit-mode.patch
dists/trunk/linux-2.6/debian/patches/bugfix/x86/
dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device-1.patch
dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device.patch
dists/trunk/linux-2.6/debian/patches/debian/drivers-ata-pata_sis-postpone-pata.patch
dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35-update.patch
dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35.patch (contents, props changed)
dists/trunk/linux-2.6/debian/patches/features/all/xen/console-hvc-overtake.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/disable-pat.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/dom0-fix-processor-throttling-when-pr-id-is-minus-1.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/fix-pci-hook.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/hardcode-xen-makefile.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/remove-4gb-warning.patch
dists/trunk/linux-2.6/debian/patches/features/all/xen/suse-20080808143035.patch
dists/trunk/linux-2.6/debian/patches/features/arm/dns323_mac_addr_read.patch
dists/trunk/linux-2.6/debian/patches/features/arm/dns323_revb1.patch
dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-mem-leak.patch
dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-workqueque.patch
dists/trunk/linux-2.6/debian/patches/features/arm/ts409-export-reset.patch
dists/trunk/linux-2.6/debian/patches/features/sparc/
dists/trunk/linux-2.6/debian/patches/features/sparc/video-sunxvr500-intergraph.patch
dists/trunk/linux-2.6/debian/patches/features/x86/drivers-hp_ilo-fixup.patch
Removed:
dists/trunk/linux-2.6/debian/patches/bugfix/alpha/
dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/mv643xx-hotplug-support.patch
dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/serial.patch
dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.34.11.patch
dists/trunk/linux-2.6/debian/patches/features/powerpc/
Modified:
dists/trunk/linux-2.6/ (props changed)
dists/trunk/linux-2.6/debian/bin/ (props changed)
dists/trunk/linux-2.6/debian/bin/abicheck.py
dists/trunk/linux-2.6/debian/bin/gencontrol.py
dists/trunk/linux-2.6/debian/changelog
dists/trunk/linux-2.6/debian/config/alpha/vserver/defines
dists/trunk/linux-2.6/debian/config/amd64/config
dists/trunk/linux-2.6/debian/config/amd64/defines
dists/trunk/linux-2.6/debian/config/amd64/vserver/defines
dists/trunk/linux-2.6/debian/config/amd64/xen/config
dists/trunk/linux-2.6/debian/config/amd64/xen/defines
dists/trunk/linux-2.6/debian/config/arm/config.iop32x
dists/trunk/linux-2.6/debian/config/arm/config.ixp4xx
dists/trunk/linux-2.6/debian/config/arm/config.orion5x
dists/trunk/linux-2.6/debian/config/arm/config.versatile
dists/trunk/linux-2.6/debian/config/config
dists/trunk/linux-2.6/debian/config/defines
dists/trunk/linux-2.6/debian/config/featureset-openvz/config
dists/trunk/linux-2.6/debian/config/featureset-vserver/config
dists/trunk/linux-2.6/debian/config/featureset-xen/config
dists/trunk/linux-2.6/debian/config/hppa/config
dists/trunk/linux-2.6/debian/config/i386/config
dists/trunk/linux-2.6/debian/config/i386/config.686-bigmem
dists/trunk/linux-2.6/debian/config/i386/defines
dists/trunk/linux-2.6/debian/config/i386/openvz/defines
dists/trunk/linux-2.6/debian/config/i386/vserver/defines
dists/trunk/linux-2.6/debian/config/i386/xen/defines
dists/trunk/linux-2.6/debian/config/ia64/defines
dists/trunk/linux-2.6/debian/config/ia64/vserver/defines
dists/trunk/linux-2.6/debian/config/powerpc/config
dists/trunk/linux-2.6/debian/config/powerpc/config.powerpc64
dists/trunk/linux-2.6/debian/config/powerpc/vserver/defines
dists/trunk/linux-2.6/debian/config/s390/defines
dists/trunk/linux-2.6/debian/config/sparc/config
dists/trunk/linux-2.6/debian/config/sparc/defines
dists/trunk/linux-2.6/debian/lib/python/debian_linux/config.py
dists/trunk/linux-2.6/debian/lib/python/debian_linux/patches.py
dists/trunk/linux-2.6/debian/patches/bugfix/all/stable/gen-patch
dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/lpar-console.patch
dists/trunk/linux-2.6/debian/patches/debian/scripts-kconfig-reportoldconfig.patch
dists/trunk/linux-2.6/debian/patches/features/all/vserver/bindmount-dev.patch
dists/trunk/linux-2.6/debian/patches/features/all/vserver/gen-patch
dists/trunk/linux-2.6/debian/patches/series/1~experimental.1
dists/trunk/linux-2.6/debian/patches/series/1~experimental.1-extra
dists/trunk/linux-2.6/debian/rules.real
dists/trunk/linux-2.6/debian/templates/control.xen-linux-system.in
Modified: dists/trunk/linux-2.6/debian/bin/abicheck.py
==============================================================================
--- dists/trunk/linux-2.6/debian/bin/abicheck.py (original)
+++ dists/trunk/linux-2.6/debian/bin/abicheck.py Sun Dec 28 16:55:48 2008
@@ -1,5 +1,6 @@
#!/usr/bin/env python
+import fnmatch
import sys
sys.path.append('debian/lib/python')
@@ -91,11 +92,21 @@
return ret
def _ignore(self, add, change, remove):
- config = self.config.merge('abi', self.arch, self.featureset, self.flavour)
- ignores = config.get('ignore-changes', None)
- if ignores is None:
- return set()
- return set(ignores.split())
+ all = set(add.keys() + change.keys() + remove.keys())
+ # TODO: let config merge this lists
+ configs = []
+ configs.append(self.config.get(('abi', self.arch, self.featureset, self.flavour), {}))
+ configs.append(self.config.get(('abi', self.arch, None, self.flavour), {}))
+ configs.append(self.config.get(('abi', self.arch, self.featureset), {}))
+ configs.append(self.config.get(('abi', self.arch), {}))
+ configs.append(self.config.get(('abi',), {}))
+ ignores = set()
+ for config in configs:
+ ignores.update(config.get('ignore-changes', []))
+ filtered = set()
+ for m in ignores:
+ filtered.update(fnmatch.filter(all, m))
+ return filtered
if __name__ == '__main__':
sys.exit(checker(*sys.argv[1:])(sys.stdout))
Modified: dists/trunk/linux-2.6/debian/bin/gencontrol.py
==============================================================================
--- dists/trunk/linux-2.6/debian/bin/gencontrol.py (original)
+++ dists/trunk/linux-2.6/debian/bin/gencontrol.py Sun Dec 28 16:55:48 2008
@@ -144,10 +144,10 @@
image_fields['Conflicts'].append(PackageRelationGroup([a]))
image_fields['Depends'].append(l_depends)
- if 'desc-parts' in config_entry_image:
+ desc_parts = self.config.get_merge('image', arch, featureset, flavour, 'desc-parts')
+ if desc_parts:
desc = image_fields['Description']
- parts = config_entry_image['desc-parts']
- for part in parts:
+ for part in desc_parts[::-1]:
desc.append(config_entry_image['desc-long-part-' + part])
desc.append_short(config_entry_image.get('desc-short-part-' + part, ''))
@@ -164,9 +164,12 @@
if config_entry_xen.get('dom0-support', True):
p = self.process_packages(self.templates['control.xen-linux-system'], vars)
l = PackageRelationGroup()
- for version in config_entry_xen['versions']:
- l.append("xen-hypervisor-%s-%s" % (version, config_entry_xen['flavour']))
- makeflags['XEN_VERSIONS'] = ' '.join(['%s-%s' % (i, config_entry_xen['flavour']) for i in config_entry_xen['versions']])
+ xen_versions = []
+ for xen_flavour in config_entry_xen['flavours']:
+ for version in config_entry_xen['versions']:
+ l.append("xen-hypervisor-%s-%s" % (version, xen_flavour))
+ xen_versions.append('%s-%s' % (version, xen_flavour))
+ makeflags['XEN_VERSIONS'] = ' '.join(xen_versions)
p[0]['Depends'].append(l)
packages_dummy.extend(p)
else:
Modified: dists/trunk/linux-2.6/debian/changelog
==============================================================================
--- dists/trunk/linux-2.6/debian/changelog (original)
+++ dists/trunk/linux-2.6/debian/changelog Sun Dec 28 16:55:48 2008
@@ -37,7 +37,601 @@
-- dann frazier <dannf at debian.org> Fri, 29 Aug 2008 17:02:18 -0600
-linux-2.6 (2.6.26-1~experimental.1) UNRELEASED; urgency=low
+linux-2.6 (2.6.26-12) unstable; urgency=high
+
+ [ Ian Campbell ]
+ * xen: fix ACPI processor throttling for when processor id is -1. (closes: #502849)
+
+ [ dann frazier ]
+ * Make sendmsg() block during UNIX garbage collection (CVE-2008-5300)
+ * Fix race conditions between inotify removal and umount (CVE-2008-5182)
+ * Fix DoS when calling svc_listen twice on the same socket while reading
+ /proc/net/atm/*vc (CVE-2008-5079)
+
+ [ Bastian Blank ]
+ * [openvz, vserver] Fix descriptions.
+ * [sparc] Enable Sun Logical Domains support. (closes: #501684)
+ * Fix coexistence of pata_marvell and ahci. (closes: #507432)
+ * [sparc] Support Intergraph graphics chips. (closes: #508108)
+
+ -- Bastian Blank <waldi at debian.org> Mon, 15 Dec 2008 12:57:18 +0100
+
+linux-2.6 (2.6.26-11) unstable; urgency=low
+
+ [ Bastian Blank ]
+ * [sparc] Reintroduce dummy PCI host controller to workaround broken X.org.
+ * [sparc] Fix size checks in PCI maps.
+ * Add stable release 2.6.26.8:
+ - netfilter: restore lost ifdef guarding defrag exception
+ - netfilter: snmp nat leaks memory in case of failure
+ - netfilter: xt_iprange: fix range inversion match
+ - ACPI: dock: avoid check _STA method
+ - ACPI: video: fix brightness allocation
+ - sparc64: Fix race in arch/sparc64/kernel/trampoline.S
+ - math-emu: Fix signalling of underflow and inexact while packing result.
+ - tcpv6: fix option space offsets with md5
+ - net: Fix netdev_run_todo dead-lock
+ - scx200_i2c: Add missing class parameter
+ - DVB: s5h1411: Power down s5h1411 when not in use
+ - DVB: s5h1411: Perform s5h1411 soft reset after tuning
+ - DVB: s5h1411: bugfix: Setting serial or parallel mode could destroy bits
+ - V4L: pvrusb2: Keep MPEG PTSs from drifting away
+ - ACPI: Always report a sync event after a lid state change
+ - ALSA: use correct lock in snd_ctl_dev_disconnect()
+ - file caps: always start with clear bprm->caps_*
+ - libertas: fix buffer overrun
+ - net: Fix recursive descent in __scm_destroy().
+ - SCSI: qla2xxx: Skip FDMI registration on ISP21xx/22xx parts.
+ (Closes: #502552)
+ - edac cell: fix incorrect edac_mode
+ - ext[234]: Avoid printk floods in the face of directory corruption
+ (CVE-2008-3528)
+ - gpiolib: fix oops in gpio_get_value_cansleep()
+ * Override ABI changes.
+ * [xen] Update description. (closes: #505961)
+ * Revert parts of 2.6.26.6 to fix resume breakage. (closes: #504167)
+ - clockevents: prevent multiple init/shutdown
+ - clockevents: broadcast fixup possible waiters
+
+ [ dann frazier ]
+ * Fix buffer overflow in hfsplus (CVE-2008-4933)
+ * Fix BUG() in hfsplus (CVE-2008-4934)
+ * Fix stack corruption in hfs (CVE-2008-5025)
+ * Fix oops in tvaudio when controlling bass/treble (CVE-2008-5033)
+
+ [ Martin Michlmayr ]
+ * [arm/iop32x, arm/ixp4xx, arm/orion5x] Enable support for more partition
+ tables, including MAC_PARTITION (requested by Benoît Knecht).
+ * leds-pca9532: Fix memory leak and properly handle errors (Sven Wegener)
+ * leds-pca9532: Move i2c work to a workqueque (Riku Voipio). (closes:
+ #506116)
+
+ -- Bastian Blank <waldi at debian.org> Wed, 26 Nov 2008 11:43:48 +0100
+
+linux-2.6 (2.6.26-10) unstable; urgency=low
+
+ [ dann frazier ]
+ * sctp: Fix possible kernel panic in sctp_sf_abort_violation (CVE-2008-4618)
+
+ [ Martin Michlmayr ]
+ * DNS-323: add support for revision B1 machines (Matthew Palmer).
+ * ext3/ext4: Add support for non-native signed/unsigned htree hash
+ algorithms (Theodore Ts'o). (closes: #493957)
+ * [arm/ixp4xx] Enable USB_ACM (closes: #504723).
+
+ [ Bastian Blank ]
+ * agp: Fix stolen memory counting on Intel G4X. (closes: #502606)
+ * Add stable release 2.6.26.7:
+ - security: avoid calling a NULL function pointer in drivers/video/tvaudio.c
+ - DVB: au0828: add support for another USB id for Hauppauge HVR950Q
+ - drm/i915: fix ioremap of a user address for non-root (CVE-2008-3831)
+ - ACPI: Ignore _BQC object when registering backlight device
+ - hwmon: (it87) Prevent power-off on Shuttle SN68PT
+ - Check mapped ranges on sysfs resource files
+ - x86: avoid dereferencing beyond stack + THREAD_SIZE
+ - PCI: disable ASPM on pre-1.1 PCIe devices
+ - PCI: disable ASPM per ACPI FADT setting
+ - V4L/DVB (9053): fix buffer overflow in uvc-video
+ - V4L/DVB (8617): uvcvideo: don't use stack-based buffers for USB transfers.
+ - V4L/DVB (8498): uvcvideo: Return sensible min and max values when querying
+ a boolean control.
+ - V4L: zr36067: Fix RGBR pixel format
+ - V4L: bttv: Prevent NULL pointer dereference in radio_open
+ - libata: fix EH action overwriting in ata_eh_reset()
+ - libata: always do follow-up SRST if hardreset returned -EAGAIN
+ - fbcon_set_all_vcs: fix kernel crash when switching the rotated consoles
+ - modules: fix module "notes" kobject leak
+ - b43legacy: Fix failure in rate-adjustment mechanism
+ - CIFS: make sure we have the right resume info before calling CIFSFindNext
+ - sched_rt.c: resch needed in rt_rq_enqueue() for the root rt_rq
+ - tty: Termios locking - sort out real_tty confusions and lock reads
+ - x86, early_ioremap: fix fencepost error
+ - x86: improve UP kernel when CPU-hotplug and SMP is enabled
+ - x86: Reserve FIRST_DEVICE_VECTOR in used_vectors bitmap.
+ * [xen] Remove pte file workaround.
+
+ [ Ian Campbell ]
+ * [xen] Disable usage of PAT. (closes: #503821)
+
+ -- Bastian Blank <waldi at debian.org> Sat, 08 Nov 2008 10:50:58 +0100
+
+linux-2.6 (2.6.26-9) unstable; urgency=low
+
+ [ Bastian Blank ]
+ * Add stable release 2.6.26.6:
+ - mm owner: fix race between swapoff and exit
+ - rtc: fix kernel panic on second use of SIGIO nofitication
+ - fbcon: fix monochrome color value calculation
+ - ALSA: snd-powermac: HP detection for 1st iMac G3 SL
+ - ALSA: snd-powermac: mixers for PowerMac G4 AGP
+ - sparc64: Fix missing devices due to PCI bridge test in
+ of_create_pci_dev().
+ - sparc64: Fix disappearing PCI devices on e3500.
+ - sparc64: Fix OOPS in psycho_pcierr_intr_other().
+ - sparc64: Fix interrupt register calculations on Psycho and Sabre.
+ - sparc64: Fix PCI error interrupt registry on PSYCHO.
+ - udp: Fix rcv socket locking
+ - sctp: Fix oops when INIT-ACK indicates that peer doesn't support AUTH
+ (CVE-2008-4576)
+ - sctp: do not enable peer features if we can't do them.
+ - ipsec: Fix pskb_expand_head corruption in xfrm_state_check_space
+ - netlink: fix overrun in attribute iteration
+ - niu: panic on reset
+ - ipv6: Fix OOPS in ip6_dst_lookup_tail().
+ - XFRM,IPv6: initialize ip6_dst_blackhole_ops.kmem_cachep
+ - af_key: Free dumping state on socket close
+ - pcmcia: Fix broken abuse of dev->driver_data
+ - clockevents: remove WARN_ON which was used to gather information
+ - ntp: fix calculation of the next jiffie to trigger RTC sync
+ - x86: HPET: read back compare register before reading counter
+ - x86: HPET fix moronic 32/64bit thinko
+ - clockevents: broadcast fixup possible waiters
+ - HPET: make minimum reprogramming delta useful
+ - clockevents: prevent endless loop lockup
+ - clockevents: prevent multiple init/shutdown
+ - clockevents: enforce reprogram in oneshot setup
+ - clockevents: prevent endless loop in periodic broadcast handler
+ - clockevents: prevent clockevent event_handler ending up handler_noop
+ - x86: fix memmap=exactmap boot argument
+ - x86: add io delay quirk for Presario F700
+ - ACPI: Avoid bogus EC timeout when EC is in Polling mode
+ - x86: fix SMP alternatives: use mutex instead of spinlock, text_poke is
+ sleepable
+ - rtc: fix deadlock
+ - mm: dirty page tracking race fix
+ - x86-64: fix overlap of modules and fixmap areas
+ - x86: PAT proper tracking of set_memory_uc and friends
+ - x86: fix oprofile + hibernation badness
+ - x86: fdiv bug detection fix
+ - rt2x00: Use ieee80211_hw->workqueue again
+ - x86: Fix 27-rc crash on vsmp due to paravirt during module load
+ - sg: disable interrupts inside sg_copy_buffer
+ - ocfs2: Increment the reference count of an already-active stack.
+ - APIC routing fix
+ - sched: fix process time monotonicity
+ - block: submit_bh() inadvertently discards barrier flag on a sync write
+ - x64, fpu: fix possible FPU leakage in error conditions
+ - x86-64: Clean up save/restore_i387() usage
+ - KVM: SVM: fix guest global tlb flushes with NPT
+ - KVM: SVM: fix random segfaults with NPT enabled
+ - ALSA: remove unneeded power_mutex lock in snd_pcm_drop
+ - ALSA: fix locking in snd_pcm_open*() and snd_rawmidi_open*()
+ - ALSA: oxygen: fix distorted output on AK4396-based cards
+ - ALSA: hda - Fix model for Dell Inspiron 1525
+ - SCSI: qla2xxx: Defer enablement of RISC interrupts until ISP
+ initialization completes.
+ - USB: fix hcd interrupt disabling
+ - smb.h: do not include linux/time.h in userspace
+ - pxa2xx_spi: fix build breakage
+ - pxa2xx_spi: chipselect bugfixes
+ - pxa2xx_spi: dma bugfixes
+ - mm: mark the correct zone as full when scanning zonelists
+ - async_tx: fix the bug in async_tx_run_dependencies
+ - drivers/mmc/card/block.c: fix refcount leak in mmc_block_open()
+ - ixgbe: initialize interrupt throttle rate
+ - i2c-dev: Return correct error code on class_create() failure
+ - x86-32: AMD c1e force timer broadcast late
+ * [x86] Update patch to detect not properly announced cmos RTC devices.
+ * [xen] Overtake hvc console by default.
+
+ [ maximilian attems ]
+ * [openvz] ip: NULL pointer dereferrence in tcp_v(4|6)_send_ack
+ (closes: #500472)
+ * [openvz] unset NF_CONNTRACK_IPV6 for now until abi bump.
+
+ [ Stephen R. Marenka ]
+ * [m68k] add patches to fix atari ethernec per Michael Schmitz:
+ atari-ethernec-IRQF_SHARED.diff and atari-ethernec-fixes.diff.
+ * [m68k] add mac-esp-fix-for-quadras-with-two-esp-chips.diff to fix macs
+ with dual scsi busses and a problem with xorg, per Finn Thain.
+ * [m68k] add atari-atari_keyb_init-operator-precedence.diff per
+ Michael Schmitz.
+ * [m68k] more mac patches, per Finn Thain.
+
+ [ Martin Michlmayr ]
+ * [arm/ixp4xx] Enable USB_ATM and USB_SPEEDTOUCH (closes: #502182).
+ * [arm/iop32x, arm/orion5x] Likewise.
+ * DNS-323: read MAC address from flash (Matthew Palmer).
+
+ [ dann frazier ]
+ * Restrict access to the DRM_I915_HWS_ADDR ioctl (CVE-2008-3831)
+ * Don't allow splicing to files opened with O_APPEND (CVE-2008-4554)
+
+ -- Bastian Blank <waldi at debian.org> Sat, 18 Oct 2008 12:14:22 +0200
+
+linux-2.6 (2.6.26-8) unstable; urgency=medium
+
+ [ dann frazier ]
+ * [x86] Fix broken LDT access in VMI (CVE-2008-4410)
+ * ata: Fix off-by-one-error that causes errors when reading a
+ block on the LBA28-LBA48 boundary
+ * [s390] prevent ptrace padding area read/write in 31-bit mode
+ (CVE-2008-1514)
+
+ [ Bastian Blank ]
+ * Fix generation of i386 Xen image information.
+ * [i386] Restrict the usage of long NOPs. (closes: #464962)
+ * Fix access to uninitialized user keyring. (closes: #500279)
+ * [x86] Fix detection of non-PNP RTC devices. (closes: #499230)
+
+ -- Bastian Blank <waldi at debian.org> Thu, 09 Oct 2008 12:07:21 +0200
+
+linux-2.6 (2.6.26-7) unstable; urgency=low
+
+ [ Bastian Blank ]
+ * [xen] Add SuSE Xen patch. (closes: #495895)
+ * Only register notifiers in braille console if used, fixes Insert key.
+ (closes: #494374)
+ * Fix ACPI EC GPE storm detection. (closes: #494546)
+ * Disable useless support for ISP1760 USB host controller.
+ (closes: #498304)
+ * rt61pci: Add a sleep after firmware upload. (closes: #498828)
+
+ [ Stephen R. Marenka ]
+ * [m68k] Set CONFIG_ATARI_ETHERNEC=m for atari, since it only works
+ in modular form.
+ * [m68k] Enable CONFIG_ADB_PMU68K=y for mac.
+ * [m68k] Add atari-aranym-nf-wrappers.diff patch to fix atari LBD
+ problems, set CONFIG_LBD=y for atari.
+
+ [ Martin Michlmayr ]
+ * [arm/orion5x] Enable CONFIG_ATALK (requested by Ben Schwarz).
+ * [arm/versatile] Enable CONFIG_VFP. (closes: #499463)
+ * ath5k: Fix bad udelay calls on AR5210 code (Nick Kossifidis).
+ * [arm] No longer disable ATH5K.
+
+ [ dann frazier ]
+ * Add missing capability checks in sbni_ioctl (CVE-2008-3525)
+
+ -- Bastian Blank <waldi at debian.org> Wed, 01 Oct 2008 09:02:30 +0200
+
+linux-2.6 (2.6.26-6) unstable; urgency=low
+
+ [ maximilian attems ]
+ * [openvz] Enable checkpointing. (closes: #497292)
+
+ [ Bastian Blank ]
+ * Allow forced module loading again. (closes: #494144)
+ * Set IEEE 802.11 (wireless) regulatory domain default to EU.
+ (closes: #497971)
+ * [i386] Enable IDE ACPI support. Override ABI changes. (closes: #470528)
+ * [i386/686-bigmem] Promote to generic subarch. (closes: #476120)
+
+ [ Martin Michlmayr ]
+ * Fix dead 21041 ethernet after ifconfig down (Thomas Bogendoerfer).
+
+ [ dann frazier ]
+ * [hppa] Enable the FPU before using it, fixes booting on A500s
+ with our CONFIG_PRINTK_TIME=y setting. (closes: #499458)
+
+ -- Bastian Blank <waldi at debian.org> Wed, 24 Sep 2008 12:06:47 +0200
+
+linux-2.6 (2.6.26-5) unstable; urgency=low
+
+ [ Martin Michlmayr ]
+ * Backport power-off method for Kurobox Pro.
+ * [arm/versatile] Really enable CONFIG_RTC_DRV_PL031 (closes: #484432).
+
+ [ Stephen R. Marenka ]
+ * [m68k] Set CONFIG_LBD=n for atari, since it conflicts with nfblock.
+
+ [ Bastian Blank ]
+ * Reenable SiS SATA support. (closes: #496603)
+ * [amd64,i386] Disable new-style SiS PATA support.
+ * Add stable release 2.6.26.4:
+ - sata_mv: don't issue two DMA commands concurrently
+ - KVM: MMU: Fix torn shadow pte
+ - x86: work around MTRR mask setting, v2
+ - nfsd: fix buffer overrun decoding NFSv4 acl (CVE-2008-3915)
+ - sunrpc: fix possible overrun on read of /proc/sys/sunrpc/transports
+ (CVE-2008-3911)
+ - r8169: balance pci_map / pci_unmap pair
+ - tg3: Fix firmware event timeouts
+ - crypto: authenc - Avoid using clobbered request pointer
+ - sparc64: Fix cmdline_memory_size handling bugs.
+ - sparc64: Fix overshoot in nid_range().
+ - ipsec: Fix deadlock in xfrm_state management. (closes: #497796)
+ - sctp: fix random memory dereference with SCTP_HMAC_IDENT option.
+ - sctp: correct bounds check in sctp_setsockopt_auth_key
+ - sch_prio: Fix nla_parse_nested_compat() regression
+ - sctp: add verification checks to SCTP_AUTH_KEY option
+ - sctp: fix potential panics in the SCTP-AUTH API.
+ - udp: Drop socket lock for encapsulated packets
+ - pkt_sched: Fix actions referencing
+ - pkt_sched: Fix return value corruption in HTB and TBF.
+ - netns: Add network namespace argument to rt6_fill_node() and
+ ipv6_dev_get_saddr()
+ - ipv6: Fix OOPS, ip -f inet6 route get fec0::1, linux-2.6.26,
+ ip6_route_output, rt6_fill_node+0x175 (CVE-2008-3686)
+ - AX.25: Fix sysctl registration if !CONFIG_AX25_DAMA_SLAVE
+ - mm: make setup_zone_migrate_reserve() aware of overlapping nodes
+ - 8250: improve workaround for UARTs that don't re-assert THRE correctly
+ - rtc_time_to_tm: fix signed/unsigned arithmetic
+ - drivers/char/random.c: fix a race which can lead to a bogus BUG()
+ - cifs: fix O_APPEND on directio mounts
+ - atl1: disable TSO by default
+ - forcedeth: fix checksum flag
+ - bio: fix bio_copy_kern() handling of bio->bv_len
+ - bio: fix __bio_copy_iov() handling of bio->bv_len
+ - ALSA: oxygen: prevent muting of nonexistent AC97 controls
+ - S390 dasd: fix data size for PSF/PRSSD command
+ - x86: fix "kernel won't boot on a Cyrix MediaGXm (Geode)"
+ - x86: work around MTRR mask setting
+ - USB: cdc-acm: don't unlock acm->mutex on error path
+ - binfmt_misc: fix false -ENOEXEC when coupled with other binary handlers
+ - fbdefio: add set_page_dirty handler to deferred IO FB
+ - eeepc-laptop: fix use after free
+ - PCI: fix reference leak in pci_get_dev_by_id()
+ - cramfs: fix named-pipe handling
+ * Override ABI changes.
+ * [hppa] Disable new-style RTC support. Override ABI changes.
+
+ [ maximilian attems ]
+ * openvz: Add upstream fixes up to 24cebf40278cb071ff8b. (closes: #497528)
+
+ -- Bastian Blank <waldi at debian.org> Wed, 10 Sep 2008 12:55:16 +0200
+
+linux-2.6 (2.6.26-4) unstable; urgency=low
+
+ [ maximilian attems ]
+ * x86: Reset ACPI_PROCFS_POWER for Lenny as buggy apps depend on it.
+ (closes: #495541)
+ * x86: ACPI: Fix thermal shutdowns
+ * openvz: Add upstream fixes up to 0f14912e3d2251aff. (closes: #494384)
+ * Add stable release 2.6.26.3:
+ - USB: fix interface unregistration logic
+ - usb-storage: unusual_devs entries for iRiver T10 and Datafab CF+SM reader
+ - usb-serial: don't release unregistered minors
+ - usb-storage: revert DMA-alignment change for Wireless USB
+ - usb-storage: automatically recognize bad residues
+ - USB: ftdi_sio: Add USB Product Id for ELV HS485
+ - qla2xxx: Set an rport's dev_loss_tmo value in a consistent manner.
+ - dccp: change L/R must have at least one byte in the dccpsf_val field
+ (CVE-2008-3276)
+ - KVM: Avoid instruction emulation when event delivery is pending
+ - cs5520: add enablebits checking
+ - acer-wmi: Fix wireless and bluetooth on early AMW0 v2 laptops
+ - USB: usb-storage: quirk around v1.11 firmware on Nikon D4
+ - radeonfb: fix accel engine hangs
+ - radeon: misc corrections
+ - sparc64: Fix global reg snapshotting on self-cpu.
+ - sparc64: Do not clobber %g7 in setcontext() trap.
+ - sparc64: Fix end-of-stack checking in save_stack_trace().
+ - sparc64: Fix recursion in stack overflow detection handling.
+ - sparc64: Make global reg dumping even more useful.
+ - sparc64: Implement IRQ stacks.
+ - sparc64: Handle stack trace attempts before irqstacks are setup.
+ - PCI: Limit VPD length for Broadcom 5708S
+ - ide: it821x in pass-through mode segfaults in 2.6.26-stable
+ - syncookies: Make sure ECN is disabled
+ - USB: ftdi_sio: add support for Luminance Stellaris Evaluation/Development
+ Kits
+ - i2c: Fix NULL pointer dereference in i2c_new_probed_device
+ - SCSI: hptiop: add more PCI device IDs
+ - SCSI: ses: fix VPD inquiry overrun
+ - SCSI: scsi_transport_spi: fix oops in revalidate
+ - CIFS: Fix compiler warning on 64-bit
+ - x86: fix spin_is_contended()
+ - matrox maven: fix a broken error path
+ - i2c: Let users select algorithm drivers manually again
+ - CIFS: properly account for new user= field in SPNEGO upcall string
+ allocation
+ - x86: fix setup code crashes on my old 486 box
+ - KVM: ia64: Fix irq disabling leak in error handling code
+ - mlock() fix return values
+ - rtl8187: Fix lockups due to concurrent access to config routine
+ - KVM: task switch: segment base is linear address
+ - KVM: task switch: use seg regs provided by subarch instead of reading
+ from GDT
+ - KVM: task switch: translate guest segment limit to virt-extension byte
+ granular field
+ - r8169: avoid thrashing PCI conf space above RTL_GIGA_MAC_VER_06
+ - sparc64: FUTEX_OP_ANDN fix
+ - posix-timers: do_schedule_next_timer: fix the setting of ->si_overrun
+ - posix-timers: fix posix_timer_event() vs dequeue_signal() race
+ - vt8623fb: fix kernel oops
+ - ide-cd: fix endianity for the error message in cdrom_read_capacity
+ - qla2xxx: Add dev_loss_tmo_callbk/terminate_rport_io callback support.
+ - random32: seeding improvement
+ - CIFS: mount of IPC$ breaks with iget patch
+ - CIFS: if get root inode fails during mount, cleanup tree connection
+ - crypto: padlock - fix VIA PadLock instruction usage with
+ irq_ts_save/restore()
+ - ipvs: Fix possible deadlock in estimator code
+ - SCSI: block: Fix miscalculation of sg_io timeout in CDROM_SEND_PACKET
+ handler.
+ - ALSA: asoc: restrict sample rate and size in Freescale MPC8610 sound
+ drivers
+ - ALSA: ASoC: fix SNDCTL_DSP_SYNC support in Freescale 8610 sound drivers
+ - USB: pl2023: Remove USB id (4348:5523) handled by ch341
+ - relay: fix "full buffer with exactly full last subbuffer" accounting
+ problem
+ - ipv6: Fix ip6_xmit to send fragments if ipfragok is true
+ - x86: amd opteron TOM2 mask val fix
+
+ [ dann frazier ]
+ * [ia64] Fix boot-time hang w/ PRINTK_TIME by ensuring that cpu0 can access
+ per-cpu vars in early boot
+ * delay calls to sched_clock() until after sched_clock_init() to prevent
+ inaccurate printk timings on ia64 and presumably other architectures
+
+ [ Ian Campbell ]
+ * [xen] import upstream fix to fb-defio driver used by Xen framebuffer.
+
+ [ Bastian Blank ]
+ * [powerpc] Enable proper RTC support. (closes: #484693)
+
+ [ Martin Michlmayr ]
+ * Add Marvell Orion fixes:
+ - sata_mv: add the Gen IIE flag to the SoC devices.
+ - sata_mv: don't avoid clearing interrupt status on SoC host adapters
+
+ [ dann frazier ]
+ * Fix overflow condition in sctp_setsockopt_auth_key (CVE-2008-3526)
+ * Fix panics that may occur if SCTP AUTH is disabled (CVE-2008-3792)
+ * [x86] Fix memory leak in the copy_user routine
+ (CVE-2008-0598, closes: #490910)
+
+ -- Bastian Blank <waldi at debian.org> Thu, 28 Aug 2008 08:46:42 +0200
+
+linux-2.6 (2.6.26-3) unstable; urgency=low
+
+ [ Bastian Blank ]
+ * Disable Emagic Audiowerk 2 soundcard support. The PCI IDs clashes with
+ many DVB cards.
+ * Update VServer patch to 2.3.0.35.
+ * [armel/versatile] Override ABI changes.
+ * [i386/686-bigmem] Add VServer image.
+
+ [ Aurelien Jarno ]
+ * [armel/versatile] Disable CONFIG_NO_HZ, CONFIG_HIGH_RES_TIMERS for
+ dynticks. (closes: #494842)
+
+ [ Martin Michlmayr ]
+ * Fix PCIe on the Kurobox Pro (Lennert Buytenhek).
+ * Fix regressions caused by the "use software GSO for SG+CSUM capable
+ netdevices" patch:
+ - loopback: Enable TSO (Herbert Xu)
+ - net: Preserve netfilter attributes in skb_gso_segment using
+ __copy_skb_header (Herbert Xu)
+
+ [ dann frazier ]
+ * [amd64] Fix typo in TOM2 mask value, preventing a hang on some opteron
+ systems. (closes: #494365)
+
+ -- Bastian Blank <waldi at debian.org> Mon, 18 Aug 2008 15:34:38 +0200
+
+linux-2.6 (2.6.26-2) unstable; urgency=low
+
+ [ Bastian Blank ]
+ * [powerpc] Install arch/powerpc/lib/crtsavres.o into the headers, it is
+ used during module linking.
+ * Add stable release 2.6.26.1:
+ - Fix off-by-one error in iov_iter_advance()
+ - ath5k: don't enable MSI, we cannot handle it yet
+ - b43legacy: Release mutex in error handling code
+ - cpufreq acpi: only call _PPC after cpufreq ACPI init funcs got called already
+ - VFS: increase pseudo-filesystem block size to PAGE_SIZE
+ - markers: fix markers read barrier for multiple probes
+ - tmpfs: fix kernel BUG in shmem_delete_inode
+ - mpc52xx_psc_spi: fix block transfer
+ - ixgbe: remove device ID for unsupported device
+ - UML - Fix boot crash
+ - eCryptfs: use page_alloc not kmalloc to get a page of memory
+ - x86: fix kernel_physical_mapping_init() for large x86 systems
+ - DVB: cx23885: SRAM changes for the 885 and 887 silicon parts
+ - DVB: cx23885: Reallocated the sram to avoid concurrent VIDB/C issues
+ - DVB: cx23885: DVB Transport cards using DVB port VIDB/TS1 did not stream
+ - DVB: cx23885: Ensure PAD_CTRL is always reset to a sensible default
+ - V4L: cx23885: Bugfix for concurrent use of /dev/video0 and /dev/video1
+ - V4L: saa7134: Copy tuner data earlier to avoid overwriting manual tuner type
+ - V4L: uvcvideo: Add support for Medion Akoya Mini E1210 integrated webcam
+ - V4L: uvcvideo: Make input device support optional
+ - V4L: uvcvideo: Don't free URB buffers on suspend
+ - V4L: uvcvideo: Use GFP_NOIO when allocating memory during resume
+ - V4L: uvcvideo: Fix a buffer overflow in format descriptor parsing
+ - DVB: dib0700: add support for Hauppauge Nova-TD Stick 52009
+ - V4L: cx18: Upgrade to newer firmware & update documentation
+ - ALSA: trident - pause s/pdif output
+ - myri10ge: do not use mgp->max_intr_slots before loading the firmware
+ - myri10ge: do not forget to setup the single slice pointers
+ - iop-adma: fix platform driver hotplug/coldplug
+ - sparc64: Do not define BIO_VMERGE_BOUNDARY.
+ - sparc64: Fix cpufreq notifier registry.
+ - sparc64: Fix lockdep issues in LDC protocol layer.
+ - tcp: Clear probes_out more aggressively in tcp_ack().
+ - ARM: fix fls() for 64-bit arguments
+ - vmlinux.lds: move __attribute__((__cold__)) functions back into final .text section
+ - rtc-at91rm9200: avoid spurious irqs
+ - ide-cd: fix oops when using growisofs
+ - x86: fix crash due to missing debugctlmsr on AMD K6-3
+ - cpusets: fix wrong domain attr updates
+ - proc: fix /proc/*/pagemap some more
+ - Fix build on COMPAT platforms when CONFIG_EPOLL is disabled
+ - markers: fix duplicate modpost entry
+ - x86, suspend, acpi: enter Big Real Mode
+ - USB: fix usb serial pm counter decrement for disconnected interfaces
+ - x86 reboot quirks: add Dell Precision WorkStation T5400
+ - Fix typos from signal_32/64.h merge
+ - rcu: fix rcu_try_flip_waitack_needed() to prevent grace-period stall
+ - Patch Upstream: x86 ptrace: fix PTRACE_GETFPXREGS error
+ - KVM: MMU: Fix potential race setting upper shadow ptes on nonpae hosts
+ - KVM: MMU: nuke shadowed pgtable pages and ptes on memslot destruction
+ - KVM: x86 emulator: Fix HLT instruction
+ - KVM: VMX: Add ept_sync_context in flush_tlb
+ - KVM: mmu_shrink: kvm_mmu_zap_page requires slots_lock to be held
+ - KVM: SVM: fix suspend/resume support
+ - KVM: VMX: Fix a wrong usage of vmcs_config
+ - isofs: fix minor filesystem corruption
+ - quota: fix possible infinite loop in quota code
+ - hdlcdrv: Fix CRC calculation.
+ - ipv6: __KERNEL__ ifdef struct ipv6_devconf
+ - ipv6: use timer pending
+ - udplite: Protection against coverage value wrap-around
+ - pxamci: trivial fix of DMA alignment register bit clearing
+ * [sparc] Install asm-sparc headers again.
+ * Force RTC on by default and set clock on startup. Override ABI changes.
+ * [i386, amd64] Make the CMOS RTC support builtin. (closes: #493567)
+ * Add stable release 2.6.26.2:
+ - sound: ensure device number is valid in snd_seq_oss_synth_make_info
+ - Ath5k: kill tasklets on shutdown
+ - Ath5k: fix memory corruption
+ - vfs: fix lookup on deleted directory
+ - ALSA: emu10k1 - Fix inverted Analog/Digital mixer switch on Audigy2
+ - ALSA: hda - Add missing Thinkpad Z60m support
+ - ALSA: hda - Fix DMA position inaccuracy
+ - ALSA: hda - Fix wrong volumes in AD1988 auto-probe mode
+ - Add compat handler for PTRACE_GETSIGINFO
+ - Bluetooth: Signal user-space for HIDP and BNEP socket errors
+ - Input: i8042 - add Acer Aspire 1360 to nomux blacklist
+ - Input: i8042 - add Gericom Bellagio to nomux blacklist
+ - Input: i8042 - add Intel D845PESV to nopnp list
+ - jbd: fix race between free buffer and commit transaction
+ - NFS: Ensure we zap only the access and acl caches when setting new acls
+ - SCSI: ch: fix ch_remove oops
+ - linear: correct disk numbering error check
+ - netfilter: xt_time: fix time's time_mt()'s use of do_div()
+ - Kprobe smoke test lockdep warning
+ - Close race in md_probe
+ - x86: io delay - add checking for NULL early param
+ - x86: idle process - add checking for NULL early param
+ - SCSI: bsg: fix bsg_mutex hang with device removal
+ - netfilter: nf_nat_sip: c= is optional for session
+ - romfs_readpage: don't report errors for pages beyond i_size
+ - ftrace: remove unneeded documentation
+
+ [ Martin Michlmayr ]
+ * METH: fix MAC address setup (Thomas Bogendoerfer)
+ * Export the reset button of the QNAP TS-409.
+ * net: use software GSO for SG+CSUM capable netdevices (Lennert Buytenhek)
+
+ [ dann frazier ]
+ * device_create interface changed between 2.6.26 and 2.6.27; adjust hpilo
+ backport appropriately. Fixes a NULL pointer dereference in ilo_probe().
+
+ -- Bastian Blank <waldi at debian.org> Fri, 08 Aug 2008 08:09:00 +0200
+
+linux-2.6 (2.6.26-1) unstable; urgency=low
* New upstream release see http://kernelnewbies.org/Linux_2_6_26
- UDF 2.50 support. (closes: #480910)
@@ -66,6 +660,7 @@
- alsa snd-hda Dell Inspiron fix (closes: #490649)
- ipw2200: queue direct scans (closes: #487721)
- better gcc-4.3 support (closes: #492301)
+ - iwl3945 monitor mode. (closes: #482387)
[ maximilian attems ]
* topconfig set CRYPTO_CTS, SND_PCSP, SND_AW2, IWL4965_LEDS, IWL3945_LEDS,
@@ -90,7 +685,6 @@
* topconfig: Enable HYSDN, no longer broken on smp.
* Add request_firmware patch for keyspan. (closes: #448900)
* [x86]: Enable dma engine. (closes: #473331)
- * iwl3945: Add monitor mode patch. (closes: #482387)
* [ppc64]: Enable IBMEBUS and EHEA. (closes: #484888)
* topconfig: Enable PROFILING across all flavours. (closes: #484885)
* 486: enable OLPC support thanks Andres Salomon for merge.
@@ -194,7 +788,7 @@
* [arm/versatile] Switch scsi/ext3/smc91x to modules now that we have proper
d-i support. Remove options defined in toplevel config file.
- -- Christian T. Steigies <cts at debian.org> Thu, 12 Jun 2008 15:31:54 +0200
+ -- Bastian Blank <waldi at debian.org> Wed, 30 Jul 2008 10:17:29 +0200
linux-2.6 (2.6.25-7) unstable; urgency=high
Modified: dists/trunk/linux-2.6/debian/config/alpha/vserver/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/alpha/vserver/defines (original)
+++ dists/trunk/linux-2.6/debian/config/alpha/vserver/defines Sun Dec 28 16:55:48 2008
@@ -1,9 +1,6 @@
[base]
flavours: alpha
-[image]
-recommends: util-vserver
-
[alpha_base]
class: Alpha
Modified: dists/trunk/linux-2.6/debian/config/amd64/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/amd64/config (original)
+++ dists/trunk/linux-2.6/debian/config/amd64/config Sun Dec 28 16:55:48 2008
@@ -118,7 +118,7 @@
##
CONFIG_ACPI=y
CONFIG_ACPI_PROCFS=y
-# CONFIG_ACPI_PROCFS_POWER is not set
+CONFIG_ACPI_PROCFS_POWER=y
CONFIG_ACPI_SYSFS_POWER=y
CONFIG_ACPI_PROC_EVENT=y
CONFIG_ACPI_AC=m
@@ -142,6 +142,11 @@
CONFIG_ACPI_SBS=m
##
+## file: drivers/ata/Kconfig
+##
+# CONFIG_PATA_SIS is not set
+
+##
## file: drivers/atm/Kconfig
##
CONFIG_ATM_DRIVERS=y
@@ -260,9 +265,6 @@
CONFIG_PPDEV=m
CONFIG_HVC_XEN=y
CONFIG_NVRAM=m
-CONFIG_RTC=y
-CONFIG_GEN_RTC=m
-CONFIG_GEN_RTC_X=y
CONFIG_DTLK=m
CONFIG_R3964=m
CONFIG_APPLICOM=m
@@ -270,7 +272,6 @@
CONFIG_RAW_DRIVER=m
CONFIG_MAX_RAW_DEVS=256
CONFIG_HPET=y
-CONFIG_HPET_RTC_IRQ=y
CONFIG_HPET_MMAP=y
CONFIG_HANGCHECK_TIMER=m
@@ -1144,6 +1145,11 @@
CONFIG_BATTERY_DS2760=m
##
+## file: drivers/rtc/Kconfig
+##
+CONFIG_RTC_DRV_CMOS=y
+
+##
## file: drivers/scsi/Kconfig
##
CONFIG_SCSI_DPT_I2O=m
Modified: dists/trunk/linux-2.6/debian/config/amd64/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/amd64/defines (original)
+++ dists/trunk/linux-2.6/debian/config/amd64/defines Sun Dec 28 16:55:48 2008
@@ -7,9 +7,6 @@
amd64
kernel-arch: x86
-[xen_base]
-enabled: false
-
[image]
suggests: grub | lilo
Modified: dists/trunk/linux-2.6/debian/config/amd64/vserver/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/amd64/vserver/defines (original)
+++ dists/trunk/linux-2.6/debian/config/amd64/vserver/defines Sun Dec 28 16:55:48 2008
@@ -2,6 +2,3 @@
flavours:
amd64
-[image]
-recommends: util-vserver
-
Modified: dists/trunk/linux-2.6/debian/config/amd64/xen/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/amd64/xen/config (original)
+++ dists/trunk/linux-2.6/debian/config/amd64/xen/config Sun Dec 28 16:55:48 2008
@@ -1,15 +1 @@
CONFIG_X86_64_XEN=y
-
-## options needed for initial pvops fedora domU support
-# CONFIG_SMP is not set
-# CONFIG_NUMA is not set
-# CONFIG_NEED_MULTIPLE_NODES is not set
-
-## x86_64 breaks with a different CONFIG_PHYSICAL_START, currently
-CONFIG_PHYSICAL_START=0x200000
-
-## x86_64 breaks with CONFIG_SPARSEMEM_VMEMMAP, currently
-# CONFIG_SPARSEMEM_VMEMMAP is not set
-
-## 32-bit emulation isn't ready yet
-# CONFIG_IA32_EMULATION is not set
Modified: dists/trunk/linux-2.6/debian/config/amd64/xen/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/amd64/xen/defines (original)
+++ dists/trunk/linux-2.6/debian/config/amd64/xen/defines Sun Dec 28 16:55:48 2008
@@ -8,5 +8,5 @@
type: plain-xen
[xen]
-flavour: amd64
+flavours: amd64
Modified: dists/trunk/linux-2.6/debian/config/arm/config.iop32x
==============================================================================
--- dists/trunk/linux-2.6/debian/config/arm/config.iop32x (original)
+++ dists/trunk/linux-2.6/debian/config/arm/config.iop32x Sun Dec 28 16:55:48 2008
@@ -646,6 +646,12 @@
CONFIG_USB=m
##
+## file: drivers/usb/atm/Kconfig
+##
+CONFIG_USB_ATM=m
+CONFIG_USB_SPEEDTOUCH=m
+
+##
## file: drivers/usb/class/Kconfig
##
CONFIG_USB_ACM=m
@@ -717,6 +723,31 @@
CONFIG_BINFMT_MISC=m
##
+## file: fs/partitions/Kconfig
+##
+CONFIG_ACORN_PARTITION=y
+# CONFIG_ACORN_PARTITION_CUMANA is not set
+# CONFIG_ACORN_PARTITION_EESOX is not set
+CONFIG_ACORN_PARTITION_ICS=y
+# CONFIG_ACORN_PARTITION_ADFS is not set
+# CONFIG_ACORN_PARTITION_POWERTEC is not set
+CONFIG_ACORN_PARTITION_RISCIX=y
+CONFIG_OSF_PARTITION=y
+CONFIG_AMIGA_PARTITION=y
+CONFIG_ATARI_PARTITION=y
+CONFIG_MAC_PARTITION=y
+CONFIG_BSD_DISKLABEL=y
+CONFIG_MINIX_SUBPARTITION=y
+CONFIG_SOLARIS_X86_PARTITION=y
+CONFIG_UNIXWARE_DISKLABEL=y
+CONFIG_LDM_PARTITION=y
+# CONFIG_LDM_DEBUG is not set
+CONFIG_SGI_PARTITION=y
+CONFIG_ULTRIX_PARTITION=y
+CONFIG_SUN_PARTITION=y
+CONFIG_EFI_PARTITION=y
+
+##
## file: init/Kconfig
##
CONFIG_LOG_BUF_SHIFT=14
Modified: dists/trunk/linux-2.6/debian/config/arm/config.ixp4xx
==============================================================================
--- dists/trunk/linux-2.6/debian/config/arm/config.ixp4xx (original)
+++ dists/trunk/linux-2.6/debian/config/arm/config.ixp4xx Sun Dec 28 16:55:48 2008
@@ -704,8 +704,15 @@
CONFIG_USB=m
##
+## file: drivers/usb/atm/Kconfig
+##
+CONFIG_USB_ATM=m
+CONFIG_USB_SPEEDTOUCH=m
+
+##
## file: drivers/usb/class/Kconfig
##
+CONFIG_USB_ACM=m
CONFIG_USB_PRINTER=m
##
@@ -775,6 +782,31 @@
CONFIG_BINFMT_MISC=m
##
+## file: fs/partitions/Kconfig
+##
+CONFIG_ACORN_PARTITION=y
+# CONFIG_ACORN_PARTITION_CUMANA is not set
+# CONFIG_ACORN_PARTITION_EESOX is not set
+CONFIG_ACORN_PARTITION_ICS=y
+# CONFIG_ACORN_PARTITION_ADFS is not set
+# CONFIG_ACORN_PARTITION_POWERTEC is not set
+CONFIG_ACORN_PARTITION_RISCIX=y
+CONFIG_OSF_PARTITION=y
+CONFIG_AMIGA_PARTITION=y
+CONFIG_ATARI_PARTITION=y
+CONFIG_MAC_PARTITION=y
+CONFIG_BSD_DISKLABEL=y
+CONFIG_MINIX_SUBPARTITION=y
+CONFIG_SOLARIS_X86_PARTITION=y
+CONFIG_UNIXWARE_DISKLABEL=y
+CONFIG_LDM_PARTITION=y
+# CONFIG_LDM_DEBUG is not set
+CONFIG_SGI_PARTITION=y
+CONFIG_ULTRIX_PARTITION=y
+CONFIG_SUN_PARTITION=y
+CONFIG_EFI_PARTITION=y
+
+##
## file: init/Kconfig
##
CONFIG_LOG_BUF_SHIFT=14
Modified: dists/trunk/linux-2.6/debian/config/arm/config.orion5x
==============================================================================
--- dists/trunk/linux-2.6/debian/config/arm/config.orion5x (original)
+++ dists/trunk/linux-2.6/debian/config/arm/config.orion5x Sun Dec 28 16:55:48 2008
@@ -141,11 +141,30 @@
# CONFIG_PATA_SCH is not set
##
+## file: drivers/bluetooth/Kconfig
+##
+CONFIG_BT_HCIUSB=m
+CONFIG_BT_HCIUSB_SCO=y
+# CONFIG_BT_HCIUART is not set
+CONFIG_BT_HCIBCM203X=m
+CONFIG_BT_HCIBPA10X=m
+CONFIG_BT_HCIBFUSB=m
+CONFIG_BT_HCIVHCI=m
+
+##
## file: drivers/char/drm/Kconfig
##
# CONFIG_DRM is not set
##
+## file: drivers/dma/Kconfig
+##
+CONFIG_DMADEVICES=y
+CONFIG_MV_XOR=y
+CONFIG_DMA_ENGINE=y
+# CONFIG_NET_DMA is not set
+
+##
## file: drivers/i2c/Kconfig
##
CONFIG_I2C=y
@@ -305,11 +324,25 @@
CONFIG_NETCONSOLE=m
##
+## file: drivers/net/appletalk/Kconfig
+##
+CONFIG_ATALK=m
+CONFIG_IPDDP=m
+CONFIG_IPDDP_ENCAP=y
+CONFIG_IPDDP_DECAP=y
+
+##
## file: drivers/net/arcnet/Kconfig
##
# CONFIG_ARCNET is not set
##
+## file: drivers/net/wireless/Kconfig
+##
+CONFIG_WLAN_PRE80211=y
+CONFIG_WLAN_80211=y
+
+##
## file: drivers/pci/Kconfig
##
CONFIG_PCI_LEGACY=y
@@ -412,6 +445,12 @@
CONFIG_USB=m
##
+## file: drivers/usb/atm/Kconfig
+##
+CONFIG_USB_ATM=m
+CONFIG_USB_SPEEDTOUCH=m
+
+##
## file: drivers/usb/class/Kconfig
##
CONFIG_USB_ACM=m
@@ -475,6 +514,44 @@
# CONFIG_DISPLAY_SUPPORT is not set
##
+## file: drivers/watchdog/Kconfig
+##
+CONFIG_WATCHDOG=y
+CONFIG_ORION5X_WATCHDOG=m
+
+##
+## file: fs/Kconfig.binfmt
+##
+CONFIG_BINFMT_ELF=y
+CONFIG_BINFMT_AOUT=m
+CONFIG_BINFMT_MISC=m
+
+##
+## file: fs/partitions/Kconfig
+##
+CONFIG_ACORN_PARTITION=y
+# CONFIG_ACORN_PARTITION_CUMANA is not set
+# CONFIG_ACORN_PARTITION_EESOX is not set
+CONFIG_ACORN_PARTITION_ICS=y
+# CONFIG_ACORN_PARTITION_ADFS is not set
+# CONFIG_ACORN_PARTITION_POWERTEC is not set
+CONFIG_ACORN_PARTITION_RISCIX=y
+CONFIG_OSF_PARTITION=y
+CONFIG_AMIGA_PARTITION=y
+CONFIG_ATARI_PARTITION=y
+CONFIG_MAC_PARTITION=y
+CONFIG_BSD_DISKLABEL=y
+CONFIG_MINIX_SUBPARTITION=y
+CONFIG_SOLARIS_X86_PARTITION=y
+CONFIG_UNIXWARE_DISKLABEL=y
+CONFIG_LDM_PARTITION=y
+# CONFIG_LDM_DEBUG is not set
+CONFIG_SGI_PARTITION=y
+CONFIG_ULTRIX_PARTITION=y
+CONFIG_SUN_PARTITION=y
+CONFIG_EFI_PARTITION=y
+
+##
## file: kernel/power/Kconfig
##
# CONFIG_PM is not set
@@ -495,6 +572,11 @@
# CONFIG_RESOURCES_64BIT is not set
##
+## file: net/atm/Kconfig
+##
+CONFIG_ATM=m
+
+##
## file: net/ax25/Kconfig
##
CONFIG_HAMRADIO=y
@@ -503,50 +585,13 @@
CONFIG_NETROM=m
CONFIG_ROSE=m
-CONFIG_DMADEVICES=y
-CONFIG_MV_XOR=y
-CONFIG_DMA_ENGINE=y
-# CONFIG_NET_DMA is not set
-
-##
-## file: fs/Kconfig.binfmt
-##
-CONFIG_BINFMT_ELF=y
-CONFIG_BINFMT_AOUT=m
-CONFIG_BINFMT_MISC=m
-
-##
-## file: net/x25/Kconfig
-##
-CONFIG_X25=m
-
-##
-## file: drivers/net/wireless/Kconfig
-##
-CONFIG_WLAN_PRE80211=y
-CONFIG_WLAN_80211=y
-
-##
-## file: net/atm/Kconfig
-##
-CONFIG_ATM=m
-
##
## file: net/wanrouter/Kconfig
##
CONFIG_WAN_ROUTER=m
##
-## file: drivers/bluetooth/Kconfig
+## file: net/x25/Kconfig
##
-CONFIG_BT_HCIUSB=m
-CONFIG_BT_HCIUSB_SCO=y
-# CONFIG_BT_HCIUART is not set
-CONFIG_BT_HCIBCM203X=m
-CONFIG_BT_HCIBPA10X=m
-CONFIG_BT_HCIBFUSB=m
-CONFIG_BT_HCIVHCI=m
-
-CONFIG_WATCHDOG=y
-CONFIG_ORION5X_WATCHDOG=m
+CONFIG_X25=m
Modified: dists/trunk/linux-2.6/debian/config/arm/config.versatile
==============================================================================
--- dists/trunk/linux-2.6/debian/config/arm/config.versatile (original)
+++ dists/trunk/linux-2.6/debian/config/arm/config.versatile Sun Dec 28 16:55:48 2008
@@ -342,6 +342,11 @@
# CONFIG_PCCARD is not set
##
+## file: drivers/rtc/Kconfig
+##
+CONFIG_RTC_DRV_PL031=y
+
+##
## file: drivers/scsi/Kconfig
##
CONFIG_SCSI_SYM53C8XX_2=m
@@ -470,6 +475,13 @@
# CONFIG_PM is not set
##
+## file: kernel/time/Kconfig
+##
+# CONFIG_TICK_ONESHOT is not set
+# CONFIG_NO_HZ is not set
+# CONFIG_HIGH_RES_TIMERS is not set
+
+##
## file: lib/Kconfig.debug
##
CONFIG_ENABLE_WARN_DEPRECATED=y
Modified: dists/trunk/linux-2.6/debian/config/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/config (original)
+++ dists/trunk/linux-2.6/debian/config/config Sun Dec 28 16:55:48 2008
@@ -98,7 +98,7 @@
CONFIG_SATA_PROMISE=m
CONFIG_SATA_SX4=m
CONFIG_SATA_SIL=m
-# CONFIG_SATA_SIS is not set
+CONFIG_SATA_SIS=m
CONFIG_SATA_ULI=m
CONFIG_SATA_VIA=m
CONFIG_SATA_VITESSE=m
@@ -783,7 +783,6 @@
CONFIG_VIDEO_HEXIUM_GEMINI=m
CONFIG_VIDEO_CAFE_CCIC=m
CONFIG_V4L_USB_DRIVERS=y
-CONFIG_USB_VIDEO_CLASS=m
CONFIG_USB_VIDEO_CLASS_INPUT_EVDEV=y
CONFIG_USB_GSPCA=m
CONFIG_VIDEO_OVCAMCHIP=m
@@ -885,6 +884,11 @@
CONFIG_VIDEO_USBVISION=m
##
+## file: drivers/media/video/uvc/Kconfig
+##
+CONFIG_USB_VIDEO_CLASS=m
+
+##
## file: drivers/media/video/zc0301/Kconfig
##
CONFIG_USB_ZC0301=m
@@ -1379,7 +1383,9 @@
##
## file: drivers/rtc/Kconfig
##
-CONFIG_RTC_CLASS=m
+CONFIG_RTC_CLASS=y
+CONFIG_RTC_HCTOSYS=y
+CONFIG_RTC_HCTOSYS_DEVICE="rtc0"
CONFIG_RTC_INTF_SYSFS=y
CONFIG_RTC_INTF_PROC=y
CONFIG_RTC_INTF_DEV=y
@@ -1593,7 +1599,7 @@
CONFIG_USB_C67X00_HCD=m
CONFIG_USB_EHCI_TT_NEWSCHED=y
CONFIG_USB_ISP116X_HCD=m
-CONFIG_USB_ISP1760_HCD=m
+# CONFIG_USB_ISP1760_HCD is not set
# CONFIG_USB_ISP1760_PCI is not set
CONFIG_USB_U132_HCD=m
CONFIG_USB_R8A66597_HCD=m
@@ -2044,7 +2050,7 @@
CONFIG_PROFILING=y
# CONFIG_MARKERS is not set
CONFIG_MODULES=y
-# CONFIG_MODULE_FORCE_LOAD is not set
+CONFIG_MODULE_FORCE_LOAD=y
CONFIG_MODULE_UNLOAD=y
CONFIG_MODULE_FORCE_UNLOAD=y
CONFIG_MODVERSIONS=y
@@ -2657,7 +2663,7 @@
CONFIG_SND_PCI=y
CONFIG_SND_AD1889=m
CONFIG_SND_ALS300=m
-CONFIG_SND_AW2=m
+# CONFIG_SND_AW2 is not set
CONFIG_SND_OXYGEN=m
CONFIG_SND_CS5530=m
CONFIG_SND_DARLA20=m
Modified: dists/trunk/linux-2.6/debian/config/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/defines (original)
+++ dists/trunk/linux-2.6/debian/config/defines Sun Dec 28 16:55:48 2008
@@ -29,14 +29,18 @@
[featureset-vserver_base]
enabled: false
+[featureset-xen_base]
+enabled: false
+
[image]
+desc-long-part-xen: This kernel also runs on a Xen hypervisor.
+ It supports only unpriviledged (domU) operation.
initramfs-generators: initramfs-tools yaird initramfs-fallback
type: plain
[xen]
-dom0-support: false
versions:
- 3.1-1
+ 3.2-1
[commands-image-initramfs-generators]
initramfs-tools: mkinitramfs-kpkg
Modified: dists/trunk/linux-2.6/debian/config/featureset-openvz/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/featureset-openvz/config (original)
+++ dists/trunk/linux-2.6/debian/config/featureset-openvz/config Sun Dec 28 16:55:48 2008
@@ -6,7 +6,7 @@
CONFIG_VZ_DEV=m
CONFIG_VE_IPTABLES=y
CONFIG_VZ_WDOG=m
-# CONFIG_VZ_CHECKPOINT is not set
+CONFIG_VZ_CHECKPOINT=m
CONFIG_SIM_FS=m
CONFIG_VZ_QUOTA=m
# CONFIG_VZ_QUOTA_UNLOAD is not set
@@ -19,3 +19,6 @@
CONFIG_BC_SWAP_ACCOUNTING=y
CONFIG_BC_PROC=y
# CONFIG_BC_DEBUG is not set
+
+# buggy
+# CONFIG_NF_CONNTRACK_IPV6 is not set
Modified: dists/trunk/linux-2.6/debian/config/featureset-vserver/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/featureset-vserver/config (original)
+++ dists/trunk/linux-2.6/debian/config/featureset-vserver/config Sun Dec 28 16:55:48 2008
@@ -1,8 +1,19 @@
+##
+## file: drivers/block/Kconfig
+##
+CONFIG_BLK_DEV_VROOT=y
+
+##
+## file: kernel/vserver/Kconfig
+##
+CONFIG_VSERVER_AUTO_LBACK=y
+# CONFIG_VSERVER_AUTO_SINGLE is not set
CONFIG_VSERVER_COWBL=y
# CONFIG_VSERVER_VTIME is not set
+# CONFIG_VSERVER_DEVICE is not set
CONFIG_VSERVER_PROC_SECURE=y
CONFIG_VSERVER_HARDCPU=y
-# CONFIG_VSERVER_IDLETIME is not set
+CONFIG_VSERVER_IDLETIME=y
CONFIG_VSERVER_IDLELIMIT=y
# CONFIG_TAGGING_NONE is not set
# CONFIG_TAGGING_UID16 is not set
@@ -11,5 +22,11 @@
# CONFIG_TAGGING_INTERN is not set
# CONFIG_TAG_NFSD is not set
CONFIG_VSERVER_PRIVACY=y
+CONFIG_VSERVER_CONTEXTS=512
+CONFIG_VSERVER_WARN=y
# CONFIG_VSERVER_DEBUG is not set
-CONFIG_BLK_DEV_VROOT=y
+
+##
+## file: net/ipv6/Kconfig
+##
+CONFIG_IPV6=y
Added: dists/trunk/linux-2.6/debian/config/featureset-vserver/defines
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/config/featureset-vserver/defines Sun Dec 28 16:55:48 2008
@@ -0,0 +1,5 @@
+[image]
+desc-long-part-vserver: This kernel includes support for Linux-VServer virtualization.
+desc-parts: vserver
+desc-short-part-vserver: Linux-VServer support
+recommends: util-vserver
Modified: dists/trunk/linux-2.6/debian/config/featureset-xen/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/featureset-xen/config (original)
+++ dists/trunk/linux-2.6/debian/config/featureset-xen/config Sun Dec 28 16:55:48 2008
@@ -1,47 +1,20 @@
CONFIG_XEN=y
-CONFIG_XENCTRL=y
+CONFIG_XEN_PRIVILEGED_GUEST=y
+CONFIG_XEN_BACKEND=y
+CONFIG_XEN_BLKDEV_BACKEND=y
+CONFIG_XEN_BLKDEV_TAP=m
CONFIG_XEN_BLKDEV_FRONTEND=y
CONFIG_HVC_XEN=y
+CONFIG_XEN_NETDEV_BACKEND=y
CONFIG_XEN_NETDEV_FRONTEND=y
-
-# CONFIG_VMI is not set
-CONFIG_PM=y
-# CONFIG_PM_DEBUG is not set
-CONFIG_PM_SLEEP_SMP=y
-CONFIG_PM_SLEEP=y
-CONFIG_SUSPEND=y
-CONFIG_SUSPEND_FREEZER=y
-# CONFIG_HIBERNATION is not set
-# CONFIG_APM is not set
-# CONFIG_CPU_FREQ is not set
-# CONFIG_PCI is not set
-# CONFIG_ISA is not set
-# CONFIG_EISA is not set
-# CONFIG_SCx200 is not set
-# CONFIG_PCCARD is not set
-# CONFIG_PARPORT is not set
-# CONFIG_BLK_DEV_FD is not set
-# CONFIG_PARIDE is not set
-# CONFIG_CDROM_PKTCDVD is not set
-# CONFIG_IDE is not set
-# CONFIG_SOUND is not set
-# CONFIG_NL80211 is not set
-# CONFIG_MAC80211 is not set
-# CONFIG_RFKILL is not set
-# CONFIG_ISDN is not set
-# CONFIG_INPUT is not set
-# CONFIG_SERIO is not set
-# CONFIG_I2C is not set
-# CONFIG_W1 is not set
-# CONFIG_HWMON is not set
-# CONFIG_VIDEO_DEV is not set
-# CONFIG_DVB_CORE is not set
-# CONFIG_DAB is not set
-# CONFIG_MTD is not set
-# CONFIG_SPI is not set
-# CONFIG_USB_SUPPORT is not set
-# CONFIG_MMC is not set
-# CONFIG_WLAN_PRE80211 is not set
-# CONFIG_WLAN_80211 is not set
-# CONFIG_ATA is not set
-# CONFIG_WAN is not set
+CONFIG_XEN_NETDEV_LOOPBACK=m
+CONFIG_XEN_PCIDEV_BACKEND=y
+CONFIG_XEN_PCIDEV_BACKEND_VPCI=y
+CONFIG_XEN_PCIDEV_FRONTEND=y
+# CONFIG_XEN_DISABLE_SERIAL is not set
+CONFIG_XEN_SCSI_BACKEND=m
+CONFIG_XEN_SCSI_FRONTEND=m
+# CONFIG_XEN_COMPAT_030002_AND_LATER is not set
+# CONCIF_XEN_COMPAT_030004_AND_LATER is not set
+CONFIG_XEN_COMPAT_030100_AND_LATER=y
+# CONFIG_XEN_COMPAT_LATEST_ONLY is not set
Added: dists/trunk/linux-2.6/debian/config/featureset-xen/defines
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/config/featureset-xen/defines Sun Dec 28 16:55:48 2008
@@ -0,0 +1,5 @@
+[image]
+desc-long-part-xenold: This kernel only runs on a Xen hypervisor.
+ It supports both priviledged (dom0) and unpriviledged (domU) operation.
+desc-parts: xenold
+desc-short-part-xenold: oldstyle Xen support
Modified: dists/trunk/linux-2.6/debian/config/hppa/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/hppa/config (original)
+++ dists/trunk/linux-2.6/debian/config/hppa/config Sun Dec 28 16:55:48 2008
@@ -439,6 +439,11 @@
# CONFIG_PNP is not set
##
+## file: drivers/rtc/Kconfig
+##
+# CONFIG_RTC_CLASS is not set
+
+##
## file: drivers/scsi/Kconfig
##
# CONFIG_SCSI_IN2000 is not set
Modified: dists/trunk/linux-2.6/debian/config/i386/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/config (original)
+++ dists/trunk/linux-2.6/debian/config/i386/config Sun Dec 28 16:55:48 2008
@@ -176,7 +176,7 @@
##
CONFIG_ACPI=y
CONFIG_ACPI_PROCFS=y
-# CONFIG_ACPI_PROCFS_POWER is not set
+CONFIG_ACPI_PROCFS_POWER=y
CONFIG_ACPI_SYSFS_POWER=y
CONFIG_ACPI_PROC_EVENT=y
CONFIG_ACPI_AC=m
@@ -203,6 +203,7 @@
##
# CONFIG_PATA_CS5535 is not set
# CONFIG_PATA_CS5536 is not set
+# CONFIG_PATA_SIS is not set
##
## file: drivers/atm/Kconfig
@@ -330,9 +331,6 @@
CONFIG_PPDEV=m
CONFIG_HVC_XEN=y
CONFIG_NVRAM=m
-CONFIG_RTC=m
-CONFIG_GEN_RTC=m
-CONFIG_GEN_RTC_X=y
CONFIG_DTLK=m
CONFIG_R3964=m
CONFIG_APPLICOM=m
@@ -343,7 +341,6 @@
CONFIG_RAW_DRIVER=m
CONFIG_MAX_RAW_DEVS=256
CONFIG_HPET=y
-CONFIG_HPET_RTC_IRQ=y
CONFIG_HPET_MMAP=y
CONFIG_HANGCHECK_TIMER=m
@@ -565,7 +562,7 @@
CONFIG_BLK_DEV_IDECD=m
CONFIG_BLK_DEV_IDETAPE=m
CONFIG_BLK_DEV_IDEFLOPPY=m
-# CONFIG_BLK_DEV_IDEACPI is not set
+CONFIG_BLK_DEV_IDEACPI=y
CONFIG_IDE_GENERIC=m
CONFIG_BLK_DEV_CMD640=y
# CONFIG_BLK_DEV_CMD640_ENHANCED is not set
@@ -1355,6 +1352,11 @@
CONFIG_BATTERY_DS2760=m
##
+## file: drivers/rtc/Kconfig
+##
+CONFIG_RTC_DRV_CMOS=y
+
+##
## file: drivers/scsi/Kconfig
##
CONFIG_SCSI_DPT_I2O=m
Modified: dists/trunk/linux-2.6/debian/config/i386/config.686-bigmem
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/config.686-bigmem (original)
+++ dists/trunk/linux-2.6/debian/config/i386/config.686-bigmem Sun Dec 28 16:55:48 2008
@@ -2,6 +2,19 @@
## file: arch/x86/Kconfig
##
CONFIG_SMP=y
+## choice: Subarchitecture Type
+# CONFIG_X86_PC is not set
+# CONFIG_X86_ELAN is not set
+# CONFIG_X86_VOYAGER is not set
+# CONFIG_X86_NUMAQ is not set
+# CONFIG_X86_SUMMIT is not set
+# CONFIG_X86_BIGSMP is not set
+# CONFIG_X86_VISWS is not set
+CONFIG_X86_GENERICARCH=y
+# CONFIG_X86_ES7000 is not set
+# CONFIG_X86_RDC321X is not set
+# CONFIG_X86_VSMP is not set
+## end choice
CONFIG_NR_CPUS=32
CONFIG_X86_MCE=y
CONFIG_X86_MCE_NONFATAL=m
Modified: dists/trunk/linux-2.6/debian/config/i386/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/defines (original)
+++ dists/trunk/linux-2.6/debian/config/i386/defines Sun Dec 28 16:55:48 2008
@@ -29,6 +29,7 @@
longclass: Pentium Pro/Celeron/Pentium II/Pentium III/Pentium 4 with 4-64G RAM
[686-bigmem_image]
+desc-parts: xen
recommends: libc6-i686
[amd64_base]
Modified: dists/trunk/linux-2.6/debian/config/i386/openvz/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/openvz/defines (original)
+++ dists/trunk/linux-2.6/debian/config/i386/openvz/defines Sun Dec 28 16:55:48 2008
@@ -5,4 +5,5 @@
[686_image]
configs:
i386/config.686-bigmem
+desc-parts: xen
Modified: dists/trunk/linux-2.6/debian/config/i386/vserver/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/vserver/defines (original)
+++ dists/trunk/linux-2.6/debian/config/i386/vserver/defines Sun Dec 28 16:55:48 2008
@@ -1,6 +1,7 @@
[base]
flavours:
686
+ 686-bigmem
[image]
recommends: util-vserver, libc6-i686
Added: dists/trunk/linux-2.6/debian/config/i386/xen/config
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/config/i386/xen/config Sun Dec 28 16:55:48 2008
@@ -0,0 +1,2 @@
+# CONFIG_X86_GENERICARCH is not set
+CONFIG_X86_XEN=y
Modified: dists/trunk/linux-2.6/debian/config/i386/xen/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/i386/xen/defines (original)
+++ dists/trunk/linux-2.6/debian/config/i386/xen/defines Sun Dec 28 16:55:48 2008
@@ -8,7 +8,9 @@
type: plain-xen
[xen]
-flavour: i386-pae
+flavours:
+ i386
+ amd64
[686_base]
class: i686
@@ -16,6 +18,7 @@
[686_image]
configs:
- i386/config.686-bigmem
+ i386/config.686-bigmem
+ i386/xen/config
recommends: libc6-xen
Modified: dists/trunk/linux-2.6/debian/config/ia64/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/ia64/defines (original)
+++ dists/trunk/linux-2.6/debian/config/ia64/defines Sun Dec 28 16:55:48 2008
@@ -1,6 +1,6 @@
[base]
featuresets:
-# vserver
+ vserver
flavours:
itanium
mckinley
Modified: dists/trunk/linux-2.6/debian/config/ia64/vserver/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/ia64/vserver/defines (original)
+++ dists/trunk/linux-2.6/debian/config/ia64/vserver/defines Sun Dec 28 16:55:48 2008
@@ -3,12 +3,3 @@
itanium
mckinley
-[image]
-recommends: util-vserver
-
-[itanium_image]
-configs: ia64/config.itanium
-
-[mckinley_image]
-configs: ia64/config.mckinley
-
Modified: dists/trunk/linux-2.6/debian/config/powerpc/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/powerpc/config (original)
+++ dists/trunk/linux-2.6/debian/config/powerpc/config Sun Dec 28 16:55:48 2008
@@ -130,8 +130,6 @@
CONFIG_PRINTER=m
CONFIG_HVC_RTAS=y
CONFIG_NVRAM=y
-CONFIG_GEN_RTC=y
-CONFIG_GEN_RTC_X=y
CONFIG_DTLK=m
CONFIG_APPLICOM=m
@@ -752,6 +750,11 @@
CONFIG_TCIC=m
##
+## file: drivers/rtc/Kconfig
+##
+CONFIG_RTC_DRV_PPC=y
+
+##
## file: drivers/scsi/Kconfig
##
CONFIG_SCSI_DPT_I2O=m
Modified: dists/trunk/linux-2.6/debian/config/powerpc/config.powerpc64
==============================================================================
--- dists/trunk/linux-2.6/debian/config/powerpc/config.powerpc64 (original)
+++ dists/trunk/linux-2.6/debian/config/powerpc/config.powerpc64 Sun Dec 28 16:55:48 2008
@@ -101,8 +101,6 @@
##
CONFIG_HVC_CONSOLE=y
CONFIG_HVCS=m
-# CONFIG_RTC is not set
-# CONFIG_GEN_RTC is not set
CONFIG_HANGCHECK_TIMER=m
##
@@ -157,11 +155,6 @@
# CONFIG_HOTPLUG_PCI is not set
##
-## file: drivers/rtc/Kconfig
-##
-CONFIG_RTC_DRV_PPC=m
-
-##
## file: drivers/scsi/Kconfig
##
CONFIG_SCSI_IBMVSCSI=m
Modified: dists/trunk/linux-2.6/debian/config/powerpc/vserver/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/powerpc/vserver/defines (original)
+++ dists/trunk/linux-2.6/debian/config/powerpc/vserver/defines Sun Dec 28 16:55:48 2008
@@ -3,9 +3,6 @@
powerpc
powerpc64
-[image]
-recommends: util-vserver
-
[powerpc_image]
configs: powerpc/config.powerpc-smp
Modified: dists/trunk/linux-2.6/debian/config/s390/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/s390/defines (original)
+++ dists/trunk/linux-2.6/debian/config/s390/defines Sun Dec 28 16:55:48 2008
@@ -1,6 +1,6 @@
[base]
featuresets:
-# vserver
+ vserver
flavours:
s390
s390-tape
@@ -10,13 +10,15 @@
[image]
desc-long-part-reader: This kernel has support to IPL (boot) from a VM reader or DASD device.
desc-long-part-tape: This kernel has support to IPL (boot) from a tape.
-desc-parts: reader
desc-short-part-tape: IPL from tape
suggests: s390-tools
[s390_base]
class: IBM S/390
+[s390_image]
+desc-parts: reader
+
[s390-tape_base]
class: IBM S/390
modules: false
@@ -30,3 +32,6 @@
[s390x_base]
class: IBM zSeries
+[s390x_image]
+desc-parts: reader
+
Modified: dists/trunk/linux-2.6/debian/config/sparc/config
==============================================================================
--- dists/trunk/linux-2.6/debian/config/sparc/config (original)
+++ dists/trunk/linux-2.6/debian/config/sparc/config Sun Dec 28 16:55:48 2008
@@ -2,6 +2,7 @@
## file: arch/sparc64/Kconfig
##
CONFIG_SECCOMP=y
+CONFIG_SUN_LDOMS=y
CONFIG_PCI=y
CONFIG_SUN_OPENPROMFS=m
@@ -13,6 +14,7 @@
# CONFIG_BLK_CPQ_CISS_DA is not set
# CONFIG_BLK_DEV_DAC960 is not set
# CONFIG_BLK_DEV_UMEM is not set
+CONFIG_SUNVDC=m
##
## file: drivers/char/Kconfig
@@ -125,6 +127,7 @@
CONFIG_HAPPYMEAL=m
CONFIG_SUNBMAC=m
CONFIG_SUNQE=m
+CONFIG_SUNVNET=m
# CONFIG_HP100 is not set
# CONFIG_HAMACHI is not set
# CONFIG_R8169 is not set
Modified: dists/trunk/linux-2.6/debian/config/sparc/defines
==============================================================================
--- dists/trunk/linux-2.6/debian/config/sparc/defines (original)
+++ dists/trunk/linux-2.6/debian/config/sparc/defines Sun Dec 28 16:55:48 2008
@@ -1,10 +1,11 @@
[base]
featuresets:
-# vserver
+ vserver
flavours:
sparc64
sparc64-smp
kernel-arch: sparc64
+kernel-header-dirs: sparc sparc64
[image]
suggests: silo, fdutils
Modified: dists/trunk/linux-2.6/debian/lib/python/debian_linux/config.py
==============================================================================
--- dists/trunk/linux-2.6/debian/lib/python/debian_linux/config.py (original)
+++ dists/trunk/linux-2.6/debian/lib/python/debian_linux/config.py Sun Dec 28 16:55:48 2008
@@ -26,6 +26,35 @@
return [j.strip() for j in re.split(self.type, i)]
class ConfigCore(dict):
+ def get_merge(self, section, arch, featureset, flavour, key, default=None):
+ temp = []
+
+ if arch and featureset and flavour:
+ temp.append(self.get((section, arch, featureset, flavour), {}).get(key))
+ temp.append(self.get((section, arch, None, flavour), {}).get(key))
+ if arch and featureset:
+ temp.append(self.get((section, arch, featureset), {}).get(key))
+ if arch:
+ temp.append(self.get((section, arch), {}).get(key))
+ if featureset:
+ temp.append(self.get((section, None, featureset), {}).get(key))
+ temp.append(self.get((section,), {}).get(key))
+
+ ret = []
+
+ for i in temp:
+ if i is None:
+ continue
+ elif isinstance(i, (list, tuple)):
+ ret.extend(i)
+ elif ret:
+ # TODO
+ return ret
+ else:
+ return i
+
+ return ret or default
+
def merge(self, section, arch = None, featureset = None, flavour = None):
ret = {}
ret.update(self.get((section,), {}))
@@ -73,6 +102,9 @@
config_name = "defines"
schemas = {
+ 'abi': {
+ 'ignore-changes': SchemaItemList(),
+ },
'base': {
'arches': SchemaItemList(),
'enabled': SchemaItemBoolean(),
@@ -90,6 +122,7 @@
},
'xen': {
'dom0-support': SchemaItemBoolean(),
+ 'flavours': SchemaItemList(),
'versions': SchemaItemList(),
}
}
Modified: dists/trunk/linux-2.6/debian/lib/python/debian_linux/patches.py
==============================================================================
--- dists/trunk/linux-2.6/debian/lib/python/debian_linux/patches.py (original)
+++ dists/trunk/linux-2.6/debian/lib/python/debian_linux/patches.py Sun Dec 28 16:55:48 2008
@@ -183,7 +183,7 @@
l = self
else:
l = self[::-1]
- for i in self:
+ for i in l:
if reverse:
print "--> Try to unapply %s." % i.name
else:
Added: dists/trunk/linux-2.6/debian/patches/bugfix/all/ext3-add-support-for-non-native-signed-unsigned-htr.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/all/ext3-add-support-for-non-native-signed-unsigned-htr.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,230 @@
+From: Theodore Ts'o <tytso at mit.edu>
+Date: Mon, 20 Oct 2008 23:16:21 -0400
+Subject: [PATCH] ext3: Add support for non-native signed/unsigned htree hash algorithms
+
+The original ext3 hash algorithms assumed that variables of type char
+were signed, as God and K&R intended. Unfortunately, this assumption
+is not true on some architectures. Userspace support for marking
+filesystems with non-native signed/unsigned chars was added two years
+ago, but the kernel-side support was never added (until now).
+
+Signed-off-by: "Theodore Ts'o" <tytso at mit.edu>
+
+diff -urN a/fs/ext3/hash.c b/fs/ext3/hash.c
+--- a/fs/ext3/hash.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext3/hash.c 2008-10-28 19:44:22.000000000 +0000
+@@ -35,11 +35,20 @@
+
+
+ /* The old legacy hash */
+-static __u32 dx_hack_hash (const char *name, int len)
++static __u32 dx_hack_hash (const char *name, int len, int unsigned_flag)
+ {
+- __u32 hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9;
++ __u32 hash, hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9;
++ const unsigned char *ucp = (const unsigned char *) name;
++ const signed char *scp = (const signed char *) name;
++ int c;
++
+ while (len--) {
+- __u32 hash = hash1 + (hash0 ^ (*name++ * 7152373));
++ if (unsigned_flag)
++ c = (int) *ucp++;
++ else
++ c = (int) *scp++;
++
++ hash = hash1 + (hash0 ^ (c * 7152373));
+
+ if (hash & 0x80000000) hash -= 0x7fffffff;
+ hash1 = hash0;
+@@ -48,10 +57,13 @@
+ return (hash0 << 1);
+ }
+
+-static void str2hashbuf(const char *msg, int len, __u32 *buf, int num)
++static void str2hashbuf(const char *msg, int len, __u32 *buf, int num,
++ int unsigned_flag)
+ {
+ __u32 pad, val;
+- int i;
++ int i, c;
++ const unsigned char *ucp = (const unsigned char *) msg;
++ const signed char *scp = (const signed char *) msg;
+
+ pad = (__u32)len | ((__u32)len << 8);
+ pad |= pad << 16;
+@@ -62,7 +74,12 @@
+ for (i=0; i < len; i++) {
+ if ((i % 4) == 0)
+ val = pad;
+- val = msg[i] + (val << 8);
++ if (unsigned_flag)
++ c = (int) ucp[i];
++ else
++ c = (int) scp[i];
++
++ val = c + (val << 8);
+ if ((i % 4) == 3) {
+ *buf++ = val;
+ val = pad;
+@@ -95,6 +112,7 @@
+ const char *p;
+ int i;
+ __u32 in[8], buf[4];
++ int unsigned_flag = 0;
+
+ /* Initialize the default seed for the hash checksum functions */
+ buf[0] = 0x67452301;
+@@ -113,13 +131,17 @@
+ }
+
+ switch (hinfo->hash_version) {
++ case DX_HASH_LEGACY_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_LEGACY:
+- hash = dx_hack_hash(name, len);
++ hash = dx_hack_hash(name, len, unsigned_flag);
+ break;
++ case DX_HASH_HALF_MD4_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_HALF_MD4:
+ p = name;
+ while (len > 0) {
+- str2hashbuf(p, len, in, 8);
++ str2hashbuf(p, len, in, 8, unsigned_flag);
+ half_md4_transform(buf, in);
+ len -= 32;
+ p += 32;
+@@ -127,10 +149,12 @@
+ minor_hash = buf[2];
+ hash = buf[1];
+ break;
++ case DX_HASH_TEA_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_TEA:
+ p = name;
+ while (len > 0) {
+- str2hashbuf(p, len, in, 4);
++ str2hashbuf(p, len, in, 4, unsigned_flag);
+ TEA_transform(buf, in);
+ len -= 16;
+ p += 16;
+diff -urN a/fs/ext3/namei.c b/fs/ext3/namei.c
+--- a/fs/ext3/namei.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext3/namei.c 2008-10-28 19:44:22.000000000 +0000
+@@ -369,6 +369,8 @@
+ goto fail;
+ }
+ hinfo->hash_version = root->info.hash_version;
++ if (hinfo->hash_version <= DX_HASH_TEA)
++ hinfo->hash_version += EXT3_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo->seed = EXT3_SB(dir->i_sb)->s_hash_seed;
+ if (dentry)
+ ext3fs_dirhash(dentry->d_name.name, dentry->d_name.len, hinfo);
+@@ -637,6 +639,9 @@
+ dir = dir_file->f_path.dentry->d_inode;
+ if (!(EXT3_I(dir)->i_flags & EXT3_INDEX_FL)) {
+ hinfo.hash_version = EXT3_SB(dir->i_sb)->s_def_hash_version;
++ if (hinfo.hash_version <= DX_HASH_TEA)
++ hinfo.hash_version +=
++ EXT3_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo.seed = EXT3_SB(dir->i_sb)->s_hash_seed;
+ count = htree_dirblock_to_tree(dir_file, dir, 0, &hinfo,
+ start_hash, start_minor_hash);
+@@ -1413,6 +1418,8 @@
+
+ /* Initialize as for dx_probe */
+ hinfo.hash_version = root->info.hash_version;
++ if (hinfo.hash_version <= DX_HASH_TEA)
++ hinfo.hash_version += EXT3_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo.seed = EXT3_SB(dir->i_sb)->s_hash_seed;
+ ext3fs_dirhash(name, namelen, &hinfo);
+ frame = frames;
+diff -urN a/fs/ext3/super.c b/fs/ext3/super.c
+--- a/fs/ext3/super.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext3/super.c 2008-10-28 19:44:22.000000000 +0000
+@@ -1712,6 +1712,21 @@
+ for (i=0; i < 4; i++)
+ sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]);
+ sbi->s_def_hash_version = es->s_def_hash_version;
++ i = le32_to_cpu(es->s_flags);
++ if (i & EXT2_FLAGS_UNSIGNED_HASH)
++ sbi->s_hash_unsigned = 3;
++ else if ((i & EXT2_FLAGS_SIGNED_HASH) == 0) {
++ char c;
++
++ c = (char) 255;
++ if (((int) c) == -1) {
++ es->s_flags |= cpu_to_le32(EXT2_FLAGS_SIGNED_HASH);
++ } else {
++ es->s_flags |= cpu_to_le32(EXT2_FLAGS_UNSIGNED_HASH);
++ sbi->s_hash_unsigned = 3;
++ }
++ sb->s_dirt = 1;
++ }
+
+ if (sbi->s_blocks_per_group > blocksize * 8) {
+ printk (KERN_ERR
+diff -urN a/include/linux/ext3_fs.h b/include/linux/ext3_fs.h
+--- a/include/linux/ext3_fs.h 2008-10-28 19:43:42.000000000 +0000
++++ b/include/linux/ext3_fs.h 2008-10-28 19:44:22.000000000 +0000
+@@ -354,6 +354,13 @@
+ #define EXT3_ORPHAN_FS 0x0004 /* Orphans being recovered */
+
+ /*
++ * Misc. filesystem flags
++ */
++#define EXT2_FLAGS_SIGNED_HASH 0x0001 /* Signed dirhash in use */
++#define EXT2_FLAGS_UNSIGNED_HASH 0x0002 /* Unsigned dirhash in use */
++#define EXT2_FLAGS_TEST_FILESYS 0x0004 /* to test development code */
++
++/*
+ * Mount flags
+ */
+ #define EXT3_MOUNT_CHECK 0x00001 /* Do mount-time checks */
+@@ -487,7 +494,23 @@
+ __u16 s_reserved_word_pad;
+ __le32 s_default_mount_opts;
+ __le32 s_first_meta_bg; /* First metablock block group */
+- __u32 s_reserved[190]; /* Padding to the end of the block */
++ __le32 s_mkfs_time; /* When the filesystem was created */
++ __le32 s_jnl_blocks[17]; /* Backup of the journal inode */
++ /* 64bit support valid if EXT4_FEATURE_COMPAT_64BIT */
++/*150*/ __le32 s_blocks_count_hi; /* Blocks count */
++ __le32 s_r_blocks_count_hi; /* Reserved blocks count */
++ __le32 s_free_blocks_count_hi; /* Free blocks count */
++ __le16 s_min_extra_isize; /* All inodes have at least # bytes */
++ __le16 s_want_extra_isize; /* New inodes should reserve # bytes */
++ __le32 s_flags; /* Miscellaneous flags */
++ __le16 s_raid_stride; /* RAID stride */
++ __le16 s_mmp_interval; /* # seconds to wait in MMP checking */
++ __le64 s_mmp_block; /* Block for multi-mount protection */
++ __le32 s_raid_stripe_width; /* blocks on all data disks (N*stride)*/
++ __u8 s_log_groups_per_flex; /* FLEX_BG group size */
++ __u8 s_reserved_char_pad2;
++ __le16 s_reserved_pad;
++ __u32 s_reserved[162]; /* Padding to the end of the block */
+ };
+
+ #ifdef __KERNEL__
+@@ -692,6 +715,9 @@
+ #define DX_HASH_LEGACY 0
+ #define DX_HASH_HALF_MD4 1
+ #define DX_HASH_TEA 2
++#define DX_HASH_LEGACY_UNSIGNED 3
++#define DX_HASH_HALF_MD4_UNSIGNED 4
++#define DX_HASH_TEA_UNSIGNED 5
+
+ #ifdef __KERNEL__
+
+diff -urN a/include/linux/ext3_fs_sb.h b/include/linux/ext3_fs_sb.h
+--- a/include/linux/ext3_fs_sb.h 2008-10-28 19:43:42.000000000 +0000
++++ b/include/linux/ext3_fs_sb.h 2008-10-28 19:44:22.000000000 +0000
+@@ -57,6 +57,7 @@
+ u32 s_next_generation;
+ u32 s_hash_seed[4];
+ int s_def_hash_version;
++ int s_hash_unsigned; /* 3 if hash should be signed, 0 if not */
+ struct percpu_counter s_freeblocks_counter;
+ struct percpu_counter s_freeinodes_counter;
+ struct percpu_counter s_dirs_counter;
Added: dists/trunk/linux-2.6/debian/patches/bugfix/all/ext4-add-support-for-non-native-signed-unsigned-htr.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/all/ext4-add-support-for-non-native-signed-unsigned-htr.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,191 @@
+From: Theodore Ts'o <tytso at mit.edu>
+Date: Mon, 20 Oct 2008 22:57:37 -0400
+Subject: [PATCH] ext4: Add support for non-native signed/unsigned htree hash algorithms
+
+The original ext3 hash algorithms assumed that variables of type char
+were signed, as God and K&R intended. Unfortunately, this assumption
+is not true on some architectures. Userspace support for marking
+filesystems with non-native signed/unsigned chars was added two years
+ago, but the kernel-side support was never added (until now).
+
+Signed-off-by: "Theodore Ts'o" <tytso at mit.edu>
+
+diff -urN a/fs/ext4/ext4.h b/fs/ext4/ext4.h
+--- a/fs/ext4/ext4.h 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext4/ext4.h 2008-10-28 19:48:22.000000000 +0000
+@@ -872,6 +872,9 @@
+ #define DX_HASH_LEGACY 0
+ #define DX_HASH_HALF_MD4 1
+ #define DX_HASH_TEA 2
++#define DX_HASH_LEGACY_UNSIGNED 3
++#define DX_HASH_HALF_MD4_UNSIGNED 4
++#define DX_HASH_TEA_UNSIGNED 5
+
+ #ifdef __KERNEL__
+
+diff -urN a/fs/ext4/ext4_sb.h b/fs/ext4/ext4_sb.h
+--- a/fs/ext4/ext4_sb.h 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext4/ext4_sb.h 2008-10-28 19:48:22.000000000 +0000
+@@ -56,6 +56,7 @@
+ u32 s_next_generation;
+ u32 s_hash_seed[4];
+ int s_def_hash_version;
++ int s_hash_unsigned; /* 3 if hash should be signed, 0 if not */
+ struct percpu_counter s_freeblocks_counter;
+ struct percpu_counter s_freeinodes_counter;
+ struct percpu_counter s_dirs_counter;
+diff -urN a/fs/ext4/hash.c b/fs/ext4/hash.c
+--- a/fs/ext4/hash.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext4/hash.c 2008-10-28 19:59:19.000000000 +0000
+@@ -35,11 +35,20 @@
+
+
+ /* The old legacy hash */
+-static __u32 dx_hack_hash (const char *name, int len)
++static __u32 dx_hack_hash(const char *name, int len, int unsigned_flag)
+ {
+- __u32 hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9;
++ __u32 hash, hash0 = 0x12a3fe2d, hash1 = 0x37abe8f9;
++ const unsigned char *ucp = (const unsigned char *) name;
++ const signed char *scp = (const signed char *) name;
++ int c;
++
+ while (len--) {
+- __u32 hash = hash1 + (hash0 ^ (*name++ * 7152373));
++ if (unsigned_flag)
++ c = (int) *ucp++;
++ else
++ c = (int) *scp++;
++
++ hash = hash1 + (hash0 ^ (c * 7152373));
+
+ if (hash & 0x80000000) hash -= 0x7fffffff;
+ hash1 = hash0;
+@@ -48,10 +57,13 @@
+ return (hash0 << 1);
+ }
+
+-static void str2hashbuf(const char *msg, int len, __u32 *buf, int num)
++static void str2hashbuf(const char *msg, int len, __u32 *buf, int num,
++ int unsigned_flag)
+ {
+ __u32 pad, val;
+- int i;
++ int i, c;
++ const unsigned char *ucp = (const unsigned char *) msg;
++ const signed char *scp = (const signed char *) msg;
+
+ pad = (__u32)len | ((__u32)len << 8);
+ pad |= pad << 16;
+@@ -62,7 +74,12 @@
+ for (i=0; i < len; i++) {
+ if ((i % 4) == 0)
+ val = pad;
+- val = msg[i] + (val << 8);
++ if (unsigned_flag)
++ c = (int) ucp[i];
++ else
++ c = (int) scp[i];
++
++ val = c + (val << 8);
+ if ((i % 4) == 3) {
+ *buf++ = val;
+ val = pad;
+@@ -95,6 +112,7 @@
+ const char *p;
+ int i;
+ __u32 in[8], buf[4];
++ int unsigned_flag = 0;
+
+ /* Initialize the default seed for the hash checksum functions */
+ buf[0] = 0x67452301;
+@@ -113,13 +131,17 @@
+ }
+
+ switch (hinfo->hash_version) {
++ case DX_HASH_LEGACY_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_LEGACY:
+- hash = dx_hack_hash(name, len);
++ hash = dx_hack_hash(name, len, unsigned_flag);
+ break;
++ case DX_HASH_HALF_MD4_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_HALF_MD4:
+ p = name;
+ while (len > 0) {
+- str2hashbuf(p, len, in, 8);
++ str2hashbuf(p, len, in, 8, unsigned_flag);
+ half_md4_transform(buf, in);
+ len -= 32;
+ p += 32;
+@@ -127,10 +149,12 @@
+ minor_hash = buf[2];
+ hash = buf[1];
+ break;
++ case DX_HASH_TEA_UNSIGNED:
++ unsigned_flag++;
+ case DX_HASH_TEA:
+ p = name;
+ while (len > 0) {
+- str2hashbuf(p, len, in, 4);
++ str2hashbuf(p, len, in, 4, unsigned_flag);
+ TEA_transform(buf, in);
+ len -= 16;
+ p += 16;
+diff -urN a/fs/ext4/namei.c b/fs/ext4/namei.c
+--- a/fs/ext4/namei.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext4/namei.c 2008-10-28 19:48:22.000000000 +0000
+@@ -361,6 +361,8 @@
+ goto fail;
+ }
+ hinfo->hash_version = root->info.hash_version;
++ if (hinfo->hash_version <= DX_HASH_TEA)
++ hinfo->hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo->seed = EXT4_SB(dir->i_sb)->s_hash_seed;
+ if (dentry)
+ ext4fs_dirhash(dentry->d_name.name, dentry->d_name.len, hinfo);
+@@ -639,6 +641,9 @@
+ dir = dir_file->f_path.dentry->d_inode;
+ if (!(EXT4_I(dir)->i_flags & EXT4_INDEX_FL)) {
+ hinfo.hash_version = EXT4_SB(dir->i_sb)->s_def_hash_version;
++ if (hinfo.hash_version <= DX_HASH_TEA)
++ hinfo.hash_version +=
++ EXT4_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed;
+ count = htree_dirblock_to_tree(dir_file, dir, 0, &hinfo,
+ start_hash, start_minor_hash);
+@@ -1415,6 +1420,8 @@
+
+ /* Initialize as for dx_probe */
+ hinfo.hash_version = root->info.hash_version;
++ if (hinfo.hash_version <= DX_HASH_TEA)
++ hinfo.hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned;
+ hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed;
+ ext4fs_dirhash(name, namelen, &hinfo);
+ frame = frames;
+diff -urN a/fs/ext4/super.c b/fs/ext4/super.c
+--- a/fs/ext4/super.c 2008-10-28 19:43:40.000000000 +0000
++++ b/fs/ext4/super.c 2008-10-28 19:48:22.000000000 +0000
+@@ -2071,6 +2071,21 @@
+ for (i=0; i < 4; i++)
+ sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]);
+ sbi->s_def_hash_version = es->s_def_hash_version;
++ i = le32_to_cpu(es->s_flags);
++ if (i & EXT2_FLAGS_UNSIGNED_HASH)
++ sbi->s_hash_unsigned = 3;
++ else if ((i & EXT2_FLAGS_SIGNED_HASH) == 0) {
++ char c;
++
++ c = (char) 255;
++ if (((int) c) == -1) {
++ es->s_flags |= cpu_to_le32(EXT2_FLAGS_SIGNED_HASH);
++ } else {
++ es->s_flags |= cpu_to_le32(EXT2_FLAGS_UNSIGNED_HASH);
++ sbi->s_hash_unsigned = 3;
++ }
++ sb->s_dirt = 1;
++ }
+
+ if (sbi->s_blocks_per_group > blocksize * 8) {
+ printk (KERN_ERR
Added: dists/trunk/linux-2.6/debian/patches/bugfix/all/sata_mv-clear_irq.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/all/sata_mv-clear_irq.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,61 @@
+From: Lennert Buytenhek <buytenh at wantstofly.org>
+Subject: [PATCH,RFC] sata_mv: don't avoid clearing interrupt status on SoC host adapters
+Date: Sun, 24 Aug 2008 05:04:29 +0200
+To: Saeed Bishara <saeed at marvell.com>, linux-ide at vger.kernel.org
+Cc: Mark Lord <liml at rtr.ca>, Jeff Garzik <jeff at garzik.org>
+
+For some reason, sata_mv doesn't clear interrupt status during init
+when it's running on an SoC host adapter. If the bootloader has
+touched the SATA controller before starting Linux, Linux can end up
+enabling the SATA interrupt with events pending, which will cause the
+interrupt to be marked as spurious and then be disabled, which then
+breaks all further accesses to the controller.
+
+This patch makes the SoC path clear interrupt status on init like in
+the non-SoC case.
+
+Signed-off-by: Lennert Buytenhek <buytenh at marvell.com>
+---
+ drivers/ata/sata_mv.c | 21 ++++++++++-----------
+ 1 files changed, 10 insertions(+), 11 deletions(-)
+
+diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
+index ad169ff..e829a3a 100644
+--- a/drivers/ata/sata_mv.c
++++ b/drivers/ata/sata_mv.c
+@@ -3131,19 +3131,18 @@ static int mv_init_host(struct ata_host *host, unsigned int board_idx)
+ writelfl(0, hc_mmio + HC_IRQ_CAUSE_OFS);
+ }
+
+- if (!IS_SOC(hpriv)) {
+- /* Clear any currently outstanding host interrupt conditions */
+- writelfl(0, mmio + hpriv->irq_cause_ofs);
++ /* Clear any currently outstanding host interrupt conditions */
++ writelfl(0, mmio + hpriv->irq_cause_ofs);
+
+- /* and unmask interrupt generation for host regs */
+- writelfl(hpriv->unmask_all_irqs, mmio + hpriv->irq_mask_ofs);
++ /* and unmask interrupt generation for host regs */
++ writelfl(hpriv->unmask_all_irqs, mmio + hpriv->irq_mask_ofs);
++
++ /*
++ * enable only global host interrupts for now.
++ * The per-port interrupts get done later as ports are set up.
++ */
++ mv_set_main_irq_mask(host, 0, PCI_ERR);
+
+- /*
+- * enable only global host interrupts for now.
+- * The per-port interrupts get done later as ports are set up.
+- */
+- mv_set_main_irq_mask(host, 0, PCI_ERR);
+- }
+ done:
+ return rc;
+ }
+--
+1.5.6.4
+--
+To unsubscribe from this list: send the line "unsubscribe linux-ide" in
+the body of a message to majordomo at vger.kernel.org
+More majordomo info at http://vger.kernel.org/majordomo-info.html
Modified: dists/trunk/linux-2.6/debian/patches/bugfix/all/stable/gen-patch
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/bugfix/all/stable/gen-patch (original)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/all/stable/gen-patch Sun Dec 28 16:55:48 2008
@@ -1,6 +1,6 @@
#!/usr/bin/env python
-import os.path, re, sys
+import os.path, re, sys, textwrap
class Version(object):
_rules = ur"^(\d+\.\d+\.\d+)\.(\d+)$"
@@ -45,6 +45,8 @@
if not line:
continue
hash, log = line.split(' ', 1)
+ log = textwrap.wrap(log, 74)
+ log = '\n '.join(log)
out.write(" - %s\n" % log)
if f.close() is not None:
raise RuntimeError
Added: dists/trunk/linux-2.6/debian/patches/bugfix/all/wireless-regulatory-default-EU.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/all/wireless-regulatory-default-EU.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,13 @@
+diff --git a/net/wireless/reg.c b/net/wireless/reg.c
+index 855bff4..5969480 100644
+--- a/net/wireless/reg.c
++++ b/net/wireless/reg.c
+@@ -29,7 +29,7 @@
+ #include <net/wireless.h>
+ #include "core.h"
+
+-static char *ieee80211_regdom = "US";
++static char *ieee80211_regdom = "EU";
+ module_param(ieee80211_regdom, charp, 0444);
+ MODULE_PARM_DESC(ieee80211_regdom, "IEEE 802.11 regulatory domain code");
+
Added: dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_ignore_pci.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_ignore_pci.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,64 @@
+diff --git a/arch/arm/mach-orion5x/common.h b/arch/arm/mach-orion5x/common.h
+index bd0f05d..c3ca3b1 100644
+--- a/arch/arm/mach-orion5x/common.h
++++ b/arch/arm/mach-orion5x/common.h
+@@ -32,6 +32,7 @@ struct pci_sys_data;
+ struct pci_bus;
+
+ void orion5x_pcie_id(u32 *dev, u32 *rev);
++void orion5x_pci_disable(void);
+ int orion5x_pci_sys_setup(int nr, struct pci_sys_data *sys);
+ struct pci_bus *orion5x_pci_sys_scan_bus(int nr, struct pci_sys_data *sys);
+ int orion5x_pci_map_irq(struct pci_dev *dev, u8 slot, u8 pin);
+diff --git a/arch/arm/mach-orion5x/kurobox_pro-setup.c b/arch/arm/mach-orion5x/kurobox_pro-setup.c
+index f5074b8..142148f 100644
+--- a/arch/arm/mach-orion5x/kurobox_pro-setup.c
++++ b/arch/arm/mach-orion5x/kurobox_pro-setup.c
+@@ -147,8 +147,10 @@ static struct hw_pci kurobox_pro_pci __initdata = {
+
+ static int __init kurobox_pro_pci_init(void)
+ {
+- if (machine_is_kurobox_pro())
++ if (machine_is_kurobox_pro()) {
++ orion5x_pci_disable();
+ pci_common_init(&kurobox_pro_pci);
++ }
+
+ return 0;
+ }
+diff --git a/arch/arm/mach-orion5x/pci.c b/arch/arm/mach-orion5x/pci.c
+index 9d5d39f..a3285da 100644
+--- a/arch/arm/mach-orion5x/pci.c
++++ b/arch/arm/mach-orion5x/pci.c
+@@ -522,6 +522,13 @@ static void __devinit rc_pci_fixup(struct pci_dev *dev)
+ }
+ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_MARVELL, PCI_ANY_ID, rc_pci_fixup);
+
++static int orion5x_pci_disabled __initdata;
++
++void __init orion5x_pci_disable(void)
++{
++ orion5x_pci_disabled = 1;
++}
++
+ int __init orion5x_pci_sys_setup(int nr, struct pci_sys_data *sys)
+ {
+ int ret = 0;
+@@ -529,7 +536,7 @@ int __init orion5x_pci_sys_setup(int nr, struct pci_sys_data *sys)
+ if (nr == 0) {
+ orion_pcie_set_local_bus_nr(PCIE_BASE, sys->busnr);
+ ret = pcie_setup(sys);
+- } else if (nr == 1) {
++ } else if (nr == 1 && !orion5x_pci_disabled) {
+ orion5x_pci_set_bus_nr(sys->busnr);
+ ret = pci_setup(sys);
+ }
+@@ -543,7 +550,7 @@ struct pci_bus __init *orion5x_pci_sys_scan_bus(int nr, struct pci_sys_data *sys
+
+ if (nr == 0) {
+ bus = pci_scan_bus(sys->busnr, &pcie_ops, sys);
+- } else if (nr == 1) {
++ } else if (nr == 1 && !orion5x_pci_disabled) {
+ bus = pci_scan_bus(sys->busnr, &pci_ops, sys);
+ } else {
+ bus = NULL;
Added: dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_powerdown.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/arm/kurobox_powerdown.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,181 @@
+2.6.26 backport of
+http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=a0087f2fcf5cb4e16502f5334168fbe304af318b
+
+[ARM] Orion: implement power-off method for Kurobox Pro
+Sylver Bruneau [Wed, 30 Apr 2008 06:14:58 +0000 (08:14 +0200)]
+This patch implements the communication with the microcontroller on the
+Kurobox Pro and Linkstation Pro/Live boards. This is allowing to send
+the commands needed to power-off the board correctly.
+
+Signed-off-by: Sylver Bruneau <sylver.bruneau at googlemail.com>
+Acked-by: Russell King <linux at arm.linux.org.uk>
+Signed-off-by: Lennert Buytenhek <buytenh at marvell.com>
+
+diff --git a/arch/arm/mach-orion5x/kurobox_pro-setup.c b/arch/arm/mach-orion5x/kurobox_pro-setup.c
+index f5074b8..7570c8b 100644
+--- a/arch/arm/mach-orion5x/kurobox_pro-setup.c
++++ b/arch/arm/mach-orion5x/kurobox_pro-setup.c
+@@ -13,10 +13,12 @@
+ #include <linux/platform_device.h>
+ #include <linux/pci.h>
+ #include <linux/irq.h>
++#include <linux/delay.h>
+ #include <linux/mtd/physmap.h>
+ #include <linux/mtd/nand.h>
+ #include <linux/mv643xx_eth.h>
+ #include <linux/i2c.h>
++#include <linux/serial_reg.h>
+ #include <linux/ata_platform.h>
+ #include <asm/mach-types.h>
+ #include <asm/gpio.h>
+@@ -179,6 +181,140 @@ static struct mv_sata_platform_data kurobox_pro_sata_data = {
+ };
+
+ /*****************************************************************************
++ * Kurobox Pro specific power off method via UART1-attached microcontroller
++ ****************************************************************************/
++
++#define UART1_REG(x) (UART1_VIRT_BASE + ((UART_##x) << 2))
++
++static int kurobox_pro_miconread(unsigned char *buf, int count)
++{
++ int i;
++ int timeout;
++
++ for (i = 0; i < count; i++) {
++ timeout = 10;
++
++ while (!(readl(UART1_REG(LSR)) & UART_LSR_DR)) {
++ if (--timeout == 0)
++ break;
++ udelay(1000);
++ }
++
++ if (timeout == 0)
++ break;
++ buf[i] = readl(UART1_REG(RX));
++ }
++
++ /* return read bytes */
++ return i;
++}
++
++static int kurobox_pro_miconwrite(const unsigned char *buf, int count)
++{
++ int i = 0;
++
++ while (count--) {
++ while (!(readl(UART1_REG(LSR)) & UART_LSR_THRE))
++ barrier();
++ writel(buf[i++], UART1_REG(TX));
++ }
++
++ return 0;
++}
++
++static int kurobox_pro_miconsend(const unsigned char *data, int count)
++{
++ int i;
++ unsigned char checksum = 0;
++ unsigned char recv_buf[40];
++ unsigned char send_buf[40];
++ unsigned char correct_ack[3];
++ int retry = 2;
++
++ /* Generate checksum */
++ for (i = 0; i < count; i++)
++ checksum -= data[i];
++
++ do {
++ /* Send data */
++ kurobox_pro_miconwrite(data, count);
++
++ /* send checksum */
++ kurobox_pro_miconwrite(&checksum, 1);
++
++ if (kurobox_pro_miconread(recv_buf, sizeof(recv_buf)) <= 3) {
++ printk(KERN_ERR ">%s: receive failed.\n", __func__);
++
++ /* send preamble to clear the receive buffer */
++ memset(&send_buf, 0xff, sizeof(send_buf));
++ kurobox_pro_miconwrite(send_buf, sizeof(send_buf));
++
++ /* make dummy reads */
++ mdelay(100);
++ kurobox_pro_miconread(recv_buf, sizeof(recv_buf));
++ } else {
++ /* Generate expected ack */
++ correct_ack[0] = 0x01;
++ correct_ack[1] = data[1];
++ correct_ack[2] = 0x00;
++
++ /* checksum Check */
++ if ((recv_buf[0] + recv_buf[1] + recv_buf[2] +
++ recv_buf[3]) & 0xFF) {
++ printk(KERN_ERR ">%s: Checksum Error : "
++ "Received data[%02x, %02x, %02x, %02x]"
++ "\n", __func__, recv_buf[0],
++ recv_buf[1], recv_buf[2], recv_buf[3]);
++ } else {
++ /* Check Received Data */
++ if (correct_ack[0] == recv_buf[0] &&
++ correct_ack[1] == recv_buf[1] &&
++ correct_ack[2] == recv_buf[2]) {
++ /* Interval for next command */
++ mdelay(10);
++
++ /* Receive ACK */
++ return 0;
++ }
++ }
++ /* Received NAK or illegal Data */
++ printk(KERN_ERR ">%s: Error : NAK or Illegal Data "
++ "Received\n", __func__);
++ }
++ } while (retry--);
++
++ /* Interval for next command */
++ mdelay(10);
++
++ return -1;
++}
++
++static void kurobox_pro_power_off(void)
++{
++ const unsigned char watchdogkill[] = {0x01, 0x35, 0x00};
++ const unsigned char shutdownwait[] = {0x00, 0x0c};
++ const unsigned char poweroff[] = {0x00, 0x06};
++ /* 38400 baud divisor */
++ const unsigned divisor = ((ORION5X_TCLK + (8 * 38400)) / (16 * 38400));
++
++ pr_info("%s: triggering power-off...\n", __func__);
++
++ /* hijack uart1 and reset into sane state (38400,8n1,even parity) */
++ writel(0x83, UART1_REG(LCR));
++ writel(divisor & 0xff, UART1_REG(DLL));
++ writel((divisor >> 8) & 0xff, UART1_REG(DLM));
++ writel(0x1b, UART1_REG(LCR));
++ writel(0x00, UART1_REG(IER));
++ writel(0x07, UART1_REG(FCR));
++ writel(0x00, UART1_REG(MCR));
++
++ /* Send the commands to shutdown the Kurobox Pro */
++ kurobox_pro_miconsend(watchdogkill, sizeof(watchdogkill)) ;
++ kurobox_pro_miconsend(shutdownwait, sizeof(shutdownwait)) ;
++ kurobox_pro_miconsend(poweroff, sizeof(poweroff));
++}
++
++/*****************************************************************************
+ * General Setup
+ ****************************************************************************/
+
+@@ -229,6 +365,9 @@ static void __init kurobox_pro_init(void)
+ i2c_register_board_info(0, &kurobox_pro_i2c_rtc, 1);
+ orion5x_eth_init(&kurobox_pro_eth_data);
+ orion5x_sata_init(&kurobox_pro_sata_data);
++
++ /* register Kurobox Pro specific power-off method */
++ pm_power_off = kurobox_pro_power_off;
+ }
+
+ #ifdef CONFIG_MACH_KUROBOX_PRO
Added: dists/trunk/linux-2.6/debian/patches/bugfix/parisc/add-lockless-fw-funcs.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/parisc/add-lockless-fw-funcs.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,135 @@
+From: Kyle McMartin <kyle at mcmartin.ca>
+Date: Tue, 29 Jul 2008 04:09:22 +0000 (-0400)
+Subject: parisc: add pdc_coproc_cfg_unlocked and set_firmware_width_unlocked
+X-Git-Url: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Fkyle%2Fparisc-2.6.git;a=commitdiff_plain;h=8735c10cff8486d58e03ee692ae9720f10cf9806
+
+parisc: add pdc_coproc_cfg_unlocked and set_firmware_width_unlocked
+
+These functions are called only when bringing up the monarch cpu,
+so it is safe to call them without taking the pdc spinlock. In the
+future, this may become relevant for lockdep, since these functions were
+taking spinlocks before start_kernel called the lockdep initializers.
+---
+
+Adjusted to apply to Debian's 2.6.26 by dann frazier <dannf at debian.org>
+
+diff -urpN linux-source-2.6.26.orig/arch/parisc/kernel/firmware.c linux-source-2.6.26/arch/parisc/kernel/firmware.c
+--- linux-source-2.6.26.orig/arch/parisc/kernel/firmware.c 2008-07-13 15:51:29.000000000 -0600
++++ linux-source-2.6.26/arch/parisc/kernel/firmware.c 2008-09-18 16:28:07.000000000 -0600
+@@ -150,26 +150,40 @@ static void convert_to_wide(unsigned lon
+ #endif
+ }
+
++#ifdef CONFIG_64BIT
++void __init set_firmware_width_unlocked(void)
++{
++ int ret;
++
++ ret = mem_pdc_call(PDC_MODEL, PDC_MODEL_CAPABILITIES,
++ __pa(pdc_result), 0);
++ convert_to_wide(pdc_result);
++ if (pdc_result[0] != NARROW_FIRMWARE)
++ parisc_narrow_firmware = 0;
++}
++
+ /**
+ * set_firmware_width - Determine if the firmware is wide or narrow.
+ *
+- * This function must be called before any pdc_* function that uses the convert_to_wide
+- * function.
++ * This function must be called before any pdc_* function that uses the
++ * convert_to_wide function.
+ */
+ void __init set_firmware_width(void)
+ {
+-#ifdef CONFIG_64BIT
+- int retval;
+ unsigned long flags;
++ spin_lock_irqsave(&pdc_lock, flags);
++ set_firmware_width_unlocked();
++ spin_unlock_irqrestore(&pdc_lock, flags);
++}
++#else
++void __init set_firmware_width_unlocked(void) {
++ return;
++}
+
+- spin_lock_irqsave(&pdc_lock, flags);
+- retval = mem_pdc_call(PDC_MODEL, PDC_MODEL_CAPABILITIES, __pa(pdc_result), 0);
+- convert_to_wide(pdc_result);
+- if(pdc_result[0] != NARROW_FIRMWARE)
+- parisc_narrow_firmware = 0;
+- spin_unlock_irqrestore(&pdc_lock, flags);
+-#endif
++void __init set_firmware_width(void) {
++ return;
+ }
++#endif /*CONFIG_64BIT*/
+
+ /**
+ * pdc_emergency_unlock - Unlock the linux pdc lock
+@@ -288,6 +302,20 @@ int pdc_chassis_warn(unsigned long *warn
+ return retval;
+ }
+
++int __init pdc_coproc_cfg_unlocked(struct pdc_coproc_cfg *pdc_coproc_info)
++{
++ int ret;
++
++ ret = mem_pdc_call(PDC_COPROC, PDC_COPROC_CFG, __pa(pdc_result));
++ convert_to_wide(pdc_result);
++ pdc_coproc_info->ccr_functional = pdc_result[0];
++ pdc_coproc_info->ccr_present = pdc_result[1];
++ pdc_coproc_info->revision = pdc_result[17];
++ pdc_coproc_info->model = pdc_result[18];
++
++ return ret;
++}
++
+ /**
+ * pdc_coproc_cfg - To identify coprocessors attached to the processor.
+ * @pdc_coproc_info: Return buffer address.
+@@ -297,19 +325,14 @@ int pdc_chassis_warn(unsigned long *warn
+ */
+ int __init pdc_coproc_cfg(struct pdc_coproc_cfg *pdc_coproc_info)
+ {
+- int retval;
++ int ret;
+ unsigned long flags;
+
+- spin_lock_irqsave(&pdc_lock, flags);
+- retval = mem_pdc_call(PDC_COPROC, PDC_COPROC_CFG, __pa(pdc_result));
+- convert_to_wide(pdc_result);
+- pdc_coproc_info->ccr_functional = pdc_result[0];
+- pdc_coproc_info->ccr_present = pdc_result[1];
+- pdc_coproc_info->revision = pdc_result[17];
+- pdc_coproc_info->model = pdc_result[18];
+- spin_unlock_irqrestore(&pdc_lock, flags);
++ spin_lock_irqsave(&pdc_lock, flags);
++ ret = pdc_coproc_cfg_unlocked(pdc_coproc_info);
++ spin_unlock_irqrestore(&pdc_lock, flags);
+
+- return retval;
++ return ret;
+ }
+
+ /**
+diff -urpN linux-source-2.6.26.orig/include/asm-parisc/pdc.h linux-source-2.6.26/include/asm-parisc/pdc.h
+--- linux-source-2.6.26.orig/include/asm-parisc/pdc.h 2008-07-13 15:51:29.000000000 -0600
++++ linux-source-2.6.26/include/asm-parisc/pdc.h 2008-09-18 16:28:07.000000000 -0600
+@@ -600,6 +600,7 @@ int pdc_chassis_info(struct pdc_chassis_
+ int pdc_chassis_disp(unsigned long disp);
+ int pdc_chassis_warn(unsigned long *warn);
+ int pdc_coproc_cfg(struct pdc_coproc_cfg *pdc_coproc_info);
++int pdc_coproc_cfg_unlocked(struct pdc_coproc_cfg *pdc_coproc_info);
+ int pdc_iodc_read(unsigned long *actcnt, unsigned long hpa, unsigned int index,
+ void *iodc_data, unsigned int iodc_data_size);
+ int pdc_system_map_find_mods(struct pdc_system_map_mod_info *pdc_mod_info,
+@@ -638,6 +639,7 @@ int pdc_mem_mem_table(struct pdc_memory_
+ #endif
+
+ void set_firmware_width(void);
++void set_firmware_width_unlocked(void);
+ int pdc_do_firm_test_reset(unsigned long ftc_bitmap);
+ int pdc_do_reset(void);
+ int pdc_soft_power_info(unsigned long *power_reg);
Added: dists/trunk/linux-2.6/debian/patches/bugfix/parisc/hijack-jump-to-start_kernel.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/parisc/hijack-jump-to-start_kernel.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,68 @@
+From: Kyle McMartin <kyle at mcmartin.ca>
+Date: Tue, 29 Jul 2008 04:11:13 +0000 (-0400)
+Subject: parisc: hijack jump to start_kernel
+X-Git-Url: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Fkyle%2Fparisc-2.6.git;a=commitdiff_plain;h=1a189c07f9f65305442d6e99efac9ac44f86bc04
+
+parisc: hijack jump to start_kernel
+
+Bang in our own start_parisc call, which initializes the PDC
+width, and turns on the FPU.
+
+Previously, if CONFIG_PRINTK_TIME was on, we'd attempt to use
+the FPU before we had enabled it, resulting in a difficult
+to diagnose panic.
+
+This patch causes init_per_cpu to redundantly set these for
+cpu0, but this is harmless.
+---
+
+diff --git a/arch/parisc/kernel/head.S b/arch/parisc/kernel/head.S
+index a84e31e..0e3d9f9 100644
+--- a/arch/parisc/kernel/head.S
++++ b/arch/parisc/kernel/head.S
+@@ -121,7 +121,7 @@ $pgt_fill_loop:
+ copy %r0,%r2
+
+ /* And the RFI Target address too */
+- load32 start_kernel,%r11
++ load32 start_parisc,%r11
+
+ /* And the initial task pointer */
+ load32 init_thread_union,%r6
+diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c
+index 39e7c5a..a59b71e 100644
+--- a/arch/parisc/kernel/setup.c
++++ b/arch/parisc/kernel/setup.c
+@@ -368,6 +368,31 @@ static int __init parisc_init(void)
+
+ return 0;
+ }
+-
+ arch_initcall(parisc_init);
+
++void start_parisc(void)
++{
++ extern void start_kernel(void);
++
++ int ret, cpunum;
++ struct pdc_coproc_cfg coproc_cfg;
++
++ cpunum = smp_processor_id();
++
++ set_firmware_width_unlocked();
++
++ ret = pdc_coproc_cfg_unlocked(&coproc_cfg);
++ if (ret >= 0 && coproc_cfg.ccr_functional) {
++ mtctl(coproc_cfg.ccr_functional, 10);
++
++ cpu_data[cpunum].fp_rev = coproc_cfg.revision;
++ cpu_data[cpunum].fp_model = coproc_cfg.model;
++
++ asm volatile ("fstd %fr0,8(%sp)");
++ } else {
++ panic("must have an fpu to boot linux");
++ }
++
++ start_kernel();
++ // not reached
++}
Modified: dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/lpar-console.patch
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/lpar-console.patch (original)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/powerpc/lpar-console.patch Sun Dec 28 16:55:48 2008
@@ -1,40 +1,39 @@
diff --git a/arch/powerpc/platforms/pseries/lpar.c b/arch/powerpc/platforms/pseries/lpar.c
-index 9235c46..626290d 100644
+index 9235c46..8a20452 100644
--- a/arch/powerpc/platforms/pseries/lpar.c
+++ b/arch/powerpc/platforms/pseries/lpar.c
-@@ -57,6 +57,7 @@ extern void pSeries_find_serial_port(void);
+@@ -204,6 +204,7 @@ void __init find_udbg_vterm(void)
+ struct device_node *stdout_node;
+ const u32 *termno;
+ const char *name;
++ int add_console;
+ /* find the boot console from /chosen/stdout */
+ if (!of_chosen)
+@@ -219,6 +220,8 @@ void __init find_udbg_vterm(void)
+ printk(KERN_WARNING "stdout node missing 'name' property!\n");
+ goto out;
+ }
++ /* The user has requested a console so this is already set up. */
++ add_console = !strstr(cmd_line, "console=");
- int vtermno; /* virtual terminal# for udbg */
-+static char *console_name;
-
- #define __ALIGNED__ __attribute__((__aligned__(sizeof(long))))
- static void udbg_hvsi_putc(char c)
-@@ -232,18 +233,24 @@ void __init find_udbg_vterm(void)
+ /* Check if it's a virtual terminal */
+ if (strncmp(name, "vty", 3) != 0)
+@@ -232,13 +235,15 @@ void __init find_udbg_vterm(void)
udbg_putc = udbg_putcLP;
udbg_getc = udbg_getcLP;
udbg_getc_poll = udbg_getc_pollLP;
- add_preferred_console("hvc", termno[0] & 0xff, NULL);
-+ console_name = "hvc";
++ if (add_console)
++ add_preferred_console("hvc", termno[0] & 0xff, NULL);
} else if (of_device_is_compatible(stdout_node, "hvterm-protocol")) {
-- vtermno = termno[0];
+ vtermno = termno[0];
udbg_putc = udbg_hvsi_putc;
udbg_getc = udbg_hvsi_getc;
udbg_getc_poll = udbg_hvsi_getc_poll;
- add_preferred_console("hvsi", termno[0] & 0xff, NULL);
-+ console_name = "hvsi";
++ if (add_console)
++ add_preferred_console("hvsi", termno[0] & 0xff, NULL);
}
out:
of_node_put(stdout_node);
- }
-
-+static void __init enable_vterm(void)
-+{
-+ if (console_name)
-+ add_preferred_console(console_name, vtermno, NULL);
-+}
-+arch_initcall(enable_vterm);
-+
- void vpa_init(int cpu)
- {
- int hwcpu = get_hard_smp_processor_id(cpu);
Added: dists/trunk/linux-2.6/debian/patches/bugfix/s390/prevent-ptrace-padding-area-read-write-in-31-bit-mode.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/s390/prevent-ptrace-padding-area-read-write-in-31-bit-mode.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,113 @@
+From 3d6e48f43340343d97839eadb1ab7b6a3ea98797 Mon Sep 17 00:00:00 2001
+From: Jarod Wilson <jwilson at redhat.com>
+Date: Tue, 9 Sep 2008 12:38:56 +0200
+Subject: S390: CVE-2008-1514: prevent ptrace padding area read/write in 31-bit mode
+
+From: Jarod Wilson <jwilson at redhat.com>
+
+commit 3d6e48f43340343d97839eadb1ab7b6a3ea98797 upstream
+
+When running a 31-bit ptrace, on either an s390 or s390x kernel,
+reads and writes into a padding area in struct user_regs_struct32
+will result in a kernel panic.
+
+This is also known as CVE-2008-1514.
+
+Test case available here:
+http://sources.redhat.com/cgi-bin/cvsweb.cgi/~checkout~/tests/ptrace-tests/tests/user-area-padding.c?cvsroot=systemtap
+
+Steps to reproduce:
+1) wget the above
+2) gcc -o user-area-padding-31bit user-area-padding.c -Wall -ggdb2 -D_GNU_SOURCE -m31
+3) ./user-area-padding-31bit
+<panic>
+
+Test status
+-----------
+Without patch, both s390 and s390x kernels panic. With patch, the test case,
+as well as the gdb testsuite, pass without incident, padding area reads
+returning zero, writes ignored.
+
+Nb: original version returned -EINVAL on write attempts, which broke the
+gdb test and made the test case slightly unhappy, Jan Kratochvil suggested
+the change to return 0 on write attempts.
+
+Signed-off-by: Jarod Wilson <jarod at redhat.com>
+Tested-by: Jan Kratochvil <jan.kratochvil at redhat.com>
+Signed-off-by: Martin Schwidefsky <schwidefsky at de.ibm.com>
+Cc: Moritz Muehlenhoff <jmm at debian.org>
+Signed-off-by: Greg Kroah-Hartman <gregkh at suse.de>
+
+---
+ arch/s390/kernel/compat_ptrace.h | 1 +
+ arch/s390/kernel/ptrace.c | 28 ++++++++++++++++++++++++++++
+ 2 files changed, 29 insertions(+)
+
+--- a/arch/s390/kernel/compat_ptrace.h
++++ b/arch/s390/kernel/compat_ptrace.h
+@@ -42,6 +42,7 @@ struct user_regs_struct32
+ u32 gprs[NUM_GPRS];
+ u32 acrs[NUM_ACRS];
+ u32 orig_gpr2;
++ /* nb: there's a 4-byte hole here */
+ s390_fp_regs fp_regs;
+ /*
+ * These per registers are in here so that gdb can modify them
+--- a/arch/s390/kernel/ptrace.c
++++ b/arch/s390/kernel/ptrace.c
+@@ -177,6 +177,13 @@ peek_user(struct task_struct *child, add
+ */
+ tmp = (addr_t) task_pt_regs(child)->orig_gpr2;
+
++ } else if (addr < (addr_t) &dummy->regs.fp_regs) {
++ /*
++ * prevent reads of padding hole between
++ * orig_gpr2 and fp_regs on s390.
++ */
++ tmp = 0;
++
+ } else if (addr < (addr_t) (&dummy->regs.fp_regs + 1)) {
+ /*
+ * floating point regs. are stored in the thread structure
+@@ -268,6 +275,13 @@ poke_user(struct task_struct *child, add
+ */
+ task_pt_regs(child)->orig_gpr2 = data;
+
++ } else if (addr < (addr_t) &dummy->regs.fp_regs) {
++ /*
++ * prevent writes of padding hole between
++ * orig_gpr2 and fp_regs on s390.
++ */
++ return 0;
++
+ } else if (addr < (addr_t) (&dummy->regs.fp_regs + 1)) {
+ /*
+ * floating point regs. are stored in the thread structure
+@@ -409,6 +423,13 @@ peek_user_emu31(struct task_struct *chil
+ */
+ tmp = *(__u32*)((addr_t) &task_pt_regs(child)->orig_gpr2 + 4);
+
++ } else if (addr < (addr_t) &dummy32->regs.fp_regs) {
++ /*
++ * prevent reads of padding hole between
++ * orig_gpr2 and fp_regs on s390.
++ */
++ tmp = 0;
++
+ } else if (addr < (addr_t) (&dummy32->regs.fp_regs + 1)) {
+ /*
+ * floating point regs. are stored in the thread structure
+@@ -488,6 +509,13 @@ poke_user_emu31(struct task_struct *chil
+ */
+ *(__u32*)((addr_t) &task_pt_regs(child)->orig_gpr2 + 4) = tmp;
+
++ } else if (addr < (addr_t) &dummy32->regs.fp_regs) {
++ /*
++ * prevent writess of padding hole between
++ * orig_gpr2 and fp_regs on s390.
++ */
++ return 0;
++
+ } else if (addr < (addr_t) (&dummy32->regs.fp_regs + 1)) {
+ /*
+ * floating point regs. are stored in the thread structure
Added: dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device-1.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device-1.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,148 @@
+From: Bjorn Helgaas <bjorn.helgaas at hp.com>
+Date: Tue, 14 Oct 2008 23:01:03 +0000 (-0600)
+Subject: x86: register a platform RTC device if PNP doesn't describe it
+X-Git-Url: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Ftorvalds%2Flinux-2.6.git;a=commitdiff_plain;h=758a7f7bb86b520aadc484f23da85e547b3bf3d8
+
+x86: register a platform RTC device if PNP doesn't describe it
+
+Most if not all x86 platforms have an RTC device, but sometimes the RTC
+is not exposed as a PNP0b00/PNP0b01/PNP0b02 device in PNPBIOS or ACPI:
+
+ http://bugzilla.kernel.org/show_bug.cgi?id=11580
+ https://bugzilla.redhat.com/show_bug.cgi?id=451188
+
+It's best if we can discover the RTC via PNP because then we know
+which flavor of device it is, where it lives, and which IRQ it uses.
+
+But if we can't, we should register a platform device using the
+compiled-in RTC_PORT/RTC_IRQ resource assumptions.
+
+Signed-off-by: Bjorn Helgaas <bjorn.helgaas at hp.com>
+Acked-by: Rafael J. Wysocki <rjw at sisk.pl>
+Acked-by: David Brownell <dbrownell at users.sourceforge.net>
+Reported-by: Rik Theys <rik.theys at esat.kuleuven.be>
+Reported-by: shr_msn at yahoo.com.tw
+Signed-off-by: Linus Torvalds <torvalds at linux-foundation.org>
+---
+
+diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c
+index 05191bb..0a23b57 100644
+--- a/arch/x86/kernel/rtc.c
++++ b/arch/x86/kernel/rtc.c
+@@ -223,11 +223,25 @@ static struct platform_device rtc_device = {
+ static __init int add_rtc_cmos(void)
+ {
+ #ifdef CONFIG_PNP
+- if (!pnp_platform_devices)
+- platform_device_register(&rtc_device);
+-#else
++ static const char *ids[] __initconst =
++ { "PNP0b00", "PNP0b01", "PNP0b02", };
++ struct pnp_dev *dev;
++ struct pnp_id *id;
++ int i;
++
++ pnp_for_each_dev(dev) {
++ for (id = dev->id; id; id = id->next) {
++ for (i = 0; i < ARRAY_SIZE(ids); i++) {
++ if (compare_pnp_id(id, ids[i]) != 0)
++ return 0;
++ }
++ }
++ }
++#endif
++
+ platform_device_register(&rtc_device);
+-#endif /* CONFIG_PNP */
++ dev_info(&rtc_device.dev,
++ "registered platform RTC device (no PNP device found)\n");
+ return 0;
+ }
+ device_initcall(add_rtc_cmos);
+
+From: Bjorn Helgaas <bjorn.helgaas at hp.com>
+Date: Tue, 14 Oct 2008 23:01:59 +0000 (-0600)
+Subject: rtc-cmos: look for PNP RTC first, then for platform RTC
+X-Git-Url: http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Ftorvalds%2Flinux-2.6.git;a=commitdiff_plain;h=72f22b1eb6ca5e4676a632a04d40d46cb61d4562
+
+rtc-cmos: look for PNP RTC first, then for platform RTC
+
+We shouldn't rely on "pnp_platform_devices" to tell us whether there
+is a PNP RTC device.
+
+I introduced "pnp_platform_devices", but I think it was a mistake.
+All it tells us is whether we found any PNPBIOS or PNPACPI devices.
+Many machines have some PNP devices, but do not describe the RTC
+via PNP. On those machines, we need to do the platform driver probe
+to find the RTC.
+
+We should just register the PNP driver and see whether it claims anything.
+If we don't find a PNP RTC, fall back to the platform driver probe.
+
+This (in conjunction with the arch/x86/kernel/rtc.c patch to add
+a platform RTC device when PNP doesn't have one) should resolve
+these issues:
+
+ http://bugzilla.kernel.org/show_bug.cgi?id=11580
+ https://bugzilla.redhat.com/show_bug.cgi?id=451188
+
+Signed-off-by: Bjorn Helgaas <bjorn.helgaas at hp.com>
+Acked-by: Rafael J. Wysocki <rjw at sisk.pl>
+Acked-by: David Brownell <dbrownell at users.sourceforge.net>
+Reported-by: Rik Theys <rik.theys at esat.kuleuven.be>
+Reported-by: shr_msn at yahoo.com.tw
+Signed-off-by: Linus Torvalds <torvalds at linux-foundation.org>
+---
+
+diff --git a/drivers/rtc/rtc-cmos.c b/drivers/rtc/rtc-cmos.c
+index 6778f82..963ad0b 100644
+--- a/drivers/rtc/rtc-cmos.c
++++ b/drivers/rtc/rtc-cmos.c
+@@ -1120,29 +1120,32 @@ static struct platform_driver cmos_platform_driver = {
+
+ static int __init cmos_init(void)
+ {
++ int retval = 0;
++
+ #ifdef CONFIG_PNP
+- if (pnp_platform_devices)
+- return pnp_register_driver(&cmos_pnp_driver);
+- else
+- return platform_driver_probe(&cmos_platform_driver,
+- cmos_platform_probe);
+-#else
+- return platform_driver_probe(&cmos_platform_driver,
+- cmos_platform_probe);
+-#endif /* CONFIG_PNP */
++ pnp_register_driver(&cmos_pnp_driver);
++#endif
++
++ if (!cmos_rtc.dev)
++ retval = platform_driver_probe(&cmos_platform_driver,
++ cmos_platform_probe);
++
++ if (retval == 0)
++ return 0;
++
++#ifdef CONFIG_PNP
++ pnp_unregister_driver(&cmos_pnp_driver);
++#endif
++ return retval;
+ }
+ module_init(cmos_init);
+
+ static void __exit cmos_exit(void)
+ {
+ #ifdef CONFIG_PNP
+- if (pnp_platform_devices)
+- pnp_unregister_driver(&cmos_pnp_driver);
+- else
+- platform_driver_unregister(&cmos_platform_driver);
+-#else
++ pnp_unregister_driver(&cmos_pnp_driver);
++#endif
+ platform_driver_unregister(&cmos_platform_driver);
+-#endif /* CONFIG_PNP */
+ }
+ module_exit(cmos_exit);
+
Added: dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/bugfix/x86/nonpnp-rtc-device.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,129 @@
+From: David Brownell <dbrownell at users.sourceforge.net>
+
+A bugzilla entry (http://bugzilla.kernel.org/show_bug.cgi?id=11580)
+reports that PNPACPI tables in some HP servers don't list RTC devices.
+
+Work around that on x86 (ignore ia64, the other user of ACPI) by
+having ACPI glue check for that case, and if necessary then setting
+up a platform device and having rtc_cmos use it.
+
+Signed-off-by: David Brownell <dbrownell at users.sourceforge.net>
+---
+ arch/x86/kernel/rtc.c | 13 ++++++++++++-
+ drivers/acpi/glue.c | 14 +++++++++++++-
+ drivers/pnp/core.c | 7 +++++++
+ drivers/rtc/rtc-cmos.c | 4 ++--
+ include/asm-x86/mc146818rtc.h | 5 +++++
+ include/linux/pnp.h | 1 +
+ 6 files changed, 40 insertions(+), 4 deletions(-)
+
+--- a/arch/x86/kernel/rtc.c
++++ b/arch/x86/kernel/rtc.c
+@@ -220,10 +220,21 @@ static struct platform_device rtc_device
+ .num_resources = ARRAY_SIZE(rtc_resources),
+ };
+
++#ifdef CONFIG_PNP
++/* PNPACPI tables sometimes omit the RTC, or are ignored */
++struct device *__init add_nonpnp_rtc_cmos(void)
++{
++ if (!rtc_device.dev.bus)
++ platform_device_register(&rtc_device);
++ return &rtc_device.dev;
++}
++#endif
++
+ static __init int add_rtc_cmos(void)
+ {
+ #ifdef CONFIG_PNP
+- if (!pnp_platform_devices)
++ /* sometimes pnpacpi=off */
++ if (!pnp_platform_devices && !rtc_device.dev.bus)
+ platform_device_register(&rtc_device);
+ #else
+ platform_device_register(&rtc_device);
+--- a/drivers/acpi/glue.c
++++ b/drivers/acpi/glue.c
+@@ -338,7 +338,19 @@ static int __init pnp_match(struct devic
+
+ static struct device *__init get_rtc_dev(void)
+ {
+- return bus_find_device(&pnp_bus_type, NULL, NULL, pnp_match);
++ struct device *rtc;
++
++ /* return RTC from PNPACPI tables */
++ rtc = bus_find_device(&pnp_bus_type, NULL, NULL, pnp_match);
++
++#ifdef ARCH_PNP_RTC_WORKAROUND
++ /* cope with buggy PNPACPI tables; seen on some HP DL3x0 servers */
++ if (!rtc) {
++ pnp_rtc_missing = true;
++ rtc = add_nonpnp_rtc_cmos();
++ }
++#endif
++ return rtc;
+ }
+
+ static int __init acpi_rtc_init(void)
+--- a/drivers/pnp/core.c
++++ b/drivers/pnp/core.c
+@@ -25,10 +25,17 @@ DEFINE_SPINLOCK(pnp_lock);
+ * ACPI or PNPBIOS should tell us about all platform devices, so we can
+ * skip some blind probes. ISAPNP typically enumerates only plug-in ISA
+ * devices, not built-in things like COM ports.
++ *
++ * Sometimes ACPI tables omit critical devices, assigning them to grab-bag
++ * nodes and forcing drivers to use a platform_device node or (yeech!) use
++ * poke-at-the-hardware algorithms for device discovery.
+ */
+ int pnp_platform_devices;
+ EXPORT_SYMBOL(pnp_platform_devices);
+
++bool pnp_rtc_missing;
++EXPORT_SYMBOL(pnp_rtc_missing);
++
+ void *pnp_alloc(long size)
+ {
+ void *result;
+--- a/drivers/rtc/rtc-cmos.c
++++ b/drivers/rtc/rtc-cmos.c
+@@ -1137,7 +1137,7 @@ static struct platform_driver cmos_platf
+ static int __init cmos_init(void)
+ {
+ #ifdef CONFIG_PNP
+- if (pnp_platform_devices)
++ if (pnp_platform_devices && !pnp_rtc_missing)
+ return pnp_register_driver(&cmos_pnp_driver);
+ else
+ return platform_driver_probe(&cmos_platform_driver,
+@@ -1152,7 +1152,7 @@ module_init(cmos_init);
+ static void __exit cmos_exit(void)
+ {
+ #ifdef CONFIG_PNP
+- if (pnp_platform_devices)
++ if (pnp_platform_devices && !pnp_rtc_missing)
+ pnp_unregister_driver(&cmos_pnp_driver);
+ else
+ platform_driver_unregister(&cmos_platform_driver);
+--- a/include/asm-x86/mc146818rtc.h
++++ b/include/asm-x86/mc146818rtc.h
+@@ -101,4 +101,9 @@ extern unsigned long mach_get_cmos_time(
+
+ #define RTC_IRQ 8
+
++#ifdef CONFIG_PNP
++#define ARCH_PNP_RTC_WORKAROUND
++extern struct device *add_nonpnp_rtc_cmos(void);
++#endif
++
+ #endif /* _ASM_MC146818RTC_H */
+--- a/include/linux/pnp.h
++++ b/include/linux/pnp.h
+@@ -420,6 +420,7 @@ int pnp_device_attach(struct pnp_dev *pn
+ void pnp_device_detach(struct pnp_dev *pnp_dev);
+ extern struct list_head pnp_global;
+ extern int pnp_platform_devices;
++extern bool pnp_rtc_missing;
+
+ /* multidevice card support */
+ struct pnp_dev *pnp_request_card_device(struct pnp_card_link *clink,
Added: dists/trunk/linux-2.6/debian/patches/debian/drivers-ata-pata_sis-postpone-pata.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/debian/drivers-ata-pata_sis-postpone-pata.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,69 @@
+diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
+index ae84949..f17c19b 100644
+--- a/drivers/ata/Kconfig
++++ b/drivers/ata/Kconfig
+@@ -172,7 +172,7 @@ config SATA_SIL
+ config SATA_SIS
+ tristate "SiS 964/965/966/180 SATA support"
+ depends on PCI
+- select PATA_SIS
++ select PATA_SIS_STUB
+ help
+ This option enables support for SiS Serial ATA on
+ SiS 964/965/966/180 and Parallel ATA on SiS 180.
+@@ -618,9 +618,13 @@ config PATA_SIL680
+
+ If unsure, say N.
+
++config PATA_SIS_STUB
++ tristate
++
+ config PATA_SIS
+ tristate "SiS PATA support"
+ depends on PCI
++ select PATA_SIS_STUB
+ help
+ This option enables support for SiS PATA controllers
+
+diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile
+index 674965f..c4bce57 100644
+--- a/drivers/ata/Makefile
++++ b/drivers/ata/Makefile
+@@ -63,7 +63,7 @@ obj-$(CONFIG_PATA_SIL680) += pata_sil680.o
+ obj-$(CONFIG_PATA_VIA) += pata_via.o
+ obj-$(CONFIG_PATA_WINBOND) += pata_sl82c105.o
+ obj-$(CONFIG_PATA_WINBOND_VLB) += pata_winbond.o
+-obj-$(CONFIG_PATA_SIS) += pata_sis.o
++obj-$(CONFIG_PATA_SIS_STUB) += pata_sis.o
+ obj-$(CONFIG_PATA_TRIFLEX) += pata_triflex.o
+ obj-$(CONFIG_PATA_IXP4XX_CF) += pata_ixp4xx_cf.o
+ obj-$(CONFIG_PATA_SCC) += pata_scc.o
+diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
+index 26345d7..2c3e3ba 100644
+--- a/drivers/ata/pata_sis.c
++++ b/drivers/ata/pata_sis.c
+@@ -826,13 +826,16 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+ }
+
+ static const struct pci_device_id sis_pci_tbl[] = {
++#ifdef CONFIG_PATA_SIS
+ { PCI_VDEVICE(SI, 0x5513), }, /* SiS 5513 */
+ { PCI_VDEVICE(SI, 0x5518), }, /* SiS 5518 */
+ { PCI_VDEVICE(SI, 0x1180), }, /* SiS 1180 */
++#endif
+
+ { }
+ };
+
++#ifdef CONFIG_PATA_SIS
+ static struct pci_driver sis_pci_driver = {
+ .name = DRV_NAME,
+ .id_table = sis_pci_tbl,
+@@ -856,6 +859,7 @@ static void __exit sis_exit(void)
+
+ module_init(sis_init);
+ module_exit(sis_exit);
++#endif
+
+ MODULE_AUTHOR("Alan Cox");
+ MODULE_DESCRIPTION("SCSI low-level driver for SiS ATA");
Modified: dists/trunk/linux-2.6/debian/patches/debian/scripts-kconfig-reportoldconfig.patch
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/debian/scripts-kconfig-reportoldconfig.patch (original)
+++ dists/trunk/linux-2.6/debian/patches/debian/scripts-kconfig-reportoldconfig.patch Sun Dec 28 16:55:48 2008
@@ -2,15 +2,9 @@
index 32e8c5a..7e14c56 100644
--- a/scripts/kconfig/Makefile
+++ b/scripts/kconfig/Makefile
-@@ -2,7 +2,7 @@
- # Kernel configuration targets
- # These targets are used from top-level makefile
-
+@@ -5,1 +5,1 @@
-PHONY += oldconfig xconfig gconfig menuconfig config silentoldconfig update-po-config
+PHONY += oldconfig xconfig gconfig menuconfig config reportoldconfig silentoldconfig updateoldconfig update-po-config
-
- Kconfig := arch/$(SRCARCH)/Kconfig
-
@@ -21,9 +21,15 @@ config: $(obj)/conf
oldconfig: $(obj)/conf
$< -o $(Kconfig)
Modified: dists/trunk/linux-2.6/debian/patches/features/all/vserver/bindmount-dev.patch
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/features/all/vserver/bindmount-dev.patch (original)
+++ dists/trunk/linux-2.6/debian/patches/features/all/vserver/bindmount-dev.patch Sun Dec 28 16:55:48 2008
@@ -7,19 +7,19 @@
+ if (!capable(CAP_SYS_ADMIN) && (old_nd.path.mnt->mnt_flags & MNT_NODEV))
+ mnt_flags |= MNT_NODEV;
+
- mnt->mnt_flags = mnt_flags;
- if (flags & MS_TAGID) {
- mnt->mnt_tag = tag;
+ err = graft_tree(mnt, &nd->path);
+ if (err) {
+ LIST_HEAD(umount_list);
@@ -1030,6 +1033,9 @@ static int do_remount(struct nameidata *
- if (nd->dentry != nd->mnt->mnt_root)
+ if (nd->path.dentry != nd->path.mnt->mnt_root)
return -EINVAL;
+ if (!capable(CAP_SYS_ADMIN))
+ mnt_flags |= MNT_NODEV;
+
down_write(&sb->s_umount);
- err = do_remount_sb(sb, flags, data, 0);
- if (!err)
+ if (flags & MS_BIND)
+ err = change_mount_flags(nd->path.mnt, flags);
@@ -1138,6 +1144,9 @@ static int do_new_mount(struct nameidata
if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT))
return -EPERM;
@@ -31,8 +31,8 @@
if (IS_ERR(mnt))
return PTR_ERR(mnt);
@@ -1489,8 +1498,6 @@ long do_mount(char *dev_name, char *dir_
- if (flags & MS_RELATIME)
- mnt_flags |= MNT_RELATIME;
+ if (flags & MS_RDONLY)
+ mnt_flags |= MNT_READONLY;
- if (!capable(CAP_SYS_ADMIN))
- mnt_flags |= MNT_NODEV;
Modified: dists/trunk/linux-2.6/debian/patches/features/all/vserver/gen-patch
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/features/all/vserver/gen-patch (original)
+++ dists/trunk/linux-2.6/debian/patches/features/all/vserver/gen-patch Sun Dec 28 16:55:48 2008
@@ -11,4 +11,4 @@
file="$(dirname $0)/$version.patch"
-filterdiff -p 1 --strip 1 --addprefix=a/ -x Makefile "$patch" | grep -v "^diff" > "$file"
+filterdiff -p 1 --strip 1 --addprefix=a/ -x Makefile -x include/linux/Kbuild -x include/linux/vserver/Kbuild "$patch" | grep -v "^diff" > "$file"
Added: dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35-update.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35-update.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,48 @@
+diff --git a/include/asm-s390/tlb.h b/include/asm-s390/tlb.h
+index 3d8a96d..656a647 100644
+--- a/include/asm-s390/tlb.h
++++ b/include/asm-s390/tlb.h
+@@ -27,6 +27,7 @@
+ #include <asm/pgalloc.h>
+ #include <asm/smp.h>
+ #include <asm/tlbflush.h>
++#include <linux/vs_memory.h>
+
+ #ifndef CONFIG_SMP
+ #define TLB_NR_PTRS 1
+diff --git a/include/linux/vserver/debug.h b/include/linux/vserver/debug.h
+index 2c989b5..c43bbb9 100644
+--- a/include/linux/vserver/debug.h
++++ b/include/linux/vserver/debug.h
+@@ -11,9 +11,6 @@
+ #define VXF_DEV "%p[%lu,%d:%d]"
+
+
+-#define __FUNC__ __func__
+-
+-
+ #define vxd_path(p) \
+ ({ static char _buffer[PATH_MAX]; \
+ d_path(p, _buffer, sizeof(_buffer)); })
+diff --git a/include/linux/vserver/pid.h b/include/linux/vserver/pid.h
+index eb0b9cb..c76a6ee 100644
+--- a/include/linux/vserver/pid.h
++++ b/include/linux/vserver/pid.h
+@@ -4,7 +4,7 @@
+ /* pid faking stuff */
+
+ #define vx_info_map_pid(v, p) \
+- __vx_info_map_pid((v), (p), __FUNC__, __FILE__, __LINE__)
++ __vx_info_map_pid((v), (p), __func__, __FILE__, __LINE__)
+ #define vx_info_map_tgid(v,p) vx_info_map_pid(v,p)
+ #define vx_map_pid(p) vx_info_map_pid(current->vx_info, p)
+ #define vx_map_tgid(p) vx_map_pid(p)
+@@ -27,7 +27,7 @@ static inline int __vx_info_map_pid(struct vx_info *vxi, int pid,
+ }
+
+ #define vx_info_rmap_pid(v, p) \
+- __vx_info_rmap_pid((v), (p), __FUNC__, __FILE__, __LINE__)
++ __vx_info_rmap_pid((v), (p), __func__, __FILE__, __LINE__)
+ #define vx_rmap_pid(p) vx_info_rmap_pid(current->vx_info, p)
+ #define vx_rmap_tgid(p) vx_rmap_pid(p)
+
Added: dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/vserver/vs2.3.0.35.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,27921 @@
+--- a/arch/alpha/Kconfig 2008-04-17 12:05:26.000000000 -0400
++++ a/arch/alpha/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -671,6 +671,8 @@ config DUMMY_CONSOLE
+ depends on VGA_HOSE
+ default y
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/alpha/kernel/entry.S 2008-04-17 11:31:21.000000000 -0400
++++ a/arch/alpha/kernel/entry.S 2008-07-16 22:41:36.000000000 -0400
+@@ -872,24 +872,15 @@ sys_getxgid:
+ .globl sys_getxpid
+ .ent sys_getxpid
+ sys_getxpid:
++ lda $sp, -16($sp)
++ stq $26, 0($sp)
+ .prologue 0
+- ldq $2, TI_TASK($8)
+
+- /* See linux/kernel/timer.c sys_getppid for discussion
+- about this loop. */
+- ldq $3, TASK_GROUP_LEADER($2)
+- ldq $4, TASK_REAL_PARENT($3)
+- ldl $0, TASK_TGID($2)
+-1: ldl $1, TASK_TGID($4)
+-#ifdef CONFIG_SMP
+- mov $4, $5
+- mb
+- ldq $3, TASK_GROUP_LEADER($2)
+- ldq $4, TASK_REAL_PARENT($3)
+- cmpeq $4, $5, $5
+- beq $5, 1b
+-#endif
+- stq $1, 80($sp)
++ lda $16, 96($sp)
++ jsr $26, do_getxpid
++ ldq $26, 0($sp)
++
++ lda $sp, 16($sp)
+ ret
+ .end sys_getxpid
+
+--- a/arch/alpha/kernel/osf_sys.c 2008-07-14 17:22:23.000000000 -0400
++++ a/arch/alpha/kernel/osf_sys.c 2008-07-16 22:41:36.000000000 -0400
+@@ -883,7 +883,7 @@ osf_gettimeofday(struct timeval32 __user
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/alpha/kernel/ptrace.c 2008-04-17 11:31:21.000000000 -0400
++++ a/arch/alpha/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -15,6 +15,7 @@
+ #include <linux/slab.h>
+ #include <linux/security.h>
+ #include <linux/signal.h>
++#include <linux/vs_base.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/pgtable.h>
+--- a/arch/alpha/kernel/systbls.S 2008-04-17 12:05:26.000000000 -0400
++++ a/arch/alpha/kernel/systbls.S 2008-07-16 22:41:36.000000000 -0400
+@@ -446,7 +446,7 @@ sys_call_table:
+ .quad sys_stat64 /* 425 */
+ .quad sys_lstat64
+ .quad sys_fstat64
+- .quad sys_ni_syscall /* sys_vserver */
++ .quad sys_vserver /* sys_vserver */
+ .quad sys_ni_syscall /* sys_mbind */
+ .quad sys_ni_syscall /* sys_get_mempolicy */
+ .quad sys_ni_syscall /* sys_set_mempolicy */
+--- a/arch/alpha/kernel/traps.c 2008-07-14 17:22:23.000000000 -0400
++++ a/arch/alpha/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -183,7 +183,8 @@ die_if_kernel(char * str, struct pt_regs
+ #ifdef CONFIG_SMP
+ printk("CPU %d ", hard_smp_processor_id());
+ #endif
+- printk("%s(%d): %s %ld\n", current->comm, task_pid_nr(current), str, err);
++ printk("%s(%d[#%u]): %s %ld\n", current->comm,
++ task_pid_nr(current), current->xid, str, err);
+ dik_show_regs(regs, r9_15);
+ add_taint(TAINT_DIE);
+ dik_show_trace((unsigned long *)(regs+1));
+--- a/arch/alpha/mm/fault.c 2008-04-17 11:31:21.000000000 -0400
++++ a/arch/alpha/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -193,8 +193,8 @@ do_page_fault(unsigned long address, uns
+ down_read(&mm->mmap_sem);
+ goto survive;
+ }
+- printk(KERN_ALERT "VM: killing process %s(%d)\n",
+- current->comm, task_pid_nr(current));
++ printk(KERN_ALERT "VM: killing process %s(%d:#%u)\n",
++ current->comm, task_pid_nr(current), current->xid);
+ if (!user_mode(regs))
+ goto no_context;
+ do_group_exit(SIGKILL);
+--- a/arch/arm/Kconfig 2008-07-14 17:22:23.000000000 -0400
++++ a/arch/arm/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -1187,6 +1187,8 @@ source "fs/Kconfig"
+
+ source "arch/arm/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/arm/kernel/calls.S 2008-07-14 17:22:23.000000000 -0400
++++ a/arch/arm/kernel/calls.S 2008-07-16 22:41:36.000000000 -0400
+@@ -322,7 +322,7 @@
+ /* 310 */ CALL(sys_request_key)
+ CALL(sys_keyctl)
+ CALL(ABI(sys_semtimedop, sys_oabi_semtimedop))
+-/* vserver */ CALL(sys_ni_syscall)
++ CALL(sys_vserver)
+ CALL(sys_ioprio_set)
+ /* 315 */ CALL(sys_ioprio_get)
+ CALL(sys_inotify_init)
+--- a/arch/arm/kernel/process.c 2008-04-17 12:05:26.000000000 -0400
++++ a/arch/arm/kernel/process.c 2008-07-16 22:41:36.000000000 -0400
+@@ -264,7 +264,8 @@ void __show_regs(struct pt_regs *regs)
+ void show_regs(struct pt_regs * regs)
+ {
+ printk("\n");
+- printk("Pid: %d, comm: %20s\n", task_pid_nr(current), current->comm);
++ printk("Pid: %d[#%u], comm: %20s\n",
++ task_pid_nr(current), current->xid, current->comm);
+ __show_regs(regs);
+ __backtrace();
+ }
+--- a/arch/arm/kernel/traps.c 2008-04-17 12:05:26.000000000 -0400
++++ a/arch/arm/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -214,8 +214,8 @@ static void __die(const char *str, int e
+ str, err, ++die_counter);
+ print_modules();
+ __show_regs(regs);
+- printk("Process %s (pid: %d, stack limit = 0x%p)\n",
+- tsk->comm, task_pid_nr(tsk), thread + 1);
++ printk("Process %s (pid: %d:#%u, stack limit = 0x%p)\n",
++ tsk->comm, task_pid_nr(tsk), tsk->xid, thread + 1);
+
+ if (!user_mode(regs) || in_interrupt()) {
+ dump_mem("Stack: ", regs->ARM_sp,
+--- a/arch/arm/mm/fault.c 2008-04-17 12:05:27.000000000 -0400
++++ a/arch/arm/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -292,7 +292,8 @@ do_page_fault(unsigned long addr, unsign
+ * happened to us that made us unable to handle
+ * the page fault gracefully.
+ */
+- printk("VM: killing process %s\n", tsk->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ tsk->comm, task_pid_nr(tsk), tsk->xid);
+ do_group_exit(SIGKILL);
+ return 0;
+ }
+--- a/arch/cris/Kconfig 2008-04-17 12:05:27.000000000 -0400
++++ a/arch/cris/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -679,6 +679,8 @@ source "drivers/usb/Kconfig"
+
+ source "arch/cris/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/frv/kernel/kernel_thread.S 2007-02-04 13:44:54.000000000 -0500
++++ a/arch/frv/kernel/kernel_thread.S 2008-07-16 22:41:36.000000000 -0400
+@@ -37,7 +37,7 @@ kernel_thread:
+
+ # start by forking the current process, but with shared VM
+ setlos.p #__NR_clone,gr7 ; syscall number
+- ori gr10,#CLONE_VM,gr8 ; first syscall arg [clone_flags]
++ ori gr10,#CLONE_KT,gr8 ; first syscall arg [clone_flags]
+ sethi.p #0xe4e4,gr9 ; second syscall arg [newsp]
+ setlo #0xe4e4,gr9
+ setlos.p #0,gr10 ; third syscall arg [parent_tidptr]
+--- a/arch/h8300/Kconfig 2008-04-17 12:05:28.000000000 -0400
++++ a/arch/h8300/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -233,6 +233,8 @@ source "fs/Kconfig"
+
+ source "arch/h8300/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/ia64/ia32/ia32_entry.S 2008-04-17 10:37:14.000000000 -0400
++++ a/arch/ia64/ia32/ia32_entry.S 2008-07-16 22:41:36.000000000 -0400
+@@ -446,7 +446,7 @@ ia32_syscall_table:
+ data8 sys_tgkill /* 270 */
+ data8 compat_sys_utimes
+ data8 sys32_fadvise64_64
+- data8 sys_ni_syscall
++ data8 sys32_vserver
+ data8 sys_ni_syscall
+ data8 sys_ni_syscall /* 275 */
+ data8 sys_ni_syscall
+--- a/arch/ia64/ia32/sys_ia32.c 2008-07-14 17:22:26.000000000 -0400
++++ a/arch/ia64/ia32/sys_ia32.c 2008-07-16 22:41:36.000000000 -0400
+@@ -1178,7 +1178,7 @@ sys32_gettimeofday (struct compat_timeva
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/ia64/Kconfig 2008-07-14 17:22:26.000000000 -0400
++++ a/arch/ia64/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -638,6 +638,8 @@ source "arch/ia64/hp/sim/Kconfig"
+
+ source "arch/ia64/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/ia64/kernel/entry.S 2008-07-14 17:22:26.000000000 -0400
++++ a/arch/ia64/kernel/entry.S 2008-07-16 22:41:36.000000000 -0400
+@@ -1619,7 +1619,7 @@ sys_call_table:
+ data8 sys_mq_notify
+ data8 sys_mq_getsetattr
+ data8 sys_kexec_load
+- data8 sys_ni_syscall // reserved for vserver
++ data8 sys_vserver
+ data8 sys_waitid // 1270
+ data8 sys_add_key
+ data8 sys_request_key
+--- a/arch/ia64/kernel/perfmon.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/ia64/kernel/perfmon.c 2008-07-16 22:41:36.000000000 -0400
+@@ -40,6 +40,7 @@
+ #include <linux/capability.h>
+ #include <linux/rcupdate.h>
+ #include <linux/completion.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/errno.h>
+ #include <asm/intrinsics.h>
+@@ -2376,7 +2377,7 @@ pfm_smpl_buffer_alloc(struct task_struct
+ */
+ insert_vm_struct(mm, vma);
+
+- mm->total_vm += size >> PAGE_SHIFT;
++ vx_vmpages_add(mm, size >> PAGE_SHIFT);
+ vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file,
+ vma_pages(vma));
+ up_write(&task->mm->mmap_sem);
+--- a/arch/ia64/kernel/process.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/ia64/kernel/process.c 2008-07-16 22:41:36.000000000 -0400
+@@ -105,8 +105,8 @@ show_regs (struct pt_regs *regs)
+ unsigned long ip = regs->cr_iip + ia64_psr(regs)->ri;
+
+ print_modules();
+- printk("\nPid: %d, CPU %d, comm: %20s\n", task_pid_nr(current),
+- smp_processor_id(), current->comm);
++ printk("\nPid: %d[#%u], CPU %d, comm: %20s\n", task_pid_nr(current),
++ current->xid, smp_processor_id(), current->comm);
+ printk("psr : %016lx ifs : %016lx ip : [<%016lx>] %s (%s)\n",
+ regs->cr_ipsr, regs->cr_ifs, ip, print_tainted(),
+ init_utsname()->release);
+--- a/arch/ia64/kernel/ptrace.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/ia64/kernel/ptrace.c 2008-07-17 21:01:48.000000000 -0400
+@@ -22,6 +22,7 @@
+ #include <linux/signal.h>
+ #include <linux/regset.h>
+ #include <linux/elf.h>
++#include <linux/vs_base.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/processor.h>
+--- a/arch/ia64/kernel/traps.c 2008-04-17 12:05:28.000000000 -0400
++++ a/arch/ia64/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -60,8 +60,9 @@ die (const char *str, struct pt_regs *re
+ put_cpu();
+
+ if (++die.lock_owner_depth < 3) {
+- printk("%s[%d]: %s %ld [%d]\n",
+- current->comm, task_pid_nr(current), str, err, ++die_counter);
++ printk("%s[%d[#%u]]: %s %ld [%d]\n",
++ current->comm, task_pid_nr(current), current->xid,
++ str, err, ++die_counter);
+ if (notify_die(DIE_OOPS, str, regs, err, 255, SIGSEGV)
+ != NOTIFY_STOP)
+ show_regs(regs);
+@@ -324,8 +325,9 @@ handle_fpu_swa (int fp_fault, struct pt_
+ if ((last.count & 15) < 5 && (ia64_fetchadd(1, &last.count, acq) & 15) < 5) {
+ last.time = current_jiffies + 5 * HZ;
+ printk(KERN_WARNING
+- "%s(%d): floating-point assist fault at ip %016lx, isr %016lx\n",
+- current->comm, task_pid_nr(current), regs->cr_iip + ia64_psr(regs)->ri, isr);
++ "%s(%d[#%u]): floating-point assist fault at ip %016lx, isr %016lx\n",
++ current->comm, task_pid_nr(current), current->xid,
++ regs->cr_iip + ia64_psr(regs)->ri, isr);
+ }
+ }
+ }
+--- a/arch/ia64/mm/fault.c 2008-04-17 12:05:28.000000000 -0400
++++ a/arch/ia64/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -10,6 +10,7 @@
+ #include <linux/interrupt.h>
+ #include <linux/kprobes.h>
+ #include <linux/kdebug.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/processor.h>
+--- a/arch/m32r/kernel/traps.c 2008-04-17 11:31:23.000000000 -0400
++++ a/arch/m32r/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -195,8 +195,9 @@ static void show_registers(struct pt_reg
+ } else {
+ printk("SPI: %08lx\n", sp);
+ }
+- printk("Process %s (pid: %d, process nr: %d, stackpage=%08lx)",
+- current->comm, task_pid_nr(current), 0xffff & i, 4096+(unsigned long)current);
++ printk("Process %s (pid: %d[#%u], process nr: %d, stackpage=%08lx)",
++ current->comm, task_pid_nr(current), current->xid,
++ 0xffff & i, 4096+(unsigned long)current);
+
+ /*
+ * When in-kernel, we also print out the stack and code at the
+--- a/arch/m68k/Kconfig 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/m68k/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -667,6 +667,8 @@ source "fs/Kconfig"
+
+ source "arch/m68k/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/m68k/kernel/ptrace.c 2008-04-17 11:31:23.000000000 -0400
++++ a/arch/m68k/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/ptrace.h>
+ #include <linux/user.h>
+ #include <linux/signal.h>
++#include <linux/vs_base.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/page.h>
+@@ -269,6 +270,8 @@ long arch_ptrace(struct task_struct *chi
+ ret = ptrace_request(child, request, addr, data);
+ break;
+ }
++ if (!vx_check(vx_task_xid(child), VS_WATCH_P | VS_IDENT))
++ goto out_tsk;
+
+ return ret;
+ out_eio:
+--- a/arch/m68k/kernel/traps.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/m68k/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -909,8 +909,8 @@ void show_registers(struct pt_regs *regs
+ printk("d4: %08lx d5: %08lx a0: %08lx a1: %08lx\n",
+ regs->d4, regs->d5, regs->a0, regs->a1);
+
+- printk("Process %s (pid: %d, task=%p)\n",
+- current->comm, task_pid_nr(current), current);
++ printk("Process %s (pid: %d[#%u], task=%p)\n",
++ current->comm, task_pid_nr(current), current->xid, current);
+ addr = (unsigned long)&fp->un;
+ printk("Frame format=%X ", regs->format);
+ switch (regs->format) {
+--- a/arch/m68knommu/Kconfig 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/m68knommu/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -725,6 +725,8 @@ source "fs/Kconfig"
+
+ source "arch/m68knommu/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/m68knommu/kernel/traps.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/m68knommu/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -79,8 +79,9 @@ void die_if_kernel(char *str, struct pt_
+ printk(KERN_EMERG "d4: %08lx d5: %08lx a0: %08lx a1: %08lx\n",
+ fp->d4, fp->d5, fp->a0, fp->a1);
+
+- printk(KERN_EMERG "Process %s (pid: %d, stackpage=%08lx)\n",
+- current->comm, current->pid, PAGE_SIZE+(unsigned long)current);
++ printk(KERN_EMERG "Process %s (pid: %d[#%u], stackpage=%08lx)\n",
++ current->comm, task_pid_nr(current), current->xid,
++ PAGE_SIZE+(unsigned long)current);
+ show_stack(NULL, (unsigned long *)(fp + 1));
+ add_taint(TAINT_DIE);
+ do_exit(SIGSEGV);
+--- a/arch/mips/Kconfig 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/mips/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -2131,6 +2131,8 @@ source "fs/Kconfig"
+
+ source "arch/mips/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/mips/kernel/linux32.c 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/mips/kernel/linux32.c 2008-07-16 22:41:36.000000000 -0400
+@@ -209,7 +209,7 @@ sys32_gettimeofday(struct compat_timeval
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/mips/kernel/ptrace.c 2008-04-17 11:31:23.000000000 -0400
++++ a/arch/mips/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -25,6 +25,7 @@
+ #include <linux/security.h>
+ #include <linux/audit.h>
+ #include <linux/seccomp.h>
++#include <linux/vs_base.h>
+
+ #include <asm/byteorder.h>
+ #include <asm/cpu.h>
+@@ -171,6 +172,9 @@ long arch_ptrace(struct task_struct *chi
+ {
+ int ret;
+
++ if (!vx_check(vx_task_xid(child), VS_WATCH_P | VS_IDENT))
++ goto out;
++
+ switch (request) {
+ /* when I and D space are separate, these will need to be fixed. */
+ case PTRACE_PEEKTEXT: /* read word at location addr. */
+--- a/arch/mips/kernel/scall32-o32.S 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/mips/kernel/scall32-o32.S 2008-07-16 22:41:36.000000000 -0400
+@@ -619,7 +619,7 @@ einval: li v0, -EINVAL
+ sys sys_mq_timedreceive 5
+ sys sys_mq_notify 2 /* 4275 */
+ sys sys_mq_getsetattr 3
+- sys sys_ni_syscall 0 /* sys_vserver */
++ sys sys_vserver 3
+ sys sys_waitid 5
+ sys sys_ni_syscall 0 /* available, was setaltroot */
+ sys sys_add_key 5 /* 4280 */
+--- a/arch/mips/kernel/scall64-64.S 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/mips/kernel/scall64-64.S 2008-07-16 22:41:36.000000000 -0400
+@@ -434,7 +434,7 @@ sys_call_table:
+ PTR sys_mq_timedreceive
+ PTR sys_mq_notify
+ PTR sys_mq_getsetattr /* 5235 */
+- PTR sys_ni_syscall /* sys_vserver */
++ PTR sys_vserver
+ PTR sys_waitid
+ PTR sys_ni_syscall /* available, was setaltroot */
+ PTR sys_add_key
+--- a/arch/mips/kernel/scall64-n32.S 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/mips/kernel/scall64-n32.S 2008-07-16 22:41:36.000000000 -0400
+@@ -360,7 +360,7 @@ EXPORT(sysn32_call_table)
+ PTR compat_sys_mq_timedreceive
+ PTR compat_sys_mq_notify
+ PTR compat_sys_mq_getsetattr
+- PTR sys_ni_syscall /* 6240, sys_vserver */
++ PTR sys32_vserver /* 6240 */
+ PTR compat_sys_waitid
+ PTR sys_ni_syscall /* available, was setaltroot */
+ PTR sys_add_key
+--- a/arch/mips/kernel/scall64-o32.S 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/mips/kernel/scall64-o32.S 2008-07-16 22:41:36.000000000 -0400
+@@ -482,7 +482,7 @@ sys_call_table:
+ PTR compat_sys_mq_timedreceive
+ PTR compat_sys_mq_notify /* 4275 */
+ PTR compat_sys_mq_getsetattr
+- PTR sys_ni_syscall /* sys_vserver */
++ PTR sys32_vserver
+ PTR sys32_waitid
+ PTR sys_ni_syscall /* available, was setaltroot */
+ PTR sys_add_key /* 4280 */
+--- a/arch/mips/kernel/traps.c 2008-07-14 17:22:27.000000000 -0400
++++ a/arch/mips/kernel/traps.c 2008-07-17 21:08:35.000000000 -0400
+@@ -324,9 +324,10 @@ void show_registers(const struct pt_regs
+
+ __show_regs(regs);
+ print_modules();
+- printk("Process %s (pid: %d, threadinfo=%p, task=%p, tls=%0*lx)\n",
+- current->comm, current->pid, current_thread_info(), current,
+- field, current_thread_info()->tp_value);
++ printk("Process %s (pid: %d:#%u, threadinfo=%p, task=%p, tls=%0*lx)\n",
++ current->comm, task_pid_nr(current), current->xid,
++ current_thread_info(), current,
++ field, current_thread_info()->tp_value);
+ if (cpu_has_userlocal) {
+ unsigned long tls;
+
+--- a/arch/mips/mm/fault.c 2008-04-17 11:31:24.000000000 -0400
++++ a/arch/mips/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -178,7 +178,8 @@ out_of_memory:
+ down_read(&mm->mmap_sem);
+ goto survive;
+ }
+- printk("VM: killing process %s\n", tsk->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ tsk->comm, tsk->pid, tsk->xid);
+ if (user_mode(regs))
+ do_group_exit(SIGKILL);
+ goto no_context;
+--- a/arch/parisc/Kconfig 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/parisc/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -278,6 +278,8 @@ source "fs/Kconfig"
+
+ source "arch/parisc/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/parisc/kernel/syscall_table.S 2008-04-17 12:05:29.000000000 -0400
++++ a/arch/parisc/kernel/syscall_table.S 2008-07-16 22:41:36.000000000 -0400
+@@ -361,7 +361,7 @@
+ ENTRY_COMP(mbind) /* 260 */
+ ENTRY_COMP(get_mempolicy)
+ ENTRY_COMP(set_mempolicy)
+- ENTRY_SAME(ni_syscall) /* 263: reserved for vserver */
++ ENTRY_DIFF(vserver)
+ ENTRY_SAME(add_key)
+ ENTRY_SAME(request_key) /* 265 */
+ ENTRY_SAME(keyctl)
+--- a/arch/parisc/kernel/sys_parisc32.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/parisc/kernel/sys_parisc32.c 2008-07-16 22:41:36.000000000 -0400
+@@ -203,11 +203,11 @@ static inline long get_ts32(struct times
+ asmlinkage int
+ sys32_gettimeofday(struct compat_timeval __user *tv, struct timezone __user *tz)
+ {
+- extern void do_gettimeofday(struct timeval *tv);
++ extern void vx_gettimeofday(struct timeval *tv);
+
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_compat_timeval(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/parisc/kernel/traps.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/parisc/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -237,8 +237,9 @@ void die_if_kernel(char *str, struct pt_
+ if (err == 0)
+ return; /* STFU */
+
+- printk(KERN_CRIT "%s (pid %d): %s (code %ld) at " RFMT "\n",
+- current->comm, task_pid_nr(current), str, err, regs->iaoq[0]);
++ printk(KERN_CRIT "%s (pid %d:#%u): %s (code %ld) at " RFMT "\n",
++ current->comm, task_pid_nr(current), current->xid,
++ str, err, regs->iaoq[0]);
+ #ifdef PRINT_USER_FAULTS
+ /* XXX for debugging only */
+ show_regs(regs);
+@@ -270,8 +271,8 @@ KERN_CRIT " || |
+ pdc_console_restart();
+
+ if (err)
+- printk(KERN_CRIT "%s (pid %d): %s (code %ld)\n",
+- current->comm, task_pid_nr(current), str, err);
++ printk(KERN_CRIT "%s (pid %d:#%u): %s (code %ld)\n",
++ current->comm, task_pid_nr(current), current->xid, str, err);
+
+ /* Wot's wrong wif bein' racy? */
+ if (current->thread.flags & PARISC_KERNEL_DEATH) {
+--- a/arch/parisc/mm/fault.c 2008-04-17 11:31:24.000000000 -0400
++++ a/arch/parisc/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -210,8 +210,9 @@ bad_area:
+
+ #ifdef PRINT_USER_FAULTS
+ printk(KERN_DEBUG "\n");
+- printk(KERN_DEBUG "do_page_fault() pid=%d command='%s' type=%lu address=0x%08lx\n",
+- task_pid_nr(tsk), tsk->comm, code, address);
++ printk(KERN_DEBUG "do_page_fault() pid=%d:#%u "
++ "command='%s' type=%lu address=0x%08lx\n",
++ task_pid_nr(tsk), tsk->xid, tsk->comm, code, address);
+ if (vma) {
+ printk(KERN_DEBUG "vm_start = 0x%08lx, vm_end = 0x%08lx\n",
+ vma->vm_start, vma->vm_end);
+@@ -261,7 +262,8 @@ no_context:
+
+ out_of_memory:
+ up_read(&mm->mmap_sem);
+- printk(KERN_CRIT "VM: killing process %s\n", current->comm);
++ printk(KERN_CRIT "VM: killing process %s(%d:#%u)\n",
++ current->comm, current->pid, current->xid);
+ if (user_mode(regs))
+ do_group_exit(SIGKILL);
+ goto no_context;
+--- a/arch/powerpc/Kconfig 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/powerpc/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -800,6 +800,8 @@ source "lib/Kconfig"
+
+ source "arch/powerpc/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ config KEYS_COMPAT
+--- a/arch/powerpc/kernel/irq.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/powerpc/kernel/irq.c 2008-07-16 22:41:36.000000000 -0400
+@@ -53,6 +53,7 @@
+ #include <linux/bootmem.h>
+ #include <linux/pci.h>
+ #include <linux/debugfs.h>
++#include <linux/vs_context.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/system.h>
+--- a/arch/powerpc/kernel/process.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/powerpc/kernel/process.c 2008-07-16 22:41:36.000000000 -0400
+@@ -464,8 +464,9 @@ void show_regs(struct pt_regs * regs)
+ #else
+ printk("DAR: "REG", DSISR: "REG"\n", regs->dar, regs->dsisr);
+ #endif
+- printk("TASK = %p[%d] '%s' THREAD: %p",
+- current, task_pid_nr(current), current->comm, task_thread_info(current));
++ printk("TASK = %p[%d,#%u] '%s' THREAD: %p",
++ current, task_pid_nr(current), current->xid,
++ current->comm, task_thread_info(current));
+
+ #ifdef CONFIG_SMP
+ printk(" CPU: %d", raw_smp_processor_id());
+--- a/arch/powerpc/kernel/sys_ppc32.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/powerpc/kernel/sys_ppc32.c 2008-07-16 22:41:36.000000000 -0400
+@@ -204,7 +204,7 @@ asmlinkage long compat_sys_gettimeofday(
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/powerpc/kernel/traps.c 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/powerpc/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -941,8 +941,9 @@ void nonrecoverable_exception(struct pt_
+
+ void trace_syscall(struct pt_regs *regs)
+ {
+- printk("Task: %p(%d), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n",
+- current, task_pid_nr(current), regs->nip, regs->link, regs->gpr[0],
++ printk("Task: %p(%d[#%u]), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n",
++ current, task_pid_nr(current), current->xid,
++ regs->nip, regs->link, regs->gpr[0],
+ regs->ccr&0x10000000?"Error=":"", regs->gpr[3], print_tainted());
+ }
+
+--- a/arch/powerpc/kernel/vdso.c 2008-07-14 17:22:28.000000000 -0400
++++ a/arch/powerpc/kernel/vdso.c 2008-07-17 21:09:13.000000000 -0400
+@@ -22,6 +22,7 @@
+ #include <linux/security.h>
+ #include <linux/bootmem.h>
+ #include <linux/lmb.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/system.h>
+--- a/arch/powerpc/mm/fault.c 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/powerpc/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -378,7 +378,8 @@ out_of_memory:
+ down_read(&mm->mmap_sem);
+ goto survive;
+ }
+- printk("VM: killing process %s\n", current->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ current->comm, current->pid, current->xid);
+ if (user_mode(regs))
+ do_group_exit(SIGKILL);
+ return SIGKILL;
+--- a/arch/ppc/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/ppc/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -1181,6 +1181,8 @@ source "lib/Kconfig"
+
+ source "arch/ppc/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/ppc/kernel/traps.c 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/ppc/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -669,8 +669,9 @@ void nonrecoverable_exception(struct pt_
+
+ void trace_syscall(struct pt_regs *regs)
+ {
+- printk("Task: %p(%d), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n",
+- current, current->pid, regs->nip, regs->link, regs->gpr[0],
++ printk("Task: %p(%d[#%u]), PC: %08lX/%08lX, Syscall: %3ld, Result: %s%ld %s\n",
++ current, current->pid, current->xid,
++ regs->nip, regs->link, regs->gpr[0],
+ regs->ccr&0x10000000?"Error=":"", regs->gpr[3], print_tainted());
+ }
+
+--- a/arch/ppc/mm/fault.c 2008-04-17 11:31:25.000000000 -0400
++++ a/arch/ppc/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -295,7 +295,8 @@ out_of_memory:
+ down_read(&mm->mmap_sem);
+ goto survive;
+ }
+- printk("VM: killing process %s\n", current->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ current->comm, current->pid, current->xid);
+ if (user_mode(regs))
+ do_group_exit(SIGKILL);
+ return SIGKILL;
+--- a/arch/s390/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/s390/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -562,6 +562,8 @@ source "fs/Kconfig"
+
+ source "arch/s390/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/s390/kernel/compat_linux.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/s390/kernel/compat_linux.c 2008-07-16 22:41:36.000000000 -0400
+@@ -566,7 +566,7 @@ asmlinkage long sys32_gettimeofday(struc
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/s390/kernel/ptrace.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/s390/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -33,6 +33,7 @@
+ #include <linux/security.h>
+ #include <linux/audit.h>
+ #include <linux/signal.h>
++#include <linux/vs_base.h>
+
+ #include <asm/segment.h>
+ #include <asm/page.h>
+--- a/arch/s390/kernel/syscalls.S 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/s390/kernel/syscalls.S 2008-07-16 22:41:36.000000000 -0400
+@@ -271,7 +271,7 @@ SYSCALL(sys_clock_settime,sys_clock_sett
+ SYSCALL(sys_clock_gettime,sys_clock_gettime,sys32_clock_gettime_wrapper) /* 260 */
+ SYSCALL(sys_clock_getres,sys_clock_getres,sys32_clock_getres_wrapper)
+ SYSCALL(sys_clock_nanosleep,sys_clock_nanosleep,sys32_clock_nanosleep_wrapper)
+-NI_SYSCALL /* reserved for vserver */
++SYSCALL(sys_vserver,sys_vserver,sys32_vserver)
+ SYSCALL(s390_fadvise64_64,sys_ni_syscall,sys32_fadvise64_64_wrapper)
+ SYSCALL(sys_statfs64,sys_statfs64,compat_sys_statfs64_wrapper)
+ SYSCALL(sys_fstatfs64,sys_fstatfs64,compat_sys_fstatfs64_wrapper)
+--- a/arch/s390/mm/fault.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/s390/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -216,7 +216,8 @@ static int do_out_of_memory(struct pt_re
+ down_read(&mm->mmap_sem);
+ return 1;
+ }
+- printk("VM: killing process %s\n", tsk->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ tsk->comm, tsk->pid, tsk->xid);
+ if (regs->psw.mask & PSW_MASK_PSTATE)
+ do_group_exit(SIGKILL);
+ do_no_context(regs, error_code, address);
+--- a/arch/sh/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sh/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -927,6 +927,8 @@ source "fs/Kconfig"
+
+ source "arch/sh/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/sh/kernel/irq.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sh/kernel/irq.c 2008-07-16 22:41:36.000000000 -0400
+@@ -11,6 +11,7 @@
+ #include <linux/module.h>
+ #include <linux/kernel_stat.h>
+ #include <linux/seq_file.h>
++#include <linux/vs_context.h>
+ #include <asm/processor.h>
+ #include <asm/machvec.h>
+ #include <asm/uaccess.h>
+--- a/arch/sh/kernel/vsyscall/vsyscall.c 2008-04-17 10:37:14.000000000 -0400
++++ a/arch/sh/kernel/vsyscall/vsyscall.c 2008-07-16 22:41:36.000000000 -0400
+@@ -19,6 +19,7 @@
+ #include <linux/elf.h>
+ #include <linux/sched.h>
+ #include <linux/err.h>
++#include <linux/vs_memory.h>
+
+ /*
+ * Should the kernel map a VDSO page into processes and pass its
+--- a/arch/sparc/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -318,6 +318,8 @@ source "fs/Kconfig"
+
+ source "arch/sparc/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/sparc/kernel/ptrace.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/signal.h>
+ #include <linux/regset.h>
+ #include <linux/elf.h>
++#include <linux/vs_base.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/system.h>
+--- a/arch/sparc/kernel/systbls.S 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc/kernel/systbls.S 2008-07-16 22:41:36.000000000 -0400
+@@ -70,7 +70,7 @@ sys_call_table:
+ /*250*/ .long sparc_mremap, sys_sysctl, sys_getsid, sys_fdatasync, sys_nfsservctl
+ /*255*/ .long sys_sync_file_range, sys_clock_settime, sys_clock_gettime, sys_clock_getres, sys_clock_nanosleep
+ /*260*/ .long sys_sched_getaffinity, sys_sched_setaffinity, sys_timer_settime, sys_timer_gettime, sys_timer_getoverrun
+-/*265*/ .long sys_timer_delete, sys_timer_create, sys_nis_syscall, sys_io_setup, sys_io_destroy
++/*265*/ .long sys_timer_delete, sys_timer_create, sys_vserver, sys_io_setup, sys_io_destroy
+ /*270*/ .long sys_io_submit, sys_io_cancel, sys_io_getevents, sys_mq_open, sys_mq_unlink
+ /*275*/ .long sys_mq_timedsend, sys_mq_timedreceive, sys_mq_notify, sys_mq_getsetattr, sys_waitid
+ /*280*/ .long sys_tee, sys_add_key, sys_request_key, sys_keyctl, sys_openat
+--- a/arch/sparc/kernel/traps.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -99,7 +99,8 @@ void die_if_kernel(char *str, struct pt_
+ " /_| \\__/ |_\\\n"
+ " \\__U_/\n");
+
+- printk("%s(%d): %s [#%d]\n", current->comm, task_pid_nr(current), str, ++die_counter);
++ printk("%s(%d[#%u]): %s [#%d]\n", current->comm,
++ task_pid_nr(current), current->xid, str, ++die_counter);
+ show_regs(regs);
+ add_taint(TAINT_DIE);
+
+--- a/arch/sparc/mm/fault.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -318,7 +318,8 @@ no_context:
+ */
+ out_of_memory:
+ up_read(&mm->mmap_sem);
+- printk("VM: killing process %s\n", tsk->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ tsk->comm, tsk->pid, tsk->xid);
+ if (from_user)
+ do_group_exit(SIGKILL);
+ goto no_context;
+--- a/arch/sparc64/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -407,6 +407,8 @@ source "fs/Kconfig"
+
+ source "arch/sparc64/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/sparc64/kernel/ptrace.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -25,6 +25,7 @@
+ #include <linux/regset.h>
+ #include <linux/compat.h>
+ #include <linux/elf.h>
++#include <linux/vs_base.h>
+
+ #include <asm/asi.h>
+ #include <asm/pgtable.h>
+--- a/arch/sparc64/kernel/sys_sparc32.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/kernel/sys_sparc32.c 2008-07-16 22:41:36.000000000 -0400
+@@ -707,7 +707,7 @@ asmlinkage long sys32_gettimeofday(struc
+ {
+ if (tv) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (put_tv32(tv, &ktv))
+ return -EFAULT;
+ }
+--- a/arch/sparc64/kernel/systbls.S 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/kernel/systbls.S 2008-07-16 22:41:36.000000000 -0400
+@@ -71,7 +71,7 @@ sys_call_table32:
+ /*250*/ .word sys32_mremap, sys32_sysctl, sys32_getsid, sys_fdatasync, sys32_nfsservctl
+ .word sys32_sync_file_range, compat_sys_clock_settime, compat_sys_clock_gettime, compat_sys_clock_getres, sys32_clock_nanosleep
+ /*260*/ .word compat_sys_sched_getaffinity, compat_sys_sched_setaffinity, sys32_timer_settime, compat_sys_timer_gettime, sys_timer_getoverrun
+- .word sys_timer_delete, compat_sys_timer_create, sys_ni_syscall, compat_sys_io_setup, sys_io_destroy
++ .word sys_timer_delete, compat_sys_timer_create, sys32_vserver, compat_sys_io_setup, sys_io_destroy
+ /*270*/ .word sys32_io_submit, sys_io_cancel, compat_sys_io_getevents, sys32_mq_open, sys_mq_unlink
+ .word compat_sys_mq_timedsend, compat_sys_mq_timedreceive, compat_sys_mq_notify, compat_sys_mq_getsetattr, compat_sys_waitid
+ /*280*/ .word sys32_tee, sys_add_key, sys_request_key, sys_keyctl, compat_sys_openat
+@@ -144,7 +144,7 @@ sys_call_table:
+ /*250*/ .word sys64_mremap, sys_sysctl, sys_getsid, sys_fdatasync, sys_nfsservctl
+ .word sys_sync_file_range, sys_clock_settime, sys_clock_gettime, sys_clock_getres, sys_clock_nanosleep
+ /*260*/ .word sys_sched_getaffinity, sys_sched_setaffinity, sys_timer_settime, sys_timer_gettime, sys_timer_getoverrun
+- .word sys_timer_delete, sys_timer_create, sys_ni_syscall, sys_io_setup, sys_io_destroy
++ .word sys_timer_delete, sys_timer_create, sys_vserver, sys_io_setup, sys_io_destroy
+ /*270*/ .word sys_io_submit, sys_io_cancel, sys_io_getevents, sys_mq_open, sys_mq_unlink
+ .word sys_mq_timedsend, sys_mq_timedreceive, sys_mq_notify, sys_mq_getsetattr, sys_waitid
+ /*280*/ .word sys_tee, sys_add_key, sys_request_key, sys_keyctl, sys_openat
+--- a/arch/sparc64/kernel/traps.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/kernel/traps.c 2008-07-16 22:41:36.000000000 -0400
+@@ -2196,7 +2196,8 @@ void die_if_kernel(char *str, struct pt_
+ " /_| \\__/ |_\\\n"
+ " \\__U_/\n");
+
+- printk("%s(%d): %s [#%d]\n", current->comm, task_pid_nr(current), str, ++die_counter);
++ printk("%s(%d[#%u]): %s [#%d]\n", current->comm,
++ task_pid_nr(current), current->xid, str, ++die_counter);
+ notify_die(DIE_OOPS, str, regs, 0, 255, SIGSEGV);
+ __asm__ __volatile__("flushw");
+ __show_regs(regs);
+--- a/arch/sparc64/mm/fault.c 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/sparc64/mm/fault.c 2008-07-16 22:41:36.000000000 -0400
+@@ -453,7 +453,8 @@ handle_kernel_fault:
+ out_of_memory:
+ insn = get_fault_insn(regs, insn);
+ up_read(&mm->mmap_sem);
+- printk("VM: killing process %s\n", current->comm);
++ printk("VM: killing process %s(%d:#%u)\n",
++ current->comm, current->pid, current->xid);
+ if (!(regs->tstate & TSTATE_PRIV))
+ do_group_exit(SIGKILL);
+ goto handle_kernel_fault;
+--- a/arch/um/Kconfig 2008-07-14 17:22:29.000000000 -0400
++++ a/arch/um/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -245,6 +245,8 @@ source "drivers/connector/Kconfig"
+
+ source "fs/Kconfig"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/um/kernel/trap.c 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/um/kernel/trap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -215,7 +215,8 @@ unsigned long segv(struct faultinfo fi,
+ current->thread.arch.faultinfo = fi;
+ force_sig_info(SIGBUS, &si, current);
+ } else if (err == -ENOMEM) {
+- printk(KERN_INFO "VM: killing process %s\n", current->comm);
++ printk(KERN_INFO "VM: killing process %s(%d:#%u)\n",
++ current->comm, task_pid_nr(current), current->xid);
+ do_exit(SIGKILL);
+ } else {
+ BUG_ON(err != -EFAULT);
+--- a/arch/v850/Kconfig 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/v850/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -344,6 +344,8 @@ source "drivers/usb/Kconfig"
+
+ source "arch/v850/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/v850/kernel/ptrace.c 2008-04-17 10:37:14.000000000 -0400
++++ a/arch/v850/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -23,6 +23,7 @@
+ #include <linux/sched.h>
+ #include <linux/ptrace.h>
+ #include <linux/signal.h>
++#include <linux/vs_base.h>
+
+ #include <asm/errno.h>
+ #include <asm/ptrace.h>
+@@ -116,6 +117,9 @@ long arch_ptrace(struct task_struct *chi
+ {
+ int rval;
+
++ if (!vx_check(vx_task_xid(child), VS_WATCH_P | VS_IDENT))
++ goto out;
++
+ switch (request) {
+ unsigned long val;
+
+--- a/arch/x86/ia32/ia32entry.S 2008-07-14 17:22:30.000000000 -0400
++++ a/arch/x86/ia32/ia32entry.S 2008-07-16 22:41:36.000000000 -0400
+@@ -677,7 +677,7 @@ ia32_sys_call_table:
+ .quad sys_tgkill /* 270 */
+ .quad compat_sys_utimes
+ .quad sys32_fadvise64_64
+- .quad quiet_ni_syscall /* sys_vserver */
++ .quad sys32_vserver
+ .quad sys_mbind
+ .quad compat_sys_get_mempolicy /* 275 */
+ .quad sys_set_mempolicy
+--- a/arch/x86/Kconfig 2008-07-14 17:22:30.000000000 -0400
++++ a/arch/x86/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -1722,6 +1722,8 @@ source "fs/Kconfig"
+
+ source "arch/x86/Kconfig.debug"
+
++source "kernel/vserver/Kconfig"
++
+ source "security/Kconfig"
+
+ source "crypto/Kconfig"
+--- a/arch/x86/kernel/syscall_table_32.S 2008-04-17 12:05:30.000000000 -0400
++++ a/arch/x86/kernel/syscall_table_32.S 2008-07-16 22:41:36.000000000 -0400
+@@ -272,7 +272,7 @@ ENTRY(sys_call_table)
+ .long sys_tgkill /* 270 */
+ .long sys_utimes
+ .long sys_fadvise64_64
+- .long sys_ni_syscall /* sys_vserver */
++ .long sys_vserver
+ .long sys_mbind
+ .long sys_get_mempolicy
+ .long sys_set_mempolicy
+--- a/Documentation/vserver/debug.txt 1969-12-31 19:00:00.000000000 -0500
++++ a/Documentation/vserver/debug.txt 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,154 @@
++
++debug_cvirt:
++
++ 2 4 "vx_map_tgid: %p/%llx: %d -> %d"
++ "vx_rmap_tgid: %p/%llx: %d -> %d"
++
++debug_dlim:
++
++ 0 1 "ALLOC (%p,#%d)%c inode (%d)"
++ "FREE (%p,#%d)%c inode"
++ 1 2 "ALLOC (%p,#%d)%c %lld bytes (%d)"
++ "FREE (%p,#%d)%c %lld bytes"
++ 2 4 "ADJUST: %lld,%lld on %ld,%ld [mult=%d]"
++ 3 8 "ext3_has_free_blocks(%p): %lu<%lu+1, %c, %u!=%u r=%d"
++ "ext3_has_free_blocks(%p): free=%lu, root=%lu"
++ "rcu_free_dl_info(%p)"
++ 4 10 "alloc_dl_info(%p,%d) = %p"
++ "dealloc_dl_info(%p)"
++ "get_dl_info(%p[#%d.%d])"
++ "put_dl_info(%p[#%d.%d])"
++ 5 20 "alloc_dl_info(%p,%d)*"
++ 6 40 "__hash_dl_info: %p[#%d]"
++ "__unhash_dl_info: %p[#%d]"
++ 7 80 "locate_dl_info(%p,#%d) = %p"
++
++debug_misc:
++
++ 0 1 "destroy_dqhash: %p [#0x%08x] c=%d"
++ "new_dqhash: %p [#0x%08x]"
++ "vroot[%d]_clr_dev: dev=%p[%lu,%d:%d]"
++ "vroot[%d]_get_real_bdev: dev=%p[%lu,%d:%d]"
++ "vroot[%d]_set_dev: dev=%p[%lu,%d:%d]"
++ "vroot_get_real_bdev not set"
++ 1 2 "cow_break_link(»%s«)"
++ "temp copy »%s«"
++ 2 4 "dentry_open(new): %p"
++ "dentry_open(old): %p"
++ "lookup_create(new): %p"
++ "old path »%s«"
++ "path_lookup(old): %d"
++ "vfs_create(new): %d"
++ "vfs_rename: %d"
++ "vfs_sendfile: %d"
++ 3 8 "fput(new_file=%p[#%d])"
++ "fput(old_file=%p[#%d])"
++ 4 10 "vx_info_kill(%p[#%d],%d,%d) = %d"
++ "vx_info_kill(%p[#%d],%d,%d)*"
++ 5 20 "vs_reboot(%p[#%d],%d)"
++ 6 40 "dropping task %p[#%u,%u] for %p[#%u,%u]"
++
++debug_net:
++
++ 2 4 "nx_addr_conflict(%p,%p) %d.%d,%d.%d"
++ 3 8 "inet_bind(%p) %d.%d.%d.%d, %d.%d.%d.%d, %d.%d.%d.%d"
++ "inet_bind(%p)* %p,%p;%lx %d.%d.%d.%d"
++ 4 10 "ip_route_connect(%p) %p,%p;%lx"
++ 5 20 "__addr_in_socket(%p,%d.%d.%d.%d) %p:%d.%d.%d.%d %p;%lx"
++ 6 40 "sk,egf: %p [#%d] (from %d)"
++ "sk,egn: %p [#%d] (from %d)"
++ "sk,req: %p [#%d] (from %d)"
++ "sk: %p [#%d] (from %d)"
++ "tw: %p [#%d] (from %d)"
++ 7 80 "__sock_recvmsg: %p[%p,%p,%p;%d]:%d/%d"
++ "__sock_sendmsg: %p[%p,%p,%p;%d]:%d/%d"
++
++debug_nid:
++
++ 0 1 "__lookup_nx_info(#%u): %p[#%u]"
++ "alloc_nx_info(%d) = %p"
++ "create_nx_info(%d) (dynamic rejected)"
++ "create_nx_info(%d) = %p (already there)"
++ "create_nx_info(%d) = %p (new)"
++ "dealloc_nx_info(%p)"
++ 1 2 "alloc_nx_info(%d)*"
++ "create_nx_info(%d)*"
++ 2 4 "get_nx_info(%p[#%d.%d])"
++ "put_nx_info(%p[#%d.%d])"
++ 3 8 "claim_nx_info(%p[#%d.%d.%d]) %p"
++ "clr_nx_info(%p[#%d.%d])"
++ "init_nx_info(%p[#%d.%d])"
++ "release_nx_info(%p[#%d.%d.%d]) %p"
++ "set_nx_info(%p[#%d.%d])"
++ 4 10 "__hash_nx_info: %p[#%d]"
++ "__nx_dynamic_id: [#%d]"
++ "__unhash_nx_info: %p[#%d.%d.%d]"
++ 5 20 "moved task %p into nxi:%p[#%d]"
++ "nx_migrate_task(%p,%p[#%d.%d.%d])"
++ "task_get_nx_info(%p)"
++ 6 40 "nx_clear_persistent(%p[#%d])"
++
++debug_quota:
++
++ 0 1 "quota_sync_dqh(%p,%d) discard inode %p"
++ 1 2 "quota_sync_dqh(%p,%d)"
++ "sync_dquots(%p,%d)"
++ "sync_dquots_dqh(%p,%d)"
++ 3 8 "do_quotactl(%p,%d,cmd=%d,id=%d,%p)"
++
++debug_switch:
++
++ 0 1 "vc: VCMD_%02d_%d[%d], %d,%p [%d,%d,%x,%x]"
++ 1 2 "vc: VCMD_%02d_%d[%d] = %08lx(%ld) [%d,%d]"
++ 4 10 "%s: (%s %s) returned %s with %d"
++
++debug_tag:
++
++ 7 80 "dx_parse_tag(»%s«): %d:#%d"
++ "dx_propagate_tag(%p[#%lu.%d]): %d,%d"
++
++debug_xid:
++
++ 0 1 "__lookup_vx_info(#%u): %p[#%u]"
++ "alloc_vx_info(%d) = %p"
++ "alloc_vx_info(%d)*"
++ "create_vx_info(%d) (dynamic rejected)"
++ "create_vx_info(%d) = %p (already there)"
++ "create_vx_info(%d) = %p (new)"
++ "dealloc_vx_info(%p)"
++ "loc_vx_info(%d) = %p (found)"
++ "loc_vx_info(%d) = %p (new)"
++ "loc_vx_info(%d) = %p (not available)"
++ 1 2 "create_vx_info(%d)*"
++ "loc_vx_info(%d)*"
++ 2 4 "get_vx_info(%p[#%d.%d])"
++ "put_vx_info(%p[#%d.%d])"
++ 3 8 "claim_vx_info(%p[#%d.%d.%d]) %p"
++ "clr_vx_info(%p[#%d.%d])"
++ "init_vx_info(%p[#%d.%d])"
++ "release_vx_info(%p[#%d.%d.%d]) %p"
++ "set_vx_info(%p[#%d.%d])"
++ 4 10 "__hash_vx_info: %p[#%d]"
++ "__unhash_vx_info: %p[#%d.%d.%d]"
++ "__vx_dynamic_id: [#%d]"
++ 5 20 "enter_vx_info(%p[#%d],%p) %p[#%d,%p]"
++ "leave_vx_info(%p[#%d,%p]) %p[#%d,%p]"
++ "moved task %p into vxi:%p[#%d]"
++ "task_get_vx_info(%p)"
++ "vx_migrate_task(%p,%p[#%d.%d])"
++ 6 40 "vx_clear_persistent(%p[#%d])"
++ "vx_exit_init(%p[#%d],%p[#%d,%d,%d])"
++ "vx_set_init(%p[#%d],%p[#%d,%d,%d])"
++ "vx_set_persistent(%p[#%d])"
++ "vx_set_reaper(%p[#%d],%p[#%d,%d])"
++ 7 80 "vx_child_reaper(%p[#%u,%u]) = %p[#%u,%u]"
++
++
++debug_limit:
++
++ n 2^n "vx_acc_cres[%5d,%s,%2d]: %5d%s"
++ "vx_cres_avail[%5d,%s,%2d]: %5ld > %5d + %5d"
++
++ m 2^m "vx_acc_page[%5d,%s,%2d]: %5d%s"
++ "vx_acc_pages[%5d,%s,%2d]: %5d += %5d"
++ "vx_pages_avail[%5d,%s,%2d]: %5ld > %5d + %5d"
+--- a/drivers/block/Kconfig 2008-04-17 12:05:31.000000000 -0400
++++ a/drivers/block/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -263,6 +263,13 @@ config BLK_DEV_CRYPTOLOOP
+ instead, which can be configured to be on-disk compatible with the
+ cryptoloop device.
+
++config BLK_DEV_VROOT
++ tristate "Virtual Root device support"
++ depends on QUOTACTL
++ ---help---
++ Saying Y here will allow you to use quota/fs ioctls on a shared
++ partition within a virtual server without compromising security.
++
+ config BLK_DEV_NBD
+ tristate "Network block device support"
+ depends on NET
+--- a/drivers/block/loop.c 2008-07-14 17:22:32.000000000 -0400
++++ a/drivers/block/loop.c 2008-07-16 22:41:36.000000000 -0400
+@@ -76,6 +76,7 @@
+ #include <linux/gfp.h>
+ #include <linux/kthread.h>
+ #include <linux/splice.h>
++#include <linux/vs_context.h>
+
+ #include <asm/uaccess.h>
+
+@@ -794,6 +795,7 @@ static int loop_set_fd(struct loop_devic
+ lo->lo_blocksize = lo_blocksize;
+ lo->lo_device = bdev;
+ lo->lo_flags = lo_flags;
++ lo->lo_xid = vx_current_xid();
+ lo->lo_backing_file = file;
+ lo->transfer = transfer_none;
+ lo->ioctl = NULL;
+@@ -915,6 +917,7 @@ static int loop_clr_fd(struct loop_devic
+ lo->lo_encrypt_key_size = 0;
+ lo->lo_flags = 0;
+ lo->lo_thread = NULL;
++ lo->lo_xid = 0;
+ memset(lo->lo_encrypt_key, 0, LO_KEY_SIZE);
+ memset(lo->lo_crypt_name, 0, LO_NAME_SIZE);
+ memset(lo->lo_file_name, 0, LO_NAME_SIZE);
+@@ -938,7 +941,7 @@ loop_set_status(struct loop_device *lo,
+ struct loop_func_table *xfer;
+
+ if (lo->lo_encrypt_key_size && lo->lo_key_owner != current->uid &&
+- !capable(CAP_SYS_ADMIN))
++ !vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_CLOOP))
+ return -EPERM;
+ if (lo->lo_state != Lo_bound)
+ return -ENXIO;
+@@ -1022,7 +1025,8 @@ loop_get_status(struct loop_device *lo,
+ memcpy(info->lo_crypt_name, lo->lo_crypt_name, LO_NAME_SIZE);
+ info->lo_encrypt_type =
+ lo->lo_encryption ? lo->lo_encryption->number : 0;
+- if (lo->lo_encrypt_key_size && capable(CAP_SYS_ADMIN)) {
++ if (lo->lo_encrypt_key_size &&
++ vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_CLOOP)) {
+ info->lo_encrypt_key_size = lo->lo_encrypt_key_size;
+ memcpy(info->lo_encrypt_key, lo->lo_encrypt_key,
+ lo->lo_encrypt_key_size);
+@@ -1331,6 +1335,9 @@ static int lo_open(struct inode *inode,
+ {
+ struct loop_device *lo = inode->i_bdev->bd_disk->private_data;
+
++ if (!vx_check(lo->lo_xid, VS_IDENT|VS_HOSTID|VS_ADMIN_P))
++ return -EACCES;
++
+ mutex_lock(&lo->lo_ctl_mutex);
+ lo->lo_refcnt++;
+ mutex_unlock(&lo->lo_ctl_mutex);
+--- a/drivers/block/Makefile 2008-04-17 12:05:31.000000000 -0400
++++ a/drivers/block/Makefile 2008-07-16 22:41:36.000000000 -0400
+@@ -29,5 +29,6 @@ obj-$(CONFIG_VIRTIO_BLK) += virtio_blk.o
+ obj-$(CONFIG_VIODASD) += viodasd.o
+ obj-$(CONFIG_BLK_DEV_SX8) += sx8.o
+ obj-$(CONFIG_BLK_DEV_UB) += ub.o
++obj-$(CONFIG_BLK_DEV_VROOT) += vroot.o
+
+ obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += xen-blkfront.o
+--- a/drivers/block/vroot.c 1969-12-31 19:00:00.000000000 -0500
++++ a/drivers/block/vroot.c 2008-08-11 23:21:07.000000000 -0400
+@@ -0,0 +1,283 @@
++/*
++ * linux/drivers/block/vroot.c
++ *
++ * written by Herbert Pötzl, 9/11/2002
++ * ported to 2.6.10 by Herbert Pötzl, 30/12/2004
++ *
++ * based on the loop.c code by Theodore Ts'o.
++ *
++ * Copyright (C) 2002-2007 by Herbert Pötzl.
++ * Redistribution of this file is permitted under the
++ * GNU General Public License.
++ *
++ */
++
++#include <linux/module.h>
++#include <linux/moduleparam.h>
++#include <linux/file.h>
++#include <linux/major.h>
++#include <linux/blkdev.h>
++
++#include <linux/vroot.h>
++#include <linux/vs_context.h>
++
++
++static int max_vroot = 8;
++
++static struct vroot_device *vroot_dev;
++static struct gendisk **disks;
++
++
++static int vroot_set_dev(
++ struct vroot_device *vr,
++ struct file *vr_file,
++ struct block_device *bdev,
++ unsigned int arg)
++{
++ struct block_device *real_bdev;
++ struct file *file;
++ struct inode *inode;
++ int error;
++
++ error = -EBUSY;
++ if (vr->vr_state != Vr_unbound)
++ goto out;
++
++ error = -EBADF;
++ file = fget(arg);
++ if (!file)
++ goto out;
++
++ error = -EINVAL;
++ inode = file->f_dentry->d_inode;
++
++
++ if (S_ISBLK(inode->i_mode)) {
++ real_bdev = inode->i_bdev;
++ vr->vr_device = real_bdev;
++ __iget(real_bdev->bd_inode);
++ } else
++ goto out_fput;
++
++ vxdprintk(VXD_CBIT(misc, 0),
++ "vroot[%d]_set_dev: dev=" VXF_DEV,
++ vr->vr_number, VXD_DEV(real_bdev));
++
++ vr->vr_state = Vr_bound;
++ error = 0;
++
++ out_fput:
++ fput(file);
++ out:
++ return error;
++}
++
++static int vroot_clr_dev(
++ struct vroot_device *vr,
++ struct file *vr_file,
++ struct block_device *bdev)
++{
++ struct block_device *real_bdev;
++
++ if (vr->vr_state != Vr_bound)
++ return -ENXIO;
++ if (vr->vr_refcnt > 1) /* we needed one fd for the ioctl */
++ return -EBUSY;
++
++ real_bdev = vr->vr_device;
++
++ vxdprintk(VXD_CBIT(misc, 0),
++ "vroot[%d]_clr_dev: dev=" VXF_DEV,
++ vr->vr_number, VXD_DEV(real_bdev));
++
++ bdput(real_bdev);
++ vr->vr_state = Vr_unbound;
++ vr->vr_device = NULL;
++ return 0;
++}
++
++
++static int vr_ioctl(struct inode *inode, struct file *file,
++ unsigned int cmd, unsigned long arg)
++{
++ struct vroot_device *vr = inode->i_bdev->bd_disk->private_data;
++ int err;
++
++ down(&vr->vr_ctl_mutex);
++ switch (cmd) {
++ case VROOT_SET_DEV:
++ err = vroot_set_dev(vr, file, inode->i_bdev, arg);
++ break;
++ case VROOT_CLR_DEV:
++ err = vroot_clr_dev(vr, file, inode->i_bdev);
++ break;
++ default:
++ err = -EINVAL;
++ break;
++ }
++ up(&vr->vr_ctl_mutex);
++ return err;
++}
++
++static int vr_open(struct inode *inode, struct file *file)
++{
++ struct vroot_device *vr = inode->i_bdev->bd_disk->private_data;
++
++ down(&vr->vr_ctl_mutex);
++ vr->vr_refcnt++;
++ up(&vr->vr_ctl_mutex);
++ return 0;
++}
++
++static int vr_release(struct inode *inode, struct file *file)
++{
++ struct vroot_device *vr = inode->i_bdev->bd_disk->private_data;
++
++ down(&vr->vr_ctl_mutex);
++ --vr->vr_refcnt;
++ up(&vr->vr_ctl_mutex);
++ return 0;
++}
++
++static struct block_device_operations vr_fops = {
++ .owner = THIS_MODULE,
++ .open = vr_open,
++ .release = vr_release,
++ .ioctl = vr_ioctl,
++};
++
++struct block_device *__vroot_get_real_bdev(struct block_device *bdev)
++{
++ struct inode *inode = bdev->bd_inode;
++ struct vroot_device *vr;
++ struct block_device *real_bdev;
++ int minor = iminor(inode);
++
++ vr = &vroot_dev[minor];
++ real_bdev = vr->vr_device;
++
++ vxdprintk(VXD_CBIT(misc, 0),
++ "vroot[%d]_get_real_bdev: dev=" VXF_DEV,
++ vr->vr_number, VXD_DEV(real_bdev));
++
++ if (vr->vr_state != Vr_bound)
++ return ERR_PTR(-ENXIO);
++
++ __iget(real_bdev->bd_inode);
++ return real_bdev;
++}
++
++/*
++ * And now the modules code and kernel interface.
++ */
++
++module_param(max_vroot, int, 0);
++
++MODULE_PARM_DESC(max_vroot, "Maximum number of vroot devices (1-256)");
++MODULE_LICENSE("GPL");
++MODULE_ALIAS_BLOCKDEV_MAJOR(VROOT_MAJOR);
++
++MODULE_AUTHOR ("Herbert Pötzl");
++MODULE_DESCRIPTION ("Virtual Root Device Mapper");
++
++
++int __init vroot_init(void)
++{
++ int err, i;
++
++ if (max_vroot < 1 || max_vroot > 256) {
++ max_vroot = MAX_VROOT_DEFAULT;
++ printk(KERN_WARNING "vroot: invalid max_vroot "
++ "(must be between 1 and 256), "
++ "using default (%d)\n", max_vroot);
++ }
++
++ if (register_blkdev(VROOT_MAJOR, "vroot"))
++ return -EIO;
++
++ err = -ENOMEM;
++ vroot_dev = kmalloc(max_vroot * sizeof(struct vroot_device), GFP_KERNEL);
++ if (!vroot_dev)
++ goto out_mem1;
++ memset(vroot_dev, 0, max_vroot * sizeof(struct vroot_device));
++
++ disks = kmalloc(max_vroot * sizeof(struct gendisk *), GFP_KERNEL);
++ if (!disks)
++ goto out_mem2;
++
++ for (i = 0; i < max_vroot; i++) {
++ disks[i] = alloc_disk(1);
++ if (!disks[i])
++ goto out_mem3;
++ disks[i]->queue = blk_alloc_queue(GFP_KERNEL);
++ if (!disks[i]->queue)
++ goto out_mem3;
++ }
++
++ for (i = 0; i < max_vroot; i++) {
++ struct vroot_device *vr = &vroot_dev[i];
++ struct gendisk *disk = disks[i];
++
++ memset(vr, 0, sizeof(*vr));
++ init_MUTEX(&vr->vr_ctl_mutex);
++ vr->vr_number = i;
++ disk->major = VROOT_MAJOR;
++ disk->first_minor = i;
++ disk->fops = &vr_fops;
++ sprintf(disk->disk_name, "vroot%d", i);
++ disk->private_data = vr;
++ }
++
++ err = register_vroot_grb(&__vroot_get_real_bdev);
++ if (err)
++ goto out_mem3;
++
++ for (i = 0; i < max_vroot; i++)
++ add_disk(disks[i]);
++ printk(KERN_INFO "vroot: loaded (max %d devices)\n", max_vroot);
++ return 0;
++
++out_mem3:
++ while (i--)
++ put_disk(disks[i]);
++ kfree(disks);
++out_mem2:
++ kfree(vroot_dev);
++out_mem1:
++ unregister_blkdev(VROOT_MAJOR, "vroot");
++ printk(KERN_ERR "vroot: ran out of memory\n");
++ return err;
++}
++
++void vroot_exit(void)
++{
++ int i;
++
++ if (unregister_vroot_grb(&__vroot_get_real_bdev))
++ printk(KERN_WARNING "vroot: cannot unregister grb\n");
++
++ for (i = 0; i < max_vroot; i++) {
++ del_gendisk(disks[i]);
++ put_disk(disks[i]);
++ }
++ unregister_blkdev(VROOT_MAJOR, "vroot");
++
++ kfree(disks);
++ kfree(vroot_dev);
++}
++
++module_init(vroot_init);
++module_exit(vroot_exit);
++
++#ifndef MODULE
++
++static int __init max_vroot_setup(char *str)
++{
++ max_vroot = simple_strtol(str, NULL, 0);
++ return 1;
++}
++
++__setup("max_vroot=", max_vroot_setup);
++
++#endif
++
+--- a/drivers/char/sysrq.c 2008-07-14 17:22:33.000000000 -0400
++++ a/drivers/char/sysrq.c 2008-07-16 22:41:36.000000000 -0400
+@@ -37,6 +37,7 @@
+ #include <linux/irq.h>
+ #include <linux/hrtimer.h>
+ #include <linux/oom.h>
++#include <linux/vserver/debug.h>
+
+ #include <asm/ptrace.h>
+ #include <asm/irq_regs.h>
+@@ -351,6 +352,21 @@ static struct sysrq_key_op sysrq_unrt_op
+ .enable_mask = SYSRQ_ENABLE_RTNICE,
+ };
+
++
++#ifdef CONFIG_VSERVER_DEBUG
++static void sysrq_handle_vxinfo(int key, struct tty_struct *tty)
++{
++ dump_vx_info_inactive((key == 'x')?0:1);
++}
++
++static struct sysrq_key_op sysrq_showvxinfo_op = {
++ .handler = sysrq_handle_vxinfo,
++ .help_msg = "conteXt",
++ .action_msg = "Show Context Info",
++ .enable_mask = SYSRQ_ENABLE_DUMP,
++};
++#endif
++
+ /* Key Operations table and lock */
+ static DEFINE_SPINLOCK(sysrq_key_table_lock);
+
+@@ -404,7 +420,11 @@ static struct sysrq_key_op *sysrq_key_ta
+ NULL, /* x */
+ /* y: May be registered on sparc64 for global register dump */
+ NULL, /* y */
+- NULL /* z */
++#ifdef CONFIG_VSERVER_DEBUG
++ &sysrq_showvxinfo_op, /* z */
++#else
++ NULL, /* z */
++#endif
+ };
+
+ /* key2index calculation, -1 on invalid index */
+@@ -416,6 +436,8 @@ static int sysrq_key_table_key2index(int
+ retval = key - '0';
+ else if ((key >= 'a') && (key <= 'z'))
+ retval = key + 10 - 'a';
++ else if ((key >= 'A') && (key <= 'Z'))
++ retval = key + 10 - 'A';
+ else
+ retval = -1;
+ return retval;
+--- a/drivers/char/tty_io.c 2008-07-14 17:22:33.000000000 -0400
++++ a/drivers/char/tty_io.c 2008-07-16 22:41:36.000000000 -0400
+@@ -105,6 +105,7 @@
+
+ #include <linux/kmod.h>
+ #include <linux/nsproxy.h>
++#include <linux/vs_pid.h>
+
+ #undef TTY_DEBUG_HANGUP
+
+@@ -3230,6 +3231,7 @@ static int tiocspgrp(struct tty_struct *
+ return -ENOTTY;
+ if (get_user(pgrp_nr, p))
+ return -EFAULT;
++ pgrp_nr = vx_rmap_pid(pgrp_nr);
+ if (pgrp_nr < 0)
+ return -EINVAL;
+ rcu_read_lock();
+--- a/drivers/infiniband/hw/ipath/ipath_user_pages.c 2008-04-17 10:37:17.000000000 -0400
++++ a/drivers/infiniband/hw/ipath/ipath_user_pages.c 2008-07-16 22:41:36.000000000 -0400
+@@ -33,6 +33,7 @@
+
+ #include <linux/mm.h>
+ #include <linux/device.h>
++#include <linux/vs_memory.h>
+
+ #include "ipath_kernel.h"
+
+@@ -61,7 +62,8 @@ static int __get_user_pages(unsigned lon
+ lock_limit = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >>
+ PAGE_SHIFT;
+
+- if (num_pages > lock_limit) {
++ if (num_pages > lock_limit ||
++ !vx_vmlocked_avail(current->mm, num_pages)) {
+ ret = -ENOMEM;
+ goto bail;
+ }
+@@ -78,7 +80,7 @@ static int __get_user_pages(unsigned lon
+ goto bail_release;
+ }
+
+- current->mm->locked_vm += num_pages;
++ vx_vmlocked_add(current->mm, num_pages);
+
+ ret = 0;
+ goto bail;
+@@ -177,7 +179,7 @@ void ipath_release_user_pages(struct pag
+
+ __ipath_release_user_pages(p, num_pages, 1);
+
+- current->mm->locked_vm -= num_pages;
++ vx_vmlocked_sub(current->mm, num_pages);
+
+ up_write(¤t->mm->mmap_sem);
+ }
+@@ -194,7 +196,7 @@ static void user_pages_account(struct wo
+ container_of(_work, struct ipath_user_pages_work, work);
+
+ down_write(&work->mm->mmap_sem);
+- work->mm->locked_vm -= work->num_pages;
++ vx_vmlocked_sub(work->mm, work->num_pages);
+ up_write(&work->mm->mmap_sem);
+ mmput(work->mm);
+ kfree(work);
+--- a/drivers/md/dm.c 2008-07-14 17:22:35.000000000 -0400
++++ a/drivers/md/dm.c 2008-07-16 22:41:36.000000000 -0400
+@@ -22,6 +22,7 @@
+ #include <linux/hdreg.h>
+ #include <linux/blktrace_api.h>
+ #include <linux/smp_lock.h>
++#include <linux/vs_base.h>
+
+ #define DM_MSG_PREFIX "core"
+
+@@ -91,6 +92,7 @@ struct mapped_device {
+ rwlock_t map_lock;
+ atomic_t holders;
+ atomic_t open_count;
++ xid_t xid;
+
+ unsigned long flags;
+
+@@ -252,6 +254,7 @@ static void __exit dm_exit(void)
+ static int dm_blk_open(struct inode *inode, struct file *file)
+ {
+ struct mapped_device *md;
++ int ret = -ENXIO;
+
+ spin_lock(&_minor_lock);
+
+@@ -260,18 +263,19 @@ static int dm_blk_open(struct inode *ino
+ goto out;
+
+ if (test_bit(DMF_FREEING, &md->flags) ||
+- test_bit(DMF_DELETING, &md->flags)) {
+- md = NULL;
++ test_bit(DMF_DELETING, &md->flags))
++ goto out;
++
++ ret = -EACCES;
++ if (!vx_check(md->xid, VS_IDENT|VS_HOSTID))
+ goto out;
+- }
+
+ dm_get(md);
+ atomic_inc(&md->open_count);
+-
++ ret = 0;
+ out:
+ spin_unlock(&_minor_lock);
+-
+- return md ? 0 : -ENXIO;
++ return ret;
+ }
+
+ static int dm_blk_close(struct inode *inode, struct file *file)
+@@ -467,6 +471,14 @@ int dm_set_geometry(struct mapped_device
+ return 0;
+ }
+
++/*
++ * Get the xid associated with a dm device
++ */
++xid_t dm_get_xid(struct mapped_device *md)
++{
++ return md->xid;
++}
++
+ /*-----------------------------------------------------------------
+ * CRUD START:
+ * A more elegant soln is in the works that uses the queue
+@@ -1022,6 +1034,7 @@ static struct mapped_device *alloc_dev(i
+ INIT_LIST_HEAD(&md->uevent_list);
+ spin_lock_init(&md->uevent_lock);
+
++ md->xid = vx_current_xid();
+ md->queue = blk_alloc_queue(GFP_KERNEL);
+ if (!md->queue)
+ goto bad_queue;
+--- a/drivers/md/dm.h 2008-07-14 17:22:35.000000000 -0400
++++ a/drivers/md/dm.h 2008-07-16 22:41:36.000000000 -0400
+@@ -66,6 +66,8 @@ void dm_put_target_type(struct target_ty
+ int dm_target_iterate(void (*iter_func)(struct target_type *tt,
+ void *param), void *param);
+
++xid_t dm_get_xid(struct mapped_device *md);
++
+ /*-----------------------------------------------------------------
+ * Useful inlines.
+ *---------------------------------------------------------------*/
+--- a/drivers/md/dm-ioctl.c 2008-04-17 12:05:33.000000000 -0400
++++ a/drivers/md/dm-ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -16,6 +16,7 @@
+ #include <linux/dm-ioctl.h>
+ #include <linux/hdreg.h>
+ #include <linux/compat.h>
++#include <linux/vs_context.h>
+
+ #include <asm/uaccess.h>
+
+@@ -101,7 +102,8 @@ static struct hash_cell *__get_name_cell
+ unsigned int h = hash_str(str);
+
+ list_for_each_entry (hc, _name_buckets + h, name_list)
+- if (!strcmp(hc->name, str)) {
++ if (vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT) &&
++ !strcmp(hc->name, str)) {
+ dm_get(hc->md);
+ return hc;
+ }
+@@ -115,7 +117,8 @@ static struct hash_cell *__get_uuid_cell
+ unsigned int h = hash_str(str);
+
+ list_for_each_entry (hc, _uuid_buckets + h, uuid_list)
+- if (!strcmp(hc->uuid, str)) {
++ if (vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT) &&
++ !strcmp(hc->uuid, str)) {
+ dm_get(hc->md);
+ return hc;
+ }
+@@ -352,6 +355,9 @@ typedef int (*ioctl_fn)(struct dm_ioctl
+
+ static int remove_all(struct dm_ioctl *param, size_t param_size)
+ {
++ if (!vx_check(0, VS_ADMIN))
++ return -EPERM;
++
+ dm_hash_remove_all(1);
+ param->data_size = 0;
+ return 0;
+@@ -399,6 +405,8 @@ static int list_devices(struct dm_ioctl
+ */
+ for (i = 0; i < NUM_BUCKETS; i++) {
+ list_for_each_entry (hc, _name_buckets + i, name_list) {
++ if (!vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT))
++ continue;
+ needed += sizeof(struct dm_name_list);
+ needed += strlen(hc->name) + 1;
+ needed += ALIGN_MASK;
+@@ -422,6 +430,8 @@ static int list_devices(struct dm_ioctl
+ */
+ for (i = 0; i < NUM_BUCKETS; i++) {
+ list_for_each_entry (hc, _name_buckets + i, name_list) {
++ if (!vx_check(dm_get_xid(hc->md), VS_WATCH_P | VS_IDENT))
++ continue;
+ if (old_nl)
+ old_nl->next = (uint32_t) ((void *) nl -
+ (void *) old_nl);
+@@ -612,10 +622,11 @@ static struct hash_cell *__find_device_h
+ if (!md)
+ goto out;
+
+- mdptr = dm_get_mdptr(md);
++ if (vx_check(dm_get_xid(md), VS_WATCH_P | VS_IDENT))
++ mdptr = dm_get_mdptr(md);
++
+ if (!mdptr)
+ dm_put(md);
+-
+ out:
+ return mdptr;
+ }
+@@ -1406,8 +1417,8 @@ static int ctl_ioctl(uint command, struc
+ ioctl_fn fn = NULL;
+ size_t param_size;
+
+- /* only root can play with this */
+- if (!capable(CAP_SYS_ADMIN))
++ /* only root and certain contexts can play with this */
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_ADMIN_MAPPER))
+ return -EACCES;
+
+ if (_IOC_TYPE(command) != DM_IOCTL)
+--- a/drivers/net/tun.c 2008-07-14 17:22:40.000000000 -0400
++++ a/drivers/net/tun.c 2008-07-17 21:09:54.000000000 -0400
+@@ -63,6 +63,7 @@
+ #include <linux/if_tun.h>
+ #include <linux/crc32.h>
+ #include <linux/nsproxy.h>
++#include <linux/vs_network.h>
+ #include <net/net_namespace.h>
+ #include <net/netns/generic.h>
+
+@@ -88,6 +89,7 @@ struct tun_struct {
+ int attached;
+ uid_t owner;
+ gid_t group;
++ nid_t nid;
+
+ wait_queue_head_t read_wait;
+ struct sk_buff_head readq;
+@@ -486,6 +488,7 @@ static void tun_setup(struct net_device
+
+ tun->owner = -1;
+ tun->group = -1;
++ tun->nid = current->nid;
+
+ dev->open = tun_net_open;
+ dev->hard_start_xmit = tun_net_xmit;
+@@ -518,6 +521,9 @@ static int tun_set_iff(struct net *net,
+ tn = net_generic(net, tun_net_id);
+ tun = tun_get_by_name(tn, ifr->ifr_name);
+ if (tun) {
++ if (!nx_check(tun->nid, VS_IDENT | VS_HOSTID | VS_ADMIN_P))
++ return -EPERM;
++
+ if (tun->attached)
+ return -EBUSY;
+
+@@ -526,7 +532,7 @@ static int tun_set_iff(struct net *net,
+ current->euid != tun->owner) ||
+ (tun->group != -1 &&
+ current->egid != tun->group)) &&
+- !capable(CAP_NET_ADMIN))
++ !cap_raised(current->cap_effective, CAP_NET_ADMIN))
+ return -EPERM;
+ }
+ else if (__dev_get_by_name(net, ifr->ifr_name))
+@@ -537,7 +543,7 @@ static int tun_set_iff(struct net *net,
+
+ err = -EINVAL;
+
+- if (!capable(CAP_NET_ADMIN))
++ if (!nx_capable(CAP_NET_ADMIN, NXC_TUN_CREATE))
+ return -EPERM;
+
+ /* Set dev type */
+@@ -688,6 +694,16 @@ static int tun_chr_ioctl(struct inode *i
+ DBG(KERN_INFO "%s: group set to %d\n", tun->dev->name, tun->group);
+ break;
+
++ case TUNSETNID:
++ if (!capable(CAP_CONTEXT))
++ return -EPERM;
++
++ /* Set nid owner of the device */
++ tun->nid = (nid_t) arg;
++
++ DBG(KERN_INFO "%s: nid owner set to %u\n", tun->dev->name, tun->nid);
++ break;
++
+ case TUNSETLINK:
+ {
+ int ret;
+--- a/fs/attr.c 2008-04-17 11:31:35.000000000 -0400
++++ a/fs/attr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -14,6 +14,9 @@
+ #include <linux/fcntl.h>
+ #include <linux/quotaops.h>
+ #include <linux/security.h>
++#include <linux/proc_fs.h>
++#include <linux/devpts_fs.h>
++#include <linux/vs_base.h>
+
+ /* Taken over from the old code... */
+
+@@ -55,6 +58,27 @@ int inode_change_ok(struct inode *inode,
+ if (!is_owner_or_cap(inode))
+ goto error;
+ }
++
++ /* Check for evil vserver activity */
++ if (vx_check(0, VS_ADMIN))
++ goto fine;
++
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ goto error;
++ }
++ switch (inode->i_sb->s_magic) {
++ case PROC_SUPER_MAGIC:
++ /* maybe allow that in the future? */
++ vxwprintk_task(1, "messing with the procfs.");
++ goto error;
++ case DEVPTS_SUPER_MAGIC:
++ /* devpts is xid tagged */
++ if (vx_check((xid_t)inode->i_tag, VS_IDENT))
++ goto fine;
++ vxwprintk_task(1, "messing with the devpts.");
++ goto error;
++ }
+ fine:
+ retval = 0;
+ error:
+@@ -78,6 +102,8 @@ int inode_setattr(struct inode * inode,
+ inode->i_uid = attr->ia_uid;
+ if (ia_valid & ATTR_GID)
+ inode->i_gid = attr->ia_gid;
++ if ((ia_valid & ATTR_TAG) && IS_TAGGED(inode))
++ inode->i_tag = attr->ia_tag;
+ if (ia_valid & ATTR_ATIME)
+ inode->i_atime = timespec_trunc(attr->ia_atime,
+ inode->i_sb->s_time_gran);
+@@ -167,7 +193,8 @@ int notify_change(struct dentry * dentry
+ error = security_inode_setattr(dentry, attr);
+ if (!error) {
+ if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid))
++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) ||
++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag))
+ error = DQUOT_TRANSFER(inode, attr) ? -EDQUOT : 0;
+ if (!error)
+ error = inode_setattr(inode, attr);
+--- a/fs/binfmt_aout.c 2008-07-14 17:22:48.000000000 -0400
++++ a/fs/binfmt_aout.c 2008-07-16 22:41:36.000000000 -0400
+@@ -24,6 +24,7 @@
+ #include <linux/binfmts.h>
+ #include <linux/personality.h>
+ #include <linux/init.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/system.h>
+ #include <asm/uaccess.h>
+--- a/fs/binfmt_elf.c 2008-07-14 17:22:48.000000000 -0400
++++ a/fs/binfmt_elf.c 2008-07-16 22:41:36.000000000 -0400
+@@ -38,6 +38,7 @@
+ #include <linux/random.h>
+ #include <linux/elf.h>
+ #include <linux/utsname.h>
++#include <linux/vs_memory.h>
+ #include <asm/uaccess.h>
+ #include <asm/param.h>
+ #include <asm/page.h>
+--- a/fs/binfmt_flat.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/binfmt_flat.c 2008-07-16 22:41:36.000000000 -0400
+@@ -35,6 +35,7 @@
+ #include <linux/init.h>
+ #include <linux/flat.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/byteorder.h>
+ #include <asm/system.h>
+--- a/fs/binfmt_som.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/binfmt_som.c 2008-07-16 22:41:36.000000000 -0400
+@@ -28,6 +28,7 @@
+ #include <linux/shm.h>
+ #include <linux/personality.h>
+ #include <linux/init.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/pgtable.h>
+--- a/fs/block_dev.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/block_dev.c 2008-07-16 22:41:36.000000000 -0400
+@@ -24,6 +24,7 @@
+ #include <linux/uio.h>
+ #include <linux/namei.h>
+ #include <linux/log2.h>
++#include <linux/vs_device.h>
+ #include <asm/uaccess.h>
+ #include "internal.h"
+
+@@ -389,6 +390,7 @@ struct block_device *bdget(dev_t dev)
+ bdev->bd_invalidated = 0;
+ inode->i_mode = S_IFBLK;
+ inode->i_rdev = dev;
++ inode->i_mdev = dev;
+ inode->i_bdev = bdev;
+ inode->i_data.a_ops = &def_blk_aops;
+ mapping_set_gfp_mask(&inode->i_data, GFP_USER);
+@@ -425,6 +427,11 @@ EXPORT_SYMBOL(bdput);
+ static struct block_device *bd_acquire(struct inode *inode)
+ {
+ struct block_device *bdev;
++ dev_t mdev;
++
++ if (!vs_map_blkdev(inode->i_rdev, &mdev, DATTR_OPEN))
++ return NULL;
++ inode->i_mdev = mdev;
+
+ spin_lock(&bdev_lock);
+ bdev = inode->i_bdev;
+@@ -435,7 +442,7 @@ static struct block_device *bd_acquire(s
+ }
+ spin_unlock(&bdev_lock);
+
+- bdev = bdget(inode->i_rdev);
++ bdev = bdget(mdev);
+ if (bdev) {
+ spin_lock(&bdev_lock);
+ if (!inode->i_bdev) {
+--- a/fs/char_dev.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/char_dev.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,8 @@
+ #include <linux/cdev.h>
+ #include <linux/mutex.h>
+ #include <linux/backing-dev.h>
++#include <linux/vs_context.h>
++#include <linux/vs_device.h>
+
+ #ifdef CONFIG_KMOD
+ #include <linux/kmod.h>
+@@ -361,14 +363,21 @@ static int chrdev_open(struct inode *ino
+ struct cdev *p;
+ struct cdev *new = NULL;
+ int ret = 0;
++ dev_t mdev;
++
++ if (!vs_map_chrdev(inode->i_rdev, &mdev, DATTR_OPEN))
++ return -EPERM;
++ inode->i_mdev = mdev;
+
+ spin_lock(&cdev_lock);
+ p = inode->i_cdev;
+ if (!p) {
+ struct kobject *kobj;
+ int idx;
++
+ spin_unlock(&cdev_lock);
+- kobj = kobj_lookup(cdev_map, inode->i_rdev, &idx);
++
++ kobj = kobj_lookup(cdev_map, mdev, &idx);
+ if (!kobj)
+ return -ENXIO;
+ new = container_of(kobj, struct cdev, kobj);
+--- a/fs/dcache.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/dcache.c 2008-07-16 22:41:36.000000000 -0400
+@@ -32,6 +32,7 @@
+ #include <linux/seqlock.h>
+ #include <linux/swap.h>
+ #include <linux/bootmem.h>
++#include <linux/vs_limit.h>
+ #include "internal.h"
+
+
+@@ -187,6 +188,7 @@ void dput(struct dentry *dentry)
+ if (!dentry)
+ return;
+
++ vx_dentry_dec(dentry);
+ repeat:
+ if (atomic_read(&dentry->d_count) == 1)
+ might_sleep();
+@@ -200,6 +202,8 @@ repeat:
+ return;
+ }
+
++ vx_dentry_dec(dentry);
++
+ /*
+ * AV: ->d_delete() is _NOT_ allowed to block now.
+ */
+@@ -291,6 +295,7 @@ static inline struct dentry * __dget_loc
+ {
+ atomic_inc(&dentry->d_count);
+ dentry_lru_remove(dentry);
++ vx_dentry_inc(dentry);
+ return dentry;
+ }
+
+@@ -888,6 +893,9 @@ struct dentry *d_alloc(struct dentry * p
+ struct dentry *dentry;
+ char *dname;
+
++ if (!vx_dentry_avail(1))
++ return NULL;
++
+ dentry = kmem_cache_alloc(dentry_cache, GFP_KERNEL);
+ if (!dentry)
+ return NULL;
+@@ -936,6 +944,7 @@ struct dentry *d_alloc(struct dentry * p
+ if (parent)
+ list_add(&dentry->d_u.d_child, &parent->d_subdirs);
+ dentry_stat.nr_dentry++;
++ vx_dentry_inc(dentry);
+ spin_unlock(&dcache_lock);
+
+ return dentry;
+@@ -1285,6 +1294,7 @@ struct dentry * __d_lookup(struct dentry
+
+ if (!d_unhashed(dentry)) {
+ atomic_inc(&dentry->d_count);
++ vx_dentry_inc(dentry);
+ found = dentry;
+ }
+ spin_unlock(&dentry->d_lock);
+--- a/fs/devpts/inode.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/devpts/inode.c 2008-07-17 17:43:28.000000000 -0400
+@@ -19,15 +19,29 @@
+ #include <linux/tty.h>
+ #include <linux/mutex.h>
+ #include <linux/idr.h>
++#include <linux/magic.h>
+ #include <linux/devpts_fs.h>
+ #include <linux/parser.h>
+ #include <linux/fsnotify.h>
+ #include <linux/seq_file.h>
+-
+-#define DEVPTS_SUPER_MAGIC 0x1cd1
++#include <linux/vs_base.h>
+
+ #define DEVPTS_DEFAULT_MODE 0600
+
++static int devpts_permission(struct inode *inode, int mask, struct nameidata *nd)
++{
++ int ret = -EACCES;
++
++ /* devpts is xid tagged */
++ if (vx_check((xid_t)inode->i_tag, VS_WATCH_P | VS_IDENT))
++ ret = generic_permission(inode, mask, NULL);
++ return ret;
++}
++
++static struct inode_operations devpts_file_inode_operations = {
++ .permission = devpts_permission,
++};
++
+ extern int pty_limit; /* Config limit on Unix98 ptys */
+ static DEFINE_IDR(allocated_ptys);
+ static DEFINE_MUTEX(allocated_ptys_lock);
+@@ -112,6 +126,25 @@ static int devpts_show_options(struct se
+ return 0;
+ }
+
++static int devpts_filter(struct dentry *de)
++{
++ /* devpts is xid tagged */
++ return vx_check((xid_t)de->d_inode->i_tag, VS_WATCH_P | VS_IDENT);
++}
++
++static int devpts_readdir(struct file * filp, void * dirent, filldir_t filldir)
++{
++ return dcache_readdir_filter(filp, dirent, filldir, devpts_filter);
++}
++
++static struct file_operations devpts_dir_operations = {
++ .open = dcache_dir_open,
++ .release = dcache_dir_close,
++ .llseek = dcache_dir_lseek,
++ .read = generic_read_dir,
++ .readdir = devpts_readdir,
++};
++
+ static const struct super_operations devpts_sops = {
+ .statfs = simple_statfs,
+ .remount_fs = devpts_remount,
+@@ -138,8 +171,10 @@ devpts_fill_super(struct super_block *s,
+ inode->i_uid = inode->i_gid = 0;
+ inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR;
+ inode->i_op = &simple_dir_inode_operations;
+- inode->i_fop = &simple_dir_operations;
++ inode->i_fop = &devpts_dir_operations;
+ inode->i_nlink = 2;
++ /* devpts is xid tagged */
++ inode->i_tag = (tag_t)vx_current_xid();
+
+ devpts_root = s->s_root = d_alloc_root(inode);
+ if (s->s_root)
+@@ -232,6 +267,9 @@ int devpts_pty_new(struct tty_struct *tt
+ inode->i_gid = config.setgid ? config.gid : current->fsgid;
+ inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
+ init_special_inode(inode, S_IFCHR|config.mode, device);
++ /* devpts is xid tagged */
++ inode->i_tag = (tag_t)vx_current_xid();
++ inode->i_op = &devpts_file_inode_operations;
+ inode->i_private = tty;
+
+ dentry = get_node(number);
+--- a/fs/exec.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/exec.c 2008-07-16 22:41:36.000000000 -0400
+@@ -254,7 +254,9 @@ static int __bprm_mm_init(struct linux_b
+ goto err;
+ }
+
+- mm->stack_vm = mm->total_vm = 1;
++ mm->total_vm = 0;
++ vx_vmpages_inc(mm);
++ mm->stack_vm = 1;
+ up_write(&mm->mmap_sem);
+
+ bprm->p = vma->vm_end - sizeof(void *);
+@@ -1447,7 +1449,7 @@ static int format_corename(char *corenam
+ /* UNIX time of coredump */
+ case 't': {
+ struct timeval tv;
+- do_gettimeofday(&tv);
++ vx_gettimeofday(&tv);
+ rc = snprintf(out_ptr, out_end - out_ptr,
+ "%lu", tv.tv_sec);
+ if (rc > out_end - out_ptr)
+--- a/fs/ext2/balloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/balloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -16,6 +16,8 @@
+ #include <linux/sched.h>
+ #include <linux/buffer_head.h>
+ #include <linux/capability.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+
+ /*
+ * balloc.c contains the blocks allocation and deallocation routines
+@@ -569,6 +571,7 @@ do_more:
+ }
+ error_return:
+ brelse(bitmap_bh);
++ DLIMIT_FREE_BLOCK(inode, freed);
+ release_blocks(sb, freed);
+ DQUOT_FREE_BLOCK(inode, freed);
+ }
+@@ -701,7 +704,6 @@ ext2_try_to_allocate(struct super_block
+ start = 0;
+ end = EXT2_BLOCKS_PER_GROUP(sb);
+ }
+-
+ BUG_ON(start > EXT2_BLOCKS_PER_GROUP(sb));
+
+ repeat:
+@@ -1251,6 +1253,11 @@ ext2_fsblk_t ext2_new_blocks(struct inod
+ *errp = -EDQUOT;
+ return 0;
+ }
++ if (DLIMIT_ALLOC_BLOCK(inode, num)) {
++ *errp = -ENOSPC;
++ DQUOT_FREE_BLOCK(inode, num);
++ return 0;
++ }
+
+ sbi = EXT2_SB(sb);
+ es = EXT2_SB(sb)->s_es;
+@@ -1408,6 +1415,7 @@ allocated:
+
+ *errp = 0;
+ brelse(bitmap_bh);
++ DLIMIT_FREE_BLOCK(inode, *count-num);
+ DQUOT_FREE_BLOCK(inode, *count-num);
+ *count = num;
+ return ret_block;
+@@ -1418,8 +1426,10 @@ out:
+ /*
+ * Undo the block allocation
+ */
+- if (!performed_allocation)
++ if (!performed_allocation) {
++ DLIMIT_FREE_BLOCK(inode, *count);
+ DQUOT_FREE_BLOCK(inode, *count);
++ }
+ brelse(bitmap_bh);
+ return 0;
+ }
+--- a/fs/ext2/ext2.h 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ext2/ext2.h 2008-07-16 22:41:36.000000000 -0400
+@@ -168,6 +168,7 @@ extern const struct file_operations ext2
+ extern const struct address_space_operations ext2_aops;
+ extern const struct address_space_operations ext2_aops_xip;
+ extern const struct address_space_operations ext2_nobh_aops;
++extern int ext2_sync_flags(struct inode *inode);
+
+ /* namei.c */
+ extern const struct inode_operations ext2_dir_inode_operations;
+--- a/fs/ext2/file.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ext2/file.c 2008-07-16 22:41:36.000000000 -0400
+@@ -86,4 +86,5 @@ const struct inode_operations ext2_file_
+ #endif
+ .setattr = ext2_setattr,
+ .permission = ext2_permission,
++ .sync_flags = ext2_sync_flags,
+ };
+--- a/fs/ext2/ialloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/ialloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -17,6 +17,8 @@
+ #include <linux/backing-dev.h>
+ #include <linux/buffer_head.h>
+ #include <linux/random.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+ #include "ext2.h"
+ #include "xattr.h"
+ #include "acl.h"
+@@ -123,6 +125,7 @@ void ext2_free_inode (struct inode * ino
+ ext2_xattr_delete_inode(inode);
+ DQUOT_FREE_INODE(inode);
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+ }
+
+ es = EXT2_SB(sb)->s_es;
+@@ -454,6 +457,11 @@ struct inode *ext2_new_inode(struct inod
+ if (!inode)
+ return ERR_PTR(-ENOMEM);
+
++ inode->i_tag = dx_current_fstag(sb);
++ if (DLIMIT_ALLOC_INODE(inode)) {
++ err = -ENOSPC;
++ goto fail_dlim;
++ }
+ ei = EXT2_I(inode);
+ sbi = EXT2_SB(sb);
+ es = sbi->s_es;
+@@ -565,7 +573,8 @@ got:
+ inode->i_blocks = 0;
+ inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC;
+ memset(ei->i_data, 0, sizeof(ei->i_data));
+- ei->i_flags = EXT2_I(dir)->i_flags & ~EXT2_BTREE_FL;
++ ei->i_flags = EXT2_I(dir)->i_flags &
++ ~(EXT2_BTREE_FL|EXT2_IXUNLINK_FL|EXT2_BARRIER_FL);
+ if (S_ISLNK(mode))
+ ei->i_flags &= ~(EXT2_IMMUTABLE_FL|EXT2_APPEND_FL);
+ /* dirsync is only applied to directories */
+@@ -610,12 +619,15 @@ fail_free_drop:
+
+ fail_drop:
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+ inode->i_flags |= S_NOQUOTA;
+ inode->i_nlink = 0;
+ iput(inode);
+ return ERR_PTR(err);
+
+ fail:
++ DLIMIT_FREE_INODE(inode);
++fail_dlim:
+ make_bad_inode(inode);
+ iput(inode);
+ return ERR_PTR(err);
+--- a/fs/ext2/inode.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -31,6 +31,7 @@
+ #include <linux/writeback.h>
+ #include <linux/buffer_head.h>
+ #include <linux/mpage.h>
++#include <linux/vs_tag.h>
+ #include "ext2.h"
+ #include "acl.h"
+ #include "xip.h"
+@@ -1010,7 +1011,7 @@ void ext2_truncate(struct inode *inode)
+ return;
+ if (ext2_inode_is_fast_symlink(inode))
+ return;
+- if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode))
+ return;
+
+ blocksize = inode->i_sb->s_blocksize;
+@@ -1148,38 +1149,72 @@ void ext2_set_inode_flags(struct inode *
+ {
+ unsigned int flags = EXT2_I(inode)->i_flags;
+
+- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK |
++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
++
++
++ if (flags & EXT2_IMMUTABLE_FL)
++ inode->i_flags |= S_IMMUTABLE;
++ if (flags & EXT2_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
++
+ if (flags & EXT2_SYNC_FL)
+ inode->i_flags |= S_SYNC;
+ if (flags & EXT2_APPEND_FL)
+ inode->i_flags |= S_APPEND;
+- if (flags & EXT2_IMMUTABLE_FL)
+- inode->i_flags |= S_IMMUTABLE;
+ if (flags & EXT2_NOATIME_FL)
+ inode->i_flags |= S_NOATIME;
+ if (flags & EXT2_DIRSYNC_FL)
+ inode->i_flags |= S_DIRSYNC;
++
++ inode->i_vflags &= ~(V_BARRIER | V_COW);
++
++ if (flags & EXT2_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ if (flags & EXT2_COW_FL)
++ inode->i_vflags |= V_COW;
+ }
+
+ /* Propagate flags from i_flags to EXT2_I(inode)->i_flags */
+ void ext2_get_inode_flags(struct ext2_inode_info *ei)
+ {
+ unsigned int flags = ei->vfs_inode.i_flags;
++ unsigned int vflags = ei->vfs_inode.i_vflags;
++
++ ei->i_flags &= ~(EXT2_SYNC_FL | EXT2_APPEND_FL |
++ EXT2_IMMUTABLE_FL | EXT2_IXUNLINK_FL |
++ EXT2_NOATIME_FL | EXT2_DIRSYNC_FL |
++ EXT2_BARRIER_FL | EXT2_COW_FL);
++
++ if (flags & S_IMMUTABLE)
++ ei->i_flags |= EXT2_IMMUTABLE_FL;
++ if (flags & S_IXUNLINK)
++ ei->i_flags |= EXT2_IXUNLINK_FL;
+
+- ei->i_flags &= ~(EXT2_SYNC_FL|EXT2_APPEND_FL|
+- EXT2_IMMUTABLE_FL|EXT2_NOATIME_FL|EXT2_DIRSYNC_FL);
+ if (flags & S_SYNC)
+ ei->i_flags |= EXT2_SYNC_FL;
+ if (flags & S_APPEND)
+ ei->i_flags |= EXT2_APPEND_FL;
+- if (flags & S_IMMUTABLE)
+- ei->i_flags |= EXT2_IMMUTABLE_FL;
+ if (flags & S_NOATIME)
+ ei->i_flags |= EXT2_NOATIME_FL;
+ if (flags & S_DIRSYNC)
+ ei->i_flags |= EXT2_DIRSYNC_FL;
++
++ if (vflags & V_BARRIER)
++ ei->i_flags |= EXT2_BARRIER_FL;
++ if (vflags & V_COW)
++ ei->i_flags |= EXT2_COW_FL;
+ }
+
++int ext2_sync_flags(struct inode *inode)
++{
++ ext2_get_inode_flags(EXT2_I(inode));
++ inode->i_ctime = CURRENT_TIME;
++ mark_inode_dirty(inode);
++ return 0;
++}
++
++
+ struct inode *ext2_iget (struct super_block *sb, unsigned long ino)
+ {
+ struct ext2_inode_info *ei;
+@@ -1187,6 +1222,8 @@ struct inode *ext2_iget (struct super_bl
+ struct ext2_inode *raw_inode;
+ struct inode *inode;
+ long ret = -EIO;
++ uid_t uid;
++ gid_t gid;
+ int n;
+
+ inode = iget_locked(sb, ino);
+@@ -1209,12 +1246,17 @@ struct inode *ext2_iget (struct super_bl
+ }
+
+ inode->i_mode = le16_to_cpu(raw_inode->i_mode);
+- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
+- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
+ if (!(test_opt (inode->i_sb, NO_UID32))) {
+- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
+- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
+ }
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid,
++ le16_to_cpu(raw_inode->i_raw_tag));
++
+ inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
+ inode->i_size = le32_to_cpu(raw_inode->i_size);
+ inode->i_atime.tv_sec = (signed)le32_to_cpu(raw_inode->i_atime);
+@@ -1310,8 +1352,8 @@ static int ext2_update_inode(struct inod
+ struct ext2_inode_info *ei = EXT2_I(inode);
+ struct super_block *sb = inode->i_sb;
+ ino_t ino = inode->i_ino;
+- uid_t uid = inode->i_uid;
+- gid_t gid = inode->i_gid;
++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag);
++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag);
+ struct buffer_head * bh;
+ struct ext2_inode * raw_inode = ext2_get_inode(sb, ino, &bh);
+ int n;
+@@ -1347,6 +1389,9 @@ static int ext2_update_inode(struct inod
+ raw_inode->i_uid_high = 0;
+ raw_inode->i_gid_high = 0;
+ }
++#ifdef CONFIG_TAGGING_INTERN
++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag);
++#endif
+ raw_inode->i_links_count = cpu_to_le16(inode->i_nlink);
+ raw_inode->i_size = cpu_to_le32(inode->i_size);
+ raw_inode->i_atime = cpu_to_le32(inode->i_atime.tv_sec);
+@@ -1433,7 +1478,8 @@ int ext2_setattr(struct dentry *dentry,
+ if (error)
+ return error;
+ if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
+- (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
++ (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) ||
++ (iattr->ia_valid & ATTR_TAG && iattr->ia_tag != inode->i_tag)) {
+ error = DQUOT_TRANSFER(inode, iattr) ? -EDQUOT : 0;
+ if (error)
+ return error;
+--- a/fs/ext2/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -14,6 +14,7 @@
+ #include <linux/compat.h>
+ #include <linux/mount.h>
+ #include <linux/smp_lock.h>
++#include <linux/mount.h>
+ #include <asm/current.h>
+ #include <asm/uaccess.h>
+
+@@ -53,6 +54,11 @@ long ext2_ioctl(struct file *filp, unsig
+ if (!S_ISDIR(inode->i_mode))
+ flags &= ~EXT2_DIRSYNC_FL;
+
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -EACCES;
++ }
++
+ mutex_lock(&inode->i_mutex);
+ /* Is it quota file? Do not allow user to mess with it */
+ if (IS_NOQUOTA(inode)) {
+@@ -68,7 +74,9 @@ long ext2_ioctl(struct file *filp, unsig
+ *
+ * This test looks nicer. Thanks to Pauline Middelink
+ */
+- if ((flags ^ oldflags) & (EXT2_APPEND_FL | EXT2_IMMUTABLE_FL)) {
++ if ((oldflags & EXT2_IMMUTABLE_FL) ||
++ ((flags ^ oldflags) & (EXT2_APPEND_FL |
++ EXT2_IMMUTABLE_FL | EXT2_IXUNLINK_FL))) {
+ if (!capable(CAP_LINUX_IMMUTABLE)) {
+ mutex_unlock(&inode->i_mutex);
+ ret = -EPERM;
+--- a/fs/ext2/namei.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ext2/namei.c 2008-07-16 22:41:36.000000000 -0400
+@@ -31,6 +31,7 @@
+ */
+
+ #include <linux/pagemap.h>
++#include <linux/vs_tag.h>
+ #include "ext2.h"
+ #include "xattr.h"
+ #include "acl.h"
+@@ -66,6 +67,7 @@ static struct dentry *ext2_lookup(struct
+ inode = ext2_iget(dir->i_sb, ino);
+ if (IS_ERR(inode))
+ return ERR_CAST(inode);
++ dx_propagate_tag(nd, inode);
+ }
+ return d_splice_alias(inode, dentry);
+ }
+@@ -391,6 +393,7 @@ const struct inode_operations ext2_dir_i
+ #endif
+ .setattr = ext2_setattr,
+ .permission = ext2_permission,
++ .sync_flags = ext2_sync_flags,
+ };
+
+ const struct inode_operations ext2_special_inode_operations = {
+@@ -402,4 +405,5 @@ const struct inode_operations ext2_speci
+ #endif
+ .setattr = ext2_setattr,
+ .permission = ext2_permission,
++ .sync_flags = ext2_sync_flags,
+ };
+--- a/fs/ext2/super.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/super.c 2008-07-16 22:41:36.000000000 -0400
+@@ -389,7 +389,8 @@ enum {
+ Opt_err_ro, Opt_nouid32, Opt_nocheck, Opt_debug,
+ Opt_oldalloc, Opt_orlov, Opt_nobh, Opt_user_xattr, Opt_nouser_xattr,
+ Opt_acl, Opt_noacl, Opt_xip, Opt_ignore, Opt_err, Opt_quota,
+- Opt_usrquota, Opt_grpquota, Opt_reservation, Opt_noreservation
++ Opt_usrquota, Opt_grpquota, Opt_reservation, Opt_noreservation,
++ Opt_tag, Opt_notag, Opt_tagid
+ };
+
+ static match_table_t tokens = {
+@@ -417,6 +418,9 @@ static match_table_t tokens = {
+ {Opt_acl, "acl"},
+ {Opt_noacl, "noacl"},
+ {Opt_xip, "xip"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
+ {Opt_grpquota, "grpquota"},
+ {Opt_ignore, "noquota"},
+ {Opt_quota, "quota"},
+@@ -487,6 +491,20 @@ static int parse_options (char * options
+ case Opt_nouid32:
+ set_opt (sbi->s_mount_opt, NO_UID32);
+ break;
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++ case Opt_notag:
++ clear_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
+ case Opt_nocheck:
+ clear_opt (sbi->s_mount_opt, CHECK);
+ break;
+@@ -829,6 +847,8 @@ static int ext2_fill_super(struct super_
+ if (!parse_options ((char *) data, sbi))
+ goto failed_mount;
+
++ if (EXT2_SB(sb)->s_mount_opt & EXT2_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED;
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((EXT2_SB(sb)->s_mount_opt & EXT2_MOUNT_POSIX_ACL) ?
+ MS_POSIXACL : 0);
+@@ -1161,6 +1181,13 @@ static int ext2_remount (struct super_bl
+ goto restore_opts;
+ }
+
++ if ((sbi->s_mount_opt & EXT2_MOUNT_TAGGED) &&
++ !(sb->s_flags & MS_TAGGED)) {
++ printk("EXT2-fs: %s: tagging not permitted on remount.\n",
++ sb->s_id);
++ return -EINVAL;
++ }
++
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((sbi->s_mount_opt & EXT2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
+--- a/fs/ext2/symlink.c 2008-04-17 10:32:27.000000000 -0400
++++ a/fs/ext2/symlink.c 2008-07-16 22:41:36.000000000 -0400
+@@ -38,6 +38,7 @@ const struct inode_operations ext2_symli
+ .listxattr = ext2_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext2_sync_flags,
+ };
+
+ const struct inode_operations ext2_fast_symlink_inode_operations = {
+@@ -49,4 +50,5 @@ const struct inode_operations ext2_fast_
+ .listxattr = ext2_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext2_sync_flags,
+ };
+--- a/fs/ext2/xattr.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext2/xattr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -60,6 +60,7 @@
+ #include <linux/mbcache.h>
+ #include <linux/quotaops.h>
+ #include <linux/rwsem.h>
++#include <linux/vs_dlimit.h>
+ #include "ext2.h"
+ #include "xattr.h"
+ #include "acl.h"
+@@ -641,8 +642,12 @@ ext2_xattr_set2(struct inode *inode, str
+ the inode. */
+ ea_bdebug(new_bh, "reusing block");
+
++ error = -ENOSPC;
++ if (DLIMIT_ALLOC_BLOCK(inode, 1))
++ goto cleanup;
+ error = -EDQUOT;
+ if (DQUOT_ALLOC_BLOCK(inode, 1)) {
++ DLIMIT_FREE_BLOCK(inode, 1);
+ unlock_buffer(new_bh);
+ goto cleanup;
+ }
+@@ -731,6 +736,7 @@ ext2_xattr_set2(struct inode *inode, str
+ le32_add_cpu(&HDR(old_bh)->h_refcount, -1);
+ if (ce)
+ mb_cache_entry_release(ce);
++ DLIMIT_FREE_BLOCK(inode, 1);
+ DQUOT_FREE_BLOCK(inode, 1);
+ mark_buffer_dirty(old_bh);
+ ea_bdebug(old_bh, "refcount now=%d",
+@@ -794,6 +800,7 @@ ext2_xattr_delete_inode(struct inode *in
+ mark_buffer_dirty(bh);
+ if (IS_SYNC(inode))
+ sync_dirty_buffer(bh);
++ DLIMIT_FREE_BLOCK(inode, 1);
+ DQUOT_FREE_BLOCK(inode, 1);
+ }
+ EXT2_I(inode)->i_file_acl = 0;
+--- a/fs/ext3/balloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/balloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -19,6 +19,8 @@
+ #include <linux/ext3_jbd.h>
+ #include <linux/quotaops.h>
+ #include <linux/buffer_head.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+
+ /*
+ * balloc.c contains the blocks allocation and deallocation routines
+@@ -675,8 +677,10 @@ void ext3_free_blocks(handle_t *handle,
+ return;
+ }
+ ext3_free_blocks_sb(handle, sb, block, count, &dquot_freed_blocks);
+- if (dquot_freed_blocks)
++ if (dquot_freed_blocks) {
++ DLIMIT_FREE_BLOCK(inode, dquot_freed_blocks);
+ DQUOT_FREE_BLOCK(inode, dquot_freed_blocks);
++ }
+ return;
+ }
+
+@@ -1415,18 +1419,33 @@ out:
+ *
+ * Check if filesystem has at least 1 free block available for allocation.
+ */
+-static int ext3_has_free_blocks(struct ext3_sb_info *sbi)
++static int ext3_has_free_blocks(struct super_block *sb)
+ {
+- ext3_fsblk_t free_blocks, root_blocks;
++ struct ext3_sb_info *sbi = EXT3_SB(sb);
++ unsigned long long free_blocks, root_blocks;
++ int cond;
+
+ free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter);
+ root_blocks = le32_to_cpu(sbi->s_es->s_r_blocks_count);
+- if (free_blocks < root_blocks + 1 && !capable(CAP_SYS_RESOURCE) &&
++
++ vxdprintk(VXD_CBIT(dlim, 3),
++ "ext3_has_free_blocks(%p): free=%llu, root=%llu",
++ sb, free_blocks, root_blocks);
++
++ DLIMIT_ADJUST_BLOCK(sb, dx_current_tag(), &free_blocks, &root_blocks);
++
++ cond = (free_blocks < root_blocks + 1 &&
++ !capable(CAP_SYS_RESOURCE) &&
+ sbi->s_resuid != current->fsuid &&
+- (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid))) {
+- return 0;
+- }
+- return 1;
++ (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid)));
++
++ vxdprintk(VXD_CBIT(dlim, 3),
++ "ext3_has_free_blocks(%p): %llu<%llu+1, %c, %u!=%u r=%d",
++ sb, free_blocks, root_blocks,
++ !capable(CAP_SYS_RESOURCE)?'1':'0',
++ sbi->s_resuid, current->fsuid, cond?0:1);
++
++ return (cond ? 0 : 1);
+ }
+
+ /**
+@@ -1443,7 +1462,7 @@ static int ext3_has_free_blocks(struct e
+ */
+ int ext3_should_retry_alloc(struct super_block *sb, int *retries)
+ {
+- if (!ext3_has_free_blocks(EXT3_SB(sb)) || (*retries)++ > 3)
++ if (!ext3_has_free_blocks(sb) || (*retries)++ > 3)
+ return 0;
+
+ jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id);
+@@ -1506,6 +1525,8 @@ ext3_fsblk_t ext3_new_blocks(handle_t *h
+ *errp = -EDQUOT;
+ return 0;
+ }
++ if (DLIMIT_ALLOC_BLOCK(inode, num))
++ goto out_dlimit;
+
+ sbi = EXT3_SB(sb);
+ es = EXT3_SB(sb)->s_es;
+@@ -1522,7 +1543,7 @@ ext3_fsblk_t ext3_new_blocks(handle_t *h
+ if (block_i && ((windowsz = block_i->rsv_window_node.rsv_goal_size) > 0))
+ my_rsv = &block_i->rsv_window_node;
+
+- if (!ext3_has_free_blocks(sbi)) {
++ if (!ext3_has_free_blocks(sb)) {
+ *errp = -ENOSPC;
+ goto out;
+ }
+@@ -1714,12 +1735,16 @@ allocated:
+ *errp = 0;
+ brelse(bitmap_bh);
+ DQUOT_FREE_BLOCK(inode, *count-num);
++ DLIMIT_FREE_BLOCK(inode, *count-num);
+ *count = num;
+ return ret_block;
+
+ io_error:
+ *errp = -EIO;
+ out:
++ if (!performed_allocation)
++ DLIMIT_FREE_BLOCK(inode, *count);
++out_dlimit:
+ if (fatal) {
+ *errp = fatal;
+ ext3_std_error(sb, fatal);
+--- a/fs/ext3/file.c 2008-04-17 10:37:23.000000000 -0400
++++ a/fs/ext3/file.c 2008-07-16 22:41:36.000000000 -0400
+@@ -134,5 +134,6 @@ const struct inode_operations ext3_file_
+ .removexattr = generic_removexattr,
+ #endif
+ .permission = ext3_permission,
++ .sync_flags = ext3_sync_flags,
+ };
+
+--- a/fs/ext3/ialloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/ialloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -23,6 +23,8 @@
+ #include <linux/buffer_head.h>
+ #include <linux/random.h>
+ #include <linux/bitops.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/byteorder.h>
+
+@@ -127,6 +129,7 @@ void ext3_free_inode (handle_t *handle,
+ ext3_xattr_delete_inode(handle, inode);
+ DQUOT_FREE_INODE(inode);
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+
+ is_directory = S_ISDIR(inode->i_mode);
+
+@@ -440,6 +443,12 @@ struct inode *ext3_new_inode(handle_t *h
+ inode = new_inode(sb);
+ if (!inode)
+ return ERR_PTR(-ENOMEM);
++
++ inode->i_tag = dx_current_fstag(sb);
++ if (DLIMIT_ALLOC_INODE(inode)) {
++ err = -ENOSPC;
++ goto out_dlimit;
++ }
+ ei = EXT3_I(inode);
+
+ sbi = EXT3_SB(sb);
+@@ -559,7 +568,8 @@ got:
+ ei->i_dir_start_lookup = 0;
+ ei->i_disksize = 0;
+
+- ei->i_flags = EXT3_I(dir)->i_flags & ~EXT3_INDEX_FL;
++ ei->i_flags = EXT3_I(dir)->i_flags &
++ ~(EXT3_INDEX_FL|EXT3_IXUNLINK_FL|EXT3_BARRIER_FL);
+ if (S_ISLNK(mode))
+ ei->i_flags &= ~(EXT3_IMMUTABLE_FL|EXT3_APPEND_FL);
+ /* dirsync only applies to directories */
+@@ -614,6 +624,8 @@ got:
+ fail:
+ ext3_std_error(sb, err);
+ out:
++ DLIMIT_FREE_INODE(inode);
++out_dlimit:
+ iput(inode);
+ ret = ERR_PTR(err);
+ really_out:
+@@ -625,6 +637,7 @@ fail_free_drop:
+
+ fail_drop:
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+ inode->i_flags |= S_NOQUOTA;
+ inode->i_nlink = 0;
+ iput(inode);
+--- a/fs/ext3/inode.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -36,6 +36,7 @@
+ #include <linux/mpage.h>
+ #include <linux/uio.h>
+ #include <linux/bio.h>
++#include <linux/vs_tag.h>
+ #include "xattr.h"
+ #include "acl.h"
+
+@@ -2302,7 +2303,7 @@ void ext3_truncate(struct inode *inode)
+ return;
+ if (ext3_inode_is_fast_symlink(inode))
+ return;
+- if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode))
+ return;
+
+ /*
+@@ -2611,36 +2612,84 @@ void ext3_set_inode_flags(struct inode *
+ {
+ unsigned int flags = EXT3_I(inode)->i_flags;
+
+- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK |
++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
++
++ if (flags & EXT3_IMMUTABLE_FL)
++ inode->i_flags |= S_IMMUTABLE;
++ if (flags & EXT3_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
++
+ if (flags & EXT3_SYNC_FL)
+ inode->i_flags |= S_SYNC;
+ if (flags & EXT3_APPEND_FL)
+ inode->i_flags |= S_APPEND;
+- if (flags & EXT3_IMMUTABLE_FL)
+- inode->i_flags |= S_IMMUTABLE;
+ if (flags & EXT3_NOATIME_FL)
+ inode->i_flags |= S_NOATIME;
+ if (flags & EXT3_DIRSYNC_FL)
+ inode->i_flags |= S_DIRSYNC;
++
++ inode->i_vflags &= ~(V_BARRIER | V_COW);
++
++ if (flags & EXT3_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ if (flags & EXT3_COW_FL)
++ inode->i_vflags |= V_COW;
+ }
+
+ /* Propagate flags from i_flags to EXT3_I(inode)->i_flags */
+ void ext3_get_inode_flags(struct ext3_inode_info *ei)
+ {
+ unsigned int flags = ei->vfs_inode.i_flags;
++ unsigned int vflags = ei->vfs_inode.i_vflags;
++
++ ei->i_flags &= ~(EXT3_SYNC_FL | EXT3_APPEND_FL |
++ EXT3_IMMUTABLE_FL | EXT3_IXUNLINK_FL |
++ EXT3_NOATIME_FL | EXT3_DIRSYNC_FL |
++ EXT3_BARRIER_FL | EXT3_COW_FL);
++
++ if (flags & S_IMMUTABLE)
++ ei->i_flags |= EXT3_IMMUTABLE_FL;
++ if (flags & S_IXUNLINK)
++ ei->i_flags |= EXT3_IXUNLINK_FL;
+
+- ei->i_flags &= ~(EXT3_SYNC_FL|EXT3_APPEND_FL|
+- EXT3_IMMUTABLE_FL|EXT3_NOATIME_FL|EXT3_DIRSYNC_FL);
+ if (flags & S_SYNC)
+ ei->i_flags |= EXT3_SYNC_FL;
+ if (flags & S_APPEND)
+ ei->i_flags |= EXT3_APPEND_FL;
+- if (flags & S_IMMUTABLE)
+- ei->i_flags |= EXT3_IMMUTABLE_FL;
+ if (flags & S_NOATIME)
+ ei->i_flags |= EXT3_NOATIME_FL;
+ if (flags & S_DIRSYNC)
+ ei->i_flags |= EXT3_DIRSYNC_FL;
++
++ if (vflags & V_BARRIER)
++ ei->i_flags |= EXT3_BARRIER_FL;
++ if (vflags & V_COW)
++ ei->i_flags |= EXT3_COW_FL;
++}
++
++int ext3_sync_flags(struct inode *inode)
++{
++ struct ext3_iloc iloc;
++ handle_t *handle;
++ int err;
++
++ handle = ext3_journal_start(inode, 1);
++ if (IS_ERR(handle))
++ return PTR_ERR(handle);
++ if (IS_SYNC(inode))
++ handle->h_sync = 1;
++ err = ext3_reserve_inode_write(handle, inode, &iloc);
++ if (err)
++ goto flags_err;
++
++ ext3_get_inode_flags(EXT3_I(inode));
++ inode->i_ctime = CURRENT_TIME;
++
++ err = ext3_mark_iloc_dirty(handle, inode, &iloc);
++flags_err:
++ ext3_journal_stop(handle);
++ return err;
+ }
+
+ struct inode *ext3_iget(struct super_block *sb, unsigned long ino)
+@@ -2652,6 +2701,8 @@ struct inode *ext3_iget(struct super_blo
+ struct inode *inode;
+ long ret;
+ int block;
++ uid_t uid;
++ gid_t gid;
+
+ inode = iget_locked(sb, ino);
+ if (!inode)
+@@ -2672,12 +2723,17 @@ struct inode *ext3_iget(struct super_blo
+ bh = iloc.bh;
+ raw_inode = ext3_raw_inode(&iloc);
+ inode->i_mode = le16_to_cpu(raw_inode->i_mode);
+- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
+- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
+ if(!(test_opt (inode->i_sb, NO_UID32))) {
+- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
+- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
+ }
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid,
++ le16_to_cpu(raw_inode->i_raw_tag));
++
+ inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
+ inode->i_size = le32_to_cpu(raw_inode->i_size);
+ inode->i_atime.tv_sec = (signed)le32_to_cpu(raw_inode->i_atime);
+@@ -2806,6 +2862,8 @@ static int ext3_do_update_inode(handle_t
+ struct ext3_inode *raw_inode = ext3_raw_inode(iloc);
+ struct ext3_inode_info *ei = EXT3_I(inode);
+ struct buffer_head *bh = iloc->bh;
++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag);
++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag);
+ int err = 0, rc, block;
+
+ /* For fields not not tracking in the in-memory inode,
+@@ -2816,29 +2874,32 @@ static int ext3_do_update_inode(handle_t
+ ext3_get_inode_flags(ei);
+ raw_inode->i_mode = cpu_to_le16(inode->i_mode);
+ if(!(test_opt(inode->i_sb, NO_UID32))) {
+- raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid));
+- raw_inode->i_gid_low = cpu_to_le16(low_16_bits(inode->i_gid));
++ raw_inode->i_uid_low = cpu_to_le16(low_16_bits(uid));
++ raw_inode->i_gid_low = cpu_to_le16(low_16_bits(gid));
+ /*
+ * Fix up interoperability with old kernels. Otherwise, old inodes get
+ * re-used with the upper 16 bits of the uid/gid intact
+ */
+ if(!ei->i_dtime) {
+ raw_inode->i_uid_high =
+- cpu_to_le16(high_16_bits(inode->i_uid));
++ cpu_to_le16(high_16_bits(uid));
+ raw_inode->i_gid_high =
+- cpu_to_le16(high_16_bits(inode->i_gid));
++ cpu_to_le16(high_16_bits(gid));
+ } else {
+ raw_inode->i_uid_high = 0;
+ raw_inode->i_gid_high = 0;
+ }
+ } else {
+ raw_inode->i_uid_low =
+- cpu_to_le16(fs_high2lowuid(inode->i_uid));
++ cpu_to_le16(fs_high2lowuid(uid));
+ raw_inode->i_gid_low =
+- cpu_to_le16(fs_high2lowgid(inode->i_gid));
++ cpu_to_le16(fs_high2lowgid(gid));
+ raw_inode->i_uid_high = 0;
+ raw_inode->i_gid_high = 0;
+ }
++#ifdef CONFIG_TAGGING_INTERN
++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag);
++#endif
+ raw_inode->i_links_count = cpu_to_le16(inode->i_nlink);
+ raw_inode->i_size = cpu_to_le32(ei->i_disksize);
+ raw_inode->i_atime = cpu_to_le32(inode->i_atime.tv_sec);
+@@ -2991,7 +3052,8 @@ int ext3_setattr(struct dentry *dentry,
+ return error;
+
+ if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) ||
++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) {
+ handle_t *handle;
+
+ /* (user+group)*(old+new) structure, inode write (sb,
+@@ -3013,6 +3075,8 @@ int ext3_setattr(struct dentry *dentry,
+ inode->i_uid = attr->ia_uid;
+ if (attr->ia_valid & ATTR_GID)
+ inode->i_gid = attr->ia_gid;
++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode))
++ inode->i_tag = attr->ia_tag;
+ error = ext3_mark_inode_dirty(handle, inode);
+ ext3_journal_stop(handle);
+ }
+--- a/fs/ext3/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -8,6 +8,7 @@
+ */
+
+ #include <linux/fs.h>
++#include <linux/mount.h>
+ #include <linux/jbd.h>
+ #include <linux/capability.h>
+ #include <linux/ext3_fs.h>
+@@ -16,6 +17,7 @@
+ #include <linux/time.h>
+ #include <linux/compat.h>
+ #include <linux/smp_lock.h>
++#include <linux/vs_tag.h>
+ #include <asm/uaccess.h>
+
+ int ext3_ioctl (struct inode * inode, struct file * filp, unsigned int cmd,
+@@ -56,6 +58,11 @@ int ext3_ioctl (struct inode * inode, st
+ if (!S_ISDIR(inode->i_mode))
+ flags &= ~EXT3_DIRSYNC_FL;
+
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -EACCES;
++ }
++
+ mutex_lock(&inode->i_mutex);
+ /* Is it quota file? Do not allow user to mess with it */
+ if (IS_NOQUOTA(inode)) {
+@@ -74,7 +81,9 @@ int ext3_ioctl (struct inode * inode, st
+ *
+ * This test looks nicer. Thanks to Pauline Middelink
+ */
+- if ((flags ^ oldflags) & (EXT3_APPEND_FL | EXT3_IMMUTABLE_FL)) {
++ if ((oldflags & EXT3_IMMUTABLE_FL) ||
++ ((flags ^ oldflags) & (EXT3_APPEND_FL |
++ EXT3_IMMUTABLE_FL | EXT3_IXUNLINK_FL))) {
+ if (!capable(CAP_LINUX_IMMUTABLE)) {
+ mutex_unlock(&inode->i_mutex);
+ err = -EPERM;
+--- a/fs/ext3/namei.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/namei.c 2008-07-16 22:41:36.000000000 -0400
+@@ -36,6 +36,7 @@
+ #include <linux/quotaops.h>
+ #include <linux/buffer_head.h>
+ #include <linux/bio.h>
++#include <linux/vs_tag.h>
+
+ #include "namei.h"
+ #include "xattr.h"
+@@ -912,6 +913,7 @@ restart:
+ if (bh)
+ ll_rw_block(READ_META, 1, &bh);
+ }
++ dx_propagate_tag(nd, inode);
+ }
+ if ((bh = bh_use[ra_ptr++]) == NULL)
+ goto next;
+@@ -2424,6 +2426,7 @@ const struct inode_operations ext3_dir_i
+ .removexattr = generic_removexattr,
+ #endif
+ .permission = ext3_permission,
++ .sync_flags = ext3_sync_flags,
+ };
+
+ const struct inode_operations ext3_special_inode_operations = {
+@@ -2435,4 +2438,5 @@ const struct inode_operations ext3_speci
+ .removexattr = generic_removexattr,
+ #endif
+ .permission = ext3_permission,
++ .sync_flags = ext3_sync_flags,
+ };
+--- a/fs/ext3/super.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/super.c 2008-07-17 17:20:52.000000000 -0400
+@@ -757,7 +757,7 @@ enum {
+ Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
+ Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
+ Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
+- Opt_grpquota
++ Opt_grpquota, Opt_tag, Opt_notag, Opt_tagid
+ };
+
+ static match_table_t tokens = {
+@@ -808,6 +808,9 @@ static match_table_t tokens = {
+ {Opt_usrquota, "usrquota"},
+ {Opt_barrier, "barrier=%u"},
+ {Opt_resize, "resize"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
+ {Opt_err, NULL},
+ };
+
+@@ -900,6 +903,20 @@ static int parse_options (char *options,
+ case Opt_nouid32:
+ set_opt (sbi->s_mount_opt, NO_UID32);
+ break;
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++ case Opt_notag:
++ clear_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
+ case Opt_nocheck:
+ clear_opt (sbi->s_mount_opt, CHECK);
+ break;
+@@ -1594,6 +1611,9 @@ static int ext3_fill_super (struct super
+ NULL, 0))
+ goto failed_mount;
+
++ if (EXT3_SB(sb)->s_mount_opt & EXT3_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED;
++
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((sbi->s_mount_opt & EXT3_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
+@@ -2432,6 +2452,13 @@ static int ext3_remount (struct super_bl
+ if (sbi->s_mount_opt & EXT3_MOUNT_ABORT)
+ ext3_abort(sb, __func__, "Abort forced by user");
+
++ if ((sbi->s_mount_opt & EXT3_MOUNT_TAGGED) &&
++ !(sb->s_flags & MS_TAGGED)) {
++ printk("EXT3-fs: %s: tagging not permitted on remount.\n",
++ sb->s_id);
++ return -EINVAL;
++ }
++
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((sbi->s_mount_opt & EXT3_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
+--- a/fs/ext3/symlink.c 2008-04-17 10:32:27.000000000 -0400
++++ a/fs/ext3/symlink.c 2008-07-16 22:41:36.000000000 -0400
+@@ -40,6 +40,7 @@ const struct inode_operations ext3_symli
+ .listxattr = ext3_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext3_sync_flags,
+ };
+
+ const struct inode_operations ext3_fast_symlink_inode_operations = {
+@@ -51,4 +52,5 @@ const struct inode_operations ext3_fast_
+ .listxattr = ext3_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext3_sync_flags,
+ };
+--- a/fs/ext3/xattr.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext3/xattr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -58,6 +58,7 @@
+ #include <linux/mbcache.h>
+ #include <linux/quotaops.h>
+ #include <linux/rwsem.h>
++#include <linux/vs_dlimit.h>
+ #include "xattr.h"
+ #include "acl.h"
+
+@@ -498,6 +499,7 @@ ext3_xattr_release_block(handle_t *handl
+ error = ext3_journal_dirty_metadata(handle, bh);
+ if (IS_SYNC(inode))
+ handle->h_sync = 1;
++ DLIMIT_FREE_BLOCK(inode, 1);
+ DQUOT_FREE_BLOCK(inode, 1);
+ ea_bdebug(bh, "refcount now=%d; releasing",
+ le32_to_cpu(BHDR(bh)->h_refcount));
+@@ -771,11 +773,14 @@ inserted:
+ if (new_bh == bs->bh)
+ ea_bdebug(new_bh, "keeping");
+ else {
++ error = -ENOSPC;
++ if (DLIMIT_ALLOC_BLOCK(inode, 1))
++ goto cleanup;
+ /* The old block is released after updating
+ the inode. */
+ error = -EDQUOT;
+ if (DQUOT_ALLOC_BLOCK(inode, 1))
+- goto cleanup;
++ goto cleanup_dlimit;
+ error = ext3_journal_get_write_access(handle,
+ new_bh);
+ if (error)
+@@ -849,6 +854,8 @@ cleanup:
+
+ cleanup_dquot:
+ DQUOT_FREE_BLOCK(inode, 1);
++cleanup_dlimit:
++ DLIMIT_FREE_BLOCK(inode, 1);
+ goto cleanup;
+
+ bad_block:
+--- a/fs/ext4/balloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/balloc.c 2008-07-17 20:27:00.000000000 -0400
+@@ -17,6 +17,8 @@
+ #include <linux/jbd2.h>
+ #include <linux/quotaops.h>
+ #include <linux/buffer_head.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+ #include "ext4.h"
+ #include "ext4_jbd2.h"
+ #include "group.h"
+@@ -861,8 +863,10 @@ void ext4_free_blocks(handle_t *handle,
+ else
+ ext4_mb_free_blocks(handle, inode, block, count,
+ metadata, &dquot_freed_blocks);
+- if (dquot_freed_blocks)
++ if (dquot_freed_blocks) {
++ DLIMIT_FREE_BLOCK(inode, dquot_freed_blocks);
+ DQUOT_FREE_BLOCK(inode, dquot_freed_blocks);
++ }
+ return;
+ }
+
+@@ -1602,18 +1606,33 @@ out:
+ *
+ * Check if filesystem has at least 1 free block available for allocation.
+ */
+-static int ext4_has_free_blocks(struct ext4_sb_info *sbi)
++static int ext4_has_free_blocks(struct super_block *sb)
+ {
++ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ ext4_fsblk_t free_blocks, root_blocks;
++ int cond;
+
+ free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter);
+ root_blocks = ext4_r_blocks_count(sbi->s_es);
+- if (free_blocks < root_blocks + 1 && !capable(CAP_SYS_RESOURCE) &&
++
++ vxdprintk(VXD_CBIT(dlim, 3),
++ "ext4_has_free_blocks(%p): free=%llu, root=%llu",
++ sb, free_blocks, root_blocks);
++
++ DLIMIT_ADJUST_BLOCK(sb, dx_current_tag(), &free_blocks, &root_blocks);
++
++ cond = (free_blocks < root_blocks + 1 &&
++ !capable(CAP_SYS_RESOURCE) &&
+ sbi->s_resuid != current->fsuid &&
+- (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid))) {
+- return 0;
+- }
+- return 1;
++ (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid)));
++
++ vxdprintk(VXD_CBIT(dlim, 3),
++ "ext4_has_free_blocks(%p): %llu<%llu+1, %c, %u!=%u r=%d",
++ sb, free_blocks, root_blocks,
++ !capable(CAP_SYS_RESOURCE)?'1':'0',
++ sbi->s_resuid, current->fsuid, cond?0:1);
++
++ return (cond ? 0 : 1);
+ }
+
+ /**
+@@ -1630,7 +1649,7 @@ static int ext4_has_free_blocks(struct e
+ */
+ int ext4_should_retry_alloc(struct super_block *sb, int *retries)
+ {
+- if (!ext4_has_free_blocks(EXT4_SB(sb)) || (*retries)++ > 3)
++ if (!ext4_has_free_blocks(sb) || (*retries)++ > 3)
+ return 0;
+
+ jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id);
+@@ -1690,6 +1709,8 @@ ext4_fsblk_t ext4_new_blocks_old(handle_
+ *errp = -EDQUOT;
+ return 0;
+ }
++ if (DLIMIT_ALLOC_BLOCK(inode, num))
++ goto out_dlimit;
+
+ sbi = EXT4_SB(sb);
+ es = EXT4_SB(sb)->s_es;
+@@ -1706,7 +1727,7 @@ ext4_fsblk_t ext4_new_blocks_old(handle_
+ if (block_i && ((windowsz = block_i->rsv_window_node.rsv_goal_size) > 0))
+ my_rsv = &block_i->rsv_window_node;
+
+- if (!ext4_has_free_blocks(sbi)) {
++ if (!ext4_has_free_blocks(sb)) {
+ *errp = -ENOSPC;
+ goto out;
+ }
+@@ -1896,12 +1917,16 @@ allocated:
+ *errp = 0;
+ brelse(bitmap_bh);
+ DQUOT_FREE_BLOCK(inode, *count-num);
++ DLIMIT_FREE_BLOCK(inode, *count-num);
+ *count = num;
+ return ret_block;
+
+ io_error:
+ *errp = -EIO;
+ out:
++ if (!performed_allocation)
++ DLIMIT_FREE_BLOCK(inode, *count);
++out_dlimit:
+ if (fatal) {
+ *errp = fatal;
+ ext4_std_error(sb, fatal);
+--- a/fs/ext4/ext4.h 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/ext4.h 2008-08-11 21:57:27.000000000 -0400
+@@ -230,8 +230,12 @@ struct ext4_group_desc
+ #define EXT4_HUGE_FILE_FL 0x00040000 /* Set to each huge file */
+ #define EXT4_EXTENTS_FL 0x00080000 /* Inode uses extents */
+ #define EXT4_EXT_MIGRATE 0x00100000 /* Inode is migrating */
++#define EXT4_IXUNLINK_FL 0x01000000 /* Immutable invert on unlink */
+ #define EXT4_RESERVED_FL 0x80000000 /* reserved for ext4 lib */
+
++#define EXT4_BARRIER_FL 0x10000000 /* Barrier for chroot() */
++#define EXT4_COW_FL 0x20000000 /* Copy on Write marker */
++
+ #define EXT4_FL_USER_VISIBLE 0x000BDFFF /* User visible flags */
+ #define EXT4_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */
+
+@@ -527,6 +531,8 @@ do { \
+ #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT 0x1000000 /* Journal Async Commit */
+ #define EXT4_MOUNT_I_VERSION 0x2000000 /* i_version support */
+ #define EXT4_MOUNT_MBALLOC 0x4000000 /* Buddy allocation support */
++#define EXT4_MOUNT_TAGGED (1<<24) /* Enable Context Tags */
++
+ /* Compatibility, for having both ext2_fs.h and ext4_fs.h included at once */
+ #ifndef _LINUX_EXT2_FS_H
+ #define clear_opt(o, opt) o &= ~EXT4_MOUNT_##opt
+@@ -1029,6 +1035,7 @@ int ext4_get_blocks_handle(handle_t *han
+ ext4_lblk_t iblock, unsigned long maxblocks,
+ struct buffer_head *bh_result,
+ int create, int extend_disksize);
++extern int ext4_sync_flags(struct inode *inode);
+
+ extern struct inode *ext4_iget(struct super_block *, unsigned long);
+ extern int ext4_write_inode (struct inode *, int);
+--- a/fs/ext4/file.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/file.c 2008-07-29 18:35:50.000000000 -0400
+@@ -152,5 +152,6 @@ const struct inode_operations ext4_file_
+ #endif
+ .permission = ext4_permission,
+ .fallocate = ext4_fallocate,
++ .sync_flags = ext4_sync_flags,
+ };
+
+--- a/fs/ext4/ialloc.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/ialloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -22,6 +22,8 @@
+ #include <linux/random.h>
+ #include <linux/bitops.h>
+ #include <linux/blkdev.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+ #include <asm/byteorder.h>
+ #include "ext4.h"
+ #include "ext4_jbd2.h"
+@@ -185,6 +187,7 @@ void ext4_free_inode (handle_t *handle,
+ ext4_xattr_delete_inode(handle, inode);
+ DQUOT_FREE_INODE(inode);
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+
+ is_directory = S_ISDIR(inode->i_mode);
+
+@@ -510,6 +513,12 @@ struct inode *ext4_new_inode(handle_t *h
+ inode = new_inode(sb);
+ if (!inode)
+ return ERR_PTR(-ENOMEM);
++
++ inode->i_tag = dx_current_fstag(sb);
++ if (DLIMIT_ALLOC_INODE(inode)) {
++ err = -ENOSPC;
++ goto out_dlimit;
++ }
+ ei = EXT4_I(inode);
+
+ sbi = EXT4_SB(sb);
+@@ -702,7 +711,8 @@ got:
+ * newly created directory and file only if -o extent mount option is
+ * specified
+ */
+- ei->i_flags = EXT4_I(dir)->i_flags & ~(EXT4_INDEX_FL|EXT4_EXTENTS_FL);
++ ei->i_flags = EXT4_I(dir)->i_flags &
++ ~(EXT4_INDEX_FL|EXT4_EXTENTS_FL|EXT4_IXUNLINK_FL|EXT4_BARRIER_FL);
+ if (S_ISLNK(mode))
+ ei->i_flags &= ~(EXT4_IMMUTABLE_FL|EXT4_APPEND_FL);
+ /* dirsync only applies to directories */
+@@ -762,6 +772,8 @@ got:
+ fail:
+ ext4_std_error(sb, err);
+ out:
++ DLIMIT_FREE_INODE(inode);
++out_dlimit:
+ iput(inode);
+ ret = ERR_PTR(err);
+ really_out:
+@@ -773,6 +785,7 @@ fail_free_drop:
+
+ fail_drop:
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+ inode->i_flags |= S_NOQUOTA;
+ inode->i_nlink = 0;
+ iput(inode);
+--- a/fs/ext4/inode.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/inode.c 2008-07-17 20:26:11.000000000 -0400
+@@ -35,6 +35,7 @@
+ #include <linux/mpage.h>
+ #include <linux/uio.h>
+ #include <linux/bio.h>
++#include <linux/vs_tag.h>
+ #include "ext4_jbd2.h"
+ #include "xattr.h"
+ #include "acl.h"
+@@ -2354,7 +2355,7 @@ void ext4_truncate(struct inode *inode)
+ return;
+ if (ext4_inode_is_fast_symlink(inode))
+ return;
+- if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode))
+ return;
+
+ /*
+@@ -2669,37 +2670,86 @@ void ext4_set_inode_flags(struct inode *
+ {
+ unsigned int flags = EXT4_I(inode)->i_flags;
+
+- inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK |
++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
++
++ if (flags & EXT4_IMMUTABLE_FL)
++ inode->i_flags |= S_IMMUTABLE;
++ if (flags & EXT4_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
++
+ if (flags & EXT4_SYNC_FL)
+ inode->i_flags |= S_SYNC;
+ if (flags & EXT4_APPEND_FL)
+ inode->i_flags |= S_APPEND;
+- if (flags & EXT4_IMMUTABLE_FL)
+- inode->i_flags |= S_IMMUTABLE;
+ if (flags & EXT4_NOATIME_FL)
+ inode->i_flags |= S_NOATIME;
+ if (flags & EXT4_DIRSYNC_FL)
+ inode->i_flags |= S_DIRSYNC;
++
++ inode->i_vflags &= ~(V_BARRIER | V_COW);
++
++ if (flags & EXT4_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ if (flags & EXT4_COW_FL)
++ inode->i_vflags |= V_COW;
+ }
+
+ /* Propagate flags from i_flags to EXT4_I(inode)->i_flags */
+ void ext4_get_inode_flags(struct ext4_inode_info *ei)
+ {
+ unsigned int flags = ei->vfs_inode.i_flags;
++ unsigned int vflags = ei->vfs_inode.i_vflags;
++
++ ei->i_flags &= ~(EXT4_SYNC_FL | EXT4_APPEND_FL |
++ EXT4_IMMUTABLE_FL | EXT4_IXUNLINK_FL |
++ EXT4_NOATIME_FL | EXT4_DIRSYNC_FL |
++ EXT4_BARRIER_FL | EXT4_COW_FL);
++
++ if (flags & S_IMMUTABLE)
++ ei->i_flags |= EXT4_IMMUTABLE_FL;
++ if (flags & S_IXUNLINK)
++ ei->i_flags |= EXT4_IXUNLINK_FL;
+
+- ei->i_flags &= ~(EXT4_SYNC_FL|EXT4_APPEND_FL|
+- EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL|EXT4_DIRSYNC_FL);
+ if (flags & S_SYNC)
+ ei->i_flags |= EXT4_SYNC_FL;
+ if (flags & S_APPEND)
+ ei->i_flags |= EXT4_APPEND_FL;
+- if (flags & S_IMMUTABLE)
+- ei->i_flags |= EXT4_IMMUTABLE_FL;
+ if (flags & S_NOATIME)
+ ei->i_flags |= EXT4_NOATIME_FL;
+ if (flags & S_DIRSYNC)
+ ei->i_flags |= EXT4_DIRSYNC_FL;
++
++ if (vflags & V_BARRIER)
++ ei->i_flags |= EXT4_BARRIER_FL;
++ if (vflags & V_COW)
++ ei->i_flags |= EXT4_COW_FL;
+ }
++
++int ext4_sync_flags(struct inode *inode)
++{
++ struct ext4_iloc iloc;
++ handle_t *handle;
++ int err;
++
++ handle = ext4_journal_start(inode, 1);
++ if (IS_ERR(handle))
++ return PTR_ERR(handle);
++ if (IS_SYNC(inode))
++ handle->h_sync = 1;
++ err = ext4_reserve_inode_write(handle, inode, &iloc);
++ if (err)
++ goto flags_err;
++
++ ext4_get_inode_flags(EXT4_I(inode));
++ inode->i_ctime = CURRENT_TIME;
++
++ err = ext4_mark_iloc_dirty(handle, inode, &iloc);
++flags_err:
++ ext4_journal_stop(handle);
++ return err;
++}
++
+ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode,
+ struct ext4_inode_info *ei)
+ {
+@@ -2732,6 +2782,8 @@ struct inode *ext4_iget(struct super_blo
+ struct inode *inode;
+ long ret;
+ int block;
++ uid_t uid;
++ gid_t gid;
+
+ inode = iget_locked(sb, ino);
+ if (!inode)
+@@ -2752,12 +2804,17 @@ struct inode *ext4_iget(struct super_blo
+ bh = iloc.bh;
+ raw_inode = ext4_raw_inode(&iloc);
+ inode->i_mode = le16_to_cpu(raw_inode->i_mode);
+- inode->i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
+- inode->i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
++ uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
++ gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
+ if(!(test_opt (inode->i_sb, NO_UID32))) {
+- inode->i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
+- inode->i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
++ uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
++ gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
+ }
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid,
++ le16_to_cpu(raw_inode->i_raw_tag));
++
+ inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
+
+ ei->i_state = 0;
+@@ -2931,6 +2988,8 @@ static int ext4_do_update_inode(handle_t
+ struct ext4_inode *raw_inode = ext4_raw_inode(iloc);
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ struct buffer_head *bh = iloc->bh;
++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag);
++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag);
+ int err = 0, rc, block;
+
+ /* For fields not not tracking in the in-memory inode,
+@@ -2941,29 +3000,32 @@ static int ext4_do_update_inode(handle_t
+ ext4_get_inode_flags(ei);
+ raw_inode->i_mode = cpu_to_le16(inode->i_mode);
+ if(!(test_opt(inode->i_sb, NO_UID32))) {
+- raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid));
+- raw_inode->i_gid_low = cpu_to_le16(low_16_bits(inode->i_gid));
++ raw_inode->i_uid_low = cpu_to_le16(low_16_bits(uid));
++ raw_inode->i_gid_low = cpu_to_le16(low_16_bits(gid));
+ /*
+ * Fix up interoperability with old kernels. Otherwise, old inodes get
+ * re-used with the upper 16 bits of the uid/gid intact
+ */
+ if(!ei->i_dtime) {
+ raw_inode->i_uid_high =
+- cpu_to_le16(high_16_bits(inode->i_uid));
++ cpu_to_le16(high_16_bits(uid));
+ raw_inode->i_gid_high =
+- cpu_to_le16(high_16_bits(inode->i_gid));
++ cpu_to_le16(high_16_bits(gid));
+ } else {
+ raw_inode->i_uid_high = 0;
+ raw_inode->i_gid_high = 0;
+ }
+ } else {
+ raw_inode->i_uid_low =
+- cpu_to_le16(fs_high2lowuid(inode->i_uid));
++ cpu_to_le16(fs_high2lowuid(uid));
+ raw_inode->i_gid_low =
+- cpu_to_le16(fs_high2lowgid(inode->i_gid));
++ cpu_to_le16(fs_high2lowgid(gid));
+ raw_inode->i_uid_high = 0;
+ raw_inode->i_gid_high = 0;
+ }
++#ifdef CONFIG_TAGGING_INTERN
++ raw_inode->i_raw_tag = cpu_to_le16(inode->i_tag);
++#endif
+ raw_inode->i_links_count = cpu_to_le16(inode->i_nlink);
+
+ EXT4_INODE_SET_XTIME(i_ctime, inode, raw_inode);
+@@ -3120,7 +3182,8 @@ int ext4_setattr(struct dentry *dentry,
+ return error;
+
+ if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) ||
++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) {
+ handle_t *handle;
+
+ /* (user+group)*(old+new) structure, inode write (sb,
+@@ -3142,6 +3205,8 @@ int ext4_setattr(struct dentry *dentry,
+ inode->i_uid = attr->ia_uid;
+ if (attr->ia_valid & ATTR_GID)
+ inode->i_gid = attr->ia_gid;
++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode))
++ inode->i_tag = attr->ia_tag;
+ error = ext4_mark_inode_dirty(handle, inode);
+ ext4_journal_stop(handle);
+ }
+--- a/fs/ext4/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/ioctl.c 2008-07-17 20:58:51.000000000 -0400
+@@ -8,12 +8,14 @@
+ */
+
+ #include <linux/fs.h>
++#include <linux/mount.h>
+ #include <linux/jbd2.h>
+ #include <linux/capability.h>
+ #include <linux/time.h>
+ #include <linux/compat.h>
+ #include <linux/smp_lock.h>
+ #include <linux/mount.h>
++#include <linux/vs_tag.h>
+ #include <asm/uaccess.h>
+ #include "ext4_jbd2.h"
+ #include "ext4.h"
+@@ -52,6 +54,11 @@ long ext4_ioctl(struct file *filp, unsig
+ if (!S_ISDIR(inode->i_mode))
+ flags &= ~EXT4_DIRSYNC_FL;
+
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -EACCES;
++ }
++
+ err = -EPERM;
+ mutex_lock(&inode->i_mutex);
+ /* Is it quota file? Do not allow user to mess with it */
+@@ -69,7 +76,9 @@ long ext4_ioctl(struct file *filp, unsig
+ *
+ * This test looks nicer. Thanks to Pauline Middelink
+ */
+- if ((flags ^ oldflags) & (EXT4_APPEND_FL | EXT4_IMMUTABLE_FL)) {
++ if ((oldflags & EXT4_IMMUTABLE_FL) ||
++ ((flags ^ oldflags) & (EXT4_APPEND_FL |
++ EXT4_IMMUTABLE_FL | EXT4_IXUNLINK_FL))) {
+ if (!capable(CAP_LINUX_IMMUTABLE))
+ goto flags_out;
+ }
+--- a/fs/ext4/namei.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/namei.c 2008-07-17 20:24:56.000000000 -0400
+@@ -34,6 +34,7 @@
+ #include <linux/quotaops.h>
+ #include <linux/buffer_head.h>
+ #include <linux/bio.h>
++#include <linux/vs_tag.h>
+ #include "ext4.h"
+ #include "ext4_jbd2.h"
+
+@@ -913,6 +914,7 @@ restart:
+ if (bh)
+ ll_rw_block(READ_META, 1, &bh);
+ }
++ dx_propagate_tag(nd, inode);
+ }
+ if ((bh = bh_use[ra_ptr++]) == NULL)
+ goto next;
+@@ -2458,6 +2460,7 @@ const struct inode_operations ext4_dir_i
+ .removexattr = generic_removexattr,
+ #endif
+ .permission = ext4_permission,
++ .sync_flags = ext4_sync_flags,
+ };
+
+ const struct inode_operations ext4_special_inode_operations = {
+@@ -2469,4 +2472,5 @@ const struct inode_operations ext4_speci
+ .removexattr = generic_removexattr,
+ #endif
+ .permission = ext4_permission,
++ .sync_flags = ext4_sync_flags,
+ };
+--- a/fs/ext4/super.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/super.c 2008-07-17 21:00:22.000000000 -0400
+@@ -895,6 +895,7 @@ enum {
+ Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
+ Opt_grpquota, Opt_extents, Opt_noextents, Opt_i_version,
+ Opt_mballoc, Opt_nomballoc, Opt_stripe,
++ Opt_tag, Opt_notag, Opt_tagid
+ };
+
+ static match_table_t tokens = {
+@@ -953,6 +954,9 @@ static match_table_t tokens = {
+ {Opt_nomballoc, "nomballoc"},
+ {Opt_stripe, "stripe=%u"},
+ {Opt_resize, "resize"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
+ {Opt_err, NULL},
+ };
+
+@@ -1045,6 +1049,20 @@ static int parse_options (char *options,
+ case Opt_nouid32:
+ set_opt (sbi->s_mount_opt, NO_UID32);
+ break;
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++ case Opt_notag:
++ clear_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ set_opt (sbi->s_mount_opt, TAGGED);
++ break;
++#endif
+ case Opt_nocheck:
+ clear_opt (sbi->s_mount_opt, CHECK);
+ break;
+@@ -1932,6 +1950,9 @@ static int ext4_fill_super (struct super
+ NULL, 0))
+ goto failed_mount;
+
++ if (EXT4_SB(sb)->s_mount_opt & EXT4_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED;
++
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
+@@ -2875,6 +2896,13 @@ static int ext4_remount (struct super_bl
+ if (sbi->s_mount_opt & EXT4_MOUNT_ABORT)
+ ext4_abort(sb, __func__, "Abort forced by user");
+
++ if ((sbi->s_mount_opt & EXT4_MOUNT_TAGGED) &&
++ !(sb->s_flags & MS_TAGGED)) {
++ printk("EXT4-fs: %s: tagging not permitted on remount.\n",
++ sb->s_id);
++ return -EINVAL;
++ }
++
+ sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+ ((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
+--- a/fs/ext4/symlink.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/symlink.c 2008-07-16 22:41:36.000000000 -0400
+@@ -40,6 +40,7 @@ const struct inode_operations ext4_symli
+ .listxattr = ext4_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext4_sync_flags,
+ };
+
+ const struct inode_operations ext4_fast_symlink_inode_operations = {
+@@ -51,4 +52,5 @@ const struct inode_operations ext4_fast_
+ .listxattr = ext4_listxattr,
+ .removexattr = generic_removexattr,
+ #endif
++ .sync_flags = ext4_sync_flags,
+ };
+--- a/fs/ext4/xattr.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ext4/xattr.c 2008-07-17 20:24:11.000000000 -0400
+@@ -56,6 +56,7 @@
+ #include <linux/mbcache.h>
+ #include <linux/quotaops.h>
+ #include <linux/rwsem.h>
++#include <linux/vs_dlimit.h>
+ #include "ext4_jbd2.h"
+ #include "ext4.h"
+ #include "xattr.h"
+@@ -490,6 +491,7 @@ ext4_xattr_release_block(handle_t *handl
+ error = ext4_journal_dirty_metadata(handle, bh);
+ if (IS_SYNC(inode))
+ handle->h_sync = 1;
++ DLIMIT_FREE_BLOCK(inode, 1);
+ DQUOT_FREE_BLOCK(inode, 1);
+ ea_bdebug(bh, "refcount now=%d; releasing",
+ le32_to_cpu(BHDR(bh)->h_refcount));
+@@ -780,11 +782,14 @@ inserted:
+ if (new_bh == bs->bh)
+ ea_bdebug(new_bh, "keeping");
+ else {
++ error = -ENOSPC;
++ if (DLIMIT_ALLOC_BLOCK(inode, 1))
++ goto cleanup;
+ /* The old block is released after updating
+ the inode. */
+ error = -EDQUOT;
+ if (DQUOT_ALLOC_BLOCK(inode, 1))
+- goto cleanup;
++ goto cleanup_dlimit;
+ error = ext4_journal_get_write_access(handle,
+ new_bh);
+ if (error)
+@@ -858,6 +863,8 @@ cleanup:
+
+ cleanup_dquot:
+ DQUOT_FREE_BLOCK(inode, 1);
++cleanup_dlimit:
++ DLIMIT_FREE_BLOCK(inode, 1);
+ goto cleanup;
+
+ bad_block:
+--- a/fs/fcntl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/fcntl.c 2008-07-17 17:22:30.000000000 -0400
+@@ -20,6 +20,7 @@
+ #include <linux/signal.h>
+ #include <linux/rcupdate.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_limit.h>
+
+ #include <asm/poll.h>
+ #include <asm/siginfo.h>
+@@ -88,6 +89,8 @@ repeat:
+ error = -EMFILE;
+ if (newfd >= current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
+ goto out;
++ if (!vx_files_avail(1))
++ goto out;
+
+ error = expand_files(files, newfd);
+ if (error < 0)
+@@ -118,9 +121,10 @@ out:
+ static int dupfd(struct file *file, unsigned int start, int cloexec)
+ {
+ int fd = locate_fd(start, cloexec);
+- if (fd >= 0)
++ if (fd >= 0) {
+ fd_install(fd, file);
+- else
++ vx_openfd_inc(fd);
++ } else
+ fput(file);
+
+ return fd;
+@@ -169,6 +173,9 @@ asmlinkage long sys_dup2(unsigned int ol
+
+ if (tofree)
+ filp_close(tofree, files);
++ else
++ vx_openfd_inc(newfd); /* fd was unused */
++
+ err = newfd;
+ out:
+ return err;
+--- a/fs/file.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/file.c 2008-07-27 14:14:05.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/spinlock.h>
+ #include <linux/rcupdate.h>
+ #include <linux/workqueue.h>
++#include <linux/vs_limit.h>
+
+ struct fdtable_defer {
+ spinlock_t lock;
+@@ -357,6 +358,8 @@ struct files_struct *dup_fd(struct files
+ struct file *f = *old_fds++;
+ if (f) {
+ get_file(f);
++ /* TODO: sum it first for check and performance */
++ vx_openfd_inc(open_files - i);
+ } else {
+ /*
+ * The fd may be claimed in the fd bitmap but not yet
+--- a/fs/file_table.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/file_table.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,8 @@
+ #include <linux/fsnotify.h>
+ #include <linux/sysctl.h>
+ #include <linux/percpu_counter.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_context.h>
+
+ #include <asm/atomic.h>
+
+@@ -126,6 +128,8 @@ struct file *get_empty_filp(void)
+ f->f_gid = tsk->fsgid;
+ eventpoll_init_file(f);
+ /* f->f_version: 0 */
++ f->f_xid = vx_current_xid();
++ vx_files_inc(f);
+ return f;
+
+ over:
+@@ -276,6 +280,8 @@ void __fput(struct file *file)
+ cdev_put(inode->i_cdev);
+ fops_put(file->f_op);
+ put_pid(file->f_owner.pid);
++ vx_files_dec(file);
++ file->f_xid = 0;
+ file_kill(file);
+ if (file->f_mode & FMODE_WRITE)
+ drop_file_write_access(file);
+@@ -343,6 +349,8 @@ void put_filp(struct file *file)
+ {
+ if (atomic_dec_and_test(&file->f_count)) {
+ security_file_free(file);
++ vx_files_dec(file);
++ file->f_xid = 0;
+ file_kill(file);
+ file_free(file);
+ }
+--- a/fs/hfsplus/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/hfsplus/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -17,6 +17,7 @@
+ #include <linux/mount.h>
+ #include <linux/sched.h>
+ #include <linux/xattr.h>
++#include <linux/mount.h>
+ #include <asm/uaccess.h>
+ #include "hfsplus_fs.h"
+
+--- a/fs/inode.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -124,6 +124,9 @@ static struct inode *alloc_inode(struct
+ struct address_space * const mapping = &inode->i_data;
+
+ inode->i_sb = sb;
++
++ /* essential because of inode slab reuse */
++ inode->i_tag = 0;
+ inode->i_blkbits = sb->s_blocksize_bits;
+ inode->i_flags = 0;
+ atomic_set(&inode->i_count, 1);
+@@ -142,6 +145,7 @@ static struct inode *alloc_inode(struct
+ inode->i_bdev = NULL;
+ inode->i_cdev = NULL;
+ inode->i_rdev = 0;
++ inode->i_mdev = 0;
+ inode->dirtied_when = 0;
+ if (security_inode_alloc(inode)) {
+ if (inode->i_sb->s_op->destroy_inode)
+@@ -246,6 +250,8 @@ void __iget(struct inode * inode)
+ inodes_stat.nr_unused--;
+ }
+
++EXPORT_SYMBOL_GPL(__iget);
++
+ /**
+ * clear_inode - clear an inode
+ * @inode: inode to clear
+@@ -1426,9 +1432,11 @@ void init_special_inode(struct inode *in
+ if (S_ISCHR(mode)) {
+ inode->i_fop = &def_chr_fops;
+ inode->i_rdev = rdev;
++ inode->i_mdev = rdev;
+ } else if (S_ISBLK(mode)) {
+ inode->i_fop = &def_blk_fops;
+ inode->i_rdev = rdev;
++ inode->i_mdev = rdev;
+ } else if (S_ISFIFO(mode))
+ inode->i_fop = &def_fifo_fops;
+ else if (S_ISSOCK(mode))
+--- a/fs/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -13,6 +13,9 @@
+ #include <linux/security.h>
+ #include <linux/module.h>
+ #include <linux/uaccess.h>
++#include <linux/proc_fs.h>
++#include <linux/vserver/inode.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/ioctls.h>
+
+--- a/fs/ioprio.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ioprio.c 2008-07-16 22:41:36.000000000 -0400
+@@ -26,6 +26,7 @@
+ #include <linux/syscalls.h>
+ #include <linux/security.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_base.h>
+
+ static int set_task_ioprio(struct task_struct *task, int ioprio)
+ {
+@@ -116,6 +117,8 @@ asmlinkage long sys_ioprio_set(int which
+ else
+ pgrp = find_vpid(who);
+ do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT))
++ continue;
+ ret = set_task_ioprio(p, ioprio);
+ if (ret)
+ break;
+@@ -205,6 +208,8 @@ asmlinkage long sys_ioprio_get(int which
+ else
+ pgrp = find_vpid(who);
+ do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT))
++ continue;
+ tmpio = get_task_ioprio(p);
+ if (tmpio < 0)
+ continue;
+--- a/fs/jfs/acl.c 2007-02-04 13:44:54.000000000 -0500
++++ a/fs/jfs/acl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -232,7 +232,8 @@ int jfs_setattr(struct dentry *dentry, s
+ return rc;
+
+ if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
+- (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
++ (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) ||
++ (iattr->ia_valid & ATTR_TAG && iattr->ia_tag != inode->i_tag)) {
+ if (DQUOT_TRANSFER(inode, iattr))
+ return -EDQUOT;
+ }
+--- a/fs/jfs/file.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/file.c 2008-07-16 22:41:36.000000000 -0400
+@@ -98,6 +98,7 @@ const struct inode_operations jfs_file_i
+ .setattr = jfs_setattr,
+ .permission = jfs_permission,
+ #endif
++ .sync_flags = jfs_sync_flags,
+ };
+
+ const struct file_operations jfs_file_operations = {
+--- a/fs/jfs/inode.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -22,6 +22,7 @@
+ #include <linux/buffer_head.h>
+ #include <linux/pagemap.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+ #include "jfs_incore.h"
+ #include "jfs_inode.h"
+ #include "jfs_filsys.h"
+@@ -155,6 +156,7 @@ void jfs_delete_inode(struct inode *inod
+ DQUOT_INIT(inode);
+ DQUOT_FREE_INODE(inode);
+ DQUOT_DROP(inode);
++ DLIMIT_FREE_INODE(inode);
+ }
+
+ clear_inode(inode);
+--- a/fs/jfs/ioctl.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/jfs/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -11,6 +11,7 @@
+ #include <linux/mount.h>
+ #include <linux/time.h>
+ #include <linux/sched.h>
++#include <linux/mount.h>
+ #include <asm/current.h>
+ #include <asm/uaccess.h>
+
+@@ -85,6 +86,11 @@ long jfs_ioctl(struct file *filp, unsign
+ if (!S_ISDIR(inode->i_mode))
+ flags &= ~JFS_DIRSYNC_FL;
+
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -EACCES;
++ }
++
+ /* Is it quota file? Do not allow user to mess with it */
+ if (IS_NOQUOTA(inode)) {
+ err = -EPERM;
+@@ -102,8 +108,8 @@ long jfs_ioctl(struct file *filp, unsign
+ * the relevant capability.
+ */
+ if ((oldflags & JFS_IMMUTABLE_FL) ||
+- ((flags ^ oldflags) &
+- (JFS_APPEND_FL | JFS_IMMUTABLE_FL))) {
++ ((flags ^ oldflags) & (JFS_APPEND_FL |
++ JFS_IMMUTABLE_FL | JFS_IXUNLINK_FL))) {
+ if (!capable(CAP_LINUX_IMMUTABLE)) {
+ mutex_unlock(&inode->i_mutex);
+ err = -EPERM;
+--- a/fs/jfs/jfs_dinode.h 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/jfs_dinode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -161,9 +161,13 @@ struct dinode {
+
+ #define JFS_APPEND_FL 0x01000000 /* writes to file may only append */
+ #define JFS_IMMUTABLE_FL 0x02000000 /* Immutable file */
++#define JFS_IXUNLINK_FL 0x04000000 /* Immutable invert on unlink */
+
+-#define JFS_FL_USER_VISIBLE 0x03F80000
+-#define JFS_FL_USER_MODIFIABLE 0x03F80000
++#define JFS_BARRIER_FL 0x10000000 /* Barrier for chroot() */
++#define JFS_COW_FL 0x20000000 /* Copy on Write marker */
++
++#define JFS_FL_USER_VISIBLE 0x07F80000
++#define JFS_FL_USER_MODIFIABLE 0x07F80000
+ #define JFS_FL_INHERIT 0x03C80000
+
+ /* These are identical to EXT[23]_IOC_GETFLAGS/SETFLAGS */
+--- a/fs/jfs/jfs_dtree.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/jfs_dtree.c 2008-07-16 22:41:36.000000000 -0400
+@@ -102,6 +102,7 @@
+
+ #include <linux/fs.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+ #include "jfs_incore.h"
+ #include "jfs_superblock.h"
+ #include "jfs_filsys.h"
+@@ -383,10 +384,10 @@ static u32 add_index(tid_t tid, struct i
+ */
+ if (DQUOT_ALLOC_BLOCK(ip, sbi->nbperpage))
+ goto clean_up;
+- if (dbAlloc(ip, 0, sbi->nbperpage, &xaddr)) {
+- DQUOT_FREE_BLOCK(ip, sbi->nbperpage);
+- goto clean_up;
+- }
++ if (DLIMIT_ALLOC_BLOCK(ip, sbi->nbperpage))
++ goto clean_up_dquot;
++ if (dbAlloc(ip, 0, sbi->nbperpage, &xaddr))
++ goto clean_up_dlimit;
+
+ /*
+ * Save the table, we're going to overwrite it with the
+@@ -480,6 +481,12 @@ static u32 add_index(tid_t tid, struct i
+
+ return index;
+
++ clean_up_dlimit:
++ DLIMIT_FREE_BLOCK(ip, sbi->nbperpage);
++
++ clean_up_dquot:
++ DQUOT_FREE_BLOCK(ip, sbi->nbperpage);
++
+ clean_up:
+
+ jfs_ip->next_index--;
+@@ -951,6 +958,7 @@ static int dtSplitUp(tid_t tid,
+ struct tlock *tlck;
+ struct lv *lv;
+ int quota_allocation = 0;
++ int dlimit_allocation = 0;
+
+ /* get split page */
+ smp = split->mp;
+@@ -1033,6 +1041,12 @@ static int dtSplitUp(tid_t tid,
+ }
+ quota_allocation += n;
+
++ if (DLIMIT_ALLOC_BLOCK(ip, n)) {
++ rc = -ENOSPC;
++ goto extendOut;
++ }
++ dlimit_allocation += n;
++
+ if ((rc = dbReAlloc(sbi->ipbmap, xaddr, (s64) xlen,
+ (s64) n, &nxaddr)))
+ goto extendOut;
+@@ -1306,6 +1320,9 @@ static int dtSplitUp(tid_t tid,
+ freeKeyName:
+ kfree(key.name);
+
++ /* Rollback dlimit allocation */
++ if (rc && dlimit_allocation)
++ DLIMIT_FREE_BLOCK(ip, dlimit_allocation);
+ /* Rollback quota allocation */
+ if (rc && quota_allocation)
+ DQUOT_FREE_BLOCK(ip, quota_allocation);
+@@ -1373,6 +1390,12 @@ static int dtSplitPage(tid_t tid, struct
+ release_metapage(rmp);
+ return -EDQUOT;
+ }
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, lengthPXD(pxd))) {
++ DQUOT_FREE_BLOCK(ip, lengthPXD(pxd));
++ release_metapage(rmp);
++ return -ENOSPC;
++ }
+
+ jfs_info("dtSplitPage: ip:0x%p smp:0x%p rmp:0x%p", ip, smp, rmp);
+
+@@ -1920,6 +1943,12 @@ static int dtSplitRoot(tid_t tid,
+ release_metapage(rmp);
+ return -EDQUOT;
+ }
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, lengthPXD(pxd))) {
++ DQUOT_FREE_BLOCK(ip, lengthPXD(pxd));
++ release_metapage(rmp);
++ return -ENOSPC;
++ }
+
+ BT_MARK_DIRTY(rmp, ip);
+ /*
+@@ -2286,6 +2315,8 @@ static int dtDeleteUp(tid_t tid, struct
+
+ xlen = lengthPXD(&fp->header.self);
+
++ /* Free dlimit allocation. */
++ DLIMIT_FREE_BLOCK(ip, xlen);
+ /* Free quota allocation. */
+ DQUOT_FREE_BLOCK(ip, xlen);
+
+@@ -2362,6 +2393,8 @@ static int dtDeleteUp(tid_t tid, struct
+
+ xlen = lengthPXD(&p->header.self);
+
++ /* Free dlimit allocation */
++ DLIMIT_FREE_BLOCK(ip, xlen);
+ /* Free quota allocation */
+ DQUOT_FREE_BLOCK(ip, xlen);
+
+--- a/fs/jfs/jfs_extent.c 2008-04-17 10:37:23.000000000 -0400
++++ a/fs/jfs/jfs_extent.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+
+ #include <linux/fs.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+ #include "jfs_incore.h"
+ #include "jfs_inode.h"
+ #include "jfs_superblock.h"
+@@ -147,6 +148,14 @@ extAlloc(struct inode *ip, s64 xlen, s64
+ return -EDQUOT;
+ }
+
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, nxlen)) {
++ DQUOT_FREE_BLOCK(ip, nxlen);
++ dbFree(ip, nxaddr, (s64) nxlen);
++ mutex_unlock(&JFS_IP(ip)->commit_mutex);
++ return -ENOSPC;
++ }
++
+ /* determine the value of the extent flag */
+ xflag = abnr ? XAD_NOTRECORDED : 0;
+
+@@ -164,6 +173,7 @@ extAlloc(struct inode *ip, s64 xlen, s64
+ */
+ if (rc) {
+ dbFree(ip, nxaddr, nxlen);
++ DLIMIT_FREE_BLOCK(ip, nxlen);
+ DQUOT_FREE_BLOCK(ip, nxlen);
+ mutex_unlock(&JFS_IP(ip)->commit_mutex);
+ return (rc);
+@@ -261,6 +271,13 @@ int extRealloc(struct inode *ip, s64 nxl
+ mutex_unlock(&JFS_IP(ip)->commit_mutex);
+ return -EDQUOT;
+ }
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, nxlen)) {
++ DQUOT_FREE_BLOCK(ip, nxlen);
++ dbFree(ip, nxaddr, (s64) nxlen);
++ up(&JFS_IP(ip)->commit_sem);
++ return -ENOSPC;
++ }
+
+ delta = nxlen - xlen;
+
+@@ -297,6 +314,7 @@ int extRealloc(struct inode *ip, s64 nxl
+ /* extend the extent */
+ if ((rc = xtExtend(0, ip, xoff + xlen, (int) nextend, 0))) {
+ dbFree(ip, xaddr + xlen, delta);
++ DLIMIT_FREE_BLOCK(ip, nxlen);
+ DQUOT_FREE_BLOCK(ip, nxlen);
+ goto exit;
+ }
+@@ -308,6 +326,7 @@ int extRealloc(struct inode *ip, s64 nxl
+ */
+ if ((rc = xtTailgate(0, ip, xoff, (int) ntail, nxaddr, 0))) {
+ dbFree(ip, nxaddr, nxlen);
++ DLIMIT_FREE_BLOCK(ip, nxlen);
+ DQUOT_FREE_BLOCK(ip, nxlen);
+ goto exit;
+ }
+--- a/fs/jfs/jfs_filsys.h 2008-04-17 10:37:23.000000000 -0400
++++ a/fs/jfs/jfs_filsys.h 2008-07-16 22:41:36.000000000 -0400
+@@ -263,6 +263,7 @@
+ #define JFS_NAME_MAX 255
+ #define JFS_PATH_MAX BPSIZE
+
++#define JFS_TAGGED 0x00800000 /* Context Tagging */
+
+ /*
+ * file system state (superblock state)
+--- a/fs/jfs/jfs_imap.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/jfs/jfs_imap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -45,6 +45,7 @@
+ #include <linux/buffer_head.h>
+ #include <linux/pagemap.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_tag.h>
+
+ #include "jfs_incore.h"
+ #include "jfs_inode.h"
+@@ -3058,6 +3059,8 @@ static int copy_from_dinode(struct dinod
+ {
+ struct jfs_inode_info *jfs_ip = JFS_IP(ip);
+ struct jfs_sb_info *sbi = JFS_SBI(ip->i_sb);
++ uid_t uid;
++ gid_t gid;
+
+ jfs_ip->fileset = le32_to_cpu(dip->di_fileset);
+ jfs_ip->mode2 = le32_to_cpu(dip->di_mode);
+@@ -3078,14 +3081,18 @@ static int copy_from_dinode(struct dinod
+ }
+ ip->i_nlink = le32_to_cpu(dip->di_nlink);
+
+- jfs_ip->saved_uid = le32_to_cpu(dip->di_uid);
++ uid = le32_to_cpu(dip->di_uid);
++ gid = le32_to_cpu(dip->di_gid);
++ ip->i_tag = INOTAG_TAG(DX_TAG(ip), uid, gid, 0);
++
++ jfs_ip->saved_uid = INOTAG_UID(DX_TAG(ip), uid, gid);
+ if (sbi->uid == -1)
+ ip->i_uid = jfs_ip->saved_uid;
+ else {
+ ip->i_uid = sbi->uid;
+ }
+
+- jfs_ip->saved_gid = le32_to_cpu(dip->di_gid);
++ jfs_ip->saved_gid = INOTAG_GID(DX_TAG(ip), uid, gid);
+ if (sbi->gid == -1)
+ ip->i_gid = jfs_ip->saved_gid;
+ else {
+@@ -3150,14 +3157,12 @@ static void copy_to_dinode(struct dinode
+ dip->di_size = cpu_to_le64(ip->i_size);
+ dip->di_nblocks = cpu_to_le64(PBLK2LBLK(ip->i_sb, ip->i_blocks));
+ dip->di_nlink = cpu_to_le32(ip->i_nlink);
+- if (sbi->uid == -1)
+- dip->di_uid = cpu_to_le32(ip->i_uid);
+- else
+- dip->di_uid = cpu_to_le32(jfs_ip->saved_uid);
+- if (sbi->gid == -1)
+- dip->di_gid = cpu_to_le32(ip->i_gid);
+- else
+- dip->di_gid = cpu_to_le32(jfs_ip->saved_gid);
++
++ dip->di_uid = cpu_to_le32(TAGINO_UID(DX_TAG(ip),
++ (sbi->uid == -1) ? ip->i_uid : jfs_ip->saved_uid, ip->i_tag));
++ dip->di_gid = cpu_to_le32(TAGINO_GID(DX_TAG(ip),
++ (sbi->gid == -1) ? ip->i_gid : jfs_ip->saved_gid, ip->i_tag));
++
+ jfs_get_inode_flags(jfs_ip);
+ /*
+ * mode2 is only needed for storing the higher order bits.
+--- a/fs/jfs/jfs_inode.c 2008-04-17 10:33:02.000000000 -0400
++++ a/fs/jfs/jfs_inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,8 @@
+
+ #include <linux/fs.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+ #include "jfs_incore.h"
+ #include "jfs_inode.h"
+ #include "jfs_filsys.h"
+@@ -30,29 +32,46 @@ void jfs_set_inode_flags(struct inode *i
+ {
+ unsigned int flags = JFS_IP(inode)->mode2;
+
+- inode->i_flags &= ~(S_IMMUTABLE | S_APPEND |
+- S_NOATIME | S_DIRSYNC | S_SYNC);
++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK |
++ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
+
+ if (flags & JFS_IMMUTABLE_FL)
+ inode->i_flags |= S_IMMUTABLE;
++ if (flags & JFS_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
++
++ if (flags & JFS_SYNC_FL)
++ inode->i_flags |= S_SYNC;
+ if (flags & JFS_APPEND_FL)
+ inode->i_flags |= S_APPEND;
+ if (flags & JFS_NOATIME_FL)
+ inode->i_flags |= S_NOATIME;
+ if (flags & JFS_DIRSYNC_FL)
+ inode->i_flags |= S_DIRSYNC;
+- if (flags & JFS_SYNC_FL)
+- inode->i_flags |= S_SYNC;
++
++ inode->i_vflags &= ~(V_BARRIER | V_COW);
++
++ if (flags & JFS_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ if (flags & JFS_COW_FL)
++ inode->i_vflags |= V_COW;
+ }
+
+ void jfs_get_inode_flags(struct jfs_inode_info *jfs_ip)
+ {
+ unsigned int flags = jfs_ip->vfs_inode.i_flags;
++ unsigned int vflags = jfs_ip->vfs_inode.i_vflags;
++
++ jfs_ip->mode2 &= ~(JFS_IMMUTABLE_FL | JFS_IXUNLINK_FL |
++ JFS_APPEND_FL | JFS_NOATIME_FL |
++ JFS_DIRSYNC_FL | JFS_SYNC_FL |
++ JFS_BARRIER_FL | JFS_COW_FL);
+
+- jfs_ip->mode2 &= ~(JFS_IMMUTABLE_FL | JFS_APPEND_FL | JFS_NOATIME_FL |
+- JFS_DIRSYNC_FL | JFS_SYNC_FL);
+ if (flags & S_IMMUTABLE)
+ jfs_ip->mode2 |= JFS_IMMUTABLE_FL;
++ if (flags & S_IXUNLINK)
++ jfs_ip->mode2 |= JFS_IXUNLINK_FL;
++
+ if (flags & S_APPEND)
+ jfs_ip->mode2 |= JFS_APPEND_FL;
+ if (flags & S_NOATIME)
+@@ -61,6 +80,19 @@ void jfs_get_inode_flags(struct jfs_inod
+ jfs_ip->mode2 |= JFS_DIRSYNC_FL;
+ if (flags & S_SYNC)
+ jfs_ip->mode2 |= JFS_SYNC_FL;
++
++ if (vflags & V_BARRIER)
++ jfs_ip->mode2 |= JFS_BARRIER_FL;
++ if (vflags & V_COW)
++ jfs_ip->mode2 |= JFS_COW_FL;
++}
++
++int jfs_sync_flags(struct inode *inode)
++{
++ jfs_get_inode_flags(JFS_IP(inode));
++ inode->i_ctime = CURRENT_TIME;
++ mark_inode_dirty(inode);
++ return 0;
+ }
+
+ /*
+@@ -108,10 +140,17 @@ struct inode *ialloc(struct inode *paren
+ jfs_inode->saved_uid = inode->i_uid;
+ jfs_inode->saved_gid = inode->i_gid;
+
++ inode->i_tag = dx_current_fstag(sb);
++ if (DLIMIT_ALLOC_INODE(inode)) {
++ iput(inode);
++ return ERR_PTR(-ENOSPC);
++ }
++
+ /*
+ * Allocate inode to quota.
+ */
+ if (DQUOT_ALLOC_INODE(inode)) {
++ DLIMIT_FREE_INODE(inode);
+ DQUOT_DROP(inode);
+ inode->i_flags |= S_NOQUOTA;
+ inode->i_nlink = 0;
+--- a/fs/jfs/jfs_inode.h 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/jfs_inode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -39,6 +39,7 @@ extern struct dentry *jfs_fh_to_dentry(s
+ extern struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid,
+ int fh_len, int fh_type);
+ extern void jfs_set_inode_flags(struct inode *);
++extern int jfs_sync_flags(struct inode *);
+ extern int jfs_get_block(struct inode *, sector_t, struct buffer_head *, int);
+
+ extern const struct address_space_operations jfs_aops;
+--- a/fs/jfs/jfs_xtree.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/jfs/jfs_xtree.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+
+ #include <linux/fs.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+ #include "jfs_incore.h"
+ #include "jfs_filsys.h"
+ #include "jfs_metapage.h"
+@@ -846,7 +847,12 @@ int xtInsert(tid_t tid, /* transaction
+ hint = 0;
+ if ((rc = DQUOT_ALLOC_BLOCK(ip, xlen)))
+ goto out;
++ if ((rc = DLIMIT_ALLOC_BLOCK(ip, xlen))) {
++ DQUOT_FREE_BLOCK(ip, xlen);
++ goto out;
++ }
+ if ((rc = dbAlloc(ip, hint, (s64) xlen, &xaddr))) {
++ DLIMIT_FREE_BLOCK(ip, xlen);
+ DQUOT_FREE_BLOCK(ip, xlen);
+ goto out;
+ }
+@@ -876,6 +882,7 @@ int xtInsert(tid_t tid, /* transaction
+ /* undo data extent allocation */
+ if (*xaddrp == 0) {
+ dbFree(ip, xaddr, (s64) xlen);
++ DLIMIT_FREE_BLOCK(ip, xlen);
+ DQUOT_FREE_BLOCK(ip, xlen);
+ }
+ return rc;
+@@ -1232,6 +1239,7 @@ xtSplitPage(tid_t tid, struct inode *ip,
+ struct tlock *tlck;
+ struct xtlock *sxtlck = NULL, *rxtlck = NULL;
+ int quota_allocation = 0;
++ int dlimit_allocation = 0;
+
+ smp = split->mp;
+ sp = XT_PAGE(ip, smp);
+@@ -1251,6 +1259,13 @@ xtSplitPage(tid_t tid, struct inode *ip,
+
+ quota_allocation += lengthPXD(pxd);
+
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, lengthPXD(pxd))) {
++ rc = -ENOSPC;
++ goto clean_up;
++ }
++ dlimit_allocation += lengthPXD(pxd);
++
+ /*
+ * allocate the new right page for the split
+ */
+@@ -1452,6 +1467,9 @@ xtSplitPage(tid_t tid, struct inode *ip,
+
+ clean_up:
+
++ /* Rollback dlimit allocation. */
++ if (dlimit_allocation)
++ DLIMIT_FREE_BLOCK(ip, dlimit_allocation);
+ /* Rollback quota allocation. */
+ if (quota_allocation)
+ DQUOT_FREE_BLOCK(ip, quota_allocation);
+@@ -1515,6 +1533,12 @@ xtSplitRoot(tid_t tid,
+ release_metapage(rmp);
+ return -EDQUOT;
+ }
++ /* Allocate blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, lengthPXD(pxd))) {
++ DQUOT_FREE_BLOCK(ip, lengthPXD(pxd));
++ release_metapage(rmp);
++ return -ENOSPC;
++ }
+
+ jfs_info("xtSplitRoot: ip:0x%p rmp:0x%p", ip, rmp);
+
+@@ -3938,6 +3962,8 @@ s64 xtTruncate(tid_t tid, struct inode *
+ else
+ ip->i_size = newsize;
+
++ /* update dlimit allocation to reflect freed blocks */
++ DLIMIT_FREE_BLOCK(ip, nfreed);
+ /* update quota allocation to reflect freed blocks */
+ DQUOT_FREE_BLOCK(ip, nfreed);
+
+--- a/fs/jfs/namei.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/namei.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/ctype.h>
+ #include <linux/quotaops.h>
+ #include <linux/exportfs.h>
++#include <linux/vs_tag.h>
+ #include "jfs_incore.h"
+ #include "jfs_superblock.h"
+ #include "jfs_inode.h"
+@@ -1468,6 +1469,7 @@ static struct dentry *jfs_lookup(struct
+ return ERR_CAST(ip);
+ }
+
++ dx_propagate_tag(nd, ip);
+ dentry = d_splice_alias(ip, dentry);
+
+ if (dentry && (JFS_SBI(dip->i_sb)->mntflag & JFS_OS2))
+@@ -1550,6 +1552,7 @@ const struct inode_operations jfs_dir_in
+ .setattr = jfs_setattr,
+ .permission = jfs_permission,
+ #endif
++ .sync_flags = jfs_sync_flags,
+ };
+
+ const struct file_operations jfs_dir_operations = {
+--- a/fs/jfs/super.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/jfs/super.c 2008-07-16 22:41:36.000000000 -0400
+@@ -195,7 +195,8 @@ static void jfs_put_super(struct super_b
+ enum {
+ Opt_integrity, Opt_nointegrity, Opt_iocharset, Opt_resize,
+ Opt_resize_nosize, Opt_errors, Opt_ignore, Opt_err, Opt_quota,
+- Opt_usrquota, Opt_grpquota, Opt_uid, Opt_gid, Opt_umask
++ Opt_usrquota, Opt_grpquota, Opt_uid, Opt_gid, Opt_umask,
++ Opt_tag, Opt_notag, Opt_tagid
+ };
+
+ static match_table_t tokens = {
+@@ -205,6 +206,10 @@ static match_table_t tokens = {
+ {Opt_resize, "resize=%u"},
+ {Opt_resize_nosize, "resize"},
+ {Opt_errors, "errors=%s"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
++ {Opt_tag, "tagxid"},
+ {Opt_ignore, "noquota"},
+ {Opt_ignore, "quota"},
+ {Opt_usrquota, "usrquota"},
+@@ -339,6 +344,20 @@ static int parse_options(char *options,
+ }
+ break;
+ }
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ *flag |= JFS_TAGGED;
++ break;
++ case Opt_notag:
++ *flag &= JFS_TAGGED;
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ *flag |= JFS_TAGGED;
++ break;
++#endif
+ default:
+ printk("jfs: Unrecognized mount option \"%s\" "
+ " or missing value\n", p);
+@@ -369,6 +388,13 @@ static int jfs_remount(struct super_bloc
+ if (!parse_options(data, sb, &newLVSize, &flag)) {
+ return -EINVAL;
+ }
++
++ if ((flag & JFS_TAGGED) && !(sb->s_flags & MS_TAGGED)) {
++ printk(KERN_ERR "JFS: %s: tagging not permitted on remount.\n",
++ sb->s_id);
++ return -EINVAL;
++ }
++
+ if (newLVSize) {
+ if (sb->s_flags & MS_RDONLY) {
+ printk(KERN_ERR
+@@ -440,6 +466,9 @@ static int jfs_fill_super(struct super_b
+ #ifdef CONFIG_JFS_POSIX_ACL
+ sb->s_flags |= MS_POSIXACL;
+ #endif
++ /* map mount option tagxid */
++ if (sbi->flag & JFS_TAGGED)
++ sb->s_flags |= MS_TAGGED;
+
+ if (newLVSize) {
+ printk(KERN_ERR "resize option for remount only\n");
+--- a/fs/jfs/xattr.c 2008-04-17 10:37:23.000000000 -0400
++++ a/fs/jfs/xattr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -23,6 +23,7 @@
+ #include <linux/posix_acl_xattr.h>
+ #include <linux/quotaops.h>
+ #include <linux/security.h>
++#include <linux/vs_dlimit.h>
+ #include "jfs_incore.h"
+ #include "jfs_superblock.h"
+ #include "jfs_dmap.h"
+@@ -263,9 +264,16 @@ static int ea_write(struct inode *ip, st
+ if (DQUOT_ALLOC_BLOCK(ip, nblocks)) {
+ return -EDQUOT;
+ }
++ /* Allocate new blocks to dlimit. */
++ if (DLIMIT_ALLOC_BLOCK(ip, nblocks)) {
++ DQUOT_FREE_BLOCK(ip, nblocks);
++ return -ENOSPC;
++ }
+
+ rc = dbAlloc(ip, INOHINT(ip), nblocks, &blkno);
+ if (rc) {
++ /*Rollback dlimit allocation. */
++ DLIMIT_FREE_BLOCK(ip, nblocks);
+ /*Rollback quota allocation. */
+ DQUOT_FREE_BLOCK(ip, nblocks);
+ return rc;
+@@ -332,6 +340,8 @@ static int ea_write(struct inode *ip, st
+
+ failed:
+ /* Rollback quota allocation. */
++ DLIMIT_FREE_BLOCK(ip, nblocks);
++ /* Rollback quota allocation. */
+ DQUOT_FREE_BLOCK(ip, nblocks);
+
+ dbFree(ip, blkno, nblocks);
+@@ -468,6 +478,7 @@ static int ea_get(struct inode *inode, s
+ s64 blkno;
+ int rc;
+ int quota_allocation = 0;
++ int dlimit_allocation = 0;
+
+ /* When fsck.jfs clears a bad ea, it doesn't clear the size */
+ if (ji->ea.flag == 0)
+@@ -543,6 +554,12 @@ static int ea_get(struct inode *inode, s
+
+ quota_allocation = blocks_needed;
+
++ /* Allocate new blocks to dlimit. */
++ rc = -ENOSPC;
++ if (DLIMIT_ALLOC_BLOCK(inode, blocks_needed))
++ goto clean_up;
++ dlimit_allocation = blocks_needed;
++
+ rc = dbAlloc(inode, INOHINT(inode), (s64) blocks_needed,
+ &blkno);
+ if (rc)
+@@ -600,6 +617,9 @@ static int ea_get(struct inode *inode, s
+ return ea_size;
+
+ clean_up:
++ /* Rollback dlimit allocation */
++ if (dlimit_allocation)
++ DLIMIT_FREE_BLOCK(inode, dlimit_allocation);
+ /* Rollback quota allocation */
+ if (quota_allocation)
+ DQUOT_FREE_BLOCK(inode, quota_allocation);
+@@ -676,8 +696,10 @@ static int ea_put(tid_t tid, struct inod
+ }
+
+ /* If old blocks exist, they must be removed from quota allocation. */
+- if (old_blocks)
++ if (old_blocks) {
++ DLIMIT_FREE_BLOCK(inode, old_blocks);
+ DQUOT_FREE_BLOCK(inode, old_blocks);
++ }
+
+ inode->i_ctime = CURRENT_TIME;
+
+--- a/fs/libfs.c 2008-08-12 01:41:51.000000000 -0400
++++ a/fs/libfs.c 2008-08-12 01:42:21.000000000 -0400
+@@ -125,7 +125,8 @@ static inline unsigned char dt_type(stru
+ * both impossible due to the lock on directory.
+ */
+
+-int dcache_readdir(struct file * filp, void * dirent, filldir_t filldir)
++static inline int do_dcache_readdir_filter(struct file *filp,
++ void *dirent, filldir_t filldir, int (*filter)(struct dentry *dentry))
+ {
+ struct dentry *dentry = filp->f_path.dentry;
+ struct dentry *cursor = filp->private_data;
+@@ -158,6 +159,8 @@ int dcache_readdir(struct file * filp, v
+ next = list_entry(p, struct dentry, d_u.d_child);
+ if (d_unhashed(next) || !next->d_inode)
+ continue;
++ if (filter && !filter(next))
++ continue;
+
+ spin_unlock(&dcache_lock);
+ if (filldir(dirent, next->d_name.name,
+@@ -176,6 +179,18 @@ int dcache_readdir(struct file * filp, v
+ return 0;
+ }
+
++int dcache_readdir(struct file *filp, void *dirent, filldir_t filldir)
++{
++ return do_dcache_readdir_filter(filp, dirent, filldir, NULL);
++}
++
++int dcache_readdir_filter(struct file *filp, void *dirent, filldir_t filldir,
++ int (*filter)(struct dentry *))
++{
++ return do_dcache_readdir_filter(filp, dirent, filldir, filter);
++}
++
++
+ ssize_t generic_read_dir(struct file *filp, char __user *buf, size_t siz, loff_t *ppos)
+ {
+ return -EISDIR;
+@@ -823,6 +838,7 @@ EXPORT_SYMBOL(dcache_dir_close);
+ EXPORT_SYMBOL(dcache_dir_lseek);
+ EXPORT_SYMBOL(dcache_dir_open);
+ EXPORT_SYMBOL(dcache_readdir);
++EXPORT_SYMBOL(dcache_readdir_filter);
+ EXPORT_SYMBOL(generic_read_dir);
+ EXPORT_SYMBOL(get_sb_pseudo);
+ EXPORT_SYMBOL(simple_write_begin);
+--- a/fs/locks.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/locks.c 2008-07-16 22:41:36.000000000 -0400
+@@ -127,6 +127,8 @@
+ #include <linux/time.h>
+ #include <linux/rcupdate.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_base.h>
++#include <linux/vs_limit.h>
+
+ #include <asm/uaccess.h>
+
+@@ -148,6 +150,8 @@ static struct kmem_cache *filelock_cache
+ /* Allocate an empty lock structure. */
+ static struct file_lock *locks_alloc_lock(void)
+ {
++ if (!vx_locks_avail(1))
++ return NULL;
+ return kmem_cache_alloc(filelock_cache, GFP_KERNEL);
+ }
+
+@@ -173,6 +177,7 @@ static void locks_free_lock(struct file_
+ BUG_ON(!list_empty(&fl->fl_block));
+ BUG_ON(!list_empty(&fl->fl_link));
+
++ vx_locks_dec(fl);
+ locks_release_private(fl);
+ kmem_cache_free(filelock_cache, fl);
+ }
+@@ -193,6 +198,7 @@ void locks_init_lock(struct file_lock *f
+ fl->fl_start = fl->fl_end = 0;
+ fl->fl_ops = NULL;
+ fl->fl_lmops = NULL;
++ fl->fl_xid = -1;
+ }
+
+ EXPORT_SYMBOL(locks_init_lock);
+@@ -247,6 +253,7 @@ void locks_copy_lock(struct file_lock *n
+ new->fl_file = fl->fl_file;
+ new->fl_ops = fl->fl_ops;
+ new->fl_lmops = fl->fl_lmops;
++ new->fl_xid = fl->fl_xid;
+
+ locks_copy_private(new, fl);
+ }
+@@ -285,6 +292,11 @@ static int flock_make_lock(struct file *
+ fl->fl_flags = FL_FLOCK;
+ fl->fl_type = type;
+ fl->fl_end = OFFSET_MAX;
++
++ vxd_assert(filp->f_xid == vx_current_xid(),
++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid());
++ fl->fl_xid = filp->f_xid;
++ vx_locks_inc(fl);
+
+ *lock = fl;
+ return 0;
+@@ -450,6 +462,7 @@ static int lease_init(struct file *filp,
+
+ fl->fl_owner = current->files;
+ fl->fl_pid = current->tgid;
++ fl->fl_xid = vx_current_xid();
+
+ fl->fl_file = filp;
+ fl->fl_flags = FL_LEASE;
+@@ -469,6 +482,11 @@ static struct file_lock *lease_alloc(str
+ if (fl == NULL)
+ return ERR_PTR(error);
+
++ fl->fl_xid = vx_current_xid();
++ if (filp)
++ vxd_assert(filp->f_xid == fl->fl_xid,
++ "f_xid(%d) == fl_xid(%d)", filp->f_xid, fl->fl_xid);
++ vx_locks_inc(fl);
+ error = lease_init(filp, type, fl);
+ if (error) {
+ locks_free_lock(fl);
+@@ -769,6 +787,7 @@ static int flock_lock_file(struct file *
+ if (found)
+ cond_resched_bkl();
+
++ new_fl->fl_xid = -1;
+ find_conflict:
+ for_each_lock(inode, before) {
+ struct file_lock *fl = *before;
+@@ -787,6 +806,7 @@ find_conflict:
+ goto out;
+ locks_copy_lock(new_fl, request);
+ locks_insert_lock(before, new_fl);
++ vx_locks_inc(new_fl);
+ new_fl = NULL;
+ error = 0;
+
+@@ -797,7 +817,8 @@ out:
+ return error;
+ }
+
+-static int __posix_lock_file(struct inode *inode, struct file_lock *request, struct file_lock *conflock)
++static int __posix_lock_file(struct inode *inode, struct file_lock *request,
++ struct file_lock *conflock, xid_t xid)
+ {
+ struct file_lock *fl;
+ struct file_lock *new_fl = NULL;
+@@ -807,6 +828,8 @@ static int __posix_lock_file(struct inod
+ struct file_lock **before;
+ int error, added = 0;
+
++ vxd_assert(xid == vx_current_xid(),
++ "xid(%d) == current(%d)", xid, vx_current_xid());
+ /*
+ * We may need two file_lock structures for this operation,
+ * so we get them in advance to avoid races.
+@@ -817,7 +840,11 @@ static int __posix_lock_file(struct inod
+ (request->fl_type != F_UNLCK ||
+ request->fl_start != 0 || request->fl_end != OFFSET_MAX)) {
+ new_fl = locks_alloc_lock();
++ new_fl->fl_xid = xid;
++ vx_locks_inc(new_fl);
+ new_fl2 = locks_alloc_lock();
++ new_fl2->fl_xid = xid;
++ vx_locks_inc(new_fl2);
+ }
+
+ lock_kernel();
+@@ -1016,7 +1043,8 @@ static int __posix_lock_file(struct inod
+ int posix_lock_file(struct file *filp, struct file_lock *fl,
+ struct file_lock *conflock)
+ {
+- return __posix_lock_file(filp->f_path.dentry->d_inode, fl, conflock);
++ return __posix_lock_file(filp->f_path.dentry->d_inode,
++ fl, conflock, filp->f_xid);
+ }
+ EXPORT_SYMBOL(posix_lock_file);
+
+@@ -1106,7 +1134,7 @@ int locks_mandatory_area(int read_write,
+ fl.fl_end = offset + count - 1;
+
+ for (;;) {
+- error = __posix_lock_file(inode, &fl, NULL);
++ error = __posix_lock_file(inode, &fl, NULL, filp->f_xid);
+ if (error != -EAGAIN)
+ break;
+ if (!(fl.fl_flags & FL_SLEEP))
+@@ -1423,6 +1451,7 @@ int generic_setlease(struct file *filp,
+
+ locks_copy_lock(new_fl, lease);
+ locks_insert_lock(before, new_fl);
++ vx_locks_inc(new_fl);
+
+ *flp = new_fl;
+ return 0;
+@@ -1753,6 +1782,11 @@ int fcntl_setlk(unsigned int fd, struct
+ if (file_lock == NULL)
+ return -ENOLCK;
+
++ vxd_assert(filp->f_xid == vx_current_xid(),
++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid());
++ file_lock->fl_xid = filp->f_xid;
++ vx_locks_inc(file_lock);
++
+ /*
+ * This might block, so we do it before checking the inode.
+ */
+@@ -1890,6 +1924,11 @@ int fcntl_setlk64(unsigned int fd, struc
+ if (file_lock == NULL)
+ return -ENOLCK;
+
++ vxd_assert(filp->f_xid == vx_current_xid(),
++ "f_xid(%d) == current(%d)", filp->f_xid, vx_current_xid());
++ file_lock->fl_xid = filp->f_xid;
++ vx_locks_inc(file_lock);
++
+ /*
+ * This might block, so we do it before checking the inode.
+ */
+@@ -2173,8 +2212,11 @@ static int locks_show(struct seq_file *f
+
+ lock_get_status(f, fl, (long)f->private, "");
+
+- list_for_each_entry(bfl, &fl->fl_block, fl_block)
++ list_for_each_entry(bfl, &fl->fl_block, fl_block) {
++ if (!vx_check(fl->fl_xid, VS_WATCH_P | VS_IDENT))
++ continue;
+ lock_get_status(f, bfl, (long)f->private, " ->");
++ }
+
+ f->private++;
+ return 0;
+--- a/fs/namei.c 2008-08-12 01:41:51.000000000 -0400
++++ a/fs/namei.c 2008-08-12 01:42:21.000000000 -0400
+@@ -31,6 +31,13 @@
+ #include <linux/file.h>
+ #include <linux/fcntl.h>
+ #include <linux/device_cgroup.h>
++#include <linux/proc_fs.h>
++#include <linux/vserver/inode.h>
++#include <linux/vs_base.h>
++#include <linux/vs_tag.h>
++#include <linux/vs_cowbl.h>
++#include <linux/vs_device.h>
++#include <linux/vs_context.h>
+ #include <asm/namei.h>
+ #include <asm/uaccess.h>
+
+@@ -226,6 +233,28 @@ int generic_permission(struct inode *ino
+ return -EACCES;
+ }
+
++static inline int dx_barrier(struct inode *inode)
++{
++ if (IS_BARRIER(inode) && !vx_check(0, VS_ADMIN)) {
++ vxwprintk_task(1, "did hit the barrier.");
++ return 1;
++ }
++ return 0;
++}
++
++static inline int dx_permission(struct inode *inode, int mask, struct nameidata *nd)
++{
++ if (dx_barrier(inode))
++ return -EACCES;
++ if (dx_notagcheck(nd) ||
++ dx_check(inode->i_tag, DX_HOSTID|DX_ADMIN|DX_WATCH|DX_IDENT))
++ return 0;
++
++ vxwprintk_task(1, "denied access to %p[#%d,%lu] »%s«.",
++ inode, inode->i_tag, inode->i_ino, vxd_cond_path(nd));
++ return -EACCES;
++}
++
+ int permission(struct inode *inode, int mask, struct nameidata *nd)
+ {
+ int retval, submask;
+@@ -247,7 +276,7 @@ int permission(struct inode *inode, int
+ /*
+ * Nobody gets write access to an immutable file.
+ */
+- if (IS_IMMUTABLE(inode))
++ if (IS_IMMUTABLE(inode) && !IS_COW(inode))
+ return -EACCES;
+ }
+
+@@ -262,6 +291,11 @@ int permission(struct inode *inode, int
+
+ /* Ordinary permission routines do not understand MAY_APPEND. */
+ submask = mask & ~MAY_APPEND;
++ if ((inode->i_sb->s_magic != DEVPTS_SUPER_MAGIC) &&
++ (inode->i_sb->s_magic != PROC_SUPER_MAGIC) &&
++ (retval = dx_permission(inode, mask, nd)))
++ return retval;
++
+ if (inode->i_op && inode->i_op->permission) {
+ retval = inode->i_op->permission(inode, submask, nd);
+ if (!retval) {
+@@ -464,6 +498,8 @@ static int exec_permission_lite(struct i
+ {
+ umode_t mode = inode->i_mode;
+
++ if (dx_barrier(inode))
++ return -EACCES;
+ if (inode->i_op && inode->i_op->permission)
+ return -EAGAIN;
+
+@@ -800,7 +836,8 @@ static __always_inline void follow_dotdo
+ if (nd->path.dentry == fs->root.dentry &&
+ nd->path.mnt == fs->root.mnt) {
+ read_unlock(&fs->lock);
+- break;
++ /* for sane '/' avoid follow_mount() */
++ return;
+ }
+ read_unlock(&fs->lock);
+ spin_lock(&dcache_lock);
+@@ -837,16 +874,39 @@ static int do_lookup(struct nameidata *n
+ {
+ struct vfsmount *mnt = nd->path.mnt;
+ struct dentry *dentry = __d_lookup(nd->path.dentry, name);
++ struct inode *inode;
+
+ if (!dentry)
+ goto need_lookup;
+ if (dentry->d_op && dentry->d_op->d_revalidate)
+ goto need_revalidate;
++ inode = dentry->d_inode;
++ if (!inode)
++ goto done;
++
++ if (inode->i_sb->s_magic == PROC_SUPER_MAGIC) {
++ struct proc_dir_entry *de = PDE(inode);
++
++ if (de && !vx_hide_check(0, de->vx_flags))
++ goto hidden;
++ } else if (inode->i_sb->s_magic == DEVPTS_SUPER_MAGIC) {
++ if (!vx_check((xid_t)inode->i_tag, VS_WATCH_P | VS_IDENT))
++ goto hidden;
++ } else {
++ if (!dx_notagcheck(nd) && !dx_check(inode->i_tag,
++ DX_WATCH | DX_ADMIN | DX_HOSTID | DX_IDENT))
++ goto hidden;
++ }
+ done:
+ path->mnt = mnt;
+ path->dentry = dentry;
+ __follow_mount(path);
+ return 0;
++hidden:
++ vxwprintk_task(1, "did lookup hidden %p[#%d,%lu] »%s«.",
++ inode, inode->i_tag, inode->i_ino, vxd_path(path));
++ dput(dentry);
++ return -ENOENT;
+
+ need_lookup:
+ dentry = real_lookup(nd->path.dentry, name, nd);
+@@ -1499,7 +1559,7 @@ static int may_delete(struct inode *dir,
+ if (IS_APPEND(dir))
+ return -EPERM;
+ if (check_sticky(dir, victim->d_inode)||IS_APPEND(victim->d_inode)||
+- IS_IMMUTABLE(victim->d_inode))
++ IS_IXORUNLINK(victim->d_inode))
+ return -EPERM;
+ if (isdir) {
+ if (!S_ISDIR(victim->d_inode->i_mode))
+@@ -1644,6 +1704,14 @@ int may_open(struct nameidata *nd, int a
+ flag &= ~O_TRUNC;
+ }
+
++#ifdef CONFIG_VSERVER_COWBL
++ if (IS_COW(inode) && (flag & FMODE_WRITE)) {
++ if (IS_COW_LINK(inode))
++ return -EMLINK;
++ inode->i_flags &= ~(S_IXUNLINK|S_IMMUTABLE);
++ mark_inode_dirty(inode);
++ }
++#endif
+ error = vfs_permission(nd, acc_mode);
+ if (error)
+ return error;
+@@ -1770,6 +1838,11 @@ struct file *do_filp_open(int dfd, const
+ int will_write;
+ int flag = open_to_namei_flags(open_flag);
+
++#ifdef CONFIG_VSERVER_COWBL
++ int rflag = flag;
++ int rmode = mode;
++restart:
++#endif
+ acc_mode = ACC_MODE(flag);
+
+ /* O_TRUNC implies we need access checks for write permissions */
+@@ -1893,6 +1966,22 @@ ok:
+ goto exit;
+ }
+ error = may_open(&nd, acc_mode, flag);
++#ifdef CONFIG_VSERVER_COWBL
++ if (error == -EMLINK) {
++ struct dentry *dentry;
++ dentry = cow_break_link(pathname);
++ if (IS_ERR(dentry)) {
++ error = PTR_ERR(dentry);
++ goto exit;
++ }
++ dput(dentry);
++ release_open_intent(&nd);
++ path_put(&nd.path);
++ flag = rflag;
++ mode = rmode;
++ goto restart;
++ }
++#endif
+ if (error) {
+ if (will_write)
+ mnt_drop_write(nd.path.mnt);
+@@ -2045,9 +2134,17 @@ int vfs_mknod(struct inode *dir, struct
+ if (error)
+ return error;
+
+- if ((S_ISCHR(mode) || S_ISBLK(mode)) && !capable(CAP_MKNOD))
++ if (!(S_ISCHR(mode) || S_ISBLK(mode)))
++ goto okay;
++
++ if (!capable(CAP_MKNOD))
+ return -EPERM;
+
++ if (S_ISCHR(mode) && !vs_chrdev_perm(dev, DATTR_CREATE))
++ return -EPERM;
++ if (S_ISBLK(mode) && !vs_blkdev_perm(dev, DATTR_CREATE))
++ return -EPERM;
++okay:
+ if (!dir->i_op || !dir->i_op->mknod)
+ return -EPERM;
+
+@@ -2509,7 +2606,7 @@ int vfs_link(struct dentry *old_dentry,
+ /*
+ * A link to an append-only or immutable file cannot be created.
+ */
+- if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
++ if (IS_APPEND(inode) || IS_IXORUNLINK(inode))
+ return -EPERM;
+ if (!dir->i_op || !dir->i_op->link)
+ return -EPERM;
+@@ -2889,6 +2986,214 @@ int vfs_follow_link(struct nameidata *nd
+ return __vfs_follow_link(nd, link);
+ }
+
++
++#ifdef CONFIG_VSERVER_COWBL
++
++#include <linux/file.h>
++
++static inline
++long do_cow_splice(struct file *in, struct file *out, size_t len)
++{
++ loff_t ppos = 0;
++
++ return do_splice_direct(in, &ppos, out, len, 0);
++}
++
++struct dentry *cow_break_link(const char *pathname)
++{
++ int ret, mode, pathlen, redo = 0;
++ struct nameidata old_nd, dir_nd;
++ struct path old_path, new_path;
++ struct dentry *dir, *res = NULL;
++ struct file *old_file;
++ struct file *new_file;
++ char *to, *path, pad='\251';
++ loff_t size;
++
++ vxdprintk(VXD_CBIT(misc, 1), "cow_break_link(»%s«)", pathname);
++ path = kmalloc(PATH_MAX, GFP_KERNEL);
++ ret = -ENOMEM;
++ if (!path)
++ goto out;
++
++ /* old_nd will have refs to dentry and mnt */
++ ret = path_lookup(pathname, LOOKUP_FOLLOW, &old_nd);
++ vxdprintk(VXD_CBIT(misc, 2), "path_lookup(old): %d", ret);
++ if (ret < 0)
++ goto out_free_path;
++
++ old_path = old_nd.path;
++ mode = old_path.dentry->d_inode->i_mode;
++
++ to = d_path(&old_path, path, PATH_MAX-2);
++ pathlen = strlen(to);
++ vxdprintk(VXD_CBIT(misc, 2), "old path »%s« [»%.*s«:%d]", to,
++ old_path.dentry->d_name.len, old_path.dentry->d_name.name,
++ old_path.dentry->d_name.len);
++
++ to[pathlen + 1] = 0;
++retry:
++ to[pathlen] = pad--;
++ ret = -EMLINK;
++ if (pad <= '\240')
++ goto out_rel_old;
++
++ vxdprintk(VXD_CBIT(misc, 1), "temp copy »%s«", to);
++ /* dir_nd will have refs to dentry and mnt */
++ ret = path_lookup(to,
++ LOOKUP_PARENT | LOOKUP_OPEN | LOOKUP_CREATE, &dir_nd);
++ vxdprintk(VXD_CBIT(misc, 2),
++ "path_lookup(new): %d", ret);
++ if (ret < 0)
++ goto retry;
++
++ /* this puppy downs the inode mutex */
++ new_path.dentry = lookup_create(&dir_nd, 0);
++ vxdprintk(VXD_CBIT(misc, 2),
++ "lookup_create(new): %p [»%.*s«:%d]", new_path.dentry,
++ new_path.dentry->d_name.len, new_path.dentry->d_name.name,
++ new_path.dentry->d_name.len);
++ if (!new_path.dentry || IS_ERR(new_path.dentry)) {
++ path_put(&dir_nd.path);
++ goto retry;
++ }
++ dir = dir_nd.path.dentry;
++
++ ret = vfs_create(dir_nd.path.dentry->d_inode, new_path.dentry, mode, &dir_nd);
++ vxdprintk(VXD_CBIT(misc, 2),
++ "vfs_create(new): %d", ret);
++ if (ret == -EEXIST) {
++ mutex_unlock(&dir->d_inode->i_mutex);
++ dput(new_path.dentry);
++ path_put(&dir_nd.path);
++ goto retry;
++ }
++ else if (ret < 0)
++ goto out_unlock_new;
++
++ /* drop out early, ret passes ENOENT */
++ ret = -ENOENT;
++ if ((redo = d_unhashed(old_path.dentry)))
++ goto out_unlock_new;
++
++ new_path.mnt = dir_nd.path.mnt;
++ dget(old_path.dentry);
++ mntget(old_path.mnt);
++ /* this one cleans up the dentry/mnt in case of failure */
++ old_file = dentry_open(old_path.dentry, old_path.mnt, O_RDONLY);
++ vxdprintk(VXD_CBIT(misc, 2),
++ "dentry_open(old): %p", old_file);
++ if (!old_file || IS_ERR(old_file)) {
++ res = IS_ERR(old_file) ? (void *) old_file : res;
++ goto out_unlock_new;
++ }
++
++ dget(new_path.dentry);
++ mntget(new_path.mnt);
++ /* this one cleans up the dentry/mnt in case of failure */
++ new_file = dentry_open(new_path.dentry, new_path.mnt, O_WRONLY);
++ vxdprintk(VXD_CBIT(misc, 2),
++ "dentry_open(new): %p", new_file);
++
++ ret = IS_ERR(new_file) ? PTR_ERR(new_file) : -ENOENT;
++ if (!new_file || IS_ERR(new_file))
++ goto out_fput_old;
++
++ size = i_size_read(old_file->f_dentry->d_inode);
++ ret = do_cow_splice(old_file, new_file, size);
++ vxdprintk(VXD_CBIT(misc, 2), "do_splice_direct: %d", ret);
++ if (ret < 0) {
++ goto out_fput_both;
++ } else if (ret < size) {
++ ret = -ENOSPC;
++ goto out_fput_both;
++ } else {
++ struct inode *old_inode = old_path.dentry->d_inode;
++ struct inode *new_inode = new_path.dentry->d_inode;
++ struct iattr attr = {
++ .ia_uid = old_inode->i_uid,
++ .ia_gid = old_inode->i_gid,
++ .ia_valid = ATTR_UID | ATTR_GID
++ };
++
++ ret = inode_setattr(new_inode, &attr);
++ if (ret)
++ goto out_fput_both;
++ }
++
++ mutex_lock(&old_path.dentry->d_inode->i_sb->s_vfs_rename_mutex);
++
++ /* drop out late */
++ ret = -ENOENT;
++ if ((redo = d_unhashed(old_path.dentry)))
++ goto out_unlock;
++
++ vxdprintk(VXD_CBIT(misc, 2),
++ "vfs_rename: [»%*s«:%d] -> [»%*s«:%d]",
++ new_path.dentry->d_name.len, new_path.dentry->d_name.name,
++ new_path.dentry->d_name.len,
++ old_path.dentry->d_name.len, old_path.dentry->d_name.name,
++ old_path.dentry->d_name.len);
++ ret = vfs_rename(dir_nd.path.dentry->d_inode, new_path.dentry,
++ old_nd.path.dentry->d_parent->d_inode, old_path.dentry);
++ vxdprintk(VXD_CBIT(misc, 2), "vfs_rename: %d", ret);
++ res = new_path.dentry;
++
++out_unlock:
++ mutex_unlock(&old_path.dentry->d_inode->i_sb->s_vfs_rename_mutex);
++
++out_fput_both:
++ vxdprintk(VXD_CBIT(misc, 3),
++ "fput(new_file=%p[#%d])", new_file,
++ atomic_read(&new_file->f_count));
++ fput(new_file);
++
++out_fput_old:
++ vxdprintk(VXD_CBIT(misc, 3),
++ "fput(old_file=%p[#%d])", old_file,
++ atomic_read(&old_file->f_count));
++ fput(old_file);
++
++out_unlock_new:
++ mutex_unlock(&dir->d_inode->i_mutex);
++ if (!ret)
++ goto out_redo;
++
++ /* error path cleanup */
++ vfs_unlink(dir->d_inode, new_path.dentry);
++ dput(new_path.dentry);
++
++out_redo:
++ if (!redo)
++ goto out_rel_both;
++ /* lookup dentry once again */
++ path_put(&old_nd.path);
++ ret = path_lookup(pathname, LOOKUP_FOLLOW, &old_nd);
++ if (ret)
++ goto out_rel_both;
++
++ new_path.dentry = old_nd.path.dentry;
++ vxdprintk(VXD_CBIT(misc, 2),
++ "path_lookup(redo): %p [»%.*s«:%d]", new_path.dentry,
++ new_path.dentry->d_name.len, new_path.dentry->d_name.name,
++ new_path.dentry->d_name.len);
++ dget(new_path.dentry);
++ res = new_path.dentry;
++
++out_rel_both:
++ path_put(&dir_nd.path);
++out_rel_old:
++ path_put(&old_nd.path);
++out_free_path:
++ kfree(path);
++out:
++ if (ret)
++ res = ERR_PTR(ret);
++ return res;
++}
++
++#endif
++
+ /* get the link contents into pagecache */
+ static char *page_getlink(struct dentry * dentry, struct page **ppage)
+ {
+--- a/fs/namespace.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/namespace.c 2008-07-17 18:08:01.000000000 -0400
+@@ -27,6 +27,11 @@
+ #include <linux/ramfs.h>
+ #include <linux/log2.h>
+ #include <linux/idr.h>
++#include <linux/vs_base.h>
++#include <linux/vs_context.h>
++#include <linux/vs_tag.h>
++#include <linux/vserver/space.h>
++#include <linux/vserver/global.h>
+ #include <asm/uaccess.h>
+ #include <asm/unistd.h>
+ #include "pnode.h"
+@@ -572,6 +577,7 @@ static struct vfsmount *clone_mnt(struct
+ mnt->mnt_root = dget(root);
+ mnt->mnt_mountpoint = mnt->mnt_root;
+ mnt->mnt_parent = mnt;
++ mnt->mnt_tag = old->mnt_tag;
+
+ if (flag & CL_SLAVE) {
+ list_add(&mnt->mnt_slave, &old->mnt_slave_list);
+@@ -684,6 +690,31 @@ static inline void mangle(struct seq_fil
+ seq_escape(m, s, " \t\n\\");
+ }
+
++static int mnt_is_reachable(struct vfsmount *mnt)
++{
++ struct path root;
++ struct dentry *point;
++ int ret;
++
++ if (mnt == mnt->mnt_ns->root)
++ return 1;
++
++ spin_lock(&vfsmount_lock);
++ root = current->fs->root;
++ point = root.dentry;
++
++ while ((mnt != mnt->mnt_parent) && (mnt != root.mnt)) {
++ point = mnt->mnt_mountpoint;
++ mnt = mnt->mnt_parent;
++ }
++
++ ret = (mnt == root.mnt) && is_subdir(point, root.dentry);
++
++ spin_unlock(&vfsmount_lock);
++
++ return ret;
++}
++
+ /*
+ * Simple .show_options callback for filesystems which don't want to
+ * implement more complex mount option showing.
+@@ -756,6 +787,7 @@ static void show_sb_opts(struct seq_file
+ { MS_SYNCHRONOUS, ",sync" },
+ { MS_DIRSYNC, ",dirsync" },
+ { MS_MANDLOCK, ",mand" },
++ { MS_TAGGED, ",tag" },
+ { 0, NULL }
+ };
+ const struct proc_fs_info *fs_infop;
+@@ -885,17 +917,27 @@ static int show_vfsstat(struct seq_file
+ struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt };
+ int err = 0;
+
+- /* device */
+- if (mnt->mnt_devname) {
+- seq_puts(m, "device ");
+- mangle(m, mnt->mnt_devname);
+- } else
+- seq_puts(m, "no device");
++ if (vx_flags(VXF_HIDE_MOUNT, 0))
++ return 0;
++ if (!mnt_is_reachable(mnt) && !vx_check(0, VS_WATCH_P))
++ return 0;
+
+- /* mount point */
+- seq_puts(m, " mounted on ");
+- seq_path(m, &mnt_path, " \t\n\\");
+- seq_putc(m, ' ');
++ if (!vx_check(0, VS_ADMIN|VS_WATCH) &&
++ mnt == current->fs->root.mnt) {
++ seq_puts(m, "device /dev/root mounted on / ");
++ } else {
++ /* device */
++ if (mnt->mnt_devname) {
++ seq_puts(m, "device ");
++ mangle(m, mnt->mnt_devname);
++ } else
++ seq_puts(m, "no device");
++
++ /* mount point */
++ seq_puts(m, " mounted on ");
++ seq_path(m, &mnt_path, " \t\n\\");
++ seq_putc(m, ' ');
++ }
+
+ /* file system type */
+ seq_puts(m, "with fstype ");
+@@ -1134,7 +1176,7 @@ asmlinkage long sys_umount(char __user *
+ goto dput_and_out;
+
+ retval = -EPERM;
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT))
+ goto dput_and_out;
+
+ retval = do_umount(nd.path.mnt, flags);
+@@ -1160,7 +1202,7 @@ asmlinkage long sys_oldumount(char __use
+
+ static int mount_is_safe(struct nameidata *nd)
+ {
+- if (capable(CAP_SYS_ADMIN))
++ if (vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT))
+ return 0;
+ return -EPERM;
+ #ifdef notyet
+@@ -1453,11 +1495,13 @@ static noinline int do_change_type(struc
+ * noinline this do_mount helper to save do_mount stack space.
+ */
+ static noinline int do_loopback(struct nameidata *nd, char *old_name,
+- int recurse)
++ tag_t tag, unsigned long flags, int mnt_flags)
+ {
+ struct nameidata old_nd;
+ struct vfsmount *mnt = NULL;
+ int err = mount_is_safe(nd);
++ int recurse = flags & MS_REC;
++
+ if (err)
+ return err;
+ if (!old_name || !*old_name)
+@@ -1491,6 +1535,7 @@ static noinline int do_loopback(struct n
+ spin_unlock(&vfsmount_lock);
+ release_mounts(&umount_list);
+ }
++ mnt->mnt_flags = mnt_flags;
+
+ out:
+ up_write(&namespace_sem);
+@@ -1522,12 +1567,12 @@ static int change_mount_flags(struct vfs
+ * noinline this do_mount helper to save do_mount stack space.
+ */
+ static noinline int do_remount(struct nameidata *nd, int flags, int mnt_flags,
+- void *data)
++ void *data, xid_t xid)
+ {
+ int err;
+ struct super_block *sb = nd->path.mnt->mnt_sb;
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_REMOUNT))
+ return -EPERM;
+
+ if (!check_mnt(nd->path.mnt))
+@@ -1568,7 +1613,7 @@ static noinline int do_move_mount(struct
+ struct path parent_path;
+ struct vfsmount *p;
+ int err = 0;
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT))
+ return -EPERM;
+ if (!old_name || !*old_name)
+ return -EINVAL;
+@@ -1651,7 +1696,7 @@ static noinline int do_new_mount(struct
+ return -EINVAL;
+
+ /* we need capabilities... */
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SECURE_MOUNT))
+ return -EPERM;
+
+ mnt = do_kern_mount(type, flags, name, data);
+@@ -1896,6 +1941,7 @@ long do_mount(char *dev_name, char *dir_
+ struct nameidata nd;
+ int retval = 0;
+ int mnt_flags = 0;
++ tag_t tag = 0;
+
+ /* Discard magic */
+ if ((flags & MS_MGC_MSK) == MS_MGC_VAL)
+@@ -1911,6 +1957,14 @@ long do_mount(char *dev_name, char *dir_
+ if (data_page)
+ ((char *)data_page)[PAGE_SIZE - 1] = 0;
+
++ retval = dx_parse_tag(data_page, &tag, 1);
++ if (retval) {
++ mnt_flags |= retval;
++ /* FIXME: bind and re-mounts get the tag flag? */
++ if (flags & (MS_BIND|MS_REMOUNT))
++ flags |= MS_TAGID;
++ }
++
+ /* Separate the per-mountpoint flags */
+ if (flags & MS_NOSUID)
+ mnt_flags |= MNT_NOSUID;
+@@ -1927,6 +1981,8 @@ long do_mount(char *dev_name, char *dir_
+ if (flags & MS_RDONLY)
+ mnt_flags |= MNT_READONLY;
+
++ if (!capable(CAP_SYS_ADMIN))
++ mnt_flags |= MNT_NODEV;
+ flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE |
+ MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT);
+
+@@ -1942,9 +1998,9 @@ long do_mount(char *dev_name, char *dir_
+
+ if (flags & MS_REMOUNT)
+ retval = do_remount(&nd, flags & ~MS_REMOUNT, mnt_flags,
+- data_page);
++ data_page, tag);
+ else if (flags & MS_BIND)
+- retval = do_loopback(&nd, dev_name, flags & MS_REC);
++ retval = do_loopback(&nd, dev_name, tag, flags, mnt_flags);
+ else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
+ retval = do_change_type(&nd, flags);
+ else if (flags & MS_MOVE)
+@@ -2017,6 +2073,7 @@ static struct mnt_namespace *dup_mnt_ns(
+ q = next_mnt(q, new_ns->root);
+ }
+ up_write(&namespace_sem);
++ atomic_inc(&vs_global_mnt_ns);
+
+ if (rootmnt)
+ mntput(rootmnt);
+@@ -2349,5 +2406,6 @@ void __put_mnt_ns(struct mnt_namespace *
+ spin_unlock(&vfsmount_lock);
+ up_write(&namespace_sem);
+ release_mounts(&umount_list);
++ atomic_dec(&vs_global_mnt_ns);
+ kfree(ns);
+ }
+--- a/fs/nfs/client.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/nfs/client.c 2008-07-16 22:41:36.000000000 -0400
+@@ -598,6 +598,9 @@ static int nfs_init_server_rpcclient(str
+ if (server->flags & NFS_MOUNT_SOFT)
+ server->client->cl_softrtry = 1;
+
++ server->client->cl_tag = 0;
++ if (server->flags & NFS_MOUNT_TAGGED)
++ server->client->cl_tag = 1;
+ return 0;
+ }
+
+@@ -763,6 +766,10 @@ static void nfs_server_set_fsinfo(struct
+ server->acdirmin = server->acdirmax = 0;
+ }
+
++ /* FIXME: needs fsinfo
++ if (server->flags & NFS_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED; */
++
+ server->maxfilesize = fsinfo->maxfilesize;
+
+ /* We're airborne Set socket buffersize */
+--- a/fs/nfs/dir.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/nfs/dir.c 2008-07-16 22:41:36.000000000 -0400
+@@ -34,6 +34,7 @@
+ #include <linux/namei.h>
+ #include <linux/mount.h>
+ #include <linux/sched.h>
++#include <linux/vs_tag.h>
+
+ #include "nfs4_fs.h"
+ #include "delegation.h"
+@@ -930,6 +931,7 @@ static struct dentry *nfs_lookup(struct
+ if (IS_ERR(res))
+ goto out_unblock_sillyrename;
+
++ dx_propagate_tag(nd, inode);
+ no_entry:
+ res = d_materialise_unique(dentry, inode);
+ if (res != NULL) {
+--- a/fs/nfs/inode.c 2008-08-12 01:41:51.000000000 -0400
++++ a/fs/nfs/inode.c 2008-08-12 01:42:21.000000000 -0400
+@@ -37,6 +37,7 @@
+ #include <linux/vfs.h>
+ #include <linux/inet.h>
+ #include <linux/nfs_xdr.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/system.h>
+ #include <asm/uaccess.h>
+@@ -314,8 +315,10 @@ nfs_fhget(struct super_block *sb, struct
+ nfsi->change_attr = fattr->change_attr;
+ inode->i_size = nfs_size_to_loff_t(fattr->size);
+ inode->i_nlink = fattr->nlink;
+- inode->i_uid = fattr->uid;
+- inode->i_gid = fattr->gid;
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), fattr->uid, fattr->gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), fattr->uid, fattr->gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), fattr->uid, fattr->gid, 0);
++ /* maybe fattr->xid someday */
+ if (fattr->valid & (NFS_ATTR_FATTR_V3 | NFS_ATTR_FATTR_V4)) {
+ /*
+ * report the blocks in 512byte units
+@@ -408,6 +411,8 @@ void nfs_setattr_update_inode(struct ino
+ inode->i_uid = attr->ia_uid;
+ if ((attr->ia_valid & ATTR_GID) != 0)
+ inode->i_gid = attr->ia_gid;
++ if ((attr->ia_valid & ATTR_TAG) && IS_TAGGED(inode))
++ inode->i_tag = attr->ia_tag;
+ spin_lock(&inode->i_lock);
+ NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+ spin_unlock(&inode->i_lock);
+@@ -849,6 +854,9 @@ static int nfs_check_inode_attributes(st
+ struct nfs_inode *nfsi = NFS_I(inode);
+ loff_t cur_size, new_isize;
+ unsigned long invalid = 0;
++ uid_t uid;
++ gid_t gid;
++ tag_t tag;
+
+
+ /* Has the inode gone and changed behind our back? */
+@@ -873,10 +881,15 @@ static int nfs_check_inode_attributes(st
+ if (cur_size != new_isize && nfsi->npages == 0)
+ invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
+
++ uid = INOTAG_UID(DX_TAG(inode), fattr->uid, fattr->gid);
++ gid = INOTAG_GID(DX_TAG(inode), fattr->uid, fattr->gid);
++ tag = INOTAG_TAG(DX_TAG(inode), fattr->uid, fattr->gid, 0);
++
+ /* Have any file permissions changed? */
+ if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO)
+- || inode->i_uid != fattr->uid
+- || inode->i_gid != fattr->gid)
++ || inode->i_uid != uid
++ || inode->i_gid != gid
++ || inode->i_tag != tag)
+ invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL;
+
+ /* Has the link count changed? */
+@@ -997,6 +1010,9 @@ static int nfs_update_inode(struct inode
+ loff_t cur_isize, new_isize;
+ unsigned long invalid = 0;
+ unsigned long now = jiffies;
++ uid_t uid;
++ gid_t gid;
++ tag_t tag;
+
+ dfprintk(VFS, "NFS: %s(%s/%ld ct=%d info=0x%x)\n",
+ __func__, inode->i_sb->s_id, inode->i_ino,
+@@ -1070,15 +1086,21 @@ static int nfs_update_inode(struct inode
+ memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime));
+ nfsi->change_attr = fattr->change_attr;
+
++ uid = INOTAG_UID(DX_TAG(inode), fattr->uid, fattr->gid);
++ gid = INOTAG_GID(DX_TAG(inode), fattr->uid, fattr->gid);
++ tag = INOTAG_TAG(DX_TAG(inode), fattr->uid, fattr->gid, 0);
++
+ if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO) ||
+- inode->i_uid != fattr->uid ||
+- inode->i_gid != fattr->gid)
++ inode->i_uid != uid ||
++ inode->i_gid != gid ||
++ inode->i_tag != tag)
+ invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+
+ inode->i_mode = fattr->mode;
+ inode->i_nlink = fattr->nlink;
+- inode->i_uid = fattr->uid;
+- inode->i_gid = fattr->gid;
++ inode->i_uid = uid;
++ inode->i_gid = gid;
++ inode->i_tag = tag;
+
+ if (fattr->valid & (NFS_ATTR_FATTR_V3 | NFS_ATTR_FATTR_V4)) {
+ /*
+--- a/fs/nfs/nfs3xdr.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/nfs/nfs3xdr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -22,6 +22,7 @@
+ #include <linux/nfs3.h>
+ #include <linux/nfs_fs.h>
+ #include <linux/nfsacl.h>
++#include <linux/vs_tag.h>
+ #include "internal.h"
+
+ #define NFSDBG_FACILITY NFSDBG_XDR
+@@ -180,7 +181,7 @@ xdr_decode_fattr(__be32 *p, struct nfs_f
+ }
+
+ static inline __be32 *
+-xdr_encode_sattr(__be32 *p, struct iattr *attr)
++xdr_encode_sattr(__be32 *p, struct iattr *attr, int tag)
+ {
+ if (attr->ia_valid & ATTR_MODE) {
+ *p++ = xdr_one;
+@@ -188,15 +189,17 @@ xdr_encode_sattr(__be32 *p, struct iattr
+ } else {
+ *p++ = xdr_zero;
+ }
+- if (attr->ia_valid & ATTR_UID) {
++ if (attr->ia_valid & ATTR_UID ||
++ (tag && (attr->ia_valid & ATTR_TAG))) {
+ *p++ = xdr_one;
+- *p++ = htonl(attr->ia_uid);
++ *p++ = htonl(TAGINO_UID(tag, attr->ia_uid, attr->ia_tag));
+ } else {
+ *p++ = xdr_zero;
+ }
+- if (attr->ia_valid & ATTR_GID) {
++ if (attr->ia_valid & ATTR_GID ||
++ (tag && (attr->ia_valid & ATTR_TAG))) {
+ *p++ = xdr_one;
+- *p++ = htonl(attr->ia_gid);
++ *p++ = htonl(TAGINO_GID(tag, attr->ia_gid, attr->ia_tag));
+ } else {
+ *p++ = xdr_zero;
+ }
+@@ -281,7 +284,8 @@ static int
+ nfs3_xdr_sattrargs(struct rpc_rqst *req, __be32 *p, struct nfs3_sattrargs *args)
+ {
+ p = xdr_encode_fhandle(p, args->fh);
+- p = xdr_encode_sattr(p, args->sattr);
++ p = xdr_encode_sattr(p, args->sattr,
++ req->rq_task->tk_client->cl_tag);
+ *p++ = htonl(args->guard);
+ if (args->guard)
+ p = xdr_encode_time3(p, &args->guardtime);
+@@ -386,7 +390,8 @@ nfs3_xdr_createargs(struct rpc_rqst *req
+ *p++ = args->verifier[0];
+ *p++ = args->verifier[1];
+ } else
+- p = xdr_encode_sattr(p, args->sattr);
++ p = xdr_encode_sattr(p, args->sattr,
++ req->rq_task->tk_client->cl_tag);
+
+ req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+ return 0;
+@@ -400,7 +405,8 @@ nfs3_xdr_mkdirargs(struct rpc_rqst *req,
+ {
+ p = xdr_encode_fhandle(p, args->fh);
+ p = xdr_encode_array(p, args->name, args->len);
+- p = xdr_encode_sattr(p, args->sattr);
++ p = xdr_encode_sattr(p, args->sattr,
++ req->rq_task->tk_client->cl_tag);
+ req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+ return 0;
+ }
+@@ -413,7 +419,8 @@ nfs3_xdr_symlinkargs(struct rpc_rqst *re
+ {
+ p = xdr_encode_fhandle(p, args->fromfh);
+ p = xdr_encode_array(p, args->fromname, args->fromlen);
+- p = xdr_encode_sattr(p, args->sattr);
++ p = xdr_encode_sattr(p, args->sattr,
++ req->rq_task->tk_client->cl_tag);
+ *p++ = htonl(args->pathlen);
+ req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+
+@@ -431,7 +438,8 @@ nfs3_xdr_mknodargs(struct rpc_rqst *req,
+ p = xdr_encode_fhandle(p, args->fh);
+ p = xdr_encode_array(p, args->name, args->len);
+ *p++ = htonl(args->type);
+- p = xdr_encode_sattr(p, args->sattr);
++ p = xdr_encode_sattr(p, args->sattr,
++ req->rq_task->tk_client->cl_tag);
+ if (args->type == NF3CHR || args->type == NF3BLK) {
+ *p++ = htonl(MAJOR(args->rdev));
+ *p++ = htonl(MINOR(args->rdev));
+--- a/fs/nfs/nfsroot.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/nfs/nfsroot.c 2008-07-16 22:41:36.000000000 -0400
+@@ -119,12 +119,12 @@ static int mount_port __initdata = 0; /
+ enum {
+ /* Options that take integer arguments */
+ Opt_port, Opt_rsize, Opt_wsize, Opt_timeo, Opt_retrans, Opt_acregmin,
+- Opt_acregmax, Opt_acdirmin, Opt_acdirmax,
++ Opt_acregmax, Opt_acdirmin, Opt_acdirmax, Opt_tagid,
+ /* Options that take no arguments */
+ Opt_soft, Opt_hard, Opt_intr,
+ Opt_nointr, Opt_posix, Opt_noposix, Opt_cto, Opt_nocto, Opt_ac,
+ Opt_noac, Opt_lock, Opt_nolock, Opt_v2, Opt_v3, Opt_udp, Opt_tcp,
+- Opt_acl, Opt_noacl,
++ Opt_acl, Opt_noacl, Opt_tag, Opt_notag,
+ /* Error token */
+ Opt_err
+ };
+@@ -161,6 +161,9 @@ static match_table_t __initdata tokens =
+ {Opt_tcp, "tcp"},
+ {Opt_acl, "acl"},
+ {Opt_noacl, "noacl"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
+ {Opt_err, NULL}
+
+ };
+@@ -272,6 +275,20 @@ static int __init root_nfs_parse(char *n
+ case Opt_noacl:
+ nfs_data.flags |= NFS_MOUNT_NOACL;
+ break;
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ nfs_data.flags |= NFS_MOUNT_TAGGED;
++ break;
++ case Opt_notag:
++ nfs_data.flags &= ~NFS_MOUNT_TAGGED;
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ nfs_data.flags |= NFS_MOUNT_TAGGED;
++ break;
++#endif
+ default:
+ printk(KERN_WARNING "Root-NFS: unknown "
+ "option: %s\n", p);
+--- a/fs/nfs/super.c 2008-07-14 17:22:49.000000000 -0400
++++ a/fs/nfs/super.c 2008-07-16 22:41:36.000000000 -0400
+@@ -50,6 +50,7 @@
+ #include <linux/nfs_xdr.h>
+ #include <linux/magic.h>
+ #include <linux/parser.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/system.h>
+ #include <asm/uaccess.h>
+@@ -502,6 +503,7 @@ static void nfs_show_mount_options(struc
+ { NFS_MOUNT_NOACL, ",noacl", "" },
+ { NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" },
+ { NFS_MOUNT_UNSHARED, ",nosharecache", ""},
++ { NFS_MOUNT_TAGGED, ",tag", "" },
+ { 0, NULL, NULL }
+ };
+ const struct proc_nfs_info *nfs_infop;
+--- a/fs/nfsd/auth.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/nfsd/auth.c 2008-07-17 17:19:21.000000000 -0400
+@@ -10,6 +10,7 @@
+ #include <linux/sunrpc/svcauth.h>
+ #include <linux/nfsd/nfsd.h>
+ #include <linux/nfsd/export.h>
++#include <linux/vs_tag.h>
+ #include "auth.h"
+
+ int nfsexp_flags(struct svc_rqst *rqstp, struct svc_export *exp)
+@@ -55,19 +56,23 @@ int nfsd_setuser(struct svc_rqst *rqstp,
+ get_group_info(cred.cr_group_info);
+
+ if (cred.cr_uid != (uid_t) -1)
+- current->fsuid = cred.cr_uid;
++ current->fsuid = INOTAG_UID(DX_TAG_NFSD, cred.cr_uid, cred.cr_gid);
+ else
+ current->fsuid = exp->ex_anon_uid;
+ if (cred.cr_gid != (gid_t) -1)
+- current->fsgid = cred.cr_gid;
++ current->fsgid = INOTAG_GID(DX_TAG_NFSD, cred.cr_uid, cred.cr_gid);
+ else
+ current->fsgid = exp->ex_anon_gid;
+
++ /* this desperately needs a tag :) */
++ current->xid = (xid_t)INOTAG_TAG(DX_TAG_NFSD, cred.cr_uid, cred.cr_gid, 0);
++
+ if (!cred.cr_group_info)
+ return -ENOMEM;
+ ret = set_current_groups(cred.cr_group_info);
+ put_group_info(cred.cr_group_info);
+- if ((cred.cr_uid)) {
++
++ if (INOTAG_UID(DX_TAG_NFSD, cred.cr_uid, cred.cr_gid)) {
+ current->cap_effective =
+ cap_drop_nfsd_set(current->cap_effective);
+ } else {
+--- a/fs/nfsd/nfs3xdr.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/nfsd/nfs3xdr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/sunrpc/svc.h>
+ #include <linux/nfsd/nfsd.h>
+ #include <linux/nfsd/xdr3.h>
++#include <linux/vs_tag.h>
+ #include "auth.h"
+
+ #define NFSDDBG_FACILITY NFSDDBG_XDR
+@@ -108,6 +109,8 @@ static __be32 *
+ decode_sattr3(__be32 *p, struct iattr *iap)
+ {
+ u32 tmp;
++ uid_t uid = 0;
++ gid_t gid = 0;
+
+ iap->ia_valid = 0;
+
+@@ -117,12 +120,15 @@ decode_sattr3(__be32 *p, struct iattr *i
+ }
+ if (*p++) {
+ iap->ia_valid |= ATTR_UID;
+- iap->ia_uid = ntohl(*p++);
++ uid = ntohl(*p++);
+ }
+ if (*p++) {
+ iap->ia_valid |= ATTR_GID;
+- iap->ia_gid = ntohl(*p++);
++ gid = ntohl(*p++);
+ }
++ iap->ia_uid = INOTAG_UID(DX_TAG_NFSD, uid, gid);
++ iap->ia_gid = INOTAG_GID(DX_TAG_NFSD, uid, gid);
++ iap->ia_tag = INOTAG_TAG(DX_TAG_NFSD, uid, gid, 0);
+ if (*p++) {
+ u64 newsize;
+
+@@ -178,8 +184,12 @@ encode_fattr3(struct svc_rqst *rqstp, __
+ *p++ = htonl(nfs3_ftypes[(stat->mode & S_IFMT) >> 12]);
+ *p++ = htonl((u32) stat->mode);
+ *p++ = htonl((u32) stat->nlink);
+- *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid));
+- *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid));
++ *p++ = htonl((u32) nfsd_ruid(rqstp,
++ TAGINO_UID(0 /* FIXME: DX_TAG(dentry->d_inode) */,
++ stat->uid, stat->tag)));
++ *p++ = htonl((u32) nfsd_rgid(rqstp,
++ TAGINO_GID(0 /* FIXME: DX_TAG(dentry->d_inode) */,
++ stat->gid, stat->tag)));
+ if (S_ISLNK(stat->mode) && stat->size > NFS3_MAXPATHLEN) {
+ p = xdr_encode_hyper(p, (u64) NFS3_MAXPATHLEN);
+ } else {
+--- a/fs/nfsd/nfs4xdr.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/nfsd/nfs4xdr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -58,6 +58,7 @@
+ #include <linux/nfs4_acl.h>
+ #include <linux/sunrpc/gss_api.h>
+ #include <linux/sunrpc/svcauth_gss.h>
++#include <linux/vs_tag.h>
+
+ #define NFSDDBG_FACILITY NFSDDBG_XDR
+
+@@ -1745,14 +1746,18 @@ out_acl:
+ WRITE32(stat.nlink);
+ }
+ if (bmval1 & FATTR4_WORD1_OWNER) {
+- status = nfsd4_encode_user(rqstp, stat.uid, &p, &buflen);
++ status = nfsd4_encode_user(rqstp,
++ TAGINO_UID(DX_TAG(dentry->d_inode),
++ stat.uid, stat.tag), &p, &buflen);
+ if (status == nfserr_resource)
+ goto out_resource;
+ if (status)
+ goto out;
+ }
+ if (bmval1 & FATTR4_WORD1_OWNER_GROUP) {
+- status = nfsd4_encode_group(rqstp, stat.gid, &p, &buflen);
++ status = nfsd4_encode_group(rqstp,
++ TAGINO_GID(DX_TAG(dentry->d_inode),
++ stat.gid, stat.tag), &p, &buflen);
+ if (status == nfserr_resource)
+ goto out_resource;
+ if (status)
+--- a/fs/nfsd/nfsxdr.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/nfsd/nfsxdr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -15,6 +15,7 @@
+ #include <linux/nfsd/nfsd.h>
+ #include <linux/nfsd/xdr.h>
+ #include <linux/mm.h>
++#include <linux/vs_tag.h>
+ #include "auth.h"
+
+ #define NFSDDBG_FACILITY NFSDDBG_XDR
+@@ -98,6 +99,8 @@ static __be32 *
+ decode_sattr(__be32 *p, struct iattr *iap)
+ {
+ u32 tmp, tmp1;
++ uid_t uid = 0;
++ gid_t gid = 0;
+
+ iap->ia_valid = 0;
+
+@@ -111,12 +114,15 @@ decode_sattr(__be32 *p, struct iattr *ia
+ }
+ if ((tmp = ntohl(*p++)) != (u32)-1) {
+ iap->ia_valid |= ATTR_UID;
+- iap->ia_uid = tmp;
++ uid = tmp;
+ }
+ if ((tmp = ntohl(*p++)) != (u32)-1) {
+ iap->ia_valid |= ATTR_GID;
+- iap->ia_gid = tmp;
++ gid = tmp;
+ }
++ iap->ia_uid = INOTAG_UID(DX_TAG_NFSD, uid, gid);
++ iap->ia_gid = INOTAG_GID(DX_TAG_NFSD, uid, gid);
++ iap->ia_tag = INOTAG_TAG(DX_TAG_NFSD, uid, gid, 0);
+ if ((tmp = ntohl(*p++)) != (u32)-1) {
+ iap->ia_valid |= ATTR_SIZE;
+ iap->ia_size = tmp;
+@@ -161,8 +167,10 @@ encode_fattr(struct svc_rqst *rqstp, __b
+ *p++ = htonl(nfs_ftypes[type >> 12]);
+ *p++ = htonl((u32) stat->mode);
+ *p++ = htonl((u32) stat->nlink);
+- *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid));
+- *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid));
++ *p++ = htonl((u32) nfsd_ruid(rqstp,
++ TAGINO_UID(DX_TAG(dentry->d_inode), stat->uid, stat->tag)));
++ *p++ = htonl((u32) nfsd_rgid(rqstp,
++ TAGINO_GID(DX_TAG(dentry->d_inode), stat->gid, stat->tag)));
+
+ if (S_ISLNK(type) && stat->size > NFS_MAXPATHLEN) {
+ *p++ = htonl(NFS_MAXPATHLEN);
+--- a/fs/ocfs2/dlm/dlmfs.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/dlm/dlmfs.c 2008-07-16 22:41:36.000000000 -0400
+@@ -43,6 +43,7 @@
+ #include <linux/init.h>
+ #include <linux/string.h>
+ #include <linux/backing-dev.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/uaccess.h>
+
+@@ -342,6 +343,7 @@ static struct inode *dlmfs_get_root_inod
+ inode->i_mode = mode;
+ inode->i_uid = current->fsuid;
+ inode->i_gid = current->fsgid;
++ inode->i_tag = dx_current_fstag(sb);
+ inode->i_blocks = 0;
+ inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info;
+ inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
+@@ -368,6 +370,7 @@ static struct inode *dlmfs_get_inode(str
+ inode->i_mode = mode;
+ inode->i_uid = current->fsuid;
+ inode->i_gid = current->fsgid;
++ inode->i_tag = dx_current_fstag(sb);
+ inode->i_blocks = 0;
+ inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info;
+ inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
+--- a/fs/ocfs2/dlmglue.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/dlmglue.c 2008-07-16 22:41:36.000000000 -0400
+@@ -1769,6 +1769,7 @@ static void __ocfs2_stuff_meta_lvb(struc
+ lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters);
+ lvb->lvb_iuid = cpu_to_be32(inode->i_uid);
+ lvb->lvb_igid = cpu_to_be32(inode->i_gid);
++ lvb->lvb_itag = cpu_to_be16(inode->i_tag);
+ lvb->lvb_imode = cpu_to_be16(inode->i_mode);
+ lvb->lvb_inlink = cpu_to_be16(inode->i_nlink);
+ lvb->lvb_iatime_packed =
+@@ -1823,6 +1824,7 @@ static void ocfs2_refresh_inode_from_lvb
+
+ inode->i_uid = be32_to_cpu(lvb->lvb_iuid);
+ inode->i_gid = be32_to_cpu(lvb->lvb_igid);
++ inode->i_tag = be16_to_cpu(lvb->lvb_itag);
+ inode->i_mode = be16_to_cpu(lvb->lvb_imode);
+ inode->i_nlink = be16_to_cpu(lvb->lvb_inlink);
+ ocfs2_unpack_timespec(&inode->i_atime,
+--- a/fs/ocfs2/dlmglue.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/dlmglue.h 2008-07-16 22:41:36.000000000 -0400
+@@ -46,7 +46,8 @@ struct ocfs2_meta_lvb {
+ __be16 lvb_inlink;
+ __be32 lvb_iattr;
+ __be32 lvb_igeneration;
+- __be32 lvb_reserved2;
++ __be16 lvb_itag;
++ __be16 lvb_reserved2;
+ };
+
+ /* ocfs2_inode_lock_full() 'arg_flags' flags */
+--- a/fs/ocfs2/file.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/file.c 2008-07-16 22:41:36.000000000 -0400
+@@ -1058,13 +1058,15 @@ int ocfs2_setattr(struct dentry *dentry,
+ mlog(0, "uid change: %d\n", attr->ia_uid);
+ if (attr->ia_valid & ATTR_GID)
+ mlog(0, "gid change: %d\n", attr->ia_gid);
++ if (attr->ia_valid & ATTR_TAG)
++ mlog(0, "tag change: %d\n", attr->ia_tag);
+ if (attr->ia_valid & ATTR_SIZE)
+ mlog(0, "size change...\n");
+ if (attr->ia_valid & (ATTR_ATIME | ATTR_MTIME | ATTR_CTIME))
+ mlog(0, "time change...\n");
+
+ #define OCFS2_VALID_ATTRS (ATTR_ATIME | ATTR_MTIME | ATTR_CTIME | ATTR_SIZE \
+- | ATTR_GID | ATTR_UID | ATTR_MODE)
++ | ATTR_GID | ATTR_UID | ATTR_TAG | ATTR_MODE)
+ if (!(attr->ia_valid & OCFS2_VALID_ATTRS)) {
+ mlog(0, "can't handle attrs: 0x%x\n", attr->ia_valid);
+ return 0;
+@@ -2233,6 +2235,7 @@ const struct inode_operations ocfs2_file
+ const struct inode_operations ocfs2_special_file_iops = {
+ .setattr = ocfs2_setattr,
+ .getattr = ocfs2_getattr,
++ .sync_flags = ocfs2_sync_flags,
+ .permission = ocfs2_permission,
+ };
+
+--- a/fs/ocfs2/inode.c 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ocfs2/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -28,6 +28,7 @@
+ #include <linux/slab.h>
+ #include <linux/highmem.h>
+ #include <linux/pagemap.h>
++#include <linux/vs_tag.h>
+
+ #include <asm/byteorder.h>
+
+@@ -42,6 +43,7 @@
+ #include "file.h"
+ #include "heartbeat.h"
+ #include "inode.h"
++#include "ioctl.h"
+ #include "journal.h"
+ #include "namei.h"
+ #include "suballoc.h"
+@@ -74,11 +76,13 @@ void ocfs2_set_inode_flags(struct inode
+ {
+ unsigned int flags = OCFS2_I(inode)->ip_attr;
+
+- inode->i_flags &= ~(S_IMMUTABLE |
++ inode->i_flags &= ~(S_IMMUTABLE | S_IXUNLINK |
+ S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
+
+ if (flags & OCFS2_IMMUTABLE_FL)
+ inode->i_flags |= S_IMMUTABLE;
++ if (flags & OCFS2_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
+
+ if (flags & OCFS2_SYNC_FL)
+ inode->i_flags |= S_SYNC;
+@@ -88,25 +92,89 @@ void ocfs2_set_inode_flags(struct inode
+ inode->i_flags |= S_NOATIME;
+ if (flags & OCFS2_DIRSYNC_FL)
+ inode->i_flags |= S_DIRSYNC;
++
++ inode->i_vflags &= ~(V_BARRIER | V_COW);
++
++ if (flags & OCFS2_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ if (flags & OCFS2_COW_FL)
++ inode->i_vflags |= V_COW;
+ }
+
+ /* Propagate flags from i_flags to OCFS2_I(inode)->ip_attr */
+ void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi)
+ {
+ unsigned int flags = oi->vfs_inode.i_flags;
++ unsigned int vflags = oi->vfs_inode.i_vflags;
++
++ oi->ip_attr &= ~(OCFS2_SYNC_FL | OCFS2_APPEND_FL |
++ OCFS2_IMMUTABLE_FL | OCFS2_IXUNLINK_FL |
++ OCFS2_NOATIME_FL | OCFS2_DIRSYNC_FL |
++ OCFS2_BARRIER_FL | OCFS2_COW_FL);
++
++ if (flags & S_IMMUTABLE)
++ oi->ip_attr |= OCFS2_IMMUTABLE_FL;
++ if (flags & S_IXUNLINK)
++ oi->ip_attr |= OCFS2_IXUNLINK_FL;
+
+- oi->ip_attr &= ~(OCFS2_SYNC_FL|OCFS2_APPEND_FL|
+- OCFS2_IMMUTABLE_FL|OCFS2_NOATIME_FL|OCFS2_DIRSYNC_FL);
+ if (flags & S_SYNC)
+ oi->ip_attr |= OCFS2_SYNC_FL;
+ if (flags & S_APPEND)
+ oi->ip_attr |= OCFS2_APPEND_FL;
+- if (flags & S_IMMUTABLE)
+- oi->ip_attr |= OCFS2_IMMUTABLE_FL;
+ if (flags & S_NOATIME)
+ oi->ip_attr |= OCFS2_NOATIME_FL;
+ if (flags & S_DIRSYNC)
+ oi->ip_attr |= OCFS2_DIRSYNC_FL;
++
++ if (vflags & V_BARRIER)
++ oi->ip_attr |= OCFS2_BARRIER_FL;
++ if (vflags & V_COW)
++ oi->ip_attr |= OCFS2_COW_FL;
++}
++
++int ocfs2_sync_flags(struct inode *inode)
++{
++ struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode);
++ struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
++ handle_t *handle = NULL;
++ struct buffer_head *bh = NULL;
++ int status;
++
++ mutex_lock(&inode->i_mutex);
++
++ status = ocfs2_inode_lock(inode, &bh, 1);
++ if (status < 0) {
++ mlog_errno(status);
++ goto bail;
++ }
++
++ status = -EROFS;
++ if (IS_RDONLY(inode))
++ goto bail_unlock;
++
++ handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
++ if (IS_ERR(handle)) {
++ status = PTR_ERR(handle);
++ mlog_errno(status);
++ goto bail_unlock;
++ }
++
++ ocfs2_set_inode_flags(inode);
++ status = ocfs2_mark_inode_dirty(handle, inode, bh);
++ if (status < 0)
++ mlog_errno(status);
++
++ ocfs2_commit_trans(osb, handle);
++bail_unlock:
++ ocfs2_inode_unlock(inode, 1);
++bail:
++ mutex_unlock(&inode->i_mutex);
++
++ if (bh)
++ brelse(bh);
++
++ mlog_exit(status);
++ return status;
+ }
+
+ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno, unsigned flags,
+@@ -219,6 +287,8 @@ int ocfs2_populate_inode(struct inode *i
+ struct super_block *sb;
+ struct ocfs2_super *osb;
+ int status = -EINVAL;
++ uid_t uid;
++ gid_t gid;
+
+ mlog_entry("(0x%p, size:%llu)\n", inode,
+ (unsigned long long)le64_to_cpu(fe->i_size));
+@@ -254,8 +324,12 @@ int ocfs2_populate_inode(struct inode *i
+ inode->i_generation = le32_to_cpu(fe->i_generation);
+ inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
+ inode->i_mode = le16_to_cpu(fe->i_mode);
+- inode->i_uid = le32_to_cpu(fe->i_uid);
+- inode->i_gid = le32_to_cpu(fe->i_gid);
++ uid = le32_to_cpu(fe->i_uid);
++ gid = le32_to_cpu(fe->i_gid);
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid,
++ /* le16_to_cpu(raw_inode->i_raw_tag)i */ 0);
+
+ /* Fast symlinks will have i_size but no allocated clusters. */
+ if (S_ISLNK(inode->i_mode) && !fe->i_clusters)
+@@ -1230,8 +1304,11 @@ int ocfs2_mark_inode_dirty(handle_t *han
+
+ fe->i_size = cpu_to_le64(i_size_read(inode));
+ fe->i_links_count = cpu_to_le16(inode->i_nlink);
+- fe->i_uid = cpu_to_le32(inode->i_uid);
+- fe->i_gid = cpu_to_le32(inode->i_gid);
++ fe->i_uid = cpu_to_le32(TAGINO_UID(DX_TAG(inode),
++ inode->i_uid, inode->i_tag));
++ fe->i_gid = cpu_to_le32(TAGINO_GID(DX_TAG(inode),
++ inode->i_gid, inode->i_tag));
++ /* i_tag = = cpu_to_le16(inode->i_tag); */
+ fe->i_mode = cpu_to_le16(inode->i_mode);
+ fe->i_atime = cpu_to_le64(inode->i_atime.tv_sec);
+ fe->i_atime_nsec = cpu_to_le32(inode->i_atime.tv_nsec);
+@@ -1259,16 +1336,25 @@ leave:
+ void ocfs2_refresh_inode(struct inode *inode,
+ struct ocfs2_dinode *fe)
+ {
++ uid_t uid;
++ gid_t gid;
++
+ spin_lock(&OCFS2_I(inode)->ip_lock);
+
+ OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
+ OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr);
++ /* OCFS2_I(inode)->ip_flags &= ~OCFS2_FL_MASK;
++ OCFS2_I(inode)->ip_flags |= le32_to_cpu(fe->i_flags) & OCFS2_FL_MASK; */
+ OCFS2_I(inode)->ip_dyn_features = le16_to_cpu(fe->i_dyn_features);
+ ocfs2_set_inode_flags(inode);
+ i_size_write(inode, le64_to_cpu(fe->i_size));
+ inode->i_nlink = le16_to_cpu(fe->i_links_count);
+- inode->i_uid = le32_to_cpu(fe->i_uid);
+- inode->i_gid = le32_to_cpu(fe->i_gid);
++ uid = le32_to_cpu(fe->i_uid);
++ gid = le32_to_cpu(fe->i_gid);
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid,
++ /* le16_to_cpu(raw_inode->i_raw_tag)i */ 0);
+ inode->i_mode = le16_to_cpu(fe->i_mode);
+ if (S_ISLNK(inode->i_mode) && le32_to_cpu(fe->i_clusters) == 0)
+ inode->i_blocks = 0;
+--- a/fs/ocfs2/inode.h 2008-04-17 12:05:40.000000000 -0400
++++ a/fs/ocfs2/inode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -143,6 +143,7 @@ int ocfs2_aio_write(struct file *file, s
+
+ void ocfs2_set_inode_flags(struct inode *inode);
+ void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi);
++int ocfs2_sync_flags(struct inode *inode);
+
+ static inline blkcnt_t ocfs2_inode_sector_count(struct inode *inode)
+ {
+--- a/fs/ocfs2/ioctl.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/ioctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -42,7 +42,7 @@ static int ocfs2_get_inode_attr(struct i
+ return status;
+ }
+
+-static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
++int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
+ unsigned mask)
+ {
+ struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode);
+@@ -67,6 +67,11 @@ static int ocfs2_set_inode_attr(struct i
+ if (!S_ISDIR(inode->i_mode))
+ flags &= ~OCFS2_DIRSYNC_FL;
+
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ goto bail_unlock;
++ }
++
+ handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
+ if (IS_ERR(handle)) {
+ status = PTR_ERR(handle);
+--- a/fs/ocfs2/ioctl.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/ioctl.h 2008-07-17 18:10:34.000000000 -0400
+@@ -10,6 +10,9 @@
+ #ifndef OCFS2_IOCTL_H
+ #define OCFS2_IOCTL_H
+
++int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
++ unsigned mask);
++
+ long ocfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
+ long ocfs2_compat_ioctl(struct file *file, unsigned cmd, unsigned long arg);
+
+--- a/fs/ocfs2/namei.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/namei.c 2008-07-17 18:13:38.000000000 -0400
+@@ -40,6 +40,7 @@
+ #include <linux/types.h>
+ #include <linux/slab.h>
+ #include <linux/highmem.h>
++#include <linux/vs_tag.h>
+
+ #define MLOG_MASK_PREFIX ML_NAMEI
+ #include <cluster/masklog.h>
+@@ -366,6 +367,9 @@ static int ocfs2_mknod_locked(struct ocf
+ u64 fe_blkno = 0;
+ u16 suballoc_bit;
+ struct inode *inode = NULL;
++ uid_t uid;
++ gid_t gid;
++ tag_t tag;
+
+ mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
+ (unsigned long)dev, dentry->d_name.len,
+@@ -425,13 +429,19 @@ static int ocfs2_mknod_locked(struct ocf
+ fe->i_blkno = cpu_to_le64(fe_blkno);
+ fe->i_suballoc_bit = cpu_to_le16(suballoc_bit);
+ fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot);
+- fe->i_uid = cpu_to_le32(current->fsuid);
++
++ tag = dx_current_fstag(osb->sb);
++ uid = current->fsuid;
+ if (dir->i_mode & S_ISGID) {
+- fe->i_gid = cpu_to_le32(dir->i_gid);
++ gid = dir->i_gid;
+ if (S_ISDIR(mode))
+ mode |= S_ISGID;
+ } else
+- fe->i_gid = cpu_to_le32(current->fsgid);
++ gid = current->fsgid;
++
++ fe->i_uid = cpu_to_le32(TAGINO_UID(DX_TAG(inode), uid, tag));
++ fe->i_gid = cpu_to_le32(TAGINO_GID(DX_TAG(inode), gid, tag));
++ inode->i_tag = tag;
+ fe->i_mode = cpu_to_le16(mode);
+ if (S_ISCHR(mode) || S_ISBLK(mode))
+ fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev));
+@@ -1917,5 +1927,6 @@ const struct inode_operations ocfs2_dir_
+ .rename = ocfs2_rename,
+ .setattr = ocfs2_setattr,
+ .getattr = ocfs2_getattr,
++ .sync_flags = ocfs2_sync_flags,
+ .permission = ocfs2_permission,
+ };
+--- a/fs/ocfs2/ocfs2_fs.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/ocfs2_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -204,9 +204,14 @@
+ #define OCFS2_NODUMP_FL (0x00000040) /* do not dump file */
+ #define OCFS2_NOATIME_FL (0x00000080) /* do not update atime */
+ #define OCFS2_DIRSYNC_FL (0x00010000) /* dirsync behaviour (directories only) */
++#define OCFS2_IXUNLINK_FL (0x00020000) /* Immutable invert on unlink */
+
+-#define OCFS2_FL_VISIBLE (0x000100FF) /* User visible flags */
+-#define OCFS2_FL_MODIFIABLE (0x000100FF) /* User modifiable flags */
++#define OCFS2_BARRIER_FL (0x01000000) /* Barrier for chroot() */
++#define OCFS2_COW_FL (0x02000000) /* Copy on Write marker */
++
++#define OCFS2_FL_VISIBLE (0x000300FF) /* User visible flags */
++#define OCFS2_FL_MODIFIABLE (0x000300FF) /* User modifiable flags */
++#define OCFS2_FL_MASK (0x030300FF)
+
+ /*
+ * Extent record flags (e_node.leaf.flags)
+--- a/fs/ocfs2/ocfs2.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/ocfs2.h 2008-07-16 22:41:36.000000000 -0400
+@@ -172,6 +172,7 @@ enum ocfs2_mount_options
+ OCFS2_MOUNT_ERRORS_PANIC = 1 << 3, /* Panic on errors */
+ OCFS2_MOUNT_DATA_WRITEBACK = 1 << 4, /* No data ordering */
+ OCFS2_MOUNT_LOCALFLOCKS = 1 << 5, /* No cluster aware user file locks */
++ OCFS2_MOUNT_TAGGED = 1 << 8, /* use tagging */
+ };
+
+ #define OCFS2_OSB_SOFT_RO 0x0001
+--- a/fs/ocfs2/super.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/ocfs2/super.c 2008-07-17 20:23:09.000000000 -0400
+@@ -154,6 +154,7 @@ enum {
+ Opt_localalloc,
+ Opt_localflocks,
+ Opt_stack,
++ Opt_tag, Opt_notag, Opt_tagid,
+ Opt_err,
+ };
+
+@@ -173,6 +174,9 @@ static match_table_t tokens = {
+ {Opt_localalloc, "localalloc=%d"},
+ {Opt_localflocks, "localflocks"},
+ {Opt_stack, "cluster_stack=%s"},
++ {Opt_tag, "tag"},
++ {Opt_notag, "notag"},
++ {Opt_tagid, "tagid=%u"},
+ {Opt_err, NULL}
+ };
+
+@@ -392,6 +396,13 @@ static int ocfs2_remount(struct super_bl
+ goto out;
+ }
+
++ if ((parsed_options.mount_opt & OCFS2_MOUNT_TAGGED) &&
++ !(sb->s_flags & MS_TAGGED)) {
++ ret = -EINVAL;
++ mlog(ML_ERROR, "Cannot change tagging on remount\n");
++ goto out;
++ }
++
+ if ((osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) !=
+ (parsed_options.mount_opt & OCFS2_MOUNT_HB_LOCAL)) {
+ ret = -EINVAL;
+@@ -725,6 +736,9 @@ static int ocfs2_fill_super(struct super
+
+ ocfs2_complete_mount_recovery(osb);
+
++ if (osb->s_mount_opt & OCFS2_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED;
++
+ if (ocfs2_mount_local(osb))
+ snprintf(nodestr, sizeof(nodestr), "local");
+ else
+@@ -918,6 +932,20 @@ static int ocfs2_parse_options(struct su
+ OCFS2_STACK_LABEL_LEN);
+ mopt->cluster_stack[OCFS2_STACK_LABEL_LEN] = '\0';
+ break;
++#ifndef CONFIG_TAGGING_NONE
++ case Opt_tag:
++ mopt->mount_opt |= OCFS2_MOUNT_TAGGED;
++ break;
++ case Opt_notag:
++ mopt->mount_opt &= ~OCFS2_MOUNT_TAGGED;
++ break;
++#endif
++#ifdef CONFIG_PROPAGATE
++ case Opt_tagid:
++ /* use args[0] */
++ mopt->mount_opt |= OCFS2_MOUNT_TAGGED;
++ break;
++#endif
+ default:
+ mlog(ML_ERROR,
+ "Unrecognized mount option \"%s\" "
+--- a/fs/open.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/open.c 2008-07-17 18:02:08.000000000 -0400
+@@ -29,22 +29,31 @@
+ #include <linux/rcupdate.h>
+ #include <linux/audit.h>
+ #include <linux/falloc.h>
++#include <linux/vs_base.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
++#include <linux/vs_cowbl.h>
+
+ int vfs_statfs(struct dentry *dentry, struct kstatfs *buf)
+ {
+ int retval = -ENODEV;
+
+ if (dentry) {
++ struct super_block *sb = dentry->d_sb;
++
+ retval = -ENOSYS;
+- if (dentry->d_sb->s_op->statfs) {
++ if (sb->s_op->statfs) {
+ memset(buf, 0, sizeof(*buf));
+ retval = security_sb_statfs(dentry);
+ if (retval)
+ return retval;
+- retval = dentry->d_sb->s_op->statfs(dentry, buf);
++ retval = sb->s_op->statfs(dentry, buf);
+ if (retval == 0 && buf->f_frsize == 0)
+ buf->f_frsize = buf->f_bsize;
+ }
++ if (!vx_check(0, VS_ADMIN|VS_WATCH))
++ vx_vsi_statfs(sb, buf);
+ }
+ return retval;
+ }
+@@ -619,6 +628,11 @@ asmlinkage long sys_fchmodat(int dfd, co
+ error = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW, &nd);
+ if (error)
+ goto out;
++
++ error = cow_check_and_break(&nd);
++ if (error)
++ goto dput_and_out;
++
+ inode = nd.path.dentry->d_inode;
+
+ error = mnt_want_write(nd.path.mnt);
+@@ -667,11 +681,11 @@ static int chown_common(struct dentry *
+ newattrs.ia_valid = ATTR_CTIME;
+ if (user != (uid_t) -1) {
+ newattrs.ia_valid |= ATTR_UID;
+- newattrs.ia_uid = user;
++ newattrs.ia_uid = dx_map_uid(user);
+ }
+ if (group != (gid_t) -1) {
+ newattrs.ia_valid |= ATTR_GID;
+- newattrs.ia_gid = group;
++ newattrs.ia_gid = dx_map_gid(group);
+ }
+ if (!S_ISDIR(inode->i_mode))
+ newattrs.ia_valid |=
+@@ -694,7 +708,11 @@ asmlinkage long sys_chown(const char __u
+ error = mnt_want_write(nd.path.mnt);
+ if (error)
+ goto out_release;
+- error = chown_common(nd.path.dentry, user, group);
++#ifdef CONFIG_VSERVER_COWBL
++ error = cow_check_and_break(&nd);
++ if (!error)
++#endif
++ error = chown_common(nd.path.dentry, user, group);
+ mnt_drop_write(nd.path.mnt);
+ out_release:
+ path_put(&nd.path);
+@@ -719,7 +737,11 @@ asmlinkage long sys_fchownat(int dfd, co
+ error = mnt_want_write(nd.path.mnt);
+ if (error)
+ goto out_release;
+- error = chown_common(nd.path.dentry, user, group);
++#ifdef CONFIG_VSERVER_COWBL
++ error = cow_check_and_break(&nd);
++ if (!error)
++#endif
++ error = chown_common(nd.path.dentry, user, group);
+ mnt_drop_write(nd.path.mnt);
+ out_release:
+ path_put(&nd.path);
+@@ -738,7 +760,11 @@ asmlinkage long sys_lchown(const char __
+ error = mnt_want_write(nd.path.mnt);
+ if (error)
+ goto out_release;
+- error = chown_common(nd.path.dentry, user, group);
++#ifdef CONFIG_VSERVER_COWBL
++ error = cow_check_and_break(&nd);
++ if (!error)
++#endif
++ error = chown_common(nd.path.dentry, user, group);
+ mnt_drop_write(nd.path.mnt);
+ out_release:
+ path_put(&nd.path);
+@@ -1016,6 +1042,7 @@ repeat:
+ else
+ FD_CLR(fd, fdt->close_on_exec);
+ files->next_fd = fd + 1;
++ vx_openfd_inc(fd);
+ #if 1
+ /* Sanity check */
+ if (fdt->fd[fd] != NULL) {
+@@ -1043,6 +1070,7 @@ static void __put_unused_fd(struct files
+ __FD_CLR(fd, fdt->open_fds);
+ if (fd < files->next_fd)
+ files->next_fd = fd;
++ vx_openfd_dec(fd);
+ }
+
+ void put_unused_fd(unsigned int fd)
+--- a/fs/proc/array.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/array.c 2008-07-17 17:40:35.000000000 -0400
+@@ -80,6 +80,8 @@
+ #include <linux/delayacct.h>
+ #include <linux/seq_file.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/processor.h>
+@@ -141,8 +143,9 @@ static const char *task_state_array[] =
+ "D (disk sleep)", /* 2 */
+ "T (stopped)", /* 4 */
+ "T (tracing stop)", /* 8 */
+- "Z (zombie)", /* 16 */
+- "X (dead)" /* 32 */
++ "H (on hold)", /* 16 */
++ "Z (zombie)", /* 32 */
++ "X (dead)", /* 64 */
+ };
+
+ static inline const char *get_task_state(struct task_struct *tsk)
+@@ -163,6 +166,7 @@ static inline void task_state(struct seq
+ struct group_info *group_info;
+ int g;
+ struct fdtable *fdt = NULL;
++/* + pid_t pid, ptgid, tppid, tgid; */
+ pid_t ppid, tpid;
+
+ rcu_read_lock();
+@@ -170,6 +174,12 @@ static inline void task_state(struct seq
+ task_tgid_nr_ns(rcu_dereference(p->real_parent), ns) : 0;
+ tpid = pid_alive(p) && p->ptrace ?
+ task_pid_nr_ns(rcu_dereference(p->parent), ns) : 0;
++/* tgid = vx_map_tgid(p->tgid);
++ pid = vx_map_pid(p->pid);
++ ptgid = vx_map_pid(pid_alive(p) ?
++ rcu_dereference(p->real_parent)->tgid : 0);
++ tppid = vx_map_pid(pid_alive(p) && p->ptrace ?
++ rcu_dereference(p->parent)->pid : 0); */
+ seq_printf(m,
+ "State:\t%s\n"
+ "Tgid:\t%d\n"
+@@ -281,7 +291,7 @@ static inline void task_sig(struct seq_f
+ }
+
+ static void render_cap_t(struct seq_file *m, const char *header,
+- kernel_cap_t *a)
++ struct vx_info *vxi, kernel_cap_t *a)
+ {
+ unsigned __capi;
+
+@@ -295,10 +305,10 @@ static void render_cap_t(struct seq_file
+
+ static inline void task_cap(struct seq_file *m, struct task_struct *p)
+ {
+- render_cap_t(m, "CapInh:\t", &p->cap_inheritable);
+- render_cap_t(m, "CapPrm:\t", &p->cap_permitted);
+- render_cap_t(m, "CapEff:\t", &p->cap_effective);
+- render_cap_t(m, "CapBnd:\t", &p->cap_bset);
++ render_cap_t(m, "CapInh:\t", p->vx_info, &p->cap_inheritable);
++ render_cap_t(m, "CapPrm:\t", p->vx_info, &p->cap_permitted);
++ render_cap_t(m, "CapEff:\t", p->vx_info, &p->cap_effective);
++ render_cap_t(m, "CapBnd:\t", p->vx_info, &p->cap_bset);
+ }
+
+ static inline void task_context_switch_counts(struct seq_file *m,
+@@ -310,6 +320,45 @@ static inline void task_context_switch_c
+ p->nivcsw);
+ }
+
++int proc_pid_nsproxy(struct seq_file *m, struct pid_namespace *ns,
++ struct pid *pid, struct task_struct *task)
++{
++ seq_printf(m, "Proxy:\t%p(%c)\n"
++ "Count:\t%u\n"
++ "uts:\t%p(%c)\n"
++ "ipc:\t%p(%c)\n"
++ "mnt:\t%p(%c)\n"
++ "pid:\t%p(%c)\n"
++ "user:\t%p(%c)\n"
++ "net:\t%p(%c)\n",
++ task->nsproxy,
++ (task->nsproxy == init_task.nsproxy ? 'I' : '-'),
++ atomic_read(&task->nsproxy->count),
++ task->nsproxy->uts_ns,
++ (task->nsproxy->uts_ns == init_task.nsproxy->uts_ns ? 'I' : '-'),
++ task->nsproxy->ipc_ns,
++ (task->nsproxy->ipc_ns == init_task.nsproxy->ipc_ns ? 'I' : '-'),
++ task->nsproxy->mnt_ns,
++ (task->nsproxy->mnt_ns == init_task.nsproxy->mnt_ns ? 'I' : '-'),
++ task->nsproxy->pid_ns,
++ (task->nsproxy->pid_ns == init_task.nsproxy->pid_ns ? 'I' : '-'),
++ task->nsproxy->user_ns,
++ (task->nsproxy->user_ns == init_task.nsproxy->user_ns ? 'I' : '-'),
++ task->nsproxy->net_ns,
++ (task->nsproxy->net_ns == init_task.nsproxy->net_ns ? 'I' : '-'));
++ return 0;
++}
++
++void task_vs_id(struct seq_file *m, struct task_struct *task)
++{
++ if (task_vx_flags(task, VXF_HIDE_VINFO, 0))
++ return;
++
++ seq_printf(m, "VxID: %d\n", vx_task_xid(task));
++ seq_printf(m, "NxID: %d\n", nx_task_nid(task));
++}
++
++
+ int proc_pid_status(struct seq_file *m, struct pid_namespace *ns,
+ struct pid *pid, struct task_struct *task)
+ {
+@@ -325,6 +374,7 @@ int proc_pid_status(struct seq_file *m,
+ task_sig(m, task);
+ task_cap(m, task);
+ cpuset_task_status_allowed(m, task);
++ task_vs_id(m, task);
+ #if defined(CONFIG_S390)
+ task_show_regs(m, task);
+ #endif
+@@ -496,6 +546,17 @@ static int do_task_stat(struct seq_file
+ /* convert nsec -> ticks */
+ start_time = nsec_to_clock_t(start_time);
+
++ /* fixup start time for virt uptime */
++ if (vx_flags(VXF_VIRT_UPTIME, 0)) {
++ unsigned long long bias =
++ current->vx_info->cvirt.bias_clock;
++
++ if (start_time > bias)
++ start_time -= bias;
++ else
++ start_time = 0;
++ }
++
+ seq_printf(m, "%d (%s) %c %d %d %d %d %d %u %lu \
+ %lu %lu %lu %lu %lu %ld %ld %ld %ld %d 0 %llu %lu %ld %lu %lu %lu %lu %lu \
+ %lu %lu %lu %lu %lu %lu %lu %lu %d %d %u %u %llu %lu %ld\n",
+--- a/fs/proc/base.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/base.c 2008-07-16 22:41:36.000000000 -0400
+@@ -77,6 +77,8 @@
+ #include <linux/oom.h>
+ #include <linux/elf.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
+ #include "internal.h"
+
+ /* NOTE:
+@@ -1412,6 +1414,8 @@ static struct inode *proc_pid_make_inode
+ inode->i_uid = task->euid;
+ inode->i_gid = task->egid;
+ }
++ /* procfs is xid tagged */
++ inode->i_tag = (tag_t)vx_task_xid(task);
+ security_task_to_inode(task, inode);
+
+ out:
+@@ -1953,6 +1957,13 @@ static struct dentry *proc_pident_lookup
+ if (!task)
+ goto out_no_task;
+
++ /* TODO: maybe we can come up with a generic approach? */
++ if (task_vx_flags(task, VXF_HIDE_VINFO, 0) &&
++ (dentry->d_name.len == 5) &&
++ (!memcmp(dentry->d_name.name, "vinfo", 5) ||
++ !memcmp(dentry->d_name.name, "ninfo", 5)))
++ goto out;
++
+ /*
+ * Yes, it does not scale. And it should not. Don't add
+ * new entries into /proc/<tgid>/ without very good reasons.
+@@ -2340,7 +2351,7 @@ out_iput:
+ static struct dentry *proc_base_lookup(struct inode *dir, struct dentry *dentry)
+ {
+ struct dentry *error;
+- struct task_struct *task = get_proc_task(dir);
++ struct task_struct *task = get_proc_task_real(dir);
+ const struct pid_entry *p, *last;
+
+ error = ERR_PTR(-ENOENT);
+@@ -2405,6 +2416,9 @@ static int proc_pid_io_accounting(struct
+ static const struct file_operations proc_task_operations;
+ static const struct inode_operations proc_task_inode_operations;
+
++extern int proc_pid_vx_info(struct task_struct *, char *);
++extern int proc_pid_nx_info(struct task_struct *, char *);
++
+ static const struct pid_entry tgid_base_stuff[] = {
+ DIR("task", S_IRUGO|S_IXUGO, task),
+ DIR("fd", S_IRUSR|S_IXUSR, fd),
+@@ -2456,6 +2470,8 @@ static const struct pid_entry tgid_base_
+ #ifdef CONFIG_CGROUPS
+ REG("cgroup", S_IRUGO, cgroup),
+ #endif
++ INF("vinfo", S_IRUGO, pid_vx_info),
++ INF("ninfo", S_IRUGO, pid_nx_info),
+ INF("oom_score", S_IRUGO, oom_score),
+ REG("oom_adj", S_IRUGO|S_IWUSR, oom_adjust),
+ #ifdef CONFIG_AUDITSYSCALL
+@@ -2471,6 +2487,7 @@ static const struct pid_entry tgid_base_
+ #ifdef CONFIG_TASK_IO_ACCOUNTING
+ INF("io", S_IRUGO, pid_io_accounting),
+ #endif
++ ONE("nsproxy", S_IRUGO, pid_nsproxy),
+ };
+
+ static int proc_tgid_base_readdir(struct file * filp,
+@@ -2667,7 +2684,7 @@ retry:
+ iter.task = NULL;
+ pid = find_ge_pid(iter.tgid, ns);
+ if (pid) {
+- iter.tgid = pid_nr_ns(pid, ns);
++ iter.tgid = pid_unmapped_nr_ns(pid, ns);
+ iter.task = pid_task(pid, PIDTYPE_PID);
+ /* What we to know is if the pid we have find is the
+ * pid of a thread_group_leader. Testing for task
+@@ -2697,7 +2714,7 @@ static int proc_pid_fill_cache(struct fi
+ struct tgid_iter iter)
+ {
+ char name[PROC_NUMBUF];
+- int len = snprintf(name, sizeof(name), "%d", iter.tgid);
++ int len = snprintf(name, sizeof(name), "%d", vx_map_tgid(iter.tgid));
+ return proc_fill_cache(filp, dirent, filldir, name, len,
+ proc_pid_instantiate, iter.task, NULL);
+ }
+@@ -2706,7 +2723,7 @@ static int proc_pid_fill_cache(struct fi
+ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir)
+ {
+ unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY;
+- struct task_struct *reaper = get_proc_task(filp->f_path.dentry->d_inode);
++ struct task_struct *reaper = get_proc_task_real(filp->f_path.dentry->d_inode);
+ struct tgid_iter iter;
+ struct pid_namespace *ns;
+
+@@ -2726,6 +2743,8 @@ int proc_pid_readdir(struct file * filp,
+ iter.task;
+ iter.tgid += 1, iter = next_tgid(ns, iter)) {
+ filp->f_pos = iter.tgid + TGID_OFFSET;
++ if (!vx_proc_task_visible(iter.task))
++ continue;
+ if (proc_pid_fill_cache(filp, dirent, filldir, iter) < 0) {
+ put_task_struct(iter.task);
+ goto out;
+@@ -2862,6 +2881,8 @@ static struct dentry *proc_task_lookup(s
+ tid = name_to_int(dentry);
+ if (tid == ~0U)
+ goto out;
++ if (vx_current_initpid(tid))
++ goto out;
+
+ ns = dentry->d_sb->s_fs_info;
+ rcu_read_lock();
+--- a/fs/proc/generic.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/generic.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/bitops.h>
+ #include <linux/spinlock.h>
+ #include <linux/completion.h>
++#include <linux/vserver/inode.h>
+ #include <asm/uaccess.h>
+
+ #include "internal.h"
+@@ -386,6 +387,8 @@ struct dentry *proc_lookup_de(struct pro
+ for (de = de->subdir; de ; de = de->next) {
+ if (de->namelen != dentry->d_name.len)
+ continue;
++ if (!vx_hide_check(0, de->vx_flags))
++ continue;
+ if (!memcmp(dentry->d_name.name, de->name, de->namelen)) {
+ unsigned int ino;
+
+@@ -394,6 +397,8 @@ struct dentry *proc_lookup_de(struct pro
+ spin_unlock(&proc_subdir_lock);
+ error = -EINVAL;
+ inode = proc_get_inode(dir->i_sb, ino, de);
++ /* generic proc entries belong to the host */
++ inode->i_tag = 0;
+ goto out_unlock;
+ }
+ }
+@@ -474,6 +479,8 @@ int proc_readdir_de(struct proc_dir_entr
+
+ /* filldir passes info to user space */
+ de_get(de);
++ if (!vx_hide_check(0, de->vx_flags))
++ goto skip;
+ spin_unlock(&proc_subdir_lock);
+ if (filldir(dirent, de->name, de->namelen, filp->f_pos,
+ de->low_ino, de->mode >> 12) < 0) {
+@@ -481,6 +488,7 @@ int proc_readdir_de(struct proc_dir_entr
+ goto out;
+ }
+ spin_lock(&proc_subdir_lock);
++ skip:
+ filp->f_pos++;
+ next = de->next;
+ de_put(de);
+@@ -595,6 +603,7 @@ static struct proc_dir_entry *__proc_cre
+ ent->nlink = nlink;
+ atomic_set(&ent->count, 1);
+ ent->pde_users = 0;
++ ent->vx_flags = IATTR_PROC_DEFAULT;
+ spin_lock_init(&ent->pde_unload_lock);
+ ent->pde_unload_completion = NULL;
+ out:
+@@ -617,7 +626,8 @@ struct proc_dir_entry *proc_symlink(cons
+ kfree(ent->data);
+ kfree(ent);
+ ent = NULL;
+- }
++ } else
++ ent->vx_flags = IATTR_PROC_SYMLINK;
+ } else {
+ kfree(ent);
+ ent = NULL;
+--- a/fs/proc/inode.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -405,6 +405,8 @@ struct inode *proc_get_inode(struct supe
+ inode->i_uid = de->uid;
+ inode->i_gid = de->gid;
+ }
++ if (de->vx_flags)
++ PROC_I(inode)->vx_flags = de->vx_flags;
+ if (de->size)
+ inode->i_size = de->size;
+ if (de->nlink)
+--- a/fs/proc/internal.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/internal.h 2008-07-16 22:41:36.000000000 -0400
+@@ -10,6 +10,7 @@
+ */
+
+ #include <linux/proc_fs.h>
++#include <linux/vs_pid.h>
+
+ extern struct proc_dir_entry proc_root;
+ #ifdef CONFIG_PROC_SYSCTL
+@@ -55,6 +56,9 @@ extern int proc_pid_status(struct seq_fi
+ struct pid *pid, struct task_struct *task);
+ extern int proc_pid_statm(struct seq_file *m, struct pid_namespace *ns,
+ struct pid *pid, struct task_struct *task);
++extern int proc_pid_nsproxy(struct seq_file *m, struct pid_namespace *ns,
++ struct pid *pid, struct task_struct *task);
++
+ extern loff_t mem_lseek(struct file *file, loff_t offset, int orig);
+
+ extern const struct file_operations proc_maps_operations;
+@@ -74,11 +78,16 @@ static inline struct pid *proc_pid(struc
+ return PROC_I(inode)->pid;
+ }
+
+-static inline struct task_struct *get_proc_task(struct inode *inode)
++static inline struct task_struct *get_proc_task_real(struct inode *inode)
+ {
+ return get_pid_task(proc_pid(inode), PIDTYPE_PID);
+ }
+
++static inline struct task_struct *get_proc_task(struct inode *inode)
++{
++ return vx_get_proc_task(inode, proc_pid(inode));
++}
++
+ static inline int proc_fd(struct inode *inode)
+ {
+ return PROC_I(inode)->fd;
+--- a/fs/proc/proc_misc.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/proc_misc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -56,6 +56,8 @@
+ #include <asm/div64.h>
+ #include "internal.h"
+
++#include <linux/vs_cvirt.h>
++
+ #define LOAD_INT(x) ((x) >> FSHIFT)
+ #define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1-1)) * 100)
+ /*
+@@ -83,22 +85,38 @@ static int proc_calc_metrics(char *page,
+ static int loadavg_read_proc(char *page, char **start, off_t off,
+ int count, int *eof, void *data)
+ {
++ unsigned long running;
++ unsigned int threads;
+ int a, b, c;
+ int len;
+ unsigned long seq;
+
+ do {
+ seq = read_seqbegin(&xtime_lock);
+- a = avenrun[0] + (FIXED_1/200);
+- b = avenrun[1] + (FIXED_1/200);
+- c = avenrun[2] + (FIXED_1/200);
++ if (vx_flags(VXF_VIRT_LOAD, 0)) {
++ struct vx_info *vxi = current->vx_info;
++
++ a = vxi->cvirt.load[0] + (FIXED_1/200);
++ b = vxi->cvirt.load[1] + (FIXED_1/200);
++ c = vxi->cvirt.load[2] + (FIXED_1/200);
++
++ running = atomic_read(&vxi->cvirt.nr_running);
++ threads = atomic_read(&vxi->cvirt.nr_threads);
++ } else {
++ a = avenrun[0] + (FIXED_1/200);
++ b = avenrun[1] + (FIXED_1/200);
++ c = avenrun[2] + (FIXED_1/200);
++
++ running = nr_running();
++ threads = nr_threads;
++ }
+ } while (read_seqretry(&xtime_lock, seq));
+
+ len = sprintf(page,"%d.%02d %d.%02d %d.%02d %ld/%d %d\n",
+ LOAD_INT(a), LOAD_FRAC(a),
+ LOAD_INT(b), LOAD_FRAC(b),
+ LOAD_INT(c), LOAD_FRAC(c),
+- nr_running(), nr_threads,
++ running, threads,
+ task_active_pid_ns(current)->last_pid);
+ return proc_calc_metrics(page, start, off, count, eof, len);
+ }
+@@ -114,6 +132,9 @@ static int uptime_read_proc(char *page,
+ do_posix_clock_monotonic_gettime(&uptime);
+ monotonic_to_bootbased(&uptime);
+ cputime_to_timespec(idletime, &idle);
++ if (vx_flags(VXF_VIRT_UPTIME, 0))
++ vx_vsi_uptime(&uptime, &idle);
++
+ len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
+ (unsigned long) uptime.tv_sec,
+ (uptime.tv_nsec / (NSEC_PER_SEC / 100)),
+@@ -145,7 +166,7 @@ static int meminfo_read_proc(char *page,
+
+ cached = global_page_state(NR_FILE_PAGES) -
+ total_swapcache_pages - i.bufferram;
+- if (cached < 0)
++ if (cached < 0 || vx_flags(VXF_VIRT_MEM, 0))
+ cached = 0;
+
+ get_vmalloc_info(&vmi);
+--- a/fs/proc/root.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/proc/root.c 2008-07-17 17:31:52.000000000 -0400
+@@ -22,6 +22,10 @@
+
+ #include "internal.h"
+
++struct proc_dir_entry *proc_virtual;
++
++extern void proc_vx_init(void);
++
+ static int proc_test_super(struct super_block *sb, void *data)
+ {
+ return sb->s_fs_info == data;
+@@ -137,6 +141,7 @@ void __init proc_root_init(void)
+ #endif
+ proc_mkdir("bus", NULL);
+ proc_sys_init();
++ proc_vx_init();
+ }
+
+ static int proc_root_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat
+--- a/fs/quota.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/quota.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/capability.h>
+ #include <linux/quotaops.h>
+ #include <linux/types.h>
++#include <linux/vs_context.h>
+
+ /* Check validity of generic quotactl commands */
+ static int generic_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t id)
+@@ -81,11 +82,11 @@ static int generic_quotactl_valid(struct
+ if (cmd == Q_GETQUOTA) {
+ if (((type == USRQUOTA && current->euid != id) ||
+ (type == GRPQUOTA && !in_egroup_p(id))) &&
+- !capable(CAP_SYS_ADMIN))
++ !vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return -EPERM;
+ }
+ else if (cmd != Q_GETFMT && cmd != Q_SYNC && cmd != Q_GETINFO)
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return -EPERM;
+
+ return 0;
+@@ -132,10 +133,10 @@ static int xqm_quotactl_valid(struct sup
+ if (cmd == Q_XGETQUOTA) {
+ if (((type == XQM_USRQUOTA && current->euid != id) ||
+ (type == XQM_GRPQUOTA && !in_egroup_p(id))) &&
+- !capable(CAP_SYS_ADMIN))
++ !vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return -EPERM;
+ } else if (cmd != Q_XGETQSTAT && cmd != Q_XQUOTASYNC) {
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return -EPERM;
+ }
+
+@@ -328,6 +329,46 @@ static int do_quotactl(struct super_bloc
+ return 0;
+ }
+
++#if defined(CONFIG_BLK_DEV_VROOT) || defined(CONFIG_BLK_DEV_VROOT_MODULE)
++
++#include <linux/vroot.h>
++#include <linux/major.h>
++#include <linux/module.h>
++#include <linux/kallsyms.h>
++#include <linux/vserver/debug.h>
++
++static vroot_grb_func *vroot_get_real_bdev = NULL;
++
++static spinlock_t vroot_grb_lock = SPIN_LOCK_UNLOCKED;
++
++int register_vroot_grb(vroot_grb_func *func) {
++ int ret = -EBUSY;
++
++ spin_lock(&vroot_grb_lock);
++ if (!vroot_get_real_bdev) {
++ vroot_get_real_bdev = func;
++ ret = 0;
++ }
++ spin_unlock(&vroot_grb_lock);
++ return ret;
++}
++EXPORT_SYMBOL(register_vroot_grb);
++
++int unregister_vroot_grb(vroot_grb_func *func) {
++ int ret = -EINVAL;
++
++ spin_lock(&vroot_grb_lock);
++ if (vroot_get_real_bdev) {
++ vroot_get_real_bdev = NULL;
++ ret = 0;
++ }
++ spin_unlock(&vroot_grb_lock);
++ return ret;
++}
++EXPORT_SYMBOL(unregister_vroot_grb);
++
++#endif
++
+ /*
+ * look up a superblock on which quota ops will be performed
+ * - use the name of a block device to find the superblock thereon
+@@ -345,6 +386,22 @@ static inline struct super_block *quotac
+ putname(tmp);
+ if (IS_ERR(bdev))
+ return ERR_CAST(bdev);
++#if defined(CONFIG_BLK_DEV_VROOT) || defined(CONFIG_BLK_DEV_VROOT_MODULE)
++ if (bdev && bdev->bd_inode &&
++ imajor(bdev->bd_inode) == VROOT_MAJOR) {
++ struct block_device *bdnew = (void *)-EINVAL;
++
++ if (vroot_get_real_bdev)
++ bdnew = vroot_get_real_bdev(bdev);
++ else
++ vxdprintk(VXD_CBIT(misc, 0),
++ "vroot_get_real_bdev not set");
++ bdput(bdev);
++ if (IS_ERR(bdnew))
++ return ERR_PTR(PTR_ERR(bdnew));
++ bdev = bdnew;
++ }
++#endif
+ sb = get_super(bdev);
+ bdput(bdev);
+ if (!sb)
+--- a/fs/reiserfs/bitmap.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/bitmap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -13,6 +13,7 @@
+ #include <linux/reiserfs_fs_sb.h>
+ #include <linux/reiserfs_fs_i.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+
+ #define PREALLOCATION_SIZE 9
+
+@@ -429,8 +430,10 @@ static void _reiserfs_free_block(struct
+ set_sb_free_blocks(rs, sb_free_blocks(rs) + 1);
+
+ journal_mark_dirty(th, s, sbh);
+- if (for_unformatted)
++ if (for_unformatted) {
++ DLIMIT_FREE_BLOCK(inode, 1);
+ DQUOT_FREE_BLOCK_NODIRTY(inode, 1);
++ }
+ }
+
+ void reiserfs_free_block(struct reiserfs_transaction_handle *th,
+@@ -1045,6 +1048,7 @@ static inline int blocknrs_and_prealloc_
+ b_blocknr_t finish = SB_BLOCK_COUNT(s) - 1;
+ int passno = 0;
+ int nr_allocated = 0;
++ int blocks;
+
+ determine_prealloc_size(hint);
+ if (!hint->formatted_node) {
+@@ -1054,19 +1058,30 @@ static inline int blocknrs_and_prealloc_
+ "reiserquota: allocating %d blocks id=%u",
+ amount_needed, hint->inode->i_uid);
+ #endif
+- quota_ret =
+- DQUOT_ALLOC_BLOCK_NODIRTY(hint->inode, amount_needed);
+- if (quota_ret) /* Quota exceeded? */
++ quota_ret = DQUOT_ALLOC_BLOCK_NODIRTY(hint->inode,
++ amount_needed);
++ if (quota_ret)
+ return QUOTA_EXCEEDED;
++ if (DLIMIT_ALLOC_BLOCK(hint->inode, amount_needed)) {
++ DQUOT_FREE_BLOCK_NODIRTY(hint->inode,
++ amount_needed);
++ return NO_DISK_SPACE;
++ }
++
+ if (hint->preallocate && hint->prealloc_size) {
+ #ifdef REISERQUOTA_DEBUG
+ reiserfs_debug(s, REISERFS_DEBUG_CODE,
+ "reiserquota: allocating (prealloc) %d blocks id=%u",
+ hint->prealloc_size, hint->inode->i_uid);
+ #endif
+- quota_ret =
+- DQUOT_PREALLOC_BLOCK_NODIRTY(hint->inode,
+- hint->prealloc_size);
++ quota_ret = DQUOT_PREALLOC_BLOCK_NODIRTY(hint->inode,
++ hint->prealloc_size);
++ if (!quota_ret &&
++ DLIMIT_ALLOC_BLOCK(hint->inode, hint->prealloc_size)) {
++ DQUOT_FREE_BLOCK_NODIRTY(hint->inode,
++ hint->prealloc_size);
++ quota_ret = 1;
++ }
+ if (quota_ret)
+ hint->preallocate = hint->prealloc_size = 0;
+ }
+@@ -1098,7 +1113,10 @@ static inline int blocknrs_and_prealloc_
+ nr_allocated,
+ hint->inode->i_uid);
+ #endif
+- DQUOT_FREE_BLOCK_NODIRTY(hint->inode, amount_needed + hint->prealloc_size - nr_allocated); /* Free not allocated blocks */
++ /* Free not allocated blocks */
++ blocks = amount_needed + hint->prealloc_size - nr_allocated;
++ DLIMIT_FREE_BLOCK(hint->inode, blocks);
++ DQUOT_FREE_BLOCK_NODIRTY(hint->inode, blocks);
+ }
+ while (nr_allocated--)
+ reiserfs_free_block(hint->th, hint->inode,
+@@ -1129,10 +1147,10 @@ static inline int blocknrs_and_prealloc_
+ REISERFS_I(hint->inode)->i_prealloc_count,
+ hint->inode->i_uid);
+ #endif
+- DQUOT_FREE_BLOCK_NODIRTY(hint->inode, amount_needed +
+- hint->prealloc_size - nr_allocated -
+- REISERFS_I(hint->inode)->
+- i_prealloc_count);
++ blocks = amount_needed + hint->prealloc_size - nr_allocated -
++ REISERFS_I(hint->inode)->i_prealloc_count;
++ DLIMIT_FREE_BLOCK(hint->inode, blocks);
++ DQUOT_FREE_BLOCK_NODIRTY(hint->inode, blocks);
+ }
+
+ return CARRY_ON;
+--- a/fs/reiserfs/file.c 2008-04-17 11:31:38.000000000 -0400
++++ a/fs/reiserfs/file.c 2008-07-16 22:41:36.000000000 -0400
+@@ -306,4 +306,5 @@ const struct inode_operations reiserfs_f
+ .listxattr = reiserfs_listxattr,
+ .removexattr = reiserfs_removexattr,
+ .permission = reiserfs_permission,
++ .sync_flags = reiserfs_sync_flags,
+ };
+--- a/fs/reiserfs/inode.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,8 @@
+ #include <linux/writeback.h>
+ #include <linux/quotaops.h>
+ #include <linux/swap.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vs_tag.h>
+
+ int reiserfs_commit_write(struct file *f, struct page *page,
+ unsigned from, unsigned to);
+@@ -54,6 +56,7 @@ void reiserfs_delete_inode(struct inode
+ * stat data deletion */
+ if (!err)
+ DQUOT_FREE_INODE(inode);
++ DLIMIT_FREE_INODE(inode);
+
+ if (journal_end(&th, inode->i_sb, jbegin_count))
+ goto out;
+@@ -1116,6 +1119,8 @@ static void init_inode(struct inode *ino
+ struct buffer_head *bh;
+ struct item_head *ih;
+ __u32 rdev;
++ uid_t uid;
++ gid_t gid;
+ //int version = ITEM_VERSION_1;
+
+ bh = PATH_PLAST_BUFFER(path);
+@@ -1139,12 +1144,13 @@ static void init_inode(struct inode *ino
+ (struct stat_data_v1 *)B_I_PITEM(bh, ih);
+ unsigned long blocks;
+
++ uid = sd_v1_uid(sd);
++ gid = sd_v1_gid(sd);
++
+ set_inode_item_key_version(inode, KEY_FORMAT_3_5);
+ set_inode_sd_version(inode, STAT_DATA_V1);
+ inode->i_mode = sd_v1_mode(sd);
+ inode->i_nlink = sd_v1_nlink(sd);
+- inode->i_uid = sd_v1_uid(sd);
+- inode->i_gid = sd_v1_gid(sd);
+ inode->i_size = sd_v1_size(sd);
+ inode->i_atime.tv_sec = sd_v1_atime(sd);
+ inode->i_mtime.tv_sec = sd_v1_mtime(sd);
+@@ -1186,11 +1192,12 @@ static void init_inode(struct inode *ino
+ // (directories and symlinks)
+ struct stat_data *sd = (struct stat_data *)B_I_PITEM(bh, ih);
+
++ uid = sd_v2_uid(sd);
++ gid = sd_v2_gid(sd);
++
+ inode->i_mode = sd_v2_mode(sd);
+ inode->i_nlink = sd_v2_nlink(sd);
+- inode->i_uid = sd_v2_uid(sd);
+ inode->i_size = sd_v2_size(sd);
+- inode->i_gid = sd_v2_gid(sd);
+ inode->i_mtime.tv_sec = sd_v2_mtime(sd);
+ inode->i_atime.tv_sec = sd_v2_atime(sd);
+ inode->i_ctime.tv_sec = sd_v2_ctime(sd);
+@@ -1220,6 +1227,10 @@ static void init_inode(struct inode *ino
+ sd_attrs_to_i_attrs(sd_v2_attrs(sd), inode);
+ }
+
++ inode->i_uid = INOTAG_UID(DX_TAG(inode), uid, gid);
++ inode->i_gid = INOTAG_GID(DX_TAG(inode), uid, gid);
++ inode->i_tag = INOTAG_TAG(DX_TAG(inode), uid, gid, 0);
++
+ pathrelse(path);
+ if (S_ISREG(inode->i_mode)) {
+ inode->i_op = &reiserfs_file_inode_operations;
+@@ -1242,13 +1253,15 @@ static void init_inode(struct inode *ino
+ static void inode2sd(void *sd, struct inode *inode, loff_t size)
+ {
+ struct stat_data *sd_v2 = (struct stat_data *)sd;
++ uid_t uid = TAGINO_UID(DX_TAG(inode), inode->i_uid, inode->i_tag);
++ gid_t gid = TAGINO_GID(DX_TAG(inode), inode->i_gid, inode->i_tag);
+ __u16 flags;
+
++ set_sd_v2_uid(sd_v2, uid);
++ set_sd_v2_gid(sd_v2, gid);
+ set_sd_v2_mode(sd_v2, inode->i_mode);
+ set_sd_v2_nlink(sd_v2, inode->i_nlink);
+- set_sd_v2_uid(sd_v2, inode->i_uid);
+ set_sd_v2_size(sd_v2, size);
+- set_sd_v2_gid(sd_v2, inode->i_gid);
+ set_sd_v2_mtime(sd_v2, inode->i_mtime.tv_sec);
+ set_sd_v2_atime(sd_v2, inode->i_atime.tv_sec);
+ set_sd_v2_ctime(sd_v2, inode->i_ctime.tv_sec);
+@@ -1771,6 +1784,10 @@ int reiserfs_new_inode(struct reiserfs_t
+
+ BUG_ON(!th->t_trans_id);
+
++ if (DLIMIT_ALLOC_INODE(inode)) {
++ err = -ENOSPC;
++ goto out_bad_dlimit;
++ }
+ if (DQUOT_ALLOC_INODE(inode)) {
+ err = -EDQUOT;
+ goto out_end_trans;
+@@ -1956,6 +1973,9 @@ int reiserfs_new_inode(struct reiserfs_t
+ DQUOT_FREE_INODE(inode);
+
+ out_end_trans:
++ DLIMIT_FREE_INODE(inode);
++
++ out_bad_dlimit:
+ journal_end(th, th->t_super, th->t_blocks_allocated);
+ /* Drop can be outside and it needs more credits so it's better to have it outside */
+ DQUOT_DROP(inode);
+@@ -2842,14 +2862,19 @@ int reiserfs_commit_write(struct file *f
+ void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode)
+ {
+ if (reiserfs_attrs(inode->i_sb)) {
+- if (sd_attrs & REISERFS_SYNC_FL)
+- inode->i_flags |= S_SYNC;
+- else
+- inode->i_flags &= ~S_SYNC;
+ if (sd_attrs & REISERFS_IMMUTABLE_FL)
+ inode->i_flags |= S_IMMUTABLE;
+ else
+ inode->i_flags &= ~S_IMMUTABLE;
++ if (sd_attrs & REISERFS_IXUNLINK_FL)
++ inode->i_flags |= S_IXUNLINK;
++ else
++ inode->i_flags &= ~S_IXUNLINK;
++
++ if (sd_attrs & REISERFS_SYNC_FL)
++ inode->i_flags |= S_SYNC;
++ else
++ inode->i_flags &= ~S_SYNC;
+ if (sd_attrs & REISERFS_APPEND_FL)
+ inode->i_flags |= S_APPEND;
+ else
+@@ -2862,6 +2887,15 @@ void sd_attrs_to_i_attrs(__u16 sd_attrs,
+ REISERFS_I(inode)->i_flags |= i_nopack_mask;
+ else
+ REISERFS_I(inode)->i_flags &= ~i_nopack_mask;
++
++ if (sd_attrs & REISERFS_BARRIER_FL)
++ inode->i_vflags |= V_BARRIER;
++ else
++ inode->i_vflags &= ~V_BARRIER;
++ if (sd_attrs & REISERFS_COW_FL)
++ inode->i_vflags |= V_COW;
++ else
++ inode->i_vflags &= ~V_COW;
+ }
+ }
+
+@@ -2872,6 +2906,11 @@ void i_attrs_to_sd_attrs(struct inode *i
+ *sd_attrs |= REISERFS_IMMUTABLE_FL;
+ else
+ *sd_attrs &= ~REISERFS_IMMUTABLE_FL;
++ if (inode->i_flags & S_IXUNLINK)
++ *sd_attrs |= REISERFS_IXUNLINK_FL;
++ else
++ *sd_attrs &= ~REISERFS_IXUNLINK_FL;
++
+ if (inode->i_flags & S_SYNC)
+ *sd_attrs |= REISERFS_SYNC_FL;
+ else
+@@ -2884,6 +2923,15 @@ void i_attrs_to_sd_attrs(struct inode *i
+ *sd_attrs |= REISERFS_NOTAIL_FL;
+ else
+ *sd_attrs &= ~REISERFS_NOTAIL_FL;
++
++ if (inode->i_vflags & V_BARRIER)
++ *sd_attrs |= REISERFS_BARRIER_FL;
++ else
++ *sd_attrs &= ~REISERFS_BARRIER_FL;
++ if (inode->i_vflags & V_COW)
++ *sd_attrs |= REISERFS_COW_FL;
++ else
++ *sd_attrs &= ~REISERFS_COW_FL;
+ }
+ }
+
+@@ -3051,6 +3099,22 @@ static ssize_t reiserfs_direct_IO(int rw
+ reiserfs_get_blocks_direct_io, NULL);
+ }
+
++int reiserfs_sync_flags(struct inode *inode)
++{
++ u16 oldflags, newflags;
++
++ oldflags = REISERFS_I(inode)->i_attrs;
++ newflags = oldflags;
++ i_attrs_to_sd_attrs(inode, &newflags);
++
++ if (oldflags ^ newflags) {
++ REISERFS_I(inode)->i_attrs = newflags;
++ inode->i_ctime = CURRENT_TIME_SEC;
++ mark_inode_dirty(inode);
++ }
++ return 0;
++}
++
+ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
+ {
+ struct inode *inode = dentry->d_inode;
+@@ -3104,9 +3168,11 @@ int reiserfs_setattr(struct dentry *dent
+ }
+
+ error = inode_change_ok(inode, attr);
++
+ if (!error) {
+ if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+- (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
++ (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid) ||
++ (ia_valid & ATTR_TAG && attr->ia_tag != inode->i_tag)) {
+ error = reiserfs_chown_xattrs(inode, attr);
+
+ if (!error) {
+@@ -3136,6 +3202,9 @@ int reiserfs_setattr(struct dentry *dent
+ inode->i_uid = attr->ia_uid;
+ if (attr->ia_valid & ATTR_GID)
+ inode->i_gid = attr->ia_gid;
++ if ((attr->ia_valid & ATTR_TAG) &&
++ IS_TAGGED(inode))
++ inode->i_tag = attr->ia_tag;
+ mark_inode_dirty(inode);
+ error =
+ journal_end(&th, inode->i_sb, jbegin_count);
+--- a/fs/reiserfs/ioctl.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/ioctl.c 2008-07-17 17:18:29.000000000 -0400
+@@ -7,6 +7,7 @@
+ #include <linux/mount.h>
+ #include <linux/reiserfs_fs.h>
+ #include <linux/time.h>
++#include <linux/mount.h>
+ #include <asm/uaccess.h>
+ #include <linux/pagemap.h>
+ #include <linux/smp_lock.h>
+@@ -23,7 +24,7 @@
+ int reiserfs_ioctl(struct inode *inode, struct file *filp, unsigned int cmd,
+ unsigned long arg)
+ {
+- unsigned int flags;
++ unsigned int flags, oldflags;
+ int err = 0;
+
+ switch (cmd) {
+@@ -43,6 +44,7 @@ int reiserfs_ioctl(struct inode *inode,
+
+ flags = REISERFS_I(inode)->i_attrs;
+ i_attrs_to_sd_attrs(inode, (__u16 *) & flags);
++ flags &= REISERFS_FL_USER_VISIBLE;
+ return put_user(flags, (int __user *)arg);
+ case REISERFS_IOC_SETFLAGS:{
+ if (!reiserfs_attrs(inode->i_sb))
+@@ -60,6 +62,10 @@ int reiserfs_ioctl(struct inode *inode,
+ err = -EFAULT;
+ goto setflags_out;
+ }
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -EACCES;
++ }
+ /*
+ * Is it quota file? Do not allow user to mess with it
+ */
+@@ -84,6 +90,10 @@ int reiserfs_ioctl(struct inode *inode,
+ goto setflags_out;
+ }
+ }
++
++ oldflags = REISERFS_I(inode)->i_attrs;
++ flags = flags & REISERFS_FL_USER_MODIFIABLE;
++ flags |= oldflags & ~REISERFS_FL_USER_MODIFIABLE;
+ sd_attrs_to_i_attrs(flags, inode);
+ REISERFS_I(inode)->i_attrs = flags;
+ inode->i_ctime = CURRENT_TIME_SEC;
+--- a/fs/reiserfs/namei.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/namei.c 2008-07-16 22:41:36.000000000 -0400
+@@ -17,6 +17,7 @@
+ #include <linux/reiserfs_acl.h>
+ #include <linux/reiserfs_xattr.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_tag.h>
+
+ #define INC_DIR_INODE_NLINK(i) if (i->i_nlink != 1) { inc_nlink(i); if (i->i_nlink >= REISERFS_LINK_MAX) i->i_nlink=1; }
+ #define DEC_DIR_INODE_NLINK(i) if (i->i_nlink != 1) drop_nlink(i);
+@@ -360,6 +361,7 @@ static struct dentry *reiserfs_lookup(st
+ reiserfs_write_unlock(dir->i_sb);
+ return ERR_PTR(-EACCES);
+ }
++ dx_propagate_tag(nd, inode);
+
+ /* Propogate the priv_object flag so we know we're in the priv tree */
+ if (is_reiserfs_priv_object(dir))
+@@ -595,6 +597,7 @@ static int new_inode_init(struct inode *
+ } else {
+ inode->i_gid = current->fsgid;
+ }
++ inode->i_tag = dx_current_fstag(inode->i_sb);
+ DQUOT_INIT(inode);
+ return 0;
+ }
+@@ -1541,6 +1544,7 @@ const struct inode_operations reiserfs_d
+ .listxattr = reiserfs_listxattr,
+ .removexattr = reiserfs_removexattr,
+ .permission = reiserfs_permission,
++ .sync_flags = reiserfs_sync_flags,
+ };
+
+ /*
+@@ -1557,6 +1561,7 @@ const struct inode_operations reiserfs_s
+ .listxattr = reiserfs_listxattr,
+ .removexattr = reiserfs_removexattr,
+ .permission = reiserfs_permission,
++ .sync_flags = reiserfs_sync_flags,
+
+ };
+
+@@ -1570,5 +1575,6 @@ const struct inode_operations reiserfs_s
+ .listxattr = reiserfs_listxattr,
+ .removexattr = reiserfs_removexattr,
+ .permission = reiserfs_permission,
++ .sync_flags = reiserfs_sync_flags,
+
+ };
+--- a/fs/reiserfs/stree.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/stree.c 2008-07-16 22:41:36.000000000 -0400
+@@ -55,6 +55,7 @@
+ #include <linux/reiserfs_fs.h>
+ #include <linux/buffer_head.h>
+ #include <linux/quotaops.h>
++#include <linux/vs_dlimit.h>
+
+ /* Does the buffer contain a disk block which is in the tree. */
+ inline int B_IS_IN_TREE(const struct buffer_head *p_s_bh)
+@@ -1297,6 +1298,7 @@ int reiserfs_delete_item(struct reiserfs
+ "reiserquota delete_item(): freeing %u, id=%u type=%c",
+ quota_cut_bytes, p_s_inode->i_uid, head2type(&s_ih));
+ #endif
++ DLIMIT_FREE_SPACE(p_s_inode, quota_cut_bytes);
+ DQUOT_FREE_SPACE_NODIRTY(p_s_inode, quota_cut_bytes);
+
+ /* Return deleted body length */
+@@ -1385,6 +1387,7 @@ void reiserfs_delete_solid_item(struct r
+ #endif
+ DQUOT_FREE_SPACE_NODIRTY(inode,
+ quota_cut_bytes);
++ DLIMIT_FREE_SPACE(inode, quota_cut_bytes);
+ }
+ break;
+ }
+@@ -1734,6 +1737,7 @@ int reiserfs_cut_from_item(struct reiser
+ "reiserquota cut_from_item(): freeing %u id=%u type=%c",
+ quota_cut_bytes, p_s_inode->i_uid, '?');
+ #endif
++ DLIMIT_FREE_SPACE(p_s_inode, quota_cut_bytes);
+ DQUOT_FREE_SPACE_NODIRTY(p_s_inode, quota_cut_bytes);
+ return n_ret_value;
+ }
+@@ -1975,6 +1979,11 @@ int reiserfs_paste_into_item(struct reis
+ pathrelse(p_s_search_path);
+ return -EDQUOT;
+ }
++ if (DLIMIT_ALLOC_SPACE(inode, n_pasted_size)) {
++ DQUOT_FREE_SPACE_NODIRTY(inode, n_pasted_size);
++ pathrelse(p_s_search_path);
++ return -ENOSPC;
++ }
+ init_tb_struct(th, &s_paste_balance, th->t_super, p_s_search_path,
+ n_pasted_size);
+ #ifdef DISPLACE_NEW_PACKING_LOCALITIES
+@@ -2027,6 +2036,7 @@ int reiserfs_paste_into_item(struct reis
+ n_pasted_size, inode->i_uid,
+ key2type(&(p_s_key->on_disk_key)));
+ #endif
++ DLIMIT_FREE_SPACE(inode, n_pasted_size);
+ DQUOT_FREE_SPACE_NODIRTY(inode, n_pasted_size);
+ return retval;
+ }
+@@ -2064,6 +2074,11 @@ int reiserfs_insert_item(struct reiserfs
+ pathrelse(p_s_path);
+ return -EDQUOT;
+ }
++ if (DLIMIT_ALLOC_SPACE(inode, quota_bytes)) {
++ DQUOT_FREE_SPACE_NODIRTY(inode, quota_bytes);
++ pathrelse(p_s_path);
++ return -ENOSPC;
++ }
+ }
+ init_tb_struct(th, &s_ins_balance, th->t_super, p_s_path,
+ IH_SIZE + ih_item_len(p_s_ih));
+@@ -2111,7 +2126,9 @@ int reiserfs_insert_item(struct reiserfs
+ "reiserquota insert_item(): freeing %u id=%u type=%c",
+ quota_bytes, inode->i_uid, head2type(p_s_ih));
+ #endif
+- if (inode)
++ if (inode) {
++ DLIMIT_FREE_SPACE(inode, quota_bytes);
+ DQUOT_FREE_SPACE_NODIRTY(inode, quota_bytes);
++ }
+ return retval;
+ }
+--- a/fs/reiserfs/super.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/super.c 2008-07-16 22:41:36.000000000 -0400
+@@ -896,6 +896,14 @@ static int reiserfs_parse_options(struct
+ {"user_xattr",.setmask = 1 << REISERFS_UNSUPPORTED_OPT},
+ {"nouser_xattr",.clrmask = 1 << REISERFS_UNSUPPORTED_OPT},
+ #endif
++#ifndef CONFIG_TAGGING_NONE
++ {"tagxid",.setmask = 1 << REISERFS_TAGGED},
++ {"tag",.setmask = 1 << REISERFS_TAGGED},
++ {"notag",.clrmask = 1 << REISERFS_TAGGED},
++#endif
++#ifdef CONFIG_PROPAGATE
++ {"tag",.arg_required = 'T',.values = NULL},
++#endif
+ #ifdef CONFIG_REISERFS_FS_POSIX_ACL
+ {"acl",.setmask = 1 << REISERFS_POSIXACL},
+ {"noacl",.clrmask = 1 << REISERFS_POSIXACL},
+@@ -1159,6 +1167,12 @@ static int reiserfs_remount(struct super
+ goto out_err;
+ }
+
++ if ((mount_options & (1 << REISERFS_TAGGED)) &&
++ !(s->s_flags & MS_TAGGED)) {
++ reiserfs_warning(s, "reiserfs: tagging not permitted on remount.");
++ return -EINVAL;
++ }
++
+ handle_attrs(s);
+
+ /* Add options that are safe here */
+@@ -1618,6 +1632,10 @@ static int reiserfs_fill_super(struct su
+ goto error;
+ }
+
++ /* map mount option tagxid */
++ if (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_TAGGED))
++ s->s_flags |= MS_TAGGED;
++
+ rs = SB_DISK_SUPER_BLOCK(s);
+ /* Let's do basic sanity check to verify that underlying device is not
+ smaller than the filesystem. If the check fails then abort and scream,
+--- a/fs/reiserfs/xattr.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/reiserfs/xattr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -35,6 +35,7 @@
+ #include <linux/namei.h>
+ #include <linux/errno.h>
+ #include <linux/fs.h>
++#include <linux/mount.h>
+ #include <linux/file.h>
+ #include <linux/pagemap.h>
+ #include <linux/xattr.h>
+--- a/fs/stat.c 2008-04-17 12:05:41.000000000 -0400
++++ a/fs/stat.c 2008-07-16 22:41:36.000000000 -0400
+@@ -26,6 +26,7 @@ void generic_fillattr(struct inode *inod
+ stat->nlink = inode->i_nlink;
+ stat->uid = inode->i_uid;
+ stat->gid = inode->i_gid;
++ stat->tag = inode->i_tag;
+ stat->rdev = inode->i_rdev;
+ stat->atime = inode->i_atime;
+ stat->mtime = inode->i_mtime;
+--- a/fs/super.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/super.c 2008-07-17 21:01:21.000000000 -0400
+@@ -38,6 +38,9 @@
+ #include <linux/kobject.h>
+ #include <linux/mutex.h>
+ #include <linux/file.h>
++#include <linux/devpts_fs.h>
++#include <linux/proc_fs.h>
++#include <linux/vs_context.h>
+ #include <asm/uaccess.h>
+ #include "internal.h"
+
+@@ -886,12 +889,18 @@ struct vfsmount *
+ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void *data)
+ {
+ struct vfsmount *mnt;
++ struct super_block *sb;
+ char *secdata = NULL;
+ int error;
+
+ if (!type)
+ return ERR_PTR(-ENODEV);
+
++ error = -EPERM;
++ if ((type->fs_flags & FS_BINARY_MOUNTDATA) &&
++ !vx_capable(CAP_SYS_ADMIN, VXC_BINARY_MOUNT))
++ goto out;
++
+ error = -ENOMEM;
+ mnt = alloc_vfsmnt(name);
+ if (!mnt)
+@@ -912,7 +921,14 @@ vfs_kern_mount(struct file_system_type *
+ goto out_free_secdata;
+ BUG_ON(!mnt->mnt_sb);
+
+- error = security_sb_kern_mount(mnt->mnt_sb, secdata);
++ sb = mnt->mnt_sb;
++ error = -EPERM;
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_BINARY_MOUNT) && !sb->s_bdev &&
++ (sb->s_magic != PROC_SUPER_MAGIC) &&
++ (sb->s_magic != DEVPTS_SUPER_MAGIC))
++ goto out_sb;
++
++ error = security_sb_kern_mount(sb, secdata);
+ if (error)
+ goto out_sb;
+
+--- a/fs/sysfs/mount.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/sysfs/mount.c 2008-07-16 22:41:36.000000000 -0400
+@@ -19,8 +19,6 @@
+
+ #include "sysfs.h"
+
+-/* Random magic number */
+-#define SYSFS_MAGIC 0x62656572
+
+ static struct vfsmount *sysfs_mount;
+ struct super_block * sysfs_sb = NULL;
+@@ -46,7 +44,7 @@ static int sysfs_fill_super(struct super
+
+ sb->s_blocksize = PAGE_CACHE_SIZE;
+ sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
+- sb->s_magic = SYSFS_MAGIC;
++ sb->s_magic = SYSFS_SUPER_MAGIC;
+ sb->s_op = &sysfs_ops;
+ sb->s_time_gran = 1;
+ sysfs_sb = sb;
+--- a/fs/utimes.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/utimes.c 2008-07-16 22:41:36.000000000 -0400
+@@ -8,6 +8,8 @@
+ #include <linux/stat.h>
+ #include <linux/utime.h>
+ #include <linux/syscalls.h>
++#include <linux/mount.h>
++#include <linux/vs_cowbl.h>
+ #include <asm/uaccess.h>
+ #include <asm/unistd.h>
+
+@@ -87,6 +89,9 @@ long do_utimes(int dfd, char __user *fil
+ if (error)
+ goto out;
+
++ error = cow_check_and_break(&nd);
++ if (error)
++ goto dput_and_out;
+ dentry = nd.path.dentry;
+ mnt = nd.path.mnt;
+ }
+--- a/fs/xattr.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xattr.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/module.h>
+ #include <linux/fsnotify.h>
+ #include <linux/audit.h>
++#include <linux/mount.h>
+ #include <asm/uaccess.h>
+
+
+--- a/fs/xfs/linux-2.6/xfs_ioctl.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_ioctl.c 2008-08-11 23:04:52.000000000 -0400
+@@ -800,6 +800,10 @@ xfs_merge_ioc_xflags(
+ xflags |= XFS_XFLAG_IMMUTABLE;
+ else
+ xflags &= ~XFS_XFLAG_IMMUTABLE;
++ if (flags & FS_IXUNLINK_FL)
++ xflags |= XFS_XFLAG_IXUNLINK;
++ else
++ xflags &= ~XFS_XFLAG_IXUNLINK;
+ if (flags & FS_APPEND_FL)
+ xflags |= XFS_XFLAG_APPEND;
+ else
+@@ -822,12 +826,16 @@ xfs_merge_ioc_xflags(
+
+ STATIC unsigned int
+ xfs_di2lxflags(
+- __uint16_t di_flags)
++ __uint16_t di_flags,
++ __uint16_t di_vflags)
+ {
+ unsigned int flags = 0;
+
+ if (di_flags & XFS_DIFLAG_IMMUTABLE)
+ flags |= FS_IMMUTABLE_FL;
++ if (di_flags & XFS_DIFLAG_IXUNLINK)
++ flags |= FS_IXUNLINK_FL;
++
+ if (di_flags & XFS_DIFLAG_APPEND)
+ flags |= FS_APPEND_FL;
+ if (di_flags & XFS_DIFLAG_SYNC)
+@@ -836,6 +844,11 @@ xfs_di2lxflags(
+ flags |= FS_NOATIME_FL;
+ if (di_flags & XFS_DIFLAG_NODUMP)
+ flags |= FS_NODUMP_FL;
++
++ if (di_vflags & XFS_DIVFLAG_BARRIER)
++ flags |= FS_BARRIER_FL;
++ if (di_vflags & XFS_DIVFLAG_COW)
++ flags |= FS_COW_FL;
+ return flags;
+ }
+
+@@ -916,7 +929,8 @@ xfs_ioc_getxflags(
+ {
+ unsigned int flags;
+
+- flags = xfs_di2lxflags(ip->i_d.di_flags);
++ flags = xfs_di2lxflags(ip->i_d.di_flags, ip->i_d.di_vflags);
++ flags &= ~(XFS_XFLAG_BARRIER | XFS_XFLAG_COW);
+ if (copy_to_user(arg, &flags, sizeof(flags)))
+ return -EFAULT;
+ return 0;
+@@ -1096,10 +1110,18 @@ xfs_ioctl(
+ case XFS_IOC_FSGETXATTRA:
+ return xfs_ioc_fsgetxattr(ip, 1, arg);
+ case XFS_IOC_FSSETXATTR:
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -XFS_ERROR(EACCES);
++ }
+ return xfs_ioc_fssetxattr(ip, filp, arg);
+ case XFS_IOC_GETXFLAGS:
+ return xfs_ioc_getxflags(ip, arg);
+ case XFS_IOC_SETXFLAGS:
++ if (IS_BARRIER(inode)) {
++ vxwprintk_task(1, "messing with the barrier.");
++ return -XFS_ERROR(EACCES);
++ }
+ return xfs_ioc_setxflags(ip, filp, arg);
+
+ case XFS_IOC_FSSETDM: {
+--- a/fs/xfs/linux-2.6/xfs_iops.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_iops.c 2008-08-12 02:03:50.000000000 -0400
+@@ -53,6 +53,7 @@
+ #include <linux/namei.h>
+ #include <linux/security.h>
+ #include <linux/falloc.h>
++#include <linux/vs_tag.h>
+
+ /*
+ * Bring the atime in the XFS inode uptodate.
+@@ -389,6 +390,7 @@ xfs_vn_lookup(
+ d_add(dentry, NULL);
+ return NULL;
+ }
++ dx_propagate_tag(nd, vn_to_inode(cvp));
+
+ return d_splice_alias(cip->i_vnode, dentry);
+ }
+@@ -619,6 +621,7 @@ xfs_vn_getattr(
+ stat->nlink = ip->i_d.di_nlink;
+ stat->uid = ip->i_d.di_uid;
+ stat->gid = ip->i_d.di_gid;
++ stat->tag = ip->i_d.di_tag;
+ stat->ino = ip->i_ino;
+ #if XFS_BIG_INUMS
+ stat->ino += mp->m_inoadd;
+@@ -668,6 +671,10 @@ xfs_vn_setattr(
+ int flags = 0;
+ int error;
+
++ error = inode_change_ok(inode, attr);
++ if (error)
++ return error;
++
+ if (ia_valid & ATTR_UID) {
+ vattr.va_mask |= XFS_AT_UID;
+ vattr.va_uid = attr->ia_uid;
+@@ -676,6 +683,10 @@ xfs_vn_setattr(
+ vattr.va_mask |= XFS_AT_GID;
+ vattr.va_gid = attr->ia_gid;
+ }
++ if ((ia_valid & ATTR_TAG) && IS_TAGGED(inode)) {
++ vattr.va_mask |= XFS_AT_TAG;
++ vattr.va_tag = attr->ia_tag;
++ }
+ if (ia_valid & ATTR_SIZE) {
+ vattr.va_mask |= XFS_AT_SIZE;
+ vattr.va_size = attr->ia_size;
+@@ -728,6 +739,57 @@ xfs_vn_truncate(
+ WARN_ON(error);
+ }
+
++/* Propagate flags from i_flags to XFS_I(inode)->di_flags */
++STATIC void
++xfs_get_inode_flags(struct inode *inode)
++{
++ xfs_inode_t *ip = XFS_I(inode);
++ unsigned int flags = inode->i_flags;
++ unsigned int vflags = inode->i_vflags;
++
++ if (flags & S_IMMUTABLE)
++ ip->i_d.di_flags |= XFS_DIFLAG_IMMUTABLE;
++ else
++ ip->i_d.di_flags &= ~XFS_DIFLAG_IMMUTABLE;
++ if (flags & S_IXUNLINK)
++ ip->i_d.di_flags |= XFS_DIFLAG_IXUNLINK;
++ else
++ ip->i_d.di_flags &= ~XFS_DIFLAG_IXUNLINK;
++
++ if (vflags & V_BARRIER)
++ ip->i_d.di_vflags |= XFS_DIVFLAG_BARRIER;
++ else
++ ip->i_d.di_vflags &= ~XFS_DIVFLAG_BARRIER;
++ if (vflags & V_COW)
++ ip->i_d.di_vflags |= XFS_DIVFLAG_COW;
++ else
++ ip->i_d.di_vflags &= ~XFS_DIVFLAG_COW;
++}
++
++STATIC int
++xfs_vn_sync_flags(struct inode *inode)
++{
++ xfs_inode_t *ip = XFS_I(inode);
++ struct bhv_vattr *vattr;
++ int error;
++
++ vattr = kmalloc(sizeof(*vattr), GFP_KERNEL);
++ if (unlikely(!vattr))
++ return -ENOMEM;
++
++ xfs_get_inode_flags(inode);
++
++ vattr->va_mask = XFS_AT_XFLAGS;
++ vattr->va_xflags = xfs_ip2xflags(ip);
++
++ error = -xfs_setattr(ip, vattr, 0, NULL);
++ if (likely(!error))
++ vn_revalidate(XFS_ITOV(ip)); /* update flags */
++
++ kfree(vattr);
++ return error;
++}
++
+ STATIC int
+ xfs_vn_setxattr(
+ struct dentry *dentry,
+@@ -883,6 +945,7 @@ const struct inode_operations xfs_inode_
+ .listxattr = xfs_vn_listxattr,
+ .removexattr = xfs_vn_removexattr,
+ .fallocate = xfs_vn_fallocate,
++ .sync_flags = xfs_vn_sync_flags,
+ };
+
+ const struct inode_operations xfs_dir_inode_operations = {
+@@ -902,6 +965,7 @@ const struct inode_operations xfs_dir_in
+ .getxattr = xfs_vn_getxattr,
+ .listxattr = xfs_vn_listxattr,
+ .removexattr = xfs_vn_removexattr,
++ .sync_flags = xfs_vn_sync_flags,
+ };
+
+ const struct inode_operations xfs_symlink_inode_operations = {
+@@ -915,4 +979,5 @@ const struct inode_operations xfs_symlin
+ .getxattr = xfs_vn_getxattr,
+ .listxattr = xfs_vn_listxattr,
+ .removexattr = xfs_vn_removexattr,
++ .sync_flags = xfs_vn_sync_flags,
+ };
+--- a/fs/xfs/linux-2.6/xfs_linux.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_linux.h 2008-07-16 22:41:36.000000000 -0400
+@@ -127,6 +127,7 @@
+ #define current_pid() (current->pid)
+ #define current_fsuid(cred) (current->fsuid)
+ #define current_fsgid(cred) (current->fsgid)
++#define current_fstag(cred,vp) (dx_current_fstag(vn_to_inode(vp)->i_sb))
+ #define current_test_flags(f) (current->flags & (f))
+ #define current_set_flags_nested(sp, f) \
+ (*(sp) = current->flags, current->flags |= (f))
+--- a/fs/xfs/linux-2.6/xfs_super.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_super.c 2008-07-16 22:41:36.000000000 -0400
+@@ -137,6 +137,9 @@ xfs_args_allocate(
+ #define MNTOPT_DMAPI "dmapi" /* DMI enabled (DMAPI / XDSM) */
+ #define MNTOPT_XDSM "xdsm" /* DMI enabled (DMAPI / XDSM) */
+ #define MNTOPT_DMI "dmi" /* DMI enabled (DMAPI / XDSM) */
++#define MNTOPT_TAGXID "tagxid" /* context tagging for inodes */
++#define MNTOPT_TAGGED "tag" /* context tagging for inodes */
++#define MNTOPT_NOTAGTAG "notag" /* do not use context tagging */
+
+ STATIC unsigned long
+ suffix_strtoul(char *s, char **endp, unsigned int base)
+@@ -355,6 +358,19 @@ xfs_parseargs(
+ } else if (!strcmp(this_char, "irixsgid")) {
+ cmn_err(CE_WARN,
+ "XFS: irixsgid is now a sysctl(2) variable, option is deprecated.");
++#ifndef CONFIG_TAGGING_NONE
++ } else if (!strcmp(this_char, MNTOPT_TAGGED)) {
++ args->flags2 |= XFSMNT2_TAGGED;
++ } else if (!strcmp(this_char, MNTOPT_NOTAGTAG)) {
++ args->flags2 &= ~XFSMNT2_TAGGED;
++ } else if (!strcmp(this_char, MNTOPT_TAGXID)) {
++ args->flags2 |= XFSMNT2_TAGGED;
++#endif
++#ifdef CONFIG_PROPAGATE
++ } else if (!strcmp(this_char, MNTOPT_TAGGED)) {
++ /* use value */
++ args->flags2 |= XFSMNT2_TAGGED;
++#endif
+ } else {
+ cmn_err(CE_WARN,
+ "XFS: unknown mount option [%s].", this_char);
+@@ -591,6 +607,7 @@ xfs_revalidate_inode(
+ inode->i_nlink = ip->i_d.di_nlink;
+ inode->i_uid = ip->i_d.di_uid;
+ inode->i_gid = ip->i_d.di_gid;
++ inode->i_tag = ip->i_d.di_tag;
+
+ switch (inode->i_mode & S_IFMT) {
+ case S_IFBLK:
+@@ -612,10 +629,16 @@ xfs_revalidate_inode(
+ inode->i_mtime.tv_nsec = ip->i_d.di_mtime.t_nsec;
+ inode->i_ctime.tv_sec = ip->i_d.di_ctime.t_sec;
+ inode->i_ctime.tv_nsec = ip->i_d.di_ctime.t_nsec;
++
+ if (ip->i_d.di_flags & XFS_DIFLAG_IMMUTABLE)
+ inode->i_flags |= S_IMMUTABLE;
+ else
+ inode->i_flags &= ~S_IMMUTABLE;
++ if (ip->i_d.di_flags & XFS_DIFLAG_IXUNLINK)
++ inode->i_flags |= S_IXUNLINK;
++ else
++ inode->i_flags &= ~S_IXUNLINK;
++
+ if (ip->i_d.di_flags & XFS_DIFLAG_APPEND)
+ inode->i_flags |= S_APPEND;
+ else
+@@ -628,6 +651,16 @@ xfs_revalidate_inode(
+ inode->i_flags |= S_NOATIME;
+ else
+ inode->i_flags &= ~S_NOATIME;
++
++ if (ip->i_d.di_vflags & XFS_DIVFLAG_BARRIER)
++ inode->i_vflags |= V_BARRIER;
++ else
++ inode->i_vflags &= ~V_BARRIER;
++ if (ip->i_d.di_vflags & XFS_DIVFLAG_COW)
++ inode->i_vflags |= V_COW;
++ else
++ inode->i_vflags &= ~V_COW;
++
+ xfs_iflags_clear(ip, XFS_IMODIFIED);
+ }
+
+@@ -1220,6 +1253,12 @@ xfs_fs_remount(
+ int error;
+
+ error = xfs_parseargs(mp, options, args, 1);
++ if ((args->flags2 & XFSMNT2_TAGGED) &&
++ !(sb->s_flags & MS_TAGGED)) {
++ printk("XFS: %s: tagging not permitted on remount.\n",
++ sb->s_id);
++ error = EINVAL;
++ }
+ if (!error)
+ error = xfs_mntupdate(mp, flags, args);
+ kmem_free(args, sizeof(*args));
+@@ -1336,6 +1375,9 @@ xfs_fs_fill_super(
+ if (error)
+ goto fail_vfsop;
+
++ if (mp->m_flags & XFS_MOUNT_TAGGED)
++ sb->s_flags |= MS_TAGGED;
++
+ sb->s_dirt = 1;
+ sb->s_magic = XFS_SB_MAGIC;
+ sb->s_blocksize = mp->m_sb.sb_blocksize;
+--- a/fs/xfs/linux-2.6/xfs_vnode.c 2008-04-17 12:05:41.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_vnode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -105,6 +105,7 @@ vn_revalidate(
+ inode->i_mode = ip->i_d.di_mode;
+ inode->i_uid = ip->i_d.di_uid;
+ inode->i_gid = ip->i_d.di_gid;
++ inode->i_tag = ip->i_d.di_tag;
+ inode->i_mtime.tv_sec = ip->i_d.di_mtime.t_sec;
+ inode->i_mtime.tv_nsec = ip->i_d.di_mtime.t_nsec;
+ inode->i_ctime.tv_sec = ip->i_d.di_ctime.t_sec;
+@@ -115,6 +116,10 @@ vn_revalidate(
+ inode->i_flags |= S_IMMUTABLE;
+ else
+ inode->i_flags &= ~S_IMMUTABLE;
++ if (xflags & XFS_XFLAG_IXUNLINK)
++ inode->i_flags |= S_IXUNLINK;
++ else
++ inode->i_flags &= ~S_IXUNLINK;
+ if (xflags & XFS_XFLAG_APPEND)
+ inode->i_flags |= S_APPEND;
+ else
+@@ -127,6 +132,15 @@ vn_revalidate(
+ inode->i_flags |= S_NOATIME;
+ else
+ inode->i_flags &= ~S_NOATIME;
++
++ if (xflags & XFS_XFLAG_BARRIER)
++ inode->i_vflags |= V_BARRIER;
++ else
++ inode->i_vflags &= ~V_BARRIER;
++ if (xflags & XFS_XFLAG_COW)
++ inode->i_vflags |= V_COW;
++ else
++ inode->i_vflags &= ~V_COW;
+ xfs_iunlock(ip, XFS_ILOCK_SHARED);
+
+ xfs_iflags_clear(ip, XFS_IMODIFIED);
+--- a/fs/xfs/linux-2.6/xfs_vnode.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/linux-2.6/xfs_vnode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -76,6 +76,7 @@ typedef struct bhv_vattr {
+ xfs_nlink_t va_nlink; /* number of references to file */
+ uid_t va_uid; /* owner user id */
+ gid_t va_gid; /* owner group id */
++ tag_t va_tag; /* owner group id */
+ xfs_ino_t va_nodeid; /* file id */
+ xfs_off_t va_size; /* file size in bytes */
+ u_long va_blocksize; /* blocksize preferred for i/o */
+@@ -124,13 +125,15 @@ typedef struct bhv_vattr {
+ #define XFS_AT_PROJID 0x04000000
+ #define XFS_AT_SIZE_NOPERM 0x08000000
+ #define XFS_AT_GENCOUNT 0x10000000
++#define XFS_AT_TAG 0x20000000
+
+ #define XFS_AT_ALL (XFS_AT_TYPE|XFS_AT_MODE|XFS_AT_UID|XFS_AT_GID|\
+ XFS_AT_FSID|XFS_AT_NODEID|XFS_AT_NLINK|XFS_AT_SIZE|\
+ XFS_AT_ATIME|XFS_AT_MTIME|XFS_AT_CTIME|XFS_AT_RDEV|\
+ XFS_AT_BLKSIZE|XFS_AT_NBLOCKS|XFS_AT_VCODE|XFS_AT_MAC|\
+ XFS_AT_ACL|XFS_AT_CAP|XFS_AT_INF|XFS_AT_XFLAGS|XFS_AT_EXTSIZE|\
+- XFS_AT_NEXTENTS|XFS_AT_ANEXTENTS|XFS_AT_PROJID|XFS_AT_GENCOUNT)
++ XFS_AT_NEXTENTS|XFS_AT_ANEXTENTS|XFS_AT_PROJID|XFS_AT_GENCOUNT\
++ XFS_AT_TAG)
+
+ #define XFS_AT_STAT (XFS_AT_TYPE|XFS_AT_MODE|XFS_AT_UID|XFS_AT_GID|\
+ XFS_AT_FSID|XFS_AT_NODEID|XFS_AT_NLINK|XFS_AT_SIZE|\
+--- a/fs/xfs/quota/xfs_qm_syscalls.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/quota/xfs_qm_syscalls.c 2008-07-16 22:41:36.000000000 -0400
+@@ -17,6 +17,7 @@
+ */
+
+ #include <linux/capability.h>
++#include <linux/vs_context.h>
+
+ #include "xfs.h"
+ #include "xfs_fs.h"
+@@ -205,7 +206,7 @@ xfs_qm_scall_quotaoff(
+ xfs_qoff_logitem_t *qoffstart;
+ int nculprits;
+
+- if (!force && !capable(CAP_SYS_ADMIN))
++ if (!force && !vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return XFS_ERROR(EPERM);
+ /*
+ * No file system can have quotas enabled on disk but not in core.
+@@ -383,7 +384,7 @@ xfs_qm_scall_trunc_qfiles(
+ int error = 0, error2 = 0;
+ xfs_inode_t *qip;
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return XFS_ERROR(EPERM);
+ if (!xfs_sb_version_hasquota(&mp->m_sb) || flags == 0) {
+ qdprintk("qtrunc flags=%x m_qflags=%x\n", flags, mp->m_qflags);
+@@ -426,7 +427,7 @@ xfs_qm_scall_quotaon(
+ uint accflags;
+ __int64_t sbflags;
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return XFS_ERROR(EPERM);
+
+ flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD);
+@@ -595,7 +596,7 @@ xfs_qm_scall_setqlim(
+ int error;
+ xfs_qcnt_t hard, soft;
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_QUOTA_CTL))
+ return XFS_ERROR(EPERM);
+
+ if ((newlim->d_fieldmask &
+--- a/fs/xfs/xfs_clnt.h 2008-04-17 12:05:42.000000000 -0400
++++ a/fs/xfs/xfs_clnt.h 2008-07-16 22:41:36.000000000 -0400
+@@ -100,5 +100,6 @@ struct xfs_mount_args {
+ * I/O size in stat(2) */
+ #define XFSMNT2_FILESTREAMS 0x00000002 /* enable the filestreams
+ * allocator */
++#define XFSMNT2_TAGGED 0x10000000 /* context tagging */
+
+ #endif /* __XFS_CLNT_H__ */
+--- a/fs/xfs/xfs_dinode.h 2008-04-17 12:05:42.000000000 -0400
++++ a/fs/xfs/xfs_dinode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -53,7 +53,9 @@ typedef struct xfs_dinode_core {
+ __be32 di_gid; /* owner's group id */
+ __be32 di_nlink; /* number of links to file */
+ __be16 di_projid; /* owner's project id */
+- __u8 di_pad[8]; /* unused, zeroed space */
++ __be16 di_tag; /* context tagging */
++ __be16 di_vflags; /* vserver specific flags */
++ __u8 di_pad[4]; /* unused, zeroed space */
+ __be16 di_flushiter; /* incremented on flush */
+ xfs_timestamp_t di_atime; /* time last accessed */
+ xfs_timestamp_t di_mtime; /* time last modified */
+@@ -136,7 +138,9 @@ typedef struct xfs_dinode
+ #define XFS_DI_NEXT_UNLINKED 0x1000000
+ #define XFS_DI_U 0x2000000
+ #define XFS_DI_A 0x4000000
+-#define XFS_DI_NUM_BITS 27
++#define XFS_DI_TAG 0x8000000
++#define XFS_DI_VFLAGS 0x10000000
++#define XFS_DI_NUM_BITS 29
+ #define XFS_DI_ALL_BITS ((1 << XFS_DI_NUM_BITS) - 1)
+ #define XFS_DI_CORE_BITS (XFS_DI_ALL_BITS & ~(XFS_DI_U|XFS_DI_A))
+
+@@ -223,6 +227,8 @@ typedef enum xfs_dinode_fmt
+ #define XFS_DIFLAG_EXTSZINHERIT_BIT 12 /* inherit inode extent size */
+ #define XFS_DIFLAG_NODEFRAG_BIT 13 /* do not reorganize/defragment */
+ #define XFS_DIFLAG_FILESTREAM_BIT 14 /* use filestream allocator */
++#define XFS_DIFLAG_IXUNLINK_BIT 15 /* Immutable inver on unlink */
++
+ #define XFS_DIFLAG_REALTIME (1 << XFS_DIFLAG_REALTIME_BIT)
+ #define XFS_DIFLAG_PREALLOC (1 << XFS_DIFLAG_PREALLOC_BIT)
+ #define XFS_DIFLAG_NEWRTBM (1 << XFS_DIFLAG_NEWRTBM_BIT)
+@@ -238,6 +244,7 @@ typedef enum xfs_dinode_fmt
+ #define XFS_DIFLAG_EXTSZINHERIT (1 << XFS_DIFLAG_EXTSZINHERIT_BIT)
+ #define XFS_DIFLAG_NODEFRAG (1 << XFS_DIFLAG_NODEFRAG_BIT)
+ #define XFS_DIFLAG_FILESTREAM (1 << XFS_DIFLAG_FILESTREAM_BIT)
++#define XFS_DIFLAG_IXUNLINK (1 << XFS_DIFLAG_IXUNLINK_BIT)
+
+ #ifdef CONFIG_XFS_RT
+ #define XFS_IS_REALTIME_INODE(ip) ((ip)->i_d.di_flags & XFS_DIFLAG_REALTIME)
+@@ -250,6 +257,10 @@ typedef enum xfs_dinode_fmt
+ XFS_DIFLAG_IMMUTABLE | XFS_DIFLAG_APPEND | XFS_DIFLAG_SYNC | \
+ XFS_DIFLAG_NOATIME | XFS_DIFLAG_NODUMP | XFS_DIFLAG_RTINHERIT | \
+ XFS_DIFLAG_PROJINHERIT | XFS_DIFLAG_NOSYMLINKS | XFS_DIFLAG_EXTSIZE | \
+- XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG | XFS_DIFLAG_FILESTREAM)
++ XFS_DIFLAG_EXTSZINHERIT | XFS_DIFLAG_NODEFRAG | XFS_DIFLAG_FILESTREAM | \
++ XFS_DIFLAG_IXUNLINK)
++
++#define XFS_DIVFLAG_BARRIER 0x01
++#define XFS_DIVFLAG_COW 0x02
+
+ #endif /* __XFS_DINODE_H__ */
+--- a/fs/xfs/xfs_fs.h 2008-04-17 12:05:42.000000000 -0400
++++ a/fs/xfs/xfs_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -67,6 +67,9 @@ struct fsxattr {
+ #define XFS_XFLAG_EXTSZINHERIT 0x00001000 /* inherit inode extent size */
+ #define XFS_XFLAG_NODEFRAG 0x00002000 /* do not defragment */
+ #define XFS_XFLAG_FILESTREAM 0x00004000 /* use filestream allocator */
++#define XFS_XFLAG_IXUNLINK 0x00008000 /* immutable invert on unlink */
++#define XFS_XFLAG_BARRIER 0x10000000 /* chroot() barrier */
++#define XFS_XFLAG_COW 0x20000000 /* copy on write mark */
+ #define XFS_XFLAG_HASATTR 0x80000000 /* no DIFLAG for this */
+
+ /*
+@@ -296,7 +299,8 @@ typedef struct xfs_bstat {
+ __s32 bs_extents; /* number of extents */
+ __u32 bs_gen; /* generation count */
+ __u16 bs_projid; /* project id */
+- unsigned char bs_pad[14]; /* pad space, unused */
++ __u16 bs_tag; /* context tagging */
++ unsigned char bs_pad[12]; /* pad space, unused */
+ __u32 bs_dmevmask; /* DMIG event mask */
+ __u16 bs_dmstate; /* DMIG state info */
+ __u16 bs_aextents; /* attribute number of extents */
+--- a/fs/xfs/xfs_ialloc.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_ialloc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -84,6 +84,8 @@ xfs_ialloc_log_di(
+ offsetof(xfs_dinode_t, di_next_unlinked),
+ offsetof(xfs_dinode_t, di_u),
+ offsetof(xfs_dinode_t, di_a),
++ offsetof(xfs_dinode_core_t, di_tag),
++ offsetof(xfs_dinode_core_t, di_vflags),
+ sizeof(xfs_dinode_t)
+ };
+
+--- a/fs/xfs/xfs_inode.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -249,6 +249,7 @@ xfs_inotobp(
+ return 0;
+ }
+
++#include <linux/vs_tag.h>
+
+ /*
+ * This routine is called to map an inode to the buffer containing
+@@ -660,15 +661,25 @@ xfs_iformat_btree(
+ void
+ xfs_dinode_from_disk(
+ xfs_icdinode_t *to,
+- xfs_dinode_core_t *from)
++ xfs_dinode_core_t *from,
++ int tagged)
+ {
++ uint32_t uid, gid, tag;
++
+ to->di_magic = be16_to_cpu(from->di_magic);
+ to->di_mode = be16_to_cpu(from->di_mode);
+ to->di_version = from ->di_version;
+ to->di_format = from->di_format;
+ to->di_onlink = be16_to_cpu(from->di_onlink);
+- to->di_uid = be32_to_cpu(from->di_uid);
+- to->di_gid = be32_to_cpu(from->di_gid);
++
++ uid = be32_to_cpu(from->di_uid);
++ gid = be32_to_cpu(from->di_gid);
++ tag = be16_to_cpu(from->di_tag);
++
++ to->di_uid = INOTAG_UID(tagged, uid, gid);
++ to->di_gid = INOTAG_GID(tagged, uid, gid);
++ to->di_tag = INOTAG_TAG(tagged, uid, gid, tag);
++
+ to->di_nlink = be32_to_cpu(from->di_nlink);
+ to->di_projid = be16_to_cpu(from->di_projid);
+ memcpy(to->di_pad, from->di_pad, sizeof(to->di_pad));
+@@ -689,21 +700,26 @@ xfs_dinode_from_disk(
+ to->di_dmevmask = be32_to_cpu(from->di_dmevmask);
+ to->di_dmstate = be16_to_cpu(from->di_dmstate);
+ to->di_flags = be16_to_cpu(from->di_flags);
++ to->di_vflags = be16_to_cpu(from->di_vflags);
+ to->di_gen = be32_to_cpu(from->di_gen);
+ }
+
+ void
+ xfs_dinode_to_disk(
+ xfs_dinode_core_t *to,
+- xfs_icdinode_t *from)
++ xfs_icdinode_t *from,
++ int tagged)
+ {
+ to->di_magic = cpu_to_be16(from->di_magic);
+ to->di_mode = cpu_to_be16(from->di_mode);
+ to->di_version = from ->di_version;
+ to->di_format = from->di_format;
+ to->di_onlink = cpu_to_be16(from->di_onlink);
+- to->di_uid = cpu_to_be32(from->di_uid);
+- to->di_gid = cpu_to_be32(from->di_gid);
++
++ to->di_uid = cpu_to_be32(TAGINO_UID(tagged, from->di_uid, from->di_tag));
++ to->di_gid = cpu_to_be32(TAGINO_GID(tagged, from->di_gid, from->di_tag));
++ to->di_tag = cpu_to_be16(TAGINO_TAG(tagged, from->di_tag));
++
+ to->di_nlink = cpu_to_be32(from->di_nlink);
+ to->di_projid = cpu_to_be16(from->di_projid);
+ memcpy(to->di_pad, from->di_pad, sizeof(to->di_pad));
+@@ -724,12 +740,14 @@ xfs_dinode_to_disk(
+ to->di_dmevmask = cpu_to_be32(from->di_dmevmask);
+ to->di_dmstate = cpu_to_be16(from->di_dmstate);
+ to->di_flags = cpu_to_be16(from->di_flags);
++ to->di_vflags = cpu_to_be16(from->di_vflags);
+ to->di_gen = cpu_to_be32(from->di_gen);
+ }
+
+ STATIC uint
+ _xfs_dic2xflags(
+- __uint16_t di_flags)
++ __uint16_t di_flags,
++ __uint16_t di_vflags)
+ {
+ uint flags = 0;
+
+@@ -740,6 +758,8 @@ _xfs_dic2xflags(
+ flags |= XFS_XFLAG_PREALLOC;
+ if (di_flags & XFS_DIFLAG_IMMUTABLE)
+ flags |= XFS_XFLAG_IMMUTABLE;
++ if (di_flags & XFS_DIFLAG_IXUNLINK)
++ flags |= XFS_XFLAG_IXUNLINK;
+ if (di_flags & XFS_DIFLAG_APPEND)
+ flags |= XFS_XFLAG_APPEND;
+ if (di_flags & XFS_DIFLAG_SYNC)
+@@ -763,7 +783,10 @@ _xfs_dic2xflags(
+ if (di_flags & XFS_DIFLAG_FILESTREAM)
+ flags |= XFS_XFLAG_FILESTREAM;
+ }
+-
++ if (di_vflags & XFS_DIVFLAG_BARRIER)
++ flags |= XFS_XFLAG_BARRIER;
++ if (di_vflags & XFS_DIVFLAG_COW)
++ flags |= XFS_XFLAG_COW;
+ return flags;
+ }
+
+@@ -773,7 +796,7 @@ xfs_ip2xflags(
+ {
+ xfs_icdinode_t *dic = &ip->i_d;
+
+- return _xfs_dic2xflags(dic->di_flags) |
++ return _xfs_dic2xflags(dic->di_flags, dic->di_vflags) |
+ (XFS_IFORK_Q(ip) ? XFS_XFLAG_HASATTR : 0);
+ }
+
+@@ -783,7 +806,7 @@ xfs_dic2xflags(
+ {
+ xfs_dinode_core_t *dic = &dip->di_core;
+
+- return _xfs_dic2xflags(be16_to_cpu(dic->di_flags)) |
++ return _xfs_dic2xflags(be16_to_cpu(dic->di_flags), be16_to_cpu(dic->di_vflags)) |
+ (XFS_DFORK_Q(dip) ? XFS_XFLAG_HASATTR : 0);
+ }
+
+@@ -878,7 +901,8 @@ xfs_iread(
+ * Otherwise, just get the truly permanent information.
+ */
+ if (dip->di_core.di_mode) {
+- xfs_dinode_from_disk(&ip->i_d, &dip->di_core);
++ xfs_dinode_from_disk(&ip->i_d, &dip->di_core,
++ mp->m_flags & XFS_MOUNT_TAGGED);
+ error = xfs_iformat(ip, dip);
+ if (error) {
+ kmem_zone_free(xfs_inode_zone, ip);
+@@ -1084,6 +1108,7 @@ xfs_ialloc(
+ ASSERT(ip->i_d.di_nlink == nlink);
+ ip->i_d.di_uid = current_fsuid(cr);
+ ip->i_d.di_gid = current_fsgid(cr);
++ ip->i_d.di_tag = current_fstag(cr, vp);
+ ip->i_d.di_projid = prid;
+ memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad));
+
+@@ -1138,6 +1163,7 @@ xfs_ialloc(
+ ip->i_d.di_dmevmask = 0;
+ ip->i_d.di_dmstate = 0;
+ ip->i_d.di_flags = 0;
++ ip->i_d.di_vflags = 0;
+ flags = XFS_ILOG_CORE;
+ switch (mode & S_IFMT) {
+ case S_IFIFO:
+@@ -1872,8 +1898,8 @@ xfs_iunlink(
+ agi_ok =
+ be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC &&
+ XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum));
+- if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IUNLINK,
+- XFS_RANDOM_IUNLINK))) {
++ if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IXUNLINK,
++ XFS_RANDOM_IXUNLINK))) {
+ XFS_CORRUPTION_ERROR("xfs_iunlink", XFS_ERRLEVEL_LOW, mp, agi);
+ xfs_trans_brelse(tp, agibp);
+ return XFS_ERROR(EFSCORRUPTED);
+@@ -1974,8 +2000,8 @@ xfs_iunlink_remove(
+ agi_ok =
+ be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC &&
+ XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum));
+- if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IUNLINK_REMOVE,
+- XFS_RANDOM_IUNLINK_REMOVE))) {
++ if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IXUNLINK_REMOVE,
++ XFS_RANDOM_IXUNLINK_REMOVE))) {
+ XFS_CORRUPTION_ERROR("xfs_iunlink_remove", XFS_ERRLEVEL_LOW,
+ mp, agi);
+ xfs_trans_brelse(tp, agibp);
+@@ -2307,6 +2333,7 @@ xfs_ifree(
+ }
+ ip->i_d.di_mode = 0; /* mark incore inode as free */
+ ip->i_d.di_flags = 0;
++ ip->i_d.di_vflags = 0;
+ ip->i_d.di_dmevmask = 0;
+ ip->i_d.di_forkoff = 0; /* mark the attr fork not in use */
+ ip->i_df.if_ext_max =
+@@ -3403,7 +3430,8 @@ xfs_iflush_int(
+ * because if the inode is dirty at all the core must
+ * be.
+ */
+- xfs_dinode_to_disk(&dip->di_core, &ip->i_d);
++ xfs_dinode_to_disk(&dip->di_core, &ip->i_d,
++ mp->m_flags & XFS_MOUNT_TAGGED);
+
+ /* Wrap, we never let the log put out DI_MAX_FLUSH */
+ if (ip->i_d.di_flushiter == DI_MAX_FLUSH)
+--- a/fs/xfs/xfs_inode.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_inode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -174,7 +174,9 @@ typedef struct xfs_icdinode {
+ __uint32_t di_gid; /* owner's group id */
+ __uint32_t di_nlink; /* number of links to file */
+ __uint16_t di_projid; /* owner's project id */
+- __uint8_t di_pad[8]; /* unused, zeroed space */
++ __uint16_t di_tag; /* context tagging */
++ __uint16_t di_vflags; /* vserver specific flags */
++ __uint8_t di_pad[4]; /* unused, zeroed space */
+ __uint16_t di_flushiter; /* incremented on flush */
+ xfs_ictimestamp_t di_atime; /* time last accessed */
+ xfs_ictimestamp_t di_mtime; /* time last modified */
+@@ -495,9 +497,9 @@ int xfs_ialloc(struct xfs_trans *, xfs_
+ xfs_nlink_t, xfs_dev_t, struct cred *, xfs_prid_t,
+ int, struct xfs_buf **, boolean_t *, xfs_inode_t **);
+ void xfs_dinode_from_disk(struct xfs_icdinode *,
+- struct xfs_dinode_core *);
++ struct xfs_dinode_core *, int tagged);
+ void xfs_dinode_to_disk(struct xfs_dinode_core *,
+- struct xfs_icdinode *);
++ struct xfs_icdinode *, int tagged);
+
+ uint xfs_ip2xflags(struct xfs_inode *);
+ uint xfs_dic2xflags(struct xfs_dinode *);
+--- a/fs/xfs/xfs_itable.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_itable.c 2008-07-16 22:41:36.000000000 -0400
+@@ -84,6 +84,7 @@ xfs_bulkstat_one_iget(
+ buf->bs_mode = dic->di_mode;
+ buf->bs_uid = dic->di_uid;
+ buf->bs_gid = dic->di_gid;
++ buf->bs_tag = dic->di_tag;
+ buf->bs_size = dic->di_size;
+ vn_atime_to_bstime(vp, &buf->bs_atime);
+ buf->bs_mtime.tv_sec = dic->di_mtime.t_sec;
+--- a/fs/xfs/xfs_log_recover.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_log_recover.c 2008-07-16 22:41:36.000000000 -0400
+@@ -2418,7 +2418,8 @@ xlog_recover_do_inode_trans(
+
+ /* The core is in in-core format */
+ xfs_dinode_to_disk(&dip->di_core,
+- (xfs_icdinode_t *)item->ri_buf[1].i_addr);
++ (xfs_icdinode_t *)item->ri_buf[1].i_addr,
++ mp->m_flags & XFS_MOUNT_TAGGED);
+
+ /* the rest is in on-disk format */
+ if (item->ri_buf[1].i_len > sizeof(xfs_dinode_core_t)) {
+--- a/fs/xfs/xfs_mount.h 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_mount.h 2008-07-16 22:41:36.000000000 -0400
+@@ -379,6 +379,7 @@ typedef struct xfs_mount {
+ #define XFS_MOUNT_FILESTREAMS (1ULL << 24) /* enable the filestreams
+ allocator */
+
++#define XFS_MOUNT_TAGGED (1ULL << 31) /* context tagging */
+
+ /*
+ * Default minimum read and write sizes.
+--- a/fs/xfs/xfs_vfsops.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_vfsops.c 2008-07-16 22:41:36.000000000 -0400
+@@ -287,6 +287,8 @@ xfs_start_flags(
+
+ if (ap->flags2 & XFSMNT2_COMPAT_IOSIZE)
+ mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE;
++ if (ap->flags2 & XFSMNT2_TAGGED)
++ mp->m_flags |= XFS_MOUNT_TAGGED;
+
+ /*
+ * no recovery flag requires a read-only mount
+@@ -399,7 +401,6 @@ xfs_finish_flags(
+ if (ap->flags & XFSMNT_PQUOTAENF)
+ mp->m_qflags |= XFS_OQUOTA_ENFD;
+ }
+-
+ return 0;
+ }
+
+--- a/fs/xfs/xfs_vnodeops.c 2008-07-14 17:22:50.000000000 -0400
++++ a/fs/xfs/xfs_vnodeops.c 2008-07-16 22:41:36.000000000 -0400
+@@ -93,6 +93,7 @@ xfs_setattr(
+ uint commit_flags=0;
+ uid_t uid=0, iuid=0;
+ gid_t gid=0, igid=0;
++ tag_t tag=0, itag=0;
+ int timeflags = 0;
+ xfs_prid_t projid=0, iprojid=0;
+ struct xfs_dquot *udqp, *gdqp, *olddquot1, *olddquot2;
+@@ -143,6 +144,7 @@ xfs_setattr(
+ (mask & (XFS_AT_UID|XFS_AT_GID|XFS_AT_PROJID))) {
+ uint qflags = 0;
+
++ /* TODO: handle tagging? */
+ if ((mask & XFS_AT_UID) && XFS_IS_UQUOTA_ON(mp)) {
+ uid = vap->va_uid;
+ qflags |= XFS_QMOPT_UQUOTA;
+@@ -222,6 +224,8 @@ xfs_setattr(
+ if (mask &
+ (XFS_AT_MODE|XFS_AT_XFLAGS|XFS_AT_EXTSIZE|XFS_AT_UID|
+ XFS_AT_GID|XFS_AT_PROJID)) {
++ /* TODO: handle tagging? */
++
+ /*
+ * CAP_FOWNER overrides the following restrictions:
+ *
+@@ -270,7 +274,7 @@ xfs_setattr(
+ * and can change the group id only to a group of which he
+ * or she is a member.
+ */
+- if (mask & (XFS_AT_UID|XFS_AT_GID|XFS_AT_PROJID)) {
++ if (mask & (XFS_AT_UID|XFS_AT_GID|XFS_AT_TAG|XFS_AT_PROJID)) {
+ /*
+ * These IDs could have changed since we last looked at them.
+ * But, we're assured that if the ownership did change
+@@ -278,10 +282,12 @@ xfs_setattr(
+ * would have changed also.
+ */
+ iuid = ip->i_d.di_uid;
+- iprojid = ip->i_d.di_projid;
+ igid = ip->i_d.di_gid;
+- gid = (mask & XFS_AT_GID) ? vap->va_gid : igid;
++ itag = ip->i_d.di_tag;
++ iprojid = ip->i_d.di_projid;
+ uid = (mask & XFS_AT_UID) ? vap->va_uid : iuid;
++ gid = (mask & XFS_AT_GID) ? vap->va_gid : igid;
++ tag = (mask & XFS_AT_TAG) ? vap->va_tag : itag;
+ projid = (mask & XFS_AT_PROJID) ? (xfs_prid_t)vap->va_projid :
+ iprojid;
+
+@@ -309,6 +315,7 @@ xfs_setattr(
+ if ((XFS_IS_UQUOTA_ON(mp) && iuid != uid) ||
+ (XFS_IS_PQUOTA_ON(mp) && iprojid != projid) ||
+ (XFS_IS_GQUOTA_ON(mp) && igid != gid)) {
++ /* TODO: handle tagging? */
+ ASSERT(tp);
+ code = XFS_QM_DQVOPCHOWNRESV(mp, tp, ip, udqp, gdqp,
+ capable(CAP_FOWNER) ?
+@@ -423,7 +430,7 @@ xfs_setattr(
+ */
+ if ((mask & XFS_AT_XFLAGS) &&
+ (ip->i_d.di_flags &
+- (XFS_DIFLAG_IMMUTABLE|XFS_DIFLAG_APPEND) ||
++ (XFS_DIFLAG_IMMUTABLE | XFS_DIFLAG_APPEND) ||
+ (vap->va_xflags &
+ (XFS_XFLAG_IMMUTABLE | XFS_XFLAG_APPEND))) &&
+ !capable(CAP_LINUX_IMMUTABLE)) {
+@@ -559,7 +566,7 @@ xfs_setattr(
+ * and can change the group id only to a group of which he
+ * or she is a member.
+ */
+- if (mask & (XFS_AT_UID|XFS_AT_GID|XFS_AT_PROJID)) {
++ if (mask & (XFS_AT_UID|XFS_AT_GID|XFS_AT_TAG|XFS_AT_PROJID)) {
+ /*
+ * CAP_FSETID overrides the following restrictions:
+ *
+@@ -575,6 +582,9 @@ xfs_setattr(
+ * Change the ownerships and register quota modifications
+ * in the transaction.
+ */
++ if (itag != tag) {
++ ip->i_d.di_tag = tag;
++ }
+ if (iuid != uid) {
+ if (XFS_IS_UQUOTA_ON(mp)) {
+ ASSERT(mask & XFS_AT_UID);
+@@ -650,11 +660,14 @@ xfs_setattr(
+ }
+ if (mask & XFS_AT_XFLAGS) {
+ uint di_flags;
++ uint di_vflags = 0;
+
+ /* can't set PREALLOC this way, just preserve it */
+ di_flags = (ip->i_d.di_flags & XFS_DIFLAG_PREALLOC);
+ if (vap->va_xflags & XFS_XFLAG_IMMUTABLE)
+ di_flags |= XFS_DIFLAG_IMMUTABLE;
++ if (vap->va_xflags & XFS_XFLAG_IXUNLINK)
++ di_flags |= XFS_DIFLAG_IXUNLINK;
+ if (vap->va_xflags & XFS_XFLAG_APPEND)
+ di_flags |= XFS_DIFLAG_APPEND;
+ if (vap->va_xflags & XFS_XFLAG_SYNC)
+@@ -683,6 +696,11 @@ xfs_setattr(
+ di_flags |= XFS_DIFLAG_EXTSIZE;
+ }
+ ip->i_d.di_flags = di_flags;
++ if (vap->va_xflags & XFS_XFLAG_BARRIER)
++ di_vflags |= XFS_DIVFLAG_BARRIER;
++ if (vap->va_xflags & XFS_XFLAG_COW)
++ di_vflags |= XFS_DIVFLAG_COW;
++ ip->i_d.di_vflags = di_vflags;
+ }
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+ timeflags |= XFS_ICHGTIME_CHG;
+--- a/include/asm-arm/tlb.h 2008-04-17 12:05:43.000000000 -0400
++++ a/include/asm-arm/tlb.h 2008-07-16 22:41:36.000000000 -0400
+@@ -28,6 +28,7 @@
+ #else /* !CONFIG_MMU */
+
+ #include <asm/pgalloc.h>
++#include <linux/vs_memory.h>
+
+ /*
+ * TLB handling. This allows us to remove pages from the page
+--- a/include/asm-blackfin/unistd.h 2008-07-14 17:22:52.000000000 -0400
++++ a/include/asm-blackfin/unistd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -278,7 +278,7 @@
+ #define __NR_tgkill 271
+ #define __NR_utimes 272
+ #define __NR_fadvise64_64 273
+- /* 274 __NR_vserver */
++#define __NR_vserver 274
+ /* 275 __NR_mbind */
+ /* 276 __NR_get_mempolicy */
+ /* 277 __NR_set_mempolicy */
+--- a/include/asm-generic/tlb.h 2008-04-17 12:05:43.000000000 -0400
++++ a/include/asm-generic/tlb.h 2008-07-16 22:41:36.000000000 -0400
+@@ -14,6 +14,7 @@
+ #define _ASM_GENERIC__TLB_H
+
+ #include <linux/swap.h>
++#include <linux/vs_memory.h>
+ #include <asm/pgalloc.h>
+ #include <asm/tlbflush.h>
+
+--- a/include/asm-ia64/tlb.h 2008-07-14 17:22:53.000000000 -0400
++++ a/include/asm-ia64/tlb.h 2008-07-16 22:41:36.000000000 -0400
+@@ -40,6 +40,7 @@
+ #include <linux/mm.h>
+ #include <linux/pagemap.h>
+ #include <linux/swap.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/pgalloc.h>
+ #include <asm/processor.h>
+--- a/include/asm-powerpc/systbl.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/asm-powerpc/systbl.h 2008-07-16 22:41:36.000000000 -0400
+@@ -260,7 +260,7 @@ COMPAT_SYS_SPU(fstatfs64)
+ SYSX(sys_ni_syscall, ppc_fadvise64_64, ppc_fadvise64_64)
+ PPC_SYS_SPU(rtas)
+ OLDSYS(debug_setcontext)
+-SYSCALL(ni_syscall)
++SYSX(sys_vserver, sys32_vserver, sys_vserver)
+ COMPAT_SYS(migrate_pages)
+ COMPAT_SYS(mbind)
+ COMPAT_SYS(get_mempolicy)
+--- a/include/asm-powerpc/unistd.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/asm-powerpc/unistd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -275,7 +275,7 @@
+ #endif
+ #define __NR_rtas 255
+ #define __NR_sys_debug_setcontext 256
+-/* Number 257 is reserved for vserver */
++#define __NR_vserver 257
+ #define __NR_migrate_pages 258
+ #define __NR_mbind 259
+ #define __NR_get_mempolicy 260
+--- a/include/asm-s390/unistd.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/asm-s390/unistd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -202,7 +202,7 @@
+ #define __NR_clock_gettime (__NR_timer_create+6)
+ #define __NR_clock_getres (__NR_timer_create+7)
+ #define __NR_clock_nanosleep (__NR_timer_create+8)
+-/* Number 263 is reserved for vserver */
++#define __NR_vserver 263
+ #define __NR_statfs64 265
+ #define __NR_fstatfs64 266
+ #define __NR_remap_file_pages 267
+--- a/include/asm-sparc/unistd.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/asm-sparc/unistd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -282,7 +282,7 @@
+ #define __NR_timer_getoverrun 264
+ #define __NR_timer_delete 265
+ #define __NR_timer_create 266
+-/* #define __NR_vserver 267 Reserved for VSERVER */
++#define __NR_vserver 267
+ #define __NR_io_setup 268
+ #define __NR_io_destroy 269
+ #define __NR_io_submit 270
+--- a/include/asm-sparc64/tlb.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/asm-sparc64/tlb.h 2008-07-16 22:41:36.000000000 -0400
+@@ -3,6 +3,7 @@
+
+ #include <linux/swap.h>
+ #include <linux/pagemap.h>
++#include <linux/vs_memory.h>
+ #include <asm/pgalloc.h>
+ #include <asm/tlbflush.h>
+ #include <asm/mmu_context.h>
+--- a/include/asm-sparc64/unistd.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/asm-sparc64/unistd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -284,7 +284,7 @@
+ #define __NR_timer_getoverrun 264
+ #define __NR_timer_delete 265
+ #define __NR_timer_create 266
+-/* #define __NR_vserver 267 Reserved for VSERVER */
++#define __NR_vserver 267
+ #define __NR_io_setup 268
+ #define __NR_io_destroy 269
+ #define __NR_io_submit 270
+--- a/include/asm-x86/unistd_64.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/asm-x86/unistd_64.h 2008-07-16 22:41:36.000000000 -0400
+@@ -535,7 +535,7 @@ __SYSCALL(__NR_tgkill, sys_tgkill)
+ #define __NR_utimes 235
+ __SYSCALL(__NR_utimes, sys_utimes)
+ #define __NR_vserver 236
+-__SYSCALL(__NR_vserver, sys_ni_syscall)
++__SYSCALL(__NR_vserver, sys_vserver)
+ #define __NR_mbind 237
+ __SYSCALL(__NR_mbind, sys_mbind)
+ #define __NR_set_mempolicy 238
+--- a/include/linux/capability.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/linux/capability.h 2008-07-16 22:41:36.000000000 -0400
+@@ -274,6 +274,7 @@ typedef struct kernel_cap_struct {
+ arbitrary SCSI commands */
+ /* Allow setting encryption key on loopback filesystem */
+ /* Allow setting zone reclaim policy */
++/* Allow the selection of a security context */
+
+ #define CAP_SYS_ADMIN 21
+
+@@ -346,7 +347,13 @@ typedef struct kernel_cap_struct {
+
+ #define CAP_MAC_ADMIN 33
+
+-#define CAP_LAST_CAP CAP_MAC_ADMIN
++/* Allow context manipulations */
++/* Allow changing context info on files */
++
++#define CAP_CONTEXT 34
++
++
++#define CAP_LAST_CAP CAP_CONTEXT
+
+ #define cap_valid(x) ((x) >= 0 && (x) <= CAP_LAST_CAP)
+
+--- a/include/linux/devpts_fs.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/linux/devpts_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -34,5 +34,4 @@ static inline void devpts_pty_kill(int n
+
+ #endif
+
+-
+ #endif /* _LINUX_DEVPTS_FS_H */
+--- a/include/linux/ext2_fs.h 2008-04-17 11:31:39.000000000 -0400
++++ a/include/linux/ext2_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -189,8 +189,12 @@ struct ext2_group_desc
+ #define EXT2_NOTAIL_FL FS_NOTAIL_FL /* file tail should not be merged */
+ #define EXT2_DIRSYNC_FL FS_DIRSYNC_FL /* dirsync behaviour (directories only) */
+ #define EXT2_TOPDIR_FL FS_TOPDIR_FL /* Top of directory hierarchies*/
++#define EXT2_IXUNLINK_FL FS_IXUNLINK_FL /* Immutable invert on unlink */
+ #define EXT2_RESERVED_FL FS_RESERVED_FL /* reserved for ext2 lib */
+
++#define EXT2_BARRIER_FL FS_BARRIER_FL /* Barrier for chroot() */
++#define EXT2_COW_FL FS_COW_FL /* Copy on Write marker */
++
+ #define EXT2_FL_USER_VISIBLE FS_FL_USER_VISIBLE /* User visible flags */
+ #define EXT2_FL_USER_MODIFIABLE FS_FL_USER_MODIFIABLE /* User modifiable flags */
+
+@@ -247,7 +251,7 @@ struct ext2_inode {
+ struct {
+ __u8 l_i_frag; /* Fragment number */
+ __u8 l_i_fsize; /* Fragment size */
+- __u16 i_pad1;
++ __u16 l_i_tag; /* Context Tag */
+ __le16 l_i_uid_high; /* these 2 fields */
+ __le16 l_i_gid_high; /* were reserved2[0] */
+ __u32 l_i_reserved2;
+@@ -279,6 +283,7 @@ struct ext2_inode {
+ #define i_gid_low i_gid
+ #define i_uid_high osd2.linux2.l_i_uid_high
+ #define i_gid_high osd2.linux2.l_i_gid_high
++#define i_raw_tag osd2.linux2.l_i_tag
+ #define i_reserved2 osd2.linux2.l_i_reserved2
+ #endif
+
+@@ -323,6 +328,7 @@ struct ext2_inode {
+ #define EXT2_MOUNT_USRQUOTA 0x020000 /* user quota */
+ #define EXT2_MOUNT_GRPQUOTA 0x040000 /* group quota */
+ #define EXT2_MOUNT_RESERVATION 0x080000 /* Preallocation */
++#define EXT2_MOUNT_TAGGED (1<<24) /* Enable Context Tags */
+
+
+ #define clear_opt(o, opt) o &= ~EXT2_MOUNT_##opt
+--- a/include/linux/ext3_fs.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/linux/ext3_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -173,10 +173,14 @@ struct ext3_group_desc
+ #define EXT3_NOTAIL_FL 0x00008000 /* file tail should not be merged */
+ #define EXT3_DIRSYNC_FL 0x00010000 /* dirsync behaviour (directories only) */
+ #define EXT3_TOPDIR_FL 0x00020000 /* Top of directory hierarchies*/
++#define EXT3_IXUNLINK_FL 0x01000000 /* Immutable invert on unlink */
+ #define EXT3_RESERVED_FL 0x80000000 /* reserved for ext3 lib */
+
+-#define EXT3_FL_USER_VISIBLE 0x0003DFFF /* User visible flags */
+-#define EXT3_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */
++#define EXT3_BARRIER_FL 0x10000000 /* Barrier for chroot() */
++#define EXT3_COW_FL 0x20000000 /* Copy on Write marker */
++
++#define EXT3_FL_USER_VISIBLE 0x0103DFFF /* User visible flags */
++#define EXT3_FL_USER_MODIFIABLE 0x010380FF /* User modifiable flags */
+
+ /*
+ * Inode dynamic state flags
+@@ -292,7 +296,7 @@ struct ext3_inode {
+ struct {
+ __u8 l_i_frag; /* Fragment number */
+ __u8 l_i_fsize; /* Fragment size */
+- __u16 i_pad1;
++ __u16 l_i_tag; /* Context Tag */
+ __le16 l_i_uid_high; /* these 2 fields */
+ __le16 l_i_gid_high; /* were reserved2[0] */
+ __u32 l_i_reserved2;
+@@ -326,6 +330,7 @@ struct ext3_inode {
+ #define i_gid_low i_gid
+ #define i_uid_high osd2.linux2.l_i_uid_high
+ #define i_gid_high osd2.linux2.l_i_gid_high
++#define i_raw_tag osd2.linux2.l_i_tag
+ #define i_reserved2 osd2.linux2.l_i_reserved2
+
+ #elif defined(__GNU__)
+@@ -380,6 +385,7 @@ struct ext3_inode {
+ #define EXT3_MOUNT_QUOTA 0x80000 /* Some quota option set */
+ #define EXT3_MOUNT_USRQUOTA 0x100000 /* "old" user quota */
+ #define EXT3_MOUNT_GRPQUOTA 0x200000 /* "old" group quota */
++#define EXT3_MOUNT_TAGGED (1<<24) /* Enable Context Tags */
+
+ /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
+ #ifndef _LINUX_EXT2_FS_H
+@@ -822,6 +828,7 @@ struct buffer_head * ext3_bread (handle_
+ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode,
+ sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result,
+ int create, int extend_disksize);
++extern int ext3_sync_flags(struct inode *inode);
+
+ extern struct inode *ext3_iget(struct super_block *, unsigned long);
+ extern int ext3_write_inode (struct inode *, int);
+--- a/include/linux/fs.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/linux/fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -126,6 +126,8 @@ extern int dir_notify_enable;
+ #define MS_RELATIME (1<<21) /* Update atime relative to mtime/ctime. */
+ #define MS_KERNMOUNT (1<<22) /* this is a kern_mount call */
+ #define MS_I_VERSION (1<<23) /* Update inode I_version field */
++#define MS_TAGGED (1<<24) /* use generic inode tagging */
++#define MS_TAGID (1<<25) /* use specific tag for this mount */
+ #define MS_ACTIVE (1<<30)
+ #define MS_NOUSER (1<<31)
+
+@@ -152,6 +154,14 @@ extern int dir_notify_enable;
+ #define S_NOCMTIME 128 /* Do not update file c/mtime */
+ #define S_SWAPFILE 256 /* Do not truncate: swapon got its bmaps */
+ #define S_PRIVATE 512 /* Inode is fs-internal */
++#define S_IXUNLINK 1024 /* Immutable Invert on unlink */
++
++/* Linux-VServer related Inode flags */
++
++#define V_VALID 1
++#define V_XATTR 2
++#define V_BARRIER 4 /* Barrier for chroot() */
++#define V_COW 8 /* Copy on Write */
+
+ /*
+ * Note that nosuid etc flags are inode-specific: setting some file-system
+@@ -174,12 +184,15 @@ extern int dir_notify_enable;
+ #define IS_DIRSYNC(inode) (__IS_FLG(inode, MS_SYNCHRONOUS|MS_DIRSYNC) || \
+ ((inode)->i_flags & (S_SYNC|S_DIRSYNC)))
+ #define IS_MANDLOCK(inode) __IS_FLG(inode, MS_MANDLOCK)
+-#define IS_NOATIME(inode) __IS_FLG(inode, MS_RDONLY|MS_NOATIME)
+-#define IS_I_VERSION(inode) __IS_FLG(inode, MS_I_VERSION)
++#define IS_NOATIME(inode) __IS_FLG(inode, MS_RDONLY|MS_NOATIME)
++#define IS_I_VERSION(inode) __IS_FLG(inode, MS_I_VERSION)
++#define IS_TAGGED(inode) __IS_FLG(inode, MS_TAGGED)
+
+ #define IS_NOQUOTA(inode) ((inode)->i_flags & S_NOQUOTA)
+ #define IS_APPEND(inode) ((inode)->i_flags & S_APPEND)
+ #define IS_IMMUTABLE(inode) ((inode)->i_flags & S_IMMUTABLE)
++#define IS_IXUNLINK(inode) ((inode)->i_flags & S_IXUNLINK)
++#define IS_IXORUNLINK(inode) ((IS_IXUNLINK(inode) ? S_IMMUTABLE : 0) ^ IS_IMMUTABLE(inode))
+ #define IS_POSIXACL(inode) __IS_FLG(inode, MS_POSIXACL)
+
+ #define IS_DEADDIR(inode) ((inode)->i_flags & S_DEAD)
+@@ -187,6 +200,16 @@ extern int dir_notify_enable;
+ #define IS_SWAPFILE(inode) ((inode)->i_flags & S_SWAPFILE)
+ #define IS_PRIVATE(inode) ((inode)->i_flags & S_PRIVATE)
+
++#define IS_BARRIER(inode) (S_ISDIR((inode)->i_mode) && ((inode)->i_vflags & V_BARRIER))
++
++#ifdef CONFIG_VSERVER_COWBL
++# define IS_COW(inode) (IS_IXUNLINK(inode) && IS_IMMUTABLE(inode))
++# define IS_COW_LINK(inode) (S_ISREG((inode)->i_mode) && ((inode)->i_nlink > 1))
++#else
++# define IS_COW(inode) (0)
++# define IS_COW_LINK(inode) (0)
++#endif
++
+ /* the read-only stuff doesn't really belong here, but any other place is
+ probably as bad and I don't want to create yet another include file. */
+
+@@ -260,11 +283,14 @@ extern int dir_notify_enable;
+ #define FS_TOPDIR_FL 0x00020000 /* Top of directory hierarchies*/
+ #define FS_EXTENT_FL 0x00080000 /* Extents */
+ #define FS_DIRECTIO_FL 0x00100000 /* Use direct i/o */
++#define FS_IXUNLINK_FL 0x01000000 /* Immutable invert on unlink */
+ #define FS_RESERVED_FL 0x80000000 /* reserved for ext2 lib */
+
+-#define FS_FL_USER_VISIBLE 0x0003DFFF /* User visible flags */
+-#define FS_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */
++#define FS_BARRIER_FL 0x10000000 /* Barrier for chroot() */
++#define FS_COW_FL 0x20000000 /* Copy on Write marker */
+
++#define FS_FL_USER_VISIBLE 0x0103DFFF /* User visible flags */
++#define FS_FL_USER_MODIFIABLE 0x010380FF /* User modifiable flags */
+
+ #define SYNC_FILE_RANGE_WAIT_BEFORE 1
+ #define SYNC_FILE_RANGE_WRITE 2
+@@ -334,6 +360,7 @@ typedef void (dio_iodone_t)(struct kiocb
+ #define ATTR_FILE 8192
+ #define ATTR_KILL_PRIV 16384
+ #define ATTR_OPEN 32768 /* Truncating from open(O_TRUNC) */
++#define ATTR_TAG 65536
+
+ /*
+ * This is the Inode Attributes structure, used for notify_change(). It
+@@ -349,6 +376,7 @@ struct iattr {
+ umode_t ia_mode;
+ uid_t ia_uid;
+ gid_t ia_gid;
++ tag_t ia_tag;
+ loff_t ia_size;
+ struct timespec ia_atime;
+ struct timespec ia_mtime;
+@@ -362,6 +390,9 @@ struct iattr {
+ struct file *ia_file;
+ };
+
++#define ATTR_FLAG_BARRIER 512 /* Barrier for chroot() */
++#define ATTR_FLAG_IXUNLINK 1024 /* Immutable invert on unlink */
++
+ /*
+ * Includes for diskquotas.
+ */
+@@ -600,7 +631,9 @@ struct inode {
+ unsigned int i_nlink;
+ uid_t i_uid;
+ gid_t i_gid;
++ tag_t i_tag;
+ dev_t i_rdev;
++ dev_t i_mdev;
+ u64 i_version;
+ loff_t i_size;
+ #ifdef __NEED_I_SIZE_ORDERED
+@@ -648,7 +681,8 @@ struct inode {
+ unsigned long i_state;
+ unsigned long dirtied_when; /* jiffies of first dirtying */
+
+- unsigned int i_flags;
++ unsigned short i_flags;
++ unsigned short i_vflags;
+
+ atomic_t i_writecount;
+ #ifdef CONFIG_SECURITY
+@@ -735,12 +769,12 @@ static inline void i_size_write(struct i
+
+ static inline unsigned iminor(const struct inode *inode)
+ {
+- return MINOR(inode->i_rdev);
++ return MINOR(inode->i_mdev);
+ }
+
+ static inline unsigned imajor(const struct inode *inode)
+ {
+- return MAJOR(inode->i_rdev);
++ return MAJOR(inode->i_mdev);
+ }
+
+ extern struct block_device *I_BDEV(struct inode *inode);
+@@ -798,6 +832,7 @@ struct file {
+ loff_t f_pos;
+ struct fown_struct f_owner;
+ unsigned int f_uid, f_gid;
++ xid_t f_xid;
+ struct file_ra_state f_ra;
+
+ u64 f_version;
+@@ -926,6 +961,7 @@ struct file_lock {
+ unsigned char fl_type;
+ loff_t fl_start;
+ loff_t fl_end;
++ xid_t fl_xid;
+
+ struct fasync_struct * fl_fasync; /* for lease break notifications */
+ unsigned long fl_break_time; /* for nonblocking lease breaks */
+@@ -1272,6 +1308,7 @@ struct inode_operations {
+ void (*truncate_range)(struct inode *, loff_t, loff_t);
+ long (*fallocate)(struct inode *inode, int mode, loff_t offset,
+ loff_t len);
++ int (*sync_flags) (struct inode *);
+ };
+
+ struct seq_file;
+@@ -1287,6 +1324,7 @@ extern ssize_t vfs_readv(struct file *,
+ unsigned long, loff_t *);
+ extern ssize_t vfs_writev(struct file *, const struct iovec __user *,
+ unsigned long, loff_t *);
++ssize_t vfs_sendfile(struct file *, struct file *, loff_t *, size_t, loff_t);
+
+ struct super_operations {
+ struct inode *(*alloc_inode)(struct super_block *sb);
+@@ -1971,6 +2009,7 @@ extern int dcache_dir_open(struct inode
+ extern int dcache_dir_close(struct inode *, struct file *);
+ extern loff_t dcache_dir_lseek(struct file *, loff_t, int);
+ extern int dcache_readdir(struct file *, void *, filldir_t);
++extern int dcache_readdir_filter(struct file *, void *, filldir_t, int (*)(struct dentry *));
+ extern int simple_getattr(struct vfsmount *, struct dentry *, struct kstat *);
+ extern int simple_statfs(struct dentry *, struct kstatfs *);
+ extern int simple_link(struct dentry *, struct inode *, struct dentry *);
+--- a/include/linux/if_tun.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/linux/if_tun.h 2008-07-16 22:41:36.000000000 -0400
+@@ -42,6 +42,7 @@
+ #define TUNSETOWNER _IOW('T', 204, int)
+ #define TUNSETLINK _IOW('T', 205, int)
+ #define TUNSETGROUP _IOW('T', 206, int)
++#define TUNSETNID _IOW('T', 215, int)
+
+ /* TUNSETIFF ifr flags */
+ #define IFF_TUN 0x0001
+--- a/include/linux/init_task.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/linux/init_task.h 2008-07-16 22:41:36.000000000 -0400
+@@ -179,6 +179,10 @@ extern struct group_info init_groups;
+ INIT_IDS \
+ INIT_TRACE_IRQFLAGS \
+ INIT_LOCKDEP \
++ .xid = 0, \
++ .vx_info = NULL, \
++ .nid = 0, \
++ .nx_info = NULL, \
+ }
+
+
+--- a/include/linux/interrupt.h 2008-07-14 17:22:54.000000000 -0400
++++ a/include/linux/interrupt.h 2008-07-16 22:41:36.000000000 -0400
+@@ -8,8 +8,8 @@
+ #include <linux/preempt.h>
+ #include <linux/cpumask.h>
+ #include <linux/irqreturn.h>
+-#include <linux/hardirq.h>
+ #include <linux/sched.h>
++#include <linux/hardirq.h>
+ #include <linux/irqflags.h>
+ #include <asm/atomic.h>
+ #include <asm/ptrace.h>
+--- a/include/linux/ipc.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/linux/ipc.h 2008-07-16 22:41:36.000000000 -0400
+@@ -93,6 +93,7 @@ struct kern_ipc_perm
+ key_t key;
+ uid_t uid;
+ gid_t gid;
++ xid_t xid;
+ uid_t cuid;
+ gid_t cgid;
+ mode_t mode;
+--- a/include/linux/loop.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/linux/loop.h 2008-07-16 22:41:36.000000000 -0400
+@@ -45,6 +45,7 @@ struct loop_device {
+ struct loop_func_table *lo_encryption;
+ __u32 lo_init[2];
+ uid_t lo_key_owner; /* Who set the key */
++ xid_t lo_xid;
+ int (*ioctl)(struct loop_device *, int cmd,
+ unsigned long arg);
+
+--- a/include/linux/magic.h 2008-04-17 11:31:39.000000000 -0400
++++ a/include/linux/magic.h 2008-07-16 22:41:36.000000000 -0400
+@@ -3,7 +3,7 @@
+
+ #define ADFS_SUPER_MAGIC 0xadf5
+ #define AFFS_SUPER_MAGIC 0xadff
+-#define AFS_SUPER_MAGIC 0x5346414F
++#define AFS_SUPER_MAGIC 0x5346414F
+ #define AUTOFS_SUPER_MAGIC 0x0187
+ #define CODA_SUPER_MAGIC 0x73757245
+ #define EFS_SUPER_MAGIC 0x414A53
+@@ -26,6 +26,7 @@
+ #define NFS_SUPER_MAGIC 0x6969
+ #define OPENPROM_SUPER_MAGIC 0x9fa1
+ #define PROC_SUPER_MAGIC 0x9fa0
++#define DEVPTS_SUPER_MAGIC 0x1cd1
+ #define QNX4_SUPER_MAGIC 0x002f /* qnx4 fs detection */
+
+ #define REISERFS_SUPER_MAGIC 0x52654973 /* used by gcc */
+--- a/include/linux/major.h 2008-04-17 10:37:24.000000000 -0400
++++ a/include/linux/major.h 2008-07-16 22:41:36.000000000 -0400
+@@ -15,6 +15,7 @@
+ #define HD_MAJOR IDE0_MAJOR
+ #define PTY_SLAVE_MAJOR 3
+ #define TTY_MAJOR 4
++#define VROOT_MAJOR 4
+ #define TTYAUX_MAJOR 5
+ #define LP_MAJOR 6
+ #define VCS_MAJOR 7
+--- a/include/linux/mm_types.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/mm_types.h 2008-07-16 22:41:36.000000000 -0400
+@@ -205,6 +205,7 @@ struct mm_struct {
+
+ /* Architecture-specific MM context */
+ mm_context_t context;
++ struct vx_info *mm_vx_info;
+
+ /* Swap token stuff */
+ /*
+--- a/include/linux/mount.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/mount.h 2008-07-16 22:47:42.000000000 -0400
+@@ -37,6 +37,10 @@ struct mnt_namespace;
+ #define MNT_UNBINDABLE 0x2000 /* if the vfsmount is a unbindable mount */
+ #define MNT_PNODE_MASK 0x3000 /* propagation flag mask */
+
++#define MNT_TAGID 0x10000
++#define MNT_NOTAG 0x20000
++#define MNT_NOTAGCHECK 0x40000
++
+ struct vfsmount {
+ struct list_head mnt_hash;
+ struct vfsmount *mnt_parent; /* fs we are mounted on */
+@@ -71,6 +75,7 @@ struct vfsmount {
+ * are held, and all mnt_writer[]s on this mount have 0 as their ->count
+ */
+ atomic_t __mnt_writers;
++ tag_t mnt_tag; /* tagging used for vfsmount */
+ };
+
+ static inline struct vfsmount *mntget(struct vfsmount *mnt)
+--- a/include/linux/net.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/net.h 2008-07-16 22:41:36.000000000 -0400
+@@ -66,6 +66,7 @@ typedef enum {
+ #define SOCK_NOSPACE 2
+ #define SOCK_PASSCRED 3
+ #define SOCK_PASSSEC 4
++#define SOCK_USER_SOCKET 5
+
+ #ifndef ARCH_HAS_SOCKET_TYPES
+ /**
+--- a/include/linux/nfs_mount.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/linux/nfs_mount.h 2008-07-16 22:41:36.000000000 -0400
+@@ -63,6 +63,7 @@ struct nfs_mount_data {
+ #define NFS_MOUNT_SECFLAVOUR 0x2000 /* 5 */
+ #define NFS_MOUNT_NORDIRPLUS 0x4000 /* 5 */
+ #define NFS_MOUNT_UNSHARED 0x8000 /* 5 */
+-#define NFS_MOUNT_FLAGMASK 0xFFFF
++#define NFS_MOUNT_TAGGED 0x10000 /* context tagging */
++#define NFS_MOUNT_FLAGMASK 0x1FFFF
+
+ #endif
+--- a/include/linux/nsproxy.h 2008-04-17 11:31:39.000000000 -0400
++++ a/include/linux/nsproxy.h 2008-07-16 22:41:36.000000000 -0400
+@@ -3,6 +3,7 @@
+
+ #include <linux/spinlock.h>
+ #include <linux/sched.h>
++#include <linux/vserver/debug.h>
+
+ struct mnt_namespace;
+ struct uts_namespace;
+@@ -63,22 +64,33 @@ static inline struct nsproxy *task_nspro
+ }
+
+ int copy_namespaces(unsigned long flags, struct task_struct *tsk);
++struct nsproxy *copy_nsproxy(struct nsproxy *orig);
+ void exit_task_namespaces(struct task_struct *tsk);
+ void switch_task_namespaces(struct task_struct *tsk, struct nsproxy *new);
+ void free_nsproxy(struct nsproxy *ns);
+ int unshare_nsproxy_namespaces(unsigned long, struct nsproxy **,
+ struct fs_struct *);
+
+-static inline void put_nsproxy(struct nsproxy *ns)
++#define get_nsproxy(n) __get_nsproxy(n, __FILE__, __LINE__)
++
++static inline void __get_nsproxy(struct nsproxy *ns,
++ const char *_file, int _line)
+ {
+- if (atomic_dec_and_test(&ns->count)) {
+- free_nsproxy(ns);
+- }
++ vxlprintk(VXD_CBIT(space, 0), "get_nsproxy(%p[%u])",
++ ns, atomic_read(&ns->count), _file, _line);
++ atomic_inc(&ns->count);
+ }
+
+-static inline void get_nsproxy(struct nsproxy *ns)
++#define put_nsproxy(n) __put_nsproxy(n, __FILE__, __LINE__)
++
++static inline void __put_nsproxy(struct nsproxy *ns,
++ const char *_file, int _line)
+ {
+- atomic_inc(&ns->count);
++ vxlprintk(VXD_CBIT(space, 0), "put_nsproxy(%p[%u])",
++ ns, atomic_read(&ns->count), _file, _line);
++ if (atomic_dec_and_test(&ns->count)) {
++ free_nsproxy(ns);
++ }
+ }
+
+ #ifdef CONFIG_CGROUP_NS
+--- a/include/linux/pid.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/pid.h 2008-07-16 22:41:36.000000000 -0400
+@@ -8,7 +8,8 @@ enum pid_type
+ PIDTYPE_PID,
+ PIDTYPE_PGID,
+ PIDTYPE_SID,
+- PIDTYPE_MAX
++ PIDTYPE_MAX,
++ PIDTYPE_REALPID
+ };
+
+ /*
+@@ -144,6 +145,7 @@ static inline pid_t pid_nr(struct pid *p
+ }
+
+ pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns);
++pid_t pid_unmapped_nr_ns(struct pid *pid, struct pid_namespace *ns);
+ pid_t pid_vnr(struct pid *pid);
+
+ #define do_each_pid_task(pid, type, task) \
+--- a/include/linux/proc_fs.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/proc_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -59,6 +59,7 @@ struct proc_dir_entry {
+ nlink_t nlink;
+ uid_t uid;
+ gid_t gid;
++ int vx_flags;
+ loff_t size;
+ const struct inode_operations *proc_iops;
+ /*
+@@ -274,16 +275,23 @@ static inline void kclist_add(struct kco
+ extern void kclist_add(struct kcore_list *, void *, size_t);
+ #endif
+
++struct vx_info;
++struct nx_info;
++
+ union proc_op {
+ int (*proc_get_link)(struct inode *, struct path *);
+ int (*proc_read)(struct task_struct *task, char *page);
+ int (*proc_show)(struct seq_file *m,
+ struct pid_namespace *ns, struct pid *pid,
+ struct task_struct *task);
++ int (*proc_vs_read)(char *page);
++ int (*proc_vxi_read)(struct vx_info *vxi, char *page);
++ int (*proc_nxi_read)(struct nx_info *nxi, char *page);
+ };
+
+ struct proc_inode {
+ struct pid *pid;
++ int vx_flags;
+ int fd;
+ union proc_op op;
+ struct proc_dir_entry *pde;
+--- a/include/linux/reiserfs_fs.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/reiserfs_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -837,6 +837,11 @@ struct stat_data_v1 {
+ #define REISERFS_COMPR_FL FS_COMPR_FL
+ #define REISERFS_NOTAIL_FL FS_NOTAIL_FL
+
++/* unfortunately reiserfs sdattr is only 16 bit */
++#define REISERFS_IXUNLINK_FL (FS_IXUNLINK_FL >> 16)
++#define REISERFS_BARRIER_FL (FS_BARRIER_FL >> 16)
++#define REISERFS_COW_FL (FS_COW_FL >> 16)
++
+ /* persistent flags that file inherits from the parent directory */
+ #define REISERFS_INHERIT_MASK ( REISERFS_IMMUTABLE_FL | \
+ REISERFS_SYNC_FL | \
+@@ -846,6 +851,9 @@ struct stat_data_v1 {
+ REISERFS_COMPR_FL | \
+ REISERFS_NOTAIL_FL )
+
++#define REISERFS_FL_USER_VISIBLE 0x80FF
++#define REISERFS_FL_USER_MODIFIABLE 0x80FF
++
+ /* Stat Data on disk (reiserfs version of UFS disk inode minus the
+ address blocks) */
+ struct stat_data {
+@@ -1911,6 +1919,7 @@ static inline void reiserfs_update_sd(st
+ void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode);
+ void i_attrs_to_sd_attrs(struct inode *inode, __u16 * sd_attrs);
+ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr);
++int reiserfs_sync_flags(struct inode *inode);
+
+ /* namei.c */
+ void set_de_name_and_namelen(struct reiserfs_dir_entry *de);
+--- a/include/linux/reiserfs_fs_sb.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/reiserfs_fs_sb.h 2008-07-16 22:41:36.000000000 -0400
+@@ -455,6 +455,7 @@ enum reiserfs_mount_options {
+ REISERFS_POSIXACL,
+ REISERFS_BARRIER_NONE,
+ REISERFS_BARRIER_FLUSH,
++ REISERFS_TAGGED,
+
+ /* Actions on error */
+ REISERFS_ERROR_PANIC,
+--- a/include/linux/sched.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/sched.h 2008-07-27 14:07:42.000000000 -0400
+@@ -71,7 +71,6 @@ struct sched_param {
+ #include <linux/fs_struct.h>
+ #include <linux/compiler.h>
+ #include <linux/completion.h>
+-#include <linux/pid.h>
+ #include <linux/percpu.h>
+ #include <linux/topology.h>
+ #include <linux/proportions.h>
+@@ -87,6 +86,7 @@ struct sched_param {
+ #include <linux/task_io_accounting.h>
+ #include <linux/kobject.h>
+ #include <linux/latencytop.h>
++#include <linux/pid.h>
+
+ #include <asm/processor.h>
+
+@@ -355,25 +355,27 @@ extern void arch_unmap_area_topdown(stru
+ * The mm counters are not protected by its page_table_lock,
+ * so must be incremented atomically.
+ */
+-#define set_mm_counter(mm, member, value) atomic_long_set(&(mm)->_##member, value)
+-#define get_mm_counter(mm, member) ((unsigned long)atomic_long_read(&(mm)->_##member))
+-#define add_mm_counter(mm, member, value) atomic_long_add(value, &(mm)->_##member)
+-#define inc_mm_counter(mm, member) atomic_long_inc(&(mm)->_##member)
+-#define dec_mm_counter(mm, member) atomic_long_dec(&(mm)->_##member)
+-
++#define __set_mm_counter(mm, member, value) \
++ atomic_long_set(&(mm)->_##member, value)
++#define get_mm_counter(mm, member) \
++ ((unsigned long)atomic_long_read(&(mm)->_##member))
+ #else /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */
+ /*
+ * The mm counters are protected by its page_table_lock,
+ * so can be incremented directly.
+ */
+-#define set_mm_counter(mm, member, value) (mm)->_##member = (value)
++#define __set_mm_counter(mm, member, value) (mm)->_##member = (value)
+ #define get_mm_counter(mm, member) ((mm)->_##member)
+-#define add_mm_counter(mm, member, value) (mm)->_##member += (value)
+-#define inc_mm_counter(mm, member) (mm)->_##member++
+-#define dec_mm_counter(mm, member) (mm)->_##member--
+
+ #endif /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */
+
++#define set_mm_counter(mm, member, value) \
++ vx_ ## member ## pages_sub((mm), (get_mm_counter(mm, member) - value))
++#define add_mm_counter(mm, member, value) \
++ vx_ ## member ## pages_add((mm), (value))
++#define inc_mm_counter(mm, member) vx_ ## member ## pages_inc((mm))
++#define dec_mm_counter(mm, member) vx_ ## member ## pages_dec((mm))
++
+ #define get_mm_rss(mm) \
+ (get_mm_counter(mm, file_rss) + get_mm_counter(mm, anon_rss))
+ #define update_hiwater_rss(mm) do { \
+@@ -1194,6 +1196,14 @@ struct task_struct {
+ #endif
+ seccomp_t seccomp;
+
++/* vserver context data */
++ struct vx_info *vx_info;
++ struct nx_info *nx_info;
++
++ xid_t xid;
++ nid_t nid;
++ tag_t tag;
++
+ /* Thread group tracking */
+ u32 parent_exec_id;
+ u32 self_exec_id;
+@@ -1382,6 +1392,11 @@ struct pid_namespace;
+ * see also pid_nr() etc in include/linux/pid.h
+ */
+
++#include <linux/vserver/base.h>
++#include <linux/vserver/context.h>
++#include <linux/vserver/debug.h>
++#include <linux/vserver/pid.h>
++
+ static inline pid_t task_pid_nr(struct task_struct *tsk)
+ {
+ return tsk->pid;
+@@ -1391,7 +1406,7 @@ pid_t task_pid_nr_ns(struct task_struct
+
+ static inline pid_t task_pid_vnr(struct task_struct *tsk)
+ {
+- return pid_vnr(task_pid(tsk));
++ return vx_map_pid(pid_vnr(task_pid(tsk)));
+ }
+
+
+@@ -1404,7 +1419,7 @@ pid_t task_tgid_nr_ns(struct task_struct
+
+ static inline pid_t task_tgid_vnr(struct task_struct *tsk)
+ {
+- return pid_vnr(task_tgid(tsk));
++ return vx_map_tgid(pid_vnr(task_tgid(tsk)));
+ }
+
+
+@@ -1707,9 +1722,9 @@ extern struct pid_namespace init_pid_ns;
+ extern struct task_struct *find_task_by_pid_type_ns(int type, int pid,
+ struct pid_namespace *ns);
+
+-static inline struct task_struct *__deprecated find_task_by_pid(pid_t nr)
++static inline struct task_struct *__deprecated find_task_by_pid(pid_t nr, struct pid_namespace *ns)
+ {
+- return find_task_by_pid_type_ns(PIDTYPE_PID, nr, &init_pid_ns);
++ return find_task_by_pid_type_ns(PIDTYPE_PID, nr, ns);
+ }
+ extern struct task_struct *find_task_by_vpid(pid_t nr);
+ extern struct task_struct *find_task_by_pid_ns(pid_t nr,
+--- a/include/linux/shmem_fs.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/shmem_fs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -8,6 +8,9 @@
+
+ #define SHMEM_NR_DIRECT 16
+
++#define TMPFS_SUPER_MAGIC 0x01021994
++
++
+ struct shmem_inode_info {
+ spinlock_t lock;
+ unsigned long flags;
+--- a/include/linux/stat.h 2008-04-17 10:33:07.000000000 -0400
++++ a/include/linux/stat.h 2008-07-16 22:41:36.000000000 -0400
+@@ -66,6 +66,7 @@ struct kstat {
+ unsigned int nlink;
+ uid_t uid;
+ gid_t gid;
++ tag_t tag;
+ dev_t rdev;
+ loff_t size;
+ struct timespec atime;
+--- a/include/linux/sunrpc/auth.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/sunrpc/auth.h 2008-07-16 22:41:36.000000000 -0400
+@@ -25,6 +25,7 @@
+ struct auth_cred {
+ uid_t uid;
+ gid_t gid;
++ tag_t tag;
+ struct group_info *group_info;
+ unsigned char machine_cred : 1;
+ };
+--- a/include/linux/sunrpc/clnt.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/sunrpc/clnt.h 2008-07-16 22:41:36.000000000 -0400
+@@ -42,7 +42,8 @@ struct rpc_clnt {
+
+ unsigned int cl_softrtry : 1,/* soft timeouts */
+ cl_discrtry : 1,/* disconnect before retry */
+- cl_autobind : 1;/* use getport() */
++ cl_autobind : 1,/* use getport() */
++ cl_tag : 1;/* context tagging */
+
+ struct rpc_rtt * cl_rtt; /* RTO estimator data */
+ const struct rpc_timeout *cl_timeout; /* Timeout strategy */
+--- a/include/linux/syscalls.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/syscalls.h 2008-07-16 22:41:36.000000000 -0400
+@@ -295,6 +295,8 @@ asmlinkage long sys_symlink(const char _
+ asmlinkage long sys_unlink(const char __user *pathname);
+ asmlinkage long sys_rename(const char __user *oldname,
+ const char __user *newname);
++asmlinkage long sys_copyfile(const char __user *from, const char __user *to,
++ umode_t mode);
+ asmlinkage long sys_chmod(const char __user *filename, mode_t mode);
+ asmlinkage long sys_fchmod(unsigned int fd, mode_t mode);
+
+--- a/include/linux/sysctl.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/sysctl.h 2008-07-16 22:41:36.000000000 -0400
+@@ -70,6 +70,7 @@ enum
+ CTL_ABI=9, /* Binary emulation */
+ CTL_CPU=10, /* CPU stuff (speed scaling, etc) */
+ CTL_ARLAN=254, /* arlan wireless driver */
++ CTL_VSERVER=4242, /* Linux-VServer debug */
+ CTL_S390DBF=5677, /* s390 debug */
+ CTL_SUNRPC=7249, /* sunrpc debug */
+ CTL_PM=9899, /* frv power management */
+@@ -104,6 +105,7 @@ enum
+
+ KERN_PANIC=15, /* int: panic timeout */
+ KERN_REALROOTDEV=16, /* real root device to mount after initrd */
++ KERN_VSHELPER=17, /* string: path to vshelper policy agent */
+
+ KERN_SPARC_REBOOT=21, /* reboot command on Sparc */
+ KERN_CTLALTDEL=22, /* int: allow ctl-alt-del to reboot */
+--- a/include/linux/sysfs.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/sysfs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -17,6 +17,8 @@
+ #include <linux/list.h>
+ #include <asm/atomic.h>
+
++#define SYSFS_SUPER_MAGIC 0x62656572
++
+ struct kobject;
+ struct module;
+
+--- a/include/linux/time.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/time.h 2008-07-16 22:41:36.000000000 -0400
+@@ -179,6 +179,9 @@ static __always_inline void timespec_add
+ a->tv_sec += __iter_div_u64_rem(a->tv_nsec + ns, NSEC_PER_SEC, &ns);
+ a->tv_nsec = ns;
+ }
++
++#include <linux/vs_time.h>
++
+ #endif /* __KERNEL__ */
+
+ #define NFDBITS __NFDBITS
+--- a/include/linux/types.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/linux/types.h 2008-07-16 22:41:36.000000000 -0400
+@@ -36,6 +36,9 @@ typedef __kernel_uid32_t uid_t;
+ typedef __kernel_gid32_t gid_t;
+ typedef __kernel_uid16_t uid16_t;
+ typedef __kernel_gid16_t gid16_t;
++typedef unsigned int xid_t;
++typedef unsigned int nid_t;
++typedef unsigned int tag_t;
+
+ typedef unsigned long uintptr_t;
+
+--- a/include/linux/vroot.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vroot.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,51 @@
++
++/*
++ * include/linux/vroot.h
++ *
++ * written by Herbert Pötzl, 9/11/2002
++ * ported to 2.6 by Herbert Pötzl, 30/12/2004
++ *
++ * Copyright (C) 2002-2007 by Herbert Pötzl.
++ * Redistribution of this file is permitted under the
++ * GNU General Public License.
++ */
++
++#ifndef _LINUX_VROOT_H
++#define _LINUX_VROOT_H
++
++
++#ifdef __KERNEL__
++
++/* Possible states of device */
++enum {
++ Vr_unbound,
++ Vr_bound,
++};
++
++struct vroot_device {
++ int vr_number;
++ int vr_refcnt;
++
++ struct semaphore vr_ctl_mutex;
++ struct block_device *vr_device;
++ int vr_state;
++};
++
++
++typedef struct block_device *(vroot_grb_func)(struct block_device *);
++
++extern int register_vroot_grb(vroot_grb_func *);
++extern int unregister_vroot_grb(vroot_grb_func *);
++
++#endif /* __KERNEL__ */
++
++#define MAX_VROOT_DEFAULT 8
++
++/*
++ * IOCTL commands --- we will commandeer 0x56 ('V')
++ */
++
++#define VROOT_SET_DEV 0x5600
++#define VROOT_CLR_DEV 0x5601
++
++#endif /* _LINUX_VROOT_H */
+--- a/include/linux/vs_base.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_base.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,10 @@
++#ifndef _VS_BASE_H
++#define _VS_BASE_H
++
++#include "vserver/base.h"
++#include "vserver/check.h"
++#include "vserver/debug.h"
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_context.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_context.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,227 @@
++#ifndef _VS_CONTEXT_H
++#define _VS_CONTEXT_H
++
++#include "vserver/base.h"
++#include "vserver/check.h"
++#include "vserver/context.h"
++#include "vserver/history.h"
++#include "vserver/debug.h"
++
++#include <linux/sched.h>
++
++
++#define get_vx_info(i) __get_vx_info(i, __FILE__, __LINE__, __HERE__)
++
++static inline struct vx_info *__get_vx_info(struct vx_info *vxi,
++ const char *_file, int _line, void *_here)
++{
++ if (!vxi)
++ return NULL;
++
++ vxlprintk(VXD_CBIT(xid, 2), "get_vx_info(%p[#%d.%d])",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ _file, _line);
++ __vxh_get_vx_info(vxi, _here);
++
++ atomic_inc(&vxi->vx_usecnt);
++ return vxi;
++}
++
++
++extern void free_vx_info(struct vx_info *);
++
++#define put_vx_info(i) __put_vx_info(i, __FILE__, __LINE__, __HERE__)
++
++static inline void __put_vx_info(struct vx_info *vxi,
++ const char *_file, int _line, void *_here)
++{
++ if (!vxi)
++ return;
++
++ vxlprintk(VXD_CBIT(xid, 2), "put_vx_info(%p[#%d.%d])",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ _file, _line);
++ __vxh_put_vx_info(vxi, _here);
++
++ if (atomic_dec_and_test(&vxi->vx_usecnt))
++ free_vx_info(vxi);
++}
++
++
++#define init_vx_info(p, i) \
++ __init_vx_info(p, i, __FILE__, __LINE__, __HERE__)
++
++static inline void __init_vx_info(struct vx_info **vxp, struct vx_info *vxi,
++ const char *_file, int _line, void *_here)
++{
++ if (vxi) {
++ vxlprintk(VXD_CBIT(xid, 3),
++ "init_vx_info(%p[#%d.%d])",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ _file, _line);
++ __vxh_init_vx_info(vxi, vxp, _here);
++
++ atomic_inc(&vxi->vx_usecnt);
++ }
++ *vxp = vxi;
++}
++
++
++#define set_vx_info(p, i) \
++ __set_vx_info(p, i, __FILE__, __LINE__, __HERE__)
++
++static inline void __set_vx_info(struct vx_info **vxp, struct vx_info *vxi,
++ const char *_file, int _line, void *_here)
++{
++ struct vx_info *vxo;
++
++ if (!vxi)
++ return;
++
++ vxlprintk(VXD_CBIT(xid, 3), "set_vx_info(%p[#%d.%d])",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ _file, _line);
++ __vxh_set_vx_info(vxi, vxp, _here);
++
++ atomic_inc(&vxi->vx_usecnt);
++ vxo = xchg(vxp, vxi);
++ BUG_ON(vxo);
++}
++
++
++#define clr_vx_info(p) __clr_vx_info(p, __FILE__, __LINE__, __HERE__)
++
++static inline void __clr_vx_info(struct vx_info **vxp,
++ const char *_file, int _line, void *_here)
++{
++ struct vx_info *vxo;
++
++ vxo = xchg(vxp, NULL);
++ if (!vxo)
++ return;
++
++ vxlprintk(VXD_CBIT(xid, 3), "clr_vx_info(%p[#%d.%d])",
++ vxo, vxo ? vxo->vx_id : 0,
++ vxo ? atomic_read(&vxo->vx_usecnt) : 0,
++ _file, _line);
++ __vxh_clr_vx_info(vxo, vxp, _here);
++
++ if (atomic_dec_and_test(&vxo->vx_usecnt))
++ free_vx_info(vxo);
++}
++
++
++#define claim_vx_info(v, p) \
++ __claim_vx_info(v, p, __FILE__, __LINE__, __HERE__)
++
++static inline void __claim_vx_info(struct vx_info *vxi,
++ struct task_struct *task,
++ const char *_file, int _line, void *_here)
++{
++ vxlprintk(VXD_CBIT(xid, 3), "claim_vx_info(%p[#%d.%d.%d]) %p",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ vxi ? atomic_read(&vxi->vx_tasks) : 0,
++ task, _file, _line);
++ __vxh_claim_vx_info(vxi, task, _here);
++
++ atomic_inc(&vxi->vx_tasks);
++}
++
++
++extern void unhash_vx_info(struct vx_info *);
++
++#define release_vx_info(v, p) \
++ __release_vx_info(v, p, __FILE__, __LINE__, __HERE__)
++
++static inline void __release_vx_info(struct vx_info *vxi,
++ struct task_struct *task,
++ const char *_file, int _line, void *_here)
++{
++ vxlprintk(VXD_CBIT(xid, 3), "release_vx_info(%p[#%d.%d.%d]) %p",
++ vxi, vxi ? vxi->vx_id : 0,
++ vxi ? atomic_read(&vxi->vx_usecnt) : 0,
++ vxi ? atomic_read(&vxi->vx_tasks) : 0,
++ task, _file, _line);
++ __vxh_release_vx_info(vxi, task, _here);
++
++ might_sleep();
++
++ if (atomic_dec_and_test(&vxi->vx_tasks))
++ unhash_vx_info(vxi);
++}
++
++
++#define task_get_vx_info(p) \
++ __task_get_vx_info(p, __FILE__, __LINE__, __HERE__)
++
++static inline struct vx_info *__task_get_vx_info(struct task_struct *p,
++ const char *_file, int _line, void *_here)
++{
++ struct vx_info *vxi;
++
++ task_lock(p);
++ vxlprintk(VXD_CBIT(xid, 5), "task_get_vx_info(%p)",
++ p, _file, _line);
++ vxi = __get_vx_info(p->vx_info, _file, _line, _here);
++ task_unlock(p);
++ return vxi;
++}
++
++
++static inline void __wakeup_vx_info(struct vx_info *vxi)
++{
++ if (waitqueue_active(&vxi->vx_wait))
++ wake_up_interruptible(&vxi->vx_wait);
++}
++
++
++#define enter_vx_info(v, s) __enter_vx_info(v, s, __FILE__, __LINE__)
++
++static inline void __enter_vx_info(struct vx_info *vxi,
++ struct vx_info_save *vxis, const char *_file, int _line)
++{
++ vxlprintk(VXD_CBIT(xid, 5), "enter_vx_info(%p[#%d],%p) %p[#%d,%p]",
++ vxi, vxi ? vxi->vx_id : 0, vxis, current,
++ current->xid, current->vx_info, _file, _line);
++ vxis->vxi = xchg(¤t->vx_info, vxi);
++ vxis->xid = current->xid;
++ current->xid = vxi ? vxi->vx_id : 0;
++}
++
++#define leave_vx_info(s) __leave_vx_info(s, __FILE__, __LINE__)
++
++static inline void __leave_vx_info(struct vx_info_save *vxis,
++ const char *_file, int _line)
++{
++ vxlprintk(VXD_CBIT(xid, 5), "leave_vx_info(%p[#%d,%p]) %p[#%d,%p]",
++ vxis, vxis->xid, vxis->vxi, current,
++ current->xid, current->vx_info, _file, _line);
++ (void)xchg(¤t->vx_info, vxis->vxi);
++ current->xid = vxis->xid;
++}
++
++
++static inline void __enter_vx_admin(struct vx_info_save *vxis)
++{
++ vxis->vxi = xchg(¤t->vx_info, NULL);
++ vxis->xid = xchg(¤t->xid, (xid_t)0);
++}
++
++static inline void __leave_vx_admin(struct vx_info_save *vxis)
++{
++ (void)xchg(¤t->xid, vxis->xid);
++ (void)xchg(¤t->vx_info, vxis->vxi);
++}
++
++extern void exit_vx_info(struct task_struct *, int);
++extern void exit_vx_info_early(struct task_struct *, int);
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_cowbl.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_cowbl.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,47 @@
++#ifndef _VS_COWBL_H
++#define _VS_COWBL_H
++
++#include <linux/fs.h>
++#include <linux/dcache.h>
++#include <linux/namei.h>
++
++extern struct dentry *cow_break_link(const char *pathname);
++
++static inline int cow_check_and_break(struct nameidata *nd)
++{
++ struct inode *inode = nd->path.dentry->d_inode;
++ int error = 0;
++
++ /* do we need this check? */
++ if (IS_RDONLY(inode))
++ return -EROFS;
++
++ if (IS_COW(inode)) {
++ if (IS_COW_LINK(inode)) {
++ struct dentry *new_dentry, *old_dentry = nd->path.dentry;
++ char *path, *buf;
++
++ buf = kmalloc(PATH_MAX, GFP_KERNEL);
++ if (!buf) {
++ return -ENOMEM;
++ }
++ path = d_path(&nd->path, buf, PATH_MAX);
++ new_dentry = cow_break_link(path);
++ kfree(buf);
++ if (!IS_ERR(new_dentry)) {
++ nd->path.dentry = new_dentry;
++ dput(old_dentry);
++ } else
++ error = PTR_ERR(new_dentry);
++ } else {
++ inode->i_flags &= ~(S_IXUNLINK | S_IMMUTABLE);
++ inode->i_ctime = CURRENT_TIME;
++ mark_inode_dirty(inode);
++ }
++ }
++ return error;
++}
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_cvirt.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_cvirt.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,50 @@
++#ifndef _VS_CVIRT_H
++#define _VS_CVIRT_H
++
++#include "vserver/cvirt.h"
++#include "vserver/context.h"
++#include "vserver/base.h"
++#include "vserver/check.h"
++#include "vserver/debug.h"
++
++
++static inline void vx_activate_task(struct task_struct *p)
++{
++ struct vx_info *vxi;
++
++ if ((vxi = p->vx_info)) {
++ vx_update_load(vxi);
++ atomic_inc(&vxi->cvirt.nr_running);
++ }
++}
++
++static inline void vx_deactivate_task(struct task_struct *p)
++{
++ struct vx_info *vxi;
++
++ if ((vxi = p->vx_info)) {
++ vx_update_load(vxi);
++ atomic_dec(&vxi->cvirt.nr_running);
++ }
++}
++
++static inline void vx_uninterruptible_inc(struct task_struct *p)
++{
++ struct vx_info *vxi;
++
++ if ((vxi = p->vx_info))
++ atomic_inc(&vxi->cvirt.nr_uninterruptible);
++}
++
++static inline void vx_uninterruptible_dec(struct task_struct *p)
++{
++ struct vx_info *vxi;
++
++ if ((vxi = p->vx_info))
++ atomic_dec(&vxi->cvirt.nr_uninterruptible);
++}
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_device.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_device.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,45 @@
++#ifndef _VS_DEVICE_H
++#define _VS_DEVICE_H
++
++#include "vserver/base.h"
++#include "vserver/device.h"
++#include "vserver/debug.h"
++
++
++#ifdef CONFIG_VSERVER_DEVICE
++
++int vs_map_device(struct vx_info *, dev_t, dev_t *, umode_t);
++
++#define vs_device_perm(v, d, m, p) \
++ ((vs_map_device(current_vx_info(), d, NULL, m) & (p)) == (p))
++
++#else
++
++static inline
++int vs_map_device(struct vx_info *vxi,
++ dev_t device, dev_t *target, umode_t mode)
++{
++ if (target)
++ *target = device;
++ return ~0;
++}
++
++#define vs_device_perm(v, d, m, p) ((p) == (p))
++
++#endif
++
++
++#define vs_map_chrdev(d, t, p) \
++ ((vs_map_device(current_vx_info(), d, t, S_IFCHR) & (p)) == (p))
++#define vs_map_blkdev(d, t, p) \
++ ((vs_map_device(current_vx_info(), d, t, S_IFBLK) & (p)) == (p))
++
++#define vs_chrdev_perm(d, p) \
++ vs_device_perm(current_vx_info(), d, S_IFCHR, p)
++#define vs_blkdev_perm(d, p) \
++ vs_device_perm(current_vx_info(), d, S_IFBLK, p)
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_dlimit.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_dlimit.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,211 @@
++#ifndef _VS_DLIMIT_H
++#define _VS_DLIMIT_H
++
++#include <linux/fs.h>
++
++#include "vserver/dlimit.h"
++#include "vserver/base.h"
++#include "vserver/debug.h"
++
++
++#define get_dl_info(i) __get_dl_info(i, __FILE__, __LINE__)
++
++static inline struct dl_info *__get_dl_info(struct dl_info *dli,
++ const char *_file, int _line)
++{
++ if (!dli)
++ return NULL;
++ vxlprintk(VXD_CBIT(dlim, 4), "get_dl_info(%p[#%d.%d])",
++ dli, dli ? dli->dl_tag : 0,
++ dli ? atomic_read(&dli->dl_usecnt) : 0,
++ _file, _line);
++ atomic_inc(&dli->dl_usecnt);
++ return dli;
++}
++
++
++#define free_dl_info(i) \
++ call_rcu(&(i)->dl_rcu, rcu_free_dl_info)
++
++#define put_dl_info(i) __put_dl_info(i, __FILE__, __LINE__)
++
++static inline void __put_dl_info(struct dl_info *dli,
++ const char *_file, int _line)
++{
++ if (!dli)
++ return;
++ vxlprintk(VXD_CBIT(dlim, 4), "put_dl_info(%p[#%d.%d])",
++ dli, dli ? dli->dl_tag : 0,
++ dli ? atomic_read(&dli->dl_usecnt) : 0,
++ _file, _line);
++ if (atomic_dec_and_test(&dli->dl_usecnt))
++ free_dl_info(dli);
++}
++
++
++#define __dlimit_char(d) ((d) ? '*' : ' ')
++
++static inline int __dl_alloc_space(struct super_block *sb,
++ tag_t tag, dlsize_t nr, const char *file, int line)
++{
++ struct dl_info *dli = NULL;
++ int ret = 0;
++
++ if (nr == 0)
++ goto out;
++ dli = locate_dl_info(sb, tag);
++ if (!dli)
++ goto out;
++
++ spin_lock(&dli->dl_lock);
++ ret = (dli->dl_space_used + nr > dli->dl_space_total);
++ if (!ret)
++ dli->dl_space_used += nr;
++ spin_unlock(&dli->dl_lock);
++ put_dl_info(dli);
++out:
++ vxlprintk(VXD_CBIT(dlim, 1),
++ "ALLOC (%p,#%d)%c %lld bytes (%d)",
++ sb, tag, __dlimit_char(dli), (long long)nr,
++ ret, file, line);
++ return ret;
++}
++
++static inline void __dl_free_space(struct super_block *sb,
++ tag_t tag, dlsize_t nr, const char *_file, int _line)
++{
++ struct dl_info *dli = NULL;
++
++ if (nr == 0)
++ goto out;
++ dli = locate_dl_info(sb, tag);
++ if (!dli)
++ goto out;
++
++ spin_lock(&dli->dl_lock);
++ if (dli->dl_space_used > nr)
++ dli->dl_space_used -= nr;
++ else
++ dli->dl_space_used = 0;
++ spin_unlock(&dli->dl_lock);
++ put_dl_info(dli);
++out:
++ vxlprintk(VXD_CBIT(dlim, 1),
++ "FREE (%p,#%d)%c %lld bytes",
++ sb, tag, __dlimit_char(dli), (long long)nr,
++ _file, _line);
++}
++
++static inline int __dl_alloc_inode(struct super_block *sb,
++ tag_t tag, const char *_file, int _line)
++{
++ struct dl_info *dli;
++ int ret = 0;
++
++ dli = locate_dl_info(sb, tag);
++ if (!dli)
++ goto out;
++
++ spin_lock(&dli->dl_lock);
++ ret = (dli->dl_inodes_used >= dli->dl_inodes_total);
++ if (!ret)
++ dli->dl_inodes_used++;
++ spin_unlock(&dli->dl_lock);
++ put_dl_info(dli);
++out:
++ vxlprintk(VXD_CBIT(dlim, 0),
++ "ALLOC (%p,#%d)%c inode (%d)",
++ sb, tag, __dlimit_char(dli), ret, _file, _line);
++ return ret;
++}
++
++static inline void __dl_free_inode(struct super_block *sb,
++ tag_t tag, const char *_file, int _line)
++{
++ struct dl_info *dli;
++
++ dli = locate_dl_info(sb, tag);
++ if (!dli)
++ goto out;
++
++ spin_lock(&dli->dl_lock);
++ if (dli->dl_inodes_used > 1)
++ dli->dl_inodes_used--;
++ else
++ dli->dl_inodes_used = 0;
++ spin_unlock(&dli->dl_lock);
++ put_dl_info(dli);
++out:
++ vxlprintk(VXD_CBIT(dlim, 0),
++ "FREE (%p,#%d)%c inode",
++ sb, tag, __dlimit_char(dli), _file, _line);
++}
++
++static inline void __dl_adjust_block(struct super_block *sb, tag_t tag,
++ unsigned long long *free_blocks, unsigned long long *root_blocks,
++ const char *_file, int _line)
++{
++ struct dl_info *dli;
++ uint64_t broot, bfree;
++
++ dli = locate_dl_info(sb, tag);
++ if (!dli)
++ return;
++
++ spin_lock(&dli->dl_lock);
++ broot = (dli->dl_space_total -
++ (dli->dl_space_total >> 10) * dli->dl_nrlmult)
++ >> sb->s_blocksize_bits;
++ bfree = (dli->dl_space_total - dli->dl_space_used)
++ >> sb->s_blocksize_bits;
++ spin_unlock(&dli->dl_lock);
++
++ vxlprintk(VXD_CBIT(dlim, 2),
++ "ADJUST: %lld,%lld on %lld,%lld [mult=%d]",
++ (long long)bfree, (long long)broot,
++ *free_blocks, *root_blocks, dli->dl_nrlmult,
++ _file, _line);
++ if (free_blocks) {
++ if (*free_blocks > bfree)
++ *free_blocks = bfree;
++ }
++ if (root_blocks) {
++ if (*root_blocks > broot)
++ *root_blocks = broot;
++ }
++ put_dl_info(dli);
++}
++
++#define DLIMIT_ALLOC_SPACE(in, bytes) \
++ __dl_alloc_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \
++ __FILE__, __LINE__ )
++
++#define DLIMIT_FREE_SPACE(in, bytes) \
++ __dl_free_space((in)->i_sb, (in)->i_tag, (dlsize_t)(bytes), \
++ __FILE__, __LINE__ )
++
++#define DLIMIT_ALLOC_BLOCK(in, nr) \
++ __dl_alloc_space((in)->i_sb, (in)->i_tag, \
++ ((dlsize_t)(nr)) << (in)->i_sb->s_blocksize_bits, \
++ __FILE__, __LINE__ )
++
++#define DLIMIT_FREE_BLOCK(in, nr) \
++ __dl_free_space((in)->i_sb, (in)->i_tag, \
++ ((dlsize_t)(nr)) << (in)->i_sb->s_blocksize_bits, \
++ __FILE__, __LINE__ )
++
++
++#define DLIMIT_ALLOC_INODE(in) \
++ __dl_alloc_inode((in)->i_sb, (in)->i_tag, __FILE__, __LINE__ )
++
++#define DLIMIT_FREE_INODE(in) \
++ __dl_free_inode((in)->i_sb, (in)->i_tag, __FILE__, __LINE__ )
++
++
++#define DLIMIT_ADJUST_BLOCK(sb, tag, fb, rb) \
++ __dl_adjust_block(sb, tag, fb, rb, __FILE__, __LINE__ )
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vserver/base.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/base.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,157 @@
++#ifndef _VX_BASE_H
++#define _VX_BASE_H
++
++
++/* context state changes */
++
++enum {
++ VSC_STARTUP = 1,
++ VSC_SHUTDOWN,
++
++ VSC_NETUP,
++ VSC_NETDOWN,
++};
++
++
++
++#define vx_task_xid(t) ((t)->xid)
++
++#define vx_current_xid() vx_task_xid(current)
++
++#define current_vx_info() (current->vx_info)
++
++
++#define nx_task_nid(t) ((t)->nid)
++
++#define nx_current_nid() nx_task_nid(current)
++
++#define current_nx_info() (current->nx_info)
++
++
++/* generic flag merging */
++
++#define vs_check_flags(v, m, f) (((v) & (m)) ^ (f))
++
++#define vs_mask_flags(v, f, m) (((v) & ~(m)) | ((f) & (m)))
++
++#define vs_mask_mask(v, f, m) (((v) & ~(m)) | ((v) & (f) & (m)))
++
++#define vs_check_bit(v, n) ((v) & (1LL << (n)))
++
++
++/* context flags */
++
++#define __vx_flags(v) ((v) ? (v)->vx_flags : 0)
++
++#define vx_current_flags() __vx_flags(current->vx_info)
++
++#define vx_info_flags(v, m, f) \
++ vs_check_flags(__vx_flags(v), m, f)
++
++#define task_vx_flags(t, m, f) \
++ ((t) && vx_info_flags((t)->vx_info, m, f))
++
++#define vx_flags(m, f) vx_info_flags(current->vx_info, m, f)
++
++
++/* context caps */
++
++#define __vx_ccaps(v) ((v) ? (v)->vx_ccaps : 0)
++
++#define vx_current_ccaps() __vx_ccaps(current->vx_info)
++
++#define vx_info_ccaps(v, c) (__vx_ccaps(v) & (c))
++
++#define vx_ccaps(c) vx_info_ccaps(current->vx_info, (c))
++
++
++
++/* network flags */
++
++#define __nx_flags(n) ((n) ? (n)->nx_flags : 0)
++
++#define nx_current_flags() __nx_flags(current->nx_info)
++
++#define nx_info_flags(n, m, f) \
++ vs_check_flags(__nx_flags(n), m, f)
++
++#define task_nx_flags(t, m, f) \
++ ((t) && nx_info_flags((t)->nx_info, m, f))
++
++#define nx_flags(m, f) nx_info_flags(current->nx_info, m, f)
++
++
++/* network caps */
++
++#define __nx_ncaps(n) ((n) ? (n)->nx_ncaps : 0)
++
++#define nx_current_ncaps() __nx_ncaps(current->nx_info)
++
++#define nx_info_ncaps(n, c) (__nx_ncaps(n) & (c))
++
++#define nx_ncaps(c) nx_info_ncaps(current->nx_info, c)
++
++
++/* context mask capabilities */
++
++#define __vx_mcaps(v) ((v) ? (v)->vx_ccaps >> 32UL : ~0 )
++
++#define vx_info_mcaps(v, c) (__vx_mcaps(v) & (c))
++
++#define vx_mcaps(c) vx_info_mcaps(current->vx_info, c)
++
++
++/* context bcap mask */
++
++#define __vx_bcaps(v) ((v)->vx_bcaps)
++
++#define vx_current_bcaps() __vx_bcaps(current->vx_info)
++
++
++/* mask given bcaps */
++
++#define vx_info_mbcaps(v, c) ((v) ? cap_intersect(__vx_bcaps(v), c) : c)
++
++#define vx_mbcaps(c) vx_info_mbcaps(current->vx_info, c)
++
++
++/* masked cap_bset */
++
++#define vx_info_cap_bset(v) vx_info_mbcaps(v, current->cap_bset)
++
++#define vx_current_cap_bset() vx_info_cap_bset(current->vx_info)
++
++#if 0
++#define vx_info_mbcap(v, b) \
++ (!vx_info_flags(v, VXF_STATE_SETUP, 0) ? \
++ vx_info_bcaps(v, b) : (b))
++
++#define task_vx_mbcap(t, b) \
++ vx_info_mbcap((t)->vx_info, (t)->b)
++
++#define vx_mbcap(b) task_vx_mbcap(current, b)
++#endif
++
++#define vx_cap_raised(v, c, f) cap_raised(vx_info_mbcaps(v, c), f)
++
++#define vx_capable(b, c) (capable(b) || \
++ (cap_raised(current->cap_effective, b) && vx_ccaps(c)))
++
++#define nx_capable(b, c) (capable(b) || \
++ (cap_raised(current->cap_effective, b) && nx_ncaps(c)))
++
++#define vx_current_initpid(n) \
++ (current->vx_info && \
++ (current->vx_info->vx_initpid == (n)))
++
++
++#define __vx_state(v) ((v) ? ((v)->vx_state) : 0)
++
++#define vx_info_state(v, m) (__vx_state(v) & (m))
++
++
++#define __nx_state(n) ((n) ? ((n)->nx_state) : 0)
++
++#define nx_info_state(n, m) (__nx_state(n) & (m))
++
++#endif
+--- a/include/linux/vserver/cacct_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cacct_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,23 @@
++#ifndef _VX_CACCT_CMD_H
++#define _VX_CACCT_CMD_H
++
++
++/* virtual host info name commands */
++
++#define VCMD_sock_stat VC_CMD(VSTAT, 5, 0)
++
++struct vcmd_sock_stat_v0 {
++ uint32_t field;
++ uint32_t count[3];
++ uint64_t total[3];
++};
++
++
++#ifdef __KERNEL__
++
++#include <linux/compiler.h>
++
++extern int vc_sock_stat(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CACCT_CMD_H */
+--- a/include/linux/vserver/cacct_def.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cacct_def.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,43 @@
++#ifndef _VX_CACCT_DEF_H
++#define _VX_CACCT_DEF_H
++
++#include <asm/atomic.h>
++#include <linux/vserver/cacct.h>
++
++
++struct _vx_sock_acc {
++ atomic_long_t count;
++ atomic_long_t total;
++};
++
++/* context sub struct */
++
++struct _vx_cacct {
++ struct _vx_sock_acc sock[VXA_SOCK_SIZE][3];
++ atomic_t slab[8];
++ atomic_t page[6][8];
++};
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++static inline void __dump_vx_cacct(struct _vx_cacct *cacct)
++{
++ int i, j;
++
++ printk("\t_vx_cacct:");
++ for (i = 0; i < 6; i++) {
++ struct _vx_sock_acc *ptr = cacct->sock[i];
++
++ printk("\t [%d] =", i);
++ for (j = 0; j < 3; j++) {
++ printk(" [%d] = %8lu, %8lu", j,
++ atomic_long_read(&ptr[j].count),
++ atomic_long_read(&ptr[j].total));
++ }
++ printk("\n");
++ }
++}
++
++#endif
++
++#endif /* _VX_CACCT_DEF_H */
+--- a/include/linux/vserver/cacct.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cacct.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,15 @@
++#ifndef _VX_CACCT_H
++#define _VX_CACCT_H
++
++
++enum sock_acc_field {
++ VXA_SOCK_UNSPEC = 0,
++ VXA_SOCK_UNIX,
++ VXA_SOCK_INET,
++ VXA_SOCK_INET6,
++ VXA_SOCK_PACKET,
++ VXA_SOCK_OTHER,
++ VXA_SOCK_SIZE /* array size */
++};
++
++#endif /* _VX_CACCT_H */
+--- a/include/linux/vserver/cacct_int.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cacct_int.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,21 @@
++#ifndef _VX_CACCT_INT_H
++#define _VX_CACCT_INT_H
++
++
++#ifdef __KERNEL__
++
++static inline
++unsigned long vx_sock_count(struct _vx_cacct *cacct, int type, int pos)
++{
++ return atomic_long_read(&cacct->sock[type][pos].count);
++}
++
++
++static inline
++unsigned long vx_sock_total(struct _vx_cacct *cacct, int type, int pos)
++{
++ return atomic_long_read(&cacct->sock[type][pos].total);
++}
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CACCT_INT_H */
+--- a/include/linux/vserver/check.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/check.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,89 @@
++#ifndef _VS_CHECK_H
++#define _VS_CHECK_H
++
++
++#define MAX_S_CONTEXT 65535 /* Arbitrary limit */
++
++#ifdef CONFIG_VSERVER_DYNAMIC_IDS
++#define MIN_D_CONTEXT 49152 /* dynamic contexts start here */
++#else
++#define MIN_D_CONTEXT 65536
++#endif
++
++/* check conditions */
++
++#define VS_ADMIN 0x0001
++#define VS_WATCH 0x0002
++#define VS_HIDE 0x0004
++#define VS_HOSTID 0x0008
++
++#define VS_IDENT 0x0010
++#define VS_EQUIV 0x0020
++#define VS_PARENT 0x0040
++#define VS_CHILD 0x0080
++
++#define VS_ARG_MASK 0x00F0
++
++#define VS_DYNAMIC 0x0100
++#define VS_STATIC 0x0200
++
++#define VS_ATR_MASK 0x0F00
++
++#ifdef CONFIG_VSERVER_PRIVACY
++#define VS_ADMIN_P (0)
++#define VS_WATCH_P (0)
++#else
++#define VS_ADMIN_P VS_ADMIN
++#define VS_WATCH_P VS_WATCH
++#endif
++
++#define VS_HARDIRQ 0x1000
++#define VS_SOFTIRQ 0x2000
++#define VS_IRQ 0x4000
++
++#define VS_IRQ_MASK 0xF000
++
++#include <linux/hardirq.h>
++
++/*
++ * check current context for ADMIN/WATCH and
++ * optionally against supplied argument
++ */
++static inline int __vs_check(int cid, int id, unsigned int mode)
++{
++ if (mode & VS_ARG_MASK) {
++ if ((mode & VS_IDENT) && (id == cid))
++ return 1;
++ }
++ if (mode & VS_ATR_MASK) {
++ if ((mode & VS_DYNAMIC) &&
++ (id >= MIN_D_CONTEXT) &&
++ (id <= MAX_S_CONTEXT))
++ return 1;
++ if ((mode & VS_STATIC) &&
++ (id > 1) && (id < MIN_D_CONTEXT))
++ return 1;
++ }
++ if (mode & VS_IRQ_MASK) {
++ if ((mode & VS_IRQ) && unlikely(in_interrupt()))
++ return 1;
++ if ((mode & VS_HARDIRQ) && unlikely(in_irq()))
++ return 1;
++ if ((mode & VS_SOFTIRQ) && unlikely(in_softirq()))
++ return 1;
++ }
++ return (((mode & VS_ADMIN) && (cid == 0)) ||
++ ((mode & VS_WATCH) && (cid == 1)) ||
++ ((mode & VS_HOSTID) && (id == 0)));
++}
++
++#define vx_check(c, m) __vs_check(vx_current_xid(), c, (m) | VS_IRQ)
++
++#define vx_weak_check(c, m) ((m) ? vx_check(c, m) : 1)
++
++
++#define nx_check(c, m) __vs_check(nx_current_nid(), c, m)
++
++#define nx_weak_check(c, m) ((m) ? nx_check(c, m) : 1)
++
++#endif
+--- a/include/linux/vserver/context_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/context_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,128 @@
++#ifndef _VX_CONTEXT_CMD_H
++#define _VX_CONTEXT_CMD_H
++
++
++/* vinfo commands */
++
++#define VCMD_task_xid VC_CMD(VINFO, 1, 0)
++
++#ifdef __KERNEL__
++extern int vc_task_xid(uint32_t);
++
++#endif /* __KERNEL__ */
++
++#define VCMD_vx_info VC_CMD(VINFO, 5, 0)
++
++struct vcmd_vx_info_v0 {
++ uint32_t xid;
++ uint32_t initpid;
++ /* more to come */
++};
++
++#ifdef __KERNEL__
++extern int vc_vx_info(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++#define VCMD_ctx_stat VC_CMD(VSTAT, 0, 0)
++
++struct vcmd_ctx_stat_v0 {
++ uint32_t usecnt;
++ uint32_t tasks;
++ /* more to come */
++};
++
++#ifdef __KERNEL__
++extern int vc_ctx_stat(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++/* context commands */
++
++#define VCMD_ctx_create_v0 VC_CMD(VPROC, 1, 0)
++#define VCMD_ctx_create VC_CMD(VPROC, 1, 1)
++
++struct vcmd_ctx_create {
++ uint64_t flagword;
++};
++
++#define VCMD_ctx_migrate_v0 VC_CMD(PROCMIG, 1, 0)
++#define VCMD_ctx_migrate VC_CMD(PROCMIG, 1, 1)
++
++struct vcmd_ctx_migrate {
++ uint64_t flagword;
++};
++
++#ifdef __KERNEL__
++extern int vc_ctx_create(uint32_t, void __user *);
++extern int vc_ctx_migrate(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* flag commands */
++
++#define VCMD_get_cflags VC_CMD(FLAGS, 1, 0)
++#define VCMD_set_cflags VC_CMD(FLAGS, 2, 0)
++
++struct vcmd_ctx_flags_v0 {
++ uint64_t flagword;
++ uint64_t mask;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_cflags(struct vx_info *, void __user *);
++extern int vc_set_cflags(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* context caps commands */
++
++#define VCMD_get_ccaps VC_CMD(FLAGS, 3, 1)
++#define VCMD_set_ccaps VC_CMD(FLAGS, 4, 1)
++
++struct vcmd_ctx_caps_v1 {
++ uint64_t ccaps;
++ uint64_t cmask;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_ccaps(struct vx_info *, void __user *);
++extern int vc_set_ccaps(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* bcaps commands */
++
++#define VCMD_get_bcaps VC_CMD(FLAGS, 9, 0)
++#define VCMD_set_bcaps VC_CMD(FLAGS, 10, 0)
++
++struct vcmd_bcaps {
++ uint64_t bcaps;
++ uint64_t bmask;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_bcaps(struct vx_info *, void __user *);
++extern int vc_set_bcaps(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* OOM badness */
++
++#define VCMD_get_badness VC_CMD(MEMCTRL, 5, 0)
++#define VCMD_set_badness VC_CMD(MEMCTRL, 6, 0)
++
++struct vcmd_badness_v0 {
++ int64_t bias;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_badness(struct vx_info *, void __user *);
++extern int vc_set_badness(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CONTEXT_CMD_H */
+--- a/include/linux/vserver/context.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/context.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,176 @@
++#ifndef _VX_CONTEXT_H
++#define _VX_CONTEXT_H
++
++#include <linux/types.h>
++#include <linux/capability.h>
++
++
++/* context flags */
++
++#define VXF_INFO_SCHED 0x00000002
++#define VXF_INFO_NPROC 0x00000004
++#define VXF_INFO_PRIVATE 0x00000008
++
++#define VXF_INFO_INIT 0x00000010
++#define VXF_INFO_HIDE 0x00000020
++#define VXF_INFO_ULIMIT 0x00000040
++#define VXF_INFO_NSPACE 0x00000080
++
++#define VXF_SCHED_HARD 0x00000100
++#define VXF_SCHED_PRIO 0x00000200
++#define VXF_SCHED_PAUSE 0x00000400
++
++#define VXF_VIRT_MEM 0x00010000
++#define VXF_VIRT_UPTIME 0x00020000
++#define VXF_VIRT_CPU 0x00040000
++#define VXF_VIRT_LOAD 0x00080000
++#define VXF_VIRT_TIME 0x00100000
++
++#define VXF_HIDE_MOUNT 0x01000000
++/* was VXF_HIDE_NETIF 0x02000000 */
++#define VXF_HIDE_VINFO 0x04000000
++
++#define VXF_STATE_SETUP (1ULL << 32)
++#define VXF_STATE_INIT (1ULL << 33)
++#define VXF_STATE_ADMIN (1ULL << 34)
++
++#define VXF_SC_HELPER (1ULL << 36)
++#define VXF_REBOOT_KILL (1ULL << 37)
++#define VXF_PERSISTENT (1ULL << 38)
++
++#define VXF_FORK_RSS (1ULL << 48)
++#define VXF_PROLIFIC (1ULL << 49)
++
++#define VXF_IGNEG_NICE (1ULL << 52)
++
++#define VXF_ONE_TIME (0x0007ULL << 32)
++
++#define VXF_INIT_SET (VXF_STATE_SETUP | VXF_STATE_INIT | VXF_STATE_ADMIN)
++
++
++/* context migration */
++
++#define VXM_SET_INIT 0x00000001
++#define VXM_SET_REAPER 0x00000002
++
++/* context caps */
++
++#define VXC_CAP_MASK 0x00000000
++
++#define VXC_SET_UTSNAME 0x00000001
++#define VXC_SET_RLIMIT 0x00000002
++
++/* was VXC_RAW_ICMP 0x00000100 */
++#define VXC_SYSLOG 0x00001000
++
++#define VXC_SECURE_MOUNT 0x00010000
++#define VXC_SECURE_REMOUNT 0x00020000
++#define VXC_BINARY_MOUNT 0x00040000
++
++#define VXC_QUOTA_CTL 0x00100000
++#define VXC_ADMIN_MAPPER 0x00200000
++#define VXC_ADMIN_CLOOP 0x00400000
++
++#define VXC_KTHREAD 0x01000000
++
++
++#ifdef __KERNEL__
++
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/rcupdate.h>
++
++#include "limit_def.h"
++#include "sched_def.h"
++#include "cvirt_def.h"
++#include "cacct_def.h"
++#include "device_def.h"
++
++struct _vx_info_pc {
++ struct _vx_sched_pc sched_pc;
++ struct _vx_cvirt_pc cvirt_pc;
++};
++
++struct vx_info {
++ struct hlist_node vx_hlist; /* linked list of contexts */
++ xid_t vx_id; /* context id */
++ atomic_t vx_usecnt; /* usage count */
++ atomic_t vx_tasks; /* tasks count */
++ struct vx_info *vx_parent; /* parent context */
++ int vx_state; /* context state */
++
++ unsigned long vx_nsmask; /* assignment mask */
++ struct nsproxy *vx_nsproxy; /* private namespace */
++ struct fs_struct *vx_fs; /* private namespace fs */
++
++ uint64_t vx_flags; /* context flags */
++ uint64_t vx_ccaps; /* context caps (vserver) */
++ kernel_cap_t vx_bcaps; /* bounding caps (system) */
++ // kernel_cap_t vx_cap_bset; /* the guest's bset */
++
++ struct task_struct *vx_reaper; /* guest reaper process */
++ pid_t vx_initpid; /* PID of guest init */
++ int64_t vx_badness_bias; /* OOM points bias */
++
++ struct _vx_limit limit; /* vserver limits */
++ struct _vx_sched sched; /* vserver scheduler */
++ struct _vx_cvirt cvirt; /* virtual/bias stuff */
++ struct _vx_cacct cacct; /* context accounting */
++
++ struct _vx_device dmap; /* default device map targets */
++
++#ifndef CONFIG_SMP
++ struct _vx_info_pc info_pc; /* per cpu data */
++#else
++ struct _vx_info_pc *ptr_pc; /* per cpu array */
++#endif
++
++ wait_queue_head_t vx_wait; /* context exit waitqueue */
++ int reboot_cmd; /* last sys_reboot() cmd */
++ int exit_code; /* last process exit code */
++
++ char vx_name[65]; /* vserver name */
++};
++
++#ifndef CONFIG_SMP
++#define vx_ptr_pc(vxi) (&(vxi)->info_pc)
++#define vx_per_cpu(vxi, v, id) vx_ptr_pc(vxi)->v
++#else
++#define vx_ptr_pc(vxi) ((vxi)->ptr_pc)
++#define vx_per_cpu(vxi, v, id) per_cpu_ptr(vx_ptr_pc(vxi), id)->v
++#endif
++
++#define vx_cpu(vxi, v) vx_per_cpu(vxi, v, smp_processor_id())
++
++
++struct vx_info_save {
++ struct vx_info *vxi;
++ xid_t xid;
++};
++
++
++/* status flags */
++
++#define VXS_HASHED 0x0001
++#define VXS_PAUSED 0x0010
++#define VXS_SHUTDOWN 0x0100
++#define VXS_HELPER 0x1000
++#define VXS_RELEASED 0x8000
++
++
++extern void claim_vx_info(struct vx_info *, struct task_struct *);
++extern void release_vx_info(struct vx_info *, struct task_struct *);
++
++extern struct vx_info *lookup_vx_info(int);
++extern struct vx_info *lookup_or_create_vx_info(int);
++
++extern int get_xid_list(int, unsigned int *, int);
++extern int xid_is_hashed(xid_t);
++
++extern int vx_migrate_task(struct task_struct *, struct vx_info *, int);
++
++extern long vs_state_change(struct vx_info *, unsigned int);
++
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CONTEXT_H */
+--- a/include/linux/vserver/cvirt_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cvirt_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,53 @@
++#ifndef _VX_CVIRT_CMD_H
++#define _VX_CVIRT_CMD_H
++
++
++/* virtual host info name commands */
++
++#define VCMD_set_vhi_name VC_CMD(VHOST, 1, 0)
++#define VCMD_get_vhi_name VC_CMD(VHOST, 2, 0)
++
++struct vcmd_vhi_name_v0 {
++ uint32_t field;
++ char name[65];
++};
++
++
++enum vhi_name_field {
++ VHIN_CONTEXT = 0,
++ VHIN_SYSNAME,
++ VHIN_NODENAME,
++ VHIN_RELEASE,
++ VHIN_VERSION,
++ VHIN_MACHINE,
++ VHIN_DOMAINNAME,
++};
++
++
++#ifdef __KERNEL__
++
++#include <linux/compiler.h>
++
++extern int vc_set_vhi_name(struct vx_info *, void __user *);
++extern int vc_get_vhi_name(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++#define VCMD_virt_stat VC_CMD(VSTAT, 3, 0)
++
++struct vcmd_virt_stat_v0 {
++ uint64_t offset;
++ uint64_t uptime;
++ uint32_t nr_threads;
++ uint32_t nr_running;
++ uint32_t nr_uninterruptible;
++ uint32_t nr_onhold;
++ uint32_t nr_forks;
++ uint32_t load[3];
++};
++
++#ifdef __KERNEL__
++extern int vc_virt_stat(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CVIRT_CMD_H */
+--- a/include/linux/vserver/cvirt_def.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cvirt_def.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,80 @@
++#ifndef _VX_CVIRT_DEF_H
++#define _VX_CVIRT_DEF_H
++
++#include <linux/jiffies.h>
++#include <linux/spinlock.h>
++#include <linux/wait.h>
++#include <linux/time.h>
++#include <asm/atomic.h>
++
++
++struct _vx_usage_stat {
++ uint64_t user;
++ uint64_t nice;
++ uint64_t system;
++ uint64_t softirq;
++ uint64_t irq;
++ uint64_t idle;
++ uint64_t iowait;
++};
++
++struct _vx_syslog {
++ wait_queue_head_t log_wait;
++ spinlock_t logbuf_lock; /* lock for the log buffer */
++
++ unsigned long log_start; /* next char to be read by syslog() */
++ unsigned long con_start; /* next char to be sent to consoles */
++ unsigned long log_end; /* most-recently-written-char + 1 */
++ unsigned long logged_chars; /* #chars since last read+clear operation */
++
++ char log_buf[1024];
++};
++
++
++/* context sub struct */
++
++struct _vx_cvirt {
++ atomic_t nr_threads; /* number of current threads */
++ atomic_t nr_running; /* number of running threads */
++ atomic_t nr_uninterruptible; /* number of uninterruptible threads */
++
++ atomic_t nr_onhold; /* processes on hold */
++ uint32_t onhold_last; /* jiffies when put on hold */
++
++ struct timeval bias_tv; /* time offset to the host */
++ struct timespec bias_idle;
++ struct timespec bias_uptime; /* context creation point */
++ uint64_t bias_clock; /* offset in clock_t */
++
++ spinlock_t load_lock; /* lock for the load averages */
++ atomic_t load_updates; /* nr of load updates done so far */
++ uint32_t load_last; /* last time load was calculated */
++ uint32_t load[3]; /* load averages 1,5,15 */
++
++ atomic_t total_forks; /* number of forks so far */
++
++ struct _vx_syslog syslog;
++};
++
++struct _vx_cvirt_pc {
++ struct _vx_usage_stat cpustat;
++};
++
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++static inline void __dump_vx_cvirt(struct _vx_cvirt *cvirt)
++{
++ printk("\t_vx_cvirt:\n");
++ printk("\t threads: %4d, %4d, %4d, %4d\n",
++ atomic_read(&cvirt->nr_threads),
++ atomic_read(&cvirt->nr_running),
++ atomic_read(&cvirt->nr_uninterruptible),
++ atomic_read(&cvirt->nr_onhold));
++ /* add rest here */
++ printk("\t total_forks = %d\n", atomic_read(&cvirt->total_forks));
++}
++
++#endif
++
++#endif /* _VX_CVIRT_DEF_H */
+--- a/include/linux/vserver/cvirt.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/cvirt.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,20 @@
++#ifndef _VX_CVIRT_H
++#define _VX_CVIRT_H
++
++
++#ifdef __KERNEL__
++
++struct timespec;
++
++void vx_vsi_uptime(struct timespec *, struct timespec *);
++
++
++struct vx_info;
++
++void vx_update_load(struct vx_info *);
++
++
++int vx_do_syslog(int, char __user *, int);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CVIRT_H */
+--- a/include/linux/vserver/debug_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/debug_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,58 @@
++#ifndef _VX_DEBUG_CMD_H
++#define _VX_DEBUG_CMD_H
++
++
++/* debug commands */
++
++#define VCMD_dump_history VC_CMD(DEBUG, 1, 0)
++
++#define VCMD_read_history VC_CMD(DEBUG, 5, 0)
++#define VCMD_read_monitor VC_CMD(DEBUG, 6, 0)
++
++struct vcmd_read_history_v0 {
++ uint32_t index;
++ uint32_t count;
++ char __user *data;
++};
++
++struct vcmd_read_monitor_v0 {
++ uint32_t index;
++ uint32_t count;
++ char __user *data;
++};
++
++
++#ifdef __KERNEL__
++
++#ifdef CONFIG_COMPAT
++
++#include <asm/compat.h>
++
++struct vcmd_read_history_v0_x32 {
++ uint32_t index;
++ uint32_t count;
++ compat_uptr_t data_ptr;
++};
++
++struct vcmd_read_monitor_v0_x32 {
++ uint32_t index;
++ uint32_t count;
++ compat_uptr_t data_ptr;
++};
++
++#endif /* CONFIG_COMPAT */
++
++extern int vc_dump_history(uint32_t);
++
++extern int vc_read_history(uint32_t, void __user *);
++extern int vc_read_monitor(uint32_t, void __user *);
++
++#ifdef CONFIG_COMPAT
++
++extern int vc_read_history_x32(uint32_t, void __user *);
++extern int vc_read_monitor_x32(uint32_t, void __user *);
++
++#endif /* CONFIG_COMPAT */
++
++#endif /* __KERNEL__ */
++#endif /* _VX_DEBUG_CMD_H */
+--- a/include/linux/vserver/debug.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/debug.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,130 @@
++#ifndef _VX_DEBUG_H
++#define _VX_DEBUG_H
++
++
++#define VXD_CBIT(n, m) (vx_debug_ ## n & (1 << (m)))
++#define VXD_CMIN(n, m) (vx_debug_ ## n > (m))
++#define VXD_MASK(n, m) (vx_debug_ ## n & (m))
++
++#define VXD_DEV(d) (d), (d)->bd_inode->i_ino, \
++ imajor((d)->bd_inode), iminor((d)->bd_inode)
++#define VXF_DEV "%p[%lu,%d:%d]"
++
++
++#define __FUNC__ __func__
++
++
++#define vxd_path(p) \
++ ({ static char _buffer[PATH_MAX]; \
++ d_path(p, _buffer, sizeof(_buffer)); })
++
++#define vxd_cond_path(n) \
++ ((n) ? vxd_path(&(n)->path) : "<null>" )
++
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++extern unsigned int vx_debug_switch;
++extern unsigned int vx_debug_xid;
++extern unsigned int vx_debug_nid;
++extern unsigned int vx_debug_tag;
++extern unsigned int vx_debug_net;
++extern unsigned int vx_debug_limit;
++extern unsigned int vx_debug_cres;
++extern unsigned int vx_debug_dlim;
++extern unsigned int vx_debug_quota;
++extern unsigned int vx_debug_cvirt;
++extern unsigned int vx_debug_space;
++extern unsigned int vx_debug_misc;
++
++
++#define VX_LOGLEVEL "vxD: "
++#define VX_PROC_FMT "%p: "
++#define VX_PROCESS current
++
++#define vxdprintk(c, f, x...) \
++ do { \
++ if (c) \
++ printk(VX_LOGLEVEL VX_PROC_FMT f "\n", \
++ VX_PROCESS , ##x); \
++ } while (0)
++
++#define vxlprintk(c, f, x...) \
++ do { \
++ if (c) \
++ printk(VX_LOGLEVEL f " @%s:%d\n", x); \
++ } while (0)
++
++#define vxfprintk(c, f, x...) \
++ do { \
++ if (c) \
++ printk(VX_LOGLEVEL f " %s@%s:%d\n", x); \
++ } while (0)
++
++
++struct vx_info;
++
++void dump_vx_info(struct vx_info *, int);
++void dump_vx_info_inactive(int);
++
++#else /* CONFIG_VSERVER_DEBUG */
++
++#define vx_debug_switch 0
++#define vx_debug_xid 0
++#define vx_debug_nid 0
++#define vx_debug_tag 0
++#define vx_debug_net 0
++#define vx_debug_limit 0
++#define vx_debug_cres 0
++#define vx_debug_dlim 0
++#define vx_debug_cvirt 0
++
++#define vxdprintk(x...) do { } while (0)
++#define vxlprintk(x...) do { } while (0)
++#define vxfprintk(x...) do { } while (0)
++
++#endif /* CONFIG_VSERVER_DEBUG */
++
++
++#ifdef CONFIG_VSERVER_WARN
++
++#define VX_WARNLEVEL KERN_WARNING "vxW: "
++#define VX_WARN_TASK "[»%s«,%u:#%u|%u|%u] "
++#define VX_WARN_XID "[xid #%u] "
++#define VX_WARN_NID "[nid #%u] "
++#define VX_WARN_TAG "[tag #%u] "
++
++#define vxwprintk(c, f, x...) \
++ do { \
++ if (c) \
++ printk(VX_WARNLEVEL f "\n", ##x); \
++ } while (0)
++
++#else /* CONFIG_VSERVER_WARN */
++
++#define vxwprintk(x...) do { } while (0)
++
++#endif /* CONFIG_VSERVER_WARN */
++
++#define vxwprintk_task(c, f, x...) \
++ vxwprintk(c, VX_WARN_TASK f, \
++ current->comm, current->pid, \
++ current->xid, current->nid, current->tag, ##x)
++#define vxwprintk_xid(c, f, x...) \
++ vxwprintk(c, VX_WARN_XID f, current->xid, x)
++#define vxwprintk_nid(c, f, x...) \
++ vxwprintk(c, VX_WARN_NID f, current->nid, x)
++#define vxwprintk_tag(c, f, x...) \
++ vxwprintk(c, VX_WARN_TAG f, current->tag, x)
++
++#ifdef CONFIG_VSERVER_DEBUG
++#define vxd_assert_lock(l) assert_spin_locked(l)
++#define vxd_assert(c, f, x...) vxlprintk(!(c), \
++ "assertion [" f "] failed.", ##x, __FILE__, __LINE__)
++#else
++#define vxd_assert_lock(l) do { } while (0)
++#define vxd_assert(c, f, x...) do { } while (0)
++#endif
++
++
++#endif /* _VX_DEBUG_H */
+--- a/include/linux/vserver/device_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/device_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,44 @@
++#ifndef _VX_DEVICE_CMD_H
++#define _VX_DEVICE_CMD_H
++
++
++/* device vserver commands */
++
++#define VCMD_set_mapping VC_CMD(DEVICE, 1, 0)
++#define VCMD_unset_mapping VC_CMD(DEVICE, 2, 0)
++
++struct vcmd_set_mapping_v0 {
++ const char __user *device;
++ const char __user *target;
++ uint32_t flags;
++};
++
++
++#ifdef __KERNEL__
++
++#ifdef CONFIG_COMPAT
++
++#include <asm/compat.h>
++
++struct vcmd_set_mapping_v0_x32 {
++ compat_uptr_t device_ptr;
++ compat_uptr_t target_ptr;
++ uint32_t flags;
++};
++
++#endif /* CONFIG_COMPAT */
++
++#include <linux/compiler.h>
++
++extern int vc_set_mapping(struct vx_info *, void __user *);
++extern int vc_unset_mapping(struct vx_info *, void __user *);
++
++#ifdef CONFIG_COMPAT
++
++extern int vc_set_mapping_x32(struct vx_info *, void __user *);
++extern int vc_unset_mapping_x32(struct vx_info *, void __user *);
++
++#endif /* CONFIG_COMPAT */
++
++#endif /* __KERNEL__ */
++#endif /* _VX_DEVICE_CMD_H */
+--- a/include/linux/vserver/device_def.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/device_def.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,17 @@
++#ifndef _VX_DEVICE_DEF_H
++#define _VX_DEVICE_DEF_H
++
++#include <linux/types.h>
++
++struct vx_dmap_target {
++ dev_t target;
++ uint32_t flags;
++};
++
++struct _vx_device {
++#ifdef CONFIG_VSERVER_DEVICE
++ struct vx_dmap_target targets[2];
++#endif
++};
++
++#endif /* _VX_DEVICE_DEF_H */
+--- a/include/linux/vserver/device.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/device.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,15 @@
++#ifndef _VX_DEVICE_H
++#define _VX_DEVICE_H
++
++
++#define DATTR_CREATE 0x00000001
++#define DATTR_OPEN 0x00000002
++
++#define DATTR_REMAP 0x00000010
++
++#define DATTR_MASK 0x00000013
++
++
++#else /* _VX_DEVICE_H */
++#warning duplicate inclusion
++#endif /* _VX_DEVICE_H */
+--- a/include/linux/vserver/dlimit_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/dlimit_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,74 @@
++#ifndef _VX_DLIMIT_CMD_H
++#define _VX_DLIMIT_CMD_H
++
++
++/* dlimit vserver commands */
++
++#define VCMD_add_dlimit VC_CMD(DLIMIT, 1, 0)
++#define VCMD_rem_dlimit VC_CMD(DLIMIT, 2, 0)
++
++#define VCMD_set_dlimit VC_CMD(DLIMIT, 5, 0)
++#define VCMD_get_dlimit VC_CMD(DLIMIT, 6, 0)
++
++struct vcmd_ctx_dlimit_base_v0 {
++ const char __user *name;
++ uint32_t flags;
++};
++
++struct vcmd_ctx_dlimit_v0 {
++ const char __user *name;
++ uint32_t space_used; /* used space in kbytes */
++ uint32_t space_total; /* maximum space in kbytes */
++ uint32_t inodes_used; /* used inodes */
++ uint32_t inodes_total; /* maximum inodes */
++ uint32_t reserved; /* reserved for root in % */
++ uint32_t flags;
++};
++
++#define CDLIM_UNSET ((uint32_t)0UL)
++#define CDLIM_INFINITY ((uint32_t)~0UL)
++#define CDLIM_KEEP ((uint32_t)~1UL)
++
++#ifdef __KERNEL__
++
++#ifdef CONFIG_COMPAT
++
++#include <asm/compat.h>
++
++struct vcmd_ctx_dlimit_base_v0_x32 {
++ compat_uptr_t name_ptr;
++ uint32_t flags;
++};
++
++struct vcmd_ctx_dlimit_v0_x32 {
++ compat_uptr_t name_ptr;
++ uint32_t space_used; /* used space in kbytes */
++ uint32_t space_total; /* maximum space in kbytes */
++ uint32_t inodes_used; /* used inodes */
++ uint32_t inodes_total; /* maximum inodes */
++ uint32_t reserved; /* reserved for root in % */
++ uint32_t flags;
++};
++
++#endif /* CONFIG_COMPAT */
++
++#include <linux/compiler.h>
++
++extern int vc_add_dlimit(uint32_t, void __user *);
++extern int vc_rem_dlimit(uint32_t, void __user *);
++
++extern int vc_set_dlimit(uint32_t, void __user *);
++extern int vc_get_dlimit(uint32_t, void __user *);
++
++#ifdef CONFIG_COMPAT
++
++extern int vc_add_dlimit_x32(uint32_t, void __user *);
++extern int vc_rem_dlimit_x32(uint32_t, void __user *);
++
++extern int vc_set_dlimit_x32(uint32_t, void __user *);
++extern int vc_get_dlimit_x32(uint32_t, void __user *);
++
++#endif /* CONFIG_COMPAT */
++
++#endif /* __KERNEL__ */
++#endif /* _VX_DLIMIT_CMD_H */
+--- a/include/linux/vserver/dlimit.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/dlimit.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,54 @@
++#ifndef _VX_DLIMIT_H
++#define _VX_DLIMIT_H
++
++#include "switch.h"
++
++
++#ifdef __KERNEL__
++
++/* keep in sync with CDLIM_INFINITY */
++
++#define DLIM_INFINITY (~0ULL)
++
++#include <linux/spinlock.h>
++#include <linux/rcupdate.h>
++
++struct super_block;
++
++struct dl_info {
++ struct hlist_node dl_hlist; /* linked list of contexts */
++ struct rcu_head dl_rcu; /* the rcu head */
++ tag_t dl_tag; /* context tag */
++ atomic_t dl_usecnt; /* usage count */
++ atomic_t dl_refcnt; /* reference count */
++
++ struct super_block *dl_sb; /* associated superblock */
++
++ spinlock_t dl_lock; /* protect the values */
++
++ unsigned long long dl_space_used; /* used space in bytes */
++ unsigned long long dl_space_total; /* maximum space in bytes */
++ unsigned long dl_inodes_used; /* used inodes */
++ unsigned long dl_inodes_total; /* maximum inodes */
++
++ unsigned int dl_nrlmult; /* non root limit mult */
++};
++
++struct rcu_head;
++
++extern void rcu_free_dl_info(struct rcu_head *);
++extern void unhash_dl_info(struct dl_info *);
++
++extern struct dl_info *locate_dl_info(struct super_block *, tag_t);
++
++
++struct kstatfs;
++
++extern void vx_vsi_statfs(struct super_block *, struct kstatfs *);
++
++typedef uint64_t dlsize_t;
++
++#endif /* __KERNEL__ */
++#else /* _VX_DLIMIT_H */
++#warning duplicate inclusion
++#endif /* _VX_DLIMIT_H */
+--- a/include/linux/vserver/global.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/global.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,20 @@
++#ifndef _VX_GLOBAL_H
++#define _VX_GLOBAL_H
++
++
++extern atomic_t vx_global_ctotal;
++extern atomic_t vx_global_cactive;
++
++extern atomic_t nx_global_ctotal;
++extern atomic_t nx_global_cactive;
++
++extern atomic_t vs_global_nsproxy;
++extern atomic_t vs_global_fs;
++extern atomic_t vs_global_mnt_ns;
++extern atomic_t vs_global_uts_ns;
++extern atomic_t vs_global_ipc_ns;
++extern atomic_t vs_global_user_ns;
++extern atomic_t vs_global_pid_ns;
++
++
++#endif /* _VX_GLOBAL_H */
+--- a/include/linux/vserver/history.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/history.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,197 @@
++#ifndef _VX_HISTORY_H
++#define _VX_HISTORY_H
++
++
++enum {
++ VXH_UNUSED = 0,
++ VXH_THROW_OOPS = 1,
++
++ VXH_GET_VX_INFO,
++ VXH_PUT_VX_INFO,
++ VXH_INIT_VX_INFO,
++ VXH_SET_VX_INFO,
++ VXH_CLR_VX_INFO,
++ VXH_CLAIM_VX_INFO,
++ VXH_RELEASE_VX_INFO,
++ VXH_ALLOC_VX_INFO,
++ VXH_DEALLOC_VX_INFO,
++ VXH_HASH_VX_INFO,
++ VXH_UNHASH_VX_INFO,
++ VXH_LOC_VX_INFO,
++ VXH_LOOKUP_VX_INFO,
++ VXH_CREATE_VX_INFO,
++};
++
++struct _vxhe_vxi {
++ struct vx_info *ptr;
++ unsigned xid;
++ unsigned usecnt;
++ unsigned tasks;
++};
++
++struct _vxhe_set_clr {
++ void *data;
++};
++
++struct _vxhe_loc_lookup {
++ unsigned arg;
++};
++
++struct _vx_hist_entry {
++ void *loc;
++ unsigned short seq;
++ unsigned short type;
++ struct _vxhe_vxi vxi;
++ union {
++ struct _vxhe_set_clr sc;
++ struct _vxhe_loc_lookup ll;
++ };
++};
++
++#ifdef CONFIG_VSERVER_HISTORY
++
++extern unsigned volatile int vxh_active;
++
++struct _vx_hist_entry *vxh_advance(void *loc);
++
++
++static inline
++void __vxh_copy_vxi(struct _vx_hist_entry *entry, struct vx_info *vxi)
++{
++ entry->vxi.ptr = vxi;
++ if (vxi) {
++ entry->vxi.usecnt = atomic_read(&vxi->vx_usecnt);
++ entry->vxi.tasks = atomic_read(&vxi->vx_tasks);
++ entry->vxi.xid = vxi->vx_id;
++ }
++}
++
++
++#define __HERE__ current_text_addr()
++
++#define __VXH_BODY(__type, __data, __here) \
++ struct _vx_hist_entry *entry; \
++ \
++ preempt_disable(); \
++ entry = vxh_advance(__here); \
++ __data; \
++ entry->type = __type; \
++ preempt_enable();
++
++
++ /* pass vxi only */
++
++#define __VXH_SMPL \
++ __vxh_copy_vxi(entry, vxi)
++
++static inline
++void __vxh_smpl(struct vx_info *vxi, int __type, void *__here)
++{
++ __VXH_BODY(__type, __VXH_SMPL, __here)
++}
++
++ /* pass vxi and data (void *) */
++
++#define __VXH_DATA \
++ __vxh_copy_vxi(entry, vxi); \
++ entry->sc.data = data
++
++static inline
++void __vxh_data(struct vx_info *vxi, void *data,
++ int __type, void *__here)
++{
++ __VXH_BODY(__type, __VXH_DATA, __here)
++}
++
++ /* pass vxi and arg (long) */
++
++#define __VXH_LONG \
++ __vxh_copy_vxi(entry, vxi); \
++ entry->ll.arg = arg
++
++static inline
++void __vxh_long(struct vx_info *vxi, long arg,
++ int __type, void *__here)
++{
++ __VXH_BODY(__type, __VXH_LONG, __here)
++}
++
++
++static inline
++void __vxh_throw_oops(void *__here)
++{
++ __VXH_BODY(VXH_THROW_OOPS, {}, __here);
++ /* prevent further acquisition */
++ vxh_active = 0;
++}
++
++
++#define vxh_throw_oops() __vxh_throw_oops(__HERE__);
++
++#define __vxh_get_vx_info(v, h) __vxh_smpl(v, VXH_GET_VX_INFO, h);
++#define __vxh_put_vx_info(v, h) __vxh_smpl(v, VXH_PUT_VX_INFO, h);
++
++#define __vxh_init_vx_info(v, d, h) \
++ __vxh_data(v, d, VXH_INIT_VX_INFO, h);
++#define __vxh_set_vx_info(v, d, h) \
++ __vxh_data(v, d, VXH_SET_VX_INFO, h);
++#define __vxh_clr_vx_info(v, d, h) \
++ __vxh_data(v, d, VXH_CLR_VX_INFO, h);
++
++#define __vxh_claim_vx_info(v, d, h) \
++ __vxh_data(v, d, VXH_CLAIM_VX_INFO, h);
++#define __vxh_release_vx_info(v, d, h) \
++ __vxh_data(v, d, VXH_RELEASE_VX_INFO, h);
++
++#define vxh_alloc_vx_info(v) \
++ __vxh_smpl(v, VXH_ALLOC_VX_INFO, __HERE__);
++#define vxh_dealloc_vx_info(v) \
++ __vxh_smpl(v, VXH_DEALLOC_VX_INFO, __HERE__);
++
++#define vxh_hash_vx_info(v) \
++ __vxh_smpl(v, VXH_HASH_VX_INFO, __HERE__);
++#define vxh_unhash_vx_info(v) \
++ __vxh_smpl(v, VXH_UNHASH_VX_INFO, __HERE__);
++
++#define vxh_loc_vx_info(v, l) \
++ __vxh_long(v, l, VXH_LOC_VX_INFO, __HERE__);
++#define vxh_lookup_vx_info(v, l) \
++ __vxh_long(v, l, VXH_LOOKUP_VX_INFO, __HERE__);
++#define vxh_create_vx_info(v, l) \
++ __vxh_long(v, l, VXH_CREATE_VX_INFO, __HERE__);
++
++extern void vxh_dump_history(void);
++
++
++#else /* CONFIG_VSERVER_HISTORY */
++
++#define __HERE__ 0
++
++#define vxh_throw_oops() do { } while (0)
++
++#define __vxh_get_vx_info(v, h) do { } while (0)
++#define __vxh_put_vx_info(v, h) do { } while (0)
++
++#define __vxh_init_vx_info(v, d, h) do { } while (0)
++#define __vxh_set_vx_info(v, d, h) do { } while (0)
++#define __vxh_clr_vx_info(v, d, h) do { } while (0)
++
++#define __vxh_claim_vx_info(v, d, h) do { } while (0)
++#define __vxh_release_vx_info(v, d, h) do { } while (0)
++
++#define vxh_alloc_vx_info(v) do { } while (0)
++#define vxh_dealloc_vx_info(v) do { } while (0)
++
++#define vxh_hash_vx_info(v) do { } while (0)
++#define vxh_unhash_vx_info(v) do { } while (0)
++
++#define vxh_loc_vx_info(v, l) do { } while (0)
++#define vxh_lookup_vx_info(v, l) do { } while (0)
++#define vxh_create_vx_info(v, l) do { } while (0)
++
++#define vxh_dump_history() do { } while (0)
++
++
++#endif /* CONFIG_VSERVER_HISTORY */
++
++#endif /* _VX_HISTORY_H */
+--- a/include/linux/vserver/inode_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/inode_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,59 @@
++#ifndef _VX_INODE_CMD_H
++#define _VX_INODE_CMD_H
++
++
++/* inode vserver commands */
++
++#define VCMD_get_iattr VC_CMD(INODE, 1, 1)
++#define VCMD_set_iattr VC_CMD(INODE, 2, 1)
++
++#define VCMD_fget_iattr VC_CMD(INODE, 3, 0)
++#define VCMD_fset_iattr VC_CMD(INODE, 4, 0)
++
++struct vcmd_ctx_iattr_v1 {
++ const char __user *name;
++ uint32_t tag;
++ uint32_t flags;
++ uint32_t mask;
++};
++
++struct vcmd_ctx_fiattr_v0 {
++ uint32_t tag;
++ uint32_t flags;
++ uint32_t mask;
++};
++
++
++#ifdef __KERNEL__
++
++
++#ifdef CONFIG_COMPAT
++
++#include <asm/compat.h>
++
++struct vcmd_ctx_iattr_v1_x32 {
++ compat_uptr_t name_ptr;
++ uint32_t tag;
++ uint32_t flags;
++ uint32_t mask;
++};
++
++#endif /* CONFIG_COMPAT */
++
++#include <linux/compiler.h>
++
++extern int vc_get_iattr(void __user *);
++extern int vc_set_iattr(void __user *);
++
++extern int vc_fget_iattr(uint32_t, void __user *);
++extern int vc_fset_iattr(uint32_t, void __user *);
++
++#ifdef CONFIG_COMPAT
++
++extern int vc_get_iattr_x32(void __user *);
++extern int vc_set_iattr_x32(void __user *);
++
++#endif /* CONFIG_COMPAT */
++
++#endif /* __KERNEL__ */
++#endif /* _VX_INODE_CMD_H */
+--- a/include/linux/vserver/inode.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/inode.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,38 @@
++#ifndef _VX_INODE_H
++#define _VX_INODE_H
++
++
++#define IATTR_TAG 0x01000000
++
++#define IATTR_ADMIN 0x00000001
++#define IATTR_WATCH 0x00000002
++#define IATTR_HIDE 0x00000004
++#define IATTR_FLAGS 0x00000007
++
++#define IATTR_BARRIER 0x00010000
++#define IATTR_IXUNLINK 0x00020000
++#define IATTR_IMMUTABLE 0x00040000
++
++#ifdef __KERNEL__
++
++
++#ifdef CONFIG_VSERVER_PROC_SECURE
++#define IATTR_PROC_DEFAULT ( IATTR_ADMIN | IATTR_HIDE )
++#define IATTR_PROC_SYMLINK ( IATTR_ADMIN )
++#else
++#define IATTR_PROC_DEFAULT ( IATTR_ADMIN )
++#define IATTR_PROC_SYMLINK ( IATTR_ADMIN )
++#endif
++
++#define vx_hide_check(c, m) (((m) & IATTR_HIDE) ? vx_check(c, m) : 1)
++
++#endif /* __KERNEL__ */
++
++/* inode ioctls */
++
++#define FIOC_GETXFLG _IOR('x', 5, long)
++#define FIOC_SETXFLG _IOW('x', 6, long)
++
++#else /* _VX_INODE_H */
++#warning duplicate inclusion
++#endif /* _VX_INODE_H */
+--- a/include/linux/vserver/limit_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/limit_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,69 @@
++#ifndef _VX_LIMIT_CMD_H
++#define _VX_LIMIT_CMD_H
++
++
++/* rlimit vserver commands */
++
++#define VCMD_get_rlimit VC_CMD(RLIMIT, 1, 0)
++#define VCMD_set_rlimit VC_CMD(RLIMIT, 2, 0)
++#define VCMD_get_rlimit_mask VC_CMD(RLIMIT, 3, 0)
++#define VCMD_reset_minmax VC_CMD(RLIMIT, 9, 0)
++
++struct vcmd_ctx_rlimit_v0 {
++ uint32_t id;
++ uint64_t minimum;
++ uint64_t softlimit;
++ uint64_t maximum;
++};
++
++struct vcmd_ctx_rlimit_mask_v0 {
++ uint32_t minimum;
++ uint32_t softlimit;
++ uint32_t maximum;
++};
++
++#define VCMD_rlimit_stat VC_CMD(VSTAT, 1, 0)
++
++struct vcmd_rlimit_stat_v0 {
++ uint32_t id;
++ uint32_t hits;
++ uint64_t value;
++ uint64_t minimum;
++ uint64_t maximum;
++};
++
++#define CRLIM_UNSET (0ULL)
++#define CRLIM_INFINITY (~0ULL)
++#define CRLIM_KEEP (~1ULL)
++
++#ifdef __KERNEL__
++
++#ifdef CONFIG_IA32_EMULATION
++
++struct vcmd_ctx_rlimit_v0_x32 {
++ uint32_t id;
++ uint64_t minimum;
++ uint64_t softlimit;
++ uint64_t maximum;
++} __attribute__ ((aligned (4)));
++
++#endif /* CONFIG_IA32_EMULATION */
++
++#include <linux/compiler.h>
++
++extern int vc_get_rlimit_mask(uint32_t, void __user *);
++extern int vc_get_rlimit(struct vx_info *, void __user *);
++extern int vc_set_rlimit(struct vx_info *, void __user *);
++extern int vc_reset_minmax(struct vx_info *, void __user *);
++
++extern int vc_rlimit_stat(struct vx_info *, void __user *);
++
++#ifdef CONFIG_IA32_EMULATION
++
++extern int vc_get_rlimit_x32(struct vx_info *, void __user *);
++extern int vc_set_rlimit_x32(struct vx_info *, void __user *);
++
++#endif /* CONFIG_IA32_EMULATION */
++
++#endif /* __KERNEL__ */
++#endif /* _VX_LIMIT_CMD_H */
+--- a/include/linux/vserver/limit_def.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/limit_def.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,47 @@
++#ifndef _VX_LIMIT_DEF_H
++#define _VX_LIMIT_DEF_H
++
++#include <asm/atomic.h>
++#include <asm/resource.h>
++
++#include "limit.h"
++
++
++struct _vx_res_limit {
++ rlim_t soft; /* Context soft limit */
++ rlim_t hard; /* Context hard limit */
++
++ rlim_atomic_t rcur; /* Current value */
++ rlim_t rmin; /* Context minimum */
++ rlim_t rmax; /* Context maximum */
++
++ atomic_t lhit; /* Limit hits */
++};
++
++/* context sub struct */
++
++struct _vx_limit {
++ struct _vx_res_limit res[NUM_LIMITS];
++};
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++static inline void __dump_vx_limit(struct _vx_limit *limit)
++{
++ int i;
++
++ printk("\t_vx_limit:");
++ for (i = 0; i < NUM_LIMITS; i++) {
++ printk("\t [%2d] = %8lu %8lu/%8lu, %8ld/%8ld, %8d\n",
++ i, (unsigned long)__rlim_get(limit, i),
++ (unsigned long)__rlim_rmin(limit, i),
++ (unsigned long)__rlim_rmax(limit, i),
++ (long)__rlim_soft(limit, i),
++ (long)__rlim_hard(limit, i),
++ atomic_read(&__rlim_lhit(limit, i)));
++ }
++}
++
++#endif
++
++#endif /* _VX_LIMIT_DEF_H */
+--- a/include/linux/vserver/limit.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/limit.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,70 @@
++#ifndef _VX_LIMIT_H
++#define _VX_LIMIT_H
++
++#define VLIMIT_NSOCK 16
++#define VLIMIT_OPENFD 17
++#define VLIMIT_ANON 18
++#define VLIMIT_SHMEM 19
++#define VLIMIT_SEMARY 20
++#define VLIMIT_NSEMS 21
++#define VLIMIT_DENTRY 22
++#define VLIMIT_MAPPED 23
++
++
++#ifdef __KERNEL__
++
++#define VLIM_NOCHECK ((1L << VLIMIT_DENTRY) | (1L << RLIMIT_RSS))
++
++/* keep in sync with CRLIM_INFINITY */
++
++#define VLIM_INFINITY (~0ULL)
++
++#include <asm/atomic.h>
++#include <asm/resource.h>
++
++#ifndef RLIM_INFINITY
++#warning RLIM_INFINITY is undefined
++#endif
++
++#define __rlim_val(l, r, v) ((l)->res[r].v)
++
++#define __rlim_soft(l, r) __rlim_val(l, r, soft)
++#define __rlim_hard(l, r) __rlim_val(l, r, hard)
++
++#define __rlim_rcur(l, r) __rlim_val(l, r, rcur)
++#define __rlim_rmin(l, r) __rlim_val(l, r, rmin)
++#define __rlim_rmax(l, r) __rlim_val(l, r, rmax)
++
++#define __rlim_lhit(l, r) __rlim_val(l, r, lhit)
++#define __rlim_hit(l, r) atomic_inc(&__rlim_lhit(l, r))
++
++typedef atomic_long_t rlim_atomic_t;
++typedef unsigned long rlim_t;
++
++#define __rlim_get(l, r) atomic_long_read(&__rlim_rcur(l, r))
++#define __rlim_set(l, r, v) atomic_long_set(&__rlim_rcur(l, r), v)
++#define __rlim_inc(l, r) atomic_long_inc(&__rlim_rcur(l, r))
++#define __rlim_dec(l, r) atomic_long_dec(&__rlim_rcur(l, r))
++#define __rlim_add(l, r, v) atomic_long_add(v, &__rlim_rcur(l, r))
++#define __rlim_sub(l, r, v) atomic_long_sub(v, &__rlim_rcur(l, r))
++
++
++#if (RLIM_INFINITY == VLIM_INFINITY)
++#define VX_VLIM(r) ((long long)(long)(r))
++#define VX_RLIM(v) ((rlim_t)(v))
++#else
++#define VX_VLIM(r) (((r) == RLIM_INFINITY) \
++ ? VLIM_INFINITY : (long long)(r))
++#define VX_RLIM(v) (((v) == VLIM_INFINITY) \
++ ? RLIM_INFINITY : (rlim_t)(v))
++#endif
++
++struct sysinfo;
++
++void vx_vsi_meminfo(struct sysinfo *);
++void vx_vsi_swapinfo(struct sysinfo *);
++
++#define NUM_LIMITS 24
++
++#endif /* __KERNEL__ */
++#endif /* _VX_LIMIT_H */
+--- a/include/linux/vserver/limit_int.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/limit_int.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,198 @@
++#ifndef _VX_LIMIT_INT_H
++#define _VX_LIMIT_INT_H
++
++#include "context.h"
++
++#ifdef __KERNEL__
++
++#define VXD_RCRES_COND(r) VXD_CBIT(cres, r)
++#define VXD_RLIMIT_COND(r) VXD_CBIT(limit, r)
++
++extern const char *vlimit_name[NUM_LIMITS];
++
++static inline void __vx_acc_cres(struct vx_info *vxi,
++ int res, int dir, void *_data, char *_file, int _line)
++{
++ if (VXD_RCRES_COND(res))
++ vxlprintk(1, "vx_acc_cres[%5d,%s,%2d]: %5ld%s (%p)",
++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res,
++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0),
++ (dir > 0) ? "++" : "--", _data, _file, _line);
++ if (!vxi)
++ return;
++
++ if (dir > 0)
++ __rlim_inc(&vxi->limit, res);
++ else
++ __rlim_dec(&vxi->limit, res);
++}
++
++static inline void __vx_add_cres(struct vx_info *vxi,
++ int res, int amount, void *_data, char *_file, int _line)
++{
++ if (VXD_RCRES_COND(res))
++ vxlprintk(1, "vx_add_cres[%5d,%s,%2d]: %5ld += %5d (%p)",
++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res,
++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0),
++ amount, _data, _file, _line);
++ if (amount == 0)
++ return;
++ if (!vxi)
++ return;
++ __rlim_add(&vxi->limit, res, amount);
++}
++
++static inline
++int __vx_cres_adjust_max(struct _vx_limit *limit, int res, rlim_t value)
++{
++ int cond = (value > __rlim_rmax(limit, res));
++
++ if (cond)
++ __rlim_rmax(limit, res) = value;
++ return cond;
++}
++
++static inline
++int __vx_cres_adjust_min(struct _vx_limit *limit, int res, rlim_t value)
++{
++ int cond = (value < __rlim_rmin(limit, res));
++
++ if (cond)
++ __rlim_rmin(limit, res) = value;
++ return cond;
++}
++
++static inline
++void __vx_cres_fixup(struct _vx_limit *limit, int res, rlim_t value)
++{
++ if (!__vx_cres_adjust_max(limit, res, value))
++ __vx_cres_adjust_min(limit, res, value);
++}
++
++
++/* return values:
++ +1 ... no limit hit
++ -1 ... over soft limit
++ 0 ... over hard limit */
++
++static inline int __vx_cres_avail(struct vx_info *vxi,
++ int res, int num, char *_file, int _line)
++{
++ struct _vx_limit *limit;
++ rlim_t value;
++
++ if (VXD_RLIMIT_COND(res))
++ vxlprintk(1, "vx_cres_avail[%5d,%s,%2d]: %5ld/%5ld > %5ld + %5d",
++ (vxi ? vxi->vx_id : -1), vlimit_name[res], res,
++ (vxi ? (long)__rlim_soft(&vxi->limit, res) : -1),
++ (vxi ? (long)__rlim_hard(&vxi->limit, res) : -1),
++ (vxi ? (long)__rlim_get(&vxi->limit, res) : 0),
++ num, _file, _line);
++ if (!vxi)
++ return 1;
++
++ limit = &vxi->limit;
++ value = __rlim_get(limit, res);
++
++ if (!__vx_cres_adjust_max(limit, res, value))
++ __vx_cres_adjust_min(limit, res, value);
++
++ if (num == 0)
++ return 1;
++
++ if (__rlim_soft(limit, res) == RLIM_INFINITY)
++ return -1;
++ if (value + num <= __rlim_soft(limit, res))
++ return -1;
++
++ if (__rlim_hard(limit, res) == RLIM_INFINITY)
++ return 1;
++ if (value + num <= __rlim_hard(limit, res))
++ return 1;
++
++ __rlim_hit(limit, res);
++ return 0;
++}
++
++
++static const int VLA_RSS[] = { RLIMIT_RSS, VLIMIT_ANON, VLIMIT_MAPPED, 0 };
++
++static inline
++rlim_t __vx_cres_array_sum(struct _vx_limit *limit, const int *array)
++{
++ rlim_t value, sum = 0;
++ int res;
++
++ while ((res = *array++)) {
++ value = __rlim_get(limit, res);
++ __vx_cres_fixup(limit, res, value);
++ sum += value;
++ }
++ return sum;
++}
++
++static inline
++rlim_t __vx_cres_array_fixup(struct _vx_limit *limit, const int *array)
++{
++ rlim_t value = __vx_cres_array_sum(limit, array + 1);
++ int res = *array;
++
++ if (value == __rlim_get(limit, res))
++ return value;
++
++ __rlim_set(limit, res, value);
++ /* now adjust min/max */
++ if (!__vx_cres_adjust_max(limit, res, value))
++ __vx_cres_adjust_min(limit, res, value);
++
++ return value;
++}
++
++static inline int __vx_cres_array_avail(struct vx_info *vxi,
++ const int *array, int num, char *_file, int _line)
++{
++ struct _vx_limit *limit;
++ rlim_t value = 0;
++ int res;
++
++ if (num == 0)
++ return 1;
++ if (!vxi)
++ return 1;
++
++ limit = &vxi->limit;
++ res = *array;
++ value = __vx_cres_array_sum(limit, array + 1);
++
++ __rlim_set(limit, res, value);
++ __vx_cres_fixup(limit, res, value);
++
++ return __vx_cres_avail(vxi, res, num, _file, _line);
++}
++
++
++static inline void vx_limit_fixup(struct _vx_limit *limit, int id)
++{
++ rlim_t value;
++ int res;
++
++ /* complex resources first */
++ if ((id < 0) || (id == RLIMIT_RSS))
++ __vx_cres_array_fixup(limit, VLA_RSS);
++
++ for (res = 0; res < NUM_LIMITS; res++) {
++ if ((id > 0) && (res != id))
++ continue;
++
++ value = __rlim_get(limit, res);
++ __vx_cres_fixup(limit, res, value);
++
++ /* not supposed to happen, maybe warn? */
++ if (__rlim_rmax(limit, res) > __rlim_hard(limit, res))
++ __rlim_rmax(limit, res) = __rlim_hard(limit, res);
++ }
++}
++
++
++#endif /* __KERNEL__ */
++#endif /* _VX_LIMIT_INT_H */
+--- a/include/linux/vserver/monitor.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/monitor.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,96 @@
++#ifndef _VX_MONITOR_H
++#define _VX_MONITOR_H
++
++#include <linux/types.h>
++
++enum {
++ VXM_UNUSED = 0,
++
++ VXM_SYNC = 0x10,
++
++ VXM_UPDATE = 0x20,
++ VXM_UPDATE_1,
++ VXM_UPDATE_2,
++
++ VXM_RQINFO_1 = 0x24,
++ VXM_RQINFO_2,
++
++ VXM_ACTIVATE = 0x40,
++ VXM_DEACTIVATE,
++ VXM_IDLE,
++
++ VXM_HOLD = 0x44,
++ VXM_UNHOLD,
++
++ VXM_MIGRATE = 0x48,
++ VXM_RESCHED,
++
++ /* all other bits are flags */
++ VXM_SCHED = 0x80,
++};
++
++struct _vxm_update_1 {
++ uint32_t tokens_max;
++ uint32_t fill_rate;
++ uint32_t interval;
++};
++
++struct _vxm_update_2 {
++ uint32_t tokens_min;
++ uint32_t fill_rate;
++ uint32_t interval;
++};
++
++struct _vxm_rqinfo_1 {
++ uint16_t running;
++ uint16_t onhold;
++ uint16_t iowait;
++ uint16_t uintr;
++ uint32_t idle_tokens;
++};
++
++struct _vxm_rqinfo_2 {
++ uint32_t norm_time;
++ uint32_t idle_time;
++ uint32_t idle_skip;
++};
++
++struct _vxm_sched {
++ uint32_t tokens;
++ uint32_t norm_time;
++ uint32_t idle_time;
++};
++
++struct _vxm_task {
++ uint16_t pid;
++ uint16_t state;
++};
++
++struct _vxm_event {
++ uint32_t jif;
++ union {
++ uint32_t seq;
++ uint32_t sec;
++ };
++ union {
++ uint32_t tokens;
++ uint32_t nsec;
++ struct _vxm_task tsk;
++ };
++};
++
++struct _vx_mon_entry {
++ uint16_t type;
++ uint16_t xid;
++ union {
++ struct _vxm_event ev;
++ struct _vxm_sched sd;
++ struct _vxm_update_1 u1;
++ struct _vxm_update_2 u2;
++ struct _vxm_rqinfo_1 q1;
++ struct _vxm_rqinfo_2 q2;
++ };
++};
++
++
++#endif /* _VX_MONITOR_H */
+--- a/include/linux/vserver/network_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/network_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,150 @@
++#ifndef _VX_NETWORK_CMD_H
++#define _VX_NETWORK_CMD_H
++
++
++/* vinfo commands */
++
++#define VCMD_task_nid VC_CMD(VINFO, 2, 0)
++
++#ifdef __KERNEL__
++extern int vc_task_nid(uint32_t);
++
++#endif /* __KERNEL__ */
++
++#define VCMD_nx_info VC_CMD(VINFO, 6, 0)
++
++struct vcmd_nx_info_v0 {
++ uint32_t nid;
++ /* more to come */
++};
++
++#ifdef __KERNEL__
++extern int vc_nx_info(struct nx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++#include <linux/in.h>
++#include <linux/in6.h>
++
++#define VCMD_net_create_v0 VC_CMD(VNET, 1, 0)
++#define VCMD_net_create VC_CMD(VNET, 1, 1)
++
++struct vcmd_net_create {
++ uint64_t flagword;
++};
++
++#define VCMD_net_migrate VC_CMD(NETMIG, 1, 0)
++
++#define VCMD_net_add VC_CMD(NETALT, 1, 0)
++#define VCMD_net_remove VC_CMD(NETALT, 2, 0)
++
++struct vcmd_net_addr_v0 {
++ uint16_t type;
++ uint16_t count;
++ struct in_addr ip[4];
++ struct in_addr mask[4];
++};
++
++#define VCMD_net_add_ipv4 VC_CMD(NETALT, 1, 1)
++#define VCMD_net_remove_ipv4 VC_CMD(NETALT, 2, 1)
++
++struct vcmd_net_addr_ipv4_v1 {
++ uint16_t type;
++ uint16_t flags;
++ struct in_addr ip;
++ struct in_addr mask;
++};
++
++#define VCMD_net_add_ipv6 VC_CMD(NETALT, 3, 1)
++#define VCMD_net_remove_ipv6 VC_CMD(NETALT, 4, 1)
++
++struct vcmd_net_addr_ipv6_v1 {
++ uint16_t type;
++ uint16_t flags;
++ uint32_t prefix;
++ struct in6_addr ip;
++ struct in6_addr mask;
++};
++
++#define VCMD_add_match_ipv4 VC_CMD(NETALT, 5, 0)
++#define VCMD_get_match_ipv4 VC_CMD(NETALT, 6, 0)
++
++struct vcmd_match_ipv4_v0 {
++ uint16_t type;
++ uint16_t flags;
++ uint16_t parent;
++ uint16_t prefix;
++ struct in_addr ip;
++ struct in_addr ip2;
++ struct in_addr mask;
++};
++
++#define VCMD_add_match_ipv6 VC_CMD(NETALT, 7, 0)
++#define VCMD_get_match_ipv6 VC_CMD(NETALT, 8, 0)
++
++struct vcmd_match_ipv6_v0 {
++ uint16_t type;
++ uint16_t flags;
++ uint16_t parent;
++ uint16_t prefix;
++ struct in6_addr ip;
++ struct in6_addr ip2;
++ struct in6_addr mask;
++};
++
++
++#ifdef __KERNEL__
++extern int vc_net_create(uint32_t, void __user *);
++extern int vc_net_migrate(struct nx_info *, void __user *);
++
++extern int vc_net_add(struct nx_info *, void __user *);
++extern int vc_net_remove(struct nx_info *, void __user *);
++
++extern int vc_net_add_ipv4(struct nx_info *, void __user *);
++extern int vc_net_remove_ipv4(struct nx_info *, void __user *);
++
++extern int vc_net_add_ipv6(struct nx_info *, void __user *);
++extern int vc_net_remove_ipv6(struct nx_info *, void __user *);
++
++extern int vc_add_match_ipv4(struct nx_info *, void __user *);
++extern int vc_get_match_ipv4(struct nx_info *, void __user *);
++
++extern int vc_add_match_ipv6(struct nx_info *, void __user *);
++extern int vc_get_match_ipv6(struct nx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* flag commands */
++
++#define VCMD_get_nflags VC_CMD(FLAGS, 5, 0)
++#define VCMD_set_nflags VC_CMD(FLAGS, 6, 0)
++
++struct vcmd_net_flags_v0 {
++ uint64_t flagword;
++ uint64_t mask;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_nflags(struct nx_info *, void __user *);
++extern int vc_set_nflags(struct nx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++
++/* network caps commands */
++
++#define VCMD_get_ncaps VC_CMD(FLAGS, 7, 0)
++#define VCMD_set_ncaps VC_CMD(FLAGS, 8, 0)
++
++struct vcmd_net_caps_v0 {
++ uint64_t ncaps;
++ uint64_t cmask;
++};
++
++#ifdef __KERNEL__
++extern int vc_get_ncaps(struct nx_info *, void __user *);
++extern int vc_set_ncaps(struct nx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_CONTEXT_CMD_H */
+--- a/include/linux/vserver/network.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/network.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,146 @@
++#ifndef _VX_NETWORK_H
++#define _VX_NETWORK_H
++
++#include <linux/types.h>
++
++
++#define MAX_N_CONTEXT 65535 /* Arbitrary limit */
++
++
++/* network flags */
++
++#define NXF_INFO_PRIVATE 0x00000008
++
++#define NXF_SINGLE_IP 0x00000100
++#define NXF_LBACK_REMAP 0x00000200
++#define NXF_LBACK_ALLOW 0x00000400
++
++#define NXF_HIDE_NETIF 0x02000000
++#define NXF_HIDE_LBACK 0x04000000
++
++#define NXF_STATE_SETUP (1ULL << 32)
++#define NXF_STATE_ADMIN (1ULL << 34)
++
++#define NXF_SC_HELPER (1ULL << 36)
++#define NXF_PERSISTENT (1ULL << 38)
++
++#define NXF_ONE_TIME (0x0005ULL << 32)
++
++
++#define NXF_INIT_SET (__nxf_init_set())
++
++static inline uint64_t __nxf_init_set(void) {
++ return NXF_STATE_ADMIN
++#ifdef CONFIG_VSERVER_AUTO_LBACK
++ | NXF_LBACK_REMAP
++ | NXF_HIDE_LBACK
++#endif
++#ifdef CONFIG_VSERVER_AUTO_SINGLE
++ | NXF_SINGLE_IP
++#endif
++ | NXF_HIDE_NETIF;
++}
++
++
++/* network caps */
++
++#define NXC_TUN_CREATE 0x00000001
++
++#define NXC_RAW_ICMP 0x00000100
++
++
++/* address types */
++
++#define NXA_TYPE_IPV4 0x0001
++#define NXA_TYPE_IPV6 0x0002
++
++#define NXA_TYPE_NONE 0x0000
++#define NXA_TYPE_ANY 0x00FF
++
++#define NXA_TYPE_ADDR 0x0010
++#define NXA_TYPE_MASK 0x0020
++#define NXA_TYPE_RANGE 0x0040
++
++#define NXA_MASK_ALL (NXA_TYPE_ADDR | NXA_TYPE_MASK | NXA_TYPE_RANGE)
++
++#define NXA_MOD_BCAST 0x0100
++#define NXA_MOD_LBACK 0x0200
++
++#define NXA_LOOPBACK 0x1000
++
++#define NXA_MASK_BIND (NXA_MASK_ALL | NXA_MOD_BCAST | NXA_MOD_LBACK)
++#define NXA_MASK_SHOW (NXA_MASK_ALL | NXA_LOOPBACK)
++
++#ifdef __KERNEL__
++
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/rcupdate.h>
++#include <linux/in.h>
++#include <linux/in6.h>
++#include <asm/atomic.h>
++
++struct nx_addr_v4 {
++ struct nx_addr_v4 *next;
++ struct in_addr ip[2];
++ struct in_addr mask;
++ uint16_t type;
++ uint16_t flags;
++};
++
++struct nx_addr_v6 {
++ struct nx_addr_v6 *next;
++ struct in6_addr ip;
++ struct in6_addr mask;
++ uint32_t prefix;
++ uint16_t type;
++ uint16_t flags;
++};
++
++struct nx_info {
++ struct hlist_node nx_hlist; /* linked list of nxinfos */
++ nid_t nx_id; /* vnet id */
++ atomic_t nx_usecnt; /* usage count */
++ atomic_t nx_tasks; /* tasks count */
++ int nx_state; /* context state */
++
++ uint64_t nx_flags; /* network flag word */
++ uint64_t nx_ncaps; /* network capabilities */
++
++ struct in_addr v4_lback; /* Loopback address */
++ struct in_addr v4_bcast; /* Broadcast address */
++ struct nx_addr_v4 v4; /* First/Single ipv4 address */
++#ifdef CONFIG_IPV6
++ struct nx_addr_v6 v6; /* First/Single ipv6 address */
++#endif
++ char nx_name[65]; /* network context name */
++};
++
++
++/* status flags */
++
++#define NXS_HASHED 0x0001
++#define NXS_SHUTDOWN 0x0100
++#define NXS_RELEASED 0x8000
++
++extern struct nx_info *lookup_nx_info(int);
++
++extern int get_nid_list(int, unsigned int *, int);
++extern int nid_is_hashed(nid_t);
++
++extern int nx_migrate_task(struct task_struct *, struct nx_info *);
++
++extern long vs_net_change(struct nx_info *, unsigned int);
++
++struct sock;
++
++
++#define NX_IPV4(n) ((n)->v4.type != NXA_TYPE_NONE)
++#ifdef CONFIG_IPV6
++#define NX_IPV6(n) ((n)->v6.type != NXA_TYPE_NONE)
++#else
++#define NX_IPV6(n) (0)
++#endif
++
++#endif /* __KERNEL__ */
++#endif /* _VX_NETWORK_H */
+--- a/include/linux/vserver/percpu.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/percpu.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,14 @@
++#ifndef _VX_PERCPU_H
++#define _VX_PERCPU_H
++
++#include "cvirt_def.h"
++#include "sched_def.h"
++
++struct _vx_percpu {
++ struct _vx_cvirt_pc cvirt;
++ struct _vx_sched_pc sched;
++};
++
++#define PERCPU_PERCTX (sizeof(struct _vx_percpu))
++
++#endif /* _VX_PERCPU_H */
+--- a/include/linux/vserver/pid.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/pid.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,51 @@
++#ifndef _VSERVER_PID_H
++#define _VSERVER_PID_H
++
++/* pid faking stuff */
++
++#define vx_info_map_pid(v, p) \
++ __vx_info_map_pid((v), (p), __FUNC__, __FILE__, __LINE__)
++#define vx_info_map_tgid(v,p) vx_info_map_pid(v,p)
++#define vx_map_pid(p) vx_info_map_pid(current->vx_info, p)
++#define vx_map_tgid(p) vx_map_pid(p)
++
++static inline int __vx_info_map_pid(struct vx_info *vxi, int pid,
++ const char *func, const char *file, int line)
++{
++ if (vx_info_flags(vxi, VXF_INFO_INIT, 0)) {
++ vxfprintk(VXD_CBIT(cvirt, 2),
++ "vx_map_tgid: %p/%llx: %d -> %d",
++ vxi, (long long)vxi->vx_flags, pid,
++ (pid && pid == vxi->vx_initpid) ? 1 : pid,
++ func, file, line);
++ if (pid == 0)
++ return 0;
++ if (pid == vxi->vx_initpid)
++ return 1;
++ }
++ return pid;
++}
++
++#define vx_info_rmap_pid(v, p) \
++ __vx_info_rmap_pid((v), (p), __FUNC__, __FILE__, __LINE__)
++#define vx_rmap_pid(p) vx_info_rmap_pid(current->vx_info, p)
++#define vx_rmap_tgid(p) vx_rmap_pid(p)
++
++static inline int __vx_info_rmap_pid(struct vx_info *vxi, int pid,
++ const char *func, const char *file, int line)
++{
++ if (vx_info_flags(vxi, VXF_INFO_INIT, 0)) {
++ vxfprintk(VXD_CBIT(cvirt, 2),
++ "vx_rmap_tgid: %p/%llx: %d -> %d",
++ vxi, (long long)vxi->vx_flags, pid,
++ (pid == 1) ? vxi->vx_initpid : pid,
++ func, file, line);
++ if ((pid == 1) && vxi->vx_initpid)
++ return vxi->vx_initpid;
++ if (pid == vxi->vx_initpid)
++ return ~0U;
++ }
++ return pid;
++}
++
++#endif
+--- a/include/linux/vserver/sched_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/sched_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,108 @@
++#ifndef _VX_SCHED_CMD_H
++#define _VX_SCHED_CMD_H
++
++
++/* sched vserver commands */
++
++#define VCMD_set_sched_v2 VC_CMD(SCHED, 1, 2)
++#define VCMD_set_sched_v3 VC_CMD(SCHED, 1, 3)
++#define VCMD_set_sched_v4 VC_CMD(SCHED, 1, 4)
++
++struct vcmd_set_sched_v2 {
++ int32_t fill_rate;
++ int32_t interval;
++ int32_t tokens;
++ int32_t tokens_min;
++ int32_t tokens_max;
++ uint64_t cpu_mask;
++};
++
++struct vcmd_set_sched_v3 {
++ uint32_t set_mask;
++ int32_t fill_rate;
++ int32_t interval;
++ int32_t tokens;
++ int32_t tokens_min;
++ int32_t tokens_max;
++ int32_t priority_bias;
++};
++
++struct vcmd_set_sched_v4 {
++ uint32_t set_mask;
++ int32_t fill_rate;
++ int32_t interval;
++ int32_t tokens;
++ int32_t tokens_min;
++ int32_t tokens_max;
++ int32_t prio_bias;
++ int32_t cpu_id;
++ int32_t bucket_id;
++};
++
++#define VCMD_set_sched VC_CMD(SCHED, 1, 5)
++#define VCMD_get_sched VC_CMD(SCHED, 2, 5)
++
++struct vcmd_sched_v5 {
++ uint32_t mask;
++ int32_t cpu_id;
++ int32_t bucket_id;
++ int32_t fill_rate[2];
++ int32_t interval[2];
++ int32_t tokens;
++ int32_t tokens_min;
++ int32_t tokens_max;
++ int32_t prio_bias;
++};
++
++#define VXSM_FILL_RATE 0x0001
++#define VXSM_INTERVAL 0x0002
++#define VXSM_FILL_RATE2 0x0004
++#define VXSM_INTERVAL2 0x0008
++#define VXSM_TOKENS 0x0010
++#define VXSM_TOKENS_MIN 0x0020
++#define VXSM_TOKENS_MAX 0x0040
++#define VXSM_PRIO_BIAS 0x0100
++
++#define VXSM_IDLE_TIME 0x0200
++#define VXSM_FORCE 0x0400
++
++#define VXSM_V3_MASK 0x0173
++#define VXSM_SET_MASK 0x01FF
++
++#define VXSM_CPU_ID 0x1000
++#define VXSM_BUCKET_ID 0x2000
++
++#define VXSM_MSEC 0x4000
++
++#define SCHED_KEEP (-2) /* only for v2 */
++
++#ifdef __KERNEL__
++
++#include <linux/compiler.h>
++
++extern int vc_set_sched_v2(struct vx_info *, void __user *);
++extern int vc_set_sched_v3(struct vx_info *, void __user *);
++extern int vc_set_sched_v4(struct vx_info *, void __user *);
++extern int vc_set_sched(struct vx_info *, void __user *);
++extern int vc_get_sched(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++#define VCMD_sched_info VC_CMD(SCHED, 3, 0)
++
++struct vcmd_sched_info {
++ int32_t cpu_id;
++ int32_t bucket_id;
++ uint64_t user_msec;
++ uint64_t sys_msec;
++ uint64_t hold_msec;
++ uint32_t token_usec;
++ int32_t vavavoom;
++};
++
++#ifdef __KERNEL__
++
++extern int vc_sched_info(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_SCHED_CMD_H */
+--- a/include/linux/vserver/sched_def.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/sched_def.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,68 @@
++#ifndef _VX_SCHED_DEF_H
++#define _VX_SCHED_DEF_H
++
++#include <linux/spinlock.h>
++#include <linux/jiffies.h>
++#include <linux/cpumask.h>
++#include <asm/atomic.h>
++#include <asm/param.h>
++
++
++/* context sub struct */
++
++struct _vx_sched {
++ spinlock_t tokens_lock; /* lock for token bucket */
++
++ int tokens; /* number of CPU tokens */
++ int fill_rate[2]; /* Fill rate: add X tokens... */
++ int interval[2]; /* Divisor: per Y jiffies */
++ int tokens_min; /* Limit: minimum for unhold */
++ int tokens_max; /* Limit: no more than N tokens */
++
++ int prio_bias; /* bias offset for priority */
++
++ unsigned update_mask; /* which features should be updated */
++ cpumask_t update; /* CPUs which should update */
++};
++
++struct _vx_sched_pc {
++ int tokens; /* number of CPU tokens */
++ int flags; /* bucket flags */
++
++ int fill_rate[2]; /* Fill rate: add X tokens... */
++ int interval[2]; /* Divisor: per Y jiffies */
++ int tokens_min; /* Limit: minimum for unhold */
++ int tokens_max; /* Limit: no more than N tokens */
++
++ int prio_bias; /* bias offset for priority */
++ int vavavoom; /* last calculated vavavoom */
++
++ unsigned long norm_time; /* last time accounted */
++ unsigned long idle_time; /* non linear time for fair sched */
++ unsigned long token_time; /* token time for accounting */
++ unsigned long onhold; /* jiffies when put on hold */
++
++ uint64_t user_ticks; /* token tick events */
++ uint64_t sys_ticks; /* token tick events */
++ uint64_t hold_ticks; /* token ticks paused */
++};
++
++
++#define VXSF_ONHOLD 0x0001
++#define VXSF_IDLE_TIME 0x0100
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++static inline void __dump_vx_sched(struct _vx_sched *sched)
++{
++ printk("\t_vx_sched:\n");
++ printk("\t tokens: %4d/%4d, %4d/%4d, %4d, %4d\n",
++ sched->fill_rate[0], sched->interval[0],
++ sched->fill_rate[1], sched->interval[1],
++ sched->tokens_min, sched->tokens_max);
++ printk("\t priority = %4d\n", sched->prio_bias);
++}
++
++#endif
++
++#endif /* _VX_SCHED_DEF_H */
+--- a/include/linux/vserver/sched.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/sched.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,26 @@
++#ifndef _VX_SCHED_H
++#define _VX_SCHED_H
++
++
++#ifdef __KERNEL__
++
++struct timespec;
++
++void vx_vsi_uptime(struct timespec *, struct timespec *);
++
++
++struct vx_info;
++
++void vx_update_load(struct vx_info *);
++
++
++int vx_tokens_recalc(struct _vx_sched_pc *,
++ unsigned long *, unsigned long *, int [2]);
++
++void vx_update_sched_param(struct _vx_sched *sched,
++ struct _vx_sched_pc *sched_pc);
++
++#endif /* __KERNEL__ */
++#else /* _VX_SCHED_H */
++#warning duplicate inclusion
++#endif /* _VX_SCHED_H */
+--- a/include/linux/vserver/signal_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/signal_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,43 @@
++#ifndef _VX_SIGNAL_CMD_H
++#define _VX_SIGNAL_CMD_H
++
++
++/* signalling vserver commands */
++
++#define VCMD_ctx_kill VC_CMD(PROCTRL, 1, 0)
++#define VCMD_wait_exit VC_CMD(EVENT, 99, 0)
++
++struct vcmd_ctx_kill_v0 {
++ int32_t pid;
++ int32_t sig;
++};
++
++struct vcmd_wait_exit_v0 {
++ int32_t reboot_cmd;
++ int32_t exit_code;
++};
++
++#ifdef __KERNEL__
++
++extern int vc_ctx_kill(struct vx_info *, void __user *);
++extern int vc_wait_exit(struct vx_info *, void __user *);
++
++#endif /* __KERNEL__ */
++
++/* process alteration commands */
++
++#define VCMD_get_pflags VC_CMD(PROCALT, 5, 0)
++#define VCMD_set_pflags VC_CMD(PROCALT, 6, 0)
++
++struct vcmd_pflags_v0 {
++ uint32_t flagword;
++ uint32_t mask;
++};
++
++#ifdef __KERNEL__
++
++extern int vc_get_pflags(uint32_t pid, void __user *);
++extern int vc_set_pflags(uint32_t pid, void __user *);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_SIGNAL_CMD_H */
+--- a/include/linux/vserver/signal.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/signal.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,14 @@
++#ifndef _VX_SIGNAL_H
++#define _VX_SIGNAL_H
++
++
++#ifdef __KERNEL__
++
++struct vx_info;
++
++int vx_info_kill(struct vx_info *, int, int);
++
++#endif /* __KERNEL__ */
++#else /* _VX_SIGNAL_H */
++#warning duplicate inclusion
++#endif /* _VX_SIGNAL_H */
+--- a/include/linux/vserver/space_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/space_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,29 @@
++#ifndef _VX_SPACE_CMD_H
++#define _VX_SPACE_CMD_H
++
++
++#define VCMD_enter_space_v0 VC_CMD(PROCALT, 1, 0)
++#define VCMD_enter_space VC_CMD(PROCALT, 1, 1)
++
++#define VCMD_set_space_v0 VC_CMD(PROCALT, 3, 0)
++#define VCMD_set_space VC_CMD(PROCALT, 3, 1)
++
++#define VCMD_get_space_mask_v0 VC_CMD(PROCALT, 4, 0)
++
++#define VCMD_get_space_mask VC_CMD(VSPACE, 0, 1)
++#define VCMD_get_space_default VC_CMD(VSPACE, 1, 0)
++
++
++struct vcmd_space_mask {
++ uint64_t mask;
++};
++
++
++#ifdef __KERNEL__
++
++extern int vc_enter_space(struct vx_info *, void __user *);
++extern int vc_set_space(struct vx_info *, void __user *);
++extern int vc_get_space_mask(void __user *, int);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_SPACE_CMD_H */
+--- a/include/linux/vserver/space.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/space.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,13 @@
++#ifndef _VX_SPACE_H
++#define _VX_SPACE_H
++
++
++#include <linux/types.h>
++
++struct vx_info;
++
++int vx_set_space(struct vx_info *vxi, unsigned long mask);
++
++#else /* _VX_SPACE_H */
++#warning duplicate inclusion
++#endif /* _VX_SPACE_H */
+--- a/include/linux/vserver/switch.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/switch.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,100 @@
++#ifndef _VX_SWITCH_H
++#define _VX_SWITCH_H
++
++#include <linux/types.h>
++
++
++#define VC_CATEGORY(c) (((c) >> 24) & 0x3F)
++#define VC_COMMAND(c) (((c) >> 16) & 0xFF)
++#define VC_VERSION(c) ((c) & 0xFFF)
++
++#define VC_CMD(c, i, v) ((((VC_CAT_ ## c) & 0x3F) << 24) \
++ | (((i) & 0xFF) << 16) | ((v) & 0xFFF))
++
++/*
++
++ Syscall Matrix V2.8
++
++ |VERSION|CREATE |MODIFY |MIGRATE|CONTROL|EXPERIM| |SPECIAL|SPECIAL|
++ |STATS |DESTROY|ALTER |CHANGE |LIMIT |TEST | | | |
++ |INFO |SETUP | |MOVE | | | | | |
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ SYSTEM |VERSION|VSETUP |VHOST | | | | |DEVICE | |
++ HOST | 00| 01| 02| 03| 04| 05| | 06| 07|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ CPU | |VPROC |PROCALT|PROCMIG|PROCTRL| | |SCHED. | |
++ PROCESS| 08| 09| 10| 11| 12| 13| | 14| 15|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ MEMORY | | | | |MEMCTRL| | |SWAP | |
++ | 16| 17| 18| 19| 20| 21| | 22| 23|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ NETWORK| |VNET |NETALT |NETMIG |NETCTL | | |SERIAL | |
++ | 24| 25| 26| 27| 28| 29| | 30| 31|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ DISK | | | |TAGMIG |DLIMIT | | |INODE | |
++ VFS | 32| 33| 34| 35| 36| 37| | 38| 39|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ OTHER |VSTAT | | | | | | |VINFO | |
++ | 40| 41| 42| 43| 44| 45| | 46| 47|
++ =======+=======+=======+=======+=======+=======+=======+ +=======+=======+
++ SPECIAL|EVENT | | | |FLAGS | | |VSPACE | |
++ | 48| 49| 50| 51| 52| 53| | 54| 55|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++ SPECIAL|DEBUG | | | |RLIMIT |SYSCALL| | |COMPAT |
++ | 56| 57| 58| 59| 60|TEST 61| | 62| 63|
++ -------+-------+-------+-------+-------+-------+-------+ +-------+-------+
++
++*/
++
++#define VC_CAT_VERSION 0
++
++#define VC_CAT_VSETUP 1
++#define VC_CAT_VHOST 2
++
++#define VC_CAT_DEVICE 6
++
++#define VC_CAT_VPROC 9
++#define VC_CAT_PROCALT 10
++#define VC_CAT_PROCMIG 11
++#define VC_CAT_PROCTRL 12
++
++#define VC_CAT_SCHED 14
++#define VC_CAT_MEMCTRL 20
++
++#define VC_CAT_VNET 25
++#define VC_CAT_NETALT 26
++#define VC_CAT_NETMIG 27
++#define VC_CAT_NETCTRL 28
++
++#define VC_CAT_TAGMIG 35
++#define VC_CAT_DLIMIT 36
++#define VC_CAT_INODE 38
++
++#define VC_CAT_VSTAT 40
++#define VC_CAT_VINFO 46
++#define VC_CAT_EVENT 48
++
++#define VC_CAT_FLAGS 52
++#define VC_CAT_VSPACE 54
++#define VC_CAT_DEBUG 56
++#define VC_CAT_RLIMIT 60
++
++#define VC_CAT_SYSTEST 61
++#define VC_CAT_COMPAT 63
++
++/* query version */
++
++#define VCMD_get_version VC_CMD(VERSION, 0, 0)
++#define VCMD_get_vci VC_CMD(VERSION, 1, 0)
++
++
++#ifdef __KERNEL__
++
++#include <linux/errno.h>
++
++
++#else /* __KERNEL__ */
++#define __user
++#endif /* __KERNEL__ */
++
++#endif /* _VX_SWITCH_H */
+--- a/include/linux/vserver/tag_cmd.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/tag_cmd.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,22 @@
++#ifndef _VX_TAG_CMD_H
++#define _VX_TAG_CMD_H
++
++
++/* vinfo commands */
++
++#define VCMD_task_tag VC_CMD(VINFO, 3, 0)
++
++#ifdef __KERNEL__
++extern int vc_task_tag(uint32_t);
++
++#endif /* __KERNEL__ */
++
++/* context commands */
++
++#define VCMD_tag_migrate VC_CMD(TAGMIG, 1, 0)
++
++#ifdef __KERNEL__
++extern int vc_tag_migrate(uint32_t);
++
++#endif /* __KERNEL__ */
++#endif /* _VX_TAG_CMD_H */
+--- a/include/linux/vserver/tag.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vserver/tag.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,143 @@
++#ifndef _DX_TAG_H
++#define _DX_TAG_H
++
++#include <linux/types.h>
++
++
++#define DX_TAG(in) (IS_TAGGED(in))
++
++
++#ifdef CONFIG_DX_TAG_NFSD
++#define DX_TAG_NFSD 1
++#else
++#define DX_TAG_NFSD 0
++#endif
++
++
++#ifdef CONFIG_TAGGING_NONE
++
++#define MAX_UID 0xFFFFFFFF
++#define MAX_GID 0xFFFFFFFF
++
++#define INOTAG_TAG(cond, uid, gid, tag) (0)
++
++#define TAGINO_UID(cond, uid, tag) (uid)
++#define TAGINO_GID(cond, gid, tag) (gid)
++
++#endif
++
++
++#ifdef CONFIG_TAGGING_GID16
++
++#define MAX_UID 0xFFFFFFFF
++#define MAX_GID 0x0000FFFF
++
++#define INOTAG_TAG(cond, uid, gid, tag) \
++ ((cond) ? (((gid) >> 16) & 0xFFFF) : 0)
++
++#define TAGINO_UID(cond, uid, tag) (uid)
++#define TAGINO_GID(cond, gid, tag) \
++ ((cond) ? (((gid) & 0xFFFF) | ((tag) << 16)) : (gid))
++
++#endif
++
++
++#ifdef CONFIG_TAGGING_ID24
++
++#define MAX_UID 0x00FFFFFF
++#define MAX_GID 0x00FFFFFF
++
++#define INOTAG_TAG(cond, uid, gid, tag) \
++ ((cond) ? ((((uid) >> 16) & 0xFF00) | (((gid) >> 24) & 0xFF)) : 0)
++
++#define TAGINO_UID(cond, uid, tag) \
++ ((cond) ? (((uid) & 0xFFFFFF) | (((tag) & 0xFF00) << 16)) : (uid))
++#define TAGINO_GID(cond, gid, tag) \
++ ((cond) ? (((gid) & 0xFFFFFF) | (((tag) & 0x00FF) << 24)) : (gid))
++
++#endif
++
++
++#ifdef CONFIG_TAGGING_UID16
++
++#define MAX_UID 0x0000FFFF
++#define MAX_GID 0xFFFFFFFF
++
++#define INOTAG_TAG(cond, uid, gid, tag) \
++ ((cond) ? (((uid) >> 16) & 0xFFFF) : 0)
++
++#define TAGINO_UID(cond, uid, tag) \
++ ((cond) ? (((uid) & 0xFFFF) | ((tag) << 16)) : (uid))
++#define TAGINO_GID(cond, gid, tag) (gid)
++
++#endif
++
++
++#ifdef CONFIG_TAGGING_INTERN
++
++#define MAX_UID 0xFFFFFFFF
++#define MAX_GID 0xFFFFFFFF
++
++#define INOTAG_TAG(cond, uid, gid, tag) \
++ ((cond) ? (tag) : 0)
++
++#define TAGINO_UID(cond, uid, tag) (uid)
++#define TAGINO_GID(cond, gid, tag) (gid)
++
++#endif
++
++
++#ifndef CONFIG_TAGGING_NONE
++#define dx_current_fstag(sb) \
++ ((sb)->s_flags & MS_TAGGED ? dx_current_tag() : 0)
++#else
++#define dx_current_fstag(sb) (0)
++#endif
++
++#ifndef CONFIG_TAGGING_INTERN
++#define TAGINO_TAG(cond, tag) (0)
++#else
++#define TAGINO_TAG(cond, tag) ((cond) ? (tag) : 0)
++#endif
++
++#define INOTAG_UID(cond, uid, gid) \
++ ((cond) ? ((uid) & MAX_UID) : (uid))
++#define INOTAG_GID(cond, uid, gid) \
++ ((cond) ? ((gid) & MAX_GID) : (gid))
++
++
++static inline uid_t dx_map_uid(uid_t uid)
++{
++ if ((uid > MAX_UID) && (uid != -1))
++ uid = -2;
++ return (uid & MAX_UID);
++}
++
++static inline gid_t dx_map_gid(gid_t gid)
++{
++ if ((gid > MAX_GID) && (gid != -1))
++ gid = -2;
++ return (gid & MAX_GID);
++}
++
++struct peer_tag {
++ int32_t xid;
++ int32_t nid;
++};
++
++#define dx_notagcheck(nd) ((nd) && (nd)->path.mnt && \
++ ((nd)->path.mnt->mnt_flags & MNT_NOTAGCHECK))
++
++int dx_parse_tag(char *string, tag_t *tag, int remove);
++
++#ifdef CONFIG_PROPAGATE
++
++void __dx_propagate_tag(struct nameidata *nd, struct inode *inode);
++
++#define dx_propagate_tag(n, i) __dx_propagate_tag(n, i)
++
++#else
++#define dx_propagate_tag(n, i) do { } while (0)
++#endif
++
++#endif /* _DX_TAG_H */
+--- a/include/linux/vs_inet6.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_inet6.h 2008-07-29 18:12:38.000000000 -0400
+@@ -0,0 +1,229 @@
++#ifndef _VS_INET6_H
++#define _VS_INET6_H
++
++#include "vserver/base.h"
++#include "vserver/network.h"
++#include "vserver/debug.h"
++
++#include <net/ipv6.h>
++
++#define NXAV6(a) NIP6((a)->ip), NIP6((a)->mask), (a)->prefix, (a)->type
++#define NXAV6_FMT "[" NIP6_FMT "/" NIP6_FMT "/%d:%04x]"
++
++
++#ifdef CONFIG_IPV6
++
++static inline
++int v6_addr_match(struct nx_addr_v6 *nxa,
++ const struct in6_addr *addr, uint16_t mask)
++{
++ switch (nxa->type & mask) {
++ case NXA_TYPE_MASK:
++ return ipv6_masked_addr_cmp(&nxa->ip, &nxa->mask, addr);
++ case NXA_TYPE_ADDR:
++ return ipv6_addr_equal(&nxa->ip, addr);
++ case NXA_TYPE_ANY:
++ return 1;
++ default:
++ return 0;
++ }
++}
++
++static inline
++int v6_addr_in_nx_info(struct nx_info *nxi,
++ const struct in6_addr *addr, uint16_t mask)
++{
++ struct nx_addr_v6 *nxa;
++
++ if (!nxi)
++ return 1;
++ for (nxa = &nxi->v6; nxa; nxa = nxa->next)
++ if (v6_addr_match(nxa, addr, mask))
++ return 1;
++ return 0;
++}
++
++static inline
++int v6_nx_addr_match(struct nx_addr_v6 *nxa, struct nx_addr_v6 *addr, uint16_t mask)
++{
++ /* FIXME: needs full range checks */
++ return v6_addr_match(nxa, &addr->ip, mask);
++}
++
++static inline
++int v6_nx_addr_in_nx_info(struct nx_info *nxi, struct nx_addr_v6 *nxa, uint16_t mask)
++{
++ struct nx_addr_v6 *ptr;
++
++ for (ptr = &nxi->v6; ptr; ptr = ptr->next)
++ if (v6_nx_addr_match(ptr, nxa, mask))
++ return 1;
++ return 0;
++}
++
++
++/*
++ * Check if a given address matches for a socket
++ *
++ * nxi: the socket's nx_info if any
++ * addr: to be verified address
++ */
++static inline
++int v6_sock_addr_match (
++ struct nx_info *nxi,
++ struct inet_sock *inet,
++ struct in6_addr *addr)
++{
++ struct sock *sk = &inet->sk;
++ struct in6_addr *saddr = inet6_rcv_saddr(sk);
++
++ if (!ipv6_addr_any(addr) &&
++ ipv6_addr_equal(saddr, addr))
++ return 1;
++ if (ipv6_addr_any(saddr))
++ return v6_addr_in_nx_info(nxi, addr, -1);
++ return 0;
++}
++
++/*
++ * check if address is covered by socket
++ *
++ * sk: the socket to check against
++ * addr: the address in question (must be != 0)
++ */
++
++static inline
++int __v6_addr_match_socket(const struct sock *sk, struct nx_addr_v6 *nxa)
++{
++ struct nx_info *nxi = sk->sk_nx_info;
++ struct in6_addr *saddr = inet6_rcv_saddr(sk);
++
++ vxdprintk(VXD_CBIT(net, 5),
++ "__v6_addr_in_socket(%p," NXAV6_FMT ") %p:" NIP6_FMT " %p;%lx",
++ sk, NXAV6(nxa), nxi, NIP6(*saddr), sk->sk_socket,
++ (sk->sk_socket?sk->sk_socket->flags:0));
++
++ if (!ipv6_addr_any(saddr)) { /* direct address match */
++ return v6_addr_match(nxa, saddr, -1);
++ } else if (nxi) { /* match against nx_info */
++ return v6_nx_addr_in_nx_info(nxi, nxa, -1);
++ } else { /* unrestricted any socket */
++ return 1;
++ }
++}
++
++
++/* inet related checks and helpers */
++
++
++struct in_ifaddr;
++struct net_device;
++struct sock;
++
++
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <net/inet_timewait_sock.h>
++
++
++int dev_in_nx_info(struct net_device *, struct nx_info *);
++int v6_dev_in_nx_info(struct net_device *, struct nx_info *);
++int nx_v6_addr_conflict(struct nx_info *, struct nx_info *);
++
++
++
++static inline
++int v6_ifa_in_nx_info(struct inet6_ifaddr *ifa, struct nx_info *nxi)
++{
++ if (!nxi)
++ return 1;
++ if (!ifa)
++ return 0;
++ return v6_addr_in_nx_info(nxi, &ifa->addr, -1);
++}
++
++static inline
++int nx_v6_ifa_visible(struct nx_info *nxi, struct inet6_ifaddr *ifa)
++{
++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0))
++ return 1;
++ if (v6_ifa_in_nx_info(ifa, nxi))
++ return 1;
++ return 0;
++}
++
++
++struct nx_v6_sock_addr {
++ struct in6_addr saddr; /* Address used for validation */
++ struct in6_addr baddr; /* Address used for socket bind */
++};
++
++static inline
++int v6_map_sock_addr(struct inet_sock *inet, struct sockaddr_in6 *addr,
++ struct nx_v6_sock_addr *nsa)
++{
++ // struct sock *sk = &inet->sk;
++ // struct nx_info *nxi = sk->sk_nx_info;
++ struct in6_addr saddr = addr->sin6_addr;
++ struct in6_addr baddr = saddr;
++
++ nsa->saddr = saddr;
++ nsa->baddr = baddr;
++ return 0;
++}
++
++static inline
++void v6_set_sock_addr(struct inet_sock *inet, struct nx_v6_sock_addr *nsa)
++{
++ // struct sock *sk = &inet->sk;
++ // struct in6_addr *saddr = inet6_rcv_saddr(sk);
++
++ // *saddr = nsa->baddr;
++ // inet->saddr = nsa->baddr;
++}
++
++static inline
++int nx_info_has_v6(struct nx_info *nxi)
++{
++ if (!nxi)
++ return 1;
++ if (NX_IPV6(nxi))
++ return 1;
++ return 0;
++}
++
++#else /* CONFIG_IPV6 */
++
++static inline
++int nx_v6_dev_visible(struct nx_info *n, struct net_device *d)
++{
++ return 1;
++}
++
++
++static inline
++int nx_v6_addr_conflict(struct nx_info *n, uint32_t a, const struct sock *s)
++{
++ return 1;
++}
++
++static inline
++int v6_ifa_in_nx_info(struct in_ifaddr *a, struct nx_info *n)
++{
++ return 1;
++}
++
++static inline
++int nx_info_has_v6(struct nx_info *nxi)
++{
++ return 0;
++}
++
++#endif /* CONFIG_IPV6 */
++
++#define current_nx_info_has_v6() \
++ nx_info_has_v6(current_nx_info())
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_inet.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_inet.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,342 @@
++#ifndef _VS_INET_H
++#define _VS_INET_H
++
++#include "vserver/base.h"
++#include "vserver/network.h"
++#include "vserver/debug.h"
++
++#define IPI_LOOPBACK htonl(INADDR_LOOPBACK)
++
++#define NXAV4(a) NIPQUAD((a)->ip[0]), NIPQUAD((a)->ip[1]), \
++ NIPQUAD((a)->mask), (a)->type
++#define NXAV4_FMT "[" NIPQUAD_FMT "-" NIPQUAD_FMT "/" NIPQUAD_FMT ":%04x]"
++
++
++static inline
++int v4_addr_match(struct nx_addr_v4 *nxa, __be32 addr, uint16_t tmask)
++{
++ __be32 ip = nxa->ip[0].s_addr;
++ __be32 mask = nxa->mask.s_addr;
++ __be32 bcast = ip | ~mask;
++ int ret = 0;
++
++ switch (nxa->type & tmask) {
++ case NXA_TYPE_MASK:
++ ret = (ip == (addr & mask));
++ break;
++ case NXA_TYPE_ADDR:
++ ret = 3;
++ if (addr == ip)
++ break;
++ /* fall through to broadcast */
++ case NXA_MOD_BCAST:
++ ret = ((tmask & NXA_MOD_BCAST) && (addr == bcast));
++ break;
++ case NXA_TYPE_RANGE:
++ ret = ((nxa->ip[0].s_addr <= addr) &&
++ (nxa->ip[1].s_addr > addr));
++ break;
++ case NXA_TYPE_ANY:
++ ret = 2;
++ break;
++ }
++
++ vxdprintk(VXD_CBIT(net, 0),
++ "v4_addr_match(%p" NXAV4_FMT "," NIPQUAD_FMT ",%04x) = %d",
++ nxa, NXAV4(nxa), NIPQUAD(addr), tmask, ret);
++ return ret;
++}
++
++static inline
++int v4_addr_in_nx_info(struct nx_info *nxi, __be32 addr, uint16_t tmask)
++{
++ struct nx_addr_v4 *nxa;
++ int ret = 1;
++
++ if (!nxi)
++ goto out;
++
++ ret = 2;
++ /* allow 127.0.0.1 when remapping lback */
++ if ((tmask & NXA_LOOPBACK) &&
++ (addr == IPI_LOOPBACK) &&
++ nx_info_flags(nxi, NXF_LBACK_REMAP, 0))
++ goto out;
++ ret = 3;
++ /* check for lback address */
++ if ((tmask & NXA_MOD_LBACK) &&
++ (nxi->v4_lback.s_addr == addr))
++ goto out;
++ ret = 4;
++ /* check for broadcast address */
++ if ((tmask & NXA_MOD_BCAST) &&
++ (nxi->v4_bcast.s_addr == addr))
++ goto out;
++ ret = 5;
++ /* check for v4 addresses */
++ for (nxa = &nxi->v4; nxa; nxa = nxa->next)
++ if (v4_addr_match(nxa, addr, tmask))
++ goto out;
++ ret = 0;
++out:
++ vxdprintk(VXD_CBIT(net, 0),
++ "v4_addr_in_nx_info(%p[#%u]," NIPQUAD_FMT ",%04x) = %d",
++ nxi, nxi ? nxi->nx_id : 0, NIPQUAD(addr), tmask, ret);
++ return ret;
++}
++
++static inline
++int v4_nx_addr_match(struct nx_addr_v4 *nxa, struct nx_addr_v4 *addr, uint16_t mask)
++{
++ /* FIXME: needs full range checks */
++ return v4_addr_match(nxa, addr->ip[0].s_addr, mask);
++}
++
++static inline
++int v4_nx_addr_in_nx_info(struct nx_info *nxi, struct nx_addr_v4 *nxa, uint16_t mask)
++{
++ struct nx_addr_v4 *ptr;
++
++ for (ptr = &nxi->v4; ptr; ptr = ptr->next)
++ if (v4_nx_addr_match(ptr, nxa, mask))
++ return 1;
++ return 0;
++}
++
++#include <net/inet_sock.h>
++
++/*
++ * Check if a given address matches for a socket
++ *
++ * nxi: the socket's nx_info if any
++ * addr: to be verified address
++ */
++static inline
++int v4_sock_addr_match (
++ struct nx_info *nxi,
++ struct inet_sock *inet,
++ __be32 addr)
++{
++ __be32 saddr = inet->rcv_saddr;
++ __be32 bcast = nxi ? nxi->v4_bcast.s_addr : INADDR_BROADCAST;
++
++ if (addr && (saddr == addr || bcast == addr))
++ return 1;
++ if (!saddr)
++ return v4_addr_in_nx_info(nxi, addr, NXA_MASK_BIND);
++ return 0;
++}
++
++
++/* inet related checks and helpers */
++
++
++struct in_ifaddr;
++struct net_device;
++struct sock;
++
++#ifdef CONFIG_INET
++
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <net/inet_sock.h>
++#include <net/inet_timewait_sock.h>
++
++
++int dev_in_nx_info(struct net_device *, struct nx_info *);
++int v4_dev_in_nx_info(struct net_device *, struct nx_info *);
++int nx_v4_addr_conflict(struct nx_info *, struct nx_info *);
++
++
++/*
++ * check if address is covered by socket
++ *
++ * sk: the socket to check against
++ * addr: the address in question (must be != 0)
++ */
++
++static inline
++int __v4_addr_match_socket(const struct sock *sk, struct nx_addr_v4 *nxa)
++{
++ struct nx_info *nxi = sk->sk_nx_info;
++ __be32 saddr = inet_rcv_saddr(sk);
++
++ vxdprintk(VXD_CBIT(net, 5),
++ "__v4_addr_in_socket(%p," NXAV4_FMT ") %p:" NIPQUAD_FMT " %p;%lx",
++ sk, NXAV4(nxa), nxi, NIPQUAD(saddr), sk->sk_socket,
++ (sk->sk_socket?sk->sk_socket->flags:0));
++
++ if (saddr) { /* direct address match */
++ return v4_addr_match(nxa, saddr, -1);
++ } else if (nxi) { /* match against nx_info */
++ return v4_nx_addr_in_nx_info(nxi, nxa, -1);
++ } else { /* unrestricted any socket */
++ return 1;
++ }
++}
++
++
++
++static inline
++int nx_dev_visible(struct nx_info *nxi, struct net_device *dev)
++{
++ vxdprintk(VXD_CBIT(net, 1), "nx_dev_visible(%p[#%u],%p »%s«) %d",
++ nxi, nxi ? nxi->nx_id : 0, dev, dev->name,
++ nxi ? dev_in_nx_info(dev, nxi) : 0);
++
++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0))
++ return 1;
++ if (dev_in_nx_info(dev, nxi))
++ return 1;
++ return 0;
++}
++
++
++static inline
++int v4_ifa_in_nx_info(struct in_ifaddr *ifa, struct nx_info *nxi)
++{
++ if (!nxi)
++ return 1;
++ if (!ifa)
++ return 0;
++ return v4_addr_in_nx_info(nxi, ifa->ifa_local, NXA_MASK_SHOW);
++}
++
++static inline
++int nx_v4_ifa_visible(struct nx_info *nxi, struct in_ifaddr *ifa)
++{
++ vxdprintk(VXD_CBIT(net, 1), "nx_v4_ifa_visible(%p[#%u],%p) %d",
++ nxi, nxi ? nxi->nx_id : 0, ifa,
++ nxi ? v4_ifa_in_nx_info(ifa, nxi) : 0);
++
++ if (!nx_info_flags(nxi, NXF_HIDE_NETIF, 0))
++ return 1;
++ if (v4_ifa_in_nx_info(ifa, nxi))
++ return 1;
++ return 0;
++}
++
++
++struct nx_v4_sock_addr {
++ __be32 saddr; /* Address used for validation */
++ __be32 baddr; /* Address used for socket bind */
++};
++
++static inline
++int v4_map_sock_addr(struct inet_sock *inet, struct sockaddr_in *addr,
++ struct nx_v4_sock_addr *nsa)
++{
++ struct sock *sk = &inet->sk;
++ struct nx_info *nxi = sk->sk_nx_info;
++ __be32 saddr = addr->sin_addr.s_addr;
++ __be32 baddr = saddr;
++
++ vxdprintk(VXD_CBIT(net, 3),
++ "inet_bind(%p)* %p,%p;%lx " NIPQUAD_FMT,
++ sk, sk->sk_nx_info, sk->sk_socket,
++ (sk->sk_socket ? sk->sk_socket->flags : 0),
++ NIPQUAD(saddr));
++
++ if (nxi) {
++ if (saddr == INADDR_ANY) {
++ if (nx_info_flags(nxi, NXF_SINGLE_IP, 0))
++ baddr = nxi->v4.ip[0].s_addr;
++ } else if (saddr == IPI_LOOPBACK) {
++ if (nx_info_flags(nxi, NXF_LBACK_REMAP, 0))
++ baddr = nxi->v4_lback.s_addr;
++ } else { /* normal address bind */
++ if (!v4_addr_in_nx_info(nxi, saddr, NXA_MASK_BIND))
++ return -EADDRNOTAVAIL;
++ }
++ }
++
++ vxdprintk(VXD_CBIT(net, 3),
++ "inet_bind(%p) " NIPQUAD_FMT ", " NIPQUAD_FMT,
++ sk, NIPQUAD(saddr), NIPQUAD(baddr));
++
++ nsa->saddr = saddr;
++ nsa->baddr = baddr;
++ return 0;
++}
++
++static inline
++void v4_set_sock_addr(struct inet_sock *inet, struct nx_v4_sock_addr *nsa)
++{
++ inet->saddr = nsa->baddr;
++ inet->rcv_saddr = nsa->baddr;
++}
++
++
++/*
++ * helper to simplify inet_lookup_listener
++ *
++ * nxi: the socket's nx_info if any
++ * addr: to be verified address
++ * saddr: socket address
++ */
++static inline int v4_inet_addr_match (
++ struct nx_info *nxi,
++ __be32 addr,
++ __be32 saddr)
++{
++ if (addr && (saddr == addr))
++ return 1;
++ if (!saddr)
++ return nxi ? v4_addr_in_nx_info(nxi, addr, NXA_MASK_BIND) : 1;
++ return 0;
++}
++
++static inline __be32 nx_map_sock_lback(struct nx_info *nxi, __be32 addr)
++{
++ if (nx_info_flags(nxi, NXF_HIDE_LBACK, 0) &&
++ (addr == nxi->v4_lback.s_addr))
++ return IPI_LOOPBACK;
++ return addr;
++}
++
++static inline
++int nx_info_has_v4(struct nx_info *nxi)
++{
++ if (!nxi)
++ return 1;
++ if (NX_IPV4(nxi))
++ return 1;
++ if (nx_info_flags(nxi, NXF_LBACK_REMAP, 0))
++ return 1;
++ return 0;
++}
++
++#else /* CONFIG_INET */
++
++static inline
++int nx_dev_visible(struct nx_info *n, struct net_device *d)
++{
++ return 1;
++}
++
++static inline
++int nx_v4_addr_conflict(struct nx_info *n, uint32_t a, const struct sock *s)
++{
++ return 1;
++}
++
++static inline
++int v4_ifa_in_nx_info(struct in_ifaddr *a, struct nx_info *n)
++{
++ return 1;
++}
++
++static inline
++int nx_info_has_v4(struct nx_info *nxi)
++{
++ return 0;
++}
++
++#endif /* CONFIG_INET */
++
++#define current_nx_info_has_v4() \
++ nx_info_has_v4(current_nx_info())
++
++#else
++// #warning duplicate inclusion
++#endif
+--- a/include/linux/vs_limit.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_limit.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,140 @@
++#ifndef _VS_LIMIT_H
++#define _VS_LIMIT_H
++
++#include "vserver/limit.h"
++#include "vserver/base.h"
++#include "vserver/context.h"
++#include "vserver/debug.h"
++#include "vserver/context.h"
++#include "vserver/limit_int.h"
++
++
++#define vx_acc_cres(v, d, p, r) \
++ __vx_acc_cres(v, r, d, p, __FILE__, __LINE__)
++
++#define vx_acc_cres_cond(x, d, p, r) \
++ __vx_acc_cres(((x) == vx_current_xid()) ? current->vx_info : 0, \
++ r, d, p, __FILE__, __LINE__)
++
++
++#define vx_add_cres(v, a, p, r) \
++ __vx_add_cres(v, r, a, p, __FILE__, __LINE__)
++#define vx_sub_cres(v, a, p, r) vx_add_cres(v, -(a), p, r)
++
++#define vx_add_cres_cond(x, a, p, r) \
++ __vx_add_cres(((x) == vx_current_xid()) ? current->vx_info : 0, \
++ r, a, p, __FILE__, __LINE__)
++#define vx_sub_cres_cond(x, a, p, r) vx_add_cres_cond(x, -(a), p, r)
++
++
++/* process and file limits */
++
++#define vx_nproc_inc(p) \
++ vx_acc_cres((p)->vx_info, 1, p, RLIMIT_NPROC)
++
++#define vx_nproc_dec(p) \
++ vx_acc_cres((p)->vx_info,-1, p, RLIMIT_NPROC)
++
++#define vx_files_inc(f) \
++ vx_acc_cres_cond((f)->f_xid, 1, f, RLIMIT_NOFILE)
++
++#define vx_files_dec(f) \
++ vx_acc_cres_cond((f)->f_xid,-1, f, RLIMIT_NOFILE)
++
++#define vx_locks_inc(l) \
++ vx_acc_cres_cond((l)->fl_xid, 1, l, RLIMIT_LOCKS)
++
++#define vx_locks_dec(l) \
++ vx_acc_cres_cond((l)->fl_xid,-1, l, RLIMIT_LOCKS)
++
++#define vx_openfd_inc(f) \
++ vx_acc_cres(current->vx_info, 1, (void *)(long)(f), VLIMIT_OPENFD)
++
++#define vx_openfd_dec(f) \
++ vx_acc_cres(current->vx_info,-1, (void *)(long)(f), VLIMIT_OPENFD)
++
++
++#define vx_cres_avail(v, n, r) \
++ __vx_cres_avail(v, r, n, __FILE__, __LINE__)
++
++
++#define vx_nproc_avail(n) \
++ vx_cres_avail(current->vx_info, n, RLIMIT_NPROC)
++
++#define vx_files_avail(n) \
++ vx_cres_avail(current->vx_info, n, RLIMIT_NOFILE)
++
++#define vx_locks_avail(n) \
++ vx_cres_avail(current->vx_info, n, RLIMIT_LOCKS)
++
++#define vx_openfd_avail(n) \
++ vx_cres_avail(current->vx_info, n, VLIMIT_OPENFD)
++
++
++/* dentry limits */
++
++#define vx_dentry_inc(d) do { \
++ if (atomic_read(&d->d_count) == 1) \
++ vx_acc_cres(current->vx_info, 1, d, VLIMIT_DENTRY); \
++ } while (0)
++
++#define vx_dentry_dec(d) do { \
++ if (atomic_read(&d->d_count) == 0) \
++ vx_acc_cres(current->vx_info,-1, d, VLIMIT_DENTRY); \
++ } while (0)
++
++#define vx_dentry_avail(n) \
++ vx_cres_avail(current->vx_info, n, VLIMIT_DENTRY)
++
++
++/* socket limits */
++
++#define vx_sock_inc(s) \
++ vx_acc_cres((s)->sk_vx_info, 1, s, VLIMIT_NSOCK)
++
++#define vx_sock_dec(s) \
++ vx_acc_cres((s)->sk_vx_info,-1, s, VLIMIT_NSOCK)
++
++#define vx_sock_avail(n) \
++ vx_cres_avail(current->vx_info, n, VLIMIT_NSOCK)
++
++
++/* ipc resource limits */
++
++#define vx_ipcmsg_add(v, u, a) \
++ vx_add_cres(v, a, u, RLIMIT_MSGQUEUE)
++
++#define vx_ipcmsg_sub(v, u, a) \
++ vx_sub_cres(v, a, u, RLIMIT_MSGQUEUE)
++
++#define vx_ipcmsg_avail(v, a) \
++ vx_cres_avail(v, a, RLIMIT_MSGQUEUE)
++
++
++#define vx_ipcshm_add(v, k, a) \
++ vx_add_cres(v, a, (void *)(long)(k), VLIMIT_SHMEM)
++
++#define vx_ipcshm_sub(v, k, a) \
++ vx_sub_cres(v, a, (void *)(long)(k), VLIMIT_SHMEM)
++
++#define vx_ipcshm_avail(v, a) \
++ vx_cres_avail(v, a, VLIMIT_SHMEM)
++
++
++#define vx_semary_inc(a) \
++ vx_acc_cres(current->vx_info, 1, a, VLIMIT_SEMARY)
++
++#define vx_semary_dec(a) \
++ vx_acc_cres(current->vx_info, -1, a, VLIMIT_SEMARY)
++
++
++#define vx_nsems_add(a,n) \
++ vx_add_cres(current->vx_info, n, a, VLIMIT_NSEMS)
++
++#define vx_nsems_sub(a,n) \
++ vx_sub_cres(current->vx_info, n, a, VLIMIT_NSEMS)
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_memory.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_memory.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,159 @@
++#ifndef _VS_MEMORY_H
++#define _VS_MEMORY_H
++
++#include "vserver/limit.h"
++#include "vserver/base.h"
++#include "vserver/context.h"
++#include "vserver/debug.h"
++#include "vserver/context.h"
++#include "vserver/limit_int.h"
++
++
++#define __acc_add_long(a, v) (*(v) += (a))
++#define __acc_inc_long(v) (++*(v))
++#define __acc_dec_long(v) (--*(v))
++
++#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS
++#define __acc_add_atomic(a, v) atomic_long_add(a, v)
++#define __acc_inc_atomic(v) atomic_long_inc(v)
++#define __acc_dec_atomic(v) atomic_long_dec(v)
++#else /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */
++#define __acc_add_atomic(a, v) __acc_add_long(a, v)
++#define __acc_inc_atomic(v) __acc_inc_long(v)
++#define __acc_dec_atomic(v) __acc_dec_long(v)
++#endif /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */
++
++
++#define vx_acc_page(m, d, v, r) do { \
++ if ((d) > 0) \
++ __acc_inc_long(&(m)->v); \
++ else \
++ __acc_dec_long(&(m)->v); \
++ __vx_acc_cres(m->mm_vx_info, r, d, m, __FILE__, __LINE__); \
++} while (0)
++
++#define vx_acc_page_atomic(m, d, v, r) do { \
++ if ((d) > 0) \
++ __acc_inc_atomic(&(m)->v); \
++ else \
++ __acc_dec_atomic(&(m)->v); \
++ __vx_acc_cres(m->mm_vx_info, r, d, m, __FILE__, __LINE__); \
++} while (0)
++
++
++#define vx_acc_pages(m, p, v, r) do { \
++ unsigned long __p = (p); \
++ __acc_add_long(__p, &(m)->v); \
++ __vx_add_cres(m->mm_vx_info, r, __p, m, __FILE__, __LINE__); \
++} while (0)
++
++#define vx_acc_pages_atomic(m, p, v, r) do { \
++ unsigned long __p = (p); \
++ __acc_add_atomic(__p, &(m)->v); \
++ __vx_add_cres(m->mm_vx_info, r, __p, m, __FILE__, __LINE__); \
++} while (0)
++
++
++
++#define vx_acc_vmpage(m, d) \
++ vx_acc_page(m, d, total_vm, RLIMIT_AS)
++#define vx_acc_vmlpage(m, d) \
++ vx_acc_page(m, d, locked_vm, RLIMIT_MEMLOCK)
++#define vx_acc_file_rsspage(m, d) \
++ vx_acc_page_atomic(m, d, _file_rss, VLIMIT_MAPPED)
++#define vx_acc_anon_rsspage(m, d) \
++ vx_acc_page_atomic(m, d, _anon_rss, VLIMIT_ANON)
++
++#define vx_acc_vmpages(m, p) \
++ vx_acc_pages(m, p, total_vm, RLIMIT_AS)
++#define vx_acc_vmlpages(m, p) \
++ vx_acc_pages(m, p, locked_vm, RLIMIT_MEMLOCK)
++#define vx_acc_file_rsspages(m, p) \
++ vx_acc_pages_atomic(m, p, _file_rss, VLIMIT_MAPPED)
++#define vx_acc_anon_rsspages(m, p) \
++ vx_acc_pages_atomic(m, p, _anon_rss, VLIMIT_ANON)
++
++#define vx_pages_add(s, r, p) __vx_add_cres(s, r, p, 0, __FILE__, __LINE__)
++#define vx_pages_sub(s, r, p) vx_pages_add(s, r, -(p))
++
++#define vx_vmpages_inc(m) vx_acc_vmpage(m, 1)
++#define vx_vmpages_dec(m) vx_acc_vmpage(m, -1)
++#define vx_vmpages_add(m, p) vx_acc_vmpages(m, p)
++#define vx_vmpages_sub(m, p) vx_acc_vmpages(m, -(p))
++
++#define vx_vmlocked_inc(m) vx_acc_vmlpage(m, 1)
++#define vx_vmlocked_dec(m) vx_acc_vmlpage(m, -1)
++#define vx_vmlocked_add(m, p) vx_acc_vmlpages(m, p)
++#define vx_vmlocked_sub(m, p) vx_acc_vmlpages(m, -(p))
++
++#define vx_file_rsspages_inc(m) vx_acc_file_rsspage(m, 1)
++#define vx_file_rsspages_dec(m) vx_acc_file_rsspage(m, -1)
++#define vx_file_rsspages_add(m, p) vx_acc_file_rsspages(m, p)
++#define vx_file_rsspages_sub(m, p) vx_acc_file_rsspages(m, -(p))
++
++#define vx_anon_rsspages_inc(m) vx_acc_anon_rsspage(m, 1)
++#define vx_anon_rsspages_dec(m) vx_acc_anon_rsspage(m, -1)
++#define vx_anon_rsspages_add(m, p) vx_acc_anon_rsspages(m, p)
++#define vx_anon_rsspages_sub(m, p) vx_acc_anon_rsspages(m, -(p))
++
++
++#define vx_pages_avail(m, p, r) \
++ __vx_cres_avail((m)->mm_vx_info, r, p, __FILE__, __LINE__)
++
++#define vx_vmpages_avail(m, p) vx_pages_avail(m, p, RLIMIT_AS)
++#define vx_vmlocked_avail(m, p) vx_pages_avail(m, p, RLIMIT_MEMLOCK)
++#define vx_anon_avail(m, p) vx_pages_avail(m, p, VLIMIT_ANON)
++#define vx_mapped_avail(m, p) vx_pages_avail(m, p, VLIMIT_MAPPED)
++
++#define vx_rss_avail(m, p) \
++ __vx_cres_array_avail((m)->mm_vx_info, VLA_RSS, p, __FILE__, __LINE__)
++
++
++enum {
++ VXPT_UNKNOWN = 0,
++ VXPT_ANON,
++ VXPT_NONE,
++ VXPT_FILE,
++ VXPT_SWAP,
++ VXPT_WRITE
++};
++
++#if 0
++#define vx_page_fault(mm, vma, type, ret)
++#else
++
++static inline
++void __vx_page_fault(struct mm_struct *mm,
++ struct vm_area_struct *vma, int type, int ret)
++{
++ struct vx_info *vxi = mm->mm_vx_info;
++ int what;
++/*
++ static char *page_type[6] =
++ { "UNKNOWN", "ANON", "NONE", "FILE", "SWAP", "WRITE" };
++ static char *page_what[4] =
++ { "FAULT_OOM", "FAULT_SIGBUS", "FAULT_MINOR", "FAULT_MAJOR" };
++*/
++
++ if (!vxi)
++ return;
++
++ what = (ret & 0x3);
++
++/* printk("[%d] page[%d][%d] %2x %s %s\n", vxi->vx_id,
++ type, what, ret, page_type[type], page_what[what]);
++*/
++ if (ret & VM_FAULT_WRITE)
++ what |= 0x4;
++ atomic_inc(&vxi->cacct.page[type][what]);
++}
++
++#define vx_page_fault(mm, vma, type, ret) __vx_page_fault(mm, vma, type, ret)
++#endif
++
++
++extern unsigned long vx_badness(struct task_struct *task, struct mm_struct *mm);
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_network.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_network.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,169 @@
++#ifndef _NX_VS_NETWORK_H
++#define _NX_VS_NETWORK_H
++
++#include "vserver/context.h"
++#include "vserver/network.h"
++#include "vserver/base.h"
++#include "vserver/check.h"
++#include "vserver/debug.h"
++
++#include <linux/sched.h>
++
++
++#define get_nx_info(i) __get_nx_info(i, __FILE__, __LINE__)
++
++static inline struct nx_info *__get_nx_info(struct nx_info *nxi,
++ const char *_file, int _line)
++{
++ if (!nxi)
++ return NULL;
++
++ vxlprintk(VXD_CBIT(nid, 2), "get_nx_info(%p[#%d.%d])",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi ? atomic_read(&nxi->nx_usecnt) : 0,
++ _file, _line);
++
++ atomic_inc(&nxi->nx_usecnt);
++ return nxi;
++}
++
++
++extern void free_nx_info(struct nx_info *);
++
++#define put_nx_info(i) __put_nx_info(i, __FILE__, __LINE__)
++
++static inline void __put_nx_info(struct nx_info *nxi, const char *_file, int _line)
++{
++ if (!nxi)
++ return;
++
++ vxlprintk(VXD_CBIT(nid, 2), "put_nx_info(%p[#%d.%d])",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi ? atomic_read(&nxi->nx_usecnt) : 0,
++ _file, _line);
++
++ if (atomic_dec_and_test(&nxi->nx_usecnt))
++ free_nx_info(nxi);
++}
++
++
++#define init_nx_info(p, i) __init_nx_info(p, i, __FILE__, __LINE__)
++
++static inline void __init_nx_info(struct nx_info **nxp, struct nx_info *nxi,
++ const char *_file, int _line)
++{
++ if (nxi) {
++ vxlprintk(VXD_CBIT(nid, 3),
++ "init_nx_info(%p[#%d.%d])",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi ? atomic_read(&nxi->nx_usecnt) : 0,
++ _file, _line);
++
++ atomic_inc(&nxi->nx_usecnt);
++ }
++ *nxp = nxi;
++}
++
++
++#define set_nx_info(p, i) __set_nx_info(p, i, __FILE__, __LINE__)
++
++static inline void __set_nx_info(struct nx_info **nxp, struct nx_info *nxi,
++ const char *_file, int _line)
++{
++ struct nx_info *nxo;
++
++ if (!nxi)
++ return;
++
++ vxlprintk(VXD_CBIT(nid, 3), "set_nx_info(%p[#%d.%d])",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi ? atomic_read(&nxi->nx_usecnt) : 0,
++ _file, _line);
++
++ atomic_inc(&nxi->nx_usecnt);
++ nxo = xchg(nxp, nxi);
++ BUG_ON(nxo);
++}
++
++#define clr_nx_info(p) __clr_nx_info(p, __FILE__, __LINE__)
++
++static inline void __clr_nx_info(struct nx_info **nxp,
++ const char *_file, int _line)
++{
++ struct nx_info *nxo;
++
++ nxo = xchg(nxp, NULL);
++ if (!nxo)
++ return;
++
++ vxlprintk(VXD_CBIT(nid, 3), "clr_nx_info(%p[#%d.%d])",
++ nxo, nxo ? nxo->nx_id : 0,
++ nxo ? atomic_read(&nxo->nx_usecnt) : 0,
++ _file, _line);
++
++ if (atomic_dec_and_test(&nxo->nx_usecnt))
++ free_nx_info(nxo);
++}
++
++
++#define claim_nx_info(v, p) __claim_nx_info(v, p, __FILE__, __LINE__)
++
++static inline void __claim_nx_info(struct nx_info *nxi,
++ struct task_struct *task, const char *_file, int _line)
++{
++ vxlprintk(VXD_CBIT(nid, 3), "claim_nx_info(%p[#%d.%d.%d]) %p",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi?atomic_read(&nxi->nx_usecnt):0,
++ nxi?atomic_read(&nxi->nx_tasks):0,
++ task, _file, _line);
++
++ atomic_inc(&nxi->nx_tasks);
++}
++
++
++extern void unhash_nx_info(struct nx_info *);
++
++#define release_nx_info(v, p) __release_nx_info(v, p, __FILE__, __LINE__)
++
++static inline void __release_nx_info(struct nx_info *nxi,
++ struct task_struct *task, const char *_file, int _line)
++{
++ vxlprintk(VXD_CBIT(nid, 3), "release_nx_info(%p[#%d.%d.%d]) %p",
++ nxi, nxi ? nxi->nx_id : 0,
++ nxi ? atomic_read(&nxi->nx_usecnt) : 0,
++ nxi ? atomic_read(&nxi->nx_tasks) : 0,
++ task, _file, _line);
++
++ might_sleep();
++
++ if (atomic_dec_and_test(&nxi->nx_tasks))
++ unhash_nx_info(nxi);
++}
++
++
++#define task_get_nx_info(i) __task_get_nx_info(i, __FILE__, __LINE__)
++
++static __inline__ struct nx_info *__task_get_nx_info(struct task_struct *p,
++ const char *_file, int _line)
++{
++ struct nx_info *nxi;
++
++ task_lock(p);
++ vxlprintk(VXD_CBIT(nid, 5), "task_get_nx_info(%p)",
++ p, _file, _line);
++ nxi = __get_nx_info(p->nx_info, _file, _line);
++ task_unlock(p);
++ return nxi;
++}
++
++
++static inline void exit_nx_info(struct task_struct *p)
++{
++ if (p->nx_info)
++ release_nx_info(p->nx_info, p);
++}
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_pid.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_pid.h 2008-07-27 14:11:13.000000000 -0400
+@@ -0,0 +1,95 @@
++#ifndef _VS_PID_H
++#define _VS_PID_H
++
++#include "vserver/base.h"
++#include "vserver/check.h"
++#include "vserver/context.h"
++#include "vserver/debug.h"
++#include "vserver/pid.h"
++#include <linux/pid_namespace.h>
++
++
++#define VXF_FAKE_INIT (VXF_INFO_INIT | VXF_STATE_INIT)
++
++static inline
++int vx_proc_task_visible(struct task_struct *task)
++{
++ if ((task->pid == 1) &&
++ !vx_flags(VXF_FAKE_INIT, VXF_FAKE_INIT))
++ /* show a blend through init */
++ goto visible;
++ if (vx_check(vx_task_xid(task), VS_WATCH | VS_IDENT))
++ goto visible;
++ return 0;
++visible:
++ return 1;
++}
++
++#define find_task_by_real_pid(pid) find_task_by_pid(pid, &init_pid_ns)
++
++#if 0
++
++static inline
++struct task_struct *vx_find_proc_task_by_pid(int pid)
++{
++ struct task_struct *task = find_task_by_real_pid(pid);
++
++ if (task && !vx_proc_task_visible(task)) {
++ vxdprintk(VXD_CBIT(misc, 6),
++ "dropping task (find) %p[#%u,%u] for %p[#%u,%u]",
++ task, task->xid, task->pid,
++ current, current->xid, current->pid);
++ task = NULL;
++ }
++ return task;
++}
++
++#endif
++
++static inline
++struct task_struct *vx_get_proc_task(struct inode *inode, struct pid *pid)
++{
++ struct task_struct *task = get_pid_task(pid, PIDTYPE_PID);
++
++ if (task && !vx_proc_task_visible(task)) {
++ vxdprintk(VXD_CBIT(misc, 6),
++ "dropping task (get) %p[#%u,%u] for %p[#%u,%u]",
++ task, task->xid, task->pid,
++ current, current->xid, current->pid);
++ put_task_struct(task);
++ task = NULL;
++ }
++ return task;
++}
++
++#if 0
++
++static inline
++struct task_struct *vx_child_reaper(struct task_struct *p)
++{
++ struct vx_info *vxi = p->vx_info;
++ struct task_struct *reaper = child_reaper(p);
++
++ if (!vxi)
++ goto out;
++
++ BUG_ON(!p->vx_info->vx_reaper);
++
++ /* child reaper for the guest reaper */
++ if (vxi->vx_reaper == p)
++ goto out;
++
++ reaper = vxi->vx_reaper;
++out:
++ vxdprintk(VXD_CBIT(xid, 7),
++ "vx_child_reaper(%p[#%u,%u]) = %p[#%u,%u]",
++ p, p->xid, p->pid, reaper, reaper->xid, reaper->pid);
++ return reaper;
++}
++
++#endif
++
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_sched.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_sched.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,110 @@
++#ifndef _VS_SCHED_H
++#define _VS_SCHED_H
++
++#include "vserver/base.h"
++#include "vserver/context.h"
++#include "vserver/sched.h"
++
++
++#define VAVAVOOM_RATIO 50
++
++#define MAX_PRIO_BIAS 20
++#define MIN_PRIO_BIAS -20
++
++
++#ifdef CONFIG_VSERVER_HARDCPU
++
++/*
++ * effective_prio - return the priority that is based on the static
++ * priority but is modified by bonuses/penalties.
++ *
++ * We scale the actual sleep average [0 .... MAX_SLEEP_AVG]
++ * into a -4 ... 0 ... +4 bonus/penalty range.
++ *
++ * Additionally, we scale another amount based on the number of
++ * CPU tokens currently held by the context, if the process is
++ * part of a context (and the appropriate SCHED flag is set).
++ * This ranges from -5 ... 0 ... +15, quadratically.
++ *
++ * So, the total bonus is -9 .. 0 .. +19
++ * We use ~50% of the full 0...39 priority range so that:
++ *
++ * 1) nice +19 interactive tasks do not preempt nice 0 CPU hogs.
++ * 2) nice -20 CPU hogs do not get preempted by nice 0 tasks.
++ * unless that context is far exceeding its CPU allocation.
++ *
++ * Both properties are important to certain workloads.
++ */
++static inline
++int vx_effective_vavavoom(struct _vx_sched_pc *sched_pc, int max_prio)
++{
++ int vavavoom, max;
++
++ /* lots of tokens = lots of vavavoom
++ * no tokens = no vavavoom */
++ if ((vavavoom = sched_pc->tokens) >= 0) {
++ max = sched_pc->tokens_max;
++ vavavoom = max - vavavoom;
++ max = max * max;
++ vavavoom = max_prio * VAVAVOOM_RATIO / 100
++ * (vavavoom*vavavoom - (max >> 2)) / max;
++ return vavavoom;
++ }
++ return 0;
++}
++
++
++static inline
++int vx_adjust_prio(struct task_struct *p, int prio, int max_user)
++{
++ struct vx_info *vxi = p->vx_info;
++ struct _vx_sched_pc *sched_pc;
++
++ if (!vxi)
++ return prio;
++
++ sched_pc = &vx_cpu(vxi, sched_pc);
++ if (vx_info_flags(vxi, VXF_SCHED_PRIO, 0)) {
++ int vavavoom = vx_effective_vavavoom(sched_pc, max_user);
++
++ sched_pc->vavavoom = vavavoom;
++ prio += vavavoom;
++ }
++ prio += sched_pc->prio_bias;
++ return prio;
++}
++
++#else /* !CONFIG_VSERVER_HARDCPU */
++
++static inline
++int vx_adjust_prio(struct task_struct *p, int prio, int max_user)
++{
++ struct vx_info *vxi = p->vx_info;
++
++ if (vxi)
++ prio += vx_cpu(vxi, sched_pc).prio_bias;
++ return prio;
++}
++
++#endif /* CONFIG_VSERVER_HARDCPU */
++
++
++static inline void vx_account_user(struct vx_info *vxi,
++ cputime_t cputime, int nice)
++{
++ if (!vxi)
++ return;
++ vx_cpu(vxi, sched_pc).user_ticks += cputime;
++}
++
++static inline void vx_account_system(struct vx_info *vxi,
++ cputime_t cputime, int idle)
++{
++ if (!vxi)
++ return;
++ vx_cpu(vxi, sched_pc).sys_ticks += cputime;
++}
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_socket.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_socket.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,67 @@
++#ifndef _VS_SOCKET_H
++#define _VS_SOCKET_H
++
++#include "vserver/debug.h"
++#include "vserver/base.h"
++#include "vserver/cacct.h"
++#include "vserver/context.h"
++#include "vserver/tag.h"
++
++
++/* socket accounting */
++
++#include <linux/socket.h>
++
++static inline int vx_sock_type(int family)
++{
++ switch (family) {
++ case PF_UNSPEC:
++ return VXA_SOCK_UNSPEC;
++ case PF_UNIX:
++ return VXA_SOCK_UNIX;
++ case PF_INET:
++ return VXA_SOCK_INET;
++ case PF_INET6:
++ return VXA_SOCK_INET6;
++ case PF_PACKET:
++ return VXA_SOCK_PACKET;
++ default:
++ return VXA_SOCK_OTHER;
++ }
++}
++
++#define vx_acc_sock(v, f, p, s) \
++ __vx_acc_sock(v, f, p, s, __FILE__, __LINE__)
++
++static inline void __vx_acc_sock(struct vx_info *vxi,
++ int family, int pos, int size, char *file, int line)
++{
++ if (vxi) {
++ int type = vx_sock_type(family);
++
++ atomic_long_inc(&vxi->cacct.sock[type][pos].count);
++ atomic_long_add(size, &vxi->cacct.sock[type][pos].total);
++ }
++}
++
++#define vx_sock_recv(sk, s) \
++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 0, s)
++#define vx_sock_send(sk, s) \
++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 1, s)
++#define vx_sock_fail(sk, s) \
++ vx_acc_sock((sk)->sk_vx_info, (sk)->sk_family, 2, s)
++
++
++#define sock_vx_init(s) do { \
++ (s)->sk_xid = 0; \
++ (s)->sk_vx_info = NULL; \
++ } while (0)
++
++#define sock_nx_init(s) do { \
++ (s)->sk_nid = 0; \
++ (s)->sk_nx_info = NULL; \
++ } while (0)
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_tag.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_tag.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,43 @@
++#ifndef _VS_TAG_H
++#define _VS_TAG_H
++
++#include <linux/vserver/tag.h>
++
++/* check conditions */
++
++#define DX_ADMIN 0x0001
++#define DX_WATCH 0x0002
++#define DX_HOSTID 0x0008
++
++#define DX_IDENT 0x0010
++
++#define DX_ARG_MASK 0x0010
++
++
++#define dx_task_tag(t) ((t)->tag)
++
++#define dx_current_tag() dx_task_tag(current)
++
++#define dx_check(c, m) __dx_check(dx_current_tag(), c, m)
++
++#define dx_weak_check(c, m) ((m) ? dx_check(c, m) : 1)
++
++
++/*
++ * check current context for ADMIN/WATCH and
++ * optionally against supplied argument
++ */
++static inline int __dx_check(tag_t cid, tag_t id, unsigned int mode)
++{
++ if (mode & DX_ARG_MASK) {
++ if ((mode & DX_IDENT) && (id == cid))
++ return 1;
++ }
++ return (((mode & DX_ADMIN) && (cid == 0)) ||
++ ((mode & DX_WATCH) && (cid == 1)) ||
++ ((mode & DX_HOSTID) && (id == 0)));
++}
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/linux/vs_time.h 1969-12-31 19:00:00.000000000 -0500
++++ a/include/linux/vs_time.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,19 @@
++#ifndef _VS_TIME_H
++#define _VS_TIME_H
++
++
++/* time faking stuff */
++
++#ifdef CONFIG_VSERVER_VTIME
++
++extern void vx_gettimeofday(struct timeval *tv);
++extern int vx_settimeofday(struct timespec *ts);
++
++#else
++#define vx_gettimeofday(t) do_gettimeofday(t)
++#define vx_settimeofday(t) do_settimeofday(t)
++#endif
++
++#else
++#warning duplicate inclusion
++#endif
+--- a/include/net/addrconf.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/net/addrconf.h 2008-07-29 18:14:29.000000000 -0400
+@@ -83,7 +83,8 @@ extern struct inet6_ifaddr *ipv6_ge
+ extern int ipv6_dev_get_saddr(struct net_device *dev,
+ const struct in6_addr *daddr,
+ unsigned int srcprefs,
+- struct in6_addr *saddr);
++ struct in6_addr *saddr,
++ struct nx_info *nxi);
+ extern int ipv6_get_lladdr(struct net_device *dev,
+ struct in6_addr *addr,
+ unsigned char banned_flags);
+--- a/include/net/af_unix.h 2008-04-17 12:05:44.000000000 -0400
++++ a/include/net/af_unix.h 2008-07-16 22:41:36.000000000 -0400
+@@ -4,6 +4,7 @@
+ #include <linux/socket.h>
+ #include <linux/un.h>
+ #include <linux/mutex.h>
++#include <linux/vs_base.h>
+ #include <net/sock.h>
+
+ extern void unix_inflight(struct file *fp);
+--- a/include/net/inet_sock.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/net/inet_sock.h 2008-07-16 22:46:02.000000000 -0400
+@@ -24,7 +24,7 @@
+ #include <net/flow.h>
+ #include <net/sock.h>
+ #include <net/request_sock.h>
+-#include <net/route.h>
++// #include <net/route.h>
+
+ /** struct ip_options - IP Options
+ *
+@@ -192,11 +192,6 @@ static inline int inet_sk_ehashfn(const
+ }
+
+
+-static inline int inet_iif(const struct sk_buff *skb)
+-{
+- return skb->rtable->rt_iif;
+-}
+-
+ static inline struct request_sock *inet_reqsk_alloc(struct request_sock_ops *ops)
+ {
+ struct request_sock *req = reqsk_alloc(ops);
+--- a/include/net/inet_timewait_sock.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/net/inet_timewait_sock.h 2008-07-16 22:41:36.000000000 -0400
+@@ -15,15 +15,14 @@
+ #ifndef _INET_TIMEWAIT_SOCK_
+ #define _INET_TIMEWAIT_SOCK_
+
++// #include <net/inet_sock.h>
++#include <net/sock.h>
+
+ #include <linux/list.h>
+ #include <linux/module.h>
+ #include <linux/timer.h>
+ #include <linux/types.h>
+ #include <linux/workqueue.h>
+-
+-#include <net/inet_sock.h>
+-#include <net/sock.h>
+ #include <net/tcp_states.h>
+ #include <net/timewait_sock.h>
+
+@@ -116,6 +115,10 @@ struct inet_timewait_sock {
+ #define tw_hash __tw_common.skc_hash
+ #define tw_prot __tw_common.skc_prot
+ #define tw_net __tw_common.skc_net
++#define tw_xid __tw_common.skc_xid
++#define tw_vx_info __tw_common.skc_vx_info
++#define tw_nid __tw_common.skc_nid
++#define tw_nx_info __tw_common.skc_nx_info
+ int tw_timeout;
+ volatile unsigned char tw_substate;
+ /* 3 bits hole, try to pack */
+--- a/include/net/route.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/net/route.h 2008-07-16 22:46:26.000000000 -0400
+@@ -28,6 +28,7 @@
+ #include <net/inetpeer.h>
+ #include <net/flow.h>
+ #include <net/sock.h>
++// #include <linux/in.h>
+ #include <linux/in_route.h>
+ #include <linux/rtnetlink.h>
+ #include <linux/route.h>
+@@ -85,6 +86,11 @@ struct ip_rt_acct
+ __u32 i_packets;
+ };
+
++static inline int inet_iif(const struct sk_buff *skb)
++{
++ return skb->rtable->rt_iif;
++}
++
+ struct rt_cache_stat
+ {
+ unsigned int in_hit;
+@@ -135,6 +141,9 @@ static inline void ip_rt_put(struct rtab
+ dst_release(&rt->u.dst);
+ }
+
++#include <linux/vs_base.h>
++#include <linux/vs_inet.h>
++
+ #define IPTOS_RT_MASK (IPTOS_TOS_MASK & ~3)
+
+ extern const __u8 ip_tos2prio[16];
+@@ -144,6 +153,9 @@ static inline char rt_tos2priority(u8 to
+ return ip_tos2prio[IPTOS_TOS(tos)>>1];
+ }
+
++extern int ip_v4_find_src(struct net *net, struct nx_info *,
++ struct rtable **, struct flowi *);
++
+ static inline int ip_route_connect(struct rtable **rp, __be32 dst,
+ __be32 src, u32 tos, int oif, u8 protocol,
+ __be16 sport, __be16 dport, struct sock *sk,
+@@ -161,7 +173,21 @@ static inline int ip_route_connect(struc
+
+ int err;
+ struct net *net = sock_net(sk);
+- if (!dst || !src) {
++ struct nx_info *nx_info = current->nx_info;
++
++ if (sk)
++ nx_info = sk->sk_nx_info;
++
++ vxdprintk(VXD_CBIT(net, 4),
++ "ip_route_connect(%p) %p,%p;%lx",
++ sk, nx_info, sk->sk_socket,
++ (sk->sk_socket?sk->sk_socket->flags:0));
++
++ err = ip_v4_find_src(net, nx_info, rp, &fl);
++ if (err)
++ return err;
++
++ if (!fl.fl4_dst || !fl.fl4_src) {
+ err = __ip_route_output_key(net, rp, &fl);
+ if (err)
+ return err;
+--- a/include/net/sock.h 2008-07-14 17:22:55.000000000 -0400
++++ a/include/net/sock.h 2008-07-16 22:47:02.000000000 -0400
+@@ -128,6 +128,10 @@ struct sock_common {
+ #ifdef CONFIG_NET_NS
+ struct net *skc_net;
+ #endif
++ xid_t skc_xid;
++ struct vx_info *skc_vx_info;
++ nid_t skc_nid;
++ struct nx_info *skc_nx_info;
+ };
+
+ /**
+@@ -211,6 +215,10 @@ struct sock {
+ #define sk_hash __sk_common.skc_hash
+ #define sk_prot __sk_common.skc_prot
+ #define sk_net __sk_common.skc_net
++#define sk_xid __sk_common.skc_xid
++#define sk_vx_info __sk_common.skc_vx_info
++#define sk_nid __sk_common.skc_nid
++#define sk_nx_info __sk_common.skc_nx_info
+ unsigned char sk_shutdown : 2,
+ sk_no_check : 2,
+ sk_userlocks : 4;
+--- a/init/main.c 2008-07-14 17:22:55.000000000 -0400
++++ a/init/main.c 2008-07-17 21:10:23.000000000 -0400
+@@ -60,6 +60,7 @@
+ #include <linux/sched.h>
+ #include <linux/signal.h>
+ #include <linux/idr.h>
++#include <linux/vserver/percpu.h>
+
+ #include <asm/io.h>
+ #include <asm/bugs.h>
+@@ -393,12 +394,14 @@ EXPORT_SYMBOL(__per_cpu_offset);
+
+ static void __init setup_per_cpu_areas(void)
+ {
+- unsigned long size, i;
++ unsigned long size, vspc, i;
+ char *ptr;
+ unsigned long nr_possible_cpus = num_possible_cpus();
+
++ vspc = PERCPU_PERCTX * CONFIG_VSERVER_CONTEXTS;
++
+ /* Copy section for each CPU (we discard the original) */
+- size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
++ size = ALIGN(PERCPU_ENOUGH_ROOM + vspc, PAGE_SIZE);
+ ptr = alloc_bootmem_pages(size * nr_possible_cpus);
+
+ for_each_possible_cpu(i) {
+--- a/ipc/mqueue.c 2008-07-14 17:22:55.000000000 -0400
++++ a/ipc/mqueue.c 2008-07-16 22:41:36.000000000 -0400
+@@ -31,6 +31,8 @@
+ #include <linux/mutex.h>
+ #include <linux/nsproxy.h>
+ #include <linux/pid.h>
++#include <linux/vs_context.h>
++#include <linux/vs_limit.h>
+
+ #include <net/sock.h>
+ #include "util.h"
+@@ -71,6 +73,7 @@ struct mqueue_inode_info {
+ struct sigevent notify;
+ struct pid* notify_owner;
+ struct user_struct *user; /* user who created, for accounting */
++ struct vx_info *vxi;
+ struct sock *notify_sock;
+ struct sk_buff *notify_cookie;
+
+@@ -119,6 +122,7 @@ static struct inode *mqueue_get_inode(st
+ struct mqueue_inode_info *info;
+ struct task_struct *p = current;
+ struct user_struct *u = p->user;
++ struct vx_info *vxi = p->vx_info;
+ unsigned long mq_bytes, mq_msg_tblsz;
+
+ inode->i_fop = &mqueue_file_operations;
+@@ -133,6 +137,7 @@ static struct inode *mqueue_get_inode(st
+ info->notify_owner = NULL;
+ info->qsize = 0;
+ info->user = NULL; /* set when all is ok */
++ info->vxi = NULL;
+ memset(&info->attr, 0, sizeof(info->attr));
+ info->attr.mq_maxmsg = DFLT_MSGMAX;
+ info->attr.mq_msgsize = DFLT_MSGSIZEMAX;
+@@ -147,22 +152,26 @@ static struct inode *mqueue_get_inode(st
+ spin_lock(&mq_lock);
+ if (u->mq_bytes + mq_bytes < u->mq_bytes ||
+ u->mq_bytes + mq_bytes >
+- p->signal->rlim[RLIMIT_MSGQUEUE].rlim_cur) {
++ p->signal->rlim[RLIMIT_MSGQUEUE].rlim_cur ||
++ !vx_ipcmsg_avail(vxi, mq_bytes)) {
+ spin_unlock(&mq_lock);
+ goto out_inode;
+ }
+ u->mq_bytes += mq_bytes;
++ vx_ipcmsg_add(vxi, u, mq_bytes);
+ spin_unlock(&mq_lock);
+
+ info->messages = kmalloc(mq_msg_tblsz, GFP_KERNEL);
+ if (!info->messages) {
+ spin_lock(&mq_lock);
+ u->mq_bytes -= mq_bytes;
++ vx_ipcmsg_sub(vxi, u, mq_bytes);
+ spin_unlock(&mq_lock);
+ goto out_inode;
+ }
+ /* all is ok */
+ info->user = get_uid(u);
++ info->vxi = get_vx_info(vxi);
+ } else if (S_ISDIR(mode)) {
+ inc_nlink(inode);
+ /* Some things misbehave if size == 0 on a directory */
+@@ -253,10 +262,14 @@ static void mqueue_delete_inode(struct i
+ (info->attr.mq_maxmsg * info->attr.mq_msgsize));
+ user = info->user;
+ if (user) {
++ struct vx_info *vxi = info->vxi;
++
+ spin_lock(&mq_lock);
+ user->mq_bytes -= mq_bytes;
++ vx_ipcmsg_sub(vxi, user, mq_bytes);
+ queues_count--;
+ spin_unlock(&mq_lock);
++ put_vx_info(vxi);
+ free_uid(user);
+ }
+ }
+--- a/ipc/msg.c 2008-07-14 17:22:55.000000000 -0400
++++ a/ipc/msg.c 2008-07-16 22:41:36.000000000 -0400
+@@ -38,6 +38,7 @@
+ #include <linux/rwsem.h>
+ #include <linux/nsproxy.h>
+ #include <linux/ipc_namespace.h>
++#include <linux/vs_base.h>
+
+ #include <asm/current.h>
+ #include <asm/uaccess.h>
+@@ -190,6 +191,7 @@ static int newque(struct ipc_namespace *
+
+ msq->q_perm.mode = msgflg & S_IRWXUGO;
+ msq->q_perm.key = key;
++ msq->q_perm.xid = vx_current_xid();
+
+ msq->q_perm.security = NULL;
+ retval = security_msg_queue_alloc(msq);
+--- a/ipc/namespace.c 2008-07-14 17:22:55.000000000 -0400
++++ a/ipc/namespace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -9,6 +9,8 @@
+ #include <linux/rcupdate.h>
+ #include <linux/nsproxy.h>
+ #include <linux/slab.h>
++#include <linux/vs_base.h>
++#include <linux/vserver/global.h>
+
+ #include "util.h"
+
+@@ -35,6 +37,7 @@ static struct ipc_namespace *clone_ipc_n
+ register_ipcns_notifier(ns);
+
+ kref_init(&ns->kref);
++ atomic_inc(&vs_global_ipc_ns);
+ return ns;
+ }
+
+@@ -101,6 +104,7 @@ void free_ipc_ns(struct kref *kref)
+ sem_exit_ns(ns);
+ msg_exit_ns(ns);
+ shm_exit_ns(ns);
++ atomic_dec(&vs_global_ipc_ns);
+ kfree(ns);
+ atomic_dec(&nr_ipc_ns);
+
+--- a/ipc/sem.c 2008-07-14 17:22:55.000000000 -0400
++++ a/ipc/sem.c 2008-07-16 22:41:36.000000000 -0400
+@@ -83,6 +83,8 @@
+ #include <linux/rwsem.h>
+ #include <linux/nsproxy.h>
+ #include <linux/ipc_namespace.h>
++#include <linux/vs_base.h>
++#include <linux/vs_limit.h>
+
+ #include <asm/uaccess.h>
+ #include "util.h"
+@@ -255,6 +257,7 @@ static int newary(struct ipc_namespace *
+
+ sma->sem_perm.mode = (semflg & S_IRWXUGO);
+ sma->sem_perm.key = key;
++ sma->sem_perm.xid = vx_current_xid();
+
+ sma->sem_perm.security = NULL;
+ retval = security_sem_alloc(sma);
+@@ -270,6 +273,9 @@ static int newary(struct ipc_namespace *
+ return id;
+ }
+ ns->used_sems += nsems;
++ /* FIXME: obsoleted? */
++ vx_semary_inc(sma);
++ vx_nsems_add(sma, nsems);
+
+ sma->sem_base = (struct sem *) &sma[1];
+ /* sma->sem_pending = NULL; */
+--- a/ipc/shm.c 2008-07-14 17:22:55.000000000 -0400
++++ a/ipc/shm.c 2008-07-16 22:41:36.000000000 -0400
+@@ -39,6 +39,8 @@
+ #include <linux/nsproxy.h>
+ #include <linux/mount.h>
+ #include <linux/ipc_namespace.h>
++#include <linux/vs_context.h>
++#include <linux/vs_limit.h>
+
+ #include <asm/uaccess.h>
+
+@@ -183,7 +185,12 @@ static void shm_open(struct vm_area_stru
+ */
+ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
+ {
+- ns->shm_tot -= (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ struct vx_info *vxi = lookup_vx_info(shp->shm_perm.xid);
++ int numpages = (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT;
++
++ vx_ipcshm_sub(vxi, shp, numpages);
++ ns->shm_tot -= numpages;
++
+ shm_rmid(ns, shp);
+ shm_unlock(shp);
+ if (!is_file_hugepages(shp->shm_file))
+@@ -193,6 +200,7 @@ static void shm_destroy(struct ipc_names
+ shp->mlock_user);
+ fput (shp->shm_file);
+ security_shm_free(shp);
++ put_vx_info(vxi);
+ ipc_rcu_putref(shp);
+ }
+
+@@ -362,11 +370,15 @@ static int newseg(struct ipc_namespace *
+ if (ns->shm_tot + numpages > ns->shm_ctlall)
+ return -ENOSPC;
+
++ if (!vx_ipcshm_avail(current->vx_info, numpages))
++ return -ENOSPC;
++
+ shp = ipc_rcu_alloc(sizeof(*shp));
+ if (!shp)
+ return -ENOMEM;
+
+ shp->shm_perm.key = key;
++ shp->shm_perm.xid = vx_current_xid();
+ shp->shm_perm.mode = (shmflg & S_IRWXUGO);
+ shp->mlock_user = NULL;
+
+@@ -419,6 +431,7 @@ static int newseg(struct ipc_namespace *
+ ns->shm_tot += numpages;
+ error = shp->shm_perm.id;
+ shm_unlock(shp);
++ vx_ipcshm_add(current->vx_info, key, numpages);
+ return error;
+
+ no_id:
+--- a/kernel/capability.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/capability.c 2008-07-16 22:41:36.000000000 -0400
+@@ -13,6 +13,7 @@
+ #include <linux/security.h>
+ #include <linux/syscalls.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vs_context.h>
+ #include <asm/uaccess.h>
+
+ /*
+@@ -241,6 +242,8 @@ static inline int cap_set_pg(int pgrp_nr
+
+ pgrp = find_vpid(pgrp_nr);
+ do_each_pid_task(pgrp, PIDTYPE_PGID, g) {
++ if (!vx_check(g->xid, VS_ADMIN_P | VS_IDENT))
++ continue;
+ target = g;
+ while_each_thread(g, target) {
+ if (!security_capset_check(target, effective,
+@@ -391,8 +394,12 @@ int __capable(struct task_struct *t, int
+ return 0;
+ }
+
++#include <linux/vserver/base.h>
+ int capable(int cap)
+ {
++ /* here for now so we don't require task locking */
++ if (vs_check_bit(VXC_CAP_MASK, cap) && !vx_mcaps(1L << cap))
++ return 0;
+ return __capable(current, cap);
+ }
+ EXPORT_SYMBOL(capable);
+--- a/kernel/compat.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/compat.c 2008-07-16 22:41:36.000000000 -0400
+@@ -845,7 +845,7 @@ asmlinkage long compat_sys_time(compat_t
+ compat_time_t i;
+ struct timeval tv;
+
+- do_gettimeofday(&tv);
++ vx_gettimeofday(&tv);
+ i = tv.tv_sec;
+
+ if (tloc) {
+@@ -869,7 +869,7 @@ asmlinkage long compat_sys_stime(compat_
+ if (err)
+ return err;
+
+- do_settimeofday(&tv);
++ vx_settimeofday(&tv);
+ return 0;
+ }
+
+--- a/kernel/exit.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/exit.c 2008-08-13 20:37:36.000000000 -0400
+@@ -45,6 +45,11 @@
+ #include <linux/resource.h>
+ #include <linux/blkdev.h>
+ #include <linux/task_io_accounting_ops.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/global.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/unistd.h>
+@@ -477,9 +482,11 @@ static void close_files(struct files_str
+ filp_close(file, files);
+ cond_resched();
+ }
++ vx_openfd_dec(i);
+ }
+ i++;
+ set >>= 1;
++ cond_resched();
+ }
+ }
+ }
+@@ -548,6 +555,7 @@ void put_fs_struct(struct fs_struct *fs)
+ path_put(&fs->pwd);
+ if (fs->altroot.dentry)
+ path_put(&fs->altroot);
++ atomic_dec(&vs_global_fs);
+ kmem_cache_free(fs_cachep, fs);
+ }
+ }
+@@ -1090,6 +1098,10 @@ NORET_TYPE void do_exit(long code)
+ if (tsk->splice_pipe)
+ __free_pipe_info(tsk->splice_pipe);
+
++ /* needs to stay after exit_notify() */
++ exit_vx_info(tsk, code);
++ exit_nx_info(tsk);
++
+ preempt_disable();
+ /* causes final put_task_struct in finish_task_switch(). */
+ tsk->state = TASK_DEAD;
+--- a/kernel/fork.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/fork.c 2008-07-16 22:41:36.000000000 -0400
+@@ -54,6 +54,11 @@
+ #include <linux/tty.h>
+ #include <linux/proc_fs.h>
+ #include <linux/blkdev.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_memory.h>
++#include <linux/vserver/global.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/pgalloc.h>
+@@ -114,6 +119,8 @@ void free_task(struct task_struct *tsk)
+ prop_local_destroy_single(&tsk->dirties);
+ free_thread_info(tsk->stack);
+ rt_mutex_debug_task_free(tsk);
++ clr_vx_info(&tsk->vx_info);
++ clr_nx_info(&tsk->nx_info);
+ free_task_struct(tsk);
+ }
+ EXPORT_SYMBOL(free_task);
+@@ -253,6 +260,8 @@ static int dup_mmap(struct mm_struct *mm
+ mm->free_area_cache = oldmm->mmap_base;
+ mm->cached_hole_size = ~0UL;
+ mm->map_count = 0;
++ __set_mm_counter(mm, file_rss, 0);
++ __set_mm_counter(mm, anon_rss, 0);
+ cpus_clear(mm->cpu_vm_mask);
+ mm->mm_rb = RB_ROOT;
+ rb_link = &mm->mm_rb.rb_node;
+@@ -264,7 +273,7 @@ static int dup_mmap(struct mm_struct *mm
+
+ if (mpnt->vm_flags & VM_DONTCOPY) {
+ long pages = vma_pages(mpnt);
+- mm->total_vm -= pages;
++ vx_vmpages_sub(mm, pages);
+ vm_stat_account(mm, mpnt->vm_flags, mpnt->vm_file,
+ -pages);
+ continue;
+@@ -375,8 +384,8 @@ static struct mm_struct * mm_init(struct
+ : MMF_DUMP_FILTER_DEFAULT;
+ mm->core_waiters = 0;
+ mm->nr_ptes = 0;
+- set_mm_counter(mm, file_rss, 0);
+- set_mm_counter(mm, anon_rss, 0);
++ __set_mm_counter(mm, file_rss, 0);
++ __set_mm_counter(mm, anon_rss, 0);
+ spin_lock_init(&mm->page_table_lock);
+ rwlock_init(&mm->ioctx_list_lock);
+ mm->ioctx_list = NULL;
+@@ -386,6 +395,7 @@ static struct mm_struct * mm_init(struct
+
+ if (likely(!mm_alloc_pgd(mm))) {
+ mm->def_flags = 0;
++ set_vx_info(&mm->mm_vx_info, p->vx_info);
+ return mm;
+ }
+
+@@ -418,6 +428,7 @@ void __mmdrop(struct mm_struct *mm)
+ BUG_ON(mm == &init_mm);
+ mm_free_pgd(mm);
+ destroy_context(mm);
++ clr_vx_info(&mm->mm_vx_info);
+ free_mm(mm);
+ }
+ EXPORT_SYMBOL_GPL(__mmdrop);
+@@ -534,6 +545,7 @@ struct mm_struct *dup_mm(struct task_str
+ goto fail_nomem;
+
+ memcpy(mm, oldmm, sizeof(*mm));
++ mm->mm_vx_info = NULL;
+
+ /* Initializing for Swap token stuff */
+ mm->token_priority = 0;
+@@ -567,6 +579,7 @@ fail_nocontext:
+ * If init_new_context() failed, we cannot use mmput() to free the mm
+ * because it calls destroy_context()
+ */
++ clr_vx_info(&mm->mm_vx_info);
+ mm_free_pgd(mm);
+ free_mm(mm);
+ return NULL;
+@@ -637,6 +650,7 @@ static struct fs_struct *__copy_fs_struc
+ fs->altroot.dentry = NULL;
+ }
+ read_unlock(&old->lock);
++ atomic_inc(&vs_global_fs);
+ }
+ return fs;
+ }
+@@ -879,6 +893,8 @@ static struct task_struct *copy_process(
+ int retval;
+ struct task_struct *p;
+ int cgroup_callbacks_done = 0;
++ struct vx_info *vxi;
++ struct nx_info *nxi;
+
+ if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS))
+ return ERR_PTR(-EINVAL);
+@@ -913,12 +929,28 @@ static struct task_struct *copy_process(
+ DEBUG_LOCKS_WARN_ON(!p->hardirqs_enabled);
+ DEBUG_LOCKS_WARN_ON(!p->softirqs_enabled);
+ #endif
++ init_vx_info(&p->vx_info, current->vx_info);
++ init_nx_info(&p->nx_info, current->nx_info);
++
++ /* check vserver memory */
++ if (p->mm && !(clone_flags & CLONE_VM)) {
++ if (vx_vmpages_avail(p->mm, p->mm->total_vm))
++ vx_pages_add(p->vx_info, RLIMIT_AS, p->mm->total_vm);
++ else
++ goto bad_fork_free;
++ }
++ if (p->mm && vx_flags(VXF_FORK_RSS, 0)) {
++ if (!vx_rss_avail(p->mm, get_mm_counter(p->mm, file_rss)))
++ goto bad_fork_cleanup_vm;
++ }
+ retval = -EAGAIN;
++ if (!vx_nproc_avail(1))
++ goto bad_fork_cleanup_vm;
+ if (atomic_read(&p->user->processes) >=
+ p->signal->rlim[RLIMIT_NPROC].rlim_cur) {
+ if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_RESOURCE) &&
+ p->user != current->nsproxy->user_ns->root_user)
+- goto bad_fork_free;
++ goto bad_fork_cleanup_vm;
+ }
+
+ atomic_inc(&p->user->__count);
+@@ -1220,6 +1252,18 @@ static struct task_struct *copy_process(
+
+ total_forks++;
+ spin_unlock(¤t->sighand->siglock);
++
++ /* p is copy of current */
++ vxi = p->vx_info;
++ if (vxi) {
++ claim_vx_info(vxi, p);
++ atomic_inc(&vxi->cvirt.nr_threads);
++ atomic_inc(&vxi->cvirt.total_forks);
++ vx_nproc_inc(p);
++ }
++ nxi = p->nx_info;
++ if (nxi)
++ claim_nx_info(nxi, p);
+ write_unlock_irq(&tasklist_lock);
+ proc_fork_connector(p);
+ cgroup_post_fork(p);
+@@ -1266,6 +1310,9 @@ bad_fork_cleanup_count:
+ put_group_info(p->group_info);
+ atomic_dec(&p->user->processes);
+ free_uid(p->user);
++bad_fork_cleanup_vm:
++ if (p->mm && !(clone_flags & CLONE_VM))
++ vx_pages_sub(p->vx_info, RLIMIT_AS, p->mm->total_vm);
+ bad_fork_free:
+ free_task(p);
+ fork_out:
+--- a/kernel/kthread.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/kthread.c 2008-07-27 14:11:48.000000000 -0400
+@@ -13,6 +13,7 @@
+ #include <linux/file.h>
+ #include <linux/module.h>
+ #include <linux/mutex.h>
++#include <linux/vs_pid.h>
+
+ #define KTHREAD_NICE_LEVEL (-5)
+
+@@ -98,7 +99,7 @@ static void create_kthread(struct kthrea
+ struct sched_param param = { .sched_priority = 0 };
+ wait_for_completion(&create->started);
+ read_lock(&tasklist_lock);
+- create->result = find_task_by_pid_ns(pid, &init_pid_ns);
++ create->result = find_task_by_real_pid(pid);
+ read_unlock(&tasklist_lock);
+ /*
+ * root may have changed our (kthreadd's) priority or CPU mask.
+--- a/kernel/Makefile 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/Makefile 2008-07-27 12:40:25.000000000 -0400
+@@ -11,6 +11,7 @@ obj-y = sched.o fork.o exec_domain.o
+ hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \
+ notifier.o ksysfs.o pm_qos_params.o sched_clock.o
+
++obj-y += vserver/
+ obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o
+ obj-$(CONFIG_STACKTRACE) += stacktrace.o
+ obj-y += time/
+--- a/kernel/nsproxy.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/nsproxy.c 2008-07-16 22:41:36.000000000 -0400
+@@ -20,6 +20,8 @@
+ #include <linux/mnt_namespace.h>
+ #include <linux/utsname.h>
+ #include <linux/pid_namespace.h>
++#include <linux/vserver/global.h>
++#include <linux/vserver/debug.h>
+ #include <net/net_namespace.h>
+ #include <linux/ipc_namespace.h>
+
+@@ -38,6 +40,9 @@ static inline struct nsproxy *clone_nspr
+ if (ns) {
+ memcpy(ns, orig, sizeof(struct nsproxy));
+ atomic_set(&ns->count, 1);
++ vxdprintk(VXD_CBIT(space, 2), "clone_nsproxy(%p[%u] = %p[1]",
++ orig, atomic_read(&orig->count), ns);
++ atomic_inc(&vs_global_nsproxy);
+ }
+ return ns;
+ }
+@@ -47,47 +52,51 @@ static inline struct nsproxy *clone_nspr
+ * Return the newly created nsproxy. Do not attach this to the task,
+ * leave it to the caller to do proper locking and attach it to task.
+ */
+-static struct nsproxy *create_new_namespaces(unsigned long flags,
+- struct task_struct *tsk, struct fs_struct *new_fs)
++static struct nsproxy *unshare_namespaces(unsigned long flags,
++ struct nsproxy *orig, struct fs_struct *new_fs)
+ {
+ struct nsproxy *new_nsp;
+ int err;
+
+- new_nsp = clone_nsproxy(tsk->nsproxy);
++ vxdprintk(VXD_CBIT(space, 4),
++ "unshare_namespaces(0x%08lx,%p,%p)",
++ flags, orig, new_fs);
++
++ new_nsp = clone_nsproxy(orig);
+ if (!new_nsp)
+ return ERR_PTR(-ENOMEM);
+
+- new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, new_fs);
++ new_nsp->mnt_ns = copy_mnt_ns(flags, orig->mnt_ns, new_fs);
+ if (IS_ERR(new_nsp->mnt_ns)) {
+ err = PTR_ERR(new_nsp->mnt_ns);
+ goto out_ns;
+ }
+
+- new_nsp->uts_ns = copy_utsname(flags, tsk->nsproxy->uts_ns);
++ new_nsp->uts_ns = copy_utsname(flags, orig->uts_ns);
+ if (IS_ERR(new_nsp->uts_ns)) {
+ err = PTR_ERR(new_nsp->uts_ns);
+ goto out_uts;
+ }
+
+- new_nsp->ipc_ns = copy_ipcs(flags, tsk->nsproxy->ipc_ns);
++ new_nsp->ipc_ns = copy_ipcs(flags, orig->ipc_ns);
+ if (IS_ERR(new_nsp->ipc_ns)) {
+ err = PTR_ERR(new_nsp->ipc_ns);
+ goto out_ipc;
+ }
+
+- new_nsp->pid_ns = copy_pid_ns(flags, task_active_pid_ns(tsk));
++ new_nsp->pid_ns = copy_pid_ns(flags, orig->pid_ns);
+ if (IS_ERR(new_nsp->pid_ns)) {
+ err = PTR_ERR(new_nsp->pid_ns);
+ goto out_pid;
+ }
+
+- new_nsp->user_ns = copy_user_ns(flags, tsk->nsproxy->user_ns);
++ new_nsp->user_ns = copy_user_ns(flags, orig->user_ns);
+ if (IS_ERR(new_nsp->user_ns)) {
+ err = PTR_ERR(new_nsp->user_ns);
+ goto out_user;
+ }
+
+- new_nsp->net_ns = copy_net_ns(flags, tsk->nsproxy->net_ns);
++ new_nsp->net_ns = copy_net_ns(flags, orig->net_ns);
+ if (IS_ERR(new_nsp->net_ns)) {
+ err = PTR_ERR(new_nsp->net_ns);
+ goto out_net;
+@@ -115,6 +124,37 @@ out_ns:
+ return ERR_PTR(err);
+ }
+
++static struct nsproxy *create_new_namespaces(int flags, struct task_struct *tsk,
++ struct fs_struct *new_fs)
++{
++ return unshare_namespaces(flags, tsk->nsproxy, new_fs);
++}
++
++/*
++ * copies the nsproxy, setting refcount to 1, and grabbing a
++ * reference to all contained namespaces.
++ */
++struct nsproxy *copy_nsproxy(struct nsproxy *orig)
++{
++ struct nsproxy *ns = clone_nsproxy(orig);
++
++ if (ns) {
++ if (ns->mnt_ns)
++ get_mnt_ns(ns->mnt_ns);
++ if (ns->uts_ns)
++ get_uts_ns(ns->uts_ns);
++ if (ns->ipc_ns)
++ get_ipc_ns(ns->ipc_ns);
++ if (ns->pid_ns)
++ get_pid_ns(ns->pid_ns);
++ if (ns->user_ns)
++ get_user_ns(ns->user_ns);
++ if (ns->net_ns)
++ get_net(ns->net_ns);
++ }
++ return ns;
++}
++
+ /*
+ * called from clone. This now handles copy for nsproxy and all
+ * namespaces therein.
+@@ -122,9 +162,12 @@ out_ns:
+ int copy_namespaces(unsigned long flags, struct task_struct *tsk)
+ {
+ struct nsproxy *old_ns = tsk->nsproxy;
+- struct nsproxy *new_ns;
++ struct nsproxy *new_ns = NULL;
+ int err = 0;
+
++ vxdprintk(VXD_CBIT(space, 7), "copy_namespaces(0x%08lx,%p[%p])",
++ flags, tsk, old_ns);
++
+ if (!old_ns)
+ return 0;
+
+@@ -167,6 +210,9 @@ int copy_namespaces(unsigned long flags,
+
+ out:
+ put_nsproxy(old_ns);
++ vxdprintk(VXD_CBIT(space, 3),
++ "copy_namespaces(0x%08lx,%p[%p]) = %d [%p]",
++ flags, tsk, old_ns, err, new_ns);
+ return err;
+ }
+
+@@ -182,7 +228,9 @@ void free_nsproxy(struct nsproxy *ns)
+ put_pid_ns(ns->pid_ns);
+ if (ns->user_ns)
+ put_user_ns(ns->user_ns);
+- put_net(ns->net_ns);
++ if (ns->net_ns)
++ put_net(ns->net_ns);
++ atomic_dec(&vs_global_nsproxy);
+ kmem_cache_free(nsproxy_cachep, ns);
+ }
+
+@@ -195,6 +243,10 @@ int unshare_nsproxy_namespaces(unsigned
+ {
+ int err = 0;
+
++ vxdprintk(VXD_CBIT(space, 4),
++ "unshare_nsproxy_namespaces(0x%08lx,[%p])",
++ unshare_flags, current->nsproxy);
++
+ if (!(unshare_flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC |
+ CLONE_NEWUSER | CLONE_NEWNET)))
+ return 0;
+--- a/kernel/pid.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/pid.c 2008-07-16 22:41:36.000000000 -0400
+@@ -35,6 +35,8 @@
+ #include <linux/pid_namespace.h>
+ #include <linux/init_task.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/global.h>
+
+ #define pid_hashfn(nr, ns) \
+ hash_long((unsigned long)nr + (unsigned long)ns, pidhash_shift)
+@@ -304,7 +306,7 @@ EXPORT_SYMBOL_GPL(find_pid_ns);
+
+ struct pid *find_vpid(int nr)
+ {
+- return find_pid_ns(nr, current->nsproxy->pid_ns);
++ return find_pid_ns(vx_rmap_pid(nr), current->nsproxy->pid_ns);
+ }
+ EXPORT_SYMBOL_GPL(find_vpid);
+
+@@ -370,6 +372,9 @@ void transfer_pid(struct task_struct *ol
+ struct task_struct *pid_task(struct pid *pid, enum pid_type type)
+ {
+ struct task_struct *result = NULL;
++
++ if (type == PIDTYPE_REALPID)
++ type = PIDTYPE_PID;
+ if (pid) {
+ struct hlist_node *first;
+ first = rcu_dereference(pid->tasks[type].first);
+@@ -393,14 +398,14 @@ EXPORT_SYMBOL(find_task_by_pid_type_ns);
+
+ struct task_struct *find_task_by_vpid(pid_t vnr)
+ {
+- return find_task_by_pid_type_ns(PIDTYPE_PID, vnr,
++ return find_task_by_pid_type_ns(PIDTYPE_PID, vx_rmap_pid(vnr),
+ current->nsproxy->pid_ns);
+ }
+ EXPORT_SYMBOL(find_task_by_vpid);
+
+ struct task_struct *find_task_by_pid_ns(pid_t nr, struct pid_namespace *ns)
+ {
+- return find_task_by_pid_type_ns(PIDTYPE_PID, nr, ns);
++ return find_task_by_pid_type_ns(PIDTYPE_PID, vx_rmap_pid(nr), ns);
+ }
+ EXPORT_SYMBOL(find_task_by_pid_ns);
+
+@@ -435,7 +440,7 @@ struct pid *find_get_pid(pid_t nr)
+ return pid;
+ }
+
+-pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns)
++pid_t pid_unmapped_nr_ns(struct pid *pid, struct pid_namespace *ns)
+ {
+ struct upid *upid;
+ pid_t nr = 0;
+@@ -448,6 +453,11 @@ pid_t pid_nr_ns(struct pid *pid, struct
+ return nr;
+ }
+
++pid_t pid_nr_ns(struct pid *pid, struct pid_namespace *ns)
++{
++ return vx_map_pid(pid_unmapped_nr_ns(pid, ns));
++}
++
+ pid_t pid_vnr(struct pid *pid)
+ {
+ return pid_nr_ns(pid, current->nsproxy->pid_ns);
+--- a/kernel/pid_namespace.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/pid_namespace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -12,6 +12,7 @@
+ #include <linux/pid_namespace.h>
+ #include <linux/syscalls.h>
+ #include <linux/err.h>
++#include <linux/vserver/global.h>
+
+ #define BITS_PER_PAGE (PAGE_SIZE*8)
+
+@@ -84,6 +85,7 @@ static struct pid_namespace *create_pid_
+ goto out_free_map;
+
+ kref_init(&ns->kref);
++ atomic_inc(&vs_global_pid_ns);
+ ns->last_pid = 0;
+ ns->child_reaper = NULL;
+ ns->level = level;
+@@ -112,6 +114,7 @@ static void destroy_pid_namespace(struct
+
+ for (i = 0; i < PIDMAP_ENTRIES; i++)
+ kfree(ns->pidmap[i].page);
++ atomic_dec(&vs_global_pid_ns);
+ kmem_cache_free(pid_ns_cachep, ns);
+ }
+
+--- a/kernel/posix-timers.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/posix-timers.c 2008-07-27 12:57:35.000000000 -0400
+@@ -46,6 +46,7 @@
+ #include <linux/wait.h>
+ #include <linux/workqueue.h>
+ #include <linux/module.h>
++#include <linux/vs_context.h>
+
+ /*
+ * Management arrays for POSIX timers. Timers are kept in slab memory
+@@ -298,6 +299,12 @@ int posix_timer_event(struct k_itimer *timr, int si_private)
+ * and re-schedules it while ->sigq is pending.
+ * Not really bad, but not that we want.
+ */
++ struct vx_info_save vxis;
++ struct vx_info *vxi;
++ int ret;
++
++ vxi = task_get_vx_info(timr->it_process);
++ enter_vx_info(vxi, &vxis);
+ timr->sigq->info.si_sys_private = si_private;
+
+ timr->sigq->info.si_signo = timr->it_sigev_signo;
+@@ -310,10 +317,11 @@ int posix_timer_event(struct k_itimer *t
+
+ if (timr->it_sigev_notify & SIGEV_THREAD_ID) {
+ struct task_struct *leader;
+- int ret = send_sigqueue(timr->sigq, timr->it_process, 0);
++
++ ret = send_sigqueue(timr->sigq, timr->it_process, 0);
+
+ if (likely(ret >= 0))
+- return ret;
++ goto out;
+
+ timr->it_sigev_notify = SIGEV_SIGNAL;
+ leader = timr->it_process->group_leader;
+@@ -321,7 +329,11 @@ int posix_timer_event(struct k_itimer *t
+ timr->it_process = leader;
+ }
+
+- return send_sigqueue(timr->sigq, timr->it_process, 1);
++ ret = send_sigqueue(timr->sigq, timr->it_process, 1);
++out:
++ leave_vx_info(&vxis);
++ put_vx_info(vxi);
++ return ret;
+ }
+ EXPORT_SYMBOL_GPL(posix_timer_event);
+
+--- a/kernel/printk.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/printk.c 2008-07-16 22:41:36.000000000 -0400
+@@ -32,6 +32,7 @@
+ #include <linux/security.h>
+ #include <linux/bootmem.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_cvirt.h>
+
+ #include <asm/uaccess.h>
+
+@@ -300,18 +301,13 @@ int do_syslog(int type, char __user *buf
+ unsigned i, j, limit, count;
+ int do_clear = 0;
+ char c;
+- int error = 0;
++ int error;
+
+ error = security_syslog(type);
+ if (error)
+ return error;
+
+- switch (type) {
+- case 0: /* Close log */
+- break;
+- case 1: /* Open log */
+- break;
+- case 2: /* Read from log */
++ if ((type >= 2) && (type <= 4)) {
+ error = -EINVAL;
+ if (!buf || len < 0)
+ goto out;
+@@ -322,6 +318,16 @@ int do_syslog(int type, char __user *buf
+ error = -EFAULT;
+ goto out;
+ }
++ }
++ if (!vx_check(0, VS_ADMIN|VS_WATCH))
++ return vx_do_syslog(type, buf, len);
++
++ switch (type) {
++ case 0: /* Close log */
++ break;
++ case 1: /* Open log */
++ break;
++ case 2: /* Read from log */
+ error = wait_event_interruptible(log_wait,
+ (log_start - log_end));
+ if (error)
+@@ -346,16 +352,6 @@ int do_syslog(int type, char __user *buf
+ do_clear = 1;
+ /* FALL THRU */
+ case 3: /* Read last kernel messages */
+- error = -EINVAL;
+- if (!buf || len < 0)
+- goto out;
+- error = 0;
+- if (!len)
+- goto out;
+- if (!access_ok(VERIFY_WRITE, buf, len)) {
+- error = -EFAULT;
+- goto out;
+- }
+ count = len;
+ if (count > log_buf_len)
+ count = log_buf_len;
+--- a/kernel/ptrace.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/ptrace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/audit.h>
+ #include <linux/pid_namespace.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_context.h>
+
+ #include <asm/pgtable.h>
+ #include <asm/uaccess.h>
+@@ -147,6 +148,11 @@ int __ptrace_may_attach(struct task_stru
+ dumpable = get_dumpable(task->mm);
+ if (!dumpable && !capable(CAP_SYS_PTRACE))
+ return -EPERM;
++ if (!vx_check(task->xid, VS_ADMIN_P|VS_IDENT))
++ return -EPERM;
++ if (!vx_check(task->xid, VS_IDENT) &&
++ !task_vx_flags(task, VXF_STATE_ADMIN, 0))
++ return -EACCES;
+
+ return security_ptrace(current, task);
+ }
+@@ -556,6 +562,10 @@ asmlinkage long sys_ptrace(long request,
+ goto out;
+ }
+
++ ret = -EPERM;
++ if (!vx_check(vx_task_xid(child), VS_WATCH_P | VS_IDENT))
++ goto out_put_task_struct;
++
+ if (request == PTRACE_ATTACH) {
+ ret = ptrace_attach(child);
+ /*
+--- a/kernel/sched.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/sched.c 2008-07-17 21:13:30.000000000 -0400
+@@ -70,6 +70,8 @@
+ #include <linux/bootmem.h>
+ #include <linux/debugfs.h>
+ #include <linux/ctype.h>
++#include <linux/vs_sched.h>
++#include <linux/vs_cvirt.h>
+
+ #include <asm/tlb.h>
+ #include <asm/irq_regs.h>
+@@ -461,6 +463,16 @@ struct root_domain {
+ static struct root_domain def_root_domain;
+
+ #endif
++ unsigned long norm_time;
++ unsigned long idle_time;
++#ifdef CONFIG_VSERVER_IDLETIME
++ int idle_skip;
++#endif
++#ifdef CONFIG_VSERVER_HARDCPU
++ struct list_head hold_queue;
++ unsigned long nr_onhold;
++ int idle_tokens;
++#endif
+
+ /*
+ * This is the main, per-CPU runqueue data structure.
+@@ -1546,6 +1558,7 @@ static void set_load_weight(struct task_
+
+ static void enqueue_task(struct rq *rq, struct task_struct *p, int wakeup)
+ {
++ // BUG_ON(p->state & TASK_ONHOLD);
+ sched_info_queued(p);
+ p->sched_class->enqueue_task(rq, p, wakeup);
+ p->se.on_rq = 1;
+@@ -1736,6 +1749,9 @@ struct migration_req {
+ struct completion done;
+ };
+
++#include "sched_mon.h"
++
++
+ /*
+ * The task's runqueue lock must be held.
+ * Returns true if you have to wait for migration thread.
+@@ -1745,6 +1761,7 @@ migrate_task(struct task_struct *p, int
+ {
+ struct rq *rq = task_rq(p);
+
++ vxm_migrate_task(p, rq, dest_cpu);
+ /*
+ * If the task is not on a runqueue (and not running), then
+ * it is sufficient to simply update the task's cpu field.
+@@ -2109,6 +2126,12 @@ static int try_to_wake_up(struct task_st
+ /* might preempt at this point */
+ rq = task_rq_lock(p, &flags);
+ old_state = p->state;
++
++ /* we need to unhold suspended tasks
++ if (old_state & TASK_ONHOLD) {
++ vx_unhold_task(p, rq);
++ old_state = p->state;
++ } */
+ if (!(old_state & state))
+ goto out;
+ if (p->se.on_rq)
+@@ -3885,13 +3908,16 @@ unsigned long long task_sched_runtime(st
+ void account_user_time(struct task_struct *p, cputime_t cputime)
+ {
+ struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat;
++ struct vx_info *vxi = p->vx_info; /* p is _always_ current */
+ cputime64_t tmp;
++ int nice = (TASK_NICE(p) > 0);
+
+ p->utime = cputime_add(p->utime, cputime);
++ vx_account_user(vxi, cputime, nice);
+
+ /* Add user time to cpustat. */
+ tmp = cputime_to_cputime64(cputime);
+- if (TASK_NICE(p) > 0)
++ if (nice)
+ cpustat->nice = cputime64_add(cpustat->nice, tmp);
+ else
+ cpustat->user = cputime64_add(cpustat->user, tmp);
+@@ -3936,6 +3962,7 @@ void account_system_time(struct task_str
+ cputime_t cputime)
+ {
+ struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat;
++ struct vx_info *vxi = p->vx_info; /* p is _always_ current */
+ struct rq *rq = this_rq();
+ cputime64_t tmp;
+
+@@ -3945,6 +3972,7 @@ void account_system_time(struct task_str
+ }
+
+ p->stime = cputime_add(p->stime, cputime);
++ vx_account_system(vxi, cputime, (p == rq->idle));
+
+ /* Add system time to cpustat. */
+ tmp = cputime_to_cputime64(cputime);
+@@ -4656,7 +4684,7 @@ asmlinkage long sys_nice(int increment)
+ nice = 19;
+
+ if (increment < 0 && !can_nice(current, nice))
+- return -EPERM;
++ return vx_flags(VXF_IGNEG_NICE, 0) ? 0 : -EPERM;
+
+ retval = security_task_setnice(current, nice);
+ if (retval)
+--- a/kernel/sched_fair.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/sched_fair.c 2008-07-17 21:17:35.000000000 -0400
+@@ -637,6 +637,9 @@ enqueue_entity(struct cfs_rq *cfs_rq, st
+ check_spread(cfs_rq, se);
+ if (se != cfs_rq->curr)
+ __enqueue_entity(cfs_rq, se);
++
++ if (entity_is_task(se))
++ vx_activate_task(task_of(se));
+ }
+
+ static void update_avg(u64 *avg, u64 sample)
+@@ -679,6 +682,8 @@ dequeue_entity(struct cfs_rq *cfs_rq, st
+
+ if (se != cfs_rq->curr)
+ __dequeue_entity(cfs_rq, se);
++ if (entity_is_task(se))
++ vx_deactivate_task(task_of(se));
+ account_entity_dequeue(cfs_rq, se);
+ }
+
+--- a/kernel/sched_hard.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/sched_hard.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,324 @@
++
++#ifdef CONFIG_VSERVER_IDLELIMIT
++
++/*
++ * vx_idle_resched - reschedule after maxidle
++ */
++static inline
++void vx_idle_resched(struct rq *rq)
++{
++ /* maybe have a better criterion for paused */
++ if (!--rq->idle_tokens && !list_empty(&rq->hold_queue))
++ set_need_resched();
++}
++
++#else /* !CONFIG_VSERVER_IDLELIMIT */
++
++#define vx_idle_resched(rq)
++
++#endif /* CONFIG_VSERVER_IDLELIMIT */
++
++
++
++#ifdef CONFIG_VSERVER_IDLETIME
++
++#define vx_set_rq_min_skip(rq, min) \
++ (rq)->idle_skip = (min)
++
++#define vx_save_min_skip(ret, min, val) \
++ __vx_save_min_skip(ret, min, val)
++
++static inline
++void __vx_save_min_skip(int ret, int *min, int val)
++{
++ if (ret > -2)
++ return;
++ if ((*min > val) || !*min)
++ *min = val;
++}
++
++static inline
++int vx_try_skip(struct rq *rq, int cpu)
++{
++ /* artificially advance time */
++ if (rq->idle_skip > 0) {
++ vxdprintk(list_empty(&rq->hold_queue),
++ "hold queue empty on cpu %d", cpu);
++ rq->idle_time += rq->idle_skip;
++ vxm_idle_skip(rq, cpu);
++ return 1;
++ }
++ return 0;
++}
++
++#else /* !CONFIG_VSERVER_IDLETIME */
++
++#define vx_set_rq_min_skip(rq, min) \
++ ({ int dummy = (min); dummy; })
++
++#define vx_save_min_skip(ret, min, val)
++
++static inline
++int vx_try_skip(struct rq *rq, int cpu)
++{
++ return 0;
++}
++
++#endif /* CONFIG_VSERVER_IDLETIME */
++
++
++
++#ifdef CONFIG_VSERVER_HARDCPU
++
++#define vx_set_rq_max_idle(rq, max) \
++ (rq)->idle_tokens = (max)
++
++#define vx_save_max_idle(ret, min, val) \
++ __vx_save_max_idle(ret, min, val)
++
++static inline
++void __vx_save_max_idle(int ret, int *min, int val)
++{
++ if (*min > val)
++ *min = val;
++}
++
++
++/*
++ * vx_hold_task - put a task on the hold queue
++ */
++static inline
++void vx_hold_task(struct task_struct *p, struct rq *rq)
++{
++ __deactivate_task(p, rq);
++ p->state |= TASK_ONHOLD;
++ /* a new one on hold */
++ rq->nr_onhold++;
++ vxm_hold_task(p, rq);
++ list_add_tail(&p->run_list, &rq->hold_queue);
++}
++
++/*
++ * vx_unhold_task - put a task back to the runqueue
++ */
++static inline
++void vx_unhold_task(struct task_struct *p, struct rq *rq)
++{
++ list_del(&p->run_list);
++ /* one less waiting */
++ rq->nr_onhold--;
++ p->state &= ~TASK_ONHOLD;
++ enqueue_task(p, rq->expired);
++ inc_nr_running(p, rq);
++ vxm_unhold_task(p, rq);
++
++ if (p->static_prio < rq->best_expired_prio)
++ rq->best_expired_prio = p->static_prio;
++}
++
++unsigned long nr_onhold(void)
++{
++ unsigned long i, sum = 0;
++
++ for_each_online_cpu(i)
++ sum += cpu_rq(i)->nr_onhold;
++
++ return sum;
++}
++
++
++
++static inline
++int __vx_tokens_avail(struct _vx_sched_pc *sched_pc)
++{
++ return sched_pc->tokens;
++}
++
++static inline
++void __vx_consume_token(struct _vx_sched_pc *sched_pc)
++{
++ sched_pc->tokens--;
++}
++
++static inline
++int vx_need_resched(struct task_struct *p, int slice, int cpu)
++{
++ struct vx_info *vxi = p->vx_info;
++
++ if (vx_info_flags(vxi, VXF_SCHED_HARD|VXF_SCHED_PRIO, 0)) {
++ struct _vx_sched_pc *sched_pc =
++ &vx_per_cpu(vxi, sched_pc, cpu);
++ int tokens;
++
++ /* maybe we can simplify that to decrement
++ the token counter unconditional? */
++
++ if ((tokens = __vx_tokens_avail(sched_pc)) > 0)
++ __vx_consume_token(sched_pc);
++
++ /* for tokens > 0, one token was consumed */
++ if (tokens < 2)
++ slice = 0;
++ }
++ vxm_need_resched(p, slice, cpu);
++ return (slice == 0);
++}
++
++
++#define vx_set_rq_time(rq, time) do { \
++ rq->norm_time = time; \
++} while (0)
++
++
++static inline
++void vx_try_unhold(struct rq *rq, int cpu)
++{
++ struct vx_info *vxi = NULL;
++ struct list_head *l, *n;
++ int maxidle = HZ;
++ int minskip = 0;
++
++ /* nothing to do? what about pause? */
++ if (list_empty(&rq->hold_queue))
++ return;
++
++ list_for_each_safe(l, n, &rq->hold_queue) {
++ int ret, delta_min[2];
++ struct _vx_sched_pc *sched_pc;
++ struct task_struct *p;
++
++ p = list_entry(l, struct task_struct, run_list);
++ /* don't bother with same context */
++ if (vxi == p->vx_info)
++ continue;
++
++ vxi = p->vx_info;
++ /* ignore paused contexts */
++ if (vx_info_flags(vxi, VXF_SCHED_PAUSE, 0))
++ continue;
++
++ sched_pc = &vx_per_cpu(vxi, sched_pc, cpu);
++
++ /* recalc tokens */
++ vxm_sched_info(sched_pc, vxi, cpu);
++ ret = vx_tokens_recalc(sched_pc,
++ &rq->norm_time, &rq->idle_time, delta_min);
++ vxm_tokens_recalc(sched_pc, rq, vxi, cpu);
++
++ if (ret > 0) {
++ /* we found a runable context */
++ vx_unhold_task(p, rq);
++ break;
++ }
++ vx_save_max_idle(ret, &maxidle, delta_min[0]);
++ vx_save_min_skip(ret, &minskip, delta_min[1]);
++ }
++ vx_set_rq_max_idle(rq, maxidle);
++ vx_set_rq_min_skip(rq, minskip);
++ vxm_rq_max_min(rq, cpu);
++}
++
++
++static inline
++int vx_schedule(struct task_struct *next, struct rq *rq, int cpu)
++{
++ struct vx_info *vxi = next->vx_info;
++ struct _vx_sched_pc *sched_pc;
++ int delta_min[2];
++ int flags, ret;
++
++ if (!vxi)
++ return 1;
++
++ flags = vxi->vx_flags;
++
++ if (unlikely(vs_check_flags(flags, VXF_SCHED_PAUSE, 0)))
++ goto put_on_hold;
++ if (!vs_check_flags(flags, VXF_SCHED_HARD | VXF_SCHED_PRIO, 0))
++ return 1;
++
++ sched_pc = &vx_per_cpu(vxi, sched_pc, cpu);
++#ifdef CONFIG_SMP
++ /* update scheduler params */
++ if (cpu_isset(cpu, vxi->sched.update)) {
++ vx_update_sched_param(&vxi->sched, sched_pc);
++ vxm_update_sched(sched_pc, vxi, cpu);
++ cpu_clear(cpu, vxi->sched.update);
++ }
++#endif
++ vxm_sched_info(sched_pc, vxi, cpu);
++ ret = vx_tokens_recalc(sched_pc,
++ &rq->norm_time, &rq->idle_time, delta_min);
++ vxm_tokens_recalc(sched_pc, rq, vxi, cpu);
++
++ if (!vs_check_flags(flags, VXF_SCHED_HARD, 0))
++ return 1;
++
++ if (unlikely(ret < 0)) {
++ vx_save_max_idle(ret, &rq->idle_tokens, delta_min[0]);
++ vx_save_min_skip(ret, &rq->idle_skip, delta_min[1]);
++ vxm_rq_max_min(rq, cpu);
++ put_on_hold:
++ vx_hold_task(next, rq);
++ return 0;
++ }
++ return 1;
++}
++
++
++#else /* CONFIG_VSERVER_HARDCPU */
++
++static inline
++void vx_hold_task(struct task_struct *p, struct rq *rq)
++{
++ return;
++}
++
++static inline
++void vx_unhold_task(struct task_struct *p, struct rq *rq)
++{
++ return;
++}
++
++unsigned long nr_onhold(void)
++{
++ return 0;
++}
++
++
++static inline
++int vx_need_resched(struct task_struct *p, int slice, int cpu)
++{
++ return (slice == 0);
++}
++
++
++#define vx_set_rq_time(rq, time)
++
++static inline
++void vx_try_unhold(struct rq *rq, int cpu)
++{
++ return;
++}
++
++static inline
++int vx_schedule(struct task_struct *next, struct rq *rq, int cpu)
++{
++ struct vx_info *vxi = next->vx_info;
++ struct _vx_sched_pc *sched_pc;
++ int delta_min[2];
++ int ret;
++
++ if (!vx_info_flags(vxi, VXF_SCHED_PRIO, 0))
++ return 1;
++
++ sched_pc = &vx_per_cpu(vxi, sched_pc, cpu);
++ vxm_sched_info(sched_pc, vxi, cpu);
++ ret = vx_tokens_recalc(sched_pc,
++ &rq->norm_time, &rq->idle_time, delta_min);
++ vxm_tokens_recalc(sched_pc, rq, vxi, cpu);
++ return 1;
++}
++
++#endif /* CONFIG_VSERVER_HARDCPU */
++
+--- a/kernel/sched_mon.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/sched_mon.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,200 @@
++
++#include <linux/vserver/monitor.h>
++
++#ifdef CONFIG_VSERVER_MONITOR
++
++#ifdef CONFIG_VSERVER_HARDCPU
++#define HARDCPU(x) (x)
++#else
++#define HARDCPU(x) (0)
++#endif
++
++#ifdef CONFIG_VSERVER_IDLETIME
++#define IDLETIME(x) (x)
++#else
++#define IDLETIME(x) (0)
++#endif
++
++struct _vx_mon_entry *vxm_advance(int cpu);
++
++
++static inline
++void __vxm_basic(struct _vx_mon_entry *entry, xid_t xid, int type)
++{
++ entry->type = type;
++ entry->xid = xid;
++}
++
++static inline
++void __vxm_sync(int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ __vxm_basic(entry, 0, VXM_SYNC);
++ entry->ev.sec = xtime.tv_sec;
++ entry->ev.nsec = xtime.tv_nsec;
++}
++
++static inline
++void __vxm_task(struct task_struct *p, int type)
++{
++ struct _vx_mon_entry *entry = vxm_advance(task_cpu(p));
++
++ __vxm_basic(entry, p->xid, type);
++ entry->ev.tsk.pid = p->pid;
++ entry->ev.tsk.state = p->state;
++}
++
++static inline
++void __vxm_sched(struct _vx_sched_pc *s, struct vx_info *vxi, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ __vxm_basic(entry, vxi->vx_id, (VXM_SCHED | s->flags));
++ entry->sd.tokens = s->tokens;
++ entry->sd.norm_time = s->norm_time;
++ entry->sd.idle_time = s->idle_time;
++}
++
++static inline
++void __vxm_rqinfo1(struct rq *q, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ entry->type = VXM_RQINFO_1;
++ entry->xid = ((unsigned long)q >> 16) & 0xffff;
++ entry->q1.running = q->nr_running;
++ entry->q1.onhold = HARDCPU(q->nr_onhold);
++ entry->q1.iowait = atomic_read(&q->nr_iowait);
++ entry->q1.uintr = q->nr_uninterruptible;
++ entry->q1.idle_tokens = IDLETIME(q->idle_tokens);
++}
++
++static inline
++void __vxm_rqinfo2(struct rq *q, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ entry->type = VXM_RQINFO_2;
++ entry->xid = (unsigned long)q & 0xffff;
++ entry->q2.norm_time = q->norm_time;
++ entry->q2.idle_time = q->idle_time;
++ entry->q2.idle_skip = IDLETIME(q->idle_skip);
++}
++
++static inline
++void __vxm_update(struct _vx_sched_pc *s, struct vx_info *vxi, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ __vxm_basic(entry, vxi->vx_id, VXM_UPDATE);
++ entry->ev.tokens = s->tokens;
++}
++
++static inline
++void __vxm_update1(struct _vx_sched_pc *s, struct vx_info *vxi, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ __vxm_basic(entry, vxi->vx_id, VXM_UPDATE_1);
++ entry->u1.tokens_max = s->tokens_max;
++ entry->u1.fill_rate = s->fill_rate[0];
++ entry->u1.interval = s->interval[0];
++}
++
++static inline
++void __vxm_update2(struct _vx_sched_pc *s, struct vx_info *vxi, int cpu)
++{
++ struct _vx_mon_entry *entry = vxm_advance(cpu);
++
++ __vxm_basic(entry, vxi->vx_id, VXM_UPDATE_2);
++ entry->u2.tokens_min = s->tokens_min;
++ entry->u2.fill_rate = s->fill_rate[1];
++ entry->u2.interval = s->interval[1];
++}
++
++
++#define vxm_activate_task(p,q) __vxm_task(p, VXM_ACTIVATE)
++#define vxm_activate_idle(p,q) __vxm_task(p, VXM_IDLE)
++#define vxm_deactivate_task(p,q) __vxm_task(p, VXM_DEACTIVATE)
++#define vxm_hold_task(p,q) __vxm_task(p, VXM_HOLD)
++#define vxm_unhold_task(p,q) __vxm_task(p, VXM_UNHOLD)
++
++static inline
++void vxm_migrate_task(struct task_struct *p, struct rq *rq, int dest)
++{
++ __vxm_task(p, VXM_MIGRATE);
++ __vxm_rqinfo1(rq, task_cpu(p));
++ __vxm_rqinfo2(rq, task_cpu(p));
++}
++
++static inline
++void vxm_idle_skip(struct rq *rq, int cpu)
++{
++ __vxm_rqinfo1(rq, cpu);
++ __vxm_rqinfo2(rq, cpu);
++}
++
++static inline
++void vxm_need_resched(struct task_struct *p, int slice, int cpu)
++{
++ if (slice)
++ return;
++
++ __vxm_task(p, VXM_RESCHED);
++}
++
++static inline
++void vxm_sync(unsigned long now, int cpu)
++{
++ if (!CONFIG_VSERVER_MONITOR_SYNC ||
++ (now % CONFIG_VSERVER_MONITOR_SYNC))
++ return;
++
++ __vxm_sync(cpu);
++}
++
++#define vxm_sched_info(s,v,c) __vxm_sched(s,v,c)
++
++static inline
++void vxm_tokens_recalc(struct _vx_sched_pc *s, struct rq *rq,
++ struct vx_info *vxi, int cpu)
++{
++ __vxm_sched(s, vxi, cpu);
++ __vxm_rqinfo2(rq, cpu);
++}
++
++static inline
++void vxm_update_sched(struct _vx_sched_pc *s, struct vx_info *vxi, int cpu)
++{
++ __vxm_sched(s, vxi, cpu);
++ __vxm_update(s, vxi, cpu);
++ __vxm_update1(s, vxi, cpu);
++ __vxm_update2(s, vxi, cpu);
++}
++
++static inline
++void vxm_rq_max_min(struct rq *rq, int cpu)
++{
++ __vxm_rqinfo1(rq, cpu);
++ __vxm_rqinfo2(rq, cpu);
++}
++
++#else /* CONFIG_VSERVER_MONITOR */
++
++#define vxm_activate_task(t,q) do { } while (0)
++#define vxm_activate_idle(t,q) do { } while (0)
++#define vxm_deactivate_task(t,q) do { } while (0)
++#define vxm_hold_task(t,q) do { } while (0)
++#define vxm_unhold_task(t,q) do { } while (0)
++#define vxm_migrate_task(t,q,d) do { } while (0)
++#define vxm_idle_skip(q,c) do { } while (0)
++#define vxm_need_resched(t,s,c) do { } while (0)
++#define vxm_sync(s,c) do { } while (0)
++#define vxm_sched_info(s,v,c) do { } while (0)
++#define vxm_tokens_recalc(s,q,v,c) do { } while (0)
++#define vxm_update_sched(s,v,c) do { } while (0)
++#define vxm_rq_max_min(q,c) do { } while (0)
++
++#endif /* CONFIG_VSERVER_MONITOR */
++
+--- a/kernel/signal.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/signal.c 2008-07-17 21:15:33.000000000 -0400
+@@ -26,6 +26,8 @@
+ #include <linux/freezer.h>
+ #include <linux/pid_namespace.h>
+ #include <linux/nsproxy.h>
++#include <linux/vs_context.h>
++#include <linux/vs_pid.h>
+
+ #include <asm/param.h>
+ #include <asm/uaccess.h>
+@@ -573,6 +575,14 @@ static int check_kill_permission(int sig
+ if (!valid_signal(sig))
+ return -EINVAL;
+
++ if ((info != SEND_SIG_NOINFO) &&
++ (is_si_special(info) || !SI_FROMUSER(info)))
++ goto skip;
++
++ vxdprintk(VXD_CBIT(misc, 7),
++ "check_kill_permission(%d,%p,%p[#%u,%u])",
++ sig, info, t, vx_task_xid(t), t->pid);
++
+ if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info)))
+ return 0;
+
+@@ -597,6 +607,18 @@ static int check_kill_permission(int sig
+ }
+ }
+
++ error = -EPERM;
++ if (t->pid == 1 && current->xid)
++ return error;
++
++ error = -ESRCH;
++ if (!vx_check(vx_task_xid(t), VS_WATCH_P | VS_IDENT)) {
++ vxdprintk(current->xid || VXD_CBIT(misc, 7),
++ "signal %d[%p] xid mismatch %p[#%u,%u] xid=#%u",
++ sig, info, t, vx_task_xid(t), t->pid, current->xid);
++ return error;
++ }
++skip:
+ return security_task_kill(t, info, sig, 0);
+ }
+
+@@ -1055,7 +1077,7 @@ int kill_pid_info(int sig, struct siginf
+ rcu_read_lock();
+ retry:
+ p = pid_task(pid, PIDTYPE_PID);
+- if (p) {
++ if (p && vx_check(vx_task_xid(p), VS_WATCH | VS_IDENT)) {
+ error = group_send_sig_info(sig, info, p);
+ if (unlikely(error == -ESRCH))
+ /*
+@@ -1093,7 +1115,7 @@ int kill_pid_info_as_uid(int sig, struct
+
+ read_lock(&tasklist_lock);
+ p = pid_task(pid, PIDTYPE_PID);
+- if (!p) {
++ if (!p || !vx_check(vx_task_xid(p), VS_IDENT)) {
+ ret = -ESRCH;
+ goto out_unlock;
+ }
+@@ -1145,7 +1167,9 @@ static int kill_something_info(int sig,
+ struct task_struct * p;
+
+ for_each_process(p) {
+- if (p->pid > 1 && !same_thread_group(p, current)) {
++ if (vx_check(vx_task_xid(p), VS_ADMIN|VS_IDENT) &&
++ p->pid > 1 && !same_thread_group(p, current) &&
++ !vx_current_initpid(p->pid)) {
+ int err = group_send_sig_info(sig, info, p);
+ ++count;
+ if (err != -EPERM)
+@@ -1809,6 +1833,11 @@ relock:
+ !signal_group_exit(signal))
+ continue;
+
++ /* virtual init is protected against user signals */
++ if ((info->si_code == SI_USER) &&
++ vx_current_initpid(current->pid))
++ continue;
++
+ if (sig_kernel_stop(signr)) {
+ /*
+ * The default action is to stop all threads in
+--- a/kernel/softirq.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/softirq.c 2008-07-16 22:41:36.000000000 -0400
+@@ -21,6 +21,7 @@
+ #include <linux/rcupdate.h>
+ #include <linux/smp.h>
+ #include <linux/tick.h>
++#include <linux/vs_context.h>
+
+ #include <asm/irq.h>
+ /*
+--- a/kernel/sys.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/sys.c 2008-07-16 22:41:36.000000000 -0400
+@@ -38,6 +38,7 @@
+ #include <linux/syscalls.h>
+ #include <linux/kprobes.h>
+ #include <linux/user_namespace.h>
++#include <linux/vs_pid.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/io.h>
+@@ -122,7 +123,10 @@ static int set_one_prio(struct task_stru
+ goto out;
+ }
+ if (niceval < task_nice(p) && !can_nice(p, niceval)) {
+- error = -EACCES;
++ if (vx_flags(VXF_IGNEG_NICE, 0))
++ error = 0;
++ else
++ error = -EACCES;
+ goto out;
+ }
+ no_nice = security_task_setnice(p, niceval);
+@@ -170,6 +174,8 @@ asmlinkage long sys_setpriority(int whic
+ else
+ pgrp = task_pgrp(current);
+ do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT))
++ continue;
+ error = set_one_prio(p, niceval, error);
+ } while_each_pid_task(pgrp, PIDTYPE_PGID, p);
+ break;
+@@ -230,6 +236,8 @@ asmlinkage long sys_getpriority(int whic
+ else
+ pgrp = task_pgrp(current);
+ do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
++ if (!vx_check(p->xid, VS_ADMIN_P | VS_IDENT))
++ continue;
+ niceval = 20 - task_nice(p);
+ if (niceval > retval)
+ retval = niceval;
+@@ -359,6 +367,9 @@ void kernel_power_off(void)
+ machine_power_off();
+ }
+ EXPORT_SYMBOL_GPL(kernel_power_off);
++
++long vs_reboot(unsigned int, void __user *);
++
+ /*
+ * Reboot system call: for obvious reasons only root may call it,
+ * and even root needs to set up some magic numbers in the registers
+@@ -389,6 +400,9 @@ asmlinkage long sys_reboot(int magic1, i
+ if ((cmd == LINUX_REBOOT_CMD_POWER_OFF) && !pm_power_off)
+ cmd = LINUX_REBOOT_CMD_HALT;
+
++ if (!vx_check(0, VS_ADMIN|VS_WATCH))
++ return vs_reboot(cmd, arg);
++
+ lock_kernel();
+ switch (cmd) {
+ case LINUX_REBOOT_CMD_RESTART:
+@@ -1361,7 +1375,7 @@ asmlinkage long sys_sethostname(char __u
+ int errno;
+ char tmp[__NEW_UTS_LEN];
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SET_UTSNAME))
+ return -EPERM;
+ if (len < 0 || len > __NEW_UTS_LEN)
+ return -EINVAL;
+@@ -1406,7 +1420,7 @@ asmlinkage long sys_setdomainname(char _
+ int errno;
+ char tmp[__NEW_UTS_LEN];
+
+- if (!capable(CAP_SYS_ADMIN))
++ if (!vx_capable(CAP_SYS_ADMIN, VXC_SET_UTSNAME))
+ return -EPERM;
+ if (len < 0 || len > __NEW_UTS_LEN)
+ return -EINVAL;
+@@ -1473,7 +1487,7 @@ asmlinkage long sys_setrlimit(unsigned i
+ return -EINVAL;
+ old_rlim = current->signal->rlim + resource;
+ if ((new_rlim.rlim_max > old_rlim->rlim_max) &&
+- !capable(CAP_SYS_RESOURCE))
++ !vx_capable(CAP_SYS_RESOURCE, VXC_SET_RLIMIT))
+ return -EPERM;
+ if (resource == RLIMIT_NOFILE && new_rlim.rlim_max > sysctl_nr_open)
+ return -EPERM;
+--- a/kernel/sysctl.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/sysctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -109,6 +109,7 @@ static int ngroups_max = NGROUPS_MAX;
+ #ifdef CONFIG_KMOD
+ extern char modprobe_path[];
+ #endif
++extern char vshelper_path[];
+ #ifdef CONFIG_CHR_DEV_SG
+ extern int sg_big_buff;
+ #endif
+@@ -477,6 +478,15 @@ static struct ctl_table kern_table[] = {
+ .strategy = &sysctl_string,
+ },
+ #endif
++ {
++ .ctl_name = KERN_VSHELPER,
++ .procname = "vshelper",
++ .data = &vshelper_path,
++ .maxlen = 256,
++ .mode = 0644,
++ .proc_handler = &proc_dostring,
++ .strategy = &sysctl_string,
++ },
+ #ifdef CONFIG_CHR_DEV_SG
+ {
+ .ctl_name = KERN_SG_BIG_BUFF,
+--- a/kernel/sysctl_check.c 2008-04-17 12:05:44.000000000 -0400
++++ a/kernel/sysctl_check.c 2008-07-16 22:41:36.000000000 -0400
+@@ -39,6 +39,7 @@ static const struct trans_ctl_table tran
+
+ { KERN_PANIC, "panic" },
+ { KERN_REALROOTDEV, "real-root-dev" },
++ { KERN_VSHELPER, "vshelper", },
+
+ { KERN_SPARC_REBOOT, "reboot-cmd" },
+ { KERN_CTLALTDEL, "ctrl-alt-del" },
+@@ -1217,6 +1218,22 @@ static const struct trans_ctl_table tran
+ {}
+ };
+
++static struct trans_ctl_table trans_vserver_table[] = {
++ { 1, "debug_switch" },
++ { 2, "debug_xid" },
++ { 3, "debug_nid" },
++ { 4, "debug_tag" },
++ { 5, "debug_net" },
++ { 6, "debug_limit" },
++ { 7, "debug_cres" },
++ { 8, "debug_dlim" },
++ { 9, "debug_quota" },
++ { 10, "debug_cvirt" },
++ { 11, "debug_space" },
++ { 12, "debug_misc" },
++ {}
++};
++
+ static const struct trans_ctl_table trans_root_table[] = {
+ { CTL_KERN, "kernel", trans_kern_table },
+ { CTL_VM, "vm", trans_vm_table },
+@@ -1233,6 +1250,7 @@ static const struct trans_ctl_table tran
+ { CTL_SUNRPC, "sunrpc", trans_sunrpc_table },
+ { CTL_PM, "pm", trans_pm_table },
+ { CTL_FRV, "frv", trans_frv_table },
++ { CTL_VSERVER, "vserver", trans_vserver_table },
+ {}
+ };
+
+--- a/kernel/time.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/time.c 2008-07-16 22:41:36.000000000 -0400
+@@ -62,6 +62,7 @@ EXPORT_SYMBOL(sys_tz);
+ asmlinkage long sys_time(time_t __user * tloc)
+ {
+ time_t i = get_seconds();
++/* FIXME: do_gettimeofday(&tv) -> vx_gettimeofday(&tv) */
+
+ if (tloc) {
+ if (put_user(i,tloc))
+@@ -91,7 +92,7 @@ asmlinkage long sys_stime(time_t __user
+ if (err)
+ return err;
+
+- do_settimeofday(&tv);
++ vx_settimeofday(&tv);
+ return 0;
+ }
+
+@@ -102,7 +103,7 @@ asmlinkage long sys_gettimeofday(struct
+ {
+ if (likely(tv != NULL)) {
+ struct timeval ktv;
+- do_gettimeofday(&ktv);
++ vx_gettimeofday(&ktv);
+ if (copy_to_user(tv, &ktv, sizeof(ktv)))
+ return -EFAULT;
+ }
+@@ -177,7 +178,7 @@ int do_sys_settimeofday(struct timespec
+ /* SMP safe, again the code in arch/foo/time.c should
+ * globally block out interrupts when it runs.
+ */
+- return do_settimeofday(tv);
++ return vx_settimeofday(tv);
+ }
+ return 0;
+ }
+@@ -309,7 +310,7 @@ void getnstimeofday(struct timespec *tv)
+ {
+ struct timeval x;
+
+- do_gettimeofday(&x);
++ vx_gettimeofday(&x);
+ tv->tv_sec = x.tv_sec;
+ tv->tv_nsec = x.tv_usec * NSEC_PER_USEC;
+ }
+--- a/kernel/timer.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/timer.c 2008-07-16 22:41:36.000000000 -0400
+@@ -37,6 +37,10 @@
+ #include <linux/delay.h>
+ #include <linux/tick.h>
+ #include <linux/kallsyms.h>
++#include <linux/vs_base.h>
++#include <linux/vs_cvirt.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/sched.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/unistd.h>
+@@ -1089,12 +1093,6 @@ asmlinkage unsigned long sys_alarm(unsig
+
+ #endif
+
+-#ifndef __alpha__
+-
+-/*
+- * The Alpha uses getxpid, getxuid, and getxgid instead. Maybe this
+- * should be moved into arch/i386 instead?
+- */
+
+ /**
+ * sys_getpid - return the thread group id of the current process
+@@ -1123,10 +1121,23 @@ asmlinkage long sys_getppid(void)
+ rcu_read_lock();
+ pid = task_tgid_vnr(current->real_parent);
+ rcu_read_unlock();
++ return vx_map_pid(pid);
++}
+
+- return pid;
++#ifdef __alpha__
++
++/*
++ * The Alpha uses getxpid, getxuid, and getxgid instead.
++ */
++
++asmlinkage long do_getxpid(long *ppid)
++{
++ *ppid = sys_getppid();
++ return sys_getpid();
+ }
+
++#else /* _alpha_ */
++
+ asmlinkage long sys_getuid(void)
+ {
+ /* Only we change this so SMP safe */
+@@ -1297,6 +1308,8 @@ int do_sysinfo(struct sysinfo *info)
+ tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
+ tp.tv_sec++;
+ }
++ if (vx_flags(VXF_VIRT_UPTIME, 0))
++ vx_vsi_uptime(&tp, NULL);
+ info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
+
+ info->loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT);
+--- a/kernel/user.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/user.c 2008-08-13 20:16:50.000000000 -0400
+@@ -243,12 +243,15 @@ static struct kobj_type uids_ktype = {
+ };
+
+ /* create /sys/kernel/uids/<uid>/cpu_share file for this user */
+-static int uids_user_create(struct user_struct *up)
++static int uids_user_create(struct user_namespace *ns, struct user_struct *up)
+ {
+ struct kobject *kobj = &up->kobj;
+- int error;
++ int error = 0;
+
+ memset(kobj, 0, sizeof(struct kobject));
++ if (ns != &init_user_ns)
++ goto done;
++
+ kobj->kset = uids_kset;
+ error = kobject_init_and_add(kobj, &uids_ktype, NULL, "%d", up->uid);
+ if (error) {
+@@ -272,7 +275,7 @@ int __init uids_sysfs_init(void)
+ if (!uids_kset)
+ return -ENOMEM;
+
+- return uids_user_create(&root_user);
++ return uids_user_create(NULL, &root_user);
+ }
+
+ /* work function to remove sysfs directory for a user and free up
+@@ -302,9 +305,11 @@ static void remove_user_sysfs_dir(struct
+ if (!remove_user)
+ goto done;
+
+- kobject_uevent(&up->kobj, KOBJ_REMOVE);
+- kobject_del(&up->kobj);
+- kobject_put(&up->kobj);
++ if (up->kobj.name) {
++ kobject_uevent(&up->kobj, KOBJ_REMOVE);
++ kobject_del(&up->kobj);
++ kobject_put(&up->kobj);
++ }
+
+ sched_destroy_user(up);
+ key_put(up->uid_keyring);
+@@ -332,7 +337,8 @@ static inline void free_user(struct user
+ #else /* CONFIG_USER_SCHED && CONFIG_SYSFS */
+
+ int uids_sysfs_init(void) { return 0; }
+-static inline int uids_user_create(struct user_struct *up) { return 0; }
++static inline int uids_user_create(struct user_namespace *ns,
++ struct user_struct *up) { return 0; }
+ static inline void uids_mutex_lock(void) { }
+ static inline void uids_mutex_unlock(void) { }
+
+@@ -409,7 +415,7 @@ struct user_struct *alloc_uid(struct use
+ if (sched_create_user(new) < 0)
+ goto out_free_user;
+
+- if (uids_user_create(new))
++ if (uids_user_create(ns, new))
+ goto out_destoy_sched;
+
+ /*
+--- a/kernel/user_namespace.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/user_namespace.c 2008-07-16 22:41:36.000000000 -0400
+@@ -10,6 +10,7 @@
+ #include <linux/nsproxy.h>
+ #include <linux/slab.h>
+ #include <linux/user_namespace.h>
++#include <linux/vserver/global.h>
+
+ /*
+ * Clone a new ns copying an original user ns, setting refcount to 1
+@@ -27,6 +28,7 @@ static struct user_namespace *clone_user
+ return ERR_PTR(-ENOMEM);
+
+ kref_init(&ns->kref);
++ atomic_inc(&vs_global_user_ns);
+
+ for (n = 0; n < UIDHASH_SZ; ++n)
+ INIT_HLIST_HEAD(ns->uidhash_table + n);
+@@ -72,6 +74,7 @@ void free_user_ns(struct kref *kref)
+
+ ns = container_of(kref, struct user_namespace, kref);
+ release_uids(ns);
++ atomic_dec(&vs_global_user_ns);
+ kfree(ns);
+ }
+ EXPORT_SYMBOL(free_user_ns);
+--- a/kernel/utsname.c 2008-07-14 17:22:55.000000000 -0400
++++ a/kernel/utsname.c 2008-07-17 21:15:55.000000000 -0400
+@@ -15,6 +15,7 @@
+ #include <linux/version.h>
+ #include <linux/err.h>
+ #include <linux/slab.h>
++#include <linux/vserver/global.h>
+
+ /*
+ * Clone a new ns copying an original utsname, setting refcount to 1
+@@ -33,6 +34,7 @@ static struct uts_namespace *clone_uts_n
+ memcpy(&ns->name, &old_ns->name, sizeof(ns->name));
+ up_read(&uts_sem);
+ kref_init(&ns->kref);
++ atomic_inc(&vs_global_uts_ns);
+ return ns;
+ }
+
+@@ -63,5 +65,6 @@ void free_uts_ns(struct kref *kref)
+ struct uts_namespace *ns;
+
+ ns = container_of(kref, struct uts_namespace, kref);
++ atomic_dec(&vs_global_uts_ns);
+ kfree(ns);
+ }
+--- a/kernel/vserver/cacct.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cacct.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,42 @@
++/*
++ * linux/kernel/vserver/cacct.c
++ *
++ * Virtual Server: Context Accounting
++ *
++ * Copyright (C) 2006-2007 Herbert Pötzl
++ *
++ * V0.01 added accounting stats
++ *
++ */
++
++#include <linux/types.h>
++#include <linux/vs_context.h>
++#include <linux/vserver/cacct_cmd.h>
++#include <linux/vserver/cacct_int.h>
++
++#include <asm/errno.h>
++#include <asm/uaccess.h>
++
++
++int vc_sock_stat(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_sock_stat_v0 vc_data;
++ int j, field;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ field = vc_data.field;
++ if ((field < 0) || (field >= VXA_SOCK_SIZE))
++ return -EINVAL;
++
++ for (j = 0; j < 3; j++) {
++ vc_data.count[j] = vx_sock_count(&vxi->cacct, field, j);
++ vc_data.total[j] = vx_sock_total(&vxi->cacct, field, j);
++ }
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
+--- a/kernel/vserver/cacct_init.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cacct_init.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,25 @@
++
++
++static inline void vx_info_init_cacct(struct _vx_cacct *cacct)
++{
++ int i, j;
++
++
++ for (i = 0; i < VXA_SOCK_SIZE; i++) {
++ for (j = 0; j < 3; j++) {
++ atomic_set(&cacct->sock[i][j].count, 0);
++ atomic_set(&cacct->sock[i][j].total, 0);
++ }
++ }
++ for (i = 0; i < 8; i++)
++ atomic_set(&cacct->slab[i], 0);
++ for (i = 0; i < 5; i++)
++ for (j = 0; j < 4; j++)
++ atomic_set(&cacct->page[i][j], 0);
++}
++
++static inline void vx_info_exit_cacct(struct _vx_cacct *cacct)
++{
++ return;
++}
++
+--- a/kernel/vserver/cacct_proc.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cacct_proc.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,53 @@
++#ifndef _VX_CACCT_PROC_H
++#define _VX_CACCT_PROC_H
++
++#include <linux/vserver/cacct_int.h>
++
++
++#define VX_SOCKA_TOP \
++ "Type\t recv #/bytes\t\t send #/bytes\t\t fail #/bytes\n"
++
++static inline int vx_info_proc_cacct(struct _vx_cacct *cacct, char *buffer)
++{
++ int i, j, length = 0;
++ static char *type[VXA_SOCK_SIZE] = {
++ "UNSPEC", "UNIX", "INET", "INET6", "PACKET", "OTHER"
++ };
++
++ length += sprintf(buffer + length, VX_SOCKA_TOP);
++ for (i = 0; i < VXA_SOCK_SIZE; i++) {
++ length += sprintf(buffer + length, "%s:", type[i]);
++ for (j = 0; j < 3; j++) {
++ length += sprintf(buffer + length,
++ "\t%10lu/%-10lu",
++ vx_sock_count(cacct, i, j),
++ vx_sock_total(cacct, i, j));
++ }
++ buffer[length++] = '\n';
++ }
++
++ length += sprintf(buffer + length, "\n");
++ length += sprintf(buffer + length,
++ "slab:\t %8u %8u %8u %8u\n",
++ atomic_read(&cacct->slab[1]),
++ atomic_read(&cacct->slab[4]),
++ atomic_read(&cacct->slab[0]),
++ atomic_read(&cacct->slab[2]));
++
++ length += sprintf(buffer + length, "\n");
++ for (i = 0; i < 5; i++) {
++ length += sprintf(buffer + length,
++ "page[%d]: %8u %8u %8u %8u\t %8u %8u %8u %8u\n", i,
++ atomic_read(&cacct->page[i][0]),
++ atomic_read(&cacct->page[i][1]),
++ atomic_read(&cacct->page[i][2]),
++ atomic_read(&cacct->page[i][3]),
++ atomic_read(&cacct->page[i][4]),
++ atomic_read(&cacct->page[i][5]),
++ atomic_read(&cacct->page[i][6]),
++ atomic_read(&cacct->page[i][7]));
++ }
++ return length;
++}
++
++#endif /* _VX_CACCT_PROC_H */
+--- a/kernel/vserver/context.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/context.c 2008-07-27 14:10:37.000000000 -0400
+@@ -0,0 +1,1010 @@
++/*
++ * linux/kernel/vserver/context.c
++ *
++ * Virtual Server: Context Support
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 context helper
++ * V0.02 vx_ctx_kill syscall command
++ * V0.03 replaced context_info calls
++ * V0.04 redesign of struct (de)alloc
++ * V0.05 rlimit basic implementation
++ * V0.06 task_xid and info commands
++ * V0.07 context flags and caps
++ * V0.08 switch to RCU based hash
++ * V0.09 revert to non RCU for now
++ * V0.10 and back to working RCU hash
++ * V0.11 and back to locking again
++ * V0.12 referenced context store
++ * V0.13 separate per cpu data
++ * V0.14 changed vcmds to vxi arg
++ * V0.15 added context stat
++ * V0.16 have __create claim() the vxi
++ * V0.17 removed older and legacy stuff
++ *
++ */
++
++#include <linux/slab.h>
++#include <linux/types.h>
++#include <linux/security.h>
++#include <linux/pid_namespace.h>
++
++#include <linux/vserver/context.h>
++#include <linux/vserver/network.h>
++#include <linux/vserver/debug.h>
++#include <linux/vserver/limit.h>
++#include <linux/vserver/limit_int.h>
++#include <linux/vserver/space.h>
++
++#include <linux/vs_context.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/context_cmd.h>
++
++#include "cvirt_init.h"
++#include "cacct_init.h"
++#include "limit_init.h"
++#include "sched_init.h"
++
++
++atomic_t vx_global_ctotal = ATOMIC_INIT(0);
++atomic_t vx_global_cactive = ATOMIC_INIT(0);
++
++
++/* now inactive context structures */
++
++static struct hlist_head vx_info_inactive = HLIST_HEAD_INIT;
++
++static spinlock_t vx_info_inactive_lock = SPIN_LOCK_UNLOCKED;
++
++
++/* __alloc_vx_info()
++
++ * allocate an initialized vx_info struct
++ * doesn't make it visible (hash) */
++
++static struct vx_info *__alloc_vx_info(xid_t xid)
++{
++ struct vx_info *new = NULL;
++ int cpu;
++
++ vxdprintk(VXD_CBIT(xid, 0), "alloc_vx_info(%d)*", xid);
++
++ /* would this benefit from a slab cache? */
++ new = kmalloc(sizeof(struct vx_info), GFP_KERNEL);
++ if (!new)
++ return 0;
++
++ memset(new, 0, sizeof(struct vx_info));
++#ifdef CONFIG_SMP
++ new->ptr_pc = alloc_percpu(struct _vx_info_pc);
++ if (!new->ptr_pc)
++ goto error;
++#endif
++ new->vx_id = xid;
++ INIT_HLIST_NODE(&new->vx_hlist);
++ atomic_set(&new->vx_usecnt, 0);
++ atomic_set(&new->vx_tasks, 0);
++ new->vx_parent = NULL;
++ new->vx_state = 0;
++ init_waitqueue_head(&new->vx_wait);
++
++ /* prepare reaper */
++ get_task_struct(init_pid_ns.child_reaper);
++ new->vx_reaper = init_pid_ns.child_reaper;
++ new->vx_badness_bias = 0;
++
++ /* rest of init goes here */
++ vx_info_init_limit(&new->limit);
++ vx_info_init_sched(&new->sched);
++ vx_info_init_cvirt(&new->cvirt);
++ vx_info_init_cacct(&new->cacct);
++
++ /* per cpu data structures */
++ for_each_possible_cpu(cpu) {
++ vx_info_init_sched_pc(
++ &vx_per_cpu(new, sched_pc, cpu), cpu);
++ vx_info_init_cvirt_pc(
++ &vx_per_cpu(new, cvirt_pc, cpu), cpu);
++ }
++
++ new->vx_flags = VXF_INIT_SET;
++ cap_set_init_eff(new->vx_bcaps);
++ new->vx_ccaps = 0;
++ // new->vx_cap_bset = current->cap_bset;
++
++ new->reboot_cmd = 0;
++ new->exit_code = 0;
++
++ vxdprintk(VXD_CBIT(xid, 0),
++ "alloc_vx_info(%d) = %p", xid, new);
++ vxh_alloc_vx_info(new);
++ atomic_inc(&vx_global_ctotal);
++ return new;
++#ifdef CONFIG_SMP
++error:
++ kfree(new);
++ return 0;
++#endif
++}
++
++/* __dealloc_vx_info()
++
++ * final disposal of vx_info */
++
++static void __dealloc_vx_info(struct vx_info *vxi)
++{
++ struct vx_info_save vxis;
++ int cpu;
++
++ vxdprintk(VXD_CBIT(xid, 0),
++ "dealloc_vx_info(%p)", vxi);
++ vxh_dealloc_vx_info(vxi);
++
++#ifdef CONFIG_VSERVER_WARN
++ enter_vx_info(vxi, &vxis);
++ vx_info_exit_limit(&vxi->limit);
++ vx_info_exit_sched(&vxi->sched);
++ vx_info_exit_cvirt(&vxi->cvirt);
++ vx_info_exit_cacct(&vxi->cacct);
++
++ for_each_possible_cpu(cpu) {
++ vx_info_exit_sched_pc(
++ &vx_per_cpu(vxi, sched_pc, cpu), cpu);
++ vx_info_exit_cvirt_pc(
++ &vx_per_cpu(vxi, cvirt_pc, cpu), cpu);
++ }
++ leave_vx_info(&vxis);
++#endif
++
++ vxi->vx_id = -1;
++ vxi->vx_state |= VXS_RELEASED;
++
++#ifdef CONFIG_SMP
++ free_percpu(vxi->ptr_pc);
++#endif
++ kfree(vxi);
++ atomic_dec(&vx_global_ctotal);
++}
++
++static void __shutdown_vx_info(struct vx_info *vxi)
++{
++ struct nsproxy *nsproxy;
++ struct fs_struct *fs;
++
++ might_sleep();
++
++ vxi->vx_state |= VXS_SHUTDOWN;
++ vs_state_change(vxi, VSC_SHUTDOWN);
++
++ nsproxy = xchg(&vxi->vx_nsproxy, NULL);
++ fs = xchg(&vxi->vx_fs, NULL);
++
++ if (nsproxy)
++ put_nsproxy(nsproxy);
++ if (fs)
++ put_fs_struct(fs);
++}
++
++/* exported stuff */
++
++void free_vx_info(struct vx_info *vxi)
++{
++ unsigned long flags;
++
++ /* check for reference counts first */
++ BUG_ON(atomic_read(&vxi->vx_usecnt));
++ BUG_ON(atomic_read(&vxi->vx_tasks));
++
++ /* context must not be hashed */
++ BUG_ON(vx_info_state(vxi, VXS_HASHED));
++
++ /* context shutdown is mandatory */
++ BUG_ON(!vx_info_state(vxi, VXS_SHUTDOWN));
++
++ BUG_ON(vxi->vx_nsproxy);
++ BUG_ON(vxi->vx_fs);
++
++ spin_lock_irqsave(&vx_info_inactive_lock, flags);
++ hlist_del(&vxi->vx_hlist);
++ spin_unlock_irqrestore(&vx_info_inactive_lock, flags);
++
++ __dealloc_vx_info(vxi);
++}
++
++
++/* hash table for vx_info hash */
++
++#define VX_HASH_SIZE 13
++
++static struct hlist_head vx_info_hash[VX_HASH_SIZE] =
++ { [0 ... VX_HASH_SIZE-1] = HLIST_HEAD_INIT };
++
++static spinlock_t vx_info_hash_lock = SPIN_LOCK_UNLOCKED;
++
++
++static inline unsigned int __hashval(xid_t xid)
++{
++ return (xid % VX_HASH_SIZE);
++}
++
++
++
++/* __hash_vx_info()
++
++ * add the vxi to the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __hash_vx_info(struct vx_info *vxi)
++{
++ struct hlist_head *head;
++
++ vxd_assert_lock(&vx_info_hash_lock);
++ vxdprintk(VXD_CBIT(xid, 4),
++ "__hash_vx_info: %p[#%d]", vxi, vxi->vx_id);
++ vxh_hash_vx_info(vxi);
++
++ /* context must not be hashed */
++ BUG_ON(vx_info_state(vxi, VXS_HASHED));
++
++ vxi->vx_state |= VXS_HASHED;
++ head = &vx_info_hash[__hashval(vxi->vx_id)];
++ hlist_add_head(&vxi->vx_hlist, head);
++ atomic_inc(&vx_global_cactive);
++}
++
++/* __unhash_vx_info()
++
++ * remove the vxi from the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __unhash_vx_info(struct vx_info *vxi)
++{
++ unsigned long flags;
++
++ vxd_assert_lock(&vx_info_hash_lock);
++ vxdprintk(VXD_CBIT(xid, 4),
++ "__unhash_vx_info: %p[#%d.%d.%d]", vxi, vxi->vx_id,
++ atomic_read(&vxi->vx_usecnt), atomic_read(&vxi->vx_tasks));
++ vxh_unhash_vx_info(vxi);
++
++ /* context must be hashed */
++ BUG_ON(!vx_info_state(vxi, VXS_HASHED));
++ /* but without tasks */
++ BUG_ON(atomic_read(&vxi->vx_tasks));
++
++ vxi->vx_state &= ~VXS_HASHED;
++ hlist_del_init(&vxi->vx_hlist);
++ spin_lock_irqsave(&vx_info_inactive_lock, flags);
++ hlist_add_head(&vxi->vx_hlist, &vx_info_inactive);
++ spin_unlock_irqrestore(&vx_info_inactive_lock, flags);
++ atomic_dec(&vx_global_cactive);
++}
++
++
++/* __lookup_vx_info()
++
++ * requires the hash_lock to be held
++ * doesn't increment the vx_refcnt */
++
++static inline struct vx_info *__lookup_vx_info(xid_t xid)
++{
++ struct hlist_head *head = &vx_info_hash[__hashval(xid)];
++ struct hlist_node *pos;
++ struct vx_info *vxi;
++
++ vxd_assert_lock(&vx_info_hash_lock);
++ hlist_for_each(pos, head) {
++ vxi = hlist_entry(pos, struct vx_info, vx_hlist);
++
++ if (vxi->vx_id == xid)
++ goto found;
++ }
++ vxi = NULL;
++found:
++ vxdprintk(VXD_CBIT(xid, 0),
++ "__lookup_vx_info(#%u): %p[#%u]",
++ xid, vxi, vxi ? vxi->vx_id : 0);
++ vxh_lookup_vx_info(vxi, xid);
++ return vxi;
++}
++
++
++/* __create_vx_info()
++
++ * create the requested context
++ * get(), claim() and hash it */
++
++static struct vx_info *__create_vx_info(int id)
++{
++ struct vx_info *new, *vxi = NULL;
++
++ vxdprintk(VXD_CBIT(xid, 1), "create_vx_info(%d)*", id);
++
++ if (!(new = __alloc_vx_info(id)))
++ return ERR_PTR(-ENOMEM);
++
++ /* required to make dynamic xids unique */
++ spin_lock(&vx_info_hash_lock);
++
++ /* static context requested */
++ if ((vxi = __lookup_vx_info(id))) {
++ vxdprintk(VXD_CBIT(xid, 0),
++ "create_vx_info(%d) = %p (already there)", id, vxi);
++ if (vx_info_flags(vxi, VXF_STATE_SETUP, 0))
++ vxi = ERR_PTR(-EBUSY);
++ else
++ vxi = ERR_PTR(-EEXIST);
++ goto out_unlock;
++ }
++ /* new context */
++ vxdprintk(VXD_CBIT(xid, 0),
++ "create_vx_info(%d) = %p (new)", id, new);
++ claim_vx_info(new, NULL);
++ __hash_vx_info(get_vx_info(new));
++ vxi = new, new = NULL;
++
++out_unlock:
++ spin_unlock(&vx_info_hash_lock);
++ vxh_create_vx_info(IS_ERR(vxi) ? NULL : vxi, id);
++ if (new)
++ __dealloc_vx_info(new);
++ return vxi;
++}
++
++
++/* exported stuff */
++
++
++void unhash_vx_info(struct vx_info *vxi)
++{
++ __shutdown_vx_info(vxi);
++ spin_lock(&vx_info_hash_lock);
++ __unhash_vx_info(vxi);
++ spin_unlock(&vx_info_hash_lock);
++ __wakeup_vx_info(vxi);
++}
++
++
++/* lookup_vx_info()
++
++ * search for a vx_info and get() it
++ * negative id means current */
++
++struct vx_info *lookup_vx_info(int id)
++{
++ struct vx_info *vxi = NULL;
++
++ if (id < 0) {
++ vxi = get_vx_info(current->vx_info);
++ } else if (id > 1) {
++ spin_lock(&vx_info_hash_lock);
++ vxi = get_vx_info(__lookup_vx_info(id));
++ spin_unlock(&vx_info_hash_lock);
++ }
++ return vxi;
++}
++
++/* xid_is_hashed()
++
++ * verify that xid is still hashed */
++
++int xid_is_hashed(xid_t xid)
++{
++ int hashed;
++
++ spin_lock(&vx_info_hash_lock);
++ hashed = (__lookup_vx_info(xid) != NULL);
++ spin_unlock(&vx_info_hash_lock);
++ return hashed;
++}
++
++#ifdef CONFIG_PROC_FS
++
++/* get_xid_list()
++
++ * get a subset of hashed xids for proc
++ * assumes size is at least one */
++
++int get_xid_list(int index, unsigned int *xids, int size)
++{
++ int hindex, nr_xids = 0;
++
++ /* only show current and children */
++ if (!vx_check(0, VS_ADMIN | VS_WATCH)) {
++ if (index > 0)
++ return 0;
++ xids[nr_xids] = vx_current_xid();
++ return 1;
++ }
++
++ for (hindex = 0; hindex < VX_HASH_SIZE; hindex++) {
++ struct hlist_head *head = &vx_info_hash[hindex];
++ struct hlist_node *pos;
++
++ spin_lock(&vx_info_hash_lock);
++ hlist_for_each(pos, head) {
++ struct vx_info *vxi;
++
++ if (--index > 0)
++ continue;
++
++ vxi = hlist_entry(pos, struct vx_info, vx_hlist);
++ xids[nr_xids] = vxi->vx_id;
++ if (++nr_xids >= size) {
++ spin_unlock(&vx_info_hash_lock);
++ goto out;
++ }
++ }
++ /* keep the lock time short */
++ spin_unlock(&vx_info_hash_lock);
++ }
++out:
++ return nr_xids;
++}
++#endif
++
++#ifdef CONFIG_VSERVER_DEBUG
++
++void dump_vx_info_inactive(int level)
++{
++ struct hlist_node *entry, *next;
++
++ hlist_for_each_safe(entry, next, &vx_info_inactive) {
++ struct vx_info *vxi =
++ list_entry(entry, struct vx_info, vx_hlist);
++
++ dump_vx_info(vxi, level);
++ }
++}
++
++#endif
++
++#if 0
++int vx_migrate_user(struct task_struct *p, struct vx_info *vxi)
++{
++ struct user_struct *new_user, *old_user;
++
++ if (!p || !vxi)
++ BUG();
++
++ if (vx_info_flags(vxi, VXF_INFO_PRIVATE, 0))
++ return -EACCES;
++
++ new_user = alloc_uid(vxi->vx_id, p->uid);
++ if (!new_user)
++ return -ENOMEM;
++
++ old_user = p->user;
++ if (new_user != old_user) {
++ atomic_inc(&new_user->processes);
++ atomic_dec(&old_user->processes);
++ p->user = new_user;
++ }
++ free_uid(old_user);
++ return 0;
++}
++#endif
++
++#if 0
++void vx_mask_cap_bset(struct vx_info *vxi, struct task_struct *p)
++{
++ // p->cap_effective &= vxi->vx_cap_bset;
++ p->cap_effective =
++ cap_intersect(p->cap_effective, vxi->cap_bset);
++ // p->cap_inheritable &= vxi->vx_cap_bset;
++ p->cap_inheritable =
++ cap_intersect(p->cap_inheritable, vxi->cap_bset);
++ // p->cap_permitted &= vxi->vx_cap_bset;
++ p->cap_permitted =
++ cap_intersect(p->cap_permitted, vxi->cap_bset);
++}
++#endif
++
++
++#include <linux/file.h>
++#include <linux/fdtable.h>
++
++static int vx_openfd_task(struct task_struct *tsk)
++{
++ struct files_struct *files = tsk->files;
++ struct fdtable *fdt;
++ const unsigned long *bptr;
++ int count, total;
++
++ /* no rcu_read_lock() because of spin_lock() */
++ spin_lock(&files->file_lock);
++ fdt = files_fdtable(files);
++ bptr = fdt->open_fds->fds_bits;
++ count = fdt->max_fds / (sizeof(unsigned long) * 8);
++ for (total = 0; count > 0; count--) {
++ if (*bptr)
++ total += hweight_long(*bptr);
++ bptr++;
++ }
++ spin_unlock(&files->file_lock);
++ return total;
++}
++
++
++/* for *space compatibility */
++
++asmlinkage long sys_unshare(unsigned long);
++
++/*
++ * migrate task to new context
++ * gets vxi, puts old_vxi on change
++ * optionally unshares namespaces (hack)
++ */
++
++int vx_migrate_task(struct task_struct *p, struct vx_info *vxi, int unshare)
++{
++ struct vx_info *old_vxi;
++ int ret = 0;
++
++ if (!p || !vxi)
++ BUG();
++
++ vxdprintk(VXD_CBIT(xid, 5),
++ "vx_migrate_task(%p,%p[#%d.%d])", p, vxi,
++ vxi->vx_id, atomic_read(&vxi->vx_usecnt));
++
++ if (vx_info_flags(vxi, VXF_INFO_PRIVATE, 0) &&
++ !vx_info_flags(vxi, VXF_STATE_SETUP, 0))
++ return -EACCES;
++
++ if (vx_info_state(vxi, VXS_SHUTDOWN))
++ return -EFAULT;
++
++ old_vxi = task_get_vx_info(p);
++ if (old_vxi == vxi)
++ goto out;
++
++// if (!(ret = vx_migrate_user(p, vxi))) {
++ {
++ int openfd;
++
++ task_lock(p);
++ openfd = vx_openfd_task(p);
++
++ if (old_vxi) {
++ atomic_dec(&old_vxi->cvirt.nr_threads);
++ atomic_dec(&old_vxi->cvirt.nr_running);
++ __rlim_dec(&old_vxi->limit, RLIMIT_NPROC);
++ /* FIXME: what about the struct files here? */
++ __rlim_sub(&old_vxi->limit, VLIMIT_OPENFD, openfd);
++ /* account for the executable */
++ __rlim_dec(&old_vxi->limit, VLIMIT_DENTRY);
++ }
++ atomic_inc(&vxi->cvirt.nr_threads);
++ atomic_inc(&vxi->cvirt.nr_running);
++ __rlim_inc(&vxi->limit, RLIMIT_NPROC);
++ /* FIXME: what about the struct files here? */
++ __rlim_add(&vxi->limit, VLIMIT_OPENFD, openfd);
++ /* account for the executable */
++ __rlim_inc(&vxi->limit, VLIMIT_DENTRY);
++
++ if (old_vxi) {
++ release_vx_info(old_vxi, p);
++ clr_vx_info(&p->vx_info);
++ }
++ claim_vx_info(vxi, p);
++ set_vx_info(&p->vx_info, vxi);
++ p->xid = vxi->vx_id;
++
++ vxdprintk(VXD_CBIT(xid, 5),
++ "moved task %p into vxi:%p[#%d]",
++ p, vxi, vxi->vx_id);
++
++ // vx_mask_cap_bset(vxi, p);
++ task_unlock(p);
++
++ /* hack for *spaces to provide compatibility */
++ if (unshare) {
++ struct nsproxy *old_nsp, *new_nsp;
++
++ ret = unshare_nsproxy_namespaces(
++ CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWUSER,
++ &new_nsp, NULL);
++ if (ret)
++ goto out;
++
++ old_nsp = xchg(&p->nsproxy, new_nsp);
++ vx_set_space(vxi, CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWUSER);
++ put_nsproxy(old_nsp);
++ }
++ }
++out:
++ put_vx_info(old_vxi);
++ return ret;
++}
++
++int vx_set_reaper(struct vx_info *vxi, struct task_struct *p)
++{
++ struct task_struct *old_reaper;
++
++ if (!vxi)
++ return -EINVAL;
++
++ vxdprintk(VXD_CBIT(xid, 6),
++ "vx_set_reaper(%p[#%d],%p[#%d,%d])",
++ vxi, vxi->vx_id, p, p->xid, p->pid);
++
++ old_reaper = vxi->vx_reaper;
++ if (old_reaper == p)
++ return 0;
++
++ /* set new child reaper */
++ get_task_struct(p);
++ vxi->vx_reaper = p;
++ put_task_struct(old_reaper);
++ return 0;
++}
++
++int vx_set_init(struct vx_info *vxi, struct task_struct *p)
++{
++ if (!vxi)
++ return -EINVAL;
++
++ vxdprintk(VXD_CBIT(xid, 6),
++ "vx_set_init(%p[#%d],%p[#%d,%d,%d])",
++ vxi, vxi->vx_id, p, p->xid, p->pid, p->tgid);
++
++ vxi->vx_flags &= ~VXF_STATE_INIT;
++ vxi->vx_initpid = p->tgid;
++ return 0;
++}
++
++void vx_exit_init(struct vx_info *vxi, struct task_struct *p, int code)
++{
++ vxdprintk(VXD_CBIT(xid, 6),
++ "vx_exit_init(%p[#%d],%p[#%d,%d,%d])",
++ vxi, vxi->vx_id, p, p->xid, p->pid, p->tgid);
++
++ vxi->exit_code = code;
++ vxi->vx_initpid = 0;
++}
++
++
++void vx_set_persistent(struct vx_info *vxi)
++{
++ vxdprintk(VXD_CBIT(xid, 6),
++ "vx_set_persistent(%p[#%d])", vxi, vxi->vx_id);
++
++ get_vx_info(vxi);
++ claim_vx_info(vxi, NULL);
++}
++
++void vx_clear_persistent(struct vx_info *vxi)
++{
++ vxdprintk(VXD_CBIT(xid, 6),
++ "vx_clear_persistent(%p[#%d])", vxi, vxi->vx_id);
++
++ release_vx_info(vxi, NULL);
++ put_vx_info(vxi);
++}
++
++void vx_update_persistent(struct vx_info *vxi)
++{
++ if (vx_info_flags(vxi, VXF_PERSISTENT, 0))
++ vx_set_persistent(vxi);
++ else
++ vx_clear_persistent(vxi);
++}
++
++
++/* task must be current or locked */
++
++void exit_vx_info(struct task_struct *p, int code)
++{
++ struct vx_info *vxi = p->vx_info;
++
++ if (vxi) {
++ atomic_dec(&vxi->cvirt.nr_threads);
++ vx_nproc_dec(p);
++
++ vxi->exit_code = code;
++ release_vx_info(vxi, p);
++ }
++}
++
++void exit_vx_info_early(struct task_struct *p, int code)
++{
++ struct vx_info *vxi = p->vx_info;
++
++ if (vxi) {
++ if (vxi->vx_initpid == p->tgid)
++ vx_exit_init(vxi, p, code);
++ if (vxi->vx_reaper == p)
++ vx_set_reaper(vxi, init_pid_ns.child_reaper);
++ }
++}
++
++
++/* vserver syscall commands below here */
++
++/* taks xid and vx_info functions */
++
++#include <asm/uaccess.h>
++
++
++int vc_task_xid(uint32_t id)
++{
++ xid_t xid;
++
++ if (id) {
++ struct task_struct *tsk;
++
++ read_lock(&tasklist_lock);
++ tsk = find_task_by_real_pid(id);
++ xid = (tsk) ? tsk->xid : -ESRCH;
++ read_unlock(&tasklist_lock);
++ } else
++ xid = vx_current_xid();
++ return xid;
++}
++
++
++int vc_vx_info(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_vx_info_v0 vc_data;
++
++ vc_data.xid = vxi->vx_id;
++ vc_data.initpid = vxi->vx_initpid;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++int vc_ctx_stat(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_stat_v0 vc_data;
++
++ vc_data.usecnt = atomic_read(&vxi->vx_usecnt);
++ vc_data.tasks = atomic_read(&vxi->vx_tasks);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++/* context functions */
++
++int vc_ctx_create(uint32_t xid, void __user *data)
++{
++ struct vcmd_ctx_create vc_data = { .flagword = VXF_INIT_SET };
++ struct vx_info *new_vxi;
++ int ret;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ if ((xid > MAX_S_CONTEXT) || (xid < 2))
++ return -EINVAL;
++
++ new_vxi = __create_vx_info(xid);
++ if (IS_ERR(new_vxi))
++ return PTR_ERR(new_vxi);
++
++ /* initial flags */
++ new_vxi->vx_flags = vc_data.flagword;
++
++ ret = -ENOEXEC;
++ if (vs_state_change(new_vxi, VSC_STARTUP))
++ goto out;
++
++ ret = vx_migrate_task(current, new_vxi, (!data));
++ if (ret)
++ goto out;
++
++ /* return context id on success */
++ ret = new_vxi->vx_id;
++
++ /* get a reference for persistent contexts */
++ if ((vc_data.flagword & VXF_PERSISTENT))
++ vx_set_persistent(new_vxi);
++out:
++ release_vx_info(new_vxi, NULL);
++ put_vx_info(new_vxi);
++ return ret;
++}
++
++
++int vc_ctx_migrate(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_migrate vc_data = { .flagword = 0 };
++ int ret;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = vx_migrate_task(current, vxi, 0);
++ if (ret)
++ return ret;
++ if (vc_data.flagword & VXM_SET_INIT)
++ ret = vx_set_init(vxi, current);
++ if (ret)
++ return ret;
++ if (vc_data.flagword & VXM_SET_REAPER)
++ ret = vx_set_reaper(vxi, current);
++ return ret;
++}
++
++
++int vc_get_cflags(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_flags_v0 vc_data;
++
++ vc_data.flagword = vxi->vx_flags;
++
++ /* special STATE flag handling */
++ vc_data.mask = vs_mask_flags(~0ULL, vxi->vx_flags, VXF_ONE_TIME);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++int vc_set_cflags(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_flags_v0 vc_data;
++ uint64_t mask, trigger;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ /* special STATE flag handling */
++ mask = vs_mask_mask(vc_data.mask, vxi->vx_flags, VXF_ONE_TIME);
++ trigger = (mask & vxi->vx_flags) ^ (mask & vc_data.flagword);
++
++ if (vxi == current->vx_info) {
++ /* if (trigger & VXF_STATE_SETUP)
++ vx_mask_cap_bset(vxi, current); */
++ if (trigger & VXF_STATE_INIT) {
++ int ret;
++
++ ret = vx_set_init(vxi, current);
++ if (ret)
++ return ret;
++ ret = vx_set_reaper(vxi, current);
++ if (ret)
++ return ret;
++ }
++ }
++
++ vxi->vx_flags = vs_mask_flags(vxi->vx_flags,
++ vc_data.flagword, mask);
++ if (trigger & VXF_PERSISTENT)
++ vx_update_persistent(vxi);
++
++ return 0;
++}
++
++
++static inline uint64_t caps_from_cap_t(kernel_cap_t c)
++{
++ uint64_t v = c.cap[0] | ((uint64_t)c.cap[1] << 32);
++
++ // printk("caps_from_cap_t(%08x:%08x) = %016llx\n", c.cap[1], c.cap[0], v);
++ return v;
++}
++
++static inline kernel_cap_t cap_t_from_caps(uint64_t v)
++{
++ kernel_cap_t c = __cap_empty_set;
++
++ c.cap[0] = v & 0xFFFFFFFF;
++ c.cap[1] = (v >> 32) & 0xFFFFFFFF;
++
++ // printk("cap_t_from_caps(%016llx) = %08x:%08x\n", v, c.cap[1], c.cap[0]);
++ return c;
++}
++
++
++static int do_get_caps(struct vx_info *vxi, uint64_t *bcaps, uint64_t *ccaps)
++{
++ if (bcaps)
++ *bcaps = caps_from_cap_t(vxi->vx_bcaps);
++ if (ccaps)
++ *ccaps = vxi->vx_ccaps;
++
++ return 0;
++}
++
++int vc_get_ccaps(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_caps_v1 vc_data;
++ int ret;
++
++ ret = do_get_caps(vxi, NULL, &vc_data.ccaps);
++ if (ret)
++ return ret;
++ vc_data.cmask = ~0ULL;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++static int do_set_caps(struct vx_info *vxi,
++ uint64_t bcaps, uint64_t bmask, uint64_t ccaps, uint64_t cmask)
++{
++ uint64_t bcold = caps_from_cap_t(vxi->vx_bcaps);
++
++#if 0
++ printk("do_set_caps(%16llx, %16llx, %16llx, %16llx)\n",
++ bcaps, bmask, ccaps, cmask);
++#endif
++ vxi->vx_bcaps = cap_t_from_caps(
++ vs_mask_flags(bcold, bcaps, bmask));
++ vxi->vx_ccaps = vs_mask_flags(vxi->vx_ccaps, ccaps, cmask);
++
++ return 0;
++}
++
++int vc_set_ccaps(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_caps_v1 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_caps(vxi, 0, 0, vc_data.ccaps, vc_data.cmask);
++}
++
++int vc_get_bcaps(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_bcaps vc_data;
++ int ret;
++
++ ret = do_get_caps(vxi, &vc_data.bcaps, NULL);
++ if (ret)
++ return ret;
++ vc_data.bmask = ~0ULL;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++int vc_set_bcaps(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_bcaps vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_caps(vxi, vc_data.bcaps, vc_data.bmask, 0, 0);
++}
++
++
++int vc_get_badness(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_badness_v0 vc_data;
++
++ vc_data.bias = vxi->vx_badness_bias;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++int vc_set_badness(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_badness_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ vxi->vx_badness_bias = vc_data.bias;
++ return 0;
++}
++
++#include <linux/module.h>
++
++EXPORT_SYMBOL_GPL(free_vx_info);
++
+--- a/kernel/vserver/cvirt.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cvirt.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,301 @@
++/*
++ * linux/kernel/vserver/cvirt.c
++ *
++ * Virtual Server: Context Virtualization
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 broken out from limit.c
++ * V0.02 added utsname stuff
++ * V0.03 changed vcmds to vxi arg
++ *
++ */
++
++#include <linux/types.h>
++#include <linux/utsname.h>
++#include <linux/vs_cvirt.h>
++#include <linux/vserver/switch.h>
++#include <linux/vserver/cvirt_cmd.h>
++
++#include <asm/uaccess.h>
++
++
++void vx_vsi_uptime(struct timespec *uptime, struct timespec *idle)
++{
++ struct vx_info *vxi = current->vx_info;
++
++ set_normalized_timespec(uptime,
++ uptime->tv_sec - vxi->cvirt.bias_uptime.tv_sec,
++ uptime->tv_nsec - vxi->cvirt.bias_uptime.tv_nsec);
++ if (!idle)
++ return;
++ set_normalized_timespec(idle,
++ idle->tv_sec - vxi->cvirt.bias_idle.tv_sec,
++ idle->tv_nsec - vxi->cvirt.bias_idle.tv_nsec);
++ return;
++}
++
++uint64_t vx_idle_jiffies(void)
++{
++ return init_task.utime + init_task.stime;
++}
++
++
++
++static inline uint32_t __update_loadavg(uint32_t load,
++ int wsize, int delta, int n)
++{
++ unsigned long long calc, prev;
++
++ /* just set it to n */
++ if (unlikely(delta >= wsize))
++ return (n << FSHIFT);
++
++ calc = delta * n;
++ calc <<= FSHIFT;
++ prev = (wsize - delta);
++ prev *= load;
++ calc += prev;
++ do_div(calc, wsize);
++ return calc;
++}
++
++
++void vx_update_load(struct vx_info *vxi)
++{
++ uint32_t now, last, delta;
++ unsigned int nr_running, nr_uninterruptible;
++ unsigned int total;
++ unsigned long flags;
++
++ spin_lock_irqsave(&vxi->cvirt.load_lock, flags);
++
++ now = jiffies;
++ last = vxi->cvirt.load_last;
++ delta = now - last;
++
++ if (delta < 5*HZ)
++ goto out;
++
++ nr_running = atomic_read(&vxi->cvirt.nr_running);
++ nr_uninterruptible = atomic_read(&vxi->cvirt.nr_uninterruptible);
++ total = nr_running + nr_uninterruptible;
++
++ vxi->cvirt.load[0] = __update_loadavg(vxi->cvirt.load[0],
++ 60*HZ, delta, total);
++ vxi->cvirt.load[1] = __update_loadavg(vxi->cvirt.load[1],
++ 5*60*HZ, delta, total);
++ vxi->cvirt.load[2] = __update_loadavg(vxi->cvirt.load[2],
++ 15*60*HZ, delta, total);
++
++ vxi->cvirt.load_last = now;
++out:
++ atomic_inc(&vxi->cvirt.load_updates);
++ spin_unlock_irqrestore(&vxi->cvirt.load_lock, flags);
++}
++
++
++/*
++ * Commands to do_syslog:
++ *
++ * 0 -- Close the log. Currently a NOP.
++ * 1 -- Open the log. Currently a NOP.
++ * 2 -- Read from the log.
++ * 3 -- Read all messages remaining in the ring buffer.
++ * 4 -- Read and clear all messages remaining in the ring buffer
++ * 5 -- Clear ring buffer.
++ * 6 -- Disable printk's to console
++ * 7 -- Enable printk's to console
++ * 8 -- Set level of messages printed to console
++ * 9 -- Return number of unread characters in the log buffer
++ * 10 -- Return size of the log buffer
++ */
++int vx_do_syslog(int type, char __user *buf, int len)
++{
++ int error = 0;
++ int do_clear = 0;
++ struct vx_info *vxi = current->vx_info;
++ struct _vx_syslog *log;
++
++ if (!vxi)
++ return -EINVAL;
++ log = &vxi->cvirt.syslog;
++
++ switch (type) {
++ case 0: /* Close log */
++ case 1: /* Open log */
++ break;
++ case 2: /* Read from log */
++ error = wait_event_interruptible(log->log_wait,
++ (log->log_start - log->log_end));
++ if (error)
++ break;
++ spin_lock_irq(&log->logbuf_lock);
++ spin_unlock_irq(&log->logbuf_lock);
++ break;
++ case 4: /* Read/clear last kernel messages */
++ do_clear = 1;
++ /* fall through */
++ case 3: /* Read last kernel messages */
++ return 0;
++
++ case 5: /* Clear ring buffer */
++ return 0;
++
++ case 6: /* Disable logging to console */
++ case 7: /* Enable logging to console */
++ case 8: /* Set level of messages printed to console */
++ break;
++
++ case 9: /* Number of chars in the log buffer */
++ return 0;
++ case 10: /* Size of the log buffer */
++ return 0;
++ default:
++ error = -EINVAL;
++ break;
++ }
++ return error;
++}
++
++
++/* virtual host info names */
++
++static char *vx_vhi_name(struct vx_info *vxi, int id)
++{
++ struct nsproxy *nsproxy;
++ struct uts_namespace *uts;
++
++
++ if (id == VHIN_CONTEXT)
++ return vxi->vx_name;
++
++ nsproxy = vxi->vx_nsproxy;
++ if (!nsproxy)
++ return NULL;
++
++ uts = nsproxy->uts_ns;
++ if (!uts)
++ return NULL;
++
++ switch (id) {
++ case VHIN_SYSNAME:
++ return uts->name.sysname;
++ case VHIN_NODENAME:
++ return uts->name.nodename;
++ case VHIN_RELEASE:
++ return uts->name.release;
++ case VHIN_VERSION:
++ return uts->name.version;
++ case VHIN_MACHINE:
++ return uts->name.machine;
++ case VHIN_DOMAINNAME:
++ return uts->name.domainname;
++ default:
++ return NULL;
++ }
++ return NULL;
++}
++
++int vc_set_vhi_name(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_vhi_name_v0 vc_data;
++ char *name;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ name = vx_vhi_name(vxi, vc_data.field);
++ if (!name)
++ return -EINVAL;
++
++ memcpy(name, vc_data.name, 65);
++ return 0;
++}
++
++int vc_get_vhi_name(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_vhi_name_v0 vc_data;
++ char *name;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ name = vx_vhi_name(vxi, vc_data.field);
++ if (!name)
++ return -EINVAL;
++
++ memcpy(vc_data.name, name, 65);
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++int vc_virt_stat(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_virt_stat_v0 vc_data;
++ struct _vx_cvirt *cvirt = &vxi->cvirt;
++ struct timespec uptime;
++
++ do_posix_clock_monotonic_gettime(&uptime);
++ set_normalized_timespec(&uptime,
++ uptime.tv_sec - cvirt->bias_uptime.tv_sec,
++ uptime.tv_nsec - cvirt->bias_uptime.tv_nsec);
++
++ vc_data.offset = timeval_to_ns(&cvirt->bias_tv);
++ vc_data.uptime = timespec_to_ns(&uptime);
++ vc_data.nr_threads = atomic_read(&cvirt->nr_threads);
++ vc_data.nr_running = atomic_read(&cvirt->nr_running);
++ vc_data.nr_uninterruptible = atomic_read(&cvirt->nr_uninterruptible);
++ vc_data.nr_onhold = atomic_read(&cvirt->nr_onhold);
++ vc_data.nr_forks = atomic_read(&cvirt->total_forks);
++ vc_data.load[0] = cvirt->load[0];
++ vc_data.load[1] = cvirt->load[1];
++ vc_data.load[2] = cvirt->load[2];
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++#ifdef CONFIG_VSERVER_VTIME
++
++/* virtualized time base */
++
++void vx_gettimeofday(struct timeval *tv)
++{
++ do_gettimeofday(tv);
++ if (!vx_flags(VXF_VIRT_TIME, 0))
++ return;
++
++ tv->tv_sec += current->vx_info->cvirt.bias_tv.tv_sec;
++ tv->tv_usec += current->vx_info->cvirt.bias_tv.tv_usec;
++
++ if (tv->tv_usec >= USEC_PER_SEC) {
++ tv->tv_sec++;
++ tv->tv_usec -= USEC_PER_SEC;
++ } else if (tv->tv_usec < 0) {
++ tv->tv_sec--;
++ tv->tv_usec += USEC_PER_SEC;
++ }
++}
++
++int vx_settimeofday(struct timespec *ts)
++{
++ struct timeval tv;
++
++ if (!vx_flags(VXF_VIRT_TIME, 0))
++ return do_settimeofday(ts);
++
++ do_gettimeofday(&tv);
++ current->vx_info->cvirt.bias_tv.tv_sec =
++ ts->tv_sec - tv.tv_sec;
++ current->vx_info->cvirt.bias_tv.tv_usec =
++ (ts->tv_nsec/NSEC_PER_USEC) - tv.tv_usec;
++ return 0;
++}
++
++#endif
++
+--- a/kernel/vserver/cvirt_init.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cvirt_init.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,69 @@
++
++
++extern uint64_t vx_idle_jiffies(void);
++
++static inline void vx_info_init_cvirt(struct _vx_cvirt *cvirt)
++{
++ uint64_t idle_jiffies = vx_idle_jiffies();
++ uint64_t nsuptime;
++
++ do_posix_clock_monotonic_gettime(&cvirt->bias_uptime);
++ nsuptime = (unsigned long long)cvirt->bias_uptime.tv_sec
++ * NSEC_PER_SEC + cvirt->bias_uptime.tv_nsec;
++ cvirt->bias_clock = nsec_to_clock_t(nsuptime);
++ cvirt->bias_tv.tv_sec = 0;
++ cvirt->bias_tv.tv_usec = 0;
++
++ jiffies_to_timespec(idle_jiffies, &cvirt->bias_idle);
++ atomic_set(&cvirt->nr_threads, 0);
++ atomic_set(&cvirt->nr_running, 0);
++ atomic_set(&cvirt->nr_uninterruptible, 0);
++ atomic_set(&cvirt->nr_onhold, 0);
++
++ spin_lock_init(&cvirt->load_lock);
++ cvirt->load_last = jiffies;
++ atomic_set(&cvirt->load_updates, 0);
++ cvirt->load[0] = 0;
++ cvirt->load[1] = 0;
++ cvirt->load[2] = 0;
++ atomic_set(&cvirt->total_forks, 0);
++
++ spin_lock_init(&cvirt->syslog.logbuf_lock);
++ init_waitqueue_head(&cvirt->syslog.log_wait);
++ cvirt->syslog.log_start = 0;
++ cvirt->syslog.log_end = 0;
++ cvirt->syslog.con_start = 0;
++ cvirt->syslog.logged_chars = 0;
++}
++
++static inline
++void vx_info_init_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc, int cpu)
++{
++ // cvirt_pc->cpustat = { 0 };
++}
++
++static inline void vx_info_exit_cvirt(struct _vx_cvirt *cvirt)
++{
++ int value;
++
++ vxwprintk_xid((value = atomic_read(&cvirt->nr_threads)),
++ "!!! cvirt: %p[nr_threads] = %d on exit.",
++ cvirt, value);
++ vxwprintk_xid((value = atomic_read(&cvirt->nr_running)),
++ "!!! cvirt: %p[nr_running] = %d on exit.",
++ cvirt, value);
++ vxwprintk_xid((value = atomic_read(&cvirt->nr_uninterruptible)),
++ "!!! cvirt: %p[nr_uninterruptible] = %d on exit.",
++ cvirt, value);
++ vxwprintk_xid((value = atomic_read(&cvirt->nr_onhold)),
++ "!!! cvirt: %p[nr_onhold] = %d on exit.",
++ cvirt, value);
++ return;
++}
++
++static inline
++void vx_info_exit_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc, int cpu)
++{
++ return;
++}
++
+--- a/kernel/vserver/cvirt_proc.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/cvirt_proc.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,135 @@
++#ifndef _VX_CVIRT_PROC_H
++#define _VX_CVIRT_PROC_H
++
++#include <linux/nsproxy.h>
++#include <linux/mnt_namespace.h>
++#include <linux/ipc_namespace.h>
++#include <linux/utsname.h>
++#include <linux/ipc.h>
++
++
++static inline
++int vx_info_proc_nsproxy(struct nsproxy *nsproxy, char *buffer)
++{
++ struct mnt_namespace *ns;
++ struct uts_namespace *uts;
++ struct ipc_namespace *ipc;
++ struct path path;
++ char *pstr, *root;
++ int length = 0;
++
++ if (!nsproxy)
++ goto out;
++
++ length += sprintf(buffer + length,
++ "NSProxy:\t%p [%p,%p,%p]\n",
++ nsproxy, nsproxy->mnt_ns,
++ nsproxy->uts_ns, nsproxy->ipc_ns);
++
++ ns = nsproxy->mnt_ns;
++ if (!ns)
++ goto skip_ns;
++
++ pstr = kmalloc(PATH_MAX, GFP_KERNEL);
++ if (!pstr)
++ goto skip_ns;
++
++ path.mnt = ns->root;
++ path.dentry = ns->root->mnt_root;
++ root = d_path(&path, pstr, PATH_MAX - 2);
++ length += sprintf(buffer + length,
++ "Namespace:\t%p [#%u]\n"
++ "RootPath:\t%s\n",
++ ns, atomic_read(&ns->count),
++ root);
++ kfree(pstr);
++skip_ns:
++
++ uts = nsproxy->uts_ns;
++ if (!uts)
++ goto skip_uts;
++
++ length += sprintf(buffer + length,
++ "SysName:\t%.*s\n"
++ "NodeName:\t%.*s\n"
++ "Release:\t%.*s\n"
++ "Version:\t%.*s\n"
++ "Machine:\t%.*s\n"
++ "DomainName:\t%.*s\n",
++ __NEW_UTS_LEN, uts->name.sysname,
++ __NEW_UTS_LEN, uts->name.nodename,
++ __NEW_UTS_LEN, uts->name.release,
++ __NEW_UTS_LEN, uts->name.version,
++ __NEW_UTS_LEN, uts->name.machine,
++ __NEW_UTS_LEN, uts->name.domainname);
++skip_uts:
++
++ ipc = nsproxy->ipc_ns;
++ if (!ipc)
++ goto skip_ipc;
++
++ length += sprintf(buffer + length,
++ "SEMS:\t\t%d %d %d %d %d\n"
++ "MSG:\t\t%d %d %d\n"
++ "SHM:\t\t%lu %lu %d %d\n",
++ ipc->sem_ctls[0], ipc->sem_ctls[1],
++ ipc->sem_ctls[2], ipc->sem_ctls[3],
++ ipc->used_sems,
++ ipc->msg_ctlmax, ipc->msg_ctlmnb, ipc->msg_ctlmni,
++ (unsigned long)ipc->shm_ctlmax,
++ (unsigned long)ipc->shm_ctlall,
++ ipc->shm_ctlmni, ipc->shm_tot);
++skip_ipc:
++out:
++ return length;
++}
++
++
++#include <linux/sched.h>
++
++#define LOAD_INT(x) ((x) >> FSHIFT)
++#define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1 - 1)) * 100)
++
++static inline
++int vx_info_proc_cvirt(struct _vx_cvirt *cvirt, char *buffer)
++{
++ int length = 0;
++ int a, b, c;
++
++ length += sprintf(buffer + length,
++ "BiasUptime:\t%lu.%02lu\n",
++ (unsigned long)cvirt->bias_uptime.tv_sec,
++ (cvirt->bias_uptime.tv_nsec / (NSEC_PER_SEC / 100)));
++
++ a = cvirt->load[0] + (FIXED_1 / 200);
++ b = cvirt->load[1] + (FIXED_1 / 200);
++ c = cvirt->load[2] + (FIXED_1 / 200);
++ length += sprintf(buffer + length,
++ "nr_threads:\t%d\n"
++ "nr_running:\t%d\n"
++ "nr_unintr:\t%d\n"
++ "nr_onhold:\t%d\n"
++ "load_updates:\t%d\n"
++ "loadavg:\t%d.%02d %d.%02d %d.%02d\n"
++ "total_forks:\t%d\n",
++ atomic_read(&cvirt->nr_threads),
++ atomic_read(&cvirt->nr_running),
++ atomic_read(&cvirt->nr_uninterruptible),
++ atomic_read(&cvirt->nr_onhold),
++ atomic_read(&cvirt->load_updates),
++ LOAD_INT(a), LOAD_FRAC(a),
++ LOAD_INT(b), LOAD_FRAC(b),
++ LOAD_INT(c), LOAD_FRAC(c),
++ atomic_read(&cvirt->total_forks));
++ return length;
++}
++
++static inline
++int vx_info_proc_cvirt_pc(struct _vx_cvirt_pc *cvirt_pc,
++ char *buffer, int cpu)
++{
++ int length = 0;
++ return length;
++}
++
++#endif /* _VX_CVIRT_PROC_H */
+--- a/kernel/vserver/debug.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/debug.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,32 @@
++/*
++ * kernel/vserver/debug.c
++ *
++ * Copyright (C) 2005-2007 Herbert Pötzl
++ *
++ * V0.01 vx_info dump support
++ *
++ */
++
++#include <linux/module.h>
++
++#include <linux/vserver/context.h>
++
++
++void dump_vx_info(struct vx_info *vxi, int level)
++{
++ printk("vx_info %p[#%d, %d.%d, %4x]\n", vxi, vxi->vx_id,
++ atomic_read(&vxi->vx_usecnt),
++ atomic_read(&vxi->vx_tasks),
++ vxi->vx_state);
++ if (level > 0) {
++ __dump_vx_limit(&vxi->limit);
++ __dump_vx_sched(&vxi->sched);
++ __dump_vx_cvirt(&vxi->cvirt);
++ __dump_vx_cacct(&vxi->cacct);
++ }
++ printk("---\n");
++}
++
++
++EXPORT_SYMBOL_GPL(dump_vx_info);
++
+--- a/kernel/vserver/device.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/device.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,443 @@
++/*
++ * linux/kernel/vserver/device.c
++ *
++ * Linux-VServer: Device Support
++ *
++ * Copyright (C) 2006 Herbert Pötzl
++ * Copyright (C) 2007 Daniel Hokka Zakrisson
++ *
++ * V0.01 device mapping basics
++ * V0.02 added defaults
++ *
++ */
++
++#include <linux/slab.h>
++#include <linux/rcupdate.h>
++#include <linux/fs.h>
++#include <linux/namei.h>
++#include <linux/hash.h>
++
++#include <asm/errno.h>
++#include <asm/uaccess.h>
++#include <linux/vserver/base.h>
++#include <linux/vserver/debug.h>
++#include <linux/vserver/context.h>
++#include <linux/vserver/device.h>
++#include <linux/vserver/device_cmd.h>
++
++
++#define DMAP_HASH_BITS 4
++
++
++struct vs_mapping {
++ union {
++ struct hlist_node hlist;
++ struct list_head list;
++ } u;
++#define dm_hlist u.hlist
++#define dm_list u.list
++ xid_t xid;
++ dev_t device;
++ struct vx_dmap_target target;
++};
++
++
++static struct hlist_head dmap_main_hash[1 << DMAP_HASH_BITS];
++
++static spinlock_t dmap_main_hash_lock = SPIN_LOCK_UNLOCKED;
++
++static struct vx_dmap_target dmap_defaults[2] = {
++ { .flags = DATTR_OPEN },
++ { .flags = DATTR_OPEN },
++};
++
++
++struct kmem_cache *dmap_cachep __read_mostly;
++
++int __init dmap_cache_init(void)
++{
++ dmap_cachep = kmem_cache_create("dmap_cache",
++ sizeof(struct vs_mapping), 0,
++ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
++ return 0;
++}
++
++__initcall(dmap_cache_init);
++
++
++static inline unsigned int __hashval(dev_t dev, int bits)
++{
++ return hash_long((unsigned long)dev, bits);
++}
++
++
++/* __hash_mapping()
++ * add the mapping to the hash table
++ */
++static inline void __hash_mapping(struct vx_info *vxi, struct vs_mapping *vdm)
++{
++ spinlock_t *hash_lock = &dmap_main_hash_lock;
++ struct hlist_head *head, *hash = dmap_main_hash;
++ int device = vdm->device;
++
++ spin_lock(hash_lock);
++ vxdprintk(VXD_CBIT(misc, 8), "__hash_mapping: %p[#%d] %08x:%08x",
++ vxi, vxi ? vxi->vx_id : 0, device, vdm->target.target);
++
++ head = &hash[__hashval(device, DMAP_HASH_BITS)];
++ hlist_add_head(&vdm->dm_hlist, head);
++ spin_unlock(hash_lock);
++}
++
++
++static inline int __mode_to_default(umode_t mode)
++{
++ switch (mode) {
++ case S_IFBLK:
++ return 0;
++ case S_IFCHR:
++ return 1;
++ default:
++ BUG();
++ }
++}
++
++
++/* __set_default()
++ * set a default
++ */
++static inline void __set_default(struct vx_info *vxi, umode_t mode,
++ struct vx_dmap_target *vdmt)
++{
++ spinlock_t *hash_lock = &dmap_main_hash_lock;
++ spin_lock(hash_lock);
++
++ if (vxi)
++ vxi->dmap.targets[__mode_to_default(mode)] = *vdmt;
++ else
++ dmap_defaults[__mode_to_default(mode)] = *vdmt;
++
++
++ spin_unlock(hash_lock);
++
++ vxdprintk(VXD_CBIT(misc, 8), "__set_default: %p[#%u] %08x %04x",
++ vxi, vxi ? vxi->vx_id : 0, vdmt->target, vdmt->flags);
++}
++
++
++/* __remove_default()
++ * remove a default
++ */
++static inline int __remove_default(struct vx_info *vxi, umode_t mode)
++{
++ spinlock_t *hash_lock = &dmap_main_hash_lock;
++ spin_lock(hash_lock);
++
++ if (vxi)
++ vxi->dmap.targets[__mode_to_default(mode)].flags = 0;
++ else /* remove == reset */
++ dmap_defaults[__mode_to_default(mode)].flags = DATTR_OPEN | mode;
++
++ spin_unlock(hash_lock);
++ return 0;
++}
++
++
++/* __find_mapping()
++ * find a mapping in the hash table
++ *
++ * caller must hold hash_lock
++ */
++static inline int __find_mapping(xid_t xid, dev_t device, umode_t mode,
++ struct vs_mapping **local, struct vs_mapping **global)
++{
++ struct hlist_head *hash = dmap_main_hash;
++ struct hlist_head *head = &hash[__hashval(device, DMAP_HASH_BITS)];
++ struct hlist_node *pos;
++ struct vs_mapping *vdm;
++
++ *local = NULL;
++ if (global)
++ *global = NULL;
++
++ hlist_for_each(pos, head) {
++ vdm = hlist_entry(pos, struct vs_mapping, dm_hlist);
++
++ if ((vdm->device == device) &&
++ !((vdm->target.flags ^ mode) & S_IFMT)) {
++ if (vdm->xid == xid) {
++ *local = vdm;
++ return 1;
++ } else if (global && vdm->xid == 0)
++ *global = vdm;
++ }
++ }
++
++ if (global && *global)
++ return 0;
++ else
++ return -ENOENT;
++}
++
++
++/* __lookup_mapping()
++ * find a mapping and store the result in target and flags
++ */
++static inline int __lookup_mapping(struct vx_info *vxi,
++ dev_t device, dev_t *target, int *flags, umode_t mode)
++{
++ spinlock_t *hash_lock = &dmap_main_hash_lock;
++ struct vs_mapping *vdm, *global;
++ struct vx_dmap_target *vdmt;
++ int ret = 0;
++ xid_t xid = vxi->vx_id;
++ int index;
++
++ spin_lock(hash_lock);
++ if (__find_mapping(xid, device, mode, &vdm, &global) > 0) {
++ ret = 1;
++ vdmt = &vdm->target;
++ goto found;
++ }
++
++ index = __mode_to_default(mode);
++ if (vxi && vxi->dmap.targets[index].flags) {
++ ret = 2;
++ vdmt = &vxi->dmap.targets[index];
++ } else if (global) {
++ ret = 3;
++ vdmt = &global->target;
++ goto found;
++ } else {
++ ret = 4;
++ vdmt = &dmap_defaults[index];
++ }
++
++found:
++ if (target && (vdmt->flags & DATTR_REMAP))
++ *target = vdmt->target;
++ else if (target)
++ *target = device;
++ if (flags)
++ *flags = vdmt->flags;
++
++ spin_unlock(hash_lock);
++
++ return ret;
++}
++
++
++/* __remove_mapping()
++ * remove a mapping from the hash table
++ */
++static inline int __remove_mapping(struct vx_info *vxi, dev_t device,
++ umode_t mode)
++{
++ spinlock_t *hash_lock = &dmap_main_hash_lock;
++ struct vs_mapping *vdm = NULL;
++ int ret = 0;
++
++ spin_lock(hash_lock);
++
++ ret = __find_mapping((vxi ? vxi->vx_id : 0), device, mode, &vdm,
++ NULL);
++ vxdprintk(VXD_CBIT(misc, 8), "__remove_mapping: %p[#%d] %08x %04x",
++ vxi, vxi ? vxi->vx_id : 0, device, mode);
++ if (ret < 0)
++ goto out;
++ hlist_del(&vdm->dm_hlist);
++
++out:
++ spin_unlock(hash_lock);
++ if (vdm)
++ kmem_cache_free(dmap_cachep, vdm);
++ return ret;
++}
++
++
++
++int vs_map_device(struct vx_info *vxi,
++ dev_t device, dev_t *target, umode_t mode)
++{
++ int ret, flags = DATTR_MASK;
++
++ if (!vxi) {
++ if (target)
++ *target = device;
++ goto out;
++ }
++ ret = __lookup_mapping(vxi, device, target, &flags, mode);
++ vxdprintk(VXD_CBIT(misc, 8), "vs_map_device: %08x target: %08x flags: %04x mode: %04x mapped=%d",
++ device, target ? *target : 0, flags, mode, ret);
++out:
++ return (flags & DATTR_MASK);
++}
++
++
++
++static int do_set_mapping(struct vx_info *vxi,
++ dev_t device, dev_t target, int flags, umode_t mode)
++{
++ if (device) {
++ struct vs_mapping *new;
++
++ new = kmem_cache_alloc(dmap_cachep, GFP_KERNEL);
++ if (!new)
++ return -ENOMEM;
++
++ INIT_HLIST_NODE(&new->dm_hlist);
++ new->device = device;
++ new->target.target = target;
++ new->target.flags = flags | mode;
++ new->xid = (vxi ? vxi->vx_id : 0);
++
++ vxdprintk(VXD_CBIT(misc, 8), "do_set_mapping: %08x target: %08x flags: %04x", device, target, flags);
++ __hash_mapping(vxi, new);
++ } else {
++ struct vx_dmap_target new = {
++ .target = target,
++ .flags = flags | mode,
++ };
++ __set_default(vxi, mode, &new);
++ }
++ return 0;
++}
++
++
++static int do_unset_mapping(struct vx_info *vxi,
++ dev_t device, dev_t target, int flags, umode_t mode)
++{
++ int ret = -EINVAL;
++
++ if (device) {
++ ret = __remove_mapping(vxi, device, mode);
++ if (ret < 0)
++ goto out;
++ } else {
++ ret = __remove_default(vxi, mode);
++ if (ret < 0)
++ goto out;
++ }
++
++out:
++ return ret;
++}
++
++
++static inline int __user_device(const char __user *name, dev_t *dev,
++ umode_t *mode)
++{
++ struct nameidata nd;
++ int ret;
++
++ if (!name) {
++ *dev = 0;
++ return 0;
++ }
++ ret = user_path_walk_link(name, &nd);
++ if (ret)
++ return ret;
++ if (nd.path.dentry->d_inode) {
++ *dev = nd.path.dentry->d_inode->i_rdev;
++ *mode = nd.path.dentry->d_inode->i_mode;
++ }
++ path_put(&nd.path);
++ return 0;
++}
++
++static inline int __mapping_mode(dev_t device, dev_t target,
++ umode_t device_mode, umode_t target_mode, umode_t *mode)
++{
++ if (device)
++ *mode = device_mode & S_IFMT;
++ else if (target)
++ *mode = target_mode & S_IFMT;
++ else
++ return -EINVAL;
++
++ /* if both given, device and target mode have to match */
++ if (device && target &&
++ ((device_mode ^ target_mode) & S_IFMT))
++ return -EINVAL;
++ return 0;
++}
++
++
++static inline int do_mapping(struct vx_info *vxi, const char __user *device_path,
++ const char __user *target_path, int flags, int set)
++{
++ dev_t device = ~0, target = ~0;
++ umode_t device_mode = 0, target_mode = 0, mode;
++ int ret;
++
++ ret = __user_device(device_path, &device, &device_mode);
++ if (ret)
++ return ret;
++ ret = __user_device(target_path, &target, &target_mode);
++ if (ret)
++ return ret;
++
++ ret = __mapping_mode(device, target,
++ device_mode, target_mode, &mode);
++ if (ret)
++ return ret;
++
++ if (set)
++ return do_set_mapping(vxi, device, target,
++ flags, mode);
++ else
++ return do_unset_mapping(vxi, device, target,
++ flags, mode);
++}
++
++
++int vc_set_mapping(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_set_mapping_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_mapping(vxi, vc_data.device, vc_data.target,
++ vc_data.flags, 1);
++}
++
++int vc_unset_mapping(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_set_mapping_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_mapping(vxi, vc_data.device, vc_data.target,
++ vc_data.flags, 0);
++}
++
++
++#ifdef CONFIG_COMPAT
++
++int vc_set_mapping_x32(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_set_mapping_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_mapping(vxi, compat_ptr(vc_data.device_ptr),
++ compat_ptr(vc_data.target_ptr), vc_data.flags, 1);
++}
++
++int vc_unset_mapping_x32(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_set_mapping_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_mapping(vxi, compat_ptr(vc_data.device_ptr),
++ compat_ptr(vc_data.target_ptr), vc_data.flags, 0);
++}
++
++#endif /* CONFIG_COMPAT */
++
++
+--- a/kernel/vserver/dlimit.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/dlimit.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,521 @@
++/*
++ * linux/kernel/vserver/dlimit.c
++ *
++ * Virtual Server: Context Disk Limits
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 initial version
++ * V0.02 compat32 splitup
++ *
++ */
++
++#include <linux/statfs.h>
++#include <linux/sched.h>
++#include <linux/vs_tag.h>
++#include <linux/vs_dlimit.h>
++#include <linux/vserver/dlimit_cmd.h>
++
++#include <asm/uaccess.h>
++
++/* __alloc_dl_info()
++
++ * allocate an initialized dl_info struct
++ * doesn't make it visible (hash) */
++
++static struct dl_info *__alloc_dl_info(struct super_block *sb, tag_t tag)
++{
++ struct dl_info *new = NULL;
++
++ vxdprintk(VXD_CBIT(dlim, 5),
++ "alloc_dl_info(%p,%d)*", sb, tag);
++
++ /* would this benefit from a slab cache? */
++ new = kmalloc(sizeof(struct dl_info), GFP_KERNEL);
++ if (!new)
++ return 0;
++
++ memset(new, 0, sizeof(struct dl_info));
++ new->dl_tag = tag;
++ new->dl_sb = sb;
++ INIT_RCU_HEAD(&new->dl_rcu);
++ INIT_HLIST_NODE(&new->dl_hlist);
++ spin_lock_init(&new->dl_lock);
++ atomic_set(&new->dl_refcnt, 0);
++ atomic_set(&new->dl_usecnt, 0);
++
++ /* rest of init goes here */
++
++ vxdprintk(VXD_CBIT(dlim, 4),
++ "alloc_dl_info(%p,%d) = %p", sb, tag, new);
++ return new;
++}
++
++/* __dealloc_dl_info()
++
++ * final disposal of dl_info */
++
++static void __dealloc_dl_info(struct dl_info *dli)
++{
++ vxdprintk(VXD_CBIT(dlim, 4),
++ "dealloc_dl_info(%p)", dli);
++
++ dli->dl_hlist.next = LIST_POISON1;
++ dli->dl_tag = -1;
++ dli->dl_sb = 0;
++
++ BUG_ON(atomic_read(&dli->dl_usecnt));
++ BUG_ON(atomic_read(&dli->dl_refcnt));
++
++ kfree(dli);
++}
++
++
++/* hash table for dl_info hash */
++
++#define DL_HASH_SIZE 13
++
++struct hlist_head dl_info_hash[DL_HASH_SIZE];
++
++static spinlock_t dl_info_hash_lock = SPIN_LOCK_UNLOCKED;
++
++
++static inline unsigned int __hashval(struct super_block *sb, tag_t tag)
++{
++ return ((tag ^ (unsigned long)sb) % DL_HASH_SIZE);
++}
++
++
++
++/* __hash_dl_info()
++
++ * add the dli to the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __hash_dl_info(struct dl_info *dli)
++{
++ struct hlist_head *head;
++
++ vxdprintk(VXD_CBIT(dlim, 6),
++ "__hash_dl_info: %p[#%d]", dli, dli->dl_tag);
++ get_dl_info(dli);
++ head = &dl_info_hash[__hashval(dli->dl_sb, dli->dl_tag)];
++ hlist_add_head_rcu(&dli->dl_hlist, head);
++}
++
++/* __unhash_dl_info()
++
++ * remove the dli from the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __unhash_dl_info(struct dl_info *dli)
++{
++ vxdprintk(VXD_CBIT(dlim, 6),
++ "__unhash_dl_info: %p[#%d]", dli, dli->dl_tag);
++ hlist_del_rcu(&dli->dl_hlist);
++ put_dl_info(dli);
++}
++
++
++/* __lookup_dl_info()
++
++ * requires the rcu_read_lock()
++ * doesn't increment the dl_refcnt */
++
++static inline struct dl_info *__lookup_dl_info(struct super_block *sb, tag_t tag)
++{
++ struct hlist_head *head = &dl_info_hash[__hashval(sb, tag)];
++ struct hlist_node *pos;
++ struct dl_info *dli;
++
++ hlist_for_each_entry_rcu(dli, pos, head, dl_hlist) {
++
++ if (dli->dl_tag == tag && dli->dl_sb == sb) {
++ return dli;
++ }
++ }
++ return NULL;
++}
++
++
++struct dl_info *locate_dl_info(struct super_block *sb, tag_t tag)
++{
++ struct dl_info *dli;
++
++ rcu_read_lock();
++ dli = get_dl_info(__lookup_dl_info(sb, tag));
++ vxdprintk(VXD_CBIT(dlim, 7),
++ "locate_dl_info(%p,#%d) = %p", sb, tag, dli);
++ rcu_read_unlock();
++ return dli;
++}
++
++void rcu_free_dl_info(struct rcu_head *head)
++{
++ struct dl_info *dli = container_of(head, struct dl_info, dl_rcu);
++ int usecnt, refcnt;
++
++ BUG_ON(!dli || !head);
++
++ usecnt = atomic_read(&dli->dl_usecnt);
++ BUG_ON(usecnt < 0);
++
++ refcnt = atomic_read(&dli->dl_refcnt);
++ BUG_ON(refcnt < 0);
++
++ vxdprintk(VXD_CBIT(dlim, 3),
++ "rcu_free_dl_info(%p)", dli);
++ if (!usecnt)
++ __dealloc_dl_info(dli);
++ else
++ printk("!!! rcu didn't free\n");
++}
++
++
++
++
++static int do_addrem_dlimit(uint32_t id, const char __user *name,
++ uint32_t flags, int add)
++{
++ struct nameidata nd;
++ int ret;
++
++ ret = user_path_walk_link(name, &nd);
++ if (!ret) {
++ struct super_block *sb;
++ struct dl_info *dli;
++
++ ret = -EINVAL;
++ if (!nd.path.dentry->d_inode)
++ goto out_release;
++ if (!(sb = nd.path.dentry->d_inode->i_sb))
++ goto out_release;
++
++ if (add) {
++ dli = __alloc_dl_info(sb, id);
++ spin_lock(&dl_info_hash_lock);
++
++ ret = -EEXIST;
++ if (__lookup_dl_info(sb, id))
++ goto out_unlock;
++ __hash_dl_info(dli);
++ dli = NULL;
++ } else {
++ spin_lock(&dl_info_hash_lock);
++ dli = __lookup_dl_info(sb, id);
++
++ ret = -ESRCH;
++ if (!dli)
++ goto out_unlock;
++ __unhash_dl_info(dli);
++ }
++ ret = 0;
++ out_unlock:
++ spin_unlock(&dl_info_hash_lock);
++ if (add && dli)
++ __dealloc_dl_info(dli);
++ out_release:
++ path_put(&nd.path);
++ }
++ return ret;
++}
++
++int vc_add_dlimit(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_base_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_addrem_dlimit(id, vc_data.name, vc_data.flags, 1);
++}
++
++int vc_rem_dlimit(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_base_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_addrem_dlimit(id, vc_data.name, vc_data.flags, 0);
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_add_dlimit_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_base_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_addrem_dlimit(id,
++ compat_ptr(vc_data.name_ptr), vc_data.flags, 1);
++}
++
++int vc_rem_dlimit_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_base_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_addrem_dlimit(id,
++ compat_ptr(vc_data.name_ptr), vc_data.flags, 0);
++}
++
++#endif /* CONFIG_COMPAT */
++
++
++static inline
++int do_set_dlimit(uint32_t id, const char __user *name,
++ uint32_t space_used, uint32_t space_total,
++ uint32_t inodes_used, uint32_t inodes_total,
++ uint32_t reserved, uint32_t flags)
++{
++ struct nameidata nd;
++ int ret;
++
++ ret = user_path_walk_link(name, &nd);
++ if (!ret) {
++ struct super_block *sb;
++ struct dl_info *dli;
++
++ ret = -EINVAL;
++ if (!nd.path.dentry->d_inode)
++ goto out_release;
++ if (!(sb = nd.path.dentry->d_inode->i_sb))
++ goto out_release;
++ if ((reserved != CDLIM_KEEP &&
++ reserved > 100) ||
++ (inodes_used != CDLIM_KEEP &&
++ inodes_used > inodes_total) ||
++ (space_used != CDLIM_KEEP &&
++ space_used > space_total))
++ goto out_release;
++
++ ret = -ESRCH;
++ dli = locate_dl_info(sb, id);
++ if (!dli)
++ goto out_release;
++
++ spin_lock(&dli->dl_lock);
++
++ if (inodes_used != CDLIM_KEEP)
++ dli->dl_inodes_used = inodes_used;
++ if (inodes_total != CDLIM_KEEP)
++ dli->dl_inodes_total = inodes_total;
++ if (space_used != CDLIM_KEEP) {
++ dli->dl_space_used = space_used;
++ dli->dl_space_used <<= 10;
++ }
++ if (space_total == CDLIM_INFINITY)
++ dli->dl_space_total = DLIM_INFINITY;
++ else if (space_total != CDLIM_KEEP) {
++ dli->dl_space_total = space_total;
++ dli->dl_space_total <<= 10;
++ }
++ if (reserved != CDLIM_KEEP)
++ dli->dl_nrlmult = (1 << 10) * (100 - reserved) / 100;
++
++ spin_unlock(&dli->dl_lock);
++
++ put_dl_info(dli);
++ ret = 0;
++
++ out_release:
++ path_put(&nd.path);
++ }
++ return ret;
++}
++
++int vc_set_dlimit(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_dlimit(id, vc_data.name,
++ vc_data.space_used, vc_data.space_total,
++ vc_data.inodes_used, vc_data.inodes_total,
++ vc_data.reserved, vc_data.flags);
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_set_dlimit_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_dlimit(id, compat_ptr(vc_data.name_ptr),
++ vc_data.space_used, vc_data.space_total,
++ vc_data.inodes_used, vc_data.inodes_total,
++ vc_data.reserved, vc_data.flags);
++}
++
++#endif /* CONFIG_COMPAT */
++
++
++static inline
++int do_get_dlimit(uint32_t id, const char __user *name,
++ uint32_t *space_used, uint32_t *space_total,
++ uint32_t *inodes_used, uint32_t *inodes_total,
++ uint32_t *reserved, uint32_t *flags)
++{
++ struct nameidata nd;
++ int ret;
++
++ ret = user_path_walk_link(name, &nd);
++ if (!ret) {
++ struct super_block *sb;
++ struct dl_info *dli;
++
++ ret = -EINVAL;
++ if (!nd.path.dentry->d_inode)
++ goto out_release;
++ if (!(sb = nd.path.dentry->d_inode->i_sb))
++ goto out_release;
++
++ ret = -ESRCH;
++ dli = locate_dl_info(sb, id);
++ if (!dli)
++ goto out_release;
++
++ spin_lock(&dli->dl_lock);
++ *inodes_used = dli->dl_inodes_used;
++ *inodes_total = dli->dl_inodes_total;
++ *space_used = dli->dl_space_used >> 10;
++ if (dli->dl_space_total == DLIM_INFINITY)
++ *space_total = CDLIM_INFINITY;
++ else
++ *space_total = dli->dl_space_total >> 10;
++
++ *reserved = 100 - ((dli->dl_nrlmult * 100 + 512) >> 10);
++ spin_unlock(&dli->dl_lock);
++
++ put_dl_info(dli);
++ ret = -EFAULT;
++
++ ret = 0;
++ out_release:
++ path_put(&nd.path);
++ }
++ return ret;
++}
++
++
++int vc_get_dlimit(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_v0 vc_data;
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_get_dlimit(id, vc_data.name,
++ &vc_data.space_used, &vc_data.space_total,
++ &vc_data.inodes_used, &vc_data.inodes_total,
++ &vc_data.reserved, &vc_data.flags);
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_get_dlimit_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_ctx_dlimit_v0_x32 vc_data;
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_get_dlimit(id, compat_ptr(vc_data.name_ptr),
++ &vc_data.space_used, &vc_data.space_total,
++ &vc_data.inodes_used, &vc_data.inodes_total,
++ &vc_data.reserved, &vc_data.flags);
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++#endif /* CONFIG_COMPAT */
++
++
++void vx_vsi_statfs(struct super_block *sb, struct kstatfs *buf)
++{
++ struct dl_info *dli;
++ __u64 blimit, bfree, bavail;
++ __u32 ifree;
++
++ dli = locate_dl_info(sb, dx_current_tag());
++ if (!dli)
++ return;
++
++ spin_lock(&dli->dl_lock);
++ if (dli->dl_inodes_total == (unsigned long)DLIM_INFINITY)
++ goto no_ilim;
++
++ /* reduce max inodes available to limit */
++ if (buf->f_files > dli->dl_inodes_total)
++ buf->f_files = dli->dl_inodes_total;
++
++ ifree = dli->dl_inodes_total - dli->dl_inodes_used;
++ /* reduce free inodes to min */
++ if (ifree < buf->f_ffree)
++ buf->f_ffree = ifree;
++
++no_ilim:
++ if (dli->dl_space_total == DLIM_INFINITY)
++ goto no_blim;
++
++ blimit = dli->dl_space_total >> sb->s_blocksize_bits;
++
++ if (dli->dl_space_total < dli->dl_space_used)
++ bfree = 0;
++ else
++ bfree = (dli->dl_space_total - dli->dl_space_used)
++ >> sb->s_blocksize_bits;
++
++ bavail = ((dli->dl_space_total >> 10) * dli->dl_nrlmult);
++ if (bavail < dli->dl_space_used)
++ bavail = 0;
++ else
++ bavail = (bavail - dli->dl_space_used)
++ >> sb->s_blocksize_bits;
++
++ /* reduce max space available to limit */
++ if (buf->f_blocks > blimit)
++ buf->f_blocks = blimit;
++
++ /* reduce free space to min */
++ if (bfree < buf->f_bfree)
++ buf->f_bfree = bfree;
++
++ /* reduce avail space to min */
++ if (bavail < buf->f_bavail)
++ buf->f_bavail = bavail;
++
++no_blim:
++ spin_unlock(&dli->dl_lock);
++ put_dl_info(dli);
++
++ return;
++}
++
++#include <linux/module.h>
++
++EXPORT_SYMBOL_GPL(locate_dl_info);
++EXPORT_SYMBOL_GPL(rcu_free_dl_info);
++
+--- a/kernel/vserver/helper.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/helper.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,199 @@
++/*
++ * linux/kernel/vserver/helper.c
++ *
++ * Virtual Context Support
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 basic helper
++ *
++ */
++
++#include <linux/kmod.h>
++#include <linux/reboot.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vserver/signal.h>
++
++
++char vshelper_path[255] = "/sbin/vshelper";
++
++
++static int do_vshelper(char *name, char *argv[], char *envp[], int sync)
++{
++ int ret;
++
++ if ((ret = call_usermodehelper(name, argv, envp, sync))) {
++ printk( KERN_WARNING
++ "%s: (%s %s) returned %s with %d\n",
++ name, argv[1], argv[2],
++ sync ? "sync" : "async", ret);
++ }
++ vxdprintk(VXD_CBIT(switch, 4),
++ "%s: (%s %s) returned %s with %d",
++ name, argv[1], argv[2], sync ? "sync" : "async", ret);
++ return ret;
++}
++
++/*
++ * vshelper path is set via /proc/sys
++ * invoked by vserver sys_reboot(), with
++ * the following arguments
++ *
++ * argv [0] = vshelper_path;
++ * argv [1] = action: "restart", "halt", "poweroff", ...
++ * argv [2] = context identifier
++ *
++ * envp [*] = type-specific parameters
++ */
++
++long vs_reboot_helper(struct vx_info *vxi, int cmd, void __user *arg)
++{
++ char id_buf[8], cmd_buf[16];
++ char uid_buf[16], pid_buf[16];
++ int ret;
++
++ char *argv[] = {vshelper_path, NULL, id_buf, 0};
++ char *envp[] = {"HOME=/", "TERM=linux",
++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin",
++ uid_buf, pid_buf, cmd_buf, 0};
++
++ if (vx_info_state(vxi, VXS_HELPER))
++ return -EAGAIN;
++ vxi->vx_state |= VXS_HELPER;
++
++ snprintf(id_buf, sizeof(id_buf)-1, "%d", vxi->vx_id);
++
++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd);
++ snprintf(uid_buf, sizeof(uid_buf)-1, "VS_UID=%d", current->uid);
++ snprintf(pid_buf, sizeof(pid_buf)-1, "VS_PID=%d", current->pid);
++
++ switch (cmd) {
++ case LINUX_REBOOT_CMD_RESTART:
++ argv[1] = "restart";
++ break;
++
++ case LINUX_REBOOT_CMD_HALT:
++ argv[1] = "halt";
++ break;
++
++ case LINUX_REBOOT_CMD_POWER_OFF:
++ argv[1] = "poweroff";
++ break;
++
++ case LINUX_REBOOT_CMD_SW_SUSPEND:
++ argv[1] = "swsusp";
++ break;
++
++ default:
++ vxi->vx_state &= ~VXS_HELPER;
++ return 0;
++ }
++
++ ret = do_vshelper(vshelper_path, argv, envp, 0);
++ vxi->vx_state &= ~VXS_HELPER;
++ __wakeup_vx_info(vxi);
++ return (ret) ? -EPERM : 0;
++}
++
++
++long vs_reboot(unsigned int cmd, void __user *arg)
++{
++ struct vx_info *vxi = current->vx_info;
++ long ret = 0;
++
++ vxdprintk(VXD_CBIT(misc, 5),
++ "vs_reboot(%p[#%d],%d)",
++ vxi, vxi ? vxi->vx_id : 0, cmd);
++
++ ret = vs_reboot_helper(vxi, cmd, arg);
++ if (ret)
++ return ret;
++
++ vxi->reboot_cmd = cmd;
++ if (vx_info_flags(vxi, VXF_REBOOT_KILL, 0)) {
++ switch (cmd) {
++ case LINUX_REBOOT_CMD_RESTART:
++ case LINUX_REBOOT_CMD_HALT:
++ case LINUX_REBOOT_CMD_POWER_OFF:
++ vx_info_kill(vxi, 0, SIGKILL);
++ vx_info_kill(vxi, 1, SIGKILL);
++ default:
++ break;
++ }
++ }
++ return 0;
++}
++
++
++/*
++ * argv [0] = vshelper_path;
++ * argv [1] = action: "startup", "shutdown"
++ * argv [2] = context identifier
++ *
++ * envp [*] = type-specific parameters
++ */
++
++long vs_state_change(struct vx_info *vxi, unsigned int cmd)
++{
++ char id_buf[8], cmd_buf[16];
++ char *argv[] = {vshelper_path, NULL, id_buf, 0};
++ char *envp[] = {"HOME=/", "TERM=linux",
++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", cmd_buf, 0};
++
++ if (!vx_info_flags(vxi, VXF_SC_HELPER, 0))
++ return 0;
++
++ snprintf(id_buf, sizeof(id_buf)-1, "%d", vxi->vx_id);
++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd);
++
++ switch (cmd) {
++ case VSC_STARTUP:
++ argv[1] = "startup";
++ break;
++ case VSC_SHUTDOWN:
++ argv[1] = "shutdown";
++ break;
++ default:
++ return 0;
++ }
++
++ return do_vshelper(vshelper_path, argv, envp, 1);
++}
++
++
++/*
++ * argv [0] = vshelper_path;
++ * argv [1] = action: "netup", "netdown"
++ * argv [2] = context identifier
++ *
++ * envp [*] = type-specific parameters
++ */
++
++long vs_net_change(struct nx_info *nxi, unsigned int cmd)
++{
++ char id_buf[8], cmd_buf[16];
++ char *argv[] = {vshelper_path, NULL, id_buf, 0};
++ char *envp[] = {"HOME=/", "TERM=linux",
++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", cmd_buf, 0};
++
++ if (!nx_info_flags(nxi, NXF_SC_HELPER, 0))
++ return 0;
++
++ snprintf(id_buf, sizeof(id_buf)-1, "%d", nxi->nx_id);
++ snprintf(cmd_buf, sizeof(cmd_buf)-1, "VS_CMD=%08x", cmd);
++
++ switch (cmd) {
++ case VSC_NETUP:
++ argv[1] = "netup";
++ break;
++ case VSC_NETDOWN:
++ argv[1] = "netdown";
++ break;
++ default:
++ return 0;
++ }
++
++ return do_vshelper(vshelper_path, argv, envp, 1);
++}
++
+--- a/kernel/vserver/history.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/history.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,258 @@
++/*
++ * kernel/vserver/history.c
++ *
++ * Virtual Context History Backtrace
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 basic structure
++ * V0.02 hash/unhash and trace
++ * V0.03 preemption fixes
++ *
++ */
++
++#include <linux/module.h>
++#include <asm/uaccess.h>
++
++#include <linux/vserver/context.h>
++#include <linux/vserver/debug.h>
++#include <linux/vserver/debug_cmd.h>
++#include <linux/vserver/history.h>
++
++
++#ifdef CONFIG_VSERVER_HISTORY
++#define VXH_SIZE CONFIG_VSERVER_HISTORY_SIZE
++#else
++#define VXH_SIZE 64
++#endif
++
++struct _vx_history {
++ unsigned int counter;
++
++ struct _vx_hist_entry entry[VXH_SIZE + 1];
++};
++
++
++DEFINE_PER_CPU(struct _vx_history, vx_history_buffer);
++
++unsigned volatile int vxh_active = 1;
++
++static atomic_t sequence = ATOMIC_INIT(0);
++
++
++/* vxh_advance()
++
++ * requires disabled preemption */
++
++struct _vx_hist_entry *vxh_advance(void *loc)
++{
++ unsigned int cpu = smp_processor_id();
++ struct _vx_history *hist = &per_cpu(vx_history_buffer, cpu);
++ struct _vx_hist_entry *entry;
++ unsigned int index;
++
++ index = vxh_active ? (hist->counter++ % VXH_SIZE) : VXH_SIZE;
++ entry = &hist->entry[index];
++
++ entry->seq = atomic_inc_return(&sequence);
++ entry->loc = loc;
++ return entry;
++}
++
++EXPORT_SYMBOL_GPL(vxh_advance);
++
++
++#define VXH_LOC_FMTS "(#%04x,*%d):%p"
++
++#define VXH_LOC_ARGS(e) (e)->seq, cpu, (e)->loc
++
++
++#define VXH_VXI_FMTS "%p[#%d,%d.%d]"
++
++#define VXH_VXI_ARGS(e) (e)->vxi.ptr, \
++ (e)->vxi.ptr ? (e)->vxi.xid : 0, \
++ (e)->vxi.ptr ? (e)->vxi.usecnt : 0, \
++ (e)->vxi.ptr ? (e)->vxi.tasks : 0
++
++void vxh_dump_entry(struct _vx_hist_entry *e, unsigned cpu)
++{
++ switch (e->type) {
++ case VXH_THROW_OOPS:
++ printk( VXH_LOC_FMTS " oops \n", VXH_LOC_ARGS(e));
++ break;
++
++ case VXH_GET_VX_INFO:
++ case VXH_PUT_VX_INFO:
++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS "\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_GET_VX_INFO) ? "get" : "put",
++ VXH_VXI_ARGS(e));
++ break;
++
++ case VXH_INIT_VX_INFO:
++ case VXH_SET_VX_INFO:
++ case VXH_CLR_VX_INFO:
++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS " @%p\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_INIT_VX_INFO) ? "init" :
++ ((e->type == VXH_SET_VX_INFO) ? "set" : "clr"),
++ VXH_VXI_ARGS(e), e->sc.data);
++ break;
++
++ case VXH_CLAIM_VX_INFO:
++ case VXH_RELEASE_VX_INFO:
++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS " @%p\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_CLAIM_VX_INFO) ? "claim" : "release",
++ VXH_VXI_ARGS(e), e->sc.data);
++ break;
++
++ case VXH_ALLOC_VX_INFO:
++ case VXH_DEALLOC_VX_INFO:
++ printk( VXH_LOC_FMTS " %s_vx_info " VXH_VXI_FMTS "\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_ALLOC_VX_INFO) ? "alloc" : "dealloc",
++ VXH_VXI_ARGS(e));
++ break;
++
++ case VXH_HASH_VX_INFO:
++ case VXH_UNHASH_VX_INFO:
++ printk( VXH_LOC_FMTS " __%s_vx_info " VXH_VXI_FMTS "\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_HASH_VX_INFO) ? "hash" : "unhash",
++ VXH_VXI_ARGS(e));
++ break;
++
++ case VXH_LOC_VX_INFO:
++ case VXH_LOOKUP_VX_INFO:
++ case VXH_CREATE_VX_INFO:
++ printk( VXH_LOC_FMTS " __%s_vx_info [#%d] -> " VXH_VXI_FMTS "\n",
++ VXH_LOC_ARGS(e),
++ (e->type == VXH_CREATE_VX_INFO) ? "create" :
++ ((e->type == VXH_LOC_VX_INFO) ? "loc" : "lookup"),
++ e->ll.arg, VXH_VXI_ARGS(e));
++ break;
++ }
++}
++
++static void __vxh_dump_history(void)
++{
++ unsigned int i, cpu;
++
++ printk("History:\tSEQ: %8x\tNR_CPUS: %d\n",
++ atomic_read(&sequence), NR_CPUS);
++
++ for (i = 0; i < VXH_SIZE; i++) {
++ for_each_online_cpu(cpu) {
++ struct _vx_history *hist =
++ &per_cpu(vx_history_buffer, cpu);
++ unsigned int index = (hist->counter - i) % VXH_SIZE;
++ struct _vx_hist_entry *entry = &hist->entry[index];
++
++ vxh_dump_entry(entry, cpu);
++ }
++ }
++}
++
++void vxh_dump_history(void)
++{
++ vxh_active = 0;
++#ifdef CONFIG_SMP
++ local_irq_enable();
++ smp_send_stop();
++ local_irq_disable();
++#endif
++ __vxh_dump_history();
++}
++
++
++/* vserver syscall commands below here */
++
++
++int vc_dump_history(uint32_t id)
++{
++ vxh_active = 0;
++ __vxh_dump_history();
++ vxh_active = 1;
++
++ return 0;
++}
++
++
++int do_read_history(struct __user _vx_hist_entry *data,
++ int cpu, uint32_t *index, uint32_t *count)
++{
++ int pos, ret = 0;
++ struct _vx_history *hist = &per_cpu(vx_history_buffer, cpu);
++ int end = hist->counter;
++ int start = end - VXH_SIZE + 2;
++ int idx = *index;
++
++ /* special case: get current pos */
++ if (!*count) {
++ *index = end;
++ return 0;
++ }
++
++ /* have we lost some data? */
++ if (idx < start)
++ idx = start;
++
++ for (pos = 0; (pos < *count) && (idx < end); pos++, idx++) {
++ struct _vx_hist_entry *entry =
++ &hist->entry[idx % VXH_SIZE];
++
++ /* send entry to userspace */
++ ret = copy_to_user(&data[pos], entry, sizeof(*entry));
++ if (ret)
++ break;
++ }
++ /* save new index and count */
++ *index = idx;
++ *count = pos;
++ return ret ? ret : (*index < end);
++}
++
++int vc_read_history(uint32_t id, void __user *data)
++{
++ struct vcmd_read_history_v0 vc_data;
++ int ret;
++
++ if (id >= NR_CPUS)
++ return -EINVAL;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_read_history((struct __user _vx_hist_entry *)vc_data.data,
++ id, &vc_data.index, &vc_data.count);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return ret;
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_read_history_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_read_history_v0_x32 vc_data;
++ int ret;
++
++ if (id >= NR_CPUS)
++ return -EINVAL;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_read_history((struct __user _vx_hist_entry *)
++ compat_ptr(vc_data.data_ptr),
++ id, &vc_data.index, &vc_data.count);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return ret;
++}
++
++#endif /* CONFIG_COMPAT */
++
+--- a/kernel/vserver/inet.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/inet.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,225 @@
++
++#include <linux/in.h>
++#include <linux/inetdevice.h>
++#include <linux/vs_inet.h>
++#include <linux/vs_inet6.h>
++#include <linux/vserver/debug.h>
++#include <net/route.h>
++#include <net/addrconf.h>
++
++
++int nx_v4_addr_conflict(struct nx_info *nxi1, struct nx_info *nxi2)
++{
++ int ret = 0;
++
++ if (!nxi1 || !nxi2 || nxi1 == nxi2)
++ ret = 1;
++ else {
++ struct nx_addr_v4 *ptr;
++
++ for (ptr = &nxi1->v4; ptr; ptr = ptr->next) {
++ if (v4_nx_addr_in_nx_info(nxi2, ptr, -1)) {
++ ret = 1;
++ break;
++ }
++ }
++ }
++
++ vxdprintk(VXD_CBIT(net, 2),
++ "nx_v4_addr_conflict(%p,%p): %d",
++ nxi1, nxi2, ret);
++
++ return ret;
++}
++
++
++#ifdef CONFIG_IPV6
++
++int nx_v6_addr_conflict(struct nx_info *nxi1, struct nx_info *nxi2)
++{
++ int ret = 0;
++
++ if (!nxi1 || !nxi2 || nxi1 == nxi2)
++ ret = 1;
++ else {
++ struct nx_addr_v6 *ptr;
++
++ for (ptr = &nxi1->v6; ptr; ptr = ptr->next) {
++ if (v6_nx_addr_in_nx_info(nxi2, ptr, -1)) {
++ ret = 1;
++ break;
++ }
++ }
++ }
++
++ vxdprintk(VXD_CBIT(net, 2),
++ "nx_v6_addr_conflict(%p,%p): %d",
++ nxi1, nxi2, ret);
++
++ return ret;
++}
++
++#endif
++
++int v4_dev_in_nx_info(struct net_device *dev, struct nx_info *nxi)
++{
++ struct in_device *in_dev;
++ struct in_ifaddr **ifap;
++ struct in_ifaddr *ifa;
++ int ret = 0;
++
++ if (!dev)
++ goto out;
++ in_dev = in_dev_get(dev);
++ if (!in_dev)
++ goto out;
++
++ for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
++ ifap = &ifa->ifa_next) {
++ if (v4_addr_in_nx_info(nxi, ifa->ifa_local, NXA_MASK_SHOW)) {
++ ret = 1;
++ break;
++ }
++ }
++ in_dev_put(in_dev);
++out:
++ return ret;
++}
++
++
++#ifdef CONFIG_IPV6
++
++int v6_dev_in_nx_info(struct net_device *dev, struct nx_info *nxi)
++{
++ struct inet6_dev *in_dev;
++ struct inet6_ifaddr **ifap;
++ struct inet6_ifaddr *ifa;
++ int ret = 0;
++
++ if (!dev)
++ goto out;
++ in_dev = in6_dev_get(dev);
++ if (!in_dev)
++ goto out;
++
++ for (ifap = &in_dev->addr_list; (ifa = *ifap) != NULL;
++ ifap = &ifa->if_next) {
++ if (v6_addr_in_nx_info(nxi, &ifa->addr, -1)) {
++ ret = 1;
++ break;
++ }
++ }
++ in6_dev_put(in_dev);
++out:
++ return ret;
++}
++
++#endif
++
++int dev_in_nx_info(struct net_device *dev, struct nx_info *nxi)
++{
++ int ret = 1;
++
++ if (!nxi)
++ goto out;
++ if (nxi->v4.type && v4_dev_in_nx_info(dev, nxi))
++ goto out;
++#ifdef CONFIG_IPV6
++ ret = 2;
++ if (nxi->v6.type && v6_dev_in_nx_info(dev, nxi))
++ goto out;
++#endif
++ ret = 0;
++out:
++ vxdprintk(VXD_CBIT(net, 3),
++ "dev_in_nx_info(%p,%p[#%d]) = %d",
++ dev, nxi, nxi ? nxi->nx_id : 0, ret);
++ return ret;
++}
++
++int ip_v4_find_src(struct net *net, struct nx_info *nxi,
++ struct rtable **rp, struct flowi *fl)
++{
++ if (!nxi)
++ return 0;
++
++ /* FIXME: handle lback only case */
++ if (!NX_IPV4(nxi))
++ return -EPERM;
++
++ vxdprintk(VXD_CBIT(net, 4),
++ "ip_v4_find_src(%p[#%u]) " NIPQUAD_FMT " -> " NIPQUAD_FMT,
++ nxi, nxi ? nxi->nx_id : 0,
++ NIPQUAD(fl->fl4_src), NIPQUAD(fl->fl4_dst));
++
++ /* single IP is unconditional */
++ if (nx_info_flags(nxi, NXF_SINGLE_IP, 0) &&
++ (fl->fl4_src == INADDR_ANY))
++ fl->fl4_src = nxi->v4.ip[0].s_addr;
++
++ if (fl->fl4_src == INADDR_ANY) {
++ struct nx_addr_v4 *ptr;
++ __be32 found = 0;
++ int err;
++
++ err = __ip_route_output_key(net, rp, fl);
++ if (!err) {
++ found = (*rp)->rt_src;
++ ip_rt_put(*rp);
++ vxdprintk(VXD_CBIT(net, 4),
++ "ip_v4_find_src(%p[#%u]) rok[%u]: " NIPQUAD_FMT,
++ nxi, nxi ? nxi->nx_id : 0, fl->oif, NIPQUAD(found));
++ if (v4_addr_in_nx_info(nxi, found, NXA_MASK_BIND))
++ goto found;
++ }
++
++ for (ptr = &nxi->v4; ptr; ptr = ptr->next) {
++ __be32 primary = ptr->ip[0].s_addr;
++ __be32 mask = ptr->mask.s_addr;
++ __be32 neta = primary & mask;
++
++ vxdprintk(VXD_CBIT(net, 4), "ip_v4_find_src(%p[#%u]) chk: "
++ NIPQUAD_FMT "/" NIPQUAD_FMT "/" NIPQUAD_FMT,
++ nxi, nxi ? nxi->nx_id : 0, NIPQUAD(primary),
++ NIPQUAD(mask), NIPQUAD(neta));
++ if ((found & mask) != neta)
++ continue;
++
++ fl->fl4_src = primary;
++ err = __ip_route_output_key(net, rp, fl);
++ vxdprintk(VXD_CBIT(net, 4),
++ "ip_v4_find_src(%p[#%u]) rok[%u]: " NIPQUAD_FMT,
++ nxi, nxi ? nxi->nx_id : 0, fl->oif, NIPQUAD(primary));
++ if (!err) {
++ found = (*rp)->rt_src;
++ ip_rt_put(*rp);
++ if (found == primary)
++ goto found;
++ }
++ }
++ /* still no source ip? */
++ found = ipv4_is_loopback(fl->fl4_dst)
++ ? IPI_LOOPBACK : nxi->v4.ip[0].s_addr;
++ found:
++ /* assign src ip to flow */
++ fl->fl4_src = found;
++
++ } else {
++ if (!v4_addr_in_nx_info(nxi, fl->fl4_src, NXA_MASK_BIND))
++ return -EPERM;
++ }
++
++ if (nx_info_flags(nxi, NXF_LBACK_REMAP, 0)) {
++ if (ipv4_is_loopback(fl->fl4_dst))
++ fl->fl4_dst = nxi->v4_lback.s_addr;
++ if (ipv4_is_loopback(fl->fl4_src))
++ fl->fl4_src = nxi->v4_lback.s_addr;
++ } else if (ipv4_is_loopback(fl->fl4_dst) &&
++ !nx_info_flags(nxi, NXF_LBACK_ALLOW, 0))
++ return -EPERM;
++
++ return 0;
++}
++
++EXPORT_SYMBOL_GPL(ip_v4_find_src);
++
+--- a/kernel/vserver/init.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/init.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,45 @@
++/*
++ * linux/kernel/init.c
++ *
++ * Virtual Server Init
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 basic structure
++ *
++ */
++
++#include <linux/init.h>
++
++int vserver_register_sysctl(void);
++void vserver_unregister_sysctl(void);
++
++
++static int __init init_vserver(void)
++{
++ int ret = 0;
++
++#ifdef CONFIG_VSERVER_DEBUG
++ vserver_register_sysctl();
++#endif
++ return ret;
++}
++
++
++static void __exit exit_vserver(void)
++{
++
++#ifdef CONFIG_VSERVER_DEBUG
++ vserver_unregister_sysctl();
++#endif
++ return;
++}
++
++/* FIXME: GFP_ZONETYPES gone
++long vx_slab[GFP_ZONETYPES]; */
++long vx_area;
++
++
++module_init(init_vserver);
++module_exit(exit_vserver);
++
+--- a/kernel/vserver/inode.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/inode.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,409 @@
++/*
++ * linux/kernel/vserver/inode.c
++ *
++ * Virtual Server: File System Support
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 separated from vcontext V0.05
++ * V0.02 moved to tag (instead of xid)
++ *
++ */
++
++#include <linux/tty.h>
++#include <linux/proc_fs.h>
++#include <linux/devpts_fs.h>
++#include <linux/fs.h>
++#include <linux/file.h>
++#include <linux/mount.h>
++#include <linux/parser.h>
++#include <linux/vserver/inode.h>
++#include <linux/vserver/inode_cmd.h>
++#include <linux/vs_base.h>
++#include <linux/vs_tag.h>
++
++#include <asm/uaccess.h>
++
++
++static int __vc_get_iattr(struct inode *in, uint32_t *tag, uint32_t *flags, uint32_t *mask)
++{
++ struct proc_dir_entry *entry;
++
++ if (!in || !in->i_sb)
++ return -ESRCH;
++
++ *flags = IATTR_TAG
++ | (IS_BARRIER(in) ? IATTR_BARRIER : 0)
++ | (IS_IXUNLINK(in) ? IATTR_IXUNLINK : 0)
++ | (IS_IMMUTABLE(in) ? IATTR_IMMUTABLE : 0);
++ *mask = IATTR_IXUNLINK | IATTR_IMMUTABLE;
++
++ if (S_ISDIR(in->i_mode))
++ *mask |= IATTR_BARRIER;
++
++ if (IS_TAGGED(in)) {
++ *tag = in->i_tag;
++ *mask |= IATTR_TAG;
++ }
++
++ switch (in->i_sb->s_magic) {
++ case PROC_SUPER_MAGIC:
++ entry = PROC_I(in)->pde;
++
++ /* check for specific inodes? */
++ if (entry)
++ *mask |= IATTR_FLAGS;
++ if (entry)
++ *flags |= (entry->vx_flags & IATTR_FLAGS);
++ else
++ *flags |= (PROC_I(in)->vx_flags & IATTR_FLAGS);
++ break;
++
++ case DEVPTS_SUPER_MAGIC:
++ *tag = in->i_tag;
++ *mask |= IATTR_TAG;
++ break;
++
++ default:
++ break;
++ }
++ return 0;
++}
++
++int vc_get_iattr(void __user *data)
++{
++ struct nameidata nd;
++ struct vcmd_ctx_iattr_v1 vc_data = { .tag = -1 };
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = user_path_walk_link(vc_data.name, &nd);
++ if (!ret) {
++ ret = __vc_get_iattr(nd.path.dentry->d_inode,
++ &vc_data.tag, &vc_data.flags, &vc_data.mask);
++ path_put(&nd.path);
++ }
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_get_iattr_x32(void __user *data)
++{
++ struct nameidata nd;
++ struct vcmd_ctx_iattr_v1_x32 vc_data = { .tag = -1 };
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = user_path_walk_link(compat_ptr(vc_data.name_ptr), &nd);
++ if (!ret) {
++ ret = __vc_get_iattr(nd.path.dentry->d_inode,
++ &vc_data.tag, &vc_data.flags, &vc_data.mask);
++ path_put(&nd.path);
++ }
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
++#endif /* CONFIG_COMPAT */
++
++
++int vc_fget_iattr(uint32_t fd, void __user *data)
++{
++ struct file *filp;
++ struct vcmd_ctx_fiattr_v0 vc_data = { .tag = -1 };
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ filp = fget(fd);
++ if (!filp || !filp->f_dentry || !filp->f_dentry->d_inode)
++ return -EBADF;
++
++ ret = __vc_get_iattr(filp->f_dentry->d_inode,
++ &vc_data.tag, &vc_data.flags, &vc_data.mask);
++
++ fput(filp);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
++
++static int __vc_set_iattr(struct dentry *de, uint32_t *tag, uint32_t *flags, uint32_t *mask)
++{
++ struct inode *in = de->d_inode;
++ int error = 0, is_proc = 0, has_tag = 0;
++ struct iattr attr = { 0 };
++
++ if (!in || !in->i_sb)
++ return -ESRCH;
++
++ is_proc = (in->i_sb->s_magic == PROC_SUPER_MAGIC);
++ if ((*mask & IATTR_FLAGS) && !is_proc)
++ return -EINVAL;
++
++ has_tag = IS_TAGGED(in) ||
++ (in->i_sb->s_magic == DEVPTS_SUPER_MAGIC);
++ if ((*mask & IATTR_TAG) && !has_tag)
++ return -EINVAL;
++
++ mutex_lock(&in->i_mutex);
++ if (*mask & IATTR_TAG) {
++ attr.ia_tag = *tag;
++ attr.ia_valid |= ATTR_TAG;
++ }
++
++ if (*mask & IATTR_FLAGS) {
++ struct proc_dir_entry *entry = PROC_I(in)->pde;
++ unsigned int iflags = PROC_I(in)->vx_flags;
++
++ iflags = (iflags & ~(*mask & IATTR_FLAGS))
++ | (*flags & IATTR_FLAGS);
++ PROC_I(in)->vx_flags = iflags;
++ if (entry)
++ entry->vx_flags = iflags;
++ }
++
++ if (*mask & (IATTR_BARRIER | IATTR_IXUNLINK | IATTR_IMMUTABLE)) {
++ if (*mask & IATTR_IMMUTABLE) {
++ if (*flags & IATTR_IMMUTABLE)
++ in->i_flags |= S_IMMUTABLE;
++ else
++ in->i_flags &= ~S_IMMUTABLE;
++ }
++ if (*mask & IATTR_IXUNLINK) {
++ if (*flags & IATTR_IXUNLINK)
++ in->i_flags |= S_IXUNLINK;
++ else
++ in->i_flags &= ~S_IXUNLINK;
++ }
++ if (S_ISDIR(in->i_mode) && (*mask & IATTR_BARRIER)) {
++ if (*flags & IATTR_BARRIER)
++ in->i_vflags |= V_BARRIER;
++ else
++ in->i_vflags &= ~V_BARRIER;
++ }
++ if (in->i_op && in->i_op->sync_flags) {
++ error = in->i_op->sync_flags(in);
++ if (error)
++ goto out;
++ }
++ }
++
++ if (attr.ia_valid) {
++ if (in->i_op && in->i_op->setattr)
++ error = in->i_op->setattr(de, &attr);
++ else {
++ error = inode_change_ok(in, &attr);
++ if (!error)
++ error = inode_setattr(in, &attr);
++ }
++ }
++
++out:
++ mutex_unlock(&in->i_mutex);
++ return error;
++}
++
++int vc_set_iattr(void __user *data)
++{
++ struct nameidata nd;
++ struct vcmd_ctx_iattr_v1 vc_data;
++ int ret;
++
++ if (!capable(CAP_LINUX_IMMUTABLE))
++ return -EPERM;
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = user_path_walk_link(vc_data.name, &nd);
++ if (!ret) {
++ ret = __vc_set_iattr(nd.path.dentry,
++ &vc_data.tag, &vc_data.flags, &vc_data.mask);
++ path_put(&nd.path);
++ }
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_set_iattr_x32(void __user *data)
++{
++ struct nameidata nd;
++ struct vcmd_ctx_iattr_v1_x32 vc_data;
++ int ret;
++
++ if (!capable(CAP_LINUX_IMMUTABLE))
++ return -EPERM;
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = user_path_walk_link(compat_ptr(vc_data.name_ptr), &nd);
++ if (!ret) {
++ ret = __vc_set_iattr(nd.path.dentry,
++ &vc_data.tag, &vc_data.flags, &vc_data.mask);
++ path_put(&nd.path);
++ }
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
++#endif /* CONFIG_COMPAT */
++
++int vc_fset_iattr(uint32_t fd, void __user *data)
++{
++ struct file *filp;
++ struct vcmd_ctx_fiattr_v0 vc_data;
++ int ret;
++
++ if (!capable(CAP_LINUX_IMMUTABLE))
++ return -EPERM;
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ filp = fget(fd);
++ if (!filp || !filp->f_dentry || !filp->f_dentry->d_inode)
++ return -EBADF;
++
++ ret = __vc_set_iattr(filp->f_dentry, &vc_data.tag,
++ &vc_data.flags, &vc_data.mask);
++
++ fput(filp);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return ret;
++}
++
++
++enum { Opt_notagcheck, Opt_tag, Opt_notag, Opt_tagid, Opt_err };
++
++static match_table_t tokens = {
++ {Opt_notagcheck, "notagcheck"},
++#ifdef CONFIG_PROPAGATE
++ {Opt_notag, "notag"},
++ {Opt_tag, "tag"},
++ {Opt_tagid, "tagid=%u"},
++#endif
++ {Opt_err, NULL}
++};
++
++
++static void __dx_parse_remove(char *string, char *opt)
++{
++ char *p = strstr(string, opt);
++ char *q = p;
++
++ if (p) {
++ while (*q != '\0' && *q != ',')
++ q++;
++ while (*q)
++ *p++ = *q++;
++ while (*p)
++ *p++ = '\0';
++ }
++}
++
++static inline
++int __dx_parse_tag(char *string, tag_t *tag, int remove)
++{
++ substring_t args[MAX_OPT_ARGS];
++ int token, option = 0;
++
++ if (!string)
++ return 0;
++
++ token = match_token(string, tokens, args);
++
++ vxdprintk(VXD_CBIT(tag, 7),
++ "dx_parse_tag(»%s«): %d:#%d",
++ string, token, option);
++
++ switch (token) {
++ case Opt_tag:
++ if (tag)
++ *tag = 0;
++ if (remove)
++ __dx_parse_remove(string, "tag");
++ return MNT_TAGID;
++ case Opt_notag:
++ if (remove)
++ __dx_parse_remove(string, "notag");
++ return MNT_NOTAG;
++ case Opt_notagcheck:
++ if (remove)
++ __dx_parse_remove(string, "notagcheck");
++ return MNT_NOTAGCHECK;
++ case Opt_tagid:
++ if (tag && !match_int(args, &option))
++ *tag = option;
++ if (remove)
++ __dx_parse_remove(string, "tagid");
++ return MNT_TAGID;
++ }
++ return 0;
++}
++
++int dx_parse_tag(char *string, tag_t *tag, int remove)
++{
++ int retval, flags = 0;
++
++ while ((retval = __dx_parse_tag(string, tag, remove)))
++ flags |= retval;
++ return flags;
++}
++
++#ifdef CONFIG_PROPAGATE
++
++void __dx_propagate_tag(struct nameidata *nd, struct inode *inode)
++{
++ tag_t new_tag = 0;
++ struct vfsmount *mnt;
++ int propagate;
++
++ if (!nd)
++ return;
++ mnt = nd->path.mnt;
++ if (!mnt)
++ return;
++
++ propagate = (mnt->mnt_flags & MNT_TAGID);
++ if (propagate)
++ new_tag = mnt->mnt_tag;
++
++ vxdprintk(VXD_CBIT(tag, 7),
++ "dx_propagate_tag(%p[#%lu.%d]): %d,%d",
++ inode, inode->i_ino, inode->i_tag,
++ new_tag, (propagate) ? 1 : 0);
++
++ if (propagate)
++ inode->i_tag = new_tag;
++}
++
++#include <linux/module.h>
++
++EXPORT_SYMBOL_GPL(__dx_propagate_tag);
++
++#endif /* CONFIG_PROPAGATE */
++
+--- a/kernel/vserver/Kconfig 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/Kconfig 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,251 @@
++#
++# Linux VServer configuration
++#
++
++menu "Linux VServer"
++
++config VSERVER_AUTO_LBACK
++ bool "Automatically Assign Loopback IP"
++ default y
++ help
++ Automatically assign a guest specific loopback
++ IP and add it to the kernel network stack on
++ startup.
++
++config VSERVER_AUTO_SINGLE
++ bool "Automatic Single IP Special Casing"
++ depends on EXPERIMENTAL
++ default y
++ help
++ This allows network contexts with a single IP to
++ automatically remap 0.0.0.0 bindings to that IP,
++ avoiding further network checks and improving
++ performance.
++
++ (note: such guests do not allow to change the ip
++ on the fly and do not show loopback addresses)
++
++config VSERVER_COWBL
++ bool "Enable COW Immutable Link Breaking"
++ default y
++ help
++ This enables the COW (Copy-On-Write) link break code.
++ It allows you to treat unified files like normal files
++ when writing to them (which will implicitely break the
++ link and create a copy of the unified file)
++
++config VSERVER_VTIME
++ bool "Enable Virtualized Guest Time"
++ depends on EXPERIMENTAL
++ default n
++ help
++ This enables per guest time offsets to allow for
++ adjusting the system clock individually per guest.
++ this adds some overhead to the time functions and
++ therefore should not be enabled without good reason.
++
++config VSERVER_DEVICE
++ bool "Enable Guest Device Mapping"
++ depends on EXPERIMENTAL
++ default n
++ help
++ This enables generic device remapping.
++
++config VSERVER_PROC_SECURE
++ bool "Enable Proc Security"
++ depends on PROC_FS
++ default y
++ help
++ This configures ProcFS security to initially hide
++ non-process entries for all contexts except the main and
++ spectator context (i.e. for all guests), which is a secure
++ default.
++
++ (note: on 1.2x the entries were visible by default)
++
++config VSERVER_HARDCPU
++ bool "Enable Hard CPU Limits"
++ default y
++ help
++ Activate the Hard CPU Limits
++
++ This will compile in code that allows the Token Bucket
++ Scheduler to put processes on hold when a context's
++ tokens are depleted (provided that its per-context
++ sched_hard flag is set).
++
++ Processes belonging to that context will not be able
++ to consume CPU resources again until a per-context
++ configured minimum of tokens has been reached.
++
++config VSERVER_IDLETIME
++ bool "Avoid idle CPUs by skipping Time"
++ depends on VSERVER_HARDCPU
++ default y
++ help
++ This option allows the scheduler to artificially
++ advance time (per cpu) when otherwise the idle
++ task would be scheduled, thus keeping the cpu
++ busy and sharing the available resources among
++ certain contexts.
++
++config VSERVER_IDLELIMIT
++ bool "Limit the IDLE task"
++ depends on VSERVER_HARDCPU
++ default n
++ help
++ Limit the idle slices, so the the next context
++ will be scheduled as soon as possible.
++
++ This might improve interactivity and latency, but
++ will also marginally increase scheduling overhead.
++
++choice
++ prompt "Persistent Inode Tagging"
++ default TAGGING_ID24
++ help
++ This adds persistent context information to filesystems
++ mounted with the tagxid option. Tagging is a requirement
++ for per-context disk limits and per-context quota.
++
++
++config TAGGING_NONE
++ bool "Disabled"
++ help
++ do not store per-context information in inodes.
++
++config TAGGING_UID16
++ bool "UID16/GID32"
++ help
++ reduces UID to 16 bit, but leaves GID at 32 bit.
++
++config TAGGING_GID16
++ bool "UID32/GID16"
++ help
++ reduces GID to 16 bit, but leaves UID at 32 bit.
++
++config TAGGING_ID24
++ bool "UID24/GID24"
++ help
++ uses the upper 8bit from UID and GID for XID tagging
++ which leaves 24bit for UID/GID each, which should be
++ more than sufficient for normal use.
++
++config TAGGING_INTERN
++ bool "UID32/GID32"
++ help
++ this uses otherwise reserved inode fields in the on
++ disk representation, which limits the use to a few
++ filesystems (currently ext2 and ext3)
++
++endchoice
++
++config TAG_NFSD
++ bool "Tag NFSD User Auth and Files"
++ default n
++ help
++ Enable this if you do want the in-kernel NFS
++ Server to use the tagging specified above.
++ (will require patched clients too)
++
++config VSERVER_PRIVACY
++ bool "Honor Privacy Aspects of Guests"
++ default n
++ help
++ When enabled, most context checks will disallow
++ access to structures assigned to a specific context,
++ like ptys or loop devices.
++
++config VSERVER_CONTEXTS
++ int "Maximum number of Contexts (1-65533)" if EMBEDDED
++ range 1 65533
++ default "768" if 64BIT
++ default "256"
++ help
++ This setting will optimize certain data structures
++ and memory allocations according to the expected
++ maximum.
++
++ note: this is not a strict upper limit.
++
++config VSERVER_WARN
++ bool "VServer Warnings"
++ default y
++ help
++ This enables various runtime warnings, which will
++ notify about potential manipulation attempts or
++ resource shortage. It is generally considered to
++ be a good idea to have that enabled.
++
++config VSERVER_DEBUG
++ bool "VServer Debugging Code"
++ default n
++ help
++ Set this to yes if you want to be able to activate
++ debugging output at runtime. It adds a very small
++ overhead to all vserver related functions and
++ increases the kernel size by about 20k.
++
++config VSERVER_HISTORY
++ bool "VServer History Tracing"
++ depends on VSERVER_DEBUG
++ default n
++ help
++ Set this to yes if you want to record the history of
++ linux-vserver activities, so they can be replayed in
++ the event of a kernel panic or oops.
++
++config VSERVER_HISTORY_SIZE
++ int "Per-CPU History Size (32-65536)"
++ depends on VSERVER_HISTORY
++ range 32 65536
++ default 64
++ help
++ This allows you to specify the number of entries in
++ the per-CPU history buffer.
++
++config VSERVER_MONITOR
++ bool "VServer Scheduling Monitor"
++ depends on VSERVER_DISABLED
++ default n
++ help
++ Set this to yes if you want to record the scheduling
++ decisions, so that they can be relayed to userspace
++ for detailed analysis.
++
++config VSERVER_MONITOR_SIZE
++ int "Per-CPU Monitor Queue Size (32-65536)"
++ depends on VSERVER_MONITOR
++ range 32 65536
++ default 1024
++ help
++ This allows you to specify the number of entries in
++ the per-CPU scheduling monitor buffer.
++
++config VSERVER_MONITOR_SYNC
++ int "Per-CPU Monitor Sync Interval (0-65536)"
++ depends on VSERVER_MONITOR
++ range 0 65536
++ default 256
++ help
++ This allows you to specify the interval in ticks
++ when a time sync entry is inserted.
++
++endmenu
++
++
++config VSERVER
++ bool
++ default y
++ select NAMESPACES
++ select UTS_NS
++ select IPC_NS
++ select USER_NS
++ select SYSVIPC
++
++config VSERVER_SECURITY
++ bool
++ depends on SECURITY
++ default y
++ select SECURITY_CAPABILITIES
++
+--- a/kernel/vserver/limit.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/limit.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,319 @@
++/*
++ * linux/kernel/vserver/limit.c
++ *
++ * Virtual Server: Context Limits
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 broken out from vcontext V0.05
++ * V0.02 changed vcmds to vxi arg
++ *
++ */
++
++#include <linux/sched.h>
++#include <linux/module.h>
++#include <linux/vs_limit.h>
++#include <linux/vserver/limit.h>
++#include <linux/vserver/limit_cmd.h>
++
++#include <asm/uaccess.h>
++
++
++const char *vlimit_name[NUM_LIMITS] = {
++ [RLIMIT_CPU] = "CPU",
++ [RLIMIT_RSS] = "RSS",
++ [RLIMIT_NPROC] = "NPROC",
++ [RLIMIT_NOFILE] = "NOFILE",
++ [RLIMIT_MEMLOCK] = "VML",
++ [RLIMIT_AS] = "VM",
++ [RLIMIT_LOCKS] = "LOCKS",
++ [RLIMIT_SIGPENDING] = "SIGP",
++ [RLIMIT_MSGQUEUE] = "MSGQ",
++
++ [VLIMIT_NSOCK] = "NSOCK",
++ [VLIMIT_OPENFD] = "OPENFD",
++ [VLIMIT_ANON] = "ANON",
++ [VLIMIT_SHMEM] = "SHMEM",
++ [VLIMIT_DENTRY] = "DENTRY",
++};
++
++EXPORT_SYMBOL_GPL(vlimit_name);
++
++#define MASK_ENTRY(x) (1 << (x))
++
++const struct vcmd_ctx_rlimit_mask_v0 vlimit_mask = {
++ /* minimum */
++ 0
++ , /* softlimit */
++ MASK_ENTRY( RLIMIT_RSS ) |
++ MASK_ENTRY( VLIMIT_ANON ) |
++ 0
++ , /* maximum */
++ MASK_ENTRY( RLIMIT_RSS ) |
++ MASK_ENTRY( RLIMIT_NPROC ) |
++ MASK_ENTRY( RLIMIT_NOFILE ) |
++ MASK_ENTRY( RLIMIT_MEMLOCK ) |
++ MASK_ENTRY( RLIMIT_AS ) |
++ MASK_ENTRY( RLIMIT_LOCKS ) |
++ MASK_ENTRY( RLIMIT_MSGQUEUE ) |
++
++ MASK_ENTRY( VLIMIT_NSOCK ) |
++ MASK_ENTRY( VLIMIT_OPENFD ) |
++ MASK_ENTRY( VLIMIT_ANON ) |
++ MASK_ENTRY( VLIMIT_SHMEM ) |
++ MASK_ENTRY( VLIMIT_DENTRY ) |
++ 0
++};
++ /* accounting only */
++uint32_t account_mask =
++ MASK_ENTRY( VLIMIT_SEMARY ) |
++ MASK_ENTRY( VLIMIT_NSEMS ) |
++ MASK_ENTRY( VLIMIT_MAPPED ) |
++ 0;
++
++
++static int is_valid_vlimit(int id)
++{
++ uint32_t mask = vlimit_mask.minimum |
++ vlimit_mask.softlimit | vlimit_mask.maximum;
++ return mask & (1 << id);
++}
++
++static int is_accounted_vlimit(int id)
++{
++ if (is_valid_vlimit(id))
++ return 1;
++ return account_mask & (1 << id);
++}
++
++
++static inline uint64_t vc_get_soft(struct vx_info *vxi, int id)
++{
++ rlim_t limit = __rlim_soft(&vxi->limit, id);
++ return VX_VLIM(limit);
++}
++
++static inline uint64_t vc_get_hard(struct vx_info *vxi, int id)
++{
++ rlim_t limit = __rlim_hard(&vxi->limit, id);
++ return VX_VLIM(limit);
++}
++
++static int do_get_rlimit(struct vx_info *vxi, uint32_t id,
++ uint64_t *minimum, uint64_t *softlimit, uint64_t *maximum)
++{
++ if (!is_valid_vlimit(id))
++ return -EINVAL;
++
++ if (minimum)
++ *minimum = CRLIM_UNSET;
++ if (softlimit)
++ *softlimit = vc_get_soft(vxi, id);
++ if (maximum)
++ *maximum = vc_get_hard(vxi, id);
++ return 0;
++}
++
++int vc_get_rlimit(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_rlimit_v0 vc_data;
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_get_rlimit(vxi, vc_data.id,
++ &vc_data.minimum, &vc_data.softlimit, &vc_data.maximum);
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++static int do_set_rlimit(struct vx_info *vxi, uint32_t id,
++ uint64_t minimum, uint64_t softlimit, uint64_t maximum)
++{
++ if (!is_valid_vlimit(id))
++ return -EINVAL;
++
++ if (maximum != CRLIM_KEEP)
++ __rlim_hard(&vxi->limit, id) = VX_RLIM(maximum);
++ if (softlimit != CRLIM_KEEP)
++ __rlim_soft(&vxi->limit, id) = VX_RLIM(softlimit);
++
++ /* clamp soft limit */
++ if (__rlim_soft(&vxi->limit, id) > __rlim_hard(&vxi->limit, id))
++ __rlim_soft(&vxi->limit, id) = __rlim_hard(&vxi->limit, id);
++
++ return 0;
++}
++
++int vc_set_rlimit(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_rlimit_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_rlimit(vxi, vc_data.id,
++ vc_data.minimum, vc_data.softlimit, vc_data.maximum);
++}
++
++#ifdef CONFIG_IA32_EMULATION
++
++int vc_set_rlimit_x32(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_rlimit_v0_x32 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_rlimit(vxi, vc_data.id,
++ vc_data.minimum, vc_data.softlimit, vc_data.maximum);
++}
++
++int vc_get_rlimit_x32(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_rlimit_v0_x32 vc_data;
++ int ret;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_get_rlimit(vxi, vc_data.id,
++ &vc_data.minimum, &vc_data.softlimit, &vc_data.maximum);
++ if (ret)
++ return ret;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++#endif /* CONFIG_IA32_EMULATION */
++
++
++int vc_get_rlimit_mask(uint32_t id, void __user *data)
++{
++ if (copy_to_user(data, &vlimit_mask, sizeof(vlimit_mask)))
++ return -EFAULT;
++ return 0;
++}
++
++
++static inline void vx_reset_minmax(struct _vx_limit *limit)
++{
++ rlim_t value;
++ int lim;
++
++ for (lim = 0; lim < NUM_LIMITS; lim++) {
++ value = __rlim_get(limit, lim);
++ __rlim_rmax(limit, lim) = value;
++ __rlim_rmin(limit, lim) = value;
++ }
++}
++
++
++int vc_reset_minmax(struct vx_info *vxi, void __user *data)
++{
++ vx_reset_minmax(&vxi->limit);
++ return 0;
++}
++
++
++int vc_rlimit_stat(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_rlimit_stat_v0 vc_data;
++ struct _vx_limit *limit = &vxi->limit;
++ int id;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ id = vc_data.id;
++ if (!is_accounted_vlimit(id))
++ return -EINVAL;
++
++ vx_limit_fixup(limit, id);
++ vc_data.hits = atomic_read(&__rlim_lhit(limit, id));
++ vc_data.value = __rlim_get(limit, id);
++ vc_data.minimum = __rlim_rmin(limit, id);
++ vc_data.maximum = __rlim_rmax(limit, id);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++void vx_vsi_meminfo(struct sysinfo *val)
++{
++ struct vx_info *vxi = current->vx_info;
++ unsigned long totalram, freeram;
++ rlim_t v;
++
++ /* we blindly accept the max */
++ v = __rlim_soft(&vxi->limit, RLIMIT_RSS);
++ totalram = (v != RLIM_INFINITY) ? v : val->totalram;
++
++ /* total minus used equals free */
++ v = __vx_cres_array_fixup(&vxi->limit, VLA_RSS);
++ freeram = (v < totalram) ? totalram - v : 0;
++
++ val->totalram = totalram;
++ val->freeram = freeram;
++ val->bufferram = 0;
++ val->totalhigh = 0;
++ val->freehigh = 0;
++ return;
++}
++
++void vx_vsi_swapinfo(struct sysinfo *val)
++{
++ struct vx_info *vxi = current->vx_info;
++ unsigned long totalswap, freeswap;
++ rlim_t v, w;
++
++ v = __rlim_soft(&vxi->limit, RLIMIT_RSS);
++ if (v == RLIM_INFINITY) {
++ val->freeswap = val->totalswap;
++ return;
++ }
++
++ /* we blindly accept the max */
++ w = __rlim_hard(&vxi->limit, RLIMIT_RSS);
++ totalswap = (w != RLIM_INFINITY) ? (w - v) : val->totalswap;
++
++ /* currently 'used' swap */
++ w = __vx_cres_array_fixup(&vxi->limit, VLA_RSS);
++ w -= (w > v) ? v : w;
++
++ /* total minus used equals free */
++ freeswap = (w < totalswap) ? totalswap - w : 0;
++
++ val->totalswap = totalswap;
++ val->freeswap = freeswap;
++ return;
++}
++
++
++unsigned long vx_badness(struct task_struct *task, struct mm_struct *mm)
++{
++ struct vx_info *vxi = mm->mm_vx_info;
++ unsigned long points;
++ rlim_t v, w;
++
++ if (!vxi)
++ return 0;
++
++ points = vxi->vx_badness_bias;
++
++ v = __vx_cres_array_fixup(&vxi->limit, VLA_RSS);
++ w = __rlim_soft(&vxi->limit, RLIMIT_RSS);
++ points += (v > w) ? (v - w) : 0;
++
++ return points;
++}
++
+--- a/kernel/vserver/limit_init.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/limit_init.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,31 @@
++
++
++static inline void vx_info_init_limit(struct _vx_limit *limit)
++{
++ int lim;
++
++ for (lim = 0; lim < NUM_LIMITS; lim++) {
++ __rlim_soft(limit, lim) = RLIM_INFINITY;
++ __rlim_hard(limit, lim) = RLIM_INFINITY;
++ __rlim_set(limit, lim, 0);
++ atomic_set(&__rlim_lhit(limit, lim), 0);
++ __rlim_rmin(limit, lim) = 0;
++ __rlim_rmax(limit, lim) = 0;
++ }
++}
++
++static inline void vx_info_exit_limit(struct _vx_limit *limit)
++{
++ rlim_t value;
++ int lim;
++
++ for (lim = 0; lim < NUM_LIMITS; lim++) {
++ if ((1 << lim) & VLIM_NOCHECK)
++ continue;
++ value = __rlim_get(limit, lim);
++ vxwprintk_xid(value,
++ "!!! limit: %p[%s,%d] = %ld on exit.",
++ limit, vlimit_name[lim], lim, (long)value);
++ }
++}
++
+--- a/kernel/vserver/limit_proc.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/limit_proc.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,57 @@
++#ifndef _VX_LIMIT_PROC_H
++#define _VX_LIMIT_PROC_H
++
++#include <linux/vserver/limit_int.h>
++
++
++#define VX_LIMIT_FMT ":\t%8ld\t%8ld/%8ld\t%8lld/%8lld\t%6d\n"
++#define VX_LIMIT_TOP \
++ "Limit\t current\t min/max\t\t soft/hard\t\thits\n"
++
++#define VX_LIMIT_ARG(r) \
++ (unsigned long)__rlim_get(limit, r), \
++ (unsigned long)__rlim_rmin(limit, r), \
++ (unsigned long)__rlim_rmax(limit, r), \
++ VX_VLIM(__rlim_soft(limit, r)), \
++ VX_VLIM(__rlim_hard(limit, r)), \
++ atomic_read(&__rlim_lhit(limit, r))
++
++static inline int vx_info_proc_limit(struct _vx_limit *limit, char *buffer)
++{
++ vx_limit_fixup(limit, -1);
++ return sprintf(buffer, VX_LIMIT_TOP
++ "PROC" VX_LIMIT_FMT
++ "VM" VX_LIMIT_FMT
++ "VML" VX_LIMIT_FMT
++ "RSS" VX_LIMIT_FMT
++ "ANON" VX_LIMIT_FMT
++ "RMAP" VX_LIMIT_FMT
++ "FILES" VX_LIMIT_FMT
++ "OFD" VX_LIMIT_FMT
++ "LOCKS" VX_LIMIT_FMT
++ "SOCK" VX_LIMIT_FMT
++ "MSGQ" VX_LIMIT_FMT
++ "SHM" VX_LIMIT_FMT
++ "SEMA" VX_LIMIT_FMT
++ "SEMS" VX_LIMIT_FMT
++ "DENT" VX_LIMIT_FMT,
++ VX_LIMIT_ARG(RLIMIT_NPROC),
++ VX_LIMIT_ARG(RLIMIT_AS),
++ VX_LIMIT_ARG(RLIMIT_MEMLOCK),
++ VX_LIMIT_ARG(RLIMIT_RSS),
++ VX_LIMIT_ARG(VLIMIT_ANON),
++ VX_LIMIT_ARG(VLIMIT_MAPPED),
++ VX_LIMIT_ARG(RLIMIT_NOFILE),
++ VX_LIMIT_ARG(VLIMIT_OPENFD),
++ VX_LIMIT_ARG(RLIMIT_LOCKS),
++ VX_LIMIT_ARG(VLIMIT_NSOCK),
++ VX_LIMIT_ARG(RLIMIT_MSGQUEUE),
++ VX_LIMIT_ARG(VLIMIT_SHMEM),
++ VX_LIMIT_ARG(VLIMIT_SEMARY),
++ VX_LIMIT_ARG(VLIMIT_NSEMS),
++ VX_LIMIT_ARG(VLIMIT_DENTRY));
++}
++
++#endif /* _VX_LIMIT_PROC_H */
++
++
+--- a/kernel/vserver/Makefile 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/Makefile 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,18 @@
++#
++# Makefile for the Linux vserver routines.
++#
++
++
++obj-y += vserver.o
++
++vserver-y := switch.o context.o space.o sched.o network.o inode.o \
++ limit.o cvirt.o cacct.o signal.o helper.o init.o \
++ dlimit.o tag.o
++
++vserver-$(CONFIG_INET) += inet.o
++vserver-$(CONFIG_PROC_FS) += proc.o
++vserver-$(CONFIG_VSERVER_DEBUG) += sysctl.o debug.o
++vserver-$(CONFIG_VSERVER_HISTORY) += history.o
++vserver-$(CONFIG_VSERVER_MONITOR) += monitor.o
++vserver-$(CONFIG_VSERVER_DEVICE) += device.o
++
+--- a/kernel/vserver/monitor.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/monitor.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,138 @@
++/*
++ * kernel/vserver/monitor.c
++ *
++ * Virtual Context Scheduler Monitor
++ *
++ * Copyright (C) 2006-2007 Herbert Pötzl
++ *
++ * V0.01 basic design
++ *
++ */
++
++#include <linux/module.h>
++#include <linux/jiffies.h>
++#include <asm/uaccess.h>
++#include <asm/atomic.h>
++
++#include <linux/vserver/monitor.h>
++#include <linux/vserver/debug_cmd.h>
++
++
++#ifdef CONFIG_VSERVER_MONITOR
++#define VXM_SIZE CONFIG_VSERVER_MONITOR_SIZE
++#else
++#define VXM_SIZE 64
++#endif
++
++struct _vx_monitor {
++ unsigned int counter;
++
++ struct _vx_mon_entry entry[VXM_SIZE+1];
++};
++
++
++DEFINE_PER_CPU(struct _vx_monitor, vx_monitor_buffer);
++
++unsigned volatile int vxm_active = 1;
++
++static atomic_t sequence = ATOMIC_INIT(0);
++
++
++/* vxm_advance()
++
++ * requires disabled preemption */
++
++struct _vx_mon_entry *vxm_advance(int cpu)
++{
++ struct _vx_monitor *mon = &per_cpu(vx_monitor_buffer, cpu);
++ struct _vx_mon_entry *entry;
++ unsigned int index;
++
++ index = vxm_active ? (mon->counter++ % VXM_SIZE) : VXM_SIZE;
++ entry = &mon->entry[index];
++
++ entry->ev.seq = atomic_inc_return(&sequence);
++ entry->ev.jif = jiffies;
++ return entry;
++}
++
++EXPORT_SYMBOL_GPL(vxm_advance);
++
++
++int do_read_monitor(struct __user _vx_mon_entry *data,
++ int cpu, uint32_t *index, uint32_t *count)
++{
++ int pos, ret = 0;
++ struct _vx_monitor *mon = &per_cpu(vx_monitor_buffer, cpu);
++ int end = mon->counter;
++ int start = end - VXM_SIZE + 2;
++ int idx = *index;
++
++ /* special case: get current pos */
++ if (!*count) {
++ *index = end;
++ return 0;
++ }
++
++ /* have we lost some data? */
++ if (idx < start)
++ idx = start;
++
++ for (pos = 0; (pos < *count) && (idx < end); pos++, idx++) {
++ struct _vx_mon_entry *entry =
++ &mon->entry[idx % VXM_SIZE];
++
++ /* send entry to userspace */
++ ret = copy_to_user(&data[pos], entry, sizeof(*entry));
++ if (ret)
++ break;
++ }
++ /* save new index and count */
++ *index = idx;
++ *count = pos;
++ return ret ? ret : (*index < end);
++}
++
++int vc_read_monitor(uint32_t id, void __user *data)
++{
++ struct vcmd_read_monitor_v0 vc_data;
++ int ret;
++
++ if (id >= NR_CPUS)
++ return -EINVAL;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_read_monitor((struct __user _vx_mon_entry *)vc_data.data,
++ id, &vc_data.index, &vc_data.count);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return ret;
++}
++
++#ifdef CONFIG_COMPAT
++
++int vc_read_monitor_x32(uint32_t id, void __user *data)
++{
++ struct vcmd_read_monitor_v0_x32 vc_data;
++ int ret;
++
++ if (id >= NR_CPUS)
++ return -EINVAL;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ ret = do_read_monitor((struct __user _vx_mon_entry *)
++ compat_ptr(vc_data.data_ptr),
++ id, &vc_data.index, &vc_data.count);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return ret;
++}
++
++#endif /* CONFIG_COMPAT */
++
+--- a/kernel/vserver/network.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/network.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,864 @@
++/*
++ * linux/kernel/vserver/network.c
++ *
++ * Virtual Server: Network Support
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 broken out from vcontext V0.05
++ * V0.02 cleaned up implementation
++ * V0.03 added equiv nx commands
++ * V0.04 switch to RCU based hash
++ * V0.05 and back to locking again
++ * V0.06 changed vcmds to nxi arg
++ * V0.07 have __create claim() the nxi
++ *
++ */
++
++#include <linux/err.h>
++#include <linux/slab.h>
++#include <linux/rcupdate.h>
++
++#include <linux/vs_network.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/network_cmd.h>
++
++
++atomic_t nx_global_ctotal = ATOMIC_INIT(0);
++atomic_t nx_global_cactive = ATOMIC_INIT(0);
++
++static struct kmem_cache *nx_addr_v4_cachep = NULL;
++static struct kmem_cache *nx_addr_v6_cachep = NULL;
++
++
++static int __init init_network(void)
++{
++ nx_addr_v4_cachep = kmem_cache_create("nx_v4_addr_cache",
++ sizeof(struct nx_addr_v4), 0,
++ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
++ nx_addr_v6_cachep = kmem_cache_create("nx_v6_addr_cache",
++ sizeof(struct nx_addr_v6), 0,
++ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
++ return 0;
++}
++
++
++/* __alloc_nx_addr_v4() */
++
++static inline struct nx_addr_v4 *__alloc_nx_addr_v4(void)
++{
++ struct nx_addr_v4 *nxa = kmem_cache_alloc(
++ nx_addr_v4_cachep, GFP_KERNEL);
++
++ if (!IS_ERR(nxa))
++ memset(nxa, 0, sizeof(*nxa));
++ return nxa;
++}
++
++/* __dealloc_nx_addr_v4() */
++
++static inline void __dealloc_nx_addr_v4(struct nx_addr_v4 *nxa)
++{
++ kmem_cache_free(nx_addr_v4_cachep, nxa);
++}
++
++/* __dealloc_nx_addr_v4_all() */
++
++static inline void __dealloc_nx_addr_v4_all(struct nx_addr_v4 *nxa)
++{
++ while (nxa) {
++ struct nx_addr_v4 *next = nxa->next;
++
++ __dealloc_nx_addr_v4(nxa);
++ nxa = next;
++ }
++}
++
++
++#ifdef CONFIG_IPV6
++
++/* __alloc_nx_addr_v6() */
++
++static inline struct nx_addr_v6 *__alloc_nx_addr_v6(void)
++{
++ struct nx_addr_v6 *nxa = kmem_cache_alloc(
++ nx_addr_v6_cachep, GFP_KERNEL);
++
++ if (!IS_ERR(nxa))
++ memset(nxa, 0, sizeof(*nxa));
++ return nxa;
++}
++
++/* __dealloc_nx_addr_v6() */
++
++static inline void __dealloc_nx_addr_v6(struct nx_addr_v6 *nxa)
++{
++ kmem_cache_free(nx_addr_v6_cachep, nxa);
++}
++
++/* __dealloc_nx_addr_v6_all() */
++
++static inline void __dealloc_nx_addr_v6_all(struct nx_addr_v6 *nxa)
++{
++ while (nxa) {
++ struct nx_addr_v6 *next = nxa->next;
++
++ __dealloc_nx_addr_v6(nxa);
++ nxa = next;
++ }
++}
++
++#endif /* CONFIG_IPV6 */
++
++/* __alloc_nx_info()
++
++ * allocate an initialized nx_info struct
++ * doesn't make it visible (hash) */
++
++static struct nx_info *__alloc_nx_info(nid_t nid)
++{
++ struct nx_info *new = NULL;
++
++ vxdprintk(VXD_CBIT(nid, 1), "alloc_nx_info(%d)*", nid);
++
++ /* would this benefit from a slab cache? */
++ new = kmalloc(sizeof(struct nx_info), GFP_KERNEL);
++ if (!new)
++ return 0;
++
++ memset(new, 0, sizeof(struct nx_info));
++ new->nx_id = nid;
++ INIT_HLIST_NODE(&new->nx_hlist);
++ atomic_set(&new->nx_usecnt, 0);
++ atomic_set(&new->nx_tasks, 0);
++ new->nx_state = 0;
++
++ new->nx_flags = NXF_INIT_SET;
++
++ /* rest of init goes here */
++
++ new->v4_lback.s_addr = htonl(INADDR_LOOPBACK);
++ new->v4_bcast.s_addr = htonl(INADDR_BROADCAST);
++
++ vxdprintk(VXD_CBIT(nid, 0),
++ "alloc_nx_info(%d) = %p", nid, new);
++ atomic_inc(&nx_global_ctotal);
++ return new;
++}
++
++/* __dealloc_nx_info()
++
++ * final disposal of nx_info */
++
++static void __dealloc_nx_info(struct nx_info *nxi)
++{
++ vxdprintk(VXD_CBIT(nid, 0),
++ "dealloc_nx_info(%p)", nxi);
++
++ nxi->nx_hlist.next = LIST_POISON1;
++ nxi->nx_id = -1;
++
++ BUG_ON(atomic_read(&nxi->nx_usecnt));
++ BUG_ON(atomic_read(&nxi->nx_tasks));
++
++ __dealloc_nx_addr_v4_all(nxi->v4.next);
++
++ nxi->nx_state |= NXS_RELEASED;
++ kfree(nxi);
++ atomic_dec(&nx_global_ctotal);
++}
++
++static void __shutdown_nx_info(struct nx_info *nxi)
++{
++ nxi->nx_state |= NXS_SHUTDOWN;
++ vs_net_change(nxi, VSC_NETDOWN);
++}
++
++/* exported stuff */
++
++void free_nx_info(struct nx_info *nxi)
++{
++ /* context shutdown is mandatory */
++ BUG_ON(nxi->nx_state != NXS_SHUTDOWN);
++
++ /* context must not be hashed */
++ BUG_ON(nxi->nx_state & NXS_HASHED);
++
++ BUG_ON(atomic_read(&nxi->nx_usecnt));
++ BUG_ON(atomic_read(&nxi->nx_tasks));
++
++ __dealloc_nx_info(nxi);
++}
++
++
++void __nx_set_lback(struct nx_info *nxi)
++{
++ int nid = nxi->nx_id;
++ __be32 lback = htonl(INADDR_LOOPBACK ^ ((nid & 0xFFFF) << 8));
++
++ nxi->v4_lback.s_addr = lback;
++}
++
++extern int __nx_inet_add_lback(__be32 addr);
++extern int __nx_inet_del_lback(__be32 addr);
++
++
++/* hash table for nx_info hash */
++
++#define NX_HASH_SIZE 13
++
++struct hlist_head nx_info_hash[NX_HASH_SIZE];
++
++static spinlock_t nx_info_hash_lock = SPIN_LOCK_UNLOCKED;
++
++
++static inline unsigned int __hashval(nid_t nid)
++{
++ return (nid % NX_HASH_SIZE);
++}
++
++
++
++/* __hash_nx_info()
++
++ * add the nxi to the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __hash_nx_info(struct nx_info *nxi)
++{
++ struct hlist_head *head;
++
++ vxd_assert_lock(&nx_info_hash_lock);
++ vxdprintk(VXD_CBIT(nid, 4),
++ "__hash_nx_info: %p[#%d]", nxi, nxi->nx_id);
++
++ /* context must not be hashed */
++ BUG_ON(nx_info_state(nxi, NXS_HASHED));
++
++ nxi->nx_state |= NXS_HASHED;
++ head = &nx_info_hash[__hashval(nxi->nx_id)];
++ hlist_add_head(&nxi->nx_hlist, head);
++ atomic_inc(&nx_global_cactive);
++}
++
++/* __unhash_nx_info()
++
++ * remove the nxi from the global hash table
++ * requires the hash_lock to be held */
++
++static inline void __unhash_nx_info(struct nx_info *nxi)
++{
++ vxd_assert_lock(&nx_info_hash_lock);
++ vxdprintk(VXD_CBIT(nid, 4),
++ "__unhash_nx_info: %p[#%d.%d.%d]", nxi, nxi->nx_id,
++ atomic_read(&nxi->nx_usecnt), atomic_read(&nxi->nx_tasks));
++
++ /* context must be hashed */
++ BUG_ON(!nx_info_state(nxi, NXS_HASHED));
++ /* but without tasks */
++ BUG_ON(atomic_read(&nxi->nx_tasks));
++
++ nxi->nx_state &= ~NXS_HASHED;
++ hlist_del(&nxi->nx_hlist);
++ atomic_dec(&nx_global_cactive);
++}
++
++
++/* __lookup_nx_info()
++
++ * requires the hash_lock to be held
++ * doesn't increment the nx_refcnt */
++
++static inline struct nx_info *__lookup_nx_info(nid_t nid)
++{
++ struct hlist_head *head = &nx_info_hash[__hashval(nid)];
++ struct hlist_node *pos;
++ struct nx_info *nxi;
++
++ vxd_assert_lock(&nx_info_hash_lock);
++ hlist_for_each(pos, head) {
++ nxi = hlist_entry(pos, struct nx_info, nx_hlist);
++
++ if (nxi->nx_id == nid)
++ goto found;
++ }
++ nxi = NULL;
++found:
++ vxdprintk(VXD_CBIT(nid, 0),
++ "__lookup_nx_info(#%u): %p[#%u]",
++ nid, nxi, nxi ? nxi->nx_id : 0);
++ return nxi;
++}
++
++
++/* __create_nx_info()
++
++ * create the requested context
++ * get(), claim() and hash it */
++
++static struct nx_info *__create_nx_info(int id)
++{
++ struct nx_info *new, *nxi = NULL;
++
++ vxdprintk(VXD_CBIT(nid, 1), "create_nx_info(%d)*", id);
++
++ if (!(new = __alloc_nx_info(id)))
++ return ERR_PTR(-ENOMEM);
++
++ /* required to make dynamic xids unique */
++ spin_lock(&nx_info_hash_lock);
++
++ /* static context requested */
++ if ((nxi = __lookup_nx_info(id))) {
++ vxdprintk(VXD_CBIT(nid, 0),
++ "create_nx_info(%d) = %p (already there)", id, nxi);
++ if (nx_info_flags(nxi, NXF_STATE_SETUP, 0))
++ nxi = ERR_PTR(-EBUSY);
++ else
++ nxi = ERR_PTR(-EEXIST);
++ goto out_unlock;
++ }
++ /* new context */
++ vxdprintk(VXD_CBIT(nid, 0),
++ "create_nx_info(%d) = %p (new)", id, new);
++ claim_nx_info(new, NULL);
++ __nx_set_lback(new);
++ __hash_nx_info(get_nx_info(new));
++ nxi = new, new = NULL;
++
++out_unlock:
++ spin_unlock(&nx_info_hash_lock);
++ if (new)
++ __dealloc_nx_info(new);
++ return nxi;
++}
++
++
++
++/* exported stuff */
++
++
++void unhash_nx_info(struct nx_info *nxi)
++{
++ __shutdown_nx_info(nxi);
++ spin_lock(&nx_info_hash_lock);
++ __unhash_nx_info(nxi);
++ spin_unlock(&nx_info_hash_lock);
++}
++
++/* lookup_nx_info()
++
++ * search for a nx_info and get() it
++ * negative id means current */
++
++struct nx_info *lookup_nx_info(int id)
++{
++ struct nx_info *nxi = NULL;
++
++ if (id < 0) {
++ nxi = get_nx_info(current->nx_info);
++ } else if (id > 1) {
++ spin_lock(&nx_info_hash_lock);
++ nxi = get_nx_info(__lookup_nx_info(id));
++ spin_unlock(&nx_info_hash_lock);
++ }
++ return nxi;
++}
++
++/* nid_is_hashed()
++
++ * verify that nid is still hashed */
++
++int nid_is_hashed(nid_t nid)
++{
++ int hashed;
++
++ spin_lock(&nx_info_hash_lock);
++ hashed = (__lookup_nx_info(nid) != NULL);
++ spin_unlock(&nx_info_hash_lock);
++ return hashed;
++}
++
++
++#ifdef CONFIG_PROC_FS
++
++/* get_nid_list()
++
++ * get a subset of hashed nids for proc
++ * assumes size is at least one */
++
++int get_nid_list(int index, unsigned int *nids, int size)
++{
++ int hindex, nr_nids = 0;
++
++ /* only show current and children */
++ if (!nx_check(0, VS_ADMIN | VS_WATCH)) {
++ if (index > 0)
++ return 0;
++ nids[nr_nids] = nx_current_nid();
++ return 1;
++ }
++
++ for (hindex = 0; hindex < NX_HASH_SIZE; hindex++) {
++ struct hlist_head *head = &nx_info_hash[hindex];
++ struct hlist_node *pos;
++
++ spin_lock(&nx_info_hash_lock);
++ hlist_for_each(pos, head) {
++ struct nx_info *nxi;
++
++ if (--index > 0)
++ continue;
++
++ nxi = hlist_entry(pos, struct nx_info, nx_hlist);
++ nids[nr_nids] = nxi->nx_id;
++ if (++nr_nids >= size) {
++ spin_unlock(&nx_info_hash_lock);
++ goto out;
++ }
++ }
++ /* keep the lock time short */
++ spin_unlock(&nx_info_hash_lock);
++ }
++out:
++ return nr_nids;
++}
++#endif
++
++
++/*
++ * migrate task to new network
++ * gets nxi, puts old_nxi on change
++ */
++
++int nx_migrate_task(struct task_struct *p, struct nx_info *nxi)
++{
++ struct nx_info *old_nxi;
++ int ret = 0;
++
++ if (!p || !nxi)
++ BUG();
++
++ vxdprintk(VXD_CBIT(nid, 5),
++ "nx_migrate_task(%p,%p[#%d.%d.%d])",
++ p, nxi, nxi->nx_id,
++ atomic_read(&nxi->nx_usecnt),
++ atomic_read(&nxi->nx_tasks));
++
++ if (nx_info_flags(nxi, NXF_INFO_PRIVATE, 0) &&
++ !nx_info_flags(nxi, NXF_STATE_SETUP, 0))
++ return -EACCES;
++
++ if (nx_info_state(nxi, NXS_SHUTDOWN))
++ return -EFAULT;
++
++ /* maybe disallow this completely? */
++ old_nxi = task_get_nx_info(p);
++ if (old_nxi == nxi)
++ goto out;
++
++ task_lock(p);
++ if (old_nxi)
++ clr_nx_info(&p->nx_info);
++ claim_nx_info(nxi, p);
++ set_nx_info(&p->nx_info, nxi);
++ p->nid = nxi->nx_id;
++ task_unlock(p);
++
++ vxdprintk(VXD_CBIT(nid, 5),
++ "moved task %p into nxi:%p[#%d]",
++ p, nxi, nxi->nx_id);
++
++ if (old_nxi)
++ release_nx_info(old_nxi, p);
++ ret = 0;
++out:
++ put_nx_info(old_nxi);
++ return ret;
++}
++
++
++void nx_set_persistent(struct nx_info *nxi)
++{
++ vxdprintk(VXD_CBIT(nid, 6),
++ "nx_set_persistent(%p[#%d])", nxi, nxi->nx_id);
++
++ get_nx_info(nxi);
++ claim_nx_info(nxi, NULL);
++}
++
++void nx_clear_persistent(struct nx_info *nxi)
++{
++ vxdprintk(VXD_CBIT(nid, 6),
++ "nx_clear_persistent(%p[#%d])", nxi, nxi->nx_id);
++
++ release_nx_info(nxi, NULL);
++ put_nx_info(nxi);
++}
++
++void nx_update_persistent(struct nx_info *nxi)
++{
++ if (nx_info_flags(nxi, NXF_PERSISTENT, 0))
++ nx_set_persistent(nxi);
++ else
++ nx_clear_persistent(nxi);
++}
++
++/* vserver syscall commands below here */
++
++/* taks nid and nx_info functions */
++
++#include <asm/uaccess.h>
++
++
++int vc_task_nid(uint32_t id)
++{
++ nid_t nid;
++
++ if (id) {
++ struct task_struct *tsk;
++
++ read_lock(&tasklist_lock);
++ tsk = find_task_by_real_pid(id);
++ nid = (tsk) ? tsk->nid : -ESRCH;
++ read_unlock(&tasklist_lock);
++ } else
++ nid = nx_current_nid();
++ return nid;
++}
++
++
++int vc_nx_info(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_nx_info_v0 vc_data;
++
++ vc_data.nid = nxi->nx_id;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++/* network functions */
++
++int vc_net_create(uint32_t nid, void __user *data)
++{
++ struct vcmd_net_create vc_data = { .flagword = NXF_INIT_SET };
++ struct nx_info *new_nxi;
++ int ret;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ if ((nid > MAX_S_CONTEXT) || (nid < 2))
++ return -EINVAL;
++
++ new_nxi = __create_nx_info(nid);
++ if (IS_ERR(new_nxi))
++ return PTR_ERR(new_nxi);
++
++ /* initial flags */
++ new_nxi->nx_flags = vc_data.flagword;
++
++ ret = -ENOEXEC;
++ if (vs_net_change(new_nxi, VSC_NETUP))
++ goto out;
++
++ ret = nx_migrate_task(current, new_nxi);
++ if (ret)
++ goto out;
++
++ /* return context id on success */
++ ret = new_nxi->nx_id;
++
++ /* get a reference for persistent contexts */
++ if ((vc_data.flagword & NXF_PERSISTENT))
++ nx_set_persistent(new_nxi);
++out:
++ release_nx_info(new_nxi, NULL);
++ put_nx_info(new_nxi);
++ return ret;
++}
++
++
++int vc_net_migrate(struct nx_info *nxi, void __user *data)
++{
++ return nx_migrate_task(current, nxi);
++}
++
++
++
++int do_add_v4_addr(struct nx_info *nxi, __be32 ip, __be32 ip2, __be32 mask,
++ uint16_t type, uint16_t flags)
++{
++ struct nx_addr_v4 *nxa = &nxi->v4;
++
++ if (NX_IPV4(nxi)) {
++ /* locate last entry */
++ for (; nxa->next; nxa = nxa->next);
++ nxa->next = __alloc_nx_addr_v4();
++ nxa = nxa->next;
++
++ if (IS_ERR(nxa))
++ return PTR_ERR(nxa);
++ }
++
++ if (nxi->v4.next)
++ /* remove single ip for ip list */
++ nxi->nx_flags &= ~NXF_SINGLE_IP;
++
++ nxa->ip[0].s_addr = ip;
++ nxa->ip[1].s_addr = ip2;
++ nxa->mask.s_addr = mask;
++ nxa->type = type;
++ nxa->flags = flags;
++ return 0;
++}
++
++
++int vc_net_add(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_v0 vc_data;
++ int index, ret = 0;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++ case NXA_TYPE_IPV4:
++ if ((vc_data.count < 1) || (vc_data.count > 4))
++ return -EINVAL;
++
++ index = 0;
++ while (index < vc_data.count) {
++ ret = do_add_v4_addr(nxi, vc_data.ip[index].s_addr, 0,
++ vc_data.mask[index].s_addr, NXA_TYPE_ADDR, 0);
++ if (ret)
++ return ret;
++ index++;
++ }
++ ret = index;
++ break;
++
++ case NXA_TYPE_IPV4|NXA_MOD_BCAST:
++ nxi->v4_bcast = vc_data.ip[0];
++ ret = 1;
++ break;
++
++ case NXA_TYPE_IPV4|NXA_MOD_LBACK:
++ nxi->v4_lback = vc_data.ip[0];
++ ret = 1;
++ break;
++
++ default:
++ ret = -EINVAL;
++ break;
++ }
++ return ret;
++}
++
++int vc_net_remove(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_v0 vc_data;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++ case NXA_TYPE_ANY:
++ __dealloc_nx_addr_v4_all(xchg(&nxi->v4.next, NULL));
++ memset(&nxi->v4, 0, sizeof(nxi->v4));
++ break;
++
++ default:
++ return -EINVAL;
++ }
++ return 0;
++}
++
++
++int vc_net_add_ipv4(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_ipv4_v1 vc_data;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++ case NXA_TYPE_ADDR:
++ case NXA_TYPE_RANGE:
++ case NXA_TYPE_MASK:
++ return do_add_v4_addr(nxi, vc_data.ip.s_addr, 0,
++ vc_data.mask.s_addr, vc_data.type, vc_data.flags);
++
++ case NXA_TYPE_ADDR | NXA_MOD_BCAST:
++ nxi->v4_bcast = vc_data.ip;
++ break;
++
++ case NXA_TYPE_ADDR | NXA_MOD_LBACK:
++ nxi->v4_lback = vc_data.ip;
++ break;
++
++ default:
++ return -EINVAL;
++ }
++ return 0;
++}
++
++int vc_net_remove_ipv4(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_ipv4_v1 vc_data;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++/* case NXA_TYPE_ADDR:
++ break; */
++
++ case NXA_TYPE_ANY:
++ __dealloc_nx_addr_v4_all(xchg(&nxi->v4.next, NULL));
++ memset(&nxi->v4, 0, sizeof(nxi->v4));
++ break;
++
++ default:
++ return -EINVAL;
++ }
++ return 0;
++}
++
++
++#ifdef CONFIG_IPV6
++
++int do_add_v6_addr(struct nx_info *nxi,
++ struct in6_addr *ip, struct in6_addr *mask,
++ uint32_t prefix, uint16_t type, uint16_t flags)
++{
++ struct nx_addr_v6 *nxa = &nxi->v6;
++
++ if (NX_IPV6(nxi)) {
++ /* locate last entry */
++ for (; nxa->next; nxa = nxa->next);
++ nxa->next = __alloc_nx_addr_v6();
++ nxa = nxa->next;
++
++ if (IS_ERR(nxa))
++ return PTR_ERR(nxa);
++ }
++
++ nxa->ip = *ip;
++ nxa->mask = *mask;
++ nxa->prefix = prefix;
++ nxa->type = type;
++ nxa->flags = flags;
++ return 0;
++}
++
++
++int vc_net_add_ipv6(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_ipv6_v1 vc_data;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++ case NXA_TYPE_ADDR:
++ case NXA_TYPE_MASK:
++ return do_add_v6_addr(nxi, &vc_data.ip, &vc_data.mask,
++ vc_data.prefix, vc_data.type, vc_data.flags);
++ default:
++ return -EINVAL;
++ }
++ return 0;
++}
++
++int vc_net_remove_ipv6(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_addr_ipv6_v1 vc_data;
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ switch (vc_data.type) {
++ case NXA_TYPE_ANY:
++ __dealloc_nx_addr_v6_all(xchg(&nxi->v6.next, NULL));
++ memset(&nxi->v6, 0, sizeof(nxi->v6));
++ break;
++
++ default:
++ return -EINVAL;
++ }
++ return 0;
++}
++
++#endif /* CONFIG_IPV6 */
++
++
++int vc_get_nflags(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_flags_v0 vc_data;
++
++ vc_data.flagword = nxi->nx_flags;
++
++ /* special STATE flag handling */
++ vc_data.mask = vs_mask_flags(~0ULL, nxi->nx_flags, NXF_ONE_TIME);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++int vc_set_nflags(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_flags_v0 vc_data;
++ uint64_t mask, trigger;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ /* special STATE flag handling */
++ mask = vs_mask_mask(vc_data.mask, nxi->nx_flags, NXF_ONE_TIME);
++ trigger = (mask & nxi->nx_flags) ^ (mask & vc_data.flagword);
++
++ nxi->nx_flags = vs_mask_flags(nxi->nx_flags,
++ vc_data.flagword, mask);
++ if (trigger & NXF_PERSISTENT)
++ nx_update_persistent(nxi);
++
++ return 0;
++}
++
++int vc_get_ncaps(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_caps_v0 vc_data;
++
++ vc_data.ncaps = nxi->nx_ncaps;
++ vc_data.cmask = ~0ULL;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++int vc_set_ncaps(struct nx_info *nxi, void __user *data)
++{
++ struct vcmd_net_caps_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ nxi->nx_ncaps = vs_mask_flags(nxi->nx_ncaps,
++ vc_data.ncaps, vc_data.cmask);
++ return 0;
++}
++
++
++#include <linux/module.h>
++
++module_init(init_network);
++
++EXPORT_SYMBOL_GPL(free_nx_info);
++EXPORT_SYMBOL_GPL(unhash_nx_info);
++
+--- a/kernel/vserver/proc.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/proc.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,1086 @@
++/*
++ * linux/kernel/vserver/proc.c
++ *
++ * Virtual Context Support
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 basic structure
++ * V0.02 adaptation vs1.3.0
++ * V0.03 proc permissions
++ * V0.04 locking/generic
++ * V0.05 next generation procfs
++ * V0.06 inode validation
++ * V0.07 generic rewrite vid
++ * V0.08 remove inode type
++ *
++ */
++
++#include <linux/proc_fs.h>
++#include <asm/unistd.h>
++
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vs_cvirt.h>
++
++#include <linux/in.h>
++#include <linux/inetdevice.h>
++#include <linux/vs_inet.h>
++#include <linux/vs_inet6.h>
++
++#include <linux/vserver/global.h>
++
++#include "cvirt_proc.h"
++#include "cacct_proc.h"
++#include "limit_proc.h"
++#include "sched_proc.h"
++#include "vci_config.h"
++
++
++static inline char *print_cap_t(char *buffer, kernel_cap_t *c)
++{
++ unsigned __capi;
++
++ CAP_FOR_EACH_U32(__capi) {
++ buffer += sprintf(buffer, "%08x",
++ c->cap[(_KERNEL_CAPABILITY_U32S-1) - __capi]);
++ }
++ return buffer;
++}
++
++
++static struct proc_dir_entry *proc_virtual;
++
++static struct proc_dir_entry *proc_virtnet;
++
++
++/* first the actual feeds */
++
++
++static int proc_vci(char *buffer)
++{
++ return sprintf(buffer,
++ "VCIVersion:\t%04x:%04x\n"
++ "VCISyscall:\t%d\n"
++ "VCIKernel:\t%08x\n",
++ VCI_VERSION >> 16,
++ VCI_VERSION & 0xFFFF,
++ __NR_vserver,
++ vci_kernel_config());
++}
++
++static int proc_virtual_info(char *buffer)
++{
++ return proc_vci(buffer);
++}
++
++static int proc_virtual_status(char *buffer)
++{
++ return sprintf(buffer,
++ "#CTotal:\t%d\n"
++ "#CActive:\t%d\n"
++ "#NSProxy:\t%d\t%d %d %d %d %d %d\n",
++ atomic_read(&vx_global_ctotal),
++ atomic_read(&vx_global_cactive),
++ atomic_read(&vs_global_nsproxy),
++ atomic_read(&vs_global_fs),
++ atomic_read(&vs_global_mnt_ns),
++ atomic_read(&vs_global_uts_ns),
++ atomic_read(&vs_global_ipc_ns),
++ atomic_read(&vs_global_user_ns),
++ atomic_read(&vs_global_pid_ns));
++}
++
++
++int proc_vxi_info(struct vx_info *vxi, char *buffer)
++{
++ int length;
++
++ length = sprintf(buffer,
++ "ID:\t%d\n"
++ "Info:\t%p\n"
++ "Init:\t%d\n"
++ "OOM:\t%lld\n",
++ vxi->vx_id,
++ vxi,
++ vxi->vx_initpid,
++ vxi->vx_badness_bias);
++ return length;
++}
++
++int proc_vxi_status(struct vx_info *vxi, char *buffer)
++{
++ char *orig = buffer;
++
++ buffer += sprintf(buffer,
++ "UseCnt:\t%d\n"
++ "Tasks:\t%d\n"
++ "Flags:\t%016llx\n",
++ atomic_read(&vxi->vx_usecnt),
++ atomic_read(&vxi->vx_tasks),
++ (unsigned long long)vxi->vx_flags);
++
++ buffer += sprintf(buffer, "BCaps:\t");
++ buffer = print_cap_t(buffer, &vxi->vx_bcaps);
++ buffer += sprintf(buffer, "\n");
++
++ buffer += sprintf(buffer,
++ "CCaps:\t%016llx\n"
++ "Spaces:\t%08lx\n",
++ (unsigned long long)vxi->vx_ccaps,
++ vxi->vx_nsmask);
++ return buffer - orig;
++}
++
++int proc_vxi_limit(struct vx_info *vxi, char *buffer)
++{
++ return vx_info_proc_limit(&vxi->limit, buffer);
++}
++
++int proc_vxi_sched(struct vx_info *vxi, char *buffer)
++{
++ int cpu, length;
++
++ length = vx_info_proc_sched(&vxi->sched, buffer);
++ for_each_online_cpu(cpu) {
++ length += vx_info_proc_sched_pc(
++ &vx_per_cpu(vxi, sched_pc, cpu),
++ buffer + length, cpu);
++ }
++ return length;
++}
++
++int proc_vxi_nsproxy(struct vx_info *vxi, char *buffer)
++{
++ return vx_info_proc_nsproxy(vxi->vx_nsproxy, buffer);
++}
++
++int proc_vxi_cvirt(struct vx_info *vxi, char *buffer)
++{
++ int cpu, length;
++
++ vx_update_load(vxi);
++ length = vx_info_proc_cvirt(&vxi->cvirt, buffer);
++ for_each_online_cpu(cpu) {
++ length += vx_info_proc_cvirt_pc(
++ &vx_per_cpu(vxi, cvirt_pc, cpu),
++ buffer + length, cpu);
++ }
++ return length;
++}
++
++int proc_vxi_cacct(struct vx_info *vxi, char *buffer)
++{
++ return vx_info_proc_cacct(&vxi->cacct, buffer);
++}
++
++
++static int proc_virtnet_info(char *buffer)
++{
++ return proc_vci(buffer);
++}
++
++static int proc_virtnet_status(char *buffer)
++{
++ return sprintf(buffer,
++ "#CTotal:\t%d\n"
++ "#CActive:\t%d\n",
++ atomic_read(&nx_global_ctotal),
++ atomic_read(&nx_global_cactive));
++}
++
++int proc_nxi_info(struct nx_info *nxi, char *buffer)
++{
++ struct nx_addr_v4 *v4a;
++#ifdef CONFIG_IPV6
++ struct nx_addr_v6 *v6a;
++#endif
++ int length, i;
++
++ length = sprintf(buffer,
++ "ID:\t%d\n"
++ "Info:\t%p\n"
++ "Bcast:\t" NIPQUAD_FMT "\n"
++ "Lback:\t" NIPQUAD_FMT "\n",
++ nxi->nx_id,
++ nxi,
++ NIPQUAD(nxi->v4_bcast.s_addr),
++ NIPQUAD(nxi->v4_lback.s_addr));
++
++ if (!NX_IPV4(nxi))
++ goto skip_v4;
++ for (i = 0, v4a = &nxi->v4; v4a; i++, v4a = v4a->next)
++ length += sprintf(buffer + length, "%d:\t" NXAV4_FMT "\n",
++ i, NXAV4(v4a));
++skip_v4:
++#ifdef CONFIG_IPV6
++ if (!NX_IPV6(nxi))
++ goto skip_v6;
++ for (i = 0, v6a = &nxi->v6; v6a; i++, v6a = v6a->next)
++ length += sprintf(buffer + length, "%d:\t" NXAV6_FMT "\n",
++ i, NXAV6(v6a));
++skip_v6:
++#endif
++ return length;
++}
++
++int proc_nxi_status(struct nx_info *nxi, char *buffer)
++{
++ int length;
++
++ length = sprintf(buffer,
++ "UseCnt:\t%d\n"
++ "Tasks:\t%d\n"
++ "Flags:\t%016llx\n"
++ "NCaps:\t%016llx\n",
++ atomic_read(&nxi->nx_usecnt),
++ atomic_read(&nxi->nx_tasks),
++ (unsigned long long)nxi->nx_flags,
++ (unsigned long long)nxi->nx_ncaps);
++ return length;
++}
++
++
++
++/* here the inode helpers */
++
++struct vs_entry {
++ int len;
++ char *name;
++ mode_t mode;
++ struct inode_operations *iop;
++ struct file_operations *fop;
++ union proc_op op;
++};
++
++static struct inode *vs_proc_make_inode(struct super_block *sb, struct vs_entry *p)
++{
++ struct inode *inode = new_inode(sb);
++
++ if (!inode)
++ goto out;
++
++ inode->i_mode = p->mode;
++ if (p->iop)
++ inode->i_op = p->iop;
++ if (p->fop)
++ inode->i_fop = p->fop;
++
++ inode->i_nlink = (p->mode & S_IFDIR) ? 2 : 1;
++ inode->i_flags |= S_IMMUTABLE;
++
++ inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
++
++ inode->i_uid = 0;
++ inode->i_gid = 0;
++ inode->i_tag = 0;
++out:
++ return inode;
++}
++
++static struct dentry *vs_proc_instantiate(struct inode *dir,
++ struct dentry *dentry, int id, void *ptr)
++{
++ struct vs_entry *p = ptr;
++ struct inode *inode = vs_proc_make_inode(dir->i_sb, p);
++ struct dentry *error = ERR_PTR(-EINVAL);
++
++ if (!inode)
++ goto out;
++
++ PROC_I(inode)->op = p->op;
++ PROC_I(inode)->fd = id;
++ d_add(dentry, inode);
++ error = NULL;
++out:
++ return error;
++}
++
++/* Lookups */
++
++typedef struct dentry *instantiate_t(struct inode *, struct dentry *, int, void *);
++
++/*
++ * Fill a directory entry.
++ *
++ * If possible create the dcache entry and derive our inode number and
++ * file type from dcache entry.
++ *
++ * Since all of the proc inode numbers are dynamically generated, the inode
++ * numbers do not exist until the inode is cache. This means creating the
++ * the dcache entry in readdir is necessary to keep the inode numbers
++ * reported by readdir in sync with the inode numbers reported
++ * by stat.
++ */
++static int proc_fill_cache(struct file *filp, void *dirent, filldir_t filldir,
++ char *name, int len, instantiate_t instantiate, int id, void *ptr)
++{
++ struct dentry *child, *dir = filp->f_dentry;
++ struct inode *inode;
++ struct qstr qname;
++ ino_t ino = 0;
++ unsigned type = DT_UNKNOWN;
++
++ qname.name = name;
++ qname.len = len;
++ qname.hash = full_name_hash(name, len);
++
++ child = d_lookup(dir, &qname);
++ if (!child) {
++ struct dentry *new;
++ new = d_alloc(dir, &qname);
++ if (new) {
++ child = instantiate(dir->d_inode, new, id, ptr);
++ if (child)
++ dput(new);
++ else
++ child = new;
++ }
++ }
++ if (!child || IS_ERR(child) || !child->d_inode)
++ goto end_instantiate;
++ inode = child->d_inode;
++ if (inode) {
++ ino = inode->i_ino;
++ type = inode->i_mode >> 12;
++ }
++ dput(child);
++end_instantiate:
++ if (!ino)
++ ino = find_inode_number(dir, &qname);
++ if (!ino)
++ ino = 1;
++ return filldir(dirent, name, len, filp->f_pos, ino, type);
++}
++
++
++
++/* get and revalidate vx_info/xid */
++
++static inline
++struct vx_info *get_proc_vx_info(struct inode *inode)
++{
++ return lookup_vx_info(PROC_I(inode)->fd);
++}
++
++static int proc_xid_revalidate(struct dentry *dentry, struct nameidata *nd)
++{
++ struct inode *inode = dentry->d_inode;
++ xid_t xid = PROC_I(inode)->fd;
++
++ if (!xid || xid_is_hashed(xid))
++ return 1;
++ d_drop(dentry);
++ return 0;
++}
++
++
++/* get and revalidate nx_info/nid */
++
++static int proc_nid_revalidate(struct dentry *dentry, struct nameidata *nd)
++{
++ struct inode *inode = dentry->d_inode;
++ nid_t nid = PROC_I(inode)->fd;
++
++ if (!nid || nid_is_hashed(nid))
++ return 1;
++ d_drop(dentry);
++ return 0;
++}
++
++
++
++#define PROC_BLOCK_SIZE (PAGE_SIZE - 1024)
++
++static ssize_t proc_vs_info_read(struct file *file, char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ struct inode *inode = file->f_dentry->d_inode;
++ unsigned long page;
++ ssize_t length = 0;
++
++ if (count > PROC_BLOCK_SIZE)
++ count = PROC_BLOCK_SIZE;
++
++ /* fade that out as soon as stable */
++ WARN_ON(PROC_I(inode)->fd);
++
++ if (!(page = __get_free_page(GFP_KERNEL)))
++ return -ENOMEM;
++
++ BUG_ON(!PROC_I(inode)->op.proc_vs_read);
++ length = PROC_I(inode)->op.proc_vs_read((char *)page);
++
++ if (length >= 0)
++ length = simple_read_from_buffer(buf, count, ppos,
++ (char *)page, length);
++
++ free_page(page);
++ return length;
++}
++
++static ssize_t proc_vx_info_read(struct file *file, char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ struct inode *inode = file->f_dentry->d_inode;
++ struct vx_info *vxi = NULL;
++ xid_t xid = PROC_I(inode)->fd;
++ unsigned long page;
++ ssize_t length = 0;
++
++ if (count > PROC_BLOCK_SIZE)
++ count = PROC_BLOCK_SIZE;
++
++ /* fade that out as soon as stable */
++ WARN_ON(!xid);
++ vxi = lookup_vx_info(xid);
++ if (!vxi)
++ goto out;
++
++ length = -ENOMEM;
++ if (!(page = __get_free_page(GFP_KERNEL)))
++ goto out_put;
++
++ BUG_ON(!PROC_I(inode)->op.proc_vxi_read);
++ length = PROC_I(inode)->op.proc_vxi_read(vxi, (char *)page);
++
++ if (length >= 0)
++ length = simple_read_from_buffer(buf, count, ppos,
++ (char *)page, length);
++
++ free_page(page);
++out_put:
++ put_vx_info(vxi);
++out:
++ return length;
++}
++
++static ssize_t proc_nx_info_read(struct file *file, char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ struct inode *inode = file->f_dentry->d_inode;
++ struct nx_info *nxi = NULL;
++ nid_t nid = PROC_I(inode)->fd;
++ unsigned long page;
++ ssize_t length = 0;
++
++ if (count > PROC_BLOCK_SIZE)
++ count = PROC_BLOCK_SIZE;
++
++ /* fade that out as soon as stable */
++ WARN_ON(!nid);
++ nxi = lookup_nx_info(nid);
++ if (!nxi)
++ goto out;
++
++ length = -ENOMEM;
++ if (!(page = __get_free_page(GFP_KERNEL)))
++ goto out_put;
++
++ BUG_ON(!PROC_I(inode)->op.proc_nxi_read);
++ length = PROC_I(inode)->op.proc_nxi_read(nxi, (char *)page);
++
++ if (length >= 0)
++ length = simple_read_from_buffer(buf, count, ppos,
++ (char *)page, length);
++
++ free_page(page);
++out_put:
++ put_nx_info(nxi);
++out:
++ return length;
++}
++
++
++
++/* here comes the lower level */
++
++
++#define NOD(NAME, MODE, IOP, FOP, OP) { \
++ .len = sizeof(NAME) - 1, \
++ .name = (NAME), \
++ .mode = MODE, \
++ .iop = IOP, \
++ .fop = FOP, \
++ .op = OP, \
++}
++
++
++#define DIR(NAME, MODE, OTYPE) \
++ NOD(NAME, (S_IFDIR | (MODE)), \
++ &proc_ ## OTYPE ## _inode_operations, \
++ &proc_ ## OTYPE ## _file_operations, { } )
++
++#define INF(NAME, MODE, OTYPE) \
++ NOD(NAME, (S_IFREG | (MODE)), NULL, \
++ &proc_vs_info_file_operations, \
++ { .proc_vs_read = &proc_##OTYPE } )
++
++#define VINF(NAME, MODE, OTYPE) \
++ NOD(NAME, (S_IFREG | (MODE)), NULL, \
++ &proc_vx_info_file_operations, \
++ { .proc_vxi_read = &proc_##OTYPE } )
++
++#define NINF(NAME, MODE, OTYPE) \
++ NOD(NAME, (S_IFREG | (MODE)), NULL, \
++ &proc_nx_info_file_operations, \
++ { .proc_nxi_read = &proc_##OTYPE } )
++
++
++static struct file_operations proc_vs_info_file_operations = {
++ .read = proc_vs_info_read,
++};
++
++static struct file_operations proc_vx_info_file_operations = {
++ .read = proc_vx_info_read,
++};
++
++static struct dentry_operations proc_xid_dentry_operations = {
++ .d_revalidate = proc_xid_revalidate,
++};
++
++static struct vs_entry vx_base_stuff[] = {
++ VINF("info", S_IRUGO, vxi_info),
++ VINF("status", S_IRUGO, vxi_status),
++ VINF("limit", S_IRUGO, vxi_limit),
++ VINF("sched", S_IRUGO, vxi_sched),
++ VINF("nsproxy", S_IRUGO, vxi_nsproxy),
++ VINF("cvirt", S_IRUGO, vxi_cvirt),
++ VINF("cacct", S_IRUGO, vxi_cacct),
++ {}
++};
++
++
++
++
++static struct dentry *proc_xid_instantiate(struct inode *dir,
++ struct dentry *dentry, int id, void *ptr)
++{
++ dentry->d_op = &proc_xid_dentry_operations;
++ return vs_proc_instantiate(dir, dentry, id, ptr);
++}
++
++static struct dentry *proc_xid_lookup(struct inode *dir,
++ struct dentry *dentry, struct nameidata *nd)
++{
++ struct vs_entry *p = vx_base_stuff;
++ struct dentry *error = ERR_PTR(-ENOENT);
++
++ for (; p->name; p++) {
++ if (p->len != dentry->d_name.len)
++ continue;
++ if (!memcmp(dentry->d_name.name, p->name, p->len))
++ break;
++ }
++ if (!p->name)
++ goto out;
++
++ error = proc_xid_instantiate(dir, dentry, PROC_I(dir)->fd, p);
++out:
++ return error;
++}
++
++static int proc_xid_readdir(struct file *filp,
++ void *dirent, filldir_t filldir)
++{
++ struct dentry *dentry = filp->f_dentry;
++ struct inode *inode = dentry->d_inode;
++ struct vs_entry *p = vx_base_stuff;
++ int size = sizeof(vx_base_stuff) / sizeof(struct vs_entry);
++ int pos, index;
++ u64 ino;
++
++ pos = filp->f_pos;
++ switch (pos) {
++ case 0:
++ ino = inode->i_ino;
++ if (filldir(dirent, ".", 1, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ case 1:
++ ino = parent_ino(dentry);
++ if (filldir(dirent, "..", 2, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ default:
++ index = pos - 2;
++ if (index >= size)
++ goto out;
++ for (p += index; p->name; p++) {
++ if (proc_fill_cache(filp, dirent, filldir, p->name, p->len,
++ vs_proc_instantiate, PROC_I(inode)->fd, p))
++ goto out;
++ pos++;
++ }
++ }
++out:
++ filp->f_pos = pos;
++ return 1;
++}
++
++
++
++static struct file_operations proc_nx_info_file_operations = {
++ .read = proc_nx_info_read,
++};
++
++static struct dentry_operations proc_nid_dentry_operations = {
++ .d_revalidate = proc_nid_revalidate,
++};
++
++static struct vs_entry nx_base_stuff[] = {
++ NINF("info", S_IRUGO, nxi_info),
++ NINF("status", S_IRUGO, nxi_status),
++ {}
++};
++
++
++static struct dentry *proc_nid_instantiate(struct inode *dir,
++ struct dentry *dentry, int id, void *ptr)
++{
++ dentry->d_op = &proc_nid_dentry_operations;
++ return vs_proc_instantiate(dir, dentry, id, ptr);
++}
++
++static struct dentry *proc_nid_lookup(struct inode *dir,
++ struct dentry *dentry, struct nameidata *nd)
++{
++ struct vs_entry *p = nx_base_stuff;
++ struct dentry *error = ERR_PTR(-ENOENT);
++
++ for (; p->name; p++) {
++ if (p->len != dentry->d_name.len)
++ continue;
++ if (!memcmp(dentry->d_name.name, p->name, p->len))
++ break;
++ }
++ if (!p->name)
++ goto out;
++
++ error = proc_nid_instantiate(dir, dentry, PROC_I(dir)->fd, p);
++out:
++ return error;
++}
++
++static int proc_nid_readdir(struct file *filp,
++ void *dirent, filldir_t filldir)
++{
++ struct dentry *dentry = filp->f_dentry;
++ struct inode *inode = dentry->d_inode;
++ struct vs_entry *p = nx_base_stuff;
++ int size = sizeof(nx_base_stuff) / sizeof(struct vs_entry);
++ int pos, index;
++ u64 ino;
++
++ pos = filp->f_pos;
++ switch (pos) {
++ case 0:
++ ino = inode->i_ino;
++ if (filldir(dirent, ".", 1, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ case 1:
++ ino = parent_ino(dentry);
++ if (filldir(dirent, "..", 2, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ default:
++ index = pos - 2;
++ if (index >= size)
++ goto out;
++ for (p += index; p->name; p++) {
++ if (proc_fill_cache(filp, dirent, filldir, p->name, p->len,
++ vs_proc_instantiate, PROC_I(inode)->fd, p))
++ goto out;
++ pos++;
++ }
++ }
++out:
++ filp->f_pos = pos;
++ return 1;
++}
++
++
++#define MAX_MULBY10 ((~0U - 9) / 10)
++
++static inline int atovid(const char *str, int len)
++{
++ int vid, c;
++
++ vid = 0;
++ while (len-- > 0) {
++ c = *str - '0';
++ str++;
++ if (c > 9)
++ return -1;
++ if (vid >= MAX_MULBY10)
++ return -1;
++ vid *= 10;
++ vid += c;
++ if (!vid)
++ return -1;
++ }
++ return vid;
++}
++
++/* now the upper level (virtual) */
++
++
++static struct file_operations proc_xid_file_operations = {
++ .read = generic_read_dir,
++ .readdir = proc_xid_readdir,
++};
++
++static struct inode_operations proc_xid_inode_operations = {
++ .lookup = proc_xid_lookup,
++};
++
++static struct vs_entry vx_virtual_stuff[] = {
++ INF("info", S_IRUGO, virtual_info),
++ INF("status", S_IRUGO, virtual_status),
++ DIR(NULL, S_IRUGO | S_IXUGO, xid),
++};
++
++
++static struct dentry *proc_virtual_lookup(struct inode *dir,
++ struct dentry *dentry, struct nameidata *nd)
++{
++ struct vs_entry *p = vx_virtual_stuff;
++ struct dentry *error = ERR_PTR(-ENOENT);
++ int id = 0;
++
++ for (; p->name; p++) {
++ if (p->len != dentry->d_name.len)
++ continue;
++ if (!memcmp(dentry->d_name.name, p->name, p->len))
++ break;
++ }
++ if (p->name)
++ goto instantiate;
++
++ id = atovid(dentry->d_name.name, dentry->d_name.len);
++ if ((id < 0) || !xid_is_hashed(id))
++ goto out;
++
++instantiate:
++ error = proc_xid_instantiate(dir, dentry, id, p);
++out:
++ return error;
++}
++
++static struct file_operations proc_nid_file_operations = {
++ .read = generic_read_dir,
++ .readdir = proc_nid_readdir,
++};
++
++static struct inode_operations proc_nid_inode_operations = {
++ .lookup = proc_nid_lookup,
++};
++
++static struct vs_entry nx_virtnet_stuff[] = {
++ INF("info", S_IRUGO, virtnet_info),
++ INF("status", S_IRUGO, virtnet_status),
++ DIR(NULL, S_IRUGO | S_IXUGO, nid),
++};
++
++
++static struct dentry *proc_virtnet_lookup(struct inode *dir,
++ struct dentry *dentry, struct nameidata *nd)
++{
++ struct vs_entry *p = nx_virtnet_stuff;
++ struct dentry *error = ERR_PTR(-ENOENT);
++ int id = 0;
++
++ for (; p->name; p++) {
++ if (p->len != dentry->d_name.len)
++ continue;
++ if (!memcmp(dentry->d_name.name, p->name, p->len))
++ break;
++ }
++ if (p->name)
++ goto instantiate;
++
++ id = atovid(dentry->d_name.name, dentry->d_name.len);
++ if ((id < 0) || !nid_is_hashed(id))
++ goto out;
++
++instantiate:
++ error = proc_nid_instantiate(dir, dentry, id, p);
++out:
++ return error;
++}
++
++
++#define PROC_MAXVIDS 32
++
++int proc_virtual_readdir(struct file *filp,
++ void *dirent, filldir_t filldir)
++{
++ struct dentry *dentry = filp->f_dentry;
++ struct inode *inode = dentry->d_inode;
++ struct vs_entry *p = vx_virtual_stuff;
++ int size = sizeof(vx_virtual_stuff) / sizeof(struct vs_entry);
++ int pos, index;
++ unsigned int xid_array[PROC_MAXVIDS];
++ char buf[PROC_NUMBUF];
++ unsigned int nr_xids, i;
++ u64 ino;
++
++ pos = filp->f_pos;
++ switch (pos) {
++ case 0:
++ ino = inode->i_ino;
++ if (filldir(dirent, ".", 1, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ case 1:
++ ino = parent_ino(dentry);
++ if (filldir(dirent, "..", 2, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ default:
++ index = pos - 2;
++ if (index >= size)
++ goto entries;
++ for (p += index; p->name; p++) {
++ if (proc_fill_cache(filp, dirent, filldir, p->name, p->len,
++ vs_proc_instantiate, 0, p))
++ goto out;
++ pos++;
++ }
++ entries:
++ index = pos - size;
++ p = &vx_virtual_stuff[size - 1];
++ nr_xids = get_xid_list(index, xid_array, PROC_MAXVIDS);
++ for (i = 0; i < nr_xids; i++) {
++ int n, xid = xid_array[i];
++ unsigned int j = PROC_NUMBUF;
++
++ n = xid;
++ do
++ buf[--j] = '0' + (n % 10);
++ while (n /= 10);
++
++ if (proc_fill_cache(filp, dirent, filldir,
++ buf + j, PROC_NUMBUF - j,
++ vs_proc_instantiate, xid, p))
++ goto out;
++ pos++;
++ }
++ }
++out:
++ filp->f_pos = pos;
++ return 0;
++}
++
++static int proc_virtual_getattr(struct vfsmount *mnt,
++ struct dentry *dentry, struct kstat *stat)
++{
++ struct inode *inode = dentry->d_inode;
++
++ generic_fillattr(inode, stat);
++ stat->nlink = 2 + atomic_read(&vx_global_cactive);
++ return 0;
++}
++
++static struct file_operations proc_virtual_dir_operations = {
++ .read = generic_read_dir,
++ .readdir = proc_virtual_readdir,
++};
++
++static struct inode_operations proc_virtual_dir_inode_operations = {
++ .getattr = proc_virtual_getattr,
++ .lookup = proc_virtual_lookup,
++};
++
++
++
++
++
++int proc_virtnet_readdir(struct file *filp,
++ void *dirent, filldir_t filldir)
++{
++ struct dentry *dentry = filp->f_dentry;
++ struct inode *inode = dentry->d_inode;
++ struct vs_entry *p = nx_virtnet_stuff;
++ int size = sizeof(nx_virtnet_stuff) / sizeof(struct vs_entry);
++ int pos, index;
++ unsigned int nid_array[PROC_MAXVIDS];
++ char buf[PROC_NUMBUF];
++ unsigned int nr_nids, i;
++ u64 ino;
++
++ pos = filp->f_pos;
++ switch (pos) {
++ case 0:
++ ino = inode->i_ino;
++ if (filldir(dirent, ".", 1, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ case 1:
++ ino = parent_ino(dentry);
++ if (filldir(dirent, "..", 2, pos, ino, DT_DIR) < 0)
++ goto out;
++ pos++;
++ /* fall through */
++ default:
++ index = pos - 2;
++ if (index >= size)
++ goto entries;
++ for (p += index; p->name; p++) {
++ if (proc_fill_cache(filp, dirent, filldir, p->name, p->len,
++ vs_proc_instantiate, 0, p))
++ goto out;
++ pos++;
++ }
++ entries:
++ index = pos - size;
++ p = &nx_virtnet_stuff[size - 1];
++ nr_nids = get_nid_list(index, nid_array, PROC_MAXVIDS);
++ for (i = 0; i < nr_nids; i++) {
++ int n, nid = nid_array[i];
++ unsigned int j = PROC_NUMBUF;
++
++ n = nid;
++ do
++ buf[--j] = '0' + (n % 10);
++ while (n /= 10);
++
++ if (proc_fill_cache(filp, dirent, filldir,
++ buf + j, PROC_NUMBUF - j,
++ vs_proc_instantiate, nid, p))
++ goto out;
++ pos++;
++ }
++ }
++out:
++ filp->f_pos = pos;
++ return 0;
++}
++
++static int proc_virtnet_getattr(struct vfsmount *mnt,
++ struct dentry *dentry, struct kstat *stat)
++{
++ struct inode *inode = dentry->d_inode;
++
++ generic_fillattr(inode, stat);
++ stat->nlink = 2 + atomic_read(&nx_global_cactive);
++ return 0;
++}
++
++static struct file_operations proc_virtnet_dir_operations = {
++ .read = generic_read_dir,
++ .readdir = proc_virtnet_readdir,
++};
++
++static struct inode_operations proc_virtnet_dir_inode_operations = {
++ .getattr = proc_virtnet_getattr,
++ .lookup = proc_virtnet_lookup,
++};
++
++
++
++void proc_vx_init(void)
++{
++ struct proc_dir_entry *ent;
++
++ ent = proc_mkdir("virtual", 0);
++ if (ent) {
++ ent->proc_fops = &proc_virtual_dir_operations;
++ ent->proc_iops = &proc_virtual_dir_inode_operations;
++ }
++ proc_virtual = ent;
++
++ ent = proc_mkdir("virtnet", 0);
++ if (ent) {
++ ent->proc_fops = &proc_virtnet_dir_operations;
++ ent->proc_iops = &proc_virtnet_dir_inode_operations;
++ }
++ proc_virtnet = ent;
++}
++
++
++
++
++/* per pid info */
++
++
++int proc_pid_vx_info(struct task_struct *p, char *buffer)
++{
++ struct vx_info *vxi;
++ char *orig = buffer;
++
++ buffer += sprintf(buffer, "XID:\t%d\n", vx_task_xid(p));
++
++ vxi = task_get_vx_info(p);
++ if (!vxi)
++ goto out;
++
++ buffer += sprintf(buffer, "BCaps:\t");
++ buffer = print_cap_t(buffer, &vxi->vx_bcaps);
++ buffer += sprintf(buffer, "\n");
++ buffer += sprintf(buffer, "CCaps:\t%016llx\n",
++ (unsigned long long)vxi->vx_ccaps);
++ buffer += sprintf(buffer, "CFlags:\t%016llx\n",
++ (unsigned long long)vxi->vx_flags);
++ buffer += sprintf(buffer, "CIPid:\t%d\n", vxi->vx_initpid);
++
++ put_vx_info(vxi);
++out:
++ return buffer - orig;
++}
++
++
++int proc_pid_nx_info(struct task_struct *p, char *buffer)
++{
++ struct nx_info *nxi;
++ struct nx_addr_v4 *v4a;
++#ifdef CONFIG_IPV6
++ struct nx_addr_v6 *v6a;
++#endif
++ char *orig = buffer;
++ int i;
++
++ buffer += sprintf(buffer, "NID:\t%d\n", nx_task_nid(p));
++
++ nxi = task_get_nx_info(p);
++ if (!nxi)
++ goto out;
++
++ buffer += sprintf(buffer, "NCaps:\t%016llx\n",
++ (unsigned long long)nxi->nx_ncaps);
++ buffer += sprintf(buffer, "NFlags:\t%016llx\n",
++ (unsigned long long)nxi->nx_flags);
++
++ buffer += sprintf(buffer,
++ "V4Root[bcast]:\t" NIPQUAD_FMT "\n",
++ NIPQUAD(nxi->v4_bcast.s_addr));
++ buffer += sprintf (buffer,
++ "V4Root[lback]:\t" NIPQUAD_FMT "\n",
++ NIPQUAD(nxi->v4_lback.s_addr));
++ if (!NX_IPV4(nxi))
++ goto skip_v4;
++ for (i = 0, v4a = &nxi->v4; v4a; i++, v4a = v4a->next)
++ buffer += sprintf(buffer, "V4Root[%d]:\t" NXAV4_FMT "\n",
++ i, NXAV4(v4a));
++skip_v4:
++#ifdef CONFIG_IPV6
++ if (!NX_IPV6(nxi))
++ goto skip_v6;
++ for (i = 0, v6a = &nxi->v6; v6a; i++, v6a = v6a->next)
++ buffer += sprintf(buffer, "V6Root[%d]:\t" NXAV6_FMT "\n",
++ i, NXAV6(v6a));
++skip_v6:
++#endif
++ put_nx_info(nxi);
++out:
++ return buffer - orig;
++}
++
+--- a/kernel/vserver/sched.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/sched.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,413 @@
++/*
++ * linux/kernel/vserver/sched.c
++ *
++ * Virtual Server: Scheduler Support
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 adapted Sam Vilains version to 2.6.3
++ * V0.02 removed legacy interface
++ * V0.03 changed vcmds to vxi arg
++ * V0.04 removed older and legacy interfaces
++ *
++ */
++
++#include <linux/vs_context.h>
++#include <linux/vs_sched.h>
++#include <linux/vserver/sched_cmd.h>
++
++#include <asm/uaccess.h>
++
++
++#define vxd_check_range(val, min, max) do { \
++ vxlprintk((val < min) || (val > max), \
++ "check_range(%ld,%ld,%ld)", \
++ (long)val, (long)min, (long)max, \
++ __FILE__, __LINE__); \
++ } while (0)
++
++
++void vx_update_sched_param(struct _vx_sched *sched,
++ struct _vx_sched_pc *sched_pc)
++{
++ unsigned int set_mask = sched->update_mask;
++
++ if (set_mask & VXSM_FILL_RATE)
++ sched_pc->fill_rate[0] = sched->fill_rate[0];
++ if (set_mask & VXSM_INTERVAL)
++ sched_pc->interval[0] = sched->interval[0];
++ if (set_mask & VXSM_FILL_RATE2)
++ sched_pc->fill_rate[1] = sched->fill_rate[1];
++ if (set_mask & VXSM_INTERVAL2)
++ sched_pc->interval[1] = sched->interval[1];
++ if (set_mask & VXSM_TOKENS)
++ sched_pc->tokens = sched->tokens;
++ if (set_mask & VXSM_TOKENS_MIN)
++ sched_pc->tokens_min = sched->tokens_min;
++ if (set_mask & VXSM_TOKENS_MAX)
++ sched_pc->tokens_max = sched->tokens_max;
++ if (set_mask & VXSM_PRIO_BIAS)
++ sched_pc->prio_bias = sched->prio_bias;
++
++ if (set_mask & VXSM_IDLE_TIME)
++ sched_pc->flags |= VXSF_IDLE_TIME;
++ else
++ sched_pc->flags &= ~VXSF_IDLE_TIME;
++
++ /* reset time */
++ sched_pc->norm_time = jiffies;
++}
++
++
++/*
++ * recalculate the context's scheduling tokens
++ *
++ * ret > 0 : number of tokens available
++ * ret < 0 : on hold, check delta_min[]
++ * -1 only jiffies
++ * -2 also idle time
++ *
++ */
++int vx_tokens_recalc(struct _vx_sched_pc *sched_pc,
++ unsigned long *norm_time, unsigned long *idle_time, int delta_min[2])
++{
++ long delta;
++ long tokens = 0;
++ int flags = sched_pc->flags;
++
++ /* how much time did pass? */
++ delta = *norm_time - sched_pc->norm_time;
++ vxd_check_range(delta, 0, INT_MAX);
++
++ if (delta >= sched_pc->interval[0]) {
++ long tokens, integral;
++
++ /* calc integral token part */
++ tokens = delta / sched_pc->interval[0];
++ integral = tokens * sched_pc->interval[0];
++ tokens *= sched_pc->fill_rate[0];
++#ifdef CONFIG_VSERVER_HARDCPU
++ delta_min[0] = delta - integral;
++ vxd_check_range(delta_min[0], 0, sched_pc->interval[0]);
++#endif
++ /* advance time */
++ sched_pc->norm_time += delta;
++
++ /* add tokens */
++ sched_pc->tokens += tokens;
++ sched_pc->token_time += tokens;
++ } else
++ delta_min[0] = delta;
++
++#ifdef CONFIG_VSERVER_IDLETIME
++ if (!(flags & VXSF_IDLE_TIME))
++ goto skip_idle;
++
++ /* how much was the idle skip? */
++ delta = *idle_time - sched_pc->idle_time;
++ vxd_check_range(delta, 0, INT_MAX);
++
++ if (delta >= sched_pc->interval[1]) {
++ long tokens, integral;
++
++ /* calc fair share token part */
++ tokens = delta / sched_pc->interval[1];
++ integral = tokens * sched_pc->interval[1];
++ tokens *= sched_pc->fill_rate[1];
++ delta_min[1] = delta - integral;
++ vxd_check_range(delta_min[1], 0, sched_pc->interval[1]);
++
++ /* advance idle time */
++ sched_pc->idle_time += integral;
++
++ /* add tokens */
++ sched_pc->tokens += tokens;
++ sched_pc->token_time += tokens;
++ } else
++ delta_min[1] = delta;
++skip_idle:
++#endif
++
++ /* clip at maximum */
++ if (sched_pc->tokens > sched_pc->tokens_max)
++ sched_pc->tokens = sched_pc->tokens_max;
++ tokens = sched_pc->tokens;
++
++ if ((flags & VXSF_ONHOLD)) {
++ /* can we unhold? */
++ if (tokens >= sched_pc->tokens_min) {
++ flags &= ~VXSF_ONHOLD;
++ sched_pc->hold_ticks +=
++ *norm_time - sched_pc->onhold;
++ } else
++ goto on_hold;
++ } else {
++ /* put on hold? */
++ if (tokens <= 0) {
++ flags |= VXSF_ONHOLD;
++ sched_pc->onhold = *norm_time;
++ goto on_hold;
++ }
++ }
++ sched_pc->flags = flags;
++ return tokens;
++
++on_hold:
++ tokens = sched_pc->tokens_min - tokens;
++ sched_pc->flags = flags;
++ BUG_ON(tokens < 0);
++
++#ifdef CONFIG_VSERVER_HARDCPU
++ /* next interval? */
++ if (!sched_pc->fill_rate[0])
++ delta_min[0] = HZ;
++ else if (tokens > sched_pc->fill_rate[0])
++ delta_min[0] += sched_pc->interval[0] *
++ tokens / sched_pc->fill_rate[0];
++ else
++ delta_min[0] = sched_pc->interval[0] - delta_min[0];
++ vxd_check_range(delta_min[0], 0, INT_MAX);
++
++#ifdef CONFIG_VSERVER_IDLETIME
++ if (!(flags & VXSF_IDLE_TIME))
++ return -1;
++
++ /* next interval? */
++ if (!sched_pc->fill_rate[1])
++ delta_min[1] = HZ;
++ else if (tokens > sched_pc->fill_rate[1])
++ delta_min[1] += sched_pc->interval[1] *
++ tokens / sched_pc->fill_rate[1];
++ else
++ delta_min[1] = sched_pc->interval[1] - delta_min[1];
++ vxd_check_range(delta_min[1], 0, INT_MAX);
++
++ return -2;
++#else
++ return -1;
++#endif /* CONFIG_VSERVER_IDLETIME */
++#else
++ return 0;
++#endif /* CONFIG_VSERVER_HARDCPU */
++}
++
++static inline unsigned long msec_to_ticks(unsigned long msec)
++{
++ return msecs_to_jiffies(msec);
++}
++
++static inline unsigned long ticks_to_msec(unsigned long ticks)
++{
++ return jiffies_to_msecs(ticks);
++}
++
++static inline unsigned long ticks_to_usec(unsigned long ticks)
++{
++ return jiffies_to_usecs(ticks);
++}
++
++
++static int do_set_sched(struct vx_info *vxi, struct vcmd_sched_v5 *data)
++{
++ unsigned int set_mask = data->mask;
++ unsigned int update_mask;
++ int i, cpu;
++
++ /* Sanity check data values */
++ if (data->tokens_max <= 0)
++ data->tokens_max = HZ;
++ if (data->tokens_min < 0)
++ data->tokens_min = HZ / 3;
++ if (data->tokens_min >= data->tokens_max)
++ data->tokens_min = data->tokens_max;
++
++ if (data->prio_bias > MAX_PRIO_BIAS)
++ data->prio_bias = MAX_PRIO_BIAS;
++ if (data->prio_bias < MIN_PRIO_BIAS)
++ data->prio_bias = MIN_PRIO_BIAS;
++
++ spin_lock(&vxi->sched.tokens_lock);
++
++ /* sync up on delayed updates */
++ for_each_cpu_mask(cpu, vxi->sched.update)
++ vx_update_sched_param(&vxi->sched,
++ &vx_per_cpu(vxi, sched_pc, cpu));
++
++ if (set_mask & VXSM_FILL_RATE)
++ vxi->sched.fill_rate[0] = data->fill_rate[0];
++ if (set_mask & VXSM_FILL_RATE2)
++ vxi->sched.fill_rate[1] = data->fill_rate[1];
++ if (set_mask & VXSM_INTERVAL)
++ vxi->sched.interval[0] = (set_mask & VXSM_MSEC) ?
++ msec_to_ticks(data->interval[0]) : data->interval[0];
++ if (set_mask & VXSM_INTERVAL2)
++ vxi->sched.interval[1] = (set_mask & VXSM_MSEC) ?
++ msec_to_ticks(data->interval[1]) : data->interval[1];
++ if (set_mask & VXSM_TOKENS)
++ vxi->sched.tokens = data->tokens;
++ if (set_mask & VXSM_TOKENS_MIN)
++ vxi->sched.tokens_min = data->tokens_min;
++ if (set_mask & VXSM_TOKENS_MAX)
++ vxi->sched.tokens_max = data->tokens_max;
++ if (set_mask & VXSM_PRIO_BIAS)
++ vxi->sched.prio_bias = data->prio_bias;
++
++ /* Sanity check rate/interval */
++ for (i = 0; i < 2; i++) {
++ if (data->fill_rate[i] < 0)
++ data->fill_rate[i] = 0;
++ if (data->interval[i] <= 0)
++ data->interval[i] = HZ;
++ }
++
++ update_mask = vxi->sched.update_mask & VXSM_SET_MASK;
++ update_mask |= (set_mask & (VXSM_SET_MASK | VXSM_IDLE_TIME));
++ vxi->sched.update_mask = update_mask;
++
++#ifdef CONFIG_SMP
++ rmb();
++ if (set_mask & VXSM_CPU_ID) {
++ vxi->sched.update = cpumask_of_cpu(data->cpu_id);
++ cpus_and(vxi->sched.update, cpu_online_map,
++ vxi->sched.update);
++ } else
++ vxi->sched.update = cpu_online_map;
++
++ /* forced reload? */
++ if (set_mask & VXSM_FORCE) {
++ for_each_cpu_mask(cpu, vxi->sched.update)
++ vx_update_sched_param(&vxi->sched,
++ &vx_per_cpu(vxi, sched_pc, cpu));
++ vxi->sched.update = CPU_MASK_NONE;
++ }
++#else
++ /* on UP we update immediately */
++ vx_update_sched_param(&vxi->sched,
++ &vx_per_cpu(vxi, sched_pc, 0));
++#endif
++
++ spin_unlock(&vxi->sched.tokens_lock);
++ return 0;
++}
++
++
++#define COPY_IDS(C) C(cpu_id); C(bucket_id)
++#define COPY_PRI(C) C(prio_bias)
++#define COPY_TOK(C) C(tokens); C(tokens_min); C(tokens_max)
++#define COPY_FRI(C) C(fill_rate[0]); C(interval[0]); \
++ C(fill_rate[1]); C(interval[1]);
++
++#define COPY_VALUE(name) vc_data.name = data->name
++
++static int do_set_sched_v4(struct vx_info *vxi, struct vcmd_set_sched_v4 *data)
++{
++ struct vcmd_sched_v5 vc_data;
++
++ vc_data.mask = data->set_mask;
++ COPY_IDS(COPY_VALUE);
++ COPY_PRI(COPY_VALUE);
++ COPY_TOK(COPY_VALUE);
++ vc_data.fill_rate[0] = vc_data.fill_rate[1] = data->fill_rate;
++ vc_data.interval[0] = vc_data.interval[1] = data->interval;
++ return do_set_sched(vxi, &vc_data);
++}
++
++int vc_set_sched_v4(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_set_sched_v4 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_sched_v4(vxi, &vc_data);
++}
++
++ /* latest interface is v5 */
++
++int vc_set_sched(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_sched_v5 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return do_set_sched(vxi, &vc_data);
++}
++
++
++#define COPY_PRI(C) C(prio_bias)
++#define COPY_TOK(C) C(tokens); C(tokens_min); C(tokens_max)
++#define COPY_FRI(C) C(fill_rate[0]); C(interval[0]); \
++ C(fill_rate[1]); C(interval[1]);
++
++#define COPY_VALUE(name) vc_data.name = data->name
++
++
++int vc_get_sched(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_sched_v5 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ if (vc_data.mask & VXSM_CPU_ID) {
++ int cpu = vc_data.cpu_id;
++ struct _vx_sched_pc *data;
++
++ if (!cpu_possible(cpu))
++ return -EINVAL;
++
++ data = &vx_per_cpu(vxi, sched_pc, cpu);
++ COPY_TOK(COPY_VALUE);
++ COPY_PRI(COPY_VALUE);
++ COPY_FRI(COPY_VALUE);
++
++ if (data->flags & VXSF_IDLE_TIME)
++ vc_data.mask |= VXSM_IDLE_TIME;
++ } else {
++ struct _vx_sched *data = &vxi->sched;
++
++ COPY_TOK(COPY_VALUE);
++ COPY_PRI(COPY_VALUE);
++ COPY_FRI(COPY_VALUE);
++ }
++
++ if (vc_data.mask & VXSM_MSEC) {
++ vc_data.interval[0] = ticks_to_msec(vc_data.interval[0]);
++ vc_data.interval[1] = ticks_to_msec(vc_data.interval[1]);
++ }
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
++
++int vc_sched_info(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_sched_info vc_data;
++ int cpu;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ cpu = vc_data.cpu_id;
++ if (!cpu_possible(cpu))
++ return -EINVAL;
++
++ if (vxi) {
++ struct _vx_sched_pc *sched_pc =
++ &vx_per_cpu(vxi, sched_pc, cpu);
++
++ vc_data.user_msec = ticks_to_msec(sched_pc->user_ticks);
++ vc_data.sys_msec = ticks_to_msec(sched_pc->sys_ticks);
++ vc_data.hold_msec = ticks_to_msec(sched_pc->hold_ticks);
++ vc_data.vavavoom = sched_pc->vavavoom;
++ }
++ vc_data.token_usec = ticks_to_usec(1);
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ return -EFAULT;
++ return 0;
++}
++
+--- a/kernel/vserver/sched_init.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/sched_init.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,50 @@
++
++static inline void vx_info_init_sched(struct _vx_sched *sched)
++{
++ static struct lock_class_key tokens_lock_key;
++
++ /* scheduling; hard code starting values as constants */
++ sched->fill_rate[0] = 1;
++ sched->interval[0] = 4;
++ sched->fill_rate[1] = 1;
++ sched->interval[1] = 8;
++ sched->tokens = HZ >> 2;
++ sched->tokens_min = HZ >> 4;
++ sched->tokens_max = HZ >> 1;
++ sched->tokens_lock = SPIN_LOCK_UNLOCKED;
++ sched->prio_bias = 0;
++
++ lockdep_set_class(&sched->tokens_lock, &tokens_lock_key);
++}
++
++static inline
++void vx_info_init_sched_pc(struct _vx_sched_pc *sched_pc, int cpu)
++{
++ sched_pc->fill_rate[0] = 1;
++ sched_pc->interval[0] = 4;
++ sched_pc->fill_rate[1] = 1;
++ sched_pc->interval[1] = 8;
++ sched_pc->tokens = HZ >> 2;
++ sched_pc->tokens_min = HZ >> 4;
++ sched_pc->tokens_max = HZ >> 1;
++ sched_pc->prio_bias = 0;
++ sched_pc->vavavoom = 0;
++ sched_pc->token_time = 0;
++ sched_pc->idle_time = 0;
++ sched_pc->norm_time = jiffies;
++
++ sched_pc->user_ticks = 0;
++ sched_pc->sys_ticks = 0;
++ sched_pc->hold_ticks = 0;
++}
++
++static inline void vx_info_exit_sched(struct _vx_sched *sched)
++{
++ return;
++}
++
++static inline
++void vx_info_exit_sched_pc(struct _vx_sched_pc *sched_pc, int cpu)
++{
++ return;
++}
+--- a/kernel/vserver/sched_proc.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/sched_proc.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,57 @@
++#ifndef _VX_SCHED_PROC_H
++#define _VX_SCHED_PROC_H
++
++
++static inline
++int vx_info_proc_sched(struct _vx_sched *sched, char *buffer)
++{
++ int length = 0;
++
++ length += sprintf(buffer,
++ "FillRate:\t%8d,%d\n"
++ "Interval:\t%8d,%d\n"
++ "TokensMin:\t%8d\n"
++ "TokensMax:\t%8d\n"
++ "PrioBias:\t%8d\n",
++ sched->fill_rate[0],
++ sched->fill_rate[1],
++ sched->interval[0],
++ sched->interval[1],
++ sched->tokens_min,
++ sched->tokens_max,
++ sched->prio_bias);
++ return length;
++}
++
++static inline
++int vx_info_proc_sched_pc(struct _vx_sched_pc *sched_pc,
++ char *buffer, int cpu)
++{
++ int length = 0;
++
++ length += sprintf(buffer + length,
++ "cpu %d: %lld %lld %lld %ld %ld", cpu,
++ (unsigned long long)sched_pc->user_ticks,
++ (unsigned long long)sched_pc->sys_ticks,
++ (unsigned long long)sched_pc->hold_ticks,
++ sched_pc->token_time,
++ sched_pc->idle_time);
++ length += sprintf(buffer + length,
++ " %c%c %d %d %d %d/%d %d/%d",
++ (sched_pc->flags & VXSF_ONHOLD) ? 'H' : 'R',
++ (sched_pc->flags & VXSF_IDLE_TIME) ? 'I' : '-',
++ sched_pc->tokens,
++ sched_pc->tokens_min,
++ sched_pc->tokens_max,
++ sched_pc->fill_rate[0],
++ sched_pc->interval[0],
++ sched_pc->fill_rate[1],
++ sched_pc->interval[1]);
++ length += sprintf(buffer + length,
++ " %d %d\n",
++ sched_pc->prio_bias,
++ sched_pc->vavavoom);
++ return length;
++}
++
++#endif /* _VX_SCHED_PROC_H */
+--- a/kernel/vserver/signal.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/signal.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,132 @@
++/*
++ * linux/kernel/vserver/signal.c
++ *
++ * Virtual Server: Signal Support
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 broken out from vcontext V0.05
++ * V0.02 changed vcmds to vxi arg
++ * V0.03 adjusted siginfo for kill
++ *
++ */
++
++#include <asm/uaccess.h>
++
++#include <linux/vs_context.h>
++#include <linux/vs_pid.h>
++#include <linux/vserver/signal_cmd.h>
++
++
++int vx_info_kill(struct vx_info *vxi, int pid, int sig)
++{
++ int retval, count = 0;
++ struct task_struct *p;
++ struct siginfo *sip = SEND_SIG_PRIV;
++
++ retval = -ESRCH;
++ vxdprintk(VXD_CBIT(misc, 4),
++ "vx_info_kill(%p[#%d],%d,%d)*",
++ vxi, vxi->vx_id, pid, sig);
++ read_lock(&tasklist_lock);
++ switch (pid) {
++ case 0:
++ case -1:
++ for_each_process(p) {
++ int err = 0;
++
++ if (vx_task_xid(p) != vxi->vx_id || p->pid <= 1 ||
++ (pid && vxi->vx_initpid == p->pid))
++ continue;
++
++ err = group_send_sig_info(sig, sip, p);
++ ++count;
++ if (err != -EPERM)
++ retval = err;
++ }
++ break;
++
++ case 1:
++ if (vxi->vx_initpid) {
++ pid = vxi->vx_initpid;
++ /* for now, only SIGINT to private init ... */
++ if (!vx_info_flags(vxi, VXF_STATE_ADMIN, 0) &&
++ /* ... as long as there are tasks left */
++ (atomic_read(&vxi->vx_tasks) > 1))
++ sig = SIGINT;
++ }
++ /* fallthrough */
++ default:
++ p = find_task_by_real_pid(pid);
++ if (p) {
++ if (vx_task_xid(p) == vxi->vx_id)
++ retval = group_send_sig_info(sig, sip, p);
++ }
++ break;
++ }
++ read_unlock(&tasklist_lock);
++ vxdprintk(VXD_CBIT(misc, 4),
++ "vx_info_kill(%p[#%d],%d,%d,%ld) = %d",
++ vxi, vxi->vx_id, pid, sig, (long)sip, retval);
++ return retval;
++}
++
++int vc_ctx_kill(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_ctx_kill_v0 vc_data;
++
++ if (copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ /* special check to allow guest shutdown */
++ if (!vx_info_flags(vxi, VXF_STATE_ADMIN, 0) &&
++ /* forbid killall pid=0 when init is present */
++ (((vc_data.pid < 1) && vxi->vx_initpid) ||
++ (vc_data.pid > 1)))
++ return -EACCES;
++
++ return vx_info_kill(vxi, vc_data.pid, vc_data.sig);
++}
++
++
++static int __wait_exit(struct vx_info *vxi)
++{
++ DECLARE_WAITQUEUE(wait, current);
++ int ret = 0;
++
++ add_wait_queue(&vxi->vx_wait, &wait);
++ set_current_state(TASK_INTERRUPTIBLE);
++
++wait:
++ if (vx_info_state(vxi,
++ VXS_SHUTDOWN | VXS_HASHED | VXS_HELPER) == VXS_SHUTDOWN)
++ goto out;
++ if (signal_pending(current)) {
++ ret = -ERESTARTSYS;
++ goto out;
++ }
++ schedule();
++ goto wait;
++
++out:
++ set_current_state(TASK_RUNNING);
++ remove_wait_queue(&vxi->vx_wait, &wait);
++ return ret;
++}
++
++
++
++int vc_wait_exit(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_wait_exit_v0 vc_data;
++ int ret;
++
++ ret = __wait_exit(vxi);
++ vc_data.reboot_cmd = vxi->reboot_cmd;
++ vc_data.exit_code = vxi->exit_code;
++
++ if (copy_to_user(data, &vc_data, sizeof(vc_data)))
++ ret = -EFAULT;
++ return ret;
++}
++
+--- a/kernel/vserver/space.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/space.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,372 @@
++/*
++ * linux/kernel/vserver/space.c
++ *
++ * Virtual Server: Context Space Support
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 broken out from context.c 0.07
++ * V0.02 added task locking for namespace
++ * V0.03 broken out vx_enter_namespace
++ * V0.04 added *space support and commands
++ *
++ */
++
++#include <linux/utsname.h>
++#include <linux/nsproxy.h>
++#include <linux/err.h>
++#include <asm/uaccess.h>
++
++#include <linux/vs_context.h>
++#include <linux/vserver/space.h>
++#include <linux/vserver/space_cmd.h>
++
++atomic_t vs_global_nsproxy = ATOMIC_INIT(0);
++atomic_t vs_global_fs = ATOMIC_INIT(0);
++atomic_t vs_global_mnt_ns = ATOMIC_INIT(0);
++atomic_t vs_global_uts_ns = ATOMIC_INIT(0);
++atomic_t vs_global_ipc_ns = ATOMIC_INIT(0);
++atomic_t vs_global_user_ns = ATOMIC_INIT(0);
++atomic_t vs_global_pid_ns = ATOMIC_INIT(0);
++
++
++/* namespace functions */
++
++#include <linux/mnt_namespace.h>
++#include <linux/user_namespace.h>
++#include <linux/pid_namespace.h>
++#include <linux/ipc_namespace.h>
++#include <net/net_namespace.h>
++
++
++static const struct vcmd_space_mask space_mask_v0 = {
++ .mask = CLONE_FS |
++ CLONE_NEWNS |
++ CLONE_NEWUTS |
++ CLONE_NEWIPC |
++ CLONE_NEWUSER |
++ 0
++};
++
++static const struct vcmd_space_mask space_mask = {
++ .mask = CLONE_FS |
++ CLONE_NEWNS |
++ CLONE_NEWUTS |
++ CLONE_NEWIPC |
++ CLONE_NEWUSER |
++#ifdef CONFIG_PID_NS
++ CLONE_NEWPID |
++#endif
++#ifdef CONFIG_NET_NS
++ CLONE_NEWNET |
++#endif
++ 0
++};
++
++static const struct vcmd_space_mask default_space_mask = {
++ .mask = CLONE_FS |
++ CLONE_NEWNS |
++ CLONE_NEWUTS |
++ CLONE_NEWIPC |
++ CLONE_NEWUSER |
++#ifdef CONFIG_PID_NS
++// CLONE_NEWPID |
++#endif
++ 0
++};
++
++/*
++ * build a new nsproxy mix
++ * assumes that both proxies are 'const'
++ * does not touch nsproxy refcounts
++ * will hold a reference on the result.
++ */
++
++struct nsproxy *vs_mix_nsproxy(struct nsproxy *old_nsproxy,
++ struct nsproxy *new_nsproxy, unsigned long mask)
++{
++ struct mnt_namespace *old_ns;
++ struct uts_namespace *old_uts;
++ struct ipc_namespace *old_ipc;
++ struct user_namespace *old_user;
++#ifdef CONFIG_PID_NS
++ struct pid_namespace *old_pid;
++#endif
++#ifdef CONFIG_NET_NS
++ struct net *old_net;
++#endif
++ struct nsproxy *nsproxy;
++
++ nsproxy = copy_nsproxy(old_nsproxy);
++ if (!nsproxy)
++ goto out;
++
++ if (mask & CLONE_NEWNS) {
++ old_ns = nsproxy->mnt_ns;
++ nsproxy->mnt_ns = new_nsproxy->mnt_ns;
++ if (nsproxy->mnt_ns)
++ get_mnt_ns(nsproxy->mnt_ns);
++ } else
++ old_ns = NULL;
++
++ if (mask & CLONE_NEWUTS) {
++ old_uts = nsproxy->uts_ns;
++ nsproxy->uts_ns = new_nsproxy->uts_ns;
++ if (nsproxy->uts_ns)
++ get_uts_ns(nsproxy->uts_ns);
++ } else
++ old_uts = NULL;
++
++ if (mask & CLONE_NEWIPC) {
++ old_ipc = nsproxy->ipc_ns;
++ nsproxy->ipc_ns = new_nsproxy->ipc_ns;
++ if (nsproxy->ipc_ns)
++ get_ipc_ns(nsproxy->ipc_ns);
++ } else
++ old_ipc = NULL;
++
++ if (mask & CLONE_NEWUSER) {
++ old_user = nsproxy->user_ns;
++ nsproxy->user_ns = new_nsproxy->user_ns;
++ if (nsproxy->user_ns)
++ get_user_ns(nsproxy->user_ns);
++ } else
++ old_user = NULL;
++
++#ifdef CONFIG_PID_NS
++ if (mask & CLONE_NEWPID) {
++ old_pid = nsproxy->pid_ns;
++ nsproxy->pid_ns = new_nsproxy->pid_ns;
++ if (nsproxy->pid_ns)
++ get_pid_ns(nsproxy->pid_ns);
++ } else
++ old_pid = NULL;
++#endif
++#ifdef CONFIG_NET_NS
++ if (mask & CLONE_NEWNET) {
++ old_net = nsproxy->net_ns;
++ nsproxy->net_ns = new_nsproxy->net_ns;
++ if (nsproxy->net_ns)
++ get_net(nsproxy->net_ns);
++ } else
++ old_net = NULL;
++#endif
++ if (old_ns)
++ put_mnt_ns(old_ns);
++ if (old_uts)
++ put_uts_ns(old_uts);
++ if (old_ipc)
++ put_ipc_ns(old_ipc);
++ if (old_user)
++ put_user_ns(old_user);
++#ifdef CONFIG_PID_NS
++ if (old_pid)
++ put_pid_ns(old_pid);
++#endif
++#ifdef CONFIG_NET_NS
++ if (old_net)
++ put_net(old_net);
++#endif
++out:
++ return nsproxy;
++}
++
++
++/*
++ * merge two nsproxy structs into a new one.
++ * will hold a reference on the result.
++ */
++
++static inline
++struct nsproxy *__vs_merge_nsproxy(struct nsproxy *old,
++ struct nsproxy *proxy, unsigned long mask)
++{
++ struct nsproxy null_proxy = { .mnt_ns = NULL };
++
++ if (!proxy)
++ return NULL;
++
++ if (mask) {
++ /* vs_mix_nsproxy returns with reference */
++ return vs_mix_nsproxy(old ? old : &null_proxy,
++ proxy, mask);
++ }
++ get_nsproxy(proxy);
++ return proxy;
++}
++
++/*
++ * merge two fs structs into a new one.
++ * will take a reference on the result.
++ */
++
++static inline
++struct fs_struct *__vs_merge_fs(struct fs_struct *old,
++ struct fs_struct *fs, unsigned long mask)
++{
++ if (!(mask & CLONE_FS)) {
++ if (old)
++ atomic_inc(&old->count);
++ return old;
++ }
++
++ if (!fs)
++ return NULL;
++
++ return copy_fs_struct(fs);
++}
++
++
++int vx_enter_space(struct vx_info *vxi, unsigned long mask)
++{
++ struct nsproxy *proxy, *proxy_cur, *proxy_new;
++ struct fs_struct *fs, *fs_cur, *fs_new;
++ int ret;
++
++ vxdprintk(VXD_CBIT(space, 8), "vx_enter_space(%p[#%u],0x%08lx)",
++ vxi, vxi->vx_id, mask);
++
++ if (vx_info_flags(vxi, VXF_INFO_PRIVATE, 0))
++ return -EACCES;
++
++ if (!mask)
++ mask = vxi->vx_nsmask;
++
++ if ((mask & vxi->vx_nsmask) != mask)
++ return -EINVAL;
++
++ proxy = vxi->vx_nsproxy;
++ fs = vxi->vx_fs;
++
++ task_lock(current);
++ fs_cur = current->fs;
++ atomic_inc(&fs_cur->count);
++ proxy_cur = current->nsproxy;
++ get_nsproxy(proxy_cur);
++ task_unlock(current);
++
++ fs_new = __vs_merge_fs(fs_cur, fs, mask);
++ if (IS_ERR(fs_new)) {
++ ret = PTR_ERR(fs_new);
++ goto out_put;
++ }
++
++ proxy_new = __vs_merge_nsproxy(proxy_cur, proxy, mask);
++ if (IS_ERR(proxy_new)) {
++ ret = PTR_ERR(proxy_new);
++ goto out_put_fs;
++ }
++
++ fs_new = xchg(¤t->fs, fs_new);
++ proxy_new = xchg(¤t->nsproxy, proxy_new);
++ ret = 0;
++
++ if (proxy_new)
++ put_nsproxy(proxy_new);
++out_put_fs:
++ if (fs_new)
++ put_fs_struct(fs_new);
++out_put:
++ if (proxy_cur)
++ put_nsproxy(proxy_cur);
++ if (fs_cur)
++ put_fs_struct(fs_cur);
++ return ret;
++}
++
++
++int vx_set_space(struct vx_info *vxi, unsigned long mask)
++{
++ struct nsproxy *proxy_vxi, *proxy_cur, *proxy_new;
++ struct fs_struct *fs_vxi, *fs_cur, *fs_new;
++ int ret;
++
++ vxdprintk(VXD_CBIT(space, 8), "vx_set_space(%p[#%u],0x%08lx)",
++ vxi, vxi->vx_id, mask);
++#if 0
++ if (!mask)
++ mask = default_space_mask.mask;
++#endif
++ if ((mask & space_mask.mask) != mask)
++ return -EINVAL;
++
++ proxy_vxi = vxi->vx_nsproxy;
++ fs_vxi = vxi->vx_fs;
++
++ task_lock(current);
++ fs_cur = current->fs;
++ atomic_inc(&fs_cur->count);
++ proxy_cur = current->nsproxy;
++ get_nsproxy(proxy_cur);
++ task_unlock(current);
++
++ fs_new = __vs_merge_fs(fs_vxi, fs_cur, mask);
++ if (IS_ERR(fs_new)) {
++ ret = PTR_ERR(fs_new);
++ goto out_put;
++ }
++
++ proxy_new = __vs_merge_nsproxy(proxy_vxi, proxy_cur, mask);
++ if (IS_ERR(proxy_new)) {
++ ret = PTR_ERR(proxy_new);
++ goto out_put_fs;
++ }
++
++ fs_new = xchg(&vxi->vx_fs, fs_new);
++ proxy_new = xchg(&vxi->vx_nsproxy, proxy_new);
++ vxi->vx_nsmask |= mask;
++ ret = 0;
++
++ if (proxy_new)
++ put_nsproxy(proxy_new);
++out_put_fs:
++ if (fs_new)
++ put_fs_struct(fs_new);
++out_put:
++ if (proxy_cur)
++ put_nsproxy(proxy_cur);
++ if (fs_cur)
++ put_fs_struct(fs_cur);
++ return ret;
++}
++
++
++int vc_enter_space(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_space_mask vc_data = { .mask = 0 };
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return vx_enter_space(vxi, vc_data.mask);
++}
++
++int vc_set_space(struct vx_info *vxi, void __user *data)
++{
++ struct vcmd_space_mask vc_data = { .mask = 0 };
++
++ if (data && copy_from_user(&vc_data, data, sizeof(vc_data)))
++ return -EFAULT;
++
++ return vx_set_space(vxi, vc_data.mask);
++}
++
++int vc_get_space_mask(void __user *data, int type)
++{
++ const struct vcmd_space_mask *mask;
++
++ if (type == 0)
++ mask = &space_mask_v0;
++ else if (type == 1)
++ mask = &space_mask;
++ else
++ mask = &default_space_mask;
++
++ vxdprintk(VXD_CBIT(space, 10),
++ "vc_get_space_mask(%d) = %08llx", type, mask->mask);
++
++ if (copy_to_user(data, mask, sizeof(struct vcmd_space_mask)))
++ return -EFAULT;
++ return 0;
++}
++
+--- a/kernel/vserver/switch.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/switch.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,537 @@
++/*
++ * linux/kernel/vserver/switch.c
++ *
++ * Virtual Server: Syscall Switch
++ *
++ * Copyright (C) 2003-2007 Herbert Pötzl
++ *
++ * V0.01 syscall switch
++ * V0.02 added signal to context
++ * V0.03 added rlimit functions
++ * V0.04 added iattr, task/xid functions
++ * V0.05 added debug/history stuff
++ * V0.06 added compat32 layer
++ * V0.07 vcmd args and perms
++ * V0.08 added status commands
++ * V0.09 added tag commands
++ * V0.10 added oom bias
++ * V0.11 added device commands
++ *
++ */
++
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vserver/switch.h>
++
++#include "vci_config.h"
++
++
++static inline
++int vc_get_version(uint32_t id)
++{
++ return VCI_VERSION;
++}
++
++static inline
++int vc_get_vci(uint32_t id)
++{
++ return vci_kernel_config();
++}
++
++#include <linux/vserver/context_cmd.h>
++#include <linux/vserver/cvirt_cmd.h>
++#include <linux/vserver/cacct_cmd.h>
++#include <linux/vserver/limit_cmd.h>
++#include <linux/vserver/network_cmd.h>
++#include <linux/vserver/sched_cmd.h>
++#include <linux/vserver/debug_cmd.h>
++#include <linux/vserver/inode_cmd.h>
++#include <linux/vserver/dlimit_cmd.h>
++#include <linux/vserver/signal_cmd.h>
++#include <linux/vserver/space_cmd.h>
++#include <linux/vserver/tag_cmd.h>
++#include <linux/vserver/device_cmd.h>
++
++#include <linux/vserver/inode.h>
++#include <linux/vserver/dlimit.h>
++
++
++#ifdef CONFIG_COMPAT
++#define __COMPAT(name, id, data, compat) \
++ (compat) ? name ## _x32(id, data) : name(id, data)
++#define __COMPAT_NO_ID(name, data, compat) \
++ (compat) ? name ## _x32(data) : name(data)
++#else
++#define __COMPAT(name, id, data, compat) \
++ name(id, data)
++#define __COMPAT_NO_ID(name, data, compat) \
++ name(data)
++#endif
++
++
++static inline
++long do_vcmd(uint32_t cmd, uint32_t id,
++ struct vx_info *vxi, struct nx_info *nxi,
++ void __user *data, int compat)
++{
++ switch (cmd) {
++
++ case VCMD_get_version:
++ return vc_get_version(id);
++ case VCMD_get_vci:
++ return vc_get_vci(id);
++
++ case VCMD_task_xid:
++ return vc_task_xid(id);
++ case VCMD_vx_info:
++ return vc_vx_info(vxi, data);
++
++ case VCMD_task_nid:
++ return vc_task_nid(id);
++ case VCMD_nx_info:
++ return vc_nx_info(nxi, data);
++
++ case VCMD_task_tag:
++ return vc_task_tag(id);
++
++ /* this is version 1 */
++ case VCMD_set_space:
++ return vc_set_space(vxi, data);
++
++ case VCMD_get_space_mask_v0:
++ return vc_get_space_mask(data, 0);
++ /* this is version 1 */
++ case VCMD_get_space_mask:
++ return vc_get_space_mask(data, 1);
++
++ case VCMD_get_space_default:
++ return vc_get_space_mask(data, -1);
++
++#ifdef CONFIG_IA32_EMULATION
++ case VCMD_get_rlimit:
++ return __COMPAT(vc_get_rlimit, vxi, data, compat);
++ case VCMD_set_rlimit:
++ return __COMPAT(vc_set_rlimit, vxi, data, compat);
++#else
++ case VCMD_get_rlimit:
++ return vc_get_rlimit(vxi, data);
++ case VCMD_set_rlimit:
++ return vc_set_rlimit(vxi, data);
++#endif
++ case VCMD_get_rlimit_mask:
++ return vc_get_rlimit_mask(id, data);
++ case VCMD_reset_minmax:
++ return vc_reset_minmax(vxi, data);
++
++ case VCMD_get_vhi_name:
++ return vc_get_vhi_name(vxi, data);
++ case VCMD_set_vhi_name:
++ return vc_set_vhi_name(vxi, data);
++
++ case VCMD_ctx_stat:
++ return vc_ctx_stat(vxi, data);
++ case VCMD_virt_stat:
++ return vc_virt_stat(vxi, data);
++ case VCMD_sock_stat:
++ return vc_sock_stat(vxi, data);
++ case VCMD_rlimit_stat:
++ return vc_rlimit_stat(vxi, data);
++
++ case VCMD_set_cflags:
++ return vc_set_cflags(vxi, data);
++ case VCMD_get_cflags:
++ return vc_get_cflags(vxi, data);
++
++ /* this is version 1 */
++ case VCMD_set_ccaps:
++ return vc_set_ccaps(vxi, data);
++ /* this is version 1 */
++ case VCMD_get_ccaps:
++ return vc_get_ccaps(vxi, data);
++ case VCMD_set_bcaps:
++ return vc_set_bcaps(vxi, data);
++ case VCMD_get_bcaps:
++ return vc_get_bcaps(vxi, data);
++
++ case VCMD_set_badness:
++ return vc_set_badness(vxi, data);
++ case VCMD_get_badness:
++ return vc_get_badness(vxi, data);
++
++ case VCMD_set_nflags:
++ return vc_set_nflags(nxi, data);
++ case VCMD_get_nflags:
++ return vc_get_nflags(nxi, data);
++
++ case VCMD_set_ncaps:
++ return vc_set_ncaps(nxi, data);
++ case VCMD_get_ncaps:
++ return vc_get_ncaps(nxi, data);
++
++ case VCMD_set_sched_v4:
++ return vc_set_sched_v4(vxi, data);
++ /* this is version 5 */
++ case VCMD_set_sched:
++ return vc_set_sched(vxi, data);
++ case VCMD_get_sched:
++ return vc_get_sched(vxi, data);
++ case VCMD_sched_info:
++ return vc_sched_info(vxi, data);
++
++ case VCMD_add_dlimit:
++ return __COMPAT(vc_add_dlimit, id, data, compat);
++ case VCMD_rem_dlimit:
++ return __COMPAT(vc_rem_dlimit, id, data, compat);
++ case VCMD_set_dlimit:
++ return __COMPAT(vc_set_dlimit, id, data, compat);
++ case VCMD_get_dlimit:
++ return __COMPAT(vc_get_dlimit, id, data, compat);
++
++ case VCMD_ctx_kill:
++ return vc_ctx_kill(vxi, data);
++
++ case VCMD_wait_exit:
++ return vc_wait_exit(vxi, data);
++
++ case VCMD_get_iattr:
++ return __COMPAT_NO_ID(vc_get_iattr, data, compat);
++ case VCMD_set_iattr:
++ return __COMPAT_NO_ID(vc_set_iattr, data, compat);
++
++ case VCMD_fget_iattr:
++ return vc_fget_iattr(id, data);
++ case VCMD_fset_iattr:
++ return vc_fset_iattr(id, data);
++
++ case VCMD_enter_space_v0:
++ return vc_enter_space(vxi, NULL);
++ /* this is version 1 */
++ case VCMD_enter_space:
++ return vc_enter_space(vxi, data);
++
++ case VCMD_ctx_create_v0:
++ return vc_ctx_create(id, NULL);
++ case VCMD_ctx_create:
++ return vc_ctx_create(id, data);
++ case VCMD_ctx_migrate_v0:
++ return vc_ctx_migrate(vxi, NULL);
++ case VCMD_ctx_migrate:
++ return vc_ctx_migrate(vxi, data);
++
++ case VCMD_net_create_v0:
++ return vc_net_create(id, NULL);
++ case VCMD_net_create:
++ return vc_net_create(id, data);
++ case VCMD_net_migrate:
++ return vc_net_migrate(nxi, data);
++
++ case VCMD_tag_migrate:
++ return vc_tag_migrate(id);
++
++ case VCMD_net_add:
++ return vc_net_add(nxi, data);
++ case VCMD_net_remove:
++ return vc_net_remove(nxi, data);
++
++ case VCMD_net_add_ipv4:
++ return vc_net_add_ipv4(nxi, data);
++ case VCMD_net_remove_ipv4:
++ return vc_net_remove_ipv4(nxi, data);
++#ifdef CONFIG_IPV6
++ case VCMD_net_add_ipv6:
++ return vc_net_add_ipv6(nxi, data);
++ case VCMD_net_remove_ipv6:
++ return vc_net_remove_ipv6(nxi, data);
++#endif
++/* case VCMD_add_match_ipv4:
++ return vc_add_match_ipv4(nxi, data);
++ case VCMD_get_match_ipv4:
++ return vc_get_match_ipv4(nxi, data);
++#ifdef CONFIG_IPV6
++ case VCMD_add_match_ipv6:
++ return vc_add_match_ipv6(nxi, data);
++ case VCMD_get_match_ipv6:
++ return vc_get_match_ipv6(nxi, data);
++#endif */
++
++#ifdef CONFIG_VSERVER_DEVICE
++ case VCMD_set_mapping:
++ return __COMPAT(vc_set_mapping, vxi, data, compat);
++ case VCMD_unset_mapping:
++ return __COMPAT(vc_unset_mapping, vxi, data, compat);
++#endif
++#ifdef CONFIG_VSERVER_HISTORY
++ case VCMD_dump_history:
++ return vc_dump_history(id);
++ case VCMD_read_history:
++ return __COMPAT(vc_read_history, id, data, compat);
++#endif
++#ifdef CONFIG_VSERVER_MONITOR
++ case VCMD_read_monitor:
++ return __COMPAT(vc_read_monitor, id, data, compat);
++#endif
++ default:
++ vxwprintk_task(1, "unimplemented VCMD_%02d_%d[%d]",
++ VC_CATEGORY(cmd), VC_COMMAND(cmd), VC_VERSION(cmd));
++ }
++ return -ENOSYS;
++}
++
++
++#define __VCMD(vcmd, _perm, _args, _flags) \
++ case VCMD_ ## vcmd: perm = _perm; \
++ args = _args; flags = _flags; break
++
++
++#define VCA_NONE 0x00
++#define VCA_VXI 0x01
++#define VCA_NXI 0x02
++
++#define VCF_NONE 0x00
++#define VCF_INFO 0x01
++#define VCF_ADMIN 0x02
++#define VCF_ARES 0x06 /* includes admin */
++#define VCF_SETUP 0x08
++
++#define VCF_ZIDOK 0x10 /* zero id okay */
++
++
++static inline
++long do_vserver(uint32_t cmd, uint32_t id, void __user *data, int compat)
++{
++ long ret;
++ int permit = -1, state = 0;
++ int perm = -1, args = 0, flags = 0;
++ struct vx_info *vxi = NULL;
++ struct nx_info *nxi = NULL;
++
++ switch (cmd) {
++ /* unpriviledged commands */
++ __VCMD(get_version, 0, VCA_NONE, 0);
++ __VCMD(get_vci, 0, VCA_NONE, 0);
++ __VCMD(get_rlimit_mask, 0, VCA_NONE, 0);
++ __VCMD(get_space_mask_v0,0, VCA_NONE, 0);
++ __VCMD(get_space_mask, 0, VCA_NONE, 0);
++ __VCMD(get_space_default,0, VCA_NONE, 0);
++
++ /* info commands */
++ __VCMD(task_xid, 2, VCA_NONE, 0);
++ __VCMD(reset_minmax, 2, VCA_VXI, 0);
++ __VCMD(vx_info, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_bcaps, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_ccaps, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_cflags, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_badness, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_vhi_name, 3, VCA_VXI, VCF_INFO);
++ __VCMD(get_rlimit, 3, VCA_VXI, VCF_INFO);
++
++ __VCMD(ctx_stat, 3, VCA_VXI, VCF_INFO);
++ __VCMD(virt_stat, 3, VCA_VXI, VCF_INFO);
++ __VCMD(sock_stat, 3, VCA_VXI, VCF_INFO);
++ __VCMD(rlimit_stat, 3, VCA_VXI, VCF_INFO);
++
++ __VCMD(task_nid, 2, VCA_NONE, 0);
++ __VCMD(nx_info, 3, VCA_NXI, VCF_INFO);
++ __VCMD(get_ncaps, 3, VCA_NXI, VCF_INFO);
++ __VCMD(get_nflags, 3, VCA_NXI, VCF_INFO);
++
++ __VCMD(task_tag, 2, VCA_NONE, 0);
++
++ __VCMD(get_iattr, 2, VCA_NONE, 0);
++ __VCMD(fget_iattr, 2, VCA_NONE, 0);
++ __VCMD(get_dlimit, 3, VCA_NONE, VCF_INFO);
++ __VCMD(get_sched, 3, VCA_VXI, VCF_INFO);
++ __VCMD(sched_info, 3, VCA_VXI, VCF_INFO | VCF_ZIDOK);
++
++ /* lower admin commands */
++ __VCMD(wait_exit, 4, VCA_VXI, VCF_INFO);
++ __VCMD(ctx_create_v0, 5, VCA_NONE, 0);
++ __VCMD(ctx_create, 5, VCA_NONE, 0);
++ __VCMD(ctx_migrate_v0, 5, VCA_VXI, VCF_ADMIN);
++ __VCMD(ctx_migrate, 5, VCA_VXI, VCF_ADMIN);
++ __VCMD(enter_space_v0, 5, VCA_VXI, VCF_ADMIN);
++ __VCMD(enter_space, 5, VCA_VXI, VCF_ADMIN);
++
++ __VCMD(net_create_v0, 5, VCA_NONE, 0);
++ __VCMD(net_create, 5, VCA_NONE, 0);
++ __VCMD(net_migrate, 5, VCA_NXI, VCF_ADMIN);
++
++ __VCMD(tag_migrate, 5, VCA_NONE, VCF_ADMIN);
++
++ /* higher admin commands */
++ __VCMD(ctx_kill, 6, VCA_VXI, VCF_ARES);
++ __VCMD(set_space, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++
++ __VCMD(set_ccaps, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_bcaps, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_cflags, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_badness, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++
++ __VCMD(set_vhi_name, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_rlimit, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_sched, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_sched_v4, 7, VCA_VXI, VCF_ARES | VCF_SETUP);
++
++ __VCMD(set_ncaps, 7, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(set_nflags, 7, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(net_add, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(net_remove, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(net_add_ipv4, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(net_remove_ipv4, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++#ifdef CONFIG_IPV6
++ __VCMD(net_add_ipv6, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++ __VCMD(net_remove_ipv6, 8, VCA_NXI, VCF_ARES | VCF_SETUP);
++#endif
++ __VCMD(set_iattr, 7, VCA_NONE, 0);
++ __VCMD(fset_iattr, 7, VCA_NONE, 0);
++ __VCMD(set_dlimit, 7, VCA_NONE, VCF_ARES);
++ __VCMD(add_dlimit, 8, VCA_NONE, VCF_ARES);
++ __VCMD(rem_dlimit, 8, VCA_NONE, VCF_ARES);
++
++#ifdef CONFIG_VSERVER_DEVICE
++ __VCMD(set_mapping, 8, VCA_VXI, VCF_ARES|VCF_ZIDOK);
++ __VCMD(unset_mapping, 8, VCA_VXI, VCF_ARES|VCF_ZIDOK);
++#endif
++ /* debug level admin commands */
++#ifdef CONFIG_VSERVER_HISTORY
++ __VCMD(dump_history, 9, VCA_NONE, 0);
++ __VCMD(read_history, 9, VCA_NONE, 0);
++#endif
++#ifdef CONFIG_VSERVER_MONITOR
++ __VCMD(read_monitor, 9, VCA_NONE, 0);
++#endif
++
++ default:
++ perm = -1;
++ }
++
++ vxdprintk(VXD_CBIT(switch, 0),
++ "vc: VCMD_%02d_%d[%d], %d,%p [%d,%d,%x,%x]",
++ VC_CATEGORY(cmd), VC_COMMAND(cmd),
++ VC_VERSION(cmd), id, data, compat,
++ perm, args, flags);
++
++ ret = -ENOSYS;
++ if (perm < 0)
++ goto out;
++
++ state = 1;
++ if (!capable(CAP_CONTEXT))
++ goto out;
++
++ state = 2;
++ /* moved here from the individual commands */
++ ret = -EPERM;
++ if ((perm > 1) && !capable(CAP_SYS_ADMIN))
++ goto out;
++
++ state = 3;
++ /* vcmd involves resource management */
++ ret = -EPERM;
++ if ((flags & VCF_ARES) && !capable(CAP_SYS_RESOURCE))
++ goto out;
++
++ state = 4;
++ /* various legacy exceptions */
++ switch (cmd) {
++ /* will go away when spectator is a cap */
++ case VCMD_ctx_migrate_v0:
++ case VCMD_ctx_migrate:
++ if (id == 1) {
++ current->xid = 1;
++ ret = 1;
++ goto out;
++ }
++ break;
++
++ /* will go away when spectator is a cap */
++ case VCMD_net_migrate:
++ if (id == 1) {
++ current->nid = 1;
++ ret = 1;
++ goto out;
++ }
++ break;
++ }
++
++ /* vcmds are fine by default */
++ permit = 1;
++
++ /* admin type vcmds require admin ... */
++ if (flags & VCF_ADMIN)
++ permit = vx_check(0, VS_ADMIN) ? 1 : 0;
++
++ /* ... but setup type vcmds override that */
++ if (!permit && (flags & VCF_SETUP))
++ permit = vx_flags(VXF_STATE_SETUP, 0) ? 2 : 0;
++
++ state = 5;
++ ret = -EPERM;
++ if (!permit)
++ goto out;
++
++ state = 6;
++ if (!id && (flags & VCF_ZIDOK))
++ goto skip_id;
++
++ ret = -ESRCH;
++ if (args & VCA_VXI) {
++ vxi = lookup_vx_info(id);
++ if (!vxi)
++ goto out;
++
++ if ((flags & VCF_ADMIN) &&
++ /* special case kill for shutdown */
++ (cmd != VCMD_ctx_kill) &&
++ /* can context be administrated? */
++ !vx_info_flags(vxi, VXF_STATE_ADMIN, 0)) {
++ ret = -EACCES;
++ goto out_vxi;
++ }
++ }
++ state = 7;
++ if (args & VCA_NXI) {
++ nxi = lookup_nx_info(id);
++ if (!nxi)
++ goto out_vxi;
++
++ if ((flags & VCF_ADMIN) &&
++ /* can context be administrated? */
++ !nx_info_flags(nxi, NXF_STATE_ADMIN, 0)) {
++ ret = -EACCES;
++ goto out_nxi;
++ }
++ }
++skip_id:
++ state = 8;
++ ret = do_vcmd(cmd, id, vxi, nxi, data, compat);
++
++out_nxi:
++ if ((args & VCA_NXI) && nxi)
++ put_nx_info(nxi);
++out_vxi:
++ if ((args & VCA_VXI) && vxi)
++ put_vx_info(vxi);
++out:
++ vxdprintk(VXD_CBIT(switch, 1),
++ "vc: VCMD_%02d_%d[%d] = %08lx(%ld) [%d,%d]",
++ VC_CATEGORY(cmd), VC_COMMAND(cmd),
++ VC_VERSION(cmd), ret, ret, state, permit);
++ return ret;
++}
++
++asmlinkage long
++sys_vserver(uint32_t cmd, uint32_t id, void __user *data)
++{
++ return do_vserver(cmd, id, data, 0);
++}
++
++#ifdef CONFIG_COMPAT
++
++asmlinkage long
++sys32_vserver(uint32_t cmd, uint32_t id, void __user *data)
++{
++ return do_vserver(cmd, id, data, 1);
++}
++
++#endif /* CONFIG_COMPAT */
+--- a/kernel/vserver/sysctl.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/sysctl.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,244 @@
++/*
++ * kernel/vserver/sysctl.c
++ *
++ * Virtual Context Support
++ *
++ * Copyright (C) 2004-2007 Herbert Pötzl
++ *
++ * V0.01 basic structure
++ *
++ */
++
++#include <linux/module.h>
++#include <linux/ctype.h>
++#include <linux/sysctl.h>
++#include <linux/parser.h>
++#include <asm/uaccess.h>
++
++
++enum {
++ CTL_DEBUG_ERROR = 0,
++ CTL_DEBUG_SWITCH = 1,
++ CTL_DEBUG_XID,
++ CTL_DEBUG_NID,
++ CTL_DEBUG_TAG,
++ CTL_DEBUG_NET,
++ CTL_DEBUG_LIMIT,
++ CTL_DEBUG_CRES,
++ CTL_DEBUG_DLIM,
++ CTL_DEBUG_QUOTA,
++ CTL_DEBUG_CVIRT,
++ CTL_DEBUG_SPACE,
++ CTL_DEBUG_MISC,
++};
++
++
++unsigned int vx_debug_switch = 0;
++unsigned int vx_debug_xid = 0;
++unsigned int vx_debug_nid = 0;
++unsigned int vx_debug_tag = 0;
++unsigned int vx_debug_net = 0;
++unsigned int vx_debug_limit = 0;
++unsigned int vx_debug_cres = 0;
++unsigned int vx_debug_dlim = 0;
++unsigned int vx_debug_quota = 0;
++unsigned int vx_debug_cvirt = 0;
++unsigned int vx_debug_space = 0;
++unsigned int vx_debug_misc = 0;
++
++
++static struct ctl_table_header *vserver_table_header;
++static ctl_table vserver_root_table[];
++
++
++void vserver_register_sysctl(void)
++{
++ if (!vserver_table_header) {
++ vserver_table_header = register_sysctl_table(vserver_root_table);
++ }
++
++}
++
++void vserver_unregister_sysctl(void)
++{
++ if (vserver_table_header) {
++ unregister_sysctl_table(vserver_table_header);
++ vserver_table_header = NULL;
++ }
++}
++
++
++static int proc_dodebug(ctl_table *table, int write,
++ struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos)
++{
++ char tmpbuf[20], *p, c;
++ unsigned int value;
++ size_t left, len;
++
++ if ((*ppos && !write) || !*lenp) {
++ *lenp = 0;
++ return 0;
++ }
++
++ left = *lenp;
++
++ if (write) {
++ if (!access_ok(VERIFY_READ, buffer, left))
++ return -EFAULT;
++ p = (char *)buffer;
++ while (left && __get_user(c, p) >= 0 && isspace(c))
++ left--, p++;
++ if (!left)
++ goto done;
++
++ if (left > sizeof(tmpbuf) - 1)
++ return -EINVAL;
++ if (copy_from_user(tmpbuf, p, left))
++ return -EFAULT;
++ tmpbuf[left] = '\0';
++
++ for (p = tmpbuf, value = 0; '0' <= *p && *p <= '9'; p++, left--)
++ value = 10 * value + (*p - '0');
++ if (*p && !isspace(*p))
++ return -EINVAL;
++ while (left && isspace(*p))
++ left--, p++;
++ *(unsigned int *)table->data = value;
++ } else {
++ if (!access_ok(VERIFY_WRITE, buffer, left))
++ return -EFAULT;
++ len = sprintf(tmpbuf, "%d", *(unsigned int *)table->data);
++ if (len > left)
++ len = left;
++ if (__copy_to_user(buffer, tmpbuf, len))
++ return -EFAULT;
++ if ((left -= len) > 0) {
++ if (put_user('\n', (char *)buffer + len))
++ return -EFAULT;
++ left--;
++ }
++ }
++
++done:
++ *lenp -= left;
++ *ppos += *lenp;
++ return 0;
++}
++
++static int zero;
++
++#define CTL_ENTRY(ctl, name) \
++ { \
++ .ctl_name = ctl, \
++ .procname = #name, \
++ .data = &vx_ ## name, \
++ .maxlen = sizeof(int), \
++ .mode = 0644, \
++ .proc_handler = &proc_dodebug, \
++ .strategy = &sysctl_intvec, \
++ .extra1 = &zero, \
++ }
++
++static ctl_table vserver_debug_table[] = {
++ CTL_ENTRY(CTL_DEBUG_SWITCH, debug_switch),
++ CTL_ENTRY(CTL_DEBUG_XID, debug_xid),
++ CTL_ENTRY(CTL_DEBUG_NID, debug_nid),
++ CTL_ENTRY(CTL_DEBUG_TAG, debug_tag),
++ CTL_ENTRY(CTL_DEBUG_NET, debug_net),
++ CTL_ENTRY(CTL_DEBUG_LIMIT, debug_limit),
++ CTL_ENTRY(CTL_DEBUG_CRES, debug_cres),
++ CTL_ENTRY(CTL_DEBUG_DLIM, debug_dlim),
++ CTL_ENTRY(CTL_DEBUG_QUOTA, debug_quota),
++ CTL_ENTRY(CTL_DEBUG_CVIRT, debug_cvirt),
++ CTL_ENTRY(CTL_DEBUG_SPACE, debug_space),
++ CTL_ENTRY(CTL_DEBUG_MISC, debug_misc),
++ { .ctl_name = 0 }
++};
++
++static ctl_table vserver_root_table[] = {
++ {
++ .ctl_name = CTL_VSERVER,
++ .procname = "vserver",
++ .mode = 0555,
++ .child = vserver_debug_table
++ },
++ { .ctl_name = 0 }
++};
++
++
++static match_table_t tokens = {
++ { CTL_DEBUG_SWITCH, "switch=%x" },
++ { CTL_DEBUG_XID, "xid=%x" },
++ { CTL_DEBUG_NID, "nid=%x" },
++ { CTL_DEBUG_TAG, "tag=%x" },
++ { CTL_DEBUG_NET, "net=%x" },
++ { CTL_DEBUG_LIMIT, "limit=%x" },
++ { CTL_DEBUG_CRES, "cres=%x" },
++ { CTL_DEBUG_DLIM, "dlim=%x" },
++ { CTL_DEBUG_QUOTA, "quota=%x" },
++ { CTL_DEBUG_CVIRT, "cvirt=%x" },
++ { CTL_DEBUG_SPACE, "space=%x" },
++ { CTL_DEBUG_MISC, "misc=%x" },
++ { CTL_DEBUG_ERROR, NULL }
++};
++
++#define HANDLE_CASE(id, name, val) \
++ case CTL_DEBUG_ ## id: \
++ vx_debug_ ## name = val; \
++ printk("vs_debug_" #name "=0x%x\n", val); \
++ break
++
++
++static int __init vs_debug_setup(char *str)
++{
++ char *p;
++ int token;
++
++ printk("vs_debug_setup(%s)\n", str);
++ while ((p = strsep(&str, ",")) != NULL) {
++ substring_t args[MAX_OPT_ARGS];
++ unsigned int value;
++
++ if (!*p)
++ continue;
++
++ token = match_token(p, tokens, args);
++ value = (token > 0) ? simple_strtoul(args[0].from, NULL, 0) : 0;
++
++ switch (token) {
++ HANDLE_CASE(SWITCH, switch, value);
++ HANDLE_CASE(XID, xid, value);
++ HANDLE_CASE(NID, nid, value);
++ HANDLE_CASE(TAG, tag, value);
++ HANDLE_CASE(NET, net, value);
++ HANDLE_CASE(LIMIT, limit, value);
++ HANDLE_CASE(CRES, cres, value);
++ HANDLE_CASE(DLIM, dlim, value);
++ HANDLE_CASE(QUOTA, quota, value);
++ HANDLE_CASE(CVIRT, cvirt, value);
++ HANDLE_CASE(SPACE, space, value);
++ HANDLE_CASE(MISC, misc, value);
++ default:
++ return -EINVAL;
++ break;
++ }
++ }
++ return 1;
++}
++
++__setup("vsdebug=", vs_debug_setup);
++
++
++
++EXPORT_SYMBOL_GPL(vx_debug_switch);
++EXPORT_SYMBOL_GPL(vx_debug_xid);
++EXPORT_SYMBOL_GPL(vx_debug_nid);
++EXPORT_SYMBOL_GPL(vx_debug_net);
++EXPORT_SYMBOL_GPL(vx_debug_limit);
++EXPORT_SYMBOL_GPL(vx_debug_cres);
++EXPORT_SYMBOL_GPL(vx_debug_dlim);
++EXPORT_SYMBOL_GPL(vx_debug_quota);
++EXPORT_SYMBOL_GPL(vx_debug_cvirt);
++EXPORT_SYMBOL_GPL(vx_debug_space);
++EXPORT_SYMBOL_GPL(vx_debug_misc);
++
+--- a/kernel/vserver/tag.c 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/tag.c 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,63 @@
++/*
++ * linux/kernel/vserver/tag.c
++ *
++ * Virtual Server: Shallow Tag Space
++ *
++ * Copyright (C) 2007 Herbert Pötzl
++ *
++ * V0.01 basic implementation
++ *
++ */
++
++#include <linux/sched.h>
++#include <linux/vserver/debug.h>
++#include <linux/vs_pid.h>
++#include <linux/vs_tag.h>
++
++#include <linux/vserver/tag_cmd.h>
++
++
++int dx_migrate_task(struct task_struct *p, tag_t tag)
++{
++ if (!p)
++ BUG();
++
++ vxdprintk(VXD_CBIT(tag, 5),
++ "dx_migrate_task(%p[#%d],#%d)", p, p->tag, tag);
++
++ task_lock(p);
++ p->tag = tag;
++ task_unlock(p);
++
++ vxdprintk(VXD_CBIT(tag, 5),
++ "moved task %p into [#%d]", p, tag);
++ return 0;
++}
++
++/* vserver syscall commands below here */
++
++/* taks xid and vx_info functions */
++
++
++int vc_task_tag(uint32_t id)
++{
++ tag_t tag;
++
++ if (id) {
++ struct task_struct *tsk;
++ read_lock(&tasklist_lock);
++ tsk = find_task_by_real_pid(id);
++ tag = (tsk) ? tsk->tag : -ESRCH;
++ read_unlock(&tasklist_lock);
++ } else
++ tag = dx_current_tag();
++ return tag;
++}
++
++
++int vc_tag_migrate(uint32_t tag)
++{
++ return dx_migrate_task(current, tag & 0xFFFF);
++}
++
++
+--- a/kernel/vserver/vci_config.h 1969-12-31 19:00:00.000000000 -0500
++++ a/kernel/vserver/vci_config.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,81 @@
++
++/* interface version */
++
++#define VCI_VERSION 0x00020303
++
++
++enum {
++ VCI_KCBIT_NO_DYNAMIC = 0,
++
++ VCI_KCBIT_PROC_SECURE = 4,
++ VCI_KCBIT_HARDCPU = 5,
++ VCI_KCBIT_IDLELIMIT = 6,
++ VCI_KCBIT_IDLETIME = 7,
++
++ VCI_KCBIT_COWBL = 8,
++ VCI_KCBIT_FULLCOWBL = 9,
++ VCI_KCBIT_SPACES = 10,
++ VCI_KCBIT_NETV2 = 11,
++
++ VCI_KCBIT_DEBUG = 16,
++ VCI_KCBIT_HISTORY = 20,
++ VCI_KCBIT_TAGGED = 24,
++ VCI_KCBIT_PPTAG = 28,
++
++ VCI_KCBIT_MORE = 31,
++};
++
++
++static inline uint32_t vci_kernel_config(void)
++{
++ return
++ (1 << VCI_KCBIT_NO_DYNAMIC) |
++
++ /* configured features */
++#ifdef CONFIG_VSERVER_PROC_SECURE
++ (1 << VCI_KCBIT_PROC_SECURE) |
++#endif
++#ifdef CONFIG_VSERVER_HARDCPU
++ (1 << VCI_KCBIT_HARDCPU) |
++#endif
++#ifdef CONFIG_VSERVER_IDLELIMIT
++ (1 << VCI_KCBIT_IDLELIMIT) |
++#endif
++#ifdef CONFIG_VSERVER_IDLETIME
++ (1 << VCI_KCBIT_IDLETIME) |
++#endif
++#ifdef CONFIG_VSERVER_COWBL
++ (1 << VCI_KCBIT_COWBL) |
++ (1 << VCI_KCBIT_FULLCOWBL) |
++#endif
++ (1 << VCI_KCBIT_SPACES) |
++ (1 << VCI_KCBIT_NETV2) |
++
++ /* debug options */
++#ifdef CONFIG_VSERVER_DEBUG
++ (1 << VCI_KCBIT_DEBUG) |
++#endif
++#ifdef CONFIG_VSERVER_HISTORY
++ (1 << VCI_KCBIT_HISTORY) |
++#endif
++
++ /* inode context tagging */
++#if defined(CONFIG_TAGGING_NONE)
++ (0 << VCI_KCBIT_TAGGED) |
++#elif defined(CONFIG_TAGGING_UID16)
++ (1 << VCI_KCBIT_TAGGED) |
++#elif defined(CONFIG_TAGGING_GID16)
++ (2 << VCI_KCBIT_TAGGED) |
++#elif defined(CONFIG_TAGGING_ID24)
++ (3 << VCI_KCBIT_TAGGED) |
++#elif defined(CONFIG_TAGGING_INTERN)
++ (4 << VCI_KCBIT_TAGGED) |
++#elif defined(CONFIG_TAGGING_RUNTIME)
++ (5 << VCI_KCBIT_TAGGED) |
++#else
++ (7 << VCI_KCBIT_TAGGED) |
++#endif
++ (1 << VCI_KCBIT_PPTAG) |
++ 0;
++}
++
+--- a/mm/filemap_xip.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/filemap_xip.c 2008-07-16 22:41:36.000000000 -0400
+@@ -14,6 +14,7 @@
+ #include <linux/uio.h>
+ #include <linux/rmap.h>
+ #include <linux/sched.h>
++#include <linux/vs_memory.h>
+ #include <asm/tlbflush.h>
+ #include <asm/io.h>
+
+--- a/mm/fremap.c 2008-04-17 12:05:44.000000000 -0400
++++ a/mm/fremap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -15,6 +15,7 @@
+ #include <linux/rmap.h>
+ #include <linux/module.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/mmu_context.h>
+ #include <asm/cacheflush.h>
+--- a/mm/hugetlb.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/hugetlb.c 2008-07-16 22:41:36.000000000 -0400
+@@ -19,6 +19,7 @@
+ #include <asm/pgtable.h>
+
+ #include <linux/hugetlb.h>
++#include <linux/vs_memory.h>
+ #include "internal.h"
+
+ const unsigned long hugetlb_zero = 0, hugetlb_infinity = ~0UL;
+--- a/mm/memory.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/memory.c 2008-07-16 22:41:36.000000000 -0400
+@@ -558,6 +558,9 @@ static int copy_pte_range(struct mm_stru
+ int progress = 0;
+ int rss[2];
+
++ if (!vx_rss_avail(dst_mm, ((end - addr)/PAGE_SIZE + 1)))
++ return -ENOMEM;
++
+ again:
+ rss[1] = rss[0] = 0;
+ dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl);
+@@ -2221,6 +2224,11 @@ static int do_swap_page(struct mm_struct
+ goto out;
+ }
+
++ if (!vx_rss_avail(mm, 1)) {
++ ret = VM_FAULT_OOM;
++ goto out;
++ }
++
+ mark_page_accessed(page);
+ lock_page(page);
+ delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
+@@ -2292,6 +2300,8 @@ static int do_anonymous_page(struct mm_s
+ /* Allocate our own private page. */
+ pte_unmap(page_table);
+
++ if (!vx_rss_avail(mm, 1))
++ goto oom;
+ if (unlikely(anon_vma_prepare(vma)))
+ goto oom;
+ page = alloc_zeroed_user_highpage_movable(vma, address);
+@@ -2606,6 +2616,7 @@ static inline int handle_pte_fault(struc
+ {
+ pte_t entry;
+ spinlock_t *ptl;
++ int ret = 0, type = VXPT_UNKNOWN;
+
+ entry = *pte;
+ if (!pte_present(entry)) {
+@@ -2633,9 +2644,12 @@ static inline int handle_pte_fault(struc
+ if (unlikely(!pte_same(*pte, entry)))
+ goto unlock;
+ if (write_access) {
+- if (!pte_write(entry))
+- return do_wp_page(mm, vma, address,
++ if (!pte_write(entry)) {
++ ret = do_wp_page(mm, vma, address,
+ pte, pmd, ptl, entry);
++ type = VXPT_WRITE;
++ goto out;
++ }
+ entry = pte_mkdirty(entry);
+ }
+ entry = pte_mkyoung(entry);
+@@ -2653,7 +2667,10 @@ static inline int handle_pte_fault(struc
+ }
+ unlock:
+ pte_unmap_unlock(pte, ptl);
+- return 0;
++ ret = 0;
++out:
++ vx_page_fault(mm, vma, type, ret);
++ return ret;
+ }
+
+ /*
+--- a/mm/mlock.c 2008-04-17 10:37:25.000000000 -0400
++++ a/mm/mlock.c 2008-07-16 22:41:36.000000000 -0400
+@@ -12,6 +12,7 @@
+ #include <linux/syscalls.h>
+ #include <linux/sched.h>
+ #include <linux/module.h>
++#include <linux/vs_memory.h>
+
+ int can_do_mlock(void)
+ {
+@@ -76,7 +77,7 @@ success:
+ ret = make_pages_present(start, end);
+ }
+
+- mm->locked_vm -= pages;
++ vx_vmlocked_sub(mm, pages);
+ out:
+ return ret;
+ }
+@@ -134,7 +135,7 @@ static int do_mlock(unsigned long start,
+
+ asmlinkage long sys_mlock(unsigned long start, size_t len)
+ {
+- unsigned long locked;
++ unsigned long locked, grow;
+ unsigned long lock_limit;
+ int error = -ENOMEM;
+
+@@ -145,8 +146,10 @@ asmlinkage long sys_mlock(unsigned long
+ len = PAGE_ALIGN(len + (start & ~PAGE_MASK));
+ start &= PAGE_MASK;
+
+- locked = len >> PAGE_SHIFT;
+- locked += current->mm->locked_vm;
++ grow = len >> PAGE_SHIFT;
++ if (!vx_vmlocked_avail(current->mm, grow))
++ goto out;
++ locked = current->mm->locked_vm + grow;
+
+ lock_limit = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur;
+ lock_limit >>= PAGE_SHIFT;
+@@ -154,6 +157,7 @@ asmlinkage long sys_mlock(unsigned long
+ /* check against resource limits */
+ if ((locked <= lock_limit) || capable(CAP_IPC_LOCK))
+ error = do_mlock(start, len, 1);
++out:
+ up_write(¤t->mm->mmap_sem);
+ return error;
+ }
+@@ -213,6 +217,8 @@ asmlinkage long sys_mlockall(int flags)
+ lock_limit >>= PAGE_SHIFT;
+
+ ret = -ENOMEM;
++ if (!vx_vmlocked_avail(current->mm, current->mm->total_vm))
++ goto out;
+ if (!(flags & MCL_CURRENT) || (current->mm->total_vm <= lock_limit) ||
+ capable(CAP_IPC_LOCK))
+ ret = do_mlockall(flags);
+--- a/mm/mmap.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/mmap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -1210,10 +1210,10 @@ munmap_back:
+ if (correct_wcount)
+ atomic_inc(&inode->i_writecount);
+ out:
+- mm->total_vm += len >> PAGE_SHIFT;
++ vx_vmpages_add(mm, len >> PAGE_SHIFT);
+ vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT);
+ if (vm_flags & VM_LOCKED) {
+- mm->locked_vm += len >> PAGE_SHIFT;
++ vx_vmlocked_add(mm, len >> PAGE_SHIFT);
+ make_pages_present(addr, addr + len);
+ }
+ if ((flags & MAP_POPULATE) && !(flags & MAP_NONBLOCK))
+@@ -1562,9 +1562,9 @@ static int acct_stack_growth(struct vm_a
+ return -ENOMEM;
+
+ /* Ok, everything looks good - let it rip */
+- mm->total_vm += grow;
++ vx_vmpages_add(mm, grow);
+ if (vma->vm_flags & VM_LOCKED)
+- mm->locked_vm += grow;
++ vx_vmlocked_add(mm, grow);
+ vm_stat_account(mm, vma->vm_flags, vma->vm_file, grow);
+ return 0;
+ }
+@@ -1735,9 +1735,9 @@ static void remove_vma_list(struct mm_st
+ do {
+ long nrpages = vma_pages(vma);
+
+- mm->total_vm -= nrpages;
++ vx_vmpages_sub(mm, nrpages);
+ if (vma->vm_flags & VM_LOCKED)
+- mm->locked_vm -= nrpages;
++ vx_vmlocked_sub(mm, nrpages);
+ vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages);
+ vma = remove_vma(vma);
+ } while (vma);
+@@ -1983,6 +1983,8 @@ unsigned long do_brk(unsigned long addr,
+ lock_limit >>= PAGE_SHIFT;
+ if (locked > lock_limit && !capable(CAP_IPC_LOCK))
+ return -EAGAIN;
++ if (!vx_vmlocked_avail(mm, len >> PAGE_SHIFT))
++ return -ENOMEM;
+ }
+
+ /*
+@@ -2009,7 +2011,8 @@ unsigned long do_brk(unsigned long addr,
+ if (mm->map_count > sysctl_max_map_count)
+ return -ENOMEM;
+
+- if (security_vm_enough_memory(len >> PAGE_SHIFT))
++ if (security_vm_enough_memory(len >> PAGE_SHIFT) ||
++ !vx_vmpages_avail(mm, len >> PAGE_SHIFT))
+ return -ENOMEM;
+
+ /* Can we just expand an old private anonymous mapping? */
+@@ -2034,9 +2037,9 @@ unsigned long do_brk(unsigned long addr,
+ vma->vm_page_prot = vm_get_page_prot(flags);
+ vma_link(mm, vma, prev, rb_link, rb_parent);
+ out:
+- mm->total_vm += len >> PAGE_SHIFT;
++ vx_vmpages_add(mm, len >> PAGE_SHIFT);
+ if (flags & VM_LOCKED) {
+- mm->locked_vm += len >> PAGE_SHIFT;
++ vx_vmlocked_add(mm, len >> PAGE_SHIFT);
+ make_pages_present(addr, addr + len);
+ }
+ return addr;
+@@ -2065,6 +2068,11 @@ void exit_mmap(struct mm_struct *mm)
+ free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, 0);
+ tlb_finish_mmu(tlb, 0, end);
+
++ set_mm_counter(mm, file_rss, 0);
++ set_mm_counter(mm, anon_rss, 0);
++ vx_vmpages_sub(mm, mm->total_vm);
++ vx_vmlocked_sub(mm, mm->locked_vm);
++
+ /*
+ * Walk the list again, actually closing and freeing it,
+ * with preemption enabled, without holding any MM locks.
+@@ -2104,7 +2112,8 @@ int insert_vm_struct(struct mm_struct *
+ if (__vma && __vma->vm_start < vma->vm_end)
+ return -ENOMEM;
+ if ((vma->vm_flags & VM_ACCOUNT) &&
+- security_vm_enough_memory_mm(mm, vma_pages(vma)))
++ (security_vm_enough_memory_mm(mm, vma_pages(vma)) ||
++ !vx_vmpages_avail(mm, vma_pages(vma))))
+ return -ENOMEM;
+ vma_link(mm, vma, prev, rb_link, rb_parent);
+ return 0;
+@@ -2180,6 +2189,8 @@ int may_expand_vm(struct mm_struct *mm,
+
+ if (cur + npages > lim)
+ return 0;
++ if (!vx_vmpages_avail(mm, npages))
++ return 0;
+ return 1;
+ }
+
+@@ -2257,7 +2268,6 @@ int install_special_mapping(struct mm_st
+ return -ENOMEM;
+ }
+
+- mm->total_vm += len >> PAGE_SHIFT;
+-
++ vx_vmpages_add(mm, len >> PAGE_SHIFT);
+ return 0;
+ }
+--- a/mm/mremap.c 2008-04-17 11:31:40.000000000 -0400
++++ a/mm/mremap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/highmem.h>
+ #include <linux/security.h>
+ #include <linux/syscalls.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/cacheflush.h>
+@@ -212,7 +213,7 @@ static unsigned long move_vma(struct vm_
+ * If this were a serious issue, we'd add a flag to do_munmap().
+ */
+ hiwater_vm = mm->hiwater_vm;
+- mm->total_vm += new_len >> PAGE_SHIFT;
++ vx_vmpages_add(mm, new_len >> PAGE_SHIFT);
+ vm_stat_account(mm, vma->vm_flags, vma->vm_file, new_len>>PAGE_SHIFT);
+
+ if (do_munmap(mm, old_addr, old_len) < 0) {
+@@ -230,7 +231,7 @@ static unsigned long move_vma(struct vm_
+ }
+
+ if (vm_flags & VM_LOCKED) {
+- mm->locked_vm += new_len >> PAGE_SHIFT;
++ vx_vmlocked_add(mm, new_len >> PAGE_SHIFT);
+ if (new_len > old_len)
+ make_pages_present(new_addr + old_len,
+ new_addr + new_len);
+@@ -341,6 +342,9 @@ unsigned long do_mremap(unsigned long ad
+ ret = -EAGAIN;
+ if (locked > lock_limit && !capable(CAP_IPC_LOCK))
+ goto out;
++ if (!vx_vmlocked_avail(current->mm,
++ (new_len - old_len) >> PAGE_SHIFT))
++ goto out;
+ }
+ if (!may_expand_vm(mm, (new_len - old_len) >> PAGE_SHIFT)) {
+ ret = -ENOMEM;
+@@ -369,10 +373,10 @@ unsigned long do_mremap(unsigned long ad
+ vma_adjust(vma, vma->vm_start,
+ addr + new_len, vma->vm_pgoff, NULL);
+
+- mm->total_vm += pages;
++ vx_vmpages_add(mm, pages);
+ vm_stat_account(mm, vma->vm_flags, vma->vm_file, pages);
+ if (vma->vm_flags & VM_LOCKED) {
+- mm->locked_vm += pages;
++ vx_vmlocked_add(mm, pages);
+ make_pages_present(addr + old_len,
+ addr + new_len);
+ }
+--- a/mm/nommu.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/nommu.c 2008-07-16 22:41:36.000000000 -0400
+@@ -1007,7 +1007,7 @@ unsigned long do_mmap_pgoff(struct file
+ realalloc += kobjsize(vma);
+ askedalloc += sizeof(*vma);
+
+- current->mm->total_vm += len >> PAGE_SHIFT;
++ vx_vmpages_add(current->mm, len >> PAGE_SHIFT);
+
+ add_nommu_vma(vma);
+
+@@ -1139,7 +1139,7 @@ int do_munmap(struct mm_struct *mm, unsi
+ kfree(vml);
+
+ update_hiwater_vm(mm);
+- mm->total_vm -= len >> PAGE_SHIFT;
++ vx_vmpages_sub(mm, len >> PAGE_SHIFT);
+
+ #ifdef DEBUG
+ show_process_blocks();
+@@ -1172,7 +1172,7 @@ void exit_mmap(struct mm_struct * mm)
+ printk("Exit_mmap:\n");
+ #endif
+
+- mm->total_vm = 0;
++ vx_vmpages_sub(mm, mm->total_vm);
+
+ while ((tmp = mm->context.vmlist)) {
+ mm->context.vmlist = tmp->next;
+--- a/mm/oom_kill.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/oom_kill.c 2008-07-16 22:41:36.000000000 -0400
+@@ -26,6 +26,7 @@
+ #include <linux/module.h>
+ #include <linux/notifier.h>
+ #include <linux/memcontrol.h>
++#include <linux/vs_memory.h>
+
+ int sysctl_panic_on_oom;
+ int sysctl_oom_kill_allocating_task;
+@@ -72,6 +73,12 @@ unsigned long badness(struct task_struct
+ points = mm->total_vm;
+
+ /*
++ * add points for context badness
++ */
++
++ points += vx_badness(p, mm);
++
++ /*
+ * After this unlock we can no longer dereference local variable `mm'
+ */
+ task_unlock(p);
+@@ -161,8 +168,8 @@ unsigned long badness(struct task_struct
+ }
+
+ #ifdef DEBUG
+- printk(KERN_DEBUG "OOMkill: task %d (%s) got %lu points\n",
+- p->pid, p->comm, points);
++ printk(KERN_DEBUG "OOMkill: task %d:#%u (%s) got %d points\n",
++ task_pid_nr(p), p->xid, p->comm, points);
+ #endif
+ return points;
+ }
+@@ -323,8 +330,8 @@ static void __oom_kill_task(struct task_
+ }
+
+ if (verbose)
+- printk(KERN_ERR "Killed process %d (%s)\n",
+- task_pid_nr(p), p->comm);
++ printk(KERN_ERR "Killed process %d:#%u (%s)\n",
++ task_pid_nr(p), p->xid, p->comm);
+
+ /*
+ * We give our sacrificial lamb high priority and access to
+@@ -404,8 +411,8 @@ static int oom_kill_process(struct task_
+ return 0;
+ }
+
+- printk(KERN_ERR "%s: kill process %d (%s) score %li or a child\n",
+- message, task_pid_nr(p), p->comm, points);
++ printk(KERN_ERR "%s: kill process %d:#%u (%s) score %li or a child\n",
++ message, task_pid_nr(p), p->xid, p->comm, points);
+
+ /* Try to kill a child first */
+ list_for_each_entry(c, &p->children, sibling) {
+--- a/mm/page_alloc.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/page_alloc.c 2008-07-17 16:47:26.000000000 -0400
+@@ -46,6 +46,8 @@
+ #include <linux/page-isolation.h>
+ #include <linux/memcontrol.h>
+ #include <linux/debugobjects.h>
++#include <linux/vs_base.h>
++#include <linux/vs_limit.h>
+
+ #include <asm/tlbflush.h>
+ #include <asm/div64.h>
+@@ -1763,6 +1765,9 @@ void si_meminfo(struct sysinfo *val)
+ val->totalhigh = totalhigh_pages;
+ val->freehigh = nr_free_highpages();
+ val->mem_unit = PAGE_SIZE;
++
++ if (vx_flags(VXF_VIRT_MEM, 0))
++ vx_vsi_meminfo(val);
+ }
+
+ EXPORT_SYMBOL(si_meminfo);
+@@ -1783,6 +1788,9 @@ void si_meminfo_node(struct sysinfo *val
+ val->freehigh = 0;
+ #endif
+ val->mem_unit = PAGE_SIZE;
++
++ if (vx_flags(VXF_VIRT_MEM, 0))
++ vx_vsi_meminfo(val);
+ }
+ #endif
+
+--- a/mm/rmap.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/rmap.c 2008-07-16 22:41:36.000000000 -0400
+@@ -49,6 +49,7 @@
+ #include <linux/module.h>
+ #include <linux/kallsyms.h>
+ #include <linux/memcontrol.h>
++#include <linux/vs_memory.h>
+
+ #include <asm/tlbflush.h>
+
+--- a/mm/shmem.c 2008-08-12 01:41:51.000000000 -0400
++++ a/mm/shmem.c 2008-08-12 01:42:21.000000000 -0400
+@@ -56,7 +56,6 @@
+ #include <asm/pgtable.h>
+
+ /* This magic number is used in glibc for posix shared memory */
+-#define TMPFS_MAGIC 0x01021994
+
+ #define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long))
+ #define ENTRIES_PER_PAGEPAGE (ENTRIES_PER_PAGE*ENTRIES_PER_PAGE)
+@@ -1716,7 +1715,7 @@ static int shmem_statfs(struct dentry *d
+ {
+ struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb);
+
+- buf->f_type = TMPFS_MAGIC;
++ buf->f_type = TMPFS_SUPER_MAGIC;
+ buf->f_bsize = PAGE_CACHE_SIZE;
+ buf->f_namelen = NAME_MAX;
+ spin_lock(&sbinfo->stat_lock);
+@@ -2284,7 +2283,7 @@ static int shmem_fill_super(struct super
+ sb->s_maxbytes = SHMEM_MAX_BYTES;
+ sb->s_blocksize = PAGE_CACHE_SIZE;
+ sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
+- sb->s_magic = TMPFS_MAGIC;
++ sb->s_magic = TMPFS_SUPER_MAGIC;
+ sb->s_op = &shmem_ops;
+ sb->s_time_gran = 1;
+ #ifdef CONFIG_TMPFS_POSIX_ACL
+--- a/mm/slab.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/slab.c 2008-07-16 22:41:36.000000000 -0400
+@@ -508,6 +508,8 @@ struct kmem_cache {
+ #define STATS_INC_FREEMISS(x) do { } while (0)
+ #endif
+
++#include "slab_vs.h"
++
+ #if DEBUG
+
+ /*
+@@ -3346,6 +3348,7 @@ retry:
+
+ obj = slab_get_obj(cachep, slabp, nodeid);
+ check_slabp(cachep, slabp);
++ vx_slab_alloc(cachep, flags);
+ l3->free_objects--;
+ /* move slabp to correct slabp list: */
+ list_del(&slabp->list);
+@@ -3418,6 +3421,7 @@ __cache_alloc_node(struct kmem_cache *ca
+ /* ___cache_alloc_node can fall back to other nodes */
+ ptr = ____cache_alloc_node(cachep, flags, nodeid);
+ out:
++ vx_slab_alloc(cachep, flags);
+ local_irq_restore(save_flags);
+ ptr = cache_alloc_debugcheck_after(cachep, flags, ptr, caller);
+
+@@ -3589,6 +3593,7 @@ static inline void __cache_free(struct k
+
+ check_irq_off();
+ objp = cache_free_debugcheck(cachep, objp, __builtin_return_address(0));
++ vx_slab_free(cachep);
+
+ /*
+ * Skip calling cache_free_alien() when the platform is not numa.
+--- a/mm/slab_vs.h 1969-12-31 19:00:00.000000000 -0500
++++ a/mm/slab_vs.h 2008-07-16 22:41:36.000000000 -0400
+@@ -0,0 +1,27 @@
++
++#include <linux/vserver/context.h>
++
++#include <linux/vs_context.h>
++
++static inline
++void vx_slab_alloc(struct kmem_cache *cachep, gfp_t flags)
++{
++ int what = gfp_zone(cachep->gfpflags);
++
++ if (!current->vx_info)
++ return;
++
++ atomic_add(cachep->buffer_size, ¤t->vx_info->cacct.slab[what]);
++}
++
++static inline
++void vx_slab_free(struct kmem_cache *cachep)
++{
++ int what = gfp_zone(cachep->gfpflags);
++
++ if (!current->vx_info)
++ return;
++
++ atomic_sub(cachep->buffer_size, ¤t->vx_info->cacct.slab[what]);
++}
++
+--- a/mm/swapfile.c 2008-07-14 17:22:56.000000000 -0400
++++ a/mm/swapfile.c 2008-07-16 22:41:36.000000000 -0400
+@@ -32,6 +32,8 @@
+ #include <asm/pgtable.h>
+ #include <asm/tlbflush.h>
+ #include <linux/swapops.h>
++#include <linux/vs_base.h>
++#include <linux/vs_memory.h>
+
+ DEFINE_SPINLOCK(swap_lock);
+ unsigned int nr_swapfiles;
+@@ -1747,6 +1749,8 @@ void si_swapinfo(struct sysinfo *val)
+ val->freeswap = nr_swap_pages + nr_to_be_unused;
+ val->totalswap = total_swap_pages + nr_to_be_unused;
+ spin_unlock(&swap_lock);
++ if (vx_flags(VXF_VIRT_MEM, 0))
++ vx_vsi_swapinfo(val);
+ }
+
+ /*
+--- a/net/core/dev.c 2008-07-14 17:22:57.000000000 -0400
++++ a/net/core/dev.c 2008-07-17 16:46:49.000000000 -0400
+@@ -120,6 +120,7 @@
+ #include <linux/ctype.h>
+ #include <linux/if_arp.h>
+ #include <linux/if_vlan.h>
++#include <linux/vs_inet.h>
+
+ #include "net-sysfs.h"
+
+@@ -2366,6 +2367,8 @@ static int dev_ifconf(struct net *net, c
+
+ total = 0;
+ for_each_netdev(net, dev) {
++ if (!nx_dev_visible(current->nx_info, dev))
++ continue;
+ for (i = 0; i < NPROTO; i++) {
+ if (gifconf_list[i]) {
+ int done;
+@@ -2434,6 +2437,9 @@ static void dev_seq_printf_stats(struct
+ {
+ struct net_device_stats *stats = dev->get_stats(dev);
+
++ if (!nx_dev_visible(current->nx_info, dev))
++ return;
++
+ seq_printf(seq, "%6s:%8lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu "
+ "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n",
+ dev->name, stats->rx_bytes, stats->rx_packets,
+--- a/net/core/rtnetlink.c 2008-07-14 17:22:57.000000000 -0400
++++ a/net/core/rtnetlink.c 2008-07-16 22:41:36.000000000 -0400
+@@ -688,6 +688,8 @@ static int rtnl_dump_ifinfo(struct sk_bu
+
+ idx = 0;
+ for_each_netdev(net, dev) {
++ if (!nx_dev_visible(skb->sk->sk_nx_info, dev))
++ continue;
+ if (idx < s_idx)
+ goto cont;
+ if (rtnl_fill_ifinfo(skb, dev, RTM_NEWLINK,
+@@ -1221,6 +1223,9 @@ void rtmsg_ifinfo(int type, struct net_d
+ struct sk_buff *skb;
+ int err = -ENOBUFS;
+
++ if (!nx_dev_visible(current->nx_info, dev))
++ return;
++
+ skb = nlmsg_new(if_nlmsg_size(dev), GFP_KERNEL);
+ if (skb == NULL)
+ goto errout;
+--- a/net/core/sock.c 2008-07-14 17:22:57.000000000 -0400
++++ a/net/core/sock.c 2008-07-17 16:45:58.000000000 -0400
+@@ -126,6 +126,10 @@
+ #include <linux/ipsec.h>
+
+ #include <linux/filter.h>
++#include <linux/vs_socket.h>
++#include <linux/vs_limit.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
+
+ #ifdef CONFIG_INET
+ #include <net/tcp.h>
+@@ -899,6 +903,8 @@ static struct sock *sk_prot_alloc(struct
+ if (!try_module_get(prot->owner))
+ goto out_free_sec;
+ }
++ sock_vx_init(sk);
++ sock_nx_init(sk);
+
+ return sk;
+
+@@ -975,6 +981,11 @@ void sk_free(struct sock *sk)
+ __func__, atomic_read(&sk->sk_omem_alloc));
+
+ put_net(sock_net(sk));
++ vx_sock_dec(sk);
++ clr_vx_info(&sk->sk_vx_info);
++ sk->sk_xid = -1;
++ clr_nx_info(&sk->sk_nx_info);
++ sk->sk_nid = -1;
+ sk_prot_free(sk->sk_prot_creator, sk);
+ }
+
+@@ -1010,6 +1021,8 @@ struct sock *sk_clone(const struct sock
+
+ /* SANITY */
+ get_net(sock_net(newsk));
++ sock_vx_init(newsk);
++ sock_nx_init(newsk);
+ sk_node_init(&newsk->sk_node);
+ sock_lock_init(newsk);
+ bh_lock_sock(newsk);
+@@ -1056,6 +1069,12 @@ struct sock *sk_clone(const struct sock
+ newsk->sk_priority = 0;
+ atomic_set(&newsk->sk_refcnt, 2);
+
++ set_vx_info(&newsk->sk_vx_info, sk->sk_vx_info);
++ newsk->sk_xid = sk->sk_xid;
++ vx_sock_inc(newsk);
++ set_nx_info(&newsk->sk_nx_info, sk->sk_nx_info);
++ newsk->sk_nid = sk->sk_nid;
++
+ /*
+ * Increment the counter in the same struct proto as the master
+ * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that
+@@ -1740,6 +1759,11 @@ void sock_init_data(struct socket *sock,
+
+ sk->sk_stamp = ktime_set(-1L, 0);
+
++ set_vx_info(&sk->sk_vx_info, current->vx_info);
++ sk->sk_xid = vx_current_xid();
++ vx_sock_inc(sk);
++ set_nx_info(&sk->sk_nx_info, current->nx_info);
++ sk->sk_nid = nx_current_nid();
+ atomic_set(&sk->sk_refcnt, 1);
+ atomic_set(&sk->sk_drops, 0);
+ }
+--- a/net/ipv4/af_inet.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/af_inet.c 2008-07-29 18:05:52.000000000 -0400
+@@ -115,6 +115,7 @@
+ #ifdef CONFIG_IP_MROUTE
+ #include <linux/mroute.h>
+ #endif
++#include <linux/vs_limit.h>
+
+ DEFINE_SNMP_STAT(struct linux_mib, net_statistics) __read_mostly;
+
+@@ -331,9 +332,12 @@ lookup_protocol:
+ }
+
+ err = -EPERM;
++ if ((protocol == IPPROTO_ICMP) &&
++ nx_capable(answer->capability, NXC_RAW_ICMP))
++ goto override;
+ if (answer->capability > 0 && !capable(answer->capability))
+ goto out_rcu_unlock;
+-
++override:
+ err = -EAFNOSUPPORT;
+ if (!inet_netns_ok(net, protocol))
+ goto out_rcu_unlock;
+@@ -451,6 +455,7 @@ int inet_bind(struct socket *sock, struc
+ struct sockaddr_in *addr = (struct sockaddr_in *)uaddr;
+ struct sock *sk = sock->sk;
+ struct inet_sock *inet = inet_sk(sk);
++ struct nx_v4_sock_addr nsa;
+ unsigned short snum;
+ int chk_addr_ret;
+ int err;
+@@ -464,7 +469,11 @@ int inet_bind(struct socket *sock, struc
+ if (addr_len < sizeof(struct sockaddr_in))
+ goto out;
+
+- chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr);
++ err = v4_map_sock_addr(inet, addr, &nsa);
++ if (err)
++ goto out;
++
++ chk_addr_ret = inet_addr_type(sock_net(sk), nsa.saddr);
+
+ /* Not specified by any standard per-se, however it breaks too
+ * many applications when removed. It is unfortunate since
+@@ -476,7 +485,7 @@ int inet_bind(struct socket *sock, struc
+ err = -EADDRNOTAVAIL;
+ if (!sysctl_ip_nonlocal_bind &&
+ !inet->freebind &&
+- addr->sin_addr.s_addr != htonl(INADDR_ANY) &&
++ nsa.saddr != htonl(INADDR_ANY) &&
+ chk_addr_ret != RTN_LOCAL &&
+ chk_addr_ret != RTN_MULTICAST &&
+ chk_addr_ret != RTN_BROADCAST)
+@@ -501,7 +510,7 @@ int inet_bind(struct socket *sock, struc
+ if (sk->sk_state != TCP_CLOSE || inet->num)
+ goto out_release_sock;
+
+- inet->rcv_saddr = inet->saddr = addr->sin_addr.s_addr;
++ v4_set_sock_addr(inet, &nsa);
+ if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST)
+ inet->saddr = 0; /* Use device */
+
+@@ -694,11 +703,13 @@ int inet_getname(struct socket *sock, st
+ peer == 1))
+ return -ENOTCONN;
+ sin->sin_port = inet->dport;
+- sin->sin_addr.s_addr = inet->daddr;
++ sin->sin_addr.s_addr =
++ nx_map_sock_lback(sk->sk_nx_info, inet->daddr);
+ } else {
+ __be32 addr = inet->rcv_saddr;
+ if (!addr)
+ addr = inet->saddr;
++ addr = nx_map_sock_lback(sk->sk_nx_info, addr);
+ sin->sin_port = inet->sport;
+ sin->sin_addr.s_addr = addr;
+ }
+--- a/net/ipv4/devinet.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/devinet.c 2008-07-29 17:27:07.000000000 -0400
+@@ -420,6 +420,7 @@ struct in_device *inetdev_by_index(struc
+ return in_dev;
+ }
+
++
+ /* Called only from RTNL semaphored context. No locks. */
+
+ struct in_ifaddr *inet_ifa_byprefix(struct in_device *in_dev, __be32 prefix,
+@@ -662,6 +663,8 @@ int devinet_ioctl(struct net *net, unsig
+ *colon = ':';
+
+ if ((in_dev = __in_dev_get_rtnl(dev)) != NULL) {
++ struct nx_info *nxi = current->nx_info;
++
+ if (tryaddrmatch) {
+ /* Matthias Andree */
+ /* compare label and address (4.4BSD style) */
+@@ -670,6 +673,8 @@ int devinet_ioctl(struct net *net, unsig
+ This is checked above. */
+ for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
+ ifap = &ifa->ifa_next) {
++ if (!nx_v4_ifa_visible(nxi, ifa))
++ continue;
+ if (!strcmp(ifr.ifr_name, ifa->ifa_label) &&
+ sin_orig.sin_addr.s_addr ==
+ ifa->ifa_address) {
+@@ -682,9 +687,12 @@ int devinet_ioctl(struct net *net, unsig
+ comparing just the label */
+ if (!ifa) {
+ for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
+- ifap = &ifa->ifa_next)
++ ifap = &ifa->ifa_next) {
++ if (!nx_v4_ifa_visible(nxi, ifa))
++ continue;
+ if (!strcmp(ifr.ifr_name, ifa->ifa_label))
+ break;
++ }
+ }
+ }
+
+@@ -835,6 +843,8 @@ static int inet_gifconf(struct net_devic
+ goto out;
+
+ for (; ifa; ifa = ifa->ifa_next) {
++ if (!nx_v4_ifa_visible(current->nx_info, ifa))
++ continue;
+ if (!buf) {
+ done += sizeof(ifr);
+ continue;
+@@ -1154,6 +1164,7 @@ static int inet_dump_ifaddr(struct sk_bu
+ struct net_device *dev;
+ struct in_device *in_dev;
+ struct in_ifaddr *ifa;
++ struct sock *sk = skb->sk;
+ int s_ip_idx, s_idx = cb->args[0];
+
+ s_ip_idx = ip_idx = cb->args[1];
+@@ -1168,6 +1179,8 @@ static int inet_dump_ifaddr(struct sk_bu
+
+ for (ifa = in_dev->ifa_list, ip_idx = 0; ifa;
+ ifa = ifa->ifa_next, ip_idx++) {
++ if (sk && !nx_v4_ifa_visible(sk->sk_nx_info, ifa))
++ continue;
+ if (ip_idx < s_ip_idx)
+ continue;
+ if (inet_fill_ifaddr(skb, ifa, NETLINK_CB(cb->skb).pid,
+--- a/net/ipv4/fib_hash.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/fib_hash.c 2008-07-29 17:58:34.000000000 -0400
+@@ -1024,7 +1024,7 @@ static int fib_seq_show(struct seq_file
+ prefix = f->fn_key;
+ mask = FZ_MASK(iter->zone);
+ flags = fib_flag_trans(fa->fa_type, mask, fi);
+- if (fi)
++ if (fi && nx_dev_visible(current->nx_info, fi->fib_dev))
+ seq_printf(seq,
+ "%s\t%08X\t%08X\t%04X\t%d\t%u\t%d\t%08X\t%d\t%u\t%u%n",
+ fi->fib_dev ? fi->fib_dev->name : "*", prefix,
+--- a/net/ipv4/inet_connection_sock.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/inet_connection_sock.c 2008-07-29 17:27:07.000000000 -0400
+@@ -47,10 +47,40 @@ void inet_get_local_port_range(int *low,
+ }
+ EXPORT_SYMBOL(inet_get_local_port_range);
+
++int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
++{
++ __be32 sk1_rcv_saddr = inet_rcv_saddr(sk1),
++ sk2_rcv_saddr = inet_rcv_saddr(sk2);
++
++ if (inet_v6_ipv6only(sk2))
++ return 0;
++
++ if (sk1_rcv_saddr &&
++ sk2_rcv_saddr &&
++ sk1_rcv_saddr == sk2_rcv_saddr)
++ return 1;
++
++ if (sk1_rcv_saddr &&
++ !sk2_rcv_saddr &&
++ v4_addr_in_nx_info(sk2->sk_nx_info, sk1_rcv_saddr, NXA_MASK_BIND))
++ return 1;
++
++ if (sk2_rcv_saddr &&
++ !sk1_rcv_saddr &&
++ v4_addr_in_nx_info(sk1->sk_nx_info, sk2_rcv_saddr, NXA_MASK_BIND))
++ return 1;
++
++ if (!sk1_rcv_saddr &&
++ !sk2_rcv_saddr &&
++ nx_v4_addr_conflict(sk1->sk_nx_info, sk2->sk_nx_info))
++ return 1;
++
++ return 0;
++}
++
+ int inet_csk_bind_conflict(const struct sock *sk,
+ const struct inet_bind_bucket *tb)
+ {
+- const __be32 sk_rcv_saddr = inet_rcv_saddr(sk);
+ struct sock *sk2;
+ struct hlist_node *node;
+ int reuse = sk->sk_reuse;
+@@ -70,9 +100,7 @@ int inet_csk_bind_conflict(const struct
+ sk->sk_bound_dev_if == sk2->sk_bound_dev_if)) {
+ if (!reuse || !sk2->sk_reuse ||
+ sk2->sk_state == TCP_LISTEN) {
+- const __be32 sk2_rcv_saddr = inet_rcv_saddr(sk2);
+- if (!sk2_rcv_saddr || !sk_rcv_saddr ||
+- sk2_rcv_saddr == sk_rcv_saddr)
++ if (ipv4_rcv_saddr_equal(sk, sk2))
+ break;
+ }
+ }
+--- a/net/ipv4/inet_diag.c 2008-04-17 12:05:44.000000000 -0400
++++ a/net/ipv4/inet_diag.c 2008-07-29 17:27:07.000000000 -0400
+@@ -34,6 +34,8 @@
+ #include <linux/stddef.h>
+
+ #include <linux/inet_diag.h>
++#include <linux/vs_network.h>
++#include <linux/vs_inet.h>
+
+ static const struct inet_diag_handler **inet_diag_table;
+
+@@ -122,8 +124,8 @@ static int inet_csk_diag_fill(struct soc
+
+ r->id.idiag_sport = inet->sport;
+ r->id.idiag_dport = inet->dport;
+- r->id.idiag_src[0] = inet->rcv_saddr;
+- r->id.idiag_dst[0] = inet->daddr;
++ r->id.idiag_src[0] = nx_map_sock_lback(sk->sk_nx_info, inet->rcv_saddr);
++ r->id.idiag_dst[0] = nx_map_sock_lback(sk->sk_nx_info, inet->daddr);
+
+ #if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+ if (r->idiag_family == AF_INET6) {
+@@ -210,8 +212,8 @@ static int inet_twsk_diag_fill(struct in
+ r->id.idiag_cookie[1] = (u32)(((unsigned long)tw >> 31) >> 1);
+ r->id.idiag_sport = tw->tw_sport;
+ r->id.idiag_dport = tw->tw_dport;
+- r->id.idiag_src[0] = tw->tw_rcv_saddr;
+- r->id.idiag_dst[0] = tw->tw_daddr;
++ r->id.idiag_src[0] = nx_map_sock_lback(tw->tw_nx_info, tw->tw_rcv_saddr);
++ r->id.idiag_dst[0] = nx_map_sock_lback(tw->tw_nx_info, tw->tw_daddr);
+ r->idiag_state = tw->tw_substate;
+ r->idiag_timer = 3;
+ r->idiag_expires = DIV_ROUND_UP(tmo * 1000, HZ);
+@@ -268,6 +270,7 @@ static int inet_diag_get_exact(struct sk
+ err = -EINVAL;
+
+ if (req->idiag_family == AF_INET) {
++ /* TODO: lback */
+ sk = inet_lookup(&init_net, hashinfo, req->id.idiag_dst[0],
+ req->id.idiag_dport, req->id.idiag_src[0],
+ req->id.idiag_sport, req->id.idiag_if);
+@@ -510,6 +513,7 @@ static int inet_csk_diag_dump(struct soc
+ } else
+ #endif
+ {
++ /* TODO: lback */
+ entry.saddr = &inet->rcv_saddr;
+ entry.daddr = &inet->daddr;
+ }
+@@ -546,6 +550,7 @@ static int inet_twsk_diag_dump(struct in
+ } else
+ #endif
+ {
++ /* TODO: lback */
+ entry.saddr = &tw->tw_rcv_saddr;
+ entry.daddr = &tw->tw_daddr;
+ }
+@@ -592,8 +597,8 @@ static int inet_diag_fill_req(struct sk_
+
+ r->id.idiag_sport = inet->sport;
+ r->id.idiag_dport = ireq->rmt_port;
+- r->id.idiag_src[0] = ireq->loc_addr;
+- r->id.idiag_dst[0] = ireq->rmt_addr;
++ r->id.idiag_src[0] = nx_map_sock_lback(sk->sk_nx_info, ireq->loc_addr);
++ r->id.idiag_dst[0] = nx_map_sock_lback(sk->sk_nx_info, ireq->rmt_addr);
+ r->idiag_expires = jiffies_to_msecs(tmo);
+ r->idiag_rqueue = 0;
+ r->idiag_wqueue = 0;
+@@ -663,6 +668,7 @@ static int inet_diag_dump_reqs(struct sk
+ continue;
+
+ if (bc) {
++ /* TODO: lback */
+ entry.saddr =
+ #if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
+ (entry.family == AF_INET6) ?
+@@ -731,6 +737,8 @@ static int inet_diag_dump(struct sk_buff
+ sk_for_each(sk, node, &hashinfo->listening_hash[i]) {
+ struct inet_sock *inet = inet_sk(sk);
+
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (num < s_num) {
+ num++;
+ continue;
+@@ -792,6 +800,8 @@ skip_listen_ht:
+ sk_for_each(sk, node, &head->chain) {
+ struct inet_sock *inet = inet_sk(sk);
+
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (num < s_num)
+ goto next_normal;
+ if (!(r->idiag_states & (1 << sk->sk_state)))
+@@ -816,6 +826,8 @@ next_normal:
+ inet_twsk_for_each(tw, node,
+ &head->twchain) {
+
++ if (!nx_check(tw->tw_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (num < s_num)
+ goto next_dying;
+ if (r->id.idiag_sport != tw->tw_sport &&
+--- a/net/ipv4/inet_hashtables.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/inet_hashtables.c 2008-07-29 17:27:07.000000000 -0400
+@@ -21,6 +21,7 @@
+
+ #include <net/inet_connection_sock.h>
+ #include <net/inet_hashtables.h>
++#include <net/route.h>
+ #include <net/ip.h>
+
+ /*
+@@ -161,11 +162,10 @@ static struct sock *inet_lookup_listener
+ const __be32 rcv_saddr = inet->rcv_saddr;
+ int score = sk->sk_family == PF_INET ? 1 : 0;
+
+- if (rcv_saddr) {
+- if (rcv_saddr != daddr)
+- continue;
++ if (v4_inet_addr_match(sk->sk_nx_info, daddr, rcv_saddr))
+ score += 2;
+- }
++ else
++ continue;
+ if (sk->sk_bound_dev_if) {
+ if (sk->sk_bound_dev_if != dif)
+ continue;
+@@ -197,7 +197,7 @@ struct sock *__inet_lookup_listener(stru
+ const struct inet_sock *inet = inet_sk((sk = __sk_head(head)));
+
+ if (inet->num == hnum && !sk->sk_node.next &&
+- (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
++ v4_inet_addr_match(sk->sk_nx_info, daddr, inet->rcv_saddr) &&
+ (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
+ !sk->sk_bound_dev_if && net_eq(sock_net(sk), net))
+ goto sherry_cache;
+--- a/net/ipv4/netfilter/nf_nat_helper.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/netfilter/nf_nat_helper.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <net/tcp.h>
+
+ #include <linux/netfilter_ipv4.h>
++#include <net/route.h>
+ #include <net/netfilter/nf_conntrack.h>
+ #include <net/netfilter/nf_conntrack_helper.h>
+ #include <net/netfilter/nf_conntrack_ecache.h>
+--- a/net/ipv4/netfilter.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/netfilter.c 2008-07-16 22:41:36.000000000 -0400
+@@ -4,7 +4,7 @@
+ #include <linux/netfilter_ipv4.h>
+ #include <linux/ip.h>
+ #include <linux/skbuff.h>
+-#include <net/route.h>
++// #include <net/route.h>
+ #include <net/xfrm.h>
+ #include <net/ip.h>
+ #include <net/netfilter/nf_queue.h>
+--- a/net/ipv4/raw.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/raw.c 2008-07-29 17:31:11.000000000 -0400
+@@ -119,7 +119,7 @@ static struct sock *__raw_v4_lookup(stru
+
+ if (net_eq(sock_net(sk), net) && inet->num == num &&
+ !(inet->daddr && inet->daddr != raddr) &&
+- !(inet->rcv_saddr && inet->rcv_saddr != laddr) &&
++ v4_sock_addr_match(sk->sk_nx_info, inet, laddr) &&
+ !(sk->sk_bound_dev_if && sk->sk_bound_dev_if != dif))
+ goto found; /* gotcha */
+ }
+@@ -373,6 +373,12 @@ static int raw_send_hdrinc(struct sock *
+ icmp_out_count(((struct icmphdr *)
+ skb_transport_header(skb))->type);
+
++ err = -EPERM;
++ if (!nx_check(0, VS_ADMIN) && !capable(CAP_NET_RAW) &&
++ sk->sk_nx_info &&
++ !v4_addr_in_nx_info(sk->sk_nx_info, iph->saddr, NXA_MASK_BIND))
++ goto error_free;
++
+ err = NF_HOOK(PF_INET, NF_INET_LOCAL_OUT, skb, NULL, rt->u.dst.dev,
+ dst_output);
+ if (err > 0)
+@@ -384,6 +390,7 @@ out:
+
+ error_fault:
+ err = -EFAULT;
++error_free:
+ kfree_skb(skb);
+ error:
+ IP_INC_STATS(IPSTATS_MIB_OUTDISCARDS);
+@@ -551,6 +558,13 @@ static int raw_sendmsg(struct kiocb *ioc
+ }
+
+ security_sk_classify_flow(sk, &fl);
++ if (sk->sk_nx_info) {
++ err = ip_v4_find_src(sock_net(sk),
++ sk->sk_nx_info, &rt, &fl);
++
++ if (err)
++ goto done;
++ }
+ err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 1);
+ }
+ if (err)
+@@ -621,17 +635,19 @@ static int raw_bind(struct sock *sk, str
+ {
+ struct inet_sock *inet = inet_sk(sk);
+ struct sockaddr_in *addr = (struct sockaddr_in *) uaddr;
++ struct nx_v4_sock_addr nsa = { 0 };
+ int ret = -EINVAL;
+ int chk_addr_ret;
+
+ if (sk->sk_state != TCP_CLOSE || addr_len < sizeof(struct sockaddr_in))
+ goto out;
+- chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr);
++ v4_map_sock_addr(inet, addr, &nsa);
++ chk_addr_ret = inet_addr_type(sock_net(sk), nsa.saddr);
+ ret = -EADDRNOTAVAIL;
+- if (addr->sin_addr.s_addr && chk_addr_ret != RTN_LOCAL &&
++ if (nsa.saddr && chk_addr_ret != RTN_LOCAL &&
+ chk_addr_ret != RTN_MULTICAST && chk_addr_ret != RTN_BROADCAST)
+ goto out;
+- inet->rcv_saddr = inet->saddr = addr->sin_addr.s_addr;
++ v4_set_sock_addr(inet, &nsa);
+ if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST)
+ inet->saddr = 0; /* Use device */
+ sk_dst_reset(sk);
+@@ -683,7 +699,8 @@ static int raw_recvmsg(struct kiocb *ioc
+ /* Copy the address. */
+ if (sin) {
+ sin->sin_family = AF_INET;
+- sin->sin_addr.s_addr = ip_hdr(skb)->saddr;
++ sin->sin_addr.s_addr =
++ nx_map_sock_lback(sk->sk_nx_info, ip_hdr(skb)->saddr);
+ sin->sin_port = 0;
+ memset(&sin->sin_zero, 0, sizeof(sin->sin_zero));
+ }
+@@ -860,7 +877,8 @@ static struct sock *raw_get_first(struct
+ struct hlist_node *node;
+
+ sk_for_each(sk, node, &state->h->ht[state->bucket])
+- if (sock_net(sk) == seq_file_net(seq))
++ if ((sock_net(sk) == seq_file_net(seq)) &&
++ nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
+ goto found;
+ }
+ sk = NULL;
+@@ -876,7 +894,8 @@ static struct sock *raw_get_next(struct
+ sk = sk_next(sk);
+ try_again:
+ ;
+- } while (sk && sock_net(sk) != seq_file_net(seq));
++ } while (sk && ((sock_net(sk) != seq_file_net(seq)) ||
++ !nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT)));
+
+ if (!sk && ++state->bucket < RAW_HTABLE_SIZE) {
+ sk = sk_head(&state->h->ht[state->bucket]);
+@@ -935,7 +954,10 @@ static void raw_sock_seq_show(struct seq
+
+ seq_printf(seq, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %d\n",
+- i, src, srcp, dest, destp, sp->sk_state,
++ i,
++ nx_map_sock_lback(current_nx_info(), src), srcp,
++ nx_map_sock_lback(current_nx_info(), dest), destp,
++ sp->sk_state,
+ atomic_read(&sp->sk_wmem_alloc),
+ atomic_read(&sp->sk_rmem_alloc),
+ 0, 0L, 0, sock_i_uid(sp), 0, sock_i_ino(sp),
+--- a/net/ipv4/syncookies.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/syncookies.c 2008-07-16 22:41:36.000000000 -0400
+@@ -18,6 +18,7 @@
+ #include <linux/cryptohash.h>
+ #include <linux/kernel.h>
+ #include <net/tcp.h>
++#include <net/route.h>
+
+ /* Timestamps: lowest 9 bits store TCP options */
+ #define TSBITS 9
+--- a/net/ipv4/tcp.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/tcp.c 2008-07-16 22:41:36.000000000 -0400
+@@ -266,6 +266,7 @@
+ #include <linux/cache.h>
+ #include <linux/err.h>
+ #include <linux/crypto.h>
++#include <linux/in.h>
+
+ #include <net/icmp.h>
+ #include <net/tcp.h>
+--- a/net/ipv4/tcp_ipv4.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/tcp_ipv4.c 2008-07-29 17:42:09.000000000 -0400
+@@ -1959,6 +1959,12 @@ static void *listening_get_next(struct s
+ req = req->dl_next;
+ while (1) {
+ while (req) {
++ vxdprintk(VXD_CBIT(net, 6),
++ "sk,req: %p [#%d] (from %d)", req->sk,
++ (req->sk)?req->sk->sk_nid:0, nx_current_nid());
++ if (req->sk &&
++ !nx_check(req->sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (req->rsk_ops->family == st->family &&
+ net_eq(sock_net(req->sk), net)) {
+ cur = req;
+@@ -1984,6 +1990,10 @@ get_req:
+ }
+ get_sk:
+ sk_for_each_from(sk, node) {
++ vxdprintk(VXD_CBIT(net, 6), "sk: %p [#%d] (from %d)",
++ sk, sk->sk_nid, nx_current_nid());
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (sk->sk_family == st->family && net_eq(sock_net(sk), net)) {
+ cur = sk;
+ goto out;
+@@ -2034,6 +2044,11 @@ static void *established_get_first(struc
+
+ read_lock_bh(lock);
+ sk_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
++ vxdprintk(VXD_CBIT(net, 6),
++ "sk,egf: %p [#%d] (from %d)",
++ sk, sk->sk_nid, nx_current_nid());
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (sk->sk_family != st->family ||
+ !net_eq(sock_net(sk), net)) {
+ continue;
+@@ -2044,6 +2059,11 @@ static void *established_get_first(struc
+ st->state = TCP_SEQ_STATE_TIME_WAIT;
+ inet_twsk_for_each(tw, node,
+ &tcp_hashinfo.ehash[st->bucket].twchain) {
++ vxdprintk(VXD_CBIT(net, 6),
++ "tw: %p [#%d] (from %d)",
++ tw, tw->tw_nid, nx_current_nid());
++ if (!nx_check(tw->tw_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (tw->tw_family != st->family ||
+ !net_eq(twsk_net(tw), net)) {
+ continue;
+@@ -2072,7 +2092,9 @@ static void *established_get_next(struct
+ tw = cur;
+ tw = tw_next(tw);
+ get_tw:
+- while (tw && (tw->tw_family != st->family || !net_eq(twsk_net(tw), net))) {
++ while (tw && (tw->tw_family != st->family ||
++ !net_eq(twsk_net(tw), net) ||
++ !nx_check(tw->tw_nid, VS_WATCH_P | VS_IDENT))) {
+ tw = tw_next(tw);
+ }
+ if (tw) {
+@@ -2093,6 +2115,11 @@ get_tw:
+ sk = sk_next(sk);
+
+ sk_for_each_from(sk, node) {
++ vxdprintk(VXD_CBIT(net, 6),
++ "sk,egn: %p [#%d] (from %d)",
++ sk, sk->sk_nid, nx_current_nid());
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (sk->sk_family == st->family && net_eq(sock_net(sk), net))
+ goto found;
+ }
+@@ -2247,9 +2274,9 @@ static void get_openreq4(struct sock *sk
+ seq_printf(f, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %u %d %p%n",
+ i,
+- ireq->loc_addr,
++ nx_map_sock_lback(current_nx_info(), ireq->loc_addr),
+ ntohs(inet_sk(sk)->sport),
+- ireq->rmt_addr,
++ nx_map_sock_lback(current_nx_info(), ireq->rmt_addr),
+ ntohs(ireq->rmt_port),
+ TCP_SYN_RECV,
+ 0, 0, /* could print option size, but that is af dependent. */
+@@ -2292,7 +2319,10 @@ static void get_tcp4_sock(struct sock *s
+
+ seq_printf(f, "%4d: %08X:%04X %08X:%04X %02X %08X:%08X %02X:%08lX "
+ "%08X %5d %8d %lu %d %p %lu %lu %u %u %d%n",
+- i, src, srcp, dest, destp, sk->sk_state,
++ i,
++ nx_map_sock_lback(current_nx_info(), src), srcp,
++ nx_map_sock_lback(current_nx_info(), dest), destp,
++ sk->sk_state,
+ tp->write_seq - tp->snd_una,
+ sk->sk_state == TCP_LISTEN ? sk->sk_ack_backlog :
+ (tp->rcv_nxt - tp->copied_seq),
+@@ -2328,7 +2358,10 @@ static void get_timewait4_sock(struct in
+
+ seq_printf(f, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %p%n",
+- i, src, srcp, dest, destp, tw->tw_substate, 0, 0,
++ i,
++ nx_map_sock_lback(current_nx_info(), src), srcp,
++ nx_map_sock_lback(current_nx_info(), dest), destp,
++ tw->tw_substate, 0, 0,
+ 3, jiffies_to_clock_t(ttd), 0, 0, 0, 0,
+ atomic_read(&tw->tw_refcnt), tw, len);
+ }
+--- a/net/ipv4/tcp_minisocks.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv4/tcp_minisocks.c 2008-07-29 17:27:07.000000000 -0400
+@@ -28,6 +28,10 @@
+ #include <net/inet_common.h>
+ #include <net/xfrm.h>
+
++#include <linux/vs_limit.h>
++#include <linux/vs_socket.h>
++#include <linux/vs_context.h>
++
+ #ifdef CONFIG_SYSCTL
+ #define SYNC_INIT 0 /* let the user enable it */
+ #else
+@@ -295,6 +299,11 @@ void tcp_time_wait(struct sock *sk, int
+ tcptw->tw_ts_recent = tp->rx_opt.ts_recent;
+ tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp;
+
++ tw->tw_xid = sk->sk_xid;
++ tw->tw_vx_info = NULL;
++ tw->tw_nid = sk->sk_nid;
++ tw->tw_nx_info = NULL;
++
+ #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+ if (tw->tw_family == PF_INET6) {
+ struct ipv6_pinfo *np = inet6_sk(sk);
+--- a/net/ipv4/udp.c 2008-08-12 01:41:51.000000000 -0400
++++ a/net/ipv4/udp.c 2008-08-12 01:42:21.000000000 -0400
+@@ -239,20 +239,14 @@ fail:
+ return error;
+ }
+
+-static int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
+-{
+- struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
+-
+- return ( !ipv6_only_sock(sk2) &&
+- (!inet1->rcv_saddr || !inet2->rcv_saddr ||
+- inet1->rcv_saddr == inet2->rcv_saddr ));
+-}
++extern int ipv4_rcv_saddr_equal(const struct sock *, const struct sock *);
+
+ int udp_v4_get_port(struct sock *sk, unsigned short snum)
+ {
+ return udp_lib_get_port(sk, snum, ipv4_rcv_saddr_equal);
+ }
+
++
+ /* UDP is nearly always wildcards out the wazoo, it makes no sense to try
+ * harder than this. -DaveM
+ */
+@@ -272,10 +266,16 @@ static struct sock *__udp4_lib_lookup(st
+ if (net_eq(sock_net(sk), net) && sk->sk_hash == hnum &&
+ !ipv6_only_sock(sk)) {
+ int score = (sk->sk_family == PF_INET ? 1 : 0);
++
+ if (inet->rcv_saddr) {
+ if (inet->rcv_saddr != daddr)
+ continue;
+ score+=2;
++ } else {
++ /* block non nx_info ips */
++ if (!v4_addr_in_nx_info(sk->sk_nx_info,
++ daddr, NXA_MASK_BIND))
++ continue;
+ }
+ if (inet->daddr) {
+ if (inet->daddr != saddr)
+@@ -301,6 +301,7 @@ static struct sock *__udp4_lib_lookup(st
+ }
+ }
+ }
++
+ if (result)
+ sock_hold(result);
+ read_unlock(&udp_hash_lock);
+@@ -322,7 +323,7 @@ static inline struct sock *udp_v4_mcast_
+ if (s->sk_hash != hnum ||
+ (inet->daddr && inet->daddr != rmt_addr) ||
+ (inet->dport != rmt_port && inet->dport) ||
+- (inet->rcv_saddr && inet->rcv_saddr != loc_addr) ||
++ !v4_sock_addr_match(sk->sk_nx_info, inet, loc_addr) ||
+ ipv6_only_sock(s) ||
+ (s->sk_bound_dev_if && s->sk_bound_dev_if != dif))
+ continue;
+@@ -656,7 +657,15 @@ int udp_sendmsg(struct kiocb *iocb, stru
+ .uli_u = { .ports =
+ { .sport = inet->sport,
+ .dport = dport } } };
++ struct nx_info *nxi = sk->sk_nx_info;
++ struct net *net = sock_net(sk);
++
+ security_sk_classify_flow(sk, &fl);
++
++ err = ip_v4_find_src(net, nxi, &rt, &fl);
++ if (err)
++ goto out;
++
+ err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 1);
+ if (err) {
+ if (err == -ENETUNREACH)
+@@ -899,7 +908,8 @@ try_again:
+ {
+ sin->sin_family = AF_INET;
+ sin->sin_port = udp_hdr(skb)->source;
+- sin->sin_addr.s_addr = ip_hdr(skb)->saddr;
++ sin->sin_addr.s_addr = nx_map_sock_lback(
++ skb->sk->sk_nx_info, ip_hdr(skb)->saddr);
+ memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
+ }
+ if (inet->cmsg_flags)
+@@ -1515,6 +1525,8 @@ static struct sock *udp_get_first(struct
+ sk_for_each(sk, node, state->hashtable + state->bucket) {
+ if (!net_eq(sock_net(sk), net))
+ continue;
++ if (!nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (sk->sk_family == state->family)
+ goto found;
+ }
+@@ -1533,7 +1545,9 @@ static struct sock *udp_get_next(struct
+ sk = sk_next(sk);
+ try_again:
+ ;
+- } while (sk && (!net_eq(sock_net(sk), net) || sk->sk_family != state->family));
++ } while (sk && (!net_eq(sock_net(sk), net) ||
++ sk->sk_family != state->family ||
++ !nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT)));
+
+ if (!sk && ++state->bucket < UDP_HTABLE_SIZE) {
+ sk = sk_head(state->hashtable + state->bucket);
+@@ -1634,7 +1648,10 @@ static void udp4_format_sock(struct sock
+
+ seq_printf(f, "%4d: %08X:%04X %08X:%04X"
+ " %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p%n",
+- bucket, src, srcp, dest, destp, sp->sk_state,
++ bucket,
++ nx_map_sock_lback(current_nx_info(), src), srcp,
++ nx_map_sock_lback(current_nx_info(), dest), destp,
++ sp->sk_state,
+ atomic_read(&sp->sk_wmem_alloc),
+ atomic_read(&sp->sk_rmem_alloc),
+ 0, 0L, 0, sock_i_uid(sp), 0, sock_i_ino(sp),
+--- a/net/ipv6/addrconf.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/addrconf.c 2008-07-29 18:55:29.000000000 -0400
+@@ -87,6 +87,8 @@
+
+ #include <linux/proc_fs.h>
+ #include <linux/seq_file.h>
++#include <linux/vs_network.h>
++#include <linux/vs_inet6.h>
+
+ /* Set to 3 to get tracing... */
+ #define ACONF_DEBUG 2
+@@ -1078,7 +1080,7 @@ out:
+
+ int ipv6_dev_get_saddr(struct net *net, struct net_device *dst_dev,
+ const struct in6_addr *daddr, unsigned int prefs,
+- struct in6_addr *saddr)
++ struct in6_addr *saddr, struct nx_info *nxi)
+ {
+ struct ipv6_saddr_score scores[2],
+ *score = &scores[0], *hiscore = &scores[1];
+@@ -1335,35 +1337,46 @@ struct inet6_ifaddr *ipv6_get_ifaddr(str
+ return ifp;
+ }
+
++extern int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2);
++
+ int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2)
+ {
+ const struct in6_addr *sk_rcv_saddr6 = &inet6_sk(sk)->rcv_saddr;
+ const struct in6_addr *sk2_rcv_saddr6 = inet6_rcv_saddr(sk2);
+- __be32 sk_rcv_saddr = inet_sk(sk)->rcv_saddr;
+ __be32 sk2_rcv_saddr = inet_rcv_saddr(sk2);
+ int sk_ipv6only = ipv6_only_sock(sk);
+ int sk2_ipv6only = inet_v6_ipv6only(sk2);
+ int addr_type = ipv6_addr_type(sk_rcv_saddr6);
+ int addr_type2 = sk2_rcv_saddr6 ? ipv6_addr_type(sk2_rcv_saddr6) : IPV6_ADDR_MAPPED;
+
+- if (!sk2_rcv_saddr && !sk_ipv6only)
++ /* FIXME: needs handling for v4 ANY */
++ if (!sk2_rcv_saddr && !sk_ipv6only && !sk2->sk_nx_info)
+ return 1;
+
+ if (addr_type2 == IPV6_ADDR_ANY &&
+- !(sk2_ipv6only && addr_type == IPV6_ADDR_MAPPED))
++ !(sk2_ipv6only && addr_type == IPV6_ADDR_MAPPED) &&
++ v6_addr_in_nx_info(sk2->sk_nx_info, sk_rcv_saddr6, -1))
+ return 1;
+
+ if (addr_type == IPV6_ADDR_ANY &&
+- !(sk_ipv6only && addr_type2 == IPV6_ADDR_MAPPED))
++ !(sk_ipv6only && addr_type2 == IPV6_ADDR_MAPPED) &&
++ (sk2_rcv_saddr6 && v6_addr_in_nx_info(sk->sk_nx_info, sk2_rcv_saddr6, -1)))
++ return 1;
++
++ if (addr_type == IPV6_ADDR_ANY &&
++ addr_type2 == IPV6_ADDR_ANY &&
++ nx_v6_addr_conflict(sk->sk_nx_info, sk2->sk_nx_info))
+ return 1;
+
+ if (sk2_rcv_saddr6 &&
++ addr_type != IPV6_ADDR_ANY &&
++ addr_type != IPV6_ADDR_ANY &&
+ ipv6_addr_equal(sk_rcv_saddr6, sk2_rcv_saddr6))
+ return 1;
+
+ if (addr_type == IPV6_ADDR_MAPPED &&
+ !sk2_ipv6only &&
+- (!sk2_rcv_saddr || !sk_rcv_saddr || sk_rcv_saddr == sk2_rcv_saddr))
++ ipv4_rcv_saddr_equal(sk, sk2))
+ return 1;
+
+ return 0;
+@@ -2935,7 +2948,10 @@ static void if6_seq_stop(struct seq_file
+ static int if6_seq_show(struct seq_file *seq, void *v)
+ {
+ struct inet6_ifaddr *ifp = (struct inet6_ifaddr *)v;
+- seq_printf(seq,
++
++ if (nx_check(0, VS_ADMIN|VS_WATCH) ||
++ v6_addr_in_nx_info(current_nx_info(), &ifp->addr, -1))
++ seq_printf(seq,
+ NIP6_SEQFMT " %02x %02x %02x %02x %8s\n",
+ NIP6(ifp->addr),
+ ifp->idev->dev->ifindex,
+@@ -3430,6 +3446,12 @@ static int inet6_dump_addr(struct sk_buf
+ struct ifmcaddr6 *ifmca;
+ struct ifacaddr6 *ifaca;
+ struct net *net = sock_net(skb->sk);
++ struct nx_info *nxi = skb->sk ? skb->sk->sk_nx_info : NULL;
++
++ /* disable ipv6 on non v6 guests */
++ if (nxi && !nx_info_has_v6(nxi))
++ return skb->len;
++
+
+ s_idx = cb->args[0];
+ s_ip_idx = ip_idx = cb->args[1];
+@@ -3451,6 +3473,8 @@ static int inet6_dump_addr(struct sk_buf
+ ifa = ifa->if_next, ip_idx++) {
+ if (ip_idx < s_ip_idx)
+ continue;
++ if (!v6_addr_in_nx_info(nxi, &ifa->addr, -1))
++ continue;
+ err = inet6_fill_ifaddr(skb, ifa,
+ NETLINK_CB(cb->skb).pid,
+ cb->nlh->nlmsg_seq,
+@@ -3464,6 +3488,8 @@ static int inet6_dump_addr(struct sk_buf
+ ifmca = ifmca->next, ip_idx++) {
+ if (ip_idx < s_ip_idx)
+ continue;
++ if (!v6_addr_in_nx_info(nxi, &ifmca->mca_addr, -1))
++ continue;
+ err = inet6_fill_ifmcaddr(skb, ifmca,
+ NETLINK_CB(cb->skb).pid,
+ cb->nlh->nlmsg_seq,
+@@ -3477,6 +3503,8 @@ static int inet6_dump_addr(struct sk_buf
+ ifaca = ifaca->aca_next, ip_idx++) {
+ if (ip_idx < s_ip_idx)
+ continue;
++ if (!v6_addr_in_nx_info(nxi, &ifaca->aca_addr, -1))
++ continue;
+ err = inet6_fill_ifacaddr(skb, ifaca,
+ NETLINK_CB(cb->skb).pid,
+ cb->nlh->nlmsg_seq,
+@@ -3760,12 +3788,19 @@ static int inet6_dump_ifinfo(struct sk_b
+ int s_idx = cb->args[0];
+ struct net_device *dev;
+ struct inet6_dev *idev;
++ struct nx_info *nxi = skb->sk ? skb->sk->sk_nx_info : NULL;
++
++ /* FIXME: maybe disable ipv6 on non v6 guests?
++ if (skb->sk && skb->sk->sk_vx_info)
++ return skb->len; */
+
+ read_lock(&dev_base_lock);
+ idx = 0;
+ for_each_netdev(net, dev) {
+ if (idx < s_idx)
+ goto cont;
++ if (!v6_dev_in_nx_info(dev, nxi))
++ goto cont;
+ if ((idev = in6_dev_get(dev)) == NULL)
+ goto cont;
+ err = inet6_fill_ifinfo(skb, idev, NETLINK_CB(cb->skb).pid,
+--- a/net/ipv6/af_inet6.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/af_inet6.c 2008-07-29 18:22:35.000000000 -0400
+@@ -43,6 +43,8 @@
+ #include <linux/netdevice.h>
+ #include <linux/icmpv6.h>
+ #include <linux/netfilter_ipv6.h>
++#include <linux/vs_inet.h>
++#include <linux/vs_inet6.h>
+
+ #include <net/ip.h>
+ #include <net/ipv6.h>
+@@ -51,6 +53,7 @@
+ #include <net/tcp.h>
+ #include <net/ipip.h>
+ #include <net/protocol.h>
++#include <net/route.h>
+ #include <net/inet_common.h>
+ #include <net/transp_v6.h>
+ #include <net/ip6_route.h>
+@@ -148,9 +151,12 @@ lookup_protocol:
+ }
+
+ err = -EPERM;
++ if ((protocol == IPPROTO_ICMPV6) &&
++ nx_capable(answer->capability, NXC_RAW_ICMP))
++ goto override;
+ if (answer->capability > 0 && !capable(answer->capability))
+ goto out_rcu_unlock;
+-
++override:
+ sock->ops = answer->ops;
+ answer_prot = answer->prot;
+ answer_no_check = answer->no_check;
+@@ -249,6 +255,7 @@ int inet6_bind(struct socket *sock, stru
+ struct inet_sock *inet = inet_sk(sk);
+ struct ipv6_pinfo *np = inet6_sk(sk);
+ struct net *net = sock_net(sk);
++ struct nx_v6_sock_addr nsa;
+ __be32 v4addr = 0;
+ unsigned short snum;
+ int addr_type = 0;
+@@ -260,6 +267,11 @@ int inet6_bind(struct socket *sock, stru
+
+ if (addr_len < SIN6_LEN_RFC2133)
+ return -EINVAL;
++
++ err = v6_map_sock_addr(inet, addr, &nsa);
++ if (err)
++ return err;
++
+ addr_type = ipv6_addr_type(&addr->sin6_addr);
+ if ((addr_type & IPV6_ADDR_MULTICAST) && sock->type == SOCK_STREAM)
+ return -EINVAL;
+@@ -283,6 +295,10 @@ int inet6_bind(struct socket *sock, stru
+ err = -EADDRNOTAVAIL;
+ goto out;
+ }
++ if (!v4_addr_in_nx_info(sk->sk_nx_info, v4addr, NXA_MASK_BIND)) {
++ err = -EADDRNOTAVAIL;
++ goto out;
++ }
+ } else {
+ if (addr_type != IPV6_ADDR_ANY) {
+ struct net_device *dev = NULL;
+@@ -308,6 +324,11 @@ int inet6_bind(struct socket *sock, stru
+ }
+ }
+
++ if (!v6_addr_in_nx_info(sk->sk_nx_info, &addr->sin6_addr, -1)) {
++ err = -EADDRNOTAVAIL;
++ goto out;
++ }
++
+ /* ipv4 addr of the socket is invalid. Only the
+ * unspecified and mapped address have a v4 equivalent.
+ */
+@@ -326,6 +347,8 @@ int inet6_bind(struct socket *sock, stru
+ }
+ }
+
++ v6_set_sock_addr(inet, &nsa);
++
+ inet->rcv_saddr = v4addr;
+ inet->saddr = v4addr;
+
+@@ -420,9 +443,11 @@ int inet6_getname(struct socket *sock, s
+ return -ENOTCONN;
+ sin->sin6_port = inet->dport;
+ ipv6_addr_copy(&sin->sin6_addr, &np->daddr);
++ /* FIXME: remap lback? */
+ if (np->sndflow)
+ sin->sin6_flowinfo = np->flow_label;
+ } else {
++ /* FIXME: remap lback? */
+ if (ipv6_addr_any(&np->rcv_saddr))
+ ipv6_addr_copy(&sin->sin6_addr, &np->saddr);
+ else
+--- a/net/ipv6/fib6_rules.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/fib6_rules.c 2008-07-29 18:23:01.000000000 -0400
+@@ -95,7 +95,7 @@ static int fib6_rule_action(struct fib_r
+ if (ipv6_dev_get_saddr(net
+ ip6_dst_idev(&rt->u.dst)->dev,
+ &flp->fl6_dst, srcprefs,
+- &saddr))
++ &saddr, NULL))
+ goto again;
+ if (!ipv6_prefix_equal(&saddr, &r->src.addr,
+ r->src.plen))
+--- a/net/ipv6/inet6_hashtables.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/inet6_hashtables.c 2008-07-29 18:12:38.000000000 -0400
+@@ -16,6 +16,7 @@
+
+ #include <linux/module.h>
+ #include <linux/random.h>
++#include <linux/vs_inet6.h>
+
+ #include <net/inet_connection_sock.h>
+ #include <net/inet_hashtables.h>
+@@ -114,6 +115,9 @@ struct sock *inet6_lookup_listener(struc
+ if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
+ continue;
+ score++;
++ } else {
++ if (!v6_addr_in_nx_info(sk->sk_nx_info, daddr, -1))
++ continue;
+ }
+ if (sk->sk_bound_dev_if) {
+ if (sk->sk_bound_dev_if != dif)
+--- a/net/ipv6/ip6_output.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/ip6_output.c 2008-07-29 18:25:00.000000000 -0400
+@@ -924,7 +924,7 @@ static int ip6_dst_lookup_tail(struct so
+ err = ipv6_dev_get_saddr(ip6_dst_idev(*dst)->dev,
+ &fl->fl6_dst,
+ sk ? inet6_sk(sk)->srcprefs : 0,
+- &fl->fl6_src);
++ &fl->fl6_src, sk->sk_nx_info);
+ if (err)
+ goto out_err_release;
+ }
+--- a/net/ipv6/Kconfig 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/Kconfig 2008-07-29 18:21:01.000000000 -0400
+@@ -4,8 +4,8 @@
+
+ # IPv6 as module will cause a CRASH if you try to unload it
+ menuconfig IPV6
+- tristate "The IPv6 protocol"
+- default m
++ bool "The IPv6 protocol"
++ default n
+ ---help---
+ This is complemental support for the IP version 6.
+ You will still be able to do traditional IPv4 networking as well.
+--- a/net/ipv6/ndisc.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/ndisc.c 2008-07-29 18:19:07.000000000 -0400
+@@ -551,7 +551,7 @@ static void ndisc_send_na(struct net_dev
+ } else {
+ if (ipv6_dev_get_saddr(dev_net(dev), dev, daddr,
+ inet6_sk(dev_net(dev)->ipv6.ndisc_sk)->srcprefs,
+- &tmpaddr))
++ &tmpaddr, NULL /* FIXME: ? */ ))
+ return;
+ src_addr = &tmpaddr;
+ }
+--- a/net/ipv6/route.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/route.c 2008-07-29 18:56:21.000000000 -0400
+@@ -2181,7 +2181,8 @@ static int rt6_fill_node(struct sk_buff
+ struct inet6_dev *idev = ip6_dst_idev(&rt->u.dst);
+ struct in6_addr saddr_buf;
+ if (ipv6_dev_get_saddr(net, idev ? idev->dev : NULL,
+- dst, 0, &saddr_buf) == 0)
++ dst, 0, &saddr_buf,
++ (skb->sk ? skb->sk->sk_nx_info : NULL)) == 0)
+ NLA_PUT(skb, RTA_PREFSRC, 16, &saddr_buf);
+ }
+
+--- a/net/ipv6/tcp_ipv6.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/tcp_ipv6.c 2008-07-29 18:12:38.000000000 -0400
+@@ -69,6 +69,7 @@
+
+ #include <linux/crypto.h>
+ #include <linux/scatterlist.h>
++#include <linux/vs_inet6.h>
+
+ static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb);
+ static void tcp_v6_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req);
+@@ -152,8 +153,15 @@ static int tcp_v6_connect(struct sock *s
+ * connect() to INADDR_ANY means loopback (BSD'ism).
+ */
+
+- if(ipv6_addr_any(&usin->sin6_addr))
+- usin->sin6_addr.s6_addr[15] = 0x1;
++ if(ipv6_addr_any(&usin->sin6_addr)) {
++ struct nx_info *nxi = sk->sk_nx_info;
++
++ if (nxi && nx_info_has_v6(nxi))
++ /* FIXME: remap lback? */
++ usin->sin6_addr = nxi->v6.ip;
++ else
++ usin->sin6_addr.s6_addr[15] = 0x1;
++ }
+
+ addr_type = ipv6_addr_type(&usin->sin6_addr);
+
+--- a/net/ipv6/udp.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/udp.c 2008-07-29 18:12:38.000000000 -0400
+@@ -49,6 +49,7 @@
+
+ #include <linux/proc_fs.h>
+ #include <linux/seq_file.h>
++#include <linux/vs_inet6.h>
+ #include "udp_impl.h"
+
+ int udp_v6_get_port(struct sock *sk, unsigned short snum)
+@@ -83,6 +84,10 @@ static struct sock *__udp6_lib_lookup(st
+ if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
+ continue;
+ score++;
++ } else {
++ /* block non nx_info ips */
++ if (!v6_addr_in_nx_info(sk->sk_nx_info, daddr, -1))
++ continue;
+ }
+ if (!ipv6_addr_any(&np->daddr)) {
+ if (!ipv6_addr_equal(&np->daddr, saddr))
+--- a/net/ipv6/xfrm6_policy.c 2008-07-14 17:22:58.000000000 -0400
++++ a/net/ipv6/xfrm6_policy.c 2008-07-29 18:21:53.000000000 -0400
+@@ -59,7 +59,7 @@ static int xfrm6_get_saddr(xfrm_address_
+ dev = ip6_dst_idev(dst)->dev;
+ ipv6_dev_get_saddr(dev_net(dev), dev,
+ (struct in6_addr *)&daddr->a6, 0,
+- (struct in6_addr *)&saddr->a6);
++ (struct in6_addr *)&saddr->a6, NULL);
+ dst_release(dst);
+ return 0;
+ }
+--- a/net/netlink/af_netlink.c 2008-07-14 17:23:03.000000000 -0400
++++ a/net/netlink/af_netlink.c 2008-07-16 22:51:45.000000000 -0400
+@@ -55,6 +55,9 @@
+ #include <linux/types.h>
+ #include <linux/audit.h>
+ #include <linux/mutex.h>
++#include <linux/vs_context.h>
++#include <linux/vs_network.h>
++#include <linux/vs_limit.h>
+
+ #include <net/net_namespace.h>
+ #include <net/sock.h>
+@@ -1760,6 +1763,8 @@ static struct sock *netlink_seq_socket_i
+ sk_for_each(s, node, &hash->table[j]) {
+ if (sock_net(s) != seq_file_net(seq))
+ continue;
++ if (!nx_check(s->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (off == pos) {
+ iter->link = i;
+ iter->hash_idx = j;
+@@ -1794,7 +1799,8 @@ static void *netlink_seq_next(struct seq
+ s = v;
+ do {
+ s = sk_next(s);
+- } while (s && sock_net(s) != seq_file_net(seq));
++ } while (s && (sock_net(s) != seq_file_net(seq) ||
++ !nx_check(s->sk_nid, VS_WATCH_P | VS_IDENT)));
+ if (s)
+ return s;
+
+@@ -1806,7 +1812,8 @@ static void *netlink_seq_next(struct seq
+
+ for (; j <= hash->mask; j++) {
+ s = sk_head(&hash->table[j]);
+- while (s && sock_net(s) != seq_file_net(seq))
++ while (s && (sock_net(s) != seq_file_net(seq) ||
++ !nx_check(s->sk_nid, VS_WATCH_P | VS_IDENT)))
+ s = sk_next(s);
+ if (s) {
+ iter->link = i;
+--- a/net/sctp/ipv6.c 2008-07-14 17:23:04.000000000 -0400
++++ a/net/sctp/ipv6.c 2008-07-29 18:16:30.000000000 -0400
+@@ -320,7 +320,8 @@ static void sctp_v6_get_saddr(struct sct
+ ipv6_dev_get_saddr(dst ? ip6_dst_idev(dst)->dev : NULL,
+ &daddr->v6.sin6_addr,
+ inet6_sk(&sk->inet.sk)->srcprefs,
+- &saddr->v6.sin6_addr);
++ &saddr->v6.sin6_addr,
++ asoc->base.sk->sk_nx_info);
+ SCTP_DEBUG_PRINTK("saddr from ipv6_get_saddr: " NIP6_FMT "\n",
+ NIP6(saddr->v6.sin6_addr));
+ return;
+--- a/net/socket.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/socket.c 2008-07-29 18:12:38.000000000 -0400
+@@ -93,6 +93,10 @@
+
+ #include <net/sock.h>
+ #include <linux/netfilter.h>
++#include <linux/vs_base.h>
++#include <linux/vs_socket.h>
++#include <linux/vs_inet.h>
++#include <linux/vs_inet6.h>
+
+ static int sock_no_open(struct inode *irrelevant, struct file *dontcare);
+ static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
+@@ -547,7 +551,7 @@ static inline int __sock_sendmsg(struct
+ struct msghdr *msg, size_t size)
+ {
+ struct sock_iocb *si = kiocb_to_siocb(iocb);
+- int err;
++ int err, len;
+
+ si->sock = sock;
+ si->scm = NULL;
+@@ -558,7 +562,22 @@ static inline int __sock_sendmsg(struct
+ if (err)
+ return err;
+
+- return sock->ops->sendmsg(iocb, sock, msg, size);
++ len = sock->ops->sendmsg(iocb, sock, msg, size);
++ if (sock->sk) {
++ if (len == size)
++ vx_sock_send(sock->sk, size);
++ else
++ vx_sock_fail(sock->sk, size);
++ }
++ vxdprintk(VXD_CBIT(net, 7),
++ "__sock_sendmsg: %p[%p,%p,%p;%d/%d]:%d/%d",
++ sock, sock->sk,
++ (sock->sk)?sock->sk->sk_nx_info:0,
++ (sock->sk)?sock->sk->sk_vx_info:0,
++ (sock->sk)?sock->sk->sk_xid:0,
++ (sock->sk)?sock->sk->sk_nid:0,
++ (unsigned int)size, len);
++ return len;
+ }
+
+ int sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
+@@ -627,7 +646,7 @@ EXPORT_SYMBOL_GPL(__sock_recv_timestamp)
+ static inline int __sock_recvmsg(struct kiocb *iocb, struct socket *sock,
+ struct msghdr *msg, size_t size, int flags)
+ {
+- int err;
++ int err, len;
+ struct sock_iocb *si = kiocb_to_siocb(iocb);
+
+ si->sock = sock;
+@@ -640,7 +659,18 @@ static inline int __sock_recvmsg(struct
+ if (err)
+ return err;
+
+- return sock->ops->recvmsg(iocb, sock, msg, size, flags);
++ len = sock->ops->recvmsg(iocb, sock, msg, size, flags);
++ if ((len >= 0) && sock->sk)
++ vx_sock_recv(sock->sk, len);
++ vxdprintk(VXD_CBIT(net, 7),
++ "__sock_recvmsg: %p[%p,%p,%p;%d/%d]:%d/%d",
++ sock, sock->sk,
++ (sock->sk)?sock->sk->sk_nx_info:0,
++ (sock->sk)?sock->sk->sk_vx_info:0,
++ (sock->sk)?sock->sk->sk_xid:0,
++ (sock->sk)?sock->sk->sk_nid:0,
++ (unsigned int)size, len);
++ return len;
+ }
+
+ int sock_recvmsg(struct socket *sock, struct msghdr *msg,
+@@ -1105,6 +1135,13 @@ static int __sock_create(struct net *net
+ if (type < 0 || type >= SOCK_MAX)
+ return -EINVAL;
+
++ if (!nx_check(0, VS_ADMIN)) {
++ if (family == PF_INET && !current_nx_info_has_v4())
++ return -EAFNOSUPPORT;
++ if (family == PF_INET6 && !current_nx_info_has_v6())
++ return -EAFNOSUPPORT;
++ }
++
+ /* Compatibility.
+
+ This uglymoron is moved from INET layer to here to avoid
+@@ -1222,6 +1259,7 @@ asmlinkage long sys_socket(int family, i
+ if (retval < 0)
+ goto out;
+
++ set_bit(SOCK_USER_SOCKET, &sock->flags);
+ retval = sock_map_fd(sock);
+ if (retval < 0)
+ goto out_release;
+@@ -1254,10 +1292,12 @@ asmlinkage long sys_socketpair(int famil
+ err = sock_create(family, type, protocol, &sock1);
+ if (err < 0)
+ goto out;
++ set_bit(SOCK_USER_SOCKET, &sock1->flags);
+
+ err = sock_create(family, type, protocol, &sock2);
+ if (err < 0)
+ goto out_release_1;
++ set_bit(SOCK_USER_SOCKET, &sock2->flags);
+
+ err = sock1->ops->socketpair(sock1, sock2);
+ if (err < 0)
+--- a/net/sunrpc/auth.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/sunrpc/auth.c 2008-07-16 22:41:36.000000000 -0400
+@@ -14,6 +14,7 @@
+ #include <linux/hash.h>
+ #include <linux/sunrpc/clnt.h>
+ #include <linux/spinlock.h>
++#include <linux/vs_tag.h>
+
+ #ifdef RPC_DEBUG
+ # define RPCDBG_FACILITY RPCDBG_AUTH
+@@ -353,6 +354,8 @@ rpcauth_lookupcred(struct rpc_auth *auth
+ struct auth_cred acred = {
+ .uid = current->fsuid,
+ .gid = current->fsgid,
++ .tag = dx_current_tag(),
++ .tag = dx_current_tag(),
+ .group_info = current->group_info,
+ };
+ struct rpc_cred *ret;
+--- a/net/sunrpc/auth_unix.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/sunrpc/auth_unix.c 2008-07-17 16:38:01.000000000 -0400
+@@ -11,12 +11,14 @@
+ #include <linux/module.h>
+ #include <linux/sunrpc/clnt.h>
+ #include <linux/sunrpc/auth.h>
++#include <linux/vs_tag.h>
+
+ #define NFS_NGROUPS 16
+
+ struct unx_cred {
+ struct rpc_cred uc_base;
+ gid_t uc_gid;
++ tag_t uc_tag;
+ gid_t uc_gids[NFS_NGROUPS];
+ };
+ #define uc_uid uc_base.cr_uid
+@@ -78,6 +80,7 @@ unx_create_cred(struct rpc_auth *auth, s
+ groups = NFS_NGROUPS;
+
+ cred->uc_gid = acred->gid;
++ cred->uc_tag = acred->tag;
+ for (i = 0; i < groups; i++)
+ cred->uc_gids[i] = GROUP_AT(acred->group_info, i);
+ if (i < NFS_NGROUPS)
+@@ -119,7 +122,9 @@ unx_match(struct auth_cred *acred, struc
+ unsigned int i;
+
+
+- if (cred->uc_uid != acred->uid || cred->uc_gid != acred->gid)
++ if (cred->uc_uid != acred->uid ||
++ cred->uc_gid != acred->gid ||
++ cred->uc_tag != acred->tag)
+ return 0;
+
+ if (acred->group_info != NULL)
+@@ -142,7 +147,7 @@ unx_marshal(struct rpc_task *task, __be3
+ struct rpc_clnt *clnt = task->tk_client;
+ struct unx_cred *cred = container_of(task->tk_msg.rpc_cred, struct unx_cred, uc_base);
+ __be32 *base, *hold;
+- int i;
++ int i, tag;
+
+ *p++ = htonl(RPC_AUTH_UNIX);
+ base = p++;
+@@ -152,9 +157,12 @@ unx_marshal(struct rpc_task *task, __be3
+ * Copy the UTS nodename captured when the client was created.
+ */
+ p = xdr_encode_array(p, clnt->cl_nodename, clnt->cl_nodelen);
++ tag = task->tk_client->cl_tag;
+
+- *p++ = htonl((u32) cred->uc_uid);
+- *p++ = htonl((u32) cred->uc_gid);
++ *p++ = htonl((u32) TAGINO_UID(tag,
++ cred->uc_uid, cred->uc_tag));
++ *p++ = htonl((u32) TAGINO_GID(tag,
++ cred->uc_gid, cred->uc_tag));
+ hold = p++;
+ for (i = 0; i < 16 && cred->uc_gids[i] != (gid_t) NOGROUP; i++)
+ *p++ = htonl((u32) cred->uc_gids[i]);
+--- a/net/sunrpc/clnt.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/sunrpc/clnt.c 2008-07-16 22:41:36.000000000 -0400
+@@ -31,6 +31,7 @@
+ #include <linux/utsname.h>
+ #include <linux/workqueue.h>
+ #include <linux/in6.h>
++#include <linux/vs_cvirt.h>
+
+ #include <linux/sunrpc/clnt.h>
+ #include <linux/sunrpc/rpc_pipe_fs.h>
+@@ -324,7 +325,9 @@ struct rpc_clnt *rpc_create(struct rpc_c
+ clnt->cl_autobind = 1;
+ if (args->flags & RPC_CLNT_CREATE_DISCRTRY)
+ clnt->cl_discrtry = 1;
+-
++ /* TODO: handle RPC_CLNT_CREATE_TAGGED
++ if (args->flags & RPC_CLNT_CREATE_TAGGED)
++ clnt->cl_tag = 1; */
+ return clnt;
+ }
+ EXPORT_SYMBOL_GPL(rpc_create);
+--- a/net/unix/af_unix.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/unix/af_unix.c 2008-07-16 22:52:51.000000000 -0400
+@@ -116,6 +116,8 @@
+ #include <linux/mount.h>
+ #include <net/checksum.h>
+ #include <linux/security.h>
++#include <linux/vs_context.h>
++#include <linux/vs_limit.h>
+
+ static struct hlist_head unix_socket_table[UNIX_HASH_SIZE + 1];
+ static DEFINE_SPINLOCK(unix_table_lock);
+@@ -260,6 +262,8 @@ static struct sock *__unix_find_socket_b
+ if (!net_eq(sock_net(s), net))
+ continue;
+
++ if (!nx_check(s->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (u->addr->len == len &&
+ !memcmp(u->addr->name, sunname, len))
+ goto found;
+@@ -2086,6 +2090,8 @@ static struct sock *unix_seq_idx(struct
+ for (s = first_unix_socket(&iter->i); s; s = next_unix_socket(&iter->i, s)) {
+ if (sock_net(s) != seq_file_net(seq))
+ continue;
++ if (!nx_check(s->sk_nid, VS_WATCH_P | VS_IDENT))
++ continue;
+ if (off == pos)
+ return s;
+ ++off;
+@@ -2111,7 +2117,8 @@ static void *unix_seq_next(struct seq_fi
+ sk = first_unix_socket(&iter->i);
+ else
+ sk = next_unix_socket(&iter->i, sk);
+- while (sk && (sock_net(sk) != seq_file_net(seq)))
++ while (sk && (sock_net(sk) != seq_file_net(seq) ||
++ !nx_check(sk->sk_nid, VS_WATCH_P | VS_IDENT)))
+ sk = next_unix_socket(&iter->i, sk);
+ return sk;
+ }
+--- a/net/x25/af_x25.c 2008-07-14 17:23:05.000000000 -0400
++++ a/net/x25/af_x25.c 2008-07-16 22:41:36.000000000 -0400
+@@ -506,7 +506,10 @@ static int x25_create(struct net *net, s
+
+ x25 = x25_sk(sk);
+
+- sock_init_data(sock, sk);
++ sk->sk_socket = sock;
++ sk->sk_type = sock->type;
++ sk->sk_sleep = &sock->wait;
++ sock->sk = sk;
+
+ x25_init_timers(sk);
+
+--- a/scripts/checksyscalls.sh 2008-04-17 11:31:42.000000000 -0400
++++ a/scripts/checksyscalls.sh 2008-07-16 22:41:36.000000000 -0400
+@@ -108,7 +108,6 @@ cat << EOF
+ #define __IGNORE_afs_syscall
+ #define __IGNORE_getpmsg
+ #define __IGNORE_putpmsg
+-#define __IGNORE_vserver
+ EOF
+ }
+
+--- a/security/commoncap.c 2008-07-14 17:23:05.000000000 -0400
++++ a/security/commoncap.c 2008-07-16 22:48:47.000000000 -0400
+@@ -26,10 +26,11 @@
+ #include <linux/sched.h>
+ #include <linux/prctl.h>
+ #include <linux/securebits.h>
++#include <linux/vs_context.h>
+
+ int cap_netlink_send(struct sock *sk, struct sk_buff *skb)
+ {
+- NETLINK_CB(skb).eff_cap = current->cap_effective;
++ NETLINK_CB(skb).eff_cap = vx_mbcaps(current->cap_effective);
+ return 0;
+ }
+
+@@ -50,9 +51,24 @@ EXPORT_SYMBOL(cap_netlink_recv);
+ */
+ int cap_capable (struct task_struct *tsk, int cap)
+ {
++ struct vx_info *vxi = tsk->vx_info;
++
++#if 0
++ printk("cap_capable() VXF_STATE_SETUP = %llx, raised = %x, eff = %08x:%08x\n",
++ vx_info_flags(vxi, VXF_STATE_SETUP, 0),
++ cap_raised(tsk->cap_effective, cap),
++ tsk->cap_effective.cap[1], tsk->cap_effective.cap[0]);
++#endif
++
++ /* special case SETUP */
++ if (vx_info_flags(vxi, VXF_STATE_SETUP, 0) &&
++ cap_raised(tsk->cap_effective, cap))
++ return 0;
++
+ /* Derived from include/linux/sched.h:capable. */
+- if (cap_raised(tsk->cap_effective, cap))
++ if (vx_cap_raised(vxi, tsk->cap_effective, cap))
+ return 0;
++
+ return -EPERM;
+ }
+
+@@ -675,7 +691,8 @@ void cap_task_reparent_to_init (struct t
+
+ int cap_syslog (int type)
+ {
+- if ((type != 3 && type != 10) && !capable(CAP_SYS_ADMIN))
++ if ((type != 3 && type != 10) &&
++ !vx_capable(CAP_SYS_ADMIN, VXC_SYSLOG))
+ return -EPERM;
+ return 0;
+ }
+--- a/security/dummy.c 2008-07-14 17:23:05.000000000 -0400
++++ a/security/dummy.c 2008-07-16 22:48:12.000000000 -0400
+@@ -29,6 +29,7 @@
+ #include <linux/file.h>
+ #include <linux/prctl.h>
+ #include <linux/securebits.h>
++#include <linux/vs_context.h>
+
+ static int dummy_ptrace (struct task_struct *parent, struct task_struct *child)
+ {
+@@ -749,7 +750,7 @@ static int dummy_sem_semop (struct sem_a
+
+ static int dummy_netlink_send (struct sock *sk, struct sk_buff *skb)
+ {
+- NETLINK_CB(skb).eff_cap = current->cap_effective;
++ NETLINK_CB(skb).eff_cap = vx_mbcaps(current->cap_effective);
+ return 0;
+ }
+
+--- a/security/selinux/hooks.c 2008-07-14 17:23:05.000000000 -0400
++++ a/security/selinux/hooks.c 2008-07-16 22:41:36.000000000 -0400
+@@ -65,7 +65,6 @@
+ #include <linux/dccp.h>
+ #include <linux/quota.h>
+ #include <linux/un.h> /* for Unix socket types */
+-#include <net/af_unix.h> /* for Unix socket types */
+ #include <linux/parser.h>
+ #include <linux/nfs_mount.h>
+ #include <net/ipv6.h>
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/console-hvc-overtake.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/console-hvc-overtake.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,71 @@
+diff --git a/drivers/xen/console/console.c b/drivers/xen/console/console.c
+index e4e5d41..03f8ce9 100644
+--- a/drivers/xen/console/console.c
++++ b/drivers/xen/console/console.c
+@@ -66,20 +66,25 @@
+ * 'xencons=tty' [XC_TTY]: Console attached to '/dev/tty[0-9]+'.
+ * 'xencons=ttyS' [XC_SERIAL]: Console attached to '/dev/ttyS[0-9]+'.
+ * 'xencons=xvc' [XC_XVC]: Console attached to '/dev/xvc0'.
+- * default: XC_XVC
++ * 'xencons=hvc' [XC_HVC]: Console attached to '/dev/hvc0'.
++ * default: XC_HVC
+ *
+ * NB. In mode XC_TTY, we create dummy consoles for tty2-63. This suppresses
+ * warnings from standard distro startup scripts.
+ */
+ static enum {
+- XC_OFF, XC_TTY, XC_SERIAL, XC_XVC
+-} xc_mode = XC_XVC;
++ XC_OFF, XC_TTY, XC_SERIAL, XC_XVC, XC_HVC
++} xc_mode = XC_HVC;
+ static int xc_num = -1;
+
+ /* /dev/xvc0 device number allocated by lanana.org. */
+ #define XEN_XVC_MAJOR 204
+ #define XEN_XVC_MINOR 191
+
++/* /dev/hvc0 device number */
++#define XEN_HVC_MAJOR 229
++#define XEN_HVC_MINOR 0
++
+ static int __init xencons_setup(char *str)
+ {
+ char *q;
+@@ -97,6 +102,9 @@ static int __init xencons_setup(char *str)
+ } else if (!strncmp(str, "xvc", 3)) {
+ xc_mode = XC_XVC;
+ str += 3;
++ } else if (!strncmp(str, "hvc", 3)) {
++ xc_mode = XC_HVC;
++ str += 3;
+ } else if (!strncmp(str, "off", 3)) {
+ xc_mode = XC_OFF;
+ str += 3;
+@@ -205,6 +213,14 @@ static int __init xen_console_init(void)
+ xc_num = 0;
+ break;
+
++ case XC_HVC:
++ strcpy(kcons_info.name, "hvc");
++ if (xc_num == -1)
++ xc_num = 0;
++ if (!is_initial_xendomain())
++ add_preferred_console(kcons_info.name, xc_num, NULL);
++ break;
++
+ case XC_SERIAL:
+ strcpy(kcons_info.name, "ttyS");
+ if (xc_num == -1)
+@@ -681,6 +697,12 @@ static int __init xencons_init(void)
+ DRV(xencons_driver)->minor_start = XEN_XVC_MINOR;
+ DRV(xencons_driver)->name_base = xc_num;
+ break;
++ case XC_HVC:
++ DRV(xencons_driver)->name = "hvc";
++ DRV(xencons_driver)->major = XEN_HVC_MAJOR;
++ DRV(xencons_driver)->minor_start = XEN_HVC_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;
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/disable-pat.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/disable-pat.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,25 @@
+diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
+index 170d743..ae1e15b 100644
+--- a/arch/x86/Kconfig
++++ b/arch/x86/Kconfig
+@@ -1141,6 +1141,7 @@ config X86_PAT
+ bool
+ prompt "x86 PAT support"
+ depends on MTRR
++ depends on !XEN
+ help
+ Use PAT attributes to setup page level cache control.
+
+diff --git a/include/asm-x86/mach-xen/asm/pgtable.h b/include/asm-x86/mach-xen/asm/pgtable.h
+index a9ff073..3c37ff4 100644
+--- a/include/asm-x86/mach-xen/asm/pgtable.h
++++ b/include/asm-x86/mach-xen/asm/pgtable.h
+@@ -74,7 +74,7 @@ extern unsigned int __kernel_page_user;
+ * PAT settings are part of the hypervisor interface, which sets the
+ * MSR to 0x050100070406 (i.e. WB, WT, UC-, UC, WC, WP [, UC, UC]).
+ */
+-#define _PAGE_CACHE_MASK (_PAGE_PCD | _PAGE_PWT | _PAGE_PAT)
++#define _PAGE_CACHE_MASK (_PAGE_PCD | _PAGE_PWT)
+ #define _PAGE_CACHE_WB (0)
+ #define _PAGE_CACHE_WT (_PAGE_PWT)
+ #define _PAGE_CACHE_WC (_PAGE_PAT)
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/dom0-fix-processor-throttling-when-pr-id-is-minus-1.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/dom0-fix-processor-throttling-when-pr-id-is-minus-1.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,32 @@
+# HG changeset patch
+# User Keir Fraser <keir.fraser at citrix.com>
+# Date 1225190351 0
+# Node ID de7f94bd650b7e00cd57191280c5f0959b2a286e
+# Parent f40f4f86d5a2c5caa0261512279f3590e95f3d91
+dom0: Fix for throttling while pr->id == -1
+
+Signed-off-by: Wei Gang <gang.wei at intel.com>
+
+diff -r f40f4f86d5a2 -r de7f94bd650b drivers/acpi/processor_core.c
+--- a/drivers/acpi/processor_core.c Mon Oct 27 13:47:07 2008 +0000
++++ b/drivers/acpi/processor_core.c Tue Oct 28 10:39:11 2008 +0000
+@@ -677,8 +677,17 @@
+ #if defined(CONFIG_CPU_FREQ) || defined(CONFIG_PROCESSOR_EXTERNAL_CONTROL)
+ acpi_processor_ppc_has_changed(pr);
+ #endif
+- acpi_processor_get_throttling_info(pr);
+- acpi_processor_get_limit_info(pr);
++
++ /*
++ * pr->id may equal to -1 while processor_cntl_external enabled.
++ * throttle and thermal module don't support this case.
++ * Tx only works when dom0 vcpu == pcpu num by far, as we give
++ * control to dom0.
++ */
++ if (pr->id != -1) {
++ acpi_processor_get_throttling_info(pr);
++ acpi_processor_get_limit_info(pr);
++ }
+
+
+ acpi_processor_power_init(pr, device);
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/fix-pci-hook.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/fix-pci-hook.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,51 @@
+diff --git a/drivers/xen/core/pci.c b/drivers/xen/core/pci.c
+index 2710c18..3fd5167 100644
+--- a/drivers/xen/core/pci.c
++++ b/drivers/xen/core/pci.c
+@@ -12,7 +12,7 @@ static int (*pci_bus_remove)(struct device *dev);
+
+ static int pci_bus_probe_wrapper(struct device *dev)
+ {
+- int r;
++ int r, r1;
+ struct pci_dev *pci_dev = to_pci_dev(dev);
+ struct physdev_manage_pci manage_pci;
+ manage_pci.bus = pci_dev->bus->number;
+@@ -23,16 +23,19 @@ static int pci_bus_probe_wrapper(struct device *dev)
+ return r;
+
+ r = pci_bus_probe(dev);
+- if (r)
+- WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
+- &manage_pci));
++ if (r) {
++ r1 = HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci);
++ if (r1 && r1 != -ENOSYS)
++ WARN_ON(1);
++ }
+
+ return r;
+ }
+
+ static int pci_bus_remove_wrapper(struct device *dev)
+ {
+- int r;
++ int r, r1;
+ struct pci_dev *pci_dev = to_pci_dev(dev);
+ struct physdev_manage_pci manage_pci;
+ manage_pci.bus = pci_dev->bus->number;
+@@ -41,8 +44,11 @@ static int pci_bus_remove_wrapper(struct device *dev)
+ r = pci_bus_remove(dev);
+ /* dev and pci_dev are no longer valid!! */
+
+- WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
+- &manage_pci));
++ r1 = HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci);
++ if (r1 && r1 != -ENOSYS)
++ WARN_ON(1);
++
+ return r;
+ }
+
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/hardcode-xen-makefile.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/hardcode-xen-makefile.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,50 @@
+diff --git a/scripts/Makefile.build b/scripts/Makefile.build
+index e89ae81..302d893 100644
+--- a/scripts/Makefile.build
++++ b/scripts/Makefile.build
+@@ -74,12 +74,6 @@ $(warning kbuild: Makefile.build is included improperly)
+ endif
+
+ ifeq ($(CONFIG_XEN),y)
+-$(objtree)/scripts/Makefile.xen: $(srctree)/scripts/Makefile.xen.awk $(srctree)/scripts/Makefile.build
+- @echo ' Updating $@'
+- $(if $(shell echo a | $(AWK) '{ print gensub(/a/, "AA", "g"); }'),\
+- ,$(error 'Your awk program does not define gensub. Use gawk or another awk with gensub'))
+- @$(AWK) -f $< $(filter-out $<,$^) >$@
+-
+ xen-src-single-used-m := $(patsubst $(srctree)/%,%,$(wildcard $(addprefix $(srctree)/,$(single-used-m:.o=-xen.c))))
+ xen-single-used-m := $(xen-src-single-used-m:-xen.c=.o)
+ single-used-m := $(filter-out $(xen-single-used-m),$(single-used-m))
+diff --git a/scripts/Makefile.xen b/scripts/Makefile.xen
+new file mode 100644
+index 0000000..c1cf128
+--- /dev/null
++++ b/scripts/Makefile.xen
+@@ -0,0 +1,27 @@
++$(obj)/%.s: $(src)/%-xen.c FORCE
++ $(call if_changed_dep,cc_s_c)
++
++$(obj)/%.i: $(src)/%-xen.c FORCE
++ $(call if_changed_dep,cc_i_c)
++
++$(obj)/%.symtypes : $(src)/%-xen.c FORCE
++ $(call if_changed_dep,cc_symtypes_c)
++
++$(obj)/%.o: $(src)/%-xen.c FORCE
++ $(call cmd,force_checksrc)
++ $(call if_changed_rule,cc_o_c)
++
++$(xen-single-used-m): $(obj)/%.o: $(src)/%-xen.c FORCE
++ $(call cmd,force_checksrc)
++ $(call if_changed_rule,cc_o_c)
++ @{ echo $(@:.o=.ko); echo $@; } > $(MODVERDIR)/$(@F:.o=.mod)
++
++$(obj)/%.lst: $(src)/%-xen.c FORCE
++ $(call if_changed_dep,cc_lst_c)
++
++$(obj)/%.s: $(src)/%-xen.S FORCE
++ $(call if_changed_dep,as_s_S)
++
++$(obj)/%.o: $(src)/%-xen.S FORCE
++ $(call if_changed_dep,as_o_S)
++
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/remove-4gb-warning.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/remove-4gb-warning.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,55 @@
+diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
+index 81f7529..c41a5c8 100644
+--- a/arch/x86/kernel/Makefile
++++ b/arch/x86/kernel/Makefile
+@@ -91,8 +91,6 @@ scx200-y += scx200_32.o
+
+ obj-$(CONFIG_OLPC) += olpc.o
+
+-obj-$(CONFIG_X86_XEN) += fixup.o
+-
+ ###
+ # 64 bit specific files
+ ifeq ($(CONFIG_X86_64),y)
+diff --git a/arch/x86/kernel/entry_32-xen.S b/arch/x86/kernel/entry_32-xen.S
+index ca66938..37f6f70 100644
+--- a/arch/x86/kernel/entry_32-xen.S
++++ b/arch/x86/kernel/entry_32-xen.S
+@@ -1238,15 +1238,8 @@ ENTRY(spurious_interrupt_bug)
+ CFI_ADJUST_CFA_OFFSET 4
+ jmp error_code
+ CFI_ENDPROC
+-#endif /* !CONFIG_XEN */
+-
+-ENTRY(fixup_4gb_segment)
+- RING0_EC_FRAME
+- pushl $do_fixup_4gb_segment
+- CFI_ADJUST_CFA_OFFSET 4
+- jmp error_code
+- CFI_ENDPROC
+ END(spurious_interrupt_bug)
++#endif /* !CONFIG_XEN */
+
+ ENTRY(kernel_thread_helper)
+ pushl $0 # fake return address for unwinder
+diff --git a/arch/x86/kernel/traps_32-xen.c b/arch/x86/kernel/traps_32-xen.c
+index c174679..0ce86ca 100644
+--- a/arch/x86/kernel/traps_32-xen.c
++++ b/arch/x86/kernel/traps_32-xen.c
+@@ -101,8 +101,6 @@ 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);
+
+@@ -1172,7 +1170,6 @@ static const trap_info_t __cpuinitconst trap_table[] = {
+ { 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
Added: dists/trunk/linux-2.6/debian/patches/features/all/xen/suse-20080808143035.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/all/xen/suse-20080808143035.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,119931 @@
+diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
+index bf07b6f..170d743 100644
+--- a/arch/x86/Kconfig
++++ b/arch/x86/Kconfig
+@@ -23,8 +23,8 @@ config X86
+ select HAVE_OPROFILE
+ select HAVE_KPROBES
+ select HAVE_KRETPROBES
+- select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64)
+- select HAVE_ARCH_KGDB if !X86_VOYAGER
++ select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) && !XEN
++ select HAVE_ARCH_KGDB if !X86_VOYAGER && !XEN
+
+ config ARCH_DEFCONFIG
+ string
+@@ -43,13 +43,16 @@ config GENERIC_CMOS_UPDATE
+
+ config CLOCKSOURCE_WATCHDOG
+ def_bool y
++ depends on !XEN
+
+ config GENERIC_CLOCKEVENTS
+ def_bool y
++ depends on !XEN
+
+ config GENERIC_CLOCKEVENTS_BROADCAST
+ def_bool y
+ depends on X86_64 || (X86_32 && X86_LOCAL_APIC)
++ depends on !XEN
+
+ config LOCKDEP_SUPPORT
+ def_bool y
+@@ -129,6 +132,7 @@ config HAVE_CPUMASK_OF_CPU_MAP
+ config ARCH_HIBERNATION_POSSIBLE
+ def_bool y
+ depends on !SMP || !X86_VOYAGER
++ depends on !XEN
+
+ config ARCH_SUSPEND_POSSIBLE
+ def_bool y
+@@ -182,18 +186,29 @@ config X86_HT
+ bool
+ depends on SMP
+ depends on (X86_32 && !(X86_VISWS || X86_VOYAGER)) || X86_64
++ depends on !XEN
+ default y
+
+ config X86_BIOS_REBOOT
+ bool
+ depends on !X86_VISWS && !X86_VOYAGER
++ depends on !XEN
+ default y
+
+ config X86_TRAMPOLINE
+ bool
+ depends on X86_SMP || (X86_VOYAGER && SMP) || (64BIT && ACPI_SLEEP)
++ depends on !XEN
+ default y
+
++config X86_NO_TSS
++ def_bool y
++ depends on XEN
++
++config X86_NO_IDT
++ def_bool y
++ depends on XEN
++
+ config KTIME_SCALAR
+ def_bool X86_32
+ source "init/Kconfig"
+@@ -239,6 +254,18 @@ config X86_PC
+ help
+ Choose this option if your computer is a standard PC or compatible.
+
++config X86_XEN
++ bool "Xen-compatible"
++ depends on X86_32
++ select XEN
++ select X86_PAE
++ select X86_UP_APIC if !SMP && XEN_PRIVILEGED_GUEST
++ select X86_UP_IOAPIC if !SMP && XEN_PRIVILEGED_GUEST
++ select SWIOTLB
++ help
++ Choose this option if you plan to run this kernel on top of the
++ Xen Hypervisor.
++
+ config X86_ELAN
+ bool "AMD Elan"
+ depends on X86_32
+@@ -334,6 +361,14 @@ config X86_RDC321X
+ as R-8610-(G).
+ If you don't have one of these chips, you should say N here.
+
++config X86_64_XEN
++ bool "Enable Xen compatible kernel"
++ depends on X86_64
++ select XEN
++ select SWIOTLB
++ help
++ This option will compile a kernel compatible with Xen hypervisor
++
+ config X86_VSMP
+ bool "Support for ScaleMP vSMP"
+ select PARAVIRT
+@@ -359,6 +394,7 @@ config SCHED_NO_NO_OMIT_FRAME_POINTER
+
+ menuconfig PARAVIRT_GUEST
+ bool "Paravirtualized guest support"
++ depends on !XEN
+ help
+ Say Y here to get to see options related to running Linux under
+ various hypervisors. This option alone does not add any kernel code.
+@@ -419,7 +455,7 @@ endif
+
+ config MEMTEST_BOOTPARAM
+ bool "Memtest boot parameter"
+- depends on X86_64
++ depends on X86_64 && !XEN
+ default y
+ help
+ This option adds a kernel parameter 'memtest', which allows memtest
+@@ -472,6 +508,7 @@ source "arch/x86/Kconfig.cpu"
+ config HPET_TIMER
+ def_bool X86_64
+ prompt "HPET Timer Support" if X86_32
++ depends on !XEN
+ help
+ Use the IA-PC HPET (High Precision Event Timer) to manage
+ time in preference to the PIT and RTC, if a HPET is
+@@ -508,7 +545,7 @@ config GART_IOMMU
+ default y
+ select SWIOTLB
+ select AGP
+- depends on X86_64 && PCI
++ depends on X86_64 && PCI && !X86_64_XEN
+ help
+ Support for full DMA access of devices with 32bit memory access only
+ on systems with more than 3GB. This is usually needed for USB,
+@@ -523,7 +560,7 @@ config GART_IOMMU
+ config CALGARY_IOMMU
+ bool "IBM Calgary IOMMU support"
+ select SWIOTLB
+- depends on X86_64 && PCI && EXPERIMENTAL
++ depends on X86_64 && PCI && !X86_64_XEN && EXPERIMENTAL
+ help
+ Support for hardware IOMMUs in IBM's xSeries x366 and x460
+ systems. Needed to run systems with more than 3GB of memory
+@@ -567,6 +604,7 @@ config NR_CPUS
+ range 2 255
+ depends on SMP
+ default "32" if X86_NUMAQ || X86_SUMMIT || X86_BIGSMP || X86_ES7000
++ default "16" if X86_64_XEN
+ default "8"
+ help
+ This allows you to specify the maximum number of CPUs which this
+@@ -598,7 +636,7 @@ source "kernel/Kconfig.preempt"
+
+ config X86_UP_APIC
+ bool "Local APIC support on uniprocessors"
+- depends on X86_32 && !SMP && !(X86_VISWS || X86_VOYAGER || X86_GENERICARCH)
++ depends on X86_32 && !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
+@@ -624,18 +662,24 @@ config X86_UP_IOAPIC
+ config X86_LOCAL_APIC
+ def_bool y
+ depends on X86_64 || (X86_32 && (X86_UP_APIC || ((X86_VISWS || SMP) && !X86_VOYAGER) || X86_GENERICARCH))
++ depends on !XEN_UNPRIVILEGED_GUEST
+
+ config X86_IO_APIC
+ def_bool y
+ depends on X86_64 || (X86_32 && (X86_UP_IOAPIC || (SMP && !(X86_VISWS || X86_VOYAGER)) || X86_GENERICARCH))
++ depends on !XEN_UNPRIVILEGED_GUEST
+
+ config X86_VISWS_APIC
+ def_bool y
+ depends on X86_32 && X86_VISWS
+
++config X86_XEN_GENAPIC
++ def_bool y
++ depends on X86_64_XEN
++
+ config X86_MCE
+ bool "Machine Check Exception"
+- depends on !X86_VOYAGER
++ depends on !(X86_VOYAGER || XEN)
+ ---help---
+ Machine Check Exception support allows the processor to notify the
+ kernel if it detects a problem (e.g. overheating, component failure).
+@@ -735,7 +779,7 @@ config I8K
+ config X86_REBOOTFIXUPS
+ def_bool n
+ prompt "Enable X86 board specific fixups for reboot"
+- depends on X86_32 && X86
++ depends on X86_32 && !XEN
+ ---help---
+ This enables chipset and/or board specific fixups to be done
+ in order to get reboot to work correctly. This is only needed on
+@@ -752,6 +796,7 @@ config X86_REBOOTFIXUPS
+
+ 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, you will be able to update the microcode on
+@@ -910,7 +955,7 @@ config X86_PAE
+ # Common NUMA Features
+ config NUMA
+ bool "Numa Memory Allocation and Scheduler Support (EXPERIMENTAL)"
+- depends on SMP
++ depends on SMP && !X86_64_XEN
+ depends on X86_64 || (X86_32 && HIGHMEM64G && (X86_NUMAQ || (X86_SUMMIT || X86_GENERICARCH) && ACPI) && EXPERIMENTAL)
+ default n if X86_PC
+ default y if (X86_NUMAQ || X86_SUMMIT)
+@@ -1003,11 +1048,11 @@ config ARCH_DISCONTIGMEM_DEFAULT
+
+ config ARCH_SPARSEMEM_DEFAULT
+ def_bool y
+- depends on X86_64
++ depends on X86_64 && !X86_64_XEN
+
+ config ARCH_SPARSEMEM_ENABLE
+ def_bool y
+- depends on X86_64 || NUMA || (EXPERIMENTAL && X86_PC)
++ depends on (X86_64 && !X86_64_XEN) || NUMA || (EXPERIMENTAL && X86_PC)
+ select SPARSEMEM_STATIC if X86_32
+ select SPARSEMEM_VMEMMAP_ENABLE if X86_64
+
+@@ -1033,6 +1078,7 @@ config HIGHPTE
+ config MATH_EMULATION
+ bool
+ prompt "Math emulation" if X86_32
++ depends on !XEN
+ ---help---
+ Linux can emulate a math coprocessor (used for floating point
+ operations) if you don't have one. 486DX and Pentium processors have
+@@ -1058,6 +1104,7 @@ config MATH_EMULATION
+
+ config MTRR
+ bool "MTRR (Memory Type Range Register) support"
++ depends on !XEN_UNPRIVILEGED_GUEST
+ ---help---
+ On Intel P6 family processors (Pentium Pro, Pentium II and later)
+ the Memory Type Range Registers (MTRRs) may be used to control
+@@ -1108,7 +1155,7 @@ config X86_PAT
+ config EFI
+ def_bool n
+ prompt "EFI runtime service support"
+- depends on ACPI
++ depends on ACPI && !XEN
+ ---help---
+ This enables the kernel to use EFI runtime services that are
+ available (such as the EFI variable services).
+@@ -1123,7 +1170,7 @@ config EFI
+ config IRQBALANCE
+ def_bool y
+ prompt "Enable kernel irq balancing"
+- depends on X86_32 && SMP && X86_IO_APIC
++ depends on X86_32 && SMP && X86_IO_APIC && !XEN
+ help
+ The default yes will allow the kernel to do irq load balancing.
+ Saying no will keep the kernel from doing irq load balancing.
+@@ -1174,6 +1221,7 @@ source kernel/Kconfig.hz
+ config KEXEC
+ bool "kexec system call"
+ depends on X86_BIOS_REBOOT
++ depends on !XEN_UNPRIVILEGED_GUEST
+ help
+ kexec is a system call that implements the ability to shutdown your
+ current kernel, and to start another kernel. It is like a reboot
+@@ -1192,6 +1240,7 @@ config CRASH_DUMP
+ bool "kernel crash dumps (EXPERIMENTAL)"
+ depends on EXPERIMENTAL
+ depends on X86_64 || (X86_32 && HIGHMEM)
++ depends on !XEN
+ help
+ Generate crash dump after being started by kexec.
+ This should be normally only set in special crash dump kernels
+@@ -1248,7 +1297,7 @@ config PHYSICAL_START
+
+ config RELOCATABLE
+ bool "Build a relocatable kernel (EXPERIMENTAL)"
+- depends on EXPERIMENTAL
++ depends on EXPERIMENTAL && !XEN
+ help
+ This builds a kernel image that retains relocation information
+ so it can be loaded someplace besides the default 1MB.
+@@ -1322,7 +1371,7 @@ config HAVE_ARCH_EARLY_PFN_TO_NID
+ depends on NUMA
+
+ menu "Power management options"
+- depends on !X86_VOYAGER
++ depends on !(X86_VOYAGER || XEN_UNPRIVILEGED_GUEST)
+
+ config ARCH_HIBERNATION_HEADER
+ def_bool y
+@@ -1339,7 +1388,7 @@ config X86_APM_BOOT
+
+ menuconfig APM
+ tristate "APM (Advanced Power Management) BIOS support"
+- depends on X86_32 && PM_SLEEP && !X86_VISWS
++ depends on X86_32 && PM_SLEEP && !(X86_VISWS || XEN)
+ ---help---
+ APM is a BIOS specification for saving power using several different
+ techniques. This is mostly useful for battery powered laptops with
+@@ -1506,6 +1555,7 @@ choice
+
+ config PCI_GOBIOS
+ bool "BIOS"
++ depends on !XEN
+
+ config PCI_GOMMCONFIG
+ bool "MMConfig"
+@@ -1517,6 +1567,13 @@ config PCI_GOOLPC
+ bool "OLPC"
+ depends on OLPC
+
++config PCI_GOXEN_FE
++ bool "Xen PCI Frontend"
++ depends on X86_XEN
++ help
++ The PCI device frontend driver allows the kernel to import arbitrary
++ PCI devices from a PCI backend to support PCI driver domains.
++
+ config PCI_GOANY
+ bool "Any"
+
+@@ -1524,7 +1581,7 @@ endchoice
+
+ config PCI_BIOS
+ def_bool y
+- depends on X86_32 && !X86_VISWS && PCI && (PCI_GOBIOS || PCI_GOANY)
++ depends on X86_32 && !(X86_VISWS || XEN) && PCI && (PCI_GOBIOS || PCI_GOANY)
+
+ # x86-64 doesn't support PCI BIOS access from long mode so always go direct.
+ config PCI_DIRECT
+@@ -1547,9 +1604,25 @@ config PCI_MMCONFIG
+ bool "Support mmconfig PCI config space access"
+ depends on X86_64 && PCI && ACPI
+
++config XEN_PCIDEV_FRONTEND
++ def_bool y
++ prompt "Xen PCI Frontend" if X86_64
++ depends on PCI && XEN && (PCI_GOXEN_FE || PCI_GOANY || X86_64)
++ select HOTPLUG
++ help
++ The PCI device frontend driver allows the kernel to import arbitrary
++ PCI devices from a PCI backend to support PCI driver domains.
++
++config XEN_PCIDEV_FE_DEBUG
++ bool "Xen PCI Frontend Debugging"
++ depends on XEN_PCIDEV_FRONTEND
++ help
++ Enables some debug statements within the PCI Frontend.
++
+ config DMAR
+ bool "Support for DMA Remapping Devices (EXPERIMENTAL)"
+ depends on X86_64 && PCI_MSI && ACPI && EXPERIMENTAL
++ depends on !XEN
+ help
+ DMA remapping (DMAR) devices support enables independent address
+ translations for Direct Memory Access (DMA) from devices.
+@@ -1589,7 +1662,7 @@ if X86_32
+
+ config ISA
+ bool "ISA support"
+- depends on !(X86_VOYAGER || X86_VISWS)
++ depends on !(X86_VOYAGER || X86_VISWS || XEN)
+ 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
+@@ -1616,7 +1689,7 @@ config EISA
+ source "drivers/eisa/Kconfig"
+
+ config MCA
+- bool "MCA support" if !(X86_VISWS || X86_VOYAGER)
++ bool "MCA support" if !(X86_VISWS || X86_VOYAGER || XEN)
+ default y if X86_VOYAGER
+ help
+ MicroChannel Architecture is found in some IBM PS/2 machines and
+diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu
+index dd138a2..25d4d15 100644
+--- a/arch/x86/Kconfig.cpu
++++ b/arch/x86/Kconfig.cpu
+@@ -340,7 +340,7 @@ config X86_PPRO_FENCE
+
+ config X86_F00F_BUG
+ def_bool y
+- depends on M586MMX || M586TSC || M586 || M486 || M386
++ depends on (M586MMX || M586TSC || M586 || M486 || M386) && !X86_NO_IDT
+
+ config X86_WP_WORKS_OK
+ def_bool y
+@@ -398,6 +398,7 @@ config X86_P6_NOP
+ config X86_TSC
+ def_bool y
+ depends on ((MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MVIAC7 || MGEODEGX1 || MGEODE_LX || MCORE2) && !X86_NUMAQ) || X86_64
++ depends on !X86_XEN && !X86_64_XEN
+
+ # this should be set for all -march=.. options where the compiler
+ # generates cmov.
+diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
+index 1836337..ab8966c 100644
+--- a/arch/x86/Kconfig.debug
++++ b/arch/x86/Kconfig.debug
+@@ -93,7 +93,7 @@ config DEBUG_RODATA
+
+ config DIRECT_GBPAGES
+ bool "Enable gbpages-mapped kernel pagetables"
+- depends on DEBUG_KERNEL && EXPERIMENTAL && X86_64
++ depends on DEBUG_KERNEL && EXPERIMENTAL && X86_64 && !XEN
+ help
+ Enable gigabyte pages support (if the CPU supports it). This can
+ improve the kernel's performance a tiny bit by reducing TLB
+@@ -137,11 +137,12 @@ config X86_FIND_SMP_CONFIG
+ config X86_MPPARSE
+ def_bool y
+ depends on (X86_32 && (X86_LOCAL_APIC && !X86_VISWS)) || X86_64
++ depends on !XEN_UNPRIVILEGED_GUEST
+
+ config DOUBLEFAULT
+ default y
+ bool "Enable doublefault exception handler" if EMBEDDED
+- depends on X86_32
++ depends on X86_32 && !X86_NO_TSS
+ help
+ This option allows trapping of rare doublefault exceptions that
+ would otherwise cause a system to silently reboot. Disabling this
+diff --git a/arch/x86/Makefile b/arch/x86/Makefile
+index 3cff3c8..862c302 100644
+--- a/arch/x86/Makefile
++++ b/arch/x86/Makefile
+@@ -129,6 +129,10 @@ mcore-$(CONFIG_X86_BIGSMP) := arch/x86/mach-default/
+ mflags-$(CONFIG_X86_SUMMIT) := -Iinclude/asm-x86/mach-summit
+ mcore-$(CONFIG_X86_SUMMIT) := arch/x86/mach-default/
+
++# Xen subarch support
++mflags-$(CONFIG_X86_XEN) := -Iinclude/asm-x86/mach-xen
++mcore-$(CONFIG_X86_XEN) := arch/x86/mach-xen/
++
+ # generic subarchitecture
+ mflags-$(CONFIG_X86_GENERICARCH):= -Iinclude/asm-x86/mach-generic
+ fcore-$(CONFIG_X86_GENERICARCH) += arch/x86/mach-generic/
+@@ -168,7 +172,7 @@ libs-y += arch/x86/lib/
+ core-y += $(fcore-y)
+
+ # Xen paravirtualization support
+-core-$(CONFIG_XEN) += arch/x86/xen/
++core-$(CONFIG_PARAVIRT_XEN) += arch/x86/xen/
+
+ # lguest paravirtualization support
+ core-$(CONFIG_LGUEST_GUEST) += arch/x86/lguest/
+@@ -202,9 +206,28 @@ endif
+
+ boot := arch/x86/boot
+
+-PHONY += zImage bzImage compressed zlilo bzlilo \
++PHONY += zImage bzImage vmlinuz compressed zlilo bzlilo \
+ zdisk bzdisk fdimage fdimage144 fdimage288 isoimage install
+
++ifdef CONFIG_XEN
++KBUILD_CPPFLAGS := -D__XEN_INTERFACE_VERSION__=$(CONFIG_XEN_INTERFACE_VERSION) \
++ -Iinclude$(if $(KBUILD_SRC),2)/asm/mach-xen $(KBUILD_CPPFLAGS)
++
++ifdef CONFIG_X86_64
++LDFLAGS_vmlinux := -e startup_64
++endif
++
++# Default kernel to build
++all: vmlinuz
++
++# KBUILD_IMAGE specifies the target image being built
++KBUILD_IMAGE := $(boot)/vmlinuz
++
++vmlinuz: vmlinux
++ $(Q)$(MAKE) $(build)=$(boot) $(KBUILD_IMAGE)
++ $(Q)mkdir -p $(objtree)/arch/$(UTS_MACHINE)/boot
++ $(Q)ln -fsn ../../x86/boot/$@ $(objtree)/arch/$(UTS_MACHINE)/boot/$@
++else
+ # Default kernel to build
+ all: bzImage
+
+@@ -227,6 +250,7 @@ zdisk bzdisk: vmlinux
+
+ fdimage fdimage144 fdimage288 isoimage: vmlinux
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) $@
++endif
+
+ install:
+ $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install
+diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile
+index 7ee102f..cb1904e 100644
+--- a/arch/x86/boot/Makefile
++++ b/arch/x86/boot/Makefile
+@@ -25,7 +25,7 @@ SVGA_MODE := -DSVGA_MODE=NORMAL_VGA
+
+ #RAMDISK := -DRAMDISK=512
+
+-targets := vmlinux.bin setup.bin setup.elf zImage bzImage
++targets := vmlinux.bin setup.bin setup.elf zImage bzImage vmlinuz vmlinux-stripped
+ subdir- := compressed
+
+ setup-y += a20.o cmdline.o copy.o cpu.o cpucheck.o edd.o
+@@ -190,5 +190,13 @@ zlilo: $(BOOTIMAGE)
+ cp System.map $(INSTALL_PATH)/
+ if [ -x /sbin/lilo ]; then /sbin/lilo; else /etc/lilo/install; fi
+
++$(obj)/vmlinuz: $(obj)/vmlinux-stripped FORCE
++ $(call if_changed,gzip)
++ @echo 'Kernel: $@ is ready' ' (#'`cat .version`')'
++
++$(obj)/vmlinux-stripped: OBJCOPYFLAGS := -g --strip-unneeded
++$(obj)/vmlinux-stripped: vmlinux FORCE
++ $(call if_changed,objcopy)
++
+ install:
+ sh $(srctree)/$(src)/install.sh $(KERNELRELEASE) $(BOOTIMAGE) System.map "$(INSTALL_PATH)"
+diff --git a/arch/x86/ia32/ia32entry-xen.S b/arch/x86/ia32/ia32entry-xen.S
+new file mode 100644
+index 0000000..3164c99
+--- /dev/null
++++ b/arch/x86/ia32/ia32entry-xen.S
+@@ -0,0 +1,673 @@
++/*
++ * Compatibility mode system call entry point for x86-64.
++ *
++ * Copyright 2000-2002 Andi Kleen, SuSE Labs.
++ */
++
++#include <asm/dwarf2.h>
++#include <asm/calling.h>
++#include <asm/asm-offsets.h>
++#include <asm/current.h>
++#include <asm/errno.h>
++#include <asm/ia32_unistd.h>
++#include <asm/thread_info.h>
++#include <asm/segment.h>
++#include <asm/irqflags.h>
++#include <linux/linkage.h>
++
++#define IA32_NR_syscalls ((ia32_syscall_end - ia32_sys_call_table)/8)
++
++ .macro IA32_ARG_FIXUP noebp=0
++ movl %edi,%r8d
++ .if \noebp
++ .else
++ movl %ebp,%r9d
++ .endif
++ xchg %ecx,%esi
++ movl %ebx,%edi
++ movl %edx,%edx /* zero extension */
++ .endm
++
++ /* clobbers %eax */
++ .macro CLEAR_RREGS
++ xorl %eax,%eax
++ movq %rax,R11(%rsp)
++ movq %rax,R10(%rsp)
++ movq %rax,R9(%rsp)
++ movq %rax,R8(%rsp)
++ .endm
++
++ .macro LOAD_ARGS32 offset
++ movl \offset(%rsp),%r11d
++ movl \offset+8(%rsp),%r10d
++ movl \offset+16(%rsp),%r9d
++ movl \offset+24(%rsp),%r8d
++ movl \offset+40(%rsp),%ecx
++ movl \offset+48(%rsp),%edx
++ movl \offset+56(%rsp),%esi
++ movl \offset+64(%rsp),%edi
++ movl \offset+72(%rsp),%eax
++ .endm
++
++ .macro CFI_STARTPROC32 simple
++ CFI_STARTPROC \simple
++ CFI_UNDEFINED r8
++ CFI_UNDEFINED r9
++ CFI_UNDEFINED r10
++ CFI_UNDEFINED r11
++ CFI_UNDEFINED r12
++ CFI_UNDEFINED r13
++ CFI_UNDEFINED r14
++ CFI_UNDEFINED r15
++ .endm
++
++/*
++ * 32bit SYSENTER instruction entry.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx Arg2
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp user stack
++ * 0(%ebp) Arg6
++ *
++ * Interrupts on.
++ *
++ * This is purely a fast path. For anything complicated we use the int 0x80
++ * path below. Set up a complete hardware stack frame to share code
++ * with the int 0x80 path.
++ */
++ENTRY(ia32_sysenter_target)
++ CFI_STARTPROC32 simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ CFI_REL_OFFSET r11,8
++ CFI_REL_OFFSET rcx,0
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ popq %rcx
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ movl %ebp,%ebp /* zero extension */
++ movl %eax,%eax
++ movl 48-THREAD_SIZE+threadinfo_sysenter_return(%rsp),%r10d
++ movl $__USER32_DS,40(%rsp)
++ movq %rbp,32(%rsp)
++ movl $__USER32_CS,16(%rsp)
++ movq %r10,8(%rsp)
++ movq %rax,(%rsp)
++ cld
++ SAVE_ARGS 0,0,1
++ /* no need to do an access_ok check here because rbp has been
++ 32bit zero extended */
++1: movl (%rbp),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz sysenter_tracesys
++sysenter_do_call:
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja ia32_badsys
++ IA32_ARG_FIXUP 1
++ call *ia32_sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++sysenter_tracesys:
++ xchgl %r9d,%ebp
++ SAVE_REST
++ CLEAR_RREGS
++ movq %r9,R9(%rsp)
++ movq $-ENOSYS,RAX(%rsp)/* ptrace can change this for a bad syscall */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ xchgl %ebp,%r9d
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja int_ret_from_sys_call /* sysenter_tracesys has set RAX(%rsp) */
++ jmp sysenter_do_call
++ CFI_ENDPROC
++ENDPROC(ia32_sysenter_target)
++
++/*
++ * 32bit SYSCALL instruction entry.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx return EIP
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp Arg2 [note: not saved in the stack frame, should not be touched]
++ * %esp user stack
++ * 0(%esp) Arg6
++ *
++ * Interrupts on.
++ *
++ * This is purely a fast path. For anything complicated we use the int 0x80
++ * path below. Set up a complete hardware stack frame to share code
++ * with the int 0x80 path.
++ */
++ENTRY(ia32_cstar_target)
++ CFI_STARTPROC32 simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ movl %eax,%eax /* zero extension */
++ movl RSP-RIP+16(%rsp),%r8d
++ SAVE_ARGS -8,1,1
++ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
++ movq %rbp,RCX-ARGOFFSET(%rsp) /* this lies slightly to ptrace */
++ movl %ebp,%ecx
++ movl $__USER32_CS,CS-ARGOFFSET(%rsp)
++ movl $__USER32_DS,SS-ARGOFFSET(%rsp)
++ /* no need to do an access_ok check here because r8 has been
++ 32bit zero extended */
++ /* hardware stack frame is complete now */
++1: movl (%r8),%r9d
++ .section __ex_table,"a"
++ .quad 1b,ia32_badarg
++ .previous
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz cstar_tracesys
++cstar_do_call:
++ cmpl $IA32_NR_syscalls-1,%eax
++ ja ia32_badsys
++ IA32_ARG_FIXUP 1
++ call *ia32_sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++cstar_tracesys:
++ xchgl %r9d,%ebp
++ SAVE_REST
++ CLEAR_RREGS
++ movq %r9,R9(%rsp)
++ movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ xchgl %ebp,%r9d
++ movl RSP-ARGOFFSET(%rsp), %r8d
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja int_ret_from_sys_call /* cstar_tracesys has set RAX(%rsp) */
++ jmp cstar_do_call
++END(ia32_cstar_target)
++
++ia32_badarg:
++ movq $-EFAULT,%rax
++ jmp ia32_sysret
++ CFI_ENDPROC
++
++/*
++ * Emulated IA32 system calls via int 0x80.
++ *
++ * Arguments:
++ * %eax System call number.
++ * %ebx Arg1
++ * %ecx Arg2
++ * %edx Arg3
++ * %esi Arg4
++ * %edi Arg5
++ * %ebp Arg6 [note: not saved in the stack frame, should not be touched]
++ *
++ * Notes:
++ * Uses the same stack frame as the x86-64 version.
++ * All registers except %eax must be saved (but ptrace may violate that)
++ * Arguments are zero extended. For system calls that want sign extension and
++ * take long arguments a wrapper is needed. Most calls can just be called
++ * directly.
++ * Assumes it is only called from user space and entered with interrupts on.
++ */
++
++ENTRY(ia32_syscall)
++ CFI_STARTPROC32 simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA rsp,SS+8-RIP+16
++ /*CFI_REL_OFFSET ss,SS-RIP+16*/
++ CFI_REL_OFFSET rsp,RSP-RIP+16
++ /*CFI_REL_OFFSET rflags,EFLAGS-RIP+16*/
++ /*CFI_REL_OFFSET cs,CS-RIP+16*/
++ CFI_REL_OFFSET rip,RIP-RIP+16
++ CFI_REL_OFFSET r11,8
++ CFI_REL_OFFSET rcx,0
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ popq %rcx
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ movl %eax,%eax
++ movq %rax,(%rsp)
++ cld
++ /* note the registers are not zero extended to the sf.
++ this could be a problem. */
++ SAVE_ARGS 0,0,1
++ GET_THREAD_INFO(%r10)
++ orl $TS_COMPAT,threadinfo_status(%r10)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%r10)
++ jnz ia32_tracesys
++ia32_do_syscall:
++ cmpl $(IA32_NR_syscalls-1),%eax
++ ja int_ret_from_sys_call /* ia32_tracesys has set RAX(%rsp) */
++ IA32_ARG_FIXUP
++ call *ia32_sys_call_table(,%rax,8) # xxx: rip relative
++ia32_sysret:
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++ia32_tracesys:
++ SAVE_REST
++ CLEAR_RREGS
++ movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */
++ movq %rsp,%rdi /* &pt_regs -> arg1 */
++ call syscall_trace_enter
++ LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ jmp ia32_do_syscall
++END(ia32_syscall)
++
++ia32_badsys:
++ movq $0,ORIG_RAX-ARGOFFSET(%rsp)
++ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
++ jmp int_ret_from_sys_call
++
++quiet_ni_syscall:
++ movq $-ENOSYS,%rax
++ ret
++ CFI_ENDPROC
++
++ .macro PTREGSCALL label, func, arg
++ .globl \label
++\label:
++ leaq \func(%rip),%rax
++ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
++ jmp ia32_ptregs_common
++ .endm
++
++ CFI_STARTPROC32
++
++ PTREGSCALL stub32_rt_sigreturn, sys32_rt_sigreturn, %rdi
++ PTREGSCALL stub32_sigreturn, sys32_sigreturn, %rdi
++ PTREGSCALL stub32_sigaltstack, sys32_sigaltstack, %rdx
++ PTREGSCALL stub32_sigsuspend, sys32_sigsuspend, %rcx
++ PTREGSCALL stub32_execve, sys32_execve, %rcx
++ PTREGSCALL stub32_fork, sys_fork, %rdi
++ PTREGSCALL stub32_clone, sys32_clone, %rdx
++ PTREGSCALL stub32_vfork, sys_vfork, %rdi
++ PTREGSCALL stub32_iopl, sys_iopl, %rsi
++ PTREGSCALL stub32_rt_sigsuspend, sys_rt_sigsuspend, %rdx
++
++ENTRY(ia32_ptregs_common)
++ popq %r11
++ CFI_ENDPROC
++ CFI_STARTPROC32 simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA rsp,SS+8-ARGOFFSET
++ CFI_REL_OFFSET rax,RAX-ARGOFFSET
++ CFI_REL_OFFSET rcx,RCX-ARGOFFSET
++ CFI_REL_OFFSET rdx,RDX-ARGOFFSET
++ CFI_REL_OFFSET rsi,RSI-ARGOFFSET
++ CFI_REL_OFFSET rdi,RDI-ARGOFFSET
++ CFI_REL_OFFSET rip,RIP-ARGOFFSET
++/* CFI_REL_OFFSET cs,CS-ARGOFFSET*/
++/* CFI_REL_OFFSET rflags,EFLAGS-ARGOFFSET*/
++ CFI_REL_OFFSET rsp,RSP-ARGOFFSET
++/* CFI_REL_OFFSET ss,SS-ARGOFFSET*/
++ SAVE_REST
++ call *%rax
++ RESTORE_REST
++ jmp ia32_sysret /* misbalances the return cache */
++ CFI_ENDPROC
++END(ia32_ptregs_common)
++
++ .section .rodata,"a"
++ .align 8
++ia32_sys_call_table:
++ .quad sys_restart_syscall
++ .quad sys_exit
++ .quad stub32_fork
++ .quad sys_read
++ .quad sys_write
++ .quad compat_sys_open /* 5 */
++ .quad sys_close
++ .quad sys32_waitpid
++ .quad sys_creat
++ .quad sys_link
++ .quad sys_unlink /* 10 */
++ .quad stub32_execve
++ .quad sys_chdir
++ .quad compat_sys_time
++ .quad sys_mknod
++ .quad sys_chmod /* 15 */
++ .quad sys_lchown16
++ .quad quiet_ni_syscall /* old break syscall holder */
++ .quad sys_stat
++ .quad sys32_lseek
++ .quad sys_getpid /* 20 */
++ .quad compat_sys_mount /* mount */
++ .quad sys_oldumount /* old_umount */
++ .quad sys_setuid16
++ .quad sys_getuid16
++ .quad compat_sys_stime /* stime */ /* 25 */
++ .quad compat_sys_ptrace /* ptrace */
++ .quad sys_alarm
++ .quad sys_fstat /* (old)fstat */
++ .quad sys_pause
++ .quad compat_sys_utime /* 30 */
++ .quad quiet_ni_syscall /* old stty syscall holder */
++ .quad quiet_ni_syscall /* old gtty syscall holder */
++ .quad sys_access
++ .quad sys_nice
++ .quad quiet_ni_syscall /* 35 */ /* old ftime syscall holder */
++ .quad sys_sync
++ .quad sys32_kill
++ .quad sys_rename
++ .quad sys_mkdir
++ .quad sys_rmdir /* 40 */
++ .quad sys_dup
++ .quad sys32_pipe
++ .quad compat_sys_times
++ .quad quiet_ni_syscall /* old prof syscall holder */
++ .quad sys_brk /* 45 */
++ .quad sys_setgid16
++ .quad sys_getgid16
++ .quad sys_signal
++ .quad sys_geteuid16
++ .quad sys_getegid16 /* 50 */
++ .quad sys_acct
++ .quad sys_umount /* new_umount */
++ .quad quiet_ni_syscall /* old lock syscall holder */
++ .quad compat_sys_ioctl
++ .quad compat_sys_fcntl64 /* 55 */
++ .quad quiet_ni_syscall /* old mpx syscall holder */
++ .quad sys_setpgid
++ .quad quiet_ni_syscall /* old ulimit syscall holder */
++ .quad sys32_olduname
++ .quad sys_umask /* 60 */
++ .quad sys_chroot
++ .quad sys32_ustat
++ .quad sys_dup2
++ .quad sys_getppid
++ .quad sys_getpgrp /* 65 */
++ .quad sys_setsid
++ .quad sys32_sigaction
++ .quad sys_sgetmask
++ .quad sys_ssetmask
++ .quad sys_setreuid16 /* 70 */
++ .quad sys_setregid16
++ .quad stub32_sigsuspend
++ .quad compat_sys_sigpending
++ .quad sys_sethostname
++ .quad compat_sys_setrlimit /* 75 */
++ .quad compat_sys_old_getrlimit /* old_getrlimit */
++ .quad compat_sys_getrusage
++ .quad sys32_gettimeofday
++ .quad sys32_settimeofday
++ .quad sys_getgroups16 /* 80 */
++ .quad sys_setgroups16
++ .quad sys32_old_select
++ .quad sys_symlink
++ .quad sys_lstat
++ .quad sys_readlink /* 85 */
++ .quad sys_uselib
++ .quad sys_swapon
++ .quad sys_reboot
++ .quad compat_sys_old_readdir
++ .quad sys32_mmap /* 90 */
++ .quad sys_munmap
++ .quad sys_truncate
++ .quad sys_ftruncate
++ .quad sys_fchmod
++ .quad sys_fchown16 /* 95 */
++ .quad sys_getpriority
++ .quad sys_setpriority
++ .quad quiet_ni_syscall /* old profil syscall holder */
++ .quad compat_sys_statfs
++ .quad compat_sys_fstatfs /* 100 */
++ .quad sys_ioperm
++ .quad compat_sys_socketcall
++ .quad sys_syslog
++ .quad compat_sys_setitimer
++ .quad compat_sys_getitimer /* 105 */
++ .quad compat_sys_newstat
++ .quad compat_sys_newlstat
++ .quad compat_sys_newfstat
++ .quad sys32_uname
++ .quad stub32_iopl /* 110 */
++ .quad sys_vhangup
++ .quad quiet_ni_syscall /* old "idle" system call */
++ .quad sys32_vm86_warning /* vm86old */
++ .quad compat_sys_wait4
++ .quad sys_swapoff /* 115 */
++ .quad compat_sys_sysinfo
++ .quad sys32_ipc
++ .quad sys_fsync
++ .quad stub32_sigreturn
++ .quad stub32_clone /* 120 */
++ .quad sys_setdomainname
++ .quad sys_uname
++ .quad sys_modify_ldt
++ .quad compat_sys_adjtimex
++ .quad sys32_mprotect /* 125 */
++ .quad compat_sys_sigprocmask
++ .quad quiet_ni_syscall /* create_module */
++ .quad sys_init_module
++ .quad sys_delete_module
++ .quad quiet_ni_syscall /* 130 get_kernel_syms */
++ .quad sys32_quotactl
++ .quad sys_getpgid
++ .quad sys_fchdir
++ .quad quiet_ni_syscall /* bdflush */
++ .quad sys_sysfs /* 135 */
++ .quad sys_personality
++ .quad quiet_ni_syscall /* for afs_syscall */
++ .quad sys_setfsuid16
++ .quad sys_setfsgid16
++ .quad sys_llseek /* 140 */
++ .quad compat_sys_getdents
++ .quad compat_sys_select
++ .quad sys_flock
++ .quad sys_msync
++ .quad compat_sys_readv /* 145 */
++ .quad compat_sys_writev
++ .quad sys_getsid
++ .quad sys_fdatasync
++ .quad sys32_sysctl /* sysctl */
++ .quad sys_mlock /* 150 */
++ .quad sys_munlock
++ .quad sys_mlockall
++ .quad sys_munlockall
++ .quad sys_sched_setparam
++ .quad sys_sched_getparam /* 155 */
++ .quad sys_sched_setscheduler
++ .quad sys_sched_getscheduler
++ .quad sys_sched_yield
++ .quad sys_sched_get_priority_max
++ .quad sys_sched_get_priority_min /* 160 */
++ .quad sys32_sched_rr_get_interval
++ .quad compat_sys_nanosleep
++ .quad sys_mremap
++ .quad sys_setresuid16
++ .quad sys_getresuid16 /* 165 */
++ .quad sys32_vm86_warning /* vm86 */
++ .quad quiet_ni_syscall /* query_module */
++ .quad sys_poll
++ .quad compat_sys_nfsservctl
++ .quad sys_setresgid16 /* 170 */
++ .quad sys_getresgid16
++ .quad sys_prctl
++ .quad stub32_rt_sigreturn
++ .quad sys32_rt_sigaction
++ .quad sys32_rt_sigprocmask /* 175 */
++ .quad sys32_rt_sigpending
++ .quad compat_sys_rt_sigtimedwait
++ .quad sys32_rt_sigqueueinfo
++ .quad stub32_rt_sigsuspend
++ .quad sys32_pread /* 180 */
++ .quad sys32_pwrite
++ .quad sys_chown16
++ .quad sys_getcwd
++ .quad sys_capget
++ .quad sys_capset
++ .quad stub32_sigaltstack
++ .quad sys32_sendfile
++ .quad quiet_ni_syscall /* streams1 */
++ .quad quiet_ni_syscall /* streams2 */
++ .quad stub32_vfork /* 190 */
++ .quad compat_sys_getrlimit
++ .quad sys32_mmap2
++ .quad sys32_truncate64
++ .quad sys32_ftruncate64
++ .quad sys32_stat64 /* 195 */
++ .quad sys32_lstat64
++ .quad sys32_fstat64
++ .quad sys_lchown
++ .quad sys_getuid
++ .quad sys_getgid /* 200 */
++ .quad sys_geteuid
++ .quad sys_getegid
++ .quad sys_setreuid
++ .quad sys_setregid
++ .quad sys_getgroups /* 205 */
++ .quad sys_setgroups
++ .quad sys_fchown
++ .quad sys_setresuid
++ .quad sys_getresuid
++ .quad sys_setresgid /* 210 */
++ .quad sys_getresgid
++ .quad sys_chown
++ .quad sys_setuid
++ .quad sys_setgid
++ .quad sys_setfsuid /* 215 */
++ .quad sys_setfsgid
++ .quad sys_pivot_root
++ .quad sys_mincore
++ .quad sys_madvise
++ .quad compat_sys_getdents64 /* 220 getdents64 */
++ .quad compat_sys_fcntl64
++ .quad quiet_ni_syscall /* tux */
++ .quad quiet_ni_syscall /* security */
++ .quad sys_gettid
++ .quad sys32_readahead /* 225 */
++ .quad sys_setxattr
++ .quad sys_lsetxattr
++ .quad sys_fsetxattr
++ .quad sys_getxattr
++ .quad sys_lgetxattr /* 230 */
++ .quad sys_fgetxattr
++ .quad sys_listxattr
++ .quad sys_llistxattr
++ .quad sys_flistxattr
++ .quad sys_removexattr /* 235 */
++ .quad sys_lremovexattr
++ .quad sys_fremovexattr
++ .quad sys_tkill
++ .quad sys_sendfile64
++ .quad compat_sys_futex /* 240 */
++ .quad compat_sys_sched_setaffinity
++ .quad compat_sys_sched_getaffinity
++ .quad sys_set_thread_area
++ .quad sys_get_thread_area
++ .quad compat_sys_io_setup /* 245 */
++ .quad sys_io_destroy
++ .quad compat_sys_io_getevents
++ .quad compat_sys_io_submit
++ .quad sys_io_cancel
++ .quad sys32_fadvise64 /* 250 */
++ .quad quiet_ni_syscall /* free_huge_pages */
++ .quad sys_exit_group
++ .quad sys32_lookup_dcookie
++ .quad sys_epoll_create
++ .quad sys_epoll_ctl /* 255 */
++ .quad sys_epoll_wait
++ .quad sys_remap_file_pages
++ .quad sys_set_tid_address
++ .quad compat_sys_timer_create
++ .quad compat_sys_timer_settime /* 260 */
++ .quad compat_sys_timer_gettime
++ .quad sys_timer_getoverrun
++ .quad sys_timer_delete
++ .quad compat_sys_clock_settime
++ .quad compat_sys_clock_gettime /* 265 */
++ .quad compat_sys_clock_getres
++ .quad compat_sys_clock_nanosleep
++ .quad compat_sys_statfs64
++ .quad compat_sys_fstatfs64
++ .quad sys_tgkill /* 270 */
++ .quad compat_sys_utimes
++ .quad sys32_fadvise64_64
++ .quad quiet_ni_syscall /* sys_vserver */
++ .quad sys_mbind
++ .quad compat_sys_get_mempolicy /* 275 */
++ .quad sys_set_mempolicy
++ .quad compat_sys_mq_open
++ .quad sys_mq_unlink
++ .quad compat_sys_mq_timedsend
++ .quad compat_sys_mq_timedreceive /* 280 */
++ .quad compat_sys_mq_notify
++ .quad compat_sys_mq_getsetattr
++ .quad compat_sys_kexec_load /* reserved for kexec */
++ .quad compat_sys_waitid
++ .quad quiet_ni_syscall /* 285: sys_altroot */
++ .quad sys_add_key
++ .quad sys_request_key
++ .quad sys_keyctl
++ .quad sys_ioprio_set
++ .quad sys_ioprio_get /* 290 */
++ .quad sys_inotify_init
++ .quad sys_inotify_add_watch
++ .quad sys_inotify_rm_watch
++ .quad sys_migrate_pages
++ .quad compat_sys_openat /* 295 */
++ .quad sys_mkdirat
++ .quad sys_mknodat
++ .quad sys_fchownat
++ .quad compat_sys_futimesat
++ .quad sys32_fstatat /* 300 */
++ .quad sys_unlinkat
++ .quad sys_renameat
++ .quad sys_linkat
++ .quad sys_symlinkat
++ .quad sys_readlinkat /* 305 */
++ .quad sys_fchmodat
++ .quad sys_faccessat
++ .quad compat_sys_pselect6
++ .quad compat_sys_ppoll
++ .quad sys_unshare /* 310 */
++ .quad compat_sys_set_robust_list
++ .quad compat_sys_get_robust_list
++ .quad sys_splice
++ .quad sys32_sync_file_range
++ .quad sys_tee /* 315 */
++ .quad compat_sys_vmsplice
++ .quad compat_sys_move_pages
++ .quad sys_getcpu
++ .quad sys_epoll_pwait
++ .quad compat_sys_utimensat /* 320 */
++ .quad compat_sys_signalfd
++ .quad sys_timerfd_create
++ .quad sys_eventfd
++ .quad sys32_fallocate
++ .quad compat_sys_timerfd_settime /* 325 */
++ .quad compat_sys_timerfd_gettime
++ia32_syscall_end:
+diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
+index 77807d4..81f7529 100644
+--- a/arch/x86/kernel/Makefile
++++ b/arch/x86/kernel/Makefile
+@@ -91,10 +91,13 @@ scx200-y += scx200_32.o
+
+ obj-$(CONFIG_OLPC) += olpc.o
+
++obj-$(CONFIG_X86_XEN) += fixup.o
++
+ ###
+ # 64 bit specific files
+ ifeq ($(CONFIG_X86_64),y)
+- obj-y += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o
++ obj-$(CONFIG_X86_LOCAL_APIC) += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o
++ obj-$(CONFIG_X86_XEN_GENAPIC) += genapic_64.o genapic_xen_64.o
+ obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o
+ obj-$(CONFIG_AUDIT) += audit_64.o
+
+@@ -103,4 +106,12 @@ ifeq ($(CONFIG_X86_64),y)
+ obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o
+
+ obj-$(CONFIG_PCI_MMCONFIG) += mmconf-fam10h_64.o
++
++ obj-$(CONFIG_XEN) += nmi_64.o
++ apic_64-$(CONFIG_XEN) += apic_32.o
++ time_64-$(CONFIG_XEN) += time_32.o
+ endif
++
++disabled-obj-$(CONFIG_XEN) := early-quirks.o genapic_flat_$(BITS).o genx2apic_uv_x.o \
++ hpet.o i8253.o i8259_$(BITS).o pci-swiotlb_64.o reboot.o smpboot.o \
++ tlb_$(BITS).o tsc_$(BITS).o tsc_sync.o vsmp_64.o
+diff --git a/arch/x86/kernel/acpi/Makefile b/arch/x86/kernel/acpi/Makefile
+index fd5ca97..52c5db9 100644
+--- a/arch/x86/kernel/acpi/Makefile
++++ b/arch/x86/kernel/acpi/Makefile
+@@ -5,6 +5,9 @@ obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup_rm.o wakeup_$(BITS).o
+
+ ifneq ($(CONFIG_ACPI_PROCESSOR),)
+ obj-y += cstate.o processor.o
++ifneq ($(CONFIG_PROCESSOR_EXTERNAL_CONTROL),)
++obj-$(CONFIG_XEN) += processor_extcntl_xen.o
++endif
+ endif
+
+ $(obj)/wakeup_rm.o: $(obj)/realmode/wakeup.bin
+@@ -12,3 +15,4 @@ $(obj)/wakeup_rm.o: $(obj)/realmode/wakeup.bin
+ $(obj)/realmode/wakeup.bin: FORCE
+ $(Q)$(MAKE) $(build)=$(obj)/realmode
+
++disabled-obj-$(CONFIG_XEN) := cstate.o wakeup_%.o
+diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
+index 33c5216..08d6740 100644
+--- a/arch/x86/kernel/acpi/boot.c
++++ b/arch/x86/kernel/acpi/boot.c
+@@ -88,7 +88,7 @@ int acpi_sci_override_gsi __initdata;
+ int acpi_skip_timer_override __initdata;
+ int acpi_use_timer_override __initdata;
+
+-#ifdef CONFIG_X86_LOCAL_APIC
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
+ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE;
+ #endif
+
+@@ -106,7 +106,7 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE;
+ */
+ 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 *__init __acpi_map_table(unsigned long phys_addr, unsigned long size)
+@@ -139,8 +139,13 @@ char *__init __acpi_map_table(unsigned long phys, unsigned long size)
+ unsigned long base, offset, mapped_size;
+ int idx;
+
++#ifndef CONFIG_XEN
+ if (phys + size < 8 * 1024 * 1024)
+ return __va(phys);
++#else
++ if (phys + size <= (NR_FIX_ISAMAPS << PAGE_SHIFT))
++ return isa_bus_to_virt(phys);
++#endif
+
+ offset = phys & (PAGE_SIZE - 1);
+ mapped_size = PAGE_SIZE - offset;
+@@ -228,12 +233,14 @@ static int __init acpi_parse_madt(struct acpi_table_header *table)
+ return -ENODEV;
+ }
+
++#ifndef CONFIG_XEN
+ if (madt->address) {
+ acpi_lapic_addr = (u64) madt->address;
+
+ printk(KERN_DEBUG PREFIX "Local APIC address 0x%08x\n",
+ madt->address);
+ }
++#endif
+
+ acpi_madt_oem_check(madt->header.oem_id, madt->header.oem_table_id);
+
+@@ -242,19 +249,23 @@ static int __init acpi_parse_madt(struct acpi_table_header *table)
+
+ static void __cpuinit acpi_register_lapic(int id, u8 enabled)
+ {
++#ifndef CONFIG_XEN
+ unsigned int ver = 0;
++#endif
+
+ if (!enabled) {
+ ++disabled_cpus;
+ return;
+ }
+
++#ifndef CONFIG_XEN
+ #ifdef CONFIG_X86_32
+ if (boot_cpu_physical_apicid != -1U)
+ ver = apic_version[boot_cpu_physical_apicid];
+ #endif
+
+ generic_processor_info(id, ver);
++#endif
+ }
+
+ static int __init
+@@ -304,6 +315,7 @@ static int __init
+ acpi_parse_lapic_addr_ovr(struct acpi_subtable_header * header,
+ const unsigned long end)
+ {
++#ifndef CONFIG_XEN
+ struct acpi_madt_local_apic_override *lapic_addr_ovr = NULL;
+
+ lapic_addr_ovr = (struct acpi_madt_local_apic_override *)header;
+@@ -312,6 +324,7 @@ acpi_parse_lapic_addr_ovr(struct acpi_subtable_header * header,
+ return -EINVAL;
+
+ acpi_lapic_addr = lapic_addr_ovr->address;
++#endif
+
+ return 0;
+ }
+@@ -769,6 +782,7 @@ static int __init acpi_parse_fadt(struct acpi_table_header *table)
+ * returns 0 on success, < 0 on error
+ */
+
++#ifndef CONFIG_XEN
+ static void __init acpi_register_lapic_address(unsigned long address)
+ {
+ mp_lapic_addr = address;
+@@ -782,6 +796,9 @@ static void __init acpi_register_lapic_address(unsigned long address)
+ #endif
+ }
+ }
++#else
++#define acpi_register_lapic_address(address)
++#endif
+
+ static int __init early_acpi_parse_madt_lapic_addr_ovr(void)
+ {
+diff --git a/arch/x86/kernel/acpi/processor.c b/arch/x86/kernel/acpi/processor.c
+index de2d2e4..6dd1d80 100644
+--- a/arch/x86/kernel/acpi/processor.c
++++ b/arch/x86/kernel/acpi/processor.c
+@@ -69,7 +69,18 @@ static void init_intel_pdc(struct acpi_processor *pr, struct cpuinfo_x86 *c)
+ /* Initialize _PDC data based on the CPU vendor */
+ void arch_acpi_processor_init_pdc(struct acpi_processor *pr)
+ {
++#ifdef CONFIG_XEN
++ /*
++ * As a work-around, just use cpu0's cpuinfo for all processors.
++ * Further work is required to expose xen hypervisor interface of
++ * getting physical cpuinfo to dom0 kernel and then
++ * arch_acpi_processor_init_pdc can set _PDC parameters according
++ * to Xen's phys information.
++ */
++ struct cpuinfo_x86 *c = &boot_cpu_data;
++#else
+ struct cpuinfo_x86 *c = &cpu_data(pr->id);
++#endif
+
+ pr->pdc = NULL;
+ if (c->x86_vendor == X86_VENDOR_INTEL)
+diff --git a/arch/x86/kernel/acpi/processor_extcntl_xen.c b/arch/x86/kernel/acpi/processor_extcntl_xen.c
+new file mode 100644
+index 0000000..cab3a42
+--- /dev/null
++++ b/arch/x86/kernel/acpi/processor_extcntl_xen.c
+@@ -0,0 +1,229 @@
++/*
++ * processor_extcntl_xen.c - interface to notify Xen
++ *
++ * Copyright (C) 2008, Intel corporation
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or (at
++ * your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful, but
++ * WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
++ * General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License along
++ * with this program; if not, write to the Free Software Foundation, Inc.,
++ * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
++ *
++ */
++
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/types.h>
++#include <linux/acpi.h>
++#include <linux/pm.h>
++#include <linux/cpu.h>
++
++#include <linux/cpufreq.h>
++#include <acpi/processor.h>
++#include <asm/hypercall.h>
++
++static int xen_processor_pmbits;
++
++static int xen_cx_notifier(struct acpi_processor *pr, int action)
++{
++ int ret, count = 0, i;
++ xen_platform_op_t op = {
++ .cmd = XENPF_set_processor_pminfo,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u.set_pminfo.id = pr->acpi_id,
++ .u.set_pminfo.type = XEN_PM_CX,
++ };
++ struct xen_processor_cx *data, *buf;
++ struct acpi_processor_cx *cx;
++
++ if (action == PROCESSOR_PM_CHANGE)
++ return -EINVAL;
++
++ /* Convert to Xen defined structure and hypercall */
++ buf = kzalloc(pr->power.count * sizeof(struct xen_processor_cx),
++ GFP_KERNEL);
++ if (!buf)
++ return -ENOMEM;
++
++ data = buf;
++ for (i = 1; i <= pr->power.count; i++) {
++ cx = &pr->power.states[i];
++ /* Skip invalid cstate entry */
++ if (!cx->valid)
++ continue;
++
++ data->type = cx->type;
++ data->latency = cx->latency;
++ data->power = cx->power;
++ data->reg.space_id = cx->reg.space_id;
++ data->reg.bit_width = cx->reg.bit_width;
++ data->reg.bit_offset = cx->reg.bit_offset;
++ data->reg.access_size = cx->reg.reserved;
++ data->reg.address = cx->reg.address;
++
++ /* Get dependency relationships */
++ if (cx->csd_count) {
++ printk("Wow! _CSD is found. Not support for now!\n");
++ kfree(buf);
++ return -EINVAL;
++ } else {
++ data->dpcnt = 0;
++ set_xen_guest_handle(data->dp, NULL);
++ }
++
++ data++;
++ count++;
++ }
++
++ if (!count) {
++ printk("No available Cx info for cpu %d\n", pr->acpi_id);
++ kfree(buf);
++ return -EINVAL;
++ }
++
++ op.u.set_pminfo.power.count = count;
++ op.u.set_pminfo.power.flags.bm_control = pr->flags.bm_control;
++ op.u.set_pminfo.power.flags.bm_check = pr->flags.bm_check;
++ op.u.set_pminfo.power.flags.has_cst = pr->flags.has_cst;
++ op.u.set_pminfo.power.flags.power_setup_done = pr->flags.power_setup_done;
++
++ set_xen_guest_handle(op.u.set_pminfo.power.states, buf);
++ ret = HYPERVISOR_platform_op(&op);
++ kfree(buf);
++ return ret;
++}
++
++static void convert_pct_reg(struct xen_pct_register *xpct,
++ struct acpi_pct_register *apct)
++{
++ xpct->descriptor = apct->descriptor;
++ xpct->length = apct->length;
++ xpct->space_id = apct->space_id;
++ xpct->bit_width = apct->bit_width;
++ xpct->bit_offset = apct->bit_offset;
++ xpct->reserved = apct->reserved;
++ xpct->address = apct->address;
++}
++
++static void convert_pss_states(struct xen_processor_px *xpss,
++ struct acpi_processor_px *apss, int state_count)
++{
++ int i;
++ for(i=0; i<state_count; i++) {
++ xpss->core_frequency = apss->core_frequency;
++ xpss->power = apss->power;
++ xpss->transition_latency = apss->transition_latency;
++ xpss->bus_master_latency = apss->bus_master_latency;
++ xpss->control = apss->control;
++ xpss->status = apss->status;
++ xpss++;
++ apss++;
++ }
++}
++
++static void convert_psd_pack(struct xen_psd_package *xpsd,
++ struct acpi_psd_package *apsd)
++{
++ xpsd->num_entries = apsd->num_entries;
++ xpsd->revision = apsd->revision;
++ xpsd->domain = apsd->domain;
++ xpsd->coord_type = apsd->coord_type;
++ xpsd->num_processors = apsd->num_processors;
++}
++
++static int xen_px_notifier(struct acpi_processor *pr, int action)
++{
++ int ret;
++ xen_platform_op_t op = {
++ .cmd = XENPF_set_processor_pminfo,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u.set_pminfo.id = pr->acpi_id,
++ .u.set_pminfo.type = XEN_PM_PX,
++ };
++ struct xen_processor_performance *perf;
++ struct xen_processor_px *states = NULL;
++ struct acpi_processor_performance *px;
++ struct acpi_psd_package *pdomain;
++
++ /* leave dynamic ppc handle in the future */
++ if (action == PROCESSOR_PM_CHANGE)
++ return 0;
++
++ perf = &op.u.set_pminfo.perf;
++ px = pr->performance;
++
++ perf->flags = XEN_PX_PPC |
++ XEN_PX_PCT |
++ XEN_PX_PSS |
++ XEN_PX_PSD;
++
++ /* ppc */
++ perf->ppc = pr->performance_platform_limit;
++
++ /* pct */
++ convert_pct_reg(&perf->control_register, &px->control_register);
++ convert_pct_reg(&perf->status_register, &px->status_register);
++
++ /* pss */
++ perf->state_count = px->state_count;
++ states = kzalloc(px->state_count*sizeof(xen_processor_px_t),GFP_KERNEL);
++ if (!states)
++ return -ENOMEM;
++ convert_pss_states(states, px->states, px->state_count);
++ set_xen_guest_handle(perf->states, states);
++
++ /* psd */
++ pdomain = &px->domain_info;
++ convert_psd_pack(&perf->domain_info, pdomain);
++ if (perf->domain_info.num_processors) {
++ if (pdomain->coord_type == DOMAIN_COORD_TYPE_SW_ALL)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_ALL;
++ else if (pdomain->coord_type == DOMAIN_COORD_TYPE_SW_ANY)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_ANY;
++ else if (pdomain->coord_type == DOMAIN_COORD_TYPE_HW_ALL)
++ perf->shared_type = CPUFREQ_SHARED_TYPE_HW;
++ } else
++ perf->shared_type = CPUFREQ_SHARED_TYPE_NONE;
++
++ ret = HYPERVISOR_platform_op(&op);
++ kfree(states);
++ return ret;
++}
++
++static int xen_tx_notifier(struct acpi_processor *pr, int action)
++{
++ return -EINVAL;
++}
++static int xen_hotplug_notifier(struct acpi_processor *pr, int event)
++{
++ return -EINVAL;
++}
++
++static struct processor_extcntl_ops xen_extcntl_ops = {
++ .hotplug = xen_hotplug_notifier,
++};
++
++void arch_acpi_processor_init_extcntl(const struct processor_extcntl_ops **ops)
++{
++ xen_processor_pmbits = (xen_start_info->flags & SIF_PM_MASK) >> 8;
++
++ if (xen_processor_pmbits & XEN_PROCESSOR_PM_CX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_IDLE] = xen_cx_notifier;
++ if (xen_processor_pmbits & XEN_PROCESSOR_PM_PX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_PERF] = xen_px_notifier;
++ if (xen_processor_pmbits & XEN_PROCESSOR_PM_TX)
++ xen_extcntl_ops.pm_ops[PM_TYPE_THR] = xen_tx_notifier;
++
++ *ops = &xen_extcntl_ops;
++}
++EXPORT_SYMBOL(arch_acpi_processor_init_extcntl);
+diff --git a/arch/x86/kernel/acpi/sleep-xen.c b/arch/x86/kernel/acpi/sleep-xen.c
+new file mode 100644
+index 0000000..26edc70
+--- /dev/null
++++ b/arch/x86/kernel/acpi/sleep-xen.c
+@@ -0,0 +1,156 @@
++/*
++ * sleep.c - x86-specific ACPI sleep support.
++ *
++ * Copyright (C) 2001-2003 Patrick Mochel
++ * Copyright (C) 2001-2003 Pavel Machek <pavel at suse.cz>
++ */
++
++#include <linux/acpi.h>
++#include <linux/bootmem.h>
++#include <linux/dmi.h>
++#include <linux/cpumask.h>
++
++#include "realmode/wakeup.h"
++#include "sleep.h"
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++unsigned long acpi_wakeup_address;
++unsigned long acpi_realmode_flags;
++
++/* address in low memory of the wakeup routine. */
++static unsigned long acpi_realmode;
++
++#ifdef CONFIG_64BIT
++static char temp_stack[10240];
++#endif
++#endif
++
++/**
++ * acpi_save_state_mem - save kernel state
++ *
++ * Create an identity mapped page table and copy the wakeup routine to
++ * low memory.
++ *
++ * Note that this is too late to change acpi_wakeup_address.
++ */
++int acpi_save_state_mem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ struct wakeup_header *header;
++
++ if (!acpi_realmode) {
++ printk(KERN_ERR "Could not allocate memory during boot, "
++ "S3 disabled\n");
++ return -ENOMEM;
++ }
++ memcpy((void *)acpi_realmode, &wakeup_code_start, WAKEUP_SIZE);
++
++ header = (struct wakeup_header *)(acpi_realmode + HEADER_OFFSET);
++ if (header->signature != 0x51ee1111) {
++ printk(KERN_ERR "wakeup header does not match\n");
++ return -EINVAL;
++ }
++
++ header->video_mode = saved_video_mode;
++
++ header->wakeup_jmp_seg = acpi_wakeup_address >> 4;
++ /* GDT[0]: GDT self-pointer */
++ header->wakeup_gdt[0] =
++ (u64)(sizeof(header->wakeup_gdt) - 1) +
++ ((u64)(acpi_wakeup_address +
++ ((char *)&header->wakeup_gdt - (char *)acpi_realmode))
++ << 16);
++ /* GDT[1]: real-mode-like code segment */
++ header->wakeup_gdt[1] = (0x009bULL << 40) +
++ ((u64)acpi_wakeup_address << 16) + 0xffff;
++ /* GDT[2]: real-mode-like data segment */
++ header->wakeup_gdt[2] = (0x0093ULL << 40) +
++ ((u64)acpi_wakeup_address << 16) + 0xffff;
++
++#ifndef CONFIG_64BIT
++ store_gdt((struct desc_ptr *)&header->pmode_gdt);
++
++ header->pmode_efer_low = nx_enabled;
++ if (header->pmode_efer_low & 1) {
++ /* This is strange, why not save efer, always? */
++ rdmsr(MSR_EFER, header->pmode_efer_low,
++ header->pmode_efer_high);
++ }
++#endif /* !CONFIG_64BIT */
++
++ header->pmode_cr0 = read_cr0();
++ header->pmode_cr4 = read_cr4();
++ header->realmode_flags = acpi_realmode_flags;
++ header->real_magic = 0x12345678;
++
++#ifndef CONFIG_64BIT
++ header->pmode_entry = (u32)&wakeup_pmode_return;
++ header->pmode_cr3 = (u32)(swsusp_pg_dir - __PAGE_OFFSET);
++ saved_magic = 0x12345678;
++#else /* CONFIG_64BIT */
++ header->trampoline_segment = setup_trampoline() >> 4;
++ init_rsp = (unsigned long)temp_stack + 4096;
++ initial_code = (unsigned long)wakeup_long64;
++ saved_magic = 0x123456789abcdef0;
++#endif /* CONFIG_64BIT */
++#endif
++
++ return 0;
++}
++
++/*
++ * acpi_restore_state - undo effects of acpi_save_state_mem
++ */
++void acpi_restore_state_mem(void)
++{
++}
++
++
++/**
++ * acpi_reserve_bootmem - do _very_ early ACPI initialisation
++ *
++ * We allocate a page from the first 1MB of memory for the wakeup
++ * routine for when we come back from a sleep state. The
++ * runtime allocator allows specification of <16MB pages, but not
++ * <1MB pages.
++ */
++void __init acpi_reserve_bootmem(void)
++{
++#ifndef CONFIG_ACPI_PV_SLEEP
++ if ((&wakeup_code_end - &wakeup_code_start) > WAKEUP_SIZE) {
++ printk(KERN_ERR
++ "ACPI: Wakeup code way too big, S3 disabled.\n");
++ return;
++ }
++
++ acpi_realmode = (unsigned long)alloc_bootmem_low(WAKEUP_SIZE);
++
++ if (!acpi_realmode) {
++ printk(KERN_ERR "ACPI: Cannot allocate lowmem, S3 disabled.\n");
++ return;
++ }
++
++ acpi_wakeup_address = virt_to_phys((void *)acpi_realmode);
++#endif
++}
++
++
++#ifndef CONFIG_ACPI_PV_SLEEP
++static int __init acpi_sleep_setup(char *str)
++{
++ while ((str != NULL) && (*str != '\0')) {
++ if (strncmp(str, "s3_bios", 7) == 0)
++ acpi_realmode_flags |= 1;
++ if (strncmp(str, "s3_mode", 7) == 0)
++ acpi_realmode_flags |= 2;
++ if (strncmp(str, "s3_beep", 7) == 0)
++ acpi_realmode_flags |= 4;
++ str = strchr(str, ',');
++ if (str != NULL)
++ str += strspn(str, ", \t");
++ }
++ return 1;
++}
++
++__setup("acpi_sleep=", acpi_sleep_setup);
++#endif /* CONFIG_ACPI_PV_SLEEP */
+diff --git a/arch/x86/kernel/apic_32-xen.c b/arch/x86/kernel/apic_32-xen.c
+new file mode 100644
+index 0000000..b8e1e9b
+--- /dev/null
++++ b/arch/x86/kernel/apic_32-xen.c
+@@ -0,0 +1,44 @@
++/*
++ * Local APIC handling stubs
++ */
++
++#include <linux/init.h>
++#include <linux/interrupt.h>
++
++#include <asm/smp.h>
++
++/*
++ * Debug level, exported for io_apic.c
++ */
++int apic_verbosity;
++
++static int __init apic_set_verbosity(char *str)
++{
++ if (strcmp("debug", str) == 0)
++ apic_verbosity = APIC_DEBUG;
++ else if (strcmp("verbose", str) == 0)
++ apic_verbosity = APIC_VERBOSE;
++ return 1;
++}
++
++__setup("apic=", apic_set_verbosity);
++
++int setup_profiling_timer(unsigned int multiplier)
++{
++ return -EINVAL;
++}
++
++/*
++ * 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;
++}
+diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c
+index 9258808..4cfdb26 100644
+--- a/arch/x86/kernel/asm-offsets_32.c
++++ b/arch/x86/kernel/asm-offsets_32.c
+@@ -19,10 +19,14 @@
+ #include <asm/bootparam.h>
+ #include <asm/elf.h>
+
++#if defined(CONFIG_XEN) || defined(CONFIG_PARAVIRT_XEN)
+ #include <xen/interface/xen.h>
++#endif
+
++#ifdef CONFIG_LGUEST_GUEST
+ #include <linux/lguest.h>
+ #include "../../../drivers/lguest/lg.h"
++#endif
+
+ /* workaround for a warning with -Wmissing-prototypes */
+ void foo(void);
+@@ -54,6 +58,7 @@ void foo(void)
+ OFFSET(TI_exec_domain, thread_info, exec_domain);
+ OFFSET(TI_flags, thread_info, flags);
+ OFFSET(TI_status, thread_info, status);
++ OFFSET(TI_cpu, thread_info, cpu);
+ OFFSET(TI_preempt_count, thread_info, preempt_count);
+ OFFSET(TI_addr_limit, thread_info, addr_limit);
+ OFFSET(TI_restart_block, thread_info, restart_block);
+@@ -91,9 +96,14 @@ void foo(void)
+ OFFSET(pbe_orig_address, pbe, orig_address);
+ OFFSET(pbe_next, pbe, next);
+
++#ifndef CONFIG_X86_NO_TSS
+ /* Offset from the sysenter stack to tss.sp0 */
+- DEFINE(TSS_sysenter_sp0, offsetof(struct tss_struct, x86_tss.sp0) -
++ DEFINE(SYSENTER_stack_sp0, offsetof(struct tss_struct, x86_tss.sp0) -
+ sizeof(struct tss_struct));
++#else
++ /* sysenter stack points directly to sp0 */
++ DEFINE(SYSENTER_stack_sp0, 0);
++#endif
+
+ DEFINE(PAGE_SIZE_asm, PAGE_SIZE);
+ DEFINE(PAGE_SHIFT_asm, PAGE_SHIFT);
+@@ -103,6 +113,11 @@ void foo(void)
+
+ OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx);
+
++#ifdef CONFIG_XEN
++ BLANK();
++ OFFSET(XEN_START_mfn_list, start_info, mfn_list);
++#endif
++
+ #ifdef CONFIG_PARAVIRT
+ BLANK();
+ OFFSET(PARAVIRT_enabled, pv_info, paravirt_enabled);
+@@ -115,7 +130,7 @@ void foo(void)
+ OFFSET(PV_CPU_read_cr0, pv_cpu_ops, read_cr0);
+ #endif
+
+-#ifdef CONFIG_XEN
++#ifdef CONFIG_PARAVIRT_XEN
+ BLANK();
+ OFFSET(XEN_vcpu_info_mask, vcpu_info, evtchn_upcall_mask);
+ OFFSET(XEN_vcpu_info_pending, vcpu_info, evtchn_upcall_pending);
+diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
+index f126c05..2a472c8 100644
+--- a/arch/x86/kernel/asm-offsets_64.c
++++ b/arch/x86/kernel/asm-offsets_64.c
+@@ -117,8 +117,10 @@ int main(void)
+ ENTRY(cr8);
+ BLANK();
+ #undef ENTRY
++#ifndef CONFIG_X86_NO_TSS
+ DEFINE(TSS_ist, offsetof(struct tss_struct, x86_tss.ist));
+ BLANK();
++#endif
+ DEFINE(crypto_tfm_ctx_offset, offsetof(struct crypto_tfm, __crt_ctx));
+ BLANK();
+ DEFINE(__NR_syscall_max, sizeof(syscalls) - 1);
+diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
+index a0c6f81..7c518c7 100644
+--- a/arch/x86/kernel/cpu/Makefile
++++ b/arch/x86/kernel/cpu/Makefile
+@@ -17,4 +17,6 @@ obj-$(CONFIG_X86_MCE) += mcheck/
+ obj-$(CONFIG_MTRR) += mtrr/
+ obj-$(CONFIG_CPU_FREQ) += cpufreq/
+
++ifneq ($(CONFIG_XEN),y)
+ obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o
++endif
+diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
+index 2458668..81ded4d 100644
+--- a/arch/x86/kernel/cpu/amd.c
++++ b/arch/x86/kernel/cpu/amd.c
+@@ -24,7 +24,7 @@
+ extern void vide(void);
+ __asm__(".align 4\nvide: ret");
+
+-#ifdef CONFIG_X86_LOCAL_APIC
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
+ #define ENABLE_C1E_MASK 0x18000000
+ #define CPUID_PROCESSOR_SIGNATURE 1
+ #define CPUID_XFAM 0x0ff00000
+@@ -297,7 +297,7 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c)
+ num_cache_leaves = 3;
+ }
+
+-#ifdef CONFIG_X86_LOCAL_APIC
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
+ if (amd_apic_timer_broken())
+ local_apic_timer_disabled = 1;
+ #endif
+diff --git a/arch/x86/kernel/cpu/common-xen.c b/arch/x86/kernel/cpu/common-xen.c
+new file mode 100644
+index 0000000..768f066
+--- /dev/null
++++ b/arch/x86/kernel/cpu/common-xen.c
+@@ -0,0 +1,741 @@
++#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/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>
++#include <asm/pat.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 gdt_page, gdt_page) = { .gdt = {
++ [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } },
++ [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } },
++ [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } },
++ [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } },
++#ifndef CONFIG_XEN
++ /*
++ * 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.
++ */
++ /* 32-bit code */
++ [GDT_ENTRY_PNPBIOS_CS32] = { { { 0x0000ffff, 0x00409a00 } } },
++ /* 16-bit code */
++ [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } },
++ /* 16-bit data */
++ [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } },
++ /* 16-bit data */
++ [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } },
++ /* 16-bit data */
++ [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } },
++ /*
++ * The APM segments have byte granularity and their bases
++ * are set at run time. All have 64k limits.
++ */
++ /* 32-bit code */
++ [GDT_ENTRY_APMBIOS_BASE] = { { { 0x0000ffff, 0x00409a00 } } },
++ /* 16-bit code */
++ [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } },
++ /* data */
++ [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } },
++
++ [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } },
++#endif
++ [GDT_ENTRY_PERCPU] = { { { 0x00000000, 0x00000000 } } },
++} };
++EXPORT_PER_CPU_SYMBOL_GPL(gdt_page);
++
++__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata;
++
++static int cachesize_override __cpuinitdata = -1;
++static int disable_x86_serial_nr __cpuinitdata = 1;
++
++struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {};
++
++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");
++ }
++}
++
++static struct cpu_dev __cpuinitdata default_cpu = {
++ .c_init = default_init,
++ .c_vendor = "Unknown",
++};
++static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu;
++
++static int __init cachesize_setup(char *str)
++{
++ get_option(&str, &cachesize_override);
++ return 1;
++}
++__setup("cachesize=", cachesize_setup);
++
++int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++{
++ 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;
++
++ /* 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 1;
++}
++
++
++void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++{
++ unsigned int n, dummy, ecx, edx, l2size;
++
++ n = cpuid_eax(0x80000000);
++
++ 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 (n < 0x80000006) /* Some chips just has a large L1. */
++ return;
++
++ 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);
++
++ /* Allow user to override all this if necessary. */
++ if (cachesize_override != -1)
++ l2size = cachesize_override;
++
++ if (l2size == 0)
++ return; /* Again, no L2 cache is possible */
++
++ c->x86_cache_size = l2size;
++
++ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
++ l2size, ecx & 0xFF);
++}
++
++/*
++ * 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)
++{
++ 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++;
++ }
++ return NULL; /* Not found */
++}
++
++
++static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c, int early)
++{
++ 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;
++}
++
++
++static int __init x86_fxsr_setup(char *s)
++{
++ setup_clear_cpu_cap(X86_FEATURE_FXSR);
++ setup_clear_cpu_cap(X86_FEATURE_XMM);
++ return 1;
++}
++__setup("nofxsr", x86_fxsr_setup);
++
++
++static int __init x86_sep_setup(char *s)
++{
++ setup_clear_cpu_cap(X86_FEATURE_SEP);
++ return 1;
++}
++__setup("nosep", x86_sep_setup);
++
++
++/* Standard macro to see if a specific flag is changeable */
++static inline int flag_is_changeable_p(u32 flag)
++{
++ 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)
++{
++ return flag_is_changeable_p(X86_EFLAGS_ID);
++}
++
++void __init cpu_detect(struct cpuinfo_x86 *c)
++{
++ /* 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]);
++
++ 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;
++ c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
++ }
++ }
++}
++static void __cpuinit early_get_cap(struct cpuinfo_x86 *c)
++{
++ u32 tfms, xlvl;
++ unsigned int ebx;
++
++ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++ if (have_cpuid_p()) {
++ /* 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;
++ }
++
++ /* 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);
++ }
++ }
++
++ }
++
++}
++
++/*
++ * 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)
++{
++ struct cpuinfo_x86 *c = &boot_cpu_data;
++
++ c->x86_cache_alignment = 32;
++ c->x86_clflush_size = 32;
++
++ if (!have_cpuid_p())
++ return;
++
++ cpu_detect(c);
++
++ get_cpu_vendor(c, 1);
++
++ if (c->x86_vendor != X86_VENDOR_UNKNOWN &&
++ cpu_devs[c->x86_vendor]->c_early_init)
++ cpu_devs[c->x86_vendor]->c_early_init(c);
++
++ early_get_cap(c);
++}
++
++static void __cpuinit generic_identify(struct cpuinfo_x86 *c)
++{
++ u32 tfms, xlvl;
++ unsigned int ebx;
++
++ if (have_cpuid_p()) {
++ /* 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, 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;
++ c->initial_apicid = (ebx >> 24) & 0xFF;
++#ifdef CONFIG_X86_HT
++ c->apicid = phys_pkg_id(c->initial_apicid, 0);
++ c->phys_proc_id = c->initial_apicid;
++#else
++ c->apicid = c->initial_apicid;
++#endif
++ if (test_cpu_cap(c, X86_FEATURE_CLFLSH))
++ 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 */
++ }
++
++ init_scattered_cpuid_features(c);
++ }
++
++}
++
++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_cpu_cap(c, X86_FEATURE_PN);
++
++ /* Disabling the serial number may affect the cpuid level */
++ c->cpuid_level = cpuid_eax(0);
++ }
++}
++
++static int __init x86_serial_nr_setup(char *s)
++{
++ disable_x86_serial_nr = 0;
++ return 1;
++}
++__setup("serialnumber", x86_serial_nr_setup);
++
++
++
++/*
++ * This does the hard work of actually picking apart the CPU stuff...
++ */
++void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++{
++ 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;
++ c->x86_clflush_size = 32;
++ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++
++ 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;
++ }
++
++ generic_identify(c);
++
++ if (this_cpu->c_identify)
++ this_cpu->c_identify(c);
++
++ /*
++ * 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);
++
++ /* Disable the PN if appropriate */
++ squash_the_stupid_serial_number(c);
++
++ /*
++ * The vendor-specific functions might have changed features. Now
++ * we do "generic changes."
++ */
++
++ /* 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);
++ }
++
++ /*
++ * 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];
++ }
++
++ /* Clear all flags overriden by options */
++ for (i = 0; i < NCAPINTS; i++)
++ c->x86_capability[i] &= ~cleared_cpu_caps[i];
++
++ /* Init Machine Check Exception if available. */
++ mcheck_init(c);
++
++ select_idle_routine(c);
++}
++
++void __init identify_boot_cpu(void)
++{
++ identify_cpu(&boot_cpu_data);
++ sysenter_setup();
++ enable_sep_cpu();
++}
++
++void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c)
++{
++ BUG_ON(c == &boot_cpu_data);
++ identify_cpu(c);
++ enable_sep_cpu();
++ 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;
++
++ if (smp_num_siblings == 1) {
++ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
++ } else if (smp_num_siblings > 1) {
++
++ if (smp_num_siblings > NR_CPUS) {
++ printk(KERN_WARNING "CPU: Unsupported number of the "
++ "siblings %d", smp_num_siblings);
++ smp_num_siblings = 1;
++ return;
++ }
++
++ index_msb = get_count_order(smp_num_siblings);
++ c->phys_proc_id = phys_pkg_id(c->initial_apicid, index_msb);
++
++ printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
++ c->phys_proc_id);
++
++ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++
++ index_msb = get_count_order(smp_num_siblings) ;
++
++ core_bits = get_count_order(c->x86_max_cores);
++
++ c->cpu_core_id = phys_pkg_id(c->initial_apicid, 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);
++ }
++}
++#endif
++
++static __init int setup_noclflush(char *arg)
++{
++ setup_clear_cpu_cap(X86_FEATURE_CLFLSH);
++ return 1;
++}
++__setup("noclflush", setup_noclflush);
++
++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");
++}
++
++static __init int setup_disablecpuid(char *arg)
++{
++ int bit;
++ if (get_option(&arg, &bit) && bit < NCAPINTS*32)
++ setup_clear_cpu_cap(bit);
++ else
++ return 0;
++ return 1;
++}
++__setup("clearcpuid=", setup_disablecpuid);
++
++cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
++
++void __init early_cpu_init(void)
++{
++ struct cpu_vendor_dev *cvdev;
++
++ for (cvdev = __x86cpuvendor_start ;
++ cvdev < __x86cpuvendor_end ;
++ cvdev++)
++ cpu_devs[cvdev->vendor] = cvdev->cpu_dev;
++
++ early_cpu_detect();
++ validate_pat_support(&boot_cpu_data);
++}
++
++/* Make sure %fs is initialized properly in idle threads */
++struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs)
++{
++ memset(regs, 0, sizeof(struct pt_regs));
++ regs->fs = __KERNEL_PERCPU;
++ return regs;
++}
++
++/* Current gdt points %fs at the "master" per-cpu area: after this,
++ * it's on the real one. */
++void switch_to_new_gdt(void)
++{
++ struct desc_ptr gdt_descr;
++ unsigned long va, frames[16];
++ int f;
++
++ gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id());
++ gdt_descr.size = GDT_SIZE - 1;
++
++ 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 + 1) / 8))
++ BUG();
++ asm("mov %0, %%fs" : : "r" (__KERNEL_PERCPU) : "memory");
++}
++
++/*
++ * 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)
++{
++ int cpu = smp_processor_id();
++ struct task_struct *curr = current;
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *t = &per_cpu(init_tss, cpu);
++#endif
++ struct thread_struct *thread = &curr->thread;
++
++ 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);
++
++ if (cpu_has_vme || cpu_has_de)
++ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
++
++ switch_to_new_gdt();
++
++ /*
++ * Set up and load the per-CPU TSS and LDT
++ */
++ atomic_inc(&init_mm.mm_count);
++ curr->active_mm = &init_mm;
++ if (curr->mm)
++ BUG();
++ enter_lazy_tlb(&init_mm, curr);
++
++ load_sp0(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);
++#endif
++
++ /* Clear %gs. */
++ asm volatile ("mov %0, %%gs" : : "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);
++
++ /*
++ * Force FPU initialization:
++ */
++ current_thread_info()->status = 0;
++ clear_used_math();
++ mxcsr_feature_mask_init();
++}
++
++#if defined(CONFIG_HOTPLUG_CPU) && !defined(CONFIG_XEN)
++void __cpuinit cpu_uninit(void)
++{
++ int cpu = raw_smp_processor_id();
++ cpu_clear(cpu, cpu_initialized);
++
++ /* lazy TLB state */
++ per_cpu(cpu_tlbstate, cpu).state = 0;
++ per_cpu(cpu_tlbstate, cpu).active_mm = &init_mm;
++}
++#endif
+diff --git a/arch/x86/kernel/cpu/mtrr/Makefile b/arch/x86/kernel/cpu/mtrr/Makefile
+index 191fc05..949d678 100644
+--- a/arch/x86/kernel/cpu/mtrr/Makefile
++++ b/arch/x86/kernel/cpu/mtrr/Makefile
+@@ -1,3 +1,4 @@
+ obj-y := main.o if.o generic.o state.o
+ obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o
+
++obj-$(CONFIG_XEN) := main.o if.o
+diff --git a/arch/x86/kernel/cpu/mtrr/main-xen.c b/arch/x86/kernel/cpu/mtrr/main-xen.c
+new file mode 100644
+index 0000000..29aa598
+--- /dev/null
++++ b/arch/x86/kernel/cpu/mtrr/main-xen.c
+@@ -0,0 +1,330 @@
++#include <linux/init.h>
++#include <linux/proc_fs.h>
++#include <linux/ctype.h>
++#include <linux/module.h>
++#include <linux/seq_file.h>
++#include <asm/uaccess.h>
++#include <linux/mutex.h>
++
++#include <asm/mtrr.h>
++#include "mtrr.h"
++
++static DEFINE_MUTEX(mtrr_mutex);
++
++void generic_get_mtrr(unsigned int reg, unsigned long *base,
++ unsigned long *size, mtrr_type * type)
++{
++ struct xen_platform_op op;
++
++ op.cmd = XENPF_read_memtype;
++ op.u.read_memtype.reg = reg;
++ if (unlikely(HYPERVISOR_platform_op(&op)))
++ memset(&op.u.read_memtype, 0, sizeof(op.u.read_memtype));
++
++ *size = op.u.read_memtype.nr_mfns;
++ *base = op.u.read_memtype.mfn;
++ *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 mtrr_usage_table[MAX_VAR_RANGES];
++
++static u64 tom2;
++
++static void __init set_num_var_ranges(void)
++{
++ struct xen_platform_op op;
++
++ for (num_var_ranges = 0; ; num_var_ranges++) {
++ op.cmd = XENPF_read_memtype;
++ op.u.read_memtype.reg = num_var_ranges;
++ if (HYPERVISOR_platform_op(&op) != 0)
++ break;
++ }
++}
++
++static void __init init_table(void)
++{
++ int i, max;
++
++ max = num_var_ranges;
++ for (i = 0; i < max; i++)
++ mtrr_usage_table[i] = 0;
++}
++
++int mtrr_add_page(unsigned long base, unsigned long size,
++ unsigned int type, bool increment)
++{
++ int error;
++ struct xen_platform_op op;
++
++ mutex_lock(&mtrr_mutex);
++
++ op.cmd = XENPF_add_memtype;
++ op.u.add_memtype.mfn = base;
++ op.u.add_memtype.nr_mfns = size;
++ op.u.add_memtype.type = type;
++ error = HYPERVISOR_platform_op(&op);
++ if (error) {
++ mutex_unlock(&mtrr_mutex);
++ BUG_ON(error > 0);
++ return error;
++ }
++
++ if (increment)
++ ++mtrr_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,
++ bool 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, lsize;
++ int error = -EINVAL;
++ struct xen_platform_op 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 (mtrr_usage_table[reg] < 1) {
++ printk(KERN_WARNING "mtrr: reg: %d has count=0\n", reg);
++ goto out;
++ }
++ if (--mtrr_usage_table[reg] < 1) {
++ op.cmd = XENPF_del_memtype;
++ op.u.del_memtype.handle = 0;
++ op.u.del_memtype.reg = reg;
++ error = HYPERVISOR_platform_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);
++
++/*
++ * Returns the effective MTRR type for the region
++ * Error returns:
++ * - 0xFE - when the range is "not entirely covered" by _any_ var range MTRR
++ * - 0xFF - when MTRR is not enabled
++ */
++u8 mtrr_type_lookup(u64 start, u64 end)
++{
++ int i, error;
++ u64 start_mfn, end_mfn, base_mfn, top_mfn;
++ u8 prev_match, curr_match;
++ struct xen_platform_op op;
++
++ if (!is_initial_xendomain())
++ return MTRR_TYPE_WRBACK;
++
++ if (!num_var_ranges)
++ return 0xFF;
++
++ start_mfn = start >> PAGE_SHIFT;
++ /* Make end inclusive end, instead of exclusive */
++ end_mfn = --end >> PAGE_SHIFT;
++
++ /* Look in fixed ranges. Just return the type as per start */
++ if (start_mfn < 0x100) {
++#if 0//todo
++ op.cmd = XENPF_read_memtype;
++ op.u.read_memtype.reg = ???;
++ error = HYPERVISOR_platform_op(&op);
++ if (!error)
++ return op.u.read_memtype.type;
++#endif
++ return MTRR_TYPE_UNCACHABLE;
++ }
++
++ /*
++ * Look in variable ranges
++ * Look of multiple ranges matching this address and pick type
++ * as per MTRR precedence
++ */
++ prev_match = 0xFF;
++ for (i = 0; i < num_var_ranges; ++i) {
++ op.cmd = XENPF_read_memtype;
++ op.u.read_memtype.reg = i;
++ error = HYPERVISOR_platform_op(&op);
++
++ if (error || !op.u.read_memtype.nr_mfns)
++ continue;
++
++ base_mfn = op.u.read_memtype.mfn;
++ top_mfn = base_mfn + op.u.read_memtype.nr_mfns - 1;
++
++ if (base_mfn > end_mfn || start_mfn > top_mfn) {
++ continue;
++ }
++
++ if (base_mfn > start_mfn || end_mfn > top_mfn) {
++ return 0xFE;
++ }
++
++ curr_match = op.u.read_memtype.type;
++ if (prev_match == 0xFF) {
++ prev_match = curr_match;
++ continue;
++ }
++
++ if (prev_match == MTRR_TYPE_UNCACHABLE ||
++ curr_match == MTRR_TYPE_UNCACHABLE) {
++ return MTRR_TYPE_UNCACHABLE;
++ }
++
++ if ((prev_match == MTRR_TYPE_WRBACK &&
++ curr_match == MTRR_TYPE_WRTHROUGH) ||
++ (prev_match == MTRR_TYPE_WRTHROUGH &&
++ curr_match == MTRR_TYPE_WRBACK)) {
++ prev_match = MTRR_TYPE_WRTHROUGH;
++ curr_match = MTRR_TYPE_WRTHROUGH;
++ }
++
++ if (prev_match != curr_match) {
++ return MTRR_TYPE_UNCACHABLE;
++ }
++ }
++
++ if (tom2) {
++ if (start >= (1ULL<<32) && (end < tom2))
++ return MTRR_TYPE_WRBACK;
++ }
++
++ if (prev_match != 0xFF)
++ return prev_match;
++
++#if 0//todo
++ op.cmd = XENPF_read_def_memtype;
++ error = HYPERVISOR_platform_op(&op);
++ if (!error)
++ return op.u.read_def_memtype.type;
++#endif
++ return MTRR_TYPE_UNCACHABLE;
++}
++
++/*
++ * Newer AMD K8s and later CPUs have a special magic MSR way to force WB
++ * for memory >4GB. Check for that here.
++ * Note this won't check if the MTRRs < 4GB where the magic bit doesn't
++ * apply to are wrong, but so far we don't know of any such case in the wild.
++ */
++#define Tom2Enabled (1U << 21)
++#define Tom2ForceMemTypeWB (1U << 22)
++
++int __init amd_special_default_mtrr(void)
++{
++ u32 l, h;
++
++ if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD)
++ return 0;
++ if (boot_cpu_data.x86 < 0xf || boot_cpu_data.x86 > 0x11)
++ return 0;
++ /* In case some hypervisor doesn't pass SYSCFG through */
++ if (rdmsr_safe(MSR_K8_SYSCFG, &l, &h) < 0)
++ return 0;
++ /*
++ * Memory between 4GB and top of mem is forced WB by this magic bit.
++ * Reserved before K8RevF, but should be zero there.
++ */
++ if ((l & (Tom2Enabled | Tom2ForceMemTypeWB)) ==
++ (Tom2Enabled | Tom2ForceMemTypeWB))
++ return 1;
++ return 0;
++}
++
++void __init mtrr_bp_init(void)
++{
++ if (amd_special_default_mtrr()) {
++ /* TOP_MEM2 */
++ rdmsrl(MSR_K8_TOP_MEM2, tom2);
++ tom2 &= 0xffffff8000000ULL;
++ }
++}
++
++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;
++
++ set_num_var_ranges();
++ init_table();
++
++ return 0;
++}
++
++subsys_initcall(mtrr_init);
+diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
+index 2685538..6bb025d 100644
+--- a/arch/x86/kernel/crash.c
++++ b/arch/x86/kernel/crash.c
+@@ -29,6 +29,7 @@
+
+ #include <mach_ipi.h>
+
++#ifndef CONFIG_XEN
+ /* This keeps a track of which one is crashing cpu. */
+ static int crashing_cpu;
+
+@@ -117,6 +118,7 @@ static void nmi_shootdown_cpus(void)
+ /* There are no cpus to shootdown */
+ }
+ #endif
++#endif /* CONFIG_XEN */
+
+ void native_machine_crash_shutdown(struct pt_regs *regs)
+ {
+@@ -131,6 +133,7 @@ void native_machine_crash_shutdown(struct pt_regs *regs)
+ /* The kernel is broken so disable interrupts */
+ local_irq_disable();
+
++#ifndef CONFIG_XEN
+ /* Make a note of crashing cpu. Will be used in NMI callback.*/
+ crashing_cpu = safe_smp_processor_id();
+ nmi_shootdown_cpus();
+@@ -138,6 +141,7 @@ void native_machine_crash_shutdown(struct pt_regs *regs)
+ #if defined(CONFIG_X86_IO_APIC)
+ disable_IO_APIC();
+ #endif
++#endif /* CONFIG_XEN */
+ #ifdef CONFIG_HPET_TIMER
+ hpet_disable();
+ #endif
+diff --git a/arch/x86/kernel/e820_32-xen.c b/arch/x86/kernel/e820_32-xen.c
+new file mode 100644
+index 0000000..99e8b4c
+--- /dev/null
++++ b/arch/x86/kernel/e820_32-xen.c
+@@ -0,0 +1,873 @@
++#include <linux/kernel.h>
++#include <linux/types.h>
++#include <linux/init.h>
++#include <linux/bootmem.h>
++#include <linux/ioport.h>
++#include <linux/string.h>
++#include <linux/kexec.h>
++#include <linux/module.h>
++#include <linux/mm.h>
++#include <linux/pfn.h>
++#include <linux/uaccess.h>
++#include <linux/suspend.h>
++
++#include <asm/pgtable.h>
++#include <asm/page.h>
++#include <asm/e820.h>
++#include <asm/setup.h>
++#include <xen/interface/memory.h>
++
++struct e820map __initdata e820;
++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;
++/* For PCI or other memory-mapped resources */
++unsigned long pci_mem_start = 0x10000000;
++#ifdef CONFIG_PCI
++EXPORT_SYMBOL(pci_mem_start);
++#endif
++extern int user_defined_memmap;
++
++static struct resource system_rom_resource = {
++ .name = "System ROM",
++ .start = 0xf0000,
++ .end = 0xfffff,
++ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
++};
++
++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
++} };
++
++static struct resource video_rom_resource = {
++ .name = "Video ROM",
++ .start = 0xc0000,
++ .end = 0xc7fff,
++ .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM
++};
++
++#define ROMSIGNATURE 0xaa55
++
++static int __init romsignature(const unsigned char *rom)
++{
++ const unsigned short * const ptr = (const unsigned short *)rom;
++ unsigned short sig;
++
++ return probe_kernel_address(ptr, sig) == 0 && sig == ROMSIGNATURE;
++}
++
++static int __init romchecksum(const unsigned char *rom, unsigned long length)
++{
++ unsigned char sum, c;
++
++ for (sum = 0; length && probe_kernel_address(rom++, c) == 0; length--)
++ sum += c;
++ return !length && !sum;
++}
++
++static void __init probe_roms(void)
++{
++ const unsigned char *rom;
++ unsigned long start, length, upper;
++ unsigned char c;
++ int i;
++
++#ifdef CONFIG_XEN
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return;
++#endif
++
++ /* video rom */
++ upper = adapter_rom_resources[0].start;
++ for (start = video_rom_resource.start; start < upper; start += 2048) {
++ rom = isa_bus_to_virt(start);
++ if (!romsignature(rom))
++ continue;
++
++ video_rom_resource.start = start;
++
++ if (probe_kernel_address(rom + 2, c) != 0)
++ continue;
++
++ /* 0 < length <= 0x7f * 512, historically */
++ length = c * 512;
++
++ /* if checksum okay, trust length byte */
++ if (length && romchecksum(rom, length))
++ video_rom_resource.end = start + length - 1;
++
++ request_resource(&iomem_resource, &video_rom_resource);
++ break;
++ }
++
++ start = (video_rom_resource.end + 1 + 2047) & ~2047UL;
++ if (start < upper)
++ start = upper;
++
++ /* system rom */
++ request_resource(&iomem_resource, &system_rom_resource);
++ upper = system_rom_resource.start;
++
++ /* check for extension rom (ignore length byte!) */
++ rom = isa_bus_to_virt((unsigned long)extension_rom_resource.start);
++ if (romsignature(rom)) {
++ length = extension_rom_resource.end - extension_rom_resource.start + 1;
++ if (romchecksum(rom, length)) {
++ request_resource(&iomem_resource, &extension_rom_resource);
++ upper = extension_rom_resource.start;
++ }
++ }
++
++ /* check for adapter roms on 2k boundaries */
++ for (i = 0; i < ARRAY_SIZE(adapter_rom_resources) && start < upper; start += 2048) {
++ rom = isa_bus_to_virt(start);
++ if (!romsignature(rom))
++ continue;
++
++ if (probe_kernel_address(rom + 2, c) != 0)
++ continue;
++
++ /* 0 < length <= 0x7f * 512, historically */
++ length = c * 512;
++
++ /* but accept any length that fits if checksum okay */
++ if (!length || start + length > upper || !romchecksum(rom, length))
++ continue;
++
++ adapter_rom_resources[i].start = start;
++ adapter_rom_resources[i].end = start + length - 1;
++ request_resource(&iomem_resource, &adapter_rom_resources[i]);
++
++ start = adapter_rom_resources[i++].end & ~2047UL;
++ }
++}
++
++#ifdef CONFIG_XEN
++static struct e820map machine_e820;
++#define e820 machine_e820
++#endif
++
++/*
++ * Request address space for all standard RAM and ROM resources
++ * and also for regions reported as reserved by the e820.
++ */
++void __init init_iomem_resources(struct resource *code_resource,
++ struct resource *data_resource,
++ struct resource *bss_resource)
++{
++ int i;
++
++ probe_roms();
++ for (i = 0; i < e820.nr_map; i++) {
++ struct resource *res;
++#ifndef CONFIG_RESOURCES_64BIT
++ if (e820.map[i].addr + e820.map[i].size > 0x100000000ULL)
++ continue;
++#endif
++ res = kzalloc(sizeof(struct resource), GFP_ATOMIC);
++ switch (e820.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->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
++ if (request_resource(&iomem_resource, res)) {
++ kfree(res);
++ continue;
++ }
++ if (e820.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);
++ request_resource(res, bss_resource);
++#endif
++#ifdef CONFIG_KEXEC
++ if (crashk_res.start != crashk_res.end)
++ request_resource(res, &crashk_res);
++#ifdef CONFIG_XEN
++ xen_machine_kexec_register_resources(res);
++#endif
++#endif
++ }
++ }
++}
++
++#undef e820
++
++#if defined(CONFIG_PM) && defined(CONFIG_HIBERNATION)
++/**
++ * e820_mark_nosave_regions - Find the ranges of physical addresses that do not
++ * correspond to e820 RAM areas and mark the corresponding pages as nosave for
++ * hibernation.
++ *
++ * 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 pfn;
++
++ pfn = PFN_DOWN(e820.map[0].addr + e820.map[0].size);
++ for (i = 1; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++
++ if (pfn < PFN_UP(ei->addr))
++ register_nosave_region(pfn, PFN_UP(ei->addr));
++
++ pfn = PFN_DOWN(ei->addr + ei->size);
++ if (ei->type != E820_RAM)
++ register_nosave_region(PFN_UP(ei->addr), pfn);
++
++ if (pfn >= max_low_pfn)
++ break;
++ }
++}
++#endif
++
++void __init add_memory_region(unsigned long long start,
++ unsigned long long size, int type)
++{
++ int x;
++
++ 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 */
++
++/*
++ * Sanitize the BIOS e820 map.
++ *
++ * Some e820 responses include overlapping entries. The following
++ * replaces the original e820 map with a new one, removing overlaps.
++ *
++ */
++int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map)
++{
++ struct change_member *change_tmp;
++ unsigned long current_type, last_type;
++ unsigned long long last_addr;
++ int chgidx, still_changing;
++ int overlap_entries;
++ int new_bios_entry;
++ int old_nr, new_nr, chg_nr;
++ int i;
++
++ /*
++ Visually we're performing the following (1,2,3,4 = memory types)...
++
++ Sample memory map (w/overlaps):
++ ____22__________________
++ ______________________4_
++ ____1111________________
++ _44_____________________
++ 11111111________________
++ ____________________33__
++ ___________44___________
++ __________33333_________
++ ______________22________
++ ___________________2222_
++ _________111111111______
++ _____________________11_
++ _________________4______
++
++ Sanitized equivalent (no overlap):
++ 1_______________________
++ _44_____________________
++ ___1____________________
++ ____22__________________
++ ______11________________
++ _________1______________
++ __________3_____________
++ ___________44___________
++ _____________33_________
++ _______________2________
++ ________________1_______
++ _________________4______
++ ___________________2____
++ ____________________33__
++ ______________________4_
++ */
++ /* if there's only one memory region, don't bother */
++ if (*pnr_map < 2) {
++ return -1;
++ }
++
++ old_nr = *pnr_map;
++
++ /* bail out if we find any unreasonable addresses in bios map */
++ for (i=0; i<old_nr; i++)
++ if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr) {
++ return -1;
++ }
++
++ /* create pointers for initial change-point information (for sorting) */
++ for (i=0; i < 2*old_nr; i++)
++ change_point[i] = &change_point_list[i];
++
++ /* record all known change-points (starting and ending addresses),
++ omitting those that are for empty memory regions */
++ chgidx = 0;
++ for (i=0; i < old_nr; i++) {
++ if (biosmap[i].size != 0) {
++ change_point[chgidx]->addr = biosmap[i].addr;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ }
++ }
++ chg_nr = chgidx; /* 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;
++ }
++ }
++ }
++
++ /* create a new bios memory map, removing overlaps */
++ overlap_entries=0; /* number of entries in the overlap table */
++ new_bios_entry=0; /* index for creating new bios map entries */
++ last_type = 0; /* start with undefined memory type */
++ last_addr = 0; /* start with 0 as last starting address */
++ /* loop through change-points, determining affect on the new bios map */
++ for (chgidx=0; chgidx < chg_nr; chgidx++)
++ {
++ /* keep track of all overlapping bios entries */
++ if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr)
++ {
++ /* add map entry to overlap list (> 1 entry implies an overlap) */
++ overlap_list[overlap_entries++]=change_point[chgidx]->pbios;
++ }
++ else
++ {
++ /* remove entry from list (order independent, so swap with last) */
++ for (i=0; i<overlap_entries; i++)
++ {
++ if (overlap_list[i] == change_point[chgidx]->pbios)
++ overlap_list[i] = overlap_list[overlap_entries-1];
++ }
++ overlap_entries--;
++ }
++ /* if there are overlapping entries, decide which "type" to use */
++ /* (larger value takes precedence -- 1=usable, 2,3,4,4+=unusable) */
++ current_type = 0;
++ for (i=0; i<overlap_entries; i++)
++ if (overlap_list[i]->type > current_type)
++ current_type = overlap_list[i]->type;
++ /* continue building up new bios map based on this information */
++ if (current_type != last_type) {
++ if (last_type != 0) {
++ new_bios[new_bios_entry].size =
++ change_point[chgidx]->addr - last_addr;
++ /* move forward only if the new size was non-zero */
++ if (new_bios[new_bios_entry].size != 0)
++ if (++new_bios_entry >= E820MAX)
++ break; /* no more space left for new bios entries */
++ }
++ if (current_type != 0) {
++ new_bios[new_bios_entry].addr = change_point[chgidx]->addr;
++ new_bios[new_bios_entry].type = current_type;
++ last_addr=change_point[chgidx]->addr;
++ }
++ last_type = current_type;
++ }
++ }
++ new_nr = new_bios_entry; /* retain count for new bios entries */
++
++ /* copy new bios mapping into original location */
++ memcpy(biosmap, new_bios, new_nr*sizeof(struct e820entry));
++ *pnr_map = new_nr;
++
++ return 0;
++}
++
++/*
++ * Copy the BIOS e820 map into a safe place.
++ *
++ * Sanity-check it while we're at it..
++ *
++ * If we're lucky and live on a modern system, the setup code
++ * will have given us a memory map that we can use to properly
++ * set up memory. If we aren't, we'll fake a memory map.
++ *
++ * We check to see that the memory map contains at least 2 elements
++ * before we'll use it, because the detection code in setup.S may
++ * not be perfect and most every PC known to man has two memory
++ * regions: one from 0 to 640k, and one from 1mb up. (The IBM
++ * thinkpad 560x, for example, does not cooperate with the memory
++ * detection code.)
++ */
++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 {
++ u64 start = biosmap->addr;
++ u64 size = biosmap->size;
++ u64 end = start + size;
++ u32 type = biosmap->type;
++
++ /* Overflow in 64 bits? Ignore the memory map. */
++ if (start > end)
++ return -1;
++
++ add_memory_region(start, size, type);
++ } while (biosmap++, --nr_map);
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ struct xen_memory_map memmap;
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ } else
++ machine_e820 = e820;
++#endif
++
++ return 0;
++}
++
++/*
++ * Find the highest page frame number we have available
++ */
++void __init propagate_e820_map(void)
++{
++ int i;
++
++ max_pfn = 0;
++
++ 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);
++ }
++}
++
++/*
++ * Register fully available low RAM pages with the bootmem allocator.
++ */
++void __init register_bootmem_low_pages(unsigned long max_low_pfn)
++{
++ int i;
++
++ 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_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;
++
++ /*
++ * .. finally, did all the rounding and playing
++ * around just make the area go away?
++ */
++ if (last_pfn <= curr_pfn)
++ continue;
++
++ size = last_pfn - curr_pfn;
++ free_bootmem(PFN_PHYS(curr_pfn), PFN_PHYS(size));
++ }
++}
++
++void __init e820_register_memory(void)
++{
++ unsigned long gapstart, gapsize, round;
++ unsigned long long last;
++ int i;
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ struct xen_memory_map memmap;
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ }
++ else
++ machine_e820 = e820;
++#define e820 machine_e820
++#endif
++
++ /*
++ * Search for the biggest gap in the low 32 bits of the e820
++ * memory space.
++ */
++ last = 0x100000000ull;
++ gapstart = 0x10000000;
++ gapsize = 0x400000;
++ i = e820.nr_map;
++ while (--i >= 0) {
++ unsigned long long start = e820.map[i].addr;
++ unsigned long long end = start + e820.map[i].size;
++
++ /*
++ * Since "last" is at most 4GB, we know we'll
++ * fit in 32 bits if this condition is true
++ */
++ if (last > end) {
++ unsigned long gap = last - end;
++
++ if (gap > gapsize) {
++ gapsize = gap;
++ gapstart = end;
++ }
++ }
++ if (start < last)
++ last = start;
++ }
++#undef e820
++
++ /*
++ * See how much we want to round up: start off with
++ * rounding to the next 1MB area.
++ */
++ round = 0x100000;
++ while ((gapsize >> 4) > round)
++ round += round;
++ /* Fun with two's complement */
++ pci_mem_start = (gapstart + round) & -round;
++
++ printk("Allocating PCI resources starting at %08lx (gap: %08lx:%08lx)\n",
++ pci_mem_start, gapstart, gapsize);
++}
++
++void __init print_memory_map(char *who)
++{
++ int i;
++
++ 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 %u\n", e820.map[i].type);
++ break;
++ }
++ }
++}
++
++void __init limit_regions(unsigned long long size)
++{
++ unsigned long long current_addr = 0;
++ int i;
++
++ print_memory_map("limit_regions start");
++ for (i = 0; i < e820.nr_map; i++) {
++ current_addr = e820.map[i].addr + e820.map[i].size;
++ if (current_addr < size)
++ continue;
++
++ if (e820.map[i].type != E820_RAM)
++ continue;
++
++ 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;
++ }
++ print_memory_map("limit_regions endfor");
++ return;
++ }
++#ifdef CONFIG_XEN
++ if (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
++ add_memory_region(current_addr, size - current_addr, E820_RAM);
++ }
++#endif
++ print_memory_map("limit_regions endfunc");
++}
++
++/*
++ * This function checks if any part of the range <start,end> is mapped
++ * with type.
++ */
++int
++e820_any_mapped(u64 start, u64 end, unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ const struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; ++i) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ return 1;
++ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(e820_any_mapped);
++
++ /*
++ * This function checks if the entire range <start,end> is mapped with type.
++ *
++ * 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)
++{
++ u64 start = s;
++ u64 end = e;
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; ++i) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ /* is the region (part) in overlap with the current region ?*/
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ /* if the region is at the beginning of <start,end> we move
++ * start to the end of the region since it's ok until there
++ */
++ if (ei->addr <= start)
++ start = ei->addr + ei->size;
++ /* if start is now at or beyond end, we're done, full
++ * coverage */
++ if (start >= end)
++ return 1; /* we're done */
++ }
++ return 0;
++}
++
++static int __init parse_memmap(char *arg)
++{
++ if (!arg)
++ return -EINVAL;
++
++ if (strcmp(arg, "exactmap") == 0) {
++#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.
++ */
++ propagate_e820_map();
++ saved_max_pfn = max_pfn;
++#endif
++ e820.nr_map = 0;
++ user_defined_memmap = 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(arg, &arg);
++ if (*arg == '@') {
++ start_at = memparse(arg+1, &arg);
++ add_memory_region(start_at, mem_size, E820_RAM);
++ } else if (*arg == '#') {
++ start_at = memparse(arg+1, &arg);
++ add_memory_region(start_at, mem_size, E820_ACPI);
++ } else if (*arg == '$') {
++ start_at = memparse(arg+1, &arg);
++ add_memory_region(start_at, mem_size, E820_RESERVED);
++ } else {
++ limit_regions(mem_size);
++ user_defined_memmap = 1;
++ }
++ }
++ return 0;
++}
++early_param("memmap", parse_memmap);
++
++#ifndef CONFIG_XEN
++void __init update_memory_range(u64 start, u64 size, unsigned old_type,
++ unsigned new_type)
++{
++ int i;
++
++ BUG_ON(old_type == new_type);
++
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ u64 final_start, final_end;
++ if (ei->type != old_type)
++ continue;
++ /* totally covered? */
++ if (ei->addr >= start && ei->size <= size) {
++ ei->type = new_type;
++ continue;
++ }
++ /* partially covered */
++ final_start = max(start, ei->addr);
++ final_end = min(start + size, ei->addr + ei->size);
++ if (final_start >= final_end)
++ continue;
++ add_memory_region(final_start, final_end - final_start,
++ new_type);
++ }
++}
++
++void __init update_e820(void)
++{
++ u8 nr_map;
++
++ nr_map = e820.nr_map;
++ if (sanitize_e820_map(e820.map, &nr_map))
++ return;
++ e820.nr_map = nr_map;
++ printk(KERN_INFO "modified physical RAM map:\n");
++ print_memory_map("modified");
++}
++#endif
+diff --git a/arch/x86/kernel/e820_64-xen.c b/arch/x86/kernel/e820_64-xen.c
+new file mode 100644
+index 0000000..371c948
+--- /dev/null
++++ b/arch/x86/kernel/e820_64-xen.c
+@@ -0,0 +1,1045 @@
++/*
++ * Handle the memory map.
++ * The functions here do the job until bootmem takes over.
++ *
++ * Getting sanitize_e820_map() in sync with i386 version by applying change:
++ * - Provisions for empty E820 memory regions (reported by certain BIOSes).
++ * Alex Achenbach <xela at slit.de>, December 2002.
++ * Venkatesh Pallipadi <venkatesh.pallipadi at intel.com>
++ *
++ */
++#include <linux/kernel.h>
++#include <linux/types.h>
++#include <linux/init.h>
++#include <linux/bootmem.h>
++#include <linux/ioport.h>
++#include <linux/string.h>
++#include <linux/kexec.h>
++#include <linux/module.h>
++#include <linux/mm.h>
++#include <linux/suspend.h>
++#include <linux/pfn.h>
++
++#include <asm/pgtable.h>
++#include <asm/page.h>
++#include <asm/e820.h>
++#include <asm/proto.h>
++#include <asm/setup.h>
++#include <asm/sections.h>
++#include <asm/kdebug.h>
++#include <xen/interface/memory.h>
++
++struct e820map e820 __initdata;
++#ifdef CONFIG_XEN
++struct e820map machine_e820;
++#endif
++
++/*
++ * PFN of last memory page.
++ */
++unsigned long end_pfn;
++
++/*
++ * end_pfn only includes RAM, while max_pfn_mapped includes all e820 entries.
++ * The direct mapping extends to max_pfn_mapped, so that we can directly access
++ * apertures, ACPI and other tables without having to play with fixmaps.
++ */
++unsigned long max_pfn_mapped;
++
++/*
++ * Last pfn which the user wants to use.
++ */
++static unsigned long __initdata end_user_pfn = MAXMEM>>PAGE_SHIFT;
++
++/*
++ * Early reserved memory areas.
++ */
++#define MAX_EARLY_RES 20
++
++struct early_res {
++ unsigned long start, end;
++ char name[16];
++};
++static struct early_res early_res[MAX_EARLY_RES] __initdata = {
++#ifndef CONFIG_XEN
++ { 0, PAGE_SIZE, "BIOS data page" }, /* BIOS data page */
++#ifdef CONFIG_X86_TRAMPOLINE
++ { TRAMPOLINE_BASE, TRAMPOLINE_BASE + 2 * PAGE_SIZE, "TRAMPOLINE" },
++#endif
++#endif
++ {}
++};
++
++void __init reserve_early(unsigned long start, unsigned long end, char *name)
++{
++ int i;
++ struct early_res *r;
++ for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
++ r = &early_res[i];
++ if (end > r->start && start < r->end)
++ panic("Overlapping early reservations %lx-%lx %s to %lx-%lx %s\n",
++ start, end - 1, name?name:"", r->start, r->end - 1, r->name);
++ }
++ if (i >= MAX_EARLY_RES)
++ panic("Too many early reservations");
++ r = &early_res[i];
++ r->start = start;
++ r->end = end;
++ if (name)
++ strncpy(r->name, name, sizeof(r->name) - 1);
++}
++
++void __init free_early(unsigned long start, unsigned long end)
++{
++ struct early_res *r;
++ int i, j;
++
++ for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
++ r = &early_res[i];
++ if (start == r->start && end == r->end)
++ break;
++ }
++ if (i >= MAX_EARLY_RES || !early_res[i].end)
++ panic("free_early on not reserved area: %lx-%lx!", start, end);
++
++ for (j = i + 1; j < MAX_EARLY_RES && early_res[j].end; j++)
++ ;
++
++ memmove(&early_res[i], &early_res[i + 1],
++ (j - 1 - i) * sizeof(struct early_res));
++
++ early_res[j - 1].end = 0;
++}
++
++void __init early_res_to_bootmem(unsigned long start, unsigned long end)
++{
++ int i;
++ unsigned long final_start, final_end;
++ for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
++ struct early_res *r = &early_res[i];
++ final_start = max(start, r->start);
++ final_end = min(end, r->end);
++ if (final_start >= final_end)
++ continue;
++ printk(KERN_INFO " early res: %d [%lx-%lx] %s\n", i,
++ final_start, final_end - 1, r->name);
++ reserve_bootmem_generic(final_start, final_end - final_start);
++ }
++}
++
++/* Check for already reserved areas */
++static inline int __init
++bad_addr(unsigned long *addrp, unsigned long size, unsigned long align)
++{
++ int i;
++ unsigned long addr = *addrp, last;
++ int changed = 0;
++again:
++ last = addr + size;
++ for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
++ struct early_res *r = &early_res[i];
++ if (last >= r->start && addr < r->end) {
++ *addrp = addr = round_up(r->end, align);
++ changed = 1;
++ goto again;
++ }
++ }
++ return changed;
++}
++
++/* Check for already reserved areas */
++static inline int __init
++bad_addr_size(unsigned long *addrp, unsigned long *sizep, unsigned long align)
++{
++ int i;
++ unsigned long addr = *addrp, last;
++ unsigned long size = *sizep;
++ int changed = 0;
++again:
++ last = addr + size;
++ for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
++ struct early_res *r = &early_res[i];
++ if (last > r->start && addr < r->start) {
++ size = r->start - addr;
++ changed = 1;
++ goto again;
++ }
++ if (last > r->end && addr < r->end) {
++ addr = round_up(r->end, align);
++ size = last - addr;
++ changed = 1;
++ goto again;
++ }
++ if (last <= r->end && addr >= r->start) {
++ (*sizep)++;
++ return 0;
++ }
++ }
++ if (changed) {
++ *addrp = addr;
++ *sizep = size;
++ }
++ return changed;
++}
++/*
++ * This function checks if any part of the range <start,end> is mapped
++ * with type.
++ */
++int
++e820_any_mapped(unsigned long start, unsigned long end, unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; i++) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++ return 1;
++ }
++ return 0;
++}
++EXPORT_SYMBOL_GPL(e820_any_mapped);
++
++/*
++ * This function checks if the entire range <start,end> is mapped with type.
++ *
++ * Note: this function only works correct if the e820 table is sorted and
++ * not-overlapping, which is the case
++ */
++int __init e820_all_mapped(unsigned long start, unsigned long end,
++ unsigned type)
++{
++ int i;
++
++#ifndef CONFIG_XEN
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++#else
++ if (!is_initial_xendomain())
++ return 0;
++ for (i = 0; i < machine_e820.nr_map; i++) {
++ const struct e820entry *ei = &machine_e820.map[i];
++#endif
++
++ if (type && ei->type != type)
++ continue;
++ /* is the region (part) in overlap with the current region ?*/
++ if (ei->addr >= end || ei->addr + ei->size <= start)
++ continue;
++
++ /* if the region is at the beginning of <start,end> we move
++ * start to the end of the region since it's ok until there
++ */
++ if (ei->addr <= start)
++ start = ei->addr + ei->size;
++ /*
++ * if start is now at or beyond end, we're done, full
++ * coverage
++ */
++ if (start >= end)
++ return 1;
++ }
++ return 0;
++}
++
++/*
++ * Find a free area with specified alignment in a specific range.
++ */
++unsigned long __init find_e820_area(unsigned long start, unsigned long end,
++ unsigned long size, unsigned long align)
++{
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long addr, last;
++ unsigned long ei_last;
++
++ if (ei->type != E820_RAM)
++ continue;
++ addr = round_up(ei->addr, align);
++ ei_last = ei->addr + ei->size;
++ if (addr < start)
++ addr = round_up(start, align);
++ if (addr >= ei_last)
++ continue;
++ while (bad_addr(&addr, size, align) && addr+size <= ei_last)
++ ;
++ last = addr + size;
++ if (last > ei_last)
++ continue;
++ if (last > end)
++ continue;
++ return addr;
++ }
++ return -1UL;
++}
++
++/*
++ * Find next free range after *start
++ */
++unsigned long __init find_e820_area_size(unsigned long start,
++ unsigned long *sizep,
++ unsigned long align)
++{
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ unsigned long addr, last;
++ unsigned long ei_last;
++
++ if (ei->type != E820_RAM)
++ continue;
++ addr = round_up(ei->addr, align);
++ ei_last = ei->addr + ei->size;
++ if (addr < start)
++ addr = round_up(start, align);
++ if (addr >= ei_last)
++ continue;
++ *sizep = ei_last - addr;
++ while (bad_addr_size(&addr, sizep, align) &&
++ addr + *sizep <= ei_last)
++ ;
++ last = addr + *sizep;
++ if (last > ei_last)
++ continue;
++ return addr;
++ }
++ return -1UL;
++
++}
++/*
++ * Find the highest page frame number we have available
++ */
++unsigned long __init e820_end_of_ram(void)
++{
++ unsigned long end_pfn;
++
++ end_pfn = find_max_pfn_with_active_regions();
++
++ if (end_pfn > max_pfn_mapped)
++ max_pfn_mapped = end_pfn;
++ if (max_pfn_mapped > MAXMEM>>PAGE_SHIFT)
++ max_pfn_mapped = MAXMEM>>PAGE_SHIFT;
++ if (end_pfn > end_user_pfn)
++ end_pfn = end_user_pfn;
++ if (end_pfn > max_pfn_mapped)
++ end_pfn = max_pfn_mapped;
++
++ printk(KERN_INFO "max_pfn_mapped = %lu\n", max_pfn_mapped);
++ return end_pfn;
++}
++
++/*
++ * Mark e820 reserved areas as busy for the resource manager.
++ */
++void __init e820_reserve_resources(struct e820entry *e820, int nr_map)
++{
++ int i;
++ struct resource *res;
++
++ res = alloc_bootmem_low(sizeof(struct resource) * nr_map);
++ for (i = 0; i < nr_map; i++) {
++ switch (e820[i].type) {
++ case E820_RAM: res->name = "System RAM"; break;
++ case E820_ACPI: res->name = "ACPI Tables"; break;
++ case E820_NVS: res->name = "ACPI Non-volatile Storage"; break;
++ default: res->name = "reserved";
++ }
++ res->start = e820[i].addr;
++ res->end = res->start + e820[i].size - 1;
++ res->flags = IORESOURCE_MEM | IORESOURCE_BUSY;
++ insert_resource(&iomem_resource, res);
++ res++;
++ }
++}
++
++#ifndef CONFIG_XEN
++/*
++ * 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)
++ register_nosave_region(PFN_DOWN(paddr),
++ PFN_UP(ei->addr));
++
++ paddr = round_down(ei->addr + ei->size, PAGE_SIZE);
++ if (ei->type != E820_RAM)
++ register_nosave_region(PFN_UP(ei->addr),
++ PFN_DOWN(paddr));
++
++ if (paddr >= (end_pfn << PAGE_SHIFT))
++ break;
++ }
++}
++#endif
++
++/*
++ * Finds an active region in the address range from start_pfn to end_pfn and
++ * returns its range in ei_startpfn and ei_endpfn for the e820 entry.
++ */
++static int __init e820_find_active_region(const struct e820entry *ei,
++ unsigned long start_pfn,
++ unsigned long end_pfn,
++ unsigned long *ei_startpfn,
++ unsigned long *ei_endpfn)
++{
++ *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)
++ return 0;
++
++ /* Check if max_pfn_mapped should be updated */
++ if (ei->type != E820_RAM && *ei_endpfn > max_pfn_mapped)
++ max_pfn_mapped = *ei_endpfn;
++
++ /* Skip if map is outside the node */
++ if (ei->type != E820_RAM || *ei_endpfn <= start_pfn ||
++ *ei_startpfn >= end_pfn)
++ return 0;
++
++ /* 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)
++ return 0;
++ if (*ei_endpfn > end_user_pfn)
++ *ei_endpfn = end_user_pfn;
++
++ return 1;
++}
++
++/* 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)
++{
++ unsigned long ei_startpfn;
++ unsigned long ei_endpfn;
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++)
++ if (e820_find_active_region(&e820.map[i],
++ start_pfn, end_pfn,
++ &ei_startpfn, &ei_endpfn))
++ add_active_range(nid, ei_startpfn, ei_endpfn);
++}
++
++/*
++ * Add a memory region to the kernel e820 map.
++ */
++void __init add_memory_region(unsigned long start, unsigned long size, int type)
++{
++ int x = e820.nr_map;
++
++ if (x == E820MAX) {
++ printk(KERN_ERR "Ooops! Too many entries in the memory map!\n");
++ return;
++ }
++
++ e820.map[x].addr = start;
++ e820.map[x].size = size;
++ e820.map[x].type = type;
++ e820.nr_map++;
++}
++
++/*
++ * Find the hole size (in bytes) in the memory range.
++ * @start: starting address of the memory range to scan
++ * @end: ending address of the memory range to scan
++ */
++unsigned long __init e820_hole_size(unsigned long start, unsigned long end)
++{
++ unsigned long start_pfn = start >> PAGE_SHIFT;
++ unsigned long end_pfn = end >> PAGE_SHIFT;
++ unsigned long ei_startpfn, ei_endpfn, ram = 0;
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ if (e820_find_active_region(&e820.map[i],
++ start_pfn, end_pfn,
++ &ei_startpfn, &ei_endpfn))
++ ram += ei_endpfn - ei_startpfn;
++ }
++ return end - start - (ram << PAGE_SHIFT);
++}
++
++static void __init e820_print_map(char *who)
++{
++ int i;
++
++ for (i = 0; i < e820.nr_map; i++) {
++ printk(KERN_INFO " %s: %016Lx - %016Lx ", who,
++ (unsigned long long) e820.map[i].addr,
++ (unsigned long long)
++ (e820.map[i].addr + e820.map[i].size));
++ switch (e820.map[i].type) {
++ case E820_RAM:
++ printk(KERN_CONT "(usable)\n");
++ break;
++ case E820_RESERVED:
++ printk(KERN_CONT "(reserved)\n");
++ break;
++ case E820_ACPI:
++ printk(KERN_CONT "(ACPI data)\n");
++ break;
++ case E820_NVS:
++ printk(KERN_CONT "(ACPI NVS)\n");
++ break;
++ default:
++ printk(KERN_CONT "type %u\n", e820.map[i].type);
++ break;
++ }
++ }
++}
++
++/*
++ * Sanitize the BIOS e820 map.
++ *
++ * Some e820 responses include overlapping entries. The following
++ * replaces the original e820 map with a new one, removing overlaps.
++ *
++ */
++static int __init sanitize_e820_map(struct e820entry *biosmap, char *pnr_map)
++{
++ struct change_member {
++ struct e820entry *pbios; /* pointer to original bios entry */
++ unsigned long long addr; /* address for this change point */
++ };
++ static struct change_member change_point_list[2*E820MAX] __initdata;
++ static struct change_member *change_point[2*E820MAX] __initdata;
++ static struct e820entry *overlap_list[E820MAX] __initdata;
++ static struct e820entry new_bios[E820MAX] __initdata;
++ struct change_member *change_tmp;
++ unsigned long current_type, last_type;
++ unsigned long long last_addr;
++ int chgidx, still_changing;
++ int overlap_entries;
++ int new_bios_entry;
++ int old_nr, new_nr, chg_nr;
++ int i;
++
++ /*
++ Visually we're performing the following
++ (1,2,3,4 = memory types)...
++
++ Sample memory map (w/overlaps):
++ ____22__________________
++ ______________________4_
++ ____1111________________
++ _44_____________________
++ 11111111________________
++ ____________________33__
++ ___________44___________
++ __________33333_________
++ ______________22________
++ ___________________2222_
++ _________111111111______
++ _____________________11_
++ _________________4______
++
++ Sanitized equivalent (no overlap):
++ 1_______________________
++ _44_____________________
++ ___1____________________
++ ____22__________________
++ ______11________________
++ _________1______________
++ __________3_____________
++ ___________44___________
++ _____________33_________
++ _______________2________
++ ________________1_______
++ _________________4______
++ ___________________2____
++ ____________________33__
++ ______________________4_
++ */
++
++ /* if there's only one memory region, don't bother */
++ if (*pnr_map < 2)
++ return -1;
++
++ old_nr = *pnr_map;
++
++ /* bail out if we find any unreasonable addresses in bios map */
++ for (i = 0; i < old_nr; i++)
++ if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr)
++ return -1;
++
++ /* create pointers for initial change-point information (for sorting) */
++ for (i = 0; i < 2 * old_nr; i++)
++ change_point[i] = &change_point_list[i];
++
++ /* record all known change-points (starting and ending addresses),
++ omitting those that are for empty memory regions */
++ chgidx = 0;
++ for (i = 0; i < old_nr; i++) {
++ if (biosmap[i].size != 0) {
++ change_point[chgidx]->addr = biosmap[i].addr;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ change_point[chgidx]->addr = biosmap[i].addr +
++ biosmap[i].size;
++ change_point[chgidx++]->pbios = &biosmap[i];
++ }
++ }
++ chg_nr = chgidx;
++
++ /* sort change-point list by memory addresses (low -> high) */
++ still_changing = 1;
++ while (still_changing) {
++ still_changing = 0;
++ for (i = 1; i < chg_nr; i++) {
++ unsigned long long curaddr, lastaddr;
++ unsigned long long curpbaddr, lastpbaddr;
++
++ curaddr = change_point[i]->addr;
++ lastaddr = change_point[i - 1]->addr;
++ curpbaddr = change_point[i]->pbios->addr;
++ lastpbaddr = change_point[i - 1]->pbios->addr;
++
++ /*
++ * swap entries, when:
++ *
++ * curaddr > lastaddr or
++ * curaddr == lastaddr and curaddr == curpbaddr and
++ * lastaddr != lastpbaddr
++ */
++ if (curaddr < lastaddr ||
++ (curaddr == lastaddr && curaddr == curpbaddr &&
++ lastaddr != lastpbaddr)) {
++ change_tmp = change_point[i];
++ change_point[i] = change_point[i-1];
++ change_point[i-1] = change_tmp;
++ still_changing = 1;
++ }
++ }
++ }
++
++ /* create a new bios memory map, removing overlaps */
++ overlap_entries = 0; /* number of entries in the overlap table */
++ new_bios_entry = 0; /* index for creating new bios map entries */
++ last_type = 0; /* start with undefined memory type */
++ last_addr = 0; /* start with 0 as last starting address */
++
++ /* loop through change-points, determining affect on the new bios map */
++ for (chgidx = 0; chgidx < chg_nr; chgidx++) {
++ /* keep track of all overlapping bios entries */
++ if (change_point[chgidx]->addr ==
++ change_point[chgidx]->pbios->addr) {
++ /*
++ * add map entry to overlap list (> 1 entry
++ * implies an overlap)
++ */
++ overlap_list[overlap_entries++] =
++ change_point[chgidx]->pbios;
++ } else {
++ /*
++ * remove entry from list (order independent,
++ * so swap with last)
++ */
++ for (i = 0; i < overlap_entries; i++) {
++ if (overlap_list[i] ==
++ change_point[chgidx]->pbios)
++ overlap_list[i] =
++ overlap_list[overlap_entries-1];
++ }
++ overlap_entries--;
++ }
++ /*
++ * if there are overlapping entries, decide which
++ * "type" to use (larger value takes precedence --
++ * 1=usable, 2,3,4,4+=unusable)
++ */
++ current_type = 0;
++ for (i = 0; i < overlap_entries; i++)
++ if (overlap_list[i]->type > current_type)
++ current_type = overlap_list[i]->type;
++ /*
++ * continue building up new bios map based on this
++ * information
++ */
++ if (current_type != last_type) {
++ if (last_type != 0) {
++ new_bios[new_bios_entry].size =
++ change_point[chgidx]->addr - last_addr;
++ /*
++ * move forward only if the new size
++ * was non-zero
++ */
++ if (new_bios[new_bios_entry].size != 0)
++ /*
++ * no more space left for new
++ * bios entries ?
++ */
++ if (++new_bios_entry >= E820MAX)
++ break;
++ }
++ 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;
++ }
++ }
++ /* retain count for new bios entries */
++ new_nr = new_bios_entry;
++
++ /* copy new bios mapping into original location */
++ memcpy(biosmap, new_bios, new_nr * sizeof(struct e820entry));
++ *pnr_map = new_nr;
++
++ return 0;
++}
++
++/*
++ * Copy the BIOS e820 map into a safe place.
++ *
++ * Sanity-check it while we're at it..
++ *
++ * If we're lucky and live on a modern system, the setup code
++ * will have given us a memory map that we can use to properly
++ * set up memory. If we aren't, we'll fake a memory map.
++ */
++static int __init copy_e820_map(struct e820entry *biosmap, int nr_map)
++{
++#ifndef CONFIG_XEN
++ /* Only one memory region (or negative)? Ignore it */
++ if (nr_map < 2)
++ return -1;
++#else
++ BUG_ON(nr_map < 1);
++#endif
++
++ do {
++ u64 start = biosmap->addr;
++ u64 size = biosmap->size;
++ u64 end = start + size;
++ u32 type = biosmap->type;
++
++ /* Overflow in 64 bits? Ignore the memory map. */
++ if (start > end)
++ return -1;
++
++ add_memory_region(start, size, type);
++ } while (biosmap++, --nr_map);
++
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain()) {
++ struct xen_memory_map memmap;
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, machine_e820.map);
++
++ if (HYPERVISOR_memory_op(XENMEM_machine_memory_map, &memmap))
++ BUG();
++ machine_e820.nr_map = memmap.nr_entries;
++ } else
++ machine_e820 = e820;
++#endif
++
++ return 0;
++}
++
++static void early_panic(char *msg)
++{
++ early_printk(msg);
++ panic(msg);
++}
++
++/* We're not void only for x86 32-bit compat */
++char * __init machine_specific_memory_setup(void)
++{
++#ifndef CONFIG_XEN
++ char *who = "BIOS-e820";
++ /*
++ * Try to copy the BIOS-supplied E820-map.
++ *
++ * Otherwise fake a memory map; one section from 0k->640k,
++ * the next section from 1mb->appropriate_mem_k
++ */
++ sanitize_e820_map(boot_params.e820_map, &boot_params.e820_entries);
++ if (copy_e820_map(boot_params.e820_map, boot_params.e820_entries) < 0)
++ early_panic("Cannot find a valid memory map");
++#else /* CONFIG_XEN */
++ char *who = "Xen";
++ int rc;
++ struct xen_memory_map memmap;
++ /*
++ * This is rather large for a stack variable but this early in
++ * the boot process we know we have plenty slack space.
++ */
++ struct e820entry map[E820MAX];
++
++ memmap.nr_entries = E820MAX;
++ set_xen_guest_handle(memmap.buffer, map);
++
++ rc = HYPERVISOR_memory_op(XENMEM_memory_map, &memmap);
++ if ( rc == -ENOSYS ) {
++ memmap.nr_entries = 1;
++ map[0].addr = 0ULL;
++ map[0].size = xen_start_info->nr_pages << PAGE_SHIFT;
++ /* 8MB slack (to balance backend allocations). */
++ map[0].size += 8 << 20;
++ map[0].type = E820_RAM;
++ rc = 0;
++ }
++ BUG_ON(rc);
++
++ sanitize_e820_map(map, (char *)&memmap.nr_entries);
++
++ if (copy_e820_map(map, (char)memmap.nr_entries) < 0)
++ early_panic("Cannot find a valid memory map");
++#endif
++ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
++ e820_print_map(who);
++
++ /* In case someone cares... */
++ return who;
++}
++
++static int __init parse_memopt(char *p)
++{
++ int i;
++ unsigned long current_end;
++ unsigned long end;
++
++ if (!p)
++ return -EINVAL;
++ end_user_pfn = memparse(p, &p);
++ end_user_pfn >>= PAGE_SHIFT;
++
++ end = end_user_pfn<<PAGE_SHIFT;
++ i = e820.nr_map-1;
++ current_end = e820.map[i].addr + e820.map[i].size;
++
++ if (current_end < end) {
++ /*
++ * The e820 map ends before our requested size so
++ * extend the final entry to the requested address.
++ */
++ if (e820.map[i].type == E820_RAM)
++ e820.map[i].size = end - e820.map[i].addr;
++ else
++ add_memory_region(current_end, end - current_end, E820_RAM);
++ }
++
++ return 0;
++}
++early_param("mem", parse_memopt);
++
++static int userdef __initdata;
++
++static int __init parse_memmap_opt(char *p)
++{
++ char *oldp;
++ unsigned long long start_at, mem_size;
++
++ 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
++ max_pfn_mapped = 0;
++ e820.nr_map = 0;
++ userdef = 1;
++ return 0;
++ }
++
++ oldp = p;
++ mem_size = memparse(p, &p);
++ if (p == oldp)
++ return -EINVAL;
++
++ userdef = 1;
++ if (*p == '@') {
++ start_at = memparse(p+1, &p);
++ add_memory_region(start_at, mem_size, E820_RAM);
++ } else if (*p == '#') {
++ start_at = memparse(p+1, &p);
++ add_memory_region(start_at, mem_size, E820_ACPI);
++ } else if (*p == '$') {
++ start_at = memparse(p+1, &p);
++ add_memory_region(start_at, mem_size, E820_RESERVED);
++ } else {
++ end_user_pfn = (mem_size >> PAGE_SHIFT);
++ }
++ return *p == '\0' ? 0 : -EINVAL;
++}
++early_param("memmap", parse_memmap_opt);
++
++void __init finish_e820_parsing(void)
++{
++ if (userdef) {
++ char nr = e820.nr_map;
++
++ if (sanitize_e820_map(e820.map, &nr) < 0)
++ early_panic("Invalid user supplied memory map");
++ e820.nr_map = nr;
++
++ printk(KERN_INFO "user-defined physical RAM map:\n");
++ e820_print_map("user");
++ }
++}
++
++#ifndef CONFIG_XEN
++void __init update_memory_range(u64 start, u64 size, unsigned old_type,
++ unsigned new_type)
++{
++ int i;
++
++ BUG_ON(old_type == new_type);
++
++ for (i = 0; i < e820.nr_map; i++) {
++ struct e820entry *ei = &e820.map[i];
++ u64 final_start, final_end;
++ if (ei->type != old_type)
++ continue;
++ /* totally covered? */
++ if (ei->addr >= start && ei->size <= size) {
++ ei->type = new_type;
++ continue;
++ }
++ /* partially covered */
++ final_start = max(start, ei->addr);
++ final_end = min(start + size, ei->addr + ei->size);
++ if (final_start >= final_end)
++ continue;
++ add_memory_region(final_start, final_end - final_start,
++ new_type);
++ }
++}
++
++void __init update_e820(void)
++{
++ u8 nr_map;
++
++ nr_map = e820.nr_map;
++ if (sanitize_e820_map(e820.map, &nr_map))
++ return;
++ e820.nr_map = nr_map;
++ printk(KERN_INFO "modified physical RAM map:\n");
++ e820_print_map("modified");
++}
++#endif
++
++unsigned long pci_mem_start = 0xaeedbabe;
++EXPORT_SYMBOL(pci_mem_start);
++
++/*
++ * Search for the biggest gap in the low 32 bits of the e820
++ * memory space. We pass this space to PCI to assign MMIO resources
++ * for hotplug or unconfigured devices in.
++ * Hopefully the BIOS let enough space left.
++ */
++__init void e820_setup_gap(struct e820entry *e820, int nr_map)
++{
++ unsigned long gapstart, gapsize, round;
++ unsigned long last;
++ int i;
++ int found = 0;
++
++ last = 0x100000000ull;
++ gapstart = 0x10000000;
++ gapsize = 0x400000;
++ i = nr_map;
++ while (--i >= 0) {
++ unsigned long long start = e820[i].addr;
++ unsigned long long end = start + e820[i].size;
++
++ /*
++ * Since "last" is at most 4GB, we know we'll
++ * fit in 32 bits if this condition is true
++ */
++ if (last > end) {
++ unsigned long gap = last - end;
++
++ if (gap > gapsize) {
++ gapsize = gap;
++ gapstart = end;
++ found = 1;
++ }
++ }
++ if (start < last)
++ last = start;
++ }
++
++ if (!found) {
++ gapstart = (end_pfn << PAGE_SHIFT) + 1024*1024;
++ printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit "
++ "address range\n"
++ KERN_ERR "PCI: Unassigned devices with 32bit resource "
++ "registers may break!\n");
++ }
++
++ /*
++ * See how much we want to round up: start off with
++ * rounding to the next 1MB area.
++ */
++ round = 0x100000;
++ while ((gapsize >> 4) > round)
++ round += round;
++ /* Fun with two's complement */
++ pci_mem_start = (gapstart + round) & -round;
++
++ printk(KERN_INFO
++ "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n",
++ pci_mem_start, gapstart, gapsize);
++}
++
++int __init arch_get_ram_range(int slot, u64 *addr, u64 *size)
++{
++ int i;
++
++ if (slot < 0 || slot >= e820.nr_map)
++ return -1;
++ for (i = slot; i < e820.nr_map; i++) {
++ if (e820.map[i].type != E820_RAM)
++ continue;
++ break;
++ }
++ if (i == e820.nr_map || e820.map[i].addr > (max_pfn << PAGE_SHIFT))
++ return -1;
++ *addr = e820.map[i].addr;
++ *size = min_t(u64, e820.map[i].size + e820.map[i].addr,
++ max_pfn << PAGE_SHIFT) - *addr;
++ return i + 1;
++}
+diff --git a/arch/x86/kernel/early_printk-xen.c b/arch/x86/kernel/early_printk-xen.c
+new file mode 100644
+index 0000000..72d0463
+--- /dev/null
++++ b/arch/x86/kernel/early_printk-xen.c
+@@ -0,0 +1,285 @@
++#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>
++#include <asm/setup.h>
++
++/* Simple VGA output */
++#define VGABASE (__ISA_IO_base + 0xb8000)
++
++#ifndef CONFIG_XEN
++static int max_ypos = 25, max_xpos = 80;
++static int current_ypos = 25, current_xpos;
++
++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) {
++ if (*s == '\n')
++ early_serial_putc('\r');
++ early_serial_putc(*s);
++ s++;
++ }
++}
++
++#define DEFAULT_BAUD 9600
++
++static __init void early_serial_init(char *s)
++{
++ unsigned char c;
++ unsigned divisor;
++ unsigned baud = DEFAULT_BAUD;
++ char *e;
++
++ if (*s == ',')
++ ++s;
++
++ if (*s) {
++ unsigned port;
++ if (!strncmp(s, "0x", 2)) {
++ early_serial_base = simple_strtoul(s, &e, 16);
++ } else {
++ static int bases[] = { 0x3f8, 0x2f8 };
++
++ if (!strncmp(s, "ttyS", 4))
++ s += 4;
++ port = simple_strtoul(s, &e, 10);
++ if (port > 1 || s == e)
++ port = 0;
++ early_serial_base = bases[port];
++ }
++ s += strcspn(s, ",");
++ if (*s == ',')
++ s++;
++ }
++
++ outb(0x3, early_serial_base + LCR); /* 8n1 */
++ outb(0, early_serial_base + IER); /* no interrupt */
++ outb(0, early_serial_base + FCR); /* no fifo */
++ outb(0x3, early_serial_base + MCR); /* DTR + RTS */
++
++ if (*s) {
++ baud = simple_strtoul(s, &e, 0);
++ if (baud == 0 || s == e)
++ baud = DEFAULT_BAUD;
++ }
++
++ divisor = 115200 / baud;
++ c = inb(early_serial_base + LCR);
++ outb(c | DLAB, early_serial_base + LCR);
++ outb(divisor & 0xff, early_serial_base + DLL);
++ outb((divisor >> 8) & 0xff, early_serial_base + DLH);
++ outb(c & ~DLAB, early_serial_base + LCR);
++}
++
++#else /* CONFIG_XEN */
++
++static void
++early_serial_write(struct console *con, const char *s, unsigned count)
++{
++ int n;
++
++ while (count > 0) {
++ n = HYPERVISOR_console_io(CONSOLEIO_write, count, (char *)s);
++ if (n <= 0)
++ break;
++ count -= n;
++ s += n;
++ }
++}
++
++static __init void early_serial_init(char *s)
++{
++}
++
++/*
++ * No early VGA console on Xen, as we do not have convenient ISA-space
++ * mappings. Someone should fix this for domain 0. For now, use fake serial.
++ */
++#define early_vga_console early_serial_console
++#define xenboot_console early_serial_console
++
++#endif
++
++static struct console early_serial_console = {
++ .name = "earlyser",
++ .write = early_serial_write,
++ .flags = CON_PRINTBUFFER,
++ .index = -1,
++};
++
++/* Console interface to a host file on AMD's SimNow! */
++
++static int simnow_fd;
++
++enum {
++ MAGIC1 = 0xBACCD00A,
++ MAGIC2 = 0xCA110000,
++ XOPEN = 5,
++ XWRITE = 4,
++};
++
++static noinline long simnow(long cmd, long a, long b, long c)
++{
++ long ret;
++ asm volatile("cpuid" :
++ "=a" (ret) :
++ "b" (a), "c" (b), "d" (c), "0" (MAGIC1), "D" (cmd + MAGIC2));
++ return ret;
++}
++
++static 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 */
++static struct console *early_console = &early_vga_console;
++static int early_console_initialized;
++
++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;
++
++static int __init setup_early_printk(char *buf)
++{
++ if (!buf)
++ return 0;
++
++ if (early_console_initialized)
++ return 0;
++ early_console_initialized = 1;
++
++ if (strstr(buf, "keep"))
++ keep_early = 1;
++
++ if (!strncmp(buf, "serial", 6)) {
++ early_serial_init(buf + 6);
++ early_console = &early_serial_console;
++ } else if (!strncmp(buf, "ttyS", 4)) {
++ early_serial_init(buf);
++ early_console = &early_serial_console;
++ } else if (!strncmp(buf, "vga", 3)) {
++#ifndef CONFIG_XEN
++ && boot_params.screen_info.orig_video_isVGA == 1) {
++ max_xpos = boot_params.screen_info.orig_video_cols;
++ max_ypos = boot_params.screen_info.orig_video_lines;
++ current_ypos = boot_params.screen_info.orig_y;
++#endif
++ early_console = &early_vga_console;
++ } else if (!strncmp(buf, "simnow", 6)) {
++ simnow_init(buf + 6);
++ early_console = &simnow_console;
++ keep_early = 1;
++#ifdef CONFIG_XEN
++ } else if (!strncmp(buf, "xen", 3)) {
++ early_console = &xenboot_console;
++#endif
++ }
++
++ if (keep_early)
++ early_console->flags &= ~CON_BOOT;
++ else
++ early_console->flags |= CON_BOOT;
++ register_console(early_console);
++ return 0;
++}
++early_param("earlyprintk", setup_early_printk);
+diff --git a/arch/x86/kernel/entry_32-xen.S b/arch/x86/kernel/entry_32-xen.S
+new file mode 100644
+index 0000000..ca66938
+--- /dev/null
++++ b/arch/x86/kernel/entry_32-xen.S
+@@ -0,0 +1,1404 @@
++/*
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ */
++
++/*
++ * 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 'syscall_exit':
++ * 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) - %fs
++ * 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.
++ */
++
++#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 <asm/processor-flags.h>
++#include "irq_vectors.h"
++#include <xen/interface/xen.h>
++
++/*
++ * 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_SYSCALL_RET (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.
++ */
++
++#define nr_syscalls ((syscall_table_size)/4)
++
++/* Pseudo-eflags. */
++NMI_MASK = 0x80000000
++
++#ifdef CONFIG_PREEMPT
++#define preempt_stop(clobbers) DISABLE_INTERRUPTS(clobbers); TRACE_IRQS_OFF
++#else
++#define preempt_stop(clobbers)
++#define resume_kernel restore_nocheck
++#endif
++
++.macro TRACE_IRQS_IRET
++#ifdef CONFIG_TRACE_IRQFLAGS
++ testl $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off?
++ jz 1f
++ TRACE_IRQS_ON
++1:
++#endif
++.endm
++
++#ifdef CONFIG_VM86
++#define resume_userspace_sig check_userspace
++#else
++#define resume_userspace_sig resume_userspace
++#endif
++
++#define SAVE_ALL \
++ cld; \
++ pushl %fs; \
++ CFI_ADJUST_CFA_OFFSET 4;\
++ /*CFI_REL_OFFSET fs, 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_PERCPU), %edx; \
++ movl %edx, %fs
++
++#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;*/\
++3: popl %fs; \
++ CFI_ADJUST_CFA_OFFSET -4;\
++ /*CFI_RESTORE fs;*/\
++.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
++
++#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
++
++#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
++
++#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
++
++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
++END(ret_from_fork)
++
++/*
++ * 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(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 $(X86_EFLAGS_VM | SEGMENT_RPL_MASK), %eax
++ cmpl $USER_RPL, %eax
++ jb resume_kernel # not returning to v8086 or userspace
++
++ENTRY(resume_userspace)
++ LOCKDEP_SYS_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
++ andl $_TIF_WORK_MASK, %ecx # is there any work to be done on
++ # int/exception return?
++ jne work_pending
++ jmp restore_all
++END(ret_from_exception)
++
++#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 $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off (exception path) ?
++ jz restore_all
++ call preempt_schedule_irq
++ jmp need_resched
++END(resume_kernel)
++#endif
++ CFI_ENDPROC
++
++ .macro test_tif ti_reg # 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(\ti_reg)
++ .endm
++
++/* 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(ia32_sysenter_target)
++ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA esp, 0
++ CFI_REGISTER esp, ebp
++ movl SYSENTER_stack_sp0(%esp),%esp
++sysenter_past_esp:
++ /*
++ * Interrupts are disabled here, but we can't trace it until
++ * enough kernel state to call TRACE_IRQS_OFF can be called - but
++ * we immediately enable interrupts at that point anyway.
++ */
++ 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
++ orl $X86_EFLAGS_IF, (%esp)
++ 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
++
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ ENABLE_INTERRUPTS(CLBR_NONE)
++
++/*
++ * Load the potential sixth argument from user stack.
++ * Careful about security.
++ */
++ cmpl $__PAGE_OFFSET-3,%ebp
++ jae syscall_fault
++1: movl (%ebp),%ebp
++ movl %ebp,PT_EBP(%esp)
++.section __ex_table,"a"
++ .align 4
++ .long 1b,syscall_fault
++.previous
++
++ GET_THREAD_INFO(%ebp)
++ test_tif %ebp
++ jnz syscall_trace_entry
++ cmpl $(nr_syscalls), %eax
++ jae syscall_badsys
++ call *sys_call_table(,%eax,4)
++ movl %eax,PT_EAX(%esp)
++ LOCKDEP_SYS_EXIT
++ DISABLE_INTERRUPTS(CLBR_ANY)
++ 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_FS(%esp), %fs
++ ENABLE_INTERRUPTS_SYSCALL_RET
++ CFI_ENDPROC
++.pushsection .fixup,"ax"
++2: movl $0,PT_FS(%esp)
++ jmp 1b
++.section __ex_table,"a"
++ .align 4
++ .long 1b,2b
++.popsection
++ENDPROC(ia32_sysenter_target)
++
++ # pv sysenter call handler stub
++ENTRY(ia32pv_sysenter_target)
++ RING0_INT_FRAME
++ movl $__USER_DS,16(%esp)
++ movl %ebp,12(%esp)
++ movl $__USER_CS,4(%esp)
++ addl $4,%esp
++ CFI_ADJUST_CFA_OFFSET -4
++ /* +5*4 is SS:ESP,EFLAGS,CS:EIP. +8 is esp0 setting. */
++ pushl (TI_sysenter_return-THREAD_SIZE+8+4*4)(%esp)
++ CFI_ADJUST_CFA_OFFSET 4
++/*
++ * Load the potential sixth argument from user stack.
++ * Careful about security.
++ */
++ cmpl $__PAGE_OFFSET-3,%ebp
++ jae syscall_fault
++1: movl (%ebp),%ebp
++.section __ex_table,"a"
++ .align 4
++ .long 1b,syscall_fault
++.previous
++ /* fall through */
++ CFI_ENDPROC
++ENDPROC(ia32pv_sysenter_target)
++
++ # 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)
++ test_tif %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:
++ LOCKDEP_SYS_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
++ testl $X86_EFLAGS_TF,PT_EFLAGS(%esp) # If tracing set singlestep flag on exit
++ jz no_singlestep
++ orl $_TIF_SINGLESTEP,TI_flags(%ebp)
++no_singlestep:
++ movl TI_flags(%ebp), %ecx
++ testw $_TIF_ALLWORK_MASK, %cx # current->work
++ jne syscall_exit_work
++
++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 $(X86_EFLAGS_VM | (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
++restore_nocheck:
++ movl PT_EFLAGS(%esp), %eax
++ testl $(X86_EFLAGS_VM|NMI_MASK), %eax
++ CFI_REMEMBER_STATE
++ 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
++ CFI_REMEMBER_STATE
++ jnz restore_all_enable_events # != 0 => enable event delivery
++#endif
++ TRACE_IRQS_IRET
++restore_nocheck_notrace:
++ RESTORE_REGS
++ addl $4, %esp # skip orig_eax/error_code
++ CFI_ADJUST_CFA_OFFSET -4
++irq_return:
++ INTERRUPT_RETURN
++.section .fixup,"ax"
++ENTRY(iret_exc)
++ pushl $0 # no error code
++ pushl $do_iret_error
++ jmp error_code
++.previous
++.section __ex_table,"a"
++ .align 4
++ .long irq_return,iret_exc
++.previous
++
++ 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
++ /*
++ * 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.
++ */
++ cmpl $0, pv_info+PARAVIRT_enabled
++ jne restore_nocheck
++#endif
++
++ /* 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
++ ALIGN
++restore_all_enable_events:
++ TRACE_IRQS_ON
++ __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: INTERRUPT_RETURN
++.section __ex_table,"a"
++ .align 4
++ .long 1b,iret_exc
++.previous
++14: __DISABLE_INTERRUPTS
++ TRACE_IRQS_OFF
++ jmp 11f
++ecrit: /**** END OF CRITICAL REGION ****/
++
++ CFI_RESTORE_STATE
++hypervisor_iret:
++ andl $~NMI_MASK, PT_EFLAGS(%esp)
++ RESTORE_REGS
++ addl $4, %esp
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp hypercall_page + (__HYPERVISOR_iret * 32)
++#endif
++ CFI_ENDPROC
++ENDPROC(system_call)
++
++ # 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
++ LOCKDEP_SYS_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
++ 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
++#ifdef CONFIG_VM86
++ testl $X86_EFLAGS_VM, 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
++
++ 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
++ xorl %edx, %edx
++ call do_notify_resume
++ jmp resume_userspace_sig
++END(work_pending)
++
++ # 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
++END(syscall_trace_entry)
++
++ # 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
++END(syscall_exit_work)
++ CFI_ENDPROC
++
++ RING0_INT_FRAME # can't unwind into user space anyway
++syscall_fault:
++ GET_THREAD_INFO(%ebp)
++ movl $-EFAULT,PT_EAX(%esp)
++ jmp resume_userspace
++END(syscall_fault)
++
++syscall_badsys:
++ movl $-ENOSYS,PT_EAX(%esp)
++ jmp resume_userspace
++END(syscall_badsys)
++ CFI_ENDPROC
++
++#ifndef CONFIG_XEN
++#define FIXUP_ESPFIX_STACK \
++ /* since we are on a wrong stack, we cant make it a C code :( */ \
++ PER_CPU(gdt_page, %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:;
++
++/*
++ * Build the entry stubs and pointer table with
++ * some assembler magic.
++ */
++.section .rodata,"a"
++ENTRY(interrupt)
++.text
++
++ENTRY(irq_entries_start)
++ RING0_INT_FRAME
++vector=0
++.rept NR_IRQS
++ ALIGN
++ .if vector
++ CFI_ADJUST_CFA_OFFSET -4
++ .endif
++1: pushl $~(vector)
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp common_interrupt
++ .previous
++ .long 1b
++ .text
++vector=vector+1
++.endr
++END(irq_entries_start)
++
++.previous
++END(interrupt)
++.previous
++
++/*
++ * 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
++ENDPROC(common_interrupt)
++ CFI_ENDPROC
++
++#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; \
++ENDPROC(name)
++
++/* The include is where all of the SMP etc. interrupts come from */
++#include "entry_arch.h"
++
++#else
++#define UNWIND_ESPFIX_STACK
++#endif
++
++KPROBE_ENTRY(page_fault)
++ RING0_EC_FRAME
++ pushl $do_page_fault
++ CFI_ADJUST_CFA_OFFSET 4
++ ALIGN
++error_code:
++ /* the function address is in %fs'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
++ 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
++ cld
++ pushl %fs
++ CFI_ADJUST_CFA_OFFSET 4
++ /*CFI_REL_OFFSET fs, 0*/
++ movl $(__KERNEL_PERCPU), %ecx
++ movl %ecx, %fs
++ UNWIND_ESPFIX_STACK
++ popl %ecx
++ CFI_ADJUST_CFA_OFFSET -4
++ /*CFI_REGISTER es, ecx*/
++ movl PT_FS(%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_FS(%esp)
++ /*CFI_REL_OFFSET fs, ES*/
++ movl $(__USER_DS), %ecx
++ movl %ecx, %ds
++ movl %ecx, %es
++ movl %esp,%eax # pt_regs pointer
++ 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.
++# We want to avoid stacking callback handlers due to events occurring
++# during handling of the last event. To do this, we keep events disabled
++# until we've done all processing. HOWEVER, we must enable events before
++# popping the stack frame (can't be done atomically) and so it would still
++# be possible to get enough handler activations to overflow the stack.
++# Although unlikely, bugs of that kind are hard to track down, so we'd
++# like to avoid the possibility.
++# So, on entry to the handler we detect whether we interrupted an
++# existing activation in its critical region -- if so, we pop the current
++# activation and restart the handler using the previous one.
++#
++# The sysexit critical region is slightly different. sysexit
++# atomically removes the entire stack frame. If we interrupt in the
++# critical region we know that the entire frame is present and correct
++# so we can simply throw away the new one.
++ENTRY(hypervisor_callback)
++ RING0_INT_FRAME
++ pushl %eax
++ CFI_ADJUST_CFA_OFFSET 4
++ SAVE_ALL
++ movl PT_EIP(%esp),%eax
++ cmpl $scrit,%eax
++ jb 11f
++ cmpl $ecrit,%eax
++ jb critical_region_fixup
++ cmpl $sysexit_scrit,%eax
++ jb 11f
++ cmpl $sysexit_ecrit,%eax
++ ja 11f
++ addl $PT_OLDESP,%esp # Remove eflags...ebx from stack frame.
++11: push %esp
++ CFI_ADJUST_CFA_OFFSET 4
++ call evtchn_do_upcall
++ add $4,%esp
++ CFI_ADJUST_CFA_OFFSET -4
++ jmp ret_from_intr
++ CFI_ENDPROC
++
++# [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:
++ movzbl critical_fixup_table-scrit(%eax),%ecx # %eax contains num bytes popped
++ cmpb $0xff,%cl # 0xff => vcpu_info critical region
++ jne 15f
++ xorl %ecx,%ecx
++15: leal (%esp,%ecx),%esi # %esi points at end of src region
++ leal PT_OLDESP(%esp),%edi # %edi points at end of dst region
++ shrl $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
++ jmp 11b
++
++.section .rodata,"a"
++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 # pop %fs
++ .byte 0x28,0x28,0x28 # add $4,%esp
++ .byte 0x2c # iret
++ .byte 0xff,0xff,0xff,0xff # movb $1,1(%esi)
++ .byte 0x00,0x00 # jmp 11b
++.previous
++
++# 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)
++ pushl %eax
++ 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
++ jz 5f
++ addl $16,%esp # EAX != 0 => Category 2 (Bad IRET)
++ jmp iret_exc
++5: addl $16,%esp # EAX == 0 => Category 1 (Bad segment)
++ RING0_INT_FRAME
++ pushl $0
++ 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
++#endif
++ CFI_ENDPROC
++
++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
++END(coprocessor_error)
++
++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
++END(simd_coprocessor_error)
++
++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
++END(device_not_available)
++
++#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_sp0+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_sp0+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 */
++
++KPROBE_ENTRY(debug)
++ RING0_INT_FRAME
++#ifndef CONFIG_XEN
++ cmpl $ia32_sysenter_target,(%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
++/*
++ * 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.
++ */
++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 $ia32_sysenter_target,(%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 $ia32_sysenter_target,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
++
++nmi_stack_fixup:
++ RING0_INT_FRAME
++ FIX_STACK(12,nmi_stack_correct, 1)
++ jmp nmi_stack_correct
++
++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
++
++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
++ jmp irq_return
++ CFI_ENDPROC
++#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
++#endif
++KPROBE_END(nmi)
++
++#ifdef CONFIG_PARAVIRT
++ENTRY(native_iret)
++ iret
++.section __ex_table,"a"
++ .align 4
++ .long native_iret, iret_exc
++.previous
++END(native_iret)
++
++ENTRY(native_irq_enable_syscall_ret)
++ sti
++ sysexit
++END(native_irq_enable_syscall_ret)
++#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
++KPROBE_END(int3)
++
++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
++END(overflow)
++
++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
++END(bounds)
++
++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
++END(invalid_op)
++
++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
++END(coprocessor_segment_overrun)
++
++ENTRY(invalid_TSS)
++ RING0_EC_FRAME
++ pushl $do_invalid_TSS
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++END(invalid_TSS)
++
++ENTRY(segment_not_present)
++ RING0_EC_FRAME
++ pushl $do_segment_not_present
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++END(segment_not_present)
++
++ENTRY(stack_segment)
++ RING0_EC_FRAME
++ pushl $do_stack_segment
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++END(stack_segment)
++
++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)
++
++ENTRY(alignment_check)
++ RING0_EC_FRAME
++ pushl $do_alignment_check
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++END(alignment_check)
++
++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
++END(divide_error)
++
++#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
++END(machine_check)
++#endif
++
++#ifndef CONFIG_XEN
++ENTRY(spurious_interrupt_bug)
++ RING0_INT_FRAME
++ pushl $0
++ CFI_ADJUST_CFA_OFFSET 4
++ pushl $do_spurious_interrupt_bug
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++#endif /* !CONFIG_XEN */
++
++ENTRY(fixup_4gb_segment)
++ RING0_EC_FRAME
++ pushl $do_fixup_4gb_segment
++ CFI_ADJUST_CFA_OFFSET 4
++ jmp error_code
++ CFI_ENDPROC
++END(spurious_interrupt_bug)
++
++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)
++
++#include <asm/alternative-asm.h>
++
++ # pv syscall call handler stub
++ENTRY(ia32pv_cstar_target)
++ RING0_INT_FRAME
++ movl $__USER_DS,16(%esp)
++ movl %ebp,%ecx
++ movl $__USER_CS,4(%esp)
++ movl 12(%esp),%ebp
++ pushl %eax # save orig_eax
++ CFI_ADJUST_CFA_OFFSET 4
++/*
++ * Load the potential sixth argument from user stack.
++ * Careful about security.
++ */
++ cmpl $__PAGE_OFFSET-4,%ebp
++ CFI_REMEMBER_STATE
++ ja cstar_fault
++1: movl (%ebp),%ebp
++.section __ex_table,"a"
++ .align 4
++ .long 1b,cstar_fault
++.previous
++ SAVE_ALL
++ GET_THREAD_INFO(%ebp)
++ test_tif %ebp
++ jnz cstar_trace_entry
++ cmpl $nr_syscalls,%eax
++ jae cstar_badsys
++.Lcstar_call:
++ btl %eax,cstar_special
++ jc .Lcstar_special
++ call *cstar_call_table(,%eax,4)
++ movl %eax,PT_EAX(%esp) # store the return value
++.Lcstar_exit:
++ movl PT_ECX(%esp),%ecx
++ movl %ecx,PT_EBP(%esp) # put user EBP back in place
++ jmp syscall_exit
++.Lcstar_special:
++ movl PT_ECX(%esp),%ecx
++ movl %ecx,PT_EBP(%esp) # put user EBP back in place
++ jmp syscall_call
++cstar_set_tif:
++ movl $cstar_clear_tif,(%esp) # replace return address
++ LOCK_PREFIX
++ orl $_TIF_CSTAR,TI_flags(%ebp)
++ jmp *sys_call_table(,%eax,4)
++cstar_clear_tif:
++ movl %eax,PT_EAX(%esp) # store the return value
++ LOCK_PREFIX
++ andl $~_TIF_CSTAR,TI_flags(%ebp)
++ jmp .Lcstar_exit
++cstar_trace_entry:
++ movl $-ENOSYS,PT_EAX(%esp)
++ cmpl $nr_syscalls,%eax
++ jae 1f
++ btl %eax,cstar_special
++ jc .Lcstar_trace_special
++1: movl %esp,%eax
++ xorl %edx,%edx
++ LOCK_PREFIX
++ orl $_TIF_CSTAR,TI_flags(%ebp)
++ call do_syscall_trace
++ LOCK_PREFIX
++ andl $~_TIF_CSTAR,TI_flags(%ebp)
++ testl %eax,%eax
++ jne .Lcstar_resume # ret != 0 -> running under PTRACE_SYSEMU,
++ # so must skip actual syscall
++ movl PT_ORIG_EAX(%esp),%eax
++ cmpl $nr_syscalls,%eax
++ jb .Lcstar_call
++ jmp .Lcstar_exit
++.Lcstar_trace_special:
++ movl PT_ECX(%esp),%ecx
++ movl %esp,%eax
++ xorl %edx,%edx
++ movl %ecx,PT_EBP(%esp) # put user EBP back in place
++ call do_syscall_trace
++ testl %eax,%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
++ jb syscall_call
++ jmp syscall_exit
++cstar_badsys:
++ movl $-ENOSYS,PT_EAX(%esp)
++.Lcstar_resume:
++ movl PT_ECX(%esp),%ecx
++ movl %ecx,PT_EBP(%esp) # put user EBP back in place
++ jmp resume_userspace
++ CFI_RESTORE_STATE
++cstar_fault:
++ movl $-EFAULT,%eax
++ SAVE_ALL
++ GET_THREAD_INFO(%ebp)
++ jmp .Lcstar_resume
++ CFI_ENDPROC
++ENDPROC(ia32pv_cstar_target)
++
++ENTRY(cstar_ret_from_fork)
++ CFI_STARTPROC
++ movl PT_ECX(%esp),%ecx
++ GET_THREAD_INFO(%ebp)
++ movl %ecx,PT_EBP(%esp) # put user EBP back in place
++ LOCK_PREFIX
++ andl $~_TIF_CSTAR,TI_flags(%ebp)
++ jmp ret_from_fork
++ CFI_ENDPROC
++END(ret_from_fork)
++
++.section .rodata,"a"
++#include "syscall_table_32.S"
++
++syscall_table_size=(.-sys_call_table)
++
++#include <asm/unistd.h>
++cstar_special:
++nr=0
++mask=0
++.rept nr_syscalls+31
++ .irp n, __NR_sigreturn, __NR_rt_sigreturn
++ .if nr == \n
++ mask = mask | (1 << (\n & 31))
++ .endif
++ .endr
++ nr = nr + 1
++ .if (nr & 31) == 0
++ .long mask
++ mask = 0
++ .endif
++.endr
++#define sys_call_table cstar_call_table
++#define sys_fork cstar_set_tif
++#define sys_clone cstar_set_tif
++#define sys_vfork cstar_set_tif
++#include "syscall_table_32.S"
++#undef sys_call_table
++#undef sys_fork
++#undef sys_clone
++#undef sys_vfork
+diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
+index c778e4f..b61ba3e 100644
+--- a/arch/x86/kernel/entry_32.S
++++ b/arch/x86/kernel/entry_32.S
+@@ -282,7 +282,7 @@ ENTRY(ia32_sysenter_target)
+ CFI_SIGNAL_FRAME
+ CFI_DEF_CFA esp, 0
+ CFI_REGISTER esp, ebp
+- movl TSS_sysenter_sp0(%esp),%esp
++ movl SYSENTER_stack_sp0(%esp),%esp
+ sysenter_past_esp:
+ /*
+ * Interrupts are disabled here, but we can't trace it until
+@@ -739,7 +739,7 @@ END(device_not_available)
+ * that sets up the real kernel stack. Check here, since we can't
+ * allow the wrong stack to be used.
+ *
+- * "TSS_sysenter_sp0+12" is because the NMI/debug handler will have
++ * "SYSENTER_stack_sp0+12" is because the NMI/debug handler will have
+ * already pushed 3 words if it hits on the sysenter instruction:
+ * eflags, cs and eip.
+ *
+@@ -751,7 +751,7 @@ END(device_not_available)
+ cmpw $__KERNEL_CS,4(%esp); \
+ jne ok; \
+ label: \
+- movl TSS_sysenter_sp0+offset(%esp),%esp; \
++ movl SYSENTER_stack_sp0+offset(%esp),%esp; \
+ CFI_DEF_CFA esp, 0; \
+ CFI_UNDEFINED eip; \
+ pushfl; \
+@@ -1017,7 +1017,7 @@ ENTRY(kernel_thread_helper)
+ CFI_ENDPROC
+ ENDPROC(kernel_thread_helper)
+
+-#ifdef CONFIG_XEN
++#ifdef CONFIG_PARAVIRT_XEN
+ /* Xen doesn't set %esp to be precisely what the normal sysenter
+ entrypoint expects, so fix it up before using the normal path. */
+ ENTRY(xen_sysenter_target)
+@@ -1108,7 +1108,7 @@ ENTRY(xen_failsafe_callback)
+ .previous
+ ENDPROC(xen_failsafe_callback)
+
+-#endif /* CONFIG_XEN */
++#endif /* CONFIG_PARAVIRT_XEN */
+
+ .section .rodata,"a"
+ #include "syscall_table_32.S"
+diff --git a/arch/x86/kernel/entry_64-xen.S b/arch/x86/kernel/entry_64-xen.S
+new file mode 100644
+index 0000000..7da37a7
+--- /dev/null
++++ b/arch/x86/kernel/entry_64-xen.S
+@@ -0,0 +1,1258 @@
++/*
++ * linux/arch/x86_64/entry.S
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Asit Mallick <asit.k.mallick at intel.com>
++ * Modified for Xen
++ */
++
++/*
++ * entry.S contains the system-call and fault low-level handling routines.
++ *
++ * NOTE: This code handles signal-recognition, which happens every time
++ * after an interrupt and after each system call.
++ *
++ * Normal syscalls and interrupts don't save a full stack frame, this is
++ * only done for syscall tracing, signals or fork/exec et.al.
++ *
++ * A note on terminology:
++ * - top of stack: Architecture defined interrupt frame from SS to RIP
++ * at the top of the kernel process stack.
++ * - partial stack frame: partially saved registers upto R11.
++ * - full stack frame: Like partial stack frame, but all register saved.
++ *
++ * 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.
++ */
++
++#include <linux/linkage.h>
++#include <asm/segment.h>
++#include <asm/cache.h>
++#include <asm/errno.h>
++#include <asm/dwarf2.h>
++#include <asm/calling.h>
++#include <asm/asm-offsets.h>
++#include <asm/msr.h>
++#include <asm/unistd.h>
++#include <asm/thread_info.h>
++#include <asm/hw_irq.h>
++#include <asm/page.h>
++#include <asm/irqflags.h>
++#include <asm/errno.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/features.h>
++
++ .code64
++
++#ifndef CONFIG_PREEMPT
++#define retint_kernel retint_restore_args
++#endif
++
++#ifdef CONFIG_PARAVIRT
++ENTRY(native_irq_enable_syscall_ret)
++ movq %gs:pda_oldrsp,%rsp
++ swapgs
++ sysretq
++#endif /* CONFIG_PARAVIRT */
++
++
++.macro TRACE_IRQS_IRETQ offset=ARGOFFSET
++#ifdef CONFIG_TRACE_IRQFLAGS
++ bt $9,EFLAGS-\offset(%rsp) /* interrupts off? */
++ jnc 1f
++ TRACE_IRQS_ON
++1:
++#endif
++.endm
++
++NMI_MASK = 0x80000000
++
++/*
++ * C code is not supposed to know about undefined top of stack. Every time
++ * a C function with an pt_regs argument is called from the SYSCALL based
++ * fast path FIXUP_TOP_OF_STACK is needed.
++ * RESTORE_TOP_OF_STACK syncs the syscall state after any possible ptregs
++ * manipulation.
++ */
++
++ /* %rsp:at FRAMEEND */
++ .macro FIXUP_TOP_OF_STACK tmp
++ movq $__USER_CS,CS(%rsp)
++ movq $-1,RCX(%rsp)
++ .endm
++
++ .macro RESTORE_TOP_OF_STACK tmp,offset=0
++ .endm
++
++ .macro FAKE_STACK_FRAME child_rip
++ /* push in order ss, rsp, eflags, cs, rip */
++ xorl %eax, %eax
++ pushq %rax /* ss */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET ss,0*/
++ pushq %rax /* rsp */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rsp,0
++ pushq $(1<<9) /* eflags - interrupts on */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET rflags,0*/
++ pushq $__KERNEL_CS /* cs */
++ CFI_ADJUST_CFA_OFFSET 8
++ /*CFI_REL_OFFSET cs,0*/
++ pushq \child_rip /* rip */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rip,0
++ pushq %rax /* orig rax */
++ CFI_ADJUST_CFA_OFFSET 8
++ .endm
++
++ .macro UNFAKE_STACK_FRAME
++ addq $8*6, %rsp
++ CFI_ADJUST_CFA_OFFSET -(6*8)
++ .endm
++
++ .macro CFI_DEFAULT_STACK start=1,adj=0
++ .if \start
++ CFI_STARTPROC simple
++ CFI_SIGNAL_FRAME
++ CFI_DEF_CFA rsp,SS+8-(\adj*ARGOFFSET)
++ .else
++ CFI_DEF_CFA_OFFSET SS+8-(\adj*ARGOFFSET)
++ .endif
++ .if \adj == 0
++ CFI_REL_OFFSET r15,R15
++ CFI_REL_OFFSET r14,R14
++ CFI_REL_OFFSET r13,R13
++ CFI_REL_OFFSET r12,R12
++ CFI_REL_OFFSET rbp,RBP
++ CFI_REL_OFFSET rbx,RBX
++ .endif
++ CFI_REL_OFFSET r11,R11
++ CFI_REL_OFFSET r10,R10
++ CFI_REL_OFFSET r9,R9
++ CFI_REL_OFFSET r8,R8
++ CFI_REL_OFFSET rax,RAX
++ CFI_REL_OFFSET rcx,RCX
++ CFI_REL_OFFSET rdx,RDX
++ CFI_REL_OFFSET rsi,RSI
++ CFI_REL_OFFSET rdi,RDI
++ CFI_REL_OFFSET rip,RIP
++ /*CFI_REL_OFFSET cs,CS*/
++ /*CFI_REL_OFFSET rflags,EFLAGS*/
++ CFI_REL_OFFSET rsp,RSP
++ /*CFI_REL_OFFSET ss,SS*/
++ .endm
++
++ /*
++ * Must be consistent with the definition in arch-x86/xen-x86_64.h:
++ * struct iret_context {
++ * u64 rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
++ * };
++ * with rax, r11, and rcx being taken care of in the hypercall stub.
++ */
++ .macro HYPERVISOR_IRET flag
++ testb $3,1*8(%rsp)
++ jnz 2f
++ testl $NMI_MASK,2*8(%rsp)
++ jnz 2f
++
++ cmpb $0,(xen_features+XENFEAT_supervisor_mode_kernel)(%rip)
++ jne 1f
++
++ /* Direct iret to kernel space. Correct CS and SS. */
++ orl $3,1*8(%rsp)
++ orl $3,4*8(%rsp)
++1: iretq
++
++2: /* Slow iret via hypervisor. */
++ andl $~NMI_MASK, 2*8(%rsp)
++ pushq $\flag
++ jmp hypercall_page + (__HYPERVISOR_iret * 32)
++ .endm
++
++/*
++ * A newly forked process directly context switches into this.
++ */
++/* rdi: prev */
++ENTRY(ret_from_fork)
++ CFI_DEFAULT_STACK
++ push kernel_eflags(%rip)
++ CFI_ADJUST_CFA_OFFSET 4
++ popf # reset kernel eflags
++ CFI_ADJUST_CFA_OFFSET -4
++ call schedule_tail
++ GET_THREAD_INFO(%rcx)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx)
++ jnz rff_trace
++rff_action:
++ RESTORE_REST
++ testl $3,CS-ARGOFFSET(%rsp) # from kernel_thread?
++ je int_ret_from_sys_call
++ testl $_TIF_IA32,threadinfo_flags(%rcx)
++ jnz int_ret_from_sys_call
++ RESTORE_TOP_OF_STACK %rdi,ARGOFFSET
++ jmp ret_from_sys_call
++rff_trace:
++ movq %rsp,%rdi
++ call syscall_trace_leave
++ GET_THREAD_INFO(%rcx)
++ jmp rff_action
++ CFI_ENDPROC
++END(ret_from_fork)
++
++/*
++ * initial frame state for interrupts and exceptions
++ */
++ .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
++ /*CFI_REL_OFFSET rflags,EFLAGS-\ref*/
++ /*CFI_REL_OFFSET cs,CS-\ref*/
++ CFI_REL_OFFSET rip,RIP-\ref
++ .endm
++
++/*
++ * System call entry. Upto 6 arguments in registers are supported.
++ *
++ * SYSCALL does not save anything on the stack and does not change the
++ * stack pointer.
++ */
++
++/*
++ * Register setup:
++ * rax system call number
++ * rdi arg0
++ * rcx return address for syscall/sysret, C arg3
++ * rsi arg1
++ * rdx arg2
++ * r10 arg3 (--> moved to rcx for C)
++ * r8 arg4
++ * r9 arg5
++ * r11 eflags for syscall/sysret, temporary for C
++ * r12-r15,rbp,rbx saved by C code, not touched.
++ *
++ * Interrupts are enabled on entry.
++ * Only called from user space.
++ *
++ * XXX if we had a free scratch register we could save the RSP into the stack frame
++ * and report it properly in ps. Unfortunately we haven't.
++ *
++ * When user can change the frames always force IRET. That is because
++ * it deals with uncanonical addresses better. SYSRET has trouble
++ * with them due to bugs in both AMD and Intel CPUs.
++ */
++
++ENTRY(system_call)
++ _frame (RIP-0x10)
++ SAVE_ARGS -8,0
++ movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
++ GET_THREAD_INFO(%rcx)
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx)
++ jnz tracesys
++ cmpq $__NR_syscall_max,%rax
++ ja badsys
++ movq %r10,%rcx
++ call *sys_call_table(,%rax,8) # XXX: rip relative
++ movq %rax,RAX-ARGOFFSET(%rsp)
++/*
++ * Syscall return path ending with SYSRET (fast path)
++ * Has incomplete stack frame and undefined top of stack.
++ */
++ret_from_sys_call:
++ movl $_TIF_ALLWORK_MASK,%edi
++ /* edi: flagmask */
++sysret_check:
++ LOCKDEP_SYS_EXIT
++ GET_THREAD_INFO(%rcx)
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ jnz sysret_careful
++ CFI_REMEMBER_STATE
++ /*
++ * sysretq will re-enable interrupts:
++ */
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET VGCF_IN_SYSCALL
++
++ CFI_RESTORE_STATE
++ /* Handle reschedules */
++ /* edx: work, edi: workmask */
++sysret_careful:
++ bt $TIF_NEED_RESCHED,%edx
++ jnc sysret_signal
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ jmp sysret_check
++
++ /* Handle a signal */
++sysret_signal:
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ testl $_TIF_DO_NOTIFY_MASK,%edx
++ jz 1f
++
++ /* Really a signal */
++ /* edx: work flags (arg3) */
++ leaq do_notify_resume(%rip),%rax
++ leaq -ARGOFFSET(%rsp),%rdi # &pt_regs -> arg1
++ xorl %esi,%esi # oldset -> arg2
++ call ptregscall_common
++1: movl $_TIF_NEED_RESCHED,%edi
++ /* Use IRET because user could have changed frame. This
++ works because ptregscall_common has called FIXUP_TOP_OF_STACK. */
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++
++badsys:
++ movq $-ENOSYS,RAX-ARGOFFSET(%rsp)
++ jmp ret_from_sys_call
++
++ /* Do syscall tracing */
++tracesys:
++ SAVE_REST
++ movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */
++ FIXUP_TOP_OF_STACK %rdi
++ movq %rsp,%rdi
++ call syscall_trace_enter
++ LOAD_ARGS ARGOFFSET /* reload args from stack in case ptrace changed it */
++ RESTORE_REST
++ cmpq $__NR_syscall_max,%rax
++ ja int_ret_from_sys_call /* RAX(%rsp) set to -ENOSYS above */
++ movq %r10,%rcx /* fixup for C */
++ call *sys_call_table(,%rax,8)
++ movq %rax,RAX-ARGOFFSET(%rsp)
++ /* Use IRET because user could have changed frame */
++
++/*
++ * Syscall return path ending with IRET.
++ * Has correct top of stack, but partial stack frame.
++ */
++ .globl int_ret_from_sys_call
++int_ret_from_sys_call:
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ testb $3,CS-ARGOFFSET(%rsp)
++ jnz 1f
++ /* Need to set the proper %ss (not NULL) for ring 3 iretq */
++ movl $__KERNEL_DS,SS-ARGOFFSET(%rsp)
++ jmp retint_restore_args # retrun from ring3 kernel
++1:
++ movl $_TIF_ALLWORK_MASK,%edi
++ /* edi: mask to check */
++int_with_check:
++ LOCKDEP_SYS_EXIT_IRQ
++ GET_THREAD_INFO(%rcx)
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ jnz int_careful
++ andl $~TS_COMPAT,threadinfo_status(%rcx)
++ jmp retint_restore_args
++
++ /* Either reschedule or signal or syscall exit tracking needed. */
++ /* First do a reschedule test. */
++ /* edx: work, edi: workmask */
++int_careful:
++ bt $TIF_NEED_RESCHED,%edx
++ jnc int_very_careful
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++
++ /* handle signals and tracing -- both require a full stack frame */
++int_very_careful:
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ SAVE_REST
++ /* Check for syscall exit trace */
++ testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edx
++ jz int_signal
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ leaq 8(%rsp),%rdi # &ptregs -> arg1
++ call syscall_trace_leave
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ andl $~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edi
++ jmp int_restore_rest
++
++int_signal:
++ testl $_TIF_DO_NOTIFY_MASK,%edx
++ jz 1f
++ movq %rsp,%rdi # &ptregs -> arg1
++ xorl %esi,%esi # oldset -> arg2
++ call do_notify_resume
++1: movl $_TIF_NEED_RESCHED,%edi
++int_restore_rest:
++ RESTORE_REST
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ jmp int_with_check
++ CFI_ENDPROC
++END(system_call)
++
++/*
++ * Certain special system calls that need to save a complete full stack frame.
++ */
++
++ .macro PTREGSCALL label,func,arg
++ .globl \label
++\label:
++ leaq \func(%rip),%rax
++ leaq -ARGOFFSET+8(%rsp),\arg /* 8 for return address */
++ jmp ptregscall_common
++END(\label)
++ .endm
++
++ CFI_STARTPROC
++
++ PTREGSCALL stub_clone, sys_clone, %r8
++ PTREGSCALL stub_fork, sys_fork, %rdi
++ PTREGSCALL stub_vfork, sys_vfork, %rdi
++ PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend, %rdx
++ PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx
++ PTREGSCALL stub_iopl, sys_iopl, %rsi
++
++ENTRY(ptregscall_common)
++ popq %r11
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_REGISTER rip, r11
++ SAVE_REST
++ movq %r11, %r15
++ CFI_REGISTER rip, r15
++ FIXUP_TOP_OF_STACK %r11
++ call *%rax
++ RESTORE_TOP_OF_STACK %r11
++ movq %r15, %r11
++ CFI_REGISTER rip, r11
++ RESTORE_REST
++ pushq %r11
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rip, 0
++ ret
++ CFI_ENDPROC
++END(ptregscall_common)
++
++ENTRY(stub_execve)
++ CFI_STARTPROC
++ popq %r11
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_REGISTER rip, r11
++ SAVE_REST
++ FIXUP_TOP_OF_STACK %r11
++ movq %rsp, %rcx
++ call sys_execve
++ RESTORE_TOP_OF_STACK %r11
++ movq %rax,RAX(%rsp)
++ RESTORE_REST
++ jmp int_ret_from_sys_call
++ CFI_ENDPROC
++END(stub_execve)
++
++/*
++ * sigreturn is special because it needs to restore all registers on return.
++ * This cannot be done with SYSRET, so use the IRET return path instead.
++ */
++ENTRY(stub_rt_sigreturn)
++ CFI_STARTPROC
++ addq $8, %rsp
++ CFI_ADJUST_CFA_OFFSET -8
++ SAVE_REST
++ movq %rsp,%rdi
++ FIXUP_TOP_OF_STACK %r11
++ call sys_rt_sigreturn
++ movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer
++ RESTORE_REST
++ jmp int_ret_from_sys_call
++ CFI_ENDPROC
++END(stub_rt_sigreturn)
++
++/* initial frame state for interrupts (and exceptions without error code) */
++#define INTR_FRAME _frame (RIP-0x10); \
++ CFI_REL_OFFSET rcx,0; \
++ CFI_REL_OFFSET r11,8
++
++/* initial frame state for exceptions with error code (and interrupts with
++ vector already pushed) */
++#define XCPT_FRAME _frame (RIP-0x18); \
++ CFI_REL_OFFSET rcx,0; \
++ CFI_REL_OFFSET r11,8
++
++/*
++ * Interrupt exit.
++ *
++ */
++
++retint_check:
++ CFI_DEFAULT_STACK adj=1
++ LOCKDEP_SYS_EXIT_IRQ
++ movl threadinfo_flags(%rcx),%edx
++ andl %edi,%edx
++ CFI_REMEMBER_STATE
++ jnz retint_careful
++retint_restore_args: /* return to kernel space */
++ movl EFLAGS-REST_SKIP(%rsp), %eax
++ shr $9, %eax # EAX[0] == IRET_EFLAGS.IF
++ GET_VCPU_INFO
++ andb evtchn_upcall_mask(%rsi),%al
++ andb $1,%al # EAX[0] == IRET_EFLAGS.IF & event_mask
++ jnz restore_all_enable_events # != 0 => enable event delivery
++
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET 0
++
++ /* edi: workmask, edx: work */
++retint_careful:
++ CFI_RESTORE_STATE
++ bt $TIF_NEED_RESCHED,%edx
++ jnc retint_signal
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ pushq %rdi
++ CFI_ADJUST_CFA_OFFSET 8
++ call schedule
++ popq %rdi
++ CFI_ADJUST_CFA_OFFSET -8
++ GET_THREAD_INFO(%rcx)
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ jmp retint_check
++
++retint_signal:
++ testl $_TIF_DO_NOTIFY_MASK,%edx
++ jz retint_restore_args
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ SAVE_REST
++ movq $-1,ORIG_RAX(%rsp)
++ xorl %esi,%esi # oldset
++ movq %rsp,%rdi # &pt_regs
++ call do_notify_resume
++ RESTORE_REST
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ movl $_TIF_NEED_RESCHED,%edi
++ GET_THREAD_INFO(%rcx)
++ jmp retint_check
++
++#ifdef CONFIG_PREEMPT
++ /* Returning to kernel space. Check if we need preemption */
++ /* rcx: threadinfo. interrupts off. */
++ENTRY(retint_kernel)
++ cmpl $0,threadinfo_preempt_count(%rcx)
++ jnz retint_restore_args
++ bt $TIF_NEED_RESCHED,threadinfo_flags(%rcx)
++ jnc retint_restore_args
++ bt $9,EFLAGS-ARGOFFSET(%rsp) /* interrupts off? */
++ jnc retint_restore_args
++ call preempt_schedule_irq
++ jmp retint_kernel /* check again */
++#endif
++
++ CFI_ENDPROC
++END(retint_check)
++
++#ifndef CONFIG_XEN
++/*
++ * APIC interrupts.
++ */
++ .macro apicinterrupt num,func
++ INTR_FRAME
++ pushq $~(\num)
++ CFI_ADJUST_CFA_OFFSET 8
++ interrupt \func
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++ENTRY(thermal_interrupt)
++ apicinterrupt THERMAL_APIC_VECTOR,smp_thermal_interrupt
++END(thermal_interrupt)
++
++ENTRY(threshold_interrupt)
++ apicinterrupt THRESHOLD_APIC_VECTOR,mce_threshold_interrupt
++END(threshold_interrupt)
++
++#ifdef CONFIG_SMP
++ENTRY(reschedule_interrupt)
++ apicinterrupt RESCHEDULE_VECTOR,smp_reschedule_interrupt
++END(reschedule_interrupt)
++
++ .macro INVALIDATE_ENTRY num
++ENTRY(invalidate_interrupt\num)
++ apicinterrupt INVALIDATE_TLB_VECTOR_START+\num,smp_invalidate_interrupt
++END(invalidate_interrupt\num)
++ .endm
++
++ INVALIDATE_ENTRY 0
++ INVALIDATE_ENTRY 1
++ INVALIDATE_ENTRY 2
++ INVALIDATE_ENTRY 3
++ INVALIDATE_ENTRY 4
++ INVALIDATE_ENTRY 5
++ INVALIDATE_ENTRY 6
++ INVALIDATE_ENTRY 7
++
++ENTRY(call_function_interrupt)
++ apicinterrupt CALL_FUNCTION_VECTOR,smp_call_function_interrupt
++END(call_function_interrupt)
++ENTRY(irq_move_cleanup_interrupt)
++ apicinterrupt IRQ_MOVE_CLEANUP_VECTOR,smp_irq_move_cleanup_interrupt
++END(irq_move_cleanup_interrupt)
++#endif
++
++ENTRY(apic_timer_interrupt)
++ apicinterrupt LOCAL_TIMER_VECTOR,smp_apic_timer_interrupt
++END(apic_timer_interrupt)
++
++ENTRY(error_interrupt)
++ apicinterrupt ERROR_APIC_VECTOR,smp_error_interrupt
++END(error_interrupt)
++
++ENTRY(spurious_interrupt)
++ apicinterrupt SPURIOUS_APIC_VECTOR,smp_spurious_interrupt
++END(spurious_interrupt)
++#endif /* !CONFIG_XEN */
++
++/*
++ * Exception entry points.
++ */
++ .macro zeroentry sym
++ INTR_FRAME
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x10,%rsp /* skip rcx and r11 */
++ CFI_ADJUST_CFA_OFFSET -0x10
++ pushq $0 /* push error code/oldrax */
++ CFI_ADJUST_CFA_OFFSET 8
++ pushq %rax /* push real oldrax to the rdi slot */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rax,0
++ leaq \sym(%rip),%rax
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++ .macro errorentry sym
++ XCPT_FRAME
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x10,%rsp /* rsp points to the error code */
++ CFI_ADJUST_CFA_OFFSET -0x10
++ pushq %rax
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rax,0
++ leaq \sym(%rip),%rax
++ jmp error_entry
++ CFI_ENDPROC
++ .endm
++
++#if 0 /* not XEN */
++ /* error code is on the stack already */
++ /* handle NMI like exceptions that can happen everywhere */
++ .macro paranoidentry sym, ist=0, irqtrace=1
++ movq (%rsp),%rcx
++ movq 8(%rsp),%r11
++ addq $0x10,%rsp /* skip rcx and r11 */
++ SAVE_ALL
++ cld
++#if 0 /* not XEN */
++ movl $1,%ebx
++ movl $MSR_GS_BASE,%ecx
++ rdmsr
++ testl %edx,%edx
++ js 1f
++ SWAPGS
++ xorl %ebx,%ebx
++1:
++#endif
++ .if \ist
++ movq %gs:pda_data_offset, %rbp
++ .endif
++ movq %rsp,%rdi
++ movq ORIG_RAX(%rsp),%rsi
++ movq $-1,ORIG_RAX(%rsp)
++ .if \ist
++ subq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
++ .endif
++ call \sym
++ .if \ist
++ addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp)
++ .endif
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ .if \irqtrace
++ TRACE_IRQS_OFF
++ .endif
++ .endm
++
++ /*
++ * "Paranoid" exit path from exception stack.
++ * Paranoid because this is used by NMIs and cannot take
++ * any kernel state for granted.
++ * We don't do kernel preemption checks here, because only
++ * NMI should be common and it does not enable IRQs and
++ * cannot get reschedule ticks.
++ *
++ * "trace" is 0 for the NMI handler only, because irq-tracing
++ * is fundamentally NMI-unsafe. (we cannot change the soft and
++ * hard flags at once, atomically)
++ */
++ .macro paranoidexit trace=1
++ /* ebx: no swapgs flag */
++paranoid_exit\trace:
++ testl %ebx,%ebx /* swapgs needed? */
++ jnz paranoid_restore\trace
++ testl $3,CS(%rsp)
++ jnz paranoid_userspace\trace
++paranoid_swapgs\trace:
++ .if \trace
++ TRACE_IRQS_IRETQ 0
++ .endif
++ SWAPGS_UNSAFE_STACK
++paranoid_restore\trace:
++ RESTORE_ALL 8
++ jmp irq_return
++paranoid_userspace\trace:
++ GET_THREAD_INFO(%rcx)
++ movl threadinfo_flags(%rcx),%ebx
++ andl $_TIF_WORK_MASK,%ebx
++ jz paranoid_swapgs\trace
++ movq %rsp,%rdi /* &pt_regs */
++ call sync_regs
++ movq %rax,%rsp /* switch stack for scheduling */
++ testl $_TIF_NEED_RESCHED,%ebx
++ jnz paranoid_schedule\trace
++ movl %ebx,%edx /* arg3: thread flags */
++ .if \trace
++ TRACE_IRQS_ON
++ .endif
++ ENABLE_INTERRUPTS(CLBR_NONE)
++ xorl %esi,%esi /* arg2: oldset */
++ movq %rsp,%rdi /* arg1: &pt_regs */
++ call do_notify_resume
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ .if \trace
++ TRACE_IRQS_OFF
++ .endif
++ jmp paranoid_userspace\trace
++paranoid_schedule\trace:
++ .if \trace
++ TRACE_IRQS_ON
++ .endif
++ ENABLE_INTERRUPTS(CLBR_ANY)
++ call schedule
++ DISABLE_INTERRUPTS(CLBR_ANY)
++ .if \trace
++ TRACE_IRQS_OFF
++ .endif
++ jmp paranoid_userspace\trace
++ CFI_ENDPROC
++ .endm
++#endif
++
++/*
++ * Exception entry point. This expects an error code/orig_rax on the stack
++ * and the exception handler in %rax.
++ */
++KPROBE_ENTRY(error_entry)
++ _frame RDI
++ CFI_REL_OFFSET rax,0
++ /* rdi slot contains rax, oldrax contains error code */
++ cld
++ subq $14*8,%rsp
++ CFI_ADJUST_CFA_OFFSET (14*8)
++ movq %rsi,13*8(%rsp)
++ CFI_REL_OFFSET rsi,RSI
++ movq 14*8(%rsp),%rsi /* load rax from rdi slot */
++ CFI_REGISTER rax,rsi
++ movq %rdx,12*8(%rsp)
++ CFI_REL_OFFSET rdx,RDX
++ movq %rcx,11*8(%rsp)
++ CFI_REL_OFFSET rcx,RCX
++ movq %rsi,10*8(%rsp) /* store rax */
++ CFI_REL_OFFSET rax,RAX
++ movq %r8, 9*8(%rsp)
++ CFI_REL_OFFSET r8,R8
++ movq %r9, 8*8(%rsp)
++ CFI_REL_OFFSET r9,R9
++ movq %r10,7*8(%rsp)
++ CFI_REL_OFFSET r10,R10
++ movq %r11,6*8(%rsp)
++ CFI_REL_OFFSET r11,R11
++ movq %rbx,5*8(%rsp)
++ CFI_REL_OFFSET rbx,RBX
++ movq %rbp,4*8(%rsp)
++ CFI_REL_OFFSET rbp,RBP
++ movq %r12,3*8(%rsp)
++ CFI_REL_OFFSET r12,R12
++ movq %r13,2*8(%rsp)
++ CFI_REL_OFFSET r13,R13
++ movq %r14,1*8(%rsp)
++ CFI_REL_OFFSET r14,R14
++ movq %r15,(%rsp)
++ CFI_REL_OFFSET r15,R15
++#if 0
++ cmpl $__KERNEL_CS,CS(%rsp)
++ CFI_REMEMBER_STATE
++ je error_kernelspace
++#endif
++error_call_handler:
++ movq %rdi, RDI(%rsp)
++ CFI_REL_OFFSET rdi,RDI
++ movq %rsp,%rdi
++ movq ORIG_RAX(%rsp),%rsi # get error code
++ movq $-1,ORIG_RAX(%rsp)
++ call *%rax
++error_exit:
++ RESTORE_REST
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ GET_THREAD_INFO(%rcx)
++ testb $3,CS-ARGOFFSET(%rsp)
++ jz retint_kernel
++ LOCKDEP_SYS_EXIT_IRQ
++ movl threadinfo_flags(%rcx),%edx
++ movl $_TIF_WORK_MASK,%edi
++ andl %edi,%edx
++ jnz retint_careful
++ jmp retint_restore_args
++
++#if 0
++ /*
++ * We need to re-write the logic here because we don't do iretq to
++ * to return to user mode. It's still possible that we get trap/fault
++ * in the kernel (when accessing buffers pointed to by system calls,
++ * for example).
++ *
++ */
++ CFI_RESTORE_STATE
++error_kernelspace:
++ incl %ebx
++ /* There are two places in the kernel that can potentially fault with
++ usergs. Handle them here. The exception handlers after
++ iret run with kernel gs again, so don't set the user space flag.
++ B stepping K8s sometimes report an truncated RIP for IRET
++ exceptions returning to compat mode. Check for these here too. */
++ leaq irq_return(%rip),%rbp
++ cmpq %rbp,RIP(%rsp)
++ je error_swapgs
++ movl %ebp,%ebp /* zero extend */
++ cmpq %rbp,RIP(%rsp)
++ je error_swapgs
++ cmpq $gs_change,RIP(%rsp)
++ je error_swapgs
++ jmp error_sti
++#endif
++ CFI_ENDPROC
++KPROBE_END(error_entry)
++
++ENTRY(hypervisor_callback)
++ zeroentry do_hypervisor_callback
++END(hypervisor_callback)
++
++/*
++ * Copied from arch/xen/i386/kernel/entry.S
++ */
++# A note on the "critical region" in our callback handler.
++# We want to avoid stacking callback handlers due to events occurring
++# during handling of the last event. To do this, we keep events disabled
++# until we've done all processing. HOWEVER, we must enable events before
++# popping the stack frame (can't be done atomically) and so it would still
++# be possible to get enough handler activations to overflow the stack.
++# Although unlikely, bugs of that kind are hard to track down, so we'd
++# like to avoid the possibility.
++# So, on entry to the handler we detect whether we interrupted an
++# existing activation in its critical region -- if so, we pop the current
++# activation and restart the handler using the previous one.
++ENTRY(do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs)
++ CFI_STARTPROC
++# Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
++# see the correct pointer to the pt_regs
++ movq %rdi, %rsp # we don't return, adjust the stack frame
++ CFI_ENDPROC
++ CFI_DEFAULT_STACK
++11: incl %gs:pda_irqcount
++ movq %rsp,%rbp
++ CFI_DEF_CFA_REGISTER rbp
++ cmovzq %gs:pda_irqstackptr,%rsp
++ pushq %rbp # backlink for old unwinder
++ call evtchn_do_upcall
++ popq %rsp
++ CFI_DEF_CFA_REGISTER rsp
++ decl %gs:pda_irqcount
++ jmp error_exit
++ CFI_ENDPROC
++END(do_hypervisor_callback)
++
++ ALIGN
++restore_all_enable_events:
++ CFI_DEFAULT_STACK adj=1
++ TRACE_IRQS_ON
++ __ENABLE_INTERRUPTS
++
++scrit: /**** START OF CRITICAL REGION ****/
++ __TEST_PENDING
++ CFI_REMEMBER_STATE
++ jnz 14f # process more events if necessary...
++ RESTORE_ARGS 0,8,0
++ HYPERVISOR_IRET 0
++
++ CFI_RESTORE_STATE
++14: __DISABLE_INTERRUPTS
++ SAVE_REST
++ movq %rsp,%rdi # set the argument again
++ jmp 11b
++ CFI_ENDPROC
++ecrit: /**** END OF CRITICAL REGION ****/
++# At this point, unlike on x86-32, we don't do the fixup to simplify the
++# code and the stack frame is more complex on x86-64.
++# When the kernel is interrupted in the critical section, the kernel
++# will do IRET in that case, and everything will be restored at that point,
++# i.e. it just resumes from the next instruction interrupted with the same context.
++
++# Hypervisor uses this for application faults while it executes.
++# We get here for two reasons:
++# 1. Fault while reloading DS, ES, FS or GS
++# 2. Fault while executing IRET
++# Category 1 we do not need to fix up as Xen has already reloaded all segment
++# registers that could be reloaded and zeroed the others.
++# Category 2 we fix up by killing the current process. We cannot use the
++# normal Linux return path in this case because if we use the IRET hypercall
++# to pop the stack frame we end up in an infinite loop of failsafe callbacks.
++# We distinguish between categories by comparing each saved segment register
++# with its current contents: any discrepancy means we in category 1.
++ENTRY(failsafe_callback)
++ _frame (RIP-0x30)
++ CFI_REL_OFFSET rcx, 0
++ CFI_REL_OFFSET r11, 8
++ movw %ds,%cx
++ cmpw %cx,0x10(%rsp)
++ CFI_REMEMBER_STATE
++ jne 1f
++ movw %es,%cx
++ cmpw %cx,0x18(%rsp)
++ jne 1f
++ movw %fs,%cx
++ cmpw %cx,0x20(%rsp)
++ jne 1f
++ movw %gs,%cx
++ cmpw %cx,0x28(%rsp)
++ jne 1f
++ /* All segments match their saved values => Category 2 (Bad IRET). */
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x30,%rsp
++ CFI_ADJUST_CFA_OFFSET -0x30
++ movq $11,%rdi /* SIGSEGV */
++ jmp do_exit
++ CFI_RESTORE_STATE
++1: /* Segment mismatch => Category 1 (Bad segment). Retry the IRET. */
++ movq (%rsp),%rcx
++ CFI_RESTORE rcx
++ movq 8(%rsp),%r11
++ CFI_RESTORE r11
++ addq $0x30,%rsp
++ CFI_ADJUST_CFA_OFFSET -0x30
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8
++ SAVE_ALL
++ jmp error_exit
++ CFI_ENDPROC
++#if 0
++ .section __ex_table,"a"
++ .align 8
++ .quad gs_change,bad_gs
++ .previous
++ .section .fixup,"ax"
++ /* running with kernelgs */
++bad_gs:
++/* swapgs */ /* switch back to user gs */
++ xorl %eax,%eax
++ movl %eax,%gs
++ jmp 2b
++ .previous
++#endif
++
++/*
++ * Create a kernel thread.
++ *
++ * C extern interface:
++ * extern long kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
++ *
++ * asm input arguments:
++ * rdi: fn, rsi: arg, rdx: flags
++ */
++ENTRY(kernel_thread)
++ CFI_STARTPROC
++ FAKE_STACK_FRAME $child_rip
++ SAVE_ALL
++
++ # rdi: flags, rsi: usp, rdx: will be &pt_regs
++ movq %rdx,%rdi
++ orq kernel_thread_flags(%rip),%rdi
++ movq $-1, %rsi
++ movq %rsp, %rdx
++
++ xorl %r8d,%r8d
++ xorl %r9d,%r9d
++
++ # clone now
++ call do_fork
++ movq %rax,RAX(%rsp)
++ xorl %edi,%edi
++
++ /*
++ * It isn't worth to check for reschedule here,
++ * so internally to the x86_64 port you can rely on kernel_thread()
++ * not to reschedule the child before returning, this avoids the need
++ * of hacks for example to fork off the per-CPU idle tasks.
++ * [Hopefully no generic code relies on the reschedule -AK]
++ */
++ RESTORE_ALL
++ UNFAKE_STACK_FRAME
++ ret
++ CFI_ENDPROC
++ENDPROC(kernel_thread)
++
++child_rip:
++ pushq $0 # fake return address
++ CFI_STARTPROC
++ /*
++ * Here we are in the child and the registers are set as they were
++ * at kernel_thread() invocation in the parent.
++ */
++ movq %rdi, %rax
++ movq %rsi, %rdi
++ call *%rax
++ # exit
++ mov %eax, %edi
++ call do_exit
++ CFI_ENDPROC
++ENDPROC(child_rip)
++
++/*
++ * execve(). This function needs to use IRET, not SYSRET, to set up all state properly.
++ *
++ * C extern interface:
++ * extern long execve(char *name, char **argv, char **envp)
++ *
++ * asm input arguments:
++ * rdi: name, rsi: argv, rdx: envp
++ *
++ * We want to fallback into:
++ * extern long sys_execve(char *name, char **argv,char **envp, struct pt_regs *regs)
++ *
++ * do_sys_execve asm fallback arguments:
++ * rdi: name, rsi: argv, rdx: envp, rcx: fake frame on the stack
++ */
++ENTRY(kernel_execve)
++ CFI_STARTPROC
++ FAKE_STACK_FRAME $0
++ SAVE_ALL
++ movq %rsp,%rcx
++ call sys_execve
++ movq %rax, RAX(%rsp)
++ RESTORE_REST
++ testq %rax,%rax
++ jne 1f
++ jmp int_ret_from_sys_call
++1: RESTORE_ARGS
++ UNFAKE_STACK_FRAME
++ ret
++ CFI_ENDPROC
++ENDPROC(kernel_execve)
++
++KPROBE_ENTRY(page_fault)
++ errorentry do_page_fault
++KPROBE_END(page_fault)
++
++ENTRY(coprocessor_error)
++ zeroentry do_coprocessor_error
++END(coprocessor_error)
++
++ENTRY(simd_coprocessor_error)
++ zeroentry do_simd_coprocessor_error
++END(simd_coprocessor_error)
++
++ENTRY(device_not_available)
++ zeroentry math_state_restore
++END(device_not_available)
++
++ /* runs on exception stack */
++KPROBE_ENTRY(debug)
++/* INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8 */
++ zeroentry do_debug
++/* paranoidexit
++ CFI_ENDPROC */
++KPROBE_END(debug)
++
++KPROBE_ENTRY(nmi)
++ zeroentry do_nmi_callback
++KPROBE_END(nmi)
++do_nmi_callback:
++ CFI_STARTPROC
++ addq $8, %rsp
++ CFI_ENDPROC
++ CFI_DEFAULT_STACK
++ call do_nmi
++ orl $NMI_MASK,EFLAGS(%rsp)
++ RESTORE_REST
++ DISABLE_INTERRUPTS(CLBR_NONE)
++ TRACE_IRQS_OFF
++ GET_THREAD_INFO(%rcx)
++ jmp retint_restore_args
++ CFI_ENDPROC
++END(do_nmi_callback)
++
++KPROBE_ENTRY(int3)
++/* INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8 */
++ zeroentry do_int3
++/* jmp paranoid_exit1
++ CFI_ENDPROC */
++KPROBE_END(int3)
++
++ENTRY(overflow)
++ zeroentry do_overflow
++END(overflow)
++
++ENTRY(bounds)
++ zeroentry do_bounds
++END(bounds)
++
++ENTRY(invalid_op)
++ zeroentry do_invalid_op
++END(invalid_op)
++
++ENTRY(coprocessor_segment_overrun)
++ zeroentry do_coprocessor_segment_overrun
++END(coprocessor_segment_overrun)
++
++ENTRY(reserved)
++ zeroentry do_reserved
++END(reserved)
++
++#if 0
++ /* runs on exception stack */
++ENTRY(double_fault)
++ XCPT_FRAME
++ paranoidentry do_double_fault
++ jmp paranoid_exit1
++ CFI_ENDPROC
++END(double_fault)
++#endif
++
++ENTRY(invalid_TSS)
++ errorentry do_invalid_TSS
++END(invalid_TSS)
++
++ENTRY(segment_not_present)
++ errorentry do_segment_not_present
++END(segment_not_present)
++
++ /* runs on exception stack */
++ENTRY(stack_segment)
++/* XCPT_FRAME
++ paranoidentry do_stack_segment */
++ errorentry do_stack_segment
++/* jmp paranoid_exit1
++ CFI_ENDPROC */
++END(stack_segment)
++
++KPROBE_ENTRY(general_protection)
++ errorentry do_general_protection
++KPROBE_END(general_protection)
++
++ENTRY(alignment_check)
++ errorentry do_alignment_check
++END(alignment_check)
++
++ENTRY(divide_error)
++ zeroentry do_divide_error
++END(divide_error)
++
++ENTRY(spurious_interrupt_bug)
++ zeroentry do_spurious_interrupt_bug
++END(spurious_interrupt_bug)
++
++#ifdef CONFIG_X86_MCE
++ /* runs on exception stack */
++ENTRY(machine_check)
++ INTR_FRAME
++ pushq $0
++ CFI_ADJUST_CFA_OFFSET 8
++ paranoidentry do_machine_check
++ jmp paranoid_exit1
++ CFI_ENDPROC
++END(machine_check)
++#endif
++
++/* Call softirq on interrupt stack. Interrupts are off. */
++ENTRY(call_softirq)
++ CFI_STARTPROC
++ push %rbp
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rbp,0
++ mov %rsp,%rbp
++ CFI_DEF_CFA_REGISTER rbp
++ incl %gs:pda_irqcount
++ cmove %gs:pda_irqstackptr,%rsp
++ push %rbp # backlink for old unwinder
++ call __do_softirq
++ leaveq
++ CFI_DEF_CFA_REGISTER rsp
++ CFI_ADJUST_CFA_OFFSET -8
++ decl %gs:pda_irqcount
++ ret
++ CFI_ENDPROC
++ENDPROC(call_softirq)
++
++KPROBE_ENTRY(ignore_sysret)
++ CFI_STARTPROC
++ mov $-ENOSYS,%eax
++ HYPERVISOR_IRET 0
++ CFI_ENDPROC
++ENDPROC(ignore_sysret)
+diff --git a/arch/x86/kernel/fixup.c b/arch/x86/kernel/fixup.c
+new file mode 100644
+index 0000000..cbd6eb5
+--- /dev/null
++++ b/arch/x86/kernel/fixup.c
+@@ -0,0 +1,88 @@
++/******************************************************************************
++ * 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 )
++
++void do_fixup_4gb_segment(struct pt_regs *regs, long error_code)
++{
++ static unsigned long printed = 0;
++ char info[100];
++ int i;
++
++ /* Ignore statically-linked init. */
++ if (current->tgid == 1)
++ return;
++
++ VOID(HYPERVISOR_vm_assist(VMASST_CMD_disable,
++ VMASST_TYPE_4gb_segments_notify));
++
++ if (test_and_set_bit(0, &printed))
++ return;
++
++ 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");
++}
++
++static int __init fixup_init(void)
++{
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_4gb_segments_notify));
++ return 0;
++}
++__initcall(fixup_init);
+diff --git a/arch/x86/kernel/genapic_64-xen.c b/arch/x86/kernel/genapic_64-xen.c
+new file mode 100644
+index 0000000..2ccd7cf
+--- /dev/null
++++ b/arch/x86/kernel/genapic_64-xen.c
+@@ -0,0 +1,123 @@
++/*
++ * Copyright 2004 James Cleverdon, IBM.
++ * Subject to the GNU Public License, v.2
++ *
++ * Generic APIC sub-arch probe layer.
++ *
++ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
++ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
++ * James Cleverdon.
++ */
++#include <linux/threads.h>
++#include <linux/cpumask.h>
++#include <linux/string.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/ctype.h>
++#include <linux/init.h>
++#include <linux/hardirq.h>
++
++#include <asm/smp.h>
++#include <asm/genapic.h>
++
++#ifdef CONFIG_ACPI
++#include <acpi/acpi_bus.h>
++#endif
++
++#ifndef CONFIG_XEN
++DEFINE_PER_CPU(int, x2apic_extra_bits);
++
++struct genapic __read_mostly *genapic = &apic_flat;
++
++static enum uv_system_type uv_system_type;
++#else
++extern struct genapic apic_xen;
++struct genapic __read_mostly *genapic = &apic_xen;
++#endif
++
++
++/*
++ * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode.
++ */
++void __init setup_apic_routing(void)
++{
++#ifndef CONFIG_XEN
++ if (uv_system_type == UV_NON_UNIQUE_APIC)
++ genapic = &apic_x2apic_uv_x;
++ else
++#ifdef CONFIG_ACPI
++ /*
++ * Quirk: some x86_64 machines can only use physical APIC mode
++ * regardless of how many processors are present (x86_64 ES7000
++ * is an example).
++ */
++ if (acpi_gbl_FADT.header.revision > FADT2_REVISION_ID &&
++ (acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL))
++ genapic = &apic_physflat;
++ else
++#endif
++
++ if (num_possible_cpus() <= 8)
++ genapic = &apic_flat;
++ else
++ genapic = &apic_physflat;
++
++#else
++ /* hardcode to xen apic functions */
++ genapic = &apic_xen;
++#endif
++ printk(KERN_INFO "Setting APIC routing to %s\n", genapic->name);
++}
++
++/* Same for both flat and physical. */
++
++#ifdef CONFIG_XEN
++extern void xen_send_IPI_shortcut(unsigned int shortcut, int vector);
++#endif
++
++void send_IPI_self(int vector)
++{
++#ifndef CONFIG_XEN
++ __send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL);
++#else
++ xen_send_IPI_shortcut(APIC_DEST_SELF, vector);
++#endif
++}
++
++int __init acpi_madt_oem_check(char *oem_id, char *oem_table_id)
++{
++#ifndef CONFIG_XEN
++ if (!strcmp(oem_id, "SGI")) {
++ if (!strcmp(oem_table_id, "UVL"))
++ uv_system_type = UV_LEGACY_APIC;
++ else if (!strcmp(oem_table_id, "UVX"))
++ uv_system_type = UV_X2APIC;
++ else if (!strcmp(oem_table_id, "UVH"))
++ uv_system_type = UV_NON_UNIQUE_APIC;
++ }
++#endif
++ return 0;
++}
++
++#ifndef CONFIG_XEN
++unsigned int read_apic_id(void)
++{
++ unsigned int id;
++
++ WARN_ON(preemptible() && num_online_cpus() > 1);
++ id = apic_read(APIC_ID);
++ if (uv_system_type >= UV_X2APIC)
++ id |= __get_cpu_var(x2apic_extra_bits);
++ return id;
++}
++
++enum uv_system_type get_uv_system_type(void)
++{
++ return uv_system_type;
++}
++
++int is_uv_system(void)
++{
++ return uv_system_type != UV_NONE;
++}
++#endif
+diff --git a/arch/x86/kernel/genapic_xen_64.c b/arch/x86/kernel/genapic_xen_64.c
+new file mode 100644
+index 0000000..a729031
+--- /dev/null
++++ b/arch/x86/kernel/genapic_xen_64.c
+@@ -0,0 +1,164 @@
++/*
++ * Copyright 2004 James Cleverdon, IBM.
++ * Subject to the GNU Public License, v.2
++ *
++ * Xen APIC subarch code. Maximum 8 CPUs, logical delivery.
++ *
++ * Hacked for x86-64 by James Cleverdon from i386 architecture code by
++ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and
++ * James Cleverdon.
++ *
++ * Hacked to pieces for Xen by Chris Wright.
++ */
++#include <linux/threads.h>
++#include <linux/cpumask.h>
++#include <linux/string.h>
++#include <linux/kernel.h>
++#include <linux/ctype.h>
++#include <linux/init.h>
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#include <asm/smp.h>
++#else
++#include <asm/apic.h>
++#endif
++#include <asm/genapic.h>
++#include <xen/evtchn.h>
++
++DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++
++static inline void __send_IPI_one(unsigned int cpu, int vector)
++{
++ int irq = per_cpu(ipi_to_irq, cpu)[vector];
++ BUG_ON(irq < 0);
++ notify_remote_via_irq(irq);
++}
++
++void xen_send_IPI_shortcut(unsigned int shortcut, int vector)
++{
++ int cpu;
++
++ switch (shortcut) {
++ case APIC_DEST_SELF:
++ __send_IPI_one(smp_processor_id(), vector);
++ break;
++ case APIC_DEST_ALLBUT:
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu == smp_processor_id())
++ continue;
++ if (cpu_isset(cpu, cpu_online_map)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ break;
++ case APIC_DEST_ALLINC:
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu_isset(cpu, cpu_online_map)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ break;
++ default:
++ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
++ vector);
++ break;
++ }
++}
++
++static cpumask_t xen_target_cpus(void)
++{
++ return cpu_online_map;
++}
++
++static cpumask_t xen_vector_allocation_domain(int cpu)
++{
++ return cpumask_of_cpu(cpu);
++}
++
++/*
++ * Set up the logical destination ID.
++ * Do nothing, not called now.
++ */
++static void xen_init_apic_ldr(void)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ return;
++}
++
++static void xen_send_IPI_allbutself(int vector)
++{
++ /*
++ * if there are no other CPUs in the system then
++ * we get an APIC send error if we try to broadcast.
++ * thus we have to avoid sending IPIs in this case.
++ */
++ Dprintk("%s\n", __FUNCTION__);
++ if (num_online_cpus() > 1)
++ xen_send_IPI_shortcut(APIC_DEST_ALLBUT, vector);
++}
++
++static void xen_send_IPI_all(int vector)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ xen_send_IPI_shortcut(APIC_DEST_ALLINC, vector);
++}
++
++static void xen_send_IPI_mask(cpumask_t cpumask, int vector)
++{
++ unsigned long mask = cpus_addr(cpumask)[0];
++ unsigned int cpu;
++ unsigned long flags;
++
++ Dprintk("%s\n", __FUNCTION__);
++ local_irq_save(flags);
++ WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]);
++
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu_isset(cpu, cpumask)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ local_irq_restore(flags);
++}
++
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++static int xen_apic_id_registered(void)
++{
++ /* better be set */
++ Dprintk("%s\n", __FUNCTION__);
++ return physid_isset(smp_processor_id(), phys_cpu_present_map);
++}
++#endif
++
++static unsigned int xen_cpu_mask_to_apicid(cpumask_t cpumask)
++{
++ Dprintk("%s\n", __FUNCTION__);
++ return cpus_addr(cpumask)[0];
++}
++
++static unsigned int phys_pkg_id(int index_msb)
++{
++ u32 ebx;
++
++ Dprintk("%s\n", __FUNCTION__);
++ ebx = cpuid_ebx(1);
++ return ((ebx >> 24) & 0xFF) >> index_msb;
++}
++
++struct genapic apic_xen = {
++ .name = "xen",
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++ .int_delivery_mode = dest_LowestPrio,
++#endif
++ .int_dest_mode = 1,
++ .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
++ .init_apic_ldr = xen_init_apic_ldr,
++ .send_IPI_all = xen_send_IPI_all,
++ .send_IPI_allbutself = xen_send_IPI_allbutself,
++ .send_IPI_mask = xen_send_IPI_mask,
++ .cpu_mask_to_apicid = xen_cpu_mask_to_apicid,
++ .phys_pkg_id = phys_pkg_id,
++};
+diff --git a/arch/x86/kernel/head64-xen.c b/arch/x86/kernel/head64-xen.c
+new file mode 100644
+index 0000000..b25ebf6
+--- /dev/null
++++ b/arch/x86/kernel/head64-xen.c
+@@ -0,0 +1,229 @@
++/*
++ * 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/start_kernel.h>
++#include <linux/io.h>
++#include <linux/module.h>
++
++#include <asm/processor.h>
++#include <asm/proto.h>
++#include <asm/smp.h>
++#include <asm/setup.h>
++#include <asm/desc.h>
++#include <asm/pgtable.h>
++#include <asm/tlbflush.h>
++#include <asm/sections.h>
++#include <asm/kdebug.h>
++#include <asm/e820.h>
++#include <asm/bios_ebda.h>
++
++unsigned long start_pfn;
++
++#ifndef CONFIG_XEN
++static void __init zap_identity_mappings(void)
++{
++ pgd_t *pgd = pgd_offset_k(0UL);
++ pgd_clear(pgd);
++ __flush_tlb_all();
++}
++
++/* Don't add a printk in there. printk relies on the PDA which is not initialized
++ yet. */
++static void __init clear_bss(void)
++{
++ memset(__bss_start, 0,
++ (unsigned long) __bss_stop - (unsigned long) __bss_start);
++}
++#endif
++
++static void __init copy_bootdata(char *real_mode_data)
++{
++#ifndef CONFIG_XEN
++ char * command_line;
++
++ memcpy(&boot_params, real_mode_data, sizeof boot_params);
++ if (boot_params.hdr.cmd_line_ptr) {
++ command_line = __va(boot_params.hdr.cmd_line_ptr);
++ memcpy(boot_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(boot_command_line, xen_start_info->cmd_line, max_cmdline);
++ boot_command_line[max_cmdline-1] = '\0';
++#endif
++}
++
++#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);
++
++#define BIOS_LOWMEM_KILOBYTES 0x413
++
++/*
++ * The BIOS places the EBDA/XBDA at the top of conventional
++ * memory, and usually decreases the reported amount of
++ * conventional memory (int 0x12) too. This also contains a
++ * workaround for Dell systems that neglect to reserve EBDA.
++ * The same workaround also avoids a problem with the AMD768MPX
++ * 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.
++ */
++static void __init reserve_ebda_region(void)
++{
++#ifndef CONFIG_XEN
++ unsigned int lowmem, ebda_addr;
++
++ /* To determine the position of the EBDA and the */
++ /* end of conventional memory, we need to look at */
++ /* the BIOS data area. In a paravirtual environment */
++ /* that area is absent. We'll just have to assume */
++ /* that the paravirt case can handle memory setup */
++ /* correctly, without our help. */
++ if (paravirt_enabled())
++ return;
++
++ /* end of low (conventional) memory */
++ lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES);
++ lowmem <<= 10;
++
++ /* start of EBDA area */
++ ebda_addr = get_bios_ebda();
++
++ /* Fixup: bios puts an EBDA in the top 64K segment */
++ /* of conventional memory, but does not adjust lowmem. */
++ if ((lowmem - ebda_addr) <= 0x10000)
++ lowmem = ebda_addr;
++
++ /* Fixup: bios does not report an EBDA at all. */
++ /* Some old Dells seem to need 4k anyhow (bugzilla 2990) */
++ if ((ebda_addr == 0) && (lowmem >= 0x9f000))
++ lowmem = 0x9f000;
++
++ /* Paranoia: should never happen, but... */
++ if ((lowmem == 0) || (lowmem >= 0x100000))
++ lowmem = 0x9f000;
++
++ /* reserve all memory between lowmem and the 1MB mark */
++ reserve_early(lowmem, 0x100000, "BIOS reserved");
++#endif
++}
++
++static void __init reserve_setup_data(void)
++{
++#ifndef CONFIG_XEN
++ struct setup_data *data;
++ unsigned long pa_data;
++ char buf[32];
++
++ if (boot_params.hdr.version < 0x0209)
++ return;
++ pa_data = boot_params.hdr.setup_data;
++ while (pa_data) {
++ data = early_ioremap(pa_data, sizeof(*data));
++ sprintf(buf, "setup data %x", data->type);
++ reserve_early(pa_data, pa_data+sizeof(*data)+data->len, buf);
++ pa_data = data->next;
++ early_iounmap(data, sizeof(*data));
++ }
++#endif
++}
++
++void __init x86_64_start_kernel(char * real_mode_data)
++{
++ struct xen_machphys_mapping mapping;
++ unsigned long machine_to_phys_nr_ents;
++ int i;
++
++ /*
++ * Build-time sanity checks on the kernel image and module
++ * area mappings. (these are purely build-time and produce no code)
++ */
++ BUILD_BUG_ON(MODULES_VADDR < KERNEL_IMAGE_START);
++ BUILD_BUG_ON(MODULES_VADDR-KERNEL_IMAGE_START < KERNEL_IMAGE_SIZE);
++ BUILD_BUG_ON(MODULES_LEN + KERNEL_IMAGE_SIZE > 2*PUD_SIZE);
++ BUILD_BUG_ON((KERNEL_IMAGE_START & ~PMD_MASK) != 0);
++ BUILD_BUG_ON((MODULES_VADDR & ~PMD_MASK) != 0);
++ BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL));
++ BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) ==
++ (__START_KERNEL & PGDIR_MASK)));
++
++ xen_setup_features();
++
++ xen_start_info = (struct start_info *)real_mode_data;
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ phys_to_machine_mapping =
++ (unsigned long *)xen_start_info->mfn_list;
++ start_pfn = (__pa(xen_start_info->pt_base) >> PAGE_SHIFT) +
++ xen_start_info->nr_pt_frames;
++
++ machine_to_phys_mapping = (unsigned long *)MACH2PHYS_VIRT_START;
++ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
++ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
++ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
++ machine_to_phys_nr_ents = mapping.max_mfn + 1;
++ }
++ while ((1UL << machine_to_phys_order) < machine_to_phys_nr_ents )
++ machine_to_phys_order++;
++
++#ifndef CONFIG_XEN
++ /* clear bss before set_intr_gate with early_idt_handler */
++ clear_bss();
++
++ /* Make NULL pointers segfault */
++ zap_identity_mappings();
++
++ /* Cleanup the over mapped high alias */
++ cleanup_highmap();
++
++ for (i = 0; i < NUM_EXCEPTION_VECTORS; i++) {
++#ifdef CONFIG_EARLY_PRINTK
++ set_intr_gate(i, &early_idt_handlers[i]);
++#else
++ set_intr_gate(i, early_idt_handler);
++#endif
++ }
++ load_idt((const struct desc_ptr *)&idt_descr);
++#endif
++
++ early_printk("Kernel alive\n");
++
++ for (i = 0; i < NR_CPUS; i++)
++ cpu_pda(i) = &boot_cpu_pda[i];
++
++ pda_init(0);
++ copy_bootdata(__va(real_mode_data));
++
++ reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS");
++
++ reserve_early(round_up(__pa_symbol(&_end), PAGE_SIZE),
++ start_pfn << PAGE_SHIFT, "Xen provided");
++
++ reserve_ebda_region();
++ reserve_setup_data();
++
++ /*
++ * At this point everything still needed from the boot loader
++ * or BIOS or kernel text should be early reserved or marked not
++ * RAM in e820. All other memory is free game.
++ */
++
++ start_kernel();
++}
+diff --git a/arch/x86/kernel/head_32-xen.S b/arch/x86/kernel/head_32-xen.S
+new file mode 100644
+index 0000000..3a4160f
+--- /dev/null
++++ b/arch/x86/kernel/head_32-xen.S
+@@ -0,0 +1,164 @@
++
++
++.text
++#include <linux/elfnote.h>
++#include <linux/threads.h>
++#include <linux/init.h>
++#include <linux/linkage.h>
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/cache.h>
++#include <asm/thread_info.h>
++#include <asm/asm-offsets.h>
++#include <asm/boot.h>
++#include <asm/dwarf2.h>
++#include <xen/interface/xen.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
++
++.section .text.head,"ax", at progbits
++#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
++
++ movl %ss,%eax
++ movl %eax,%fs # gets reset once there's real percpu
++
++ /* 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 GS
++ movl %eax,%gs
++
++ cld # gcc2 wants the direction flag cleared at all times
++
++ pushl $0 # fake return address for unwinder
++ jmp i386_start_kernel
++
++#define HYPERCALL_PAGE_OFFSET 0x1000
++.org HYPERCALL_PAGE_OFFSET
++ENTRY(hypercall_page)
++ CFI_STARTPROC
++.skip 0x1000
++ CFI_ENDPROC
++
++/*
++ * Real beginning of normal "text" segment
++ */
++ENTRY(stext)
++ENTRY(_stext)
++
++/*
++ * BSS section
++ */
++.section ".bss.page_aligned","wa"
++ .align PAGE_SIZE_asm
++ENTRY(swapper_pg_fixmap)
++ .fill 1024,4,0
++ENTRY(empty_zero_page)
++ .fill 4096,1,0
++
++/*
++ * This starts the data section.
++ */
++.data
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++/*
++ * __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 + 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 <= 0x030002 */
++
++
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS, .asciz "linux")
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION, .asciz "2.6")
++ ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION, .asciz "xen-3.0")
++ ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE, .long __PAGE_OFFSET)
++#if CONFIG_XEN_COMPAT <= 0x030002
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long __PAGE_OFFSET)
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .long 0)
++#endif
++ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .long startup_32)
++ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long hypercall_page)
++ ELFNOTE(Xen, XEN_ELFNOTE_HV_START_LOW, .long HYPERVISOR_VIRT_START)
++ 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")
++ ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID, .quad _PAGE_PRESENT, _PAGE_PRESENT)
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz "no")
++ ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID, .long _PAGE_PRESENT, _PAGE_PRESENT)
++#endif
++ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz "generic")
++ ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long 1)
+diff --git a/arch/x86/kernel/head_64-xen.S b/arch/x86/kernel/head_64-xen.S
+new file mode 100644
+index 0000000..fdd4408
+--- /dev/null
++++ b/arch/x86/kernel/head_64-xen.S
+@@ -0,0 +1,211 @@
++/*
++ * linux/arch/x86_64/kernel/head.S -- start in 32bit and switch to 64bit
++ *
++ * Copyright (C) 2000 Andrea Arcangeli <andrea at suse.de> SuSE
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ * Copyright (C) 2000 Karsten Keil <kkeil at suse.de>
++ * Copyright (C) 2001,2002 Andi Kleen <ak at suse.de>
++ * Copyright (C) 2005 Eric Biederman <ebiederm at xmission.com>
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen
++ */
++
++
++#include <linux/linkage.h>
++#include <linux/threads.h>
++#include <linux/init.h>
++#include <linux/elfnote.h>
++#include <asm/desc.h>
++#include <asm/segment.h>
++#include <asm/page.h>
++#include <asm/msr.h>
++#include <asm/cache.h>
++#include <asm/dwarf2.h>
++#include <xen/interface/elfnote.h>
++
++ .section .text.head, "ax", @progbits
++ .code64
++ .globl startup_64
++startup_64:
++ movq $(init_thread_union+THREAD_SIZE-8),%rsp
++
++ /* rsi is pointer to startup info structure.
++ pass it to C */
++ movq %rsi,%rdi
++ pushq $0 # fake return address
++ jmp x86_64_start_kernel
++
++#ifdef CONFIG_ACPI_SLEEP
++.org 0xf00
++ .globl pGDT32
++pGDT32:
++ .word gdt_end-cpu_gdt_table-1
++ .long cpu_gdt_table-__START_KERNEL_map
++#endif
++
++.balign PAGE_SIZE
++
++#define NEXT_PAGE(name) \
++ .balign PAGE_SIZE; \
++ phys_##name = . - .text.head; \
++ENTRY(name)
++
++NEXT_PAGE(init_level4_pgt)
++ .fill 512,8,0
++ /*
++ * We update two pgd entries to make kernel and user pgd consistent
++ * at pgd_populate(). It can be used for kernel modules. So we place
++ * this page here for those cases to avoid memory corruption.
++ * We also use this page to establish the initial mapping for the
++ * vsyscall area.
++ */
++ .fill 512,8,0
++
++NEXT_PAGE(level3_kernel_pgt)
++ .fill 512,8,0
++
++ /*
++ * This is used for vsyscall area mapping as we have a different
++ * level4 page table for user.
++ */
++NEXT_PAGE(level3_user_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(level2_kernel_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(level2_fixmap_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(level1_fixmap_pgt)
++ .fill 512,8,0
++
++NEXT_PAGE(hypercall_page)
++ CFI_STARTPROC
++ .rept 0x1000 / 0x20
++ .skip 1 /* push %rcx */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rcx,0
++ .skip 2 /* push %r11 */
++ CFI_ADJUST_CFA_OFFSET 8
++ CFI_REL_OFFSET rcx,0
++ .skip 5 /* mov $#,%eax */
++ .skip 2 /* syscall */
++ .skip 2 /* pop %r11 */
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE r11
++ .skip 1 /* pop %rcx */
++ CFI_ADJUST_CFA_OFFSET -8
++ CFI_RESTORE rcx
++ .align 0x20,0 /* ret */
++ .endr
++ CFI_ENDPROC
++
++#undef NEXT_PAGE
++
++ .data
++/* Just dummy symbol to allow compilation. Not used in sleep path */
++#ifdef CONFIG_ACPI_SLEEP
++ .align PAGE_SIZE
++ENTRY(wakeup_level4_pgt)
++ .fill 512,8,0
++#endif
++
++ .data
++
++ .align 16
++ .globl cpu_gdt_descr
++cpu_gdt_descr:
++ .word gdt_end-cpu_gdt_table-1
++gdt:
++ .quad cpu_gdt_table
++#ifdef CONFIG_SMP
++ .rept NR_CPUS-1
++ .word 0
++ .quad 0
++ .endr
++#endif
++
++/* We need valid kernel segments for data and code in long mode too
++ * IRET will check the segment types kkeil 2000/10/28
++ * Also sysret mandates a special GDT layout
++ */
++
++ .section .data.page_aligned, "aw"
++ .align PAGE_SIZE
++
++/* The TLS descriptors are currently at a different place compared to i386.
++ Hopefully nobody expects them at a fixed place (Wine?) */
++
++ENTRY(cpu_gdt_table)
++ .quad 0x0000000000000000 /* NULL descriptor */
++ .quad 0x00cf9b000000ffff /* __KERNEL32_CS */
++ .quad 0x00af9b000000ffff /* __KERNEL_CS */
++ .quad 0x00cf93000000ffff /* __KERNEL_DS */
++ .quad 0x00cffb000000ffff /* __USER32_CS */
++ .quad 0x00cff3000000ffff /* __USER_DS, __USER32_DS */
++ .quad 0x00affb000000ffff /* __USER_CS */
++ .quad 0x0 /* unused */
++ .quad 0,0 /* TSS */
++ .quad 0,0 /* LDT */
++ .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 */
++
++ /* zero the remaining page */
++ .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0
++
++ .section .bss.page_aligned, "aw", @nobits
++ .align PAGE_SIZE
++ENTRY(empty_zero_page)
++ .skip PAGE_SIZE
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++/*
++ * __xen_guest information
++ */
++.macro utoh value
++ i = 64
++ .rept 16
++ i = i - 4
++ .byte '0' + ((((\value) >> i) & 0xf) > 9) * ('0' - 'A' + 10) + (((\value) >> i) & 0xf)
++ .endr
++.endm
++
++.section __xen_guest
++ .ascii "GUEST_OS=linux,GUEST_VER=2.6"
++ .ascii ",XEN_VER=xen-3.0"
++ .ascii ",VIRT_BASE=0x"
++ utoh __START_KERNEL_map
++ .ascii ",ELF_PADDR_OFFSET=0x"
++ utoh __START_KERNEL_map
++ .ascii ",VIRT_ENTRY=0x"
++ utoh (__START_KERNEL_map + __PHYSICAL_START)
++ .ascii ",HYPERCALL_PAGE=0x"
++ utoh (phys_hypercall_page >> PAGE_SHIFT)
++ .ascii ",FEATURES=writable_page_tables"
++ .ascii "|writable_descriptor_tables"
++ .ascii "|auto_translated_physmap"
++ .ascii "|supervisor_mode_kernel"
++ .ascii ",LOADER=generic"
++ .byte 0
++#endif /* CONFIG_XEN_COMPAT <= 0x030002 */
++
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS, .asciz "linux")
++ ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION, .asciz "2.6")
++ ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION, .asciz "xen-3.0")
++ ELFNOTE(Xen, XEN_ELFNOTE_VIRT_BASE, .quad __START_KERNEL_map)
++#if CONFIG_XEN_COMPAT <= 0x030002
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad __START_KERNEL_map)
++#else
++ ELFNOTE(Xen, XEN_ELFNOTE_PADDR_OFFSET, .quad 0)
++#endif
++ ELFNOTE(Xen, XEN_ELFNOTE_ENTRY, .quad startup_64)
++ ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .quad hypercall_page)
++ ELFNOTE(Xen, XEN_ELFNOTE_L1_MFN_VALID, .quad _PAGE_PRESENT, _PAGE_PRESENT)
++ ELFNOTE(Xen, XEN_ELFNOTE_FEATURES, .asciz "writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel")
++ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz "generic")
++ ELFNOTE(Xen, XEN_ELFNOTE_SUSPEND_CANCEL, .long 1)
+diff --git a/arch/x86/kernel/init_task-xen.c b/arch/x86/kernel/init_task-xen.c
+new file mode 100644
+index 0000000..420e4c7
+--- /dev/null
++++ b/arch/x86/kernel/init_task-xen.c
+@@ -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 signal_struct init_signals = INIT_SIGNALS(init_signals);
++static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
++#ifdef CONFIG_X86_XEN
++#define swapper_pg_dir ((pgd_t *)NULL)
++#endif
++struct mm_struct init_mm = INIT_MM(init_mm);
++#undef swapper_pg_dir
++EXPORT_UNUSED_SYMBOL(init_mm); /* will be removed in 2.6.26 */
++
++/*
++ * 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
++/*
++ * no more per-task TSS's. The TSS size is kept cacheline-aligned
++ * so they are allowed to end up in the .data.cacheline_aligned
++ * section. Since TSS's are completely CPU-local, we want them
++ * on exact cacheline boundaries, to eliminate cacheline ping-pong.
++ */
++DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS;
++#endif
++
+diff --git a/arch/x86/kernel/io_apic_32-xen.c b/arch/x86/kernel/io_apic_32-xen.c
+new file mode 100644
+index 0000000..5646681
+--- /dev/null
++++ b/arch/x86/kernel/io_apic_32-xen.c
+@@ -0,0 +1,2913 @@
++/*
++ * 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/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>
++#include <linux/kthread.h>
++#include <linux/jiffies.h> /* time_after() */
++
++#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>
++
++#include <mach_apic.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;
++
++#define clear_IO_APIC() ((void)0)
++#else
++int (*ioapic_renumber_irq)(int ioapic, int irq);
++atomic_t irq_mis_count;
++#endif /* CONFIG_XEN */
++
++/* 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);
++
++#ifndef CONFIG_XEN
++int timer_over_8254 __initdata = 1;
++#endif
++
++/*
++ * 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];
++
++/* I/O APIC entries */
++struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
++int nr_ioapics;
++
++/* MP IRQ source entries */
++struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
++
++/* # of MP IRQ source entries */
++int mp_irq_entries;
++
++#ifndef CONFIG_XEN
++static int disable_timer_pin_1 __initdata;
++#endif
++
++/*
++ * 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];
++
++#ifndef CONFIG_XEN
++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);
++}
++#endif
++
++static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg)
++{
++#ifndef CONFIG_XEN
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ return readl(&io_apic->data);
++#else
++ 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;
++#endif
++}
++
++static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
++{
++#ifndef CONFIG_XEN
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++#else
++ struct physdev_apic apic_op;
++
++ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
++ apic_op.reg = reg;
++ apic_op.value = value;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op));
++#endif
++}
++
++#ifndef CONFIG_XEN
++/*
++ * 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 __iomem *io_apic = io_apic_base(apic);
++ if (sis_apic_bug)
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++}
++#else
++#define io_apic_modify io_apic_write
++#endif
++
++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);
++}
++
++#ifndef CONFIG_XEN
++/*
++ * 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!
++ */
++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);
++}
++#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;
++}
++
++#ifndef CONFIG_XEN
++/*
++ * 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;
++
++ /* 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);
++}
++
++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;
++ }
++ irq_desc[irq].affinity = 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>
++
++#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(per_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;
++
++ 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 or balancing disabled ? */
++ if (!irq_desc[j].action || irq_balancing_disabled(j))
++ 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;
++
++ /* if imbalance is less than approx 10% of max load, then
++ * observe diminishing returns action. - quit
++ */
++ if (imbalance < (max_cpu_irq >> 3))
++ 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 accumulated 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, per_cpu(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)) {
++ /* 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);
++ return;
++}
++
++static int balanced_irq(void *unused)
++{
++ int i;
++ unsigned long prev_balance_time = jiffies;
++ long time_remaining = balanced_irq_interval;
++
++ /* 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));
++ }
++
++ set_freezable();
++ 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 (!IS_ERR(kthread_run(balanced_irq, NULL, "kirqd")))
++ return 0;
++ 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 __devinit 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 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_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 (test_bit(lbus, mp_bus_not_pci) &&
++ (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 (test_bit(lbus, mp_bus_not_pci) &&
++ (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 (!test_bit(lbus, mp_bus_not_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
++
++#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
++/*
++ * 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;
++}
++#endif
++
++/* 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)
++
++/* 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) default_ISA_polarity(idx)
++
++/* 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) default_ISA_polarity(idx)
++
++static int 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 */
++ {
++ polarity = test_bit(bus, mp_bus_not_pci)?
++ default_ISA_polarity(idx):
++ default_PCI_polarity(idx);
++ 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 */
++ {
++ trigger = test_bit(bus, mp_bus_not_pci)?
++ default_ISA_trigger(idx):
++ default_PCI_trigger(idx);
++#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
++ switch (mp_bus_id_to_type[bus])
++ {
++ case MP_BUS_ISA: /* ISA pin */
++ {
++ /* set before the switch */
++ break;
++ }
++ case MP_BUS_EISA: /* EISA pin */
++ {
++ trigger = default_EISA_trigger(idx);
++ break;
++ }
++ case MP_BUS_PCI: /* PCI pin */
++ {
++ /* set before the switch */
++ break;
++ }
++ case MP_BUS_MCA: /* MCA pin */
++ {
++ trigger = default_MCA_trigger(idx);
++ break;
++ }
++ default:
++ {
++ printk(KERN_WARNING "broken BIOS!!\n");
++ trigger = 1;
++ break;
++ }
++ }
++#endif
++ 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");
++
++ 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;
++
++#ifndef CONFIG_XEN
++ /*
++ * For MPS mode, so far only needed by ES7000 platform
++ */
++ if (ioapic_renumber_irq)
++ irq = ioapic_renumber_irq(apic, irq);
++#endif
++ }
++
++ /*
++ * 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. */
++static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly; /* = { FIRST_DEVICE_VECTOR , 0 }; */
++
++static int __assign_irq_vector(int irq)
++{
++ int vector;
++ struct physdev_irq irq_op;
++
++ BUG_ON((unsigned)irq >= NR_IRQ_VECTORS);
++
++ 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;
++
++ vector = irq_op.vector;
++ irq_vector[irq] = vector;
++
++ return vector;
++}
++
++static int assign_irq_vector(int irq)
++{
++ unsigned long flags;
++ int vector;
++
++ spin_lock_irqsave(&vector_lock, flags);
++ vector = __assign_irq_vector(irq);
++ spin_unlock_irqrestore(&vector_lock, flags);
++
++ return vector;
++}
++
++#ifndef CONFIG_XEN
++static struct irq_chip ioapic_chip;
++
++#define IOAPIC_AUTO -1
++#define IOAPIC_EDGE 0
++#define IOAPIC_LEVEL 1
++
++static void ioapic_register_intr(int irq, int vector, unsigned long trigger)
++{
++ if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) ||
++ trigger == IOAPIC_LEVEL) {
++ irq_desc[irq].status |= IRQ_LEVEL;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_fasteoi_irq, "fasteoi");
++ } else {
++ irq_desc[irq].status &= ~IRQ_LEVEL;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_edge_irq, "edge");
++ }
++ set_intr_gate(vector, interrupt[irq]);
++}
++#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;
++
++ 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;
++ }
++
++ if (!first_notcon) {
++ apic_printk(APIC_VERBOSE, " not connected.\n");
++ first_notcon = 1;
++ }
++
++ 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);
++ }
++ ioapic_write_entry(apic, pin, entry);
++ }
++ }
++
++ 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;
++
++ 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_chip;
++ set_irq_handler(0, handle_edge_irq);
++
++ /*
++ * Add it to the IO-APIC irq-routing table:
++ */
++ ioapic_write_entry(apic, pin, entry);
++
++ enable_8259A_irq(0);
++}
++
++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);
++
++ printk(KERN_DEBUG ".... register #01: %08X\n", reg_01.raw);
++ printk(KERN_DEBUG "....... : max redirection entries: %04X\n", reg_01.bits.entries);
++
++ printk(KERN_DEBUG "....... : PRQ implemented: %X\n", reg_01.bits.PRQ);
++ printk(KERN_DEBUG "....... : IO APIC version: %04X\n", reg_01.bits.version);
++
++ /*
++ * 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);
++ }
++
++ /*
++ * 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);
++ }
++
++ 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;
++
++ entry = ioapic_read_entry(apic, i);
++
++ 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
++ );
++ }
++ }
++ 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");
++ }
++
++ printk(KERN_INFO ".................................... done.\n");
++
++ return;
++}
++
++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());
++ printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v,
++ GET_APIC_ID(read_apic_id()));
++ v = apic_read(APIC_LVR);
++ printk(KERN_INFO "... APIC VERSION: %08x\n", v);
++ ver = GET_APIC_VERSION(v);
++ maxlvt = lapic_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 /* !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;
++ 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.
++ */
++ 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;
++
++ 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(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
++}
++
++/*
++ * 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
++int no_timer_check __initdata;
++
++static int __init notimercheck(char *s)
++{
++ no_timer_check = 1;
++ return 1;
++}
++__setup("no_timer_check", notimercheck);
++
++/*
++ * 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;
++ unsigned long flags;
++
++ if (no_timer_check)
++ return 1;
++
++ local_save_flags(flags);
++ local_irq_enable();
++ /* Let ten ticks pass... */
++ mdelay((10 * 1000) / HZ);
++ local_irq_restore(flags);
++
++ /*
++ * 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 (time_after(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.
++ */
++
++/*
++ * 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)
++{
++ 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;
++}
++
++static void ack_ioapic_irq(unsigned int irq)
++{
++ move_native_irq(irq);
++ ack_APIC_irq();
++}
++
++static void ack_ioapic_quirk_irq(unsigned int irq)
++{
++ unsigned long v;
++ int i;
++
++ move_native_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 = irq_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);
++ }
++}
++
++static int ioapic_retrigger_irq(unsigned int irq)
++{
++ send_IPI_self(irq_vector[irq]);
++
++ return 1;
++}
++
++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,
++};
++#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++) {
++ if (IO_APIC_IRQ(irq) && !irq_vector[irq]) {
++ /*
++ * 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
++ }
++ }
++}
++
++#ifndef CONFIG_XEN
++/*
++ * The local APIC irq-chip implementation:
++ */
++
++static void ack_apic(unsigned int irq)
++{
++ ack_APIC_irq();
++}
++
++static void mask_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED);
++}
++
++static void unmask_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write_around(APIC_LVT0, v & ~APIC_LVT_MASKED);
++}
++
++static struct irq_chip lapic_chip __read_mostly = {
++ .name = "local-APIC-edge",
++ .mask = mask_lapic_irq,
++ .unmask = unmask_lapic_irq,
++ .eoi = ack_apic,
++};
++
++static void __init 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 ...");
++
++ enable_NMI_through_LVT0();
++
++ 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 __init unlock_ExtINT_logic(void)
++{
++ int apic, pin, i;
++ struct IO_APIC_route_entry entry0, entry1;
++ unsigned char save_control, save_freq_select;
++
++ 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;
++ }
++
++ entry0 = ioapic_read_entry(apic, pin);
++ 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;
++
++ ioapic_write_entry(apic, pin, entry1);
++
++ 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);
++
++ ioapic_write_entry(apic, pin, entry0);
++}
++
++/*
++ * 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 __init check_timer(void)
++{
++ int apic1, pin1, apic2, pin2;
++ int vector;
++ unsigned long flags;
++
++ local_irq_save(flags);
++
++ /*
++ * 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;
++
++ 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);
++ goto out;
++ }
++ 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();
++ }
++ goto out;
++ }
++ /*
++ * 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...");
++
++ disable_8259A_irq(0);
++ set_irq_chip_and_handler_name(0, &lapic_chip, handle_fasteoi_irq,
++ "fasteoi");
++ apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */
++ enable_8259A_irq(0);
++
++ if (timer_irq_works()) {
++ printk(" works.\n");
++ goto out;
++ }
++ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector);
++ printk(" failed.\n");
++
++ printk(KERN_INFO "...trying to set up timer as ExtINT IRQ...");
++
++ 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");
++ goto out;
++ }
++ 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");
++out:
++ local_irq_restore(flags);
++}
++#else
++int timer_uses_ioapic_pin_0 = 0;
++#define check_timer() ((void)0)
++#endif
++
++/*
++ *
++ * IRQ's that are handled by the PIC in the MPS IOAPIC case.
++ * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ.
++ * Linux doesn't really care, as it's not actually used
++ * for any interrupt handling anyway.
++ */
++#define PIC_IRQS (1 << PIC_CASCADE_IR)
++
++void __init setup_IO_APIC(void)
++{
++#ifndef CONFIG_XEN
++ int i;
++
++ /* Reserve all the system vectors. */
++ for (i = FIRST_SYSTEM_VECTOR; i < NR_VECTORS; i++)
++ set_bit(i, used_vectors);
++#endif
++
++ 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");
++
++ /*
++ * 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();
++}
++
++#ifndef CONFIG_XEN
++static int __init setup_disable_8254_timer(char *s)
++{
++ timer_over_8254 = -1;
++ return 1;
++}
++static int __init setup_enable_8254_timer(char *s)
++{
++ timer_over_8254 = 2;
++ return 1;
++}
++
++__setup("disable_8254_timer", setup_disable_8254_timer);
++__setup("enable_8254_timer", setup_enable_8254_timer);
++#endif
++
++/*
++ * Called after all the initialization is done. If we didnt find any
++ * APIC bugs then we can allow the modify fast path
++ */
++
++static int __init io_apic_bug_finalize(void)
++{
++ if(sis_apic_bug == -1)
++ sis_apic_bug = 0;
++ if (is_initial_xendomain()) {
++ struct xen_platform_op op = { .cmd = XENPF_platform_quirk };
++ op.u.platform_quirk.quirk_id = sis_apic_bug ?
++ QUIRK_IOAPIC_BAD_REGSEL : QUIRK_IOAPIC_GOOD_REGSEL;
++ VOID(HYPERVISOR_platform_op(&op));
++ }
++ return 0;
++}
++
++late_initcall(io_apic_bug_finalize);
++
++struct sysfs_ioapic_data {
++ struct sys_device dev;
++ struct IO_APIC_route_entry entry[0];
++};
++static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
++
++static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ 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);
++
++ return 0;
++}
++
++static int ioapic_resume(struct sys_device *dev)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ unsigned long flags;
++ union IO_APIC_reg_00 reg_00;
++ int i;
++
++ data = container_of(dev, struct sysfs_ioapic_data, dev);
++ entry = data->entry;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(dev->id, 0);
++ if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) {
++ reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid;
++ io_apic_write(dev->id, 0, reg_00.raw);
++ }
++ 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;
++}
++
++static struct sysdev_class ioapic_sysdev_class = {
++ .name = "ioapic",
++ .suspend = ioapic_suspend,
++ .resume = ioapic_resume,
++};
++
++static int __init ioapic_init_sysfs(void)
++{
++ struct sys_device * dev;
++ int i, size, error = 0;
++
++ error = sysdev_class_register(&ioapic_sysdev_class);
++ if (error)
++ return error;
++
++ for (i = 0; i < nr_ioapics; i++ ) {
++ size = sizeof(struct sys_device) + nr_ioapic_registers[i]
++ * sizeof(struct IO_APIC_route_entry);
++ mp_ioapic_data[i] = kmalloc(size, GFP_KERNEL);
++ if (!mp_ioapic_data[i]) {
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ memset(mp_ioapic_data[i], 0, size);
++ dev = &mp_ioapic_data[i]->dev;
++ dev->id = i;
++ dev->cls = &ioapic_sysdev_class;
++ error = sysdev_register(dev);
++ if (error) {
++ kfree(mp_ioapic_data[i]);
++ mp_ioapic_data[i] = NULL;
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ }
++
++ return 0;
++}
++
++device_initcall(ioapic_init_sysfs);
++
++#ifndef CONFIG_XEN
++/*
++ * Dynamic irq allocate and deallocation
++ */
++int create_irq(void)
++{
++ /* Allocate an unused irq */
++ int irq, new, vector = 0;
++ unsigned long flags;
++
++ 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);
++
++ if (irq >= 0) {
++ set_intr_gate(vector, interrupt[irq]);
++ dynamic_irq_init(irq);
++ }
++ return irq;
++}
++
++void destroy_irq(unsigned int irq)
++{
++ unsigned long flags;
++
++ dynamic_irq_cleanup(irq);
++
++ spin_lock_irqsave(&vector_lock, flags);
++ clear_bit(irq_vector[irq], used_vectors);
++ irq_vector[irq] = 0;
++ spin_unlock_irqrestore(&vector_lock, flags);
++}
++#endif
++
++/*
++ * MSI message composition
++ */
++#if defined(CONFIG_PCI_MSI) && !defined(CONFIG_XEN)
++static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
++{
++ int vector;
++ unsigned dest;
++
++ 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;
++}
++
++#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;
++
++ vector = assign_irq_vector(irq);
++ if (vector < 0)
++ return;
++
++ dest = cpu_mask_to_apicid(mask);
++
++ 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);
++ irq_desc[irq].affinity = 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_ioapic_irq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_msi_irq_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc)
++{
++ struct msi_msg msg;
++ int irq, ret;
++ irq = create_irq();
++ if (irq < 0)
++ return irq;
++
++ ret = msi_compose_msg(dev, irq, &msg);
++ if (ret < 0) {
++ destroy_irq(irq);
++ return ret;
++ }
++
++ set_irq_msi(irq, desc);
++ 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)
++{
++ destroy_irq(irq);
++}
++
++#endif /* CONFIG_PCI_MSI */
++
++/*
++ * Hypertransport interrupt support
++ */
++#ifdef CONFIG_HT_IRQ
++
++#ifdef CONFIG_SMP
++
++static void target_ht_irq(unsigned int irq, unsigned int dest)
++{
++ struct ht_irq_msg msg;
++ fetch_ht_irq_msg(irq, &msg);
++
++ msg.address_lo &= ~(HT_IRQ_LOW_DEST_ID_MASK);
++ msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK);
++
++ msg.address_lo |= 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;
++
++ 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);
++ irq_desc[irq].affinity = mask;
++}
++#endif
++
++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)
++{
++ 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;
++}
++#endif /* CONFIG_HT_IRQ */
++
++/* --------------------------------------------------------------------------
++ ACPI-based IOAPIC Configuration
++ -------------------------------------------------------------------------- */
++
++#ifdef CONFIG_ACPI
++
++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;
++
++ /*
++ * 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.
++ */
++
++ if (physids_empty(apic_id_map))
++ apic_id_map = ioapic_phys_id_map(phys_cpu_present_map);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(ioapic, 0);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ 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;
++ }
++
++ /*
++ * 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)) {
++
++ for (i = 0; i < get_physical_broadcast(); i++) {
++ if (!check_apicid_used(apic_id_map, i))
++ break;
++ }
++
++ if (i == get_physical_broadcast())
++ panic("Max apic_id exceeded!\n");
++
++ printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, "
++ "trying %d\n", ioapic, apic_id, i);
++
++ apic_id = i;
++ }
++
++ tmp = apicid_to_cpu_present(apic_id);
++ physids_or(apic_id_map, apic_id_map, tmp);
++
++ 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);
++
++ /* Sanity check */
++ if (reg_00.bits.ID != apic_id) {
++ printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic);
++ return -1;
++ }
++ }
++
++ apic_printk(APIC_VERBOSE, KERN_INFO
++ "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id);
++#endif /* !CONFIG_XEN */
++
++ return apic_id;
++}
++
++
++int __init io_apic_get_version (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return reg_01.bits.version;
++}
++
++
++int __init io_apic_get_redir_entries (int ioapic)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return reg_01.bits.entries;
++}
++
++
++int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low)
++{
++ struct IO_APIC_route_entry entry;
++
++ if (!IO_APIC_IRQ(irq)) {
++ printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
++ ioapic);
++ return -EINVAL;
++ }
++
++ /*
++ * Generate a PCI IRQ routing entry and program the IOAPIC accordingly.
++ * Note that we mask (disable) IRQs now -- these get enabled when the
++ * corresponding device driver registers for this IRQ.
++ */
++
++ memset(&entry,0,sizeof(entry));
++
++ entry.delivery_mode = INT_DELIVERY_MODE;
++ entry.dest_mode = INT_DEST_MODE;
++ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS);
++ entry.trigger = edge_level;
++ entry.polarity = active_high_low;
++ entry.mask = 1;
++
++ /*
++ * IRQs < 16 are already in the irq_2_pin[] map
++ */
++ if (irq >= 16)
++ add_pin_to_irq(irq, ioapic, pin);
++
++ entry.vector = assign_irq_vector(irq);
++
++ apic_printk(APIC_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);
++
++ ioapic_register_intr(irq, entry.vector, edge_level);
++
++ if (!ioapic && (irq < 16))
++ disable_8259A_irq(irq);
++
++ ioapic_write_entry(ioapic, pin, entry);
++
++ return 0;
++}
++
++int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
++{
++ int i;
++
++ if (skip_ioapic_setup)
++ return -1;
++
++ for (i = 0; i < mp_irq_entries; i++)
++ if (mp_irqs[i].mpc_irqtype == mp_INT &&
++ mp_irqs[i].mpc_srcbusirq == bus_irq)
++ break;
++ if (i >= mp_irq_entries)
++ return -1;
++
++ *trigger = irq_trigger(i);
++ *polarity = irq_polarity(i);
++ return 0;
++}
++
++#endif /* CONFIG_ACPI */
++
++#ifndef CONFIG_XEN
++static int __init parse_disable_timer_pin_1(char *arg)
++{
++ disable_timer_pin_1 = 1;
++ return 0;
++}
++early_param("disable_timer_pin_1", parse_disable_timer_pin_1);
++
++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);
++#endif
++
++static int __init parse_noapic(char *arg)
++{
++ /* disable IO-APIC */
++ disable_ioapic_setup();
++ return 0;
++}
++early_param("noapic", parse_noapic);
+diff --git a/arch/x86/kernel/io_apic_64-xen.c b/arch/x86/kernel/io_apic_64-xen.c
+new file mode 100644
+index 0000000..ad00f4d
+--- /dev/null
++++ b/arch/x86/kernel/io_apic_64-xen.c
+@@ -0,0 +1,2385 @@
++/*
++ * 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/pci.h>
++#include <linux/mc146818rtc.h>
++#include <linux/acpi.h>
++#include <linux/sysdev.h>
++#include <linux/msi.h>
++#include <linux/htirq.h>
++#include <linux/dmar.h>
++#include <linux/jiffies.h>
++#ifdef CONFIG_ACPI
++#include <acpi/acpi_bus.h>
++#endif
++#include <linux/bootmem.h>
++
++#include <asm/idle.h>
++#include <asm/io.h>
++#include <asm/smp.h>
++#include <asm/desc.h>
++#include <asm/proto.h>
++#include <asm/acpi.h>
++#include <asm/dma.h>
++#include <asm/nmi.h>
++#include <asm/msidef.h>
++#include <asm/hypertransport.h>
++
++#include <mach_ipi.h>
++#include <mach_apic.h>
++
++struct irq_cfg {
++#ifndef CONFIG_XEN
++ cpumask_t domain;
++ cpumask_t old_domain;
++#endif
++ unsigned move_cleanup_count;
++ u8 vector;
++ u8 move_in_progress : 1;
++};
++
++/* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
++struct irq_cfg irq_cfg[NR_IRQS] __read_mostly;
++
++static int assign_irq_vector(int irq, cpumask_t mask);
++
++#define __apicdebuginit __init
++
++int sis_apic_bug; /* not actually supported, dummy for compile */
++
++#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;
++
++#define clear_IO_APIC() ((void)0)
++#else
++static int no_timer_check;
++
++static int disable_timer_pin_1 __initdata;
++
++int timer_over_8254 __initdata = 1;
++
++/* 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);
++DEFINE_SPINLOCK(vector_lock);
++
++/*
++ * # of IRQ routing registers
++ */
++int nr_ioapic_registers[MAX_IO_APICS];
++
++/* I/O APIC entries */
++struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS];
++int nr_ioapics;
++
++/* MP IRQ source entries */
++struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES];
++
++/* # of MP IRQ source entries */
++int mp_irq_entries;
++
++/*
++ * 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 {
++ short apic, pin, next;
++} irq_2_pin[PIN_MAP_SIZE];
++
++#ifndef CONFIG_XEN
++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);
++}
++#endif
++
++static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg)
++{
++#ifndef CONFIG_XEN
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ return readl(&io_apic->data);
++#else
++ 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;
++#endif
++}
++
++static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value)
++{
++#ifndef CONFIG_XEN
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ writel(reg, &io_apic->index);
++ writel(value, &io_apic->data);
++#else
++ struct physdev_apic apic_op;
++
++ apic_op.apic_physbase = mp_ioapics[apic].mpc_apicaddr;
++ apic_op.reg = reg;
++ apic_op.value = value;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_apic_write, &apic_op));
++#endif
++}
++
++#ifdef CONFIG_XEN
++#define io_apic_modify io_apic_write
++#else
++/*
++ * 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);
++}
++
++static bool io_apic_level_ack_pending(unsigned int irq)
++{
++ struct irq_pin_list *entry;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ entry = irq_2_pin + irq;
++ for (;;) {
++ unsigned int reg;
++ int pin;
++
++ pin = entry->pin;
++ if (pin == -1)
++ break;
++ reg = io_apic_read(entry->apic, 0x10 + pin*2);
++ /* Is the remote IRR bit set? */
++ if ((reg >> 14) & 1) {
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ return true;
++ }
++ if (!entry->next)
++ break;
++ entry = irq_2_pin + entry->next;
++ }
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return false;
++}
++#endif
++
++/*
++ * 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)
++{
++#ifndef CONFIG_XEN
++ struct io_apic __iomem *io_apic = io_apic_base(apic);
++ readl(&io_apic->data);
++#endif
++}
++
++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);
++}
++
++#ifndef CONFIG_XEN
++/*
++ * 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!
++ */
++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);
++}
++
++#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;
++
++ 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;
++ }
++}
++
++static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ unsigned long flags;
++ unsigned int dest;
++ cpumask_t tmp;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ return;
++
++ if (assign_irq_vector(irq, mask))
++ return;
++
++ cpus_and(tmp, cfg->domain, mask);
++ dest = cpu_mask_to_apicid(tmp);
++
++ /*
++ * Only the high 8 bits are valid.
++ */
++ dest = SET_APIC_LOGICAL_ID(dest);
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __target_IO_APIC_irq(irq, dest, cfg->vector);
++ irq_desc[irq].affinity = mask;
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++#endif
++#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;
++
++ BUG_ON(irq >= NR_IRQS);
++ while (entry->next)
++ entry = irq_2_pin + entry->next;
++
++ if (entry->pin != -1) {
++ entry->next = first_free_entry;
++ entry = irq_2_pin + entry->next;
++ if (++first_free_entry >= PIN_MAP_SIZE)
++ panic("io_apic.c: ran out of irq_2_pin entries!");
++ }
++ entry->apic = apic;
++ entry->pin = pin;
++}
++
++#ifndef CONFIG_XEN
++#define __DO_ACTION(R, ACTION, FINAL) \
++ \
++{ \
++ int pin; \
++ struct irq_pin_list *entry = irq_2_pin + irq; \
++ \
++ BUG_ON(irq >= NR_IRQS); \
++ for (;;) { \
++ unsigned int reg; \
++ pin = entry->pin; \
++ if (pin == -1) \
++ break; \
++ reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \
++ reg ACTION; \
++ io_apic_modify(entry->apic, reg); \
++ FINAL; \
++ if (!entry->next) \
++ break; \
++ entry = irq_2_pin + entry->next; \
++ } \
++}
++
++#define DO_ACTION(name,R,ACTION, FINAL) \
++ \
++ static void name##_IO_APIC_irq (unsigned int irq) \
++ __DO_ACTION(R, ACTION, FINAL)
++
++DO_ACTION( __mask, 0, |= 0x00010000, io_apic_sync(entry->apic) )
++ /* mask = 1 */
++DO_ACTION( __unmask, 0, &= 0xfffeffff, )
++ /* mask = 0 */
++
++static void mask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __mask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++static void unmask_IO_APIC_irq (unsigned int irq)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ __unmask_IO_APIC_irq(irq);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++}
++
++static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin)
++{
++ struct IO_APIC_route_entry entry;
++
++ /* 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);
++}
++
++static void clear_IO_APIC (void)
++{
++ int apic, pin;
++
++ for (apic = 0; apic < nr_ioapics; apic++)
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++)
++ clear_IO_APIC_pin(apic, pin);
++}
++
++#endif /* !CONFIG_XEN */
++
++int skip_ioapic_setup;
++int ioapic_force;
++
++static int __init parse_noapic(char *str)
++{
++ disable_ioapic_setup();
++ return 0;
++}
++early_param("noapic", parse_noapic);
++
++#ifndef CONFIG_XEN
++/* Actually the next is obsolete, but keep it for paranoid reasons -AK */
++static int __init disable_timer_pin_setup(char *arg)
++{
++ disable_timer_pin_1 = 1;
++ return 1;
++}
++__setup("disable_timer_pin_1", disable_timer_pin_setup);
++
++static int __init setup_disable_8254_timer(char *s)
++{
++ timer_over_8254 = -1;
++ return 1;
++}
++static int __init setup_enable_8254_timer(char *s)
++{
++ timer_over_8254 = 2;
++ return 1;
++}
++
++__setup("disable_8254_timer", setup_disable_8254_timer);
++__setup("enable_8254_timer", setup_enable_8254_timer);
++#endif /* !CONFIG_XEN */
++
++
++/*
++ * Find the IRQ entry number of a certain pin.
++ */
++static int find_irq_entry(int apic, int pin, int type)
++{
++ int i;
++
++ for (i = 0; i < mp_irq_entries; i++)
++ if (mp_irqs[i].mpc_irqtype == type &&
++ (mp_irqs[i].mpc_dstapic == mp_ioapics[apic].mpc_apicid ||
++ mp_irqs[i].mpc_dstapic == MP_APIC_ALL) &&
++ mp_irqs[i].mpc_dstirq == pin)
++ return i;
++
++ return -1;
++}
++
++#ifndef CONFIG_XEN
++/*
++ * Find the pin to which IRQ[irq] (ISA) is connected
++ */
++static int __init find_isa_irq_pin(int irq, int type)
++{
++ int i;
++
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
++
++ if (test_bit(lbus, mp_bus_not_pci) &&
++ (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 (test_bit(lbus, mp_bus_not_pci) &&
++ (mp_irqs[i].mpc_irqtype == type) &&
++ (mp_irqs[i].mpc_srcbusirq == irq))
++ break;
++ }
++ if (i < mp_irq_entries) {
++ int apic;
++ for(apic = 0; apic < nr_ioapics; apic++) {
++ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic)
++ return apic;
++ }
++ }
++
++ return -1;
++}
++#endif
++
++/*
++ * Find a specific PCI IRQ entry.
++ * Not an __init, possibly needed by modules
++ */
++static int pin_2_irq(int idx, int apic, int pin);
++
++int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin)
++{
++ int apic, i, best_guess = -1;
++
++ apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n",
++ bus, slot, pin);
++ if (mp_bus_id_to_pci_bus[bus] == -1) {
++ apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
++ return -1;
++ }
++ for (i = 0; i < mp_irq_entries; i++) {
++ int lbus = mp_irqs[i].mpc_srcbus;
++
++ for (apic = 0; apic < nr_ioapics; apic++)
++ if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic ||
++ mp_irqs[i].mpc_dstapic == MP_APIC_ALL)
++ break;
++
++ if (!test_bit(lbus, mp_bus_not_pci) &&
++ !mp_irqs[i].mpc_irqtype &&
++ (bus == lbus) &&
++ (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) {
++ int irq = pin_2_irq(i,apic,mp_irqs[i].mpc_dstirq);
++
++ if (!(apic || IO_APIC_IRQ(irq)))
++ continue;
++
++ if (pin == (mp_irqs[i].mpc_srcbusirq & 3))
++ return irq;
++ /*
++ * Use the first all-but-pin matching entry as a
++ * best-guess fuzzy result for broken mptables.
++ */
++ if (best_guess < 0)
++ best_guess = irq;
++ }
++ }
++ BUG_ON(best_guess >= NR_IRQS);
++ return best_guess;
++}
++
++/* 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)
++
++static int 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 */
++ 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;
++ 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 */
++ 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;
++ 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");
++
++ 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;
++}
++
++static int __assign_irq_vector(int irq, cpumask_t mask)
++{
++ struct physdev_irq irq_op;
++ struct irq_cfg *cfg;
++
++ BUG_ON((unsigned)irq >= NR_IRQS);
++ cfg = &irq_cfg[irq];
++
++ if ((cfg->move_in_progress) || cfg->move_cleanup_count)
++ return -EBUSY;
++
++ if (cfg->vector)
++ return 0;
++
++ irq_op.irq = irq;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op))
++ return -ENOSPC;
++
++ cfg->vector = irq_op.vector;
++
++ return 0;
++}
++
++static int assign_irq_vector(int irq, cpumask_t mask)
++{
++ int err;
++ unsigned long flags;
++
++ spin_lock_irqsave(&vector_lock, flags);
++ err = __assign_irq_vector(irq, mask);
++ spin_unlock_irqrestore(&vector_lock, flags);
++ return err;
++}
++
++#ifndef CONFIG_XEN
++static void __clear_irq_vector(int irq)
++{
++ struct irq_cfg *cfg;
++ cpumask_t mask;
++ int cpu, vector;
++
++ BUG_ON((unsigned)irq >= NR_IRQS);
++ cfg = &irq_cfg[irq];
++ BUG_ON(!cfg->vector);
++
++ vector = cfg->vector;
++ cpus_and(mask, cfg->domain, cpu_online_map);
++ for_each_cpu_mask(cpu, mask)
++ per_cpu(vector_irq, cpu)[vector] = -1;
++
++ cfg->vector = 0;
++ cpus_clear(cfg->domain);
++}
++
++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_IRQS; ++irq) {
++ if (!cpu_isset(cpu, irq_cfg[irq].domain))
++ continue;
++ vector = irq_cfg[irq].vector;
++ 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_cfg[irq].domain))
++ per_cpu(vector_irq, cpu)[vector] = -1;
++ }
++}
++
++static struct irq_chip ioapic_chip;
++
++static void ioapic_register_intr(int irq, unsigned long trigger)
++{
++ if (trigger) {
++ irq_desc[irq].status |= IRQ_LEVEL;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_fasteoi_irq, "fasteoi");
++ } else {
++ irq_desc[irq].status &= ~IRQ_LEVEL;
++ set_irq_chip_and_handler_name(irq, &ioapic_chip,
++ handle_edge_irq, "edge");
++ }
++}
++#else
++#define ioapic_register_intr(irq,trigger) ((void)0)
++#endif /* !CONFIG_XEN */
++
++static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq,
++ int trigger, int polarity)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ struct IO_APIC_route_entry entry;
++ cpumask_t mask;
++
++ if (!IO_APIC_IRQ(irq))
++ return;
++
++ mask = TARGET_CPUS;
++ if (assign_irq_vector(irq, mask))
++ return;
++
++#ifndef CONFIG_XEN
++ cpus_and(mask, cfg->domain, mask);
++#endif
++
++ apic_printk(APIC_VERBOSE,KERN_DEBUG
++ "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> "
++ "IRQ %d Mode:%i Active:%i)\n",
++ apic, mp_ioapics[apic].mpc_apicid, pin, cfg->vector,
++ irq, trigger, polarity);
++
++ /*
++ * 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.dest = cpu_mask_to_apicid(mask);
++ entry.mask = 0; /* enable IRQ */
++ entry.trigger = trigger;
++ entry.polarity = polarity;
++ entry.vector = cfg->vector;
++
++ /* Mask level triggered irqs.
++ * Use IRQ_DELAYED_DISABLE for edge triggered irqs.
++ */
++ if (trigger)
++ entry.mask = 1;
++
++ ioapic_register_intr(irq, trigger);
++ if (irq < 16)
++ disable_8259A_irq(irq);
++
++ ioapic_write_entry(apic, pin, entry);
++}
++
++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) {
++ 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;
++ }
++ if (!first_notcon) {
++ apic_printk(APIC_VERBOSE, " not connected.\n");
++ first_notcon = 1;
++ }
++
++ irq = pin_2_irq(idx, apic, pin);
++ add_pin_to_irq(irq, apic, pin);
++
++ setup_IO_APIC_irq(apic, pin, irq,
++ irq_trigger(idx), irq_polarity(idx));
++ }
++ }
++
++ if (!first_notcon)
++ apic_printk(APIC_VERBOSE, " not connected.\n");
++}
++
++#ifndef CONFIG_XEN
++/*
++ * Set up the 8259A-master output pin as broadcast to all
++ * CPUs.
++ */
++static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector)
++{
++ struct IO_APIC_route_entry entry;
++
++ memset(&entry, 0, sizeof(entry));
++
++ disable_8259A_irq(0);
++
++ /* mask LVT0 */
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++
++ /*
++ * We use logical delivery to get the timer IRQ
++ * to the first CPU.
++ */
++ entry.dest_mode = INT_DEST_MODE;
++ entry.mask = 0; /* unmask IRQ now */
++ entry.dest = 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 ...
++ */
++ set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge");
++
++ /*
++ * Add it to the IO-APIC irq-routing table:
++ */
++ ioapic_write_entry(apic, pin, entry);
++
++ enable_8259A_irq(0);
++}
++
++void __apicdebuginit print_IO_APIC(void)
++{
++ int apic, i;
++ union IO_APIC_reg_00 reg_00;
++ union IO_APIC_reg_01 reg_01;
++ union IO_APIC_reg_02 reg_02;
++ unsigned long flags;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries);
++ for (i = 0; i < nr_ioapics; i++)
++ printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n",
++ mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]);
++
++ /*
++ * We are a bit conservative about what we expect. We have to
++ * know about every hardware change ASAP.
++ */
++ printk(KERN_INFO "testing the IO APIC.......................\n");
++
++ for (apic = 0; apic < nr_ioapics; apic++) {
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(apic, 0);
++ reg_01.raw = io_apic_read(apic, 1);
++ if (reg_01.bits.version >= 0x10)
++ reg_02.raw = io_apic_read(apic, 2);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ printk("\n");
++ printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mpc_apicid);
++ printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw);
++ printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID);
++
++ printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)®_01);
++ printk(KERN_DEBUG "....... : max redirection entries: %04X\n", reg_01.bits.entries);
++
++ 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 >= 0x10) {
++ printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw);
++ printk(KERN_DEBUG "....... : arbitration: %02X\n", reg_02.bits.arbitration);
++ }
++
++ printk(KERN_DEBUG ".... IRQ redirection table:\n");
++
++ printk(KERN_DEBUG " NR Dst Mask Trig IRR Pol"
++ " Stat Dmod Deli Vect: \n");
++
++ for (i = 0; i <= reg_01.bits.entries; i++) {
++ struct IO_APIC_route_entry entry;
++
++ entry = ioapic_read_entry(apic, i);
++
++ printk(KERN_DEBUG " %02x %03X ",
++ i,
++ entry.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
++ );
++ }
++ }
++ 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");
++ }
++
++ printk(KERN_INFO ".................................... done.\n");
++
++ return;
++}
++
++static __apicdebuginit void print_APIC_bitfield (int base)
++{
++ unsigned int v;
++ int i, j;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
++ for (i = 0; i < 8; i++) {
++ v = apic_read(base + i*0x10);
++ for (j = 0; j < 32; j++) {
++ if (v & (1<<j))
++ printk("1");
++ else
++ printk("0");
++ }
++ printk("\n");
++ }
++}
++
++void __apicdebuginit print_local_APIC(void * dummy)
++{
++ unsigned int v, ver, maxlvt;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
++ smp_processor_id(), hard_smp_processor_id());
++ printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(read_apic_id()));
++ v = apic_read(APIC_LVR);
++ printk(KERN_INFO "... APIC VERSION: %08x\n", v);
++ ver = GET_APIC_VERSION(v);
++ maxlvt = lapic_get_maxlvt();
++
++ v = apic_read(APIC_TASKPRI);
++ printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK);
++
++ v = apic_read(APIC_ARBPRI);
++ printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v,
++ v & APIC_ARBPRI_MASK);
++ v = apic_read(APIC_PROCPRI);
++ printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v);
++
++ v = apic_read(APIC_EOI);
++ printk(KERN_DEBUG "... APIC EOI: %08x\n", v);
++ v = apic_read(APIC_RRR);
++ printk(KERN_DEBUG "... APIC RRR: %08x\n", v);
++ v = apic_read(APIC_LDR);
++ printk(KERN_DEBUG "... APIC LDR: %08x\n", v);
++ v = apic_read(APIC_DFR);
++ printk(KERN_DEBUG "... APIC DFR: %08x\n", v);
++ v = apic_read(APIC_SPIV);
++ printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
++
++ printk(KERN_DEBUG "... APIC ISR field:\n");
++ print_APIC_bitfield(APIC_ISR);
++ printk(KERN_DEBUG "... APIC TMR field:\n");
++ print_APIC_bitfield(APIC_TMR);
++ printk(KERN_DEBUG "... APIC IRR field:\n");
++ print_APIC_bitfield(APIC_IRR);
++
++ v = apic_read(APIC_ESR);
++ printk(KERN_DEBUG "... APIC ESR: %08x\n", v);
++
++ v = apic_read(APIC_ICR);
++ printk(KERN_DEBUG "... APIC ICR: %08x\n", v);
++ v = apic_read(APIC_ICR2);
++ printk(KERN_DEBUG "... APIC ICR2: %08x\n", v);
++
++ v = apic_read(APIC_LVTT);
++ printk(KERN_DEBUG "... APIC LVTT: %08x\n", v);
++
++ if (maxlvt > 3) { /* PC is LVT#4. */
++ v = apic_read(APIC_LVTPC);
++ printk(KERN_DEBUG "... APIC LVTPC: %08x\n", v);
++ }
++ v = apic_read(APIC_LVT0);
++ printk(KERN_DEBUG "... APIC LVT0: %08x\n", v);
++ v = apic_read(APIC_LVT1);
++ printk(KERN_DEBUG "... APIC LVT1: %08x\n", v);
++
++ if (maxlvt > 2) { /* ERR is LVT#3. */
++ v = apic_read(APIC_LVTERR);
++ printk(KERN_DEBUG "... APIC LVTERR: %08x\n", v);
++ }
++
++ v = apic_read(APIC_TMICT);
++ printk(KERN_DEBUG "... APIC TMICT: %08x\n", v);
++ v = apic_read(APIC_TMCCT);
++ printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
++ v = apic_read(APIC_TDCR);
++ printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
++ printk("\n");
++}
++
++void print_all_local_APICs (void)
++{
++ on_each_cpu(print_local_APIC, NULL, 1, 1);
++}
++
++void __apicdebuginit print_PIC(void)
++{
++ unsigned int v;
++ unsigned long flags;
++
++ if (apic_verbosity == APIC_QUIET)
++ return;
++
++ printk(KERN_DEBUG "\nprinting PIC contents\n");
++
++ spin_lock_irqsave(&i8259A_lock, flags);
++
++ v = inb(0xa1) << 8 | inb(0x21);
++ printk(KERN_DEBUG "... PIC IMR: %04x\n", v);
++
++ v = inb(0xa0) << 8 | inb(0x20);
++ printk(KERN_DEBUG "... PIC IRR: %04x\n", v);
++
++ outb(0x0b,0xa0);
++ outb(0x0b,0x20);
++ v = inb(0xa0) << 8 | inb(0x20);
++ outb(0x0a,0xa0);
++ outb(0x0a,0x20);
++
++ spin_unlock_irqrestore(&i8259A_lock, flags);
++
++ printk(KERN_DEBUG "... PIC ISR: %04x\n", v);
++
++ v = inb(0x4d1) << 8 | inb(0x4d0);
++ printk(KERN_DEBUG "... PIC ELCR: %04x\n", v);
++}
++#endif /* !CONFIG_XEN */
++
++void __init enable_IO_APIC(void)
++{
++ union IO_APIC_reg_01 reg_01;
++#ifndef CONFIG_XEN
++ int i8259_apic, i8259_pin;
++#endif
++ int i, apic;
++ unsigned long flags;
++
++ for (i = 0; i < PIN_MAP_SIZE; i++) {
++ irq_2_pin[i].pin = -1;
++ irq_2_pin[i].next = 0;
++ }
++
++ /*
++ * The number of IO-APIC IRQ registers (== #pins):
++ */
++ for (apic = 0; apic < nr_ioapics; apic++) {
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(apic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++ nr_ioapic_registers[apic] = reg_01.bits.entries+1;
++ }
++#ifndef CONFIG_XEN
++ for(apic = 0; apic < nr_ioapics; apic++) {
++ int pin;
++ /* See if any of the pins is in ExtINT mode */
++ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) {
++ struct IO_APIC_route_entry entry;
++ 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.
++ */
++ if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) {
++ ioapic_i8259.apic = apic;
++ ioapic_i8259.pin = pin;
++ goto found_i8259;
++ }
++ }
++ }
++ found_i8259:
++ /* Look to see what if the MP table has reported the ExtINT */
++ i8259_pin = find_isa_irq_pin(0, mp_ExtINT);
++ i8259_apic = find_isa_irq_apic(0, mp_ExtINT);
++ /* Trust the MP table if nothing is setup in the hardware */
++ if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) {
++ printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n");
++ ioapic_i8259.pin = i8259_pin;
++ ioapic_i8259.apic = i8259_apic;
++ }
++ /* Complain if the MP table and the hardware disagree */
++ if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) &&
++ (i8259_pin >= 0) && (ioapic_i8259.pin >= 0))
++ {
++ printk(KERN_WARNING "ExtINT in hardware and MP table differ\n");
++ }
++#endif
++
++ /*
++ * Do not trust the IO-APIC being empty at bootup
++ */
++ clear_IO_APIC();
++}
++
++/*
++ * Not an __init, needed by the reboot code
++ */
++void disable_IO_APIC(void)
++{
++ /*
++ * Clear the IO-APIC before rebooting:
++ */
++ clear_IO_APIC();
++
++#ifndef CONFIG_XEN
++ /*
++ * If the i8259 is routed through an IOAPIC
++ * Put that IOAPIC in virtual wire mode
++ * so legacy interrupts can be delivered.
++ */
++ if (ioapic_i8259.pin != -1) {
++ struct IO_APIC_route_entry entry;
++
++ 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 = GET_APIC_ID(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
++}
++
++/*
++ * There is a nasty bug in some older SMP boards, their mptable lies
++ * about the timer IRQ. We do the following to work around the situation:
++ *
++ * - timer IRQ defaults to IO-APIC IRQ
++ * - if this function detects that timer IRQs are defunct, then we fall
++ * back to ISA timer IRQs
++ */
++#ifndef CONFIG_XEN
++static int __init timer_irq_works(void)
++{
++ unsigned long t1 = jiffies;
++ unsigned long flags;
++
++ local_save_flags(flags);
++ local_irq_enable();
++ /* Let ten ticks pass... */
++ mdelay((10 * 1000) / HZ);
++ local_irq_restore(flags);
++
++ /*
++ * Expect a few ticks at least, to be sure some possible
++ * glue logic does not lock up after one or two first
++ * ticks in a non-ExtINT mode. Also the local APIC
++ * might have cached one ExtINT interrupt. Finally, at
++ * least one tick may be lost due to delays.
++ */
++
++ /* jiffies wrap? */
++ if (time_after(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_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;
++}
++
++static int ioapic_retrigger_irq(unsigned int irq)
++{
++ struct irq_cfg *cfg = &irq_cfg[irq];
++ cpumask_t mask;
++ unsigned long flags;
++
++ spin_lock_irqsave(&vector_lock, flags);
++ mask = cpumask_of_cpu(first_cpu(cfg->domain));
++ send_IPI_mask(mask, cfg->vector);
++ spin_unlock_irqrestore(&vector_lock, flags);
++
++ 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.
++ */
++
++#ifdef CONFIG_SMP
++asmlinkage void smp_irq_move_cleanup_interrupt(void)
++{
++ unsigned vector, me;
++ ack_APIC_irq();
++ exit_idle();
++ irq_enter();
++
++ me = smp_processor_id();
++ for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
++ unsigned int irq;
++ struct irq_desc *desc;
++ struct irq_cfg *cfg;
++ irq = __get_cpu_var(vector_irq)[vector];
++ if (irq >= NR_IRQS)
++ continue;
++
++ desc = irq_desc + irq;
++ cfg = irq_cfg + irq;
++ spin_lock(&desc->lock);
++ if (!cfg->move_cleanup_count)
++ goto unlock;
++
++ if ((vector == cfg->vector) && cpu_isset(me, cfg->domain))
++ goto unlock;
++
++ __get_cpu_var(vector_irq)[vector] = -1;
++ cfg->move_cleanup_count--;
++unlock:
++ spin_unlock(&desc->lock);
++ }
++
++ irq_exit();
++}
++
++static void irq_complete_move(unsigned int irq)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ unsigned vector, me;
++
++ if (likely(!cfg->move_in_progress))
++ return;
++
++ vector = ~get_irq_regs()->orig_ax;
++ me = smp_processor_id();
++ if ((vector == cfg->vector) && cpu_isset(me, cfg->domain)) {
++ cpumask_t cleanup_mask;
++
++ cpus_and(cleanup_mask, cfg->old_domain, cpu_online_map);
++ cfg->move_cleanup_count = cpus_weight(cleanup_mask);
++ send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
++ cfg->move_in_progress = 0;
++ }
++}
++#else
++static inline void irq_complete_move(unsigned int irq) {}
++#endif
++
++static void ack_apic_edge(unsigned int irq)
++{
++ irq_complete_move(irq);
++ move_native_irq(irq);
++ ack_APIC_irq();
++}
++
++static void ack_apic_level(unsigned int irq)
++{
++ int do_unmask_irq = 0;
++
++ irq_complete_move(irq);
++#ifdef CONFIG_GENERIC_PENDING_IRQ
++ /* 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
++
++ /*
++ * We must acknowledge the irq before we move it or the acknowledge will
++ * not propagate properly.
++ */
++ ack_APIC_irq();
++
++ /* Now we can move and renable the irq */
++ if (unlikely(do_unmask_irq)) {
++ /* Only migrate the irq if the ack has been received.
++ *
++ * On rare occasions the broadcast level triggered ack gets
++ * delayed going to ioapics, and if we reprogram the
++ * vector while Remote IRR is still set the irq will never
++ * fire again.
++ *
++ * To prevent this scenario we read the Remote IRR bit
++ * of the ioapic. This has two effects.
++ * - On any sane system the read of the ioapic will
++ * flush writes (and acks) going to the ioapic from
++ * this cpu.
++ * - We get to see if the ACK has actually been delivered.
++ *
++ * Based on failed experiments of reprogramming the
++ * ioapic entry from outside of irq context starting
++ * with masking the ioapic entry and then polling until
++ * Remote IRR was clear before reprogramming the
++ * ioapic I don't trust the Remote IRR bit to be
++ * completey accurate.
++ *
++ * However there appears to be no other way to plug
++ * this race, so if the Remote IRR bit is not
++ * accurate and is causing problems then it is a hardware bug
++ * and you can go talk to the chipset vendor about it.
++ */
++ if (!io_apic_level_ack_pending(irq))
++ move_masked_irq(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_irq,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++#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++) {
++ if (IO_APIC_IRQ(irq) && !irq_cfg[irq].vector) {
++ /*
++ * 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
++ }
++ }
++}
++
++#ifndef CONFIG_XEN
++static void enable_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED);
++}
++
++static void disable_lapic_irq (unsigned int irq)
++{
++ unsigned long v;
++
++ v = apic_read(APIC_LVT0);
++ apic_write(APIC_LVT0, v | APIC_LVT_MASKED);
++}
++
++static void ack_lapic_irq (unsigned int irq)
++{
++ ack_APIC_irq();
++}
++
++static void end_lapic_irq (unsigned int i) { /* nothing */ }
++
++static struct hw_interrupt_type lapic_irq_type __read_mostly = {
++ .name = "local-APIC",
++ .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 __init setup_nmi(void)
++{
++ /*
++ * Dirty trick to enable the NMI watchdog ...
++ * We put the 8259A master into AEOI mode and
++ * unmask on all local APICs LVT0 as NMI.
++ *
++ * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire')
++ * is from Maciej W. Rozycki - so we do not have to EOI from
++ * the NMI handler or the timer interrupt.
++ */
++ printk(KERN_INFO "activating NMI Watchdog ...");
++
++ enable_NMI_through_LVT0();
++
++ printk(" done.\n");
++}
++
++/*
++ * This looks a bit hackish but it's about the only one way of sending
++ * a few INTA cycles to 8259As and any associated glue logic. ICR does
++ * not support the ExtINT mode, unfortunately. We need to send these
++ * cycles as some i82489DX-based boards have glue logic that keeps the
++ * 8259A interrupt line asserted until INTA. --macro
++ */
++static inline void __init unlock_ExtINT_logic(void)
++{
++ int apic, pin, i;
++ struct IO_APIC_route_entry entry0, entry1;
++ unsigned char save_control, save_freq_select;
++
++ pin = find_isa_irq_pin(8, mp_INT);
++ apic = find_isa_irq_apic(8, mp_INT);
++ if (pin == -1)
++ return;
++
++ entry0 = ioapic_read_entry(apic, pin);
++
++ 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 = hard_smp_processor_id();
++ entry1.delivery_mode = dest_ExtINT;
++ entry1.polarity = entry0.polarity;
++ entry1.trigger = 0;
++ entry1.vector = 0;
++
++ ioapic_write_entry(apic, pin, entry1);
++
++ 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);
++
++ ioapic_write_entry(apic, pin, entry0);
++}
++
++/*
++ * This code may look a bit paranoid, but it's supposed to cooperate with
++ * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ
++ * is so screwy. Thanks to Brian Perkins for testing/hacking this beast
++ * fanatically on his truly buggy board.
++ *
++ * FIXME: really need to revamp this for modern platforms only.
++ */
++static inline void __init check_timer(void)
++{
++ struct irq_cfg *cfg = irq_cfg + 0;
++ int apic1, pin1, apic2, pin2;
++ unsigned long flags;
++
++ local_irq_save(flags);
++
++ /*
++ * get/set the timer IRQ vector:
++ */
++ disable_8259A_irq(0);
++ assign_irq_vector(0, TARGET_CPUS);
++
++ /*
++ * Subtle, code in do_timer_interrupt() expects an AEOI
++ * mode for the 8259A whenever interrupts are routed
++ * through I/O APICs. Also IRQ0 has to be enabled in
++ * the 8259A which implies the virtual wire has to be
++ * disabled in the local APIC.
++ */
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT);
++ init_8259A(1);
++ if (timer_over_8254 > 0)
++ enable_8259A_irq(0);
++
++ pin1 = find_isa_irq_pin(0, mp_INT);
++ apic1 = find_isa_irq_apic(0, mp_INT);
++ pin2 = ioapic_i8259.pin;
++ apic2 = ioapic_i8259.apic;
++
++ apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n",
++ cfg->vector, apic1, pin1, apic2, pin2);
++
++ if (pin1 != -1) {
++ /*
++ * Ok, does IRQ0 through the IOAPIC work?
++ */
++ unmask_IO_APIC_irq(0);
++ if (!no_timer_check && timer_irq_works()) {
++ nmi_watchdog_default();
++ if (nmi_watchdog == NMI_IO_APIC) {
++ disable_8259A_irq(0);
++ setup_nmi();
++ enable_8259A_irq(0);
++ }
++ if (disable_timer_pin_1 > 0)
++ clear_IO_APIC_pin(0, pin1);
++ goto out;
++ }
++ clear_IO_APIC_pin(apic1, pin1);
++ apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: 8254 timer not "
++ "connected to IO-APIC\n");
++ }
++
++ apic_printk(APIC_VERBOSE,KERN_INFO "...trying to set up timer (IRQ0) "
++ "through the 8259A ... ");
++ if (pin2 != -1) {
++ apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...",
++ apic2, pin2);
++ /*
++ * legacy devices should be connected to IO APIC #0
++ */
++ setup_ExtINT_IRQ0_pin(apic2, pin2, cfg->vector);
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ nmi_watchdog_default();
++ if (nmi_watchdog == NMI_IO_APIC) {
++ setup_nmi();
++ }
++ goto out;
++ }
++ /*
++ * Cleanup, just in case ...
++ */
++ clear_IO_APIC_pin(apic2, pin2);
++ }
++ apic_printk(APIC_VERBOSE," failed.\n");
++
++ if (nmi_watchdog == NMI_IO_APIC) {
++ printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n");
++ nmi_watchdog = 0;
++ }
++
++ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ...");
++
++ disable_8259A_irq(0);
++ irq_desc[0].chip = &lapic_irq_type;
++ apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector); /* Fixed mode */
++ enable_8259A_irq(0);
++
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ goto out;
++ }
++ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector);
++ apic_printk(APIC_VERBOSE," failed.\n");
++
++ apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ...");
++
++ init_8259A(0);
++ make_8259A_irq(0);
++ apic_write(APIC_LVT0, APIC_DM_EXTINT);
++
++ unlock_ExtINT_logic();
++
++ if (timer_irq_works()) {
++ apic_printk(APIC_VERBOSE," works.\n");
++ goto out;
++ }
++ apic_printk(APIC_VERBOSE," failed :(.\n");
++ panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n");
++out:
++ local_irq_restore(flags);
++}
++
++static int __init notimercheck(char *s)
++{
++ no_timer_check = 1;
++ return 1;
++}
++__setup("no_timer_check", notimercheck);
++#else
++#define check_timer() ((void)0)
++int timer_uses_ioapic_pin_0 = 0;
++#endif /* !CONFIG_XEN */
++
++/*
++ *
++ * IRQs that are handled by the PIC in the MPS IOAPIC case.
++ * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ.
++ * Linux doesn't really care, as it's not actually used
++ * for any interrupt handling anyway.
++ */
++#define PIC_IRQS (1<<2)
++
++void __init setup_IO_APIC(void)
++{
++ enable_IO_APIC();
++
++ if (acpi_ioapic)
++ io_apic_irqs = ~0; /* all IRQs go through IOAPIC */
++ else
++ io_apic_irqs = ~PIC_IRQS;
++
++ apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n");
++
++#ifndef CONFIG_XEN
++ sync_Arb_IDs();
++#endif /* !CONFIG_XEN */
++ setup_IO_APIC_irqs();
++ init_IO_APIC_traps();
++ check_timer();
++ if (!acpi_ioapic)
++ print_IO_APIC();
++}
++
++struct sysfs_ioapic_data {
++ struct sys_device dev;
++ struct IO_APIC_route_entry entry[0];
++};
++static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS];
++
++static int ioapic_suspend(struct sys_device *dev, pm_message_t state)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ 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 ++ )
++ *entry = ioapic_read_entry(dev->id, i);
++
++ return 0;
++}
++
++static int ioapic_resume(struct sys_device *dev)
++{
++ struct IO_APIC_route_entry *entry;
++ struct sysfs_ioapic_data *data;
++ unsigned long flags;
++ union IO_APIC_reg_00 reg_00;
++ int i;
++
++ data = container_of(dev, struct sysfs_ioapic_data, dev);
++ entry = data->entry;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_00.raw = io_apic_read(dev->id, 0);
++ if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) {
++ reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid;
++ io_apic_write(dev->id, 0, reg_00.raw);
++ }
++ 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;
++}
++
++static struct sysdev_class ioapic_sysdev_class = {
++ .name = "ioapic",
++ .suspend = ioapic_suspend,
++ .resume = ioapic_resume,
++};
++
++static int __init ioapic_init_sysfs(void)
++{
++ struct sys_device * dev;
++ int i, size, error;
++
++ error = sysdev_class_register(&ioapic_sysdev_class);
++ if (error)
++ return error;
++
++ for (i = 0; i < nr_ioapics; i++ ) {
++ size = sizeof(struct sys_device) + nr_ioapic_registers[i]
++ * sizeof(struct IO_APIC_route_entry);
++ mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL);
++ if (!mp_ioapic_data[i]) {
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ dev = &mp_ioapic_data[i]->dev;
++ dev->id = i;
++ dev->cls = &ioapic_sysdev_class;
++ error = sysdev_register(dev);
++ if (error) {
++ kfree(mp_ioapic_data[i]);
++ mp_ioapic_data[i] = NULL;
++ printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i);
++ continue;
++ }
++ }
++
++ return 0;
++}
++
++device_initcall(ioapic_init_sysfs);
++
++#ifndef CONFIG_XEN
++/*
++ * Dynamic irq allocate and deallocation
++ */
++int create_irq(void)
++{
++ /* Allocate an unused irq */
++ int irq;
++ int new;
++ unsigned long flags;
++
++ irq = -ENOSPC;
++ spin_lock_irqsave(&vector_lock, flags);
++ for (new = (NR_IRQS - 1); new >= 0; new--) {
++ if (platform_legacy_irq(new))
++ continue;
++ if (irq_cfg[new].vector != 0)
++ continue;
++ if (__assign_irq_vector(new, TARGET_CPUS) == 0)
++ irq = new;
++ break;
++ }
++ spin_unlock_irqrestore(&vector_lock, flags);
++
++ if (irq >= 0) {
++ dynamic_irq_init(irq);
++ }
++ return irq;
++}
++
++void destroy_irq(unsigned int irq)
++{
++ unsigned long flags;
++
++ dynamic_irq_cleanup(irq);
++
++ spin_lock_irqsave(&vector_lock, flags);
++ __clear_irq_vector(irq);
++ spin_unlock_irqrestore(&vector_lock, flags);
++}
++#endif
++
++/*
++ * MSI message composition
++ */
++#if defined(CONFIG_PCI_MSI) && !defined(CONFIG_XEN)
++static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ int err;
++ unsigned dest;
++ cpumask_t tmp;
++
++ tmp = TARGET_CPUS;
++ err = assign_irq_vector(irq, tmp);
++ if (!err) {
++ cpus_and(tmp, cfg->domain, tmp);
++ 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(cfg->vector);
++ }
++ return err;
++}
++
++#ifdef CONFIG_SMP
++static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ struct msi_msg msg;
++ unsigned int dest;
++ cpumask_t tmp;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ return;
++
++ if (assign_irq_vector(irq, mask))
++ return;
++
++ cpus_and(tmp, cfg->domain, mask);
++ dest = cpu_mask_to_apicid(tmp);
++
++ read_msi_msg(irq, &msg);
++
++ msg.data &= ~MSI_DATA_VECTOR_MASK;
++ msg.data |= MSI_DATA_VECTOR(cfg->vector);
++ msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
++ msg.address_lo |= MSI_ADDR_DEST_ID(dest);
++
++ write_msi_msg(irq, &msg);
++ irq_desc[irq].affinity = 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(struct pci_dev *dev, struct msi_desc *desc)
++{
++ struct msi_msg msg;
++ int irq, ret;
++ irq = create_irq();
++ if (irq < 0)
++ return irq;
++
++ ret = msi_compose_msg(dev, irq, &msg);
++ if (ret < 0) {
++ destroy_irq(irq);
++ return ret;
++ }
++
++ set_irq_msi(irq, desc);
++ 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)
++{
++ destroy_irq(irq);
++}
++
++#ifdef CONFIG_DMAR
++#ifdef CONFIG_SMP
++static void dmar_msi_set_affinity(unsigned int irq, cpumask_t mask)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ struct msi_msg msg;
++ unsigned int dest;
++ cpumask_t tmp;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ return;
++
++ if (assign_irq_vector(irq, mask))
++ return;
++
++ cpus_and(tmp, cfg->domain, mask);
++ dest = cpu_mask_to_apicid(tmp);
++
++ dmar_msi_read(irq, &msg);
++
++ msg.data &= ~MSI_DATA_VECTOR_MASK;
++ msg.data |= MSI_DATA_VECTOR(cfg->vector);
++ msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK;
++ msg.address_lo |= MSI_ADDR_DEST_ID(dest);
++
++ dmar_msi_write(irq, &msg);
++ irq_desc[irq].affinity = mask;
++}
++#endif /* CONFIG_SMP */
++
++struct irq_chip dmar_msi_type = {
++ .name = "DMAR_MSI",
++ .unmask = dmar_msi_unmask,
++ .mask = dmar_msi_mask,
++ .ack = ack_apic_edge,
++#ifdef CONFIG_SMP
++ .set_affinity = dmar_msi_set_affinity,
++#endif
++ .retrigger = ioapic_retrigger_irq,
++};
++
++int arch_setup_dmar_msi(unsigned int irq)
++{
++ int ret;
++ struct msi_msg msg;
++
++ ret = msi_compose_msg(NULL, irq, &msg);
++ if (ret < 0)
++ return ret;
++ dmar_msi_write(irq, &msg);
++ set_irq_chip_and_handler_name(irq, &dmar_msi_type, handle_edge_irq,
++ "edge");
++ return 0;
++}
++#endif
++
++#endif /* CONFIG_PCI_MSI */
++/*
++ * 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)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ unsigned int dest;
++ cpumask_t tmp;
++
++ cpus_and(tmp, mask, cpu_online_map);
++ if (cpus_empty(tmp))
++ return;
++
++ if (assign_irq_vector(irq, mask))
++ return;
++
++ cpus_and(tmp, cfg->domain, mask);
++ dest = cpu_mask_to_apicid(tmp);
++
++ target_ht_irq(irq, dest, cfg->vector);
++ irq_desc[irq].affinity = 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)
++{
++ struct irq_cfg *cfg = irq_cfg + irq;
++ int err;
++ cpumask_t tmp;
++
++ tmp = TARGET_CPUS;
++ err = assign_irq_vector(irq, tmp);
++ if (!err) {
++ struct ht_irq_msg msg;
++ unsigned dest;
++
++ cpus_and(tmp, cfg->domain, 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(cfg->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 err;
++}
++#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)
++{
++ union IO_APIC_reg_01 reg_01;
++ unsigned long flags;
++
++ spin_lock_irqsave(&ioapic_lock, flags);
++ reg_01.raw = io_apic_read(ioapic, 1);
++ spin_unlock_irqrestore(&ioapic_lock, flags);
++
++ return reg_01.bits.entries;
++}
++
++
++int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
++{
++ if (!IO_APIC_IRQ(irq)) {
++ apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
++ ioapic);
++ return -EINVAL;
++ }
++
++ /*
++ * IRQs < 16 are already in the irq_2_pin[] map
++ */
++ if (irq >= 16)
++ add_pin_to_irq(irq, ioapic, pin);
++
++ setup_IO_APIC_irq(ioapic, pin, irq, triggering, polarity);
++
++ return 0;
++}
++
++
++int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
++{
++ int i;
++
++ if (skip_ioapic_setup)
++ return -1;
++
++ for (i = 0; i < mp_irq_entries; i++)
++ if (mp_irqs[i].mpc_irqtype == mp_INT &&
++ mp_irqs[i].mpc_srcbusirq == bus_irq)
++ break;
++ if (i >= mp_irq_entries)
++ return -1;
++
++ *trigger = irq_trigger(i);
++ *polarity = irq_polarity(i);
++ return 0;
++}
++
++#endif /* CONFIG_ACPI */
++
++#ifndef CONFIG_XEN
++/*
++ * This function currently is only a helper for the i386 smp boot process where
++ * we need to reprogram the ioredtbls to cater for the cpus which have come online
++ * so mask in all cases should simply be TARGET_CPUS
++ */
++#ifdef CONFIG_SMP
++void __init setup_ioapic_dest(void)
++{
++ int pin, ioapic, irq, irq_entry;
++
++ if (skip_ioapic_setup == 1)
++ return;
++
++ for (ioapic = 0; ioapic < nr_ioapics; ioapic++) {
++ for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) {
++ irq_entry = find_irq_entry(ioapic, pin, mp_INT);
++ if (irq_entry == -1)
++ continue;
++ irq = pin_2_irq(irq_entry, ioapic, pin);
++
++ /* 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_cfg[irq].vector)
++ setup_IO_APIC_irq(ioapic, pin, irq,
++ irq_trigger(irq_entry),
++ irq_polarity(irq_entry));
++ else
++ set_ioapic_affinity_irq(irq, TARGET_CPUS);
++ }
++
++ }
++}
++#endif
++
++#define IOAPIC_RESOURCE_NAME_SIZE 11
++
++static struct resource *ioapic_resources;
++
++static struct resource * __init ioapic_setup_resources(void)
++{
++ unsigned long n;
++ struct resource *res;
++ char *mem;
++ int i;
++
++ if (nr_ioapics <= 0)
++ return NULL;
++
++ n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource);
++ n *= nr_ioapics;
++
++ mem = alloc_bootmem(n);
++ res = (void *)mem;
++
++ if (mem != NULL) {
++ mem += sizeof(struct resource) * nr_ioapics;
++
++ for (i = 0; i < nr_ioapics; i++) {
++ res[i].name = mem;
++ res[i].flags = IORESOURCE_MEM | IORESOURCE_BUSY;
++ sprintf(mem, "IOAPIC %u", i);
++ mem += IOAPIC_RESOURCE_NAME_SIZE;
++ }
++ }
++
++ ioapic_resources = res;
++
++ return res;
++}
++
++void __init ioapic_init_mappings(void)
++{
++ unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0;
++ struct resource *ioapic_res;
++ int i;
++
++ ioapic_res = ioapic_setup_resources();
++ for (i = 0; i < nr_ioapics; i++) {
++ if (smp_found_config) {
++ ioapic_phys = mp_ioapics[i].mpc_apicaddr;
++ } else {
++ ioapic_phys = (unsigned long)
++ alloc_bootmem_pages(PAGE_SIZE);
++ ioapic_phys = __pa(ioapic_phys);
++ }
++ set_fixmap_nocache(idx, ioapic_phys);
++ apic_printk(APIC_VERBOSE,
++ "mapped IOAPIC to %016lx (%016lx)\n",
++ __fix_to_virt(idx), ioapic_phys);
++ idx++;
++
++ if (ioapic_res != NULL) {
++ ioapic_res->start = ioapic_phys;
++ ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
++ ioapic_res++;
++ }
++ }
++}
++
++static int __init ioapic_insert_resources(void)
++{
++ int i;
++ struct resource *r = ioapic_resources;
++
++ if (!r) {
++ printk(KERN_ERR
++ "IO APIC resources could be not be allocated.\n");
++ return -1;
++ }
++
++ for (i = 0; i < nr_ioapics; i++) {
++ insert_resource(&iomem_resource, r);
++ r++;
++ }
++
++ return 0;
++}
++
++/* Insert the IO APIC resources after PCI initialization has occured to handle
++ * IO APICS that are mapped in on a BAR in PCI space. */
++late_initcall(ioapic_insert_resources);
++#endif /* !CONFIG_XEN */
+diff --git a/arch/x86/kernel/ioport-xen.c b/arch/x86/kernel/ioport-xen.c
+new file mode 100644
+index 0000000..af83044
+--- /dev/null
++++ b/arch/x86/kernel/ioport-xen.c
+@@ -0,0 +1,112 @@
++/*
++ * This contains the io-permission bitmap code - written by obz, with changes
++ * by Linus. 32/64 bits code unification by Miguel Botón.
++ */
++
++#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/stddef.h>
++#include <linux/slab.h>
++#include <linux/thread_info.h>
++#include <linux/syscalls.h>
++#include <xen/interface/physdev.h>
++
++/* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
++static void set_bitmap(unsigned long *bitmap, unsigned int base,
++ unsigned int extent, int new_value)
++{
++ unsigned int i;
++
++ for (i = base; i < base + extent; i++) {
++ if (new_value)
++ __set_bit(i, bitmap);
++ else
++ __clear_bit(i, bitmap);
++ }
++}
++
++/*
++ * this changes the io permissions bitmap in the current task.
++ */
++asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on)
++{
++ struct thread_struct * t = ¤t->thread;
++ struct physdev_set_iobitmap set_iobitmap;
++
++ if ((from + num <= from) || (from + num > IO_BITMAP_BITS))
++ return -EINVAL;
++ if (turn_on && !capable(CAP_SYS_RAWIO))
++ return -EPERM;
++
++ /*
++ * If it's the first ioperm() call in this thread's lifetime, set the
++ * IO bitmap up. ioperm() is much less timing critical than clone(),
++ * this is why we delay this operation until now:
++ */
++ if (!t->io_bitmap_ptr) {
++ unsigned long *bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
++
++ if (!bitmap)
++ return -ENOMEM;
++
++ memset(bitmap, 0xff, IO_BITMAP_BYTES);
++ t->io_bitmap_ptr = bitmap;
++ set_thread_flag(TIF_IO_BITMAP);
++
++ set_xen_guest_handle(set_iobitmap.bitmap, (char *)bitmap);
++ set_iobitmap.nr_ports = IO_BITMAP_BITS;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &set_iobitmap));
++ }
++
++ set_bitmap(t->io_bitmap_ptr, from, num, !turn_on);
++
++ return 0;
++}
++
++/*
++ * sys_iopl has to be used when you want to access the IO ports
++ * beyond the 0x3ff range: to get the full 65536 ports bitmapped
++ * you'd need 8kB of bitmaps/process, which is a bit excessive.
++ */
++static int do_iopl(unsigned int level, struct thread_struct *t)
++{
++ unsigned int old = t->iopl >> 12;
++
++ if (level > 3)
++ return -EINVAL;
++ /* Trying to gain more privileges? */
++ if (level > old) {
++ if (!capable(CAP_SYS_RAWIO))
++ return -EPERM;
++ }
++
++ return 0;
++}
++
++#ifdef CONFIG_X86_32
++asmlinkage long sys_iopl(unsigned long regsp)
++{
++ struct pt_regs *regs = (struct pt_regs *)®sp;
++ unsigned int level = regs->bx;
++#else
++asmlinkage long sys_iopl(unsigned int level, struct pt_regs *regs)
++{
++#endif
++ struct thread_struct *t = ¤t->thread;
++ int rc;
++
++ rc = do_iopl(level, t);
++ if (rc < 0)
++ goto out;
++
++ t->iopl = level << 12;
++ set_iopl_mask(t->iopl);
++out:
++ return rc;
++}
+diff --git a/arch/x86/kernel/ipi-xen.c b/arch/x86/kernel/ipi-xen.c
+new file mode 100644
+index 0000000..6fb39f3
+--- /dev/null
++++ b/arch/x86/kernel/ipi-xen.c
+@@ -0,0 +1,232 @@
++#include <linux/cpumask.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++
++#include <linux/mm.h>
++#include <linux/delay.h>
++#include <linux/spinlock.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>
++
++#include <asm/smp.h>
++#include <asm/mtrr.h>
++#include <asm/tlbflush.h>
++#include <asm/mmu_context.h>
++#include <asm/apic.h>
++#include <asm/proto.h>
++
++#ifdef CONFIG_X86_32
++#ifndef CONFIG_XEN
++#include <mach_apic.h>
++/*
++ * 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)
++{
++ 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;
++}
++
++static inline int __prepare_ICR2(unsigned int mask)
++{
++ return SET_APIC_DEST_FIELD(mask);
++}
++#else
++#include <xen/evtchn.h>
++
++DECLARE_PER_CPU(int, ipi_to_irq[NR_IPIS]);
++
++static inline void __send_IPI_one(unsigned int cpu, int vector)
++{
++ int irq = per_cpu(ipi_to_irq, cpu)[vector];
++ BUG_ON(irq < 0);
++ notify_remote_via_irq(irq);
++}
++#endif
++
++void __send_IPI_shortcut(unsigned int shortcut, int vector)
++{
++#ifndef CONFIG_XEN
++ /*
++ * 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
++ * of the value read we use an atomic rmw access to avoid costly
++ * cli/sti. Otherwise we use an even cheaper single atomic write
++ * to the APIC.
++ */
++ unsigned int cfg;
++
++ /*
++ * Wait for idle.
++ */
++ apic_wait_icr_idle();
++
++ /*
++ * No need to touch the target chip field
++ */
++ cfg = __prepare_ICR(shortcut, vector);
++
++ /*
++ * Send the IPI. The write to APIC_ICR fires this off.
++ */
++ apic_write_around(APIC_ICR, cfg);
++#else
++ int cpu;
++
++ switch (shortcut) {
++ case APIC_DEST_SELF:
++ __send_IPI_one(smp_processor_id(), vector);
++ break;
++ case APIC_DEST_ALLBUT:
++ for (cpu = 0; cpu < NR_CPUS; ++cpu) {
++ if (cpu == smp_processor_id())
++ continue;
++ if (cpu_isset(cpu, cpu_online_map)) {
++ __send_IPI_one(cpu, vector);
++ }
++ }
++ break;
++ default:
++ printk("XXXXXX __send_IPI_shortcut %08x vector %d\n", shortcut,
++ vector);
++ break;
++ }
++#endif
++}
++
++void send_IPI_self(int vector)
++{
++ __send_IPI_shortcut(APIC_DEST_SELF, vector);
++}
++
++#ifndef CONFIG_XEN
++/*
++ * This is used to send an IPI with no shorthand notation (the destination is
++ * specified in bits 56 to 63 of the ICR).
++ */
++static inline void __send_IPI_dest_field(unsigned long mask, int vector)
++{
++ unsigned long cfg;
++
++ /*
++ * Wait for idle.
++ */
++ if (unlikely(vector == NMI_VECTOR))
++ safe_apic_wait_icr_idle();
++ else
++ apic_wait_icr_idle();
++
++ /*
++ * prepare target chip field
++ */
++ cfg = __prepare_ICR2(mask);
++ apic_write_around(APIC_ICR2, cfg);
++
++ /*
++ * program the ICR
++ */
++ cfg = __prepare_ICR(0, vector);
++
++ /*
++ * Send the IPI. The write to APIC_ICR fires this off.
++ */
++ apic_write_around(APIC_ICR, cfg);
++}
++#endif
++
++/*
++ * This is only used on smaller machines.
++ */
++void send_IPI_mask_bitmask(cpumask_t cpumask, int vector)
++{
++ unsigned long mask = cpus_addr(cpumask)[0];
++ unsigned long flags;
++#ifdef CONFIG_XEN
++ unsigned int cpu;
++#endif
++
++ local_irq_save(flags);
++ WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]);
++#ifndef CONFIG_XEN
++ __send_IPI_dest_field(mask, vector);
++#else
++ for (cpu = 0; cpu < NR_CPUS; ++cpu)
++ if (cpu_isset(cpu, cpumask))
++ __send_IPI_one(cpu, vector);
++#endif
++ local_irq_restore(flags);
++}
++
++void send_IPI_mask_sequence(cpumask_t mask, int vector)
++{
++#ifndef CONFIG_XEN
++ unsigned long flags;
++ unsigned int query_cpu;
++
++ /*
++ * Hack. The clustered APIC addressing mode doesn't allow us to send
++ * to an arbitrary mask, so I do a unicasts to each CPU instead. This
++ * should be modified to do 1 message per cluster ID - mbligh
++ */
++
++ local_irq_save(flags);
++ for_each_possible_cpu(query_cpu) {
++ if (cpu_isset(query_cpu, mask)) {
++ __send_IPI_dest_field(cpu_to_logical_apicid(query_cpu),
++ vector);
++ }
++ }
++ local_irq_restore(flags);
++#else
++ send_IPI_mask_bitmask(mask, vector);
++#endif
++}
++
++/* must come after the send_IPI functions above for inlining */
++#include <mach_ipi.h>
++
++#ifndef CONFIG_XEN
++static int convert_apicid_to_cpu(int apic_id)
++{
++ int i;
++
++ for_each_possible_cpu(i) {
++ if (per_cpu(x86_cpu_to_apicid, i) == apic_id)
++ return i;
++ }
++ return -1;
++}
++
++int safe_smp_processor_id(void)
++{
++ int apicid, cpuid;
++
++ if (!boot_cpu_has(X86_FEATURE_APIC))
++ return 0;
++
++ apicid = hard_smp_processor_id();
++ if (apicid == BAD_APICID)
++ return 0;
++
++ cpuid = convert_apicid_to_cpu(apicid);
++
++ return cpuid >= 0 ? cpuid : 0;
++}
++#endif
++#endif
+diff --git a/arch/x86/kernel/irq_32-xen.c b/arch/x86/kernel/irq_32-xen.c
+new file mode 100644
+index 0000000..274718d
+--- /dev/null
++++ b/arch/x86/kernel/irq_32-xen.c
+@@ -0,0 +1,383 @@
++/*
++ * 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.)
++ */
++
++#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/apic.h>
++#include <asm/uaccess.h>
++
++DEFINE_PER_CPU_SHARED_ALIGNED(irq_cpustat_t, irq_stat);
++EXPORT_PER_CPU_SYMBOL(irq_stat);
++
++DEFINE_PER_CPU(struct pt_regs *, irq_regs);
++EXPORT_PER_CPU_SYMBOL(irq_regs);
++
++/*
++ * '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(KERN_ERR "unexpected IRQ trap at vector %02x\n", irq);
++
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
++ /*
++ * 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
++ */
++ if (cpu_has_apic)
++ ack_APIC_irq();
++#endif
++}
++
++#ifdef CONFIG_4KSTACKS
++/*
++ * per-CPU IRQ handling contexts (thread information and stack)
++ */
++union irq_ctx {
++ struct thread_info tinfo;
++ u32 stack[THREAD_SIZE/sizeof(u32)];
++};
++
++static union irq_ctx *hardirq_ctx[NR_CPUS] __read_mostly;
++static union irq_ctx *softirq_ctx[NR_CPUS] __read_mostly;
++#endif
++
++/*
++ * do_IRQ handles all normal device IRQ's (the special
++ * SMP cross-CPU interrupts have their own specific
++ * handlers).
++ */
++unsigned int do_IRQ(struct pt_regs *regs)
++{
++ struct pt_regs *old_regs;
++ /* high bit used in ret_from_ code */
++ int irq = ~regs->orig_ax;
++ struct irq_desc *desc = irq_desc + irq;
++#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",
++ __func__, irq);
++ BUG();
++ }
++
++ old_regs = set_irq_regs(regs);
++ irq_enter();
++#ifdef CONFIG_DEBUG_STACKOVERFLOW
++ /* Debugging check for stack overflow: is there less than 1KB free? */
++ {
++ long sp;
++
++ __asm__ __volatile__("andl %%esp,%0" :
++ "=r" (sp) : "0" (THREAD_SIZE - 1));
++ if (unlikely(sp < (sizeof(struct thread_info) + STACK_WARN))) {
++ printk("do_IRQ: stack overflow: %ld\n",
++ sp - sizeof(struct thread_info));
++ dump_stack();
++ }
++ }
++#endif
++
++#ifdef CONFIG_4KSTACKS
++
++ curctx = (union irq_ctx *) current_thread_info();
++ irqctx = hardirq_ctx[smp_processor_id()];
++
++ /*
++ * 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, bx;
++
++ /* 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;
++
++ /*
++ * 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 *%%edi \n"
++ " movl %%ebx,%%esp \n"
++ : "=a" (arg1), "=d" (arg2), "=b" (bx)
++ : "0" (irq), "1" (desc), "2" (isp),
++ "D" (desc->handle_irq)
++ : "memory", "cc", "ecx"
++ );
++ } else
++#endif
++ desc->handle_irq(irq, desc);
++
++ irq_exit();
++ set_irq_regs(old_regs);
++ return 1;
++}
++
++#ifdef CONFIG_4KSTACKS
++
++static char softirq_stack[NR_CPUS * THREAD_SIZE]
++ __attribute__((__section__(".bss.page_aligned")));
++
++static char hardirq_stack[NR_CPUS * THREAD_SIZE]
++ __attribute__((__section__(".bss.page_aligned")));
++
++/*
++ * allocate per-cpu stacks for hardirq and for softirq processing
++ */
++void irq_ctx_init(int cpu)
++{
++ union irq_ctx *irqctx;
++
++ if (hardirq_ctx[cpu])
++ return;
++
++ 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);
++
++ hardirq_ctx[cpu] = irqctx;
++
++ 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);
++
++ softirq_ctx[cpu] = irqctx;
++
++ printk("CPU %u irqstacks, hard=%p soft=%p\n",
++ cpu,hardirq_ctx[cpu],softirq_ctx[cpu]);
++}
++
++void irq_ctx_exit(int cpu)
++{
++ hardirq_ctx[cpu] = NULL;
++}
++
++asmlinkage void do_softirq(void)
++{
++ unsigned long flags;
++ struct thread_info *curctx;
++ union irq_ctx *irqctx;
++ u32 *isp;
++
++ if (in_interrupt())
++ return;
++
++ local_irq_save(flags);
++
++ 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());
++ }
++
++ local_irq_restore(flags);
++}
++#endif
++
++/*
++ * Interrupt statistics:
++ */
++
++#ifndef CONFIG_XEN
++atomic_t irq_err_count;
++#endif
++
++/*
++ * /proc/interrupts printing:
++ */
++
++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) {
++ unsigned any_count = 0;
++
++ spin_lock_irqsave(&irq_desc[i].lock, flags);
++#ifndef CONFIG_SMP
++ any_count = kstat_irqs(i);
++#else
++ for_each_online_cpu(j)
++ any_count |= kstat_cpu(j).irqs[i];
++#endif
++ action = irq_desc[i].action;
++ if (!action && !any_count)
++ goto skip;
++ seq_printf(p, "%3d: ",i);
++#ifndef CONFIG_SMP
++ seq_printf(p, "%10u ", kstat_irqs(i));
++#else
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
++#endif
++ seq_printf(p, " %8s", irq_desc[i].chip->name);
++ seq_printf(p, "-%-8s", irq_desc[i].name);
++
++ if (action) {
++ seq_printf(p, " %s", action->name);
++ while ((action = action->next) != NULL)
++ 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_printf(p, " Non-maskable interrupts\n");
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
++ seq_printf(p, "LOC: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).apic_timer_irqs);
++ seq_printf(p, " Local timer interrupts\n");
++#endif
++#ifdef CONFIG_SMP
++ seq_printf(p, "RES: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).irq_resched_count);
++ seq_printf(p, " Rescheduling interrupts\n");
++ seq_printf(p, "CAL: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).irq_call_count);
++ seq_printf(p, " function call interrupts\n");
++#ifndef CONFIG_XEN
++ seq_printf(p, "TLB: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).irq_tlb_count);
++ seq_printf(p, " TLB shootdowns\n");
++#endif
++#endif
++#ifdef CONFIG_X86_MCE
++ seq_printf(p, "TRM: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).irq_thermal_count);
++ seq_printf(p, " Thermal event interrupts\n");
++#endif
++#ifndef CONFIG_XEN
++#ifdef CONFIG_X86_LOCAL_APIC
++ seq_printf(p, "SPU: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ",
++ per_cpu(irq_stat,j).irq_spurious_count);
++ seq_printf(p, " Spurious interrupts\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
++#endif
++ }
++ return 0;
++}
++
++#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);
++ }
++
++#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 --git a/arch/x86/kernel/irq_64-xen.c b/arch/x86/kernel/irq_64-xen.c
+new file mode 100644
+index 0000000..9e30794
+--- /dev/null
++++ b/arch/x86/kernel/irq_64-xen.c
+@@ -0,0 +1,264 @@
++/*
++ * 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>
++#include <asm/smp.h>
++
++#ifndef CONFIG_XEN
++atomic_t irq_err_count;
++#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(KERN_WARNING "unexpected IRQ trap at irq %02x\n", irq);
++}
++
++#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)task_stack_page(current);
++ static unsigned long warned = -60*HZ;
++
++ if (regs->sp >= curbase && regs->sp <= curbase + THREAD_SIZE &&
++ regs->sp < curbase + sizeof(struct thread_info) + 128 &&
++ time_after(jiffies, warned + 60*HZ)) {
++ printk("do_IRQ: %s near stack overflow (cur:%Lx,sp:%lx)\n",
++ current->comm, curbase, regs->sp);
++ 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) {
++ unsigned any_count = 0;
++
++ spin_lock_irqsave(&irq_desc[i].lock, flags);
++#ifndef CONFIG_SMP
++ any_count = kstat_irqs(i);
++#else
++ for_each_online_cpu(j)
++ any_count |= kstat_cpu(j).irqs[i];
++#endif
++ action = irq_desc[i].action;
++ if (!action && !any_count)
++ goto skip;
++ seq_printf(p, "%3d: ",i);
++#ifndef CONFIG_SMP
++ seq_printf(p, "%10u ", kstat_irqs(i));
++#else
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", kstat_cpu(j).irqs[i]);
++#endif
++ seq_printf(p, " %8s", irq_desc[i].chip->name);
++ seq_printf(p, "-%-8s", irq_desc[i].name);
++
++ if (action) {
++ seq_printf(p, " %s", action->name);
++ while ((action = action->next) != NULL)
++ 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_printf(p, " Non-maskable interrupts\n");
++#ifndef CONFIG_XEN
++ seq_printf(p, "LOC: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->apic_timer_irqs);
++ seq_printf(p, " Local timer interrupts\n");
++#endif
++#ifdef CONFIG_SMP
++ seq_printf(p, "RES: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_resched_count);
++ seq_printf(p, " Rescheduling interrupts\n");
++ seq_printf(p, "CAL: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_call_count);
++ seq_printf(p, " function call interrupts\n");
++#ifndef CONFIG_XEN
++ seq_printf(p, "TLB: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_tlb_count);
++ seq_printf(p, " TLB shootdowns\n");
++#endif
++#endif
++#ifdef CONFIG_X86_MCE
++ seq_printf(p, "TRM: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_thermal_count);
++ seq_printf(p, " Thermal event interrupts\n");
++ seq_printf(p, "THR: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_threshold_count);
++ seq_printf(p, " Threshold APIC interrupts\n");
++#endif
++#ifndef CONFIG_XEN
++ seq_printf(p, "SPU: ");
++ for_each_online_cpu(j)
++ seq_printf(p, "%10u ", cpu_pda(j)->irq_spurious_count);
++ seq_printf(p, " Spurious interrupts\n");
++ seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count));
++#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)
++{
++ struct pt_regs *old_regs = set_irq_regs(regs);
++
++ /* high bit used in ret_from_ code */
++ unsigned irq = ~regs->orig_ax;
++
++ exit_idle();
++ irq_enter();
++
++#ifdef CONFIG_DEBUG_STACKOVERFLOW
++ stack_overflow_check(regs);
++#endif
++
++ if (likely(irq < NR_IRQS))
++ generic_handle_irq(irq);
++ else {
++#ifndef CONFIG_XEN
++ if (!disable_apic)
++ ack_APIC_irq();
++#endif
++ if (printk_ratelimit())
++ printk(KERN_EMERG "%s: %d.%d No irq handler for irq\n",
++ __func__, smp_processor_id(), irq);
++ }
++
++ irq_exit();
++
++ set_irq_regs(old_regs);
++ 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;
++ int break_affinity = 0;
++ int set_affinity = 1;
++
++ if (irq == 2)
++ continue;
++
++ /* interrupt's are disabled at this point */
++ spin_lock(&irq_desc[irq].lock);
++
++ if (!irq_has_action(irq) ||
++ cpus_equal(irq_desc[irq].affinity, map)) {
++ spin_unlock(&irq_desc[irq].lock);
++ continue;
++ }
++
++ cpus_and(mask, irq_desc[irq].affinity, map);
++ if (cpus_empty(mask)) {
++ break_affinity = 1;
++ mask = map;
++ }
++
++ if (irq_desc[irq].chip->mask)
++ irq_desc[irq].chip->mask(irq);
++
++ if (irq_desc[irq].chip->set_affinity)
++ irq_desc[irq].chip->set_affinity(irq, mask);
++ else if (!(warned++))
++ set_affinity = 0;
++
++ if (irq_desc[irq].chip->unmask)
++ irq_desc[irq].chip->unmask(irq);
++
++ spin_unlock(&irq_desc[irq].lock);
++
++ if (break_affinity && set_affinity)
++ /*printk("Broke affinity for irq %i\n", irq)*/;
++ else if (!set_affinity)
++ 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);
++}
+diff --git a/arch/x86/kernel/ldt-xen.c b/arch/x86/kernel/ldt-xen.c
+new file mode 100644
+index 0000000..33b6e3a
+--- /dev/null
++++ b/arch/x86/kernel/ldt-xen.c
+@@ -0,0 +1,272 @@
++/*
++ * Copyright (C) 1992 Krishna Balasubramanian and Linus Torvalds
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ * Copyright (C) 2002 Andi Kleen
++ *
++ * This handles calls from both 32bit and 64bit mode.
++ */
++
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/string.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/vmalloc.h>
++
++#include <asm/uaccess.h>
++#include <asm/system.h>
++#include <asm/ldt.h>
++#include <asm/desc.h>
++#include <asm/mmu_context.h>
++
++#ifdef CONFIG_SMP
++static void flush_ldt(void *null)
++{
++ if (current->active_mm)
++ load_LDT(¤t->active_mm->context);
++}
++#endif
++
++static int alloc_ldt(mm_context_t *pc, int mincount, int reload)
++{
++ void *oldldt, *newldt;
++ int oldsize;
++
++ if (mincount <= pc->size)
++ return 0;
++ oldsize = pc->size;
++ mincount = (mincount + (PAGE_SIZE / LDT_ENTRY_SIZE - 1)) &
++ (~(PAGE_SIZE / LDT_ENTRY_SIZE - 1));
++ if (mincount * LDT_ENTRY_SIZE > PAGE_SIZE)
++ newldt = vmalloc(mincount * LDT_ENTRY_SIZE);
++ else
++ newldt = (void *)__get_free_page(GFP_KERNEL);
++
++ if (!newldt)
++ return -ENOMEM;
++
++ if (oldsize)
++ memcpy(newldt, pc->ldt, oldsize * LDT_ENTRY_SIZE);
++ oldldt = pc->ldt;
++ memset(newldt + oldsize * LDT_ENTRY_SIZE, 0,
++ (mincount - oldsize) * LDT_ENTRY_SIZE);
++
++#ifdef CONFIG_X86_64
++ /* CHECKME: Do we really need this ? */
++ wmb();
++#endif
++ pc->ldt = newldt;
++ wmb();
++ pc->size = mincount;
++ wmb();
++
++ if (reload) {
++#ifdef CONFIG_SMP
++ cpumask_t mask;
++
++ preempt_disable();
++#endif
++ make_pages_readonly(newldt,
++ (mincount * 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
++ put_page(virt_to_page(oldldt));
++ }
++ return 0;
++}
++
++static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
++{
++ int err = alloc_ldt(new, old->size, 0);
++
++ if (err < 0)
++ return err;
++ memcpy(new->ldt, old->ldt, old->size * LDT_ENTRY_SIZE);
++ make_pages_readonly(new->ldt,
++ (new->size * LDT_ENTRY_SIZE) / PAGE_SIZE,
++ XENFEAT_writable_descriptor_tables);
++ return 0;
++}
++
++/*
++ * we do not have to muck with descriptors here, that is
++ * done in switch_mm() as needed.
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm)
++{
++ struct mm_struct *old_mm;
++ int retval = 0;
++
++ memset(&mm->context, 0, sizeof(mm->context));
++ mutex_init(&mm->context.lock);
++ old_mm = current->mm;
++ if (old_mm)
++ mm->context.vdso = old_mm->context.vdso;
++ if (old_mm && old_mm->context.size > 0) {
++ mutex_lock(&old_mm->context.lock);
++ retval = copy_ldt(&mm->context, &old_mm->context);
++ mutex_unlock(&old_mm->context.lock);
++ }
++ return retval;
++}
++
++/*
++ * No need to lock the MM as we are the last user
++ *
++ * 64bit: Don't touch the LDT register - we're already in the next thread.
++ */
++void destroy_context(struct mm_struct *mm)
++{
++ if (mm->context.size) {
++ /* CHECKME: Can this ever happen ? */
++ 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
++ put_page(virt_to_page(mm->context.ldt));
++ mm->context.size = 0;
++ }
++}
++
++static int read_ldt(void __user *ptr, unsigned long bytecount)
++{
++ int err;
++ unsigned long size;
++ struct mm_struct *mm = current->mm;
++
++ if (!mm->context.size)
++ return 0;
++ if (bytecount > LDT_ENTRY_SIZE * LDT_ENTRIES)
++ bytecount = LDT_ENTRY_SIZE * LDT_ENTRIES;
++
++ mutex_lock(&mm->context.lock);
++ size = mm->context.size * LDT_ENTRY_SIZE;
++ if (size > bytecount)
++ size = bytecount;
++
++ err = 0;
++ if (copy_to_user(ptr, mm->context.ldt, size))
++ err = -EFAULT;
++ mutex_unlock(&mm->context.lock);
++ if (err < 0)
++ goto error_return;
++ if (size != bytecount) {
++ /* zero-fill the rest */
++ if (clear_user(ptr + size, bytecount - size) != 0) {
++ err = -EFAULT;
++ goto error_return;
++ }
++ }
++ return bytecount;
++error_return:
++ return err;
++}
++
++static int read_default_ldt(void __user *ptr, unsigned long bytecount)
++{
++ /* CHECKME: Can we use _one_ random number ? */
++#ifdef CONFIG_X86_32
++ unsigned long size = 5 * sizeof(struct desc_struct);
++#else
++ unsigned long size = 128;
++#endif
++ if (bytecount > size)
++ bytecount = size;
++ if (clear_user(ptr, bytecount))
++ return -EFAULT;
++ return bytecount;
++}
++
++static int write_ldt(void __user *ptr, unsigned long bytecount, int oldmode)
++{
++ struct mm_struct *mm = current->mm;
++ struct desc_struct ldt;
++ 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;
++ }
++
++ mutex_lock(&mm->context.lock);
++ 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;
++ }
++
++ /* Allow LDTs to be cleared by the user. */
++ if (ldt_info.base_addr == 0 && ldt_info.limit == 0) {
++ if (oldmode || LDT_empty(&ldt_info)) {
++ memset(&ldt, 0, sizeof(ldt));
++ goto install;
++ }
++ }
++
++ fill_ldt(&ldt, &ldt_info);
++ if (oldmode)
++ ldt.avl = 0;
++
++ /* Install the new entry ... */
++install:
++ error = write_ldt_entry(mm->context.ldt, ldt_info.entry_number, &ldt);
++
++out_unlock:
++ mutex_unlock(&mm->context.lock);
++out:
++ return error;
++}
++
++asmlinkage int sys_modify_ldt(int func, void __user *ptr,
++ unsigned long bytecount)
++{
++ int ret = -ENOSYS;
++
++ switch (func) {
++ case 0:
++ ret = read_ldt(ptr, bytecount);
++ break;
++ case 1:
++ ret = write_ldt(ptr, bytecount, 1);
++ break;
++ case 2:
++ ret = read_default_ldt(ptr, bytecount);
++ break;
++ case 0x11:
++ ret = write_ldt(ptr, bytecount, 0);
++ break;
++ }
++ return ret;
++}
+diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c
+index d0b234c..c748fc8 100644
+--- a/arch/x86/kernel/machine_kexec_32.c
++++ b/arch/x86/kernel/machine_kexec_32.c
+@@ -21,6 +21,10 @@
+ #include <asm/desc.h>
+ #include <asm/system.h>
+
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#endif
++
+ #define PAGE_ALIGNED __attribute__ ((__aligned__(PAGE_SIZE)))
+ static u32 kexec_pgd[1024] PAGE_ALIGNED;
+ #ifdef CONFIG_X86_PAE
+@@ -30,48 +34,55 @@ static u32 kexec_pmd1[1024] PAGE_ALIGNED;
+ static u32 kexec_pte0[1024] PAGE_ALIGNED;
+ static u32 kexec_pte1[1024] PAGE_ALIGNED;
+
+-static void set_idt(void *newidt, __u16 limit)
+-{
+- struct desc_ptr curidt;
++#ifdef CONFIG_XEN
+
+- /* ia32 supports unaliged loads & stores */
+- curidt.size = limit;
+- curidt.address = (unsigned long)newidt;
++#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
+
+- load_idt(&curidt);
+-};
++#if PAGES_NR > KEXEC_XEN_NO_PAGES
++#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
++#endif
+
++#if PA_CONTROL_PAGE != 0
++#error PA_CONTROL_PAGE is non zero - Xen support will break
++#endif
+
+-static void set_gdt(void *newgdt, __u16 limit)
++void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
+ {
+- struct desc_ptr curgdt;
++ void *control_page;
+
+- /* ia32 supports unaligned loads & stores */
+- curgdt.size = limit;
+- curgdt.address = (unsigned long)newgdt;
++ memset(xki->page_list, 0, sizeof(xki->page_list));
+
+- load_gdt(&curgdt);
+-};
++ control_page = page_address(image->control_code_page);
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
+
+-static void load_segments(void)
++ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
++ xki->page_list[PA_PGD] = __ma(kexec_pgd);
++#ifdef CONFIG_X86_PAE
++ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
++ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
++#endif
++ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
++ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
++
++}
++
++int __init machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus)
+ {
+-#define __STR(X) #X
+-#define STR(X) __STR(X)
+-
+- __asm__ __volatile__ (
+- "\tljmp $"STR(__KERNEL_CS)",$1f\n"
+- "\t1:\n"
+- "\tmovl $"STR(__KERNEL_DS)",%%eax\n"
+- "\tmovl %%eax,%%ds\n"
+- "\tmovl %%eax,%%es\n"
+- "\tmovl %%eax,%%fs\n"
+- "\tmovl %%eax,%%gs\n"
+- "\tmovl %%eax,%%ss\n"
+- ::: "eax", "memory");
+-#undef STR
+-#undef __STR
++ int k;
++
++ /* The per-cpu crash note resources belong to the hypervisor resource */
++ for (k = 0; k < nr_phys_cpus; k++)
++ request_resource(hypervisor, phys_cpus + k);
++
++ return 0;
+ }
+
++void machine_kexec_register_resources(struct resource *res) { ; }
++
++#endif /* CONFIG_XEN */
++
+ /*
+ * A architecture hook called to validate the
+ * proposed image and prepare the control pages
+@@ -98,6 +109,7 @@ void machine_kexec_cleanup(struct kimage *image)
+ {
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * Do not allocate memory (or fail in any way) in machine_kexec().
+ * We are past the point of no return, committed to rebooting now.
+@@ -128,26 +140,10 @@ NORET_TYPE void machine_kexec(struct kimage *image)
+ page_list[PA_PTE_1] = __pa(kexec_pte1);
+ page_list[VA_PTE_1] = (unsigned long)kexec_pte1;
+
+- /* The segment registers are funny things, they have both a
+- * visible and an invisible part. Whenever the visible part is
+- * set to a specific selector, the invisible part is loaded
+- * with from a table in memory. At no other time is the
+- * descriptor table in memory accessed.
+- *
+- * I take advantage of this here by force loading the
+- * segments, before I zap the gdt with an invalid value.
+- */
+- load_segments();
+- /* The gdt & idt are now invalid.
+- * If you want to load them you must set up your own idt & gdt.
+- */
+- set_gdt(phys_to_virt(0),0);
+- set_idt(phys_to_virt(0),0);
+-
+- /* now call it */
+ relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
+ image->start, cpu_has_pae);
+ }
++#endif
+
+ void arch_crash_save_vmcoreinfo(void)
+ {
+diff --git a/arch/x86/kernel/machine_kexec_64.c b/arch/x86/kernel/machine_kexec_64.c
+index 576a03d..99ebe85 100644
+--- a/arch/x86/kernel/machine_kexec_64.c
++++ b/arch/x86/kernel/machine_kexec_64.c
+@@ -25,6 +25,117 @@ static u64 kexec_pud1[512] PAGE_ALIGNED;
+ static u64 kexec_pmd1[512] PAGE_ALIGNED;
+ static u64 kexec_pte1[512] PAGE_ALIGNED;
+
++#ifdef CONFIG_XEN
++
++/* In the case of Xen, override hypervisor functions to be able to create
++ * a regular identity mapping page table...
++ */
++
++#include <xen/interface/kexec.h>
++#include <xen/interface/memory.h>
++
++#define x__pmd(x) ((pmd_t) { (x) } )
++#define x__pud(x) ((pud_t) { (x) } )
++#define x__pgd(x) ((pgd_t) { (x) } )
++
++#define x_pmd_val(x) ((x).pmd)
++#define x_pud_val(x) ((x).pud)
++#define x_pgd_val(x) ((x).pgd)
++
++static inline void x_set_pmd(pmd_t *dst, pmd_t val)
++{
++ x_pmd_val(*dst) = x_pmd_val(val);
++}
++
++static inline void x_set_pud(pud_t *dst, pud_t val)
++{
++ x_pud_val(*dst) = phys_to_machine(x_pud_val(val));
++}
++
++static inline void x_pud_clear (pud_t *pud)
++{
++ x_pud_val(*pud) = 0;
++}
++
++static inline void x_set_pgd(pgd_t *dst, pgd_t val)
++{
++ x_pgd_val(*dst) = phys_to_machine(x_pgd_val(val));
++}
++
++static inline void x_pgd_clear (pgd_t * pgd)
++{
++ x_pgd_val(*pgd) = 0;
++}
++
++#define X__PAGE_KERNEL_LARGE_EXEC \
++ _PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_PSE
++#define X_KERNPG_TABLE _PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY
++
++#define __ma(x) (pfn_to_mfn(__pa((x)) >> PAGE_SHIFT) << PAGE_SHIFT)
++
++#if PAGES_NR > KEXEC_XEN_NO_PAGES
++#error PAGES_NR is greater than KEXEC_XEN_NO_PAGES - Xen support will break
++#endif
++
++#if PA_CONTROL_PAGE != 0
++#error PA_CONTROL_PAGE is non zero - Xen support will break
++#endif
++
++void machine_kexec_setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
++{
++ void *control_page;
++ void *table_page;
++
++ memset(xki->page_list, 0, sizeof(xki->page_list));
++
++ control_page = page_address(image->control_code_page) + PAGE_SIZE;
++ memcpy(control_page, relocate_kernel, PAGE_SIZE);
++
++ table_page = page_address(image->control_code_page);
++
++ xki->page_list[PA_CONTROL_PAGE] = __ma(control_page);
++ xki->page_list[PA_TABLE_PAGE] = __ma(table_page);
++
++ xki->page_list[PA_PGD] = __ma(kexec_pgd);
++ xki->page_list[PA_PUD_0] = __ma(kexec_pud0);
++ xki->page_list[PA_PUD_1] = __ma(kexec_pud1);
++ xki->page_list[PA_PMD_0] = __ma(kexec_pmd0);
++ xki->page_list[PA_PMD_1] = __ma(kexec_pmd1);
++ xki->page_list[PA_PTE_0] = __ma(kexec_pte0);
++ xki->page_list[PA_PTE_1] = __ma(kexec_pte1);
++}
++
++int __init machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus)
++{
++ int k;
++
++ /* The per-cpu crash note resources belong to the hypervisor resource */
++ for (k = 0; k < nr_phys_cpus; k++)
++ request_resource(hypervisor, phys_cpus + k);
++
++ return 0;
++}
++
++#else /* CONFIG_XEN */
++
++#define x__pmd(x) __pmd(x)
++#define x__pud(x) __pud(x)
++#define x__pgd(x) __pgd(x)
++
++#define x_set_pmd(x, y) set_pmd(x, y)
++#define x_set_pud(x, y) set_pud(x, y)
++#define x_set_pgd(x, y) set_pgd(x, y)
++
++#define x_pud_clear(x) pud_clear(x)
++#define x_pgd_clear(x) pgd_clear(x)
++
++#define X__PAGE_KERNEL_LARGE_EXEC __PAGE_KERNEL_LARGE_EXEC
++#define X_KERNPG_TABLE _KERNPG_TABLE
++
++#endif /* CONFIG_XEN */
++
+ static void init_level2_page(pmd_t *level2p, unsigned long addr)
+ {
+ unsigned long end_addr;
+@@ -32,7 +143,7 @@ static void init_level2_page(pmd_t *level2p, unsigned long addr)
+ addr &= PAGE_MASK;
+ end_addr = addr + PUD_SIZE;
+ while (addr < end_addr) {
+- set_pmd(level2p++, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC));
++ x_set_pmd(level2p++, x__pmd(addr | X__PAGE_KERNEL_LARGE_EXEC));
+ addr += PMD_SIZE;
+ }
+ }
+@@ -57,12 +168,12 @@ static int init_level3_page(struct kimage *image, pud_t *level3p,
+ }
+ level2p = (pmd_t *)page_address(page);
+ init_level2_page(level2p, addr);
+- set_pud(level3p++, __pud(__pa(level2p) | _KERNPG_TABLE));
++ x_set_pud(level3p++, x__pud(__pa(level2p) | X_KERNPG_TABLE));
+ addr += PUD_SIZE;
+ }
+ /* clear the unused entries */
+ while (addr < end_addr) {
+- pud_clear(level3p++);
++ x_pud_clear(level3p++);
+ addr += PUD_SIZE;
+ }
+ out:
+@@ -93,12 +204,12 @@ static int init_level4_page(struct kimage *image, pgd_t *level4p,
+ if (result) {
+ goto out;
+ }
+- set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE));
++ x_set_pgd(level4p++, x__pgd(__pa(level3p) | X_KERNPG_TABLE));
+ addr += PGDIR_SIZE;
+ }
+ /* clear the unused entries */
+ while (addr < end_addr) {
+- pgd_clear(level4p++);
++ x_pgd_clear(level4p++);
+ addr += PGDIR_SIZE;
+ }
+ out:
+@@ -109,49 +220,14 @@ out:
+ static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
+ {
+ pgd_t *level4p;
+- level4p = (pgd_t *)__va(start_pgtable);
+- return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT);
+-}
+-
+-static void set_idt(void *newidt, u16 limit)
+-{
+- struct desc_ptr curidt;
+-
+- /* x86-64 supports unaliged loads & stores */
+- curidt.size = limit;
+- curidt.address = (unsigned long)newidt;
+-
+- __asm__ __volatile__ (
+- "lidtq %0\n"
+- : : "m" (curidt)
+- );
+-};
++ unsigned long x_end_pfn = end_pfn;
+
++#ifdef CONFIG_XEN
++ x_end_pfn = HYPERVISOR_memory_op(XENMEM_maximum_ram_page, NULL);
++#endif
+
+-static void set_gdt(void *newgdt, u16 limit)
+-{
+- struct desc_ptr curgdt;
+-
+- /* x86-64 supports unaligned loads & stores */
+- curgdt.size = limit;
+- curgdt.address = (unsigned long)newgdt;
+-
+- __asm__ __volatile__ (
+- "lgdtq %0\n"
+- : : "m" (curgdt)
+- );
+-};
+-
+-static void load_segments(void)
+-{
+- __asm__ __volatile__ (
+- "\tmovl %0,%%ds\n"
+- "\tmovl %0,%%es\n"
+- "\tmovl %0,%%ss\n"
+- "\tmovl %0,%%fs\n"
+- "\tmovl %0,%%gs\n"
+- : : "a" (__KERNEL_DS) : "memory"
+- );
++ level4p = (pgd_t *)__va(start_pgtable);
++ return init_level4_page(image, level4p, 0, x_end_pfn << PAGE_SHIFT);
+ }
+
+ int machine_kexec_prepare(struct kimage *image)
+@@ -175,6 +251,7 @@ void machine_kexec_cleanup(struct kimage *image)
+ return;
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * Do not allocate memory (or fail in any way) in machine_kexec().
+ * We are past the point of no return, committed to rebooting now.
+@@ -210,26 +287,10 @@ NORET_TYPE void machine_kexec(struct kimage *image)
+ page_list[PA_TABLE_PAGE] =
+ (unsigned long)__pa(page_address(image->control_code_page));
+
+- /* The segment registers are funny things, they have both a
+- * visible and an invisible part. Whenever the visible part is
+- * set to a specific selector, the invisible part is loaded
+- * with from a table in memory. At no other time is the
+- * descriptor table in memory accessed.
+- *
+- * I take advantage of this here by force loading the
+- * segments, before I zap the gdt with an invalid value.
+- */
+- load_segments();
+- /* The gdt & idt are now invalid.
+- * If you want to load them you must set up your own idt & gdt.
+- */
+- set_gdt(phys_to_virt(0),0);
+- set_idt(phys_to_virt(0),0);
+-
+- /* now call it */
+ relocate_kernel((unsigned long)image->head, (unsigned long)page_list,
+ image->start);
+ }
++#endif
+
+ void arch_crash_save_vmcoreinfo(void)
+ {
+diff --git a/arch/x86/kernel/microcode-xen.c b/arch/x86/kernel/microcode-xen.c
+new file mode 100644
+index 0000000..42c6517
+--- /dev/null
++++ b/arch/x86/kernel/microcode-xen.c
+@@ -0,0 +1,210 @@
++/*
++ * Intel CPU Microcode Update Driver for Linux
++ *
++ * Copyright (C) 2000-2006 Tigran Aivazian <tigran at aivazian.fsnet.co.uk>
++ * 2006 Shaohua Li <shaohua.li at intel.com>
++ *
++ * 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.
++ */
++
++//#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/fs.h>
++#include <linux/mutex.h>
++#include <linux/cpu.h>
++#include <linux/firmware.h>
++#include <linux/platform_device.h>
++
++#include <asm/msr.h>
++#include <asm/uaccess.h>
++#include <asm/processor.h>
++
++MODULE_DESCRIPTION("Intel CPU (IA-32) Microcode Update Driver");
++MODULE_AUTHOR("Tigran Aivazian <tigran at aivazian.fsnet.co.uk>");
++MODULE_LICENSE("GPL");
++
++static int verbose;
++module_param(verbose, int, 0644);
++
++#define MICROCODE_VERSION "1.14a-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 */
++
++/* no concurrent ->write()s are allowed on /dev/cpu/microcode */
++static DEFINE_MUTEX(microcode_mutex);
++
++#ifdef CONFIG_MICROCODE_OLD_INTERFACE
++static int do_microcode_update (const void __user *ubuf, size_t len)
++{
++ int err;
++ void *kbuf;
++
++ kbuf = vmalloc(len);
++ if (!kbuf)
++ return -ENOMEM;
++
++ if (copy_from_user(kbuf, ubuf, len) == 0) {
++ struct xen_platform_op op;
++
++ op.cmd = XENPF_microcode_update;
++ set_xen_guest_handle(op.u.microcode.data, kbuf);
++ op.u.microcode.length = len;
++ err = HYPERVISOR_platform_op(&op);
++ } else
++ err = -EFAULT;
++
++ vfree(kbuf);
++
++ return err;
++}
++
++static int microcode_open (struct inode *unused1, struct file *unused2)
++{
++ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
++}
++
++static ssize_t microcode_write (struct file *file, const char __user *buf, size_t len, loff_t *ppos)
++{
++ ssize_t ret;
++
++ if (len < MC_HEADER_SIZE) {
++ printk(KERN_ERR "microcode: not enough data\n");
++ return -EINVAL;
++ }
++
++ mutex_lock(µcode_mutex);
++
++ ret = do_microcode_update(buf, len);
++ if (!ret)
++ ret = (ssize_t)len;
++
++ mutex_unlock(µcode_mutex);
++
++ return ret;
++}
++
++static const 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_dev_init (void)
++{
++ int error;
++
++ 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 void microcode_dev_exit (void)
++{
++ misc_deregister(µcode_dev);
++}
++
++MODULE_ALIAS_MISCDEV(MICROCODE_MINOR);
++#else
++#define microcode_dev_init() 0
++#define microcode_dev_exit() do { } while(0)
++#endif
++
++/* fake device for request_firmware */
++static struct platform_device *microcode_pdev;
++
++static int request_microcode(void)
++{
++ char name[30];
++ const struct cpuinfo_x86 *c = &boot_cpu_data;
++ const struct firmware *firmware;
++ int error;
++ struct xen_platform_op op;
++
++ sprintf(name,"intel-ucode/%02x-%02x-%02x",
++ c->x86, c->x86_model, c->x86_mask);
++ error = request_firmware(&firmware, name, µcode_pdev->dev);
++ if (error) {
++ pr_debug("microcode: ucode data file %s load failed\n", name);
++ return error;
++ }
++
++ op.cmd = XENPF_microcode_update;
++ set_xen_guest_handle(op.u.microcode.data, firmware->data);
++ op.u.microcode.length = firmware->size;
++ error = HYPERVISOR_platform_op(&op);
++
++ release_firmware(firmware);
++
++ if (error)
++ pr_debug("ucode load failed\n");
++
++ return error;
++}
++
++static int __init microcode_init (void)
++{
++ int error;
++
++ error = microcode_dev_init();
++ if (error)
++ return error;
++ microcode_pdev = platform_device_register_simple("microcode", -1,
++ NULL, 0);
++ if (IS_ERR(microcode_pdev)) {
++ microcode_dev_exit();
++ return PTR_ERR(microcode_pdev);
++ }
++
++ request_microcode();
++
++ printk(KERN_INFO
++ "IA-32 Microcode Update Driver: v" MICROCODE_VERSION " <tigran at aivazian.fsnet.co.uk>\n");
++ return 0;
++}
++
++static void __exit microcode_exit (void)
++{
++ microcode_dev_exit();
++ platform_device_unregister(microcode_pdev);
++}
++
++module_init(microcode_init)
++module_exit(microcode_exit)
+diff --git a/arch/x86/kernel/mpparse-xen.c b/arch/x86/kernel/mpparse-xen.c
+new file mode 100644
+index 0000000..abc8f05
+--- /dev/null
++++ b/arch/x86/kernel/mpparse-xen.c
+@@ -0,0 +1,1111 @@
++/*
++ * 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>
++ * (c) 2008 Alexey Starikovskiy <astarikovskiy at suse.de>
++ */
++
++#include <linux/mm.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/bootmem.h>
++#include <linux/kernel_stat.h>
++#include <linux/mc146818rtc.h>
++#include <linux/bitops.h>
++#include <linux/acpi.h>
++#include <linux/module.h>
++
++#include <asm/smp.h>
++#include <asm/mtrr.h>
++#include <asm/mpspec.h>
++#include <asm/pgalloc.h>
++#include <asm/io_apic.h>
++#include <asm/proto.h>
++#include <asm/acpi.h>
++#include <asm/bios_ebda.h>
++
++#include <mach_apic.h>
++#ifdef CONFIG_X86_32
++#include <mach_mpparse.h>
++#endif
++
++/* Have we found an MP table */
++int smp_found_config;
++
++/*
++ * Various Linux-internal data structures created from the
++ * MP-table.
++ */
++#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
++int mp_bus_id_to_type[MAX_MP_BUSSES];
++#endif
++
++DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES);
++int mp_bus_id_to_pci_bus[MAX_MP_BUSSES] = {[0 ... MAX_MP_BUSSES - 1] = -1 };
++
++static int mp_current_pci_id;
++
++int pic_mode;
++
++/*
++ * Intel MP BIOS table parsing routines:
++ */
++
++/*
++ * Checksum an MP configuration block.
++ */
++
++static int __init mpf_checksum(unsigned char *mp, int len)
++{
++ int sum = 0;
++
++ while (len--)
++ sum += *mp++;
++
++ return sum & 0xFF;
++}
++
++#ifdef CONFIG_X86_NUMAQ
++/*
++ * 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 ....
++ */
++
++static int mpc_record;
++static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY]
++ __cpuinitdata;
++#endif
++
++#ifndef CONFIG_XEN
++static void __cpuinit MP_processor_info(struct mpc_config_processor *m)
++{
++ int apicid;
++ char *bootup_cpu = "";
++
++ if (!(m->mpc_cpuflag & CPU_ENABLED)) {
++ disabled_cpus++;
++ return;
++ }
++#ifdef CONFIG_X86_NUMAQ
++ apicid = mpc_apic_id(m, translation_table[mpc_record]);
++#else
++ apicid = m->mpc_apicid;
++#endif
++ if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) {
++ bootup_cpu = " (Bootup-CPU)";
++ boot_cpu_physical_apicid = m->mpc_apicid;
++ }
++
++ printk(KERN_INFO "Processor #%d%s\n", m->mpc_apicid, bootup_cpu);
++ generic_processor_info(apicid, m->mpc_apicver);
++}
++#else
++static void __cpuinit MP_processor_info (struct mpc_config_processor *m)
++{
++ num_processors++;
++}
++#endif /* CONFIG_XEN */
++
++static void __init MP_bus_info(struct mpc_config_bus *m)
++{
++ char str[7];
++
++ memcpy(str, m->mpc_bustype, 6);
++ str[6] = 0;
++
++#ifdef CONFIG_X86_NUMAQ
++ mpc_oem_bus_info(m, str, translation_table[mpc_record]);
++#else
++ Dprintk("Bus #%d is %s\n", m->mpc_busid, str);
++#endif
++
++#if MAX_MP_BUSSES < 256
++ 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;
++ }
++#endif
++
++ if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA) - 1) == 0) {
++ set_bit(m->mpc_busid, mp_bus_not_pci);
++#if defined(CONFIG_EISA) || defined (CONFIG_MCA)
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
++#endif
++ } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) {
++#ifdef CONFIG_X86_NUMAQ
++ mpc_oem_pci_bus(m, translation_table[mpc_record]);
++#endif
++ clear_bit(m->mpc_busid, mp_bus_not_pci);
++ mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id;
++ mp_current_pci_id++;
++#if defined(CONFIG_EISA) || defined (CONFIG_MCA)
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI;
++ } 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_MCA, sizeof(BUSTYPE_MCA) - 1) == 0) {
++ mp_bus_id_to_type[m->mpc_busid] = MP_BUS_MCA;
++#endif
++ } else
++ printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str);
++}
++
++#ifdef CONFIG_X86_IO_APIC
++
++static int bad_ioapic(unsigned long address)
++{
++ 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 table, skipping!\n");
++ return 1;
++ }
++ return 0;
++}
++
++static void __init MP_ioapic_info(struct mpc_config_ioapic *m)
++{
++ if (!(m->mpc_flags & MPC_APIC_USABLE))
++ return;
++
++ printk(KERN_INFO "I/O APIC #%d Version %d at 0x%X.\n",
++ m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr);
++
++ if (bad_ioapic(m->mpc_apicaddr))
++ return;
++
++ mp_ioapics[nr_ioapics] = *m;
++ nr_ioapics++;
++}
++
++static void __init MP_intsrc_info(struct mpc_config_intsrc *m)
++{
++ mp_irqs[mp_irq_entries] = *m;
++ Dprintk("Int: type %d, pol %d, trig %d, bus %d,"
++ " IRQ %02x, APIC ID %x, APIC INT %02x\n",
++ m->mpc_irqtype, m->mpc_irqflag & 3,
++ (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus,
++ m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq);
++ if (++mp_irq_entries == MAX_IRQ_SOURCES)
++ panic("Max # of irq sources exceeded!!\n");
++}
++
++#endif
++
++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);
++}
++
++#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 (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);
++}
++
++/*
++ * Read/parse the MPC oem tables
++ */
++
++static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable,
++ unsigned short oemsize)
++{
++ 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");
++ 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;
++ }
++ }
++ }
++}
++
++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 */
++
++/*
++ * Read/parse the MPC
++ */
++
++static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early)
++{
++ 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 "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(KERN_ERR "MPTABLE: checksum error!\n");
++ return 0;
++ }
++ if (mpc->mpc_spec != 0x01 && mpc->mpc_spec != 0x04) {
++ printk(KERN_ERR "MPTABLE: bad table version (%d)!!\n",
++ mpc->mpc_spec);
++ return 0;
++ }
++ if (!mpc->mpc_lapic) {
++ printk(KERN_ERR "MPTABLE: null local APIC address!\n");
++ return 0;
++ }
++ memcpy(oem, mpc->mpc_oem, 8);
++ oem[8] = 0;
++ printk(KERN_INFO "MPTABLE: OEM ID: %s ", oem);
++
++ memcpy(str, mpc->mpc_productid, 12);
++ str[12] = 0;
++ printk("Product ID: %s ", str);
++
++#ifdef CONFIG_X86_32
++ mps_oem_check(mpc, oem, str);
++#endif
++ printk(KERN_INFO "MPTABLE: Product ID: %s ", str);
++
++#ifndef CONFIG_XEN
++ printk(KERN_INFO "MPTABLE: APIC at: 0x%X\n", mpc->mpc_lapic);
++
++ /* save the local APIC address, it might be non-default */
++ if (!acpi_lapic)
++ mp_lapic_addr = mpc->mpc_lapic;
++#endif
++
++ if (early)
++ return 1;
++
++ /*
++ * Now process the configuration blocks.
++ */
++#ifdef CONFIG_X86_NUMAQ
++ mpc_record = 0;
++#endif
++ 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:
++ {
++#ifdef CONFIG_X86_IO_APIC
++ struct mpc_config_ioapic *m =
++ (struct mpc_config_ioapic *)mpt;
++ MP_ioapic_info(m);
++#endif
++ mpt += sizeof(struct mpc_config_ioapic);
++ count += sizeof(struct mpc_config_ioapic);
++ break;
++ }
++ case MP_INTSRC:
++ {
++#ifdef CONFIG_X86_IO_APIC
++ struct mpc_config_intsrc *m =
++ (struct mpc_config_intsrc *)mpt;
++
++ MP_intsrc_info(m);
++#endif
++ mpt += sizeof(struct mpc_config_intsrc);
++ count += sizeof(struct mpc_config_intsrc);
++ 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:
++ /* wrong mptable */
++ printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n");
++ printk(KERN_ERR "type %x\n", *mpt);
++ print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16,
++ 1, mpc, mpc->mpc_length, 1);
++ count = mpc->mpc_length;
++ break;
++ }
++#ifdef CONFIG_X86_NUMAQ
++ ++mpc_record;
++#endif
++ }
++ setup_apic_routing();
++ if (!num_processors)
++ printk(KERN_ERR "MPTABLE: no processors registered!\n");
++ return num_processors;
++}
++
++#ifdef CONFIG_X86_IO_APIC
++
++static int __init ELCR_trigger(unsigned int irq)
++{
++ unsigned int port;
++
++ port = 0x4d0 + (irq >> 3);
++ return (inb(port) >> (irq & 7)) & 1;
++}
++
++static void __init construct_default_ioirq_mptable(int mpc_default_type)
++{
++ struct mpc_config_intsrc intsrc;
++ int i;
++ int ELCR_fallback = 0;
++
++ intsrc.mpc_type = MP_INTSRC;
++ intsrc.mpc_irqflag = 0; /* conforming */
++ intsrc.mpc_srcbus = 0;
++ intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid;
++
++ intsrc.mpc_irqtype = mp_INT;
++
++ /*
++ * If true, we have an ISA/PCI system with no IRQ entries
++ * in the MP table. To prevent the PCI interrupts from being set up
++ * incorrectly, we try to use the ELCR. The sanity check to see if
++ * there is good ELCR data is very simple - IRQ0, 1, 2 and 13 can
++ * never be level sensitive, so we simply see if the ELCR agrees.
++ * If it does, we assume it's valid.
++ */
++ if (mpc_default_type == 5) {
++ printk(KERN_INFO "ISA/PCI bus type with no IRQ information... "
++ "falling back to ELCR\n");
++
++ if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) ||
++ ELCR_trigger(13))
++ printk(KERN_ERR "ELCR contains invalid data... "
++ "not using ELCR\n");
++ else {
++ printk(KERN_INFO
++ "Using ELCR to identify PCI interrupts\n");
++ ELCR_fallback = 1;
++ }
++ }
++
++ for (i = 0; i < 16; i++) {
++ switch (mpc_default_type) {
++ case 2:
++ if (i == 0 || i == 13)
++ continue; /* IRQ0 & IRQ13 not connected */
++ /* fall through */
++ default:
++ if (i == 2)
++ continue; /* IRQ2 is never connected */
++ }
++
++ if (ELCR_fallback) {
++ /*
++ * If the ELCR indicates a level-sensitive interrupt, we
++ * copy that information over to the MP table in the
++ * irqflag field (level sensitive, active high polarity).
++ */
++ if (ELCR_trigger(i))
++ intsrc.mpc_irqflag = 13;
++ else
++ intsrc.mpc_irqflag = 0;
++ }
++
++ intsrc.mpc_srcbusirq = i;
++ intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */
++ MP_intsrc_info(&intsrc);
++ }
++
++ intsrc.mpc_irqtype = mp_ExtINT;
++ intsrc.mpc_srcbusirq = 0;
++ intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */
++ MP_intsrc_info(&intsrc);
++}
++
++#endif
++
++static inline void __init construct_default_ISA_mptable(int mpc_default_type)
++{
++ struct mpc_config_processor processor;
++ struct mpc_config_bus bus;
++#ifdef CONFIG_X86_IO_APIC
++ struct mpc_config_ioapic ioapic;
++#endif
++ struct mpc_config_lintsrc lintsrc;
++ int linttypes[2] = { mp_ExtINT, mp_NMI };
++ int i;
++
++#ifndef CONFIG_XEN
++ /*
++ * local APIC has default address
++ */
++ mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
++#endif
++
++ /*
++ * 2 CPUs, numbered 0 & 1.
++ */
++ processor.mpc_type = MP_PROCESSOR;
++ /* Either an integrated APIC or a discrete 82489DX. */
++ processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01;
++ processor.mpc_cpuflag = CPU_ENABLED;
++ processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) |
++ (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask;
++ processor.mpc_featureflag = boot_cpu_data.x86_capability[0];
++ processor.mpc_reserved[0] = 0;
++ processor.mpc_reserved[1] = 0;
++ for (i = 0; i < 2; i++) {
++ processor.mpc_apicid = i;
++ MP_processor_info(&processor);
++ }
++
++ bus.mpc_type = MP_BUS;
++ bus.mpc_busid = 0;
++ switch (mpc_default_type) {
++ default:
++ printk(KERN_ERR "???\nUnknown standard configuration %d\n",
++ mpc_default_type);
++ /* fall through */
++ case 1:
++ case 5:
++ memcpy(bus.mpc_bustype, "ISA ", 6);
++ break;
++ case 2:
++ case 6:
++ case 3:
++ memcpy(bus.mpc_bustype, "EISA ", 6);
++ break;
++ case 4:
++ case 7:
++ memcpy(bus.mpc_bustype, "MCA ", 6);
++ }
++ MP_bus_info(&bus);
++ if (mpc_default_type > 4) {
++ bus.mpc_busid = 1;
++ memcpy(bus.mpc_bustype, "PCI ", 6);
++ MP_bus_info(&bus);
++ }
++
++#ifdef CONFIG_X86_IO_APIC
++ 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);
++#endif
++ lintsrc.mpc_type = MP_LINTSRC;
++ lintsrc.mpc_irqflag = 0; /* conforming */
++ lintsrc.mpc_srcbusid = 0;
++ lintsrc.mpc_srcbusirq = 0;
++ lintsrc.mpc_destapic = MP_APIC_ALL;
++ for (i = 0; i < 2; i++) {
++ lintsrc.mpc_irqtype = linttypes[i];
++ lintsrc.mpc_destapiclint = i;
++ MP_lintsrc_info(&lintsrc);
++ }
++}
++
++static struct intel_mp_floating *mpf_found;
++
++/*
++ * Scan the memory blocks for an SMP configuration block.
++ */
++static void __init __get_smp_config(unsigned early)
++{
++ struct intel_mp_floating *mpf = mpf_found;
++
++ if (acpi_lapic && early)
++ return;
++ /*
++ * ACPI supports both logical (e.g. Hyper-Threading) and physical
++ * processors, where MPS only supports physical.
++ */
++ if (acpi_lapic && acpi_ioapic) {
++ printk(KERN_INFO "Using ACPI (MADT) for SMP configuration "
++ "information\n");
++ return;
++ } else if (acpi_lapic)
++ printk(KERN_INFO "Using ACPI for processor (LAPIC) "
++ "configuration information\n");
++
++ printk(KERN_INFO "Intel MultiProcessor Specification v1.%d\n",
++ mpf->mpf_specification);
++#ifdef CONFIG_X86_32
++ 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;
++ }
++#endif
++ /*
++ * Now see if we need to read further.
++ */
++ if (mpf->mpf_feature1 != 0) {
++ if (early) {
++#ifndef CONFIG_XEN
++ /*
++ * local APIC has default address
++ */
++ mp_lapic_addr = APIC_DEFAULT_PHYS_BASE;
++#endif
++ return;
++ }
++
++ 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), early)) {
++ 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 (early)
++ return;
++#ifdef CONFIG_X86_IO_APIC
++ /*
++ * If there are no explicit MP IRQ entries, then we are
++ * broken. We set up most of the low 16 IO-APIC pins to
++ * ISA defaults and hope it will work.
++ */
++ if (!mp_irq_entries) {
++ struct mpc_config_bus bus;
++
++ printk(KERN_ERR "BIOS bug, no explicit IRQ entries, "
++ "using default mptable. "
++ "(tell your hw vendor)\n");
++
++ bus.mpc_type = MP_BUS;
++ bus.mpc_busid = 0;
++ memcpy(bus.mpc_bustype, "ISA ", 6);
++ MP_bus_info(&bus);
++
++ construct_default_ioirq_mptable(0);
++ }
++#endif
++ } else
++ BUG();
++
++ if (!early)
++ printk(KERN_INFO "Processors: %d\n", num_processors);
++ /*
++ * Only use the first configuration found.
++ */
++}
++
++void __init early_get_smp_config(void)
++{
++ __get_smp_config(1);
++}
++
++void __init get_smp_config(void)
++{
++ __get_smp_config(0);
++}
++
++static int __init smp_scan_config(unsigned long base, unsigned long length,
++ unsigned reserve)
++{
++ unsigned int *bp = isa_bus_to_virt(base);
++ struct intel_mp_floating *mpf;
++
++ Dprintk("Scan SMP from %p for %ld bytes.\n", bp, length);
++ BUILD_BUG_ON(sizeof(*mpf) != 16);
++
++ while (length > 0) {
++ mpf = (struct intel_mp_floating *)bp;
++ if ((*bp == SMP_MAGIC_IDENT) &&
++ (mpf->mpf_length == 1) &&
++ !mpf_checksum((unsigned char *)bp, 16) &&
++ ((mpf->mpf_specification == 1)
++ || (mpf->mpf_specification == 4))) {
++
++ smp_found_config = 1;
++ mpf_found = mpf;
++#ifdef CONFIG_X86_32
++#ifndef CONFIG_XEN
++ printk(KERN_INFO "found SMP MP-table at [%p] %08lx\n",
++ mpf, virt_to_phys(mpf));
++ reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE,
++ BOOTMEM_DEFAULT);
++ 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,
++ BOOTMEM_DEFAULT);
++ }
++#else
++ printk(KERN_INFO "found SMP MP-table at [%p] %08lx\n",
++ mpf, ((void *)bp - isa_bus_to_virt(base)) + base);
++#endif
++#elif !defined(CONFIG_XEN)
++ if (!reserve)
++ return 1;
++
++ reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE);
++ if (mpf->mpf_physptr)
++ reserve_bootmem_generic(mpf->mpf_physptr,
++ PAGE_SIZE);
++#endif
++ return 1;
++ }
++ bp += 4;
++ length -= 16;
++ }
++ return 0;
++}
++
++static void __init __find_smp_config(unsigned reserve)
++{
++#ifndef CONFIG_XEN
++ unsigned int address;
++#endif
++
++ /*
++ * 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, reserve) ||
++ smp_scan_config(639 * 0x400, 0x400, reserve) ||
++ smp_scan_config(0xF0000, 0x10000, reserve))
++ 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.
++ */
++
++#ifndef CONFIG_XEN
++ address = get_bios_ebda();
++ if (address)
++ smp_scan_config(address, 0x400, reserve);
++#endif
++}
++
++void __init early_find_smp_config(void)
++{
++ __find_smp_config(0);
++}
++
++void __init find_smp_config(void)
++{
++ __find_smp_config(1);
++}
++
++/* --------------------------------------------------------------------------
++ ACPI-based MP Configuration
++ -------------------------------------------------------------------------- */
++
++/*
++ * Keep this outside and initialized to 0, for !CONFIG_ACPI builds:
++ */
++int es7000_plat;
++
++#ifdef CONFIG_ACPI
++
++#ifdef CONFIG_X86_IO_APIC
++
++#define MP_ISA_BUS 0
++
++extern struct mp_ioapic_routing 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_base)
++ && (gsi <= mp_ioapic_routing[i].gsi_end))
++ return i;
++ }
++
++ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
++ return -1;
++}
++
++static u8 __init uniq_ioapic_id(u8 id)
++{
++#ifdef CONFIG_X86_32
++#ifndef CONFIG_XEN
++ if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) &&
++ !APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
++ return io_apic_get_unique_id(nr_ioapics, id);
++ else
++#endif
++ return id;
++#else
++ int i;
++ DECLARE_BITMAP(used, 256);
++ bitmap_zero(used, 256);
++ for (i = 0; i < nr_ioapics; i++) {
++ struct mpc_config_ioapic *ia = &mp_ioapics[i];
++ __set_bit(ia->mpc_apicid, used);
++ }
++ if (!test_bit(id, used))
++ return id;
++ return find_first_zero_bit(used, 256);
++#endif
++}
++
++void __init mp_register_ioapic(int id, u32 address, u32 gsi_base)
++{
++ int idx = 0;
++
++ if (bad_ioapic(address))
++ return;
++
++ idx = nr_ioapics;
++
++ mp_ioapics[idx].mpc_type = MP_IOAPIC;
++ mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE;
++ mp_ioapics[idx].mpc_apicaddr = address;
++
++#ifndef CONFIG_XEN
++ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
++#endif
++ mp_ioapics[idx].mpc_apicid = uniq_ioapic_id(id);
++#ifdef CONFIG_X86_32
++ mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx);
++#else
++ mp_ioapics[idx].mpc_apicver = 0;
++#endif
++ /*
++ * 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(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_base, mp_ioapic_routing[idx].gsi_end);
++
++ nr_ioapics++;
++}
++
++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_base;
++
++ /*
++ * 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# */
++
++ MP_intsrc_info(&intsrc);
++}
++
++void __init mp_config_acpi_legacy_irqs(void)
++{
++ struct mpc_config_intsrc intsrc;
++ int i = 0;
++ int ioapic = -1;
++
++#if defined (CONFIG_MCA) || defined (CONFIG_EISA)
++ /*
++ * Fabricate the legacy ISA bus (bus #31).
++ */
++ mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA;
++#endif
++ set_bit(MP_ISA_BUS, mp_bus_not_pci);
++ Dprintk("Bus #%d is ISA\n", MP_ISA_BUS);
++
++ /*
++ * Older generations of ES7000 have no legacy identity mappings
++ */
++ if (es7000_plat == 1)
++ return;
++
++ /*
++ * Locate the IOAPIC that manages the ISA IRQs (0-15).
++ */
++ ioapic = mp_find_ioapic(0);
++ if (ioapic < 0)
++ return;
++
++ intsrc.mpc_type = MP_INTSRC;
++ intsrc.mpc_irqflag = 0; /* Conforming */
++ intsrc.mpc_srcbus = MP_ISA_BUS;
++#ifdef CONFIG_X86_IO_APIC
++ intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid;
++#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;
++
++ for (idx = 0; idx < mp_irq_entries; idx++) {
++ struct mpc_config_intsrc *irq = mp_irqs + idx;
++
++ /* Do we already have a mapping for this ISA IRQ? */
++ if (irq->mpc_srcbus == MP_ISA_BUS
++ && irq->mpc_srcbusirq == i)
++ break;
++
++ /* Do we already have a mapping for this IOAPIC pin */
++ if ((irq->mpc_dstapic == intsrc.mpc_dstapic) &&
++ (irq->mpc_dstirq == i))
++ break;
++ }
++
++ if (idx != mp_irq_entries) {
++ printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i);
++ continue; /* IRQ already used */
++ }
++
++ intsrc.mpc_irqtype = mp_INT;
++ intsrc.mpc_srcbusirq = i; /* Identity mapped */
++ intsrc.mpc_dstirq = i;
++
++ MP_intsrc_info(&intsrc);
++ }
++}
++
++int mp_register_gsi(u32 gsi, int triggering, int polarity)
++{
++ int ioapic;
++ int ioapic_pin;
++#ifdef CONFIG_X86_32
++#define MAX_GSI_NUM 4096
++#define IRQ_COMPRESSION_START 64
++
++ static int pci_irq = IRQ_COMPRESSION_START;
++ /*
++ * Mapping between Global System Interrupts, which
++ * represent all possible interrupts, and IRQs
++ * assigned to actual devices.
++ */
++ static int gsi_to_irq[MAX_GSI_NUM];
++#else
++
++ if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC)
++ return gsi;
++#endif
++
++ /* Don't set up the ACPI SCI because it's already set up */
++ if (acpi_gbl_FADT.sci_interrupt == gsi)
++ return gsi;
++
++ ioapic = mp_find_ioapic(gsi);
++ if (ioapic < 0) {
++ printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi);
++ return gsi;
++ }
++
++ ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_base;
++
++#if defined(CONFIG_X86_32) && !defined(CONFIG_XEN)
++ if (ioapic_renumber_irq)
++ gsi = ioapic_renumber_irq(ioapic, gsi);
++#endif
++
++ /*
++ * Avoid pin reprogramming. PRTs typically include entries
++ * with redundant pin->gsi mappings (but unique PCI devices);
++ * we only program the IOAPIC on the first.
++ */
++ if (ioapic_pin > MP_MAX_IOAPIC_PIN) {
++ printk(KERN_ERR "Invalid reference to IOAPIC pin "
++ "%d-%d\n", mp_ioapic_routing[ioapic].apic_id,
++ ioapic_pin);
++ return gsi;
++ }
++ if (test_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed)) {
++ Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n",
++ mp_ioapic_routing[ioapic].apic_id, ioapic_pin);
++#ifdef CONFIG_X86_32
++ return (gsi < IRQ_COMPRESSION_START ? gsi : gsi_to_irq[gsi]);
++#else
++ return gsi;
++#endif
++ }
++
++ set_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed);
++#ifdef CONFIG_X86_32
++ /*
++ * For GSI >= 64, use IRQ compression
++ */
++ if ((gsi >= IRQ_COMPRESSION_START)
++ && (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.
++ */
++ gsi = pci_irq++;
++ /*
++ * Don't assign IRQ used by ACPI SCI
++ */
++ if (gsi == acpi_gbl_FADT.sci_interrupt)
++ gsi = pci_irq++;
++ gsi_to_irq[irq] = gsi;
++ } else {
++ printk(KERN_ERR "GSI %u is too high\n", gsi);
++ return gsi;
++ }
++ }
++#endif
++ 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 --git a/arch/x86/kernel/nmi_32.c b/arch/x86/kernel/nmi_32.c
+index 84160f7..60dbfd7 100644
+--- a/arch/x86/kernel/nmi_32.c
++++ b/arch/x86/kernel/nmi_32.c
+@@ -29,7 +29,10 @@
+
+ #include "mach_traps.h"
+
+-int unknown_nmi_panic;
++extern void die_nmi(struct pt_regs *, const char *msg);
++
++#ifndef CONFIG_XEN
++
+ int nmi_watchdog_enabled;
+
+ static cpumask_t backtrace_mask = CPU_MASK_NONE;
+@@ -314,8 +317,6 @@ void touch_nmi_watchdog(void)
+ }
+ EXPORT_SYMBOL(touch_nmi_watchdog);
+
+-extern void die_nmi(struct pt_regs *, const char *msg);
+-
+ notrace __kprobes int
+ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
+ {
+@@ -388,8 +389,12 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
+ return rc;
+ }
+
++#endif /* CONFIG_XEN */
++
+ #ifdef CONFIG_SYSCTL
+
++int unknown_nmi_panic;
++
+ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
+ {
+ unsigned char reason = get_nmi_reason();
+@@ -400,6 +405,7 @@ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
+ return 0;
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * proc handler for /proc/sys/kernel/nmi
+ */
+@@ -438,6 +444,7 @@ int proc_nmi_enabled(struct ctl_table *table, int write, struct file *file,
+ }
+ return 0;
+ }
++#endif
+
+ #endif
+
+@@ -450,6 +457,7 @@ int do_nmi_callback(struct pt_regs *regs, int cpu)
+ return 0;
+ }
+
++#ifndef CONFIG_XEN
+ void __trigger_all_cpu_backtrace(void)
+ {
+ int i;
+@@ -465,3 +473,4 @@ void __trigger_all_cpu_backtrace(void)
+
+ EXPORT_SYMBOL(nmi_active);
+ EXPORT_SYMBOL(nmi_watchdog);
++#endif
+diff --git a/arch/x86/kernel/nmi_64.c b/arch/x86/kernel/nmi_64.c
+index 5a29ded..ddd2120 100644
+--- a/arch/x86/kernel/nmi_64.c
++++ b/arch/x86/kernel/nmi_64.c
+@@ -28,10 +28,12 @@
+
+ #include <mach_traps.h>
+
+-int unknown_nmi_panic;
+-int nmi_watchdog_enabled;
+ int panic_on_unrecovered_nmi;
+
++#ifndef CONFIG_XEN
++
++int nmi_watchdog_enabled;
++
+ static cpumask_t backtrace_mask = CPU_MASK_NONE;
+
+ /* nmi_active:
+@@ -383,6 +385,8 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
+ return rc;
+ }
+
++#endif /* CONFIG_XEN */
++
+ static unsigned ignore_nmis;
+
+ asmlinkage notrace __kprobes void
+@@ -409,6 +413,8 @@ void restart_nmi(void)
+
+ #ifdef CONFIG_SYSCTL
+
++int unknown_nmi_panic;
++
+ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
+ {
+ unsigned char reason = get_nmi_reason();
+@@ -419,6 +425,7 @@ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu)
+ return 0;
+ }
+
++#ifndef CONFIG_XEN
+ /*
+ * proc handler for /proc/sys/kernel/nmi
+ */
+@@ -453,6 +460,7 @@ int proc_nmi_enabled(struct ctl_table *table, int write, struct file *file,
+ }
+ return 0;
+ }
++#endif
+
+ #endif
+
+@@ -465,6 +473,7 @@ int do_nmi_callback(struct pt_regs *regs, int cpu)
+ return 0;
+ }
+
++#ifndef CONFIG_XEN
+ void __trigger_all_cpu_backtrace(void)
+ {
+ int i;
+@@ -480,3 +489,4 @@ void __trigger_all_cpu_backtrace(void)
+
+ EXPORT_SYMBOL(nmi_active);
+ EXPORT_SYMBOL(nmi_watchdog);
++#endif
+diff --git a/arch/x86/kernel/pci-dma-xen.c b/arch/x86/kernel/pci-dma-xen.c
+new file mode 100644
+index 0000000..ada8f02
+--- /dev/null
++++ b/arch/x86/kernel/pci-dma-xen.c
+@@ -0,0 +1,599 @@
++#include <linux/dma-mapping.h>
++#include <linux/dmar.h>
++#include <linux/bootmem.h>
++#include <linux/pci.h>
++
++#include <asm/proto.h>
++#include <asm/dma.h>
++#include <asm/gart.h>
++#include <asm/calgary.h>
++
++int forbid_dac __read_mostly;
++EXPORT_SYMBOL(forbid_dac);
++
++const struct dma_mapping_ops *dma_ops;
++EXPORT_SYMBOL(dma_ops);
++
++static int iommu_sac_force __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
++
++int iommu_merge __read_mostly = 0;
++
++int no_iommu __read_mostly;
++/* Set this to 1 if there is a HW IOMMU in the system */
++int iommu_detected __read_mostly = 0;
++
++/* 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);
++
++dma_addr_t bad_dma_address __read_mostly = 0;
++EXPORT_SYMBOL(bad_dma_address);
++
++/* Dummy device used for NULL arguments (normally ISA). Better would
++ be probably a smaller DMA mask, but this is bug-to-bug compatible
++ to older i386. */
++struct device fallback_dev = {
++ .bus_id = "fallback device",
++ .coherent_dma_mask = DMA_32BIT_MASK,
++ .dma_mask = &fallback_dev.coherent_dma_mask,
++};
++
++int dma_set_mask(struct device *dev, u64 mask)
++{
++ if (!dev->dma_mask || !dma_supported(dev, mask))
++ return -EIO;
++
++ *dev->dma_mask = mask;
++
++ return 0;
++}
++EXPORT_SYMBOL(dma_set_mask);
++
++#if defined(CONFIG_X86_64) && !defined(CONFIG_XEN)
++static __initdata void *dma32_bootmem_ptr;
++static unsigned long dma32_bootmem_size __initdata = (128ULL<<20);
++
++static int __init parse_dma32_size_opt(char *p)
++{
++ if (!p)
++ return -EINVAL;
++ dma32_bootmem_size = memparse(p, &p);
++ return 0;
++}
++early_param("dma32_size", parse_dma32_size_opt);
++
++void __init dma32_reserve_bootmem(void)
++{
++ unsigned long size, align;
++ if (end_pfn <= MAX_DMA32_PFN)
++ return;
++
++ align = 64ULL<<20;
++ size = round_up(dma32_bootmem_size, align);
++ dma32_bootmem_ptr = __alloc_bootmem_nopanic(size, align,
++ __pa(MAX_DMA_ADDRESS));
++ if (dma32_bootmem_ptr)
++ dma32_bootmem_size = size;
++ else
++ dma32_bootmem_size = 0;
++}
++static void __init dma32_free_bootmem(void)
++{
++ int node;
++
++ if (end_pfn <= MAX_DMA32_PFN)
++ return;
++
++ if (!dma32_bootmem_ptr)
++ return;
++
++ for_each_online_node(node)
++ free_bootmem_node(NODE_DATA(node), __pa(dma32_bootmem_ptr),
++ dma32_bootmem_size);
++
++ dma32_bootmem_ptr = NULL;
++ dma32_bootmem_size = 0;
++}
++#else
++#define dma32_free_bootmem() ((void)0)
++#endif
++
++static const struct dma_mapping_ops swiotlb_dma_ops = {
++ .mapping_error = swiotlb_dma_mapping_error,
++ .map_single = swiotlb_map_single_phys,
++ .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 = swiotlb_dma_supported
++};
++
++void __init pci_iommu_alloc(void)
++{
++ /* free the range so iommu could get some range less than 4G */
++ dma32_free_bootmem();
++ /*
++ * The order of these functions is important for
++ * fall-back/fail-over reasons
++ */
++#ifdef CONFIG_GART_IOMMU
++ gart_iommu_hole_init();
++#endif
++
++#ifdef CONFIG_CALGARY_IOMMU
++ detect_calgary();
++#endif
++
++ detect_intel_iommu();
++
++#ifdef CONFIG_SWIOTLB
++ swiotlb_init();
++ if (swiotlb) {
++ printk(KERN_INFO "PCI-DMA: Using software bounce buffering for IO (SWIOTLB)\n");
++ dma_ops = &swiotlb_dma_ops;
++ }
++#endif
++}
++
++/*
++ * See <Documentation/x86_64/boot-options.txt> for the iommu kernel parameter
++ * documentation.
++ */
++static __init int iommu_setup(char *p)
++{
++ iommu_merge = 1;
++
++ if (!p)
++ return -EINVAL;
++
++ while (*p) {
++ if (!strncmp(p, "off", 3))
++ no_iommu = 1;
++ /* gart_parse_options has more force support */
++ if (!strncmp(p, "force", 5))
++ force_iommu = 1;
++ if (!strncmp(p, "noforce", 7)) {
++ iommu_merge = 0;
++ force_iommu = 0;
++ }
++
++ if (!strncmp(p, "biomerge", 8)) {
++ iommu_bio_merge = 4096;
++ iommu_merge = 1;
++ force_iommu = 1;
++ }
++ if (!strncmp(p, "panic", 5))
++ panic_on_overflow = 1;
++ if (!strncmp(p, "nopanic", 7))
++ panic_on_overflow = 0;
++ if (!strncmp(p, "merge", 5)) {
++ iommu_merge = 1;
++ force_iommu = 1;
++ }
++ if (!strncmp(p, "nomerge", 7))
++ iommu_merge = 0;
++ if (!strncmp(p, "forcesac", 8))
++ iommu_sac_force = 1;
++ if (!strncmp(p, "allowdac", 8))
++ forbid_dac = 0;
++ if (!strncmp(p, "nodac", 5))
++ forbid_dac = -1;
++ if (!strncmp(p, "usedac", 6)) {
++ forbid_dac = -1;
++ return 1;
++ }
++#ifdef CONFIG_SWIOTLB
++ if (!strncmp(p, "soft", 4))
++ swiotlb = 1;
++#endif
++
++#ifdef CONFIG_GART_IOMMU
++ gart_parse_options(p);
++#endif
++
++#ifdef CONFIG_CALGARY_IOMMU
++ if (!strncmp(p, "calgary", 7))
++ use_calgary = 1;
++#endif /* CONFIG_CALGARY_IOMMU */
++
++ p += strcspn(p, ",");
++ if (*p == ',')
++ ++p;
++ }
++ return 0;
++}
++early_param("iommu", iommu_setup);
++
++static int check_pages_physically_contiguous(unsigned long pfn,
++ unsigned int offset,
++ size_t length)
++{
++ unsigned long next_mfn;
++ int i;
++ int nr_pages;
++
++ next_mfn = pfn_to_mfn(pfn);
++ nr_pages = (offset + length + PAGE_SIZE-1) >> PAGE_SHIFT;
++
++ for (i = 1; i < nr_pages; i++) {
++ if (pfn_to_mfn(++pfn) != ++next_mfn)
++ return 0;
++ }
++ return 1;
++}
++
++int range_straddles_page_boundary(paddr_t p, size_t size)
++{
++ extern unsigned long *contiguous_bitmap;
++ unsigned long pfn = p >> PAGE_SHIFT;
++ unsigned int offset = p & ~PAGE_MASK;
++
++ if (offset + size <= PAGE_SIZE)
++ return 0;
++ if (test_bit(pfn, contiguous_bitmap))
++ return 0;
++ if (check_pages_physically_contiguous(pfn, offset, size))
++ return 0;
++ return 1;
++}
++
++#ifdef CONFIG_X86_32
++int dma_declare_coherent_memory(struct device *dev, dma_addr_t bus_addr,
++ dma_addr_t device_addr, size_t size, int flags)
++{
++ void __iomem *mem_base = NULL;
++ int pages = size >> PAGE_SHIFT;
++ int bitmap_size = BITS_TO_LONGS(pages) * sizeof(long);
++
++ if ((flags & (DMA_MEMORY_MAP | DMA_MEMORY_IO)) == 0)
++ goto out;
++ if (!size)
++ goto out;
++ if (dev->dma_mem)
++ goto out;
++
++ /* FIXME: this routine just ignores DMA_MEMORY_INCLUDES_CHILDREN */
++
++ mem_base = ioremap(bus_addr, size);
++ if (!mem_base)
++ goto out;
++
++ 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;
++
++ dev->dma_mem->virt_base = mem_base;
++ dev->dma_mem->device_base = device_addr;
++ dev->dma_mem->size = pages;
++ dev->dma_mem->flags = flags;
++
++ if (flags & DMA_MEMORY_MAP)
++ return DMA_MEMORY_MAP;
++
++ return DMA_MEMORY_IO;
++
++ free1_out:
++ kfree(dev->dma_mem);
++ out:
++ if (mem_base)
++ iounmap(mem_base);
++ return 0;
++}
++EXPORT_SYMBOL(dma_declare_coherent_memory);
++
++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);
++
++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 pos, err;
++ int pages = (size + (device_addr & ~PAGE_MASK) + PAGE_SIZE - 1);
++
++ pages >>= PAGE_SHIFT;
++
++ if (!mem)
++ return ERR_PTR(-EINVAL);
++
++ 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);
++
++static int dma_alloc_from_coherent_mem(struct device *dev, ssize_t size,
++ dma_addr_t *dma_handle, void **ret)
++{
++ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
++ int order = get_order(size);
++
++ 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);
++ }
++ if (mem->flags & DMA_MEMORY_EXCLUSIVE)
++ *ret = NULL;
++ }
++ return (mem != NULL);
++}
++
++static int dma_release_coherent(struct device *dev, int order, void *vaddr)
++{
++ struct dma_coherent_mem *mem = dev ? dev->dma_mem : NULL;
++
++ 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);
++ return 1;
++ }
++ return 0;
++}
++#else
++#define dma_alloc_from_coherent_mem(dev, size, handle, ret) (0)
++#define dma_release_coherent(dev, order, vaddr) (0)
++#endif /* CONFIG_X86_32 */
++
++int dma_supported(struct device *dev, u64 mask)
++{
++#ifdef CONFIG_PCI
++ if (mask > 0xffffffff && forbid_dac > 0) {
++ printk(KERN_INFO "PCI: Disallowing DAC for device %s\n",
++ dev->bus_id);
++ return 0;
++ }
++#endif
++
++ if (dma_ops->dma_supported)
++ return dma_ops->dma_supported(dev, mask);
++
++ /* Copied from i386. Doesn't make much sense, because it will
++ only work for pci_alloc_coherent.
++ The caller just has to use GFP_DMA in this case. */
++ if (mask < DMA_24BIT_MASK)
++ return 0;
++
++ /* Tell the device to use SAC when IOMMU force is on. This
++ allows the driver to use cheaper accesses in some cases.
++
++ Problem with this is that if we overflow the IOMMU area and
++ return DAC as fallback address the device may not handle it
++ correctly.
++
++ As a special case some controllers have a 39bit address
++ mode that is as efficient as 32bit (aic79xx). Don't force
++ SAC for these. Assume all masks <= 40 bits are of this
++ type. Normally this doesn't make any difference, but gives
++ more gentle handling of IOMMU overflow. */
++ if (iommu_sac_force && (mask >= DMA_40BIT_MASK)) {
++ printk(KERN_INFO "%s: Force SAC with mask %Lx\n",
++ dev->bus_id, mask);
++ return 0;
++ }
++
++ return 1;
++}
++EXPORT_SYMBOL(dma_supported);
++
++/* Allocate DMA memory on node near device */
++static struct page *
++dma_alloc_pages(struct device *dev, gfp_t gfp, unsigned order)
++{
++ int node;
++
++ node = dev_to_node(dev);
++
++ return alloc_pages_node(node, gfp, order);
++}
++
++/*
++ * Allocate memory for a coherent mapping.
++ */
++void *
++dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle,
++ gfp_t gfp)
++{
++ void *memory = NULL;
++ struct page *page;
++ unsigned long dma_mask = 0;
++ int noretry = 0;
++ unsigned int order = get_order(size);
++
++ /* ignore region specifiers */
++ gfp &= ~(__GFP_DMA | __GFP_HIGHMEM | __GFP_DMA32);
++
++ if (dma_alloc_from_coherent_mem(dev, size, dma_handle, &memory))
++ return memory;
++
++ if (!dev) {
++ dev = &fallback_dev;
++ gfp |= GFP_DMA;
++ }
++ dma_mask = dev->coherent_dma_mask;
++ if (dma_mask == 0)
++ dma_mask = (gfp & GFP_DMA) ? DMA_24BIT_MASK : DMA_32BIT_MASK;
++
++ /* Device not DMA able */
++ if (dev->dma_mask == NULL)
++ return NULL;
++
++ /* Don't invoke OOM killer or retry in lower 16MB DMA zone */
++ if (gfp & __GFP_DMA)
++ noretry = 1;
++
++#ifdef CONFIG_XEN
++ gfp &= ~(__GFP_DMA | __GFP_DMA32);
++#else
++#ifdef CONFIG_X86_64
++ /* Why <=? Even when the mask is smaller than 4GB it is often
++ larger than 16MB and in this case we have a chance of
++ finding fitting memory in the next higher zone first. If
++ not retry with true GFP_DMA. -AK */
++ if (dma_mask <= DMA_32BIT_MASK && !(gfp & GFP_DMA)) {
++ gfp |= GFP_DMA32;
++#endif
++
++ again:
++#endif
++ page = dma_alloc_pages(dev,
++ noretry ? gfp | __GFP_NORETRY : gfp, get_order(size));
++ if (page == NULL)
++ return NULL;
++
++#ifndef CONFIG_XEN
++ {
++ int high, mmu;
++ dma_addr_t bus = page_to_phys(page);
++ memory = page_address(page);
++ high = (bus + size) >= dma_mask;
++ mmu = high;
++ if (force_iommu && !(gfp & GFP_DMA))
++ mmu = 1;
++ else if (high) {
++ free_pages((unsigned long)memory, order);
++
++ /* Don't use the 16MB ZONE_DMA unless absolutely
++ needed. It's better to use remapping first. */
++ if (dma_mask < DMA_32BIT_MASK && !(gfp & GFP_DMA)) {
++ gfp = (gfp & ~GFP_DMA32) | GFP_DMA;
++ goto again;
++ }
++
++ /* Let low level make its own zone decisions */
++ gfp &= ~(GFP_DMA32|GFP_DMA);
++
++ if (dma_ops->alloc_coherent)
++ return dma_ops->alloc_coherent(dev, size,
++ dma_handle, gfp);
++ return NULL;
++ }
++
++ memset(memory, 0, size);
++ if (!mmu) {
++ *dma_handle = bus;
++ return memory;
++ }
++ }
++
++ if (dma_ops->alloc_coherent) {
++ free_pages((unsigned long)memory, order);
++ gfp &= ~(GFP_DMA|GFP_DMA32);
++ return dma_ops->alloc_coherent(dev, size, dma_handle, gfp);
++ }
++
++ if (dma_ops->map_simple) {
++ *dma_handle = dma_ops->map_simple(dev, virt_to_bus(memory),
++ size,
++ PCI_DMA_BIDIRECTIONAL);
++ if (*dma_handle != bad_dma_address)
++ return memory;
++ }
++#else
++ memory = page_address(page);
++ if (xen_create_contiguous_region((unsigned long)memory, order,
++ fls64(dma_mask)) == 0) {
++ memset(memory, 0, size);
++ *dma_handle = virt_to_bus(memory);
++ return memory;
++ }
++#endif
++
++ if (panic_on_overflow)
++ panic("dma_alloc_coherent: IOMMU overflow by %lu bytes\n",
++ (unsigned long)size);
++ free_pages((unsigned long)memory, get_order(size));
++ return NULL;
++}
++EXPORT_SYMBOL(dma_alloc_coherent);
++
++/*
++ * Unmap coherent memory.
++ * The caller must ensure that the device has finished accessing the mapping.
++ */
++void dma_free_coherent(struct device *dev, size_t size,
++ void *vaddr, dma_addr_t bus)
++{
++ int order = get_order(size);
++ WARN_ON(irqs_disabled()); /* for portability */
++ if (dma_release_coherent(dev, order, vaddr))
++ return;
++#ifndef CONFIG_XEN
++ if (dma_ops->unmap_single)
++ dma_ops->unmap_single(dev, bus, size, 0);
++#endif
++ xen_destroy_contiguous_region((unsigned long)vaddr, order);
++ free_pages((unsigned long)vaddr, order);
++}
++EXPORT_SYMBOL(dma_free_coherent);
++
++static int __init pci_iommu_init(void)
++{
++#ifdef CONFIG_CALGARY_IOMMU
++ calgary_iommu_init();
++#endif
++
++ intel_iommu_init();
++
++#ifdef CONFIG_GART_IOMMU
++ gart_iommu_init();
++#endif
++
++ no_iommu_init();
++ return 0;
++}
++
++void pci_iommu_shutdown(void)
++{
++ gart_iommu_shutdown();
++}
++/* Must execute after PCI subsystem */
++fs_initcall(pci_iommu_init);
++
++#ifdef CONFIG_PCI
++/* Many VIA bridges seem to corrupt data for DAC. Disable it here */
++
++static __devinit void via_no_dac(struct pci_dev *dev)
++{
++ if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) {
++ printk(KERN_INFO "PCI: VIA PCI bridge detected."
++ "Disabling DAC.\n");
++ forbid_dac = 1;
++ }
++}
++DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, via_no_dac);
++#endif
+diff --git a/arch/x86/kernel/pci-nommu-xen.c b/arch/x86/kernel/pci-nommu-xen.c
+new file mode 100644
+index 0000000..7e92955
+--- /dev/null
++++ b/arch/x86/kernel/pci-nommu-xen.c
+@@ -0,0 +1,103 @@
++#include <linux/dma-mapping.h>
++#include <linux/dmar.h>
++#include <linux/bootmem.h>
++#include <linux/pci.h>
++
++#include <xen/gnttab.h>
++
++#include <asm/proto.h>
++#include <asm/dma.h>
++#include <asm/swiotlb.h>
++#include <asm/tlbflush.h>
++#include <asm/gnttab_dma.h>
++#include <asm/bug.h>
++
++#define IOMMU_BUG_ON(test) \
++do { \
++ if (unlikely(test)) { \
++ printk(KERN_ALERT "Fatal DMA error! " \
++ "Please use 'swiotlb=force'\n"); \
++ BUG(); \
++ } \
++} while (0)
++
++static int
++gnttab_map_sg(struct device *hwdev, struct scatterlist *sgl, int nents,
++ int direction)
++{
++ unsigned int i;
++ struct scatterlist *sg;
++
++ WARN_ON(nents == 0 || sgl->length == 0);
++
++ for_each_sg(sgl, sg, nents, i) {
++ BUG_ON(!sg_page(sg));
++ sg->dma_address =
++ gnttab_dma_map_page(sg_page(sg)) + sg->offset;
++ sg->dma_length = sg->length;
++ IOMMU_BUG_ON(address_needs_mapping(
++ hwdev, sg->dma_address));
++ IOMMU_BUG_ON(range_straddles_page_boundary(
++ page_to_pseudophys(sg_page(sg)) + sg->offset,
++ sg->length));
++ }
++
++ return nents;
++}
++
++static void
++gnttab_unmap_sg(struct device *hwdev, struct scatterlist *sgl, int nents,
++ int direction)
++{
++ unsigned int i;
++ struct scatterlist *sg;
++
++ for_each_sg(sgl, sg, nents, i)
++ gnttab_dma_unmap_page(sg->dma_address);
++}
++
++static dma_addr_t
++gnttab_map_single(struct device *dev, phys_addr_t paddr, size_t size,
++ int direction)
++{
++ dma_addr_t dma;
++
++ WARN_ON(size == 0);
++
++ dma = gnttab_dma_map_page(pfn_to_page(paddr >> PAGE_SHIFT)) +
++ offset_in_page(paddr);
++ IOMMU_BUG_ON(range_straddles_page_boundary(paddr, size));
++ IOMMU_BUG_ON(address_needs_mapping(dev, dma));
++
++ return dma;
++}
++
++static void
++gnttab_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
++ int direction)
++{
++ gnttab_dma_unmap_page(dma_addr);
++}
++
++static int nommu_mapping_error(dma_addr_t dma_addr)
++{
++ return (dma_addr == bad_dma_address);
++}
++
++static const struct dma_mapping_ops nommu_dma_ops = {
++ .map_single = gnttab_map_single,
++ .unmap_single = gnttab_unmap_single,
++ .map_sg = gnttab_map_sg,
++ .unmap_sg = gnttab_unmap_sg,
++ .dma_supported = swiotlb_dma_supported,
++ .mapping_error = nommu_mapping_error
++};
++
++void __init no_iommu_init(void)
++{
++ if (dma_ops)
++ return;
++
++ force_iommu = 0; /* no HW IOMMU */
++ dma_ops = &nommu_dma_ops;
++}
+diff --git a/arch/x86/kernel/pcspeaker.c b/arch/x86/kernel/pcspeaker.c
+index bc1f2d3..a2c342f 100644
+--- a/arch/x86/kernel/pcspeaker.c
++++ b/arch/x86/kernel/pcspeaker.c
+@@ -7,6 +7,11 @@ static __init int add_pcspkr(void)
+ struct platform_device *pd;
+ int ret;
+
++#ifdef CONFIG_XEN
++ if (!is_initial_xendomain())
++ return 0;
++#endif
++
+ pd = platform_device_alloc("pcspkr", -1);
+ if (!pd)
+ return -ENOMEM;
+diff --git a/arch/x86/kernel/process-xen.c b/arch/x86/kernel/process-xen.c
+new file mode 100644
+index 0000000..88f164b
+--- /dev/null
++++ b/arch/x86/kernel/process-xen.c
+@@ -0,0 +1,188 @@
++#include <linux/errno.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/slab.h>
++#include <linux/sched.h>
++#include <linux/module.h>
++#include <linux/pm.h>
++
++struct kmem_cache *task_xstate_cachep;
++
++int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
++{
++ *dst = *src;
++ if (src->thread.xstate) {
++ dst->thread.xstate = kmem_cache_alloc(task_xstate_cachep,
++ GFP_KERNEL);
++ if (!dst->thread.xstate)
++ return -ENOMEM;
++ WARN_ON((unsigned long)dst->thread.xstate & 15);
++ memcpy(dst->thread.xstate, src->thread.xstate, xstate_size);
++ }
++ return 0;
++}
++
++void free_thread_xstate(struct task_struct *tsk)
++{
++ if (tsk->thread.xstate) {
++ kmem_cache_free(task_xstate_cachep, tsk->thread.xstate);
++ tsk->thread.xstate = NULL;
++ }
++}
++
++void free_thread_info(struct thread_info *ti)
++{
++ free_thread_xstate(ti->task);
++ free_pages((unsigned long)ti, get_order(THREAD_SIZE));
++}
++
++void arch_task_cache_init(void)
++{
++ task_xstate_cachep =
++ kmem_cache_create("task_xstate", xstate_size,
++ __alignof__(union thread_xstate),
++ SLAB_PANIC, NULL);
++}
++
++static void do_nothing(void *unused)
++{
++}
++
++/*
++ * cpu_idle_wait - Used to ensure that all the CPUs discard old value of
++ * pm_idle and update to new pm_idle value. Required while changing pm_idle
++ * handler on SMP systems.
++ *
++ * Caller must have changed pm_idle to the new value before the call. Old
++ * pm_idle value will not be used by any CPU after the return of this function.
++ */
++void cpu_idle_wait(void)
++{
++ smp_mb();
++ /* kick all the CPUs so that they exit out of pm_idle */
++ smp_call_function(do_nothing, NULL, 0, 1);
++}
++EXPORT_SYMBOL_GPL(cpu_idle_wait);
++
++#ifndef CONFIG_XEN
++/*
++ * This uses new MONITOR/MWAIT instructions on P4 processors with PNI,
++ * which can obviate IPI to trigger checking of need_resched.
++ * We execute MONITOR against need_resched and enter optimized wait state
++ * through MWAIT. Whenever someone changes need_resched, we would be woken
++ * up from MWAIT (without an IPI).
++ *
++ * New with Core Duo processors, MWAIT can take some hints based on CPU
++ * capability.
++ */
++void mwait_idle_with_hints(unsigned long ax, unsigned long cx)
++{
++ if (!need_resched()) {
++ __monitor((void *)¤t_thread_info()->flags, 0, 0);
++ smp_mb();
++ if (!need_resched())
++ __mwait(ax, cx);
++ }
++}
++
++/* Default MONITOR/MWAIT with no hints, used for default C1 state */
++static void mwait_idle(void)
++{
++ if (!need_resched()) {
++ __monitor((void *)¤t_thread_info()->flags, 0, 0);
++ smp_mb();
++ if (!need_resched())
++ __sti_mwait(0, 0);
++ else
++ local_irq_enable();
++ } else
++ local_irq_enable();
++}
++#endif
++
++/*
++ * On SMP it's slightly faster (but much more power-consuming!)
++ * to poll the ->work.need_resched flag instead of waiting for the
++ * cross-CPU IPI to arrive. Use this option with caution.
++ */
++static void poll_idle(void)
++{
++ local_irq_enable();
++ cpu_relax();
++}
++
++#ifndef CONFIG_XEN
++/*
++ * mwait selection logic:
++ *
++ * It depends on the CPU. For AMD CPUs that support MWAIT this is
++ * wrong. Family 0x10 and 0x11 CPUs will enter C1 on HLT. Powersavings
++ * then depend on a clock divisor and current Pstate of the core. If
++ * all cores of a processor are in halt state (C1) the processor can
++ * enter the C1E (C1 enhanced) state. If mwait is used this will never
++ * happen.
++ *
++ * idle=mwait overrides this decision and forces the usage of mwait.
++ */
++static int __cpuinit mwait_usable(const struct cpuinfo_x86 *c)
++{
++ if (force_mwait)
++ return 1;
++
++ if (c->x86_vendor == X86_VENDOR_AMD) {
++ switch(c->x86) {
++ case 0x10:
++ case 0x11:
++ return 0;
++ }
++ }
++ return 1;
++}
++#endif
++
++void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
++{
++#ifndef CONFIG_XEN
++ static int selected;
++
++ if (selected)
++ return;
++#ifdef CONFIG_X86_SMP
++ if (pm_idle == poll_idle && smp_num_siblings > 1) {
++ printk(KERN_WARNING "WARNING: polling idle and HT enabled,"
++ " performance may degrade.\n");
++ }
++#endif
++ if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) {
++ /*
++ * Skip, if setup has overridden idle.
++ * One CPU supports mwait => All CPUs supports mwait
++ */
++ if (!pm_idle) {
++ printk(KERN_INFO "using mwait in idle threads.\n");
++ pm_idle = mwait_idle;
++ }
++ }
++ selected = 1;
++#endif
++}
++
++static int __init idle_setup(char *str)
++{
++ if (!strcmp(str, "poll")) {
++ printk("using polling idle threads.\n");
++ pm_idle = poll_idle;
++ }
++#ifndef CONFIG_XEN
++ else if (!strcmp(str, "mwait"))
++ force_mwait = 1;
++#endif
++ else
++ return -1;
++
++ boot_option_idle_override = 1;
++ return 0;
++}
++early_param("idle", idle_setup);
++
+diff --git a/arch/x86/kernel/process_32-xen.c b/arch/x86/kernel/process_32-xen.c
+new file mode 100644
+index 0000000..86dc5ed
+--- /dev/null
++++ b/arch/x86/kernel/process_32-xen.c
+@@ -0,0 +1,775 @@
++/*
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
++ */
++
++/*
++ * 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/fs.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/elfcore.h>
++#include <linux/smp.h>
++#include <linux/stddef.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/user.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>
++#include <linux/tick.h>
++#include <linux/percpu.h>
++#include <linux/prctl.h>
++
++#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>
++#ifdef CONFIG_MATH_EMULATION
++#include <asm/math_emu.h>
++#endif
++
++#include <xen/interface/physdev.h>
++#include <xen/interface/vcpu.h>
++#include <xen/cpu_hotplug.h>
++
++#include <linux/err.h>
++
++#include <asm/tlbflush.h>
++#include <asm/cpu.h>
++#include <asm/kdebug.h>
++
++asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
++asmlinkage void cstar_ret_from_fork(void) __asm__("cstar_ret_from_fork");
++
++static int hlt_counter;
++
++unsigned long boot_option_idle_override = 0;
++EXPORT_SYMBOL(boot_option_idle_override);
++
++DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task;
++EXPORT_PER_CPU_SYMBOL(current_task);
++
++DEFINE_PER_CPU(int, cpu_number);
++EXPORT_PER_CPU_SYMBOL(cpu_number);
++
++/*
++ * Return saved PC of a blocked thread.
++ */
++unsigned long thread_saved_pc(struct task_struct *tsk)
++{
++ return ((unsigned long *)tsk->thread.sp)[3];
++}
++
++/*
++ * Powermanagement idle function, if any..
++ */
++void (*pm_idle)(void);
++EXPORT_SYMBOL(pm_idle);
++
++void disable_hlt(void)
++{
++ hlt_counter++;
++}
++
++EXPORT_SYMBOL(disable_hlt);
++
++void enable_hlt(void)
++{
++ hlt_counter--;
++}
++
++EXPORT_SYMBOL(enable_hlt);
++
++static void xen_idle(void)
++{
++ current_thread_info()->status &= ~TS_POLLING;
++ /*
++ * TS_POLLING-cleared state must be visible before we
++ * test NEED_RESCHED:
++ */
++ smp_mb();
++
++ 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
++
++#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();
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
++ cpu_bringup();
++}
++#else
++static inline void play_dead(void)
++{
++ BUG();
++}
++#endif /* CONFIG_HOTPLUG_CPU */
++
++/*
++ * The idle thread. There's no useful work to be
++ * done, so just try to conserve power and have a
++ * low exit latency (ie sit in a loop waiting for
++ * somebody to say that they'd like to reschedule)
++ */
++void cpu_idle(void)
++{
++ int cpu = smp_processor_id();
++
++ current_thread_info()->status |= TS_POLLING;
++
++ /* endless idle loop with no priority at all */
++ while (1) {
++ tick_nohz_stop_sched_tick();
++ while (!need_resched()) {
++ void (*idle)(void);
++
++ check_pgt_cache();
++ rmb();
++ idle = xen_idle; /* no alternatives */
++
++ if (rcu_pending(cpu))
++ rcu_check_callbacks(cpu, 0);
++
++ if (cpu_is_offline(cpu))
++ play_dead();
++
++ local_irq_disable();
++ __get_cpu_var(irq_stat).idle_timestamp = jiffies;
++ idle();
++ }
++ tick_nohz_restart_sched_tick();
++ preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
++ }
++}
++
++void __show_registers(struct pt_regs *regs, int all)
++{
++ unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L;
++ unsigned long d0, d1, d2, d3, d6, d7;
++ unsigned long sp;
++ unsigned short ss, gs;
++
++ if (user_mode_vm(regs)) {
++ sp = regs->sp;
++ ss = regs->ss & 0xffff;
++ savesegment(gs, gs);
++ } else {
++ sp = (unsigned long) (®s->sp);
++ savesegment(ss, ss);
++ savesegment(gs, gs);
++ }
++
++ printk("\n");
++ printk("Pid: %d, comm: %s %s (%s %.*s)\n",
++ task_pid_nr(current), current->comm,
++ print_tainted(), init_utsname()->release,
++ (int)strcspn(init_utsname()->version, " "),
++ init_utsname()->version);
++
++ printk("EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n",
++ (u16)regs->cs, regs->ip, regs->flags,
++ smp_processor_id());
++ print_symbol("EIP is at %s\n", regs->ip);
++
++ printk("EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n",
++ regs->ax, regs->bx, regs->cx, regs->dx);
++ printk("ESI: %08lx EDI: %08lx EBP: %08lx ESP: %08lx\n",
++ regs->si, regs->di, regs->bp, sp);
++ printk(" DS: %04x ES: %04x FS: %04x GS: %04x SS: %04x\n",
++ (u16)regs->ds, (u16)regs->es, (u16)regs->fs, gs, ss);
++
++ if (!all)
++ return;
++
++ 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);
++
++ get_debugreg(d0, 0);
++ get_debugreg(d1, 1);
++ get_debugreg(d2, 2);
++ get_debugreg(d3, 3);
++ printk("DR0: %08lx DR1: %08lx DR2: %08lx DR3: %08lx\n",
++ d0, d1, d2, d3);
++
++ get_debugreg(d6, 6);
++ get_debugreg(d7, 7);
++ printk("DR6: %08lx DR7: %08lx\n",
++ d6, d7);
++}
++
++void show_regs(struct pt_regs *regs)
++{
++ __show_registers(regs, 1);
++ show_trace(NULL, regs, ®s->sp, regs->bp);
++}
++
++/*
++ * This gets run with %bx containing the
++ * function to call, and %dx containing
++ * the "args".
++ */
++extern void kernel_thread_helper(void);
++
++/*
++ * Create a kernel thread
++ */
++int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
++{
++ struct pt_regs regs;
++
++ memset(®s, 0, sizeof(regs));
++
++ regs.bx = (unsigned long) fn;
++ regs.dx = (unsigned long) arg;
++
++ regs.ds = __USER_DS;
++ regs.es = __USER_DS;
++ regs.fs = __KERNEL_PERCPU;
++ regs.orig_ax = -1;
++ regs.ip = (unsigned long) kernel_thread_helper;
++ regs.cs = __KERNEL_CS | get_kernel_rpl();
++ regs.flags = X86_EFLAGS_IF | X86_EFLAGS_SF | X86_EFLAGS_PF | 0x2;
++
++ /* Ok, create the new process.. */
++ return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL);
++}
++EXPORT_SYMBOL(kernel_thread);
++
++/*
++ * 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;
++ struct physdev_set_iobitmap set_iobitmap;
++ memset(&set_iobitmap, 0, sizeof(set_iobitmap));
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &set_iobitmap));
++ kfree(t->io_bitmap_ptr);
++ t->io_bitmap_ptr = NULL;
++ clear_thread_flag(TIF_IO_BITMAP);
++ }
++}
++
++void flush_thread(void)
++{
++ struct task_struct *tsk = current;
++
++ 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));
++ clear_tsk_thread_flag(tsk, TIF_DEBUG);
++ /*
++ * Forget coprocessor state..
++ */
++ tsk->fpu_counter = 0;
++ clear_fpu(tsk);
++ clear_used_math();
++}
++
++void release_thread(struct task_struct *dead_task)
++{
++ BUG_ON(dead_task->mm);
++ release_vm86_irqs(dead_task);
++}
++
++/*
++ * This gets called before we allocate a new thread and copy
++ * the current task into it.
++ */
++void prepare_to_copy(struct task_struct *tsk)
++{
++ unlazy_fpu(tsk);
++}
++
++int copy_thread(int nr, unsigned long clone_flags, unsigned long sp,
++ unsigned long unused,
++ struct task_struct * p, struct pt_regs * regs)
++{
++ struct pt_regs * childregs;
++ struct task_struct *tsk;
++ int err;
++
++ childregs = task_pt_regs(p);
++ *childregs = *regs;
++ childregs->ax = 0;
++ childregs->sp = sp;
++
++ p->thread.sp = (unsigned long) childregs;
++ p->thread.sp0 = (unsigned long) (childregs+1);
++
++ p->thread.ip = (unsigned long) ret_from_fork;
++
++ savesegment(gs, p->thread.gs);
++
++ tsk = current;
++ if (test_tsk_thread_flag(tsk, TIF_CSTAR))
++ p->thread.ip = (unsigned long) cstar_ret_from_fork;
++ 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;
++ }
++ set_tsk_thread_flag(p, TIF_IO_BITMAP);
++ }
++
++ err = 0;
++
++ /*
++ * Set a new TLS for the child thread?
++ */
++ if (clone_flags & CLONE_SETTLS)
++ err = do_set_thread_area(p, -1,
++ (struct user_desc __user *)childregs->si, 0);
++
++ p->thread.iopl = current->thread.iopl;
++
++ if (err && p->thread.io_bitmap_ptr) {
++ kfree(p->thread.io_bitmap_ptr);
++ p->thread.io_bitmap_max = 0;
++ }
++ return err;
++}
++
++void
++start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp)
++{
++ __asm__("movl %0, %%gs" :: "r"(0));
++ regs->fs = 0;
++ set_fs(USER_DS);
++ regs->ds = __USER_DS;
++ regs->es = __USER_DS;
++ regs->ss = __USER_DS;
++ regs->cs = __USER_CS;
++ regs->ip = new_ip;
++ regs->sp = new_sp;
++ /*
++ * Free the old FP and other extended state
++ */
++ free_thread_xstate(current);
++}
++EXPORT_SYMBOL_GPL(start_thread);
++
++static void hard_disable_TSC(void)
++{
++ write_cr4(read_cr4() | X86_CR4_TSD);
++}
++
++void disable_TSC(void)
++{
++#ifdef CONFIG_SECCOMP_DISABLE_TSC
++ preempt_disable();
++ if (!test_and_set_thread_flag(TIF_NOTSC))
++ /*
++ * Must flip the CPU state synchronously with
++ * TIF_NOTSC in the current running context.
++ */
++ hard_disable_TSC();
++ preempt_enable();
++#endif
++}
++
++static void hard_enable_TSC(void)
++{
++ write_cr4(read_cr4() & ~X86_CR4_TSD);
++}
++
++static void enable_TSC(void)
++{
++ preempt_disable();
++ if (test_and_clear_thread_flag(TIF_NOTSC))
++ /*
++ * Must flip the CPU state synchronously with
++ * TIF_NOTSC in the current running context.
++ */
++ hard_enable_TSC();
++ preempt_enable();
++}
++
++int get_tsc_mode(unsigned long adr)
++{
++ unsigned int val;
++
++ if (test_thread_flag(TIF_NOTSC))
++ val = PR_TSC_SIGSEGV;
++ else
++ val = PR_TSC_ENABLE;
++
++ return put_user(val, (unsigned int __user *)adr);
++}
++
++int set_tsc_mode(unsigned int val)
++{
++ if (val == PR_TSC_SIGSEGV)
++ disable_TSC();
++ else if (val == PR_TSC_ENABLE)
++ enable_TSC();
++ else
++ return -EINVAL;
++
++ return 0;
++}
++
++static noinline void
++__switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p)
++{
++ struct thread_struct *prev, *next;
++ unsigned long debugctl;
++
++ prev = &prev_p->thread;
++ next = &next_p->thread;
++
++ debugctl = prev->debugctlmsr;
++ if (next->ds_area_msr != prev->ds_area_msr) {
++ /* we clear debugctl to make sure DS
++ * is not in use when we change it */
++ debugctl = 0;
++ update_debugctlmsr(0);
++ wrmsr(MSR_IA32_DS_AREA, next->ds_area_msr, 0);
++ }
++
++ if (next->debugctlmsr != debugctl)
++ update_debugctlmsr(next->debugctlmsr);
++
++ if (test_tsk_thread_flag(next_p, TIF_DEBUG)) {
++ 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);
++ }
++
++ if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
++ test_tsk_thread_flag(next_p, TIF_NOTSC)) {
++ /* prev and next are different */
++ if (test_tsk_thread_flag(next_p, TIF_NOTSC))
++ hard_disable_TSC();
++ else
++ hard_enable_TSC();
++ }
++
++#ifdef X86_BTS
++ if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
++ ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);
++
++ if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
++ ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
++#endif
++}
++
++/*
++ * 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 %ax) will be the "prev" task after
++ * the task-switch, and shows up in ret_from_fork in entry.S,
++ * for example.
++ */
++struct task_struct * __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
++{
++ struct thread_struct *prev = &prev_p->thread,
++ *next = &next_p->thread;
++ int cpu = smp_processor_id();
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *tss = &per_cpu(init_tss, cpu);
++#endif
++#if CONFIG_XEN_COMPAT > 0x030002
++ struct physdev_set_iopl iopl_op;
++ struct physdev_set_iobitmap iobmp_op;
++#else
++ struct physdev_op _pdo[2], *pdo = _pdo;
++#define iopl_op pdo->u.set_iopl
++#define iobmp_op pdo->u.set_iobitmap
++#endif
++ multicall_entry_t _mcl[8], *mcl = _mcl;
++
++ /* XEN NOTE: FS/GS saved in switch_mm(), not here. */
++
++ /*
++ * This is basically '__unlazy_fpu', except that we queue a
++ * multicall to indicate FPU task switch, rather than
++ * synchronously trapping to Xen.
++ */
++ if (task_thread_info(prev_p)->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
++
++ /*
++ * Reload sp0.
++ * This is load_sp0(tss, next) with a multicall.
++ */
++ mcl->op = __HYPERVISOR_stack_switch;
++ mcl->args[0] = __KERNEL_DS;
++ mcl->args[1] = next->sp0;
++ 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].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->iopl != next->iopl)) {
++ iopl_op.iopl = (next->iopl == 0) ? 1 : (next->iopl >> 12) & 3;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iopl;
++ mcl->args[1] = (unsigned long)&iopl_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iopl;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
++ set_xen_guest_handle(iobmp_op.bitmap,
++ (char *)next->io_bitmap_ptr);
++ iobmp_op.nr_ports = next->io_bitmap_ptr ? IO_BITMAP_BITS : 0;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iobitmap;
++ mcl->args[1] = (unsigned long)&iobmp_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iobitmap;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ BUG_ON(pdo > _pdo + ARRAY_SIZE(_pdo));
++#endif
++ BUG_ON(mcl > _mcl + ARRAY_SIZE(_mcl));
++ if (unlikely(HYPERVISOR_multicall_check(_mcl, mcl - _mcl, NULL)))
++ BUG();
++
++ /* we're going to use this soon, after a few expensive things */
++ if (next_p->fpu_counter > 5)
++ prefetch(next->xstate);
++
++ /*
++ * Now maybe handle debug registers
++ */
++ if (unlikely(task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV ||
++ task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT))
++ __switch_to_xtra(prev_p, next_p);
++
++ /*
++ * Leave lazy mode, flushing any hypercalls made here.
++ * This must be done before restoring TLS segments so
++ * the GDT and LDT are properly updated, and must be
++ * done before math_state_restore, so the TS bit is up
++ * to date.
++ */
++ arch_leave_lazy_cpu_mode();
++
++ /* 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
++ *
++ * tsk_used_math() checks prevent calling math_state_restore(),
++ * which can sleep in the case of !tsk_used_math()
++ */
++ if (tsk_used_math(next_p) && next_p->fpu_counter > 5)
++ math_state_restore();
++
++ /*
++ * Restore %gs if needed (which is common)
++ */
++ if (prev->gs | next->gs)
++ loadsegment(gs, next->gs);
++
++ x86_write_percpu(current_task, next_p);
++
++ return prev_p;
++}
++
++asmlinkage int sys_fork(struct pt_regs regs)
++{
++ return do_fork(SIGCHLD, regs.sp, ®s, 0, NULL, NULL);
++}
++
++asmlinkage int sys_clone(struct pt_regs regs)
++{
++ unsigned long clone_flags;
++ unsigned long newsp;
++ int __user *parent_tidptr, *child_tidptr;
++
++ clone_flags = regs.bx;
++ newsp = regs.cx;
++ parent_tidptr = (int __user *)regs.dx;
++ child_tidptr = (int __user *)regs.di;
++ if (!newsp)
++ newsp = regs.sp;
++ return do_fork(clone_flags, newsp, ®s, 0, parent_tidptr, child_tidptr);
++}
++
++/*
++ * 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 int sys_vfork(struct pt_regs regs)
++{
++ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs.sp, ®s, 0, NULL, NULL);
++}
++
++/*
++ * sys_execve() executes a new program.
++ */
++asmlinkage int sys_execve(struct pt_regs regs)
++{
++ int error;
++ char * filename;
++
++ filename = getname((char __user *) regs.bx);
++ error = PTR_ERR(filename);
++ if (IS_ERR(filename))
++ goto out;
++ error = do_execve(filename,
++ (char __user * __user *) regs.cx,
++ (char __user * __user *) regs.dx,
++ ®s);
++ if (error == 0) {
++ /* Make sure we don't return using sysenter.. */
++ set_thread_flag(TIF_IRET);
++ }
++ putname(filename);
++out:
++ return error;
++}
++
++#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)
++{
++ unsigned long bp, sp, ip;
++ 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);
++ sp = p->thread.sp;
++ if (!stack_page || sp < stack_page || sp > top_esp+stack_page)
++ return 0;
++ /* include/asm-i386/system.h:switch_to() pushes bp last. */
++ bp = *(unsigned long *) sp;
++ do {
++ if (bp < stack_page || bp > top_ebp+stack_page)
++ return 0;
++ ip = *(unsigned long *) (bp+4);
++ if (!in_sched_functions(ip))
++ return ip;
++ bp = *(unsigned long *) bp;
++ } while (count++ < 16);
++ return 0;
++}
++
++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;
++}
++
++unsigned long arch_randomize_brk(struct mm_struct *mm)
++{
++ unsigned long range_end = mm->brk + 0x02000000;
++ return randomize_range(mm->brk, range_end, 0) ? : mm->brk;
++}
+diff --git a/arch/x86/kernel/process_64-xen.c b/arch/x86/kernel/process_64-xen.c
+new file mode 100644
+index 0000000..893fb36
+--- /dev/null
++++ b/arch/x86/kernel/process_64-xen.c
+@@ -0,0 +1,923 @@
++/*
++ * 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/fs.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/interrupt.h>
++#include <linux/utsname.h>
++#include <linux/delay.h>
++#include <linux/module.h>
++#include <linux/ptrace.h>
++#include <linux/random.h>
++#include <linux/notifier.h>
++#include <linux/kprobes.h>
++#include <linux/kdebug.h>
++#include <linux/tick.h>
++#include <linux/prctl.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 <xen/interface/platform.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/vcpu.h>
++#include <asm/desc.h>
++#include <asm/proto.h>
++#include <asm/hardirq.h>
++#include <asm/ia32.h>
++#include <asm/idle.h>
++
++#include <xen/cpu_hotplug.h>
++
++asmlinkage extern void ret_from_fork(void);
++
++unsigned long kernel_thread_flags = CLONE_VM | CLONE_UNTRACED;
++
++unsigned long boot_option_idle_override = 0;
++EXPORT_SYMBOL(boot_option_idle_override);
++
++/*
++ * Powermanagement idle function, if any..
++ */
++void (*pm_idle)(void);
++EXPORT_SYMBOL(pm_idle);
++
++static ATOMIC_NOTIFIER_HEAD(idle_notifier);
++
++void idle_notifier_register(struct notifier_block *n)
++{
++ atomic_notifier_chain_register(&idle_notifier, n);
++}
++
++void enter_idle(void)
++{
++ write_pda(isidle, 1);
++ atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL);
++}
++
++static void __exit_idle(void)
++{
++ if (test_and_clear_bit_pda(0, isidle) == 0)
++ return;
++ atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL);
++}
++
++/* Called from interrupts to signify idle end */
++void exit_idle(void)
++{
++ /* idle loop has pid 0 */
++ if (current->pid)
++ return;
++ __exit_idle();
++}
++
++static void xen_idle(void)
++{
++ current_thread_info()->status &= ~TS_POLLING;
++ /*
++ * TS_POLLING-cleared state must be visible before we
++ * test NEED_RESCHED:
++ */
++ smp_mb();
++ if (!need_resched())
++ safe_halt(); /* enables interrupts racelessly */
++ else
++ local_irq_enable();
++ current_thread_info()->status |= TS_POLLING;
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++static inline void play_dead(void)
++{
++ idle_task_exit();
++ local_irq_disable();
++ cpu_clear(smp_processor_id(), cpu_initialized);
++ preempt_enable_no_resched();
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
++ cpu_bringup();
++}
++#else
++static inline void play_dead(void)
++{
++ BUG();
++}
++#endif /* CONFIG_HOTPLUG_CPU */
++
++/*
++ * The idle thread. There's no useful work to be
++ * done, so just try to conserve power and have a
++ * low exit latency (ie sit in a loop waiting for
++ * somebody to say that they'd like to reschedule)
++ */
++void cpu_idle(void)
++{
++ current_thread_info()->status |= TS_POLLING;
++ /* endless idle loop with no priority at all */
++ while (1) {
++ tick_nohz_stop_sched_tick();
++ while (!need_resched()) {
++ void (*idle)(void);
++
++ rmb();
++ idle = xen_idle; /* no alternatives */
++ 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();
++ 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();
++ }
++
++ tick_nohz_restart_sched_tick();
++ preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
++ }
++}
++
++/* 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 long d0, d1, d2, d3, d6, d7;
++ 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(),
++ init_utsname()->release,
++ (int)strcspn(init_utsname()->version, " "),
++ init_utsname()->version);
++ printk("RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->ip);
++ printk_address(regs->ip, 1);
++ printk("RSP: %04lx:%016lx EFLAGS: %08lx\n", regs->ss, regs->sp,
++ regs->flags);
++ printk("RAX: %016lx RBX: %016lx RCX: %016lx\n",
++ regs->ax, regs->bx, regs->cx);
++ printk("RDX: %016lx RSI: %016lx RDI: %016lx\n",
++ regs->dx, regs->si, regs->di);
++ printk("RBP: %016lx R08: %016lx R09: %016lx\n",
++ regs->bp, regs->r8, regs->r9);
++ printk("R10: %016lx R11: %016lx R12: %016lx\n",
++ regs->r10, regs->r11, regs->r12);
++ printk("R13: %016lx R14: %016lx R15: %016lx\n",
++ regs->r13, regs->r14, regs->r15);
++
++ asm("mov %%ds,%0" : "=r" (ds));
++ asm("mov %%cs,%0" : "=r" (cs));
++ asm("mov %%es,%0" : "=r" (es));
++ asm("mov %%fs,%0" : "=r" (fsindex));
++ asm("mov %%gs,%0" : "=r" (gsindex));
++
++ rdmsrl(MSR_FS_BASE, fs);
++ rdmsrl(MSR_GS_BASE, gs);
++ rdmsrl(MSR_KERNEL_GS_BASE, shadowgs);
++
++ printk("FS: %016lx(%04x) GS:%016lx(%04x) knlGS:%016lx\n",
++ fs,fsindex,gs,gsindex,shadowgs);
++ printk("CS: %04x DS: %04x ES: %04x\n", cs, ds, es);
++
++ get_debugreg(d0, 0);
++ get_debugreg(d1, 1);
++ get_debugreg(d2, 2);
++ printk("DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2);
++ get_debugreg(d3, 3);
++ get_debugreg(d6, 6);
++ get_debugreg(d7, 7);
++ printk("DR3: %016lx DR6: %016lx DR7: %016lx\n", d3, d6, d7);
++}
++
++void show_regs(struct pt_regs *regs)
++{
++ printk("CPU %d:", smp_processor_id());
++ __show_regs(regs);
++ show_trace(NULL, regs, (void *)(regs + 1), regs->bp);
++}
++
++/*
++ * Free current thread data structures etc..
++ */
++void exit_thread(void)
++{
++ struct task_struct *me = current;
++ struct thread_struct *t = &me->thread;
++
++ if (me->thread.io_bitmap_ptr) {
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *tss = &per_cpu(init_tss, get_cpu());
++#endif
++#ifdef CONFIG_XEN
++ struct physdev_set_iobitmap iobmp_op;
++ memset(&iobmp_op, 0, sizeof(iobmp_op));
++#endif
++
++ kfree(t->io_bitmap_ptr);
++ t->io_bitmap_ptr = NULL;
++ 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
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap,
++ &iobmp_op));
++#endif
++ t->io_bitmap_max = 0;
++ }
++}
++
++void load_gs_index(unsigned gs)
++{
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, gs));
++}
++
++void flush_thread(void)
++{
++ struct task_struct *tsk = current;
++
++ if (test_tsk_thread_flag(tsk, TIF_ABI_PENDING)) {
++ clear_tsk_thread_flag(tsk, TIF_ABI_PENDING);
++ if (test_tsk_thread_flag(tsk, TIF_IA32)) {
++ clear_tsk_thread_flag(tsk, TIF_IA32);
++ } else {
++ set_tsk_thread_flag(tsk, TIF_IA32);
++ current_thread_info()->status |= TS_COMPAT;
++ }
++ }
++ clear_tsk_thread_flag(tsk, 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..
++ */
++ tsk->fpu_counter = 0;
++ clear_fpu(tsk);
++ clear_used_math();
++}
++
++void release_thread(struct task_struct *dead_task)
++{
++ if (dead_task->mm) {
++ if (dead_task->mm->context.size) {
++ printk("WARNING: dead process %8s still has LDT? <%p/%d>\n",
++ dead_task->comm,
++ dead_task->mm->context.ldt,
++ dead_task->mm->context.size);
++ BUG();
++ }
++ }
++}
++
++static inline void set_32bit_tls(struct task_struct *t, int tls, u32 addr)
++{
++ struct user_desc ud = {
++ .base_addr = addr,
++ .limit = 0xfffff,
++ .seg_32bit = 1,
++ .limit_in_pages = 1,
++ .useable = 1,
++ };
++ struct desc_struct *desc = t->thread.tls_array;
++ desc += tls;
++ fill_ldt(desc, &ud);
++}
++
++static inline u32 read_32bit_tls(struct task_struct *t, int tls)
++{
++ return get_desc_base(&t->thread.tls_array[tls]);
++}
++
++/*
++ * This gets called before we allocate a new thread and copy
++ * the current task into it.
++ */
++void prepare_to_copy(struct task_struct *tsk)
++{
++ unlazy_fpu(tsk);
++}
++
++int copy_thread(int nr, unsigned long clone_flags, unsigned long sp,
++ unsigned long unused,
++ struct task_struct * p, struct pt_regs * regs)
++{
++ int err;
++ struct pt_regs * childregs;
++ struct task_struct *me = current;
++
++ childregs = ((struct pt_regs *)
++ (THREAD_SIZE + task_stack_page(p))) - 1;
++ *childregs = *regs;
++
++ childregs->ax = 0;
++ childregs->sp = sp;
++ if (sp == ~0UL)
++ childregs->sp = (unsigned long)childregs;
++
++ p->thread.sp = (unsigned long) childregs;
++ p->thread.sp0 = (unsigned long) (childregs+1);
++ p->thread.usersp = me->thread.usersp;
++
++ set_tsk_thread_flag(p, TIF_FORK);
++
++ p->thread.fs = me->thread.fs;
++ p->thread.gs = me->thread.gs;
++
++ asm("mov %%gs,%0" : "=m" (p->thread.gsindex));
++ asm("mov %%fs,%0" : "=m" (p->thread.fsindex));
++ asm("mov %%es,%0" : "=m" (p->thread.es));
++ asm("mov %%ds,%0" : "=m" (p->thread.ds));
++
++ if (unlikely(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);
++ }
++
++ /*
++ * Set a new TLS for the child thread?
++ */
++ if (clone_flags & CLONE_SETTLS) {
++#ifdef CONFIG_IA32_EMULATION
++ if (test_thread_flag(TIF_IA32))
++ err = do_set_thread_area(p, -1,
++ (struct user_desc __user *)childregs->si, 0);
++ else
++#endif
++ err = do_arch_prctl(p, ARCH_SET_FS, childregs->r8);
++ if (err)
++ goto out;
++ }
++ p->thread.iopl = current->thread.iopl;
++
++ err = 0;
++out:
++ if (err && p->thread.io_bitmap_ptr) {
++ kfree(p->thread.io_bitmap_ptr);
++ p->thread.io_bitmap_max = 0;
++ }
++ return err;
++}
++
++void
++start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp)
++{
++ asm volatile("movl %0, %%fs; movl %0, %%es; movl %0, %%ds" :: "r"(0));
++ load_gs_index(0);
++ regs->ip = new_ip;
++ regs->sp = new_sp;
++ write_pda(oldrsp, new_sp);
++ regs->cs = __USER_CS;
++ regs->ss = __USER_DS;
++ regs->flags = 0x200;
++ set_fs(USER_DS);
++ /*
++ * Free the old FP and other extended state
++ */
++ free_thread_xstate(current);
++}
++EXPORT_SYMBOL_GPL(start_thread);
++
++static void hard_disable_TSC(void)
++{
++ write_cr4(read_cr4() | X86_CR4_TSD);
++}
++
++void disable_TSC(void)
++{
++#ifdef CONFIG_SECCOMP_DISABLE_TSC
++ preempt_disable();
++ if (!test_and_set_thread_flag(TIF_NOTSC))
++ /*
++ * Must flip the CPU state synchronously with
++ * TIF_NOTSC in the current running context.
++ */
++ hard_disable_TSC();
++ preempt_enable();
++#endif
++}
++
++static void hard_enable_TSC(void)
++{
++ write_cr4(read_cr4() & ~X86_CR4_TSD);
++}
++
++static void enable_TSC(void)
++{
++ preempt_disable();
++ if (test_and_clear_thread_flag(TIF_NOTSC))
++ /*
++ * Must flip the CPU state synchronously with
++ * TIF_NOTSC in the current running context.
++ */
++ hard_enable_TSC();
++ preempt_enable();
++}
++
++int get_tsc_mode(unsigned long adr)
++{
++ unsigned int val;
++
++ if (test_thread_flag(TIF_NOTSC))
++ val = PR_TSC_SIGSEGV;
++ else
++ val = PR_TSC_ENABLE;
++
++ return put_user(val, (unsigned int __user *)adr);
++}
++
++int set_tsc_mode(unsigned int val)
++{
++ if (val == PR_TSC_SIGSEGV)
++ disable_TSC();
++ else if (val == PR_TSC_ENABLE)
++ enable_TSC();
++ else
++ return -EINVAL;
++
++ return 0;
++}
++
++/*
++ * 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;
++ unsigned long debugctl;
++
++ prev = &prev_p->thread,
++ next = &next_p->thread;
++
++ debugctl = prev->debugctlmsr;
++ if (next->ds_area_msr != prev->ds_area_msr) {
++ /* we clear debugctl to make sure DS
++ * is not in use when we change it */
++ debugctl = 0;
++ update_debugctlmsr(0);
++ wrmsrl(MSR_IA32_DS_AREA, next->ds_area_msr);
++ }
++
++ if (next->debugctlmsr != debugctl)
++ update_debugctlmsr(next->debugctlmsr);
++
++ 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);
++ }
++
++ if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
++ test_tsk_thread_flag(next_p, TIF_NOTSC)) {
++ /* prev and next are different */
++ if (test_tsk_thread_flag(next_p, TIF_NOTSC))
++ hard_disable_TSC();
++ else
++ hard_enable_TSC();
++ }
++
++#ifdef X86_BTS
++ if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
++ ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);
++
++ if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
++ ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
++#endif
++}
++
++/*
++ * 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.
++ */
++struct task_struct *
++__switch_to(struct task_struct *prev_p, struct task_struct *next_p)
++{
++ struct thread_struct *prev = &prev_p->thread,
++ *next = &next_p->thread;
++ int cpu = smp_processor_id();
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *tss = &per_cpu(init_tss, cpu);
++#endif
++#if CONFIG_XEN_COMPAT > 0x030002
++ struct physdev_set_iopl iopl_op;
++ struct physdev_set_iobitmap iobmp_op;
++#else
++ struct physdev_op _pdo[2], *pdo = _pdo;
++#define iopl_op pdo->u.set_iopl
++#define iobmp_op pdo->u.set_iobitmap
++#endif
++ multicall_entry_t _mcl[8], *mcl = _mcl;
++
++ /* we're going to use this soon, after a few expensive things */
++ if (next_p->fpu_counter>5)
++ prefetch(next->xstate);
++
++ /*
++ * This is basically '__unlazy_fpu', except that we queue a
++ * multicall to indicate FPU task switch, rather than
++ * synchronously trapping to Xen.
++ * The AMD workaround requires it to be after DS reload, or
++ * after DS has been cleared, which we do in __prepare_arch_switch.
++ */
++ if (task_thread_info(prev_p)->status & TS_USEDFPU) {
++ __save_init_fpu(prev_p); /* _not_ save_init_fpu() */
++ mcl->op = __HYPERVISOR_fpu_taskswitch;
++ mcl->args[0] = 1;
++ mcl++;
++ } else
++ prev_p->fpu_counter = 0;
++
++ /*
++ * Reload sp0.
++ * This is load_sp0(tss, next) with a multicall.
++ */
++ mcl->op = __HYPERVISOR_stack_switch;
++ mcl->args[0] = __KERNEL_DS;
++ mcl->args[1] = next->sp0;
++ 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].a != prev->tls_array[i].a || \
++ next->tls_array[i].b != prev->tls_array[i].b)) { \
++ mcl->op = __HYPERVISOR_update_descriptor; \
++ mcl->args[0] = virt_to_machine( \
++ &get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i]);\
++ mcl->args[1] = *(u64 *)&next->tls_array[i]; \
++ mcl++; \
++ } \
++} while (0)
++ C(0); C(1); C(2);
++#undef C
++
++ if (unlikely(prev->iopl != next->iopl)) {
++ iopl_op.iopl = (next->iopl == 0) ? 1 : (next->iopl >> 12) & 3;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iopl;
++ mcl->args[1] = (unsigned long)&iopl_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iopl;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++ if (unlikely(prev->io_bitmap_ptr || next->io_bitmap_ptr)) {
++ set_xen_guest_handle(iobmp_op.bitmap,
++ (char *)next->io_bitmap_ptr);
++ iobmp_op.nr_ports = next->io_bitmap_ptr ? IO_BITMAP_BITS : 0;
++#if CONFIG_XEN_COMPAT > 0x030002
++ mcl->op = __HYPERVISOR_physdev_op;
++ mcl->args[0] = PHYSDEVOP_set_iobitmap;
++ mcl->args[1] = (unsigned long)&iobmp_op;
++#else
++ mcl->op = __HYPERVISOR_physdev_op_compat;
++ pdo->cmd = PHYSDEVOP_set_iobitmap;
++ mcl->args[0] = (unsigned long)pdo++;
++#endif
++ mcl++;
++ }
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ BUG_ON(pdo > _pdo + ARRAY_SIZE(_pdo));
++#endif
++ BUG_ON(mcl > _mcl + ARRAY_SIZE(_mcl));
++ if (unlikely(HYPERVISOR_multicall_check(_mcl, mcl - _mcl, NULL)))
++ BUG();
++
++ /*
++ * Switch DS and ES.
++ * This won't pick up thread selector changes, but I guess that is ok.
++ */
++ if (unlikely(next->es))
++ loadsegment(es, next->es);
++
++ if (unlikely(next->ds))
++ loadsegment(ds, next->ds);
++
++ /*
++ * Switch FS and GS.
++ */
++ if (unlikely(next->fsindex))
++ loadsegment(fs, next->fsindex);
++
++ if (next->fs)
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_FS, next->fs));
++
++ if (unlikely(next->gsindex))
++ load_gs_index(next->gsindex);
++
++ if (next->gs)
++ WARN_ON(HYPERVISOR_set_segment_base(SEGBASE_GS_USER, next->gs));
++
++ /*
++ * Switch the PDA context.
++ */
++ prev->usersp = read_pda(oldrsp);
++ write_pda(oldrsp, next->usersp);
++ 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);
++
++ /*
++ * Build time only check to make sure the stack_canary is at
++ * offset 40 in the pda; this is a gcc ABI requirement
++ */
++ BUILD_BUG_ON(offsetof(struct x8664_pda, stack_canary) != 40);
++#endif
++
++ /*
++ * Now maybe reload the debug registers
++ */
++ if (unlikely(task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT ||
++ task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV))
++ __switch_to_xtra(prev_p, next_p);
++
++ /* 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
++ *
++ * tsk_used_math() checks prevent calling math_state_restore(),
++ * which can sleep in the case of !tsk_used_math()
++ */
++ if (tsk_used_math(next_p) && next_p->fpu_counter > 5)
++ math_state_restore();
++ return prev_p;
++}
++
++/*
++ * sys_execve() executes a new program.
++ */
++asmlinkage
++long sys_execve(char __user *name, char __user * __user *argv,
++ char __user * __user *envp, struct pt_regs *regs)
++{
++ long error;
++ char * filename;
++
++ filename = getname(name);
++ error = PTR_ERR(filename);
++ if (IS_ERR(filename))
++ return error;
++ error = do_execve(filename, argv, envp, regs);
++ putname(filename);
++ return error;
++}
++
++void set_personality_64bit(void)
++{
++ /* inherit personality from parent */
++
++ /* Make sure to be in 64bit mode */
++ clear_thread_flag(TIF_IA32);
++
++ /* TBD: overwrites user setup. Should have two bits.
++ But 64bit processes have always behaved this way,
++ so it's not too bad. The main problem is just that
++ 32bit childs are affected again. */
++ current->personality &= ~READ_IMPLIES_EXEC;
++}
++
++asmlinkage long sys_fork(struct pt_regs *regs)
++{
++ return do_fork(SIGCHLD, regs->sp, regs, 0, NULL, NULL);
++}
++
++asmlinkage long
++sys_clone(unsigned long clone_flags, unsigned long newsp,
++ void __user *parent_tid, void __user *child_tid, struct pt_regs *regs)
++{
++ if (!newsp)
++ newsp = regs->sp;
++ return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid);
++}
++
++/*
++ * This is trivial, and on the face of it looks like it
++ * could equally well be done in user mode.
++ *
++ * Not so, for quite unobvious reasons - register pressure.
++ * In user mode vfork() cannot have a stack frame, and if
++ * done by calling the "clone()" system call directly, you
++ * do not have enough call-clobbered registers to hold all
++ * the information you need.
++ */
++asmlinkage long sys_vfork(struct pt_regs *regs)
++{
++ return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->sp, regs, 0,
++ NULL, NULL);
++}
++
++unsigned long get_wchan(struct task_struct *p)
++{
++ unsigned long stack;
++ u64 fp,ip;
++ int count = 0;
++
++ if (!p || p == current || p->state==TASK_RUNNING)
++ return 0;
++ stack = (unsigned long)task_stack_page(p);
++ if (p->thread.sp < stack || p->thread.sp > stack+THREAD_SIZE)
++ return 0;
++ fp = *(u64 *)(p->thread.sp);
++ do {
++ if (fp < (unsigned long)stack ||
++ fp > (unsigned long)stack+THREAD_SIZE)
++ return 0;
++ ip = *(u64 *)(fp+8);
++ if (!in_sched_functions(ip))
++ return ip;
++ fp = *(u64 *)fp;
++ } while (count++ < 16);
++ return 0;
++}
++
++long do_arch_prctl(struct task_struct *task, int code, unsigned long addr)
++{
++ int ret = 0;
++ int doit = task == current;
++ int cpu;
++
++ switch (code) {
++ case ARCH_SET_GS:
++ if (addr >= TASK_SIZE_OF(task))
++ return -EPERM;
++ cpu = get_cpu();
++ /* handle small bases via the GDT because that's faster to
++ switch. */
++ if (addr <= 0xffffffff) {
++ set_32bit_tls(task, GS_TLS, addr);
++ if (doit) {
++ load_TLS(&task->thread, cpu);
++ load_gs_index(GS_TLS_SEL);
++ }
++ task->thread.gsindex = GS_TLS_SEL;
++ task->thread.gs = 0;
++ } else {
++ task->thread.gsindex = 0;
++ task->thread.gs = addr;
++ if (doit) {
++ load_gs_index(0);
++ ret = HYPERVISOR_set_segment_base(
++ SEGBASE_GS_USER, addr);
++ }
++ }
++ put_cpu();
++ break;
++ case ARCH_SET_FS:
++ /* Not strictly needed for fs, but do it for symmetry
++ with gs */
++ if (addr >= TASK_SIZE_OF(task))
++ return -EPERM;
++ cpu = get_cpu();
++ /* handle small bases via the GDT because that's faster to
++ switch. */
++ if (addr <= 0xffffffff) {
++ set_32bit_tls(task, FS_TLS, addr);
++ if (doit) {
++ load_TLS(&task->thread, cpu);
++ asm volatile("movl %0,%%fs" :: "r"(FS_TLS_SEL));
++ }
++ task->thread.fsindex = FS_TLS_SEL;
++ task->thread.fs = 0;
++ } else {
++ task->thread.fsindex = 0;
++ task->thread.fs = addr;
++ if (doit) {
++ /* set the selector to 0 to not confuse
++ __switch_to */
++ asm volatile("movl %0,%%fs" :: "r" (0));
++ ret = HYPERVISOR_set_segment_base(SEGBASE_FS,
++ addr);
++ }
++ }
++ put_cpu();
++ break;
++ case ARCH_GET_FS: {
++ unsigned long base;
++ if (task->thread.fsindex == FS_TLS_SEL)
++ base = read_32bit_tls(task, FS_TLS);
++ else if (doit)
++ rdmsrl(MSR_FS_BASE, base);
++ else
++ base = task->thread.fs;
++ ret = put_user(base, (unsigned long __user *)addr);
++ break;
++ }
++ case ARCH_GET_GS: {
++ unsigned long base;
++ unsigned gsindex;
++ if (task->thread.gsindex == GS_TLS_SEL)
++ base = read_32bit_tls(task, GS_TLS);
++ else if (doit) {
++ asm("movl %%gs,%0" : "=r" (gsindex));
++ if (gsindex)
++ rdmsrl(MSR_KERNEL_GS_BASE, base);
++ else
++ base = task->thread.gs;
++ }
++ else
++ base = task->thread.gs;
++ ret = put_user(base, (unsigned long __user *)addr);
++ break;
++ }
++
++ default:
++ ret = -EINVAL;
++ break;
++ }
++
++ return ret;
++}
++
++long sys_arch_prctl(int code, unsigned long addr)
++{
++ return do_arch_prctl(current, code, addr);
++}
++
++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;
++}
++
++unsigned long arch_randomize_brk(struct mm_struct *mm)
++{
++ unsigned long range_end = mm->brk + 0x02000000;
++ return randomize_range(mm->brk, range_end, 0) ? : mm->brk;
++}
+diff --git a/arch/x86/kernel/quirks-xen.c b/arch/x86/kernel/quirks-xen.c
+new file mode 100644
+index 0000000..fddd924
+--- /dev/null
++++ b/arch/x86/kernel/quirks-xen.c
+@@ -0,0 +1,403 @@
++/*
++ * This file contains work-arounds for x86 and x86_64 platform bugs.
++ */
++#include <linux/pci.h>
++#include <linux/irq.h>
++
++#if defined(CONFIG_X86_IO_APIC) && (defined(CONFIG_SMP) || defined(CONFIG_XEN)) && defined(CONFIG_PCI)
++
++static void __devinit quirk_intel_irqbalance(struct pci_dev *dev)
++{
++ u8 config, rev;
++ u16 word;
++
++ /* BIOS may enable hardware IRQ balancing for
++ * E7520/E7320/E7525(revision ID 0x9 and below)
++ * based platforms.
++ * Disable SW irqbalance/affinity on those platforms.
++ */
++ pci_read_config_byte(dev, PCI_CLASS_REVISION, &rev);
++ if (rev > 0x9)
++ return;
++
++ /* enable access to config space*/
++ pci_read_config_byte(dev, 0xf4, &config);
++ pci_write_config_byte(dev, 0xf4, config|0x2);
++
++ /*
++ * read xTPR register. We may not have a pci_dev for device 8
++ * because it might be hidden until the above write.
++ */
++ pci_bus_read_config_word(dev->bus, PCI_DEVFN(8, 0), 0x4c, &word);
++
++ if (!(word & (1 << 13))) {
++ struct xen_platform_op op;
++
++ dev_info(&dev->dev, "Intel E7520/7320/7525 detected; "
++ "disabling irq balancing and affinity\n");
++ op.cmd = XENPF_platform_quirk;
++ op.u.platform_quirk.quirk_id = QUIRK_NOIRQBALANCING;
++ WARN_ON(HYPERVISOR_platform_op(&op));
++ }
++
++ /* 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
++
++#if defined(CONFIG_HPET_TIMER)
++#include <asm/hpet.h>
++
++unsigned long force_hpet_address;
++
++static enum {
++ NONE_FORCE_HPET_RESUME,
++ OLD_ICH_FORCE_HPET_RESUME,
++ ICH_FORCE_HPET_RESUME,
++ VT8237_FORCE_HPET_RESUME,
++ NVIDIA_FORCE_HPET_RESUME,
++} force_hpet_resume_type;
++
++static void __iomem *rcba_base;
++
++static void ich_force_hpet_resume(void)
++{
++ u32 val;
++
++ if (!force_hpet_address)
++ return;
++
++ if (rcba_base == NULL)
++ BUG();
++
++ /* read the Function Disable register, dword mode only */
++ val = readl(rcba_base + 0x3404);
++ if (!(val & 0x80)) {
++ /* HPET disabled in HPTC. Trying to enable */
++ writel(val | 0x80, rcba_base + 0x3404);
++ }
++
++ val = readl(rcba_base + 0x3404);
++ if (!(val & 0x80))
++ BUG();
++ else
++ printk(KERN_DEBUG "Force enabled HPET at resume\n");
++
++ return;
++}
++
++static void ich_force_enable_hpet(struct pci_dev *dev)
++{
++ u32 val;
++ u32 uninitialized_var(rcba);
++ int err = 0;
++
++ if (hpet_address || force_hpet_address)
++ return;
++
++ pci_read_config_dword(dev, 0xF0, &rcba);
++ rcba &= 0xFFFFC000;
++ if (rcba == 0) {
++ dev_printk(KERN_DEBUG, &dev->dev, "RCBA disabled; "
++ "cannot force enable HPET\n");
++ return;
++ }
++
++ /* use bits 31:14, 16 kB aligned */
++ rcba_base = ioremap_nocache(rcba, 0x4000);
++ if (rcba_base == NULL) {
++ dev_printk(KERN_DEBUG, &dev->dev, "ioremap failed; "
++ "cannot force enable HPET\n");
++ return;
++ }
++
++ /* read the Function Disable register, dword mode only */
++ val = readl(rcba_base + 0x3404);
++
++ if (val & 0x80) {
++ /* HPET is enabled in HPTC. Just not reported by BIOS */
++ val = val & 0x3;
++ force_hpet_address = 0xFED00000 | (val << 12);
++ dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at "
++ "0x%lx\n", force_hpet_address);
++ iounmap(rcba_base);
++ return;
++ }
++
++ /* HPET disabled in HPTC. Trying to enable */
++ writel(val | 0x80, rcba_base + 0x3404);
++
++ val = readl(rcba_base + 0x3404);
++ if (!(val & 0x80)) {
++ err = 1;
++ } else {
++ val = val & 0x3;
++ force_hpet_address = 0xFED00000 | (val << 12);
++ }
++
++ if (err) {
++ force_hpet_address = 0;
++ iounmap(rcba_base);
++ dev_printk(KERN_DEBUG, &dev->dev,
++ "Failed to force enable HPET\n");
++ } else {
++ force_hpet_resume_type = ICH_FORCE_HPET_RESUME;
++ dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at "
++ "0x%lx\n", force_hpet_address);
++ }
++}
++
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ESB2_0,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH6_1,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH7_0,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH7_1,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH7_31,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH8_1,
++ ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH9_7,
++ ich_force_enable_hpet);
++
++
++static struct pci_dev *cached_dev;
++
++static void old_ich_force_hpet_resume(void)
++{
++ u32 val;
++ u32 uninitialized_var(gen_cntl);
++
++ if (!force_hpet_address || !cached_dev)
++ return;
++
++ pci_read_config_dword(cached_dev, 0xD0, &gen_cntl);
++ gen_cntl &= (~(0x7 << 15));
++ gen_cntl |= (0x4 << 15);
++
++ pci_write_config_dword(cached_dev, 0xD0, gen_cntl);
++ pci_read_config_dword(cached_dev, 0xD0, &gen_cntl);
++ val = gen_cntl >> 15;
++ val &= 0x7;
++ if (val == 0x4)
++ printk(KERN_DEBUG "Force enabled HPET at resume\n");
++ else
++ BUG();
++}
++
++static void old_ich_force_enable_hpet(struct pci_dev *dev)
++{
++ u32 val;
++ u32 uninitialized_var(gen_cntl);
++
++ if (hpet_address || force_hpet_address)
++ return;
++
++ pci_read_config_dword(dev, 0xD0, &gen_cntl);
++ /*
++ * Bit 17 is HPET enable bit.
++ * Bit 16:15 control the HPET base address.
++ */
++ val = gen_cntl >> 15;
++ val &= 0x7;
++ if (val & 0x4) {
++ val &= 0x3;
++ force_hpet_address = 0xFED00000 | (val << 12);
++ dev_printk(KERN_DEBUG, &dev->dev, "HPET at 0x%lx\n",
++ force_hpet_address);
++ return;
++ }
++
++ /*
++ * HPET is disabled. Trying enabling at FED00000 and check
++ * whether it sticks
++ */
++ gen_cntl &= (~(0x7 << 15));
++ gen_cntl |= (0x4 << 15);
++ pci_write_config_dword(dev, 0xD0, gen_cntl);
++
++ pci_read_config_dword(dev, 0xD0, &gen_cntl);
++
++ val = gen_cntl >> 15;
++ val &= 0x7;
++ if (val & 0x4) {
++ /* HPET is enabled in HPTC. Just not reported by BIOS */
++ val &= 0x3;
++ force_hpet_address = 0xFED00000 | (val << 12);
++ dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at "
++ "0x%lx\n", force_hpet_address);
++ cached_dev = dev;
++ force_hpet_resume_type = OLD_ICH_FORCE_HPET_RESUME;
++ return;
++ }
++
++ dev_printk(KERN_DEBUG, &dev->dev, "Failed to force enable HPET\n");
++}
++
++/*
++ * Undocumented chipset features. Make sure that the user enforced
++ * this.
++ */
++static void old_ich_force_enable_hpet_user(struct pci_dev *dev)
++{
++ if (hpet_force_user)
++ old_ich_force_enable_hpet(dev);
++}
++
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801CA_0,
++ old_ich_force_enable_hpet_user);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801CA_12,
++ old_ich_force_enable_hpet_user);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801DB_0,
++ old_ich_force_enable_hpet_user);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801DB_12,
++ old_ich_force_enable_hpet_user);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801EB_0,
++ old_ich_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801EB_12,
++ old_ich_force_enable_hpet);
++
++
++static void vt8237_force_hpet_resume(void)
++{
++ u32 val;
++
++ if (!force_hpet_address || !cached_dev)
++ return;
++
++ val = 0xfed00000 | 0x80;
++ pci_write_config_dword(cached_dev, 0x68, val);
++
++ pci_read_config_dword(cached_dev, 0x68, &val);
++ if (val & 0x80)
++ printk(KERN_DEBUG "Force enabled HPET at resume\n");
++ else
++ BUG();
++}
++
++static void vt8237_force_enable_hpet(struct pci_dev *dev)
++{
++ u32 uninitialized_var(val);
++
++ if (!hpet_force_user || hpet_address || force_hpet_address)
++ return;
++
++ pci_read_config_dword(dev, 0x68, &val);
++ /*
++ * Bit 7 is HPET enable bit.
++ * Bit 31:10 is HPET base address (contrary to what datasheet claims)
++ */
++ if (val & 0x80) {
++ force_hpet_address = (val & ~0x3ff);
++ dev_printk(KERN_DEBUG, &dev->dev, "HPET at 0x%lx\n",
++ force_hpet_address);
++ return;
++ }
++
++ /*
++ * HPET is disabled. Trying enabling at FED00000 and check
++ * whether it sticks
++ */
++ val = 0xfed00000 | 0x80;
++ pci_write_config_dword(dev, 0x68, val);
++
++ pci_read_config_dword(dev, 0x68, &val);
++ if (val & 0x80) {
++ force_hpet_address = (val & ~0x3ff);
++ dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at "
++ "0x%lx\n", force_hpet_address);
++ cached_dev = dev;
++ force_hpet_resume_type = VT8237_FORCE_HPET_RESUME;
++ return;
++ }
++
++ dev_printk(KERN_DEBUG, &dev->dev, "Failed to force enable HPET\n");
++}
++
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8235,
++ vt8237_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8237,
++ vt8237_force_enable_hpet);
++
++/*
++ * Undocumented chipset feature taken from LinuxBIOS.
++ */
++static void nvidia_force_hpet_resume(void)
++{
++ pci_write_config_dword(cached_dev, 0x44, 0xfed00001);
++ printk(KERN_DEBUG "Force enabled HPET at resume\n");
++}
++
++static void nvidia_force_enable_hpet(struct pci_dev *dev)
++{
++ u32 uninitialized_var(val);
++
++ if (!hpet_force_user || hpet_address || force_hpet_address)
++ return;
++
++ pci_write_config_dword(dev, 0x44, 0xfed00001);
++ pci_read_config_dword(dev, 0x44, &val);
++ force_hpet_address = val & 0xfffffffe;
++ force_hpet_resume_type = NVIDIA_FORCE_HPET_RESUME;
++ dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at 0x%lx\n",
++ force_hpet_address);
++ cached_dev = dev;
++ return;
++}
++
++/* ISA Bridges */
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0050,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0051,
++ nvidia_force_enable_hpet);
++
++/* LPC bridges */
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0260,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0360,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0361,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0362,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0363,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0364,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0365,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0366,
++ nvidia_force_enable_hpet);
++DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, 0x0367,
++ nvidia_force_enable_hpet);
++
++void force_hpet_resume(void)
++{
++ switch (force_hpet_resume_type) {
++ case ICH_FORCE_HPET_RESUME:
++ ich_force_hpet_resume();
++ return;
++ case OLD_ICH_FORCE_HPET_RESUME:
++ old_ich_force_hpet_resume();
++ return;
++ case VT8237_FORCE_HPET_RESUME:
++ vt8237_force_hpet_resume();
++ return;
++ case NVIDIA_FORCE_HPET_RESUME:
++ nvidia_force_hpet_resume();
++ return;
++ default:
++ break;
++ }
++}
++
++#endif
+diff --git a/arch/x86/kernel/relocate_kernel_32.S b/arch/x86/kernel/relocate_kernel_32.S
+index c30fe25..64beabd 100644
+--- a/arch/x86/kernel/relocate_kernel_32.S
++++ b/arch/x86/kernel/relocate_kernel_32.S
+@@ -155,14 +155,45 @@ relocate_new_kernel:
+ movl PTR(PA_PGD)(%ebp), %eax
+ movl %eax, %cr3
+
++ /* setup idt */
++ movl %edi, %eax
++ addl $(idt_48 - relocate_kernel), %eax
++ lidtl (%eax)
++
++ /* setup gdt */
++ movl %edi, %eax
++ addl $(gdt - relocate_kernel), %eax
++ movl %edi, %esi
++ addl $((gdt_48 - relocate_kernel) + 2), %esi
++ movl %eax, (%esi)
++
++ movl %edi, %eax
++ addl $(gdt_48 - relocate_kernel), %eax
++ lgdtl (%eax)
++
++ /* setup data segment registers */
++ mov $(gdt_ds - gdt), %eax
++ mov %eax, %ds
++ mov %eax, %es
++ mov %eax, %fs
++ mov %eax, %gs
++ mov %eax, %ss
++
+ /* setup a new stack at the end of the physical control page */
+ lea PAGE_SIZE(%edi), %esp
+
+- /* jump to identity mapped page */
+- movl %edi, %eax
+- addl $(identity_mapped - relocate_kernel), %eax
+- pushl %eax
+- ret
++ /* load new code segment and jump to identity mapped page */
++ movl %edi, %esi
++ xorl %eax, %eax
++ pushl %eax
++ pushl %esi
++ pushl %eax
++ movl $(gdt_cs - gdt), %eax
++ pushl %eax
++ movl %edi, %eax
++ addl $(identity_mapped - relocate_kernel),%eax
++ pushl %eax
++ iretl
+
+ identity_mapped:
+ /* store the start address on the stack */
+@@ -250,3 +281,20 @@ identity_mapped:
+ xorl %edi, %edi
+ xorl %ebp, %ebp
+ ret
++
++ .align 16
++gdt:
++ .quad 0x0000000000000000 /* NULL descriptor */
++gdt_cs:
++ .quad 0x00cf9a000000ffff /* kernel 4GB code at 0x00000000 */
++gdt_ds:
++ .quad 0x00cf92000000ffff /* kernel 4GB data at 0x00000000 */
++gdt_end:
++
++gdt_48:
++ .word gdt_end - gdt - 1 /* limit */
++ .long 0 /* base - filled in by code above */
++
++idt_48:
++ .word 0 /* limit */
++ .long 0 /* base */
+diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S
+index f5afe66..4261d0e 100644
+--- a/arch/x86/kernel/relocate_kernel_64.S
++++ b/arch/x86/kernel/relocate_kernel_64.S
+@@ -160,13 +160,39 @@ relocate_new_kernel:
+ movq PTR(PA_PGD)(%rsi), %r9
+ movq %r9, %cr3
+
++ /* setup idt */
++ movq %r8, %rax
++ addq $(idt_80 - relocate_kernel), %rax
++ lidtq (%rax)
++
++ /* setup gdt */
++ movq %r8, %rax
++ addq $(gdt - relocate_kernel), %rax
++ movq %r8, %r9
++ addq $((gdt_80 - relocate_kernel) + 2), %r9
++ movq %rax, (%r9)
++
++ movq %r8, %rax
++ addq $(gdt_80 - relocate_kernel), %rax
++ lgdtq (%rax)
++
++ /* setup data segment registers */
++ xorl %eax, %eax
++ movl %eax, %ds
++ movl %eax, %es
++ movl %eax, %fs
++ movl %eax, %gs
++ movl %eax, %ss
++
+ /* setup a new stack at the end of the physical control page */
+ lea PAGE_SIZE(%r8), %rsp
+
+- /* jump to identity mapped page */
+- addq $(identity_mapped - relocate_kernel), %r8
+- pushq %r8
+- ret
++ /* load new code segment and jump to identity mapped page */
++ movq %r8, %rax
++ addq $(identity_mapped - relocate_kernel), %rax
++ pushq $(gdt_cs - gdt)
++ pushq %rax
++ lretq
+
+ identity_mapped:
+ /* store the start address on the stack */
+@@ -262,5 +288,19 @@ identity_mapped:
+ xorq %r13, %r13
+ xorq %r14, %r14
+ xorq %r15, %r15
+-
+ ret
++
++ .align 16
++gdt:
++ .quad 0x0000000000000000 /* NULL descriptor */
++gdt_cs:
++ .quad 0x00af9a000000ffff
++gdt_end:
++
++gdt_80:
++ .word gdt_end - gdt - 1 /* limit */
++ .quad 0 /* base - filled in by code above */
++
++idt_80:
++ .word 0 /* limit */
++ .quad 0 /* base */
+diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c
+index 05191bb..6dec56f 100644
+--- a/arch/x86/kernel/rtc.c
++++ b/arch/x86/kernel/rtc.c
+@@ -181,6 +181,10 @@ unsigned long read_persistent_clock(void)
+ {
+ unsigned long retval, flags;
+
++#ifdef CONFIG_XEN
++ if (!is_initial_xendomain())
++ return xen_read_persistent_clock();
++#endif
+ spin_lock_irqsave(&rtc_lock, flags);
+ retval = get_wallclock();
+ spin_unlock_irqrestore(&rtc_lock, flags);
+@@ -190,6 +194,10 @@ unsigned long read_persistent_clock(void)
+
+ int update_persistent_clock(struct timespec now)
+ {
++#ifdef CONFIG_XEN
++ if (xen_update_persistent_clock() < 0 || xen_independent_wallclock())
++ return 0;
++#endif
+ return set_rtc_mmss(now.tv_sec);
+ }
+
+diff --git a/arch/x86/kernel/setup-xen.c b/arch/x86/kernel/setup-xen.c
+new file mode 100644
+index 0000000..22cdc5d
+--- /dev/null
++++ b/arch/x86/kernel/setup-xen.c
+@@ -0,0 +1,141 @@
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/bootmem.h>
++#include <linux/percpu.h>
++#include <asm/smp.h>
++#include <asm/percpu.h>
++#include <asm/sections.h>
++#include <asm/processor.h>
++#include <asm/setup.h>
++#include <asm/topology.h>
++#include <asm/mpspec.h>
++#include <asm/apicdef.h>
++
++#ifdef CONFIG_X86_LOCAL_APIC
++unsigned int num_processors;
++unsigned disabled_cpus __cpuinitdata;
++/* Processor that is doing the boot up */
++unsigned int boot_cpu_physical_apicid = -1U;
++EXPORT_SYMBOL(boot_cpu_physical_apicid);
++
++DEFINE_PER_CPU(u16, x86_cpu_to_apicid) = BAD_APICID;
++EXPORT_PER_CPU_SYMBOL(x86_cpu_to_apicid);
++
++/* Bitmask of physically existing CPUs */
++physid_mask_t phys_cpu_present_map;
++#endif
++
++#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_X86_SMP)
++/*
++ * Copy data used in early init routines from the initial arrays to the
++ * per cpu data areas. These arrays then become expendable and the
++ * *_early_ptr's are zeroed indicating that the static arrays are gone.
++ */
++static void __init setup_per_cpu_maps(void)
++{
++#ifndef CONFIG_XEN
++ int cpu;
++
++ for_each_possible_cpu(cpu) {
++ per_cpu(x86_cpu_to_apicid, cpu) = x86_cpu_to_apicid_init[cpu];
++ per_cpu(x86_bios_cpu_apicid, cpu) =
++ x86_bios_cpu_apicid_init[cpu];
++#ifdef CONFIG_NUMA
++ per_cpu(x86_cpu_to_node_map, cpu) =
++ x86_cpu_to_node_map_init[cpu];
++#endif
++ }
++
++ /* indicate the early static arrays will soon be gone */
++ x86_cpu_to_apicid_early_ptr = NULL;
++ x86_bios_cpu_apicid_early_ptr = NULL;
++#ifdef CONFIG_NUMA
++ x86_cpu_to_node_map_early_ptr = NULL;
++#endif
++#endif
++}
++
++#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP
++cpumask_t *cpumask_of_cpu_map __read_mostly;
++EXPORT_SYMBOL(cpumask_of_cpu_map);
++
++/* requires nr_cpu_ids to be initialized */
++static void __init setup_cpumask_of_cpu(void)
++{
++ int i;
++
++ /* alloc_bootmem zeroes memory */
++ cpumask_of_cpu_map = alloc_bootmem_low(sizeof(cpumask_t) * nr_cpu_ids);
++ for (i = 0; i < nr_cpu_ids; i++)
++ cpu_set(i, cpumask_of_cpu_map[i]);
++}
++#else
++static inline void setup_cpumask_of_cpu(void) { }
++#endif
++
++#ifdef CONFIG_X86_32
++/*
++ * Great future not-so-futuristic plan: make i386 and x86_64 do it
++ * the same way
++ */
++unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
++EXPORT_SYMBOL(__per_cpu_offset);
++#endif
++
++/*
++ * 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, highest_cpu = 0;
++ unsigned long size;
++
++#ifdef CONFIG_HOTPLUG_CPU
++ prefill_possible_map();
++#endif
++
++ /* Copy section for each CPU (we discard the original) */
++ size = PERCPU_ENOUGH_ROOM;
++ printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n",
++ size);
++
++ for_each_possible_cpu(i) {
++ char *ptr;
++#ifndef CONFIG_NEED_MULTIPLE_NODES
++ ptr = alloc_bootmem_pages(size);
++#else
++ int node = early_cpu_to_node(i);
++ if (!node_online(node) || !NODE_DATA(node)) {
++ ptr = alloc_bootmem_pages(size);
++ printk(KERN_INFO
++ "cpu %d has no node or node-local memory\n", i);
++ }
++ else
++ ptr = alloc_bootmem_pages_node(NODE_DATA(node), size);
++#endif
++ if (!ptr)
++ panic("Cannot allocate cpu data for CPU %d\n", i);
++#ifdef CONFIG_X86_64
++ cpu_pda(i)->data_offset = ptr - __per_cpu_start;
++#else
++ __per_cpu_offset[i] = ptr - __per_cpu_start;
++#endif
++ memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
++
++ highest_cpu = i;
++ }
++
++ nr_cpu_ids = highest_cpu + 1;
++ printk(KERN_DEBUG "NR_CPUS: %d, nr_cpu_ids: %d\n", NR_CPUS, nr_cpu_ids);
++
++ /* Setup percpu data maps */
++ setup_per_cpu_maps();
++
++ /* Setup cpumask_of_cpu map */
++ setup_cpumask_of_cpu();
++}
++
++#endif
+diff --git a/arch/x86/kernel/setup64-xen.c b/arch/x86/kernel/setup64-xen.c
+new file mode 100644
+index 0000000..589ccbd
+--- /dev/null
++++ b/arch/x86/kernel/setup64-xen.c
+@@ -0,0 +1,368 @@
++/*
++ * 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 <linux/kgdb.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>
++#include <asm/setup.h>
++#include <asm/genapic.h>
++#ifdef CONFIG_XEN
++#include <asm/hypervisor.h>
++#endif
++
++#ifndef CONFIG_DEBUG_BOOT_PARAMS
++struct boot_params __initdata boot_params;
++#else
++struct boot_params boot_params;
++#endif
++
++cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE;
++
++struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly;
++EXPORT_SYMBOL(_cpu_pda);
++struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned;
++
++#ifndef CONFIG_X86_NO_IDT
++struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table };
++#endif
++
++char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned")));
++
++unsigned long __supported_pte_mask __read_mostly = ~0UL;
++EXPORT_SYMBOL(__supported_pte_mask);
++
++static int do_not_nx __cpuinitdata = 0;
++
++/* noexec=on|off
++Control non executable mappings for 64bit processes.
++
++on Enable(default)
++off Disable
++*/
++static int __init nonx_setup(char *str)
++{
++ 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);
++
++int force_personality32 = 0;
++
++/* noexec32=on|off
++Control non executable heap for 32bit processes.
++To control the stack too use noexec=off
++
++on PROT_READ does not imply PROT_EXEC for 32bit processes (default)
++off PROT_READ implies PROT_EXEC
++*/
++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_XEN
++static void __init_refok switch_pt(int cpu)
++{
++ if (cpu == 0)
++ xen_init_pt();
++ xen_pt_switch(__pa_symbol(init_level4_pgt));
++ xen_new_user_pt(__pa_symbol(__user_pgd(init_level4_pgt)));
++}
++#define switch_pt() switch_pt(cpu)
++
++static void __cpuinit cpu_gdt_init(const struct desc_ptr *gdt_descr)
++{
++ unsigned long frames[16];
++ unsigned long va;
++ int f;
++
++ for (va = gdt_descr->address, f = 0;
++ va < gdt_descr->address + gdt_descr->size;
++ va += PAGE_SIZE, f++) {
++ frames[f] = virt_to_mfn(va);
++ make_page_readonly(
++ (void *)va, XENFEAT_writable_descriptor_tables);
++ }
++ if (HYPERVISOR_set_gdt(frames, (gdt_descr->size + 1) /
++ sizeof (struct desc_struct)))
++ BUG();
++}
++#else
++static void switch_pt(void)
++{
++ asm volatile("movq %0,%%cr3" :: "r" (__pa_symbol(&init_level4_pgt)));
++}
++
++static void __cpuinit cpu_gdt_init(const struct desc_ptr *gdt_descr)
++{
++ load_gdt(gdt_descr);
++ load_idt(idt_descr);
++}
++#endif
++
++void pda_init(int cpu)
++{
++ struct x8664_pda *pda = cpu_pda(cpu);
++
++ /* Setup up data that may be needed in __get_free_pages early */
++ asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0));
++#ifndef CONFIG_XEN
++ /* Memory clobbers used to order PDA accessed */
++ mb();
++ wrmsrl(MSR_GS_BASE, pda);
++ mb();
++#else
++ if (HYPERVISOR_set_segment_base(SEGBASE_GS_KERNEL,
++ (unsigned long)pda))
++ BUG();
++#endif
++ pda->cpunumber = cpu;
++ pda->irqcount = -1;
++ pda->kernelstack =
++ (unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE;
++ pda->active_mm = &init_mm;
++ pda->mmu_state = 0;
++
++ if (cpu == 0) {
++ /* others are initialized in smpboot.c */
++ pda->pcurrent = &init_task;
++ pda->irqstackptr = boot_cpu_stack;
++ } else {
++ pda->irqstackptr = (char *)
++ __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER);
++ if (!pda->irqstackptr)
++ panic("cannot allocate irqstack for cpu %d", cpu);
++ }
++
++ switch_pt();
++
++ pda->irqstackptr += IRQSTACKSIZE-64;
++}
++
++#ifndef CONFIG_X86_NO_TSS
++char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]
++__attribute__((section(".bss.page_aligned")));
++#endif
++
++extern asmlinkage void ignore_sysret(void);
++
++/* 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);
++ wrmsrl(MSR_CSTAR, ignore_sysret);
++
++ /* Flags to clear on syscall */
++ wrmsrl(MSR_SYSCALL_MASK,
++ X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL);
++#endif
++#ifdef CONFIG_IA32_EMULATION
++ syscall32_cpu_init ();
++#else
++ {
++ static const struct callback_register cstar = {
++ .type = CALLBACKTYPE_syscall32,
++ .address = (unsigned long)ignore_sysret
++ };
++ if (HYPERVISOR_callback_op(CALLBACKOP_register, &cstar))
++ printk(KERN_WARN "Unable to register CSTAR callback\n");
++ }
++#endif
++}
++
++void __cpuinit check_efer(void)
++{
++ unsigned long efer;
++
++ rdmsrl(MSR_EFER, efer);
++ if (!(efer & EFER_NX) || do_not_nx) {
++ __supported_pte_mask &= ~_PAGE_NX;
++ }
++}
++
++unsigned long kernel_eflags;
++
++#ifndef CONFIG_X86_NO_TSS
++/*
++ * Copies of the original ist values from the tss are only accessed during
++ * debugging, no special alignment required.
++ */
++DEFINE_PER_CPU(struct orig_ist, orig_ist);
++#endif
++
++/*
++ * cpu_init() initializes state that is per-CPU. Some data is already
++ * initialized (naturally) in the bootstrap process, such as the GDT
++ * and IDT. We reload them nevertheless, this function acts as a
++ * 'CPU state barrier', nothing should get across.
++ * A lot of state is already set up in PDA init.
++ */
++void __cpuinit cpu_init (void)
++{
++ int cpu = stack_smp_processor_id();
++#ifndef CONFIG_X86_NO_TSS
++ struct tss_struct *t = &per_cpu(init_tss, cpu);
++ struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu);
++ unsigned long v;
++ char *estacks = NULL;
++ unsigned i;
++#endif
++ struct task_struct *me;
++
++ /* CPU 0 is initialised in head64.c */
++ if (cpu != 0) {
++ pda_init(cpu);
++ }
++#ifndef CONFIG_X86_NO_TSS
++ else
++ estacks = boot_exception_stacks;
++#endif
++
++ me = current;
++
++ if (cpu_test_and_set(cpu, cpu_initialized))
++ panic("CPU#%d already initialized!\n", cpu);
++
++ printk("Initializing CPU#%d\n", cpu);
++
++ clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
++
++ /*
++ * Initialize the per-CPU GDT with the boot GDT,
++ * and set up the GDT descriptor:
++ */
++#ifndef CONFIG_XEN
++ if (cpu)
++ memcpy(get_cpu_gdt_table(cpu), cpu_gdt_table, GDT_SIZE);
++#endif
++
++ cpu_gdt_descr[cpu].size = GDT_SIZE;
++ cpu_gdt_init(&cpu_gdt_descr[cpu]);
++
++ memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8);
++ syscall_init();
++
++ wrmsrl(MSR_FS_BASE, 0);
++ wrmsrl(MSR_KERNEL_GS_BASE, 0);
++ barrier();
++
++ check_efer();
++
++#ifndef CONFIG_X86_NO_TSS
++ /*
++ * set up and load the per-CPU TSS
++ */
++ for (v = 0; v < N_EXCEPTION_STACKS; v++) {
++ 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->x86_tss.ist[v] = (unsigned long)estacks;
++ }
++
++ t->x86_tss.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);
++
++#ifdef CONFIG_KGDB
++ /*
++ * If the kgdb is connected no debug regs should be altered. This
++ * is only applicable when KGDB and a KGDB I/O module are built
++ * into the kernel and you are using early debugging with
++ * kgdbwait. KGDB will control the kernel HW breakpoint registers.
++ */
++ if (kgdb_connected && arch_kgdb_ops.correct_hw_break)
++ arch_kgdb_ops.correct_hw_break();
++ else {
++#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);
++#ifdef CONFIG_KGDB
++ /* If the kgdb is connected no debug regs should be altered. */
++ }
++#endif
++
++ fpu_init();
++
++ raw_local_save_flags(kernel_eflags);
++
++ if (is_uv_system())
++ uv_cpu_init();
++}
+diff --git a/arch/x86/kernel/setup_32-xen.c b/arch/x86/kernel/setup_32-xen.c
+new file mode 100644
+index 0000000..e7a488a
+--- /dev/null
++++ b/arch/x86/kernel/setup_32-xen.c
+@@ -0,0 +1,1154 @@
++/*
++ * 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
++ */
++
++#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/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/iscsi_ibft.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>
++#include <linux/pci.h>
++#include <linux/init_ohci1394_dma.h>
++#include <linux/kvm_para.h>
++
++#include <video/edid.h>
++
++#include <asm/mtrr.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/hypervisor.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/memory.h>
++#include <xen/features.h>
++#include <xen/firmware.h>
++#include <xen/xencons.h>
++#include <setup_arch.h>
++#include <asm/bios_ebda.h>
++#include <asm/cacheflush.h>
++#include <asm/processor.h>
++
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#endif
++
++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 */
++};
++
++extern char hypercall_page[PAGE_SIZE];
++EXPORT_SYMBOL(hypercall_page);
++
++/*
++ * Machine setup..
++ */
++static struct resource data_resource = {
++ .name = "Kernel data",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
++};
++
++static struct resource code_resource = {
++ .name = "Kernel code",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
++};
++
++static struct resource bss_resource = {
++ .name = "Kernel bss",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_BUSY | IORESOURCE_MEM
++};
++
++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 = 0x0060,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO
++}, {
++ .name = "keyboard",
++ .start = 0x0064,
++ .end = 0x0064,
++ .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
++} };
++
++/* 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);
++
++unsigned int def_to_bigsmp;
++
++#ifndef CONFIG_X86_PAE
++unsigned long mmu_cr4_features;
++#else
++unsigned long mmu_cr4_features = X86_CR4_PAE;
++#endif
++
++/* for MCA, but anyone else can use it if they want */
++unsigned int machine_id;
++unsigned int machine_submodel_id;
++unsigned int BIOS_revision;
++
++/* Boot loader ID as an integer, for the benefit of proc_dointvec */
++int bootloader_type;
++
++/* user-defined highmem size */
++static unsigned int highmem_pages = -1;
++
++/*
++ * Setup options
++ */
++struct screen_info screen_info;
++EXPORT_SYMBOL(screen_info);
++struct apm_info apm_info;
++EXPORT_SYMBOL(apm_info);
++struct edid_info edid_info;
++EXPORT_SYMBOL_GPL(edid_info);
++#ifndef CONFIG_XEN
++#define copy_edid() (edid_info = boot_params.edid_info)
++#endif
++struct ist_info ist_info;
++#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \
++ defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE)
++EXPORT_SYMBOL(ist_info);
++#endif
++
++extern void early_cpu_init(void);
++extern int root_mountflags;
++
++unsigned long saved_video_mode;
++
++#define RAMDISK_IMAGE_START_MASK 0x07FF
++#define RAMDISK_PROMPT_FLAG 0x8000
++#define RAMDISK_LOAD_FLAG 0x4000
++
++static char __initdata command_line[COMMAND_LINE_SIZE];
++
++#ifndef CONFIG_DEBUG_BOOT_PARAMS
++struct boot_params __initdata boot_params;
++#else
++struct boot_params boot_params;
++#endif
++
++/*
++ * Point at the empty zero page to start with. We map the real shared_info
++ * page as soon as fixmap is up and running.
++ */
++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);
++
++/* Raw start-of-day parameters from the hypervisor. */
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++struct edd edd;
++#ifdef CONFIG_EDD_MODULE
++EXPORT_SYMBOL(edd);
++#endif
++#ifndef CONFIG_XEN
++/**
++ * copy_edd() - Copy the BIOS EDD information
++ * from boot_params into a safe place.
++ *
++ */
++static inline void copy_edd(void)
++{
++ memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer,
++ sizeof(edd.mbr_signature));
++ memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info));
++ edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries;
++ edd.edd_info_nr = boot_params.eddbuf_entries;
++}
++#endif
++#else
++static inline void copy_edd(void)
++{
++}
++#endif
++
++int __initdata user_defined_memmap;
++
++/*
++ * "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 int __init parse_mem(char *arg)
++{
++ if (!arg)
++ return -EINVAL;
++
++ if (strcmp(arg, "nopentium") == 0) {
++ setup_clear_cpu_cap(X86_FEATURE_PSE);
++ } 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(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 */
++
++/*
++ * 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 int __init parse_highmem(char *arg)
++{
++ if (!arg)
++ return -EINVAL;
++
++ highmem_pages = memparse(arg, &arg) >> PAGE_SHIFT;
++ return 0;
++}
++early_param("highmem", parse_highmem);
++
++/*
++ * 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;
++
++ __VMALLOC_RESERVE = memparse(arg, &arg);
++ return 0;
++}
++early_param("vmalloc", parse_vmalloc);
++
++#ifndef CONFIG_XEN
++/*
++ * 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;
++
++ if (!arg)
++ return -EINVAL;
++
++ address = memparse(arg, &arg);
++ reserve_top_address(address);
++ return 0;
++}
++early_param("reservetop", parse_reservetop);
++#endif
++
++/*
++ * Determine low and high memory ranges:
++ */
++unsigned long __init find_max_low_pfn(void)
++{
++ unsigned long max_low_pfn;
++
++ 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 HIGHMEM64G enabled kernel.\n");
++ else
++ printk(KERN_WARNING "Use a HIGHMEM enabled kernel.\n");
++ max_pfn = MAXMEM_PFN;
++#else /* !CONFIG_HIGHMEM */
++#ifndef CONFIG_HIGHMEM64G
++ 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 HIGHMEM64G enabled kernel.\n");
++ }
++#endif /* !CONFIG_HIGHMEM64G */
++#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;
++ }
++#else
++ if (highmem_pages)
++ printk(KERN_ERR "ignoring highmem size on non-highmem kernel!\n");
++#endif
++ }
++ return max_low_pfn;
++}
++
++#ifndef CONFIG_XEN
++#define BIOS_LOWMEM_KILOBYTES 0x413
++
++/*
++ * The BIOS places the EBDA/XBDA at the top of conventional
++ * memory, and usually decreases the reported amount of
++ * conventional memory (int 0x12) too. This also contains a
++ * workaround for Dell systems that neglect to reserve EBDA.
++ * The same workaround also avoids a problem with the AMD768MPX
++ * 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.
++ */
++static void __init reserve_ebda_region(void)
++{
++ unsigned int lowmem, ebda_addr;
++
++ /* To determine the position of the EBDA and the */
++ /* end of conventional memory, we need to look at */
++ /* the BIOS data area. In a paravirtual environment */
++ /* that area is absent. We'll just have to assume */
++ /* that the paravirt case can handle memory setup */
++ /* correctly, without our help. */
++ if (paravirt_enabled())
++ return;
++
++ /* end of low (conventional) memory */
++ lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES);
++ lowmem <<= 10;
++
++ /* start of EBDA area */
++ ebda_addr = get_bios_ebda();
++
++ /* Fixup: bios puts an EBDA in the top 64K segment */
++ /* of conventional memory, but does not adjust lowmem. */
++ if ((lowmem - ebda_addr) <= 0x10000)
++ lowmem = ebda_addr;
++
++ /* Fixup: bios does not report an EBDA at all. */
++ /* Some old Dells seem to need 4k anyhow (bugzilla 2990) */
++ if ((ebda_addr == 0) && (lowmem >= 0x9f000))
++ lowmem = 0x9f000;
++
++ /* Paranoia: should never happen, but... */
++ if ((lowmem == 0) || (lowmem >= 0x100000))
++ lowmem = 0x9f000;
++
++ /* reserve all memory between lowmem and the 1MB mark */
++ reserve_bootmem(lowmem, 0x100000 - lowmem, BOOTMEM_DEFAULT);
++}
++#endif
++
++#ifndef CONFIG_NEED_MULTIPLE_NODES
++static void __init setup_bootmem_allocator(void);
++static unsigned long __init setup_memory(void)
++{
++ /*
++ * 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;
++
++ 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;
++ }
++ 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));
++
++ setup_bootmem_allocator();
++
++ return max_low_pfn;
++}
++
++static void __init zone_sizes_init(void)
++{
++ unsigned long max_zone_pfns[MAX_NR_ZONES];
++ memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
++ max_zone_pfns[ZONE_DMA] =
++ virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
++ 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
++
++ 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 */
++
++static inline unsigned long long get_total_mem(void)
++{
++ unsigned long long total;
++
++ total = max_low_pfn - min_low_pfn;
++#ifdef CONFIG_HIGHMEM
++ total += highend_pfn - highstart_pfn;
++#endif
++
++ return total << PAGE_SHIFT;
++}
++
++#ifdef CONFIG_KEXEC
++#ifndef CONFIG_XEN
++static void __init reserve_crashkernel(void)
++{
++ unsigned long long total_mem;
++ unsigned long long crash_size, crash_base;
++ int ret;
++
++ total_mem = get_total_mem();
++
++ ret = parse_crashkernel(boot_command_line, total_mem,
++ &crash_size, &crash_base);
++ if (ret == 0 && crash_size > 0) {
++ if (crash_base > 0) {
++ printk(KERN_INFO "Reserving %ldMB of memory at %ldMB "
++ "for crashkernel (System RAM: %ldMB)\n",
++ (unsigned long)(crash_size >> 20),
++ (unsigned long)(crash_base >> 20),
++ (unsigned long)(total_mem >> 20));
++
++ if (reserve_bootmem(crash_base, crash_size,
++ BOOTMEM_EXCLUSIVE) < 0) {
++ printk(KERN_INFO "crashkernel reservation "
++ "failed - memory is in use\n");
++ return;
++ }
++
++ crashk_res.start = crash_base;
++ crashk_res.end = crash_base + crash_size - 1;
++ } else
++ printk(KERN_INFO "crashkernel reservation failed - "
++ "you have to specify a base address\n");
++ }
++}
++#else
++#define reserve_crashkernel xen_machine_kexec_setup_resources
++#endif
++#else
++static inline void __init reserve_crashkernel(void)
++{}
++#endif
++
++#ifdef CONFIG_BLK_DEV_INITRD
++
++static bool do_relocate_initrd = false;
++
++static void __init reserve_initrd(void)
++{
++ unsigned long ramdisk_image = __pa(xen_start_info->mod_start);
++ unsigned long ramdisk_size = xen_start_info->mod_len;
++ unsigned long ramdisk_end = ramdisk_image + ramdisk_size;
++ unsigned long end_of_lowmem = max_low_pfn << PAGE_SHIFT;
++ unsigned long ramdisk_here;
++
++ initrd_start = 0;
++
++ if (!xen_start_info->mod_start || !ramdisk_size)
++ return; /* No initrd provided by bootloader */
++
++ if (ramdisk_end < ramdisk_image) {
++ printk(KERN_ERR "initrd wraps around end of memory, "
++ "disabling initrd\n");
++ return;
++ }
++ if (ramdisk_size >= end_of_lowmem/2) {
++ printk(KERN_ERR "initrd too large to handle, "
++ "disabling initrd\n");
++ return;
++ }
++ if (ramdisk_end <= end_of_lowmem) {
++ /* All in lowmem, easy case */
++ reserve_bootmem(ramdisk_image, ramdisk_size, BOOTMEM_DEFAULT);
++ initrd_start = ramdisk_image + PAGE_OFFSET;
++ initrd_end = initrd_start+ramdisk_size;
++ return;
++ }
++
++ /* We need to move the initrd down into lowmem */
++ ramdisk_here = (end_of_lowmem - ramdisk_size) & PAGE_MASK;
++
++ /* Note: this includes all the lowmem currently occupied by
++ the initrd, we rely on that fact to keep the data intact. */
++ reserve_bootmem(ramdisk_here, ramdisk_size, BOOTMEM_DEFAULT);
++ initrd_start = ramdisk_here + PAGE_OFFSET;
++ initrd_end = initrd_start + ramdisk_size;
++
++ do_relocate_initrd = true;
++}
++
++#define MAX_MAP_CHUNK (NR_FIX_BTMAPS << PAGE_SHIFT)
++
++static void __init relocate_initrd(void)
++{
++ unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
++ unsigned long ramdisk_size = boot_params.hdr.ramdisk_size;
++ unsigned long end_of_lowmem = max_low_pfn << PAGE_SHIFT;
++ unsigned long ramdisk_here;
++ unsigned long slop, clen, mapaddr;
++ char *p, *q;
++
++ if (!do_relocate_initrd)
++ return;
++
++ ramdisk_here = initrd_start - PAGE_OFFSET;
++
++ q = (char *)initrd_start;
++
++ /* Copy any lowmem portion of the initrd */
++ if (ramdisk_image < end_of_lowmem) {
++ clen = end_of_lowmem - ramdisk_image;
++ p = (char *)__va(ramdisk_image);
++ memcpy(q, p, clen);
++ q += clen;
++ ramdisk_image += clen;
++ ramdisk_size -= clen;
++ }
++
++ /* Copy the highmem portion of the initrd */
++ while (ramdisk_size) {
++ slop = ramdisk_image & ~PAGE_MASK;
++ clen = ramdisk_size;
++ if (clen > MAX_MAP_CHUNK-slop)
++ clen = MAX_MAP_CHUNK-slop;
++ mapaddr = ramdisk_image & PAGE_MASK;
++ p = early_ioremap(mapaddr, clen+slop);
++ memcpy(q, p+slop, clen);
++ early_iounmap(p, clen+slop);
++ q += clen;
++ ramdisk_image += clen;
++ ramdisk_size -= clen;
++ }
++}
++
++#endif /* CONFIG_BLK_DEV_INITRD */
++
++void __init setup_bootmem_allocator(void)
++{
++ 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);
++
++ /*
++ * 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.
++ */
++ reserve_bootmem(__pa_symbol(_text), (PFN_PHYS(min_low_pfn) +
++ bootmap_size + PAGE_SIZE-1) - __pa_symbol(_text),
++ BOOTMEM_DEFAULT);
++
++#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, BOOTMEM_DEFAULT);
++
++ /* reserve EBDA region */
++ reserve_ebda_region();
++
++#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, BOOTMEM_DEFAULT);
++#endif
++#ifdef CONFIG_ACPI_SLEEP
++ /*
++ * Reserve low memory region for sleep support.
++ */
++ acpi_reserve_bootmem();
++#endif
++#endif /* !CONFIG_XEN */
++
++#ifdef CONFIG_BLK_DEV_INITRD
++ reserve_initrd();
++#endif
++ numa_kva_reserve();
++ reserve_crashkernel();
++
++ reserve_ibft_region();
++}
++
++/*
++ * 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.
++ */
++static void __init remapped_pgdat_init(void)
++{
++ int nid;
++
++ for_each_online_node(nid) {
++ if (nid != 0)
++ memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
++ }
++}
++
++#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();
++}
++
++#ifdef CONFIG_NUMA
++/*
++ * In the golden day, when everything among i386 and x86_64 will be
++ * integrated, this will not live here
++ */
++void *x86_cpu_to_node_map_early_ptr;
++int x86_cpu_to_node_map_init[NR_CPUS] = {
++ [0 ... NR_CPUS-1] = NUMA_NO_NODE
++};
++DEFINE_PER_CPU(int, x86_cpu_to_node_map) = NUMA_NO_NODE;
++#endif
++
++/*
++ * 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 __init setup_arch(char **cmdline_p)
++{
++ int i, j, k, fpp;
++ struct physdev_set_iopl set_iopl;
++ unsigned long max_low_pfn;
++ unsigned long p2m_pages;
++
++ /* 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);
++
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_4gb_segments));
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables));
++
++ memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data));
++ pre_setup_arch_hook();
++ early_cpu_init();
++ early_ioremap_init();
++#ifdef CONFIG_SMP
++ prefill_possible_map();
++#endif
++
++#ifdef CONFIG_EFI
++ if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature,
++ "EL32", 4))
++ efi_enabled = 1;
++#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);
++ screen_info = boot_params.screen_info;
++ copy_edid();
++ apm_info.bios = boot_params.apm_bios_info;
++ ist_info = boot_params.ist_info;
++ saved_video_mode = boot_params.hdr.vid_mode;
++ if( boot_params.sys_desc_table.length != 0 ) {
++ set_mca_bus(boot_params.sys_desc_table.table[3] & 0x2);
++ machine_id = boot_params.sys_desc_table.table[0];
++ machine_submodel_id = boot_params.sys_desc_table.table[1];
++ BIOS_revision = boot_params.sys_desc_table.table[2];
++ }
++ bootloader_type = boot_params.hdr.type_of_loader;
++
++ if (is_initial_xendomain()) {
++ const struct dom0_vga_console_info *info =
++ (void *)((char *)xen_start_info +
++ xen_start_info->console.dom0.info_off);
++
++ dom0_init_screen_info(info,
++ xen_start_info->console.dom0.info_size);
++ xen_start_info->console.domU.mfn = 0;
++ xen_start_info->console.domU.evtchn = 0;
++ } else
++ screen_info.orig_video_isVGA = 0;
++
++#ifdef CONFIG_BLK_DEV_RAM
++ rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK;
++ rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0);
++ rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0);
++#endif
++
++ ARCH_SETUP
++
++ printk(KERN_INFO "BIOS-provided physical RAM map:\n");
++ print_memory_map(memory_setup());
++
++ copy_edd();
++
++ if (!boot_params.hdr.root_flags)
++ 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;
++
++ 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;
++ bss_resource.start = virt_to_phys(&__bss_start);
++ bss_resource.end = virt_to_phys(&__bss_stop)-1;
++
++ if ((i = MAX_GUEST_CMDLINE) > COMMAND_LINE_SIZE)
++ i = COMMAND_LINE_SIZE;
++ memcpy(boot_command_line, xen_start_info->cmd_line, i);
++ boot_command_line[i - 1] = '\0';
++ parse_early_param();
++
++ if (user_defined_memmap) {
++ printk(KERN_INFO "user-defined physical RAM map:\n");
++ print_memory_map("user");
++ }
++
++ strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
++ *cmdline_p = command_line;
++
++ if (efi_enabled)
++ efi_init();
++
++ /* update e820 for memory not covered by WB MTRRs */
++ propagate_e820_map();
++ mtrr_bp_init();
++#ifndef CONFIG_XEN
++ if (mtrr_trim_uncached_memory(max_pfn))
++ propagate_e820_map();
++#endif
++
++ max_low_pfn = setup_memory();
++
++#ifdef CONFIG_KVM_CLOCK
++ kvmclock_init();
++#endif
++
++#ifdef CONFIG_VMI
++ /*
++ * Must be after max_low_pfn is determined, and before kernel
++ * pagetables are setup.
++ */
++ vmi_init();
++#endif
++ kvm_guest_init();
++
++ /*
++ * NOTE: before this point _nobody_ is allowed to allocate
++ * any memory using the bootmem allocator. Although the
++ * allocator 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();
++
++ /*
++ * NOTE: On x86-32, only from this point on, fixmaps are ready for use.
++ */
++
++#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT
++ if (init_ohci1394_dma_early)
++ init_ohci1394_dma_on_all_controllers();
++#endif
++
++ remapped_pgdat_init();
++ sparse_init();
++ zone_sizes_init();
++
++#ifdef CONFIG_X86_FIND_SMP_CONFIG
++ /*
++ * Find and reserve possible boot-time SMP configuration:
++ */
++ find_smp_config();
++#endif
++
++ p2m_pages = max_pfn;
++ if (xen_start_info->nr_pages > max_pfn) {
++ /*
++ * the max_pfn was shrunk (probably by mem= or highmem=
++ * kernel parameter); shrink reservation with the HV
++ */
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ unsigned int difference;
++ int ret;
++
++ difference = xen_start_info->nr_pages - max_pfn;
++
++ set_xen_guest_handle(reservation.extent_start,
++ ((unsigned long *)xen_start_info->mfn_list) + max_pfn);
++ reservation.nr_extents = difference;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ BUG_ON (ret != difference);
++ }
++ else if (max_pfn > xen_start_info->nr_pages)
++ p2m_pages = xen_start_info->nr_pages;
++
++ /* 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,
++ p2m_pages * sizeof(unsigned long));
++ free_bootmem(
++ __pa(xen_start_info->mfn_list),
++ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
++ sizeof(unsigned long))));
++
++ /*
++ * Initialise the list of the frames that specify the list of
++ * frames that make up the p2m table. Used by save/restore
++ */
++ pfn_to_mfn_frame_list_list = alloc_bootmem_low_pages(PAGE_SIZE);
++
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i=0, j=0, k=-1; i< max_pfn; i+=fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ 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;
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
++ }
++
++ /* Mark all ISA DMA channels in-use - using them wouldn't work. */
++ for (i = 0; i < MAX_DMA_CHANNELS; ++i)
++ if (i != 4 && request_dma(i, "xen") != 0)
++ BUG();
++
++ /*
++ * NOTE: at this point the bootmem allocator is fully available.
++ */
++
++#ifdef CONFIG_BLK_DEV_INITRD
++ relocate_initrd();
++#endif
++
++ paravirt_post_allocator_init();
++
++ if (is_initial_xendomain())
++ dmi_scan_machine();
++
++ io_delay_init();
++
++#if defined(CONFIG_X86_SMP) && !defined(CONFIG_XEN)
++ /*
++ * setup to use the early static init tables during kernel startup
++ * X86_SMP will exclude sub-arches that don't deal well with it.
++ */
++ x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init;
++ x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init;
++#ifdef CONFIG_NUMA
++ x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init;
++#endif
++#endif
++
++#ifdef CONFIG_X86_GENERICARCH
++ generic_apic_probe();
++#endif
++
++ set_iopl.iopl = 1;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++
++#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
++
++#ifndef CONFIG_XEN
++ early_quirks();
++#endif
++
++#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
++
++ e820_register_memory();
++ e820_mark_nosave_regions();
++
++ 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
++ }
++}
++
++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;
++}
++
++/*
++ * 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;
++
++ /* Nothing to do if not running in dom0. */
++ if (!is_initial_xendomain())
++ return 0;
++
++ printk(KERN_INFO "Setting up standard PCI resources\n");
++ init_iomem_resources(&code_resource, &data_resource, &bss_resource);
++
++ request_resource(&iomem_resource, &video_ram_resource);
++
++ /* 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]);
++ return 0;
++}
++
++subsys_initcall(request_standard_resources);
+diff --git a/arch/x86/kernel/setup_64-xen.c b/arch/x86/kernel/setup_64-xen.c
+new file mode 100644
+index 0000000..a815ad9
+--- /dev/null
++++ b/arch/x86/kernel/setup_64-xen.c
+@@ -0,0 +1,1434 @@
++/*
++ * Copyright (C) 1995 Linus Torvalds
++ */
++
++/*
++ * This file handles the architecture-dependent parts of initialization
++ */
++
++#include <linux/errno.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/stddef.h>
++#include <linux/unistd.h>
++#include <linux/ptrace.h>
++#include <linux/slab.h>
++#include <linux/user.h>
++#include <linux/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 <asm/pci-direct.h>
++#include <linux/efi.h>
++#include <linux/acpi.h>
++#include <linux/kallsyms.h>
++#include <linux/edd.h>
++#include <linux/iscsi_ibft.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 <linux/sort.h>
++#include <linux/uaccess.h>
++#include <linux/init_ohci1394_dma.h>
++#include <linux/kvm_para.h>
++
++#include <asm/mtrr.h>
++#include <asm/uaccess.h>
++#include <asm/system.h>
++#include <asm/vsyscall.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/gart.h>
++#include <asm/mpspec.h>
++#include <asm/mmu_context.h>
++#include <asm/proto.h>
++#include <asm/setup.h>
++#include <asm/numa.h>
++#include <asm/sections.h>
++#include <asm/dmi.h>
++#include <asm/cacheflush.h>
++#include <asm/mce.h>
++#include <asm/ds.h>
++#include <asm/topology.h>
++#include <asm/pat.h>
++
++#include <mach_apic.h>
++#ifdef CONFIG_XEN
++#include <linux/percpu.h>
++#include <xen/interface/physdev.h>
++#include "setup_arch_pre.h"
++#include <asm/hypervisor.h>
++#include <xen/interface/nmi.h>
++#include <xen/features.h>
++#include <xen/firmware.h>
++#include <xen/xencons.h>
++#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
++#define PFN_PHYS(x) ((x) << PAGE_SHIFT)
++#include <asm/mach-xen/setup_arch_post.h>
++#include <xen/interface/memory.h>
++
++#ifdef CONFIG_XEN
++#include <xen/interface/kexec.h>
++#endif
++
++extern unsigned long start_pfn;
++extern struct edid_info edid_info;
++
++shared_info_t *HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++EXPORT_SYMBOL(HYPERVISOR_shared_info);
++
++extern char hypercall_page[PAGE_SIZE];
++EXPORT_SYMBOL(hypercall_page);
++
++static int xen_panic_event(struct notifier_block *, unsigned long, void *);
++static struct notifier_block xen_panic_block = {
++ xen_panic_event, NULL, 0 /* try to go last */
++};
++
++unsigned long *phys_to_machine_mapping;
++unsigned long *pfn_to_mfn_frame_list_list, *pfn_to_mfn_frame_list[512];
++
++EXPORT_SYMBOL(phys_to_machine_mapping);
++
++DEFINE_PER_CPU(multicall_entry_t, multicall_list[8]);
++DEFINE_PER_CPU(int, nr_multicall_ents);
++
++/* Raw start-of-day parameters from the hypervisor. */
++start_info_t *xen_start_info;
++EXPORT_SYMBOL(xen_start_info);
++#endif
++
++/*
++ * Machine setup..
++ */
++
++struct cpuinfo_x86 boot_cpu_data __read_mostly;
++EXPORT_SYMBOL(boot_cpu_data);
++
++__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata;
++
++unsigned long mmu_cr4_features;
++
++/* Boot loader ID as an integer, for the benefit of proc_dointvec */
++int bootloader_type;
++
++unsigned long saved_video_mode;
++
++int force_mwait __cpuinitdata;
++
++/*
++ * Early DMI memory
++ */
++int dmi_alloc_index;
++char dmi_alloc_data[DMI_MAX_DATA];
++
++/*
++ * Setup options
++ */
++struct screen_info screen_info;
++EXPORT_SYMBOL(screen_info);
++struct sys_desc_table_struct {
++ unsigned short length;
++ unsigned char table[0];
++};
++
++struct edid_info edid_info;
++EXPORT_SYMBOL_GPL(edid_info);
++
++extern int root_mountflags;
++
++char __initdata command_line[COMMAND_LINE_SIZE];
++
++static struct resource standard_io_resources[] = {
++ { .name = "dma1", .start = 0x00, .end = 0x1f,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "pic1", .start = 0x20, .end = 0x21,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "timer0", .start = 0x40, .end = 0x43,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "timer1", .start = 0x50, .end = 0x53,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "keyboard", .start = 0x60, .end = 0x60,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "keyboard", .start = 0x64, .end = 0x64,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "dma page reg", .start = 0x80, .end = 0x8f,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "pic2", .start = 0xa0, .end = 0xa1,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "dma2", .start = 0xc0, .end = 0xdf,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO },
++ { .name = "fpu", .start = 0xf0, .end = 0xff,
++ .flags = IORESOURCE_BUSY | IORESOURCE_IO }
++};
++
++#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM)
++
++static struct resource data_resource = {
++ .name = "Kernel data",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_RAM,
++};
++static struct resource code_resource = {
++ .name = "Kernel code",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_RAM,
++};
++static struct resource bss_resource = {
++ .name = "Kernel bss",
++ .start = 0,
++ .end = 0,
++ .flags = IORESOURCE_RAM,
++};
++
++static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c);
++
++#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)
++{
++ char *end;
++ if (!arg)
++ return -EINVAL;
++ elfcorehdr_addr = memparse(arg, &end);
++ return end > arg ? 0 : -EINVAL;
++}
++early_param("elfcorehdr", setup_elfcorehdr);
++#endif
++
++#ifndef CONFIG_NUMA
++static void __init
++contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
++{
++ unsigned long bootmap_size, bootmap;
++
++ bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT;
++ bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size,
++ PAGE_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);
++ early_res_to_bootmem(0, xen_start_info->nr_pages<<PAGE_SHIFT);
++#else
++ free_bootmem_with_active_regions(0, end_pfn);
++ early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT);
++#endif
++ reserve_bootmem(bootmap, bootmap_size, BOOTMEM_DEFAULT);
++}
++#endif
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++struct edd edd;
++#ifdef CONFIG_EDD_MODULE
++EXPORT_SYMBOL(edd);
++#endif
++#ifndef CONFIG_XEN
++/**
++ * copy_edd() - Copy the BIOS EDD information
++ * from boot_params into a safe place.
++ *
++ */
++static inline void copy_edd(void)
++{
++ memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer,
++ sizeof(edd.mbr_signature));
++ memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info));
++ edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries;
++ edd.edd_info_nr = boot_params.eddbuf_entries;
++}
++#endif
++#else
++static inline void copy_edd(void)
++{
++}
++#endif
++
++#ifdef CONFIG_KEXEC
++#ifndef CONFIG_XEN
++static void __init reserve_crashkernel(void)
++{
++ unsigned long long total_mem;
++ unsigned long long crash_size, crash_base;
++ int ret;
++
++ total_mem = ((unsigned long long)max_low_pfn - min_low_pfn) << PAGE_SHIFT;
++
++ ret = parse_crashkernel(boot_command_line, total_mem,
++ &crash_size, &crash_base);
++ if (ret == 0 && crash_size) {
++ if (crash_base <= 0) {
++ printk(KERN_INFO "crashkernel reservation failed - "
++ "you have to specify a base address\n");
++ return;
++ }
++
++ if (reserve_bootmem(crash_base, crash_size,
++ BOOTMEM_EXCLUSIVE) < 0) {
++ printk(KERN_INFO "crashkernel reservation failed - "
++ "memory is in use\n");
++ return;
++ }
++
++ printk(KERN_INFO "Reserving %ldMB of memory at %ldMB "
++ "for crashkernel (System RAM: %ldMB)\n",
++ (unsigned long)(crash_size >> 20),
++ (unsigned long)(crash_base >> 20),
++ (unsigned long)(total_mem >> 20));
++ crashk_res.start = crash_base;
++ crashk_res.end = crash_base + crash_size - 1;
++ insert_resource(&iomem_resource, &crashk_res);
++ }
++}
++#else
++#define reserve_crashkernel xen_machine_kexec_setup_resources
++#endif
++#else
++static inline void __init reserve_crashkernel(void)
++{}
++#endif
++
++/* Overridden in paravirt.c if CONFIG_PARAVIRT */
++void __attribute__((weak)) __init memory_setup(void)
++{
++ machine_specific_memory_setup();
++}
++
++static void __init parse_setup_data(void)
++{
++ struct setup_data *data;
++ unsigned long pa_data;
++
++ if (boot_params.hdr.version < 0x0209)
++ return;
++ pa_data = boot_params.hdr.setup_data;
++ while (pa_data) {
++ data = early_ioremap(pa_data, PAGE_SIZE);
++ switch (data->type) {
++ default:
++ break;
++ }
++#ifndef CONFIG_DEBUG_BOOT_PARAMS
++ free_early(pa_data, pa_data+sizeof(*data)+data->len);
++#endif
++ pa_data = data->next;
++ early_iounmap(data, PAGE_SIZE);
++ }
++}
++
++#ifdef CONFIG_PCI_MMCONFIG
++extern void __cpuinit fam10h_check_enable_mmcfg(void);
++extern void __init check_enable_amd_mmconf_dmi(void);
++#else
++void __cpuinit fam10h_check_enable_mmcfg(void)
++{
++}
++void __init check_enable_amd_mmconf_dmi(void)
++{
++}
++#endif
++
++/*
++ * setup_arch - architecture-specific boot-time initializations
++ *
++ * Note: On x86_64, fixmaps are ready for use even before this is called.
++ */
++void __init setup_arch(char **cmdline_p)
++{
++ unsigned i;
++
++#ifdef CONFIG_XEN
++ extern struct e820map machine_e820;
++
++ printk(KERN_INFO "Command line: %s\n", boot_command_line);
++
++ /* Register a call for panic conditions. */
++ atomic_notifier_chain_register(&panic_notifier_list, &xen_panic_block);
++
++ WARN_ON(HYPERVISOR_vm_assist(VMASST_CMD_enable,
++ VMASST_TYPE_writable_pagetables));
++
++ early_ioremap_init();
++
++ ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
++ screen_info = boot_params.screen_info;
++
++ if (is_initial_xendomain()) {
++ const struct dom0_vga_console_info *info =
++ (void *)((char *)xen_start_info +
++ xen_start_info->console.dom0.info_off);
++
++ dom0_init_screen_info(info,
++ xen_start_info->console.dom0.info_size);
++ xen_start_info->console.domU.mfn = 0;
++ xen_start_info->console.domU.evtchn = 0;
++ } else
++ screen_info.orig_video_isVGA = 0;
++
++ copy_edid();
++#else
++ printk(KERN_INFO "Command line: %s\n", boot_command_line);
++
++ ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev);
++ screen_info = boot_params.screen_info;
++ edid_info = boot_params.edid_info;
++#endif /* !CONFIG_XEN */
++ saved_video_mode = boot_params.hdr.vid_mode;
++ bootloader_type = boot_params.hdr.type_of_loader;
++
++#ifdef CONFIG_BLK_DEV_RAM
++ rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK;
++ rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0);
++ rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0);
++#endif
++#ifdef CONFIG_EFI
++ if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature,
++ "EL64", 4))
++ efi_enabled = 1;
++#endif
++
++ ARCH_SETUP
++
++ memory_setup();
++ copy_edd();
++
++ if (!boot_params.hdr.root_flags)
++ root_mountflags &= ~MS_RDONLY;
++ init_mm.start_code = (unsigned long) &_text;
++ init_mm.end_code = (unsigned long) &_etext;
++ init_mm.end_data = (unsigned long) &_edata;
++ init_mm.brk = (unsigned long) &_end;
++
++ code_resource.start = virt_to_phys(&_text);
++ code_resource.end = virt_to_phys(&_etext)-1;
++ data_resource.start = virt_to_phys(&_etext);
++ data_resource.end = virt_to_phys(&_edata)-1;
++ bss_resource.start = virt_to_phys(&__bss_start);
++ bss_resource.end = virt_to_phys(&__bss_stop)-1;
++
++ early_identify_cpu(&boot_cpu_data);
++
++ strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
++ *cmdline_p = command_line;
++
++ parse_setup_data();
++
++ parse_early_param();
++
++#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT
++ if (init_ohci1394_dma_early)
++ init_ohci1394_dma_on_all_controllers();
++#endif
++
++ finish_e820_parsing();
++
++ /* after parse_early_param, so could debug it */
++ insert_resource(&iomem_resource, &code_resource);
++ insert_resource(&iomem_resource, &data_resource);
++ insert_resource(&iomem_resource, &bss_resource);
++
++ early_gart_iommu_check();
++
++ e820_register_active_regions(0, 0, -1UL);
++ /*
++ * partially used pages are not usable - thus
++ * we are rounding upwards:
++ */
++ end_pfn = e820_end_of_ram();
++ /* update e820 for memory not covered by WB MTRRs */
++ mtrr_bp_init();
++#ifndef CONFIG_XEN
++ if (mtrr_trim_uncached_memory(end_pfn)) {
++ e820_register_active_regions(0, 0, -1UL);
++ end_pfn = e820_end_of_ram();
++ }
++#endif
++
++ num_physpages = end_pfn;
++ max_mapnr = end_pfn;
++
++ check_efer();
++
++ max_pfn_mapped = init_memory_mapping(0, (max_pfn_mapped << PAGE_SHIFT));
++ if (efi_enabled)
++ efi_init();
++
++#ifndef CONFIG_XEN
++ vsmp_init();
++#endif
++
++ if (is_initial_xendomain())
++ dmi_scan_machine();
++
++ io_delay_init();
++
++#ifdef CONFIG_KVM_CLOCK
++ kvmclock_init();
++#endif
++
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ /* setup to use the early static init tables during kernel startup */
++ x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init;
++ x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init;
++#ifdef CONFIG_NUMA
++ x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init;
++#endif
++#endif
++
++ /* 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
++
++#ifdef CONFIG_NUMA
++ numa_initmem_init(0, end_pfn);
++#else
++ contig_initmem_init(0, end_pfn);
++#endif
++
++#ifndef CONFIG_XEN
++ dma32_reserve_bootmem();
++
++#ifdef CONFIG_ACPI_SLEEP
++ /*
++ * Reserve low memory region for sleep support.
++ */
++ acpi_reserve_bootmem();
++#endif
++
++ if (efi_enabled)
++ efi_reserve_bootmem();
++#endif
++
++#ifdef CONFIG_BLK_DEV_INITRD
++#ifdef CONFIG_XEN
++ if (xen_start_info->mod_start) {
++ unsigned long ramdisk_image = __pa(xen_start_info->mod_start);
++ unsigned long ramdisk_size = xen_start_info->mod_len;
++#else
++ if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) {
++ unsigned long ramdisk_image = boot_params.hdr.ramdisk_image;
++ unsigned long ramdisk_size = boot_params.hdr.ramdisk_size;
++#endif
++ unsigned long ramdisk_end = ramdisk_image + ramdisk_size;
++ unsigned long end_of_mem = end_pfn << PAGE_SHIFT;
++
++ if (ramdisk_end <= end_of_mem) {
++ /*
++ * don't need to reserve again, already reserved early
++ * in x86_64_start_kernel, and early_res_to_bootmem
++ * convert that to reserved in bootmem
++ */
++ initrd_start = ramdisk_image + PAGE_OFFSET;
++ initrd_end = initrd_start+ramdisk_size;
++#ifdef CONFIG_XEN
++ initrd_below_start_ok = 1;
++#endif
++ } else {
++ free_bootmem(ramdisk_image, ramdisk_size);
++ printk(KERN_ERR "initrd extends beyond end of memory "
++ "(0x%08lx > 0x%08lx)\ndisabling initrd\n",
++ ramdisk_end, end_of_mem);
++ initrd_start = 0;
++ }
++ }
++#endif
++ reserve_crashkernel();
++
++ reserve_ibft_region();
++
++ paging_init();
++ map_vsyscall();
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * Find and reserve possible boot-time SMP configuration:
++ */
++ find_smp_config();
++#endif
++#ifdef CONFIG_XEN
++ {
++ int i, j, k, fpp;
++ unsigned long p2m_pages;
++
++ p2m_pages = end_pfn;
++ if (xen_start_info->nr_pages > end_pfn) {
++ /*
++ * the end_pfn was shrunk (probably by mem= or highmem=
++ * kernel parameter); shrink reservation with the HV
++ */
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ unsigned int difference;
++ int ret;
++
++ difference = xen_start_info->nr_pages - end_pfn;
++
++ set_xen_guest_handle(reservation.extent_start,
++ ((unsigned long *)xen_start_info->mfn_list) + end_pfn);
++ reservation.nr_extents = difference;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ BUG_ON (ret != difference);
++ }
++ else if (end_pfn > xen_start_info->nr_pages)
++ p2m_pages = xen_start_info->nr_pages;
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Make sure we have a large enough P->M table. */
++ phys_to_machine_mapping = alloc_bootmem_pages(
++ end_pfn * sizeof(unsigned long));
++ memset(phys_to_machine_mapping, ~0,
++ end_pfn * sizeof(unsigned long));
++ memcpy(phys_to_machine_mapping,
++ (unsigned long *)xen_start_info->mfn_list,
++ p2m_pages * sizeof(unsigned long));
++ free_bootmem(
++ __pa(xen_start_info->mfn_list),
++ PFN_PHYS(PFN_UP(xen_start_info->nr_pages *
++ sizeof(unsigned long))));
++
++ /*
++ * Initialise the list of the frames that specify the
++ * list of frames that make up the p2m table. Used by
++ * save/restore.
++ */
++ pfn_to_mfn_frame_list_list = alloc_bootmem_pages(PAGE_SIZE);
++
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i=0, j=0, k=-1; i< end_pfn; i+=fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ BUG_ON(k>=fpp);
++ pfn_to_mfn_frame_list[k] =
++ alloc_bootmem_pages(PAGE_SIZE);
++ pfn_to_mfn_frame_list_list[k] =
++ virt_to_mfn(pfn_to_mfn_frame_list[k]);
++ j=0;
++ }
++ pfn_to_mfn_frame_list[k][j] =
++ virt_to_mfn(&phys_to_machine_mapping[i]);
++ }
++ HYPERVISOR_shared_info->arch.max_pfn = end_pfn;
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
++ }
++
++ /* Mark all ISA DMA channels in-use - using them wouldn't work. */
++ for (i = 0; i < MAX_DMA_CHANNELS; ++i)
++ if (i != 4 && request_dma(i, "xen") != 0)
++ BUG();
++ }
++
++#ifdef CONFIG_ACPI
++ if (!is_initial_xendomain()) {
++ acpi_disabled = 1;
++ acpi_ht = 0;
++ }
++#endif
++#endif
++
++#ifndef CONFIG_XEN
++ early_quirks();
++#endif
++
++#ifdef CONFIG_ACPI
++ /*
++ * Initialize the ACPI boot-time table parser (gets the RSDP and SDT).
++ * Call this early for SRAT node setup.
++ */
++ acpi_boot_table_init();
++
++ /*
++ * Read APIC and some other early information from ACPI tables.
++ */
++ acpi_boot_init();
++#endif
++
++ init_cpu_to_node();
++
++#ifdef CONFIG_X86_LOCAL_APIC
++ /*
++ * get boot-time SMP configuration:
++ */
++ if (smp_found_config)
++ get_smp_config();
++#ifndef CONFIG_XEN
++ init_apic_mappings();
++ ioapic_init_mappings();
++#endif
++#endif
++#if defined(CONFIG_XEN) && defined(CONFIG_SMP) && !defined(CONFIG_HOTPLUG_CPU)
++ prefill_possible_map();
++#endif
++
++ kvm_guest_init();
++
++ /*
++ * We trust e820 completely. No explicit ROM probing in memory.
++ */
++#ifdef CONFIG_XEN
++ if (is_initial_xendomain())
++ e820_reserve_resources(machine_e820.map, machine_e820.nr_map);
++#else
++ e820_reserve_resources(e820.map, e820.nr_map);
++ e820_mark_nosave_regions();
++#endif
++
++ /* 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;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++
++ if (is_initial_xendomain()) {
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ conswitchp = &vga_con;
++#elif defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++#endif
++ } else {
++#if defined(CONFIG_VT) && defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++ }
++ }
++#else /* CONFIG_XEN */
++
++#ifdef CONFIG_VT
++#if defined(CONFIG_VGA_CONSOLE)
++ if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY))
++ conswitchp = &vga_con;
++#elif defined(CONFIG_DUMMY_CONSOLE)
++ conswitchp = &dummy_con;
++#endif
++#endif
++
++#endif /* !CONFIG_XEN */
++
++ /* do this before identify_cpu for boot cpu */
++ check_enable_amd_mmconf_dmi();
++}
++
++#ifdef CONFIG_XEN
++static int
++xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
++{
++ HYPERVISOR_shutdown(SHUTDOWN_crash);
++ /* we're never actually going to get here... */
++ return NOTIFY_DONE;
++}
++#endif /* !CONFIG_XEN */
++
++
++static int __cpuinit get_model_name(struct cpuinfo_x86 *c)
++{
++ unsigned int *v;
++
++ if (c->extended_cpuid_level < 0x80000004)
++ return 0;
++
++ v = (unsigned int *) c->x86_model_id;
++ cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]);
++ cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]);
++ cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]);
++ c->x86_model_id[48] = 0;
++ return 1;
++}
++
++
++static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c)
++{
++ unsigned int n, dummy, eax, ebx, ecx, edx;
++
++ n = c->extended_cpuid_level;
++
++ if (n >= 0x80000005) {
++ cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
++ printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), "
++ "D cache %dK (%d bytes/line)\n",
++ edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
++ c->x86_cache_size = (ecx>>24) + (edx>>24);
++ /* On K8 L1 TLB is inclusive, so don't count it */
++ c->x86_tlbsize = 0;
++ }
++
++ if (n >= 0x80000006) {
++ cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
++ ecx = cpuid_ecx(0x80000006);
++ c->x86_cache_size = ecx >> 16;
++ c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
++
++ printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
++ c->x86_cache_size, ecx & 0xFF);
++ }
++ if (n >= 0x80000008) {
++ cpuid(0x80000008, &eax, &dummy, &dummy, &dummy);
++ c->x86_virt_bits = (eax >> 8) & 0xff;
++ c->x86_phys_bits = eax & 0xff;
++ }
++}
++
++#ifdef CONFIG_NUMA
++static int __cpuinit nearby_node(int apicid)
++{
++ int i, node;
++
++ for (i = apicid - 1; i >= 0; i--) {
++ node = apicid_to_node[i];
++ if (node != NUMA_NO_NODE && node_online(node))
++ return node;
++ }
++ for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) {
++ node = apicid_to_node[i];
++ if (node != NUMA_NO_NODE && node_online(node))
++ return node;
++ }
++ return first_node(node_online_map); /* Shouldn't happen */
++}
++#endif
++
++/*
++ * On a AMD dual core setup the lower bits of the APIC id distingush the cores.
++ * Assumes number of cores is a power of two.
++ */
++static void __cpuinit 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
++ bits = c->x86_coreid_bits;
++
++ /* Low order bits define the core id (index of core in socket) */
++ c->cpu_core_id = c->initial_apicid & ((1 << bits)-1);
++ /* Convert the initial APIC ID into the socket ID */
++ c->phys_proc_id = c->initial_apicid >> bits;
++
++#ifdef CONFIG_NUMA
++ node = c->phys_proc_id;
++ if (apicid_to_node[apicid] != NUMA_NO_NODE)
++ node = apicid_to_node[apicid];
++ if (!node_online(node)) {
++ /* Two possibilities here:
++ - The CPU is missing memory and no node was created.
++ In that case try picking one from a nearby CPU
++ - The APIC IDs differ from the HyperTransport node IDs
++ which the K8 northbridge parsing fills in.
++ Assume they are all increased by a constant offset,
++ but in the same order as the HT nodeids.
++ If that doesn't result in a usable node fall back to the
++ path for the previous case. */
++
++ int ht_nodeid = c->initial_apicid;
++
++ if (ht_nodeid >= 0 &&
++ apicid_to_node[ht_nodeid] != NUMA_NO_NODE)
++ node = apicid_to_node[ht_nodeid];
++ /* Pick a nearby node */
++ if (!node_online(node))
++ node = nearby_node(apicid);
++ }
++ numa_set_node(cpu, node);
++
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
++#endif
++}
++
++static void __cpuinit early_init_amd_mc(struct cpuinfo_x86 *c)
++{
++#ifdef CONFIG_SMP
++ unsigned bits, ecx;
++
++ /* Multi core CPU? */
++ if (c->extended_cpuid_level < 0x80000008)
++ return;
++
++ ecx = cpuid_ecx(0x80000008);
++
++ c->x86_max_cores = (ecx & 0xff) + 1;
++
++ /* CPU telling us the core id bits shift? */
++ bits = (ecx >> 12) & 0xF;
++
++ /* Otherwise recompute */
++ if (bits == 0) {
++ while ((1 << bits) < c->x86_max_cores)
++ bits++;
++ }
++
++ c->x86_coreid_bits = bits;
++
++#endif
++}
++
++#define ENABLE_C1E_MASK 0x18000000
++#define CPUID_PROCESSOR_SIGNATURE 1
++#define CPUID_XFAM 0x0ff00000
++#define CPUID_XFAM_K8 0x00000000
++#define CPUID_XFAM_10H 0x00100000
++#define CPUID_XFAM_11H 0x00200000
++#define CPUID_XMOD 0x000f0000
++#define CPUID_XMOD_REV_F 0x00040000
++
++#ifndef CONFIG_XEN
++/* AMD systems with C1E don't have a working lAPIC timer. Check for that. */
++static __cpuinit int amd_apic_timer_broken(void)
++{
++ u32 lo, hi, eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE);
++
++ switch (eax & CPUID_XFAM) {
++ case CPUID_XFAM_K8:
++ if ((eax & CPUID_XMOD) < CPUID_XMOD_REV_F)
++ break;
++ case CPUID_XFAM_10H:
++ case CPUID_XFAM_11H:
++ rdmsr(MSR_K8_ENABLE_C1E, lo, hi);
++ if (lo & ENABLE_C1E_MASK)
++ return 1;
++ break;
++ default:
++ /* err on the side of caution */
++ return 1;
++ }
++ return 0;
++}
++#endif
++
++static void __cpuinit early_init_amd(struct cpuinfo_x86 *c)
++{
++ early_init_amd_mc(c);
++
++ /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */
++ if (c->x86_power & (1<<8))
++ set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
++}
++
++static void __cpuinit init_amd(struct cpuinfo_x86 *c)
++{
++ unsigned level;
++
++#ifdef CONFIG_SMP
++ unsigned long value;
++
++ /*
++ * Disable TLB flush filter by setting HWCR.FFDIS on K8
++ * bit 6 of msr C001_0015
++ *
++ * Errata 63 for SH-B3 steppings
++ * Errata 122 for all steppings (F+ have it disabled by default)
++ */
++ if (c->x86 == 15) {
++ rdmsrl(MSR_K8_HWCR, value);
++ value |= 1 << 6;
++ wrmsrl(MSR_K8_HWCR, value);
++ }
++#endif
++
++ /* Bit 31 in normal CPUID used for nonstandard 3DNow ID;
++ 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */
++ clear_cpu_cap(c, 0*32+31);
++
++ /* 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_cpu_cap(c, X86_FEATURE_REP_GOOD);
++ if (c->x86 == 0x10 || c->x86 == 0x11)
++ set_cpu_cap(c, X86_FEATURE_REP_GOOD);
++
++ /* Enable workaround for FXSAVE leak */
++ if (c->x86 >= 6)
++ set_cpu_cap(c, X86_FEATURE_FXSAVE_LEAK);
++
++ 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);
++
++ /* Multi core CPU? */
++ if (c->extended_cpuid_level >= 0x80000008)
++ amd_detect_cmp(c);
++
++ if (c->extended_cpuid_level >= 0x80000006 &&
++ (cpuid_edx(0x80000006) & 0xf000))
++ num_cache_leaves = 4;
++ else
++ num_cache_leaves = 3;
++
++ if (c->x86 == 0xf || c->x86 == 0x10 || c->x86 == 0x11)
++ set_cpu_cap(c, X86_FEATURE_K8);
++
++ /* MFENCE stops RDTSC speculation */
++ set_cpu_cap(c, X86_FEATURE_MFENCE_RDTSC);
++
++ if (c->x86 == 0x10)
++ fam10h_check_enable_mmcfg();
++
++#ifndef CONFIG_XEN
++ if (amd_apic_timer_broken())
++ disable_apic_timer = 1;
++
++ if (c == &boot_cpu_data && c->x86 >= 0xf && c->x86 <= 0x11) {
++ unsigned long long tseg;
++
++ /*
++ * Split up direct mapping around the TSEG SMM area.
++ * Don't do it for gbpages because there seems very little
++ * benefit in doing so.
++ */
++ if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg) &&
++ (tseg >> PMD_SHIFT) < (max_pfn_mapped >> (PMD_SHIFT-PAGE_SHIFT)))
++ set_memory_4k((unsigned long)__va(tseg), 1);
++ }
++#endif
++}
++
++void __cpuinit detect_ht(struct cpuinfo_x86 *c)
++{
++#ifdef CONFIG_SMP
++ u32 eax, ebx, ecx, edx;
++ int index_msb, core_bits;
++
++ cpuid(1, &eax, &ebx, &ecx, &edx);
++
++
++ if (!cpu_has(c, X86_FEATURE_HT))
++ return;
++ if (cpu_has(c, X86_FEATURE_CMP_LEGACY))
++ goto out;
++
++ smp_num_siblings = (ebx & 0xff0000) >> 16;
++
++ if (smp_num_siblings == 1) {
++ printk(KERN_INFO "CPU: Hyper-Threading is disabled\n");
++ } else if (smp_num_siblings > 1) {
++
++ if (smp_num_siblings > NR_CPUS) {
++ printk(KERN_WARNING "CPU: Unsupported number of "
++ "siblings %d", smp_num_siblings);
++ smp_num_siblings = 1;
++ return;
++ }
++
++ index_msb = get_count_order(smp_num_siblings);
++ c->phys_proc_id = phys_pkg_id(index_msb);
++
++ smp_num_siblings = smp_num_siblings / c->x86_max_cores;
++
++ index_msb = get_count_order(smp_num_siblings);
++
++ core_bits = get_count_order(c->x86_max_cores);
++
++ c->cpu_core_id = phys_pkg_id(index_msb) &
++ ((1 << core_bits) - 1);
++ }
++out:
++ if ((c->x86_max_cores * smp_num_siblings) > 1) {
++ printk(KERN_INFO "CPU: Physical Processor ID: %d\n",
++ c->phys_proc_id);
++ printk(KERN_INFO "CPU: Processor Core ID: %d\n",
++ c->cpu_core_id);
++ }
++
++#endif
++}
++
++/*
++ * find out the number of processor cores on the die
++ */
++static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c)
++{
++ unsigned int eax, t;
++
++ if (c->cpuid_level < 4)
++ return 1;
++
++ cpuid_count(4, 0, &eax, &t, &t, &t);
++
++ if (eax & 0x1f)
++ return ((eax >> 26) + 1);
++ else
++ return 1;
++}
++
++static void __cpuinit srat_detect_node(void)
++{
++#ifdef CONFIG_NUMA
++ unsigned node;
++ int cpu = smp_processor_id();
++ int apicid = hard_smp_processor_id();
++
++ /* Don't do the funky fallback heuristics the AMD version employs
++ for now. */
++ node = apicid_to_node[apicid];
++ if (node == NUMA_NO_NODE || !node_online(node))
++ node = first_node(node_online_map);
++ numa_set_node(cpu, node);
++
++ printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node);
++#endif
++}
++
++static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
++{
++ if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
++ (c->x86 == 0x6 && c->x86_model >= 0x0e))
++ set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
++}
++
++static void __cpuinit init_intel(struct cpuinfo_x86 *c)
++{
++ /* Cache sizes */
++ unsigned n;
++
++ init_intel_cacheinfo(c);
++ if (c->cpuid_level > 9) {
++ unsigned eax = cpuid_eax(10);
++ /* Check for version and the number of counters */
++ if ((eax & 0xff) && (((eax>>8) & 0xff) > 1))
++ set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON);
++ }
++
++ if (cpu_has_ds) {
++ unsigned int l1, l2;
++ rdmsr(MSR_IA32_MISC_ENABLE, l1, l2);
++ if (!(l1 & (1<<11)))
++ set_cpu_cap(c, X86_FEATURE_BTS);
++ if (!(l1 & (1<<12)))
++ set_cpu_cap(c, X86_FEATURE_PEBS);
++ }
++
++
++ if (cpu_has_bts)
++ ds_init_intel(c);
++
++ n = c->extended_cpuid_level;
++ if (n >= 0x80000008) {
++ unsigned eax = cpuid_eax(0x80000008);
++ c->x86_virt_bits = (eax >> 8) & 0xff;
++ c->x86_phys_bits = eax & 0xff;
++ /* CPUID workaround for Intel 0F34 CPU */
++ if (c->x86_vendor == X86_VENDOR_INTEL &&
++ c->x86 == 0xF && c->x86_model == 0x3 &&
++ c->x86_mask == 0x4)
++ c->x86_phys_bits = 36;
++ }
++
++ if (c->x86 == 15)
++ c->x86_cache_alignment = c->x86_clflush_size * 2;
++ if (c->x86 == 6)
++ set_cpu_cap(c, X86_FEATURE_REP_GOOD);
++ set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
++ c->x86_max_cores = intel_num_cpu_cores(c);
++
++ srat_detect_node();
++}
++
++static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c)
++{
++ if (c->x86 == 0x6 && c->x86_model >= 0xf)
++ set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
++}
++
++static void __cpuinit init_centaur(struct cpuinfo_x86 *c)
++{
++ /* Cache sizes */
++ unsigned n;
++
++ 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;
++ }
++
++ if (c->x86 == 0x6 && c->x86_model >= 0xf) {
++ c->x86_cache_alignment = c->x86_clflush_size * 2;
++ set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
++ set_cpu_cap(c, X86_FEATURE_REP_GOOD);
++ }
++ set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
++}
++
++static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c)
++{
++ char *v = c->x86_vendor_id;
++
++ if (!strcmp(v, "AuthenticAMD"))
++ c->x86_vendor = X86_VENDOR_AMD;
++ else if (!strcmp(v, "GenuineIntel"))
++ c->x86_vendor = X86_VENDOR_INTEL;
++ else if (!strcmp(v, "CentaurHauls"))
++ c->x86_vendor = X86_VENDOR_CENTAUR;
++ else
++ c->x86_vendor = X86_VENDOR_UNKNOWN;
++}
++
++/* 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. */
++static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c)
++{
++ u32 tfms, xlvl;
++
++ 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->x86_coreid_bits = 0;
++ c->extended_cpuid_level = 0;
++ memset(&c->x86_capability, 0, sizeof c->x86_capability);
++
++ /* Get vendor name */
++ cpuid(0x00000000, (unsigned int *)&c->cpuid_level,
++ (unsigned int *)&c->x86_vendor_id[0],
++ (unsigned int *)&c->x86_vendor_id[8],
++ (unsigned int *)&c->x86_vendor_id[4]);
++
++ get_cpu_vendor(c);
++
++ /* Initialize the standard set of capabilities */
++ /* Note that the vendor-specific code below might override */
++
++ /* Intel-defined flags: level 0x00000001 */
++ if (c->cpuid_level >= 0x00000001) {
++ __u32 misc;
++ cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4],
++ &c->x86_capability[0]);
++ c->x86 = (tfms >> 8) & 0xf;
++ c->x86_model = (tfms >> 4) & 0xf;
++ c->x86_mask = tfms & 0xf;
++ if (c->x86 == 0xf)
++ c->x86 += (tfms >> 20) & 0xff;
++ if (c->x86 >= 0x6)
++ c->x86_model += ((tfms >> 16) & 0xF) << 4;
++ if (test_cpu_cap(c, X86_FEATURE_CLFLSH))
++ c->x86_clflush_size = ((misc >> 8) & 0xff) * 8;
++ } else {
++ /* Have CPUID level 0 only - unheard of */
++ c->x86 = 4;
++ }
++
++ c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff;
++#ifdef CONFIG_SMP
++ c->phys_proc_id = c->initial_apicid;
++#endif
++ /* AMD-defined flags: level 0x80000001 */
++ xlvl = cpuid_eax(0x80000000);
++ c->extended_cpuid_level = xlvl;
++ if ((xlvl & 0xffff0000) == 0x80000000) {
++ if (xlvl >= 0x80000001) {
++ c->x86_capability[1] = cpuid_edx(0x80000001);
++ c->x86_capability[6] = cpuid_ecx(0x80000001);
++ }
++ if (xlvl >= 0x80000004)
++ get_model_name(c); /* Default name */
++ }
++
++ /* Transmeta-defined flags: level 0x80860001 */
++ xlvl = cpuid_eax(0x80860000);
++ if ((xlvl & 0xffff0000) == 0x80860000) {
++ /* Don't set x86_cpuid_level here for now to not confuse. */
++ if (xlvl >= 0x80860001)
++ c->x86_capability[2] = cpuid_edx(0x80860001);
++ }
++
++ c->extended_cpuid_level = cpuid_eax(0x80000000);
++ if (c->extended_cpuid_level >= 0x80000007)
++ c->x86_power = cpuid_edx(0x80000007);
++
++ switch (c->x86_vendor) {
++ case X86_VENDOR_AMD:
++ early_init_amd(c);
++ break;
++ case X86_VENDOR_INTEL:
++ early_init_intel(c);
++ break;
++ case X86_VENDOR_CENTAUR:
++ early_init_centaur(c);
++ break;
++ }
++
++ validate_pat_support(c);
++}
++
++/*
++ * This does the hard work of actually picking apart the CPU stuff...
++ */
++void __cpuinit identify_cpu(struct cpuinfo_x86 *c)
++{
++ int i;
++
++ early_identify_cpu(c);
++
++ init_scattered_cpuid_features(c);
++
++ c->apicid = phys_pkg_id(0);
++
++ /*
++ * Vendor-specific initialization. In this section we
++ * canonicalize the feature flags, meaning if there are
++ * features a certain CPU supports which CPUID doesn't
++ * tell us, CPUID claiming incorrect flags, or other bugs,
++ * we handle them here.
++ *
++ * At the end of this section, c->x86_capability better
++ * indicate the features this CPU genuinely supports!
++ */
++ switch (c->x86_vendor) {
++ case X86_VENDOR_AMD:
++ init_amd(c);
++ break;
++
++ case X86_VENDOR_INTEL:
++ init_intel(c);
++ break;
++
++ case X86_VENDOR_CENTAUR:
++ init_centaur(c);
++ break;
++
++ case X86_VENDOR_UNKNOWN:
++ default:
++ display_cacheinfo(c);
++ break;
++ }
++
++ detect_ht(c);
++
++ /*
++ * On SMP, boot_cpu_data holds the common feature set between
++ * all CPUs; so make sure that we indicate which features are
++ * common between the CPUs. The first time this routine gets
++ * executed, c == &boot_cpu_data.
++ */
++ if (c != &boot_cpu_data) {
++ /* AND the already accumulated flags with these */
++ for (i = 0; i < NCAPINTS; i++)
++ boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
++ }
++
++ /* Clear all flags overriden by options */
++ for (i = 0; i < NCAPINTS; i++)
++ c->x86_capability[i] &= ~cleared_cpu_caps[i];
++
++#ifdef CONFIG_X86_MCE
++ mcheck_init(c);
++#endif
++ select_idle_routine(c);
++
++#ifdef CONFIG_NUMA
++ numa_add_cpu(smp_processor_id());
++#endif
++
++}
++
++void __cpuinit identify_boot_cpu(void)
++{
++ identify_cpu(&boot_cpu_data);
++}
++
++void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c)
++{
++ BUG_ON(c == &boot_cpu_data);
++ identify_cpu(c);
++ mtrr_ap_init();
++}
++
++static __init int setup_noclflush(char *arg)
++{
++ setup_clear_cpu_cap(X86_FEATURE_CLFLSH);
++ return 1;
++}
++__setup("noclflush", setup_noclflush);
++
++void __cpuinit print_cpu_info(struct cpuinfo_x86 *c)
++{
++ if (c->x86_model_id[0])
++ printk(KERN_CONT "%s", c->x86_model_id);
++
++ if (c->x86_mask || c->cpuid_level >= 0)
++ printk(KERN_CONT " stepping %02x\n", c->x86_mask);
++ else
++ printk(KERN_CONT "\n");
++}
++
++static __init int setup_disablecpuid(char *arg)
++{
++ int bit;
++ if (get_option(&arg, &bit) && bit < NCAPINTS*32)
++ setup_clear_cpu_cap(bit);
++ else
++ return 0;
++ return 1;
++}
++__setup("clearcpuid=", setup_disablecpuid);
+diff --git a/arch/x86/kernel/smp-xen.c b/arch/x86/kernel/smp-xen.c
+new file mode 100644
+index 0000000..fb737c7
+--- /dev/null
++++ b/arch/x86/kernel/smp-xen.c
+@@ -0,0 +1,329 @@
++/*
++ * 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.
++ *
++ * i386 and x86_64 integration by Glauber Costa <gcosta at redhat.com>
++ *
++ * This code is released under the GNU General Public License version 2 or
++ * later.
++ */
++
++#include <linux/init.h>
++
++#include <linux/mm.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/kernel_stat.h>
++#include <linux/mc146818rtc.h>
++#include <linux/cache.h>
++#include <linux/interrupt.h>
++#include <linux/cpu.h>
++
++#include <asm/mtrr.h>
++#include <asm/tlbflush.h>
++#include <asm/mmu_context.h>
++#include <asm/proto.h>
++#include <mach_ipi.h>
++#include <xen/evtchn.h>
++/*
++ * 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 that's 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.
++ */
++
++/*
++ * 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 xen_smp_send_reschedule(int cpu)
++{
++ if (unlikely(cpu_is_offline(cpu))) {
++ WARN_ON(1);
++ return;
++ }
++ send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR);
++}
++
++/*
++ * Structure and data for smp_call_function(). This is designed to minimise
++ * static memory requirements. It also looks cleaner.
++ */
++static DEFINE_SPINLOCK(call_lock);
++
++struct call_data_struct {
++ void (*func) (void *info);
++ void *info;
++ atomic_t started;
++ atomic_t finished;
++ int wait;
++};
++
++void lock_ipi_call_lock(void)
++{
++ spin_lock_irq(&call_lock);
++}
++
++void unlock_ipi_call_lock(void)
++{
++ spin_unlock_irq(&call_lock);
++}
++
++static struct call_data_struct *call_data;
++
++static void __smp_call_function(void (*func) (void *info), void *info,
++ int nonatomic, int wait)
++{
++ struct call_data_struct data;
++ int cpus = num_online_cpus() - 1;
++
++ if (!cpus)
++ return;
++
++ data.func = func;
++ data.info = info;
++ atomic_set(&data.started, 0);
++ data.wait = wait;
++ if (wait)
++ atomic_set(&data.finished, 0);
++
++ call_data = &data;
++ mb();
++
++ /* Send a message to all other CPUs and wait for them to respond */
++ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ cpu_relax();
++
++ if (wait)
++ while (atomic_read(&data.finished) != cpus)
++ cpu_relax();
++}
++
++
++/**
++ * smp_call_function_mask(): Run a function on a set of other CPUs.
++ * @mask: The set of cpus to run on. Must not include the current cpu.
++ * @func: The function to run. This must be fast and non-blocking.
++ * @info: An arbitrary pointer to pass to the function.
++ * @wait: If true, wait (atomically) until function has completed on other CPUs.
++ *
++ * Returns 0 on success, else a negative status code.
++ *
++ * If @wait is true, then returns once @func has returned; otherwise
++ * it returns just before the target cpu calls @func.
++ *
++ * You must not call this function with disabled interrupts or from a
++ * hardware interrupt handler or from a bottom half handler.
++ */
++int
++xen_smp_call_function_mask(cpumask_t mask,
++ void (*func)(void *), void *info,
++ int wait)
++{
++ struct call_data_struct data;
++ cpumask_t allbutself;
++ int cpus;
++
++ /* Can deadlock when called with interrupts disabled */
++ WARN_ON(irqs_disabled());
++
++ /* Holding any lock stops cpus from going down. */
++ spin_lock(&call_lock);
++
++ allbutself = cpu_online_map;
++ cpu_clear(smp_processor_id(), allbutself);
++
++ cpus_and(mask, mask, allbutself);
++ cpus = cpus_weight(mask);
++
++ if (!cpus) {
++ spin_unlock(&call_lock);
++ return 0;
++ }
++
++ 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 other CPUs */
++ if (cpus_equal(mask, allbutself) &&
++ cpus_equal(cpu_online_map, cpu_callout_map))
++ send_IPI_allbutself(CALL_FUNCTION_VECTOR);
++ else
++ send_IPI_mask(mask, CALL_FUNCTION_VECTOR);
++
++ /* Wait for response */
++ while (atomic_read(&data.started) != cpus)
++ cpu_relax();
++
++ if (wait)
++ while (atomic_read(&data.finished) != cpus)
++ cpu_relax();
++ spin_unlock(&call_lock);
++
++ return 0;
++}
++
++static void stop_this_cpu(void *dummy)
++{
++ local_irq_disable();
++ /*
++ * Remove this CPU:
++ */
++ cpu_clear(smp_processor_id(), cpu_online_map);
++ disable_all_local_evtchn();
++ if (hlt_works(smp_processor_id()))
++ for (;;) halt();
++ for (;;);
++}
++
++/*
++ * this function calls the 'stop' function on all other CPUs in the system.
++ */
++
++void xen_smp_send_stop(void)
++{
++ int nolock;
++ unsigned long flags;
++
++ /* Don't deadlock on the call lock in panic */
++ nolock = !spin_trylock(&call_lock);
++ local_irq_save(flags);
++ __smp_call_function(stop_this_cpu, NULL, 0, 0);
++ if (!nolock)
++ spin_unlock(&call_lock);
++ disable_all_local_evtchn();
++ local_irq_restore(flags);
++}
++
++/*
++ * 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)
++{
++#ifdef CONFIG_X86_32
++ __get_cpu_var(irq_stat).irq_resched_count++;
++#else
++ add_pda(irq_resched_count, 1);
++#endif
++ return IRQ_HANDLED;
++}
++
++irqreturn_t smp_call_function_interrupt(int irq, void *dev_id)
++{
++ void (*func) (void *info) = call_data->func;
++ void *info = call_data->info;
++ int wait = call_data->wait;
++
++ /*
++ * 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);
++#ifdef CONFIG_X86_32
++ __get_cpu_var(irq_stat).irq_call_count++;
++#else
++ add_pda(irq_call_count, 1);
++#endif
++ irq_exit();
++
++ if (wait) {
++ mb();
++ atomic_inc(&call_data->finished);
++ }
++
++ return IRQ_HANDLED;
++}
+diff --git a/arch/x86/kernel/time_32-xen.c b/arch/x86/kernel/time_32-xen.c
+new file mode 100644
+index 0000000..6d4ea72
+--- /dev/null
++++ b/arch/x86/kernel/time_32-xen.c
+@@ -0,0 +1,1003 @@
++/*
++ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
++ *
++ * This file contains the PC-specific time handling details:
++ * reading the RTC at bootup, etc..
++ * 1994-07-02 Alan Modra
++ * fixed set_rtc_mmss, fixed time.year for >= 2000, new mktime
++ * 1995-03-26 Markus Kuhn
++ * fixed 500 ms bug at call to set_rtc_mmss, fixed DS12887
++ * precision CMOS clock update
++ * 1996-05-03 Ingo Molnar
++ * fixed time warps in do_[slow|fast]_gettimeoffset()
++ * 1997-09-10 Updated NTP code according to technical memorandum Jan '96
++ * "A Kernel Model for Precision Timekeeping" by Dave Mills
++ * 1998-09-05 (Various)
++ * More robust do_fast_gettimeoffset() algorithm implemented
++ * (works with APM, Cyrix 6x86MX and Centaur C6),
++ * monotonic gettimeofday() with fast_get_timeoffset(),
++ * drift-proof precision TSC calibration on boot
++ * (C. Scott Ananian <cananian at alumni.princeton.edu>, Andrew D.
++ * Balsa <andrebalsa at altern.org>, Philip Gladstone <philip at raptor.com>;
++ * ported from 2.0.35 Jumbo-9 by Michael Krause <m.krause at tu-harburg.de>).
++ * 1998-12-16 Andrea Arcangeli
++ * Fixed Jumbo-9 code in 2.1.131: do_gettimeofday was missing 1 jiffy
++ * because was not accounting lost_ticks.
++ * 1998-12-24 Copyright (C) 1998 Andrea Arcangeli
++ * Fixed a xtime SMP race (we need the xtime_lock rw spinlock to
++ * serialize accesses to xtime/lost_ticks).
++ */
++
++#include <linux/init.h>
++#include <linux/interrupt.h>
++#include <linux/time.h>
++#include <linux/mca.h>
++#include <linux/sysctl.h>
++#include <linux/percpu.h>
++#include <linux/kernel_stat.h>
++#include <linux/posix-timers.h>
++#include <linux/cpufreq.h>
++#include <linux/clocksource.h>
++#include <linux/sysdev.h>
++
++#include <asm/delay.h>
++#include <asm/time.h>
++
++#include <xen/evtchn.h>
++#include <xen/sysctl.h>
++#include <xen/interface/vcpu.h>
++
++#include <asm/i8253.h>
++DEFINE_SPINLOCK(i8253_lock);
++EXPORT_SYMBOL(i8253_lock);
++
++#ifdef CONFIG_X86_64
++#include <asm/vsyscall.h>
++volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
++#endif
++
++#define XEN_SHIFT 22
++
++unsigned int cpu_khz; /* Detected as we calibrate the TSC */
++EXPORT_SYMBOL(cpu_khz);
++
++/* 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 void __clock_was_set(struct work_struct *unused)
++{
++ clock_was_set();
++}
++static DECLARE_WORK(clock_was_set_work, __clock_was_set);
++
++/*
++ * GCC 4.3 can turn loops over an induction variable into division. We do
++ * not support arbitrary 64-bit division, and so must break the induction.
++ */
++#define clobber_induction_variable(v) asm ( "" : "+r" (v) )
++
++static inline void __normalize_time(time_t *sec, s64 *nsec)
++{
++ while (*nsec >= NSEC_PER_SEC) {
++ clobber_induction_variable(*nsec);
++ (*nsec) -= NSEC_PER_SEC;
++ (*sec)++;
++ }
++ while (*nsec < 0) {
++ clobber_induction_variable(*nsec);
++ (*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);
++
++int xen_independent_wallclock(void)
++{
++ return 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);
++
++#if 0
++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,
++};
++#endif
++
++/*
++ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
++ * yielding a 64-bit result.
++ */
++static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
++{
++ u64 product;
++#ifdef __i386__
++ u32 tmp1, tmp2;
++#endif
++
++ if (shift < 0)
++ delta >>= -shift;
++ else
++ delta <<= shift;
++
++#ifdef __i386__
++ __asm__ (
++ "mul %5 ; "
++ "mov %4,%%eax ; "
++ "mov %%edx,%4 ; "
++ "mul %5 ; "
++ "xor %5,%5 ; "
++ "add %4,%%eax ; "
++ "adc %5,%%edx ; "
++ : "=A" (product), "=r" (tmp1), "=r" (tmp2)
++ : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
++#else
++ __asm__ (
++ "mul %%rdx ; shrd $32,%%rdx,%%rax"
++ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
++#endif
++
++ return product;
++}
++
++static inline u64 get64(volatile u64 *ptr)
++{
++#ifndef CONFIG_64BIT
++ return cmpxchg64(ptr, 0, 0);
++#else
++ return *ptr;
++#endif
++}
++
++static inline u64 get64_local(volatile u64 *ptr)
++{
++#ifndef CONFIG_64BIT
++ return cmpxchg64_local(ptr, 0, 0);
++#else
++ return *ptr;
++#endif
++}
++
++#if 0 /* defined (__i386__) */
++int read_current_timer(unsigned long *timer_val)
++{
++ rdtscl(*timer_val);
++ return 0;
++}
++#endif
++
++static void init_cpu_khz(void)
++{
++ u64 __cpu_khz = 1000000ULL << 32;
++ struct vcpu_time_info *info = &vcpu_info(0)->time;
++ do_div(__cpu_khz, info->tsc_to_system_mul);
++ if (info->tsc_shift < 0)
++ cpu_khz = __cpu_khz << -info->tsc_shift;
++ else
++ cpu_khz = __cpu_khz >> info->tsc_shift;
++}
++
++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);
++}
++
++static void __update_wallclock(time_t sec, long nsec)
++{
++ long wtm_nsec, xtime_nsec;
++ time_t wtm_sec, xtime_sec;
++ u64 tmp, wc_nsec;
++
++ /* Adjust wall-clock time base. */
++ wc_nsec = processed_system_time;
++ wc_nsec += sec * (u64)NSEC_PER_SEC;
++ wc_nsec += nsec;
++
++ /* Split wallclock base into seconds and nanoseconds. */
++ tmp = wc_nsec;
++ xtime_nsec = do_div(tmp, 1000000000);
++ xtime_sec = (time_t)tmp;
++
++ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
++ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
++
++ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
++ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
++
++ ntp_clear();
++}
++
++static void update_wallclock(void)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++
++ 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);
++}
++
++/*
++ * Reads a consistent set of time-base values from Xen, into a shadow data
++ * area.
++ */
++static void get_time_values_from_xen(unsigned int cpu)
++{
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
++
++ src = &vcpu_info(cpu)->time;
++ dst = &per_cpu(shadow_time, cpu);
++
++ 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(unsigned int cpu)
++{
++ struct vcpu_time_info *src;
++ struct shadow_time_info *dst;
++
++ src = &vcpu_info(cpu)->time;
++ dst = &per_cpu(shadow_time, cpu);
++
++ rmb();
++ return (dst->version == src->version);
++}
++
++static void sync_xen_wallclock(unsigned long dummy);
++static DEFINE_TIMER(sync_xen_wallclock_timer, sync_xen_wallclock, 0, 0);
++static void sync_xen_wallclock(unsigned long dummy)
++{
++ time_t sec;
++ s64 nsec;
++ struct xen_platform_op op;
++
++ BUG_ON(!is_initial_xendomain());
++ if (!ntp_synced() || independent_wallclock)
++ return;
++
++ write_seqlock_irq(&xtime_lock);
++
++ sec = xtime.tv_sec;
++ nsec = xtime.tv_nsec;
++ __normalize_time(&sec, &nsec);
++
++ op.cmd = XENPF_settime;
++ op.u.settime.secs = sec;
++ op.u.settime.nsecs = nsec;
++ op.u.settime.system_time = processed_system_time;
++ WARN_ON(HYPERVISOR_platform_op(&op));
++
++ update_wallclock();
++
++ write_sequnlock_irq(&xtime_lock);
++
++ /* Once per minute. */
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 60*HZ);
++}
++
++static unsigned long long local_clock(void)
++{
++ unsigned int cpu = get_cpu();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ u64 time;
++ u32 local_time_version;
++
++ do {
++ local_time_version = shadow->version;
++ barrier();
++ time = shadow->system_timestamp + get_nsec_offset(shadow);
++ if (!time_values_up_to_date(cpu))
++ get_time_values_from_xen(cpu);
++ barrier();
++ } while (local_time_version != shadow->version);
++
++ put_cpu();
++
++ return time;
++}
++
++/*
++ * Runstate accounting
++ */
++static void get_runstate_snapshot(struct vcpu_runstate_info *res)
++{
++ u64 state_time;
++ struct vcpu_runstate_info *state;
++
++ BUG_ON(preemptible());
++
++ state = &__get_cpu_var(runstate);
++
++ do {
++ state_time = get64_local(&state->state_entry_time);
++ *res = *state;
++ } while (get64_local(&state->state_entry_time) != state_time);
++
++ WARN_ON_ONCE(res->state != RUNSTATE_running);
++}
++
++/*
++ * Xen sched_clock implementation. Returns the number of unstolen
++ * nanoseconds, which is nanoseconds the VCPU spent in RUNNING+BLOCKED
++ * states.
++ */
++unsigned long long sched_clock(void)
++{
++ struct vcpu_runstate_info runstate;
++ cycle_t now;
++ u64 ret;
++ s64 offset;
++
++ /*
++ * Ideally sched_clock should be called on a per-cpu basis
++ * anyway, so preempt should already be disabled, but that's
++ * not current practice at the moment.
++ */
++ preempt_disable();
++
++ now = local_clock();
++
++ get_runstate_snapshot(&runstate);
++
++ offset = now - runstate.state_entry_time;
++ if (offset < 0)
++ offset = 0;
++
++ ret = offset + runstate.time[RUNSTATE_running]
++ + runstate.time[RUNSTATE_blocked];
++
++ preempt_enable();
++
++ return ret;
++}
++
++unsigned long profile_pc(struct pt_regs *regs)
++{
++ unsigned long pc = instruction_pointer(regs);
++
++#if defined(CONFIG_SMP) || defined(__x86_64__)
++# ifdef __i386__
++ if (!v8086_mode(regs) && SEGMENT_IS_KERNEL_CODE(regs->cs)
++# else
++ if (!user_mode(regs)
++# endif
++ && in_lock_functions(pc)) {
++# ifdef CONFIG_FRAME_POINTER
++ return ((unsigned long *)regs->bp)[1];
++# else
++# ifdef __i386__
++ unsigned long *sp = (unsigned long *)®s->sp;
++# else
++ unsigned long *sp = (unsigned long *)regs->sp;
++# endif
++
++ /* Return address is either directly at stack pointer
++ or above a saved flags. 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
++
++ return pc;
++}
++EXPORT_SYMBOL(profile_pc);
++
++/*
++ * This is the same as the above, except we _also_ save the current
++ * Time Stamp Counter value at the time of the timer interrupt, so that
++ * we later on can estimate the time of day more exactly.
++ */
++irqreturn_t timer_interrupt(int irq, void *dev_id)
++{
++ s64 delta, delta_cpu, stolen, blocked;
++ unsigned int i, cpu = smp_processor_id();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++ struct vcpu_runstate_info runstate;
++
++ /* Keep nmi watchdog up to date */
++#ifdef __i386__
++ per_cpu(irq_stat, smp_processor_id()).irq0_irqs++;
++#else
++ add_pda(irq0_irqs, 1);
++#endif
++
++ /*
++ * Here we are in the timer irq handler. We just have irqs locally
++ * disabled but we don't know if the timer_bh is running on the other
++ * CPU. We need to avoid to SMP race with it. NOTE: we don' t need
++ * the irq version of write_lock because as just said we have irq
++ * locally disabled. -arca
++ */
++ write_seqlock(&xtime_lock);
++
++ do {
++ get_time_values_from_xen(cpu);
++
++ /* 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);
++
++ get_runstate_snapshot(&runstate);
++ } 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/%u: Time went backwards: "
++ "delta=%lld delta_cpu=%lld shadow=%lld "
++ "off=%lld processed=%lld cpu_processed=%lld\n",
++ cpu, delta, delta_cpu, shadow->system_timestamp,
++ (s64)get_nsec_offset(shadow),
++ processed_system_time,
++ per_cpu(processed_system_time, cpu));
++ for (i = 0; i < num_online_cpus(); i++)
++ printk(" %d: %lld\n", i,
++ per_cpu(processed_system_time, i));
++ }
++
++ /* System-wide jiffy work. */
++ if (delta >= NS_PER_TICK) {
++ do_div(delta, NS_PER_TICK);
++ processed_system_time += delta * NS_PER_TICK;
++ while (delta > HZ) {
++ do_timer(HZ);
++ delta -= HZ;
++ }
++ do_timer(delta);
++ }
++
++ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
++ update_wallclock();
++ if (keventd_up())
++ schedule_work(&clock_was_set_work);
++ }
++
++ write_sequnlock(&xtime_lock);
++
++ /*
++ * Account stolen ticks.
++ * HACK: Passing NULL to account_steal_time()
++ * ensures that the ticks are accounted as stolen.
++ */
++ stolen = runstate.time[RUNSTATE_runnable]
++ + runstate.time[RUNSTATE_offline]
++ - per_cpu(processed_stolen_time, cpu);
++ 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);
++ }
++
++ /*
++ * Account blocked ticks.
++ * HACK: Passing idle_task to account_steal_time()
++ * ensures that the ticks are accounted as idle/wait.
++ */
++ blocked = runstate.time[RUNSTATE_blocked]
++ - per_cpu(processed_blocked_time, cpu);
++ 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);
++ }
++
++ /* Account user/system ticks. */
++ if (delta_cpu > 0) {
++ do_div(delta_cpu, NS_PER_TICK);
++ per_cpu(processed_system_time, cpu) += delta_cpu * NS_PER_TICK;
++ if (user_mode_vm(get_irq_regs()))
++ account_user_time(current, (cputime_t)delta_cpu);
++ else
++ account_system_time(current, HARDIRQ_OFFSET,
++ (cputime_t)delta_cpu);
++ }
++
++ /* Offlined for more than a few seconds? Avoid lockup warnings. */
++ if (stolen > 5*HZ)
++ touch_softlockup_watchdog();
++
++ /* Local timer processing (see update_process_times()). */
++ run_local_timers();
++ if (rcu_pending(cpu))
++ rcu_check_callbacks(cpu, user_mode_vm(get_irq_regs()));
++ scheduler_tick();
++ run_posix_cpu_timers(current);
++ profile_tick(CPU_PROFILING);
++
++ return IRQ_HANDLED;
++}
++
++void mark_tsc_unstable(char *reason)
++{
++#ifndef CONFIG_XEN /* XXX Should tell the hypervisor about this fact. */
++ tsc_unstable = 1;
++#endif
++}
++EXPORT_SYMBOL_GPL(mark_tsc_unstable);
++
++static cycle_t cs_last;
++
++static cycle_t xen_clocksource_read(void)
++{
++ cycle_t ret = local_clock();
++
++#ifdef CONFIG_SMP
++ for (;;) {
++ cycle_t last = get64(&cs_last);
++
++ if ((s64)(ret - last) < 0) {
++ if (last - ret > permitted_clock_jitter
++ && printk_ratelimit()) {
++ unsigned int cpu = get_cpu();
++ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
++
++ printk(KERN_WARNING "clocksource/%u: "
++ "Time went backwards: "
++ "ret=%Lx delta=%Ld shadow=%Lx offset=%Lx\n",
++ cpu, ret, ret - last,
++ shadow->system_timestamp,
++ get_nsec_offset(shadow));
++ put_cpu();
++ }
++ ret = last;
++ }
++ if (cmpxchg64(&cs_last, last, ret) == last)
++ break;
++ }
++#endif
++
++ return ret;
++}
++
++static void xen_clocksource_resume(void)
++{
++ extern void time_resume(void);
++
++ time_resume();
++ cs_last = local_clock();
++}
++
++static struct clocksource clocksource_xen = {
++ .name = "xen",
++ .rating = 400,
++ .read = xen_clocksource_read,
++ .mask = CLOCKSOURCE_MASK(64),
++ .mult = 1 << XEN_SHIFT, /* time directly in nanoseconds */
++ .shift = XEN_SHIFT,
++ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
++ .resume = xen_clocksource_resume,
++};
++
++static void init_missing_ticks_accounting(unsigned int cpu)
++{
++ struct vcpu_register_runstate_memory_area area;
++ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
++ int rc;
++
++ memset(runstate, 0, sizeof(*runstate));
++
++ area.addr.v = runstate;
++ rc = HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
++ WARN_ON(rc && rc != -ENOSYS);
++
++ per_cpu(processed_blocked_time, cpu) =
++ runstate->time[RUNSTATE_blocked];
++ per_cpu(processed_stolen_time, cpu) =
++ runstate->time[RUNSTATE_runnable] +
++ runstate->time[RUNSTATE_offline];
++}
++
++unsigned long xen_read_persistent_clock(void)
++{
++ const shared_info_t *s = HYPERVISOR_shared_info;
++ u32 version, sec, nsec;
++ u64 delta;
++
++ do {
++ version = s->wc_version;
++ rmb();
++ sec = s->wc_sec;
++ nsec = s->wc_nsec;
++ rmb();
++ } while ((s->wc_version & 1) | (version ^ s->wc_version));
++
++ delta = local_clock() + (u64)sec * NSEC_PER_SEC + nsec;
++ do_div(delta, NSEC_PER_SEC);
++
++ return delta;
++}
++
++int xen_update_persistent_clock(void)
++{
++ if (!is_initial_xendomain())
++ return -1;
++ mod_timer(&sync_xen_wallclock_timer, jiffies + 1);
++ return 0;
++}
++
++/* Dynamically-mapped IRQ. */
++DEFINE_PER_CPU(int, timer_irq);
++
++static void setup_cpu0_timer_irq(void)
++{
++ per_cpu(timer_irq, 0) =
++ bind_virq_to_irqhandler(
++ VIRQ_TIMER,
++ 0,
++ timer_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ "timer0",
++ NULL);
++ BUG_ON(per_cpu(timer_irq, 0) < 0);
++}
++
++static struct vcpu_set_periodic_timer xen_set_periodic_tick = {
++ .period_ns = NS_PER_TICK
++};
++
++void __init time_init(void)
++{
++ init_cpu_khz();
++ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
++ cpu_khz / 1000, cpu_khz % 1000);
++
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, 0,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
++ }
++
++ get_time_values_from_xen(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);
++
++ clocksource_register(&clocksource_xen);
++
++ update_wallclock();
++
++#ifndef CONFIG_X86_64
++ use_tsc_delay();
++#endif
++
++ /* Cannot request_irq() until kmem is initialised. */
++ late_time_init = setup_cpu0_timer_irq;
++}
++
++/* Convert jiffies to system time. */
++u64 jiffies_to_st(unsigned long j)
++{
++ unsigned long seq;
++ long delta;
++ u64 st;
++
++ do {
++ seq = read_seqbegin(&xtime_lock);
++ delta = j - jiffies;
++ if (delta < 1) {
++ /* Triggers in some wrap-around cases, but that's okay:
++ * we just end up with a shorter timeout. */
++ st = processed_system_time + NS_PER_TICK;
++ } else if (((unsigned long)delta >> (BITS_PER_LONG-3)) != 0) {
++ /* Very long timeout means there is no pending timer.
++ * We indicate this to Xen by passing zero timeout. */
++ st = 0;
++ } else {
++ st = processed_system_time + delta * (u64)NS_PER_TICK;
++ }
++ } while (read_seqretry(&xtime_lock, seq));
++
++ return st;
++}
++EXPORT_SYMBOL(jiffies_to_st);
++
++/*
++ * stop_hz_timer / start_hz_timer - enter/exit 'tickless mode' on an idle cpu
++ * These functions are based on implementations from arch/s390/kernel/time.c
++ */
++static void stop_hz_timer(void)
++{
++ struct vcpu_set_singleshot_timer singleshot;
++ unsigned int cpu = smp_processor_id();
++ unsigned long j;
++ int rc;
++
++ 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. */
++
++ smp_mb();
++
++ /* Leave ourselves in tick mode if rcu or softirq or timer pending. */
++ if (rcu_needs_cpu(cpu) || local_softirq_pending() ||
++ (j = next_timer_interrupt(), time_before_eq(j, jiffies))) {
++ cpu_clear(cpu, nohz_cpu_mask);
++ j = jiffies + 1;
++ }
++
++ singleshot.timeout_abs_ns = jiffies_to_st(j) + NS_PER_TICK/2;
++ singleshot.flags = 0;
++ rc = HYPERVISOR_vcpu_op(VCPUOP_set_singleshot_timer, cpu, &singleshot);
++#if CONFIG_XEN_COMPAT <= 0x030004
++ if (rc) {
++ BUG_ON(rc != -ENOSYS);
++ rc = HYPERVISOR_set_timer_op(singleshot.timeout_abs_ns);
++ }
++#endif
++ BUG_ON(rc);
++}
++
++static void start_hz_timer(void)
++{
++ cpu_clear(smp_processor_id(), nohz_cpu_mask);
++}
++
++void xen_safe_halt(void)
++{
++ stop_hz_timer();
++ /* Blocking includes an implicit local_irq_enable(). */
++ HYPERVISOR_block();
++ start_hz_timer();
++}
++EXPORT_SYMBOL(xen_safe_halt);
++
++void xen_halt(void)
++{
++ if (irqs_disabled())
++ VOID(HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL));
++}
++EXPORT_SYMBOL(xen_halt);
++
++/* No locking required. Interrupts are disabled on all CPUs. */
++void time_resume(void)
++{
++ unsigned int cpu;
++
++ init_cpu_khz();
++
++ for_each_online_cpu(cpu) {
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
++ }
++ get_time_values_from_xen(cpu);
++ per_cpu(processed_system_time, cpu) =
++ per_cpu(shadow_time, 0).system_timestamp;
++ init_missing_ticks_accounting(cpu);
++ }
++
++ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
++
++ update_wallclock();
++}
++
++#ifdef CONFIG_SMP
++static char timer_name[NR_CPUS][15];
++
++int __cpuinit local_setup_timer(unsigned int cpu)
++{
++ int seq, irq;
++
++ BUG_ON(cpu == 0);
++
++ switch (HYPERVISOR_vcpu_op(VCPUOP_set_periodic_timer, cpu,
++ &xen_set_periodic_tick)) {
++ case 0:
++#if CONFIG_XEN_COMPAT <= 0x030004
++ case -ENOSYS:
++#endif
++ break;
++ default:
++ BUG();
++ }
++
++ 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%u", cpu);
++ irq = bind_virq_to_irqhandler(VIRQ_TIMER,
++ cpu,
++ timer_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ timer_name[cpu],
++ NULL);
++ if (irq < 0)
++ return irq;
++ per_cpu(timer_irq, cpu) = irq;
++
++ return 0;
++}
++
++void __cpuexit local_teardown_timer(unsigned int cpu)
++{
++ BUG_ON(cpu == 0);
++ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
++}
++#endif
++
++#ifdef CONFIG_CPU_FREQ
++static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
++ void *data)
++{
++ struct cpufreq_freqs *freq = data;
++ struct xen_platform_op op;
++
++ if (cpu_has(&cpu_data(freq->cpu), X86_FEATURE_CONSTANT_TSC))
++ return 0;
++
++ if (val == CPUFREQ_PRECHANGE)
++ return 0;
++
++ op.cmd = XENPF_change_freq;
++ op.u.change_freq.flags = 0;
++ op.u.change_freq.cpu = freq->cpu;
++ op.u.change_freq.freq = (u64)freq->new * 1000;
++ WARN_ON(HYPERVISOR_platform_op(&op));
++
++ return 0;
++}
++
++static struct notifier_block time_cpufreq_notifier_block = {
++ .notifier_call = time_cpufreq_notifier
++};
++
++static int __init cpufreq_time_setup(void)
++{
++ if (!cpufreq_register_notifier(&time_cpufreq_notifier_block,
++ CPUFREQ_TRANSITION_NOTIFIER)) {
++ printk(KERN_ERR "failed to set up cpufreq notifier\n");
++ return -ENODEV;
++ }
++ return 0;
++}
++
++core_initcall(cpufreq_time_setup);
++#endif
++
++/*
++ * /proc/sys/xen: This really belongs in another file. It can stay here for
++ * now however.
++ */
++static ctl_table xen_subtable[] = {
++ {
++ .ctl_name = CTL_XEN_INDEPENDENT_WALLCLOCK,
++ .procname = "independent_wallclock",
++ .data = &independent_wallclock,
++ .maxlen = sizeof(independent_wallclock),
++ .mode = 0644,
++ .strategy = sysctl_data,
++ .proc_handler = proc_dointvec
++ },
++ {
++ .ctl_name = CTL_XEN_PERMITTED_CLOCK_JITTER,
++ .procname = "permitted_clock_jitter",
++ .data = &permitted_clock_jitter,
++ .maxlen = sizeof(permitted_clock_jitter),
++ .mode = 0644,
++ .strategy = sysctl_data,
++ .proc_handler = proc_doulongvec_minmax
++ },
++ { }
++};
++static ctl_table xen_table[] = {
++ {
++ .ctl_name = CTL_XEN,
++ .procname = "xen",
++ .mode = 0555,
++ .child = xen_subtable
++ },
++ { }
++};
++static int __init xen_sysctl_init(void)
++{
++ (void)register_sysctl_table(xen_table);
++ return 0;
++}
++__initcall(xen_sysctl_init);
+diff --git a/arch/x86/kernel/traps_32-xen.c b/arch/x86/kernel/traps_32-xen.c
+new file mode 100644
+index 0000000..c174679
+--- /dev/null
++++ b/arch/x86/kernel/traps_32-xen.c
+@@ -0,0 +1,1240 @@
++/*
++ * 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/interrupt.h>
++#include <linux/kallsyms.h>
++#include <linux/spinlock.h>
++#include <linux/highmem.h>
++#include <linux/kprobes.h>
++#include <linux/uaccess.h>
++#include <linux/utsname.h>
++#include <linux/kdebug.h>
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/ptrace.h>
++#include <linux/string.h>
++#include <linux/unwind.h>
++#include <linux/delay.h>
++#include <linux/errno.h>
++#include <linux/kexec.h>
++#include <linux/sched.h>
++#include <linux/timer.h>
++#include <linux/init.h>
++#include <linux/bug.h>
++#include <linux/nmi.h>
++#include <linux/mm.h>
++
++#ifdef CONFIG_EISA
++#include <linux/ioport.h>
++#include <linux/eisa.h>
++#endif
++
++#ifdef CONFIG_MCA
++#include <linux/mca.h>
++#endif
++
++#if defined(CONFIG_EDAC)
++#include <linux/edac.h>
++#endif
++
++#include <asm/arch_hooks.h>
++#include <asm/stacktrace.h>
++#include <asm/processor.h>
++#include <asm/debugreg.h>
++#include <asm/atomic.h>
++#include <asm/system.h>
++#include <asm/unwind.h>
++#include <asm/desc.h>
++#include <asm/i387.h>
++#include <asm/nmi.h>
++#include <asm/smp.h>
++#include <asm/io.h>
++
++#include "mach_traps.h"
++
++int panic_on_unrecovered_nmi;
++
++#ifndef CONFIG_XEN
++DECLARE_BITMAP(used_vectors, NR_VECTORS);
++EXPORT_SYMBOL_GPL(used_vectors);
++#endif
++
++asmlinkage int system_call(void);
++
++/* Do we ignore FPU interrupts ? */
++char ignore_fpu_irq;
++
++#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.
++ */
++gate_desc idt_table[256]
++ __attribute__((__section__(".data.idt"))) = { { { { 0, 0 } } }, };
++#endif
++
++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);
++
++int kstack_depth_to_print = 24;
++static unsigned int code_bytes = 64;
++
++void printk_address(unsigned long address, int reliable)
++{
++#ifdef CONFIG_KALLSYMS
++ char namebuf[KSYM_NAME_LEN];
++ unsigned long offset = 0;
++ unsigned long symsize;
++ const char *symname;
++ char reliab[4] = "";
++ char *delim = ":";
++ char *modname;
++
++ symname = kallsyms_lookup(address, &symsize, &offset,
++ &modname, namebuf);
++ if (!symname) {
++ printk(" [<%08lx>]\n", address);
++ return;
++ }
++ if (!reliable)
++ strcpy(reliab, "? ");
++
++ if (!modname)
++ modname = delim = "";
++ printk(" [<%08lx>] %s%s%s%s%s+0x%lx/0x%lx\n",
++ address, reliab, delim, modname, delim, symname, offset, symsize);
++#else
++ printk(" [<%08lx>]\n", address);
++#endif
++}
++
++static inline int valid_stack_ptr(struct thread_info *tinfo, void *p, unsigned size)
++{
++ return p > (void *)tinfo &&
++ p <= (void *)tinfo + THREAD_SIZE - size;
++}
++
++/* The form of the top of the frame on the stack */
++struct stack_frame {
++ struct stack_frame *next_frame;
++ unsigned long return_address;
++};
++
++static inline unsigned long
++print_context_stack(struct thread_info *tinfo,
++ unsigned long *stack, unsigned long bp,
++ const struct stacktrace_ops *ops, void *data)
++{
++ struct stack_frame *frame = (struct stack_frame *)bp;
++
++ while (valid_stack_ptr(tinfo, stack, sizeof(*stack))) {
++ unsigned long addr;
++
++ addr = *stack;
++ if (__kernel_text_address(addr)) {
++ if ((unsigned long) stack == bp + 4) {
++ ops->address(data, addr, 1);
++ frame = frame->next_frame;
++ bp = (unsigned long) frame;
++ } else {
++ ops->address(data, addr, bp == 0);
++ }
++ }
++ stack++;
++ }
++ return bp;
++}
++
++#define MSG(msg) ops->warning(data, msg)
++
++void dump_trace(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *stack, unsigned long bp,
++ const struct stacktrace_ops *ops, void *data)
++{
++ if (!task)
++ task = current;
++
++ if (!stack) {
++ unsigned long dummy;
++
++ stack = &dummy;
++ if (task != current)
++ stack = (unsigned long *)task->thread.sp;
++ }
++
++#ifdef CONFIG_FRAME_POINTER
++ if (!bp) {
++ if (task == current) {
++ /* Grab bp right from our regs */
++ asm("movl %%ebp, %0" : "=r" (bp) :);
++ } else {
++ /* bp is the last reg pushed by switch_to */
++ bp = *(unsigned long *) task->thread.sp;
++ }
++ }
++#endif
++
++ while (1) {
++ struct thread_info *context;
++
++ context = (struct thread_info *)
++ ((unsigned long)stack & (~(THREAD_SIZE - 1)));
++ bp = print_context_stack(context, stack, bp, ops, data);
++ /*
++ * Should be after the line below, but somewhere
++ * in early boot context comes out corrupted and we
++ * can't reference it:
++ */
++ if (ops->stack(data, "IRQ") < 0)
++ break;
++ stack = (unsigned long *)context->previous_esp;
++ if (!stack)
++ break;
++ touch_nmi_watchdog();
++ }
++}
++EXPORT_SYMBOL(dump_trace);
++
++static void
++print_trace_warning_symbol(void *data, char *msg, unsigned long symbol)
++{
++ printk(data);
++ print_symbol(msg, symbol);
++ printk("\n");
++}
++
++static void print_trace_warning(void *data, char *msg)
++{
++ printk("%s%s\n", (char *)data, msg);
++}
++
++static int print_trace_stack(void *data, char *name)
++{
++ return 0;
++}
++
++/*
++ * Print one address/symbol entries per line.
++ */
++static void print_trace_address(void *data, unsigned long addr, int reliable)
++{
++ printk("%s [<%08lx>] ", (char *)data, addr);
++ if (!reliable)
++ printk("? ");
++ print_symbol("%s\n", addr);
++ touch_nmi_watchdog();
++}
++
++static const struct stacktrace_ops print_trace_ops = {
++ .warning = print_trace_warning,
++ .warning_symbol = print_trace_warning_symbol,
++ .stack = print_trace_stack,
++ .address = print_trace_address,
++};
++
++static void
++show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *stack, unsigned long bp, char *log_lvl)
++{
++ dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl);
++ printk("%s =======================\n", log_lvl);
++}
++
++void show_trace(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *stack, unsigned long bp)
++{
++ show_trace_log_lvl(task, regs, stack, bp, "");
++}
++
++static void
++show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
++ unsigned long *sp, unsigned long bp, char *log_lvl)
++{
++ unsigned long *stack;
++ int i;
++
++ if (sp == NULL) {
++ if (task)
++ sp = (unsigned long *)task->thread.sp;
++ else
++ sp = (unsigned long *)&sp;
++ }
++
++ stack = sp;
++ 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, sp, bp, log_lvl);
++}
++
++void show_stack(struct task_struct *task, unsigned long *sp)
++{
++ printk(" ");
++ show_stack_log_lvl(task, NULL, sp, 0, "");
++}
++
++/*
++ * The architecture-independent dump_stack generator
++ */
++void dump_stack(void)
++{
++ unsigned long bp = 0;
++ unsigned long stack;
++
++#ifdef CONFIG_FRAME_POINTER
++ if (!bp)
++ asm("movl %%ebp, %0" : "=r" (bp):);
++#endif
++
++ 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);
++
++ show_trace(current, NULL, &stack, bp);
++}
++
++EXPORT_SYMBOL(dump_stack);
++
++void show_registers(struct pt_regs *regs)
++{
++ int i;
++
++ print_modules();
++ __show_registers(regs, 0);
++
++ printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)",
++ TASK_COMM_LEN, current->comm, task_pid_nr(current),
++ current_thread_info(), current, task_thread_info(current));
++ /*
++ * When in-kernel, we also print out the stack and code at the
++ * time of the fault..
++ */
++ if (!user_mode_vm(regs)) {
++ unsigned int code_prologue = code_bytes * 43 / 64;
++ unsigned int code_len = code_bytes;
++ unsigned char c;
++ u8 *ip;
++
++ printk("\n" KERN_EMERG "Stack: ");
++ show_stack_log_lvl(NULL, regs, ®s->sp, 0, KERN_EMERG);
++
++ printk(KERN_EMERG "Code: ");
++
++ ip = (u8 *)regs->ip - code_prologue;
++ if (ip < (u8 *)PAGE_OFFSET ||
++ probe_kernel_address(ip, c)) {
++ /* try starting at EIP */
++ ip = (u8 *)regs->ip;
++ code_len = code_len - code_prologue + 1;
++ }
++ for (i = 0; i < code_len; i++, ip++) {
++ if (ip < (u8 *)PAGE_OFFSET ||
++ probe_kernel_address(ip, c)) {
++ printk(" Bad EIP value.");
++ break;
++ }
++ if (ip == (u8 *)regs->ip)
++ printk("<%02x> ", c);
++ else
++ printk("%02x ", c);
++ }
++ }
++ printk("\n");
++}
++
++int is_valid_bugaddr(unsigned long ip)
++{
++ unsigned short ud2;
++
++ if (ip < PAGE_OFFSET)
++ return 0;
++ if (probe_kernel_address((unsigned short *)ip, ud2))
++ return 0;
++
++ return ud2 == 0x0b0f;
++}
++
++static int die_counter;
++
++int __kprobes __die(const char *str, struct pt_regs *regs, long err)
++{
++ unsigned short ss;
++ unsigned long sp;
++
++ printk(KERN_EMERG "%s: %04lx [#%d] ", str, err & 0xffff, ++die_counter);
++#ifdef CONFIG_PREEMPT
++ printk("PREEMPT ");
++#endif
++#ifdef CONFIG_SMP
++ printk("SMP ");
++#endif
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ printk("DEBUG_PAGEALLOC");
++#endif
++ printk("\n");
++
++ 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 */
++ sp = (unsigned long) (®s->sp);
++ savesegment(ss, ss);
++ if (user_mode(regs)) {
++ sp = regs->sp;
++ ss = regs->ss & 0xffff;
++ }
++ printk(KERN_EMERG "EIP: [<%08lx>] ", regs->ip);
++ print_symbol("%s", regs->ip);
++ printk(" SS:ESP %04x:%08lx\n", ss, sp);
++
++ return 0;
++ }
++
++ return 1;
++}
++
++/*
++ * This is gone through when something in the kernel has done something bad
++ * and is about to be terminated:
++ */
++void die(const char *str, struct pt_regs *regs, long err)
++{
++ static struct {
++ raw_spinlock_t lock;
++ u32 lock_owner;
++ int lock_owner_depth;
++ } die = {
++ .lock = __RAW_SPIN_LOCK_UNLOCKED,
++ .lock_owner = -1,
++ .lock_owner_depth = 0
++ };
++ unsigned long flags;
++
++ oops_enter();
++
++ if (die.lock_owner != raw_smp_processor_id()) {
++ console_verbose();
++ raw_local_irq_save(flags);
++ __raw_spin_lock(&die.lock);
++ die.lock_owner = smp_processor_id();
++ die.lock_owner_depth = 0;
++ bust_spinlocks(1);
++ } else {
++ raw_local_irq_save(flags);
++ }
++
++ if (++die.lock_owner_depth < 3) {
++ report_bug(regs->ip, regs);
++
++ if (__die(str, regs, err))
++ regs = NULL;
++ } else {
++ printk(KERN_EMERG "Recursive die() failure, output suppressed\n");
++ }
++
++ bust_spinlocks(0);
++ die.lock_owner = -1;
++ add_taint(TAINT_DIE);
++ __raw_spin_unlock(&die.lock);
++ raw_local_irq_restore(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);
++}
++
++static inline void
++die_if_kernel(const char *str, struct pt_regs *regs, long err)
++{
++ if (!user_mode_vm(regs))
++ die(str, regs, err);
++}
++
++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;
++
++ if (regs->flags & X86_VM_MASK) {
++ if (vm86)
++ goto vm86_trap;
++ goto trap_signal;
++ }
++
++ if (!user_mode(regs))
++ goto kernel_trap;
++
++trap_signal:
++ /*
++ * We want error_code and trap_no set for userspace faults and
++ * kernelspace faults which result in die(), but not
++ * kernelspace faults which are fixed up. die() gives the
++ * process no chance to handle the signal and notice the
++ * kernel fault information, so that won't result in polluting
++ * the information about previously queued, but not yet
++ * delivered, faults. See also do_general_protection below.
++ */
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
++
++ if (info)
++ force_sig_info(signr, info, tsk);
++ else
++ force_sig(signr, tsk);
++ return;
++
++kernel_trap:
++ if (!fixup_exception(regs)) {
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
++ die(str, regs, error_code);
++ }
++ return;
++
++vm86_trap:
++ if (handle_vm86_trap((struct kernel_vm86_regs *) regs,
++ error_code, trapnr))
++ goto trap_signal;
++ return;
++}
++
++#define DO_ERROR(trapnr, signr, str, name) \
++void do_##name(struct pt_regs *regs, long error_code) \
++{ \
++ trace_hardirqs_fixup(); \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \
++}
++
++#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr, irq) \
++void do_##name(struct pt_regs *regs, long error_code) \
++{ \
++ siginfo_t info; \
++ if (irq) \
++ local_irq_enable(); \
++ 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); \
++}
++
++#define DO_VM86_ERROR(trapnr, signr, str, name) \
++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); \
++}
++
++#define DO_VM86_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
++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; \
++ trace_hardirqs_fixup(); \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ do_trap(trapnr, signr, str, 1, regs, error_code, &info); \
++}
++
++DO_VM86_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip)
++#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->ip, 0)
++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, 0)
++DO_ERROR_INFO(32, SIGILL, "iret exception", iret_error, ILL_BADSTK, 0, 1)
++
++void __kprobes do_general_protection(struct pt_regs * regs,
++ long error_code)
++{
++ struct thread_struct *thread;
++
++ thread = ¤t->thread;
++
++ if (regs->flags & X86_VM_MASK)
++ goto gp_in_vm86;
++
++ if (!user_mode(regs))
++ goto gp_in_kernel;
++
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
++
++ if (show_unhandled_signals && unhandled_signal(current, SIGSEGV) &&
++ printk_ratelimit()) {
++ printk(KERN_INFO
++ "%s[%d] general protection ip:%lx sp:%lx error:%lx",
++ current->comm, task_pid_nr(current),
++ regs->ip, regs->sp, error_code);
++ print_vma_addr(" in ", regs->ip);
++ printk("\n");
++ }
++
++ 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)) {
++ current->thread.error_code = error_code;
++ current->thread.trap_no = 13;
++ if (notify_die(DIE_GPF, "general protection fault", regs,
++ error_code, 13, SIGSEGV) == NOTIFY_STOP)
++ return;
++ die("general protection fault", regs, error_code);
++ }
++}
++
++static notrace __kprobes void
++mem_parity_error(unsigned char reason, struct pt_regs *regs)
++{
++ 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 defined(CONFIG_EDAC)
++ if (edac_handler_set()) {
++ edac_atomic_assert_error();
++ return;
++ }
++#endif
++
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
++
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
++
++ /* Clear and disable the memory parity error line. */
++ clear_mem_error(reason);
++}
++
++static notrace __kprobes void
++io_check_error(unsigned char reason, struct pt_regs *regs)
++{
++ printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n");
++ show_registers(regs);
++
++ /* Re-enable the IOCK line, wait for a few seconds */
++ clear_io_check_error(reason);
++}
++
++static notrace __kprobes void
++unknown_nmi_error(unsigned char reason, struct pt_regs *regs)
++{
++ if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
++ return;
++#ifdef CONFIG_MCA
++ /*
++ * Might actually be able to figure out what the guilty party
++ * is:
++ */
++ if (MCA_bus) {
++ mca_handle_nmi();
++ return;
++ }
++#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");
++
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
++}
++
++static DEFINE_SPINLOCK(nmi_print_lock);
++
++void notrace __kprobes die_nmi(struct pt_regs *regs, const char *msg)
++{
++ 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, ip %08lx, registers:\n",
++ smp_processor_id(), regs->ip);
++ show_registers(regs);
++ console_silent();
++ spin_unlock(&nmi_print_lock);
++ bust_spinlocks(0);
++
++ /*
++ * If we are in kernel we are probably nested up pretty bad
++ * and might aswell get out now while we still can:
++ */
++ if (!user_mode_vm(regs)) {
++ current->thread.trap_no = 2;
++ crash_kexec(regs);
++ }
++
++ do_exit(SIGSEGV);
++}
++
++static notrace __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
++#ifndef CONFIG_XEN
++ /*
++ * Ok, so this is none of the documented NMI sources,
++ * so it must be the NMI watchdog.
++ */
++ if (nmi_watchdog_tick(regs, reason))
++ return;
++#endif
++ if (!do_nmi_callback(regs, smp_processor_id()))
++ unknown_nmi_error(reason, regs);
++#else
++ unknown_nmi_error(reason, regs);
++#endif
++
++ 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 int ignore_nmis;
++
++notrace __kprobes void do_nmi(struct pt_regs *regs, long error_code)
++{
++ int cpu;
++
++ nmi_enter();
++
++ cpu = smp_processor_id();
++
++ ++nmi_count(cpu);
++
++ if (!ignore_nmis)
++ default_do_nmi(regs);
++
++ nmi_exit();
++}
++
++void stop_nmi(void)
++{
++ acpi_nmi_disable();
++ ignore_nmis++;
++}
++
++void restart_nmi(void)
++{
++ ignore_nmis--;
++ acpi_nmi_enable();
++}
++
++#ifdef CONFIG_KPROBES
++void __kprobes do_int3(struct pt_regs *regs, long error_code)
++{
++ trace_hardirqs_fixup();
++
++ 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
++
++/*
++ * 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)
++ */
++void __kprobes do_debug(struct pt_regs *regs, long error_code)
++{
++ struct task_struct *tsk = current;
++ unsigned int condition;
++
++ trace_hardirqs_fixup();
++
++ get_debugreg(condition, 6);
++
++ /*
++ * The processor cleared BTF, so don't mark that we need it set.
++ */
++ clear_tsk_thread_flag(tsk, TIF_DEBUGCTLMSR);
++ tsk->thread.debugctlmsr = 0;
++
++ 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->flags & X86_EFLAGS_IF)
++ local_irq_enable();
++
++ /* Mask out spurious debug traps due to lazy DR7 setting */
++ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
++ if (!tsk->thread.debugreg7)
++ goto clear_dr7;
++ }
++
++ if (regs->flags & X86_VM_MASK)
++ goto debug_vm86;
++
++ /* Save debug status register where ptrace can see it */
++ tsk->thread.debugreg6 = condition;
++
++ /*
++ * Single-stepping through TF: make sure we ignore any events in
++ * kernel space (but re-enable TF when returning to user mode).
++ */
++ if (condition & DR_STEP) {
++ /*
++ * We already checked v86 mode above, so we can
++ * check for kernel mode by just checking the CPL
++ * of CS.
++ */
++ if (!user_mode(regs))
++ goto clear_TF_reenable;
++ }
++
++ /* Ok, finally something we can handle */
++ send_sigtrap(tsk, regs, error_code);
++
++ /*
++ * Disable additional traps. They'll be re-enabled when
++ * the signal is delivered.
++ */
++clear_dr7:
++ set_debugreg(0, 7);
++ return;
++
++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->flags &= ~X86_EFLAGS_TF;
++ return;
++}
++
++/*
++ * 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 *ip)
++{
++ struct task_struct *task;
++ unsigned short cwd;
++ unsigned short swd;
++ siginfo_t info;
++
++ /*
++ * 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 = ip;
++ /*
++ * (~cwd & swd) will mask out exceptions that are not set to unmasked
++ * status. 0x3f is the exception bits in these regs, 0x200 is the
++ * C1 reg you need in case of a stack fault, 0x040 is the stack
++ * fault bit. We should only be taking one exception at a time,
++ * so if this combination doesn't produce any single exception,
++ * then we have a bad program that isn't syncronizing its FPU usage
++ * and it will suffer the consequences since we won't be able to
++ * fully reproduce the context of the exception
++ */
++ cwd = get_fpu_cwd(task);
++ swd = get_fpu_swd(task);
++ switch (swd & ~cwd & 0x3f) {
++ case 0x000: /* No unmasked exception */
++ return;
++ default: /* Multiple exceptions */
++ break;
++ case 0x001: /* Invalid Op */
++ /*
++ * swd & 0x240 == 0x040: Stack Underflow
++ * swd & 0x240 == 0x240: Stack Overflow
++ * User must clear the SF bit (0x40) if set
++ */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++void do_coprocessor_error(struct pt_regs *regs, long error_code)
++{
++ ignore_fpu_irq = 1;
++ math_error((void __user *)regs->ip);
++}
++
++static void simd_math_error(void __user *ip)
++{
++ struct task_struct *task;
++ unsigned short mxcsr;
++ siginfo_t info;
++
++ /*
++ * 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 = ip;
++ /*
++ * The SIMD FPU exceptions are handled a little differently, as there
++ * is only a single status/control register. Thus, to determine which
++ * unmasked exception was caught we must mask the exception mask bits
++ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
++ */
++ mxcsr = get_fpu_mxcsr(task);
++ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++void do_simd_coprocessor_error(struct pt_regs *regs, long error_code)
++{
++ if (cpu_has_xmm) {
++ /* Handle SIMD FPU exceptions on PIII+ processors. */
++ ignore_fpu_irq = 1;
++ simd_math_error((void __user *)regs->ip);
++ return;
++ }
++ /*
++ * Handle strange cache flush from user space exception
++ * in all other cases. This is undocumented behaviour.
++ */
++ if (regs->flags & X86_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);
++}
++
++#ifndef CONFIG_XEN
++void do_spurious_interrupt_bug(struct pt_regs *regs, long error_code)
++{
++#if 0
++ /* No need to warn about this any longer. */
++ printk(KERN_INFO "Ignoring P6 Local APIC Spurious Interrupt Bug...\n");
++#endif
++}
++
++unsigned long patch_espfix_desc(unsigned long uesp, unsigned long kesp)
++{
++ struct desc_struct *gdt = __get_cpu_var(gdt_page).gdt;
++ 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
++
++/*
++ * 'math_state_restore()' saves the current math information in the
++ * old math state array, and gets the new ones from the current task
++ *
++ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
++ * Don't touch unless you *really* know how it works.
++ *
++ * Must be called with kernel preemption disabled (in this case,
++ * local interrupts are disabled at the call-site in entry.S).
++ */
++asmlinkage void math_state_restore(void)
++{
++ struct thread_info *thread = current_thread_info();
++ struct task_struct *tsk = thread->task;
++
++ if (!tsk_used_math(tsk)) {
++ local_irq_enable();
++ /*
++ * does a slab alloc which can sleep
++ */
++ if (init_fpu(tsk)) {
++ /*
++ * ran out of memory!
++ */
++ do_group_exit(SIGKILL);
++ return;
++ }
++ local_irq_disable();
++ }
++
++ /* NB. 'clts' is done for us by Xen during virtual trap. */
++ restore_fpu(tsk);
++ thread->status |= TS_USEDFPU; /* So we fnsave on switch_to() */
++ tsk->fpu_counter++;
++}
++EXPORT_SYMBOL_GPL(math_state_restore);
++
++#ifndef CONFIG_MATH_EMULATION
++
++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();
++}
++
++#endif /* CONFIG_MATH_EMULATION */
++
++/*
++ * 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 const trap_info_t __cpuinitconst trap_table[] = {
++ { 0, 0, __KERNEL_CS, (unsigned long)divide_error },
++ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
++ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
++ { 4, 3, __KERNEL_CS, (unsigned long)overflow },
++ { 5, 0, __KERNEL_CS, (unsigned long)bounds },
++ { 6, 0, __KERNEL_CS, (unsigned long)invalid_op },
++ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
++ { 9, 0, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun },
++ { 10, 0, __KERNEL_CS, (unsigned long)invalid_TSS },
++ { 11, 0, __KERNEL_CS, (unsigned long)segment_not_present },
++ { 12, 0, __KERNEL_CS, (unsigned long)stack_segment },
++ { 13, 0, __KERNEL_CS, (unsigned long)general_protection },
++ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
++ { 15, 0, __KERNEL_CS, (unsigned long)fixup_4gb_segment },
++ { 16, 0, __KERNEL_CS, (unsigned long)coprocessor_error },
++ { 17, 0, __KERNEL_CS, (unsigned long)alignment_check },
++#ifdef CONFIG_X86_MCE
++ { 18, 0, __KERNEL_CS, (unsigned long)machine_check },
++#endif
++ { 19, 0, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
++ { SYSCALL_VECTOR, 3, __KERNEL_CS, (unsigned long)system_call },
++ { 0, 0, 0, 0 }
++};
++
++void __init trap_init(void)
++{
++ int ret;
++
++ ret = HYPERVISOR_set_trap_table(trap_table);
++ if (ret)
++ printk("HYPERVISOR_set_trap_table failed: error %d\n", ret);
++
++ if (cpu_has_fxsr) {
++ 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");
++ }
++
++ init_thread_xstate();
++ /*
++ * Should be a barrier for any external CPU state:
++ */
++ cpu_init();
++}
++
++void __cpuinit smp_trap_init(trap_info_t *trap_ctxt)
++{
++ const trap_info_t *t = trap_table;
++
++ for (t = trap_table; t->address; t++) {
++ trap_ctxt[t->vector].flags = t->flags;
++ trap_ctxt[t->vector].cs = t->cs;
++ trap_ctxt[t->vector].address = t->address;
++ }
++}
++
++static int __init kstack_setup(char *s)
++{
++ kstack_depth_to_print = simple_strtoul(s, NULL, 0);
++
++ return 1;
++}
++__setup("kstack=", kstack_setup);
++
++static int __init code_bytes_setup(char *s)
++{
++ code_bytes = simple_strtoul(s, NULL, 0);
++ if (code_bytes > 8192)
++ code_bytes = 8192;
++
++ return 1;
++}
++__setup("code_bytes=", code_bytes_setup);
+diff --git a/arch/x86/kernel/traps_64-xen.c b/arch/x86/kernel/traps_64-xen.c
+new file mode 100644
+index 0000000..f2c5672
+--- /dev/null
++++ b/arch/x86/kernel/traps_64-xen.c
+@@ -0,0 +1,1244 @@
++/*
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
++ *
++ * Pentium III FXSR, SSE support
++ * Gareth Hughes <gareth at valinux.com>, May 2000
++ */
++
++/*
++ * 'Traps.c' handles hardware traps and faults after we have saved some
++ * state in 'entry.S'.
++ */
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/ptrace.h>
++#include <linux/timer.h>
++#include <linux/mm.h>
++#include <linux/init.h>
++#include <linux/delay.h>
++#include <linux/spinlock.h>
++#include <linux/interrupt.h>
++#include <linux/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 <linux/kdebug.h>
++#include <linux/utsname.h>
++
++#include <mach_traps.h>
++
++#if defined(CONFIG_EDAC)
++#include <linux/edac.h>
++#endif
++
++#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/processor.h>
++#include <asm/unwind.h>
++#include <asm/smp.h>
++#include <asm/pgalloc.h>
++#include <asm/pda.h>
++#include <asm/proto.h>
++#include <asm/nmi.h>
++#include <asm/stacktrace.h>
++
++asmlinkage void divide_error(void);
++asmlinkage void debug(void);
++asmlinkage void nmi(void);
++asmlinkage void int3(void);
++asmlinkage void overflow(void);
++asmlinkage void bounds(void);
++asmlinkage void invalid_op(void);
++asmlinkage void device_not_available(void);
++asmlinkage void double_fault(void);
++asmlinkage void coprocessor_segment_overrun(void);
++asmlinkage void invalid_TSS(void);
++asmlinkage void segment_not_present(void);
++asmlinkage void stack_segment(void);
++asmlinkage void general_protection(void);
++asmlinkage void page_fault(void);
++asmlinkage void coprocessor_error(void);
++asmlinkage void simd_coprocessor_error(void);
++asmlinkage void reserved(void);
++asmlinkage void alignment_check(void);
++asmlinkage void machine_check(void);
++asmlinkage void spurious_interrupt_bug(void);
++
++static unsigned int code_bytes = 64;
++
++static inline void conditional_sti(struct pt_regs *regs)
++{
++ if (regs->flags & X86_EFLAGS_IF)
++ local_irq_enable();
++}
++
++static inline void preempt_conditional_sti(struct pt_regs *regs)
++{
++ inc_preempt_count();
++ if (regs->flags & X86_EFLAGS_IF)
++ local_irq_enable();
++}
++
++static inline void preempt_conditional_cli(struct pt_regs *regs)
++{
++ if (regs->flags & X86_EFLAGS_IF)
++ local_irq_disable();
++ /* Make sure to not schedule here because we could be running
++ on an exception stack. */
++ dec_preempt_count();
++}
++
++int kstack_depth_to_print = 12;
++
++void printk_address(unsigned long address, int reliable)
++{
++#ifdef CONFIG_KALLSYMS
++ unsigned long offset = 0, symsize;
++ const char *symname;
++ char *modname;
++ char *delim = ":";
++ char namebuf[KSYM_NAME_LEN];
++ char reliab[4] = "";
++
++ symname = kallsyms_lookup(address, &symsize, &offset,
++ &modname, namebuf);
++ if (!symname) {
++ printk(" [<%016lx>]\n", address);
++ return;
++ }
++ if (!reliable)
++ strcpy(reliab, "? ");
++
++ if (!modname)
++ modname = delim = "";
++ printk(" [<%016lx>] %s%s%s%s%s+0x%lx/0x%lx\n",
++ address, reliab, delim, modname, delim, symname, offset, symsize);
++#else
++ printk(" [<%016lx>]\n", address);
++#endif
++}
++
++static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
++ unsigned *usedp, char **idp)
++{
++#ifndef CONFIG_X86_NO_TSS
++ static char ids[][8] = {
++ [DEBUG_STACK - 1] = "#DB",
++ [NMI_STACK - 1] = "NMI",
++ [DOUBLEFAULT_STACK - 1] = "#DF",
++ [STACKFAULT_STACK - 1] = "#SS",
++ [MCE_STACK - 1] = "#MC",
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ [N_EXCEPTION_STACKS ... N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
++#endif
++ };
++ unsigned k;
++
++ /*
++ * Iterate over all exception stacks, and figure out whether
++ * 'stack' is in one of them:
++ */
++ for (k = 0; k < N_EXCEPTION_STACKS; k++) {
++ unsigned long end = per_cpu(orig_ist, cpu).ist[k];
++ /*
++ * Is 'stack' above this exception frame's end?
++ * If yes then skip to the next frame.
++ */
++ if (stack >= end)
++ continue;
++ /*
++ * Is 'stack' above this exception frame's start address?
++ * If yes then we found the right frame.
++ */
++ if (stack >= end - EXCEPTION_STKSZ) {
++ /*
++ * Make sure we only iterate through an exception
++ * stack once. If it comes up for the second time
++ * then there's something wrong going on - just
++ * break out and return NULL:
++ */
++ if (*usedp & (1U << k))
++ break;
++ *usedp |= 1U << k;
++ *idp = ids[k];
++ return (unsigned long *)end;
++ }
++ /*
++ * If this is a debug stack, and if it has a larger size than
++ * the usual exception stacks, then 'stack' might still
++ * be within the lower portion of the debug stack:
++ */
++#if DEBUG_STKSZ > EXCEPTION_STKSZ
++ if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
++ unsigned j = N_EXCEPTION_STACKS - 1;
++
++ /*
++ * Black magic. A large debug stack is composed of
++ * multiple exception stack entries, which we
++ * iterate through now. Dont look:
++ */
++ do {
++ ++j;
++ end -= EXCEPTION_STKSZ;
++ ids[j][4] = '1' + (j - N_EXCEPTION_STACKS);
++ } while (stack < end - EXCEPTION_STKSZ);
++ if (*usedp & (1U << j))
++ break;
++ *usedp |= 1U << j;
++ *idp = ids[j];
++ return (unsigned long *)end;
++ }
++#endif
++ }
++#endif
++ return NULL;
++}
++
++#define MSG(txt) ops->warning(data, txt)
++
++/*
++ * x86-64 can have up to three kernel stacks:
++ * process stack
++ * interrupt stack
++ * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
++ */
++
++static inline int valid_stack_ptr(struct thread_info *tinfo,
++ void *p, unsigned int size, void *end)
++{
++ void *t = tinfo;
++ if (end) {
++ if (p < end && p >= (end-THREAD_SIZE))
++ return 1;
++ else
++ return 0;
++ }
++ return p > t && p < t + THREAD_SIZE - size;
++}
++
++/* The form of the top of the frame on the stack */
++struct stack_frame {
++ struct stack_frame *next_frame;
++ unsigned long return_address;
++};
++
++
++static inline unsigned long print_context_stack(struct thread_info *tinfo,
++ unsigned long *stack, unsigned long bp,
++ const struct stacktrace_ops *ops, void *data,
++ unsigned long *end)
++{
++ struct stack_frame *frame = (struct stack_frame *)bp;
++
++ while (valid_stack_ptr(tinfo, stack, sizeof(*stack), end)) {
++ unsigned long addr;
++
++ addr = *stack;
++ if (__kernel_text_address(addr)) {
++ if ((unsigned long) stack == bp + 8) {
++ ops->address(data, addr, 1);
++ frame = frame->next_frame;
++ bp = (unsigned long) frame;
++ } else {
++ ops->address(data, addr, bp == 0);
++ }
++ }
++ stack++;
++ }
++ return bp;
++}
++
++void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
++ unsigned long *stack, unsigned long bp,
++ const 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;
++ tinfo = task_thread_info(tsk);
++
++ if (!stack) {
++ unsigned long dummy;
++ stack = &dummy;
++ if (tsk && tsk != current)
++ stack = (unsigned long *)tsk->thread.sp;
++ }
++
++#ifdef CONFIG_FRAME_POINTER
++ if (!bp) {
++ if (tsk == current) {
++ /* Grab bp right from our regs */
++ asm("movq %%rbp, %0" : "=r" (bp):);
++ } else {
++ /* bp is the last reg pushed by switch_to */
++ bp = *(unsigned long *) tsk->thread.sp;
++ }
++ }
++#endif
++
++
++
++ /*
++ * Print function call entries in all stacks, starting at the
++ * current stack address. If the stacks consist of nested
++ * exceptions
++ */
++ for (;;) {
++ char *id;
++ unsigned long *estack_end;
++ estack_end = in_exception_stack(cpu, (unsigned long)stack,
++ &used, &id);
++
++ if (estack_end) {
++ if (ops->stack(data, id) < 0)
++ break;
++
++ bp = print_context_stack(tinfo, stack, bp, ops,
++ data, estack_end);
++ ops->stack(data, "<EOE>");
++ /*
++ * We link to the next stack via the
++ * second-to-last pointer (index -2 to end) in the
++ * exception stack:
++ */
++ stack = (unsigned long *) estack_end[-2];
++ continue;
++ }
++ if (irqstack_end) {
++ unsigned long *irqstack;
++ irqstack = irqstack_end -
++ (IRQSTACKSIZE - 64) / sizeof(*irqstack);
++
++ if (stack >= irqstack && stack < irqstack_end) {
++ if (ops->stack(data, "IRQ") < 0)
++ break;
++ bp = print_context_stack(tinfo, stack, bp,
++ ops, data, irqstack_end);
++ /*
++ * We link to the next stack (which would be
++ * the process stack normally) the last
++ * pointer (index -1 to end) in the IRQ stack:
++ */
++ stack = (unsigned long *) (irqstack_end[-1]);
++ irqstack_end = NULL;
++ ops->stack(data, "EOI");
++ continue;
++ }
++ }
++ break;
++ }
++
++ /*
++ * This handles the process stack:
++ */
++ bp = print_context_stack(tinfo, stack, bp, ops, data, NULL);
++ 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, int reliable)
++{
++ touch_nmi_watchdog();
++ printk_address(addr, reliable);
++}
++
++static const 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,
++ unsigned long bp)
++{
++ printk("\nCall Trace:\n");
++ dump_trace(tsk, regs, stack, bp, &print_trace_ops, NULL);
++ printk("\n");
++}
++
++static void
++_show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *sp,
++ unsigned long bp)
++{
++ unsigned long *stack;
++ int i;
++ 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);
++
++ // debugging aid: "show_stack(NULL, NULL);" prints the
++ // back trace for this cpu.
++
++ if (sp == NULL) {
++ if (tsk)
++ sp = (unsigned long *)tsk->thread.sp;
++ else
++ sp = (unsigned long *)&sp;
++ }
++
++ stack = sp;
++ for(i=0; i < kstack_depth_to_print; i++) {
++ if (stack >= irqstack && stack <= irqstack_end) {
++ if (stack == irqstack_end) {
++ stack = (unsigned long *) (irqstack_end[-1]);
++ printk(" <EOI> ");
++ }
++ } else {
++ if (((long) stack & (THREAD_SIZE-1)) == 0)
++ break;
++ }
++ if (i && ((i % 4) == 0))
++ printk("\n");
++ printk(" %016lx", *stack++);
++ touch_nmi_watchdog();
++ }
++ show_trace(tsk, regs, sp, bp);
++}
++
++void show_stack(struct task_struct *tsk, unsigned long * sp)
++{
++ _show_stack(tsk, NULL, sp, 0);
++}
++
++/*
++ * The architecture-independent dump_stack generator
++ */
++void dump_stack(void)
++{
++ unsigned long dummy;
++ unsigned long bp = 0;
++
++#ifdef CONFIG_FRAME_POINTER
++ if (!bp)
++ asm("movq %%rbp, %0" : "=r" (bp):);
++#endif
++
++ 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);
++ show_trace(NULL, NULL, &dummy, bp);
++}
++
++EXPORT_SYMBOL(dump_stack);
++
++void show_registers(struct pt_regs *regs)
++{
++ int i;
++ unsigned long sp;
++ const int cpu = smp_processor_id();
++ struct task_struct *cur = cpu_pda(cpu)->pcurrent;
++ u8 *ip;
++ unsigned int code_prologue = code_bytes * 43 / 64;
++ unsigned int code_len = code_bytes;
++
++ sp = regs->sp;
++ ip = (u8 *) regs->ip - code_prologue;
++ printk("CPU %d ", cpu);
++ __show_regs(regs);
++ printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
++ cur->comm, cur->pid, task_thread_info(cur), cur);
++
++ /*
++ * When in-kernel, we also print out the stack and code at the
++ * time of the fault..
++ */
++ if (!user_mode(regs)) {
++ unsigned char c;
++ printk("Stack: ");
++ _show_stack(NULL, regs, (unsigned long *)sp, regs->bp);
++ printk("\n");
++
++ printk(KERN_EMERG "Code: ");
++ if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) {
++ /* try starting at RIP */
++ ip = (u8 *) regs->ip;
++ code_len = code_len - code_prologue + 1;
++ }
++ for (i = 0; i < code_len; i++, ip++) {
++ if (ip < (u8 *)PAGE_OFFSET ||
++ probe_kernel_address(ip, c)) {
++ printk(" Bad RIP value.");
++ break;
++ }
++ if (ip == (u8 *)regs->ip)
++ printk("<%02x> ", c);
++ else
++ printk("%02x ", c);
++ }
++ }
++ printk("\n");
++}
++
++int is_valid_bugaddr(unsigned long ip)
++{
++ unsigned short ud2;
++
++ if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
++ return 0;
++
++ return ud2 == 0x0b0f;
++}
++
++static raw_spinlock_t die_lock = __RAW_SPIN_LOCK_UNLOCKED;
++static int die_owner = -1;
++static unsigned int die_nest_count;
++
++unsigned __kprobes long oops_begin(void)
++{
++ int cpu;
++ unsigned long flags;
++
++ oops_enter();
++
++ /* racy, but better than risking deadlock. */
++ raw_local_irq_save(flags);
++ cpu = smp_processor_id();
++ if (!__raw_spin_trylock(&die_lock)) {
++ if (cpu == die_owner)
++ /* nested oops. should stop eventually */;
++ else
++ __raw_spin_lock(&die_lock);
++ }
++ die_nest_count++;
++ die_owner = cpu;
++ console_verbose();
++ bust_spinlocks(1);
++ return flags;
++}
++
++void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr)
++{
++ die_owner = -1;
++ bust_spinlocks(0);
++ die_nest_count--;
++ if (!die_nest_count)
++ /* Nest count reaches zero, release the lock. */
++ __raw_spin_unlock(&die_lock);
++ raw_local_irq_restore(flags);
++ if (!regs) {
++ oops_exit();
++ return;
++ }
++ if (panic_on_oops)
++ panic("Fatal exception");
++ oops_exit();
++ do_exit(signr);
++}
++
++int __kprobes __die(const char * str, struct pt_regs * regs, long err)
++{
++ static int die_counter;
++ printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff,++die_counter);
++#ifdef CONFIG_PREEMPT
++ printk("PREEMPT ");
++#endif
++#ifdef CONFIG_SMP
++ printk("SMP ");
++#endif
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ printk("DEBUG_PAGEALLOC");
++#endif
++ printk("\n");
++ if (notify_die(DIE_OOPS, str, regs, err, current->thread.trap_no, SIGSEGV) == NOTIFY_STOP)
++ return 1;
++ show_registers(regs);
++ add_taint(TAINT_DIE);
++ /* Executive summary in case the oops scrolled away */
++ printk(KERN_ALERT "RIP ");
++ printk_address(regs->ip, 1);
++ printk(" RSP <%016lx>\n", regs->sp);
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
++ return 0;
++}
++
++void die(const char * str, struct pt_regs * regs, long err)
++{
++ unsigned long flags = oops_begin();
++
++ if (!user_mode(regs))
++ report_bug(regs->ip, regs);
++
++ if (__die(str, regs, err))
++ regs = NULL;
++ oops_end(flags, regs, SIGSEGV);
++}
++
++#ifdef CONFIG_SYSCTL
++notrace __kprobes void
++die_nmi(char *str, struct pt_regs *regs, int do_panic)
++{
++ unsigned long flags;
++
++ if (notify_die(DIE_NMIWATCHDOG, str, regs, 0, 2, SIGINT) ==
++ NOTIFY_STOP)
++ return;
++
++ flags = oops_begin();
++ /*
++ * We are in trouble anyway, lets at least try
++ * to get a message out.
++ */
++ printk(str, smp_processor_id());
++ show_registers(regs);
++ if (kexec_should_crash(current))
++ crash_kexec(regs);
++ if (do_panic || panic_on_oops)
++ panic("Non maskable interrupt");
++ oops_end(flags, NULL, SIGBUS);
++ nmi_exit();
++ local_irq_enable();
++ do_exit(SIGBUS);
++}
++#endif
++
++static void __kprobes do_trap(int trapnr, int signr, char *str,
++ struct pt_regs * regs, long error_code,
++ siginfo_t *info)
++{
++ struct task_struct *tsk = current;
++
++ if (user_mode(regs)) {
++ /*
++ * We want error_code and trap_no set for userspace
++ * faults and kernelspace faults which result in
++ * die(), but not kernelspace faults which are fixed
++ * up. die() gives the process no chance to handle
++ * the signal and notice the kernel fault information,
++ * so that won't result in polluting the information
++ * about previously queued, but not yet delivered,
++ * faults. See also do_general_protection below.
++ */
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
++
++ if (show_unhandled_signals && unhandled_signal(tsk, signr) &&
++ printk_ratelimit()) {
++ printk(KERN_INFO
++ "%s[%d] trap %s ip:%lx sp:%lx error:%lx",
++ tsk->comm, tsk->pid, str,
++ regs->ip, regs->sp, error_code);
++ print_vma_addr(" in ", regs->ip);
++ printk("\n");
++ }
++
++ if (info)
++ force_sig_info(signr, info, tsk);
++ else
++ force_sig(signr, tsk);
++ return;
++ }
++
++
++ if (!fixup_exception(regs)) {
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = trapnr;
++ die(str, regs, error_code);
++ }
++ return;
++}
++
++#define DO_ERROR(trapnr, signr, str, name) \
++asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ conditional_sti(regs); \
++ do_trap(trapnr, signr, str, regs, error_code, NULL); \
++}
++
++#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \
++asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
++{ \
++ siginfo_t info; \
++ info.si_signo = signr; \
++ info.si_errno = 0; \
++ info.si_code = sicode; \
++ info.si_addr = (void __user *)siaddr; \
++ trace_hardirqs_fixup(); \
++ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
++ == NOTIFY_STOP) \
++ return; \
++ conditional_sti(regs); \
++ do_trap(trapnr, signr, str, regs, error_code, &info); \
++}
++
++DO_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip)
++DO_ERROR( 4, SIGSEGV, "overflow", overflow)
++DO_ERROR( 5, SIGSEGV, "bounds", bounds)
++DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip)
++DO_ERROR( 7, SIGSEGV, "device not available", device_not_available)
++DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun)
++DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
++DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
++DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
++DO_ERROR(18, SIGSEGV, "reserved", reserved)
++
++/* Runs on IST stack */
++asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code)
++{
++ if (notify_die(DIE_TRAP, "stack segment", regs, error_code,
++ 12, SIGBUS) == NOTIFY_STOP)
++ return;
++ preempt_conditional_sti(regs);
++ do_trap(12, SIGBUS, "stack segment", regs, error_code, NULL);
++ preempt_conditional_cli(regs);
++}
++
++asmlinkage void do_double_fault(struct pt_regs * regs, long error_code)
++{
++ static const char str[] = "double fault";
++ struct task_struct *tsk = current;
++
++ /* Return not checked because double check cannot be ignored */
++ notify_die(DIE_TRAP, str, regs, error_code, 8, SIGSEGV);
++
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 8;
++
++ /* This is always a kernel trap and never fixable (and thus must
++ never return). */
++ for (;;)
++ die(str, regs, error_code);
++}
++
++asmlinkage void __kprobes do_general_protection(struct pt_regs * regs,
++ long error_code)
++{
++ struct task_struct *tsk = current;
++
++ conditional_sti(regs);
++
++ if (user_mode(regs)) {
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 13;
++
++ if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
++ printk_ratelimit()) {
++ printk(KERN_INFO
++ "%s[%d] general protection ip:%lx sp:%lx error:%lx",
++ tsk->comm, tsk->pid,
++ regs->ip, regs->sp, error_code);
++ print_vma_addr(" in ", regs->ip);
++ printk("\n");
++ }
++
++ force_sig(SIGSEGV, tsk);
++ return;
++ }
++
++ if (fixup_exception(regs))
++ return;
++
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 13;
++ if (notify_die(DIE_GPF, "general protection fault", regs,
++ error_code, 13, SIGSEGV) == NOTIFY_STOP)
++ return;
++ die("general protection fault", regs, error_code);
++}
++
++static notrace __kprobes void
++mem_parity_error(unsigned char reason, struct pt_regs * regs)
++{
++ 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 defined(CONFIG_EDAC)
++ if(edac_handler_set()) {
++ edac_atomic_assert_error();
++ return;
++ }
++#endif
++
++ if (panic_on_unrecovered_nmi)
++ panic("NMI: Not continuing");
++
++ printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
++
++ /* Clear and disable the memory parity error line. */
++ clear_mem_error(reason);
++}
++
++static notrace __kprobes void
++io_check_error(unsigned char reason, struct pt_regs * regs)
++{
++ printk("NMI: IOCK error (debug interrupt?)\n");
++ show_registers(regs);
++
++ /* Re-enable the IOCK line, wait for a few seconds */
++ clear_io_check_error(reason);
++}
++
++static notrace __kprobes void
++unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
++{
++ if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
++ return;
++ 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.
++ Nested NMIs are prevented by the CPU. */
++asmlinkage notrace __kprobes void default_do_nmi(struct pt_regs *regs)
++{
++ unsigned char reason = 0;
++ int cpu;
++
++ cpu = smp_processor_id();
++
++ /* Only the BSP gets external NMIs from the system. */
++ if (!cpu)
++ reason = get_nmi_reason();
++
++ if (!(reason & 0xc0)) {
++ if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT)
++ == NOTIFY_STOP)
++ return;
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
++ /*
++ * Ok, so this is none of the documented NMI sources,
++ * so it must be the NMI watchdog.
++ */
++ if (nmi_watchdog_tick(regs,reason))
++ return;
++#endif
++ if (!do_nmi_callback(regs,cpu))
++ unknown_nmi_error(reason, regs);
++
++ return;
++ }
++ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP)
++ return;
++
++ /* AK: following checks seem to be broken on modern chipsets. FIXME */
++
++ if (reason & 0x80)
++ mem_parity_error(reason, regs);
++ if (reason & 0x40)
++ io_check_error(reason, regs);
++}
++
++/* runs on IST stack. */
++asmlinkage void __kprobes do_int3(struct pt_regs * regs, long error_code)
++{
++ trace_hardirqs_fixup();
++
++ if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) {
++ return;
++ }
++ preempt_conditional_sti(regs);
++ do_trap(3, SIGTRAP, "int3", regs, error_code, NULL);
++ preempt_conditional_cli(regs);
++}
++
++/* Help handler running on IST stack to switch back to user stack
++ for scheduling or signal handling. The actual stack switch is done in
++ entry.S */
++asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
++{
++ struct pt_regs *regs = eregs;
++ /* Did already sync */
++ if (eregs == (struct pt_regs *)eregs->sp)
++ ;
++ /* Exception from user space */
++ else if (user_mode(eregs))
++ regs = task_pt_regs(current);
++ /* Exception from kernel and interrupts are enabled. Move to
++ kernel process stack. */
++ else if (eregs->flags & X86_EFLAGS_IF)
++ regs = (struct pt_regs *)(eregs->sp -= sizeof(struct pt_regs));
++ if (eregs != regs)
++ *regs = *eregs;
++ return regs;
++}
++
++/* runs on IST stack. */
++asmlinkage void __kprobes do_debug(struct pt_regs * regs,
++ unsigned long error_code)
++{
++ unsigned long condition;
++ struct task_struct *tsk = current;
++ siginfo_t info;
++
++ trace_hardirqs_fixup();
++
++ get_debugreg(condition, 6);
++
++ /*
++ * The processor cleared BTF, so don't mark that we need it set.
++ */
++ clear_tsk_thread_flag(tsk, TIF_DEBUGCTLMSR);
++ tsk->thread.debugctlmsr = 0;
++
++ if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code,
++ SIGTRAP) == NOTIFY_STOP)
++ return;
++
++ preempt_conditional_sti(regs);
++
++ /* Mask out spurious debug traps due to lazy DR7 setting */
++ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) {
++ if (!tsk->thread.debugreg7) {
++ goto clear_dr7;
++ }
++ }
++
++ tsk->thread.debugreg6 = condition;
++
++
++ /*
++ * Single-stepping through TF: make sure we ignore any events in
++ * kernel space (but re-enable TF when returning to user mode).
++ */
++ if (condition & DR_STEP) {
++ if (!user_mode(regs))
++ goto clear_TF_reenable;
++ }
++
++ /* Ok, finally something we can handle */
++ tsk->thread.trap_no = 1;
++ tsk->thread.error_code = error_code;
++ info.si_signo = SIGTRAP;
++ info.si_errno = 0;
++ info.si_code = TRAP_BRKPT;
++ info.si_addr = user_mode(regs) ? (void __user *)regs->ip : NULL;
++ force_sig_info(SIGTRAP, &info, tsk);
++
++clear_dr7:
++ set_debugreg(0UL, 7);
++ preempt_conditional_cli(regs);
++ return;
++
++clear_TF_reenable:
++ set_tsk_thread_flag(tsk, TIF_SINGLESTEP);
++ regs->flags &= ~X86_EFLAGS_TF;
++ preempt_conditional_cli(regs);
++}
++
++static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr)
++{
++ if (fixup_exception(regs))
++ return 1;
++
++ notify_die(DIE_GPF, str, regs, 0, trapnr, SIGFPE);
++ /* Illegal floating point operation in the kernel */
++ current->thread.trap_no = trapnr;
++ die(str, regs, 0);
++ return 0;
++}
++
++/*
++ * Note that we play around with the 'TS' bit in an attempt to get
++ * the correct behaviour even in the presence of the asynchronous
++ * IRQ13 behaviour
++ */
++asmlinkage void do_coprocessor_error(struct pt_regs *regs)
++{
++ void __user *ip = (void __user *)(regs->ip);
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short cwd, swd;
++
++ conditional_sti(regs);
++ if (!user_mode(regs) &&
++ kernel_math_error(regs, "kernel x87 math error", 16))
++ return;
++
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 16;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = ip;
++ /*
++ * (~cwd & swd) will mask out exceptions that are not set to unmasked
++ * status. 0x3f is the exception bits in these regs, 0x200 is the
++ * C1 reg you need in case of a stack fault, 0x040 is the stack
++ * fault bit. We should only be taking one exception at a time,
++ * so if this combination doesn't produce any single exception,
++ * then we have a bad program that isn't synchronizing its FPU usage
++ * and it will suffer the consequences since we won't be able to
++ * fully reproduce the context of the exception
++ */
++ cwd = get_fpu_cwd(task);
++ swd = get_fpu_swd(task);
++ switch (swd & ~cwd & 0x3f) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ /*
++ * swd & 0x240 == 0x040: Stack Underflow
++ * swd & 0x240 == 0x240: Stack Overflow
++ * User must clear the SF bit (0x40) if set
++ */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++asmlinkage void bad_intr(void)
++{
++ printk("bad interrupt");
++}
++
++asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs)
++{
++ void __user *ip = (void __user *)(regs->ip);
++ struct task_struct * task;
++ siginfo_t info;
++ unsigned short mxcsr;
++
++ conditional_sti(regs);
++ if (!user_mode(regs) &&
++ kernel_math_error(regs, "kernel simd math error", 19))
++ return;
++
++ /*
++ * Save the info for the exception handler and clear the error.
++ */
++ task = current;
++ save_init_fpu(task);
++ task->thread.trap_no = 19;
++ task->thread.error_code = 0;
++ info.si_signo = SIGFPE;
++ info.si_errno = 0;
++ info.si_code = __SI_FAULT;
++ info.si_addr = ip;
++ /*
++ * The SIMD FPU exceptions are handled a little differently, as there
++ * is only a single status/control register. Thus, to determine which
++ * unmasked exception was caught we must mask the exception mask bits
++ * at 0x1f80, and then use these to mask the exception bits at 0x3f.
++ */
++ mxcsr = get_fpu_mxcsr(task);
++ switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) {
++ case 0x000:
++ default:
++ break;
++ case 0x001: /* Invalid Op */
++ info.si_code = FPE_FLTINV;
++ break;
++ case 0x002: /* Denormalize */
++ case 0x010: /* Underflow */
++ info.si_code = FPE_FLTUND;
++ break;
++ case 0x004: /* Zero Divide */
++ info.si_code = FPE_FLTDIV;
++ break;
++ case 0x008: /* Overflow */
++ info.si_code = FPE_FLTOVF;
++ break;
++ case 0x020: /* Precision */
++ info.si_code = FPE_FLTRES;
++ break;
++ }
++ force_sig_info(SIGFPE, &info, task);
++}
++
++asmlinkage void do_spurious_interrupt_bug(struct pt_regs * regs)
++{
++}
++
++#ifndef CONFIG_XEN
++asmlinkage void __attribute__((weak)) smp_thermal_interrupt(void)
++{
++}
++
++asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void)
++{
++}
++#endif
++
++/*
++ * 'math_state_restore()' saves the current math information in the
++ * old math state array, and gets the new ones from the current task
++ *
++ * Careful.. There are problems with IBM-designed IRQ13 behaviour.
++ * Don't touch unless you *really* know how it works.
++ */
++asmlinkage void math_state_restore(void)
++{
++ struct task_struct *me = current;
++
++ if (!used_math()) {
++ local_irq_enable();
++ /*
++ * does a slab alloc which can sleep
++ */
++ if (init_fpu(me)) {
++ /*
++ * ran out of memory!
++ */
++ do_group_exit(SIGKILL);
++ return;
++ }
++ local_irq_disable();
++ }
++
++ /* clts(); */ /* 'clts' is done for us by Xen during virtual trap. */
++
++ restore_fpu_checking(&me->thread.xstate->fxsave);
++ task_thread_info(me)->status |= TS_USEDFPU;
++ me->fpu_counter++;
++}
++EXPORT_SYMBOL_GPL(math_state_restore);
++
++
++/*
++ * NB. All these are "interrupt gates" (i.e. events_mask is set) because we
++ * specify <dpl>|4 in the second field.
++ */
++static const trap_info_t __cpuinitconst trap_table[] = {
++ { 0, 0|4, __KERNEL_CS, (unsigned long)divide_error },
++ { 1, 0|4, __KERNEL_CS, (unsigned long)debug },
++ { 3, 3|4, __KERNEL_CS, (unsigned long)int3 },
++ { 4, 3|4, __KERNEL_CS, (unsigned long)overflow },
++ { 5, 0|4, __KERNEL_CS, (unsigned long)bounds },
++ { 6, 0|4, __KERNEL_CS, (unsigned long)invalid_op },
++ { 7, 0|4, __KERNEL_CS, (unsigned long)device_not_available },
++ { 9, 0|4, __KERNEL_CS, (unsigned long)coprocessor_segment_overrun},
++ { 10, 0|4, __KERNEL_CS, (unsigned long)invalid_TSS },
++ { 11, 0|4, __KERNEL_CS, (unsigned long)segment_not_present },
++ { 12, 0|4, __KERNEL_CS, (unsigned long)stack_segment },
++ { 13, 0|4, __KERNEL_CS, (unsigned long)general_protection },
++ { 14, 0|4, __KERNEL_CS, (unsigned long)page_fault },
++ { 15, 0|4, __KERNEL_CS, (unsigned long)spurious_interrupt_bug },
++ { 16, 0|4, __KERNEL_CS, (unsigned long)coprocessor_error },
++ { 17, 0|4, __KERNEL_CS, (unsigned long)alignment_check },
++#ifdef CONFIG_X86_MCE
++ { 18, 0|4, __KERNEL_CS, (unsigned long)machine_check },
++#endif
++ { 19, 0|4, __KERNEL_CS, (unsigned long)simd_coprocessor_error },
++#ifdef CONFIG_IA32_EMULATION
++ { IA32_SYSCALL_VECTOR, 3, __KERNEL_CS, (unsigned long)ia32_syscall},
++#endif
++ { 0, 0, 0, 0 }
++};
++
++void __init trap_init(void)
++{
++ int ret;
++
++ ret = HYPERVISOR_set_trap_table(trap_table);
++ if (ret)
++ printk("HYPERVISOR_set_trap_table failed: error %d\n", ret);
++
++ /*
++ * initialize the per thread extended state:
++ */
++ init_thread_xstate();
++ /*
++ * Should be a barrier for any external CPU state.
++ */
++ cpu_init();
++}
++
++void __cpuinit smp_trap_init(trap_info_t *trap_ctxt)
++{
++ const trap_info_t *t = trap_table;
++
++ for (t = trap_table; t->address; t++) {
++ trap_ctxt[t->vector].flags = t->flags;
++ trap_ctxt[t->vector].cs = t->cs;
++ trap_ctxt[t->vector].address = t->address;
++ }
++}
++
++
++static int __init oops_setup(char *s)
++{
++ if (!s)
++ return -EINVAL;
++ if (!strcmp(s, "panic"))
++ panic_on_oops = 1;
++ return 0;
++}
++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 0;
++}
++early_param("kstack", kstack_setup);
++
++
++static int __init code_bytes_setup(char *s)
++{
++ code_bytes = simple_strtoul(s, NULL, 0);
++ if (code_bytes > 8192)
++ code_bytes = 8192;
++
++ return 1;
++}
++__setup("code_bytes=", code_bytes_setup);
+diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c
+index 38f566f..fc500c7 100644
+--- a/arch/x86/kernel/vm86_32.c
++++ b/arch/x86/kernel/vm86_32.c
+@@ -124,7 +124,9 @@ static int copy_vm86_regs_from_user(struct kernel_vm86_regs *regs,
+
+ struct pt_regs *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;
+
+@@ -147,12 +149,16 @@ struct pt_regs *save_v86_state(struct kernel_vm86_regs *regs)
+ do_exit(SIGSEGV);
+ }
+
++#ifndef CONFIG_X86_NO_TSS
+ tss = &per_cpu(init_tss, get_cpu());
++#endif
+ current->thread.sp0 = current->thread.saved_sp0;
+ current->thread.sysenter_cs = __KERNEL_CS;
+ load_sp0(tss, ¤t->thread);
+ current->thread.saved_sp0 = 0;
++#ifndef CONFIG_X86_NO_TSS
+ put_cpu();
++#endif
+
+ ret = KVM86->regs32;
+
+@@ -279,7 +285,9 @@ out:
+
+ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ struct tss_struct *tss;
++#endif
+ /*
+ * make sure the vm86() system call doesn't try to do anything silly
+ */
+@@ -324,12 +332,16 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk
+ tsk->thread.saved_fs = info->regs32->fs;
+ savesegment(gs, tsk->thread.saved_gs);
+
++#ifndef CONFIG_X86_NO_TSS
+ tss = &per_cpu(init_tss, get_cpu());
++#endif
+ tsk->thread.sp0 = (unsigned long) &info->VM86_TSS_ESP0;
+ if (cpu_has_sep)
+ tsk->thread.sysenter_cs = 0;
+ load_sp0(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 --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S
+index ce5ed08..050a8eb 100644
+--- a/arch/x86/kernel/vmlinux_32.lds.S
++++ b/arch/x86/kernel/vmlinux_32.lds.S
+@@ -29,6 +29,12 @@ PHDRS {
+ SECTIONS
+ {
+ . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR;
++
++#if defined(CONFIG_XEN) && CONFIG_XEN_COMPAT <= 0x030002
++#undef LOAD_OFFSET
++#define LOAD_OFFSET 0
++#endif
++
+ phys_startup_32 = startup_32 - LOAD_OFFSET;
+
+ .text.head : AT(ADDR(.text.head) - LOAD_OFFSET) {
+diff --git a/arch/x86/kernel/vsyscall_64-xen.c b/arch/x86/kernel/vsyscall_64-xen.c
+new file mode 100644
+index 0000000..c942370
+--- /dev/null
++++ b/arch/x86/kernel/vsyscall_64-xen.c
+@@ -0,0 +1,319 @@
++/*
++ * 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 <linux/clocksource.h>
++#include <linux/getcpu.h>
++#include <linux/cpu.h>
++#include <linux/smp.h>
++#include <linux/notifier.h>
++
++#include <asm/vsyscall.h>
++#include <asm/pgtable.h>
++#include <asm/page.h>
++#include <asm/unistd.h>
++#include <asm/fixmap.h>
++#include <asm/errno.h>
++#include <asm/io.h>
++#include <asm/segment.h>
++#include <asm/desc.h>
++#include <asm/topology.h>
++#include <asm/vgtod.h>
++
++#define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr)))
++#define __syscall_clobber "r11","cx","memory"
++
++/*
++ * vsyscall_gtod_data contains data that is :
++ * - readonly from vsyscalls
++ * - written by timer interrupt or systcl (/proc/sys/kernel/vsyscall64)
++ * Try to keep this structure as small as possible to avoid cache line ping pongs
++ */
++int __vgetcpu_mode __section_vgetcpu_mode;
++
++struct vsyscall_gtod_data __vsyscall_gtod_data __section_vsyscall_gtod_data =
++{
++ .lock = SEQLOCK_UNLOCKED,
++ .sysctl_enabled = 1,
++};
++
++void update_vsyscall_tz(void)
++{
++ unsigned long flags;
++
++ write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
++ /* sys_tz has changed */
++ vsyscall_gtod_data.sys_tz = sys_tz;
++ write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
++}
++
++void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
++{
++ unsigned long flags;
++
++ write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
++ /* copy vsyscall data */
++ vsyscall_gtod_data.clock.vread = clock->vread;
++ vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
++ vsyscall_gtod_data.clock.mask = clock->mask;
++ vsyscall_gtod_data.clock.mult = clock->mult;
++ vsyscall_gtod_data.clock.shift = clock->shift;
++ vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec;
++ vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
++ vsyscall_gtod_data.wall_to_monotonic = wall_to_monotonic;
++ write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
++}
++
++/* 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 = __vsyscall_gtod_data.sys_tz;
++}
++
++static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz)
++{
++ int ret;
++ asm volatile("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("syscall"
++ : "=a" (secs)
++ : "0" (__NR_time),"D" (t) : __syscall_clobber);
++ return secs;
++}
++
++static __always_inline void do_vgettimeofday(struct timeval * tv)
++{
++ cycle_t now, base, mask, cycle_delta;
++ unsigned seq;
++ unsigned long mult, shift, nsec;
++ cycle_t (*vread)(void);
++ do {
++ seq = read_seqbegin(&__vsyscall_gtod_data.lock);
++
++ vread = __vsyscall_gtod_data.clock.vread;
++ if (unlikely(!__vsyscall_gtod_data.sysctl_enabled || !vread)) {
++ gettimeofday(tv,NULL);
++ return;
++ }
++ now = vread();
++ base = __vsyscall_gtod_data.clock.cycle_last;
++ mask = __vsyscall_gtod_data.clock.mask;
++ mult = __vsyscall_gtod_data.clock.mult;
++ shift = __vsyscall_gtod_data.clock.shift;
++
++ tv->tv_sec = __vsyscall_gtod_data.wall_time_sec;
++ nsec = __vsyscall_gtod_data.wall_time_nsec;
++ } while (read_seqretry(&__vsyscall_gtod_data.lock, seq));
++
++ /* calculate interval: */
++ cycle_delta = (now - base) & mask;
++ /* convert to nsecs: */
++ nsec += (cycle_delta * mult) >> shift;
++
++ while (nsec >= NSEC_PER_SEC) {
++ tv->tv_sec += 1;
++ nsec -= NSEC_PER_SEC;
++ }
++ tv->tv_usec = nsec / NSEC_PER_USEC;
++}
++
++int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * 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)
++{
++ struct timeval tv;
++ time_t result;
++ if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
++ return time_syscall(t);
++
++ vgettimeofday(&tv, NULL);
++ result = tv.tv_sec;
++ if (t)
++ *t = result;
++ return result;
++}
++
++/* Fast way to get current CPU and node.
++ This helps to do per node and per CPU caches in user space.
++ The result is not guaranteed without CPU affinity, but usually
++ works out because the scheduler tries to keep a thread on the same
++ CPU.
++
++ tcache must point to a two element sized long array.
++ All arguments can be NULL. */
++long __vsyscall(2)
++vgetcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache)
++{
++ unsigned int p;
++ unsigned long j = 0;
++
++ /* Fast cache - only recompute value once per jiffies and avoid
++ relatively costly rdtscp/cpuid otherwise.
++ This works because the scheduler usually keeps the process
++ on the same CPU and this syscall doesn't guarantee its
++ results anyways.
++ We do this here because otherwise user space would do it on
++ its own in a likely inferior way (no access to jiffies).
++ If you don't like it pass NULL. */
++ if (tcache && tcache->blob[0] == (j = __jiffies)) {
++ p = tcache->blob[1];
++ } else if (__vgetcpu_mode == VGETCPU_RDTSCP) {
++ /* Load per CPU data from RDTSCP */
++ native_read_tscp(&p);
++ } else {
++ /* Load per CPU data from GDT */
++ asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG));
++ }
++ if (tcache) {
++ tcache->blob[0] = j;
++ tcache->blob[1] = p;
++ }
++ if (cpu)
++ *cpu = p & 0xfff;
++ if (node)
++ *node = p >> 12;
++ return 0;
++}
++
++static long __vsyscall(3) venosys_1(void)
++{
++ return -ENOSYS;
++}
++
++#ifdef CONFIG_SYSCTL
++
++static int
++vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp,
++ void __user *buffer, size_t *lenp, loff_t *ppos)
++{
++ return proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
++}
++
++static ctl_table kernel_table2[] = {
++ { .procname = "vsyscall64",
++ .data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
++ .mode = 0644,
++ .proc_handler = vsyscall_sysctl_change },
++ {}
++};
++
++static ctl_table kernel_root_table2[] = {
++ { .ctl_name = CTL_KERN, .procname = "kernel", .mode = 0555,
++ .child = kernel_table2 },
++ {}
++};
++#endif
++
++/* 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)
++{
++ unsigned long d;
++ unsigned long node = 0;
++#ifdef CONFIG_NUMA
++ node = cpu_to_node(cpu);
++#endif
++ if (cpu_has(&cpu_data(cpu), X86_FEATURE_RDTSCP))
++ write_rdtscp_aux((node << 12) | cpu);
++
++ /* Store cpu number in limit so that it can be loaded quickly
++ in user space in vgetcpu.
++ 12 bits for the CPU and 8 bits for the node. */
++ d = 0x0f40000000000ULL;
++ d |= cpu;
++ d |= (node & 0xf) << 12;
++ d |= (node >> 4) << 48;
++ if (HYPERVISOR_update_descriptor(virt_to_machine(get_cpu_gdt_table(cpu)
++ + GDT_ENTRY_PER_CPU),
++ d))
++ BUG();
++}
++
++static void __cpuinit cpu_vsyscall_init(void *arg)
++{
++ /* preemption should be already off */
++ vsyscall_set_cpu(raw_smp_processor_id());
++}
++
++static int __cpuinit
++cpu_vsyscall_notifier(struct notifier_block *n, unsigned long action, void *arg)
++{
++ long cpu = (long)arg;
++ if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN)
++ smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 0, 1);
++ return NOTIFY_DONE;
++}
++
++void __init map_vsyscall(void)
++{
++ extern char __vsyscall_0;
++ unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0);
++
++ /* Note that VSYSCALL_MAPPED_PAGES must agree with the code below. */
++ __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL);
++}
++
++static int __init vsyscall_init(void)
++{
++ BUG_ON(((unsigned long) &vgettimeofday !=
++ VSYSCALL_ADDR(__NR_vgettimeofday)));
++ BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime));
++ BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)));
++ BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu));
++#ifdef CONFIG_XEN
++ vsyscall_gtod_data.sysctl_enabled = 0; /* disable vgettimeofay() */
++ if (boot_cpu_has(X86_FEATURE_RDTSCP))
++ vgetcpu_mode = VGETCPU_RDTSCP;
++ else
++ vgetcpu_mode = VGETCPU_LSL;
++#endif
++#ifdef CONFIG_SYSCTL
++ register_sysctl_table(kernel_root_table2);
++#endif
++ on_each_cpu(cpu_vsyscall_init, NULL, 0, 1);
++ hotcpu_notifier(cpu_vsyscall_notifier, 0);
++ return 0;
++}
++
++__initcall(vsyscall_init);
+diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
+index 8d45fab..dc1e3fc 100644
+--- a/arch/x86/kvm/Kconfig
++++ b/arch/x86/kvm/Kconfig
+@@ -7,6 +7,7 @@ config HAVE_KVM
+ menuconfig VIRTUALIZATION
+ bool "Virtualization"
+ depends on HAVE_KVM || X86
++ depends on !XEN
+ default y
+ ---help---
+ Say Y here to get to see options for using your Linux host to run other
+diff --git a/arch/x86/lib/Makefile b/arch/x86/lib/Makefile
+index 76f60f5..b41635e 100644
+--- a/arch/x86/lib/Makefile
++++ b/arch/x86/lib/Makefile
+@@ -24,3 +24,5 @@ else
+ lib-y += memmove_64.o memset_64.o
+ lib-y += copy_user_64.o rwlock_64.o copy_user_nocache_64.o
+ endif
++
++lib-$(CONFIG_XEN_SCRUB_PAGES) += scrub.o
+diff --git a/arch/x86/lib/scrub.c b/arch/x86/lib/scrub.c
+new file mode 100644
+index 0000000..f333ae8
+--- /dev/null
++++ b/arch/x86/lib/scrub.c
+@@ -0,0 +1,21 @@
++#include <asm/cpufeature.h>
++#include <asm/page.h>
++#include <asm/processor.h>
++
++void scrub_pages(void *v, unsigned int count)
++{
++ if (likely(cpu_has_xmm2)) {
++ unsigned long n = count * (PAGE_SIZE / sizeof(long) / 4);
++
++ for (; n--; v += sizeof(long) * 4)
++ asm("movnti %1,(%0)\n\t"
++ "movnti %1,%c2(%0)\n\t"
++ "movnti %1,2*%c2(%0)\n\t"
++ "movnti %1,3*%c2(%0)\n\t"
++ : : "r" (v), "r" (0L), "i" (sizeof(long))
++ : "memory");
++ asm volatile("sfence" : : : "memory");
++ } else
++ for (; count--; v += PAGE_SIZE)
++ clear_page(v);
++}
+diff --git a/arch/x86/mach-xen/Makefile b/arch/x86/mach-xen/Makefile
+new file mode 100644
+index 0000000..012fe34
+--- /dev/null
++++ b/arch/x86/mach-xen/Makefile
+@@ -0,0 +1,5 @@
++#
++# Makefile for the linux kernel.
++#
++
++obj-y := setup.o
+diff --git a/arch/x86/mach-xen/setup.c b/arch/x86/mach-xen/setup.c
+new file mode 100644
+index 0000000..6fed8f1
+--- /dev/null
++++ b/arch/x86/mach-xen/setup.c
+@@ -0,0 +1,172 @@
++/*
++ * Machine specific setup for generic
++ */
++
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/interrupt.h>
++#include <linux/module.h>
++#include <asm/acpi.h>
++#include <asm/arch_hooks.h>
++#include <asm/e820.h>
++#include <asm/setup.h>
++#include <asm/fixmap.h>
++#include <asm/pgtable.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)
++{
++ get_option(&str, &no_broadcast);
++ printk ("Using %s mode\n", no_broadcast ? "No IPI Broadcast" :
++ "IPI Broadcast");
++ return 1;
++}
++
++__setup("no_ipi_broadcast", no_ipi_broadcast);
++
++static int __init print_ipi_mode(void)
++{
++ 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)
++{
++ 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((unsigned long long)xen_start_info->nr_pages);
++ /* 8MB slack (to balance backend allocations). */
++ map[0].size += 8ULL << 20;
++ map[0].type = E820_RAM;
++ rc = 0;
++ }
++ BUG_ON(rc);
++
++ sanitize_e820_map(map, (char *)&memmap.nr_entries);
++
++ BUG_ON(copy_e820_map(map, (char)memmap.nr_entries) < 0);
++
++ return "Xen";
++}
++
++
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void nmi(void);
++
++unsigned long *machine_to_phys_mapping = (void *)MACH2PHYS_VIRT_START;
++EXPORT_SYMBOL(machine_to_phys_mapping);
++unsigned int machine_to_phys_order;
++EXPORT_SYMBOL(machine_to_phys_order);
++
++void __init pre_setup_arch_hook(void)
++{
++ struct xen_machphys_mapping mapping;
++ unsigned long machine_to_phys_nr_ents;
++ struct xen_platform_parameters pp;
++
++ init_mm.pgd = swapper_pg_dir = (pgd_t *)xen_start_info->pt_base;
++
++ xen_setup_features();
++
++ if (HYPERVISOR_xen_version(XENVER_platform_parameters, &pp) == 0) {
++ hypervisor_virt_start = pp.virt_start;
++ reserve_top_address(0UL - pp.virt_start);
++ }
++
++ if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
++ machine_to_phys_mapping = (unsigned long *)mapping.v_start;
++ machine_to_phys_nr_ents = mapping.max_mfn + 1;
++ } else
++ machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
++ machine_to_phys_order = fls(machine_to_phys_nr_ents - 1);
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ phys_to_machine_mapping =
++ (unsigned long *)xen_start_info->mfn_list;
++}
++
++void __init machine_specific_arch_setup(void)
++{
++ int ret;
++ static struct callback_register __initdata event = {
++ .type = CALLBACKTYPE_event,
++ .address = { __KERNEL_CS, (unsigned long)hypervisor_callback },
++ };
++ static struct callback_register __initdata failsafe = {
++ .type = CALLBACKTYPE_failsafe,
++ .address = { __KERNEL_CS, (unsigned long)failsafe_callback },
++ };
++ static struct callback_register __initdata nmi_cb = {
++ .type = CALLBACKTYPE_nmi,
++ .address = { __KERNEL_CS, (unsigned long)nmi },
++ };
++
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS)
++ ret = HYPERVISOR_set_callbacks(
++ event.address.cs, event.address.eip,
++ failsafe.address.cs, failsafe.address.eip);
++#endif
++ BUG_ON(ret);
++
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS) {
++ static struct xennmi_callback __initdata cb = {
++ .handler_address = (unsigned long)nmi
++ };
++
++ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++ }
++#endif
++
++ /* Do an early initialization of the fixmap area */
++ {
++ extern pte_t swapper_pg_fixmap[PTRS_PER_PTE];
++ unsigned long addr = __fix_to_virt(FIX_EARLYCON_MEM_BASE);
++ pud_t *pud = pud_offset(swapper_pg_dir + pgd_index(addr), addr);
++ pmd_t *pmd = pmd_offset(pud, addr);
++
++ make_lowmem_page_readonly(swapper_pg_fixmap, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa_symbol(swapper_pg_fixmap) | _PAGE_TABLE));
++ }
++}
+diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
+index b7b3e4c..785fe11 100644
+--- a/arch/x86/mm/Makefile
++++ b/arch/x86/mm/Makefile
+@@ -15,3 +15,5 @@ obj-$(CONFIG_NUMA) += numa_64.o
+ obj-$(CONFIG_K8_NUMA) += k8topology_64.o
+ obj-$(CONFIG_ACPI_NUMA) += srat_64.o
+ endif
++
++obj-$(CONFIG_XEN) += hypervisor.o
+diff --git a/arch/x86/mm/fault-xen.c b/arch/x86/mm/fault-xen.c
+new file mode 100644
+index 0000000..b9b5594
+--- /dev/null
++++ b/arch/x86/mm/fault-xen.c
+@@ -0,0 +1,1024 @@
++/*
++ * Copyright (C) 1995 Linus Torvalds
++ * Copyright (C) 2001,2002 Andi Kleen, SuSE Labs.
++ */
++
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/smp.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/tty.h>
++#include <linux/vt_kern.h> /* For unblank_screen() */
++#include <linux/compiler.h>
++#include <linux/highmem.h>
++#include <linux/bootmem.h> /* for max_low_pfn */
++#include <linux/vmalloc.h>
++#include <linux/module.h>
++#include <linux/kprobes.h>
++#include <linux/uaccess.h>
++#include <linux/kdebug.h>
++
++#include <asm/system.h>
++#include <asm/desc.h>
++#include <asm/segment.h>
++#include <asm/pgalloc.h>
++#include <asm/smp.h>
++#include <asm/tlbflush.h>
++#include <asm/proto.h>
++#include <asm-generic/sections.h>
++
++/*
++ * Page fault error code bits
++ * 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
++ */
++#define PF_PROT (1<<0)
++#define PF_WRITE (1<<1)
++#define PF_USER (1<<2)
++#define PF_RSVD (1<<3)
++#define PF_INSTR (1<<4)
++
++static inline int notify_page_fault(struct pt_regs *regs)
++{
++#ifdef CONFIG_KPROBES
++ int ret = 0;
++
++ /* kprobe_running() needs smp_processor_id() */
++#ifdef CONFIG_X86_32
++ if (!user_mode_vm(regs)) {
++#else
++ if (!user_mode(regs)) {
++#endif
++ preempt_disable();
++ if (kprobe_running() && kprobe_fault_handler(regs, 14))
++ ret = 1;
++ preempt_enable();
++ }
++
++ return ret;
++#else
++ return 0;
++#endif
++}
++
++/*
++ * X86_32
++ * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
++ * Check that here and ignore it.
++ *
++ * X86_64
++ * Sometimes the CPU reports invalid exceptions on prefetch.
++ * Check that here and ignore it.
++ *
++ * Opcode checker based on code by Richard Brunner
++ */
++static int is_prefetch(struct pt_regs *regs, unsigned long addr,
++ unsigned long error_code)
++{
++ unsigned char *instr;
++ int scan_more = 1;
++ int prefetch = 0;
++ unsigned char *max_instr;
++
++ /*
++ * If it was a exec (instruction fetch) fault on NX page, then
++ * do not ignore the fault:
++ */
++ if (error_code & PF_INSTR)
++ return 0;
++
++ instr = (unsigned char *)convert_ip_to_linear(current, regs);
++ max_instr = instr + 15;
++
++ if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE)
++ return 0;
++
++ while (scan_more && instr < max_instr) {
++ unsigned char opcode;
++ unsigned char instr_hi;
++ unsigned char instr_lo;
++
++ if (probe_kernel_address(instr, opcode))
++ break;
++
++ instr_hi = opcode & 0xf0;
++ instr_lo = opcode & 0x0f;
++ instr++;
++
++ switch (instr_hi) {
++ case 0x20:
++ case 0x30:
++ /*
++ * Values 0x26,0x2E,0x36,0x3E are valid x86 prefixes.
++ * In X86_64 long mode, the CPU will signal invalid
++ * opcode if some of these prefixes are present so
++ * X86_64 will never get here anyway
++ */
++ scan_more = ((instr_lo & 7) == 0x6);
++ break;
++#ifdef CONFIG_X86_64
++ case 0x40:
++ /*
++ * In AMD64 long mode 0x40..0x4F are valid REX prefixes
++ * Need to figure out under what instruction mode the
++ * instruction was issued. Could check the LDT for lm,
++ * but for now it's good enough to assume that long
++ * mode only uses well known segments or kernel.
++ */
++ scan_more = (!user_mode(regs)) || (regs->cs == __USER_CS);
++ break;
++#endif
++ case 0x60:
++ /* 0x64 thru 0x67 are valid prefixes in all modes. */
++ scan_more = (instr_lo & 0xC) == 0x4;
++ break;
++ case 0xF0:
++ /* 0xF0, 0xF2, 0xF3 are valid prefixes in all modes. */
++ scan_more = !instr_lo || (instr_lo>>1) == 1;
++ break;
++ case 0x00:
++ /* Prefetch instruction is 0x0F0D or 0x0F18 */
++ scan_more = 0;
++
++ if (probe_kernel_address(instr, opcode))
++ break;
++ prefetch = (instr_lo == 0xF) &&
++ (opcode == 0x0D || opcode == 0x18);
++ break;
++ default:
++ scan_more = 0;
++ break;
++ }
++ }
++ return prefetch;
++}
++
++static void force_sig_info_fault(int si_signo, int si_code,
++ unsigned long address, struct task_struct *tsk)
++{
++ siginfo_t info;
++
++ info.si_signo = si_signo;
++ info.si_errno = 0;
++ info.si_code = si_code;
++ info.si_addr = (void __user *)address;
++ force_sig_info(si_signo, &info, tsk);
++}
++
++#ifdef CONFIG_X86_64
++static int bad_address(void *p)
++{
++ unsigned long dummy;
++ return probe_kernel_address((unsigned long *)p, dummy);
++}
++#endif
++
++static void dump_pagetable(unsigned long address)
++{
++#ifdef CONFIG_X86_32
++ __typeof__(pte_val(__pte(0))) page;
++
++ page = read_cr3();
++ page = ((__typeof__(page) *) __va(page))[address >> PGDIR_SHIFT];
++#ifdef CONFIG_X86_PAE
++ printk("*pdpt = %016Lx ", page);
++ if ((page & _PAGE_PRESENT)
++ && mfn_to_local_pfn(page >> PAGE_SHIFT) < max_low_pfn) {
++ page = mfn_to_pfn(page >> PAGE_SHIFT);
++ page <<= PAGE_SHIFT;
++ page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
++ & (PTRS_PER_PMD - 1)];
++ printk(KERN_CONT "*pde = %016Lx ", page);
++ page &= ~_PAGE_NX;
++ }
++#else
++ printk("*pde = %08lx ", page);
++#endif
++
++ /*
++ * We must not directly access the pte in the highpte
++ * case if the page table is located in highmem.
++ * And let's rather not kmap-atomic the pte, just in case
++ * it's allocated already.
++ */
++ if ((page & _PAGE_PRESENT)
++ && mfn_to_local_pfn(page >> PAGE_SHIFT) < max_low_pfn
++ && !(page & _PAGE_PSE)) {
++ page = mfn_to_pfn(page >> PAGE_SHIFT);
++ page <<= PAGE_SHIFT;
++ page = ((__typeof__(page) *) __va(page))[(address >> PAGE_SHIFT)
++ & (PTRS_PER_PTE - 1)];
++ printk(KERN_CONT "*pte = %0*Lx ", sizeof(page)*2, (u64)page);
++ }
++
++ printk(KERN_CONT "\n");
++#else /* CONFIG_X86_64 */
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ pgd = (pgd_t *)read_cr3();
++
++ pgd = __va((unsigned long)pgd & PHYSICAL_PAGE_MASK);
++ pgd += pgd_index(address);
++ if (bad_address(pgd)) goto bad;
++ printk("PGD %lx ", pgd_val(*pgd));
++ if (!pgd_present(*pgd)) goto ret;
++
++ pud = pud_offset(pgd, address);
++ if (bad_address(pud)) goto bad;
++ printk(KERN_CONT "PUD %lx ", pud_val(*pud));
++ if (!pud_present(*pud) || pud_large(*pud))
++ goto ret;
++
++ pmd = pmd_offset(pud, address);
++ if (bad_address(pmd)) goto bad;
++ printk(KERN_CONT "PMD %lx ", pmd_val(*pmd));
++ if (!pmd_present(*pmd) || pmd_large(*pmd)) goto ret;
++
++ pte = pte_offset_kernel(pmd, address);
++ if (bad_address(pte)) goto bad;
++ printk(KERN_CONT "PTE %lx", pte_val(*pte));
++ret:
++ printk(KERN_CONT "\n");
++ return;
++bad:
++ printk("BAD\n");
++#endif
++}
++
++#ifdef CONFIG_X86_32
++static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
++{
++ 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;
++
++ if (!pgd_present(*pgd_k))
++ return NULL;
++
++ /*
++ * set_pgd(pgd, *pgd_k); here would be useless on PAE
++ * and redundant with the set_pmd() on non-PAE. As would
++ * set_pud.
++ */
++
++ pud = pud_offset(pgd, address);
++ pud_k = pud_offset(pgd_k, address);
++ if (!pud_present(*pud_k))
++ return NULL;
++
++ pmd = pmd_offset(pud, address);
++ pmd_k = pmd_offset(pud_k, address);
++ if (!pmd_present(*pmd_k))
++ return NULL;
++ if (!pmd_present(*pmd)) {
++#if CONFIG_XEN_COMPAT > 0x030002
++ set_pmd(pmd, *pmd_k);
++#else
++ /*
++ * When running on older Xen we must launder *pmd_k through
++ * pmd_val() to ensure that _PAGE_PRESENT is correctly set.
++ */
++ set_pmd(pmd, __pmd(pmd_val(*pmd_k)));
++#endif
++ arch_flush_lazy_mmu_mode();
++ } else
++ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
++ return pmd_k;
++}
++#endif
++
++#ifdef CONFIG_X86_64
++static const char errata93_warning[] =
++KERN_ERR "******* Your BIOS seems to not contain a fix for K8 errata #93\n"
++KERN_ERR "******* Working around it, but it may cause SEGVs or burn power.\n"
++KERN_ERR "******* Please consider a BIOS update.\n"
++KERN_ERR "******* Disabling USB legacy in the BIOS may also help.\n";
++#endif
++
++/* Workaround for K8 erratum #93 & buggy BIOS.
++ BIOS SMM functions are required to use a specific workaround
++ to avoid corruption of the 64bit RIP register on C stepping K8.
++ A lot of BIOS that didn't get tested properly miss this.
++ The OS sees this as a page fault with the upper 32bits of RIP cleared.
++ Try to work around it here.
++ Note we only handle faults in kernel here.
++ Does nothing for X86_32
++ */
++static int is_errata93(struct pt_regs *regs, unsigned long address)
++{
++#ifdef CONFIG_X86_64
++ static int warned;
++ if (address != regs->ip)
++ return 0;
++ if ((address >> 32) != 0)
++ return 0;
++ address |= 0xffffffffUL << 32;
++ if ((address >= (u64)_stext && address <= (u64)_etext) ||
++ (address >= MODULES_VADDR && address <= MODULES_END)) {
++ if (!warned) {
++ printk(errata93_warning);
++ warned = 1;
++ }
++ regs->ip = address;
++ return 1;
++ }
++#endif
++ return 0;
++}
++
++/*
++ * Work around K8 erratum #100 K8 in compat mode occasionally jumps to illegal
++ * addresses >4GB. We catch this in the page fault handler because these
++ * addresses are not reachable. Just detect this case and return. Any code
++ * segment in LDT is compatibility mode.
++ */
++static int is_errata100(struct pt_regs *regs, unsigned long address)
++{
++#ifdef CONFIG_X86_64
++ if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) &&
++ (address >> 32))
++ return 1;
++#endif
++ return 0;
++}
++
++void do_invalid_op(struct pt_regs *, unsigned long);
++
++static int is_f00f_bug(struct pt_regs *regs, unsigned long address)
++{
++#ifdef CONFIG_X86_F00F_BUG
++ unsigned long nr;
++ /*
++ * Pentium F0 0F C7 C8 bug workaround.
++ */
++ if (boot_cpu_data.f00f_bug) {
++ nr = (address - idt_descr.address) >> 3;
++
++ if (nr == 6) {
++ do_invalid_op(regs, 0);
++ return 1;
++ }
++ }
++#endif
++ return 0;
++}
++
++static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
++ unsigned long address)
++{
++#ifdef CONFIG_X86_32
++ if (!oops_may_print())
++ return;
++#endif
++
++#ifdef CONFIG_X86_PAE
++ if (error_code & PF_INSTR) {
++ unsigned int level;
++ pte_t *pte = lookup_address(address, &level);
++
++ if (pte && pte_present(*pte) && !pte_exec(*pte))
++ printk(KERN_CRIT "kernel tried to execute "
++ "NX-protected page - exploit attempt? "
++ "(uid: %d)\n", current->uid);
++ }
++#endif
++
++ printk(KERN_ALERT "BUG: unable to handle kernel ");
++ if (address < PAGE_SIZE)
++ printk(KERN_CONT "NULL pointer dereference");
++ else
++ printk(KERN_CONT "paging request");
++#ifdef CONFIG_X86_32
++ printk(KERN_CONT " at %08lx\n", address);
++#else
++ printk(KERN_CONT " at %016lx\n", address);
++#endif
++ printk(KERN_ALERT "IP:");
++ printk_address(regs->ip, 1);
++ dump_pagetable(address);
++}
++
++#ifdef CONFIG_X86_64
++static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
++ unsigned long error_code)
++{
++ unsigned long flags = oops_begin();
++ struct task_struct *tsk;
++
++ printk(KERN_ALERT "%s: Corrupted page table at address %lx\n",
++ current->comm, address);
++ dump_pagetable(address);
++ tsk = current;
++ tsk->thread.cr2 = address;
++ tsk->thread.trap_no = 14;
++ tsk->thread.error_code = error_code;
++ if (__die("Bad pagetable", regs, error_code))
++ regs = NULL;
++ oops_end(flags, regs, SIGKILL);
++}
++#endif
++
++static int spurious_fault_check(unsigned long error_code, pte_t *pte)
++{
++ if ((error_code & PF_WRITE) && !pte_write(*pte))
++ return 0;
++ if ((error_code & PF_INSTR) && !pte_exec(*pte))
++ return 0;
++
++ return 1;
++}
++
++/*
++ * Handle a spurious fault caused by a stale TLB entry. This allows
++ * us to lazily refresh the TLB when increasing the permissions of a
++ * kernel page (RO -> RW or NX -> X). Doing it eagerly is very
++ * expensive since that implies doing a full cross-processor TLB
++ * flush, even if no stale TLB entries exist on other processors.
++ * There are no security implications to leaving a stale TLB when
++ * increasing the permissions on a page.
++ */
++static int spurious_fault(unsigned long address,
++ unsigned long error_code)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ /* Reserved-bit violation or user access to kernel space? */
++ if (error_code & (PF_USER | PF_RSVD))
++ return 0;
++
++ pgd = init_mm.pgd + pgd_index(address);
++ if (!pgd_present(*pgd))
++ return 0;
++
++ pud = pud_offset(pgd, address);
++ if (!pud_present(*pud))
++ return 0;
++
++ if (pud_large(*pud))
++ return spurious_fault_check(error_code, (pte_t *) pud);
++
++ pmd = pmd_offset(pud, address);
++ if (!pmd_present(*pmd))
++ return 0;
++
++ if (pmd_large(*pmd))
++ return spurious_fault_check(error_code, (pte_t *) pmd);
++
++ pte = pte_offset_kernel(pmd, address);
++ if (!pte_present(*pte))
++ return 0;
++
++ return spurious_fault_check(error_code, pte);
++}
++
++/*
++ * X86_32
++ * Handle a fault on the vmalloc or module mapping area
++ *
++ * X86_64
++ * Handle a fault on the vmalloc area
++ *
++ * This assumes no large pages in there.
++ */
++static int vmalloc_fault(unsigned long address)
++{
++#ifdef CONFIG_X86_32
++ unsigned long pgd_paddr;
++ pmd_t *pmd_k;
++ pte_t *pte_k;
++
++ /* Make sure we are in vmalloc area */
++ if (!(address >= VMALLOC_START && address < VMALLOC_END))
++ return -1;
++
++ /*
++ * 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;
++#else
++ pgd_t *pgd, *pgd_ref;
++ pud_t *pud, *pud_ref;
++ pmd_t *pmd, *pmd_ref;
++ pte_t *pte, *pte_ref;
++
++ /* Make sure we are in vmalloc area */
++ if (!(address >= VMALLOC_START && address < VMALLOC_END))
++ return -1;
++
++ /* Copy kernel mappings over when needed. This can also
++ happen within a race in page table update. In the later
++ case just flush. */
++
++ /* On Xen the line below does not always work. Needs investigating! */
++ /*pgd = pgd_offset(current->mm ?: &init_mm, address);*/
++ pgd = __va(read_cr3() & PHYSICAL_PAGE_MASK);
++ pgd += pgd_index(address);
++ pgd_ref = pgd_offset_k(address);
++ if (pgd_none(*pgd_ref))
++ return -1;
++ if (pgd_none(*pgd))
++ set_pgd(pgd, *pgd_ref);
++ else
++ BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
++
++ /* Below here mismatches are bugs because these lower tables
++ are shared */
++
++ pud = pud_offset(pgd, address);
++ pud_ref = pud_offset(pgd_ref, address);
++ if (pud_none(*pud_ref))
++ return -1;
++ if (pud_none(*pud) || pud_page_vaddr(*pud) != pud_page_vaddr(*pud_ref))
++ BUG();
++ pmd = pmd_offset(pud, address);
++ pmd_ref = pmd_offset(pud_ref, address);
++ if (pmd_none(*pmd_ref))
++ return -1;
++ if (pmd_none(*pmd) || pmd_page(*pmd) != pmd_page(*pmd_ref))
++ BUG();
++ pte_ref = pte_offset_kernel(pmd_ref, address);
++ if (!pte_present(*pte_ref))
++ return -1;
++ pte = pte_offset_kernel(pmd, address);
++ /* Don't use pte_page here, because the mappings can point
++ outside mem_map, and the NUMA hash lookup cannot handle
++ that. */
++ if (!pte_present(*pte) || pte_pfn(*pte) != pte_pfn(*pte_ref))
++ BUG();
++ return 0;
++#endif
++}
++
++int show_unhandled_signals = 1;
++
++/*
++ * This routine handles page faults. It determines the address,
++ * and the problem, and then passes it off to one of the appropriate
++ * routines.
++ */
++#ifdef CONFIG_X86_64
++asmlinkage
++#endif
++void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code)
++{
++ struct task_struct *tsk;
++ struct mm_struct *mm;
++ struct vm_area_struct *vma;
++ unsigned long address;
++ int write, si_code;
++ int fault;
++#ifdef CONFIG_X86_64
++ unsigned long flags;
++#endif
++
++ /*
++ * We can fault from pretty much anywhere, with unknown IRQ state.
++ */
++ trace_hardirqs_fixup();
++
++ /* Set the "privileged fault" bit to something sane. */
++ if (user_mode_vm(regs))
++ error_code |= PF_USER;
++ else
++ error_code &= ~PF_USER;
++
++ tsk = current;
++ mm = tsk->mm;
++ prefetchw(&mm->mmap_sem);
++
++ /* get the address */
++ address = read_cr2();
++
++ si_code = SEGV_MAPERR;
++
++ if (notify_page_fault(regs))
++ return;
++
++ /*
++ * 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.
++ */
++#ifdef CONFIG_X86_32
++ if (unlikely(address >= TASK_SIZE)) {
++#else
++ if (unlikely(address >= TASK_SIZE64)) {
++#endif
++ /* Faults in hypervisor area can never be patched up. */
++#if defined(CONFIG_X86_XEN)
++ if (address >= hypervisor_virt_start)
++ goto bad_area_nosemaphore;
++#elif defined(CONFIG_X86_64_XEN)
++ /* Faults in hypervisor area are never spurious. */
++ if (address >= HYPERVISOR_VIRT_START
++ && address < HYPERVISOR_VIRT_END)
++ goto bad_area_nosemaphore;
++#endif
++ if (!(error_code & (PF_RSVD|PF_USER|PF_PROT)) &&
++ vmalloc_fault(address) >= 0)
++ return;
++
++ /* Can handle a stale RO->RW TLB */
++ if (spurious_fault(address, error_code))
++ return;
++
++ /*
++ * Don't take the mm semaphore here. If we fixup a prefetch
++ * fault we could otherwise deadlock.
++ */
++ goto bad_area_nosemaphore;
++ }
++
++
++#ifdef CONFIG_X86_32
++ /* It's safe to allow irq's after cr2 has been saved and the vmalloc
++ fault has been handled. */
++ if (regs->flags & (X86_EFLAGS_IF | X86_VM_MASK))
++ local_irq_enable();
++
++ /*
++ * 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;
++#else /* CONFIG_X86_64 */
++ if (likely(regs->flags & X86_EFLAGS_IF))
++ local_irq_enable();
++
++ if (unlikely(error_code & PF_RSVD))
++ pgtable_bad(address, 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.
++ */
++ if (unlikely(in_atomic() || !mm))
++ goto bad_area_nosemaphore;
++
++ /*
++ * User-mode registers count as a user access even for any
++ * potential system fault or CPU buglet.
++ */
++ if (user_mode_vm(regs))
++ error_code |= PF_USER;
++again:
++#endif
++ /* 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. Unfortunately, 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 possibility of a deadlock.
++ * Attempt to lock the address space, if we cannot we then validate the
++ * source. If this is invalid we can skip the address space check,
++ * thus avoiding the deadlock.
++ */
++ if (!down_read_trylock(&mm->mmap_sem)) {
++ if ((error_code & PF_USER) == 0 &&
++ !search_exception_tables(regs->ip))
++ goto bad_area_nosemaphore;
++ down_read(&mm->mmap_sem);
++ }
++
++ 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 & PF_USER) {
++ /*
++ * Accessing the stack below %sp is always a bug.
++ * The large cushion allows instructions like enter
++ * and pusha to work. ("enter $65535,$31" pushes
++ * 32 pointers and then decrements %sp by 65535.)
++ */
++ if (address + 65536 + 32 * sizeof(unsigned long) < regs->sp)
++ 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 & (PF_PROT|PF_WRITE)) {
++ default: /* 3: write, present */
++ /* fall through */
++ case PF_WRITE: /* write, not present */
++ if (!(vma->vm_flags & VM_WRITE))
++ goto bad_area;
++ write++;
++ break;
++ case PF_PROT: /* read, present */
++ goto bad_area;
++ case 0: /* read, not present */
++ if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
++ goto bad_area;
++ }
++
++#ifdef CONFIG_X86_32
++survive:
++#endif
++ /*
++ * If for any reason at all we couldn't handle the fault,
++ * make sure we exit gracefully rather than endlessly redo
++ * the fault.
++ */
++ fault = handle_mm_fault(mm, vma, address, write);
++ if (unlikely(fault & VM_FAULT_ERROR)) {
++ if (fault & VM_FAULT_OOM)
++ goto out_of_memory;
++ else if (fault & VM_FAULT_SIGBUS)
++ goto do_sigbus;
++ BUG();
++ }
++ if (fault & VM_FAULT_MAJOR)
++ tsk->maj_flt++;
++ else
++ tsk->min_flt++;
++
++#ifdef CONFIG_X86_32
++ /*
++ * Did it hit the DOS screen memory VA from vm86 mode?
++ */
++ if (v8086_mode(regs)) {
++ unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
++ if (bit < 32)
++ tsk->thread.screen_bitmap |= 1 << bit;
++ }
++#endif
++ up_read(&mm->mmap_sem);
++ return;
++
++/*
++ * Something tried to access memory that isn't in our memory map..
++ * Fix it, but check if it's kernel or user first..
++ */
++bad_area:
++ up_read(&mm->mmap_sem);
++
++bad_area_nosemaphore:
++ /* User mode accesses just cause a SIGSEGV */
++ if (error_code & PF_USER) {
++ /*
++ * It's possible to have interrupts off here.
++ */
++ local_irq_enable();
++
++ /*
++ * Valid to do another page fault here because this one came
++ * from user space.
++ */
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ if (is_errata100(regs, address))
++ return;
++
++ if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
++ printk_ratelimit()) {
++ printk(
++#ifdef CONFIG_X86_32
++ "%s%s[%d]: segfault at %lx ip %08lx sp %08lx error %lx",
++#else
++ "%s%s[%d]: segfault at %lx ip %lx sp %lx error %lx",
++#endif
++ task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
++ tsk->comm, task_pid_nr(tsk), address, regs->ip,
++ regs->sp, error_code);
++ print_vma_addr(" in ", regs->ip);
++ printk("\n");
++ }
++
++ 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;
++ }
++
++ if (is_f00f_bug(regs, address))
++ return;
++
++no_context:
++ /* Are we prepared to handle this kernel fault? */
++ if (fixup_exception(regs))
++ return;
++
++ /*
++ * X86_32
++ * Valid to do another page fault here, because if this fault
++ * had been triggered by is_prefetch fixup_exception would have
++ * handled it.
++ *
++ * X86_64
++ * Hall of shame of CPU/BIOS bugs.
++ */
++ if (is_prefetch(regs, address, error_code))
++ return;
++
++ if (is_errata93(regs, address))
++ return;
++
++/*
++ * Oops. The kernel tried to access some bad page. We'll have to
++ * terminate things with extreme prejudice.
++ */
++#ifdef CONFIG_X86_32
++ bust_spinlocks(1);
++#else
++ flags = oops_begin();
++#endif
++
++ show_fault_oops(regs, error_code, address);
++
++ tsk->thread.cr2 = address;
++ tsk->thread.trap_no = 14;
++ tsk->thread.error_code = error_code;
++
++#ifdef CONFIG_X86_32
++ die("Oops", regs, error_code);
++ bust_spinlocks(0);
++ do_exit(SIGKILL);
++#else
++ if (__die("Oops", regs, error_code))
++ regs = NULL;
++ /* Executive summary in case the body of the oops scrolled away */
++ printk(KERN_EMERG "CR2: %016lx\n", address);
++ oops_end(flags, regs, SIGKILL);
++#endif
++
++/*
++ * 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_global_init(tsk)) {
++ yield();
++#ifdef CONFIG_X86_32
++ down_read(&mm->mmap_sem);
++ goto survive;
++#else
++ goto again;
++#endif
++ }
++
++ printk("VM: killing process %s\n", tsk->comm);
++ if (error_code & PF_USER)
++ do_group_exit(SIGKILL);
++ goto no_context;
++
++do_sigbus:
++ up_read(&mm->mmap_sem);
++
++ /* Kernel mode? Handle exceptions or die */
++ if (!(error_code & PF_USER))
++ goto no_context;
++#ifdef CONFIG_X86_32
++ /* User space => ok to do another page fault */
++ if (is_prefetch(regs, address, error_code))
++ return;
++#endif
++ tsk->thread.cr2 = address;
++ tsk->thread.error_code = error_code;
++ tsk->thread.trap_no = 14;
++ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
++}
++
++DEFINE_SPINLOCK(pgd_lock);
++LIST_HEAD(pgd_list);
++
++void vmalloc_sync_all(void)
++{
++#ifdef CONFIG_X86_32
++ /*
++ * Note that races in the updates of insync and start aren't
++ * problematic: insync can only get set bits added, and updates to
++ * start are only improving performance (without affecting correctness
++ * if undone).
++ * XEN: To work on PAE, we need to iterate over PMDs rather than PGDs.
++ * This change works just fine with 2-level paging too.
++ */
++#define sync_index(a) ((a) >> PMD_SHIFT)
++ static DECLARE_BITMAP(insync, PTRS_PER_PGD*PTRS_PER_PMD);
++ static unsigned long start = TASK_SIZE;
++ unsigned long address;
++
++ if (SHARED_KERNEL_PMD)
++ return;
++
++ BUILD_BUG_ON(TASK_SIZE & ~PMD_MASK);
++ for (address = start;
++ address < hypervisor_virt_start;
++ address += PMD_SIZE) {
++ if (!test_bit(sync_index(address), insync)) {
++ unsigned long flags;
++ struct page *page;
++
++ spin_lock_irqsave(&pgd_lock, flags);
++ /* XEN: failure path assumes non-empty pgd_list. */
++ if (unlikely(list_empty(&pgd_list))) {
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ return;
++ }
++ list_for_each_entry(page, &pgd_list, lru) {
++ if (!vmalloc_sync_one(page_address(page),
++ address))
++ break;
++ }
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ if (!page)
++ set_bit(sync_index(address), insync);
++ }
++ if (address == start && test_bit(sync_index(address), insync))
++ start = address + PMD_SIZE;
++ }
++#else /* CONFIG_X86_64 */
++ /*
++ * Note that races in the updates of insync and start aren't
++ * problematic: insync can only get set bits added, and updates to
++ * start are only improving performance (without affecting correctness
++ * if undone).
++ */
++ static DECLARE_BITMAP(insync, PTRS_PER_PGD);
++ static unsigned long start = VMALLOC_START & PGDIR_MASK;
++ unsigned long address;
++
++ for (address = start; address <= VMALLOC_END; address += PGDIR_SIZE) {
++ if (!test_bit(pgd_index(address), insync)) {
++ const pgd_t *pgd_ref = pgd_offset_k(address);
++ unsigned long flags;
++ struct page *page;
++
++ if (pgd_none(*pgd_ref))
++ continue;
++ spin_lock_irqsave(&pgd_lock, flags);
++ list_for_each_entry(page, &pgd_list, lru) {
++ pgd_t *pgd;
++ pgd = (pgd_t *)page_address(page) + pgd_index(address);
++ if (pgd_none(*pgd))
++ set_pgd(pgd, *pgd_ref);
++ else
++ BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
++ }
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ set_bit(pgd_index(address), insync);
++ }
++ if (address == start)
++ start = address + PGDIR_SIZE;
++ }
++#endif
++}
+diff --git a/arch/x86/mm/highmem_32-xen.c b/arch/x86/mm/highmem_32-xen.c
+new file mode 100644
+index 0000000..cdd792f
+--- /dev/null
++++ b/arch/x86/mm/highmem_32-xen.c
+@@ -0,0 +1,157 @@
++#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);
++}
++
++static void debug_kmap_atomic_prot(enum km_type type)
++{
++#ifdef CONFIG_DEBUG_HIGHMEM
++ static unsigned warn_count = 10;
++
++ if (unlikely(warn_count == 0))
++ return;
++
++ if (unlikely(in_interrupt())) {
++ if (in_irq()) {
++ if (type != KM_IRQ0 && type != KM_IRQ1 &&
++ type != KM_BIO_SRC_IRQ && type != KM_BIO_DST_IRQ &&
++ type != KM_BOUNCE_READ) {
++ WARN_ON(1);
++ warn_count--;
++ }
++ } else if (!irqs_disabled()) { /* softirq */
++ if (type != KM_IRQ0 && type != KM_IRQ1 &&
++ type != KM_SOFTIRQ0 && type != KM_SOFTIRQ1 &&
++ type != KM_SKB_SUNRPC_DATA &&
++ type != KM_SKB_DATA_SOFTIRQ &&
++ type != KM_BOUNCE_READ) {
++ WARN_ON(1);
++ warn_count--;
++ }
++ }
++ }
++
++ if (type == KM_IRQ0 || type == KM_IRQ1 || type == KM_BOUNCE_READ ||
++ type == KM_BIO_SRC_IRQ || type == KM_BIO_DST_IRQ) {
++ if (!irqs_disabled()) {
++ WARN_ON(1);
++ warn_count--;
++ }
++ } else if (type == KM_SOFTIRQ0 || type == KM_SOFTIRQ1) {
++ if (irq_count() == 0 && !irqs_disabled()) {
++ WARN_ON(1);
++ warn_count--;
++ }
++ }
++#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.
++ *
++ * However when holding an atomic kmap is is not legal to sleep, so atomic
++ * kmaps are appropriate for short, tight code paths only.
++ */
++void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot)
++{
++ enum fixed_addresses idx;
++ unsigned long vaddr;
++
++ /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */
++ pagefault_disable();
++
++ if (!PageHighMem(page))
++ return page_address(page);
++
++ debug_kmap_atomic_prot(type);
++
++ idx = type + KM_TYPE_NR*smp_processor_id();
++ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
++ BUG_ON(!pte_none(*(kmap_pte-idx)));
++ set_pte_at(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
++ arch_flush_lazy_mmu_mode();
++
++ return (void *)vaddr;
++}
++
++void *kmap_atomic(struct page *page, enum km_type type)
++{
++ return kmap_atomic_prot(page, type, kmap_prot);
++}
++
++void kunmap_atomic(void *kvaddr, enum km_type type)
++{
++ unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
++ enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id();
++
++ /*
++ * 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);
++ else {
++#ifdef CONFIG_DEBUG_HIGHMEM
++ BUG_ON(vaddr < PAGE_OFFSET);
++ BUG_ON(vaddr >= (unsigned long)high_memory);
++#endif
++ }
++
++ arch_flush_lazy_mmu_mode();
++ pagefault_enable();
++}
++
++/* 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;
++
++ pagefault_disable();
++
++ 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));
++ arch_flush_lazy_mmu_mode();
++
++ return (void*) vaddr;
++}
++
++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);
+diff --git a/arch/x86/mm/hypervisor.c b/arch/x86/mm/hypervisor.c
+new file mode 100644
+index 0000000..6ce5ab0
+--- /dev/null
++++ b/arch/x86/mm/hypervisor.c
+@@ -0,0 +1,603 @@
++/******************************************************************************
++ * 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>
++#include <linux/highmem.h>
++
++void xen_l1_entry_update(pte_t *ptr, pte_t val)
++{
++ mmu_update_t u;
++#ifdef CONFIG_HIGHPTE
++ u.ptr = ((unsigned long)ptr >= (unsigned long)high_memory) ?
++ arbitrary_virt_to_machine(ptr) : virt_to_machine(ptr);
++#else
++ u.ptr = virt_to_machine(ptr);
++#endif
++ u.val = __pte_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
++EXPORT_SYMBOL_GPL(xen_l1_entry_update);
++
++void xen_l2_entry_update(pmd_t *ptr, pmd_t val)
++{
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = __pmd_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
++
++#if defined(CONFIG_X86_PAE) || defined(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 = __pud_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
++#endif
++
++#ifdef CONFIG_X86_64
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val)
++{
++ mmu_update_t u;
++ u.ptr = virt_to_machine(ptr);
++ u.val = __pgd_val(val);
++ BUG_ON(HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF) < 0);
++}
++#endif /* CONFIG_X86_64 */
++
++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);
++}
++
++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);
++}
++
++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);
++
++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);
++
++#ifdef CONFIG_SMP
++
++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);
++}
++
++void xen_tlb_flush_mask(cpumask_t *mask)
++{
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_TLB_FLUSH_MULTI;
++ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++
++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);
++}
++
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr)
++{
++ struct mmuext_op op;
++ if ( cpus_empty(*mask) )
++ return;
++ op.cmd = MMUEXT_INVLPG_MULTI;
++ op.arg1.linear_addr = ptr & PAGE_MASK;
++ set_xen_guest_handle(op.arg2.vcpumask, mask->bits);
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++
++#endif /* CONFIG_SMP */
++
++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);
++}
++
++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(const void *ptr, unsigned int ents)
++{
++ struct mmuext_op op;
++ op.cmd = MMUEXT_SET_LDT;
++ op.arg1.linear_addr = (unsigned long)ptr;
++ op.arg2.nr_ents = ents;
++ BUG_ON(HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF) < 0);
++}
++
++/*
++ * Bitmap is indexed by page number. If bit is set, the page is part of a
++ * xen_create_contiguous_region() area of memory.
++ */
++unsigned long *contiguous_bitmap;
++
++static void contiguous_bitmap_set(
++ unsigned long first_page, unsigned long nr_pages)
++{
++ unsigned long start_off, end_off, curr_idx, end_idx;
++
++ curr_idx = first_page / BITS_PER_LONG;
++ start_off = first_page & (BITS_PER_LONG-1);
++ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
++ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
++
++ if (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;
++ }
++}
++
++static void contiguous_bitmap_clear(
++ unsigned long first_page, unsigned long nr_pages)
++{
++ unsigned long start_off, end_off, curr_idx, end_idx;
++
++ curr_idx = first_page / BITS_PER_LONG;
++ start_off = first_page & (BITS_PER_LONG-1);
++ end_idx = (first_page + nr_pages) / BITS_PER_LONG;
++ end_off = (first_page + nr_pages) & (BITS_PER_LONG-1);
++
++ if (curr_idx == end_idx) {
++ contiguous_bitmap[curr_idx] &=
++ -(1UL<<end_off) | ((1UL<<start_off)-1);
++ } else {
++ contiguous_bitmap[curr_idx] &= (1UL<<start_off)-1;
++ while ( ++curr_idx != end_idx )
++ contiguous_bitmap[curr_idx] = 0;
++ contiguous_bitmap[curr_idx] &= -(1UL<<end_off);
++ }
++}
++
++/* Protected by balloon_lock. */
++#define MAX_CONTIG_ORDER 9 /* 2MB */
++static unsigned long discontig_frames[1<<MAX_CONTIG_ORDER];
++static unsigned long limited_frames[1<<MAX_CONTIG_ORDER];
++static multicall_entry_t cr_mcl[1<<MAX_CONTIG_ORDER];
++
++/* Ensure multi-page extents are contiguous in machine memory. */
++int xen_create_contiguous_region(
++ unsigned long vstart, unsigned int order, unsigned int address_bits)
++{
++ unsigned long *in_frames = discontig_frames, out_frame;
++ unsigned long frame, flags;
++ unsigned int i;
++ int rc, success;
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .nr_extents = 1UL << order,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .nr_extents = 1,
++ .extent_order = order,
++ .address_bits = address_bits,
++ .domid = DOMID_SELF
++ }
++ };
++
++ /*
++ * Currently an auto-translated guest will not perform I/O, nor will
++ * it require PAE page directories below 4GB. Therefore any calls to
++ * this function are redundant and can be ignored.
++ */
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 0;
++
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return -ENOMEM;
++
++ set_xen_guest_handle(exchange.in.extent_start, in_frames);
++ set_xen_guest_handle(exchange.out.extent_start, &out_frame);
++
++ scrub_pages((void *)vstart, 1 << order);
++
++ balloon_lock(flags);
++
++ /* 1. Zap current PTEs, remembering MFNs. */
++ for (i = 0; i < (1U<<order); i++) {
++ in_frames[i] = pfn_to_mfn((__pa(vstart) >> PAGE_SHIFT) + i);
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ __pte_ma(0), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
++ INVALID_P2M_ENTRY);
++ }
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
++
++ /* 2. Get a new contiguous memory extent. */
++ out_frame = __pa(vstart) >> PAGE_SHIFT;
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == (1UL << order));
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != (1UL << order))
++ BUG();
++ success = (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) == 1);
++ if (!success) {
++ /* Couldn't get special memory: fall back to normal. */
++ for (i = 0; i < (1U<<order); i++)
++ in_frames[i] = (__pa(vstart)>>PAGE_SHIFT) + i;
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.in) != (1UL<<order))
++ BUG();
++ }
++ }
++#endif
++
++ /* 3. Map the new extent in place of old pages. */
++ for (i = 0; i < (1U<<order); i++) {
++ frame = success ? (out_frame + i) : in_frames[i];
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
++ }
++
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
++
++ if (success)
++ contiguous_bitmap_set(__pa(vstart) >> PAGE_SHIFT,
++ 1UL << order);
++
++ balloon_unlock(flags);
++
++ return success ? 0 : -ENOMEM;
++}
++EXPORT_SYMBOL_GPL(xen_create_contiguous_region);
++
++void xen_destroy_contiguous_region(unsigned long vstart, unsigned int order)
++{
++ unsigned long *out_frames = discontig_frames, in_frame;
++ unsigned long frame, flags;
++ unsigned int i;
++ int rc, success;
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .nr_extents = 1,
++ .extent_order = order,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .nr_extents = 1UL << order,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ }
++ };
++
++ if (xen_feature(XENFEAT_auto_translated_physmap) ||
++ !test_bit(__pa(vstart) >> PAGE_SHIFT, contiguous_bitmap))
++ return;
++
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return;
++
++ set_xen_guest_handle(exchange.in.extent_start, &in_frame);
++ set_xen_guest_handle(exchange.out.extent_start, out_frames);
++
++ scrub_pages((void *)vstart, 1 << order);
++
++ balloon_lock(flags);
++
++ contiguous_bitmap_clear(__pa(vstart) >> PAGE_SHIFT, 1UL << order);
++
++ /* 1. Find start MFN of contiguous extent. */
++ in_frame = pfn_to_mfn(__pa(vstart) >> PAGE_SHIFT);
++
++ /* 2. Zap current PTEs. */
++ for (i = 0; i < (1U<<order); i++) {
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ __pte_ma(0), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i,
++ INVALID_P2M_ENTRY);
++ out_frames[i] = (__pa(vstart) >> PAGE_SHIFT) + i;
++ }
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
++
++ /* 3. Do the exchange for non-contiguous MFNs. */
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == 1);
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != 1)
++ BUG();
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) != (1UL << order))
++ BUG();
++ success = 1;
++ }
++#endif
++
++ /* 4. Map new pages in place of old pages. */
++ for (i = 0; i < (1U<<order); i++) {
++ frame = success ? out_frames[i] : (in_frame + i);
++ MULTI_update_va_mapping(cr_mcl + i, vstart + (i*PAGE_SIZE),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++ set_phys_to_machine((__pa(vstart)>>PAGE_SHIFT)+i, frame);
++ }
++
++ cr_mcl[i - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, i, NULL))
++ BUG();
++
++ balloon_unlock(flags);
++}
++EXPORT_SYMBOL_GPL(xen_destroy_contiguous_region);
++
++int xen_limit_pages_to_max_mfn(
++ struct page *pages, unsigned int order, unsigned int address_bits)
++{
++ unsigned long flags, frame;
++ unsigned long *in_frames = discontig_frames, *out_frames = limited_frames;
++ struct page *page;
++ unsigned int i, n, nr_mcl;
++ int rc, success;
++ DECLARE_BITMAP(limit_map, 1 << MAX_CONTIG_ORDER);
++
++ struct xen_memory_exchange exchange = {
++ .in = {
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ },
++ .out = {
++ .extent_order = 0,
++ .address_bits = address_bits,
++ .domid = DOMID_SELF
++ }
++ };
++
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 0;
++
++ if (unlikely(order > MAX_CONTIG_ORDER))
++ return -ENOMEM;
++
++ bitmap_zero(limit_map, 1U << order);
++ set_xen_guest_handle(exchange.in.extent_start, in_frames);
++ set_xen_guest_handle(exchange.out.extent_start, out_frames);
++
++ /* 0. Scrub the pages. */
++ for (i = 0, n = 0; i < 1U<<order ; i++) {
++ page = &pages[i];
++ if (!(pfn_to_mfn(page_to_pfn(page)) >> (address_bits - PAGE_SHIFT)))
++ continue;
++ __set_bit(i, limit_map);
++
++ if (!PageHighMem(page))
++ scrub_pages(page_address(page), 1);
++#ifdef CONFIG_XEN_SCRUB_PAGES
++ else {
++ scrub_pages(kmap(page), 1);
++ kunmap(page);
++ ++n;
++ }
++#endif
++ }
++ if (bitmap_empty(limit_map, 1U << order))
++ return 0;
++
++ if (n)
++ kmap_flush_unused();
++
++ balloon_lock(flags);
++
++ /* 1. Zap current PTEs (if any), remembering MFNs. */
++ for (i = 0, n = 0, nr_mcl = 0; i < (1U<<order); i++) {
++ if(!test_bit(i, limit_map))
++ continue;
++ page = &pages[i];
++
++ out_frames[n] = page_to_pfn(page);
++ in_frames[n] = pfn_to_mfn(out_frames[n]);
++
++ if (!PageHighMem(page))
++ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
++ (unsigned long)page_address(page),
++ __pte_ma(0), 0);
++
++ set_phys_to_machine(out_frames[n], INVALID_P2M_ENTRY);
++ ++n;
++ }
++ if (nr_mcl && HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
++ BUG();
++
++ /* 2. Get new memory below the required limit. */
++ exchange.in.nr_extents = n;
++ exchange.out.nr_extents = n;
++ rc = HYPERVISOR_memory_op(XENMEM_exchange, &exchange);
++ success = (exchange.nr_exchanged == n);
++ BUG_ON(!success && ((exchange.nr_exchanged != 0) || (rc == 0)));
++ BUG_ON(success && (rc != 0));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ /* Compatibility when XENMEM_exchange is unsupported. */
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &exchange.in) != n)
++ BUG();
++ if (HYPERVISOR_memory_op(XENMEM_populate_physmap,
++ &exchange.out) != n)
++ BUG();
++ success = 1;
++ }
++#endif
++
++ /* 3. Map the new pages in place of old pages. */
++ for (i = 0, n = 0, nr_mcl = 0; i < (1U<<order); i++) {
++ if(!test_bit(i, limit_map))
++ continue;
++ page = &pages[i];
++
++ frame = success ? out_frames[n] : in_frames[n];
++
++ if (!PageHighMem(page))
++ MULTI_update_va_mapping(cr_mcl + nr_mcl++,
++ (unsigned long)page_address(page),
++ pfn_pte_ma(frame, PAGE_KERNEL), 0);
++
++ set_phys_to_machine(page_to_pfn(page), frame);
++ ++n;
++ }
++ if (nr_mcl) {
++ cr_mcl[nr_mcl - 1].args[MULTI_UVMFLAGS_INDEX] = order
++ ? UVMF_TLB_FLUSH|UVMF_ALL
++ : UVMF_INVLPG|UVMF_ALL;
++ if (HYPERVISOR_multicall_check(cr_mcl, nr_mcl, NULL))
++ BUG();
++ }
++
++ balloon_unlock(flags);
++
++ return success ? 0 : -ENOMEM;
++}
++EXPORT_SYMBOL_GPL(xen_limit_pages_to_max_mfn);
++
++int write_ldt_entry(struct desc_struct *ldt, int entry, const void *desc)
++{
++ maddr_t mach_lp = arbitrary_virt_to_machine(ldt + entry);
++ return HYPERVISOR_update_descriptor(mach_lp, *(const u64*)desc);
++}
++
++#define MAX_BATCHED_FULL_PTES 32
++
++int xen_change_pte_range(struct mm_struct *mm, pmd_t *pmd,
++ unsigned long addr, unsigned long end, pgprot_t newprot,
++ int dirty_accountable)
++{
++ int rc = 0, i = 0;
++ mmu_update_t u[MAX_BATCHED_FULL_PTES];
++ pte_t *pte;
++ spinlock_t *ptl;
++
++ if (!xen_feature(XENFEAT_mmu_pt_update_preserve_ad))
++ return 0;
++
++ pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
++ do {
++ if (pte_present(*pte)) {
++ pte_t ptent = pte_modify(*pte, newprot);
++
++ if (dirty_accountable && pte_dirty(ptent))
++ ptent = pte_mkwrite(ptent);
++ u[i].ptr = (__pmd_val(*pmd) & PHYSICAL_PAGE_MASK)
++ | ((unsigned long)pte & ~PAGE_MASK)
++ | MMU_PT_UPDATE_PRESERVE_AD;
++ u[i].val = __pte_val(ptent);
++ if (++i == MAX_BATCHED_FULL_PTES) {
++ if ((rc = HYPERVISOR_mmu_update(
++ &u[0], i, NULL, DOMID_SELF)) != 0)
++ break;
++ i = 0;
++ }
++ }
++ } while (pte++, addr += PAGE_SIZE, addr != end);
++ if (i)
++ rc = HYPERVISOR_mmu_update( &u[0], i, NULL, DOMID_SELF);
++ pte_unmap_unlock(pte - 1, ptl);
++ BUG_ON(rc && rc != -ENOSYS);
++ return !rc;
++}
+diff --git a/arch/x86/mm/init_32-xen.c b/arch/x86/mm/init_32-xen.c
+new file mode 100644
+index 0000000..d76241c
+--- /dev/null
++++ b/arch/x86/mm/init_32-xen.c
+@@ -0,0 +1,838 @@
++/*
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ */
++
++#include <linux/module.h>
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/hugetlb.h>
++#include <linux/swap.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/highmem.h>
++#include <linux/pagemap.h>
++#include <linux/pci.h>
++#include <linux/pfn.h>
++#include <linux/poison.h>
++#include <linux/bootmem.h>
++#include <linux/slab.h>
++#include <linux/proc_fs.h>
++#include <linux/memory_hotplug.h>
++#include <linux/initrd.h>
++#include <linux/cpumask.h>
++#include <linux/dma-mapping.h>
++#include <linux/scatterlist.h>
++
++#include <asm/asm.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/bugs.h>
++#include <asm/tlb.h>
++#include <asm/tlbflush.h>
++#include <asm/pgalloc.h>
++#include <asm/sections.h>
++#include <asm/hypervisor.h>
++#include <asm/swiotlb.h>
++#include <asm/setup.h>
++#include <asm/cacheflush.h>
++
++extern unsigned long *contiguous_bitmap;
++
++unsigned int __VMALLOC_RESERVE = 128 << 20;
++
++unsigned long max_pfn_mapped;
++
++DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
++unsigned long highstart_pfn, highend_pfn;
++
++static noinline int do_test_wp_bit(void);
++
++/*
++ * Creates a middle page table and puts a pointer to it in the
++ * given global directory entry. This only returns the gd entry
++ * in non-PAE compilation mode, since the middle layer is folded.
++ */
++static pmd_t * __init one_md_table_init(pgd_t *pgd)
++{
++ pud_t *pud;
++ pmd_t *pmd_table;
++
++#ifdef CONFIG_X86_PAE
++ if (!(__pgd_val(*pgd) & _PAGE_PRESENT)) {
++ pmd_table = (pmd_t *) alloc_bootmem_low_pages(PAGE_SIZE);
++
++ paravirt_alloc_pmd(&init_mm, __pa(pmd_table) >> PAGE_SHIFT);
++ make_lowmem_page_readonly(pmd_table, XENFEAT_writable_page_tables);
++ set_pgd(pgd, __pgd(__pa(pmd_table) | _PAGE_PRESENT));
++ pud = pud_offset(pgd, 0);
++ BUG_ON(pmd_table != pmd_offset(pud, 0));
++ }
++#endif
++ pud = pud_offset(pgd, 0);
++ pmd_table = pmd_offset(pud, 0);
++
++ return pmd_table;
++}
++
++/*
++ * Create a page table and place a pointer to it in a middle page
++ * directory entry:
++ */
++static pte_t * __init one_page_table_init(pmd_t *pmd)
++{
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (pmd_none(*pmd)) {
++#else
++ if (!(__pmd_val(*pmd) & _PAGE_PRESENT)) {
++#endif
++ pte_t *page_table = NULL;
++
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ page_table = (pte_t *) alloc_bootmem_pages(PAGE_SIZE);
++#endif
++ if (!page_table) {
++ page_table =
++ (pte_t *)alloc_bootmem_low_pages(PAGE_SIZE);
++ }
++
++ paravirt_alloc_pte(&init_mm, __pa(page_table) >> PAGE_SHIFT);
++ make_lowmem_page_readonly(page_table,
++ XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(page_table) | _PAGE_TABLE));
++ BUG_ON(page_table != pte_offset_kernel(pmd, 0));
++ }
++
++ return pte_offset_kernel(pmd, 0);
++}
++
++/*
++ * This function initializes a certain range of kernel virtual memory
++ * with new bootmem page tables, everywhere page tables are missing in
++ * the given range.
++ *
++ * NOTE: The pagetables are allocated contiguous on the physical space
++ * so we can cache the place of the first one and move around without
++ * checking the pgd every time.
++ */
++static void __init
++page_table_range_init(unsigned long start, unsigned long end, pgd_t *pgd_base)
++{
++ int pgd_idx, pmd_idx;
++ unsigned long vaddr;
++ pgd_t *pgd;
++ pmd_t *pmd;
++
++ vaddr = start;
++ pgd_idx = pgd_index(vaddr);
++ pmd_idx = pmd_index(vaddr);
++ pgd = pgd_base + pgd_idx;
++
++ for ( ; (pgd_idx < PTRS_PER_PGD) && (vaddr != end); pgd++, pgd_idx++) {
++ pmd = one_md_table_init(pgd);
++ pmd = pmd + pmd_index(vaddr);
++ for (; (pmd_idx < PTRS_PER_PMD) && (vaddr != end);
++ pmd++, pmd_idx++) {
++ if (vaddr < hypervisor_virt_start)
++ one_page_table_init(pmd);
++
++ vaddr += PMD_SIZE;
++ }
++ pmd_idx = 0;
++ }
++}
++
++static inline int is_kernel_text(unsigned long addr)
++{
++ if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
++ return 1;
++ return 0;
++}
++
++/*
++ * 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)
++{
++ int pgd_idx, pmd_idx, pte_ofs;
++ unsigned long pfn;
++ pgd_t *pgd;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ unsigned long max_ram_pfn = xen_start_info->nr_pages;
++ if (max_ram_pfn > max_low_pfn)
++ max_ram_pfn = max_low_pfn;
++
++ pgd_idx = pgd_index(PAGE_OFFSET);
++ pgd = pgd_base + pgd_idx;
++ pfn = 0;
++ pmd_idx = pmd_index(PAGE_OFFSET);
++ pte_ofs = pte_index(PAGE_OFFSET);
++
++ 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 addr = pfn * PAGE_SIZE + PAGE_OFFSET;
++
++ if (addr >= hypervisor_virt_start)
++ continue;
++
++ /*
++ * Map with big pages if possible, otherwise
++ * create normal page tables:
++ *
++ * Don't use a large page for the first 2/4MB of memory
++ * because there are often fixed size MTRRs in there
++ * and overlapping MTRRs into large pages can cause
++ * slowdowns.
++ */
++ if (cpu_has_pse && !(pgd_idx == 0 && pmd_idx == 0)) {
++ unsigned int addr2;
++ pgprot_t prot = PAGE_KERNEL_LARGE;
++
++ addr2 = (pfn + PTRS_PER_PTE-1) * PAGE_SIZE +
++ PAGE_OFFSET + PAGE_SIZE-1;
++
++ if (is_kernel_text(addr) ||
++ is_kernel_text(addr2))
++ prot = PAGE_KERNEL_LARGE_EXEC;
++
++ set_pmd(pmd, pfn_pmd(pfn, prot));
++
++ pfn += PTRS_PER_PTE;
++ max_pfn_mapped = pfn;
++ continue;
++ }
++ pte = one_page_table_init(pmd);
++
++ for (pte += pte_ofs;
++ pte_ofs < PTRS_PER_PTE && pfn < max_low_pfn;
++ pte++, pfn++, pte_ofs++, addr += PAGE_SIZE) {
++ pgprot_t prot = PAGE_KERNEL;
++
++ /* XEN: Only map initial RAM allocation. */
++ if ((pfn >= max_ram_pfn) || pte_present(*pte))
++ continue;
++ if (is_kernel_text(addr))
++ prot = PAGE_KERNEL_EXEC;
++
++ set_pte(pte, pfn_pte(pfn, prot));
++ }
++ max_pfn_mapped = pfn;
++ pte_ofs = 0;
++ }
++ pmd_idx = 0;
++ }
++}
++
++#ifndef CONFIG_XEN
++
++static inline int page_kills_ppro(unsigned long pagenr)
++{
++ if (pagenr >= 0x70000 && pagenr <= 0x7003F)
++ return 1;
++ return 0;
++}
++
++#else
++
++#define page_kills_ppro(p) 0
++
++#endif
++
++/*
++ * devmem_is_allowed() checks to see if /dev/mem access to a certain address
++ * is valid. The argument is a physical page number.
++ *
++ *
++ * On x86, access has to be given to the first megabyte of ram because that area
++ * contains bios code and data regions used by X and dosemu and similar apps.
++ * Access has to be given to non-kernel-ram areas as well, these contain the PCI
++ * mmio resources as well as potential bios/acpi data regions.
++ */
++int devmem_is_allowed(unsigned long pagenr)
++{
++ if (pagenr <= 256)
++ return 1;
++ if (mfn_to_local_pfn(pagenr) >= max_pfn)
++ return 1;
++ return 0;
++}
++
++#ifdef CONFIG_HIGHMEM
++pte_t *kmap_pte;
++pgprot_t kmap_prot;
++
++static inline pte_t *kmap_get_fixmap_pte(unsigned long vaddr)
++{
++ return pte_offset_kernel(pmd_offset(pud_offset(pgd_offset_k(vaddr),
++ vaddr), vaddr), vaddr);
++}
++
++static void __init kmap_init(void)
++{
++ unsigned long kmap_vstart;
++
++ /*
++ * Cache the first kmap pte:
++ */
++ kmap_vstart = __fix_to_virt(FIX_KMAP_BEGIN);
++ kmap_pte = kmap_get_fixmap_pte(kmap_vstart);
++
++ kmap_prot = PAGE_KERNEL;
++}
++
++static void __init permanent_kmaps_init(pgd_t *pgd_base)
++{
++ unsigned long vaddr;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ vaddr = PKMAP_BASE;
++ page_table_range_init(vaddr, vaddr + PAGE_SIZE*LAST_PKMAP, pgd_base);
++
++ 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;
++}
++
++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);
++ init_page_count(page);
++ if (pfn < xen_start_info->nr_pages)
++ __free_page(page);
++ totalhigh_pages++;
++ } else
++ SetPageReserved(page);
++}
++
++#ifndef CONFIG_NUMA
++static void __init set_highmem_pages_init(int bad_ppro)
++{
++ int pfn;
++
++ for (pfn = highstart_pfn; pfn < highend_pfn; pfn++) {
++ /*
++ * Holes under sparsemem might not have no mem_map[]:
++ */
++ if (pfn_valid(pfn))
++ add_one_highpage_init(pfn_to_page(pfn), pfn, bad_ppro);
++ }
++ totalram_pages += totalhigh_pages;
++}
++#endif /* !CONFIG_NUMA */
++
++#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 */
++
++pteval_t __PAGE_KERNEL = _PAGE_KERNEL;
++EXPORT_SYMBOL(__PAGE_KERNEL);
++
++pteval_t __PAGE_KERNEL_EXEC = _PAGE_KERNEL_EXEC;
++
++pgd_t *swapper_pg_dir;
++
++static void __init xen_pagetable_setup_start(pgd_t *base)
++{
++}
++
++static void __init xen_pagetable_setup_done(pgd_t *base)
++{
++}
++
++/*
++ * Build a proper pagetable for the kernel mappings. Up until this
++ * point, we've been running on some set of pagetables constructed by
++ * the boot process.
++ *
++ * If we're booting on native hardware, this will be a pagetable
++ * constructed in arch/x86/kernel/head_32.S. The root of the
++ * pagetable will be swapper_pg_dir.
++ *
++ * If we're booting paravirtualized under a hypervisor, then there are
++ * more options: we may already be running PAE, and the pagetable may
++ * or may not be based in swapper_pg_dir. In any case,
++ * paravirt_pagetable_setup_start() will set up swapper_pg_dir
++ * appropriately for the rest of the initialization to work.
++ *
++ * In general, pagetable_init() assumes that the pagetable may already
++ * be partially populated, and so it avoids stomping on any existing
++ * mappings.
++ */
++static void __init pagetable_init(void)
++{
++ pgd_t *pgd_base = (pgd_t *)xen_start_info->pt_base;
++ unsigned long vaddr, end;
++
++ xen_pagetable_setup_start(pgd_base);
++
++ /* Enable PSE if available */
++ if (cpu_has_pse)
++ set_in_cr4(X86_CR4_PSE);
++
++ /* Enable PGE if available */
++ if (cpu_has_pge) {
++ set_in_cr4(X86_CR4_PGE);
++ __PAGE_KERNEL |= _PAGE_GLOBAL;
++ __PAGE_KERNEL_EXEC |= _PAGE_GLOBAL;
++ }
++
++ kernel_physical_mapping_init(pgd_base);
++ remap_numa_kva();
++
++ /*
++ * Fixed mappings, only the page table structure has to be
++ * created - mappings will be set by set_fixmap():
++ */
++ early_ioremap_clear();
++ vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
++ end = (FIXADDR_TOP + PMD_SIZE - 1) & PMD_MASK;
++ page_table_range_init(vaddr, end, pgd_base);
++ early_ioremap_reset();
++
++ permanent_kmaps_init(pgd_base);
++
++ xen_pagetable_setup_done(pgd_base);
++}
++
++#if defined(CONFIG_ACPI_SLEEP) && !defined(CONFIG_XEN)
++/*
++ * ACPI suspend needs this for resume, because things like the intel-agp
++ * driver might have split up a kernel 4MB mapping.
++ */
++char 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 /* !CONFIG_ACPI_SLEEP */
++static inline void save_pg_dir(void)
++{
++}
++#endif /* !CONFIG_ACPI_SLEEP */
++
++void zap_low_mappings(void)
++{
++ int i;
++
++ /*
++ * Zap initial low-memory mappings.
++ *
++ * Note that "pgd_clear()" doesn't do it for
++ * us, because pgd_clear() is a no-op on i386.
++ */
++ for (i = 0; i < KERNEL_PGD_BOUNDARY; 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();
++}
++
++int nx_enabled;
++
++pteval_t __supported_pte_mask __read_mostly = ~_PAGE_NX;
++EXPORT_SYMBOL_GPL(__supported_pte_mask);
++
++#ifdef CONFIG_X86_PAE
++
++static int disable_nx __initdata;
++
++/*
++ * noexec = on|off
++ *
++ * Control non executable mappings.
++ *
++ * on Enable
++ * off Disable
++ */
++static int __init noexec_setup(char *str)
++{
++ 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;
++ }
++ }
++
++ return 0;
++}
++early_param("noexec", noexec_setup);
++
++static void __init set_nx(void)
++{
++ unsigned int v[4], l, h;
++
++ if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) {
++ cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]);
++
++ if ((v[3] & (1 << 20)) && !disable_nx) {
++ rdmsr(MSR_EFER, l, h);
++ l |= EFER_NX;
++ wrmsr(MSR_EFER, l, h);
++ nx_enabled = 1;
++ __supported_pte_mask |= _PAGE_NX;
++ }
++ }
++}
++#endif
++
++/*
++ * paging_init() sets up the page tables - note that the first 8MB are
++ * already mapped by head.S.
++ *
++ * This routines also unmaps the page at virtual kernel address 0, so
++ * that we can trap those pesky NULL-reference errors in the kernel.
++ */
++void __init paging_init(void)
++{
++ int i;
++
++#ifdef CONFIG_X86_PAE
++ set_nx();
++ if (nx_enabled)
++ printk(KERN_INFO "NX (Execute Disable) protection: active\n");
++#endif
++ pagetable_init();
++
++ __flush_tlb_all();
++
++ kmap_init();
++
++ /* Switch to the real shared_info page, and clear the
++ * dummy page. */
++ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++ memset(empty_zero_page, 0, sizeof(empty_zero_page));
++
++ /* Setup mapping of lower 1st MB */
++ for (i = 0; i < NR_FIX_ISAMAPS; i++)
++ if (is_initial_xendomain())
++ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
++ else
++ __set_fixmap(FIX_ISAMAP_BEGIN - i,
++ virt_to_machine(empty_zero_page),
++ PAGE_KERNEL_RO);
++}
++
++/*
++ * Test if the WP bit works in supervisor mode. It isn't supported on 386's
++ * and also on some strange 486's. All 586+'s are OK. This used to involve
++ * black magic jumps to work around some nasty CPU bugs, but fortunately the
++ * switch to using exceptions got rid of all that.
++ */
++static void __init test_wp_bit(void)
++{
++ printk(KERN_INFO
++ "Checking if this processor honours the WP bit even in supervisor mode...");
++
++ /* Any page-aligned address will do, the test is non-destructive */
++ __set_fixmap(FIX_WP_TEST, __pa(&swapper_pg_dir), PAGE_READONLY);
++ boot_cpu_data.wp_works_ok = do_test_wp_bit();
++ clear_fixmap(FIX_WP_TEST);
++
++ if (!boot_cpu_data.wp_works_ok) {
++ printk(KERN_CONT "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(KERN_CONT "Ok.\n");
++ }
++}
++
++static struct kcore_list kcore_mem, kcore_vmalloc;
++
++void __init mem_init(void)
++{
++ int codesize, reservedpages, datasize, initsize;
++ int tmp, bad_ppro;
++ unsigned long pfn;
++
++ contiguous_bitmap = alloc_bootmem_low_pages(
++ (max_low_pfn + 2*BITS_PER_LONG) >> 3);
++ BUG_ON(!contiguous_bitmap);
++ memset(contiguous_bitmap, 0, (max_low_pfn + 2*BITS_PER_LONG) >> 3);
++
++ pci_iommu_alloc();
++
++#ifdef CONFIG_FLATMEM
++ BUG_ON(!mem_map);
++#endif
++ bad_ppro = ppro_with_ram_bug();
++
++#ifdef CONFIG_HIGHMEM
++ /* check that fixmap and pkmap do not overlap */
++ if (PKMAP_BASE + LAST_PKMAP*PAGE_SIZE >= FIXADDR_START) {
++ printk(KERN_ERR
++ "fixmap and kmap areas overlap - this will crash\n");
++ printk(KERN_ERR "pkstart: %lxh pkend: %lxh fixstart %lxh\n",
++ PKMAP_BASE, PKMAP_BASE + LAST_PKMAP*PAGE_SIZE,
++ FIXADDR_START);
++ BUG();
++ }
++#endif
++ /* this will put all low memory onto the freelists */
++ totalram_pages += free_all_bootmem();
++ /* XEN: init and count low-mem pages outside initial allocation. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_low_pfn; pfn++) {
++ ClearPageReserved(pfn_to_page(pfn));
++ init_page_count(pfn_to_page(pfn));
++ totalram_pages++;
++ }
++
++ reservedpages = 0;
++ for (tmp = 0; tmp < max_low_pfn; tmp++)
++ /*
++ * Only count reserved RAM pages:
++ */
++ if (page_is_ram(tmp) && PageReserved(pfn_to_page(tmp)))
++ reservedpages++;
++
++ set_highmem_pages_init(bad_ppro);
++
++ codesize = (unsigned long) &_etext - (unsigned long) &_text;
++ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
++ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
++
++ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
++ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
++ VMALLOC_END-VMALLOC_START);
++
++ printk(KERN_INFO "Memory: %luk/%luk available (%dk kernel code, "
++ "%dk reserved, %dk data, %dk init, %ldk highmem)\n",
++ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
++ num_physpages << (PAGE_SHIFT-10),
++ codesize >> 10,
++ reservedpages << (PAGE_SHIFT-10),
++ datasize >> 10,
++ initsize >> 10,
++ (unsigned long) (totalhigh_pages << (PAGE_SHIFT-10))
++ );
++
++#if 1 /* double-sanity-check paranoia */
++ printk(KERN_INFO "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,
++
++#ifdef CONFIG_HIGHMEM
++ PKMAP_BASE, PKMAP_BASE+LAST_PKMAP*PAGE_SIZE,
++ (LAST_PKMAP*PAGE_SIZE) >> 10,
++#endif
++
++ VMALLOC_START, VMALLOC_END,
++ (VMALLOC_END - VMALLOC_START) >> 20,
++
++ (unsigned long)__va(0), (unsigned long)high_memory,
++ ((unsigned long)high_memory - (unsigned long)__va(0)) >> 20,
++
++ (unsigned long)&__init_begin, (unsigned long)&__init_end,
++ ((unsigned long)&__init_end -
++ (unsigned long)&__init_begin) >> 10,
++
++ (unsigned long)&_etext, (unsigned long)&_edata,
++ ((unsigned long)&_edata - (unsigned long)&_etext) >> 10,
++
++ (unsigned long)&_text, (unsigned long)&_etext,
++ ((unsigned long)&_etext - (unsigned long)&_text) >> 10);
++
++#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 */
++
++ if (boot_cpu_data.wp_works_ok < 0)
++ test_wp_bit();
++
++ cpa_init();
++ save_pg_dir();
++ zap_low_mappings();
++
++ SetPagePinned(virt_to_page(init_mm.pgd));
++}
++
++#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);
++}
++#endif
++
++/*
++ * 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 noinline int do_test_wp_bit(void)
++{
++ char tmp_reg;
++ int flag;
++
++ __asm__ __volatile__(
++ " movb %0, %1 \n"
++ "1: movb %1, %0 \n"
++ " xorl %2, %2 \n"
++ "2: \n"
++ _ASM_EXTABLE(1b,2b)
++ :"=m" (*(char *)fix_to_virt(FIX_WP_TEST)),
++ "=q" (tmp_reg),
++ "=r" (flag)
++ :"2" (1)
++ :"memory");
++
++ return flag;
++}
++
++#ifdef CONFIG_DEBUG_RODATA
++const int rodata_test_data = 0xC3;
++EXPORT_SYMBOL_GPL(rodata_test_data);
++
++void mark_rodata_ro(void)
++{
++ unsigned long start = PFN_ALIGN(_text);
++ unsigned long size = PFN_ALIGN(_etext) - start;
++
++ set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT);
++ printk(KERN_INFO "Write protecting the kernel text: %luk\n",
++ size >> 10);
++
++#ifdef CONFIG_CPA_DEBUG
++ printk(KERN_INFO "Testing CPA: Reverting %lx-%lx\n",
++ start, start+size);
++ set_pages_rw(virt_to_page(start), size>>PAGE_SHIFT);
++
++ printk(KERN_INFO "Testing CPA: write protecting again\n");
++ set_pages_ro(virt_to_page(start), size>>PAGE_SHIFT);
++#endif
++ start += size;
++ size = (unsigned long)__end_rodata - start;
++ set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT);
++ printk(KERN_INFO "Write protecting the kernel read-only data: %luk\n",
++ size >> 10);
++ rodata_test();
++
++#ifdef CONFIG_CPA_DEBUG
++ printk(KERN_INFO "Testing CPA: undo %lx-%lx\n", start, start + size);
++ set_pages_rw(virt_to_page(start), size >> PAGE_SHIFT);
++
++ printk(KERN_INFO "Testing CPA: write protecting again\n");
++ set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT);
++#endif
++}
++#endif
++
++void free_init_pages(char *what, unsigned long begin, unsigned long end)
++{
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ /*
++ * If debugging page accesses then do not free this memory but
++ * mark them not present - any buggy init-section access will
++ * create a kernel page fault:
++ */
++ printk(KERN_INFO "debug: unmapping init memory %08lx..%08lx\n",
++ begin, PAGE_ALIGN(end));
++ set_memory_np(begin, (end - begin) >> PAGE_SHIFT);
++#else
++ unsigned long addr;
++
++ /*
++ * We just marked the kernel text read only above, now that
++ * we are going to free part of that, we need to make that
++ * writeable first.
++ */
++ set_memory_rw(begin, (end - begin) >> PAGE_SHIFT);
++
++ 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: %luk freed\n", what, (end - begin) >> 10);
++#endif
++}
++
++void free_initmem(void)
++{
++ free_init_pages("unused kernel memory",
++ (unsigned long)(&__init_begin),
++ (unsigned long)(&__init_end));
++}
++
++#ifdef CONFIG_BLK_DEV_INITRD
++void free_initrd_mem(unsigned long start, unsigned long end)
++{
++ free_init_pages("initrd memory", start, end);
++}
++#endif
+diff --git a/arch/x86/mm/init_64-xen.c b/arch/x86/mm/init_64-xen.c
+new file mode 100644
+index 0000000..18a01a5
+--- /dev/null
++++ b/arch/x86/mm/init_64-xen.c
+@@ -0,0 +1,1397 @@
++/*
++ * linux/arch/x86_64/mm/init.c
++ *
++ * Copyright (C) 1995 Linus Torvalds
++ * Copyright (C) 2000 Pavel Machek <pavel at suse.cz>
++ * Copyright (C) 2002,2003 Andi Kleen <ak at suse.de>
++ *
++ * Jun Nakajima <jun.nakajima at intel.com>
++ * Modified for Xen.
++ */
++
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/types.h>
++#include <linux/ptrace.h>
++#include <linux/mman.h>
++#include <linux/mm.h>
++#include <linux/swap.h>
++#include <linux/smp.h>
++#include <linux/init.h>
++#include <linux/pagemap.h>
++#include <linux/bootmem.h>
++#include <linux/proc_fs.h>
++#include <linux/pci.h>
++#include <linux/pfn.h>
++#include <linux/poison.h>
++#include <linux/dma-mapping.h>
++#include <linux/module.h>
++#include <linux/memory_hotplug.h>
++#include <linux/nmi.h>
++
++#include <asm/processor.h>
++#include <asm/system.h>
++#include <asm/uaccess.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
++#include <asm/dma.h>
++#include <asm/fixmap.h>
++#include <asm/e820.h>
++#include <asm/apic.h>
++#include <asm/tlb.h>
++#include <asm/mmu_context.h>
++#include <asm/proto.h>
++#include <asm/smp.h>
++#include <asm/sections.h>
++#include <asm/kdebug.h>
++#include <asm/numa.h>
++#include <asm/cacheflush.h>
++
++#include <xen/features.h>
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++unsigned int __kernel_page_user;
++EXPORT_SYMBOL(__kernel_page_user);
++#endif
++
++int after_bootmem;
++
++extern unsigned long *contiguous_bitmap;
++
++DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
++extern unsigned long start_pfn;
++
++extern pmd_t level2_fixmap_pgt[PTRS_PER_PMD];
++extern pte_t level1_fixmap_pgt[PTRS_PER_PTE];
++
++int direct_gbpages __meminitdata
++#ifdef CONFIG_DIRECT_GBPAGES
++ = 1
++#endif
++;
++
++#ifndef CONFIG_XEN
++static int __init parse_direct_gbpages_off(char *arg)
++{
++ direct_gbpages = 0;
++ return 0;
++}
++early_param("nogbpages", parse_direct_gbpages_off);
++
++static int __init parse_direct_gbpages_on(char *arg)
++{
++ direct_gbpages = 1;
++ return 0;
++}
++early_param("gbpages", parse_direct_gbpages_on);
++#endif
++
++/*
++ * Use this until direct mapping is established, i.e. before __va() is
++ * available in init_memory_mapping().
++ */
++
++#define addr_to_page(addr, page) \
++ (addr) &= PHYSICAL_PAGE_MASK; \
++ (page) = ((unsigned long *) ((unsigned long) \
++ (((mfn_to_pfn((addr) >> PAGE_SHIFT)) << PAGE_SHIFT) + \
++ __START_KERNEL_map)))
++
++pmd_t *__init early_get_pmd(unsigned long va)
++{
++ unsigned long addr;
++ unsigned long *page = (unsigned long *)init_level4_pgt;
++
++ addr = page[pgd_index(va)];
++ addr_to_page(addr, page);
++
++ addr = page[pud_index(va)];
++ addr_to_page(addr, page);
++
++ return (pmd_t *)&page[pmd_index(va)];
++}
++
++void __meminit early_make_page_readonly(void *va, unsigned int feature)
++{
++ unsigned long addr, _va = (unsigned long)va;
++ pte_t pte, *ptep;
++ unsigned long *page = (unsigned long *) init_level4_pgt;
++
++ BUG_ON(after_bootmem);
++
++ if (xen_feature(feature))
++ return;
++
++ addr = (unsigned long) page[pgd_index(_va)];
++ addr_to_page(addr, page);
++
++ addr = page[pud_index(_va)];
++ addr_to_page(addr, page);
++
++ addr = page[pmd_index(_va)];
++ addr_to_page(addr, page);
++
++ ptep = (pte_t *) &page[pte_index(_va)];
++
++ pte.pte = ptep->pte & ~_PAGE_RW;
++ if (HYPERVISOR_update_va_mapping(_va, pte, 0))
++ BUG();
++}
++
++/*
++ * NOTE: pagetable_init alloc all the fixmap pagetables contiguous on the
++ * physical space so we can cache the place of the first one and move
++ * around without checking the pgd every time.
++ */
++
++void show_mem(void)
++{
++ long i, total = 0, reserved = 0;
++ long shared = 0, cached = 0;
++ struct page *page;
++ pg_data_t *pgdat;
++
++ printk(KERN_INFO "Mem-info:\n");
++ show_free_areas();
++ for_each_online_pgdat(pgdat) {
++ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
++ /*
++ * This loop can take a while with 256 GB and
++ * 4k pages so defer the NMI watchdog:
++ */
++ if (unlikely(i % MAX_ORDER_NR_PAGES == 0))
++ touch_nmi_watchdog();
++
++ if (!pfn_valid(pgdat->node_start_pfn + i))
++ continue;
++
++ page = pfn_to_page(pgdat->node_start_pfn + i);
++ total++;
++ if (PageReserved(page))
++ reserved++;
++ else if (PageSwapCache(page))
++ cached++;
++ else if (page_count(page))
++ shared += page_count(page) - 1;
++ }
++ }
++ printk(KERN_INFO "%lu pages of RAM\n", total);
++ printk(KERN_INFO "%lu reserved pages\n", reserved);
++ printk(KERN_INFO "%lu pages shared\n", shared);
++ printk(KERN_INFO "%lu pages swap cached\n", cached);
++}
++
++static unsigned long __meminitdata table_start;
++static unsigned long __meminitdata table_end;
++
++static __init void *spp_getpage(void)
++{
++ void *ptr;
++
++ if (after_bootmem)
++ ptr = (void *) get_zeroed_page(GFP_ATOMIC);
++ else if (start_pfn < table_end) {
++ ptr = __va(start_pfn << PAGE_SHIFT);
++ start_pfn++;
++ memset(ptr, 0, PAGE_SIZE);
++ } else
++ ptr = alloc_bootmem_pages(PAGE_SIZE);
++
++ if (!ptr || ((unsigned long)ptr & ~PAGE_MASK)) {
++ panic("set_pte_phys: cannot allocate page data %s\n",
++ after_bootmem ? "after bootmem" : "");
++ }
++
++ pr_debug("spp_getpage %p\n", ptr);
++
++ return ptr;
++}
++
++#define pgd_offset_u(address) (__user_pgd(init_level4_pgt) + pgd_index(address))
++#define pud_offset_u(address) (level3_user_pgt + pud_index(address))
++
++static __init void
++set_pte_phys(unsigned long vaddr, unsigned long phys, pgprot_t prot, int user_mode)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++
++ pr_debug("set_pte_phys %lx to %lx\n", vaddr, phys);
++
++ pgd = (user_mode ? pgd_offset_u(vaddr) : pgd_offset_k(vaddr));
++ if (pgd_none(*pgd)) {
++ printk(KERN_ERR
++ "PGD FIXMAP MISSING, it should be setup in head.S!\n");
++ return;
++ }
++ pud = (user_mode ? pud_offset_u(vaddr) : pud_offset(pgd, vaddr));
++ if (pud_none(*pud)) {
++ pmd = (pmd_t *) spp_getpage();
++ make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
++ if (pmd != pmd_offset(pud, 0)) {
++ printk(KERN_ERR "PAGETABLE BUG #01! %p <-> %p\n",
++ pmd, pmd_offset(pud, 0));
++ return;
++ }
++ }
++ pmd = pmd_offset(pud, vaddr);
++ if (pmd_none(*pmd)) {
++ pte = (pte_t *) spp_getpage();
++ make_page_readonly(pte, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
++ if (pte != pte_offset_kernel(pmd, 0)) {
++ printk(KERN_ERR "PAGETABLE BUG #02!\n");
++ return;
++ }
++ }
++ if (pgprot_val(prot))
++ new_pte = pfn_pte(phys >> PAGE_SHIFT, prot);
++ else
++ new_pte = __pte(0);
++
++ pte = pte_offset_kernel(pmd, vaddr);
++ if (!pte_none(*pte) && __pte_val(new_pte) &&
++ __pte_val(*pte) != (__pte_val(new_pte) & __supported_pte_mask))
++ pte_ERROR(*pte);
++ set_pte(pte, new_pte);
++
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
++}
++
++static __init void
++set_pte_phys_ma(unsigned long vaddr, unsigned long phys, pgprot_t prot)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++
++ pr_debug("set_pte_phys_ma %lx to %lx\n", vaddr, phys);
++
++ pgd = pgd_offset_k(vaddr);
++ if (pgd_none(*pgd)) {
++ printk(KERN_ERR
++ "PGD FIXMAP MISSING, it should be setup in head.S!\n");
++ return;
++ }
++ pud = pud_offset(pgd, vaddr);
++ if (pud_none(*pud)) {
++ pmd = (pmd_t *) spp_getpage();
++ make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE | _PAGE_USER));
++ if (pmd != pmd_offset(pud, 0)) {
++ printk(KERN_ERR "PAGETABLE BUG #01! %p <-> %p\n",
++ pmd, pmd_offset(pud, 0));
++ }
++ }
++ pmd = pmd_offset(pud, vaddr);
++ if (pmd_none(*pmd)) {
++ pte = (pte_t *) spp_getpage();
++ make_page_readonly(pte, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE | _PAGE_USER));
++ if (pte != pte_offset_kernel(pmd, 0)) {
++ printk(KERN_ERR "PAGETABLE BUG #02!\n");
++ return;
++ }
++ }
++ new_pte = pfn_pte_ma(phys >> PAGE_SHIFT, prot);
++
++ pte = pte_offset_kernel(pmd, vaddr);
++ if (!pte_none(*pte) && __pte_val(new_pte) &&
++#ifdef CONFIG_ACPI
++ /* __acpi_map_table() fails to properly call clear_fixmap() */
++ (vaddr < __fix_to_virt(FIX_ACPI_END) ||
++ vaddr > __fix_to_virt(FIX_ACPI_BEGIN)) &&
++#endif
++ __pte_val(*pte) != (__pte_val(new_pte) & __supported_pte_mask))
++ pte_ERROR(*pte);
++ set_pte(pte, new_pte);
++
++ /*
++ * It's enough to flush this one mapping.
++ * (PGE mappings get flushed as well)
++ */
++ __flush_tlb_one(vaddr);
++}
++
++#ifndef CONFIG_XEN
++/*
++ * The head.S code sets up the kernel high mapping:
++ *
++ * from __START_KERNEL_map to __START_KERNEL_map + size (== _end-_text)
++ *
++ * phys_addr holds the negative offset to the kernel, which is added
++ * to the compile time generated pmds. This results in invalid pmds up
++ * to the point where we hit the physaddr 0 mapping.
++ *
++ * We limit the mappings to the region from _text to _end. _end is
++ * rounded up to the 2MB boundary. This catches the invalid pmds as
++ * well, as they are located before _text:
++ */
++void __init cleanup_highmap(void)
++{
++ unsigned long vaddr = __START_KERNEL_map;
++ unsigned long end = round_up((unsigned long)_end, PMD_SIZE) - 1;
++ pmd_t *pmd = level2_kernel_pgt;
++ pmd_t *last_pmd = pmd + PTRS_PER_PMD;
++
++ for (; pmd < last_pmd; pmd++, vaddr += PMD_SIZE) {
++ if (pmd_none(*pmd))
++ continue;
++ if (vaddr < (unsigned long) _text || vaddr > end)
++ set_pmd(pmd, __pmd(0));
++ }
++}
++#endif
++
++/* NOTE: this is meant to be run only at boot */
++void __init __set_fixmap(enum fixed_addresses idx, unsigned long phys, pgprot_t prot)
++{
++ unsigned long address = __fix_to_virt(idx);
++
++ if (idx >= __end_of_fixed_addresses) {
++ printk(KERN_ERR "Invalid __set_fixmap\n");
++ return;
++ }
++ switch (idx) {
++ case VSYSCALL_LAST_PAGE ... VSYSCALL_FIRST_PAGE:
++ set_pte_phys(address, phys, prot, 0);
++ set_pte_phys(address, phys, prot, 1);
++ break;
++ case FIX_EARLYCON_MEM_BASE:
++ xen_l1_entry_update(level1_fixmap_pgt + pte_index(address),
++ pfn_pte_ma(phys >> PAGE_SHIFT, prot));
++ break;
++ default:
++ set_pte_phys_ma(address, phys, prot);
++ break;
++ }
++}
++
++static __meminit void *alloc_static_page(unsigned long *phys)
++{
++ unsigned long va = (start_pfn << PAGE_SHIFT) + __START_KERNEL_map;
++
++ if (after_bootmem) {
++ void *adr = (void *)get_zeroed_page(GFP_ATOMIC);
++
++ *phys = __pa(adr);
++
++ return adr;
++ }
++
++ *phys = start_pfn << PAGE_SHIFT;
++ start_pfn++;
++ memset((void *)va, 0, PAGE_SIZE);
++ return (void *)va;
++}
++
++#define PTE_SIZE PAGE_SIZE
++
++static inline int __meminit make_readonly(unsigned long paddr)
++{
++ extern char __vsyscall_0;
++ int readonly = 0;
++
++ /* Make new page tables read-only. */
++ if (!xen_feature(XENFEAT_writable_page_tables)
++ && (paddr >= (table_start << PAGE_SHIFT))
++ && (paddr < (table_end << PAGE_SHIFT)))
++ readonly = 1;
++ /* Make old page tables read-only. */
++ if (!xen_feature(XENFEAT_writable_page_tables)
++ && (paddr >= (xen_start_info->pt_base - __START_KERNEL_map))
++ && (paddr < (start_pfn << PAGE_SHIFT)))
++ readonly = 1;
++
++ /*
++ * No need for writable mapping of kernel image. This also ensures that
++ * page and descriptor tables embedded inside don't have writable
++ * mappings. Exclude the vsyscall area here, allowing alternative
++ * instruction patching to work.
++ */
++ if ((paddr >= __pa_symbol(&_text)) && (paddr < __pa_symbol(&_end))
++ && !(paddr >= __pa_symbol(&__vsyscall_0)
++ && paddr < __pa_symbol(&__vsyscall_0) + PAGE_SIZE))
++ readonly = 1;
++
++ return readonly;
++}
++
++#ifndef CONFIG_XEN
++/* Must run before zap_low_mappings */
++__meminit void *early_ioremap(unsigned long addr, unsigned long size)
++{
++ pmd_t *pmd, *last_pmd;
++ unsigned long vaddr;
++ int i, pmds;
++
++ pmds = ((addr & ~PMD_MASK) + size + ~PMD_MASK) / PMD_SIZE;
++ vaddr = __START_KERNEL_map;
++ pmd = level2_kernel_pgt;
++ last_pmd = level2_kernel_pgt + PTRS_PER_PMD - 1;
++
++ for (; pmd <= last_pmd; pmd++, vaddr += PMD_SIZE) {
++ for (i = 0; i < pmds; i++) {
++ if (pmd_present(pmd[i]))
++ goto continue_outer_loop;
++ }
++ vaddr += addr & ~PMD_MASK;
++ addr &= PMD_MASK;
++
++ for (i = 0; i < pmds; i++, addr += PMD_SIZE)
++ set_pmd(pmd+i, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC));
++ __flush_tlb_all();
++
++ return (void *)vaddr;
++continue_outer_loop:
++ ;
++ }
++ printk("early_ioremap(0x%lx, %lu) failed\n", addr, size);
++ return NULL;
++}
++
++/*
++ * To avoid virtual aliases later:
++ */
++__meminit void early_iounmap(void *addr, unsigned long size)
++{
++ unsigned long vaddr;
++ pmd_t *pmd;
++ int i, pmds;
++
++ vaddr = (unsigned long)addr;
++ pmds = ((vaddr & ~PMD_MASK) + size + ~PMD_MASK) / PMD_SIZE;
++ pmd = level2_kernel_pgt + pmd_index(vaddr);
++
++ for (i = 0; i < pmds; i++)
++ pmd_clear(pmd + i);
++
++ __flush_tlb_all();
++}
++#endif
++
++static unsigned long __meminit
++phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end)
++{
++ int i = pmd_index(address);
++
++ for (; i < PTRS_PER_PMD; i++) {
++ unsigned long pte_phys;
++ pmd_t *pmd = pmd_page + i;
++ pte_t *pte, *pte_save;
++ int k;
++
++ if (address >= end)
++ break;
++
++ if (__pmd_val(*pmd)) {
++ address += PMD_SIZE;
++ continue;
++ }
++
++ pte = alloc_static_page(&pte_phys);
++ pte_save = pte;
++ for (k = 0; k < PTRS_PER_PTE; pte++, k++, address += PTE_SIZE) {
++ unsigned long pteval = address | _PAGE_NX | _KERNPG_TABLE;
++
++ if (address >= (after_bootmem
++ ? end
++ : xen_start_info->nr_pages << PAGE_SHIFT))
++ pteval = 0;
++ else if (make_readonly(address))
++ pteval &= ~_PAGE_RW;
++ set_pte(pte, __pte(pteval & __supported_pte_mask));
++ }
++ if (!after_bootmem) {
++ early_make_page_readonly(pte_save, XENFEAT_writable_page_tables);
++ *pmd = __pmd(pte_phys | _KERNPG_TABLE);
++ } else {
++ make_page_readonly(pte_save, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(pte_phys | _KERNPG_TABLE));
++ }
++ }
++ return address;
++}
++
++static unsigned long __meminit
++phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end)
++{
++ pmd_t *pmd = pmd_offset(pud, 0);
++ unsigned long last_map_addr;
++
++ spin_lock(&init_mm.page_table_lock);
++ last_map_addr = phys_pmd_init(pmd, address, end);
++ spin_unlock(&init_mm.page_table_lock);
++ __flush_tlb_all();
++ return last_map_addr;
++}
++
++static unsigned long __meminit
++phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end)
++{
++ unsigned long last_map_addr = end;
++ int i = pud_index(addr);
++
++ 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;
++
++ if (addr >= end)
++ break;
++
++ if (__pud_val(*pud)) {
++ if (!pud_large(*pud))
++ last_map_addr = phys_pmd_update(pud, addr, end);
++ continue;
++ }
++
++ if (direct_gbpages) {
++ set_pte((pte_t *)pud,
++ pfn_pte(addr >> PAGE_SHIFT, PAGE_KERNEL_LARGE));
++ last_map_addr = (addr & PUD_MASK) + PUD_SIZE;
++ continue;
++ }
++
++ pmd = alloc_static_page(&pmd_phys);
++
++ spin_lock(&init_mm.page_table_lock);
++ *pud = __pud(pmd_phys | _KERNPG_TABLE);
++ last_map_addr = phys_pmd_init(pmd, addr, end);
++ spin_unlock(&init_mm.page_table_lock);
++
++ early_make_page_readonly(pmd, XENFEAT_writable_page_tables);
++ }
++ __flush_tlb_all();
++
++ return last_map_addr >> PAGE_SHIFT;
++}
++
++void __init xen_init_pt(void)
++{
++ unsigned long addr, *page;
++
++ /* Find the initial pte page that was built for us. */
++ page = (unsigned long *)xen_start_info->pt_base;
++ addr = page[pgd_index(__START_KERNEL_map)];
++ addr_to_page(addr, page);
++ addr = page[pud_index(__START_KERNEL_map)];
++ addr_to_page(addr, page);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ /* On Xen 3.0.2 and older we may need to explicitly specify _PAGE_USER
++ in kernel PTEs. We check that here. */
++ if (HYPERVISOR_xen_version(XENVER_version, NULL) <= 0x30000) {
++ unsigned long *pg;
++ pte_t pte;
++
++ /* Mess with the initial mapping of page 0. It's not needed. */
++ BUILD_BUG_ON(__START_KERNEL <= __START_KERNEL_map);
++ addr = page[pmd_index(__START_KERNEL_map)];
++ addr_to_page(addr, pg);
++ pte.pte = pg[pte_index(__START_KERNEL_map)];
++ BUG_ON(!(pte.pte & _PAGE_PRESENT));
++
++ /* If _PAGE_USER isn't set, we obviously do not need it. */
++ if (pte.pte & _PAGE_USER) {
++ /* _PAGE_USER is needed, but is it set implicitly? */
++ pte.pte &= ~_PAGE_USER;
++ if ((HYPERVISOR_update_va_mapping(__START_KERNEL_map,
++ pte, 0) != 0) ||
++ !(pg[pte_index(__START_KERNEL_map)] & _PAGE_USER))
++ /* We need to explicitly specify _PAGE_USER. */
++ __kernel_page_user = _PAGE_USER;
++ }
++ }
++#endif
++
++ /* Construct mapping of initial pte page in our own directories. */
++ init_level4_pgt[pgd_index(__START_KERNEL_map)] =
++ __pgd(__pa_symbol(level3_kernel_pgt) | _PAGE_TABLE);
++ level3_kernel_pgt[pud_index(__START_KERNEL_map)] =
++ __pud(__pa_symbol(level2_kernel_pgt) | _PAGE_TABLE);
++ memcpy(level2_kernel_pgt, page, PAGE_SIZE);
++
++ __user_pgd(init_level4_pgt)[pgd_index(VSYSCALL_START)] =
++ __pgd(__pa_symbol(level3_user_pgt) | _PAGE_TABLE);
++
++ /* Do an early initialization of the fixmap area. */
++ addr = __fix_to_virt(FIX_EARLYCON_MEM_BASE);
++ level3_kernel_pgt[pud_index(addr)] =
++ __pud(__pa_symbol(level2_fixmap_pgt) | _PAGE_TABLE);
++ level2_fixmap_pgt[pmd_index(addr)] =
++ __pmd(__pa_symbol(level1_fixmap_pgt) | _PAGE_TABLE);
++
++ early_make_page_readonly(init_level4_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(__user_pgd(init_level4_pgt),
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level3_kernel_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level3_user_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level2_kernel_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level2_fixmap_pgt,
++ XENFEAT_writable_page_tables);
++ early_make_page_readonly(level1_fixmap_pgt,
++ XENFEAT_writable_page_tables);
++
++ if (!xen_feature(XENFEAT_writable_page_tables)) {
++ xen_pgd_pin(__pa_symbol(init_level4_pgt));
++ xen_pgd_pin(__pa_symbol(__user_pgd(init_level4_pgt)));
++ }
++}
++
++static void __init extend_init_mapping(unsigned long tables_space)
++{
++ unsigned long va = __START_KERNEL_map;
++ unsigned long start = start_pfn;
++ unsigned long phys, addr, *pte_page;
++ pmd_t *pmd;
++ pte_t *pte, new_pte;
++ unsigned long *page = (unsigned long *)init_level4_pgt;
++
++ addr = page[pgd_index(va)];
++ addr_to_page(addr, page);
++ addr = page[pud_index(va)];
++ addr_to_page(addr, page);
++
++ /* Kill mapping of low 1MB. */
++ while (va < (unsigned long)&_text) {
++ if (HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0))
++ BUG();
++ va += PAGE_SIZE;
++ }
++
++ /* Ensure init mappings cover kernel text/data and initial tables. */
++ while (va < (__START_KERNEL_map
++ + (start_pfn << PAGE_SHIFT)
++ + tables_space)) {
++ pmd = (pmd_t *)&page[pmd_index(va)];
++ if (pmd_none(*pmd)) {
++ pte_page = alloc_static_page(&phys);
++ early_make_page_readonly(
++ pte_page, XENFEAT_writable_page_tables);
++ set_pmd(pmd, __pmd(phys | _KERNPG_TABLE));
++ } else {
++ addr = page[pmd_index(va)];
++ addr_to_page(addr, pte_page);
++ }
++ pte = (pte_t *)&pte_page[pte_index(va)];
++ if (pte_none(*pte)) {
++ new_pte = pfn_pte(
++ (va - __START_KERNEL_map) >> PAGE_SHIFT,
++ __pgprot(_KERNPG_TABLE));
++ xen_l1_entry_update(pte, new_pte);
++ }
++ va += PAGE_SIZE;
++ }
++
++ /* Finally, blow away any spurious initial mappings. */
++ while (1) {
++ pmd = (pmd_t *)&page[pmd_index(va)];
++ if (pmd_none(*pmd))
++ break;
++ if (HYPERVISOR_update_va_mapping(va, __pte_ma(0), 0))
++ BUG();
++ va += PAGE_SIZE;
++ }
++
++ if (start_pfn > start)
++ reserve_early(start << PAGE_SHIFT,
++ start_pfn << PAGE_SHIFT, "INITMAP");
++}
++
++static void __init find_early_table_space(unsigned long end)
++{
++ unsigned long puds, pmds, ptes, tables;
++
++ puds = (end + PUD_SIZE - 1) >> PUD_SHIFT;
++ pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
++ ptes = (end + PTE_SIZE - 1) >> PAGE_SHIFT;
++
++ tables = round_up(puds * 8, PAGE_SIZE) +
++ round_up(pmds * 8, PAGE_SIZE) +
++ round_up(ptes * 8, PAGE_SIZE);
++
++ extend_init_mapping(tables);
++
++ table_start = start_pfn;
++ table_end = table_start + (tables>>PAGE_SHIFT);
++
++ early_printk("kernel direct mapping tables up to %lx @ %lx-%lx\n",
++ end, table_start << PAGE_SHIFT,
++ (table_start << PAGE_SHIFT) + tables);
++}
++
++static void __init xen_finish_init_mapping(void)
++{
++ unsigned long i, start, end;
++
++ /* Re-vector virtual addresses pointing into the initial
++ mapping to the just-established permanent ones. */
++ xen_start_info = __va(__pa(xen_start_info));
++ xen_start_info->pt_base = (unsigned long)
++ __va(__pa(xen_start_info->pt_base));
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ phys_to_machine_mapping =
++ __va(__pa(xen_start_info->mfn_list));
++ xen_start_info->mfn_list = (unsigned long)
++ phys_to_machine_mapping;
++ }
++ if (xen_start_info->mod_start)
++ xen_start_info->mod_start = (unsigned long)
++ __va(__pa(xen_start_info->mod_start));
++
++ /* Destroy the Xen-created mappings beyond the kernel image as
++ * well as the temporary mappings created above. Prevents
++ * overlap with modules area (if init mapping is very big).
++ */
++ start = PAGE_ALIGN((unsigned long)_end);
++ end = __START_KERNEL_map + (table_end << PAGE_SHIFT);
++ for (; start < end; start += PAGE_SIZE)
++ if (HYPERVISOR_update_va_mapping(start, __pte_ma(0), 0))
++ BUG();
++
++ /* Allocate pte's for initial fixmaps from 'start_pfn' allocator. */
++ table_end = ~0UL;
++
++ /*
++ * Prefetch pte's for the bt_ioremap() area. It gets used before the
++ * boot-time allocator is online, so allocate-on-demand would fail.
++ */
++ early_ioremap_clear();
++ for (i = FIX_BTMAP_END; i <= FIX_BTMAP_BEGIN; i++)
++ __set_fixmap(i, 0, __pgprot(0));
++ early_ioremap_reset();
++
++ /* Switch to the real shared_info page, and clear the dummy page. */
++ set_fixmap(FIX_SHARED_INFO, xen_start_info->shared_info);
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++ memset(empty_zero_page, 0, sizeof(empty_zero_page));
++
++ /* Set up mapping of lowest 1MB of physical memory. */
++ for (i = 0; i < NR_FIX_ISAMAPS; i++)
++ if (is_initial_xendomain())
++ set_fixmap(FIX_ISAMAP_BEGIN - i, i * PAGE_SIZE);
++ else
++ __set_fixmap(FIX_ISAMAP_BEGIN - i,
++ virt_to_mfn(empty_zero_page)
++ << PAGE_SHIFT,
++ PAGE_KERNEL_RO);
++
++ /* Disable the 'start_pfn' allocator. */
++ table_end = start_pfn;
++}
++
++static void __init init_gbpages(void)
++{
++ if (direct_gbpages && cpu_has_gbpages)
++ printk(KERN_INFO "Using GB pages for direct mapping\n");
++ else
++ direct_gbpages = 0;
++}
++
++#ifdef CONFIG_MEMTEST_BOOTPARAM
++
++static void __init memtest(unsigned long start_phys, unsigned long size,
++ unsigned pattern)
++{
++ unsigned long i;
++ unsigned long *start;
++ unsigned long start_bad;
++ unsigned long last_bad;
++ unsigned long val;
++ unsigned long start_phys_aligned;
++ unsigned long count;
++ unsigned long incr;
++
++ switch (pattern) {
++ case 0:
++ val = 0UL;
++ break;
++ case 1:
++ val = -1UL;
++ break;
++ case 2:
++ val = 0x5555555555555555UL;
++ break;
++ case 3:
++ val = 0xaaaaaaaaaaaaaaaaUL;
++ break;
++ default:
++ return;
++ }
++
++ incr = sizeof(unsigned long);
++ start_phys_aligned = ALIGN(start_phys, incr);
++ count = (size - (start_phys_aligned - start_phys))/incr;
++ start = __va(start_phys_aligned);
++ start_bad = 0;
++ last_bad = 0;
++
++ for (i = 0; i < count; i++)
++ start[i] = val;
++ for (i = 0; i < count; i++, start++, start_phys_aligned += incr) {
++ if (*start != val) {
++ if (start_phys_aligned == last_bad + incr) {
++ last_bad += incr;
++ } else {
++ if (start_bad) {
++ printk(KERN_CONT "\n %016lx bad mem addr %016lx - %016lx reserved",
++ val, start_bad, last_bad + incr);
++ reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
++ }
++ start_bad = last_bad = start_phys_aligned;
++ }
++ }
++ }
++ if (start_bad) {
++ printk(KERN_CONT "\n %016lx bad mem addr %016lx - %016lx reserved",
++ val, start_bad, last_bad + incr);
++ reserve_early(start_bad, last_bad - start_bad, "BAD RAM");
++ }
++
++}
++
++static int memtest_pattern __initdata = CONFIG_MEMTEST_BOOTPARAM_VALUE;
++
++static int __init parse_memtest(char *arg)
++{
++ if (arg)
++ memtest_pattern = simple_strtoul(arg, NULL, 0);
++ return 0;
++}
++
++early_param("memtest", parse_memtest);
++
++static void __init early_memtest(unsigned long start, unsigned long end)
++{
++ u64 t_start, t_size;
++ unsigned pattern;
++
++ if (!memtest_pattern)
++ return;
++
++ printk(KERN_INFO "early_memtest: pattern num %d", memtest_pattern);
++ for (pattern = 0; pattern < memtest_pattern; pattern++) {
++ t_start = start;
++ t_size = 0;
++ while (t_start < end) {
++ t_start = find_e820_area_size(t_start, &t_size, 1);
++
++ /* done ? */
++ if (t_start >= end)
++ break;
++ if (t_start + t_size > end)
++ t_size = end - t_start;
++
++ printk(KERN_CONT "\n %016llx - %016llx pattern %d",
++ (unsigned long long)t_start,
++ (unsigned long long)t_start + t_size, pattern);
++
++ memtest(t_start, t_size, pattern);
++
++ t_start += t_size;
++ }
++ }
++ printk(KERN_CONT "\n");
++}
++#else
++static void __init early_memtest(unsigned long start, unsigned long end)
++{
++}
++#endif
++
++/*
++ * Setup the direct mapping of the physical memory at PAGE_OFFSET.
++ * This runs before bootmem is initialized and gets pages directly from
++ * the physical memory. To access them they are temporarily mapped.
++ */
++unsigned long __init_refok init_memory_mapping(unsigned long start, unsigned long end)
++{
++ unsigned long next, last_map_addr = end;
++ unsigned long start_phys = start, end_phys = end;
++
++ printk(KERN_INFO "init_memory_mapping\n");
++
++ /*
++ * Find space for the kernel direct mapping tables.
++ *
++ * Later we should allocate these tables in the local node of the
++ * memory mapped. Unfortunately this is done currently before the
++ * nodes are discovered.
++ */
++ if (!after_bootmem) {
++ init_gbpages();
++ find_early_table_space(end);
++ }
++
++ start = (unsigned long)__va(start);
++ end = (unsigned long)__va(end);
++
++ for (; start < end; start = next) {
++ pgd_t *pgd = pgd_offset_k(start);
++ unsigned long pud_phys;
++ pud_t *pud;
++
++ if (after_bootmem)
++ pud = pud_offset(pgd, start & PGDIR_MASK);
++ else
++ pud = alloc_static_page(&pud_phys);
++ next = start + PGDIR_SIZE;
++ if (next > end)
++ next = end;
++ last_map_addr = phys_pud_init(pud, __pa(start), __pa(next));
++ if (!after_bootmem) {
++ early_make_page_readonly(pud, XENFEAT_writable_page_tables);
++ set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys));
++ }
++ }
++
++ if (!after_bootmem) {
++ BUG_ON(start_pfn != table_end);
++ xen_finish_init_mapping();
++ }
++
++ __flush_tlb_all();
++
++ if (!after_bootmem)
++ reserve_early(table_start << PAGE_SHIFT,
++ table_end << PAGE_SHIFT, "PGTABLE");
++
++ if (!after_bootmem)
++ early_memtest(start_phys, end_phys);
++
++ return last_map_addr;
++}
++
++#ifndef CONFIG_NUMA
++void __init paging_init(void)
++{
++ unsigned long max_zone_pfns[MAX_NR_ZONES];
++
++ memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
++ max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN;
++ max_zone_pfns[ZONE_DMA32] = MAX_DMA32_PFN;
++ max_zone_pfns[ZONE_NORMAL] = end_pfn;
++
++ memory_present(0, 0, end_pfn);
++ sparse_init();
++ free_area_init_nodes(max_zone_pfns);
++
++ SetPagePinned(virt_to_page(init_mm.pgd));
++}
++#endif
++
++/*
++ * Memory hotplug specific functions
++ */
++#ifdef CONFIG_MEMORY_HOTPLUG
++/*
++ * 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 + ZONE_NORMAL;
++ unsigned long last_mapped_pfn, start_pfn = start >> PAGE_SHIFT;
++ unsigned long nr_pages = size >> PAGE_SHIFT;
++ int ret;
++
++ last_mapped_pfn = init_memory_mapping(start, start + size-1);
++ if (last_mapped_pfn > max_pfn_mapped)
++ max_pfn_mapped = last_mapped_pfn;
++
++ ret = __add_pages(zone, start_pfn, nr_pages);
++ WARN_ON(1);
++
++ return ret;
++}
++EXPORT_SYMBOL_GPL(arch_add_memory);
++
++#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 */
++
++/*
++ * devmem_is_allowed() checks to see if /dev/mem access to a certain address
++ * is valid. The argument is a physical page number.
++ *
++ *
++ * On x86, access has to be given to the first megabyte of ram because that area
++ * contains bios code and data regions used by X and dosemu and similar apps.
++ * Access has to be given to non-kernel-ram areas as well, these contain the PCI
++ * mmio resources as well as potential bios/acpi data regions.
++ */
++int devmem_is_allowed(unsigned long pagenr)
++{
++ if (pagenr <= 256)
++ return 1;
++ if (mfn_to_local_pfn(pagenr) >= max_pfn)
++ return 1;
++ return 0;
++}
++
++
++static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel,
++ kcore_modules, kcore_vsyscall;
++
++void __init mem_init(void)
++{
++ long codesize, reservedpages, datasize, initsize;
++ unsigned long pfn;
++
++ contiguous_bitmap = alloc_bootmem_low_pages(
++ (end_pfn + 2*BITS_PER_LONG) >> 3);
++ BUG_ON(!contiguous_bitmap);
++ memset(contiguous_bitmap, 0, (end_pfn + 2*BITS_PER_LONG) >> 3);
++
++ pci_iommu_alloc();
++
++ /* clear_bss() already clear the empty_zero_page */
++
++ reservedpages = 0;
++
++ /* this will put all low memory onto the freelists */
++#ifdef CONFIG_NUMA
++ totalram_pages = numa_free_all_bootmem();
++#else
++ totalram_pages = free_all_bootmem();
++#endif
++ /* XEN: init and count pages outside initial allocation. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
++ ClearPageReserved(pfn_to_page(pfn));
++ init_page_count(pfn_to_page(pfn));
++ totalram_pages++;
++ }
++ reservedpages = end_pfn - totalram_pages -
++ absent_pages_in_range(0, end_pfn);
++ after_bootmem = 1;
++
++ codesize = (unsigned long) &_etext - (unsigned long) &_text;
++ datasize = (unsigned long) &_edata - (unsigned long) &_etext;
++ initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
++
++ /* Register memory areas for /proc/kcore */
++ kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
++ kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
++ VMALLOC_END-VMALLOC_START);
++ kclist_add(&kcore_kernel, &_stext, _end - _stext);
++ kclist_add(&kcore_modules, (void *)MODULES_VADDR, MODULES_LEN);
++ kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START,
++ VSYSCALL_END - VSYSCALL_START);
++
++ printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, "
++ "%ldk reserved, %ldk data, %ldk init)\n",
++ (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
++ end_pfn << (PAGE_SHIFT-10),
++ codesize >> 10,
++ reservedpages << (PAGE_SHIFT-10),
++ datasize >> 10,
++ initsize >> 10);
++
++ cpa_init();
++}
++
++void free_init_pages(char *what, unsigned long begin, unsigned long end)
++{
++ unsigned long addr = begin;
++
++ if (addr >= end)
++ return;
++
++ /*
++ * If debugging page accesses then do not free this memory but
++ * mark them not present - any buggy init-section access will
++ * create a kernel page fault:
++ */
++#ifdef CONFIG_DEBUG_PAGEALLOC
++ printk(KERN_INFO "debug: unmapping init memory %08lx..%08lx\n",
++ begin, PAGE_ALIGN(end));
++ set_memory_np(begin, (end - begin) >> PAGE_SHIFT);
++#else
++ printk(KERN_INFO "Freeing %s: %luk freed\n", what, (end - begin) >> 10);
++
++ for (; addr < end; addr += PAGE_SIZE) {
++ ClearPageReserved(virt_to_page(addr));
++ init_page_count(virt_to_page(addr));
++ memset((void *)(addr & ~(PAGE_SIZE-1)),
++ POISON_FREE_INITMEM, PAGE_SIZE);
++ if (addr >= __START_KERNEL_map) {
++ /* make_readonly() reports all kernel addresses. */
++ if (HYPERVISOR_update_va_mapping((unsigned long)__va(__pa(addr)),
++ pfn_pte(__pa(addr) >> PAGE_SHIFT,
++ PAGE_KERNEL),
++ 0))
++ BUG();
++ if (HYPERVISOR_update_va_mapping(addr, __pte(0), 0))
++ BUG();
++ }
++ free_page(addr);
++ totalram_pages++;
++ }
++#endif
++}
++
++void free_initmem(void)
++{
++ free_init_pages("unused kernel memory",
++ (unsigned long)(&__init_begin),
++ (unsigned long)(&__init_end));
++}
++
++#ifdef CONFIG_DEBUG_RODATA
++const int rodata_test_data = 0xC3;
++EXPORT_SYMBOL_GPL(rodata_test_data);
++
++void mark_rodata_ro(void)
++{
++ unsigned long start = PFN_ALIGN(_stext), end = PFN_ALIGN(__end_rodata);
++
++ printk(KERN_INFO "Write protecting the kernel read-only data: %luk\n",
++ (end - start) >> 10);
++ set_memory_ro(start, (end - start) >> PAGE_SHIFT);
++
++ /*
++ * The rodata section (but not the kernel text!) should also be
++ * not-executable.
++ */
++ start = ((unsigned long)__start_rodata + PAGE_SIZE - 1) & PAGE_MASK;
++ set_memory_nx(start, (end - start) >> PAGE_SHIFT);
++
++ rodata_test();
++
++#ifdef CONFIG_CPA_DEBUG
++ printk(KERN_INFO "Testing CPA: undo %lx-%lx\n", start, end);
++ set_memory_rw(start, (end-start) >> PAGE_SHIFT);
++
++ printk(KERN_INFO "Testing CPA: again\n");
++ set_memory_ro(start, (end-start) >> PAGE_SHIFT);
++#endif
++}
++
++#endif
++
++#ifdef CONFIG_BLK_DEV_INITRD
++void free_initrd_mem(unsigned long start, unsigned long end)
++{
++ free_init_pages("initrd memory", start, end);
++}
++#endif
++
++void __init reserve_bootmem_generic(unsigned long phys, unsigned len)
++{
++#ifdef CONFIG_NUMA
++ int nid, next_nid;
++#endif
++ unsigned long pfn = phys >> PAGE_SHIFT;
++
++ if (pfn >= end_pfn) {
++ /*
++ * This can happen with kdump kernels when accessing
++ * firmware tables:
++ */
++ if (pfn < max_pfn_mapped)
++ 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
++ nid = phys_to_nid(phys);
++ next_nid = phys_to_nid(phys + len - 1);
++ if (nid == next_nid)
++ reserve_bootmem_node(NODE_DATA(nid), phys, len, BOOTMEM_DEFAULT);
++ else
++ reserve_bootmem(phys, len, BOOTMEM_DEFAULT);
++#else
++ reserve_bootmem(phys, len, BOOTMEM_DEFAULT);
++#endif
++
++#ifndef CONFIG_XEN
++ if (phys+len <= MAX_DMA_PFN*PAGE_SIZE) {
++ static unsigned long dma_reserve __initdata;
++
++ dma_reserve += len / PAGE_SIZE;
++ set_dma_reserve(dma_reserve);
++ }
++#endif
++}
++
++int kern_addr_valid(unsigned long addr)
++{
++ unsigned long above = ((long)addr) >> __VIRTUAL_MASK_SHIFT;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++ pte_t *pte;
++
++ if (above != 0 && above != -1UL)
++ return 0;
++
++ pgd = pgd_offset_k(addr);
++ if (pgd_none(*pgd))
++ return 0;
++
++ pud = pud_offset(pgd, addr);
++ if (pud_none(*pud))
++ return 0;
++
++ pmd = pmd_offset(pud, addr);
++ if (pmd_none(*pmd))
++ return 0;
++
++ if (pmd_large(*pmd))
++ return pfn_valid(pmd_pfn(*pmd));
++
++ pte = pte_offset_kernel(pmd, addr);
++ if (pte_none(*pte))
++ return 0;
++
++ return pfn_valid(pte_pfn(*pte));
++}
++
++/*
++ * 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_START + (VSYSCALL_MAPPED_PAGES * PAGE_SIZE),
++ .vm_page_prot = PAGE_READONLY_EXEC,
++ .vm_flags = VM_READ | VM_EXEC
++};
++
++struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
++{
++#ifdef CONFIG_IA32_EMULATION
++ if (test_tsk_thread_flag(tsk, TIF_IA32))
++ return NULL;
++#endif
++ return &gate_vma;
++}
++
++int in_gate_area(struct task_struct *task, unsigned long addr)
++{
++ struct vm_area_struct *vma = get_gate_vma(task);
++
++ if (!vma)
++ return 0;
++
++ return (addr >= vma->vm_start) && (addr < vma->vm_end);
++}
++
++/*
++ * Use this when you have no reliable task/vma, typically from interrupt
++ * context. It is less reliable than using the task's vma and may give
++ * false positives:
++ */
++int in_gate_area_no_task(unsigned long addr)
++{
++ return (addr >= VSYSCALL_START) && (addr < VSYSCALL_END);
++}
++
++const char *arch_vma_name(struct vm_area_struct *vma)
++{
++ if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso)
++ return "[vdso]";
++ if (vma == &gate_vma)
++ return "[vsyscall]";
++ return NULL;
++}
++
++#ifdef CONFIG_SPARSEMEM_VMEMMAP
++/*
++ * Initialise the sparsemem vmemmap using huge-pages at the PMD level.
++ */
++static long __meminitdata addr_start, addr_end;
++static void __meminitdata *p_start, *p_end;
++static int __meminitdata node_start;
++
++int __meminit
++vmemmap_populate(struct page *start_page, unsigned long size, int node)
++{
++ unsigned long addr = (unsigned long)start_page;
++ unsigned long end = (unsigned long)(start_page + size);
++ unsigned long next;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++
++ for (; addr < end; addr = next) {
++ next = pmd_addr_end(addr, end);
++
++ pgd = vmemmap_pgd_populate(addr, node);
++ if (!pgd)
++ return -ENOMEM;
++
++ pud = vmemmap_pud_populate(pgd, addr, node);
++ if (!pud)
++ return -ENOMEM;
++
++ pmd = pmd_offset(pud, addr);
++ if (pmd_none(*pmd)) {
++ pte_t entry;
++ void *p;
++
++ p = vmemmap_alloc_block(PMD_SIZE, node);
++ if (!p)
++ return -ENOMEM;
++
++ entry = pfn_pte(__pa(p) >> PAGE_SHIFT,
++ PAGE_KERNEL_LARGE);
++ set_pmd(pmd, __pmd(pte_val(entry)));
++
++ /* check to see if we have contiguous blocks */
++ if (p_end != p || node_start != node) {
++ if (p_start)
++ printk(KERN_DEBUG " [%lx-%lx] PMD -> [%p-%p] on node %d\n",
++ addr_start, addr_end-1, p_start, p_end-1, node_start);
++ addr_start = addr;
++ node_start = node;
++ p_start = p;
++ }
++ addr_end = addr + PMD_SIZE;
++ p_end = p + PMD_SIZE;
++ } else {
++ vmemmap_verify((pte_t *)pmd, node, addr, next);
++ }
++ }
++ return 0;
++}
++
++void __meminit vmemmap_populate_print_last(void)
++{
++ if (p_start) {
++ printk(KERN_DEBUG " [%lx-%lx] PMD -> [%p-%p] on node %d\n",
++ addr_start, addr_end-1, p_start, p_end-1, node_start);
++ p_start = NULL;
++ p_end = NULL;
++ node_start = 0;
++ }
++}
++#endif
+diff --git a/arch/x86/mm/ioremap-xen.c b/arch/x86/mm/ioremap-xen.c
+new file mode 100644
+index 0000000..95c214f
+--- /dev/null
++++ b/arch/x86/mm/ioremap-xen.c
+@@ -0,0 +1,819 @@
++/*
++ * 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
++ */
++
++#include <linux/bootmem.h>
++#include <linux/init.h>
++#include <linux/io.h>
++#include <linux/module.h>
++#include <linux/pfn.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++
++#include <asm/cacheflush.h>
++#include <asm/e820.h>
++#include <asm/fixmap.h>
++#include <asm/pgtable.h>
++#include <asm/tlbflush.h>
++#include <asm/pgalloc.h>
++#include <asm/pat.h>
++
++#ifdef CONFIG_X86_64
++
++#ifndef CONFIG_XEN
++unsigned long __phys_addr(unsigned long x)
++{
++ if (x >= __START_KERNEL_map)
++ return x - __START_KERNEL_map + phys_base;
++ return x - PAGE_OFFSET;
++}
++EXPORT_SYMBOL(__phys_addr);
++#endif
++
++static inline int phys_addr_valid(unsigned long addr)
++{
++ return addr < (1UL << boot_cpu_data.x86_phys_bits);
++}
++
++#else
++
++static inline int phys_addr_valid(unsigned long addr)
++{
++ return 1;
++}
++
++#endif
++
++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;
++
++ BUG_ON(!pte_none(*pte));
++
++ (*v)->ptr = ((u64)pfn_to_mfn(page_to_pfn(pmd_page)) <<
++ PAGE_SHIFT) | ((unsigned long)pte & ~PAGE_MASK);
++ (*v)++;
++
++ return 0;
++}
++
++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)
++{
++ int rc;
++ unsigned long i, start_address;
++ mmu_update_t *u, *v, *w;
++
++ u = v = w = (mmu_update_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT);
++ if (u == NULL)
++ return -ENOMEM;
++
++ start_address = address;
++
++ flush_cache_all();
++
++ 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;
++ }
++
++ /*
++ * Fill in the machine address: PTE ptr is done later by
++ * apply_to_page_range().
++ */
++ v->val = __pte_val(pfn_pte_ma(mfn, prot)) | _PAGE_IO;
++
++ mfn++;
++ address += PAGE_SIZE;
++ v++;
++ }
++
++ 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;
++ }
++
++ rc = 0;
++
++ out:
++ flush_tlb_all();
++
++ free_page((unsigned long)u);
++
++ return rc;
++}
++
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return remap_pfn_range(vma, address, mfn, size, prot);
++
++ if (domid == DOMID_SELF)
++ return -EINVAL;
++
++ vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP;
++
++ vma->vm_mm->context.has_foreign_mappings = 1;
++
++ return __direct_remap_pfn_range(
++ vma->vm_mm, address, mfn, size, prot, domid);
++}
++EXPORT_SYMBOL(direct_remap_pfn_range);
++
++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);
++
++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;
++}
++
++int create_lookup_pte_addr(struct mm_struct *mm,
++ unsigned long address,
++ uint64_t *ptep)
++{
++ return apply_to_page_range(mm, address, PAGE_SIZE,
++ lookup_pte_fn, ptep);
++}
++
++EXPORT_SYMBOL(create_lookup_pte_addr);
++
++static int noop_fn(
++ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
++{
++ return 0;
++}
++
++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);
++
++#ifdef CONFIG_X86_32
++int __init page_is_ram(unsigned long pagenr)
++{
++ resource_size_t addr, end;
++ int i;
++
++#ifndef CONFIG_XEN
++ /*
++ * A special case is the first 4Kb of memory;
++ * This is a BIOS owned area, not kernel ram, but generally
++ * not listed as such in the E820 table.
++ */
++ if (pagenr == 0)
++ return 0;
++
++ /*
++ * Second special case: Some BIOSen report the PC BIOS
++ * area (640->1Mb) as ram even though it is not.
++ */
++ if (pagenr >= (BIOS_BEGIN >> PAGE_SHIFT) &&
++ pagenr < (BIOS_END >> PAGE_SHIFT))
++ return 0;
++#endif
++
++ for (i = 0; i < e820.nr_map; i++) {
++ /*
++ * Not usable memory:
++ */
++ if (e820.map[i].type != E820_RAM)
++ continue;
++ 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;
++}
++#endif
++
++/*
++ * Fix up the linear direct mapping of the kernel to avoid cache attribute
++ * conflicts.
++ */
++int ioremap_change_attr(unsigned long vaddr, unsigned long size,
++ unsigned long prot_val)
++{
++ unsigned long nrpages = size >> PAGE_SHIFT;
++ int err;
++
++ switch (prot_val) {
++ case _PAGE_CACHE_UC:
++ default:
++ err = _set_memory_uc(vaddr, nrpages);
++ break;
++ case _PAGE_CACHE_WC:
++ err = _set_memory_wc(vaddr, nrpages);
++ break;
++ case _PAGE_CACHE_WB:
++ err = _set_memory_wb(vaddr, nrpages);
++ break;
++ }
++
++ return err;
++}
++
++/*
++ * 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.
++ */
++static void __iomem *__ioremap_caller(resource_size_t phys_addr,
++ unsigned long size, unsigned long prot_val, void *caller)
++{
++ unsigned long mfn, offset, vaddr;
++ resource_size_t last_addr;
++ struct vm_struct *area;
++ unsigned long new_prot_val;
++ pgprot_t prot;
++ int retval;
++ 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;
++
++ if (!phys_addr_valid(phys_addr)) {
++ printk(KERN_WARNING "ioremap: invalid physical address %llx\n",
++ (unsigned long long)phys_addr);
++ WARN_ON_ONCE(1);
++ return NULL;
++ }
++
++ /*
++ * Don't remap the low PCI/ISA area, it's always mapped..
++ */
++ if (is_initial_xendomain() && last_addr < ISA_END_ADDRESS)
++ return (__force void __iomem *)isa_bus_to_virt((unsigned long)phys_addr);
++
++ /*
++ * Don't allow anybody to remap normal RAM that we're using..
++ */
++ for (mfn = PFN_DOWN(phys_addr); mfn < PFN_UP(last_addr); mfn++) {
++ unsigned long pfn = mfn_to_local_pfn(mfn);
++
++ if (pfn_valid(pfn)) {
++ if (!PageReserved(pfn_to_page(pfn)))
++ return NULL;
++ domid = DOMID_SELF;
++ }
++ }
++ WARN_ON_ONCE(domid == DOMID_SELF);
++
++ /*
++ * Mappings have to be page-aligned
++ */
++ offset = phys_addr & ~PAGE_MASK;
++ phys_addr &= PAGE_MASK;
++ size = PAGE_ALIGN(last_addr+1) - phys_addr;
++
++ retval = reserve_memtype(phys_addr, phys_addr + size,
++ prot_val, &new_prot_val);
++ if (retval) {
++ pr_debug("Warning: reserve_memtype returned %d\n", retval);
++ return NULL;
++ }
++
++ if (prot_val != new_prot_val) {
++ /*
++ * Do not fallback to certain memory types with certain
++ * requested type:
++ * - request is uc-, return cannot be write-back
++ * - request is uc-, return cannot be write-combine
++ * - request is write-combine, return cannot be write-back
++ */
++ if ((prot_val == _PAGE_CACHE_UC_MINUS &&
++ (new_prot_val == _PAGE_CACHE_WB ||
++ new_prot_val == _PAGE_CACHE_WC)) ||
++ (prot_val == _PAGE_CACHE_WC &&
++ new_prot_val == _PAGE_CACHE_WB)) {
++ pr_debug(
++ "ioremap error for 0x%llx-0x%llx, requested 0x%lx, got 0x%lx\n",
++ (unsigned long long)phys_addr,
++ (unsigned long long)(phys_addr + size),
++ prot_val, new_prot_val);
++ free_memtype(phys_addr, phys_addr + size);
++ return NULL;
++ }
++ prot_val = new_prot_val;
++ }
++
++ switch (prot_val) {
++ case _PAGE_CACHE_UC:
++ default:
++ prot = PAGE_KERNEL_NOCACHE;
++ break;
++ case _PAGE_CACHE_UC_MINUS:
++ prot = PAGE_KERNEL_UC_MINUS;
++ break;
++ case _PAGE_CACHE_WC:
++ prot = PAGE_KERNEL_WC;
++ break;
++ case _PAGE_CACHE_WB:
++ prot = PAGE_KERNEL;
++ break;
++ }
++
++ /*
++ * Ok, go for it..
++ */
++ area = get_vm_area_caller(size, VM_IOREMAP, caller);
++ if (!area)
++ return NULL;
++ area->phys_addr = phys_addr;
++ vaddr = (unsigned long) area->addr;
++ if (__direct_remap_pfn_range(&init_mm, vaddr, PFN_DOWN(phys_addr),
++ size, prot, domid)) {
++ free_memtype(phys_addr, phys_addr + size);
++ free_vm_area(area);
++ return NULL;
++ }
++
++ if (ioremap_change_attr(vaddr, size, prot_val) < 0) {
++ free_memtype(phys_addr, phys_addr + size);
++ vunmap(area->addr);
++ return NULL;
++ }
++
++ return (void __iomem *) (vaddr + offset);
++}
++
++/**
++ * 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.
++ */
++void __iomem *ioremap_nocache(resource_size_t phys_addr, unsigned long size)
++{
++ /*
++ * Ideally, this should be:
++ * pat_wc_enabled ? _PAGE_CACHE_UC : _PAGE_CACHE_UC_MINUS;
++ *
++ * Till we fix all X drivers to use ioremap_wc(), we will use
++ * UC MINUS.
++ */
++ unsigned long val = _PAGE_CACHE_UC_MINUS;
++
++ return __ioremap_caller(phys_addr, size, val,
++ __builtin_return_address(0));
++}
++EXPORT_SYMBOL(ioremap_nocache);
++
++/**
++ * ioremap_wc - map memory into CPU space write combined
++ * @offset: bus address of the memory
++ * @size: size of the resource to map
++ *
++ * This version of ioremap ensures that the memory is marked write combining.
++ * Write combining allows faster writes to some hardware devices.
++ *
++ * Must be freed with iounmap.
++ */
++void __iomem *ioremap_wc(unsigned long phys_addr, unsigned long size)
++{
++ if (pat_wc_enabled)
++ return __ioremap_caller(phys_addr, size, _PAGE_CACHE_WC,
++ __builtin_return_address(0));
++ else
++ return ioremap_nocache(phys_addr, size);
++}
++EXPORT_SYMBOL(ioremap_wc);
++
++void __iomem *ioremap_cache(resource_size_t phys_addr, unsigned long size)
++{
++ return __ioremap_caller(phys_addr, size, _PAGE_CACHE_WB,
++ __builtin_return_address(0));
++}
++EXPORT_SYMBOL(ioremap_cache);
++
++#ifndef CONFIG_XEN
++static void __iomem *ioremap_default(resource_size_t phys_addr,
++ unsigned long size)
++{
++ unsigned long flags;
++ void *ret;
++ int err;
++
++ /*
++ * - WB for WB-able memory and no other conflicting mappings
++ * - UC_MINUS for non-WB-able memory with no other conflicting mappings
++ * - Inherit from confliting mappings otherwise
++ */
++ err = reserve_memtype(phys_addr, phys_addr + size, -1, &flags);
++ if (err < 0)
++ return NULL;
++
++ ret = (void *) __ioremap_caller(phys_addr, size, flags,
++ __builtin_return_address(0));
++
++ free_memtype(phys_addr, phys_addr + size);
++ return (void __iomem *)ret;
++}
++#endif
++
++/**
++ * 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;
++
++ if ((void __force *)addr <= high_memory)
++ return;
++
++ /*
++ * __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;
++
++ addr = (volatile void __iomem *)
++ (PAGE_MASK & (unsigned long __force)addr);
++
++ /* 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);
++
++ if (!p) {
++ printk(KERN_ERR "iounmap: bad address %p\n", addr);
++ dump_stack();
++ return;
++ }
++
++ free_memtype(p->phys_addr, p->phys_addr + get_vm_area_size(p));
++
++ /* Finally remove it */
++ o = remove_vm_area((void *)addr);
++ BUG_ON(p != o || o == NULL);
++ kfree(p);
++}
++EXPORT_SYMBOL(iounmap);
++
++#ifndef CONFIG_XEN
++/*
++ * Convert a physical pointer to a virtual kernel pointer for /dev/mem
++ * access
++ */
++void *xlate_dev_mem_ptr(unsigned long phys)
++{
++ void *addr;
++ unsigned long start = phys & PAGE_MASK;
++
++ /* If page is RAM, we can use __va. Otherwise ioremap and unmap. */
++ if (page_is_ram(start >> PAGE_SHIFT))
++ return __va(phys);
++
++ addr = (void *)ioremap_default(start, PAGE_SIZE);
++ if (addr)
++ addr = (void *)((unsigned long)addr | (phys & ~PAGE_MASK));
++
++ return addr;
++}
++
++void unxlate_dev_mem_ptr(unsigned long phys, void *addr)
++{
++ if (page_is_ram(phys >> PAGE_SHIFT))
++ return;
++
++ iounmap((void __iomem *)((unsigned long)addr & PAGE_MASK));
++ return;
++}
++#endif
++
++int __initdata early_ioremap_debug;
++
++static int __init early_ioremap_debug_setup(char *str)
++{
++ early_ioremap_debug = 1;
++
++ return 0;
++}
++early_param("early_ioremap_debug", early_ioremap_debug_setup);
++
++static __initdata int after_paging_init;
++static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)]
++ __section(.bss.page_aligned);
++
++#ifdef CONFIG_X86_32
++static inline pmd_t * __init early_ioremap_pmd(unsigned long addr)
++{
++ /* Don't assume we're using swapper_pg_dir at this point */
++ pgd_t *base = __va(read_cr3());
++ pgd_t *pgd = &base[pgd_index(addr)];
++ pud_t *pud = pud_offset(pgd, addr);
++ pmd_t *pmd = pmd_offset(pud, addr);
++
++ return pmd;
++}
++#else
++#define early_ioremap_pmd early_get_pmd
++#undef make_lowmem_page_readonly
++#define make_lowmem_page_readonly early_make_page_readonly
++#endif
++
++static inline pte_t * __init early_ioremap_pte(unsigned long addr)
++{
++ return &bm_pte[pte_index(addr)];
++}
++
++void __init early_ioremap_init(void)
++{
++ pmd_t *pmd;
++
++ if (early_ioremap_debug)
++ printk(KERN_INFO "early_ioremap_init()\n");
++
++ pmd = early_ioremap_pmd(fix_to_virt(FIX_BTMAP_BEGIN));
++ memset(bm_pte, 0, sizeof(bm_pte));
++ make_lowmem_page_readonly(bm_pte, XENFEAT_writable_page_tables);
++ pmd_populate_kernel(&init_mm, pmd, bm_pte);
++
++ /*
++ * The boot-ioremap range spans multiple pmds, for which
++ * we are not prepared:
++ */
++ if (pmd != early_ioremap_pmd(fix_to_virt(FIX_BTMAP_END))) {
++ WARN_ON(1);
++ printk(KERN_WARNING "pmd %p != %p\n",
++ pmd, early_ioremap_pmd(fix_to_virt(FIX_BTMAP_END)));
++ printk(KERN_WARNING "fix_to_virt(FIX_BTMAP_BEGIN): %08lx\n",
++ fix_to_virt(FIX_BTMAP_BEGIN));
++ printk(KERN_WARNING "fix_to_virt(FIX_BTMAP_END): %08lx\n",
++ fix_to_virt(FIX_BTMAP_END));
++
++ printk(KERN_WARNING "FIX_BTMAP_END: %d\n", FIX_BTMAP_END);
++ printk(KERN_WARNING "FIX_BTMAP_BEGIN: %d\n",
++ FIX_BTMAP_BEGIN);
++ }
++}
++
++void __init early_ioremap_clear(void)
++{
++ pmd_t *pmd;
++
++ if (early_ioremap_debug)
++ printk(KERN_INFO "early_ioremap_clear()\n");
++
++ pmd = early_ioremap_pmd(fix_to_virt(FIX_BTMAP_BEGIN));
++ pmd_clear(pmd);
++ make_lowmem_page_writable(bm_pte, XENFEAT_writable_page_tables);
++ /* paravirt_release_pte(__pa(bm_pte) >> PAGE_SHIFT); */
++ __flush_tlb_all();
++}
++
++void __init early_ioremap_reset(void)
++{
++ enum fixed_addresses idx;
++ unsigned long addr, phys;
++ pte_t *pte;
++
++ after_paging_init = 1;
++ for (idx = FIX_BTMAP_BEGIN; idx >= FIX_BTMAP_END; idx--) {
++ addr = fix_to_virt(idx);
++ pte = early_ioremap_pte(addr);
++ if (pte_present(*pte)) {
++ phys = __pte_val(*pte) & PAGE_MASK;
++ set_fixmap(idx, phys);
++ }
++ }
++}
++
++static void __init __early_set_fixmap(enum fixed_addresses idx,
++ unsigned long phys, pgprot_t flags)
++{
++ unsigned long addr = __fix_to_virt(idx);
++ pte_t *pte;
++
++ if (idx >= __end_of_fixed_addresses) {
++ BUG();
++ return;
++ }
++ pte = early_ioremap_pte(addr);
++ if (pgprot_val(flags))
++ set_pte(pte, pfn_pte_ma(phys >> PAGE_SHIFT, flags));
++ else
++ pte_clear(NULL, addr, pte);
++ __flush_tlb_one(addr);
++}
++
++static inline void __init early_set_fixmap(enum fixed_addresses idx,
++ unsigned long phys)
++{
++ if (after_paging_init)
++ set_fixmap(idx, phys);
++ else
++ __early_set_fixmap(idx, phys, PAGE_KERNEL);
++}
++
++static inline void __init early_clear_fixmap(enum fixed_addresses idx)
++{
++ if (after_paging_init)
++ clear_fixmap(idx);
++ else
++ __early_set_fixmap(idx, 0, __pgprot(0));
++}
++
++
++int __initdata early_ioremap_nested;
++
++static int __init check_early_ioremap_leak(void)
++{
++ if (!early_ioremap_nested)
++ return 0;
++
++ printk(KERN_WARNING
++ "Debug warning: early ioremap leak of %d areas detected.\n",
++ early_ioremap_nested);
++ printk(KERN_WARNING
++ "please boot with early_ioremap_debug and report the dmesg.\n");
++ WARN_ON(1);
++
++ return 1;
++}
++late_initcall(check_early_ioremap_leak);
++
++void __init *early_ioremap(unsigned long phys_addr, unsigned long size)
++{
++ unsigned long offset, last_addr;
++ unsigned int nrpages, nesting;
++ enum fixed_addresses idx0, idx;
++
++ WARN_ON(system_state != SYSTEM_BOOTING);
++
++ nesting = early_ioremap_nested;
++ if (early_ioremap_debug) {
++ printk(KERN_INFO "early_ioremap(%08lx, %08lx) [%d] => ",
++ phys_addr, size, nesting);
++ dump_stack();
++ }
++
++ /* Don't allow wraparound or zero size */
++ last_addr = phys_addr + size - 1;
++ if (!size || last_addr < phys_addr) {
++ WARN_ON(1);
++ return NULL;
++ }
++
++ if (nesting >= FIX_BTMAPS_NESTING) {
++ WARN_ON(1);
++ return NULL;
++ }
++ early_ioremap_nested++;
++ /*
++ * Mappings have to be page-aligned
++ */
++ offset = phys_addr & ~PAGE_MASK;
++ phys_addr &= PAGE_MASK;
++ size = PAGE_ALIGN(last_addr) - phys_addr;
++
++ /*
++ * Mappings have to fit in the FIX_BTMAP area.
++ */
++ nrpages = size >> PAGE_SHIFT;
++ if (nrpages > NR_FIX_BTMAPS) {
++ WARN_ON(1);
++ return NULL;
++ }
++
++ /*
++ * Ok, go for it..
++ */
++ idx0 = FIX_BTMAP_BEGIN - NR_FIX_BTMAPS*nesting;
++ idx = idx0;
++ while (nrpages > 0) {
++ early_set_fixmap(idx, phys_addr);
++ phys_addr += PAGE_SIZE;
++ --idx;
++ --nrpages;
++ }
++ if (early_ioremap_debug)
++ printk(KERN_CONT "%08lx + %08lx\n", offset, fix_to_virt(idx0));
++
++ return (void *) (offset + fix_to_virt(idx0));
++}
++
++void __init early_iounmap(void *addr, unsigned long size)
++{
++ unsigned long virt_addr;
++ unsigned long offset;
++ unsigned int nrpages;
++ enum fixed_addresses idx;
++ int nesting;
++
++ nesting = --early_ioremap_nested;
++ if (WARN_ON(nesting < 0))
++ return;
++
++ if (early_ioremap_debug) {
++ printk(KERN_INFO "early_iounmap(%p, %08lx) [%d]\n", addr,
++ size, nesting);
++ dump_stack();
++ }
++
++ virt_addr = (unsigned long)addr;
++ if (virt_addr < fix_to_virt(FIX_BTMAP_BEGIN)) {
++ WARN_ON(1);
++ return;
++ }
++ offset = virt_addr & ~PAGE_MASK;
++ nrpages = PAGE_ALIGN(offset + size - 1) >> PAGE_SHIFT;
++
++ idx = FIX_BTMAP_BEGIN - NR_FIX_BTMAPS*nesting;
++ while (nrpages > 0) {
++ early_clear_fixmap(idx);
++ --idx;
++ --nrpages;
++ }
++}
++
++void __this_fixmap_does_not_exist(void)
++{
++ WARN_ON(1);
++}
+diff --git a/arch/x86/mm/pageattr-xen.c b/arch/x86/mm/pageattr-xen.c
+new file mode 100644
+index 0000000..9bf43e1
+--- /dev/null
++++ b/arch/x86/mm/pageattr-xen.c
+@@ -0,0 +1,1141 @@
++/*
++ * Copyright 2002 Andi Kleen, SuSE Labs.
++ * Thanks to Ben LaHaise for precious feedback.
++ */
++#include <linux/highmem.h>
++#include <linux/bootmem.h>
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/mm.h>
++#include <linux/interrupt.h>
++#include <linux/seq_file.h>
++#include <linux/debugfs.h>
++
++#include <asm/e820.h>
++#include <asm/processor.h>
++#include <asm/tlbflush.h>
++#include <asm/sections.h>
++#include <asm/uaccess.h>
++#include <asm/pgalloc.h>
++#include <asm/proto.h>
++#include <asm/pat.h>
++
++/*
++ * The current flushing context - we pass it instead of 5 arguments:
++ */
++struct cpa_data {
++ unsigned long vaddr;
++ pgprot_t mask_set;
++ pgprot_t mask_clr;
++ int numpages;
++ int flushtlb;
++ unsigned long pfn;
++ unsigned force_split : 1;
++};
++
++#ifdef CONFIG_X86_64
++
++static inline unsigned long highmap_start_pfn(void)
++{
++ return __pa(_text) >> PAGE_SHIFT;
++}
++
++static inline unsigned long highmap_end_pfn(void)
++{
++ return __pa(round_up((unsigned long)_end, PMD_SIZE)) >> PAGE_SHIFT;
++}
++
++#endif
++
++#ifdef CONFIG_DEBUG_PAGEALLOC
++# define debug_pagealloc 1
++#else
++# define debug_pagealloc 0
++#endif
++
++static inline int
++within(unsigned long addr, unsigned long start, unsigned long end)
++{
++ return addr >= start && addr < end;
++}
++
++/*
++ * Flushing functions
++ */
++
++/**
++ * clflush_cache_range - flush a cache range with clflush
++ * @addr: virtual start address
++ * @size: number of bytes to flush
++ *
++ * clflush is an unordered instruction which needs fencing with mfence
++ * to avoid ordering issues.
++ */
++void clflush_cache_range(void *vaddr, unsigned int size)
++{
++ void *vend = vaddr + size - 1;
++
++ mb();
++
++ for (; vaddr < vend; vaddr += boot_cpu_data.x86_clflush_size)
++ clflush(vaddr);
++ /*
++ * Flush any possible final partial cacheline:
++ */
++ clflush(vend);
++
++ mb();
++}
++
++static void __cpa_flush_all(void *arg)
++{
++ unsigned long cache = (unsigned long)arg;
++
++ /*
++ * Flush all to work around Errata in early athlons regarding
++ * large page flushing.
++ */
++ __flush_tlb_all();
++
++ if (cache && boot_cpu_data.x86_model >= 4)
++ wbinvd();
++}
++
++static void cpa_flush_all(unsigned long cache)
++{
++ BUG_ON(irqs_disabled());
++
++ on_each_cpu(__cpa_flush_all, (void *) cache, 1, 1);
++}
++
++static void __cpa_flush_range(void *arg)
++{
++ /*
++ * We could optimize that further and do individual per page
++ * tlb invalidates for a low number of pages. Caveat: we must
++ * flush the high aliases on 64bit as well.
++ */
++ __flush_tlb_all();
++}
++
++static void cpa_flush_range(unsigned long start, int numpages, int cache)
++{
++ unsigned int i, level;
++ unsigned long addr;
++
++ BUG_ON(irqs_disabled());
++ WARN_ON(PAGE_ALIGN(start) != start);
++
++ on_each_cpu(__cpa_flush_range, NULL, 1, 1);
++
++ if (!cache)
++ return;
++
++ /*
++ * We only need to flush on one CPU,
++ * clflush is a MESI-coherent instruction that
++ * will cause all other CPUs to flush the same
++ * cachelines:
++ */
++ for (i = 0, addr = start; i < numpages; i++, addr += PAGE_SIZE) {
++ pte_t *pte = lookup_address(addr, &level);
++
++ /*
++ * Only flush present addresses:
++ */
++ if (pte && (__pte_val(*pte) & _PAGE_PRESENT))
++ clflush_cache_range((void *) addr, PAGE_SIZE);
++ }
++}
++
++/*
++ * Certain areas of memory on x86 require very specific protection flags,
++ * for example the BIOS area or kernel text. Callers don't always get this
++ * right (again, ioremap() on BIOS memory is not uncommon) so this function
++ * checks and fixes these known static required protection bits.
++ */
++static inline pgprot_t static_protections(pgprot_t prot, unsigned long address,
++ unsigned long pfn)
++{
++ pgprot_t forbidden = __pgprot(0);
++
++#ifndef CONFIG_XEN
++ /*
++ * The BIOS area between 640k and 1Mb needs to be executable for
++ * PCI BIOS based config access (CONFIG_PCI_GOBIOS) support.
++ */
++ if (within(pfn, BIOS_BEGIN >> PAGE_SHIFT, BIOS_END >> PAGE_SHIFT))
++ pgprot_val(forbidden) |= _PAGE_NX;
++#endif
++
++ /*
++ * The kernel text needs to be executable for obvious reasons
++ * Does not cover __inittext since that is gone later on. On
++ * 64bit we do not enforce !NX on the low mapping
++ */
++ if (within(address, (unsigned long)_text, (unsigned long)_etext))
++ pgprot_val(forbidden) |= _PAGE_NX;
++
++ /*
++ * The .rodata section needs to be read-only. Using the pfn
++ * catches all aliases.
++ */
++ if (within(pfn, __pa((unsigned long)__start_rodata) >> PAGE_SHIFT,
++ __pa((unsigned long)__end_rodata) >> PAGE_SHIFT))
++ pgprot_val(forbidden) |= _PAGE_RW;
++
++ prot = __pgprot(pgprot_val(prot) & ~pgprot_val(forbidden));
++
++ return prot;
++}
++
++/*
++ * Lookup the page table entry for a virtual address. Return a pointer
++ * to the entry and the level of the mapping.
++ *
++ * Note: We return pud and pmd either when the entry is marked large
++ * or when the present bit is not set. Otherwise we would return a
++ * pointer to a nonexisting mapping.
++ */
++pte_t *lookup_address(unsigned long address, unsigned int *level)
++{
++ pgd_t *pgd = pgd_offset_k(address);
++ pud_t *pud;
++ pmd_t *pmd;
++
++ *level = PG_LEVEL_NONE;
++
++ if (pgd_none(*pgd))
++ return NULL;
++
++ pud = pud_offset(pgd, address);
++ if (pud_none(*pud))
++ return NULL;
++
++ *level = PG_LEVEL_1G;
++ if (pud_large(*pud) || !pud_present(*pud))
++ return (pte_t *)pud;
++
++ pmd = pmd_offset(pud, address);
++ if (pmd_none(*pmd))
++ return NULL;
++
++ *level = PG_LEVEL_2M;
++ if (pmd_large(*pmd) || !pmd_present(*pmd))
++ return (pte_t *)pmd;
++
++ *level = PG_LEVEL_4K;
++
++ return pte_offset_kernel(pmd, address);
++}
++
++/*
++ * Set the new pmd in all the pgds we know about:
++ */
++static void __set_pmd_pte(pte_t *kpte, unsigned long address,
++ unsigned int level, pte_t pte)
++{
++ /* change init_mm */
++ switch(level) {
++ case PG_LEVEL_2M:
++ xen_l2_entry_update((pmd_t *)kpte, __pmd_ma(__pte_val(pte)));
++ break;
++#ifdef CONFIG_X86_64
++ case PG_LEVEL_1G:
++ xen_l3_entry_update((pud_t *)kpte, __pud_ma(__pte_val(pte)));
++ break;
++#endif
++ default:
++ BUG();
++ }
++#ifdef CONFIG_X86_32
++ if (!SHARED_KERNEL_PMD) {
++ struct page *page;
++
++ list_for_each_entry(page, &pgd_list, lru) {
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++
++ pgd = (pgd_t *)page_address(page) + pgd_index(address);
++ pud = pud_offset(pgd, address);
++ pmd = pmd_offset(pud, address);
++ xen_l2_entry_update(pmd, __pmd_ma(__pte_val(pte)));
++ }
++ }
++#endif
++}
++
++static int
++try_preserve_large_page(pte_t *kpte, unsigned long address,
++ struct cpa_data *cpa)
++{
++ unsigned long nextpage_addr, numpages, pmask, psize, flags, addr, pfn;
++ pte_t new_pte, old_pte, *tmp;
++ pgprot_t old_prot, new_prot;
++ int i, do_split = 1;
++ unsigned int level;
++
++ if (cpa->force_split)
++ return 1;
++
++ spin_lock_irqsave(&pgd_lock, flags);
++ /*
++ * Check for races, another CPU might have split this page
++ * up already:
++ */
++ tmp = lookup_address(address, &level);
++ if (tmp != kpte)
++ goto out_unlock;
++
++ switch (level) {
++ case PG_LEVEL_2M:
++ psize = PMD_PAGE_SIZE;
++ pmask = PMD_PAGE_MASK;
++ break;
++#ifdef CONFIG_X86_64
++ case PG_LEVEL_1G:
++ psize = PUD_PAGE_SIZE;
++ pmask = PUD_PAGE_MASK;
++ break;
++#endif
++ default:
++ do_split = -EINVAL;
++ goto out_unlock;
++ }
++
++ /*
++ * Calculate the number of pages, which fit into this large
++ * page starting at address:
++ */
++ nextpage_addr = (address + psize) & pmask;
++ numpages = (nextpage_addr - address) >> PAGE_SHIFT;
++ if (numpages < cpa->numpages)
++ cpa->numpages = numpages;
++
++ /*
++ * We are safe now. Check whether the new pgprot is the same:
++ */
++ old_pte = *kpte;
++ old_prot = new_prot = pte_pgprot(old_pte);
++
++ pgprot_val(new_prot) &= ~pgprot_val(cpa->mask_clr);
++ pgprot_val(new_prot) |= pgprot_val(cpa->mask_set);
++
++ /*
++ * old_pte points to the large page base address. So we need
++ * to add the offset of the virtual address:
++ */
++ pfn = pte_pfn(old_pte) + ((address & (psize - 1)) >> PAGE_SHIFT);
++ cpa->pfn = pfn;
++
++ new_prot = static_protections(new_prot, address, pfn);
++
++ /*
++ * We need to check the full range, whether
++ * static_protection() requires a different pgprot for one of
++ * the pages in the range we try to preserve:
++ */
++ if (pfn < max_mapnr) {
++ addr = address + PAGE_SIZE;
++ for (i = 1; i < cpa->numpages && ++pfn < max_mapnr;
++ i++, addr += PAGE_SIZE) {
++ pgprot_t chk_prot = static_protections(new_prot, addr, pfn);
++
++ if (pgprot_val(chk_prot) != pgprot_val(new_prot))
++ goto out_unlock;
++ }
++ }
++
++ /*
++ * If there are no changes, return. maxpages has been updated
++ * above:
++ */
++ if (pgprot_val(new_prot) == pgprot_val(old_prot)) {
++ do_split = 0;
++ goto out_unlock;
++ }
++
++ /*
++ * We need to change the attributes. Check, whether we can
++ * change the large page in one go. We request a split, when
++ * the address is not aligned and the number of pages is
++ * smaller than the number of pages in the large page. Note
++ * that we limited the number of possible pages already to
++ * the number of pages in the large page.
++ */
++ if (address == (nextpage_addr - psize) && cpa->numpages == numpages) {
++ /*
++ * The address is aligned and the number of pages
++ * covers the full page.
++ */
++ new_pte = pfn_pte_ma(__pte_mfn(old_pte), canon_pgprot(new_prot));
++ __set_pmd_pte(kpte, address, level, new_pte);
++ cpa->flushtlb = 1;
++ do_split = 0;
++ }
++
++out_unlock:
++ spin_unlock_irqrestore(&pgd_lock, flags);
++
++ return do_split;
++}
++
++static LIST_HEAD(page_pool);
++static unsigned long pool_size, pool_pages, pool_low;
++static unsigned long pool_used, pool_failed;
++
++static void cpa_fill_pool(struct page **ret)
++{
++ gfp_t gfp = GFP_KERNEL;
++ unsigned long flags;
++ struct page *p;
++
++ /*
++ * Avoid recursion (on debug-pagealloc) and also signal
++ * our priority to get to these pagetables:
++ */
++ if (current->flags & PF_MEMALLOC)
++ return;
++ current->flags |= PF_MEMALLOC;
++
++ /*
++ * Allocate atomically from atomic contexts:
++ */
++ if (in_atomic() || irqs_disabled() || debug_pagealloc)
++ gfp = GFP_ATOMIC | __GFP_NORETRY | __GFP_NOWARN;
++
++ while (pool_pages < pool_size || (ret && !*ret)) {
++ p = alloc_pages(gfp, 0);
++ if (!p) {
++ pool_failed++;
++ break;
++ }
++ /*
++ * If the call site needs a page right now, provide it:
++ */
++ if (ret && !*ret) {
++ *ret = p;
++ continue;
++ }
++ spin_lock_irqsave(&pgd_lock, flags);
++ list_add(&p->lru, &page_pool);
++ pool_pages++;
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ }
++
++ current->flags &= ~PF_MEMALLOC;
++}
++
++#define SHIFT_MB (20 - PAGE_SHIFT)
++#define ROUND_MB_GB ((1 << 10) - 1)
++#define SHIFT_MB_GB 10
++#define POOL_PAGES_PER_GB 16
++
++void __init cpa_init(void)
++{
++ struct sysinfo si;
++ unsigned long gb;
++
++ si_meminfo(&si);
++ /*
++ * Calculate the number of pool pages:
++ *
++ * Convert totalram (nr of pages) to MiB and round to the next
++ * GiB. Shift MiB to Gib and multiply the result by
++ * POOL_PAGES_PER_GB:
++ */
++ if (debug_pagealloc) {
++ gb = ((si.totalram >> SHIFT_MB) + ROUND_MB_GB) >> SHIFT_MB_GB;
++ pool_size = POOL_PAGES_PER_GB * gb;
++ } else {
++ pool_size = 1;
++ }
++ pool_low = pool_size;
++
++ cpa_fill_pool(NULL);
++ printk(KERN_DEBUG
++ "CPA: page pool initialized %lu of %lu pages preallocated\n",
++ pool_pages, pool_size);
++}
++
++static int split_large_page(pte_t *kpte, unsigned long address)
++{
++ unsigned long flags, mfn, mfninc = 1;
++ unsigned int i, level;
++ pte_t *pbase, *tmp;
++ pgprot_t ref_prot;
++ struct page *base;
++
++ /*
++ * Get a page from the pool. The pool list is protected by the
++ * pgd_lock, which we have to take anyway for the split
++ * operation:
++ */
++ spin_lock_irqsave(&pgd_lock, flags);
++ if (list_empty(&page_pool)) {
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ base = NULL;
++ cpa_fill_pool(&base);
++ if (!base)
++ return -ENOMEM;
++ spin_lock_irqsave(&pgd_lock, flags);
++ } else {
++ base = list_first_entry(&page_pool, struct page, lru);
++ list_del(&base->lru);
++ pool_pages--;
++
++ if (pool_pages < pool_low)
++ pool_low = pool_pages;
++ }
++
++ /*
++ * Check for races, another CPU might have split this page
++ * up for us already:
++ */
++ tmp = lookup_address(address, &level);
++ if (tmp != kpte)
++ goto out_unlock;
++
++ pbase = (pte_t *)page_address(base);
++ paravirt_alloc_pte(&init_mm, page_to_pfn(base));
++ ref_prot = pte_pgprot(pte_clrhuge(*kpte));
++
++#ifdef CONFIG_X86_64
++ if (level == PG_LEVEL_1G) {
++ mfninc = PMD_PAGE_SIZE >> PAGE_SHIFT;
++ pgprot_val(ref_prot) |= _PAGE_PSE;
++ }
++#endif
++
++ /*
++ * Get the target mfn from the original entry:
++ */
++ mfn = __pte_mfn(*kpte);
++ for (i = 0; i < PTRS_PER_PTE; i++, mfn += mfninc)
++ set_pte(&pbase[i], pfn_pte_ma(mfn, ref_prot));
++
++ /*
++ * Install the new, split up pagetable. Important details here:
++ *
++ * On Intel the NX bit of all levels must be cleared to make a
++ * page executable. See section 4.13.2 of Intel 64 and IA-32
++ * Architectures Software Developer's Manual).
++ *
++ * Mark the entry present. The current mapping might be
++ * set to not present, which we preserved above.
++ */
++ if (HYPERVISOR_update_va_mapping((unsigned long)pbase,
++ mk_pte(base, PAGE_KERNEL_RO), 0))
++ BUG();
++ ref_prot = pte_pgprot(pte_mkexec(pte_clrhuge(*kpte)));
++ pgprot_val(ref_prot) |= _PAGE_PRESENT;
++ __set_pmd_pte(kpte, address, level, mk_pte(base, ref_prot));
++ base = NULL;
++
++out_unlock:
++ /*
++ * If we dropped out via the lookup_address check under
++ * pgd_lock then stick the page back into the pool:
++ */
++ if (base) {
++ list_add(&base->lru, &page_pool);
++ pool_pages++;
++ } else
++ pool_used++;
++ spin_unlock_irqrestore(&pgd_lock, flags);
++
++ return 0;
++}
++
++static int __change_page_attr(struct cpa_data *cpa, int primary)
++{
++ unsigned long address = cpa->vaddr;
++ int do_split, err;
++ unsigned int level;
++ pte_t *kpte, old_pte;
++
++repeat:
++ kpte = lookup_address(address, &level);
++ if (!kpte)
++ return 0;
++
++ old_pte = *kpte;
++ if (!__pte_val(old_pte)) {
++ if (!primary)
++ return 0;
++ printk(KERN_WARNING "CPA: called for zero pte. "
++ "vaddr = %lx cpa->vaddr = %lx\n", address,
++ cpa->vaddr);
++ WARN_ON(1);
++ return -EINVAL;
++ }
++
++ if (level == PG_LEVEL_4K) {
++ pte_t new_pte;
++ pgprot_t new_prot = pte_pgprot(old_pte);
++ unsigned long mfn = __pte_mfn(old_pte);
++
++ pgprot_val(new_prot) &= ~pgprot_val(cpa->mask_clr);
++ pgprot_val(new_prot) |= pgprot_val(cpa->mask_set);
++
++ new_prot = static_protections(new_prot, address,
++ mfn_to_local_pfn(mfn));
++
++ /*
++ * We need to keep the mfn from the existing PTE,
++ * after all we're only going to change it's attributes
++ * not the memory it points to
++ */
++ new_pte = pfn_pte_ma(mfn, canon_pgprot(new_prot));
++ cpa->pfn = mfn_to_local_pfn(mfn);
++ /*
++ * Do we really change anything ?
++ */
++ if (__pte_val(old_pte) != __pte_val(new_pte)) {
++ set_pte_atomic(kpte, new_pte);
++ cpa->flushtlb = 1;
++ }
++ cpa->numpages = 1;
++ return 0;
++ }
++
++ /*
++ * Check, whether we can keep the large page intact
++ * and just change the pte:
++ */
++ do_split = try_preserve_large_page(kpte, address, cpa);
++ /*
++ * When the range fits into the existing large page,
++ * return. cp->numpages and cpa->tlbflush have been updated in
++ * try_large_page:
++ */
++ if (do_split <= 0)
++ return do_split;
++
++ /*
++ * We have to split the large page:
++ */
++ err = split_large_page(kpte, address);
++ if (!err) {
++ cpa->flushtlb = 1;
++ goto repeat;
++ }
++
++ return err;
++}
++
++static int __change_page_attr_set_clr(struct cpa_data *cpa, int checkalias);
++
++static int cpa_process_alias(struct cpa_data *cpa)
++{
++ struct cpa_data alias_cpa;
++ int ret = 0;
++
++ if (cpa->pfn > max_pfn_mapped)
++ return 0;
++
++ /*
++ * No need to redo, when the primary call touched the direct
++ * mapping already:
++ */
++ if (!within(cpa->vaddr, PAGE_OFFSET,
++ PAGE_OFFSET + (max_pfn_mapped << PAGE_SHIFT))) {
++
++ alias_cpa = *cpa;
++ alias_cpa.vaddr = (unsigned long) __va(cpa->pfn << PAGE_SHIFT);
++
++ ret = __change_page_attr_set_clr(&alias_cpa, 0);
++ }
++
++#ifdef CONFIG_X86_64
++ if (ret)
++ return ret;
++ /*
++ * No need to redo, when the primary call touched the high
++ * mapping already:
++ */
++ if (within(cpa->vaddr, (unsigned long) _text, (unsigned long) _end))
++ return 0;
++
++ /*
++ * If the physical address is inside the kernel map, we need
++ * to touch the high mapped kernel as well:
++ */
++ if (!within(cpa->pfn, highmap_start_pfn(), highmap_end_pfn()))
++ return 0;
++
++ alias_cpa = *cpa;
++ alias_cpa.vaddr =
++ (cpa->pfn << PAGE_SHIFT) + __START_KERNEL_map;
++
++ /*
++ * The high mapping range is imprecise, so ignore the return value.
++ */
++ __change_page_attr_set_clr(&alias_cpa, 0);
++#endif
++ return ret;
++}
++
++static int __change_page_attr_set_clr(struct cpa_data *cpa, int checkalias)
++{
++ int ret, numpages = cpa->numpages;
++
++ while (numpages) {
++ /*
++ * Store the remaining nr of pages for the large page
++ * preservation check.
++ */
++ cpa->numpages = numpages;
++
++ ret = __change_page_attr(cpa, checkalias);
++ if (ret)
++ return ret;
++
++ if (checkalias) {
++ ret = cpa_process_alias(cpa);
++ if (ret)
++ return ret;
++ }
++
++ /*
++ * Adjust the number of pages with the result of the
++ * CPA operation. Either a large page has been
++ * preserved or a single page update happened.
++ */
++ BUG_ON(cpa->numpages > numpages);
++ numpages -= cpa->numpages;
++ cpa->vaddr += cpa->numpages * PAGE_SIZE;
++ }
++ return 0;
++}
++
++static inline int cache_attr(pgprot_t attr)
++{
++ return pgprot_val(attr) &
++ (_PAGE_PAT | _PAGE_PAT_LARGE | _PAGE_PWT | _PAGE_PCD);
++}
++
++static int change_page_attr_set_clr(unsigned long addr, int numpages,
++ pgprot_t mask_set, pgprot_t mask_clr,
++ int force_split)
++{
++ struct cpa_data cpa;
++ int ret, cache, checkalias;
++
++ /*
++ * Check, if we are requested to change a not supported
++ * feature:
++ */
++ mask_set = canon_pgprot(mask_set);
++ mask_clr = canon_pgprot(mask_clr);
++ if (!pgprot_val(mask_set) && !pgprot_val(mask_clr) && !force_split)
++ return 0;
++
++ /* Ensure we are PAGE_SIZE aligned */
++ if (addr & ~PAGE_MASK) {
++ addr &= PAGE_MASK;
++ /*
++ * People should not be passing in unaligned addresses:
++ */
++ WARN_ON_ONCE(1);
++ }
++
++ cpa.vaddr = addr;
++ cpa.numpages = numpages;
++ cpa.mask_set = mask_set;
++ cpa.mask_clr = mask_clr;
++ cpa.flushtlb = 0;
++ cpa.force_split = force_split;
++
++ /* No alias checking for _NX bit modifications */
++ checkalias = (pgprot_val(mask_set) | pgprot_val(mask_clr)) != _PAGE_NX;
++
++ ret = __change_page_attr_set_clr(&cpa, checkalias);
++
++ /*
++ * Check whether we really changed something:
++ */
++ if (!cpa.flushtlb)
++ goto out;
++
++ /*
++ * No need to flush, when we did not set any of the caching
++ * attributes:
++ */
++ cache = cache_attr(mask_set);
++
++ /*
++ * On success we use clflush, when the CPU supports it to
++ * avoid the wbindv. If the CPU does not support it and in the
++ * error case we fall back to cpa_flush_all (which uses
++ * wbindv):
++ */
++ if (!ret && cpu_has_clflush)
++ cpa_flush_range(addr, numpages, cache);
++ else
++ cpa_flush_all(cache);
++
++out:
++ cpa_fill_pool(NULL);
++
++ return ret;
++}
++
++static inline int change_page_attr_set(unsigned long addr, int numpages,
++ pgprot_t mask)
++{
++ return change_page_attr_set_clr(addr, numpages, mask, __pgprot(0), 0);
++}
++
++static inline int change_page_attr_clear(unsigned long addr, int numpages,
++ pgprot_t mask)
++{
++ return change_page_attr_set_clr(addr, numpages, __pgprot(0), mask, 0);
++}
++
++int _set_memory_uc(unsigned long addr, int numpages)
++{
++ /*
++ * for now UC MINUS. see comments in ioremap_nocache()
++ */
++ return change_page_attr_set(addr, numpages,
++ __pgprot(_PAGE_CACHE_UC_MINUS));
++}
++
++int set_memory_uc(unsigned long addr, int numpages)
++{
++ /*
++ * for now UC MINUS. see comments in ioremap_nocache()
++ */
++ if (reserve_memtype(addr, addr + numpages * PAGE_SIZE,
++ _PAGE_CACHE_UC_MINUS, NULL))
++ return -EINVAL;
++
++ return _set_memory_uc(addr, numpages);
++}
++EXPORT_SYMBOL(set_memory_uc);
++
++int _set_memory_wc(unsigned long addr, int numpages)
++{
++ return change_page_attr_set(addr, numpages,
++ __pgprot(_PAGE_CACHE_WC));
++}
++
++int set_memory_wc(unsigned long addr, int numpages)
++{
++ if (!pat_wc_enabled)
++ return set_memory_uc(addr, numpages);
++
++ if (reserve_memtype(addr, addr + numpages * PAGE_SIZE,
++ _PAGE_CACHE_WC, NULL))
++ return -EINVAL;
++
++ return _set_memory_wc(addr, numpages);
++}
++EXPORT_SYMBOL(set_memory_wc);
++
++int _set_memory_wb(unsigned long addr, int numpages)
++{
++ return change_page_attr_clear(addr, numpages,
++ __pgprot(_PAGE_CACHE_MASK));
++}
++
++int set_memory_wb(unsigned long addr, int numpages)
++{
++ free_memtype(addr, addr + numpages * PAGE_SIZE);
++
++ return _set_memory_wb(addr, numpages);
++}
++EXPORT_SYMBOL(set_memory_wb);
++
++int set_memory_x(unsigned long addr, int numpages)
++{
++ return change_page_attr_clear(addr, numpages, __pgprot(_PAGE_NX));
++}
++EXPORT_SYMBOL(set_memory_x);
++
++int set_memory_nx(unsigned long addr, int numpages)
++{
++ return change_page_attr_set(addr, numpages, __pgprot(_PAGE_NX));
++}
++EXPORT_SYMBOL(set_memory_nx);
++
++int set_memory_ro(unsigned long addr, int numpages)
++{
++ return change_page_attr_clear(addr, numpages, __pgprot(_PAGE_RW));
++}
++
++int set_memory_rw(unsigned long addr, int numpages)
++{
++ return change_page_attr_set(addr, numpages, __pgprot(_PAGE_RW));
++}
++
++int set_memory_np(unsigned long addr, int numpages)
++{
++ return change_page_attr_clear(addr, numpages, __pgprot(_PAGE_PRESENT));
++}
++
++int set_memory_4k(unsigned long addr, int numpages)
++{
++ return change_page_attr_set_clr(addr, numpages, __pgprot(0),
++ __pgprot(0), 1);
++}
++
++int set_pages_uc(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_uc(addr, numpages);
++}
++EXPORT_SYMBOL(set_pages_uc);
++
++int set_pages_wb(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_wb(addr, numpages);
++}
++EXPORT_SYMBOL(set_pages_wb);
++
++int set_pages_x(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_x(addr, numpages);
++}
++EXPORT_SYMBOL(set_pages_x);
++
++int set_pages_nx(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_nx(addr, numpages);
++}
++EXPORT_SYMBOL(set_pages_nx);
++
++int set_pages_ro(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_ro(addr, numpages);
++}
++
++int set_pages_rw(struct page *page, int numpages)
++{
++ unsigned long addr = (unsigned long)page_address(page);
++
++ return set_memory_rw(addr, numpages);
++}
++
++#ifdef CONFIG_DEBUG_PAGEALLOC
++
++static int __set_pages_p(struct page *page, int numpages)
++{
++ struct cpa_data cpa = { .vaddr = (unsigned long) page_address(page),
++ .numpages = numpages,
++ .mask_set = __pgprot(_PAGE_PRESENT | _PAGE_RW),
++ .mask_clr = __pgprot(0)};
++
++ return __change_page_attr_set_clr(&cpa, 1);
++}
++
++static int __set_pages_np(struct page *page, int numpages)
++{
++ struct cpa_data cpa = { .vaddr = (unsigned long) page_address(page),
++ .numpages = numpages,
++ .mask_set = __pgprot(0),
++ .mask_clr = __pgprot(_PAGE_PRESENT | _PAGE_RW)};
++
++ return __change_page_attr_set_clr(&cpa, 1);
++}
++
++void kernel_map_pages(struct page *page, int numpages, int enable)
++{
++ if (PageHighMem(page))
++ return;
++ if (!enable) {
++ debug_check_no_locks_freed(page_address(page),
++ numpages * PAGE_SIZE);
++ }
++
++ /*
++ * If page allocator is not up yet then do not call c_p_a():
++ */
++ if (!debug_pagealloc_enabled)
++ return;
++
++ /*
++ * The return value is ignored as the calls cannot fail.
++ * Large pages are kept enabled at boot time, and are
++ * split up quickly with DEBUG_PAGEALLOC. If a splitup
++ * fails here (due to temporary memory shortage) no damage
++ * is done because we just keep the largepage intact up
++ * to the next attempt when it will likely be split up:
++ */
++ if (enable)
++ __set_pages_p(page, numpages);
++ else
++ __set_pages_np(page, numpages);
++
++ /*
++ * We should perform an IPI and flush all tlbs,
++ * but that can deadlock->flush only current cpu:
++ */
++ __flush_tlb_all();
++
++ /*
++ * Try to refill the page pool here. We can do this only after
++ * the tlb flush.
++ */
++ cpa_fill_pool(NULL);
++}
++
++#ifdef CONFIG_DEBUG_FS
++static int dpa_show(struct seq_file *m, void *v)
++{
++ seq_puts(m, "DEBUG_PAGEALLOC\n");
++ seq_printf(m, "pool_size : %lu\n", pool_size);
++ seq_printf(m, "pool_pages : %lu\n", pool_pages);
++ seq_printf(m, "pool_low : %lu\n", pool_low);
++ seq_printf(m, "pool_used : %lu\n", pool_used);
++ seq_printf(m, "pool_failed : %lu\n", pool_failed);
++
++ return 0;
++}
++
++static int dpa_open(struct inode *inode, struct file *filp)
++{
++ return single_open(filp, dpa_show, NULL);
++}
++
++static const struct file_operations dpa_fops = {
++ .open = dpa_open,
++ .read = seq_read,
++ .llseek = seq_lseek,
++ .release = single_release,
++};
++
++static int __init debug_pagealloc_proc_init(void)
++{
++ struct dentry *de;
++
++ de = debugfs_create_file("debug_pagealloc", 0600, NULL, NULL,
++ &dpa_fops);
++ if (!de)
++ return -ENOMEM;
++
++ return 0;
++}
++__initcall(debug_pagealloc_proc_init);
++#endif
++
++#ifdef CONFIG_HIBERNATION
++
++bool kernel_page_present(struct page *page)
++{
++ unsigned int level;
++ pte_t *pte;
++
++ if (PageHighMem(page))
++ return false;
++
++ pte = lookup_address((unsigned long)page_address(page), &level);
++ return (__pte_val(*pte) & _PAGE_PRESENT);
++}
++
++#endif /* CONFIG_HIBERNATION */
++
++#endif /* CONFIG_DEBUG_PAGEALLOC */
++
++static inline int in_secondary_range(unsigned long va)
++{
++#ifdef CONFIG_X86_64
++ return va >= VMALLOC_START && va < VMALLOC_END;
++#else
++ return va >= (unsigned long)high_memory;
++#endif
++}
++
++static void __make_page_readonly(unsigned long va)
++{
++ pte_t *pte;
++ unsigned int level;
++
++ pte = lookup_address(va, &level);
++ BUG_ON(!pte || level != PG_LEVEL_4K);
++ if (HYPERVISOR_update_va_mapping(va, pte_wrprotect(*pte), 0))
++ BUG();
++ if (in_secondary_range(va)) {
++ unsigned long pfn = pte_pfn(*pte);
++
++#ifdef CONFIG_HIGHMEM
++ if (pfn >= highstart_pfn)
++ kmap_flush_unused(); /* flush stale writable kmaps */
++ else
++#endif
++ __make_page_readonly((unsigned long)__va(pfn << PAGE_SHIFT));
++ }
++}
++
++static void __make_page_writable(unsigned long va)
++{
++ pte_t *pte;
++ unsigned int level;
++
++ pte = lookup_address(va, &level);
++ BUG_ON(!pte || level != PG_LEVEL_4K);
++ if (HYPERVISOR_update_va_mapping(va, pte_mkwrite(*pte), 0))
++ BUG();
++ if (in_secondary_range(va)) {
++ unsigned long pfn = pte_pfn(*pte);
++
++#ifdef CONFIG_HIGHMEM
++ if (pfn < highstart_pfn)
++#endif
++ __make_page_writable((unsigned long)__va(pfn << PAGE_SHIFT));
++ }
++}
++
++void make_page_readonly(void *va, unsigned int feature)
++{
++ if (!xen_feature(feature))
++ __make_page_readonly((unsigned long)va);
++}
++
++void make_page_writable(void *va, unsigned int feature)
++{
++ if (!xen_feature(feature))
++ __make_page_writable((unsigned long)va);
++}
++
++void make_pages_readonly(void *va, unsigned int nr, unsigned int feature)
++{
++ unsigned long addr;
++
++ if (xen_feature(feature))
++ return;
++
++ for (addr = (unsigned long)va; nr--; addr += PAGE_SIZE)
++ __make_page_readonly(addr);
++}
++
++void make_pages_writable(void *va, unsigned int nr, unsigned int feature)
++{
++ unsigned long addr;
++
++ if (xen_feature(feature))
++ return;
++
++ for (addr = (unsigned long)va; nr--; addr += PAGE_SIZE)
++ __make_page_writable(addr);
++}
++
++/*
++ * The testcases use internal knowledge of the implementation that shouldn't
++ * be exposed to the rest of the kernel. Include these directly here.
++ */
++#ifdef CONFIG_CPA_DEBUG
++#include "pageattr-test.c"
++#endif
+diff --git a/arch/x86/mm/pat-xen.c b/arch/x86/mm/pat-xen.c
+new file mode 100644
+index 0000000..6fc94ce
+--- /dev/null
++++ b/arch/x86/mm/pat-xen.c
+@@ -0,0 +1,608 @@
++/*
++ * Handle caching attributes in page tables (PAT)
++ *
++ * Authors: Venkatesh Pallipadi <venkatesh.pallipadi at intel.com>
++ * Suresh B Siddha <suresh.b.siddha at intel.com>
++ *
++ * Loosely based on earlier PAT patchset from Eric Biederman and Andi Kleen.
++ */
++
++#include <linux/mm.h>
++#include <linux/kernel.h>
++#include <linux/gfp.h>
++#include <linux/fs.h>
++#include <linux/bootmem.h>
++
++#include <asm/msr.h>
++#include <asm/tlbflush.h>
++#include <asm/processor.h>
++#include <asm/page.h>
++#include <asm/pgtable.h>
++#include <asm/pat.h>
++#include <asm/e820.h>
++#include <asm/cacheflush.h>
++#include <asm/fcntl.h>
++#include <asm/mtrr.h>
++#include <asm/io.h>
++
++#ifdef CONFIG_X86_PAT
++int __read_mostly pat_wc_enabled = 1;
++
++void __cpuinit pat_disable(char *reason)
++{
++ pat_wc_enabled = 0;
++ printk(KERN_INFO "%s\n", reason);
++}
++
++static int __init nopat(char *str)
++{
++ pat_disable("PAT support disabled.");
++ return 0;
++}
++early_param("nopat", nopat);
++#endif
++
++static u64 __read_mostly boot_pat_state;
++
++enum {
++ PAT_UC = 0, /* uncached */
++ PAT_WC = 1, /* Write combining */
++ PAT_WT = 4, /* Write Through */
++ PAT_WP = 5, /* Write Protected */
++ PAT_WB = 6, /* Write Back (default) */
++ PAT_UC_MINUS = 7, /* UC, but can be overriden by MTRR */
++};
++
++#define PAT(x,y) ((u64)PAT_ ## y << ((x)*8))
++
++void pat_init(void)
++{
++ u64 pat;
++
++ if (!pat_wc_enabled)
++ return;
++
++ /* Paranoia check. */
++ if (!cpu_has_pat) {
++ printk(KERN_ERR "PAT enabled, but CPU feature cleared\n");
++ /*
++ * Panic if this happens on the secondary CPU, and we
++ * switched to PAT on the boot CPU. We have no way to
++ * undo PAT.
++ */
++ BUG_ON(boot_pat_state);
++ }
++
++#ifndef CONFIG_XEN
++ /* Set PWT to Write-Combining. All other bits stay the same */
++ /*
++ * PTE encoding used in Linux:
++ * PAT
++ * |PCD
++ * ||PWT
++ * |||
++ * 000 WB _PAGE_CACHE_WB
++ * 001 WC _PAGE_CACHE_WC
++ * 010 UC- _PAGE_CACHE_UC_MINUS
++ * 011 UC _PAGE_CACHE_UC
++ * PAT bit unused
++ */
++ pat = PAT(0,WB) | PAT(1,WC) | PAT(2,UC_MINUS) | PAT(3,UC) |
++ PAT(4,WB) | PAT(5,WC) | PAT(6,UC_MINUS) | PAT(7,UC);
++
++ /* Boot CPU check */
++ if (!boot_pat_state)
++ rdmsrl(MSR_IA32_CR_PAT, boot_pat_state);
++
++ wrmsrl(MSR_IA32_CR_PAT, pat);
++#else
++ /*
++ * PAT settings are part of the hypervisor interface, and their
++ * assignment cannot be changed.
++ */
++ rdmsrl(MSR_IA32_CR_PAT, pat);
++ if (!boot_pat_state)
++ boot_pat_state = pat;
++#endif
++ printk(KERN_INFO "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n",
++ smp_processor_id(), boot_pat_state, pat);
++}
++
++#undef PAT
++
++static char *cattr_name(unsigned long flags)
++{
++ switch (flags & _PAGE_CACHE_MASK) {
++ case _PAGE_CACHE_UC: return "uncached";
++ case _PAGE_CACHE_UC_MINUS: return "uncached-minus";
++ case _PAGE_CACHE_WB: return "write-back";
++ case _PAGE_CACHE_WC: return "write-combining";
++ case _PAGE_CACHE_WP: return "write-protected";
++ case _PAGE_CACHE_WT: return "write-through";
++ default: return "broken";
++ }
++}
++
++/*
++ * The global memtype list keeps track of memory type for specific
++ * physical memory areas. Conflicting memory types in different
++ * mappings can cause CPU cache corruption. To avoid this we keep track.
++ *
++ * The list is sorted based on starting address and can contain multiple
++ * entries for each address (this allows reference counting for overlapping
++ * areas). All the aliases have the same cache attributes of course.
++ * Zero attributes are represented as holes.
++ *
++ * Currently the data structure is a list because the number of mappings
++ * are expected to be relatively small. If this should be a problem
++ * it could be changed to a rbtree or similar.
++ *
++ * memtype_lock protects the whole list.
++ */
++
++struct memtype {
++ u64 start;
++ u64 end;
++ unsigned long type;
++ struct list_head nd;
++};
++
++static LIST_HEAD(memtype_list);
++static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */
++
++/*
++ * Does intersection of PAT memory type and MTRR memory type and returns
++ * the resulting memory type as PAT understands it.
++ * (Type in pat and mtrr will not have same value)
++ * The intersection is based on "Effective Memory Type" tables in IA-32
++ * SDM vol 3a
++ */
++static int pat_x_mtrr_type(u64 start, u64 end, unsigned long prot,
++ unsigned long *ret_prot)
++{
++ unsigned long pat_type;
++ u8 mtrr_type;
++
++ pat_type = prot & _PAGE_CACHE_MASK;
++ prot &= (~_PAGE_CACHE_MASK);
++
++ /*
++ * We return the PAT request directly for types where PAT takes
++ * precedence with respect to MTRR and for UC_MINUS.
++ * Consistency checks with other PAT requests is done later
++ * while going through memtype list.
++ */
++ if (pat_type == _PAGE_CACHE_WC) {
++ *ret_prot = prot | _PAGE_CACHE_WC;
++ return 0;
++ } else if (pat_type == _PAGE_CACHE_UC_MINUS) {
++ *ret_prot = prot | _PAGE_CACHE_UC_MINUS;
++ return 0;
++ } else if (pat_type == _PAGE_CACHE_UC) {
++ *ret_prot = prot | _PAGE_CACHE_UC;
++ return 0;
++ }
++
++ /*
++ * Look for MTRR hint to get the effective type in case where PAT
++ * request is for WB.
++ */
++ mtrr_type = mtrr_type_lookup(start, end);
++
++ if (mtrr_type == MTRR_TYPE_UNCACHABLE) {
++ *ret_prot = prot | _PAGE_CACHE_UC;
++ } else if (mtrr_type == MTRR_TYPE_WRCOMB) {
++ *ret_prot = prot | _PAGE_CACHE_WC;
++ } else {
++ *ret_prot = prot | _PAGE_CACHE_WB;
++ }
++
++ return 0;
++}
++
++/*
++ * req_type typically has one of the:
++ * - _PAGE_CACHE_WB
++ * - _PAGE_CACHE_WC
++ * - _PAGE_CACHE_UC_MINUS
++ * - _PAGE_CACHE_UC
++ *
++ * req_type will have a special case value '-1', when requester want to inherit
++ * the memory type from mtrr (if WB), existing PAT, defaulting to UC_MINUS.
++ *
++ * If ret_type is NULL, function will return an error if it cannot reserve the
++ * region with req_type. If ret_type is non-null, function will return
++ * available type in ret_type in case of no error. In case of any error
++ * it will return a negative return value.
++ */
++int reserve_memtype(u64 start, u64 end, unsigned long req_type,
++ unsigned long *ret_type)
++{
++ struct memtype *new_entry = NULL;
++ struct memtype *parse;
++ unsigned long actual_type;
++ int err = 0;
++
++ /* Only track when pat_wc_enabled */
++ if (!pat_wc_enabled) {
++ /* This is identical to page table setting without PAT */
++ if (ret_type) {
++ if (req_type == -1) {
++ *ret_type = _PAGE_CACHE_WB;
++ } else {
++ *ret_type = req_type;
++ }
++ }
++ return 0;
++ }
++
++ /* Low ISA region is always mapped WB in page table. No need to track */
++ if (start >= ISA_START_ADDRESS && (end - 1) <= ISA_END_ADDRESS) {
++ if (ret_type)
++ *ret_type = _PAGE_CACHE_WB;
++
++ return 0;
++ }
++
++ if (req_type == -1) {
++ /*
++ * Call mtrr_lookup to get the type hint. This is an
++ * optimization for /dev/mem mmap'ers into WB memory (BIOS
++ * tools and ACPI tools). Use WB request for WB memory and use
++ * UC_MINUS otherwise.
++ */
++ u8 mtrr_type = mtrr_type_lookup(start, end);
++
++ if (mtrr_type == MTRR_TYPE_WRBACK) {
++ req_type = _PAGE_CACHE_WB;
++ actual_type = _PAGE_CACHE_WB;
++ } else {
++ req_type = _PAGE_CACHE_UC_MINUS;
++ actual_type = _PAGE_CACHE_UC_MINUS;
++ }
++ } else {
++ req_type &= _PAGE_CACHE_MASK;
++ err = pat_x_mtrr_type(start, end, req_type, &actual_type);
++ }
++
++ if (err) {
++ if (ret_type)
++ *ret_type = actual_type;
++
++ return -EINVAL;
++ }
++
++ new_entry = kmalloc(sizeof(struct memtype), GFP_KERNEL);
++ if (!new_entry)
++ return -ENOMEM;
++
++ new_entry->start = start;
++ new_entry->end = end;
++ new_entry->type = actual_type;
++
++ if (ret_type)
++ *ret_type = actual_type;
++
++ spin_lock(&memtype_lock);
++
++ /* Search for existing mapping that overlaps the current range */
++ list_for_each_entry(parse, &memtype_list, nd) {
++ struct memtype *saved_ptr;
++
++ if (parse->start >= end) {
++ pr_debug("New Entry\n");
++ list_add(&new_entry->nd, parse->nd.prev);
++ new_entry = NULL;
++ break;
++ }
++
++ if (start <= parse->start && end >= parse->start) {
++ if (actual_type != parse->type && ret_type) {
++ actual_type = parse->type;
++ *ret_type = actual_type;
++ new_entry->type = actual_type;
++ }
++
++ if (actual_type != parse->type) {
++ printk(
++ KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n",
++ current->comm, current->pid,
++ start, end,
++ cattr_name(actual_type),
++ cattr_name(parse->type));
++ err = -EBUSY;
++ break;
++ }
++
++ saved_ptr = parse;
++ /*
++ * Check to see whether the request overlaps more
++ * than one entry in the list
++ */
++ list_for_each_entry_continue(parse, &memtype_list, nd) {
++ if (end <= parse->start) {
++ break;
++ }
++
++ if (actual_type != parse->type) {
++ printk(
++ KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n",
++ current->comm, current->pid,
++ start, end,
++ cattr_name(actual_type),
++ cattr_name(parse->type));
++ err = -EBUSY;
++ break;
++ }
++ }
++
++ if (err) {
++ break;
++ }
++
++ pr_debug("Overlap at 0x%Lx-0x%Lx\n",
++ saved_ptr->start, saved_ptr->end);
++ /* No conflict. Go ahead and add this new entry */
++ list_add(&new_entry->nd, saved_ptr->nd.prev);
++ new_entry = NULL;
++ break;
++ }
++
++ if (start < parse->end) {
++ if (actual_type != parse->type && ret_type) {
++ actual_type = parse->type;
++ *ret_type = actual_type;
++ new_entry->type = actual_type;
++ }
++
++ if (actual_type != parse->type) {
++ printk(
++ KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n",
++ current->comm, current->pid,
++ start, end,
++ cattr_name(actual_type),
++ cattr_name(parse->type));
++ err = -EBUSY;
++ break;
++ }
++
++ saved_ptr = parse;
++ /*
++ * Check to see whether the request overlaps more
++ * than one entry in the list
++ */
++ list_for_each_entry_continue(parse, &memtype_list, nd) {
++ if (end <= parse->start) {
++ break;
++ }
++
++ if (actual_type != parse->type) {
++ printk(
++ KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n",
++ current->comm, current->pid,
++ start, end,
++ cattr_name(actual_type),
++ cattr_name(parse->type));
++ err = -EBUSY;
++ break;
++ }
++ }
++
++ if (err) {
++ break;
++ }
++
++ pr_debug(KERN_INFO "Overlap at 0x%Lx-0x%Lx\n",
++ saved_ptr->start, saved_ptr->end);
++ /* No conflict. Go ahead and add this new entry */
++ list_add(&new_entry->nd, &saved_ptr->nd);
++ new_entry = NULL;
++ break;
++ }
++ }
++
++ if (err) {
++ printk(KERN_INFO
++ "reserve_memtype failed 0x%Lx-0x%Lx, track %s, req %s\n",
++ start, end, cattr_name(new_entry->type),
++ cattr_name(req_type));
++ kfree(new_entry);
++ spin_unlock(&memtype_lock);
++ return err;
++ }
++
++ if (new_entry) {
++ /* No conflict. Not yet added to the list. Add to the tail */
++ list_add_tail(&new_entry->nd, &memtype_list);
++ pr_debug("New Entry\n");
++ }
++
++ if (ret_type) {
++ pr_debug(
++ "reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s, ret %s\n",
++ start, end, cattr_name(actual_type),
++ cattr_name(req_type), cattr_name(*ret_type));
++ } else {
++ pr_debug(
++ "reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s\n",
++ start, end, cattr_name(actual_type),
++ cattr_name(req_type));
++ }
++
++ spin_unlock(&memtype_lock);
++ return err;
++}
++
++int free_memtype(u64 start, u64 end)
++{
++ struct memtype *ml;
++ int err = -EINVAL;
++
++ /* Only track when pat_wc_enabled */
++ if (!pat_wc_enabled) {
++ return 0;
++ }
++
++ /* Low ISA region is always mapped WB. No need to track */
++ if (start >= ISA_START_ADDRESS && end <= ISA_END_ADDRESS) {
++ return 0;
++ }
++
++ spin_lock(&memtype_lock);
++ list_for_each_entry(ml, &memtype_list, nd) {
++ if (ml->start == start && ml->end == end) {
++ list_del(&ml->nd);
++ kfree(ml);
++ err = 0;
++ break;
++ }
++ }
++ spin_unlock(&memtype_lock);
++
++ if (err) {
++ printk(KERN_INFO "%s:%d freeing invalid memtype %Lx-%Lx\n",
++ current->comm, current->pid, start, end);
++ }
++
++ pr_debug("free_memtype request 0x%Lx-0x%Lx\n", start, end);
++ return err;
++}
++
++
++/*
++ * /dev/mem mmap interface. The memtype used for mapping varies:
++ * - Use UC for mappings with O_SYNC flag
++ * - Without O_SYNC flag, if there is any conflict in reserve_memtype,
++ * inherit the memtype from existing mapping.
++ * - Else use UC_MINUS memtype (for backward compatibility with existing
++ * X drivers.
++ */
++pgprot_t phys_mem_access_prot(struct file *file, unsigned long mfn,
++ unsigned long size, pgprot_t vma_prot)
++{
++ return vma_prot;
++}
++
++#ifdef CONFIG_NONPROMISC_DEVMEM
++/* This check is done in drivers/char/mem.c in case of NONPROMISC_DEVMEM*/
++static inline int range_is_allowed(unsigned long mfn, unsigned long size)
++{
++ return 1;
++}
++#else
++static inline int range_is_allowed(unsigned long mfn, unsigned long size)
++{
++ u64 from = ((u64)mfn) << PAGE_SHIFT;
++ u64 to = from + size;
++ u64 cursor = from;
++
++ while (cursor < to) {
++ if (!devmem_is_allowed(mfn)) {
++ printk(KERN_INFO
++ "Program %s tried to access /dev/mem between %Lx->%Lx.\n",
++ current->comm, from, to);
++ return 0;
++ }
++ cursor += PAGE_SIZE;
++ mfn++;
++ }
++ return 1;
++}
++#endif /* CONFIG_NONPROMISC_DEVMEM */
++
++int phys_mem_access_prot_allowed(struct file *file, unsigned long mfn,
++ unsigned long size, pgprot_t *vma_prot)
++{
++ u64 addr = (u64)mfn << PAGE_SHIFT;
++ unsigned long sz, flags = _PAGE_CACHE_UC_MINUS;
++ int retval;
++
++ if (!range_is_allowed(mfn, size))
++ return 0;
++
++ if (file->f_flags & O_SYNC) {
++ flags = _PAGE_CACHE_UC;
++ }
++
++#ifndef CONFIG_X86_32
++#ifndef CONFIG_XEN /* Xen sets correct MTRR type on non-RAM for us. */
++ /*
++ * On the PPro and successors, the MTRRs are used to set
++ * memory types for physical addresses outside main memory,
++ * so blindly setting UC or PWT on those pages is wrong.
++ * For Pentiums and earlier, the surround logic should disable
++ * caching for the high addresses through the KEN pin, but
++ * we maintain the tradition of paranoia in this code.
++ */
++ if (!pat_wc_enabled &&
++ ! ( test_bit(X86_FEATURE_MTRR, boot_cpu_data.x86_capability) ||
++ test_bit(X86_FEATURE_K6_MTRR, boot_cpu_data.x86_capability) ||
++ test_bit(X86_FEATURE_CYRIX_ARR, boot_cpu_data.x86_capability) ||
++ test_bit(X86_FEATURE_CENTAUR_MCR, boot_cpu_data.x86_capability)) &&
++ (pfn << PAGE_SHIFT) >= __pa(high_memory)) {
++ flags = _PAGE_CACHE_UC;
++ }
++#endif
++#endif
++
++ /*
++ * With O_SYNC, we can only take UC mapping. Fail if we cannot.
++ * Without O_SYNC, we want to get
++ * - WB for WB-able memory and no other conflicting mappings
++ * - UC_MINUS for non-WB-able memory with no other conflicting mappings
++ * - Inherit from confliting mappings otherwise
++ */
++ if (flags != _PAGE_CACHE_UC_MINUS) {
++ retval = reserve_memtype(addr, addr + size, flags, NULL);
++ } else {
++ retval = reserve_memtype(addr, addr + size, -1, &flags);
++ }
++
++ if (retval < 0)
++ return 0;
++
++ for (sz = 0; sz < size; ++mfn, sz += PAGE_SIZE) {
++ unsigned long pfn = mfn_to_local_pfn(mfn);
++
++ if (pfn <= max_pfn_mapped &&
++ ioremap_change_attr((unsigned long)__va(pfn << PAGE_SHIFT),
++ PAGE_SIZE, flags) < 0) {
++ free_memtype(addr, addr + size);
++ printk(KERN_INFO
++ "%s:%d /dev/mem ioremap_change_attr failed %s for %Lx-%Lx\n",
++ current->comm, current->pid,
++ cattr_name(flags),
++ addr, addr + size);
++ return 0;
++ }
++ }
++
++ *vma_prot = __pgprot((pgprot_val(*vma_prot) & ~_PAGE_CACHE_MASK) |
++ flags);
++ return 1;
++}
++
++void map_devmem(unsigned long mfn, unsigned long size, pgprot_t vma_prot)
++{
++ u64 addr = (u64)mfn << PAGE_SHIFT;
++ unsigned long flags;
++ unsigned long want_flags = (pgprot_val(vma_prot) & _PAGE_CACHE_MASK);
++
++ reserve_memtype(addr, addr + size, want_flags, &flags);
++ if (flags != want_flags) {
++ printk(KERN_INFO
++ "%s:%d /dev/mem expected mapping type %s for %Lx-%Lx, got %s\n",
++ current->comm, current->pid,
++ cattr_name(want_flags),
++ addr, (unsigned long long)(addr + size),
++ cattr_name(flags));
++ }
++}
++
++void unmap_devmem(unsigned long mfn, unsigned long size, pgprot_t vma_prot)
++{
++ u64 addr = (u64)mfn << PAGE_SHIFT;
++
++ free_memtype(addr, addr + size);
++}
++
+diff --git a/arch/x86/mm/pgtable-xen.c b/arch/x86/mm/pgtable-xen.c
+new file mode 100644
+index 0000000..ec4fbf7
+--- /dev/null
++++ b/arch/x86/mm/pgtable-xen.c
+@@ -0,0 +1,699 @@
++#include <linux/mm.h>
++#include <linux/module.h>
++#include <xen/features.h>
++#include <asm/pgalloc.h>
++#include <asm/pgtable.h>
++#include <asm/tlb.h>
++#include <asm/hypervisor.h>
++#include <asm/mmu_context.h>
++
++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;
++}
++
++pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
++{
++ struct page *pte;
++
++#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);
++#endif
++ if (pte) {
++ pgtable_page_ctor(pte);
++ SetPageForeign(pte, __pte_free);
++ init_page_count(pte);
++ }
++ return pte;
++}
++
++void __pte_free(pgtable_t pte)
++{
++ if (!PageHighMem(pte)) {
++ unsigned long va = (unsigned long)page_address(pte);
++ unsigned int level;
++ pte_t *ptep = lookup_address(va, &level);
++
++ BUG_ON(!ptep || level != PG_LEVEL_4K || !pte_present(*ptep));
++ if (!pte_write(*ptep)
++ && HYPERVISOR_update_va_mapping(va,
++ mk_pte(pte, PAGE_KERNEL),
++ 0))
++ BUG();
++ } else
++#ifdef CONFIG_HIGHPTE
++ ClearPagePinned(pte);
++#else
++ BUG();
++#endif
++
++ ClearPageForeign(pte);
++ init_page_count(pte);
++ pgtable_page_dtor(pte);
++ __free_page(pte);
++}
++
++void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte)
++{
++ pgtable_page_dtor(pte);
++ paravirt_release_pte(page_to_pfn(pte));
++ tlb_remove_page(tlb, pte);
++}
++
++#if PAGETABLE_LEVELS > 2
++pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long address)
++{
++ struct page *pmd;
++
++ pmd = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
++ if (!pmd)
++ return NULL;
++ SetPageForeign(pmd, __pmd_free);
++ init_page_count(pmd);
++ return page_address(pmd);
++}
++
++void __pmd_free(pgtable_t pmd)
++{
++ unsigned long va = (unsigned long)page_address(pmd);
++ unsigned int level;
++ pte_t *ptep = lookup_address(va, &level);
++
++ BUG_ON(!ptep || level != PG_LEVEL_4K || !pte_present(*ptep));
++ if (!pte_write(*ptep)
++ && HYPERVISOR_update_va_mapping(va, mk_pte(pmd, PAGE_KERNEL), 0))
++ BUG();
++
++ ClearPageForeign(pmd);
++ init_page_count(pmd);
++ __free_page(pmd);
++}
++
++void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd)
++{
++ paravirt_release_pmd(__pa(pmd) >> PAGE_SHIFT);
++ tlb_remove_page(tlb, virt_to_page(pmd));
++}
++
++#if PAGETABLE_LEVELS > 3
++void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pud)
++{
++ paravirt_release_pud(__pa(pud) >> PAGE_SHIFT);
++ tlb_remove_page(tlb, virt_to_page(pud));
++}
++#endif /* PAGETABLE_LEVELS > 3 */
++#endif /* PAGETABLE_LEVELS > 2 */
++
++#ifndef CONFIG_X86_64
++#define TASK_SIZE64 TASK_SIZE
++#endif
++
++static void _pin_lock(struct mm_struct *mm, int lock) {
++ if (lock)
++ spin_lock(&mm->page_table_lock);
++#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS
++ /* While mm->page_table_lock protects us against insertions and
++ * removals of higher level page table pages, it doesn't protect
++ * against updates of pte-s. Such updates, however, require the
++ * pte pages to be in consistent state (unpinned+writable or
++ * pinned+readonly). The pinning and attribute changes, however
++ * cannot be done atomically, which is why such updates must be
++ * prevented from happening concurrently.
++ * Note that no pte lock can ever elsewhere be acquired nesting
++ * with an already acquired one in the same mm, or with the mm's
++ * page_table_lock already acquired, as that would break in the
++ * non-split case (where all these are actually resolving to the
++ * one page_table_lock). Thus acquiring all of them here is not
++ * going to result in dead locks, and the order of acquires
++ * doesn't matter.
++ */
++ {
++ pgd_t *pgd = mm->pgd;
++ unsigned g;
++
++ for (g = 0; g <= ((TASK_SIZE64-1) / PGDIR_SIZE); g++, pgd++) {
++ pud_t *pud;
++ unsigned u;
++
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, 0);
++ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
++ pmd_t *pmd;
++ unsigned m;
++
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, 0);
++ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
++ spinlock_t *ptl;
++
++ if (pmd_none(*pmd))
++ continue;
++ ptl = pte_lockptr(0, pmd);
++ if (lock)
++ spin_lock(ptl);
++ else
++ spin_unlock(ptl);
++ }
++ }
++ }
++ }
++#endif
++ if (!lock)
++ spin_unlock(&mm->page_table_lock);
++}
++#define pin_lock(mm) _pin_lock(mm, 1)
++#define pin_unlock(mm) _pin_lock(mm, 0)
++
++#define PIN_BATCH sizeof(void *)
++static DEFINE_PER_CPU(multicall_entry_t[PIN_BATCH], pb_mcl);
++
++static inline unsigned int pgd_walk_set_prot(struct page *page, pgprot_t flags,
++ unsigned int cpu, unsigned int seq)
++{
++ unsigned long pfn = page_to_pfn(page);
++
++ if (PageHighMem(page)) {
++ if (pgprot_val(flags) & _PAGE_RW)
++ ClearPagePinned(page);
++ else
++ SetPagePinned(page);
++ } else {
++ MULTI_update_va_mapping(per_cpu(pb_mcl, cpu) + seq,
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, flags), 0);
++ if (unlikely(++seq == PIN_BATCH)) {
++ if (unlikely(HYPERVISOR_multicall_check(per_cpu(pb_mcl, cpu),
++ PIN_BATCH, NULL)))
++ BUG();
++ seq = 0;
++ }
++ }
++
++ return seq;
++}
++
++static void pgd_walk(pgd_t *pgd_base, pgprot_t flags)
++{
++ pgd_t *pgd = pgd_base;
++ pud_t *pud;
++ pmd_t *pmd;
++ int g,u,m;
++ unsigned int cpu, seq;
++ multicall_entry_t *mcl;
++
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return;
++
++ cpu = get_cpu();
++
++ /*
++ * Cannot iterate up to USER_PTRS_PER_PGD on x86-64 as these pagetables
++ * may not be the 'current' task's pagetables (e.g., current may be
++ * 32-bit, but the pagetables may be for a 64-bit task).
++ * Subtracting 1 from TASK_SIZE64 means the loop limit is correct
++ * regardless of whether TASK_SIZE64 is a multiple of PGDIR_SIZE.
++ */
++ for (g = 0, seq = 0; g <= ((TASK_SIZE64-1) / PGDIR_SIZE); g++, pgd++) {
++ if (pgd_none(*pgd))
++ continue;
++ pud = pud_offset(pgd, 0);
++ if (PTRS_PER_PUD > 1) /* not folded */
++ seq = pgd_walk_set_prot(virt_to_page(pud),flags,cpu,seq);
++ for (u = 0; u < PTRS_PER_PUD; u++, pud++) {
++ if (pud_none(*pud))
++ continue;
++ pmd = pmd_offset(pud, 0);
++ if (PTRS_PER_PMD > 1) /* not folded */
++ seq = pgd_walk_set_prot(virt_to_page(pmd),flags,cpu,seq);
++ for (m = 0; m < PTRS_PER_PMD; m++, pmd++) {
++ if (pmd_none(*pmd))
++ continue;
++ seq = pgd_walk_set_prot(pmd_page(*pmd),flags,cpu,seq);
++ }
++ }
++ }
++
++ mcl = per_cpu(pb_mcl, cpu);
++#ifdef CONFIG_X86_64
++ if (unlikely(seq > PIN_BATCH - 2)) {
++ if (unlikely(HYPERVISOR_multicall_check(mcl, seq, NULL)))
++ BUG();
++ seq = 0;
++ }
++ MULTI_update_va_mapping(mcl + seq,
++ (unsigned long)__user_pgd(pgd_base),
++ pfn_pte(virt_to_phys(__user_pgd(pgd_base))>>PAGE_SHIFT, flags),
++ 0);
++ MULTI_update_va_mapping(mcl + seq + 1,
++ (unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH);
++ if (unlikely(HYPERVISOR_multicall_check(mcl, seq + 2, NULL)))
++ BUG();
++#else
++ if (likely(seq != 0)) {
++ MULTI_update_va_mapping(per_cpu(pb_mcl, cpu) + seq,
++ (unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH);
++ if (unlikely(HYPERVISOR_multicall_check(per_cpu(pb_mcl, cpu),
++ seq + 1, NULL)))
++ BUG();
++ } else if(HYPERVISOR_update_va_mapping((unsigned long)pgd_base,
++ pfn_pte(virt_to_phys(pgd_base)>>PAGE_SHIFT, flags),
++ UVMF_TLB_FLUSH))
++ BUG();
++#endif
++
++ put_cpu();
++}
++
++static void __pgd_pin(pgd_t *pgd)
++{
++ pgd_walk(pgd, PAGE_KERNEL_RO);
++ kmap_flush_unused();
++ xen_pgd_pin(__pa(pgd)); /* kernel */
++#ifdef CONFIG_X86_64
++ xen_pgd_pin(__pa(__user_pgd(pgd))); /* user */
++#endif
++ SetPagePinned(virt_to_page(pgd));
++}
++
++static void __pgd_unpin(pgd_t *pgd)
++{
++ xen_pgd_unpin(__pa(pgd));
++#ifdef CONFIG_X86_64
++ xen_pgd_unpin(__pa(__user_pgd(pgd)));
++#endif
++ pgd_walk(pgd, PAGE_KERNEL);
++ ClearPagePinned(virt_to_page(pgd));
++}
++
++static void pgd_test_and_unpin(pgd_t *pgd)
++{
++ if (PagePinned(virt_to_page(pgd)))
++ __pgd_unpin(pgd);
++}
++
++void mm_pin(struct mm_struct *mm)
++{
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ pin_lock(mm);
++ __pgd_pin(mm->pgd);
++ pin_unlock(mm);
++}
++
++void mm_unpin(struct mm_struct *mm)
++{
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ pin_lock(mm);
++ __pgd_unpin(mm->pgd);
++ pin_unlock(mm);
++}
++
++void mm_pin_all(void)
++{
++ struct page *page;
++ unsigned long flags;
++
++ if (xen_feature(XENFEAT_writable_page_tables))
++ return;
++
++ /*
++ * Allow uninterrupted access to the pgd_list. Also protects
++ * __pgd_pin() by disabling preemption.
++ * All other CPUs must be at a safe point (e.g., in stop_machine
++ * or offlined entirely).
++ */
++ spin_lock_irqsave(&pgd_lock, flags);
++ list_for_each_entry(page, &pgd_list, lru) {
++ if (!PagePinned(page))
++ __pgd_pin((pgd_t *)page_address(page));
++ }
++ spin_unlock_irqrestore(&pgd_lock, flags);
++}
++
++void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm)
++{
++ if (!PagePinned(virt_to_page(mm->pgd)))
++ mm_pin(mm);
++}
++
++void arch_exit_mmap(struct mm_struct *mm)
++{
++ struct task_struct *tsk = current;
++
++ task_lock(tsk);
++
++ /*
++ * We aggressively remove defunct pgd from cr3. We execute unmap_vmas()
++ * *much* faster this way, as no tlb flushes means bigger wrpt batches.
++ */
++ if (tsk->active_mm == mm) {
++ tsk->active_mm = &init_mm;
++ atomic_inc(&init_mm.mm_count);
++
++ switch_mm(mm, &init_mm, tsk);
++
++ atomic_dec(&mm->mm_count);
++ BUG_ON(atomic_read(&mm->mm_count) == 0);
++ }
++
++ task_unlock(tsk);
++
++ if (PagePinned(virt_to_page(mm->pgd))
++ && atomic_read(&mm->mm_count) == 1
++ && !mm->context.has_foreign_mappings)
++ mm_unpin(mm);
++}
++
++static inline void pgd_list_add(pgd_t *pgd)
++{
++ struct page *page = virt_to_page(pgd);
++
++ list_add(&page->lru, &pgd_list);
++}
++
++static inline void pgd_list_del(pgd_t *pgd)
++{
++ struct page *page = virt_to_page(pgd);
++
++ list_del(&page->lru);
++}
++
++#define UNSHARED_PTRS_PER_PGD \
++ (SHARED_KERNEL_PMD ? KERNEL_PGD_BOUNDARY : PTRS_PER_PGD)
++
++static void pgd_ctor(void *p)
++{
++ pgd_t *pgd = p;
++ unsigned long flags;
++
++ pgd_test_and_unpin(pgd);
++
++ /* Clear usermode parts of PGD */
++ memset(pgd, 0, KERNEL_PGD_BOUNDARY*sizeof(pgd_t));
++
++ spin_lock_irqsave(&pgd_lock, flags);
++
++ /* If the pgd points to a shared pagetable level (either the
++ ptes in non-PAE, or shared PMD in PAE), then just copy the
++ references from swapper_pg_dir. */
++ if (PAGETABLE_LEVELS == 2 ||
++ (PAGETABLE_LEVELS == 3 && SHARED_KERNEL_PMD) ||
++ PAGETABLE_LEVELS == 4) {
++ clone_pgd_range(pgd + KERNEL_PGD_BOUNDARY,
++ swapper_pg_dir + KERNEL_PGD_BOUNDARY,
++ KERNEL_PGD_PTRS);
++ paravirt_alloc_pmd_clone(__pa(pgd) >> PAGE_SHIFT,
++ __pa(swapper_pg_dir) >> PAGE_SHIFT,
++ KERNEL_PGD_BOUNDARY,
++ KERNEL_PGD_PTRS);
++ }
++
++#ifdef CONFIG_X86_64
++ /*
++ * Set level3_user_pgt for vsyscall area
++ */
++ __user_pgd(pgd)[pgd_index(VSYSCALL_START)] =
++ __pgd(__pa_symbol(level3_user_pgt) | _PAGE_TABLE);
++#endif
++
++#ifndef CONFIG_X86_PAE
++ /* list required to sync kernel mapping updates */
++ if (!SHARED_KERNEL_PMD)
++ pgd_list_add(pgd);
++#endif
++
++ spin_unlock_irqrestore(&pgd_lock, flags);
++}
++
++static void pgd_dtor(void *pgd)
++{
++ unsigned long flags; /* can be called from interrupt context */
++
++ if (!SHARED_KERNEL_PMD) {
++ spin_lock_irqsave(&pgd_lock, flags);
++ pgd_list_del(pgd);
++ spin_unlock_irqrestore(&pgd_lock, flags);
++ }
++
++ pgd_test_and_unpin(pgd);
++}
++
++/*
++ * 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.
++ * -- wli
++ */
++
++#ifdef CONFIG_X86_PAE
++/*
++ * Mop up any pmd pages which may still be attached to the pgd.
++ * Normally they will be freed by munmap/exit_mmap, but any pmd we
++ * preallocate which never got a corresponding vma will need to be
++ * freed manually.
++ */
++static void pgd_mop_up_pmds(struct mm_struct *mm, pgd_t *pgdp)
++{
++ int i;
++
++ for(i = 0; i < UNSHARED_PTRS_PER_PGD; i++) {
++ pgd_t pgd = pgdp[i];
++
++ if (__pgd_val(pgd) != 0) {
++ pmd_t *pmd = (pmd_t *)pgd_page_vaddr(pgd);
++
++ pgdp[i] = xen_make_pgd(0);
++
++ paravirt_release_pmd(pgd_val(pgd) >> PAGE_SHIFT);
++ pmd_free(mm, pmd);
++ }
++ }
++
++ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb))
++ xen_destroy_contiguous_region((unsigned long)pgdp, 0);
++}
++
++/*
++ * In PAE mode, we need to do a cr3 reload (=tlb flush) when
++ * updating the top-level pagetable entries to guarantee the
++ * processor notices the update. Since this is expensive, and
++ * all 4 top-level entries are used almost immediately in a
++ * new process's life, we just pre-populate them here.
++ *
++ * Also, if we're in a paravirt environment where the kernel pmd is
++ * not shared between pagetables (!SHARED_KERNEL_PMDS), we allocate
++ * and initialize the kernel pmds here.
++ */
++static int pgd_prepopulate_pmd(struct mm_struct *mm, pgd_t *pgd)
++{
++ pud_t *pud;
++ pmd_t *pmds[UNSHARED_PTRS_PER_PGD];
++ unsigned long addr, flags;
++ int i;
++
++ /*
++ * 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.
++ */
++ for (addr = i = 0; i < UNSHARED_PTRS_PER_PGD; i++, addr += PUD_SIZE) {
++ pmds[i] = pmd_alloc_one(mm, addr);
++ if (!pmds[i])
++ goto out_oom;
++ }
++
++ spin_lock_irqsave(&pgd_lock, flags);
++
++ /* Protect against save/restore: move below 4GB under pgd_lock. */
++ if (!xen_feature(XENFEAT_pae_pgdir_above_4gb)
++ && xen_create_contiguous_region((unsigned long)pgd, 0, 32)) {
++ spin_unlock_irqrestore(&pgd_lock, flags);
++out_oom:
++ while (i--)
++ pmd_free(mm, pmds[i]);
++ return 0;
++ }
++
++ /* Copy kernel pmd contents and write-protect the new pmds. */
++ pud = pud_offset(pgd, 0);
++ for (addr = i = 0; i < UNSHARED_PTRS_PER_PGD;
++ i++, pud++, addr += PUD_SIZE) {
++ if (i >= KERNEL_PGD_BOUNDARY) {
++ memcpy(pmds[i],
++ (pmd_t *)pgd_page_vaddr(swapper_pg_dir[i]),
++ sizeof(pmd_t) * PTRS_PER_PMD);
++ make_lowmem_page_readonly(
++ pmds[i], XENFEAT_writable_page_tables);
++ }
++
++ /* It is safe to poke machine addresses of pmds under the pgd_lock. */
++ pud_populate(mm, pud, pmds[i]);
++ }
++
++ /* List required to sync kernel mapping updates and
++ * to pin/unpin on save/restore. */
++ pgd_list_add(pgd);
++
++ spin_unlock_irqrestore(&pgd_lock, flags);
++
++ return 1;
++}
++
++void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd)
++{
++ struct page *page = virt_to_page(pmd);
++ unsigned long pfn = page_to_pfn(page);
++
++ paravirt_alloc_pmd(mm, __pa(pmd) >> PAGE_SHIFT);
++
++ /* Note: almost everything apart from _PAGE_PRESENT is
++ reserved at the pmd (PDPT) level. */
++ if (PagePinned(virt_to_page(mm->pgd))) {
++ BUG_ON(PageHighMem(page));
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, PAGE_KERNEL_RO), 0));
++ set_pud(pudp, __pud(__pa(pmd) | _PAGE_PRESENT));
++ } else
++ *pudp = __pud(__pa(pmd) | _PAGE_PRESENT);
++
++ /*
++ * According to Intel App note "TLBs, Paging-Structure Caches,
++ * and Their Invalidation", April 2007, document 317080-001,
++ * section 8.1: in PAE mode we explicitly have to flush the
++ * TLB via cr3 if the top-level pgd is changed...
++ */
++ if (mm == current->active_mm)
++ xen_tlb_flush();
++}
++#else /* !CONFIG_X86_PAE */
++/* No need to prepopulate any pagetable entries in non-PAE modes. */
++static int pgd_prepopulate_pmd(struct mm_struct *mm, pgd_t *pgd)
++{
++ return 1;
++}
++
++static void pgd_mop_up_pmds(struct mm_struct *mm, pgd_t *pgd)
++{
++}
++#endif /* CONFIG_X86_PAE */
++
++#ifdef CONFIG_X86_64
++/* We allocate two contiguous pages for kernel and user. */
++#define PGD_ORDER 1
++#else
++#define PGD_ORDER 0
++#endif
++
++pgd_t *pgd_alloc(struct mm_struct *mm)
++{
++ pgd_t *pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, PGD_ORDER);
++
++ /* so that alloc_pd can use it */
++ mm->pgd = pgd;
++ if (pgd)
++ pgd_ctor(pgd);
++
++ if (pgd && !pgd_prepopulate_pmd(mm, pgd)) {
++ free_pages((unsigned long)pgd, PGD_ORDER);
++ pgd = NULL;
++ }
++
++ return pgd;
++}
++
++void pgd_free(struct mm_struct *mm, pgd_t *pgd)
++{
++ /*
++ * 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_dtor(pgd);
++
++ pgd_mop_up_pmds(mm, pgd);
++ free_pages((unsigned long)pgd, PGD_ORDER);
++}
++
++/* blktap and gntdev need this, as otherwise they would implicitly (and
++ * needlessly, as they never use it) reference init_mm. */
++pte_t xen_ptep_get_and_clear_full(struct vm_area_struct *vma,
++ unsigned long addr, pte_t *ptep, int full)
++{
++ return ptep_get_and_clear_full(vma->vm_mm, addr, ptep, full);
++}
++EXPORT_SYMBOL_GPL(xen_ptep_get_and_clear_full);
++
++int ptep_set_access_flags(struct vm_area_struct *vma,
++ unsigned long address, pte_t *ptep,
++ pte_t entry, int dirty)
++{
++ int changed = !pte_same(*ptep, entry);
++
++ if (changed && dirty) {
++ if (likely(vma->vm_mm == current->mm)) {
++ if (HYPERVISOR_update_va_mapping(address,
++ entry,
++ (unsigned long)vma->vm_mm->cpu_vm_mask.bits|
++ UVMF_INVLPG|UVMF_MULTI))
++ BUG();
++ } else {
++ xen_l1_entry_update(ptep, entry);
++ flush_tlb_page(vma, address);
++ }
++ }
++
++ return changed;
++}
++
++int ptep_test_and_clear_young(struct vm_area_struct *vma,
++ unsigned long addr, pte_t *ptep)
++{
++ int ret = 0;
++
++ if (pte_young(*ptep))
++ ret = test_and_clear_bit(_PAGE_BIT_ACCESSED,
++ &ptep->pte);
++
++ if (ret)
++ pte_update(vma->vm_mm, addr, ptep);
++
++ return ret;
++}
++
++int ptep_clear_flush_young(struct vm_area_struct *vma,
++ unsigned long address, pte_t *ptep)
++{
++ pte_t pte = *ptep;
++ int young = pte_young(pte);
++
++ pte = pte_mkold(pte);
++ if (PagePinned(virt_to_page(vma->vm_mm->pgd)))
++ ptep_set_access_flags(vma, address, ptep, pte, young);
++ else if (young)
++ ptep->pte_low = pte.pte_low;
++
++ return young;
++}
+diff --git a/arch/x86/mm/pgtable_32-xen.c b/arch/x86/mm/pgtable_32-xen.c
+new file mode 100644
+index 0000000..7636e18
+--- /dev/null
++++ b/arch/x86/mm/pgtable_32-xen.c
+@@ -0,0 +1,185 @@
++#include <linux/sched.h>
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/nmi.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 <linux/quicklist.h>
++
++#include <asm/system.h>
++#include <asm/pgtable.h>
++#include <asm/pgalloc.h>
++#include <asm/fixmap.h>
++#include <asm/e820.h>
++#include <asm/tlb.h>
++#include <asm/tlbflush.h>
++#include <asm/io.h>
++#include <asm/mmu_context.h>
++
++#include <xen/features.h>
++#include <asm/hypervisor.h>
++
++void show_mem(void)
++{
++ int total = 0, reserved = 0;
++ int shared = 0, cached = 0;
++ int highmem = 0;
++ struct page *page;
++ pg_data_t *pgdat;
++ unsigned long i;
++ unsigned long flags;
++
++ printk(KERN_INFO "Mem-info:\n");
++ show_free_areas();
++ for_each_online_pgdat(pgdat) {
++ pgdat_resize_lock(pgdat, &flags);
++ for (i = 0; i < pgdat->node_spanned_pages; ++i) {
++ if (unlikely(i % MAX_ORDER_NR_PAGES == 0))
++ touch_nmi_watchdog();
++ 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);
++
++ 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));
++}
++
++/*
++ * 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)
++{
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++
++ 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);
++}
++
++static int fixmaps;
++unsigned long hypervisor_virt_start = HYPERVISOR_VIRT_START;
++unsigned long __FIXADDR_TOP = (HYPERVISOR_VIRT_START - PAGE_SIZE);
++EXPORT_SYMBOL(__FIXADDR_TOP);
++
++void __set_fixmap (enum fixed_addresses idx, maddr_t phys, pgprot_t flags)
++{
++ unsigned long address = __fix_to_virt(idx);
++ pte_t pte;
++
++ if (idx >= __end_of_fixed_addresses) {
++ BUG();
++ return;
++ }
++ switch (idx) {
++ case FIX_WP_TEST:
++ case FIX_VDSO:
++ pte = pfn_pte(phys >> PAGE_SHIFT, flags);
++ break;
++ default:
++ pte = pfn_pte_ma(phys >> PAGE_SHIFT, flags);
++ break;
++ }
++ if (HYPERVISOR_update_va_mapping(address, pte,
++ UVMF_INVLPG|UVMF_ALL))
++ BUG();
++ fixmaps++;
++}
++
++/**
++ * 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 __init reserve_top_address(unsigned long reserve)
++{
++ BUG_ON(fixmaps > 0);
++ printk(KERN_INFO "Reserving virtual address space above 0x%08x\n",
++ (int)-reserve);
++ __FIXADDR_TOP = -reserve - PAGE_SIZE;
++ __VMALLOC_RESERVE += reserve;
++}
++
++void make_lowmem_page_readonly(void *va, unsigned int feature)
++{
++ pte_t *pte;
++ unsigned int level;
++ int rc;
++
++ if (xen_feature(feature))
++ return;
++
++ pte = lookup_address((unsigned long)va, &level);
++ BUG_ON(!pte || level != PG_LEVEL_4K || !pte_present(*pte));
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_wrprotect(*pte), 0);
++ BUG_ON(rc);
++}
++
++void make_lowmem_page_writable(void *va, unsigned int feature)
++{
++ pte_t *pte;
++ unsigned int level;
++ int rc;
++
++ if (xen_feature(feature))
++ return;
++
++ pte = lookup_address((unsigned long)va, &level);
++ BUG_ON(!pte || level != PG_LEVEL_4K || !pte_present(*pte));
++ rc = HYPERVISOR_update_va_mapping(
++ (unsigned long)va, pte_mkwrite(*pte), 0);
++ BUG_ON(rc);
++}
+diff --git a/arch/x86/oprofile/Makefile b/arch/x86/oprofile/Makefile
+index 30f3eb3..8b3a60b 100644
+--- a/arch/x86/oprofile/Makefile
++++ b/arch/x86/oprofile/Makefile
+@@ -6,7 +6,14 @@ DRIVER_OBJS = $(addprefix ../../../drivers/oprofile/, \
+ oprofilefs.o oprofile_stats.o \
+ timer_int.o )
+
++ifdef CONFIG_XEN
++XENOPROF_COMMON_OBJS = $(addprefix ../../../drivers/xen/xenoprof/, \
++ xenoprofile.o)
++oprofile-y := $(DRIVER_OBJS) \
++ $(XENOPROF_COMMON_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 --git a/arch/x86/oprofile/xenoprof.c b/arch/x86/oprofile/xenoprof.c
+new file mode 100644
+index 0000000..cb6ae1e
+--- /dev/null
++++ b/arch/x86/oprofile/xenoprof.c
+@@ -0,0 +1,179 @@
++/**
++ * @file xenoprof.c
++ *
++ * @remark Copyright 2002 OProfile authors
++ * @remark Read the file COPYING
++ *
++ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon and Jose Renato Santos for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ *
++ * x86-specific part
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ */
++
++#include <linux/init.h>
++#include <linux/oprofile.h>
++#include <linux/sched.h>
++#include <asm/pgtable.h>
++
++#include <xen/driver_util.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/xenoprof.h>
++#include <xen/xenoprof.h>
++#include "op_counter.h"
++
++static unsigned int num_events = 0;
++
++void __init xenoprof_arch_init_counter(struct xenoprof_init *init)
++{
++ num_events = init->num_events;
++ /* 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;
++ init->num_events = num_events;
++ }
++}
++
++void xenoprof_arch_counter(void)
++{
++ int i;
++ struct xenoprof_counter counter;
++
++ 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;
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_counter,
++ &counter));
++ }
++}
++
++void xenoprof_arch_start(void)
++{
++ /* nothing */
++}
++
++void xenoprof_arch_stop(void)
++{
++ /* nothing */
++}
++
++void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer * sbuf)
++{
++ if (sbuf->buffer) {
++ vunmap(sbuf->buffer);
++ sbuf->buffer = NULL;
++ }
++}
++
++int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer * get_buffer,
++ struct xenoprof_shared_buffer * sbuf)
++{
++ int npages, ret;
++ struct vm_struct *area;
++
++ sbuf->buffer = NULL;
++ if ( (ret = HYPERVISOR_xenoprof_op(XENOPROF_get_buffer, get_buffer)) )
++ return ret;
++
++ npages = (get_buffer->bufsize * get_buffer->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_gmaddr >> PAGE_SHIFT,
++ npages * PAGE_SIZE, __pgprot(_KERNPG_TABLE),
++ DOMID_SELF)) ) {
++ vunmap(area->addr);
++ return ret;
++ }
++
++ sbuf->buffer = area->addr;
++ return ret;
++}
++
++int xenoprof_arch_set_passive(struct xenoprof_passive * pdomain,
++ struct xenoprof_shared_buffer * sbuf)
++{
++ int ret;
++ int npages;
++ struct vm_struct *area;
++ pgprot_t prot = __pgprot(_KERNPG_TABLE);
++
++ sbuf->buffer = NULL;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_passive, pdomain);
++ if (ret)
++ goto out;
++
++ npages = (pdomain->bufsize * pdomain->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,
++ pdomain->buf_gmaddr >> PAGE_SHIFT,
++ npages * PAGE_SIZE, prot, DOMID_SELF);
++ if (ret) {
++ vunmap(area->addr);
++ goto out;
++ }
++ sbuf->buffer = area->addr;
++
++out:
++ return ret;
++}
++
++struct op_counter_config counter_config[OP_MAX_COUNTER];
++
++int xenoprof_create_files(struct super_block * sb, struct dentry * root)
++{
++ unsigned int i;
++
++ 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);
++ }
++
++ return 0;
++}
++
++int __init oprofile_arch_init(struct oprofile_operations * ops)
++{
++ return xenoprofile_init(ops);
++}
++
++void oprofile_arch_exit(void)
++{
++ xenoprofile_exit();
++}
+diff --git a/arch/x86/pci/Makefile b/arch/x86/pci/Makefile
+index c5c8e48..785c043 100644
+--- a/arch/x86/pci/Makefile
++++ b/arch/x86/pci/Makefile
+@@ -3,3 +3,7 @@ include ${srctree}/arch/x86/pci/Makefile_32
+ else
+ include ${srctree}/arch/x86/pci/Makefile_64
+ endif
++
++# pcifront should be after mmconfig.o and direct.o as it should only
++# take over if direct access to the PCI bus is unavailable
++obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront.o
+diff --git a/arch/x86/pci/irq-xen.c b/arch/x86/pci/irq-xen.c
+new file mode 100644
+index 0000000..49e15f3
+--- /dev/null
++++ b/arch/x86/pci/irq-xen.c
+@@ -0,0 +1,1241 @@
++/*
++ * Low-Level PCI Support for PC -- Routing of Interrupts
++ *
++ * (c) 1999--2000 Martin Mares <mj at ucw.cz>
++ */
++
++#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>
++
++#include "pci.h"
++
++#define PIRQ_SIGNATURE (('$' << 0) + ('P' << 8) + ('I' << 16) + ('R' << 24))
++#define PIRQ_VERSION 0x0100
++
++static int broken_hp_bios_irq9;
++static int acer_tm360_irqrouting;
++
++static struct irq_routing_table *pirq_table;
++
++static int pirq_enable_irq(struct pci_dev *dev);
++
++/*
++ * 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;
++
++static int pirq_penalty[16] = {
++ 1000000, 1000000, 1000000, 1000, 1000, 0, 1000, 1000,
++ 0, 0, 0, 0, 1000, 100000, 100000, 100000
++};
++
++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);
++};
++
++struct irq_router_handler {
++ u16 vendor;
++ int (*probe)(struct irq_router *r, struct pci_dev *router, u16 device);
++};
++
++int (*pcibios_enable_irq)(struct pci_dev *dev) = NULL;
++void (*pcibios_disable_irq)(struct pci_dev *dev) = NULL;
++
++/*
++ * Check passed address for the PCI IRQ Routing Table signature
++ * and perform checksum verification.
++ */
++
++static inline struct irq_routing_table * pirq_check_routing_table(u8 *addr)
++{
++ struct irq_routing_table *rt;
++ int i;
++ u8 sum;
++
++ 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;
++ }
++ return NULL;
++}
++
++
++
++/*
++ * Search 0xf0000 -- 0xfffff for the PCI IRQ Routing Table.
++ */
++
++static struct irq_routing_table * __init pirq_find_routing_table(void)
++{
++ u8 *addr;
++ struct irq_routing_table *rt;
++
++#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 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)
++{
++ struct irq_routing_table *rt = pirq_table;
++ u8 busmap[256];
++ int i;
++ struct irq_info *e;
++
++ 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");
++ }
++#endif
++ busmap[e->bus] = 1;
++ }
++ for(i = 1; i < 256; i++) {
++ int node;
++ if (!busmap[i] || pci_find_bus(0, i))
++ continue;
++ node = get_mp_bus_to_node(i);
++ if (pci_scan_bus_on_node(i, &pci_root_ops, node))
++ printk(KERN_INFO "PCI: Discovered primary peer "
++ "bus %02x [IRQ]\n", i);
++ }
++ pcibios_last_bus = -1;
++}
++
++/*
++ * Code for querying and setting of IRQ routes on various interrupt routers.
++ */
++
++void eisa_set_level_irq(unsigned int irq)
++{
++ unsigned char mask = 1 << (irq & 7);
++ unsigned int port = 0x4d0 + (irq >> 3);
++ unsigned char val;
++ static u16 eisa_irq_mask;
++
++ if (irq >= 16 || (1 << irq) & eisa_irq_mask)
++ 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);
++ }
++}
++
++/*
++ * Common IRQ routing practice: nibbles in config space,
++ * offset by some magic constant.
++ */
++static unsigned int read_config_nybble(struct pci_dev *router, unsigned offset, unsigned nr)
++{
++ u8 x;
++ unsigned reg = offset + (nr >> 1);
++
++ pci_read_config_byte(router, reg, &x);
++ return (nr & 1) ? (x >> 4) : (x & 0xf);
++}
++
++static void write_config_nybble(struct pci_dev *router, unsigned offset, unsigned nr, unsigned int val)
++{
++ u8 x;
++ unsigned reg = offset + (nr >> 1);
++
++ pci_read_config_byte(router, reg, &x);
++ x = (nr & 1) ? ((x & 0x0f) | (val << 4)) : ((x & 0xf0) | val);
++ pci_write_config_byte(router, reg, x);
++}
++
++/*
++ * 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 const unsigned char irqmap[16] = { 0, 9, 3, 10, 4, 5, 7, 6, 1, 11, 0, 12, 0, 14, 0, 15 };
++
++ WARN_ON_ONCE(pirq > 16);
++ return irqmap[read_config_nybble(router, 0x48, pirq-1)];
++}
++
++static int pirq_ali_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ 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];
++
++ WARN_ON_ONCE(pirq > 16);
++ if (val) {
++ write_config_nybble(router, 0x48, pirq-1, val);
++ return 1;
++ }
++ return 0;
++}
++
++/*
++ * 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;
++
++ pci_read_config_byte(router, pirq, &x);
++ return (x < 16) ? x : 0;
++}
++
++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;
++}
++
++/*
++ * 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)
++{
++ return read_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq);
++}
++
++static int pirq_via_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ write_config_nybble(router, 0x55, pirq == 4 ? 5 : pirq, irq);
++ return 1;
++}
++
++/*
++ * 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 const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
++
++ WARN_ON_ONCE(pirq > 5);
++ return read_config_nybble(router, 0x55, pirqmap[pirq-1]);
++}
++
++static int pirq_via586_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 };
++
++ WARN_ON_ONCE(pirq > 5);
++ write_config_nybble(router, 0x55, pirqmap[pirq-1], irq);
++ return 1;
++}
++
++/*
++ * 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)
++{
++ static const unsigned char pirqmap[4] = { 1, 0, 2, 3 };
++
++ WARN_ON_ONCE(pirq > 4);
++ return read_config_nybble(router,0x43, pirqmap[pirq-1]);
++}
++
++static int pirq_ite_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ static const unsigned char pirqmap[4] = { 1, 0, 2, 3 };
++
++ WARN_ON_ONCE(pirq > 4);
++ write_config_nybble(router, 0x43, pirqmap[pirq-1], irq);
++ return 1;
++}
++
++/*
++ * 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);
++}
++
++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;
++}
++
++/*
++ * 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)
++{
++ return read_config_nybble(router, 0x5C, (pirq-1)^1);
++}
++
++static int pirq_cyrix_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ write_config_nybble(router, 0x5C, (pirq-1)^1, irq);
++ return 1;
++}
++
++/*
++ * 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.
++ *
++ * 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.
++ *
++ * Onchip routing for router rev-id 0x04 (try-and-error observation)
++ *
++ * 0x60/0x61/0x62/0x63: 1xEHCI and 3xOHCI (companion) USB-HCs
++ * bit 6-4 are probably unused, not like 5595
++ */
++
++#define PIRQ_SIS_IRQ_MASK 0x0f
++#define PIRQ_SIS_IRQ_DISABLE 0x80
++#define PIRQ_SIS_USB_ENABLE 0x40
++
++static int pirq_sis_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++{
++ u8 x;
++ int reg;
++
++ 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 pirq_sis_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ u8 x;
++ int reg;
++
++ 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;
++}
++
++
++/*
++ * 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 int pirq_vlsi_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++{
++ WARN_ON_ONCE(pirq >= 9);
++ if (pirq > 8) {
++ printk(KERN_INFO "VLSI router pirq escape (%d)\n", pirq);
++ return 0;
++ }
++ return read_config_nybble(router, 0x74, pirq-1);
++}
++
++static int pirq_vlsi_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ WARN_ON_ONCE(pirq >= 9);
++ 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;
++}
++
++/*
++ * 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)
++{
++ outb(pirq, 0xc00);
++ return inb(0xc01) & 0xf;
++}
++
++static int pirq_serverworks_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ outb(pirq, 0xc00);
++ outb(irq, 0xc01);
++ return 1;
++}
++
++/* 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)
++{
++ u8 irq;
++ irq = 0;
++ if (pirq <= 4)
++ {
++ irq = read_config_nybble(router, 0x56, pirq - 1);
++ }
++ printk(KERN_INFO "AMD756: dev %04x:%04x, router pirq : %d get irq : %2d\n",
++ dev->vendor, dev->device, pirq, irq);
++ return irq;
++}
++
++static int pirq_amd756_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ 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);
++ }
++ return 1;
++}
++
++/*
++ * PicoPower PT86C523
++ */
++static int pirq_pico_get(struct pci_dev *router, struct pci_dev *dev, int pirq)
++{
++ outb(0x10 + ((pirq - 1) >> 1), 0x24);
++ return ((pirq - 1) & 1) ? (inb(0x26) >> 4) : (inb(0x26) & 0xf);
++}
++
++static int pirq_pico_set(struct pci_dev *router, struct pci_dev *dev, int pirq,
++ int irq)
++{
++ unsigned int x;
++ outb(0x10 + ((pirq - 1) >> 1), 0x24);
++ x = inb(0x26);
++ x = ((pirq - 1) & 1) ? ((x & 0x0f) | (irq << 4)) : ((x & 0xf0) | (irq));
++ outb(x, 0x26);
++ return 1;
++}
++
++#ifdef CONFIG_PCI_BIOS
++
++static int pirq_bios_set(struct pci_dev *router, struct pci_dev *dev, int pirq, int irq)
++{
++ struct pci_dev *bridge;
++ int pin = pci_get_interrupt_pin(dev, &bridge);
++ return pcibios_set_irq_routing(bridge, pin, irq);
++}
++
++#endif
++
++static __init int intel_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++{
++ 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) },
++ { },
++ };
++
++ /* 440GX has a proprietary PIRQ router -- don't use it */
++ if (pci_dev_present(pirq_440gx))
++ 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:
++ case PCI_DEVICE_ID_INTEL_ICH9_0:
++ case PCI_DEVICE_ID_INTEL_ICH9_1:
++ case PCI_DEVICE_ID_INTEL_ICH9_2:
++ case PCI_DEVICE_ID_INTEL_ICH9_3:
++ case PCI_DEVICE_ID_INTEL_ICH9_4:
++ case PCI_DEVICE_ID_INTEL_ICH9_5:
++ case PCI_DEVICE_ID_INTEL_TOLAPAI_0:
++ case PCI_DEVICE_ID_INTEL_ICH10_0:
++ case PCI_DEVICE_ID_INTEL_ICH10_1:
++ case PCI_DEVICE_ID_INTEL_ICH10_2:
++ case PCI_DEVICE_ID_INTEL_ICH10_3:
++ r->name = "PIIX/ICH";
++ r->get = pirq_piix_get;
++ r->set = pirq_piix_set;
++ return 1;
++ }
++ return 0;
++}
++
++static __init int via_router_probe(struct irq_router *r,
++ struct pci_dev *router, u16 device)
++{
++ /* FIXME: We should move some of the quirk fixup stuff here */
++
++ /*
++ * workarounds 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;
++ case PCI_DEVICE_ID_VIA_8237:
++ /**
++ * Asus a7v600 bios wrongly reports 8237
++ * as 586-compatible
++ */
++ device = PCI_DEVICE_ID_VIA_8237;
++ break;
++ }
++ }
++
++ 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;
++ }
++ 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;
++ }
++ return 0;
++}
++
++
++static __init int serverworks_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++{
++ 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;
++}
++
++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;
++}
++
++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;
++ }
++ 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;
++}
++
++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)
++{
++ 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;
++}
++
++static __init int amd_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++{
++ 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;
++}
++
++static __init int pico_router_probe(struct irq_router *r, struct pci_dev *router, u16 device)
++{
++ switch (device) {
++ case PCI_DEVICE_ID_PICOPOWER_PT86C523:
++ r->name = "PicoPower PT86C523";
++ r->get = pirq_pico_get;
++ r->set = pirq_pico_set;
++ return 1;
++
++ case PCI_DEVICE_ID_PICOPOWER_PT86C523BBP:
++ r->name = "PicoPower PT86C523 rev. BB+";
++ r->get = pirq_pico_get;
++ r->set = pirq_pico_set;
++ return 1;
++ }
++ return 0;
++}
++
++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 },
++ { PCI_VENDOR_ID_PICOPOWER, pico_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)
++{
++ struct irq_routing_table *rt = pirq_table;
++ struct irq_router_handler *h;
++
++#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
++
++ /* 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);
++
++ pirq_router_dev = pci_get_bus_and_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;
++ }
++
++ 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;
++ }
++ 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));
++
++ /* The device remains referenced for the kernel lifetime */
++}
++
++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;
++}
++
++static int pcibios_lookup_irq(struct pci_dev *dev, int assign)
++{
++ 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;
++
++ /* Find IRQ routing entry */
++
++ 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;
++
++ /* Work around broken HP Pavilion Notebooks which assign USB to
++ IRQ 9 even though it is actually wired to IRQ 11 */
++
++ 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);
++ }
++
++ /* 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);
++ }
++
++ /*
++ * 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;
++ }
++ }
++ 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;
++ }
++ }
++
++ 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));
++
++ /* 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));
++ }
++ }
++ 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) {
++ 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(const 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);
++ }
++ 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(const 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);
++ }
++ return 0;
++}
++
++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"),
++ },
++ },
++ { }
++};
++
++static int __init pcibios_irq_init(void)
++{
++ 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();
++
++#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;
++ }
++
++ pcibios_enable_irq = pirq_enable_irq;
++
++ pcibios_fixup_irqs();
++ return 0;
++}
++
++subsys_initcall(pcibios_irq_init);
++
++
++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);
++}
++
++static int pirq_enable_irq(struct pci_dev *dev)
++{
++ u8 pin;
++ struct pci_dev *temp_dev;
++
++ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin);
++ if (pin && !pcibios_lookup_irq(dev, 1) && !dev->irq) {
++ char *msg = "";
++
++ pin--; /* interrupt pins are numbered starting from 1 */
++
++ if (io_apic_assign_pci_irqs) {
++ int irq;
++
++ 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) {
++ 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.";
++
++ /* 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;
++
++ printk(KERN_WARNING "PCI: No IRQ known for interrupt pin %c of device %s.%s\n",
++ 'A' + pin, pci_name(dev), msg);
++ }
++ return 0;
++}
+diff --git a/arch/x86/pci/pcifront.c b/arch/x86/pci/pcifront.c
+new file mode 100644
+index 0000000..7009115
+--- /dev/null
++++ b/arch/x86/pci/pcifront.c
+@@ -0,0 +1,55 @@
++/*
++ * PCI Frontend Stub - puts some "dummy" functions in to the Linux x86 PCI core
++ * to support the Xen PCI Frontend's operation
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <asm/acpi.h>
++#include "pci.h"
++
++static int pcifront_enable_irq(struct pci_dev *dev)
++{
++ u8 irq;
++ pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
++ dev->irq = irq;
++
++ return 0;
++}
++
++extern u8 pci_cache_line_size;
++
++static int __init pcifront_x86_stub_init(void)
++{
++ struct cpuinfo_x86 *c = &boot_cpu_data;
++
++ /* Only install our method if we haven't found real hardware already */
++ if (raw_pci_ops)
++ return 0;
++
++ printk(KERN_INFO "PCI: setting up Xen PCI frontend stub\n");
++
++ /* Copied from arch/i386/pci/common.c */
++ pci_cache_line_size = 32 >> 2;
++ if (c->x86 >= 6 && c->x86_vendor == X86_VENDOR_AMD)
++ pci_cache_line_size = 64 >> 2; /* K7 & K8 */
++ else if (c->x86 > 6 && c->x86_vendor == X86_VENDOR_INTEL)
++ pci_cache_line_size = 128 >> 2; /* P4 */
++
++ /* On x86, we need to disable the normal IRQ routing table and
++ * just ask the backend
++ */
++ pcibios_enable_irq = pcifront_enable_irq;
++ pcibios_disable_irq = NULL;
++
++#ifdef CONFIG_ACPI
++ /* Keep ACPI out of the picture */
++ acpi_noirq = 1;
++#endif
++
++ return 0;
++}
++
++arch_initcall(pcifront_x86_stub_init);
+diff --git a/arch/x86/power/Makefile b/arch/x86/power/Makefile
+index 9ff4d5b..c12a175 100644
+--- a/arch/x86/power/Makefile
++++ b/arch/x86/power/Makefile
+@@ -1,2 +1,4 @@
+ obj-$(CONFIG_PM_SLEEP) += cpu_$(BITS).o
+ obj-$(CONFIG_HIBERNATION) += hibernate_$(BITS).o hibernate_asm_$(BITS).o
++
++disabled-obj-$(CONFIG_XEN) := cpu_$(BITS).o
+diff --git a/arch/x86/power/cpu_32.c b/arch/x86/power/cpu_32.c
+index 7dc5d5c..c1d8769 100644
+--- a/arch/x86/power/cpu_32.c
++++ b/arch/x86/power/cpu_32.c
+@@ -65,6 +65,7 @@ static void do_fpu_end(void)
+
+ static void fix_processor_context(void)
+ {
++#ifndef CONFIG_X86_NO_TSS
+ int cpu = smp_processor_id();
+ struct tss_struct *t = &per_cpu(init_tss, cpu);
+
+@@ -74,6 +75,7 @@ static void fix_processor_context(void)
+ * 386 hardware has concept of busy TSS or some
+ * similar stupidity.
+ */
++#endif
+
+ load_TR_desc(); /* This does ltr */
+ load_LDT(¤t->active_mm->context); /* This does lldt */
+diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile
+index b7ad9f8..005214e 100644
+--- a/arch/x86/vdso/Makefile
++++ b/arch/x86/vdso/Makefile
+@@ -65,6 +65,9 @@ obj-$(VDSO32-y) += vdso32-syms.lds
+ vdso32.so-$(CONFIG_X86_32) += int80
+ vdso32.so-$(CONFIG_COMPAT) += syscall
+ vdso32.so-$(VDSO32-y) += sysenter
++xen-vdso32-$(subst 1,$(CONFIG_COMPAT),$(shell expr $(CONFIG_XEN_COMPAT)0 '<' 0x0302000)) += int80
++xen-vdso32-$(CONFIG_X86_32) += syscall
++vdso32.so-$(CONFIG_XEN) += $(xen-vdso32-y)
+
+ vdso32-images = $(vdso32.so-y:%=vdso32-%.so)
+
+diff --git a/arch/x86/vdso/vdso32-setup-xen.c b/arch/x86/vdso/vdso32-setup-xen.c
+new file mode 100644
+index 0000000..2a5fb16
+--- /dev/null
++++ b/arch/x86/vdso/vdso32-setup-xen.c
+@@ -0,0 +1,514 @@
++/*
++ * (C) Copyright 2002 Linus Torvalds
++ * Portions based on the vdso-randomization code from exec-shield:
++ * Copyright(C) 2005-2006, Red Hat, Inc., Ingo Molnar
++ *
++ * This file contains the needed initializations to support sysenter.
++ */
++
++#include <linux/init.h>
++#include <linux/smp.h>
++#include <linux/thread_info.h>
++#include <linux/sched.h>
++#include <linux/gfp.h>
++#include <linux/string.h>
++#include <linux/elf.h>
++#include <linux/mm.h>
++#include <linux/err.h>
++#include <linux/module.h>
++
++#include <asm/cpufeature.h>
++#include <asm/msr.h>
++#include <asm/pgtable.h>
++#include <asm/unistd.h>
++#include <asm/elf.h>
++#include <asm/tlbflush.h>
++#include <asm/vdso.h>
++#include <asm/proto.h>
++
++#include <xen/interface/callback.h>
++
++enum {
++ VDSO_DISABLED = 0,
++ VDSO_ENABLED = 1,
++ VDSO_COMPAT = 2,
++};
++
++#ifdef CONFIG_COMPAT_VDSO
++#define VDSO_DEFAULT VDSO_COMPAT
++#else
++#define VDSO_DEFAULT VDSO_ENABLED
++#endif
++
++#ifdef CONFIG_X86_64
++#define vdso_enabled sysctl_vsyscall32
++#define arch_setup_additional_pages syscall32_setup_pages
++#endif
++
++/*
++ * This is the difference between the prelinked addresses in the vDSO images
++ * and the VDSO_HIGH_BASE address where CONFIG_COMPAT_VDSO places the vDSO
++ * in the user address space.
++ */
++#define VDSO_ADDR_ADJUST (VDSO_HIGH_BASE - (unsigned long)VDSO32_PRELINK)
++
++/*
++ * Should the kernel map a VDSO page into processes and pass its
++ * address down to glibc upon exec()?
++ */
++unsigned int __read_mostly vdso_enabled = VDSO_DEFAULT;
++
++static int __init vdso_setup(char *s)
++{
++ vdso_enabled = simple_strtoul(s, NULL, 0);
++
++ return 1;
++}
++
++/*
++ * For consistency, the argument vdso32=[012] affects the 32-bit vDSO
++ * behavior on both 64-bit and 32-bit kernels.
++ * On 32-bit kernels, vdso=[012] means the same thing.
++ */
++__setup("vdso32=", vdso_setup);
++
++#ifdef CONFIG_X86_32
++__setup_param("vdso=", vdso32_setup, vdso_setup, 0);
++
++EXPORT_SYMBOL_GPL(vdso_enabled);
++#endif
++
++static __init void reloc_symtab(Elf32_Ehdr *ehdr,
++ unsigned offset, unsigned size)
++{
++ Elf32_Sym *sym = (void *)ehdr + offset;
++ unsigned nsym = size / sizeof(*sym);
++ unsigned i;
++
++ for(i = 0; i < nsym; i++, sym++) {
++ if (sym->st_shndx == SHN_UNDEF ||
++ sym->st_shndx == SHN_ABS)
++ continue; /* skip */
++
++ if (sym->st_shndx > SHN_LORESERVE) {
++ printk(KERN_INFO "VDSO: unexpected st_shndx %x\n",
++ sym->st_shndx);
++ continue;
++ }
++
++ switch(ELF_ST_TYPE(sym->st_info)) {
++ case STT_OBJECT:
++ case STT_FUNC:
++ case STT_SECTION:
++ case STT_FILE:
++ sym->st_value += VDSO_ADDR_ADJUST;
++ }
++ }
++}
++
++static __init void reloc_dyn(Elf32_Ehdr *ehdr, unsigned offset)
++{
++ Elf32_Dyn *dyn = (void *)ehdr + offset;
++
++ for(; dyn->d_tag != DT_NULL; dyn++)
++ switch(dyn->d_tag) {
++ case DT_PLTGOT:
++ case DT_HASH:
++ case DT_STRTAB:
++ case DT_SYMTAB:
++ case DT_RELA:
++ case DT_INIT:
++ case DT_FINI:
++ case DT_REL:
++ case DT_DEBUG:
++ case DT_JMPREL:
++ case DT_VERSYM:
++ case DT_VERDEF:
++ case DT_VERNEED:
++ case DT_ADDRRNGLO ... DT_ADDRRNGHI:
++ /* definitely pointers needing relocation */
++ dyn->d_un.d_ptr += VDSO_ADDR_ADJUST;
++ break;
++
++ case DT_ENCODING ... OLD_DT_LOOS-1:
++ case DT_LOOS ... DT_HIOS-1:
++ /* Tags above DT_ENCODING are pointers if
++ they're even */
++ if (dyn->d_tag >= DT_ENCODING &&
++ (dyn->d_tag & 1) == 0)
++ dyn->d_un.d_ptr += VDSO_ADDR_ADJUST;
++ break;
++
++ case DT_VERDEFNUM:
++ case DT_VERNEEDNUM:
++ case DT_FLAGS_1:
++ case DT_RELACOUNT:
++ case DT_RELCOUNT:
++ case DT_VALRNGLO ... DT_VALRNGHI:
++ /* definitely not pointers */
++ break;
++
++ case OLD_DT_LOOS ... DT_LOOS-1:
++ case DT_HIOS ... DT_VALRNGLO-1:
++ default:
++ if (dyn->d_tag > DT_ENCODING)
++ printk(KERN_INFO "VDSO: unexpected DT_tag %x\n",
++ dyn->d_tag);
++ break;
++ }
++}
++
++static __init void relocate_vdso(Elf32_Ehdr *ehdr)
++{
++ Elf32_Phdr *phdr;
++ Elf32_Shdr *shdr;
++ int i;
++
++ BUG_ON(memcmp(ehdr->e_ident, ELFMAG, SELFMAG) != 0 ||
++ !elf_check_arch_ia32(ehdr) ||
++ ehdr->e_type != ET_DYN);
++
++ ehdr->e_entry += VDSO_ADDR_ADJUST;
++
++ /* rebase phdrs */
++ phdr = (void *)ehdr + ehdr->e_phoff;
++ for (i = 0; i < ehdr->e_phnum; i++) {
++ phdr[i].p_vaddr += VDSO_ADDR_ADJUST;
++
++ /* relocate dynamic stuff */
++ if (phdr[i].p_type == PT_DYNAMIC)
++ reloc_dyn(ehdr, phdr[i].p_offset);
++ }
++
++ /* rebase sections */
++ shdr = (void *)ehdr + ehdr->e_shoff;
++ for(i = 0; i < ehdr->e_shnum; i++) {
++ if (!(shdr[i].sh_flags & SHF_ALLOC))
++ continue;
++
++ shdr[i].sh_addr += VDSO_ADDR_ADJUST;
++
++ if (shdr[i].sh_type == SHT_SYMTAB ||
++ shdr[i].sh_type == SHT_DYNSYM)
++ reloc_symtab(ehdr, shdr[i].sh_offset,
++ shdr[i].sh_size);
++ }
++}
++
++/*
++ * These symbols are defined by vdso32.S to mark the bounds
++ * of the ELF DSO images included therein.
++ */
++extern const char vdso32_default_start, vdso32_default_end;
++extern const char vdso32_sysenter_start, vdso32_sysenter_end;
++static struct page *vdso32_pages[1];
++
++
++#ifdef CONFIG_X86_64
++
++#if CONFIG_XEN_COMPAT < 0x030200
++static int use_int80 = 1;
++#endif
++static int use_sysenter __read_mostly = -1;
++
++#define vdso32_sysenter() (use_sysenter > 0)
++
++/* May not be __init: called during resume */
++void syscall32_cpu_init(void)
++{
++#ifdef CONFIG_XEN
++ static const struct callback_register cstar = {
++ .type = CALLBACKTYPE_syscall32,
++ .address = (unsigned long)ia32_cstar_target
++ };
++ static const struct callback_register sysenter = {
++ .type = CALLBACKTYPE_sysenter,
++ .address = (unsigned long)ia32_sysenter_target
++ };
++
++ if ((HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) < 0) ||
++ (HYPERVISOR_callback_op(CALLBACKOP_register, &cstar) < 0))
++#if CONFIG_XEN_COMPAT < 0x030200
++ return;
++ use_int80 = 0;
++#else
++ BUG();
++#endif
++#endif
++
++ if (use_sysenter < 0) {
++ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
++ use_sysenter = 1;
++ if (boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR)
++ use_sysenter = 1;
++ }
++}
++
++#define compat_uses_vma 1
++
++static inline void map_compat_vdso(int map)
++{
++}
++
++#else /* CONFIG_X86_32 */
++
++#define vdso32_sysenter() (boot_cpu_has(X86_FEATURE_SEP))
++
++extern asmlinkage void ia32pv_cstar_target(void);
++static const struct callback_register cstar = {
++ .type = CALLBACKTYPE_syscall32,
++ .address = { __KERNEL_CS, (unsigned long)ia32pv_cstar_target },
++};
++
++void __cpuinit enable_sep_cpu(void)
++{
++ extern asmlinkage void ia32pv_sysenter_target(void);
++ static struct callback_register __cpuinitdata sysenter = {
++ .type = CALLBACKTYPE_sysenter,
++ .address = { __KERNEL_CS, (unsigned long)ia32pv_sysenter_target },
++ };
++
++ if (boot_cpu_has(X86_FEATURE_SYSCALL)) {
++ if (HYPERVISOR_callback_op(CALLBACKOP_register, &cstar) != 0)
++ BUG();
++ return;
++ }
++
++ if (!boot_cpu_has(X86_FEATURE_SEP))
++ return;
++
++ if (xen_feature(XENFEAT_supervisor_mode_kernel))
++ sysenter.address.eip = (unsigned long)ia32_sysenter_target;
++
++ switch (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter)) {
++ case 0:
++ break;
++#if CONFIG_XEN_COMPAT < 0x030200
++ case -ENOSYS:
++ sysenter.type = CALLBACKTYPE_sysenter_deprecated;
++ if (HYPERVISOR_callback_op(CALLBACKOP_register, &sysenter) == 0)
++ break;
++#endif
++ default:
++ setup_clear_cpu_cap(X86_FEATURE_SEP);
++ break;
++ }
++}
++
++static struct vm_area_struct gate_vma;
++
++static int __init gate_vma_init(void)
++{
++ gate_vma.vm_mm = NULL;
++ gate_vma.vm_start = FIXADDR_USER_START;
++ gate_vma.vm_end = FIXADDR_USER_END;
++ gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC;
++ gate_vma.vm_page_prot = __P101;
++ /*
++ * Make sure the vDSO gets into every core dump.
++ * Dumping its contents makes post-mortem fully interpretable later
++ * without matching up the same kernel and hardware config to see
++ * what PC values meant.
++ */
++ gate_vma.vm_flags |= VM_ALWAYSDUMP;
++ return 0;
++}
++
++#define compat_uses_vma 0
++
++static void map_compat_vdso(int map)
++{
++ static int vdso_mapped;
++
++ if (map == vdso_mapped)
++ return;
++
++ vdso_mapped = map;
++
++ __set_fixmap(FIX_VDSO, page_to_pfn(vdso32_pages[0]) << PAGE_SHIFT,
++ map ? PAGE_READONLY_EXEC : PAGE_NONE);
++
++ /* flush stray tlbs */
++ flush_tlb_all();
++}
++
++#endif /* CONFIG_X86_64 */
++
++int __init sysenter_setup(void)
++{
++ void *syscall_page = (void *)get_zeroed_page(GFP_ATOMIC);
++ const void *vsyscall;
++ size_t vsyscall_len;
++
++ vdso32_pages[0] = virt_to_page(syscall_page);
++
++#ifdef CONFIG_X86_32
++ gate_vma_init();
++#endif
++
++#if defined(CONFIG_X86_64) && CONFIG_XEN_COMPAT < 0x030200
++ if (use_int80) {
++ extern const char vdso32_int80_start, vdso32_int80_end;
++
++ vsyscall = &vdso32_int80_start;
++ vsyscall_len = &vdso32_int80_end - &vdso32_int80_start;
++ } else
++#elif defined(CONFIG_X86_32)
++ if (boot_cpu_has(X86_FEATURE_SYSCALL)
++ && (boot_cpu_data.x86_vendor != X86_VENDOR_AMD
++ || HYPERVISOR_callback_op(CALLBACKOP_register, &cstar) != 0))
++ setup_clear_cpu_cap(X86_FEATURE_SYSCALL);
++ barrier(); /* until clear_bit()'s constraints are correct ... */
++ if (boot_cpu_has(X86_FEATURE_SYSCALL)) {
++ extern const char vdso32_syscall_start, vdso32_syscall_end;
++
++ vsyscall = &vdso32_syscall_start;
++ vsyscall_len = &vdso32_syscall_end - &vdso32_syscall_start;
++ } else
++#endif
++ if (!vdso32_sysenter()) {
++ vsyscall = &vdso32_default_start;
++ vsyscall_len = &vdso32_default_end - &vdso32_default_start;
++ } else {
++ vsyscall = &vdso32_sysenter_start;
++ vsyscall_len = &vdso32_sysenter_end - &vdso32_sysenter_start;
++ }
++
++ memcpy(syscall_page, vsyscall, vsyscall_len);
++ relocate_vdso(syscall_page);
++
++ return 0;
++}
++
++/* Setup a VMA at program startup for the vsyscall page */
++int arch_setup_additional_pages(struct linux_binprm *bprm, int exstack)
++{
++ struct mm_struct *mm = current->mm;
++ unsigned long addr;
++ int ret = 0;
++ bool compat;
++
++ if (vdso_enabled == VDSO_DISABLED)
++ return 0;
++
++ down_write(&mm->mmap_sem);
++
++ /* Test compat mode once here, in case someone
++ changes it via sysctl */
++ compat = (vdso_enabled == VDSO_COMPAT);
++
++ map_compat_vdso(compat);
++
++ if (compat)
++ addr = VDSO_HIGH_BASE;
++ else {
++ addr = get_unmapped_area(NULL, 0, PAGE_SIZE, 0, 0);
++ if (IS_ERR_VALUE(addr)) {
++ ret = addr;
++ goto up_fail;
++ }
++ }
++
++ if (compat_uses_vma || !compat) {
++ /*
++ * MAYWRITE to allow gdb to COW and set breakpoints
++ *
++ * Make sure the vDSO gets into every core dump.
++ * Dumping its contents makes post-mortem fully
++ * interpretable later without matching up the same
++ * kernel and hardware config to see what PC values
++ * meant.
++ */
++ ret = install_special_mapping(mm, addr, PAGE_SIZE,
++ VM_READ|VM_EXEC|
++ VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
++ VM_ALWAYSDUMP,
++ vdso32_pages);
++
++ if (ret)
++ goto up_fail;
++ }
++
++ current->mm->context.vdso = (void *)addr;
++ current_thread_info()->sysenter_return =
++ VDSO32_SYMBOL(addr, SYSENTER_RETURN);
++
++ up_fail:
++ up_write(&mm->mmap_sem);
++
++ return ret;
++}
++
++#ifdef CONFIG_X86_64
++
++/*
++ * This must be done early in case we have an initrd containing 32-bit
++ * binaries (e.g., hotplug). This could be pushed upstream.
++ */
++core_initcall(sysenter_setup);
++
++#ifdef CONFIG_SYSCTL
++/* Register vsyscall32 into the ABI table */
++#include <linux/sysctl.h>
++
++static ctl_table abi_table2[] = {
++ {
++ .procname = "vsyscall32",
++ .data = &sysctl_vsyscall32,
++ .maxlen = sizeof(int),
++ .mode = 0644,
++ .proc_handler = proc_dointvec
++ },
++ {}
++};
++
++static ctl_table abi_root_table2[] = {
++ {
++ .ctl_name = CTL_ABI,
++ .procname = "abi",
++ .mode = 0555,
++ .child = abi_table2
++ },
++ {}
++};
++
++static __init int ia32_binfmt_init(void)
++{
++ register_sysctl_table(abi_root_table2);
++ return 0;
++}
++__initcall(ia32_binfmt_init);
++#endif
++
++#else /* CONFIG_X86_32 */
++
++const char *arch_vma_name(struct vm_area_struct *vma)
++{
++ if (vma->vm_mm && vma->vm_start == (long)vma->vm_mm->context.vdso)
++ return "[vdso]";
++ return NULL;
++}
++
++struct vm_area_struct *get_gate_vma(struct task_struct *tsk)
++{
++ struct mm_struct *mm = tsk->mm;
++
++ /* Check to see if this task was created in compat vdso mode */
++ if (mm && mm->context.vdso == (void *)VDSO_HIGH_BASE)
++ return &gate_vma;
++ return NULL;
++}
++
++int in_gate_area(struct task_struct *task, unsigned long addr)
++{
++ const struct vm_area_struct *vma = get_gate_vma(task);
++
++ return vma && addr >= vma->vm_start && addr < vma->vm_end;
++}
++
++int in_gate_area_no_task(unsigned long addr)
++{
++ return 0;
++}
++
++#endif /* CONFIG_X86_64 */
+diff --git a/arch/x86/vdso/vdso32.S b/arch/x86/vdso/vdso32.S
+index 1e36f72..8719357 100644
+--- a/arch/x86/vdso/vdso32.S
++++ b/arch/x86/vdso/vdso32.S
+@@ -16,4 +16,16 @@ vdso32_sysenter_start:
+ .incbin "arch/x86/vdso/vdso32-sysenter.so"
+ vdso32_sysenter_end:
+
++#if defined(CONFIG_X86_64_XEN) && CONFIG_XEN_COMPAT < 0x030200
++ .globl vdso32_int80_start, vdso32_int80_end
++vdso32_int80_start:
++ .incbin "arch/x86/vdso/vdso32-int80.so"
++vdso32_int80_end:
++#elif defined(CONFIG_X86_XEN)
++ .globl vdso32_syscall_start, vdso32_syscall_end
++vdso32_syscall_start:
++ .incbin "arch/x86/vdso/vdso32-syscall.so"
++vdso32_syscall_end:
++#endif
++
+ __FINIT
+diff --git a/arch/x86/vdso/vdso32/note.S b/arch/x86/vdso/vdso32/note.S
+index c83f257..b6ed8cd 100644
+--- a/arch/x86/vdso/vdso32/note.S
++++ b/arch/x86/vdso/vdso32/note.S
+@@ -13,7 +13,7 @@ ELFNOTE_START(Linux, 0, "a")
+ .long LINUX_VERSION_CODE
+ ELFNOTE_END
+
+-#ifdef CONFIG_XEN
++#if defined(CONFIG_X86_XEN) || defined(CONFIG_PARAVIRT_XEN)
+ /*
+ * 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
+@@ -37,8 +37,12 @@ ELFNOTE_END
+
+ ELFNOTE_START(GNU, 2, "a")
+ .long 1 /* ncaps */
++#ifdef CONFIG_PARAVIRT_XEN
+ VDSO32_NOTE_MASK: /* Symbol used by arch/x86/xen/setup.c */
+ .long 0 /* mask */
++#else
++ .long 1 << VDSO_NOTE_NONEGSEG_BIT /* mask */
++#endif
+ .byte VDSO_NOTE_NONEGSEG_BIT; .asciz "nosegneg" /* bit, name */
+ ELFNOTE_END
+ #endif
+diff --git a/arch/x86/vdso/vdso32/syscall.S b/arch/x86/vdso/vdso32/syscall.S
+index 5415b56..0a27d17 100644
+--- a/arch/x86/vdso/vdso32/syscall.S
++++ b/arch/x86/vdso/vdso32/syscall.S
+@@ -19,8 +19,10 @@ __kernel_vsyscall:
+ .Lpush_ebp:
+ movl %ecx, %ebp
+ syscall
++#ifndef CONFIG_XEN
+ movl $__USER32_DS, %ecx
+ movl %ecx, %ss
++#endif
+ movl %ebp, %ecx
+ popl %ebp
+ .Lpop_ebp:
+diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig
+index 6c388e5..9c73619 100644
+--- a/arch/x86/xen/Kconfig
++++ b/arch/x86/xen/Kconfig
+@@ -2,7 +2,7 @@
+ # This Kconfig describes xen options
+ #
+
+-config XEN
++config PARAVIRT_XEN
+ bool "Xen guest support"
+ select PARAVIRT
+ select PARAVIRT_CLOCK
+diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S
+index 6ec3b4f..0ada2b2 100644
+--- a/arch/x86/xen/xen-head.S
++++ b/arch/x86/xen/xen-head.S
+@@ -1,7 +1,7 @@
+ /* Xen-specific pieces of head.S, intended to be included in the right
+ place in head.S */
+
+-#ifdef CONFIG_XEN
++#ifdef CONFIG_PARAVIRT_XEN
+
+ #include <linux/elfnote.h>
+ #include <linux/init.h>
+@@ -33,4 +33,4 @@ ENTRY(hypercall_page)
+ ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE, .asciz "yes")
+ ELFNOTE(Xen, XEN_ELFNOTE_LOADER, .asciz "generic")
+
+-#endif /*CONFIG_XEN */
++#endif /* CONFIG_PARAVIRT_XEN */
+diff --git a/drivers/Makefile b/drivers/Makefile
+index f65deda..d28a2ae 100644
+--- a/drivers/Makefile
++++ b/drivers/Makefile
+@@ -16,7 +16,7 @@ obj-$(CONFIG_ACPI) += acpi/
+ obj-$(CONFIG_PNP) += pnp/
+ obj-$(CONFIG_ARM_AMBA) += amba/
+
+-obj-$(CONFIG_XEN) += xen/
++obj-$(CONFIG_PARAVIRT_XEN) += xen/
+
+ # char/ comes before serial/ etc so that the VT console is the boot-time
+ # default.
+@@ -34,6 +34,7 @@ obj-y += base/ block/ misc/ mfd/ net/ media/
+ obj-$(CONFIG_NUBUS) += nubus/
+ obj-$(CONFIG_ATM) += atm/
+ obj-y += macintosh/
++obj-$(CONFIG_XEN) += xen/
+ obj-$(CONFIG_IDE) += ide/
+ obj-$(CONFIG_SCSI) += scsi/
+ obj-$(CONFIG_ATA) += ata/
+diff --git a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
+index c52fca8..ed77ac2 100644
+--- a/drivers/acpi/Kconfig
++++ b/drivers/acpi/Kconfig
+@@ -352,6 +352,7 @@ config ACPI_SYSTEM
+ config X86_PM_TIMER
+ bool "Power Management Timer Support" if EMBEDDED
+ depends on X86
++ depends on !XEN
+ default y
+ help
+ The Power Management Timer is available on all ACPI-capable,
+@@ -405,4 +406,13 @@ config ACPI_SBS
+ This driver adds support for the Smart Battery System, another
+ type of access to battery information, found on some laptops.
+
++config ACPI_PV_SLEEP
++ bool
++ depends on X86 && XEN && ACPI_SLEEP
++ default y
++
++config PROCESSOR_EXTERNAL_CONTROL
++ bool
++ depends on X86 && XEN
++ default y
+ endif # ACPI
+diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile
+index 40b0fca..f48cd06 100644
+--- a/drivers/acpi/Makefile
++++ b/drivers/acpi/Makefile
+@@ -34,6 +34,9 @@ processor-objs += processor_core.o processor_throttling.o \
+ ifdef CONFIG_CPU_FREQ
+ processor-objs += processor_perflib.o
+ endif
++ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++processor-objs += processor_perflib.o processor_extcntl.o
++endif
+
+ obj-y += sleep/
+ obj-y += bus.o glue.o
+diff --git a/drivers/acpi/hardware/hwsleep.c b/drivers/acpi/hardware/hwsleep.c
+index dba3cfb..6d11299 100644
+--- a/drivers/acpi/hardware/hwsleep.c
++++ b/drivers/acpi/hardware/hwsleep.c
+@@ -256,7 +256,11 @@ acpi_status asmlinkage acpi_enter_sleep_state(u8 sleep_state)
+ u32 PM1Bcontrol;
+ struct acpi_bit_register_info *sleep_type_reg_info;
+ struct acpi_bit_register_info *sleep_enable_reg_info;
++#if !(defined(CONFIG_XEN) && defined(CONFIG_X86))
+ u32 in_value;
++#else
++ int err;
++#endif
+ struct acpi_object_list arg_list;
+ union acpi_object arg;
+ acpi_status status;
+@@ -366,6 +370,7 @@ acpi_status asmlinkage acpi_enter_sleep_state(u8 sleep_state)
+
+ ACPI_FLUSH_CPU_CACHE();
+
++#if !(defined(CONFIG_XEN) && defined(CONFIG_X86))
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
+ PM1Acontrol);
+ if (ACPI_FAILURE(status)) {
+@@ -412,6 +417,16 @@ acpi_status asmlinkage acpi_enter_sleep_state(u8 sleep_state)
+ /* Spin until we wake */
+
+ } while (!in_value);
++#else
++ /* PV ACPI just need check hypercall return value */
++ err = acpi_notify_hypervisor_state(sleep_state,
++ PM1Acontrol, PM1Bcontrol);
++ if (err) {
++ ACPI_DEBUG_PRINT((ACPI_DB_ERROR,
++ "Hypervisor failure [%d]\n", err));
++ return_ACPI_STATUS(AE_ERROR);
++ }
++#endif
+
+ return_ACPI_STATUS(AE_OK);
+ }
+@@ -430,6 +445,7 @@ ACPI_EXPORT_SYMBOL(acpi_enter_sleep_state)
+ * THIS FUNCTION MUST BE CALLED WITH INTERRUPTS DISABLED
+ *
+ ******************************************************************************/
++#ifndef CONFIG_XEN
+ acpi_status asmlinkage acpi_enter_sleep_state_s4bios(void)
+ {
+ u32 in_value;
+@@ -479,6 +495,7 @@ acpi_status asmlinkage acpi_enter_sleep_state_s4bios(void)
+ }
+
+ ACPI_EXPORT_SYMBOL(acpi_enter_sleep_state_s4bios)
++#endif
+
+ /*******************************************************************************
+ *
+diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c
+index 9dd0fa9..a409bbd 100644
+--- a/drivers/acpi/processor_core.c
++++ b/drivers/acpi/processor_core.c
+@@ -572,7 +572,8 @@ static int acpi_processor_get_info(struct acpi_processor *pr, unsigned has_uid)
+ */
+ if (pr->id == -1) {
+ if (ACPI_FAILURE
+- (acpi_processor_hotadd_init(pr->handle, &pr->id))) {
++ (acpi_processor_hotadd_init(pr->handle, &pr->id)) &&
++ !processor_cntl_external()) {
+ return -ENODEV;
+ }
+ }
+@@ -614,7 +615,7 @@ static int acpi_processor_get_info(struct acpi_processor *pr, unsigned has_uid)
+ return 0;
+ }
+
+-static void *processor_device_array[NR_CPUS];
++static void *processor_device_array[NR_ACPI_CPUS];
+
+ static int __cpuinit acpi_processor_start(struct acpi_device *device)
+ {
+@@ -622,31 +623,46 @@ static int __cpuinit acpi_processor_start(struct acpi_device *device)
+ acpi_status status = AE_OK;
+ struct acpi_processor *pr;
+
++ processor_extcntl_init();
+
+ pr = acpi_driver_data(device);
+
+ result = acpi_processor_get_info(pr, device->flags.unique_id);
+- if (result) {
++ if (result ||
++ ((pr->id == -1) && !processor_cntl_external())) {
+ /* Processor is physically not present */
+ return 0;
+ }
+
+- BUG_ON((pr->id >= nr_cpu_ids) || (pr->id < 0));
++ BUG_ON(!processor_cntl_external() &&
++ ((pr->id >= nr_cpu_ids) || (pr->id < 0)));
+
+ /*
+ * Buggy BIOS check
+ * ACPI id of processors can be reported wrongly by the BIOS.
+ * Don't trust it blindly
+ */
++#ifdef CONFIG_XEN
++ BUG_ON(pr->acpi_id >= NR_ACPI_CPUS);
++ if (processor_device_array[pr->acpi_id] != NULL &&
++ processor_device_array[pr->acpi_id] != device) {
++#else
+ if (processor_device_array[pr->id] != NULL &&
+ processor_device_array[pr->id] != device) {
++#endif /* CONFIG_XEN */
+ printk(KERN_WARNING "BIOS reported wrong ACPI id "
+ "for the processor\n");
+ return -ENODEV;
+ }
++#ifdef CONFIG_XEN
++ processor_device_array[pr->acpi_id] = device;
++ if (pr->id != -1)
++ processors[pr->id] = pr;
++#else
+ processor_device_array[pr->id] = device;
+
+ processors[pr->id] = pr;
++#endif /* CONFIG_XEN */
+
+ result = acpi_processor_add_fs(device);
+ if (result)
+@@ -658,7 +674,7 @@ static int __cpuinit acpi_processor_start(struct acpi_device *device)
+ /* _PDC call should be done before doing anything else (if reqd.). */
+ arch_acpi_processor_init_pdc(pr);
+ acpi_processor_set_pdc(pr);
+-#ifdef CONFIG_CPU_FREQ
++#if defined(CONFIG_CPU_FREQ) || defined(CONFIG_PROCESSOR_EXTERNAL_CONTROL)
+ acpi_processor_ppc_has_changed(pr);
+ #endif
+ acpi_processor_get_throttling_info(pr);
+@@ -667,6 +683,10 @@ static int __cpuinit acpi_processor_start(struct acpi_device *device)
+
+ acpi_processor_power_init(pr, device);
+
++ result = processor_extcntl_prepare(pr);
++ if (result)
++ goto end;
++
+ pr->cdev = thermal_cooling_device_register("Processor", device,
+ &processor_cooling_ops);
+ if (IS_ERR(pr->cdev)) {
+@@ -819,8 +839,14 @@ static int acpi_processor_remove(struct acpi_device *device, int type)
+ pr->cdev = NULL;
+ }
+
++#ifdef CONFIG_XEN
++ if (pr->id != -1)
++ processors[pr->id] = NULL;
++ processor_device_array[pr->acpi_id] = NULL;
++#else
+ processors[pr->id] = NULL;
+ processor_device_array[pr->id] = NULL;
++#endif /* CONFIG_XEN */
+ kfree(pr);
+
+ return 0;
+@@ -880,6 +906,10 @@ int acpi_processor_device_add(acpi_handle handle, struct acpi_device **device)
+ if (!pr)
+ return -ENODEV;
+
++ if (processor_cntl_external())
++ processor_notify_external(pr,
++ PROCESSOR_HOTPLUG, HOTPLUG_TYPE_ADD);
++
+ if ((pr->id >= 0) && (pr->id < nr_cpu_ids)) {
+ kobject_uevent(&(*device)->dev.kobj, KOBJ_ONLINE);
+ }
+@@ -919,6 +949,10 @@ static void __ref acpi_processor_hotplug_notify(acpi_handle handle,
+ break;
+ }
+
++ if (processor_cntl_external())
++ processor_notify_external(pr,
++ PROCESSOR_HOTPLUG, HOTPLUG_TYPE_ADD);
++
+ if (pr->id >= 0 && (pr->id < nr_cpu_ids)) {
+ kobject_uevent(&device->dev.kobj, KOBJ_OFFLINE);
+ break;
+@@ -948,8 +982,18 @@ static void __ref acpi_processor_hotplug_notify(acpi_handle handle,
+ return;
+ }
+
++#ifdef CONFIG_XEN
++ if ((pr->id >= 0) && (pr->id < nr_cpu_ids)
++ && (cpu_present(pr->id)))
++#else
+ if ((pr->id < nr_cpu_ids) && (cpu_present(pr->id)))
++#endif /* CONFIG_XEN */
+ kobject_uevent(&device->dev.kobj, KOBJ_OFFLINE);
++
++ if (processor_cntl_external())
++ processor_notify_external(pr, PROCESSOR_HOTPLUG,
++ HOTPLUG_TYPE_REMOVE);
++
+ break;
+ default:
+ ACPI_DEBUG_PRINT((ACPI_DB_INFO,
+@@ -1014,6 +1058,11 @@ static acpi_status acpi_processor_hotadd_init(acpi_handle handle, int *p_cpu)
+
+ static int acpi_processor_handle_eject(struct acpi_processor *pr)
+ {
++#ifdef CONFIG_XEN
++ if (pr->id == -1)
++ return (0);
++#endif /* CONFIG_XEN */
++
+ if (cpu_online(pr->id)) {
+ return (-EINVAL);
+ }
+diff --git a/drivers/acpi/processor_extcntl.c b/drivers/acpi/processor_extcntl.c
+new file mode 100644
+index 0000000..83d4ff0
+--- /dev/null
++++ b/drivers/acpi/processor_extcntl.c
+@@ -0,0 +1,224 @@
++/*
++ * processor_extcntl.c - channel to external control logic
++ *
++ * Copyright (C) 2008, Intel corporation
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or (at
++ * your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful, but
++ * WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
++ * General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License along
++ * with this program; if not, write to the Free Software Foundation, Inc.,
++ * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
++ *
++ */
++
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/types.h>
++#include <linux/acpi.h>
++#include <linux/pm.h>
++#include <linux/cpu.h>
++
++#include <acpi/processor.h>
++
++#define ACPI_PROCESSOR_COMPONENT 0x01000000
++#define ACPI_PROCESSOR_CLASS "processor"
++#define _COMPONENT ACPI_PROCESSOR_COMPONENT
++ACPI_MODULE_NAME("processor_extcntl")
++
++static int processor_extcntl_parse_csd(struct acpi_processor *pr);
++static int processor_extcntl_get_performance(struct acpi_processor *pr);
++/*
++ * External processor control logic may register with its own set of
++ * ops to get ACPI related notification. One example is like VMM.
++ */
++const struct processor_extcntl_ops *processor_extcntl_ops;
++EXPORT_SYMBOL(processor_extcntl_ops);
++
++static int processor_notify_smm(void)
++{
++ acpi_status status;
++ static int is_done = 0;
++
++ /* only need successfully notify BIOS once */
++ /* avoid double notification which may lead to unexpected result */
++ if (is_done)
++ return 0;
++
++ /* Can't write pstate_cnt to smi_cmd if either value is zero */
++ if (!acpi_gbl_FADT.smi_command || !acpi_gbl_FADT.pstate_control) {
++ ACPI_DEBUG_PRINT((ACPI_DB_INFO,"No SMI port or pstate_cnt\n"));
++ return 0;
++ }
++
++ ACPI_DEBUG_PRINT((ACPI_DB_INFO,
++ "Writing pstate_cnt [0x%x] to smi_cmd [0x%x]\n",
++ acpi_gbl_FADT.pstate_control, acpi_gbl_FADT.smi_command));
++
++ status = acpi_os_write_port(acpi_gbl_FADT.smi_command,
++ acpi_gbl_FADT.pstate_control, 8);
++ if (ACPI_FAILURE(status))
++ return status;
++
++ is_done = 1;
++
++ return 0;
++}
++
++int processor_notify_external(struct acpi_processor *pr, int event, int type)
++{
++ int ret = -EINVAL;
++
++ if (!processor_cntl_external())
++ return -EINVAL;
++
++ switch (event) {
++ case PROCESSOR_PM_INIT:
++ case PROCESSOR_PM_CHANGE:
++ if ((type >= PM_TYPE_MAX) ||
++ !processor_extcntl_ops->pm_ops[type])
++ break;
++
++ ret = processor_extcntl_ops->pm_ops[type](pr, event);
++ break;
++ case PROCESSOR_HOTPLUG:
++ if (processor_extcntl_ops->hotplug)
++ ret = processor_extcntl_ops->hotplug(pr, type);
++ break;
++ default:
++ printk(KERN_ERR "Unsupport processor events %d.\n", event);
++ break;
++ }
++
++ return ret;
++}
++
++/*
++ * External control logic can decide to grab full or part of physical
++ * processor control bits. Take a VMM for example, physical processors
++ * are owned by VMM and thus existence information like hotplug is
++ * always required to be notified to VMM. Similar is processor idle
++ * state which is also necessarily controlled by VMM. But for other
++ * control bits like performance/throttle states, VMM may choose to
++ * control or not upon its own policy.
++ */
++void processor_extcntl_init(void)
++{
++ if (!processor_extcntl_ops)
++ arch_acpi_processor_init_extcntl(&processor_extcntl_ops);
++}
++
++/*
++ * This is called from ACPI processor init, and targeted to hold
++ * some tricky housekeeping jobs to satisfy external control model.
++ * For example, we may put dependency parse stub here for idle
++ * and performance state. Those information may be not available
++ * if splitting from dom0 control logic like cpufreq driver.
++ */
++int processor_extcntl_prepare(struct acpi_processor *pr)
++{
++ /* parse cstate dependency information */
++ if (processor_pm_external())
++ processor_extcntl_parse_csd(pr);
++
++ /* Initialize performance states */
++ if (processor_pmperf_external())
++ processor_extcntl_get_performance(pr);
++
++ return 0;
++}
++
++/*
++ * Currently no _CSD is implemented which is why existing ACPI code
++ * doesn't parse _CSD at all. But to keep interface complete with
++ * external control logic, we put a placeholder here for future
++ * compatibility.
++ */
++static int processor_extcntl_parse_csd(struct acpi_processor *pr)
++{
++ int i;
++
++ for (i = 0; i < pr->power.count; i++) {
++ if (!pr->power.states[i].valid)
++ continue;
++
++ /* No dependency by default */
++ pr->power.states[i].domain_info = NULL;
++ pr->power.states[i].csd_count = 0;
++ }
++
++ return 0;
++}
++
++/*
++ * Existing ACPI module does parse performance states at some point,
++ * when acpi-cpufreq driver is loaded which however is something
++ * we'd like to disable to avoid confliction with external control
++ * logic. So we have to collect raw performance information here
++ * when ACPI processor object is found and started.
++ */
++static int processor_extcntl_get_performance(struct acpi_processor *pr)
++{
++ int ret;
++ struct acpi_processor_performance *perf;
++ struct acpi_psd_package *pdomain;
++
++ if (pr->performance)
++ return -EBUSY;
++
++ perf = kzalloc(sizeof(struct acpi_processor_performance), GFP_KERNEL);
++ if (!perf)
++ return -ENOMEM;
++
++ pr->performance = perf;
++ /* Get basic performance state information */
++ ret = acpi_processor_get_performance_info(pr);
++ if (ret < 0)
++ goto err_out;
++
++ /*
++ * Well, here we need retrieve performance dependency information
++ * from _PSD object. The reason why existing interface is not used
++ * is due to the reason that existing interface sticks to Linux cpu
++ * id to construct some bitmap, however we want to split ACPI
++ * processor objects from Linux cpu id logic. For example, even
++ * when Linux is configured as UP, we still want to parse all ACPI
++ * processor objects to external logic. In this case, it's preferred
++ * to use ACPI ID instead.
++ */
++ pr->performance->domain_info.num_processors = 0;
++ ret = acpi_processor_get_psd(pr);
++ if (ret < 0)
++ goto err_out;
++
++ /* Some sanity check */
++ pdomain = &pr->performance->domain_info;
++ if ((pdomain->revision != ACPI_PSD_REV0_REVISION) ||
++ (pdomain->num_entries != ACPI_PSD_REV0_ENTRIES) ||
++ ((pdomain->coord_type != DOMAIN_COORD_TYPE_SW_ALL) &&
++ (pdomain->coord_type != DOMAIN_COORD_TYPE_SW_ANY) &&
++ (pdomain->coord_type != DOMAIN_COORD_TYPE_HW_ALL))) {
++ ret = -EINVAL;
++ goto err_out;
++ }
++
++ /* Last step is to notify BIOS that external logic exists */
++ processor_notify_smm();
++
++ processor_notify_external(pr, PROCESSOR_PM_INIT, PM_TYPE_PERF);
++
++ return 0;
++err_out:
++ pr->performance = NULL;
++ kfree(perf);
++ return ret;
++}
+diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
+index 556ee15..c19961b 100644
+--- a/drivers/acpi/processor_idle.c
++++ b/drivers/acpi/processor_idle.c
+@@ -952,7 +952,8 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
+ */
+ cx.entry_method = ACPI_CSTATE_HALT;
+ snprintf(cx.desc, ACPI_CX_DESC_LEN, "ACPI HLT");
+- } else {
++ /* This doesn't apply to external control case */
++ } else if (!processor_pm_external()) {
+ continue;
+ }
+ } else {
+@@ -976,6 +977,12 @@ static int acpi_processor_get_power_info_cst(struct acpi_processor *pr)
+
+ cx.power = obj->integer.value;
+
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ /* cache control methods to notify external logic */
++ if (processor_pm_external())
++ memcpy(&cx.reg, reg, sizeof(*reg));
++#endif
++
+ current_count++;
+ memcpy(&(pr->power.states[current_count]), &cx, sizeof(cx));
+
+@@ -1313,12 +1320,16 @@ int acpi_processor_cst_has_changed(struct acpi_processor *pr)
+ return -ENODEV;
+
+ /* Fall back to the default idle loop */
+- pm_idle = pm_idle_save;
++ if (!processor_pm_external())
++ pm_idle = pm_idle_save;
+ synchronize_sched(); /* Relies on interrupts forcing exit from idle. */
+
+ pr->flags.power = 0;
+ result = acpi_processor_get_power_info(pr);
+- if ((pr->flags.power == 1) && (pr->flags.power_setup_done))
++ if (processor_pm_external())
++ processor_notify_external(pr,
++ PROCESSOR_PM_CHANGE, PM_TYPE_IDLE);
++ else if ((pr->flags.power == 1) && (pr->flags.power_setup_done))
+ pm_idle = acpi_processor_idle;
+
+ return result;
+@@ -1754,6 +1765,13 @@ int acpi_processor_cst_has_changed(struct acpi_processor *pr)
+ if (!pr->flags.power_setup_done)
+ return -ENODEV;
+
++ if (processor_pm_external()) {
++ acpi_processor_get_power_info(pr);
++ processor_notify_external(pr,
++ PROCESSOR_PM_CHANGE, PM_TYPE_IDLE);
++ return ret;
++ }
++
+ cpuidle_pause_and_lock();
+ cpuidle_disable_device(&pr->power.dev);
+ acpi_processor_get_power_info(pr);
+@@ -1828,7 +1846,7 @@ int __cpuinit acpi_processor_power_init(struct acpi_processor *pr,
+ printk(")\n");
+
+ #ifndef CONFIG_CPU_IDLE
+- if (pr->id == 0) {
++ if (!processor_pm_external() && (pr->id == 0)) {
+ pm_idle_save = pm_idle;
+ pm_idle = acpi_processor_idle;
+ }
+@@ -1842,6 +1860,11 @@ int __cpuinit acpi_processor_power_init(struct acpi_processor *pr,
+ acpi_driver_data(device));
+ if (!entry)
+ return -EIO;
++
++ if (processor_pm_external())
++ processor_notify_external(pr,
++ PROCESSOR_PM_INIT, PM_TYPE_IDLE);
++
+ return 0;
+ }
+
+diff --git a/drivers/acpi/processor_perflib.c b/drivers/acpi/processor_perflib.c
+index 8c06a53..63d0c34 100644
+--- a/drivers/acpi/processor_perflib.c
++++ b/drivers/acpi/processor_perflib.c
+@@ -80,6 +80,7 @@ MODULE_PARM_DESC(ignore_ppc, "If the frequency of your machine gets wrongly" \
+
+ static int acpi_processor_ppc_status;
+
++#ifdef CONFIG_CPU_FREQ
+ static int acpi_processor_ppc_notifier(struct notifier_block *nb,
+ unsigned long event, void *data)
+ {
+@@ -122,6 +123,7 @@ static int acpi_processor_ppc_notifier(struct notifier_block *nb,
+ static struct notifier_block acpi_ppc_notifier_block = {
+ .notifier_call = acpi_processor_ppc_notifier,
+ };
++#endif /* CONFIG_CPU_FREQ */
+
+ static int acpi_processor_get_platform_limit(struct acpi_processor *pr)
+ {
+@@ -166,9 +168,15 @@ int acpi_processor_ppc_has_changed(struct acpi_processor *pr)
+ if (ret < 0)
+ return (ret);
+ else
++#ifdef CONFIG_CPU_FREQ
+ return cpufreq_update_policy(pr->id);
++#elif CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ return processor_notify_external(pr,
++ PROCESSOR_PM_CHANGE, PM_TYPE_PERF);
++#endif
+ }
+
++#ifdef CONFIG_CPU_FREQ
+ void acpi_processor_ppc_init(void)
+ {
+ if (!cpufreq_register_notifier
+@@ -187,6 +195,7 @@ void acpi_processor_ppc_exit(void)
+
+ acpi_processor_ppc_status &= ~PPC_REGISTERED;
+ }
++#endif /* CONFIG_CPU_FREQ */
+
+ static int acpi_processor_get_performance_control(struct acpi_processor *pr)
+ {
+@@ -328,7 +337,10 @@ static int acpi_processor_get_performance_states(struct acpi_processor *pr)
+ return result;
+ }
+
+-static int acpi_processor_get_performance_info(struct acpi_processor *pr)
++#ifndef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++static
++#endif
++int acpi_processor_get_performance_info(struct acpi_processor *pr)
+ {
+ int result = 0;
+ acpi_status status = AE_OK;
+@@ -356,6 +368,7 @@ static int acpi_processor_get_performance_info(struct acpi_processor *pr)
+ return 0;
+ }
+
++#ifdef CONFIG_CPU_FREQ
+ int acpi_processor_notify_smm(struct module *calling_module)
+ {
+ acpi_status status;
+@@ -416,6 +429,7 @@ int acpi_processor_notify_smm(struct module *calling_module)
+ }
+
+ EXPORT_SYMBOL(acpi_processor_notify_smm);
++#endif /* CONFIG_CPU_FREQ */
+
+ #ifdef CONFIG_X86_ACPI_CPUFREQ_PROC_INTF
+ /* /proc/acpi/processor/../performance interface (DEPRECATED) */
+@@ -507,7 +521,10 @@ static void acpi_cpufreq_remove_file(struct acpi_processor *pr)
+ }
+ #endif /* CONFIG_X86_ACPI_CPUFREQ_PROC_INTF */
+
+-static int acpi_processor_get_psd(struct acpi_processor *pr)
++#ifndef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++static
++#endif
++int acpi_processor_get_psd(struct acpi_processor *pr)
+ {
+ int result = 0;
+ acpi_status status = AE_OK;
+diff --git a/drivers/acpi/sleep/main.c b/drivers/acpi/sleep/main.c
+index 495c63a..278f72a 100644
+--- a/drivers/acpi/sleep/main.c
++++ b/drivers/acpi/sleep/main.c
+@@ -31,6 +31,7 @@ static u32 acpi_target_sleep_state = ACPI_STATE_S0;
+ static int acpi_sleep_prepare(u32 acpi_state)
+ {
+ #ifdef CONFIG_ACPI_SLEEP
++#ifndef CONFIG_ACPI_PV_SLEEP
+ /* do we have a wakeup address for S2 and S3? */
+ if (acpi_state == ACPI_STATE_S3) {
+ if (!acpi_wakeup_address) {
+@@ -40,6 +41,7 @@ static int acpi_sleep_prepare(u32 acpi_state)
+ (acpi_physical_address)acpi_wakeup_address);
+
+ }
++#endif
+ ACPI_FLUSH_CPU_CACHE();
+ acpi_enable_wakeup_device_prep(acpi_state);
+ #endif
+@@ -136,7 +138,14 @@ static int acpi_suspend_enter(suspend_state_t pm_state)
+ break;
+
+ case ACPI_STATE_S3:
++#ifdef CONFIG_ACPI_PV_SLEEP
++ /* Hyperviosr will save and restore CPU context
++ * and then we can skip low level housekeeping here.
++ */
++ acpi_enter_sleep_state(acpi_state);
++#else
+ do_suspend_lowlevel();
++#endif
+ break;
+ }
+
+@@ -186,7 +195,7 @@ static void acpi_suspend_finish(void)
+
+ acpi_target_sleep_state = ACPI_STATE_S0;
+
+-#ifdef CONFIG_X86
++#if defined(CONFIG_X86) && !defined(CONFIG_ACPI_PV_SLEEP)
+ if (init_8259A_after_S1) {
+ printk("Broken toshiba laptop -> kicking interrupts\n");
+ init_8259A(0);
+diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
+index 0d1d213..463d112 100644
+--- a/drivers/block/Kconfig
++++ b/drivers/block/Kconfig
+@@ -417,9 +417,9 @@ config XILINX_SYSACE
+ help
+ Include support for the Xilinx SystemACE CompactFlash interface
+
+-config XEN_BLKDEV_FRONTEND
++config XEN_BLKFRONT
+ tristate "Xen virtual block device support"
+- depends on XEN
++ depends on PARAVIRT_XEN
+ default y
+ help
+ This driver implements the front-end of the Xen virtual
+diff --git a/drivers/block/Makefile b/drivers/block/Makefile
+index 5e58430..ef8e7de 100644
+--- a/drivers/block/Makefile
++++ b/drivers/block/Makefile
+@@ -30,4 +30,4 @@ obj-$(CONFIG_VIODASD) += viodasd.o
+ obj-$(CONFIG_BLK_DEV_SX8) += sx8.o
+ obj-$(CONFIG_BLK_DEV_UB) += ub.o
+
+-obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += xen-blkfront.o
++obj-$(CONFIG_XEN_BLKFRONT) += xen-blkfront.o
+diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
+index f2fff57..e1a7fcd 100644
+--- a/drivers/block/xen-blkfront.c
++++ b/drivers/block/xen-blkfront.c
+@@ -981,7 +981,6 @@ static struct xenbus_device_id blkfront_ids[] = {
+
+ static struct xenbus_driver blkfront = {
+ .name = "vbd",
+- .owner = THIS_MODULE,
+ .ids = blkfront_ids,
+ .probe = blkfront_probe,
+ .remove = blkfront_remove,
+diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig
+index 2d854bb..2a40e89 100644
+--- a/drivers/char/Kconfig
++++ b/drivers/char/Kconfig
+@@ -622,7 +622,7 @@ config HVC_BEAT
+
+ config HVC_XEN
+ bool "Xen Hypervisor Console support"
+- depends on XEN
++ depends on PARAVIRT_XEN
+ select HVC_DRIVER
+ default y
+ help
+@@ -1029,7 +1029,7 @@ config MAX_RAW_DEVS
+ config HPET
+ bool "HPET - High Precision Event Timer" if (X86 || IA64)
+ default n
+- depends on ACPI
++ depends on ACPI && !XEN
+ help
+ If you say Y here, you will have a miscdevice named "/dev/hpet/". Each
+ open selects one of the timers supported by the HPET. The timers are
+diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c
+index df70264..719bbb6 100644
+--- a/drivers/char/agp/intel-agp.c
++++ b/drivers/char/agp/intel-agp.c
+@@ -245,6 +245,13 @@ static void *i8xx_alloc_pages(void)
+ if (page == NULL)
+ return NULL;
+
++#ifdef CONFIG_XEN
++ if (xen_create_contiguous_region((unsigned long)page_address(page), 2, 32)) {
++ __free_pages(page, 2);
++ return NULL;
++ }
++#endif
++
+ if (set_pages_uc(page, 4) < 0) {
+ set_pages_wb(page, 4);
+ __free_pages(page, 2);
+@@ -264,6 +271,9 @@ static void i8xx_destroy_pages(void *addr)
+
+ page = virt_to_page(addr);
+ set_pages_wb(page, 4);
++#ifdef CONFIG_XEN
++ xen_destroy_contiguous_region((unsigned long)page_address(page), 2);
++#endif
+ put_page(page);
+ __free_pages(page, 2);
+ atomic_dec(&agp_bridge->current_memory_agp);
+diff --git a/drivers/char/mem.c b/drivers/char/mem.c
+index 934ffaf..6ed61ea 100644
+--- a/drivers/char/mem.c
++++ b/drivers/char/mem.c
+@@ -109,6 +109,7 @@ void __attribute__((weak)) unxlate_dev_mem_ptr(unsigned long phys, void *addr)
+ {
+ }
+
++#ifndef ARCH_HAS_DEV_MEM
+ /*
+ * This funcion reads the *physical* memory. The f_pos points directly to the
+ * memory location.
+@@ -253,6 +254,7 @@ static ssize_t write_mem(struct file * file, const char __user * buf,
+ *ppos += written;
+ return written;
+ }
++#endif
+
+ int __attribute__((weak)) phys_mem_access_prot_allowed(struct file *file,
+ unsigned long pfn, unsigned long size, pgprot_t *vma_prot)
+@@ -368,6 +370,9 @@ static int mmap_mem(struct file * file, struct vm_area_struct * vma)
+ static int mmap_kmem(struct file * file, struct vm_area_struct * vma)
+ {
+ unsigned long pfn;
++#ifdef CONFIG_XEN
++ unsigned long i, count;
++#endif
+
+ /* Turn a kernel-virtual address into a physical page frame */
+ pfn = __pa((u64)vma->vm_pgoff << PAGE_SHIFT) >> PAGE_SHIFT;
+@@ -382,6 +387,13 @@ static int mmap_kmem(struct file * file, struct vm_area_struct * vma)
+ if (!pfn_valid(pfn))
+ return -EIO;
+
++#ifdef CONFIG_XEN
++ count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ for (i = 0; i < count; i++)
++ if ((pfn + i) != mfn_to_local_pfn(pfn_to_mfn(pfn + i)))
++ return -EIO;
++#endif
++
+ vma->vm_pgoff = pfn;
+ return mmap_mem(file, vma);
+ }
+@@ -798,6 +810,7 @@ static int open_port(struct inode * inode, struct file * filp)
+ #define open_kmem open_mem
+ #define open_oldmem open_mem
+
++#ifndef ARCH_HAS_DEV_MEM
+ static const struct file_operations mem_fops = {
+ .llseek = memory_lseek,
+ .read = read_mem,
+@@ -806,6 +819,9 @@ static const struct file_operations mem_fops = {
+ .open = open_mem,
+ .get_unmapped_area = get_unmapped_area_mem,
+ };
++#else
++extern const struct file_operations mem_fops;
++#endif
+
+ #ifdef CONFIG_DEVKMEM
+ static const struct file_operations kmem_fops = {
+diff --git a/drivers/char/tpm/Kconfig b/drivers/char/tpm/Kconfig
+index 3738cfa..692573e 100644
+--- a/drivers/char/tpm/Kconfig
++++ b/drivers/char/tpm/Kconfig
+@@ -57,4 +57,13 @@ config TCG_INFINEON
+ Further information on this driver and the supported hardware
+ can be found at http://www.prosec.rub.de/tpm
+
++config TCG_XEN
++ tristate "XEN TPM Interface"
++ depends on XEN
++ ---help---
++ If you want to make TPM support available to a Xen user domain,
++ say Yes and it will be accessible from within Linux.
++ To compile this driver as a module, choose M here; the module
++ will be called tpm_xenu.
++
+ endif # TCG_TPM
+diff --git a/drivers/char/tpm/Makefile b/drivers/char/tpm/Makefile
+index ea3a1e0..b5cea0a 100644
+--- a/drivers/char/tpm/Makefile
++++ b/drivers/char/tpm/Makefile
+@@ -9,3 +9,5 @@ obj-$(CONFIG_TCG_TIS) += tpm_tis.o
+ obj-$(CONFIG_TCG_NSC) += tpm_nsc.o
+ obj-$(CONFIG_TCG_ATMEL) += tpm_atmel.o
+ obj-$(CONFIG_TCG_INFINEON) += tpm_infineon.o
++obj-$(CONFIG_TCG_XEN) += tpm_xenu.o
++tpm_xenu-y = tpm_xen.o tpm_vtpm.o
+diff --git a/drivers/char/tpm/tpm.h b/drivers/char/tpm/tpm.h
+index e885148..557550a 100644
+--- a/drivers/char/tpm/tpm.h
++++ b/drivers/char/tpm/tpm.h
+@@ -107,6 +107,9 @@ struct tpm_chip {
+ struct dentry **bios_dir;
+
+ struct list_head list;
++#ifdef CONFIG_XEN
++ void *priv;
++#endif
+ void (*release) (struct device *);
+ };
+
+@@ -124,6 +127,18 @@ static inline void tpm_write_index(int base, int index, int value)
+ outb(value & 0xFF, base+1);
+ }
+
++#ifdef CONFIG_XEN
++static inline void *chip_get_private(const struct tpm_chip *chip)
++{
++ return chip->priv;
++}
++
++static inline void chip_set_private(struct tpm_chip *chip, void *priv)
++{
++ chip->priv = priv;
++}
++#endif
++
+ 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 --git a/drivers/char/tpm/tpm_vtpm.c b/drivers/char/tpm/tpm_vtpm.c
+new file mode 100644
+index 0000000..1f6982e
+--- /dev/null
++++ b/drivers/char/tpm/tpm_vtpm.c
+@@ -0,0 +1,542 @@
++/*
++ * Copyright (C) 2006 IBM Corporation
++ *
++ * Authors:
++ * Stefan Berger <stefanb at us.ibm.com>
++ *
++ * Generic device driver part for device drivers in a virtualized
++ * environment.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License as
++ * published by the Free Software Foundation, version 2 of the
++ * License.
++ *
++ */
++
++#include <asm/uaccess.h>
++#include <linux/list.h>
++#include <linux/device.h>
++#include <linux/interrupt.h>
++#include <linux/platform_device.h>
++#include "tpm.h"
++#include "tpm_vtpm.h"
++
++/* read status bits */
++enum {
++ STATUS_BUSY = 0x01,
++ STATUS_DATA_AVAIL = 0x02,
++ STATUS_READY = 0x04
++};
++
++struct transmission {
++ struct list_head next;
++
++ unsigned char *request;
++ size_t request_len;
++ size_t request_buflen;
++
++ unsigned char *response;
++ size_t response_len;
++ size_t response_buflen;
++
++ unsigned int flags;
++};
++
++enum {
++ TRANSMISSION_FLAG_WAS_QUEUED = 0x1
++};
++
++
++enum {
++ DATAEX_FLAG_QUEUED_ONLY = 0x1
++};
++
++
++/* local variables */
++
++/* local function prototypes */
++static int _vtpm_send_queued(struct tpm_chip *chip);
++
++
++/* =============================================================
++ * Some utility functions
++ * =============================================================
++ */
++static void vtpm_state_init(struct vtpm_state *vtpms)
++{
++ vtpms->current_request = NULL;
++ spin_lock_init(&vtpms->req_list_lock);
++ init_waitqueue_head(&vtpms->req_wait_queue);
++ INIT_LIST_HEAD(&vtpms->queued_requests);
++
++ vtpms->current_response = NULL;
++ spin_lock_init(&vtpms->resp_list_lock);
++ init_waitqueue_head(&vtpms->resp_wait_queue);
++
++ vtpms->disconnect_time = jiffies;
++}
++
++
++static inline struct transmission *transmission_alloc(void)
++{
++ return kzalloc(sizeof(struct transmission), GFP_ATOMIC);
++}
++
++static unsigned char *
++transmission_set_req_buffer(struct transmission *t,
++ unsigned char *buffer, size_t len)
++{
++ if (t->request_buflen < len) {
++ kfree(t->request);
++ t->request = kmalloc(len, GFP_KERNEL);
++ if (!t->request) {
++ t->request_buflen = 0;
++ return NULL;
++ }
++ t->request_buflen = len;
++ }
++
++ memcpy(t->request, buffer, len);
++ t->request_len = len;
++
++ return t->request;
++}
++
++static unsigned char *
++transmission_set_res_buffer(struct transmission *t,
++ const unsigned char *buffer, size_t len)
++{
++ if (t->response_buflen < len) {
++ kfree(t->response);
++ t->response = kmalloc(len, GFP_ATOMIC);
++ if (!t->response) {
++ t->response_buflen = 0;
++ return NULL;
++ }
++ t->response_buflen = len;
++ }
++
++ memcpy(t->response, buffer, len);
++ t->response_len = len;
++
++ return t->response;
++}
++
++static inline void transmission_free(struct transmission *t)
++{
++ kfree(t->request);
++ kfree(t->response);
++ kfree(t);
++}
++
++/* =============================================================
++ * Interface with the lower layer driver
++ * =============================================================
++ */
++/*
++ * Lower layer uses this function to make a response available.
++ */
++int vtpm_vd_recv(const struct tpm_chip *chip,
++ const unsigned char *buffer, size_t count,
++ void *ptr)
++{
++ unsigned long flags;
++ int ret_size = 0;
++ struct transmission *t;
++ struct vtpm_state *vtpms;
++
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ /*
++ * The list with requests must contain one request
++ * only and the element there must be the one that
++ * was passed to me from the front-end.
++ */
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ if (vtpms->current_request != ptr) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return 0;
++ }
++
++ if ((t = vtpms->current_request)) {
++ transmission_free(t);
++ vtpms->current_request = NULL;
++ }
++
++ t = transmission_alloc();
++ if (t) {
++ if (!transmission_set_res_buffer(t, buffer, count)) {
++ transmission_free(t);
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return -ENOMEM;
++ }
++ ret_size = count;
++ vtpms->current_response = t;
++ wake_up_interruptible(&vtpms->resp_wait_queue);
++ }
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++
++ return ret_size;
++}
++
++
++/*
++ * Lower layer indicates its status (connected/disconnected)
++ */
++void vtpm_vd_status(const struct tpm_chip *chip, u8 vd_status)
++{
++ struct vtpm_state *vtpms;
++
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ vtpms->vd_status = vd_status;
++ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
++ vtpms->disconnect_time = jiffies;
++ }
++}
++
++/* =============================================================
++ * Interface with the generic TPM driver
++ * =============================================================
++ */
++static int vtpm_recv(struct tpm_chip *chip, u8 *buf, size_t count)
++{
++ int rc = 0;
++ unsigned long flags;
++ struct vtpm_state *vtpms;
++
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ /*
++ * Check if the previous operation only queued the command
++ * In this case there won't be a response, so I just
++ * return from here and reset that flag. In any other
++ * case I should receive a response from the back-end.
++ */
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ if ((vtpms->flags & DATAEX_FLAG_QUEUED_ONLY) != 0) {
++ vtpms->flags &= ~DATAEX_FLAG_QUEUED_ONLY;
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ /*
++ * The first few commands (measurements) must be
++ * queued since it might not be possible to talk to the
++ * TPM, yet.
++ * Return a response of up to 30 '0's.
++ */
++
++ count = min_t(size_t, count, 30);
++ memset(buf, 0x0, count);
++ return count;
++ }
++ /*
++ * Check whether something is in the responselist and if
++ * there's nothing in the list wait for something to appear.
++ */
++
++ if (!vtpms->current_response) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ interruptible_sleep_on_timeout(&vtpms->resp_wait_queue,
++ 1000);
++ spin_lock_irqsave(&vtpms->resp_list_lock ,flags);
++ }
++
++ if (vtpms->current_response) {
++ struct transmission *t = vtpms->current_response;
++ vtpms->current_response = NULL;
++ rc = min(count, t->response_len);
++ memcpy(buf, t->response, rc);
++ transmission_free(t);
++ }
++
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return rc;
++}
++
++static int vtpm_send(struct tpm_chip *chip, u8 *buf, size_t count)
++{
++ int rc = 0;
++ unsigned long flags;
++ struct transmission *t = transmission_alloc();
++ struct vtpm_state *vtpms;
++
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ if (!t)
++ return -ENOMEM;
++ /*
++ * If there's a current request, it must be the
++ * previous request that has timed out.
++ */
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ if (vtpms->current_request != NULL) {
++ printk("WARNING: Sending although there is a request outstanding.\n"
++ " Previous request must have timed out.\n");
++ transmission_free(vtpms->current_request);
++ vtpms->current_request = NULL;
++ }
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++
++ /*
++ * Queue the packet if the driver below is not
++ * ready, yet, or there is any packet already
++ * in the queue.
++ * If the driver below is ready, unqueue all
++ * packets first before sending our current
++ * packet.
++ * For each unqueued packet, except for the
++ * last (=current) packet, call the function
++ * tpm_xen_recv to wait for the response to come
++ * back.
++ */
++ if ((vtpms->vd_status & TPM_VD_STATUS_CONNECTED) == 0) {
++ if (time_after(jiffies,
++ vtpms->disconnect_time + HZ * 10)) {
++ rc = -ENOENT;
++ } else {
++ goto queue_it;
++ }
++ } else {
++ /*
++ * Send all queued packets.
++ */
++ if (_vtpm_send_queued(chip) == 0) {
++
++ vtpms->current_request = t;
++
++ rc = vtpm_vd_send(vtpms->tpm_private,
++ buf,
++ count,
++ t);
++ /*
++ * The generic TPM driver will call
++ * the function to receive the response.
++ */
++ if (rc < 0) {
++ vtpms->current_request = NULL;
++ goto queue_it;
++ }
++ } else {
++queue_it:
++ if (!transmission_set_req_buffer(t, buf, count)) {
++ transmission_free(t);
++ rc = -ENOMEM;
++ goto exit;
++ }
++ /*
++ * An error occurred. Don't event try
++ * to send the current request. Just
++ * queue it.
++ */
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ vtpms->flags |= DATAEX_FLAG_QUEUED_ONLY;
++ list_add_tail(&t->next, &vtpms->queued_requests);
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++ }
++ }
++
++exit:
++ return rc;
++}
++
++
++/*
++ * Send all queued requests.
++ */
++static int _vtpm_send_queued(struct tpm_chip *chip)
++{
++ int rc;
++ int error = 0;
++ long flags;
++ unsigned char buffer[1];
++ struct vtpm_state *vtpms;
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++
++ while (!list_empty(&vtpms->queued_requests)) {
++ /*
++ * Need to dequeue them.
++ * Read the result into a dummy buffer.
++ */
++ struct transmission *qt = (struct transmission *)
++ vtpms->queued_requests.next;
++ list_del(&qt->next);
++ vtpms->current_request = qt;
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++
++ rc = vtpm_vd_send(vtpms->tpm_private,
++ qt->request,
++ qt->request_len,
++ qt);
++
++ if (rc < 0) {
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ if ((qt = vtpms->current_request) != NULL) {
++ /*
++ * requeue it at the beginning
++ * of the list
++ */
++ list_add(&qt->next,
++ &vtpms->queued_requests);
++ }
++ vtpms->current_request = NULL;
++ error = 1;
++ break;
++ }
++ /*
++ * After this point qt is not valid anymore!
++ * It is freed when the front-end is delivering
++ * the data by calling tpm_recv
++ */
++ /*
++ * Receive response into provided dummy buffer
++ */
++ rc = vtpm_recv(chip, buffer, sizeof(buffer));
++ spin_lock_irqsave(&vtpms->req_list_lock, flags);
++ }
++
++ spin_unlock_irqrestore(&vtpms->req_list_lock, flags);
++
++ return error;
++}
++
++static void vtpm_cancel(struct tpm_chip *chip)
++{
++ unsigned long flags;
++ struct vtpm_state *vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
++
++ if (!vtpms->current_response && vtpms->current_request) {
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ interruptible_sleep_on(&vtpms->resp_wait_queue);
++ spin_lock_irqsave(&vtpms->resp_list_lock,flags);
++ }
++
++ if (vtpms->current_response) {
++ struct transmission *t = vtpms->current_response;
++ vtpms->current_response = NULL;
++ transmission_free(t);
++ }
++
++ spin_unlock_irqrestore(&vtpms->resp_list_lock,flags);
++}
++
++static u8 vtpm_status(struct tpm_chip *chip)
++{
++ u8 rc = 0;
++ unsigned long flags;
++ struct vtpm_state *vtpms;
++
++ vtpms = (struct vtpm_state *)chip_get_private(chip);
++
++ spin_lock_irqsave(&vtpms->resp_list_lock, flags);
++ /*
++ * Data are available if:
++ * - there's a current response
++ * - the last packet was queued only (this is fake, but necessary to
++ * get the generic TPM layer to call the receive function.)
++ */
++ if (vtpms->current_response ||
++ 0 != (vtpms->flags & DATAEX_FLAG_QUEUED_ONLY)) {
++ rc = STATUS_DATA_AVAIL;
++ } else if (!vtpms->current_response && !vtpms->current_request) {
++ rc = STATUS_READY;
++ }
++
++ spin_unlock_irqrestore(&vtpms->resp_list_lock, flags);
++ return rc;
++}
++
++static struct file_operations vtpm_ops = {
++ .owner = THIS_MODULE,
++ .llseek = no_llseek,
++ .open = tpm_open,
++ .read = tpm_read,
++ .write = tpm_write,
++ .release = tpm_release,
++};
++
++static DEVICE_ATTR(pubek, S_IRUGO, tpm_show_pubek, NULL);
++static DEVICE_ATTR(pcrs, S_IRUGO, tpm_show_pcrs, NULL);
++static DEVICE_ATTR(enabled, S_IRUGO, tpm_show_enabled, NULL);
++static DEVICE_ATTR(active, S_IRUGO, tpm_show_active, NULL);
++static DEVICE_ATTR(owned, S_IRUGO, tpm_show_owned, NULL);
++static DEVICE_ATTR(temp_deactivated, S_IRUGO, tpm_show_temp_deactivated,
++ NULL);
++static DEVICE_ATTR(caps, S_IRUGO, tpm_show_caps, NULL);
++static DEVICE_ATTR(cancel, S_IWUSR |S_IWGRP, NULL, tpm_store_cancel);
++
++static struct attribute *vtpm_attrs[] = {
++ &dev_attr_pubek.attr,
++ &dev_attr_pcrs.attr,
++ &dev_attr_enabled.attr,
++ &dev_attr_active.attr,
++ &dev_attr_owned.attr,
++ &dev_attr_temp_deactivated.attr,
++ &dev_attr_caps.attr,
++ &dev_attr_cancel.attr,
++ NULL,
++};
++
++static struct attribute_group vtpm_attr_grp = { .attrs = vtpm_attrs };
++
++#define TPM_LONG_TIMEOUT (10 * 60 * HZ)
++
++static struct tpm_vendor_specific tpm_vtpm = {
++ .recv = vtpm_recv,
++ .send = vtpm_send,
++ .cancel = vtpm_cancel,
++ .status = vtpm_status,
++ .req_complete_mask = STATUS_BUSY | STATUS_DATA_AVAIL,
++ .req_complete_val = STATUS_DATA_AVAIL,
++ .req_canceled = STATUS_READY,
++ .attr_group = &vtpm_attr_grp,
++ .miscdev = {
++ .fops = &vtpm_ops,
++ },
++ .duration = {
++ TPM_LONG_TIMEOUT,
++ TPM_LONG_TIMEOUT,
++ TPM_LONG_TIMEOUT,
++ },
++};
++
++struct tpm_chip *init_vtpm(struct device *dev,
++ struct tpm_private *tp)
++{
++ long rc;
++ struct tpm_chip *chip;
++ struct vtpm_state *vtpms;
++
++ vtpms = kzalloc(sizeof(struct vtpm_state), GFP_KERNEL);
++ if (!vtpms)
++ return ERR_PTR(-ENOMEM);
++
++ vtpm_state_init(vtpms);
++ vtpms->tpm_private = tp;
++
++ chip = tpm_register_hardware(dev, &tpm_vtpm);
++ if (!chip) {
++ rc = -ENODEV;
++ goto err_free_mem;
++ }
++
++ chip_set_private(chip, vtpms);
++
++ return chip;
++
++err_free_mem:
++ kfree(vtpms);
++
++ return ERR_PTR(rc);
++}
++
++void cleanup_vtpm(struct device *dev)
++{
++ struct tpm_chip *chip = dev_get_drvdata(dev);
++ struct vtpm_state *vtpms = (struct vtpm_state*)chip_get_private(chip);
++ tpm_remove_hardware(dev);
++ kfree(vtpms);
++}
+diff --git a/drivers/char/tpm/tpm_vtpm.h b/drivers/char/tpm/tpm_vtpm.h
+new file mode 100644
+index 0000000..77aa342
+--- /dev/null
++++ b/drivers/char/tpm/tpm_vtpm.h
+@@ -0,0 +1,55 @@
++#ifndef TPM_VTPM_H
++#define TPM_VTPM_H
++
++struct tpm_chip;
++struct tpm_private;
++
++struct vtpm_state {
++ struct transmission *current_request;
++ spinlock_t req_list_lock;
++ wait_queue_head_t req_wait_queue;
++
++ struct list_head queued_requests;
++
++ struct transmission *current_response;
++ spinlock_t resp_list_lock;
++ wait_queue_head_t resp_wait_queue; // processes waiting for responses
++
++ u8 vd_status;
++ u8 flags;
++
++ unsigned long disconnect_time;
++
++ /*
++ * The following is a private structure of the underlying
++ * driver. It is passed as parameter in the send function.
++ */
++ struct tpm_private *tpm_private;
++};
++
++
++enum vdev_status {
++ TPM_VD_STATUS_DISCONNECTED = 0x0,
++ TPM_VD_STATUS_CONNECTED = 0x1
++};
++
++/* this function is called from tpm_vtpm.c */
++int vtpm_vd_send(struct tpm_private * tp,
++ const u8 * buf, size_t count, void *ptr);
++
++/* these functions are offered by tpm_vtpm.c */
++struct tpm_chip *init_vtpm(struct device *,
++ struct tpm_private *);
++void cleanup_vtpm(struct device *);
++int vtpm_vd_recv(const struct tpm_chip* chip,
++ const unsigned char *buffer, size_t count, void *ptr);
++void vtpm_vd_status(const struct tpm_chip *, u8 status);
++
++static inline struct tpm_private *tpm_private_from_dev(struct device *dev)
++{
++ struct tpm_chip *chip = dev_get_drvdata(dev);
++ struct vtpm_state *vtpms = chip_get_private(chip);
++ return vtpms->tpm_private;
++}
++
++#endif
+diff --git a/drivers/char/tpm/tpm_xen.c b/drivers/char/tpm/tpm_xen.c
+new file mode 100644
+index 0000000..0f4cd80
+--- /dev/null
++++ b/drivers/char/tpm/tpm_xen.c
+@@ -0,0 +1,720 @@
++/*
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from drivers/xen/netfront/netfront.c
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/errno.h>
++#include <linux/err.h>
++#include <linux/interrupt.h>
++#include <linux/mutex.h>
++#include <asm/uaccess.h>
++#include <xen/evtchn.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/tpmif.h>
++#include <xen/gnttab.h>
++#include <xen/xenbus.h>
++#include "tpm.h"
++#include "tpm_vtpm.h"
++
++#undef DEBUG
++
++/* local structures */
++struct tpm_private {
++ struct tpm_chip *chip;
++
++ tpmif_tx_interface_t *tx;
++ atomic_t refcnt;
++ unsigned int irq;
++ u8 is_connected;
++ u8 is_suspended;
++
++ spinlock_t tx_lock;
++
++ struct tx_buffer *tx_buffers[TPMIF_TX_RING_SIZE];
++
++ atomic_t tx_busy;
++ void *tx_remember;
++
++ domid_t backend_id;
++ wait_queue_head_t wait_q;
++
++ struct xenbus_device *dev;
++ int ring_ref;
++};
++
++struct tx_buffer {
++ unsigned int size; // available space in data
++ unsigned int len; // used space in data
++ unsigned char *data; // pointer to a page
++};
++
++
++/* locally visible variables */
++static grant_ref_t gref_head;
++static struct tpm_private *my_priv;
++
++/* local function prototypes */
++static irqreturn_t tpmif_int(int irq,
++ void *tpm_priv);
++static void tpmif_rx_action(unsigned long unused);
++static int tpmif_connect(struct xenbus_device *dev,
++ struct tpm_private *tp,
++ domid_t domid);
++static DECLARE_TASKLET(tpmif_rx_tasklet, tpmif_rx_action, 0);
++static int tpmif_allocate_tx_buffers(struct tpm_private *tp);
++static void tpmif_free_tx_buffers(struct tpm_private *tp);
++static void tpmif_set_connected_state(struct tpm_private *tp,
++ u8 newstate);
++static int tpm_xmit(struct tpm_private *tp,
++ const u8 * buf, size_t count, int userbuffer,
++ void *remember);
++static void destroy_tpmring(struct tpm_private *tp);
++void __exit tpmif_exit(void);
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("xen_tpm_fr (%s:%d) " fmt, __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_tpm_fr: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_tpm_fr: " fmt, ##args)
++
++#define GRANT_INVALID_REF 0
++
++
++static inline int
++tx_buffer_copy(struct tx_buffer *txb, const u8 *src, int len,
++ int isuserbuffer)
++{
++ int copied = len;
++
++ if (len > txb->size)
++ copied = txb->size;
++ if (isuserbuffer) {
++ if (copy_from_user(txb->data, src, copied))
++ return -EFAULT;
++ } else {
++ memcpy(txb->data, src, copied);
++ }
++ txb->len = len;
++ return copied;
++}
++
++static inline struct tx_buffer *tx_buffer_alloc(void)
++{
++ struct tx_buffer *txb;
++
++ txb = kzalloc(sizeof(struct tx_buffer), GFP_KERNEL);
++ if (!txb)
++ return NULL;
++
++ txb->len = 0;
++ txb->size = PAGE_SIZE;
++ txb->data = (unsigned char *)__get_free_page(GFP_KERNEL);
++ if (txb->data == NULL) {
++ kfree(txb);
++ txb = NULL;
++ }
++
++ return txb;
++}
++
++
++static inline void tx_buffer_free(struct tx_buffer *txb)
++{
++ if (txb) {
++ free_page((long)txb->data);
++ kfree(txb);
++ }
++}
++
++/**************************************************************
++ Utility function for the tpm_private structure
++**************************************************************/
++static void tpm_private_init(struct tpm_private *tp)
++{
++ spin_lock_init(&tp->tx_lock);
++ init_waitqueue_head(&tp->wait_q);
++ atomic_set(&tp->refcnt, 1);
++}
++
++static void tpm_private_put(void)
++{
++ if (!atomic_dec_and_test(&my_priv->refcnt))
++ return;
++
++ tpmif_free_tx_buffers(my_priv);
++ kfree(my_priv);
++ my_priv = NULL;
++}
++
++static struct tpm_private *tpm_private_get(void)
++{
++ int err;
++
++ if (my_priv) {
++ atomic_inc(&my_priv->refcnt);
++ return my_priv;
++ }
++
++ my_priv = kzalloc(sizeof(struct tpm_private), GFP_KERNEL);
++ if (!my_priv)
++ return NULL;
++
++ tpm_private_init(my_priv);
++ err = tpmif_allocate_tx_buffers(my_priv);
++ if (err < 0)
++ tpm_private_put();
++
++ return my_priv;
++}
++
++/**************************************************************
++
++ The interface to let the tpm plugin register its callback
++ function and send data to another partition using this module
++
++**************************************************************/
++
++static DEFINE_MUTEX(suspend_lock);
++/*
++ * Send data via this module by calling this function
++ */
++int vtpm_vd_send(struct tpm_private *tp,
++ const u8 * buf, size_t count, void *ptr)
++{
++ int sent;
++
++ mutex_lock(&suspend_lock);
++ sent = tpm_xmit(tp, buf, count, 0, ptr);
++ mutex_unlock(&suspend_lock);
++
++ return sent;
++}
++
++/**************************************************************
++ XENBUS support code
++**************************************************************/
++
++static int setup_tpmring(struct xenbus_device *dev,
++ struct tpm_private *tp)
++{
++ tpmif_tx_interface_t *sring;
++ int err;
++
++ tp->ring_ref = GRANT_INVALID_REF;
++
++ sring = (void *)__get_free_page(GFP_KERNEL);
++ if (!sring) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
++ return -ENOMEM;
++ }
++ tp->tx = sring;
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(tp->tx));
++ if (err < 0) {
++ free_page((unsigned long)sring);
++ tp->tx = NULL;
++ xenbus_dev_fatal(dev, err, "allocating grant reference");
++ goto fail;
++ }
++ tp->ring_ref = err;
++
++ err = tpmif_connect(dev, tp, dev->otherend_id);
++ if (err)
++ goto fail;
++
++ return 0;
++fail:
++ destroy_tpmring(tp);
++ return err;
++}
++
++
++static void destroy_tpmring(struct tpm_private *tp)
++{
++ tpmif_set_connected_state(tp, 0);
++
++ if (tp->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(tp->ring_ref, (unsigned long)tp->tx);
++ tp->ring_ref = GRANT_INVALID_REF;
++ tp->tx = NULL;
++ }
++
++ if (tp->irq)
++ unbind_from_irqhandler(tp->irq, tp);
++
++ tp->irq = 0;
++}
++
++
++static int talk_to_backend(struct xenbus_device *dev,
++ struct tpm_private *tp)
++{
++ const char *message = NULL;
++ int err;
++ struct xenbus_transaction xbt;
++
++ err = setup_tpmring(dev, tp);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "setting up ring");
++ goto out;
++ }
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_tpmring;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename,
++ "ring-ref","%u", tp->ring_ref);
++ if (err) {
++ message = "writing ring-ref";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(tp->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err) {
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_tpmring;
++ }
++
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ return 0;
++
++abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ if (message)
++ xenbus_dev_error(dev, err, "%s", message);
++destroy_tpmring:
++ destroy_tpmring(tp);
++out:
++ return err;
++}
++
++/**
++ * Callback received when the backend's state changes.
++ */
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ DPRINTK("\n");
++
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ break;
++
++ case XenbusStateConnected:
++ tpmif_set_connected_state(tp, 1);
++ break;
++
++ case XenbusStateClosing:
++ tpmif_set_connected_state(tp, 0);
++ xenbus_frontend_closed(dev);
++ break;
++
++ case XenbusStateClosed:
++ tpmif_set_connected_state(tp, 0);
++ if (tp->is_suspended == 0)
++ device_unregister(&dev->dev);
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
++
++static int tpmfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ int handle;
++ struct tpm_private *tp = tpm_private_get();
++
++ if (!tp)
++ return -ENOMEM;
++
++ tp->chip = init_vtpm(&dev->dev, tp);
++ if (IS_ERR(tp->chip))
++ return PTR_ERR(tp->chip);
++
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "handle", "%i", &handle);
++ if (XENBUS_EXIST_ERR(err))
++ return err;
++
++ if (err < 0) {
++ xenbus_dev_fatal(dev,err,"reading virtual-device");
++ return err;
++ }
++
++ tp->dev = dev;
++
++ err = talk_to_backend(dev, tp);
++ if (err) {
++ tpm_private_put();
++ return err;
++ }
++
++ return 0;
++}
++
++
++static int tpmfront_remove(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ destroy_tpmring(tp);
++ cleanup_vtpm(&dev->dev);
++ return 0;
++}
++
++static int tpmfront_suspend(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ u32 ctr;
++
++ /* Take the lock, preventing any application from sending. */
++ mutex_lock(&suspend_lock);
++ tp->is_suspended = 1;
++
++ for (ctr = 0; atomic_read(&tp->tx_busy); ctr++) {
++ if ((ctr % 10) == 0)
++ printk("TPM-FE [INFO]: Waiting for outstanding "
++ "request.\n");
++ /* Wait for a request to be responded to. */
++ interruptible_sleep_on_timeout(&tp->wait_q, 100);
++ }
++
++ return 0;
++}
++
++static int tpmfront_suspend_finish(struct tpm_private *tp)
++{
++ tp->is_suspended = 0;
++ /* Allow applications to send again. */
++ mutex_unlock(&suspend_lock);
++ return 0;
++}
++
++static int tpmfront_suspend_cancel(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ return tpmfront_suspend_finish(tp);
++}
++
++static int tpmfront_resume(struct xenbus_device *dev)
++{
++ struct tpm_private *tp = tpm_private_from_dev(&dev->dev);
++ destroy_tpmring(tp);
++ return talk_to_backend(dev, tp);
++}
++
++static int tpmif_connect(struct xenbus_device *dev,
++ struct tpm_private *tp,
++ domid_t domid)
++{
++ int err;
++
++ tp->backend_id = domid;
++
++ err = bind_listening_port_to_irqhandler(
++ domid, tpmif_int, IRQF_SAMPLE_RANDOM, "tpmif", tp);
++ if (err <= 0) {
++ WPRINTK("bind_listening_port_to_irqhandler failed "
++ "(err=%d)\n", err);
++ return err;
++ }
++ tp->irq = err;
++
++ return 0;
++}
++
++static struct xenbus_device_id tpmfront_ids[] = {
++ { "vtpm" },
++ { "" }
++};
++
++static struct xenbus_driver tpmfront = {
++ .name = "vtpm",
++ .ids = tpmfront_ids,
++ .probe = tpmfront_probe,
++ .remove = tpmfront_remove,
++ .resume = tpmfront_resume,
++ .otherend_changed = backend_changed,
++ .suspend = tpmfront_suspend,
++ .suspend_cancel = tpmfront_suspend_cancel,
++};
++
++static int __init init_tpm_xenbus(void)
++{
++ return xenbus_register_frontend(&tpmfront);
++}
++
++static int tpmif_allocate_tx_buffers(struct tpm_private *tp)
++{
++ unsigned int i;
++
++ for (i = 0; i < TPMIF_TX_RING_SIZE; i++) {
++ tp->tx_buffers[i] = tx_buffer_alloc();
++ if (!tp->tx_buffers[i]) {
++ tpmif_free_tx_buffers(tp);
++ return -ENOMEM;
++ }
++ }
++ return 0;
++}
++
++static void tpmif_free_tx_buffers(struct tpm_private *tp)
++{
++ unsigned int i;
++
++ for (i = 0; i < TPMIF_TX_RING_SIZE; i++)
++ tx_buffer_free(tp->tx_buffers[i]);
++}
++
++static void tpmif_rx_action(unsigned long priv)
++{
++ struct tpm_private *tp = (struct tpm_private *)priv;
++ int i = 0;
++ unsigned int received;
++ unsigned int offset = 0;
++ u8 *buffer;
++ tpmif_tx_request_t *tx = &tp->tx->ring[i].req;
++
++ atomic_set(&tp->tx_busy, 0);
++ wake_up_interruptible(&tp->wait_q);
++
++ received = tx->size;
++
++ buffer = kmalloc(received, GFP_ATOMIC);
++ if (!buffer)
++ return;
++
++ for (i = 0; i < TPMIF_TX_RING_SIZE && offset < received; i++) {
++ struct tx_buffer *txb = tp->tx_buffers[i];
++ tpmif_tx_request_t *tx;
++ unsigned int tocopy;
++
++ tx = &tp->tx->ring[i].req;
++ tocopy = tx->size;
++ if (tocopy > PAGE_SIZE)
++ tocopy = PAGE_SIZE;
++
++ memcpy(&buffer[offset], txb->data, tocopy);
++
++ gnttab_release_grant_reference(&gref_head, tx->ref);
++
++ offset += tocopy;
++ }
++
++ vtpm_vd_recv(tp->chip, buffer, received, tp->tx_remember);
++ kfree(buffer);
++}
++
++
++static irqreturn_t tpmif_int(int irq, void *tpm_priv)
++{
++ struct tpm_private *tp = tpm_priv;
++ unsigned long flags;
++
++ spin_lock_irqsave(&tp->tx_lock, flags);
++ tpmif_rx_tasklet.data = (unsigned long)tp;
++ tasklet_schedule(&tpmif_rx_tasklet);
++ spin_unlock_irqrestore(&tp->tx_lock, flags);
++
++ return IRQ_HANDLED;
++}
++
++
++static int tpm_xmit(struct tpm_private *tp,
++ const u8 * buf, size_t count, int isuserbuffer,
++ void *remember)
++{
++ tpmif_tx_request_t *tx;
++ TPMIF_RING_IDX i;
++ unsigned int offset = 0;
++
++ spin_lock_irq(&tp->tx_lock);
++
++ if (unlikely(atomic_read(&tp->tx_busy))) {
++ printk("tpm_xmit: There's an outstanding request/response "
++ "on the way!\n");
++ spin_unlock_irq(&tp->tx_lock);
++ return -EBUSY;
++ }
++
++ if (tp->is_connected != 1) {
++ spin_unlock_irq(&tp->tx_lock);
++ return -EIO;
++ }
++
++ for (i = 0; count > 0 && i < TPMIF_TX_RING_SIZE; i++) {
++ struct tx_buffer *txb = tp->tx_buffers[i];
++ int copied;
++
++ if (!txb) {
++ DPRINTK("txb (i=%d) is NULL. buffers initilized?\n"
++ "Not transmitting anything!\n", i);
++ spin_unlock_irq(&tp->tx_lock);
++ return -EFAULT;
++ }
++
++ copied = tx_buffer_copy(txb, &buf[offset], count,
++ isuserbuffer);
++ if (copied < 0) {
++ /* An error occurred */
++ spin_unlock_irq(&tp->tx_lock);
++ return copied;
++ }
++ count -= copied;
++ offset += copied;
++
++ tx = &tp->tx->ring[i].req;
++ tx->addr = virt_to_machine(txb->data);
++ tx->size = txb->len;
++ tx->unused = 0;
++
++ DPRINTK("First 4 characters sent by TPM-FE are "
++ "0x%02x 0x%02x 0x%02x 0x%02x\n",
++ txb->data[0],txb->data[1],txb->data[2],txb->data[3]);
++
++ /* Get the granttable reference for this page. */
++ tx->ref = gnttab_claim_grant_reference(&gref_head);
++ if (tx->ref == -ENOSPC) {
++ spin_unlock_irq(&tp->tx_lock);
++ DPRINTK("Grant table claim reference failed in "
++ "func:%s line:%d file:%s\n",
++ __FUNCTION__, __LINE__, __FILE__);
++ return -ENOSPC;
++ }
++ gnttab_grant_foreign_access_ref(tx->ref,
++ tp->backend_id,
++ virt_to_mfn(txb->data),
++ 0 /*RW*/);
++ wmb();
++ }
++
++ atomic_set(&tp->tx_busy, 1);
++ tp->tx_remember = remember;
++
++ mb();
++
++ notify_remote_via_irq(tp->irq);
++
++ spin_unlock_irq(&tp->tx_lock);
++ return offset;
++}
++
++
++static void tpmif_notify_upperlayer(struct tpm_private *tp)
++{
++ /* Notify upper layer about the state of the connection to the BE. */
++ vtpm_vd_status(tp->chip, (tp->is_connected
++ ? TPM_VD_STATUS_CONNECTED
++ : TPM_VD_STATUS_DISCONNECTED));
++}
++
++
++static void tpmif_set_connected_state(struct tpm_private *tp, u8 is_connected)
++{
++ /*
++ * Don't notify upper layer if we are in suspend mode and
++ * should disconnect - assumption is that we will resume
++ * The mutex keeps apps from sending.
++ */
++ if (is_connected == 0 && tp->is_suspended == 1)
++ return;
++
++ /*
++ * Unlock the mutex if we are connected again
++ * after being suspended - now resuming.
++ * This also removes the suspend state.
++ */
++ if (is_connected == 1 && tp->is_suspended == 1)
++ tpmfront_suspend_finish(tp);
++
++ if (is_connected != tp->is_connected) {
++ tp->is_connected = is_connected;
++ tpmif_notify_upperlayer(tp);
++ }
++}
++
++
++
++/* =================================================================
++ * Initialization function.
++ * =================================================================
++ */
++
++
++static int __init tpmif_init(void)
++{
++ struct tpm_private *tp;
++
++ if (is_initial_xendomain())
++ return -EPERM;
++
++ tp = tpm_private_get();
++ if (!tp)
++ return -ENOMEM;
++
++ IPRINTK("Initialising the vTPM driver.\n");
++ if (gnttab_alloc_grant_references(TPMIF_TX_RING_SIZE,
++ &gref_head) < 0) {
++ tpm_private_put();
++ return -EFAULT;
++ }
++
++ init_tpm_xenbus();
++ return 0;
++}
++
++
++module_init(tpmif_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/char/tty_io.c b/drivers/char/tty_io.c
+index 7501310..cae44ce 100644
+--- a/drivers/char/tty_io.c
++++ b/drivers/char/tty_io.c
+@@ -135,6 +135,8 @@ LIST_HEAD(tty_drivers); /* linked list of tty drivers */
+ DEFINE_MUTEX(tty_mutex);
+ EXPORT_SYMBOL(tty_mutex);
+
++int console_use_vt = 1;
++
+ #ifdef CONFIG_UNIX98_PTYS
+ extern struct tty_driver *ptm_driver; /* Unix98 pty masters; for /dev/ptmx */
+ static int ptmx_open(struct inode *, struct file *);
+@@ -2696,7 +2698,7 @@ retry_open:
+ goto got_driver;
+ }
+ #ifdef CONFIG_VT
+- if (device == MKDEV(TTY_MAJOR, 0)) {
++ if (console_use_vt && device == MKDEV(TTY_MAJOR, 0)) {
+ extern struct tty_driver *console_driver;
+ driver = console_driver;
+ index = fg_console;
+@@ -4191,6 +4193,8 @@ static int __init tty_init(void)
+ #endif
+
+ #ifdef CONFIG_VT
++ if (!console_use_vt)
++ goto out_vt;
+ cdev_init(&vc0_cdev, &console_fops);
+ if (cdev_add(&vc0_cdev, MKDEV(TTY_MAJOR, 0), 1) ||
+ register_chrdev_region(MKDEV(TTY_MAJOR, 0), 1, "/dev/vc/0") < 0)
+@@ -4198,6 +4202,7 @@ static int __init tty_init(void)
+ device_create(tty_class, NULL, MKDEV(TTY_MAJOR, 0), "tty0");
+
+ vty_init();
++ out_vt:
+ #endif
+ return 0;
+ }
+diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig
+index 5f076ae..abf7e90 100644
+--- a/drivers/cpufreq/Kconfig
++++ b/drivers/cpufreq/Kconfig
+@@ -1,5 +1,6 @@
+ config CPU_FREQ
+ bool "CPU Frequency scaling"
++ depends on !PROCESSOR_EXTERNAL_CONTROL
+ help
+ CPU Frequency scaling allows you to change the clock speed of
+ CPUs on the fly. This is a nice method to save power, because
+diff --git a/drivers/cpuidle/Kconfig b/drivers/cpuidle/Kconfig
+index 7dbc4a8..5f1f297 100644
+--- a/drivers/cpuidle/Kconfig
++++ b/drivers/cpuidle/Kconfig
+@@ -1,6 +1,7 @@
+
+ config CPU_IDLE
+ bool "CPU idle PM support"
++ depends on !PROCESSOR_EXTERNAL_CONTROL
+ default ACPI
+ help
+ CPU idle is a generic framework for supporting software-controlled
+diff --git a/drivers/ide/ide-lib.c b/drivers/ide/ide-lib.c
+index 47af80d..c80eb78 100644
+--- a/drivers/ide/ide-lib.c
++++ b/drivers/ide/ide-lib.c
+@@ -328,6 +328,7 @@ void ide_toggle_bounce(ide_drive_t *drive, int on)
+ {
+ u64 addr = BLK_BOUNCE_HIGH; /* dma64_addr_t */
+
++#ifndef CONFIG_XEN
+ if (!PCI_DMA_BUS_IS_PHYS) {
+ addr = BLK_BOUNCE_ANY;
+ } else if (on && drive->media == ide_disk) {
+@@ -336,6 +337,16 @@ void ide_toggle_bounce(ide_drive_t *drive, int on)
+ if (dev && dev->dma_mask)
+ addr = *dev->dma_mask;
+ }
++#else
++ if (on && drive->media == ide_disk) {
++ struct device *dev = drive->hwif->dev;
++
++ if (!PCI_DMA_BUS_IS_PHYS)
++ addr = BLK_BOUNCE_ANY;
++ else if (dev && dev->dma_mask)
++ addr = *dev->dma_mask;
++ }
++#endif
+
+ if (drive->queue)
+ blk_queue_bounce_limit(drive->queue, addr);
+diff --git a/drivers/input/xen-kbdfront.c b/drivers/input/xen-kbdfront.c
+index 0f47f46..e4ec9c7 100644
+--- a/drivers/input/xen-kbdfront.c
++++ b/drivers/input/xen-kbdfront.c
+@@ -309,7 +309,6 @@ static struct xenbus_device_id xenkbd_ids[] = {
+
+ static struct xenbus_driver xenkbd = {
+ .name = "vkbd",
+- .owner = THIS_MODULE,
+ .ids = xenkbd_ids,
+ .probe = xenkbd_probe,
+ .remove = xenkbd_remove,
+diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
+index f4182cf..d8fb853 100644
+--- a/drivers/net/Kconfig
++++ b/drivers/net/Kconfig
+@@ -2597,9 +2597,9 @@ source "drivers/atm/Kconfig"
+
+ source "drivers/s390/net/Kconfig"
+
+-config XEN_NETDEV_FRONTEND
++config XEN_NETFRONT
+ tristate "Xen network device frontend driver"
+- depends on XEN
++ depends on PARAVIRT_XEN
+ default y
+ help
+ The network device frontend driver allows the kernel to
+diff --git a/drivers/net/Makefile b/drivers/net/Makefile
+index dcbfe84..31ef67f 100644
+--- a/drivers/net/Makefile
++++ b/drivers/net/Makefile
+@@ -139,7 +139,7 @@ obj-$(CONFIG_PPPOL2TP) += pppox.o pppol2tp.o
+ obj-$(CONFIG_SLIP) += slip.o
+ obj-$(CONFIG_SLHC) += slhc.o
+
+-obj-$(CONFIG_XEN_NETDEV_FRONTEND) += xen-netfront.o
++obj-$(CONFIG_XEN_NETFRONT) += xen-netfront.o
+
+ obj-$(CONFIG_DUMMY) += dummy.o
+ obj-$(CONFIG_IFB) += ifb.o
+diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
+index d26f69b..f4c6989 100644
+--- a/drivers/net/xen-netfront.c
++++ b/drivers/net/xen-netfront.c
+@@ -36,8 +36,6 @@
+ #include <linux/skbuff.h>
+ #include <linux/ethtool.h>
+ #include <linux/if_ether.h>
+-#include <linux/tcp.h>
+-#include <linux/udp.h>
+ #include <linux/moduleparam.h>
+ #include <linux/mm.h>
+ #include <net/ip.h>
+@@ -754,45 +752,6 @@ static RING_IDX xennet_fill_frags(struct netfront_info *np,
+ return cons;
+ }
+
+-static int skb_checksum_setup(struct sk_buff *skb)
+-{
+- struct iphdr *iph;
+- unsigned char *th;
+- int err = -EPROTO;
+-
+- if (skb->protocol != htons(ETH_P_IP))
+- goto out;
+-
+- iph = (void *)skb->data;
+- th = skb->data + 4 * iph->ihl;
+- if (th >= skb_tail_pointer(skb))
+- goto out;
+-
+- skb->csum_start = th - skb->head;
+- switch (iph->protocol) {
+- case IPPROTO_TCP:
+- skb->csum_offset = offsetof(struct tcphdr, check);
+- break;
+- case IPPROTO_UDP:
+- skb->csum_offset = offsetof(struct udphdr, check);
+- break;
+- default:
+- if (net_ratelimit())
+- printk(KERN_ERR "Attempting to checksum a non-"
+- "TCP/UDP packet, dropping a protocol"
+- " %d packet", iph->protocol);
+- goto out;
+- }
+-
+- if ((th + skb->csum_offset + 2) > skb_tail_pointer(skb))
+- goto out;
+-
+- err = 0;
+-
+-out:
+- return err;
+-}
+-
+ static int handle_incoming_queue(struct net_device *dev,
+ struct sk_buff_head *rxq)
+ {
+@@ -1773,7 +1732,6 @@ static int __devexit xennet_remove(struct xenbus_device *dev)
+
+ static struct xenbus_driver netfront = {
+ .name = "vif",
+- .owner = THIS_MODULE,
+ .ids = netfront_ids,
+ .probe = netfront_probe,
+ .remove = __devexit_p(xennet_remove),
+diff --git a/drivers/oprofile/buffer_sync.c b/drivers/oprofile/buffer_sync.c
+index 9304c45..d92f444 100644
+--- a/drivers/oprofile/buffer_sync.c
++++ b/drivers/oprofile/buffer_sync.c
+@@ -6,6 +6,10 @@
+ *
+ * @author John Levon <levon at movementarian.org>
+ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ *
+ * This is the core of the buffer management. Each
+ * CPU buffer is processed and entered into the
+ * global event buffer. Such processing is necessary
+@@ -40,6 +44,9 @@ static cpumask_t marked_cpus = CPU_MASK_NONE;
+ static DEFINE_SPINLOCK(task_mortuary);
+ static void process_task_mortuary(void);
+
++#ifdef CONFIG_XEN
++static int cpu_current_domain[NR_CPUS];
++#endif
+
+ /* Take ownership of the task struct and place it on the
+ * list for processing. Only after two full buffer syncs
+@@ -148,6 +155,13 @@ static void end_sync(void)
+ int sync_start(void)
+ {
+ int err;
++#ifdef CONFIG_XEN
++ int i;
++
++ for (i = 0; i < NR_CPUS; i++) {
++ cpu_current_domain[i] = COORDINATOR_DOMAIN;
++ }
++#endif
+
+ start_cpu_work();
+
+@@ -274,15 +288,33 @@ static void add_cpu_switch(int i)
+ last_cookie = INVALID_COOKIE;
+ }
+
+-static void add_kernel_ctx_switch(unsigned int in_kernel)
++static void add_cpu_mode_switch(unsigned int cpu_mode)
+ {
+ add_event_entry(ESCAPE_CODE);
+- if (in_kernel)
+- add_event_entry(KERNEL_ENTER_SWITCH_CODE);
+- else
+- add_event_entry(KERNEL_EXIT_SWITCH_CODE);
++ switch (cpu_mode) {
++ case CPU_MODE_USER:
++ add_event_entry(USER_ENTER_SWITCH_CODE);
++ break;
++ case CPU_MODE_KERNEL:
++ add_event_entry(KERNEL_ENTER_SWITCH_CODE);
++ break;
++ case CPU_MODE_XEN:
++ add_event_entry(XEN_ENTER_SWITCH_CODE);
++ break;
++ default:
++ break;
++ }
+ }
+-
++
++#ifdef CONFIG_XEN
++static void add_domain_switch(unsigned long domain_id)
++{
++ add_event_entry(ESCAPE_CODE);
++ add_event_entry(DOMAIN_SWITCH_CODE);
++ add_event_entry(domain_id);
++}
++#endif
++
+ static void
+ add_user_ctx_switch(struct task_struct const * task, unsigned long cookie)
+ {
+@@ -347,9 +379,9 @@ static int add_us_sample(struct mm_struct * mm, struct op_sample * s)
+ * for later lookup from userspace.
+ */
+ static int
+-add_sample(struct mm_struct * mm, struct op_sample * s, int in_kernel)
++add_sample(struct mm_struct * mm, struct op_sample * s, int cpu_mode)
+ {
+- if (in_kernel) {
++ if (cpu_mode >= CPU_MODE_KERNEL) {
+ add_sample_entry(s->eip, s->event);
+ return 1;
+ } else if (mm) {
+@@ -495,15 +527,24 @@ void sync_buffer(int cpu)
+ struct mm_struct *mm = NULL;
+ struct task_struct * new;
+ unsigned long cookie = 0;
+- int in_kernel = 1;
++ int cpu_mode = 1;
+ unsigned int i;
+ sync_buffer_state state = sb_buffer_start;
+ unsigned long available;
++ int domain_switch = 0;
+
+ mutex_lock(&buffer_mutex);
+
+ add_cpu_switch(cpu);
+
++#ifdef CONFIG_XEN
++ /* We need to assign the first samples in this CPU buffer to the
++ same domain that we were processing at the last sync_buffer */
++ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
++ add_domain_switch(cpu_current_domain[cpu]);
++ }
++#endif
++
+ /* Remember, only we can modify tail_pos */
+
+ available = get_slots(cpu_buf);
+@@ -511,16 +552,20 @@ void sync_buffer(int cpu)
+ for (i = 0; i < available; ++i) {
+ struct op_sample * s = &cpu_buf->buffer[cpu_buf->tail_pos];
+
+- if (is_code(s->eip)) {
+- if (s->event <= CPU_IS_KERNEL) {
+- /* kernel/userspace switch */
+- in_kernel = s->event;
++ if (is_code(s->eip) && !domain_switch) {
++ if (s->event <= CPU_MODE_XEN) {
++ /* xen/kernel/userspace switch */
++ cpu_mode = s->event;
+ if (state == sb_buffer_start)
+ state = sb_sample_start;
+- add_kernel_ctx_switch(s->event);
++ add_cpu_mode_switch(s->event);
+ } else if (s->event == CPU_TRACE_BEGIN) {
+ state = sb_bt_start;
+ add_trace_begin();
++#ifdef CONFIG_XEN
++ } else if (s->event == CPU_DOMAIN_SWITCH) {
++ domain_switch = 1;
++#endif
+ } else {
+ struct mm_struct * oldmm = mm;
+
+@@ -534,8 +579,18 @@ void sync_buffer(int cpu)
+ add_user_ctx_switch(new, cookie);
+ }
+ } else {
++#ifdef CONFIG_XEN
++ if (domain_switch) {
++ cpu_current_domain[cpu] = s->eip;
++ add_domain_switch(s->eip);
++ domain_switch = 0;
++ } else if (cpu_current_domain[cpu] !=
++ COORDINATOR_DOMAIN) {
++ add_sample_entry(s->eip, s->event);
++ } else
++#endif
+ if (state >= sb_bt_start &&
+- !add_sample(mm, s, in_kernel)) {
++ !add_sample(mm, s, cpu_mode)) {
+ if (state == sb_bt_start) {
+ state = sb_bt_ignore;
+ atomic_inc(&oprofile_stats.bt_lost_no_mapping);
+@@ -547,6 +602,13 @@ void sync_buffer(int cpu)
+ }
+ release_mm(mm);
+
++#ifdef CONFIG_XEN
++ /* We reset domain to COORDINATOR at each CPU switch */
++ if (cpu_current_domain[cpu] != COORDINATOR_DOMAIN) {
++ add_domain_switch(COORDINATOR_DOMAIN);
++ }
++#endif
++
+ mark_done(cpu);
+
+ mutex_unlock(&buffer_mutex);
+diff --git a/drivers/oprofile/cpu_buffer.c b/drivers/oprofile/cpu_buffer.c
+index 2450b3a..0afaee1 100644
+--- a/drivers/oprofile/cpu_buffer.c
++++ b/drivers/oprofile/cpu_buffer.c
+@@ -6,6 +6,10 @@
+ *
+ * @author John Levon <levon at movementarian.org>
+ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ *
+ * Each CPU has a local buffer that stores PC value/event
+ * pairs. We also log context switches when we notice them.
+ * Eventually each CPU's buffer is processed into the global
+@@ -34,6 +38,12 @@ static void wq_sync_buffer(struct work_struct *work);
+ #define DEFAULT_TIMER_EXPIRE (HZ / 10)
+ static int work_enabled;
+
++#ifndef CONFIG_XEN
++#define current_domain COORDINATOR_DOMAIN
++#else
++static int32_t current_domain = COORDINATOR_DOMAIN;
++#endif
++
+ void free_cpu_buffers(void)
+ {
+ int i;
+@@ -57,7 +67,7 @@ int alloc_cpu_buffers(void)
+ goto fail;
+
+ b->last_task = NULL;
+- b->last_is_kernel = -1;
++ b->last_cpu_mode = -1;
+ b->tracing = 0;
+ b->buffer_size = buffer_size;
+ b->tail_pos = 0;
+@@ -115,7 +125,7 @@ void cpu_buffer_reset(struct oprofile_cpu_buffer * cpu_buf)
+ * collected will populate the buffer with proper
+ * values to initialize the buffer
+ */
+- cpu_buf->last_is_kernel = -1;
++ cpu_buf->last_cpu_mode = -1;
+ cpu_buf->last_task = NULL;
+ }
+
+@@ -165,13 +175,13 @@ add_code(struct oprofile_cpu_buffer * buffer, unsigned long value)
+ * because of the head/tail separation of the writer and reader
+ * of the CPU buffer.
+ *
+- * is_kernel is needed because on some architectures you cannot
++ * cpu_mode is needed because on some architectures you cannot
+ * tell if you are in kernel or user space simply by looking at
+- * pc. We tag this in the buffer by generating kernel enter/exit
+- * events whenever is_kernel changes
++ * pc. We tag this in the buffer by generating kernel/user (and xen)
++ * enter events whenever cpu_mode changes
+ */
+ static int log_sample(struct oprofile_cpu_buffer * cpu_buf, unsigned long pc,
+- int is_kernel, unsigned long event)
++ int cpu_mode, unsigned long event)
+ {
+ struct task_struct * task;
+
+@@ -187,18 +197,18 @@ static int log_sample(struct oprofile_cpu_buffer * cpu_buf, unsigned long pc,
+ return 0;
+ }
+
+- is_kernel = !!is_kernel;
+-
+ task = current;
+
+ /* notice a switch from user->kernel or vice versa */
+- if (cpu_buf->last_is_kernel != is_kernel) {
+- cpu_buf->last_is_kernel = is_kernel;
+- add_code(cpu_buf, is_kernel);
++ if (cpu_buf->last_cpu_mode != cpu_mode) {
++ cpu_buf->last_cpu_mode = cpu_mode;
++ add_code(cpu_buf, cpu_mode);
+ }
+-
++
+ /* notice a task switch */
+- if (cpu_buf->last_task != task) {
++ /* if not processing other domain samples */
++ if ((cpu_buf->last_task != task) &&
++ (current_domain == COORDINATOR_DOMAIN)) {
+ cpu_buf->last_task = task;
+ add_code(cpu_buf, (unsigned long)task);
+ }
+@@ -282,6 +292,27 @@ void oprofile_add_trace(unsigned long pc)
+ add_sample(cpu_buf, pc, 0);
+ }
+
++#ifdef CONFIG_XEN
++int oprofile_add_domain_switch(int32_t domain_id)
++{
++ struct oprofile_cpu_buffer *cpu_buf = &__get_cpu_var(cpu_buffer);
++
++ /* should have space for switching into and out of domain
++ (2 slots each) plus one sample and one cpu mode switch */
++ if (((nr_available_slots(cpu_buf) < 6) &&
++ (domain_id != COORDINATOR_DOMAIN)) ||
++ (nr_available_slots(cpu_buf) < 2))
++ return 0;
++
++ add_code(cpu_buf, CPU_DOMAIN_SWITCH);
++ add_sample(cpu_buf, domain_id, 0);
++
++ current_domain = domain_id;
++
++ return 1;
++}
++#endif
++
+ /*
+ * This serves to avoid cpu buffer overflow, and makes sure
+ * the task mortuary progresses
+diff --git a/drivers/oprofile/cpu_buffer.h b/drivers/oprofile/cpu_buffer.h
+index c3e366b..d4c03d9 100644
+--- a/drivers/oprofile/cpu_buffer.h
++++ b/drivers/oprofile/cpu_buffer.h
+@@ -37,7 +37,7 @@ struct oprofile_cpu_buffer {
+ volatile unsigned long tail_pos;
+ unsigned long buffer_size;
+ struct task_struct * last_task;
+- int last_is_kernel;
++ int last_cpu_mode;
+ int tracing;
+ struct op_sample * buffer;
+ unsigned long sample_received;
+@@ -53,7 +53,10 @@ DECLARE_PER_CPU(struct oprofile_cpu_buffer, cpu_buffer);
+ void cpu_buffer_reset(struct oprofile_cpu_buffer * cpu_buf);
+
+ /* transient events for the CPU buffer -> event buffer */
+-#define CPU_IS_KERNEL 1
+-#define CPU_TRACE_BEGIN 2
++#define CPU_MODE_USER 0
++#define CPU_MODE_KERNEL 1
++#define CPU_MODE_XEN 2
++#define CPU_TRACE_BEGIN 3
++#define CPU_DOMAIN_SWITCH 4
+
+ #endif /* OPROFILE_CPU_BUFFER_H */
+diff --git a/drivers/oprofile/event_buffer.h b/drivers/oprofile/event_buffer.h
+index 5076ed1..494a3a5 100644
+--- a/drivers/oprofile/event_buffer.h
++++ b/drivers/oprofile/event_buffer.h
+@@ -23,6 +23,9 @@ void wake_up_buffer_waiter(void);
+ #define INVALID_COOKIE ~0UL
+ #define NO_COOKIE 0UL
+
++/* Constant used to refer to coordinator domain (Xen) */
++#define COORDINATOR_DOMAIN -1
++
+ extern const struct file_operations event_buffer_fops;
+
+ /* mutex between sync_cpu_buffers() and the
+diff --git a/drivers/oprofile/oprof.c b/drivers/oprofile/oprof.c
+index 2c64517..fd56e7b 100644
+--- a/drivers/oprofile/oprof.c
++++ b/drivers/oprofile/oprof.c
+@@ -5,6 +5,10 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ */
+
+ #include <linux/kernel.h>
+@@ -33,6 +37,34 @@ static DEFINE_MUTEX(start_mutex);
+ */
+ static int timer = 0;
+
++#ifdef CONFIG_XEN
++int oprofile_set_active(int active_domains[], unsigned int adomains)
++{
++ int err;
++
++ if (!oprofile_ops.set_active)
++ return -EINVAL;
++
++ mutex_lock(&start_mutex);
++ err = oprofile_ops.set_active(active_domains, adomains);
++ mutex_unlock(&start_mutex);
++ return err;
++}
++
++int oprofile_set_passive(int passive_domains[], unsigned int pdomains)
++{
++ int err;
++
++ if (!oprofile_ops.set_passive)
++ return -EINVAL;
++
++ mutex_lock(&start_mutex);
++ err = oprofile_ops.set_passive(passive_domains, pdomains);
++ mutex_unlock(&start_mutex);
++ return err;
++}
++#endif
++
+ int oprofile_setup(void)
+ {
+ int err;
+diff --git a/drivers/oprofile/oprof.h b/drivers/oprofile/oprof.h
+index 1832365..4ae4ff9 100644
+--- a/drivers/oprofile/oprof.h
++++ b/drivers/oprofile/oprof.h
+@@ -35,5 +35,8 @@ void oprofile_create_files(struct super_block * sb, struct dentry * root);
+ void oprofile_timer_init(struct oprofile_operations * ops);
+
+ int oprofile_set_backtrace(unsigned long depth);
++
++int oprofile_set_active(int active_domains[], unsigned int adomains);
++int oprofile_set_passive(int passive_domains[], unsigned int pdomains);
+
+ #endif /* OPROF_H */
+diff --git a/drivers/oprofile/oprofile_files.c b/drivers/oprofile/oprofile_files.c
+index ef953ba..fcbdff8 100644
+--- a/drivers/oprofile/oprofile_files.c
++++ b/drivers/oprofile/oprofile_files.c
+@@ -5,15 +5,21 @@
+ * @remark Read the file COPYING
+ *
+ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
+ */
+
+ #include <linux/fs.h>
+ #include <linux/oprofile.h>
++#include <asm/uaccess.h>
++#include <linux/ctype.h>
+
+ #include "event_buffer.h"
+ #include "oprofile_stats.h"
+ #include "oprof.h"
+-
++
+ unsigned long fs_buffer_size = 131072;
+ unsigned long fs_cpu_buffer_size = 8192;
+ unsigned long fs_buffer_watershed = 32768; /* FIXME: tune */
+@@ -117,11 +123,208 @@ static ssize_t dump_write(struct file * file, char const __user * buf, size_t co
+ static const struct file_operations dump_fops = {
+ .write = dump_write,
+ };
+-
++
++#ifdef CONFIG_XEN
++
++#define TMPBUFSIZE 512
++
++static unsigned int adomains = 0;
++static int active_domains[MAX_OPROF_DOMAINS + 1];
++static DEFINE_MUTEX(adom_mutex);
++
++static ssize_t adomain_write(struct file * file, char const __user * buf,
++ size_t count, loff_t * offset)
++{
++ char *tmpbuf;
++ char *startp, *endp;
++ int i;
++ unsigned long val;
++ ssize_t retval = count;
++
++ if (*offset)
++ return -EINVAL;
++ if (count > TMPBUFSIZE - 1)
++ return -EINVAL;
++
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
++
++ if (copy_from_user(tmpbuf, buf, count)) {
++ kfree(tmpbuf);
++ return -EFAULT;
++ }
++ tmpbuf[count] = 0;
++
++ mutex_lock(&adom_mutex);
++
++ startp = tmpbuf;
++ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
++ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
++ val = simple_strtoul(startp, &endp, 0);
++ if (endp == startp)
++ break;
++ while (ispunct(*endp) || isspace(*endp))
++ endp++;
++ active_domains[i] = val;
++ if (active_domains[i] != val)
++ /* Overflow, force error below */
++ i = MAX_OPROF_DOMAINS + 1;
++ startp = endp;
++ }
++ /* Force error on trailing junk */
++ adomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
++
++ kfree(tmpbuf);
++
++ if (adomains > MAX_OPROF_DOMAINS
++ || oprofile_set_active(active_domains, adomains)) {
++ adomains = 0;
++ retval = -EINVAL;
++ }
++
++ mutex_unlock(&adom_mutex);
++ return retval;
++}
++
++static ssize_t adomain_read(struct file * file, char __user * buf,
++ size_t count, loff_t * offset)
++{
++ char * tmpbuf;
++ size_t len;
++ int i;
++ ssize_t retval;
++
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
++
++ mutex_lock(&adom_mutex);
++
++ len = 0;
++ for (i = 0; i < adomains; i++)
++ len += snprintf(tmpbuf + len,
++ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
++ "%u ", active_domains[i]);
++ WARN_ON(len > TMPBUFSIZE);
++ if (len != 0 && len <= TMPBUFSIZE)
++ tmpbuf[len-1] = '\n';
++
++ mutex_unlock(&adom_mutex);
++
++ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
++
++ kfree(tmpbuf);
++ return retval;
++}
++
++
++static struct file_operations active_domain_ops = {
++ .read = adomain_read,
++ .write = adomain_write,
++};
++
++static unsigned int pdomains = 0;
++static int passive_domains[MAX_OPROF_DOMAINS];
++static DEFINE_MUTEX(pdom_mutex);
++
++static ssize_t pdomain_write(struct file * file, char const __user * buf,
++ size_t count, loff_t * offset)
++{
++ char *tmpbuf;
++ char *startp, *endp;
++ int i;
++ unsigned long val;
++ ssize_t retval = count;
++
++ if (*offset)
++ return -EINVAL;
++ if (count > TMPBUFSIZE - 1)
++ return -EINVAL;
++
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
++
++ if (copy_from_user(tmpbuf, buf, count)) {
++ kfree(tmpbuf);
++ return -EFAULT;
++ }
++ tmpbuf[count] = 0;
++
++ mutex_lock(&pdom_mutex);
++
++ startp = tmpbuf;
++ /* Parse one more than MAX_OPROF_DOMAINS, for easy error checking */
++ for (i = 0; i <= MAX_OPROF_DOMAINS; i++) {
++ val = simple_strtoul(startp, &endp, 0);
++ if (endp == startp)
++ break;
++ while (ispunct(*endp) || isspace(*endp))
++ endp++;
++ passive_domains[i] = val;
++ if (passive_domains[i] != val)
++ /* Overflow, force error below */
++ i = MAX_OPROF_DOMAINS + 1;
++ startp = endp;
++ }
++ /* Force error on trailing junk */
++ pdomains = *startp ? MAX_OPROF_DOMAINS + 1 : i;
++
++ kfree(tmpbuf);
++
++ if (pdomains > MAX_OPROF_DOMAINS
++ || oprofile_set_passive(passive_domains, pdomains)) {
++ pdomains = 0;
++ retval = -EINVAL;
++ }
++
++ mutex_unlock(&pdom_mutex);
++ return retval;
++}
++
++static ssize_t pdomain_read(struct file * file, char __user * buf,
++ size_t count, loff_t * offset)
++{
++ char * tmpbuf;
++ size_t len;
++ int i;
++ ssize_t retval;
++
++ if (!(tmpbuf = kmalloc(TMPBUFSIZE, GFP_KERNEL)))
++ return -ENOMEM;
++
++ mutex_lock(&pdom_mutex);
++
++ len = 0;
++ for (i = 0; i < pdomains; i++)
++ len += snprintf(tmpbuf + len,
++ len < TMPBUFSIZE ? TMPBUFSIZE - len : 0,
++ "%u ", passive_domains[i]);
++ WARN_ON(len > TMPBUFSIZE);
++ if (len != 0 && len <= TMPBUFSIZE)
++ tmpbuf[len-1] = '\n';
++
++ mutex_unlock(&pdom_mutex);
++
++ retval = simple_read_from_buffer(buf, count, offset, tmpbuf, len);
++
++ kfree(tmpbuf);
++ return retval;
++}
++
++static struct file_operations passive_domain_ops = {
++ .read = pdomain_read,
++ .write = pdomain_write,
++};
++
++#endif /* CONFIG_XEN */
++
+ void oprofile_create_files(struct super_block * sb, struct dentry * root)
+ {
+ oprofilefs_create_file(sb, root, "enable", &enable_fops);
+ oprofilefs_create_file_perm(sb, root, "dump", &dump_fops, 0666);
++#ifdef CONFIG_XEN
++ oprofilefs_create_file(sb, root, "active_domains", &active_domain_ops);
++ oprofilefs_create_file(sb, root, "passive_domains", &passive_domain_ops);
++#endif
+ 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 --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
+index e1ca425..aa5d3fd 100644
+--- a/drivers/pci/Kconfig
++++ b/drivers/pci/Kconfig
+@@ -45,7 +45,7 @@ config PCI_DEBUG
+ 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 --git a/drivers/pci/bus.c b/drivers/pci/bus.c
+index 529d9d7..0fd6e10 100644
+--- a/drivers/pci/bus.c
++++ b/drivers/pci/bus.c
+@@ -17,6 +17,8 @@
+
+ #include "pci.h"
+
++extern int pci_mem_align;
++
+ /**
+ * pci_bus_alloc_resource - allocate a resource from a parent bus
+ * @bus: PCI bus
+@@ -44,6 +46,13 @@ pci_bus_alloc_resource(struct pci_bus *bus, struct resource *res,
+
+ type_mask |= IORESOURCE_IO | IORESOURCE_MEM;
+
++#ifdef CONFIG_XEN
++ /* If the boot parameter 'pci-mem-align' was specified then we need to
++ align the memory addresses, at page size alignment. */
++ if (pci_mem_align && (align < (PAGE_SIZE-1)))
++ align = PAGE_SIZE - 1;
++#endif
++
+ for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
+ struct resource *r = bus->resource[i];
+ if (!r)
+diff --git a/drivers/pci/msi-xen.c b/drivers/pci/msi-xen.c
+new file mode 100644
+index 0000000..373dd98
+--- /dev/null
++++ b/drivers/pci/msi-xen.c
+@@ -0,0 +1,719 @@
++/*
++ * File: msi.c
++ * Purpose: PCI Message Signaled Interrupt (MSI)
++ *
++ * Copyright (C) 2003-2004 Intel
++ * Copyright (C) Tom Long Nguyen (tom.l.nguyen at intel.com)
++ */
++
++#include <linux/err.h>
++#include <linux/mm.h>
++#include <linux/irq.h>
++#include <linux/interrupt.h>
++#include <linux/init.h>
++#include <linux/ioport.h>
++#include <linux/pci.h>
++#include <linux/proc_fs.h>
++#include <linux/msi.h>
++#include <linux/smp.h>
++
++#include <asm/errno.h>
++#include <asm/io.h>
++
++#include "pci.h"
++#include "msi.h"
++
++static int pci_msi_enable = 1;
++
++static LIST_HEAD(msi_dev_head);
++DEFINE_SPINLOCK(msi_dev_lock);
++
++struct msi_dev_list {
++ struct pci_dev *dev;
++ struct list_head list;
++ spinlock_t pirq_list_lock;
++ struct list_head pirq_list_head;
++};
++
++struct msi_pirq_entry {
++ struct list_head list;
++ int pirq;
++ int entry_nr;
++};
++
++/* Arch hooks */
++
++int __attribute__ ((weak))
++arch_msi_check_device(struct pci_dev *dev, int nvec, int type)
++{
++ return 0;
++}
++
++#ifndef CONFIG_XEN
++int __attribute__ ((weak))
++arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *entry)
++{
++ return 0;
++}
++
++int __attribute__ ((weak))
++arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
++{
++ struct msi_desc *entry;
++ int ret;
++
++ list_for_each_entry(entry, &dev->msi_list, list) {
++ ret = arch_setup_msi_irq(dev, entry);
++ if (ret)
++ return ret;
++ }
++
++ return 0;
++}
++
++void __attribute__ ((weak)) arch_teardown_msi_irq(unsigned int irq)
++{
++ return;
++}
++
++void __attribute__ ((weak))
++arch_teardown_msi_irqs(struct pci_dev *dev)
++{
++ struct msi_desc *entry;
++
++ list_for_each_entry(entry, &dev->msi_list, list) {
++ if (entry->irq != 0)
++ arch_teardown_msi_irq(entry->irq);
++ }
++}
++#endif
++
++static void msi_set_enable(struct pci_dev *dev, int enable)
++{
++ int pos;
++ u16 control;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ if (pos) {
++ pci_read_config_word(dev, pos + PCI_MSI_FLAGS, &control);
++ control &= ~PCI_MSI_FLAGS_ENABLE;
++ if (enable)
++ control |= PCI_MSI_FLAGS_ENABLE;
++ pci_write_config_word(dev, pos + PCI_MSI_FLAGS, control);
++ }
++}
++
++static void msix_set_enable(struct pci_dev *dev, int enable)
++{
++ int pos;
++ u16 control;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ if (pos) {
++ pci_read_config_word(dev, pos + PCI_MSIX_FLAGS, &control);
++ control &= ~PCI_MSIX_FLAGS_ENABLE;
++ if (enable)
++ control |= PCI_MSIX_FLAGS_ENABLE;
++ pci_write_config_word(dev, pos + PCI_MSIX_FLAGS, control);
++ }
++}
++
++static struct msi_dev_list *get_msi_dev_pirq_list(struct pci_dev *dev)
++{
++ struct msi_dev_list *msi_dev_list, *ret = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&msi_dev_lock, flags);
++
++ list_for_each_entry(msi_dev_list, &msi_dev_head, list)
++ if ( msi_dev_list->dev == dev )
++ ret = msi_dev_list;
++
++ if ( ret ) {
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return ret;
++ }
++
++ /* Has not allocate msi_dev until now. */
++ ret = kmalloc(sizeof(struct msi_dev_list), GFP_ATOMIC);
++
++ /* Failed to allocate msi_dev structure */
++ if ( !ret ) {
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return NULL;
++ }
++
++ spin_lock_init(&ret->pirq_list_lock);
++ INIT_LIST_HEAD(&ret->pirq_list_head);
++ list_add_tail(&ret->list, &msi_dev_head);
++ spin_unlock_irqrestore(&msi_dev_lock, flags);
++ return ret;
++}
++
++static int attach_pirq_entry(int pirq, int entry_nr,
++ struct msi_dev_list *msi_dev_entry)
++{
++ struct msi_pirq_entry *entry = kmalloc(sizeof(*entry), GFP_ATOMIC);
++ unsigned long flags;
++
++ if (!entry)
++ return -ENOMEM;
++ entry->pirq = pirq;
++ entry->entry_nr = entry_nr;
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_add_tail(&entry->list, &msi_dev_entry->pirq_list_head);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ return 0;
++}
++
++static void detach_pirq_entry(int entry_nr,
++ struct msi_dev_list *msi_dev_entry)
++{
++ unsigned long flags;
++ struct msi_pirq_entry *pirq_entry;
++
++ list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
++ if (pirq_entry->entry_nr == entry_nr) {
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_del(&pirq_entry->list);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ kfree(pirq_entry);
++ return;
++ }
++ }
++}
++
++/*
++ * pciback will provide device's owner
++ */
++static int (*get_owner)(struct pci_dev *dev);
++
++int register_msi_get_owner(int (*func)(struct pci_dev *dev))
++{
++ if (get_owner) {
++ printk(KERN_WARNING "register msi_get_owner again\n");
++ return -EEXIST;
++ }
++ get_owner = func;
++ return 0;
++}
++EXPORT_SYMBOL(register_msi_get_owner);
++
++int unregister_msi_get_owner(int (*func)(struct pci_dev *dev))
++{
++ if (get_owner != func)
++ return -EINVAL;
++ get_owner = NULL;
++ return 0;
++}
++EXPORT_SYMBOL(unregister_msi_get_owner);
++
++static int msi_get_dev_owner(struct pci_dev *dev)
++{
++ int owner;
++
++ BUG_ON(!is_initial_xendomain());
++ if (get_owner && (owner = get_owner(dev)) >= 0) {
++ printk(KERN_INFO "get owner for dev %x get %x \n",
++ dev->devfn, owner);
++ return owner;
++ }
++
++ return DOMID_SELF;
++}
++
++static int msi_unmap_pirq(struct pci_dev *dev, int pirq)
++{
++ struct physdev_unmap_pirq unmap;
++ int rc;
++
++ unmap.domid = msi_get_dev_owner(dev);
++ unmap.pirq = pirq;
++
++ if ((rc = HYPERVISOR_physdev_op(PHYSDEVOP_unmap_pirq, &unmap)))
++ printk(KERN_WARNING "unmap irq %x failed\n", pirq);
++
++ if (rc < 0)
++ return rc;
++ return 0;
++}
++
++/*
++ * Protected by msi_lock
++ */
++static int msi_map_pirq_to_vector(struct pci_dev *dev, int pirq,
++ int entry_nr, int msi)
++{
++ struct physdev_map_pirq map_irq;
++ int rc;
++ domid_t domid = DOMID_SELF;
++
++ domid = msi_get_dev_owner(dev);
++
++ map_irq.domid = domid;
++ map_irq.type = MAP_PIRQ_TYPE_MSI;
++ map_irq.index = -1;
++ map_irq.pirq = pirq;
++ map_irq.msi_info.bus = dev->bus->number;
++ map_irq.msi_info.devfn = dev->devfn;
++ map_irq.msi_info.entry_nr = entry_nr;
++ map_irq.msi_info.msi = msi;
++
++ if ((rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq)))
++ printk(KERN_WARNING "map irq failed\n");
++
++ if (rc < 0)
++ return rc;
++
++ return map_irq.pirq;
++}
++
++static int msi_map_vector(struct pci_dev *dev, int entry_nr, int msi)
++{
++ return msi_map_pirq_to_vector(dev, -1, entry_nr, msi);
++}
++
++static void pci_intx_for_msi(struct pci_dev *dev, int enable)
++{
++ if (!(dev->dev_flags & PCI_DEV_FLAGS_MSI_INTX_DISABLE_BUG))
++ pci_intx(dev, enable);
++}
++
++static void __pci_restore_msi_state(struct pci_dev *dev)
++{
++ int pirq;
++
++ if (!dev->msi_enabled)
++ return;
++
++ pirq = msi_map_pirq_to_vector(dev, dev->irq, 0, 1);
++ if (pirq < 0)
++ return;
++
++ pci_intx_for_msi(dev, 0);
++ msi_set_enable(dev, 0);
++}
++
++static void __pci_restore_msix_state(struct pci_dev *dev)
++{
++ unsigned long flags;
++ struct msi_dev_list *msi_dev_entry;
++ struct msi_pirq_entry *pirq_entry, *tmp;
++
++ if (!dev->msix_enabled)
++ return;
++
++ msi_dev_entry = get_msi_dev_pirq_list(dev);
++
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ list_for_each_entry_safe(pirq_entry, tmp,
++ &msi_dev_entry->pirq_list_head, list)
++ msi_map_pirq_to_vector(dev, pirq_entry->pirq, pirq_entry->entry_nr, 0);
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++
++ pci_intx_for_msi(dev, 0);
++ msix_set_enable(dev, 0);
++}
++
++void pci_restore_msi_state(struct pci_dev *dev)
++{
++ __pci_restore_msi_state(dev);
++ __pci_restore_msix_state(dev);
++}
++EXPORT_SYMBOL_GPL(pci_restore_msi_state);
++
++/**
++ * msi_capability_init - configure device's MSI capability structure
++ * @dev: pointer to the pci_dev data structure of MSI device function
++ *
++ * Setup the MSI capability structure of device function with a single
++ * MSI irq, regardless of device function is capable of handling
++ * multiple messages. A return of zero indicates the successful setup
++ * of an entry zero with the new MSI irq or non-zero for otherwise.
++ **/
++static int msi_capability_init(struct pci_dev *dev)
++{
++ int pos, pirq;
++ u16 control;
++
++ msi_set_enable(dev, 0); /* Ensure msi is disabled as I set it up */
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++
++ pirq = msi_map_vector(dev, 0, 1);
++ if (pirq < 0)
++ return -EBUSY;
++
++ /* Set MSI enabled bits */
++ pci_intx_for_msi(dev, 0);
++ msi_set_enable(dev, 1);
++ dev->msi_enabled = 1;
++
++ dev->irq = pirq;
++ return 0;
++}
++
++/**
++ * msix_capability_init - configure device's MSI-X capability
++ * @dev: pointer to the pci_dev data structure of MSI-X device function
++ * @entries: pointer to an array of struct msix_entry entries
++ * @nvec: number of @entries
++ *
++ * Setup the MSI-X capability structure of device function with a
++ * single MSI-X irq. A return of zero indicates the successful setup of
++ * requested MSI-X entries with allocated irqs or non-zero for otherwise.
++ **/
++static int msix_capability_init(struct pci_dev *dev,
++ struct msix_entry *entries, int nvec)
++{
++ int pirq, i, j, mapped;
++ struct msi_dev_list *msi_dev_entry = get_msi_dev_pirq_list(dev);
++ struct msi_pirq_entry *pirq_entry;
++
++ if (!msi_dev_entry)
++ return -ENOMEM;
++
++ msix_set_enable(dev, 0);/* Ensure msix is disabled as I set it up */
++
++ /* MSI-X Table Initialization */
++ for (i = 0; i < nvec; i++) {
++ mapped = 0;
++ list_for_each_entry(pirq_entry, &msi_dev_entry->pirq_list_head, list) {
++ if (pirq_entry->entry_nr == entries[i].entry) {
++ printk(KERN_WARNING "msix entry %d for dev %02x:%02x:%01x are \
++ not freed before acquire again.\n", entries[i].entry,
++ dev->bus->number, PCI_SLOT(dev->devfn),
++ PCI_FUNC(dev->devfn));
++ (entries + i)->vector = pirq_entry->pirq;
++ mapped = 1;
++ break;
++ }
++ }
++ if (mapped)
++ continue;
++ pirq = msi_map_vector(dev, entries[i].entry, 0);
++ if (pirq < 0)
++ break;
++ attach_pirq_entry(pirq, entries[i].entry, msi_dev_entry);
++ (entries + i)->vector = pirq;
++ }
++
++ if (i != nvec) {
++ int avail = i - 1;
++ for (j = --i; j >= 0; j--) {
++ msi_unmap_pirq(dev, entries[j].vector);
++ detach_pirq_entry(entries[j].entry, msi_dev_entry);
++ entries[j].vector = 0;
++ }
++ /* If we had some success report the number of irqs
++ * we succeeded in setting up.
++ */
++ if (avail <= 0)
++ avail = -EBUSY;
++ return avail;
++ }
++
++ pci_intx_for_msi(dev, 0);
++ msix_set_enable(dev, 1);
++ dev->msix_enabled = 1;
++
++ return 0;
++}
++
++/**
++ * pci_msi_check_device - check whether MSI may be enabled on a device
++ * @dev: pointer to the pci_dev data structure of MSI device function
++ * @nvec: how many MSIs have been requested ?
++ * @type: are we checking for MSI or MSI-X ?
++ *
++ * Look at global flags, the device itself, and its parent busses
++ * to determine if MSI/-X are supported for the device. If MSI/-X is
++ * supported return 0, else return an error code.
++ **/
++static int pci_msi_check_device(struct pci_dev* dev, int nvec, int type)
++{
++ struct pci_bus *bus;
++ int ret;
++
++ /* MSI must be globally enabled and supported by the device */
++ if (!pci_msi_enable || !dev || dev->no_msi)
++ return -EINVAL;
++
++ /*
++ * You can't ask to have 0 or less MSIs configured.
++ * a) it's stupid ..
++ * b) the list manipulation code assumes nvec >= 1.
++ */
++ if (nvec < 1)
++ return -ERANGE;
++
++ /* Any bridge which does NOT route MSI transactions from it's
++ * secondary bus to it's primary bus must set NO_MSI flag on
++ * the secondary pci_bus.
++ * We expect only arch-specific PCI host bus controller driver
++ * or quirks for specific PCI bridges to be setting NO_MSI.
++ */
++ for (bus = dev->bus; bus; bus = bus->parent)
++ if (bus->bus_flags & PCI_BUS_FLAGS_NO_MSI)
++ return -EINVAL;
++
++ ret = arch_msi_check_device(dev, nvec, type);
++ if (ret)
++ return ret;
++
++ if (!pci_find_capability(dev, type))
++ return -EINVAL;
++
++ return 0;
++}
++
++/**
++ * pci_enable_msi - configure device's MSI capability structure
++ * @dev: pointer to the pci_dev data structure of MSI device function
++ *
++ * Setup the MSI capability structure of device function with
++ * a single MSI irq upon its software driver call to request for
++ * MSI mode enabled on its hardware device function. A return of zero
++ * indicates the successful setup of an entry zero with the new MSI
++ * vector or non-zero for otherwise.
++ **/
++extern int pci_frontend_enable_msi(struct pci_dev *dev);
++int pci_enable_msi(struct pci_dev* dev)
++{
++ struct pci_bus *bus;
++ int temp, status;
++
++ for (bus = dev->bus; bus; bus = bus->parent)
++ if (bus->bus_flags & PCI_BUS_FLAGS_NO_MSI)
++ return -EINVAL;
++
++ status = pci_msi_check_device(dev, 1, PCI_CAP_ID_MSI);
++ if (status)
++ return status;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain())
++ {
++ int ret;
++
++ temp = dev->irq;
++ ret = pci_frontend_enable_msi(dev);
++ if (ret)
++ return ret;
++
++ dev->irq_old = temp;
++
++ return ret;
++ }
++#endif
++
++ temp = dev->irq;
++
++ /* Check whether driver already requested for MSI-X irqs */
++ if (dev->msix_enabled) {
++ printk(KERN_INFO "PCI: %s: Can't enable MSI. "
++ "Device already has MSI-X enabled\n",
++ pci_name(dev));
++ return -EINVAL;
++ }
++
++ status = msi_capability_init(dev);
++ if ( !status )
++ dev->irq_old = temp;
++
++ return status;
++}
++EXPORT_SYMBOL(pci_enable_msi);
++
++extern void pci_frontend_disable_msi(struct pci_dev* dev);
++void pci_msi_shutdown(struct pci_dev* dev)
++{
++ int pirq;
++
++ if (!pci_msi_enable || !dev)
++ return;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ pci_frontend_disable_msi(dev);
++ dev->irq = dev->irq_old;
++ return;
++ }
++#endif
++
++ if (!dev->msi_enabled)
++ return;
++
++ pirq = dev->irq;
++ /* Restore dev->irq to its default pin-assertion vector */
++ dev->irq = dev->irq_old;
++ msi_unmap_pirq(dev, pirq);
++
++ /* Disable MSI mode */
++ msi_set_enable(dev, 0);
++ pci_intx_for_msi(dev, 1);
++ dev->msi_enabled = 0;
++}
++void pci_disable_msi(struct pci_dev* dev)
++{
++ pci_msi_shutdown(dev);
++}
++EXPORT_SYMBOL(pci_disable_msi);
++
++/**
++ * pci_enable_msix - configure device's MSI-X capability structure
++ * @dev: pointer to the pci_dev data structure of MSI-X device function
++ * @entries: pointer to an array of MSI-X entries
++ * @nvec: number of MSI-X irqs requested for allocation by device driver
++ *
++ * Setup the MSI-X capability structure of device function with the number
++ * of requested irqs upon its software driver call to request for
++ * MSI-X mode enabled on its hardware device function. A return of zero
++ * indicates the successful configuration of MSI-X capability structure
++ * with new allocated MSI-X irqs. A return of < 0 indicates a failure.
++ * Or a return of > 0 indicates that driver request is exceeding the number
++ * of irqs available. Driver should use the returned value to re-send
++ * its request.
++ **/
++extern int pci_frontend_enable_msix(struct pci_dev *dev,
++ struct msix_entry *entries, int nvec);
++int pci_enable_msix(struct pci_dev* dev, struct msix_entry *entries, int nvec)
++{
++ int status, pos, nr_entries;
++ int i, j, temp;
++ u16 control;
++
++ if (!entries)
++ return -EINVAL;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ int ret;
++
++ ret = pci_frontend_enable_msix(dev, entries, nvec);
++ if (ret) {
++ printk("get %x from pci_frontend_enable_msix\n", ret);
++ return ret;
++ }
++
++ return 0;
++ }
++#endif
++
++ status = pci_msi_check_device(dev, nvec, PCI_CAP_ID_MSIX);
++ if (status)
++ return status;
++
++ pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
++ pci_read_config_word(dev, msi_control_reg(pos), &control);
++ nr_entries = multi_msix_capable(control);
++ if (nvec > nr_entries)
++ return -EINVAL;
++
++ /* Check for any invalid entries */
++ for (i = 0; i < nvec; i++) {
++ if (entries[i].entry >= nr_entries)
++ return -EINVAL; /* invalid entry */
++ for (j = i + 1; j < nvec; j++) {
++ if (entries[i].entry == entries[j].entry)
++ return -EINVAL; /* duplicate entry */
++ }
++ }
++
++ temp = dev->irq;
++ /* Check whether driver already requested for MSI vector */
++ if (dev->msi_enabled) {
++ printk(KERN_INFO "PCI: %s: Can't enable MSI-X. "
++ "Device already has an MSI irq assigned\n",
++ pci_name(dev));
++ return -EINVAL;
++ }
++
++ status = msix_capability_init(dev, entries, nvec);
++
++ if ( !status )
++ dev->irq_old = temp;
++
++ return status;
++}
++EXPORT_SYMBOL(pci_enable_msix);
++
++extern void pci_frontend_disable_msix(struct pci_dev* dev);
++void pci_msix_shutdown(struct pci_dev* dev)
++{
++ if (!pci_msi_enable)
++ return;
++ if (!dev)
++ return;
++
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ if (!is_initial_xendomain()) {
++ pci_frontend_disable_msix(dev);
++ dev->irq = dev->irq_old;
++ return;
++ }
++#endif
++
++ if (!dev->msix_enabled)
++ return;
++
++ msi_remove_pci_irq_vectors(dev);
++
++ /* Disable MSI mode */
++ msix_set_enable(dev, 0);
++ pci_intx_for_msi(dev, 1);
++ dev->msix_enabled = 0;
++}
++void pci_disable_msix(struct pci_dev* dev)
++{
++ pci_msix_shutdown(dev);
++}
++EXPORT_SYMBOL(pci_disable_msix);
++
++/**
++ * msi_remove_pci_irq_vectors - reclaim MSI(X) irqs to unused state
++ * @dev: pointer to the pci_dev data structure of MSI(X) device function
++ *
++ * Being called during hotplug remove, from which the device function
++ * is hot-removed. All previous assigned MSI/MSI-X irqs, if
++ * allocated for this device function, are reclaimed to unused state,
++ * which may be used later on.
++ **/
++void msi_remove_pci_irq_vectors(struct pci_dev* dev)
++{
++ unsigned long flags;
++ struct msi_dev_list *msi_dev_entry;
++ struct msi_pirq_entry *pirq_entry, *tmp;
++
++ if (!pci_msi_enable || !dev)
++ return;
++
++ msi_dev_entry = get_msi_dev_pirq_list(dev);
++
++ spin_lock_irqsave(&msi_dev_entry->pirq_list_lock, flags);
++ if (!list_empty(&msi_dev_entry->pirq_list_head))
++ {
++ printk(KERN_WARNING "msix pirqs for dev %02x:%02x:%01x are not freed \
++ before acquire again.\n", dev->bus->number, PCI_SLOT(dev->devfn),
++ PCI_FUNC(dev->devfn));
++ list_for_each_entry_safe(pirq_entry, tmp,
++ &msi_dev_entry->pirq_list_head, list) {
++ msi_unmap_pirq(dev, pirq_entry->pirq);
++ list_del(&pirq_entry->list);
++ kfree(pirq_entry);
++ }
++ }
++ spin_unlock_irqrestore(&msi_dev_entry->pirq_list_lock, flags);
++ dev->irq = dev->irq_old;
++}
++
++void pci_no_msi(void)
++{
++ pci_msi_enable = 0;
++}
++
++void pci_msi_init_pci_dev(struct pci_dev *dev)
++{
++ INIT_LIST_HEAD(&dev->msi_list);
++}
+diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
+index e4548ab..f507d4a 100644
+--- a/drivers/pci/pci.c
++++ b/drivers/pci/pci.c
+@@ -354,7 +354,12 @@ pci_find_parent_resource(const struct pci_dev *dev, struct resource *res)
+ * Restore the BAR values for a given device, so as to make it
+ * accessible by its driver.
+ */
++#ifndef CONFIG_XEN
+ static void
++#else
++EXPORT_SYMBOL_GPL(pci_restore_bars);
++void
++#endif
+ pci_restore_bars(struct pci_dev *dev)
+ {
+ int i, numres;
+diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
+index c14de8e..0b4d559 100644
+--- a/drivers/pci/quirks.c
++++ b/drivers/pci/quirks.c
+@@ -24,6 +24,42 @@
+ #include <linux/kallsyms.h>
+ #include "pci.h"
+
++#ifdef CONFIG_XEN
++/* A global flag which signals if we should page-align PCI mem windows. */
++int pci_mem_align = 0;
++
++static int __init set_pci_mem_align(char *str)
++{
++ pci_mem_align = 1;
++ return 1;
++}
++__setup("pci-mem-align", set_pci_mem_align);
++
++/* This quirk function enables us to force all memory resources which are
++ * assigned to PCI devices, to be page-aligned.
++ */
++static void __devinit quirk_align_mem_resources(struct pci_dev *dev)
++{
++ int i;
++ struct resource *r;
++ resource_size_t old_start;
++
++ if (!pci_mem_align)
++ return;
++
++ for (i=0; i < DEVICE_COUNT_RESOURCE; i++) {
++ r = &dev->resource[i];
++ if ((r == NULL) || !(r->flags & IORESOURCE_MEM))
++ continue;
++
++ old_start = r->start;
++ r->start = (r->start + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1);
++ r->end = r->end - (old_start - r->start);
++ }
++}
++DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, quirk_align_mem_resources);
++#endif
++
+ /* The Mellanox Tavor device gives false positive parity errors
+ * Mark this device with a broken_parity_status, to allow
+ * PCI scanning code to "skip" this now blacklisted device.
+diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
+index 9bc4276..5d4ac28 100644
+--- a/drivers/serial/Kconfig
++++ b/drivers/serial/Kconfig
+@@ -12,6 +12,7 @@ menu "Serial drivers"
+ config SERIAL_8250
+ tristate "8250/16550 and compatible serial support"
+ depends on (BROKEN || !SPARC)
++ depends on !XEN_DISABLE_SERIAL
+ select SERIAL_CORE
+ ---help---
+ This selects whether you want to include the driver for the standard
+diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig
+index e0c5f96..1ccdae9 100644
+--- a/drivers/video/Kconfig
++++ b/drivers/video/Kconfig
+@@ -1583,7 +1583,7 @@ config FB_CYBLA
+ tristate "Cyberblade/i1 support"
+ depends on FB && PCI && X86_32 && !64BIT
+ select FB_CFB_IMAGEBLIT
+- select VIDEO_SELECT
++ select VIDEO_SELECT if !XEN
+ ---help---
+ This driver is supposed to support the Trident Cyberblade/i1
+ graphics core integrated in the VIA VT8601A North Bridge,
+@@ -1976,7 +1976,7 @@ config FB_VIRTUAL
+
+ config XEN_FBDEV_FRONTEND
+ tristate "Xen virtual frame buffer support"
+- depends on FB && XEN
++ depends on FB && PARAVIRT_XEN
+ select FB_SYS_FILLRECT
+ select FB_SYS_COPYAREA
+ select FB_SYS_IMAGEBLIT
+diff --git a/drivers/video/console/Kconfig b/drivers/video/console/Kconfig
+index 06f87b0..54376c9 100644
+--- a/drivers/video/console/Kconfig
++++ b/drivers/video/console/Kconfig
+@@ -46,6 +46,7 @@ config VGACON_SOFT_SCROLLBACK_SIZE
+ config VIDEO_SELECT
+ bool "Video mode selection support"
+ depends on X86 && VGA_CONSOLE
++ depends on !XEN
+ ---help---
+ This enables support for text mode selection on kernel startup. If
+ you want to take advantage of some high-resolution text mode your
+diff --git a/drivers/video/xen-fbfront.c b/drivers/video/xen-fbfront.c
+index 619a6f8..695c800 100644
+--- a/drivers/video/xen-fbfront.c
++++ b/drivers/video/xen-fbfront.c
+@@ -519,7 +519,6 @@ static struct xenbus_device_id xenfb_ids[] = {
+
+ static struct xenbus_driver xenfb = {
+ .name = "vfb",
+- .owner = THIS_MODULE,
+ .ids = xenfb_ids,
+ .probe = xenfb_probe,
+ .remove = xenfb_remove,
+diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
+index 4b75a16..4a524a2 100644
+--- a/drivers/xen/Kconfig
++++ b/drivers/xen/Kconfig
+@@ -1,6 +1,307 @@
++#
++# This Kconfig describe xen options
++#
++
++config XEN
++ bool
++
++if XEN
++config XEN_INTERFACE_VERSION
++ hex
++ default 0x00030207
++
++menu "XEN"
++
++config XEN_PRIVILEGED_GUEST
++ bool "Privileged Guest (domain 0)"
++ help
++ Support for privileged operation (domain 0)
++
++config XEN_UNPRIVILEGED_GUEST
++ def_bool !XEN_PRIVILEGED_GUEST
++
++config XEN_PRIVCMD
++ def_bool y
++ depends on PROC_FS
++
++config XEN_XENBUS_DEV
++ def_bool y
++ depends on PROC_FS
++
++config XEN_NETDEV_ACCEL_SFC_UTIL
++ depends on X86
++ tristate
++
++config XEN_BACKEND
++ tristate "Backend driver support"
++ default XEN_PRIVILEGED_GUEST
++ help
++ Support for backend device drivers that provide I/O services
++ to other virtual machines.
++
++config XEN_BLKDEV_BACKEND
++ tristate "Block-device backend driver"
++ depends on XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The block-device backend driver allows the kernel to export its
++ block devices to other guests via a high-performance shared-memory
++ interface.
++
++config XEN_BLKDEV_TAP
++ tristate "Block-device tap backend driver"
++ depends on XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The block tap driver is an alternative to the block back driver
++ and allows VM block requests to be redirected to userspace through
++ a device interface. The tap allows user-space development of
++ high-performance block backends, where disk images may be implemented
++ as files, in memory, or on other hosts across the network. This
++ driver can safely coexist with the existing blockback driver.
++
++config XEN_NETDEV_BACKEND
++ tristate "Network-device backend driver"
++ depends on XEN_BACKEND && NET
++ default XEN_BACKEND
++ help
++ The network-device backend driver allows the kernel to export its
++ network devices to other guests via a high-performance shared-memory
++ interface.
++
++config XEN_NETDEV_PIPELINED_TRANSMITTER
++ bool "Pipelined transmitter (DANGEROUS)"
++ depends on XEN_NETDEV_BACKEND
++ help
++ If the net backend is a dumb domain, such as a transparent Ethernet
++ bridge with no local IP interface, it is safe to say Y here to get
++ slightly lower network overhead.
++ If the backend has a local IP interface; or may be doing smart things
++ like reassembling packets to perform firewall filtering; or if you
++ are unsure; or if you experience network hangs when this option is
++ enabled; then you must say N here.
++
++config XEN_NETDEV_ACCEL_SFC_BACKEND
++ tristate "Network-device backend driver acceleration for Solarflare NICs"
++ depends on XEN_NETDEV_BACKEND && SFC && SFC_RESOURCE && X86
++ select XEN_NETDEV_ACCEL_SFC_UTIL
++ default m
++
++config XEN_NETDEV_LOOPBACK
++ tristate "Network-device loopback driver"
++ depends on XEN_NETDEV_BACKEND
++ help
++ A two-interface loopback device to emulate a local netfront-netback
++ connection. If unsure, it is probably safe to say N here.
++
++config XEN_PCIDEV_BACKEND
++ tristate "PCI-device backend driver"
++ depends on PCI && XEN_BACKEND
++ default XEN_BACKEND
++ help
++ The PCI device backend driver allows the kernel to export arbitrary
++ PCI devices to other guests. If you select this to be a module, you
++ will need to make sure no other driver has bound to the device(s)
++ you want to make visible to other guests.
++
++choice
++ prompt "PCI Backend Mode"
++ depends on XEN_PCIDEV_BACKEND
++ default XEN_PCIDEV_BACKEND_VPCI if !IA64
++ default XEN_PCIDEV_BACKEND_CONTROLLER if IA64
++
++config XEN_PCIDEV_BACKEND_VPCI
++ bool "Virtual PCI"
++ ---help---
++ This PCI Backend hides the true PCI topology and makes the frontend
++ think there is a single PCI bus with only the exported devices on it.
++ For example, a device at 03:05.0 will be re-assigned to 00:00.0. A
++ second device at 02:1a.1 will be re-assigned to 00:01.1.
++
++config XEN_PCIDEV_BACKEND_PASS
++ bool "Passthrough"
++ ---help---
++ This PCI Backend provides a real view of the PCI topology to the
++ frontend (for example, a device at 06:01.b will still appear at
++ 06:01.b to the frontend). This is similar to how Xen 2.0.x exposed
++ PCI devices to its driver domains. This may be required for drivers
++ which depend on finding their hardward in certain bus/slot
++ locations.
++
++config XEN_PCIDEV_BACKEND_SLOT
++ bool "Slot"
++ ---help---
++ This PCI Backend hides the true PCI topology and makes the frontend
++ think there is a single PCI bus with only the exported devices on it.
++ Contrary to the virtual PCI backend, a function becomes a new slot.
++ For example, a device at 03:05.2 will be re-assigned to 00:00.0. A
++ second device at 02:1a.1 will be re-assigned to 00:01.0.
++
++config XEN_PCIDEV_BACKEND_CONTROLLER
++ bool "Controller"
++ depends on IA64
++ ---help---
++ This PCI backend virtualizes the PCI bus topology by providing a
++ virtual bus per PCI root device. Devices which are physically under
++ the same root bus will appear on the same virtual bus. For systems
++ with complex I/O addressing, this is the only backend which supports
++ extended I/O port spaces and MMIO translation offsets. This backend
++ also supports slot virtualization. For example, a device at
++ 0000:01:02.1 will be re-assigned to 0000:00:00.0. A second device
++ at 0000:02:05.0 (behind a P2P bridge on bus 0000:01) will be
++ re-assigned to 0000:00:01.0. A third device at 0000:16:05.0 (under
++ a different PCI root bus) will be re-assigned to 0000:01:00.0.
++
++endchoice
++
++config XEN_PCIDEV_BE_DEBUG
++ bool "PCI Backend Debugging"
++ depends on XEN_PCIDEV_BACKEND
++
++config XEN_TPMDEV_BACKEND
++ tristate "TPM-device backend driver"
++ depends on XEN_BACKEND
++ help
++ The TPM-device backend driver
++
++config XEN_SCSI_BACKEND
++ tristate "SCSI backend driver"
++ depends on SCSI && XEN_BACKEND
++ default m
++ help
++ The SCSI backend driver allows the kernel to export its SCSI Devices
++ to other guests via a high-performance shared-memory interface.
++
++config XEN_BLKDEV_FRONTEND
++ tristate "Block-device frontend driver"
++ default y
++ help
++ The block-device frontend driver allows the kernel to access block
++ devices mounted within another guest OS. Unless you are building a
++ dedicated device-driver domain, or your master control domain
++ (domain 0), then you almost certainly want to say Y here.
++
++config XEN_NETDEV_FRONTEND
++ tristate "Network-device frontend driver"
++ depends on NET
++ default y
++ help
++ The network-device frontend driver allows the kernel to access
++ network interfaces within another guest OS. Unless you are building a
++ dedicated device-driver domain, or your master control domain
++ (domain 0), then you almost certainly want to say Y here.
++
++config XEN_NETDEV_ACCEL_SFC_FRONTEND
++ tristate "Network-device frontend driver acceleration for Solarflare NICs"
++ depends on XEN_NETDEV_FRONTEND && X86
++ select XEN_NETDEV_ACCEL_SFC_UTIL
++ default m
++
++config XEN_SCSI_FRONTEND
++ tristate "SCSI frontend driver"
++ depends on SCSI
++ default m
++ help
++ The SCSI frontend driver allows the kernel to access SCSI Devices
++ within another guest OS.
++
++config XEN_GRANT_DEV
++ tristate "User-space granted page access driver"
++ default XEN_PRIVILEGED_GUEST
++ help
++ Device for accessing (in user-space) pages that have been granted
++ by other domains.
++
++config XEN_FRAMEBUFFER
++ tristate "Framebuffer-device frontend driver"
++ depends on FB
++ select FB_CFB_FILLRECT
++ select FB_CFB_COPYAREA
++ select FB_CFB_IMAGEBLIT
++ default y
++ help
++ The framebuffer-device frontend drivers allows the kernel to create a
++ virtual framebuffer. This framebuffer can be viewed in another
++ domain. Unless this domain has access to a real video card, you
++ probably want to say Y here.
++
++config XEN_KEYBOARD
++ tristate "Keyboard-device frontend driver"
++ depends on XEN_FRAMEBUFFER && INPUT
++ default y
++ help
++ The keyboard-device frontend driver allows the kernel to create a
++ virtual keyboard. This keyboard can then be driven by another
++ domain. If you've said Y to CONFIG_XEN_FRAMEBUFFER, you probably
++ want to say Y here.
++
++config XEN_DISABLE_SERIAL
++ bool "Disable serial port drivers"
++ default y
++ help
++ Disable serial port drivers, allowing the Xen console driver
++ to provide a serial console at ttyS0.
++
++config XEN_SYSFS
++ tristate "Export Xen attributes in sysfs"
++ depends on SYSFS
++ select SYS_HYPERVISOR
++ default y
++ help
++ Xen hypervisor attributes will show up under /sys/hypervisor/.
++
++choice
++ prompt "Xen version compatibility"
++ default XEN_COMPAT_030002_AND_LATER
++
++ config XEN_COMPAT_030002_AND_LATER
++ bool "3.0.2 and later"
++
++ config XEN_COMPAT_030004_AND_LATER
++ bool "3.0.4 and later"
++
++ config XEN_COMPAT_030100_AND_LATER
++ bool "3.1.0 and later"
++
++ config XEN_COMPAT_LATEST_ONLY
++ bool "no compatibility code"
++
++endchoice
++
++config XEN_COMPAT
++ hex
++ default 0xffffff if XEN_COMPAT_LATEST_ONLY
++ default 0x030100 if XEN_COMPAT_030100_AND_LATER
++ default 0x030004 if XEN_COMPAT_030004_AND_LATER
++ default 0x030002 if XEN_COMPAT_030002_AND_LATER
++ default 0
++
++endmenu
++
++config HAVE_IRQ_IGNORE_UNHANDLED
++ def_bool y
++
++config GENERIC_HARDIRQS_NO__DO_IRQ
++ def_bool y
++
++config NO_IDLE_HZ
++ def_bool y
++
++config XEN_SMPBOOT
++ def_bool y
++ depends on SMP && !PPC_XEN
++
++config XEN_XENCOMM
++ bool
++
++config XEN_DEVMEM
++ def_bool y
++
++endif
++
+ config XEN_BALLOON
+- bool "Xen memory balloon driver"
+- depends on XEN
++ bool "Xen memory balloon driver" if PARAVIRT_XEN
++ depends on (XEN && !PPC_XEN) || PARAVIRT_XEN
+ default y
+ help
+ The balloon driver allows the Xen domain to request more memory from
+@@ -8,12 +309,14 @@ config XEN_BALLOON
+ return unneeded memory to the system.
+
+ config XEN_SCRUB_PAGES
+- bool "Scrub pages before returning them to system"
+- depends on XEN_BALLOON
++ bool "Scrub memory before freeing it to Xen"
++ depends on XEN || XEN_BALLOON
+ default y
+ help
+- Scrub pages before returning them to the system for reuse by
+- other domains. This makes sure that any confidential data
+- is not accidentally visible to other domains. Is it more
+- secure, but slightly less efficient.
++ Erase memory contents before freeing it back to Xen's global
++ pool. This ensures that any secrets contained within that
++ memory (e.g., private keys) cannot be found by other guests that
++ may be running on the machine. Most people will want to say Y here.
++ If security is not a concern then you may increase performance by
++ saying N.
+ If in doubt, say yes.
+diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
+index 37af04f..af5d633 100644
+--- a/drivers/xen/Makefile
++++ b/drivers/xen/Makefile
+@@ -1,4 +1,31 @@
+-obj-y += grant-table.o features.o events.o
+-obj-y += xenbus/
+-obj-$(CONFIG_XEN_XENCOMM) += xencomm.o
+-obj-$(CONFIG_XEN_BALLOON) += balloon.o
++obj-$(CONFIG_PARAVIRT_XEN) += grant-table.o features.o events.o
++xen-xencomm-$(CONFIG_PARAVIRT_XEN) := xencomm.o
++xen-balloon-$(CONFIG_PARAVIRT_XEN) := balloon.o
++
++xen-balloon-$(CONFIG_XEN) := balloon/
++obj-$(CONFIG_XEN) += core/
++obj-$(CONFIG_XEN) += console/
++obj-$(CONFIG_XEN) += evtchn/
++obj-y += xenbus/
++obj-$(CONFIG_XEN) += char/
++
++obj-$(CONFIG_XEN) += util.o
++obj-$(CONFIG_XEN_XENCOMM) += $(xen-xencomm-y)
++obj-$(CONFIG_XEN_BALLOON) += $(xen-balloon-y)
++obj-$(CONFIG_XEN_BLKDEV_BACKEND) += blkback/
++obj-$(CONFIG_XEN_BLKDEV_TAP) += blktap/
++obj-$(CONFIG_XEN_NETDEV_BACKEND) += netback/
++obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmback/
++obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += blkfront/
++obj-$(CONFIG_XEN_NETDEV_FRONTEND) += netfront/
++obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback/
++obj-$(CONFIG_XEN_PCIDEV_FRONTEND) += pcifront/
++obj-$(CONFIG_XEN_FRAMEBUFFER) += fbfront/
++obj-$(CONFIG_XEN_KEYBOARD) += fbfront/
++obj-$(CONFIG_XEN_SCSI_BACKEND) += scsiback/
++obj-$(CONFIG_XEN_SCSI_FRONTEND) += scsifront/
++obj-$(CONFIG_XEN_PRIVCMD) += privcmd/
++obj-$(CONFIG_XEN_GRANT_DEV) += gntdev/
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL) += sfc_netutil/
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND) += sfc_netfront/
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND) += sfc_netback/
+diff --git a/drivers/xen/balloon/Makefile b/drivers/xen/balloon/Makefile
+new file mode 100644
+index 0000000..3fc3d0b
+--- /dev/null
++++ b/drivers/xen/balloon/Makefile
+@@ -0,0 +1,2 @@
++
++obj-y := balloon.o sysfs.o
+diff --git a/drivers/xen/balloon/balloon.c b/drivers/xen/balloon/balloon.c
+new file mode 100644
+index 0000000..3aae38b
+--- /dev/null
++++ b/drivers/xen/balloon/balloon.c
+@@ -0,0 +1,724 @@
++/******************************************************************************
++ * balloon.c
++ *
++ * Xen balloon driver - enables returning/claiming memory to/from Xen.
++ *
++ * Copyright (c) 2003, B Dragovic
++ * Copyright (c) 2003-2004, M Williamson, K Fraser
++ * Copyright (c) 2005 Dan M. Smith, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <linux/smp_lock.h>
++#include <linux/pagemap.h>
++#include <linux/bootmem.h>
++#include <linux/highmem.h>
++#include <linux/vmalloc.h>
++#include <linux/mutex.h>
++#include <xen/xen_proc.h>
++#include <asm/hypervisor.h>
++#include <xen/balloon.h>
++#include <xen/interface/memory.h>
++#include <asm/maddr.h>
++#include <asm/page.h>
++#include <asm/pgalloc.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/tlb.h>
++#include <linux/highmem.h>
++#include <linux/list.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#ifdef CONFIG_PROC_FS
++static struct proc_dir_entry *balloon_pde;
++#endif
++
++static DEFINE_MUTEX(balloon_mutex);
++
++/*
++ * Protects atomic reservation decrease/increase against concurrent increases.
++ * Also protects non-atomic updates of current_pages and driver_pages, and
++ * balloon lists.
++ */
++DEFINE_SPINLOCK(balloon_lock);
++
++struct balloon_stats balloon_stats;
++
++/* We increase/decrease in batches which fit in a page */
++static unsigned long frame_list[PAGE_SIZE / sizeof(unsigned long)];
++
++/* VM /proc information for memory */
++extern unsigned long totalram_pages;
++
++#if !defined(MODULE) && defined(CONFIG_HIGHMEM)
++extern unsigned long totalhigh_pages;
++#define inc_totalhigh_pages() (totalhigh_pages++)
++#define dec_totalhigh_pages() (totalhigh_pages--)
++#else
++#define inc_totalhigh_pages() ((void)0)
++#define dec_totalhigh_pages() ((void)0)
++#endif
++
++/* List of ballooned pages, threaded through the mem_map array. */
++static LIST_HEAD(ballooned_pages);
++
++/* Main work function, always executed in process context. */
++static void balloon_process(struct work_struct *unused);
++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
++ want the kernel to try too hard since that can trigger the oom killer. */
++#define GFP_BALLOON \
++ (GFP_HIGHUSER|__GFP_NOWARN|__GFP_NORETRY|__GFP_NOMEMALLOC|__GFP_COLD)
++
++#define PAGE_TO_LIST(p) (&(p)->lru)
++#define LIST_TO_PAGE(l) list_entry((l), struct page, lru)
++#define UNLIST_PAGE(p) \
++ do { \
++ list_del(PAGE_TO_LIST(p)); \
++ PAGE_TO_LIST(p)->next = NULL; \
++ PAGE_TO_LIST(p)->prev = NULL; \
++ } while(0)
++
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_mem: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_mem: " fmt, ##args)
++
++/* balloon_append: add the given page to the balloon. */
++static void balloon_append(struct page *page)
++{
++ /* Lowmem is re-populated first, so highmem pages go at list tail. */
++ if (PageHighMem(page)) {
++ list_add_tail(PAGE_TO_LIST(page), &ballooned_pages);
++ bs.balloon_high++;
++ dec_totalhigh_pages();
++ } else {
++ list_add(PAGE_TO_LIST(page), &ballooned_pages);
++ bs.balloon_low++;
++ }
++}
++
++/* balloon_retrieve: rescue a page from the balloon, if it is not empty. */
++static struct page *balloon_retrieve(void)
++{
++ struct page *page;
++
++ if (list_empty(&ballooned_pages))
++ return NULL;
++
++ page = LIST_TO_PAGE(ballooned_pages.next);
++ UNLIST_PAGE(page);
++
++ if (PageHighMem(page)) {
++ bs.balloon_high--;
++ inc_totalhigh_pages();
++ }
++ else
++ bs.balloon_low--;
++
++ return page;
++}
++
++static struct page *balloon_first_page(void)
++{
++ if (list_empty(&ballooned_pages))
++ return NULL;
++ return LIST_TO_PAGE(ballooned_pages.next);
++}
++
++static struct page *balloon_next_page(struct page *page)
++{
++ struct list_head *next = PAGE_TO_LIST(page)->next;
++ if (next == &ballooned_pages)
++ return NULL;
++ return LIST_TO_PAGE(next);
++}
++
++static inline void balloon_free_page(struct page *page)
++{
++#ifndef MODULE
++ if (put_page_testzero(page))
++ free_cold_page(page);
++#else
++ /* free_cold_page() is not being exported. */
++ __free_page(page);
++#endif
++}
++
++static void balloon_alarm(unsigned long unused)
++{
++ schedule_work(&balloon_worker);
++}
++
++static unsigned long current_target(void)
++{
++ unsigned long target = min(bs.target_pages, bs.hard_limit);
++ if (target > (bs.current_pages + bs.balloon_low + bs.balloon_high))
++ target = bs.current_pages + bs.balloon_low + bs.balloon_high;
++ return target;
++}
++
++static unsigned long minimum_target(void)
++{
++#ifndef CONFIG_XEN
++#define max_pfn num_physpages
++#endif
++ unsigned long min_pages, curr_pages = current_target();
++
++#define MB2PAGES(mb) ((mb) << (20 - PAGE_SHIFT))
++ /* Simple continuous piecewiese linear function:
++ * max MiB -> min MiB gradient
++ * 0 0
++ * 16 16
++ * 32 24
++ * 128 72 (1/2)
++ * 512 168 (1/4)
++ * 2048 360 (1/8)
++ * 8192 552 (1/32)
++ * 32768 1320
++ * 131072 4392
++ */
++ if (max_pfn < MB2PAGES(128))
++ min_pages = MB2PAGES(8) + (max_pfn >> 1);
++ else if (max_pfn < MB2PAGES(512))
++ min_pages = MB2PAGES(40) + (max_pfn >> 2);
++ else if (max_pfn < MB2PAGES(2048))
++ min_pages = MB2PAGES(104) + (max_pfn >> 3);
++ else
++ min_pages = MB2PAGES(296) + (max_pfn >> 5);
++#undef MB2PAGES
++
++ /* Don't enforce growth */
++ return min(min_pages, curr_pages);
++#ifndef CONFIG_XEN
++#undef max_pfn
++#endif
++}
++
++static int increase_reservation(unsigned long nr_pages)
++{
++ unsigned long pfn, i, flags;
++ struct page *page;
++ long rc;
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++
++ if (nr_pages > ARRAY_SIZE(frame_list))
++ nr_pages = ARRAY_SIZE(frame_list);
++
++ balloon_lock(flags);
++
++ page = balloon_first_page();
++ for (i = 0; i < nr_pages; i++) {
++ BUG_ON(page == NULL);
++ frame_list[i] = page_to_pfn(page);;
++ page = balloon_next_page(page);
++ }
++
++ set_xen_guest_handle(reservation.extent_start, frame_list);
++ reservation.nr_extents = nr_pages;
++ rc = HYPERVISOR_memory_op(
++ XENMEM_populate_physmap, &reservation);
++ if (rc < nr_pages) {
++ if (rc > 0) {
++ int ret;
++
++ /* We hit the Xen hard limit: reprobe. */
++ reservation.nr_extents = rc;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ BUG_ON(ret != rc);
++ }
++ if (rc >= 0)
++ bs.hard_limit = (bs.current_pages + rc -
++ bs.driver_pages);
++ goto out;
++ }
++
++ for (i = 0; i < nr_pages; i++) {
++ page = balloon_retrieve();
++ BUG_ON(page == NULL);
++
++ pfn = page_to_pfn(page);
++ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap) &&
++ phys_to_machine_mapping_valid(pfn));
++
++ set_phys_to_machine(pfn, frame_list[i]);
++
++#ifdef CONFIG_XEN
++ /* Link back into the page tables if not highmem. */
++ if (pfn < max_low_pfn) {
++ int ret;
++ ret = HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte_ma(frame_list[i], PAGE_KERNEL),
++ 0);
++ BUG_ON(ret);
++ }
++#endif
++
++ /* Relinquish the page back to the allocator. */
++ ClearPageReserved(page);
++ init_page_count(page);
++ balloon_free_page(page);
++ }
++
++ bs.current_pages += nr_pages;
++ totalram_pages = bs.current_pages;
++
++ out:
++ balloon_unlock(flags);
++
++ return 0;
++}
++
++static int decrease_reservation(unsigned long nr_pages)
++{
++ unsigned long pfn, i, flags;
++ struct page *page;
++ void *v;
++ int need_sleep = 0;
++ int ret;
++ struct xen_memory_reservation reservation = {
++ .address_bits = 0,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++
++ if (nr_pages > ARRAY_SIZE(frame_list))
++ nr_pages = ARRAY_SIZE(frame_list);
++
++ for (i = 0; i < nr_pages; i++) {
++ if ((page = alloc_page(GFP_BALLOON)) == NULL) {
++ nr_pages = i;
++ need_sleep = 1;
++ break;
++ }
++
++ pfn = page_to_pfn(page);
++ frame_list[i] = pfn_to_mfn(pfn);
++
++ if (!PageHighMem(page)) {
++ v = phys_to_virt(pfn << PAGE_SHIFT);
++ scrub_pages(v, 1);
++#ifdef CONFIG_XEN
++ ret = HYPERVISOR_update_va_mapping(
++ (unsigned long)v, __pte_ma(0), 0);
++ BUG_ON(ret);
++#endif
++ }
++#ifdef CONFIG_XEN_SCRUB_PAGES
++ else {
++ v = kmap(page);
++ scrub_pages(v, 1);
++ kunmap(page);
++ }
++#endif
++ }
++
++#ifdef CONFIG_XEN
++ /* Ensure that ballooned highmem pages don't have kmaps. */
++ kmap_flush_unused();
++ flush_tlb_all();
++#endif
++
++ balloon_lock(flags);
++
++ /* No more mappings: invalidate P2M and add to balloon. */
++ for (i = 0; i < nr_pages; i++) {
++ pfn = mfn_to_pfn(frame_list[i]);
++ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
++ balloon_append(pfn_to_page(pfn));
++ }
++
++ set_xen_guest_handle(reservation.extent_start, frame_list);
++ reservation.nr_extents = nr_pages;
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
++ BUG_ON(ret != nr_pages);
++
++ bs.current_pages -= nr_pages;
++ totalram_pages = bs.current_pages;
++
++ balloon_unlock(flags);
++
++ return need_sleep;
++}
++
++/*
++ * We avoid multiple worker processes conflicting via the balloon mutex.
++ * We may of course race updates of the target counts (which are protected
++ * by the balloon lock), or with changes to the Xen hard limit, but we will
++ * recover from these in time.
++ */
++static void balloon_process(struct work_struct *unused)
++{
++ int need_sleep = 0;
++ long credit;
++
++ mutex_lock(&balloon_mutex);
++
++ do {
++ credit = current_target() - bs.current_pages;
++ if (credit > 0)
++ need_sleep = (increase_reservation(credit) != 0);
++ if (credit < 0)
++ need_sleep = (decrease_reservation(-credit) != 0);
++
++#ifndef CONFIG_PREEMPT
++ if (need_resched())
++ schedule();
++#endif
++ } while ((credit != 0) && !need_sleep);
++
++ /* Schedule more work if there is some still to be done. */
++ if (current_target() != bs.current_pages)
++ mod_timer(&balloon_timer, jiffies + HZ);
++
++ mutex_unlock(&balloon_mutex);
++}
++
++/* Resets the Xen limit, sets new target, and kicks off processing. */
++void balloon_set_new_target(unsigned long target)
++{
++ /* No need for lock. Not read-modify-write updates. */
++ bs.hard_limit = ~0UL;
++ bs.target_pages = max(target, minimum_target());
++ schedule_work(&balloon_worker);
++}
++
++static struct xenbus_watch target_watch =
++{
++ .node = "memory/target"
++};
++
++/* React to a change in the target key */
++static void watch_target(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ unsigned long long new_target;
++ int err;
++
++ err = xenbus_scanf(XBT_NIL, "memory", "target", "%llu", &new_target);
++ if (err != 1) {
++ /* This is ok (for domain0 at least) - so just return */
++ return;
++ }
++
++ /* The given memory/target value is in KiB, so it needs converting to
++ * pages. PAGE_SHIFT converts bytes to pages, hence PAGE_SHIFT - 10.
++ */
++ balloon_set_new_target(new_target >> (PAGE_SHIFT - 10));
++}
++
++static int balloon_init_watcher(struct notifier_block *notifier,
++ unsigned long event,
++ void *data)
++{
++ int err;
++
++ err = register_xenbus_watch(&target_watch);
++ if (err)
++ printk(KERN_ERR "Failed to set balloon watcher\n");
++
++ return NOTIFY_DONE;
++}
++
++#ifdef CONFIG_PROC_FS
++static int balloon_write(struct file *file, const char __user *buffer,
++ unsigned long count, void *data)
++{
++ char memstring[64], *endchar;
++ unsigned long long target_bytes;
++
++ if (!capable(CAP_SYS_ADMIN))
++ return -EPERM;
++
++ if (count <= 1)
++ return -EBADMSG; /* runt */
++ if (count > sizeof(memstring))
++ return -EFBIG; /* too long */
++
++ if (copy_from_user(memstring, buffer, count))
++ return -EFAULT;
++ memstring[sizeof(memstring)-1] = '\0';
++
++ target_bytes = memparse(memstring, &endchar);
++ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
++
++ return count;
++}
++
++static int balloon_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len;
++
++ len = sprintf(
++ page,
++ "Current allocation: %8lu kB\n"
++ "Requested target: %8lu kB\n"
++ "Low-mem balloon: %8lu kB\n"
++ "High-mem balloon: %8lu kB\n"
++ "Driver pages: %8lu kB\n"
++ "Xen hard limit: ",
++ PAGES2KB(bs.current_pages), PAGES2KB(bs.target_pages),
++ PAGES2KB(bs.balloon_low), PAGES2KB(bs.balloon_high),
++ PAGES2KB(bs.driver_pages));
++
++ if (bs.hard_limit != ~0UL)
++ len += sprintf(page + len, "%8lu kB\n",
++ PAGES2KB(bs.hard_limit));
++ else
++ len += sprintf(page + len, " ??? kB\n");
++
++ *eof = 1;
++ return len;
++}
++#endif
++
++static struct notifier_block xenstore_notifier;
++
++static int __init balloon_init(void)
++{
++#if defined(CONFIG_X86) && defined(CONFIG_XEN)
++ unsigned long pfn;
++ struct page *page;
++#endif
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ IPRINTK("Initialising balloon driver.\n");
++
++#ifdef CONFIG_XEN
++ bs.current_pages = min(xen_start_info->nr_pages, max_pfn);
++ totalram_pages = bs.current_pages;
++#else
++ bs.current_pages = totalram_pages;
++#endif
++ bs.target_pages = bs.current_pages;
++ bs.balloon_low = 0;
++ bs.balloon_high = 0;
++ bs.driver_pages = 0UL;
++ bs.hard_limit = ~0UL;
++
++ init_timer(&balloon_timer);
++ balloon_timer.data = 0;
++ balloon_timer.function = balloon_alarm;
++
++#ifdef CONFIG_PROC_FS
++ if ((balloon_pde = create_xen_proc_entry("balloon", 0644)) == NULL) {
++ WPRINTK("Unable to create /proc/xen/balloon.\n");
++ return -1;
++ }
++
++ balloon_pde->read_proc = balloon_read;
++ balloon_pde->write_proc = balloon_write;
++#endif
++ balloon_sysfs_init();
++
++#if defined(CONFIG_X86) && defined(CONFIG_XEN)
++ /* Initialise the balloon with excess memory space. */
++ for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
++ page = pfn_to_page(pfn);
++ if (!PageReserved(page))
++ balloon_append(page);
++ }
++#endif
++
++ target_watch.callback = watch_target;
++ xenstore_notifier.notifier_call = balloon_init_watcher;
++
++ register_xenstore_notifier(&xenstore_notifier);
++
++ return 0;
++}
++
++subsys_initcall(balloon_init);
++
++static void __exit balloon_exit(void)
++{
++ /* XXX - release balloon here */
++ return;
++}
++
++module_exit(balloon_exit);
++
++void balloon_update_driver_allowance(long delta)
++{
++ unsigned long flags;
++
++ balloon_lock(flags);
++ bs.driver_pages += delta;
++ balloon_unlock(flags);
++}
++
++#ifdef CONFIG_XEN
++static int dealloc_pte_fn(
++ pte_t *pte, struct page *pmd_page, unsigned long addr, void *data)
++{
++ unsigned long mfn = pte_mfn(*pte);
++ int ret;
++ struct xen_memory_reservation reservation = {
++ .nr_extents = 1,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ set_xen_guest_handle(reservation.extent_start, &mfn);
++ set_pte_at(&init_mm, addr, pte, __pte_ma(0));
++ set_phys_to_machine(__pa(addr) >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation, &reservation);
++ BUG_ON(ret != 1);
++ return 0;
++}
++#endif
++
++struct page **alloc_empty_pages_and_pagevec(int nr_pages)
++{
++ unsigned long flags;
++ void *v;
++ struct page *page, **pagevec;
++ int i, ret;
++
++ pagevec = kmalloc(sizeof(page) * nr_pages, GFP_KERNEL);
++ if (pagevec == NULL)
++ return NULL;
++
++ for (i = 0; i < nr_pages; i++) {
++ page = pagevec[i] = alloc_page(GFP_KERNEL|__GFP_COLD);
++ if (page == NULL)
++ goto err;
++
++ v = page_address(page);
++ scrub_pages(v, 1);
++
++ balloon_lock(flags);
++
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ unsigned long gmfn = page_to_pfn(page);
++ struct xen_memory_reservation reservation = {
++ .nr_extents = 1,
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ set_xen_guest_handle(reservation.extent_start, &gmfn);
++ ret = HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation);
++ if (ret == 1)
++ ret = 0; /* success */
++ } else {
++#ifdef CONFIG_XEN
++ ret = apply_to_page_range(&init_mm, (unsigned long)v,
++ PAGE_SIZE, dealloc_pte_fn,
++ NULL);
++#else
++ /* Cannot handle non-auto translate mode. */
++ ret = 1;
++#endif
++ }
++
++ if (ret != 0) {
++ balloon_unlock(flags);
++ balloon_free_page(page);
++ goto err;
++ }
++
++ totalram_pages = --bs.current_pages;
++
++ balloon_unlock(flags);
++ }
++
++ out:
++ schedule_work(&balloon_worker);
++#ifdef CONFIG_XEN
++ flush_tlb_all();
++#endif
++ return pagevec;
++
++ err:
++ balloon_lock(flags);
++ while (--i >= 0)
++ balloon_append(pagevec[i]);
++ balloon_unlock(flags);
++ kfree(pagevec);
++ pagevec = NULL;
++ goto out;
++}
++
++void free_empty_pages_and_pagevec(struct page **pagevec, int nr_pages)
++{
++ unsigned long flags;
++ int i;
++
++ if (pagevec == NULL)
++ return;
++
++ balloon_lock(flags);
++ for (i = 0; i < nr_pages; i++) {
++ BUG_ON(page_count(pagevec[i]) != 1);
++ balloon_append(pagevec[i]);
++ }
++ balloon_unlock(flags);
++
++ kfree(pagevec);
++
++ schedule_work(&balloon_worker);
++}
++
++void balloon_release_driver_page(struct page *page)
++{
++ unsigned long flags;
++
++ balloon_lock(flags);
++ balloon_append(page);
++ bs.driver_pages--;
++ balloon_unlock(flags);
++
++ schedule_work(&balloon_worker);
++}
++
++EXPORT_SYMBOL_GPL(balloon_update_driver_allowance);
++EXPORT_SYMBOL_GPL(alloc_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(free_empty_pages_and_pagevec);
++EXPORT_SYMBOL_GPL(balloon_release_driver_page);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/balloon/common.h b/drivers/xen/balloon/common.h
+new file mode 100644
+index 0000000..4496d21
+--- /dev/null
++++ b/drivers/xen/balloon/common.h
+@@ -0,0 +1,58 @@
++/******************************************************************************
++ * balloon/common.h
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_BALLOON_COMMON_H__
++#define __XEN_BALLOON_COMMON_H__
++
++#define PAGES2KB(_p) ((_p)<<(PAGE_SHIFT-10))
++
++struct balloon_stats {
++ /* We aim for 'current allocation' == 'target allocation'. */
++ unsigned long current_pages;
++ unsigned long target_pages;
++ /* We may hit the hard limit in Xen. If we do then we remember it. */
++ unsigned long hard_limit;
++ /*
++ * Drivers may alter the memory reservation independently, but they
++ * must inform the balloon driver so we avoid hitting the hard limit.
++ */
++ unsigned long driver_pages;
++ /* Number of pages in high- and low-memory balloons. */
++ unsigned long balloon_low;
++ unsigned long balloon_high;
++};
++
++extern struct balloon_stats balloon_stats;
++#define bs balloon_stats
++
++int balloon_sysfs_init(void);
++void balloon_sysfs_exit(void);
++
++void balloon_set_new_target(unsigned long target);
++
++#endif /* __XEN_BALLOON_COMMON_H__ */
+diff --git a/drivers/xen/balloon/sysfs.c b/drivers/xen/balloon/sysfs.c
+new file mode 100644
+index 0000000..e972ec3
+--- /dev/null
++++ b/drivers/xen/balloon/sysfs.c
+@@ -0,0 +1,171 @@
++/******************************************************************************
++ * balloon/sysfs.c
++ *
++ * Xen balloon driver - sysfs interfaces.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/capability.h>
++#include <linux/errno.h>
++#include <linux/stat.h>
++#include <linux/string.h>
++#include <linux/sysdev.h>
++#include <linux/module.h>
++#include "common.h"
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#define BALLOON_CLASS_NAME "xen_memory"
++
++#define BALLOON_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct sys_device *dev, \
++ char *buf) \
++ { \
++ return sprintf(buf, format, ##args); \
++ } \
++ static SYSDEV_ATTR(name, S_IRUGO, show_##name, NULL)
++
++BALLOON_SHOW(current_kb, "%lu\n", PAGES2KB(bs.current_pages));
++BALLOON_SHOW(low_kb, "%lu\n", PAGES2KB(bs.balloon_low));
++BALLOON_SHOW(high_kb, "%lu\n", PAGES2KB(bs.balloon_high));
++BALLOON_SHOW(hard_limit_kb,
++ (bs.hard_limit!=~0UL) ? "%lu\n" : "???\n",
++ (bs.hard_limit!=~0UL) ? PAGES2KB(bs.hard_limit) : 0);
++BALLOON_SHOW(driver_kb, "%lu\n", PAGES2KB(bs.driver_pages));
++
++static ssize_t show_target_kb(struct sys_device *dev, char *buf)
++{
++ return sprintf(buf, "%lu\n", PAGES2KB(bs.target_pages));
++}
++
++static ssize_t store_target_kb(struct sys_device *dev,
++ const char *buf,
++ size_t count)
++{
++ char memstring[64], *endchar;
++ unsigned long long target_bytes;
++
++ if (!capable(CAP_SYS_ADMIN))
++ return -EPERM;
++
++ if (count <= 1)
++ return -EBADMSG; /* runt */
++ if (count > sizeof(memstring))
++ return -EFBIG; /* too long */
++ strcpy(memstring, buf);
++
++ target_bytes = memparse(memstring, &endchar);
++ balloon_set_new_target(target_bytes >> PAGE_SHIFT);
++
++ return count;
++}
++
++static SYSDEV_ATTR(target_kb, S_IRUGO | S_IWUSR,
++ show_target_kb, store_target_kb);
++
++static struct sysdev_attribute *balloon_attrs[] = {
++ &attr_target_kb,
++};
++
++static struct attribute *balloon_info_attrs[] = {
++ &attr_current_kb.attr,
++ &attr_low_kb.attr,
++ &attr_high_kb.attr,
++ &attr_hard_limit_kb.attr,
++ &attr_driver_kb.attr,
++ NULL
++};
++
++static struct attribute_group balloon_info_group = {
++ .name = "info",
++ .attrs = balloon_info_attrs,
++};
++
++static struct sysdev_class balloon_sysdev_class = {
++ .name = BALLOON_CLASS_NAME,
++};
++
++static struct sys_device balloon_sysdev;
++
++static int register_balloon(struct sys_device *sysdev)
++{
++ int i, error;
++
++ error = sysdev_class_register(&balloon_sysdev_class);
++ if (error)
++ return error;
++
++ sysdev->id = 0;
++ sysdev->cls = &balloon_sysdev_class;
++
++ error = sysdev_register(sysdev);
++ if (error) {
++ sysdev_class_unregister(&balloon_sysdev_class);
++ return error;
++ }
++
++ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++) {
++ error = sysdev_create_file(sysdev, balloon_attrs[i]);
++ if (error)
++ goto fail;
++ }
++
++ error = sysfs_create_group(&sysdev->kobj, &balloon_info_group);
++ if (error)
++ goto fail;
++
++ return 0;
++
++ fail:
++ while (--i >= 0)
++ sysdev_remove_file(sysdev, balloon_attrs[i]);
++ sysdev_unregister(sysdev);
++ sysdev_class_unregister(&balloon_sysdev_class);
++ return error;
++}
++
++static void unregister_balloon(struct sys_device *sysdev)
++{
++ int i;
++
++ sysfs_remove_group(&sysdev->kobj, &balloon_info_group);
++ for (i = 0; i < ARRAY_SIZE(balloon_attrs); i++)
++ sysdev_remove_file(sysdev, balloon_attrs[i]);
++ sysdev_unregister(sysdev);
++ sysdev_class_unregister(&balloon_sysdev_class);
++}
++
++int balloon_sysfs_init(void)
++{
++ return register_balloon(&balloon_sysdev);
++}
++
++void balloon_sysfs_exit(void)
++{
++ unregister_balloon(&balloon_sysdev);
++}
+diff --git a/drivers/xen/blkback/Makefile b/drivers/xen/blkback/Makefile
+new file mode 100644
+index 0000000..8bab63d
+--- /dev/null
++++ b/drivers/xen/blkback/Makefile
+@@ -0,0 +1,3 @@
++obj-$(CONFIG_XEN_BLKDEV_BACKEND) := blkbk.o
++
++blkbk-y := blkback.o xenbus.o interface.o vbd.o
+diff --git a/drivers/xen/blkback/blkback.c b/drivers/xen/blkback/blkback.c
+new file mode 100644
+index 0000000..2cae987
+--- /dev/null
++++ b/drivers/xen/blkback/blkback.c
+@@ -0,0 +1,645 @@
++/******************************************************************************
++ * arch/xen/drivers/blkif/backend/main.c
++ *
++ * Back-end of the driver for virtual block devices. This portion of the
++ * driver exports a 'unified' block-device interface that can be accessed
++ * by any operating system that implements a compatible front end. A
++ * reference front-end implementation can be found in:
++ * arch/xen/drivers/blkif/frontend
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Copyright (c) 2005, Christopher Clark
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/freezer.h>
++#include <linux/list.h>
++#include <linux/delay.h>
++#include <xen/balloon.h>
++#include <asm/hypervisor.h>
++#include "common.h"
++
++/*
++ * These are rather arbitrary. They are fairly large because adjacent requests
++ * pulled from a communication ring are quite likely to end up being part of
++ * the same scatter/gather request at the disc.
++ *
++ * ** TRY INCREASING 'blkif_reqs' IF WRITE SPEEDS SEEM TOO LOW **
++ *
++ * This will increase the chances of being able to write whole tracks.
++ * 64 should be enough to keep us competitive with Linux.
++ */
++static int blkif_reqs = 64;
++module_param_named(reqs, blkif_reqs, int, 0);
++MODULE_PARM_DESC(reqs, "Number of blkback requests to allocate");
++
++/* Run-time switchable: /sys/module/blkback/parameters/ */
++static unsigned int log_stats = 0;
++static unsigned int debug_lvl = 0;
++module_param(log_stats, int, 0644);
++module_param(debug_lvl, int, 0644);
++
++/*
++ * Each outstanding request that we've passed to the lower device layers has a
++ * 'pending_req' allocated to it. Each buffer_head that completes decrements
++ * the pendcnt towards zero. When it hits zero, the specified domain has a
++ * response queued for it, with the saved 'id' passed back.
++ */
++typedef struct {
++ blkif_t *blkif;
++ u64 id;
++ int nr_pages;
++ atomic_t pendcnt;
++ unsigned short operation;
++ int status;
++ struct list_head free_list;
++} pending_req_t;
++
++static pending_req_t *pending_reqs;
++static struct list_head pending_free;
++static DEFINE_SPINLOCK(pending_free_lock);
++static DECLARE_WAIT_QUEUE_HEAD(pending_free_wq);
++
++#define BLKBACK_INVALID_HANDLE (~0)
++
++static struct page **pending_pages;
++static grant_handle_t *pending_grant_handles;
++
++static inline int vaddr_pagenr(pending_req_t *req, int seg)
++{
++ return (req - pending_reqs) * BLKIF_MAX_SEGMENTS_PER_REQUEST + seg;
++}
++
++static inline unsigned long vaddr(pending_req_t *req, int seg)
++{
++ unsigned long pfn = page_to_pfn(pending_pages[vaddr_pagenr(req, seg)]);
++ return (unsigned long)pfn_to_kaddr(pfn);
++}
++
++#define pending_handle(_req, _seg) \
++ (pending_grant_handles[vaddr_pagenr(_req, _seg)])
++
++
++static int do_block_io_op(blkif_t *blkif);
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req);
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st);
++
++/******************************************************************
++ * misc small helpers
++ */
++static pending_req_t* alloc_req(void)
++{
++ pending_req_t *req = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
++ }
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return req;
++}
++
++static void free_req(pending_req_t *req)
++{
++ unsigned long flags;
++ int was_empty;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ if (was_empty)
++ wake_up(&pending_free_wq);
++}
++
++static void unplug_queue(blkif_t *blkif)
++{
++ if (blkif->plug == NULL)
++ return;
++ if (blkif->plug->unplug_fn)
++ blkif->plug->unplug_fn(blkif->plug);
++ kobject_put(&blkif->plug->kobj);
++ blkif->plug = NULL;
++}
++
++static void plug_queue(blkif_t *blkif, struct bio *bio)
++{
++ struct request_queue *q = bdev_get_queue(bio->bi_bdev);
++
++ if (q == blkif->plug)
++ return;
++ unplug_queue(blkif);
++ WARN_ON(test_bit(QUEUE_FLAG_DEAD, &q->queue_flags));
++ kobject_get(&q->kobj);
++ blkif->plug = q;
++}
++
++static void fast_flush_area(pending_req_t *req)
++{
++ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ unsigned int i, invcount = 0;
++ grant_handle_t handle;
++ int ret;
++
++ for (i = 0; i < req->nr_pages; i++) {
++ handle = pending_handle(req, i);
++ if (handle == BLKBACK_INVALID_HANDLE)
++ continue;
++ gnttab_set_unmap_op(&unmap[invcount], vaddr(req, i),
++ GNTMAP_host_map, handle);
++ pending_handle(req, i) = BLKBACK_INVALID_HANDLE;
++ invcount++;
++ }
++
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(ret);
++}
++
++/******************************************************************
++ * SCHEDULER FUNCTIONS
++ */
++
++static void print_stats(blkif_t *blkif)
++{
++ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d | br %4d\n",
++ current->comm, blkif->st_oo_req,
++ blkif->st_rd_req, blkif->st_wr_req, blkif->st_br_req);
++ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
++ blkif->st_rd_req = 0;
++ blkif->st_wr_req = 0;
++ blkif->st_oo_req = 0;
++}
++
++int blkif_schedule(void *arg)
++{
++ blkif_t *blkif = arg;
++
++ blkif_get(blkif);
++
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: started\n", current->comm);
++
++ while (!kthread_should_stop()) {
++ if (try_to_freeze())
++ continue;
++
++ wait_event_interruptible(
++ blkif->wq,
++ blkif->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
++
++ blkif->waiting_reqs = 0;
++ smp_mb(); /* clear flag *before* checking for work */
++
++ if (do_block_io_op(blkif))
++ blkif->waiting_reqs = 1;
++ unplug_queue(blkif);
++
++ if (log_stats && time_after(jiffies, blkif->st_print))
++ print_stats(blkif);
++ }
++
++ if (log_stats)
++ print_stats(blkif);
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: exiting\n", current->comm);
++
++ blkif->xenblkd = NULL;
++ blkif_put(blkif);
++
++ return 0;
++}
++
++/******************************************************************
++ * COMPLETION CALLBACK -- Called as bh->b_end_io()
++ */
++
++static void __end_block_io_op(pending_req_t *pending_req, int error)
++{
++ /* An error fails the entire request. */
++ if ((pending_req->operation == BLKIF_OP_WRITE_BARRIER) &&
++ (error == -EOPNOTSUPP)) {
++ DPRINTK("blkback: write barrier op failed, not supported\n");
++ blkback_barrier(XBT_NIL, pending_req->blkif->be, 0);
++ pending_req->status = BLKIF_RSP_EOPNOTSUPP;
++ } else if (error) {
++ DPRINTK("Buffer not up-to-date at end of operation, "
++ "error=%d\n", error);
++ pending_req->status = BLKIF_RSP_ERROR;
++ }
++
++ if (atomic_dec_and_test(&pending_req->pendcnt)) {
++ fast_flush_area(pending_req);
++ make_response(pending_req->blkif, pending_req->id,
++ pending_req->operation, pending_req->status);
++ blkif_put(pending_req->blkif);
++ free_req(pending_req);
++ }
++}
++
++static void end_block_io_op(struct bio *bio, int error)
++{
++ __end_block_io_op(bio->bi_private, error);
++ bio_put(bio);
++}
++
++
++/******************************************************************************
++ * NOTIFICATION FROM GUEST OS.
++ */
++
++static void blkif_notify_work(blkif_t *blkif)
++{
++ blkif->waiting_reqs = 1;
++ wake_up(&blkif->wq);
++}
++
++irqreturn_t blkif_be_int(int irq, void *dev_id)
++{
++ blkif_notify_work(dev_id);
++ return IRQ_HANDLED;
++}
++
++
++
++/******************************************************************
++ * DOWNWARD CALLS -- These interface with the block-device layer proper.
++ */
++
++static int do_block_io_op(blkif_t *blkif)
++{
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ blkif_request_t req;
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int more_to_do = 0;
++
++ rc = blk_rings->common.req_cons;
++ rp = blk_rings->common.sring->req_prod;
++ rmb(); /* Ensure we see queued requests up to 'rp'. */
++
++ while (rc != rp) {
++
++ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc))
++ break;
++
++ pending_req = alloc_req();
++ if (NULL == pending_req) {
++ blkif->st_oo_req++;
++ more_to_do = 1;
++ break;
++ }
++
++ if (kthread_should_stop()) {
++ more_to_do = 1;
++ break;
++ }
++
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc), sizeof(req));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.req_cons = ++rc; /* before make_response() */
++
++ /* Apply all sanity checks to /private copy/ of request. */
++ barrier();
++
++ switch (req.operation) {
++ case BLKIF_OP_READ:
++ blkif->st_rd_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++ case BLKIF_OP_WRITE_BARRIER:
++ blkif->st_br_req++;
++ /* fall through */
++ case BLKIF_OP_WRITE:
++ blkif->st_wr_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++ default:
++ /* A good sign something is wrong: sleep for a while to
++ * avoid excessive CPU consumption by a bad guest. */
++ msleep(1);
++ DPRINTK("error: unknown block io operation [%d]\n",
++ req.operation);
++ make_response(blkif, req.id, req.operation,
++ BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ break;
++ }
++
++ /* Yield point for this unbounded loop. */
++ cond_resched();
++ }
++
++ return more_to_do;
++}
++
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req)
++{
++ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
++ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ struct phys_req preq;
++ struct {
++ unsigned long buf; unsigned int nsec;
++ } seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ unsigned int nseg;
++ struct bio *bio = NULL, *biolist[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ int ret, i, nbio = 0;
++ int operation;
++
++ switch (req->operation) {
++ case BLKIF_OP_READ:
++ operation = READ;
++ break;
++ case BLKIF_OP_WRITE:
++ operation = WRITE;
++ break;
++ case BLKIF_OP_WRITE_BARRIER:
++ operation = WRITE_BARRIER;
++ break;
++ default:
++ operation = 0; /* make gcc happy */
++ BUG();
++ }
++
++ /* Check that number of segments is sane. */
++ nseg = req->nr_segments;
++ if (unlikely(nseg == 0 && operation != WRITE_BARRIER) ||
++ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
++ DPRINTK("Bad number of segments in request (%d)\n", nseg);
++ goto fail_response;
++ }
++
++ preq.dev = req->handle;
++ preq.sector_number = req->sector_number;
++ preq.nr_sects = 0;
++
++ pending_req->blkif = blkif;
++ pending_req->id = req->id;
++ pending_req->operation = req->operation;
++ pending_req->status = BLKIF_RSP_OKAY;
++ pending_req->nr_pages = nseg;
++
++ for (i = 0; i < nseg; i++) {
++ uint32_t flags;
++
++ seg[i].nsec = req->seg[i].last_sect -
++ req->seg[i].first_sect + 1;
++
++ if ((req->seg[i].last_sect >= (PAGE_SIZE >> 9)) ||
++ (req->seg[i].last_sect < req->seg[i].first_sect))
++ goto fail_response;
++ preq.nr_sects += seg[i].nsec;
++
++ flags = GNTMAP_host_map;
++ if (operation != READ)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[i], vaddr(pending_req, i), flags,
++ req->seg[i].gref, blkif->domid);
++ }
++
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, nseg);
++ BUG_ON(ret);
++
++ for (i = 0; i < nseg; i++) {
++ if (unlikely(map[i].status != 0)) {
++ DPRINTK("invalid buffer -- could not remap it\n");
++ map[i].handle = BLKBACK_INVALID_HANDLE;
++ ret |= 1;
++ }
++
++ pending_handle(pending_req, i) = map[i].handle;
++
++ if (ret)
++ continue;
++
++ set_phys_to_machine(__pa(vaddr(
++ pending_req, i)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr >> PAGE_SHIFT));
++ seg[i].buf = map[i].dev_bus_addr |
++ (req->seg[i].first_sect << 9);
++ }
++
++ if (ret)
++ goto fail_flush;
++
++ if (vbd_translate(&preq, blkif, operation) != 0) {
++ DPRINTK("access denied: %s of [%llu,%llu] on dev=%04x\n",
++ operation == READ ? "read" : "write",
++ preq.sector_number,
++ preq.sector_number + preq.nr_sects, preq.dev);
++ goto fail_flush;
++ }
++
++ for (i = 0; i < nseg; i++) {
++ if (((int)preq.sector_number|(int)seg[i].nsec) &
++ ((bdev_hardsect_size(preq.bdev) >> 9) - 1)) {
++ DPRINTK("Misaligned I/O request from domain %d",
++ blkif->domid);
++ goto fail_put_bio;
++ }
++
++ while ((bio == NULL) ||
++ (bio_add_page(bio,
++ virt_to_page(vaddr(pending_req, i)),
++ seg[i].nsec << 9,
++ seg[i].buf & ~PAGE_MASK) == 0)) {
++ bio = biolist[nbio++] = bio_alloc(GFP_KERNEL, nseg-i);
++ if (unlikely(bio == NULL))
++ goto fail_put_bio;
++
++ bio->bi_bdev = preq.bdev;
++ bio->bi_private = pending_req;
++ bio->bi_end_io = end_block_io_op;
++ bio->bi_sector = preq.sector_number;
++ }
++
++ preq.sector_number += seg[i].nsec;
++ }
++
++ if (!bio) {
++ BUG_ON(operation != WRITE_BARRIER);
++ bio = biolist[nbio++] = bio_alloc(GFP_KERNEL, 0);
++ if (unlikely(bio == NULL))
++ goto fail_put_bio;
++
++ bio->bi_bdev = preq.bdev;
++ bio->bi_private = pending_req;
++ bio->bi_end_io = end_block_io_op;
++ bio->bi_sector = -1;
++ }
++
++ plug_queue(blkif, bio);
++ atomic_set(&pending_req->pendcnt, nbio);
++ blkif_get(blkif);
++
++ for (i = 0; i < nbio; i++)
++ submit_bio(operation, biolist[i]);
++
++ if (operation == READ)
++ blkif->st_rd_sect += preq.nr_sects;
++ else if (operation == WRITE || operation == WRITE_BARRIER)
++ blkif->st_wr_sect += preq.nr_sects;
++
++ return;
++
++ fail_put_bio:
++ for (i = 0; i < (nbio-1); i++)
++ bio_put(biolist[i]);
++ fail_flush:
++ fast_flush_area(pending_req);
++ fail_response:
++ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ msleep(1); /* back off a bit */
++}
++
++
++
++/******************************************************************
++ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
++ */
++
++
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st)
++{
++ blkif_response_t resp;
++ unsigned long flags;
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ int more_to_do = 0;
++ int notify;
++
++ resp.id = id;
++ resp.operation = op;
++ resp.status = st;
++
++ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
++ /* Place on the response ring for the relevant domain. */
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(RING_GET_RESPONSE(&blk_rings->native, blk_rings->native.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32, blk_rings->x86_32.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64, blk_rings->x86_64.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.rsp_prod_pvt++;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
++ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
++ /*
++ * Tail check for pending requests. Allows frontend to avoid
++ * notifications if requests are already in flight (lower
++ * overheads and promotes batching).
++ */
++ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
++
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
++ more_to_do = 1;
++ }
++
++ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
++
++ if (more_to_do)
++ blkif_notify_work(blkif);
++ if (notify)
++ notify_remote_via_irq(blkif->irq);
++}
++
++static int __init blkif_init(void)
++{
++ int i, mmap_pages;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ mmap_pages = blkif_reqs * BLKIF_MAX_SEGMENTS_PER_REQUEST;
++
++ pending_reqs = kmalloc(sizeof(pending_reqs[0]) *
++ blkif_reqs, GFP_KERNEL);
++ pending_grant_handles = kmalloc(sizeof(pending_grant_handles[0]) *
++ mmap_pages, GFP_KERNEL);
++ pending_pages = alloc_empty_pages_and_pagevec(mmap_pages);
++
++ if (!pending_reqs || !pending_grant_handles || !pending_pages)
++ goto out_of_memory;
++
++ for (i = 0; i < mmap_pages; i++)
++ pending_grant_handles[i] = BLKBACK_INVALID_HANDLE;
++
++ blkif_interface_init();
++
++ memset(pending_reqs, 0, sizeof(pending_reqs));
++ INIT_LIST_HEAD(&pending_free);
++
++ for (i = 0; i < blkif_reqs; i++)
++ list_add_tail(&pending_reqs[i].free_list, &pending_free);
++
++ blkif_xenbus_init();
++
++ return 0;
++
++ out_of_memory:
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, mmap_pages);
++ printk("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++}
++
++module_init(blkif_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/blkback/common.h b/drivers/xen/blkback/common.h
+new file mode 100644
+index 0000000..018d8d7
+--- /dev/null
++++ b/drivers/xen/blkback/common.h
+@@ -0,0 +1,139 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __BLKIF__BACKEND__COMMON_H__
++#define __BLKIF__BACKEND__COMMON_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/blkdev.h>
++#include <linux/vmalloc.h>
++#include <linux/wait.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/blkif.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++struct vbd {
++ blkif_vdev_t handle; /* what the domain refers to this vbd as */
++ unsigned char readonly; /* Non-zero -> read-only */
++ unsigned char type; /* VDISK_xxx */
++ u32 pdevice; /* phys device that this vbd maps to */
++ struct block_device *bdev;
++};
++
++struct backend_info;
++
++typedef struct blkif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
++ /* Comms information. */
++ enum blkif_protocol blk_protocol;
++ blkif_back_rings_t blk_rings;
++ struct vm_struct *blk_ring_area;
++ /* The VBD attached to this interface. */
++ struct vbd vbd;
++ /* Back pointer to the backend_info. */
++ struct backend_info *be;
++ /* Private fields. */
++ spinlock_t blk_ring_lock;
++ atomic_t refcnt;
++
++ wait_queue_head_t wq;
++ struct task_struct *xenblkd;
++ unsigned int waiting_reqs;
++ struct request_queue *plug;
++
++ /* statistics */
++ unsigned long st_print;
++ int st_rd_req;
++ int st_wr_req;
++ int st_oo_req;
++ int st_br_req;
++ int st_rd_sect;
++ int st_wr_sect;
++
++ wait_queue_head_t waiting_to_free;
++
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++} blkif_t;
++
++blkif_t *blkif_alloc(domid_t domid);
++void blkif_disconnect(blkif_t *blkif);
++void blkif_free(blkif_t *blkif);
++int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn);
++
++#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define blkif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
++
++/* Create a vbd. */
++int vbd_create(blkif_t *blkif, blkif_vdev_t vdevice, unsigned major,
++ unsigned minor, int readonly, int cdrom);
++void vbd_free(struct vbd *vbd);
++
++unsigned long long vbd_size(struct vbd *vbd);
++unsigned int vbd_info(struct vbd *vbd);
++unsigned long vbd_secsize(struct vbd *vbd);
++
++struct phys_req {
++ unsigned short dev;
++ unsigned short nr_sects;
++ struct block_device *bdev;
++ blkif_sector_t sector_number;
++};
++
++int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation);
++
++void blkif_interface_init(void);
++
++void blkif_xenbus_init(void);
++
++irqreturn_t blkif_be_int(int irq, void *dev_id);
++int blkif_schedule(void *arg);
++
++int blkback_barrier(struct xenbus_transaction xbt,
++ struct backend_info *be, int state);
++
++#endif /* __BLKIF__BACKEND__COMMON_H__ */
+diff --git a/drivers/xen/blkback/interface.c b/drivers/xen/blkback/interface.c
+new file mode 100644
+index 0000000..cd38c85
+--- /dev/null
++++ b/drivers/xen/blkback/interface.c
+@@ -0,0 +1,181 @@
++/******************************************************************************
++ * arch/xen/drivers/blkif/backend/interface.c
++ *
++ * Block-device interface management.
++ *
++ * Copyright (c) 2004, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "common.h"
++#include <xen/evtchn.h>
++#include <linux/kthread.h>
++
++static struct kmem_cache *blkif_cachep;
++
++blkif_t *blkif_alloc(domid_t domid)
++{
++ blkif_t *blkif;
++
++ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
++ if (!blkif)
++ return ERR_PTR(-ENOMEM);
++
++ memset(blkif, 0, sizeof(*blkif));
++ blkif->domid = domid;
++ spin_lock_init(&blkif->blk_ring_lock);
++ atomic_set(&blkif->refcnt, 1);
++ init_waitqueue_head(&blkif->wq);
++ blkif->st_print = jiffies;
++ init_waitqueue_head(&blkif->waiting_to_free);
++
++ return blkif;
++}
++
++static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, shared_page, blkif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
++ }
++
++ blkif->shmem_ref = shared_page;
++ blkif->shmem_handle = op.handle;
++
++ return 0;
++}
++
++static void unmap_frontend_page(blkif_t *blkif)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, blkif->shmem_handle);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
++
++int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn)
++{
++ int err;
++
++ /* Already connected through? */
++ if (blkif->irq)
++ return 0;
++
++ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
++ return -ENOMEM;
++
++ err = map_frontend_page(blkif, shared_page);
++ if (err) {
++ free_vm_area(blkif->blk_ring_area);
++ return err;
++ }
++
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ {
++ blkif_sring_t *sring;
++ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_32:
++ {
++ blkif_x86_32_sring_t *sring_x86_32;
++ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_64:
++ {
++ blkif_x86_64_sring_t *sring_x86_64;
++ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
++ break;
++ }
++ default:
++ BUG();
++ }
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ blkif->domid, evtchn, blkif_be_int, 0, "blkif-backend", blkif);
++ if (err < 0)
++ {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ return err;
++ }
++ blkif->irq = err;
++
++ return 0;
++}
++
++void blkif_disconnect(blkif_t *blkif)
++{
++ if (blkif->xenblkd) {
++ kthread_stop(blkif->xenblkd);
++ blkif->xenblkd = NULL;
++ }
++
++ atomic_dec(&blkif->refcnt);
++ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
++ atomic_inc(&blkif->refcnt);
++
++ if (blkif->irq) {
++ unbind_from_irqhandler(blkif->irq, blkif);
++ blkif->irq = 0;
++ }
++
++ if (blkif->blk_rings.common.sring) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ }
++}
++
++void blkif_free(blkif_t *blkif)
++{
++ if (!atomic_dec_and_test(&blkif->refcnt))
++ BUG();
++ kmem_cache_free(blkif_cachep, blkif);
++}
++
++void __init blkif_interface_init(void)
++{
++ blkif_cachep = kmem_cache_create("blkif_cache", sizeof(blkif_t),
++ 0, 0, NULL);
++}
+diff --git a/drivers/xen/blkback/vbd.c b/drivers/xen/blkback/vbd.c
+new file mode 100644
+index 0000000..2594419
+--- /dev/null
++++ b/drivers/xen/blkback/vbd.c
+@@ -0,0 +1,118 @@
++/******************************************************************************
++ * blkback/vbd.c
++ *
++ * Routines for managing virtual block devices (VBDs).
++ *
++ * Copyright (c) 2003-2005, Keir Fraser & Steve Hand
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "common.h"
++
++#define vbd_sz(_v) ((_v)->bdev->bd_part ? \
++ (_v)->bdev->bd_part->nr_sects : (_v)->bdev->bd_disk->capacity)
++
++unsigned long long vbd_size(struct vbd *vbd)
++{
++ return vbd_sz(vbd);
++}
++
++unsigned int vbd_info(struct vbd *vbd)
++{
++ return vbd->type | (vbd->readonly?VDISK_READONLY:0);
++}
++
++unsigned long vbd_secsize(struct vbd *vbd)
++{
++ return bdev_hardsect_size(vbd->bdev);
++}
++
++int vbd_create(blkif_t *blkif, blkif_vdev_t handle, unsigned major,
++ unsigned minor, int readonly, int cdrom)
++{
++ struct vbd *vbd;
++ struct block_device *bdev;
++
++ vbd = &blkif->vbd;
++ vbd->handle = handle;
++ vbd->readonly = readonly;
++ vbd->type = 0;
++
++ vbd->pdevice = MKDEV(major, minor);
++
++ bdev = open_by_devnum(vbd->pdevice,
++ vbd->readonly ? FMODE_READ : FMODE_WRITE);
++
++ if (IS_ERR(bdev)) {
++ DPRINTK("vbd_creat: device %08x could not be opened.\n",
++ vbd->pdevice);
++ return -ENOENT;
++ }
++
++ vbd->bdev = bdev;
++
++ if (vbd->bdev->bd_disk == NULL) {
++ DPRINTK("vbd_creat: device %08x doesn't exist.\n",
++ vbd->pdevice);
++ vbd_free(vbd);
++ return -ENOENT;
++ }
++
++ if (vbd->bdev->bd_disk->flags & GENHD_FL_CD || cdrom)
++ vbd->type |= VDISK_CDROM;
++ if (vbd->bdev->bd_disk->flags & GENHD_FL_REMOVABLE)
++ vbd->type |= VDISK_REMOVABLE;
++
++ DPRINTK("Successful creation of handle=%04x (dom=%u)\n",
++ handle, blkif->domid);
++ return 0;
++}
++
++void vbd_free(struct vbd *vbd)
++{
++ if (vbd->bdev)
++ blkdev_put(vbd->bdev);
++ vbd->bdev = NULL;
++}
++
++int vbd_translate(struct phys_req *req, blkif_t *blkif, int operation)
++{
++ struct vbd *vbd = &blkif->vbd;
++ int rc = -EACCES;
++
++ if ((operation != READ) && vbd->readonly)
++ goto out;
++
++ if (unlikely((req->sector_number + req->nr_sects) > vbd_sz(vbd)))
++ goto out;
++
++ req->dev = vbd->pdevice;
++ req->bdev = vbd->bdev;
++ rc = 0;
++
++ out:
++ return rc;
++}
+diff --git a/drivers/xen/blkback/xenbus.c b/drivers/xen/blkback/xenbus.c
+new file mode 100644
+index 0000000..42288db
+--- /dev/null
++++ b/drivers/xen/blkback/xenbus.c
+@@ -0,0 +1,541 @@
++/* Xenbus code for blkif backend
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ Copyright (C) 2005 XenSource Ltd
++
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
++
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
++
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
++
++#include <stdarg.h>
++#include <linux/module.h>
++#include <linux/kthread.h>
++#include "common.h"
++
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ pr_debug("blkback/xenbus (%s:%d) " fmt ".\n", \
++ __FUNCTION__, __LINE__, ##args)
++
++struct backend_info
++{
++ struct xenbus_device *dev;
++ blkif_t *blkif;
++ struct xenbus_watch backend_watch;
++ unsigned major;
++ unsigned minor;
++ char *mode;
++};
++
++static void connect(struct backend_info *);
++static int connect_ring(struct backend_info *);
++static void backend_changed(struct xenbus_watch *, const char **,
++ unsigned int);
++
++static int blkback_name(blkif_t *blkif, char *buf)
++{
++ char *devpath, *devname;
++ struct xenbus_device *dev = blkif->be->dev;
++
++ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
++ if (IS_ERR(devpath))
++ return PTR_ERR(devpath);
++
++ if ((devname = strstr(devpath, "/dev/")) != NULL)
++ devname += strlen("/dev/");
++ else
++ devname = devpath;
++
++ snprintf(buf, TASK_COMM_LEN, "blkback.%d.%s", blkif->domid, devname);
++ kfree(devpath);
++
++ return 0;
++}
++
++static void update_blkif_status(blkif_t *blkif)
++{
++ int err;
++ char name[TASK_COMM_LEN];
++
++ /* Not ready to connect? */
++ if (!blkif->irq || !blkif->vbd.bdev)
++ return;
++
++ /* Already connected? */
++ if (blkif->be->dev->state == XenbusStateConnected)
++ return;
++
++ /* Attempt to connect: exit if we fail to. */
++ connect(blkif->be);
++ if (blkif->be->dev->state != XenbusStateConnected)
++ return;
++
++ err = blkback_name(blkif, name);
++ if (err) {
++ xenbus_dev_error(blkif->be->dev, err, "get blkback dev name");
++ return;
++ }
++
++ blkif->xenblkd = kthread_run(blkif_schedule, blkif, name);
++ if (IS_ERR(blkif->xenblkd)) {
++ err = PTR_ERR(blkif->xenblkd);
++ blkif->xenblkd = NULL;
++ xenbus_dev_error(blkif->be->dev, err, "start xenblkd");
++ }
++}
++
++
++/****************************************************************
++ * sysfs interface for VBD I/O requests
++ */
++
++#define VBD_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct device *_dev, \
++ struct device_attribute *attr, \
++ char *buf) \
++ { \
++ struct xenbus_device *dev = to_xenbus_device(_dev); \
++ struct backend_info *be = dev->dev.driver_data; \
++ \
++ return sprintf(buf, format, ##args); \
++ } \
++ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
++
++VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
++VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
++VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
++VBD_SHOW(br_req, "%d\n", be->blkif->st_br_req);
++VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
++VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
++
++static struct attribute *vbdstat_attrs[] = {
++ &dev_attr_oo_req.attr,
++ &dev_attr_rd_req.attr,
++ &dev_attr_wr_req.attr,
++ &dev_attr_br_req.attr,
++ &dev_attr_rd_sect.attr,
++ &dev_attr_wr_sect.attr,
++ NULL
++};
++
++static struct attribute_group vbdstat_group = {
++ .name = "statistics",
++ .attrs = vbdstat_attrs,
++};
++
++VBD_SHOW(physical_device, "%x:%x\n", be->major, be->minor);
++VBD_SHOW(mode, "%s\n", be->mode);
++
++int xenvbd_sysfs_addif(struct xenbus_device *dev)
++{
++ int error;
++
++ error = device_create_file(&dev->dev, &dev_attr_physical_device);
++ if (error)
++ goto fail1;
++
++ error = device_create_file(&dev->dev, &dev_attr_mode);
++ if (error)
++ goto fail2;
++
++ error = sysfs_create_group(&dev->dev.kobj, &vbdstat_group);
++ if (error)
++ goto fail3;
++
++ return 0;
++
++fail3: sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
++fail2: device_remove_file(&dev->dev, &dev_attr_mode);
++fail1: device_remove_file(&dev->dev, &dev_attr_physical_device);
++ return error;
++}
++
++void xenvbd_sysfs_delif(struct xenbus_device *dev)
++{
++ sysfs_remove_group(&dev->dev.kobj, &vbdstat_group);
++ device_remove_file(&dev->dev, &dev_attr_mode);
++ device_remove_file(&dev->dev, &dev_attr_physical_device);
++}
++
++static int blkback_remove(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ DPRINTK("");
++
++ if (be->major || be->minor)
++ xenvbd_sysfs_delif(dev);
++
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
++
++ if (be->blkif) {
++ blkif_disconnect(be->blkif);
++ vbd_free(&be->blkif->vbd);
++ blkif_free(be->blkif);
++ be->blkif = NULL;
++ }
++
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
++}
++
++int blkback_barrier(struct xenbus_transaction xbt,
++ struct backend_info *be, int state)
++{
++ struct xenbus_device *dev = be->dev;
++ int err;
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-barrier",
++ "%d", state);
++ if (err)
++ xenbus_dev_fatal(dev, err, "writing feature-barrier");
++
++ return err;
++}
++
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures, and watch the store waiting for the hotplug scripts to tell us
++ * the device's physical major and minor numbers. Switch to InitWait.
++ */
++static int blkback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ be->blkif = blkif_alloc(dev->otherend_id);
++ if (IS_ERR(be->blkif)) {
++ err = PTR_ERR(be->blkif);
++ be->blkif = NULL;
++ xenbus_dev_fatal(dev, err, "creating block interface");
++ goto fail;
++ }
++
++ /* setup back pointer */
++ be->blkif->be = be;
++
++ err = xenbus_watch_path2(dev, dev->nodename, "physical-device",
++ &be->backend_watch, backend_changed);
++ if (err)
++ goto fail;
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++
++ return 0;
++
++fail:
++ DPRINTK("failed");
++ blkback_remove(dev);
++ return err;
++}
++
++
++/**
++ * Callback received when the hotplug scripts have placed the physical-device
++ * node. Read it and the mode node, and create a vbd. If the frontend is
++ * ready, connect.
++ */
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int err;
++ unsigned major;
++ unsigned minor;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
++ int cdrom = 0;
++ char *device_type;
++
++ DPRINTK("");
++
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "physical-device", "%x:%x",
++ &major, &minor);
++ if (XENBUS_EXIST_ERR(err)) {
++ /* Since this watch will fire once immediately after it is
++ registered, we expect this. Ignore it, and wait for the
++ hotplug scripts. */
++ return;
++ }
++ if (err != 2) {
++ xenbus_dev_fatal(dev, err, "reading physical-device");
++ return;
++ }
++
++ if ((be->major || be->minor) &&
++ ((be->major != major) || (be->minor != minor))) {
++ printk(KERN_WARNING
++ "blkback: changing physical device (from %x:%x to "
++ "%x:%x) not supported.\n", be->major, be->minor,
++ major, minor);
++ return;
++ }
++
++ be->mode = xenbus_read(XBT_NIL, dev->nodename, "mode", NULL);
++ if (IS_ERR(be->mode)) {
++ err = PTR_ERR(be->mode);
++ be->mode = NULL;
++ xenbus_dev_fatal(dev, err, "reading mode");
++ return;
++ }
++
++ device_type = xenbus_read(XBT_NIL, dev->otherend, "device-type", NULL);
++ if (!IS_ERR(device_type)) {
++ cdrom = strcmp(device_type, "cdrom") == 0;
++ kfree(device_type);
++ }
++
++ if (be->major == 0 && be->minor == 0) {
++ /* Front end dir is a number, which is used as the handle. */
++
++ char *p = strrchr(dev->otherend, '/') + 1;
++ long handle = simple_strtoul(p, NULL, 0);
++
++ be->major = major;
++ be->minor = minor;
++
++ err = vbd_create(be->blkif, handle, major, minor,
++ (NULL == strchr(be->mode, 'w')), cdrom);
++ if (err) {
++ be->major = be->minor = 0;
++ xenbus_dev_fatal(dev, err, "creating vbd structure");
++ return;
++ }
++
++ err = xenvbd_sysfs_addif(dev);
++ if (err) {
++ vbd_free(&be->blkif->vbd);
++ be->major = be->minor = 0;
++ xenbus_dev_fatal(dev, err, "creating sysfs entries");
++ return;
++ }
++
++ /* We're potentially connected now */
++ update_blkif_status(be->blkif);
++ }
++}
++
++
++/**
++ * Callback received when the frontend's state changes.
++ */
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
++
++ DPRINTK("%s", xenbus_strstate(frontend_state));
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
++
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ /* Ensure we connect even when two watches fire in
++ close successsion and we miss the intermediate value
++ of frontend_state. */
++ if (dev->state == XenbusStateConnected)
++ break;
++
++ err = connect_ring(be);
++ if (err)
++ break;
++ update_blkif_status(be->blkif);
++ break;
++
++ case XenbusStateClosing:
++ blkif_disconnect(be->blkif);
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
++}
++
++
++/* ** Connection ** */
++
++
++/**
++ * Write the physical details regarding the block device to the store, and
++ * switch to Connected state.
++ */
++static void connect(struct backend_info *be)
++{
++ struct xenbus_transaction xbt;
++ int err;
++ struct xenbus_device *dev = be->dev;
++
++ DPRINTK("%s", dev->otherend);
++
++ /* Supply the information about the device the frontend needs */
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ return;
++ }
++
++ err = blkback_barrier(xbt, be, 1);
++ if (err)
++ goto abort;
++
++ err = xenbus_printf(xbt, dev->nodename, "sectors", "%llu",
++ vbd_size(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/sectors",
++ dev->nodename);
++ goto abort;
++ }
++
++ /* FIXME: use a typename instead */
++ err = xenbus_printf(xbt, dev->nodename, "info", "%u",
++ vbd_info(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/info",
++ dev->nodename);
++ goto abort;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "sector-size", "%lu",
++ vbd_secsize(&be->blkif->vbd));
++ if (err) {
++ xenbus_dev_fatal(dev, err, "writing %s/sector-size",
++ dev->nodename);
++ goto abort;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err)
++ xenbus_dev_fatal(dev, err, "ending transaction");
++
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(dev, err, "switching to Connected state",
++ dev->nodename);
++
++ return;
++ abort:
++ xenbus_transaction_end(xbt, 1);
++}
++
++
++static int connect_ring(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ char protocol[64] = "";
++ int err;
++
++ DPRINTK("%s", dev->otherend);
++
++ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
++
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
++ "%63s", protocol, NULL);
++ if (err)
++ strcpy(protocol, "unspecified, assuming native");
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
++ else {
++ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
++ return -1;
++ }
++ printk(KERN_INFO
++ "blkback: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
++ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
++
++ /* Map the shared frame, irq etc. */
++ err = blkif_map(be->blkif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
++ ring_ref, evtchn);
++ return err;
++ }
++
++ return 0;
++}
++
++
++/* ** Driver Registration ** */
++
++
++static const struct xenbus_device_id blkback_ids[] = {
++ { "vbd" },
++ { "" }
++};
++
++
++static struct xenbus_driver blkback = {
++ .name = "vbd",
++ .ids = blkback_ids,
++ .probe = blkback_probe,
++ .remove = blkback_remove,
++ .otherend_changed = frontend_changed
++};
++
++
++void blkif_xenbus_init(void)
++{
++ if (xenbus_register_backend(&blkback))
++ BUG();
++}
+diff --git a/drivers/xen/blkfront/Makefile b/drivers/xen/blkfront/Makefile
+new file mode 100644
+index 0000000..182ef65
+--- /dev/null
++++ b/drivers/xen/blkfront/Makefile
+@@ -0,0 +1,5 @@
++
++obj-$(CONFIG_XEN_BLKDEV_FRONTEND) := xenblk.o
++
++xenblk-objs := blkfront.o vbd.o
++
+diff --git a/drivers/xen/blkfront/blkfront.c b/drivers/xen/blkfront/blkfront.c
+new file mode 100644
+index 0000000..ce68e2c
+--- /dev/null
++++ b/drivers/xen/blkfront/blkfront.c
+@@ -0,0 +1,929 @@
++/******************************************************************************
++ * blkfront.c
++ *
++ * XenLinux virtual block-device driver.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004, Christian Limpach
++ * Copyright (c) 2004, Andrew Warfield
++ * Copyright (c) 2005, Christopher Clark
++ * Copyright (c) 2005, XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/version.h>
++#include "block.h"
++#include <linux/cdrom.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <scsi/scsi.h>
++#include <xen/evtchn.h>
++#include <xen/xenbus.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/protocols.h>
++#include <xen/gnttab.h>
++#include <asm/hypervisor.h>
++#include <asm/maddr.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#define BLKIF_STATE_DISCONNECTED 0
++#define BLKIF_STATE_CONNECTED 1
++#define BLKIF_STATE_SUSPENDED 2
++
++#define MAXIMUM_OUTSTANDING_BLOCK_REQS \
++ (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE)
++#define GRANT_INVALID_REF 0
++
++static void connect(struct blkfront_info *);
++static void blkfront_closing(struct xenbus_device *);
++static int blkfront_remove(struct xenbus_device *);
++static int talk_to_backend(struct xenbus_device *, struct blkfront_info *);
++static int setup_blkring(struct xenbus_device *, struct blkfront_info *);
++
++static void kick_pending_request_queues(struct blkfront_info *);
++
++static irqreturn_t blkif_int(int irq, void *dev_id);
++static void blkif_restart_queue(struct work_struct *arg);
++static void blkif_recover(struct blkfront_info *);
++static void blkif_completion(struct blk_shadow *);
++static void blkif_free(struct blkfront_info *, int);
++
++
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and the ring buffer for communication with the backend, and
++ * inform the backend of the appropriate details for those. Switch to
++ * Initialised state.
++ */
++static int blkfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err, vdevice, i;
++ struct blkfront_info *info;
++
++ /* FIXME: Use dynamic device id if this is not set. */
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "virtual-device", "%i", &vdevice);
++ if (err != 1) {
++ /* go looking in the extended area instead */
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext",
++ "%i", &vdevice);
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading virtual-device");
++ return err;
++ }
++ }
++
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (!info) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
++ }
++
++ info->xbdev = dev;
++ info->vdevice = vdevice;
++ info->connected = BLKIF_STATE_DISCONNECTED;
++ INIT_WORK(&info->work, blkif_restart_queue);
++
++ for (i = 0; i < BLK_RING_SIZE; i++)
++ info->shadow[i].req.id = i+1;
++ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
++
++ /* Front end dir is a number, which is used as the id. */
++ info->handle = simple_strtoul(strrchr(dev->nodename,'/')+1, NULL, 0);
++ dev->dev.driver_data = info;
++
++ err = talk_to_backend(dev, info);
++ if (err) {
++ kfree(info);
++ dev->dev.driver_data = NULL;
++ return err;
++ }
++
++ return 0;
++}
++
++
++/**
++ * We are reconnecting to the backend, due to a suspend/resume, or a backend
++ * driver restart. We tear down our blkif structure and recreate it, but
++ * leave the device-layer structures intact so that this is transparent to the
++ * rest of the kernel.
++ */
++static int blkfront_resume(struct xenbus_device *dev)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++ int err;
++
++ DPRINTK("blkfront_resume: %s\n", dev->nodename);
++
++ blkif_free(info, info->connected == BLKIF_STATE_CONNECTED);
++
++ err = talk_to_backend(dev, info);
++ if (info->connected == BLKIF_STATE_SUSPENDED && !err)
++ blkif_recover(info);
++
++ return err;
++}
++
++
++/* Common code used when first setting up, and when resuming. */
++static int talk_to_backend(struct xenbus_device *dev,
++ struct blkfront_info *info)
++{
++ const char *message = NULL;
++ struct xenbus_transaction xbt;
++ int err;
++
++ /* Create shared ring, alloc event channel. */
++ err = setup_blkring(dev, info);
++ if (err)
++ goto out;
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_blkring;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename,
++ "ring-ref","%u", info->ring_ref);
++ if (err) {
++ message = "writing ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
++ XEN_IO_PROTO_ABI_NATIVE);
++ if (err) {
++ message = "writing protocol";
++ goto abort_transaction;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_blkring;
++ }
++
++ xenbus_switch_state(dev, XenbusStateInitialised);
++
++ return 0;
++
++ abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ if (message)
++ xenbus_dev_fatal(dev, err, "%s", message);
++ destroy_blkring:
++ blkif_free(info, 0);
++ out:
++ return err;
++}
++
++
++static int setup_blkring(struct xenbus_device *dev,
++ struct blkfront_info *info)
++{
++ blkif_sring_t *sring;
++ int err;
++
++ info->ring_ref = GRANT_INVALID_REF;
++
++ sring = (blkif_sring_t *)__get_free_page(GFP_NOIO | __GFP_HIGH);
++ if (!sring) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
++ return -ENOMEM;
++ }
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring));
++ if (err < 0) {
++ free_page((unsigned long)sring);
++ info->ring.sring = NULL;
++ goto fail;
++ }
++ info->ring_ref = err;
++
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, blkif_int, IRQF_SAMPLE_RANDOM, "blkif", info);
++ if (err <= 0) {
++ xenbus_dev_fatal(dev, err,
++ "bind_listening_port_to_irqhandler");
++ goto fail;
++ }
++ info->irq = err;
++
++ return 0;
++fail:
++ blkif_free(info, 0);
++ return err;
++}
++
++
++/**
++ * Callback received when the backend's state changes.
++ */
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++ struct block_device *bd;
++
++ DPRINTK("blkfront:backend_changed.\n");
++
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateConnected:
++ connect(info);
++ break;
++
++ case XenbusStateClosing:
++ bd = bdget_disk(info->gd, 0);
++ if (bd == NULL)
++ xenbus_dev_fatal(dev, -ENODEV, "bdget failed");
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
++ down(&bd->bd_sem);
++#else
++ mutex_lock(&bd->bd_mutex);
++#endif
++ if (info->users > 0)
++ xenbus_dev_error(dev, -EBUSY,
++ "Device in use; refusing to close");
++ else
++ blkfront_closing(dev);
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17)
++ up(&bd->bd_sem);
++#else
++ mutex_unlock(&bd->bd_mutex);
++#endif
++ bdput(bd);
++ break;
++ }
++}
++
++
++/* ** Connection ** */
++
++
++/*
++ * Invoked when the backend is finally 'ready' (and has told produced
++ * the details about the physical device - #sectors, size, etc).
++ */
++static void connect(struct blkfront_info *info)
++{
++ unsigned long long sectors;
++ unsigned long sector_size;
++ unsigned int binfo;
++ int err;
++
++ if ((info->connected == BLKIF_STATE_CONNECTED) ||
++ (info->connected == BLKIF_STATE_SUSPENDED) )
++ return;
++
++ DPRINTK("blkfront.c:connect:%s.\n", info->xbdev->otherend);
++
++ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
++ "sectors", "%Lu", §ors,
++ "info", "%u", &binfo,
++ "sector-size", "%lu", §or_size,
++ NULL);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err,
++ "reading backend fields at %s",
++ info->xbdev->otherend);
++ return;
++ }
++
++ err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
++ "feature-barrier", "%lu", &info->feature_barrier,
++ NULL);
++ if (err)
++ info->feature_barrier = 0;
++
++ err = xlvbd_add(sectors, info->vdevice, binfo, sector_size, info);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s",
++ info->xbdev->otherend);
++ return;
++ }
++
++ err = xlvbd_sysfs_addif(info);
++ if (err) {
++ xenbus_dev_fatal(info->xbdev, err, "xlvbd_sysfs_addif at %s",
++ info->xbdev->otherend);
++ return;
++ }
++
++ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
++
++ /* Kick pending requests. */
++ spin_lock_irq(&blkif_io_lock);
++ info->connected = BLKIF_STATE_CONNECTED;
++ kick_pending_request_queues(info);
++ spin_unlock_irq(&blkif_io_lock);
++
++ add_disk(info->gd);
++
++ info->is_ready = 1;
++}
++
++/**
++ * Handle the change of state of the backend to Closing. We must delete our
++ * device-layer structures now, to ensure that writes are flushed through to
++ * the backend. Once is this done, we can switch to Closed in
++ * acknowledgement.
++ */
++static void blkfront_closing(struct xenbus_device *dev)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++ unsigned long flags;
++
++ DPRINTK("blkfront_closing: %s removed\n", dev->nodename);
++
++ if (info->rq == NULL)
++ goto out;
++
++ spin_lock_irqsave(&blkif_io_lock, flags);
++ /* No more blkif_request(). */
++ blk_stop_queue(info->rq);
++ /* No more gnttab callback work. */
++ gnttab_cancel_free_callback(&info->callback);
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
++
++ /* Flush gnttab callback work. Must be done with no locks held. */
++ flush_scheduled_work();
++
++ xlvbd_sysfs_delif(info);
++
++ xlvbd_del(info);
++
++ out:
++ xenbus_frontend_closed(dev);
++}
++
++
++static int blkfront_remove(struct xenbus_device *dev)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++
++ DPRINTK("blkfront_remove: %s removed\n", dev->nodename);
++
++ blkif_free(info, 0);
++
++ kfree(info);
++
++ return 0;
++}
++
++
++static inline int GET_ID_FROM_FREELIST(
++ struct blkfront_info *info)
++{
++ unsigned long free = info->shadow_free;
++ BUG_ON(free > BLK_RING_SIZE);
++ info->shadow_free = info->shadow[free].req.id;
++ info->shadow[free].req.id = 0x0fffffee; /* debug */
++ return free;
++}
++
++static inline void ADD_ID_TO_FREELIST(
++ struct blkfront_info *info, unsigned long id)
++{
++ info->shadow[id].req.id = info->shadow_free;
++ info->shadow[id].request = 0;
++ info->shadow_free = id;
++}
++
++static inline void flush_requests(struct blkfront_info *info)
++{
++ int notify;
++
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify);
++
++ if (notify)
++ notify_remote_via_irq(info->irq);
++}
++
++static void kick_pending_request_queues(struct blkfront_info *info)
++{
++ if (!RING_FULL(&info->ring)) {
++ /* Re-enable calldowns. */
++ blk_start_queue(info->rq);
++ /* Kick things off immediately. */
++ do_blkif_request(info->rq);
++ }
++}
++
++static void blkif_restart_queue(struct work_struct *arg)
++{
++ struct blkfront_info *info = container_of(arg, struct blkfront_info, work);
++ spin_lock_irq(&blkif_io_lock);
++ if (info->connected == BLKIF_STATE_CONNECTED)
++ kick_pending_request_queues(info);
++ spin_unlock_irq(&blkif_io_lock);
++}
++
++static void blkif_restart_queue_callback(void *arg)
++{
++ struct blkfront_info *info = (struct blkfront_info *)arg;
++ schedule_work(&info->work);
++}
++
++int blkif_open(struct inode *inode, struct file *filep)
++{
++ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
++ info->users++;
++ return 0;
++}
++
++
++int blkif_release(struct inode *inode, struct file *filep)
++{
++ struct blkfront_info *info = inode->i_bdev->bd_disk->private_data;
++ info->users--;
++ if (info->users == 0) {
++ /* Check whether we have been instructed to close. We will
++ have ignored this request initially, as the device was
++ still mounted. */
++ struct xenbus_device * dev = info->xbdev;
++ enum xenbus_state state = xenbus_read_driver_state(dev->otherend);
++
++ if (state == XenbusStateClosing && info->is_ready)
++ blkfront_closing(dev);
++ }
++ return 0;
++}
++
++
++int blkif_ioctl(struct inode *inode, struct file *filep,
++ unsigned command, unsigned long argument)
++{
++ int i;
++
++ DPRINTK_IOCTL("command: 0x%x, argument: 0x%lx, dev: 0x%04x\n",
++ command, (long)argument, inode->i_rdev);
++
++ switch (command) {
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
++ case HDIO_GETGEO: {
++ struct block_device *bd = inode->i_bdev;
++ struct hd_geometry geo;
++ int ret;
++
++ if (!argument)
++ return -EINVAL;
++
++ geo.start = get_start_sect(bd);
++ ret = blkif_getgeo(bd, &geo);
++ if (ret)
++ return ret;
++
++ if (copy_to_user((struct hd_geometry __user *)argument, &geo,
++ sizeof(geo)))
++ return -EFAULT;
++
++ return 0;
++ }
++#endif
++ case CDROMMULTISESSION:
++ DPRINTK("FIXME: support multisession CDs later\n");
++ for (i = 0; i < sizeof(struct cdrom_multisession); i++)
++ if (put_user(0, (char __user *)(argument + i)))
++ return -EFAULT;
++ return 0;
++
++ case CDROM_GET_CAPABILITY: {
++ struct blkfront_info *info =
++ inode->i_bdev->bd_disk->private_data;
++ struct gendisk *gd = info->gd;
++ if (gd->flags & GENHD_FL_CD)
++ return 0;
++ return -EINVAL;
++ }
++ default:
++ /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
++ command);*/
++ return -EINVAL; /* same return as native Linux */
++ }
++
++ return 0;
++}
++
++
++int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
++{
++ /* We don't have real geometry info, but let's at least return
++ values consistent with the size of the device */
++ sector_t nsect = get_capacity(bd->bd_disk);
++ sector_t cylinders = nsect;
++
++ hg->heads = 0xff;
++ hg->sectors = 0x3f;
++ sector_div(cylinders, hg->heads * hg->sectors);
++ hg->cylinders = cylinders;
++ if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect)
++ hg->cylinders = 0xffff;
++ return 0;
++}
++
++
++/*
++ * blkif_queue_request
++ *
++ * request block io
++ *
++ * id: for guest use only.
++ * operation: BLKIF_OP_{READ,WRITE,PROBE}
++ * buffer: buffer to read/write into. this should be a
++ * virtual address in the guest os.
++ */
++static int blkif_queue_request(struct request *req)
++{
++ struct blkfront_info *info = req->rq_disk->private_data;
++ unsigned long buffer_mfn;
++ blkif_request_t *ring_req;
++ struct bio_vec *bvec;
++ struct req_iterator iter;
++ unsigned long id;
++ unsigned int fsect, lsect;
++ int ref;
++ grant_ref_t gref_head;
++
++ if (unlikely(info->connected != BLKIF_STATE_CONNECTED))
++ return 1;
++
++ if (gnttab_alloc_grant_references(
++ BLKIF_MAX_SEGMENTS_PER_REQUEST, &gref_head) < 0) {
++ gnttab_request_free_callback(
++ &info->callback,
++ blkif_restart_queue_callback,
++ info,
++ BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ return 1;
++ }
++
++ /* Fill out a communications ring structure. */
++ ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
++ id = GET_ID_FROM_FREELIST(info);
++ info->shadow[id].request = (unsigned long)req;
++
++ ring_req->id = id;
++ ring_req->sector_number = (blkif_sector_t)req->sector;
++ ring_req->handle = info->handle;
++
++ ring_req->operation = rq_data_dir(req) ?
++ BLKIF_OP_WRITE : BLKIF_OP_READ;
++ if (blk_barrier_rq(req))
++ ring_req->operation = BLKIF_OP_WRITE_BARRIER;
++
++ ring_req->nr_segments = 0;
++ rq_for_each_segment(bvec, req, iter) {
++ BUG_ON(ring_req->nr_segments
++ == BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ buffer_mfn = page_to_phys(bvec->bv_page) >> PAGE_SHIFT;
++ fsect = bvec->bv_offset >> 9;
++ lsect = fsect + (bvec->bv_len >> 9) - 1;
++ /* install a grant reference. */
++ ref = gnttab_claim_grant_reference(&gref_head);
++ BUG_ON(ref == -ENOSPC);
++
++ gnttab_grant_foreign_access_ref(
++ ref,
++ info->xbdev->otherend_id,
++ buffer_mfn,
++ rq_data_dir(req) ? GTF_readonly : 0 );
++
++ info->shadow[id].frame[ring_req->nr_segments] =
++ mfn_to_pfn(buffer_mfn);
++
++ ring_req->seg[ring_req->nr_segments] =
++ (struct blkif_request_segment) {
++ .gref = ref,
++ .first_sect = fsect,
++ .last_sect = lsect };
++
++ ring_req->nr_segments++;
++ }
++
++ info->ring.req_prod_pvt++;
++
++ /* Keep a private copy so we can reissue requests when recovering. */
++ info->shadow[id].req = *ring_req;
++
++ gnttab_free_grant_references(gref_head);
++
++ return 0;
++}
++
++/*
++ * do_blkif_request
++ * read a block; request is in a request queue
++ */
++void do_blkif_request(struct request_queue *rq)
++{
++ struct blkfront_info *info = NULL;
++ struct request *req;
++ int queued;
++
++ DPRINTK("Entered do_blkif_request\n");
++
++ queued = 0;
++
++ while ((req = elv_next_request(rq)) != NULL) {
++ info = req->rq_disk->private_data;
++ if (!blk_fs_request(req)) {
++ end_request(req, 0);
++ continue;
++ }
++
++ if (RING_FULL(&info->ring))
++ goto wait;
++
++ DPRINTK("do_blk_req %p: cmd %p, sec %llx, "
++ "(%u/%li) buffer:%p [%s]\n",
++ req, req->cmd, (long long)req->sector,
++ req->current_nr_sectors,
++ req->nr_sectors, req->buffer,
++ rq_data_dir(req) ? "write" : "read");
++
++
++ blkdev_dequeue_request(req);
++ if (blkif_queue_request(req)) {
++ blk_requeue_request(rq, req);
++ wait:
++ /* Avoid pointless unplugs. */
++ blk_stop_queue(rq);
++ break;
++ }
++
++ queued++;
++ }
++
++ if (queued != 0)
++ flush_requests(info);
++}
++
++
++static irqreturn_t blkif_int(int irq, void *dev_id)
++{
++ struct request *req;
++ blkif_response_t *bret;
++ RING_IDX i, rp;
++ unsigned long flags;
++ struct blkfront_info *info = (struct blkfront_info *)dev_id;
++
++ spin_lock_irqsave(&blkif_io_lock, flags);
++
++ if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
++ return IRQ_HANDLED;
++ }
++
++ again:
++ rp = info->ring.sring->rsp_prod;
++ rmb(); /* Ensure we see queued responses up to 'rp'. */
++
++ for (i = info->ring.rsp_cons; i != rp; i++) {
++ unsigned long id;
++ int ret;
++
++ bret = RING_GET_RESPONSE(&info->ring, i);
++ id = bret->id;
++ req = (struct request *)info->shadow[id].request;
++
++ blkif_completion(&info->shadow[id]);
++
++ ADD_ID_TO_FREELIST(info, id);
++
++ ret = bret->status == BLKIF_RSP_OKAY ? 0 : -EIO;
++ switch (bret->operation) {
++ case BLKIF_OP_WRITE_BARRIER:
++ if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
++ printk("blkfront: %s: write barrier op failed\n",
++ info->gd->disk_name);
++ ret = -EOPNOTSUPP;
++ info->feature_barrier = 0;
++ xlvbd_barrier(info);
++ }
++ /* fall through */
++ case BLKIF_OP_READ:
++ case BLKIF_OP_WRITE:
++ if (unlikely(bret->status != BLKIF_RSP_OKAY))
++ DPRINTK("Bad return from blkdev data "
++ "request: %x\n", bret->status);
++
++ ret = __blk_end_request(req, ret, blk_rq_bytes(req));
++ BUG_ON(ret);
++ break;
++ default:
++ BUG();
++ }
++ }
++
++ info->ring.rsp_cons = i;
++
++ if (i != info->ring.req_prod_pvt) {
++ int more_to_do;
++ RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
++ if (more_to_do)
++ goto again;
++ } else
++ info->ring.sring->rsp_event = i + 1;
++
++ kick_pending_request_queues(info);
++
++ spin_unlock_irqrestore(&blkif_io_lock, flags);
++
++ return IRQ_HANDLED;
++}
++
++static void blkif_free(struct blkfront_info *info, int suspend)
++{
++ /* Prevent new requests being issued until we fix things up. */
++ spin_lock_irq(&blkif_io_lock);
++ info->connected = suspend ?
++ BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
++ /* No more blkif_request(). */
++ if (info->rq)
++ blk_stop_queue(info->rq);
++ /* No more gnttab callback work. */
++ gnttab_cancel_free_callback(&info->callback);
++ spin_unlock_irq(&blkif_io_lock);
++
++ /* Flush gnttab callback work. Must be done with no locks held. */
++ flush_scheduled_work();
++
++ /* Free resources associated with old device channel. */
++ if (info->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(info->ring_ref,
++ (unsigned long)info->ring.sring);
++ info->ring_ref = GRANT_INVALID_REF;
++ info->ring.sring = NULL;
++ }
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
++}
++
++static void blkif_completion(struct blk_shadow *s)
++{
++ int i;
++ for (i = 0; i < s->req.nr_segments; i++)
++ gnttab_end_foreign_access(s->req.seg[i].gref, 0UL);
++}
++
++static void blkif_recover(struct blkfront_info *info)
++{
++ int i;
++ blkif_request_t *req;
++ struct blk_shadow *copy;
++ int j;
++
++ /* Stage 1: Make a safe copy of the shadow state. */
++ copy = kmalloc(sizeof(info->shadow), GFP_NOIO | __GFP_NOFAIL | __GFP_HIGH);
++ memcpy(copy, info->shadow, sizeof(info->shadow));
++
++ /* Stage 2: Set up free list. */
++ memset(&info->shadow, 0, sizeof(info->shadow));
++ for (i = 0; i < BLK_RING_SIZE; i++)
++ info->shadow[i].req.id = i+1;
++ info->shadow_free = info->ring.req_prod_pvt;
++ info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;
++
++ /* Stage 3: Find pending requests and requeue them. */
++ for (i = 0; i < BLK_RING_SIZE; i++) {
++ /* Not in use? */
++ if (copy[i].request == 0)
++ continue;
++
++ /* Grab a request slot and copy shadow state into it. */
++ req = RING_GET_REQUEST(
++ &info->ring, info->ring.req_prod_pvt);
++ *req = copy[i].req;
++
++ /* We get a new request id, and must reset the shadow state. */
++ req->id = GET_ID_FROM_FREELIST(info);
++ memcpy(&info->shadow[req->id], ©[i], sizeof(copy[i]));
++
++ /* Rewrite any grant references invalidated by susp/resume. */
++ for (j = 0; j < req->nr_segments; j++)
++ gnttab_grant_foreign_access_ref(
++ req->seg[j].gref,
++ info->xbdev->otherend_id,
++ pfn_to_mfn(info->shadow[req->id].frame[j]),
++ rq_data_dir((struct request *)
++ info->shadow[req->id].request) ?
++ GTF_readonly : 0);
++ info->shadow[req->id].req = *req;
++
++ info->ring.req_prod_pvt++;
++ }
++
++ kfree(copy);
++
++ (void)xenbus_switch_state(info->xbdev, XenbusStateConnected);
++
++ spin_lock_irq(&blkif_io_lock);
++
++ /* Now safe for us to use the shared ring */
++ info->connected = BLKIF_STATE_CONNECTED;
++
++ /* Send off requeued requests */
++ flush_requests(info);
++
++ /* Kick any other new requests queued since we resumed */
++ kick_pending_request_queues(info);
++
++ spin_unlock_irq(&blkif_io_lock);
++}
++
++int blkfront_is_ready(struct xenbus_device *dev)
++{
++ struct blkfront_info *info = dev->dev.driver_data;
++
++ return info->is_ready;
++}
++
++
++/* ** Driver Registration ** */
++
++
++static const struct xenbus_device_id blkfront_ids[] = {
++ { "vbd" },
++ { "" }
++};
++MODULE_ALIAS("xen:vbd");
++
++static struct xenbus_driver blkfront = {
++ .name = "vbd",
++ .ids = blkfront_ids,
++ .probe = blkfront_probe,
++ .remove = blkfront_remove,
++ .resume = blkfront_resume,
++ .otherend_changed = backend_changed,
++ .is_ready = blkfront_is_ready,
++};
++
++
++static int __init xlblk_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ return xenbus_register_frontend(&blkfront);
++}
++module_init(xlblk_init);
++
++
++static void __exit xlblk_exit(void)
++{
++ return xenbus_unregister_driver(&blkfront);
++}
++module_exit(xlblk_exit);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/blkfront/block.h b/drivers/xen/blkfront/block.h
+new file mode 100644
+index 0000000..ee87a40
+--- /dev/null
++++ b/drivers/xen/blkfront/block.h
+@@ -0,0 +1,157 @@
++/******************************************************************************
++ * block.h
++ *
++ * Shared definitions between all levels of XenLinux Virtual block devices.
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004-2005, Christian Limpach
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_DRIVERS_BLOCK_H__
++#define __XEN_DRIVERS_BLOCK_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/fs.h>
++#include <linux/hdreg.h>
++#include <linux/blkdev.h>
++#include <linux/major.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/blkif.h>
++#include <xen/interface/io/ring.h>
++#include <asm/io.h>
++#include <asm/atomic.h>
++#include <asm/uaccess.h>
++
++#define DPRINTK(_f, _a...) pr_debug(_f, ## _a)
++
++#if 0
++#define DPRINTK_IOCTL(_f, _a...) printk(KERN_ALERT _f, ## _a)
++#else
++#define DPRINTK_IOCTL(_f, _a...) ((void)0)
++#endif
++
++struct xlbd_type_info
++{
++ int partn_shift;
++ int disks_per_major;
++ char *devname;
++ char *diskname;
++};
++
++struct xlbd_major_info
++{
++ int major;
++ int index;
++ int usage;
++ struct xlbd_type_info *type;
++};
++
++struct blk_shadow {
++ blkif_request_t req;
++ unsigned long request;
++ unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++
++#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
++
++/*
++ * We have one of these per vbd, whether ide, scsi or 'other'. They
++ * hang in private_data off the gendisk structure. We may end up
++ * putting all kinds of interesting stuff here :-)
++ */
++struct blkfront_info
++{
++ struct xenbus_device *xbdev;
++ struct gendisk *gd;
++ int vdevice;
++ blkif_vdev_t handle;
++ int connected;
++ int ring_ref;
++ blkif_front_ring_t ring;
++ unsigned int irq;
++ struct xlbd_major_info *mi;
++ struct request_queue *rq;
++ struct work_struct work;
++ struct gnttab_free_callback callback;
++ struct blk_shadow shadow[BLK_RING_SIZE];
++ unsigned long shadow_free;
++ int feature_barrier;
++ int is_ready;
++
++ /**
++ * The number of people holding this device open. We won't allow a
++ * hot-unplug unless this is 0.
++ */
++ int users;
++};
++
++extern spinlock_t blkif_io_lock;
++
++extern int blkif_open(struct inode *inode, struct file *filep);
++extern int blkif_release(struct inode *inode, struct file *filep);
++extern int blkif_ioctl(struct inode *inode, struct file *filep,
++ unsigned command, unsigned long argument);
++extern int blkif_getgeo(struct block_device *, struct hd_geometry *);
++extern int blkif_check(dev_t dev);
++extern int blkif_revalidate(dev_t dev);
++extern void do_blkif_request (struct request_queue *rq);
++
++/* Virtual block-device subsystem. */
++/* Note that xlvbd_add doesn't call add_disk for you: you're expected
++ to call add_disk on info->gd once the disk is properly connected
++ up. */
++int xlvbd_add(blkif_sector_t capacity, int device,
++ u16 vdisk_info, u16 sector_size, struct blkfront_info *info);
++void xlvbd_del(struct blkfront_info *info);
++int xlvbd_barrier(struct blkfront_info *info);
++
++#ifdef CONFIG_SYSFS
++int xlvbd_sysfs_addif(struct blkfront_info *info);
++void xlvbd_sysfs_delif(struct blkfront_info *info);
++#else
++static inline int xlvbd_sysfs_addif(struct blkfront_info *info)
++{
++ return 0;
++}
++
++static inline void xlvbd_sysfs_delif(struct blkfront_info *info)
++{
++ ;
++}
++#endif
++
++#endif /* __XEN_DRIVERS_BLOCK_H__ */
+diff --git a/drivers/xen/blkfront/vbd.c b/drivers/xen/blkfront/vbd.c
+new file mode 100644
+index 0000000..ad56e46
+--- /dev/null
++++ b/drivers/xen/blkfront/vbd.c
+@@ -0,0 +1,440 @@
++/******************************************************************************
++ * vbd.c
++ *
++ * XenLinux virtual block-device driver (xvd).
++ *
++ * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
++ * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
++ * Copyright (c) 2004-2005, Christian Limpach
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "block.h"
++#include <linux/blkdev.h>
++#include <linux/list.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#define BLKIF_MAJOR(dev) ((dev)>>8)
++#define BLKIF_MINOR(dev) ((dev) & 0xff)
++
++#define EXT_SHIFT 28
++#define EXTENDED (1<<EXT_SHIFT)
++#define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
++#define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
++
++/*
++ * For convenience we distinguish between ide, scsi and 'other' (i.e.,
++ * potentially combinations of the two) in the naming scheme and in a few other
++ * places.
++ */
++
++#define NUM_IDE_MAJORS 10
++#define NUM_SCSI_MAJORS 17
++#define NUM_VBD_MAJORS 2
++
++static struct xlbd_type_info xlbd_ide_type = {
++ .partn_shift = 6,
++ .disks_per_major = 2,
++ .devname = "ide",
++ .diskname = "hd",
++};
++
++static struct xlbd_type_info xlbd_scsi_type = {
++ .partn_shift = 4,
++ .disks_per_major = 16,
++ .devname = "sd",
++ .diskname = "sd",
++};
++
++static struct xlbd_type_info xlbd_vbd_type = {
++ .partn_shift = 4,
++ .disks_per_major = 16,
++ .devname = "xvd",
++ .diskname = "xvd",
++};
++
++static struct xlbd_type_info xlbd_vbd_type_ext = {
++ .partn_shift = 8,
++ .disks_per_major = 256,
++ .devname = "xvd",
++ .diskname = "xvd",
++};
++
++static struct xlbd_major_info *major_info[NUM_IDE_MAJORS + NUM_SCSI_MAJORS +
++ NUM_VBD_MAJORS];
++
++#define XLBD_MAJOR_IDE_START 0
++#define XLBD_MAJOR_SCSI_START (NUM_IDE_MAJORS)
++#define XLBD_MAJOR_VBD_START (NUM_IDE_MAJORS + NUM_SCSI_MAJORS)
++
++#define XLBD_MAJOR_IDE_RANGE XLBD_MAJOR_IDE_START ... XLBD_MAJOR_SCSI_START - 1
++#define XLBD_MAJOR_SCSI_RANGE XLBD_MAJOR_SCSI_START ... XLBD_MAJOR_VBD_START - 1
++#define XLBD_MAJOR_VBD_RANGE XLBD_MAJOR_VBD_START ... XLBD_MAJOR_VBD_START + NUM_VBD_MAJORS - 1
++
++static struct block_device_operations xlvbd_block_fops =
++{
++ .owner = THIS_MODULE,
++ .open = blkif_open,
++ .release = blkif_release,
++ .ioctl = blkif_ioctl,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++ .getgeo = blkif_getgeo
++#endif
++};
++
++DEFINE_SPINLOCK(blkif_io_lock);
++
++static struct xlbd_major_info *
++xlbd_alloc_major_info(int major, int minor, int index)
++{
++ struct xlbd_major_info *ptr;
++ int do_register;
++
++ ptr = kzalloc(sizeof(struct xlbd_major_info), GFP_KERNEL);
++ if (ptr == NULL)
++ return NULL;
++
++ ptr->major = major;
++ do_register = 1;
++
++ switch (index) {
++ case XLBD_MAJOR_IDE_RANGE:
++ ptr->type = &xlbd_ide_type;
++ ptr->index = index - XLBD_MAJOR_IDE_START;
++ break;
++ case XLBD_MAJOR_SCSI_RANGE:
++ ptr->type = &xlbd_scsi_type;
++ ptr->index = index - XLBD_MAJOR_SCSI_START;
++ break;
++ case XLBD_MAJOR_VBD_RANGE:
++ ptr->index = 0;
++ if ((index - XLBD_MAJOR_VBD_START) == 0)
++ ptr->type = &xlbd_vbd_type;
++ else
++ ptr->type = &xlbd_vbd_type_ext;
++
++ /*
++ * if someone already registered block major 202,
++ * don't try to register it again
++ */
++ if (major_info[XLBD_MAJOR_VBD_START] != NULL)
++ do_register = 0;
++ break;
++ }
++
++ if (do_register) {
++ if (register_blkdev(ptr->major, ptr->type->devname)) {
++ kfree(ptr);
++ return NULL;
++ }
++
++ printk("xen-vbd: registered block device major %i\n", ptr->major);
++ }
++
++ major_info[index] = ptr;
++ return ptr;
++}
++
++static struct xlbd_major_info *
++xlbd_get_major_info(int major, int minor, int vdevice)
++{
++ struct xlbd_major_info *mi;
++ int index;
++
++ switch (major) {
++ case IDE0_MAJOR: index = 0; break;
++ case IDE1_MAJOR: index = 1; break;
++ case IDE2_MAJOR: index = 2; break;
++ case IDE3_MAJOR: index = 3; break;
++ case IDE4_MAJOR: index = 4; break;
++ case IDE5_MAJOR: index = 5; break;
++ case IDE6_MAJOR: index = 6; break;
++ case IDE7_MAJOR: index = 7; break;
++ case IDE8_MAJOR: index = 8; break;
++ case IDE9_MAJOR: index = 9; break;
++ case SCSI_DISK0_MAJOR: index = 10; break;
++ case SCSI_DISK1_MAJOR ... SCSI_DISK7_MAJOR:
++ index = 11 + major - SCSI_DISK1_MAJOR;
++ break;
++ case SCSI_DISK8_MAJOR ... SCSI_DISK15_MAJOR:
++ index = 18 + major - SCSI_DISK8_MAJOR;
++ break;
++ case SCSI_CDROM_MAJOR: index = 26; break;
++ default:
++ if (!VDEV_IS_EXTENDED(vdevice))
++ index = 27;
++ else
++ index = 28;
++ break;
++ }
++
++ mi = ((major_info[index] != NULL) ? major_info[index] :
++ xlbd_alloc_major_info(major, minor, index));
++ if (mi)
++ mi->usage++;
++ return mi;
++}
++
++static void
++xlbd_put_major_info(struct xlbd_major_info *mi)
++{
++ mi->usage--;
++ /* XXX: release major if 0 */
++}
++
++static int
++xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size)
++{
++ struct request_queue *rq;
++
++ rq = blk_init_queue(do_blkif_request, &blkif_io_lock);
++ if (rq == NULL)
++ return -1;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
++ elevator_init(rq, "noop");
++#else
++ elevator_init(rq, &elevator_noop);
++#endif
++
++ /* Hard sector size and max sectors impersonate the equiv. hardware. */
++ blk_queue_hardsect_size(rq, sector_size);
++ blk_queue_max_sectors(rq, 512);
++
++ /* Each segment in a request is up to an aligned page in size. */
++ blk_queue_segment_boundary(rq, PAGE_SIZE - 1);
++ blk_queue_max_segment_size(rq, PAGE_SIZE);
++
++ /* Ensure a merged request will fit in a single I/O ring slot. */
++ blk_queue_max_phys_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
++ blk_queue_max_hw_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
++
++ /* Make sure buffer addresses are sector-aligned. */
++ blk_queue_dma_alignment(rq, 511);
++
++ /* Make sure we don't use bounce buffers. */
++ blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);
++
++ gd->queue = rq;
++
++ return 0;
++}
++
++int
++xlvbd_add(blkif_sector_t capacity, int vdevice, u16 vdisk_info,
++ u16 sector_size, struct blkfront_info *info)
++{
++ int major, minor;
++ struct gendisk *gd;
++ struct xlbd_major_info *mi;
++ int nr_minors = 1;
++ int err = -ENODEV;
++ unsigned int offset;
++
++ if ((vdevice>>EXT_SHIFT) > 1) {
++ /* this is above the extended range; something is wrong */
++ printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", vdevice);
++ return -ENODEV;
++ }
++
++ if (!VDEV_IS_EXTENDED(vdevice)) {
++ major = BLKIF_MAJOR(vdevice);
++ minor = BLKIF_MINOR(vdevice);
++ }
++ else {
++ major = 202;
++ minor = BLKIF_MINOR_EXT(vdevice);
++ }
++
++ BUG_ON(info->gd != NULL);
++ BUG_ON(info->mi != NULL);
++ BUG_ON(info->rq != NULL);
++
++ mi = xlbd_get_major_info(major, minor, vdevice);
++ if (mi == NULL)
++ goto out;
++ info->mi = mi;
++
++ if ((minor & ((1 << mi->type->partn_shift) - 1)) == 0)
++ nr_minors = 1 << mi->type->partn_shift;
++
++ gd = alloc_disk(nr_minors);
++ if (gd == NULL)
++ goto out;
++
++ offset = mi->index * mi->type->disks_per_major +
++ (minor >> mi->type->partn_shift);
++ if (nr_minors > 1) {
++ if (offset < 26) {
++ sprintf(gd->disk_name, "%s%c",
++ mi->type->diskname, 'a' + offset );
++ }
++ else {
++ sprintf(gd->disk_name, "%s%c%c",
++ mi->type->diskname,
++ 'a' + ((offset/26)-1), 'a' + (offset%26) );
++ }
++ }
++ else {
++ if (offset < 26) {
++ sprintf(gd->disk_name, "%s%c%d",
++ mi->type->diskname,
++ 'a' + offset,
++ minor & ((1 << mi->type->partn_shift) - 1));
++ }
++ else {
++ sprintf(gd->disk_name, "%s%c%c%d",
++ mi->type->diskname,
++ 'a' + ((offset/26)-1), 'a' + (offset%26),
++ minor & ((1 << mi->type->partn_shift) - 1));
++ }
++ }
++
++ gd->major = mi->major;
++ gd->first_minor = minor;
++ gd->fops = &xlvbd_block_fops;
++ gd->private_data = info;
++ gd->driverfs_dev = &(info->xbdev->dev);
++ set_capacity(gd, capacity);
++
++ if (xlvbd_init_blk_queue(gd, sector_size)) {
++ del_gendisk(gd);
++ goto out;
++ }
++
++ info->rq = gd->queue;
++ info->gd = gd;
++
++ if (info->feature_barrier)
++ xlvbd_barrier(info);
++
++ if (vdisk_info & VDISK_READONLY)
++ set_disk_ro(gd, 1);
++
++ if (vdisk_info & VDISK_REMOVABLE)
++ gd->flags |= GENHD_FL_REMOVABLE;
++
++ if (vdisk_info & VDISK_CDROM)
++ gd->flags |= GENHD_FL_CD;
++
++ return 0;
++
++ out:
++ if (mi)
++ xlbd_put_major_info(mi);
++ info->mi = NULL;
++ return err;
++}
++
++void
++xlvbd_del(struct blkfront_info *info)
++{
++ if (info->mi == NULL)
++ return;
++
++ BUG_ON(info->gd == NULL);
++ del_gendisk(info->gd);
++ put_disk(info->gd);
++ info->gd = NULL;
++
++ xlbd_put_major_info(info->mi);
++ info->mi = NULL;
++
++ BUG_ON(info->rq == NULL);
++ blk_cleanup_queue(info->rq);
++ info->rq = NULL;
++}
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++int
++xlvbd_barrier(struct blkfront_info *info)
++{
++ int err;
++
++ err = blk_queue_ordered(info->rq,
++ info->feature_barrier ? QUEUE_ORDERED_DRAIN : QUEUE_ORDERED_NONE, NULL);
++ if (err)
++ return err;
++ printk(KERN_INFO "blkfront: %s: barriers %s\n",
++ info->gd->disk_name, info->feature_barrier ? "enabled" : "disabled");
++ return 0;
++}
++#else
++int
++xlvbd_barrier(struct blkfront_info *info)
++{
++ printk(KERN_INFO "blkfront: %s: barriers disabled\n", info->gd->disk_name);
++ return -ENOSYS;
++}
++#endif
++
++#ifdef CONFIG_SYSFS
++static ssize_t show_media(struct device *dev,
++ struct device_attribute *attr, char *buf)
++{
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct blkfront_info *info = xendev->dev.driver_data;
++
++ if (info->gd->flags & GENHD_FL_CD)
++ return sprintf(buf, "cdrom\n");
++ return sprintf(buf, "disk\n");
++}
++
++static struct device_attribute xlvbd_attrs[] = {
++ __ATTR(media, S_IRUGO, show_media, NULL),
++};
++
++int xlvbd_sysfs_addif(struct blkfront_info *info)
++{
++ int i;
++ int error = 0;
++
++ for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++) {
++ error = device_create_file(info->gd->driverfs_dev,
++ &xlvbd_attrs[i]);
++ if (error)
++ goto fail;
++ }
++ return 0;
++
++fail:
++ while (--i >= 0)
++ device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]);
++ return error;
++}
++
++void xlvbd_sysfs_delif(struct blkfront_info *info)
++{
++ int i;
++
++ for (i = 0; i < ARRAY_SIZE(xlvbd_attrs); i++)
++ device_remove_file(info->gd->driverfs_dev, &xlvbd_attrs[i]);
++}
++
++#endif /* CONFIG_SYSFS */
+diff --git a/drivers/xen/blktap/Makefile b/drivers/xen/blktap/Makefile
+new file mode 100644
+index 0000000..f10cc4f
+--- /dev/null
++++ b/drivers/xen/blktap/Makefile
+@@ -0,0 +1,5 @@
++LINUXINCLUDE += -I../xen/include/public/io
++
++obj-$(CONFIG_XEN_BLKDEV_TAP) := xenblktap.o
++
++xenblktap-y := xenbus.o interface.o blktap.o
+diff --git a/drivers/xen/blktap/blktap.c b/drivers/xen/blktap/blktap.c
+new file mode 100644
+index 0000000..e18eeec
+--- /dev/null
++++ b/drivers/xen/blktap/blktap.c
+@@ -0,0 +1,1677 @@
++/******************************************************************************
++ * drivers/xen/blktap/blktap.c
++ *
++ * Back-end driver for user level virtual block devices. This portion of the
++ * driver exports a 'unified' block-device interface that can be accessed
++ * by any operating system that implements a compatible front end. Requests
++ * are remapped to a user-space memory region.
++ *
++ * Based on the blkback driver code.
++ *
++ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
++ *
++ * Clean ups and fix ups:
++ * Copyright (c) 2006, Steven Rostedt - Red Hat, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/freezer.h>
++#include <linux/list.h>
++#include <asm/hypervisor.h>
++#include "common.h"
++#include <xen/balloon.h>
++#include <xen/driver_util.h>
++#include <linux/kernel.h>
++#include <linux/fs.h>
++#include <linux/mm.h>
++#include <linux/errno.h>
++#include <linux/major.h>
++#include <linux/gfp.h>
++#include <linux/poll.h>
++#include <linux/delay.h>
++#include <asm/tlbflush.h>
++
++#define MAX_TAP_DEV 256 /*the maximum number of tapdisk ring devices */
++#define MAX_DEV_NAME 100 /*the max tapdisk ring device name e.g. blktap0 */
++
++/*
++ * The maximum number of requests that can be outstanding at any time
++ * is determined by
++ *
++ * [mmap_alloc * MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST]
++ *
++ * where mmap_alloc < MAX_DYNAMIC_MEM.
++ *
++ * TODO:
++ * mmap_alloc is initialised to 2 and should be adjustable on the fly via
++ * sysfs.
++ */
++#define BLK_RING_SIZE __RING_SIZE((blkif_sring_t *)0, PAGE_SIZE)
++#define MAX_DYNAMIC_MEM BLK_RING_SIZE
++#define MAX_PENDING_REQS BLK_RING_SIZE
++#define MMAP_PAGES (MAX_PENDING_REQS * BLKIF_MAX_SEGMENTS_PER_REQUEST)
++#define MMAP_VADDR(_start, _req,_seg) \
++ (_start + \
++ ((_req) * BLKIF_MAX_SEGMENTS_PER_REQUEST * PAGE_SIZE) + \
++ ((_seg) * PAGE_SIZE))
++static int blkif_reqs = MAX_PENDING_REQS;
++static int mmap_pages = MMAP_PAGES;
++
++#define RING_PAGES 1 /* BLKTAP - immediately before the mmap area, we
++ * have a bunch of pages reserved for shared
++ * memory rings.
++ */
++
++/*Data struct handed back to userspace for tapdisk device to VBD mapping*/
++typedef struct domid_translate {
++ unsigned short domid;
++ unsigned short busid;
++} domid_translate_t ;
++
++typedef struct domid_translate_ext {
++ unsigned short domid;
++ u32 busid;
++} domid_translate_ext_t ;
++
++/*Data struct associated with each of the tapdisk devices*/
++typedef struct tap_blkif {
++ struct vm_area_struct *vma; /*Shared memory area */
++ unsigned long rings_vstart; /*Kernel memory mapping */
++ unsigned long user_vstart; /*User memory mapping */
++ unsigned long dev_inuse; /*One process opens device at a time. */
++ unsigned long dev_pending; /*In process of being opened */
++ unsigned long ring_ok; /*make this ring->state */
++ blkif_front_ring_t ufe_ring; /*Rings up to user space. */
++ wait_queue_head_t wait; /*for poll */
++ unsigned long mode; /*current switching mode */
++ int minor; /*Minor number for tapdisk device */
++ pid_t pid; /*tapdisk process id */
++ enum { RUNNING, CLEANSHUTDOWN } status; /*Detect a clean userspace
++ shutdown */
++ unsigned long *idx_map; /*Record the user ring id to kern
++ [req id, idx] tuple */
++ blkif_t *blkif; /*Associate blkif with tapdev */
++ struct domid_translate_ext trans; /*Translation from domid to bus. */
++} tap_blkif_t;
++
++static struct tap_blkif *tapfds[MAX_TAP_DEV];
++static int blktap_next_minor;
++
++module_param(blkif_reqs, int, 0);
++/* Run-time switchable: /sys/module/blktap/parameters/ */
++static unsigned int log_stats = 0;
++static unsigned int debug_lvl = 0;
++module_param(log_stats, int, 0644);
++module_param(debug_lvl, int, 0644);
++
++/*
++ * Each outstanding request that we've passed to the lower device layers has a
++ * 'pending_req' allocated to it. Each buffer_head that completes decrements
++ * the pendcnt towards zero. When it hits zero, the specified domain has a
++ * response queued for it, with the saved 'id' passed back.
++ */
++typedef struct {
++ blkif_t *blkif;
++ u64 id;
++ unsigned short mem_idx;
++ int nr_pages;
++ atomic_t pendcnt;
++ unsigned short operation;
++ int status;
++ struct list_head free_list;
++ int inuse;
++} pending_req_t;
++
++static pending_req_t *pending_reqs[MAX_PENDING_REQS];
++static struct list_head pending_free;
++static DEFINE_SPINLOCK(pending_free_lock);
++static DECLARE_WAIT_QUEUE_HEAD (pending_free_wq);
++static int alloc_pending_reqs;
++
++typedef unsigned int PEND_RING_IDX;
++
++static inline int MASK_PEND_IDX(int i) {
++ return (i & (MAX_PENDING_REQS-1));
++}
++
++static inline unsigned int RTN_PEND_IDX(pending_req_t *req, int idx) {
++ return (req - pending_reqs[idx]);
++}
++
++#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
++
++#define BLKBACK_INVALID_HANDLE (~0)
++
++static struct page **foreign_pages[MAX_DYNAMIC_MEM];
++static inline unsigned long idx_to_kaddr(
++ unsigned int mmap_idx, unsigned int req_idx, unsigned int sg_idx)
++{
++ unsigned int arr_idx = req_idx*BLKIF_MAX_SEGMENTS_PER_REQUEST + sg_idx;
++ unsigned long pfn = page_to_pfn(foreign_pages[mmap_idx][arr_idx]);
++ return (unsigned long)pfn_to_kaddr(pfn);
++}
++
++static unsigned short mmap_alloc = 0;
++static unsigned short mmap_lock = 0;
++static unsigned short mmap_inuse = 0;
++
++/******************************************************************
++ * GRANT HANDLES
++ */
++
++/* When using grant tables to map a frame for device access then the
++ * handle returned must be used to unmap the frame. This is needed to
++ * drop the ref count on the frame.
++ */
++struct grant_handle_pair
++{
++ grant_handle_t kernel;
++ grant_handle_t user;
++};
++#define INVALID_GRANT_HANDLE 0xFFFF
++
++static struct grant_handle_pair
++ pending_grant_handles[MAX_DYNAMIC_MEM][MMAP_PAGES];
++#define pending_handle(_id, _idx, _i) \
++ (pending_grant_handles[_id][((_idx) * BLKIF_MAX_SEGMENTS_PER_REQUEST) \
++ + (_i)])
++
++
++static int blktap_read_ufe_ring(tap_blkif_t *info); /*local prototypes*/
++
++#define BLKTAP_MINOR 0 /*/dev/xen/blktap has a dynamic major */
++#define BLKTAP_DEV_DIR "/dev/xen"
++
++static int blktap_major;
++
++/* blktap IOCTLs: */
++#define BLKTAP_IOCTL_KICK_FE 1
++#define BLKTAP_IOCTL_KICK_BE 2 /* currently unused */
++#define BLKTAP_IOCTL_SETMODE 3
++#define BLKTAP_IOCTL_SENDPID 4
++#define BLKTAP_IOCTL_NEWINTF 5
++#define BLKTAP_IOCTL_MINOR 6
++#define BLKTAP_IOCTL_MAJOR 7
++#define BLKTAP_QUERY_ALLOC_REQS 8
++#define BLKTAP_IOCTL_FREEINTF 9
++#define BLKTAP_IOCTL_NEWINTF_EXT 50
++#define BLKTAP_IOCTL_PRINT_IDXS 100
++
++/* blktap switching modes: (Set with BLKTAP_IOCTL_SETMODE) */
++#define BLKTAP_MODE_PASSTHROUGH 0x00000000 /* default */
++#define BLKTAP_MODE_INTERCEPT_FE 0x00000001
++#define BLKTAP_MODE_INTERCEPT_BE 0x00000002 /* unimp. */
++
++#define BLKTAP_MODE_INTERPOSE \
++ (BLKTAP_MODE_INTERCEPT_FE | BLKTAP_MODE_INTERCEPT_BE)
++
++
++static inline int BLKTAP_MODE_VALID(unsigned long arg)
++{
++ return ((arg == BLKTAP_MODE_PASSTHROUGH ) ||
++ (arg == BLKTAP_MODE_INTERCEPT_FE) ||
++ (arg == BLKTAP_MODE_INTERPOSE ));
++}
++
++/* Requests passing through the tap to userspace are re-assigned an ID.
++ * We must record a mapping between the BE [IDX,ID] tuple and the userspace
++ * ring ID.
++ */
++
++static inline unsigned long MAKE_ID(domid_t fe_dom, PEND_RING_IDX idx)
++{
++ return ((fe_dom << 16) | MASK_PEND_IDX(idx));
++}
++
++extern inline PEND_RING_IDX ID_TO_IDX(unsigned long id)
++{
++ return (PEND_RING_IDX)(id & 0x0000ffff);
++}
++
++extern inline int ID_TO_MIDX(unsigned long id)
++{
++ return (int)(id >> 16);
++}
++
++#define INVALID_REQ 0xdead0000
++
++/*TODO: Convert to a free list*/
++static inline int GET_NEXT_REQ(unsigned long *idx_map)
++{
++ int i;
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ if (idx_map[i] == INVALID_REQ)
++ return i;
++
++ return INVALID_REQ;
++}
++
++static inline int OFFSET_TO_USR_IDX(int offset)
++{
++ return offset / BLKIF_MAX_SEGMENTS_PER_REQUEST;
++}
++
++static inline int OFFSET_TO_SEG(int offset)
++{
++ return offset % BLKIF_MAX_SEGMENTS_PER_REQUEST;
++}
++
++
++#define BLKTAP_INVALID_HANDLE(_g) \
++ (((_g->kernel) == INVALID_GRANT_HANDLE) && \
++ ((_g->user) == INVALID_GRANT_HANDLE))
++
++#define BLKTAP_INVALIDATE_HANDLE(_g) do { \
++ (_g)->kernel = INVALID_GRANT_HANDLE; (_g)->user = INVALID_GRANT_HANDLE; \
++ } while(0)
++
++
++/******************************************************************
++ * BLKTAP VM OPS
++ */
++
++static int blktap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
++{
++ /*
++ * if the page has not been mapped in by the driver then return
++ * VM_FAULT_SIGBUS to the domain.
++ */
++
++ return VM_FAULT_SIGBUS;
++}
++
++static pte_t blktap_clear_pte(struct vm_area_struct *vma,
++ unsigned long uvaddr,
++ pte_t *ptep, int is_fullmm)
++{
++ pte_t copy;
++ tap_blkif_t *info;
++ int offset, seg, usr_idx, pending_idx, mmap_idx;
++ unsigned long uvstart = vma->vm_start + (RING_PAGES << PAGE_SHIFT);
++ unsigned long kvaddr;
++ struct page **map;
++ struct page *pg;
++ struct grant_handle_pair *khandle;
++ struct gnttab_unmap_grant_ref unmap[2];
++ int count = 0;
++
++ /*
++ * If the address is before the start of the grant mapped region or
++ * if vm_file is NULL (meaning mmap failed and we have nothing to do)
++ */
++ if (uvaddr < uvstart || vma->vm_file == NULL)
++ return xen_ptep_get_and_clear_full(vma, uvaddr, ptep,
++ is_fullmm);
++
++ info = vma->vm_file->private_data;
++ map = vma->vm_private_data;
++
++ /* TODO Should these be changed to if statements? */
++ BUG_ON(!info);
++ BUG_ON(!info->idx_map);
++ BUG_ON(!map);
++
++ offset = (int) ((uvaddr - uvstart) >> PAGE_SHIFT);
++ usr_idx = OFFSET_TO_USR_IDX(offset);
++ seg = OFFSET_TO_SEG(offset);
++
++ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
++ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
++
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, seg);
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ClearPageReserved(pg);
++ map[offset + RING_PAGES] = NULL;
++
++ khandle = &pending_handle(mmap_idx, pending_idx, seg);
++
++ if (khandle->kernel != INVALID_GRANT_HANDLE) {
++ gnttab_set_unmap_op(&unmap[count], kvaddr,
++ GNTMAP_host_map, khandle->kernel);
++ count++;
++
++ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
++ INVALID_P2M_ENTRY);
++ }
++
++ if (khandle->user != INVALID_GRANT_HANDLE) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++
++ copy = *ptep;
++ gnttab_set_unmap_op(&unmap[count], virt_to_machine(ptep),
++ GNTMAP_host_map
++ | GNTMAP_application_map
++ | GNTMAP_contains_pte,
++ khandle->user);
++ count++;
++ } else {
++ BUG_ON(!xen_feature(XENFEAT_auto_translated_physmap));
++
++ /* USING SHADOW PAGE TABLES. */
++ copy = xen_ptep_get_and_clear_full(vma, uvaddr, ptep,
++ is_fullmm);
++ }
++
++ if (count) {
++ BLKTAP_INVALIDATE_HANDLE(khandle);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
++ unmap, count))
++ BUG();
++ }
++
++ return copy;
++}
++
++struct vm_operations_struct blktap_vm_ops = {
++ fault: blktap_fault,
++ zap_pte: blktap_clear_pte,
++};
++
++/******************************************************************
++ * BLKTAP FILE OPS
++ */
++
++/*Function Declarations*/
++static tap_blkif_t *get_next_free_dev(void);
++static int blktap_open(struct inode *inode, struct file *filp);
++static int blktap_release(struct inode *inode, struct file *filp);
++static int blktap_mmap(struct file *filp, struct vm_area_struct *vma);
++static int blktap_ioctl(struct inode *inode, struct file *filp,
++ unsigned int cmd, unsigned long arg);
++static unsigned int blktap_poll(struct file *file, poll_table *wait);
++
++static const struct file_operations blktap_fops = {
++ .owner = THIS_MODULE,
++ .poll = blktap_poll,
++ .ioctl = blktap_ioctl,
++ .open = blktap_open,
++ .release = blktap_release,
++ .mmap = blktap_mmap,
++};
++
++
++static tap_blkif_t *get_next_free_dev(void)
++{
++ struct class *class;
++ tap_blkif_t *info;
++ int minor;
++
++ /*
++ * This is called only from the ioctl, which
++ * means we should always have interrupts enabled.
++ */
++ BUG_ON(irqs_disabled());
++
++ spin_lock_irq(&pending_free_lock);
++
++ /* tapfds[0] is always NULL */
++
++ for (minor = 1; minor < blktap_next_minor; minor++) {
++ info = tapfds[minor];
++ /* we could have failed a previous attempt. */
++ if (!info ||
++ ((info->dev_inuse == 0) &&
++ (info->dev_pending == 0)) ) {
++ info->dev_pending = 1;
++ goto found;
++ }
++ }
++ info = NULL;
++ minor = -1;
++
++ /*
++ * We didn't find free device. If we can still allocate
++ * more, then we grab the next device minor that is
++ * available. This is done while we are still under
++ * the protection of the pending_free_lock.
++ */
++ if (blktap_next_minor < MAX_TAP_DEV)
++ minor = blktap_next_minor++;
++found:
++ spin_unlock_irq(&pending_free_lock);
++
++ if (!info && minor > 0) {
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (unlikely(!info)) {
++ /*
++ * If we failed here, try to put back
++ * the next minor number. But if one
++ * was just taken, then we just lose this
++ * minor. We can try to allocate this
++ * minor again later.
++ */
++ spin_lock_irq(&pending_free_lock);
++ if (blktap_next_minor == minor+1)
++ blktap_next_minor--;
++ spin_unlock_irq(&pending_free_lock);
++ goto out;
++ }
++
++ info->minor = minor;
++ /*
++ * Make sure that we have a minor before others can
++ * see us.
++ */
++ wmb();
++ tapfds[minor] = info;
++
++ if ((class = get_xen_class()) != NULL)
++ device_create(class, NULL, MKDEV(blktap_major, minor),
++ "blktap%d", minor);
++ }
++
++out:
++ return info;
++}
++
++int dom_to_devid(domid_t domid, int xenbus_id, blkif_t *blkif)
++{
++ tap_blkif_t *info;
++ int i;
++
++ for (i = 1; i < blktap_next_minor; i++) {
++ info = tapfds[i];
++ if ( info &&
++ (info->trans.domid == domid) &&
++ (info->trans.busid == xenbus_id) ) {
++ info->blkif = blkif;
++ info->status = RUNNING;
++ return i;
++ }
++ }
++ return -1;
++}
++
++void signal_tapdisk(int idx)
++{
++ tap_blkif_t *info;
++ struct task_struct *ptask;
++
++ /*
++ * if the userland tools set things up wrong, this could be negative;
++ * just don't try to signal in this case
++ */
++ if (idx < 0)
++ return;
++
++ info = tapfds[idx];
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
++ return;
++
++ if (info->pid > 0) {
++ ptask = find_task_by_pid(info->pid);
++ if (ptask)
++ info->status = CLEANSHUTDOWN;
++ }
++ info->blkif = NULL;
++
++ return;
++}
++
++static int blktap_open(struct inode *inode, struct file *filp)
++{
++ blkif_sring_t *sring;
++ int idx = iminor(inode) - BLKTAP_MINOR;
++ tap_blkif_t *info;
++ int i;
++
++ /* ctrl device, treat differently */
++ if (!idx)
++ return 0;
++
++ info = tapfds[idx];
++
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info) {
++ WPRINTK("Unable to open device /dev/xen/blktap%d\n",
++ idx);
++ return -ENODEV;
++ }
++
++ DPRINTK("Opening device /dev/xen/blktap%d\n",idx);
++
++ /*Only one process can access device at a time*/
++ if (test_and_set_bit(0, &info->dev_inuse))
++ return -EBUSY;
++
++ info->dev_pending = 0;
++
++ /* Allocate the fe ring. */
++ sring = (blkif_sring_t *)get_zeroed_page(GFP_KERNEL);
++ if (sring == NULL)
++ goto fail_nomem;
++
++ SetPageReserved(virt_to_page(sring));
++
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ufe_ring, sring, PAGE_SIZE);
++
++ filp->private_data = info;
++ info->vma = NULL;
++
++ info->idx_map = kmalloc(sizeof(unsigned long) * MAX_PENDING_REQS,
++ GFP_KERNEL);
++
++ if (info->idx_map == NULL)
++ goto fail_nomem;
++
++ if (idx > 0) {
++ init_waitqueue_head(&info->wait);
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ info->idx_map[i] = INVALID_REQ;
++ }
++
++ DPRINTK("Tap open: device /dev/xen/blktap%d\n",idx);
++ return 0;
++
++ fail_nomem:
++ return -ENOMEM;
++}
++
++static int blktap_release(struct inode *inode, struct file *filp)
++{
++ tap_blkif_t *info = filp->private_data;
++
++ /* check for control device */
++ if (!info)
++ return 0;
++
++ info->dev_inuse = 0;
++ DPRINTK("Freeing device [/dev/xen/blktap%d]\n",info->minor);
++
++ /* Free the ring page. */
++ ClearPageReserved(virt_to_page(info->ufe_ring.sring));
++ free_page((unsigned long) info->ufe_ring.sring);
++
++ /* Clear any active mappings and free foreign map table */
++ if (info->vma) {
++ zap_page_range(
++ info->vma, info->vma->vm_start,
++ info->vma->vm_end - info->vma->vm_start, NULL);
++
++ kfree(info->vma->vm_private_data);
++
++ info->vma = NULL;
++ }
++
++ if (info->idx_map) {
++ kfree(info->idx_map);
++ info->idx_map = NULL;
++ }
++
++ if ( (info->status != CLEANSHUTDOWN) && (info->blkif != NULL) ) {
++ if (info->blkif->xenblkd != NULL) {
++ kthread_stop(info->blkif->xenblkd);
++ info->blkif->xenblkd = NULL;
++ }
++ info->status = CLEANSHUTDOWN;
++ }
++
++ return 0;
++}
++
++
++/* Note on mmap:
++ * We need to map pages to user space in a way that will allow the block
++ * subsystem set up direct IO to them. This couldn't be done before, because
++ * there isn't really a sane way to translate a user virtual address down to a
++ * physical address when the page belongs to another domain.
++ *
++ * My first approach was to map the page in to kernel memory, add an entry
++ * for it in the physical frame list (using alloc_lomem_region as in blkback)
++ * and then attempt to map that page up to user space. This is disallowed
++ * by xen though, which realizes that we don't really own the machine frame
++ * underlying the physical page.
++ *
++ * The new approach is to provide explicit support for this in xen linux.
++ * The VMA now has a flag, VM_FOREIGN, to indicate that it contains pages
++ * mapped from other vms. vma->vm_private_data is set up as a mapping
++ * from pages to actual page structs. There is a new clause in get_user_pages
++ * that does the right thing for this sort of mapping.
++ */
++static int blktap_mmap(struct file *filp, struct vm_area_struct *vma)
++{
++ int size;
++ struct page **map;
++ int i;
++ tap_blkif_t *info = filp->private_data;
++ int ret;
++
++ if (info == NULL) {
++ WPRINTK("blktap: mmap, retrieving idx failed\n");
++ return -ENOMEM;
++ }
++
++ vma->vm_flags |= VM_RESERVED;
++ vma->vm_ops = &blktap_vm_ops;
++
++ size = vma->vm_end - vma->vm_start;
++ if (size != ((mmap_pages + RING_PAGES) << PAGE_SHIFT)) {
++ WPRINTK("you _must_ map exactly %d pages!\n",
++ mmap_pages + RING_PAGES);
++ return -EAGAIN;
++ }
++
++ size >>= PAGE_SHIFT;
++ info->rings_vstart = vma->vm_start;
++ info->user_vstart = info->rings_vstart + (RING_PAGES << PAGE_SHIFT);
++
++ /* Map the ring pages to the start of the region and reserve it. */
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ ret = vm_insert_page(vma, vma->vm_start,
++ virt_to_page(info->ufe_ring.sring));
++ else
++ ret = remap_pfn_range(vma, vma->vm_start,
++ __pa(info->ufe_ring.sring) >> PAGE_SHIFT,
++ PAGE_SIZE, vma->vm_page_prot);
++ if (ret) {
++ WPRINTK("Mapping user ring failed!\n");
++ goto fail;
++ }
++
++ /* Mark this VM as containing foreign pages, and set up mappings. */
++ map = kzalloc(((vma->vm_end - vma->vm_start) >> PAGE_SHIFT)
++ * sizeof(struct page *),
++ GFP_KERNEL);
++ if (map == NULL) {
++ WPRINTK("Couldn't alloc VM_FOREIGN map.\n");
++ goto fail;
++ }
++
++ for (i = 0; i < ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT); i++)
++ map[i] = NULL;
++
++ vma->vm_private_data = map;
++ vma->vm_flags |= VM_FOREIGN;
++ vma->vm_flags |= VM_DONTCOPY;
++
++#ifdef CONFIG_X86
++ vma->vm_mm->context.has_foreign_mappings = 1;
++#endif
++
++ info->vma = vma;
++ info->ring_ok = 1;
++ return 0;
++ fail:
++ /* Clear any active mappings. */
++ zap_page_range(vma, vma->vm_start,
++ vma->vm_end - vma->vm_start, NULL);
++
++ return -ENOMEM;
++}
++
++
++static int blktap_ioctl(struct inode *inode, struct file *filp,
++ unsigned int cmd, unsigned long arg)
++{
++ tap_blkif_t *info = filp->private_data;
++
++ switch(cmd) {
++ case BLKTAP_IOCTL_KICK_FE:
++ {
++ /* There are fe messages to process. */
++ return blktap_read_ufe_ring(info);
++ }
++ case BLKTAP_IOCTL_SETMODE:
++ {
++ if (info) {
++ if (BLKTAP_MODE_VALID(arg)) {
++ info->mode = arg;
++ /* XXX: may need to flush rings here. */
++ DPRINTK("blktap: set mode to %lx\n",
++ arg);
++ return 0;
++ }
++ }
++ return 0;
++ }
++ case BLKTAP_IOCTL_PRINT_IDXS:
++ {
++ if (info) {
++ printk("User Rings: \n-----------\n");
++ printk("UF: rsp_cons: %2d, req_prod_prv: %2d "
++ "| req_prod: %2d, rsp_prod: %2d\n",
++ info->ufe_ring.rsp_cons,
++ info->ufe_ring.req_prod_pvt,
++ info->ufe_ring.sring->req_prod,
++ info->ufe_ring.sring->rsp_prod);
++ }
++ return 0;
++ }
++ case BLKTAP_IOCTL_SENDPID:
++ {
++ if (info) {
++ info->pid = (pid_t)arg;
++ DPRINTK("blktap: pid received %d\n",
++ info->pid);
++ }
++ return 0;
++ }
++ case BLKTAP_IOCTL_NEWINTF:
++ {
++ uint64_t val = (uint64_t)arg;
++ domid_translate_t *tr = (domid_translate_t *)&val;
++
++ DPRINTK("NEWINTF Req for domid %d and bus id %d\n",
++ tr->domid, tr->busid);
++ info = get_next_free_dev();
++ if (!info) {
++ WPRINTK("Error initialising /dev/xen/blktap - "
++ "No more devices\n");
++ return -1;
++ }
++ info->trans.domid = tr->domid;
++ info->trans.busid = tr->busid;
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_NEWINTF_EXT:
++ {
++ void __user *udata = (void __user *) arg;
++ domid_translate_ext_t tr;
++
++ if (copy_from_user(&tr, udata, sizeof(domid_translate_ext_t)))
++ return -EFAULT;
++
++ DPRINTK("NEWINTF_EXT Req for domid %d and bus id %d\n",
++ tr.domid, tr.busid);
++ info = get_next_free_dev();
++ if (!info) {
++ WPRINTK("Error initialising /dev/xen/blktap - "
++ "No more devices\n");
++ return -1;
++ }
++ info->trans.domid = tr.domid;
++ info->trans.busid = tr.busid;
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_FREEINTF:
++ {
++ unsigned long dev = arg;
++ unsigned long flags;
++
++ info = tapfds[dev];
++
++ if ((dev > MAX_TAP_DEV) || !info)
++ return 0; /* should this be an error? */
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (info->dev_pending)
++ info->dev_pending = 0;
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++
++ return 0;
++ }
++ case BLKTAP_IOCTL_MINOR:
++ {
++ unsigned long dev = arg;
++
++ info = tapfds[dev];
++
++ if ((dev > MAX_TAP_DEV) || !info)
++ return -EINVAL;
++
++ return info->minor;
++ }
++ case BLKTAP_IOCTL_MAJOR:
++ return blktap_major;
++
++ case BLKTAP_QUERY_ALLOC_REQS:
++ {
++ WPRINTK("BLKTAP_QUERY_ALLOC_REQS ioctl: %d/%d\n",
++ alloc_pending_reqs, blkif_reqs);
++ return (alloc_pending_reqs/blkif_reqs) * 100;
++ }
++ }
++ return -ENOIOCTLCMD;
++}
++
++static unsigned int blktap_poll(struct file *filp, poll_table *wait)
++{
++ tap_blkif_t *info = filp->private_data;
++
++ /* do not work on the control device */
++ if (!info)
++ return 0;
++
++ poll_wait(filp, &info->wait, wait);
++ if (info->ufe_ring.req_prod_pvt != info->ufe_ring.sring->req_prod) {
++ RING_PUSH_REQUESTS(&info->ufe_ring);
++ return POLLIN | POLLRDNORM;
++ }
++ return 0;
++}
++
++void blktap_kick_user(int idx)
++{
++ tap_blkif_t *info;
++
++ info = tapfds[idx];
++
++ if ((idx < 0) || (idx > MAX_TAP_DEV) || !info)
++ return;
++
++ wake_up_interruptible(&info->wait);
++
++ return;
++}
++
++static int do_block_io_op(blkif_t *blkif);
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req);
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st);
++
++/******************************************************************
++ * misc small helpers
++ */
++static int req_increase(void)
++{
++ int i, j;
++
++ if (mmap_alloc >= MAX_PENDING_REQS || mmap_lock)
++ return -EINVAL;
++
++ pending_reqs[mmap_alloc] = kzalloc(sizeof(pending_req_t)
++ * blkif_reqs, GFP_KERNEL);
++ foreign_pages[mmap_alloc] = alloc_empty_pages_and_pagevec(mmap_pages);
++
++ if (!pending_reqs[mmap_alloc] || !foreign_pages[mmap_alloc])
++ goto out_of_memory;
++
++ DPRINTK("%s: reqs=%d, pages=%d\n",
++ __FUNCTION__, blkif_reqs, mmap_pages);
++
++ for (i = 0; i < MAX_PENDING_REQS; i++) {
++ list_add_tail(&pending_reqs[mmap_alloc][i].free_list,
++ &pending_free);
++ pending_reqs[mmap_alloc][i].mem_idx = mmap_alloc;
++ for (j = 0; j < BLKIF_MAX_SEGMENTS_PER_REQUEST; j++)
++ BLKTAP_INVALIDATE_HANDLE(&pending_handle(mmap_alloc,
++ i, j));
++ }
++
++ mmap_alloc++;
++ DPRINTK("# MMAPs increased to %d\n",mmap_alloc);
++ return 0;
++
++ out_of_memory:
++ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
++ kfree(pending_reqs[mmap_alloc]);
++ WPRINTK("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++}
++
++static void mmap_req_del(int mmap)
++{
++ BUG_ON(!spin_is_locked(&pending_free_lock));
++
++ kfree(pending_reqs[mmap]);
++ pending_reqs[mmap] = NULL;
++
++ free_empty_pages_and_pagevec(foreign_pages[mmap_alloc], mmap_pages);
++ foreign_pages[mmap] = NULL;
++
++ mmap_lock = 0;
++ DPRINTK("# MMAPs decreased to %d\n",mmap_alloc);
++ mmap_alloc--;
++}
++
++static pending_req_t* alloc_req(void)
++{
++ pending_req_t *req = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
++ }
++
++ if (req) {
++ req->inuse = 1;
++ alloc_pending_reqs++;
++ }
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++
++ return req;
++}
++
++static void free_req(pending_req_t *req)
++{
++ unsigned long flags;
++ int was_empty;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++
++ alloc_pending_reqs--;
++ req->inuse = 0;
++ if (mmap_lock && (req->mem_idx == mmap_alloc-1)) {
++ mmap_inuse--;
++ if (mmap_inuse == 0) mmap_req_del(mmap_alloc-1);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return;
++ }
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++
++ if (was_empty)
++ wake_up(&pending_free_wq);
++}
++
++static void fast_flush_area(pending_req_t *req, int k_idx, int u_idx,
++ int tapidx)
++{
++ struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
++ unsigned int i, invcount = 0;
++ struct grant_handle_pair *khandle;
++ uint64_t ptep;
++ int ret, mmap_idx;
++ unsigned long kvaddr, uvaddr;
++ tap_blkif_t *info;
++
++
++ info = tapfds[tapidx];
++
++ if ((tapidx < 0) || (tapidx > MAX_TAP_DEV) || !info) {
++ WPRINTK("fast_flush: Couldn't get info!\n");
++ return;
++ }
++
++ if (info->vma != NULL &&
++ xen_feature(XENFEAT_auto_translated_physmap)) {
++ down_write(&info->vma->vm_mm->mmap_sem);
++ zap_page_range(info->vma,
++ MMAP_VADDR(info->user_vstart, u_idx, 0),
++ req->nr_pages << PAGE_SHIFT, NULL);
++ up_write(&info->vma->vm_mm->mmap_sem);
++ return;
++ }
++
++ mmap_idx = req->mem_idx;
++
++ for (i = 0; i < req->nr_pages; i++) {
++ kvaddr = idx_to_kaddr(mmap_idx, k_idx, i);
++ uvaddr = MMAP_VADDR(info->user_vstart, u_idx, i);
++
++ khandle = &pending_handle(mmap_idx, k_idx, i);
++
++ if (khandle->kernel != INVALID_GRANT_HANDLE) {
++ gnttab_set_unmap_op(&unmap[invcount],
++ idx_to_kaddr(mmap_idx, k_idx, i),
++ GNTMAP_host_map, khandle->kernel);
++ invcount++;
++
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(mmap_idx, k_idx, i))
++ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++ }
++
++ if (khandle->user != INVALID_GRANT_HANDLE) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++ if (create_lookup_pte_addr(
++ info->vma->vm_mm,
++ MMAP_VADDR(info->user_vstart, u_idx, i),
++ &ptep) !=0) {
++ WPRINTK("Couldn't get a pte addr!\n");
++ return;
++ }
++
++ gnttab_set_unmap_op(&unmap[invcount], ptep,
++ GNTMAP_host_map
++ | GNTMAP_application_map
++ | GNTMAP_contains_pte,
++ khandle->user);
++ invcount++;
++ }
++
++ BLKTAP_INVALIDATE_HANDLE(khandle);
++ }
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(ret);
++
++ if (info->vma != NULL && !xen_feature(XENFEAT_auto_translated_physmap))
++ zap_page_range(info->vma,
++ MMAP_VADDR(info->user_vstart, u_idx, 0),
++ req->nr_pages << PAGE_SHIFT, NULL);
++}
++
++/******************************************************************
++ * SCHEDULER FUNCTIONS
++ */
++
++static void print_stats(blkif_t *blkif)
++{
++ printk(KERN_DEBUG "%s: oo %3d | rd %4d | wr %4d\n",
++ current->comm, blkif->st_oo_req,
++ blkif->st_rd_req, blkif->st_wr_req);
++ blkif->st_print = jiffies + msecs_to_jiffies(10 * 1000);
++ blkif->st_rd_req = 0;
++ blkif->st_wr_req = 0;
++ blkif->st_oo_req = 0;
++}
++
++int tap_blkif_schedule(void *arg)
++{
++ blkif_t *blkif = arg;
++
++ blkif_get(blkif);
++
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: started\n", current->comm);
++
++ while (!kthread_should_stop()) {
++ if (try_to_freeze())
++ continue;
++
++ wait_event_interruptible(
++ blkif->wq,
++ blkif->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
++
++ blkif->waiting_reqs = 0;
++ smp_mb(); /* clear flag *before* checking for work */
++
++ if (do_block_io_op(blkif))
++ blkif->waiting_reqs = 1;
++
++ if (log_stats && time_after(jiffies, blkif->st_print))
++ print_stats(blkif);
++ }
++
++ if (log_stats)
++ print_stats(blkif);
++ if (debug_lvl)
++ printk(KERN_DEBUG "%s: exiting\n", current->comm);
++
++ blkif->xenblkd = NULL;
++ blkif_put(blkif);
++
++ return 0;
++}
++
++/******************************************************************
++ * COMPLETION CALLBACK -- Called by user level ioctl()
++ */
++
++static int blktap_read_ufe_ring(tap_blkif_t *info)
++{
++ /* This is called to read responses from the UFE ring. */
++ RING_IDX i, j, rp;
++ blkif_response_t *resp;
++ blkif_t *blkif=NULL;
++ int pending_idx, usr_idx, mmap_idx;
++ pending_req_t *pending_req;
++
++ if (!info)
++ return 0;
++
++ /* We currently only forward packets in INTERCEPT_FE mode. */
++ if (!(info->mode & BLKTAP_MODE_INTERCEPT_FE))
++ return 0;
++
++ /* for each outstanding message on the UFEring */
++ rp = info->ufe_ring.sring->rsp_prod;
++ rmb();
++
++ for (i = info->ufe_ring.rsp_cons; i != rp; i++) {
++ blkif_response_t res;
++ resp = RING_GET_RESPONSE(&info->ufe_ring, i);
++ memcpy(&res, resp, sizeof(res));
++ mb(); /* rsp_cons read by RING_FULL() in do_block_io_op(). */
++ ++info->ufe_ring.rsp_cons;
++
++ /*retrieve [usr_idx] to [mmap_idx,pending_idx] mapping*/
++ usr_idx = (int)res.id;
++ pending_idx = MASK_PEND_IDX(ID_TO_IDX(info->idx_map[usr_idx]));
++ mmap_idx = ID_TO_MIDX(info->idx_map[usr_idx]);
++
++ if ( (mmap_idx >= mmap_alloc) ||
++ (ID_TO_IDX(info->idx_map[usr_idx]) >= MAX_PENDING_REQS) )
++ WPRINTK("Incorrect req map"
++ "[%d], internal map [%d,%d (%d)]\n",
++ usr_idx, mmap_idx,
++ ID_TO_IDX(info->idx_map[usr_idx]),
++ MASK_PEND_IDX(
++ ID_TO_IDX(info->idx_map[usr_idx])));
++
++ pending_req = &pending_reqs[mmap_idx][pending_idx];
++ blkif = pending_req->blkif;
++
++ for (j = 0; j < pending_req->nr_pages; j++) {
++
++ unsigned long kvaddr, uvaddr;
++ struct page **map = info->vma->vm_private_data;
++ struct page *pg;
++ int offset;
++
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, j);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, j);
++
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ClearPageReserved(pg);
++ offset = (uvaddr - info->vma->vm_start)
++ >> PAGE_SHIFT;
++ map[offset] = NULL;
++ }
++ fast_flush_area(pending_req, pending_idx, usr_idx, info->minor);
++ info->idx_map[usr_idx] = INVALID_REQ;
++ make_response(blkif, pending_req->id, res.operation,
++ res.status);
++ blkif_put(pending_req->blkif);
++ free_req(pending_req);
++ }
++
++ return 0;
++}
++
++
++/******************************************************************************
++ * NOTIFICATION FROM GUEST OS.
++ */
++
++static void blkif_notify_work(blkif_t *blkif)
++{
++ blkif->waiting_reqs = 1;
++ wake_up(&blkif->wq);
++}
++
++irqreturn_t tap_blkif_be_int(int irq, void *dev_id)
++{
++ blkif_notify_work(dev_id);
++ return IRQ_HANDLED;
++}
++
++
++
++/******************************************************************
++ * DOWNWARD CALLS -- These interface with the block-device layer proper.
++ */
++static int print_dbug = 1;
++static int do_block_io_op(blkif_t *blkif)
++{
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ blkif_request_t req;
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int more_to_do = 0;
++ tap_blkif_t *info;
++
++ rc = blk_rings->common.req_cons;
++ rp = blk_rings->common.sring->req_prod;
++ rmb(); /* Ensure we see queued requests up to 'rp'. */
++
++ /*Check blkif has corresponding UE ring*/
++ if (blkif->dev_num < 0) {
++ /*oops*/
++ if (print_dbug) {
++ WPRINTK("Corresponding UE "
++ "ring does not exist!\n");
++ print_dbug = 0; /*We only print this message once*/
++ }
++ return 0;
++ }
++
++ info = tapfds[blkif->dev_num];
++
++ if (blkif->dev_num > MAX_TAP_DEV || !info || !info->dev_inuse) {
++ if (print_dbug) {
++ WPRINTK("Can't get UE info!\n");
++ print_dbug = 0;
++ }
++ return 0;
++ }
++
++ while (rc != rp) {
++
++ if (RING_FULL(&info->ufe_ring)) {
++ WPRINTK("RING_FULL! More to do\n");
++ more_to_do = 1;
++ break;
++ }
++
++ if (RING_REQUEST_CONS_OVERFLOW(&blk_rings->common, rc)) {
++ WPRINTK("RING_REQUEST_CONS_OVERFLOW!"
++ " More to do\n");
++ more_to_do = 1;
++ break;
++ }
++
++ pending_req = alloc_req();
++ if (NULL == pending_req) {
++ blkif->st_oo_req++;
++ more_to_do = 1;
++ break;
++ }
++
++ if (kthread_should_stop()) {
++ more_to_do = 1;
++ break;
++ }
++
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(&req, RING_GET_REQUEST(&blk_rings->native, rc),
++ sizeof(req));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ blkif_get_x86_32_req(&req, RING_GET_REQUEST(&blk_rings->x86_32, rc));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ blkif_get_x86_64_req(&req, RING_GET_REQUEST(&blk_rings->x86_64, rc));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.req_cons = ++rc; /* before make_response() */
++
++ /* Apply all sanity checks to /private copy/ of request. */
++ barrier();
++
++ switch (req.operation) {
++ case BLKIF_OP_READ:
++ blkif->st_rd_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++
++ case BLKIF_OP_WRITE:
++ blkif->st_wr_req++;
++ dispatch_rw_block_io(blkif, &req, pending_req);
++ break;
++
++ default:
++ /* A good sign something is wrong: sleep for a while to
++ * avoid excessive CPU consumption by a bad guest. */
++ msleep(1);
++ WPRINTK("unknown operation [%d]\n",
++ req.operation);
++ make_response(blkif, req.id, req.operation,
++ BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ break;
++ }
++
++ /* Yield point for this unbounded loop. */
++ cond_resched();
++ }
++
++ blktap_kick_user(blkif->dev_num);
++
++ return more_to_do;
++}
++
++static void dispatch_rw_block_io(blkif_t *blkif,
++ blkif_request_t *req,
++ pending_req_t *pending_req)
++{
++ extern void ll_rw_block(int rw, int nr, struct buffer_head * bhs[]);
++ int op, operation = (req->operation == BLKIF_OP_WRITE) ? WRITE : READ;
++ struct gnttab_map_grant_ref map[BLKIF_MAX_SEGMENTS_PER_REQUEST*2];
++ unsigned int nseg;
++ int ret, i, nr_sects = 0;
++ tap_blkif_t *info;
++ blkif_request_t *target;
++ int pending_idx = RTN_PEND_IDX(pending_req,pending_req->mem_idx);
++ int usr_idx;
++ uint16_t mmap_idx = pending_req->mem_idx;
++
++ if (blkif->dev_num < 0 || blkif->dev_num > MAX_TAP_DEV)
++ goto fail_response;
++
++ info = tapfds[blkif->dev_num];
++ if (info == NULL)
++ goto fail_response;
++
++ /* Check we have space on user ring - should never fail. */
++ usr_idx = GET_NEXT_REQ(info->idx_map);
++ if (usr_idx == INVALID_REQ) {
++ BUG();
++ goto fail_response;
++ }
++
++ /* Check that number of segments is sane. */
++ nseg = req->nr_segments;
++ if ( unlikely(nseg == 0) ||
++ unlikely(nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST) ) {
++ WPRINTK("Bad number of segments in request (%d)\n", nseg);
++ goto fail_response;
++ }
++
++ /* Make sure userspace is ready. */
++ if (!info->ring_ok) {
++ WPRINTK("blktap: ring not ready for requests!\n");
++ goto fail_response;
++ }
++
++ if (RING_FULL(&info->ufe_ring)) {
++ WPRINTK("blktap: fe_ring is full, can't add "
++ "IO Request will be dropped. %d %d\n",
++ RING_SIZE(&info->ufe_ring),
++ RING_SIZE(&blkif->blk_rings.common));
++ goto fail_response;
++ }
++
++ pending_req->blkif = blkif;
++ pending_req->id = req->id;
++ pending_req->operation = operation;
++ pending_req->status = BLKIF_RSP_OKAY;
++ pending_req->nr_pages = nseg;
++ op = 0;
++ for (i = 0; i < nseg; i++) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ uint64_t ptep;
++ uint32_t flags;
++
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++
++ flags = GNTMAP_host_map;
++ if (operation == WRITE)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[op], kvaddr, flags,
++ req->seg[i].gref, blkif->domid);
++ op++;
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Now map it to user. */
++ ret = create_lookup_pte_addr(info->vma->vm_mm,
++ uvaddr, &ptep);
++ if (ret) {
++ WPRINTK("Couldn't get a pte addr!\n");
++ goto fail_flush;
++ }
++
++ flags = GNTMAP_host_map | GNTMAP_application_map
++ | GNTMAP_contains_pte;
++ if (operation == WRITE)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[op], ptep, flags,
++ req->seg[i].gref, blkif->domid);
++ op++;
++ }
++
++ nr_sects += (req->seg[i].last_sect -
++ req->seg[i].first_sect + 1);
++ }
++
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, op);
++ BUG_ON(ret);
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ for (i = 0; i < (nseg*2); i+=2) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ unsigned long offset;
++ struct page *pg;
++
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i/2);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i/2);
++
++ if (unlikely(map[i].status != 0)) {
++ WPRINTK("invalid kernel buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i].handle = INVALID_GRANT_HANDLE;
++ }
++
++ if (unlikely(map[i+1].status != 0)) {
++ WPRINTK("invalid user buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i+1].handle = INVALID_GRANT_HANDLE;
++ }
++
++ pending_handle(mmap_idx, pending_idx, i/2).kernel
++ = map[i].handle;
++ pending_handle(mmap_idx, pending_idx, i/2).user
++ = map[i+1].handle;
++
++ if (ret)
++ continue;
++
++ set_phys_to_machine(__pa(kvaddr) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr
++ >> PAGE_SHIFT));
++ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ((struct page **)info->vma->vm_private_data)[offset] =
++ pg;
++ }
++ } else {
++ for (i = 0; i < nseg; i++) {
++ unsigned long uvaddr;
++ unsigned long kvaddr;
++ unsigned long offset;
++ struct page *pg;
++
++ uvaddr = MMAP_VADDR(info->user_vstart, usr_idx, i);
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++
++ if (unlikely(map[i].status != 0)) {
++ WPRINTK("invalid kernel buffer -- "
++ "could not remap it\n");
++ ret |= 1;
++ map[i].handle = INVALID_GRANT_HANDLE;
++ }
++
++ pending_handle(mmap_idx, pending_idx, i).kernel
++ = map[i].handle;
++
++ if (ret)
++ continue;
++
++ offset = (uvaddr - info->vma->vm_start) >> PAGE_SHIFT;
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ ((struct page **)info->vma->vm_private_data)[offset] =
++ pg;
++ }
++ }
++
++ if (ret)
++ goto fail_flush;
++
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ down_write(&info->vma->vm_mm->mmap_sem);
++ /* Mark mapped pages as reserved: */
++ for (i = 0; i < req->nr_segments; i++) {
++ unsigned long kvaddr;
++ struct page *pg;
++
++ kvaddr = idx_to_kaddr(mmap_idx, pending_idx, i);
++ pg = pfn_to_page(__pa(kvaddr) >> PAGE_SHIFT);
++ SetPageReserved(pg);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ ret = vm_insert_page(info->vma,
++ MMAP_VADDR(info->user_vstart,
++ usr_idx, i), pg);
++ if (ret) {
++ up_write(&info->vma->vm_mm->mmap_sem);
++ goto fail_flush;
++ }
++ }
++ }
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ up_write(&info->vma->vm_mm->mmap_sem);
++
++ /*record [mmap_idx,pending_idx] to [usr_idx] mapping*/
++ info->idx_map[usr_idx] = MAKE_ID(mmap_idx, pending_idx);
++
++ blkif_get(blkif);
++ /* Finally, write the request message to the user ring. */
++ target = RING_GET_REQUEST(&info->ufe_ring,
++ info->ufe_ring.req_prod_pvt);
++ memcpy(target, req, sizeof(*req));
++ target->id = usr_idx;
++ wmb(); /* blktap_poll() reads req_prod_pvt asynchronously */
++ info->ufe_ring.req_prod_pvt++;
++
++ if (operation == READ)
++ blkif->st_rd_sect += nr_sects;
++ else if (operation == WRITE)
++ blkif->st_wr_sect += nr_sects;
++
++ return;
++
++ fail_flush:
++ WPRINTK("Reached Fail_flush\n");
++ fast_flush_area(pending_req, pending_idx, usr_idx, blkif->dev_num);
++ fail_response:
++ make_response(blkif, req->id, req->operation, BLKIF_RSP_ERROR);
++ free_req(pending_req);
++ msleep(1); /* back off a bit */
++}
++
++
++
++/******************************************************************
++ * MISCELLANEOUS SETUP / TEARDOWN / DEBUGGING
++ */
++
++
++static void make_response(blkif_t *blkif, u64 id,
++ unsigned short op, int st)
++{
++ blkif_response_t resp;
++ unsigned long flags;
++ blkif_back_rings_t *blk_rings = &blkif->blk_rings;
++ int more_to_do = 0;
++ int notify;
++
++ resp.id = id;
++ resp.operation = op;
++ resp.status = st;
++
++ spin_lock_irqsave(&blkif->blk_ring_lock, flags);
++ /* Place on the response ring for the relevant domain. */
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ memcpy(RING_GET_RESPONSE(&blk_rings->native,
++ blk_rings->native.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_32:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_32,
++ blk_rings->x86_32.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ case BLKIF_PROTOCOL_X86_64:
++ memcpy(RING_GET_RESPONSE(&blk_rings->x86_64,
++ blk_rings->x86_64.rsp_prod_pvt),
++ &resp, sizeof(resp));
++ break;
++ default:
++ BUG();
++ }
++ blk_rings->common.rsp_prod_pvt++;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify);
++
++ if (blk_rings->common.rsp_prod_pvt == blk_rings->common.req_cons) {
++ /*
++ * Tail check for pending requests. Allows frontend to avoid
++ * notifications if requests are already in flight (lower
++ * overheads and promotes batching).
++ */
++ RING_FINAL_CHECK_FOR_REQUESTS(&blk_rings->common, more_to_do);
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&blk_rings->common)) {
++ more_to_do = 1;
++ }
++
++ spin_unlock_irqrestore(&blkif->blk_ring_lock, flags);
++ if (more_to_do)
++ blkif_notify_work(blkif);
++ if (notify)
++ notify_remote_via_irq(blkif->irq);
++}
++
++static int __init blkif_init(void)
++{
++ int i, ret;
++ struct class *class;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ INIT_LIST_HEAD(&pending_free);
++ for(i = 0; i < 2; i++) {
++ ret = req_increase();
++ if (ret)
++ break;
++ }
++ if (i == 0)
++ return ret;
++
++ tap_blkif_interface_init();
++
++ alloc_pending_reqs = 0;
++
++ tap_blkif_xenbus_init();
++
++ /* Dynamically allocate a major for this device */
++ ret = register_chrdev(0, "blktap", &blktap_fops);
++
++ if (ret < 0) {
++ WPRINTK("Couldn't register /dev/xen/blktap\n");
++ return -ENOMEM;
++ }
++
++ blktap_major = ret;
++
++ /* tapfds[0] is always NULL */
++ blktap_next_minor++;
++
++ DPRINTK("Created misc_dev [/dev/xen/blktap%d]\n",i);
++
++ /* Make sure the xen class exists */
++ if ((class = get_xen_class()) != NULL) {
++ /*
++ * This will allow udev to create the blktap ctrl device.
++ * We only want to create blktap0 first. We don't want
++ * to flood the sysfs system with needless blktap devices.
++ * We only create the device when a request of a new device is
++ * made.
++ */
++ device_create(class, NULL, MKDEV(blktap_major, 0), "blktap0");
++ } else {
++ /* this is bad, but not fatal */
++ WPRINTK("blktap: sysfs xen_class not created\n");
++ }
++
++ DPRINTK("Blktap device successfully created\n");
++
++ return 0;
++}
++
++module_init(blkif_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/blktap/common.h b/drivers/xen/blktap/common.h
+new file mode 100644
+index 0000000..fc7e013
+--- /dev/null
++++ b/drivers/xen/blktap/common.h
+@@ -0,0 +1,121 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __BLKIF__BACKEND__COMMON_H__
++#define __BLKIF__BACKEND__COMMON_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/blkdev.h>
++#include <linux/vmalloc.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/blkif.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++
++#define DPRINTK(_f, _a...) pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++#define WPRINTK(fmt, args...) printk(KERN_WARNING "blk_tap: " fmt, ##args)
++
++struct backend_info;
++
++typedef struct blkif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
++ /* Comms information. */
++ enum blkif_protocol blk_protocol;
++ blkif_back_rings_t blk_rings;
++ struct vm_struct *blk_ring_area;
++ /* Back pointer to the backend_info. */
++ struct backend_info *be;
++ /* Private fields. */
++ spinlock_t blk_ring_lock;
++ atomic_t refcnt;
++
++ wait_queue_head_t wq;
++ struct task_struct *xenblkd;
++ unsigned int waiting_reqs;
++ struct request_queue *plug;
++
++ /* statistics */
++ unsigned long st_print;
++ int st_rd_req;
++ int st_wr_req;
++ int st_oo_req;
++ int st_rd_sect;
++ int st_wr_sect;
++
++ wait_queue_head_t waiting_to_free;
++
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++
++ int dev_num;
++ uint64_t sectors;
++} blkif_t;
++
++blkif_t *tap_alloc_blkif(domid_t domid);
++void tap_blkif_free(blkif_t *blkif);
++int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
++ unsigned int evtchn);
++void tap_blkif_unmap(blkif_t *blkif);
++
++#define blkif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define blkif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
++
++
++struct phys_req {
++ unsigned short dev;
++ unsigned short nr_sects;
++ struct block_device *bdev;
++ blkif_sector_t sector_number;
++};
++
++void tap_blkif_interface_init(void);
++
++void tap_blkif_xenbus_init(void);
++
++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 --git a/drivers/xen/blktap/interface.c b/drivers/xen/blktap/interface.c
+new file mode 100644
+index 0000000..fc662bd
+--- /dev/null
++++ b/drivers/xen/blktap/interface.c
+@@ -0,0 +1,174 @@
++/******************************************************************************
++ * drivers/xen/blktap/interface.c
++ *
++ * Block-device interface management.
++ *
++ * Copyright (c) 2004, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++
++ */
++
++#include "common.h"
++#include <xen/evtchn.h>
++
++static struct kmem_cache *blkif_cachep;
++
++blkif_t *tap_alloc_blkif(domid_t domid)
++{
++ blkif_t *blkif;
++
++ blkif = kmem_cache_alloc(blkif_cachep, GFP_KERNEL);
++ if (!blkif)
++ return ERR_PTR(-ENOMEM);
++
++ memset(blkif, 0, sizeof(*blkif));
++ blkif->domid = domid;
++ spin_lock_init(&blkif->blk_ring_lock);
++ atomic_set(&blkif->refcnt, 1);
++ init_waitqueue_head(&blkif->wq);
++ blkif->st_print = jiffies;
++ init_waitqueue_head(&blkif->waiting_to_free);
++
++ return blkif;
++}
++
++static int map_frontend_page(blkif_t *blkif, unsigned long shared_page)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, shared_page, blkif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
++ }
++
++ blkif->shmem_ref = shared_page;
++ blkif->shmem_handle = op.handle;
++
++ return 0;
++}
++
++static void unmap_frontend_page(blkif_t *blkif)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr,
++ GNTMAP_host_map, blkif->shmem_handle);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
++
++int tap_blkif_map(blkif_t *blkif, unsigned long shared_page,
++ unsigned int evtchn)
++{
++ int err;
++
++ /* Already connected through? */
++ if (blkif->irq)
++ return 0;
++
++ if ( (blkif->blk_ring_area = alloc_vm_area(PAGE_SIZE)) == NULL )
++ return -ENOMEM;
++
++ err = map_frontend_page(blkif, shared_page);
++ if (err) {
++ free_vm_area(blkif->blk_ring_area);
++ return err;
++ }
++
++ switch (blkif->blk_protocol) {
++ case BLKIF_PROTOCOL_NATIVE:
++ {
++ blkif_sring_t *sring;
++ sring = (blkif_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.native, sring, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_32:
++ {
++ blkif_x86_32_sring_t *sring_x86_32;
++ sring_x86_32 = (blkif_x86_32_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_32, sring_x86_32, PAGE_SIZE);
++ break;
++ }
++ case BLKIF_PROTOCOL_X86_64:
++ {
++ blkif_x86_64_sring_t *sring_x86_64;
++ sring_x86_64 = (blkif_x86_64_sring_t *)blkif->blk_ring_area->addr;
++ BACK_RING_INIT(&blkif->blk_rings.x86_64, sring_x86_64, PAGE_SIZE);
++ break;
++ }
++ default:
++ BUG();
++ }
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ blkif->domid, evtchn, tap_blkif_be_int,
++ 0, "blkif-backend", blkif);
++ if (err < 0) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ return err;
++ }
++ blkif->irq = err;
++
++ return 0;
++}
++
++void tap_blkif_unmap(blkif_t *blkif)
++{
++ if (blkif->irq) {
++ unbind_from_irqhandler(blkif->irq, blkif);
++ blkif->irq = 0;
++ }
++ if (blkif->blk_rings.common.sring) {
++ unmap_frontend_page(blkif);
++ free_vm_area(blkif->blk_ring_area);
++ blkif->blk_rings.common.sring = NULL;
++ }
++}
++
++void tap_blkif_free(blkif_t *blkif)
++{
++ atomic_dec(&blkif->refcnt);
++ wait_event(blkif->waiting_to_free, atomic_read(&blkif->refcnt) == 0);
++
++ tap_blkif_unmap(blkif);
++ kmem_cache_free(blkif_cachep, blkif);
++}
++
++void __init tap_blkif_interface_init(void)
++{
++ blkif_cachep = kmem_cache_create("blktapif_cache", sizeof(blkif_t),
++ 0, 0, NULL);
++}
+diff --git a/drivers/xen/blktap/xenbus.c b/drivers/xen/blktap/xenbus.c
+new file mode 100644
+index 0000000..197b10a
+--- /dev/null
++++ b/drivers/xen/blktap/xenbus.c
+@@ -0,0 +1,477 @@
++/* drivers/xen/blktap/xenbus.c
++ *
++ * Xenbus code for blktap
++ *
++ * Copyright (c) 2004-2005, Andrew Warfield and Julian Chesterfield
++ *
++ * Based on the blkback xenbus code:
++ *
++ * Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ * Copyright (C) 2005 XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <stdarg.h>
++#include <linux/module.h>
++#include <linux/kthread.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++
++struct backend_info
++{
++ struct xenbus_device *dev;
++ blkif_t *blkif;
++ struct xenbus_watch backend_watch;
++ int xenbus_id;
++ int group_added;
++};
++
++
++static void connect(struct backend_info *);
++static int connect_ring(struct backend_info *);
++static int blktap_remove(struct xenbus_device *dev);
++static int blktap_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id);
++static void tap_backend_changed(struct xenbus_watch *, const char **,
++ unsigned int);
++static void tap_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state);
++
++static int strsep_len(const char *str, char c, unsigned int len)
++{
++ unsigned int i;
++
++ for (i = 0; str[i]; i++)
++ if (str[i] == c) {
++ if (len == 0)
++ return i;
++ len--;
++ }
++ return (len == 0) ? i : -ERANGE;
++}
++
++static long get_id(const char *str)
++{
++ int len,end;
++ const char *ptr;
++ char *tptr, num[10];
++
++ len = strsep_len(str, '/', 2);
++ end = strlen(str);
++ if ( (len < 0) || (end < 0) ) return -1;
++
++ ptr = str + len + 1;
++ strncpy(num,ptr,end - len);
++ tptr = num + (end - (len + 1));
++ *tptr = '\0';
++ DPRINTK("Get_id called for %s (%s)\n",str,num);
++
++ return simple_strtol(num, NULL, 10);
++}
++
++static int blktap_name(blkif_t *blkif, char *buf)
++{
++ char *devpath, *devname;
++ struct xenbus_device *dev = blkif->be->dev;
++
++ devpath = xenbus_read(XBT_NIL, dev->nodename, "dev", NULL);
++ if (IS_ERR(devpath))
++ return PTR_ERR(devpath);
++
++ if ((devname = strstr(devpath, "/dev/")) != NULL)
++ devname += strlen("/dev/");
++ else
++ devname = devpath;
++
++ snprintf(buf, TASK_COMM_LEN, "blktap.%d.%s", blkif->domid, devname);
++ kfree(devpath);
++
++ return 0;
++}
++
++/****************************************************************
++ * sysfs interface for I/O requests of blktap device
++ */
++
++#define VBD_SHOW(name, format, args...) \
++ static ssize_t show_##name(struct device *_dev, \
++ struct device_attribute *attr, \
++ char *buf) \
++ { \
++ struct xenbus_device *dev = to_xenbus_device(_dev); \
++ struct backend_info *be = dev->dev.driver_data; \
++ \
++ return sprintf(buf, format, ##args); \
++ } \
++ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL)
++
++VBD_SHOW(oo_req, "%d\n", be->blkif->st_oo_req);
++VBD_SHOW(rd_req, "%d\n", be->blkif->st_rd_req);
++VBD_SHOW(wr_req, "%d\n", be->blkif->st_wr_req);
++VBD_SHOW(rd_sect, "%d\n", be->blkif->st_rd_sect);
++VBD_SHOW(wr_sect, "%d\n", be->blkif->st_wr_sect);
++
++static struct attribute *tapstat_attrs[] = {
++ &dev_attr_oo_req.attr,
++ &dev_attr_rd_req.attr,
++ &dev_attr_wr_req.attr,
++ &dev_attr_rd_sect.attr,
++ &dev_attr_wr_sect.attr,
++ NULL
++};
++
++static struct attribute_group tapstat_group = {
++ .name = "statistics",
++ .attrs = tapstat_attrs,
++};
++
++int xentap_sysfs_addif(struct xenbus_device *dev)
++{
++ int err;
++ struct backend_info *be = dev->dev.driver_data;
++ err = sysfs_create_group(&dev->dev.kobj, &tapstat_group);
++ if (!err)
++ be->group_added = 1;
++ return err;
++}
++
++void xentap_sysfs_delif(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ sysfs_remove_group(&dev->dev.kobj, &tapstat_group);
++ be->group_added = 0;
++}
++
++static int blktap_remove(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ if (be->group_added)
++ xentap_sysfs_delif(be->dev);
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
++ if (be->blkif) {
++ if (be->blkif->xenblkd)
++ kthread_stop(be->blkif->xenblkd);
++ signal_tapdisk(be->blkif->dev_num);
++ tap_blkif_free(be->blkif);
++ be->blkif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
++}
++
++static void tap_update_blkif_status(blkif_t *blkif)
++{
++ int err;
++ char name[TASK_COMM_LEN];
++
++ /* Not ready to connect? */
++ if(!blkif->irq || !blkif->sectors) {
++ return;
++ }
++
++ /* Already connected? */
++ if (blkif->be->dev->state == XenbusStateConnected)
++ return;
++
++ /* Attempt to connect: exit if we fail to. */
++ connect(blkif->be);
++ if (blkif->be->dev->state != XenbusStateConnected)
++ return;
++
++ err = blktap_name(blkif, name);
++ if (err) {
++ xenbus_dev_error(blkif->be->dev, err, "get blktap dev name");
++ return;
++ }
++
++ if (!blkif->be->group_added) {
++ err = xentap_sysfs_addif(blkif->be->dev);
++ if (err) {
++ xenbus_dev_fatal(blkif->be->dev, err,
++ "creating sysfs entries");
++ return;
++ }
++ }
++
++ blkif->xenblkd = kthread_run(tap_blkif_schedule, blkif, name);
++ if (IS_ERR(blkif->xenblkd)) {
++ err = PTR_ERR(blkif->xenblkd);
++ blkif->xenblkd = NULL;
++ xenbus_dev_fatal(blkif->be->dev, err, "start xenblkd");
++ WPRINTK("Error starting thread\n");
++ }
++}
++
++/**
++ * Entry point to this code when a new device is created. Allocate
++ * the basic structures, and watch the store waiting for the
++ * user-space program to tell us the physical device info. Switch to
++ * InitWait.
++ */
++static int blktap_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++
++ be->dev = dev;
++ dev->dev.driver_data = be;
++ be->xenbus_id = get_id(dev->nodename);
++
++ be->blkif = tap_alloc_blkif(dev->otherend_id);
++ if (IS_ERR(be->blkif)) {
++ err = PTR_ERR(be->blkif);
++ be->blkif = NULL;
++ xenbus_dev_fatal(dev, err, "creating block interface");
++ goto fail;
++ }
++
++ /* setup back pointer */
++ be->blkif->be = be;
++ be->blkif->sectors = 0;
++
++ /* set a watch on disk info, waiting for userspace to update details*/
++ err = xenbus_watch_path2(dev, dev->nodename, "info",
++ &be->backend_watch, tap_backend_changed);
++ if (err)
++ goto fail;
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++ return 0;
++
++fail:
++ DPRINTK("blktap probe failed\n");
++ blktap_remove(dev);
++ return err;
++}
++
++
++/**
++ * Callback received when the user space code has placed the device
++ * information in xenstore.
++ */
++static void tap_backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int err;
++ unsigned long info;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
++
++ /**
++ * Check to see whether userspace code has opened the image
++ * and written sector
++ * and disk info to xenstore
++ */
++ err = xenbus_gather(XBT_NIL, dev->nodename, "info", "%lu", &info,
++ NULL);
++ if (XENBUS_EXIST_ERR(err))
++ return;
++ if (err) {
++ xenbus_dev_error(dev, err, "getting info");
++ return;
++ }
++
++ DPRINTK("Userspace update on disk info, %lu\n",info);
++
++ err = xenbus_gather(XBT_NIL, dev->nodename, "sectors", "%llu",
++ &be->blkif->sectors, NULL);
++
++ /* Associate tap dev with domid*/
++ be->blkif->dev_num = dom_to_devid(be->blkif->domid, be->xenbus_id,
++ be->blkif);
++ DPRINTK("Thread started for domid [%d], connecting disk\n",
++ be->blkif->dev_num);
++
++ tap_update_blkif_status(be->blkif);
++}
++
++/**
++ * Callback received when the frontend's state changes.
++ */
++static void tap_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
++
++ DPRINTK("\n");
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
++
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ /* Ensure we connect even when two watches fire in
++ close successsion and we miss the intermediate value
++ of frontend_state. */
++ if (dev->state == XenbusStateConnected)
++ break;
++
++ err = connect_ring(be);
++ if (err)
++ break;
++ tap_update_blkif_status(be->blkif);
++ break;
++
++ case XenbusStateClosing:
++ if (be->blkif->xenblkd) {
++ kthread_stop(be->blkif->xenblkd);
++ be->blkif->xenblkd = NULL;
++ }
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
++}
++
++
++/**
++ * Switch to Connected state.
++ */
++static void connect(struct backend_info *be)
++{
++ int err;
++
++ struct xenbus_device *dev = be->dev;
++
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(dev, err, "switching to Connected state",
++ dev->nodename);
++
++ return;
++}
++
++
++static int connect_ring(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ char protocol[64];
++ int err;
++
++ DPRINTK("%s\n", dev->otherend);
++
++ err = xenbus_gather(XBT_NIL, dev->otherend, "ring-ref", "%lu",
++ &ring_ref, "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
++
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ err = xenbus_gather(XBT_NIL, dev->otherend, "protocol",
++ "%63s", protocol, NULL);
++ if (err)
++ strcpy(protocol, "unspecified, assuming native");
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_NATIVE))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_NATIVE;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_32))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_32;
++ else if (0 == strcmp(protocol, XEN_IO_PROTO_ABI_X86_64))
++ be->blkif->blk_protocol = BLKIF_PROTOCOL_X86_64;
++ else {
++ xenbus_dev_fatal(dev, err, "unknown fe protocol %s", protocol);
++ return -1;
++ }
++ printk(KERN_INFO
++ "blktap: ring-ref %ld, event-channel %d, protocol %d (%s)\n",
++ ring_ref, evtchn, be->blkif->blk_protocol, protocol);
++
++ /* Map the shared frame, irq etc. */
++ err = tap_blkif_map(be->blkif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "mapping ring-ref %lu port %u",
++ ring_ref, evtchn);
++ return err;
++ }
++
++ return 0;
++}
++
++
++/* ** Driver Registration ** */
++
++
++static const struct xenbus_device_id blktap_ids[] = {
++ { "tap" },
++ { "" }
++};
++
++
++static struct xenbus_driver blktap = {
++ .name = "tap",
++ .ids = blktap_ids,
++ .probe = blktap_probe,
++ .remove = blktap_remove,
++ .otherend_changed = tap_frontend_changed
++};
++
++
++void tap_blkif_xenbus_init(void)
++{
++ if (xenbus_register_backend(&blktap))
++ BUG();
++}
+diff --git a/drivers/xen/char/Makefile b/drivers/xen/char/Makefile
+new file mode 100644
+index 0000000..13604ad
+--- /dev/null
++++ b/drivers/xen/char/Makefile
+@@ -0,0 +1 @@
++obj-$(CONFIG_XEN_DEVMEM) := mem.o
+diff --git a/drivers/xen/char/mem.c b/drivers/xen/char/mem.c
+new file mode 100644
+index 0000000..24b5318
+--- /dev/null
++++ b/drivers/xen/char/mem.c
+@@ -0,0 +1,242 @@
++/*
++ * Originally from linux/drivers/char/mem.c
++ *
++ * Copyright (C) 1991, 1992 Linus Torvalds
++ *
++ * Added devfs support.
++ * Jan-11-1998, C. Scott Ananian <cananian at alumni.princeton.edu>
++ * Shared /dev/zero mmaping support, Feb 2000, Kanoj Sarcar <kanoj at sgi.com>
++ */
++
++#include <linux/mm.h>
++#include <linux/miscdevice.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/mman.h>
++#include <linux/random.h>
++#include <linux/init.h>
++#include <linux/raw.h>
++#include <linux/tty.h>
++#include <linux/capability.h>
++#include <linux/ptrace.h>
++#include <linux/device.h>
++#include <asm/pgalloc.h>
++#include <asm/uaccess.h>
++#include <asm/io.h>
++#include <asm/hypervisor.h>
++
++static inline int uncached_access(struct file *file)
++{
++ if (file->f_flags & O_SYNC)
++ return 1;
++ /* Xen sets correct MTRR type on non-RAM for us. */
++ return 0;
++}
++
++static inline int range_is_allowed(unsigned long pfn, unsigned long size)
++{
++#ifdef CONFIG_NONPROMISC_DEVMEM
++ u64 from = ((u64)pfn) << PAGE_SHIFT;
++ u64 to = from + size;
++ u64 cursor = from;
++
++ while (cursor < to) {
++ if (!devmem_is_allowed(pfn)) {
++ printk(KERN_INFO
++ "Program %s tried to access /dev/mem between %Lx->%Lx.\n",
++ current->comm, from, to);
++ return 0;
++ }
++ cursor += PAGE_SIZE;
++ pfn++;
++ }
++#endif
++ return 1;
++}
++
++/*
++ * This funcion reads the *physical* memory. The f_pos points directly to the
++ * memory location.
++ */
++static ssize_t read_mem(struct file * file, char __user * buf,
++ size_t count, loff_t *ppos)
++{
++ unsigned long p = *ppos, ignored;
++ ssize_t read = 0, sz;
++ void __iomem *v;
++
++ while (count > 0) {
++ /*
++ * Handle first page in case it's not aligned
++ */
++ if (-p & (PAGE_SIZE - 1))
++ sz = -p & (PAGE_SIZE - 1);
++ else
++ sz = PAGE_SIZE;
++
++ sz = min_t(unsigned long, sz, count);
++
++ if (!range_is_allowed(p >> PAGE_SHIFT, count))
++ return -EPERM;
++
++ v = ioremap(p, sz);
++ if (IS_ERR(v) || v == NULL) {
++ /*
++ * Some programs (e.g., dmidecode) groove off into
++ * weird RAM areas where no tables can possibly exist
++ * (because Xen will have stomped on them!). These
++ * programs get rather upset if we let them know that
++ * Xen failed their access, so we fake out a read of
++ * all zeroes.
++ */
++ if (clear_user(buf, count))
++ return -EFAULT;
++ read += count;
++ break;
++ }
++
++ ignored = copy_to_user(buf, v, sz);
++ iounmap(v);
++ if (ignored)
++ return -EFAULT;
++ buf += sz;
++ p += sz;
++ count -= sz;
++ read += sz;
++ }
++
++ *ppos += read;
++ return read;
++}
++
++static ssize_t write_mem(struct file * file, const char __user * buf,
++ size_t count, loff_t *ppos)
++{
++ unsigned long p = *ppos, ignored;
++ ssize_t written = 0, sz;
++ void __iomem *v;
++
++ while (count > 0) {
++ /*
++ * Handle first page in case it's not aligned
++ */
++ if (-p & (PAGE_SIZE - 1))
++ sz = -p & (PAGE_SIZE - 1);
++ else
++ sz = PAGE_SIZE;
++
++ sz = min_t(unsigned long, sz, count);
++
++ if (!range_is_allowed(p >> PAGE_SHIFT, sz))
++ return -EPERM;
++
++ v = ioremap(p, sz);
++ if (v == NULL)
++ break;
++ if (IS_ERR(v)) {
++ if (written == 0)
++ return PTR_ERR(v);
++ break;
++ }
++
++ ignored = copy_from_user(v, buf, sz);
++ iounmap(v);
++ if (ignored) {
++ written += sz - ignored;
++ if (written)
++ break;
++ return -EFAULT;
++ }
++ buf += sz;
++ p += sz;
++ count -= sz;
++ written += sz;
++ }
++
++ *ppos += written;
++ return written;
++}
++
++#ifndef ARCH_HAS_DEV_MEM_MMAP_MEM
++static void mmap_mem_open(struct vm_area_struct *vma)
++{
++ map_devmem(vma->vm_pgoff, vma->vm_end - vma->vm_start,
++ vma->vm_page_prot);
++}
++
++static void mmap_mem_close(struct vm_area_struct *vma)
++{
++ unmap_devmem(vma->vm_pgoff, vma->vm_end - vma->vm_start,
++ vma->vm_page_prot);
++}
++
++static struct vm_operations_struct mmap_mem_ops = {
++ .open = mmap_mem_open,
++ .close = mmap_mem_close
++};
++
++static int xen_mmap_mem(struct file * file, struct vm_area_struct * vma)
++{
++ size_t size = vma->vm_end - vma->vm_start;
++
++ if (uncached_access(file))
++ vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
++
++ if (!range_is_allowed(vma->vm_pgoff, size))
++ return -EPERM;
++
++ if (!phys_mem_access_prot_allowed(file, vma->vm_pgoff, size,
++ &vma->vm_page_prot))
++ return -EINVAL;
++
++ vma->vm_ops = &mmap_mem_ops;
++
++ /* We want to return the real error code, not EAGAIN. */
++ return direct_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff,
++ size, vma->vm_page_prot, DOMID_IO);
++}
++#endif
++
++/*
++ * The memory devices use the full 32/64 bits of the offset, and so we cannot
++ * check against negative addresses: they are ok. The return value is weird,
++ * though, in that case (0).
++ *
++ * also note that seeking relative to the "end of file" isn't supported:
++ * it has no meaning, so it returns -EINVAL.
++ */
++static loff_t memory_lseek(struct file * file, loff_t offset, int orig)
++{
++ loff_t ret;
++
++ mutex_lock(&file->f_path.dentry->d_inode->i_mutex);
++ switch (orig) {
++ case 0:
++ file->f_pos = offset;
++ ret = file->f_pos;
++ force_successful_syscall_return();
++ break;
++ case 1:
++ file->f_pos += offset;
++ ret = file->f_pos;
++ force_successful_syscall_return();
++ break;
++ default:
++ ret = -EINVAL;
++ }
++ mutex_unlock(&file->f_path.dentry->d_inode->i_mutex);
++ return ret;
++}
++
++static int open_mem(struct inode * inode, struct file * filp)
++{
++ return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
++}
++
++const struct file_operations mem_fops = {
++ .llseek = memory_lseek,
++ .read = read_mem,
++ .write = write_mem,
++ .mmap = xen_mmap_mem,
++ .open = open_mem,
++};
+diff --git a/drivers/xen/console/Makefile b/drivers/xen/console/Makefile
+new file mode 100644
+index 0000000..35de3e9
+--- /dev/null
++++ b/drivers/xen/console/Makefile
+@@ -0,0 +1,2 @@
++
++obj-y := console.o xencons_ring.o
+diff --git a/drivers/xen/console/console.c b/drivers/xen/console/console.c
+new file mode 100644
+index 0000000..e4e5d41
+--- /dev/null
++++ b/drivers/xen/console/console.c
+@@ -0,0 +1,727 @@
++/******************************************************************************
++ * console.c
++ *
++ * Virtual console driver.
++ *
++ * Copyright (c) 2002-2004, K A Fraser.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/errno.h>
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <linux/tty.h>
++#include <linux/tty_flip.h>
++#include <linux/serial.h>
++#include <linux/major.h>
++#include <linux/ptrace.h>
++#include <linux/ioport.h>
++#include <linux/mm.h>
++#include <linux/slab.h>
++#include <linux/init.h>
++#include <linux/console.h>
++#include <linux/bootmem.h>
++#include <linux/sysrq.h>
++#include <linux/screen_info.h>
++#include <linux/vt.h>
++#include <asm/io.h>
++#include <asm/irq.h>
++#include <asm/uaccess.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/event_channel.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/xenbus.h>
++#include <xen/xencons.h>
++
++/*
++ * Modes:
++ * 'xencons=off' [XC_OFF]: Console is disabled.
++ * 'xencons=tty' [XC_TTY]: Console attached to '/dev/tty[0-9]+'.
++ * 'xencons=ttyS' [XC_SERIAL]: Console attached to '/dev/ttyS[0-9]+'.
++ * 'xencons=xvc' [XC_XVC]: Console attached to '/dev/xvc0'.
++ * default: XC_XVC
++ *
++ * NB. In mode XC_TTY, we create dummy consoles for tty2-63. This suppresses
++ * warnings from standard distro startup scripts.
++ */
++static enum {
++ XC_OFF, XC_TTY, XC_SERIAL, XC_XVC
++} xc_mode = XC_XVC;
++static int xc_num = -1;
++
++/* /dev/xvc0 device number allocated by lanana.org. */
++#define XEN_XVC_MAJOR 204
++#define XEN_XVC_MINOR 191
++
++static int __init xencons_setup(char *str)
++{
++ char *q;
++ int n;
++ extern int console_use_vt;
++
++ console_use_vt = 1;
++ if (!strncmp(str, "ttyS", 4)) {
++ xc_mode = XC_SERIAL;
++ str += 4;
++ } else if (!strncmp(str, "tty", 3)) {
++ xc_mode = XC_TTY;
++ str += 3;
++ console_use_vt = 0;
++ } else if (!strncmp(str, "xvc", 3)) {
++ xc_mode = XC_XVC;
++ str += 3;
++ } else if (!strncmp(str, "off", 3)) {
++ xc_mode = XC_OFF;
++ str += 3;
++ }
++
++ n = simple_strtol(str, &q, 10);
++ if (q != str)
++ xc_num = n;
++
++ return 1;
++}
++__setup("xencons=", xencons_setup);
++
++/* The kernel and user-land drivers share a common transmit buffer. */
++static unsigned int wbuf_size = 4096;
++#define WBUF_MASK(_i) ((_i)&(wbuf_size-1))
++static char *wbuf;
++static unsigned int wc, wp; /* write_cons, write_prod */
++
++static int __init xencons_bufsz_setup(char *str)
++{
++ unsigned int goal;
++ goal = simple_strtoul(str, NULL, 0);
++ if (goal) {
++ goal = roundup_pow_of_two(goal);
++ if (wbuf_size < goal)
++ wbuf_size = goal;
++ }
++ return 1;
++}
++__setup("xencons_bufsz=", xencons_bufsz_setup);
++
++/* This lock protects accesses to the common transmit buffer. */
++static DEFINE_SPINLOCK(xencons_lock);
++
++/* Common transmit-kick routine. */
++static void __xencons_tx_flush(void);
++
++static struct tty_driver *xencons_driver;
++
++/******************** Kernel console driver ********************************/
++
++static void kcons_write(struct console *c, const char *s, unsigned int count)
++{
++ int i = 0;
++ unsigned long flags;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++
++ while (i < count) {
++ for (; i < count; i++) {
++ if ((wp - wc) >= (wbuf_size - 1))
++ break;
++ if ((wbuf[WBUF_MASK(wp++)] = s[i]) == '\n')
++ wbuf[WBUF_MASK(wp++)] = '\r';
++ }
++
++ __xencons_tx_flush();
++ }
++
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static void kcons_write_dom0(struct console *c, const char *s, unsigned int count)
++{
++
++ while (count > 0) {
++ int rc;
++ rc = HYPERVISOR_console_io( CONSOLEIO_write, count, (char *)s);
++ if (rc <= 0)
++ break;
++ count -= rc;
++ s += rc;
++ }
++}
++
++static struct tty_driver *kcons_device(struct console *c, int *index)
++{
++ *index = 0;
++ return xencons_driver;
++}
++
++static struct console kcons_info = {
++ .device = kcons_device,
++ .flags = CON_PRINTBUFFER | CON_ENABLED,
++ .index = -1,
++};
++
++static int __init xen_console_init(void)
++{
++ if (!is_running_on_xen())
++ goto out;
++
++ if (is_initial_xendomain()) {
++ kcons_info.write = kcons_write_dom0;
++ } else {
++ if (!xen_start_info->console.domU.evtchn)
++ goto out;
++ kcons_info.write = kcons_write;
++ }
++
++ switch (xc_mode) {
++ case XC_XVC:
++ strcpy(kcons_info.name, "xvc");
++ if (xc_num == -1)
++ xc_num = 0;
++ break;
++
++ case XC_SERIAL:
++ strcpy(kcons_info.name, "ttyS");
++ if (xc_num == -1)
++ xc_num = 0;
++ break;
++
++ case XC_TTY:
++ strcpy(kcons_info.name, "tty");
++ if (xc_num == -1)
++ xc_num = 1;
++ break;
++
++ default:
++ goto out;
++ }
++
++ wbuf = alloc_bootmem(wbuf_size);
++
++ register_console(&kcons_info);
++
++ out:
++ return 0;
++}
++console_initcall(xen_console_init);
++
++/*** Useful function for console debugging -- goes straight to Xen. ***/
++asmlinkage int xprintk(const char *fmt, ...)
++{
++ va_list args;
++ int printk_len;
++ static char printk_buf[1024];
++
++ /* Emit the output into the temporary buffer */
++ va_start(args, fmt);
++ printk_len = vsnprintf(printk_buf, sizeof(printk_buf), fmt, args);
++ va_end(args);
++
++ /* Send the processed output directly to Xen. */
++ kcons_write_dom0(NULL, printk_buf, printk_len);
++
++ return 0;
++}
++
++/*** Forcibly flush console data before dying. ***/
++void xencons_force_flush(void)
++{
++ int sz;
++
++ /* Emergency console is synchronous, so there's nothing to flush. */
++ if (!is_running_on_xen() ||
++ is_initial_xendomain() ||
++ !xen_start_info->console.domU.evtchn)
++ return;
++
++ /* Spin until console data is flushed through to the daemon. */
++ while (wc != wp) {
++ int sent = 0;
++ if ((sz = wp - wc) == 0)
++ continue;
++ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
++ if (sent > 0)
++ wc += sent;
++ }
++}
++
++
++void __init dom0_init_screen_info(const struct dom0_vga_console_info *info, size_t size)
++{
++ /* This is drawn from a dump from vgacon:startup in
++ * standard Linux. */
++ screen_info.orig_video_mode = 3;
++ screen_info.orig_video_isVGA = 1;
++ screen_info.orig_video_lines = 25;
++ screen_info.orig_video_cols = 80;
++ screen_info.orig_video_ega_bx = 3;
++ screen_info.orig_video_points = 16;
++ screen_info.orig_y = screen_info.orig_video_lines - 1;
++
++ switch (info->video_type) {
++ case XEN_VGATYPE_TEXT_MODE_3:
++ if (size < offsetof(struct dom0_vga_console_info, u.text_mode_3)
++ + sizeof(info->u.text_mode_3))
++ break;
++ screen_info.orig_video_lines = info->u.text_mode_3.rows;
++ screen_info.orig_video_cols = info->u.text_mode_3.columns;
++ screen_info.orig_x = info->u.text_mode_3.cursor_x;
++ screen_info.orig_y = info->u.text_mode_3.cursor_y;
++ screen_info.orig_video_points =
++ info->u.text_mode_3.font_height;
++ break;
++
++ case XEN_VGATYPE_VESA_LFB:
++ if (size < offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.gbl_caps))
++ break;
++ screen_info.orig_video_isVGA = VIDEO_TYPE_VLFB;
++ screen_info.lfb_width = info->u.vesa_lfb.width;
++ screen_info.lfb_height = info->u.vesa_lfb.height;
++ screen_info.lfb_depth = info->u.vesa_lfb.bits_per_pixel;
++ screen_info.lfb_base = info->u.vesa_lfb.lfb_base;
++ screen_info.lfb_size = info->u.vesa_lfb.lfb_size;
++ screen_info.lfb_linelength = info->u.vesa_lfb.bytes_per_line;
++ screen_info.red_size = info->u.vesa_lfb.red_size;
++ screen_info.red_pos = info->u.vesa_lfb.red_pos;
++ screen_info.green_size = info->u.vesa_lfb.green_size;
++ screen_info.green_pos = info->u.vesa_lfb.green_pos;
++ screen_info.blue_size = info->u.vesa_lfb.blue_size;
++ screen_info.blue_pos = info->u.vesa_lfb.blue_pos;
++ screen_info.rsvd_size = info->u.vesa_lfb.rsvd_size;
++ screen_info.rsvd_pos = info->u.vesa_lfb.rsvd_pos;
++ if (size >= offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.gbl_caps)
++ + sizeof(info->u.vesa_lfb.gbl_caps))
++ screen_info.capabilities = info->u.vesa_lfb.gbl_caps;
++ if (size >= offsetof(struct dom0_vga_console_info,
++ u.vesa_lfb.mode_attrs)
++ + sizeof(info->u.vesa_lfb.mode_attrs))
++ screen_info.vesa_attributes = info->u.vesa_lfb.mode_attrs;
++ break;
++ }
++}
++
++
++/******************** User-space console driver (/dev/console) ************/
++
++#define DRV(_d) (_d)
++#define DUMMY_TTY(_tty) ((xc_mode == XC_TTY) && \
++ ((_tty)->index != (xc_num - 1)))
++
++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)
++{
++ int i;
++ unsigned long flags;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ if (xencons_tty == NULL)
++ goto out;
++
++ for (i = 0; i < len; i++) {
++#ifdef CONFIG_MAGIC_SYSRQ
++ if (sysrq_on()) {
++ static unsigned long sysrq_requested;
++
++ if (buf[i] == '\x0f') { /* ^O */
++ if (!sysrq_requested) {
++ sysrq_requested = jiffies;
++ continue; /* don't print sysrq key */
++ }
++ sysrq_requested = 0;
++ } else if (sysrq_requested) {
++ unsigned long sysrq_timeout =
++ sysrq_requested + HZ*2;
++ sysrq_requested = 0;
++ if (time_before(jiffies, sysrq_timeout)) {
++ spin_unlock_irqrestore(
++ &xencons_lock, flags);
++ handle_sysrq(buf[i], xencons_tty);
++ spin_lock_irqsave(
++ &xencons_lock, flags);
++ continue;
++ }
++ }
++ }
++#endif
++ tty_insert_flip_char(xencons_tty, buf[i], 0);
++ }
++ tty_flip_buffer_push(xencons_tty);
++
++ out:
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static void __xencons_tx_flush(void)
++{
++ int sent, sz, work_done = 0;
++
++ if (x_char) {
++ if (is_initial_xendomain())
++ kcons_write_dom0(NULL, &x_char, 1);
++ else
++ while (x_char)
++ if (xencons_ring_send(&x_char, 1) == 1)
++ break;
++ x_char = 0;
++ work_done = 1;
++ }
++
++ while (wc != wp) {
++ sz = wp - wc;
++ if (sz > (wbuf_size - WBUF_MASK(wc)))
++ sz = wbuf_size - WBUF_MASK(wc);
++ if (is_initial_xendomain()) {
++ kcons_write_dom0(NULL, &wbuf[WBUF_MASK(wc)], sz);
++ wc += sz;
++ } else {
++ sent = xencons_ring_send(&wbuf[WBUF_MASK(wc)], sz);
++ if (sent == 0)
++ break;
++ wc += sent;
++ }
++ work_done = 1;
++ }
++
++ if (work_done && (xencons_tty != NULL)) {
++ wake_up_interruptible(&xencons_tty->write_wait);
++ if ((xencons_tty->flags & (1 << TTY_DO_WRITE_WAKEUP)) &&
++ (xencons_tty->ldisc.write_wakeup != NULL))
++ (xencons_tty->ldisc.write_wakeup)(xencons_tty);
++ }
++}
++
++void xencons_tx(void)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++/* Privileged receive callback and transmit kicker. */
++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);
++
++ xencons_tx();
++
++ return IRQ_HANDLED;
++}
++
++static int xencons_write_room(struct tty_struct *tty)
++{
++ return wbuf_size - (wp - wc);
++}
++
++static int xencons_chars_in_buffer(struct tty_struct *tty)
++{
++ return wp - wc;
++}
++
++static void xencons_send_xchar(struct tty_struct *tty, char ch)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ x_char = ch;
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static void xencons_throttle(struct tty_struct *tty)
++{
++ if (DUMMY_TTY(tty))
++ return;
++
++ if (I_IXOFF(tty))
++ xencons_send_xchar(tty, STOP_CHAR(tty));
++}
++
++static void xencons_unthrottle(struct tty_struct *tty)
++{
++ if (DUMMY_TTY(tty))
++ return;
++
++ if (I_IXOFF(tty)) {
++ if (x_char != 0)
++ x_char = 0;
++ else
++ xencons_send_xchar(tty, START_CHAR(tty));
++ }
++}
++
++static void xencons_flush_buffer(struct tty_struct *tty)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ wc = wp = 0;
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static inline int __xencons_put_char(int ch)
++{
++ char _ch = (char)ch;
++ if ((wp - wc) == wbuf_size)
++ return 0;
++ wbuf[WBUF_MASK(wp++)] = _ch;
++ return 1;
++}
++
++static int xencons_write(
++ struct tty_struct *tty,
++ const unsigned char *buf,
++ int count)
++{
++ int i;
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return count;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++
++ for (i = 0; i < count; i++)
++ if (!__xencons_put_char(buf[i]))
++ break;
++
++ if (i != 0)
++ __xencons_tx_flush();
++
++ spin_unlock_irqrestore(&xencons_lock, flags);
++
++ return i;
++}
++
++static int xencons_put_char(struct tty_struct *tty, u_char ch)
++{
++ unsigned long flags;
++ int ret;
++
++ if (DUMMY_TTY(tty))
++ return 0;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ ret = __xencons_put_char(ch);
++ spin_unlock_irqrestore(&xencons_lock, flags);
++ return ret;
++}
++
++static void xencons_flush_chars(struct tty_struct *tty)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static void xencons_wait_until_sent(struct tty_struct *tty, int timeout)
++{
++ unsigned long orig_jiffies = jiffies;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ while (tty_chars_in_buffer(tty)) {
++ set_current_state(TASK_INTERRUPTIBLE);
++ schedule_timeout(1);
++ if (signal_pending(current))
++ break;
++ if (timeout && time_after(jiffies, orig_jiffies + timeout))
++ break;
++ }
++
++ set_current_state(TASK_RUNNING);
++}
++
++static int xencons_open(struct tty_struct *tty, struct file *filp)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return 0;
++
++ spin_lock_irqsave(&xencons_lock, flags);
++ tty->driver_data = NULL;
++ if (xencons_tty == NULL)
++ xencons_tty = tty;
++ __xencons_tx_flush();
++ spin_unlock_irqrestore(&xencons_lock, flags);
++
++ return 0;
++}
++
++static void xencons_close(struct tty_struct *tty, struct file *filp)
++{
++ unsigned long flags;
++
++ if (DUMMY_TTY(tty))
++ return;
++
++ mutex_lock(&tty_mutex);
++
++ if (tty->count != 1) {
++ mutex_unlock(&tty_mutex);
++ return;
++ }
++
++ /* Prevent other threads from re-opening this tty. */
++ set_bit(TTY_CLOSING, &tty->flags);
++ mutex_unlock(&tty_mutex);
++
++ tty->closing = 1;
++ tty_wait_until_sent(tty, 0);
++ tty_driver_flush_buffer(tty);
++ if (tty->ldisc.flush_buffer != NULL)
++ tty->ldisc.flush_buffer(tty);
++ tty->closing = 0;
++ spin_lock_irqsave(&xencons_lock, flags);
++ xencons_tty = NULL;
++ spin_unlock_irqrestore(&xencons_lock, flags);
++}
++
++static const struct tty_operations xencons_ops = {
++ .open = xencons_open,
++ .close = xencons_close,
++ .write = xencons_write,
++ .write_room = xencons_write_room,
++ .put_char = xencons_put_char,
++ .flush_chars = xencons_flush_chars,
++ .chars_in_buffer = xencons_chars_in_buffer,
++ .send_xchar = xencons_send_xchar,
++ .flush_buffer = xencons_flush_buffer,
++ .throttle = xencons_throttle,
++ .unthrottle = xencons_unthrottle,
++ .wait_until_sent = xencons_wait_until_sent,
++};
++
++static int __init xencons_init(void)
++{
++ int rc;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ if (xc_mode == XC_OFF)
++ return 0;
++
++ if (!is_initial_xendomain()) {
++ rc = xencons_ring_init();
++ if (rc)
++ return rc;
++ }
++
++ xencons_driver = alloc_tty_driver((xc_mode == XC_TTY) ?
++ MAX_NR_CONSOLES : 1);
++ if (xencons_driver == NULL)
++ return -ENOMEM;
++
++ DRV(xencons_driver)->name = "xencons";
++ DRV(xencons_driver)->major = TTY_MAJOR;
++ DRV(xencons_driver)->type = TTY_DRIVER_TYPE_SERIAL;
++ DRV(xencons_driver)->subtype = SERIAL_TYPE_NORMAL;
++ DRV(xencons_driver)->init_termios = tty_std_termios;
++ DRV(xencons_driver)->flags =
++ TTY_DRIVER_REAL_RAW |
++ TTY_DRIVER_RESET_TERMIOS;
++ DRV(xencons_driver)->termios = xencons_termios;
++ DRV(xencons_driver)->termios_locked = xencons_termios_locked;
++
++ switch (xc_mode) {
++ case XC_XVC:
++ DRV(xencons_driver)->name = "xvc";
++ DRV(xencons_driver)->major = XEN_XVC_MAJOR;
++ DRV(xencons_driver)->minor_start = XEN_XVC_MINOR;
++ DRV(xencons_driver)->name_base = xc_num;
++ break;
++ case XC_SERIAL:
++ DRV(xencons_driver)->name = "ttyS";
++ DRV(xencons_driver)->minor_start = 64 + xc_num;
++ DRV(xencons_driver)->name_base = xc_num;
++ break;
++ default:
++ DRV(xencons_driver)->name = "tty";
++ DRV(xencons_driver)->minor_start = 1;
++ DRV(xencons_driver)->name_base = 1;
++ break;
++ }
++
++ tty_set_operations(xencons_driver, &xencons_ops);
++
++ if ((rc = tty_register_driver(DRV(xencons_driver))) != 0) {
++ printk("WARNING: Failed to register Xen virtual "
++ "console driver as '%s%d'\n",
++ DRV(xencons_driver)->name,
++ DRV(xencons_driver)->name_base);
++ put_tty_driver(xencons_driver);
++ xencons_driver = NULL;
++ return rc;
++ }
++
++ if (is_initial_xendomain()) {
++ xencons_priv_irq = bind_virq_to_irqhandler(
++ VIRQ_CONSOLE,
++ 0,
++ xencons_priv_interrupt,
++ 0,
++ "console",
++ NULL);
++ BUG_ON(xencons_priv_irq < 0);
++ }
++
++ printk("Xen virtual console successfully installed as %s%d\n",
++ DRV(xencons_driver)->name, xc_num);
++
++ return 0;
++}
++
++module_init(xencons_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/console/xencons_ring.c b/drivers/xen/console/xencons_ring.c
+new file mode 100644
+index 0000000..a775e2d
+--- /dev/null
++++ b/drivers/xen/console/xencons_ring.c
+@@ -0,0 +1,143 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/errno.h>
++#include <linux/signal.h>
++#include <linux/sched.h>
++#include <linux/interrupt.h>
++#include <linux/tty.h>
++#include <linux/tty_flip.h>
++#include <linux/serial.h>
++#include <linux/major.h>
++#include <linux/ptrace.h>
++#include <linux/ioport.h>
++#include <linux/mm.h>
++#include <linux/slab.h>
++
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/xencons.h>
++#include <linux/wait.h>
++#include <linux/interrupt.h>
++#include <linux/sched.h>
++#include <linux/err.h>
++#include <xen/interface/io/console.h>
++
++static int xencons_irq;
++
++static inline struct xencons_interface *xencons_interface(void)
++{
++ return mfn_to_virt(xen_start_info->console.domU.mfn);
++}
++
++static inline void notify_daemon(void)
++{
++ /* Use evtchn: this is called early, before irq is set up. */
++ notify_remote_via_evtchn(xen_start_info->console.domU.evtchn);
++}
++
++int xencons_ring_send(const char *data, unsigned len)
++{
++ int sent = 0;
++ struct xencons_interface *intf = xencons_interface();
++ XENCONS_RING_IDX cons, prod;
++
++ cons = intf->out_cons;
++ prod = intf->out_prod;
++ mb();
++ BUG_ON((prod - cons) > sizeof(intf->out));
++
++ while ((sent < len) && ((prod - cons) < sizeof(intf->out)))
++ intf->out[MASK_XENCONS_IDX(prod++, intf->out)] = data[sent++];
++
++ wmb();
++ intf->out_prod = prod;
++
++ notify_daemon();
++
++ return sent;
++}
++
++static irqreturn_t handle_input(int irq, void *unused)
++{
++ struct xencons_interface *intf = xencons_interface();
++ XENCONS_RING_IDX cons, prod;
++
++ cons = intf->in_cons;
++ prod = intf->in_prod;
++ mb();
++ BUG_ON((prod - cons) > sizeof(intf->in));
++
++ while (cons != prod) {
++ xencons_rx(intf->in+MASK_XENCONS_IDX(cons,intf->in), 1);
++ cons++;
++ }
++
++ mb();
++ intf->in_cons = cons;
++
++ notify_daemon();
++
++ xencons_tx();
++
++ return IRQ_HANDLED;
++}
++
++int xencons_ring_init(void)
++{
++ int irq;
++
++ if (xencons_irq)
++ unbind_from_irqhandler(xencons_irq, NULL);
++ xencons_irq = 0;
++
++ if (!is_running_on_xen() ||
++ is_initial_xendomain() ||
++ !xen_start_info->console.domU.evtchn)
++ return -ENODEV;
++
++ irq = bind_caller_port_to_irqhandler(
++ xen_start_info->console.domU.evtchn,
++ handle_input, 0, "xencons", NULL);
++ if (irq < 0) {
++ printk(KERN_ERR "XEN console request irq failed %i\n", irq);
++ return irq;
++ }
++
++ xencons_irq = irq;
++
++ /* In case we have in-flight data after save/restore... */
++ notify_daemon();
++
++ return 0;
++}
++
++void xencons_resume(void)
++{
++ (void)xencons_ring_init();
++}
+diff --git a/drivers/xen/core/Makefile b/drivers/xen/core/Makefile
+new file mode 100644
+index 0000000..546f0b2
+--- /dev/null
++++ b/drivers/xen/core/Makefile
+@@ -0,0 +1,15 @@
++#
++# Makefile for the linux kernel.
++#
++
++obj-y := evtchn.o gnttab.o features.o reboot.o machine_reboot.o firmware.o
++
++obj-$(CONFIG_PCI) += pci.o
++obj-$(CONFIG_PROC_FS) += xen_proc.o
++obj-$(CONFIG_SYS_HYPERVISOR) += hypervisor_sysfs.o
++obj-$(CONFIG_HOTPLUG_CPU) += cpu_hotplug.o
++obj-$(CONFIG_XEN_SYSFS) += xen_sysfs.o
++obj-$(CONFIG_XEN_SMPBOOT) += smpboot.o
++obj-$(CONFIG_X86_SMP) += spinlock.o
++obj-$(CONFIG_KEXEC) += machine_kexec.o
++obj-$(CONFIG_XEN_XENCOMM) += xencomm.o
+diff --git a/drivers/xen/core/cpu_hotplug.c b/drivers/xen/core/cpu_hotplug.c
+new file mode 100644
+index 0000000..f9f49d8
+--- /dev/null
++++ b/drivers/xen/core/cpu_hotplug.c
+@@ -0,0 +1,173 @@
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/notifier.h>
++#include <linux/cpu.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/xenbus.h>
++
++/*
++ * Set of CPUs that remote admin software will allow us to bring online.
++ * Notified to us via xenbus.
++ */
++static cpumask_t xenbus_allowed_cpumask;
++
++/* Set of CPUs that local admin will allow us to bring online. */
++static cpumask_t local_allowed_cpumask = CPU_MASK_ALL;
++
++static int local_cpu_hotplug_request(void)
++{
++ /*
++ * We assume a CPU hotplug request comes from local admin if it is made
++ * via a userspace process (i.e., one with a real mm_struct).
++ */
++ return (current->mm != NULL);
++}
++
++static void vcpu_hotplug(unsigned int cpu)
++{
++ int err;
++ char dir[32], state[32];
++
++ if ((cpu >= NR_CPUS) || !cpu_possible(cpu))
++ return;
++
++ sprintf(dir, "cpu/%u", cpu);
++ err = xenbus_scanf(XBT_NIL, dir, "availability", "%s", state);
++ if (err != 1) {
++ printk(KERN_ERR "XENBUS: Unable to read cpu state\n");
++ return;
++ }
++
++ if (strcmp(state, "online") == 0) {
++ cpu_set(cpu, xenbus_allowed_cpumask);
++ (void)cpu_up(cpu);
++ } else if (strcmp(state, "offline") == 0) {
++ cpu_clear(cpu, xenbus_allowed_cpumask);
++ (void)cpu_down(cpu);
++ } else {
++ printk(KERN_ERR "XENBUS: unknown state(%s) on CPU%d\n",
++ state, cpu);
++ }
++}
++
++static void handle_vcpu_hotplug_event(
++ struct xenbus_watch *watch, const char **vec, unsigned int len)
++{
++ unsigned int cpu;
++ char *cpustr;
++ const char *node = vec[XS_WATCH_PATH];
++
++ if ((cpustr = strstr(node, "cpu/")) != NULL) {
++ sscanf(cpustr, "cpu/%u", &cpu);
++ vcpu_hotplug(cpu);
++ }
++}
++
++static int smpboot_cpu_notify(struct notifier_block *notifier,
++ unsigned long action, void *hcpu)
++{
++ unsigned int cpu = (long)hcpu;
++
++ /*
++ * We do this in a callback notifier rather than __cpu_disable()
++ * because local_cpu_hotplug_request() does not work in the latter
++ * as it's always executed from within a stopmachine kthread.
++ */
++ if ((action == CPU_DOWN_PREPARE) && local_cpu_hotplug_request())
++ cpu_clear(cpu, local_allowed_cpumask);
++
++ return NOTIFY_OK;
++}
++
++static int setup_cpu_watcher(struct notifier_block *notifier,
++ unsigned long event, void *data)
++{
++ unsigned int i;
++
++ static struct xenbus_watch cpu_watch = {
++ .node = "cpu",
++ .callback = handle_vcpu_hotplug_event,
++ .flags = XBWF_new_thread };
++ (void)register_xenbus_watch(&cpu_watch);
++
++ if (!is_initial_xendomain()) {
++ for_each_possible_cpu(i)
++ vcpu_hotplug(i);
++ printk(KERN_INFO "Brought up %ld CPUs\n",
++ (long)num_online_cpus());
++ }
++
++ return NOTIFY_DONE;
++}
++
++static int __init setup_vcpu_hotplug_event(void)
++{
++ static struct notifier_block hotplug_cpu = {
++ .notifier_call = smpboot_cpu_notify };
++ static struct notifier_block xsn_cpu = {
++ .notifier_call = setup_cpu_watcher };
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ register_cpu_notifier(&hotplug_cpu);
++ register_xenstore_notifier(&xsn_cpu);
++
++ return 0;
++}
++
++arch_initcall(setup_vcpu_hotplug_event);
++
++int smp_suspend(void)
++{
++ unsigned int cpu;
++ int err;
++
++ for_each_online_cpu(cpu) {
++ if (cpu == 0)
++ continue;
++ err = cpu_down(cpu);
++ if (err) {
++ printk(KERN_CRIT "Failed to take all CPUs "
++ "down: %d.\n", err);
++ for_each_possible_cpu(cpu)
++ vcpu_hotplug(cpu);
++ return err;
++ }
++ }
++
++ return 0;
++}
++
++void smp_resume(void)
++{
++ unsigned int cpu;
++
++ for_each_possible_cpu(cpu)
++ vcpu_hotplug(cpu);
++}
++
++int cpu_up_check(unsigned int cpu)
++{
++ int rc = 0;
++
++ if (local_cpu_hotplug_request()) {
++ cpu_set(cpu, local_allowed_cpumask);
++ if (!cpu_isset(cpu, xenbus_allowed_cpumask)) {
++ printk("%s: attempt to bring up CPU %u disallowed by "
++ "remote admin.\n", __FUNCTION__, cpu);
++ rc = -EBUSY;
++ }
++ } else if (!cpu_isset(cpu, local_allowed_cpumask) ||
++ !cpu_isset(cpu, xenbus_allowed_cpumask)) {
++ rc = -EBUSY;
++ }
++
++ return rc;
++}
++
++void init_xenbus_allowed_cpumask(void)
++{
++ xenbus_allowed_cpumask = cpu_present_map;
++}
+diff --git a/drivers/xen/core/evtchn.c b/drivers/xen/core/evtchn.c
+new file mode 100644
+index 0000000..544631c
+--- /dev/null
++++ b/drivers/xen/core/evtchn.c
+@@ -0,0 +1,1093 @@
++/******************************************************************************
++ * evtchn.c
++ *
++ * Communication via Xen event channels.
++ *
++ * Copyright (c) 2002-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/module.h>
++#include <linux/irq.h>
++#include <linux/interrupt.h>
++#include <linux/sched.h>
++#include <linux/kernel_stat.h>
++#include <linux/version.h>
++#include <asm/atomic.h>
++#include <asm/system.h>
++#include <asm/ptrace.h>
++#include <asm/synch_bitops.h>
++#include <xen/evtchn.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <asm/hypervisor.h>
++#include <linux/mc146818rtc.h> /* RTC_IRQ */
++
++/*
++ * This lock protects updates to the following mapping and reference-count
++ * arrays. The lock does not need to be acquired to read the mapping tables.
++ */
++static DEFINE_SPINLOCK(irq_mapping_update_lock);
++
++/* IRQ <-> event-channel mappings. */
++static int evtchn_to_irq[NR_EVENT_CHANNELS] = {
++ [0 ... NR_EVENT_CHANNELS-1] = -1 };
++
++/* Packed IRQ information: binding type, sub-type index, and event channel. */
++static u32 irq_info[NR_IRQS];
++
++/* Binding types. */
++enum {
++ IRQT_UNBOUND,
++ IRQT_PIRQ,
++ IRQT_VIRQ,
++ IRQT_IPI,
++ IRQT_LOCAL_PORT,
++ IRQT_CALLER_PORT
++};
++
++/* Constructor for packed IRQ information. */
++static inline u32 mk_irq_info(u32 type, u32 index, u32 evtchn)
++{
++ return ((type << 24) | (index << 16) | evtchn);
++}
++
++/* Convenient shorthand for packed representation of an unbound IRQ. */
++#define IRQ_UNBOUND mk_irq_info(IRQT_UNBOUND, 0, 0)
++
++/*
++ * Accessors for packed IRQ information.
++ */
++
++static inline unsigned int evtchn_from_irq(int irq)
++{
++ return (u16)(irq_info[irq]);
++}
++
++static inline unsigned int index_from_irq(int irq)
++{
++ return (u8)(irq_info[irq] >> 16);
++}
++
++static inline unsigned int type_from_irq(int irq)
++{
++ return (u8)(irq_info[irq] >> 24);
++}
++
++/* IRQ <-> VIRQ mapping. */
++DEFINE_PER_CPU(int, virq_to_irq[NR_VIRQS]) = {[0 ... NR_VIRQS-1] = -1};
++
++/* IRQ <-> IPI mapping. */
++#ifndef NR_IPIS
++#define NR_IPIS 1
++#endif
++DEFINE_PER_CPU(int, ipi_to_irq[NR_IPIS]) = {[0 ... NR_IPIS-1] = -1};
++
++/* Reference counts for bindings to IRQs. */
++static int irq_bindcount[NR_IRQS];
++
++/* Bitmap indicating which PIRQs require Xen to be notified on unmask. */
++static DECLARE_BITMAP(pirq_needs_eoi, NR_PIRQS);
++
++#ifdef CONFIG_SMP
++
++static u8 cpu_evtchn[NR_EVENT_CHANNELS];
++static unsigned long cpu_evtchn_mask[NR_CPUS][NR_EVENT_CHANNELS/BITS_PER_LONG];
++
++static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
++ unsigned int idx)
++{
++ return (sh->evtchn_pending[idx] &
++ cpu_evtchn_mask[cpu][idx] &
++ ~sh->evtchn_mask[idx]);
++}
++
++static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ int irq = evtchn_to_irq[chn];
++
++ BUG_ON(!test_bit(chn, s->evtchn_mask));
++
++ if (irq != -1)
++ irq_desc[irq].affinity = cpumask_of_cpu(cpu);
++
++ clear_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu_evtchn[chn]]);
++ set_bit(chn, (unsigned long *)cpu_evtchn_mask[cpu]);
++ cpu_evtchn[chn] = cpu;
++}
++
++static void init_evtchn_cpu_bindings(void)
++{
++ int i;
++
++ /* By default all event channels notify CPU#0. */
++ for (i = 0; i < NR_IRQS; i++)
++ irq_desc[i].affinity = cpumask_of_cpu(0);
++
++ memset(cpu_evtchn, 0, sizeof(cpu_evtchn));
++ memset(cpu_evtchn_mask[0], ~0, sizeof(cpu_evtchn_mask[0]));
++}
++
++static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
++{
++ return cpu_evtchn[evtchn];
++}
++
++#else
++
++static inline unsigned long active_evtchns(unsigned int cpu, shared_info_t *sh,
++ unsigned int idx)
++{
++ return (sh->evtchn_pending[idx] & ~sh->evtchn_mask[idx]);
++}
++
++static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
++{
++}
++
++static void init_evtchn_cpu_bindings(void)
++{
++}
++
++static inline unsigned int cpu_from_evtchn(unsigned int evtchn)
++{
++ return 0;
++}
++
++#endif
++
++/* Upcall to generic IRQ layer. */
++#ifdef CONFIG_X86
++extern unsigned int do_IRQ(struct pt_regs *regs);
++void __init xen_init_IRQ(void);
++void __init init_IRQ(void)
++{
++ irq_ctx_init(0);
++ xen_init_IRQ();
++}
++#if defined (__i386__)
++static inline void exit_idle(void) {}
++#elif defined (__x86_64__)
++#include <asm/idle.h>
++#endif
++#define do_IRQ(irq, regs) do { \
++ (regs)->orig_ax = ~(irq); \
++ do_IRQ((regs)); \
++} while (0)
++#endif
++
++/* Xen will never allocate port zero for any purpose. */
++#define VALID_EVTCHN(chn) ((chn) != 0)
++
++/*
++ * Force a proper event-channel callback from Xen after clearing the
++ * callback mask. We do this in a very simple manner, by making a call
++ * down into Xen. The pending flag will be checked by Xen on return.
++ */
++void force_evtchn_callback(void)
++{
++ VOID(HYPERVISOR_xen_version(0, NULL));
++}
++/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
++EXPORT_SYMBOL(force_evtchn_callback);
++
++static DEFINE_PER_CPU(unsigned int, upcall_count) = { 0 };
++static DEFINE_PER_CPU(unsigned int, last_processed_l1i) = { BITS_PER_LONG - 1 };
++static DEFINE_PER_CPU(unsigned int, last_processed_l2i) = { BITS_PER_LONG - 1 };
++
++/* NB. Interrupts are disabled on entry. */
++asmlinkage void evtchn_do_upcall(struct pt_regs *regs)
++{
++ unsigned long l1, l2;
++ unsigned long masked_l1, masked_l2;
++ unsigned int l1i, l2i, port, count;
++ int irq;
++ unsigned int cpu = smp_processor_id();
++ shared_info_t *s = HYPERVISOR_shared_info;
++ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
++
++
++ do {
++ /* Avoid a callback storm when we reenable delivery. */
++ vcpu_info->evtchn_upcall_pending = 0;
++
++ /* Nested invocations bail immediately. */
++ if (unlikely(per_cpu(upcall_count, cpu)++))
++ return;
++
++#ifndef CONFIG_X86 /* No need for a barrier -- XCHG is a barrier on x86. */
++ /* Clear master flag /before/ clearing selector flag. */
++ wmb();
++#endif
++ l1 = xchg(&vcpu_info->evtchn_pending_sel, 0);
++
++ l1i = per_cpu(last_processed_l1i, cpu);
++ l2i = per_cpu(last_processed_l2i, cpu);
++
++ while (l1 != 0) {
++
++ l1i = (l1i + 1) % BITS_PER_LONG;
++ masked_l1 = l1 & ((~0UL) << l1i);
++
++ if (masked_l1 == 0) { /* if we masked out all events, wrap around to the beginning */
++ l1i = BITS_PER_LONG - 1;
++ l2i = BITS_PER_LONG - 1;
++ continue;
++ }
++ l1i = __ffs(masked_l1);
++
++ do {
++ l2 = active_evtchns(cpu, s, l1i);
++
++ l2i = (l2i + 1) % BITS_PER_LONG;
++ masked_l2 = l2 & ((~0UL) << l2i);
++
++ if (masked_l2 == 0) { /* if we masked out all events, move on */
++ l2i = BITS_PER_LONG - 1;
++ break;
++ }
++
++ l2i = __ffs(masked_l2);
++
++ /* process port */
++ port = (l1i * BITS_PER_LONG) + l2i;
++ if ((irq = evtchn_to_irq[port]) != -1)
++ do_IRQ(irq, regs);
++ else {
++ exit_idle();
++ evtchn_device_upcall(port);
++ }
++
++ /* if this is the final port processed, we'll pick up here+1 next time */
++ per_cpu(last_processed_l1i, cpu) = l1i;
++ per_cpu(last_processed_l2i, cpu) = l2i;
++
++ } while (l2i != BITS_PER_LONG - 1);
++
++ l2 = active_evtchns(cpu, s, l1i);
++ if (l2 == 0) /* we handled all ports, so we can clear the selector bit */
++ l1 &= ~(1UL << l1i);
++
++ }
++
++ /* If there were nested callbacks then we have more to do. */
++ count = per_cpu(upcall_count, cpu);
++ per_cpu(upcall_count, cpu) = 0;
++ } while (unlikely(count != 1));
++}
++
++static int find_unbound_irq(void)
++{
++ static int warned;
++ int dynirq, irq;
++
++ for (dynirq = 0; dynirq < NR_DYNIRQS; dynirq++) {
++ irq = dynirq_to_irq(dynirq);
++ if (irq_bindcount[irq] == 0)
++ return irq;
++ }
++
++ if (!warned) {
++ warned = 1;
++ printk(KERN_WARNING "No available IRQ to bind to: "
++ "increase NR_DYNIRQS.\n");
++ }
++
++ return -ENOSPC;
++}
++
++static int bind_caller_port_to_irq(unsigned int caller_port)
++{
++ int irq;
++
++ spin_lock(&irq_mapping_update_lock);
++
++ if ((irq = evtchn_to_irq[caller_port]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
++
++ evtchn_to_irq[caller_port] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_CALLER_PORT, 0, caller_port);
++ }
++
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
++}
++
++static int bind_local_port_to_irq(unsigned int local_port)
++{
++ int irq;
++
++ spin_lock(&irq_mapping_update_lock);
++
++ BUG_ON(evtchn_to_irq[local_port] != -1);
++
++ if ((irq = find_unbound_irq()) < 0) {
++ struct evtchn_close close = { .port = local_port };
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
++ BUG();
++ goto out;
++ }
++
++ evtchn_to_irq[local_port] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_LOCAL_PORT, 0, local_port);
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
++}
++
++static int bind_listening_port_to_irq(unsigned int remote_domain)
++{
++ struct evtchn_alloc_unbound alloc_unbound;
++ int err;
++
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = remote_domain;
++
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++
++ return err ? : bind_local_port_to_irq(alloc_unbound.port);
++}
++
++static int bind_interdomain_evtchn_to_irq(unsigned int remote_domain,
++ unsigned int remote_port)
++{
++ struct evtchn_bind_interdomain bind_interdomain;
++ int err;
++
++ bind_interdomain.remote_dom = remote_domain;
++ bind_interdomain.remote_port = remote_port;
++
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
++ &bind_interdomain);
++
++ return err ? : bind_local_port_to_irq(bind_interdomain.local_port);
++}
++
++static int bind_virq_to_irq(unsigned int virq, unsigned int cpu)
++{
++ struct evtchn_bind_virq bind_virq;
++ int evtchn, irq;
++
++ spin_lock(&irq_mapping_update_lock);
++
++ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
++
++ bind_virq.virq = virq;
++ bind_virq.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq) != 0)
++ BUG();
++ evtchn = bind_virq.port;
++
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
++
++ per_cpu(virq_to_irq, cpu)[virq] = irq;
++
++ bind_evtchn_to_cpu(evtchn, cpu);
++ }
++
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
++}
++
++static int bind_ipi_to_irq(unsigned int ipi, unsigned int cpu)
++{
++ struct evtchn_bind_ipi bind_ipi;
++ int evtchn, irq;
++
++ spin_lock(&irq_mapping_update_lock);
++
++ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1) {
++ if ((irq = find_unbound_irq()) < 0)
++ goto out;
++
++ bind_ipi.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
++ &bind_ipi) != 0)
++ BUG();
++ evtchn = bind_ipi.port;
++
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
++
++ per_cpu(ipi_to_irq, cpu)[ipi] = irq;
++
++ bind_evtchn_to_cpu(evtchn, cpu);
++ }
++
++ irq_bindcount[irq]++;
++
++ out:
++ spin_unlock(&irq_mapping_update_lock);
++ return irq;
++}
++
++static void unbind_from_irq(unsigned int irq)
++{
++ struct evtchn_close close;
++ unsigned int cpu;
++ int evtchn = evtchn_from_irq(irq);
++
++ spin_lock(&irq_mapping_update_lock);
++
++ if ((--irq_bindcount[irq] == 0) && VALID_EVTCHN(evtchn)) {
++ close.port = evtchn;
++ if ((type_from_irq(irq) != IRQT_CALLER_PORT) &&
++ HYPERVISOR_event_channel_op(EVTCHNOP_close, &close))
++ BUG();
++
++ switch (type_from_irq(irq)) {
++ case IRQT_VIRQ:
++ per_cpu(virq_to_irq, cpu_from_evtchn(evtchn))
++ [index_from_irq(irq)] = -1;
++ break;
++ case IRQT_IPI:
++ per_cpu(ipi_to_irq, cpu_from_evtchn(evtchn))
++ [index_from_irq(irq)] = -1;
++ break;
++ default:
++ break;
++ }
++
++ /* Closed ports are implicitly re-bound to VCPU0. */
++ bind_evtchn_to_cpu(evtchn, 0);
++
++ evtchn_to_irq[evtchn] = -1;
++ irq_info[irq] = IRQ_UNBOUND;
++
++ /* Zap stats across IRQ changes of use. */
++ for_each_possible_cpu(cpu)
++ kstat_cpu(cpu).irqs[irq] = 0;
++ }
++
++ spin_unlock(&irq_mapping_update_lock);
++}
++
++int bind_caller_port_to_irqhandler(
++ unsigned int caller_port,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_caller_port_to_irq(caller_port);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_caller_port_to_irqhandler);
++
++int bind_listening_port_to_irqhandler(
++ unsigned int remote_domain,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_listening_port_to_irq(remote_domain);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_listening_port_to_irqhandler);
++
++int bind_interdomain_evtchn_to_irqhandler(
++ unsigned int remote_domain,
++ unsigned int remote_port,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_interdomain_evtchn_to_irq(remote_domain, remote_port);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_interdomain_evtchn_to_irqhandler);
++
++int bind_virq_to_irqhandler(
++ unsigned int virq,
++ unsigned int cpu,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_virq_to_irq(virq, cpu);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_virq_to_irqhandler);
++
++int bind_ipi_to_irqhandler(
++ unsigned int ipi,
++ unsigned int cpu,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id)
++{
++ int irq, retval;
++
++ irq = bind_ipi_to_irq(ipi, cpu);
++ if (irq < 0)
++ return irq;
++
++ retval = request_irq(irq, handler, irqflags, devname, dev_id);
++ if (retval != 0) {
++ unbind_from_irq(irq);
++ return retval;
++ }
++
++ return irq;
++}
++EXPORT_SYMBOL_GPL(bind_ipi_to_irqhandler);
++
++void unbind_from_irqhandler(unsigned int irq, void *dev_id)
++{
++ free_irq(irq, dev_id);
++ unbind_from_irq(irq);
++}
++EXPORT_SYMBOL_GPL(unbind_from_irqhandler);
++
++#ifdef CONFIG_SMP
++void rebind_evtchn_to_cpu(int port, unsigned int cpu)
++{
++ struct evtchn_bind_vcpu ebv = { .port = port, .vcpu = cpu };
++ int masked;
++
++ masked = test_and_set_evtchn_mask(port);
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &ebv) == 0)
++ bind_evtchn_to_cpu(port, cpu);
++ if (!masked)
++ unmask_evtchn(port);
++}
++
++static void rebind_irq_to_cpu(unsigned int irq, unsigned int tcpu)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ rebind_evtchn_to_cpu(evtchn, tcpu);
++}
++
++static void set_affinity_irq(unsigned int irq, cpumask_t dest)
++{
++ unsigned tcpu = first_cpu(dest);
++ rebind_irq_to_cpu(irq, tcpu);
++}
++#endif
++
++int resend_irq_on_evtchn(unsigned int irq)
++{
++ int masked, evtchn = evtchn_from_irq(irq);
++ shared_info_t *s = HYPERVISOR_shared_info;
++
++ if (!VALID_EVTCHN(evtchn))
++ return 1;
++
++ masked = test_and_set_evtchn_mask(evtchn);
++ synch_set_bit(evtchn, s->evtchn_pending);
++ if (!masked)
++ unmask_evtchn(evtchn);
++
++ return 1;
++}
++
++/*
++ * Interface to generic handling in irq.c
++ */
++
++static unsigned int startup_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ unmask_evtchn(evtchn);
++ return 0;
++}
++
++static void unmask_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ unmask_evtchn(evtchn);
++}
++
++static void mask_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ mask_evtchn(evtchn);
++}
++
++static void ack_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ move_native_irq(irq);
++
++ if (VALID_EVTCHN(evtchn)) {
++ mask_evtchn(evtchn);
++ clear_evtchn(evtchn);
++ }
++}
++
++static void end_dynirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn) && !(irq_desc[irq].status & IRQ_DISABLED))
++ unmask_evtchn(evtchn);
++}
++
++static struct irq_chip dynirq_chip = {
++ .name = "Dynamic-irq",
++ .startup = startup_dynirq,
++ .mask = mask_dynirq,
++ .unmask = unmask_dynirq,
++ .mask_ack = ack_dynirq,
++ .ack = ack_dynirq,
++ .end = end_dynirq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_affinity_irq,
++#endif
++ .retrigger = resend_irq_on_evtchn,
++};
++
++static inline void pirq_unmask_notify(int pirq)
++{
++ struct physdev_eoi eoi = { .irq = pirq };
++ if (unlikely(test_bit(pirq, pirq_needs_eoi)))
++ VOID(HYPERVISOR_physdev_op(PHYSDEVOP_eoi, &eoi));
++}
++
++static inline void pirq_query_unmask(int pirq)
++{
++ struct physdev_irq_status_query irq_status;
++ irq_status.irq = pirq;
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status))
++ irq_status.flags = 0;
++ clear_bit(pirq, pirq_needs_eoi);
++ if (irq_status.flags & XENIRQSTAT_needs_eoi)
++ set_bit(pirq, pirq_needs_eoi);
++}
++
++/*
++ * On startup, if there is no action associated with the IRQ then we are
++ * probing. In this case we should not share with others as it will confuse us.
++ */
++#define probing_irq(_irq) (irq_desc[(_irq)].action == NULL)
++
++static unsigned int startup_pirq(unsigned int irq)
++{
++ struct evtchn_bind_pirq bind_pirq;
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ goto out;
++
++ bind_pirq.pirq = irq;
++ /* NB. We are happy to share unless we are probing. */
++ bind_pirq.flags = probing_irq(irq) ? 0 : BIND_PIRQ__WILL_SHARE;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_pirq, &bind_pirq) != 0) {
++ if (!probing_irq(irq))
++ printk(KERN_INFO "Failed to obtain physical IRQ %d\n",
++ irq);
++ return 0;
++ }
++ evtchn = bind_pirq.port;
++
++ pirq_query_unmask(irq_to_pirq(irq));
++
++ evtchn_to_irq[evtchn] = irq;
++ bind_evtchn_to_cpu(evtchn, 0);
++ irq_info[irq] = mk_irq_info(IRQT_PIRQ, irq, evtchn);
++
++ out:
++ unmask_evtchn(evtchn);
++ pirq_unmask_notify(irq_to_pirq(irq));
++
++ 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 unmask_pirq(unsigned int irq)
++{
++ startup_pirq(irq);
++}
++
++static void mask_pirq(unsigned int irq)
++{
++}
++
++static void ack_pirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ move_native_irq(irq);
++
++ if (VALID_EVTCHN(evtchn)) {
++ mask_evtchn(evtchn);
++ clear_evtchn(evtchn);
++ }
++}
++
++static void end_pirq(unsigned int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if ((irq_desc[irq].status & (IRQ_DISABLED|IRQ_PENDING)) ==
++ (IRQ_DISABLED|IRQ_PENDING)) {
++ shutdown_pirq(irq);
++ } else if (VALID_EVTCHN(evtchn)) {
++ unmask_evtchn(evtchn);
++ pirq_unmask_notify(irq_to_pirq(irq));
++ }
++}
++
++static struct irq_chip pirq_chip = {
++ .name = "Phys-irq",
++ .typename = "Phys-irq",
++ .startup = startup_pirq,
++ .shutdown = shutdown_pirq,
++ .mask = mask_pirq,
++ .unmask = unmask_pirq,
++ .mask_ack = ack_pirq,
++ .ack = ack_pirq,
++ .end = end_pirq,
++#ifdef CONFIG_SMP
++ .set_affinity = set_affinity_irq,
++#endif
++ .retrigger = resend_irq_on_evtchn,
++};
++
++int irq_ignore_unhandled(unsigned int irq)
++{
++ struct physdev_irq_status_query irq_status = { .irq = irq };
++
++ if (!is_running_on_xen())
++ return 0;
++
++ if (HYPERVISOR_physdev_op(PHYSDEVOP_irq_status_query, &irq_status))
++ return 0;
++ return !!(irq_status.flags & XENIRQSTAT_shared);
++}
++
++void notify_remote_via_irq(int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ notify_remote_via_evtchn(evtchn);
++}
++EXPORT_SYMBOL_GPL(notify_remote_via_irq);
++
++int irq_to_evtchn_port(int irq)
++{
++ return evtchn_from_irq(irq);
++}
++EXPORT_SYMBOL_GPL(irq_to_evtchn_port);
++
++void mask_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ synch_set_bit(port, s->evtchn_mask);
++}
++EXPORT_SYMBOL_GPL(mask_evtchn);
++
++void unmask_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned int cpu = smp_processor_id();
++ vcpu_info_t *vcpu_info = &s->vcpu_info[cpu];
++
++ BUG_ON(!irqs_disabled());
++
++ /* Slow path (hypercall) if this is a non-local port. */
++ if (unlikely(cpu != cpu_from_evtchn(port))) {
++ struct evtchn_unmask unmask = { .port = port };
++ VOID(HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask));
++ return;
++ }
++
++ synch_clear_bit(port, s->evtchn_mask);
++
++ /* Did we miss an interrupt 'edge'? Re-fire if so. */
++ if (synch_test_bit(port, s->evtchn_pending) &&
++ !synch_test_and_set_bit(port / BITS_PER_LONG,
++ &vcpu_info->evtchn_pending_sel))
++ vcpu_info->evtchn_upcall_pending = 1;
++}
++EXPORT_SYMBOL_GPL(unmask_evtchn);
++
++void disable_all_local_evtchn(void)
++{
++ unsigned i, cpu = smp_processor_id();
++ shared_info_t *s = HYPERVISOR_shared_info;
++
++ for (i = 0; i < NR_EVENT_CHANNELS; ++i)
++ if (cpu_from_evtchn(i) == cpu)
++ synch_set_bit(i, &s->evtchn_mask[0]);
++}
++
++/* Clear an irq's pending state, in preparation for polling on it. */
++void xen_clear_irq_pending(int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ clear_evtchn(evtchn);
++}
++
++/* Set an irq's pending state, to avoid blocking on it. */
++void xen_set_irq_pending(int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn))
++ set_evtchn(evtchn);
++}
++
++/* Test an irq's pending state. */
++int xen_test_irq_pending(int irq)
++{
++ int evtchn = evtchn_from_irq(irq);
++
++ return VALID_EVTCHN(evtchn) && test_evtchn(evtchn);
++}
++
++/* Poll waiting for an irq to become pending. In the usual case, the
++ irq will be disabled so it won't deliver an interrupt. */
++void xen_poll_irq(int irq)
++{
++ evtchn_port_t evtchn = evtchn_from_irq(irq);
++
++ if (VALID_EVTCHN(evtchn)
++ && HYPERVISOR_poll(&evtchn, 1,
++ jiffies ^ (1UL << (BITS_PER_LONG - 1))))
++ BUG();
++}
++
++static void restore_cpu_virqs(unsigned int cpu)
++{
++ struct evtchn_bind_virq bind_virq;
++ int virq, irq, evtchn;
++
++ for (virq = 0; virq < NR_VIRQS; virq++) {
++ if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1)
++ continue;
++
++ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_VIRQ, virq, 0));
++
++ /* Get a new binding from Xen. */
++ bind_virq.virq = virq;
++ bind_virq.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq) != 0)
++ BUG();
++ evtchn = bind_virq.port;
++
++ /* Record the new mapping. */
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn);
++ bind_evtchn_to_cpu(evtchn, cpu);
++
++ /* Ready for use. */
++ unmask_evtchn(evtchn);
++ }
++}
++
++static void restore_cpu_ipis(unsigned int cpu)
++{
++ struct evtchn_bind_ipi bind_ipi;
++ int ipi, irq, evtchn;
++
++ for (ipi = 0; ipi < NR_IPIS; ipi++) {
++ if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1)
++ continue;
++
++ BUG_ON(irq_info[irq] != mk_irq_info(IRQT_IPI, ipi, 0));
++
++ /* Get a new binding from Xen. */
++ bind_ipi.vcpu = cpu;
++ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi,
++ &bind_ipi) != 0)
++ BUG();
++ evtchn = bind_ipi.port;
++
++ /* Record the new mapping. */
++ evtchn_to_irq[evtchn] = irq;
++ irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn);
++ bind_evtchn_to_cpu(evtchn, cpu);
++
++ /* Ready for use. */
++ unmask_evtchn(evtchn);
++
++ }
++}
++
++void irq_resume(void)
++{
++ unsigned int cpu, pirq, irq, evtchn;
++
++ init_evtchn_cpu_bindings();
++
++ /* New event-channel space is not 'live' yet. */
++ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
++ mask_evtchn(evtchn);
++
++ /* Check that no PIRQs are still bound. */
++ for (pirq = 0; pirq < NR_PIRQS; pirq++)
++ BUG_ON(irq_info[pirq_to_irq(pirq)] != IRQ_UNBOUND);
++
++ /* No IRQ <-> event-channel mappings. */
++ for (irq = 0; irq < NR_IRQS; irq++)
++ irq_info[irq] &= ~0xFFFF; /* zap event-channel binding */
++ for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++)
++ evtchn_to_irq[evtchn] = -1;
++
++ for_each_possible_cpu(cpu) {
++ restore_cpu_virqs(cpu);
++ restore_cpu_ipis(cpu);
++ }
++
++}
++
++void __init xen_init_IRQ(void)
++{
++ unsigned int i;
++
++ init_evtchn_cpu_bindings();
++
++ /* No event channels are 'live' right now. */
++ for (i = 0; i < NR_EVENT_CHANNELS; i++)
++ mask_evtchn(i);
++
++ /* No IRQ -> event-channel mappings. */
++ for (i = 0; i < NR_IRQS; i++)
++ irq_info[i] = IRQ_UNBOUND;
++
++ /* Dynamic IRQ space is currently unbound. Zero the refcnts. */
++ for (i = 0; i < NR_DYNIRQS; i++) {
++ irq_bindcount[dynirq_to_irq(i)] = 0;
++
++ 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;
++ 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. */
++ for (i = 0; i < NR_PIRQS; i++) {
++ irq_bindcount[pirq_to_irq(i)] = 1;
++
++#ifdef RTC_IRQ
++ /* If not domain 0, force our RTC driver to fail its probe. */
++ if ((i == RTC_IRQ) && !is_initial_xendomain())
++ continue;
++#endif
++
++ irq_desc[pirq_to_irq(i)].status = IRQ_DISABLED;
++ irq_desc[pirq_to_irq(i)].action = NULL;
++ irq_desc[pirq_to_irq(i)].depth = 1;
++ set_irq_chip_and_handler_name(pirq_to_irq(i), &pirq_chip,
++ handle_level_irq, "level");
++ }
++}
+diff --git a/drivers/xen/core/features.c b/drivers/xen/core/features.c
+new file mode 100644
+index 0000000..4c2b022
+--- /dev/null
++++ b/drivers/xen/core/features.c
+@@ -0,0 +1,34 @@
++/******************************************************************************
++ * features.c
++ *
++ * Xen feature flags.
++ *
++ * Copyright (c) 2006, Ian Campbell, XenSource Inc.
++ */
++#include <linux/types.h>
++#include <linux/cache.h>
++#include <linux/module.h>
++#include <asm/hypervisor.h>
++#include <xen/features.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++u8 xen_features[XENFEAT_NR_SUBMAPS * 32] __read_mostly;
++/* Not a GPL symbol: used in ubiquitous macros, so too restrictive. */
++EXPORT_SYMBOL(xen_features);
++
++void xen_setup_features(void)
++{
++ xen_feature_info_t fi;
++ int i, j;
++
++ for (i = 0; i < XENFEAT_NR_SUBMAPS; i++) {
++ fi.submap_idx = i;
++ if (HYPERVISOR_xen_version(XENVER_get_features, &fi) < 0)
++ break;
++ for (j=0; j<32; j++)
++ xen_features[i*32+j] = !!(fi.submap & 1<<j);
++ }
++}
+diff --git a/drivers/xen/core/firmware.c b/drivers/xen/core/firmware.c
+new file mode 100644
+index 0000000..91cbbd1
+--- /dev/null
++++ b/drivers/xen/core/firmware.c
+@@ -0,0 +1,74 @@
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/init.h>
++#include <linux/edd.h>
++#include <video/edid.h>
++#include <xen/interface/platform.h>
++#include <asm/hypervisor.h>
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++void __init copy_edd(void)
++{
++ int ret;
++ struct xen_platform_op op;
++
++ if (!is_initial_xendomain())
++ return;
++
++ op.cmd = XENPF_firmware_info;
++
++ op.u.firmware_info.type = XEN_FW_DISK_INFO;
++ for (op.u.firmware_info.index = 0;
++ edd.edd_info_nr < EDDMAXNR;
++ op.u.firmware_info.index++) {
++ struct edd_info *info = edd.edd_info + edd.edd_info_nr;
++
++ info->params.length = sizeof(info->params);
++ set_xen_guest_handle(op.u.firmware_info.u.disk_info.edd_params,
++ &info->params);
++ ret = HYPERVISOR_platform_op(&op);
++ if (ret)
++ break;
++
++#define C(x) info->x = op.u.firmware_info.u.disk_info.x
++ C(device);
++ C(version);
++ C(interface_support);
++ C(legacy_max_cylinder);
++ C(legacy_max_head);
++ C(legacy_sectors_per_track);
++#undef C
++
++ edd.edd_info_nr++;
++ }
++
++ op.u.firmware_info.type = XEN_FW_DISK_MBR_SIGNATURE;
++ for (op.u.firmware_info.index = 0;
++ edd.mbr_signature_nr < EDD_MBR_SIG_MAX;
++ op.u.firmware_info.index++) {
++ ret = HYPERVISOR_platform_op(&op);
++ if (ret)
++ break;
++ edd.mbr_signature[edd.mbr_signature_nr++] =
++ op.u.firmware_info.u.disk_mbr_signature.mbr_signature;
++ }
++}
++#endif
++
++void __init copy_edid(void)
++{
++#if defined(CONFIG_FIRMWARE_EDID) && defined(CONFIG_X86)
++ struct xen_platform_op op;
++
++ if (!is_initial_xendomain())
++ return;
++
++ op.cmd = XENPF_firmware_info;
++ op.u.firmware_info.index = 0;
++ op.u.firmware_info.type = XEN_FW_VBEDDC_INFO;
++ set_xen_guest_handle(op.u.firmware_info.u.vbeddc_info.edid,
++ edid_info.dummy);
++ if (HYPERVISOR_platform_op(&op) != 0)
++ memset(edid_info.dummy, 0x13, sizeof(edid_info.dummy));
++#endif
++}
+diff --git a/drivers/xen/core/gnttab.c b/drivers/xen/core/gnttab.c
+new file mode 100644
+index 0000000..6d3af45
+--- /dev/null
++++ b/drivers/xen/core/gnttab.c
+@@ -0,0 +1,769 @@
++/******************************************************************************
++ * gnttab.c
++ *
++ * Granting foreign access to our memory reservation.
++ *
++ * Copyright (c) 2005-2006, Christopher Clark
++ * Copyright (c) 2004-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/module.h>
++#include <linux/sched.h>
++#include <linux/mm.h>
++#include <linux/seqlock.h>
++#include <xen/interface/xen.h>
++#include <xen/gnttab.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/synch_bitops.h>
++#include <asm/io.h>
++#include <xen/interface/memory.h>
++#include <xen/driver_util.h>
++#include <asm/gnttab_dma.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++/* External tools reserve first few grant table entries. */
++#define NR_RESERVED_ENTRIES 8
++#define GNTTAB_LIST_END 0xffffffff
++#define ENTRIES_PER_GRANT_FRAME (PAGE_SIZE / sizeof(grant_entry_t))
++
++static grant_ref_t **gnttab_list;
++static unsigned int nr_grant_frames;
++static unsigned int boot_max_nr_grant_frames;
++static int gnttab_free_count;
++static grant_ref_t gnttab_free_head;
++static DEFINE_SPINLOCK(gnttab_list_lock);
++
++static struct grant_entry *shared;
++
++static struct gnttab_free_callback *gnttab_free_callback_list;
++
++static int gnttab_expand(unsigned int req_entries);
++
++#define RPP (PAGE_SIZE / sizeof(grant_ref_t))
++#define gnttab_entry(entry) (gnttab_list[(entry) / RPP][(entry) % RPP])
++
++#define nr_freelist_frames(grant_frames) \
++ (((grant_frames) * ENTRIES_PER_GRANT_FRAME + RPP - 1) / RPP)
++
++static int get_free_entries(int count)
++{
++ unsigned long flags;
++ int ref, rc;
++ grant_ref_t head;
++
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++
++ if ((gnttab_free_count < count) &&
++ ((rc = gnttab_expand(count - gnttab_free_count)) < 0)) {
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++ return rc;
++ }
++
++ ref = head = gnttab_free_head;
++ gnttab_free_count -= count;
++ while (count-- > 1)
++ head = gnttab_entry(head);
++ gnttab_free_head = gnttab_entry(head);
++ gnttab_entry(head) = GNTTAB_LIST_END;
++
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++
++ return ref;
++}
++
++#define get_free_entry() get_free_entries(1)
++
++static void do_free_callbacks(void)
++{
++ struct gnttab_free_callback *callback, *next;
++
++ callback = gnttab_free_callback_list;
++ gnttab_free_callback_list = NULL;
++
++ while (callback != NULL) {
++ next = callback->next;
++ if (gnttab_free_count >= callback->count) {
++ callback->next = NULL;
++ callback->fn(callback->arg);
++ } else {
++ callback->next = gnttab_free_callback_list;
++ gnttab_free_callback_list = callback;
++ }
++ callback = next;
++ }
++}
++
++static inline void check_free_callbacks(void)
++{
++ if (unlikely(gnttab_free_callback_list))
++ do_free_callbacks();
++}
++
++static void put_free_entry(grant_ref_t ref)
++{
++ unsigned long flags;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ gnttab_entry(ref) = gnttab_free_head;
++ gnttab_free_head = ref;
++ gnttab_free_count++;
++ check_free_callbacks();
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++
++/*
++ * Public grant-issuing interface functions
++ */
++
++int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
++ int flags)
++{
++ int ref;
++
++ if (unlikely((ref = get_free_entry()) < 0))
++ return -ENOSPC;
++
++ shared[ref].frame = frame;
++ shared[ref].domid = domid;
++ wmb();
++ BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing));
++ shared[ref].flags = GTF_permit_access | flags;
++
++ return ref;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access);
++
++void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
++ unsigned long frame, int flags)
++{
++ shared[ref].frame = frame;
++ shared[ref].domid = domid;
++ wmb();
++ BUG_ON(flags & (GTF_accept_transfer | GTF_reading | GTF_writing));
++ shared[ref].flags = GTF_permit_access | flags;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_access_ref);
++
++
++int gnttab_query_foreign_access(grant_ref_t ref)
++{
++ u16 nflags;
++
++ nflags = shared[ref].flags;
++
++ return (nflags & (GTF_reading|GTF_writing));
++}
++EXPORT_SYMBOL_GPL(gnttab_query_foreign_access);
++
++int gnttab_end_foreign_access_ref(grant_ref_t ref)
++{
++ u16 flags, nflags;
++
++ nflags = shared[ref].flags;
++ do {
++ if ((flags = nflags) & (GTF_reading|GTF_writing)) {
++ printk(KERN_DEBUG "WARNING: g.e. still in use!\n");
++ return 0;
++ }
++ } while ((nflags = synch_cmpxchg_subword(&shared[ref].flags, flags, 0)) !=
++ flags);
++
++ return 1;
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_access_ref);
++
++void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page)
++{
++ if (gnttab_end_foreign_access_ref(ref)) {
++ put_free_entry(ref);
++ if (page != 0)
++ free_page(page);
++ } else {
++ /* XXX This needs to be fixed so that the ref and page are
++ placed on a list to be freed up later. */
++ printk(KERN_DEBUG
++ "WARNING: leaking g.e. and page still in use!\n");
++ }
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_access);
++
++int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn)
++{
++ int ref;
++
++ if (unlikely((ref = get_free_entry()) < 0))
++ return -ENOSPC;
++ gnttab_grant_foreign_transfer_ref(ref, domid, pfn);
++
++ return ref;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer);
++
++void gnttab_grant_foreign_transfer_ref(grant_ref_t ref, domid_t domid,
++ unsigned long pfn)
++{
++ shared[ref].frame = pfn;
++ shared[ref].domid = domid;
++ wmb();
++ shared[ref].flags = GTF_accept_transfer;
++}
++EXPORT_SYMBOL_GPL(gnttab_grant_foreign_transfer_ref);
++
++unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref)
++{
++ unsigned long frame;
++ u16 flags;
++
++ /*
++ * If a transfer is not even yet started, try to reclaim the grant
++ * reference and return failure (== 0).
++ */
++ while (!((flags = shared[ref].flags) & GTF_transfer_committed)) {
++ if (synch_cmpxchg_subword(&shared[ref].flags, flags, 0) == flags)
++ return 0;
++ cpu_relax();
++ }
++
++ /* If a transfer is in progress then wait until it is completed. */
++ while (!(flags & GTF_transfer_completed)) {
++ flags = shared[ref].flags;
++ cpu_relax();
++ }
++
++ /* Read the frame number /after/ reading completion status. */
++ rmb();
++ frame = shared[ref].frame;
++ BUG_ON(frame == 0);
++
++ return frame;
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer_ref);
++
++unsigned long gnttab_end_foreign_transfer(grant_ref_t ref)
++{
++ unsigned long frame = gnttab_end_foreign_transfer_ref(ref);
++ put_free_entry(ref);
++ return frame;
++}
++EXPORT_SYMBOL_GPL(gnttab_end_foreign_transfer);
++
++void gnttab_free_grant_reference(grant_ref_t ref)
++{
++ put_free_entry(ref);
++}
++EXPORT_SYMBOL_GPL(gnttab_free_grant_reference);
++
++void gnttab_free_grant_references(grant_ref_t head)
++{
++ grant_ref_t ref;
++ unsigned long flags;
++ int count = 1;
++ if (head == GNTTAB_LIST_END)
++ return;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ ref = head;
++ while (gnttab_entry(ref) != GNTTAB_LIST_END) {
++ ref = gnttab_entry(ref);
++ count++;
++ }
++ gnttab_entry(ref) = gnttab_free_head;
++ gnttab_free_head = head;
++ gnttab_free_count += count;
++ check_free_callbacks();
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_free_grant_references);
++
++int gnttab_alloc_grant_references(u16 count, grant_ref_t *head)
++{
++ int h = get_free_entries(count);
++
++ if (h < 0)
++ return -ENOSPC;
++
++ *head = h;
++
++ return 0;
++}
++EXPORT_SYMBOL_GPL(gnttab_alloc_grant_references);
++
++int gnttab_empty_grant_references(const grant_ref_t *private_head)
++{
++ return (*private_head == GNTTAB_LIST_END);
++}
++EXPORT_SYMBOL_GPL(gnttab_empty_grant_references);
++
++int gnttab_claim_grant_reference(grant_ref_t *private_head)
++{
++ grant_ref_t g = *private_head;
++ if (unlikely(g == GNTTAB_LIST_END))
++ return -ENOSPC;
++ *private_head = gnttab_entry(g);
++ return g;
++}
++EXPORT_SYMBOL_GPL(gnttab_claim_grant_reference);
++
++void gnttab_release_grant_reference(grant_ref_t *private_head,
++ grant_ref_t release)
++{
++ gnttab_entry(release) = *private_head;
++ *private_head = release;
++}
++EXPORT_SYMBOL_GPL(gnttab_release_grant_reference);
++
++void gnttab_request_free_callback(struct gnttab_free_callback *callback,
++ void (*fn)(void *), void *arg, u16 count)
++{
++ unsigned long flags;
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ if (callback->next)
++ goto out;
++ callback->fn = fn;
++ callback->arg = arg;
++ callback->count = count;
++ callback->next = gnttab_free_callback_list;
++ gnttab_free_callback_list = callback;
++ check_free_callbacks();
++out:
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_request_free_callback);
++
++void gnttab_cancel_free_callback(struct gnttab_free_callback *callback)
++{
++ struct gnttab_free_callback **pcb;
++ unsigned long flags;
++
++ spin_lock_irqsave(&gnttab_list_lock, flags);
++ for (pcb = &gnttab_free_callback_list; *pcb; pcb = &(*pcb)->next) {
++ if (*pcb == callback) {
++ *pcb = callback->next;
++ break;
++ }
++ }
++ spin_unlock_irqrestore(&gnttab_list_lock, flags);
++}
++EXPORT_SYMBOL_GPL(gnttab_cancel_free_callback);
++
++static int grow_gnttab_list(unsigned int more_frames)
++{
++ unsigned int new_nr_grant_frames, extra_entries, i;
++ unsigned int nr_glist_frames, new_nr_glist_frames;
++
++ new_nr_grant_frames = nr_grant_frames + more_frames;
++ extra_entries = more_frames * ENTRIES_PER_GRANT_FRAME;
++
++ nr_glist_frames = nr_freelist_frames(nr_grant_frames);
++ new_nr_glist_frames = nr_freelist_frames(new_nr_grant_frames);
++ for (i = nr_glist_frames; i < new_nr_glist_frames; i++) {
++ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_ATOMIC);
++ if (!gnttab_list[i])
++ goto grow_nomem;
++ }
++
++ for (i = ENTRIES_PER_GRANT_FRAME * nr_grant_frames;
++ i < ENTRIES_PER_GRANT_FRAME * new_nr_grant_frames - 1; i++)
++ gnttab_entry(i) = i + 1;
++
++ gnttab_entry(i) = gnttab_free_head;
++ gnttab_free_head = ENTRIES_PER_GRANT_FRAME * nr_grant_frames;
++ gnttab_free_count += extra_entries;
++
++ nr_grant_frames = new_nr_grant_frames;
++
++ check_free_callbacks();
++
++ return 0;
++
++grow_nomem:
++ for ( ; i >= nr_glist_frames; i--)
++ free_page((unsigned long) gnttab_list[i]);
++ return -ENOMEM;
++}
++
++static unsigned int __max_nr_grant_frames(void)
++{
++ struct gnttab_query_size query;
++ int rc;
++
++ query.dom = DOMID_SELF;
++
++ rc = HYPERVISOR_grant_table_op(GNTTABOP_query_size, &query, 1);
++ if ((rc < 0) || (query.status != GNTST_okay))
++ return 4; /* Legacy max supported number of frames */
++
++ return query.max_nr_frames;
++}
++
++static inline unsigned int max_nr_grant_frames(void)
++{
++ unsigned int xen_max = __max_nr_grant_frames();
++
++ if (xen_max > boot_max_nr_grant_frames)
++ return boot_max_nr_grant_frames;
++ return xen_max;
++}
++
++#ifdef CONFIG_XEN
++
++static DEFINE_SEQLOCK(gnttab_dma_lock);
++
++#ifdef CONFIG_X86
++static int map_pte_fn(pte_t *pte, struct page *pmd_page,
++ unsigned long addr, void *data)
++{
++ unsigned long **frames = (unsigned long **)data;
++
++ set_pte_at(&init_mm, addr, pte, pfn_pte_ma((*frames)[0], PAGE_KERNEL));
++ (*frames)++;
++ return 0;
++}
++
++static int unmap_pte_fn(pte_t *pte, struct page *pmd_page,
++ unsigned long addr, void *data)
++{
++
++ set_pte_at(&init_mm, addr, pte, __pte(0));
++ return 0;
++}
++
++void *arch_gnttab_alloc_shared(unsigned long *frames)
++{
++ struct vm_struct *area;
++ area = alloc_vm_area(PAGE_SIZE * max_nr_grant_frames());
++ BUG_ON(area == NULL);
++ return area->addr;
++}
++#endif /* CONFIG_X86 */
++
++static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
++{
++ struct gnttab_setup_table setup;
++ unsigned long *frames;
++ unsigned int nr_gframes = end_idx + 1;
++ int rc;
++
++ frames = kmalloc(nr_gframes * sizeof(unsigned long), GFP_ATOMIC);
++ if (!frames)
++ return -ENOMEM;
++
++ setup.dom = DOMID_SELF;
++ setup.nr_frames = nr_gframes;
++ set_xen_guest_handle(setup.frame_list, frames);
++
++ rc = HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup, 1);
++ if (rc == -ENOSYS) {
++ kfree(frames);
++ return -ENOSYS;
++ }
++
++ BUG_ON(rc || setup.status);
++
++ if (shared == NULL)
++ shared = arch_gnttab_alloc_shared(frames);
++
++#ifdef CONFIG_X86
++ rc = apply_to_page_range(&init_mm, (unsigned long)shared,
++ PAGE_SIZE * nr_gframes,
++ map_pte_fn, &frames);
++ BUG_ON(rc);
++ frames -= nr_gframes; /* adjust after map_pte_fn() */
++#endif /* CONFIG_X86 */
++
++ kfree(frames);
++
++ return 0;
++}
++
++static void gnttab_page_free(struct page *page)
++{
++ ClearPageForeign(page);
++ gnttab_reset_grant_page(page);
++ put_page(page);
++}
++
++/*
++ * Must not be called with IRQs off. This should only be used on the
++ * slow path.
++ *
++ * Copy a foreign granted page to local memory.
++ */
++int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep)
++{
++ struct gnttab_unmap_and_replace unmap;
++ mmu_update_t mmu;
++ struct page *page;
++ struct page *new_page;
++ void *new_addr;
++ void *addr;
++ paddr_t pfn;
++ maddr_t mfn;
++ maddr_t new_mfn;
++ int err;
++
++ page = *pagep;
++ if (!get_page_unless_zero(page))
++ return -ENOENT;
++
++ err = -ENOMEM;
++ new_page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (!new_page)
++ goto out;
++
++ new_addr = page_address(new_page);
++ addr = page_address(page);
++ memcpy(new_addr, addr, PAGE_SIZE);
++
++ pfn = page_to_pfn(page);
++ mfn = pfn_to_mfn(pfn);
++ new_mfn = virt_to_mfn(new_addr);
++
++ write_seqlock(&gnttab_dma_lock);
++
++ /* Make seq visible before checking page_mapped. */
++ smp_mb();
++
++ /* Has the page been DMA-mapped? */
++ if (unlikely(page_mapped(page))) {
++ write_sequnlock(&gnttab_dma_lock);
++ put_page(new_page);
++ err = -EBUSY;
++ goto out;
++ }
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap))
++ set_phys_to_machine(pfn, new_mfn);
++
++ gnttab_set_replace_op(&unmap, (unsigned long)addr,
++ (unsigned long)new_addr, ref);
++
++ err = HYPERVISOR_grant_table_op(GNTTABOP_unmap_and_replace,
++ &unmap, 1);
++ BUG_ON(err);
++ BUG_ON(unmap.status);
++
++ write_sequnlock(&gnttab_dma_lock);
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ set_phys_to_machine(page_to_pfn(new_page), INVALID_P2M_ENTRY);
++
++ mmu.ptr = (new_mfn << PAGE_SHIFT) | MMU_MACHPHYS_UPDATE;
++ mmu.val = pfn;
++ err = HYPERVISOR_mmu_update(&mmu, 1, NULL, DOMID_SELF);
++ BUG_ON(err);
++ }
++
++ new_page->mapping = page->mapping;
++ new_page->index = page->index;
++ set_bit(PG_foreign, &new_page->flags);
++ *pagep = new_page;
++
++ SetPageForeign(page, gnttab_page_free);
++ page->mapping = NULL;
++
++out:
++ put_page(page);
++ return err;
++}
++EXPORT_SYMBOL_GPL(gnttab_copy_grant_page);
++
++void gnttab_reset_grant_page(struct page *page)
++{
++ init_page_count(page);
++ reset_page_mapcount(page);
++}
++EXPORT_SYMBOL_GPL(gnttab_reset_grant_page);
++
++/*
++ * Keep track of foreign pages marked as PageForeign so that we don't
++ * return them to the remote domain prematurely.
++ *
++ * PageForeign pages are pinned down by increasing their mapcount.
++ *
++ * All other pages are simply returned as is.
++ */
++void __gnttab_dma_map_page(struct page *page)
++{
++ unsigned int seq;
++
++ if (!is_running_on_xen() || !PageForeign(page))
++ return;
++
++ do {
++ seq = read_seqbegin(&gnttab_dma_lock);
++
++ if (gnttab_dma_local_pfn(page))
++ break;
++
++ atomic_set(&page->_mapcount, 0);
++
++ /* Make _mapcount visible before read_seqretry. */
++ smp_mb();
++ } while (unlikely(read_seqretry(&gnttab_dma_lock, seq)));
++}
++
++int gnttab_resume(void)
++{
++ if (max_nr_grant_frames() < nr_grant_frames)
++ return -ENOSYS;
++ return gnttab_map(0, nr_grant_frames - 1);
++}
++
++int gnttab_suspend(void)
++{
++#ifdef CONFIG_X86
++ apply_to_page_range(&init_mm, (unsigned long)shared,
++ PAGE_SIZE * nr_grant_frames,
++ unmap_pte_fn, NULL);
++#endif
++ return 0;
++}
++
++#else /* !CONFIG_XEN */
++
++#include <platform-pci.h>
++
++static unsigned long resume_frames;
++
++static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
++{
++ struct xen_add_to_physmap xatp;
++ unsigned int i = end_idx;
++
++ /* Loop backwards, so that the first hypercall has the largest index,
++ * ensuring that the table will grow only once.
++ */
++ do {
++ xatp.domid = DOMID_SELF;
++ xatp.idx = i;
++ xatp.space = XENMAPSPACE_grant_table;
++ xatp.gpfn = (resume_frames >> PAGE_SHIFT) + i;
++ if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
++ BUG();
++ } while (i-- > start_idx);
++
++ return 0;
++}
++
++int gnttab_resume(void)
++{
++ unsigned int max_nr_gframes, nr_gframes;
++
++ nr_gframes = nr_grant_frames;
++ max_nr_gframes = max_nr_grant_frames();
++ if (max_nr_gframes < nr_gframes)
++ return -ENOSYS;
++
++ if (!resume_frames) {
++ resume_frames = alloc_xen_mmio(PAGE_SIZE * max_nr_gframes);
++ shared = ioremap(resume_frames, PAGE_SIZE * max_nr_gframes);
++ if (shared == NULL) {
++ printk("error to ioremap gnttab share frames\n");
++ return -1;
++ }
++ }
++
++ gnttab_map(0, nr_gframes - 1);
++
++ return 0;
++}
++
++#endif /* !CONFIG_XEN */
++
++static int gnttab_expand(unsigned int req_entries)
++{
++ int rc;
++ unsigned int cur, extra;
++
++ cur = nr_grant_frames;
++ extra = ((req_entries + (ENTRIES_PER_GRANT_FRAME-1)) /
++ ENTRIES_PER_GRANT_FRAME);
++ if (cur + extra > max_nr_grant_frames())
++ return -ENOSPC;
++
++ if ((rc = gnttab_map(cur, cur + extra - 1)) == 0)
++ rc = grow_gnttab_list(extra);
++
++ return rc;
++}
++
++int __devinit gnttab_init(void)
++{
++ int i;
++ unsigned int max_nr_glist_frames, nr_glist_frames;
++ unsigned int nr_init_grefs;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ nr_grant_frames = 1;
++ boot_max_nr_grant_frames = __max_nr_grant_frames();
++
++ /* Determine the maximum number of frames required for the
++ * grant reference free list on the current hypervisor.
++ */
++ max_nr_glist_frames = nr_freelist_frames(boot_max_nr_grant_frames);
++
++ gnttab_list = kmalloc(max_nr_glist_frames * sizeof(grant_ref_t *),
++ GFP_KERNEL);
++ if (gnttab_list == NULL)
++ return -ENOMEM;
++
++ nr_glist_frames = nr_freelist_frames(nr_grant_frames);
++ for (i = 0; i < nr_glist_frames; i++) {
++ gnttab_list[i] = (grant_ref_t *)__get_free_page(GFP_KERNEL);
++ if (gnttab_list[i] == NULL)
++ goto ini_nomem;
++ }
++
++ if (gnttab_resume() < 0)
++ return -ENODEV;
++
++ nr_init_grefs = nr_grant_frames * ENTRIES_PER_GRANT_FRAME;
++
++ for (i = NR_RESERVED_ENTRIES; i < nr_init_grefs - 1; i++)
++ gnttab_entry(i) = i + 1;
++
++ gnttab_entry(nr_init_grefs - 1) = GNTTAB_LIST_END;
++ gnttab_free_count = nr_init_grefs - NR_RESERVED_ENTRIES;
++ gnttab_free_head = NR_RESERVED_ENTRIES;
++
++ return 0;
++
++ ini_nomem:
++ for (i--; i >= 0; i--)
++ free_page((unsigned long)gnttab_list[i]);
++ kfree(gnttab_list);
++ return -ENOMEM;
++}
++
++#ifdef CONFIG_XEN
++core_initcall(gnttab_init);
++#endif
+diff --git a/drivers/xen/core/hypervisor_sysfs.c b/drivers/xen/core/hypervisor_sysfs.c
+new file mode 100644
+index 0000000..f9ce6a5
+--- /dev/null
++++ b/drivers/xen/core/hypervisor_sysfs.c
+@@ -0,0 +1,57 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
++
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/kobject.h>
++#include <xen/hypervisor_sysfs.h>
++#include <asm/hypervisor.h>
++
++static ssize_t hyp_sysfs_show(struct kobject *kobj,
++ struct attribute *attr,
++ char *buffer)
++{
++ struct hyp_sysfs_attr *hyp_attr;
++ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
++ if (hyp_attr->show)
++ return hyp_attr->show(hyp_attr, buffer);
++ return 0;
++}
++
++static ssize_t hyp_sysfs_store(struct kobject *kobj,
++ struct attribute *attr,
++ const char *buffer,
++ size_t len)
++{
++ struct hyp_sysfs_attr *hyp_attr;
++ hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr);
++ if (hyp_attr->store)
++ return hyp_attr->store(hyp_attr, buffer, len);
++ return 0;
++}
++
++static struct sysfs_ops hyp_sysfs_ops = {
++ .show = hyp_sysfs_show,
++ .store = hyp_sysfs_store,
++};
++
++static struct kobj_type hyp_sysfs_kobj_type = {
++ .sysfs_ops = &hyp_sysfs_ops,
++};
++
++static int __init hypervisor_subsys_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ hypervisor_kobj->ktype = &hyp_sysfs_kobj_type;
++ return 0;
++}
++
++device_initcall(hypervisor_subsys_init);
+diff --git a/drivers/xen/core/machine_kexec.c b/drivers/xen/core/machine_kexec.c
+new file mode 100644
+index 0000000..1c14a39
+--- /dev/null
++++ b/drivers/xen/core/machine_kexec.c
+@@ -0,0 +1,216 @@
++/*
++ * drivers/xen/core/machine_kexec.c
++ * handle transition of Linux booting another kernel
++ */
++
++#include <linux/kexec.h>
++#include <xen/interface/kexec.h>
++#include <linux/mm.h>
++#include <linux/bootmem.h>
++
++extern void machine_kexec_setup_load_arg(xen_kexec_image_t *xki,
++ struct kimage *image);
++extern int machine_kexec_setup_resources(struct resource *hypervisor,
++ struct resource *phys_cpus,
++ int nr_phys_cpus);
++extern void machine_kexec_register_resources(struct resource *res);
++
++static int __initdata xen_max_nr_phys_cpus;
++static struct resource xen_hypervisor_res;
++static struct resource *xen_phys_cpus;
++
++size_t vmcoreinfo_size_xen;
++unsigned long paddr_vmcoreinfo_xen;
++
++void __init xen_machine_kexec_setup_resources(void)
++{
++ xen_kexec_range_t range;
++ struct resource *res;
++ int k = 0;
++
++ if (strstr(boot_command_line, "crashkernel="))
++ printk(KERN_WARNING "Ignoring crashkernel command line, "
++ "parameter will be supplied by xen\n");
++
++ if (!is_initial_xendomain())
++ return;
++
++ /* determine maximum number of physical cpus */
++
++ while (1) {
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CPU;
++ range.nr = k;
++
++ if(HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ break;
++
++ k++;
++ }
++
++ if (k == 0)
++ return;
++
++ xen_max_nr_phys_cpus = k;
++
++ /* allocate xen_phys_cpus */
++
++ xen_phys_cpus = alloc_bootmem_low(k * sizeof(struct resource));
++ BUG_ON(xen_phys_cpus == NULL);
++
++ /* fill in xen_phys_cpus with per-cpu crash note information */
++
++ for (k = 0; k < xen_max_nr_phys_cpus; k++) {
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CPU;
++ range.nr = k;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
++
++ res = xen_phys_cpus + k;
++
++ memset(res, 0, sizeof(*res));
++ res->name = "Crash note";
++ res->start = range.start;
++ res->end = range.start + range.size - 1;
++ res->flags = IORESOURCE_BUSY | IORESOURCE_MEM;
++ }
++
++ /* fill in xen_hypervisor_res with hypervisor machine address range */
++
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_XEN;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
++
++ xen_hypervisor_res.name = "Hypervisor code and data";
++ xen_hypervisor_res.start = range.start;
++ xen_hypervisor_res.end = range.start + range.size - 1;
++ xen_hypervisor_res.flags = IORESOURCE_BUSY | IORESOURCE_MEM;
++#ifdef CONFIG_X86_64
++ insert_resource(iomem_resource, &xen_hypervisor_res);
++#endif
++
++ /* fill in crashk_res if range is reserved by hypervisor */
++
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_CRASH;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ goto err;
++
++ if (range.size) {
++ crashk_res.start = range.start;
++ crashk_res.end = range.start + range.size - 1;
++ }
++
++ /* get physical address of vmcoreinfo */
++ memset(&range, 0, sizeof(range));
++ range.range = KEXEC_RANGE_MA_VMCOREINFO;
++
++ if (HYPERVISOR_kexec_op(KEXEC_CMD_kexec_get_range, &range))
++ return;
++
++ if (range.size) {
++ paddr_vmcoreinfo_xen = range.start;
++ vmcoreinfo_size_xen = range.size;
++ }
++
++ if (machine_kexec_setup_resources(&xen_hypervisor_res, xen_phys_cpus,
++ xen_max_nr_phys_cpus))
++ goto err;
++
++ return;
++
++ err:
++ /*
++ * It isn't possible to free xen_phys_cpus this early in the
++ * boot. Failure at this stage is unexpected and the amount of
++ * memory is small therefore we tolerate the potential leak.
++ */
++ xen_max_nr_phys_cpus = 0;
++ return;
++}
++
++#ifndef CONFIG_X86_64
++void __init xen_machine_kexec_register_resources(struct resource *res)
++{
++ request_resource(res, &xen_hypervisor_res);
++ machine_kexec_register_resources(res);
++}
++#endif
++
++static void setup_load_arg(xen_kexec_image_t *xki, struct kimage *image)
++{
++ machine_kexec_setup_load_arg(xki, image);
++
++ xki->indirection_page = image->head;
++ xki->start_address = image->start;
++}
++
++/*
++ * Load the image into xen so xen can kdump itself
++ * This might have been done in prepare, but prepare
++ * is currently called too early. It might make sense
++ * to move prepare, but for now, just add an extra hook.
++ */
++int xen_machine_kexec_load(struct kimage *image)
++{
++ xen_kexec_load_t xkl;
++
++ memset(&xkl, 0, sizeof(xkl));
++ xkl.type = image->type;
++ setup_load_arg(&xkl.image, image);
++ return HYPERVISOR_kexec_op(KEXEC_CMD_kexec_load, &xkl);
++}
++
++/*
++ * Unload the image that was stored by machine_kexec_load()
++ * This might have been done in machine_kexec_cleanup() but it
++ * is called too late, and its possible xen could try and kdump
++ * using resources that have been freed.
++ */
++void xen_machine_kexec_unload(struct kimage *image)
++{
++ xen_kexec_load_t xkl;
++
++ memset(&xkl, 0, sizeof(xkl));
++ xkl.type = image->type;
++ WARN_ON(HYPERVISOR_kexec_op(KEXEC_CMD_kexec_unload, &xkl));
++}
++
++/*
++ * Do not allocate memory (or fail in any way) in machine_kexec().
++ * We are past the point of no return, committed to rebooting now.
++ *
++ * This has the hypervisor move to the prefered reboot CPU,
++ * stop all CPUs and kexec. That is it combines machine_shutdown()
++ * and machine_kexec() in Linux kexec terms.
++ */
++NORET_TYPE void machine_kexec(struct kimage *image)
++{
++ xen_kexec_exec_t xke;
++
++ memset(&xke, 0, sizeof(xke));
++ xke.type = image->type;
++ VOID(HYPERVISOR_kexec_op(KEXEC_CMD_kexec, &xke));
++ panic("KEXEC_CMD_kexec hypercall should not return\n");
++}
++
++void machine_shutdown(void)
++{
++ /* do nothing */
++}
++
++
++/*
++ * Local variables:
++ * c-file-style: "linux"
++ * indent-tabs-mode: t
++ * c-indent-level: 8
++ * c-basic-offset: 8
++ * tab-width: 8
++ * End:
++ */
+diff --git a/drivers/xen/core/machine_reboot.c b/drivers/xen/core/machine_reboot.c
+new file mode 100644
+index 0000000..5a13599
+--- /dev/null
++++ b/drivers/xen/core/machine_reboot.c
+@@ -0,0 +1,252 @@
++#include <linux/version.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/unistd.h>
++#include <linux/module.h>
++#include <linux/reboot.h>
++#include <linux/sysrq.h>
++#include <linux/stringify.h>
++#include <linux/stop_machine.h>
++#include <asm/irq.h>
++#include <asm/mmu_context.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <linux/cpu.h>
++#include <xen/gnttab.h>
++#include <xen/xencons.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/interface/vcpu.h>
++
++#if defined(__i386__) || defined(__x86_64__)
++
++/*
++ * Power off function, if any
++ */
++void (*pm_power_off)(void);
++EXPORT_SYMBOL(pm_power_off);
++
++void machine_emergency_restart(void)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ HYPERVISOR_shutdown(SHUTDOWN_reboot);
++}
++
++void machine_restart(char * __unused)
++{
++ machine_emergency_restart();
++}
++
++void machine_halt(void)
++{
++ machine_power_off();
++}
++
++void machine_power_off(void)
++{
++ /* We really want to get pending console data out before we die. */
++ xencons_force_flush();
++ if (pm_power_off)
++ pm_power_off();
++ HYPERVISOR_shutdown(SHUTDOWN_poweroff);
++}
++
++int reboot_thru_bios = 0; /* for dmi_scan.c */
++EXPORT_SYMBOL(machine_restart);
++EXPORT_SYMBOL(machine_halt);
++EXPORT_SYMBOL(machine_power_off);
++
++static void pre_suspend(void)
++{
++ HYPERVISOR_shared_info = (shared_info_t *)empty_zero_page;
++ WARN_ON(HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
++ __pte_ma(0), 0));
++
++ xen_start_info->store_mfn = mfn_to_pfn(xen_start_info->store_mfn);
++ xen_start_info->console.domU.mfn =
++ mfn_to_pfn(xen_start_info->console.domU.mfn);
++}
++
++static void post_suspend(int suspend_cancelled)
++{
++ int i, j, k, fpp;
++ unsigned long shinfo_mfn;
++ extern unsigned long max_pfn;
++ extern unsigned long *pfn_to_mfn_frame_list_list;
++ extern unsigned long *pfn_to_mfn_frame_list[];
++
++ if (suspend_cancelled) {
++ xen_start_info->store_mfn =
++ pfn_to_mfn(xen_start_info->store_mfn);
++ xen_start_info->console.domU.mfn =
++ pfn_to_mfn(xen_start_info->console.domU.mfn);
++ } else {
++#ifdef CONFIG_SMP
++ cpu_initialized_map = cpu_online_map;
++#endif
++ }
++
++ shinfo_mfn = xen_start_info->shared_info >> PAGE_SHIFT;
++ if (HYPERVISOR_update_va_mapping(fix_to_virt(FIX_SHARED_INFO),
++ pfn_pte_ma(shinfo_mfn, PAGE_KERNEL),
++ 0))
++ BUG();
++ HYPERVISOR_shared_info = (shared_info_t *)fix_to_virt(FIX_SHARED_INFO);
++
++ memset(empty_zero_page, 0, PAGE_SIZE);
++
++ fpp = PAGE_SIZE/sizeof(unsigned long);
++ for (i = 0, j = 0, k = -1; i < max_pfn; i += fpp, j++) {
++ if ((j % fpp) == 0) {
++ k++;
++ pfn_to_mfn_frame_list_list[k] =
++ virt_to_mfn(pfn_to_mfn_frame_list[k]);
++ j = 0;
++ }
++ pfn_to_mfn_frame_list[k][j] =
++ virt_to_mfn(&phys_to_machine_mapping[i]);
++ }
++ HYPERVISOR_shared_info->arch.max_pfn = max_pfn;
++ HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
++ virt_to_mfn(pfn_to_mfn_frame_list_list);
++}
++
++#else /* !(defined(__i386__) || defined(__x86_64__)) */
++
++#ifndef HAVE_XEN_PRE_SUSPEND
++#define xen_pre_suspend() ((void)0)
++#endif
++
++#ifndef HAVE_XEN_POST_SUSPEND
++#define xen_post_suspend(x) ((void)0)
++#endif
++
++#define switch_idle_mm() ((void)0)
++#define mm_pin_all() ((void)0)
++#define pre_suspend() xen_pre_suspend()
++#define post_suspend(x) xen_post_suspend(x)
++
++#endif
++
++struct suspend {
++ int fast_suspend;
++ void (*resume_notifier)(int);
++};
++
++static int take_machine_down(void *_suspend)
++{
++ struct suspend *suspend = _suspend;
++ int suspend_cancelled, err;
++ extern void time_resume(void);
++
++ if (suspend->fast_suspend) {
++ BUG_ON(!irqs_disabled());
++ } else {
++ BUG_ON(irqs_disabled());
++
++ for (;;) {
++ err = smp_suspend();
++ if (err)
++ return err;
++
++ xenbus_suspend();
++ preempt_disable();
++
++ if (num_online_cpus() == 1)
++ break;
++
++ preempt_enable();
++ xenbus_suspend_cancel();
++ }
++
++ local_irq_disable();
++ }
++
++ mm_pin_all();
++ gnttab_suspend();
++ pre_suspend();
++
++ /*
++ * This hypercall returns 1 if suspend was cancelled or the domain was
++ * merely checkpointed, and 0 if it is resuming in a new domain.
++ */
++ suspend_cancelled = HYPERVISOR_suspend(virt_to_mfn(xen_start_info));
++
++ suspend->resume_notifier(suspend_cancelled);
++ post_suspend(suspend_cancelled);
++ gnttab_resume();
++ if (!suspend_cancelled) {
++ irq_resume();
++#ifdef __x86_64__
++ /*
++ * Older versions of Xen do not save/restore the user %cr3.
++ * We do it here just in case, but there's no need if we are
++ * in fast-suspend mode as that implies a new enough Xen.
++ */
++ if (!suspend->fast_suspend) {
++ struct mmuext_op op;
++ op.cmd = MMUEXT_NEW_USER_BASEPTR;
++ op.arg1.mfn = pfn_to_mfn(__pa(__user_pgd(
++ current->active_mm->pgd)) >> PAGE_SHIFT);
++ if (HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF))
++ BUG();
++ }
++#endif
++ }
++ time_resume();
++
++ if (!suspend->fast_suspend)
++ local_irq_enable();
++
++ return suspend_cancelled;
++}
++
++int __xen_suspend(int fast_suspend, void (*resume_notifier)(int))
++{
++ int err, suspend_cancelled;
++ struct suspend suspend;
++
++ BUG_ON(smp_processor_id() != 0);
++ BUG_ON(in_interrupt());
++
++#if defined(__i386__) || defined(__x86_64__)
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ printk(KERN_WARNING "Cannot suspend in "
++ "auto_translated_physmap mode.\n");
++ return -EOPNOTSUPP;
++ }
++#endif
++
++ /* If we are definitely UP then 'slow mode' is actually faster. */
++ if (num_possible_cpus() == 1)
++ fast_suspend = 0;
++
++ suspend.fast_suspend = fast_suspend;
++ suspend.resume_notifier = resume_notifier;
++
++ if (fast_suspend) {
++ xenbus_suspend();
++ err = stop_machine_run(take_machine_down, &suspend, 0);
++ if (err < 0)
++ xenbus_suspend_cancel();
++ } else {
++ err = take_machine_down(&suspend);
++ }
++
++ if (err < 0)
++ return err;
++
++ suspend_cancelled = err;
++ if (!suspend_cancelled) {
++ xencons_resume();
++ xenbus_resume();
++ } else {
++ xenbus_suspend_cancel();
++ }
++
++ if (!fast_suspend)
++ smp_resume();
++
++ return 0;
++}
+diff --git a/drivers/xen/core/pci.c b/drivers/xen/core/pci.c
+new file mode 100644
+index 0000000..2710c18
+--- /dev/null
++++ b/drivers/xen/core/pci.c
+@@ -0,0 +1,63 @@
++/*
++ * vim:shiftwidth=8:noexpandtab
++ */
++
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <xen/interface/physdev.h>
++
++static int (*pci_bus_probe)(struct device *dev);
++static int (*pci_bus_remove)(struct device *dev);
++
++static int pci_bus_probe_wrapper(struct device *dev)
++{
++ int r;
++ struct pci_dev *pci_dev = to_pci_dev(dev);
++ struct physdev_manage_pci manage_pci;
++ manage_pci.bus = pci_dev->bus->number;
++ manage_pci.devfn = pci_dev->devfn;
++
++ r = HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_add, &manage_pci);
++ if (r && r != -ENOSYS)
++ return r;
++
++ r = pci_bus_probe(dev);
++ if (r)
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci));
++
++ return r;
++}
++
++static int pci_bus_remove_wrapper(struct device *dev)
++{
++ int r;
++ struct pci_dev *pci_dev = to_pci_dev(dev);
++ struct physdev_manage_pci manage_pci;
++ manage_pci.bus = pci_dev->bus->number;
++ manage_pci.devfn = pci_dev->devfn;
++
++ r = pci_bus_remove(dev);
++ /* dev and pci_dev are no longer valid!! */
++
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_manage_pci_remove,
++ &manage_pci));
++ return r;
++}
++
++static int __init hook_pci_bus(void)
++{
++ if (!is_running_on_xen() || !is_initial_xendomain())
++ return 0;
++
++ pci_bus_probe = pci_bus_type.probe;
++ pci_bus_type.probe = pci_bus_probe_wrapper;
++
++ pci_bus_remove = pci_bus_type.remove;
++ pci_bus_type.remove = pci_bus_remove_wrapper;
++
++ return 0;
++}
++
++core_initcall(hook_pci_bus);
+diff --git a/drivers/xen/core/reboot.c b/drivers/xen/core/reboot.c
+new file mode 100644
+index 0000000..d4ffdaf
+--- /dev/null
++++ b/drivers/xen/core/reboot.c
+@@ -0,0 +1,337 @@
++#define __KERNEL_SYSCALLS__
++#include <linux/version.h>
++#include <linux/kernel.h>
++#include <linux/unistd.h>
++#include <linux/module.h>
++#include <linux/reboot.h>
++#include <linux/sched.h>
++#include <linux/sysrq.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++#include <linux/kmod.h>
++#include <linux/slab.h>
++#include <linux/workqueue.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#undef handle_sysrq
++#endif
++
++MODULE_LICENSE("Dual BSD/GPL");
++
++#define SHUTDOWN_INVALID -1
++#define SHUTDOWN_POWEROFF 0
++#define SHUTDOWN_SUSPEND 2
++#define SHUTDOWN_RESUMING 3
++#define SHUTDOWN_HALT 4
++
++/* Ignore multiple shutdown requests. */
++static int shutting_down = SHUTDOWN_INVALID;
++
++/* Was last suspend request cancelled? */
++static int suspend_cancelled;
++
++/* Can we leave APs online when we suspend? */
++static int fast_suspend;
++
++static void __shutdown_handler(struct work_struct *unused);
++static DECLARE_DELAYED_WORK(shutdown_work, __shutdown_handler);
++
++static int setup_suspend_evtchn(void);
++
++int __xen_suspend(int fast_suspend, void (*resume_notifier)(int));
++
++static int shutdown_process(void *__unused)
++{
++ static char *envp[] = { "HOME=/", "TERM=linux",
++ "PATH=/sbin:/usr/sbin:/bin:/usr/bin", NULL };
++ static char *poweroff_argv[] = { "/sbin/poweroff", NULL };
++
++ extern asmlinkage long sys_reboot(int magic1, int magic2,
++ unsigned int cmd, void *arg);
++
++ if ((shutting_down == SHUTDOWN_POWEROFF) ||
++ (shutting_down == SHUTDOWN_HALT)) {
++ if (call_usermodehelper("/sbin/poweroff", poweroff_argv,
++ envp, 0) < 0) {
++#ifdef CONFIG_XEN
++ sys_reboot(LINUX_REBOOT_MAGIC1,
++ LINUX_REBOOT_MAGIC2,
++ LINUX_REBOOT_CMD_POWER_OFF,
++ NULL);
++#endif /* CONFIG_XEN */
++ }
++ }
++
++ shutting_down = SHUTDOWN_INVALID; /* could try again */
++
++ return 0;
++}
++
++static void xen_resume_notifier(int _suspend_cancelled)
++{
++ int old_state = xchg(&shutting_down, SHUTDOWN_RESUMING);
++ BUG_ON(old_state != SHUTDOWN_SUSPEND);
++ suspend_cancelled = _suspend_cancelled;
++}
++
++static int xen_suspend(void *__unused)
++{
++ int err, old_state;
++
++ daemonize("suspend");
++ err = set_cpus_allowed(current, cpumask_of_cpu(0));
++ if (err) {
++ printk(KERN_ERR "Xen suspend can't run on CPU0 (%d)\n", err);
++ goto fail;
++ }
++
++ do {
++ err = __xen_suspend(fast_suspend, xen_resume_notifier);
++ if (err) {
++ printk(KERN_ERR "Xen suspend failed (%d)\n", err);
++ goto fail;
++ }
++ if (!suspend_cancelled)
++ setup_suspend_evtchn();
++ old_state = cmpxchg(
++ &shutting_down, SHUTDOWN_RESUMING, SHUTDOWN_INVALID);
++ } while (old_state == SHUTDOWN_SUSPEND);
++
++ switch (old_state) {
++ case SHUTDOWN_INVALID:
++ case SHUTDOWN_SUSPEND:
++ BUG();
++ case SHUTDOWN_RESUMING:
++ break;
++ default:
++ schedule_delayed_work(&shutdown_work, 0);
++ break;
++ }
++
++ return 0;
++
++ fail:
++ old_state = xchg(&shutting_down, SHUTDOWN_INVALID);
++ BUG_ON(old_state != SHUTDOWN_SUSPEND);
++ return 0;
++}
++
++static void switch_shutdown_state(int new_state)
++{
++ int prev_state, old_state = SHUTDOWN_INVALID;
++
++ /* We only drive shutdown_state into an active state. */
++ if (new_state == SHUTDOWN_INVALID)
++ return;
++
++ do {
++ /* We drop this transition if already in an active state. */
++ if ((old_state != SHUTDOWN_INVALID) &&
++ (old_state != SHUTDOWN_RESUMING))
++ return;
++ /* Attempt to transition. */
++ prev_state = old_state;
++ old_state = cmpxchg(&shutting_down, old_state, new_state);
++ } while (old_state != prev_state);
++
++ /* Either we kick off the work, or we leave it to xen_suspend(). */
++ if (old_state == SHUTDOWN_INVALID)
++ schedule_delayed_work(&shutdown_work, 0);
++ else
++ BUG_ON(old_state != SHUTDOWN_RESUMING);
++}
++
++static void __shutdown_handler(struct work_struct *unused)
++{
++ int err;
++
++ err = kernel_thread((shutting_down == SHUTDOWN_SUSPEND) ?
++ xen_suspend : shutdown_process,
++ NULL, CLONE_FS | CLONE_FILES);
++
++ if (err < 0) {
++ printk(KERN_WARNING "Error creating shutdown process (%d): "
++ "retrying...\n", -err);
++ schedule_delayed_work(&shutdown_work, HZ/2);
++ }
++}
++
++static void shutdown_handler(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ extern void ctrl_alt_del(void);
++ char *str;
++ struct xenbus_transaction xbt;
++ int err, new_state = SHUTDOWN_INVALID;
++
++ if ((shutting_down != SHUTDOWN_INVALID) &&
++ (shutting_down != SHUTDOWN_RESUMING))
++ return;
++
++ again:
++ err = xenbus_transaction_start(&xbt);
++ if (err)
++ return;
++
++ str = (char *)xenbus_read(xbt, "control", "shutdown", NULL);
++ /* Ignore read errors and empty reads. */
++ if (XENBUS_IS_ERR_READ(str)) {
++ xenbus_transaction_end(xbt, 1);
++ return;
++ }
++
++ xenbus_write(xbt, "control", "shutdown", "");
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN) {
++ kfree(str);
++ goto again;
++ }
++
++ if (strcmp(str, "poweroff") == 0)
++ new_state = SHUTDOWN_POWEROFF;
++ else if (strcmp(str, "reboot") == 0)
++ ctrl_alt_del();
++ else if (strcmp(str, "suspend") == 0)
++ new_state = SHUTDOWN_SUSPEND;
++ else if (strcmp(str, "halt") == 0)
++ new_state = SHUTDOWN_HALT;
++ else
++ printk("Ignoring shutdown request: %s\n", str);
++
++ switch_shutdown_state(new_state);
++
++ kfree(str);
++}
++
++static void sysrq_handler(struct xenbus_watch *watch, const char **vec,
++ unsigned int len)
++{
++ char sysrq_key = '\0';
++ struct xenbus_transaction xbt;
++ int err;
++
++ again:
++ err = xenbus_transaction_start(&xbt);
++ if (err)
++ return;
++ if (!xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key)) {
++ printk(KERN_ERR "Unable to read sysrq code in "
++ "control/sysrq\n");
++ xenbus_transaction_end(xbt, 1);
++ return;
++ }
++
++ if (sysrq_key != '\0')
++ xenbus_printf(xbt, "control", "sysrq", "%c", '\0');
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++
++#ifdef CONFIG_MAGIC_SYSRQ
++ if (sysrq_key != '\0')
++ handle_sysrq(sysrq_key, NULL);
++#endif
++}
++
++static struct xenbus_watch shutdown_watch = {
++ .node = "control/shutdown",
++ .callback = shutdown_handler
++};
++
++static struct xenbus_watch sysrq_watch = {
++ .node = "control/sysrq",
++ .callback = sysrq_handler
++};
++
++static irqreturn_t suspend_int(int irq, void* dev_id)
++{
++ switch_shutdown_state(SHUTDOWN_SUSPEND);
++ return IRQ_HANDLED;
++}
++
++static int setup_suspend_evtchn(void)
++{
++ static int irq;
++ int port;
++ char portstr[16];
++
++ if (irq > 0)
++ unbind_from_irqhandler(irq, NULL);
++
++ irq = bind_listening_port_to_irqhandler(0, suspend_int, 0, "suspend",
++ NULL);
++ if (irq <= 0)
++ return -1;
++
++ port = irq_to_evtchn_port(irq);
++ printk(KERN_INFO "suspend: event channel %d\n", port);
++ sprintf(portstr, "%d", port);
++ xenbus_write(XBT_NIL, "device/suspend", "event-channel", portstr);
++
++ return 0;
++}
++
++static int setup_shutdown_watcher(void)
++{
++ int err;
++
++ xenbus_scanf(XBT_NIL, "control",
++ "platform-feature-multiprocessor-suspend",
++ "%d", &fast_suspend);
++
++ err = register_xenbus_watch(&shutdown_watch);
++ if (err) {
++ printk(KERN_ERR "Failed to set shutdown watcher\n");
++ return err;
++ }
++
++ err = register_xenbus_watch(&sysrq_watch);
++ if (err) {
++ printk(KERN_ERR "Failed to set sysrq watcher\n");
++ return err;
++ }
++
++ /* suspend event channel */
++ err = setup_suspend_evtchn();
++ if (err) {
++ printk(KERN_ERR "Failed to register suspend event channel\n");
++ return err;
++ }
++
++ return 0;
++}
++
++#ifdef CONFIG_XEN
++
++static int shutdown_event(struct notifier_block *notifier,
++ unsigned long event,
++ void *data)
++{
++ setup_shutdown_watcher();
++ return NOTIFY_DONE;
++}
++
++static int __init setup_shutdown_event(void)
++{
++ static struct notifier_block xenstore_notifier = {
++ .notifier_call = shutdown_event
++ };
++ register_xenstore_notifier(&xenstore_notifier);
++
++ return 0;
++}
++
++subsys_initcall(setup_shutdown_event);
++
++#else /* !defined(CONFIG_XEN) */
++
++int xen_reboot_init(void)
++{
++ return setup_shutdown_watcher();
++}
++
++#endif /* !defined(CONFIG_XEN) */
+diff --git a/drivers/xen/core/smpboot.c b/drivers/xen/core/smpboot.c
+new file mode 100644
+index 0000000..7181188
+--- /dev/null
++++ b/drivers/xen/core/smpboot.c
+@@ -0,0 +1,451 @@
++/*
++ * Xen SMP booting functions
++ *
++ * See arch/i386/kernel/smpboot.c for copyright and credits for derived
++ * portions of this file.
++ */
++
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/kernel.h>
++#include <linux/mm.h>
++#include <linux/sched.h>
++#include <linux/kernel_stat.h>
++#include <linux/smp_lock.h>
++#include <linux/irq.h>
++#include <linux/bootmem.h>
++#include <linux/notifier.h>
++#include <linux/cpu.h>
++#include <linux/percpu.h>
++#include <asm/desc.h>
++#include <asm/arch_hooks.h>
++#include <asm/pgalloc.h>
++#include <xen/evtchn.h>
++#include <xen/interface/vcpu.h>
++#include <xen/cpu_hotplug.h>
++#include <xen/xenbus.h>
++
++extern irqreturn_t smp_reschedule_interrupt(int, void *);
++extern irqreturn_t smp_call_function_interrupt(int, void *);
++
++extern int local_setup_timer(unsigned int cpu);
++extern void local_teardown_timer(unsigned int cpu);
++
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void system_call(void);
++extern void smp_trap_init(trap_info_t *);
++
++/* Number of siblings per CPU package */
++int smp_num_siblings = 1;
++
++cpumask_t cpu_online_map;
++EXPORT_SYMBOL(cpu_online_map);
++cpumask_t cpu_possible_map;
++EXPORT_SYMBOL(cpu_possible_map);
++cpumask_t cpu_initialized_map;
++
++DEFINE_PER_CPU(struct cpuinfo_x86, cpu_info);
++EXPORT_PER_CPU_SYMBOL(cpu_info);
++
++#ifdef CONFIG_HOTPLUG_CPU
++DEFINE_PER_CPU(int, cpu_state) = { 0 };
++#endif
++
++static DEFINE_PER_CPU(int, resched_irq);
++static DEFINE_PER_CPU(int, callfunc_irq);
++static char resched_name[NR_CPUS][15];
++static char callfunc_name[NR_CPUS][15];
++
++#ifdef CONFIG_X86_LOCAL_APIC
++#define set_cpu_to_apicid(cpu, apicid) (per_cpu(x86_cpu_to_apicid, cpu) = (apicid))
++#else
++#define set_cpu_to_apicid(cpu, apicid)
++#endif
++
++DEFINE_PER_CPU(cpumask_t, cpu_sibling_map);
++DEFINE_PER_CPU(cpumask_t, cpu_core_map);
++EXPORT_PER_CPU_SYMBOL(cpu_core_map);
++
++void __init prefill_possible_map(void)
++{
++ int i, rc;
++
++ for_each_possible_cpu(i)
++ if (i != smp_processor_id())
++ return;
++
++ for (i = 0; i < NR_CPUS; i++) {
++ rc = HYPERVISOR_vcpu_op(VCPUOP_is_up, i, NULL);
++ if (rc >= 0)
++ cpu_set(i, cpu_possible_map);
++ }
++}
++
++void __init smp_alloc_memory(void)
++{
++}
++
++static inline void
++set_cpu_sibling_map(unsigned int cpu)
++{
++ cpu_data(cpu).phys_proc_id = cpu;
++ cpu_data(cpu).cpu_core_id = 0;
++
++ per_cpu(cpu_sibling_map, cpu) = cpumask_of_cpu(cpu);
++ per_cpu(cpu_core_map, cpu) = cpumask_of_cpu(cpu);
++
++ cpu_data(cpu).booted_cores = 1;
++}
++
++static void
++remove_siblinginfo(unsigned int cpu)
++{
++ cpu_data(cpu).phys_proc_id = BAD_APICID;
++ cpu_data(cpu).cpu_core_id = BAD_APICID;
++
++ cpus_clear(per_cpu(cpu_sibling_map, cpu));
++ cpus_clear(per_cpu(cpu_core_map, cpu));
++
++ cpu_data(cpu).booted_cores = 0;
++}
++
++static int __cpuinit xen_smp_intr_init(unsigned int cpu)
++{
++ int rc;
++
++ per_cpu(resched_irq, cpu) = per_cpu(callfunc_irq, cpu) = -1;
++
++ sprintf(resched_name[cpu], "resched%u", cpu);
++ rc = bind_ipi_to_irqhandler(RESCHEDULE_VECTOR,
++ cpu,
++ smp_reschedule_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ resched_name[cpu],
++ NULL);
++ if (rc < 0)
++ goto fail;
++ per_cpu(resched_irq, cpu) = rc;
++
++ sprintf(callfunc_name[cpu], "callfunc%u", cpu);
++ rc = bind_ipi_to_irqhandler(CALL_FUNCTION_VECTOR,
++ cpu,
++ smp_call_function_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ callfunc_name[cpu],
++ NULL);
++ if (rc < 0)
++ goto fail;
++ per_cpu(callfunc_irq, cpu) = rc;
++
++ rc = xen_spinlock_init(cpu);
++ if (rc < 0)
++ goto fail;
++
++ if ((cpu != 0) && ((rc = local_setup_timer(cpu)) != 0))
++ goto fail;
++
++ return 0;
++
++ fail:
++ if (per_cpu(resched_irq, cpu) >= 0)
++ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
++ if (per_cpu(callfunc_irq, cpu) >= 0)
++ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
++ xen_spinlock_cleanup(cpu);
++ return rc;
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++static void __cpuexit xen_smp_intr_exit(unsigned int cpu)
++{
++ if (cpu != 0)
++ local_teardown_timer(cpu);
++
++ unbind_from_irqhandler(per_cpu(resched_irq, cpu), NULL);
++ unbind_from_irqhandler(per_cpu(callfunc_irq, cpu), NULL);
++ xen_spinlock_cleanup(cpu);
++}
++#endif
++
++void __cpuinit cpu_bringup(void)
++{
++ cpu_init();
++#ifdef __i386__
++ identify_secondary_cpu(¤t_cpu_data);
++#else
++ identify_cpu(¤t_cpu_data);
++#endif
++ touch_softlockup_watchdog();
++ preempt_disable();
++ local_irq_enable();
++}
++
++static void __cpuinit cpu_bringup_and_idle(void)
++{
++ cpu_bringup();
++ cpu_idle();
++}
++
++static void __cpuinit cpu_initialize_context(unsigned int cpu)
++{
++ /* vcpu_guest_context_t is too large to allocate on the stack.
++ * Hence we allocate statically and protect it with a lock */
++ static vcpu_guest_context_t ctxt;
++ static DEFINE_SPINLOCK(ctxt_lock);
++
++ struct task_struct *idle = idle_task(cpu);
++
++ if (cpu_test_and_set(cpu, cpu_initialized_map))
++ return;
++
++ spin_lock(&ctxt_lock);
++
++ memset(&ctxt, 0, sizeof(ctxt));
++
++ ctxt.flags = VGCF_IN_KERNEL;
++ ctxt.user_regs.ds = __USER_DS;
++ ctxt.user_regs.es = __USER_DS;
++ ctxt.user_regs.fs = 0;
++ ctxt.user_regs.gs = 0;
++ ctxt.user_regs.ss = __KERNEL_DS;
++ ctxt.user_regs.eip = (unsigned long)cpu_bringup_and_idle;
++ ctxt.user_regs.eflags = X86_EFLAGS_IF | 0x1000; /* IOPL_RING1 */
++
++ memset(&ctxt.fpu_ctxt, 0, sizeof(ctxt.fpu_ctxt));
++
++ smp_trap_init(ctxt.trap_ctxt);
++
++ ctxt.ldt_ents = 0;
++ ctxt.gdt_frames[0] = virt_to_mfn(get_cpu_gdt_table(cpu));
++ ctxt.gdt_ents = GDT_SIZE / 8;
++
++ ctxt.user_regs.cs = __KERNEL_CS;
++ ctxt.user_regs.esp = idle->thread.sp0 - sizeof(struct pt_regs);
++
++ ctxt.kernel_ss = __KERNEL_DS;
++ ctxt.kernel_sp = idle->thread.sp0;
++
++ ctxt.event_callback_eip = (unsigned long)hypervisor_callback;
++ ctxt.failsafe_callback_eip = (unsigned long)failsafe_callback;
++#ifdef __i386__
++ ctxt.event_callback_cs = __KERNEL_CS;
++ ctxt.failsafe_callback_cs = __KERNEL_CS;
++
++ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(swapper_pg_dir));
++
++ ctxt.user_regs.fs = __KERNEL_PERCPU;
++#else /* __x86_64__ */
++ ctxt.syscall_callback_eip = (unsigned long)system_call;
++
++ ctxt.ctrlreg[3] = xen_pfn_to_cr3(virt_to_mfn(init_level4_pgt));
++
++ ctxt.gs_base_kernel = (unsigned long)(cpu_pda(cpu));
++#endif
++
++ if (HYPERVISOR_vcpu_op(VCPUOP_initialise, cpu, &ctxt))
++ BUG();
++
++ spin_unlock(&ctxt_lock);
++}
++
++void __init smp_prepare_cpus(unsigned int max_cpus)
++{
++ unsigned int cpu;
++ struct task_struct *idle;
++ int apicid;
++ struct vcpu_get_physid cpu_id;
++#ifdef __x86_64__
++ struct desc_ptr *gdt_descr;
++#endif
++ void *gdt_addr;
++
++ apicid = 0;
++ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, 0, &cpu_id) == 0)
++ apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
++ boot_cpu_data.apicid = apicid;
++ cpu_data(0) = boot_cpu_data;
++
++ set_cpu_to_apicid(0, apicid);
++
++ current_thread_info()->cpu = 0;
++
++ for (cpu = 0; cpu < NR_CPUS; cpu++) {
++ cpus_clear(per_cpu(cpu_sibling_map, cpu));
++ cpus_clear(per_cpu(cpu_core_map, cpu));
++ }
++
++ set_cpu_sibling_map(0);
++
++ if (xen_smp_intr_init(0))
++ BUG();
++
++ cpu_initialized_map = cpumask_of_cpu(0);
++
++ /* Restrict the possible_map according to max_cpus. */
++ while ((num_possible_cpus() > 1) && (num_possible_cpus() > max_cpus)) {
++ for (cpu = NR_CPUS-1; !cpu_isset(cpu, cpu_possible_map); cpu--)
++ continue;
++ cpu_clear(cpu, cpu_possible_map);
++ }
++
++ for_each_possible_cpu (cpu) {
++ if (cpu == 0)
++ continue;
++
++ idle = fork_idle(cpu);
++ if (IS_ERR(idle))
++ panic("failed fork for CPU %d", cpu);
++
++#ifdef __x86_64__
++ gdt_descr = &cpu_gdt_descr[cpu];
++ gdt_descr->address = get_zeroed_page(GFP_KERNEL);
++ if (unlikely(!gdt_descr->address)) {
++ printk(KERN_CRIT "CPU%d failed to allocate GDT\n",
++ cpu);
++ continue;
++ }
++ gdt_descr->size = GDT_SIZE;
++ memcpy((void *)gdt_descr->address, cpu_gdt_table, GDT_SIZE);
++ gdt_addr = (void *)gdt_descr->address;
++#else
++ init_gdt(cpu);
++ gdt_addr = get_cpu_gdt_table(cpu);
++#endif
++ make_page_readonly(gdt_addr, XENFEAT_writable_descriptor_tables);
++
++ apicid = cpu;
++ if (HYPERVISOR_vcpu_op(VCPUOP_get_physid, cpu, &cpu_id) == 0)
++ apicid = xen_vcpu_physid_to_x86_apicid(cpu_id.phys_id);
++ cpu_data(cpu) = boot_cpu_data;
++ cpu_data(cpu).cpu_index = cpu;
++ cpu_data(cpu).apicid = apicid;
++
++ set_cpu_to_apicid(cpu, apicid);
++
++#ifdef __x86_64__
++ cpu_pda(cpu)->pcurrent = idle;
++ cpu_pda(cpu)->cpunumber = cpu;
++ clear_ti_thread_flag(task_thread_info(idle), TIF_FORK);
++#else
++ per_cpu(current_task, cpu) = idle;
++#endif
++
++ irq_ctx_init(cpu);
++
++#ifdef CONFIG_HOTPLUG_CPU
++ if (is_initial_xendomain())
++ cpu_set(cpu, cpu_present_map);
++#else
++ cpu_set(cpu, cpu_present_map);
++#endif
++ }
++
++ 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:
++ */
++ if (!skip_ioapic_setup && nr_ioapics)
++ setup_IO_APIC();
++#endif
++}
++
++void __init smp_prepare_boot_cpu(void)
++{
++#ifdef __i386__
++ init_gdt(smp_processor_id());
++ switch_to_new_gdt();
++#endif
++ prefill_possible_map();
++}
++
++#ifdef CONFIG_HOTPLUG_CPU
++
++/*
++ * Initialize cpu_present_map late to skip SMP boot code in init/main.c.
++ * But do it early enough to catch critical for_each_present_cpu() loops
++ * in i386-specific code.
++ */
++static int __init initialize_cpu_present_map(void)
++{
++ cpu_present_map = cpu_possible_map;
++ return 0;
++}
++core_initcall(initialize_cpu_present_map);
++
++int __cpuexit __cpu_disable(void)
++{
++ cpumask_t map = cpu_online_map;
++ unsigned int cpu = smp_processor_id();
++
++ if (cpu == 0)
++ return -EBUSY;
++
++ remove_siblinginfo(cpu);
++
++ cpu_clear(cpu, map);
++ fixup_irqs(map);
++ cpu_clear(cpu, cpu_online_map);
++
++ return 0;
++}
++
++void __cpuexit __cpu_die(unsigned int cpu)
++{
++ while (HYPERVISOR_vcpu_op(VCPUOP_is_up, cpu, NULL)) {
++ current->state = TASK_UNINTERRUPTIBLE;
++ schedule_timeout(HZ/10);
++ }
++
++ xen_smp_intr_exit(cpu);
++
++ if (num_online_cpus() == 1)
++ alternatives_smp_switch(0);
++}
++
++#endif /* CONFIG_HOTPLUG_CPU */
++
++int __cpuinit __cpu_up(unsigned int cpu)
++{
++ int rc;
++
++ rc = cpu_up_check(cpu);
++ if (rc)
++ return rc;
++
++ cpu_initialize_context(cpu);
++
++ if (num_online_cpus() == 1)
++ alternatives_smp_switch(1);
++
++ /* This must be done before setting cpu_online_map */
++ set_cpu_sibling_map(cpu);
++ wmb();
++
++ rc = xen_smp_intr_init(cpu);
++ if (rc) {
++ remove_siblinginfo(cpu);
++ return rc;
++ }
++
++ cpu_set(cpu, cpu_online_map);
++
++ rc = HYPERVISOR_vcpu_op(VCPUOP_up, cpu, NULL);
++ BUG_ON(rc);
++
++ return 0;
++}
++
++void __init smp_cpus_done(unsigned int max_cpus)
++{
++}
++
++#ifndef CONFIG_X86_LOCAL_APIC
++int setup_profiling_timer(unsigned int multiplier)
++{
++ return -EINVAL;
++}
++#endif
+diff --git a/drivers/xen/core/spinlock.c b/drivers/xen/core/spinlock.c
+new file mode 100644
+index 0000000..27b6e8d
+--- /dev/null
++++ b/drivers/xen/core/spinlock.c
+@@ -0,0 +1,149 @@
++/*
++ * Xen spinlock functions
++ *
++ * See arch/x86/xen/smp.c for copyright and credits for derived
++ * portions of this file.
++ */
++
++#include <linux/init.h>
++#include <linux/irq.h>
++#include <linux/kernel.h>
++#include <linux/kernel_stat.h>
++#include <linux/module.h>
++#include <xen/evtchn.h>
++
++extern irqreturn_t smp_reschedule_interrupt(int, void *);
++
++static DEFINE_PER_CPU(int, spinlock_irq) = -1;
++static char spinlock_name[NR_CPUS][15];
++
++struct spinning {
++ raw_spinlock_t *lock;
++ unsigned int ticket;
++};
++static DEFINE_PER_CPU(struct spinning, spinning);
++static DEFINE_PER_CPU(struct spinning, spinning_bh);
++static DEFINE_PER_CPU(struct spinning, spinning_irq);
++
++int __cpuinit xen_spinlock_init(unsigned int cpu)
++{
++ int rc;
++
++ sprintf(spinlock_name[cpu], "spinlock%u", cpu);
++ rc = bind_ipi_to_irqhandler(SPIN_UNLOCK_VECTOR,
++ cpu,
++ smp_reschedule_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ spinlock_name[cpu],
++ NULL);
++ if (rc < 0)
++ return rc;
++
++ disable_irq(rc); /* make sure it's never delivered */
++ per_cpu(spinlock_irq, cpu) = rc;
++
++ return 0;
++}
++
++void __cpuinit xen_spinlock_cleanup(unsigned int cpu)
++{
++ if (per_cpu(spinlock_irq, cpu) >= 0)
++ unbind_from_irqhandler(per_cpu(spinlock_irq, cpu), NULL);
++ per_cpu(spinlock_irq, cpu) = -1;
++}
++
++int xen_spin_wait(raw_spinlock_t *lock, unsigned int token)
++{
++ struct spinning *spinning;
++ int rc = 0, irq = __get_cpu_var(spinlock_irq);
++
++ /* If kicker interrupt not initialized yet, just spin. */
++ if (irq < 0)
++ return 0;
++
++ token >>= TICKET_SHIFT;
++
++ /* announce we're spinning */
++ spinning = &__get_cpu_var(spinning);
++ if (spinning->lock) {
++ BUG_ON(spinning->lock == lock);
++ if(raw_irqs_disabled()) {
++ BUG_ON(__get_cpu_var(spinning_bh).lock == lock);
++ spinning = &__get_cpu_var(spinning_irq);
++ } else {
++ BUG_ON(!in_softirq());
++ spinning = &__get_cpu_var(spinning_bh);
++ }
++ BUG_ON(spinning->lock);
++ }
++ spinning->ticket = token;
++ smp_wmb();
++ spinning->lock = lock;
++
++ /* clear pending */
++ xen_clear_irq_pending(irq);
++
++ do {
++ /* Check again to make sure it didn't become free while
++ * we weren't looking. */
++ if ((lock->slock & ((1U << TICKET_SHIFT) - 1)) == token) {
++ /* If we interrupted another spinlock while it was
++ * blocking, make sure it doesn't block (again)
++ * without rechecking the lock. */
++ if (spinning != &__get_cpu_var(spinning))
++ xen_set_irq_pending(irq);
++ rc = 1;
++ break;
++ }
++
++ /* block until irq becomes pending */
++ xen_poll_irq(irq);
++ } while (!xen_test_irq_pending(irq));
++
++ /* Leave the irq pending so that any interrupted blocker will
++ * re-check. */
++ kstat_this_cpu.irqs[irq] += !rc;
++
++ /* announce we're done */
++ spinning->lock = NULL;
++
++ return rc;
++}
++EXPORT_SYMBOL(xen_spin_wait);
++
++int xen_spin_wait_flags(raw_spinlock_t *lock, unsigned int *token,
++ unsigned int flags)
++{
++ return xen_spin_wait(lock, *token);//todo
++}
++EXPORT_SYMBOL(xen_spin_wait_flags);
++
++static inline int spinning(const struct spinning *spinning, unsigned int cpu,
++ raw_spinlock_t *lock, unsigned int ticket)
++{
++ if (spinning->lock != lock)
++ return 0;
++ smp_rmb();
++ if (spinning->ticket != ticket)
++ return 0;
++ notify_remote_via_irq(per_cpu(spinlock_irq, cpu));
++ return 1;
++}
++
++void xen_spin_kick(raw_spinlock_t *lock, unsigned int token)
++{
++ unsigned int cpu;
++
++ token &= (1U << TICKET_SHIFT) - 1;
++ for_each_online_cpu(cpu) {
++ if (spinning(&per_cpu(spinning, cpu), cpu, lock, token))
++ return;
++ if (in_interrupt()
++ && spinning(&per_cpu(spinning_bh, cpu), cpu, lock, token))
++ return;
++ if (raw_irqs_disabled()
++ && spinning(&per_cpu(spinning_irq, cpu), cpu, lock, token))
++ return;
++ }
++}
++EXPORT_SYMBOL(xen_spin_kick);
+diff --git a/drivers/xen/core/xen_proc.c b/drivers/xen/core/xen_proc.c
+new file mode 100644
+index 0000000..4f7c1be
+--- /dev/null
++++ b/drivers/xen/core/xen_proc.c
+@@ -0,0 +1,23 @@
++
++#include <linux/module.h>
++#include <linux/proc_fs.h>
++#include <xen/xen_proc.h>
++
++static struct proc_dir_entry *xen_base;
++
++struct proc_dir_entry *create_xen_proc_entry(const char *name, mode_t mode)
++{
++ if ( xen_base == NULL )
++ if ( (xen_base = proc_mkdir("xen", NULL)) == NULL )
++ panic("Couldn't create /proc/xen");
++ return create_proc_entry(name, mode, xen_base);
++}
++
++EXPORT_SYMBOL_GPL(create_xen_proc_entry);
++
++void remove_xen_proc_entry(const char *name)
++{
++ remove_proc_entry(name, xen_base);
++}
++
++EXPORT_SYMBOL_GPL(remove_xen_proc_entry);
+diff --git a/drivers/xen/core/xen_sysfs.c b/drivers/xen/core/xen_sysfs.c
+new file mode 100644
+index 0000000..9e19882
+--- /dev/null
++++ b/drivers/xen/core/xen_sysfs.c
+@@ -0,0 +1,412 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
++
++#include <linux/err.h>
++#include <linux/kernel.h>
++#include <linux/module.h>
++#include <linux/init.h>
++#include <asm/hypervisor.h>
++#include <xen/features.h>
++#include <xen/hypervisor_sysfs.h>
++#include <xen/xenbus.h>
++#include <xen/interface/kexec.h>
++
++MODULE_LICENSE("GPL");
++MODULE_AUTHOR("Mike D. Day <ncmike at us.ibm.com>");
++
++static ssize_t type_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ return sprintf(buffer, "xen\n");
++}
++
++HYPERVISOR_ATTR_RO(type);
++
++static int __init xen_sysfs_type_init(void)
++{
++ return sysfs_create_file(hypervisor_kobj, &type_attr.attr);
++}
++
++static void xen_sysfs_type_destroy(void)
++{
++ sysfs_remove_file(hypervisor_kobj, &type_attr.attr);
++}
++
++/* xen version attributes */
++static ssize_t major_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
++ if (version)
++ return sprintf(buffer, "%d\n", version >> 16);
++ return -ENODEV;
++}
++
++HYPERVISOR_ATTR_RO(major);
++
++static ssize_t minor_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int version = HYPERVISOR_xen_version(XENVER_version, NULL);
++ if (version)
++ return sprintf(buffer, "%d\n", version & 0xff);
++ return -ENODEV;
++}
++
++HYPERVISOR_ATTR_RO(minor);
++
++static ssize_t extra_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ char *extra;
++
++ extra = kmalloc(XEN_EXTRAVERSION_LEN, GFP_KERNEL);
++ if (extra) {
++ ret = HYPERVISOR_xen_version(XENVER_extraversion, extra);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", extra);
++ kfree(extra);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(extra);
++
++static struct attribute *version_attrs[] = {
++ &major_attr.attr,
++ &minor_attr.attr,
++ &extra_attr.attr,
++ NULL
++};
++
++static struct attribute_group version_group = {
++ .name = "version",
++ .attrs = version_attrs,
++};
++
++static int __init xen_sysfs_version_init(void)
++{
++ return sysfs_create_group(hypervisor_kobj, &version_group);
++}
++
++static void xen_sysfs_version_destroy(void)
++{
++ sysfs_remove_group(hypervisor_kobj, &version_group);
++}
++
++/* UUID */
++
++static ssize_t uuid_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ char *vm, *val;
++ int ret;
++ extern int xenstored_ready;
++
++ if (!xenstored_ready)
++ return -EBUSY;
++
++ vm = xenbus_read(XBT_NIL, "vm", "", NULL);
++ if (IS_ERR(vm))
++ return PTR_ERR(vm);
++ val = xenbus_read(XBT_NIL, vm, "uuid", NULL);
++ kfree(vm);
++ if (IS_ERR(val))
++ return PTR_ERR(val);
++ ret = sprintf(buffer, "%s\n", val);
++ kfree(val);
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(uuid);
++
++static int __init xen_sysfs_uuid_init(void)
++{
++ return sysfs_create_file(hypervisor_kobj, &uuid_attr.attr);
++}
++
++static void xen_sysfs_uuid_destroy(void)
++{
++ sysfs_remove_file(hypervisor_kobj, &uuid_attr.attr);
++}
++
++/* xen compilation attributes */
++
++static ssize_t compiler_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
++
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compiler);
++ kfree(info);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(compiler);
++
++static ssize_t compiled_by_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
++
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compile_by);
++ kfree(info);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(compiled_by);
++
++static ssize_t compile_date_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_compile_info *info;
++
++ info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL);
++ if (info) {
++ ret = HYPERVISOR_xen_version(XENVER_compile_info, info);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", info->compile_date);
++ kfree(info);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(compile_date);
++
++static struct attribute *xen_compile_attrs[] = {
++ &compiler_attr.attr,
++ &compiled_by_attr.attr,
++ &compile_date_attr.attr,
++ NULL
++};
++
++static struct attribute_group xen_compilation_group = {
++ .name = "compilation",
++ .attrs = xen_compile_attrs,
++};
++
++int __init static xen_compilation_init(void)
++{
++ return sysfs_create_group(hypervisor_kobj, &xen_compilation_group);
++}
++
++static void xen_compilation_destroy(void)
++{
++ sysfs_remove_group(hypervisor_kobj, &xen_compilation_group);
++}
++
++/* xen properties info */
++
++static ssize_t capabilities_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ char *caps;
++
++ caps = kmalloc(XEN_CAPABILITIES_INFO_LEN, GFP_KERNEL);
++ if (caps) {
++ ret = HYPERVISOR_xen_version(XENVER_capabilities, caps);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", caps);
++ kfree(caps);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(capabilities);
++
++static ssize_t changeset_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ char *cset;
++
++ cset = kmalloc(XEN_CHANGESET_INFO_LEN, GFP_KERNEL);
++ if (cset) {
++ ret = HYPERVISOR_xen_version(XENVER_changeset, cset);
++ if (!ret)
++ ret = sprintf(buffer, "%s\n", cset);
++ kfree(cset);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(changeset);
++
++static ssize_t virtual_start_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_platform_parameters *parms;
++
++ parms = kmalloc(sizeof(struct xen_platform_parameters), GFP_KERNEL);
++ if (parms) {
++ ret = HYPERVISOR_xen_version(XENVER_platform_parameters,
++ parms);
++ if (!ret)
++ ret = sprintf(buffer, "%lx\n", parms->virt_start);
++ kfree(parms);
++ }
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(virtual_start);
++
++static ssize_t pagesize_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ int ret;
++
++ ret = HYPERVISOR_xen_version(XENVER_pagesize, NULL);
++ if (ret > 0)
++ ret = sprintf(buffer, "%x\n", ret);
++
++ return ret;
++}
++
++HYPERVISOR_ATTR_RO(pagesize);
++
++/* eventually there will be several more features to export */
++static ssize_t xen_feature_show(int index, char *buffer)
++{
++ int ret = -ENOMEM;
++ struct xen_feature_info *info;
++
++ info = kmalloc(sizeof(struct xen_feature_info), GFP_KERNEL);
++ if (info) {
++ info->submap_idx = index;
++ ret = HYPERVISOR_xen_version(XENVER_get_features, info);
++ if (!ret)
++ ret = sprintf(buffer, "%d\n", info->submap);
++ kfree(info);
++ }
++
++ return ret;
++}
++
++static ssize_t writable_pt_show(struct hyp_sysfs_attr *attr, char *buffer)
++{
++ return xen_feature_show(XENFEAT_writable_page_tables, buffer);
++}
++
++HYPERVISOR_ATTR_RO(writable_pt);
++
++static struct attribute *xen_properties_attrs[] = {
++ &capabilities_attr.attr,
++ &changeset_attr.attr,
++ &virtual_start_attr.attr,
++ &pagesize_attr.attr,
++ &writable_pt_attr.attr,
++ NULL
++};
++
++static struct attribute_group xen_properties_group = {
++ .name = "properties",
++ .attrs = xen_properties_attrs,
++};
++
++static int __init xen_properties_init(void)
++{
++ return sysfs_create_group(hypervisor_kobj, &xen_properties_group);
++}
++
++static void xen_properties_destroy(void)
++{
++ sysfs_remove_group(hypervisor_kobj, &xen_properties_group);
++}
++
++#ifdef CONFIG_KEXEC
++
++static ssize_t vmcoreinfo_show(struct hyp_sysfs_attr *attr, char *page)
++{
++ return sprintf(page, "%lx %zx\n",
++ paddr_vmcoreinfo_xen, vmcoreinfo_size_xen);
++}
++
++HYPERVISOR_ATTR_RO(vmcoreinfo);
++
++static int __init xen_sysfs_vmcoreinfo_init(void)
++{
++ return sysfs_create_file(&hypervisor_subsys.kobj,
++ &vmcoreinfo_attr.attr);
++}
++
++static void xen_sysfs_vmcoreinfo_destroy(void)
++{
++ sysfs_remove_file(&hypervisor_subsys.kobj, &vmcoreinfo_attr.attr);
++}
++
++#else
++
++#define xen_sysfs_vmcoreinfo_init() 0
++#define xen_sysfs_vmcoreinfo_destroy() ((void)0)
++
++#endif
++
++static int __init hyper_sysfs_init(void)
++{
++ int ret;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ ret = xen_sysfs_type_init();
++ if (ret)
++ goto out;
++ ret = xen_sysfs_version_init();
++ if (ret)
++ goto version_out;
++ ret = xen_compilation_init();
++ if (ret)
++ goto comp_out;
++ ret = xen_sysfs_uuid_init();
++ if (ret)
++ goto uuid_out;
++ ret = xen_properties_init();
++ if (ret)
++ goto prop_out;
++ ret = xen_sysfs_vmcoreinfo_init();
++ if (!ret)
++ goto out;
++
++ xen_properties_destroy();
++prop_out:
++ xen_sysfs_uuid_destroy();
++uuid_out:
++ xen_compilation_destroy();
++comp_out:
++ xen_sysfs_version_destroy();
++version_out:
++ xen_sysfs_type_destroy();
++out:
++ return ret;
++}
++
++static void __exit hyper_sysfs_exit(void)
++{
++ xen_sysfs_vmcoreinfo_destroy();
++ xen_properties_destroy();
++ xen_compilation_destroy();
++ xen_sysfs_uuid_destroy();
++ xen_sysfs_version_destroy();
++ xen_sysfs_type_destroy();
++
++}
++
++module_init(hyper_sysfs_init);
++module_exit(hyper_sysfs_exit);
+diff --git a/drivers/xen/core/xencomm.c b/drivers/xen/core/xencomm.c
+new file mode 100644
+index 0000000..e2e35f1
+--- /dev/null
++++ b/drivers/xen/core/xencomm.c
+@@ -0,0 +1,229 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Hollis Blanchard <hollisb at us.ibm.com>
++ */
++
++#include <linux/gfp.h>
++#include <linux/mm.h>
++#include <asm/page.h>
++#include <xen/xencomm.h>
++#include <xen/interface/xen.h>
++#ifdef __ia64__
++#include <asm/xen/xencomm.h> /* for is_kern_addr() */
++#endif
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++static int xencomm_init(struct xencomm_desc *desc,
++ void *buffer, unsigned long bytes)
++{
++ unsigned long recorded = 0;
++ int i = 0;
++
++ while ((recorded < bytes) && (i < desc->nr_addrs)) {
++ unsigned long vaddr = (unsigned long)buffer + recorded;
++ unsigned long paddr;
++ int offset;
++ int chunksz;
++
++ offset = vaddr % PAGE_SIZE; /* handle partial pages */
++ chunksz = min(PAGE_SIZE - offset, bytes - recorded);
++
++ paddr = xencomm_vtop(vaddr);
++ if (paddr == ~0UL) {
++ printk("%s: couldn't translate vaddr %lx\n",
++ __func__, vaddr);
++ return -EINVAL;
++ }
++
++ desc->address[i++] = paddr;
++ recorded += chunksz;
++ }
++
++ if (recorded < bytes) {
++ printk("%s: could only translate %ld of %ld bytes\n",
++ __func__, recorded, bytes);
++ return -ENOSPC;
++ }
++
++ /* mark remaining addresses invalid (just for safety) */
++ while (i < desc->nr_addrs)
++ desc->address[i++] = XENCOMM_INVALID;
++
++ desc->magic = XENCOMM_MAGIC;
++
++ return 0;
++}
++
++static struct xencomm_desc *xencomm_alloc(gfp_t gfp_mask,
++ void *buffer, unsigned long bytes)
++{
++ struct xencomm_desc *desc;
++ unsigned long buffer_ulong = (unsigned long)buffer;
++ unsigned long start = buffer_ulong & PAGE_MASK;
++ unsigned long end = (buffer_ulong + bytes) | ~PAGE_MASK;
++ unsigned long nr_addrs = (end - start + 1) >> PAGE_SHIFT;
++ unsigned long size = sizeof(*desc) +
++ sizeof(desc->address[0]) * nr_addrs;
++
++ /*
++ * slab allocator returns at least sizeof(void*) aligned pointer.
++ * When sizeof(*desc) > sizeof(void*), struct xencomm_desc might
++ * cross page boundary.
++ */
++ if (sizeof(*desc) > sizeof(void*)) {
++ unsigned long order = get_order(size);
++ desc = (struct xencomm_desc *)__get_free_pages(gfp_mask,
++ order);
++ if (desc == NULL)
++ return NULL;
++
++ desc->nr_addrs =
++ ((PAGE_SIZE << order) - sizeof(struct xencomm_desc)) /
++ sizeof(*desc->address);
++ } else {
++ desc = kmalloc(size, gfp_mask);
++ if (desc == NULL)
++ return NULL;
++
++ desc->nr_addrs = nr_addrs;
++ }
++ return desc;
++}
++
++void xencomm_free(struct xencomm_handle *desc)
++{
++ if (desc && !((ulong)desc & XENCOMM_INLINE_FLAG)) {
++ struct xencomm_desc *desc__ = (struct xencomm_desc*)desc;
++ if (sizeof(*desc__) > sizeof(void*)) {
++ unsigned long size = sizeof(*desc__) +
++ sizeof(desc__->address[0]) * desc__->nr_addrs;
++ unsigned long order = get_order(size);
++ free_pages((unsigned long)__va(desc), order);
++ } else
++ kfree(__va(desc));
++ }
++}
++
++static int xencomm_create(void *buffer, unsigned long bytes, struct xencomm_desc **ret, gfp_t gfp_mask)
++{
++ struct xencomm_desc *desc;
++ int rc;
++
++ pr_debug("%s: %p[%ld]\n", __func__, buffer, bytes);
++
++ if (bytes == 0) {
++ /* don't create a descriptor; Xen recognizes NULL. */
++ BUG_ON(buffer != NULL);
++ *ret = NULL;
++ return 0;
++ }
++
++ BUG_ON(buffer == NULL); /* 'bytes' is non-zero */
++
++ desc = xencomm_alloc(gfp_mask, buffer, bytes);
++ if (!desc) {
++ printk("%s failure\n", "xencomm_alloc");
++ return -ENOMEM;
++ }
++
++ rc = xencomm_init(desc, buffer, bytes);
++ if (rc) {
++ printk("%s failure: %d\n", "xencomm_init", rc);
++ xencomm_free((struct xencomm_handle *)__pa(desc));
++ return rc;
++ }
++
++ *ret = desc;
++ return 0;
++}
++
++/* check if memory address is within VMALLOC region */
++static int is_phys_contiguous(unsigned long addr)
++{
++ if (!is_kernel_addr(addr))
++ return 0;
++
++ return (addr < VMALLOC_START) || (addr >= VMALLOC_END);
++}
++
++static struct xencomm_handle *xencomm_create_inline(void *ptr)
++{
++ unsigned long paddr;
++
++ BUG_ON(!is_phys_contiguous((unsigned long)ptr));
++
++ paddr = (unsigned long)xencomm_pa(ptr);
++ BUG_ON(paddr & XENCOMM_INLINE_FLAG);
++ return (struct xencomm_handle *)(paddr | XENCOMM_INLINE_FLAG);
++}
++
++/* "mini" routine, for stack-based communications: */
++static int xencomm_create_mini(void *buffer,
++ unsigned long bytes, struct xencomm_mini *xc_desc,
++ struct xencomm_desc **ret)
++{
++ int rc = 0;
++ struct xencomm_desc *desc;
++ BUG_ON(((unsigned long)xc_desc) % sizeof(*xc_desc) != 0);
++
++ desc = (void *)xc_desc;
++
++ desc->nr_addrs = XENCOMM_MINI_ADDRS;
++
++ if (!(rc = xencomm_init(desc, buffer, bytes)))
++ *ret = desc;
++
++ return rc;
++}
++
++struct xencomm_handle *xencomm_map(void *ptr, unsigned long bytes)
++{
++ int rc;
++ struct xencomm_desc *desc;
++
++ if (is_phys_contiguous((unsigned long)ptr))
++ return xencomm_create_inline(ptr);
++
++ rc = xencomm_create(ptr, bytes, &desc, GFP_KERNEL);
++
++ if (rc || desc == NULL)
++ return NULL;
++
++ return xencomm_pa(desc);
++}
++
++struct xencomm_handle *__xencomm_map_no_alloc(void *ptr, unsigned long bytes,
++ struct xencomm_mini *xc_desc)
++{
++ int rc;
++ struct xencomm_desc *desc = NULL;
++
++ if (is_phys_contiguous((unsigned long)ptr))
++ return xencomm_create_inline(ptr);
++
++ rc = xencomm_create_mini(ptr, bytes, xc_desc,
++ &desc);
++
++ if (rc)
++ return NULL;
++
++ return xencomm_pa(desc);
++}
+diff --git a/drivers/xen/evtchn/Makefile b/drivers/xen/evtchn/Makefile
+new file mode 100644
+index 0000000..7b082a0
+--- /dev/null
++++ b/drivers/xen/evtchn/Makefile
+@@ -0,0 +1,2 @@
++
++obj-y := evtchn.o
+diff --git a/drivers/xen/evtchn/evtchn.c b/drivers/xen/evtchn/evtchn.c
+new file mode 100644
+index 0000000..f814a8f
+--- /dev/null
++++ b/drivers/xen/evtchn/evtchn.c
+@@ -0,0 +1,560 @@
++/******************************************************************************
++ * evtchn.c
++ *
++ * Driver for receiving and demuxing event-channel signals.
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ * Multi-process extensions Copyright (c) 2004, Steven Smith
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/fs.h>
++#include <linux/errno.h>
++#include <linux/miscdevice.h>
++#include <linux/major.h>
++#include <linux/proc_fs.h>
++#include <linux/stat.h>
++#include <linux/poll.h>
++#include <linux/irq.h>
++#include <linux/init.h>
++#include <linux/gfp.h>
++#include <linux/mutex.h>
++#include <linux/cpu.h>
++#include <xen/evtchn.h>
++#include <xen/public/evtchn.h>
++
++struct per_user_data {
++ /* Notification ring, accessed via /dev/xen/evtchn. */
++#define EVTCHN_RING_SIZE (PAGE_SIZE / sizeof(evtchn_port_t))
++#define EVTCHN_RING_MASK(_i) ((_i)&(EVTCHN_RING_SIZE-1))
++ evtchn_port_t *ring;
++ unsigned int ring_cons, ring_prod, ring_overflow;
++ struct mutex ring_cons_mutex; /* protect against concurrent readers */
++
++ /* Processes wait on this queue when ring is empty. */
++ wait_queue_head_t evtchn_wait;
++ struct fasync_struct *evtchn_async_queue;
++
++ int bind_cpu;
++ int nr_event_wrong_delivery;
++};
++
++/* Who's bound to each port? */
++static struct per_user_data *port_user[NR_EVENT_CHANNELS];
++static spinlock_t port_user_lock;
++
++void evtchn_device_upcall(int port)
++{
++ struct per_user_data *u;
++
++ spin_lock(&port_user_lock);
++
++ mask_evtchn(port);
++ clear_evtchn(port);
++
++ if ((u = port_user[port]) != NULL) {
++ if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) {
++ u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port;
++ wmb(); /* Ensure ring contents visible */
++ if (u->ring_cons == u->ring_prod++) {
++ wake_up_interruptible(&u->evtchn_wait);
++ kill_fasync(&u->evtchn_async_queue,
++ SIGIO, POLL_IN);
++ }
++ } else {
++ u->ring_overflow = 1;
++ }
++ }
++
++ spin_unlock(&port_user_lock);
++}
++
++static void evtchn_check_wrong_delivery(struct per_user_data *u)
++{
++ evtchn_port_t port;
++ unsigned int current_cpu = smp_processor_id();
++
++ /* Delivered to correct CPU? All is good. */
++ if (u->bind_cpu == current_cpu) {
++ u->nr_event_wrong_delivery = 0;
++ return;
++ }
++
++ /* Tolerate up to 100 consecutive misdeliveries. */
++ if (++u->nr_event_wrong_delivery < 100)
++ return;
++
++ spin_lock_irq(&port_user_lock);
++
++ for (port = 0; port < NR_EVENT_CHANNELS; port++)
++ if (port_user[port] == u)
++ rebind_evtchn_to_cpu(port, current_cpu);
++
++ u->bind_cpu = current_cpu;
++ u->nr_event_wrong_delivery = 0;
++
++ spin_unlock_irq(&port_user_lock);
++}
++
++static ssize_t evtchn_read(struct file *file, char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ int rc;
++ unsigned int c, p, bytes1 = 0, bytes2 = 0;
++ struct per_user_data *u = file->private_data;
++
++ /* Whole number of ports. */
++ count &= ~(sizeof(evtchn_port_t)-1);
++
++ if (count == 0)
++ return 0;
++
++ if (count > PAGE_SIZE)
++ count = PAGE_SIZE;
++
++ for (;;) {
++ mutex_lock(&u->ring_cons_mutex);
++
++ rc = -EFBIG;
++ if (u->ring_overflow)
++ goto unlock_out;
++
++ if ((c = u->ring_cons) != (p = u->ring_prod))
++ break;
++
++ mutex_unlock(&u->ring_cons_mutex);
++
++ if (file->f_flags & O_NONBLOCK)
++ return -EAGAIN;
++
++ rc = wait_event_interruptible(
++ u->evtchn_wait, u->ring_cons != u->ring_prod);
++ if (rc)
++ return rc;
++ }
++
++ /* Byte lengths of two chunks. Chunk split (if any) is at ring wrap. */
++ if (((c ^ p) & EVTCHN_RING_SIZE) != 0) {
++ bytes1 = (EVTCHN_RING_SIZE - EVTCHN_RING_MASK(c)) *
++ sizeof(evtchn_port_t);
++ bytes2 = EVTCHN_RING_MASK(p) * sizeof(evtchn_port_t);
++ } else {
++ bytes1 = (p - c) * sizeof(evtchn_port_t);
++ bytes2 = 0;
++ }
++
++ /* Truncate chunks according to caller's maximum byte count. */
++ if (bytes1 > count) {
++ bytes1 = count;
++ bytes2 = 0;
++ } else if ((bytes1 + bytes2) > count) {
++ bytes2 = count - bytes1;
++ }
++
++ rc = -EFAULT;
++ rmb(); /* Ensure that we see the port before we copy it. */
++ if (copy_to_user(buf, &u->ring[EVTCHN_RING_MASK(c)], bytes1) ||
++ ((bytes2 != 0) &&
++ copy_to_user(&buf[bytes1], &u->ring[0], bytes2)))
++ goto unlock_out;
++
++ evtchn_check_wrong_delivery(u);
++
++ u->ring_cons += (bytes1 + bytes2) / sizeof(evtchn_port_t);
++ rc = bytes1 + bytes2;
++
++ unlock_out:
++ mutex_unlock(&u->ring_cons_mutex);
++ return rc;
++}
++
++static ssize_t evtchn_write(struct file *file, const char __user *buf,
++ size_t count, loff_t *ppos)
++{
++ int rc, i;
++ evtchn_port_t *kbuf = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
++ struct per_user_data *u = file->private_data;
++
++ if (kbuf == NULL)
++ return -ENOMEM;
++
++ /* Whole number of ports. */
++ count &= ~(sizeof(evtchn_port_t)-1);
++
++ rc = 0;
++ if (count == 0)
++ goto out;
++
++ if (count > PAGE_SIZE)
++ count = PAGE_SIZE;
++
++ rc = -EFAULT;
++ if (copy_from_user(kbuf, buf, count) != 0)
++ goto out;
++
++ spin_lock_irq(&port_user_lock);
++ for (i = 0; i < (count/sizeof(evtchn_port_t)); i++)
++ if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u))
++ unmask_evtchn(kbuf[i]);
++ spin_unlock_irq(&port_user_lock);
++
++ rc = count;
++
++ out:
++ free_page((unsigned long)kbuf);
++ return rc;
++}
++
++static unsigned int next_bind_cpu(cpumask_t map)
++{
++ static unsigned int bind_cpu;
++ bind_cpu = next_cpu(bind_cpu, map);
++ if (bind_cpu >= NR_CPUS)
++ bind_cpu = first_cpu(map);
++ return bind_cpu;
++}
++
++static void evtchn_bind_to_user(struct per_user_data *u, int port)
++{
++ spin_lock_irq(&port_user_lock);
++
++ BUG_ON(port_user[port] != NULL);
++ port_user[port] = u;
++
++ if (u->bind_cpu == -1)
++ u->bind_cpu = next_bind_cpu(cpu_online_map);
++
++ rebind_evtchn_to_cpu(port, u->bind_cpu);
++
++ unmask_evtchn(port);
++
++ spin_unlock_irq(&port_user_lock);
++}
++
++static long evtchn_ioctl(struct file *file,
++ unsigned int cmd, unsigned long arg)
++{
++ int rc;
++ struct per_user_data *u = file->private_data;
++ void __user *uarg = (void __user *) arg;
++
++ switch (cmd) {
++ case IOCTL_EVTCHN_BIND_VIRQ: {
++ struct ioctl_evtchn_bind_virq bind;
++ struct evtchn_bind_virq bind_virq;
++
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
++
++ bind_virq.virq = bind.virq;
++ bind_virq.vcpu = 0;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
++ &bind_virq);
++ if (rc != 0)
++ break;
++
++ rc = bind_virq.port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
++
++ case IOCTL_EVTCHN_BIND_INTERDOMAIN: {
++ struct ioctl_evtchn_bind_interdomain bind;
++ struct evtchn_bind_interdomain bind_interdomain;
++
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
++
++ bind_interdomain.remote_dom = bind.remote_domain;
++ bind_interdomain.remote_port = bind.remote_port;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
++ &bind_interdomain);
++ if (rc != 0)
++ break;
++
++ rc = bind_interdomain.local_port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
++
++ case IOCTL_EVTCHN_BIND_UNBOUND_PORT: {
++ struct ioctl_evtchn_bind_unbound_port bind;
++ struct evtchn_alloc_unbound alloc_unbound;
++
++ rc = -EFAULT;
++ if (copy_from_user(&bind, uarg, sizeof(bind)))
++ break;
++
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = bind.remote_domain;
++ rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++ if (rc != 0)
++ break;
++
++ rc = alloc_unbound.port;
++ evtchn_bind_to_user(u, rc);
++ break;
++ }
++
++ case IOCTL_EVTCHN_UNBIND: {
++ struct ioctl_evtchn_unbind unbind;
++ struct evtchn_close close;
++ int ret;
++
++ rc = -EFAULT;
++ if (copy_from_user(&unbind, uarg, sizeof(unbind)))
++ break;
++
++ rc = -EINVAL;
++ if (unbind.port >= NR_EVENT_CHANNELS)
++ break;
++
++ spin_lock_irq(&port_user_lock);
++
++ rc = -ENOTCONN;
++ if (port_user[unbind.port] != u) {
++ spin_unlock_irq(&port_user_lock);
++ break;
++ }
++
++ port_user[unbind.port] = NULL;
++ mask_evtchn(unbind.port);
++ rebind_evtchn_to_cpu(unbind.port, 0);
++
++ spin_unlock_irq(&port_user_lock);
++
++ close.port = unbind.port;
++ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ BUG_ON(ret);
++
++ rc = 0;
++ break;
++ }
++
++ case IOCTL_EVTCHN_NOTIFY: {
++ struct ioctl_evtchn_notify notify;
++
++ rc = -EFAULT;
++ if (copy_from_user(¬ify, uarg, sizeof(notify)))
++ break;
++
++ if (notify.port >= NR_EVENT_CHANNELS) {
++ rc = -EINVAL;
++ } else if (port_user[notify.port] != u) {
++ rc = -ENOTCONN;
++ } else {
++ notify_remote_via_evtchn(notify.port);
++ rc = 0;
++ }
++ break;
++ }
++
++ case IOCTL_EVTCHN_RESET: {
++ /* Initialise the ring to empty. Clear errors. */
++ mutex_lock(&u->ring_cons_mutex);
++ spin_lock_irq(&port_user_lock);
++ u->ring_cons = u->ring_prod = u->ring_overflow = 0;
++ spin_unlock_irq(&port_user_lock);
++ mutex_unlock(&u->ring_cons_mutex);
++ rc = 0;
++ break;
++ }
++
++ default:
++ rc = -ENOSYS;
++ break;
++ }
++
++ return rc;
++}
++
++static unsigned int evtchn_poll(struct file *file, poll_table *wait)
++{
++ unsigned int mask = POLLOUT | POLLWRNORM;
++ struct per_user_data *u = file->private_data;
++
++ poll_wait(file, &u->evtchn_wait, wait);
++ if (u->ring_cons != u->ring_prod)
++ mask |= POLLIN | POLLRDNORM;
++ if (u->ring_overflow)
++ mask = POLLERR;
++ return mask;
++}
++
++static int evtchn_fasync(int fd, struct file *filp, int on)
++{
++ struct per_user_data *u = filp->private_data;
++ return fasync_helper(fd, filp, on, &u->evtchn_async_queue);
++}
++
++static int evtchn_open(struct inode *inode, struct file *filp)
++{
++ struct per_user_data *u;
++
++ if ((u = kmalloc(sizeof(*u), GFP_KERNEL)) == NULL)
++ return -ENOMEM;
++
++ memset(u, 0, sizeof(*u));
++ init_waitqueue_head(&u->evtchn_wait);
++
++ u->ring = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
++ if (u->ring == NULL) {
++ kfree(u);
++ return -ENOMEM;
++ }
++
++ mutex_init(&u->ring_cons_mutex);
++
++ filp->private_data = u;
++
++ u->bind_cpu = -1;
++
++ return 0;
++}
++
++static int evtchn_release(struct inode *inode, struct file *filp)
++{
++ int i;
++ struct per_user_data *u = filp->private_data;
++ struct evtchn_close close;
++
++ spin_lock_irq(&port_user_lock);
++
++ free_page((unsigned long)u->ring);
++
++ for (i = 0; i < NR_EVENT_CHANNELS; i++) {
++ int ret;
++ if (port_user[i] != u)
++ continue;
++
++ port_user[i] = NULL;
++ mask_evtchn(i);
++ rebind_evtchn_to_cpu(i, 0);
++
++ close.port = i;
++ ret = HYPERVISOR_event_channel_op(EVTCHNOP_close, &close);
++ BUG_ON(ret);
++ }
++
++ spin_unlock_irq(&port_user_lock);
++
++ kfree(u);
++
++ return 0;
++}
++
++static const struct file_operations evtchn_fops = {
++ .owner = THIS_MODULE,
++ .read = evtchn_read,
++ .write = evtchn_write,
++ .unlocked_ioctl = evtchn_ioctl,
++ .poll = evtchn_poll,
++ .fasync = evtchn_fasync,
++ .open = evtchn_open,
++ .release = evtchn_release,
++};
++
++static struct miscdevice evtchn_miscdev = {
++ .minor = MISC_DYNAMIC_MINOR,
++ .name = "evtchn",
++ .fops = &evtchn_fops,
++};
++
++static int __cpuinit evtchn_cpu_notify(struct notifier_block *nfb,
++ unsigned long action, void *hcpu)
++{
++ int hotcpu = (unsigned long)hcpu;
++ cpumask_t map = cpu_online_map;
++ int port, newcpu;
++ struct per_user_data *u;
++
++ switch (action) {
++ case CPU_DOWN_PREPARE:
++ cpu_clear(hotcpu, map);
++ spin_lock_irq(&port_user_lock);
++ for (port = 0; port < NR_EVENT_CHANNELS; port++) {
++ if ((u = port_user[port]) != NULL &&
++ u->bind_cpu == hotcpu &&
++ (newcpu = next_bind_cpu(map)) < NR_CPUS) {
++ rebind_evtchn_to_cpu(port, newcpu);
++ u->bind_cpu = newcpu;
++ }
++ }
++ spin_unlock_irq(&port_user_lock);
++ break;
++ default:
++ return NOTIFY_DONE;
++ }
++ return NOTIFY_OK;
++}
++
++static struct notifier_block __cpuinitdata evtchn_cpu_nfb = {
++ .notifier_call = evtchn_cpu_notify
++};
++
++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;
++ }
++
++ register_cpu_notifier(&evtchn_cpu_nfb);
++
++ printk("Event-channel device installed.\n");
++
++ return 0;
++}
++
++static void __exit evtchn_cleanup(void)
++{
++ misc_deregister(&evtchn_miscdev);
++ unregister_cpu_notifier(&evtchn_cpu_nfb);
++}
++
++module_init(evtchn_init);
++module_exit(evtchn_cleanup);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/fbfront/Makefile b/drivers/xen/fbfront/Makefile
+new file mode 100644
+index 0000000..e2b8909
+--- /dev/null
++++ b/drivers/xen/fbfront/Makefile
+@@ -0,0 +1,2 @@
++obj-$(CONFIG_XEN_FRAMEBUFFER) := xenfb.o
++obj-$(CONFIG_XEN_KEYBOARD) += xenkbd.o
+diff --git a/drivers/xen/fbfront/xenfb.c b/drivers/xen/fbfront/xenfb.c
+new file mode 100644
+index 0000000..1564b88
+--- /dev/null
++++ b/drivers/xen/fbfront/xenfb.c
+@@ -0,0 +1,884 @@
++/*
++ * 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/mutex.h>
++#include <linux/freezer.h>
++#include <asm/hypervisor.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/fbif.h>
++#include <xen/interface/io/protocols.h>
++#include <xen/xenbus.h>
++#include <linux/kthread.h>
++
++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 */
++
++ int irq;
++ struct xenfb_page *page;
++ unsigned long *mfns;
++ int update_wanted; /* XENFB_TYPE_UPDATE wanted */
++ int feature_resize; /* Backend has resize feature */
++ struct xenfb_resize resize;
++ int resize_dpy;
++ spinlock_t resize_lock;
++
++ struct xenbus_device *xbdev;
++};
++
++/*
++ * There are three locks:
++ * spinlock resize_lock protecting resize_dpy and resize
++ * spinlock dirty_lock protecting the dirty rectangle
++ * mutex mm_lock protecting mappings.
++ *
++ * How the dirty and mapping locks work together
++ *
++ * The problem is that dirty rectangle and mappings aren't
++ * independent: the dirty rectangle must cover all faulted pages in
++ * mappings. We need to prove that our locking maintains this
++ * invariant.
++ *
++ * There are several kinds of critical regions:
++ *
++ * 1. Holding only dirty_lock: xenfb_refresh(). May run in
++ * interrupts. Extends the dirty rectangle. Trivially preserves
++ * invariant.
++ *
++ * 2. Holding only mm_lock: xenfb_mmap() and xenfb_vm_close(). Touch
++ * only mappings. The former creates unfaulted pages. Preserves
++ * invariant. The latter removes pages. Preserves invariant.
++ *
++ * 3. Holding both locks: xenfb_vm_fault(). Extends the dirty
++ * rectangle and updates mappings consistently. Preserves
++ * invariant.
++ *
++ * 4. The ugliest one: xenfb_update_screen(). Clear the dirty
++ * rectangle and update mappings consistently.
++ *
++ * We can't simply hold both locks, because zap_page_range() cannot
++ * be called with a spinlock held.
++ *
++ * Therefore, we first clear the dirty rectangle with both locks
++ * held. Then we unlock dirty_lock and update the mappings.
++ * Critical regions that hold only dirty_lock may interfere with
++ * that. This can only be region 1: xenfb_refresh(). But that
++ * just extends the dirty rectangle, which can't harm the
++ * invariant.
++ *
++ * But FIXME: the invariant is too weak. It misses that the fault
++ * record in mappings must be consistent with the mapping of pages in
++ * the associated address space! __do_fault() updates the PTE after
++ * xenfb_vm_fault() returns, i.e. outside the critical region. This
++ * allows the following race:
++ *
++ * X writes to some address in the Xen frame buffer
++ * Fault - call __do_fault()
++ * call xenfb_vm_fault()
++ * grab mm_lock
++ * map->faults++;
++ * release mm_lock
++ * return back to do_no_page()
++ * (preempted, or SMP)
++ * Xen worker thread runs.
++ * grab mm_lock
++ * look at mappings
++ * find this mapping, zaps its pages (but page not in pte yet)
++ * clear map->faults
++ * releases mm_lock
++ * (back to X process)
++ * put page in X's pte
++ *
++ * Oh well, we wont be updating the writes to this page anytime soon.
++ */
++#define MB_ (1024*1024)
++#define XENFB_DEFAULT_FB_LEN (XENFB_WIDTH * XENFB_HEIGHT * XENFB_DEPTH / 8)
++
++enum {KPARAM_MEM, KPARAM_WIDTH, KPARAM_HEIGHT, KPARAM_CNT};
++static int video[KPARAM_CNT] = {2, XENFB_WIDTH, XENFB_HEIGHT};
++module_param_array(video, int, NULL, 0);
++MODULE_PARM_DESC(video,
++ "Size of video memory in MB and width,height in pixels, default = (2,800,600)");
++
++static int xenfb_fps = 20;
++
++static int xenfb_remove(struct xenbus_device *);
++static void xenfb_init_shared_page(struct xenfb_info *, struct fb_info *);
++static int xenfb_connect_backend(struct xenbus_device *, struct xenfb_info *);
++static void xenfb_disconnect_backend(struct xenfb_info *);
++
++static void xenfb_send_event(struct xenfb_info *info,
++ union xenfb_out_event *event)
++{
++ __u32 prod;
++
++ 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_irq(info->irq);
++}
++
++static void xenfb_do_update(struct xenfb_info *info,
++ int x, int y, int w, int h)
++{
++ union xenfb_out_event event;
++
++ memset(&event, 0, sizeof(event));
++ event.type = XENFB_TYPE_UPDATE;
++ event.update.x = x;
++ event.update.y = y;
++ event.update.width = w;
++ event.update.height = h;
++
++ /* caller ensures !xenfb_queue_full() */
++ xenfb_send_event(info, &event);
++}
++
++static void xenfb_do_resize(struct xenfb_info *info)
++{
++ union xenfb_out_event event;
++
++ memset(&event, 0, sizeof(event));
++ event.resize = info->resize;
++
++ /* caller ensures !xenfb_queue_full() */
++ xenfb_send_event(info, &event);
++}
++
++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);
++
++ if (x2 < x1 || y2 < y1) {
++ printk("xenfb_update_screen bogus rect %d %d %d %d\n",
++ x1, x2, y1, y2);
++ WARN_ON(1);
++ }
++ xenfb_do_update(info, x1, y1, x2 - x1, y2 - y1);
++}
++
++static void xenfb_handle_resize_dpy(struct xenfb_info *info)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->resize_lock, flags);
++ if (info->resize_dpy) {
++ if (!xenfb_queue_full(info)) {
++ info->resize_dpy = 0;
++ xenfb_do_resize(info);
++ }
++ }
++ spin_unlock_irqrestore(&info->resize_lock, flags);
++}
++
++static int xenfb_thread(void *data)
++{
++ struct xenfb_info *info = data;
++
++ while (!kthread_should_stop()) {
++ xenfb_handle_resize_dpy(info);
++ if (info->dirty) {
++ info->dirty = 0;
++ xenfb_update_screen(info);
++ }
++ wait_event_interruptible(info->wq,
++ kthread_should_stop() || info->dirty);
++ try_to_freeze();
++ }
++ return 0;
++}
++
++static int 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;
++ 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;
++ info->dirty = 1;
++
++ 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 int xenfb_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
++{
++ struct xenfb_mapping *map = vma->vm_private_data;
++ struct xenfb_info *info = map->info;
++ int pgnr = ((long)vmf->virtual_address - vma->vm_start) >> PAGE_SHIFT;
++ unsigned long flags;
++ struct page *page;
++ int y1, y2;
++
++ if (pgnr >= info->nr_pages)
++ return VM_FAULT_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);
++
++ vmf->page = page;
++
++ return VM_FAULT_MINOR;
++}
++
++static struct vm_operations_struct xenfb_vm_ops = {
++ .open = xenfb_vm_open,
++ .close = xenfb_vm_close,
++ .fault = xenfb_vm_fault,
++};
++
++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 int
++xenfb_check_var(struct fb_var_screeninfo *var, struct fb_info *info)
++{
++ struct xenfb_info *xenfb_info;
++ int required_mem_len;
++
++ xenfb_info = info->par;
++
++ if (!xenfb_info->feature_resize) {
++ if (var->xres == video[KPARAM_WIDTH] &&
++ var->yres == video[KPARAM_HEIGHT] &&
++ var->bits_per_pixel == xenfb_info->page->depth) {
++ return 0;
++ }
++ return -EINVAL;
++ }
++
++ /* Can't resize past initial width and height */
++ if (var->xres > video[KPARAM_WIDTH] || var->yres > video[KPARAM_HEIGHT])
++ return -EINVAL;
++
++ required_mem_len = var->xres * var->yres * (xenfb_info->page->depth / 8);
++ if (var->bits_per_pixel == xenfb_info->page->depth &&
++ var->xres <= info->fix.line_length / (XENFB_DEPTH / 8) &&
++ required_mem_len <= info->fix.smem_len) {
++ var->xres_virtual = var->xres;
++ var->yres_virtual = var->yres;
++ return 0;
++ }
++ return -EINVAL;
++}
++
++static int xenfb_set_par(struct fb_info *info)
++{
++ struct xenfb_info *xenfb_info;
++ unsigned long flags;
++
++ xenfb_info = info->par;
++
++ spin_lock_irqsave(&xenfb_info->resize_lock, flags);
++ xenfb_info->resize.type = XENFB_TYPE_RESIZE;
++ xenfb_info->resize.width = info->var.xres;
++ xenfb_info->resize.height = info->var.yres;
++ xenfb_info->resize.stride = info->fix.line_length;
++ xenfb_info->resize.depth = info->var.bits_per_pixel;
++ xenfb_info->resize.offset = 0;
++ xenfb_info->resize_dpy = 1;
++ spin_unlock_irqrestore(&xenfb_info->resize_lock, flags);
++ return 0;
++}
++
++static struct fb_ops xenfb_fb_ops = {
++ .owner = THIS_MODULE,
++ .fb_setcolreg = xenfb_setcolreg,
++ .fb_fillrect = xenfb_fillrect,
++ .fb_copyarea = xenfb_copyarea,
++ .fb_imageblit = xenfb_imageblit,
++ .fb_mmap = xenfb_mmap,
++ .fb_check_var = xenfb_check_var,
++ .fb_set_par = xenfb_set_par,
++};
++
++static irqreturn_t xenfb_event_handler(int rq, void *dev_id)
++{
++ /*
++ * 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_irq(info->irq);
++ }
++ 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 fb_size;
++ int val;
++ int ret;
++
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (info == NULL) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
++ }
++
++ /* Limit kernel param videoram amount to what is in xenstore */
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "videoram", "%d", &val) == 1) {
++ if (val < video[KPARAM_MEM])
++ video[KPARAM_MEM] = val;
++ }
++
++ /* If requested res does not fit in available memory, use default */
++ fb_size = video[KPARAM_MEM] * MB_;
++ if (video[KPARAM_WIDTH] * video[KPARAM_HEIGHT] * XENFB_DEPTH/8 > fb_size) {
++ video[KPARAM_WIDTH] = XENFB_WIDTH;
++ video[KPARAM_HEIGHT] = XENFB_HEIGHT;
++ fb_size = XENFB_DEFAULT_FB_LEN;
++ }
++
++ dev->dev.driver_data = info;
++ info->xbdev = dev;
++ info->irq = -1;
++ info->x1 = info->y1 = INT_MAX;
++ spin_lock_init(&info->dirty_lock);
++ spin_lock_init(&info->resize_lock);
++ mutex_init(&info->mm_lock);
++ init_waitqueue_head(&info->wq);
++ init_timer(&info->refresh);
++ info->refresh.function = xenfb_timer;
++ info->refresh.data = (unsigned long)info;
++ INIT_LIST_HEAD(&info->mappings);
++
++ info->fb = vmalloc(fb_size);
++ if (info->fb == NULL)
++ goto error_nomem;
++ memset(info->fb, 0, fb_size);
++
++ info->nr_pages = (fb_size + PAGE_SIZE - 1) >> PAGE_SHIFT;
++
++ info->pages = kmalloc(sizeof(struct page *) * info->nr_pages,
++ GFP_KERNEL);
++ if (info->pages == NULL)
++ goto error_nomem;
++
++ 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 | __GFP_ZERO);
++ if (!info->page)
++ goto error_nomem;
++
++ fb_info = framebuffer_alloc(sizeof(u32) * 256, NULL);
++ /* see fishy hackery below */
++ if (fb_info == NULL)
++ goto error_nomem;
++
++ /* FIXME fishy hackery */
++ fb_info->pseudo_palette = fb_info->par;
++ fb_info->par = info;
++ /* /FIXME */
++ fb_info->screen_base = info->fb;
++
++ fb_info->fbops = &xenfb_fb_ops;
++ fb_info->var.xres_virtual = fb_info->var.xres = video[KPARAM_WIDTH];
++ fb_info->var.yres_virtual = fb_info->var.yres = video[KPARAM_HEIGHT];
++ fb_info->var.bits_per_pixel = XENFB_DEPTH;
++
++ 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 = fb_info->var.xres * (XENFB_DEPTH / 8);
++ fb_info->fix.smem_start = 0;
++ fb_info->fix.smem_len = fb_size;
++ strcpy(fb_info->fix.id, "xen");
++ fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
++ fb_info->fix.accel = FB_ACCEL_NONE;
++
++ fb_info->flags = FBINFO_FLAG_DEFAULT;
++
++ ret = fb_alloc_cmap(&fb_info->cmap, 256, 0);
++ if (ret < 0) {
++ framebuffer_release(fb_info);
++ xenbus_dev_fatal(dev, ret, "fb_alloc_cmap");
++ goto error;
++ }
++
++ xenfb_init_shared_page(info, fb_info);
++
++ 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, info->fb_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,
++ struct fb_info * fb_info)
++{
++ int i;
++ int epd = PAGE_SIZE / sizeof(info->mfns[0]);
++
++ 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);
++
++ for (i = 0; i * epd < info->nr_pages; i++)
++ info->page->pd[i] = vmalloc_to_mfn(&info->mfns[i * epd]);
++
++ info->page->width = fb_info->var.xres;
++ info->page->height = fb_info->var.yres;
++ info->page->depth = fb_info->var.bits_per_pixel;
++ info->page->line_length = fb_info->fix.line_length;
++ info->page->mem_length = fb_info->fix.smem_len;
++ info->page->in_cons = info->page->in_prod = 0;
++ info->page->out_cons = info->page->out_prod = 0;
++}
++
++static int xenfb_connect_backend(struct xenbus_device *dev,
++ struct xenfb_info *info)
++{
++ int ret;
++ struct xenbus_transaction xbt;
++
++ ret = bind_listening_port_to_irqhandler(
++ dev->otherend_id, xenfb_event_handler, 0, "xenfb", info);
++ if (ret < 0) {
++ xenbus_dev_fatal(dev, ret,
++ "bind_listening_port_to_irqhandler");
++ return ret;
++ }
++ info->irq = ret;
++
++ again:
++ ret = xenbus_transaction_start(&xbt);
++ if (ret) {
++ xenbus_dev_fatal(dev, ret, "starting transaction");
++ return ret;
++ }
++ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
++ virt_to_mfn(info->page));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
++ XEN_IO_PROTO_ABI_NATIVE);
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "feature-update", "1");
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_transaction_end(xbt, 0);
++ if (ret) {
++ if (ret == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, ret, "completing transaction");
++ return ret;
++ }
++
++ 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 XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateInitWait:
++ InitWait:
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
++
++ case XenbusStateConnected:
++ /*
++ * Work around xenbus race condition: If backend goes
++ * through InitWait to Connected fast enough, we can
++ * get Connected twice here.
++ */
++ if (dev->state != XenbusStateConnected)
++ goto InitWait; /* no InitWait seen yet, fudge it */
++
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "request-update", "%d", &val) < 0)
++ val = 0;
++ if (val)
++ info->update_wanted = 1;
++
++ if (xenbus_scanf(XBT_NIL, dev->otherend,
++ "feature-resize", "%d", &val) < 0)
++ val = 0;
++ info->feature_resize = val;
++ break;
++
++ case XenbusStateClosing:
++ // FIXME is this safe in any dev->state?
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
++
++static const struct xenbus_device_id xenfb_ids[] = {
++ { "vfb" },
++ { "" }
++};
++MODULE_ALIAS("xen:vfb");
++
++static struct xenbus_driver xenfb_driver = {
++ .name = "vfb",
++ .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_driver);
++}
++
++static void __exit xenfb_cleanup(void)
++{
++ return xenbus_unregister_driver(&xenfb_driver);
++}
++
++module_init(xenfb_init);
++module_exit(xenfb_cleanup);
++
++MODULE_LICENSE("GPL");
+diff --git a/drivers/xen/fbfront/xenkbd.c b/drivers/xen/fbfront/xenkbd.c
+new file mode 100644
+index 0000000..425318b
+--- /dev/null
++++ b/drivers/xen/fbfront/xenkbd.c
+@@ -0,0 +1,353 @@
++/*
++ * 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 *kbd;
++ struct input_dev *ptr;
++ struct xenkbd_page *page;
++ int irq;
++ struct xenbus_device *xbdev;
++ char phys[32];
++};
++
++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 xenkbd_info *info = dev_id;
++ struct xenkbd_page *page = info->page;
++ __u32 cons, prod;
++
++ prod = page->in_prod;
++ if (prod == page->in_cons)
++ return IRQ_HANDLED;
++ rmb(); /* ensure we see ring contents up to prod */
++ for (cons = page->in_cons; cons != prod; cons++) {
++ union xenkbd_in_event *event;
++ struct input_dev *dev;
++ event = &XENKBD_IN_RING_REF(page, cons);
++
++ dev = info->ptr;
++ switch (event->type) {
++ case XENKBD_TYPE_MOTION:
++ if (event->motion.rel_z)
++ input_report_rel(dev, REL_WHEEL,
++ -event->motion.rel_z);
++ input_report_rel(dev, REL_X, event->motion.rel_x);
++ input_report_rel(dev, REL_Y, event->motion.rel_y);
++ break;
++ case XENKBD_TYPE_KEY:
++ dev = NULL;
++ if (test_bit(event->key.keycode, info->kbd->keybit))
++ dev = info->kbd;
++ if (test_bit(event->key.keycode, info->ptr->keybit))
++ dev = info->ptr;
++ if (dev)
++ input_report_key(dev, event->key.keycode,
++ event->key.pressed);
++ else
++ printk("xenkbd: unhandled keycode 0x%x\n",
++ event->key.keycode);
++ break;
++ case XENKBD_TYPE_POS:
++ if (event->pos.rel_z)
++ input_report_rel(dev, REL_WHEEL,
++ -event->pos.rel_z);
++ input_report_abs(dev, ABS_X, event->pos.abs_x);
++ input_report_abs(dev, ABS_Y, event->pos.abs_y);
++ break;
++ }
++ if (dev)
++ input_sync(dev);
++ }
++ mb(); /* ensure we got ring contents */
++ page->in_cons = cons;
++ notify_remote_via_irq(info->irq);
++
++ 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 *kbd, *ptr;
++
++ info = kzalloc(sizeof(*info), GFP_KERNEL);
++ if (!info) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
++ return -ENOMEM;
++ }
++ dev->dev.driver_data = info;
++ info->xbdev = dev;
++ snprintf(info->phys, sizeof(info->phys), "xenbus/%s", dev->nodename);
++
++ 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;
++
++ /* keyboard */
++ kbd = input_allocate_device();
++ if (!kbd)
++ goto error_nomem;
++ kbd->name = "Xen Virtual Keyboard";
++ kbd->phys = info->phys;
++ kbd->id.bustype = BUS_PCI;
++ kbd->id.vendor = 0x5853;
++ kbd->id.product = 0xffff;
++ kbd->evbit[0] = BIT(EV_KEY);
++ for (i = KEY_ESC; i < KEY_UNKNOWN; i++)
++ set_bit(i, kbd->keybit);
++ for (i = KEY_OK; i < KEY_MAX; i++)
++ set_bit(i, kbd->keybit);
++
++ ret = input_register_device(kbd);
++ if (ret) {
++ input_free_device(kbd);
++ xenbus_dev_fatal(dev, ret, "input_register_device(kbd)");
++ goto error;
++ }
++ info->kbd = kbd;
++
++ /* pointing device */
++ ptr = input_allocate_device();
++ if (!ptr)
++ goto error_nomem;
++ ptr->name = "Xen Virtual Pointer";
++ ptr->phys = info->phys;
++ ptr->id.bustype = BUS_PCI;
++ ptr->id.vendor = 0x5853;
++ ptr->id.product = 0xfffe;
++ ptr->evbit[0] = BIT(EV_KEY) | BIT(EV_REL) | BIT(EV_ABS);
++ for (i = BTN_LEFT; i <= BTN_TASK; i++)
++ set_bit(i, ptr->keybit);
++ ptr->relbit[0] = BIT(REL_X) | BIT(REL_Y) | BIT(REL_WHEEL);
++ input_set_abs_params(ptr, ABS_X, 0, XENFB_WIDTH, 0, 0);
++ input_set_abs_params(ptr, ABS_Y, 0, XENFB_HEIGHT, 0, 0);
++
++ ret = input_register_device(ptr);
++ if (ret) {
++ input_free_device(ptr);
++ xenbus_dev_fatal(dev, ret, "input_register_device(ptr)");
++ goto error;
++ }
++ info->ptr = ptr;
++
++ ret = xenkbd_connect_backend(dev, info);
++ if (ret < 0)
++ goto error;
++
++ 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);
++ info->page->in_cons = info->page->in_prod = 0;
++ info->page->out_cons = info->page->out_prod = 0;
++ 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->kbd);
++ input_unregister_device(info->ptr);
++ free_page((unsigned long)info->page);
++ kfree(info);
++ return 0;
++}
++
++static int xenkbd_connect_backend(struct xenbus_device *dev,
++ struct xenkbd_info *info)
++{
++ int ret;
++ struct xenbus_transaction xbt;
++
++ ret = bind_listening_port_to_irqhandler(
++ dev->otherend_id, input_handler, 0, "xenkbd", info);
++ if (ret < 0) {
++ xenbus_dev_fatal(dev, ret,
++ "bind_listening_port_to_irqhandler");
++ return ret;
++ }
++ info->irq = ret;
++
++ again:
++ ret = xenbus_transaction_start(&xbt);
++ if (ret) {
++ xenbus_dev_fatal(dev, ret, "starting transaction");
++ return ret;
++ }
++ ret = xenbus_printf(xbt, dev->nodename, "page-ref", "%lu",
++ virt_to_mfn(info->page));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (ret)
++ goto error_xenbus;
++ ret = xenbus_transaction_end(xbt, 0);
++ if (ret) {
++ if (ret == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, ret, "completing transaction");
++ return ret;
++ }
++
++ 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;
++}
++
++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;
++
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
++
++ 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;
++
++ 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 */
++
++ /* Set input abs params to match backend screen res */
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "width", "%d", &val) > 0 )
++ input_set_abs_params(info->ptr, ABS_X, 0, val, 0, 0);
++
++ if (xenbus_scanf(XBT_NIL, info->xbdev->otherend,
++ "height", "%d", &val) > 0 )
++ input_set_abs_params(info->ptr, ABS_Y, 0, val, 0, 0);
++
++ break;
++
++ case XenbusStateClosing:
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
++
++static const struct xenbus_device_id xenkbd_ids[] = {
++ { "vkbd" },
++ { "" }
++};
++MODULE_ALIAS("xen:vkbd");
++
++static struct xenbus_driver xenkbd_driver = {
++ .name = "vkbd",
++ .ids = xenkbd_ids,
++ .probe = xenkbd_probe,
++ .remove = xenkbd_remove,
++ .resume = xenkbd_resume,
++ .otherend_changed = xenkbd_backend_changed,
++};
++
++static int __init xenkbd_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(&xenkbd_driver);
++}
++
++static void __exit xenkbd_cleanup(void)
++{
++ return xenbus_unregister_driver(&xenkbd_driver);
++}
++
++module_init(xenkbd_init);
++module_exit(xenkbd_cleanup);
++
++MODULE_LICENSE("GPL");
+diff --git a/drivers/xen/gntdev/Makefile b/drivers/xen/gntdev/Makefile
+new file mode 100644
+index 0000000..8bd8c62
+--- /dev/null
++++ b/drivers/xen/gntdev/Makefile
+@@ -0,0 +1 @@
++obj-$(CONFIG_XEN_GRANT_DEV) := gntdev.o
+diff --git a/drivers/xen/gntdev/gntdev.c b/drivers/xen/gntdev/gntdev.c
+new file mode 100644
+index 0000000..fccf8d3
+--- /dev/null
++++ b/drivers/xen/gntdev/gntdev.c
+@@ -0,0 +1,1075 @@
++/******************************************************************************
++ * gntdev.c
++ *
++ * Device for accessing (in user-space) pages that have been granted by other
++ * domains.
++ *
++ * Copyright (c) 2006-2007, D G Murray.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ */
++
++#include <asm/atomic.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/init.h>
++#include <linux/fs.h>
++#include <linux/device.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <asm/uaccess.h>
++#include <asm/io.h>
++#include <xen/gnttab.h>
++#include <asm/hypervisor.h>
++#include <xen/balloon.h>
++#include <xen/evtchn.h>
++#include <xen/driver_util.h>
++
++#include <linux/types.h>
++#include <xen/public/gntdev.h>
++
++
++#define DRIVER_AUTHOR "Derek G. Murray <Derek.Murray at cl.cam.ac.uk>"
++#define DRIVER_DESC "User-space granted page access driver"
++
++MODULE_LICENSE("GPL");
++MODULE_AUTHOR(DRIVER_AUTHOR);
++MODULE_DESCRIPTION(DRIVER_DESC);
++
++#define MAX_GRANTS_LIMIT 1024
++#define DEFAULT_MAX_GRANTS 128
++
++/* A slot can be in one of three states:
++ *
++ * 0. GNTDEV_SLOT_INVALID:
++ * This slot is not associated with a grant reference, and is therefore free
++ * to be overwritten by a new grant reference.
++ *
++ * 1. GNTDEV_SLOT_NOT_YET_MAPPED:
++ * This slot is associated with a grant reference (via the
++ * IOCTL_GNTDEV_MAP_GRANT_REF ioctl), but it has not yet been mmap()-ed.
++ *
++ * 2. GNTDEV_SLOT_MAPPED:
++ * This slot is associated with a grant reference, and has been mmap()-ed.
++ */
++typedef enum gntdev_slot_state {
++ GNTDEV_SLOT_INVALID = 0,
++ GNTDEV_SLOT_NOT_YET_MAPPED,
++ GNTDEV_SLOT_MAPPED
++} gntdev_slot_state_t;
++
++#define GNTDEV_INVALID_HANDLE -1
++#define GNTDEV_FREE_LIST_INVALID -1
++/* Each opened instance of gntdev is associated with a list of grants,
++ * represented by an array of elements of the following type,
++ * gntdev_grant_info_t.
++ */
++typedef struct gntdev_grant_info {
++ gntdev_slot_state_t state;
++ union {
++ uint32_t free_list_index;
++ struct {
++ domid_t domid;
++ grant_ref_t ref;
++ grant_handle_t kernel_handle;
++ grant_handle_t user_handle;
++ uint64_t dev_bus_addr;
++ } valid;
++ } u;
++} gntdev_grant_info_t;
++
++/* Private data structure, which is stored in the file pointer for files
++ * associated with this device.
++ */
++typedef struct gntdev_file_private_data {
++
++ /* Array of grant information. */
++ gntdev_grant_info_t *grants;
++ uint32_t grants_size;
++
++ /* Read/write semaphore used to protect the grants array. */
++ struct rw_semaphore grants_sem;
++
++ /* An array of indices of free slots in the grants array.
++ * N.B. An entry in this list may temporarily have the value
++ * GNTDEV_FREE_LIST_INVALID if the corresponding slot has been removed
++ * from the list by the contiguous allocator, but the list has not yet
++ * been compressed. However, this is not visible across invocations of
++ * the device.
++ */
++ int32_t *free_list;
++
++ /* The number of free slots in the grants array. */
++ uint32_t free_list_size;
++
++ /* Read/write semaphore used to protect the free list. */
++ struct rw_semaphore free_list_sem;
++
++ /* Index of the next slot after the most recent contiguous allocation,
++ * for use in a next-fit allocator.
++ */
++ uint32_t next_fit_index;
++
++ /* Used to map grants into the kernel, before mapping them into user
++ * space.
++ */
++ struct page **foreign_pages;
++
++} gntdev_file_private_data_t;
++
++/* Module lifecycle operations. */
++static int __init gntdev_init(void);
++static void __exit gntdev_exit(void);
++
++module_init(gntdev_init);
++module_exit(gntdev_exit);
++
++/* File operations. */
++static int gntdev_open(struct inode *inode, struct file *flip);
++static int gntdev_release(struct inode *inode, struct file *flip);
++static int gntdev_mmap(struct file *flip, struct vm_area_struct *vma);
++static long gntdev_ioctl(struct file *flip,
++ unsigned int cmd, unsigned long arg);
++
++static const struct file_operations gntdev_fops = {
++ .owner = THIS_MODULE,
++ .open = gntdev_open,
++ .release = gntdev_release,
++ .mmap = gntdev_mmap,
++ .unlocked_ioctl = gntdev_ioctl
++};
++
++/* VM operations. */
++static void gntdev_vma_close(struct vm_area_struct *vma);
++static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
++ pte_t *ptep, int is_fullmm);
++
++static struct vm_operations_struct gntdev_vmops = {
++ .close = gntdev_vma_close,
++ .zap_pte = gntdev_clear_pte
++};
++
++/* Global variables. */
++
++/* The driver major number, for use when unregistering the driver. */
++static int gntdev_major;
++
++#define GNTDEV_NAME "gntdev"
++
++/* Memory mapping functions
++ * ------------------------
++ *
++ * Every granted page is mapped into both kernel and user space, and the two
++ * following functions return the respective virtual addresses of these pages.
++ *
++ * When shadow paging is disabled, the granted page is mapped directly into
++ * user space; when it is enabled, it is mapped into the kernel and remapped
++ * into user space using vm_insert_page() (see gntdev_mmap(), below).
++ */
++
++/* Returns the virtual address (in user space) of the @page_index'th page
++ * in the given VM area.
++ */
++static inline unsigned long get_user_vaddr (struct vm_area_struct *vma,
++ int page_index)
++{
++ return (unsigned long) vma->vm_start + (page_index << PAGE_SHIFT);
++}
++
++/* Returns the virtual address (in kernel space) of the @slot_index'th page
++ * mapped by the gntdev instance that owns the given private data struct.
++ */
++static inline unsigned long get_kernel_vaddr (gntdev_file_private_data_t *priv,
++ int slot_index)
++{
++ unsigned long pfn;
++ void *kaddr;
++ pfn = page_to_pfn(priv->foreign_pages[slot_index]);
++ kaddr = pfn_to_kaddr(pfn);
++ return (unsigned long) kaddr;
++}
++
++/* Helper functions. */
++
++/* Adds information about a grant reference to the list of grants in the file's
++ * private data structure. Returns non-zero on failure. On success, sets the
++ * value of *offset to the offset that should be mmap()-ed in order to map the
++ * grant reference.
++ */
++static int add_grant_reference(struct file *flip,
++ struct ioctl_gntdev_grant_ref *op,
++ uint64_t *offset)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++
++ uint32_t slot_index;
++
++ if (unlikely(private_data->free_list_size == 0)) {
++ return -ENOMEM;
++ }
++
++ slot_index = private_data->free_list[--private_data->free_list_size];
++ private_data->free_list[private_data->free_list_size]
++ = GNTDEV_FREE_LIST_INVALID;
++
++ /* Copy the grant information into file's private data. */
++ private_data->grants[slot_index].state = GNTDEV_SLOT_NOT_YET_MAPPED;
++ private_data->grants[slot_index].u.valid.domid = op->domid;
++ private_data->grants[slot_index].u.valid.ref = op->ref;
++
++ /* The offset is calculated as the index of the chosen entry in the
++ * file's private data's array of grant information. This is then
++ * shifted to give an offset into the virtual "file address space".
++ */
++ *offset = slot_index << PAGE_SHIFT;
++
++ return 0;
++}
++
++/* Adds the @count grant references to the contiguous range in the slot array
++ * beginning at @first_slot. It is assumed that @first_slot was returned by a
++ * previous invocation of find_contiguous_free_range(), during the same
++ * invocation of the driver.
++ */
++static int add_grant_references(struct file *flip,
++ int count,
++ struct ioctl_gntdev_grant_ref *ops,
++ uint32_t first_slot)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++ int i;
++
++ for (i = 0; i < count; ++i) {
++
++ /* First, mark the slot's entry in the free list as invalid. */
++ int free_list_index =
++ private_data->grants[first_slot+i].u.free_list_index;
++ private_data->free_list[free_list_index] =
++ GNTDEV_FREE_LIST_INVALID;
++
++ /* Now, update the slot. */
++ private_data->grants[first_slot+i].state =
++ GNTDEV_SLOT_NOT_YET_MAPPED;
++ private_data->grants[first_slot+i].u.valid.domid =
++ ops[i].domid;
++ private_data->grants[first_slot+i].u.valid.ref = ops[i].ref;
++ }
++
++ return 0;
++}
++
++/* Scans through the free list for @flip, removing entries that are marked as
++ * GNTDEV_SLOT_INVALID. This will reduce the recorded size of the free list to
++ * the number of valid entries.
++ */
++static void compress_free_list(struct file *flip)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++ int i, j = 0, old_size, slot_index;
++
++ old_size = private_data->free_list_size;
++ for (i = 0; i < old_size; ++i) {
++ if (private_data->free_list[i] != GNTDEV_FREE_LIST_INVALID) {
++ if (i > j) {
++ slot_index = private_data->free_list[i];
++ private_data->free_list[j] = slot_index;
++ private_data->grants[slot_index].u
++ .free_list_index = j;
++ private_data->free_list[i]
++ = GNTDEV_FREE_LIST_INVALID;
++ }
++ ++j;
++ } else {
++ --private_data->free_list_size;
++ }
++ }
++}
++
++/* Searches the grant array in the private data of @flip for a range of
++ * @num_slots contiguous slots in the GNTDEV_SLOT_INVALID state.
++ *
++ * Returns the index of the first slot if a range is found, otherwise -ENOMEM.
++ */
++static int find_contiguous_free_range(struct file *flip,
++ uint32_t num_slots)
++{
++ gntdev_file_private_data_t *private_data
++ = (gntdev_file_private_data_t *) flip->private_data;
++
++ int i;
++ int start_index = private_data->next_fit_index;
++ int range_start = 0, range_length;
++
++ if (private_data->free_list_size < num_slots) {
++ return -ENOMEM;
++ }
++
++ /* First search from the start_index to the end of the array. */
++ range_length = 0;
++ for (i = start_index; i < private_data->grants_size; ++i) {
++ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
++ if (range_length == 0) {
++ range_start = i;
++ }
++ ++range_length;
++ if (range_length == num_slots) {
++ return range_start;
++ }
++ }
++ }
++
++ /* Now search from the start of the array to the start_index. */
++ range_length = 0;
++ for (i = 0; i < start_index; ++i) {
++ if (private_data->grants[i].state == GNTDEV_SLOT_INVALID) {
++ if (range_length == 0) {
++ range_start = i;
++ }
++ ++range_length;
++ if (range_length == num_slots) {
++ return range_start;
++ }
++ }
++ }
++
++ return -ENOMEM;
++}
++
++static int init_private_data(gntdev_file_private_data_t *priv,
++ uint32_t max_grants)
++{
++ int i;
++
++ /* Allocate space for the kernel-mapping of granted pages. */
++ priv->foreign_pages =
++ alloc_empty_pages_and_pagevec(max_grants);
++ if (!priv->foreign_pages)
++ goto nomem_out;
++
++ /* Allocate the grant list and free-list. */
++ priv->grants = kmalloc(max_grants * sizeof(gntdev_grant_info_t),
++ GFP_KERNEL);
++ if (!priv->grants)
++ goto nomem_out2;
++ priv->free_list = kmalloc(max_grants * sizeof(int32_t), GFP_KERNEL);
++ if (!priv->free_list)
++ goto nomem_out3;
++
++ /* Initialise the free-list, which contains all slots at first. */
++ for (i = 0; i < max_grants; ++i) {
++ priv->free_list[max_grants - i - 1] = i;
++ priv->grants[i].state = GNTDEV_SLOT_INVALID;
++ priv->grants[i].u.free_list_index = max_grants - i - 1;
++ }
++ priv->grants_size = max_grants;
++ priv->free_list_size = max_grants;
++ priv->next_fit_index = 0;
++
++ return 0;
++
++nomem_out3:
++ kfree(priv->grants);
++nomem_out2:
++ free_empty_pages_and_pagevec(priv->foreign_pages, max_grants);
++nomem_out:
++ return -ENOMEM;
++
++}
++
++/* Interface functions. */
++
++/* Initialises the driver. Called when the module is loaded. */
++static int __init gntdev_init(void)
++{
++ struct class *class;
++ struct device *device;
++
++ if (!is_running_on_xen()) {
++ printk(KERN_ERR "You must be running Xen to use gntdev\n");
++ return -ENODEV;
++ }
++
++ gntdev_major = register_chrdev(0, GNTDEV_NAME, &gntdev_fops);
++ if (gntdev_major < 0)
++ {
++ printk(KERN_ERR "Could not register gntdev device\n");
++ return -ENOMEM;
++ }
++
++ /* Note that if the sysfs code fails, we will still initialise the
++ * device, and output the major number so that the device can be
++ * created manually using mknod.
++ */
++ if ((class = get_xen_class()) == NULL) {
++ printk(KERN_ERR "Error setting up xen_class\n");
++ printk(KERN_ERR "gntdev created with major number = %d\n",
++ gntdev_major);
++ return 0;
++ }
++
++ device = device_create(class, NULL, MKDEV(gntdev_major, 0),
++ GNTDEV_NAME);
++ if (IS_ERR(device)) {
++ printk(KERN_ERR "Error creating gntdev device in xen_class\n");
++ printk(KERN_ERR "gntdev created with major number = %d\n",
++ gntdev_major);
++ return 0;
++ }
++
++ return 0;
++}
++
++/* Cleans up and unregisters the driver. Called when the driver is unloaded.
++ */
++static void __exit gntdev_exit(void)
++{
++ struct class *class;
++ if ((class = get_xen_class()) != NULL)
++ device_destroy(class, MKDEV(gntdev_major, 0));
++ unregister_chrdev(gntdev_major, GNTDEV_NAME);
++}
++
++/* Called when the device is opened. */
++static int gntdev_open(struct inode *inode, struct file *flip)
++{
++ gntdev_file_private_data_t *private_data;
++
++ try_module_get(THIS_MODULE);
++
++ /* Allocate space for the per-instance private data. */
++ private_data = kmalloc(sizeof(*private_data), GFP_KERNEL);
++ if (!private_data)
++ goto nomem_out;
++
++ /* These will be lazily initialised by init_private_data. */
++ private_data->grants = NULL;
++ private_data->free_list = NULL;
++ private_data->foreign_pages = NULL;
++
++ init_rwsem(&private_data->grants_sem);
++ init_rwsem(&private_data->free_list_sem);
++
++ flip->private_data = private_data;
++
++ return 0;
++
++nomem_out:
++ return -ENOMEM;
++}
++
++/* Called when the device is closed.
++ */
++static int gntdev_release(struct inode *inode, struct file *flip)
++{
++ if (flip->private_data) {
++ gntdev_file_private_data_t *private_data =
++ (gntdev_file_private_data_t *) flip->private_data;
++ if (private_data->foreign_pages)
++ free_empty_pages_and_pagevec
++ (private_data->foreign_pages,
++ private_data->grants_size);
++ if (private_data->grants)
++ kfree(private_data->grants);
++ if (private_data->free_list)
++ kfree(private_data->free_list);
++ kfree(private_data);
++ }
++ module_put(THIS_MODULE);
++ return 0;
++}
++
++/* Called when an attempt is made to mmap() the device. The private data from
++ * @flip contains the list of grant references that can be mapped. The vm_pgoff
++ * field of @vma contains the index into that list that refers to the grant
++ * reference that will be mapped. Only mappings that are a multiple of
++ * PAGE_SIZE are handled.
++ */
++static int gntdev_mmap (struct file *flip, struct vm_area_struct *vma)
++{
++ struct gnttab_map_grant_ref op;
++ unsigned long slot_index = vma->vm_pgoff;
++ unsigned long kernel_vaddr, user_vaddr;
++ uint32_t size = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ uint64_t ptep;
++ int ret;
++ int flags;
++ int i;
++ struct page *page;
++ gntdev_file_private_data_t *private_data = flip->private_data;
++
++ if (unlikely(!private_data)) {
++ printk(KERN_ERR "File's private data is NULL.\n");
++ return -EINVAL;
++ }
++
++ /* Test to make sure that the grants array has been initialised. */
++ down_read(&private_data->grants_sem);
++ if (unlikely(!private_data->grants)) {
++ up_read(&private_data->grants_sem);
++ printk(KERN_ERR "Attempted to mmap before ioctl.\n");
++ return -EINVAL;
++ }
++ up_read(&private_data->grants_sem);
++
++ if (unlikely((size <= 0) ||
++ (size + slot_index) > private_data->grants_size)) {
++ printk(KERN_ERR "Invalid number of pages or offset"
++ "(num_pages = %d, first_slot = %ld).\n",
++ size, slot_index);
++ return -ENXIO;
++ }
++
++ if ((vma->vm_flags & VM_WRITE) && !(vma->vm_flags & VM_SHARED)) {
++ printk(KERN_ERR "Writable mappings must be shared.\n");
++ return -EINVAL;
++ }
++
++ /* Slots must be in the NOT_YET_MAPPED state. */
++ down_write(&private_data->grants_sem);
++ for (i = 0; i < size; ++i) {
++ if (private_data->grants[slot_index + i].state !=
++ GNTDEV_SLOT_NOT_YET_MAPPED) {
++ printk(KERN_ERR "Slot (index = %ld) is in the wrong "
++ "state (%d).\n", slot_index + i,
++ private_data->grants[slot_index + i].state);
++ up_write(&private_data->grants_sem);
++ return -EINVAL;
++ }
++ }
++
++ /* Install the hook for unmapping. */
++ vma->vm_ops = &gntdev_vmops;
++
++ /* The VM area contains pages from another VM. */
++ vma->vm_flags |= VM_FOREIGN;
++ vma->vm_private_data = kzalloc(size * sizeof(struct page *),
++ GFP_KERNEL);
++ if (vma->vm_private_data == NULL) {
++ printk(KERN_ERR "Couldn't allocate mapping structure for VM "
++ "area.\n");
++ return -ENOMEM;
++ }
++
++ /* This flag prevents Bad PTE errors when the memory is unmapped. */
++ vma->vm_flags |= VM_RESERVED;
++
++ /* This flag prevents this VM area being copied on a fork(). A better
++ * behaviour might be to explicitly carry out the appropriate mappings
++ * on fork(), but I don't know if there's a hook for this.
++ */
++ vma->vm_flags |= VM_DONTCOPY;
++
++#ifdef CONFIG_X86
++ /* This flag ensures that the page tables are not unpinned before the
++ * VM area is unmapped. Therefore Xen still recognises the PTE as
++ * belonging to an L1 pagetable, and the grant unmap operation will
++ * succeed, even if the process does not exit cleanly.
++ */
++ vma->vm_mm->context.has_foreign_mappings = 1;
++#endif
++
++ for (i = 0; i < size; ++i) {
++
++ flags = GNTMAP_host_map;
++ if (!(vma->vm_flags & VM_WRITE))
++ flags |= GNTMAP_readonly;
++
++ kernel_vaddr = get_kernel_vaddr(private_data, slot_index + i);
++ user_vaddr = get_user_vaddr(vma, i);
++ page = pfn_to_page(__pa(kernel_vaddr) >> PAGE_SHIFT);
++
++ gnttab_set_map_op(&op, kernel_vaddr, flags,
++ private_data->grants[slot_index+i]
++ .u.valid.ref,
++ private_data->grants[slot_index+i]
++ .u.valid.domid);
++
++ /* Carry out the mapping of the grant reference. */
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status) {
++ printk(KERN_ERR "Error mapping the grant reference "
++ "into the kernel (%d). domid = %d; ref = %d\n",
++ op.status,
++ private_data->grants[slot_index+i]
++ .u.valid.domid,
++ private_data->grants[slot_index+i]
++ .u.valid.ref);
++ goto undo_map_out;
++ }
++
++ /* Store a reference to the page that will be mapped into user
++ * space.
++ */
++ ((struct page **) vma->vm_private_data)[i] = page;
++
++ /* Mark mapped page as reserved. */
++ SetPageReserved(page);
++
++ /* Record the grant handle, for use in the unmap operation. */
++ private_data->grants[slot_index+i].u.valid.kernel_handle =
++ op.handle;
++ private_data->grants[slot_index+i].u.valid.dev_bus_addr =
++ op.dev_bus_addr;
++
++ private_data->grants[slot_index+i].state = GNTDEV_SLOT_MAPPED;
++ private_data->grants[slot_index+i].u.valid.user_handle =
++ GNTDEV_INVALID_HANDLE;
++
++ /* Now perform the mapping to user space. */
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++
++ /* NOT USING SHADOW PAGE TABLES. */
++ /* In this case, we map the grant(s) straight into user
++ * space.
++ */
++
++ /* Get the machine address of the PTE for the user
++ * page.
++ */
++ if ((ret = create_lookup_pte_addr(vma->vm_mm,
++ vma->vm_start
++ + (i << PAGE_SHIFT),
++ &ptep)))
++ {
++ printk(KERN_ERR "Error obtaining PTE pointer "
++ "(%d).\n", ret);
++ goto undo_map_out;
++ }
++
++ /* Configure the map operation. */
++
++ /* The reference is to be used by host CPUs. */
++ flags = GNTMAP_host_map;
++
++ /* Specifies a user space mapping. */
++ flags |= GNTMAP_application_map;
++
++ /* The map request contains the machine address of the
++ * PTE to update.
++ */
++ flags |= GNTMAP_contains_pte;
++
++ if (!(vma->vm_flags & VM_WRITE))
++ flags |= GNTMAP_readonly;
++
++ gnttab_set_map_op(&op, ptep, flags,
++ private_data->grants[slot_index+i]
++ .u.valid.ref,
++ private_data->grants[slot_index+i]
++ .u.valid.domid);
++
++ /* Carry out the mapping of the grant reference. */
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status) {
++ printk(KERN_ERR "Error mapping the grant "
++ "reference into user space (%d). domid "
++ "= %d; ref = %d\n", op.status,
++ private_data->grants[slot_index+i].u
++ .valid.domid,
++ private_data->grants[slot_index+i].u
++ .valid.ref);
++ goto undo_map_out;
++ }
++
++ /* Record the grant handle, for use in the unmap
++ * operation.
++ */
++ private_data->grants[slot_index+i].u.
++ valid.user_handle = op.handle;
++
++ /* Update p2m structure with the new mapping. */
++ set_phys_to_machine(__pa(kernel_vaddr) >> PAGE_SHIFT,
++ FOREIGN_FRAME(private_data->
++ grants[slot_index+i]
++ .u.valid.dev_bus_addr
++ >> PAGE_SHIFT));
++ } else {
++ /* USING SHADOW PAGE TABLES. */
++ /* In this case, we simply insert the page into the VM
++ * area. */
++ ret = vm_insert_page(vma, user_vaddr, page);
++ }
++
++ }
++
++ up_write(&private_data->grants_sem);
++ return 0;
++
++undo_map_out:
++ /* If we have a mapping failure, the unmapping will be taken care of
++ * by do_mmap_pgoff(), which will eventually call gntdev_clear_pte().
++ * All we need to do here is free the vma_private_data.
++ */
++ kfree(vma->vm_private_data);
++
++ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
++ * to NULL on failure. However, we need this in gntdev_clear_pte() to
++ * unmap the grants. Therefore, we smuggle a reference to the file's
++ * private data in the VM area's private data pointer.
++ */
++ vma->vm_private_data = private_data;
++
++ up_write(&private_data->grants_sem);
++
++ return -ENOMEM;
++}
++
++static pte_t gntdev_clear_pte(struct vm_area_struct *vma, unsigned long addr,
++ pte_t *ptep, int is_fullmm)
++{
++ int slot_index, ret;
++ pte_t copy;
++ struct gnttab_unmap_grant_ref op;
++ gntdev_file_private_data_t *private_data;
++
++ /* THIS IS VERY UNPLEASANT: do_mmap_pgoff() will set the vma->vm_file
++ * to NULL on failure. However, we need this in gntdev_clear_pte() to
++ * unmap the grants. Therefore, we smuggle a reference to the file's
++ * private data in the VM area's private data pointer.
++ */
++ if (vma->vm_file) {
++ private_data = (gntdev_file_private_data_t *)
++ vma->vm_file->private_data;
++ } else if (vma->vm_private_data) {
++ private_data = (gntdev_file_private_data_t *)
++ vma->vm_private_data;
++ } else {
++ private_data = NULL; /* gcc warning */
++ BUG();
++ }
++
++ /* Calculate the grant relating to this PTE. */
++ slot_index = vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT);
++
++ /* Only unmap grants if the slot has been mapped. This could be being
++ * called from a failing mmap().
++ */
++ if (private_data->grants[slot_index].state == GNTDEV_SLOT_MAPPED) {
++
++ /* First, we clear the user space mapping, if it has been made.
++ */
++ if (private_data->grants[slot_index].u.valid.user_handle !=
++ GNTDEV_INVALID_HANDLE &&
++ !xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* NOT USING SHADOW PAGE TABLES. */
++
++ /* Copy the existing value of the PTE for returning. */
++ copy = *ptep;
++
++ gnttab_set_unmap_op(&op, virt_to_machine(ptep),
++ GNTMAP_contains_pte,
++ private_data->grants[slot_index]
++ .u.valid.user_handle);
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, &op, 1);
++ BUG_ON(ret);
++ if (op.status)
++ printk("User unmap grant status = %d\n",
++ op.status);
++ } else {
++ /* USING SHADOW PAGE TABLES. */
++ copy = xen_ptep_get_and_clear_full(vma, addr, ptep, is_fullmm);
++ }
++
++ /* Finally, we unmap the grant from kernel space. */
++ gnttab_set_unmap_op(&op,
++ get_kernel_vaddr(private_data, slot_index),
++ GNTMAP_host_map,
++ private_data->grants[slot_index].u.valid
++ .kernel_handle);
++ ret = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref,
++ &op, 1);
++ BUG_ON(ret);
++ if (op.status)
++ printk("Kernel unmap grant status = %d\n", op.status);
++
++
++ /* Return slot to the not-yet-mapped state, so that it may be
++ * mapped again, or removed by a subsequent ioctl.
++ */
++ private_data->grants[slot_index].state =
++ GNTDEV_SLOT_NOT_YET_MAPPED;
++
++ /* Invalidate the physical to machine mapping for this page. */
++ set_phys_to_machine(__pa(get_kernel_vaddr(private_data,
++ slot_index))
++ >> PAGE_SHIFT, INVALID_P2M_ENTRY);
++
++ } else {
++ copy = xen_ptep_get_and_clear_full(vma, addr, ptep, is_fullmm);
++ }
++
++ return copy;
++}
++
++/* "Destructor" for a VM area.
++ */
++static void gntdev_vma_close(struct vm_area_struct *vma) {
++ if (vma->vm_private_data) {
++ kfree(vma->vm_private_data);
++ }
++}
++
++/* Called when an ioctl is made on the device.
++ */
++static long gntdev_ioctl(struct file *flip,
++ unsigned int cmd, unsigned long arg)
++{
++ int rc = 0;
++ gntdev_file_private_data_t *private_data =
++ (gntdev_file_private_data_t *) flip->private_data;
++
++ /* On the first invocation, we will lazily initialise the grant array
++ * and free-list.
++ */
++ if (unlikely(!private_data->grants)
++ && likely(cmd != IOCTL_GNTDEV_SET_MAX_GRANTS)) {
++ down_write(&private_data->grants_sem);
++
++ if (unlikely(private_data->grants)) {
++ up_write(&private_data->grants_sem);
++ goto private_data_initialised;
++ }
++
++ /* Just use the default. Setting to a non-default is handled
++ * in the ioctl switch.
++ */
++ rc = init_private_data(private_data, DEFAULT_MAX_GRANTS);
++
++ up_write(&private_data->grants_sem);
++
++ if (rc) {
++ printk (KERN_ERR "Initialising gntdev private data "
++ "failed.\n");
++ return rc;
++ }
++ }
++
++private_data_initialised:
++ switch (cmd) {
++ case IOCTL_GNTDEV_MAP_GRANT_REF:
++ {
++ struct ioctl_gntdev_map_grant_ref op;
++ down_write(&private_data->grants_sem);
++ down_write(&private_data->free_list_sem);
++
++ if ((rc = copy_from_user(&op, (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto map_out;
++ }
++ if (unlikely(op.count <= 0)) {
++ rc = -EINVAL;
++ goto map_out;
++ }
++
++ if (op.count == 1) {
++ if ((rc = add_grant_reference(flip, &op.refs[0],
++ &op.index)) < 0) {
++ printk(KERN_ERR "Adding grant reference "
++ "failed (%d).\n", rc);
++ goto map_out;
++ }
++ } else {
++ struct ioctl_gntdev_grant_ref *refs, *u;
++ refs = kmalloc(op.count * sizeof(*refs), GFP_KERNEL);
++ if (!refs) {
++ rc = -ENOMEM;
++ goto map_out;
++ }
++ u = ((struct ioctl_gntdev_map_grant_ref *)arg)->refs;
++ if ((rc = copy_from_user(refs,
++ (void __user *)u,
++ sizeof(*refs) * op.count))) {
++ printk(KERN_ERR "Copying refs from user failed"
++ " (%d).\n", rc);
++ rc = -EINVAL;
++ goto map_out;
++ }
++ if ((rc = find_contiguous_free_range(flip, op.count))
++ < 0) {
++ printk(KERN_ERR "Finding contiguous range "
++ "failed (%d).\n", rc);
++ kfree(refs);
++ goto map_out;
++ }
++ op.index = rc << PAGE_SHIFT;
++ if ((rc = add_grant_references(flip, op.count,
++ refs, rc))) {
++ printk(KERN_ERR "Adding grant references "
++ "failed (%d).\n", rc);
++ kfree(refs);
++ goto map_out;
++ }
++ compress_free_list(flip);
++ kfree(refs);
++ }
++ if ((rc = copy_to_user((void __user *) arg,
++ &op,
++ sizeof(op)))) {
++ printk(KERN_ERR "Copying result back to user failed "
++ "(%d)\n", rc);
++ rc = -EFAULT;
++ goto map_out;
++ }
++ map_out:
++ up_write(&private_data->grants_sem);
++ up_write(&private_data->free_list_sem);
++ return rc;
++ }
++ case IOCTL_GNTDEV_UNMAP_GRANT_REF:
++ {
++ struct ioctl_gntdev_unmap_grant_ref op;
++ int i, start_index;
++
++ down_write(&private_data->grants_sem);
++ down_write(&private_data->free_list_sem);
++
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto unmap_out;
++ }
++
++ start_index = op.index >> PAGE_SHIFT;
++
++ /* First, check that all pages are in the NOT_YET_MAPPED
++ * state.
++ */
++ for (i = 0; i < op.count; ++i) {
++ if (unlikely
++ (private_data->grants[start_index + i].state
++ != GNTDEV_SLOT_NOT_YET_MAPPED)) {
++ if (private_data->grants[start_index + i].state
++ == GNTDEV_SLOT_INVALID) {
++ printk(KERN_ERR
++ "Tried to remove an invalid "
++ "grant at offset 0x%x.",
++ (start_index + i)
++ << PAGE_SHIFT);
++ rc = -EINVAL;
++ } else {
++ printk(KERN_ERR
++ "Tried to remove a grant which "
++ "is currently mmap()-ed at "
++ "offset 0x%x.",
++ (start_index + i)
++ << PAGE_SHIFT);
++ rc = -EBUSY;
++ }
++ goto unmap_out;
++ }
++ }
++
++ /* Unmap pages and add them to the free list.
++ */
++ for (i = 0; i < op.count; ++i) {
++ private_data->grants[start_index+i].state =
++ GNTDEV_SLOT_INVALID;
++ private_data->grants[start_index+i].u.free_list_index =
++ private_data->free_list_size;
++ private_data->free_list[private_data->free_list_size] =
++ start_index + i;
++ ++private_data->free_list_size;
++ }
++
++ unmap_out:
++ up_write(&private_data->grants_sem);
++ up_write(&private_data->free_list_sem);
++ return rc;
++ }
++ case IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR:
++ {
++ struct ioctl_gntdev_get_offset_for_vaddr op;
++ struct vm_area_struct *vma;
++ unsigned long vaddr;
++
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto get_offset_out;
++ }
++ vaddr = (unsigned long)op.vaddr;
++
++ down_read(¤t->mm->mmap_sem);
++ vma = find_vma(current->mm, vaddr);
++ if (vma == NULL) {
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ if ((!vma->vm_ops) || (vma->vm_ops != &gntdev_vmops)) {
++ printk(KERN_ERR "The vaddr specified does not belong "
++ "to a gntdev instance: %#lx\n", vaddr);
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ if (vma->vm_start != vaddr) {
++ printk(KERN_ERR "The vaddr specified in an "
++ "IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR must be at "
++ "the start of the VM area. vma->vm_start = "
++ "%#lx; vaddr = %#lx\n",
++ vma->vm_start, vaddr);
++ rc = -EFAULT;
++ goto get_offset_unlock_out;
++ }
++ op.offset = vma->vm_pgoff << PAGE_SHIFT;
++ op.count = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
++ up_read(¤t->mm->mmap_sem);
++ if ((rc = copy_to_user((void __user *) arg,
++ &op,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto get_offset_out;
++ }
++ goto get_offset_out;
++ get_offset_unlock_out:
++ up_read(¤t->mm->mmap_sem);
++ get_offset_out:
++ return rc;
++ }
++ case IOCTL_GNTDEV_SET_MAX_GRANTS:
++ {
++ struct ioctl_gntdev_set_max_grants op;
++ if ((rc = copy_from_user(&op,
++ (void __user *) arg,
++ sizeof(op)))) {
++ rc = -EFAULT;
++ goto set_max_out;
++ }
++ down_write(&private_data->grants_sem);
++ if (private_data->grants) {
++ rc = -EBUSY;
++ goto set_max_unlock_out;
++ }
++ if (op.count > MAX_GRANTS_LIMIT) {
++ rc = -EINVAL;
++ goto set_max_unlock_out;
++ }
++ rc = init_private_data(private_data, op.count);
++ set_max_unlock_out:
++ up_write(&private_data->grants_sem);
++ set_max_out:
++ return rc;
++ }
++ default:
++ return -ENOIOCTLCMD;
++ }
++
++ return 0;
++}
+diff --git a/drivers/xen/netback/Makefile b/drivers/xen/netback/Makefile
+new file mode 100644
+index 0000000..2bb2677
+--- /dev/null
++++ b/drivers/xen/netback/Makefile
+@@ -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 accel.o
++netloop-y := loopback.o
+diff --git a/drivers/xen/netback/accel.c b/drivers/xen/netback/accel.c
+new file mode 100644
+index 0000000..d3c6ece
+--- /dev/null
++++ b/drivers/xen/netback/accel.c
+@@ -0,0 +1,269 @@
++/******************************************************************************
++ * drivers/xen/netback/accel.c
++ *
++ * Interface between backend virtual network device and accelerated plugin.
++ *
++ * Copyright (C) 2007 Solarflare Communications, Inc
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/list.h>
++#include <asm/atomic.h>
++#include <xen/xenbus.h>
++#include <linux/mutex.h>
++
++#include "common.h"
++
++#if 0
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ printk("netback/accel (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
++#endif
++
++/*
++ * A list of available netback accelerator plugin modules (each list
++ * entry is of type struct netback_accelerator)
++ */
++static struct list_head accelerators_list;
++/* Lock used to protect access to accelerators_list */
++DEFINE_MUTEX(accelerators_mutex);
++
++/*
++ * Compare a backend to an accelerator, and decide if they are
++ * compatible (i.e. if the accelerator should be used by the
++ * backend)
++ */
++static int match_accelerator(struct xenbus_device *xendev,
++ struct backend_info *be,
++ struct netback_accelerator *accelerator)
++{
++ int rc = 0;
++ char *eth_name = xenbus_read(XBT_NIL, xendev->nodename, "accel", NULL);
++
++ if (IS_ERR(eth_name)) {
++ /* Probably means not present */
++ DPRINTK("%s: no match due to xenbus_read accel error %ld\n",
++ __FUNCTION__, PTR_ERR(eth_name));
++ return 0;
++ } else {
++ if (!strcmp(eth_name, accelerator->eth_name))
++ rc = 1;
++ kfree(eth_name);
++ return rc;
++ }
++}
++
++
++static void do_probe(struct backend_info *be,
++ struct netback_accelerator *accelerator,
++ struct xenbus_device *xendev)
++{
++ be->accelerator = accelerator;
++ atomic_inc(&be->accelerator->use_count);
++ if (be->accelerator->hooks->probe(xendev) != 0) {
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
++ }
++}
++
++
++/*
++ * Notify suitable backends that a new accelerator is available and
++ * connected. This will also notify the accelerator plugin module
++ * that it is being used for a device through the probe hook.
++ */
++static int netback_accelerator_probe_backend(struct device *dev, void *arg)
++{
++ struct netback_accelerator *accelerator =
++ (struct netback_accelerator *)arg;
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++
++ if (!strcmp("vif", xendev->devicetype)) {
++ struct backend_info *be = xendev->dev.driver_data;
++
++ if (match_accelerator(xendev, be, accelerator) &&
++ try_module_get(accelerator->hooks->owner)) {
++ do_probe(be, accelerator, xendev);
++ }
++ }
++ return 0;
++}
++
++
++/*
++ * Notify suitable backends that an accelerator is unavailable.
++ */
++static int netback_accelerator_remove_backend(struct device *dev, void *arg)
++{
++ struct xenbus_device *xendev = to_xenbus_device(dev);
++ struct netback_accelerator *accelerator =
++ (struct netback_accelerator *)arg;
++
++ if (!strcmp("vif", xendev->devicetype)) {
++ struct backend_info *be = xendev->dev.driver_data;
++
++ if (be->accelerator == accelerator) {
++ be->accelerator->hooks->remove(xendev);
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
++ }
++ }
++ return 0;
++}
++
++
++
++/*
++ * Entry point for an netback accelerator plugin module. Called to
++ * advertise its presence, and connect to any suitable backends.
++ */
++int netback_connect_accelerator(unsigned version, int id, const char *eth_name,
++ struct netback_accel_hooks *hooks)
++{
++ struct netback_accelerator *new_accelerator;
++ unsigned eth_name_len;
++
++ if (version != NETBACK_ACCEL_VERSION) {
++ if (version > NETBACK_ACCEL_VERSION) {
++ /* Caller has higher version number, leave it
++ up to them to decide whether to continue.
++ They can recall with a lower number if
++ they're happy to be compatible with us */
++ return NETBACK_ACCEL_VERSION;
++ } else {
++ /* We have a more recent version than caller.
++ Currently reject, but may in future be able
++ to be backwardly compatible */
++ return -EPROTO;
++ }
++ }
++
++ new_accelerator =
++ kmalloc(sizeof(struct netback_accelerator), GFP_KERNEL);
++ if (!new_accelerator) {
++ DPRINTK("%s: failed to allocate memory for accelerator\n",
++ __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ new_accelerator->id = id;
++
++ eth_name_len = strlen(eth_name)+1;
++ new_accelerator->eth_name = kmalloc(eth_name_len, GFP_KERNEL);
++ if (!new_accelerator->eth_name) {
++ DPRINTK("%s: failed to allocate memory for eth_name string\n",
++ __FUNCTION__);
++ kfree(new_accelerator);
++ return -ENOMEM;
++ }
++ strlcpy(new_accelerator->eth_name, eth_name, eth_name_len);
++
++ new_accelerator->hooks = hooks;
++
++ atomic_set(&new_accelerator->use_count, 0);
++
++ mutex_lock(&accelerators_mutex);
++ list_add(&new_accelerator->link, &accelerators_list);
++
++ /* tell existing backends about new plugin */
++ xenbus_for_each_backend(new_accelerator,
++ netback_accelerator_probe_backend);
++
++ mutex_unlock(&accelerators_mutex);
++
++ return 0;
++
++}
++EXPORT_SYMBOL_GPL(netback_connect_accelerator);
++
++
++/*
++ * Disconnect an accelerator plugin module that has previously been
++ * connected.
++ */
++void netback_disconnect_accelerator(int id, const char *eth_name)
++{
++ struct netback_accelerator *accelerator, *next;
++
++ mutex_lock(&accelerators_mutex);
++ list_for_each_entry_safe(accelerator, next, &accelerators_list, link) {
++ if (!strcmp(eth_name, accelerator->eth_name)) {
++ xenbus_for_each_backend
++ (accelerator, netback_accelerator_remove_backend);
++ BUG_ON(atomic_read(&accelerator->use_count) != 0);
++ list_del(&accelerator->link);
++ kfree(accelerator->eth_name);
++ kfree(accelerator);
++ break;
++ }
++ }
++ mutex_unlock(&accelerators_mutex);
++}
++EXPORT_SYMBOL_GPL(netback_disconnect_accelerator);
++
++
++void netback_probe_accelerators(struct backend_info *be,
++ struct xenbus_device *dev)
++{
++ struct netback_accelerator *accelerator;
++
++ /*
++ * Check list of accelerators to see if any is suitable, and
++ * use it if it is.
++ */
++ mutex_lock(&accelerators_mutex);
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(dev, be, accelerator) &&
++ try_module_get(accelerator->hooks->owner)) {
++ do_probe(be, accelerator, dev);
++ break;
++ }
++ }
++ mutex_unlock(&accelerators_mutex);
++}
++
++
++void netback_remove_accelerators(struct backend_info *be,
++ struct xenbus_device *dev)
++{
++ mutex_lock(&accelerators_mutex);
++ /* Notify the accelerator (if any) of this device's removal */
++ if (be->accelerator != NULL) {
++ be->accelerator->hooks->remove(dev);
++ atomic_dec(&be->accelerator->use_count);
++ module_put(be->accelerator->hooks->owner);
++ be->accelerator = NULL;
++ }
++ mutex_unlock(&accelerators_mutex);
++}
++
++
++void netif_accel_init(void)
++{
++ INIT_LIST_HEAD(&accelerators_list);
++}
+diff --git a/drivers/xen/netback/common.h b/drivers/xen/netback/common.h
+new file mode 100644
+index 0000000..525ac84
+--- /dev/null
++++ b/drivers/xen/netback/common.h
+@@ -0,0 +1,217 @@
++/******************************************************************************
++ * arch/xen/drivers/netif/backend/common.h
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __NETIF__BACKEND__COMMON_H__
++#define __NETIF__BACKEND__COMMON_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/ip.h>
++#include <linux/in.h>
++#include <linux/netdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/wait.h>
++#include <xen/evtchn.h>
++#include <xen/interface/io/netif.h>
++#include <asm/io.h>
++#include <asm/pgalloc.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "xen_net: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "xen_net: " fmt, ##args)
++
++typedef struct netif_st {
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++
++ u8 fe_dev_addr[6];
++
++ /* Physical parameters of the comms window. */
++ grant_handle_t tx_shmem_handle;
++ grant_ref_t tx_shmem_ref;
++ grant_handle_t rx_shmem_handle;
++ grant_ref_t rx_shmem_ref;
++ unsigned int irq;
++
++ /* The shared rings and indexes. */
++ netif_tx_back_ring_t tx;
++ netif_rx_back_ring_t rx;
++ struct vm_struct *tx_comms_area;
++ struct vm_struct *rx_comms_area;
++
++ /* Set of features that can be turned on in dev->features. */
++ int features;
++
++ /* Internal feature information. */
++ u8 can_queue:1; /* can queue packets for receiver? */
++ u8 copying_receiver:1; /* copy packets to receiver? */
++
++ /* Allow netif_be_start_xmit() to peek ahead in the rx request ring. */
++ RING_IDX rx_req_cons_peek;
++
++ /* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
++ unsigned long credit_bytes;
++ unsigned long credit_usec;
++ unsigned long remaining_credit;
++ struct timer_list credit_timeout;
++
++ /* Enforce draining of the transmit queue. */
++ struct timer_list tx_queue_timeout;
++
++ /* Miscellaneous private stuff. */
++ struct list_head list; /* scheduling list */
++ atomic_t refcnt;
++ struct net_device *dev;
++ struct net_device_stats stats;
++
++ unsigned int carrier;
++
++ wait_queue_head_t waiting_to_free;
++} netif_t;
++
++/*
++ * Implement our own carrier flag: the network stack's version causes delays
++ * when the carrier is re-enabled (in particular, dev_activate() may not
++ * immediately be called, which can cause packet loss; also the etherbridge
++ * can be rather lazy in activating its port).
++ */
++#define netback_carrier_on(netif) ((netif)->carrier = 1)
++#define netback_carrier_off(netif) ((netif)->carrier = 0)
++#define netback_carrier_ok(netif) ((netif)->carrier)
++
++enum {
++ NETBK_DONT_COPY_SKB,
++ NETBK_DELAYED_COPY_SKB,
++ NETBK_ALWAYS_COPY_SKB,
++};
++
++extern int netbk_copy_skb_mode;
++
++/* Function pointers into netback accelerator plugin modules */
++struct netback_accel_hooks {
++ struct module *owner;
++ int (*probe)(struct xenbus_device *dev);
++ int (*remove)(struct xenbus_device *dev);
++};
++
++/* Structure to track the state of a netback accelerator plugin */
++struct netback_accelerator {
++ struct list_head link;
++ int id;
++ char *eth_name;
++ atomic_t use_count;
++ struct netback_accel_hooks *hooks;
++};
++
++struct backend_info {
++ struct xenbus_device *dev;
++ netif_t *netif;
++ enum xenbus_state frontend_state;
++
++ /* State relating to the netback accelerator */
++ void *netback_accel_priv;
++ /* The accelerator that this backend is currently using */
++ struct netback_accelerator *accelerator;
++};
++
++#define NETBACK_ACCEL_VERSION 0x00010001
++
++/*
++ * Connect an accelerator plugin module to netback. Returns zero on
++ * success, < 0 on error, > 0 (with highest version number supported)
++ * if version mismatch.
++ */
++extern int netback_connect_accelerator(unsigned version,
++ int id, const char *eth_name,
++ struct netback_accel_hooks *hooks);
++/* Disconnect a previously connected accelerator plugin module */
++extern void netback_disconnect_accelerator(int id, const char *eth_name);
++
++
++extern
++void netback_probe_accelerators(struct backend_info *be,
++ struct xenbus_device *dev);
++extern
++void netback_remove_accelerators(struct backend_info *be,
++ struct xenbus_device *dev);
++extern
++void netif_accel_init(void);
++
++
++#define NET_TX_RING_SIZE __RING_SIZE((netif_tx_sring_t *)0, PAGE_SIZE)
++#define NET_RX_RING_SIZE __RING_SIZE((netif_rx_sring_t *)0, PAGE_SIZE)
++
++void netif_disconnect(netif_t *netif);
++
++netif_t *netif_alloc(domid_t domid, unsigned int handle);
++int netif_map(netif_t *netif, unsigned long tx_ring_ref,
++ unsigned long rx_ring_ref, unsigned int evtchn);
++
++#define netif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define netif_put(_b) \
++ do { \
++ if ( atomic_dec_and_test(&(_b)->refcnt) ) \
++ wake_up(&(_b)->waiting_to_free); \
++ } while (0)
++
++void netif_xenbus_init(void);
++
++#define netif_schedulable(netif) \
++ (netif_running((netif)->dev) && netback_carrier_ok(netif))
++
++void netif_schedule_work(netif_t *netif);
++void netif_deschedule_work(netif_t *netif);
++
++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);
++
++static inline int netbk_can_queue(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ return netif->can_queue;
++}
++
++static inline int netbk_can_sg(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ return netif->features & NETIF_F_SG;
++}
++
++#endif /* __NETIF__BACKEND__COMMON_H__ */
+diff --git a/drivers/xen/netback/interface.c b/drivers/xen/netback/interface.c
+new file mode 100644
+index 0000000..610891f
+--- /dev/null
++++ b/drivers/xen/netback/interface.c
+@@ -0,0 +1,336 @@
++/******************************************************************************
++ * arch/xen/drivers/netif/backend/interface.c
++ *
++ * Network-device interface management.
++ *
++ * Copyright (c) 2004-2005, Keir Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "common.h"
++#include <linux/ethtool.h>
++#include <linux/rtnetlink.h>
++
++/*
++ * Module parameter 'queue_length':
++ *
++ * Enables queuing in the network stack when a client has run out of receive
++ * descriptors. Although this feature can improve receive bandwidth by avoiding
++ * packet loss, it can also result in packets sitting in the 'tx_queue' for
++ * unbounded time. This is bad if those packets hold onto foreign resources.
++ * For example, consider a packet that holds onto resources belonging to the
++ * guest for which it is queued (e.g., packet received on vif1.0, destined for
++ * vif1.1 which is not activated in the guest): in this situation the guest
++ * will never be destroyed, unless vif1.1 is taken down. To avoid this, we
++ * run a timer (tx_queue_timeout) to drain the queue when the interface is
++ * blocked.
++ */
++static unsigned long netbk_queue_length = 32;
++module_param_named(queue_length, netbk_queue_length, ulong, 0);
++
++static void __netif_up(netif_t *netif)
++{
++ enable_irq(netif->irq);
++ netif_schedule_work(netif);
++}
++
++static void __netif_down(netif_t *netif)
++{
++ disable_irq(netif->irq);
++ netif_deschedule_work(netif);
++}
++
++static int net_open(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ if (netback_carrier_ok(netif)) {
++ __netif_up(netif);
++ netif_start_queue(dev);
++ }
++ return 0;
++}
++
++static int net_close(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ if (netback_carrier_ok(netif))
++ __netif_down(netif);
++ netif_stop_queue(dev);
++ return 0;
++}
++
++static int netbk_change_mtu(struct net_device *dev, int mtu)
++{
++ int max = netbk_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
++
++ if (mtu > max)
++ return -EINVAL;
++ dev->mtu = mtu;
++ return 0;
++}
++
++static int netbk_set_sg(struct net_device *dev, u32 data)
++{
++ if (data) {
++ netif_t *netif = netdev_priv(dev);
++
++ if (!(netif->features & NETIF_F_SG))
++ return -ENOSYS;
++ }
++
++ return ethtool_op_set_sg(dev, data);
++}
++
++static int netbk_set_tso(struct net_device *dev, u32 data)
++{
++ if (data) {
++ netif_t *netif = netdev_priv(dev);
++
++ if (!(netif->features & NETIF_F_TSO))
++ return -ENOSYS;
++ }
++
++ return ethtool_op_set_tso(dev, data);
++}
++
++static struct ethtool_ops network_ethtool_ops =
++{
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = netbk_set_sg,
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = netbk_set_tso,
++ .get_link = ethtool_op_get_link,
++};
++
++netif_t *netif_alloc(domid_t domid, unsigned int handle)
++{
++ int err = 0;
++ struct net_device *dev;
++ netif_t *netif;
++ char name[IFNAMSIZ] = {};
++
++ snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle);
++ dev = alloc_netdev(sizeof(netif_t), name, ether_setup);
++ if (dev == NULL) {
++ DPRINTK("Could not create netif: out of memory\n");
++ return ERR_PTR(-ENOMEM);
++ }
++
++ netif = netdev_priv(dev);
++ memset(netif, 0, sizeof(*netif));
++ netif->domid = domid;
++ netif->handle = handle;
++ atomic_set(&netif->refcnt, 1);
++ init_waitqueue_head(&netif->waiting_to_free);
++ netif->dev = dev;
++
++ netback_carrier_off(netif);
++
++ netif->credit_bytes = netif->remaining_credit = ~0UL;
++ netif->credit_usec = 0UL;
++ init_timer(&netif->credit_timeout);
++ /* Initialize 'expires' now: it's used to track the credit window. */
++ netif->credit_timeout.expires = jiffies;
++
++ init_timer(&netif->tx_queue_timeout);
++
++ dev->hard_start_xmit = netif_be_start_xmit;
++ dev->get_stats = netif_be_get_stats;
++ dev->open = net_open;
++ dev->stop = net_close;
++ dev->change_mtu = netbk_change_mtu;
++ dev->features = NETIF_F_IP_CSUM;
++
++ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
++
++ dev->tx_queue_len = netbk_queue_length;
++
++ /*
++ * Initialise a dummy MAC address. We choose the numerically
++ * largest non-broadcast address to prevent the address getting
++ * stolen by an Ethernet bridge for STP purposes.
++ * (FE:FF:FF:FF:FF:FF)
++ */
++ memset(dev->dev_addr, 0xFF, ETH_ALEN);
++ dev->dev_addr[0] &= ~0x01;
++
++ rtnl_lock();
++ err = register_netdevice(dev);
++ rtnl_unlock();
++ if (err) {
++ DPRINTK("Could not register new net device %s: err=%d\n",
++ dev->name, err);
++ free_netdev(dev);
++ return ERR_PTR(err);
++ }
++
++ DPRINTK("Successfully created netif\n");
++ return netif;
++}
++
++static int map_frontend_pages(
++ netif_t *netif, grant_ref_t tx_ring_ref, grant_ref_t rx_ring_ref)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)netif->tx_comms_area->addr,
++ GNTMAP_host_map, tx_ring_ref, netif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Gnttab failure mapping tx_ring_ref!\n");
++ return op.status;
++ }
++
++ netif->tx_shmem_ref = tx_ring_ref;
++ netif->tx_shmem_handle = op.handle;
++
++ gnttab_set_map_op(&op, (unsigned long)netif->rx_comms_area->addr,
++ GNTMAP_host_map, rx_ring_ref, netif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Gnttab failure mapping rx_ring_ref!\n");
++ return op.status;
++ }
++
++ netif->rx_shmem_ref = rx_ring_ref;
++ netif->rx_shmem_handle = op.handle;
++
++ return 0;
++}
++
++static void unmap_frontend_pages(netif_t *netif)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)netif->tx_comms_area->addr,
++ GNTMAP_host_map, netif->tx_shmem_handle);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++
++ gnttab_set_unmap_op(&op, (unsigned long)netif->rx_comms_area->addr,
++ GNTMAP_host_map, netif->rx_shmem_handle);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
++
++int netif_map(netif_t *netif, unsigned long tx_ring_ref,
++ unsigned long rx_ring_ref, unsigned int evtchn)
++{
++ int err = -ENOMEM;
++ netif_tx_sring_t *txs;
++ netif_rx_sring_t *rxs;
++
++ /* Already connected through? */
++ if (netif->irq)
++ return 0;
++
++ netif->tx_comms_area = alloc_vm_area(PAGE_SIZE);
++ if (netif->tx_comms_area == NULL)
++ return -ENOMEM;
++ netif->rx_comms_area = alloc_vm_area(PAGE_SIZE);
++ if (netif->rx_comms_area == NULL)
++ goto err_rx;
++
++ err = map_frontend_pages(netif, tx_ring_ref, rx_ring_ref);
++ if (err)
++ goto err_map;
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ netif->domid, evtchn, netif_be_int, 0,
++ netif->dev->name, netif);
++ if (err < 0)
++ goto err_hypervisor;
++ netif->irq = err;
++ disable_irq(netif->irq);
++
++ txs = (netif_tx_sring_t *)netif->tx_comms_area->addr;
++ BACK_RING_INIT(&netif->tx, txs, PAGE_SIZE);
++
++ rxs = (netif_rx_sring_t *)
++ ((char *)netif->rx_comms_area->addr);
++ BACK_RING_INIT(&netif->rx, rxs, PAGE_SIZE);
++
++ netif->rx_req_cons_peek = 0;
++
++ netif_get(netif);
++
++ rtnl_lock();
++ netback_carrier_on(netif);
++ if (netif_running(netif->dev))
++ __netif_up(netif);
++ rtnl_unlock();
++
++ return 0;
++err_hypervisor:
++ unmap_frontend_pages(netif);
++err_map:
++ free_vm_area(netif->rx_comms_area);
++err_rx:
++ free_vm_area(netif->tx_comms_area);
++ return err;
++}
++
++void netif_disconnect(netif_t *netif)
++{
++ if (netback_carrier_ok(netif)) {
++ rtnl_lock();
++ netback_carrier_off(netif);
++ netif_carrier_off(netif->dev); /* discard queued packets */
++ if (netif_running(netif->dev))
++ __netif_down(netif);
++ rtnl_unlock();
++ netif_put(netif);
++ }
++
++ atomic_dec(&netif->refcnt);
++ wait_event(netif->waiting_to_free, atomic_read(&netif->refcnt) == 0);
++
++ del_timer_sync(&netif->credit_timeout);
++ del_timer_sync(&netif->tx_queue_timeout);
++
++ if (netif->irq)
++ unbind_from_irqhandler(netif->irq, netif);
++
++ unregister_netdev(netif->dev);
++
++ if (netif->tx.sring) {
++ unmap_frontend_pages(netif);
++ free_vm_area(netif->tx_comms_area);
++ free_vm_area(netif->rx_comms_area);
++ }
++
++ free_netdev(netif->dev);
++}
+diff --git a/drivers/xen/netback/loopback.c b/drivers/xen/netback/loopback.c
+new file mode 100644
+index 0000000..1f70cd8
+--- /dev/null
++++ b/drivers/xen/netback/loopback.c
+@@ -0,0 +1,325 @@
++/******************************************************************************
++ * netback/loopback.c
++ *
++ * A two-interface loopback device to emulate a local netfront-netback
++ * connection. This ensures that local packet delivery looks identical
++ * to inter-domain delivery. Most importantly, packets delivered locally
++ * originating from other domains will get *copied* when they traverse this
++ * driver. This prevents unbounded delays in socket-buffer queues from
++ * causing the netback driver to "seize up".
++ *
++ * This driver creates a symmetric pair of loopback interfaces with names
++ * vif0.0 and veth0. The intention is that 'vif0.0' is bound to an Ethernet
++ * bridge, just like a proper netback interface, while a local IP interface
++ * is configured on 'veth0'.
++ *
++ * As with a real netback interface, vif0.0 is configured with a suitable
++ * dummy MAC address. No default is provided for veth0: a reasonable strategy
++ * is to transfer eth0's MAC address to veth0, and give eth0 a dummy address
++ * (to avoid confusing the Etherbridge).
++ *
++ * Copyright (c) 2005 K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/module.h>
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/skbuff.h>
++#include <linux/ethtool.h>
++#include <net/dst.h>
++#include <net/xfrm.h> /* secpath_reset() */
++#include <asm/hypervisor.h> /* is_initial_xendomain() */
++#include <../net/core/kmap_skb.h> /* k{,un}map_skb_frag() */
++
++static int nloopbacks = -1;
++module_param(nloopbacks, int, 0);
++MODULE_PARM_DESC(nloopbacks, "Number of netback-loopback devices to create");
++
++struct net_private {
++ struct net_device *loopback_dev;
++ struct net_device_stats stats;
++};
++
++static int loopback_open(struct net_device *dev)
++{
++ struct net_private *np = netdev_priv(dev);
++ memset(&np->stats, 0, sizeof(np->stats));
++ netif_start_queue(dev);
++ return 0;
++}
++
++static int loopback_close(struct net_device *dev)
++{
++ netif_stop_queue(dev);
++ return 0;
++}
++
++#ifdef CONFIG_X86
++static int is_foreign(unsigned long pfn)
++{
++ /* NB. Play it safe for auto-translation mode. */
++ return (xen_feature(XENFEAT_auto_translated_physmap) ||
++ (phys_to_machine_mapping[pfn] & FOREIGN_FRAME_BIT));
++}
++#else
++/* How to detect a foreign mapping? Play it safe. */
++#define is_foreign(pfn) (1)
++#endif
++
++static int skb_remove_foreign_references(struct sk_buff *skb)
++{
++ struct page *page;
++ unsigned long pfn;
++ int i, off;
++ char *vaddr;
++
++ BUG_ON(skb_shinfo(skb)->frag_list);
++
++ if (skb_cloned(skb) &&
++ unlikely(pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
++ return 0;
++
++ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
++ pfn = page_to_pfn(skb_shinfo(skb)->frags[i].page);
++ if (!is_foreign(pfn))
++ continue;
++
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!page))
++ return 0;
++
++ vaddr = kmap_skb_frag(&skb_shinfo(skb)->frags[i]);
++ off = skb_shinfo(skb)->frags[i].page_offset;
++ memcpy(page_address(page) + off,
++ vaddr + off,
++ skb_shinfo(skb)->frags[i].size);
++ kunmap_skb_frag(vaddr);
++
++ put_page(skb_shinfo(skb)->frags[i].page);
++ skb_shinfo(skb)->frags[i].page = page;
++ }
++
++ return 1;
++}
++
++static int loopback_start_xmit(struct sk_buff *skb, struct net_device *dev)
++{
++ struct net_private *np = netdev_priv(dev);
++
++ if (!skb_remove_foreign_references(skb)) {
++ np->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
++ }
++
++ dst_release(skb->dst);
++ skb->dst = NULL;
++
++ skb_orphan(skb);
++
++ np->stats.tx_bytes += skb->len;
++ np->stats.tx_packets++;
++
++ /* Switch to loopback context. */
++ dev = np->loopback_dev;
++ np = netdev_priv(dev);
++
++ np->stats.rx_bytes += skb->len;
++ np->stats.rx_packets++;
++
++ if (skb->ip_summed == CHECKSUM_PARTIAL) {
++ /* Defer checksum calculation. */
++ skb->proto_csum_blank = 1;
++ /* Must be a local packet: assert its integrity. */
++ skb->proto_data_valid = 1;
++ }
++
++ skb->ip_summed = skb->proto_data_valid ?
++ CHECKSUM_UNNECESSARY : CHECKSUM_NONE;
++
++ skb->pkt_type = PACKET_HOST; /* overridden by eth_type_trans() */
++ skb->protocol = eth_type_trans(skb, dev);
++ skb->dev = dev;
++ dev->last_rx = jiffies;
++
++ /* Flush netfilter context: rx'ed skbuffs not expected to have any. */
++ nf_reset(skb);
++ secpath_reset(skb);
++
++ netif_rx(skb);
++
++ return 0;
++}
++
++static struct net_device_stats *loopback_get_stats(struct net_device *dev)
++{
++ struct net_private *np = netdev_priv(dev);
++ return &np->stats;
++}
++
++static struct ethtool_ops network_ethtool_ops =
++{
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = ethtool_op_set_sg,
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = ethtool_op_set_tso,
++ .get_link = ethtool_op_get_link,
++};
++
++/*
++ * Nothing to do here. Virtual interface is point-to-point and the
++ * physical interface is probably promiscuous anyway.
++ */
++static void loopback_set_multicast_list(struct net_device *dev)
++{
++}
++
++static void loopback_construct(struct net_device *dev, struct net_device *lo)
++{
++ struct net_private *np = netdev_priv(dev);
++
++ np->loopback_dev = lo;
++
++ dev->open = loopback_open;
++ dev->stop = loopback_close;
++ dev->hard_start_xmit = loopback_start_xmit;
++ dev->get_stats = loopback_get_stats;
++ dev->set_multicast_list = loopback_set_multicast_list;
++ dev->change_mtu = NULL; /* allow arbitrary mtu */
++
++ dev->tx_queue_len = 0;
++
++ dev->features = (NETIF_F_HIGHDMA |
++ NETIF_F_LLTX |
++ NETIF_F_TSO |
++ NETIF_F_SG |
++ NETIF_F_IP_CSUM);
++
++ SET_ETHTOOL_OPS(dev, &network_ethtool_ops);
++
++ /*
++ * We do not set a jumbo MTU on the interface. Otherwise the network
++ * stack will try to send large packets that will get dropped by the
++ * Ethernet bridge (unless the physical Ethernet interface is
++ * configured to transfer jumbo packets). If a larger MTU is desired
++ * then the system administrator can specify it using the 'ifconfig'
++ * command.
++ */
++ /*dev->mtu = 16*1024;*/
++}
++
++static int __init make_loopback(int i)
++{
++ struct net_device *dev1, *dev2;
++ char dev_name[IFNAMSIZ];
++ int err = -ENOMEM;
++
++ sprintf(dev_name, "vif0.%d", i);
++ dev1 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
++ if (!dev1)
++ return err;
++
++ sprintf(dev_name, "veth%d", i);
++ dev2 = alloc_netdev(sizeof(struct net_private), dev_name, ether_setup);
++ if (!dev2)
++ goto fail_netdev2;
++
++ loopback_construct(dev1, dev2);
++ loopback_construct(dev2, dev1);
++
++ /*
++ * Initialise a dummy MAC address for the 'dummy backend' interface. We
++ * choose the numerically largest non-broadcast address to prevent the
++ * address getting stolen by an Ethernet bridge for STP purposes.
++ */
++ memset(dev1->dev_addr, 0xFF, ETH_ALEN);
++ dev1->dev_addr[0] &= ~0x01;
++
++ if ((err = register_netdev(dev1)) != 0)
++ goto fail;
++
++ if ((err = register_netdev(dev2)) != 0) {
++ unregister_netdev(dev1);
++ goto fail;
++ }
++
++ return 0;
++
++ fail:
++ free_netdev(dev2);
++ fail_netdev2:
++ free_netdev(dev1);
++ return err;
++}
++
++static void __exit clean_loopback(int i)
++{
++ struct net_device *dev1, *dev2;
++ char dev_name[IFNAMSIZ];
++
++ sprintf(dev_name, "vif0.%d", i);
++ dev1 = dev_get_by_name(&init_net, dev_name);
++ sprintf(dev_name, "veth%d", i);
++ dev2 = dev_get_by_name(&init_net, dev_name);
++ if (dev1 && dev2) {
++ unregister_netdev(dev2);
++ unregister_netdev(dev1);
++ free_netdev(dev2);
++ free_netdev(dev1);
++ }
++}
++
++static int __init loopback_init(void)
++{
++ int i, err = 0;
++
++ if (nloopbacks == -1)
++ nloopbacks = is_initial_xendomain() ? 4 : 0;
++
++ for (i = 0; i < nloopbacks; i++)
++ if ((err = make_loopback(i)) != 0)
++ break;
++
++ return err;
++}
++
++module_init(loopback_init);
++
++static void __exit loopback_exit(void)
++{
++ int i;
++
++ for (i = nloopbacks; i-- > 0; )
++ clean_loopback(i);
++}
++
++module_exit(loopback_exit);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/netback/netback.c b/drivers/xen/netback/netback.c
+new file mode 100644
+index 0000000..e73df86
+--- /dev/null
++++ b/drivers/xen/netback/netback.c
+@@ -0,0 +1,1618 @@
++/******************************************************************************
++ * drivers/xen/netback/netback.c
++ *
++ * Back-end of the driver for virtual network devices. This portion of the
++ * driver exports a 'unified' network-device interface that can be accessed
++ * by any operating system that implements a compatible front end. A
++ * reference front-end implementation can be found in:
++ * drivers/xen/netfront/netfront.c
++ *
++ * Copyright (c) 2002-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include "common.h"
++#include <xen/balloon.h>
++#include <xen/interface/memory.h>
++
++/*define NETBE_DEBUG_INTERRUPT*/
++
++/* extra field used in struct page */
++#define netif_page_index(pg) (*(long *)&(pg)->mapping)
++
++struct netbk_rx_meta {
++ skb_frag_t frag;
++ int id;
++ u8 copy:1;
++};
++
++struct netbk_tx_pending_inuse {
++ struct list_head list;
++ unsigned long alloc_time;
++};
++
++static void netif_idx_release(u16 pending_idx);
++static void netif_page_release(struct page *page);
++static void make_tx_response(netif_t *netif,
++ netif_tx_request_t *txp,
++ s8 st);
++static netif_rx_response_t *make_rx_response(netif_t *netif,
++ u16 id,
++ s8 st,
++ u16 offset,
++ u16 size,
++ u16 flags);
++
++static void net_tx_action(unsigned long unused);
++static DECLARE_TASKLET(net_tx_tasklet, net_tx_action, 0);
++
++static void net_rx_action(unsigned long unused);
++static DECLARE_TASKLET(net_rx_tasklet, net_rx_action, 0);
++
++static struct timer_list net_timer;
++static struct timer_list netbk_tx_pending_timer;
++
++#define MAX_PENDING_REQS 256
++
++static struct sk_buff_head rx_queue;
++
++static struct page **mmap_pages;
++static inline unsigned long idx_to_pfn(unsigned int idx)
++{
++ return page_to_pfn(mmap_pages[idx]);
++}
++
++static inline unsigned long idx_to_kaddr(unsigned int idx)
++{
++ return (unsigned long)pfn_to_kaddr(idx_to_pfn(idx));
++}
++
++#define PKT_PROT_LEN 64
++
++static struct pending_tx_info {
++ netif_tx_request_t req;
++ netif_t *netif;
++} pending_tx_info[MAX_PENDING_REQS];
++static u16 pending_ring[MAX_PENDING_REQS];
++typedef unsigned int PEND_RING_IDX;
++#define MASK_PEND_IDX(_i) ((_i)&(MAX_PENDING_REQS-1))
++static PEND_RING_IDX pending_prod, pending_cons;
++#define NR_PENDING_REQS (MAX_PENDING_REQS - pending_prod + pending_cons)
++
++/* Freed TX SKBs get batched on this ring before return to pending_ring. */
++static u16 dealloc_ring[MAX_PENDING_REQS];
++static PEND_RING_IDX dealloc_prod, dealloc_cons;
++
++/* Doubly-linked list of in-use pending entries. */
++static struct netbk_tx_pending_inuse pending_inuse[MAX_PENDING_REQS];
++static LIST_HEAD(pending_inuse_head);
++
++static struct sk_buff_head tx_queue;
++
++static grant_handle_t grant_tx_handle[MAX_PENDING_REQS];
++static gnttab_unmap_grant_ref_t tx_unmap_ops[MAX_PENDING_REQS];
++static gnttab_map_grant_ref_t tx_map_ops[MAX_PENDING_REQS];
++
++static struct list_head net_schedule_list;
++static spinlock_t net_schedule_list_lock;
++
++#define MAX_MFN_ALLOC 64
++static unsigned long mfn_list[MAX_MFN_ALLOC];
++static unsigned int alloc_index = 0;
++
++/* Setting this allows the safe use of this driver without netloop. */
++static int MODPARM_copy_skb = 1;
++module_param_named(copy_skb, MODPARM_copy_skb, bool, 0);
++MODULE_PARM_DESC(copy_skb, "Copy data received from netfront without netloop");
++
++int netbk_copy_skb_mode;
++
++static inline unsigned long alloc_mfn(void)
++{
++ BUG_ON(alloc_index == 0);
++ return mfn_list[--alloc_index];
++}
++
++static int check_mfn(int nr)
++{
++ struct xen_memory_reservation reservation = {
++ .extent_order = 0,
++ .domid = DOMID_SELF
++ };
++ int rc;
++
++ if (likely(alloc_index >= nr))
++ return 0;
++
++ set_xen_guest_handle(reservation.extent_start, mfn_list + alloc_index);
++ reservation.nr_extents = MAX_MFN_ALLOC - alloc_index;
++ rc = HYPERVISOR_memory_op(XENMEM_increase_reservation, &reservation);
++ if (likely(rc > 0))
++ alloc_index += rc;
++
++ return alloc_index >= nr ? 0 : -ENOMEM;
++}
++
++static inline void maybe_schedule_tx_action(void)
++{
++ smp_mb();
++ if ((NR_PENDING_REQS < (MAX_PENDING_REQS/2)) &&
++ !list_empty(&net_schedule_list))
++ tasklet_schedule(&net_tx_tasklet);
++}
++
++static struct sk_buff *netbk_copy_skb(struct sk_buff *skb)
++{
++ struct skb_shared_info *ninfo;
++ struct sk_buff *nskb;
++ unsigned long offset;
++ int ret;
++ int len;
++ int headlen;
++
++ BUG_ON(skb_shinfo(skb)->frag_list != NULL);
++
++ nskb = alloc_skb(SKB_MAX_HEAD(0), GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!nskb))
++ goto err;
++
++ skb_reserve(nskb, 16 + NET_IP_ALIGN);
++ headlen = skb_end_pointer(nskb) - nskb->data;
++ if (headlen > skb_headlen(skb))
++ headlen = skb_headlen(skb);
++ ret = skb_copy_bits(skb, 0, __skb_put(nskb, headlen), headlen);
++ BUG_ON(ret);
++
++ ninfo = skb_shinfo(nskb);
++ ninfo->gso_size = skb_shinfo(skb)->gso_size;
++ ninfo->gso_type = skb_shinfo(skb)->gso_type;
++
++ offset = headlen;
++ len = skb->len - headlen;
++
++ nskb->len = skb->len;
++ nskb->data_len = len;
++ nskb->truesize += len;
++
++ while (len) {
++ struct page *page;
++ int copy;
++ int zero;
++
++ if (unlikely(ninfo->nr_frags >= MAX_SKB_FRAGS)) {
++ dump_stack();
++ goto err_free;
++ }
++
++ copy = len >= PAGE_SIZE ? PAGE_SIZE : len;
++ zero = len >= PAGE_SIZE ? 0 : __GFP_ZERO;
++
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN | zero);
++ if (unlikely(!page))
++ goto err_free;
++
++ ret = skb_copy_bits(skb, offset, page_address(page), copy);
++ BUG_ON(ret);
++
++ ninfo->frags[ninfo->nr_frags].page = page;
++ ninfo->frags[ninfo->nr_frags].page_offset = 0;
++ ninfo->frags[ninfo->nr_frags].size = copy;
++ ninfo->nr_frags++;
++
++ offset += copy;
++ len -= copy;
++ }
++
++#ifdef NET_SKBUFF_DATA_USES_OFFSET
++ offset = 0;
++#else
++ offset = nskb->data - skb->data;
++#endif
++
++ nskb->transport_header = skb->transport_header + offset;
++ nskb->network_header = skb->network_header + offset;
++ nskb->mac_header = skb->mac_header + offset;
++
++ return nskb;
++
++ err_free:
++ kfree_skb(nskb);
++ err:
++ return NULL;
++}
++
++static inline int netbk_max_required_rx_slots(netif_t *netif)
++{
++ if (netif->features & (NETIF_F_SG|NETIF_F_TSO))
++ return MAX_SKB_FRAGS + 2; /* header + extra_info + frags */
++ return 1; /* all in one */
++}
++
++static inline int netbk_queue_full(netif_t *netif)
++{
++ RING_IDX peek = netif->rx_req_cons_peek;
++ RING_IDX needed = netbk_max_required_rx_slots(netif);
++
++ return ((netif->rx.sring->req_prod - peek) < needed) ||
++ ((netif->rx.rsp_prod_pvt + NET_RX_RING_SIZE - peek) < needed);
++}
++
++static void tx_queue_callback(unsigned long data)
++{
++ netif_t *netif = (netif_t *)data;
++ if (netif_schedulable(netif))
++ netif_wake_queue(netif->dev);
++}
++
++int netif_be_start_xmit(struct sk_buff *skb, struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++
++ BUG_ON(skb->dev != dev);
++
++ /* Drop the packet if the target domain has no receive buffers. */
++ if (unlikely(!netif_schedulable(netif) || netbk_queue_full(netif)))
++ goto drop;
++
++ /*
++ * Copy the packet here if it's destined for a flipping interface
++ * but isn't flippable (e.g. extra references to data).
++ * XXX For now we also copy skbuffs whose head crosses a page
++ * boundary, because netbk_gop_skb can't handle them.
++ */
++ if (!netif->copying_receiver ||
++ ((skb_headlen(skb) + offset_in_page(skb->data)) >= PAGE_SIZE)) {
++ struct sk_buff *nskb = netbk_copy_skb(skb);
++ if ( unlikely(nskb == NULL) )
++ goto drop;
++ /* Copy only the header fields we use in this driver. */
++ nskb->dev = skb->dev;
++ nskb->ip_summed = skb->ip_summed;
++ nskb->proto_data_valid = skb->proto_data_valid;
++ dev_kfree_skb(skb);
++ skb = nskb;
++ }
++
++ netif->rx_req_cons_peek += skb_shinfo(skb)->nr_frags + 1 +
++ !!skb_shinfo(skb)->gso_size;
++ netif_get(netif);
++
++ if (netbk_can_queue(dev) && netbk_queue_full(netif)) {
++ netif->rx.sring->req_event = netif->rx_req_cons_peek +
++ netbk_max_required_rx_slots(netif);
++ mb(); /* request notification /then/ check & stop the queue */
++ if (netbk_queue_full(netif)) {
++ netif_stop_queue(dev);
++ /*
++ * Schedule 500ms timeout to restart the queue, thus
++ * ensuring that an inactive queue will be drained.
++ * Packets will be immediately be dropped until more
++ * receive buffers become available (see
++ * netbk_queue_full() check above).
++ */
++ netif->tx_queue_timeout.data = (unsigned long)netif;
++ netif->tx_queue_timeout.function = tx_queue_callback;
++ __mod_timer(&netif->tx_queue_timeout, jiffies + HZ/2);
++ }
++ }
++
++ skb_queue_tail(&rx_queue, skb);
++ tasklet_schedule(&net_rx_tasklet);
++
++ return 0;
++
++ drop:
++ netif->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
++}
++
++#if 0
++static void xen_network_done_notify(void)
++{
++ static struct net_device *eth0_dev = NULL;
++ if (unlikely(eth0_dev == NULL))
++ eth0_dev = __dev_get_by_name(&init_net, "eth0");
++ netif_rx_schedule(eth0_dev, ???);
++}
++/*
++ * Add following to poll() function in NAPI driver (Tigon3 is example):
++ * if ( xen_network_done() )
++ * tg3_enable_ints(tp);
++ */
++int xen_network_done(void)
++{
++ return skb_queue_empty(&rx_queue);
++}
++#endif
++
++struct netrx_pending_operations {
++ unsigned trans_prod, trans_cons;
++ unsigned mmu_prod, mmu_mcl;
++ unsigned mcl_prod, mcl_cons;
++ unsigned copy_prod, copy_cons;
++ unsigned meta_prod, meta_cons;
++ mmu_update_t *mmu;
++ gnttab_transfer_t *trans;
++ gnttab_copy_t *copy;
++ multicall_entry_t *mcl;
++ struct netbk_rx_meta *meta;
++};
++
++/* Set up the grant operations for this fragment. If it's a flipping
++ interface, we also set up the unmap request from here. */
++static u16 netbk_gop_frag(netif_t *netif, struct netbk_rx_meta *meta,
++ int i, struct netrx_pending_operations *npo,
++ struct page *page, unsigned long size,
++ unsigned long offset)
++{
++ mmu_update_t *mmu;
++ gnttab_transfer_t *gop;
++ gnttab_copy_t *copy_gop;
++ multicall_entry_t *mcl;
++ netif_rx_request_t *req;
++ unsigned long old_mfn, new_mfn;
++
++ old_mfn = virt_to_mfn(page_address(page));
++
++ req = RING_GET_REQUEST(&netif->rx, netif->rx.req_cons + i);
++ if (netif->copying_receiver) {
++ /* The fragment needs to be copied rather than
++ flipped. */
++ meta->copy = 1;
++ copy_gop = npo->copy + npo->copy_prod++;
++ copy_gop->flags = GNTCOPY_dest_gref;
++ if (PageForeign(page)) {
++ struct pending_tx_info *src_pend =
++ &pending_tx_info[netif_page_index(page)];
++ copy_gop->source.domid = src_pend->netif->domid;
++ copy_gop->source.u.ref = src_pend->req.gref;
++ copy_gop->flags |= GNTCOPY_source_gref;
++ } else {
++ copy_gop->source.domid = DOMID_SELF;
++ copy_gop->source.u.gmfn = old_mfn;
++ }
++ copy_gop->source.offset = offset;
++ copy_gop->dest.domid = netif->domid;
++ copy_gop->dest.offset = 0;
++ copy_gop->dest.u.ref = req->gref;
++ copy_gop->len = size;
++ } else {
++ meta->copy = 0;
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ new_mfn = alloc_mfn();
++
++ /*
++ * Set the new P2M table entry before
++ * reassigning the old data page. Heed the
++ * comment in pgtable-2level.h:pte_page(). :-)
++ */
++ set_phys_to_machine(page_to_pfn(page), new_mfn);
++
++ mcl = npo->mcl + npo->mcl_prod++;
++ MULTI_update_va_mapping(mcl,
++ (unsigned long)page_address(page),
++ pfn_pte_ma(new_mfn, PAGE_KERNEL),
++ 0);
++
++ mmu = npo->mmu + npo->mmu_prod++;
++ mmu->ptr = ((maddr_t)new_mfn << PAGE_SHIFT) |
++ MMU_MACHPHYS_UPDATE;
++ mmu->val = page_to_pfn(page);
++ }
++
++ gop = npo->trans + npo->trans_prod++;
++ gop->mfn = old_mfn;
++ gop->domid = netif->domid;
++ gop->ref = req->gref;
++ }
++ return req->id;
++}
++
++static void netbk_gop_skb(struct sk_buff *skb,
++ struct netrx_pending_operations *npo)
++{
++ netif_t *netif = netdev_priv(skb->dev);
++ int nr_frags = skb_shinfo(skb)->nr_frags;
++ int i;
++ int extra;
++ struct netbk_rx_meta *head_meta, *meta;
++
++ head_meta = npo->meta + npo->meta_prod++;
++ head_meta->frag.page_offset = skb_shinfo(skb)->gso_type;
++ head_meta->frag.size = skb_shinfo(skb)->gso_size;
++ extra = !!head_meta->frag.size + 1;
++
++ for (i = 0; i < nr_frags; i++) {
++ meta = npo->meta + npo->meta_prod++;
++ meta->frag = skb_shinfo(skb)->frags[i];
++ meta->id = netbk_gop_frag(netif, meta, i + extra, npo,
++ meta->frag.page,
++ meta->frag.size,
++ meta->frag.page_offset);
++ }
++
++ /*
++ * This must occur at the end to ensure that we don't trash skb_shinfo
++ * until we're done. We know that the head doesn't cross a page
++ * boundary because such packets get copied in netif_be_start_xmit.
++ */
++ head_meta->id = netbk_gop_frag(netif, head_meta, 0, npo,
++ virt_to_page(skb->data),
++ skb_headlen(skb),
++ offset_in_page(skb->data));
++
++ netif->rx.req_cons += nr_frags + extra;
++}
++
++static inline void netbk_free_pages(int nr_frags, struct netbk_rx_meta *meta)
++{
++ int i;
++
++ for (i = 0; i < nr_frags; i++)
++ put_page(meta[i].frag.page);
++}
++
++/* This is a twin to netbk_gop_skb. Assume that netbk_gop_skb was
++ used to set up the operations on the top of
++ netrx_pending_operations, which have since been done. Check that
++ they didn't give any errors and advance over them. */
++static int netbk_check_gop(int nr_frags, domid_t domid,
++ struct netrx_pending_operations *npo)
++{
++ multicall_entry_t *mcl;
++ gnttab_transfer_t *gop;
++ gnttab_copy_t *copy_op;
++ int status = NETIF_RSP_OKAY;
++ int i;
++
++ for (i = 0; i <= nr_frags; i++) {
++ if (npo->meta[npo->meta_cons + i].copy) {
++ copy_op = npo->copy + npo->copy_cons++;
++ if (copy_op->status != GNTST_okay) {
++ DPRINTK("Bad status %d from copy to DOM%d.\n",
++ copy_op->status, domid);
++ status = NETIF_RSP_ERROR;
++ }
++ } else {
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ mcl = npo->mcl + npo->mcl_cons++;
++ /* The update_va_mapping() must not fail. */
++ BUG_ON(mcl->result != 0);
++ }
++
++ gop = npo->trans + npo->trans_cons++;
++ /* Check the reassignment error code. */
++ if (gop->status != 0) {
++ DPRINTK("Bad status %d from grant transfer to DOM%u\n",
++ gop->status, domid);
++ /*
++ * Page no longer belongs to us unless
++ * GNTST_bad_page, but that should be
++ * a fatal error anyway.
++ */
++ BUG_ON(gop->status == GNTST_bad_page);
++ status = NETIF_RSP_ERROR;
++ }
++ }
++ }
++
++ return status;
++}
++
++static void netbk_add_frag_responses(netif_t *netif, int status,
++ struct netbk_rx_meta *meta, int nr_frags)
++{
++ int i;
++ unsigned long offset;
++
++ for (i = 0; i < nr_frags; i++) {
++ int id = meta[i].id;
++ int flags = (i == nr_frags - 1) ? 0 : NETRXF_more_data;
++
++ if (meta[i].copy)
++ offset = 0;
++ else
++ offset = meta[i].frag.page_offset;
++ make_rx_response(netif, id, status, offset,
++ meta[i].frag.size, flags);
++ }
++}
++
++static void net_rx_action(unsigned long unused)
++{
++ netif_t *netif = NULL;
++ s8 status;
++ u16 id, irq, flags;
++ netif_rx_response_t *resp;
++ multicall_entry_t *mcl;
++ struct sk_buff_head rxq;
++ struct sk_buff *skb;
++ int notify_nr = 0;
++ int ret;
++ int nr_frags;
++ int count;
++ unsigned long offset;
++
++ /*
++ * Putting hundreds of bytes on the stack is considered rude.
++ * Static works because a tasklet can only be on one CPU at any time.
++ */
++ static multicall_entry_t rx_mcl[NET_RX_RING_SIZE+3];
++ static mmu_update_t rx_mmu[NET_RX_RING_SIZE];
++ static gnttab_transfer_t grant_trans_op[NET_RX_RING_SIZE];
++ static gnttab_copy_t grant_copy_op[NET_RX_RING_SIZE];
++ static unsigned char rx_notify[NR_IRQS];
++ static u16 notify_list[NET_RX_RING_SIZE];
++ static struct netbk_rx_meta meta[NET_RX_RING_SIZE];
++
++ struct netrx_pending_operations npo = {
++ mmu: rx_mmu,
++ trans: grant_trans_op,
++ copy: grant_copy_op,
++ mcl: rx_mcl,
++ meta: meta};
++
++ skb_queue_head_init(&rxq);
++
++ count = 0;
++
++ while ((skb = skb_dequeue(&rx_queue)) != NULL) {
++ nr_frags = skb_shinfo(skb)->nr_frags;
++ *(int *)skb->cb = nr_frags;
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap) &&
++ !((netif_t *)netdev_priv(skb->dev))->copying_receiver &&
++ check_mfn(nr_frags + 1)) {
++ /* Memory squeeze? Back off for an arbitrary while. */
++ if ( net_ratelimit() )
++ WPRINTK("Memory squeeze in netback "
++ "driver.\n");
++ mod_timer(&net_timer, jiffies + HZ);
++ skb_queue_head(&rx_queue, skb);
++ break;
++ }
++
++ netbk_gop_skb(skb, &npo);
++
++ count += nr_frags + 1;
++
++ __skb_queue_tail(&rxq, skb);
++
++ /* Filled the batch queue? */
++ if (count + MAX_SKB_FRAGS >= NET_RX_RING_SIZE)
++ break;
++ }
++
++ BUG_ON(npo.meta_prod > ARRAY_SIZE(meta));
++
++ npo.mmu_mcl = npo.mcl_prod;
++ if (npo.mcl_prod) {
++ BUG_ON(xen_feature(XENFEAT_auto_translated_physmap));
++ BUG_ON(npo.mmu_prod > ARRAY_SIZE(rx_mmu));
++ mcl = npo.mcl + npo.mcl_prod++;
++
++ BUG_ON(mcl[-1].op != __HYPERVISOR_update_va_mapping);
++ mcl[-1].args[MULTI_UVMFLAGS_INDEX] = UVMF_TLB_FLUSH|UVMF_ALL;
++
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)rx_mmu;
++ mcl->args[1] = npo.mmu_prod;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
++ }
++
++ if (npo.trans_prod) {
++ BUG_ON(npo.trans_prod > ARRAY_SIZE(grant_trans_op));
++ mcl = npo.mcl + npo.mcl_prod++;
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = GNTTABOP_transfer;
++ mcl->args[1] = (unsigned long)grant_trans_op;
++ mcl->args[2] = npo.trans_prod;
++ }
++
++ if (npo.copy_prod) {
++ BUG_ON(npo.copy_prod > ARRAY_SIZE(grant_copy_op));
++ mcl = npo.mcl + npo.mcl_prod++;
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = GNTTABOP_copy;
++ mcl->args[1] = (unsigned long)grant_copy_op;
++ mcl->args[2] = npo.copy_prod;
++ }
++
++ /* Nothing to do? */
++ if (!npo.mcl_prod)
++ return;
++
++ BUG_ON(npo.mcl_prod > ARRAY_SIZE(rx_mcl));
++
++ ret = HYPERVISOR_multicall(npo.mcl, npo.mcl_prod);
++ BUG_ON(ret != 0);
++ /* The mmu_machphys_update() must not fail. */
++ BUG_ON(npo.mmu_mcl && npo.mcl[npo.mmu_mcl].result != 0);
++
++ while ((skb = __skb_dequeue(&rxq)) != NULL) {
++ nr_frags = *(int *)skb->cb;
++
++ netif = netdev_priv(skb->dev);
++ /* We can't rely on skb_release_data to release the
++ pages used by fragments for us, since it tries to
++ touch the pages in the fraglist. If we're in
++ flipping mode, that doesn't work. In copying mode,
++ we still have access to all of the pages, and so
++ it's safe to let release_data deal with it. */
++ /* (Freeing the fragments is safe since we copy
++ non-linear skbs destined for flipping interfaces) */
++ if (!netif->copying_receiver) {
++ atomic_set(&(skb_shinfo(skb)->dataref), 1);
++ skb_shinfo(skb)->frag_list = NULL;
++ skb_shinfo(skb)->nr_frags = 0;
++ netbk_free_pages(nr_frags, meta + npo.meta_cons + 1);
++ }
++
++ netif->stats.tx_bytes += skb->len;
++ netif->stats.tx_packets++;
++
++ status = netbk_check_gop(nr_frags, netif->domid, &npo);
++
++ id = meta[npo.meta_cons].id;
++ flags = nr_frags ? NETRXF_more_data : 0;
++
++ 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;
++
++ if (meta[npo.meta_cons].copy)
++ offset = 0;
++ else
++ offset = offset_in_page(skb->data);
++ resp = make_rx_response(netif, id, status, offset,
++ skb_headlen(skb), flags);
++
++ if (meta[npo.meta_cons].frag.size) {
++ struct netif_extra_info *gso =
++ (struct netif_extra_info *)
++ RING_GET_RESPONSE(&netif->rx,
++ netif->rx.rsp_prod_pvt++);
++
++ resp->flags |= NETRXF_extra_info;
++
++ gso->u.gso.size = meta[npo.meta_cons].frag.size;
++ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
++ gso->u.gso.pad = 0;
++ gso->u.gso.features = 0;
++
++ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
++ gso->flags = 0;
++ }
++
++ netbk_add_frag_responses(netif, status,
++ meta + npo.meta_cons + 1,
++ nr_frags);
++
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->rx, ret);
++ irq = netif->irq;
++ if (ret && !rx_notify[irq]) {
++ rx_notify[irq] = 1;
++ notify_list[notify_nr++] = irq;
++ }
++
++ if (netif_queue_stopped(netif->dev) &&
++ netif_schedulable(netif) &&
++ !netbk_queue_full(netif))
++ netif_wake_queue(netif->dev);
++
++ netif_put(netif);
++ dev_kfree_skb(skb);
++ npo.meta_cons += nr_frags + 1;
++ }
++
++ while (notify_nr != 0) {
++ irq = notify_list[--notify_nr];
++ rx_notify[irq] = 0;
++ notify_remote_via_irq(irq);
++ }
++
++ /* More work to do? */
++ if (!skb_queue_empty(&rx_queue) && !timer_pending(&net_timer))
++ tasklet_schedule(&net_rx_tasklet);
++#if 0
++ else
++ xen_network_done_notify();
++#endif
++}
++
++static void net_alarm(unsigned long unused)
++{
++ tasklet_schedule(&net_rx_tasklet);
++}
++
++static void netbk_tx_pending_timeout(unsigned long unused)
++{
++ tasklet_schedule(&net_tx_tasklet);
++}
++
++struct net_device_stats *netif_be_get_stats(struct net_device *dev)
++{
++ netif_t *netif = netdev_priv(dev);
++ return &netif->stats;
++}
++
++static int __on_net_schedule_list(netif_t *netif)
++{
++ return netif->list.next != NULL;
++}
++
++static void remove_from_net_schedule_list(netif_t *netif)
++{
++ spin_lock_irq(&net_schedule_list_lock);
++ if (likely(__on_net_schedule_list(netif))) {
++ list_del(&netif->list);
++ netif->list.next = NULL;
++ netif_put(netif);
++ }
++ spin_unlock_irq(&net_schedule_list_lock);
++}
++
++static void add_to_net_schedule_list_tail(netif_t *netif)
++{
++ if (__on_net_schedule_list(netif))
++ return;
++
++ spin_lock_irq(&net_schedule_list_lock);
++ if (!__on_net_schedule_list(netif) &&
++ likely(netif_schedulable(netif))) {
++ list_add_tail(&netif->list, &net_schedule_list);
++ netif_get(netif);
++ }
++ spin_unlock_irq(&net_schedule_list_lock);
++}
++
++/*
++ * Note on CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER:
++ * If this driver is pipelining transmit requests then we can be very
++ * aggressive in avoiding new-packet notifications -- frontend only needs to
++ * send a notification if there are no outstanding unreceived responses.
++ * If we may be buffer transmit buffers for any reason then we must be rather
++ * more conservative and treat this as the final check for pending work.
++ */
++void netif_schedule_work(netif_t *netif)
++{
++ int more_to_do;
++
++#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
++ more_to_do = RING_HAS_UNCONSUMED_REQUESTS(&netif->tx);
++#else
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
++#endif
++
++ if (more_to_do) {
++ add_to_net_schedule_list_tail(netif);
++ maybe_schedule_tx_action();
++ }
++}
++
++void netif_deschedule_work(netif_t *netif)
++{
++ remove_from_net_schedule_list(netif);
++}
++
++
++static void tx_add_credit(netif_t *netif)
++{
++ unsigned long max_burst, max_credit;
++
++ /*
++ * Allow a burst big enough to transmit a jumbo packet of up to 128kB.
++ * Otherwise the interface can seize up due to insufficient credit.
++ */
++ max_burst = RING_GET_REQUEST(&netif->tx, netif->tx.req_cons)->size;
++ max_burst = min(max_burst, 131072UL);
++ max_burst = max(max_burst, netif->credit_bytes);
++
++ /* Take care that adding a new chunk of credit doesn't wrap to zero. */
++ max_credit = netif->remaining_credit + netif->credit_bytes;
++ if (max_credit < netif->remaining_credit)
++ max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */
++
++ netif->remaining_credit = min(max_credit, max_burst);
++}
++
++static void tx_credit_callback(unsigned long data)
++{
++ netif_t *netif = (netif_t *)data;
++ tx_add_credit(netif);
++ netif_schedule_work(netif);
++}
++
++static inline int copy_pending_req(PEND_RING_IDX pending_idx)
++{
++ return gnttab_copy_grant_page(grant_tx_handle[pending_idx],
++ &mmap_pages[pending_idx]);
++}
++
++inline static void net_tx_action_dealloc(void)
++{
++ struct netbk_tx_pending_inuse *inuse, *n;
++ gnttab_unmap_grant_ref_t *gop;
++ u16 pending_idx;
++ PEND_RING_IDX dc, dp;
++ netif_t *netif;
++ int ret;
++ LIST_HEAD(list);
++
++ dc = dealloc_cons;
++ gop = tx_unmap_ops;
++
++ /*
++ * Free up any grants we have finished using
++ */
++ do {
++ dp = dealloc_prod;
++
++ /* Ensure we see all indices enqueued by netif_idx_release(). */
++ smp_rmb();
++
++ while (dc != dp) {
++ unsigned long pfn;
++
++ pending_idx = dealloc_ring[MASK_PEND_IDX(dc++)];
++ list_move_tail(&pending_inuse[pending_idx].list, &list);
++
++ pfn = idx_to_pfn(pending_idx);
++ /* Already unmapped? */
++ if (!phys_to_machine_mapping_valid(pfn))
++ continue;
++
++ gnttab_set_unmap_op(gop, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map,
++ grant_tx_handle[pending_idx]);
++ gop++;
++ }
++
++ if (netbk_copy_skb_mode != NETBK_DELAYED_COPY_SKB ||
++ list_empty(&pending_inuse_head))
++ break;
++
++ /* Copy any entries that have been pending for too long. */
++ list_for_each_entry_safe(inuse, n, &pending_inuse_head, list) {
++ if (time_after(inuse->alloc_time + HZ / 2, jiffies))
++ break;
++
++ switch (copy_pending_req(inuse - pending_inuse)) {
++ case 0:
++ list_move_tail(&inuse->list, &list);
++ continue;
++ case -EBUSY:
++ list_del_init(&inuse->list);
++ continue;
++ case -ENOENT:
++ continue;
++ }
++
++ break;
++ }
++ } while (dp != dealloc_prod);
++
++ dealloc_cons = dc;
++
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, tx_unmap_ops, gop - tx_unmap_ops);
++ BUG_ON(ret);
++
++ list_for_each_entry_safe(inuse, n, &list, list) {
++ pending_idx = inuse - pending_inuse;
++
++ netif = pending_tx_info[pending_idx].netif;
++
++ make_tx_response(netif, &pending_tx_info[pending_idx].req,
++ NETIF_RSP_OKAY);
++
++ /* Ready for next use. */
++ gnttab_reset_grant_page(mmap_pages[pending_idx]);
++
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++
++ netif_put(netif);
++
++ list_del_init(&inuse->list);
++ }
++}
++
++static void netbk_tx_err(netif_t *netif, netif_tx_request_t *txp, RING_IDX end)
++{
++ RING_IDX cons = netif->tx.req_cons;
++
++ do {
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ if (cons >= end)
++ break;
++ txp = RING_GET_REQUEST(&netif->tx, cons++);
++ } while (1);
++ netif->tx.req_cons = cons;
++ netif_schedule_work(netif);
++ netif_put(netif);
++}
++
++static int netbk_count_requests(netif_t *netif, netif_tx_request_t *first,
++ netif_tx_request_t *txp, int work_to_do)
++{
++ RING_IDX cons = netif->tx.req_cons;
++ int frags = 0;
++
++ if (!(first->flags & NETTXF_more_data))
++ return 0;
++
++ do {
++ if (frags >= work_to_do) {
++ DPRINTK("Need more frags\n");
++ return -frags;
++ }
++
++ if (unlikely(frags >= MAX_SKB_FRAGS)) {
++ DPRINTK("Too many frags\n");
++ return -frags;
++ }
++
++ memcpy(txp, RING_GET_REQUEST(&netif->tx, cons + frags),
++ sizeof(*txp));
++ if (txp->size > first->size) {
++ DPRINTK("Frags galore\n");
++ return -frags;
++ }
++
++ first->size -= txp->size;
++ frags++;
++
++ if (unlikely((txp->offset + txp->size) > PAGE_SIZE)) {
++ DPRINTK("txp->offset: %x, size: %u\n",
++ txp->offset, txp->size);
++ return -frags;
++ }
++ } while ((txp++)->flags & NETTXF_more_data);
++
++ return frags;
++}
++
++static gnttab_map_grant_ref_t *netbk_get_requests(netif_t *netif,
++ struct sk_buff *skb,
++ netif_tx_request_t *txp,
++ gnttab_map_grant_ref_t *mop)
++{
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ skb_frag_t *frags = shinfo->frags;
++ unsigned long pending_idx = *((u16 *)skb->data);
++ int i, start;
++
++ /* Skip first skb fragment if it is on same page as header fragment. */
++ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
++
++ for (i = start; i < shinfo->nr_frags; i++, txp++) {
++ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons++)];
++
++ gnttab_set_map_op(mop++, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map | GNTMAP_readonly,
++ txp->gref, netif->domid);
++
++ memcpy(&pending_tx_info[pending_idx].req, txp, sizeof(*txp));
++ netif_get(netif);
++ pending_tx_info[pending_idx].netif = netif;
++ frags[i].page = (void *)pending_idx;
++ }
++
++ return mop;
++}
++
++static int netbk_tx_check_mop(struct sk_buff *skb,
++ gnttab_map_grant_ref_t **mopp)
++{
++ gnttab_map_grant_ref_t *mop = *mopp;
++ int pending_idx = *((u16 *)skb->data);
++ netif_t *netif = pending_tx_info[pending_idx].netif;
++ netif_tx_request_t *txp;
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ int i, err, start;
++
++ /* Check status of header. */
++ err = mop->status;
++ if (unlikely(err)) {
++ txp = &pending_tx_info[pending_idx].req;
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++ netif_put(netif);
++ } else {
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(pending_idx)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(mop->dev_bus_addr >> PAGE_SHIFT));
++ grant_tx_handle[pending_idx] = mop->handle;
++ }
++
++ /* Skip first skb fragment if it is on same page as header fragment. */
++ start = ((unsigned long)shinfo->frags[0].page == pending_idx);
++
++ for (i = start; i < nr_frags; i++) {
++ int j, newerr;
++
++ pending_idx = (unsigned long)shinfo->frags[i].page;
++
++ /* Check error status: if okay then remember grant handle. */
++ newerr = (++mop)->status;
++ if (likely(!newerr)) {
++ set_phys_to_machine(
++ __pa(idx_to_kaddr(pending_idx))>>PAGE_SHIFT,
++ FOREIGN_FRAME(mop->dev_bus_addr>>PAGE_SHIFT));
++ grant_tx_handle[pending_idx] = mop->handle;
++ /* Had a previous error? Invalidate this fragment. */
++ if (unlikely(err))
++ netif_idx_release(pending_idx);
++ continue;
++ }
++
++ /* Error on this fragment: respond to client with an error. */
++ txp = &pending_tx_info[pending_idx].req;
++ make_tx_response(netif, txp, NETIF_RSP_ERROR);
++ pending_ring[MASK_PEND_IDX(pending_prod++)] = pending_idx;
++ netif_put(netif);
++
++ /* Not the first error? Preceding frags already invalidated. */
++ if (err)
++ continue;
++
++ /* First error: invalidate header and preceding fragments. */
++ pending_idx = *((u16 *)skb->data);
++ netif_idx_release(pending_idx);
++ for (j = start; j < i; j++) {
++ pending_idx = (unsigned long)shinfo->frags[i].page;
++ netif_idx_release(pending_idx);
++ }
++
++ /* Remember the error: invalidate all subsequent fragments. */
++ err = newerr;
++ }
++
++ *mopp = mop + 1;
++ return err;
++}
++
++static void netbk_fill_frags(struct sk_buff *skb)
++{
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ int i;
++
++ for (i = 0; i < nr_frags; i++) {
++ skb_frag_t *frag = shinfo->frags + i;
++ netif_tx_request_t *txp;
++ unsigned long pending_idx;
++
++ pending_idx = (unsigned long)frag->page;
++
++ pending_inuse[pending_idx].alloc_time = jiffies;
++ list_add_tail(&pending_inuse[pending_idx].list,
++ &pending_inuse_head);
++
++ txp = &pending_tx_info[pending_idx].req;
++ frag->page = virt_to_page(idx_to_kaddr(pending_idx));
++ frag->size = txp->size;
++ frag->page_offset = txp->offset;
++
++ skb->len += txp->size;
++ skb->data_len += txp->size;
++ skb->truesize += txp->size;
++ }
++}
++
++int netbk_get_extras(netif_t *netif, struct netif_extra_info *extras,
++ int work_to_do)
++{
++ struct netif_extra_info extra;
++ RING_IDX cons = netif->tx.req_cons;
++
++ do {
++ if (unlikely(work_to_do-- <= 0)) {
++ DPRINTK("Missing extra info\n");
++ return -EBADR;
++ }
++
++ memcpy(&extra, RING_GET_REQUEST(&netif->tx, cons),
++ sizeof(extra));
++ if (unlikely(!extra.type ||
++ extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
++ netif->tx.req_cons = ++cons;
++ DPRINTK("Invalid extra type: %d\n", extra.type);
++ return -EINVAL;
++ }
++
++ memcpy(&extras[extra.type - 1], &extra, sizeof(extra));
++ netif->tx.req_cons = ++cons;
++ } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE);
++
++ return work_to_do;
++}
++
++static int netbk_set_skb_gso(struct sk_buff *skb, struct netif_extra_info *gso)
++{
++ if (!gso->u.gso.size) {
++ DPRINTK("GSO size must not be zero.\n");
++ return -EINVAL;
++ }
++
++ /* Currently only TCPv4 S.O. is supported. */
++ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
++ DPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
++ return -EINVAL;
++ }
++
++ skb_shinfo(skb)->gso_size = gso->u.gso.size;
++ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
++
++ /* Header must be checked, and gso_segs computed. */
++ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
++ skb_shinfo(skb)->gso_segs = 0;
++
++ return 0;
++}
++
++/* Called after netfront has transmitted */
++static void net_tx_action(unsigned long unused)
++{
++ struct list_head *ent;
++ struct sk_buff *skb;
++ netif_t *netif;
++ netif_tx_request_t txreq;
++ netif_tx_request_t txfrags[MAX_SKB_FRAGS];
++ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
++ u16 pending_idx;
++ RING_IDX i;
++ gnttab_map_grant_ref_t *mop;
++ unsigned int data_len;
++ int ret, work_to_do;
++
++ if (dealloc_cons != dealloc_prod)
++ net_tx_action_dealloc();
++
++ mop = tx_map_ops;
++ while (((NR_PENDING_REQS + MAX_SKB_FRAGS) < MAX_PENDING_REQS) &&
++ !list_empty(&net_schedule_list)) {
++ /* Get a netif from the list with work to do. */
++ ent = net_schedule_list.next;
++ netif = list_entry(ent, netif_t, list);
++ netif_get(netif);
++ remove_from_net_schedule_list(netif);
++
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, work_to_do);
++ if (!work_to_do) {
++ netif_put(netif);
++ continue;
++ }
++
++ i = netif->tx.req_cons;
++ rmb(); /* Ensure that we see the request before we copy it. */
++ memcpy(&txreq, RING_GET_REQUEST(&netif->tx, i), sizeof(txreq));
++
++ /* Credit-based scheduling. */
++ if (txreq.size > netif->remaining_credit) {
++ unsigned long now = jiffies;
++ unsigned long next_credit =
++ netif->credit_timeout.expires +
++ msecs_to_jiffies(netif->credit_usec / 1000);
++
++ /* Timer could already be pending in rare cases. */
++ if (timer_pending(&netif->credit_timeout)) {
++ netif_put(netif);
++ continue;
++ }
++
++ /* Passed the point where we can replenish credit? */
++ if (time_after_eq(now, next_credit)) {
++ netif->credit_timeout.expires = now;
++ tx_add_credit(netif);
++ }
++
++ /* Still too big to send right now? Set a callback. */
++ if (txreq.size > netif->remaining_credit) {
++ netif->credit_timeout.data =
++ (unsigned long)netif;
++ netif->credit_timeout.function =
++ tx_credit_callback;
++ __mod_timer(&netif->credit_timeout,
++ next_credit);
++ netif_put(netif);
++ continue;
++ }
++ }
++ netif->remaining_credit -= txreq.size;
++
++ work_to_do--;
++ netif->tx.req_cons = ++i;
++
++ memset(extras, 0, sizeof(extras));
++ if (txreq.flags & NETTXF_extra_info) {
++ work_to_do = netbk_get_extras(netif, extras,
++ work_to_do);
++ i = netif->tx.req_cons;
++ if (unlikely(work_to_do < 0)) {
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++ }
++
++ ret = netbk_count_requests(netif, &txreq, txfrags, work_to_do);
++ if (unlikely(ret < 0)) {
++ netbk_tx_err(netif, &txreq, i - ret);
++ continue;
++ }
++ i += ret;
++
++ if (unlikely(txreq.size < ETH_HLEN)) {
++ DPRINTK("Bad packet size: %d\n", txreq.size);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++
++ /* No crossing a page as the payload mustn't fragment. */
++ if (unlikely((txreq.offset + txreq.size) > PAGE_SIZE)) {
++ DPRINTK("txreq.offset: %x, size: %u, end: %lu\n",
++ txreq.offset, txreq.size,
++ (txreq.offset &~PAGE_MASK) + txreq.size);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++
++ pending_idx = pending_ring[MASK_PEND_IDX(pending_cons)];
++
++ data_len = (txreq.size > PKT_PROT_LEN &&
++ ret < MAX_SKB_FRAGS) ?
++ PKT_PROT_LEN : txreq.size;
++
++ skb = alloc_skb(data_len + 16 + NET_IP_ALIGN,
++ GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(skb == NULL)) {
++ DPRINTK("Can't allocate a skb in start_xmit.\n");
++ netbk_tx_err(netif, &txreq, i);
++ break;
++ }
++
++ /* Packets passed to netif_rx() must have some headroom. */
++ skb_reserve(skb, 16 + NET_IP_ALIGN);
++
++ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
++ struct netif_extra_info *gso;
++ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
++
++ if (netbk_set_skb_gso(skb, gso)) {
++ kfree_skb(skb);
++ netbk_tx_err(netif, &txreq, i);
++ continue;
++ }
++ }
++
++ gnttab_set_map_op(mop, idx_to_kaddr(pending_idx),
++ GNTMAP_host_map | GNTMAP_readonly,
++ txreq.gref, netif->domid);
++ mop++;
++
++ memcpy(&pending_tx_info[pending_idx].req,
++ &txreq, sizeof(txreq));
++ pending_tx_info[pending_idx].netif = netif;
++ *((u16 *)skb->data) = pending_idx;
++
++ __skb_put(skb, data_len);
++
++ skb_shinfo(skb)->nr_frags = ret;
++ if (data_len < txreq.size) {
++ skb_shinfo(skb)->nr_frags++;
++ skb_shinfo(skb)->frags[0].page =
++ (void *)(unsigned long)pending_idx;
++ } else {
++ /* Discriminate from any valid pending_idx value. */
++ skb_shinfo(skb)->frags[0].page = (void *)~0UL;
++ }
++
++ __skb_queue_tail(&tx_queue, skb);
++
++ pending_cons++;
++
++ mop = netbk_get_requests(netif, skb, txfrags, mop);
++
++ netif->tx.req_cons = i;
++ netif_schedule_work(netif);
++
++ if ((mop - tx_map_ops) >= ARRAY_SIZE(tx_map_ops))
++ break;
++ }
++
++ if (mop == tx_map_ops)
++ return;
++
++ ret = HYPERVISOR_grant_table_op(
++ GNTTABOP_map_grant_ref, tx_map_ops, mop - tx_map_ops);
++ BUG_ON(ret);
++
++ mop = tx_map_ops;
++ while ((skb = __skb_dequeue(&tx_queue)) != NULL) {
++ netif_tx_request_t *txp;
++
++ pending_idx = *((u16 *)skb->data);
++ netif = pending_tx_info[pending_idx].netif;
++ txp = &pending_tx_info[pending_idx].req;
++
++ /* Check the remap error code. */
++ if (unlikely(netbk_tx_check_mop(skb, &mop))) {
++ DPRINTK("netback grant failed.\n");
++ skb_shinfo(skb)->nr_frags = 0;
++ kfree_skb(skb);
++ continue;
++ }
++
++ data_len = skb->len;
++ memcpy(skb->data,
++ (void *)(idx_to_kaddr(pending_idx)|txp->offset),
++ data_len);
++ if (data_len < txp->size) {
++ /* Append the packet payload as a fragment. */
++ txp->offset += data_len;
++ txp->size -= data_len;
++ } else {
++ /* Schedule a response immediately. */
++ netif_idx_release(pending_idx);
++ }
++
++ /*
++ * Old frontends do not assert data_validated but we
++ * can infer it from csum_blank so test both flags.
++ */
++ if (txp->flags & (NETTXF_data_validated|NETTXF_csum_blank)) {
++ skb->ip_summed = CHECKSUM_UNNECESSARY;
++ skb->proto_data_valid = 1;
++ } else {
++ skb->ip_summed = CHECKSUM_NONE;
++ skb->proto_data_valid = 0;
++ }
++ skb->proto_csum_blank = !!(txp->flags & NETTXF_csum_blank);
++
++ netbk_fill_frags(skb);
++
++ skb->dev = netif->dev;
++ skb->protocol = eth_type_trans(skb, skb->dev);
++
++ netif->stats.rx_bytes += skb->len;
++ netif->stats.rx_packets++;
++
++ if (unlikely(netbk_copy_skb_mode == NETBK_ALWAYS_COPY_SKB) &&
++ unlikely(skb_linearize(skb))) {
++ DPRINTK("Can't linearize skb in net_tx_action.\n");
++ kfree_skb(skb);
++ continue;
++ }
++
++ netif_rx(skb);
++ netif->dev->last_rx = jiffies;
++ }
++
++ if (netbk_copy_skb_mode == NETBK_DELAYED_COPY_SKB &&
++ !list_empty(&pending_inuse_head)) {
++ struct netbk_tx_pending_inuse *oldest;
++
++ oldest = list_entry(pending_inuse_head.next,
++ struct netbk_tx_pending_inuse, list);
++ mod_timer(&netbk_tx_pending_timer, oldest->alloc_time + HZ);
++ }
++}
++
++static void netif_idx_release(u16 pending_idx)
++{
++ static DEFINE_SPINLOCK(_lock);
++ unsigned long flags;
++
++ spin_lock_irqsave(&_lock, flags);
++ dealloc_ring[MASK_PEND_IDX(dealloc_prod)] = pending_idx;
++ /* Sync with net_tx_action_dealloc: insert idx /then/ incr producer. */
++ smp_wmb();
++ dealloc_prod++;
++ spin_unlock_irqrestore(&_lock, flags);
++
++ tasklet_schedule(&net_tx_tasklet);
++}
++
++static void netif_page_release(struct page *page)
++{
++ netif_idx_release(netif_page_index(page));
++}
++
++irqreturn_t netif_be_int(int irq, void *dev_id)
++{
++ netif_t *netif = dev_id;
++
++ add_to_net_schedule_list_tail(netif);
++ maybe_schedule_tx_action();
++
++ if (netif_schedulable(netif) && !netbk_queue_full(netif))
++ netif_wake_queue(netif->dev);
++
++ return IRQ_HANDLED;
++}
++
++static void make_tx_response(netif_t *netif,
++ netif_tx_request_t *txp,
++ s8 st)
++{
++ RING_IDX i = netif->tx.rsp_prod_pvt;
++ netif_tx_response_t *resp;
++ int notify;
++
++ resp = RING_GET_RESPONSE(&netif->tx, i);
++ resp->id = txp->id;
++ resp->status = st;
++
++ if (txp->flags & NETTXF_extra_info)
++ RING_GET_RESPONSE(&netif->tx, ++i)->status = NETIF_RSP_NULL;
++
++ netif->tx.rsp_prod_pvt = ++i;
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&netif->tx, notify);
++ if (notify)
++ notify_remote_via_irq(netif->irq);
++
++#ifdef CONFIG_XEN_NETDEV_PIPELINED_TRANSMITTER
++ if (i == netif->tx.req_cons) {
++ int more_to_do;
++ RING_FINAL_CHECK_FOR_REQUESTS(&netif->tx, more_to_do);
++ if (more_to_do)
++ add_to_net_schedule_list_tail(netif);
++ }
++#endif
++}
++
++static netif_rx_response_t *make_rx_response(netif_t *netif,
++ u16 id,
++ s8 st,
++ u16 offset,
++ u16 size,
++ u16 flags)
++{
++ RING_IDX i = netif->rx.rsp_prod_pvt;
++ netif_rx_response_t *resp;
++
++ resp = RING_GET_RESPONSE(&netif->rx, i);
++ resp->offset = offset;
++ resp->flags = flags;
++ resp->id = id;
++ resp->status = (s16)size;
++ if (st < 0)
++ resp->status = (s16)st;
++
++ netif->rx.rsp_prod_pvt = ++i;
++
++ return resp;
++}
++
++#ifdef NETBE_DEBUG_INTERRUPT
++static irqreturn_t netif_be_dbg(int irq, void *dev_id)
++{
++ struct list_head *ent;
++ netif_t *netif;
++ int i = 0;
++
++ printk(KERN_ALERT "netif_schedule_list:\n");
++ spin_lock_irq(&net_schedule_list_lock);
++
++ list_for_each (ent, &net_schedule_list) {
++ netif = list_entry(ent, netif_t, list);
++ printk(KERN_ALERT " %d: private(rx_req_cons=%08x "
++ "rx_resp_prod=%08x\n",
++ i, netif->rx.req_cons, netif->rx.rsp_prod_pvt);
++ printk(KERN_ALERT " tx_req_cons=%08x tx_resp_prod=%08x)\n",
++ netif->tx.req_cons, netif->tx.rsp_prod_pvt);
++ printk(KERN_ALERT " shared(rx_req_prod=%08x "
++ "rx_resp_prod=%08x\n",
++ netif->rx.sring->req_prod, netif->rx.sring->rsp_prod);
++ printk(KERN_ALERT " rx_event=%08x tx_req_prod=%08x\n",
++ netif->rx.sring->rsp_event, netif->tx.sring->req_prod);
++ printk(KERN_ALERT " tx_resp_prod=%08x, tx_event=%08x)\n",
++ netif->tx.sring->rsp_prod, netif->tx.sring->rsp_event);
++ i++;
++ }
++
++ spin_unlock_irq(&net_schedule_list_lock);
++ printk(KERN_ALERT " ** End of netif_schedule_list **\n");
++
++ return IRQ_HANDLED;
++}
++#endif
++
++static int __init netback_init(void)
++{
++ int i;
++ struct page *page;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ /* We can increase reservation by this much in net_rx_action(). */
++ balloon_update_driver_allowance(NET_RX_RING_SIZE);
++
++ skb_queue_head_init(&rx_queue);
++ skb_queue_head_init(&tx_queue);
++
++ init_timer(&net_timer);
++ net_timer.data = 0;
++ net_timer.function = net_alarm;
++
++ init_timer(&netbk_tx_pending_timer);
++ netbk_tx_pending_timer.data = 0;
++ netbk_tx_pending_timer.function = netbk_tx_pending_timeout;
++
++ mmap_pages = alloc_empty_pages_and_pagevec(MAX_PENDING_REQS);
++ if (mmap_pages == NULL) {
++ printk("%s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ for (i = 0; i < MAX_PENDING_REQS; i++) {
++ page = mmap_pages[i];
++ SetPageForeign(page, netif_page_release);
++ netif_page_index(page) = i;
++ INIT_LIST_HEAD(&pending_inuse[i].list);
++ }
++
++ pending_cons = 0;
++ pending_prod = MAX_PENDING_REQS;
++ for (i = 0; i < MAX_PENDING_REQS; i++)
++ pending_ring[i] = i;
++
++ spin_lock_init(&net_schedule_list_lock);
++ INIT_LIST_HEAD(&net_schedule_list);
++
++ netbk_copy_skb_mode = NETBK_DONT_COPY_SKB;
++ if (MODPARM_copy_skb) {
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_and_replace,
++ NULL, 0))
++ netbk_copy_skb_mode = NETBK_ALWAYS_COPY_SKB;
++ else
++ netbk_copy_skb_mode = NETBK_DELAYED_COPY_SKB;
++ }
++
++ netif_accel_init();
++
++ netif_xenbus_init();
++
++#ifdef NETBE_DEBUG_INTERRUPT
++ (void)bind_virq_to_irqhandler(VIRQ_DEBUG,
++ 0,
++ netif_be_dbg,
++ IRQF_SHARED,
++ "net-be-dbg",
++ &netif_be_dbg);
++#endif
++
++ return 0;
++}
++
++module_init(netback_init);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/netback/xenbus.c b/drivers/xen/netback/xenbus.c
+new file mode 100644
+index 0000000..33df68a
+--- /dev/null
++++ b/drivers/xen/netback/xenbus.c
+@@ -0,0 +1,446 @@
++/* Xenbus code for netif backend
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++ Copyright (C) 2005 XenSource Ltd
++
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
++
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
++
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
++
++#include <stdarg.h>
++#include <linux/module.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++#if 0
++#undef DPRINTK
++#define DPRINTK(fmt, args...) \
++ printk("netback/xenbus (%s:%d) " fmt ".\n", __FUNCTION__, __LINE__, ##args)
++#endif
++
++
++static int connect_rings(struct backend_info *);
++static void connect(struct backend_info *);
++static void backend_create_netif(struct backend_info *be);
++
++static int netback_remove(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ netback_remove_accelerators(be, dev);
++
++ if (be->netif) {
++ kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
++ netif_disconnect(be->netif);
++ be->netif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
++}
++
++
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and switch to InitWait.
++ */
++static int netback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ const char *message;
++ struct xenbus_transaction xbt;
++ int err;
++ int sg;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ sg = 1;
++ if (netbk_copy_skb_mode == NETBK_ALWAYS_COPY_SKB)
++ sg = 0;
++
++ do {
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto fail;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", sg);
++ if (err) {
++ message = "writing feature-sg";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4",
++ "%d", sg);
++ if (err) {
++ message = "writing feature-gso-tcpv4";
++ goto abort_transaction;
++ }
++
++ /* We support rx-copy path. */
++ err = xenbus_printf(xbt, dev->nodename,
++ "feature-rx-copy", "%d", 1);
++ if (err) {
++ message = "writing feature-rx-copy";
++ goto abort_transaction;
++ }
++
++ /*
++ * We don't support rx-flip path (except old guests who don't
++ * grok this feature flag).
++ */
++ err = xenbus_printf(xbt, dev->nodename,
++ "feature-rx-flip", "%d", 0);
++ if (err) {
++ message = "writing feature-rx-flip";
++ goto abort_transaction;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ } while (err == -EAGAIN);
++
++ if (err) {
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto fail;
++ }
++
++ netback_probe_accelerators(be, dev);
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++
++ /* This kicks hotplug scripts, so do it immediately. */
++ backend_create_netif(be);
++
++ return 0;
++
++abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, err, "%s", message);
++fail:
++ DPRINTK("failed");
++ netback_remove(dev);
++ return err;
++}
++
++
++/**
++ * Handle the creation of the hotplug script environment. We add the script
++ * and vif variables to the environment, for the benefit of the vif-* hotplug
++ * scripts.
++ */
++static int netback_uevent(struct xenbus_device *xdev, struct kobj_uevent_env *env)
++{
++ struct backend_info *be = xdev->dev.driver_data;
++ netif_t *netif = be->netif;
++ char *val;
++
++ DPRINTK("netback_uevent");
++
++ val = xenbus_read(XBT_NIL, xdev->nodename, "script", NULL);
++ if (IS_ERR(val)) {
++ int err = PTR_ERR(val);
++ xenbus_dev_fatal(xdev, err, "reading script");
++ return err;
++ }
++ else {
++ add_uevent_var(env, "script=%s", val);
++ kfree(val);
++ }
++
++ add_uevent_var(env, "vif=%s", netif->dev->name);
++
++ return 0;
++}
++
++
++static void backend_create_netif(struct backend_info *be)
++{
++ int err;
++ long handle;
++ struct xenbus_device *dev = be->dev;
++
++ if (be->netif != NULL)
++ return;
++
++ err = xenbus_scanf(XBT_NIL, dev->nodename, "handle", "%li", &handle);
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading handle");
++ return;
++ }
++
++ be->netif = netif_alloc(dev->otherend_id, handle);
++ if (IS_ERR(be->netif)) {
++ err = PTR_ERR(be->netif);
++ be->netif = NULL;
++ xenbus_dev_fatal(dev, err, "creating interface");
++ return;
++ }
++
++ kobject_uevent(&dev->dev.kobj, KOBJ_ONLINE);
++}
++
++
++/**
++ * Callback received when the frontend's state changes.
++ */
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ DPRINTK("%s", xenbus_strstate(frontend_state));
++
++ be->frontend_state = frontend_state;
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (dev->state == XenbusStateClosed) {
++ printk(KERN_INFO "%s: %s: prepare for reconnect\n",
++ __FUNCTION__, dev->nodename);
++ xenbus_switch_state(dev, XenbusStateInitWait);
++ }
++ break;
++
++ case XenbusStateInitialised:
++ break;
++
++ case XenbusStateConnected:
++ backend_create_netif(be);
++ if (be->netif)
++ connect(be);
++ break;
++
++ case XenbusStateClosing:
++ if (be->netif) {
++ kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
++ netif_disconnect(be->netif);
++ be->netif = NULL;
++ }
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
++}
++
++
++static void xen_net_read_rate(struct xenbus_device *dev,
++ unsigned long *bytes, unsigned long *usec)
++{
++ char *s, *e;
++ unsigned long b, u;
++ char *ratestr;
++
++ /* Default to unlimited bandwidth. */
++ *bytes = ~0UL;
++ *usec = 0;
++
++ ratestr = xenbus_read(XBT_NIL, dev->nodename, "rate", NULL);
++ if (IS_ERR(ratestr))
++ return;
++
++ s = ratestr;
++ b = simple_strtoul(s, &e, 10);
++ if ((s == e) || (*e != ','))
++ goto fail;
++
++ s = e + 1;
++ u = simple_strtoul(s, &e, 10);
++ if ((s == e) || (*e != '\0'))
++ goto fail;
++
++ *bytes = b;
++ *usec = u;
++
++ kfree(ratestr);
++ return;
++
++ fail:
++ WPRINTK("Failed to parse network rate limit. Traffic unlimited.\n");
++ kfree(ratestr);
++}
++
++static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++{
++ char *s, *e, *macstr;
++ int i;
++
++ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
++ if (IS_ERR(macstr))
++ return PTR_ERR(macstr);
++
++ for (i = 0; i < ETH_ALEN; i++) {
++ mac[i] = simple_strtoul(s, &e, 16);
++ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
++ kfree(macstr);
++ return -ENOENT;
++ }
++ s = e+1;
++ }
++
++ kfree(macstr);
++ return 0;
++}
++
++static void connect(struct backend_info *be)
++{
++ int err;
++ struct xenbus_device *dev = be->dev;
++
++ err = connect_rings(be);
++ if (err)
++ return;
++
++ err = xen_net_read_mac(dev, be->netif->fe_dev_addr);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
++ return;
++ }
++
++ xen_net_read_rate(dev, &be->netif->credit_bytes,
++ &be->netif->credit_usec);
++ be->netif->remaining_credit = be->netif->credit_bytes;
++
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ netif_wake_queue(be->netif->dev);
++}
++
++
++static int connect_rings(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long tx_ring_ref, rx_ring_ref;
++ unsigned int evtchn, rx_copy;
++ int err;
++ int val;
++
++ DPRINTK("");
++
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "tx-ring-ref", "%lu", &tx_ring_ref,
++ "rx-ring-ref", "%lu", &rx_ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
++
++ err = xenbus_scanf(XBT_NIL, dev->otherend, "request-rx-copy", "%u",
++ &rx_copy);
++ if (err == -ENOENT) {
++ err = 0;
++ rx_copy = 0;
++ }
++ if (err < 0) {
++ xenbus_dev_fatal(dev, err, "reading %s/request-rx-copy",
++ dev->otherend);
++ return err;
++ }
++ be->netif->copying_receiver = !!rx_copy;
++
++ if (be->netif->dev->tx_queue_len != 0) {
++ if (xenbus_scanf(XBT_NIL, dev->otherend,
++ "feature-rx-notify", "%d", &val) < 0)
++ val = 0;
++ if (val)
++ be->netif->can_queue = 1;
++ else
++ /* Must be non-zero for pfifo_fast to work. */
++ be->netif->dev->tx_queue_len = 1;
++ }
++
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-sg", "%d", &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features |= NETIF_F_SG;
++ be->netif->dev->features |= NETIF_F_SG;
++ }
++
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-gso-tcpv4", "%d",
++ &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features |= NETIF_F_TSO;
++ be->netif->dev->features |= NETIF_F_TSO;
++ }
++
++ if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-no-csum-offload",
++ "%d", &val) < 0)
++ val = 0;
++ if (val) {
++ be->netif->features &= ~NETIF_F_IP_CSUM;
++ be->netif->dev->features &= ~NETIF_F_IP_CSUM;
++ }
++
++ /* Map the shared frame, irq etc. */
++ err = netif_map(be->netif, tx_ring_ref, rx_ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_fatal(dev, err,
++ "mapping shared-frames %lu/%lu port %u",
++ tx_ring_ref, rx_ring_ref, evtchn);
++ return err;
++ }
++ return 0;
++}
++
++
++/* ** Driver Registration ** */
++
++
++static const struct xenbus_device_id netback_ids[] = {
++ { "vif" },
++ { "" }
++};
++
++
++static struct xenbus_driver netback = {
++ .name = "vif",
++ .ids = netback_ids,
++ .probe = netback_probe,
++ .remove = netback_remove,
++ .uevent = netback_uevent,
++ .otherend_changed = frontend_changed,
++};
++
++
++void netif_xenbus_init(void)
++{
++ if (xenbus_register_backend(&netback))
++ BUG();
++}
+diff --git a/drivers/xen/netfront/Makefile b/drivers/xen/netfront/Makefile
+new file mode 100644
+index 0000000..9c0c6ad
+--- /dev/null
++++ b/drivers/xen/netfront/Makefile
+@@ -0,0 +1,4 @@
++
++obj-$(CONFIG_XEN_NETDEV_FRONTEND) := xennet.o
++
++xennet-objs := netfront.o accel.o
+diff --git a/drivers/xen/netfront/accel.c b/drivers/xen/netfront/accel.c
+new file mode 100644
+index 0000000..6ecd424
+--- /dev/null
++++ b/drivers/xen/netfront/accel.c
+@@ -0,0 +1,824 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
++ *
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/netdevice.h>
++#include <linux/skbuff.h>
++#include <linux/list.h>
++#include <linux/mutex.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++
++#include "netfront.h"
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("netfront/accel (%s:%d) " fmt, \
++ __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "netfront/accel: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "netfront/accel: " fmt, ##args)
++
++static int netfront_remove_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev);
++static int netfront_load_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev,
++ const char *frontend);
++
++/*
++ * List of all netfront accelerator plugin modules available. Each
++ * list entry is of type struct netfront_accelerator.
++ */
++static struct list_head accelerators_list;
++
++/* Lock to protect access to accelerators_list */
++static spinlock_t accelerators_lock;
++
++/* Workqueue to process acceleration configuration changes */
++struct workqueue_struct *accel_watch_workqueue;
++
++/* Mutex to prevent concurrent loads and suspends, etc. */
++DEFINE_MUTEX(accelerator_mutex);
++
++void netif_init_accel(void)
++{
++ INIT_LIST_HEAD(&accelerators_list);
++ spin_lock_init(&accelerators_lock);
++
++ accel_watch_workqueue = create_workqueue("net_accel");
++}
++
++void netif_exit_accel(void)
++{
++ struct netfront_accelerator *accelerator, *tmp;
++ unsigned long flags;
++
++ flush_workqueue(accel_watch_workqueue);
++ destroy_workqueue(accel_watch_workqueue);
++
++ spin_lock_irqsave(&accelerators_lock, flags);
++
++ list_for_each_entry_safe(accelerator, tmp, &accelerators_list, link) {
++ BUG_ON(!list_empty(&accelerator->vif_states));
++
++ list_del(&accelerator->link);
++ kfree(accelerator->frontend);
++ kfree(accelerator);
++ }
++
++ spin_unlock_irqrestore(&accelerators_lock, flags);
++}
++
++
++/*
++ * Watch the configured accelerator and change plugin if it's modified
++ */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++static void accel_watch_work(struct work_struct *context)
++#else
++static void accel_watch_work(void *context)
++#endif
++{
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ struct netfront_accel_vif_state *vif_state =
++ container_of(context, struct netfront_accel_vif_state,
++ accel_work);
++#else
++ struct netfront_accel_vif_state *vif_state =
++ (struct netfront_accel_vif_state *)context;
++#endif
++ struct netfront_info *np = vif_state->np;
++ char *accel_frontend;
++ int accel_len, rc = -1;
++
++ mutex_lock(&accelerator_mutex);
++
++ accel_frontend = xenbus_read(XBT_NIL, np->xbdev->otherend,
++ "accel-frontend", &accel_len);
++ if (IS_ERR(accel_frontend)) {
++ accel_frontend = NULL;
++ netfront_remove_accelerator(np, np->xbdev);
++ } else {
++ /* If this is the first time, request the accelerator,
++ otherwise only request one if it has changed */
++ if (vif_state->accel_frontend == NULL) {
++ rc = netfront_load_accelerator(np, np->xbdev,
++ accel_frontend);
++ } else {
++ if (strncmp(vif_state->accel_frontend, accel_frontend,
++ accel_len)) {
++ netfront_remove_accelerator(np, np->xbdev);
++ rc = netfront_load_accelerator(np, np->xbdev,
++ accel_frontend);
++ }
++ }
++ }
++
++ /* Get rid of previous state and replace with the new name */
++ if (vif_state->accel_frontend != NULL)
++ kfree(vif_state->accel_frontend);
++ vif_state->accel_frontend = accel_frontend;
++
++ mutex_unlock(&accelerator_mutex);
++
++ if (rc == 0) {
++ DPRINTK("requesting module %s\n", accel_frontend);
++ request_module("%s", accel_frontend);
++ /*
++ * Module should now call netfront_accelerator_loaded() once
++ * it's up and running, and we can continue from there
++ */
++ }
++}
++
++
++static void accel_watch_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ struct netfront_accel_vif_state *vif_state =
++ container_of(watch, struct netfront_accel_vif_state,
++ accel_watch);
++ queue_work(accel_watch_workqueue, &vif_state->accel_work);
++}
++
++
++void netfront_accelerator_add_watch(struct netfront_info *np)
++{
++ int err;
++
++ /* Check we're not trying to overwrite an existing watch */
++ BUG_ON(np->accel_vif_state.accel_watch.node != NULL);
++
++ /* Get a watch on the accelerator plugin */
++ err = xenbus_watch_path2(np->xbdev, np->xbdev->otherend,
++ "accel-frontend",
++ &np->accel_vif_state.accel_watch,
++ accel_watch_changed);
++ if (err) {
++ DPRINTK("%s: Failed to register accel watch: %d\n",
++ __FUNCTION__, err);
++ np->accel_vif_state.accel_watch.node = NULL;
++ }
++}
++
++
++static
++void netfront_accelerator_remove_watch(struct netfront_info *np)
++{
++ struct netfront_accel_vif_state *vif_state = &np->accel_vif_state;
++
++ /* Get rid of watch on accelerator plugin */
++ if (vif_state->accel_watch.node != NULL) {
++ unregister_xenbus_watch(&vif_state->accel_watch);
++ kfree(vif_state->accel_watch.node);
++ vif_state->accel_watch.node = NULL;
++
++ flush_workqueue(accel_watch_workqueue);
++
++ /* Clean up any state left from watch */
++ if (vif_state->accel_frontend != NULL) {
++ kfree(vif_state->accel_frontend);
++ vif_state->accel_frontend = NULL;
++ }
++ }
++}
++
++
++/*
++ * Initialise the accel_vif_state field in the netfront state
++ */
++void init_accelerator_vif(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ np->accelerator = NULL;
++
++ /* It's assumed that these things don't change */
++ np->accel_vif_state.np = np;
++ np->accel_vif_state.dev = dev;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work);
++#else
++ INIT_WORK(&np->accel_vif_state.accel_work, accel_watch_work,
++ &np->accel_vif_state);
++#endif
++}
++
++
++/*
++ * Compare a frontend description string against an accelerator to see
++ * if they match. Would ultimately be nice to replace the string with
++ * a unique numeric identifier for each accelerator.
++ */
++static int match_accelerator(const char *frontend,
++ struct netfront_accelerator *accelerator)
++{
++ return strcmp(frontend, accelerator->frontend) == 0;
++}
++
++
++/*
++ * Add a frontend vif to the list of vifs that is using a netfront
++ * accelerator plugin module.
++ */
++static void add_accelerator_vif(struct netfront_accelerator *accelerator,
++ struct netfront_info *np)
++{
++ unsigned long flags;
++
++ /* Need lock to write list */
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++
++ if (np->accelerator == NULL) {
++ np->accelerator = accelerator;
++
++ list_add(&np->accel_vif_state.link, &accelerator->vif_states);
++ } else {
++ /*
++ * May get here legitimately if suspend_cancel is
++ * called, but in that case configuration should not
++ * have changed
++ */
++ BUG_ON(np->accelerator != accelerator);
++ }
++
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++}
++
++
++/*
++ * Initialise the state to track an accelerator plugin module.
++ */
++static int init_accelerator(const char *frontend,
++ struct netfront_accelerator **result,
++ struct netfront_accel_hooks *hooks)
++{
++ struct netfront_accelerator *accelerator =
++ kmalloc(sizeof(struct netfront_accelerator), GFP_KERNEL);
++ unsigned long flags;
++ int frontend_len;
++
++ if (!accelerator) {
++ DPRINTK("no memory for accelerator\n");
++ return -ENOMEM;
++ }
++
++ frontend_len = strlen(frontend) + 1;
++ accelerator->frontend = kmalloc(frontend_len, GFP_KERNEL);
++ if (!accelerator->frontend) {
++ DPRINTK("no memory for accelerator\n");
++ kfree(accelerator);
++ return -ENOMEM;
++ }
++ strlcpy(accelerator->frontend, frontend, frontend_len);
++
++ INIT_LIST_HEAD(&accelerator->vif_states);
++ spin_lock_init(&accelerator->vif_states_lock);
++
++ accelerator->hooks = hooks;
++
++ spin_lock_irqsave(&accelerators_lock, flags);
++ list_add(&accelerator->link, &accelerators_list);
++ spin_unlock_irqrestore(&accelerators_lock, flags);
++
++ *result = accelerator;
++
++ return 0;
++}
++
++
++/*
++ * Modify the hooks stored in the per-vif state to match that in the
++ * netfront accelerator's state.
++ */
++static void
++accelerator_set_vif_state_hooks(struct netfront_accel_vif_state *vif_state)
++{
++ /* This function must be called with the vif_states_lock held */
++
++ DPRINTK("%p\n",vif_state);
++
++ /* Make sure there are no data path operations going on */
++ napi_disable(&vif_state->np->napi);
++ netif_tx_lock_bh(vif_state->np->netdev);
++
++ vif_state->hooks = vif_state->np->accelerator->hooks;
++
++ netif_tx_unlock_bh(vif_state->np->netdev);
++ napi_enable(&vif_state->np->napi);
++}
++
++
++static void accelerator_probe_new_vif(struct netfront_info *np,
++ struct xenbus_device *dev,
++ struct netfront_accelerator *accelerator)
++{
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
++
++ DPRINTK("\n");
++
++ /* Include this frontend device on the accelerator's list */
++ add_accelerator_vif(accelerator, np);
++
++ hooks = accelerator->hooks;
++
++ if (hooks) {
++ if (hooks->new_device(np->netdev, dev) == 0) {
++ spin_lock_irqsave
++ (&accelerator->vif_states_lock, flags);
++
++ accelerator_set_vif_state_hooks(&np->accel_vif_state);
++
++ spin_unlock_irqrestore
++ (&accelerator->vif_states_lock, flags);
++ }
++ }
++
++ return;
++}
++
++
++/*
++ * Request that a particular netfront accelerator plugin is loaded.
++ * Usually called as a result of the vif configuration specifying
++ * which one to use. Must be called with accelerator_mutex held
++ */
++static int netfront_load_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev,
++ const char *frontend)
++{
++ struct netfront_accelerator *accelerator;
++ int rc = 0;
++
++ DPRINTK(" %s\n", frontend);
++
++ /*
++ * Look at list of loaded accelerators to see if the requested
++ * one is already there
++ */
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ accelerator_probe_new_vif(np, dev, accelerator);
++ return 0;
++ }
++ }
++
++ /* Couldn't find it, so create a new one and load the module */
++ if ((rc = init_accelerator(frontend, &accelerator, NULL)) < 0) {
++ return rc;
++ }
++
++ /* Include this frontend device on the accelerator's list */
++ add_accelerator_vif(accelerator, np);
++
++ return rc;
++}
++
++
++/*
++ * Go through all the netfront vifs and see if they have requested
++ * this accelerator. Notify the accelerator plugin of the relevant
++ * device if so. Called when an accelerator plugin module is first
++ * loaded and connects to netfront.
++ */
++static void
++accelerator_probe_vifs(struct netfront_accelerator *accelerator,
++ struct netfront_accel_hooks *hooks)
++{
++ struct netfront_accel_vif_state *vif_state, *tmp;
++ unsigned long flags;
++
++ DPRINTK("%p\n", accelerator);
++
++ /*
++ * Store the hooks for future calls to probe a new device, and
++ * to wire into the vif_state once the accelerator plugin is
++ * ready to accelerate each vif
++ */
++ BUG_ON(hooks == NULL);
++ accelerator->hooks = hooks;
++
++ /*
++ * currently hold accelerator_mutex, so don't need
++ * vif_states_lock to read the list
++ */
++ list_for_each_entry_safe(vif_state, tmp, &accelerator->vif_states,
++ link) {
++ struct netfront_info *np = vif_state->np;
++
++ if (hooks->new_device(np->netdev, vif_state->dev) == 0) {
++ spin_lock_irqsave
++ (&accelerator->vif_states_lock, flags);
++
++ accelerator_set_vif_state_hooks(vif_state);
++
++ spin_unlock_irqrestore
++ (&accelerator->vif_states_lock, flags);
++ }
++ }
++}
++
++
++/*
++ * Called by the netfront accelerator plugin module when it has loaded
++ */
++int netfront_accelerator_loaded(int version, const char *frontend,
++ struct netfront_accel_hooks *hooks)
++{
++ struct netfront_accelerator *accelerator;
++
++ if (is_initial_xendomain())
++ return -EINVAL;
++
++ if (version != NETFRONT_ACCEL_VERSION) {
++ if (version > NETFRONT_ACCEL_VERSION) {
++ /* Caller has higher version number, leave it
++ up to them to decide whether to continue.
++ They can re-call with a lower number if
++ they're happy to be compatible with us */
++ return NETFRONT_ACCEL_VERSION;
++ } else {
++ /* We have a more recent version than caller.
++ Currently reject, but may in future be able
++ to be backwardly compatible */
++ return -EPROTO;
++ }
++ }
++
++ mutex_lock(&accelerator_mutex);
++
++ /*
++ * Look through list of accelerators to see if it has already
++ * been requested
++ */
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ accelerator_probe_vifs(accelerator, hooks);
++ goto out;
++ }
++ }
++
++ /*
++ * If it wasn't in the list, add it now so that when it is
++ * requested the caller will find it
++ */
++ DPRINTK("Couldn't find matching accelerator (%s)\n",
++ frontend);
++
++ init_accelerator(frontend, &accelerator, hooks);
++
++ out:
++ mutex_unlock(&accelerator_mutex);
++ return 0;
++}
++EXPORT_SYMBOL_GPL(netfront_accelerator_loaded);
++
++
++/*
++ * Remove the hooks from a single vif state.
++ */
++static void
++accelerator_remove_single_hook(struct netfront_accelerator *accelerator,
++ struct netfront_accel_vif_state *vif_state)
++{
++ /* Make sure there are no data path operations going on */
++ napi_disable(&vif_state->np->napi);
++ netif_tx_lock_bh(vif_state->np->netdev);
++
++ /*
++ * Remove the hooks, but leave the vif_state on the
++ * accelerator's list as that signifies this vif is
++ * interested in using that accelerator if it becomes
++ * available again
++ */
++ vif_state->hooks = NULL;
++
++ netif_tx_unlock_bh(vif_state->np->netdev);
++ napi_enable(&vif_state->np->napi);
++}
++
++
++/*
++ * Safely remove the accelerator function hooks from a netfront state.
++ */
++static void accelerator_remove_hooks(struct netfront_accelerator *accelerator)
++{
++ struct netfront_accel_hooks *hooks;
++ struct netfront_accel_vif_state *vif_state, *tmp;
++ unsigned long flags;
++
++ /* Mutex is held so don't need vif_states_lock to iterate list */
++ list_for_each_entry_safe(vif_state, tmp,
++ &accelerator->vif_states,
++ link) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++
++ if(vif_state->hooks) {
++ hooks = vif_state->hooks;
++
++ /* Last chance to get statistics from the accelerator */
++ hooks->get_stats(vif_state->np->netdev,
++ &vif_state->np->stats);
++
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ flags);
++
++ accelerator_remove_single_hook(accelerator, vif_state);
++
++ accelerator->hooks->remove(vif_state->dev);
++ } else {
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ flags);
++ }
++ }
++
++ accelerator->hooks = NULL;
++}
++
++
++/*
++ * Called by a netfront accelerator when it is unloaded. This safely
++ * removes the hooks into the plugin and blocks until all devices have
++ * finished using it, so on return it is safe to unload.
++ */
++void netfront_accelerator_stop(const char *frontend)
++{
++ struct netfront_accelerator *accelerator;
++ unsigned long flags;
++
++ mutex_lock(&accelerator_mutex);
++ spin_lock_irqsave(&accelerators_lock, flags);
++
++ list_for_each_entry(accelerator, &accelerators_list, link) {
++ if (match_accelerator(frontend, accelerator)) {
++ spin_unlock_irqrestore(&accelerators_lock, flags);
++
++ accelerator_remove_hooks(accelerator);
++
++ goto out;
++ }
++ }
++ spin_unlock_irqrestore(&accelerators_lock, flags);
++ out:
++ mutex_unlock(&accelerator_mutex);
++}
++EXPORT_SYMBOL_GPL(netfront_accelerator_stop);
++
++
++/* Helper for call_remove and do_suspend */
++static int do_remove(struct netfront_info *np, struct xenbus_device *dev,
++ unsigned long *lock_flags)
++{
++ struct netfront_accelerator *accelerator = np->accelerator;
++ struct netfront_accel_hooks *hooks;
++ int rc = 0;
++
++ if (np->accel_vif_state.hooks) {
++ hooks = np->accel_vif_state.hooks;
++
++ /* Last chance to get statistics from the accelerator */
++ hooks->get_stats(np->netdev, &np->stats);
++
++ spin_unlock_irqrestore(&accelerator->vif_states_lock,
++ *lock_flags);
++
++ /*
++ * Try and do the opposite of accelerator_probe_new_vif
++ * to ensure there's no state pointing back at the
++ * netdev
++ */
++ accelerator_remove_single_hook(accelerator,
++ &np->accel_vif_state);
++
++ rc = accelerator->hooks->remove(dev);
++
++ spin_lock_irqsave(&accelerator->vif_states_lock, *lock_flags);
++ }
++
++ return rc;
++}
++
++
++static int netfront_remove_accelerator(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_vif_state *tmp_vif_state;
++ unsigned long flags;
++ int rc = 0;
++
++ /* Check that we've got a device that was accelerated */
++ if (np->accelerator == NULL)
++ return rc;
++
++ accelerator = np->accelerator;
++
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++
++ list_for_each_entry(tmp_vif_state, &accelerator->vif_states,
++ link) {
++ if (tmp_vif_state == &np->accel_vif_state) {
++ list_del(&np->accel_vif_state.link);
++ break;
++ }
++ }
++
++ rc = do_remove(np, dev, &flags);
++
++ np->accelerator = NULL;
++
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++
++ return rc;
++}
++
++
++int netfront_accelerator_call_remove(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ int rc;
++ netfront_accelerator_remove_watch(np);
++ mutex_lock(&accelerator_mutex);
++ rc = netfront_remove_accelerator(np, dev);
++ mutex_unlock(&accelerator_mutex);
++ return rc;
++}
++
++
++int netfront_accelerator_suspend(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ unsigned long flags;
++ int rc = 0;
++
++ netfront_accelerator_remove_watch(np);
++
++ mutex_lock(&accelerator_mutex);
++
++ /* Check that we've got a device that was accelerated */
++ if (np->accelerator == NULL)
++ goto out;
++
++ /*
++ * Call the remove accelerator hook, but leave the vif_state
++ * on the accelerator's list in case there is a suspend_cancel.
++ */
++ spin_lock_irqsave(&np->accelerator->vif_states_lock, flags);
++
++ rc = do_remove(np, dev, &flags);
++
++ spin_unlock_irqrestore(&np->accelerator->vif_states_lock, flags);
++ out:
++ mutex_unlock(&accelerator_mutex);
++ return rc;
++}
++
++
++int netfront_accelerator_suspend_cancel(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ /*
++ * Setting the watch will cause it to fire and probe the
++ * accelerator, so no need to call accelerator_probe_new_vif()
++ * directly here
++ */
++ if (dev->state == XenbusStateConnected)
++ netfront_accelerator_add_watch(np);
++ return 0;
++}
++
++
++void netfront_accelerator_resume(struct netfront_info *np,
++ struct xenbus_device *dev)
++{
++ struct netfront_accel_vif_state *accel_vif_state = NULL;
++ spinlock_t *vif_states_lock;
++ unsigned long flags;
++
++ mutex_lock(&accelerator_mutex);
++
++ /* Check that we've got a device that was accelerated */
++ if(np->accelerator == NULL)
++ goto out;
++
++ /* Find the vif_state from the accelerator's list */
++ list_for_each_entry(accel_vif_state, &np->accelerator->vif_states,
++ link) {
++ if (accel_vif_state->dev == dev) {
++ BUG_ON(accel_vif_state != &np->accel_vif_state);
++
++ vif_states_lock = &np->accelerator->vif_states_lock;
++ spin_lock_irqsave(vif_states_lock, flags);
++
++ /*
++ * Remove it from the accelerator's list so
++ * state is consistent for probing new vifs
++ * when they get connected
++ */
++ list_del(&accel_vif_state->link);
++ np->accelerator = NULL;
++
++ spin_unlock_irqrestore(vif_states_lock, flags);
++
++ break;
++ }
++ }
++
++ out:
++ mutex_unlock(&accelerator_mutex);
++ return;
++}
++
++
++int netfront_check_accelerator_queue_ready(struct net_device *dev,
++ struct netfront_info *np)
++{
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ int rc = 1;
++ unsigned long flags;
++
++ accelerator = np->accelerator;
++
++ /* Call the check_ready accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ rc = np->accel_vif_state.hooks->check_ready(dev);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++ }
++
++ return rc;
++}
++
++
++void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np,
++ struct net_device *dev)
++{
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
++
++ accelerator = np->accelerator;
++
++ /* Call the stop_napi_interrupts accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator != NULL) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ np->accel_vif_state.hooks->stop_napi_irq(dev);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++ }
++}
++
++
++int netfront_accelerator_call_get_stats(struct netfront_info *np,
++ struct net_device *dev)
++{
++ struct netfront_accelerator *accelerator;
++ struct netfront_accel_hooks *hooks;
++ unsigned long flags;
++ int rc = 0;
++
++ accelerator = np->accelerator;
++
++ /* Call the get_stats accelerator hook. */
++ if (np->accel_vif_state.hooks && accelerator != NULL) {
++ spin_lock_irqsave(&accelerator->vif_states_lock, flags);
++ hooks = np->accel_vif_state.hooks;
++ if (hooks && np->accelerator == accelerator)
++ rc = np->accel_vif_state.hooks->get_stats(dev,
++ &np->stats);
++ spin_unlock_irqrestore(&accelerator->vif_states_lock, flags);
++ }
++ return rc;
++}
++
+diff --git a/drivers/xen/netfront/netfront.c b/drivers/xen/netfront/netfront.c
+new file mode 100644
+index 0000000..be91fee
+--- /dev/null
++++ b/drivers/xen/netfront/netfront.c
+@@ -0,0 +1,2227 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
++ *
++ * Copyright (c) 2002-2005, K A Fraser
++ * Copyright (c) 2005, XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/module.h>
++#include <linux/version.h>
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/netdevice.h>
++#include <linux/inetdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/skbuff.h>
++#include <linux/init.h>
++#include <linux/bitops.h>
++#include <linux/ethtool.h>
++#include <linux/in.h>
++#include <linux/if_ether.h>
++#include <linux/io.h>
++#include <linux/moduleparam.h>
++#include <net/sock.h>
++#include <net/pkt_sched.h>
++#include <net/arp.h>
++#include <net/route.h>
++#include <asm/uaccess.h>
++#include <xen/evtchn.h>
++#include <xen/xenbus.h>
++#include <xen/interface/io/netif.h>
++#include <xen/interface/memory.h>
++#include <xen/balloon.h>
++#include <asm/page.h>
++#include <asm/maddr.h>
++#include <asm/uaccess.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
++
++struct netfront_cb {
++ struct page *page;
++ unsigned offset;
++};
++
++#define NETFRONT_SKB_CB(skb) ((struct netfront_cb *)((skb)->cb))
++
++#include "netfront.h"
++
++/*
++ * Mutually-exclusive module options to select receive data path:
++ * rx_copy : Packets are copied by network backend into local memory
++ * rx_flip : Page containing packet data is transferred to our ownership
++ * For fully-virtualised guests there is no option - copying must be used.
++ * For paravirtualised guests, flipping is the default.
++ */
++#ifdef CONFIG_XEN
++static int MODPARM_rx_copy = 0;
++module_param_named(rx_copy, MODPARM_rx_copy, bool, 0);
++MODULE_PARM_DESC(rx_copy, "Copy packets from network card (rather than flip)");
++static int MODPARM_rx_flip = 0;
++module_param_named(rx_flip, MODPARM_rx_flip, bool, 0);
++MODULE_PARM_DESC(rx_flip, "Flip packets from network card (rather than copy)");
++#else
++static const int MODPARM_rx_copy = 1;
++static const int MODPARM_rx_flip = 0;
++#endif
++
++#define RX_COPY_THRESHOLD 256
++
++/* If we don't have GSO, fake things up so that we never try to use it. */
++#if defined(NETIF_F_GSO)
++#define HAVE_GSO 1
++#define HAVE_TSO 1 /* TSO is a subset of GSO */
++#define HAVE_CSUM_OFFLOAD 1
++static inline void dev_disable_gso_features(struct net_device *dev)
++{
++ /* Turn off all GSO bits except ROBUST. */
++ dev->features &= (1 << NETIF_F_GSO_SHIFT) - 1;
++ dev->features |= NETIF_F_GSO_ROBUST;
++}
++#elif defined(NETIF_F_TSO)
++#define HAVE_GSO 0
++#define HAVE_TSO 1
++
++/* Some older kernels cannot cope with incorrect checksums,
++ * particularly in netfilter. I'm not sure there is 100% correlation
++ * with the presence of NETIF_F_TSO but it appears to be a good first
++ * approximiation.
++ */
++#define HAVE_CSUM_OFFLOAD 0
++
++#define gso_size tso_size
++#define gso_segs tso_segs
++static inline void dev_disable_gso_features(struct net_device *dev)
++{
++ /* Turn off all TSO bits. */
++ dev->features &= ~NETIF_F_TSO;
++}
++static inline int skb_is_gso(const struct sk_buff *skb)
++{
++ return skb_shinfo(skb)->tso_size;
++}
++static inline int skb_gso_ok(struct sk_buff *skb, int features)
++{
++ return (features & NETIF_F_TSO);
++}
++
++static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
++{
++ return skb_is_gso(skb) &&
++ (!skb_gso_ok(skb, dev->features) ||
++ unlikely(skb->ip_summed != CHECKSUM_PARTIAL));
++}
++#else
++#define HAVE_GSO 0
++#define HAVE_TSO 0
++#define HAVE_CSUM_OFFLOAD 0
++#define netif_needs_gso(dev, skb) 0
++#define dev_disable_gso_features(dev) ((void)0)
++#define ethtool_op_set_tso(dev, data) (-ENOSYS)
++#endif
++
++#define GRANT_INVALID_REF 0
++
++struct netfront_rx_info {
++ struct netif_rx_response rx;
++ struct netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
++};
++
++/*
++ * Implement our own carrier flag: the network stack's version causes delays
++ * when the carrier is re-enabled (in particular, dev_activate() may not
++ * immediately be called, which can cause packet loss).
++ */
++#define netfront_carrier_on(netif) ((netif)->carrier = 1)
++#define netfront_carrier_off(netif) ((netif)->carrier = 0)
++#define netfront_carrier_ok(netif) ((netif)->carrier)
++
++/*
++ * Access macros for acquiring freeing slots in tx_skbs[].
++ */
++
++static inline void add_id_to_freelist(struct sk_buff **list, unsigned short id)
++{
++ list[id] = list[0];
++ list[0] = (void *)(unsigned long)id;
++}
++
++static inline unsigned short get_id_from_freelist(struct sk_buff **list)
++{
++ unsigned int id = (unsigned int)(unsigned long)list[0];
++ list[0] = list[id];
++ return id;
++}
++
++static inline int xennet_rxidx(RING_IDX idx)
++{
++ return idx & (NET_RX_RING_SIZE - 1);
++}
++
++static inline struct sk_buff *xennet_get_rx_skb(struct netfront_info *np,
++ RING_IDX ri)
++{
++ int i = xennet_rxidx(ri);
++ struct sk_buff *skb = np->rx_skbs[i];
++ np->rx_skbs[i] = NULL;
++ return skb;
++}
++
++static inline grant_ref_t xennet_get_rx_ref(struct netfront_info *np,
++ RING_IDX ri)
++{
++ int i = xennet_rxidx(ri);
++ grant_ref_t ref = np->grant_rx_ref[i];
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
++ return ref;
++}
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("netfront (%s:%d) " fmt, \
++ __FUNCTION__, __LINE__, ##args)
++#define IPRINTK(fmt, args...) \
++ printk(KERN_INFO "netfront: " fmt, ##args)
++#define WPRINTK(fmt, args...) \
++ printk(KERN_WARNING "netfront: " fmt, ##args)
++
++static int setup_device(struct xenbus_device *, struct netfront_info *);
++static struct net_device *create_netdev(struct xenbus_device *);
++
++static void end_access(int, void *);
++static void netif_disconnect_backend(struct netfront_info *);
++
++static int network_connect(struct net_device *);
++static void network_tx_buf_gc(struct net_device *);
++static void network_alloc_rx_buffers(struct net_device *);
++static void send_fake_arp(struct net_device *);
++
++static irqreturn_t netif_int(int irq, void *dev_id);
++
++#ifdef CONFIG_SYSFS
++static int xennet_sysfs_addif(struct net_device *netdev);
++static void xennet_sysfs_delif(struct net_device *netdev);
++#else /* !CONFIG_SYSFS */
++#define xennet_sysfs_addif(dev) (0)
++#define xennet_sysfs_delif(dev) do { } while(0)
++#endif
++
++static inline int xennet_can_sg(struct net_device *dev)
++{
++ return dev->features & NETIF_F_SG;
++}
++
++/**
++ * Entry point to this code when a new device is created. Allocate the basic
++ * structures and the ring buffers for communication with the backend, and
++ * inform the backend of the appropriate details for those.
++ */
++static int __devinit netfront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ struct net_device *netdev;
++ struct netfront_info *info;
++
++ netdev = create_netdev(dev);
++ if (IS_ERR(netdev)) {
++ err = PTR_ERR(netdev);
++ xenbus_dev_fatal(dev, err, "creating netdev");
++ return err;
++ }
++
++ info = netdev_priv(netdev);
++ dev->dev.driver_data = info;
++
++ err = register_netdev(info->netdev);
++ if (err) {
++ printk(KERN_WARNING "%s: register_netdev err=%d\n",
++ __FUNCTION__, err);
++ goto fail;
++ }
++
++ err = xennet_sysfs_addif(info->netdev);
++ if (err) {
++ unregister_netdev(info->netdev);
++ printk(KERN_WARNING "%s: add sysfs failed err=%d\n",
++ __FUNCTION__, err);
++ goto fail;
++ }
++
++ return 0;
++
++ fail:
++ free_netdev(netdev);
++ dev->dev.driver_data = NULL;
++ return err;
++}
++
++static int __devexit netfront_remove(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++
++ DPRINTK("%s\n", dev->nodename);
++
++ netfront_accelerator_call_remove(info, dev);
++
++ netif_disconnect_backend(info);
++
++ del_timer_sync(&info->rx_refill_timer);
++
++ xennet_sysfs_delif(info->netdev);
++
++ unregister_netdev(info->netdev);
++
++ free_netdev(info->netdev);
++
++ return 0;
++}
++
++
++static int netfront_suspend(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++ return netfront_accelerator_suspend(info, dev);
++}
++
++
++static int netfront_suspend_cancel(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++ return netfront_accelerator_suspend_cancel(info, dev);
++}
++
++
++/**
++ * We are reconnecting to the backend, due to a suspend/resume, or a backend
++ * driver restart. We tear down our netif structure and recreate it, but
++ * leave the device-layer structures intact so that this is transparent to the
++ * rest of the kernel.
++ */
++static int netfront_resume(struct xenbus_device *dev)
++{
++ struct netfront_info *info = dev->dev.driver_data;
++
++ DPRINTK("%s\n", dev->nodename);
++
++ netfront_accelerator_resume(info, dev);
++
++ netif_disconnect_backend(info);
++ return 0;
++}
++
++static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++{
++ char *s, *e, *macstr;
++ int i;
++
++ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
++ if (IS_ERR(macstr))
++ return PTR_ERR(macstr);
++
++ for (i = 0; i < ETH_ALEN; i++) {
++ mac[i] = simple_strtoul(s, &e, 16);
++ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
++ kfree(macstr);
++ return -ENOENT;
++ }
++ s = e+1;
++ }
++
++ kfree(macstr);
++ return 0;
++}
++
++/* Common code used when first setting up, and when resuming. */
++static int talk_to_backend(struct xenbus_device *dev,
++ struct netfront_info *info)
++{
++ const char *message;
++ struct xenbus_transaction xbt;
++ int err;
++
++ /* Read mac only in the first setup. */
++ if (!is_valid_ether_addr(info->mac)) {
++ err = xen_net_read_mac(dev, info->mac);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "parsing %s/mac",
++ dev->nodename);
++ goto out;
++ }
++ }
++
++ /* Create shared ring, alloc event channel. */
++ err = setup_device(dev, info);
++ if (err)
++ goto out;
++
++ /* This will load an accelerator if one is configured when the
++ * watch fires */
++ netfront_accelerator_add_watch(info);
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ goto destroy_ring;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "tx-ring-ref","%u",
++ info->tx_ring_ref);
++ if (err) {
++ message = "writing tx ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename, "rx-ring-ref","%u",
++ info->rx_ring_ref);
++ if (err) {
++ message = "writing rx ring-ref";
++ goto abort_transaction;
++ }
++ err = xenbus_printf(xbt, dev->nodename,
++ "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++ if (err) {
++ message = "writing event-channel";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "request-rx-copy", "%u",
++ info->copying_receiver);
++ if (err) {
++ message = "writing request-rx-copy";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-rx-notify", "%d", 1);
++ if (err) {
++ message = "writing feature-rx-notify";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-no-csum-offload",
++ "%d", !HAVE_CSUM_OFFLOAD);
++ if (err) {
++ message = "writing feature-no-csum-offload";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", 1);
++ if (err) {
++ message = "writing feature-sg";
++ goto abort_transaction;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4", "%d",
++ HAVE_TSO);
++ if (err) {
++ message = "writing feature-gso-tcpv4";
++ goto abort_transaction;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto destroy_ring;
++ }
++
++ return 0;
++
++ abort_transaction:
++ xenbus_transaction_end(xbt, 1);
++ xenbus_dev_fatal(dev, err, "%s", message);
++ destroy_ring:
++ netfront_accelerator_call_remove(info, dev);
++ netif_disconnect_backend(info);
++ out:
++ return err;
++}
++
++static int setup_device(struct xenbus_device *dev, struct netfront_info *info)
++{
++ struct netif_tx_sring *txs;
++ struct netif_rx_sring *rxs;
++ int err;
++ struct net_device *netdev = info->netdev;
++
++ info->tx_ring_ref = GRANT_INVALID_REF;
++ info->rx_ring_ref = GRANT_INVALID_REF;
++ info->rx.sring = NULL;
++ info->tx.sring = NULL;
++ info->irq = 0;
++
++ txs = (struct netif_tx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
++ if (!txs) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err, "allocating tx ring page");
++ goto fail;
++ }
++ SHARED_RING_INIT(txs);
++ FRONT_RING_INIT(&info->tx, txs, PAGE_SIZE);
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(txs));
++ if (err < 0) {
++ free_page((unsigned long)txs);
++ goto fail;
++ }
++ info->tx_ring_ref = err;
++
++ rxs = (struct netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
++ if (!rxs) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err, "allocating rx ring page");
++ goto fail;
++ }
++ SHARED_RING_INIT(rxs);
++ FRONT_RING_INIT(&info->rx, rxs, PAGE_SIZE);
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(rxs));
++ if (err < 0) {
++ free_page((unsigned long)rxs);
++ goto fail;
++ }
++ info->rx_ring_ref = err;
++
++ memcpy(netdev->dev_addr, info->mac, ETH_ALEN);
++
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, netif_int, IRQF_SAMPLE_RANDOM, netdev->name,
++ netdev);
++ if (err < 0)
++ goto fail;
++ info->irq = err;
++
++ return 0;
++
++ fail:
++ return err;
++}
++
++/**
++ * Callback received when the backend's state changes.
++ */
++static void backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct netfront_info *np = dev->dev.driver_data;
++ struct net_device *netdev = np->netdev;
++
++ DPRINTK("%s\n", xenbus_strstate(backend_state));
++
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ case XenbusStateConnected:
++ case XenbusStateReconfiguring:
++ case XenbusStateReconfigured:
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateInitWait:
++ if (dev->state != XenbusStateInitialising)
++ break;
++ if (network_connect(netdev) != 0)
++ break;
++ xenbus_switch_state(dev, XenbusStateConnected);
++ send_fake_arp(netdev);
++ break;
++
++ case XenbusStateClosing:
++ xenbus_frontend_closed(dev);
++ break;
++ }
++}
++
++/** Send a packet on a net device to encourage switches to learn the
++ * MAC. We send a fake ARP request.
++ *
++ * @param dev device
++ * @return 0 on success, error code otherwise
++ */
++static void send_fake_arp(struct net_device *dev)
++{
++#ifdef CONFIG_INET
++ struct sk_buff *skb;
++ u32 src_ip, dst_ip;
++
++ dst_ip = INADDR_BROADCAST;
++ src_ip = inet_select_addr(dev, dst_ip, RT_SCOPE_LINK);
++
++ /* No IP? Then nothing to do. */
++ if (src_ip == 0)
++ return;
++
++ skb = arp_create(ARPOP_REPLY, ETH_P_ARP,
++ dst_ip, dev, src_ip,
++ /*dst_hw*/ NULL, /*src_hw*/ NULL,
++ /*target_hw*/ dev->dev_addr);
++ if (skb == NULL)
++ return;
++
++ dev_queue_xmit(skb);
++#endif
++}
++
++static inline int netfront_tx_slot_available(struct netfront_info *np)
++{
++ return ((np->tx.req_prod_pvt - np->tx.rsp_cons) <
++ (TX_MAX_TARGET - MAX_SKB_FRAGS - 2));
++}
++
++
++static inline void network_maybe_wake_tx(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++
++ if (unlikely(netif_queue_stopped(dev)) &&
++ netfront_tx_slot_available(np) &&
++ likely(netif_running(dev)) &&
++ netfront_check_accelerator_queue_ready(dev, np))
++ netif_wake_queue(dev);
++}
++
++
++int netfront_check_queue_ready(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++
++ return unlikely(netif_queue_stopped(dev)) &&
++ netfront_tx_slot_available(np) &&
++ likely(netif_running(dev));
++}
++EXPORT_SYMBOL(netfront_check_queue_ready);
++
++
++static int network_open(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++
++ memset(&np->stats, 0, sizeof(np->stats));
++ napi_enable(&np->napi);
++
++ spin_lock_bh(&np->rx_lock);
++ if (netfront_carrier_ok(np)) {
++ network_alloc_rx_buffers(dev);
++ np->rx.sring->rsp_event = np->rx.rsp_cons + 1;
++ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)){
++ netfront_accelerator_call_stop_napi_irq(np, dev);
++
++ netif_rx_schedule(dev, &np->napi);
++ }
++ }
++ spin_unlock_bh(&np->rx_lock);
++
++ network_maybe_wake_tx(dev);
++
++ return 0;
++}
++
++static void network_tx_buf_gc(struct net_device *dev)
++{
++ RING_IDX cons, prod;
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct sk_buff *skb;
++
++ BUG_ON(!netfront_carrier_ok(np));
++
++ do {
++ prod = np->tx.sring->rsp_prod;
++ rmb(); /* Ensure we see responses up to 'rp'. */
++
++ for (cons = np->tx.rsp_cons; cons != prod; cons++) {
++ struct netif_tx_response *txrsp;
++
++ txrsp = RING_GET_RESPONSE(&np->tx, cons);
++ if (txrsp->status == NETIF_RSP_NULL)
++ continue;
++
++ id = txrsp->id;
++ skb = np->tx_skbs[id];
++ if (unlikely(gnttab_query_foreign_access(
++ np->grant_tx_ref[id]) != 0)) {
++ printk(KERN_ALERT "network_tx_buf_gc: warning "
++ "-- grant still in use by backend "
++ "domain.\n");
++ BUG();
++ }
++ gnttab_end_foreign_access_ref(np->grant_tx_ref[id]);
++ gnttab_release_grant_reference(
++ &np->gref_tx_head, np->grant_tx_ref[id]);
++ np->grant_tx_ref[id] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->tx_skbs, id);
++ dev_kfree_skb_irq(skb);
++ }
++
++ np->tx.rsp_cons = prod;
++
++ /*
++ * Set a new event, then check for race with update of tx_cons.
++ * Note that it is essential to schedule a callback, no matter
++ * how few buffers are pending. Even if there is space in the
++ * transmit ring, higher layers may be blocked because too much
++ * data is outstanding: in such cases notification from Xen is
++ * likely to be the only kick that we'll get.
++ */
++ np->tx.sring->rsp_event =
++ prod + ((np->tx.sring->req_prod - prod) >> 1) + 1;
++ mb();
++ } while ((cons == prod) && (prod != np->tx.sring->rsp_prod));
++
++ network_maybe_wake_tx(dev);
++}
++
++static void rx_refill_timeout(unsigned long data)
++{
++ struct net_device *dev = (struct net_device *)data;
++ struct netfront_info *np = netdev_priv(dev);
++
++ netfront_accelerator_call_stop_napi_irq(np, dev);
++
++ netif_rx_schedule(dev, &np->napi);
++}
++
++static void network_alloc_rx_buffers(struct net_device *dev)
++{
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct sk_buff *skb;
++ struct page *page;
++ int i, batch_target, notify;
++ RING_IDX req_prod = np->rx.req_prod_pvt;
++ struct xen_memory_reservation reservation;
++ grant_ref_t ref;
++ unsigned long pfn;
++ void *vaddr;
++ int nr_flips;
++ netif_rx_request_t *req;
++
++ if (unlikely(!netfront_carrier_ok(np)))
++ return;
++
++ /*
++ * Allocate skbuffs greedily, even though we batch updates to the
++ * receive ring. This creates a less bursty demand on the memory
++ * allocator, so should reduce the chance of failed allocation requests
++ * both for ourself and for other kernel subsystems.
++ */
++ batch_target = np->rx_target - (req_prod - np->rx.rsp_cons);
++ for (i = skb_queue_len(&np->rx_batch); i < batch_target; i++) {
++ /*
++ * Allocate an skb and a page. Do not use __dev_alloc_skb as
++ * that will allocate page-sized buffers which is not
++ * necessary here.
++ * 16 bytes added as necessary headroom for netif_receive_skb.
++ */
++ skb = alloc_skb(RX_COPY_THRESHOLD + 16 + NET_IP_ALIGN,
++ GFP_ATOMIC | __GFP_NOWARN);
++ if (unlikely(!skb))
++ goto no_skb;
++
++ page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
++ if (!page) {
++ kfree_skb(skb);
++no_skb:
++ /* Any skbuffs queued for refill? Force them out. */
++ if (i != 0)
++ goto refill;
++ /* Could not allocate any skbuffs. Try again later. */
++ mod_timer(&np->rx_refill_timer,
++ jiffies + (HZ/10));
++ break;
++ }
++
++ skb_reserve(skb, 16 + NET_IP_ALIGN); /* mimic dev_alloc_skb() */
++ skb_shinfo(skb)->frags[0].page = page;
++ skb_shinfo(skb)->nr_frags = 1;
++ __skb_queue_tail(&np->rx_batch, skb);
++ }
++
++ /* Is the batch large enough to be worthwhile? */
++ if (i < (np->rx_target/2)) {
++ if (req_prod > np->rx.sring->req_prod)
++ goto push;
++ return;
++ }
++
++ /* Adjust our fill target if we risked running out of buffers. */
++ if (((req_prod - np->rx.sring->rsp_prod) < (np->rx_target / 4)) &&
++ ((np->rx_target *= 2) > np->rx_max_target))
++ np->rx_target = np->rx_max_target;
++
++ refill:
++ for (nr_flips = i = 0; ; i++) {
++ if ((skb = __skb_dequeue(&np->rx_batch)) == NULL)
++ break;
++
++ skb->dev = dev;
++
++ id = xennet_rxidx(req_prod + i);
++
++ BUG_ON(np->rx_skbs[id]);
++ np->rx_skbs[id] = skb;
++
++ ref = gnttab_claim_grant_reference(&np->gref_rx_head);
++ BUG_ON((signed short)ref < 0);
++ np->grant_rx_ref[id] = ref;
++
++ pfn = page_to_pfn(skb_shinfo(skb)->frags[0].page);
++ vaddr = page_address(skb_shinfo(skb)->frags[0].page);
++
++ req = RING_GET_REQUEST(&np->rx, req_prod + i);
++ if (!np->copying_receiver) {
++ gnttab_grant_foreign_transfer_ref(ref,
++ np->xbdev->otherend_id,
++ pfn);
++ np->rx_pfn_array[nr_flips] = pfn_to_mfn(pfn);
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remove this page before passing
++ * back to Xen. */
++ set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
++ MULTI_update_va_mapping(np->rx_mcl+i,
++ (unsigned long)vaddr,
++ __pte(0), 0);
++ }
++ nr_flips++;
++ } else {
++ gnttab_grant_foreign_access_ref(ref,
++ np->xbdev->otherend_id,
++ pfn_to_mfn(pfn),
++ 0);
++ }
++
++ req->id = id;
++ req->gref = ref;
++ }
++
++ if ( nr_flips != 0 ) {
++ /* Tell the ballon driver what is going on. */
++ balloon_update_driver_allowance(i);
++
++ set_xen_guest_handle(reservation.extent_start,
++ np->rx_pfn_array);
++ reservation.nr_extents = nr_flips;
++ reservation.extent_order = 0;
++ reservation.address_bits = 0;
++ reservation.domid = DOMID_SELF;
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* After all PTEs have been zapped, flush the TLB. */
++ np->rx_mcl[i-1].args[MULTI_UVMFLAGS_INDEX] =
++ UVMF_TLB_FLUSH|UVMF_ALL;
++
++ /* Give away a batch of pages. */
++ np->rx_mcl[i].op = __HYPERVISOR_memory_op;
++ np->rx_mcl[i].args[0] = XENMEM_decrease_reservation;
++ np->rx_mcl[i].args[1] = (unsigned long)&reservation;
++
++ /* Zap PTEs and give away pages in one big
++ * multicall. */
++ if (unlikely(HYPERVISOR_multicall(np->rx_mcl, i+1)))
++ BUG();
++
++ /* Check return status of HYPERVISOR_memory_op(). */
++ if (unlikely(np->rx_mcl[i].result != i))
++ panic("Unable to reduce memory reservation\n");
++ while (nr_flips--)
++ BUG_ON(np->rx_mcl[nr_flips].result);
++ } else {
++ if (HYPERVISOR_memory_op(XENMEM_decrease_reservation,
++ &reservation) != i)
++ panic("Unable to reduce memory reservation\n");
++ }
++ } else {
++ wmb();
++ }
++
++ /* Above is a suitable barrier to ensure backend will see requests. */
++ np->rx.req_prod_pvt = req_prod + i;
++ push:
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->rx, notify);
++ if (notify)
++ notify_remote_via_irq(np->irq);
++}
++
++static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev,
++ struct netif_tx_request *tx)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ char *data = skb->data;
++ unsigned long mfn;
++ RING_IDX prod = np->tx.req_prod_pvt;
++ int frags = skb_shinfo(skb)->nr_frags;
++ unsigned int offset = offset_in_page(data);
++ unsigned int len = skb_headlen(skb);
++ unsigned int id;
++ grant_ref_t ref;
++ int i;
++
++ while (len > PAGE_SIZE - offset) {
++ tx->size = PAGE_SIZE - offset;
++ tx->flags |= NETTXF_more_data;
++ len -= tx->size;
++ data += tx->size;
++ offset = 0;
++
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb_get(skb);
++ tx = RING_GET_REQUEST(&np->tx, prod++);
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
++
++ mfn = virt_to_mfn(data);
++ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
++ mfn, GTF_readonly);
++
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = offset;
++ tx->size = len;
++ tx->flags = 0;
++ }
++
++ for (i = 0; i < frags; i++) {
++ skb_frag_t *frag = skb_shinfo(skb)->frags + i;
++
++ tx->flags |= NETTXF_more_data;
++
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb_get(skb);
++ tx = RING_GET_REQUEST(&np->tx, prod++);
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
++
++ mfn = pfn_to_mfn(page_to_pfn(frag->page));
++ gnttab_grant_foreign_access_ref(ref, np->xbdev->otherend_id,
++ mfn, GTF_readonly);
++
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = frag->page_offset;
++ tx->size = frag->size;
++ tx->flags = 0;
++ }
++
++ np->tx.req_prod_pvt = prod;
++}
++
++static int network_start_xmit(struct sk_buff *skb, struct net_device *dev)
++{
++ unsigned short id;
++ struct netfront_info *np = netdev_priv(dev);
++ struct netif_tx_request *tx;
++ struct netif_extra_info *extra;
++ char *data = skb->data;
++ RING_IDX i;
++ grant_ref_t ref;
++ unsigned long mfn;
++ int notify;
++ int frags = skb_shinfo(skb)->nr_frags;
++ unsigned int offset = offset_in_page(data);
++ unsigned int len = skb_headlen(skb);
++
++ /* Check the fast path, if hooks are available */
++ if (np->accel_vif_state.hooks &&
++ np->accel_vif_state.hooks->start_xmit(skb, dev)) {
++ /* Fast path has sent this packet */
++ return 0;
++ }
++
++ frags += (offset + len + PAGE_SIZE - 1) / PAGE_SIZE;
++ if (unlikely(frags > MAX_SKB_FRAGS + 1)) {
++ printk(KERN_ALERT "xennet: skb rides the rocket: %d frags\n",
++ frags);
++ dump_stack();
++ goto drop;
++ }
++
++ spin_lock_irq(&np->tx_lock);
++
++ if (unlikely(!netfront_carrier_ok(np) ||
++ (frags > 1 && !xennet_can_sg(dev)) ||
++ netif_needs_gso(dev, skb))) {
++ spin_unlock_irq(&np->tx_lock);
++ goto drop;
++ }
++
++ i = np->tx.req_prod_pvt;
++
++ id = get_id_from_freelist(np->tx_skbs);
++ np->tx_skbs[id] = skb;
++
++ tx = RING_GET_REQUEST(&np->tx, i);
++
++ tx->id = id;
++ ref = gnttab_claim_grant_reference(&np->gref_tx_head);
++ BUG_ON((signed short)ref < 0);
++ mfn = virt_to_mfn(data);
++ gnttab_grant_foreign_access_ref(
++ ref, np->xbdev->otherend_id, mfn, GTF_readonly);
++ tx->gref = np->grant_tx_ref[id] = ref;
++ tx->offset = offset;
++ tx->size = len;
++
++ tx->flags = 0;
++ extra = NULL;
++
++ 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? */
++ tx->flags |= NETTXF_data_validated;
++#endif
++
++#if HAVE_TSO
++ if (skb_shinfo(skb)->gso_size) {
++ struct netif_extra_info *gso = (struct netif_extra_info *)
++ RING_GET_REQUEST(&np->tx, ++i);
++
++ if (extra)
++ extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE;
++ else
++ tx->flags |= NETTXF_extra_info;
++
++ gso->u.gso.size = skb_shinfo(skb)->gso_size;
++ gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
++ gso->u.gso.pad = 0;
++ gso->u.gso.features = 0;
++
++ gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
++ gso->flags = 0;
++ extra = gso;
++ }
++#endif
++
++ np->tx.req_prod_pvt = i + 1;
++
++ xennet_make_frags(skb, dev, tx);
++ tx->size = skb->len;
++
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&np->tx, notify);
++ if (notify)
++ notify_remote_via_irq(np->irq);
++
++ np->stats.tx_bytes += skb->len;
++ np->stats.tx_packets++;
++ dev->trans_start = jiffies;
++
++ /* Note: It is not safe to access skb after network_tx_buf_gc()! */
++ network_tx_buf_gc(dev);
++
++ if (!netfront_tx_slot_available(np))
++ netif_stop_queue(dev);
++
++ spin_unlock_irq(&np->tx_lock);
++
++ return 0;
++
++ drop:
++ np->stats.tx_dropped++;
++ dev_kfree_skb(skb);
++ return 0;
++}
++
++static irqreturn_t netif_int(int irq, void *dev_id)
++{
++ struct net_device *dev = dev_id;
++ struct netfront_info *np = netdev_priv(dev);
++ unsigned long flags;
++
++ spin_lock_irqsave(&np->tx_lock, flags);
++
++ if (likely(netfront_carrier_ok(np))) {
++ network_tx_buf_gc(dev);
++ /* Under tx_lock: protects access to rx shared-ring indexes. */
++ if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx)) {
++ netfront_accelerator_call_stop_napi_irq(np, dev);
++
++ netif_rx_schedule(dev, &np->napi);
++ dev->last_rx = jiffies;
++ }
++ }
++
++ spin_unlock_irqrestore(&np->tx_lock, flags);
++
++ return IRQ_HANDLED;
++}
++
++static void xennet_move_rx_slot(struct netfront_info *np, struct sk_buff *skb,
++ grant_ref_t ref)
++{
++ int new = xennet_rxidx(np->rx.req_prod_pvt);
++
++ BUG_ON(np->rx_skbs[new]);
++ np->rx_skbs[new] = skb;
++ np->grant_rx_ref[new] = ref;
++ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->id = new;
++ RING_GET_REQUEST(&np->rx, np->rx.req_prod_pvt)->gref = ref;
++ np->rx.req_prod_pvt++;
++}
++
++int xennet_get_extras(struct netfront_info *np,
++ struct netif_extra_info *extras, RING_IDX rp)
++
++{
++ struct netif_extra_info *extra;
++ RING_IDX cons = np->rx.rsp_cons;
++ int err = 0;
++
++ do {
++ struct sk_buff *skb;
++ grant_ref_t ref;
++
++ if (unlikely(cons + 1 == rp)) {
++ if (net_ratelimit())
++ WPRINTK("Missing extra info\n");
++ err = -EBADR;
++ break;
++ }
++
++ extra = (struct netif_extra_info *)
++ RING_GET_RESPONSE(&np->rx, ++cons);
++
++ if (unlikely(!extra->type ||
++ extra->type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
++ if (net_ratelimit())
++ WPRINTK("Invalid extra type: %d\n",
++ extra->type);
++ err = -EINVAL;
++ } else {
++ memcpy(&extras[extra->type - 1], extra,
++ sizeof(*extra));
++ }
++
++ skb = xennet_get_rx_skb(np, cons);
++ ref = xennet_get_rx_ref(np, cons);
++ xennet_move_rx_slot(np, skb, ref);
++ } while (extra->flags & XEN_NETIF_EXTRA_FLAG_MORE);
++
++ np->rx.rsp_cons = cons;
++ return err;
++}
++
++static int xennet_get_responses(struct netfront_info *np,
++ struct netfront_rx_info *rinfo, RING_IDX rp,
++ struct sk_buff_head *list,
++ int *pages_flipped_p)
++{
++ int pages_flipped = *pages_flipped_p;
++ struct mmu_update *mmu;
++ struct multicall_entry *mcl;
++ struct netif_rx_response *rx = &rinfo->rx;
++ struct netif_extra_info *extras = rinfo->extras;
++ RING_IDX cons = np->rx.rsp_cons;
++ struct sk_buff *skb = xennet_get_rx_skb(np, cons);
++ grant_ref_t ref = xennet_get_rx_ref(np, cons);
++ int max = MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD);
++ int frags = 1;
++ int err = 0;
++ unsigned long ret;
++
++ if (rx->flags & NETRXF_extra_info) {
++ err = xennet_get_extras(np, extras, rp);
++ cons = np->rx.rsp_cons;
++ }
++
++ for (;;) {
++ unsigned long mfn;
++
++ if (unlikely(rx->status < 0 ||
++ rx->offset + rx->status > PAGE_SIZE)) {
++ if (net_ratelimit())
++ WPRINTK("rx->offset: %x, size: %u\n",
++ rx->offset, rx->status);
++ xennet_move_rx_slot(np, skb, ref);
++ err = -EINVAL;
++ goto next;
++ }
++
++ /*
++ * This definitely indicates a bug, either in this driver or in
++ * the backend driver. In future this should flag the bad
++ * situation to the system controller to reboot the backed.
++ */
++ if (ref == GRANT_INVALID_REF) {
++ if (net_ratelimit())
++ WPRINTK("Bad rx response id %d.\n", rx->id);
++ err = -EINVAL;
++ goto next;
++ }
++
++ if (!np->copying_receiver) {
++ /* Memory pressure, insufficient buffer
++ * headroom, ... */
++ if (!(mfn = gnttab_end_foreign_transfer_ref(ref))) {
++ if (net_ratelimit())
++ WPRINTK("Unfulfilled rx req "
++ "(id=%d, st=%d).\n",
++ rx->id, rx->status);
++ xennet_move_rx_slot(np, skb, ref);
++ err = -ENOMEM;
++ goto next;
++ }
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remap the page. */
++ struct page *page =
++ skb_shinfo(skb)->frags[0].page;
++ unsigned long pfn = page_to_pfn(page);
++ void *vaddr = page_address(page);
++
++ mcl = np->rx_mcl + pages_flipped;
++ mmu = np->rx_mmu + pages_flipped;
++
++ MULTI_update_va_mapping(mcl,
++ (unsigned long)vaddr,
++ pfn_pte_ma(mfn,
++ PAGE_KERNEL),
++ 0);
++ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
++ | MMU_MACHPHYS_UPDATE;
++ mmu->val = pfn;
++
++ set_phys_to_machine(pfn, mfn);
++ }
++ pages_flipped++;
++ } else {
++ ret = gnttab_end_foreign_access_ref(ref);
++ BUG_ON(!ret);
++ }
++
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++
++ __skb_queue_tail(list, skb);
++
++next:
++ if (!(rx->flags & NETRXF_more_data))
++ break;
++
++ if (cons + frags == rp) {
++ if (net_ratelimit())
++ WPRINTK("Need more frags\n");
++ err = -ENOENT;
++ break;
++ }
++
++ rx = RING_GET_RESPONSE(&np->rx, cons + frags);
++ skb = xennet_get_rx_skb(np, cons + frags);
++ ref = xennet_get_rx_ref(np, cons + frags);
++ frags++;
++ }
++
++ if (unlikely(frags > max)) {
++ if (net_ratelimit())
++ WPRINTK("Too many frags\n");
++ err = -E2BIG;
++ }
++
++ if (unlikely(err))
++ np->rx.rsp_cons = cons + frags;
++
++ *pages_flipped_p = pages_flipped;
++
++ return err;
++}
++
++static RING_IDX xennet_fill_frags(struct netfront_info *np,
++ struct sk_buff *skb,
++ struct sk_buff_head *list)
++{
++ struct skb_shared_info *shinfo = skb_shinfo(skb);
++ int nr_frags = shinfo->nr_frags;
++ RING_IDX cons = np->rx.rsp_cons;
++ skb_frag_t *frag = shinfo->frags + nr_frags;
++ struct sk_buff *nskb;
++
++ while ((nskb = __skb_dequeue(list))) {
++ struct netif_rx_response *rx =
++ RING_GET_RESPONSE(&np->rx, ++cons);
++
++ frag->page = skb_shinfo(nskb)->frags[0].page;
++ frag->page_offset = rx->offset;
++ frag->size = rx->status;
++
++ skb->data_len += rx->status;
++
++ skb_shinfo(nskb)->nr_frags = 0;
++ kfree_skb(nskb);
++
++ frag++;
++ nr_frags++;
++ }
++
++ shinfo->nr_frags = nr_frags;
++ return cons;
++}
++
++static int xennet_set_skb_gso(struct sk_buff *skb,
++ struct netif_extra_info *gso)
++{
++ if (!gso->u.gso.size) {
++ if (net_ratelimit())
++ WPRINTK("GSO size must not be zero.\n");
++ return -EINVAL;
++ }
++
++ /* Currently only TCPv4 S.O. is supported. */
++ if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
++ if (net_ratelimit())
++ WPRINTK("Bad GSO type %d.\n", gso->u.gso.type);
++ return -EINVAL;
++ }
++
++#if HAVE_TSO
++ skb_shinfo(skb)->gso_size = gso->u.gso.size;
++#if HAVE_GSO
++ skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
++
++ /* Header must be checked, and gso_segs computed. */
++ skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
++#endif
++ skb_shinfo(skb)->gso_segs = 0;
++
++ return 0;
++#else
++ if (net_ratelimit())
++ WPRINTK("GSO unsupported by this kernel.\n");
++ return -EINVAL;
++#endif
++}
++
++static int netif_poll(struct napi_struct *napi, int budget)
++{
++ struct netfront_info *np = container_of(napi, struct netfront_info, napi);
++ struct net_device *dev = np->netdev;
++ struct sk_buff *skb;
++ struct netfront_rx_info rinfo;
++ struct netif_rx_response *rx = &rinfo.rx;
++ struct netif_extra_info *extras = rinfo.extras;
++ RING_IDX i, rp;
++ struct multicall_entry *mcl;
++ int work_done, more_to_do = 1, accel_more_to_do = 1;
++ struct sk_buff_head rxq;
++ struct sk_buff_head errq;
++ struct sk_buff_head tmpq;
++ unsigned long flags;
++ unsigned int len;
++ int pages_flipped = 0;
++ int err;
++
++ spin_lock(&np->rx_lock); /* no need for spin_lock_bh() in ->poll() */
++
++ if (unlikely(!netfront_carrier_ok(np))) {
++ spin_unlock(&np->rx_lock);
++ return 0;
++ }
++
++ skb_queue_head_init(&rxq);
++ skb_queue_head_init(&errq);
++ skb_queue_head_init(&tmpq);
++
++ rp = np->rx.sring->rsp_prod;
++ rmb(); /* Ensure we see queued responses up to 'rp'. */
++
++ i = np->rx.rsp_cons;
++ work_done = 0;
++ while ((i != rp) && (work_done < budget)) {
++ memcpy(rx, RING_GET_RESPONSE(&np->rx, i), sizeof(*rx));
++ memset(extras, 0, sizeof(rinfo.extras));
++
++ err = xennet_get_responses(np, &rinfo, rp, &tmpq,
++ &pages_flipped);
++
++ if (unlikely(err)) {
++err:
++ while ((skb = __skb_dequeue(&tmpq)))
++ __skb_queue_tail(&errq, skb);
++ np->stats.rx_errors++;
++ i = np->rx.rsp_cons;
++ continue;
++ }
++
++ skb = __skb_dequeue(&tmpq);
++
++ if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
++ struct netif_extra_info *gso;
++ gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
++
++ if (unlikely(xennet_set_skb_gso(skb, gso))) {
++ __skb_queue_head(&tmpq, skb);
++ np->rx.rsp_cons += skb_queue_len(&tmpq);
++ goto err;
++ }
++ }
++
++ NETFRONT_SKB_CB(skb)->page = skb_shinfo(skb)->frags[0].page;
++ NETFRONT_SKB_CB(skb)->offset = rx->offset;
++
++ len = rx->status;
++ if (len > RX_COPY_THRESHOLD)
++ len = RX_COPY_THRESHOLD;
++ skb_put(skb, len);
++
++ if (rx->status > len) {
++ skb_shinfo(skb)->frags[0].page_offset =
++ rx->offset + len;
++ skb_shinfo(skb)->frags[0].size = rx->status - len;
++ skb->data_len = rx->status - len;
++ } else {
++ skb_shinfo(skb)->frags[0].page = NULL;
++ skb_shinfo(skb)->nr_frags = 0;
++ }
++
++ i = xennet_fill_frags(np, skb, &tmpq);
++
++ /*
++ * Truesize must approximates the size of true data plus
++ * any supervisor overheads. Adding hypervisor overheads
++ * has been shown to significantly reduce achievable
++ * bandwidth with the default receive buffer size. It is
++ * therefore not wise to account for it here.
++ *
++ * After alloc_skb(RX_COPY_THRESHOLD), truesize is set to
++ * RX_COPY_THRESHOLD + the supervisor overheads. Here, we
++ * add the size of the data pulled in xennet_fill_frags().
++ *
++ * We also adjust for any unused space in the main data
++ * area by subtracting (RX_COPY_THRESHOLD - len). This is
++ * especially important with drivers which split incoming
++ * packets into header and data, using only 66 bytes of
++ * the main data area (see the e1000 driver for example.)
++ * On such systems, without this last adjustement, our
++ * achievable receive throughout using the standard receive
++ * buffer size was cut by 25%(!!!).
++ */
++ skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len);
++ skb->len += skb->data_len;
++
++ /*
++ * Old backends do not assert data_validated but we
++ * can infer it from csum_blank so test both flags.
++ */
++ if (rx->flags & (NETRXF_data_validated|NETRXF_csum_blank))
++ skb->ip_summed = CHECKSUM_UNNECESSARY;
++ else
++ skb->ip_summed = CHECKSUM_NONE;
++#ifdef CONFIG_XEN
++ skb->proto_data_valid = (skb->ip_summed != CHECKSUM_NONE);
++ skb->proto_csum_blank = !!(rx->flags & NETRXF_csum_blank);
++#endif
++ np->stats.rx_packets++;
++ np->stats.rx_bytes += skb->len;
++
++ __skb_queue_tail(&rxq, skb);
++
++ np->rx.rsp_cons = ++i;
++ work_done++;
++ }
++
++ if (pages_flipped) {
++ /* Some pages are no longer absent... */
++ balloon_update_driver_allowance(-pages_flipped);
++
++ /* Do all the remapping work and M2P updates. */
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ mcl = np->rx_mcl + pages_flipped;
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)np->rx_mmu;
++ mcl->args[1] = pages_flipped;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
++ err = HYPERVISOR_multicall_check(np->rx_mcl,
++ pages_flipped + 1,
++ NULL);
++ BUG_ON(err);
++ }
++ }
++
++ __skb_queue_purge(&errq);
++
++ while ((skb = __skb_dequeue(&rxq)) != NULL) {
++ struct page *page = NETFRONT_SKB_CB(skb)->page;
++ void *vaddr = page_address(page);
++ unsigned offset = NETFRONT_SKB_CB(skb)->offset;
++
++ memcpy(skb->data, vaddr + offset, skb_headlen(skb));
++
++ if (page != skb_shinfo(skb)->frags[0].page)
++ __free_page(page);
++
++ /* Ethernet work: Delayed to here as it peeks the header. */
++ skb->protocol = eth_type_trans(skb, dev);
++
++ /* Pass it up. */
++ netif_receive_skb(skb);
++ dev->last_rx = jiffies;
++ }
++
++ /* If we get a callback with very few responses, reduce fill target. */
++ /* NB. Note exponential increase, linear decrease. */
++ if (((np->rx.req_prod_pvt - np->rx.sring->rsp_prod) >
++ ((3*np->rx_target) / 4)) &&
++ (--np->rx_target < np->rx_min_target))
++ np->rx_target = np->rx_min_target;
++
++ network_alloc_rx_buffers(dev);
++
++ if (work_done < budget) {
++ /* there's some spare capacity, try the accelerated path */
++ int accel_budget = budget - work_done;
++ int accel_budget_start = accel_budget;
++
++ if (np->accel_vif_state.hooks) {
++ accel_more_to_do =
++ np->accel_vif_state.hooks->netdev_poll
++ (dev, &accel_budget);
++ work_done += (accel_budget_start - accel_budget);
++ } else
++ accel_more_to_do = 0;
++ }
++
++ if (work_done < budget) {
++ local_irq_save(flags);
++
++ RING_FINAL_CHECK_FOR_RESPONSES(&np->rx, more_to_do);
++
++ if (!more_to_do && !accel_more_to_do &&
++ np->accel_vif_state.hooks) {
++ /*
++ * Slow path has nothing more to do, see if
++ * fast path is likewise
++ */
++ accel_more_to_do =
++ np->accel_vif_state.hooks->start_napi_irq(dev);
++ }
++
++ if (!more_to_do && !accel_more_to_do)
++ __netif_rx_complete(dev, napi);
++
++ local_irq_restore(flags);
++ }
++
++ spin_unlock(&np->rx_lock);
++
++ return work_done;
++}
++
++static void netif_release_tx_bufs(struct netfront_info *np)
++{
++ struct sk_buff *skb;
++ int i;
++
++ for (i = 1; i <= NET_TX_RING_SIZE; i++) {
++ if ((unsigned long)np->tx_skbs[i] < PAGE_OFFSET)
++ continue;
++
++ skb = np->tx_skbs[i];
++ gnttab_end_foreign_access_ref(np->grant_tx_ref[i]);
++ gnttab_release_grant_reference(
++ &np->gref_tx_head, np->grant_tx_ref[i]);
++ np->grant_tx_ref[i] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->tx_skbs, i);
++ dev_kfree_skb_irq(skb);
++ }
++}
++
++static void netif_release_rx_bufs_flip(struct netfront_info *np)
++{
++ struct mmu_update *mmu = np->rx_mmu;
++ struct multicall_entry *mcl = np->rx_mcl;
++ struct sk_buff_head free_list;
++ struct sk_buff *skb;
++ unsigned long mfn;
++ int xfer = 0, noxfer = 0, unused = 0;
++ int id, ref, rc;
++
++ skb_queue_head_init(&free_list);
++
++ spin_lock_bh(&np->rx_lock);
++
++ for (id = 0; id < NET_RX_RING_SIZE; id++) {
++ if ((ref = np->grant_rx_ref[id]) == GRANT_INVALID_REF) {
++ unused++;
++ continue;
++ }
++
++ skb = np->rx_skbs[id];
++ mfn = gnttab_end_foreign_transfer_ref(ref);
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++ np->grant_rx_ref[id] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->rx_skbs, id);
++
++ if (0 == mfn) {
++ struct page *page = skb_shinfo(skb)->frags[0].page;
++ balloon_release_driver_page(page);
++ skb_shinfo(skb)->nr_frags = 0;
++ dev_kfree_skb(skb);
++ noxfer++;
++ continue;
++ }
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Remap the page. */
++ struct page *page = skb_shinfo(skb)->frags[0].page;
++ unsigned long pfn = page_to_pfn(page);
++ void *vaddr = page_address(page);
++
++ MULTI_update_va_mapping(mcl, (unsigned long)vaddr,
++ pfn_pte_ma(mfn, PAGE_KERNEL),
++ 0);
++ mcl++;
++ mmu->ptr = ((maddr_t)mfn << PAGE_SHIFT)
++ | MMU_MACHPHYS_UPDATE;
++ mmu->val = pfn;
++ mmu++;
++
++ set_phys_to_machine(pfn, mfn);
++ }
++ __skb_queue_tail(&free_list, skb);
++ xfer++;
++ }
++
++ DPRINTK("%s: %d xfer, %d noxfer, %d unused\n",
++ __FUNCTION__, xfer, noxfer, unused);
++
++ if (xfer) {
++ /* Some pages are no longer absent... */
++ balloon_update_driver_allowance(-xfer);
++
++ if (!xen_feature(XENFEAT_auto_translated_physmap)) {
++ /* Do all the remapping work and M2P updates. */
++ mcl->op = __HYPERVISOR_mmu_update;
++ mcl->args[0] = (unsigned long)np->rx_mmu;
++ mcl->args[1] = mmu - np->rx_mmu;
++ mcl->args[2] = 0;
++ mcl->args[3] = DOMID_SELF;
++ mcl++;
++ rc = HYPERVISOR_multicall_check(
++ np->rx_mcl, mcl - np->rx_mcl, NULL);
++ BUG_ON(rc);
++ }
++ }
++
++ __skb_queue_purge(&free_list);
++
++ spin_unlock_bh(&np->rx_lock);
++}
++
++static void netif_release_rx_bufs_copy(struct netfront_info *np)
++{
++ struct sk_buff *skb;
++ int i, ref;
++ int busy = 0, inuse = 0;
++
++ spin_lock_bh(&np->rx_lock);
++
++ for (i = 0; i < NET_RX_RING_SIZE; i++) {
++ ref = np->grant_rx_ref[i];
++
++ if (ref == GRANT_INVALID_REF)
++ continue;
++
++ inuse++;
++
++ skb = np->rx_skbs[i];
++
++ if (!gnttab_end_foreign_access_ref(ref))
++ {
++ busy++;
++ continue;
++ }
++
++ gnttab_release_grant_reference(&np->gref_rx_head, ref);
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
++ add_id_to_freelist(np->rx_skbs, i);
++
++ dev_kfree_skb(skb);
++ }
++
++ if (busy)
++ DPRINTK("%s: Unable to release %d of %d inuse grant references out of %ld total.\n",
++ __FUNCTION__, busy, inuse, NET_RX_RING_SIZE);
++
++ spin_unlock_bh(&np->rx_lock);
++}
++
++static int network_close(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ netif_stop_queue(np->netdev);
++ napi_disable(&np->napi);
++ return 0;
++}
++
++
++static struct net_device_stats *network_get_stats(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++
++ netfront_accelerator_call_get_stats(np, dev);
++ return &np->stats;
++}
++
++static int xennet_set_mac_address(struct net_device *dev, void *p)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ struct sockaddr *addr = p;
++
++ if (netif_running(dev))
++ return -EBUSY;
++
++ if (!is_valid_ether_addr(addr->sa_data))
++ return -EADDRNOTAVAIL;
++
++ memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
++ memcpy(np->mac, addr->sa_data, ETH_ALEN);
++
++ return 0;
++}
++
++static int xennet_change_mtu(struct net_device *dev, int mtu)
++{
++ int max = xennet_can_sg(dev) ? 65535 - ETH_HLEN : ETH_DATA_LEN;
++
++ if (mtu > max)
++ return -EINVAL;
++ dev->mtu = mtu;
++ return 0;
++}
++
++static int xennet_set_sg(struct net_device *dev, u32 data)
++{
++ if (data) {
++ struct netfront_info *np = netdev_priv(dev);
++ int val;
++
++ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend, "feature-sg",
++ "%d", &val) < 0)
++ val = 0;
++ if (!val)
++ return -ENOSYS;
++ } else if (dev->mtu > ETH_DATA_LEN)
++ dev->mtu = ETH_DATA_LEN;
++
++ return ethtool_op_set_sg(dev, data);
++}
++
++static int xennet_set_tso(struct net_device *dev, u32 data)
++{
++ if (data) {
++ struct netfront_info *np = netdev_priv(dev);
++ int val;
++
++ if (xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-gso-tcpv4", "%d", &val) < 0)
++ val = 0;
++ if (!val)
++ return -ENOSYS;
++ }
++
++ return ethtool_op_set_tso(dev, data);
++}
++
++static void xennet_set_features(struct net_device *dev)
++{
++ dev_disable_gso_features(dev);
++ xennet_set_sg(dev, 0);
++
++ /* We need checksum offload to enable scatter/gather and TSO. */
++ if (!(dev->features & NETIF_F_IP_CSUM))
++ return;
++
++ if (xennet_set_sg(dev, 1))
++ return;
++
++ /* Before 2.6.9 TSO seems to be unreliable so do not enable it
++ * on older kernels.
++ */
++ if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,9))
++ xennet_set_tso(dev, 1);
++}
++
++static int network_connect(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ int i, requeue_idx, err;
++ struct sk_buff *skb;
++ grant_ref_t ref;
++ netif_rx_request_t *req;
++ unsigned int feature_rx_copy, feature_rx_flip;
++
++ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-rx-copy", "%u", &feature_rx_copy);
++ if (err != 1)
++ feature_rx_copy = 0;
++ err = xenbus_scanf(XBT_NIL, np->xbdev->otherend,
++ "feature-rx-flip", "%u", &feature_rx_flip);
++ if (err != 1)
++ feature_rx_flip = 1;
++
++ /*
++ * Copy packets on receive path if:
++ * (a) This was requested by user, and the backend supports it; or
++ * (b) Flipping was requested, but this is unsupported by the backend.
++ */
++ np->copying_receiver = ((MODPARM_rx_copy && feature_rx_copy) ||
++ (MODPARM_rx_flip && !feature_rx_flip));
++
++ err = talk_to_backend(np->xbdev, np);
++ if (err)
++ return err;
++
++ xennet_set_features(dev);
++
++ DPRINTK("device %s has %sing receive path.\n",
++ dev->name, np->copying_receiver ? "copy" : "flipp");
++
++ spin_lock_bh(&np->rx_lock);
++ spin_lock_irq(&np->tx_lock);
++
++ /*
++ * Recovery procedure:
++ * NB. Freelist index entries are always going to be less than
++ * PAGE_OFFSET, whereas pointers to skbs will always be equal or
++ * greater than PAGE_OFFSET: we use this property to distinguish
++ * them.
++ */
++
++ /* Step 1: Discard all pending TX packet fragments. */
++ netif_release_tx_bufs(np);
++
++ /* Step 2: Rebuild the RX buffer freelist and the RX ring itself. */
++ for (requeue_idx = 0, i = 0; i < NET_RX_RING_SIZE; i++) {
++ if (!np->rx_skbs[i])
++ continue;
++
++ skb = np->rx_skbs[requeue_idx] = xennet_get_rx_skb(np, i);
++ ref = np->grant_rx_ref[requeue_idx] = xennet_get_rx_ref(np, i);
++ req = RING_GET_REQUEST(&np->rx, requeue_idx);
++
++ if (!np->copying_receiver) {
++ gnttab_grant_foreign_transfer_ref(
++ ref, np->xbdev->otherend_id,
++ page_to_pfn(skb_shinfo(skb)->frags->page));
++ } else {
++ gnttab_grant_foreign_access_ref(
++ ref, np->xbdev->otherend_id,
++ pfn_to_mfn(page_to_pfn(skb_shinfo(skb)->
++ frags->page)),
++ 0);
++ }
++ req->gref = ref;
++ req->id = requeue_idx;
++
++ requeue_idx++;
++ }
++
++ np->rx.req_prod_pvt = requeue_idx;
++
++ /*
++ * Step 3: All public and private state should now be sane. Get
++ * ready to start sending and receiving packets and give the driver
++ * domain a kick because we've probably just requeued some
++ * packets.
++ */
++ netfront_carrier_on(np);
++ notify_remote_via_irq(np->irq);
++ network_tx_buf_gc(dev);
++ network_alloc_rx_buffers(dev);
++
++ spin_unlock_irq(&np->tx_lock);
++ spin_unlock_bh(&np->rx_lock);
++
++ return 0;
++}
++
++static void netif_uninit(struct net_device *dev)
++{
++ struct netfront_info *np = netdev_priv(dev);
++ netif_release_tx_bufs(np);
++ if (np->copying_receiver)
++ netif_release_rx_bufs_copy(np);
++ else
++ netif_release_rx_bufs_flip(np);
++ gnttab_free_grant_references(np->gref_tx_head);
++ gnttab_free_grant_references(np->gref_rx_head);
++}
++
++static struct ethtool_ops network_ethtool_ops =
++{
++ .get_tx_csum = ethtool_op_get_tx_csum,
++ .set_tx_csum = ethtool_op_set_tx_csum,
++ .get_sg = ethtool_op_get_sg,
++ .set_sg = xennet_set_sg,
++#if HAVE_TSO
++ .get_tso = ethtool_op_get_tso,
++ .set_tso = xennet_set_tso,
++#endif
++ .get_link = ethtool_op_get_link,
++};
++
++#ifdef CONFIG_SYSFS
++static ssize_t show_rxbuf_min(struct device *dev,
++ struct device_attribute *attr, char *buf)
++{
++ struct netfront_info *info = netdev_priv(to_net_dev(dev));
++
++ return sprintf(buf, "%u\n", info->rx_min_target);
++}
++
++static ssize_t store_rxbuf_min(struct device *dev,
++ struct device_attribute *attr,
++ const char *buf, size_t len)
++{
++ struct net_device *netdev = to_net_dev(dev);
++ struct netfront_info *np = netdev_priv(netdev);
++ char *endp;
++ unsigned long target;
++
++ if (!capable(CAP_NET_ADMIN))
++ return -EPERM;
++
++ target = simple_strtoul(buf, &endp, 0);
++ if (endp == buf)
++ return -EBADMSG;
++
++ if (target < RX_MIN_TARGET)
++ target = RX_MIN_TARGET;
++ if (target > RX_MAX_TARGET)
++ target = RX_MAX_TARGET;
++
++ spin_lock_bh(&np->rx_lock);
++ if (target > np->rx_max_target)
++ np->rx_max_target = target;
++ np->rx_min_target = target;
++ if (target > np->rx_target)
++ np->rx_target = target;
++
++ network_alloc_rx_buffers(netdev);
++
++ spin_unlock_bh(&np->rx_lock);
++ return len;
++}
++
++static ssize_t show_rxbuf_max(struct device *dev,
++ struct device_attribute *attr, char *buf)
++{
++ struct netfront_info *info = netdev_priv(to_net_dev(dev));
++
++ return sprintf(buf, "%u\n", info->rx_max_target);
++}
++
++static ssize_t store_rxbuf_max(struct device *dev,
++ struct device_attribute *attr,
++ const char *buf, size_t len)
++{
++ struct net_device *netdev = to_net_dev(dev);
++ struct netfront_info *np = netdev_priv(netdev);
++ char *endp;
++ unsigned long target;
++
++ if (!capable(CAP_NET_ADMIN))
++ return -EPERM;
++
++ target = simple_strtoul(buf, &endp, 0);
++ if (endp == buf)
++ return -EBADMSG;
++
++ if (target < RX_MIN_TARGET)
++ target = RX_MIN_TARGET;
++ if (target > RX_MAX_TARGET)
++ target = RX_MAX_TARGET;
++
++ spin_lock_bh(&np->rx_lock);
++ if (target < np->rx_min_target)
++ np->rx_min_target = target;
++ np->rx_max_target = target;
++ if (target < np->rx_target)
++ np->rx_target = target;
++
++ network_alloc_rx_buffers(netdev);
++
++ spin_unlock_bh(&np->rx_lock);
++ return len;
++}
++
++static ssize_t show_rxbuf_cur(struct device *dev,
++ struct device_attribute *attr, char *buf)
++{
++ struct netfront_info *info = netdev_priv(to_net_dev(dev));
++
++ return sprintf(buf, "%u\n", info->rx_target);
++}
++
++static struct device_attribute xennet_attrs[] = {
++ __ATTR(rxbuf_min, S_IRUGO|S_IWUSR, show_rxbuf_min, store_rxbuf_min),
++ __ATTR(rxbuf_max, S_IRUGO|S_IWUSR, show_rxbuf_max, store_rxbuf_max),
++ __ATTR(rxbuf_cur, S_IRUGO, show_rxbuf_cur, NULL),
++};
++
++static int xennet_sysfs_addif(struct net_device *netdev)
++{
++ int i;
++ int error = 0;
++
++ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++) {
++ error = device_create_file(&netdev->dev,
++ &xennet_attrs[i]);
++ if (error)
++ goto fail;
++ }
++ return 0;
++
++ fail:
++ while (--i >= 0)
++ device_remove_file(&netdev->dev, &xennet_attrs[i]);
++ return error;
++}
++
++static void xennet_sysfs_delif(struct net_device *netdev)
++{
++ int i;
++
++ for (i = 0; i < ARRAY_SIZE(xennet_attrs); i++)
++ device_remove_file(&netdev->dev, &xennet_attrs[i]);
++}
++
++#endif /* CONFIG_SYSFS */
++
++
++/*
++ * Nothing to do here. Virtual interface is point-to-point and the
++ * physical interface is probably promiscuous anyway.
++ */
++static void network_set_multicast_list(struct net_device *dev)
++{
++}
++
++static struct net_device * __devinit create_netdev(struct xenbus_device *dev)
++{
++ int i, err = 0;
++ struct net_device *netdev = NULL;
++ struct netfront_info *np = NULL;
++
++ netdev = alloc_etherdev(sizeof(struct netfront_info));
++ if (!netdev) {
++ printk(KERN_WARNING "%s> alloc_etherdev failed.\n",
++ __FUNCTION__);
++ return ERR_PTR(-ENOMEM);
++ }
++
++ np = netdev_priv(netdev);
++ np->xbdev = dev;
++
++ spin_lock_init(&np->tx_lock);
++ spin_lock_init(&np->rx_lock);
++
++ init_accelerator_vif(np, dev);
++
++ skb_queue_head_init(&np->rx_batch);
++ np->rx_target = RX_DFL_MIN_TARGET;
++ np->rx_min_target = RX_DFL_MIN_TARGET;
++ np->rx_max_target = RX_MAX_TARGET;
++
++ init_timer(&np->rx_refill_timer);
++ np->rx_refill_timer.data = (unsigned long)netdev;
++ np->rx_refill_timer.function = rx_refill_timeout;
++
++ /* Initialise {tx,rx}_skbs as a free chain containing every entry. */
++ for (i = 0; i <= NET_TX_RING_SIZE; i++) {
++ np->tx_skbs[i] = (void *)((unsigned long) i+1);
++ np->grant_tx_ref[i] = GRANT_INVALID_REF;
++ }
++
++ for (i = 0; i < NET_RX_RING_SIZE; i++) {
++ np->rx_skbs[i] = NULL;
++ np->grant_rx_ref[i] = GRANT_INVALID_REF;
++ }
++
++ /* A grant for every tx ring slot */
++ if (gnttab_alloc_grant_references(TX_MAX_TARGET,
++ &np->gref_tx_head) < 0) {
++ printk(KERN_ALERT "#### netfront can't alloc tx grant refs\n");
++ err = -ENOMEM;
++ goto exit;
++ }
++ /* A grant for every rx ring slot */
++ if (gnttab_alloc_grant_references(RX_MAX_TARGET,
++ &np->gref_rx_head) < 0) {
++ printk(KERN_ALERT "#### netfront can't alloc rx grant refs\n");
++ err = -ENOMEM;
++ goto exit_free_tx;
++ }
++
++ netdev->open = network_open;
++ netdev->hard_start_xmit = network_start_xmit;
++ netdev->stop = network_close;
++ netdev->get_stats = network_get_stats;
++ netif_napi_add(netdev, &np->napi, netif_poll, 64);
++ netdev->set_multicast_list = network_set_multicast_list;
++ netdev->uninit = netif_uninit;
++ netdev->set_mac_address = xennet_set_mac_address;
++ netdev->change_mtu = xennet_change_mtu;
++ netdev->features = NETIF_F_IP_CSUM;
++
++ SET_ETHTOOL_OPS(netdev, &network_ethtool_ops);
++ SET_NETDEV_DEV(netdev, &dev->dev);
++
++ np->netdev = netdev;
++
++ netfront_carrier_off(np);
++
++ return netdev;
++
++ exit_free_tx:
++ gnttab_free_grant_references(np->gref_tx_head);
++ exit:
++ free_netdev(netdev);
++ return ERR_PTR(err);
++}
++
++#ifdef CONFIG_INET
++/*
++ * We use this notifier to send out a fake ARP reply to reset switches and
++ * router ARP caches when an IP interface is brought up on a VIF.
++ */
++static int
++inetdev_notify(struct notifier_block *this, unsigned long event, void *ptr)
++{
++ struct in_ifaddr *ifa = (struct in_ifaddr *)ptr;
++ struct net_device *dev = ifa->ifa_dev->dev;
++
++ /* UP event and is it one of our devices? */
++ if (event == NETDEV_UP && dev->open == network_open)
++ send_fake_arp(dev);
++
++ return NOTIFY_DONE;
++}
++
++static struct notifier_block notifier_inetdev = {
++ .notifier_call = inetdev_notify,
++ .next = NULL,
++ .priority = 0
++};
++#endif
++
++
++static void netif_disconnect_backend(struct netfront_info *info)
++{
++ /* Stop old i/f to prevent errors whilst we rebuild the state. */
++ spin_lock_bh(&info->rx_lock);
++ spin_lock_irq(&info->tx_lock);
++ netfront_carrier_off(info);
++ spin_unlock_irq(&info->tx_lock);
++ spin_unlock_bh(&info->rx_lock);
++
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info->netdev);
++ info->irq = 0;
++
++ end_access(info->tx_ring_ref, info->tx.sring);
++ end_access(info->rx_ring_ref, info->rx.sring);
++ info->tx_ring_ref = GRANT_INVALID_REF;
++ info->rx_ring_ref = GRANT_INVALID_REF;
++ info->tx.sring = NULL;
++ info->rx.sring = NULL;
++}
++
++
++static void end_access(int ref, void *page)
++{
++ if (ref != GRANT_INVALID_REF)
++ gnttab_end_foreign_access(ref, (unsigned long)page);
++}
++
++
++/* ** Driver registration ** */
++
++
++static const struct xenbus_device_id netfront_ids[] = {
++ { "vif" },
++ { "" }
++};
++MODULE_ALIAS("xen:vif");
++
++
++static struct xenbus_driver netfront_driver = {
++ .name = "vif",
++ .ids = netfront_ids,
++ .probe = netfront_probe,
++ .remove = __devexit_p(netfront_remove),
++ .suspend = netfront_suspend,
++ .suspend_cancel = netfront_suspend_cancel,
++ .resume = netfront_resume,
++ .otherend_changed = backend_changed,
++};
++
++
++static int __init netif_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++#ifdef CONFIG_XEN
++ if (MODPARM_rx_flip && MODPARM_rx_copy) {
++ WPRINTK("Cannot specify both rx_copy and rx_flip.\n");
++ return -EINVAL;
++ }
++
++ if (!MODPARM_rx_flip && !MODPARM_rx_copy)
++ MODPARM_rx_flip = 1; /* Default is to flip. */
++#endif
++
++ netif_init_accel();
++
++ IPRINTK("Initialising virtual ethernet driver.\n");
++
++#ifdef CONFIG_INET
++ (void)register_inetaddr_notifier(¬ifier_inetdev);
++#endif
++
++ return xenbus_register_frontend(&netfront_driver);
++}
++module_init(netif_init);
++
++
++static void __exit netif_exit(void)
++{
++#ifdef CONFIG_INET
++ unregister_inetaddr_notifier(¬ifier_inetdev);
++#endif
++
++ netif_exit_accel();
++
++ return xenbus_unregister_driver(&netfront_driver);
++}
++module_exit(netif_exit);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/netfront/netfront.h b/drivers/xen/netfront/netfront.h
+new file mode 100644
+index 0000000..4cfa983
+--- /dev/null
++++ b/drivers/xen/netfront/netfront.h
+@@ -0,0 +1,276 @@
++/******************************************************************************
++ * Virtual network driver for conversing with remote driver backends.
++ *
++ * Copyright (c) 2002-2005, K A Fraser
++ * Copyright (c) 2005, XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef NETFRONT_H
++#define NETFRONT_H
++
++#include <xen/interface/io/netif.h>
++#include <linux/netdevice.h>
++#include <linux/skbuff.h>
++#include <linux/list.h>
++
++#define NET_TX_RING_SIZE __RING_SIZE((struct netif_tx_sring *)0, PAGE_SIZE)
++#define NET_RX_RING_SIZE __RING_SIZE((struct netif_rx_sring *)0, PAGE_SIZE)
++
++#include <xen/xenbus.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++/*
++ * Function pointer table for hooks into a network acceleration
++ * plugin. These are called at appropriate points from the netfront
++ * driver
++ */
++struct netfront_accel_hooks {
++ /*
++ * new_device: Accelerator hook to ask the plugin to support a
++ * new network interface
++ */
++ int (*new_device)(struct net_device *net_dev, struct xenbus_device *dev);
++ /*
++ * remove: Opposite of new_device
++ */
++ int (*remove)(struct xenbus_device *dev);
++ /*
++ * The net_device is being polled, check the accelerated
++ * hardware for any pending packets
++ */
++ int (*netdev_poll)(struct net_device *dev, int *pbudget);
++ /*
++ * start_xmit: Used to give the accelerated plugin the option
++ * of sending a packet. Returns non-zero if has done so, or
++ * zero to decline and force the packet onto normal send
++ * path
++ */
++ int (*start_xmit)(struct sk_buff *skb, struct net_device *dev);
++ /*
++ * start/stop_napi_interrupts Used by netfront to indicate
++ * when napi interrupts should be enabled or disabled
++ */
++ int (*start_napi_irq)(struct net_device *dev);
++ void (*stop_napi_irq)(struct net_device *dev);
++ /*
++ * Called before re-enabling the TX queue to check the fast
++ * path has slots too
++ */
++ int (*check_ready)(struct net_device *dev);
++ /*
++ * Get the fastpath network statistics
++ */
++ int (*get_stats)(struct net_device *dev,
++ struct net_device_stats *stats);
++};
++
++
++/* Version of API/protocol for communication between netfront and
++ acceleration plugin supported */
++#define NETFRONT_ACCEL_VERSION 0x00010003
++
++/*
++ * Per-netfront device state for the accelerator. This is used to
++ * allow efficient per-netfront device access to the accelerator
++ * hooks
++ */
++struct netfront_accel_vif_state {
++ struct list_head link;
++
++ struct xenbus_device *dev;
++ struct netfront_info *np;
++ struct netfront_accel_hooks *hooks;
++
++ /* Watch on the accelerator configuration value */
++ struct xenbus_watch accel_watch;
++ /* Work item to process change in accelerator */
++ struct work_struct accel_work;
++ /* The string from xenbus last time accel_watch fired */
++ char *accel_frontend;
++};
++
++/*
++ * Per-accelerator state stored in netfront. These form a list that
++ * is used to track which devices are accelerated by which plugins,
++ * and what plugins are available/have been requested
++ */
++struct netfront_accelerator {
++ /* Used to make a list */
++ struct list_head link;
++ /* ID of the accelerator */
++ int id;
++ /*
++ * String describing the accelerator. Currently this is the
++ * name of the accelerator module. This is provided by the
++ * backend accelerator through xenstore
++ */
++ char *frontend;
++ /* The hooks into the accelerator plugin module */
++ struct netfront_accel_hooks *hooks;
++
++ /*
++ * List of per-netfront device state (struct
++ * netfront_accel_vif_state) for each netfront device that is
++ * using this accelerator
++ */
++ struct list_head vif_states;
++ spinlock_t vif_states_lock;
++};
++
++struct netfront_info {
++ struct list_head list;
++ struct net_device *netdev;
++
++ struct net_device_stats stats;
++
++ struct netif_tx_front_ring tx;
++ struct netif_rx_front_ring rx;
++
++ spinlock_t tx_lock;
++ spinlock_t rx_lock;
++
++ struct napi_struct napi;
++
++ unsigned int irq;
++ unsigned int copying_receiver;
++ unsigned int carrier;
++
++ /* Receive-ring batched refills. */
++#define RX_MIN_TARGET 8
++#define RX_DFL_MIN_TARGET 64
++#define RX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
++ unsigned rx_min_target, rx_max_target, rx_target;
++ struct sk_buff_head rx_batch;
++
++ struct timer_list rx_refill_timer;
++
++ /*
++ * {tx,rx}_skbs store outstanding skbuffs. The first entry in tx_skbs
++ * is an index into a chain of free entries.
++ */
++ struct sk_buff *tx_skbs[NET_TX_RING_SIZE+1];
++ struct sk_buff *rx_skbs[NET_RX_RING_SIZE];
++
++#define TX_MAX_TARGET min_t(int, NET_RX_RING_SIZE, 256)
++ grant_ref_t gref_tx_head;
++ grant_ref_t grant_tx_ref[NET_TX_RING_SIZE + 1];
++ grant_ref_t gref_rx_head;
++ grant_ref_t grant_rx_ref[NET_RX_RING_SIZE];
++
++ struct xenbus_device *xbdev;
++ int tx_ring_ref;
++ int rx_ring_ref;
++ u8 mac[ETH_ALEN];
++
++ unsigned long rx_pfn_array[NET_RX_RING_SIZE];
++ struct multicall_entry rx_mcl[NET_RX_RING_SIZE+1];
++ struct mmu_update rx_mmu[NET_RX_RING_SIZE];
++
++ /* Private pointer to state internal to accelerator module */
++ void *accel_priv;
++ /* The accelerator used by this netfront device */
++ struct netfront_accelerator *accelerator;
++ /* The accelerator state for this netfront device */
++ struct netfront_accel_vif_state accel_vif_state;
++};
++
++
++/* Exported Functions */
++
++/*
++ * Called by an accelerator plugin module when it has loaded.
++ *
++ * frontend: the string describing the accelerator, currently the module name
++ * hooks: the hooks for netfront to use to call into the accelerator
++ * version: the version of API between frontend and plugin requested
++ *
++ * return: 0 on success, <0 on error, >0 (with version supported) on
++ * version mismatch
++ */
++extern int netfront_accelerator_loaded(int version, const char *frontend,
++ struct netfront_accel_hooks *hooks);
++
++/*
++ * Called by an accelerator plugin module when it is about to unload.
++ *
++ * frontend: the string describing the accelerator. Must match the
++ * one passed to netfront_accelerator_loaded()
++ */
++extern void netfront_accelerator_stop(const char *frontend);
++
++/*
++ * Called by an accelerator before waking the net device's TX queue to
++ * ensure the slow path has available slots. Returns true if OK to
++ * wake, false if still busy
++ */
++extern int netfront_check_queue_ready(struct net_device *net_dev);
++
++
++/* Internal-to-netfront Functions */
++
++/*
++ * Call into accelerator and check to see if it has tx space before we
++ * wake the net device's TX queue. Returns true if OK to wake, false
++ * if still busy
++ */
++extern
++int netfront_check_accelerator_queue_ready(struct net_device *dev,
++ struct netfront_info *np);
++extern
++int netfront_accelerator_call_remove(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++int netfront_accelerator_suspend(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++int netfront_accelerator_suspend_cancel(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++void netfront_accelerator_resume(struct netfront_info *np,
++ struct xenbus_device *dev);
++extern
++void netfront_accelerator_call_stop_napi_irq(struct netfront_info *np,
++ struct net_device *dev);
++extern
++int netfront_accelerator_call_get_stats(struct netfront_info *np,
++ struct net_device *dev);
++extern
++void netfront_accelerator_add_watch(struct netfront_info *np);
++
++extern
++void netif_init_accel(void);
++extern
++void netif_exit_accel(void);
++
++extern
++void init_accelerator_vif(struct netfront_info *np,
++ struct xenbus_device *dev);
++#endif /* NETFRONT_H */
+diff --git a/drivers/xen/pciback/Makefile b/drivers/xen/pciback/Makefile
+new file mode 100644
+index 0000000..ee107aa
+--- /dev/null
++++ b/drivers/xen/pciback/Makefile
+@@ -0,0 +1,15 @@
++obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o
++
++pciback-y := pci_stub.o pciback_ops.o xenbus.o
++pciback-y += conf_space.o conf_space_header.o \
++ conf_space_capability.o \
++ conf_space_capability_vpd.o \
++ conf_space_capability_pm.o \
++ conf_space_quirks.o
++pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o
++pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o
++
++ccflags-$(CONFIG_XEN_PCIDEV_BE_DEBUG) += -DDEBUG
+diff --git a/drivers/xen/pciback/conf_space.c b/drivers/xen/pciback/conf_space.c
+new file mode 100644
+index 0000000..4c99534
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space.c
+@@ -0,0 +1,426 @@
++/*
++ * PCI Backend - Functions for creating a virtual configuration space for
++ * exported PCI Devices.
++ * It's dangerous to allow PCI Driver Domains to change their
++ * device's resources (memory, i/o ports, interrupts). We need to
++ * restrict changes to certain PCI Configuration registers:
++ * BARs, INTERRUPT_PIN, most registers in the header...
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
++
++#define DEFINE_PCI_CONFIG(op,size,type) \
++int pciback_##op##_config_##size \
++(struct pci_dev *dev, int offset, type value, void *data) \
++{ \
++ return pci_##op##_config_##size (dev, offset, value); \
++}
++
++DEFINE_PCI_CONFIG(read, byte, u8 *)
++DEFINE_PCI_CONFIG(read, word, u16 *)
++DEFINE_PCI_CONFIG(read, dword, u32 *)
++
++DEFINE_PCI_CONFIG(write, byte, u8)
++DEFINE_PCI_CONFIG(write, word, u16)
++DEFINE_PCI_CONFIG(write, dword, u32)
++
++static int conf_space_read(struct pci_dev *dev,
++ struct config_field_entry *entry, int offset,
++ u32 * value)
++{
++ int ret = 0;
++ struct config_field *field = entry->field;
++
++ *value = 0;
++
++ switch (field->size) {
++ case 1:
++ if (field->u.b.read)
++ ret = field->u.b.read(dev, offset, (u8 *) value,
++ entry->data);
++ break;
++ case 2:
++ if (field->u.w.read)
++ ret = field->u.w.read(dev, offset, (u16 *) value,
++ entry->data);
++ break;
++ case 4:
++ if (field->u.dw.read)
++ ret = field->u.dw.read(dev, offset, value, entry->data);
++ break;
++ }
++ return ret;
++}
++
++static int conf_space_write(struct pci_dev *dev,
++ struct config_field_entry *entry, int offset,
++ u32 value)
++{
++ int ret = 0;
++ struct config_field *field = entry->field;
++
++ switch (field->size) {
++ case 1:
++ if (field->u.b.write)
++ ret = field->u.b.write(dev, offset, (u8) value,
++ entry->data);
++ break;
++ case 2:
++ if (field->u.w.write)
++ ret = field->u.w.write(dev, offset, (u16) value,
++ entry->data);
++ break;
++ case 4:
++ if (field->u.dw.write)
++ ret = field->u.dw.write(dev, offset, value,
++ entry->data);
++ break;
++ }
++ return ret;
++}
++
++static inline u32 get_mask(int size)
++{
++ if (size == 1)
++ return 0xff;
++ else if (size == 2)
++ return 0xffff;
++ else
++ return 0xffffffff;
++}
++
++static inline int valid_request(int offset, int size)
++{
++ /* Validate request (no un-aligned requests) */
++ if ((size == 1 || size == 2 || size == 4) && (offset % size) == 0)
++ return 1;
++ return 0;
++}
++
++static inline u32 merge_value(u32 val, u32 new_val, u32 new_val_mask,
++ int offset)
++{
++ if (offset >= 0) {
++ new_val_mask <<= (offset * 8);
++ new_val <<= (offset * 8);
++ } else {
++ new_val_mask >>= (offset * -8);
++ new_val >>= (offset * -8);
++ }
++ val = (val & ~new_val_mask) | (new_val & new_val_mask);
++
++ return val;
++}
++
++static int pcibios_err_to_errno(int err)
++{
++ switch (err) {
++ case PCIBIOS_SUCCESSFUL:
++ return XEN_PCI_ERR_success;
++ case PCIBIOS_DEVICE_NOT_FOUND:
++ return XEN_PCI_ERR_dev_not_found;
++ case PCIBIOS_BAD_REGISTER_NUMBER:
++ return XEN_PCI_ERR_invalid_offset;
++ case PCIBIOS_FUNC_NOT_SUPPORTED:
++ return XEN_PCI_ERR_not_implemented;
++ case PCIBIOS_SET_FAILED:
++ return XEN_PCI_ERR_access_denied;
++ }
++ return err;
++}
++
++int pciback_config_read(struct pci_dev *dev, int offset, int size,
++ u32 * ret_val)
++{
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
++ int req_start, req_end, field_start, field_end;
++ /* if read fails for any reason, return 0 (as if device didn't respond) */
++ u32 value = 0, tmp_val;
++
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x\n",
++ pci_name(dev), size, offset);
++
++ if (!valid_request(offset, size)) {
++ err = XEN_PCI_ERR_invalid_offset;
++ goto out;
++ }
++
++ /* Get the real value first, then modify as appropriate */
++ switch (size) {
++ case 1:
++ err = pci_read_config_byte(dev, offset, (u8 *) & value);
++ break;
++ case 2:
++ err = pci_read_config_word(dev, offset, (u16 *) & value);
++ break;
++ case 4:
++ err = pci_read_config_dword(dev, offset, &value);
++ break;
++ }
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++
++ req_start = offset;
++ req_end = offset + size;
++ field_start = OFFSET(cfg_entry);
++ field_end = OFFSET(cfg_entry) + field->size;
++
++ if ((req_start >= field_start && req_start < field_end)
++ || (req_end > field_start && req_end <= field_end)) {
++ err = conf_space_read(dev, cfg_entry, field_start,
++ &tmp_val);
++ if (err)
++ goto out;
++
++ value = merge_value(value, tmp_val,
++ get_mask(field->size),
++ field_start - req_start);
++ }
++ }
++
++ out:
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: read %d bytes at 0x%x = %x\n",
++ pci_name(dev), size, offset, value);
++
++ *ret_val = value;
++ return pcibios_err_to_errno(err);
++}
++
++int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value)
++{
++ int err = 0, handled = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
++ u32 tmp_val;
++ int req_start, req_end, field_start, field_end;
++
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG
++ "pciback: %s: write request %d bytes at 0x%x = %x\n",
++ pci_name(dev), size, offset, value);
++
++ if (!valid_request(offset, size))
++ return XEN_PCI_ERR_invalid_offset;
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++
++ req_start = offset;
++ req_end = offset + size;
++ field_start = OFFSET(cfg_entry);
++ field_end = OFFSET(cfg_entry) + field->size;
++
++ if ((req_start >= field_start && req_start < field_end)
++ || (req_end > field_start && req_end <= field_end)) {
++ tmp_val = 0;
++
++ err = pciback_config_read(dev, field_start,
++ field->size, &tmp_val);
++ if (err)
++ break;
++
++ tmp_val = merge_value(tmp_val, value, get_mask(size),
++ req_start - field_start);
++
++ err = conf_space_write(dev, cfg_entry, field_start,
++ tmp_val);
++
++ /* handled is set true here, but not every byte
++ * may have been written! Properly detecting if
++ * every byte is handled is unnecessary as the
++ * flag is used to detect devices that need
++ * special helpers to work correctly.
++ */
++ handled = 1;
++ }
++ }
++
++ if (!handled && !err) {
++ /* By default, anything not specificially handled above is
++ * read-only. The permissive flag changes this behavior so
++ * that anything not specifically handled above is writable.
++ * This means that some fields may still be read-only because
++ * they have entries in the config_field list that intercept
++ * the write and do nothing. */
++ if (dev_data->permissive) {
++ switch (size) {
++ case 1:
++ err = pci_write_config_byte(dev, offset,
++ (u8) value);
++ break;
++ case 2:
++ err = pci_write_config_word(dev, offset,
++ (u16) value);
++ break;
++ case 4:
++ err = pci_write_config_dword(dev, offset,
++ (u32) value);
++ break;
++ }
++ } else if (!dev_data->warned_on_write) {
++ dev_data->warned_on_write = 1;
++ dev_warn(&dev->dev, "Driver tried to write to a "
++ "read-only configuration space field at offset "
++ "0x%x, size %d. This may be harmless, but if "
++ "you have problems with your device:\n"
++ "1) see permissive attribute in sysfs\n"
++ "2) report problems to the xen-devel "
++ "mailing list along with details of your "
++ "device obtained from lspci.\n", offset, size);
++ }
++ }
++
++ return pcibios_err_to_errno(err);
++}
++
++void pciback_config_free_dyn_fields(struct pci_dev *dev)
++{
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry, *t;
++ struct config_field *field;
++
++ dev_dbg(&dev->dev,
++ "free-ing dynamically allocated virtual configuration space fields\n");
++
++ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++
++ if (field->clean) {
++ field->clean(field);
++
++ if (cfg_entry->data)
++ kfree(cfg_entry->data);
++
++ list_del(&cfg_entry->list);
++ kfree(cfg_entry);
++ }
++
++ }
++}
++
++void pciback_config_reset_dev(struct pci_dev *dev)
++{
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ struct config_field *field;
++
++ dev_dbg(&dev->dev, "resetting virtual configuration space\n");
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++
++ if (field->reset)
++ field->reset(dev, OFFSET(cfg_entry), cfg_entry->data);
++ }
++}
++
++void pciback_config_free_dev(struct pci_dev *dev)
++{
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry, *t;
++ struct config_field *field;
++
++ dev_dbg(&dev->dev, "free-ing virtual configuration space fields\n");
++
++ list_for_each_entry_safe(cfg_entry, t, &dev_data->config_fields, list) {
++ list_del(&cfg_entry->list);
++
++ field = cfg_entry->field;
++
++ if (field->release)
++ field->release(dev, OFFSET(cfg_entry), cfg_entry->data);
++
++ kfree(cfg_entry);
++ }
++}
++
++int pciback_config_add_field_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int base_offset)
++{
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++ void *tmp;
++
++ cfg_entry = kmalloc(sizeof(*cfg_entry), GFP_KERNEL);
++ if (!cfg_entry) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ cfg_entry->data = NULL;
++ cfg_entry->field = field;
++ cfg_entry->base_offset = base_offset;
++
++ /* silently ignore duplicate fields */
++ err = pciback_field_is_dup(dev,OFFSET(cfg_entry));
++ if (err)
++ goto out;
++
++ if (field->init) {
++ tmp = field->init(dev, OFFSET(cfg_entry));
++
++ if (IS_ERR(tmp)) {
++ err = PTR_ERR(tmp);
++ goto out;
++ }
++
++ cfg_entry->data = tmp;
++ }
++
++ dev_dbg(&dev->dev, "added config field at offset 0x%02x\n",
++ OFFSET(cfg_entry));
++ list_add_tail(&cfg_entry->list, &dev_data->config_fields);
++
++ out:
++ if (err)
++ kfree(cfg_entry);
++
++ return err;
++}
++
++/* This sets up the device's virtual configuration space to keep track of
++ * certain registers (like the base address registers (BARs) so that we can
++ * keep the client from manipulating them directly.
++ */
++int pciback_config_init_dev(struct pci_dev *dev)
++{
++ int err = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++
++ dev_dbg(&dev->dev, "initializing virtual configuration space\n");
++
++ INIT_LIST_HEAD(&dev_data->config_fields);
++
++ err = pciback_config_header_add_fields(dev);
++ if (err)
++ goto out;
++
++ err = pciback_config_capability_add_fields(dev);
++ if (err)
++ goto out;
++
++ err = pciback_config_quirks_init(dev);
++
++ out:
++ return err;
++}
++
++int pciback_config_init(void)
++{
++ return pciback_config_capability_init();
++}
+diff --git a/drivers/xen/pciback/conf_space.h b/drivers/xen/pciback/conf_space.h
+new file mode 100644
+index 0000000..76fe5d8
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space.h
+@@ -0,0 +1,126 @@
++/*
++ * PCI Backend - Common data structures for overriding the configuration space
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#ifndef __XEN_PCIBACK_CONF_SPACE_H__
++#define __XEN_PCIBACK_CONF_SPACE_H__
++
++#include <linux/list.h>
++#include <linux/err.h>
++
++/* conf_field_init can return an errno in a ptr with ERR_PTR() */
++typedef void *(*conf_field_init) (struct pci_dev * dev, int offset);
++typedef void (*conf_field_reset) (struct pci_dev * dev, int offset, void *data);
++typedef void (*conf_field_free) (struct pci_dev * dev, int offset, void *data);
++
++typedef int (*conf_dword_write) (struct pci_dev * dev, int offset, u32 value,
++ void *data);
++typedef int (*conf_word_write) (struct pci_dev * dev, int offset, u16 value,
++ void *data);
++typedef int (*conf_byte_write) (struct pci_dev * dev, int offset, u8 value,
++ void *data);
++typedef int (*conf_dword_read) (struct pci_dev * dev, int offset, u32 * value,
++ void *data);
++typedef int (*conf_word_read) (struct pci_dev * dev, int offset, u16 * value,
++ void *data);
++typedef int (*conf_byte_read) (struct pci_dev * dev, int offset, u8 * value,
++ void *data);
++
++/* These are the fields within the configuration space which we
++ * are interested in intercepting reads/writes to and changing their
++ * values.
++ */
++struct config_field {
++ unsigned int offset;
++ unsigned int size;
++ unsigned int mask;
++ conf_field_init init;
++ conf_field_reset reset;
++ conf_field_free release;
++ void (*clean) (struct config_field * field);
++ union {
++ struct {
++ conf_dword_write write;
++ conf_dword_read read;
++ } dw;
++ struct {
++ conf_word_write write;
++ conf_word_read read;
++ } w;
++ struct {
++ conf_byte_write write;
++ conf_byte_read read;
++ } b;
++ } u;
++ struct list_head list;
++};
++
++struct config_field_entry {
++ struct list_head list;
++ struct config_field *field;
++ unsigned int base_offset;
++ void *data;
++};
++
++#define OFFSET(cfg_entry) ((cfg_entry)->base_offset+(cfg_entry)->field->offset)
++
++/* Add fields to a device - the add_fields macro expects to get a pointer to
++ * the first entry in an array (of which the ending is marked by size==0)
++ */
++int pciback_config_add_field_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int offset);
++
++static inline int pciback_config_add_field(struct pci_dev *dev,
++ struct config_field *field)
++{
++ return pciback_config_add_field_offset(dev, field, 0);
++}
++
++static inline int pciback_config_add_fields(struct pci_dev *dev,
++ struct config_field *field)
++{
++ int i, err = 0;
++ for (i = 0; field[i].size != 0; i++) {
++ err = pciback_config_add_field(dev, &field[i]);
++ if (err)
++ break;
++ }
++ return err;
++}
++
++static inline int pciback_config_add_fields_offset(struct pci_dev *dev,
++ struct config_field *field,
++ unsigned int offset)
++{
++ int i, err = 0;
++ for (i = 0; field[i].size != 0; i++) {
++ err = pciback_config_add_field_offset(dev, &field[i], offset);
++ if (err)
++ break;
++ }
++ return err;
++}
++
++/* Read/Write the real configuration space */
++int pciback_read_config_byte(struct pci_dev *dev, int offset, u8 * value,
++ void *data);
++int pciback_read_config_word(struct pci_dev *dev, int offset, u16 * value,
++ void *data);
++int pciback_read_config_dword(struct pci_dev *dev, int offset, u32 * value,
++ void *data);
++int pciback_write_config_byte(struct pci_dev *dev, int offset, u8 value,
++ void *data);
++int pciback_write_config_word(struct pci_dev *dev, int offset, u16 value,
++ void *data);
++int pciback_write_config_dword(struct pci_dev *dev, int offset, u32 value,
++ void *data);
++
++int pciback_config_capability_init(void);
++
++int pciback_config_header_add_fields(struct pci_dev *dev);
++int pciback_config_capability_add_fields(struct pci_dev *dev);
++
++#endif /* __XEN_PCIBACK_CONF_SPACE_H__ */
+diff --git a/drivers/xen/pciback/conf_space_capability.c b/drivers/xen/pciback/conf_space_capability.c
+new file mode 100644
+index 0000000..832a0c8
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_capability.c
+@@ -0,0 +1,71 @@
++/*
++ * PCI Backend - Handles the virtual fields found on the capability lists
++ * in the configuration space.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_capability.h"
++
++static LIST_HEAD(capabilities);
++
++static struct config_field caplist_header[] = {
++ {
++ .offset = PCI_CAP_LIST_ID,
++ .size = 2, /* encompass PCI_CAP_LIST_ID & PCI_CAP_LIST_NEXT */
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = NULL,
++ },
++ {
++ .size = 0,
++ },
++};
++
++static inline void register_capability(struct pciback_config_capability *cap)
++{
++ list_add_tail(&cap->cap_list, &capabilities);
++}
++
++int pciback_config_capability_add_fields(struct pci_dev *dev)
++{
++ int err = 0;
++ struct pciback_config_capability *cap;
++ int cap_offset;
++
++ list_for_each_entry(cap, &capabilities, cap_list) {
++ cap_offset = pci_find_capability(dev, cap->capability);
++ if (cap_offset) {
++ dev_dbg(&dev->dev, "Found capability 0x%x at 0x%x\n",
++ cap->capability, cap_offset);
++
++ err = pciback_config_add_fields_offset(dev,
++ caplist_header,
++ cap_offset);
++ if (err)
++ goto out;
++ err = pciback_config_add_fields_offset(dev,
++ cap->fields,
++ cap_offset);
++ if (err)
++ goto out;
++ }
++ }
++
++ out:
++ return err;
++}
++
++extern struct pciback_config_capability pciback_config_capability_vpd;
++extern struct pciback_config_capability pciback_config_capability_pm;
++
++int pciback_config_capability_init(void)
++{
++ register_capability(&pciback_config_capability_vpd);
++ register_capability(&pciback_config_capability_pm);
++
++ return 0;
++}
+diff --git a/drivers/xen/pciback/conf_space_capability.h b/drivers/xen/pciback/conf_space_capability.h
+new file mode 100644
+index 0000000..02bfcad
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_capability.h
+@@ -0,0 +1,23 @@
++/*
++ * PCI Backend - Data structures for special overlays for structures on
++ * the capability list.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#ifndef __PCIBACK_CONFIG_CAPABILITY_H__
++#define __PCIBACK_CONFIG_CAPABILITY_H__
++
++#include <linux/pci.h>
++#include <linux/list.h>
++
++struct pciback_config_capability {
++ struct list_head cap_list;
++
++ int capability;
++
++ /* If the device has the capability found above, add these fields */
++ struct config_field *fields;
++};
++
++#endif
+diff --git a/drivers/xen/pciback/conf_space_capability_msi.c b/drivers/xen/pciback/conf_space_capability_msi.c
+new file mode 100644
+index 0000000..46f5faf
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_capability_msi.c
+@@ -0,0 +1,69 @@
++/*
++ * PCI Backend -- Configuration overlay for MSI capability
++ */
++#include <linux/pci.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
++#include <xen/interface/io/pciif.h>
++#include "pciback.h"
++
++int pciback_enable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++ int otherend = pdev->xdev->otherend_id;
++ int status;
++
++ status = pci_enable_msi(dev);
++
++ if (status) {
++ printk("error enable msi for guest %x status %x\n", otherend, status);
++ op->value = 0;
++ return XEN_PCI_ERR_op_failed;
++ }
++
++ op->value = dev->irq;
++ return 0;
++}
++
++int pciback_disable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++ pci_disable_msi(dev);
++
++ op->value = dev->irq;
++ return 0;
++}
++
++int pciback_enable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++ int result;
++
++ if (op->value > SH_INFO_MAX_VEC)
++ return -EINVAL;
++ else {
++ struct msix_entry entries[op->value];
++ int i;
++
++ for (i = 0; i < op->value; i++) {
++ entries[i].entry = op->msix_entries[i].entry;
++ entries[i].vector = op->msix_entries[i].vector;
++ }
++
++ result = pci_enable_msix(dev, entries, op->value);
++ op->value = result;
++ }
++
++ return result;
++}
++
++int pciback_disable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op)
++{
++
++ pci_disable_msix(dev);
++
++ op->value = dev->irq;
++ return 0;
++}
++
+diff --git a/drivers/xen/pciback/conf_space_capability_pm.c b/drivers/xen/pciback/conf_space_capability_pm.c
+new file mode 100644
+index 0000000..1cdc6e9
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_capability_pm.c
+@@ -0,0 +1,128 @@
++/*
++ * PCI Backend - Configuration space overlay for power management
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/pci.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
++
++static int pm_caps_read(struct pci_dev *dev, int offset, u16 *value,
++ void *data)
++{
++ int err;
++ u16 real_value;
++
++ err = pci_read_config_word(dev, offset, &real_value);
++ if (err)
++ goto out;
++
++ *value = real_value & ~PCI_PM_CAP_PME_MASK;
++
++ out:
++ return err;
++}
++
++/* PM_OK_BITS specifies the bits that the driver domain is allowed to change.
++ * Can't allow driver domain to enable PMEs - they're shared */
++#define PM_OK_BITS (PCI_PM_CTRL_PME_STATUS|PCI_PM_CTRL_DATA_SEL_MASK)
++
++static int pm_ctrl_write(struct pci_dev *dev, int offset, u16 new_value,
++ void *data)
++{
++ int err;
++ u16 old_value;
++ pci_power_t new_state, old_state;
++
++ err = pci_read_config_word(dev, offset, &old_value);
++ if (err)
++ goto out;
++
++ old_state = (pci_power_t)(old_value & PCI_PM_CTRL_STATE_MASK);
++ new_state = (pci_power_t)(new_value & PCI_PM_CTRL_STATE_MASK);
++
++ new_value &= PM_OK_BITS;
++ if ((old_value & PM_OK_BITS) != new_value) {
++ new_value = (old_value & ~PM_OK_BITS) | new_value;
++ err = pci_write_config_word(dev, offset, new_value);
++ if (err)
++ goto out;
++ }
++
++ /* Let pci core handle the power management change */
++ dev_dbg(&dev->dev, "set power state to %x\n", new_state);
++ err = pci_set_power_state(dev, new_state);
++ if (err) {
++ err = PCIBIOS_SET_FAILED;
++ goto out;
++ }
++
++ /*
++ * Device may lose PCI config info on D3->D0 transition. This
++ * is a problem for some guests which will not reset BARs. Even
++ * those that have a go will be foiled by our BAR-write handler
++ * which will discard the write! Since Linux won't re-init
++ * the config space automatically in all cases, we do it here.
++ * Future: Should we re-initialise all first 64 bytes of config space?
++ */
++ if (new_state == PCI_D0 &&
++ (old_state == PCI_D3hot || old_state == PCI_D3cold) &&
++ !(old_value & PCI_PM_CTRL_NO_SOFT_RESET))
++ pci_restore_bars(dev);
++
++ out:
++ return err;
++}
++
++/* Ensure PMEs are disabled */
++static void *pm_ctrl_init(struct pci_dev *dev, int offset)
++{
++ int err;
++ u16 value;
++
++ err = pci_read_config_word(dev, offset, &value);
++ if (err)
++ goto out;
++
++ if (value & PCI_PM_CTRL_PME_ENABLE) {
++ value &= ~PCI_PM_CTRL_PME_ENABLE;
++ err = pci_write_config_word(dev, offset, value);
++ }
++
++ out:
++ return ERR_PTR(err);
++}
++
++static struct config_field caplist_pm[] = {
++ {
++ .offset = PCI_PM_PMC,
++ .size = 2,
++ .u.w.read = pm_caps_read,
++ },
++ {
++ .offset = PCI_PM_CTRL,
++ .size = 2,
++ .init = pm_ctrl_init,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = pm_ctrl_write,
++ },
++ {
++ .offset = PCI_PM_PPB_EXTENSIONS,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .offset = PCI_PM_DATA_REGISTER,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .size = 0,
++ },
++};
++
++struct pciback_config_capability pciback_config_capability_pm = {
++ .capability = PCI_CAP_ID_PM,
++ .fields = caplist_pm,
++};
+diff --git a/drivers/xen/pciback/conf_space_capability_vpd.c b/drivers/xen/pciback/conf_space_capability_vpd.c
+new file mode 100644
+index 0000000..cc6b853
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_capability_vpd.c
+@@ -0,0 +1,42 @@
++/*
++ * PCI Backend - Configuration space overlay for Vital Product Data
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/pci.h>
++#include "conf_space.h"
++#include "conf_space_capability.h"
++
++static int vpd_address_write(struct pci_dev *dev, int offset, u16 value,
++ void *data)
++{
++ /* Disallow writes to the vital product data */
++ if (value & PCI_VPD_ADDR_F)
++ return PCIBIOS_SET_FAILED;
++ else
++ return pci_write_config_word(dev, offset, value);
++}
++
++static struct config_field caplist_vpd[] = {
++ {
++ .offset = PCI_VPD_ADDR,
++ .size = 2,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = vpd_address_write,
++ },
++ {
++ .offset = PCI_VPD_DATA,
++ .size = 4,
++ .u.dw.read = pciback_read_config_dword,
++ .u.dw.write = NULL,
++ },
++ {
++ .size = 0,
++ },
++};
++
++struct pciback_config_capability pciback_config_capability_vpd = {
++ .capability = PCI_CAP_ID_VPD,
++ .fields = caplist_vpd,
++};
+diff --git a/drivers/xen/pciback/conf_space_header.c b/drivers/xen/pciback/conf_space_header.c
+new file mode 100644
+index 0000000..88e3e4a
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_header.c
+@@ -0,0 +1,323 @@
++/*
++ * PCI Backend - Handles the virtual fields in the configuration space headers.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++
++struct pci_bar_info {
++ u32 val;
++ u32 len_val;
++ int which;
++};
++
++#define is_enable_cmd(value) ((value)&(PCI_COMMAND_MEMORY|PCI_COMMAND_IO))
++#define is_master_cmd(value) ((value)&PCI_COMMAND_MASTER)
++
++static int command_write(struct pci_dev *dev, int offset, u16 value, void *data)
++{
++ int err;
++
++ if (!atomic_read(&dev->enable_cnt) && is_enable_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: enable\n",
++ pci_name(dev));
++ err = pci_enable_device(dev);
++ if (err)
++ return err;
++ } else if (atomic_read(&dev->enable_cnt) && !is_enable_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: disable\n",
++ pci_name(dev));
++ pci_disable_device(dev);
++ }
++
++ if (!dev->is_busmaster && is_master_cmd(value)) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG "pciback: %s: set bus master\n",
++ pci_name(dev));
++ pci_set_master(dev);
++ }
++
++ if (value & PCI_COMMAND_INVALIDATE) {
++ if (unlikely(verbose_request))
++ printk(KERN_DEBUG
++ "pciback: %s: enable memory-write-invalidate\n",
++ pci_name(dev));
++ err = pci_set_mwi(dev);
++ if (err) {
++ printk(KERN_WARNING
++ "pciback: %s: cannot enable memory-write-invalidate (%d)\n",
++ pci_name(dev), err);
++ value &= ~PCI_COMMAND_INVALIDATE;
++ }
++ }
++
++ return pci_write_config_word(dev, offset, value);
++}
++
++static int rom_write(struct pci_dev *dev, int offset, u32 value, void *data)
++{
++ struct pci_bar_info *bar = data;
++
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
++
++ /* A write to obtain the length must happen as a 32-bit write.
++ * This does not (yet) support writing individual bytes
++ */
++ if (value == ~PCI_ROM_ADDRESS_ENABLE)
++ bar->which = 1;
++ else {
++ u32 tmpval;
++ pci_read_config_dword(dev, offset, &tmpval);
++ if (tmpval != bar->val && value == bar->val) {
++ /* Allow restoration of bar value. */
++ pci_write_config_dword(dev, offset, bar->val);
++ }
++ bar->which = 0;
++ }
++
++ /* Do we need to support enabling/disabling the rom address here? */
++
++ return 0;
++}
++
++/* For the BARs, only allow writes which write ~0 or
++ * the correct resource information
++ * (Needed for when the driver probes the resource usage)
++ */
++static int bar_write(struct pci_dev *dev, int offset, u32 value, void *data)
++{
++ struct pci_bar_info *bar = data;
++
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
++
++ /* A write to obtain the length must happen as a 32-bit write.
++ * This does not (yet) support writing individual bytes
++ */
++ if (value == ~0)
++ bar->which = 1;
++ else {
++ u32 tmpval;
++ pci_read_config_dword(dev, offset, &tmpval);
++ if (tmpval != bar->val && value == bar->val) {
++ /* Allow restoration of bar value. */
++ pci_write_config_dword(dev, offset, bar->val);
++ }
++ bar->which = 0;
++ }
++
++ return 0;
++}
++
++static int bar_read(struct pci_dev *dev, int offset, u32 * value, void *data)
++{
++ struct pci_bar_info *bar = data;
++
++ if (unlikely(!bar)) {
++ printk(KERN_WARNING "pciback: driver data not found for %s\n",
++ pci_name(dev));
++ return XEN_PCI_ERR_op_failed;
++ }
++
++ *value = bar->which ? bar->len_val : bar->val;
++
++ return 0;
++}
++
++static inline void read_dev_bar(struct pci_dev *dev,
++ struct pci_bar_info *bar_info, int offset,
++ u32 len_mask)
++{
++ pci_read_config_dword(dev, offset, &bar_info->val);
++ pci_write_config_dword(dev, offset, len_mask);
++ pci_read_config_dword(dev, offset, &bar_info->len_val);
++ pci_write_config_dword(dev, offset, bar_info->val);
++}
++
++static void *bar_init(struct pci_dev *dev, int offset)
++{
++ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
++
++ if (!bar)
++ return ERR_PTR(-ENOMEM);
++
++ read_dev_bar(dev, bar, offset, ~0);
++ bar->which = 0;
++
++ return bar;
++}
++
++static void *rom_init(struct pci_dev *dev, int offset)
++{
++ struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL);
++
++ if (!bar)
++ return ERR_PTR(-ENOMEM);
++
++ read_dev_bar(dev, bar, offset, ~PCI_ROM_ADDRESS_ENABLE);
++ bar->which = 0;
++
++ return bar;
++}
++
++static void bar_reset(struct pci_dev *dev, int offset, void *data)
++{
++ struct pci_bar_info *bar = data;
++
++ bar->which = 0;
++}
++
++static void bar_release(struct pci_dev *dev, int offset, void *data)
++{
++ kfree(data);
++}
++
++static int interrupt_read(struct pci_dev *dev, int offset, u8 * value,
++ void *data)
++{
++ *value = (u8) dev->irq;
++
++ return 0;
++}
++
++static int bist_write(struct pci_dev *dev, int offset, u8 value, void *data)
++{
++ u8 cur_value;
++ int err;
++
++ err = pci_read_config_byte(dev, offset, &cur_value);
++ if (err)
++ goto out;
++
++ if ((cur_value & ~PCI_BIST_START) == (value & ~PCI_BIST_START)
++ || value == PCI_BIST_START)
++ err = pci_write_config_byte(dev, offset, value);
++
++ out:
++ return err;
++}
++
++static struct config_field header_common[] = {
++ {
++ .offset = PCI_COMMAND,
++ .size = 2,
++ .u.w.read = pciback_read_config_word,
++ .u.w.write = command_write,
++ },
++ {
++ .offset = PCI_INTERRUPT_LINE,
++ .size = 1,
++ .u.b.read = interrupt_read,
++ },
++ {
++ .offset = PCI_INTERRUPT_PIN,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ /* Any side effects of letting driver domain control cache line? */
++ .offset = PCI_CACHE_LINE_SIZE,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ .u.b.write = pciback_write_config_byte,
++ },
++ {
++ .offset = PCI_LATENCY_TIMER,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ },
++ {
++ .offset = PCI_BIST,
++ .size = 1,
++ .u.b.read = pciback_read_config_byte,
++ .u.b.write = bist_write,
++ },
++ {
++ .size = 0,
++ },
++};
++
++#define CFG_FIELD_BAR(reg_offset) \
++ { \
++ .offset = reg_offset, \
++ .size = 4, \
++ .init = bar_init, \
++ .reset = bar_reset, \
++ .release = bar_release, \
++ .u.dw.read = bar_read, \
++ .u.dw.write = bar_write, \
++ }
++
++#define CFG_FIELD_ROM(reg_offset) \
++ { \
++ .offset = reg_offset, \
++ .size = 4, \
++ .init = rom_init, \
++ .reset = bar_reset, \
++ .release = bar_release, \
++ .u.dw.read = bar_read, \
++ .u.dw.write = rom_write, \
++ }
++
++static struct config_field header_0[] = {
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_2),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_3),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_4),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_5),
++ CFG_FIELD_ROM(PCI_ROM_ADDRESS),
++ {
++ .size = 0,
++ },
++};
++
++static struct config_field header_1[] = {
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_0),
++ CFG_FIELD_BAR(PCI_BASE_ADDRESS_1),
++ CFG_FIELD_ROM(PCI_ROM_ADDRESS1),
++ {
++ .size = 0,
++ },
++};
++
++int pciback_config_header_add_fields(struct pci_dev *dev)
++{
++ int err;
++
++ err = pciback_config_add_fields(dev, header_common);
++ if (err)
++ goto out;
++
++ switch (dev->hdr_type) {
++ case PCI_HEADER_TYPE_NORMAL:
++ err = pciback_config_add_fields(dev, header_0);
++ break;
++
++ case PCI_HEADER_TYPE_BRIDGE:
++ err = pciback_config_add_fields(dev, header_1);
++ break;
++
++ default:
++ err = -EINVAL;
++ printk(KERN_ERR "pciback: %s: Unsupported header type %d!\n",
++ pci_name(dev), dev->hdr_type);
++ break;
++ }
++
++ out:
++ return err;
++}
+diff --git a/drivers/xen/pciback/conf_space_quirks.c b/drivers/xen/pciback/conf_space_quirks.c
+new file mode 100644
+index 0000000..525ec95
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_quirks.c
+@@ -0,0 +1,126 @@
++/*
++ * PCI Backend - Handle special overlays for broken devices.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Author: Chris Bookholt <hap10 at epoch.ncsc.mil>
++ */
++
++#include <linux/kernel.h>
++#include <linux/pci.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
++
++LIST_HEAD(pciback_quirks);
++
++struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev)
++{
++ struct pciback_config_quirk *tmp_quirk;
++
++ list_for_each_entry(tmp_quirk, &pciback_quirks, quirks_list)
++ if (pci_match_id(&tmp_quirk->devid, dev))
++ goto out;
++ tmp_quirk = NULL;
++ printk(KERN_DEBUG
++ "quirk didn't match any device pciback knows about\n");
++ out:
++ return tmp_quirk;
++}
++
++static inline void register_quirk(struct pciback_config_quirk *quirk)
++{
++ list_add_tail(&quirk->quirks_list, &pciback_quirks);
++}
++
++int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg)
++{
++ int ret = 0;
++ struct pciback_dev_data *dev_data = pci_get_drvdata(dev);
++ struct config_field_entry *cfg_entry;
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ if ( OFFSET(cfg_entry) == reg) {
++ ret = 1;
++ break;
++ }
++ }
++ return ret;
++}
++
++int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
++ *field)
++{
++ int err = 0;
++
++ switch (field->size) {
++ case 1:
++ field->u.b.read = pciback_read_config_byte;
++ field->u.b.write = pciback_write_config_byte;
++ break;
++ case 2:
++ field->u.w.read = pciback_read_config_word;
++ field->u.w.write = pciback_write_config_word;
++ break;
++ case 4:
++ field->u.dw.read = pciback_read_config_dword;
++ field->u.dw.write = pciback_write_config_dword;
++ break;
++ default:
++ err = -EINVAL;
++ goto out;
++ }
++
++ pciback_config_add_field(dev, field);
++
++ out:
++ return err;
++}
++
++int pciback_config_quirks_init(struct pci_dev *dev)
++{
++ struct pciback_config_quirk *quirk;
++ int ret = 0;
++
++ quirk = kzalloc(sizeof(*quirk), GFP_ATOMIC);
++ if (!quirk) {
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ quirk->devid.vendor = dev->vendor;
++ quirk->devid.device = dev->device;
++ quirk->devid.subvendor = dev->subsystem_vendor;
++ quirk->devid.subdevice = dev->subsystem_device;
++ quirk->devid.class = 0;
++ quirk->devid.class_mask = 0;
++ quirk->devid.driver_data = 0UL;
++
++ quirk->pdev = dev;
++
++ register_quirk(quirk);
++ out:
++ return ret;
++}
++
++void pciback_config_field_free(struct config_field *field)
++{
++ kfree(field);
++}
++
++int pciback_config_quirk_release(struct pci_dev *dev)
++{
++ struct pciback_config_quirk *quirk;
++ int ret = 0;
++
++ quirk = pciback_find_quirk(dev);
++ if (!quirk) {
++ ret = -ENXIO;
++ goto out;
++ }
++
++ list_del(&quirk->quirks_list);
++ kfree(quirk);
++
++ out:
++ return ret;
++}
+diff --git a/drivers/xen/pciback/conf_space_quirks.h b/drivers/xen/pciback/conf_space_quirks.h
+new file mode 100644
+index 0000000..acd0e1a
+--- /dev/null
++++ b/drivers/xen/pciback/conf_space_quirks.h
+@@ -0,0 +1,35 @@
++/*
++ * PCI Backend - Data structures for special overlays for broken devices.
++ *
++ * Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Chris Bookholt <hap10 at epoch.ncsc.mil>
++ */
++
++#ifndef __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
++#define __XEN_PCIBACK_CONF_SPACE_QUIRKS_H__
++
++#include <linux/pci.h>
++#include <linux/list.h>
++
++struct pciback_config_quirk {
++ struct list_head quirks_list;
++ struct pci_device_id devid;
++ struct pci_dev *pdev;
++};
++
++struct pciback_config_quirk *pciback_find_quirk(struct pci_dev *dev);
++
++int pciback_config_quirks_add_field(struct pci_dev *dev, struct config_field
++ *field);
++
++int pciback_config_quirks_remove_field(struct pci_dev *dev, int reg);
++
++int pciback_config_quirks_init(struct pci_dev *dev);
++
++void pciback_config_field_free(struct config_field *field);
++
++int pciback_config_quirk_release(struct pci_dev *dev);
++
++int pciback_field_is_dup(struct pci_dev *dev, unsigned int reg);
++
++#endif
+diff --git a/drivers/xen/pciback/controller.c b/drivers/xen/pciback/controller.c
+new file mode 100644
+index 0000000..20e45a4
+--- /dev/null
++++ b/drivers/xen/pciback/controller.c
+@@ -0,0 +1,408 @@
++/*
++ * Copyright (C) 2007 Hewlett-Packard Development Company, L.P.
++ * Alex Williamson <alex.williamson at hp.com>
++ *
++ * PCI "Controller" Backend - virtualize PCI bus topology based on PCI
++ * controllers. Devices under the same PCI controller are exposed on the
++ * same virtual domain:bus. Within a bus, device slots are virtualized
++ * to compact the bus.
++ *
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
++ */
++
++#include <linux/acpi.h>
++#include <linux/list.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
++
++#define PCI_MAX_BUSSES 255
++#define PCI_MAX_SLOTS 32
++
++struct controller_dev_entry {
++ struct list_head list;
++ struct pci_dev *dev;
++ unsigned int devfn;
++};
++
++struct controller_list_entry {
++ struct list_head list;
++ struct pci_controller *controller;
++ unsigned int domain;
++ unsigned int bus;
++ unsigned int next_devfn;
++ struct list_head dev_list;
++};
++
++struct controller_dev_data {
++ struct list_head list;
++ unsigned int next_domain;
++ unsigned int next_bus;
++ spinlock_t lock;
++};
++
++struct walk_info {
++ struct pciback_device *pdev;
++ int resource_count;
++ int root_num;
++};
++
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_dev_entry *dev_entry;
++ struct controller_list_entry *cntrl_entry;
++ struct pci_dev *dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->domain != domain ||
++ cntrl_entry->bus != bus)
++ continue;
++
++ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
++ if (devfn == dev_entry->devfn) {
++ dev = dev_entry->dev;
++ goto found;
++ }
++ }
++ }
++found:
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ return dev;
++}
++
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_dev_entry *dev_entry;
++ struct controller_list_entry *cntrl_entry;
++ struct pci_controller *dev_controller = PCI_CONTROLLER(dev);
++ unsigned long flags;
++ int ret = 0, found = 0;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ /* Look to see if we already have a domain:bus for this controller */
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->controller == dev_controller) {
++ found = 1;
++ break;
++ }
++ }
++
++ if (!found) {
++ cntrl_entry = kmalloc(sizeof(*cntrl_entry), GFP_ATOMIC);
++ if (!cntrl_entry) {
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ cntrl_entry->controller = dev_controller;
++ cntrl_entry->next_devfn = PCI_DEVFN(0, 0);
++
++ cntrl_entry->domain = dev_data->next_domain;
++ cntrl_entry->bus = dev_data->next_bus++;
++ if (dev_data->next_bus > PCI_MAX_BUSSES) {
++ dev_data->next_domain++;
++ dev_data->next_bus = 0;
++ }
++
++ INIT_LIST_HEAD(&cntrl_entry->dev_list);
++
++ list_add_tail(&cntrl_entry->list, &dev_data->list);
++ }
++
++ if (PCI_SLOT(cntrl_entry->next_devfn) > PCI_MAX_SLOTS) {
++ /*
++ * While it seems unlikely, this can actually happen if
++ * a controller has P2P bridges under it.
++ */
++ xenbus_dev_fatal(pdev->xdev, -ENOSPC, "Virtual bus %04x:%02x "
++ "is full, no room to export %04x:%02x:%02x.%x",
++ cntrl_entry->domain, cntrl_entry->bus,
++ pci_domain_nr(dev->bus), dev->bus->number,
++ PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn));
++ ret = -ENOSPC;
++ goto out;
++ }
++
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_ATOMIC);
++ if (!dev_entry) {
++ if (list_empty(&cntrl_entry->dev_list)) {
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
++ ret = -ENOMEM;
++ goto out;
++ }
++
++ dev_entry->dev = dev;
++ dev_entry->devfn = cntrl_entry->next_devfn;
++
++ list_add_tail(&dev_entry->list, &cntrl_entry->dev_list);
++
++ cntrl_entry->next_devfn += PCI_DEVFN(1, 0);
++
++out:
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ /* TODO: Publish virtual domain:bus:slot.func here. */
++
++ return ret;
++}
++
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry;
++ struct controller_dev_entry *dev_entry = NULL;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ if (cntrl_entry->controller != PCI_CONTROLLER(dev))
++ continue;
++
++ list_for_each_entry(dev_entry, &cntrl_entry->dev_list, list) {
++ if (dev_entry->dev == dev) {
++ found_dev = dev_entry->dev;
++ break;
++ }
++ }
++ }
++
++ if (!found_dev) {
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++ return;
++ }
++
++ list_del(&dev_entry->list);
++ kfree(dev_entry);
++
++ if (list_empty(&cntrl_entry->dev_list)) {
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
++
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++ pcistub_put_pci_dev(found_dev);
++}
++
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ struct controller_dev_data *dev_data;
++
++ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
++ if (!dev_data)
++ return -ENOMEM;
++
++ spin_lock_init(&dev_data->lock);
++
++ INIT_LIST_HEAD(&dev_data->list);
++
++ /* Starting domain:bus numbers */
++ dev_data->next_domain = 0;
++ dev_data->next_bus = 0;
++
++ pdev->pci_dev_data = dev_data;
++
++ return 0;
++}
++
++static acpi_status write_xenbus_resource(struct acpi_resource *res, void *data)
++{
++ struct walk_info *info = data;
++ struct acpi_resource_address64 addr;
++ acpi_status status;
++ int i, len, err;
++ char str[32], tmp[3];
++ unsigned char *ptr, *buf;
++
++ status = acpi_resource_to_address64(res, &addr);
++
++ /* Do we care about this range? Let's check. */
++ if (!ACPI_SUCCESS(status) ||
++ !(addr.resource_type == ACPI_MEMORY_RANGE ||
++ addr.resource_type == ACPI_IO_RANGE) ||
++ !addr.address_length || addr.producer_consumer != ACPI_PRODUCER)
++ return AE_OK;
++
++ /*
++ * Furthermore, we really only care to tell the guest about
++ * address ranges that require address translation of some sort.
++ */
++ if (!(addr.resource_type == ACPI_MEMORY_RANGE &&
++ addr.info.mem.translation) &&
++ !(addr.resource_type == ACPI_IO_RANGE &&
++ addr.info.io.translation))
++ return AE_OK;
++
++ /* Store the resource in xenbus for the guest */
++ len = snprintf(str, sizeof(str), "root-%d-resource-%d",
++ info->root_num, info->resource_count);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return AE_OK;
++
++ buf = kzalloc((sizeof(*res) * 2) + 1, GFP_KERNEL);
++ if (!buf)
++ return AE_OK;
++
++ /* Clean out resource_source */
++ res->data.address64.resource_source.index = 0xFF;
++ res->data.address64.resource_source.string_length = 0;
++ res->data.address64.resource_source.string_ptr = NULL;
++
++ ptr = (unsigned char *)res;
++
++ /* Turn the acpi_resource into an ASCII byte stream */
++ for (i = 0; i < sizeof(*res); i++) {
++ snprintf(tmp, sizeof(tmp), "%02x", ptr[i]);
++ strncat(buf, tmp, 2);
++ }
++
++ err = xenbus_printf(XBT_NIL, info->pdev->xdev->nodename,
++ str, "%s", buf);
++
++ if (!err)
++ info->resource_count++;
++
++ kfree(buf);
++
++ return AE_OK;
++}
++
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_root_cb)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry;
++ int i, root_num, len, err = 0;
++ unsigned int domain, bus;
++ char str[64];
++ struct walk_info info;
++
++ spin_lock(&dev_data->lock);
++
++ list_for_each_entry(cntrl_entry, &dev_data->list, list) {
++ /* First publish all the domain:bus info */
++ err = publish_root_cb(pdev, cntrl_entry->domain,
++ cntrl_entry->bus);
++ if (err)
++ goto out;
++
++ /*
++ * Now figure out which root-%d this belongs to
++ * so we can associate resources with it.
++ */
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", &root_num);
++
++ if (err != 1)
++ goto out;
++
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ str, "%x:%x", &domain, &bus);
++ if (err != 2)
++ goto out;
++
++ /* Is this the one we just published? */
++ if (domain == cntrl_entry->domain &&
++ bus == cntrl_entry->bus)
++ break;
++ }
++
++ if (i == root_num)
++ goto out;
++
++ info.pdev = pdev;
++ info.resource_count = 0;
++ info.root_num = i;
++
++ /* Let ACPI do the heavy lifting on decoding resources */
++ acpi_walk_resources(cntrl_entry->controller->acpi_handle,
++ METHOD_NAME__CRS, write_xenbus_resource,
++ &info);
++
++ /* No resouces. OK. On to the next one */
++ if (!info.resource_count)
++ continue;
++
++ /* Store the number of resources we wrote for this root-%d */
++ len = snprintf(str, sizeof(str), "root-%d-resources", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%d", info.resource_count);
++ if (err)
++ goto out;
++ }
++
++ /* Finally, write some magic to synchronize with the guest. */
++ len = snprintf(str, sizeof(str), "root-resource-magic");
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%lx", (sizeof(struct acpi_resource) * 2) + 1);
++
++out:
++ spin_unlock(&dev_data->lock);
++
++ return err;
++}
++
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ struct controller_dev_data *dev_data = pdev->pci_dev_data;
++ struct controller_list_entry *cntrl_entry, *c;
++ struct controller_dev_entry *dev_entry, *d;
++
++ list_for_each_entry_safe(cntrl_entry, c, &dev_data->list, list) {
++ list_for_each_entry_safe(dev_entry, d,
++ &cntrl_entry->dev_list, list) {
++ list_del(&dev_entry->list);
++ pcistub_put_pci_dev(dev_entry->dev);
++ kfree(dev_entry);
++ }
++ list_del(&cntrl_entry->list);
++ kfree(cntrl_entry);
++ }
++
++ kfree(dev_data);
++ pdev->pci_dev_data = NULL;
++}
+diff --git a/drivers/xen/pciback/passthrough.c b/drivers/xen/pciback/passthrough.c
+new file mode 100644
+index 0000000..73b5390
+--- /dev/null
++++ b/drivers/xen/pciback/passthrough.c
+@@ -0,0 +1,166 @@
++/*
++ * PCI Backend - Provides restricted access to the real PCI bus topology
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/list.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
++
++struct passthrough_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct list_head dev_list;
++ spinlock_t lock;
++};
++
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry;
++ struct pci_dev *dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
++ if (domain == (unsigned int)pci_domain_nr(dev_entry->dev->bus)
++ && bus == (unsigned int)dev_entry->dev->bus->number
++ && devfn == dev_entry->dev->devfn) {
++ dev = dev_entry->dev;
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ return dev;
++}
++
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry;
++ unsigned long flags;
++ unsigned int domain, bus, devfn;
++ int err;
++
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
++ if (!dev_entry)
++ return -ENOMEM;
++ dev_entry->dev = dev;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++ list_add_tail(&dev_entry->list, &dev_data->dev_list);
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ /* Publish this device. */
++ domain = (unsigned int)pci_domain_nr(dev->bus);
++ bus = (unsigned int)dev->bus->number;
++ devfn = dev->devfn;
++ err = publish_cb(pdev, domain, bus, devfn, devid);
++
++ return err;
++}
++
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *t;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&dev_data->lock, flags);
++
++ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
++ if (dev_entry->dev == dev) {
++ list_del(&dev_entry->list);
++ found_dev = dev_entry->dev;
++ kfree(dev_entry);
++ }
++ }
++
++ spin_unlock_irqrestore(&dev_data->lock, flags);
++
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
++}
++
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ struct passthrough_dev_data *dev_data;
++
++ dev_data = kmalloc(sizeof(*dev_data), GFP_KERNEL);
++ if (!dev_data)
++ return -ENOMEM;
++
++ spin_lock_init(&dev_data->lock);
++
++ INIT_LIST_HEAD(&dev_data->dev_list);
++
++ pdev->pci_dev_data = dev_data;
++
++ return 0;
++}
++
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_root_cb)
++{
++ int err = 0;
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *e;
++ struct pci_dev *dev;
++ int found;
++ unsigned int domain, bus;
++
++ spin_lock(&dev_data->lock);
++
++ list_for_each_entry(dev_entry, &dev_data->dev_list, list) {
++ /* Only publish this device as a root if none of its
++ * parent bridges are exported
++ */
++ found = 0;
++ dev = dev_entry->dev->bus->self;
++ for (; !found && dev != NULL; dev = dev->bus->self) {
++ list_for_each_entry(e, &dev_data->dev_list, list) {
++ if (dev == e->dev) {
++ found = 1;
++ break;
++ }
++ }
++ }
++
++ domain = (unsigned int)pci_domain_nr(dev_entry->dev->bus);
++ bus = (unsigned int)dev_entry->dev->bus->number;
++
++ if (!found) {
++ err = publish_root_cb(pdev, domain, bus);
++ if (err)
++ break;
++ }
++ }
++
++ spin_unlock(&dev_data->lock);
++
++ return err;
++}
++
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ struct passthrough_dev_data *dev_data = pdev->pci_dev_data;
++ struct pci_dev_entry *dev_entry, *t;
++
++ list_for_each_entry_safe(dev_entry, t, &dev_data->dev_list, list) {
++ list_del(&dev_entry->list);
++ pcistub_put_pci_dev(dev_entry->dev);
++ kfree(dev_entry);
++ }
++
++ kfree(dev_data);
++ pdev->pci_dev_data = NULL;
++}
+diff --git a/drivers/xen/pciback/pci_stub.c b/drivers/xen/pciback/pci_stub.c
+new file mode 100644
+index 0000000..f4bef49
+--- /dev/null
++++ b/drivers/xen/pciback/pci_stub.c
+@@ -0,0 +1,949 @@
++/*
++ * PCI Stub Driver - Grabs devices in backend to be exported later
++ *
++ * Ryan Wilson <hap9 at epoch.ncsc.mil>
++ * Chris Bookholt <hap10 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/kref.h>
++#include <asm/atomic.h>
++#include "pciback.h"
++#include "conf_space.h"
++#include "conf_space_quirks.h"
++
++static char *pci_devs_to_hide = NULL;
++module_param_named(hide, pci_devs_to_hide, charp, 0444);
++
++struct pcistub_device_id {
++ struct list_head slot_list;
++ int domain;
++ unsigned char bus;
++ unsigned int devfn;
++};
++static LIST_HEAD(pcistub_device_ids);
++static DEFINE_SPINLOCK(device_ids_lock);
++
++struct pcistub_device {
++ struct kref kref;
++ struct list_head dev_list;
++ spinlock_t lock;
++
++ struct pci_dev *dev;
++ struct pciback_device *pdev; /* non-NULL if struct pci_dev is in use */
++};
++
++/* Access to pcistub_devices & seized_devices lists and the initialize_devices
++ * flag must be locked with pcistub_devices_lock
++ */
++static DEFINE_SPINLOCK(pcistub_devices_lock);
++static LIST_HEAD(pcistub_devices);
++
++/* wait for device_initcall before initializing our devices
++ * (see pcistub_init_devices_late)
++ */
++static int initialize_devices = 0;
++static LIST_HEAD(seized_devices);
++
++static struct pcistub_device *pcistub_device_alloc(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
++
++ dev_dbg(&dev->dev, "pcistub_device_alloc\n");
++
++ psdev = kzalloc(sizeof(*psdev), GFP_ATOMIC);
++ if (!psdev)
++ return NULL;
++
++ psdev->dev = pci_dev_get(dev);
++ if (!psdev->dev) {
++ kfree(psdev);
++ return NULL;
++ }
++
++ kref_init(&psdev->kref);
++ spin_lock_init(&psdev->lock);
++
++ return psdev;
++}
++
++/* Don't call this directly as it's called by pcistub_device_put */
++static void pcistub_device_release(struct kref *kref)
++{
++ struct pcistub_device *psdev;
++
++ psdev = container_of(kref, struct pcistub_device, kref);
++
++ dev_dbg(&psdev->dev->dev, "pcistub_device_release\n");
++
++ /* Clean-up the device */
++ pciback_reset_device(psdev->dev);
++ pciback_config_free_dyn_fields(psdev->dev);
++ pciback_config_free_dev(psdev->dev);
++ kfree(pci_get_drvdata(psdev->dev));
++ pci_set_drvdata(psdev->dev, NULL);
++
++ pci_dev_put(psdev->dev);
++
++ kfree(psdev);
++}
++
++static inline void pcistub_device_get(struct pcistub_device *psdev)
++{
++ kref_get(&psdev->kref);
++}
++
++static inline void pcistub_device_put(struct pcistub_device *psdev)
++{
++ kref_put(&psdev->kref, pcistub_device_release);
++}
++
++static struct pcistub_device *pcistub_device_find(int domain, int bus,
++ int slot, int func)
++{
++ struct pcistub_device *psdev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev != NULL
++ && domain == pci_domain_nr(psdev->dev->bus)
++ && bus == psdev->dev->bus->number
++ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
++ pcistub_device_get(psdev);
++ goto out;
++ }
++ }
++
++ /* didn't find it */
++ psdev = NULL;
++
++ out:
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return psdev;
++}
++
++static struct pci_dev *pcistub_device_get_pci_dev(struct pciback_device *pdev,
++ struct pcistub_device *psdev)
++{
++ struct pci_dev *pci_dev = NULL;
++ unsigned long flags;
++
++ pcistub_device_get(psdev);
++
++ spin_lock_irqsave(&psdev->lock, flags);
++ if (!psdev->pdev) {
++ psdev->pdev = pdev;
++ pci_dev = psdev->dev;
++ }
++ spin_unlock_irqrestore(&psdev->lock, flags);
++
++ if (!pci_dev)
++ pcistub_device_put(psdev);
++
++ return pci_dev;
++}
++
++struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
++ int domain, int bus,
++ int slot, int func)
++{
++ struct pcistub_device *psdev;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev != NULL
++ && domain == pci_domain_nr(psdev->dev->bus)
++ && bus == psdev->dev->bus->number
++ && PCI_DEVFN(slot, func) == psdev->dev->devfn) {
++ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return found_dev;
++}
++
++struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
++ struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_dev = pcistub_device_get_pci_dev(pdev, psdev);
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return found_dev;
++}
++
++void pcistub_put_pci_dev(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev, *found_psdev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_psdev = psdev;
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ /* Cleanup our device
++ * (so it's ready for the next domain)
++ */
++ pciback_reset_device(found_psdev->dev);
++ pciback_config_free_dyn_fields(found_psdev->dev);
++ pciback_config_reset_dev(found_psdev->dev);
++
++ spin_lock_irqsave(&found_psdev->lock, flags);
++ found_psdev->pdev = NULL;
++ spin_unlock_irqrestore(&found_psdev->lock, flags);
++
++ pcistub_device_put(found_psdev);
++}
++
++static int __devinit pcistub_match_one(struct pci_dev *dev,
++ struct pcistub_device_id *pdev_id)
++{
++ /* Match the specified device by domain, bus, slot, func and also if
++ * any of the device's parent bridges match.
++ */
++ for (; dev != NULL; dev = dev->bus->self) {
++ if (pci_domain_nr(dev->bus) == pdev_id->domain
++ && dev->bus->number == pdev_id->bus
++ && dev->devfn == pdev_id->devfn)
++ return 1;
++
++ /* Sometimes topmost bridge links to itself. */
++ if (dev == dev->bus->self)
++ break;
++ }
++
++ return 0;
++}
++
++static int __devinit pcistub_match(struct pci_dev *dev)
++{
++ struct pcistub_device_id *pdev_id;
++ unsigned long flags;
++ int found = 0;
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(pdev_id, &pcistub_device_ids, slot_list) {
++ if (pcistub_match_one(dev, pdev_id)) {
++ found = 1;
++ break;
++ }
++ }
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return found;
++}
++
++static int __devinit pcistub_init_device(struct pci_dev *dev)
++{
++ struct pciback_dev_data *dev_data;
++ int err = 0;
++
++ dev_dbg(&dev->dev, "initializing...\n");
++
++ /* The PCI backend is not intended to be a module (or to work with
++ * removable PCI devices (yet). If it were, pciback_config_free()
++ * would need to be called somewhere to free the memory allocated
++ * here and then to call kfree(pci_get_drvdata(psdev->dev)).
++ */
++ dev_data = kzalloc(sizeof(*dev_data), GFP_ATOMIC);
++ if (!dev_data) {
++ err = -ENOMEM;
++ goto out;
++ }
++ pci_set_drvdata(dev, dev_data);
++
++ dev_dbg(&dev->dev, "initializing config\n");
++ err = pciback_config_init_dev(dev);
++ if (err)
++ goto out;
++
++ /* HACK: Force device (& ACPI) to determine what IRQ it's on - we
++ * must do this here because pcibios_enable_device may specify
++ * the pci device's true irq (and possibly its other resources)
++ * if they differ from what's in the configuration space.
++ * This makes the assumption that the device's resources won't
++ * change after this point (otherwise this code may break!)
++ */
++ dev_dbg(&dev->dev, "enabling device\n");
++ err = pci_enable_device(dev);
++ if (err)
++ goto config_release;
++
++ /* Now disable the device (this also ensures some private device
++ * data is setup before we export)
++ */
++ dev_dbg(&dev->dev, "reset device\n");
++ pciback_reset_device(dev);
++
++ return 0;
++
++ config_release:
++ pciback_config_free_dev(dev);
++
++ out:
++ pci_set_drvdata(dev, NULL);
++ kfree(dev_data);
++ return err;
++}
++
++/*
++ * Because some initialization still happens on
++ * devices during fs_initcall, we need to defer
++ * full initialization of our devices until
++ * device_initcall.
++ */
++static int __init pcistub_init_devices_late(void)
++{
++ struct pcistub_device *psdev;
++ unsigned long flags;
++ int err = 0;
++
++ pr_debug("pciback: pcistub_init_devices_late\n");
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ while (!list_empty(&seized_devices)) {
++ psdev = container_of(seized_devices.next,
++ struct pcistub_device, dev_list);
++ list_del(&psdev->dev_list);
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ err = pcistub_init_device(psdev->dev);
++ if (err) {
++ dev_err(&psdev->dev->dev,
++ "error %d initializing device\n", err);
++ kfree(psdev);
++ psdev = NULL;
++ }
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (psdev)
++ list_add_tail(&psdev->dev_list, &pcistub_devices);
++ }
++
++ initialize_devices = 1;
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ return 0;
++}
++
++static int __devinit pcistub_seize(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
++ unsigned long flags;
++ int err = 0;
++
++ psdev = pcistub_device_alloc(dev);
++ if (!psdev)
++ return -ENOMEM;
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (initialize_devices) {
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ /* don't want irqs disabled when calling pcistub_init_device */
++ err = pcistub_init_device(psdev->dev);
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ if (!err)
++ list_add(&psdev->dev_list, &pcistub_devices);
++ } else {
++ dev_dbg(&dev->dev, "deferring initialization\n");
++ list_add(&psdev->dev_list, &seized_devices);
++ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ if (err)
++ pcistub_device_put(psdev);
++
++ return err;
++}
++
++static int __devinit pcistub_probe(struct pci_dev *dev,
++ const struct pci_device_id *id)
++{
++ int err = 0;
++
++ dev_dbg(&dev->dev, "probing...\n");
++
++ if (pcistub_match(dev)) {
++
++ if (dev->hdr_type != PCI_HEADER_TYPE_NORMAL
++ && dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) {
++ dev_err(&dev->dev, "can't export pci devices that "
++ "don't have a normal (0) or bridge (1) "
++ "header type!\n");
++ err = -ENODEV;
++ goto out;
++ }
++
++ dev_info(&dev->dev, "seizing device\n");
++ err = pcistub_seize(dev);
++ } else
++ /* Didn't find the device */
++ err = -ENODEV;
++
++ out:
++ return err;
++}
++
++static void pcistub_remove(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev, *found_psdev = NULL;
++ unsigned long flags;
++
++ dev_dbg(&dev->dev, "removing\n");
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++
++ pciback_config_quirk_release(dev);
++
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (psdev->dev == dev) {
++ found_psdev = psdev;
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ if (found_psdev) {
++ dev_dbg(&dev->dev, "found device to remove - in use? %p\n",
++ found_psdev->pdev);
++
++ if (found_psdev->pdev) {
++ printk(KERN_WARNING "pciback: ****** removing device "
++ "%s while still in-use! ******\n",
++ pci_name(found_psdev->dev));
++ printk(KERN_WARNING "pciback: ****** driver domain may "
++ "still access this device's i/o resources!\n");
++ printk(KERN_WARNING "pciback: ****** shutdown driver "
++ "domain before binding device\n");
++ printk(KERN_WARNING "pciback: ****** to other drivers "
++ "or domains\n");
++
++ pciback_release_pci_dev(found_psdev->pdev,
++ found_psdev->dev);
++ }
++
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ list_del(&found_psdev->dev_list);
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++
++ /* the final put for releasing from the list */
++ pcistub_device_put(found_psdev);
++ }
++}
++
++static struct pci_device_id pcistub_ids[] = {
++ {
++ .vendor = PCI_ANY_ID,
++ .device = PCI_ANY_ID,
++ .subvendor = PCI_ANY_ID,
++ .subdevice = PCI_ANY_ID,
++ },
++ {0,},
++};
++
++/*
++ * Note: There is no MODULE_DEVICE_TABLE entry here because this isn't
++ * for a normal device. I don't want it to be loaded automatically.
++ */
++
++static struct pci_driver pciback_pci_driver = {
++ .name = "pciback",
++ .id_table = pcistub_ids,
++ .probe = pcistub_probe,
++ .remove = pcistub_remove,
++};
++
++static inline int str_to_slot(const char *buf, int *domain, int *bus,
++ int *slot, int *func)
++{
++ int err;
++
++ err = sscanf(buf, " %x:%x:%x.%x", domain, bus, slot, func);
++ if (err == 4)
++ return 0;
++ else if (err < 0)
++ return -EINVAL;
++
++ /* try again without domain */
++ *domain = 0;
++ err = sscanf(buf, " %x:%x.%x", bus, slot, func);
++ if (err == 3)
++ return 0;
++
++ return -EINVAL;
++}
++
++static inline int str_to_quirk(const char *buf, int *domain, int *bus, int
++ *slot, int *func, int *reg, int *size, int *mask)
++{
++ int err;
++
++ err =
++ sscanf(buf, " %04x:%02x:%02x.%1x-%08x:%1x:%08x", domain, bus, slot,
++ func, reg, size, mask);
++ if (err == 7)
++ return 0;
++ return -EINVAL;
++}
++
++static int pcistub_device_id_add(int domain, int bus, int slot, int func)
++{
++ struct pcistub_device_id *pci_dev_id;
++ unsigned long flags;
++
++ pci_dev_id = kmalloc(sizeof(*pci_dev_id), GFP_KERNEL);
++ if (!pci_dev_id)
++ return -ENOMEM;
++
++ pci_dev_id->domain = domain;
++ pci_dev_id->bus = bus;
++ pci_dev_id->devfn = PCI_DEVFN(slot, func);
++
++ pr_debug("pciback: wants to seize %04x:%02x:%02x.%01x\n",
++ domain, bus, slot, func);
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_add_tail(&pci_dev_id->slot_list, &pcistub_device_ids);
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return 0;
++}
++
++static int pcistub_device_id_remove(int domain, int bus, int slot, int func)
++{
++ struct pcistub_device_id *pci_dev_id, *t;
++ int devfn = PCI_DEVFN(slot, func);
++ int err = -ENOENT;
++ unsigned long flags;
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry_safe(pci_dev_id, t, &pcistub_device_ids, slot_list) {
++
++ if (pci_dev_id->domain == domain
++ && pci_dev_id->bus == bus && pci_dev_id->devfn == devfn) {
++ /* Don't break; here because it's possible the same
++ * slot could be in the list more than once
++ */
++ list_del(&pci_dev_id->slot_list);
++ kfree(pci_dev_id);
++
++ err = 0;
++
++ pr_debug("pciback: removed %04x:%02x:%02x.%01x from "
++ "seize list\n", domain, bus, slot, func);
++ }
++ }
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return err;
++}
++
++static int pcistub_reg_add(int domain, int bus, int slot, int func, int reg,
++ int size, int mask)
++{
++ int err = 0;
++ struct pcistub_device *psdev;
++ struct pci_dev *dev;
++ struct config_field *field;
++
++ psdev = pcistub_device_find(domain, bus, slot, func);
++ if (!psdev || !psdev->dev) {
++ err = -ENODEV;
++ goto out;
++ }
++ dev = psdev->dev;
++
++ field = kzalloc(sizeof(*field), GFP_ATOMIC);
++ if (!field) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ field->offset = reg;
++ field->size = size;
++ field->mask = mask;
++ field->init = NULL;
++ field->reset = NULL;
++ field->release = NULL;
++ field->clean = pciback_config_field_free;
++
++ err = pciback_config_quirks_add_field(dev, field);
++ if (err)
++ kfree(field);
++ out:
++ return err;
++}
++
++static ssize_t pcistub_slot_add(struct device_driver *drv, const char *buf,
++ size_t count)
++{
++ int domain, bus, slot, func;
++ int err;
++
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
++
++ err = pcistub_device_id_add(domain, bus, slot, func);
++
++ out:
++ if (!err)
++ err = count;
++ return err;
++}
++
++DRIVER_ATTR(new_slot, S_IWUSR, NULL, pcistub_slot_add);
++
++static ssize_t pcistub_slot_remove(struct device_driver *drv, const char *buf,
++ size_t count)
++{
++ int domain, bus, slot, func;
++ int err;
++
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
++
++ err = pcistub_device_id_remove(domain, bus, slot, func);
++
++ out:
++ if (!err)
++ err = count;
++ return err;
++}
++
++DRIVER_ATTR(remove_slot, S_IWUSR, NULL, pcistub_slot_remove);
++
++static ssize_t pcistub_slot_show(struct device_driver *drv, char *buf)
++{
++ struct pcistub_device_id *pci_dev_id;
++ size_t count = 0;
++ unsigned long flags;
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(pci_dev_id, &pcistub_device_ids, slot_list) {
++ if (count >= PAGE_SIZE)
++ break;
++
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "%04x:%02x:%02x.%01x\n",
++ pci_dev_id->domain, pci_dev_id->bus,
++ PCI_SLOT(pci_dev_id->devfn),
++ PCI_FUNC(pci_dev_id->devfn));
++ }
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return count;
++}
++
++DRIVER_ATTR(slots, S_IRUSR, pcistub_slot_show, NULL);
++
++static ssize_t pcistub_quirk_add(struct device_driver *drv, const char *buf,
++ size_t count)
++{
++ int domain, bus, slot, func, reg, size, mask;
++ int err;
++
++ err = str_to_quirk(buf, &domain, &bus, &slot, &func, ®, &size,
++ &mask);
++ if (err)
++ goto out;
++
++ err = pcistub_reg_add(domain, bus, slot, func, reg, size, mask);
++
++ out:
++ if (!err)
++ err = count;
++ return err;
++}
++
++static ssize_t pcistub_quirk_show(struct device_driver *drv, char *buf)
++{
++ int count = 0;
++ unsigned long flags;
++ extern struct list_head pciback_quirks;
++ struct pciback_config_quirk *quirk;
++ struct pciback_dev_data *dev_data;
++ struct config_field *field;
++ struct config_field_entry *cfg_entry;
++
++ spin_lock_irqsave(&device_ids_lock, flags);
++ list_for_each_entry(quirk, &pciback_quirks, quirks_list) {
++ if (count >= PAGE_SIZE)
++ goto out;
++
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "%02x:%02x.%01x\n\t%04x:%04x:%04x:%04x\n",
++ quirk->pdev->bus->number,
++ PCI_SLOT(quirk->pdev->devfn),
++ PCI_FUNC(quirk->pdev->devfn),
++ quirk->devid.vendor, quirk->devid.device,
++ quirk->devid.subvendor,
++ quirk->devid.subdevice);
++
++ dev_data = pci_get_drvdata(quirk->pdev);
++
++ list_for_each_entry(cfg_entry, &dev_data->config_fields, list) {
++ field = cfg_entry->field;
++ if (count >= PAGE_SIZE)
++ goto out;
++
++ count += scnprintf(buf + count, PAGE_SIZE - count,
++ "\t\t%08x:%01x:%08x\n",
++ cfg_entry->base_offset + field->offset,
++ field->size, field->mask);
++ }
++ }
++
++ out:
++ spin_unlock_irqrestore(&device_ids_lock, flags);
++
++ return count;
++}
++
++DRIVER_ATTR(quirks, S_IRUSR | S_IWUSR, pcistub_quirk_show, pcistub_quirk_add);
++
++static ssize_t permissive_add(struct device_driver *drv, const char *buf,
++ size_t count)
++{
++ int domain, bus, slot, func;
++ int err;
++ struct pcistub_device *psdev;
++ struct pciback_dev_data *dev_data;
++ err = str_to_slot(buf, &domain, &bus, &slot, &func);
++ if (err)
++ goto out;
++ psdev = pcistub_device_find(domain, bus, slot, func);
++ if (!psdev) {
++ err = -ENODEV;
++ goto out;
++ }
++ if (!psdev->dev) {
++ err = -ENODEV;
++ goto release;
++ }
++ dev_data = pci_get_drvdata(psdev->dev);
++ /* the driver data for a device should never be null at this point */
++ if (!dev_data) {
++ err = -ENXIO;
++ goto release;
++ }
++ if (!dev_data->permissive) {
++ dev_data->permissive = 1;
++ /* Let user know that what they're doing could be unsafe */
++ dev_warn(&psdev->dev->dev,
++ "enabling permissive mode configuration space accesses!\n");
++ dev_warn(&psdev->dev->dev,
++ "permissive mode is potentially unsafe!\n");
++ }
++ release:
++ pcistub_device_put(psdev);
++ out:
++ if (!err)
++ err = count;
++ return err;
++}
++
++static ssize_t permissive_show(struct device_driver *drv, char *buf)
++{
++ struct pcistub_device *psdev;
++ struct pciback_dev_data *dev_data;
++ size_t count = 0;
++ unsigned long flags;
++ spin_lock_irqsave(&pcistub_devices_lock, flags);
++ list_for_each_entry(psdev, &pcistub_devices, dev_list) {
++ if (count >= PAGE_SIZE)
++ break;
++ if (!psdev->dev)
++ continue;
++ dev_data = pci_get_drvdata(psdev->dev);
++ if (!dev_data || !dev_data->permissive)
++ continue;
++ count +=
++ scnprintf(buf + count, PAGE_SIZE - count, "%s\n",
++ pci_name(psdev->dev));
++ }
++ spin_unlock_irqrestore(&pcistub_devices_lock, flags);
++ return count;
++}
++
++DRIVER_ATTR(permissive, S_IRUSR | S_IWUSR, permissive_show, permissive_add);
++
++#ifdef CONFIG_PCI_MSI
++
++int pciback_get_owner(struct pci_dev *dev)
++{
++ struct pcistub_device *psdev;
++
++ psdev = pcistub_device_find(pci_domain_nr(dev->bus), dev->bus->number,
++ PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn));
++ /* XXX will other domain has pciback support ??? */
++ if (!psdev || !psdev->pdev) {
++ printk(KERN_WARNING "no ownder\n");
++ return -1;
++ }
++ return psdev->pdev->xdev->otherend_id;
++}
++#endif
++
++static void pcistub_exit(void)
++{
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_new_slot);
++ driver_remove_file(&pciback_pci_driver.driver,
++ &driver_attr_remove_slot);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_slots);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_quirks);
++ driver_remove_file(&pciback_pci_driver.driver, &driver_attr_permissive);
++
++ pci_unregister_driver(&pciback_pci_driver);
++ WARN_ON(unregister_msi_get_owner(pciback_get_owner));
++}
++
++static int __init pcistub_init(void)
++{
++ int pos = 0;
++ int err = 0;
++ int domain, bus, slot, func;
++ int parsed;
++
++ if (pci_devs_to_hide && *pci_devs_to_hide) {
++ do {
++ parsed = 0;
++
++ err = sscanf(pci_devs_to_hide + pos,
++ " (%x:%x:%x.%x) %n",
++ &domain, &bus, &slot, &func, &parsed);
++ if (err != 4) {
++ domain = 0;
++ err = sscanf(pci_devs_to_hide + pos,
++ " (%x:%x.%x) %n",
++ &bus, &slot, &func, &parsed);
++ if (err != 3)
++ goto parse_error;
++ }
++
++ err = pcistub_device_id_add(domain, bus, slot, func);
++ if (err)
++ goto out;
++
++ /* if parsed<=0, we've reached the end of the string */
++ pos += parsed;
++ } while (parsed > 0 && pci_devs_to_hide[pos]);
++ }
++
++ /* If we're the first PCI Device Driver to register, we're the
++ * first one to get offered PCI devices as they become
++ * available (and thus we can be the first to grab them)
++ */
++ err = pci_register_driver(&pciback_pci_driver);
++ if (err < 0)
++ goto out;
++
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_new_slot);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_remove_slot);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_slots);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_quirks);
++ if (!err)
++ err = driver_create_file(&pciback_pci_driver.driver,
++ &driver_attr_permissive);
++
++ if (!err)
++ err = register_msi_get_owner(pciback_get_owner);
++ if (err)
++ pcistub_exit();
++
++ out:
++ return err;
++
++ parse_error:
++ printk(KERN_ERR "pciback: Error parsing pci_devs_to_hide at \"%s\"\n",
++ pci_devs_to_hide + pos);
++ return -EINVAL;
++}
++
++#ifndef MODULE
++/*
++ * fs_initcall happens before device_initcall
++ * so pciback *should* get called first (b/c we
++ * want to suck up any device before other drivers
++ * get a chance by being the first pci device
++ * driver to register)
++ */
++fs_initcall(pcistub_init);
++#endif
++
++static int __init pciback_init(void)
++{
++ int err;
++
++ err = pciback_config_init();
++ if (err)
++ return err;
++
++#ifdef MODULE
++ err = pcistub_init();
++ if (err < 0)
++ return err;
++#endif
++
++ pcistub_init_devices_late();
++ err = pciback_xenbus_register();
++ if (err)
++ pcistub_exit();
++
++ return err;
++}
++
++static void __exit pciback_cleanup(void)
++{
++ pciback_xenbus_unregister();
++ pcistub_exit();
++}
++
++module_init(pciback_init);
++module_exit(pciback_cleanup);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/pciback/pciback.h b/drivers/xen/pciback/pciback.h
+new file mode 100644
+index 0000000..14c85a7
+--- /dev/null
++++ b/drivers/xen/pciback/pciback.h
+@@ -0,0 +1,111 @@
++/*
++ * PCI Backend Common Data Structures & Function Declarations
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_PCIBACK_H__
++#define __XEN_PCIBACK_H__
++
++#include <linux/pci.h>
++#include <linux/interrupt.h>
++#include <xen/xenbus.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/workqueue.h>
++#include <asm/atomic.h>
++#include <xen/interface/io/pciif.h>
++
++struct pci_dev_entry {
++ struct list_head list;
++ struct pci_dev *dev;
++};
++
++#define _PDEVF_op_active (0)
++#define PDEVF_op_active (1<<(_PDEVF_op_active))
++
++struct pciback_device {
++ void *pci_dev_data;
++ spinlock_t dev_lock;
++
++ struct xenbus_device *xdev;
++
++ struct xenbus_watch be_watch;
++ u8 be_watching;
++
++ int evtchn_irq;
++
++ struct vm_struct *sh_area;
++ struct xen_pci_sharedinfo *sh_info;
++
++ unsigned long flags;
++
++ struct work_struct op_work;
++};
++
++struct pciback_dev_data {
++ struct list_head config_fields;
++ int permissive;
++ int warned_on_write;
++};
++
++/* Get/Put PCI Devices that are hidden from the PCI Backend Domain */
++struct pci_dev *pcistub_get_pci_dev_by_slot(struct pciback_device *pdev,
++ int domain, int bus,
++ int slot, int func);
++struct pci_dev *pcistub_get_pci_dev(struct pciback_device *pdev,
++ struct pci_dev *dev);
++void pcistub_put_pci_dev(struct pci_dev *dev);
++
++/* Ensure a device is turned off or reset */
++void pciback_reset_device(struct pci_dev *pdev);
++
++/* Access a virtual configuration space for a PCI device */
++int pciback_config_init(void);
++int pciback_config_init_dev(struct pci_dev *dev);
++void pciback_config_free_dyn_fields(struct pci_dev *dev);
++void pciback_config_reset_dev(struct pci_dev *dev);
++void pciback_config_free_dev(struct pci_dev *dev);
++int pciback_config_read(struct pci_dev *dev, int offset, int size,
++ u32 * ret_val);
++int pciback_config_write(struct pci_dev *dev, int offset, int size, u32 value);
++
++/* Handle requests for specific devices from the frontend */
++typedef int (*publish_pci_dev_cb) (struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn, unsigned int devid);
++typedef int (*publish_pci_root_cb) (struct pciback_device * pdev,
++ unsigned int domain, unsigned int bus);
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb);
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev);
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn);
++int pciback_init_devices(struct pciback_device *pdev);
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb cb);
++void pciback_release_devices(struct pciback_device *pdev);
++
++/* Handles events from front-end */
++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);
++
++#ifdef CONFIG_PCI_MSI
++int pciback_enable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
++
++int pciback_disable_msi(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
++
++
++int pciback_enable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
++
++int pciback_disable_msix(struct pciback_device *pdev,
++ struct pci_dev *dev, struct xen_pci_op *op);
++#endif
++extern int verbose_request;
++#endif
+diff --git a/drivers/xen/pciback/pciback_ops.c b/drivers/xen/pciback/pciback_ops.c
+new file mode 100644
+index 0000000..a3663df
+--- /dev/null
++++ b/drivers/xen/pciback/pciback_ops.c
+@@ -0,0 +1,117 @@
++/*
++ * PCI Backend Operations - respond to PCI requests from Frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <asm/bitops.h>
++#include <xen/evtchn.h>
++#include "pciback.h"
++
++int verbose_request = 0;
++module_param(verbose_request, int, 0644);
++
++/* Ensure a device is "turned off" and ready to be exported.
++ * (Also see pciback_config_reset to ensure virtual configuration space is
++ * ready to be re-exported)
++ */
++void pciback_reset_device(struct pci_dev *dev)
++{
++ u16 cmd;
++
++ /* Disable devices (but not bridges) */
++ if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) {
++ pci_disable_device(dev);
++
++ pci_write_config_word(dev, PCI_COMMAND, 0);
++
++ atomic_set(&dev->enable_cnt, 0);
++ dev->is_busmaster = 0;
++ } else {
++ pci_read_config_word(dev, PCI_COMMAND, &cmd);
++ if (cmd & (PCI_COMMAND_INVALIDATE)) {
++ cmd &= ~(PCI_COMMAND_INVALIDATE);
++ pci_write_config_word(dev, PCI_COMMAND, cmd);
++
++ dev->is_busmaster = 0;
++ }
++ }
++}
++
++static inline void test_and_schedule_op(struct pciback_device *pdev)
++{
++ /* Check that frontend is requesting an operation and that we are not
++ * already processing a request */
++ if (test_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags)
++ && !test_and_set_bit(_PDEVF_op_active, &pdev->flags))
++ schedule_work(&pdev->op_work);
++}
++
++/* 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(struct work_struct *work)
++{
++ struct pciback_device *pdev = container_of(work, struct pciback_device, op_work);
++ struct pci_dev *dev;
++ struct xen_pci_op *op = &pdev->sh_info->op;
++
++ dev = pciback_get_pci_dev(pdev, op->domain, op->bus, op->devfn);
++
++ if (dev == NULL)
++ op->err = XEN_PCI_ERR_dev_not_found;
++ else
++ {
++ switch (op->cmd)
++ {
++ case XEN_PCI_OP_conf_read:
++ op->err = pciback_config_read(dev,
++ op->offset, op->size, &op->value);
++ break;
++ case XEN_PCI_OP_conf_write:
++ op->err = pciback_config_write(dev,
++ op->offset, op->size, op->value);
++ break;
++#ifdef CONFIG_PCI_MSI
++ case XEN_PCI_OP_enable_msi:
++ op->err = pciback_enable_msi(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_disable_msi:
++ op->err = pciback_disable_msi(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_enable_msix:
++ op->err = pciback_enable_msix(pdev, dev, op);
++ break;
++ case XEN_PCI_OP_disable_msix:
++ op->err = pciback_disable_msix(pdev, dev, op);
++ break;
++#endif
++ default:
++ op->err = XEN_PCI_ERR_not_implemented;
++ break;
++ }
++ }
++ /* Tell the driver domain that we're done. */
++ wmb();
++ clear_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
++ notify_remote_via_irq(pdev->evtchn_irq);
++
++ /* Mark that we're done. */
++ smp_mb__before_clear_bit(); /* /after/ clearing PCIF_active */
++ clear_bit(_PDEVF_op_active, &pdev->flags);
++ smp_mb__after_clear_bit(); /* /before/ final check for work */
++
++ /* Check to see if the driver domain tried to start another request in
++ * between clearing _XEN_PCIF_active and clearing _PDEVF_op_active. */
++ test_and_schedule_op(pdev);
++}
++
++irqreturn_t pciback_handle_event(int irq, void *dev_id)
++{
++ struct pciback_device *pdev = dev_id;
++
++ test_and_schedule_op(pdev);
++
++ return IRQ_HANDLED;
++}
+diff --git a/drivers/xen/pciback/slot.c b/drivers/xen/pciback/slot.c
+new file mode 100644
+index 0000000..5a96092
+--- /dev/null
++++ b/drivers/xen/pciback/slot.c
+@@ -0,0 +1,157 @@
++/*
++ * PCI Backend - Provides a Virtual PCI bus (with real devices)
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil> (vpci.c)
++ * Author: Tristan Gingold <tristan.gingold at bull.net>, from vpci.c
++ */
++
++#include <linux/list.h>
++#include <linux/slab.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
++
++/* There are at most 32 slots in a pci bus. */
++#define PCI_SLOT_MAX 32
++
++#define PCI_BUS_NBR 2
++
++struct slot_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct pci_dev *slots[PCI_BUS_NBR][PCI_SLOT_MAX];
++ spinlock_t lock;
++};
++
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct pci_dev *dev = NULL;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ unsigned long flags;
++
++ if (domain != 0 || PCI_FUNC(devfn) != 0)
++ return NULL;
++
++ if (PCI_SLOT(devfn) >= PCI_SLOT_MAX || bus >= PCI_BUS_NBR)
++ return NULL;
++
++ spin_lock_irqsave(&slot_dev->lock, flags);
++ dev = slot_dev->slots[bus][PCI_SLOT(devfn)];
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
++
++ return dev;
++}
++
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
++{
++ int err = 0, slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ unsigned long flags;
++
++ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
++ err = -EFAULT;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Can't export bridges on the virtual PCI bus");
++ goto out;
++ }
++
++ spin_lock_irqsave(&slot_dev->lock, flags);
++
++ /* Assign to a new slot on the virtual PCI bus */
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (slot_dev->slots[bus][slot] == NULL) {
++ printk(KERN_INFO
++ "pciback: slot: %s: assign to virtual slot %d, bus %d\n",
++ pci_name(dev), slot, bus);
++ slot_dev->slots[bus][slot] = dev;
++ goto unlock;
++ }
++ }
++
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "No more space on root virtual PCI bus");
++
++ unlock:
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
++
++ /* Publish this device. */
++ if(!err)
++ err = publish_cb(pdev, 0, 0, PCI_DEVFN(slot, 0), devid);
++
++ out:
++ return err;
++}
++
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++{
++ int slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&slot_dev->lock, flags);
++
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (slot_dev->slots[bus][slot] == dev) {
++ slot_dev->slots[bus][slot] = NULL;
++ found_dev = dev;
++ goto out;
++ }
++ }
++
++ out:
++ spin_unlock_irqrestore(&slot_dev->lock, flags);
++
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
++}
++
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ int slot, bus;
++ struct slot_dev_data *slot_dev;
++
++ slot_dev = kmalloc(sizeof(*slot_dev), GFP_KERNEL);
++ if (!slot_dev)
++ return -ENOMEM;
++
++ spin_lock_init(&slot_dev->lock);
++
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++)
++ slot_dev->slots[bus][slot] = NULL;
++
++ pdev->pci_dev_data = slot_dev;
++
++ return 0;
++}
++
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_cb)
++{
++ /* The Virtual PCI bus has only one root */
++ return publish_cb(pdev, 0, 0);
++}
++
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ int slot, bus;
++ struct slot_dev_data *slot_dev = pdev->pci_dev_data;
++ struct pci_dev *dev;
++
++ for (bus = 0; bus < PCI_BUS_NBR; bus++)
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ dev = slot_dev->slots[bus][slot];
++ if (dev != NULL)
++ pcistub_put_pci_dev(dev);
++ }
++
++ kfree(slot_dev);
++ pdev->pci_dev_data = NULL;
++}
+diff --git a/drivers/xen/pciback/vpci.c b/drivers/xen/pciback/vpci.c
+new file mode 100644
+index 0000000..09aa2e7
+--- /dev/null
++++ b/drivers/xen/pciback/vpci.c
+@@ -0,0 +1,212 @@
++/*
++ * PCI Backend - Provides a Virtual PCI bus (with real devices)
++ * to the frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++
++#include <linux/list.h>
++#include <linux/slab.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pciback.h"
++
++#define PCI_SLOT_MAX 32
++
++struct vpci_dev_data {
++ /* Access to dev_list must be protected by lock */
++ struct list_head dev_list[PCI_SLOT_MAX];
++ spinlock_t lock;
++};
++
++static inline struct list_head *list_first(struct list_head *head)
++{
++ return head->next;
++}
++
++struct pci_dev *pciback_get_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn)
++{
++ struct pci_dev_entry *entry;
++ struct pci_dev *dev = NULL;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ unsigned long flags;
++
++ if (domain != 0 || bus != 0)
++ return NULL;
++
++ if (PCI_SLOT(devfn) < PCI_SLOT_MAX) {
++ spin_lock_irqsave(&vpci_dev->lock, flags);
++
++ list_for_each_entry(entry,
++ &vpci_dev->dev_list[PCI_SLOT(devfn)],
++ list) {
++ if (PCI_FUNC(entry->dev->devfn) == PCI_FUNC(devfn)) {
++ dev = entry->dev;
++ break;
++ }
++ }
++
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
++ }
++ return dev;
++}
++
++static inline int match_slot(struct pci_dev *l, struct pci_dev *r)
++{
++ if (pci_domain_nr(l->bus) == pci_domain_nr(r->bus)
++ && l->bus == r->bus && PCI_SLOT(l->devfn) == PCI_SLOT(r->devfn))
++ return 1;
++
++ return 0;
++}
++
++int pciback_add_pci_dev(struct pciback_device *pdev, struct pci_dev *dev,
++ int devid, publish_pci_dev_cb publish_cb)
++{
++ int err = 0, slot, func;
++ struct pci_dev_entry *t, *dev_entry;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ unsigned long flags;
++
++ if ((dev->class >> 24) == PCI_BASE_CLASS_BRIDGE) {
++ err = -EFAULT;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Can't export bridges on the virtual PCI bus");
++ goto out;
++ }
++
++ dev_entry = kmalloc(sizeof(*dev_entry), GFP_KERNEL);
++ if (!dev_entry) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error adding entry to virtual PCI bus");
++ goto out;
++ }
++
++ dev_entry->dev = dev;
++
++ spin_lock_irqsave(&vpci_dev->lock, flags);
++
++ /* Keep multi-function devices together on the virtual PCI bus */
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (!list_empty(&vpci_dev->dev_list[slot])) {
++ t = list_entry(list_first(&vpci_dev->dev_list[slot]),
++ struct pci_dev_entry, list);
++
++ if (match_slot(dev, t->dev)) {
++ pr_info("pciback: vpci: %s: "
++ "assign to virtual slot %d func %d\n",
++ pci_name(dev), slot,
++ PCI_FUNC(dev->devfn));
++ list_add_tail(&dev_entry->list,
++ &vpci_dev->dev_list[slot]);
++ func = PCI_FUNC(dev->devfn);
++ goto unlock;
++ }
++ }
++ }
++
++ /* Assign to a new slot on the virtual PCI bus */
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ if (list_empty(&vpci_dev->dev_list[slot])) {
++ printk(KERN_INFO
++ "pciback: vpci: %s: assign to virtual slot %d\n",
++ pci_name(dev), slot);
++ list_add_tail(&dev_entry->list,
++ &vpci_dev->dev_list[slot]);
++ func = PCI_FUNC(dev->devfn);
++ goto unlock;
++ }
++ }
++
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "No more space on root virtual PCI bus");
++
++ unlock:
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
++
++ /* Publish this device. */
++ if(!err)
++ err = publish_cb(pdev, 0, 0, PCI_DEVFN(slot, func), devid);
++
++ out:
++ return err;
++}
++
++void pciback_release_pci_dev(struct pciback_device *pdev, struct pci_dev *dev)
++{
++ int slot;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++ struct pci_dev *found_dev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&vpci_dev->lock, flags);
++
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ struct pci_dev_entry *e, *tmp;
++ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
++ list) {
++ if (e->dev == dev) {
++ list_del(&e->list);
++ found_dev = e->dev;
++ kfree(e);
++ goto out;
++ }
++ }
++ }
++
++ out:
++ spin_unlock_irqrestore(&vpci_dev->lock, flags);
++
++ if (found_dev)
++ pcistub_put_pci_dev(found_dev);
++}
++
++int pciback_init_devices(struct pciback_device *pdev)
++{
++ int slot;
++ struct vpci_dev_data *vpci_dev;
++
++ vpci_dev = kmalloc(sizeof(*vpci_dev), GFP_KERNEL);
++ if (!vpci_dev)
++ return -ENOMEM;
++
++ spin_lock_init(&vpci_dev->lock);
++
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ INIT_LIST_HEAD(&vpci_dev->dev_list[slot]);
++ }
++
++ pdev->pci_dev_data = vpci_dev;
++
++ return 0;
++}
++
++int pciback_publish_pci_roots(struct pciback_device *pdev,
++ publish_pci_root_cb publish_cb)
++{
++ /* The Virtual PCI bus has only one root */
++ return publish_cb(pdev, 0, 0);
++}
++
++void pciback_release_devices(struct pciback_device *pdev)
++{
++ int slot;
++ struct vpci_dev_data *vpci_dev = pdev->pci_dev_data;
++
++ for (slot = 0; slot < PCI_SLOT_MAX; slot++) {
++ struct pci_dev_entry *e, *tmp;
++ list_for_each_entry_safe(e, tmp, &vpci_dev->dev_list[slot],
++ list) {
++ list_del(&e->list);
++ pcistub_put_pci_dev(e->dev);
++ kfree(e);
++ }
++ }
++
++ kfree(vpci_dev);
++ pdev->pci_dev_data = NULL;
++}
+diff --git a/drivers/xen/pciback/xenbus.c b/drivers/xen/pciback/xenbus.c
+new file mode 100644
+index 0000000..91f3245
+--- /dev/null
++++ b/drivers/xen/pciback/xenbus.c
+@@ -0,0 +1,702 @@
++/*
++ * PCI Backend Xenbus Setup - handles setup with frontend and xend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/list.h>
++#include <linux/vmalloc.h>
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++#include "pciback.h"
++
++#define INVALID_EVTCHN_IRQ (-1)
++
++static struct pciback_device *alloc_pdev(struct xenbus_device *xdev)
++{
++ struct pciback_device *pdev;
++
++ pdev = kzalloc(sizeof(struct pciback_device), GFP_KERNEL);
++ if (pdev == NULL)
++ goto out;
++ dev_dbg(&xdev->dev, "allocated pdev @ 0x%p\n", pdev);
++
++ pdev->xdev = xdev;
++ xdev->dev.driver_data = pdev;
++
++ spin_lock_init(&pdev->dev_lock);
++
++ pdev->sh_area = NULL;
++ pdev->sh_info = NULL;
++ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
++ pdev->be_watching = 0;
++
++ INIT_WORK(&pdev->op_work, pciback_do_op);
++
++ if (pciback_init_devices(pdev)) {
++ kfree(pdev);
++ pdev = NULL;
++ }
++ out:
++ return pdev;
++}
++
++static void pciback_disconnect(struct pciback_device *pdev)
++{
++ spin_lock(&pdev->dev_lock);
++
++ /* Ensure the guest can't trigger our handler before removing devices */
++ if (pdev->evtchn_irq != INVALID_EVTCHN_IRQ) {
++ unbind_from_irqhandler(pdev->evtchn_irq, pdev);
++ pdev->evtchn_irq = INVALID_EVTCHN_IRQ;
++ }
++
++ /* If the driver domain started an op, make sure we complete it or
++ * delete it before releasing the shared memory */
++ flush_scheduled_work();
++
++ if (pdev->sh_info != NULL) {
++ xenbus_unmap_ring_vfree(pdev->xdev, pdev->sh_area);
++ pdev->sh_info = NULL;
++ }
++
++ spin_unlock(&pdev->dev_lock);
++}
++
++static void free_pdev(struct pciback_device *pdev)
++{
++ if (pdev->be_watching)
++ unregister_xenbus_watch(&pdev->be_watch);
++
++ pciback_disconnect(pdev);
++
++ pciback_release_devices(pdev);
++
++ pdev->xdev->dev.driver_data = NULL;
++ pdev->xdev = NULL;
++
++ kfree(pdev);
++}
++
++static int pciback_do_attach(struct pciback_device *pdev, int gnt_ref,
++ int remote_evtchn)
++{
++ int err = 0;
++ struct vm_struct *area;
++
++ dev_dbg(&pdev->xdev->dev,
++ "Attaching to frontend resources - gnt_ref=%d evtchn=%d\n",
++ gnt_ref, remote_evtchn);
++
++ area = xenbus_map_ring_valloc(pdev->xdev, gnt_ref);
++ if (IS_ERR(area)) {
++ err = PTR_ERR(area);
++ goto out;
++ }
++ pdev->sh_area = area;
++ pdev->sh_info = area->addr;
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ pdev->xdev->otherend_id, remote_evtchn, pciback_handle_event,
++ IRQF_SAMPLE_RANDOM, "pciback", pdev);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error binding event channel to IRQ");
++ goto out;
++ }
++ pdev->evtchn_irq = err;
++ err = 0;
++
++ dev_dbg(&pdev->xdev->dev, "Attached!\n");
++ out:
++ return err;
++}
++
++static int pciback_attach(struct pciback_device *pdev)
++{
++ int err = 0;
++ int gnt_ref, remote_evtchn;
++ char *magic = NULL;
++
++ spin_lock(&pdev->dev_lock);
++
++ /* Make sure we only do this setup once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitialised)
++ goto out;
++
++ /* Wait for frontend to state that it has published the configuration */
++ if (xenbus_read_driver_state(pdev->xdev->otherend) !=
++ XenbusStateInitialised)
++ goto out;
++
++ dev_dbg(&pdev->xdev->dev, "Reading frontend config\n");
++
++ err = xenbus_gather(XBT_NIL, pdev->xdev->otherend,
++ "pci-op-ref", "%u", &gnt_ref,
++ "event-channel", "%u", &remote_evtchn,
++ "magic", NULL, &magic, NULL);
++ if (err) {
++ /* If configuration didn't get read correctly, wait longer */
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading configuration from frontend");
++ goto out;
++ }
++
++ if (magic == NULL || strcmp(magic, XEN_PCI_MAGIC) != 0) {
++ xenbus_dev_fatal(pdev->xdev, -EFAULT,
++ "version mismatch (%s/%s) with pcifront - "
++ "halting pciback",
++ magic, XEN_PCI_MAGIC);
++ goto out;
++ }
++
++ err = pciback_do_attach(pdev, gnt_ref, remote_evtchn);
++ if (err)
++ goto out;
++
++ dev_dbg(&pdev->xdev->dev, "Connecting...\n");
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
++ if (err)
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to connected state!");
++
++ dev_dbg(&pdev->xdev->dev, "Connected? %d\n", err);
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ if (magic)
++ kfree(magic);
++
++ return err;
++}
++
++static int pciback_publish_pci_dev(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus,
++ unsigned int devfn, unsigned int devid)
++{
++ int err;
++ int len;
++ char str[64];
++
++ len = snprintf(str, sizeof(str), "vdev-%d", devid);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%04x:%02x:%02x.%02x", domain, bus,
++ PCI_SLOT(devfn), PCI_FUNC(devfn));
++
++ out:
++ return err;
++}
++
++static int pciback_export_device(struct pciback_device *pdev,
++ int domain, int bus, int slot, int func,
++ int devid)
++{
++ struct pci_dev *dev;
++ int err = 0;
++
++ dev_dbg(&pdev->xdev->dev, "exporting dom %x bus %x slot %x func %x\n",
++ domain, bus, slot, func);
++
++ dev = pcistub_get_pci_dev_by_slot(pdev, domain, bus, slot, func);
++ if (!dev) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Couldn't locate PCI device "
++ "(%04x:%02x:%02x.%01x)! "
++ "perhaps already in-use?",
++ domain, bus, slot, func);
++ goto out;
++ }
++
++ err = pciback_add_pci_dev(pdev, dev, devid, pciback_publish_pci_dev);
++ if (err)
++ goto out;
++
++ /* TODO: It'd be nice to export a bridge and have all of its children
++ * get exported with it. This may be best done in xend (which will
++ * have to calculate resource usage anyway) but we probably want to
++ * put something in here to ensure that if a bridge gets given to a
++ * driver domain, that all devices under that bridge are not given
++ * to other driver domains (as he who controls the bridge can disable
++ * it and stop the other devices from working).
++ */
++ out:
++ return err;
++}
++
++static int pciback_remove_device(struct pciback_device *pdev,
++ int domain, int bus, int slot, int func)
++{
++ int err = 0;
++ struct pci_dev *dev;
++
++ dev_dbg(&pdev->xdev->dev, "removing dom %x bus %x slot %x func %x\n",
++ domain, bus, slot, func);
++
++ dev = pciback_get_pci_dev(pdev, domain, bus, PCI_DEVFN(slot, func));
++ if (!dev) {
++ err = -EINVAL;
++ dev_dbg(&pdev->xdev->dev, "Couldn't locate PCI device "
++ "(%04x:%02x:%02x.%01x)! not owned by this domain\n",
++ domain, bus, slot, func);
++ goto out;
++ }
++
++ pciback_release_pci_dev(pdev, dev);
++
++ out:
++ return err;
++}
++
++static int pciback_publish_pci_root(struct pciback_device *pdev,
++ unsigned int domain, unsigned int bus)
++{
++ unsigned int d, b;
++ int i, root_num, len, err;
++ char str[64];
++
++ dev_dbg(&pdev->xdev->dev, "Publishing pci roots\n");
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", &root_num);
++ if (err == 0 || err == -ENOENT)
++ root_num = 0;
++ else if (err < 0)
++ goto out;
++
++ /* Verify that we haven't already published this pci root */
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ str, "%x:%x", &d, &b);
++ if (err < 0)
++ goto out;
++ if (err != 2) {
++ err = -EINVAL;
++ goto out;
++ }
++
++ if (d == domain && b == bus) {
++ err = 0;
++ goto out;
++ }
++ }
++
++ len = snprintf(str, sizeof(str), "root-%d", root_num);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ dev_dbg(&pdev->xdev->dev, "writing root %d at %04x:%02x\n",
++ root_num, domain, bus);
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, str,
++ "%04x:%02x", domain, bus);
++ if (err)
++ goto out;
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename,
++ "root_num", "%d", (root_num + 1));
++
++ out:
++ return err;
++}
++
++static int pciback_reconfigure(struct pciback_device *pdev)
++{
++ int err = 0;
++ int num_devs;
++ int domain, bus, slot, func;
++ int substate;
++ int i, len;
++ char state_str[64];
++ char dev_str[64];
++
++ spin_lock(&pdev->dev_lock);
++
++ dev_dbg(&pdev->xdev->dev, "Reconfiguring device ...\n");
++
++ /* Make sure we only reconfigure once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateReconfiguring)
++ goto out;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of devices");
++ goto out;
++ }
++
++ for (i = 0; i < num_devs; i++) {
++ len = snprintf(state_str, sizeof(state_str), "state-%d", i);
++ if (unlikely(len >= (sizeof(state_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, state_str,
++ "%d", &substate);
++ if (err != 1)
++ substate = XenbusStateUnknown;
++
++ switch (substate) {
++ case XenbusStateInitialising:
++ dev_dbg(&pdev->xdev->dev, "Attaching dev-%d ...\n", i);
++
++ len = snprintf(dev_str, sizeof(dev_str), "dev-%d", i);
++ if (unlikely(len >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while "
++ "reading configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ dev_str, "%x:%x:%x.%x",
++ &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device "
++ "configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
++
++ err = pciback_export_device(pdev, domain, bus, slot,
++ func, i);
++ if (err)
++ goto out;
++
++ /* Publish pci roots. */
++ err = pciback_publish_pci_roots(pdev, pciback_publish_pci_root);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error while publish PCI root"
++ "buses for frontend");
++ goto out;
++ }
++
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename,
++ state_str, "%d",
++ XenbusStateInitialised);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching substate of "
++ "dev-%d\n", i);
++ goto out;
++ }
++ break;
++
++ case XenbusStateClosing:
++ dev_dbg(&pdev->xdev->dev, "Detaching dev-%d ...\n", i);
++
++ len = snprintf(dev_str, sizeof(dev_str), "vdev-%d", i);
++ if (unlikely(len >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while "
++ "reading configuration");
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename,
++ dev_str, "%x:%x:%x.%x",
++ &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device "
++ "configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
++
++ err = pciback_remove_device(pdev, domain, bus, slot,
++ func);
++ if(err)
++ goto out;
++
++ /* TODO: If at some point we implement support for pci
++ * root hot-remove on pcifront side, we'll need to
++ * remove unnecessary xenstore nodes of pci roots here.
++ */
++
++ break;
++
++ default:
++ break;
++ }
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateReconfigured);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to reconfigured state!");
++ goto out;
++ }
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ return 0;
++}
++
++static void pciback_frontend_changed(struct xenbus_device *xdev,
++ enum xenbus_state fe_state)
++{
++ struct pciback_device *pdev = xdev->dev.driver_data;
++
++ dev_dbg(&xdev->dev, "fe state changed %d\n", fe_state);
++
++ switch (fe_state) {
++ case XenbusStateInitialised:
++ pciback_attach(pdev);
++ break;
++
++ case XenbusStateReconfiguring:
++ pciback_reconfigure(pdev);
++ break;
++
++ case XenbusStateConnected:
++ /* pcifront switched its state from reconfiguring to connected.
++ * Then switch to connected state.
++ */
++ xenbus_switch_state(xdev, XenbusStateConnected);
++ break;
++
++ case XenbusStateClosing:
++ pciback_disconnect(pdev);
++ xenbus_switch_state(xdev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ pciback_disconnect(pdev);
++ xenbus_switch_state(xdev, XenbusStateClosed);
++ if (xenbus_dev_is_online(xdev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ dev_dbg(&xdev->dev, "frontend is gone! unregister device\n");
++ device_unregister(&xdev->dev);
++ break;
++
++ default:
++ break;
++ }
++}
++
++static int pciback_setup_backend(struct pciback_device *pdev)
++{
++ /* Get configuration from xend (if available now) */
++ int domain, bus, slot, func;
++ int err = 0;
++ int i, num_devs;
++ char dev_str[64];
++ char state_str[64];
++
++ spin_lock(&pdev->dev_lock);
++
++ /* It's possible we could get the call to setup twice, so make sure
++ * we're not already connected.
++ */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitWait)
++ goto out;
++
++ dev_dbg(&pdev->xdev->dev, "getting be setup\n");
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of devices");
++ goto out;
++ }
++
++ for (i = 0; i < num_devs; i++) {
++ int l = snprintf(dev_str, sizeof(dev_str), "dev-%d", i);
++ if (unlikely(l >= (sizeof(dev_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->nodename, dev_str,
++ "%x:%x:%x.%x", &domain, &bus, &slot, &func);
++ if (err < 0) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading device configuration");
++ goto out;
++ }
++ if (err != 4) {
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error parsing pci device "
++ "configuration");
++ goto out;
++ }
++
++ err = pciback_export_device(pdev, domain, bus, slot, func, i);
++ if (err)
++ goto out;
++
++ /* Switch substate of this device. */
++ l = snprintf(state_str, sizeof(state_str), "state-%d", i);
++ if (unlikely(l >= (sizeof(state_str) - 1))) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "String overflow while reading "
++ "configuration");
++ goto out;
++ }
++ err = xenbus_printf(XBT_NIL, pdev->xdev->nodename, state_str,
++ "%d", XenbusStateInitialised);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err, "Error switching "
++ "substate of dev-%d\n", i);
++ goto out;
++ }
++ }
++
++ err = pciback_publish_pci_roots(pdev, pciback_publish_pci_root);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error while publish PCI root buses "
++ "for frontend");
++ goto out;
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
++ if (err)
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error switching to initialised state!");
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ if (!err)
++ /* see if pcifront is already configured (if not, we'll wait) */
++ pciback_attach(pdev);
++
++ return err;
++}
++
++static void pciback_be_watch(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ struct pciback_device *pdev =
++ container_of(watch, struct pciback_device, be_watch);
++
++ switch (xenbus_read_driver_state(pdev->xdev->nodename)) {
++ case XenbusStateInitWait:
++ pciback_setup_backend(pdev);
++ break;
++
++ default:
++ break;
++ }
++}
++
++static int pciback_xenbus_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err = 0;
++ struct pciback_device *pdev = alloc_pdev(dev);
++
++ if (pdev == NULL) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(dev, err,
++ "Error allocating pciback_device struct");
++ goto out;
++ }
++
++ /* wait for xend to configure us */
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto out;
++
++ /* watch the backend node for backend configuration information */
++ err = xenbus_watch_path(dev, dev->nodename, &pdev->be_watch,
++ pciback_be_watch);
++ if (err)
++ goto out;
++ pdev->be_watching = 1;
++
++ /* We need to force a call to our callback here in case
++ * xend already configured us!
++ */
++ pciback_be_watch(&pdev->be_watch, NULL, 0);
++
++ out:
++ return err;
++}
++
++static int pciback_xenbus_remove(struct xenbus_device *dev)
++{
++ struct pciback_device *pdev = dev->dev.driver_data;
++
++ if (pdev != NULL)
++ free_pdev(pdev);
++
++ return 0;
++}
++
++static const struct xenbus_device_id xenpci_ids[] = {
++ {"pci"},
++ {{0}},
++};
++
++static struct xenbus_driver xenbus_pciback_driver = {
++ .name = "pciback",
++ .ids = xenpci_ids,
++ .probe = pciback_xenbus_probe,
++ .remove = pciback_xenbus_remove,
++ .otherend_changed = pciback_frontend_changed,
++};
++
++int __init pciback_xenbus_register(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ return xenbus_register_backend(&xenbus_pciback_driver);
++}
++
++void __exit pciback_xenbus_unregister(void)
++{
++ xenbus_unregister_driver(&xenbus_pciback_driver);
++}
+diff --git a/drivers/xen/pcifront/Makefile b/drivers/xen/pcifront/Makefile
+new file mode 100644
+index 0000000..4ceb18a
+--- /dev/null
++++ b/drivers/xen/pcifront/Makefile
+@@ -0,0 +1,5 @@
++obj-y += pcifront.o
++
++pcifront-y := pci_op.o xenbus.o pci.o
++
++ccflags-$(CONFIG_XEN_PCIDEV_FE_DEBUG) += -DDEBUG
+diff --git a/drivers/xen/pcifront/pci.c b/drivers/xen/pcifront/pci.c
+new file mode 100644
+index 0000000..4239f00
+--- /dev/null
++++ b/drivers/xen/pcifront/pci.c
+@@ -0,0 +1,46 @@
++/*
++ * PCI Frontend Operations - ensure only one PCI frontend runs at a time
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include "pcifront.h"
++
++DEFINE_SPINLOCK(pcifront_dev_lock);
++static struct pcifront_device *pcifront_dev = NULL;
++
++int pcifront_connect(struct pcifront_device *pdev)
++{
++ int err = 0;
++
++ spin_lock(&pcifront_dev_lock);
++
++ if (!pcifront_dev) {
++ dev_info(&pdev->xdev->dev, "Installing PCI frontend\n");
++ pcifront_dev = pdev;
++ }
++ else {
++ dev_err(&pdev->xdev->dev, "PCI frontend already installed!\n");
++ err = -EEXIST;
++ }
++
++ spin_unlock(&pcifront_dev_lock);
++
++ return err;
++}
++
++void pcifront_disconnect(struct pcifront_device *pdev)
++{
++ spin_lock(&pcifront_dev_lock);
++
++ if (pdev == pcifront_dev) {
++ dev_info(&pdev->xdev->dev,
++ "Disconnecting PCI Frontend Buses\n");
++ pcifront_dev = NULL;
++ }
++
++ spin_unlock(&pcifront_dev_lock);
++}
+diff --git a/drivers/xen/pcifront/pci_op.c b/drivers/xen/pcifront/pci_op.c
+new file mode 100644
+index 0000000..136facc
+--- /dev/null
++++ b/drivers/xen/pcifront/pci_op.c
+@@ -0,0 +1,557 @@
++/*
++ * PCI Frontend Operations - Communicates with frontend
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/version.h>
++#include <linux/init.h>
++#include <linux/pci.h>
++#include <linux/spinlock.h>
++#include <linux/time.h>
++#include <xen/evtchn.h>
++#include "pcifront.h"
++
++static int verbose_request = 0;
++module_param(verbose_request, int, 0644);
++
++#ifdef __ia64__
++static void pcifront_init_sd(struct pcifront_sd *sd,
++ unsigned int domain, unsigned int bus,
++ struct pcifront_device *pdev)
++{
++ int err, i, j, k, len, root_num, res_count;
++ struct acpi_resource res;
++ unsigned int d, b, byte;
++ unsigned long magic;
++ char str[64], tmp[3];
++ unsigned char *buf, *bufp;
++ u8 *ptr;
++
++ memset(sd, 0, sizeof(*sd));
++
++ sd->segment = domain;
++ sd->node = -1; /* Revisit for NUMA */
++ sd->platform_data = pdev;
++
++ /* Look for resources for this controller in xenbus. */
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "root_num",
++ "%d", &root_num);
++ if (err != 1)
++ return;
++
++ for (i = 0; i < root_num; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%x:%x", &d, &b);
++ if (err != 2)
++ return;
++
++ if (d == domain && b == bus)
++ break;
++ }
++
++ if (i == root_num)
++ return;
++
++ len = snprintf(str, sizeof(str), "root-resource-magic");
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%lx", &magic);
++
++ if (err != 1)
++ return; /* No resources, nothing to do */
++
++ if (magic != (sizeof(res) * 2) + 1) {
++ printk(KERN_WARNING "pcifront: resource magic mismatch\n");
++ return;
++ }
++
++ len = snprintf(str, sizeof(str), "root-%d-resources", i);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ str, "%d", &res_count);
++
++ if (err != 1)
++ return; /* No resources, nothing to do */
++
++ sd->window = kzalloc(sizeof(*sd->window) * res_count, GFP_KERNEL);
++ if (!sd->window)
++ return;
++
++ /* magic is also the size of the byte stream in xenbus */
++ buf = kmalloc(magic, GFP_KERNEL);
++ if (!buf) {
++ kfree(sd->window);
++ sd->window = NULL;
++ return;
++ }
++
++ /* Read the resources out of xenbus */
++ for (j = 0; j < res_count; j++) {
++ memset(&res, 0, sizeof(res));
++ memset(buf, 0, magic);
++
++ len = snprintf(str, sizeof(str), "root-%d-resource-%d", i, j);
++ if (unlikely(len >= (sizeof(str) - 1)))
++ return;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%s", buf);
++ if (err != 1) {
++ printk(KERN_WARNING "pcifront: error reading "
++ "resource %d on bus %04x:%02x\n",
++ j, domain, bus);
++ continue;
++ }
++
++ bufp = buf;
++ ptr = (u8 *)&res;
++ memset(tmp, 0, sizeof(tmp));
++
++ /* Copy ASCII byte stream into structure */
++ for (k = 0; k < magic - 1; k += 2) {
++ memcpy(tmp, bufp, 2);
++ bufp += 2;
++
++ sscanf(tmp, "%02x", &byte);
++ *ptr = byte;
++ ptr++;
++ }
++
++ xen_add_resource(sd, domain, bus, &res);
++ sd->windows++;
++ }
++ kfree(buf);
++}
++#endif
++
++static int errno_to_pcibios_err(int errno)
++{
++ switch (errno) {
++ case XEN_PCI_ERR_success:
++ return PCIBIOS_SUCCESSFUL;
++
++ case XEN_PCI_ERR_dev_not_found:
++ return PCIBIOS_DEVICE_NOT_FOUND;
++
++ case XEN_PCI_ERR_invalid_offset:
++ case XEN_PCI_ERR_op_failed:
++ return PCIBIOS_BAD_REGISTER_NUMBER;
++
++ case XEN_PCI_ERR_not_implemented:
++ return PCIBIOS_FUNC_NOT_SUPPORTED;
++
++ case XEN_PCI_ERR_access_denied:
++ return PCIBIOS_SET_FAILED;
++ }
++ return errno;
++}
++
++static int do_pci_op(struct pcifront_device *pdev, struct xen_pci_op *op)
++{
++ int err = 0;
++ struct xen_pci_op *active_op = &pdev->sh_info->op;
++ unsigned long irq_flags;
++ evtchn_port_t port = pdev->evtchn;
++ s64 ns, ns_timeout;
++ struct timeval tv;
++
++ spin_lock_irqsave(&pdev->sh_info_lock, irq_flags);
++
++ memcpy(active_op, op, sizeof(struct xen_pci_op));
++
++ /* Go */
++ wmb();
++ set_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
++ notify_remote_via_evtchn(port);
++
++ /*
++ * We set a poll timeout of 3 seconds but give up on return after
++ * 2 seconds. It is better to time out too late rather than too early
++ * (in the latter case we end up continually re-executing poll() with a
++ * timeout in the past). 1s difference gives plenty of slack for error.
++ */
++ do_gettimeofday(&tv);
++ ns_timeout = timeval_to_ns(&tv) + 2 * (s64)NSEC_PER_SEC;
++
++ clear_evtchn(port);
++
++ while (test_bit(_XEN_PCIF_active,
++ (unsigned long *)&pdev->sh_info->flags)) {
++ if (HYPERVISOR_poll(&port, 1, jiffies + 3*HZ))
++ BUG();
++ clear_evtchn(port);
++ do_gettimeofday(&tv);
++ ns = timeval_to_ns(&tv);
++ if (ns > ns_timeout) {
++ dev_err(&pdev->xdev->dev,
++ "pciback not responding!!!\n");
++ clear_bit(_XEN_PCIF_active,
++ (unsigned long *)&pdev->sh_info->flags);
++ err = XEN_PCI_ERR_dev_not_found;
++ goto out;
++ }
++ }
++
++ memcpy(op, active_op, sizeof(struct xen_pci_op));
++
++ err = op->err;
++ out:
++ spin_unlock_irqrestore(&pdev->sh_info_lock, irq_flags);
++ return err;
++}
++
++/* Access to this function is spinlocked in drivers/pci/access.c */
++static int pcifront_bus_read(struct pci_bus *bus, unsigned int devfn,
++ int where, int size, u32 * val)
++{
++ int err = 0;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_conf_read,
++ .domain = pci_domain_nr(bus),
++ .bus = bus->number,
++ .devfn = devfn,
++ .offset = where,
++ .size = size,
++ };
++ struct pcifront_sd *sd = bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev,
++ "read dev=%04x:%02x:%02x.%01x - offset %x size %d\n",
++ pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
++ PCI_FUNC(devfn), where, size);
++
++ err = do_pci_op(pdev, &op);
++
++ if (likely(!err)) {
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev, "read got back value %x\n",
++ op.value);
++
++ *val = op.value;
++ } else if (err == -ENODEV) {
++ /* No device here, pretend that it just returned 0 */
++ err = 0;
++ *val = 0;
++ }
++
++ return errno_to_pcibios_err(err);
++}
++
++/* Access to this function is spinlocked in drivers/pci/access.c */
++static int pcifront_bus_write(struct pci_bus *bus, unsigned int devfn,
++ int where, int size, u32 val)
++{
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_conf_write,
++ .domain = pci_domain_nr(bus),
++ .bus = bus->number,
++ .devfn = devfn,
++ .offset = where,
++ .size = size,
++ .value = val,
++ };
++ struct pcifront_sd *sd = bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ if (verbose_request)
++ dev_info(&pdev->xdev->dev,
++ "write dev=%04x:%02x:%02x.%01x - "
++ "offset %x size %d val %x\n",
++ pci_domain_nr(bus), bus->number,
++ PCI_SLOT(devfn), PCI_FUNC(devfn), where, size, val);
++
++ return errno_to_pcibios_err(do_pci_op(pdev, &op));
++}
++
++struct pci_ops pcifront_bus_ops = {
++ .read = pcifront_bus_read,
++ .write = pcifront_bus_write,
++};
++
++#ifdef CONFIG_PCI_MSI
++int pci_frontend_enable_msix(struct pci_dev *dev,
++ struct msix_entry *entries,
++ int nvec)
++{
++ int err;
++ int i;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_enable_msix,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ .value = nvec,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ if (nvec > SH_INFO_MAX_VEC) {
++ printk("too much vector for pci frontend%x\n", nvec);
++ return -EINVAL;
++ }
++
++ for (i = 0; i < nvec; i++) {
++ op.msix_entries[i].entry = entries[i].entry;
++ op.msix_entries[i].vector = entries[i].vector;
++ }
++
++ err = do_pci_op(pdev, &op);
++
++ if (!err) {
++ if (!op.value) {
++ /* we get the result */
++ for ( i = 0; i < nvec; i++)
++ entries[i].vector = op.msix_entries[i].vector;
++ return 0;
++ }
++ else {
++ printk("enable msix get value %x\n", op.value);
++ return op.value;
++ }
++ }
++ else {
++ printk("enable msix get err %x\n", err);
++ return err;
++ }
++}
++
++void pci_frontend_disable_msix(struct pci_dev* dev)
++{
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_disable_msix,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ err = do_pci_op(pdev, &op);
++
++ /* What should do for error ? */
++ if (err)
++ printk("pci_disable_msix get err %x\n", err);
++}
++
++int pci_frontend_enable_msi(struct pci_dev *dev)
++{
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_enable_msi,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ err = do_pci_op(pdev, &op);
++ if (likely(!err)) {
++ dev->irq = op.value;
++ }
++ else {
++ printk("pci frontend enable msi failed for dev %x:%x \n",
++ op.bus, op.devfn);
++ err = -EINVAL;
++ }
++ return err;
++}
++
++void pci_frontend_disable_msi(struct pci_dev* dev)
++{
++ int err;
++ struct xen_pci_op op = {
++ .cmd = XEN_PCI_OP_disable_msi,
++ .domain = pci_domain_nr(dev->bus),
++ .bus = dev->bus->number,
++ .devfn = dev->devfn,
++ };
++ struct pcifront_sd *sd = dev->bus->sysdata;
++ struct pcifront_device *pdev = pcifront_get_pdev(sd);
++
++ err = do_pci_op(pdev, &op);
++ if (err == XEN_PCI_ERR_dev_not_found) {
++ /* XXX No response from backend, what shall we do? */
++ printk("get no response from backend for disable MSI\n");
++ return;
++ }
++ if (likely(!err))
++ dev->irq = op.value;
++ else
++ /* how can pciback notify us fail? */
++ printk("get fake response frombackend \n");
++}
++#endif /* CONFIG_PCI_MSI */
++
++/* Claim resources for the PCI frontend as-is, backend won't allow changes */
++static void pcifront_claim_resource(struct pci_dev *dev, void *data)
++{
++ struct pcifront_device *pdev = data;
++ int i;
++ struct resource *r;
++
++ for (i = 0; i < PCI_NUM_RESOURCES; i++) {
++ r = &dev->resource[i];
++
++ if (!r->parent && r->start && r->flags) {
++ dev_dbg(&pdev->xdev->dev, "claiming resource %s/%d\n",
++ pci_name(dev), i);
++ pci_claim_resource(dev, i);
++ }
++ }
++}
++
++int __devinit pcifront_scan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus)
++{
++ struct pci_bus *b;
++ struct pcifront_sd *sd = NULL;
++ struct pci_bus_entry *bus_entry = NULL;
++ int err = 0;
++
++#ifndef CONFIG_PCI_DOMAINS
++ if (domain != 0) {
++ dev_err(&pdev->xdev->dev,
++ "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
++ dev_err(&pdev->xdev->dev,
++ "Please compile with CONFIG_PCI_DOMAINS\n");
++ err = -EINVAL;
++ goto err_out;
++ }
++#endif
++
++ dev_info(&pdev->xdev->dev, "Creating PCI Frontend Bus %04x:%02x\n",
++ domain, bus);
++
++ bus_entry = kmalloc(sizeof(*bus_entry), GFP_KERNEL);
++ sd = kmalloc(sizeof(*sd), GFP_KERNEL);
++ if (!bus_entry || !sd) {
++ err = -ENOMEM;
++ goto err_out;
++ }
++ pcifront_init_sd(sd, domain, bus, pdev);
++
++ b = pci_scan_bus_parented(&pdev->xdev->dev, bus,
++ &pcifront_bus_ops, sd);
++ if (!b) {
++ dev_err(&pdev->xdev->dev,
++ "Error creating PCI Frontend Bus!\n");
++ err = -ENOMEM;
++ goto err_out;
++ }
++
++ pcifront_setup_root_resources(b, sd);
++ bus_entry->bus = b;
++
++ list_add(&bus_entry->list, &pdev->root_buses);
++
++ /* Claim resources before going "live" with our devices */
++ pci_walk_bus(b, pcifront_claim_resource, pdev);
++
++ pci_bus_add_devices(b);
++
++ return 0;
++
++ err_out:
++ kfree(bus_entry);
++ kfree(sd);
++
++ return err;
++}
++
++int __devinit pcifront_rescan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus)
++{
++ struct pci_bus *b;
++ struct pci_dev *d;
++ unsigned int devfn;
++
++#ifndef CONFIG_PCI_DOMAINS
++ if (domain != 0) {
++ dev_err(&pdev->xdev->dev,
++ "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
++ dev_err(&pdev->xdev->dev,
++ "Please compile with CONFIG_PCI_DOMAINS\n");
++ return -EINVAL;
++ }
++#endif
++
++ dev_info(&pdev->xdev->dev, "Rescanning PCI Frontend Bus %04x:%02x\n",
++ domain, bus);
++
++ b = pci_find_bus(domain, bus);
++ if(!b)
++ /* If the bus is unknown, create it. */
++ return pcifront_scan_root(pdev, domain, bus);
++
++ /* Rescan the bus for newly attached functions and add.
++ * We omit handling of PCI bridge attachment because pciback prevents
++ * bridges from being exported.
++ */
++ for (devfn = 0; devfn < 0x100; devfn++) {
++ d = pci_get_slot(b, devfn);
++ if(d) {
++ /* Device is already known. */
++ pci_dev_put(d);
++ continue;
++ }
++
++ d = pci_scan_single_device(b, devfn);
++ if (d) {
++ int err;
++
++ dev_info(&pdev->xdev->dev, "New device on "
++ "%04x:%02x:%02x.%02x found.\n", domain, bus,
++ PCI_SLOT(devfn), PCI_FUNC(devfn));
++ err = pci_bus_add_device(d);
++ if (err)
++ dev_err(&pdev->xdev->dev,
++ "error %d adding device, continuing.\n",
++ err);
++ }
++ }
++
++ return 0;
++}
++
++static void free_root_bus_devs(struct pci_bus *bus)
++{
++ struct pci_dev *dev;
++
++ while (!list_empty(&bus->devices)) {
++ dev = container_of(bus->devices.next, struct pci_dev,
++ bus_list);
++ dev_dbg(&dev->dev, "removing device\n");
++ pci_remove_bus_device(dev);
++ }
++}
++
++void pcifront_free_roots(struct pcifront_device *pdev)
++{
++ struct pci_bus_entry *bus_entry, *t;
++
++ dev_dbg(&pdev->xdev->dev, "cleaning up root buses\n");
++
++ list_for_each_entry_safe(bus_entry, t, &pdev->root_buses, list) {
++ list_del(&bus_entry->list);
++
++ free_root_bus_devs(bus_entry->bus);
++
++ kfree(bus_entry->bus->sysdata);
++
++ device_unregister(bus_entry->bus->bridge);
++ pci_remove_bus(bus_entry->bus);
++
++ kfree(bus_entry);
++ }
++}
+diff --git a/drivers/xen/pcifront/pcifront.h b/drivers/xen/pcifront/pcifront.h
+new file mode 100644
+index 0000000..0ebac13
+--- /dev/null
++++ b/drivers/xen/pcifront/pcifront.h
+@@ -0,0 +1,42 @@
++/*
++ * PCI Frontend - Common data structures & function declarations
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_PCIFRONT_H__
++#define __XEN_PCIFRONT_H__
++
++#include <linux/spinlock.h>
++#include <linux/pci.h>
++#include <xen/xenbus.h>
++#include <xen/interface/io/pciif.h>
++#include <xen/pcifront.h>
++
++struct pci_bus_entry {
++ struct list_head list;
++ struct pci_bus *bus;
++};
++
++struct pcifront_device {
++ struct xenbus_device *xdev;
++ struct list_head root_buses;
++ spinlock_t dev_lock;
++
++ int evtchn;
++ int gnt_ref;
++
++ /* Lock this when doing any operations in sh_info */
++ spinlock_t sh_info_lock;
++ struct xen_pci_sharedinfo *sh_info;
++};
++
++int pcifront_connect(struct pcifront_device *pdev);
++void pcifront_disconnect(struct pcifront_device *pdev);
++
++int pcifront_scan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus);
++int pcifront_rescan_root(struct pcifront_device *pdev,
++ unsigned int domain, unsigned int bus);
++void pcifront_free_roots(struct pcifront_device *pdev);
++
++#endif /* __XEN_PCIFRONT_H__ */
+diff --git a/drivers/xen/pcifront/xenbus.c b/drivers/xen/pcifront/xenbus.c
+new file mode 100644
+index 0000000..659b07b
+--- /dev/null
++++ b/drivers/xen/pcifront/xenbus.c
+@@ -0,0 +1,450 @@
++/*
++ * PCI Frontend Xenbus Setup - handles setup with backend (imports page/evtchn)
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/mm.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include "pcifront.h"
++
++#define INVALID_GRANT_REF (0)
++#define INVALID_EVTCHN (-1)
++
++static struct pcifront_device *alloc_pdev(struct xenbus_device *xdev)
++{
++ struct pcifront_device *pdev;
++
++ pdev = kzalloc(sizeof(struct pcifront_device), GFP_KERNEL);
++ if (pdev == NULL)
++ goto out;
++
++ pdev->sh_info =
++ (struct xen_pci_sharedinfo *)__get_free_page(GFP_KERNEL);
++ if (pdev->sh_info == NULL) {
++ kfree(pdev);
++ pdev = NULL;
++ goto out;
++ }
++ pdev->sh_info->flags = 0;
++
++ xdev->dev.driver_data = pdev;
++ pdev->xdev = xdev;
++
++ INIT_LIST_HEAD(&pdev->root_buses);
++
++ spin_lock_init(&pdev->dev_lock);
++ spin_lock_init(&pdev->sh_info_lock);
++
++ pdev->evtchn = INVALID_EVTCHN;
++ pdev->gnt_ref = INVALID_GRANT_REF;
++
++ dev_dbg(&xdev->dev, "Allocated pdev @ 0x%p pdev->sh_info @ 0x%p\n",
++ pdev, pdev->sh_info);
++ out:
++ return pdev;
++}
++
++static void free_pdev(struct pcifront_device *pdev)
++{
++ dev_dbg(&pdev->xdev->dev, "freeing pdev @ 0x%p\n", pdev);
++
++ pcifront_free_roots(pdev);
++
++ if (pdev->evtchn != INVALID_EVTCHN)
++ xenbus_free_evtchn(pdev->xdev, pdev->evtchn);
++
++ if (pdev->gnt_ref != INVALID_GRANT_REF)
++ gnttab_end_foreign_access(pdev->gnt_ref,
++ (unsigned long)pdev->sh_info);
++
++ pdev->xdev->dev.driver_data = NULL;
++
++ kfree(pdev);
++}
++
++static int pcifront_publish_info(struct pcifront_device *pdev)
++{
++ int err = 0;
++ struct xenbus_transaction trans;
++
++ err = xenbus_grant_ring(pdev->xdev, virt_to_mfn(pdev->sh_info));
++ if (err < 0)
++ goto out;
++
++ pdev->gnt_ref = err;
++
++ err = xenbus_alloc_evtchn(pdev->xdev, &pdev->evtchn);
++ if (err)
++ goto out;
++
++ do_publish:
++ err = xenbus_transaction_start(&trans);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error writing configuration for backend "
++ "(start transaction)");
++ goto out;
++ }
++
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "pci-op-ref", "%u", pdev->gnt_ref);
++ if (!err)
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "event-channel", "%u", pdev->evtchn);
++ if (!err)
++ err = xenbus_printf(trans, pdev->xdev->nodename,
++ "magic", XEN_PCI_MAGIC);
++
++ if (err) {
++ xenbus_transaction_end(trans, 1);
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error writing configuration for backend");
++ goto out;
++ } else {
++ err = xenbus_transaction_end(trans, 0);
++ if (err == -EAGAIN)
++ goto do_publish;
++ else if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error completing transaction "
++ "for backend");
++ goto out;
++ }
++ }
++
++ xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
++
++ dev_dbg(&pdev->xdev->dev, "publishing successful!\n");
++
++ out:
++ return err;
++}
++
++static int __devinit pcifront_try_connect(struct pcifront_device *pdev)
++{
++ int err = -EFAULT;
++ int i, num_roots, len;
++ char str[64];
++ unsigned int domain, bus;
++
++ spin_lock(&pdev->dev_lock);
++
++ /* Only connect once */
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateInitialised)
++ goto out;
++
++ err = pcifront_connect(pdev);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error connecting PCI Frontend");
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ "root_num", "%d", &num_roots);
++ if (err == -ENOENT) {
++ xenbus_dev_error(pdev->xdev, err,
++ "No PCI Roots found, trying 0000:00");
++ err = pcifront_scan_root(pdev, 0, 0);
++ num_roots = 0;
++ } else if (err != 1) {
++ if (err == 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI roots");
++ goto out;
++ }
++
++ for (i = 0; i < num_roots; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x", &domain, &bus);
++ if (err != 2) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI root %d", i);
++ goto out;
++ }
++
++ err = pcifront_scan_root(pdev, domain, bus);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error scanning PCI root %04x:%02x",
++ domain, bus);
++ goto out;
++ }
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
++ if (err)
++ goto out;
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++ return err;
++}
++
++static int pcifront_try_disconnect(struct pcifront_device *pdev)
++{
++ int err = 0;
++ enum xenbus_state prev_state;
++
++ spin_lock(&pdev->dev_lock);
++
++ prev_state = xenbus_read_driver_state(pdev->xdev->nodename);
++
++ if (prev_state >= XenbusStateClosing)
++ goto out;
++
++ if(prev_state == XenbusStateConnected) {
++ pcifront_free_roots(pdev);
++ pcifront_disconnect(pdev);
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateClosed);
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++
++ return err;
++}
++
++static int __devinit pcifront_attach_devices(struct pcifront_device *pdev)
++{
++ int err = -EFAULT;
++ int i, num_roots, len;
++ unsigned int domain, bus;
++ char str[64];
++
++ spin_lock(&pdev->dev_lock);
++
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateReconfiguring)
++ goto out;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
++ "root_num", "%d", &num_roots);
++ if (err == -ENOENT) {
++ xenbus_dev_error(pdev->xdev, err,
++ "No PCI Roots found, trying 0000:00");
++ err = pcifront_rescan_root(pdev, 0, 0);
++ num_roots = 0;
++ } else if (err != 1) {
++ if (err == 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI roots");
++ goto out;
++ }
++
++ for (i = 0; i < num_roots; i++) {
++ len = snprintf(str, sizeof(str), "root-%d", i);
++ if (unlikely(len >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x", &domain, &bus);
++ if (err != 2) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI root %d", i);
++ goto out;
++ }
++
++ err = pcifront_rescan_root(pdev, domain, bus);
++ if (err) {
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error scanning PCI root %04x:%02x",
++ domain, bus);
++ goto out;
++ }
++ }
++
++ xenbus_switch_state(pdev->xdev, XenbusStateConnected);
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++ return err;
++}
++
++static int pcifront_detach_devices(struct pcifront_device *pdev)
++{
++ int err = 0;
++ int i, num_devs;
++ unsigned int domain, bus, slot, func;
++ struct pci_bus *pci_bus;
++ struct pci_dev *pci_dev;
++ char str[64];
++
++ spin_lock(&pdev->dev_lock);
++
++ if (xenbus_read_driver_state(pdev->xdev->nodename) !=
++ XenbusStateConnected)
++ goto out;
++
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "num_devs", "%d",
++ &num_devs);
++ if (err != 1) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading number of PCI devices");
++ goto out;
++ }
++
++ /* Find devices being detached and remove them. */
++ for (i = 0; i < num_devs; i++) {
++ int l, state;
++ l = snprintf(str, sizeof(str), "state-%d", i);
++ if (unlikely(l >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str, "%d",
++ &state);
++ if (err != 1)
++ state = XenbusStateUnknown;
++
++ if (state != XenbusStateClosing)
++ continue;
++
++ /* Remove device. */
++ l = snprintf(str, sizeof(str), "vdev-%d", i);
++ if (unlikely(l >= (sizeof(str) - 1))) {
++ err = -ENOMEM;
++ goto out;
++ }
++ err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
++ "%x:%x:%x.%x", &domain, &bus, &slot, &func);
++ if (err != 4) {
++ if (err >= 0)
++ err = -EINVAL;
++ xenbus_dev_fatal(pdev->xdev, err,
++ "Error reading PCI device %d", i);
++ goto out;
++ }
++
++ pci_bus = pci_find_bus(domain, bus);
++ if(!pci_bus) {
++ dev_dbg(&pdev->xdev->dev, "Cannot get bus %04x:%02x\n",
++ domain, bus);
++ continue;
++ }
++ pci_dev = pci_get_slot(pci_bus, PCI_DEVFN(slot, func));
++ if(!pci_dev) {
++ dev_dbg(&pdev->xdev->dev,
++ "Cannot get PCI device %04x:%02x:%02x.%02x\n",
++ domain, bus, slot, func);
++ continue;
++ }
++ pci_remove_bus_device(pci_dev);
++ pci_dev_put(pci_dev);
++
++ dev_dbg(&pdev->xdev->dev,
++ "PCI device %04x:%02x:%02x.%02x removed.\n",
++ domain, bus, slot, func);
++ }
++
++ err = xenbus_switch_state(pdev->xdev, XenbusStateReconfiguring);
++
++ out:
++ spin_unlock(&pdev->dev_lock);
++ return err;
++}
++
++static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
++ enum xenbus_state be_state)
++{
++ struct pcifront_device *pdev = xdev->dev.driver_data;
++
++ switch (be_state) {
++ case XenbusStateUnknown:
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateConnected:
++ pcifront_try_connect(pdev);
++ break;
++
++ case XenbusStateClosing:
++ dev_warn(&xdev->dev, "backend going away!\n");
++ pcifront_try_disconnect(pdev);
++ break;
++
++ case XenbusStateReconfiguring:
++ pcifront_detach_devices(pdev);
++ break;
++
++ case XenbusStateReconfigured:
++ pcifront_attach_devices(pdev);
++ break;
++ }
++}
++
++static int pcifront_xenbus_probe(struct xenbus_device *xdev,
++ const struct xenbus_device_id *id)
++{
++ int err = 0;
++ struct pcifront_device *pdev = alloc_pdev(xdev);
++
++ if (pdev == NULL) {
++ err = -ENOMEM;
++ xenbus_dev_fatal(xdev, err,
++ "Error allocating pcifront_device struct");
++ goto out;
++ }
++
++ err = pcifront_publish_info(pdev);
++
++ out:
++ return err;
++}
++
++static int pcifront_xenbus_remove(struct xenbus_device *xdev)
++{
++ if (xdev->dev.driver_data)
++ free_pdev(xdev->dev.driver_data);
++
++ return 0;
++}
++
++static const struct xenbus_device_id xenpci_ids[] = {
++ {"pci"},
++ {{0}},
++};
++MODULE_ALIAS("xen:pci");
++
++static struct xenbus_driver xenbus_pcifront_driver = {
++ .name = "pcifront",
++ .ids = xenpci_ids,
++ .probe = pcifront_xenbus_probe,
++ .remove = pcifront_xenbus_remove,
++ .otherend_changed = pcifront_backend_changed,
++};
++
++static int __init pcifront_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ return xenbus_register_frontend(&xenbus_pcifront_driver);
++}
++
++/* Initialize after the Xen PCI Frontend Stub is initialized */
++subsys_initcall(pcifront_init);
+diff --git a/drivers/xen/privcmd/Makefile b/drivers/xen/privcmd/Makefile
+new file mode 100644
+index 0000000..a90321c
+--- /dev/null
++++ b/drivers/xen/privcmd/Makefile
+@@ -0,0 +1,3 @@
++
++obj-y += privcmd.o
++obj-$(CONFIG_COMPAT) += compat_privcmd.o
+diff --git a/drivers/xen/privcmd/compat_privcmd.c b/drivers/xen/privcmd/compat_privcmd.c
+new file mode 100644
+index 0000000..8022eec
+--- /dev/null
++++ b/drivers/xen/privcmd/compat_privcmd.c
+@@ -0,0 +1,72 @@
++/*
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ * Copyright (C) IBM Corp. 2006
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ */
++
++#include <linux/compat.h>
++#include <linux/ioctl.h>
++#include <linux/syscalls.h>
++#include <asm/hypervisor.h>
++#include <asm/uaccess.h>
++#include <xen/public/privcmd.h>
++#include <xen/compat_ioctl.h>
++
++int privcmd_ioctl_32(int fd, unsigned int cmd, unsigned long arg)
++{
++ int ret;
++
++ switch (cmd) {
++ case IOCTL_PRIVCMD_MMAP_32: {
++ struct privcmd_mmap *p;
++ struct privcmd_mmap_32 *p32;
++ struct privcmd_mmap_32 n32;
++
++ p32 = compat_ptr(arg);
++ p = compat_alloc_user_space(sizeof(*p));
++ if (copy_from_user(&n32, p32, sizeof(n32)) ||
++ put_user(n32.num, &p->num) ||
++ put_user(n32.dom, &p->dom) ||
++ put_user(compat_ptr(n32.entry), &p->entry))
++ return -EFAULT;
++
++ ret = sys_ioctl(fd, IOCTL_PRIVCMD_MMAP, (unsigned long)p);
++ }
++ break;
++ case IOCTL_PRIVCMD_MMAPBATCH_32: {
++ struct privcmd_mmapbatch *p;
++ struct privcmd_mmapbatch_32 *p32;
++ struct privcmd_mmapbatch_32 n32;
++
++ p32 = compat_ptr(arg);
++ p = compat_alloc_user_space(sizeof(*p));
++ if (copy_from_user(&n32, p32, sizeof(n32)) ||
++ put_user(n32.num, &p->num) ||
++ put_user(n32.dom, &p->dom) ||
++ put_user(n32.addr, &p->addr) ||
++ put_user(compat_ptr(n32.arr), &p->arr))
++ return -EFAULT;
++
++ ret = sys_ioctl(fd, IOCTL_PRIVCMD_MMAPBATCH, (unsigned long)p);
++ }
++ break;
++ default:
++ ret = -EINVAL;
++ break;
++ }
++ return ret;
++}
+diff --git a/drivers/xen/privcmd/privcmd.c b/drivers/xen/privcmd/privcmd.c
+new file mode 100644
+index 0000000..233b9c4
+--- /dev/null
++++ b/drivers/xen/privcmd/privcmd.c
+@@ -0,0 +1,354 @@
++/******************************************************************************
++ * privcmd.c
++ *
++ * Interface to privileged domain-0 commands.
++ *
++ * Copyright (c) 2002-2004, K A Fraser, B Dragovic
++ */
++
++#include <linux/kernel.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <linux/errno.h>
++#include <linux/mm.h>
++#include <linux/mman.h>
++#include <linux/swap.h>
++#include <linux/smp_lock.h>
++#include <linux/highmem.h>
++#include <linux/pagemap.h>
++#include <linux/seq_file.h>
++#include <asm/hypervisor.h>
++
++#include <asm/pgalloc.h>
++#include <asm/pgtable.h>
++#include <asm/uaccess.h>
++#include <asm/tlb.h>
++#include <asm/hypervisor.h>
++#include <xen/public/privcmd.h>
++#include <xen/interface/xen.h>
++#include <xen/xen_proc.h>
++#include <xen/features.h>
++
++static struct proc_dir_entry *privcmd_intf;
++static struct proc_dir_entry *capabilities_intf;
++
++#ifndef HAVE_ARCH_PRIVCMD_MMAP
++static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma);
++#endif
++
++static long privcmd_ioctl(struct file *file,
++ unsigned int cmd, unsigned long data)
++{
++ int ret = -ENOSYS;
++ void __user *udata = (void __user *) data;
++
++ switch (cmd) {
++ case IOCTL_PRIVCMD_HYPERCALL: {
++ privcmd_hypercall_t hypercall;
++
++ if (copy_from_user(&hypercall, udata, sizeof(hypercall)))
++ return -EFAULT;
++
++#if defined(__i386__)
++ if (hypercall.op >= (PAGE_SIZE >> 5))
++ break;
++ __asm__ __volatile__ (
++ "pushl %%ebx; pushl %%ecx; pushl %%edx; "
++ "pushl %%esi; pushl %%edi; "
++ "movl 8(%%eax),%%ebx ;"
++ "movl 16(%%eax),%%ecx ;"
++ "movl 24(%%eax),%%edx ;"
++ "movl 32(%%eax),%%esi ;"
++ "movl 40(%%eax),%%edi ;"
++ "movl (%%eax),%%eax ;"
++ "shll $5,%%eax ;"
++ "addl $hypercall_page,%%eax ;"
++ "call *%%eax ;"
++ "popl %%edi; popl %%esi; popl %%edx; "
++ "popl %%ecx; popl %%ebx"
++ : "=a" (ret) : "0" (&hypercall) : "memory" );
++#elif defined (__x86_64__)
++ if (hypercall.op < (PAGE_SIZE >> 5)) {
++ long ign1, ign2, ign3;
++ __asm__ __volatile__ (
++ "movq %8,%%r10; movq %9,%%r8;"
++ "shll $5,%%eax ;"
++ "addq $hypercall_page,%%rax ;"
++ "call *%%rax"
++ : "=a" (ret), "=D" (ign1),
++ "=S" (ign2), "=d" (ign3)
++ : "0" ((unsigned int)hypercall.op),
++ "1" (hypercall.arg[0]),
++ "2" (hypercall.arg[1]),
++ "3" (hypercall.arg[2]),
++ "g" (hypercall.arg[3]),
++ "g" (hypercall.arg[4])
++ : "r8", "r10", "memory" );
++ }
++#else
++ ret = privcmd_hypercall(&hypercall);
++#endif
++ }
++ break;
++
++ case IOCTL_PRIVCMD_MMAP: {
++#define MMAP_NR_PER_PAGE (int)((PAGE_SIZE-sizeof(struct list_head))/sizeof(privcmd_mmap_entry_t))
++ privcmd_mmap_t mmapcmd;
++ privcmd_mmap_entry_t *msg;
++ privcmd_mmap_entry_t __user *p;
++ struct mm_struct *mm = current->mm;
++ struct vm_area_struct *vma;
++ unsigned long va;
++ int i, rc;
++ LIST_HEAD(pagelist);
++ struct list_head *l,*l2;
++
++ if (!is_initial_xendomain())
++ return -EPERM;
++
++ if (copy_from_user(&mmapcmd, udata, sizeof(mmapcmd)))
++ return -EFAULT;
++
++ p = mmapcmd.entry;
++ for (i = 0; i < mmapcmd.num;) {
++ int nr = min(mmapcmd.num - i, MMAP_NR_PER_PAGE);
++
++ rc = -ENOMEM;
++ l = (struct list_head *) __get_free_page(GFP_KERNEL);
++ if (l == NULL)
++ goto mmap_out;
++
++ INIT_LIST_HEAD(l);
++ list_add_tail(l, &pagelist);
++ msg = (privcmd_mmap_entry_t*)(l + 1);
++
++ rc = -EFAULT;
++ if (copy_from_user(msg, p, nr*sizeof(*msg)))
++ goto mmap_out;
++ i += nr;
++ p += nr;
++ }
++
++ l = pagelist.next;
++ msg = (privcmd_mmap_entry_t*)(l + 1);
++
++ down_write(&mm->mmap_sem);
++
++ vma = find_vma(mm, msg->va);
++ rc = -EINVAL;
++ if (!vma || (msg->va != vma->vm_start) ||
++ !privcmd_enforce_singleshot_mapping(vma))
++ goto mmap_out;
++
++ va = vma->vm_start;
++
++ i = 0;
++ list_for_each(l, &pagelist) {
++ int nr = i + min(mmapcmd.num - i, MMAP_NR_PER_PAGE);
++
++ msg = (privcmd_mmap_entry_t*)(l + 1);
++ while (i<nr) {
++
++ /* Do not allow range to wrap the address space. */
++ rc = -EINVAL;
++ if ((msg->npages > (LONG_MAX >> PAGE_SHIFT)) ||
++ ((unsigned long)(msg->npages << PAGE_SHIFT) >= -va))
++ goto mmap_out;
++
++ /* Range chunks must be contiguous in va space. */
++ if ((msg->va != va) ||
++ ((msg->va+(msg->npages<<PAGE_SHIFT)) > vma->vm_end))
++ goto mmap_out;
++
++ if ((rc = direct_remap_pfn_range(
++ vma,
++ msg->va & PAGE_MASK,
++ msg->mfn,
++ msg->npages << PAGE_SHIFT,
++ vma->vm_page_prot,
++ mmapcmd.dom)) < 0)
++ goto mmap_out;
++
++ va += msg->npages << PAGE_SHIFT;
++ msg++;
++ i++;
++ }
++ }
++
++ rc = 0;
++
++ mmap_out:
++ up_write(&mm->mmap_sem);
++ list_for_each_safe(l,l2,&pagelist)
++ free_page((unsigned long)l);
++ ret = rc;
++ }
++#undef MMAP_NR_PER_PAGE
++ break;
++
++ case IOCTL_PRIVCMD_MMAPBATCH: {
++#define MMAPBATCH_NR_PER_PAGE (unsigned long)((PAGE_SIZE-sizeof(struct list_head))/sizeof(unsigned long))
++ privcmd_mmapbatch_t m;
++ struct mm_struct *mm = current->mm;
++ struct vm_area_struct *vma;
++ xen_pfn_t __user *p;
++ unsigned long addr, *mfn, nr_pages;
++ int i;
++ LIST_HEAD(pagelist);
++ struct list_head *l, *l2;
++
++ if (!is_initial_xendomain())
++ return -EPERM;
++
++ if (copy_from_user(&m, udata, sizeof(m)))
++ return -EFAULT;
++
++ nr_pages = m.num;
++ if ((m.num <= 0) || (nr_pages > (LONG_MAX >> PAGE_SHIFT)))
++ return -EINVAL;
++
++ p = m.arr;
++ for (i=0; i<nr_pages; ) {
++ int nr = min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
++
++ ret = -ENOMEM;
++ l = (struct list_head *)__get_free_page(GFP_KERNEL);
++ if (l == NULL)
++ goto mmapbatch_out;
++
++ INIT_LIST_HEAD(l);
++ list_add_tail(l, &pagelist);
++
++ mfn = (unsigned long*)(l + 1);
++ ret = -EFAULT;
++ if (copy_from_user(mfn, p, nr*sizeof(*mfn)))
++ goto mmapbatch_out;
++
++ i += nr; p+= nr;
++ }
++
++ down_write(&mm->mmap_sem);
++
++ vma = find_vma(mm, m.addr);
++ ret = -EINVAL;
++ if (!vma ||
++ (m.addr != vma->vm_start) ||
++ ((m.addr + (nr_pages << PAGE_SHIFT)) != vma->vm_end) ||
++ !privcmd_enforce_singleshot_mapping(vma)) {
++ up_write(&mm->mmap_sem);
++ goto mmapbatch_out;
++ }
++
++ p = m.arr;
++ addr = m.addr;
++ i = 0;
++ ret = 0;
++ list_for_each(l, &pagelist) {
++ int nr = i + min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
++ mfn = (unsigned long *)(l + 1);
++
++ while (i<nr) {
++ if(direct_remap_pfn_range(vma, addr & PAGE_MASK,
++ *mfn, PAGE_SIZE,
++ vma->vm_page_prot, m.dom) < 0) {
++ *mfn |= 0xf0000000U;
++ ret++;
++ }
++ mfn++; i++; addr += PAGE_SIZE;
++ }
++ }
++
++ up_write(&mm->mmap_sem);
++ if (ret > 0) {
++ p = m.arr;
++ i = 0;
++ ret = 0;
++ list_for_each(l, &pagelist) {
++ int nr = min(nr_pages - i, MMAPBATCH_NR_PER_PAGE);
++ mfn = (unsigned long *)(l + 1);
++ if (copy_to_user(p, mfn, nr*sizeof(*mfn)))
++ ret = -EFAULT;
++ i += nr; p += nr;
++ }
++ }
++ mmapbatch_out:
++ list_for_each_safe(l,l2,&pagelist)
++ free_page((unsigned long)l);
++#undef MMAPBATCH_NR_PER_PAGE
++ }
++ break;
++
++ default:
++ ret = -EINVAL;
++ break;
++ }
++
++ return ret;
++}
++
++#ifndef HAVE_ARCH_PRIVCMD_MMAP
++static int privcmd_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
++{
++ return VM_FAULT_SIGBUS;
++}
++
++static struct vm_operations_struct privcmd_vm_ops = {
++ .fault = privcmd_fault
++};
++
++static int privcmd_mmap(struct file * file, struct vm_area_struct * vma)
++{
++ /* Unsupported for auto-translate guests. */
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return -ENOSYS;
++
++ /* DONTCOPY is essential for Xen as copy_page_range is broken. */
++ vma->vm_flags |= VM_RESERVED | VM_IO | VM_PFNMAP | VM_DONTCOPY;
++ vma->vm_ops = &privcmd_vm_ops;
++ vma->vm_private_data = NULL;
++
++ return 0;
++}
++
++static int privcmd_enforce_singleshot_mapping(struct vm_area_struct *vma)
++{
++ return (xchg(&vma->vm_private_data, (void *)1) == NULL);
++}
++#endif
++
++static const struct file_operations privcmd_file_ops = {
++ .unlocked_ioctl = privcmd_ioctl,
++ .mmap = privcmd_mmap,
++};
++
++static int capabilities_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len = 0;
++ *page = 0;
++
++ if (is_initial_xendomain())
++ len = sprintf( page, "control_d\n" );
++
++ *eof = 1;
++ return len;
++}
++
++static int __init privcmd_init(void)
++{
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ privcmd_intf = create_xen_proc_entry("privcmd", 0400);
++ if (privcmd_intf != NULL)
++ privcmd_intf->proc_fops = &privcmd_file_ops;
++
++ capabilities_intf = create_xen_proc_entry("capabilities", 0400 );
++ if (capabilities_intf != NULL)
++ capabilities_intf->read_proc = capabilities_read;
++
++ return 0;
++}
++
++__initcall(privcmd_init);
+diff --git a/drivers/xen/scsiback/Makefile b/drivers/xen/scsiback/Makefile
+new file mode 100644
+index 0000000..56271df
+--- /dev/null
++++ b/drivers/xen/scsiback/Makefile
+@@ -0,0 +1,4 @@
++obj-$(CONFIG_XEN_SCSI_BACKEND) := xen-scsibk.o
++
++xen-scsibk-y := interface.o scsiback.o xenbus.o translate.o emulate.o
++
+diff --git a/drivers/xen/scsiback/common.h b/drivers/xen/scsiback/common.h
+new file mode 100644
+index 0000000..497218a
+--- /dev/null
++++ b/drivers/xen/scsiback/common.h
+@@ -0,0 +1,181 @@
++/*
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __SCSIIF__BACKEND__COMMON_H__
++#define __SCSIIF__BACKEND__COMMON_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <linux/vmalloc.h>
++#include <linux/wait.h>
++#include <linux/sched.h>
++#include <linux/kthread.h>
++#include <linux/blkdev.h>
++#include <linux/list.h>
++#include <linux/kthread.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi_dbg.h>
++#include <scsi/scsi_eh.h>
++#include <asm/io.h>
++#include <asm/setup.h>
++#include <asm/pgalloc.h>
++#include <asm/delay.h>
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#include <xen/xenbus.h>
++#include <xen/interface/io/ring.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/vscsiif.h>
++
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++struct ids_tuple {
++ unsigned int hst; /* host */
++ unsigned int chn; /* channel */
++ unsigned int tgt; /* target */
++ unsigned int lun; /* LUN */
++};
++
++struct v2p_entry {
++ struct ids_tuple v; /* translate from */
++ struct scsi_device *sdev; /* translate to */
++ struct list_head l;
++};
++
++struct vscsibk_info {
++ struct xenbus_device *dev;
++
++ domid_t domid;
++ unsigned int evtchn;
++ unsigned int irq;
++
++ struct vscsiif_back_ring ring;
++ struct vm_struct *ring_area;
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++
++ spinlock_t ring_lock;
++ atomic_t nr_unreplied_reqs;
++
++ spinlock_t v2p_lock;
++ struct list_head v2p_entry_lists;
++
++ struct task_struct *kthread;
++ wait_queue_head_t waiting_to_free;
++ wait_queue_head_t wq;
++ unsigned int waiting_reqs;
++ struct page **mmap_pages;
++
++};
++
++typedef struct {
++ unsigned char act;
++ struct vscsibk_info *info;
++ struct scsi_device *sdev;
++
++ uint16_t rqid;
++
++ uint8_t nr_segments;
++ uint8_t cmnd[VSCSIIF_MAX_COMMAND_SIZE];
++ uint8_t cmd_len;
++
++ uint8_t sc_data_direction;
++ uint16_t timeout_per_command;
++
++ uint32_t request_bufflen;
++ struct scatterlist *sgl;
++ grant_ref_t gref[VSCSIIF_SG_TABLESIZE];
++
++ int32_t rslt;
++ uint32_t resid;
++ uint8_t sense_buffer[VSCSIIF_SENSE_BUFFERSIZE];
++
++ struct list_head free_list;
++} pending_req_t;
++
++
++
++#define scsiback_get(_b) (atomic_inc(&(_b)->nr_unreplied_reqs))
++#define scsiback_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->nr_unreplied_reqs)) \
++ wake_up(&(_b)->waiting_to_free);\
++ } while (0)
++
++#define VSCSIIF_TIMEOUT (900*HZ)
++
++
++irqreturn_t scsiback_intr(int, void *);
++int scsiback_init_sring(struct vscsibk_info *info,
++ unsigned long ring_ref, unsigned int evtchn);
++int scsiback_schedule(void *data);
++
++
++struct vscsibk_info *vscsibk_info_alloc(domid_t domid);
++void scsiback_free(struct vscsibk_info *info);
++void scsiback_disconnect(struct vscsibk_info *info);
++int __init scsiback_interface_init(void);
++void scsiback_interface_exit(void);
++int scsiback_xenbus_init(void);
++void scsiback_xenbus_unregister(void);
++
++void scsiback_init_translation_table(struct vscsibk_info *info);
++
++int scsiback_add_translation_entry(struct vscsibk_info *info,
++ struct scsi_device *sdev, struct ids_tuple *v);
++
++int scsiback_del_translation_entry(struct vscsibk_info *info,
++ struct ids_tuple *v);
++struct scsi_device *scsiback_do_translation(struct vscsibk_info *info,
++ struct ids_tuple *v);
++void scsiback_release_translation_entry(struct vscsibk_info *info);
++
++
++void scsiback_cmd_exec(pending_req_t *pending_req);
++void scsiback_do_resp_with_sense(char *sense_buffer, int32_t result,
++ uint32_t resid, pending_req_t *pending_req);
++void scsiback_fast_flush_area(pending_req_t *req);
++
++void scsiback_rsp_emulation(pending_req_t *pending_req);
++void scsiback_req_emulation_or_cmdexec(pending_req_t *pending_req);
++void scsiback_emulation_init(void);
++
++
++#endif /* __SCSIIF__BACKEND__COMMON_H__ */
+diff --git a/drivers/xen/scsiback/emulate.c b/drivers/xen/scsiback/emulate.c
+new file mode 100644
+index 0000000..7ba3274
+--- /dev/null
++++ b/drivers/xen/scsiback/emulate.c
+@@ -0,0 +1,453 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_device.h>
++#include "common.h"
++
++/* Following SCSI commands are not defined in scsi/scsi.h */
++#define EXTENDED_COPY 0x83 /* EXTENDED COPY command */
++#define REPORT_ALIASES 0xa3 /* REPORT ALIASES command */
++#define CHANGE_ALIASES 0xa4 /* CHANGE ALIASES command */
++#define SET_PRIORITY 0xa4 /* SET PRIORITY command */
++
++
++/*
++ The bitmap in order to control emulation.
++ (Bit 3 to 7 are reserved for future use.)
++*/
++#define VSCSIIF_NEED_CMD_EXEC 0x01 /* If this bit is set, cmd exec */
++ /* is required. */
++#define VSCSIIF_NEED_EMULATE_REQBUF 0x02 /* If this bit is set, need */
++ /* emulation reqest buff before */
++ /* cmd exec. */
++#define VSCSIIF_NEED_EMULATE_RSPBUF 0x04 /* If this bit is set, need */
++ /* emulation resp buff after */
++ /* cmd exec. */
++
++/* Additional Sense Code (ASC) used */
++#define NO_ADDITIONAL_SENSE 0x0
++#define LOGICAL_UNIT_NOT_READY 0x4
++#define UNRECOVERED_READ_ERR 0x11
++#define PARAMETER_LIST_LENGTH_ERR 0x1a
++#define INVALID_OPCODE 0x20
++#define ADDR_OUT_OF_RANGE 0x21
++#define INVALID_FIELD_IN_CDB 0x24
++#define INVALID_FIELD_IN_PARAM_LIST 0x26
++#define POWERON_RESET 0x29
++#define SAVING_PARAMS_UNSUP 0x39
++#define THRESHOLD_EXCEEDED 0x5d
++#define LOW_POWER_COND_ON 0x5e
++
++
++
++/* Number os SCSI op_code */
++#define VSCSI_MAX_SCSI_OP_CODE 256
++static unsigned char bitmap[VSCSI_MAX_SCSI_OP_CODE];
++
++
++
++/*
++ Emulation routines for each SCSI op_code.
++*/
++static void (*pre_function[VSCSI_MAX_SCSI_OP_CODE])(pending_req_t *, void *);
++static void (*post_function[VSCSI_MAX_SCSI_OP_CODE])(pending_req_t *, void *);
++
++
++static const int check_condition_result =
++ (DRIVER_SENSE << 24) | SAM_STAT_CHECK_CONDITION;
++
++static void scsiback_mk_sense_buffer(uint8_t *data, uint8_t key,
++ uint8_t asc, uint8_t asq)
++{
++ data[0] = 0x70; /* fixed, current */
++ data[2] = key;
++ data[7] = 0xa; /* implies 18 byte sense buffer */
++ data[12] = asc;
++ data[13] = asq;
++}
++
++static void resp_not_supported_cmd(pending_req_t *pending_req, void *data)
++{
++ scsiback_mk_sense_buffer(pending_req->sense_buffer, ILLEGAL_REQUEST,
++ INVALID_OPCODE, 0);
++ pending_req->resid = 0;
++ pending_req->rslt = check_condition_result;
++}
++
++
++static int __copy_to_sg(struct scatterlist *sgl, unsigned int nr_sg,
++ void *buf, unsigned int buflen)
++{
++ struct scatterlist *sg;
++ void *from = buf;
++ void *to;
++ unsigned int from_rest = buflen;
++ unsigned int to_capa;
++ unsigned int copy_size = 0;
++ unsigned int i;
++ unsigned long pfn;
++
++ for_each_sg (sgl, sg, nr_sg, i) {
++ if (sg_page(sg) == NULL) {
++ printk(KERN_WARNING "%s: inconsistent length field in "
++ "scatterlist\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ to_capa = sg->length;
++ copy_size = min_t(unsigned int, to_capa, from_rest);
++
++ pfn = page_to_pfn(sg_page(sg));
++ to = pfn_to_kaddr(pfn) + (sg->offset);
++ memcpy(to, from, copy_size);
++
++ from_rest -= copy_size;
++ if (from_rest == 0) {
++ return 0;
++ }
++
++ from += copy_size;
++ }
++
++ printk(KERN_WARNING "%s: no space in scatterlist\n",
++ __FUNCTION__);
++ return -ENOMEM;
++}
++
++static int __copy_from_sg(struct scatterlist *sgl, unsigned int nr_sg,
++ void *buf, unsigned int buflen)
++{
++ struct scatterlist *sg;
++ void *from;
++ void *to = buf;
++ unsigned int from_rest;
++ unsigned int to_capa = buflen;
++ unsigned int copy_size;
++ unsigned int i;
++ unsigned long pfn;
++
++ for_each_sg (sgl, sg, nr_sg, i) {
++ if (sg_page(sg) == NULL) {
++ printk(KERN_WARNING "%s: inconsistent length field in "
++ "scatterlist\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ from_rest = sg->length;
++ if ((from_rest > 0) && (to_capa < from_rest)) {
++ printk(KERN_WARNING
++ "%s: no space in destination buffer\n",
++ __FUNCTION__);
++ return -ENOMEM;
++ }
++ copy_size = from_rest;
++
++ pfn = page_to_pfn(sg_page(sg));
++ from = pfn_to_kaddr(pfn) + (sg->offset);
++ memcpy(to, from, copy_size);
++
++ to_capa -= copy_size;
++ to += copy_size;
++ }
++
++ return 0;
++}
++
++static int __nr_luns_under_host(struct vscsibk_info *info)
++{
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++ int lun_cnt = 0;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ lun_cnt++;
++ }
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++
++ return (lun_cnt);
++}
++
++
++/* REPORT LUNS Define*/
++#define VSCSI_REPORT_LUNS_HEADER 8
++#define VSCSI_REPORT_LUNS_RETRY 3
++
++/* quoted scsi_debug.c/resp_report_luns() */
++static void __report_luns(pending_req_t *pending_req, void *data)
++{
++ struct vscsibk_info *info = pending_req->info;
++ unsigned int channel = pending_req->sdev->channel;
++ unsigned int target = pending_req->sdev->id;
++ unsigned int nr_seg = pending_req->nr_segments;
++ unsigned char *cmd = (unsigned char *)pending_req->cmnd;
++
++ unsigned char *buff = NULL;
++ unsigned char alloc_len;
++ unsigned int alloc_luns = 0;
++ unsigned int req_bufflen = 0;
++ unsigned int actual_len = 0;
++ unsigned int retry_cnt = 0;
++ int select_report = (int)cmd[2];
++ int i, lun_cnt = 0, lun, upper, err = 0;
++
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ struct scsi_lun *one_lun;
++
++ req_bufflen = cmd[9] + (cmd[8] << 8) + (cmd[7] << 16) + (cmd[6] << 24);
++ if ((req_bufflen < 4) || (select_report != 0))
++ goto fail;
++
++ alloc_luns = __nr_luns_under_host(info);
++ alloc_len = sizeof(struct scsi_lun) * alloc_luns
++ + VSCSI_REPORT_LUNS_HEADER;
++retry:
++ if ((buff = kmalloc(alloc_len, GFP_KERNEL)) == NULL) {
++ printk(KERN_ERR "scsiback:%s kmalloc err\n", __FUNCTION__);
++ goto fail;
++ }
++
++ memset(buff, 0, alloc_len);
++
++ one_lun = (struct scsi_lun *) &buff[8];
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == channel) &&
++ (entry->v.tgt == target)) {
++
++ /* check overflow */
++ if (lun_cnt >= alloc_luns) {
++ spin_unlock_irqrestore(&info->v2p_lock,
++ flags);
++
++ if (retry_cnt < VSCSI_REPORT_LUNS_RETRY) {
++ retry_cnt++;
++ if (buff)
++ kfree(buff);
++ goto retry;
++ }
++
++ goto fail;
++ }
++
++ lun = entry->v.lun;
++ upper = (lun >> 8) & 0x3f;
++ if (upper)
++ one_lun[lun_cnt].scsi_lun[0] = upper;
++ one_lun[lun_cnt].scsi_lun[1] = lun & 0xff;
++ lun_cnt++;
++ }
++ }
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++
++ buff[2] = ((sizeof(struct scsi_lun) * lun_cnt) >> 8) & 0xff;
++ buff[3] = (sizeof(struct scsi_lun) * lun_cnt) & 0xff;
++
++ actual_len = lun_cnt * sizeof(struct scsi_lun)
++ + VSCSI_REPORT_LUNS_HEADER;
++ req_bufflen = 0;
++ for (i = 0; i < nr_seg; i++)
++ req_bufflen += pending_req->sgl[i].length;
++
++ err = __copy_to_sg(pending_req->sgl, nr_seg, buff,
++ min(req_bufflen, actual_len));
++ if (err)
++ goto fail;
++
++ memset(pending_req->sense_buffer, 0, VSCSIIF_SENSE_BUFFERSIZE);
++ pending_req->rslt = 0x00;
++ pending_req->resid = req_bufflen - min(req_bufflen, actual_len);
++
++ kfree(buff);
++ return;
++
++fail:
++ scsiback_mk_sense_buffer(pending_req->sense_buffer, ILLEGAL_REQUEST,
++ INVALID_FIELD_IN_CDB, 0);
++ pending_req->rslt = check_condition_result;
++ pending_req->resid = 0;
++ if (buff)
++ kfree(buff);
++ return;
++}
++
++
++
++int __pre_do_emulation(pending_req_t *pending_req, void *data)
++{
++ uint8_t op_code = pending_req->cmnd[0];
++
++ if ((bitmap[op_code] & VSCSIIF_NEED_EMULATE_REQBUF) &&
++ pre_function[op_code] != NULL) {
++ pre_function[op_code](pending_req, data);
++ }
++
++ /*
++ 0: no need for native driver call, so should return immediately.
++ 1: non emulation or should call native driver
++ after modifing the request buffer.
++ */
++ return !!(bitmap[op_code] & VSCSIIF_NEED_CMD_EXEC);
++}
++
++void scsiback_rsp_emulation(pending_req_t *pending_req)
++{
++ uint8_t op_code = pending_req->cmnd[0];
++
++ if ((bitmap[op_code] & VSCSIIF_NEED_EMULATE_RSPBUF) &&
++ post_function[op_code] != NULL) {
++ post_function[op_code](pending_req, NULL);
++ }
++
++ return;
++}
++
++
++void scsiback_req_emulation_or_cmdexec(pending_req_t *pending_req)
++{
++ if (__pre_do_emulation(pending_req, NULL)) {
++ scsiback_cmd_exec(pending_req);
++ }
++ else {
++ scsiback_fast_flush_area(pending_req);
++ scsiback_do_resp_with_sense(pending_req->sense_buffer,
++ pending_req->rslt, pending_req->resid, pending_req);
++ }
++}
++
++
++/*
++ Following are not customizable functions.
++*/
++void scsiback_emulation_init(void)
++{
++ int i;
++
++ /* Initialize to default state */
++ for (i = 0; i < VSCSI_MAX_SCSI_OP_CODE; i++) {
++ bitmap[i] = (VSCSIIF_NEED_EMULATE_REQBUF |
++ VSCSIIF_NEED_EMULATE_RSPBUF);
++ pre_function[i] = resp_not_supported_cmd;
++ post_function[i] = NULL;
++ /* means,
++ - no need for pre-emulation
++ - no need for post-emulation
++ - call native driver
++ */
++ }
++
++ /*
++ Register appropriate functions below as you need.
++ (See scsi/scsi.h for definition of SCSI op_code.)
++ */
++
++ /*
++ This command is Non emulation.
++ */
++ bitmap[TEST_UNIT_READY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[TEST_UNIT_READY] = NULL;
++ post_function[TEST_UNIT_READY] = NULL;
++
++ bitmap[REZERO_UNIT] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[REZERO_UNIT] = NULL;
++ post_function[REZERO_UNIT] = NULL;
++
++ bitmap[REQUEST_SENSE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[REQUEST_SENSE] = NULL;
++ post_function[REQUEST_SENSE] = NULL;
++
++ bitmap[FORMAT_UNIT] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[FORMAT_UNIT] = NULL;
++ post_function[FORMAT_UNIT] = NULL;
++
++ bitmap[READ_BLOCK_LIMITS] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_BLOCK_LIMITS] = NULL;
++ post_function[READ_BLOCK_LIMITS] = NULL;
++
++ bitmap[READ_6] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_6] = NULL;
++ post_function[READ_6] = NULL;
++
++ bitmap[WRITE_6] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_6] = NULL;
++ post_function[WRITE_6] = NULL;
++
++ bitmap[WRITE_FILEMARKS] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_FILEMARKS] = NULL;
++ post_function[WRITE_FILEMARKS] = NULL;
++
++ bitmap[SPACE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[SPACE] = NULL;
++ post_function[SPACE] = NULL;
++
++ bitmap[INQUIRY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[INQUIRY] = NULL;
++ post_function[INQUIRY] = NULL;
++
++ bitmap[ERASE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[ERASE] = NULL;
++ post_function[ERASE] = NULL;
++
++ bitmap[MODE_SENSE] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[MODE_SENSE] = NULL;
++ post_function[MODE_SENSE] = NULL;
++
++ bitmap[SEND_DIAGNOSTIC] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[SEND_DIAGNOSTIC] = NULL;
++ post_function[SEND_DIAGNOSTIC] = NULL;
++
++ bitmap[READ_CAPACITY] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_CAPACITY] = NULL;
++ post_function[READ_CAPACITY] = NULL;
++
++ bitmap[READ_10] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[READ_10] = NULL;
++ post_function[READ_10] = NULL;
++
++ bitmap[WRITE_10] = VSCSIIF_NEED_CMD_EXEC;
++ pre_function[WRITE_10] = NULL;
++ post_function[WRITE_10] = NULL;
++
++ /*
++ This command is Full emulation.
++ */
++ pre_function[REPORT_LUNS] = __report_luns;
++ bitmap[REPORT_LUNS] = (VSCSIIF_NEED_EMULATE_REQBUF |
++ VSCSIIF_NEED_EMULATE_RSPBUF);
++
++ return;
++}
+diff --git a/drivers/xen/scsiback/interface.c b/drivers/xen/scsiback/interface.c
+new file mode 100644
+index 0000000..5c51c3b
+--- /dev/null
++++ b/drivers/xen/scsiback/interface.c
+@@ -0,0 +1,182 @@
++/*
++ * interface management.
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include "common.h"
++
++#include <xen/evtchn.h>
++#include <linux/kthread.h>
++
++
++static struct kmem_cache *scsiback_cachep;
++
++struct vscsibk_info *vscsibk_info_alloc(domid_t domid)
++{
++ struct vscsibk_info *info;
++
++ info = kmem_cache_alloc(scsiback_cachep, GFP_KERNEL);
++ if (!info)
++ return ERR_PTR(-ENOMEM);
++
++ memset(info, 0, sizeof(*info));
++ info->domid = domid;
++ spin_lock_init(&info->ring_lock);
++ atomic_set(&info->nr_unreplied_reqs, 0);
++ init_waitqueue_head(&info->wq);
++ init_waitqueue_head(&info->waiting_to_free);
++
++ return info;
++}
++
++static int map_frontend_page( struct vscsibk_info *info,
++ unsigned long ring_ref)
++{
++ struct gnttab_map_grant_ref op;
++ int err;
++
++ gnttab_set_map_op(&op, (unsigned long)info->ring_area->addr,
++ GNTMAP_host_map, ring_ref,
++ info->domid);
++
++ err = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1);
++ BUG_ON(err);
++
++ if (op.status) {
++ printk(KERN_ERR "scsiback: Grant table operation failure !\n");
++ return op.status;
++ }
++
++ info->shmem_ref = ring_ref;
++ info->shmem_handle = op.handle;
++
++ return (GNTST_okay);
++}
++
++static void unmap_frontend_page(struct vscsibk_info *info)
++{
++ struct gnttab_unmap_grant_ref op;
++ int err;
++
++ gnttab_set_unmap_op(&op, (unsigned long)info->ring_area->addr,
++ GNTMAP_host_map, info->shmem_handle);
++
++ err = HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1);
++ BUG_ON(err);
++
++}
++
++int scsiback_init_sring(struct vscsibk_info *info,
++ unsigned long ring_ref, unsigned int evtchn)
++{
++ struct vscsiif_sring *sring;
++ int err;
++
++ if (info->irq) {
++ printk(KERN_ERR "scsiback: Already connected through?\n");
++ return -1;
++ }
++
++ info->ring_area = alloc_vm_area(PAGE_SIZE);
++ if (!info)
++ return -ENOMEM;
++
++ err = map_frontend_page(info, ring_ref);
++ if (err)
++ goto free_vm;
++
++ sring = (struct vscsiif_sring *) info->ring_area->addr;
++ BACK_RING_INIT(&info->ring, sring, PAGE_SIZE);
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ info->domid, evtchn,
++ scsiback_intr, 0, "vscsiif-backend", info);
++
++ if (err < 0)
++ goto unmap_page;
++
++ info->irq = err;
++
++ return 0;
++
++unmap_page:
++ unmap_frontend_page(info);
++free_vm:
++ free_vm_area(info->ring_area);
++
++ return err;
++}
++
++void scsiback_disconnect(struct vscsibk_info *info)
++{
++ if (info->kthread) {
++ kthread_stop(info->kthread);
++ info->kthread = NULL;
++ }
++
++ wait_event(info->waiting_to_free,
++ atomic_read(&info->nr_unreplied_reqs) == 0);
++
++ if (info->irq) {
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
++ }
++
++ if (info->ring.sring) {
++ unmap_frontend_page(info);
++ free_vm_area(info->ring_area);
++ info->ring.sring = NULL;
++ }
++}
++
++void scsiback_free(struct vscsibk_info *info)
++{
++ kmem_cache_free(scsiback_cachep, info);
++}
++
++int __init scsiback_interface_init(void)
++{
++ scsiback_cachep = kmem_cache_create("vscsiif_cache",
++ sizeof(struct vscsibk_info), 0, 0, NULL);
++ if (!scsiback_cachep) {
++ printk(KERN_ERR "scsiback: can't init scsi cache\n");
++ return -ENOMEM;
++ }
++
++ return 0;
++}
++
++void scsiback_interface_exit(void)
++{
++ kmem_cache_destroy(scsiback_cachep);
++}
+diff --git a/drivers/xen/scsiback/scsiback.c b/drivers/xen/scsiback/scsiback.c
+new file mode 100644
+index 0000000..6efe62f
+--- /dev/null
++++ b/drivers/xen/scsiback/scsiback.c
+@@ -0,0 +1,706 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/spinlock.h>
++#include <linux/kthread.h>
++#include <linux/list.h>
++#include <linux/delay.h>
++#include <xen/balloon.h>
++#include <asm/hypervisor.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi_dbg.h>
++#include <scsi/scsi_eh.h>
++
++#include "common.h"
++
++
++struct list_head pending_free;
++DEFINE_SPINLOCK(pending_free_lock);
++DECLARE_WAIT_QUEUE_HEAD(pending_free_wq);
++
++int vscsiif_reqs = VSCSIIF_BACK_MAX_PENDING_REQS;
++module_param_named(reqs, vscsiif_reqs, int, 0);
++MODULE_PARM_DESC(reqs, "Number of scsiback requests to allocate");
++
++static unsigned int log_print_stat = 0;
++module_param(log_print_stat, int, 0644);
++
++#define SCSIBACK_INVALID_HANDLE (~0)
++
++static pending_req_t *pending_reqs;
++static struct page **pending_pages;
++static grant_handle_t *pending_grant_handles;
++
++static int vaddr_pagenr(pending_req_t *req, int seg)
++{
++ return (req - pending_reqs) * VSCSIIF_SG_TABLESIZE + seg;
++}
++
++static unsigned long vaddr(pending_req_t *req, int seg)
++{
++ unsigned long pfn = page_to_pfn(pending_pages[vaddr_pagenr(req, seg)]);
++ return (unsigned long)pfn_to_kaddr(pfn);
++}
++
++#define pending_handle(_req, _seg) \
++ (pending_grant_handles[vaddr_pagenr(_req, _seg)])
++
++
++void scsiback_fast_flush_area(pending_req_t *req)
++{
++ struct gnttab_unmap_grant_ref unmap[VSCSIIF_SG_TABLESIZE];
++ unsigned int i, invcount = 0;
++ grant_handle_t handle;
++ int err;
++
++ if (req->nr_segments) {
++ for (i = 0; i < req->nr_segments; i++) {
++ handle = pending_handle(req, i);
++ if (handle == SCSIBACK_INVALID_HANDLE)
++ continue;
++ gnttab_set_unmap_op(&unmap[i], vaddr(req, i),
++ GNTMAP_host_map, handle);
++ pending_handle(req, i) = SCSIBACK_INVALID_HANDLE;
++ invcount++;
++ }
++
++ err = HYPERVISOR_grant_table_op(
++ GNTTABOP_unmap_grant_ref, unmap, invcount);
++ BUG_ON(err);
++ kfree(req->sgl);
++ }
++
++ return;
++}
++
++
++static pending_req_t * alloc_req(struct vscsibk_info *info)
++{
++ pending_req_t *req = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++ if (!list_empty(&pending_free)) {
++ req = list_entry(pending_free.next, pending_req_t, free_list);
++ list_del(&req->free_list);
++ }
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ return req;
++}
++
++
++static void free_req(pending_req_t *req)
++{
++ unsigned long flags;
++ int was_empty;
++
++ spin_lock_irqsave(&pending_free_lock, flags);
++ was_empty = list_empty(&pending_free);
++ list_add(&req->free_list, &pending_free);
++ spin_unlock_irqrestore(&pending_free_lock, flags);
++ if (was_empty)
++ wake_up(&pending_free_wq);
++}
++
++
++static void scsiback_notify_work(struct vscsibk_info *info)
++{
++ info->waiting_reqs = 1;
++ wake_up(&info->wq);
++}
++
++void scsiback_do_resp_with_sense(char *sense_buffer, int32_t result,
++ uint32_t resid, pending_req_t *pending_req)
++{
++ vscsiif_response_t *ring_res;
++ struct vscsibk_info *info = pending_req->info;
++ int notify;
++ int more_to_do = 1;
++ unsigned long flags;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ spin_lock_irqsave(&info->ring_lock, flags);
++
++ ring_res = RING_GET_RESPONSE(&info->ring, info->ring.rsp_prod_pvt);
++ info->ring.rsp_prod_pvt++;
++
++ ring_res->rslt = result;
++ ring_res->rqid = pending_req->rqid;
++
++ if (sense_buffer != NULL) {
++ memcpy(ring_res->sense_buffer, sense_buffer,
++ VSCSIIF_SENSE_BUFFERSIZE);
++ ring_res->sense_len = VSCSIIF_SENSE_BUFFERSIZE;
++ } else {
++ ring_res->sense_len = 0;
++ }
++
++ ring_res->residual_len = resid;
++
++ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&info->ring, notify);
++ if (info->ring.rsp_prod_pvt == info->ring.req_cons) {
++ RING_FINAL_CHECK_FOR_REQUESTS(&info->ring, more_to_do);
++ } else if (RING_HAS_UNCONSUMED_REQUESTS(&info->ring)) {
++ more_to_do = 1;
++ }
++
++ spin_unlock_irqrestore(&info->ring_lock, flags);
++
++ if (more_to_do)
++ scsiback_notify_work(info);
++
++ if (notify)
++ notify_remote_via_irq(info->irq);
++
++ free_req(pending_req);
++}
++
++static void scsiback_print_status(char *sense_buffer, int errors,
++ pending_req_t *pending_req)
++{
++ struct scsi_device *sdev = pending_req->sdev;
++
++ printk(KERN_ERR "scsiback: %d:%d:%d:%d ",sdev->host->host_no,
++ sdev->channel, sdev->id, sdev->lun);
++ printk(KERN_ERR "status = 0x%02x, message = 0x%02x, host = 0x%02x, driver = 0x%02x\n",
++ status_byte(errors), msg_byte(errors),
++ host_byte(errors), driver_byte(errors));
++
++ printk(KERN_ERR "scsiback: cmnd[0]=0x%02X\n",
++ pending_req->cmnd[0]);
++
++ if (CHECK_CONDITION & status_byte(errors))
++ __scsi_print_sense("scsiback", sense_buffer, SCSI_SENSE_BUFFERSIZE);
++}
++
++
++static void scsiback_cmd_done(struct request *req, int errors)
++{
++ pending_req_t *pending_req = req->end_io_data;
++ unsigned char *sense_buffer;
++ unsigned int resid;
++
++ sense_buffer = req->sense;
++ resid = req->data_len;
++
++ if (errors != 0) {
++ if (log_print_stat)
++ scsiback_print_status(sense_buffer, errors, pending_req);
++ }
++
++ scsiback_rsp_emulation(pending_req);
++
++ scsiback_fast_flush_area(pending_req);
++ scsiback_do_resp_with_sense(sense_buffer, errors, resid, pending_req);
++ scsiback_put(pending_req->info);
++
++ __blk_put_request(req->q, req);
++}
++
++
++static int scsiback_gnttab_data_map(vscsiif_request_t *ring_req,
++ pending_req_t *pending_req)
++{
++ u32 flags;
++ int write;
++ int i, err = 0;
++ unsigned int data_len = 0;
++ struct gnttab_map_grant_ref map[VSCSIIF_SG_TABLESIZE];
++ struct vscsibk_info *info = pending_req->info;
++
++ int data_dir = (int)pending_req->sc_data_direction;
++ unsigned int nr_segments = (unsigned int)pending_req->nr_segments;
++
++ write = (data_dir == DMA_TO_DEVICE);
++
++ if (nr_segments) {
++ struct scatterlist *sg;
++
++ /* free of (sgl) in fast_flush_area()*/
++ pending_req->sgl = kmalloc(sizeof(struct scatterlist) * nr_segments,
++ GFP_KERNEL);
++ if (!pending_req->sgl) {
++ printk(KERN_ERR "scsiback: %s: kmalloc() error.\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ sg_init_table(pending_req->sgl, nr_segments);
++
++ for (i = 0; i < nr_segments; i++) {
++ flags = GNTMAP_host_map;
++ if (write)
++ flags |= GNTMAP_readonly;
++ gnttab_set_map_op(&map[i], vaddr(pending_req, i), flags,
++ ring_req->seg[i].gref,
++ info->domid);
++ }
++
++ err = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, nr_segments);
++ BUG_ON(err);
++
++ for_each_sg (pending_req->sgl, sg, nr_segments, i) {
++ if (unlikely(map[i].status != 0)) {
++ printk(KERN_ERR "scsiback: invalid buffer -- could not remap it\n");
++ map[i].handle = SCSIBACK_INVALID_HANDLE;
++ err |= 1;
++ }
++
++ pending_handle(pending_req, i) = map[i].handle;
++
++ if (err)
++ continue;
++
++ set_phys_to_machine(__pa(vaddr(
++ pending_req, i)) >> PAGE_SHIFT,
++ FOREIGN_FRAME(map[i].dev_bus_addr >> PAGE_SHIFT));
++
++ sg_set_page(sg, virt_to_page(vaddr(pending_req, i)),
++ ring_req->seg[i].length,
++ ring_req->seg[i].offset);
++ data_len += sg->length;
++
++ barrier();
++ if (sg->offset >= PAGE_SIZE ||
++ sg->length > PAGE_SIZE ||
++ sg->offset + sg->length > PAGE_SIZE)
++ err |= 1;
++
++ }
++
++ if (err)
++ goto fail_flush;
++ }
++
++ pending_req->request_bufflen = data_len;
++
++ return 0;
++
++fail_flush:
++ scsiback_fast_flush_area(pending_req);
++ return -ENOMEM;
++}
++
++/* quoted scsi_lib.c/scsi_merge_bio */
++static int scsiback_merge_bio(struct request *rq, struct bio *bio)
++{
++ struct request_queue *q = rq->q;
++
++ bio->bi_flags &= ~(1 << BIO_SEG_VALID);
++ if (rq_data_dir(rq) == WRITE)
++ bio->bi_rw |= (1 << BIO_RW);
++
++ blk_queue_bounce(q, &bio);
++
++ return blk_rq_append_bio(q, rq, bio);
++}
++
++
++/* quoted scsi_lib.c/scsi_bi_endio */
++static void scsiback_bi_endio(struct bio *bio, int error)
++{
++ bio_put(bio);
++}
++
++
++
++/* quoted scsi_lib.c/scsi_req_map_sg . */
++static int request_map_sg(struct request *rq, pending_req_t *pending_req, unsigned int count)
++{
++ struct request_queue *q = rq->q;
++ int nr_pages;
++ unsigned int nsegs = count;
++ unsigned int data_len = 0, len, bytes, off;
++ struct scatterlist *sg;
++ struct page *page;
++ struct bio *bio = NULL;
++ int i, err, nr_vecs = 0;
++
++ for_each_sg (pending_req->sgl, sg, nsegs, i) {
++ page = sg_page(sg);
++ off = sg->offset;
++ len = sg->length;
++ data_len += len;
++
++ nr_pages = (len + off + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ while (len > 0) {
++ bytes = min_t(unsigned int, len, PAGE_SIZE - off);
++
++ if (!bio) {
++ nr_vecs = min_t(int, BIO_MAX_PAGES, nr_pages);
++ nr_pages -= nr_vecs;
++ bio = bio_alloc(GFP_KERNEL, nr_vecs);
++ if (!bio) {
++ err = -ENOMEM;
++ goto free_bios;
++ }
++ bio->bi_end_io = scsiback_bi_endio;
++ }
++
++ if (bio_add_pc_page(q, bio, page, bytes, off) !=
++ bytes) {
++ bio_put(bio);
++ err = -EINVAL;
++ goto free_bios;
++ }
++
++ if (bio->bi_vcnt >= nr_vecs) {
++ err = scsiback_merge_bio(rq, bio);
++ if (err) {
++ bio_endio(bio, 0);
++ goto free_bios;
++ }
++ bio = NULL;
++ }
++
++ page++;
++ len -= bytes;
++ off = 0;
++ }
++ }
++
++ rq->buffer = rq->data = NULL;
++ rq->data_len = data_len;
++
++ return 0;
++
++free_bios:
++ while ((bio = rq->bio) != NULL) {
++ rq->bio = bio->bi_next;
++ /*
++ * call endio instead of bio_put incase it was bounced
++ */
++ bio_endio(bio, 0);
++ }
++
++ return err;
++}
++
++
++void scsiback_cmd_exec(pending_req_t *pending_req)
++{
++ int cmd_len = (int)pending_req->cmd_len;
++ int data_dir = (int)pending_req->sc_data_direction;
++ unsigned int nr_segments = (unsigned int)pending_req->nr_segments;
++ unsigned int timeout;
++ struct request *rq;
++ int write;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ /* because it doesn't timeout backend earlier than frontend.*/
++ if (pending_req->timeout_per_command)
++ timeout = pending_req->timeout_per_command * HZ;
++ else
++ timeout = VSCSIIF_TIMEOUT;
++
++ write = (data_dir == DMA_TO_DEVICE);
++ rq = blk_get_request(pending_req->sdev->request_queue, write, GFP_KERNEL);
++
++ rq->cmd_type = REQ_TYPE_BLOCK_PC;
++ rq->cmd_len = cmd_len;
++ memcpy(rq->cmd, pending_req->cmnd, cmd_len);
++
++ memset(pending_req->sense_buffer, 0, VSCSIIF_SENSE_BUFFERSIZE);
++ rq->sense = pending_req->sense_buffer;
++ rq->sense_len = 0;
++
++ /* not allowed to retry in backend. */
++ rq->retries = 0;
++ rq->timeout = timeout;
++ rq->end_io_data = pending_req;
++
++ if (nr_segments) {
++
++ if (request_map_sg(rq, pending_req, nr_segments)) {
++ printk(KERN_ERR "scsiback: SG Request Map Error\n");
++ return;
++ }
++ }
++
++ scsiback_get(pending_req->info);
++ blk_execute_rq_nowait(rq->q, NULL, rq, 1, scsiback_cmd_done);
++
++ return ;
++}
++
++
++static void scsiback_device_reset_exec(pending_req_t *pending_req)
++{
++ struct vscsibk_info *info = pending_req->info;
++ int err;
++ struct scsi_device *sdev = pending_req->sdev;
++
++ scsiback_get(info);
++ err = scsi_reset_provider(sdev, SCSI_TRY_RESET_DEVICE);
++
++ scsiback_do_resp_with_sense(NULL, err, 0, pending_req);
++ scsiback_put(info);
++
++ return;
++}
++
++
++irqreturn_t scsiback_intr(int irq, void *dev_id)
++{
++ scsiback_notify_work((struct vscsibk_info *)dev_id);
++ return IRQ_HANDLED;
++}
++
++static int prepare_pending_reqs(struct vscsibk_info *info,
++ vscsiif_request_t *ring_req, pending_req_t *pending_req)
++{
++ struct scsi_device *sdev;
++ struct ids_tuple vir;
++ int err = -EINVAL;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ pending_req->rqid = ring_req->rqid;
++ pending_req->act = ring_req->act;
++
++ pending_req->info = info;
++
++ vir.chn = ring_req->channel;
++ vir.tgt = ring_req->id;
++ vir.lun = ring_req->lun;
++
++ rmb();
++ sdev = scsiback_do_translation(info, &vir);
++ if (!sdev) {
++ pending_req->sdev = NULL;
++ DPRINTK("scsiback: doesn't exist.\n");
++ err = -ENODEV;
++ goto invalid_value;
++ }
++ pending_req->sdev = sdev;
++
++ /* request range check from frontend */
++ pending_req->sc_data_direction = ring_req->sc_data_direction;
++ barrier();
++ if ((pending_req->sc_data_direction != DMA_BIDIRECTIONAL) &&
++ (pending_req->sc_data_direction != DMA_TO_DEVICE) &&
++ (pending_req->sc_data_direction != DMA_FROM_DEVICE) &&
++ (pending_req->sc_data_direction != DMA_NONE)) {
++ DPRINTK("scsiback: invalid parameter data_dir = %d\n",
++ pending_req->sc_data_direction);
++ err = -EINVAL;
++ goto invalid_value;
++ }
++
++ pending_req->nr_segments = ring_req->nr_segments;
++ barrier();
++ if (pending_req->nr_segments > VSCSIIF_SG_TABLESIZE) {
++ DPRINTK("scsiback: invalid parameter nr_seg = %d\n",
++ pending_req->nr_segments);
++ err = -EINVAL;
++ goto invalid_value;
++ }
++
++ pending_req->cmd_len = ring_req->cmd_len;
++ barrier();
++ if (pending_req->cmd_len > VSCSIIF_MAX_COMMAND_SIZE) {
++ DPRINTK("scsiback: invalid parameter cmd_len = %d\n",
++ pending_req->cmd_len);
++ err = -EINVAL;
++ goto invalid_value;
++ }
++ memcpy(pending_req->cmnd, ring_req->cmnd, pending_req->cmd_len);
++
++ pending_req->timeout_per_command = ring_req->timeout_per_command;
++
++ if(scsiback_gnttab_data_map(ring_req, pending_req)) {
++ DPRINTK("scsiback: invalid buffer\n");
++ err = -EINVAL;
++ goto invalid_value;
++ }
++
++ return 0;
++
++invalid_value:
++ return err;
++}
++
++
++static int scsiback_do_cmd_fn(struct vscsibk_info *info)
++{
++ struct vscsiif_back_ring *ring = &info->ring;
++ vscsiif_request_t *ring_req;
++
++ pending_req_t *pending_req;
++ RING_IDX rc, rp;
++ int err, more_to_do = 0;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ rc = ring->req_cons;
++ rp = ring->sring->req_prod;
++ rmb();
++
++ while ((rc != rp)) {
++ if (RING_REQUEST_CONS_OVERFLOW(ring, rc))
++ break;
++ pending_req = alloc_req(info);
++ if (NULL == pending_req) {
++ more_to_do = 1;
++ break;
++ }
++
++ ring_req = RING_GET_REQUEST(ring, rc);
++ ring->req_cons = ++rc;
++
++ err = prepare_pending_reqs(info, ring_req,
++ pending_req);
++ if (err == -EINVAL) {
++ scsiback_do_resp_with_sense(NULL, (DRIVER_ERROR << 24),
++ 0, pending_req);
++ continue;
++ } else if (err == -ENODEV) {
++ scsiback_do_resp_with_sense(NULL, (DID_NO_CONNECT << 16),
++ 0, pending_req);
++ continue;
++ }
++
++ if (pending_req->act == VSCSIIF_ACT_SCSI_CDB) {
++ scsiback_req_emulation_or_cmdexec(pending_req);
++ } else if (pending_req->act == VSCSIIF_ACT_SCSI_RESET) {
++ scsiback_device_reset_exec(pending_req);
++ } else {
++ printk(KERN_ERR "scsiback: invalid parameter for request\n");
++ scsiback_do_resp_with_sense(NULL, (DRIVER_ERROR << 24),
++ 0, pending_req);
++ continue;
++ }
++ }
++
++ if (RING_HAS_UNCONSUMED_REQUESTS(ring))
++ more_to_do = 1;
++
++ /* Yield point for this unbounded loop. */
++ cond_resched();
++
++ return more_to_do;
++}
++
++
++int scsiback_schedule(void *data)
++{
++ struct vscsibk_info *info = (struct vscsibk_info *)data;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ while (!kthread_should_stop()) {
++ wait_event_interruptible(
++ info->wq,
++ info->waiting_reqs || kthread_should_stop());
++ wait_event_interruptible(
++ pending_free_wq,
++ !list_empty(&pending_free) || kthread_should_stop());
++
++ info->waiting_reqs = 0;
++ smp_mb();
++
++ if (scsiback_do_cmd_fn(info))
++ info->waiting_reqs = 1;
++ }
++
++ return 0;
++}
++
++
++static int __init scsiback_init(void)
++{
++ int i, mmap_pages;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ mmap_pages = vscsiif_reqs * VSCSIIF_SG_TABLESIZE;
++
++ pending_reqs = kmalloc(sizeof(pending_reqs[0]) *
++ vscsiif_reqs, GFP_KERNEL);
++ pending_grant_handles = kmalloc(sizeof(pending_grant_handles[0]) *
++ mmap_pages, GFP_KERNEL);
++ pending_pages = alloc_empty_pages_and_pagevec(mmap_pages);
++
++ if (!pending_reqs || !pending_grant_handles || !pending_pages)
++ goto out_of_memory;
++
++ for (i = 0; i < mmap_pages; i++)
++ pending_grant_handles[i] = SCSIBACK_INVALID_HANDLE;
++
++ if (scsiback_interface_init() < 0)
++ goto out_of_kmem;
++
++ memset(pending_reqs, 0, sizeof(pending_reqs));
++ INIT_LIST_HEAD(&pending_free);
++
++ for (i = 0; i < vscsiif_reqs; i++)
++ list_add_tail(&pending_reqs[i].free_list, &pending_free);
++
++ if (scsiback_xenbus_init())
++ goto out_of_xenbus;
++
++ scsiback_emulation_init();
++
++ return 0;
++
++out_of_xenbus:
++ scsiback_xenbus_unregister();
++out_of_kmem:
++ scsiback_interface_exit();
++out_of_memory:
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, mmap_pages);
++ printk(KERN_ERR "scsiback: %s: out of memory\n", __FUNCTION__);
++ return -ENOMEM;
++}
++
++static void __exit scsiback_exit(void)
++{
++ scsiback_xenbus_unregister();
++ scsiback_interface_exit();
++ kfree(pending_reqs);
++ kfree(pending_grant_handles);
++ free_empty_pages_and_pagevec(pending_pages, (vscsiif_reqs * VSCSIIF_SG_TABLESIZE));
++
++}
++
++module_init(scsiback_init);
++module_exit(scsiback_exit);
++
++MODULE_DESCRIPTION("Xen SCSI backend driver");
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/scsiback/translate.c b/drivers/xen/scsiback/translate.c
+new file mode 100644
+index 0000000..25510c5
+--- /dev/null
++++ b/drivers/xen/scsiback/translate.c
+@@ -0,0 +1,168 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/list.h>
++#include <linux/gfp.h>
++
++#include "common.h"
++
++/*
++ Initialize the translation entry list
++*/
++void scsiback_init_translation_table(struct vscsibk_info *info)
++{
++ INIT_LIST_HEAD(&info->v2p_entry_lists);
++ spin_lock_init(&info->v2p_lock);
++}
++
++
++/*
++ Add a new translation entry
++*/
++int scsiback_add_translation_entry(struct vscsibk_info *info,
++ struct scsi_device *sdev, struct ids_tuple *v)
++{
++ int err = 0;
++ struct v2p_entry *entry;
++ struct v2p_entry *new;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++
++ /* Check double assignment to identical virtual ID */
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ printk(KERN_WARNING "scsiback: Virtual ID is already used. "
++ "Assignment was not performed.\n");
++ err = -EEXIST;
++ goto out;
++ }
++
++ }
++
++ /* Create a new translation entry and add to the list */
++ if ((new = kmalloc(sizeof(struct v2p_entry), GFP_ATOMIC)) == NULL) {
++ printk(KERN_ERR "scsiback: %s: kmalloc() error.\n", __FUNCTION__);
++ err = -ENOMEM;
++ goto out;
++ }
++ new->v = *v;
++ new->sdev = sdev;
++ list_add_tail(&new->l, head);
++
++out:
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return err;
++}
++
++
++/*
++ Delete the translation entry specfied
++*/
++int scsiback_del_translation_entry(struct vscsibk_info *info,
++ struct ids_tuple *v)
++{
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ /* Find out the translation entry specified */
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ goto found;
++ }
++ }
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return 1;
++
++found:
++ /* Delete the translation entry specfied */
++ scsi_device_put(entry->sdev);
++ list_del(&entry->l);
++ kfree(entry);
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return 0;
++}
++
++
++/*
++ Perform virtual to physical translation
++*/
++struct scsi_device *scsiback_do_translation(struct vscsibk_info *info,
++ struct ids_tuple *v)
++{
++ struct v2p_entry *entry;
++ struct list_head *head = &(info->v2p_entry_lists);
++ struct scsi_device *sdev = NULL;
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry(entry, head, l) {
++ if ((entry->v.chn == v->chn) &&
++ (entry->v.tgt == v->tgt) &&
++ (entry->v.lun == v->lun)) {
++ sdev = entry->sdev;
++ goto out;
++ }
++ }
++out:
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return sdev;
++}
++
++
++/*
++ Release the translation entry specfied
++*/
++void scsiback_release_translation_entry(struct vscsibk_info *info)
++{
++ struct v2p_entry *entry, *tmp;
++ struct list_head *head = &(info->v2p_entry_lists);
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->v2p_lock, flags);
++ list_for_each_entry_safe(entry, tmp, head, l) {
++ scsi_device_put(entry->sdev);
++ list_del(&entry->l);
++ kfree(entry);
++ }
++
++ spin_unlock_irqrestore(&info->v2p_lock, flags);
++ return;
++
++}
+diff --git a/drivers/xen/scsiback/xenbus.c b/drivers/xen/scsiback/xenbus.c
+new file mode 100644
+index 0000000..31ffe92
+--- /dev/null
++++ b/drivers/xen/scsiback/xenbus.c
+@@ -0,0 +1,367 @@
++/*
++ * Xen SCSI backend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * Based on the blkback driver code.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <stdarg.h>
++#include <linux/module.h>
++#include <linux/kthread.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <scsi/scsi_device.h>
++
++#include "common.h"
++
++struct backend_info
++{
++ struct xenbus_device *dev;
++ struct vscsibk_info *info;
++};
++
++
++static int __vscsiif_name(struct backend_info *be, char *buf)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned int domid, id;
++
++ sscanf(dev->nodename, "backend/vscsi/%u/%u", &domid, &id);
++ snprintf(buf, TASK_COMM_LEN, "vscsi.%u.%u", be->info->domid, id);
++
++ return 0;
++}
++
++static int scsiback_map(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ int err;
++ char name[TASK_COMM_LEN];
++
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "reading %s ring", dev->otherend);
++ return err;
++ }
++
++ err = scsiback_init_sring(be->info, ring_ref, evtchn);
++ if (err)
++ return err;
++
++ err = __vscsiif_name(be, name);
++ if (err) {
++ xenbus_dev_error(dev, err, "get scsiback dev name");
++ return err;
++ }
++
++ be->info->kthread = kthread_run(scsiback_schedule, be->info, name);
++ if (IS_ERR(be->info->kthread)) {
++ err = PTR_ERR(be->info->kthread);
++ be->info->kthread = NULL;
++ xenbus_dev_error(be->dev, err, "start vscsiif");
++ return err;
++ }
++
++ return 0;
++}
++
++
++struct scsi_device *scsiback_get_scsi_device(struct ids_tuple *phy)
++{
++ struct Scsi_Host *shost;
++ struct scsi_device *sdev = NULL;
++
++ shost = scsi_host_lookup(phy->hst);
++ if (IS_ERR(shost)) {
++ printk(KERN_ERR "scsiback: host%d doesn't exist.\n",
++ phy->hst);
++ return NULL;
++ }
++ sdev = scsi_device_lookup(shost, phy->chn, phy->tgt, phy->lun);
++ if (!sdev) {
++ printk(KERN_ERR "scsiback: %d:%d:%d:%d doesn't exist.\n",
++ phy->hst, phy->chn, phy->tgt, phy->lun);
++ scsi_host_put(shost);
++ return NULL;
++ }
++
++ scsi_host_put(shost);
++ return (sdev);
++}
++
++#define VSCSIBACK_OP_ADD_OR_DEL_LUN 1
++#define VSCSIBACK_OP_UPDATEDEV_STATE 2
++
++
++static void scsiback_do_lun_hotplug(struct backend_info *be, int op)
++{
++ int i, err = 0;
++ struct ids_tuple phy, vir;
++ int device_state;
++ char str[64], state_str[64];
++ char **dir;
++ unsigned int dir_n = 0;
++ struct xenbus_device *dev = be->dev;
++ struct scsi_device *sdev;
++
++ dir = xenbus_directory(XBT_NIL, dev->nodename, "vscsi-devs", &dir_n);
++ if (IS_ERR(dir))
++ return;
++
++ for (i = 0; i < dir_n; i++) {
++
++ /* read status */
++ snprintf(state_str, sizeof(state_str), "vscsi-devs/%s/state", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, state_str, "%u",
++ &device_state);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
++
++ /* physical SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/p-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, str,
++ "%u:%u:%u:%u", &phy.hst, &phy.chn, &phy.tgt, &phy.lun);
++ if (XENBUS_EXIST_ERR(err)) {
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ continue;
++ }
++
++ /* virtual SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/v-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->nodename, str,
++ "%u:%u:%u:%u", &vir.hst, &vir.chn, &vir.tgt, &vir.lun);
++ if (XENBUS_EXIST_ERR(err)) {
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ continue;
++ }
++
++ switch (op) {
++ case VSCSIBACK_OP_ADD_OR_DEL_LUN:
++ if (device_state == XenbusStateInitialising) {
++ sdev = scsiback_get_scsi_device(&phy);
++ if (!sdev)
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ else {
++ err = scsiback_add_translation_entry(be->info, sdev, &vir);
++ if (!err) {
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateInitialised)) {
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ scsiback_del_translation_entry(be->info, &vir);
++ }
++ } else {
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ }
++ }
++ }
++
++ if (device_state == XenbusStateClosing) {
++ if (!scsiback_del_translation_entry(be->info, &vir)) {
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed))
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ }
++ }
++ break;
++
++ case VSCSIBACK_OP_UPDATEDEV_STATE:
++ if (device_state == XenbusStateInitialised) {
++ /* modify vscsi-devs/dev-x/state */
++ if (xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateConnected)) {
++ printk(KERN_ERR "scsiback: xenbus_printf error %s\n", state_str);
++ scsiback_del_translation_entry(be->info, &vir);
++ xenbus_printf(XBT_NIL, dev->nodename, state_str,
++ "%d", XenbusStateClosed);
++ }
++ }
++ break;
++ /*When it is necessary, processing is added here.*/
++ default:
++ break;
++ }
++ }
++
++ kfree(dir);
++ return ;
++}
++
++
++static void scsiback_frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ break;
++ case XenbusStateInitialised:
++ err = scsiback_map(be);
++ if (err)
++ break;
++
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_ADD_OR_DEL_LUN);
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ break;
++ case XenbusStateConnected:
++
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_UPDATEDEV_STATE);
++
++ if (dev->state == XenbusStateConnected)
++ break;
++
++ xenbus_switch_state(dev, XenbusStateConnected);
++
++ break;
++
++ case XenbusStateClosing:
++ scsiback_disconnect(be->info);
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ if (xenbus_dev_is_online(dev))
++ break;
++ /* fall through if not online */
++ case XenbusStateUnknown:
++ device_unregister(&dev->dev);
++ break;
++
++ case XenbusStateReconfiguring:
++ scsiback_do_lun_hotplug(be, VSCSIBACK_OP_ADD_OR_DEL_LUN);
++
++ xenbus_switch_state(dev, XenbusStateReconfigured);
++
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
++}
++
++
++static int scsiback_remove(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ if (be->info) {
++ scsiback_disconnect(be->info);
++ scsiback_release_translation_entry(be->info);
++ scsiback_free(be->info);
++ be->info = NULL;
++ }
++
++ kfree(be);
++ dev->dev.driver_data = NULL;
++
++ return 0;
++}
++
++
++static int scsiback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++
++ DPRINTK("%p %d\n", dev, dev->otherend_id);
++
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ be->info = vscsibk_info_alloc(dev->otherend_id);
++ if (IS_ERR(be->info)) {
++ err = PTR_ERR(be->info);
++ be->info = NULL;
++ xenbus_dev_fatal(dev, err, "creating scsihost interface");
++ goto fail;
++ }
++
++ be->info->dev = dev;
++ be->info->irq = 0;
++
++ scsiback_init_translation_table(be->info);
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err)
++ goto fail;
++
++ return 0;
++
++
++fail:
++ printk(KERN_WARNING "scsiback: %s failed\n",__FUNCTION__);
++ scsiback_remove(dev);
++
++ return err;
++}
++
++
++static struct xenbus_device_id scsiback_ids[] = {
++ { "vscsi" },
++ { "" }
++};
++
++static struct xenbus_driver scsiback = {
++ .name = "vscsi",
++ .ids = scsiback_ids,
++ .probe = scsiback_probe,
++ .remove = scsiback_remove,
++ .otherend_changed = scsiback_frontend_changed
++};
++
++int scsiback_xenbus_init(void)
++{
++ return xenbus_register_backend(&scsiback);
++}
++
++void scsiback_xenbus_unregister(void)
++{
++ xenbus_unregister_driver(&scsiback);
++}
+diff --git a/drivers/xen/scsifront/Makefile b/drivers/xen/scsifront/Makefile
+new file mode 100644
+index 0000000..58ee185
+--- /dev/null
++++ b/drivers/xen/scsifront/Makefile
+@@ -0,0 +1,3 @@
++
++obj-$(CONFIG_XEN_SCSI_FRONTEND) := xenscsi.o
++xenscsi-objs := scsifront.o xenbus.o
+diff --git a/drivers/xen/scsifront/common.h b/drivers/xen/scsifront/common.h
+new file mode 100644
+index 0000000..5b0cd8c
+--- /dev/null
++++ b/drivers/xen/scsifront/common.h
+@@ -0,0 +1,129 @@
++/*
++ * Xen SCSI frontend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_DRIVERS_SCSIFRONT_H__
++#define __XEN_DRIVERS_SCSIFRONT_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/kernel.h>
++#include <linux/device.h>
++#include <linux/kthread.h>
++#include <linux/wait.h>
++#include <linux/interrupt.h>
++#include <linux/spinlock.h>
++#include <linux/sched.h>
++#include <linux/blkdev.h>
++#include <scsi/scsi_cmnd.h>
++#include <scsi/scsi_device.h>
++#include <scsi/scsi.h>
++#include <scsi/scsi_host.h>
++#include <xen/xenbus.h>
++#include <xen/gnttab.h>
++#include <xen/evtchn.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/io/ring.h>
++#include <xen/interface/io/vscsiif.h>
++#include <asm/delay.h>
++
++
++#define GRANT_INVALID_REF 0
++#define VSCSI_IN_ABORT 1
++#define VSCSI_IN_RESET 2
++
++/* tuning point*/
++#define VSCSIIF_DEFAULT_CMD_PER_LUN 10
++#define VSCSIIF_MAX_TARGET 64
++#define VSCSIIF_MAX_LUN 255
++
++#define VSCSIIF_RING_SIZE \
++ __RING_SIZE((struct vscsiif_sring *)0, PAGE_SIZE)
++#define VSCSIIF_MAX_REQS VSCSIIF_RING_SIZE
++
++struct vscsifrnt_shadow {
++ uint16_t next_free;
++
++ /* command between backend and frontend
++ * VSCSIIF_ACT_SCSI_CDB or VSCSIIF_ACT_SCSI_RESET */
++ unsigned char act;
++
++ /* do reset function */
++ wait_queue_head_t wq_reset; /* reset work queue */
++ int wait_reset; /* reset work queue condition */
++ int32_t rslt_reset; /* reset response status */
++ /* (SUCESS or FAILED) */
++
++ /* for DMA_TO_DEVICE(1), DMA_FROM_DEVICE(2), DMA_NONE(3)
++ requests */
++ unsigned int sc_data_direction;
++
++ /* Number of pieces of scatter-gather */
++ unsigned int nr_segments;
++
++ /* requested struct scsi_cmnd is stored from kernel */
++ unsigned long req_scsi_cmnd;
++ int gref[VSCSIIF_SG_TABLESIZE];
++};
++
++struct vscsifrnt_info {
++ struct xenbus_device *dev;
++
++ struct Scsi_Host *host;
++
++ spinlock_t io_lock;
++ spinlock_t shadow_lock;
++ unsigned int evtchn;
++ unsigned int irq;
++
++ grant_ref_t ring_ref;
++ struct vscsiif_front_ring ring;
++ struct vscsiif_response ring_res;
++
++ struct vscsifrnt_shadow shadow[VSCSIIF_MAX_REQS];
++ uint32_t shadow_free;
++
++ struct task_struct *kthread;
++ wait_queue_head_t wq;
++ unsigned int waiting_resp;
++
++};
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++int scsifront_xenbus_init(void);
++void scsifront_xenbus_unregister(void);
++int scsifront_schedule(void *data);
++irqreturn_t scsifront_intr(int irq, void *dev_id);
++int scsifront_cmd_done(struct vscsifrnt_info *info);
++
++
++#endif /* __XEN_DRIVERS_SCSIFRONT_H__ */
+diff --git a/drivers/xen/scsifront/scsifront.c b/drivers/xen/scsifront/scsifront.c
+new file mode 100644
+index 0000000..a899885
+--- /dev/null
++++ b/drivers/xen/scsifront/scsifront.c
+@@ -0,0 +1,478 @@
++/*
++ * Xen SCSI frontend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++
++#include <linux/version.h>
++#include "common.h"
++
++static int get_id_from_freelist(struct vscsifrnt_info *info)
++{
++ unsigned long flags;
++ uint32_t free;
++
++ spin_lock_irqsave(&info->shadow_lock, flags);
++
++ free = info->shadow_free;
++ BUG_ON(free > VSCSIIF_MAX_REQS);
++ info->shadow_free = info->shadow[free].next_free;
++ info->shadow[free].next_free = 0x0fff;
++
++ info->shadow[free].wait_reset = 0;
++
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
++
++ return free;
++}
++
++static void add_id_to_freelist(struct vscsifrnt_info *info, uint32_t id)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->shadow_lock, flags);
++
++ info->shadow[id].next_free = info->shadow_free;
++ info->shadow[id].req_scsi_cmnd = 0;
++ info->shadow_free = id;
++
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
++}
++
++
++struct vscsiif_request * scsifront_pre_request(struct vscsifrnt_info *info)
++{
++ struct vscsiif_front_ring *ring = &(info->ring);
++ vscsiif_request_t *ring_req;
++ uint32_t id;
++
++ ring_req = RING_GET_REQUEST(&(info->ring), ring->req_prod_pvt);
++
++ ring->req_prod_pvt++;
++
++ id = get_id_from_freelist(info); /* use id by response */
++ ring_req->rqid = (uint16_t)id;
++
++ return ring_req;
++}
++
++
++static void scsifront_notify_work(struct vscsifrnt_info *info)
++{
++ info->waiting_resp = 1;
++ wake_up(&info->wq);
++}
++
++
++static void scsifront_do_request(struct vscsifrnt_info *info)
++{
++ struct vscsiif_front_ring *ring = &(info->ring);
++ unsigned int irq = info->irq;
++ int notify;
++
++ RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(ring, notify);
++ if (notify)
++ notify_remote_via_irq(irq);
++}
++
++irqreturn_t scsifront_intr(int irq, void *dev_id)
++{
++ scsifront_notify_work((struct vscsifrnt_info *)dev_id);
++ return IRQ_HANDLED;
++}
++
++
++static void scsifront_gnttab_done(struct vscsifrnt_shadow *s, uint32_t id)
++{
++ int i;
++
++ if (s->sc_data_direction == DMA_NONE)
++ return;
++
++ if (s->nr_segments) {
++ for (i = 0; i < s->nr_segments; i++) {
++ if (unlikely(gnttab_query_foreign_access(
++ s->gref[i]) != 0)) {
++ printk(KERN_ALERT "scsifront: "
++ "grant still in use by backend.\n");
++ BUG();
++ }
++ gnttab_end_foreign_access(s->gref[i], 0UL);
++ }
++ }
++
++ return;
++}
++
++
++static void scsifront_cdb_cmd_done(struct vscsifrnt_info *info,
++ vscsiif_response_t *ring_res)
++{
++ struct scsi_cmnd *sc;
++ uint32_t id;
++ uint8_t sense_len;
++
++ id = ring_res->rqid;
++ sc = (struct scsi_cmnd *)info->shadow[id].req_scsi_cmnd;
++
++ if (sc == NULL)
++ BUG();
++
++ scsifront_gnttab_done(&info->shadow[id], id);
++ add_id_to_freelist(info, id);
++
++ sc->result = ring_res->rslt;
++ scsi_set_resid(sc, ring_res->residual_len);
++
++ if (ring_res->sense_len > VSCSIIF_SENSE_BUFFERSIZE)
++ sense_len = VSCSIIF_SENSE_BUFFERSIZE;
++ else
++ sense_len = ring_res->sense_len;
++
++ if (sense_len)
++ memcpy(sc->sense_buffer, ring_res->sense_buffer, sense_len);
++
++ sc->scsi_done(sc);
++
++ return;
++}
++
++
++static void scsifront_sync_cmd_done(struct vscsifrnt_info *info,
++ vscsiif_response_t *ring_res)
++{
++ uint16_t id = ring_res->rqid;
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->shadow_lock, flags);
++ info->shadow[id].wait_reset = 1;
++ info->shadow[id].rslt_reset = ring_res->rslt;
++ spin_unlock_irqrestore(&info->shadow_lock, flags);
++
++ wake_up(&(info->shadow[id].wq_reset));
++}
++
++
++int scsifront_cmd_done(struct vscsifrnt_info *info)
++{
++ vscsiif_response_t *ring_res;
++
++ RING_IDX i, rp;
++ int more_to_do = 0;
++ unsigned long flags;
++
++ spin_lock_irqsave(&info->io_lock, flags);
++
++ rp = info->ring.sring->rsp_prod;
++ rmb();
++ for (i = info->ring.rsp_cons; i != rp; i++) {
++
++ ring_res = RING_GET_RESPONSE(&info->ring, i);
++
++ if (info->shadow[ring_res->rqid].act == VSCSIIF_ACT_SCSI_CDB)
++ scsifront_cdb_cmd_done(info, ring_res);
++ else
++ scsifront_sync_cmd_done(info, ring_res);
++ }
++
++ info->ring.rsp_cons = i;
++
++ if (i != info->ring.req_prod_pvt) {
++ RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
++ } else {
++ info->ring.sring->rsp_event = i + 1;
++ }
++
++ spin_unlock_irqrestore(&info->io_lock, flags);
++
++
++ /* Yield point for this unbounded loop. */
++ cond_resched();
++
++ return more_to_do;
++}
++
++
++
++
++int scsifront_schedule(void *data)
++{
++ struct vscsifrnt_info *info = (struct vscsifrnt_info *)data;
++
++ while (!kthread_should_stop()) {
++ wait_event_interruptible(
++ info->wq,
++ info->waiting_resp || kthread_should_stop());
++
++ info->waiting_resp = 0;
++ smp_mb();
++
++ if (scsifront_cmd_done(info))
++ info->waiting_resp = 1;
++ }
++
++ return 0;
++}
++
++
++
++static int map_data_for_request(struct vscsifrnt_info *info,
++ struct scsi_cmnd *sc, vscsiif_request_t *ring_req, uint32_t id)
++{
++ grant_ref_t gref_head;
++ struct page *page;
++ int err, ref, ref_cnt = 0;
++ int write = (sc->sc_data_direction == DMA_TO_DEVICE);
++ unsigned int i, nr_pages, off, len, bytes;
++ unsigned long buffer_pfn;
++
++ if (sc->sc_data_direction == DMA_NONE)
++ return 0;
++
++ err = gnttab_alloc_grant_references(VSCSIIF_SG_TABLESIZE, &gref_head);
++ if (err) {
++ printk(KERN_ERR "scsifront: gnttab_alloc_grant_references() error\n");
++ return -ENOMEM;
++ }
++
++ if (scsi_bufflen(sc)) {
++ /* quoted scsi_lib.c/scsi_req_map_sg . */
++ struct scatterlist *sg, *sgl = scsi_sglist(sc);
++ unsigned int data_len = scsi_bufflen(sc);
++
++ nr_pages = (data_len + sgl->offset + PAGE_SIZE - 1) >> PAGE_SHIFT;
++ if (nr_pages > VSCSIIF_SG_TABLESIZE) {
++ printk(KERN_ERR "scsifront: Unable to map request_buffer for command!\n");
++ ref_cnt = (-E2BIG);
++ goto big_to_sg;
++ }
++
++ for_each_sg (sgl, sg, scsi_sg_count(sc), i) {
++ page = sg_page(sg);
++ off = sg->offset;
++ len = sg->length;
++
++ buffer_pfn = page_to_phys(page) >> PAGE_SHIFT;
++
++ while (len > 0 && data_len > 0) {
++ /*
++ * sg sends a scatterlist that is larger than
++ * the data_len it wants transferred for certain
++ * IO sizes
++ */
++ bytes = min_t(unsigned int, len, PAGE_SIZE - off);
++ bytes = min(bytes, data_len);
++
++ ref = gnttab_claim_grant_reference(&gref_head);
++ BUG_ON(ref == -ENOSPC);
++
++ gnttab_grant_foreign_access_ref(ref, info->dev->otherend_id,
++ buffer_pfn, write);
++
++ info->shadow[id].gref[ref_cnt] = ref;
++ ring_req->seg[ref_cnt].gref = ref;
++ ring_req->seg[ref_cnt].offset = (uint16_t)off;
++ ring_req->seg[ref_cnt].length = (uint16_t)bytes;
++
++ buffer_pfn++;
++ len -= bytes;
++ data_len -= bytes;
++ off = 0;
++ ref_cnt++;
++ }
++ }
++ }
++
++big_to_sg:
++
++ gnttab_free_grant_references(gref_head);
++
++ return ref_cnt;
++}
++
++static int scsifront_queuecommand(struct scsi_cmnd *sc,
++ void (*done)(struct scsi_cmnd *))
++{
++ struct vscsifrnt_info *info =
++ (struct vscsifrnt_info *) sc->device->host->hostdata;
++ vscsiif_request_t *ring_req;
++ int ref_cnt;
++ uint16_t rqid;
++
++ if (RING_FULL(&info->ring)) {
++ goto out_host_busy;
++ }
++
++ sc->scsi_done = done;
++ sc->result = 0;
++
++ ring_req = scsifront_pre_request(info);
++ rqid = ring_req->rqid;
++ ring_req->act = VSCSIIF_ACT_SCSI_CDB;
++
++ ring_req->id = sc->device->id;
++ ring_req->lun = sc->device->lun;
++ ring_req->channel = sc->device->channel;
++ ring_req->cmd_len = sc->cmd_len;
++
++ BUG_ON(sc->cmd_len > VSCSIIF_MAX_COMMAND_SIZE);
++
++ if ( sc->cmd_len )
++ memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len);
++ else
++ memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE);
++
++ ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction;
++ ring_req->timeout_per_command = (sc->timeout_per_command / HZ);
++
++ info->shadow[rqid].req_scsi_cmnd = (unsigned long)sc;
++ info->shadow[rqid].sc_data_direction = sc->sc_data_direction;
++ info->shadow[rqid].act = ring_req->act;
++
++ ref_cnt = map_data_for_request(info, sc, ring_req, rqid);
++ if (ref_cnt < 0) {
++ add_id_to_freelist(info, rqid);
++ if (ref_cnt == (-ENOMEM))
++ goto out_host_busy;
++ else {
++ sc->result = (DID_ERROR << 16);
++ goto out_fail_command;
++ }
++ }
++
++ ring_req->nr_segments = (uint8_t)ref_cnt;
++ info->shadow[rqid].nr_segments = ref_cnt;
++
++ scsifront_do_request(info);
++
++ return 0;
++
++out_host_busy:
++ return SCSI_MLQUEUE_HOST_BUSY;
++
++out_fail_command:
++ done(sc);
++ return 0;
++}
++
++
++static int scsifront_eh_abort_handler(struct scsi_cmnd *sc)
++{
++ return (FAILED);
++}
++
++/* vscsi supports only device_reset, because it is each of LUNs */
++static int scsifront_dev_reset_handler(struct scsi_cmnd *sc)
++{
++ struct Scsi_Host *host = sc->device->host;
++ struct vscsifrnt_info *info =
++ (struct vscsifrnt_info *) sc->device->host->hostdata;
++
++ vscsiif_request_t *ring_req;
++ uint16_t rqid;
++ int err;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12)
++ spin_lock_irq(host->host_lock);
++#endif
++
++ ring_req = scsifront_pre_request(info);
++ ring_req->act = VSCSIIF_ACT_SCSI_RESET;
++
++ rqid = ring_req->rqid;
++ info->shadow[rqid].act = VSCSIIF_ACT_SCSI_RESET;
++
++ ring_req->channel = sc->device->channel;
++ ring_req->id = sc->device->id;
++ ring_req->lun = sc->device->lun;
++ ring_req->cmd_len = sc->cmd_len;
++
++ if ( sc->cmd_len )
++ memcpy(ring_req->cmnd, sc->cmnd, sc->cmd_len);
++ else
++ memset(ring_req->cmnd, 0, VSCSIIF_MAX_COMMAND_SIZE);
++
++ ring_req->sc_data_direction = (uint8_t)sc->sc_data_direction;
++ ring_req->timeout_per_command = (sc->timeout_per_command / HZ);
++ ring_req->nr_segments = 0;
++
++ scsifront_do_request(info);
++
++ spin_unlock_irq(host->host_lock);
++ wait_event_interruptible(info->shadow[rqid].wq_reset,
++ info->shadow[rqid].wait_reset);
++ spin_lock_irq(host->host_lock);
++
++ err = info->shadow[rqid].rslt_reset;
++
++ add_id_to_freelist(info, rqid);
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,12)
++ spin_unlock_irq(host->host_lock);
++#endif
++ return (err);
++}
++
++
++struct scsi_host_template scsifront_sht = {
++ .module = THIS_MODULE,
++ .name = "Xen SCSI frontend driver",
++ .queuecommand = scsifront_queuecommand,
++ .eh_abort_handler = scsifront_eh_abort_handler,
++ .eh_device_reset_handler= scsifront_dev_reset_handler,
++ .cmd_per_lun = VSCSIIF_DEFAULT_CMD_PER_LUN,
++ .can_queue = VSCSIIF_MAX_REQS,
++ .this_id = -1,
++ .sg_tablesize = VSCSIIF_SG_TABLESIZE,
++ .use_clustering = DISABLE_CLUSTERING,
++ .proc_name = "scsifront",
++};
++
++
++static int __init scsifront_init(void)
++{
++ int err;
++
++ if (!is_running_on_xen())
++ return -ENODEV;
++
++ err = scsifront_xenbus_init();
++
++ return err;
++}
++
++static void __exit scsifront_exit(void)
++{
++ scsifront_xenbus_unregister();
++}
++
++module_init(scsifront_init);
++module_exit(scsifront_exit);
++
++MODULE_DESCRIPTION("Xen SCSI frontend driver");
++MODULE_LICENSE("GPL");
+diff --git a/drivers/xen/scsifront/xenbus.c b/drivers/xen/scsifront/xenbus.c
+new file mode 100644
+index 0000000..fe92bfa
+--- /dev/null
++++ b/drivers/xen/scsifront/xenbus.c
+@@ -0,0 +1,420 @@
++/*
++ * Xen SCSI frontend driver
++ *
++ * Copyright (c) 2008, FUJITSU Limited
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++
++#include <linux/version.h>
++#include "common.h"
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,11)
++ #define DEFAULT_TASK_COMM_LEN 16
++#else
++ #define DEFAULT_TASK_COMM_LEN TASK_COMM_LEN
++#endif
++
++extern struct scsi_host_template scsifront_sht;
++
++static void scsifront_free(struct vscsifrnt_info *info)
++{
++ struct Scsi_Host *host = info->host;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,14)
++ if (host->shost_state != SHOST_DEL) {
++#else
++ if (!test_bit(SHOST_DEL, &host->shost_state)) {
++#endif
++ scsi_remove_host(info->host);
++ }
++
++ if (info->ring_ref != GRANT_INVALID_REF) {
++ gnttab_end_foreign_access(info->ring_ref,
++ (unsigned long)info->ring.sring);
++ info->ring_ref = GRANT_INVALID_REF;
++ info->ring.sring = NULL;
++ }
++
++ if (info->irq)
++ unbind_from_irqhandler(info->irq, info);
++ info->irq = 0;
++
++ scsi_host_put(info->host);
++}
++
++
++static int scsifront_alloc_ring(struct vscsifrnt_info *info)
++{
++ struct xenbus_device *dev = info->dev;
++ struct vscsiif_sring *sring;
++ int err = -ENOMEM;
++
++
++ info->ring_ref = GRANT_INVALID_REF;
++
++ /***** Frontend to Backend ring start *****/
++ sring = (struct vscsiif_sring *) __get_free_page(GFP_KERNEL);
++ if (!sring) {
++ xenbus_dev_fatal(dev, err, "fail to allocate shared ring (Front to Back)");
++ return err;
++ }
++ SHARED_RING_INIT(sring);
++ FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
++
++ err = xenbus_grant_ring(dev, virt_to_mfn(sring));
++ if (err < 0) {
++ free_page((unsigned long) sring);
++ info->ring.sring = NULL;
++ xenbus_dev_fatal(dev, err, "fail to grant shared ring (Front to Back)");
++ goto free_sring;
++ }
++ info->ring_ref = err;
++
++ err = bind_listening_port_to_irqhandler(
++ dev->otherend_id, scsifront_intr,
++ IRQF_SAMPLE_RANDOM, "scsifront", info);
++
++ if (err <= 0) {
++ xenbus_dev_fatal(dev, err, "bind_listening_port_to_irqhandler");
++ goto free_sring;
++ }
++ info->irq = err;
++
++ return 0;
++
++/* free resource */
++free_sring:
++ scsifront_free(info);
++
++ return err;
++}
++
++
++static int scsifront_init_ring(struct vscsifrnt_info *info)
++{
++ struct xenbus_device *dev = info->dev;
++ struct xenbus_transaction xbt;
++ int err;
++
++ DPRINTK("%s\n",__FUNCTION__);
++
++ err = scsifront_alloc_ring(info);
++ if (err)
++ return err;
++ DPRINTK("%u %u\n", info->ring_ref, info->evtchn);
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "starting transaction");
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "ring-ref", "%u",
++ info->ring_ref);
++ if (err) {
++ xenbus_dev_fatal(dev, err, "%s", "writing ring-ref");
++ goto fail;
++ }
++
++ err = xenbus_printf(xbt, dev->nodename, "event-channel", "%u",
++ irq_to_evtchn_port(info->irq));
++
++ if (err) {
++ xenbus_dev_fatal(dev, err, "%s", "writing event-channel");
++ goto fail;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err) {
++ if (err == -EAGAIN)
++ goto again;
++ xenbus_dev_fatal(dev, err, "completing transaction");
++ goto free_sring;
++ }
++
++ return 0;
++
++fail:
++ xenbus_transaction_end(xbt, 1);
++free_sring:
++ /* free resource */
++ scsifront_free(info);
++
++ return err;
++}
++
++
++static int scsifront_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ struct vscsifrnt_info *info;
++ struct Scsi_Host *host;
++ int i, err = -ENOMEM;
++ char name[DEFAULT_TASK_COMM_LEN];
++
++ host = scsi_host_alloc(&scsifront_sht, sizeof(*info));
++ if (!host) {
++ xenbus_dev_fatal(dev, err, "fail to allocate scsi host");
++ return err;
++ }
++ info = (struct vscsifrnt_info *) host->hostdata;
++ info->host = host;
++
++
++ dev->dev.driver_data = info;
++ info->dev = dev;
++
++ for (i = 0; i < VSCSIIF_MAX_REQS; i++) {
++ info->shadow[i].next_free = i + 1;
++ init_waitqueue_head(&(info->shadow[i].wq_reset));
++ info->shadow[i].wait_reset = 0;
++ }
++ info->shadow[VSCSIIF_MAX_REQS - 1].next_free = 0x0fff;
++
++ err = scsifront_init_ring(info);
++ if (err) {
++ scsi_host_put(host);
++ return err;
++ }
++
++ init_waitqueue_head(&info->wq);
++ spin_lock_init(&info->io_lock);
++ spin_lock_init(&info->shadow_lock);
++
++ snprintf(name, DEFAULT_TASK_COMM_LEN, "vscsiif.%d", info->host->host_no);
++
++ info->kthread = kthread_run(scsifront_schedule, info, name);
++ if (IS_ERR(info->kthread)) {
++ err = PTR_ERR(info->kthread);
++ info->kthread = NULL;
++ printk(KERN_ERR "scsifront: kthread start err %d\n", err);
++ goto free_sring;
++ }
++
++ host->max_id = VSCSIIF_MAX_TARGET;
++ host->max_channel = 0;
++ host->max_lun = VSCSIIF_MAX_LUN;
++ host->max_sectors = (VSCSIIF_SG_TABLESIZE - 1) * PAGE_SIZE / 512;
++
++ err = scsi_add_host(host, &dev->dev);
++ if (err) {
++ printk(KERN_ERR "scsifront: fail to add scsi host %d\n", err);
++ goto free_sring;
++ }
++
++ xenbus_switch_state(dev, XenbusStateInitialised);
++
++ return 0;
++
++free_sring:
++ /* free resource */
++ scsifront_free(info);
++ return err;
++}
++
++static int scsifront_remove(struct xenbus_device *dev)
++{
++ struct vscsifrnt_info *info = dev->dev.driver_data;
++
++ DPRINTK("%s: %s removed\n",__FUNCTION__ ,dev->nodename);
++
++ if (info->kthread) {
++ kthread_stop(info->kthread);
++ info->kthread = NULL;
++ }
++
++ scsifront_free(info);
++
++ return 0;
++}
++
++
++static int scsifront_disconnect(struct vscsifrnt_info *info)
++{
++ struct xenbus_device *dev = info->dev;
++ struct Scsi_Host *host = info->host;
++
++ DPRINTK("%s: %s disconnect\n",__FUNCTION__ ,dev->nodename);
++
++ /*
++ When this function is executed, all devices of
++ Frontend have been deleted.
++ Therefore, it need not block I/O before remove_host.
++ */
++
++ scsi_remove_host(host);
++ xenbus_frontend_closed(dev);
++
++ return 0;
++}
++
++#define VSCSIFRONT_OP_ADD_LUN 1
++#define VSCSIFRONT_OP_DEL_LUN 2
++
++static void scsifront_do_lun_hotplug(struct vscsifrnt_info *info, int op)
++{
++ struct xenbus_device *dev = info->dev;
++ int i, err = 0;
++ char str[64], state_str[64];
++ char **dir;
++ unsigned int dir_n = 0;
++ unsigned int device_state;
++ unsigned int hst, chn, tgt, lun;
++ struct scsi_device *sdev;
++
++ dir = xenbus_directory(XBT_NIL, dev->otherend, "vscsi-devs", &dir_n);
++ if (IS_ERR(dir))
++ return;
++
++ for (i = 0; i < dir_n; i++) {
++ /* read status */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/state", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->otherend, str, "%u",
++ &device_state);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
++
++ /* virtual SCSI device */
++ snprintf(str, sizeof(str), "vscsi-devs/%s/v-dev", dir[i]);
++ err = xenbus_scanf(XBT_NIL, dev->otherend, str,
++ "%u:%u:%u:%u", &hst, &chn, &tgt, &lun);
++ if (XENBUS_EXIST_ERR(err))
++ continue;
++
++ /* front device state path */
++ snprintf(state_str, sizeof(state_str), "vscsi-devs/%s/state", dir[i]);
++
++ switch (op) {
++ case VSCSIFRONT_OP_ADD_LUN:
++ if (device_state == XenbusStateInitialised) {
++ sdev = scsi_device_lookup(info->host, chn, tgt, lun);
++ if (sdev) {
++ printk(KERN_ERR "scsifront: Device already in use.\n");
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateClosed);
++ } else {
++ scsi_add_device(info->host, chn, tgt, lun);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateConnected);
++ }
++ }
++ break;
++ case VSCSIFRONT_OP_DEL_LUN:
++ if (device_state == XenbusStateClosing) {
++ sdev = scsi_device_lookup(info->host, chn, tgt, lun);
++ if (sdev) {
++ scsi_remove_device(sdev);
++ scsi_device_put(sdev);
++ xenbus_printf(XBT_NIL, dev->nodename,
++ state_str, "%d", XenbusStateClosed);
++ }
++ }
++ break;
++ default:
++ break;
++ }
++ }
++
++ kfree(dir);
++ return;
++}
++
++
++
++
++static void scsifront_backend_changed(struct xenbus_device *dev,
++ enum xenbus_state backend_state)
++{
++ struct vscsifrnt_info *info = dev->dev.driver_data;
++
++ DPRINTK("%p %u %u\n", dev, dev->state, backend_state);
++
++ switch (backend_state) {
++ case XenbusStateUnknown:
++ case XenbusStateInitialising:
++ case XenbusStateInitWait:
++ case XenbusStateClosed:
++ break;
++
++ case XenbusStateInitialised:
++ break;
++
++ case XenbusStateConnected:
++ if (xenbus_read_driver_state(dev->nodename) ==
++ XenbusStateInitialised) {
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN);
++ }
++
++ if (dev->state == XenbusStateConnected)
++ break;
++
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
++
++ case XenbusStateClosing:
++ scsifront_disconnect(info);
++ break;
++
++ case XenbusStateReconfiguring:
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_DEL_LUN);
++ xenbus_switch_state(dev, XenbusStateReconfiguring);
++ break;
++
++ case XenbusStateReconfigured:
++ scsifront_do_lun_hotplug(info, VSCSIFRONT_OP_ADD_LUN);
++ xenbus_switch_state(dev, XenbusStateConnected);
++ break;
++ }
++}
++
++
++static struct xenbus_device_id scsifront_ids[] = {
++ { "vscsi" },
++ { "" }
++};
++
++
++static struct xenbus_driver scsifront_driver = {
++ .name = "vscsi",
++ .ids = scsifront_ids,
++ .probe = scsifront_probe,
++ .remove = scsifront_remove,
++/* .resume = scsifront_resume, */
++ .otherend_changed = scsifront_backend_changed,
++};
++
++int scsifront_xenbus_init(void)
++{
++ return xenbus_register_frontend(&scsifront_driver);
++}
++
++void scsifront_xenbus_unregister(void)
++{
++ xenbus_unregister_driver(&scsifront_driver);
++}
++
+diff --git a/drivers/xen/sfc_netback/Makefile b/drivers/xen/sfc_netback/Makefile
+new file mode 100644
+index 0000000..f5c911b
+--- /dev/null
++++ b/drivers/xen/sfc_netback/Makefile
+@@ -0,0 +1,12 @@
++EXTRA_CFLAGS += -Idrivers/xen/sfc_netback -Idrivers/xen/sfc_netutil -Idrivers/xen/netback -Idrivers/net/sfc
++EXTRA_CFLAGS += -D__ci_driver__
++EXTRA_CFLAGS += -DEFX_USE_KCOMPAT
++EXTRA_CFLAGS += -Werror
++
++ifdef GCOV
++EXTRA_CFLAGS += -fprofile-arcs -ftest-coverage -DEFX_GCOV
++endif
++
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_BACKEND) := sfc_netback.o
++
++sfc_netback-objs := accel.o accel_fwd.o accel_msg.o accel_solarflare.o accel_xenbus.o accel_debugfs.o
+diff --git a/drivers/xen/sfc_netback/accel.c b/drivers/xen/sfc_netback/accel.c
+new file mode 100644
+index 0000000..8cc94ab
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel.c
+@@ -0,0 +1,129 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include "accel.h"
++#include "accel_msg_iface.h"
++#include "accel_solarflare.h"
++
++#include <linux/notifier.h>
++
++#ifdef EFX_GCOV
++#include "gcov.h"
++#endif
++
++static int netback_accel_netdev_event(struct notifier_block *nb,
++ unsigned long event, void *ptr)
++{
++ struct net_device *net_dev = (struct net_device *)ptr;
++ struct netback_accel *bend;
++
++ if ((event == NETDEV_UP) || (event == NETDEV_DOWN)) {
++ mutex_lock(&bend_list_mutex);
++ bend = bend_list;
++ while (bend != NULL) {
++ mutex_lock(&bend->bend_mutex);
++ /*
++ * This happens when the shared pages have
++ * been unmapped, but the bend not yet removed
++ * from list
++ */
++ if (bend->shared_page == NULL)
++ goto next;
++
++ if (bend->net_dev->ifindex == net_dev->ifindex)
++ netback_accel_set_interface_state
++ (bend, event == NETDEV_UP);
++
++ next:
++ mutex_unlock(&bend->bend_mutex);
++ bend = bend->next_bend;
++ }
++ mutex_unlock(&bend_list_mutex);
++ }
++
++ return NOTIFY_DONE;
++}
++
++
++static struct notifier_block netback_accel_netdev_notifier = {
++ .notifier_call = netback_accel_netdev_event,
++};
++
++
++unsigned sfc_netback_max_pages = NETBACK_ACCEL_DEFAULT_MAX_BUF_PAGES;
++module_param_named(max_pages, sfc_netback_max_pages, uint, 0644);
++MODULE_PARM_DESC(max_pages,
++ "The number of buffer pages to enforce on each guest");
++
++/* Initialise subsystems need for the accelerated fast path */
++static int __init netback_accel_init(void)
++{
++ int rc = 0;
++
++#ifdef EFX_GCOV
++ gcov_provider_init(THIS_MODULE);
++#endif
++
++ rc = netback_accel_init_fwd();
++
++ if (rc == 0)
++ netback_accel_debugfs_init();
++
++ if (rc == 0)
++ rc = netback_accel_sf_init();
++
++ if (rc == 0)
++ rc = register_netdevice_notifier
++ (&netback_accel_netdev_notifier);
++
++ /*
++ * What if no device was found, shouldn't we clean up stuff
++ * we've allocated for acceleration subsystem?
++ */
++
++ return rc;
++}
++
++module_init(netback_accel_init);
++
++static void __exit netback_accel_exit(void)
++{
++ unregister_netdevice_notifier(&netback_accel_netdev_notifier);
++
++ netback_accel_sf_shutdown();
++
++ netback_accel_shutdown_bends();
++
++ netback_accel_debugfs_fini();
++
++ netback_accel_shutdown_fwd();
++
++#ifdef EFX_GCOV
++ gcov_provider_fini(THIS_MODULE);
++#endif
++}
++
++module_exit(netback_accel_exit);
++
++MODULE_LICENSE("GPL");
+diff --git a/drivers/xen/sfc_netback/accel.h b/drivers/xen/sfc_netback/accel.h
+new file mode 100644
+index 0000000..f790d94
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel.h
+@@ -0,0 +1,393 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETBACK_ACCEL_H
++#define NETBACK_ACCEL_H
++
++#include <linux/slab.h>
++#include <linux/ip.h>
++#include <linux/tcp.h>
++#include <linux/udp.h>
++#include <linux/in.h>
++#include <linux/netdevice.h>
++#include <linux/etherdevice.h>
++#include <linux/mutex.h>
++#include <linux/wait.h>
++
++#include <xen/xenbus.h>
++
++#include "accel_shared_fifo.h"
++#include "accel_msg_iface.h"
++#include "accel_util.h"
++
++/**************************************************************************
++ * Datatypes
++ **************************************************************************/
++
++#define NETBACK_ACCEL_DEFAULT_MAX_FILTERS (8)
++#define NETBACK_ACCEL_DEFAULT_MAX_MCASTS (8)
++#define NETBACK_ACCEL_DEFAULT_MAX_BUF_PAGES (384)
++/* Variable to store module parameter for max_buf_pages */
++extern unsigned sfc_netback_max_pages;
++
++#define NETBACK_ACCEL_STATS 1
++
++#if NETBACK_ACCEL_STATS
++#define NETBACK_ACCEL_STATS_OP(x) x
++#else
++#define NETBACK_ACCEL_STATS_OP(x)
++#endif
++
++/*! Statistics for a given backend */
++struct netback_accel_stats {
++ /*! Number of eventq wakeup events */
++ u64 evq_wakeups;
++ /*! Number of eventq timeout events */
++ u64 evq_timeouts;
++ /*! Number of filters used */
++ u32 num_filters;
++ /*! Number of buffer pages registered */
++ u32 num_buffer_pages;
++};
++
++
++/* Debug fs nodes for each of the above stats */
++struct netback_accel_dbfs {
++ struct dentry *evq_wakeups;
++ struct dentry *evq_timeouts;
++ struct dentry *num_filters;
++ struct dentry *num_buffer_pages;
++};
++
++
++/*! Resource limits for a given NIC */
++struct netback_accel_limits {
++ int max_filters; /*!< Max. number of filters to use. */
++ int max_mcasts; /*!< Max. number of mcast subscriptions */
++ int max_buf_pages; /*!< Max. number of pages of NIC buffers */
++};
++
++
++/*! The state for an instance of the back end driver. */
++struct netback_accel {
++ /*! mutex to protect this state */
++ struct mutex bend_mutex;
++
++ /*! Watches on xenstore */
++ struct xenbus_watch domu_accel_watch;
++ struct xenbus_watch config_accel_watch;
++
++ /*! Pointer to whatever device cookie ties us in to the hypervisor */
++ void *hdev_data;
++
++ /*! FIFO indices. Next page is msg FIFOs */
++ struct net_accel_shared_page *shared_page;
++
++ /*! Defer control message processing */
++ struct work_struct handle_msg;
++
++ /*! Identifies other end VM and interface.*/
++ int far_end;
++ int vif_num;
++
++ /*!< To unmap the shared pages */
++ void *sh_pages_unmap;
++
++ /* Resource tracking */
++ /*! Limits on H/W & Dom0 resources */
++ struct netback_accel_limits quotas;
++
++ /* Hardware resources */
++ /*! The H/W type of associated NIC */
++ enum net_accel_hw_type hw_type;
++ /*! State of allocation */
++ int hw_state;
++ /*! Index into ci_driver.nics[] for this interface */
++ int nic_index;
++ /*! How to set up the acceleration for this hardware */
++ int (*accel_setup)(struct netback_accel *);
++ /*! And how to stop it. */
++ void (*accel_shutdown)(struct netback_accel *);
++
++ /*! The physical/real net_dev for this interface */
++ struct net_device *net_dev;
++
++ /*! Magic pointer to locate state in fowarding table */
++ void *fwd_priv;
++
++ /*! Message FIFO */
++ sh_msg_fifo2 to_domU;
++ /*! Message FIFO */
++ sh_msg_fifo2 from_domU;
++
++ /*! General notification channel id */
++ int msg_channel;
++ /*! General notification channel irq */
++ int msg_channel_irq;
++
++ /*! Event channel id dedicated to network packet interrupts. */
++ int net_channel;
++ /*! Event channel irq dedicated to network packets interrupts */
++ int net_channel_irq;
++
++ /*! The MAC address the frontend goes by. */
++ u8 mac[ETH_ALEN];
++ /*! Driver name of associated NIC */
++ char *nicname;
++
++ /*! Array of pointers to buffer pages mapped */
++ grant_handle_t *buffer_maps;
++ u64 *buffer_addrs;
++ /*! Index into buffer_maps */
++ int buffer_maps_index;
++ /*! Max number of pages that domU is allowed/will request to map */
++ int max_pages;
++
++ /*! Pointer to hardware specific private area */
++ void *accel_hw_priv;
++
++ /*! Wait queue for changes in accelstate. */
++ wait_queue_head_t state_wait_queue;
++
++ /*! Current state of the frontend according to the xenbus
++ * watch. */
++ XenbusState frontend_state;
++
++ /*! Current state of this backend. */
++ XenbusState backend_state;
++
++ /*! Non-zero if the backend is being removed. */
++ int removing;
++
++ /*! Non-zero if the setup_vnic has been called. */
++ int vnic_is_setup;
++
++#if NETBACK_ACCEL_STATS
++ struct netback_accel_stats stats;
++#endif
++#if defined(CONFIG_DEBUG_FS)
++ char *dbfs_dir_name;
++ struct dentry *dbfs_dir;
++ struct netback_accel_dbfs dbfs;
++#endif
++
++ /*! List */
++ struct netback_accel *next_bend;
++};
++
++
++/*
++ * Values for netback_accel.hw_state. States of resource allocation
++ * we can go through
++ */
++/*! No hardware has yet been allocated. */
++#define NETBACK_ACCEL_RES_NONE (0)
++/*! Hardware has been allocated. */
++#define NETBACK_ACCEL_RES_ALLOC (1)
++#define NETBACK_ACCEL_RES_FILTER (2)
++#define NETBACK_ACCEL_RES_HWINFO (3)
++
++/*! Filtering specification. This assumes that for VNIC support we
++ * will always want wildcard entries, so only specifies the
++ * destination IP/port
++ */
++struct netback_accel_filter_spec {
++ /*! Internal, used to access efx_vi API */
++ void *filter_handle;
++
++ /*! Destination IP in network order */
++ u32 destip_be;
++ /*! Destination port in network order */
++ u16 destport_be;
++ /*! Mac address */
++ u8 mac[ETH_ALEN];
++ /*! TCP or UDP */
++ u8 proto;
++};
++
++
++/**************************************************************************
++ * From accel.c
++ **************************************************************************/
++
++/*! \brief Start up all the acceleration plugins
++ *
++ * \return 0 on success, an errno on failure
++ */
++extern int netback_accel_init_accel(void);
++
++/*! \brief Shut down all the acceleration plugins
++ */
++extern void netback_accel_shutdown_accel(void);
++
++
++/**************************************************************************
++ * From accel_fwd.c
++ **************************************************************************/
++
++/*! \brief Init the forwarding infrastructure
++ * \return 0 on success, or -ENOMEM if it couldn't get memory for the
++ * forward table
++ */
++extern int netback_accel_init_fwd(void);
++
++/*! \brief Shut down the forwarding and free memory. */
++extern void netback_accel_shutdown_fwd(void);
++
++/*! Initialise each nic port's fowarding table */
++extern void *netback_accel_init_fwd_port(void);
++extern void netback_accel_shutdown_fwd_port(void *fwd_priv);
++
++/*! \brief Add an entry to the forwarding table.
++ * \param mac : MAC address, used as hash key
++ * \param ctxt : value to associate with key (can be NULL, see
++ * netback_accel_fwd_set_context)
++ * \return 0 on success, -ENOMEM if table was full and could no grow it
++ */
++extern int netback_accel_fwd_add(const __u8 *mac, void *context,
++ void *fwd_priv);
++
++/*! \brief Remove an entry from the forwarding table.
++ * \param mac : the MAC address to remove
++ * \return nothing: it is not an error if the mac was not in the table
++ */
++extern void netback_accel_fwd_remove(const __u8 *mac, void *fwd_priv);
++
++/*! \brief Set the context pointer for an existing fwd table entry.
++ * \param mac : key that is already present in the table
++ * \param context : new value to associate with key
++ * \return 0 on success, -ENOENT if mac not present in table.
++ */
++extern int netback_accel_fwd_set_context(const __u8 *mac, void *context,
++ void *fwd_priv);
++
++/**************************************************************************
++ * From accel_msg.c
++ **************************************************************************/
++
++
++/*! \brief Send the start-of-day message that handshakes with the VNIC
++ * and tells it its MAC address.
++ *
++ * \param bend The back end driver data structure
++ * \param version The version of communication to use, e.g. NET_ACCEL_MSG_VERSION
++ */
++extern void netback_accel_msg_tx_hello(struct netback_accel *bend,
++ unsigned version);
++
++/*! \brief Send a "there's a new local mac address" message
++ *
++ * \param bend The back end driver data structure for the vnic to send
++ * the message to
++ * \param mac Pointer to the new mac address
++ */
++extern void netback_accel_msg_tx_new_localmac(struct netback_accel *bend,
++ const void *mac);
++
++/*! \brief Send a "a mac address that was local has gone away" message
++ *
++ * \param bend The back end driver data structure for the vnic to send
++ * the message to
++ * \param mac Pointer to the old mac address
++ */
++extern void netback_accel_msg_tx_old_localmac(struct netback_accel *bend,
++ const void *mac);
++
++extern void netback_accel_set_interface_state(struct netback_accel *bend,
++ int up);
++
++/*! \brief Process the message queue for a bend that has just
++ * interrupted.
++ *
++ * Demultiplexs an interrupt from the front end driver, taking
++ * messages from the fifo and taking appropriate action.
++ *
++ * \param bend The back end driver data structure
++ */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++extern void netback_accel_msg_rx_handler(struct work_struct *arg);
++#else
++extern void netback_accel_msg_rx_handler(void *bend_void);
++#endif
++
++/**************************************************************************
++ * From accel_xenbus.c
++ **************************************************************************/
++/*! List of all the bends currently in existence. */
++extern struct netback_accel *bend_list;
++extern struct mutex bend_list_mutex;
++
++/*! \brief Probe a new network interface. */
++extern int netback_accel_probe(struct xenbus_device *dev);
++
++/*! \brief Remove a network interface. */
++extern int netback_accel_remove(struct xenbus_device *dev);
++
++/*! \brief Shutdown all accelerator backends */
++extern void netback_accel_shutdown_bends(void);
++
++/*! \brief Initiate the xenbus state teardown handshake */
++extern void netback_accel_set_closing(struct netback_accel *bend);
++
++/**************************************************************************
++ * From accel_debugfs.c
++ **************************************************************************/
++/*! Global statistics */
++struct netback_accel_global_stats {
++ /*! Number of TX packets seen through driverlink */
++ u64 dl_tx_packets;
++ /*! Number of TX packets seen through driverlink we didn't like */
++ u64 dl_tx_bad_packets;
++ /*! Number of RX packets seen through driverlink */
++ u64 dl_rx_packets;
++ /*! Number of mac addresses we are forwarding to */
++ u32 num_fwds;
++};
++
++/*! Debug fs entries for each of the above stats */
++struct netback_accel_global_dbfs {
++ struct dentry *dl_tx_packets;
++ struct dentry *dl_tx_bad_packets;
++ struct dentry *dl_rx_packets;
++ struct dentry *num_fwds;
++};
++
++#if NETBACK_ACCEL_STATS
++extern struct netback_accel_global_stats global_stats;
++#endif
++
++/*! \brief Initialise the debugfs root and populate with global stats */
++extern void netback_accel_debugfs_init(void);
++
++/*! \brief Remove our debugfs root directory */
++extern void netback_accel_debugfs_fini(void);
++
++/*! \brief Add per-bend statistics to debug fs */
++extern int netback_accel_debugfs_create(struct netback_accel *bend);
++/*! \brief Remove per-bend statistics from debug fs */
++extern int netback_accel_debugfs_remove(struct netback_accel *bend);
++
++#endif /* NETBACK_ACCEL_H */
++
++
+diff --git a/drivers/xen/sfc_netback/accel_debugfs.c b/drivers/xen/sfc_netback/accel_debugfs.c
+new file mode 100644
+index 0000000..6527c4b
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_debugfs.c
+@@ -0,0 +1,148 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/fs.h>
++#include <linux/debugfs.h>
++
++#include "accel.h"
++
++#if defined(CONFIG_DEBUG_FS)
++static struct dentry *sfc_debugfs_root = NULL;
++#endif
++
++#if NETBACK_ACCEL_STATS
++struct netback_accel_global_stats global_stats;
++#if defined(CONFIG_DEBUG_FS)
++static struct netback_accel_global_dbfs global_dbfs;
++#endif
++#endif
++
++void netback_accel_debugfs_init(void)
++{
++#if defined(CONFIG_DEBUG_FS)
++ sfc_debugfs_root = debugfs_create_dir("sfc_netback", NULL);
++ if (sfc_debugfs_root == NULL)
++ return;
++
++ global_dbfs.num_fwds = debugfs_create_u32
++ ("num_fwds", S_IRUSR | S_IRGRP | S_IROTH,
++ sfc_debugfs_root, &global_stats.num_fwds);
++ global_dbfs.dl_tx_packets = debugfs_create_u64
++ ("dl_tx_packets", S_IRUSR | S_IRGRP | S_IROTH,
++ sfc_debugfs_root, &global_stats.dl_tx_packets);
++ global_dbfs.dl_rx_packets = debugfs_create_u64
++ ("dl_rx_packets", S_IRUSR | S_IRGRP | S_IROTH,
++ sfc_debugfs_root, &global_stats.dl_rx_packets);
++ global_dbfs.dl_tx_bad_packets = debugfs_create_u64
++ ("dl_tx_bad_packets", S_IRUSR | S_IRGRP | S_IROTH,
++ sfc_debugfs_root, &global_stats.dl_tx_bad_packets);
++#endif
++}
++
++
++void netback_accel_debugfs_fini(void)
++{
++#if defined(CONFIG_DEBUG_FS)
++ debugfs_remove(global_dbfs.num_fwds);
++ debugfs_remove(global_dbfs.dl_tx_packets);
++ debugfs_remove(global_dbfs.dl_rx_packets);
++ debugfs_remove(global_dbfs.dl_tx_bad_packets);
++
++ debugfs_remove(sfc_debugfs_root);
++#endif
++}
++
++
++int netback_accel_debugfs_create(struct netback_accel *bend)
++{
++#if defined(CONFIG_DEBUG_FS)
++ /* Smallest length is 7 (vif0.0\n) */
++ int length = 7, temp;
++
++ if (sfc_debugfs_root == NULL)
++ return -ENOENT;
++
++ /* Work out length of string representation of far_end and vif_num */
++ temp = bend->far_end;
++ while (temp > 9) {
++ length++;
++ temp = temp / 10;
++ }
++ temp = bend->vif_num;
++ while (temp > 9) {
++ length++;
++ temp = temp / 10;
++ }
++
++ bend->dbfs_dir_name = kmalloc(length, GFP_KERNEL);
++ if (bend->dbfs_dir_name == NULL)
++ return -ENOMEM;
++ sprintf(bend->dbfs_dir_name, "vif%d.%d", bend->far_end, bend->vif_num);
++
++ bend->dbfs_dir = debugfs_create_dir(bend->dbfs_dir_name,
++ sfc_debugfs_root);
++ if (bend->dbfs_dir == NULL) {
++ kfree(bend->dbfs_dir_name);
++ return -ENOMEM;
++ }
++
++#if NETBACK_ACCEL_STATS
++ bend->dbfs.evq_wakeups = debugfs_create_u64
++ ("evq_wakeups", S_IRUSR | S_IRGRP | S_IROTH,
++ bend->dbfs_dir, &bend->stats.evq_wakeups);
++ bend->dbfs.evq_timeouts = debugfs_create_u64
++ ("evq_timeouts", S_IRUSR | S_IRGRP | S_IROTH,
++ bend->dbfs_dir, &bend->stats.evq_timeouts);
++ bend->dbfs.num_filters = debugfs_create_u32
++ ("num_filters", S_IRUSR | S_IRGRP | S_IROTH,
++ bend->dbfs_dir, &bend->stats.num_filters);
++ bend->dbfs.num_buffer_pages = debugfs_create_u32
++ ("num_buffer_pages", S_IRUSR | S_IRGRP | S_IROTH,
++ bend->dbfs_dir, &bend->stats.num_buffer_pages);
++#endif
++#endif
++ return 0;
++}
++
++
++int netback_accel_debugfs_remove(struct netback_accel *bend)
++{
++#if defined(CONFIG_DEBUG_FS)
++ if (bend->dbfs_dir != NULL) {
++#if NETBACK_ACCEL_STATS
++ debugfs_remove(bend->dbfs.evq_wakeups);
++ debugfs_remove(bend->dbfs.evq_timeouts);
++ debugfs_remove(bend->dbfs.num_filters);
++ debugfs_remove(bend->dbfs.num_buffer_pages);
++#endif
++ debugfs_remove(bend->dbfs_dir);
++ }
++
++ if (bend->dbfs_dir_name)
++ kfree(bend->dbfs_dir_name);
++#endif
++ return 0;
++}
++
++
+diff --git a/drivers/xen/sfc_netback/accel_fwd.c b/drivers/xen/sfc_netback/accel_fwd.c
+new file mode 100644
+index 0000000..88ecc61
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_fwd.c
+@@ -0,0 +1,425 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include "accel.h"
++#include "accel_cuckoo_hash.h"
++#include "accel_util.h"
++#include "accel_solarflare.h"
++
++#include "driverlink_api.h"
++
++#include <linux/if_arp.h>
++#include <linux/skbuff.h>
++#include <linux/list.h>
++
++/* State stored in the forward table */
++struct fwd_struct {
++ struct list_head link; /* Forms list */
++ void * context;
++ __u8 valid;
++ __u8 mac[ETH_ALEN];
++};
++
++/* Max value we support */
++#define NUM_FWDS_BITS 8
++#define NUM_FWDS (1 << NUM_FWDS_BITS)
++#define FWD_MASK (NUM_FWDS - 1)
++
++struct port_fwd {
++ /* Make a list */
++ struct list_head link;
++ /* Hash table to store the fwd_structs */
++ cuckoo_hash_table fwd_hash_table;
++ /* The array of fwd_structs */
++ struct fwd_struct *fwd_array;
++ /* Linked list of entries in use. */
++ struct list_head fwd_list;
++ /* Could do something clever with a reader/writer lock. */
++ spinlock_t fwd_lock;
++ /* Make find_free_entry() a bit faster by caching this */
++ int last_free_index;
++};
++
++/*
++ * This is unlocked as it's only called from dl probe and remove,
++ * which are themselves synchronised. Could get rid of it entirely as
++ * it's never iterated, but useful for debug
++ */
++static struct list_head port_fwds;
++
++
++/* Search the fwd_array for an unused entry */
++static int fwd_find_free_entry(struct port_fwd *fwd_set)
++{
++ int index = fwd_set->last_free_index;
++
++ do {
++ if (!fwd_set->fwd_array[index].valid) {
++ fwd_set->last_free_index = index;
++ return index;
++ }
++ index++;
++ if (index >= NUM_FWDS)
++ index = 0;
++ } while (index != fwd_set->last_free_index);
++
++ return -ENOMEM;
++}
++
++
++/* Look up a MAC in the hash table. Caller should hold table lock. */
++static inline struct fwd_struct *fwd_find_entry(const __u8 *mac,
++ struct port_fwd *fwd_set)
++{
++ cuckoo_hash_value value;
++ cuckoo_hash_mac_key key = cuckoo_mac_to_key(mac);
++
++ if (cuckoo_hash_lookup(&fwd_set->fwd_hash_table,
++ (cuckoo_hash_key *)(&key),
++ &value)) {
++ struct fwd_struct *fwd = &fwd_set->fwd_array[value];
++ DPRINTK_ON(memcmp(fwd->mac, mac, ETH_ALEN) != 0);
++ return fwd;
++ }
++
++ return NULL;
++}
++
++
++/* Initialise each nic port's fowarding table */
++void *netback_accel_init_fwd_port(void)
++{
++ struct port_fwd *fwd_set;
++
++ fwd_set = kzalloc(sizeof(struct port_fwd), GFP_KERNEL);
++ if (fwd_set == NULL) {
++ return NULL;
++ }
++
++ spin_lock_init(&fwd_set->fwd_lock);
++
++ fwd_set->fwd_array = kzalloc(sizeof (struct fwd_struct) * NUM_FWDS,
++ GFP_KERNEL);
++ if (fwd_set->fwd_array == NULL) {
++ kfree(fwd_set);
++ return NULL;
++ }
++
++ if (cuckoo_hash_init(&fwd_set->fwd_hash_table, NUM_FWDS_BITS, 8) != 0) {
++ kfree(fwd_set->fwd_array);
++ kfree(fwd_set);
++ return NULL;
++ }
++
++ INIT_LIST_HEAD(&fwd_set->fwd_list);
++
++ list_add(&fwd_set->link, &port_fwds);
++
++ return fwd_set;
++}
++
++
++void netback_accel_shutdown_fwd_port(void *fwd_priv)
++{
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++
++ BUG_ON(fwd_priv == NULL);
++
++ BUG_ON(list_empty(&port_fwds));
++ list_del(&fwd_set->link);
++
++ BUG_ON(!list_empty(&fwd_set->fwd_list));
++
++ cuckoo_hash_destroy(&fwd_set->fwd_hash_table);
++ kfree(fwd_set->fwd_array);
++ kfree(fwd_set);
++}
++
++
++int netback_accel_init_fwd()
++{
++ INIT_LIST_HEAD(&port_fwds);
++ return 0;
++}
++
++
++void netback_accel_shutdown_fwd()
++{
++ BUG_ON(!list_empty(&port_fwds));
++}
++
++
++/*
++ * Add an entry to the forwarding table. Returns -ENOMEM if no
++ * space.
++ */
++int netback_accel_fwd_add(const __u8 *mac, void *context, void *fwd_priv)
++{
++ struct fwd_struct *fwd;
++ int rc = 0, index;
++ unsigned long flags;
++ cuckoo_hash_mac_key key = cuckoo_mac_to_key(mac);
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++ DECLARE_MAC_BUF(buf);
++
++ BUG_ON(fwd_priv == NULL);
++
++ DPRINTK("Adding mac %s\n", print_mac(buf, mac));
++
++ spin_lock_irqsave(&fwd_set->fwd_lock, flags);
++
++ if ((rc = fwd_find_free_entry(fwd_set)) < 0 ) {
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++ return rc;
++ }
++
++ index = rc;
++
++ /* Shouldn't already be in the table */
++ if (cuckoo_hash_lookup(&fwd_set->fwd_hash_table,
++ (cuckoo_hash_key *)(&key), &rc) != 0) {
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++ EPRINTK("MAC address %s already accelerated.\n",
++ print_mac(buf, mac));
++ return -EEXIST;
++ }
++
++ if ((rc = cuckoo_hash_add(&fwd_set->fwd_hash_table,
++ (cuckoo_hash_key *)(&key), index, 1)) == 0) {
++ fwd = &fwd_set->fwd_array[index];
++ fwd->valid = 1;
++ fwd->context = context;
++ memcpy(fwd->mac, mac, ETH_ALEN);
++ list_add(&fwd->link, &fwd_set->fwd_list);
++ NETBACK_ACCEL_STATS_OP(global_stats.num_fwds++);
++ }
++
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++
++ /*
++ * No need to tell frontend that this mac address is local -
++ * it should auto-discover through packets on fastpath what is
++ * local and what is not, and just being on same server
++ * doesn't make it local (it could be on a different
++ * bridge)
++ */
++
++ return rc;
++}
++
++
++/* remove an entry from the forwarding tables. */
++void netback_accel_fwd_remove(const __u8 *mac, void *fwd_priv)
++{
++ struct fwd_struct *fwd;
++ unsigned long flags;
++ cuckoo_hash_mac_key key = cuckoo_mac_to_key(mac);
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++ DECLARE_MAC_BUF(buf);
++
++ DPRINTK("Removing mac %s\n", print_mac(buf, mac));
++
++ BUG_ON(fwd_priv == NULL);
++
++ spin_lock_irqsave(&fwd_set->fwd_lock, flags);
++
++ fwd = fwd_find_entry(mac, fwd_set);
++ if (fwd != NULL) {
++ BUG_ON(list_empty(&fwd_set->fwd_list));
++ list_del(&fwd->link);
++
++ fwd->valid = 0;
++ cuckoo_hash_remove(&fwd_set->fwd_hash_table,
++ (cuckoo_hash_key *)(&key));
++ NETBACK_ACCEL_STATS_OP(global_stats.num_fwds--);
++ }
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++
++ /*
++ * No need to tell frontend that this is no longer present -
++ * the frontend is currently only interested in remote
++ * addresses and it works these out (mostly) by itself
++ */
++}
++
++
++/* Set the context pointer for a hash table entry. */
++int netback_accel_fwd_set_context(const __u8 *mac, void *context,
++ void *fwd_priv)
++{
++ struct fwd_struct *fwd;
++ unsigned long flags;
++ int rc = -ENOENT;
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++
++ BUG_ON(fwd_priv == NULL);
++
++ spin_lock_irqsave(&fwd_set->fwd_lock, flags);
++ fwd = fwd_find_entry(mac, fwd_set);
++ if (fwd != NULL) {
++ fwd->context = context;
++ rc = 0;
++ }
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++ return rc;
++}
++
++
++/**************************************************************************
++ * Process a received packet
++ **************************************************************************/
++
++/*
++ * Returns whether or not we have a match in our forward table for the
++ * this skb. Must be called with appropriate fwd_lock already held
++ */
++static struct netback_accel *for_a_vnic(struct netback_pkt_buf *skb,
++ struct port_fwd *fwd_set)
++{
++ struct fwd_struct *fwd;
++ struct netback_accel *retval = NULL;
++
++ fwd = fwd_find_entry(skb->mac.raw, fwd_set);
++ if (fwd != NULL)
++ retval = fwd->context;
++ return retval;
++}
++
++
++static inline int packet_is_arp_reply(struct sk_buff *skb)
++{
++ return skb->protocol == ntohs(ETH_P_ARP)
++ && arp_hdr(skb)->ar_op == ntohs(ARPOP_REPLY);
++}
++
++
++static inline void hdr_to_filt(struct ethhdr *ethhdr, struct iphdr *ip,
++ struct netback_accel_filter_spec *spec)
++{
++ spec->proto = ip->protocol;
++ spec->destip_be = ip->daddr;
++ memcpy(spec->mac, ethhdr->h_source, ETH_ALEN);
++
++ if (ip->protocol == IPPROTO_TCP) {
++ struct tcphdr *tcp = (struct tcphdr *)((char *)ip + 4 * ip->ihl);
++ spec->destport_be = tcp->dest;
++ } else {
++ struct udphdr *udp = (struct udphdr *)((char *)ip + 4 * ip->ihl);
++ EPRINTK_ON(ip->protocol != IPPROTO_UDP);
++ spec->destport_be = udp->dest;
++ }
++}
++
++
++static inline int netback_accel_can_filter(struct netback_pkt_buf *skb)
++{
++ return (skb->protocol == htons(ETH_P_IP) &&
++ ((skb->nh.iph->protocol == IPPROTO_TCP) ||
++ (skb->nh.iph->protocol == IPPROTO_UDP)));
++}
++
++
++static inline void netback_accel_filter_packet(struct netback_accel *bend,
++ struct netback_pkt_buf *skb)
++{
++ struct netback_accel_filter_spec fs;
++ struct ethhdr *eh = (struct ethhdr *)(skb->mac.raw);
++
++ hdr_to_filt(eh, skb->nh.iph, &fs);
++
++ netback_accel_filter_check_add(bend, &fs);
++}
++
++
++/*
++ * Receive a packet and do something appropriate with it. Return true
++ * to take exclusive ownership of the packet. This is verging on
++ * solarflare specific
++ */
++void netback_accel_rx_packet(struct netback_pkt_buf *skb, void *fwd_priv)
++{
++ struct netback_accel *bend;
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++ unsigned long flags;
++
++ BUG_ON(fwd_priv == NULL);
++
++ /* Checking for bcast is cheaper so do that first */
++ if (is_broadcast_ether_addr(skb->mac.raw)) {
++ /* pass through the slow path by not claiming ownership */
++ return;
++ } else if (is_multicast_ether_addr(skb->mac.raw)) {
++ /* pass through the slow path by not claiming ownership */
++ return;
++ } else {
++ /* It is unicast */
++ spin_lock_irqsave(&fwd_set->fwd_lock, flags);
++ /* We insert filter to pass it off to a VNIC */
++ if ((bend = for_a_vnic(skb, fwd_set)) != NULL)
++ if (netback_accel_can_filter(skb))
++ netback_accel_filter_packet(bend, skb);
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++ }
++ return;
++}
++
++
++void netback_accel_tx_packet(struct sk_buff *skb, void *fwd_priv)
++{
++ __u8 *mac;
++ unsigned long flags;
++ struct port_fwd *fwd_set = (struct port_fwd *)fwd_priv;
++ struct fwd_struct *fwd;
++
++ BUG_ON(fwd_priv == NULL);
++
++ if (is_broadcast_ether_addr(skb_mac_header(skb))
++ && packet_is_arp_reply(skb)) {
++ DECLARE_MAC_BUF(buf);
++
++ /*
++ * update our fast path forwarding to reflect this
++ * gratuitous ARP
++ */
++ mac = skb_mac_header(skb)+ETH_ALEN;
++
++ DPRINTK("%s: found gratuitous ARP for %s\n",
++ __FUNCTION__, print_mac(buf, mac));
++
++ spin_lock_irqsave(&fwd_set->fwd_lock, flags);
++ /*
++ * Might not be local, but let's tell them all it is,
++ * and they can restore the fastpath if they continue
++ * to get packets that way
++ */
++ list_for_each_entry(fwd, &fwd_set->fwd_list, link) {
++ struct netback_accel *bend = fwd->context;
++ if (bend != NULL)
++ netback_accel_msg_tx_new_localmac(bend, mac);
++ }
++
++ spin_unlock_irqrestore(&fwd_set->fwd_lock, flags);
++ }
++ return;
++}
+diff --git a/drivers/xen/sfc_netback/accel_msg.c b/drivers/xen/sfc_netback/accel_msg.c
+new file mode 100644
+index 0000000..4ed8617
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_msg.c
+@@ -0,0 +1,392 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <xen/evtchn.h>
++
++#include "accel.h"
++#include "accel_msg_iface.h"
++#include "accel_util.h"
++#include "accel_solarflare.h"
++
++/* Send a HELLO to front end to start things off */
++void netback_accel_msg_tx_hello(struct netback_accel *bend, unsigned version)
++{
++ unsigned long lock_state;
++ struct net_accel_msg *msg =
++ net_accel_msg_start_send(bend->shared_page,
++ &bend->to_domU, &lock_state);
++ /* The queue _cannot_ be full, we're the first users. */
++ EPRINTK_ON(msg == NULL);
++
++ if (msg != NULL) {
++ net_accel_msg_init(msg, NET_ACCEL_MSG_HELLO);
++ msg->u.hello.version = version;
++ msg->u.hello.max_pages = bend->quotas.max_buf_pages;
++ VPRINTK("Sending hello to channel %d\n", bend->msg_channel);
++ net_accel_msg_complete_send_notify(bend->shared_page,
++ &bend->to_domU,
++ &lock_state,
++ bend->msg_channel_irq);
++ }
++}
++
++/* Send a local mac message to vnic */
++static void netback_accel_msg_tx_localmac(struct netback_accel *bend,
++ int type, const void *mac)
++{
++ unsigned long lock_state;
++ struct net_accel_msg *msg;
++ DECLARE_MAC_BUF(buf);
++
++ BUG_ON(bend == NULL || mac == NULL);
++
++ VPRINTK("Sending local mac message: %s\n", print_mac(buf, mac));
++
++ msg = net_accel_msg_start_send(bend->shared_page, &bend->to_domU,
++ &lock_state);
++
++ if (msg != NULL) {
++ net_accel_msg_init(msg, NET_ACCEL_MSG_LOCALMAC);
++ msg->u.localmac.flags = type;
++ memcpy(msg->u.localmac.mac, mac, ETH_ALEN);
++ net_accel_msg_complete_send_notify(bend->shared_page,
++ &bend->to_domU,
++ &lock_state,
++ bend->msg_channel_irq);
++ } else {
++ /*
++ * TODO if this happens we may leave a domU
++ * fastpathing packets when they should be delivered
++ * locally. Solution is get domU to timeout entries
++ * in its fastpath lookup table when it receives no RX
++ * traffic
++ */
++ EPRINTK("%s: saw full queue, may need ARP timer to recover\n",
++ __FUNCTION__);
++ }
++}
++
++/* Send an add local mac message to vnic */
++void netback_accel_msg_tx_new_localmac(struct netback_accel *bend,
++ const void *mac)
++{
++ netback_accel_msg_tx_localmac(bend, NET_ACCEL_MSG_ADD, mac);
++}
++
++
++static int netback_accel_msg_rx_buffer_map(struct netback_accel *bend,
++ struct net_accel_msg *msg)
++{
++ int log2_pages, rc;
++
++ /* Can only allocate in power of two */
++ log2_pages = log2_ge(msg->u.mapbufs.pages, 0);
++ if (msg->u.mapbufs.pages != pow2(log2_pages)) {
++ EPRINTK("%s: Can only alloc bufs in power of 2 sizes (%d)\n",
++ __FUNCTION__, msg->u.mapbufs.pages);
++ rc = -EINVAL;
++ goto err_out;
++ }
++
++ /*
++ * Sanity. Assumes NET_ACCEL_MSG_MAX_PAGE_REQ is same for
++ * both directions/domains
++ */
++ if (msg->u.mapbufs.pages > NET_ACCEL_MSG_MAX_PAGE_REQ) {
++ EPRINTK("%s: too many pages in a single message: %d %d\n",
++ __FUNCTION__, msg->u.mapbufs.pages,
++ NET_ACCEL_MSG_MAX_PAGE_REQ);
++ rc = -EINVAL;
++ goto err_out;
++ }
++
++ if ((rc = netback_accel_add_buffers(bend, msg->u.mapbufs.pages,
++ log2_pages, msg->u.mapbufs.grants,
++ &msg->u.mapbufs.buf)) < 0) {
++ goto err_out;
++ }
++
++ msg->id |= NET_ACCEL_MSG_REPLY;
++
++ return 0;
++
++ err_out:
++ EPRINTK("%s: err_out\n", __FUNCTION__);
++ msg->id |= NET_ACCEL_MSG_ERROR | NET_ACCEL_MSG_REPLY;
++ return rc;
++}
++
++
++/* Hint from frontend that one of our filters is out of date */
++static int netback_accel_process_fastpath(struct netback_accel *bend,
++ struct net_accel_msg *msg)
++{
++ struct netback_accel_filter_spec spec;
++
++ if (msg->u.fastpath.flags & NET_ACCEL_MSG_REMOVE) {
++ /*
++ * Would be nice to BUG() this but would leave us
++ * vulnerable to naughty frontend
++ */
++ EPRINTK_ON(msg->u.fastpath.flags & NET_ACCEL_MSG_ADD);
++
++ memcpy(spec.mac, msg->u.fastpath.mac, ETH_ALEN);
++ spec.destport_be = msg->u.fastpath.port;
++ spec.destip_be = msg->u.fastpath.ip;
++ spec.proto = msg->u.fastpath.proto;
++
++ netback_accel_filter_remove_spec(bend, &spec);
++ }
++
++ return 0;
++}
++
++
++/* Flow control for message queues */
++inline void set_queue_not_full(struct netback_accel *bend)
++{
++ if (!test_and_set_bit(NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL_B,
++ (unsigned long *)&bend->shared_page->aflags))
++ notify_remote_via_irq(bend->msg_channel_irq);
++ else
++ VPRINTK("queue not full bit already set, not signalling\n");
++}
++
++
++/* Flow control for message queues */
++inline void set_queue_full(struct netback_accel *bend)
++{
++ if (!test_and_set_bit(NET_ACCEL_MSG_AFLAGS_QUEUE0FULL_B,
++ (unsigned long *)&bend->shared_page->aflags))
++ notify_remote_via_irq(bend->msg_channel_irq);
++ else
++ VPRINTK("queue full bit already set, not signalling\n");
++}
++
++
++void netback_accel_set_interface_state(struct netback_accel *bend, int up)
++{
++ bend->shared_page->net_dev_up = up;
++ if (!test_and_set_bit(NET_ACCEL_MSG_AFLAGS_NETUPDOWN_B,
++ (unsigned long *)&bend->shared_page->aflags))
++ notify_remote_via_irq(bend->msg_channel_irq);
++ else
++ VPRINTK("interface up/down bit already set, not signalling\n");
++}
++
++
++static int check_rx_hello_version(unsigned version)
++{
++ /* Should only happen if there's been a version mismatch */
++ BUG_ON(version == NET_ACCEL_MSG_VERSION);
++
++ if (version > NET_ACCEL_MSG_VERSION) {
++ /* Newer protocol, we must refuse */
++ return -EPROTO;
++ }
++
++ if (version < NET_ACCEL_MSG_VERSION) {
++ /*
++ * We are newer, so have discretion to accept if we
++ * wish. For now however, just reject
++ */
++ return -EPROTO;
++ }
++
++ return -EINVAL;
++}
++
++
++static int process_rx_msg(struct netback_accel *bend,
++ struct net_accel_msg *msg)
++{
++ int err = 0;
++
++ switch (msg->id) {
++ case NET_ACCEL_MSG_REPLY | NET_ACCEL_MSG_HELLO:
++ /* Reply to a HELLO; mark ourselves as connected */
++ DPRINTK("got Hello reply, version %.8x\n",
++ msg->u.hello.version);
++
++ /*
++ * Check that we've not successfully done this
++ * already. NB no check at the moment that this reply
++ * comes after we've actually sent a HELLO as that's
++ * not possible with the current code structure
++ */
++ if (bend->hw_state != NETBACK_ACCEL_RES_NONE)
++ return -EPROTO;
++
++ /* Store max_pages for accel_setup */
++ if (msg->u.hello.max_pages > bend->quotas.max_buf_pages) {
++ EPRINTK("More pages than quota allows (%d > %d)\n",
++ msg->u.hello.max_pages,
++ bend->quotas.max_buf_pages);
++ /* Force it down to the quota */
++ msg->u.hello.max_pages = bend->quotas.max_buf_pages;
++ }
++ bend->max_pages = msg->u.hello.max_pages;
++
++ /* Set up the hardware visible to the other end */
++ err = bend->accel_setup(bend);
++ if (err) {
++ /* This is fatal */
++ DPRINTK("Hello gave accel_setup error %d\n", err);
++ netback_accel_set_closing(bend);
++ } else {
++ /*
++ * Now add the context so that packet
++ * forwarding will commence
++ */
++ netback_accel_fwd_set_context(bend->mac, bend,
++ bend->fwd_priv);
++ }
++ break;
++ case NET_ACCEL_MSG_REPLY | NET_ACCEL_MSG_HELLO | NET_ACCEL_MSG_ERROR:
++ EPRINTK("got Hello error, versions us:%.8x them:%.8x\n",
++ NET_ACCEL_MSG_VERSION, msg->u.hello.version);
++
++ if (bend->hw_state != NETBACK_ACCEL_RES_NONE)
++ return -EPROTO;
++
++ if (msg->u.hello.version != NET_ACCEL_MSG_VERSION) {
++ /* Error is due to version mismatch */
++ err = check_rx_hello_version(msg->u.hello.version);
++ if (err == 0) {
++ /*
++ * It's OK to be compatible, send
++ * another hello with compatible version
++ */
++ netback_accel_msg_tx_hello
++ (bend, msg->u.hello.version);
++ } else {
++ /*
++ * Tell frontend that we're not going to
++ * send another HELLO by going to Closing.
++ */
++ netback_accel_set_closing(bend);
++ }
++ }
++ break;
++ case NET_ACCEL_MSG_MAPBUF:
++ VPRINTK("Got mapped buffers request %d\n",
++ msg->u.mapbufs.reqid);
++
++ if (bend->hw_state == NETBACK_ACCEL_RES_NONE)
++ return -EPROTO;
++
++ /*
++ * Frontend wants a buffer table entry for the
++ * supplied pages
++ */
++ err = netback_accel_msg_rx_buffer_map(bend, msg);
++ if (net_accel_msg_reply_notify(bend->shared_page,
++ bend->msg_channel_irq,
++ &bend->to_domU, msg)) {
++ /*
++ * This is fatal as we can't tell the frontend
++ * about the problem through the message
++ * queue, and so would otherwise stalemate
++ */
++ netback_accel_set_closing(bend);
++ }
++ break;
++ case NET_ACCEL_MSG_FASTPATH:
++ DPRINTK("Got fastpath request\n");
++
++ if (bend->hw_state == NETBACK_ACCEL_RES_NONE)
++ return -EPROTO;
++
++ err = netback_accel_process_fastpath(bend, msg);
++ break;
++ default:
++ EPRINTK("Huh? Message code is %x\n", msg->id);
++ err = -EPROTO;
++ break;
++ }
++ return err;
++}
++
++
++/* Demultiplex an IRQ from the frontend driver. */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++void netback_accel_msg_rx_handler(struct work_struct *arg)
++#else
++void netback_accel_msg_rx_handler(void *bend_void)
++#endif
++{
++ struct net_accel_msg msg;
++ int err, queue_was_full = 0;
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ struct netback_accel *bend =
++ container_of(arg, struct netback_accel, handle_msg);
++#else
++ struct netback_accel *bend = (struct netback_accel *)bend_void;
++#endif
++
++ mutex_lock(&bend->bend_mutex);
++
++ /*
++ * This happens when the shared pages have been unmapped, but
++ * the workqueue not flushed yet
++ */
++ if (bend->shared_page == NULL)
++ goto done;
++
++ if ((bend->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_TO_DOM0_MASK) != 0) {
++ if (bend->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL) {
++ /* We've been told there may now be space. */
++ clear_bit(NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL_B,
++ (unsigned long *)&bend->shared_page->aflags);
++ }
++
++ if (bend->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_QUEUEUFULL) {
++ clear_bit(NET_ACCEL_MSG_AFLAGS_QUEUEUFULL_B,
++ (unsigned long *)&bend->shared_page->aflags);
++ queue_was_full = 1;
++ }
++ }
++
++ while ((err = net_accel_msg_recv(bend->shared_page, &bend->from_domU,
++ &msg)) == 0) {
++ err = process_rx_msg(bend, &msg);
++
++ if (err != 0) {
++ EPRINTK("%s: Error %d\n", __FUNCTION__, err);
++ goto err;
++ }
++ }
++
++ err:
++ /* There will be space now if we can make any. */
++ if (queue_was_full)
++ set_queue_not_full(bend);
++ done:
++ mutex_unlock(&bend->bend_mutex);
++
++ return;
++}
+diff --git a/drivers/xen/sfc_netback/accel_solarflare.c b/drivers/xen/sfc_netback/accel_solarflare.c
+new file mode 100644
+index 0000000..e398a96
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_solarflare.c
+@@ -0,0 +1,1253 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include "common.h"
++
++#include "accel.h"
++#include "accel_solarflare.h"
++#include "accel_msg_iface.h"
++#include "accel_util.h"
++
++#include "accel_cuckoo_hash.h"
++
++#include "ci/driver/resource/efx_vi.h"
++
++#include "ci/efrm/nic_table.h"
++#include "ci/efhw/public.h"
++
++#include <xen/evtchn.h>
++#include <xen/driver_util.h>
++#include <linux/list.h>
++#include <linux/mutex.h>
++
++#include "driverlink_api.h"
++
++#define SF_XEN_RX_USR_BUF_SIZE 2048
++
++struct falcon_bend_accel_priv {
++ struct efx_vi_state *efx_vih;
++
++ /*! Array of pointers to dma_map state, used so VNIC can
++ * request their removal in a single message
++ */
++ struct efx_vi_dma_map_state **dma_maps;
++ /*! Index into dma_maps */
++ int dma_maps_index;
++
++ /*! Serialises access to filters */
++ spinlock_t filter_lock;
++ /*! Bitmap of which filters are free */
++ unsigned long free_filters;
++ /*! Used for index normalisation */
++ u32 filter_idx_mask;
++ struct netback_accel_filter_spec *fspecs;
++ cuckoo_hash_table filter_hash_table;
++
++ u32 txdmaq_gnt;
++ u32 rxdmaq_gnt;
++ u32 doorbell_gnt;
++ u32 evq_rptr_gnt;
++ u32 evq_mem_gnts[EF_HW_FALCON_EVQ_PAGES];
++ u32 evq_npages;
++};
++
++/* Forward declaration */
++static int netback_accel_filter_init(struct netback_accel *);
++static void netback_accel_filter_shutdown(struct netback_accel *);
++
++/**************************************************************************
++ *
++ * Driverlink stuff
++ *
++ **************************************************************************/
++
++struct driverlink_port {
++ struct list_head link;
++ enum net_accel_hw_type type;
++ struct net_device *net_dev;
++ struct efx_dl_device *efx_dl_dev;
++ int nic_index;
++ void *fwd_priv;
++};
++
++static struct list_head dl_ports;
++
++/* This mutex protects global state, such as the dl_ports list */
++DEFINE_MUTEX(accel_mutex);
++
++static int init_done = 0;
++
++/* The DL callbacks */
++
++
++#if defined(EFX_USE_FASTCALL)
++static enum efx_veto fastcall
++#else
++static enum efx_veto
++#endif
++bend_dl_tx_packet(struct efx_dl_device *efx_dl_dev,
++ struct sk_buff *skb)
++{
++ struct driverlink_port *port = efx_dl_dev->priv;
++
++ BUG_ON(port == NULL);
++
++ NETBACK_ACCEL_STATS_OP(global_stats.dl_tx_packets++);
++ if (skb_mac_header_was_set(skb))
++ netback_accel_tx_packet(skb, port->fwd_priv);
++ else {
++ DPRINTK("Ignoring packet with missing mac address\n");
++ NETBACK_ACCEL_STATS_OP(global_stats.dl_tx_bad_packets++);
++ }
++ return EFX_ALLOW_PACKET;
++}
++
++/* EFX_USE_FASTCALL */
++#if defined(EFX_USE_FASTCALL)
++static enum efx_veto fastcall
++#else
++static enum efx_veto
++#endif
++bend_dl_rx_packet(struct efx_dl_device *efx_dl_dev,
++ const char *pkt_buf, int pkt_len)
++{
++ struct driverlink_port *port = efx_dl_dev->priv;
++ struct netback_pkt_buf pkt;
++ struct ethhdr *eh;
++
++ BUG_ON(port == NULL);
++
++ pkt.mac.raw = (char *)pkt_buf;
++ pkt.nh.raw = (char *)pkt_buf + ETH_HLEN;
++ eh = (struct ethhdr *)pkt_buf;
++ pkt.protocol = eh->h_proto;
++
++ NETBACK_ACCEL_STATS_OP(global_stats.dl_rx_packets++);
++ netback_accel_rx_packet(&pkt, port->fwd_priv);
++ return EFX_ALLOW_PACKET;
++}
++
++
++/* Callbacks we'd like to get from the netdriver through driverlink */
++struct efx_dl_callbacks bend_dl_callbacks =
++ {
++ .tx_packet = bend_dl_tx_packet,
++ .rx_packet = bend_dl_rx_packet,
++ };
++
++
++static struct netback_accel_hooks accel_hooks = {
++ THIS_MODULE,
++ &netback_accel_probe,
++ &netback_accel_remove
++};
++
++
++/*
++ * Handy helper which given an efx_dl_device works out which
++ * efab_nic_t index into efrm_nic_table.nics[] it corresponds to
++ */
++static int efx_device_to_efab_nic_index(struct efx_dl_device *efx_dl_dev)
++{
++ int i;
++
++ for (i = 0; i < EFHW_MAX_NR_DEVS; i++) {
++ struct efhw_nic *nic = efrm_nic_table.nic[i];
++
++ /*
++ * It's possible for the nic structure to have not
++ * been initialised if the resource driver failed its
++ * driverlink probe
++ */
++ if (nic == NULL || nic->net_driver_dev == NULL)
++ continue;
++
++ /* Work out if these are talking about the same NIC */
++ if (nic->net_driver_dev->pci_dev == efx_dl_dev->pci_dev)
++ return i;
++ }
++
++ return -1;
++}
++
++
++/* Driver link probe - register our callbacks */
++static int bend_dl_probe(struct efx_dl_device *efx_dl_dev,
++ const struct net_device *net_dev,
++ const struct efx_dl_device_info *dev_info,
++ const char* silicon_rev)
++{
++ int rc;
++ enum net_accel_hw_type type;
++ struct driverlink_port *port;
++
++ DPRINTK("%s: %s\n", __FUNCTION__, silicon_rev);
++
++ if (strcmp(silicon_rev, "falcon/a1") == 0)
++ type = NET_ACCEL_MSG_HWTYPE_FALCON_A;
++ else if (strcmp(silicon_rev, "falcon/b0") == 0)
++ type = NET_ACCEL_MSG_HWTYPE_FALCON_B;
++ else {
++ EPRINTK("%s: unsupported silicon %s\n", __FUNCTION__,
++ silicon_rev);
++ rc = -EINVAL;
++ goto fail1;
++ }
++
++ port = kmalloc(sizeof(struct driverlink_port), GFP_KERNEL);
++ if (port == NULL) {
++ EPRINTK("%s: no memory for dl probe\n", __FUNCTION__);
++ rc = -ENOMEM;
++ goto fail1;
++ }
++
++ port->efx_dl_dev = efx_dl_dev;
++ efx_dl_dev->priv = port;
++
++ port->nic_index = efx_device_to_efab_nic_index(efx_dl_dev);
++ if (port->nic_index < 0) {
++ /*
++ * This can happen in theory if the resource driver
++ * failed to initialise properly
++ */
++ EPRINTK("%s: nic structure not found\n", __FUNCTION__);
++ rc = -EINVAL;
++ goto fail2;
++ }
++
++ port->fwd_priv = netback_accel_init_fwd_port();
++ if (port->fwd_priv == NULL) {
++ EPRINTK("%s: failed to set up forwarding for port\n",
++ __FUNCTION__);
++ rc = -ENOMEM;
++ goto fail2;
++ }
++
++ rc = efx_dl_register_callbacks(efx_dl_dev, &bend_dl_callbacks);
++ if (rc != 0) {
++ EPRINTK("%s: register_callbacks failed\n", __FUNCTION__);
++ goto fail3;
++ }
++
++ port->type = type;
++ port->net_dev = (struct net_device *)net_dev;
++
++ mutex_lock(&accel_mutex);
++ list_add(&port->link, &dl_ports);
++ mutex_unlock(&accel_mutex);
++
++ rc = netback_connect_accelerator(NETBACK_ACCEL_VERSION, 0,
++ port->net_dev->name, &accel_hooks);
++
++ if (rc < 0) {
++ EPRINTK("Xen netback accelerator version mismatch\n");
++ goto fail4;
++ } else if (rc > 0) {
++ /*
++ * In future may want to add backwards compatibility
++ * and accept certain subsets of previous versions
++ */
++ EPRINTK("Xen netback accelerator version mismatch\n");
++ goto fail4;
++ }
++
++ return 0;
++
++ fail4:
++ mutex_lock(&accel_mutex);
++ list_del(&port->link);
++ mutex_unlock(&accel_mutex);
++
++ efx_dl_unregister_callbacks(efx_dl_dev, &bend_dl_callbacks);
++ fail3:
++ netback_accel_shutdown_fwd_port(port->fwd_priv);
++ fail2:
++ efx_dl_dev->priv = NULL;
++ kfree(port);
++ fail1:
++ return rc;
++}
++
++
++static void bend_dl_remove(struct efx_dl_device *efx_dl_dev)
++{
++ struct driverlink_port *port;
++
++ DPRINTK("Unregistering driverlink callbacks.\n");
++
++ mutex_lock(&accel_mutex);
++
++ port = (struct driverlink_port *)efx_dl_dev->priv;
++
++ BUG_ON(list_empty(&dl_ports));
++ BUG_ON(port == NULL);
++ BUG_ON(port->efx_dl_dev != efx_dl_dev);
++
++ netback_disconnect_accelerator(0, port->net_dev->name);
++
++ list_del(&port->link);
++
++ mutex_unlock(&accel_mutex);
++
++ efx_dl_unregister_callbacks(efx_dl_dev, &bend_dl_callbacks);
++ netback_accel_shutdown_fwd_port(port->fwd_priv);
++
++ efx_dl_dev->priv = NULL;
++ kfree(port);
++
++ return;
++}
++
++
++static struct efx_dl_driver bend_dl_driver =
++ {
++ .name = "SFC Xen backend",
++ .probe = bend_dl_probe,
++ .remove = bend_dl_remove,
++ };
++
++
++int netback_accel_sf_init(void)
++{
++ int rc, nic_i;
++ struct efhw_nic *nic;
++
++ INIT_LIST_HEAD(&dl_ports);
++
++ rc = efx_dl_register_driver(&bend_dl_driver);
++ /* If we couldn't find the NET driver, give up */
++ if (rc == -ENOENT)
++ return rc;
++
++ if (rc == 0) {
++ EFRM_FOR_EACH_NIC(nic_i, nic)
++ falcon_nic_set_rx_usr_buf_size(nic,
++ SF_XEN_RX_USR_BUF_SIZE);
++ }
++
++ init_done = (rc == 0);
++ return rc;
++}
++
++
++void netback_accel_sf_shutdown(void)
++{
++ if (!init_done)
++ return;
++ DPRINTK("Unregistering driverlink driver\n");
++
++ /*
++ * This will trigger removal callbacks for all the devices, which
++ * will unregister their callbacks, disconnect from netfront, etc.
++ */
++ efx_dl_unregister_driver(&bend_dl_driver);
++}
++
++
++int netback_accel_sf_hwtype(struct netback_accel *bend)
++{
++ struct driverlink_port *port;
++
++ mutex_lock(&accel_mutex);
++
++ list_for_each_entry(port, &dl_ports, link) {
++ if (strcmp(bend->nicname, port->net_dev->name) == 0) {
++ bend->hw_type = port->type;
++ bend->accel_setup = netback_accel_setup_vnic_hw;
++ bend->accel_shutdown = netback_accel_shutdown_vnic_hw;
++ bend->fwd_priv = port->fwd_priv;
++ /* This is just needed to pass to efx_vi_alloc */
++ bend->nic_index = port->nic_index;
++ bend->net_dev = port->net_dev;
++ mutex_unlock(&accel_mutex);
++ return 0;
++ }
++ }
++
++ mutex_unlock(&accel_mutex);
++
++ EPRINTK("Failed to identify backend device '%s' with a NIC\n",
++ bend->nicname);
++
++ return -ENOENT;
++}
++
++
++/****************************************************************************
++ * Resource management code
++ ***************************************************************************/
++
++static int alloc_page_state(struct netback_accel *bend, int max_pages)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv;
++
++ if (max_pages < 0 || max_pages > bend->quotas.max_buf_pages) {
++ EPRINTK("%s: invalid max_pages: %d\n", __FUNCTION__, max_pages);
++ return -EINVAL;
++ }
++
++ accel_hw_priv = kzalloc(sizeof(struct falcon_bend_accel_priv),
++ GFP_KERNEL);
++ if (accel_hw_priv == NULL) {
++ EPRINTK("%s: no memory for accel_hw_priv\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ accel_hw_priv->dma_maps = kzalloc
++ (sizeof(struct efx_vi_dma_map_state **) *
++ (max_pages / NET_ACCEL_MSG_MAX_PAGE_REQ), GFP_KERNEL);
++ if (accel_hw_priv->dma_maps == NULL) {
++ EPRINTK("%s: no memory for dma_maps\n", __FUNCTION__);
++ kfree(accel_hw_priv);
++ return -ENOMEM;
++ }
++
++ bend->buffer_maps = kzalloc(sizeof(struct vm_struct *) * max_pages,
++ GFP_KERNEL);
++ if (bend->buffer_maps == NULL) {
++ EPRINTK("%s: no memory for buffer_maps\n", __FUNCTION__);
++ kfree(accel_hw_priv->dma_maps);
++ kfree(accel_hw_priv);
++ return -ENOMEM;
++ }
++
++ bend->buffer_addrs = kzalloc(sizeof(u64) * max_pages, GFP_KERNEL);
++ if (bend->buffer_addrs == NULL) {
++ kfree(bend->buffer_maps);
++ kfree(accel_hw_priv->dma_maps);
++ kfree(accel_hw_priv);
++ return -ENOMEM;
++ }
++
++ bend->accel_hw_priv = accel_hw_priv;
++
++ return 0;
++}
++
++
++static int free_page_state(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv;
++
++ DPRINTK("%s: %p\n", __FUNCTION__, bend);
++
++ accel_hw_priv = bend->accel_hw_priv;
++
++ if (accel_hw_priv) {
++ kfree(accel_hw_priv->dma_maps);
++ kfree(bend->buffer_maps);
++ kfree(bend->buffer_addrs);
++ kfree(accel_hw_priv);
++ bend->accel_hw_priv = NULL;
++ bend->max_pages = 0;
++ }
++
++ return 0;
++}
++
++
++/* The timeout event callback for the event q */
++static void bend_evq_timeout(void *context, int is_timeout)
++{
++ struct netback_accel *bend = (struct netback_accel *)context;
++ if (is_timeout) {
++ /* Pass event to vnic front end driver */
++ VPRINTK("timeout event to %d\n", bend->net_channel);
++ NETBACK_ACCEL_STATS_OP(bend->stats.evq_timeouts++);
++ notify_remote_via_irq(bend->net_channel_irq);
++ } else {
++ /* It's a wakeup event, used by Falcon */
++ VPRINTK("wakeup to %d\n", bend->net_channel);
++ NETBACK_ACCEL_STATS_OP(bend->stats.evq_wakeups++);
++ notify_remote_via_irq(bend->net_channel_irq);
++ }
++}
++
++
++/*
++ * Create the eventq and associated gubbins for communication with the
++ * front end vnic driver
++ */
++static int ef_get_vnic(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv;
++ int rc = 0;
++
++ BUG_ON(bend->hw_state != NETBACK_ACCEL_RES_NONE);
++
++ /* Allocate page related state and accel_hw_priv */
++ rc = alloc_page_state(bend, bend->max_pages);
++ if (rc != 0) {
++ EPRINTK("Failed to allocate page state: %d\n", rc);
++ return rc;
++ }
++
++ accel_hw_priv = bend->accel_hw_priv;
++
++ rc = efx_vi_alloc(&accel_hw_priv->efx_vih, bend->nic_index);
++ if (rc != 0) {
++ EPRINTK("%s: efx_vi_alloc failed %d\n", __FUNCTION__, rc);
++ free_page_state(bend);
++ return rc;
++ }
++
++ rc = efx_vi_eventq_register_callback(accel_hw_priv->efx_vih,
++ bend_evq_timeout,
++ bend);
++ if (rc != 0) {
++ EPRINTK("%s: register_callback failed %d\n", __FUNCTION__, rc);
++ efx_vi_free(accel_hw_priv->efx_vih);
++ free_page_state(bend);
++ return rc;
++ }
++
++ bend->hw_state = NETBACK_ACCEL_RES_ALLOC;
++
++ return 0;
++}
++
++
++static void ef_free_vnic(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++
++ BUG_ON(bend->hw_state != NETBACK_ACCEL_RES_ALLOC);
++
++ efx_vi_eventq_kill_callback(accel_hw_priv->efx_vih);
++
++ DPRINTK("Hardware is freeable. Will proceed.\n");
++
++ efx_vi_free(accel_hw_priv->efx_vih);
++ accel_hw_priv->efx_vih = NULL;
++
++ VPRINTK("Free page state...\n");
++ free_page_state(bend);
++
++ bend->hw_state = NETBACK_ACCEL_RES_NONE;
++}
++
++
++static inline void ungrant_or_crash(grant_ref_t gntref, int domain) {
++ if (net_accel_ungrant_page(gntref) == -EBUSY)
++ net_accel_shutdown_remote(domain);
++}
++
++
++static void netback_accel_release_hwinfo(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ int i;
++
++ DPRINTK("Remove dma q grants %d %d\n", accel_hw_priv->txdmaq_gnt,
++ accel_hw_priv->rxdmaq_gnt);
++ ungrant_or_crash(accel_hw_priv->txdmaq_gnt, bend->far_end);
++ ungrant_or_crash(accel_hw_priv->rxdmaq_gnt, bend->far_end);
++
++ DPRINTK("Remove doorbell grant %d\n", accel_hw_priv->doorbell_gnt);
++ ungrant_or_crash(accel_hw_priv->doorbell_gnt, bend->far_end);
++
++ if (bend->hw_type == NET_ACCEL_MSG_HWTYPE_FALCON_A) {
++ DPRINTK("Remove rptr grant %d\n", accel_hw_priv->evq_rptr_gnt);
++ ungrant_or_crash(accel_hw_priv->evq_rptr_gnt, bend->far_end);
++ }
++
++ for (i = 0; i < accel_hw_priv->evq_npages; i++) {
++ DPRINTK("Remove evq grant %d\n", accel_hw_priv->evq_mem_gnts[i]);
++ ungrant_or_crash(accel_hw_priv->evq_mem_gnts[i], bend->far_end);
++ }
++
++ bend->hw_state = NETBACK_ACCEL_RES_FILTER;
++
++ return;
++}
++
++
++static int ef_bend_hwinfo_falcon_common(struct netback_accel *bend,
++ struct net_accel_hw_falcon_b *hwinfo)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ struct efx_vi_hw_resource_metadata res_mdata;
++ struct efx_vi_hw_resource res_array[EFX_VI_HW_RESOURCE_MAXSIZE];
++ int rc, len = EFX_VI_HW_RESOURCE_MAXSIZE, i, pfn = 0;
++ unsigned long txdmaq_pfn = 0, rxdmaq_pfn = 0;
++
++ rc = efx_vi_hw_resource_get_phys(accel_hw_priv->efx_vih, &res_mdata,
++ res_array, &len);
++ if (rc != 0) {
++ DPRINTK("%s: resource_get_phys returned %d\n",
++ __FUNCTION__, rc);
++ return rc;
++ }
++
++ if (res_mdata.version != 0)
++ return -EPROTO;
++
++ hwinfo->nic_arch = res_mdata.nic_arch;
++ hwinfo->nic_variant = res_mdata.nic_variant;
++ hwinfo->nic_revision = res_mdata.nic_revision;
++
++ hwinfo->evq_order = res_mdata.evq_order;
++ hwinfo->evq_offs = res_mdata.evq_offs;
++ hwinfo->evq_capacity = res_mdata.evq_capacity;
++ hwinfo->instance = res_mdata.instance;
++ hwinfo->rx_capacity = res_mdata.rx_capacity;
++ hwinfo->tx_capacity = res_mdata.tx_capacity;
++
++ VPRINTK("evq_order %d evq_offs %d evq_cap %d inst %d rx_cap %d tx_cap %d\n",
++ hwinfo->evq_order, hwinfo->evq_offs, hwinfo->evq_capacity,
++ hwinfo->instance, hwinfo->rx_capacity, hwinfo->tx_capacity);
++
++ for (i = 0; i < len; i++) {
++ struct efx_vi_hw_resource *res = &(res_array[i]);
++ switch (res->type) {
++ case EFX_VI_HW_RESOURCE_TXDMAQ:
++ txdmaq_pfn = page_to_pfn(virt_to_page(res->address));
++ break;
++ case EFX_VI_HW_RESOURCE_RXDMAQ:
++ rxdmaq_pfn = page_to_pfn(virt_to_page(res->address));
++ break;
++ case EFX_VI_HW_RESOURCE_EVQTIMER:
++ break;
++ case EFX_VI_HW_RESOURCE_EVQRPTR:
++ case EFX_VI_HW_RESOURCE_EVQRPTR_OFFSET:
++ hwinfo->evq_rptr = res->address;
++ break;
++ case EFX_VI_HW_RESOURCE_EVQMEMKVA:
++ accel_hw_priv->evq_npages = 1 << res_mdata.evq_order;
++ pfn = page_to_pfn(virt_to_page(res->address));
++ break;
++ case EFX_VI_HW_RESOURCE_BELLPAGE:
++ hwinfo->doorbell_mfn = res->address;
++ break;
++ default:
++ EPRINTK("%s: Unknown hardware resource type %d\n",
++ __FUNCTION__, res->type);
++ break;
++ }
++ }
++
++ VPRINTK("Passing txdmaq page pfn %lx\n", txdmaq_pfn);
++ accel_hw_priv->txdmaq_gnt = hwinfo->txdmaq_gnt =
++ net_accel_grant_page(bend->hdev_data, pfn_to_mfn(txdmaq_pfn),
++ 0);
++
++ VPRINTK("Passing rxdmaq page pfn %lx\n", rxdmaq_pfn);
++ accel_hw_priv->rxdmaq_gnt = hwinfo->rxdmaq_gnt =
++ net_accel_grant_page(bend->hdev_data, pfn_to_mfn(rxdmaq_pfn),
++ 0);
++
++ VPRINTK("Passing doorbell page mfn %x\n", hwinfo->doorbell_mfn);
++ /* Make the relevant H/W pages mappable by the far end */
++ accel_hw_priv->doorbell_gnt = hwinfo->doorbell_gnt =
++ net_accel_grant_page(bend->hdev_data, hwinfo->doorbell_mfn, 1);
++
++ /* Now do the same for the memory pages */
++ /* Convert the page + length we got back for the evq to grants. */
++ for (i = 0; i < accel_hw_priv->evq_npages; i++) {
++ accel_hw_priv->evq_mem_gnts[i] = hwinfo->evq_mem_gnts[i] =
++ net_accel_grant_page(bend->hdev_data, pfn_to_mfn(pfn), 0);
++ VPRINTK("Got grant %u for evq pfn %x\n", hwinfo->evq_mem_gnts[i],
++ pfn);
++ pfn++;
++ }
++
++ return 0;
++}
++
++
++static int ef_bend_hwinfo_falcon_a(struct netback_accel *bend,
++ struct net_accel_hw_falcon_a *hwinfo)
++{
++ int rc;
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++
++ if ((rc = ef_bend_hwinfo_falcon_common(bend, &hwinfo->common)) != 0)
++ return rc;
++
++ /*
++ * Note that unlike the above, where the message field is the
++ * page number, here evq_rptr is the entire address because
++ * it is currently a pointer into the densely mapped timer page.
++ */
++ VPRINTK("Passing evq_rptr pfn %x for rptr %x\n",
++ hwinfo->common.evq_rptr >> PAGE_SHIFT,
++ hwinfo->common.evq_rptr);
++ rc = net_accel_grant_page(bend->hdev_data,
++ hwinfo->common.evq_rptr >> PAGE_SHIFT, 0);
++ if (rc < 0)
++ return rc;
++
++ accel_hw_priv->evq_rptr_gnt = hwinfo->evq_rptr_gnt = rc;
++ VPRINTK("evq_rptr_gnt got %d\n", hwinfo->evq_rptr_gnt);
++
++ return 0;
++}
++
++
++static int ef_bend_hwinfo_falcon_b(struct netback_accel *bend,
++ struct net_accel_hw_falcon_b *hwinfo)
++{
++ return ef_bend_hwinfo_falcon_common(bend, hwinfo);
++}
++
++
++/*
++ * Fill in the message with a description of the hardware resources, based on
++ * the H/W type
++ */
++static int netback_accel_hwinfo(struct netback_accel *bend,
++ struct net_accel_msg_hw *msgvi)
++{
++ int rc = 0;
++
++ BUG_ON(bend->hw_state != NETBACK_ACCEL_RES_FILTER);
++
++ msgvi->type = bend->hw_type;
++ switch (bend->hw_type) {
++ case NET_ACCEL_MSG_HWTYPE_FALCON_A:
++ rc = ef_bend_hwinfo_falcon_a(bend, &msgvi->resources.falcon_a);
++ break;
++ case NET_ACCEL_MSG_HWTYPE_FALCON_B:
++ rc = ef_bend_hwinfo_falcon_b(bend, &msgvi->resources.falcon_b);
++ break;
++ case NET_ACCEL_MSG_HWTYPE_NONE:
++ /* Nothing to do. The slow path should just work. */
++ break;
++ }
++
++ if (rc == 0)
++ bend->hw_state = NETBACK_ACCEL_RES_HWINFO;
++
++ return rc;
++}
++
++
++/* Allocate hardware resources and make them available to the client domain */
++int netback_accel_setup_vnic_hw(struct netback_accel *bend)
++{
++ struct net_accel_msg msg;
++ int err;
++
++ /* Allocate the event queue, VI and so on. */
++ err = ef_get_vnic(bend);
++ if (err) {
++ EPRINTK("Failed to allocate hardware resource for bend:"
++ "error %d\n", err);
++ return err;
++ }
++
++ /* Set up the filter management */
++ err = netback_accel_filter_init(bend);
++ if (err) {
++ EPRINTK("Filter setup failed, error %d", err);
++ ef_free_vnic(bend);
++ return err;
++ }
++
++ net_accel_msg_init(&msg, NET_ACCEL_MSG_SETHW);
++
++ /*
++ * Extract the low-level hardware info we will actually pass to the
++ * other end, and set up the grants/ioremap permissions needed
++ */
++ err = netback_accel_hwinfo(bend, &msg.u.hw);
++
++ if (err != 0) {
++ netback_accel_filter_shutdown(bend);
++ ef_free_vnic(bend);
++ return err;
++ }
++
++ /* Send the message, this is a reply to a hello-reply */
++ err = net_accel_msg_reply_notify(bend->shared_page,
++ bend->msg_channel_irq,
++ &bend->to_domU, &msg);
++
++ /*
++ * The message should succeed as it's logically a reply and we
++ * guarantee space for replies, but a misbehaving frontend
++ * could result in that behaviour, so be tolerant
++ */
++ if (err != 0) {
++ netback_accel_release_hwinfo(bend);
++ netback_accel_filter_shutdown(bend);
++ ef_free_vnic(bend);
++ }
++
++ return err;
++}
++
++
++/* Free hardware resources */
++void netback_accel_shutdown_vnic_hw(struct netback_accel *bend)
++{
++ /*
++ * Only try and release resources if accel_hw_priv was setup,
++ * otherwise there is nothing to do as we're on "null-op"
++ * acceleration
++ */
++ switch (bend->hw_state) {
++ case NETBACK_ACCEL_RES_HWINFO:
++ VPRINTK("Release hardware resources\n");
++ netback_accel_release_hwinfo(bend);
++ /* deliberate drop through */
++ case NETBACK_ACCEL_RES_FILTER:
++ VPRINTK("Free filters...\n");
++ netback_accel_filter_shutdown(bend);
++ /* deliberate drop through */
++ case NETBACK_ACCEL_RES_ALLOC:
++ VPRINTK("Free vnic...\n");
++ ef_free_vnic(bend);
++ /* deliberate drop through */
++ case NETBACK_ACCEL_RES_NONE:
++ break;
++ default:
++ BUG();
++ }
++}
++
++/**************************************************************************
++ *
++ * Buffer table stuff
++ *
++ **************************************************************************/
++
++/*
++ * Undo any allocation that netback_accel_msg_rx_buffer_map() has made
++ * if it fails half way through
++ */
++static inline void buffer_map_cleanup(struct netback_accel *bend, int i)
++{
++ while (i > 0) {
++ i--;
++ bend->buffer_maps_index--;
++ net_accel_unmap_device_page(bend->hdev_data,
++ bend->buffer_maps[bend->buffer_maps_index],
++ bend->buffer_addrs[bend->buffer_maps_index]);
++ }
++}
++
++
++int netback_accel_add_buffers(struct netback_accel *bend, int pages, int log2_pages,
++ u32 *grants, u32 *buf_addr_out)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ unsigned long long addr_array[NET_ACCEL_MSG_MAX_PAGE_REQ];
++ int rc, i, index;
++ u64 dev_bus_addr;
++
++ /* Make sure we can't overflow the dma_maps array */
++ if (accel_hw_priv->dma_maps_index >=
++ bend->max_pages / NET_ACCEL_MSG_MAX_PAGE_REQ) {
++ EPRINTK("%s: too many buffer table allocations: %d %d\n",
++ __FUNCTION__, accel_hw_priv->dma_maps_index,
++ bend->max_pages / NET_ACCEL_MSG_MAX_PAGE_REQ);
++ return -EINVAL;
++ }
++
++ /* Make sure we can't overflow the buffer_maps array */
++ if (bend->buffer_maps_index + pages > bend->max_pages) {
++ EPRINTK("%s: too many pages mapped: %d + %d > %d\n",
++ __FUNCTION__, bend->buffer_maps_index,
++ pages, bend->max_pages);
++ return -EINVAL;
++ }
++
++ for (i = 0; i < pages; i++) {
++ VPRINTK("%s: mapping page %d\n", __FUNCTION__, i);
++ rc = net_accel_map_device_page
++ (bend->hdev_data, grants[i],
++ &bend->buffer_maps[bend->buffer_maps_index],
++ &dev_bus_addr);
++
++ if (rc != 0) {
++ EPRINTK("error in net_accel_map_device_page\n");
++ buffer_map_cleanup(bend, i);
++ return rc;
++ }
++
++ bend->buffer_addrs[bend->buffer_maps_index] = dev_bus_addr;
++
++ bend->buffer_maps_index++;
++
++ addr_array[i] = dev_bus_addr;
++ }
++
++ VPRINTK("%s: mapping dma addresses to vih %p\n", __FUNCTION__,
++ accel_hw_priv->efx_vih);
++
++ index = accel_hw_priv->dma_maps_index;
++ if ((rc = efx_vi_dma_map_addrs(accel_hw_priv->efx_vih, addr_array, pages,
++ &(accel_hw_priv->dma_maps[index]))) < 0) {
++ EPRINTK("error in dma_map_pages\n");
++ buffer_map_cleanup(bend, i);
++ return rc;
++ }
++
++ accel_hw_priv->dma_maps_index++;
++ NETBACK_ACCEL_STATS_OP(bend->stats.num_buffer_pages += pages);
++
++ //DPRINTK("%s: getting map address\n", __FUNCTION__);
++
++ *buf_addr_out = efx_vi_dma_get_map_addr(accel_hw_priv->efx_vih,
++ accel_hw_priv->dma_maps[index]);
++
++ //DPRINTK("%s: done\n", __FUNCTION__);
++
++ return 0;
++}
++
++
++int netback_accel_remove_buffers(struct netback_accel *bend)
++{
++ /* Only try to free buffers if accel_hw_priv was setup */
++ if (bend->hw_state != NETBACK_ACCEL_RES_NONE) {
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ int i;
++
++ efx_vi_reset(accel_hw_priv->efx_vih);
++
++ while (accel_hw_priv->dma_maps_index > 0) {
++ accel_hw_priv->dma_maps_index--;
++ i = accel_hw_priv->dma_maps_index;
++ efx_vi_dma_unmap_addrs(accel_hw_priv->efx_vih,
++ accel_hw_priv->dma_maps[i]);
++ }
++
++ while (bend->buffer_maps_index > 0) {
++ VPRINTK("Unmapping granted buffer %d\n",
++ bend->buffer_maps_index);
++ bend->buffer_maps_index--;
++ i = bend->buffer_maps_index;
++ net_accel_unmap_device_page(bend->hdev_data,
++ bend->buffer_maps[i],
++ bend->buffer_addrs[i]);
++ }
++
++ NETBACK_ACCEL_STATS_OP(bend->stats.num_buffer_pages = 0);
++ }
++
++ return 0;
++}
++
++/**************************************************************************
++ *
++ * Filter stuff
++ *
++ **************************************************************************/
++
++static int netback_accel_filter_init(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ int i, rc;
++
++ BUG_ON(bend->hw_state != NETBACK_ACCEL_RES_ALLOC);
++
++ spin_lock_init(&accel_hw_priv->filter_lock);
++
++ if ((rc = cuckoo_hash_init(&accel_hw_priv->filter_hash_table,
++ 5 /* space for 32 filters */, 8)) != 0) {
++ EPRINTK("Failed to initialise filter hash table\n");
++ return rc;
++ }
++
++ accel_hw_priv->fspecs = kzalloc(sizeof(struct netback_accel_filter_spec) *
++ bend->quotas.max_filters,
++ GFP_KERNEL);
++
++ if (accel_hw_priv->fspecs == NULL) {
++ EPRINTK("No memory for filter specs.\n");
++ cuckoo_hash_destroy(&accel_hw_priv->filter_hash_table);
++ return -ENOMEM;
++ }
++
++ for (i = 0; i < bend->quotas.max_filters; i++) {
++ accel_hw_priv->free_filters |= (1 << i);
++ }
++
++ /* Base mask on highest set bit in max_filters */
++ accel_hw_priv->filter_idx_mask = (1 << fls(bend->quotas.max_filters)) - 1;
++ VPRINTK("filter setup: max is %x mask is %x\n",
++ bend->quotas.max_filters, accel_hw_priv->filter_idx_mask);
++
++ bend->hw_state = NETBACK_ACCEL_RES_FILTER;
++
++ return 0;
++}
++
++
++static inline void make_filter_key(cuckoo_hash_ip_key *key,
++ struct netback_accel_filter_spec *filt)
++
++{
++ key->local_ip = filt->destip_be;
++ key->local_port = filt->destport_be;
++ key->proto = filt->proto;
++}
++
++
++static inline
++void netback_accel_free_filter(struct falcon_bend_accel_priv *accel_hw_priv,
++ int filter)
++{
++ cuckoo_hash_ip_key filter_key;
++
++ if (!(accel_hw_priv->free_filters & (1 << filter))) {
++ efx_vi_filter_stop(accel_hw_priv->efx_vih,
++ accel_hw_priv->fspecs[filter].filter_handle);
++ make_filter_key(&filter_key, &(accel_hw_priv->fspecs[filter]));
++ if (cuckoo_hash_remove(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)&filter_key)) {
++ EPRINTK("%s: Couldn't find filter to remove from table\n",
++ __FUNCTION__);
++ BUG();
++ }
++ }
++}
++
++
++static void netback_accel_filter_shutdown(struct netback_accel *bend)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ int i;
++ unsigned long flags;
++
++ BUG_ON(bend->hw_state != NETBACK_ACCEL_RES_FILTER);
++
++ spin_lock_irqsave(&accel_hw_priv->filter_lock, flags);
++
++ BUG_ON(accel_hw_priv->fspecs == NULL);
++
++ for (i = 0; i < bend->quotas.max_filters; i++) {
++ netback_accel_free_filter(accel_hw_priv, i);
++ }
++
++ kfree(accel_hw_priv->fspecs);
++ accel_hw_priv->fspecs = NULL;
++ accel_hw_priv->free_filters = 0;
++
++ cuckoo_hash_destroy(&accel_hw_priv->filter_hash_table);
++
++ spin_unlock_irqrestore(&accel_hw_priv->filter_lock, flags);
++
++ bend->hw_state = NETBACK_ACCEL_RES_ALLOC;
++}
++
++
++/*! Suggest a filter to replace when we want to insert a new one and have
++ * none free.
++ */
++static unsigned get_victim_filter(struct netback_accel *bend)
++{
++ /*
++ * We could attempt to get really clever, and may do at some
++ * point, but random replacement is v. cheap and low on
++ * pathological worst cases.
++ */
++ unsigned index, cycles;
++
++ rdtscl(cycles);
++
++ /*
++ * Some doubt about the quality of the bottom few bits, so
++ * throw 'em * away
++ */
++ index = (cycles >> 4) & ((struct falcon_bend_accel_priv *)
++ bend->accel_hw_priv)->filter_idx_mask;
++ /*
++ * We don't enforce that the number of filters is a power of
++ * two, but the masking gets us to within one subtraction of a
++ * valid index
++ */
++ if (index >= bend->quotas.max_filters)
++ index -= bend->quotas.max_filters;
++ DPRINTK("backend %s->%d has no free filters. Filter %d will be evicted\n",
++ bend->nicname, bend->far_end, index);
++ return index;
++}
++
++
++/* Add a filter for the specified IP/port to the backend */
++int
++netback_accel_filter_check_add(struct netback_accel *bend,
++ struct netback_accel_filter_spec *filt)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ struct netback_accel_filter_spec *fs;
++ unsigned filter_index;
++ unsigned long flags;
++ int rc, recycling = 0;
++ cuckoo_hash_ip_key filter_key, evict_key;
++
++ BUG_ON(filt->proto != IPPROTO_TCP && filt->proto != IPPROTO_UDP);
++
++ DPRINTK("Will add %s filter for dst ip %08x and dst port %d\n",
++ (filt->proto == IPPROTO_TCP) ? "TCP" : "UDP",
++ be32_to_cpu(filt->destip_be), be16_to_cpu(filt->destport_be));
++
++ spin_lock_irqsave(&accel_hw_priv->filter_lock, flags);
++ /*
++ * Check to see if we're already filtering this IP address and
++ * port. Happens if you insert a filter mid-stream as there
++ * are many packets backed up to be delivered to dom0 already
++ */
++ make_filter_key(&filter_key, filt);
++ if (cuckoo_hash_lookup(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)(&filter_key),
++ &filter_index)) {
++ DPRINTK("Found matching filter %d already in table\n",
++ filter_index);
++ rc = -1;
++ goto out;
++ }
++
++ if (accel_hw_priv->free_filters == 0) {
++ filter_index = get_victim_filter(bend);
++ recycling = 1;
++ } else {
++ filter_index = __ffs(accel_hw_priv->free_filters);
++ clear_bit(filter_index, &accel_hw_priv->free_filters);
++ }
++
++ fs = &accel_hw_priv->fspecs[filter_index];
++
++ if (recycling) {
++ DPRINTK("Removing filter index %d handle %p\n", filter_index,
++ fs->filter_handle);
++
++ if ((rc = efx_vi_filter_stop(accel_hw_priv->efx_vih,
++ fs->filter_handle)) != 0) {
++ EPRINTK("Couldn't clear NIC filter table entry %d\n", rc);
++ }
++
++ make_filter_key(&evict_key, fs);
++ if (cuckoo_hash_remove(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)&evict_key)) {
++ EPRINTK("Couldn't find filter to remove from table\n");
++ BUG();
++ }
++ NETBACK_ACCEL_STATS_OP(bend->stats.num_filters--);
++ }
++
++ /* Update the filter spec with new details */
++ *fs = *filt;
++
++ if ((rc = cuckoo_hash_add(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)&filter_key, filter_index,
++ 1)) != 0) {
++ EPRINTK("Error (%d) adding filter to table\n", rc);
++ accel_hw_priv->free_filters |= (1 << filter_index);
++ goto out;
++ }
++
++ rc = efx_vi_filter(accel_hw_priv->efx_vih, filt->proto, filt->destip_be,
++ filt->destport_be,
++ (struct filter_resource_t **)&fs->filter_handle);
++
++ if (rc != 0) {
++ EPRINTK("Hardware filter insertion failed. Error %d\n", rc);
++ accel_hw_priv->free_filters |= (1 << filter_index);
++ cuckoo_hash_remove(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)&filter_key);
++ rc = -1;
++ goto out;
++ }
++
++ NETBACK_ACCEL_STATS_OP(bend->stats.num_filters++);
++
++ VPRINTK("%s: success index %d handle %p\n", __FUNCTION__, filter_index,
++ fs->filter_handle);
++
++ rc = filter_index;
++ out:
++ spin_unlock_irqrestore(&accel_hw_priv->filter_lock, flags);
++ return rc;
++}
++
++
++/* Remove a filter entry for the specific device and IP/port */
++static void netback_accel_filter_remove(struct netback_accel *bend,
++ int filter_index)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++
++ BUG_ON(accel_hw_priv->free_filters & (1 << filter_index));
++ netback_accel_free_filter(accel_hw_priv, filter_index);
++ accel_hw_priv->free_filters |= (1 << filter_index);
++}
++
++
++/* Remove a filter entry for the specific device and IP/port */
++void netback_accel_filter_remove_spec(struct netback_accel *bend,
++ struct netback_accel_filter_spec *filt)
++{
++ struct falcon_bend_accel_priv *accel_hw_priv = bend->accel_hw_priv;
++ unsigned filter_found;
++ unsigned long flags;
++ cuckoo_hash_ip_key filter_key;
++ struct netback_accel_filter_spec *fs;
++
++ if (filt->proto == IPPROTO_TCP) {
++ DPRINTK("Remove TCP filter for dst ip %08x and dst port %d\n",
++ be32_to_cpu(filt->destip_be),
++ be16_to_cpu(filt->destport_be));
++ } else if (filt->proto == IPPROTO_UDP) {
++ DPRINTK("Remove UDP filter for dst ip %08x and dst port %d\n",
++ be32_to_cpu(filt->destip_be),
++ be16_to_cpu(filt->destport_be));
++ } else {
++ /*
++ * This could be provoked by an evil frontend, so can't
++ * BUG(), but harmless as it should fail tests below
++ */
++ DPRINTK("Non-TCP/UDP filter dst ip %08x and dst port %d\n",
++ be32_to_cpu(filt->destip_be),
++ be16_to_cpu(filt->destport_be));
++ }
++
++ spin_lock_irqsave(&accel_hw_priv->filter_lock, flags);
++
++ make_filter_key(&filter_key, filt);
++ if (!cuckoo_hash_lookup(&accel_hw_priv->filter_hash_table,
++ (cuckoo_hash_key *)(&filter_key),
++ &filter_found)) {
++ EPRINTK("Couldn't find matching filter already in table\n");
++ goto out;
++ }
++
++ /* Do a full check to make sure we've not had a hash collision */
++ fs = &accel_hw_priv->fspecs[filter_found];
++ if (fs->destip_be == filt->destip_be &&
++ fs->destport_be == filt->destport_be &&
++ fs->proto == filt->proto &&
++ !memcmp(fs->mac, filt->mac, ETH_ALEN)) {
++ netback_accel_filter_remove(bend, filter_found);
++ } else {
++ EPRINTK("Entry in hash table does not match filter spec\n");
++ goto out;
++ }
++
++ out:
++ spin_unlock_irqrestore(&accel_hw_priv->filter_lock, flags);
++}
+diff --git a/drivers/xen/sfc_netback/accel_solarflare.h b/drivers/xen/sfc_netback/accel_solarflare.h
+new file mode 100644
+index 0000000..84d2146
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_solarflare.h
+@@ -0,0 +1,88 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETBACK_ACCEL_SOLARFLARE_H
++#define NETBACK_ACCEL_SOLARFLARE_H
++
++#include "accel.h"
++#include "accel_msg_iface.h"
++
++#include "driverlink_api.h"
++
++#define MAX_NICS 5
++#define MAX_PORTS 2
++
++
++extern int netback_accel_sf_init(void);
++extern void netback_accel_sf_shutdown(void);
++extern int netback_accel_sf_hwtype(struct netback_accel *bend);
++
++extern int netback_accel_sf_char_init(void);
++extern void netback_accel_sf_char_shutdown(void);
++
++extern int netback_accel_setup_vnic_hw(struct netback_accel *bend);
++extern void netback_accel_shutdown_vnic_hw(struct netback_accel *bend);
++
++extern int netback_accel_add_buffers(struct netback_accel *bend, int pages,
++ int log2_pages, u32 *grants,
++ u32 *buf_addr_out);
++extern int netback_accel_remove_buffers(struct netback_accel *bend);
++
++
++/* Add a filter for the specified IP/port to the backend */
++extern int
++netback_accel_filter_check_add(struct netback_accel *bend,
++ struct netback_accel_filter_spec *filt);
++/* Remove a filter entry for the specific device and IP/port */
++extern
++void netback_accel_filter_remove_index(struct netback_accel *bend,
++ int filter_index);
++extern
++void netback_accel_filter_remove_spec(struct netback_accel *bend,
++ struct netback_accel_filter_spec *filt);
++
++/* This is designed to look a bit like a skb */
++struct netback_pkt_buf {
++ union {
++ unsigned char *raw;
++ } mac;
++ union {
++ struct iphdr *iph;
++ struct arphdr *arph;
++ unsigned char *raw;
++ } nh;
++ int protocol;
++};
++
++/*! \brief Handle a received packet: insert fast path filters as necessary
++ * \param skb The packet buffer
++ */
++extern void netback_accel_rx_packet(struct netback_pkt_buf *skb, void *fwd_priv);
++
++/*! \brief Handle a transmitted packet: update fast path filters as necessary
++ * \param skb The packet buffer
++ */
++extern void netback_accel_tx_packet(struct sk_buff *skb, void *fwd_priv);
++
++#endif /* NETBACK_ACCEL_SOLARFLARE_H */
+diff --git a/drivers/xen/sfc_netback/accel_xenbus.c b/drivers/xen/sfc_netback/accel_xenbus.c
+new file mode 100644
+index 0000000..c98a235
+--- /dev/null
++++ b/drivers/xen/sfc_netback/accel_xenbus.c
+@@ -0,0 +1,829 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <xen/evtchn.h>
++#include <linux/mutex.h>
++
++/* drivers/xen/netback/common.h */
++#include "common.h"
++
++#include "accel.h"
++#include "accel_solarflare.h"
++#include "accel_util.h"
++
++#define NODENAME_PATH_FMT "backend/vif/%d/%d"
++
++#define NETBACK_ACCEL_FROM_XENBUS_DEVICE(_dev) (struct netback_accel *) \
++ ((struct backend_info *)(_dev)->dev.driver_data)->netback_accel_priv
++
++/* List of all the bends currently in existence. */
++struct netback_accel *bend_list = NULL;
++DEFINE_MUTEX(bend_list_mutex);
++
++/* Put in bend_list. Must hold bend_list_mutex */
++static void link_bend(struct netback_accel *bend)
++{
++ bend->next_bend = bend_list;
++ bend_list = bend;
++}
++
++/* Remove from bend_list, Must hold bend_list_mutex */
++static void unlink_bend(struct netback_accel *bend)
++{
++ struct netback_accel *tmp = bend_list;
++ struct netback_accel *prev = NULL;
++ while (tmp != NULL) {
++ if (tmp == bend) {
++ if (prev != NULL)
++ prev->next_bend = bend->next_bend;
++ else
++ bend_list = bend->next_bend;
++ return;
++ }
++ prev = tmp;
++ tmp = tmp->next_bend;
++ }
++}
++
++
++/* Demultiplex a message IRQ from the frontend driver. */
++static irqreturn_t msgirq_from_frontend(int irq, void *context)
++{
++ struct xenbus_device *dev = context;
++ struct netback_accel *bend = NETBACK_ACCEL_FROM_XENBUS_DEVICE(dev);
++ VPRINTK("irq %d from device %s\n", irq, dev->nodename);
++ schedule_work(&bend->handle_msg);
++ return IRQ_HANDLED;
++}
++
++
++/*
++ * Demultiplex an IRQ from the frontend driver. This is never used
++ * functionally, but we need it to pass to the bind function, and may
++ * get called spuriously
++ */
++static irqreturn_t netirq_from_frontend(int irq, void *context)
++{
++ VPRINTK("netirq %d from device %s\n", irq,
++ ((struct xenbus_device *)context)->nodename);
++
++ return IRQ_HANDLED;
++}
++
++
++/* Read the limits values of the xenbus structure. */
++static
++void cfg_hw_quotas(struct xenbus_device *dev, struct netback_accel *bend)
++{
++ int err = xenbus_gather
++ (XBT_NIL, dev->nodename,
++ "limits/max-filters", "%d", &bend->quotas.max_filters,
++ "limits/max-buf-pages", "%d", &bend->quotas.max_buf_pages,
++ "limits/max-mcasts", "%d", &bend->quotas.max_mcasts,
++ NULL);
++ if (err) {
++ /*
++ * TODO what if they have previously been set by the
++ * user? This will overwrite with defaults. Maybe
++ * not what we want to do, but useful in startup
++ * case
++ */
++ DPRINTK("Failed to read quotas from xenbus, using defaults\n");
++ bend->quotas.max_filters = NETBACK_ACCEL_DEFAULT_MAX_FILTERS;
++ bend->quotas.max_buf_pages = sfc_netback_max_pages;
++ bend->quotas.max_mcasts = NETBACK_ACCEL_DEFAULT_MAX_MCASTS;
++ }
++
++ return;
++}
++
++
++static void bend_config_accel_change(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ struct netback_accel *bend;
++
++ bend = container_of(watch, struct netback_accel, config_accel_watch);
++
++ mutex_lock(&bend->bend_mutex);
++ if (bend->config_accel_watch.node != NULL) {
++ struct xenbus_device *dev =
++ (struct xenbus_device *)bend->hdev_data;
++ DPRINTK("Watch matched, got dev %p otherend %p\n",
++ dev, dev->otherend);
++ if(!xenbus_exists(XBT_NIL, watch->node, "")) {
++ DPRINTK("Ignoring watch as otherend seems invalid\n");
++ goto out;
++ }
++
++ cfg_hw_quotas(dev, bend);
++ }
++ out:
++ mutex_unlock(&bend->bend_mutex);
++ return;
++}
++
++
++/*
++ * Setup watch on "limits" in the backend vif info to know when
++ * configuration has been set
++ */
++static int setup_config_accel_watch(struct xenbus_device *dev,
++ struct netback_accel *bend)
++{
++ int err;
++
++ VPRINTK("Setting watch on %s/%s\n", dev->nodename, "limits");
++
++ err = xenbus_watch_path2(dev, dev->nodename, "limits",
++ &bend->config_accel_watch,
++ bend_config_accel_change);
++
++ if (err) {
++ EPRINTK("%s: Failed to register xenbus watch: %d\n",
++ __FUNCTION__, err);
++ bend->config_accel_watch.node = NULL;
++ return err;
++ }
++ return 0;
++}
++
++
++static int
++cfg_frontend_info(struct xenbus_device *dev, struct netback_accel *bend,
++ int *grants)
++{
++ /* Get some info from xenbus on the event channel and shmem grant */
++ int err = xenbus_gather(XBT_NIL, dev->otherend,
++ "accel-msg-channel", "%u", &bend->msg_channel,
++ "accel-ctrl-page", "%d", &(grants[0]),
++ "accel-msg-page", "%d", &(grants[1]),
++ "accel-net-channel", "%u", &bend->net_channel,
++ NULL);
++ if (err)
++ EPRINTK("failed to read event channels or shmem grant: %d\n",
++ err);
++ else
++ DPRINTK("got event chan %d and net chan %d from frontend\n",
++ bend->msg_channel, bend->net_channel);
++ return err;
++}
++
++
++/* Setup all the comms needed to chat with the front end driver */
++static int setup_vnic(struct xenbus_device *dev)
++{
++ struct netback_accel *bend;
++ int grants[2], err, msgs_per_queue;
++
++ bend = NETBACK_ACCEL_FROM_XENBUS_DEVICE(dev);
++
++ err = cfg_frontend_info(dev, bend, grants);
++ if (err)
++ goto fail1;
++
++ /*
++ * If we get here, both frontend Connected and configuration
++ * options available. All is well.
++ */
++
++ /* Get the hardware quotas for the VNIC in question. */
++ cfg_hw_quotas(dev, bend);
++
++ /* Set up the deferred work handlers */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ INIT_WORK(&bend->handle_msg,
++ netback_accel_msg_rx_handler);
++#else
++ INIT_WORK(&bend->handle_msg,
++ netback_accel_msg_rx_handler,
++ (void*)bend);
++#endif
++
++ /* Request the frontend mac */
++ err = net_accel_xen_net_read_mac(dev, bend->mac);
++ if (err)
++ goto fail2;
++
++ /* Set up the shared page. */
++ bend->shared_page = net_accel_map_grants_contig(dev, grants, 2,
++ &bend->sh_pages_unmap);
++
++ if (bend->shared_page == NULL) {
++ EPRINTK("failed to map shared page for %s\n", dev->otherend);
++ err = -ENOMEM;
++ goto fail2;
++ }
++
++ /* Initialise the shared page(s) used for comms */
++ net_accel_msg_init_page(bend->shared_page, PAGE_SIZE,
++ bend->net_dev->flags & IFF_UP);
++
++ msgs_per_queue = (PAGE_SIZE/2) / sizeof(struct net_accel_msg);
++
++ net_accel_msg_init_queue
++ (&bend->to_domU, &bend->shared_page->queue0,
++ (struct net_accel_msg *)((__u8*)bend->shared_page + PAGE_SIZE),
++ msgs_per_queue);
++
++ net_accel_msg_init_queue
++ (&bend->from_domU, &bend->shared_page->queue1,
++ (struct net_accel_msg *)((__u8*)bend->shared_page +
++ (3 * PAGE_SIZE / 2)),
++ msgs_per_queue);
++
++ /* Bind the message event channel to a handler
++ *
++ * Note that we will probably get a spurious interrupt when we
++ * do this, so it must not be done until we have set up
++ * everything we need to handle it.
++ */
++ err = bind_interdomain_evtchn_to_irqhandler(dev->otherend_id,
++ bend->msg_channel,
++ msgirq_from_frontend,
++ 0,
++ "netback_accel",
++ dev);
++ if (err < 0) {
++ EPRINTK("failed to bind event channel: %d\n", err);
++ goto fail3;
++ }
++ else
++ bend->msg_channel_irq = err;
++
++ /* TODO: No need to bind this evtchn to an irq. */
++ err = bind_interdomain_evtchn_to_irqhandler(dev->otherend_id,
++ bend->net_channel,
++ netirq_from_frontend,
++ 0,
++ "netback_accel",
++ dev);
++ if (err < 0) {
++ EPRINTK("failed to bind net channel: %d\n", err);
++ goto fail4;
++ }
++ else
++ bend->net_channel_irq = err;
++
++ /*
++ * Grab ourselves an entry in the forwarding hash table. We do
++ * this now so we don't have the embarassmesnt of sorting out
++ * an allocation failure while at IRQ. Because we pass NULL as
++ * the context, the actual hash lookup will succeed for this
++ * NIC, but the check for somewhere to forward to will
++ * fail. This is necessary to prevent forwarding before
++ * hardware resources are set up
++ */
++ err = netback_accel_fwd_add(bend->mac, NULL, bend->fwd_priv);
++ if (err) {
++ EPRINTK("failed to add to fwd hash table\n");
++ goto fail5;
++ }
++
++ /*
++ * Say hello to frontend. Important to do this straight after
++ * obtaining the message queue as otherwise we are vulnerable
++ * to an evil frontend sending a HELLO-REPLY before we've sent
++ * the HELLO and confusing us
++ */
++ netback_accel_msg_tx_hello(bend, NET_ACCEL_MSG_VERSION);
++ return 0;
++
++ fail5:
++ unbind_from_irqhandler(bend->net_channel_irq, dev);
++ fail4:
++ unbind_from_irqhandler(bend->msg_channel_irq, dev);
++ fail3:
++ net_accel_unmap_grants_contig(dev, bend->sh_pages_unmap);
++ bend->shared_page = NULL;
++ bend->sh_pages_unmap = NULL;
++ fail2:
++ fail1:
++ return err;
++}
++
++
++static int read_nicname(struct xenbus_device *dev, struct netback_accel *bend)
++{
++ int len;
++
++ /* nic name used to select interface used for acceleration */
++ bend->nicname = xenbus_read(XBT_NIL, dev->nodename, "accel", &len);
++ if (IS_ERR(bend->nicname))
++ return PTR_ERR(bend->nicname);
++
++ return 0;
++}
++
++static const char *frontend_name = "sfc_netfront";
++
++static int publish_frontend_name(struct xenbus_device *dev)
++{
++ struct xenbus_transaction tr;
++ int err;
++
++ /* Publish the name of the frontend driver */
++ do {
++ err = xenbus_transaction_start(&tr);
++ if (err != 0) {
++ EPRINTK("%s: transaction start failed\n", __FUNCTION__);
++ return err;
++ }
++ err = xenbus_printf(tr, dev->nodename, "accel-frontend",
++ "%s", frontend_name);
++ if (err != 0) {
++ EPRINTK("%s: xenbus_printf failed\n", __FUNCTION__);
++ xenbus_transaction_end(tr, 1);
++ return err;
++ }
++ err = xenbus_transaction_end(tr, 0);
++ } while (err == -EAGAIN);
++
++ if (err != 0) {
++ EPRINTK("failed to end frontend name transaction\n");
++ return err;
++ }
++ return 0;
++}
++
++
++static int unpublish_frontend_name(struct xenbus_device *dev)
++{
++ struct xenbus_transaction tr;
++ int err;
++
++ do {
++ err = xenbus_transaction_start(&tr);
++ if (err != 0)
++ break;
++ err = xenbus_rm(tr, dev->nodename, "accel-frontend");
++ if (err != 0) {
++ xenbus_transaction_end(tr, 1);
++ break;
++ }
++ err = xenbus_transaction_end(tr, 0);
++ } while (err == -EAGAIN);
++
++ return err;
++}
++
++
++static void cleanup_vnic(struct netback_accel *bend)
++{
++ struct xenbus_device *dev;
++
++ dev = (struct xenbus_device *)bend->hdev_data;
++
++ DPRINTK("%s: bend %p dev %p\n", __FUNCTION__, bend, dev);
++
++ DPRINTK("%s: Remove %p's mac from fwd table...\n",
++ __FUNCTION__, bend);
++ netback_accel_fwd_remove(bend->mac, bend->fwd_priv);
++
++ /* Free buffer table allocations */
++ netback_accel_remove_buffers(bend);
++
++ DPRINTK("%s: Release hardware resources...\n", __FUNCTION__);
++ if (bend->accel_shutdown)
++ bend->accel_shutdown(bend);
++
++ if (bend->net_channel_irq) {
++ unbind_from_irqhandler(bend->net_channel_irq, dev);
++ bend->net_channel_irq = 0;
++ }
++
++ if (bend->msg_channel_irq) {
++ unbind_from_irqhandler(bend->msg_channel_irq, dev);
++ bend->msg_channel_irq = 0;
++ }
++
++ if (bend->sh_pages_unmap) {
++ DPRINTK("%s: Unmap grants %p\n", __FUNCTION__,
++ bend->sh_pages_unmap);
++ net_accel_unmap_grants_contig(dev, bend->sh_pages_unmap);
++ bend->sh_pages_unmap = NULL;
++ bend->shared_page = NULL;
++ }
++}
++
++
++/*************************************************************************/
++
++/*
++ * The following code handles accelstate changes between the frontend
++ * and the backend. It calls setup_vnic and cleanup_vnic in matching
++ * pairs in response to transitions.
++ *
++ * Valid state transitions for Dom0 are as follows:
++ *
++ * Closed->Init on probe or in response to Init from domU
++ * Closed->Closing on error/remove
++ *
++ * Init->Connected in response to Connected from domU
++ * Init->Closing on error/remove or in response to Closing from domU
++ *
++ * Connected->Closing on error/remove or in response to Closing from domU
++ *
++ * Closing->Closed in response to Closed from domU
++ *
++ */
++
++
++static void netback_accel_frontend_changed(struct xenbus_device *dev,
++ XenbusState frontend_state)
++{
++ struct netback_accel *bend = NETBACK_ACCEL_FROM_XENBUS_DEVICE(dev);
++ XenbusState backend_state;
++
++ DPRINTK("%s: changing from %s to %s. nodename %s, otherend %s\n",
++ __FUNCTION__, xenbus_strstate(bend->frontend_state),
++ xenbus_strstate(frontend_state),dev->nodename, dev->otherend);
++
++ /*
++ * Ignore duplicate state changes. This can happen if the
++ * frontend changes state twice in quick succession and the
++ * first watch fires in the backend after the second
++ * transition has completed.
++ */
++ if (bend->frontend_state == frontend_state)
++ return;
++
++ bend->frontend_state = frontend_state;
++ backend_state = bend->backend_state;
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ if (backend_state == XenbusStateClosed &&
++ !bend->removing)
++ backend_state = XenbusStateInitialising;
++ break;
++
++ case XenbusStateConnected:
++ if (backend_state == XenbusStateInitialising) {
++ if (!bend->vnic_is_setup &&
++ setup_vnic(dev) == 0) {
++ bend->vnic_is_setup = 1;
++ backend_state = XenbusStateConnected;
++ } else {
++ backend_state = XenbusStateClosing;
++ }
++ }
++ break;
++
++ case XenbusStateInitWait:
++ case XenbusStateInitialised:
++ default:
++ DPRINTK("Unknown state %s (%d) from frontend.\n",
++ xenbus_strstate(frontend_state), frontend_state);
++ /* Unknown state. Fall through. */
++ case XenbusStateClosing:
++ if (backend_state != XenbusStateClosed)
++ backend_state = XenbusStateClosing;
++
++ /*
++ * The bend will now persist (with watches active) in
++ * case the frontend comes back again, eg. after
++ * frontend module reload or suspend/resume
++ */
++
++ break;
++
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ if (bend->vnic_is_setup) {
++ bend->vnic_is_setup = 0;
++ cleanup_vnic(bend);
++ }
++
++ if (backend_state == XenbusStateClosing)
++ backend_state = XenbusStateClosed;
++ break;
++ }
++
++ if (backend_state != bend->backend_state) {
++ DPRINTK("Switching from state %s (%d) to %s (%d)\n",
++ xenbus_strstate(bend->backend_state),
++ bend->backend_state,
++ xenbus_strstate(backend_state), backend_state);
++ bend->backend_state = backend_state;
++ net_accel_update_state(dev, backend_state);
++ }
++
++ wake_up(&bend->state_wait_queue);
++}
++
++
++/* accelstate on the frontend's xenbus node has changed */
++static void bend_domu_accel_change(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int state;
++ struct netback_accel *bend;
++
++ bend = container_of(watch, struct netback_accel, domu_accel_watch);
++ if (bend->domu_accel_watch.node != NULL) {
++ struct xenbus_device *dev =
++ (struct xenbus_device *)bend->hdev_data;
++ VPRINTK("Watch matched, got dev %p otherend %p\n",
++ dev, dev->otherend);
++ /*
++ * dev->otherend != NULL check to protect against
++ * watch firing when domain goes away and we haven't
++ * yet cleaned up
++ */
++ if (!dev->otherend ||
++ !xenbus_exists(XBT_NIL, watch->node, "") ||
++ strncmp(dev->otherend, vec[XS_WATCH_PATH],
++ strlen(dev->otherend))) {
++ DPRINTK("Ignoring watch as otherend seems invalid\n");
++ return;
++ }
++
++ mutex_lock(&bend->bend_mutex);
++
++ xenbus_scanf(XBT_NIL, dev->otherend, "accelstate", "%d",
++ &state);
++ netback_accel_frontend_changed(dev, state);
++
++ mutex_unlock(&bend->bend_mutex);
++ }
++}
++
++/* Setup watch on frontend's accelstate */
++static int setup_domu_accel_watch(struct xenbus_device *dev,
++ struct netback_accel *bend)
++{
++ int err;
++
++ VPRINTK("Setting watch on %s/%s\n", dev->otherend, "accelstate");
++
++ err = xenbus_watch_path2(dev, dev->otherend, "accelstate",
++ &bend->domu_accel_watch,
++ bend_domu_accel_change);
++ if (err) {
++ EPRINTK("%s: Failed to register xenbus watch: %d\n",
++ __FUNCTION__, err);
++ goto fail;
++ }
++ return 0;
++ fail:
++ bend->domu_accel_watch.node = NULL;
++ return err;
++}
++
++
++int netback_accel_probe(struct xenbus_device *dev)
++{
++ struct netback_accel *bend;
++ struct backend_info *binfo;
++ int err;
++
++ DPRINTK("%s: passed device %s\n", __FUNCTION__, dev->nodename);
++
++ /* Allocate structure to store all our state... */
++ bend = kzalloc(sizeof(struct netback_accel), GFP_KERNEL);
++ if (bend == NULL) {
++ DPRINTK("%s: no memory for bend\n", __FUNCTION__);
++ return -ENOMEM;
++ }
++
++ mutex_init(&bend->bend_mutex);
++
++ mutex_lock(&bend->bend_mutex);
++
++ /* ...and store it where we can get at it */
++ binfo = (struct backend_info *) dev->dev.driver_data;
++ binfo->netback_accel_priv = bend;
++ /* And vice-versa */
++ bend->hdev_data = dev;
++
++ DPRINTK("%s: Adding bend %p to list\n", __FUNCTION__, bend);
++
++ init_waitqueue_head(&bend->state_wait_queue);
++ bend->vnic_is_setup = 0;
++ bend->frontend_state = XenbusStateUnknown;
++ bend->backend_state = XenbusStateClosed;
++ bend->removing = 0;
++
++ sscanf(dev->nodename, NODENAME_PATH_FMT, &bend->far_end,
++ &bend->vif_num);
++
++ err = read_nicname(dev, bend);
++ if (err) {
++ /*
++ * Technically not an error, just means we're not
++ * supposed to accelerate this
++ */
++ DPRINTK("failed to get device name\n");
++ goto fail_nicname;
++ }
++
++ /*
++ * Look up the device name in the list of NICs provided by
++ * driverlink to get the hardware type.
++ */
++ err = netback_accel_sf_hwtype(bend);
++ if (err) {
++ /*
++ * Technically not an error, just means we're not
++ * supposed to accelerate this, probably belongs to
++ * some other backend
++ */
++ DPRINTK("failed to match device name\n");
++ goto fail_init_type;
++ }
++
++ err = publish_frontend_name(dev);
++ if (err)
++ goto fail_publish;
++
++ err = netback_accel_debugfs_create(bend);
++ if (err)
++ goto fail_debugfs;
++
++ mutex_unlock(&bend->bend_mutex);
++
++ err = setup_config_accel_watch(dev, bend);
++ if (err)
++ goto fail_config_watch;
++
++ err = setup_domu_accel_watch(dev, bend);
++ if (err)
++ goto fail_domu_watch;
++
++ /*
++ * Indicate to the other end that we're ready to start unless
++ * the watch has already fired.
++ */
++ mutex_lock(&bend->bend_mutex);
++ if (bend->backend_state == XenbusStateClosed) {
++ bend->backend_state = XenbusStateInitialising;
++ net_accel_update_state(dev, XenbusStateInitialising);
++ }
++ mutex_unlock(&bend->bend_mutex);
++
++ mutex_lock(&bend_list_mutex);
++ link_bend(bend);
++ mutex_unlock(&bend_list_mutex);
++
++ return 0;
++
++fail_domu_watch:
++
++ unregister_xenbus_watch(&bend->config_accel_watch);
++ kfree(bend->config_accel_watch.node);
++fail_config_watch:
++
++ /*
++ * Flush the scheduled work queue before freeing bend to get
++ * rid of any pending netback_accel_msg_rx_handler()
++ */
++ flush_scheduled_work();
++
++ mutex_lock(&bend->bend_mutex);
++ net_accel_update_state(dev, XenbusStateUnknown);
++ netback_accel_debugfs_remove(bend);
++fail_debugfs:
++
++ unpublish_frontend_name(dev);
++fail_publish:
++
++ /* No need to reverse netback_accel_sf_hwtype. */
++fail_init_type:
++
++ kfree(bend->nicname);
++fail_nicname:
++ binfo->netback_accel_priv = NULL;
++ mutex_unlock(&bend->bend_mutex);
++ kfree(bend);
++ return err;
++}
++
++
++int netback_accel_remove(struct xenbus_device *dev)
++{
++ struct backend_info *binfo;
++ struct netback_accel *bend;
++ int frontend_state;
++
++ binfo = (struct backend_info *) dev->dev.driver_data;
++ bend = (struct netback_accel *) binfo->netback_accel_priv;
++
++ DPRINTK("%s: dev %p bend %p\n", __FUNCTION__, dev, bend);
++
++ BUG_ON(bend == NULL);
++
++ mutex_lock(&bend_list_mutex);
++ unlink_bend(bend);
++ mutex_unlock(&bend_list_mutex);
++
++ mutex_lock(&bend->bend_mutex);
++
++ /* Reject any requests to connect. */
++ bend->removing = 1;
++
++ /*
++ * Switch to closing to tell the other end that we're going
++ * away.
++ */
++ if (bend->backend_state != XenbusStateClosing) {
++ bend->backend_state = XenbusStateClosing;
++ net_accel_update_state(dev, XenbusStateClosing);
++ }
++
++ frontend_state = (int)XenbusStateUnknown;
++ xenbus_scanf(XBT_NIL, dev->otherend, "accelstate", "%d",
++ &frontend_state);
++
++ mutex_unlock(&bend->bend_mutex);
++
++ /*
++ * Wait until this end goes to the closed state. This happens
++ * in response to the other end going to the closed state.
++ * Don't bother doing this if the other end is already closed
++ * because if it is then there is nothing to do.
++ */
++ if (frontend_state != (int)XenbusStateClosed &&
++ frontend_state != (int)XenbusStateUnknown)
++ wait_event(bend->state_wait_queue,
++ bend->backend_state == XenbusStateClosed);
++
++ unregister_xenbus_watch(&bend->domu_accel_watch);
++ kfree(bend->domu_accel_watch.node);
++
++ unregister_xenbus_watch(&bend->config_accel_watch);
++ kfree(bend->config_accel_watch.node);
++
++ /*
++ * Flush the scheduled work queue before freeing bend to get
++ * rid of any pending netback_accel_msg_rx_handler()
++ */
++ flush_scheduled_work();
++
++ mutex_lock(&bend->bend_mutex);
++
++ /* Tear down the vnic if it was set up. */
++ if (bend->vnic_is_setup) {
++ bend->vnic_is_setup = 0;
++ cleanup_vnic(bend);
++ }
++
++ bend->backend_state = XenbusStateUnknown;
++ net_accel_update_state(dev, XenbusStateUnknown);
++
++ netback_accel_debugfs_remove(bend);
++
++ unpublish_frontend_name(dev);
++
++ kfree(bend->nicname);
++
++ binfo->netback_accel_priv = NULL;
++
++ mutex_unlock(&bend->bend_mutex);
++
++ kfree(bend);
++
++ return 0;
++}
++
++
++void netback_accel_shutdown_bends(void)
++{
++ mutex_lock(&bend_list_mutex);
++ /*
++ * I think we should have had a remove callback for all
++ * interfaces before being allowed to unload the module
++ */
++ BUG_ON(bend_list != NULL);
++ mutex_unlock(&bend_list_mutex);
++}
++
++
++void netback_accel_set_closing(struct netback_accel *bend)
++{
++
++ bend->backend_state = XenbusStateClosing;
++ net_accel_update_state((struct xenbus_device *)bend->hdev_data,
++ XenbusStateClosing);
++}
+diff --git a/drivers/xen/sfc_netback/ci/compat.h b/drivers/xen/sfc_netback/ci/compat.h
+new file mode 100644
+index 0000000..79f96f2
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat.h
+@@ -0,0 +1,53 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Compatability layer. Provides definitions of fundamental
++ * types and definitions that are used throughout CI source
++ * code. It does not introduce any link time dependencies,
++ * or include any unnecessary system headers.
++ */
++/*! \cidoxg_include_ci */
++
++#ifndef __CI_COMPAT_H__
++#define __CI_COMPAT_H__
++
++#ifdef __cplusplus
++extern "C" {
++#endif
++
++#include <ci/compat/primitive.h>
++#include <ci/compat/sysdep.h>
++#include <ci/compat/utils.h>
++
++
++#ifdef __cplusplus
++}
++#endif
++
++#endif /* __CI_COMPAT_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/gcc.h b/drivers/xen/sfc_netback/ci/compat/gcc.h
+new file mode 100644
+index 0000000..0cf77c4
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/gcc.h
+@@ -0,0 +1,158 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_GCC_H__
++#define __CI_COMPAT_GCC_H__
++
++
++#define CI_HAVE_INT64
++
++
++#if defined(__linux__) && defined(__KERNEL__)
++
++# include <linux/types.h>
++
++typedef __u64 ci_uint64;
++typedef __s64 ci_int64;
++# if BITS_PER_LONG == 32
++typedef __s32 ci_ptr_arith_t;
++typedef __u32 ci_uintptr_t;
++# else
++typedef __s64 ci_ptr_arith_t;
++typedef __u64 ci_uintptr_t;
++# endif
++
++
++/* it's not obvious to me why the below is wrong for x64_64, but
++ * gcc seems to complain on this platform
++ */
++# if defined(__ia64__)
++# define CI_PRId64 "ld"
++# define CI_PRIi64 "li"
++# define CI_PRIo64 "lo"
++# define CI_PRIu64 "lu"
++# define CI_PRIx64 "lx"
++# define CI_PRIX64 "lX"
++# else
++# define CI_PRId64 "lld"
++# define CI_PRIi64 "lli"
++# define CI_PRIo64 "llo"
++# define CI_PRIu64 "llu"
++# define CI_PRIx64 "llx"
++# define CI_PRIX64 "llX"
++# endif
++
++# define CI_PRId32 "d"
++# define CI_PRIi32 "i"
++# define CI_PRIo32 "o"
++# define CI_PRIu32 "u"
++# define CI_PRIx32 "x"
++# define CI_PRIX32 "X"
++
++#else
++
++# include <stdint.h>
++# include <inttypes.h>
++
++typedef uint64_t ci_uint64;
++typedef int64_t ci_int64;
++typedef intptr_t ci_ptr_arith_t;
++typedef uintptr_t ci_uintptr_t;
++
++# define CI_PRId64 PRId64
++# define CI_PRIi64 PRIi64
++# define CI_PRIo64 PRIo64
++# define CI_PRIu64 PRIu64
++# define CI_PRIx64 PRIx64
++# define CI_PRIX64 PRIX64
++
++# define CI_PRId32 PRId32
++# define CI_PRIi32 PRIi32
++# define CI_PRIo32 PRIo32
++# define CI_PRIu32 PRIu32
++# define CI_PRIx32 PRIx32
++# define CI_PRIX32 PRIX32
++
++#endif
++
++
++typedef ci_uint64 ci_fixed_descriptor_t;
++
++#define from_fixed_descriptor(desc) ((ci_uintptr_t)(desc))
++#define to_fixed_descriptor(desc) ((ci_fixed_descriptor_t)(ci_uintptr_t)(desc))
++
++
++#if __GNUC__ >= 3 && !defined(__cplusplus)
++/*
++** Checks that [p_mbr] has the same type as [&c_type::mbr_name].
++*/
++# define CI_CONTAINER(c_type, mbr_name, p_mbr) \
++ __builtin_choose_expr( \
++ __builtin_types_compatible_p(__typeof__(&((c_type*)0)->mbr_name), \
++ __typeof__(p_mbr)), \
++ __CI_CONTAINER(c_type, mbr_name, p_mbr), (void)0)
++
++# define ci_restrict __restrict__
++#endif
++
++
++#if !defined(__KERNEL__) || defined(__unix__)
++#define CI_HAVE_NPRINTF 1
++#endif
++
++
++/* At what version was this introduced? */
++#if __GNUC__ >= 3 || (__GNUC__ == 2 && __GNUC_MINOR__ > 91)
++# define CI_LIKELY(t) __builtin_expect((t), 1)
++# define CI_UNLIKELY(t) __builtin_expect((t), 0)
++#endif
++
++/**********************************************************************
++ * Attributes
++ */
++#if __GNUC__ >= 3 && defined(NDEBUG)
++# define CI_HF __attribute__((visibility("hidden")))
++# define CI_HV __attribute__((visibility("hidden")))
++#else
++# define CI_HF
++# define CI_HV
++#endif
++
++#if __GNUC__ >= 4 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)
++# define ci_noinline static __attribute__((__noinline__))
++/* (Linux 2.6 defines its own "noinline", so we use the "__noinline__" form) */
++#else
++# define ci_noinline static
++#endif
++
++#define CI_ALIGN(x) __attribute__ ((aligned (x)))
++
++#define CI_PRINTF_LIKE(a,b) __attribute__((format(printf,a,b)))
++
++#endif /* __CI_COMPAT_GCC_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/gcc_x86.h b/drivers/xen/sfc_netback/ci/compat/gcc_x86.h
+new file mode 100644
+index 0000000..438f0ba
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/gcc_x86.h
+@@ -0,0 +1,115 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_GCC_X86_H__
++#define __CI_COMPAT_GCC_X86_H__
++
++/*
++** The facts:
++**
++** SSE sfence
++** SSE2 lfence, mfence, pause
++*/
++
++/*
++ Barriers to enforce ordering with respect to:
++
++ normal memory use: ci_wmb, ci_rmb, ci_wmb
++ IO bus access use: ci_wiob, ci_riob, ci_iob
++*/
++#if defined(__x86_64__)
++# define ci_x86_mb() __asm__ __volatile__ ("lock; addl $0,0(%%rsp)":::"memory")
++#else
++# define ci_x86_mb() __asm__ __volatile__ ("lock; addl $0,0(%%esp)":::"memory")
++#endif
++
++/* ?? measure the impact of latency of sfence on a modern processor before we
++ take a decision on how to integrate with respect to writecombining */
++
++/* DJR: I don't think we need to add "memory" here. It means the asm does
++** something to memory that GCC doesn't understand. But all this does is
++** commit changes that GCC thinks have already happened. NB. GCC will not
++** reorder across a __volatile__ __asm__ anyway.
++*/
++#define ci_gcc_fence() __asm__ __volatile__ ("")
++
++#if __GNUC__ >= 3 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 96)
++# define ci_x86_sfence() __asm__ __volatile__ ("sfence")
++# define ci_x86_lfence() __asm__ __volatile__ ("lfence")
++# define ci_x86_mfence() __asm__ __volatile__ ("mfence")
++#else
++# define ci_x86_sfence() __asm__ __volatile__ (".byte 0x0F, 0xAE, 0xF8")
++# define ci_x86_lfence() __asm__ __volatile__ (".byte 0x0F, 0xAE, 0xE8")
++# define ci_x86_mfence() __asm__ __volatile__ (".byte 0x0F, 0xAE, 0xF0")
++#endif
++
++
++/* x86 processors to P4 Xeon store in-order unless executing streaming
++ extensions or when using writecombining
++
++ Hence we do not define ci_wmb to use sfence by default. Requirement is that
++ we do not use writecombining to memory and any code which uses SSE
++ extensions must call sfence directly
++
++ We need to track non intel clones which may support out of order store.
++
++*/
++
++#if CI_CPU_OOS
++# if CI_CPU_HAS_SSE
++# define ci_wmb() ci_x86_sfence()
++# else
++# define ci_wmb() ci_x86_mb()
++# endif
++#else
++# define ci_wmb() ci_gcc_fence()
++#endif
++
++#if CI_CPU_HAS_SSE2
++# define ci_rmb() ci_x86_lfence()
++# define ci_mb() ci_x86_mfence()
++# define ci_riob() ci_x86_lfence()
++# define ci_wiob() ci_x86_sfence()
++# define ci_iob() ci_x86_mfence()
++#else
++# if CI_CPU_HAS_SSE
++# define ci_wiob() ci_x86_sfence()
++# else
++# define ci_wiob() ci_x86_mb()
++# endif
++# define ci_rmb() ci_x86_mb()
++# define ci_mb() ci_x86_mb()
++# define ci_riob() ci_x86_mb()
++# define ci_iob() ci_x86_mb()
++#endif
++
++typedef unsigned long ci_phys_addr_t;
++#define ci_phys_addr_fmt "%lx"
++
++#endif /* __CI_COMPAT_GCC_X86_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/primitive.h b/drivers/xen/sfc_netback/ci/compat/primitive.h
+new file mode 100644
+index 0000000..3e58685
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/primitive.h
+@@ -0,0 +1,77 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_PRIMITIVE_H__
++#define __CI_COMPAT_PRIMITIVE_H__
++
++
++/**********************************************************************
++ * Primitive types.
++ */
++
++typedef unsigned char ci_uint8;
++typedef char ci_int8;
++
++typedef unsigned short ci_uint16;
++typedef short ci_int16;
++
++typedef unsigned int ci_uint32;
++typedef int ci_int32;
++
++/* 64-bit support is platform dependent. */
++
++
++/**********************************************************************
++ * Other fancy types.
++ */
++
++typedef ci_uint8 ci_octet;
++
++typedef enum {
++ CI_FALSE = 0,
++ CI_TRUE
++} ci_boolean_t;
++
++
++/**********************************************************************
++ * Some nice types you'd always assumed were standards.
++ * (Really, they are SYSV "standards".)
++ */
++
++#ifdef _WIN32
++typedef unsigned long ulong;
++typedef unsigned int uint;
++typedef char* caddr_t;
++#elif defined(__linux__) && defined(__KERNEL__)
++#include <linux/types.h>
++#elif defined(__linux__)
++#include <sys/types.h>
++#endif
++
++
++#endif /* __CI_COMPAT_PRIMITIVE_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/sysdep.h b/drivers/xen/sfc_netback/ci/compat/sysdep.h
+new file mode 100644
+index 0000000..7f7423c
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/sysdep.h
+@@ -0,0 +1,166 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_SYSDEP_H__
++#define __CI_COMPAT_SYSDEP_H__
++
++
++/**********************************************************************
++ * Platform definition fixups.
++ */
++
++#if defined(__ci_ul_driver__) && !defined(__ci_driver__)
++# define __ci_driver__
++#endif
++
++#if defined(__ci_driver__) && !defined(__ci_ul_driver__) && \
++ !defined(__KERNEL__)
++# define __KERNEL__
++#endif
++
++
++/**********************************************************************
++ * Sanity checks (no cheating!)
++ */
++
++#if defined(__KERNEL__) && !defined(__ci_driver__)
++# error Insane.
++#endif
++
++#if defined(__KERNEL__) && defined(__ci_ul_driver__)
++# error Madness.
++#endif
++
++#if defined(__unix__) && defined(_WIN32)
++# error Strange.
++#endif
++
++#if defined(__GNUC__) && defined(_MSC_VER)
++# error Crazy.
++#endif
++
++
++/**********************************************************************
++ * Compiler and processor dependencies.
++ */
++
++#if defined(__GNUC__)
++
++# include <ci/compat/gcc.h>
++
++# if defined(__i386__)
++# include <ci/compat/x86.h>
++# include <ci/compat/gcc_x86.h>
++# elif defined(__x86_64__)
++# include <ci/compat/x86_64.h>
++# include <ci/compat/gcc_x86.h>
++# elif defined(__PPC__)
++# include <ci/compat/ppc.h>
++# include <ci/compat/gcc_ppc.h>
++# elif defined(__ia64__)
++# include <ci/compat/ia64.h>
++# include <ci/compat/gcc_ia64.h>
++# else
++# error Unknown processor - GNU C
++# endif
++
++#elif defined(_MSC_VER)
++
++# include <ci/compat/msvc.h>
++
++# if defined(__i386__)
++# include <ci/compat/x86.h>
++# include <ci/compat/msvc_x86.h>
++# elif defined(__x86_64__)
++# include <ci/compat/x86_64.h>
++# include <ci/compat/msvc_x86_64.h>
++# else
++# error Unknown processor MSC
++# endif
++
++#elif defined(__PGI)
++
++# include <ci/compat/x86.h>
++# include <ci/compat/pg_x86.h>
++
++#elif defined(__INTEL_COMPILER)
++
++/* Intel compilers v7 claim to be very gcc compatible. */
++# if __INTEL_COMPILER >= 700
++# include <ci/compat/gcc.h>
++# include <ci/compat/x86.h>
++# include <ci/compat/gcc_x86.h>
++# else
++# error Old Intel compiler not supported. Yet.
++# endif
++
++#else
++# error Unknown compiler.
++#endif
++
++
++/**********************************************************************
++ * Misc stuff (that probably shouldn't be here).
++ */
++
++#ifdef __sun
++# ifdef __KERNEL__
++# define _KERNEL
++# define _SYSCALL32
++# ifdef _LP64
++# define _SYSCALL32_IMPL
++# endif
++# else
++# define _REENTRANT
++# endif
++#endif
++
++
++/**********************************************************************
++ * Defaults for anything left undefined.
++ */
++
++#ifndef CI_LIKELY
++# define CI_LIKELY(t) (t)
++# define CI_UNLIKELY(t) (t)
++#endif
++
++#ifndef ci_restrict
++# define ci_restrict
++#endif
++
++#ifndef ci_inline
++# define ci_inline static inline
++#endif
++
++#ifndef ci_noinline
++# define ci_noinline static
++#endif
++
++#endif /* __CI_COMPAT_SYSDEP_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/utils.h b/drivers/xen/sfc_netback/ci/compat/utils.h
+new file mode 100644
+index 0000000..34d4c99
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/utils.h
+@@ -0,0 +1,269 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Handy utility macros.
++ * \date 2003/01/17
++ */
++
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_UTILS_H__
++#define __CI_COMPAT_UTILS_H__
++
++
++/**********************************************************************
++ * Alignment -- [align] must be a power of 2.
++ **********************************************************************/
++
++ /*! Align forward onto next boundary. */
++
++#define CI_ALIGN_FWD(p, align) (((p)+(align)-1u) & ~((align)-1u))
++
++
++ /*! Align back onto prev boundary. */
++
++#define CI_ALIGN_BACK(p, align) ((p) & ~((align)-1u))
++
++
++ /*! How far to next boundary? */
++
++#define CI_ALIGN_NEEDED(p, align, signed_t) (-(signed_t)(p) & ((align)-1u))
++
++
++ /*! How far beyond prev boundary? */
++
++#define CI_OFFSET(p, align) ((p) & ((align)-1u))
++
++
++ /*! Does object fit in gap before next boundary? */
++
++#define CI_FITS(p, size, align, signed_t) \
++ (CI_ALIGN_NEEDED((p) + 1, (align), signed_t) + 1 >= (size))
++
++
++ /*! Align forward onto next boundary. */
++
++#define CI_PTR_ALIGN_FWD(p, align) \
++ ((char*) CI_ALIGN_FWD(((ci_ptr_arith_t)(p)), ((ci_ptr_arith_t)(align))))
++
++ /*! Align back onto prev boundary. */
++
++#define CI_PTR_ALIGN_BACK(p, align) \
++ ((char*) CI_ALIGN_BACK(((ci_ptr_arith_t)(p)), ((ci_ptr_arith_t)(align))))
++
++ /*! How far to next boundary? */
++
++#define CI_PTR_ALIGN_NEEDED(p, align) \
++ CI_ALIGN_NEEDED(((ci_ptr_arith_t)(p)), ((ci_ptr_arith_t)(align)), \
++ ci_ptr_arith_t)
++
++ /*! How far to next boundary? NZ = not zero i.e. give align if on boundary */
++
++#define CI_PTR_ALIGN_NEEDED_NZ(p, align) \
++ ((align) - (((char*)p) - \
++ ((char*) CI_ALIGN_BACK(((ci_ptr_arith_t)(p)), ((ci_ptr_arith_t)(align))))))
++
++ /*! How far beyond prev boundary? */
++
++#define CI_PTR_OFFSET(p, align) \
++ CI_OFFSET(((ci_ptr_arith_t)(p)), ((ci_ptr_arith_t)(align)))
++
++
++ /* Same as CI_ALIGN_FWD and CI_ALIGN_BACK. */
++
++#define CI_ROUND_UP(i, align) (((i)+(align)-1u) & ~((align)-1u))
++
++#define CI_ROUND_DOWN(i, align) ((i) & ~((align)-1u))
++
++
++/**********************************************************************
++ * Byte-order
++ **********************************************************************/
++
++/* These are not flags. They are enumeration values for use with
++ * CI_MY_BYTE_ORDER. */
++#define CI_BIG_ENDIAN 1
++#define CI_LITTLE_ENDIAN 0
++
++/*
++** Note that these byte-swapping primitives may leave junk in bits above
++** the range they operate on.
++**
++** The CI_BSWAP_nn() routines require that bits above [nn] are zero. Use
++** CI_BSWAPM_nn(x) if this cannot be guaranteed.
++*/
++
++/* ?? May be able to improve on some of these with inline assembler on some
++** platforms.
++*/
++
++#define CI_BSWAP_16(v) ((((v) & 0xff) << 8) | ((v) >> 8))
++#define CI_BSWAPM_16(v) ((((v) & 0xff) << 8) | (((v) & 0xff00) >> 8))
++
++#define CI_BSWAP_32(v) (((v) >> 24) | \
++ (((v) & 0x00ff0000) >> 8) | \
++ (((v) & 0x0000ff00) << 8) | \
++ ((v) << 24))
++#define CI_BSWAPM_32(v) ((((v) & 0xff000000) >> 24) | \
++ (((v) & 0x00ff0000) >> 8) | \
++ (((v) & 0x0000ff00) << 8) | \
++ ((v) << 24))
++
++#define CI_BSWAP_64(v) (((v) >> 56) | \
++ (((v) & 0x00ff000000000000) >> 40) | \
++ (((v) & 0x0000ff0000000000) >> 24) | \
++ (((v) & 0x000000ff00000000) >> 8) | \
++ (((v) & 0x00000000ff000000) << 8) | \
++ (((v) & 0x0000000000ff0000) << 24) | \
++ (((v) & 0x000000000000ff00) << 40) | \
++ ((v) << 56))
++
++# define CI_BSWAPPED_16_IF(c,v) ((c) ? CI_BSWAP_16(v) : (v))
++# define CI_BSWAPPED_32_IF(c,v) ((c) ? CI_BSWAP_32(v) : (v))
++# define CI_BSWAPPED_64_IF(c,v) ((c) ? CI_BSWAP_64(v) : (v))
++# define CI_BSWAP_16_IF(c,v) do{ if((c)) (v) = CI_BSWAP_16(v); }while(0)
++# define CI_BSWAP_32_IF(c,v) do{ if((c)) (v) = CI_BSWAP_32(v); }while(0)
++# define CI_BSWAP_64_IF(c,v) do{ if((c)) (v) = CI_BSWAP_64(v); }while(0)
++
++#if (CI_MY_BYTE_ORDER == CI_LITTLE_ENDIAN)
++# define CI_BSWAP_LE16(v) (v)
++# define CI_BSWAP_LE32(v) (v)
++# define CI_BSWAP_LE64(v) (v)
++# define CI_BSWAP_BE16(v) CI_BSWAP_16(v)
++# define CI_BSWAP_BE32(v) CI_BSWAP_32(v)
++# define CI_BSWAP_BE64(v) CI_BSWAP_64(v)
++# define CI_BSWAPM_LE16(v) (v)
++# define CI_BSWAPM_LE32(v) (v)
++# define CI_BSWAPM_LE64(v) (v)
++# define CI_BSWAPM_BE16(v) CI_BSWAPM_16(v)
++# define CI_BSWAPM_BE32(v) CI_BSWAPM_32(v)
++#elif (CI_MY_BYTE_ORDER == CI_BIG_ENDIAN)
++# define CI_BSWAP_BE16(v) (v)
++# define CI_BSWAP_BE32(v) (v)
++# define CI_BSWAP_BE64(v) (v)
++# define CI_BSWAP_LE16(v) CI_BSWAP_16(v)
++# define CI_BSWAP_LE32(v) CI_BSWAP_32(v)
++# define CI_BSWAP_LE64(v) CI_BSWAP_64(v)
++# define CI_BSWAPM_BE16(v) (v)
++# define CI_BSWAPM_BE32(v) (v)
++# define CI_BSWAPM_BE64(v) (v)
++# define CI_BSWAPM_LE16(v) CI_BSWAPM_16(v)
++# define CI_BSWAPM_LE32(v) CI_BSWAPM_32(v)
++#else
++# error Bad endian.
++#endif
++
++
++/**********************************************************************
++ * Get pointer to struct from pointer to member
++ **********************************************************************/
++
++#define CI_MEMBER_OFFSET(c_type, mbr_name) \
++ ((ci_uint32) (ci_uintptr_t)(&((c_type*)0)->mbr_name))
++
++#define CI_MEMBER_SIZE(c_type, mbr_name) \
++ sizeof(((c_type*)0)->mbr_name)
++
++#define __CI_CONTAINER(c_type, mbr_name, p_mbr) \
++ ( (c_type*) ((char*)(p_mbr) - CI_MEMBER_OFFSET(c_type, mbr_name)) )
++
++#ifndef CI_CONTAINER
++# define CI_CONTAINER(t,m,p) __CI_CONTAINER(t,m,p)
++#endif
++
++
++/**********************************************************************
++ * Structure member initialiser.
++ **********************************************************************/
++
++#ifndef CI_STRUCT_MBR
++# define CI_STRUCT_MBR(name, val) .name = val
++#endif
++
++
++/**********************************************************************
++ * min / max
++ **********************************************************************/
++
++#define CI_MIN(x,y) (((x) < (y)) ? (x) : (y))
++#define CI_MAX(x,y) (((x) > (y)) ? (x) : (y))
++
++/**********************************************************************
++ * abs
++ **********************************************************************/
++
++#define CI_ABS(x) (((x) < 0) ? -(x) : (x))
++
++/**********************************************************************
++ * Conditional debugging
++ **********************************************************************/
++
++#ifdef NDEBUG
++# define CI_DEBUG(x)
++# define CI_NDEBUG(x) x
++# define CI_IF_DEBUG(y,n) (n)
++# define CI_DEBUG_ARG(x)
++#else
++# define CI_DEBUG(x) x
++# define CI_NDEBUG(x)
++# define CI_IF_DEBUG(y,n) (y)
++# define CI_DEBUG_ARG(x) ,x
++#endif
++
++#ifdef __KERNEL__
++#define CI_KERNEL_ARG(x) ,x
++#else
++#define CI_KERNEL_ARG(x)
++#endif
++
++#ifdef _WIN32
++# define CI_KERNEL_ARG_WIN(x) CI_KERNEL_ARG(x)
++# define CI_ARG_WIN(x) ,x
++#else
++# define CI_KERNEL_ARG_WIN(x)
++# define CI_ARG_WIN(x)
++#endif
++
++#ifdef __unix__
++# define CI_KERNEL_ARG_UNIX(x) CI_KERNEL_ARG(x)
++# define CI_ARG_UNIX(x) ,x
++#else
++# define CI_KERNEL_ARG_UNIX(x)
++# define CI_ARG_UNIX(x)
++#endif
++
++#ifdef __linux__
++# define CI_KERNEL_ARG_LINUX(x) CI_KERNEL_ARG(x)
++# define CI_ARG_LINUX(x) ,x
++#else
++# define CI_KERNEL_ARG_LINUX(x)
++# define CI_ARG_LINUX(x)
++#endif
++
++
++#endif /* __CI_COMPAT_UTILS_H__ */
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/x86.h b/drivers/xen/sfc_netback/ci/compat/x86.h
+new file mode 100644
+index 0000000..2c1dfb3
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/x86.h
+@@ -0,0 +1,48 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_compat */
++
++#ifndef __CI_COMPAT_X86_H__
++#define __CI_COMPAT_X86_H__
++
++
++#define CI_MY_BYTE_ORDER CI_LITTLE_ENDIAN
++
++#define CI_WORD_SIZE 4
++#define CI_PTR_SIZE 4
++
++#define CI_PAGE_SIZE 4096
++#define CI_PAGE_SHIFT 12
++#define CI_PAGE_MASK (~(CI_PAGE_SIZE - 1))
++
++#define CI_CPU_HAS_SSE 1 /* SSE extensions supported */
++#define CI_CPU_HAS_SSE2 0 /* SSE2 extensions supported */
++#define CI_CPU_OOS 0 /* CPU does out of order stores */
++
++
++#endif /* __CI_COMPAT_X86_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/compat/x86_64.h b/drivers/xen/sfc_netback/ci/compat/x86_64.h
+new file mode 100644
+index 0000000..c09f540
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/compat/x86_64.h
+@@ -0,0 +1,54 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Arch stuff for AMD x86_64.
++ * \date 2004/08/17
++ */
++
++/*! \cidoxg_include_ci_compat */
++#ifndef __CI_COMPAT_X86_64_H__
++#define __CI_COMPAT_X86_64_H__
++
++
++#define CI_MY_BYTE_ORDER CI_LITTLE_ENDIAN
++
++#define CI_WORD_SIZE 8
++#define CI_PTR_SIZE 8
++
++#define CI_PAGE_SIZE 4096
++#define CI_PAGE_SHIFT 12
++#define CI_PAGE_MASK (~(CI_PAGE_SIZE - 1))
++
++#define CI_CPU_HAS_SSE 1 /* SSE extensions supported */
++
++/* SSE2 disabled while investigating BUG1060 */
++#define CI_CPU_HAS_SSE2 0 /* SSE2 extensions supported */
++#define CI_CPU_OOS 0 /* CPU does out of order stores */
++
++
++#endif /* __CI_COMPAT_X86_64_H__ */
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/driver/resource/efx_vi.h b/drivers/xen/sfc_netback/ci/driver/resource/efx_vi.h
+new file mode 100644
+index 0000000..3c74fc8
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/driver/resource/efx_vi.h
+@@ -0,0 +1,276 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file contains public EFX VI API to Solarflare resource manager.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_DRIVER_RESOURCE_EFX_VI_H__
++#define __CI_DRIVER_RESOURCE_EFX_VI_H__
++
++/* Default size of event queue in the efx_vi resource. Copied from
++ * CI_CFG_NETIF_EVENTQ_SIZE */
++#define EFX_VI_EVENTQ_SIZE_DEFAULT 1024
++
++extern int efx_vi_eventq_size;
++
++/**************************************************************************
++ * efx_vi_state types, allocation and free
++ **************************************************************************/
++
++/*! Handle for refering to a efx_vi */
++struct efx_vi_state;
++
++/*!
++ * Allocate an efx_vi, including event queue and pt_endpoint
++ *
++ * \param vih_out Pointer to a handle that is set on success
++ * \param nic_index Index of NIC to apply this resource to
++ * \return Zero on success (and vih_out set), non-zero on failure.
++ */
++extern int
++efx_vi_alloc(struct efx_vi_state **vih_out, int nic_index);
++
++/*!
++ * Free a previously allocated efx_vi
++ *
++ * \param vih The handle of the efx_vi to free
++ */
++extern void
++efx_vi_free(struct efx_vi_state *vih);
++
++/*!
++ * Reset a previously allocated efx_vi
++ *
++ * \param vih The handle of the efx_vi to reset
++ */
++extern void
++efx_vi_reset(struct efx_vi_state *vih);
++
++/**************************************************************************
++ * efx_vi_eventq types and functions
++ **************************************************************************/
++
++/*!
++ * Register a function to receive callbacks when event queue timeouts
++ * or wakeups occur. Only one function per efx_vi can be registered
++ * at once.
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param callback The function to callback
++ * \param context An argument to pass to the callback function
++ * \return Zero on success, non-zero on failure.
++ */
++extern int
++efx_vi_eventq_register_callback(struct efx_vi_state *vih,
++ void (*callback)(void *context, int is_timeout),
++ void *context);
++
++/*!
++ * Remove the current eventq timeout or wakeup callback function
++ *
++ * \param vih The handle to identify the efx_vi
++ * \return Zero on success, non-zero on failure
++ */
++extern int
++efx_vi_eventq_kill_callback(struct efx_vi_state *vih);
++
++/**************************************************************************
++ * efx_vi_dma_map types and functions
++ **************************************************************************/
++
++/*!
++ * Handle for refering to a efx_vi
++ */
++struct efx_vi_dma_map_state;
++
++/*!
++ * Map a list of buffer pages so they are registered with the hardware
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param addrs An array of page pointers to map
++ * \param n_addrs Length of the page pointer array. Must be a power of two.
++ * \param dmh_out Set on success to a handle used to refer to this mapping
++ * \return Zero on success, non-zero on failure.
++ */
++extern int
++efx_vi_dma_map_pages(struct efx_vi_state *vih, struct page **pages,
++ int n_pages, struct efx_vi_dma_map_state **dmh_out);
++extern int
++efx_vi_dma_map_addrs(struct efx_vi_state *vih,
++ unsigned long long *dev_bus_addrs, int n_pages,
++ struct efx_vi_dma_map_state **dmh_out);
++
++/*!
++ * Unmap a previously mapped set of pages so they are no longer registered
++ * with the hardware.
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param dmh The handle to identify the dma mapping
++ */
++extern void
++efx_vi_dma_unmap_pages(struct efx_vi_state *vih,
++ struct efx_vi_dma_map_state *dmh);
++extern void
++efx_vi_dma_unmap_addrs(struct efx_vi_state *vih,
++ struct efx_vi_dma_map_state *dmh);
++
++/*!
++ * Retrieve the buffer address of the mapping
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param dmh The handle to identify the buffer mapping
++ * \return The buffer address on success, or zero on failure
++ */
++extern unsigned
++efx_vi_dma_get_map_addr(struct efx_vi_state *vih,
++ struct efx_vi_dma_map_state *dmh);
++
++/**************************************************************************
++ * efx_vi filter functions
++ **************************************************************************/
++
++#define EFX_VI_STATIC_FILTERS 32
++
++/*! Handle to refer to a filter instance */
++struct filter_resource_t;
++
++/*!
++ * Allocate and add a filter
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param protocol The protocol of the new filter: UDP or TCP
++ * \param ip_addr_be32 The local ip address of the filter
++ * \param port_le16 The local port of the filter
++ * \param fh_out Set on success to be a handle to refer to this filter
++ * \return Zero on success, non-zero on failure.
++ */
++extern int
++efx_vi_filter(struct efx_vi_state *vih, int protocol, unsigned ip_addr_be32,
++ int port_le16, struct filter_resource_t **fh_out);
++
++/*!
++ * Remove a filter and free resources associated with it
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param fh The handle to identify the filter
++ * \return Zero on success, non-zero on failure
++ */
++extern int
++efx_vi_filter_stop(struct efx_vi_state *vih, struct filter_resource_t *fh);
++
++/**************************************************************************
++ * efx_vi hw resources types and functions
++ **************************************************************************/
++
++/*! Constants for the type field in efx_vi_hw_resource */
++#define EFX_VI_HW_RESOURCE_TXDMAQ 0x0 /* PFN of TX DMA Q */
++#define EFX_VI_HW_RESOURCE_RXDMAQ 0x1 /* PFN of RX DMA Q */
++#define EFX_VI_HW_RESOURCE_TXBELL 0x2 /* PFN of TX Doorbell (EF1) */
++#define EFX_VI_HW_RESOURCE_RXBELL 0x3 /* PFN of RX Doorbell (EF1) */
++#define EFX_VI_HW_RESOURCE_EVQTIMER 0x4 /* Address of event q timer */
++
++/* Address of event q pointer (EF1) */
++#define EFX_VI_HW_RESOURCE_EVQPTR 0x5
++/* Address of register pointer (Falcon A) */
++#define EFX_VI_HW_RESOURCE_EVQRPTR 0x6
++/* Offset of register pointer (Falcon B) */
++#define EFX_VI_HW_RESOURCE_EVQRPTR_OFFSET 0x7
++/* Address of mem KVA */
++#define EFX_VI_HW_RESOURCE_EVQMEMKVA 0x8
++/* PFN of doorbell page (Falcon) */
++#define EFX_VI_HW_RESOURCE_BELLPAGE 0x9
++
++/*! How large an array to allocate for the get_() functions - smaller
++ than the total number of constants as some are mutually exclusive */
++#define EFX_VI_HW_RESOURCE_MAXSIZE 0x7
++
++/*! Constants for the mem_type field in efx_vi_hw_resource */
++#define EFX_VI_HW_RESOURCE_IOBUFFER 0 /* Host memory */
++#define EFX_VI_HW_RESOURCE_PERIPHERAL 1 /* Card memory/registers */
++
++/*!
++ * Data structure providing information on a hardware resource mapping
++ */
++struct efx_vi_hw_resource {
++ u8 type; /*!< What this resource represents */
++ u8 mem_type; /*!< What type of memory is it in, eg,
++ * host or iomem */
++ u8 more_to_follow; /*!< Is this part of a multi-region resource */
++ u32 length; /*!< Length of the resource in bytes */
++ unsigned long address; /*!< Address of this resource */
++};
++
++/*!
++ * Metadata concerning the list of hardware resource mappings
++ */
++struct efx_vi_hw_resource_metadata {
++ int version;
++ int evq_order;
++ int evq_offs;
++ int evq_capacity;
++ int instance;
++ unsigned rx_capacity;
++ unsigned tx_capacity;
++ int nic_arch;
++ int nic_revision;
++ char nic_variant;
++};
++
++/*!
++ * Obtain a list of hardware resource mappings, using virtual addresses
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param mdata Pointer to a structure to receive the metadata
++ * \param hw_res_array An array to receive the list of hardware resources
++ * \param length The length of hw_res_array. Updated on success to contain
++ * the number of entries in the supplied array that were used.
++ * \return Zero on success, non-zero on failure
++ */
++extern int
++efx_vi_hw_resource_get_virt(struct efx_vi_state *vih,
++ struct efx_vi_hw_resource_metadata *mdata,
++ struct efx_vi_hw_resource *hw_res_array,
++ int *length);
++
++/*!
++ * Obtain a list of hardware resource mappings, using physical addresses
++ *
++ * \param vih The handle to identify the efx_vi
++ * \param mdata Pointer to a structure to receive the metadata
++ * \param hw_res_array An array to receive the list of hardware resources
++ * \param length The length of hw_res_array. Updated on success to contain
++ * the number of entries in the supplied array that were used.
++ * \return Zero on success, non-zero on failure
++ */
++extern int
++efx_vi_hw_resource_get_phys(struct efx_vi_state *vih,
++ struct efx_vi_hw_resource_metadata *mdata,
++ struct efx_vi_hw_resource *hw_res_array,
++ int *length);
++
++#endif /* __CI_DRIVER_RESOURCE_EFX_VI_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/common.h b/drivers/xen/sfc_netback/ci/efhw/common.h
+new file mode 100644
+index 0000000..b1e460d
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/common.h
+@@ -0,0 +1,102 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides API of the efhw library which may be used both from
++ * the kernel and from the user-space code.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_COMMON_H__
++#define __CI_EFHW_COMMON_H__
++
++#include <ci/efhw/common_sysdep.h>
++
++enum efhw_arch {
++ EFHW_ARCH_FALCON,
++ EFHW_ARCH_SIENA,
++};
++
++typedef uint32_t efhw_buffer_addr_t;
++#define EFHW_BUFFER_ADDR_FMT "[ba:%"PRIx32"]"
++
++/*! Comment? */
++typedef union {
++ uint64_t u64;
++ struct {
++ uint32_t a;
++ uint32_t b;
++ } opaque;
++ struct {
++ uint32_t code;
++ uint32_t status;
++ } ev1002;
++} efhw_event_t;
++
++/* Flags for TX/RX queues */
++#define EFHW_VI_JUMBO_EN 0x01 /*! scatter RX over multiple desc */
++#define EFHW_VI_ISCSI_RX_HDIG_EN 0x02 /*! iscsi rx header digest */
++#define EFHW_VI_ISCSI_TX_HDIG_EN 0x04 /*! iscsi tx header digest */
++#define EFHW_VI_ISCSI_RX_DDIG_EN 0x08 /*! iscsi rx data digest */
++#define EFHW_VI_ISCSI_TX_DDIG_EN 0x10 /*! iscsi tx data digest */
++#define EFHW_VI_TX_PHYS_ADDR_EN 0x20 /*! TX physical address mode */
++#define EFHW_VI_RX_PHYS_ADDR_EN 0x40 /*! RX physical address mode */
++#define EFHW_VI_RM_WITH_INTERRUPT 0x80 /*! VI with an interrupt */
++#define EFHW_VI_TX_IP_CSUM_DIS 0x100 /*! enable ip checksum generation */
++#define EFHW_VI_TX_TCPUDP_CSUM_DIS 0x200 /*! enable tcp/udp checksum
++ generation */
++#define EFHW_VI_TX_TCPUDP_ONLY 0x400 /*! drop non-tcp/udp packets */
++
++/* Types of hardware filter */
++/* Each of these values implicitly selects scatter filters on B0 - or in
++ EFHW_IP_FILTER_TYPE_NOSCAT_B0_MASK if a non-scatter filter is required */
++#define EFHW_IP_FILTER_TYPE_UDP_WILDCARD (0) /* dest host only */
++#define EFHW_IP_FILTER_TYPE_UDP_FULL (1) /* dest host and port */
++#define EFHW_IP_FILTER_TYPE_TCP_WILDCARD (2) /* dest based filter */
++#define EFHW_IP_FILTER_TYPE_TCP_FULL (3) /* src filter */
++/* Same again, but with RSS (for B0 only) */
++#define EFHW_IP_FILTER_TYPE_UDP_WILDCARD_RSS_B0 (4)
++#define EFHW_IP_FILTER_TYPE_UDP_FULL_RSS_B0 (5)
++#define EFHW_IP_FILTER_TYPE_TCP_WILDCARD_RSS_B0 (6)
++#define EFHW_IP_FILTER_TYPE_TCP_FULL_RSS_B0 (7)
++
++#define EFHW_IP_FILTER_TYPE_FULL_MASK (0x1) /* Mask for full / wildcard */
++#define EFHW_IP_FILTER_TYPE_TCP_MASK (0x2) /* Mask for TCP type */
++#define EFHW_IP_FILTER_TYPE_RSS_B0_MASK (0x4) /* Mask for B0 RSS enable */
++#define EFHW_IP_FILTER_TYPE_NOSCAT_B0_MASK (0x8) /* Mask for B0 SCATTER dsbl */
++
++#define EFHW_IP_FILTER_TYPE_MASK (0xffff) /* Mask of types above */
++
++#define EFHW_IP_FILTER_BROADCAST (0x10000) /* driverlink filter
++ support */
++
++#endif /* __CI_EFHW_COMMON_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/common_sysdep.h b/drivers/xen/sfc_netback/ci/efhw/common_sysdep.h
+new file mode 100644
+index 0000000..aeaccf0
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/common_sysdep.h
+@@ -0,0 +1,67 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides version-independent Linux kernel API for
++ * userland-to-kernel interfaces.
++ * Only kernels >=2.6.9 are supported.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_COMMON_LINUX_H__
++#define __CI_EFHW_COMMON_LINUX_H__
++
++#include <linux/types.h>
++#include <linux/version.h>
++
++/* Dirty hack, but Linux kernel does not provide DMA_ADDR_T_FMT */
++#if BITS_PER_LONG == 64 || defined(CONFIG_HIGHMEM64G)
++#define DMA_ADDR_T_FMT "%llx"
++#else
++#define DMA_ADDR_T_FMT "%x"
++#endif
++
++/* Linux kernel also does not provide PRIx32... Sigh. */
++#define PRIx32 "x"
++#define PRIx64 "llx"
++
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,19)
++enum {
++ false = 0,
++ true = 1
++};
++
++typedef _Bool bool;
++#endif /* LINUX_VERSION_CODE < 2.6.19 */
++
++#endif /* __CI_EFHW_COMMON_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/debug.h b/drivers/xen/sfc_netback/ci/efhw/debug.h
+new file mode 100644
+index 0000000..89d7a60
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/debug.h
+@@ -0,0 +1,84 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides debug-related API for efhw library using Linux kernel
++ * primitives.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_DEBUG_LINUX_H__
++#define __CI_EFHW_DEBUG_LINUX_H__
++
++#define EFHW_PRINTK_PREFIX "[sfc efhw] "
++
++#define EFHW_PRINTK(level, fmt, ...) \
++ printk(level EFHW_PRINTK_PREFIX fmt "\n", __VA_ARGS__)
++
++/* Following macros should be used with non-zero format parameters
++ * due to __VA_ARGS__ limitations. Use "%s" with __FUNCTION__ if you can't
++ * find better parameters. */
++#define EFHW_ERR(fmt, ...) EFHW_PRINTK(KERN_ERR, fmt, __VA_ARGS__)
++#define EFHW_WARN(fmt, ...) EFHW_PRINTK(KERN_WARNING, fmt, __VA_ARGS__)
++#define EFHW_NOTICE(fmt, ...) EFHW_PRINTK(KERN_NOTICE, fmt, __VA_ARGS__)
++#if 0 && !defined(NDEBUG)
++#define EFHW_TRACE(fmt, ...) EFHW_PRINTK(KERN_DEBUG, fmt, __VA_ARGS__)
++#else
++#define EFHW_TRACE(fmt, ...)
++#endif
++
++#ifndef NDEBUG
++#define EFHW_ASSERT(cond) BUG_ON((cond) == 0)
++#define EFHW_DO_DEBUG(expr) expr
++#else
++#define EFHW_ASSERT(cond)
++#define EFHW_DO_DEBUG(expr)
++#endif
++
++#define EFHW_TEST(expr) \
++ do { \
++ if (unlikely(!(expr))) \
++ BUG(); \
++ } while (0)
++
++/* Build time asserts. We paste the line number into the type name
++ * so that the macro can be used more than once per file even if the
++ * compiler objects to multiple identical typedefs. Collisions
++ * between use in different header files is still possible. */
++#ifndef EFHW_BUILD_ASSERT
++#define __EFHW_BUILD_ASSERT_NAME(_x) __EFHW_BUILD_ASSERT_ILOATHECPP(_x)
++#define __EFHW_BUILD_ASSERT_ILOATHECPP(_x) __EFHW_BUILD_ASSERT__ ##_x
++#define EFHW_BUILD_ASSERT(e) \
++ typedef char __EFHW_BUILD_ASSERT_NAME(__LINE__)[(e) ? 1 : -1]
++#endif
++
++#endif /* __CI_EFHW_DEBUG_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/efhw_config.h b/drivers/xen/sfc_netback/ci/efhw/efhw_config.h
+new file mode 100644
+index 0000000..8837ed4
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/efhw_config.h
+@@ -0,0 +1,43 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides some limits used in both kernel and userland code.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_EFAB_CONFIG_H__
++#define __CI_EFHW_EFAB_CONFIG_H__
++
++#define EFHW_MAX_NR_DEVS 5 /* max number of efhw devices supported */
++
++#endif /* __CI_EFHW_EFAB_CONFIG_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/efhw_types.h b/drivers/xen/sfc_netback/ci/efhw/efhw_types.h
+new file mode 100644
+index 0000000..e0d590e
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/efhw_types.h
+@@ -0,0 +1,342 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides struct efhw_nic and some related types.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_EFAB_TYPES_H__
++#define __CI_EFHW_EFAB_TYPES_H__
++
++#include <ci/efhw/efhw_config.h>
++#include <ci/efhw/hardware_sysdep.h>
++#include <ci/efhw/iopage_types.h>
++#include <ci/efhw/sysdep.h>
++
++/*--------------------------------------------------------------------
++ *
++ * hardware limits used in the types
++ *
++ *--------------------------------------------------------------------*/
++
++#define EFHW_KEVENTQ_MAX 8
++
++/*--------------------------------------------------------------------
++ *
++ * forward type declarations
++ *
++ *--------------------------------------------------------------------*/
++
++struct efhw_nic;
++
++/*--------------------------------------------------------------------
++ *
++ * Managed interface
++ *
++ *--------------------------------------------------------------------*/
++
++struct efhw_buffer_table_allocation{
++ unsigned base;
++ unsigned order;
++};
++
++struct eventq_resource_hardware {
++ /*!iobuffer allocated for eventq - can be larger than eventq */
++ efhw_iopages_t iobuff;
++ unsigned iobuff_off;
++ struct efhw_buffer_table_allocation buf_tbl_alloc;
++ int capacity; /*!< capacity of event queue */
++};
++
++/*--------------------------------------------------------------------
++ *
++ * event queues and event driven callbacks
++ *
++ *--------------------------------------------------------------------*/
++
++struct efhw_keventq {
++ volatile int lock;
++ caddr_t evq_base;
++ int32_t evq_ptr;
++ uint32_t evq_mask;
++ unsigned instance;
++ struct eventq_resource_hardware hw;
++ struct efhw_ev_handler *ev_handlers;
++};
++
++/**********************************************************************
++ * Portable HW interface. ***************************************
++ **********************************************************************/
++
++/*--------------------------------------------------------------------
++ *
++ * EtherFabric Functional units - configuration and control
++ *
++ *--------------------------------------------------------------------*/
++
++struct efhw_func_ops {
++
++ /*-------------- Initialisation ------------ */
++
++ /*! close down all hardware functional units - leaves NIC in a safe
++ state for driver unload */
++ void (*close_hardware) (struct efhw_nic *nic);
++
++ /*! initialise all hardware functional units */
++ int (*init_hardware) (struct efhw_nic *nic,
++ struct efhw_ev_handler *,
++ const uint8_t *mac_addr);
++
++ /*-------------- Interrupt support ------------ */
++
++ /*! Main interrupt routine
++ ** This function returns,
++ ** - zero, if the IRQ was not generated by EF1
++ ** - non-zero, if EF1 was the source of the IRQ
++ **
++ **
++ ** opaque is an OS provided pointer for use by the OS callbacks
++ ** e.g in Windows used to indicate DPC scheduled
++ */
++ int (*interrupt) (struct efhw_nic *nic);
++
++ /*! Enable given interrupt mask for the given IRQ unit */
++ void (*interrupt_enable) (struct efhw_nic *nic, uint idx);
++
++ /*! Disable given interrupt mask for the given IRQ unit */
++ void (*interrupt_disable) (struct efhw_nic *nic, uint idx);
++
++ /*! Set interrupt moderation strategy for the given IRQ unit
++ ** val is in usec
++ */
++ void (*set_interrupt_moderation)(struct efhw_nic *nic,
++ uint idx, uint val);
++
++ /*-------------- Event support ------------ */
++
++ /*! Enable the given event queue
++ depending on the underlying implementation (EF1 or Falcon) then
++ either a q_base_addr in host memory, or a buffer base id should
++ be proivded
++ */
++ void (*event_queue_enable) (struct efhw_nic *nic,
++ uint evq, /* evnt queue index */
++ uint evq_size, /* units of #entries */
++ dma_addr_t q_base_addr, uint buf_base_id);
++
++ /*! Disable the given event queue (and any associated timer) */
++ void (*event_queue_disable) (struct efhw_nic *nic, uint evq,
++ int timer_only);
++
++ /*! request wakeup from the NIC on a given event Q */
++ void (*wakeup_request) (struct efhw_nic *nic, dma_addr_t q_base_addr,
++ int next_i, int evq);
++
++ /*! Push a SW event on a given eventQ */
++ void (*sw_event) (struct efhw_nic *nic, int data, int evq);
++
++ /*-------------- Filter support ------------ */
++
++ /*! Setup a given filter - The software can request a filter_i,
++ * but some EtherFabric implementations will override with
++ * a more suitable index
++ */
++ int (*ipfilter_set) (struct efhw_nic *nic, int type,
++ int *filter_i, int dmaq,
++ unsigned saddr_be32, unsigned sport_be16,
++ unsigned daddr_be32, unsigned dport_be16);
++
++ /*! Attach a given filter to a DMAQ */
++ void (*ipfilter_attach) (struct efhw_nic *nic, int filter_idx,
++ int dmaq_idx);
++
++ /*! Detach a filter from its DMAQ */
++ void (*ipfilter_detach) (struct efhw_nic *nic, int filter_idx);
++
++ /*! Clear down a given filter */
++ void (*ipfilter_clear) (struct efhw_nic *nic, int filter_idx);
++
++ /*-------------- DMA support ------------ */
++
++ /*! Initialise NIC state for a given TX DMAQ */
++ void (*dmaq_tx_q_init) (struct efhw_nic *nic,
++ uint dmaq, uint evq, uint owner, uint tag,
++ uint dmaq_size, uint buf_idx, uint flags);
++
++ /*! Initialise NIC state for a given RX DMAQ */
++ void (*dmaq_rx_q_init) (struct efhw_nic *nic,
++ uint dmaq, uint evq, uint owner, uint tag,
++ uint dmaq_size, uint buf_idx, uint flags);
++
++ /*! Disable a given TX DMAQ */
++ void (*dmaq_tx_q_disable) (struct efhw_nic *nic, uint dmaq);
++
++ /*! Disable a given RX DMAQ */
++ void (*dmaq_rx_q_disable) (struct efhw_nic *nic, uint dmaq);
++
++ /*! Flush a given TX DMA channel */
++ int (*flush_tx_dma_channel) (struct efhw_nic *nic, uint dmaq);
++
++ /*! Flush a given RX DMA channel */
++ int (*flush_rx_dma_channel) (struct efhw_nic *nic, uint dmaq);
++
++ /*-------------- Buffer table Support ------------ */
++
++ /*! Initialise a buffer table page */
++ void (*buffer_table_set) (struct efhw_nic *nic,
++ dma_addr_t dma_addr,
++ uint bufsz, uint region,
++ int own_id, int buffer_id);
++
++ /*! Initialise a block of buffer table pages */
++ void (*buffer_table_set_n) (struct efhw_nic *nic, int buffer_id,
++ dma_addr_t dma_addr,
++ uint bufsz, uint region,
++ int n_pages, int own_id);
++
++ /*! Clear a block of buffer table pages */
++ void (*buffer_table_clear) (struct efhw_nic *nic, int buffer_id,
++ int num);
++
++ /*! Commit a buffer table update */
++ void (*buffer_table_commit) (struct efhw_nic *nic);
++
++};
++
++
++/*----------------------------------------------------------------------------
++ *
++ * NIC type
++ *
++ *---------------------------------------------------------------------------*/
++
++struct efhw_device_type {
++ int arch; /* enum efhw_arch */
++ char variant; /* 'A', 'B', ... */
++ int revision; /* 0, 1, ... */
++};
++
++
++/*----------------------------------------------------------------------------
++ *
++ * EtherFabric NIC instance - nic.c for HW independent functions
++ *
++ *---------------------------------------------------------------------------*/
++
++/*! */
++struct efhw_nic {
++ /*! zero base index in efrm_nic_table.nic array */
++ volatile int index;
++ int ifindex; /*!< OS level nic index */
++#ifdef HAS_NET_NAMESPACE
++ struct net *nd_net;
++#endif
++
++ struct efhw_device_type devtype;
++
++ /*! Options that can be set by user. */
++ unsigned options;
++# define NIC_OPT_EFTEST 0x1 /* owner is an eftest app */
++
++# define NIC_OPT_DEFAULT 0
++
++ /*! Internal flags that indicate hardware properties at runtime. */
++ unsigned flags;
++# define NIC_FLAG_NO_INTERRUPT 0x01 /* to be set at init time only */
++# define NIC_FLAG_TRY_MSI 0x02
++# define NIC_FLAG_MSI 0x04
++# define NIC_FLAG_OS_IRQ_EN 0x08
++# define NIC_FLAG_10G 0x10
++
++ unsigned mtu; /*!< MAC MTU (includes MAC hdr) */
++
++ /* hardware resources */
++
++ /*! I/O address of the start of the bar */
++ efhw_ioaddr_t bar_ioaddr;
++
++ /*! Bar number of control aperture. */
++ unsigned ctr_ap_bar;
++ /*! Length of control aperture in bytes. */
++ unsigned ctr_ap_bytes;
++
++ uint8_t mac_addr[ETH_ALEN]; /*!< mac address */
++
++ /*! EtherFabric Functional Units -- functions */
++ const struct efhw_func_ops *efhw_func;
++
++ /* Value read from FPGA version register. Zero for asic. */
++ unsigned fpga_version;
++
++ /*! This lock protects a number of misc NIC resources. It should
++ * only be used for things that can be at the bottom of the lock
++ * order. ie. You mustn't attempt to grab any other lock while
++ * holding this one.
++ */
++ spinlock_t *reg_lock;
++ spinlock_t the_reg_lock;
++
++ int buf_commit_outstanding; /*!< outstanding buffer commits */
++
++ /*! interrupt callbacks (hard-irq) */
++ void (*irq_handler) (struct efhw_nic *, int unit);
++
++ /*! event queues per driver */
++ struct efhw_keventq evq[EFHW_KEVENTQ_MAX];
++
++/* for marking when we are not using an IRQ unit
++ - 0 is a valid offset to an IRQ unit on EF1! */
++#define EFHW_IRQ_UNIT_UNUSED 0xffff
++ /*! interrupt unit in use */
++ unsigned int irq_unit[EFHW_KEVENTQ_MAX];
++ efhw_iopage_t irq_iobuff; /*!< Falcon SYSERR interrupt */
++
++ /* The new driverlink infrastructure. */
++ struct efx_dl_device *net_driver_dev;
++ struct efx_dlfilt_cb_s *dlfilter_cb;
++
++ /*! Bit masks of the sizes of event queues and dma queues supported
++ * by the nic. */
++ unsigned evq_sizes;
++ unsigned rxq_sizes;
++ unsigned txq_sizes;
++
++ /* Size of filter table (including odd and even banks). */
++ unsigned filter_tbl_size;
++};
++
++
++#define EFHW_KVA(nic) ((nic)->bar_ioaddr)
++
++
++#endif /* __CI_EFHW_EFHW_TYPES_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/hardware_sysdep.h b/drivers/xen/sfc_netback/ci/efhw/hardware_sysdep.h
+new file mode 100644
+index 0000000..a31cf41
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/hardware_sysdep.h
+@@ -0,0 +1,84 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides version-independent Linux kernel API for header files
++ * with hardware-related definitions (in ci/driver/efab/hardware*).
++ * Only kernels >=2.6.9 are supported.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_HARDWARE_LINUX_H__
++#define __CI_EFHW_HARDWARE_LINUX_H__
++
++#include <asm/io.h>
++
++#ifdef __LITTLE_ENDIAN
++#define EFHW_IS_LITTLE_ENDIAN
++#elif __BIG_ENDIAN
++#define EFHW_IS_BIG_ENDIAN
++#else
++#error Unknown endianness
++#endif
++
++#ifndef mmiowb
++ #if defined(__i386__) || defined(__x86_64__)
++ #define mmiowb()
++ #elif defined(__ia64__)
++ #ifndef ia64_mfa
++ #define ia64_mfa() asm volatile ("mf.a" ::: "memory")
++ #endif
++ #define mmiowb ia64_mfa
++ #else
++ #error "Need definition for mmiowb()"
++ #endif
++#endif
++
++typedef char *efhw_ioaddr_t;
++
++#ifndef readq
++static inline uint64_t __readq(void __iomem *addr)
++{
++ return *(volatile uint64_t *)addr;
++}
++#define readq(x) __readq(x)
++#endif
++
++#ifndef writeq
++static inline void __writeq(uint64_t v, void __iomem *addr)
++{
++ *(volatile uint64_t *)addr = v;
++}
++#define writeq(val, addr) __writeq((val), (addr))
++#endif
++
++#endif /* __CI_EFHW_HARDWARE_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/iopage_types.h b/drivers/xen/sfc_netback/ci/efhw/iopage_types.h
+new file mode 100644
+index 0000000..565d513
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/iopage_types.h
+@@ -0,0 +1,188 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides efhw_page_t and efhw_iopage_t for Linux kernel.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_IOPAGE_LINUX_H__
++#define __CI_EFHW_IOPAGE_LINUX_H__
++
++#include <linux/gfp.h>
++#include <linux/hardirq.h>
++#include <ci/efhw/debug.h>
++
++/*--------------------------------------------------------------------
++ *
++ * efhw_page_t: A single page of memory. Directly mapped in the driver,
++ * and can be mapped to userlevel.
++ *
++ *--------------------------------------------------------------------*/
++
++typedef struct {
++ unsigned long kva;
++} efhw_page_t;
++
++static inline int efhw_page_alloc(efhw_page_t *p)
++{
++ p->kva = __get_free_page(in_interrupt()? GFP_ATOMIC : GFP_KERNEL);
++ return p->kva ? 0 : -ENOMEM;
++}
++
++static inline int efhw_page_alloc_zeroed(efhw_page_t *p)
++{
++ p->kva = get_zeroed_page(in_interrupt()? GFP_ATOMIC : GFP_KERNEL);
++ return p->kva ? 0 : -ENOMEM;
++}
++
++static inline void efhw_page_free(efhw_page_t *p)
++{
++ free_page(p->kva);
++ EFHW_DO_DEBUG(memset(p, 0, sizeof(*p)));
++}
++
++static inline char *efhw_page_ptr(efhw_page_t *p)
++{
++ return (char *)p->kva;
++}
++
++static inline unsigned efhw_page_pfn(efhw_page_t *p)
++{
++ return (unsigned)(__pa(p->kva) >> PAGE_SHIFT);
++}
++
++static inline void efhw_page_mark_invalid(efhw_page_t *p)
++{
++ p->kva = 0;
++}
++
++static inline int efhw_page_is_valid(efhw_page_t *p)
++{
++ return p->kva != 0;
++}
++
++static inline void efhw_page_init_from_va(efhw_page_t *p, void *va)
++{
++ p->kva = (unsigned long)va;
++}
++
++/*--------------------------------------------------------------------
++ *
++ * efhw_iopage_t: A single page of memory. Directly mapped in the driver,
++ * and can be mapped to userlevel. Can also be accessed by the NIC.
++ *
++ *--------------------------------------------------------------------*/
++
++typedef struct {
++ efhw_page_t p;
++ dma_addr_t dma_addr;
++} efhw_iopage_t;
++
++static inline dma_addr_t efhw_iopage_dma_addr(efhw_iopage_t *p)
++{
++ return p->dma_addr;
++}
++
++#define efhw_iopage_ptr(iop) efhw_page_ptr(&(iop)->p)
++#define efhw_iopage_pfn(iop) efhw_page_pfn(&(iop)->p)
++#define efhw_iopage_mark_invalid(iop) efhw_page_mark_invalid(&(iop)->p)
++#define efhw_iopage_is_valid(iop) efhw_page_is_valid(&(iop)->p)
++
++/*--------------------------------------------------------------------
++ *
++ * efhw_iopages_t: A set of pages that are contiguous in physical memory.
++ * Directly mapped in the driver, and can be mapped to userlevel. Can also
++ * be accessed by the NIC.
++ *
++ * NB. The O/S may be unwilling to allocate many, or even any of these. So
++ * only use this type where the NIC really needs a physically contiguous
++ * buffer.
++ *
++ *--------------------------------------------------------------------*/
++
++typedef struct {
++ caddr_t kva;
++ unsigned order;
++ dma_addr_t dma_addr;
++} efhw_iopages_t;
++
++static inline caddr_t efhw_iopages_ptr(efhw_iopages_t *p)
++{
++ return p->kva;
++}
++
++static inline unsigned efhw_iopages_pfn(efhw_iopages_t *p)
++{
++ return (unsigned)(__pa(p->kva) >> PAGE_SHIFT);
++}
++
++static inline dma_addr_t efhw_iopages_dma_addr(efhw_iopages_t *p)
++{
++ return p->dma_addr;
++}
++
++static inline unsigned efhw_iopages_size(efhw_iopages_t *p)
++{
++ return 1u << (p->order + PAGE_SHIFT);
++}
++
++/* efhw_iopage_t <-> efhw_iopages_t conversions for handling physically
++ * contiguous allocations in iobufsets for iSCSI. This allows the
++ * essential information about contiguous allocations from
++ * efhw_iopages_alloc() to be saved away in the efhw_iopage_t array in an
++ * iobufset. (Changing the iobufset resource to use a union type would
++ * involve a lot of code changes, and make the iobufset's metadata larger
++ * which could be bad as it's supposed to fit into a single page on some
++ * platforms.)
++ */
++static inline void
++efhw_iopage_init_from_iopages(efhw_iopage_t *iopage,
++ efhw_iopages_t *iopages, unsigned pageno)
++{
++ iopage->p.kva = ((unsigned long)efhw_iopages_ptr(iopages))
++ + (pageno * PAGE_SIZE);
++ iopage->dma_addr = efhw_iopages_dma_addr(iopages) +
++ (pageno * PAGE_SIZE);
++}
++
++static inline void
++efhw_iopages_init_from_iopage(efhw_iopages_t *iopages,
++ efhw_iopage_t *iopage, unsigned order)
++{
++ iopages->kva = (caddr_t) efhw_iopage_ptr(iopage);
++ EFHW_ASSERT(iopages->kva);
++ iopages->order = order;
++ iopages->dma_addr = efhw_iopage_dma_addr(iopage);
++}
++
++#endif /* __CI_EFHW_IOPAGE_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/public.h b/drivers/xen/sfc_netback/ci/efhw/public.h
+new file mode 100644
+index 0000000..76c1c59
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/public.h
+@@ -0,0 +1,83 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides public API of efhw library exported from the SFC
++ * resource driver.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_PUBLIC_H__
++#define __CI_EFHW_PUBLIC_H__
++
++#include <ci/efhw/common.h>
++#include <ci/efhw/efhw_types.h>
++
++/*! Returns true if we have some EtherFabric functional units -
++ whether configured or not */
++static inline int efhw_nic_have_functional_units(struct efhw_nic *nic)
++{
++ return nic->efhw_func != 0;
++}
++
++/*! Returns true if the EtherFabric functional units have been configured */
++static inline int efhw_nic_have_hw(struct efhw_nic *nic)
++{
++ return efhw_nic_have_functional_units(nic) && (EFHW_KVA(nic) != 0);
++}
++
++/*! Helper function to allocate the iobuffer needed by an eventq
++ * - it ensures the eventq has the correct alignment for the NIC
++ *
++ * \param rm Event-queue resource manager
++ * \param instance Event-queue instance (index)
++ * \param buf_bytes Requested size of eventq
++ * \return < 0 if iobuffer allocation fails
++ */
++int efhw_nic_event_queue_alloc_iobuffer(struct efhw_nic *nic,
++ struct eventq_resource_hardware *h,
++ int evq_instance, unsigned buf_bytes);
++
++extern void falcon_nic_set_rx_usr_buf_size(struct efhw_nic *,
++ int rx_usr_buf_size);
++
++extern void
++falcon_nic_rx_filter_ctl_set(struct efhw_nic *nic, uint32_t tcp_full,
++ uint32_t tcp_wild,
++ uint32_t udp_full, uint32_t udp_wild);
++
++extern void
++falcon_nic_rx_filter_ctl_get(struct efhw_nic *nic, uint32_t *tcp_full,
++ uint32_t *tcp_wild,
++ uint32_t *udp_full, uint32_t *udp_wild);
++
++#endif /* __CI_EFHW_PUBLIC_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efhw/sysdep.h b/drivers/xen/sfc_netback/ci/efhw/sysdep.h
+new file mode 100644
+index 0000000..8368bdd
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efhw/sysdep.h
+@@ -0,0 +1,72 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides version-independent Linux kernel API for efhw library.
++ * Only kernels >=2.6.9 are supported.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFHW_SYSDEP_LINUX_H__
++#define __CI_EFHW_SYSDEP_LINUX_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/spinlock.h>
++#include <linux/delay.h>
++#include <linux/if_ether.h>
++
++#include <linux/netdevice.h> /* necessary for etherdevice.h on some kernels */
++#include <linux/etherdevice.h>
++
++#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,21)
++static inline int is_local_ether_addr(const u8 *addr)
++{
++ return (0x02 & addr[0]);
++}
++#endif
++
++typedef unsigned long irq_flags_t;
++
++#define spin_lock_destroy(l_) do {} while (0)
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,24)
++#define HAS_NET_NAMESPACE
++#endif
++
++/* Funny, but linux has round_up for x86 only, defined in
++ * x86-specific header */
++#ifndef round_up
++#define round_up(x, y) (((x) + (y) - 1) & ~((y)-1))
++#endif
++
++#endif /* __CI_EFHW_SYSDEP_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efrm/nic_table.h b/drivers/xen/sfc_netback/ci/efrm/nic_table.h
+new file mode 100644
+index 0000000..702719e
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efrm/nic_table.h
+@@ -0,0 +1,98 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides public API for NIC table.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFRM_NIC_TABLE_H__
++#define __CI_EFRM_NIC_TABLE_H__
++
++#include <ci/efhw/efhw_types.h>
++#include <ci/efrm/sysdep.h>
++
++/*--------------------------------------------------------------------
++ *
++ * struct efrm_nic_table - top level driver object keeping all NICs -
++ * implemented in driver_object.c
++ *
++ *--------------------------------------------------------------------*/
++
++/*! Comment? */
++struct efrm_nic_table {
++ /*! nics attached to this driver */
++ struct efhw_nic *nic[EFHW_MAX_NR_DEVS];
++ /*! pointer to an arbitrary struct efhw_nic if one exists;
++ * for code which does not care which NIC it wants but
++ * still needs one. Note you cannot assume nic[0] exists. */
++ struct efhw_nic *a_nic;
++ uint32_t nic_count; /*!< number of nics attached to this driver */
++ spinlock_t lock; /*!< lock for table modifications */
++ atomic_t ref_count; /*!< refcount for users of nic table */
++};
++
++/* Resource driver structures used by other drivers as well */
++extern struct efrm_nic_table efrm_nic_table;
++
++static inline void efrm_nic_table_hold(void)
++{
++ atomic_inc(&efrm_nic_table.ref_count);
++}
++
++static inline void efrm_nic_table_rele(void)
++{
++ atomic_dec(&efrm_nic_table.ref_count);
++}
++
++static inline int efrm_nic_table_held(void)
++{
++ return (atomic_read(&efrm_nic_table.ref_count) != 0);
++}
++
++/* Run code block _x multiple times with variable nic set to each
++ * registered NIC in turn.
++ * DO NOT "break" out of this loop early. */
++#define EFRM_FOR_EACH_NIC(_nic_i, _nic) \
++ for ((_nic_i) = (efrm_nic_table_hold(), 0); \
++ (_nic_i) < EFHW_MAX_NR_DEVS || (efrm_nic_table_rele(), 0); \
++ (_nic_i)++) \
++ if (((_nic) = efrm_nic_table.nic[_nic_i]))
++
++#define EFRM_FOR_EACH_NIC_IN_SET(_set, _i, _nic) \
++ for ((_i) = (efrm_nic_table_hold(), 0); \
++ (_i) < EFHW_MAX_NR_DEVS || (efrm_nic_table_rele(), 0); \
++ ++(_i)) \
++ if (((_nic) = efrm_nic_table.nic[_i]) && \
++ efrm_nic_set_read((_set), (_i)))
++
++#endif /* __CI_EFRM_NIC_TABLE_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efrm/sysdep.h b/drivers/xen/sfc_netback/ci/efrm/sysdep.h
+new file mode 100644
+index 0000000..97013b9
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efrm/sysdep.h
+@@ -0,0 +1,54 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides Linux-like system-independent API for efrm library.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFRM_SYSDEP_H__
++#define __CI_EFRM_SYSDEP_H__
++
++/* Spinlocks are defined in efhw/sysdep.h */
++#include <ci/efhw/sysdep.h>
++
++#if defined(__linux__) && defined(__KERNEL__)
++
++# include <ci/efrm/sysdep_linux.h>
++
++#else
++
++# include <ci/efrm/sysdep_ci2linux.h>
++
++#endif
++
++#endif /* __CI_EFRM_SYSDEP_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/efrm/sysdep_linux.h b/drivers/xen/sfc_netback/ci/efrm/sysdep_linux.h
+new file mode 100644
+index 0000000..3ca7854
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/efrm/sysdep_linux.h
+@@ -0,0 +1,248 @@
++/****************************************************************************
++ * Driver for Solarflare network controllers -
++ * resource management for Xen backend, OpenOnload, etc
++ * (including support for SFE4001 10GBT NIC)
++ *
++ * This file provides version-independent Linux kernel API for efrm library.
++ * Only kernels >=2.6.9 are supported.
++ *
++ * Copyright 2005-2007: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Kfifo API is partially stolen from linux-2.6.22/include/linux/list.h
++ * Copyright (C) 2004 Stelian Pop <stelian at popies.net>
++ *
++ * Developed and maintained by Solarflare Communications:
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * Certain parts of the driver were implemented by
++ * Alexandra Kossovsky <Alexandra.Kossovsky at oktetlabs.ru>
++ * OKTET Labs Ltd, Russia,
++ * http://oktetlabs.ru, <info at oktetlabs.ru>
++ * by request of Solarflare Communications
++ *
++ *
++ * 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, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef __CI_EFRM_SYSDEP_LINUX_H__
++#define __CI_EFRM_SYSDEP_LINUX_H__
++
++#include <linux/version.h>
++#include <linux/list.h>
++#include <linux/vmalloc.h>
++#include <linux/errno.h>
++#include <linux/string.h>
++#include <linux/workqueue.h>
++#include <linux/gfp.h>
++#include <linux/slab.h>
++#include <linux/hardirq.h>
++#include <linux/kernel.h>
++#include <linux/if_ether.h>
++#include <linux/completion.h>
++#include <linux/in.h>
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++/* get roundup_pow_of_two(), which was in kernel.h in early kernel versions */
++#include <linux/log2.h>
++#endif
++
++/********************************************************************
++ *
++ * List API
++ *
++ ********************************************************************/
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
++static inline void
++list_replace_init(struct list_head *old, struct list_head *new)
++{
++ new->next = old->next;
++ new->next->prev = new;
++ new->prev = old->prev;
++ new->prev->next = new;
++ INIT_LIST_HEAD(old);
++}
++#endif
++
++static inline struct list_head *list_pop(struct list_head *list)
++{
++ struct list_head *link = list->next;
++ list_del(link);
++ return link;
++}
++
++static inline struct list_head *list_pop_tail(struct list_head *list)
++{
++ struct list_head *link = list->prev;
++ list_del(link);
++ return link;
++}
++
++/********************************************************************
++ *
++ * Workqueue API
++ *
++ ********************************************************************/
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)
++#define NEED_OLD_WORK_API
++
++/**
++ * The old and new work function prototypes just change
++ * the type of the pointer in the only argument, so it's
++ * safe to cast one function type to the other
++ */
++typedef void (*efrm_old_work_func_t) (void *p);
++
++#undef INIT_WORK
++#define INIT_WORK(_work, _func) \
++ do { \
++ INIT_LIST_HEAD(&(_work)->entry); \
++ (_work)->pending = 0; \
++ PREPARE_WORK((_work), \
++ (efrm_old_work_func_t) (_func), \
++ (_work)); \
++ } while (0)
++
++#endif
++
++/********************************************************************
++ *
++ * Kfifo API
++ *
++ ********************************************************************/
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,10)
++
++#if !defined(RHEL_RELEASE_CODE) || (RHEL_RELEASE_CODE < 1029)
++typedef unsigned gfp_t;
++#endif
++
++#define HAS_NO_KFIFO
++
++struct kfifo {
++ unsigned char *buffer; /* the buffer holding the data */
++ unsigned int size; /* the size of the allocated buffer */
++ unsigned int in; /* data is added at offset (in % size) */
++ unsigned int out; /* data is extracted from off. (out % size) */
++ spinlock_t *lock; /* protects concurrent modifications */
++};
++
++extern struct kfifo *kfifo_init(unsigned char *buffer, unsigned int size,
++ gfp_t gfp_mask, spinlock_t *lock);
++extern struct kfifo *kfifo_alloc(unsigned int size, gfp_t gfp_mask,
++ spinlock_t *lock);
++extern void kfifo_free(struct kfifo *fifo);
++extern unsigned int __kfifo_put(struct kfifo *fifo,
++ unsigned char *buffer, unsigned int len);
++extern unsigned int __kfifo_get(struct kfifo *fifo,
++ unsigned char *buffer, unsigned int len);
++
++/**
++ * kfifo_put - puts some data into the FIFO
++ * @fifo: the fifo to be used.
++ * @buffer: the data to be added.
++ * @len: the length of the data to be added.
++ *
++ * This function copies at most @len bytes from the @buffer into
++ * the FIFO depending on the free space, and returns the number of
++ * bytes copied.
++ */
++static inline unsigned int
++kfifo_put(struct kfifo *fifo, unsigned char *buffer, unsigned int len)
++{
++ unsigned long flags;
++ unsigned int ret;
++
++ spin_lock_irqsave(fifo->lock, flags);
++
++ ret = __kfifo_put(fifo, buffer, len);
++
++ spin_unlock_irqrestore(fifo->lock, flags);
++
++ return ret;
++}
++
++/**
++ * kfifo_get - gets some data from the FIFO
++ * @fifo: the fifo to be used.
++ * @buffer: where the data must be copied.
++ * @len: the size of the destination buffer.
++ *
++ * This function copies at most @len bytes from the FIFO into the
++ * @buffer and returns the number of copied bytes.
++ */
++static inline unsigned int
++kfifo_get(struct kfifo *fifo, unsigned char *buffer, unsigned int len)
++{
++ unsigned long flags;
++ unsigned int ret;
++
++ spin_lock_irqsave(fifo->lock, flags);
++
++ ret = __kfifo_get(fifo, buffer, len);
++
++ /*
++ * optimization: if the FIFO is empty, set the indices to 0
++ * so we don't wrap the next time
++ */
++ if (fifo->in == fifo->out)
++ fifo->in = fifo->out = 0;
++
++ spin_unlock_irqrestore(fifo->lock, flags);
++
++ return ret;
++}
++
++/**
++ * __kfifo_len - returns the number of bytes available in the FIFO, no locking version
++ * @fifo: the fifo to be used.
++ */
++static inline unsigned int __kfifo_len(struct kfifo *fifo)
++{
++ return fifo->in - fifo->out;
++}
++
++/**
++ * kfifo_len - returns the number of bytes available in the FIFO
++ * @fifo: the fifo to be used.
++ */
++static inline unsigned int kfifo_len(struct kfifo *fifo)
++{
++ unsigned long flags;
++ unsigned int ret;
++
++ spin_lock_irqsave(fifo->lock, flags);
++
++ ret = __kfifo_len(fifo);
++
++ spin_unlock_irqrestore(fifo->lock, flags);
++
++ return ret;
++}
++
++#else
++#include <linux/kfifo.h>
++#endif
++
++static inline void kfifo_vfree(struct kfifo *fifo)
++{
++ vfree(fifo->buffer);
++ kfree(fifo);
++}
++
++#endif /* __CI_EFRM_SYSDEP_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netback/ci/tools/config.h b/drivers/xen/sfc_netback/ci/tools/config.h
+new file mode 100644
+index 0000000..fb802f9
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/config.h
+@@ -0,0 +1,49 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_tools */
++
++#ifndef __CI_TOOLS_CONFIG_H__
++#define __CI_TOOLS_CONFIG_H__
++
++
++/**********************************************************************
++ * Debugging.
++ */
++
++#define CI_INCLUDE_ASSERT_VALID 0
++
++/* Set non-zero to allow info about who has allocated what to appear in
++ * /proc/drivers/level5/mem.
++ * However - Note that doing so can lead to segfault when you unload the
++ * driver, and other weirdness. i.e. I don't think the code for is quite
++ * right (written by Oktet, hacked by gel), but it does work well enough to be
++ * useful.
++ */
++#define CI_MEMLEAK_DEBUG_ALLOC_TABLE 0
++
++
++#endif /* __CI_TOOLS_CONFIG_H__ */
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/tools/debug.h b/drivers/xen/sfc_netback/ci/tools/debug.h
+new file mode 100644
+index 0000000..a25c2c4
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/debug.h
+@@ -0,0 +1,336 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_tools */
++
++#ifndef __CI_TOOLS_DEBUG_H__
++#define __CI_TOOLS_DEBUG_H__
++
++#define CI_LOG_E(x) x /* errors */
++#define CI_LOG_W(x) x /* warnings */
++#define CI_LOG_I(x) x /* information */
++#define CI_LOG_V(x) x /* verbose */
++
++/* Build time asserts. We paste the line number into the type name
++ * so that the macro can be used more than once per file even if the
++ * compiler objects to multiple identical typedefs. Collisions
++ * between use in different header files is still possible. */
++#ifndef CI_BUILD_ASSERT
++#define __CI_BUILD_ASSERT_NAME(_x) __CI_BUILD_ASSERT_ILOATHECPP(_x)
++#define __CI_BUILD_ASSERT_ILOATHECPP(_x) __CI_BUILD_ASSERT__ ##_x
++#define CI_BUILD_ASSERT(e)\
++ typedef char __CI_BUILD_ASSERT_NAME(__LINE__)[(e)?1:-1]
++#endif
++
++
++#ifdef NDEBUG
++
++# define _ci_check(exp, file, line)
++# define _ci_assert2(e, x, y, file, line)
++# define _ci_assert(exp, file, line)
++# define _ci_assert_equal(exp1, exp2, file, line)
++# define _ci_assert_equiv(exp1, exp2, file, line)
++# define _ci_assert_nequal(exp1, exp2, file, line)
++# define _ci_assert_le(exp1, exp2, file, line)
++# define _ci_assert_lt(exp1, exp2, file, line)
++# define _ci_assert_ge(exp1, exp2, file, line)
++# define _ci_assert_gt(exp1, exp2, file, line)
++# define _ci_assert_impl(exp1, exp2, file, line)
++
++# define _ci_verify(exp, file, line) \
++ do { \
++ (void)(exp); \
++ } while (0)
++
++# define CI_DEBUG_TRY(exp) \
++ do { \
++ (void)(exp); \
++ } while (0)
++
++#define CI_TRACE(exp,fmt)
++#define CI_TRACE_INT(integer)
++#define CI_TRACE_INT32(integer)
++#define CI_TRACE_INT64(integer)
++#define CI_TRACE_UINT(integer)
++#define CI_TRACE_UINT32(integer)
++#define CI_TRACE_UINT64(integer)
++#define CI_TRACE_HEX(integer)
++#define CI_TRACE_HEX32(integer)
++#define CI_TRACE_HEX64(integer)
++#define CI_TRACE_PTR(pointer)
++#define CI_TRACE_STRING(string)
++#define CI_TRACE_MAC(mac)
++#define CI_TRACE_IP(ip_be32)
++#define CI_TRACE_ARP(arp_pkt)
++
++#else
++
++# define _CI_ASSERT_FMT "\nfrom %s:%d"
++
++# define _ci_check(exp, file, line) \
++ do { \
++ if (CI_UNLIKELY(!(exp))) \
++ ci_warn(("ci_check(%s)"_CI_ASSERT_FMT, #exp, \
++ (file), (line))); \
++ } while (0)
++
++/*
++ * NOTE: ci_fail() emits the file and line where the assert is actually
++ * coded.
++ */
++
++# define _ci_assert(exp, file, line) \
++ do { \
++ if (CI_UNLIKELY(!(exp))) \
++ ci_fail(("ci_assert(%s)"_CI_ASSERT_FMT, #exp, \
++ (file), (line))); \
++ } while (0)
++
++# define _ci_assert2(e, x, y, file, line) do { \
++ if(CI_UNLIKELY( ! (e) )) \
++ ci_fail(("ci_assert(%s)\nwhere [%s=%"CI_PRIx64"] " \
++ "[%s=%"CI_PRIx64"]\nat %s:%d\nfrom %s:%d", #e \
++ , #x, (ci_uint64)(ci_uintptr_t)(x) \
++ , #y, (ci_uint64)(ci_uintptr_t)(y), \
++ __FILE__, __LINE__, (file), (line))); \
++ } while (0)
++
++# define _ci_verify(exp, file, line) \
++ do { \
++ if (CI_UNLIKELY(!(exp))) \
++ ci_fail(("ci_verify(%s)"_CI_ASSERT_FMT, #exp, \
++ (file), (line))); \
++ } while (0)
++
++# define _ci_assert_equal(x, y, f, l) _ci_assert2((x)==(y), x, y, (f), (l))
++# define _ci_assert_nequal(x, y, f, l) _ci_assert2((x)!=(y), x, y, (f), (l))
++# define _ci_assert_le(x, y, f, l) _ci_assert2((x)<=(y), x, y, (f), (l))
++# define _ci_assert_lt(x, y, f, l) _ci_assert2((x)< (y), x, y, (f), (l))
++# define _ci_assert_ge(x, y, f, l) _ci_assert2((x)>=(y), x, y, (f), (l))
++# define _ci_assert_gt(x, y, f, l) _ci_assert2((x)> (y), x, y, (f), (l))
++# define _ci_assert_or(x, y, f, l) _ci_assert2((x)||(y), x, y, (f), (l))
++# define _ci_assert_impl(x, y, f, l) _ci_assert2(!(x) || (y), x, y, (f), (l))
++# define _ci_assert_equiv(x, y, f, l) _ci_assert2(!(x)== !(y), x, y, (f), (l))
++
++#define _ci_assert_equal_msg(exp1, exp2, msg, file, line) \
++ do { \
++ if (CI_UNLIKELY((exp1)!=(exp2))) \
++ ci_fail(("ci_assert_equal_msg(%s == %s) were " \
++ "(%"CI_PRIx64":%"CI_PRIx64") with msg[%c%c%c%c]" \
++ _CI_ASSERT_FMT, #exp1, #exp2, \
++ (ci_uint64)(ci_uintptr_t)(exp1), \
++ (ci_uint64)(ci_uintptr_t)(exp2), \
++ (((ci_uint32)msg) >> 24) && 0xff, \
++ (((ci_uint32)msg) >> 16) && 0xff, \
++ (((ci_uint32)msg) >> 8 ) && 0xff, \
++ (((ci_uint32)msg) ) && 0xff, \
++ (file), (line))); \
++ } while (0)
++
++# define CI_DEBUG_TRY(exp) CI_TRY(exp)
++
++#define CI_TRACE(exp,fmt) \
++ ci_log("%s:%d:%s] " #exp "=" fmt, \
++ __FILE__, __LINE__, __FUNCTION__, (exp))
++
++
++#define CI_TRACE_INT(integer) \
++ ci_log("%s:%d:%s] " #integer "=%d", \
++ __FILE__, __LINE__, __FUNCTION__, (integer))
++
++
++#define CI_TRACE_INT32(integer) \
++ ci_log("%s:%d:%s] " #integer "=%d", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_int32)integer))
++
++
++#define CI_TRACE_INT64(integer) \
++ ci_log("%s:%d:%s] " #integer "=%lld", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_int64)integer))
++
++
++#define CI_TRACE_UINT(integer) \
++ ci_log("%s:%d:%s] " #integer "=%ud", \
++ __FILE__, __LINE__, __FUNCTION__, (integer))
++
++
++#define CI_TRACE_UINT32(integer) \
++ ci_log("%s:%d:%s] " #integer "=%ud", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_uint32)integer))
++
++
++#define CI_TRACE_UINT64(integer) \
++ ci_log("%s:%d:%s] " #integer "=%ulld", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_uint64)integer))
++
++
++#define CI_TRACE_HEX(integer) \
++ ci_log("%s:%d:%s] " #integer "=0x%x", \
++ __FILE__, __LINE__, __FUNCTION__, (integer))
++
++
++#define CI_TRACE_HEX32(integer) \
++ ci_log("%s:%d:%s] " #integer "=0x%x", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_uint32)integer))
++
++
++#define CI_TRACE_HEX64(integer) \
++ ci_log("%s:%d:%s] " #integer "=0x%llx", \
++ __FILE__, __LINE__, __FUNCTION__, ((ci_uint64)integer))
++
++
++#define CI_TRACE_PTR(pointer) \
++ ci_log("%s:%d:%s] " #pointer "=0x%p", \
++ __FILE__, __LINE__, __FUNCTION__, (pointer))
++
++
++#define CI_TRACE_STRING(string) \
++ ci_log("%s:%d:%s] " #string "=%s", \
++ __FILE__, __LINE__, __FUNCTION__, (string))
++
++
++#define CI_TRACE_MAC(mac) \
++ ci_log("%s:%d:%s] " #mac "=" CI_MAC_PRINTF_FORMAT, \
++ __FILE__, __LINE__, __FUNCTION__, CI_MAC_PRINTF_ARGS(mac))
++
++
++#define CI_TRACE_IP(ip_be32) \
++ ci_log("%s:%d:%s] " #ip_be32 "=" CI_IP_PRINTF_FORMAT, __FILE__, \
++ __LINE__, __FUNCTION__, CI_IP_PRINTF_ARGS(&(ip_be32)))
++
++
++#define CI_TRACE_ARP(arp_pkt) \
++ ci_log("%s:%d:%s]\n"CI_ARP_PRINTF_FORMAT, \
++ __FILE__, __LINE__, __FUNCTION__, CI_ARP_PRINTF_ARGS(arp_pkt))
++
++#endif /* NDEBUG */
++
++#define ci_check(exp) \
++ _ci_check(exp, __FILE__, __LINE__)
++
++#define ci_assert(exp) \
++ _ci_assert(exp, __FILE__, __LINE__)
++
++#define ci_verify(exp) \
++ _ci_verify(exp, __FILE__, __LINE__)
++
++#define ci_assert_equal(exp1, exp2) \
++ _ci_assert_equal(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_equal_msg(exp1, exp2, msg) \
++ _ci_assert_equal_msg(exp1, exp2, msg, __FILE__, __LINE__)
++
++#define ci_assert_nequal(exp1, exp2) \
++ _ci_assert_nequal(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_le(exp1, exp2) \
++ _ci_assert_le(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_lt(exp1, exp2) \
++ _ci_assert_lt(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_ge(exp1, exp2) \
++ _ci_assert_ge(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_gt(exp1, exp2) \
++ _ci_assert_gt(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_impl(exp1, exp2) \
++ _ci_assert_impl(exp1, exp2, __FILE__, __LINE__)
++
++#define ci_assert_equiv(exp1, exp2) \
++ _ci_assert_equiv(exp1, exp2, __FILE__, __LINE__)
++
++
++#define CI_TEST(exp) \
++ do{ \
++ if( CI_UNLIKELY(!(exp)) ) \
++ ci_fail(("CI_TEST(%s)", #exp)); \
++ }while(0)
++
++
++#define CI_TRY(exp) \
++ do{ \
++ int _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(_trc < 0) ) \
++ ci_sys_fail(#exp, _trc); \
++ }while(0)
++
++
++#define CI_TRY_RET(exp) \
++ do{ \
++ int _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(_trc < 0) ) { \
++ ci_log("%s returned %d at %s:%d", #exp, _trc, __FILE__, __LINE__); \
++ return _trc; \
++ } \
++ }while(0)
++
++#define CI_LOGLEVEL_TRY_RET(logfn, exp) \
++ do{ \
++ int _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(_trc < 0) ) { \
++ logfn (ci_log("%s returned %d at %s:%d", #exp, _trc, __FILE__, __LINE__)); \
++ return _trc; \
++ } \
++ }while(0)
++
++
++#define CI_SOCK_TRY(exp) \
++ do{ \
++ ci_sock_err_t _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(!ci_sock_errok(_trc)) ) \
++ ci_sys_fail(#exp, _trc.val); \
++ }while(0)
++
++
++#define CI_SOCK_TRY_RET(exp) \
++ do{ \
++ ci_sock_err_t _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(!ci_sock_errok(_trc)) ) { \
++ ci_log("%s returned %d at %s:%d", #exp, _trc.val, __FILE__, __LINE__); \
++ return ci_sock_errcode(_trc); \
++ } \
++ }while(0)
++
++
++#define CI_SOCK_TRY_SOCK_RET(exp) \
++ do{ \
++ ci_sock_err_t _trc; \
++ _trc=(exp); \
++ if( CI_UNLIKELY(!ci_sock_errok(_trc)) ) { \
++ ci_log("%s returned %d at %s:%d", #exp, _trc.val, __FILE__, __LINE__); \
++ return _trc; \
++ } \
++ }while(0)
++
++#endif /* __CI_TOOLS_DEBUG_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/tools/log.h b/drivers/xen/sfc_netback/ci/tools/log.h
+new file mode 100644
+index 0000000..03a6e11
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/log.h
+@@ -0,0 +1,262 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Functions for logging and pretty-printing.
++ * \date 2002/08/07
++ */
++
++/*! \cidoxg_include_ci_tools */
++
++#ifndef __CI_TOOLS_LOG_H__
++#define __CI_TOOLS_LOG_H__
++
++#include <stdarg.h>
++
++
++/**********************************************************************
++ * Logging.
++ */
++
++/* size of internal log buffer */
++#define CI_LOG_MAX_LINE 512
++/* uses of ci_log must ensure that all trace messages are shorter than this */
++#define CI_LOG_MAX_MSG_LENGTH (CI_LOG_MAX_LINE-50)
++
++extern void ci_vlog(const char* fmt, va_list args) CI_HF;
++extern void ci_log(const char* fmt, ...) CI_PRINTF_LIKE(1,2) CI_HF;
++
++ /*! Set the prefix for log messages.
++ **
++ ** Uses the storage pointed to by \em prefix. Therefore \em prefix must
++ ** be allocated on the heap, or statically.
++ */
++extern void ci_set_log_prefix(const char* prefix) CI_HF;
++
++typedef void (*ci_log_fn_t)(const char* msg);
++extern ci_log_fn_t ci_log_fn CI_HV;
++
++/* Log functions. */
++extern void ci_log_null(const char* msg) CI_HF;
++extern void ci_log_stderr(const char* msg) CI_HF;
++extern void ci_log_stdout(const char* msg) CI_HF;
++extern void ci_log_syslog(const char* msg) CI_HF;
++
++/*! Call the following to install special logging behaviours. */
++extern void ci_log_buffer_till_fail(void) CI_HF;
++extern void ci_log_buffer_till_exit(void) CI_HF;
++
++extern void __ci_log_unique(const char* msg) CI_HF;
++extern ci_log_fn_t __ci_log_unique_fn CI_HV;
++ci_inline void ci_log_uniquify(void) {
++ if( ci_log_fn != __ci_log_unique ) {
++ __ci_log_unique_fn = ci_log_fn;
++ ci_log_fn = __ci_log_unique;
++ }
++}
++
++extern void ci_log_file(const char* msg) CI_HF;
++extern int ci_log_file_fd CI_HV;
++
++extern void __ci_log_nth(const char* msg) CI_HF;
++extern ci_log_fn_t __ci_log_nth_fn CI_HV;
++extern int ci_log_nth_n CI_HV; /* default 100 */
++ci_inline void ci_log_nth(void) {
++ if( ci_log_fn != __ci_log_nth ) {
++ __ci_log_nth_fn = ci_log_fn;
++ ci_log_fn = __ci_log_nth;
++ }
++}
++
++extern int ci_log_level CI_HV;
++
++extern int ci_log_options CI_HV;
++#define CI_LOG_PID 0x1
++#define CI_LOG_TID 0x2
++#define CI_LOG_TIME 0x4
++#define CI_LOG_DELTA 0x8
++
++/**********************************************************************
++ * Used to define which mode we are in
++ */
++#if (defined(_WIN32) && !defined(__KERNEL__))
++typedef enum {
++ ci_log_md_NULL=0,
++ ci_log_md_ioctl,
++ ci_log_md_stderr,
++ ci_log_md_stdout,
++ ci_log_md_file,
++ ci_log_md_serial,
++ ci_log_md_syslog,
++ ci_log_md_pidfile
++} ci_log_mode_t;
++extern ci_log_mode_t ci_log_mode;
++#endif
++
++/**********************************************************************
++ * Pretty-printing.
++ */
++
++extern char ci_printable_char(char c) CI_HF;
++
++extern void (*ci_hex_dump_formatter)(char* buf, const ci_octet* s,
++ int i, int off, int len) CI_HV;
++extern void ci_hex_dump_format_octets(char*,const ci_octet*,int,int,int) CI_HF;
++extern void ci_hex_dump_format_dwords(char*,const ci_octet*,int,int,int) CI_HF;
++
++extern void ci_hex_dump_row(char* buf, volatile const void* s, int len,
++ ci_ptr_arith_t address) CI_HF;
++ /*!< A row contains up to 16 bytes. Row starts at [address & 15u], so
++ ** therefore [len + (address & 15u)] must be <= 16.
++ */
++
++extern void ci_hex_dump(ci_log_fn_t, volatile const void*,
++ int len, ci_ptr_arith_t address) CI_HF;
++
++extern int ci_hex_dump_to_raw(const char* src_hex, void* buf,
++ unsigned* addr_out_opt, int* skip) CI_HF;
++ /*!< Recovers raw data from a single line of a hex dump. [buf] must be at
++ ** least 16 bytes long. Returns the number of bytes written to [buf] (in
++ ** range 1 -> 16), or -1 if [src_hex] doesn't contain hex data. Does not
++ ** cope with missing bytes at the start of a line.
++ */
++
++extern int ci_format_eth_addr(char* buf, const void* eth_mac_addr,
++ char sep) CI_HF;
++ /*!< This will write 18 characters to <buf> including terminating null.
++ ** Returns number of bytes written excluding null. If [sep] is zero, ':'
++ ** is used.
++ */
++
++extern int ci_parse_eth_addr(void* eth_mac_addr,
++ const char* str, char sep) CI_HF;
++ /*!< If [sep] is zero, absolutely any separator is accepted (even
++ ** inconsistent separators). Returns 0 on success, -1 on error.
++ */
++
++extern int ci_format_ip4_addr(char* buf, unsigned addr_be32) CI_HF;
++ /*!< Formats the IP address (in network endian) in dotted-quad. Returns
++ ** the number of bytes written (up to 15), excluding the null. [buf]
++ ** must be at least 16 bytes long.
++ */
++
++
++/**********************************************************************
++ * Error checking.
++ */
++
++extern void (*ci_fail_stop_fn)(void) CI_HV;
++
++extern void ci_fail_stop(void) CI_HF;
++extern void ci_fail_hang(void) CI_HF;
++extern void ci_fail_bomb(void) CI_HF;
++extern void ci_backtrace(void) CI_HF;
++
++#if defined __linux__ && !defined __KERNEL__
++extern void ci_fail_abort (void) CI_HF;
++#endif
++
++#ifdef __GNUC__
++extern void
++__ci_fail(const char*, ...) CI_PRINTF_LIKE(1,2) CI_HF;
++#else
++# if _PREFAST_
++ extern void _declspec(noreturn) __ci_fail(const char* fmt, ...);
++# else
++ extern void __ci_fail(const char* fmt, ...);
++# endif
++
++#endif
++
++#define ci_warn(x) \
++ do{ ci_log("WARN at %s:%d", __FILE__, __LINE__); }while(0)
++
++#define ci_fail(x) \
++ do{ ci_log("FAIL at %s:%d", __FILE__, __LINE__); __ci_fail x; }while(0)
++
++extern void __ci_sys_fail(const char* fn, int rc,
++ const char* file, int line) CI_HF;
++#define ci_sys_fail(fn, rc) __ci_sys_fail(fn, rc, __FILE__, __LINE__)
++
++/**********************************************************************
++ * Logging to buffer (src/citools/log_buffer.c)
++ */
++
++/*! Divert ci_log() messages to the log buffer
++ * normally they go to the system console */
++extern void ci_log_buffer_till_fail(void) CI_HF;
++
++/*! Dump the contents of the log buffer to the system console */
++extern void ci_log_buffer_dump(void) CI_HF;
++
++
++/**********************************************************************
++ * Some useful pretty-printing.
++ */
++
++#ifdef __linux__
++# define CI_SOCKCALL_FLAGS_FMT "%s%s%s%s%s%s%s%s%s%s%s"
++
++# define CI_SOCKCALL_FLAGS_PRI_ARG(x) \
++ (((x) & MSG_OOB ) ? "OOB " :""), \
++ (((x) & MSG_PEEK ) ? "PEEK " :""), \
++ (((x) & MSG_DONTROUTE ) ? "DONTROUTE " :""), \
++ (((x) & MSG_EOR ) ? "EOR " :""), \
++ (((x) & MSG_CTRUNC ) ? "CTRUNC " :""), \
++ (((x) & MSG_TRUNC ) ? "TRUNC " :""), \
++ (((x) & MSG_WAITALL ) ? "WAITALL " :""), \
++ (((x) & MSG_DONTWAIT ) ? "DONTWAIT " :""), \
++ (((x) & MSG_NOSIGNAL ) ? "NOSIGNAL " :""), \
++ (((x) & MSG_ERRQUEUE ) ? "ERRQUEUE " :""), \
++ (((x) & MSG_CONFIRM ) ? "CONFIRM " :"")
++#endif
++
++#ifdef _WIN32
++# define CI_SOCKCALL_FLAGS_FMT "%s%s%s"
++
++# define CI_SOCKCALL_FLAGS_PRI_ARG(x) \
++ (((x) & MSG_OOB ) ? "OOB " :""), \
++ (((x) & MSG_PEEK ) ? "PEEK " :""), \
++ (((x) & MSG_DONTROUTE ) ? "DONTROUTE " :"")
++#endif
++
++#ifdef __sun__
++# define CI_SOCKCALL_FLAGS_FMT "%s%s%s%s%s%s%s%s%s"
++
++# define CI_SOCKCALL_FLAGS_PRI_ARG(x) \
++ (((x) & MSG_OOB ) ? "OOB " :""), \
++ (((x) & MSG_PEEK ) ? "PEEK " :""), \
++ (((x) & MSG_DONTROUTE ) ? "DONTROUTE " :""), \
++ (((x) & MSG_EOR ) ? "EOR " :""), \
++ (((x) & MSG_CTRUNC ) ? "CTRUNC " :""), \
++ (((x) & MSG_TRUNC ) ? "TRUNC " :""), \
++ (((x) & MSG_WAITALL ) ? "WAITALL " :""), \
++ (((x) & MSG_DONTWAIT ) ? "DONTWAIT " :""), \
++ (((x) & MSG_NOTIFICATION) ? "NOTIFICATION" :"")
++#endif
++
++#endif /* __CI_TOOLS_LOG_H__ */
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/tools/platform/gcc_x86.h b/drivers/xen/sfc_netback/ci/tools/platform/gcc_x86.h
+new file mode 100644
+index 0000000..50a057c
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/platform/gcc_x86.h
+@@ -0,0 +1,361 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_tools_platform */
++
++#ifndef __CI_TOOLS_GCC_X86_H__
++#define __CI_TOOLS_GCC_X86_H__
++
++
++/**********************************************************************
++ * Free-running cycle counters.
++ */
++
++#define CI_HAVE_FRC64
++#define CI_HAVE_FRC32
++
++#define ci_frc32(pval) __asm__ __volatile__("rdtsc" : "=a" (*pval) : : "edx")
++
++#if defined(__x86_64__)
++ci_inline void ci_frc64(ci_uint64* pval) {
++ /* temp fix until we figure how to get this out in one bite */
++ ci_uint64 low, high;
++ __asm__ __volatile__("rdtsc" : "=a" (low) , "=d" (high));
++ *pval = (high << 32) | low;
++}
++
++#else
++#define ci_frc64(pval) __asm__ __volatile__("rdtsc" : "=A" (*pval))
++#endif
++
++#define ci_frc_flush() /* ?? Need a pipeline barrier. */
++
++
++/**********************************************************************
++ * Atomic integer.
++ */
++
++/*
++** int ci_atomic_read(a) { return a->n; }
++** void ci_atomic_set(a, v) { a->n = v; }
++** void ci_atomic_inc(a) { ++a->n; }
++** void ci_atomic_dec(a) { --a->n; }
++** int ci_atomic_inc_and_test(a) { return ++a->n == 0; }
++** int ci_atomic_dec_and_test(a) { return --a->n == 0; }
++** void ci_atomic_and(a, v) { a->n &= v; }
++** void ci_atomic_or(a, v) { a->n |= v; }
++*/
++
++typedef struct { volatile ci_int32 n; } ci_atomic_t;
++
++#define CI_ATOMIC_INITIALISER(i) {(i)}
++
++static inline ci_int32 ci_atomic_read(const ci_atomic_t* a) { return a->n; }
++static inline void ci_atomic_set(ci_atomic_t* a, int v) { a->n = v; ci_wmb(); }
++
++static inline void ci_atomic_inc(ci_atomic_t* a)
++{ __asm__ __volatile__("lock; incl %0" : "+m" (a->n)); }
++
++
++static inline void ci_atomic_dec(ci_atomic_t* a)
++{ __asm__ __volatile__("lock; decl %0" : "+m" (a->n)); }
++
++static inline int ci_atomic_inc_and_test(ci_atomic_t* a) {
++ char r;
++ __asm__ __volatile__("lock; incl %0; sete %1"
++ : "+m" (a->n), "=qm" (r));
++ return r;
++}
++
++static inline int ci_atomic_dec_and_test(ci_atomic_t* a) {
++ char r;
++ __asm__ __volatile__("lock; decl %0; sete %1"
++ : "+m" (a->n), "=qm" (r));
++ return r;
++}
++
++ci_inline int
++ci_atomic_xadd (ci_atomic_t *a, int v) {
++ __asm__ ("lock xadd %0, %1" : "=r" (v), "+m" (a->n) : "0" (v));
++ return v;
++}
++ci_inline int
++ci_atomic_xchg (ci_atomic_t *a, int v) {
++ __asm__ ("lock xchg %0, %1" : "=r" (v), "+m" (a->n) : "0" (v));
++ return v;
++}
++
++ci_inline void ci_atomic32_or(volatile ci_uint32* p, ci_uint32 mask)
++{ __asm__ __volatile__("lock; orl %1, %0" : "+m" (*p) : "ir" (mask)); }
++
++ci_inline void ci_atomic32_and(volatile ci_uint32* p, ci_uint32 mask)
++{ __asm__ __volatile__("lock; andl %1, %0" : "+m" (*p) : "ir" (mask)); }
++
++ci_inline void ci_atomic32_add(volatile ci_uint32* p, ci_uint32 v)
++{ __asm__ __volatile__("lock; addl %1, %0" : "+m" (*p) : "ir" (v)); }
++
++#define ci_atomic_or(a, v) ci_atomic32_or ((ci_uint32*) &(a)->n, (v))
++#define ci_atomic_and(a, v) ci_atomic32_and((ci_uint32*) &(a)->n, (v))
++#define ci_atomic_add(a, v) ci_atomic32_add((ci_uint32*) &(a)->n, (v))
++
++extern int ci_glibc_uses_nptl (void) CI_HF;
++extern int ci_glibc_nptl_broken(void) CI_HF;
++extern int ci_glibc_gs_get_is_multihreaded_offset (void) CI_HF;
++extern int ci_glibc_gs_is_multihreaded_offset CI_HV;
++
++#if !defined(__x86_64__)
++#ifdef __GLIBC__
++/* Returns non-zero if the calling process might be mulithreaded, returns 0 if
++ * it definitely isn't (i.e. if reimplementing this function for other
++ * architectures and platforms, you can safely just return 1).
++ */
++static inline int ci_is_multithreaded (void) {
++
++ while (1) {
++ if (ci_glibc_gs_is_multihreaded_offset >= 0) {
++ /* NPTL keeps a variable that tells us this hanging off gs (i.e. in thread-
++ * local storage); just return this
++ */
++ int r;
++ __asm__ __volatile__ ("movl %%gs:(%1), %0"
++ : "=r" (r)
++ : "r" (ci_glibc_gs_is_multihreaded_offset));
++ return r;
++ }
++
++ if (ci_glibc_gs_is_multihreaded_offset == -2) {
++ /* This means we've already determined that the libc version is NOT good
++ * for our funky "is multithreaded" hack
++ */
++ return 1;
++ }
++
++ /* If we get here, it means this is the first time the function has been
++ * called -- detect the libc version and go around again.
++ */
++ ci_glibc_gs_is_multihreaded_offset = ci_glibc_gs_get_is_multihreaded_offset ();
++
++ /* Go around again. We do the test here rather than at the top so that we go
++ * quicker in the common the case
++ */
++ }
++}
++
++#else /* def __GLIBC__ */
++
++#define ci_is_multithreaded() 1 /* ?? Is the the POSIX way of finding out */
++ /* whether the appication is single */
++ /* threaded? */
++
++#endif /* def __GLIBC__ */
++
++#else /* defined __x86_64__ */
++
++static inline int ci_is_multithreaded (void) {
++ /* Now easy way to tell on x86_64; so assume we're multithreaded */
++ return 1;
++}
++
++#endif /* defined __x86_64__ */
++
++
++/**********************************************************************
++ * Compare and swap.
++ */
++
++#define CI_HAVE_COMPARE_AND_SWAP
++
++ci_inline int ci_cas32_succeed(volatile ci_int32* p, ci_int32 oldval,
++ ci_int32 newval) {
++ char ret;
++ ci_int32 prevval;
++ __asm__ __volatile__("lock; cmpxchgl %3, %1; sete %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++ci_inline int ci_cas32_fail(volatile ci_int32* p, ci_int32 oldval,
++ ci_int32 newval) {
++ char ret;
++ ci_int32 prevval;
++ __asm__ __volatile__("lock; cmpxchgl %3, %1; setne %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++#ifdef __x86_64__
++ci_inline int ci_cas64_succeed(volatile ci_int64* p, ci_int64 oldval,
++ ci_int64 newval) {
++ char ret;
++ ci_int64 prevval;
++ __asm__ __volatile__("lock; cmpxchgq %3, %1; sete %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++ci_inline int ci_cas64_fail(volatile ci_int64* p, ci_int64 oldval,
++ ci_int64 newval) {
++ char ret;
++ ci_int64 prevval;
++ __asm__ __volatile__("lock; cmpxchgq %3, %1; setne %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++#endif
++
++ci_inline int ci_cas32u_succeed(volatile ci_uint32* p, ci_uint32 oldval, ci_uint32 newval) {
++ char ret;
++ ci_uint32 prevval;
++ __asm__ __volatile__("lock; cmpxchgl %3, %1; sete %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++ci_inline int ci_cas32u_fail(volatile ci_uint32* p, ci_uint32 oldval, ci_uint32 newval) {
++ char ret;
++ ci_uint32 prevval;
++ __asm__ __volatile__("lock; cmpxchgl %3, %1; setne %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++ci_inline int ci_cas64u_succeed(volatile ci_uint64* p, ci_uint64 oldval,
++ ci_uint64 newval) {
++ char ret;
++ ci_uint64 prevval;
++ __asm__ __volatile__("lock; cmpxchgq %3, %1; sete %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++ci_inline int ci_cas64u_fail(volatile ci_uint64* p, ci_uint64 oldval,
++ ci_uint64 newval) {
++ char ret;
++ ci_uint64 prevval;
++ __asm__ __volatile__("lock; cmpxchgq %3, %1; setne %0"
++ : "=q"(ret), "+m"(*p), "=a"(prevval)
++ : "r"(newval), "a"(oldval));
++ return ret;
++}
++
++#ifdef __x86_64__
++
++# define ci_cas_uintptr_succeed(p,o,n) \
++ ci_cas64u_succeed((volatile ci_uint64*) (p), (o), (n))
++# define ci_cas_uintptr_fail(p,o,n) \
++ ci_cas64u_fail((volatile ci_uint64*) (p), (o), (n))
++
++#else
++
++# define ci_cas_uintptr_succeed(p,o,n) \
++ ci_cas32u_succeed((volatile ci_uint32*) (p), (o), (n))
++# define ci_cas_uintptr_fail(p,o,n) \
++ ci_cas32u_fail((volatile ci_uint32*) (p), (o), (n))
++
++#endif
++
++
++/**********************************************************************
++ * Atomic bit field.
++ */
++
++typedef ci_uint32 ci_bits;
++#define CI_BITS_N 32u
++
++#define CI_BITS_DECLARE(name, n) \
++ ci_bits name[((n) + CI_BITS_N - 1u) / CI_BITS_N]
++
++ci_inline void ci_bits_clear_all(volatile ci_bits* b, int n_bits)
++{ memset((void*) b, 0, (n_bits+CI_BITS_N-1u) / CI_BITS_N * sizeof(ci_bits)); }
++
++ci_inline void ci_bit_set(volatile ci_bits* b, int i) {
++ __asm__ __volatile__("lock; btsl %1, %0"
++ : "=m" (*b)
++ : "Ir" (i));
++}
++
++ci_inline void ci_bit_clear(volatile ci_bits* b, int i) {
++ __asm__ __volatile__("lock; btrl %1, %0"
++ : "=m" (*b)
++ : "Ir" (i));
++}
++
++ci_inline int ci_bit_test(volatile ci_bits* b, int i) {
++ char rc;
++ __asm__("btl %2, %1; setc %0"
++ : "=r" (rc)
++ : "m" (*b), "Ir" (i));
++ return rc;
++}
++
++ci_inline int ci_bit_test_and_set(volatile ci_bits* b, int i) {
++ char rc;
++ __asm__ __volatile__("lock; btsl %2, %1; setc %0"
++ : "=r" (rc), "+m" (*b)
++ : "Ir" (i));
++ return rc;
++}
++
++ci_inline int ci_bit_test_and_clear(volatile ci_bits* b, int i) {
++ char rc;
++ __asm__ __volatile__("lock; btrl %2, %1; setc %0"
++ : "=r" (rc), "+m" (*b)
++ : "Ir" (i));
++ return rc;
++}
++
++/* These mask ops only work within a single ci_bits word. */
++#define ci_bit_mask_set(b,m) ci_atomic32_or((b), (m))
++#define ci_bit_mask_clear(b,m) ci_atomic32_and((b), ~(m))
++
++
++/**********************************************************************
++ * Misc.
++ */
++
++#if __GNUC__ >= 3
++# define ci_spinloop_pause() __asm__("pause")
++#else
++# define ci_spinloop_pause() __asm__(".byte 0xf3, 0x90")
++#endif
++
++
++#define CI_HAVE_ADDC32
++#define ci_add_carry32(sum, v) __asm__("addl %1, %0 ;" \
++ "adcl $0, %0 ;" \
++ : "=r" (sum) \
++ : "g" ((ci_uint32) v), "0" (sum))
++
++
++#endif /* __CI_TOOLS_GCC_X86_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/tools/platform/linux_kernel.h b/drivers/xen/sfc_netback/ci/tools/platform/linux_kernel.h
+new file mode 100644
+index 0000000..8a87c52
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/platform/linux_kernel.h
+@@ -0,0 +1,362 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++
++/*! \cidoxg_include_ci_tools_platform */
++
++#ifndef __CI_TOOLS_LINUX_KERNEL_H__
++#define __CI_TOOLS_LINUX_KERNEL_H__
++
++/**********************************************************************
++ * Need to know the kernel version.
++ */
++
++#ifndef LINUX_VERSION_CODE
++# include <linux/version.h>
++# ifndef UTS_RELEASE
++ /* 2.6.18 onwards defines UTS_RELEASE in a separate header */
++# include <linux/utsrelease.h>
++# endif
++#endif
++
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,0) || \
++ LINUX_VERSION_CODE >= KERNEL_VERSION(2,7,0)
++# error "Linux 2.6 required"
++#endif
++
++
++#include <linux/slab.h> /* kmalloc / kfree */
++#include <linux/vmalloc.h> /* vmalloc / vfree */
++#include <linux/interrupt.h>/* in_interrupt() */
++#include <linux/in.h>
++#include <linux/in6.h>
++#include <linux/spinlock.h>
++#include <linux/highmem.h>
++#include <linux/smp_lock.h>
++#include <linux/ctype.h>
++#include <linux/uio.h>
++#include <asm/current.h>
++#include <asm/errno.h>
++#include <asm/kmap_types.h>
++#include <asm/semaphore.h>
++
++#include <ci/tools/config.h>
++
++#define ci_in_irq in_irq
++#define ci_in_interrupt in_interrupt
++#define ci_in_atomic in_atomic
++
++
++/**********************************************************************
++ * Misc stuff.
++ */
++
++#ifdef BUG
++# define CI_BOMB BUG
++#endif
++
++ci_inline void* __ci_alloc(size_t n)
++{ return kmalloc(n, (in_interrupt() ? GFP_ATOMIC : GFP_KERNEL)); }
++
++ci_inline void* __ci_atomic_alloc(size_t n)
++{ return kmalloc(n, GFP_ATOMIC ); }
++
++ci_inline void __ci_free(void* p) { return kfree(p); }
++ci_inline void* __ci_vmalloc(size_t n) { return vmalloc(n); }
++ci_inline void __ci_vfree(void* p) { return vfree(p); }
++
++
++#if CI_MEMLEAK_DEBUG_ALLOC_TABLE
++ #define ci_alloc(s) ci_alloc_memleak_debug (s, __FILE__, __LINE__)
++ #define ci_atomic_alloc(s) ci_atomic_alloc_memleak_debug(s, __FILE__, __LINE__)
++ #define ci_free ci_free_memleak_debug
++ #define ci_vmalloc(s) ci_vmalloc_memleak_debug (s, __FILE__,__LINE__)
++ #define ci_vfree ci_vfree_memleak_debug
++ #define ci_alloc_fn ci_alloc_fn_memleak_debug
++ #define ci_vmalloc_fn ci_vmalloc_fn_memleak_debug
++#else /* !CI_MEMLEAK_DEBUG_ALLOC_TABLE */
++ #define ci_alloc_fn __ci_alloc
++ #define ci_vmalloc_fn __ci_vmalloc
++#endif
++
++#ifndef ci_alloc
++ #define ci_atomic_alloc __ci_atomic_alloc
++ #define ci_alloc __ci_alloc
++ #define ci_free __ci_free
++ #define ci_vmalloc __ci_vmalloc
++ #define ci_vmalloc_fn __ci_vmalloc
++ #define ci_vfree __ci_vfree
++#endif
++
++#define ci_sprintf sprintf
++#define ci_vsprintf vsprintf
++#define ci_snprintf snprintf
++#define ci_vsnprintf vsnprintf
++#define ci_sscanf sscanf
++
++
++#define CI_LOG_FN_DEFAULT ci_log_syslog
++
++
++/*--------------------------------------------------------------------
++ *
++ * irqs_disabled - needed for kmap helpers on some kernels
++ *
++ *--------------------------------------------------------------------*/
++#ifdef irqs_disabled
++# define ci_irqs_disabled irqs_disabled
++#else
++# if defined(__i386__) | defined(__x86_64__)
++# define ci_irqs_disabled(x) \
++ ({ \
++ unsigned long flags; \
++ local_save_flags(flags); \
++ !(flags & (1<<9)); \
++ })
++# else
++# error "Need to implement irqs_disabled() for your architecture"
++# endif
++#endif
++
++
++/**********************************************************************
++ * kmap helpers.
++ *
++ * Use ci_k(un)map for code paths which are not in an atomic context.
++ * For atomic code you need to use ci_k(un)map_in_atomic. This will grab
++ * one of the per-CPU kmap slots.
++ *
++ * NB in_interrupt != in_irq. If you don't know the difference then
++ * don't use kmap_in_atomic
++ *
++ * 2.4 allocates kmap slots by function. We are going to re-use the
++ * skb module's slot - we also use the same interlock
++ *
++ * 2.6 allocates kmap slots by type as well as by function. We are
++ * going to use the currently (2.6.10) unsused SOFTIRQ slot
++ *
++ */
++
++ci_inline void* ci_kmap(struct page *page) {
++ CI_DEBUG(if( ci_in_atomic() | ci_in_interrupt() | ci_in_irq() ) BUG());
++ return kmap(page);
++}
++
++ci_inline void ci_kunmap(struct page *page) {
++ kunmap(page);
++}
++
++#define CI_KM_SLOT KM_SOFTIRQ0
++
++
++typedef struct semaphore ci_semaphore_t;
++
++ci_inline void
++ci_sem_init (ci_semaphore_t *sem, int val) {
++ sema_init (sem, val);
++}
++
++ci_inline void
++ci_sem_down (ci_semaphore_t *sem) {
++ down (sem);
++}
++
++ci_inline int
++ci_sem_trydown (ci_semaphore_t *sem) {
++ return down_trylock (sem);
++}
++
++ci_inline void
++ci_sem_up (ci_semaphore_t *sem) {
++ up (sem);
++}
++
++ci_inline int
++ci_sem_get_count(ci_semaphore_t *sem) {
++ return sem->count.counter;
++}
++
++ci_inline void* ci_kmap_in_atomic(struct page *page)
++{
++ CI_DEBUG(if( ci_in_irq() ) BUG());
++
++ /* iSCSI can call without in_interrupt() but with irqs_disabled()
++ and in a context that can't sleep, so we need to check that
++ too */
++ if(ci_in_interrupt() || ci_irqs_disabled())
++ return kmap_atomic(page, CI_KM_SLOT);
++ else
++ return kmap(page);
++}
++
++ci_inline void ci_kunmap_in_atomic(struct page *page, void* kaddr)
++{
++ CI_DEBUG(if( ci_in_irq() ) BUG());
++
++ /* iSCSI can call without in_interrupt() but with irqs_disabled()
++ and in a context that can't sleep, so we need to check that
++ too */
++ if(ci_in_interrupt() || ci_irqs_disabled())
++ kunmap_atomic(kaddr, CI_KM_SLOT);
++ else
++ kunmap(page);
++}
++
++/**********************************************************************
++ * spinlock implementation: used by <ci/tools/spinlock.h>
++ */
++
++#define CI_HAVE_SPINLOCKS
++
++typedef ci_uintptr_t ci_lock_holder_t;
++#define ci_lock_thisthread (ci_lock_holder_t)current
++#define ci_lock_no_holder (ci_lock_holder_t)NULL
++
++typedef spinlock_t ci_lock_i;
++typedef spinlock_t ci_irqlock_i;
++typedef unsigned long ci_irqlock_state_t;
++
++#define IRQLOCK_CYCLES 500000
++
++#define ci_lock_ctor_i(l) spin_lock_init(l)
++#define ci_lock_dtor_i(l) do{}while(0)
++#define ci_lock_lock_i(l) spin_lock(l)
++#define ci_lock_trylock_i(l) spin_trylock(l)
++#define ci_lock_unlock_i(l) spin_unlock(l)
++
++#define ci_irqlock_ctor_i(l) spin_lock_init(l)
++#define ci_irqlock_dtor_i(l) do{}while(0)
++#define ci_irqlock_lock_i(l,s) spin_lock_irqsave(l,*(s))
++#define ci_irqlock_unlock_i(l,s) spin_unlock_irqrestore(l, *(s))
++
++
++/**********************************************************************
++ * register access
++ */
++
++#include <asm/io.h>
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,9)
++typedef volatile void __iomem* ioaddr_t;
++#else
++typedef unsigned long ioaddr_t;
++#endif
++
++
++
++/**********************************************************************
++ * thread implementation -- kernel dependancies probably should be
++ * moved to driver/linux_kernel.h
++ */
++
++#define ci_linux_daemonize(name) daemonize(name)
++
++#include <linux/workqueue.h>
++
++
++typedef struct {
++ void* (*fn)(void* arg);
++ void* arg;
++ const char* name;
++ int thrd_id;
++ struct completion exit_event;
++ struct work_struct keventd_witem;
++} ci_kernel_thread_t;
++
++
++typedef ci_kernel_thread_t* cithread_t;
++
++
++extern int cithread_create(cithread_t* tid, void* (*fn)(void*), void* arg,
++ const char* name);
++extern int cithread_detach(cithread_t kt);
++extern int cithread_join(cithread_t kt);
++
++
++/* Kernel sysctl variables. */
++extern int sysctl_tcp_wmem[3];
++extern int sysctl_tcp_rmem[3];
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
++#define LINUX_HAS_SYSCTL_MEM_MAX
++extern ci_uint32 sysctl_wmem_max;
++extern ci_uint32 sysctl_rmem_max;
++#endif
++
++
++/*--------------------------------------------------------------------
++ *
++ * ci_bigbuf_t: An abstraction of a large buffer. Needed because in the
++ * Linux kernel, large buffers need to be allocated with vmalloc(), whereas
++ * smaller buffers should use kmalloc(). This abstraction chooses the
++ * appropriate mechansim.
++ *
++ *--------------------------------------------------------------------*/
++
++typedef struct {
++ char* p;
++ int is_vmalloc;
++} ci_bigbuf_t;
++
++
++ci_inline int ci_bigbuf_alloc(ci_bigbuf_t* bb, size_t bytes) {
++ if( bytes >= CI_PAGE_SIZE && ! ci_in_atomic() ) {
++ bb->is_vmalloc = 1;
++ if( (bb->p = vmalloc(bytes)) ) return 0;
++ }
++ bb->is_vmalloc = 0;
++ bb->p = kmalloc(bytes, ci_in_interrupt() ? GFP_ATOMIC : GFP_KERNEL);
++ return bb->p ? 0 : -ENOMEM;
++}
++
++ci_inline void ci_bigbuf_free(ci_bigbuf_t* bb) {
++ if( bb->is_vmalloc ) vfree(bb->p);
++ else kfree(bb->p);
++}
++
++ci_inline char* ci_bigbuf_ptr(ci_bigbuf_t* bb)
++{ return bb->p; }
++
++/**********************************************************************
++ * struct iovec abstraction (for Windows port)
++ */
++
++typedef struct iovec ci_iovec;
++
++/* Accessors for buffer/length */
++#define CI_IOVEC_BASE(i) ((i)->iov_base)
++#define CI_IOVEC_LEN(i) ((i)->iov_len)
++
++/**********************************************************************
++ * Signals
++ */
++
++ci_inline void
++ci_send_sig(int signum)
++{
++ send_sig(signum, current, 0);
++}
++
++#endif /* __CI_TOOLS_LINUX_KERNEL_H__ */
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netback/ci/tools/sysdep.h b/drivers/xen/sfc_netback/ci/tools/sysdep.h
+new file mode 100644
+index 0000000..9be16dd
+--- /dev/null
++++ b/drivers/xen/sfc_netback/ci/tools/sysdep.h
+@@ -0,0 +1,132 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*! \cidoxg_include_ci_tools */
++
++#ifndef __CI_TOOLS_SYSDEP_H__
++#define __CI_TOOLS_SYSDEP_H__
++
++/* Make this header self-sufficient */
++#include <ci/compat.h>
++#include <ci/tools/log.h>
++#include <ci/tools/debug.h>
++
++
++/**********************************************************************
++ * Platform dependencies.
++ */
++
++#if defined(__KERNEL__)
++
++# if defined(__linux__)
++# include <ci/tools/platform/linux_kernel.h>
++# elif defined(_WIN32)
++# include <ci/tools/platform/win32_kernel.h>
++# elif defined(__sun__)
++# include <ci/tools/platform/sunos_kernel.h>
++# else
++# error Unknown platform.
++# endif
++
++#elif defined(_WIN32)
++
++# include <ci/tools/platform/win32.h>
++
++#elif defined(__unix__)
++
++# include <ci/tools/platform/unix.h>
++
++#else
++
++# error Unknown platform.
++
++#endif
++
++#if defined(__linux__)
++/*! Linux sendfile() support enable/disable. */
++# define CI_HAVE_SENDFILE /* provide sendfile i/f */
++
++# define CI_HAVE_OS_NOPAGE
++#endif
++
++#if defined(__sun__)
++# define CI_HAVE_SENDFILE /* provide sendfile i/f */
++# define CI_HAVE_SENDFILEV /* provide sendfilev i/f */
++
++# define CI_IOCTL_SENDFILE /* use efrm CI_SENDFILEV ioctl */
++#endif
++
++#if defined(_WIN32)
++typedef ci_uint32 ci_uerr_t; /* range of OS user-mode return codes */
++typedef ci_uint32 ci_kerr_t; /* range of OS kernel-mode return codes */
++#elif defined(__unix__)
++typedef ci_int32 ci_uerr_t; /* range of OS user-mode return codes */
++typedef ci_int32 ci_kerr_t; /* range of OS kernel-mode return codes */
++#endif
++
++
++/**********************************************************************
++ * Compiler and processor dependencies.
++ */
++
++#if defined(__GNUC__)
++
++#if defined(__i386__) || defined(__x86_64__)
++# include <ci/tools/platform/gcc_x86.h>
++#elif defined(__PPC__)
++# include <ci/tools/platform/gcc_ppc.h>
++#elif defined(__ia64__)
++# include <ci/tools/platform/gcc_ia64.h>
++#else
++# error Unknown processor.
++#endif
++
++#elif defined(_MSC_VER)
++
++#if defined(__i386__)
++# include <ci/tools/platform/msvc_x86.h>
++# elif defined(__x86_64__)
++# include <ci/tools/platform/msvc_x86_64.h>
++#else
++# error Unknown processor.
++#endif
++
++#elif defined(__PGI)
++
++# include <ci/tools/platform/pg_x86.h>
++
++#elif defined(__INTEL_COMPILER)
++
++/* Intel compilers v7 claim to be very gcc compatible. */
++# include <ci/tools/platform/gcc_x86.h>
++
++#else
++# error Unknown compiler.
++#endif
++
++
++#endif /* __CI_TOOLS_SYSDEP_H__ */
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netfront/Makefile b/drivers/xen/sfc_netfront/Makefile
+new file mode 100644
+index 0000000..0e4a54b
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/Makefile
+@@ -0,0 +1,11 @@
++EXTRA_CFLAGS += -Idrivers/xen/sfc_netfront -Idrivers/xen/sfc_netutil -Idrivers/xen/netfront
++EXTRA_CFLAGS += -D__ci_driver__
++EXTRA_CFLAGS += -Werror
++
++ifdef GCOV
++EXTRA_CFLAGS += -fprofile-arcs -ftest-coverage -DEFX_GCOV
++endif
++
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_FRONTEND) := sfc_netfront.o
++
++sfc_netfront-objs := accel_msg.o accel_bufs.o accel_netfront.o accel_vi.o accel_xenbus.o accel_tso.o accel_ssr.o accel_debugfs.o falcon_event.o falcon_vi.o pt_tx.o vi_init.o
+diff --git a/drivers/xen/sfc_netfront/accel.h b/drivers/xen/sfc_netfront/accel.h
+new file mode 100644
+index 0000000..abdd837
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel.h
+@@ -0,0 +1,475 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETFRONT_ACCEL_H
++#define NETFRONT_ACCEL_H
++
++#include "accel_msg_iface.h"
++#include "accel_cuckoo_hash.h"
++#include "accel_bufs.h"
++
++#include "etherfabric/ef_vi.h"
++
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++
++#include <linux/kernel.h>
++#include <linux/list.h>
++
++enum netfront_accel_post_status {
++ NETFRONT_ACCEL_STATUS_GOOD,
++ NETFRONT_ACCEL_STATUS_BUSY,
++ NETFRONT_ACCEL_STATUS_CANT
++};
++
++#define NETFRONT_ACCEL_STATS 1
++#if NETFRONT_ACCEL_STATS
++#define NETFRONT_ACCEL_STATS_OP(x) x
++#else
++#define NETFRONT_ACCEL_STATS_OP(x)
++#endif
++
++
++enum netfront_accel_msg_state {
++ NETFRONT_ACCEL_MSG_NONE = 0,
++ NETFRONT_ACCEL_MSG_HELLO = 1,
++ NETFRONT_ACCEL_MSG_HW = 2
++};
++
++
++typedef struct {
++ u32 in_progress;
++ u32 total_len;
++ struct sk_buff *skb;
++} netfront_accel_jumbo_state;
++
++
++struct netfront_accel_ssr_state {
++ /** List of tracked connections. */
++ struct list_head conns;
++
++ /** Free efx_ssr_conn instances. */
++ struct list_head free_conns;
++};
++
++
++struct netfront_accel_netdev_stats {
++ /* Fastpath stats. */
++ u32 fastpath_rx_pkts;
++ u32 fastpath_rx_bytes;
++ u32 fastpath_rx_errors;
++ u32 fastpath_tx_pkts;
++ u32 fastpath_tx_bytes;
++ u32 fastpath_tx_errors;
++};
++
++
++struct netfront_accel_netdev_dbfs {
++ struct dentry *fastpath_rx_pkts;
++ struct dentry *fastpath_rx_bytes;
++ struct dentry *fastpath_rx_errors;
++ struct dentry *fastpath_tx_pkts;
++ struct dentry *fastpath_tx_bytes;
++ struct dentry *fastpath_tx_errors;
++};
++
++
++struct netfront_accel_stats {
++ /** Fast path events */
++ u64 fastpath_tx_busy;
++
++ /** TX DMA queue status */
++ u64 fastpath_tx_completions;
++
++ /** The number of events processed. */
++ u64 event_count;
++
++ /** Number of frame trunc events seen on fastpath */
++ u64 fastpath_frm_trunc;
++
++ /** Number of no rx descriptor trunc events seen on fastpath */
++ u64 rx_no_desc_trunc;
++
++ /** The number of misc bad events (e.g. RX_DISCARD) processed. */
++ u64 bad_event_count;
++
++ /** Number of events dealt with in poll loop */
++ u32 events_per_poll_max;
++ u32 events_per_poll_tx_max;
++ u32 events_per_poll_rx_max;
++
++ /** Largest number of concurrently outstanding tx descriptors */
++ u32 fastpath_tx_pending_max;
++
++ /** The number of events since the last interrupts. */
++ u32 event_count_since_irq;
++
++ /** The max number of events between interrupts. */
++ u32 events_per_irq_max;
++
++ /** The number of interrupts. */
++ u64 irq_count;
++
++ /** The number of useless interrupts. */
++ u64 useless_irq_count;
++
++ /** The number of polls scheduled. */
++ u64 poll_schedule_count;
++
++ /** The number of polls called. */
++ u64 poll_call_count;
++
++ /** The number of rechecks. */
++ u64 poll_reschedule_count;
++
++ /** Number of times we've called netif_stop_queue/netif_wake_queue */
++ u64 queue_stops;
++ u64 queue_wakes;
++
++ /** SSR stats */
++ u64 ssr_bursts;
++ u64 ssr_drop_stream;
++ u64 ssr_misorder;
++ u64 ssr_slow_start;
++ u64 ssr_merges;
++ u64 ssr_too_many;
++ u64 ssr_new_stream;
++};
++
++
++struct netfront_accel_dbfs {
++ struct dentry *fastpath_tx_busy;
++ struct dentry *fastpath_tx_completions;
++ struct dentry *fastpath_tx_pending_max;
++ struct dentry *fastpath_frm_trunc;
++ struct dentry *rx_no_desc_trunc;
++ struct dentry *event_count;
++ struct dentry *bad_event_count;
++ struct dentry *events_per_poll_max;
++ struct dentry *events_per_poll_rx_max;
++ struct dentry *events_per_poll_tx_max;
++ struct dentry *event_count_since_irq;
++ struct dentry *events_per_irq_max;
++ struct dentry *irq_count;
++ struct dentry *useless_irq_count;
++ struct dentry *poll_schedule_count;
++ struct dentry *poll_call_count;
++ struct dentry *poll_reschedule_count;
++ struct dentry *queue_stops;
++ struct dentry *queue_wakes;
++ struct dentry *ssr_bursts;
++ struct dentry *ssr_drop_stream;
++ struct dentry *ssr_misorder;
++ struct dentry *ssr_slow_start;
++ struct dentry *ssr_merges;
++ struct dentry *ssr_too_many;
++ struct dentry *ssr_new_stream;
++};
++
++
++typedef struct netfront_accel_vnic {
++ struct netfront_accel_vnic *next;
++
++ struct mutex vnic_mutex;
++
++ spinlock_t tx_lock;
++
++ struct netfront_accel_bufpages bufpages;
++ struct netfront_accel_bufinfo *rx_bufs;
++ struct netfront_accel_bufinfo *tx_bufs;
++
++ /** Hardware & VI state */
++ ef_vi vi;
++
++ ef_vi_state *vi_state;
++
++ ef_eventq_state evq_state;
++
++ void *evq_mapping;
++
++ /** Hardware dependant state */
++ union {
++ struct {
++ /** Falcon A or B */
++ enum net_accel_hw_type type;
++ u32 *evq_rptr;
++ u32 *doorbell;
++ void *evq_rptr_mapping;
++ void *doorbell_mapping;
++ void *txdmaq_mapping;
++ void *rxdmaq_mapping;
++ } falcon;
++ } hw;
++
++ /** RX DMA queue status */
++ u32 rx_dma_level;
++
++ /** Number of RX descriptors waiting to be pushed to the card. */
++ u32 rx_dma_batched;
++#define NETFRONT_ACCEL_RX_DESC_BATCH 16
++
++ /**
++ * Hash table of remote mac addresses to decide whether to try
++ * fast path
++ */
++ cuckoo_hash_table fastpath_table;
++ spinlock_t table_lock;
++
++ /** the local mac address of virtual interface we're accelerating */
++ u8 mac[ETH_ALEN];
++
++ int rx_pkt_stride;
++ int rx_skb_stride;
++
++ /**
++ * Keep track of fragments of jumbo packets as events are
++ * delivered by NIC
++ */
++ netfront_accel_jumbo_state jumbo_state;
++
++ struct net_device *net_dev;
++
++ /** These two gate the enabling of fast path operations */
++ int frontend_ready;
++ int backend_netdev_up;
++
++ int irq_enabled;
++ spinlock_t irq_enabled_lock;
++
++ int tx_enabled;
++
++ int poll_enabled;
++
++ /** A spare slot for a TX packet. This is treated as an extension
++ * of the DMA queue. */
++ struct sk_buff *tx_skb;
++
++ /** Keep track of fragments of SSR packets */
++ struct netfront_accel_ssr_state ssr_state;
++
++ struct xenbus_device *dev;
++
++ /** Event channel for messages */
++ int msg_channel;
++ int msg_channel_irq;
++
++ /** Event channel for network interrupts. */
++ int net_channel;
++ int net_channel_irq;
++
++ struct net_accel_shared_page *shared_page;
++
++ grant_ref_t ctrl_page_gnt;
++ grant_ref_t msg_page_gnt;
++
++ /** Message Qs, 1 each way. */
++ sh_msg_fifo2 to_dom0;
++ sh_msg_fifo2 from_dom0;
++
++ enum netfront_accel_msg_state msg_state;
++
++ /** Watch on accelstate */
++ struct xenbus_watch backend_accel_watch;
++ /** Watch on frontend's MAC address */
++ struct xenbus_watch mac_address_watch;
++
++ /** Work to process received irq/msg */
++ struct work_struct msg_from_bend;
++
++ /** Wait queue for changes in accelstate. */
++ wait_queue_head_t state_wait_queue;
++
++ /** The current accelstate of this driver. */
++ XenbusState frontend_state;
++
++ /** The most recent accelstate seen by the xenbus watch. */
++ XenbusState backend_state;
++
++ /** Non-zero if we should reject requests to connect. */
++ int removing;
++
++ /** Non-zero if the domU shared state has been initialised. */
++ int domU_state_is_setup;
++
++ /** Non-zero if the dom0 shared state has been initialised. */
++ int dom0_state_is_setup;
++
++ /* Those statistics that are added to the netdev stats */
++ struct netfront_accel_netdev_stats netdev_stats;
++ struct netfront_accel_netdev_stats stats_last_read;
++#ifdef CONFIG_DEBUG_FS
++ struct netfront_accel_netdev_dbfs netdev_dbfs;
++#endif
++
++ /* These statistics are internal and optional */
++#if NETFRONT_ACCEL_STATS
++ struct netfront_accel_stats stats;
++#ifdef CONFIG_DEBUG_FS
++ struct netfront_accel_dbfs dbfs;
++#endif
++#endif
++
++ /** Debufs fs dir for this interface */
++ struct dentry *dbfs_dir;
++} netfront_accel_vnic;
++
++
++/* Module parameters */
++extern unsigned sfc_netfront_max_pages;
++extern unsigned sfc_netfront_buffer_split;
++
++extern const char *frontend_name;
++extern struct netfront_accel_hooks accel_hooks;
++extern struct workqueue_struct *netfront_accel_workqueue;
++
++
++extern
++void netfront_accel_vi_ctor(netfront_accel_vnic *vnic);
++
++extern
++int netfront_accel_vi_init(netfront_accel_vnic *vnic,
++ struct net_accel_msg_hw *hw_msg);
++
++extern
++void netfront_accel_vi_dtor(netfront_accel_vnic *vnic);
++
++
++/**
++ * Add new buffers which have been registered with the NIC.
++ *
++ * @v vnic The vnic instance to process the response.
++ *
++ * The buffers contained in the message are added to the buffer pool.
++ */
++extern
++void netfront_accel_vi_add_bufs(netfront_accel_vnic *vnic, int is_rx);
++
++/**
++ * Put a packet on the tx DMA queue.
++ *
++ * @v vnic The vnic instance to accept the packet.
++ * @v skb A sk_buff to send.
++ *
++ * Attempt to send a packet. On success, the skb is owned by the DMA
++ * queue and will be released when the completion event arrives.
++ */
++extern enum netfront_accel_post_status
++netfront_accel_vi_tx_post(netfront_accel_vnic *vnic,
++ struct sk_buff *skb);
++
++
++/**
++ * Process events in response to an interrupt.
++ *
++ * @v vnic The vnic instance to poll.
++ * @v rx_packets The maximum number of rx packets to process.
++ * @ret rx_done The number of rx packets processed.
++ *
++ * The vnic will process events until there are no more events
++ * remaining or the specified number of rx packets has been processed.
++ * The split from the interrupt call is to allow Linux NAPI
++ * polling.
++ */
++extern
++int netfront_accel_vi_poll(netfront_accel_vnic *vnic, int rx_packets);
++
++
++/**
++ * Iterate over the fragments of a packet buffer.
++ *
++ * @v skb The packet buffer to examine.
++ * @v idx A variable name for the fragment index.
++ * @v data A variable name for the address of the fragment data.
++ * @v length A variable name for the fragment length.
++ * @v code A section of code to execute for each fragment.
++ *
++ * This macro iterates over the fragments in a packet buffer and
++ * executes the code for each of them.
++ */
++#define NETFRONT_ACCEL_PKTBUFF_FOR_EACH_FRAGMENT(skb, frag_idx, \
++ frag_data, frag_len, \
++ code) \
++ do { \
++ int frag_idx; \
++ void *frag_data; \
++ unsigned int frag_len; \
++ \
++ frag_data = skb->data; \
++ frag_len = skb_headlen(skb); \
++ frag_idx = 0; \
++ while (1) { /* For each fragment */ \
++ code; \
++ if (frag_idx >= skb_shinfo(skb)->nr_frags) { \
++ break; \
++ } else { \
++ skb_frag_t *fragment; \
++ fragment = &skb_shinfo(skb)->frags[frag_idx]; \
++ frag_len = fragment->size; \
++ frag_data = ((void*)page_address(fragment->page) \
++ + fragment->page_offset); \
++ }; \
++ frag_idx++; \
++ } \
++ } while(0)
++
++static inline
++void netfront_accel_disable_net_interrupts(netfront_accel_vnic *vnic)
++{
++ mask_evtchn(vnic->net_channel);
++}
++
++static inline
++void netfront_accel_enable_net_interrupts(netfront_accel_vnic *vnic)
++{
++ unmask_evtchn(vnic->net_channel);
++}
++
++void netfront_accel_msg_tx_fastpath(netfront_accel_vnic *vnic, const void *mac,
++ u32 ip, u16 port, u8 protocol);
++
++/* Process an IRQ received from back end driver */
++irqreturn_t netfront_accel_msg_channel_irq_from_bend(int irq, void *context);
++irqreturn_t netfront_accel_net_channel_irq_from_bend(int irq, void *context);
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++extern void netfront_accel_msg_from_bend(struct work_struct *context);
++#else
++extern void netfront_accel_msg_from_bend(void *context);
++#endif
++
++extern void vnic_stop_fastpath(netfront_accel_vnic *vnic);
++
++extern int netfront_accel_probe(struct net_device *net_dev,
++ struct xenbus_device *dev);
++extern int netfront_accel_remove(struct xenbus_device *dev);
++extern void netfront_accel_set_closing(netfront_accel_vnic *vnic);
++
++extern int netfront_accel_vi_enable_interrupts(netfront_accel_vnic *vnic);
++
++extern void netfront_accel_debugfs_init(void);
++extern void netfront_accel_debugfs_fini(void);
++extern int netfront_accel_debugfs_create(netfront_accel_vnic *vnic);
++extern int netfront_accel_debugfs_remove(netfront_accel_vnic *vnic);
++
++#endif /* NETFRONT_ACCEL_H */
+diff --git a/drivers/xen/sfc_netfront/accel_bufs.c b/drivers/xen/sfc_netfront/accel_bufs.c
+new file mode 100644
+index 0000000..f96f73c
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_bufs.c
+@@ -0,0 +1,393 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <xen/gnttab.h>
++
++#include "accel_bufs.h"
++#include "accel_util.h"
++
++#include "accel.h"
++
++
++static int
++netfront_accel_alloc_buf_desc_blocks(struct netfront_accel_bufinfo *manager,
++ int pages)
++{
++ manager->desc_blocks =
++ kzalloc(sizeof(struct netfront_accel_pkt_desc *) *
++ NETFRONT_ACCEL_BUF_NUM_BLOCKS(pages), GFP_KERNEL);
++ if (manager->desc_blocks == NULL) {
++ return -ENOMEM;
++ }
++
++ return 0;
++}
++
++static int
++netfront_accel_alloc_buf_lists(struct netfront_accel_bufpages *bufpages,
++ int pages)
++{
++ bufpages->page_list = kmalloc(pages * sizeof(void *), GFP_KERNEL);
++ if (bufpages->page_list == NULL) {
++ return -ENOMEM;
++ }
++
++ bufpages->grant_list = kzalloc(pages * sizeof(grant_ref_t), GFP_KERNEL);
++ if (bufpages->grant_list == NULL) {
++ kfree(bufpages->page_list);
++ bufpages->page_list = NULL;
++ return -ENOMEM;
++ }
++
++ return 0;
++}
++
++
++int netfront_accel_alloc_buffer_mem(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *rx_manager,
++ struct netfront_accel_bufinfo *tx_manager,
++ int pages)
++{
++ int n, rc;
++
++ if ((rc = netfront_accel_alloc_buf_desc_blocks
++ (rx_manager, pages - (pages / sfc_netfront_buffer_split))) < 0) {
++ goto rx_fail;
++ }
++
++ if ((rc = netfront_accel_alloc_buf_desc_blocks
++ (tx_manager, pages / sfc_netfront_buffer_split)) < 0) {
++ goto tx_fail;
++ }
++
++ if ((rc = netfront_accel_alloc_buf_lists(bufpages, pages)) < 0) {
++ goto lists_fail;
++ }
++
++ for (n = 0; n < pages; n++) {
++ void *tmp = (void*)__get_free_page(GFP_KERNEL);
++ if (tmp == NULL)
++ break;
++
++ bufpages->page_list[n] = tmp;
++ }
++
++ if (n != pages) {
++ EPRINTK("%s: not enough pages: %d != %d\n", __FUNCTION__, n,
++ pages);
++ for (; n >= 0; n--)
++ free_page((unsigned long)(bufpages->page_list[n]));
++ rc = -ENOMEM;
++ goto pages_fail;
++ }
++
++ bufpages->max_pages = pages;
++ bufpages->page_reqs = 0;
++
++ return 0;
++
++ pages_fail:
++ kfree(bufpages->page_list);
++ kfree(bufpages->grant_list);
++
++ bufpages->page_list = NULL;
++ bufpages->grant_list = NULL;
++ lists_fail:
++ kfree(tx_manager->desc_blocks);
++ tx_manager->desc_blocks = NULL;
++
++ tx_fail:
++ kfree(rx_manager->desc_blocks);
++ rx_manager->desc_blocks = NULL;
++ rx_fail:
++ return rc;
++}
++
++
++void netfront_accel_free_buffer_mem(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *rx_manager,
++ struct netfront_accel_bufinfo *tx_manager)
++{
++ int i;
++
++ for (i = 0; i < bufpages->max_pages; i++) {
++ if (bufpages->grant_list[i] != 0)
++ net_accel_ungrant_page(bufpages->grant_list[i]);
++ free_page((unsigned long)(bufpages->page_list[i]));
++ }
++
++ if (bufpages->max_pages) {
++ kfree(bufpages->page_list);
++ kfree(bufpages->grant_list);
++ kfree(rx_manager->desc_blocks);
++ kfree(tx_manager->desc_blocks);
++ }
++}
++
++
++/*
++ * Allocate memory for the buffer manager and create a lock. If no
++ * lock is supplied its own is allocated.
++ */
++struct netfront_accel_bufinfo *netfront_accel_init_bufs(spinlock_t *lock)
++{
++ struct netfront_accel_bufinfo *res = kmalloc(sizeof(*res), GFP_KERNEL);
++ if (res != NULL) {
++ res->npages = res->nused = 0;
++ res->first_free = -1;
++
++ if (lock == NULL) {
++ res->lock = kmalloc(sizeof(*res->lock), GFP_KERNEL);
++ if (res->lock == NULL) {
++ kfree(res);
++ return NULL;
++ }
++ spin_lock_init(res->lock);
++ res->internally_locked = 1;
++ } else {
++ res->lock = lock;
++ res->internally_locked = 0;
++ }
++
++ res->desc_blocks = NULL;
++ }
++
++ return res;
++}
++
++
++void netfront_accel_fini_bufs(struct netfront_accel_bufinfo *bufs)
++{
++ if (bufs->internally_locked)
++ kfree(bufs->lock);
++ kfree(bufs);
++}
++
++
++int netfront_accel_buf_map_request(struct xenbus_device *dev,
++ struct netfront_accel_bufpages *bufpages,
++ struct net_accel_msg *msg,
++ int pages, int offset)
++{
++ int i, mfn;
++ int err;
++
++ net_accel_msg_init(msg, NET_ACCEL_MSG_MAPBUF);
++
++ BUG_ON(pages > NET_ACCEL_MSG_MAX_PAGE_REQ);
++
++ msg->u.mapbufs.pages = pages;
++
++ for (i = 0; i < msg->u.mapbufs.pages; i++) {
++ /*
++ * This can happen if we tried to send this message
++ * earlier but the queue was full.
++ */
++ if (bufpages->grant_list[offset+i] != 0) {
++ msg->u.mapbufs.grants[i] =
++ bufpages->grant_list[offset+i];
++ continue;
++ }
++
++ mfn = virt_to_mfn(bufpages->page_list[offset+i]);
++ VPRINTK("%s: Granting page %d, mfn %08x\n",
++ __FUNCTION__, i, mfn);
++
++ bufpages->grant_list[offset+i] =
++ net_accel_grant_page(dev, mfn, 0);
++ msg->u.mapbufs.grants[i] = bufpages->grant_list[offset+i];
++
++ if (msg->u.mapbufs.grants[i] < 0) {
++ EPRINTK("%s: Failed to grant buffer: %d\n",
++ __FUNCTION__, msg->u.mapbufs.grants[i]);
++ err = -EIO;
++ goto error;
++ }
++ }
++
++ /* This is interpreted on return as the offset in the the page_list */
++ msg->u.mapbufs.reqid = offset;
++
++ return 0;
++
++error:
++ /* Ungrant all the pages we've successfully granted. */
++ for (i--; i >= 0; i--) {
++ net_accel_ungrant_page(bufpages->grant_list[offset+i]);
++ bufpages->grant_list[offset+i] = 0;
++ }
++ return err;
++}
++
++
++/* Process a response to a buffer request. */
++int netfront_accel_add_bufs(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *manager,
++ struct net_accel_msg *msg)
++{
++ int msg_pages, page_offset, i, newtot;
++ int old_block_count, new_block_count;
++ u32 msg_buf;
++ unsigned long flags;
++
++ VPRINTK("%s: manager %p msg %p\n", __FUNCTION__, manager, msg);
++
++ BUG_ON(msg->id != (NET_ACCEL_MSG_MAPBUF | NET_ACCEL_MSG_REPLY));
++
++ msg_pages = msg->u.mapbufs.pages;
++ msg_buf = msg->u.mapbufs.buf;
++ page_offset = msg->u.mapbufs.reqid;
++
++ spin_lock_irqsave(manager->lock, flags);
++ newtot = manager->npages + msg_pages;
++ old_block_count =
++ (manager->npages + NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK - 1) >>
++ NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT;
++ new_block_count =
++ (newtot + NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK - 1) >>
++ NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT;
++
++ for (i = old_block_count; i < new_block_count; i++) {
++ struct netfront_accel_pkt_desc *block;
++ if (manager->desc_blocks[i] != NULL) {
++ VPRINTK("Not needed\n");
++ continue;
++ }
++ block = kzalloc(NETFRONT_ACCEL_BUFS_PER_BLOCK *
++ sizeof(netfront_accel_pkt_desc), GFP_ATOMIC);
++ if (block == NULL) {
++ spin_unlock_irqrestore(manager->lock, flags);
++ return -ENOMEM;
++ }
++ manager->desc_blocks[i] = block;
++ }
++ for (i = manager->npages; i < newtot; i++) {
++ int k, j = i - manager->npages;
++ int block_num;
++ int block_idx;
++ struct netfront_accel_pkt_desc *pkt;
++
++ block_num = i >> NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT;
++ block_idx = (NETFRONT_ACCEL_BUFS_PER_PAGE*i)
++ & (NETFRONT_ACCEL_BUFS_PER_BLOCK-1);
++
++ pkt = manager->desc_blocks[block_num] + block_idx;
++
++ for (k = 0; k < NETFRONT_ACCEL_BUFS_PER_PAGE; k++) {
++ BUG_ON(page_offset + j >= bufpages->max_pages);
++
++ pkt[k].buf_id = NETFRONT_ACCEL_BUFS_PER_PAGE * i + k;
++ pkt[k].pkt_kva = bufpages->page_list[page_offset + j] +
++ (PAGE_SIZE/NETFRONT_ACCEL_BUFS_PER_PAGE) * k;
++ pkt[k].pkt_buff_addr = msg_buf +
++ (PAGE_SIZE/NETFRONT_ACCEL_BUFS_PER_PAGE) *
++ (NETFRONT_ACCEL_BUFS_PER_PAGE * j + k);
++ pkt[k].next_free = manager->first_free;
++ manager->first_free = pkt[k].buf_id;
++ *(int*)(pkt[k].pkt_kva) = pkt[k].buf_id;
++
++ VPRINTK("buf %d desc %p kva %p buffaddr %x\n",
++ pkt[k].buf_id, &(pkt[k]), pkt[k].pkt_kva,
++ pkt[k].pkt_buff_addr);
++ }
++ }
++ manager->npages = newtot;
++ spin_unlock_irqrestore(manager->lock, flags);
++ VPRINTK("Added %d pages. Total is now %d\n", msg_pages,
++ manager->npages);
++ return 0;
++}
++
++
++netfront_accel_pkt_desc *
++netfront_accel_buf_find(struct netfront_accel_bufinfo *manager, u16 id)
++{
++ netfront_accel_pkt_desc *pkt;
++ int block_num = id >> NETFRONT_ACCEL_BUFS_PER_BLOCK_SHIFT;
++ int block_idx = id & (NETFRONT_ACCEL_BUFS_PER_BLOCK - 1);
++ BUG_ON(id >= manager->npages * NETFRONT_ACCEL_BUFS_PER_PAGE);
++ BUG_ON(block_idx >= NETFRONT_ACCEL_BUFS_PER_BLOCK);
++ pkt = manager->desc_blocks[block_num] + block_idx;
++ return pkt;
++}
++
++
++/* Allocate a buffer from the buffer manager */
++netfront_accel_pkt_desc *
++netfront_accel_buf_get(struct netfront_accel_bufinfo *manager)
++{
++ int bufno = -1;
++ netfront_accel_pkt_desc *buf = NULL;
++ unsigned long flags = 0;
++
++ /* Any spare? */
++ if (manager->first_free == -1)
++ return NULL;
++ /* Take lock */
++ if (manager->internally_locked)
++ spin_lock_irqsave(manager->lock, flags);
++ bufno = manager->first_free;
++ if (bufno != -1) {
++ buf = netfront_accel_buf_find(manager, bufno);
++ manager->first_free = buf->next_free;
++ manager->nused++;
++ }
++ /* Release lock */
++ if (manager->internally_locked)
++ spin_unlock_irqrestore(manager->lock, flags);
++
++ /* Tell the world */
++ VPRINTK("Allocated buffer %i, buffaddr %x\n", bufno,
++ buf->pkt_buff_addr);
++
++ return buf;
++}
++
++
++/* Release a buffer back to the buffer manager pool */
++int netfront_accel_buf_put(struct netfront_accel_bufinfo *manager, u16 id)
++{
++ netfront_accel_pkt_desc *buf = netfront_accel_buf_find(manager, id);
++ unsigned long flags = 0;
++ unsigned was_empty = 0;
++ int bufno = id;
++
++ VPRINTK("Freeing buffer %i\n", id);
++ BUG_ON(id == (u16)-1);
++
++ if (manager->internally_locked)
++ spin_lock_irqsave(manager->lock, flags);
++
++ if (manager->first_free == -1)
++ was_empty = 1;
++
++ buf->next_free = manager->first_free;
++ manager->first_free = bufno;
++ manager->nused--;
++
++ if (manager->internally_locked)
++ spin_unlock_irqrestore(manager->lock, flags);
++
++ return was_empty;
++}
+diff --git a/drivers/xen/sfc_netfront/accel_bufs.h b/drivers/xen/sfc_netfront/accel_bufs.h
+new file mode 100644
+index 0000000..4ff3eaa
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_bufs.h
+@@ -0,0 +1,181 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETFRONT_ACCEL_BUFS_H
++#define NETFRONT_ACCEL_BUFS_H
++
++#include <linux/skbuff.h>
++#include <linux/spinlock.h>
++#include <xen/xenbus.h>
++
++#include "accel_msg_iface.h"
++
++
++/*! Buffer descriptor structure */
++typedef struct netfront_accel_pkt_desc {
++ int buf_id;
++ u32 pkt_buff_addr;
++ void *pkt_kva;
++ /* This is the socket buffer currently married to this buffer */
++ struct sk_buff *skb;
++ int next_free;
++} netfront_accel_pkt_desc;
++
++
++#define NETFRONT_ACCEL_DEFAULT_BUF_PAGES (384)
++#define NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT (4)
++#define NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK \
++ (1 << (NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT))
++#define NETFRONT_ACCEL_BUFS_PER_PAGE_SHIFT (1)
++#define NETFRONT_ACCEL_BUFS_PER_PAGE \
++ (1 << (NETFRONT_ACCEL_BUFS_PER_PAGE_SHIFT))
++#define NETFRONT_ACCEL_BUFS_PER_BLOCK_SHIFT \
++ (NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK_SHIFT + \
++ NETFRONT_ACCEL_BUFS_PER_PAGE_SHIFT)
++#define NETFRONT_ACCEL_BUFS_PER_BLOCK \
++ (1 << NETFRONT_ACCEL_BUFS_PER_BLOCK_SHIFT)
++#define NETFRONT_ACCEL_BUF_NUM_BLOCKS(max_pages) \
++ (((max_pages)+NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK-1) / \
++ NETFRONT_ACCEL_BUF_PAGES_PER_BLOCK)
++
++/*! Buffer management structure. */
++struct netfront_accel_bufinfo {
++ /* number added to this manager */
++ unsigned npages;
++ /* number currently used from this manager */
++ unsigned nused;
++
++ int first_free;
++
++ int internally_locked;
++ spinlock_t *lock;
++
++ /*
++ * array of pointers (length NETFRONT_ACCEL_BUF_NUM_BLOCKS) to
++ * pkt descs
++ */
++ struct netfront_accel_pkt_desc **desc_blocks;
++};
++
++
++struct netfront_accel_bufpages {
++ /* length of lists of pages/grants */
++ int max_pages;
++ /* list of pages allocated for network buffers */
++ void **page_list;
++ /* list of grants for the above pages */
++ grant_ref_t *grant_list;
++
++ /* number of page requests that have been made */
++ unsigned page_reqs;
++};
++
++
++/*! Allocate memory for the buffer manager, set up locks etc.
++ * Optionally takes a lock to use, if not supplied it makes its own.
++ *
++ * \return pointer to netfront_accel_bufinfo structure that represents the
++ * buffer manager
++ */
++extern struct netfront_accel_bufinfo *
++netfront_accel_init_bufs(spinlock_t *lock);
++
++/*! Allocate memory for the buffers
++ */
++extern int
++netfront_accel_alloc_buffer_mem(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *rx_res,
++ struct netfront_accel_bufinfo *tx_res,
++ int pages);
++extern void
++netfront_accel_free_buffer_mem(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *rx_res,
++ struct netfront_accel_bufinfo *tx_res);
++
++/*! Release memory for the buffer manager, buffers, etc.
++ *
++ * \param manager pointer to netfront_accel_bufinfo structure that
++ * represents the buffer manager
++ */
++extern void netfront_accel_fini_bufs(struct netfront_accel_bufinfo *manager);
++
++/*! Release a buffer.
++ *
++ * \param manager The buffer manager which owns the buffer.
++ * \param id The buffer identifier.
++ */
++extern int netfront_accel_buf_put(struct netfront_accel_bufinfo *manager,
++ u16 id);
++
++/*! Get the packet descriptor associated with a buffer id.
++ *
++ * \param manager The buffer manager which owns the buffer.
++ * \param id The buffer identifier.
++ *
++ * The returned value is the packet descriptor for this buffer.
++ */
++extern netfront_accel_pkt_desc *
++netfront_accel_buf_find(struct netfront_accel_bufinfo *manager, u16 id);
++
++
++/*! Fill out a message request for some buffers to be mapped by the
++ * back end driver
++ *
++ * \param manager The buffer manager
++ * \param msg Pointer to an ef_msg to complete.
++ * \return 0 on success
++ */
++extern int
++netfront_accel_buf_map_request(struct xenbus_device *dev,
++ struct netfront_accel_bufpages *bufpages,
++ struct net_accel_msg *msg,
++ int pages, int offset);
++
++/*! Process a response to a buffer request.
++ *
++ * Deal with a received message from the back end in response to our
++ * request for buffers
++ *
++ * \param manager The buffer manager
++ * \param msg The received message from the back end describing new
++ * buffers
++ * \return 0 on success
++ */
++extern int
++netfront_accel_add_bufs(struct netfront_accel_bufpages *bufpages,
++ struct netfront_accel_bufinfo *manager,
++ struct net_accel_msg *msg);
++
++
++/*! Allocate a buffer from the buffer manager
++ *
++ * \param manager The buffer manager data structure
++ * \param id On exit, the id of the buffer allocated
++ * \return Pointer to buffer descriptor.
++ */
++struct netfront_accel_pkt_desc *
++netfront_accel_buf_get(struct netfront_accel_bufinfo *manager);
++
++#endif /* NETFRONT_ACCEL_BUFS_H */
++
+diff --git a/drivers/xen/sfc_netfront/accel_debugfs.c b/drivers/xen/sfc_netfront/accel_debugfs.c
+new file mode 100644
+index 0000000..d5ef238
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_debugfs.c
+@@ -0,0 +1,211 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/fs.h>
++#include <linux/debugfs.h>
++
++#include "accel.h"
++
++#if defined(CONFIG_DEBUG_FS)
++static struct dentry *sfc_debugfs_root = NULL;
++#endif
++
++void netfront_accel_debugfs_init(void)
++{
++#if defined(CONFIG_DEBUG_FS)
++ sfc_debugfs_root = debugfs_create_dir(frontend_name, NULL);
++#endif
++}
++
++
++void netfront_accel_debugfs_fini(void)
++{
++#if defined(CONFIG_DEBUG_FS)
++ if (sfc_debugfs_root)
++ debugfs_remove(sfc_debugfs_root);
++#endif
++}
++
++
++int netfront_accel_debugfs_create(netfront_accel_vnic *vnic)
++{
++#if defined(CONFIG_DEBUG_FS)
++ if (sfc_debugfs_root == NULL)
++ return -ENOENT;
++
++ vnic->dbfs_dir = debugfs_create_dir(vnic->net_dev->name,
++ sfc_debugfs_root);
++ if (vnic->dbfs_dir == NULL)
++ return -ENOMEM;
++
++ vnic->netdev_dbfs.fastpath_rx_pkts = debugfs_create_u32
++ ("fastpath_rx_pkts", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_rx_pkts);
++ vnic->netdev_dbfs.fastpath_rx_bytes = debugfs_create_u32
++ ("fastpath_rx_bytes", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_rx_bytes);
++ vnic->netdev_dbfs.fastpath_rx_errors = debugfs_create_u32
++ ("fastpath_rx_errors", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_rx_errors);
++ vnic->netdev_dbfs.fastpath_tx_pkts = debugfs_create_u32
++ ("fastpath_tx_pkts", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_tx_pkts);
++ vnic->netdev_dbfs.fastpath_tx_bytes = debugfs_create_u32
++ ("fastpath_tx_bytes", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_tx_bytes);
++ vnic->netdev_dbfs.fastpath_tx_errors = debugfs_create_u32
++ ("fastpath_tx_errors", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->netdev_stats.fastpath_tx_errors);
++
++#if NETFRONT_ACCEL_STATS
++ vnic->dbfs.irq_count = debugfs_create_u64
++ ("irq_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.irq_count);
++ vnic->dbfs.useless_irq_count = debugfs_create_u64
++ ("useless_irq_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.useless_irq_count);
++ vnic->dbfs.poll_schedule_count = debugfs_create_u64
++ ("poll_schedule_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.poll_schedule_count);
++ vnic->dbfs.poll_call_count = debugfs_create_u64
++ ("poll_call_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.poll_call_count);
++ vnic->dbfs.poll_reschedule_count = debugfs_create_u64
++ ("poll_reschedule_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.poll_reschedule_count);
++ vnic->dbfs.queue_stops = debugfs_create_u64
++ ("queue_stops", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.queue_stops);
++ vnic->dbfs.queue_wakes = debugfs_create_u64
++ ("queue_wakes", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.queue_wakes);
++ vnic->dbfs.ssr_bursts = debugfs_create_u64
++ ("ssr_bursts", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_bursts);
++ vnic->dbfs.ssr_drop_stream = debugfs_create_u64
++ ("ssr_drop_stream", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_drop_stream);
++ vnic->dbfs.ssr_misorder = debugfs_create_u64
++ ("ssr_misorder", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_misorder);
++ vnic->dbfs.ssr_slow_start = debugfs_create_u64
++ ("ssr_slow_start", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_slow_start);
++ vnic->dbfs.ssr_merges = debugfs_create_u64
++ ("ssr_merges", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_merges);
++ vnic->dbfs.ssr_too_many = debugfs_create_u64
++ ("ssr_too_many", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_too_many);
++ vnic->dbfs.ssr_new_stream = debugfs_create_u64
++ ("ssr_new_stream", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.ssr_new_stream);
++
++ vnic->dbfs.fastpath_tx_busy = debugfs_create_u64
++ ("fastpath_tx_busy", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.fastpath_tx_busy);
++ vnic->dbfs.fastpath_tx_completions = debugfs_create_u64
++ ("fastpath_tx_completions", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.fastpath_tx_completions);
++ vnic->dbfs.fastpath_tx_pending_max = debugfs_create_u32
++ ("fastpath_tx_pending_max", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.fastpath_tx_pending_max);
++ vnic->dbfs.event_count = debugfs_create_u64
++ ("event_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.event_count);
++ vnic->dbfs.bad_event_count = debugfs_create_u64
++ ("bad_event_count", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.bad_event_count);
++ vnic->dbfs.event_count_since_irq = debugfs_create_u32
++ ("event_count_since_irq", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.event_count_since_irq);
++ vnic->dbfs.events_per_irq_max = debugfs_create_u32
++ ("events_per_irq_max", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.events_per_irq_max);
++ vnic->dbfs.fastpath_frm_trunc = debugfs_create_u64
++ ("fastpath_frm_trunc", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.fastpath_frm_trunc);
++ vnic->dbfs.rx_no_desc_trunc = debugfs_create_u64
++ ("rx_no_desc_trunc", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.rx_no_desc_trunc);
++ vnic->dbfs.events_per_poll_max = debugfs_create_u32
++ ("events_per_poll_max", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.events_per_poll_max);
++ vnic->dbfs.events_per_poll_rx_max = debugfs_create_u32
++ ("events_per_poll_rx_max", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.events_per_poll_rx_max);
++ vnic->dbfs.events_per_poll_tx_max = debugfs_create_u32
++ ("events_per_poll_tx_max", S_IRUSR | S_IRGRP | S_IROTH,
++ vnic->dbfs_dir, &vnic->stats.events_per_poll_tx_max);
++#endif
++#endif
++ return 0;
++}
++
++
++int netfront_accel_debugfs_remove(netfront_accel_vnic *vnic)
++{
++#if defined(CONFIG_DEBUG_FS)
++ if (vnic->dbfs_dir != NULL) {
++ debugfs_remove(vnic->netdev_dbfs.fastpath_rx_pkts);
++ debugfs_remove(vnic->netdev_dbfs.fastpath_rx_bytes);
++ debugfs_remove(vnic->netdev_dbfs.fastpath_rx_errors);
++ debugfs_remove(vnic->netdev_dbfs.fastpath_tx_pkts);
++ debugfs_remove(vnic->netdev_dbfs.fastpath_tx_bytes);
++ debugfs_remove(vnic->netdev_dbfs.fastpath_tx_errors);
++
++#if NETFRONT_ACCEL_STATS
++ debugfs_remove(vnic->dbfs.irq_count);
++ debugfs_remove(vnic->dbfs.useless_irq_count);
++ debugfs_remove(vnic->dbfs.poll_schedule_count);
++ debugfs_remove(vnic->dbfs.poll_call_count);
++ debugfs_remove(vnic->dbfs.poll_reschedule_count);
++ debugfs_remove(vnic->dbfs.queue_stops);
++ debugfs_remove(vnic->dbfs.queue_wakes);
++ debugfs_remove(vnic->dbfs.ssr_bursts);
++ debugfs_remove(vnic->dbfs.ssr_drop_stream);
++ debugfs_remove(vnic->dbfs.ssr_misorder);
++ debugfs_remove(vnic->dbfs.ssr_slow_start);
++ debugfs_remove(vnic->dbfs.ssr_merges);
++ debugfs_remove(vnic->dbfs.ssr_too_many);
++ debugfs_remove(vnic->dbfs.ssr_new_stream);
++
++ debugfs_remove(vnic->dbfs.fastpath_tx_busy);
++ debugfs_remove(vnic->dbfs.fastpath_tx_completions);
++ debugfs_remove(vnic->dbfs.fastpath_tx_pending_max);
++ debugfs_remove(vnic->dbfs.event_count);
++ debugfs_remove(vnic->dbfs.bad_event_count);
++ debugfs_remove(vnic->dbfs.event_count_since_irq);
++ debugfs_remove(vnic->dbfs.events_per_irq_max);
++ debugfs_remove(vnic->dbfs.fastpath_frm_trunc);
++ debugfs_remove(vnic->dbfs.rx_no_desc_trunc);
++ debugfs_remove(vnic->dbfs.events_per_poll_max);
++ debugfs_remove(vnic->dbfs.events_per_poll_rx_max);
++ debugfs_remove(vnic->dbfs.events_per_poll_tx_max);
++#endif
++ debugfs_remove(vnic->dbfs_dir);
++ }
++#endif
++ return 0;
++}
+diff --git a/drivers/xen/sfc_netfront/accel_msg.c b/drivers/xen/sfc_netfront/accel_msg.c
+new file mode 100644
+index 0000000..ca18379
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_msg.c
+@@ -0,0 +1,571 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/stddef.h>
++#include <linux/errno.h>
++
++#include <xen/xenbus.h>
++
++#include "accel.h"
++#include "accel_msg_iface.h"
++#include "accel_util.h"
++#include "accel_bufs.h"
++
++#include "netfront.h" /* drivers/xen/netfront/netfront.h */
++
++static void vnic_start_interrupts(netfront_accel_vnic *vnic)
++{
++ unsigned long flags;
++
++ /* Prime our interrupt */
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags);
++ if (!netfront_accel_vi_enable_interrupts(vnic)) {
++ struct netfront_info *np = netdev_priv(vnic->net_dev);
++
++ /* Cripes, that was quick, better pass it up */
++ netfront_accel_disable_net_interrupts(vnic);
++ vnic->irq_enabled = 0;
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.poll_schedule_count++);
++ netif_rx_schedule(vnic->net_dev, &np->napi);
++ } else {
++ /*
++ * Nothing yet, make sure we get interrupts through
++ * back end
++ */
++ vnic->irq_enabled = 1;
++ netfront_accel_enable_net_interrupts(vnic);
++ }
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++}
++
++
++static void vnic_stop_interrupts(netfront_accel_vnic *vnic)
++{
++ unsigned long flags;
++
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags);
++ netfront_accel_disable_net_interrupts(vnic);
++ vnic->irq_enabled = 0;
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++}
++
++
++static void vnic_start_fastpath(netfront_accel_vnic *vnic)
++{
++ struct net_device *net_dev = vnic->net_dev;
++ struct netfront_info *np = netdev_priv(net_dev);
++ unsigned long flags;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ spin_lock_irqsave(&vnic->tx_lock, flags);
++ vnic->tx_enabled = 1;
++ spin_unlock_irqrestore(&vnic->tx_lock, flags);
++
++ napi_disable(&np->napi);
++ vnic->poll_enabled = 1;
++ napi_enable(&np->napi);
++
++ vnic_start_interrupts(vnic);
++}
++
++
++void vnic_stop_fastpath(netfront_accel_vnic *vnic)
++{
++ struct net_device *net_dev = vnic->net_dev;
++ struct netfront_info *np = (struct netfront_info *)netdev_priv(net_dev);
++ unsigned long flags1, flags2;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ vnic_stop_interrupts(vnic);
++
++ spin_lock_irqsave(&vnic->tx_lock, flags1);
++ vnic->tx_enabled = 0;
++ spin_lock_irqsave(&np->tx_lock, flags2);
++ if (vnic->tx_skb != NULL) {
++ dev_kfree_skb_any(vnic->tx_skb);
++ vnic->tx_skb = NULL;
++ if (netfront_check_queue_ready(net_dev)) {
++ netif_wake_queue(net_dev);
++ NETFRONT_ACCEL_STATS_OP
++ (vnic->stats.queue_wakes++);
++ }
++ }
++ spin_unlock_irqrestore(&np->tx_lock, flags2);
++ spin_unlock_irqrestore(&vnic->tx_lock, flags1);
++
++ /* Must prevent polls and hold lock to modify poll_enabled */
++ napi_disable(&np->napi);
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags1);
++ vnic->poll_enabled = 0;
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags1);
++ napi_enable(&np->napi);
++}
++
++
++static void netfront_accel_interface_up(netfront_accel_vnic *vnic)
++{
++
++ if (!vnic->backend_netdev_up) {
++ vnic->backend_netdev_up = 1;
++
++ if (vnic->frontend_ready)
++ vnic_start_fastpath(vnic);
++ }
++}
++
++
++static void netfront_accel_interface_down(netfront_accel_vnic *vnic)
++{
++
++ if (vnic->backend_netdev_up) {
++ vnic->backend_netdev_up = 0;
++
++ if (vnic->frontend_ready)
++ vnic_stop_fastpath(vnic);
++ }
++}
++
++
++static int vnic_add_bufs(netfront_accel_vnic *vnic,
++ struct net_accel_msg *msg)
++{
++ int rc, offset;
++ struct netfront_accel_bufinfo *bufinfo;
++
++ BUG_ON(msg->u.mapbufs.pages > NET_ACCEL_MSG_MAX_PAGE_REQ);
++
++ offset = msg->u.mapbufs.reqid;
++
++ if (offset < vnic->bufpages.max_pages -
++ (vnic->bufpages.max_pages / sfc_netfront_buffer_split)) {
++ bufinfo = vnic->rx_bufs;
++ } else
++ bufinfo = vnic->tx_bufs;
++
++ /* Queue up some Rx buffers to start things off. */
++ if ((rc = netfront_accel_add_bufs(&vnic->bufpages, bufinfo, msg)) == 0) {
++ netfront_accel_vi_add_bufs(vnic, bufinfo == vnic->rx_bufs);
++
++ if (offset + msg->u.mapbufs.pages == vnic->bufpages.max_pages) {
++ VPRINTK("%s: got all buffers back\n", __FUNCTION__);
++ vnic->frontend_ready = 1;
++ if (vnic->backend_netdev_up)
++ vnic_start_fastpath(vnic);
++ } else {
++ VPRINTK("%s: got buffers back %d %d\n", __FUNCTION__,
++ offset, msg->u.mapbufs.pages);
++ }
++ }
++
++ return rc;
++}
++
++
++/* The largest [o] such that (1u << o) <= n. Requires n > 0. */
++
++inline unsigned log2_le(unsigned long n) {
++ unsigned order = 1;
++ while ((1ul << order) <= n) ++order;
++ return (order - 1);
++}
++
++static int vnic_send_buffer_requests(netfront_accel_vnic *vnic,
++ struct netfront_accel_bufpages *bufpages)
++{
++ int pages, offset, rc = 0, sent = 0;
++ struct net_accel_msg msg;
++
++ while (bufpages->page_reqs < bufpages->max_pages) {
++ offset = bufpages->page_reqs;
++
++ pages = pow2(log2_le(bufpages->max_pages -
++ bufpages->page_reqs));
++ pages = pages < NET_ACCEL_MSG_MAX_PAGE_REQ ?
++ pages : NET_ACCEL_MSG_MAX_PAGE_REQ;
++
++ BUG_ON(offset < 0);
++ BUG_ON(pages <= 0);
++
++ rc = netfront_accel_buf_map_request(vnic->dev, bufpages,
++ &msg, pages, offset);
++ if (rc == 0) {
++ rc = net_accel_msg_send(vnic->shared_page,
++ &vnic->to_dom0, &msg);
++ if (rc < 0) {
++ VPRINTK("%s: queue full, stopping for now\n",
++ __FUNCTION__);
++ break;
++ }
++ sent++;
++ } else {
++ EPRINTK("%s: problem with grant, stopping for now\n",
++ __FUNCTION__);
++ break;
++ }
++
++ bufpages->page_reqs += pages;
++ }
++
++ if (sent)
++ net_accel_msg_notify(vnic->msg_channel_irq);
++
++ return rc;
++}
++
++
++/*
++ * In response to dom0 saying "my queue is full", we reply with this
++ * when it is no longer full
++ */
++inline void vnic_set_queue_not_full(netfront_accel_vnic *vnic)
++{
++
++ if (test_and_set_bit(NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL_B,
++ (unsigned long *)&vnic->shared_page->aflags))
++ notify_remote_via_irq(vnic->msg_channel_irq);
++ else
++ VPRINTK("queue not full bit already set, not signalling\n");
++}
++
++/*
++ * Notify dom0 that the queue we want to use is full, it should
++ * respond by setting MSG_AFLAGS_QUEUEUNOTFULL in due course
++ */
++inline void vnic_set_queue_full(netfront_accel_vnic *vnic)
++{
++
++ if (!test_and_set_bit(NET_ACCEL_MSG_AFLAGS_QUEUEUFULL_B,
++ (unsigned long *)&vnic->shared_page->aflags))
++ notify_remote_via_irq(vnic->msg_channel_irq);
++ else
++ VPRINTK("queue full bit already set, not signalling\n");
++}
++
++
++static int vnic_check_hello_version(unsigned version)
++{
++ if (version > NET_ACCEL_MSG_VERSION) {
++ /* Newer protocol, we must refuse */
++ return -EPROTO;
++ }
++
++ if (version < NET_ACCEL_MSG_VERSION) {
++ /*
++ * We are newer, so have discretion to accept if we
++ * wish. For now however, just reject
++ */
++ return -EPROTO;
++ }
++
++ BUG_ON(version != NET_ACCEL_MSG_VERSION);
++ return 0;
++}
++
++
++static int vnic_process_hello_msg(netfront_accel_vnic *vnic,
++ struct net_accel_msg *msg)
++{
++ int err = 0;
++ unsigned pages = sfc_netfront_max_pages;
++
++ if (vnic_check_hello_version(msg->u.hello.version) < 0) {
++ msg->id = NET_ACCEL_MSG_HELLO | NET_ACCEL_MSG_REPLY
++ | NET_ACCEL_MSG_ERROR;
++ msg->u.hello.version = NET_ACCEL_MSG_VERSION;
++ } else {
++ vnic->backend_netdev_up
++ = vnic->shared_page->net_dev_up;
++
++ msg->id = NET_ACCEL_MSG_HELLO | NET_ACCEL_MSG_REPLY;
++ msg->u.hello.version = NET_ACCEL_MSG_VERSION;
++ if (msg->u.hello.max_pages &&
++ msg->u.hello.max_pages < pages)
++ pages = msg->u.hello.max_pages;
++ msg->u.hello.max_pages = pages;
++
++ /* Half of pages for rx, half for tx */
++ err = netfront_accel_alloc_buffer_mem(&vnic->bufpages,
++ vnic->rx_bufs,
++ vnic->tx_bufs,
++ pages);
++ if (err)
++ msg->id |= NET_ACCEL_MSG_ERROR;
++ }
++
++ /* Send reply */
++ net_accel_msg_reply_notify(vnic->shared_page, vnic->msg_channel_irq,
++ &vnic->to_dom0, msg);
++ return err;
++}
++
++
++static int vnic_process_localmac_msg(netfront_accel_vnic *vnic,
++ struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ cuckoo_hash_mac_key key;
++
++ if (msg->u.localmac.flags & NET_ACCEL_MSG_ADD) {
++ DECLARE_MAC_BUF(buf);
++
++ DPRINTK("MAC has moved, could be local: %s\n",
++ print_mac(buf, msg->u.localmac.mac));
++ key = cuckoo_mac_to_key(msg->u.localmac.mac);
++ spin_lock_irqsave(&vnic->table_lock, flags);
++ /* Try to remove it, not a big deal if not there */
++ cuckoo_hash_remove(&vnic->fastpath_table,
++ (cuckoo_hash_key *)&key);
++ spin_unlock_irqrestore(&vnic->table_lock, flags);
++ }
++
++ return 0;
++}
++
++
++static
++int vnic_process_rx_msg(netfront_accel_vnic *vnic,
++ struct net_accel_msg *msg)
++{
++ int err;
++
++ switch (msg->id) {
++ case NET_ACCEL_MSG_HELLO:
++ /* Hello, reply with Reply */
++ DPRINTK("got Hello, with version %.8x\n",
++ msg->u.hello.version);
++ BUG_ON(vnic->msg_state != NETFRONT_ACCEL_MSG_NONE);
++ err = vnic_process_hello_msg(vnic, msg);
++ if (err == 0)
++ vnic->msg_state = NETFRONT_ACCEL_MSG_HELLO;
++ break;
++ case NET_ACCEL_MSG_SETHW:
++ /* Hardware info message */
++ DPRINTK("got H/W info\n");
++ BUG_ON(vnic->msg_state != NETFRONT_ACCEL_MSG_HELLO);
++ err = netfront_accel_vi_init(vnic, &msg->u.hw);
++ if (err == 0)
++ vnic->msg_state = NETFRONT_ACCEL_MSG_HW;
++ break;
++ case NET_ACCEL_MSG_MAPBUF | NET_ACCEL_MSG_REPLY:
++ VPRINTK("Got mapped buffers back\n");
++ BUG_ON(vnic->msg_state != NETFRONT_ACCEL_MSG_HW);
++ err = vnic_add_bufs(vnic, msg);
++ break;
++ case NET_ACCEL_MSG_MAPBUF | NET_ACCEL_MSG_REPLY | NET_ACCEL_MSG_ERROR:
++ /* No buffers. Can't use the fast path. */
++ EPRINTK("Got mapped buffers error. Cannot accelerate.\n");
++ BUG_ON(vnic->msg_state != NETFRONT_ACCEL_MSG_HW);
++ err = -EIO;
++ break;
++ case NET_ACCEL_MSG_LOCALMAC:
++ /* Should be add, remove not currently used */
++ EPRINTK_ON(!(msg->u.localmac.flags & NET_ACCEL_MSG_ADD));
++ BUG_ON(vnic->msg_state != NETFRONT_ACCEL_MSG_HW);
++ err = vnic_process_localmac_msg(vnic, msg);
++ break;
++ default:
++ EPRINTK("Huh? Message code is 0x%x\n", msg->id);
++ err = -EPROTO;
++ break;
++ }
++
++ return err;
++}
++
++
++/* Process an IRQ received from back end driver */
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++void netfront_accel_msg_from_bend(struct work_struct *context)
++#else
++void netfront_accel_msg_from_bend(void *context)
++#endif
++{
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ netfront_accel_vnic *vnic =
++ container_of(context, netfront_accel_vnic, msg_from_bend);
++#else
++ netfront_accel_vnic *vnic = (netfront_accel_vnic *)context;
++#endif
++ struct net_accel_msg msg;
++ int err, queue_was_full = 0;
++
++ mutex_lock(&vnic->vnic_mutex);
++
++ /*
++ * This happens when the shared pages have been unmapped but
++ * the workqueue has yet to be flushed
++ */
++ if (!vnic->dom0_state_is_setup)
++ goto unlock_out;
++
++ while ((vnic->shared_page->aflags & NET_ACCEL_MSG_AFLAGS_TO_DOMU_MASK)
++ != 0) {
++ if (vnic->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL) {
++ /* We've been told there may now be space. */
++ clear_bit(NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL_B,
++ (unsigned long *)&vnic->shared_page->aflags);
++ }
++
++ if (vnic->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_QUEUE0FULL) {
++ /*
++ * There will be space at the end of this
++ * function if we can make any.
++ */
++ clear_bit(NET_ACCEL_MSG_AFLAGS_QUEUE0FULL_B,
++ (unsigned long *)&vnic->shared_page->aflags);
++ queue_was_full = 1;
++ }
++
++ if (vnic->shared_page->aflags &
++ NET_ACCEL_MSG_AFLAGS_NETUPDOWN) {
++ DPRINTK("%s: net interface change\n", __FUNCTION__);
++ clear_bit(NET_ACCEL_MSG_AFLAGS_NETUPDOWN_B,
++ (unsigned long *)&vnic->shared_page->aflags);
++ if (vnic->shared_page->net_dev_up)
++ netfront_accel_interface_up(vnic);
++ else
++ netfront_accel_interface_down(vnic);
++ }
++ }
++
++ /* Pull msg out of shared memory */
++ while ((err = net_accel_msg_recv(vnic->shared_page, &vnic->from_dom0,
++ &msg)) == 0) {
++ err = vnic_process_rx_msg(vnic, &msg);
++
++ if (err != 0)
++ goto done;
++ }
++
++ /*
++ * Send any pending buffer map request messages that we can,
++ * and mark domU->dom0 as full if necessary.
++ */
++ if (vnic->msg_state == NETFRONT_ACCEL_MSG_HW &&
++ vnic->bufpages.page_reqs < vnic->bufpages.max_pages) {
++ if (vnic_send_buffer_requests(vnic, &vnic->bufpages) == -ENOSPC)
++ vnic_set_queue_full(vnic);
++ }
++
++ /*
++ * If there are no messages then this is not an error. It
++ * just means that we've finished processing the queue.
++ */
++ if (err == -ENOENT)
++ err = 0;
++ done:
++ /* We will now have made space in the dom0->domU queue if we can */
++ if (queue_was_full)
++ vnic_set_queue_not_full(vnic);
++
++ if (err != 0) {
++ EPRINTK("%s returned %d\n", __FUNCTION__, err);
++ netfront_accel_set_closing(vnic);
++ }
++
++ unlock_out:
++ mutex_unlock(&vnic->vnic_mutex);
++
++ return;
++}
++
++
++irqreturn_t netfront_accel_msg_channel_irq_from_bend(int irq, void *context)
++{
++ netfront_accel_vnic *vnic = (netfront_accel_vnic *)context;
++ VPRINTK("irq %d from device %s\n", irq, vnic->dev->nodename);
++
++ queue_work(netfront_accel_workqueue, &vnic->msg_from_bend);
++
++ return IRQ_HANDLED;
++}
++
++/* Process an interrupt received from the NIC via backend */
++irqreturn_t netfront_accel_net_channel_irq_from_bend(int irq, void *context)
++{
++ netfront_accel_vnic *vnic = (netfront_accel_vnic *)context;
++ struct net_device *net_dev = vnic->net_dev;
++ unsigned long flags;
++
++ VPRINTK("net irq %d from device %s\n", irq, vnic->dev->nodename);
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.irq_count++);
++
++ BUG_ON(net_dev==NULL);
++
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags);
++ if (vnic->irq_enabled) {
++ struct netfront_info *np = netdev_priv(net_dev);
++
++ netfront_accel_disable_net_interrupts(vnic);
++ vnic->irq_enabled = 0;
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++
++#if NETFRONT_ACCEL_STATS
++ vnic->stats.poll_schedule_count++;
++ if (vnic->stats.event_count_since_irq >
++ vnic->stats.events_per_irq_max)
++ vnic->stats.events_per_irq_max =
++ vnic->stats.event_count_since_irq;
++ vnic->stats.event_count_since_irq = 0;
++#endif
++ netif_rx_schedule(net_dev, &np->napi);
++ }
++ else {
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.useless_irq_count++);
++ DPRINTK("%s: irq when disabled\n", __FUNCTION__);
++ }
++
++ return IRQ_HANDLED;
++}
++
++
++void netfront_accel_msg_tx_fastpath(netfront_accel_vnic *vnic, const void *mac,
++ u32 ip, u16 port, u8 protocol)
++{
++ unsigned long lock_state;
++ struct net_accel_msg *msg;
++
++ msg = net_accel_msg_start_send(vnic->shared_page, &vnic->to_dom0,
++ &lock_state);
++
++ if (msg == NULL)
++ return;
++
++ net_accel_msg_init(msg, NET_ACCEL_MSG_FASTPATH);
++ msg->u.fastpath.flags = NET_ACCEL_MSG_REMOVE;
++ memcpy(msg->u.fastpath.mac, mac, ETH_ALEN);
++
++ msg->u.fastpath.port = port;
++ msg->u.fastpath.ip = ip;
++ msg->u.fastpath.proto = protocol;
++
++ net_accel_msg_complete_send_notify(vnic->shared_page, &vnic->to_dom0,
++ &lock_state, vnic->msg_channel_irq);
++}
+diff --git a/drivers/xen/sfc_netfront/accel_netfront.c b/drivers/xen/sfc_netfront/accel_netfront.c
+new file mode 100644
+index 0000000..6bb995d
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_netfront.c
+@@ -0,0 +1,319 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/skbuff.h>
++#include <linux/netdevice.h>
++
++/* drivers/xen/netfront/netfront.h */
++#include "netfront.h"
++
++#include "accel.h"
++#include "accel_bufs.h"
++#include "accel_util.h"
++#include "accel_msg_iface.h"
++#include "accel_ssr.h"
++
++#ifdef EFX_GCOV
++#include "gcov.h"
++#endif
++
++#define NETFRONT_ACCEL_VNIC_FROM_NETDEV(_nd) \
++ ((netfront_accel_vnic *)((struct netfront_info *)netdev_priv(net_dev))->accel_priv)
++
++static int netfront_accel_netdev_start_xmit(struct sk_buff *skb,
++ struct net_device *net_dev)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++ struct netfront_info *np =
++ (struct netfront_info *)netdev_priv(net_dev);
++ int handled, rc;
++ unsigned long flags1, flags2;
++
++ BUG_ON(vnic == NULL);
++
++ /* Take our tx lock and hold for the duration */
++ spin_lock_irqsave(&vnic->tx_lock, flags1);
++
++ if (!vnic->tx_enabled) {
++ rc = 0;
++ goto unlock_out;
++ }
++
++ handled = netfront_accel_vi_tx_post(vnic, skb);
++ if (handled == NETFRONT_ACCEL_STATUS_BUSY) {
++ BUG_ON(vnic->net_dev != net_dev);
++ DPRINTK("%s stopping queue\n", __FUNCTION__);
++
++ /* Netfront's lock protects tx_skb */
++ spin_lock_irqsave(&np->tx_lock, flags2);
++ BUG_ON(vnic->tx_skb != NULL);
++ vnic->tx_skb = skb;
++ netif_stop_queue(net_dev);
++ spin_unlock_irqrestore(&np->tx_lock, flags2);
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.queue_stops++);
++ }
++
++ if (handled == NETFRONT_ACCEL_STATUS_CANT)
++ rc = 0;
++ else
++ rc = 1;
++
++unlock_out:
++ spin_unlock_irqrestore(&vnic->tx_lock, flags1);
++
++ return rc;
++}
++
++
++static int netfront_accel_netdev_poll(struct net_device *net_dev, int *budget)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++ int rx_allowed = *budget, rx_done;
++
++ BUG_ON(vnic == NULL);
++
++ /* Can check this without lock as modifier excludes polls */
++ if (!vnic->poll_enabled)
++ return 0;
++
++ rx_done = netfront_accel_vi_poll(vnic, rx_allowed);
++ *budget -= rx_done;
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.poll_call_count++);
++
++ VPRINTK("%s: done %d allowed %d\n",
++ __FUNCTION__, rx_done, rx_allowed);
++
++ netfront_accel_ssr_end_of_burst(vnic, &vnic->ssr_state);
++
++ if (rx_done < rx_allowed) {
++ return 0; /* Done */
++ }
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.poll_reschedule_count++);
++
++ return 1; /* More to do. */
++}
++
++
++/*
++ * Process request from netfront to start napi interrupt
++ * mode. (i.e. enable interrupts as it's finished polling)
++ */
++static int netfront_accel_start_napi_interrupts(struct net_device *net_dev)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++ unsigned long flags;
++
++ BUG_ON(vnic == NULL);
++
++ /*
++ * Can check this without lock as writer excludes poll before
++ * modifying
++ */
++ if (!vnic->poll_enabled)
++ return 0;
++
++ if (!netfront_accel_vi_enable_interrupts(vnic)) {
++ /*
++ * There was something there, tell caller we had
++ * something to do.
++ */
++ return 1;
++ }
++
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags);
++ vnic->irq_enabled = 1;
++ netfront_accel_enable_net_interrupts(vnic);
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++
++ return 0;
++}
++
++
++/*
++ * Process request from netfront to stop napi interrupt
++ * mode. (i.e. disable interrupts as it's starting to poll
++ */
++static void netfront_accel_stop_napi_interrupts(struct net_device *net_dev)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++ unsigned long flags;
++
++ BUG_ON(vnic == NULL);
++
++ spin_lock_irqsave(&vnic->irq_enabled_lock, flags);
++
++ if (!vnic->poll_enabled) {
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++ return;
++ }
++
++ netfront_accel_disable_net_interrupts(vnic);
++ vnic->irq_enabled = 0;
++ spin_unlock_irqrestore(&vnic->irq_enabled_lock, flags);
++}
++
++
++static int netfront_accel_check_ready(struct net_device *net_dev)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++
++ BUG_ON(vnic == NULL);
++
++ /* This is protected by netfront's lock */
++ return vnic->tx_skb == NULL;
++}
++
++
++static int netfront_accel_get_stats(struct net_device *net_dev,
++ struct net_device_stats *stats)
++{
++ netfront_accel_vnic *vnic = NETFRONT_ACCEL_VNIC_FROM_NETDEV(net_dev);
++ struct netfront_accel_netdev_stats now;
++
++ BUG_ON(vnic == NULL);
++
++ now.fastpath_rx_pkts = vnic->netdev_stats.fastpath_rx_pkts;
++ now.fastpath_rx_bytes = vnic->netdev_stats.fastpath_rx_bytes;
++ now.fastpath_rx_errors = vnic->netdev_stats.fastpath_rx_errors;
++ now.fastpath_tx_pkts = vnic->netdev_stats.fastpath_tx_pkts;
++ now.fastpath_tx_bytes = vnic->netdev_stats.fastpath_tx_bytes;
++ now.fastpath_tx_errors = vnic->netdev_stats.fastpath_tx_errors;
++
++ stats->rx_packets += (now.fastpath_rx_pkts -
++ vnic->stats_last_read.fastpath_rx_pkts);
++ stats->rx_bytes += (now.fastpath_rx_bytes -
++ vnic->stats_last_read.fastpath_rx_bytes);
++ stats->rx_errors += (now.fastpath_rx_errors -
++ vnic->stats_last_read.fastpath_rx_errors);
++ stats->tx_packets += (now.fastpath_tx_pkts -
++ vnic->stats_last_read.fastpath_tx_pkts);
++ stats->tx_bytes += (now.fastpath_tx_bytes -
++ vnic->stats_last_read.fastpath_tx_bytes);
++ stats->tx_errors += (now.fastpath_tx_errors -
++ vnic->stats_last_read.fastpath_tx_errors);
++
++ vnic->stats_last_read = now;
++
++ return 0;
++}
++
++
++struct netfront_accel_hooks accel_hooks = {
++ .new_device = &netfront_accel_probe,
++ .remove = &netfront_accel_remove,
++ .netdev_poll = &netfront_accel_netdev_poll,
++ .start_xmit = &netfront_accel_netdev_start_xmit,
++ .start_napi_irq = &netfront_accel_start_napi_interrupts,
++ .stop_napi_irq = &netfront_accel_stop_napi_interrupts,
++ .check_ready = &netfront_accel_check_ready,
++ .get_stats = &netfront_accel_get_stats
++};
++
++
++unsigned sfc_netfront_max_pages = NETFRONT_ACCEL_DEFAULT_BUF_PAGES;
++module_param_named (max_pages, sfc_netfront_max_pages, uint, 0644);
++MODULE_PARM_DESC(max_pages, "Number of buffer pages to request");
++
++unsigned sfc_netfront_buffer_split = 2;
++module_param_named (buffer_split, sfc_netfront_buffer_split, uint, 0644);
++MODULE_PARM_DESC(buffer_split,
++ "Fraction of buffers to use for TX, rest for RX");
++
++
++const char *frontend_name = "sfc_netfront";
++
++struct workqueue_struct *netfront_accel_workqueue;
++
++static int __init netfront_accel_init(void)
++{
++ int rc;
++#ifdef EFX_GCOV
++ gcov_provider_init(THIS_MODULE);
++#endif
++
++ /*
++ * If we're running on dom0, netfront hasn't initialised
++ * itself, so we need to keep away
++ */
++ if (is_initial_xendomain())
++ return 0;
++
++ if (!is_pow2(sizeof(struct net_accel_msg)))
++ EPRINTK("%s: bad structure size\n", __FUNCTION__);
++
++ netfront_accel_workqueue = create_workqueue(frontend_name);
++
++ netfront_accel_debugfs_init();
++
++ rc = netfront_accelerator_loaded(NETFRONT_ACCEL_VERSION,
++ frontend_name, &accel_hooks);
++
++ if (rc < 0) {
++ EPRINTK("Xen netfront accelerator version mismatch\n");
++ return -EINVAL;
++ }
++
++ if (rc > 0) {
++ /*
++ * In future may want to add backwards compatibility
++ * and accept certain subsets of previous versions
++ */
++ EPRINTK("Xen netfront accelerator version mismatch\n");
++ return -EINVAL;
++ }
++
++ return 0;
++}
++module_init(netfront_accel_init);
++
++static void __exit netfront_accel_exit(void)
++{
++ if (is_initial_xendomain())
++ return;
++
++ DPRINTK("%s: unhooking\n", __FUNCTION__);
++
++ /* Unhook from normal netfront */
++ netfront_accelerator_stop(frontend_name);
++
++ DPRINTK("%s: done\n", __FUNCTION__);
++
++ netfront_accel_debugfs_fini();
++
++ flush_workqueue(netfront_accel_workqueue);
++
++ destroy_workqueue(netfront_accel_workqueue);
++
++#ifdef EFX_GCOV
++ gcov_provider_fini(THIS_MODULE);
++#endif
++ return;
++}
++module_exit(netfront_accel_exit);
++
++MODULE_LICENSE("GPL");
++
+diff --git a/drivers/xen/sfc_netfront/accel_ssr.c b/drivers/xen/sfc_netfront/accel_ssr.c
+new file mode 100644
+index 0000000..9c44144
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_ssr.c
+@@ -0,0 +1,308 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/socket.h>
++#include <linux/in.h>
++#include <linux/ip.h>
++#include <linux/tcp.h>
++#include <linux/list.h>
++#include <net/ip.h>
++#include <net/checksum.h>
++
++#include "accel.h"
++#include "accel_util.h"
++#include "accel_bufs.h"
++
++#include "accel_ssr.h"
++
++static inline int list_valid(struct list_head *lh) {
++ return(lh->next != NULL);
++}
++
++static void netfront_accel_ssr_deliver (struct netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st,
++ struct netfront_accel_ssr_conn *c);
++
++/** Construct an efx_ssr_state.
++ *
++ * @v st The SSR state (per channel per port)
++ * @v port The port.
++ */
++void netfront_accel_ssr_init(struct netfront_accel_ssr_state *st) {
++ unsigned i;
++
++ INIT_LIST_HEAD(&st->conns);
++ INIT_LIST_HEAD(&st->free_conns);
++ for (i = 0; i < 8; ++i) {
++ struct netfront_accel_ssr_conn *c =
++ kmalloc(sizeof(*c), GFP_KERNEL);
++ if (c == NULL) break;
++ c->n_in_order_pkts = 0;
++ c->skb = NULL;
++ list_add(&c->link, &st->free_conns);
++ }
++
++}
++
++
++/** Destructor for an efx_ssr_state.
++ *
++ * @v st The SSR state (per channel per port)
++ */
++void netfront_accel_ssr_fini(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st) {
++ struct netfront_accel_ssr_conn *c;
++
++ /* Return cleanly if efx_ssr_init() not previously called */
++ BUG_ON(list_valid(&st->conns) != list_valid(&st->free_conns));
++ if (! list_valid(&st->conns))
++ return;
++
++ while ( ! list_empty(&st->free_conns)) {
++ c = list_entry(st->free_conns.prev,
++ struct netfront_accel_ssr_conn, link);
++ list_del(&c->link);
++ BUG_ON(c->skb != NULL);
++ kfree(c);
++ }
++ while ( ! list_empty(&st->conns)) {
++ c = list_entry(st->conns.prev,
++ struct netfront_accel_ssr_conn, link);
++ list_del(&c->link);
++ if (c->skb)
++ netfront_accel_ssr_deliver(vnic, st, c);
++ kfree(c);
++ }
++}
++
++
++/** Calc IP checksum and deliver to the OS
++ *
++ * @v st The SSR state (per channel per port)
++ * @v c The SSR connection state
++ */
++static void netfront_accel_ssr_deliver(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st,
++ struct netfront_accel_ssr_conn *c) {
++ BUG_ON(c->skb == NULL);
++
++ /*
++ * If we've chained packets together, recalculate the IP
++ * checksum.
++ */
++ if (skb_shinfo(c->skb)->frag_list) {
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_bursts);
++ c->iph->check = 0;
++ c->iph->check = ip_fast_csum((unsigned char *) c->iph,
++ c->iph->ihl);
++ }
++
++ VPRINTK("%s: %d\n", __FUNCTION__, c->skb->len);
++
++ netif_receive_skb(c->skb);
++ c->skb = NULL;
++}
++
++
++/** Push held skbs down into network stack.
++ *
++ * @v st SSR state
++ *
++ * Only called if we are tracking one or more connections.
++ */
++void __netfront_accel_ssr_end_of_burst(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st) {
++ struct netfront_accel_ssr_conn *c;
++
++ BUG_ON(list_empty(&st->conns));
++
++ list_for_each_entry(c, &st->conns, link)
++ if (c->skb)
++ netfront_accel_ssr_deliver(vnic, st, c);
++
++ /* Time-out connections that have received no traffic for 20ms. */
++ c = list_entry(st->conns.prev, struct netfront_accel_ssr_conn,
++ link);
++ if (jiffies - c->last_pkt_jiffies > (HZ / 50 + 1)) {
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_drop_stream);
++ list_del(&c->link);
++ list_add(&c->link, &st->free_conns);
++ }
++}
++
++
++/** Process SKB and decide whether to dispatch it to the stack now or
++ * later.
++ *
++ * @v st SSR state
++ * @v skb SKB to exmaine
++ * @ret rc 0 => deliver SKB to kernel now, otherwise the SKB belongs
++ * us.
++ */
++int netfront_accel_ssr_skb(struct netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st,
++ struct sk_buff *skb) {
++ int data_length, dont_merge;
++ struct netfront_accel_ssr_conn *c;
++ struct iphdr *iph;
++ struct tcphdr *th;
++ unsigned th_seq;
++
++ BUG_ON(skb_shinfo(skb)->frag_list != NULL);
++ BUG_ON(skb->next != NULL);
++
++ /* We're not interested if it isn't TCP over IPv4. */
++ iph = (struct iphdr *) skb->data;
++ if (skb->protocol != htons(ETH_P_IP) ||
++ iph->protocol != IPPROTO_TCP) {
++ return 0;
++ }
++
++ /* Ignore segments that fail csum or are fragmented. */
++ if (unlikely((skb->ip_summed - CHECKSUM_UNNECESSARY) |
++ (iph->frag_off & htons(IP_MF | IP_OFFSET)))) {
++ return 0;
++ }
++
++ th = (struct tcphdr*)(skb->data + iph->ihl * 4);
++ data_length = ntohs(iph->tot_len) - iph->ihl * 4 - th->doff * 4;
++ th_seq = ntohl(th->seq);
++ dont_merge = (data_length == 0) | th->urg | th->syn | th->rst;
++
++ list_for_each_entry(c, &st->conns, link) {
++ if ((c->saddr - iph->saddr) |
++ (c->daddr - iph->daddr) |
++ (c->source - th->source) |
++ (c->dest - th->dest ))
++ continue;
++
++ /* Re-insert at head of list to reduce lookup time. */
++ list_del(&c->link);
++ list_add(&c->link, &st->conns);
++ c->last_pkt_jiffies = jiffies;
++
++ if (unlikely(th_seq - c->next_seq)) {
++ /* Out-of-order, so start counting again. */
++ if (c->skb)
++ netfront_accel_ssr_deliver(vnic, st, c);
++ c->n_in_order_pkts = 0;
++ c->next_seq = th_seq + data_length;
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_misorder);
++ return 0;
++ }
++ c->next_seq = th_seq + data_length;
++
++ if (++c->n_in_order_pkts < 300) {
++ /* May be in slow-start, so don't merge. */
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_slow_start);
++ return 0;
++ }
++
++ if (unlikely(dont_merge)) {
++ if (c->skb)
++ netfront_accel_ssr_deliver(vnic, st, c);
++ return 0;
++ }
++
++ if (c->skb) {
++ c->iph->tot_len = ntohs(c->iph->tot_len);
++ c->iph->tot_len += data_length;
++ c->iph->tot_len = htons(c->iph->tot_len);
++ c->th->ack_seq = th->ack_seq;
++ c->th->fin |= th->fin;
++ c->th->psh |= th->psh;
++ c->th->window = th->window;
++
++ /* Remove the headers from this skb. */
++ skb_pull(skb, skb->len - data_length);
++
++ /*
++ * Tack the new skb onto the head skb's frag_list.
++ * This is exactly the format that fragmented IP
++ * datagrams are reassembled into.
++ */
++ BUG_ON(skb->next != 0);
++ if ( ! skb_shinfo(c->skb)->frag_list)
++ skb_shinfo(c->skb)->frag_list = skb;
++ else
++ c->skb_tail->next = skb;
++ c->skb_tail = skb;
++ c->skb->len += skb->len;
++ c->skb->data_len += skb->len;
++ c->skb->truesize += skb->truesize;
++
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_merges);
++
++ /*
++ * If the next packet might push this super-packet
++ * over the limit for an IP packet, deliver it now.
++ * This is slightly conservative, but close enough.
++ */
++ if (c->skb->len +
++ (PAGE_SIZE / NETFRONT_ACCEL_BUFS_PER_PAGE)
++ > 16384)
++ netfront_accel_ssr_deliver(vnic, st, c);
++
++ return 1;
++ }
++ else {
++ c->iph = iph;
++ c->th = th;
++ c->skb = skb;
++ return 1;
++ }
++ }
++
++ /* We're not yet tracking this connection. */
++
++ if (dont_merge) {
++ return 0;
++ }
++
++ if (list_empty(&st->free_conns)) {
++ c = list_entry(st->conns.prev,
++ struct netfront_accel_ssr_conn,
++ link);
++ if (c->skb) {
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_too_many);
++ return 0;
++ }
++ }
++ else {
++ c = list_entry(st->free_conns.next,
++ struct netfront_accel_ssr_conn,
++ link);
++ }
++ list_del(&c->link);
++ list_add(&c->link, &st->conns);
++ c->saddr = iph->saddr;
++ c->daddr = iph->daddr;
++ c->source = th->source;
++ c->dest = th->dest;
++ c->next_seq = th_seq + data_length;
++ c->n_in_order_pkts = 0;
++ BUG_ON(c->skb != NULL);
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.ssr_new_stream);
++ return 0;
++}
+diff --git a/drivers/xen/sfc_netfront/accel_ssr.h b/drivers/xen/sfc_netfront/accel_ssr.h
+new file mode 100644
+index 0000000..1d10f46
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_ssr.h
+@@ -0,0 +1,88 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETFRONT_ACCEL_SSR_H
++#define NETFRONT_ACCEL_SSR_H
++
++#include <linux/skbuff.h>
++#include <linux/ip.h>
++#include <linux/tcp.h>
++#include <linux/list.h>
++
++#include "accel.h"
++
++/** State for Soft Segment Reassembly (SSR). */
++
++struct netfront_accel_ssr_conn {
++ struct list_head link;
++
++ unsigned saddr, daddr;
++ unsigned short source, dest;
++
++ /** Number of in-order packets we've seen with payload. */
++ unsigned n_in_order_pkts;
++
++ /** Next in-order sequence number. */
++ unsigned next_seq;
++
++ /** Time we last saw a packet on this connection. */
++ unsigned long last_pkt_jiffies;
++
++ /** The SKB we are currently holding. If NULL, then all following
++ * fields are undefined.
++ */
++ struct sk_buff *skb;
++
++ /** The tail of the frag_list of SKBs we're holding. Only valid
++ * after at least one merge.
++ */
++ struct sk_buff *skb_tail;
++
++ /** The IP header of the skb we are holding. */
++ struct iphdr *iph;
++
++ /** The TCP header of the skb we are holding. */
++ struct tcphdr *th;
++};
++
++extern void netfront_accel_ssr_init(struct netfront_accel_ssr_state *st);
++extern void netfront_accel_ssr_fini(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st);
++
++extern void
++__netfront_accel_ssr_end_of_burst(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st);
++
++extern int netfront_accel_ssr_skb(netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st,
++ struct sk_buff *skb);
++
++static inline void
++netfront_accel_ssr_end_of_burst (netfront_accel_vnic *vnic,
++ struct netfront_accel_ssr_state *st) {
++ if ( ! list_empty(&st->conns) )
++ __netfront_accel_ssr_end_of_burst(vnic, st);
++}
++
++#endif /* NETFRONT_ACCEL_SSR_H */
+diff --git a/drivers/xen/sfc_netfront/accel_tso.c b/drivers/xen/sfc_netfront/accel_tso.c
+new file mode 100644
+index 0000000..004980f
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_tso.c
+@@ -0,0 +1,509 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/pci.h>
++#include <linux/tcp.h>
++#include <linux/ip.h>
++#include <linux/in.h>
++#include <linux/if_ether.h>
++
++#include "accel.h"
++#include "accel_util.h"
++
++#include "accel_tso.h"
++
++#define ETH_HDR_LEN(skb) skb_network_offset(skb)
++#define SKB_TCP_OFF(skb) skb_transport_offset(skb)
++#define SKB_IP_OFF(skb) skb_network_offset(skb)
++
++/*
++ * Set a maximum number of buffers in each output packet to make life
++ * a little simpler - if this is reached it will just move on to
++ * another packet
++ */
++#define ACCEL_TSO_MAX_BUFFERS (6)
++
++/** TSO State.
++ *
++ * The state used during segmentation. It is put into this data structure
++ * just to make it easy to pass into inline functions.
++ */
++struct netfront_accel_tso_state {
++ /** bytes of data we've yet to segment */
++ unsigned remaining_len;
++
++ /** current sequence number */
++ unsigned seqnum;
++
++ /** remaining space in current packet */
++ unsigned packet_space;
++
++ /** List of packets to be output, containing the buffers and
++ * iovecs to describe each packet
++ */
++ struct netfront_accel_tso_output_packet *output_packets;
++
++ /** Total number of buffers in output_packets */
++ unsigned buffers;
++
++ /** Total number of packets in output_packets */
++ unsigned packets;
++
++ /** Input Fragment Cursor.
++ *
++ * Where we are in the current fragment of the incoming SKB. These
++ * values get updated in place when we split a fragment over
++ * multiple packets.
++ */
++ struct {
++ /** address of current position */
++ void *addr;
++ /** remaining length */
++ unsigned int len;
++ } ifc; /* == ifc Input Fragment Cursor */
++
++ /** Parameters.
++ *
++ * These values are set once at the start of the TSO send and do
++ * not get changed as the routine progresses.
++ */
++ struct {
++ /* the number of bytes of header */
++ unsigned int header_length;
++
++ /* The number of bytes to put in each outgoing segment. */
++ int full_packet_size;
++
++ /* Current IP ID, host endian. */
++ unsigned ip_id;
++
++ /* Max size of each output packet payload */
++ int gso_size;
++ } p;
++};
++
++
++/**
++ * Verify that our various assumptions about sk_buffs and the conditions
++ * under which TSO will be attempted hold true.
++ *
++ * @v skb The sk_buff to check.
++ */
++static inline void tso_check_safe(struct sk_buff *skb) {
++ EPRINTK_ON(skb->protocol != htons (ETH_P_IP));
++ EPRINTK_ON(((struct ethhdr*) skb->data)->h_proto != htons (ETH_P_IP));
++ EPRINTK_ON(ip_hdr(skb)->protocol != IPPROTO_TCP);
++ EPRINTK_ON((SKB_TCP_OFF(skb) + tcp_hdrlen(skb)) > skb_headlen(skb));
++}
++
++
++
++/** Parse the SKB header and initialise state. */
++static inline void tso_start(struct netfront_accel_tso_state *st,
++ struct sk_buff *skb) {
++
++ /*
++ * All ethernet/IP/TCP headers combined size is TCP header size
++ * plus offset of TCP header relative to start of packet.
++ */
++ st->p.header_length = tcp_hdrlen(skb) + SKB_TCP_OFF(skb);
++ st->p.full_packet_size = (st->p.header_length
++ + skb_shinfo(skb)->gso_size);
++ st->p.gso_size = skb_shinfo(skb)->gso_size;
++
++ st->p.ip_id = htons(ip_hdr(skb)->id);
++ st->seqnum = ntohl(tcp_hdr(skb)->seq);
++
++ EPRINTK_ON(tcp_hdr(skb)->urg);
++ EPRINTK_ON(tcp_hdr(skb)->syn);
++ EPRINTK_ON(tcp_hdr(skb)->rst);
++
++ st->remaining_len = skb->len - st->p.header_length;
++
++ st->output_packets = NULL;
++ st->buffers = 0;
++ st->packets = 0;
++
++ VPRINTK("Starting new TSO: hl %d ps %d gso %d seq %x len %d\n",
++ st->p.header_length, st->p.full_packet_size, st->p.gso_size,
++ st->seqnum, skb->len);
++}
++
++/**
++ * Add another NIC mapped buffer onto an output packet
++ */
++static inline int tso_start_new_buffer(netfront_accel_vnic *vnic,
++ struct netfront_accel_tso_state *st,
++ int first)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct netfront_accel_pkt_desc *buf;
++
++ /* Get a mapped packet buffer */
++ buf = netfront_accel_buf_get(vnic->tx_bufs);
++ if (buf == NULL) {
++ DPRINTK("%s: No buffer for TX\n", __FUNCTION__);
++ return -1;
++ }
++
++ /* Store a bit of meta-data at the end */
++ tso_buf =(struct netfront_accel_tso_buffer *)
++ (buf->pkt_kva + NETFRONT_ACCEL_TSO_BUF_LENGTH
++ + sizeof(struct netfront_accel_tso_output_packet));
++
++ tso_buf->buf = buf;
++
++ tso_buf->length = 0;
++
++ if (first) {
++ struct netfront_accel_tso_output_packet *output_packet
++ = (struct netfront_accel_tso_output_packet *)
++ (buf->pkt_kva + NETFRONT_ACCEL_TSO_BUF_LENGTH);
++ output_packet->next = st->output_packets;
++ st->output_packets = output_packet;
++ tso_buf->next = NULL;
++ st->output_packets->tso_bufs = tso_buf;
++ st->output_packets->tso_bufs_len = 1;
++ } else {
++ tso_buf->next = st->output_packets->tso_bufs;
++ st->output_packets->tso_bufs = tso_buf;
++ st->output_packets->tso_bufs_len ++;
++ }
++
++ BUG_ON(st->output_packets->tso_bufs_len > ACCEL_TSO_MAX_BUFFERS);
++
++ st->buffers ++;
++
++ /*
++ * Store the context, set to NULL, last packet buffer will get
++ * non-NULL later
++ */
++ tso_buf->buf->skb = NULL;
++
++ return 0;
++}
++
++
++/* Generate a new header, and prepare for the new packet.
++ *
++ * @v vnic VNIC
++ * @v skb Socket buffer
++ * @v st TSO state
++ * @ret rc 0 on success, or -1 if failed to alloc header
++ */
++
++static inline
++int tso_start_new_packet(netfront_accel_vnic *vnic,
++ struct sk_buff *skb,
++ struct netfront_accel_tso_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct iphdr *tsoh_iph;
++ struct tcphdr *tsoh_th;
++ unsigned ip_length;
++
++ if (tso_start_new_buffer(vnic, st, 1) < 0) {
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_busy++);
++ return -1;
++ }
++
++ /* This has been set up by tso_start_new_buffer() */
++ tso_buf = st->output_packets->tso_bufs;
++
++ /* Copy in the header */
++ memcpy(tso_buf->buf->pkt_kva, skb->data, st->p.header_length);
++ tso_buf->length = st->p.header_length;
++
++ tsoh_th = (struct tcphdr*)
++ (tso_buf->buf->pkt_kva + SKB_TCP_OFF(skb));
++ tsoh_iph = (struct iphdr*)
++ (tso_buf->buf->pkt_kva + SKB_IP_OFF(skb));
++
++ /* Set to zero to encourage falcon to fill these in */
++ tsoh_th->check = 0;
++ tsoh_iph->check = 0;
++
++ tsoh_th->seq = htonl(st->seqnum);
++ st->seqnum += st->p.gso_size;
++
++ if (st->remaining_len > st->p.gso_size) {
++ /* This packet will not finish the TSO burst. */
++ ip_length = st->p.full_packet_size - ETH_HDR_LEN(skb);
++ tsoh_th->fin = 0;
++ tsoh_th->psh = 0;
++ } else {
++ /* This packet will be the last in the TSO burst. */
++ ip_length = (st->p.header_length - ETH_HDR_LEN(skb)
++ + st->remaining_len);
++ tsoh_th->fin = tcp_hdr(skb)->fin;
++ tsoh_th->psh = tcp_hdr(skb)->psh;
++ }
++
++ tsoh_iph->tot_len = htons(ip_length);
++
++ /* Linux leaves suitable gaps in the IP ID space for us to fill. */
++ tsoh_iph->id = st->p.ip_id++;
++ tsoh_iph->id = htons(tsoh_iph->id);
++
++ st->packet_space = st->p.gso_size;
++
++ st->packets++;
++
++ return 0;
++}
++
++
++
++static inline void tso_get_fragment(struct netfront_accel_tso_state *st,
++ int len, void *addr)
++{
++ st->ifc.len = len;
++ st->ifc.addr = addr;
++ return;
++}
++
++
++static inline void tso_unwind(netfront_accel_vnic *vnic,
++ struct netfront_accel_tso_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct netfront_accel_tso_output_packet *output_packet;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ while (st->output_packets != NULL) {
++ output_packet = st->output_packets;
++ st->output_packets = output_packet->next;
++ while (output_packet->tso_bufs != NULL) {
++ tso_buf = output_packet->tso_bufs;
++ output_packet->tso_bufs = tso_buf->next;
++
++ st->buffers --;
++ output_packet->tso_bufs_len --;
++
++ netfront_accel_buf_put(vnic->tx_bufs,
++ tso_buf->buf->buf_id);
++ }
++ }
++ BUG_ON(st->buffers != 0);
++}
++
++
++
++static inline
++void tso_fill_packet_with_fragment(netfront_accel_vnic *vnic,
++ struct netfront_accel_tso_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ int n, space;
++
++ BUG_ON(st->output_packets == NULL);
++ BUG_ON(st->output_packets->tso_bufs == NULL);
++
++ tso_buf = st->output_packets->tso_bufs;
++
++ if (st->ifc.len == 0) return;
++ if (st->packet_space == 0) return;
++ if (tso_buf->length == NETFRONT_ACCEL_TSO_BUF_LENGTH) return;
++
++ n = min(st->ifc.len, st->packet_space);
++
++ space = NETFRONT_ACCEL_TSO_BUF_LENGTH - tso_buf->length;
++ n = min(n, space);
++
++ st->packet_space -= n;
++ st->remaining_len -= n;
++ st->ifc.len -= n;
++
++ memcpy(tso_buf->buf->pkt_kva + tso_buf->length, st->ifc.addr, n);
++
++ tso_buf->length += n;
++
++ BUG_ON(tso_buf->length > NETFRONT_ACCEL_TSO_BUF_LENGTH);
++
++ st->ifc.addr += n;
++
++ return;
++}
++
++
++int netfront_accel_enqueue_skb_tso(netfront_accel_vnic *vnic,
++ struct sk_buff *skb)
++{
++ struct netfront_accel_tso_state state;
++ struct netfront_accel_tso_buffer *tso_buf = NULL;
++ struct netfront_accel_tso_output_packet *reversed_list = NULL;
++ struct netfront_accel_tso_output_packet *tmp_pkt;
++ ef_iovec iovecs[ACCEL_TSO_MAX_BUFFERS];
++ int frag_i, rc, dma_id;
++ skb_frag_t *f;
++
++ tso_check_safe(skb);
++
++ if (skb->ip_summed != CHECKSUM_PARTIAL)
++ EPRINTK("Trying to TSO send a packet without HW checksum\n");
++
++ tso_start(&state, skb);
++
++ /*
++ * Setup the first payload fragment. If the skb header area
++ * contains exactly the headers and all payload is in the frag
++ * list things are little simpler
++ */
++ if (skb_headlen(skb) == state.p.header_length) {
++ /* Grab the first payload fragment. */
++ BUG_ON(skb_shinfo(skb)->nr_frags < 1);
++ frag_i = 0;
++ f = &skb_shinfo(skb)->frags[frag_i];
++ tso_get_fragment(&state, f->size,
++ page_address(f->page) + f->page_offset);
++ } else {
++ int hl = state.p.header_length;
++ tso_get_fragment(&state, skb_headlen(skb) - hl,
++ skb->data + hl);
++ frag_i = -1;
++ }
++
++ if (tso_start_new_packet(vnic, skb, &state) < 0) {
++ DPRINTK("%s: out of first start-packet memory\n",
++ __FUNCTION__);
++ goto unwind;
++ }
++
++ while (1) {
++ tso_fill_packet_with_fragment(vnic, &state);
++
++ /* Move onto the next fragment? */
++ if (state.ifc.len == 0) {
++ if (++frag_i >= skb_shinfo(skb)->nr_frags)
++ /* End of payload reached. */
++ break;
++ f = &skb_shinfo(skb)->frags[frag_i];
++ tso_get_fragment(&state, f->size,
++ page_address(f->page) +
++ f->page_offset);
++ }
++
++ /* Start a new buffer? */
++ if ((state.output_packets->tso_bufs->length ==
++ NETFRONT_ACCEL_TSO_BUF_LENGTH) &&
++ tso_start_new_buffer(vnic, &state, 0)) {
++ DPRINTK("%s: out of start-buffer memory\n",
++ __FUNCTION__);
++ goto unwind;
++ }
++
++ /* Start at new packet? */
++ if ((state.packet_space == 0 ||
++ ((state.output_packets->tso_bufs_len >=
++ ACCEL_TSO_MAX_BUFFERS) &&
++ (state.output_packets->tso_bufs->length >=
++ NETFRONT_ACCEL_TSO_BUF_LENGTH))) &&
++ tso_start_new_packet(vnic, skb, &state) < 0) {
++ DPRINTK("%s: out of start-packet memory\n",
++ __FUNCTION__);
++ goto unwind;
++ }
++
++ }
++
++ /* Check for space */
++ if (ef_vi_transmit_space(&vnic->vi) < state.buffers) {
++ DPRINTK("%s: Not enough TX space (%d)\n",
++ __FUNCTION__, state.buffers);
++ goto unwind;
++ }
++
++ /*
++ * Store the skb context in the most recent buffer (i.e. the
++ * last buffer that will be sent)
++ */
++ state.output_packets->tso_bufs->buf->skb = skb;
++
++ /* Reverse the list of packets as we construct it on a stack */
++ while (state.output_packets != NULL) {
++ tmp_pkt = state.output_packets;
++ state.output_packets = tmp_pkt->next;
++ tmp_pkt->next = reversed_list;
++ reversed_list = tmp_pkt;
++ }
++
++ /* Pass off to hardware */
++ while (reversed_list != NULL) {
++ tmp_pkt = reversed_list;
++ reversed_list = tmp_pkt->next;
++
++ BUG_ON(tmp_pkt->tso_bufs_len > ACCEL_TSO_MAX_BUFFERS);
++ BUG_ON(tmp_pkt->tso_bufs_len == 0);
++
++ dma_id = tmp_pkt->tso_bufs->buf->buf_id;
++
++ /*
++ * Make an iovec of the buffers in the list, reversing
++ * the buffers as we go as they are constructed on a
++ * stack
++ */
++ tso_buf = tmp_pkt->tso_bufs;
++ for (frag_i = tmp_pkt->tso_bufs_len - 1;
++ frag_i >= 0;
++ frag_i--) {
++ iovecs[frag_i].iov_base = tso_buf->buf->pkt_buff_addr;
++ iovecs[frag_i].iov_len = tso_buf->length;
++ tso_buf = tso_buf->next;
++ }
++
++ rc = ef_vi_transmitv(&vnic->vi, iovecs, tmp_pkt->tso_bufs_len,
++ dma_id);
++ /*
++ * We checked for space already, so it really should
++ * succeed
++ */
++ BUG_ON(rc != 0);
++ }
++
++ /* Track number of tx fastpath stats */
++ vnic->netdev_stats.fastpath_tx_bytes += skb->len;
++ vnic->netdev_stats.fastpath_tx_pkts += state.packets;
++#if NETFRONT_ACCEL_STATS
++ {
++ unsigned n;
++ n = vnic->netdev_stats.fastpath_tx_pkts -
++ vnic->stats.fastpath_tx_completions;
++ if (n > vnic->stats.fastpath_tx_pending_max)
++ vnic->stats.fastpath_tx_pending_max = n;
++ }
++#endif
++
++ return NETFRONT_ACCEL_STATUS_GOOD;
++
++ unwind:
++ tso_unwind(vnic, &state);
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_busy++);
++
++ return NETFRONT_ACCEL_STATUS_BUSY;
++}
++
++
++
+diff --git a/drivers/xen/sfc_netfront/accel_tso.h b/drivers/xen/sfc_netfront/accel_tso.h
+new file mode 100644
+index 0000000..b9c3ca8
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_tso.h
+@@ -0,0 +1,57 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETFRONT_ACCEL_TSO_H
++#define NETFRONT_ACCEL_TSO_H
++
++#include "accel_bufs.h"
++
++/* Track the buffers used in each output packet */
++struct netfront_accel_tso_buffer {
++ struct netfront_accel_tso_buffer *next;
++ struct netfront_accel_pkt_desc *buf;
++ unsigned length;
++};
++
++/* Track the output packets formed from each input packet */
++struct netfront_accel_tso_output_packet {
++ struct netfront_accel_tso_output_packet *next;
++ struct netfront_accel_tso_buffer *tso_bufs;
++ unsigned tso_bufs_len;
++};
++
++
++/*
++ * Max available space in a buffer for data once meta-data has taken
++ * its place
++ */
++#define NETFRONT_ACCEL_TSO_BUF_LENGTH \
++ ((PAGE_SIZE / NETFRONT_ACCEL_BUFS_PER_PAGE) \
++ - sizeof(struct netfront_accel_tso_buffer) \
++ - sizeof(struct netfront_accel_tso_output_packet))
++
++int netfront_accel_enqueue_skb_tso(netfront_accel_vnic *vnic,
++ struct sk_buff *skb);
++
++#endif /* NETFRONT_ACCEL_TSO_H */
+diff --git a/drivers/xen/sfc_netfront/accel_vi.c b/drivers/xen/sfc_netfront/accel_vi.c
+new file mode 100644
+index 0000000..4dc8e9e
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_vi.c
+@@ -0,0 +1,1197 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/if_ether.h>
++#include <linux/ip.h>
++#include <net/checksum.h>
++#include <asm/io.h>
++
++#include "accel.h"
++#include "accel_util.h"
++#include "accel_bufs.h"
++#include "accel_tso.h"
++#include "accel_ssr.h"
++#include "netfront.h"
++
++#include "etherfabric/ef_vi.h"
++
++/*
++ * Max available space in a buffer for data once meta-data has taken
++ * its place
++ */
++#define NETFRONT_ACCEL_TX_BUF_LENGTH \
++ ((PAGE_SIZE / NETFRONT_ACCEL_BUFS_PER_PAGE) \
++ - sizeof(struct netfront_accel_tso_buffer))
++
++#define ACCEL_TX_MAX_BUFFERS (6)
++#define ACCEL_VI_POLL_EVENTS (8)
++
++static
++int netfront_accel_vi_init_fini(netfront_accel_vnic *vnic,
++ struct net_accel_msg_hw *hw_msg)
++{
++ struct ef_vi_nic_type nic_type;
++ struct net_accel_hw_falcon_b *hw_info;
++ void *io_kva, *evq_base, *rx_dma_kva, *tx_dma_kva, *doorbell_kva;
++ u32 *evq_gnts;
++ u32 evq_order;
++ int vi_state_size;
++ u8 vi_data[VI_MAPPINGS_SIZE];
++
++ if (hw_msg == NULL)
++ goto fini;
++
++ /* And create the local macs table lock */
++ spin_lock_init(&vnic->table_lock);
++
++ /* Create fastpath table, initial size 8, key length 8 */
++ if (cuckoo_hash_init(&vnic->fastpath_table, 3, 8)) {
++ EPRINTK("failed to allocate fastpath table\n");
++ goto fail_cuckoo;
++ }
++
++ vnic->hw.falcon.type = hw_msg->type;
++
++ switch (hw_msg->type) {
++ case NET_ACCEL_MSG_HWTYPE_FALCON_A:
++ hw_info = &hw_msg->resources.falcon_a.common;
++ /* Need the extra rptr register page on A1 */
++ io_kva = net_accel_map_iomem_page
++ (vnic->dev, hw_msg->resources.falcon_a.evq_rptr_gnt,
++ &vnic->hw.falcon.evq_rptr_mapping);
++ if (io_kva == NULL) {
++ EPRINTK("%s: evq_rptr permission failed\n", __FUNCTION__);
++ goto evq_rptr_fail;
++ }
++
++ vnic->hw.falcon.evq_rptr = io_kva +
++ (hw_info->evq_rptr & (PAGE_SIZE - 1));
++ break;
++ case NET_ACCEL_MSG_HWTYPE_FALCON_B:
++ hw_info = &hw_msg->resources.falcon_b;
++ break;
++ default:
++ goto bad_type;
++ }
++
++ /**** Event Queue ****/
++
++ /* Map the event queue pages */
++ evq_gnts = hw_info->evq_mem_gnts;
++ evq_order = hw_info->evq_order;
++
++ EPRINTK_ON(hw_info->evq_offs != 0);
++
++ DPRINTK("Will map evq %d pages\n", 1 << evq_order);
++
++ evq_base =
++ net_accel_map_grants_contig(vnic->dev, evq_gnts, 1 << evq_order,
++ &vnic->evq_mapping);
++ if (evq_base == NULL) {
++ EPRINTK("%s: evq_base failed\n", __FUNCTION__);
++ goto evq_fail;
++ }
++
++ /**** Doorbells ****/
++ /* Set up the doorbell mappings. */
++ doorbell_kva =
++ net_accel_map_iomem_page(vnic->dev, hw_info->doorbell_gnt,
++ &vnic->hw.falcon.doorbell_mapping);
++ if (doorbell_kva == NULL) {
++ EPRINTK("%s: doorbell permission failed\n", __FUNCTION__);
++ goto doorbell_fail;
++ }
++ vnic->hw.falcon.doorbell = doorbell_kva;
++
++ /* On Falcon_B we get the rptr from the doorbell page */
++ if (hw_msg->type == NET_ACCEL_MSG_HWTYPE_FALCON_B) {
++ vnic->hw.falcon.evq_rptr =
++ (u32 *)((char *)vnic->hw.falcon.doorbell
++ + hw_info->evq_rptr);
++ }
++
++ /**** DMA Queue ****/
++
++ /* Set up the DMA Queues from the message. */
++ tx_dma_kva = net_accel_map_grants_contig
++ (vnic->dev, &(hw_info->txdmaq_gnt), 1,
++ &vnic->hw.falcon.txdmaq_mapping);
++ if (tx_dma_kva == NULL) {
++ EPRINTK("%s: TX dma failed\n", __FUNCTION__);
++ goto tx_dma_fail;
++ }
++
++ rx_dma_kva = net_accel_map_grants_contig
++ (vnic->dev, &(hw_info->rxdmaq_gnt), 1,
++ &vnic->hw.falcon.rxdmaq_mapping);
++ if (rx_dma_kva == NULL) {
++ EPRINTK("%s: RX dma failed\n", __FUNCTION__);
++ goto rx_dma_fail;
++ }
++
++ /* Full confession */
++ DPRINTK("Mapped H/W"
++ " Tx DMAQ grant %x -> %p\n"
++ " Rx DMAQ grant %x -> %p\n"
++ " EVQ grant %x -> %p\n",
++ hw_info->txdmaq_gnt, tx_dma_kva,
++ hw_info->rxdmaq_gnt, rx_dma_kva,
++ evq_gnts[0], evq_base
++ );
++
++ memset(vi_data, 0, sizeof(vi_data));
++
++ /* TODO BUG11305: convert efhw_arch to ef_vi_arch
++ * e.g.
++ * arch = ef_vi_arch_from_efhw_arch(hw_info->nic_arch);
++ * assert(arch >= 0);
++ * nic_type.arch = arch;
++ */
++ nic_type.arch = (unsigned char)hw_info->nic_arch;
++ nic_type.variant = (char)hw_info->nic_variant;
++ nic_type.revision = (unsigned char)hw_info->nic_revision;
++
++ ef_vi_init_mapping_evq(vi_data, nic_type, hw_info->instance,
++ 1 << (evq_order + PAGE_SHIFT), evq_base,
++ (void *)0xdeadbeef);
++
++ ef_vi_init_mapping_vi(vi_data, nic_type, hw_info->rx_capacity,
++ hw_info->tx_capacity, hw_info->instance,
++ doorbell_kva, rx_dma_kva, tx_dma_kva, 0);
++
++ vi_state_size = ef_vi_calc_state_bytes(hw_info->rx_capacity,
++ hw_info->tx_capacity);
++ vnic->vi_state = (ef_vi_state *)kmalloc(vi_state_size, GFP_KERNEL);
++ if (vnic->vi_state == NULL) {
++ EPRINTK("%s: kmalloc for VI state failed\n", __FUNCTION__);
++ goto vi_state_fail;
++ }
++ ef_vi_init(&vnic->vi, vi_data, vnic->vi_state, &vnic->evq_state, 0);
++
++ ef_eventq_state_init(&vnic->vi);
++
++ ef_vi_state_init(&vnic->vi);
++
++ return 0;
++
++fini:
++ kfree(vnic->vi_state);
++ vnic->vi_state = NULL;
++vi_state_fail:
++ net_accel_unmap_grants_contig(vnic->dev, vnic->hw.falcon.rxdmaq_mapping);
++rx_dma_fail:
++ net_accel_unmap_grants_contig(vnic->dev, vnic->hw.falcon.txdmaq_mapping);
++tx_dma_fail:
++ net_accel_unmap_iomem_page(vnic->dev, vnic->hw.falcon.doorbell_mapping);
++ vnic->hw.falcon.doorbell = NULL;
++doorbell_fail:
++ net_accel_unmap_grants_contig(vnic->dev, vnic->evq_mapping);
++evq_fail:
++ if (vnic->hw.falcon.type == NET_ACCEL_MSG_HWTYPE_FALCON_A)
++ net_accel_unmap_iomem_page(vnic->dev,
++ vnic->hw.falcon.evq_rptr_mapping);
++ vnic->hw.falcon.evq_rptr = NULL;
++evq_rptr_fail:
++bad_type:
++ cuckoo_hash_destroy(&vnic->fastpath_table);
++fail_cuckoo:
++ return -EIO;
++}
++
++
++void netfront_accel_vi_ctor(netfront_accel_vnic *vnic)
++{
++ /* Just mark the VI as uninitialised. */
++ vnic->vi_state = NULL;
++}
++
++
++int netfront_accel_vi_init(netfront_accel_vnic *vnic, struct net_accel_msg_hw *hw_msg)
++{
++ BUG_ON(hw_msg == NULL);
++ return netfront_accel_vi_init_fini(vnic, hw_msg);
++}
++
++
++void netfront_accel_vi_dtor(netfront_accel_vnic *vnic)
++{
++ if (vnic->vi_state != NULL)
++ netfront_accel_vi_init_fini(vnic, NULL);
++}
++
++
++static
++void netfront_accel_vi_post_rx(netfront_accel_vnic *vnic, u16 id,
++ netfront_accel_pkt_desc *buf)
++{
++
++ int idx = vnic->rx_dma_batched;
++
++#if 0
++ VPRINTK("Posting buffer %d (0x%08x) for rx at index %d, space is %d\n",
++ id, buf->pkt_buff_addr, idx, ef_vi_receive_space(&vnic->vi));
++#endif
++ /* Set up a virtual buffer descriptor */
++ ef_vi_receive_init(&vnic->vi, buf->pkt_buff_addr, id,
++ /*rx_bytes=max*/0);
++
++ idx++;
++
++ vnic->rx_dma_level++;
++
++ /*
++ * Only push the descriptor to the card if we've reached the
++ * batch size. Otherwise, the descriptors can sit around for
++ * a while. There will be plenty available.
++ */
++ if (idx >= NETFRONT_ACCEL_RX_DESC_BATCH ||
++ vnic->rx_dma_level < NETFRONT_ACCEL_RX_DESC_BATCH) {
++#if 0
++ VPRINTK("Flushing %d rx descriptors.\n", idx);
++#endif
++
++ /* Push buffer to hardware */
++ ef_vi_receive_push(&vnic->vi);
++
++ idx = 0;
++ }
++
++ vnic->rx_dma_batched = idx;
++}
++
++
++inline
++void netfront_accel_vi_post_rx_or_free(netfront_accel_vnic *vnic, u16 id,
++ netfront_accel_pkt_desc *buf)
++{
++
++ VPRINTK("%s: %d\n", __FUNCTION__, id);
++
++ if (ef_vi_receive_space(&vnic->vi) <= vnic->rx_dma_batched) {
++ VPRINTK("RX space is full\n");
++ netfront_accel_buf_put(vnic->rx_bufs, id);
++ return;
++ }
++
++ VPRINTK("Completed buffer %d is reposted\n", id);
++ netfront_accel_vi_post_rx(vnic, id, buf);
++
++ /*
++ * Let's see if there's any more to be pushed out to the NIC
++ * while we're here
++ */
++ while (ef_vi_receive_space(&vnic->vi) > vnic->rx_dma_batched) {
++ /* Try to allocate a buffer. */
++ buf = netfront_accel_buf_get(vnic->rx_bufs);
++ if (buf == NULL)
++ break;
++
++ /* Add it to the rx dma queue. */
++ netfront_accel_vi_post_rx(vnic, buf->buf_id, buf);
++ }
++}
++
++
++void netfront_accel_vi_add_bufs(netfront_accel_vnic *vnic, int is_rx)
++{
++
++ while (is_rx &&
++ ef_vi_receive_space(&vnic->vi) > vnic->rx_dma_batched) {
++ netfront_accel_pkt_desc *buf;
++
++ VPRINTK("%s: %d\n", __FUNCTION__, vnic->rx_dma_level);
++
++ /* Try to allocate a buffer. */
++ buf = netfront_accel_buf_get(vnic->rx_bufs);
++
++ if (buf == NULL)
++ break;
++
++ /* Add it to the rx dma queue. */
++ netfront_accel_vi_post_rx(vnic, buf->buf_id, buf);
++ }
++
++ VPRINTK("%s: done\n", __FUNCTION__);
++}
++
++
++struct netfront_accel_multi_state {
++ unsigned remaining_len;
++
++ unsigned buffers;
++
++ struct netfront_accel_tso_buffer *output_buffers;
++
++ /* Where we are in the current fragment of the SKB. */
++ struct {
++ /* address of current position */
++ void *addr;
++ /* remaining length */
++ unsigned int len;
++ } ifc; /* == Input Fragment Cursor */
++};
++
++
++static inline void multi_post_start(struct netfront_accel_multi_state *st,
++ struct sk_buff *skb)
++{
++ st->remaining_len = skb->len;
++ st->output_buffers = NULL;
++ st->buffers = 0;
++ st->ifc.len = skb_headlen(skb);
++ st->ifc.addr = skb->data;
++}
++
++static int multi_post_start_new_buffer(netfront_accel_vnic *vnic,
++ struct netfront_accel_multi_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct netfront_accel_pkt_desc *buf;
++
++ /* Get a mapped packet buffer */
++ buf = netfront_accel_buf_get(vnic->tx_bufs);
++ if (buf == NULL) {
++ DPRINTK("%s: No buffer for TX\n", __FUNCTION__);
++ return -1;
++ }
++
++ /* Store a bit of meta-data at the end */
++ tso_buf = (struct netfront_accel_tso_buffer *)
++ (buf->pkt_kva + NETFRONT_ACCEL_TX_BUF_LENGTH);
++
++ tso_buf->buf = buf;
++
++ tso_buf->length = 0;
++
++ tso_buf->next = st->output_buffers;
++ st->output_buffers = tso_buf;
++ st->buffers++;
++
++ BUG_ON(st->buffers >= ACCEL_TX_MAX_BUFFERS);
++
++ /*
++ * Store the context, set to NULL, last packet buffer will get
++ * non-NULL later
++ */
++ tso_buf->buf->skb = NULL;
++
++ return 0;
++}
++
++
++static void
++multi_post_fill_buffer_with_fragment(netfront_accel_vnic *vnic,
++ struct netfront_accel_multi_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ unsigned n, space;
++
++ BUG_ON(st->output_buffers == NULL);
++ tso_buf = st->output_buffers;
++
++ if (st->ifc.len == 0) return;
++ if (tso_buf->length == NETFRONT_ACCEL_TX_BUF_LENGTH) return;
++
++ BUG_ON(tso_buf->length > NETFRONT_ACCEL_TX_BUF_LENGTH);
++
++ space = NETFRONT_ACCEL_TX_BUF_LENGTH - tso_buf->length;
++ n = min(st->ifc.len, space);
++
++ memcpy(tso_buf->buf->pkt_kva + tso_buf->length, st->ifc.addr, n);
++
++ st->remaining_len -= n;
++ st->ifc.len -= n;
++ tso_buf->length += n;
++ st->ifc.addr += n;
++
++ BUG_ON(tso_buf->length > NETFRONT_ACCEL_TX_BUF_LENGTH);
++
++ return;
++}
++
++
++static inline void multi_post_unwind(netfront_accel_vnic *vnic,
++ struct netfront_accel_multi_state *st)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ while (st->output_buffers != NULL) {
++ tso_buf = st->output_buffers;
++ st->output_buffers = tso_buf->next;
++ st->buffers--;
++ netfront_accel_buf_put(vnic->tx_bufs, tso_buf->buf->buf_id);
++ }
++ BUG_ON(st->buffers != 0);
++}
++
++
++static enum netfront_accel_post_status
++netfront_accel_enqueue_skb_multi(netfront_accel_vnic *vnic, struct sk_buff *skb)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct netfront_accel_multi_state state;
++ ef_iovec iovecs[ACCEL_TX_MAX_BUFFERS];
++ skb_frag_t *f;
++ int frag_i, rc, dma_id;
++
++ multi_post_start(&state, skb);
++
++ frag_i = -1;
++
++ if (skb->ip_summed == CHECKSUM_PARTIAL) {
++ /* Set to zero to encourage falcon to work it out for us */
++ *(u16*)(skb->head + skb->csum_start + skb->csum_offset) = 0;
++ }
++
++ if (multi_post_start_new_buffer(vnic, &state)) {
++ DPRINTK("%s: out of buffers\n", __FUNCTION__);
++ goto unwind;
++ }
++
++ while (1) {
++ multi_post_fill_buffer_with_fragment(vnic, &state);
++
++ /* Move onto the next fragment? */
++ if (state.ifc.len == 0) {
++ if (++frag_i >= skb_shinfo(skb)->nr_frags)
++ /* End of payload reached. */
++ break;
++ f = &skb_shinfo(skb)->frags[frag_i];
++ state.ifc.len = f->size;
++ state.ifc.addr = page_address(f->page) + f->page_offset;
++ }
++
++ /* Start a new buffer? */
++ if ((state.output_buffers->length ==
++ NETFRONT_ACCEL_TX_BUF_LENGTH) &&
++ multi_post_start_new_buffer(vnic, &state)) {
++ DPRINTK("%s: out of buffers\n", __FUNCTION__);
++ goto unwind;
++ }
++ }
++
++ /* Check for space */
++ if (ef_vi_transmit_space(&vnic->vi) < state.buffers) {
++ DPRINTK("%s: Not enough TX space (%d)\n", __FUNCTION__, state.buffers);
++ goto unwind;
++ }
++
++ /* Store the skb in what will be the last buffer's context */
++ state.output_buffers->buf->skb = skb;
++ /* Remember dma_id of what will be the last buffer */
++ dma_id = state.output_buffers->buf->buf_id;
++
++ /*
++ * Make an iovec of the buffers in the list, reversing the
++ * buffers as we go as they are constructed on a stack
++ */
++ tso_buf = state.output_buffers;
++ for (frag_i = state.buffers-1; frag_i >= 0; frag_i--) {
++ iovecs[frag_i].iov_base = tso_buf->buf->pkt_buff_addr;
++ iovecs[frag_i].iov_len = tso_buf->length;
++ tso_buf = tso_buf->next;
++ }
++
++ rc = ef_vi_transmitv(&vnic->vi, iovecs, state.buffers, dma_id);
++
++ /* Track number of tx fastpath stats */
++ vnic->netdev_stats.fastpath_tx_bytes += skb->len;
++ vnic->netdev_stats.fastpath_tx_pkts ++;
++#if NETFRONT_ACCEL_STATS
++ {
++ u32 n;
++ n = vnic->netdev_stats.fastpath_tx_pkts -
++ (u32)vnic->stats.fastpath_tx_completions;
++ if (n > vnic->stats.fastpath_tx_pending_max)
++ vnic->stats.fastpath_tx_pending_max = n;
++ }
++#endif
++ return NETFRONT_ACCEL_STATUS_GOOD;
++
++unwind:
++ multi_post_unwind(vnic, &state);
++
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_busy++);
++
++ return NETFRONT_ACCEL_STATUS_BUSY;
++}
++
++
++static enum netfront_accel_post_status
++netfront_accel_enqueue_skb_single(netfront_accel_vnic *vnic, struct sk_buff *skb)
++{
++ struct netfront_accel_tso_buffer *tso_buf;
++ struct netfront_accel_pkt_desc *buf;
++ u8 *kva;
++ int rc;
++
++ if (ef_vi_transmit_space(&vnic->vi) < 1) {
++ DPRINTK("%s: No TX space\n", __FUNCTION__);
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_busy++);
++ return NETFRONT_ACCEL_STATUS_BUSY;
++ }
++
++ buf = netfront_accel_buf_get(vnic->tx_bufs);
++ if (buf == NULL) {
++ DPRINTK("%s: No buffer for TX\n", __FUNCTION__);
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_busy++);
++ return NETFRONT_ACCEL_STATUS_BUSY;
++ }
++
++ /* Track number of tx fastpath stats */
++ vnic->netdev_stats.fastpath_tx_pkts++;
++ vnic->netdev_stats.fastpath_tx_bytes += skb->len;
++
++#if NETFRONT_ACCEL_STATS
++ {
++ u32 n;
++ n = vnic->netdev_stats.fastpath_tx_pkts -
++ (u32)vnic->stats.fastpath_tx_completions;
++ if (n > vnic->stats.fastpath_tx_pending_max)
++ vnic->stats.fastpath_tx_pending_max = n;
++ }
++#endif
++
++ /* Store the context */
++ buf->skb = skb;
++
++ kva = buf->pkt_kva;
++
++ if (skb->ip_summed == CHECKSUM_PARTIAL) {
++ /* Set to zero to encourage falcon to work it out for us */
++ *(u16*)(skb->head + skb->csum_start + skb->csum_offset) = 0;
++ }
++ NETFRONT_ACCEL_PKTBUFF_FOR_EACH_FRAGMENT
++ (skb, idx, frag_data, frag_len, {
++ /* Copy in payload */
++ VPRINTK("*** Copying %d bytes to %p\n", frag_len, kva);
++ memcpy(kva, frag_data, frag_len);
++ kva += frag_len;
++ });
++
++ VPRINTK("%s: id %d pkt %p kva %p buff_addr 0x%08x\n", __FUNCTION__,
++ buf->buf_id, buf, buf->pkt_kva, buf->pkt_buff_addr);
++
++
++ /* Set up the TSO meta-data for a single buffer/packet */
++ tso_buf = (struct netfront_accel_tso_buffer *)
++ (buf->pkt_kva + NETFRONT_ACCEL_TX_BUF_LENGTH);
++ tso_buf->next = NULL;
++ tso_buf->buf = buf;
++ tso_buf->length = skb->len;
++
++ rc = ef_vi_transmit(&vnic->vi, buf->pkt_buff_addr, skb->len,
++ buf->buf_id);
++ /* We checked for space already, so it really should succeed */
++ BUG_ON(rc != 0);
++
++ return NETFRONT_ACCEL_STATUS_GOOD;
++}
++
++
++enum netfront_accel_post_status
++netfront_accel_vi_tx_post(netfront_accel_vnic *vnic, struct sk_buff *skb)
++{
++ struct ethhdr *pkt_eth_hdr;
++ struct iphdr *pkt_ipv4_hdr;
++ int value, try_fastpath;
++
++ /*
++ * This assumes that the data field points to the dest mac
++ * address.
++ */
++ cuckoo_hash_mac_key key = cuckoo_mac_to_key(skb->data);
++
++ /*
++ * NB very important that all things that could return "CANT"
++ * are tested before things that return "BUSY" as if it it
++ * returns "BUSY" it is assumed that it won't return "CANT"
++ * next time it is tried
++ */
++
++ /*
++ * Do a fastpath send if fast path table lookup returns true.
++ * We do this without the table lock and so may get the wrong
++ * answer, but current opinion is that's not a big problem
++ */
++ try_fastpath = cuckoo_hash_lookup(&vnic->fastpath_table,
++ (cuckoo_hash_key *)(&key), &value);
++
++ if (!try_fastpath) {
++ DECLARE_MAC_BUF(buf);
++
++ VPRINTK("try fast path false for mac: %s\n",
++ print_mac(buf, skb->data));
++
++ return NETFRONT_ACCEL_STATUS_CANT;
++ }
++
++ /* Check to see if the packet can be sent. */
++ if (skb_headlen(skb) < sizeof(*pkt_eth_hdr) + sizeof(*pkt_ipv4_hdr)) {
++ EPRINTK("%s: Packet header is too small\n", __FUNCTION__);
++ return NETFRONT_ACCEL_STATUS_CANT;
++ }
++
++ pkt_eth_hdr = (void*)skb->data;
++ pkt_ipv4_hdr = (void*)(pkt_eth_hdr+1);
++
++ if (be16_to_cpu(pkt_eth_hdr->h_proto) != ETH_P_IP) {
++ DPRINTK("%s: Packet is not IPV4 (ether_type=0x%04x)\n", __FUNCTION__,
++ be16_to_cpu(pkt_eth_hdr->h_proto));
++ return NETFRONT_ACCEL_STATUS_CANT;
++ }
++
++ if (pkt_ipv4_hdr->protocol != IPPROTO_TCP &&
++ pkt_ipv4_hdr->protocol != IPPROTO_UDP) {
++ DPRINTK("%s: Packet is not TCP/UDP (ip_protocol=0x%02x)\n",
++ __FUNCTION__, pkt_ipv4_hdr->protocol);
++ return NETFRONT_ACCEL_STATUS_CANT;
++ }
++
++ VPRINTK("%s: %d bytes, gso %d\n", __FUNCTION__, skb->len,
++ skb_shinfo(skb)->gso_size);
++
++ if (skb_shinfo(skb)->gso_size) {
++ return netfront_accel_enqueue_skb_tso(vnic, skb);
++ }
++
++ if (skb->len <= NETFRONT_ACCEL_TX_BUF_LENGTH) {
++ return netfront_accel_enqueue_skb_single(vnic, skb);
++ }
++
++ return netfront_accel_enqueue_skb_multi(vnic, skb);
++}
++
++
++/*
++ * Copy the data to required end destination. NB. len is the total new
++ * length of the socket buffer, not the amount of data to copy
++ */
++inline
++int ef_vnic_copy_to_skb(netfront_accel_vnic *vnic, struct sk_buff *skb,
++ struct netfront_accel_pkt_desc *buf, int len)
++{
++ int i, extra = len - skb->len;
++ char c;
++ int pkt_stride = vnic->rx_pkt_stride;
++ int skb_stride = vnic->rx_skb_stride;
++ char *skb_start;
++
++ /*
++ * This pulls stuff into the cache - have seen performance
++ * benefit in this, but disabled by default
++ */
++ skb_start = skb->data;
++ if (pkt_stride) {
++ for (i = 0; i < len; i += pkt_stride) {
++ c += ((volatile char*)(buf->pkt_kva))[i];
++ }
++ }
++ if (skb_stride) {
++ for (i = skb->len; i < len ; i += skb_stride) {
++ c += ((volatile char*)(skb_start))[i];
++ }
++ }
++
++ if (skb_tailroom(skb) >= extra) {
++ memcpy(skb_put(skb, extra), buf->pkt_kva, extra);
++ return 0;
++ }
++
++ return -ENOSPC;
++}
++
++
++static void discard_jumbo_state(netfront_accel_vnic *vnic)
++{
++
++ if (vnic->jumbo_state.skb != NULL) {
++ dev_kfree_skb_any(vnic->jumbo_state.skb);
++
++ vnic->jumbo_state.skb = NULL;
++ }
++ vnic->jumbo_state.in_progress = 0;
++}
++
++
++static void netfront_accel_vi_rx_complete(netfront_accel_vnic *vnic,
++ struct sk_buff *skb)
++{
++ cuckoo_hash_mac_key key;
++ unsigned long flags;
++ int value;
++ struct net_device *net_dev;
++
++
++ key = cuckoo_mac_to_key(skb->data + ETH_ALEN);
++
++ /*
++ * If this is a MAC address that we want to do fast path TX
++ * to, and we don't already, add it to the fastpath table.
++ * The initial lookup is done without the table lock and so
++ * may get the wrong answer, but current opinion is that's not
++ * a big problem
++ */
++ if (is_valid_ether_addr(skb->data + ETH_ALEN) &&
++ !cuckoo_hash_lookup(&vnic->fastpath_table, (cuckoo_hash_key *)&key,
++ &value)) {
++ spin_lock_irqsave(&vnic->table_lock, flags);
++
++ cuckoo_hash_add_check(&vnic->fastpath_table,
++ (cuckoo_hash_key *)&key,
++ 1, 1);
++
++ spin_unlock_irqrestore(&vnic->table_lock, flags);
++ }
++
++ if (compare_ether_addr(skb->data, vnic->mac)) {
++ struct iphdr *ip = (struct iphdr *)(skb->data + ETH_HLEN);
++ u16 port;
++ DECLARE_MAC_BUF(buf);
++
++ DPRINTK("%s: saw wrong MAC address %s\n",
++ __FUNCTION__, print_mac(buf, skb->data));
++
++ if (ip->protocol == IPPROTO_TCP) {
++ struct tcphdr *tcp = (struct tcphdr *)
++ ((char *)ip + 4 * ip->ihl);
++ port = tcp->dest;
++ } else {
++ struct udphdr *udp = (struct udphdr *)
++ ((char *)ip + 4 * ip->ihl);
++ EPRINTK_ON(ip->protocol != IPPROTO_UDP);
++ port = udp->dest;
++ }
++
++ netfront_accel_msg_tx_fastpath(vnic, skb->data,
++ ip->daddr, port,
++ ip->protocol);
++ }
++
++ net_dev = vnic->net_dev;
++ skb->dev = net_dev;
++ skb->protocol = eth_type_trans(skb, net_dev);
++ /* CHECKSUM_UNNECESSARY as hardware has done it already */
++ skb->ip_summed = CHECKSUM_UNNECESSARY;
++
++ if (!netfront_accel_ssr_skb(vnic, &vnic->ssr_state, skb))
++ netif_receive_skb(skb);
++}
++
++
++static int netfront_accel_vi_poll_process_rx(netfront_accel_vnic *vnic,
++ ef_event *ev)
++{
++ struct netfront_accel_bufinfo *bufinfo = vnic->rx_bufs;
++ struct netfront_accel_pkt_desc *buf = NULL;
++ struct sk_buff *skb;
++ int id, len, sop = 0, cont = 0;
++
++ VPRINTK("Rx event.\n");
++ /*
++ * Complete the receive operation, and get the request id of
++ * the buffer
++ */
++ id = ef_vi_receive_done(&vnic->vi, ev);
++
++ if (id < 0 || id >= bufinfo->npages*NETFRONT_ACCEL_BUFS_PER_PAGE) {
++ EPRINTK("Rx packet %d is invalid\n", id);
++ /* Carry on round the loop if more events */
++ goto bad_packet;
++ }
++ /* Get our buffer descriptor */
++ buf = netfront_accel_buf_find(bufinfo, id);
++
++ len = EF_EVENT_RX_BYTES(*ev);
++
++ /* An RX buffer has been removed from the DMA ring. */
++ vnic->rx_dma_level--;
++
++ if (EF_EVENT_TYPE(*ev) == EF_EVENT_TYPE_RX) {
++ sop = EF_EVENT_RX_SOP(*ev);
++ cont = EF_EVENT_RX_CONT(*ev);
++
++ skb = vnic->jumbo_state.skb;
++
++ VPRINTK("Rx packet %d: %d bytes so far; sop %d; cont %d\n",
++ id, len, sop, cont);
++
++ if (sop) {
++ if (!vnic->jumbo_state.in_progress) {
++ vnic->jumbo_state.in_progress = 1;
++ BUG_ON(vnic->jumbo_state.skb != NULL);
++ } else {
++ /*
++ * This fragment shows a missing tail in
++ * previous one, but is itself possibly OK
++ */
++ DPRINTK("sop and in_progress => no tail\n");
++
++ /* Release the socket buffer we already had */
++ discard_jumbo_state(vnic);
++
++ /* Now start processing this fragment */
++ vnic->jumbo_state.in_progress = 1;
++ skb = NULL;
++ }
++ } else if (!vnic->jumbo_state.in_progress) {
++ DPRINTK("!sop and !in_progress => missing head\n");
++ goto missing_head;
++ }
++
++ if (!cont) {
++ /* Update state for next time */
++ vnic->jumbo_state.in_progress = 0;
++ vnic->jumbo_state.skb = NULL;
++ } else if (!vnic->jumbo_state.in_progress) {
++ DPRINTK("cont and !in_progress => missing head\n");
++ goto missing_head;
++ }
++
++ if (skb == NULL) {
++ BUG_ON(!sop);
++
++ if (!cont)
++ skb = alloc_skb(len+NET_IP_ALIGN, GFP_ATOMIC);
++ else
++ skb = alloc_skb(vnic->net_dev->mtu+NET_IP_ALIGN,
++ GFP_ATOMIC);
++
++ if (skb == NULL) {
++ DPRINTK("%s: Couldn't get an rx skb.\n",
++ __FUNCTION__);
++ netfront_accel_vi_post_rx_or_free(vnic, (u16)id, buf);
++ /*
++ * Dropping this fragment means we
++ * should discard the rest too
++ */
++ discard_jumbo_state(vnic);
++
++ /* Carry on round the loop if more events */
++ return 0;
++ }
++
++ }
++
++ /* Copy the data to required end destination */
++ if (ef_vnic_copy_to_skb(vnic, skb, buf, len) != 0) {
++ /*
++ * No space in the skb - suggests > MTU packet
++ * received
++ */
++ EPRINTK("%s: Rx packet too large (%d)\n",
++ __FUNCTION__, len);
++ netfront_accel_vi_post_rx_or_free(vnic, (u16)id, buf);
++ discard_jumbo_state(vnic);
++ return 0;
++ }
++
++ /* Put the buffer back in the DMA queue. */
++ netfront_accel_vi_post_rx_or_free(vnic, (u16)id, buf);
++
++ if (cont) {
++ vnic->jumbo_state.skb = skb;
++
++ return 0;
++ } else {
++ /* Track number of rx fastpath packets */
++ vnic->netdev_stats.fastpath_rx_pkts++;
++ vnic->netdev_stats.fastpath_rx_bytes += len;
++
++ netfront_accel_vi_rx_complete(vnic, skb);
++
++ return 1;
++ }
++ } else {
++ BUG_ON(EF_EVENT_TYPE(*ev) != EF_EVENT_TYPE_RX_DISCARD);
++
++ if (EF_EVENT_RX_DISCARD_TYPE(*ev)
++ == EF_EVENT_RX_DISCARD_TRUNC) {
++ DPRINTK("%s: " EF_EVENT_FMT
++ " buffer %d FRM_TRUNC q_id %d\n",
++ __FUNCTION__, EF_EVENT_PRI_ARG(*ev), id,
++ EF_EVENT_RX_DISCARD_Q_ID(*ev) );
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.fastpath_frm_trunc);
++ } else if (EF_EVENT_RX_DISCARD_TYPE(*ev)
++ == EF_EVENT_RX_DISCARD_OTHER) {
++ DPRINTK("%s: " EF_EVENT_FMT
++ " buffer %d RX_DISCARD_OTHER q_id %d\n",
++ __FUNCTION__, EF_EVENT_PRI_ARG(*ev), id,
++ EF_EVENT_RX_DISCARD_Q_ID(*ev) );
++ /*
++ * Probably tail of packet for which error has
++ * already been logged, so don't count in
++ * stats
++ */
++ } else {
++ EPRINTK("%s: " EF_EVENT_FMT
++ " buffer %d rx discard type %d q_id %d\n",
++ __FUNCTION__, EF_EVENT_PRI_ARG(*ev), id,
++ EF_EVENT_RX_DISCARD_TYPE(*ev),
++ EF_EVENT_RX_DISCARD_Q_ID(*ev) );
++ NETFRONT_ACCEL_STATS_OP(++vnic->stats.bad_event_count);
++ }
++ }
++
++ /* discard type drops through here */
++
++bad_packet:
++ /* Release the socket buffer we already had */
++ discard_jumbo_state(vnic);
++
++missing_head:
++ BUG_ON(vnic->jumbo_state.in_progress != 0);
++ BUG_ON(vnic->jumbo_state.skb != NULL);
++
++ if (id >= 0 && id < bufinfo->npages*NETFRONT_ACCEL_BUFS_PER_PAGE)
++ /* Put the buffer back in the DMA queue. */
++ netfront_accel_vi_post_rx_or_free(vnic, (u16)id, buf);
++
++ vnic->netdev_stats.fastpath_rx_errors++;
++
++ DPRINTK("%s experienced bad packet/missing fragment error: %d \n",
++ __FUNCTION__, ev->rx.flags);
++
++ return 0;
++}
++
++
++static void netfront_accel_vi_not_busy(netfront_accel_vnic *vnic)
++{
++ struct netfront_info *np = ((struct netfront_info *)
++ netdev_priv(vnic->net_dev));
++ struct sk_buff *skb;
++ int handled;
++ unsigned long flags;
++
++ /*
++ * TODO if we could safely check tx_skb == NULL and return
++ * early without taking the lock, that would obviously help
++ * performance
++ */
++
++ /* Take the netfront lock which protects tx_skb. */
++ spin_lock_irqsave(&np->tx_lock, flags);
++ if (vnic->tx_skb != NULL) {
++ DPRINTK("%s trying to send spare buffer\n", __FUNCTION__);
++
++ skb = vnic->tx_skb;
++ vnic->tx_skb = NULL;
++
++ spin_unlock_irqrestore(&np->tx_lock, flags);
++
++ handled = netfront_accel_vi_tx_post(vnic, skb);
++
++ spin_lock_irqsave(&np->tx_lock, flags);
++
++ if (handled != NETFRONT_ACCEL_STATUS_BUSY) {
++ DPRINTK("%s restarting tx\n", __FUNCTION__);
++ if (netfront_check_queue_ready(vnic->net_dev)) {
++ netif_wake_queue(vnic->net_dev);
++ NETFRONT_ACCEL_STATS_OP
++ (vnic->stats.queue_wakes++);
++ }
++ } else {
++ vnic->tx_skb = skb;
++ }
++
++ /*
++ * Should never get a CANT, as it checks that before
++ * deciding it was BUSY first time round
++ */
++ BUG_ON(handled == NETFRONT_ACCEL_STATUS_CANT);
++ }
++ spin_unlock_irqrestore(&np->tx_lock, flags);
++}
++
++
++static void netfront_accel_vi_tx_complete(netfront_accel_vnic *vnic,
++ struct netfront_accel_tso_buffer *tso_buf,
++ int is_last)
++{
++ struct netfront_accel_tso_buffer *next;
++
++ /*
++ * We get a single completion for every call to
++ * ef_vi_transmitv so handle any other buffers which are part
++ * of the same packet
++ */
++ while (tso_buf != NULL) {
++ if (tso_buf->buf->skb != NULL) {
++ dev_kfree_skb_any(tso_buf->buf->skb);
++ tso_buf->buf->skb = NULL;
++ }
++
++ next = tso_buf->next;
++
++ netfront_accel_buf_put(vnic->tx_bufs, tso_buf->buf->buf_id);
++
++ tso_buf = next;
++ }
++
++ /*
++ * If this was the last one in the batch, we try and send any
++ * pending tx_skb. There should now be buffers and
++ * descriptors
++ */
++ if (is_last)
++ netfront_accel_vi_not_busy(vnic);
++}
++
++
++static void netfront_accel_vi_poll_process_tx(netfront_accel_vnic *vnic,
++ ef_event *ev)
++{
++ struct netfront_accel_pkt_desc *buf;
++ struct netfront_accel_tso_buffer *tso_buf;
++ ef_request_id ids[EF_VI_TRANSMIT_BATCH];
++ int i, n_ids;
++ unsigned long flags;
++
++ /* Get the request ids for this tx completion event. */
++ n_ids = ef_vi_transmit_unbundle(&vnic->vi, ev, ids);
++
++ /* Take the tx buffer spin lock and hold for the duration */
++ spin_lock_irqsave(&vnic->tx_lock, flags);
++
++ for (i = 0; i < n_ids; ++i) {
++ VPRINTK("Tx packet %d complete\n", ids[i]);
++ buf = netfront_accel_buf_find(vnic->tx_bufs, ids[i]);
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.fastpath_tx_completions++);
++
++ tso_buf = (struct netfront_accel_tso_buffer *)
++ (buf->pkt_kva + NETFRONT_ACCEL_TX_BUF_LENGTH);
++ BUG_ON(tso_buf->buf != buf);
++
++ netfront_accel_vi_tx_complete(vnic, tso_buf, i == (n_ids-1));
++ }
++
++ spin_unlock_irqrestore(&vnic->tx_lock, flags);
++}
++
++
++int netfront_accel_vi_poll(netfront_accel_vnic *vnic, int rx_packets)
++{
++ ef_event ev[ACCEL_VI_POLL_EVENTS];
++ int rx_remain = rx_packets, rc, events, i;
++#if NETFRONT_ACCEL_STATS
++ int n_evs_polled = 0, rx_evs_polled = 0, tx_evs_polled = 0;
++#endif
++ BUG_ON(rx_packets <= 0);
++
++ events = ef_eventq_poll(&vnic->vi, ev,
++ min(rx_remain, ACCEL_VI_POLL_EVENTS));
++ i = 0;
++ NETFRONT_ACCEL_STATS_OP(n_evs_polled += events);
++
++ VPRINTK("%s: %d events\n", __FUNCTION__, events);
++
++ /* Loop over each event */
++ while (events) {
++ VPRINTK("%s: Event "EF_EVENT_FMT", index %lu\n", __FUNCTION__,
++ EF_EVENT_PRI_ARG(ev[i]),
++ (unsigned long)(vnic->vi.evq_state->evq_ptr));
++
++ if ((EF_EVENT_TYPE(ev[i]) == EF_EVENT_TYPE_RX) ||
++ (EF_EVENT_TYPE(ev[i]) == EF_EVENT_TYPE_RX_DISCARD)) {
++ rc = netfront_accel_vi_poll_process_rx(vnic, &ev[i]);
++ rx_remain -= rc;
++ BUG_ON(rx_remain < 0);
++ NETFRONT_ACCEL_STATS_OP(rx_evs_polled++);
++ } else if (EF_EVENT_TYPE(ev[i]) == EF_EVENT_TYPE_TX) {
++ netfront_accel_vi_poll_process_tx(vnic, &ev[i]);
++ NETFRONT_ACCEL_STATS_OP(tx_evs_polled++);
++ } else if (EF_EVENT_TYPE(ev[i]) ==
++ EF_EVENT_TYPE_RX_NO_DESC_TRUNC) {
++ DPRINTK("%s: RX_NO_DESC_TRUNC " EF_EVENT_FMT "\n",
++ __FUNCTION__, EF_EVENT_PRI_ARG(ev[i]));
++ discard_jumbo_state(vnic);
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.rx_no_desc_trunc++);
++ } else {
++ EPRINTK("Unexpected event " EF_EVENT_FMT "\n",
++ EF_EVENT_PRI_ARG(ev[i]));
++ NETFRONT_ACCEL_STATS_OP(vnic->stats.bad_event_count++);
++ }
++
++ i++;
++
++ /* Carry on round the loop if more events and more space */
++ if (i == events) {
++ if (rx_remain == 0)
++ break;
++
++ events = ef_eventq_poll(&vnic->vi, ev,
++ min(rx_remain,
++ ACCEL_VI_POLL_EVENTS));
++ i = 0;
++ NETFRONT_ACCEL_STATS_OP(n_evs_polled += events);
++ }
++ }
++
++#if NETFRONT_ACCEL_STATS
++ vnic->stats.event_count += n_evs_polled;
++ vnic->stats.event_count_since_irq += n_evs_polled;
++ if (n_evs_polled > vnic->stats.events_per_poll_max)
++ vnic->stats.events_per_poll_max = n_evs_polled;
++ if (rx_evs_polled > vnic->stats.events_per_poll_rx_max)
++ vnic->stats.events_per_poll_rx_max = rx_evs_polled;
++ if (tx_evs_polled > vnic->stats.events_per_poll_tx_max)
++ vnic->stats.events_per_poll_tx_max = tx_evs_polled;
++#endif
++
++ return rx_packets - rx_remain;
++}
++
++
++int netfront_accel_vi_enable_interrupts(netfront_accel_vnic *vnic)
++{
++ u32 sw_evq_ptr;
++
++ VPRINTK("%s: checking for event on %p\n", __FUNCTION__, &vnic->vi.evq_state);
++
++ BUG_ON(vnic == NULL);
++ BUG_ON(vnic->vi.evq_state == NULL);
++
++ /* Do a quick check for an event. */
++ if (ef_eventq_has_event(&vnic->vi)) {
++ VPRINTK("%s: found event\n", __FUNCTION__);
++ return 0;
++ }
++
++ VPRINTK("evq_ptr=0x%08x evq_mask=0x%08x\n",
++ vnic->evq_state.evq_ptr, vnic->vi.evq_mask);
++
++ /* Request a wakeup from the hardware. */
++ sw_evq_ptr = vnic->evq_state.evq_ptr & vnic->vi.evq_mask;
++
++ BUG_ON(vnic->hw.falcon.evq_rptr == NULL);
++
++ VPRINTK("Requesting wakeup at 0x%08x, rptr %p\n", sw_evq_ptr,
++ vnic->hw.falcon.evq_rptr);
++ *(volatile u32 *)(vnic->hw.falcon.evq_rptr) = (sw_evq_ptr >> 3);
++
++ return 1;
++}
+diff --git a/drivers/xen/sfc_netfront/accel_xenbus.c b/drivers/xen/sfc_netfront/accel_xenbus.c
+new file mode 100644
+index 0000000..0b04b4f
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/accel_xenbus.c
+@@ -0,0 +1,776 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/stddef.h>
++#include <linux/errno.h>
++
++#include <xen/xenbus.h>
++#include <xen/evtchn.h>
++#include <xen/gnttab.h>
++
++#include "accel.h"
++#include "accel_util.h"
++#include "accel_msg_iface.h"
++#include "accel_bufs.h"
++#include "accel_ssr.h"
++/* drivers/xen/netfront/netfront.h */
++#include "netfront.h"
++
++void netfront_accel_set_closing(netfront_accel_vnic *vnic)
++{
++
++ vnic->frontend_state = XenbusStateClosing;
++ net_accel_update_state(vnic->dev, XenbusStateClosing);
++}
++
++
++static void mac_address_change(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ netfront_accel_vnic *vnic;
++ struct xenbus_device *dev;
++ int rc;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ vnic = container_of(watch, netfront_accel_vnic,
++ mac_address_watch);
++ dev = vnic->dev;
++
++ rc = net_accel_xen_net_read_mac(dev, vnic->mac);
++
++ if (rc != 0)
++ EPRINTK("%s: failed to read mac (%d)\n", __FUNCTION__, rc);
++}
++
++
++static int setup_mac_address_watch(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ int err;
++
++ DPRINTK("Setting watch on %s/%s\n", dev->nodename, "mac");
++
++ err = xenbus_watch_path2(dev, dev->nodename, "mac",
++ &vnic->mac_address_watch,
++ mac_address_change);
++ if (err) {
++ EPRINTK("%s: Failed to register xenbus watch: %d\n",
++ __FUNCTION__, err);
++ goto fail;
++ }
++
++ return 0;
++ fail:
++ vnic->mac_address_watch.node = NULL;
++ return err;
++}
++
++
++/* Grant access to some pages and publish through xenbus */
++static int make_named_grant(struct xenbus_device *dev, void *page,
++ const char *name, grant_ref_t *gnt_ref)
++{
++ struct xenbus_transaction tr;
++ int err;
++ grant_ref_t gnt;
++
++ gnt = net_accel_grant_page(dev, virt_to_mfn(page), 0);
++ if (gnt < 0)
++ return gnt;
++
++ do {
++ err = xenbus_transaction_start(&tr);
++ if (err != 0) {
++ EPRINTK("%s: transaction start failed %d\n",
++ __FUNCTION__, err);
++ return err;
++ }
++ err = xenbus_printf(tr, dev->nodename, name, "%d", gnt);
++ if (err != 0) {
++ EPRINTK("%s: xenbus_printf failed %d\n", __FUNCTION__,
++ err);
++ xenbus_transaction_end(tr, 1);
++ return err;
++ }
++ err = xenbus_transaction_end(tr, 0);
++ } while (err == -EAGAIN);
++
++ if (err != 0) {
++ EPRINTK("%s: transaction end failed %d\n", __FUNCTION__, err);
++ return err;
++ }
++
++ *gnt_ref = gnt;
++
++ return 0;
++}
++
++
++static int remove_named_grant(struct xenbus_device *dev,
++ const char *name, grant_ref_t gnt_ref)
++{
++ struct xenbus_transaction tr;
++ int err;
++
++ net_accel_ungrant_page(gnt_ref);
++
++ do {
++ err = xenbus_transaction_start(&tr);
++ if (err != 0) {
++ EPRINTK("%s: transaction start failed %d\n",
++ __FUNCTION__, err);
++ return err;
++ }
++ err = xenbus_rm(tr, dev->nodename, name);
++ if (err != 0) {
++ EPRINTK("%s: xenbus_rm failed %d\n", __FUNCTION__,
++ err);
++ xenbus_transaction_end(tr, 1);
++ return err;
++ }
++ err = xenbus_transaction_end(tr, 0);
++ } while (err == -EAGAIN);
++
++ if (err != 0) {
++ EPRINTK("%s: transaction end failed %d\n", __FUNCTION__, err);
++ return err;
++ }
++
++ return 0;
++}
++
++
++static
++netfront_accel_vnic *netfront_accel_vnic_ctor(struct net_device *net_dev,
++ struct xenbus_device *dev)
++{
++ struct netfront_info *np =
++ (struct netfront_info *)netdev_priv(net_dev);
++ netfront_accel_vnic *vnic;
++ int err;
++
++ /*
++ * A bug in earlier versions of Xen accel plugin system meant
++ * you could be probed twice for the same device on suspend
++ * cancel. Be tolerant of that.
++ */
++ if (np->accel_priv != NULL)
++ return ERR_PTR(-EALREADY);
++
++ /* Alloc mem for state */
++ vnic = kzalloc(sizeof(netfront_accel_vnic), GFP_KERNEL);
++ if (vnic == NULL) {
++ EPRINTK("%s: no memory for vnic state\n", __FUNCTION__);
++ return ERR_PTR(-ENOMEM);
++ }
++
++ spin_lock_init(&vnic->tx_lock);
++
++ mutex_init(&vnic->vnic_mutex);
++ mutex_lock(&vnic->vnic_mutex);
++
++ /* Store so state can be retrieved from device */
++ BUG_ON(np->accel_priv != NULL);
++ np->accel_priv = vnic;
++ vnic->dev = dev;
++ vnic->net_dev = net_dev;
++ spin_lock_init(&vnic->irq_enabled_lock);
++ netfront_accel_ssr_init(&vnic->ssr_state);
++
++ init_waitqueue_head(&vnic->state_wait_queue);
++ vnic->backend_state = XenbusStateUnknown;
++ vnic->frontend_state = XenbusStateClosed;
++ vnic->removing = 0;
++ vnic->domU_state_is_setup = 0;
++ vnic->dom0_state_is_setup = 0;
++ vnic->poll_enabled = 0;
++ vnic->tx_enabled = 0;
++ vnic->tx_skb = NULL;
++
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,20)
++ INIT_WORK(&vnic->msg_from_bend, netfront_accel_msg_from_bend);
++#else
++ INIT_WORK(&vnic->msg_from_bend, netfront_accel_msg_from_bend, vnic);
++#endif
++
++ netfront_accel_debugfs_create(vnic);
++
++ mutex_unlock(&vnic->vnic_mutex);
++
++ err = net_accel_xen_net_read_mac(dev, vnic->mac);
++ if (err)
++ goto fail_mac;
++
++ /* Setup a watch on the frontend's MAC address */
++ err = setup_mac_address_watch(dev, vnic);
++ if (err)
++ goto fail_mac;
++
++ return vnic;
++
++fail_mac:
++
++ mutex_lock(&vnic->vnic_mutex);
++
++ netfront_accel_debugfs_remove(vnic);
++
++ netfront_accel_ssr_fini(vnic, &vnic->ssr_state);
++
++ EPRINTK_ON(vnic->tx_skb != NULL);
++
++ vnic->frontend_state = XenbusStateUnknown;
++ net_accel_update_state(dev, XenbusStateUnknown);
++
++ mutex_unlock(&vnic->vnic_mutex);
++
++ np->accel_priv = NULL;
++ kfree(vnic);
++
++ return ERR_PTR(err);
++}
++
++
++static void netfront_accel_vnic_dtor(netfront_accel_vnic *vnic)
++{
++ struct net_device *net_dev = vnic->net_dev;
++ struct netfront_info *np =
++ (struct netfront_info *)netdev_priv(net_dev);
++
++ /*
++ * Now we don't hold the lock any more it is safe to remove
++ * this watch and synchonrise with the completion of
++ * watches
++ */
++ DPRINTK("%s: unregistering xenbus mac watch\n", __FUNCTION__);
++ unregister_xenbus_watch(&vnic->mac_address_watch);
++ kfree(vnic->mac_address_watch.node);
++
++ flush_workqueue(netfront_accel_workqueue);
++
++ mutex_lock(&vnic->vnic_mutex);
++
++ netfront_accel_debugfs_remove(vnic);
++
++ netfront_accel_ssr_fini(vnic, &vnic->ssr_state);
++
++ EPRINTK_ON(vnic->tx_skb != NULL);
++
++ vnic->frontend_state = XenbusStateUnknown;
++ net_accel_update_state(vnic->dev, XenbusStateUnknown);
++
++ mutex_unlock(&vnic->vnic_mutex);
++
++ np->accel_priv = NULL;
++ kfree(vnic);
++}
++
++
++static int vnic_setup_domU_shared_state(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ struct xenbus_transaction tr;
++ int err;
++ int msgs_per_queue;
++
++
++ DPRINTK("Setting up domU shared state.\n");
++
++ msgs_per_queue = (PAGE_SIZE/2) / sizeof(struct net_accel_msg);
++
++ /* Allocate buffer state */
++ vnic->tx_bufs = netfront_accel_init_bufs(&vnic->tx_lock);
++ if (vnic->tx_bufs == NULL) {
++ err = -ENOMEM;
++ EPRINTK("%s: Failed to allocate tx buffers\n", __FUNCTION__);
++ goto fail_tx_bufs;
++ }
++
++ vnic->rx_bufs = netfront_accel_init_bufs(NULL);
++ if (vnic->rx_bufs == NULL) {
++ err = -ENOMEM;
++ EPRINTK("%s: Failed to allocate rx buffers\n", __FUNCTION__);
++ goto fail_rx_bufs;
++ }
++
++ /*
++ * This allocates two pages, one for the shared page and one
++ * for the message queue.
++ */
++ vnic->shared_page = (struct net_accel_shared_page *)
++ __get_free_pages(GFP_KERNEL, 1);
++ if (vnic->shared_page == NULL) {
++ EPRINTK("%s: no memory for shared pages\n", __FUNCTION__);
++ err = -ENOMEM;
++ goto fail_shared_page;
++ }
++
++ net_accel_msg_init_queue
++ (&vnic->from_dom0, &vnic->shared_page->queue0,
++ (struct net_accel_msg *)((u8*)vnic->shared_page + PAGE_SIZE),
++ msgs_per_queue);
++
++ net_accel_msg_init_queue
++ (&vnic->to_dom0, &vnic->shared_page->queue1,
++ (struct net_accel_msg *)((u8*)vnic->shared_page +
++ (3 * PAGE_SIZE / 2)),
++ msgs_per_queue);
++
++ vnic->msg_state = NETFRONT_ACCEL_MSG_NONE;
++
++ err = make_named_grant(dev, vnic->shared_page, "accel-ctrl-page",
++ &vnic->ctrl_page_gnt);
++ if (err) {
++ EPRINTK("couldn't make ctrl-page named grant\n");
++ goto fail_ctrl_page_grant;
++ }
++
++ err = make_named_grant(dev, (u8*)vnic->shared_page + PAGE_SIZE,
++ "accel-msg-page", &vnic->msg_page_gnt);
++ if (err) {
++ EPRINTK("couldn't make msg-page named grant\n");
++ goto fail_msg_page_grant;
++ }
++
++ /* Create xenbus msg event channel */
++ err = bind_listening_port_to_irqhandler
++ (dev->otherend_id, netfront_accel_msg_channel_irq_from_bend,
++ IRQF_SAMPLE_RANDOM, "vnicctrl", vnic);
++ if (err < 0) {
++ EPRINTK("Couldn't bind msg event channel\n");
++ goto fail_msg_irq;
++ }
++ vnic->msg_channel_irq = err;
++ vnic->msg_channel = irq_to_evtchn_port(vnic->msg_channel_irq);
++
++ /* Create xenbus net event channel */
++ err = bind_listening_port_to_irqhandler
++ (dev->otherend_id, netfront_accel_net_channel_irq_from_bend,
++ IRQF_SAMPLE_RANDOM, "vnicfront", vnic);
++ if (err < 0) {
++ EPRINTK("Couldn't bind net event channel\n");
++ goto fail_net_irq;
++ }
++ vnic->net_channel_irq = err;
++ vnic->net_channel = irq_to_evtchn_port(vnic->net_channel_irq);
++ /* Want to ensure we don't get interrupts before we're ready */
++ netfront_accel_disable_net_interrupts(vnic);
++
++ DPRINTK("otherend %d has msg ch %u (%u) and net ch %u (%u)\n",
++ dev->otherend_id, vnic->msg_channel, vnic->msg_channel_irq,
++ vnic->net_channel, vnic->net_channel_irq);
++
++ do {
++ err = xenbus_transaction_start(&tr);
++ if (err != 0) {
++ EPRINTK("%s: Transaction start failed %d\n",
++ __FUNCTION__, err);
++ goto fail_transaction;
++ }
++
++ err = xenbus_printf(tr, dev->nodename, "accel-msg-channel",
++ "%u", vnic->msg_channel);
++ if (err != 0) {
++ EPRINTK("%s: event channel xenbus write failed %d\n",
++ __FUNCTION__, err);
++ xenbus_transaction_end(tr, 1);
++ goto fail_transaction;
++ }
++
++ err = xenbus_printf(tr, dev->nodename, "accel-net-channel",
++ "%u", vnic->net_channel);
++ if (err != 0) {
++ EPRINTK("%s: net channel xenbus write failed %d\n",
++ __FUNCTION__, err);
++ xenbus_transaction_end(tr, 1);
++ goto fail_transaction;
++ }
++
++ err = xenbus_transaction_end(tr, 0);
++ } while (err == -EAGAIN);
++
++ if (err != 0) {
++ EPRINTK("%s: Transaction end failed %d\n", __FUNCTION__, err);
++ goto fail_transaction;
++ }
++
++ DPRINTK("Completed setting up domU shared state\n");
++
++ return 0;
++
++fail_transaction:
++
++ unbind_from_irqhandler(vnic->net_channel_irq, vnic);
++fail_net_irq:
++
++ unbind_from_irqhandler(vnic->msg_channel_irq, vnic);
++fail_msg_irq:
++
++ remove_named_grant(dev, "accel-ctrl-page", vnic->ctrl_page_gnt);
++fail_msg_page_grant:
++
++ remove_named_grant(dev, "accel-msg-page", vnic->msg_page_gnt);
++fail_ctrl_page_grant:
++
++ free_pages((unsigned long)vnic->shared_page, 1);
++ vnic->shared_page = NULL;
++fail_shared_page:
++
++ netfront_accel_fini_bufs(vnic->rx_bufs);
++fail_rx_bufs:
++
++ netfront_accel_fini_bufs(vnic->tx_bufs);
++fail_tx_bufs:
++
++ /* Undo the memory allocation created when we got the HELLO */
++ netfront_accel_free_buffer_mem(&vnic->bufpages,
++ vnic->rx_bufs,
++ vnic->tx_bufs);
++
++ DPRINTK("Failed to setup domU shared state with code %d\n", err);
++
++ return err;
++}
++
++
++static void vnic_remove_domU_shared_state(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ struct xenbus_transaction tr;
++
++ /*
++ * Don't remove any watches because we currently hold the
++ * mutex and the watches take the mutex.
++ */
++
++ DPRINTK("%s: removing event channel irq handlers %d %d\n",
++ __FUNCTION__, vnic->net_channel_irq, vnic->msg_channel_irq);
++ do {
++ if (xenbus_transaction_start(&tr) != 0)
++ break;
++ xenbus_rm(tr, dev->nodename, "accel-msg-channel");
++ xenbus_rm(tr, dev->nodename, "accel-net-channel");
++ } while (xenbus_transaction_end(tr, 0) == -EAGAIN);
++
++ unbind_from_irqhandler(vnic->net_channel_irq, vnic);
++ unbind_from_irqhandler(vnic->msg_channel_irq, vnic);
++
++ /* ungrant pages for msg channel */
++ remove_named_grant(dev, "accel-ctrl-page", vnic->ctrl_page_gnt);
++ remove_named_grant(dev, "accel-msg-page", vnic->msg_page_gnt);
++ free_pages((unsigned long)vnic->shared_page, 1);
++ vnic->shared_page = NULL;
++
++ /* ungrant pages for buffers, and free buffer memory */
++ netfront_accel_free_buffer_mem(&vnic->bufpages,
++ vnic->rx_bufs,
++ vnic->tx_bufs);
++ netfront_accel_fini_bufs(vnic->rx_bufs);
++ netfront_accel_fini_bufs(vnic->tx_bufs);
++}
++
++
++static void vnic_setup_dom0_shared_state(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ DPRINTK("Setting up dom0 shared state\n");
++
++ netfront_accel_vi_ctor(vnic);
++
++ /*
++ * Message processing will be enabled when this function
++ * returns, but we might have missed an interrupt. Schedule a
++ * check just in case.
++ */
++ queue_work(netfront_accel_workqueue, &vnic->msg_from_bend);
++}
++
++
++static void vnic_remove_dom0_shared_state(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ DPRINTK("Removing dom0 shared state\n");
++
++ vnic_stop_fastpath(vnic);
++
++ netfront_accel_vi_dtor(vnic);
++}
++
++
++/*************************************************************************/
++
++/*
++ * The following code handles accelstate changes between the frontend
++ * and the backend. In response to transitions, calls the following
++ * functions in matching pairs:
++ *
++ * vnic_setup_domU_shared_state
++ * vnic_remove_domU_shared_state
++ *
++ * vnic_setup_dom0_shared_state
++ * vnic_remove_dom0_shared_state
++ *
++ * Valid state transitions for DomU are as follows:
++ *
++ * Closed->Init on probe or in response to Init from dom0
++ *
++ * Init->Connected in response to Init from dom0
++ * Init->Closing on error providing dom0 is in Init
++ * Init->Closed on remove or in response to Closing from dom0
++ *
++ * Connected->Closing on error/remove
++ * Connected->Closed in response to Closing from dom0
++ *
++ * Closing->Closed in response to Closing from dom0
++ *
++ */
++
++
++/* Function to deal with Xenbus accel state change in backend */
++static void netfront_accel_backend_accel_changed(netfront_accel_vnic *vnic,
++ XenbusState backend_state)
++{
++ struct xenbus_device *dev = vnic->dev;
++ XenbusState frontend_state;
++ int state;
++
++ DPRINTK("%s: changing from %s to %s. nodename %s, otherend %s\n",
++ __FUNCTION__, xenbus_strstate(vnic->backend_state),
++ xenbus_strstate(backend_state), dev->nodename, dev->otherend);
++
++ /*
++ * Ignore duplicate state changes. This can happen if the
++ * backend changes state twice in quick succession and the
++ * first watch fires in the frontend after the second
++ * transition has completed.
++ */
++ if (vnic->backend_state == backend_state)
++ return;
++
++ vnic->backend_state = backend_state;
++ frontend_state = vnic->frontend_state;
++
++ switch (backend_state) {
++ case XenbusStateInitialising:
++ /*
++ * It's possible for us to miss the closed state from
++ * dom0, so do the work here.
++ */
++ if (vnic->domU_state_is_setup) {
++ vnic_remove_domU_shared_state(dev, vnic);
++ vnic->domU_state_is_setup = 0;
++ }
++
++ if (frontend_state != XenbusStateInitialising) {
++ /* Make sure the backend doesn't go away. */
++ frontend_state = XenbusStateInitialising;
++ net_accel_update_state(dev, frontend_state);
++ xenbus_scanf(XBT_NIL, dev->otherend, "accelstate", "%d", &state);
++ backend_state = (XenbusState)state;
++ if (backend_state != XenbusStateInitialising)
++ break;
++ }
++
++ /* Start the new connection. */
++ if (!vnic->removing) {
++ BUG_ON(vnic->domU_state_is_setup);
++ if (vnic_setup_domU_shared_state(dev, vnic) == 0) {
++ vnic->domU_state_is_setup = 1;
++ frontend_state = XenbusStateConnected;
++ } else
++ frontend_state = XenbusStateClosing;
++ }
++ break;
++ case XenbusStateConnected:
++ if (vnic->domU_state_is_setup &&
++ !vnic->dom0_state_is_setup) {
++ vnic_setup_dom0_shared_state(dev, vnic);
++ vnic->dom0_state_is_setup = 1;
++ }
++ break;
++ default:
++ case XenbusStateClosing:
++ if (vnic->dom0_state_is_setup) {
++ vnic_remove_dom0_shared_state(dev, vnic);
++ vnic->dom0_state_is_setup = 0;
++ }
++ frontend_state = XenbusStateClosed;
++ break;
++ case XenbusStateUnknown:
++ case XenbusStateClosed:
++ if (vnic->domU_state_is_setup) {
++ vnic_remove_domU_shared_state(dev, vnic);
++ vnic->domU_state_is_setup = 0;
++ }
++ break;
++ }
++
++ if (frontend_state != vnic->frontend_state) {
++ DPRINTK("Switching from state %s (%d) to %s (%d)\n",
++ xenbus_strstate(vnic->frontend_state),
++ vnic->frontend_state,
++ xenbus_strstate(frontend_state), frontend_state);
++ vnic->frontend_state = frontend_state;
++ net_accel_update_state(dev, frontend_state);
++ }
++
++ wake_up(&vnic->state_wait_queue);
++}
++
++
++static void backend_accel_state_change(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int state;
++ netfront_accel_vnic *vnic;
++ struct xenbus_device *dev;
++
++ DPRINTK("%s\n", __FUNCTION__);
++
++ vnic = container_of(watch, struct netfront_accel_vnic,
++ backend_accel_watch);
++
++ mutex_lock(&vnic->vnic_mutex);
++
++ dev = vnic->dev;
++
++ state = (int)XenbusStateUnknown;
++ xenbus_scanf(XBT_NIL, dev->otherend, "accelstate", "%d", &state);
++ netfront_accel_backend_accel_changed(vnic, state);
++
++ mutex_unlock(&vnic->vnic_mutex);
++}
++
++
++static int setup_dom0_accel_watch(struct xenbus_device *dev,
++ netfront_accel_vnic *vnic)
++{
++ int err;
++
++ DPRINTK("Setting watch on %s/%s\n", dev->otherend, "accelstate");
++
++ err = xenbus_watch_path2(dev, dev->otherend, "accelstate",
++ &vnic->backend_accel_watch,
++ backend_accel_state_change);
++ if (err) {
++ EPRINTK("%s: Failed to register xenbus watch: %d\n",
++ __FUNCTION__, err);
++ goto fail;
++ }
++ return 0;
++ fail:
++ vnic->backend_accel_watch.node = NULL;
++ return err;
++}
++
++
++int netfront_accel_probe(struct net_device *net_dev, struct xenbus_device *dev)
++{
++ netfront_accel_vnic *vnic;
++ int err;
++
++ DPRINTK("Probe passed device %s\n", dev->nodename);
++
++ vnic = netfront_accel_vnic_ctor(net_dev, dev);
++ if (IS_ERR(vnic))
++ return PTR_ERR(vnic);
++
++ /*
++ * Setup a watch on the backend accel state. This sets things
++ * going.
++ */
++ err = setup_dom0_accel_watch(dev, vnic);
++ if (err) {
++ netfront_accel_vnic_dtor(vnic);
++ EPRINTK("%s: probe failed with code %d\n", __FUNCTION__, err);
++ return err;
++ }
++
++ /*
++ * Indicate to the other end that we're ready to start unless
++ * the watch has already fired.
++ */
++ mutex_lock(&vnic->vnic_mutex);
++ VPRINTK("setup success, updating accelstate\n");
++ if (vnic->frontend_state == XenbusStateClosed) {
++ vnic->frontend_state = XenbusStateInitialising;
++ net_accel_update_state(dev, XenbusStateInitialising);
++ }
++ mutex_unlock(&vnic->vnic_mutex);
++
++ DPRINTK("Probe done device %s\n", dev->nodename);
++
++ return 0;
++}
++
++
++int netfront_accel_remove(struct xenbus_device *dev)
++{
++ struct netfront_info *np =
++ (struct netfront_info *)dev->dev.driver_data;
++ netfront_accel_vnic *vnic = (netfront_accel_vnic *)np->accel_priv;
++
++ DPRINTK("%s %s\n", __FUNCTION__, dev->nodename);
++
++ BUG_ON(vnic == NULL);
++
++ mutex_lock(&vnic->vnic_mutex);
++
++ /* Reject any attempts to connect. */
++ vnic->removing = 1;
++
++ /* Close any existing connection. */
++ if (vnic->frontend_state == XenbusStateConnected) {
++ vnic->frontend_state = XenbusStateClosing;
++ net_accel_update_state(dev, XenbusStateClosing);
++ }
++
++ mutex_unlock(&vnic->vnic_mutex);
++
++ DPRINTK("%s waiting for release of %s\n", __FUNCTION__, dev->nodename);
++
++ /*
++ * Wait for the xenbus watch to release the shared resources.
++ * This indicates that dom0 has made the transition
++ * Closing->Closed or that dom0 was in Closed or Init and no
++ * resources were mapped.
++ */
++ wait_event(vnic->state_wait_queue,
++ !vnic->domU_state_is_setup);
++
++ /*
++ * Now we don't need this watch anymore it is safe to remove
++ * it (and so synchronise with it completing if outstanding)
++ */
++ DPRINTK("%s: unregistering xenbus accel watch\n",
++ __FUNCTION__);
++ unregister_xenbus_watch(&vnic->backend_accel_watch);
++ kfree(vnic->backend_accel_watch.node);
++
++ netfront_accel_vnic_dtor(vnic);
++
++ DPRINTK("%s done %s\n", __FUNCTION__, dev->nodename);
++
++ return 0;
++}
+diff --git a/drivers/xen/sfc_netfront/ef_vi_falcon.h b/drivers/xen/sfc_netfront/ef_vi_falcon.h
+new file mode 100644
+index 0000000..0f41726
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/ef_vi_falcon.h
+@@ -0,0 +1,172 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author slp
++ * \brief Falcon specific definitions
++ * \date 2004/08
++ */
++
++#ifndef __EF_VI_FALCON_H__
++#define __EF_VI_FALCON_H__
++
++#define EFHW_4K 0x00001000u
++#define EFHW_8K 0x00002000u
++
++/* include the autogenerated register definitions */
++
++#include "ef_vi_falcon_core.h"
++#include "ef_vi_falcon_desc.h"
++#include "ef_vi_falcon_event.h"
++
++
++/*----------------------------------------------------------------------------
++ *
++ * Helpers to turn bit shifts into dword shifts and check that the bit fields
++ * haven't overflown the dword etc. Aim is to preserve consistency with the
++ * autogenerated headers - once stable we could hard code.
++ *
++ *---------------------------------------------------------------------------*/
++
++/* mask constructors */
++#define __FALCON_MASK(WIDTH,T) ((((T)1) << (WIDTH)) - 1)
++#define __EFVI_MASK32(WIDTH) __FALCON_MASK((WIDTH),uint32_t)
++#define __EFVI_MASK64(WIDTH) __FALCON_MASK((WIDTH),uint64_t)
++
++#define __EFVI_FALCON_MASKFIELD32(LBN, WIDTH) ((uint32_t) \
++ (__EFVI_MASK32(WIDTH) << (LBN)))
++
++/* constructors for fields which span the first and second dwords */
++#define __LW(LBN) (32 - LBN)
++#define LOW(v, LBN, WIDTH) ((uint32_t) \
++ (((v) & __EFVI_MASK64(__LW((LBN)))) << (LBN)))
++#define HIGH(v, LBN, WIDTH) ((uint32_t)(((v) >> __LW((LBN))) & \
++ __EFVI_MASK64((WIDTH - __LW((LBN))))))
++/* constructors for fields within the second dword */
++#define __DW2(LBN) ((LBN) - 32)
++
++/* constructors for fields which span the second and third dwords */
++#define __LW2(LBN) (64 - LBN)
++#define LOW2(v, LBN, WIDTH) ((uint32_t) \
++ (((v) & __EFVI_MASK64(__LW2((LBN)))) << ((LBN) - 32)))
++#define HIGH2(v, LBN, WIDTH) ((uint32_t) \
++ (((v) >> __LW2((LBN))) & __EFVI_MASK64((WIDTH - __LW2((LBN))))))
++
++/* constructors for fields within the third dword */
++#define __DW3(LBN) ((LBN) - 64)
++
++
++/* constructors for fields which span the third and fourth dwords */
++#define __LW3(LBN) (96 - LBN)
++#define LOW3(v, LBN, WIDTH) ((uint32_t) \
++ (((v) & __EFVI_MASK64(__LW3((LBN)))) << ((LBN) - 64)))
++#define HIGH3(v, LBN, WIDTH) ((unit32_t) \
++ (((v) >> __LW3((LBN))) & __EFVI_MASK64((WIDTH - __LW3((LBN))))))
++
++/* constructors for fields within the fourth dword */
++#define __DW4(LBN) ((LBN) - 96)
++
++/* checks that the autogenerated headers our consistent with our model */
++#define WIDTHCHCK(a, b) ef_assert((a) == (b))
++#define RANGECHCK(v, WIDTH) \
++ ef_assert(((uint64_t)(v) & ~(__EFVI_MASK64((WIDTH)))) == 0)
++
++/* fields within the first dword */
++#define DWCHCK(LBN, WIDTH) ef_assert(((LBN) >= 0) &&(((LBN)+(WIDTH)) <= 32))
++
++/* fields which span the first and second dwords */
++#define LWCHK(LBN, WIDTH) ef_assert(WIDTH >= __LW(LBN))
++
++/*----------------------------------------------------------------------------
++ *
++ * Buffer virtual addresses (4K buffers)
++ *
++ *---------------------------------------------------------------------------*/
++
++/* Form a buffer virtual address from buffer ID and offset. If the offset
++** is larger than the buffer size, then the buffer indexed will be
++** calculated appropriately. It is the responsibility of the caller to
++** ensure that they have valid buffers programmed at that address.
++*/
++#define EFVI_FALCON_VADDR_4K_S (12)
++#define EFVI_FALCON_VADDR_M 0xfffff /* post shift mask */
++
++
++#define EFVI_FALCON_BUFFER_4K_ADDR(id,off) \
++ (((id) << EFVI_FALCON_VADDR_4K_S) + (off))
++
++#define EFVI_FALCON_BUFFER_4K_PAGE(vaddr) \
++ (((vaddr) >> EFVI_FALCON_VADDR_4K_S) & EFVI_FALCON_VADDR_M)
++
++#define EFVI_FALCON_BUFFER_4K_OFF(vaddr) \
++ ((vaddr) & __EFVI_MASK32(EFVI_FALCON_VADDR_4K_S))
++
++
++/*----------------------------------------------------------------------------
++ *
++ * Masks
++ *
++ *---------------------------------------------------------------------------*/
++
++#define EFVI_FALCON_CLOCK_ASIC_HZ (125000)
++#define EFVI_FALCON_CLOCK_FPGA_HZ (62500)
++#define EFVI_FALCON_CLOCK_HZ EFVI_FALCON_CLOCK_ASIC_HZ
++
++
++/*----------------------------------------------------------------------------
++ *
++ * Timers
++ *
++ *---------------------------------------------------------------------------*/
++
++/* Event-Queue Timer granularity - measured in us
++ Given by: 4096 * 3 cycle * clock period */
++
++#define EFVI_FALCON_EVQTIMER_PERIOD_US ((4096 * 3 * 1000) / EFVI_FALCON_CLOCK_HZ)
++
++/* mode bits */
++#define EFVI_FALCON_TIMER_MODE_DIS 0 /* disabled */
++#define EFVI_FALCON_TIMER_MODE_RUN 1 /* started counting right away */
++#define EFVI_FALCON_TIMER_MODE_HOLD 2 /* trigger mode (user queues) */
++
++#define EFVI_FALCON_EVQTIMER_HOLD (EFVI_FALCON_TIMER_MODE_HOLD << TIMER_MODE_LBN)
++#define EFVI_FALCON_EVQTIMER_RUN (EFVI_FALCON_TIMER_MODE_RUN << TIMER_MODE_LBN)
++#define EFVI_FALCON_EVQTIMER_DISABLE (EFVI_FALCON_TIMER_MODE_DIS << TIMER_MODE_LBN)
++
++
++/* ---- efhw_event_t helpers --- */
++
++#define EFVI_FALCON_EVENT_CODE(evp) \
++ ((evp)->u64 & EFVI_FALCON_EVENT_CODE_MASK)
++
++#define EFVI_FALCON_EVENT_SW_DATA_MASK 0x0000ffff
++
++#define __EFVI_FALCON_OPEN_MASK(WIDTH) ((((uint64_t)1) << (WIDTH)) - 1)
++
++#define EFVI_FALCON_EVENT_CODE_MASK \
++ (__EFVI_FALCON_OPEN_MASK(EV_CODE_WIDTH) << EV_CODE_LBN)
++
++
++#endif /* __EF_VI_FALCON_H__ */
+diff --git a/drivers/xen/sfc_netfront/ef_vi_falcon_core.h b/drivers/xen/sfc_netfront/ef_vi_falcon_core.h
+new file mode 100644
+index 0000000..089f42a
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/ef_vi_falcon_core.h
+@@ -0,0 +1,1075 @@
++
++#define EFVI_FALCON_EXTENDED_P_BAR 1
++
++//////////////---- Bus Interface Unit Registers C Header ----//////////////
++#define IOM_IND_ADR_REG_OFST 0x0 // IO-mapped indirect access address register
++ #define IOM_AUTO_ADR_INC_EN_LBN 16
++ #define IOM_AUTO_ADR_INC_EN_WIDTH 1
++ #define IOM_IND_ADR_LBN 0
++ #define IOM_IND_ADR_WIDTH 16
++#define IOM_IND_DAT_REG_OFST 0x4 // IO-mapped indirect access data register
++ #define IOM_IND_DAT_LBN 0
++ #define IOM_IND_DAT_WIDTH 32
++#define ADR_REGION_REG_KER_OFST 0x0 // Address region register
++#define ADR_REGION_REG_OFST 0x0 // Address region register
++ #define ADR_REGION3_LBN 96
++ #define ADR_REGION3_WIDTH 18
++ #define ADR_REGION2_LBN 64
++ #define ADR_REGION2_WIDTH 18
++ #define ADR_REGION1_LBN 32
++ #define ADR_REGION1_WIDTH 18
++ #define ADR_REGION0_LBN 0
++ #define ADR_REGION0_WIDTH 18
++#define INT_EN_REG_KER_OFST 0x10 // Kernel driver Interrupt enable register
++ #define KER_INT_CHAR_LBN 4
++ #define KER_INT_CHAR_WIDTH 1
++ #define KER_INT_KER_LBN 3
++ #define KER_INT_KER_WIDTH 1
++ #define ILL_ADR_ERR_INT_EN_KER_LBN 2
++ #define ILL_ADR_ERR_INT_EN_KER_WIDTH 1
++ #define SRM_PERR_INT_EN_KER_LBN 1
++ #define SRM_PERR_INT_EN_KER_WIDTH 1
++ #define DRV_INT_EN_KER_LBN 0
++ #define DRV_INT_EN_KER_WIDTH 1
++#define INT_EN_REG_CHAR_OFST 0x20 // Char Driver interrupt enable register
++ #define CHAR_INT_CHAR_LBN 4
++ #define CHAR_INT_CHAR_WIDTH 1
++ #define CHAR_INT_KER_LBN 3
++ #define CHAR_INT_KER_WIDTH 1
++ #define ILL_ADR_ERR_INT_EN_CHAR_LBN 2
++ #define ILL_ADR_ERR_INT_EN_CHAR_WIDTH 1
++ #define SRM_PERR_INT_EN_CHAR_LBN 1
++ #define SRM_PERR_INT_EN_CHAR_WIDTH 1
++ #define DRV_INT_EN_CHAR_LBN 0
++ #define DRV_INT_EN_CHAR_WIDTH 1
++#define INT_ADR_REG_KER_OFST 0x30 // Interrupt host address for Kernel driver
++ #define INT_ADR_KER_LBN 0
++ #define INT_ADR_KER_WIDTH 64
++ #define DRV_INT_KER_LBN 32
++ #define DRV_INT_KER_WIDTH 1
++ #define EV_FF_HALF_INT_KER_LBN 3
++ #define EV_FF_HALF_INT_KER_WIDTH 1
++ #define EV_FF_FULL_INT_KER_LBN 2
++ #define EV_FF_FULL_INT_KER_WIDTH 1
++ #define ILL_ADR_ERR_INT_KER_LBN 1
++ #define ILL_ADR_ERR_INT_KER_WIDTH 1
++ #define SRAM_PERR_INT_KER_LBN 0
++ #define SRAM_PERR_INT_KER_WIDTH 1
++#define INT_ADR_REG_CHAR_OFST 0x40 // Interrupt host address for Char driver
++ #define INT_ADR_CHAR_LBN 0
++ #define INT_ADR_CHAR_WIDTH 64
++ #define DRV_INT_CHAR_LBN 32
++ #define DRV_INT_CHAR_WIDTH 1
++ #define EV_FF_HALF_INT_CHAR_LBN 3
++ #define EV_FF_HALF_INT_CHAR_WIDTH 1
++ #define EV_FF_FULL_INT_CHAR_LBN 2
++ #define EV_FF_FULL_INT_CHAR_WIDTH 1
++ #define ILL_ADR_ERR_INT_CHAR_LBN 1
++ #define ILL_ADR_ERR_INT_CHAR_WIDTH 1
++ #define SRAM_PERR_INT_CHAR_LBN 0
++ #define SRAM_PERR_INT_CHAR_WIDTH 1
++#define INT_ISR0_B0_OFST 0x90 // B0 only
++#define INT_ISR1_B0_OFST 0xA0
++#define INT_ACK_REG_KER_A1_OFST 0x50 // Kernel interrupt acknowledge register
++ #define RESERVED_LBN 0
++ #define RESERVED_WIDTH 32
++#define INT_ACK_REG_CHAR_A1_OFST 0x60 // CHAR interrupt acknowledge register
++ #define RESERVED_LBN 0
++ #define RESERVED_WIDTH 32
++//////////////---- Global CSR Registers C Header ----//////////////
++#define STRAP_REG_KER_OFST 0x200 // ASIC strap status register
++#define STRAP_REG_OFST 0x200 // ASIC strap status register
++ #define ONCHIP_SRAM_LBN 16
++ #define ONCHIP_SRAM_WIDTH 0
++ #define STRAP_ISCSI_EN_LBN 3
++ #define STRAP_ISCSI_EN_WIDTH 1
++ #define STRAP_PINS_LBN 0
++ #define STRAP_PINS_WIDTH 3
++#define GPIO_CTL_REG_KER_OFST 0x210 // GPIO control register
++#define GPIO_CTL_REG_OFST 0x210 // GPIO control register
++ #define GPIO_OEN_LBN 24
++ #define GPIO_OEN_WIDTH 4
++ #define GPIO_OUT_LBN 16
++ #define GPIO_OUT_WIDTH 4
++ #define GPIO_IN_LBN 8
++ #define GPIO_IN_WIDTH 4
++ #define GPIO_PWRUP_VALUE_LBN 0
++ #define GPIO_PWRUP_VALUE_WIDTH 4
++#define GLB_CTL_REG_KER_OFST 0x220 // Global control register
++#define GLB_CTL_REG_OFST 0x220 // Global control register
++ #define SWRST_LBN 0
++ #define SWRST_WIDTH 1
++#define FATAL_INTR_REG_KER_OFST 0x230 // Fatal interrupt register for Kernel
++ #define PCI_BUSERR_INT_KER_EN_LBN 43
++ #define PCI_BUSERR_INT_KER_EN_WIDTH 1
++ #define SRAM_OOB_INT_KER_EN_LBN 42
++ #define SRAM_OOB_INT_KER_EN_WIDTH 1
++ #define BUFID_OOB_INT_KER_EN_LBN 41
++ #define BUFID_OOB_INT_KER_EN_WIDTH 1
++ #define MEM_PERR_INT_KER_EN_LBN 40
++ #define MEM_PERR_INT_KER_EN_WIDTH 1
++ #define RBUF_OWN_INT_KER_EN_LBN 39
++ #define RBUF_OWN_INT_KER_EN_WIDTH 1
++ #define TBUF_OWN_INT_KER_EN_LBN 38
++ #define TBUF_OWN_INT_KER_EN_WIDTH 1
++ #define RDESCQ_OWN_INT_KER_EN_LBN 37
++ #define RDESCQ_OWN_INT_KER_EN_WIDTH 1
++ #define TDESCQ_OWN_INT_KER_EN_LBN 36
++ #define TDESCQ_OWN_INT_KER_EN_WIDTH 1
++ #define EVQ_OWN_INT_KER_EN_LBN 35
++ #define EVQ_OWN_INT_KER_EN_WIDTH 1
++ #define EVFF_OFLO_INT_KER_EN_LBN 34
++ #define EVFF_OFLO_INT_KER_EN_WIDTH 1
++ #define ILL_ADR_INT_KER_EN_LBN 33
++ #define ILL_ADR_INT_KER_EN_WIDTH 1
++ #define SRM_PERR_INT_KER_EN_LBN 32
++ #define SRM_PERR_INT_KER_EN_WIDTH 1
++ #define PCI_BUSERR_INT_KER_LBN 11
++ #define PCI_BUSERR_INT_KER_WIDTH 1
++ #define SRAM_OOB_INT_KER_LBN 10
++ #define SRAM_OOB_INT_KER_WIDTH 1
++ #define BUFID_OOB_INT_KER_LBN 9
++ #define BUFID_OOB_INT_KER_WIDTH 1
++ #define MEM_PERR_INT_KER_LBN 8
++ #define MEM_PERR_INT_KER_WIDTH 1
++ #define RBUF_OWN_INT_KER_LBN 7
++ #define RBUF_OWN_INT_KER_WIDTH 1
++ #define TBUF_OWN_INT_KER_LBN 6
++ #define TBUF_OWN_INT_KER_WIDTH 1
++ #define RDESCQ_OWN_INT_KER_LBN 5
++ #define RDESCQ_OWN_INT_KER_WIDTH 1
++ #define TDESCQ_OWN_INT_KER_LBN 4
++ #define TDESCQ_OWN_INT_KER_WIDTH 1
++ #define EVQ_OWN_INT_KER_LBN 3
++ #define EVQ_OWN_INT_KER_WIDTH 1
++ #define EVFF_OFLO_INT_KER_LBN 2
++ #define EVFF_OFLO_INT_KER_WIDTH 1
++ #define ILL_ADR_INT_KER_LBN 1
++ #define ILL_ADR_INT_KER_WIDTH 1
++ #define SRM_PERR_INT_KER_LBN 0
++ #define SRM_PERR_INT_KER_WIDTH 1
++#define FATAL_INTR_REG_OFST 0x240 // Fatal interrupt register for Char
++ #define PCI_BUSERR_INT_CHAR_EN_LBN 43
++ #define PCI_BUSERR_INT_CHAR_EN_WIDTH 1
++ #define SRAM_OOB_INT_CHAR_EN_LBN 42
++ #define SRAM_OOB_INT_CHAR_EN_WIDTH 1
++ #define BUFID_OOB_INT_CHAR_EN_LBN 41
++ #define BUFID_OOB_INT_CHAR_EN_WIDTH 1
++ #define MEM_PERR_INT_CHAR_EN_LBN 40
++ #define MEM_PERR_INT_CHAR_EN_WIDTH 1
++ #define RBUF_OWN_INT_CHAR_EN_LBN 39
++ #define RBUF_OWN_INT_CHAR_EN_WIDTH 1
++ #define TBUF_OWN_INT_CHAR_EN_LBN 38
++ #define TBUF_OWN_INT_CHAR_EN_WIDTH 1
++ #define RDESCQ_OWN_INT_CHAR_EN_LBN 37
++ #define RDESCQ_OWN_INT_CHAR_EN_WIDTH 1
++ #define TDESCQ_OWN_INT_CHAR_EN_LBN 36
++ #define TDESCQ_OWN_INT_CHAR_EN_WIDTH 1
++ #define EVQ_OWN_INT_CHAR_EN_LBN 35
++ #define EVQ_OWN_INT_CHAR_EN_WIDTH 1
++ #define EVFF_OFLO_INT_CHAR_EN_LBN 34
++ #define EVFF_OFLO_INT_CHAR_EN_WIDTH 1
++ #define ILL_ADR_INT_CHAR_EN_LBN 33
++ #define ILL_ADR_INT_CHAR_EN_WIDTH 1
++ #define SRM_PERR_INT_CHAR_EN_LBN 32
++ #define SRM_PERR_INT_CHAR_EN_WIDTH 1
++ #define FATAL_INTR_REG_EN_BITS 0xffffffffffffffffULL
++ #define PCI_BUSERR_INT_CHAR_LBN 11
++ #define PCI_BUSERR_INT_CHAR_WIDTH 1
++ #define SRAM_OOB_INT_CHAR_LBN 10
++ #define SRAM_OOB_INT_CHAR_WIDTH 1
++ #define BUFID_OOB_INT_CHAR_LBN 9
++ #define BUFID_OOB_INT_CHAR_WIDTH 1
++ #define MEM_PERR_INT_CHAR_LBN 8
++ #define MEM_PERR_INT_CHAR_WIDTH 1
++ #define RBUF_OWN_INT_CHAR_LBN 7
++ #define RBUF_OWN_INT_CHAR_WIDTH 1
++ #define TBUF_OWN_INT_CHAR_LBN 6
++ #define TBUF_OWN_INT_CHAR_WIDTH 1
++ #define RDESCQ_OWN_INT_CHAR_LBN 5
++ #define RDESCQ_OWN_INT_CHAR_WIDTH 1
++ #define TDESCQ_OWN_INT_CHAR_LBN 4
++ #define TDESCQ_OWN_INT_CHAR_WIDTH 1
++ #define EVQ_OWN_INT_CHAR_LBN 3
++ #define EVQ_OWN_INT_CHAR_WIDTH 1
++ #define EVFF_OFLO_INT_CHAR_LBN 2
++ #define EVFF_OFLO_INT_CHAR_WIDTH 1
++ #define ILL_ADR_INT_CHAR_LBN 1
++ #define ILL_ADR_INT_CHAR_WIDTH 1
++ #define SRM_PERR_INT_CHAR_LBN 0
++ #define SRM_PERR_INT_CHAR_WIDTH 1
++#define DP_CTRL_REG_OFST 0x250 // Datapath control register
++ #define FLS_EVQ_ID_LBN 0
++ #define FLS_EVQ_ID_WIDTH 12
++#define MEM_STAT_REG_KER_OFST 0x260 // Memory status register
++#define MEM_STAT_REG_OFST 0x260 // Memory status register
++ #define MEM_PERR_VEC_LBN 53
++ #define MEM_PERR_VEC_WIDTH 38
++ #define MBIST_CORR_LBN 38
++ #define MBIST_CORR_WIDTH 15
++ #define MBIST_ERR_LBN 0
++ #define MBIST_ERR_WIDTH 38
++#define DEBUG_REG_KER_OFST 0x270 // Debug register
++#define DEBUG_REG_OFST 0x270 // Debug register
++ #define DEBUG_BLK_SEL2_LBN 47
++ #define DEBUG_BLK_SEL2_WIDTH 3
++ #define DEBUG_BLK_SEL1_LBN 44
++ #define DEBUG_BLK_SEL1_WIDTH 3
++ #define DEBUG_BLK_SEL0_LBN 41
++ #define DEBUG_BLK_SEL0_WIDTH 3
++ #define MISC_DEBUG_ADDR_LBN 36
++ #define MISC_DEBUG_ADDR_WIDTH 5
++ #define SERDES_DEBUG_ADDR_LBN 31
++ #define SERDES_DEBUG_ADDR_WIDTH 5
++ #define EM_DEBUG_ADDR_LBN 26
++ #define EM_DEBUG_ADDR_WIDTH 5
++ #define SR_DEBUG_ADDR_LBN 21
++ #define SR_DEBUG_ADDR_WIDTH 5
++ #define EV_DEBUG_ADDR_LBN 16
++ #define EV_DEBUG_ADDR_WIDTH 5
++ #define RX_DEBUG_ADDR_LBN 11
++ #define RX_DEBUG_ADDR_WIDTH 5
++ #define TX_DEBUG_ADDR_LBN 6
++ #define TX_DEBUG_ADDR_WIDTH 5
++ #define BIU_DEBUG_ADDR_LBN 1
++ #define BIU_DEBUG_ADDR_WIDTH 5
++ #define DEBUG_EN_LBN 0
++ #define DEBUG_EN_WIDTH 1
++#define DRIVER_REG0_KER_OFST 0x280 // Driver scratch register 0
++#define DRIVER_REG0_OFST 0x280 // Driver scratch register 0
++ #define DRIVER_DW0_LBN 0
++ #define DRIVER_DW0_WIDTH 32
++#define DRIVER_REG1_KER_OFST 0x290 // Driver scratch register 1
++#define DRIVER_REG1_OFST 0x290 // Driver scratch register 1
++ #define DRIVER_DW1_LBN 0
++ #define DRIVER_DW1_WIDTH 32
++#define DRIVER_REG2_KER_OFST 0x2A0 // Driver scratch register 2
++#define DRIVER_REG2_OFST 0x2A0 // Driver scratch register 2
++ #define DRIVER_DW2_LBN 0
++ #define DRIVER_DW2_WIDTH 32
++#define DRIVER_REG3_KER_OFST 0x2B0 // Driver scratch register 3
++#define DRIVER_REG3_OFST 0x2B0 // Driver scratch register 3
++ #define DRIVER_DW3_LBN 0
++ #define DRIVER_DW3_WIDTH 32
++#define DRIVER_REG4_KER_OFST 0x2C0 // Driver scratch register 4
++#define DRIVER_REG4_OFST 0x2C0 // Driver scratch register 4
++ #define DRIVER_DW4_LBN 0
++ #define DRIVER_DW4_WIDTH 32
++#define DRIVER_REG5_KER_OFST 0x2D0 // Driver scratch register 5
++#define DRIVER_REG5_OFST 0x2D0 // Driver scratch register 5
++ #define DRIVER_DW5_LBN 0
++ #define DRIVER_DW5_WIDTH 32
++#define DRIVER_REG6_KER_OFST 0x2E0 // Driver scratch register 6
++#define DRIVER_REG6_OFST 0x2E0 // Driver scratch register 6
++ #define DRIVER_DW6_LBN 0
++ #define DRIVER_DW6_WIDTH 32
++#define DRIVER_REG7_KER_OFST 0x2F0 // Driver scratch register 7
++#define DRIVER_REG7_OFST 0x2F0 // Driver scratch register 7
++ #define DRIVER_DW7_LBN 0
++ #define DRIVER_DW7_WIDTH 32
++#define ALTERA_BUILD_REG_OFST 0x300 // Altera build register
++#define ALTERA_BUILD_REG_OFST 0x300 // Altera build register
++ #define ALTERA_BUILD_VER_LBN 0
++ #define ALTERA_BUILD_VER_WIDTH 32
++
++/* so called CSR spare register
++ - contains separate parity enable bits for the various internal memory blocks */
++#define MEM_PARITY_ERR_EN_REG_KER 0x310
++#define MEM_PARITY_ALL_BLOCKS_EN_LBN 64
++#define MEM_PARITY_ALL_BLOCKS_EN_WIDTH 38
++#define MEM_PARITY_TX_DATA_EN_LBN 72
++#define MEM_PARITY_TX_DATA_EN_WIDTH 2
++
++//////////////---- Event & Timer Module Registers C Header ----//////////////
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define EVQ_RPTR_REG_KER_OFST 0x11B00 // Event queue read pointer register
++#else
++#define EVQ_RPTR_REG_KER_OFST 0x1B00 // Event queue read pointer register
++#endif
++
++#define EVQ_RPTR_REG_OFST 0xFA0000 // Event queue read pointer register array.
++ #define EVQ_RPTR_LBN 0
++ #define EVQ_RPTR_WIDTH 15
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define EVQ_PTR_TBL_KER_OFST 0x11A00 // Event queue pointer table for kernel access
++#else
++#define EVQ_PTR_TBL_KER_OFST 0x1A00 // Event queue pointer table for kernel access
++#endif
++
++#define EVQ_PTR_TBL_CHAR_OFST 0xF60000 // Event queue pointer table for char direct access
++ #define EVQ_WKUP_OR_INT_EN_LBN 39
++ #define EVQ_WKUP_OR_INT_EN_WIDTH 1
++ #define EVQ_NXT_WPTR_LBN 24
++ #define EVQ_NXT_WPTR_WIDTH 15
++ #define EVQ_EN_LBN 23
++ #define EVQ_EN_WIDTH 1
++ #define EVQ_SIZE_LBN 20
++ #define EVQ_SIZE_WIDTH 3
++ #define EVQ_BUF_BASE_ID_LBN 0
++ #define EVQ_BUF_BASE_ID_WIDTH 20
++#define TIMER_CMD_REG_KER_OFST 0x420 // Timer table for kernel access. Page-mapped
++#define TIMER_CMD_REG_PAGE4_OFST 0x8420 // Timer table for user-level access. Page-mapped. For lowest 1K queues.
++#define TIMER_CMD_REG_PAGE123K_OFST 0x1000420 // Timer table for user-level access. Page-mapped. For upper 3K queues.
++#define TIMER_TBL_OFST 0xF70000 // Timer table for char driver direct access
++ #define TIMER_MODE_LBN 12
++ #define TIMER_MODE_WIDTH 2
++ #define TIMER_VAL_LBN 0
++ #define TIMER_VAL_WIDTH 12
++ #define TIMER_MODE_INT_HLDOFF 2
++ #define EVQ_BUF_SIZE_LBN 0
++ #define EVQ_BUF_SIZE_WIDTH 1
++#define DRV_EV_REG_KER_OFST 0x440 // Driver generated event register
++#define DRV_EV_REG_OFST 0x440 // Driver generated event register
++ #define DRV_EV_QID_LBN 64
++ #define DRV_EV_QID_WIDTH 12
++ #define DRV_EV_DATA_LBN 0
++ #define DRV_EV_DATA_WIDTH 64
++#define EVQ_CTL_REG_KER_OFST 0x450 // Event queue control register
++#define EVQ_CTL_REG_OFST 0x450 // Event queue control register
++ #define RX_EVQ_WAKEUP_MASK_B0_LBN 15
++ #define RX_EVQ_WAKEUP_MASK_B0_WIDTH 6
++ #define EVQ_OWNERR_CTL_LBN 14
++ #define EVQ_OWNERR_CTL_WIDTH 1
++ #define EVQ_FIFO_AF_TH_LBN 8
++ #define EVQ_FIFO_AF_TH_WIDTH 6
++ #define EVQ_FIFO_NOTAF_TH_LBN 0
++ #define EVQ_FIFO_NOTAF_TH_WIDTH 6
++//////////////---- SRAM Module Registers C Header ----//////////////
++#define BUF_TBL_CFG_REG_KER_OFST 0x600 // Buffer table configuration register
++#define BUF_TBL_CFG_REG_OFST 0x600 // Buffer table configuration register
++ #define BUF_TBL_MODE_LBN 3
++ #define BUF_TBL_MODE_WIDTH 1
++#define SRM_RX_DC_CFG_REG_KER_OFST 0x610 // SRAM receive descriptor cache configuration register
++#define SRM_RX_DC_CFG_REG_OFST 0x610 // SRAM receive descriptor cache configuration register
++ #define SRM_RX_DC_BASE_ADR_LBN 0
++ #define SRM_RX_DC_BASE_ADR_WIDTH 21
++#define SRM_TX_DC_CFG_REG_KER_OFST 0x620 // SRAM transmit descriptor cache configuration register
++#define SRM_TX_DC_CFG_REG_OFST 0x620 // SRAM transmit descriptor cache configuration register
++ #define SRM_TX_DC_BASE_ADR_LBN 0
++ #define SRM_TX_DC_BASE_ADR_WIDTH 21
++#define SRM_CFG_REG_KER_OFST 0x630 // SRAM configuration register
++#define SRM_CFG_REG_OFST 0x630 // SRAM configuration register
++ #define SRAM_OOB_ADR_INTEN_LBN 5
++ #define SRAM_OOB_ADR_INTEN_WIDTH 1
++ #define SRAM_OOB_BUF_INTEN_LBN 4
++ #define SRAM_OOB_BUF_INTEN_WIDTH 1
++ #define SRAM_BT_INIT_EN_LBN 3
++ #define SRAM_BT_INIT_EN_WIDTH 1
++ #define SRM_NUM_BANK_LBN 2
++ #define SRM_NUM_BANK_WIDTH 1
++ #define SRM_BANK_SIZE_LBN 0
++ #define SRM_BANK_SIZE_WIDTH 2
++#define BUF_TBL_UPD_REG_KER_OFST 0x650 // Buffer table update register
++#define BUF_TBL_UPD_REG_OFST 0x650 // Buffer table update register
++ #define BUF_UPD_CMD_LBN 63
++ #define BUF_UPD_CMD_WIDTH 1
++ #define BUF_CLR_CMD_LBN 62
++ #define BUF_CLR_CMD_WIDTH 1
++ #define BUF_CLR_END_ID_LBN 32
++ #define BUF_CLR_END_ID_WIDTH 20
++ #define BUF_CLR_START_ID_LBN 0
++ #define BUF_CLR_START_ID_WIDTH 20
++#define SRM_UPD_EVQ_REG_KER_OFST 0x660 // Buffer table update register
++#define SRM_UPD_EVQ_REG_OFST 0x660 // Buffer table update register
++ #define SRM_UPD_EVQ_ID_LBN 0
++ #define SRM_UPD_EVQ_ID_WIDTH 12
++#define SRAM_PARITY_REG_KER_OFST 0x670 // SRAM parity register.
++#define SRAM_PARITY_REG_OFST 0x670 // SRAM parity register.
++ #define FORCE_SRAM_PERR_LBN 0
++ #define FORCE_SRAM_PERR_WIDTH 1
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define BUF_HALF_TBL_KER_OFST 0x18000 // Buffer table in half buffer table mode direct access by kernel driver
++#else
++#define BUF_HALF_TBL_KER_OFST 0x8000 // Buffer table in half buffer table mode direct access by kernel driver
++#endif
++
++
++#define BUF_HALF_TBL_OFST 0x800000 // Buffer table in half buffer table mode direct access by char driver
++ #define BUF_ADR_HBUF_ODD_LBN 44
++ #define BUF_ADR_HBUF_ODD_WIDTH 20
++ #define BUF_OWNER_ID_HBUF_ODD_LBN 32
++ #define BUF_OWNER_ID_HBUF_ODD_WIDTH 12
++ #define BUF_ADR_HBUF_EVEN_LBN 12
++ #define BUF_ADR_HBUF_EVEN_WIDTH 20
++ #define BUF_OWNER_ID_HBUF_EVEN_LBN 0
++ #define BUF_OWNER_ID_HBUF_EVEN_WIDTH 12
++
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define BUF_FULL_TBL_KER_OFST 0x18000 // Buffer table in full buffer table mode direct access by kernel driver
++#else
++#define BUF_FULL_TBL_KER_OFST 0x8000 // Buffer table in full buffer table mode direct access by kernel driver
++#endif
++
++
++
++
++#define BUF_FULL_TBL_OFST 0x800000 // Buffer table in full buffer table mode direct access by char driver
++ #define IP_DAT_BUF_SIZE_LBN 50
++ #define IP_DAT_BUF_SIZE_WIDTH 1
++ #define BUF_ADR_REGION_LBN 48
++ #define BUF_ADR_REGION_WIDTH 2
++ #define BUF_ADR_FBUF_LBN 14
++ #define BUF_ADR_FBUF_WIDTH 34
++ #define BUF_OWNER_ID_FBUF_LBN 0
++ #define BUF_OWNER_ID_FBUF_WIDTH 14
++#define SRM_DBG_REG_OFST 0x3000000 // SRAM debug access
++ #define SRM_DBG_LBN 0
++ #define SRM_DBG_WIDTH 64
++//////////////---- RX Datapath Registers C Header ----//////////////
++
++#define RX_CFG_REG_KER_OFST 0x800 // Receive configuration register
++#define RX_CFG_REG_OFST 0x800 // Receive configuration register
++
++#if !defined(FALCON_64K_RXFIFO) && !defined(FALCON_PRE_02020029)
++# if !defined(FALCON_128K_RXFIFO)
++# define FALCON_128K_RXFIFO
++# endif
++#endif
++
++#if defined(FALCON_128K_RXFIFO)
++
++/* new for B0 */
++ #define RX_TOEP_TCP_SUPPRESS_B0_LBN 48
++ #define RX_TOEP_TCP_SUPPRESS_B0_WIDTH 1
++ #define RX_INGR_EN_B0_LBN 47
++ #define RX_INGR_EN_B0_WIDTH 1
++ #define RX_TOEP_IPV4_B0_LBN 46
++ #define RX_TOEP_IPV4_B0_WIDTH 1
++ #define RX_HASH_ALG_B0_LBN 45
++ #define RX_HASH_ALG_B0_WIDTH 1
++ #define RX_HASH_INSERT_HDR_B0_LBN 44
++ #define RX_HASH_INSERT_HDR_B0_WIDTH 1
++/* moved for B0 */
++ #define RX_DESC_PUSH_EN_B0_LBN 43
++ #define RX_DESC_PUSH_EN_B0_WIDTH 1
++ #define RX_RDW_PATCH_EN_LBN 42 /* Non head of line blocking */
++ #define RX_RDW_PATCH_EN_WIDTH 1
++ #define RX_PCI_BURST_SIZE_B0_LBN 39
++ #define RX_PCI_BURST_SIZE_B0_WIDTH 3
++ #define RX_OWNERR_CTL_B0_LBN 38
++ #define RX_OWNERR_CTL_B0_WIDTH 1
++ #define RX_XON_TX_TH_B0_LBN 33
++ #define RX_XON_TX_TH_B0_WIDTH 5
++ #define RX_XOFF_TX_TH_B0_LBN 28
++ #define RX_XOFF_TX_TH_B0_WIDTH 5
++ #define RX_USR_BUF_SIZE_B0_LBN 19
++ #define RX_USR_BUF_SIZE_B0_WIDTH 9
++ #define RX_XON_MAC_TH_B0_LBN 10
++ #define RX_XON_MAC_TH_B0_WIDTH 9
++ #define RX_XOFF_MAC_TH_B0_LBN 1
++ #define RX_XOFF_MAC_TH_B0_WIDTH 9
++ #define RX_XOFF_MAC_EN_B0_LBN 0
++ #define RX_XOFF_MAC_EN_B0_WIDTH 1
++
++#elif !defined(FALCON_PRE_02020029)
++/* new for B0 */
++ #define RX_TOEP_TCP_SUPPRESS_B0_LBN 46
++ #define RX_TOEP_TCP_SUPPRESS_B0_WIDTH 1
++ #define RX_INGR_EN_B0_LBN 45
++ #define RX_INGR_EN_B0_WIDTH 1
++ #define RX_TOEP_IPV4_B0_LBN 44
++ #define RX_TOEP_IPV4_B0_WIDTH 1
++ #define RX_HASH_ALG_B0_LBN 43
++ #define RX_HASH_ALG_B0_WIDTH 41
++ #define RX_HASH_INSERT_HDR_B0_LBN 42
++ #define RX_HASH_INSERT_HDR_B0_WIDTH 1
++/* moved for B0 */
++ #define RX_DESC_PUSH_EN_B0_LBN 41
++ #define RX_DESC_PUSH_EN_B0_WIDTH 1
++ #define RX_PCI_BURST_SIZE_B0_LBN 37
++ #define RX_PCI_BURST_SIZE_B0_WIDTH 3
++ #define RX_OWNERR_CTL_B0_LBN 36
++ #define RX_OWNERR_CTL_B0_WIDTH 1
++ #define RX_XON_TX_TH_B0_LBN 31
++ #define RX_XON_TX_TH_B0_WIDTH 5
++ #define RX_XOFF_TX_TH_B0_LBN 26
++ #define RX_XOFF_TX_TH_B0_WIDTH 5
++ #define RX_USR_BUF_SIZE_B0_LBN 17
++ #define RX_USR_BUF_SIZE_B0_WIDTH 9
++ #define RX_XON_MAC_TH_B0_LBN 9
++ #define RX_XON_MAC_TH_B0_WIDTH 8
++ #define RX_XOFF_MAC_TH_B0_LBN 1
++ #define RX_XOFF_MAC_TH_B0_WIDTH 8
++ #define RX_XOFF_MAC_EN_B0_LBN 0
++ #define RX_XOFF_MAC_EN_B0_WIDTH 1
++
++#else
++/* new for B0 */
++ #define RX_TOEP_TCP_SUPPRESS_B0_LBN 44
++ #define RX_TOEP_TCP_SUPPRESS_B0_WIDTH 1
++ #define RX_INGR_EN_B0_LBN 43
++ #define RX_INGR_EN_B0_WIDTH 1
++ #define RX_TOEP_IPV4_B0_LBN 42
++ #define RX_TOEP_IPV4_B0_WIDTH 1
++ #define RX_HASH_ALG_B0_LBN 41
++ #define RX_HASH_ALG_B0_WIDTH 41
++ #define RX_HASH_INSERT_HDR_B0_LBN 40
++ #define RX_HASH_INSERT_HDR_B0_WIDTH 1
++/* moved for B0 */
++ #define RX_DESC_PUSH_EN_B0_LBN 35
++ #define RX_DESC_PUSH_EN_B0_WIDTH 1
++ #define RX_PCI_BURST_SIZE_B0_LBN 35
++ #define RX_PCI_BURST_SIZE_B0_WIDTH 2
++ #define RX_OWNERR_CTL_B0_LBN 34
++ #define RX_OWNERR_CTL_B0_WIDTH 1
++ #define RX_XON_TX_TH_B0_LBN 29
++ #define RX_XON_TX_TH_B0_WIDTH 5
++ #define RX_XOFF_TX_TH_B0_LBN 24
++ #define RX_XOFF_TX_TH_B0_WIDTH 5
++ #define RX_USR_BUF_SIZE_B0_LBN 15
++ #define RX_USR_BUF_SIZE_B0_WIDTH 9
++ #define RX_XON_MAC_TH_B0_LBN 8
++ #define RX_XON_MAC_TH_B0_WIDTH 7
++ #define RX_XOFF_MAC_TH_B0_LBN 1
++ #define RX_XOFF_MAC_TH_B0_WIDTH 7
++ #define RX_XOFF_MAC_EN_B0_LBN 0
++ #define RX_XOFF_MAC_EN_B0_WIDTH 1
++
++#endif
++
++/* A0/A1 */
++ #define RX_PUSH_EN_A1_LBN 35
++ #define RX_PUSH_EN_A1_WIDTH 1
++ #define RX_PCI_BURST_SIZE_A1_LBN 31
++ #define RX_PCI_BURST_SIZE_A1_WIDTH 3
++ #define RX_OWNERR_CTL_A1_LBN 30
++ #define RX_OWNERR_CTL_A1_WIDTH 1
++ #define RX_XON_TX_TH_A1_LBN 25
++ #define RX_XON_TX_TH_A1_WIDTH 5
++ #define RX_XOFF_TX_TH_A1_LBN 20
++ #define RX_XOFF_TX_TH_A1_WIDTH 5
++ #define RX_USR_BUF_SIZE_A1_LBN 11
++ #define RX_USR_BUF_SIZE_A1_WIDTH 9
++ #define RX_XON_MAC_TH_A1_LBN 6
++ #define RX_XON_MAC_TH_A1_WIDTH 5
++ #define RX_XOFF_MAC_TH_A1_LBN 1
++ #define RX_XOFF_MAC_TH_A1_WIDTH 5
++ #define RX_XOFF_MAC_EN_A1_LBN 0
++ #define RX_XOFF_MAC_EN_A1_WIDTH 1
++
++#define RX_FILTER_CTL_REG_OFST 0x810 // Receive filter control registers
++ #define SCATTER_ENBL_NO_MATCH_Q_B0_LBN 40
++ #define SCATTER_ENBL_NO_MATCH_Q_B0_WIDTH 1
++ #define UDP_FULL_SRCH_LIMIT_LBN 32
++ #define UDP_FULL_SRCH_LIMIT_WIDTH 8
++ #define NUM_KER_LBN 24
++ #define NUM_KER_WIDTH 2
++ #define UDP_WILD_SRCH_LIMIT_LBN 16
++ #define UDP_WILD_SRCH_LIMIT_WIDTH 8
++ #define TCP_WILD_SRCH_LIMIT_LBN 8
++ #define TCP_WILD_SRCH_LIMIT_WIDTH 8
++ #define TCP_FULL_SRCH_LIMIT_LBN 0
++ #define TCP_FULL_SRCH_LIMIT_WIDTH 8
++#define RX_FLUSH_DESCQ_REG_KER_OFST 0x820 // Receive flush descriptor queue register
++#define RX_FLUSH_DESCQ_REG_OFST 0x820 // Receive flush descriptor queue register
++ #define RX_FLUSH_DESCQ_CMD_LBN 24
++ #define RX_FLUSH_DESCQ_CMD_WIDTH 1
++ #define RX_FLUSH_EVQ_ID_LBN 12
++ #define RX_FLUSH_EVQ_ID_WIDTH 12
++ #define RX_FLUSH_DESCQ_LBN 0
++ #define RX_FLUSH_DESCQ_WIDTH 12
++#define RX_DESC_UPD_REG_KER_OFST 0x830 // Kernel receive descriptor update register. Page-mapped
++#define RX_DESC_UPD_REG_PAGE4_OFST 0x8830 // Char & user receive descriptor update register. Page-mapped. For lowest 1K queues.
++#define RX_DESC_UPD_REG_PAGE123K_OFST 0x1000830 // Char & user receive descriptor update register. Page-mapped. For upper 3K queues.
++ #define RX_DESC_WPTR_LBN 96
++ #define RX_DESC_WPTR_WIDTH 12
++ #define RX_DESC_PUSH_CMD_LBN 95
++ #define RX_DESC_PUSH_CMD_WIDTH 1
++ #define RX_DESC_LBN 0
++ #define RX_DESC_WIDTH 64
++ #define RX_KER_DESC_LBN 0
++ #define RX_KER_DESC_WIDTH 64
++ #define RX_USR_DESC_LBN 0
++ #define RX_USR_DESC_WIDTH 32
++#define RX_DC_CFG_REG_KER_OFST 0x840 // Receive descriptor cache configuration register
++#define RX_DC_CFG_REG_OFST 0x840 // Receive descriptor cache configuration register
++ #define RX_DC_SIZE_LBN 0
++ #define RX_DC_SIZE_WIDTH 2
++#define RX_DC_PF_WM_REG_KER_OFST 0x850 // Receive descriptor cache pre-fetch watermark register
++#define RX_DC_PF_WM_REG_OFST 0x850 // Receive descriptor cache pre-fetch watermark register
++ #define RX_DC_PF_LWM_LO_LBN 0
++ #define RX_DC_PF_LWM_LO_WIDTH 6
++
++#define RX_RSS_TKEY_B0_OFST 0x860 // RSS Toeplitz hash key (B0 only)
++
++#define RX_NODESC_DROP_REG 0x880
++ #define RX_NODESC_DROP_CNT_LBN 0
++ #define RX_NODESC_DROP_CNT_WIDTH 16
++
++#define XM_TX_CFG_REG_OFST 0x1230
++ #define XM_AUTO_PAD_LBN 5
++ #define XM_AUTO_PAD_WIDTH 1
++
++#define RX_FILTER_TBL0_OFST 0xF00000 // Receive filter table - even entries
++ #define RSS_EN_0_B0_LBN 110
++ #define RSS_EN_0_B0_WIDTH 1
++ #define SCATTER_EN_0_B0_LBN 109
++ #define SCATTER_EN_0_B0_WIDTH 1
++ #define TCP_UDP_0_LBN 108
++ #define TCP_UDP_0_WIDTH 1
++ #define RXQ_ID_0_LBN 96
++ #define RXQ_ID_0_WIDTH 12
++ #define DEST_IP_0_LBN 64
++ #define DEST_IP_0_WIDTH 32
++ #define DEST_PORT_TCP_0_LBN 48
++ #define DEST_PORT_TCP_0_WIDTH 16
++ #define SRC_IP_0_LBN 16
++ #define SRC_IP_0_WIDTH 32
++ #define SRC_TCP_DEST_UDP_0_LBN 0
++ #define SRC_TCP_DEST_UDP_0_WIDTH 16
++#define RX_FILTER_TBL1_OFST 0xF00010 // Receive filter table - odd entries
++ #define RSS_EN_1_B0_LBN 110
++ #define RSS_EN_1_B0_WIDTH 1
++ #define SCATTER_EN_1_B0_LBN 109
++ #define SCATTER_EN_1_B0_WIDTH 1
++ #define TCP_UDP_1_LBN 108
++ #define TCP_UDP_1_WIDTH 1
++ #define RXQ_ID_1_LBN 96
++ #define RXQ_ID_1_WIDTH 12
++ #define DEST_IP_1_LBN 64
++ #define DEST_IP_1_WIDTH 32
++ #define DEST_PORT_TCP_1_LBN 48
++ #define DEST_PORT_TCP_1_WIDTH 16
++ #define SRC_IP_1_LBN 16
++ #define SRC_IP_1_WIDTH 32
++ #define SRC_TCP_DEST_UDP_1_LBN 0
++ #define SRC_TCP_DEST_UDP_1_WIDTH 16
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define RX_DESC_PTR_TBL_KER_OFST 0x11800 // Receive descriptor pointer kernel access
++#else
++#define RX_DESC_PTR_TBL_KER_OFST 0x1800 // Receive descriptor pointer kernel access
++#endif
++
++
++#define RX_DESC_PTR_TBL_OFST 0xF40000 // Receive descriptor pointer table
++ #define RX_ISCSI_DDIG_EN_LBN 88
++ #define RX_ISCSI_DDIG_EN_WIDTH 1
++ #define RX_ISCSI_HDIG_EN_LBN 87
++ #define RX_ISCSI_HDIG_EN_WIDTH 1
++ #define RX_DESC_PREF_ACT_LBN 86
++ #define RX_DESC_PREF_ACT_WIDTH 1
++ #define RX_DC_HW_RPTR_LBN 80
++ #define RX_DC_HW_RPTR_WIDTH 6
++ #define RX_DESCQ_HW_RPTR_LBN 68
++ #define RX_DESCQ_HW_RPTR_WIDTH 12
++ #define RX_DESCQ_SW_WPTR_LBN 56
++ #define RX_DESCQ_SW_WPTR_WIDTH 12
++ #define RX_DESCQ_BUF_BASE_ID_LBN 36
++ #define RX_DESCQ_BUF_BASE_ID_WIDTH 20
++ #define RX_DESCQ_EVQ_ID_LBN 24
++ #define RX_DESCQ_EVQ_ID_WIDTH 12
++ #define RX_DESCQ_OWNER_ID_LBN 10
++ #define RX_DESCQ_OWNER_ID_WIDTH 14
++ #define RX_DESCQ_LABEL_LBN 5
++ #define RX_DESCQ_LABEL_WIDTH 5
++ #define RX_DESCQ_SIZE_LBN 3
++ #define RX_DESCQ_SIZE_WIDTH 2
++ #define RX_DESCQ_TYPE_LBN 2
++ #define RX_DESCQ_TYPE_WIDTH 1
++ #define RX_DESCQ_JUMBO_LBN 1
++ #define RX_DESCQ_JUMBO_WIDTH 1
++ #define RX_DESCQ_EN_LBN 0
++ #define RX_DESCQ_EN_WIDTH 1
++
++
++#define RX_RSS_INDIR_TBL_B0_OFST 0xFB0000 // RSS indirection table (B0 only)
++ #define RX_RSS_INDIR_ENT_B0_LBN 0
++ #define RX_RSS_INDIR_ENT_B0_WIDTH 6
++
++//////////////---- TX Datapath Registers C Header ----//////////////
++#define TX_FLUSH_DESCQ_REG_KER_OFST 0xA00 // Transmit flush descriptor queue register
++#define TX_FLUSH_DESCQ_REG_OFST 0xA00 // Transmit flush descriptor queue register
++ #define TX_FLUSH_DESCQ_CMD_LBN 12
++ #define TX_FLUSH_DESCQ_CMD_WIDTH 1
++ #define TX_FLUSH_DESCQ_LBN 0
++ #define TX_FLUSH_DESCQ_WIDTH 12
++#define TX_DESC_UPD_REG_KER_OFST 0xA10 // Kernel transmit descriptor update register. Page-mapped
++#define TX_DESC_UPD_REG_PAGE4_OFST 0x8A10 // Char & user transmit descriptor update register. Page-mapped
++#define TX_DESC_UPD_REG_PAGE123K_OFST 0x1000A10 // Char & user transmit descriptor update register. Page-mapped
++ #define TX_DESC_WPTR_LBN 96
++ #define TX_DESC_WPTR_WIDTH 12
++ #define TX_DESC_PUSH_CMD_LBN 95
++ #define TX_DESC_PUSH_CMD_WIDTH 1
++ #define TX_DESC_LBN 0
++ #define TX_DESC_WIDTH 95
++ #define TX_KER_DESC_LBN 0
++ #define TX_KER_DESC_WIDTH 64
++ #define TX_USR_DESC_LBN 0
++ #define TX_USR_DESC_WIDTH 64
++#define TX_DC_CFG_REG_KER_OFST 0xA20 // Transmit descriptor cache configuration register
++#define TX_DC_CFG_REG_OFST 0xA20 // Transmit descriptor cache configuration register
++ #define TX_DC_SIZE_LBN 0
++ #define TX_DC_SIZE_WIDTH 2
++
++#if EFVI_FALCON_EXTENDED_P_BAR
++#define TX_DESC_PTR_TBL_KER_OFST 0x11900 // Transmit descriptor pointer.
++#else
++#define TX_DESC_PTR_TBL_KER_OFST 0x1900 // Transmit descriptor pointer.
++#endif
++
++
++#define TX_DESC_PTR_TBL_OFST 0xF50000 // Transmit descriptor pointer
++ #define TX_NON_IP_DROP_DIS_B0_LBN 91
++ #define TX_NON_IP_DROP_DIS_B0_WIDTH 1
++ #define TX_IP_CHKSM_DIS_B0_LBN 90
++ #define TX_IP_CHKSM_DIS_B0_WIDTH 1
++ #define TX_TCP_CHKSM_DIS_B0_LBN 89
++ #define TX_TCP_CHKSM_DIS_B0_WIDTH 1
++ #define TX_DESCQ_EN_LBN 88
++ #define TX_DESCQ_EN_WIDTH 1
++ #define TX_ISCSI_DDIG_EN_LBN 87
++ #define TX_ISCSI_DDIG_EN_WIDTH 1
++ #define TX_ISCSI_HDIG_EN_LBN 86
++ #define TX_ISCSI_HDIG_EN_WIDTH 1
++ #define TX_DC_HW_RPTR_LBN 80
++ #define TX_DC_HW_RPTR_WIDTH 6
++ #define TX_DESCQ_HW_RPTR_LBN 68
++ #define TX_DESCQ_HW_RPTR_WIDTH 12
++ #define TX_DESCQ_SW_WPTR_LBN 56
++ #define TX_DESCQ_SW_WPTR_WIDTH 12
++ #define TX_DESCQ_BUF_BASE_ID_LBN 36
++ #define TX_DESCQ_BUF_BASE_ID_WIDTH 20
++ #define TX_DESCQ_EVQ_ID_LBN 24
++ #define TX_DESCQ_EVQ_ID_WIDTH 12
++ #define TX_DESCQ_OWNER_ID_LBN 10
++ #define TX_DESCQ_OWNER_ID_WIDTH 14
++ #define TX_DESCQ_LABEL_LBN 5
++ #define TX_DESCQ_LABEL_WIDTH 5
++ #define TX_DESCQ_SIZE_LBN 3
++ #define TX_DESCQ_SIZE_WIDTH 2
++ #define TX_DESCQ_TYPE_LBN 1
++ #define TX_DESCQ_TYPE_WIDTH 2
++ #define TX_DESCQ_FLUSH_LBN 0
++ #define TX_DESCQ_FLUSH_WIDTH 1
++#define TX_CFG_REG_KER_OFST 0xA50 // Transmit configuration register
++#define TX_CFG_REG_OFST 0xA50 // Transmit configuration register
++ #define TX_IP_ID_P1_OFS_LBN 32
++ #define TX_IP_ID_P1_OFS_WIDTH 15
++ #define TX_IP_ID_P0_OFS_LBN 16
++ #define TX_IP_ID_P0_OFS_WIDTH 15
++ #define TX_TURBO_EN_LBN 3
++ #define TX_TURBO_EN_WIDTH 1
++ #define TX_OWNERR_CTL_LBN 2
++ #define TX_OWNERR_CTL_WIDTH 2
++ #define TX_NON_IP_DROP_DIS_LBN 1
++ #define TX_NON_IP_DROP_DIS_WIDTH 1
++ #define TX_IP_ID_REP_EN_LBN 0
++ #define TX_IP_ID_REP_EN_WIDTH 1
++#define TX_RESERVED_REG_KER_OFST 0xA80 // Transmit configuration register
++#define TX_RESERVED_REG_OFST 0xA80 // Transmit configuration register
++ #define TX_CSR_PUSH_EN_LBN 89
++ #define TX_CSR_PUSH_EN_WIDTH 1
++ #define TX_RX_SPACER_LBN 64
++ #define TX_RX_SPACER_WIDTH 8
++ #define TX_SW_EV_EN_LBN 59
++ #define TX_SW_EV_EN_WIDTH 1
++ #define TX_RX_SPACER_EN_LBN 57
++ #define TX_RX_SPACER_EN_WIDTH 1
++ #define TX_CSR_PREF_WD_TMR_LBN 24
++ #define TX_CSR_PREF_WD_TMR_WIDTH 16
++ #define TX_CSR_ONLY1TAG_LBN 21
++ #define TX_CSR_ONLY1TAG_WIDTH 1
++ #define TX_PREF_THRESHOLD_LBN 19
++ #define TX_PREF_THRESHOLD_WIDTH 2
++ #define TX_ONE_PKT_PER_Q_LBN 18
++ #define TX_ONE_PKT_PER_Q_WIDTH 1
++ #define TX_DIS_NON_IP_EV_LBN 17
++ #define TX_DIS_NON_IP_EV_WIDTH 1
++ #define TX_DMA_SPACER_LBN 8
++ #define TX_DMA_SPACER_WIDTH 8
++ #define TX_FLUSH_MIN_LEN_EN_B0_LBN 7
++ #define TX_FLUSH_MIN_LEN_EN_B0_WIDTH 1
++ #define TX_TCP_DIS_A1_LBN 7
++ #define TX_TCP_DIS_A1_WIDTH 1
++ #define TX_IP_DIS_A1_LBN 6
++ #define TX_IP_DIS_A1_WIDTH 1
++ #define TX_MAX_CPL_LBN 2
++ #define TX_MAX_CPL_WIDTH 2
++ #define TX_MAX_PREF_LBN 0
++ #define TX_MAX_PREF_WIDTH 2
++#define TX_VLAN_REG_OFST 0xAE0 // Transmit VLAN tag register
++ #define TX_VLAN_EN_LBN 127
++ #define TX_VLAN_EN_WIDTH 1
++ #define TX_VLAN7_PORT1_EN_LBN 125
++ #define TX_VLAN7_PORT1_EN_WIDTH 1
++ #define TX_VLAN7_PORT0_EN_LBN 124
++ #define TX_VLAN7_PORT0_EN_WIDTH 1
++ #define TX_VLAN7_LBN 112
++ #define TX_VLAN7_WIDTH 12
++ #define TX_VLAN6_PORT1_EN_LBN 109
++ #define TX_VLAN6_PORT1_EN_WIDTH 1
++ #define TX_VLAN6_PORT0_EN_LBN 108
++ #define TX_VLAN6_PORT0_EN_WIDTH 1
++ #define TX_VLAN6_LBN 96
++ #define TX_VLAN6_WIDTH 12
++ #define TX_VLAN5_PORT1_EN_LBN 93
++ #define TX_VLAN5_PORT1_EN_WIDTH 1
++ #define TX_VLAN5_PORT0_EN_LBN 92
++ #define TX_VLAN5_PORT0_EN_WIDTH 1
++ #define TX_VLAN5_LBN 80
++ #define TX_VLAN5_WIDTH 12
++ #define TX_VLAN4_PORT1_EN_LBN 77
++ #define TX_VLAN4_PORT1_EN_WIDTH 1
++ #define TX_VLAN4_PORT0_EN_LBN 76
++ #define TX_VLAN4_PORT0_EN_WIDTH 1
++ #define TX_VLAN4_LBN 64
++ #define TX_VLAN4_WIDTH 12
++ #define TX_VLAN3_PORT1_EN_LBN 61
++ #define TX_VLAN3_PORT1_EN_WIDTH 1
++ #define TX_VLAN3_PORT0_EN_LBN 60
++ #define TX_VLAN3_PORT0_EN_WIDTH 1
++ #define TX_VLAN3_LBN 48
++ #define TX_VLAN3_WIDTH 12
++ #define TX_VLAN2_PORT1_EN_LBN 45
++ #define TX_VLAN2_PORT1_EN_WIDTH 1
++ #define TX_VLAN2_PORT0_EN_LBN 44
++ #define TX_VLAN2_PORT0_EN_WIDTH 1
++ #define TX_VLAN2_LBN 32
++ #define TX_VLAN2_WIDTH 12
++ #define TX_VLAN1_PORT1_EN_LBN 29
++ #define TX_VLAN1_PORT1_EN_WIDTH 1
++ #define TX_VLAN1_PORT0_EN_LBN 28
++ #define TX_VLAN1_PORT0_EN_WIDTH 1
++ #define TX_VLAN1_LBN 16
++ #define TX_VLAN1_WIDTH 12
++ #define TX_VLAN0_PORT1_EN_LBN 13
++ #define TX_VLAN0_PORT1_EN_WIDTH 1
++ #define TX_VLAN0_PORT0_EN_LBN 12
++ #define TX_VLAN0_PORT0_EN_WIDTH 1
++ #define TX_VLAN0_LBN 0
++ #define TX_VLAN0_WIDTH 12
++#define TX_FIL_CTL_REG_OFST 0xAF0 // Transmit filter control register
++ #define TX_MADR1_FIL_EN_LBN 65
++ #define TX_MADR1_FIL_EN_WIDTH 1
++ #define TX_MADR0_FIL_EN_LBN 64
++ #define TX_MADR0_FIL_EN_WIDTH 1
++ #define TX_IPFIL31_PORT1_EN_LBN 63
++ #define TX_IPFIL31_PORT1_EN_WIDTH 1
++ #define TX_IPFIL31_PORT0_EN_LBN 62
++ #define TX_IPFIL31_PORT0_EN_WIDTH 1
++ #define TX_IPFIL30_PORT1_EN_LBN 61
++ #define TX_IPFIL30_PORT1_EN_WIDTH 1
++ #define TX_IPFIL30_PORT0_EN_LBN 60
++ #define TX_IPFIL30_PORT0_EN_WIDTH 1
++ #define TX_IPFIL29_PORT1_EN_LBN 59
++ #define TX_IPFIL29_PORT1_EN_WIDTH 1
++ #define TX_IPFIL29_PORT0_EN_LBN 58
++ #define TX_IPFIL29_PORT0_EN_WIDTH 1
++ #define TX_IPFIL28_PORT1_EN_LBN 57
++ #define TX_IPFIL28_PORT1_EN_WIDTH 1
++ #define TX_IPFIL28_PORT0_EN_LBN 56
++ #define TX_IPFIL28_PORT0_EN_WIDTH 1
++ #define TX_IPFIL27_PORT1_EN_LBN 55
++ #define TX_IPFIL27_PORT1_EN_WIDTH 1
++ #define TX_IPFIL27_PORT0_EN_LBN 54
++ #define TX_IPFIL27_PORT0_EN_WIDTH 1
++ #define TX_IPFIL26_PORT1_EN_LBN 53
++ #define TX_IPFIL26_PORT1_EN_WIDTH 1
++ #define TX_IPFIL26_PORT0_EN_LBN 52
++ #define TX_IPFIL26_PORT0_EN_WIDTH 1
++ #define TX_IPFIL25_PORT1_EN_LBN 51
++ #define TX_IPFIL25_PORT1_EN_WIDTH 1
++ #define TX_IPFIL25_PORT0_EN_LBN 50
++ #define TX_IPFIL25_PORT0_EN_WIDTH 1
++ #define TX_IPFIL24_PORT1_EN_LBN 49
++ #define TX_IPFIL24_PORT1_EN_WIDTH 1
++ #define TX_IPFIL24_PORT0_EN_LBN 48
++ #define TX_IPFIL24_PORT0_EN_WIDTH 1
++ #define TX_IPFIL23_PORT1_EN_LBN 47
++ #define TX_IPFIL23_PORT1_EN_WIDTH 1
++ #define TX_IPFIL23_PORT0_EN_LBN 46
++ #define TX_IPFIL23_PORT0_EN_WIDTH 1
++ #define TX_IPFIL22_PORT1_EN_LBN 45
++ #define TX_IPFIL22_PORT1_EN_WIDTH 1
++ #define TX_IPFIL22_PORT0_EN_LBN 44
++ #define TX_IPFIL22_PORT0_EN_WIDTH 1
++ #define TX_IPFIL21_PORT1_EN_LBN 43
++ #define TX_IPFIL21_PORT1_EN_WIDTH 1
++ #define TX_IPFIL21_PORT0_EN_LBN 42
++ #define TX_IPFIL21_PORT0_EN_WIDTH 1
++ #define TX_IPFIL20_PORT1_EN_LBN 41
++ #define TX_IPFIL20_PORT1_EN_WIDTH 1
++ #define TX_IPFIL20_PORT0_EN_LBN 40
++ #define TX_IPFIL20_PORT0_EN_WIDTH 1
++ #define TX_IPFIL19_PORT1_EN_LBN 39
++ #define TX_IPFIL19_PORT1_EN_WIDTH 1
++ #define TX_IPFIL19_PORT0_EN_LBN 38
++ #define TX_IPFIL19_PORT0_EN_WIDTH 1
++ #define TX_IPFIL18_PORT1_EN_LBN 37
++ #define TX_IPFIL18_PORT1_EN_WIDTH 1
++ #define TX_IPFIL18_PORT0_EN_LBN 36
++ #define TX_IPFIL18_PORT0_EN_WIDTH 1
++ #define TX_IPFIL17_PORT1_EN_LBN 35
++ #define TX_IPFIL17_PORT1_EN_WIDTH 1
++ #define TX_IPFIL17_PORT0_EN_LBN 34
++ #define TX_IPFIL17_PORT0_EN_WIDTH 1
++ #define TX_IPFIL16_PORT1_EN_LBN 33
++ #define TX_IPFIL16_PORT1_EN_WIDTH 1
++ #define TX_IPFIL16_PORT0_EN_LBN 32
++ #define TX_IPFIL16_PORT0_EN_WIDTH 1
++ #define TX_IPFIL15_PORT1_EN_LBN 31
++ #define TX_IPFIL15_PORT1_EN_WIDTH 1
++ #define TX_IPFIL15_PORT0_EN_LBN 30
++ #define TX_IPFIL15_PORT0_EN_WIDTH 1
++ #define TX_IPFIL14_PORT1_EN_LBN 29
++ #define TX_IPFIL14_PORT1_EN_WIDTH 1
++ #define TX_IPFIL14_PORT0_EN_LBN 28
++ #define TX_IPFIL14_PORT0_EN_WIDTH 1
++ #define TX_IPFIL13_PORT1_EN_LBN 27
++ #define TX_IPFIL13_PORT1_EN_WIDTH 1
++ #define TX_IPFIL13_PORT0_EN_LBN 26
++ #define TX_IPFIL13_PORT0_EN_WIDTH 1
++ #define TX_IPFIL12_PORT1_EN_LBN 25
++ #define TX_IPFIL12_PORT1_EN_WIDTH 1
++ #define TX_IPFIL12_PORT0_EN_LBN 24
++ #define TX_IPFIL12_PORT0_EN_WIDTH 1
++ #define TX_IPFIL11_PORT1_EN_LBN 23
++ #define TX_IPFIL11_PORT1_EN_WIDTH 1
++ #define TX_IPFIL11_PORT0_EN_LBN 22
++ #define TX_IPFIL11_PORT0_EN_WIDTH 1
++ #define TX_IPFIL10_PORT1_EN_LBN 21
++ #define TX_IPFIL10_PORT1_EN_WIDTH 1
++ #define TX_IPFIL10_PORT0_EN_LBN 20
++ #define TX_IPFIL10_PORT0_EN_WIDTH 1
++ #define TX_IPFIL9_PORT1_EN_LBN 19
++ #define TX_IPFIL9_PORT1_EN_WIDTH 1
++ #define TX_IPFIL9_PORT0_EN_LBN 18
++ #define TX_IPFIL9_PORT0_EN_WIDTH 1
++ #define TX_IPFIL8_PORT1_EN_LBN 17
++ #define TX_IPFIL8_PORT1_EN_WIDTH 1
++ #define TX_IPFIL8_PORT0_EN_LBN 16
++ #define TX_IPFIL8_PORT0_EN_WIDTH 1
++ #define TX_IPFIL7_PORT1_EN_LBN 15
++ #define TX_IPFIL7_PORT1_EN_WIDTH 1
++ #define TX_IPFIL7_PORT0_EN_LBN 14
++ #define TX_IPFIL7_PORT0_EN_WIDTH 1
++ #define TX_IPFIL6_PORT1_EN_LBN 13
++ #define TX_IPFIL6_PORT1_EN_WIDTH 1
++ #define TX_IPFIL6_PORT0_EN_LBN 12
++ #define TX_IPFIL6_PORT0_EN_WIDTH 1
++ #define TX_IPFIL5_PORT1_EN_LBN 11
++ #define TX_IPFIL5_PORT1_EN_WIDTH 1
++ #define TX_IPFIL5_PORT0_EN_LBN 10
++ #define TX_IPFIL5_PORT0_EN_WIDTH 1
++ #define TX_IPFIL4_PORT1_EN_LBN 9
++ #define TX_IPFIL4_PORT1_EN_WIDTH 1
++ #define TX_IPFIL4_PORT0_EN_LBN 8
++ #define TX_IPFIL4_PORT0_EN_WIDTH 1
++ #define TX_IPFIL3_PORT1_EN_LBN 7
++ #define TX_IPFIL3_PORT1_EN_WIDTH 1
++ #define TX_IPFIL3_PORT0_EN_LBN 6
++ #define TX_IPFIL3_PORT0_EN_WIDTH 1
++ #define TX_IPFIL2_PORT1_EN_LBN 5
++ #define TX_IPFIL2_PORT1_EN_WIDTH 1
++ #define TX_IPFIL2_PORT0_EN_LBN 4
++ #define TX_IPFIL2_PORT0_EN_WIDTH 1
++ #define TX_IPFIL1_PORT1_EN_LBN 3
++ #define TX_IPFIL1_PORT1_EN_WIDTH 1
++ #define TX_IPFIL1_PORT0_EN_LBN 2
++ #define TX_IPFIL1_PORT0_EN_WIDTH 1
++ #define TX_IPFIL0_PORT1_EN_LBN 1
++ #define TX_IPFIL0_PORT1_EN_WIDTH 1
++ #define TX_IPFIL0_PORT0_EN_LBN 0
++ #define TX_IPFIL0_PORT0_EN_WIDTH 1
++#define TX_IPFIL_TBL_OFST 0xB00 // Transmit IP source address filter table
++ #define TX_IPFIL_MASK_LBN 32
++ #define TX_IPFIL_MASK_WIDTH 32
++ #define TX_IP_SRC_ADR_LBN 0
++ #define TX_IP_SRC_ADR_WIDTH 32
++#define TX_PACE_REG_A1_OFST 0xF80000 // Transmit pace control register
++#define TX_PACE_REG_B0_OFST 0xA90 // Transmit pace control register
++ #define TX_PACE_SB_AF_LBN 19
++ #define TX_PACE_SB_AF_WIDTH 10
++ #define TX_PACE_SB_NOTAF_LBN 9
++ #define TX_PACE_SB_NOTAF_WIDTH 10
++ #define TX_PACE_FB_BASE_LBN 5
++ #define TX_PACE_FB_BASE_WIDTH 4
++ #define TX_PACE_BIN_TH_LBN 0
++ #define TX_PACE_BIN_TH_WIDTH 5
++#define TX_PACE_TBL_A1_OFST 0xF80040 // Transmit pacing table
++#define TX_PACE_TBL_FIRST_QUEUE_A1 4
++#define TX_PACE_TBL_B0_OFST 0xF80000 // Transmit pacing table
++#define TX_PACE_TBL_FIRST_QUEUE_B0 0
++ #define TX_PACE_LBN 0
++ #define TX_PACE_WIDTH 5
++
++//////////////---- EE/Flash Registers C Header ----//////////////
++#define EE_SPI_HCMD_REG_KER_OFST 0x100 // SPI host command register
++#define EE_SPI_HCMD_REG_OFST 0x100 // SPI host command register
++ #define EE_SPI_HCMD_CMD_EN_LBN 31
++ #define EE_SPI_HCMD_CMD_EN_WIDTH 1
++ #define EE_WR_TIMER_ACTIVE_LBN 28
++ #define EE_WR_TIMER_ACTIVE_WIDTH 1
++ #define EE_SPI_HCMD_SF_SEL_LBN 24
++ #define EE_SPI_HCMD_SF_SEL_WIDTH 1
++ #define EE_SPI_HCMD_DABCNT_LBN 16
++ #define EE_SPI_HCMD_DABCNT_WIDTH 5
++ #define EE_SPI_HCMD_READ_LBN 15
++ #define EE_SPI_HCMD_READ_WIDTH 1
++ #define EE_SPI_HCMD_DUBCNT_LBN 12
++ #define EE_SPI_HCMD_DUBCNT_WIDTH 2
++ #define EE_SPI_HCMD_ADBCNT_LBN 8
++ #define EE_SPI_HCMD_ADBCNT_WIDTH 2
++ #define EE_SPI_HCMD_ENC_LBN 0
++ #define EE_SPI_HCMD_ENC_WIDTH 8
++#define EE_SPI_HADR_REG_KER_OFST 0X110 // SPI host address register
++#define EE_SPI_HADR_REG_OFST 0X110 // SPI host address register
++ #define EE_SPI_HADR_DUBYTE_LBN 24
++ #define EE_SPI_HADR_DUBYTE_WIDTH 8
++ #define EE_SPI_HADR_ADR_LBN 0
++ #define EE_SPI_HADR_ADR_WIDTH 24
++#define EE_SPI_HDATA_REG_KER_OFST 0x120 // SPI host data register
++#define EE_SPI_HDATA_REG_OFST 0x120 // SPI host data register
++ #define EE_SPI_HDATA3_LBN 96
++ #define EE_SPI_HDATA3_WIDTH 32
++ #define EE_SPI_HDATA2_LBN 64
++ #define EE_SPI_HDATA2_WIDTH 32
++ #define EE_SPI_HDATA1_LBN 32
++ #define EE_SPI_HDATA1_WIDTH 32
++ #define EE_SPI_HDATA0_LBN 0
++ #define EE_SPI_HDATA0_WIDTH 32
++#define EE_BASE_PAGE_REG_KER_OFST 0x130 // Expansion ROM base mirror register
++#define EE_BASE_PAGE_REG_OFST 0x130 // Expansion ROM base mirror register
++ #define EE_EXP_ROM_WINDOW_BASE_LBN 16
++ #define EE_EXP_ROM_WINDOW_BASE_WIDTH 13
++ #define EE_EXPROM_MASK_LBN 0
++ #define EE_EXPROM_MASK_WIDTH 13
++#define EE_VPD_CFG0_REG_KER_OFST 0X140 // SPI/VPD configuration register
++#define EE_VPD_CFG0_REG_OFST 0X140 // SPI/VPD configuration register
++ #define EE_SF_FASTRD_EN_LBN 127
++ #define EE_SF_FASTRD_EN_WIDTH 1
++ #define EE_SF_CLOCK_DIV_LBN 120
++ #define EE_SF_CLOCK_DIV_WIDTH 7
++ #define EE_VPD_WIP_POLL_LBN 119
++ #define EE_VPD_WIP_POLL_WIDTH 1
++ #define EE_VPDW_LENGTH_LBN 80
++ #define EE_VPDW_LENGTH_WIDTH 15
++ #define EE_VPDW_BASE_LBN 64
++ #define EE_VPDW_BASE_WIDTH 15
++ #define EE_VPD_WR_CMD_EN_LBN 56
++ #define EE_VPD_WR_CMD_EN_WIDTH 8
++ #define EE_VPD_BASE_LBN 32
++ #define EE_VPD_BASE_WIDTH 24
++ #define EE_VPD_LENGTH_LBN 16
++ #define EE_VPD_LENGTH_WIDTH 13
++ #define EE_VPD_AD_SIZE_LBN 8
++ #define EE_VPD_AD_SIZE_WIDTH 5
++ #define EE_VPD_ACCESS_ON_LBN 5
++ #define EE_VPD_ACCESS_ON_WIDTH 1
++#define EE_VPD_SW_CNTL_REG_KER_OFST 0X150 // VPD access SW control register
++#define EE_VPD_SW_CNTL_REG_OFST 0X150 // VPD access SW control register
++ #define EE_VPD_CYCLE_PENDING_LBN 31
++ #define EE_VPD_CYCLE_PENDING_WIDTH 1
++ #define EE_VPD_CYC_WRITE_LBN 28
++ #define EE_VPD_CYC_WRITE_WIDTH 1
++ #define EE_VPD_CYC_ADR_LBN 0
++ #define EE_VPD_CYC_ADR_WIDTH 15
++#define EE_VPD_SW_DATA_REG_KER_OFST 0x160 // VPD access SW data register
++#define EE_VPD_SW_DATA_REG_OFST 0x160 // VPD access SW data register
++ #define EE_VPD_CYC_DAT_LBN 0
++ #define EE_VPD_CYC_DAT_WIDTH 32
+diff --git a/drivers/xen/sfc_netfront/ef_vi_falcon_desc.h b/drivers/xen/sfc_netfront/ef_vi_falcon_desc.h
+new file mode 100644
+index 0000000..8c8404c
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/ef_vi_falcon_desc.h
+@@ -0,0 +1,43 @@
++//////////////---- Descriptors C Headers ----//////////////
++// Receive Kernel IP Descriptor
++ #define RX_KER_BUF_SIZE_LBN 48
++ #define RX_KER_BUF_SIZE_WIDTH 14
++ #define RX_KER_BUF_REGION_LBN 46
++ #define RX_KER_BUF_REGION_WIDTH 2
++ #define RX_KER_BUF_REGION0_DECODE 0
++ #define RX_KER_BUF_REGION1_DECODE 1
++ #define RX_KER_BUF_REGION2_DECODE 2
++ #define RX_KER_BUF_REGION3_DECODE 3
++ #define RX_KER_BUF_ADR_LBN 0
++ #define RX_KER_BUF_ADR_WIDTH 46
++// Receive User IP Descriptor
++ #define RX_USR_2BYTE_OFS_LBN 20
++ #define RX_USR_2BYTE_OFS_WIDTH 12
++ #define RX_USR_BUF_ID_LBN 0
++ #define RX_USR_BUF_ID_WIDTH 20
++// Transmit Kernel IP Descriptor
++ #define TX_KER_PORT_LBN 63
++ #define TX_KER_PORT_WIDTH 1
++ #define TX_KER_CONT_LBN 62
++ #define TX_KER_CONT_WIDTH 1
++ #define TX_KER_BYTE_CNT_LBN 48
++ #define TX_KER_BYTE_CNT_WIDTH 14
++ #define TX_KER_BUF_REGION_LBN 46
++ #define TX_KER_BUF_REGION_WIDTH 2
++ #define TX_KER_BUF_REGION0_DECODE 0
++ #define TX_KER_BUF_REGION1_DECODE 1
++ #define TX_KER_BUF_REGION2_DECODE 2
++ #define TX_KER_BUF_REGION3_DECODE 3
++ #define TX_KER_BUF_ADR_LBN 0
++ #define TX_KER_BUF_ADR_WIDTH 46
++// Transmit User IP Descriptor
++ #define TX_USR_PORT_LBN 47
++ #define TX_USR_PORT_WIDTH 1
++ #define TX_USR_CONT_LBN 46
++ #define TX_USR_CONT_WIDTH 1
++ #define TX_USR_BYTE_CNT_LBN 33
++ #define TX_USR_BYTE_CNT_WIDTH 13
++ #define TX_USR_BUF_ID_LBN 13
++ #define TX_USR_BUF_ID_WIDTH 20
++ #define TX_USR_BYTE_OFS_LBN 0
++ #define TX_USR_BYTE_OFS_WIDTH 13
+diff --git a/drivers/xen/sfc_netfront/ef_vi_falcon_event.h b/drivers/xen/sfc_netfront/ef_vi_falcon_event.h
+new file mode 100644
+index 0000000..abb63f3
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/ef_vi_falcon_event.h
+@@ -0,0 +1,123 @@
++//////////////---- Events Format C Header ----//////////////
++//////////////---- Event entry ----//////////////
++ #define EV_CODE_LBN 60
++ #define EV_CODE_WIDTH 4
++ #define RX_IP_EV_DECODE 0
++ #define TX_IP_EV_DECODE 2
++ #define DRIVER_EV_DECODE 5
++ #define GLOBAL_EV_DECODE 6
++ #define DRV_GEN_EV_DECODE 7
++ #define EV_DATA_LBN 0
++ #define EV_DATA_WIDTH 60
++//////////////---- Receive IP events for both Kernel & User event queues ----//////////////
++ #define RX_EV_PKT_OK_LBN 56
++ #define RX_EV_PKT_OK_WIDTH 1
++ #define RX_EV_BUF_OWNER_ID_ERR_LBN 54
++ #define RX_EV_BUF_OWNER_ID_ERR_WIDTH 1
++ #define RX_EV_IP_HDR_CHKSUM_ERR_LBN 52
++ #define RX_EV_IP_HDR_CHKSUM_ERR_WIDTH 1
++ #define RX_EV_TCP_UDP_CHKSUM_ERR_LBN 51
++ #define RX_EV_TCP_UDP_CHKSUM_ERR_WIDTH 1
++ #define RX_EV_ETH_CRC_ERR_LBN 50
++ #define RX_EV_ETH_CRC_ERR_WIDTH 1
++ #define RX_EV_FRM_TRUNC_LBN 49
++ #define RX_EV_FRM_TRUNC_WIDTH 1
++ #define RX_EV_DRIB_NIB_LBN 48
++ #define RX_EV_DRIB_NIB_WIDTH 1
++ #define RX_EV_TOBE_DISC_LBN 47
++ #define RX_EV_TOBE_DISC_WIDTH 1
++ #define RX_EV_PKT_TYPE_LBN 44
++ #define RX_EV_PKT_TYPE_WIDTH 3
++ #define RX_EV_PKT_TYPE_ETH_DECODE 0
++ #define RX_EV_PKT_TYPE_LLC_DECODE 1
++ #define RX_EV_PKT_TYPE_JUMBO_DECODE 2
++ #define RX_EV_PKT_TYPE_VLAN_DECODE 3
++ #define RX_EV_PKT_TYPE_VLAN_LLC_DECODE 4
++ #define RX_EV_PKT_TYPE_VLAN_JUMBO_DECODE 5
++ #define RX_EV_HDR_TYPE_LBN 42
++ #define RX_EV_HDR_TYPE_WIDTH 2
++ #define RX_EV_HDR_TYPE_TCP_IPV4_DECODE 0
++ #define RX_EV_HDR_TYPE_UDP_IPV4_DECODE 1
++ #define RX_EV_HDR_TYPE_OTHER_IP_DECODE 2
++ #define RX_EV_HDR_TYPE_NON_IP_DECODE 3
++ #define RX_EV_DESC_Q_EMPTY_LBN 41
++ #define RX_EV_DESC_Q_EMPTY_WIDTH 1
++ #define RX_EV_MCAST_HASH_MATCH_LBN 40
++ #define RX_EV_MCAST_HASH_MATCH_WIDTH 1
++ #define RX_EV_MCAST_PKT_LBN 39
++ #define RX_EV_MCAST_PKT_WIDTH 1
++ #define RX_EV_Q_LABEL_LBN 32
++ #define RX_EV_Q_LABEL_WIDTH 5
++ #define RX_JUMBO_CONT_LBN 31
++ #define RX_JUMBO_CONT_WIDTH 1
++ #define RX_SOP_LBN 15
++ #define RX_SOP_WIDTH 1
++ #define RX_PORT_LBN 30
++ #define RX_PORT_WIDTH 1
++ #define RX_EV_BYTE_CNT_LBN 16
++ #define RX_EV_BYTE_CNT_WIDTH 14
++ #define RX_iSCSI_PKT_OK_LBN 14
++ #define RX_iSCSI_PKT_OK_WIDTH 1
++ #define RX_ISCSI_DDIG_ERR_LBN 13
++ #define RX_ISCSI_DDIG_ERR_WIDTH 1
++ #define RX_ISCSI_HDIG_ERR_LBN 12
++ #define RX_ISCSI_HDIG_ERR_WIDTH 1
++ #define RX_EV_DESC_PTR_LBN 0
++ #define RX_EV_DESC_PTR_WIDTH 12
++//////////////---- Transmit IP events for both Kernel & User event queues ----//////////////
++ #define TX_EV_PKT_ERR_LBN 38
++ #define TX_EV_PKT_ERR_WIDTH 1
++ #define TX_EV_PKT_TOO_BIG_LBN 37
++ #define TX_EV_PKT_TOO_BIG_WIDTH 1
++ #define TX_EV_Q_LABEL_LBN 32
++ #define TX_EV_Q_LABEL_WIDTH 5
++ #define TX_EV_PORT_LBN 16
++ #define TX_EV_PORT_WIDTH 1
++ #define TX_EV_WQ_FF_FULL_LBN 15
++ #define TX_EV_WQ_FF_FULL_WIDTH 1
++ #define TX_EV_BUF_OWNER_ID_ERR_LBN 14
++ #define TX_EV_BUF_OWNER_ID_ERR_WIDTH 1
++ #define TX_EV_COMP_LBN 12
++ #define TX_EV_COMP_WIDTH 1
++ #define TX_EV_DESC_PTR_LBN 0
++ #define TX_EV_DESC_PTR_WIDTH 12
++//////////////---- Char or Kernel driver events ----//////////////
++ #define DRIVER_EV_SUB_CODE_LBN 56
++ #define DRIVER_EV_SUB_CODE_WIDTH 4
++ #define TX_DESCQ_FLS_DONE_EV_DECODE 0x0
++ #define RX_DESCQ_FLS_DONE_EV_DECODE 0x1
++ #define EVQ_INIT_DONE_EV_DECODE 0x2
++ #define EVQ_NOT_EN_EV_DECODE 0x3
++ #define RX_DESCQ_FLSFF_OVFL_EV_DECODE 0x4
++ #define SRM_UPD_DONE_EV_DECODE 0x5
++ #define WAKE_UP_EV_DECODE 0x6
++ #define TX_PKT_NON_TCP_UDP_DECODE 0x9
++ #define TIMER_EV_DECODE 0xA
++ #define RX_DSC_ERROR_EV_DECODE 0xE
++ #define DRIVER_EV_TX_DESCQ_ID_LBN 0
++ #define DRIVER_EV_TX_DESCQ_ID_WIDTH 12
++ #define DRIVER_EV_RX_DESCQ_ID_LBN 0
++ #define DRIVER_EV_RX_DESCQ_ID_WIDTH 12
++ #define DRIVER_EV_EVQ_ID_LBN 0
++ #define DRIVER_EV_EVQ_ID_WIDTH 12
++ #define DRIVER_TMR_ID_LBN 0
++ #define DRIVER_TMR_ID_WIDTH 12
++ #define DRIVER_EV_SRM_UPD_LBN 0
++ #define DRIVER_EV_SRM_UPD_WIDTH 2
++ #define SRM_CLR_EV_DECODE 0
++ #define SRM_UPD_EV_DECODE 1
++ #define SRM_ILLCLR_EV_DECODE 2
++//////////////---- Global events. Sent to both event queue 0 and 4. ----//////////////
++ #define XFP_PHY_INTR_LBN 10
++ #define XFP_PHY_INTR_WIDTH 1
++ #define XG_PHY_INTR_LBN 9
++ #define XG_PHY_INTR_WIDTH 1
++ #define G_PHY1_INTR_LBN 8
++ #define G_PHY1_INTR_WIDTH 1
++ #define G_PHY0_INTR_LBN 7
++ #define G_PHY0_INTR_WIDTH 1
++//////////////---- Driver generated events ----//////////////
++ #define DRV_GEN_EV_CODE_LBN 60
++ #define DRV_GEN_EV_CODE_WIDTH 4
++ #define DRV_GEN_EV_DATA_LBN 0
++ #define DRV_GEN_EV_DATA_WIDTH 60
+diff --git a/drivers/xen/sfc_netfront/ef_vi_internal.h b/drivers/xen/sfc_netfront/ef_vi_internal.h
+new file mode 100644
+index 0000000..396ae46
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/ef_vi_internal.h
+@@ -0,0 +1,256 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Really-and-truely-honestly internal stuff for libef.
++ * \date 2004/06/13
++ */
++
++/*! \cidoxg_include_ci_ul */
++#ifndef __CI_EF_VI_INTERNAL_H__
++#define __CI_EF_VI_INTERNAL_H__
++
++
++/* These flags share space with enum ef_vi_flags. */
++#define EF_VI_BUG5692_WORKAROUND 0x10000
++
++
++/* ***********************************************************************
++ * COMPILATION CONTROL FLAGS (see ef_vi.h for "workaround" controls)
++ */
++
++#define EF_VI_DO_MAGIC_CHECKS 1
++
++
++/**********************************************************************
++ * Headers
++ */
++
++#include <etherfabric/ef_vi.h>
++#include "sysdep.h"
++#include "ef_vi_falcon.h"
++
++
++/**********************************************************************
++ * Debugging.
++ */
++
++#ifndef NDEBUG
++
++# define _ef_assert(exp, file, line) BUG_ON(!(exp));
++
++# define _ef_assert2(exp, x, y, file, line) do { \
++ if (unlikely(!(exp))) \
++ BUG(); \
++ } while (0)
++
++#else
++
++# define _ef_assert(exp, file, line)
++# define _ef_assert2(e, x, y, file, line)
++
++#endif
++
++#define ef_assert(a) do{ _ef_assert((a),__FILE__,__LINE__); } while(0)
++#define ef_assert_equal(a,b) _ef_assert2((a)==(b),(a),(b),__FILE__,__LINE__)
++#define ef_assert_eq ef_assert_equal
++#define ef_assert_lt(a,b) _ef_assert2((a)<(b),(a),(b),__FILE__,__LINE__)
++#define ef_assert_le(a,b) _ef_assert2((a)<=(b),(a),(b),__FILE__,__LINE__)
++#define ef_assert_nequal(a,b) _ef_assert2((a)!=(b),(a),(b),__FILE__,__LINE__)
++#define ef_assert_ne ef_assert_nequal
++#define ef_assert_ge(a,b) _ef_assert2((a)>=(b),(a),(b),__FILE__,__LINE__)
++#define ef_assert_gt(a,b) _ef_assert2((a)>(b),(a),(b),__FILE__,__LINE__)
++
++/**********************************************************************
++ * Debug checks. ******************************************************
++ **********************************************************************/
++
++#ifdef NDEBUG
++# define EF_VI_MAGIC_SET(p, type)
++# define EF_VI_CHECK_VI(p)
++# define EF_VI_CHECK_EVENT_Q(p)
++# define EF_VI_CHECK_IOBUFSET(p)
++# define EF_VI_CHECK_FILTER(p)
++# define EF_VI_CHECK_SHMBUF(p)
++# define EF_VI_CHECK_PT_EP(p)
++#else
++# define EF_VI 0x3
++# define EF_EPLOCK 0x6
++# define EF_IOBUFSET 0x9
++# define EF_FILTER 0xa
++# define EF_SHMBUF 0x11
++
++# define EF_VI_MAGIC(p, type) \
++ (((unsigned)(type) << 28) | \
++ (((unsigned)(intptr_t)(p)) & 0x0fffffffu))
++
++# if !EF_VI_DO_MAGIC_CHECKS
++# define EF_VI_MAGIC_SET(p, type)
++# define EF_VI_MAGIC_CHECK(p, type)
++# else
++# define EF_VI_MAGIC_SET(p, type) \
++ do { \
++ (p)->magic = EF_VI_MAGIC((p), (type)); \
++ } while (0)
++
++# define EF_VI_MAGIC_OKAY(p, type) \
++ ((p)->magic == EF_VI_MAGIC((p), (type)))
++
++# define EF_VI_MAGIC_CHECK(p, type) \
++ ef_assert(EF_VI_MAGIC_OKAY((p), (type)))
++
++#endif /* EF_VI_DO_MAGIC_CHECKS */
++
++# define EF_VI_CHECK_VI(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_VI);
++
++# define EF_VI_CHECK_EVENT_Q(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_VI); \
++ ef_assert((p)->evq_base); \
++ ef_assert((p)->evq_mask);
++
++# define EF_VI_CHECK_PT_EP(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_VI); \
++ ef_assert((p)->ep_state);
++
++# define EF_VI_CHECK_IOBUFSET(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_IOBUFSET)
++
++# define EF_VI_CHECK_FILTER(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_FILTER);
++
++# define EF_VI_CHECK_SHMBUF(p) \
++ ef_assert(p); \
++ EF_VI_MAGIC_CHECK((p), EF_SHMBUF);
++
++#endif
++
++#ifndef NDEBUG
++# define EF_DRIVER_MAGIC 0x00f00ba4
++# define EF_ASSERT_THIS_DRIVER_VALID(driver) \
++ do{ ef_assert(driver); \
++ EF_VI_MAGIC_CHECK((driver), EF_DRIVER_MAGIC); \
++ ef_assert((driver)->init); }while(0)
++
++# define EF_ASSERT_DRIVER_VALID() EF_ASSERT_THIS_DRIVER_VALID(&ci_driver)
++#else
++# define EF_ASSERT_THIS_DRIVER_VALID(driver)
++# define EF_ASSERT_DRIVER_VALID()
++#endif
++
++
++/* *************************************
++ * Power of 2 FIFO
++ */
++
++#define EF_VI_FIFO2_M(f, x) ((x) & ((f)->fifo_mask))
++#define ef_vi_fifo2_valid(f) ((f) && (f)->fifo && (f)->fifo_mask > 0 && \
++ (f)->fifo_rd_i <= (f)->fifo_mask && \
++ (f)->fifo_wr_i <= (f)->fifo_mask && \
++ EF_VI_IS_POW2((f)->fifo_mask+1u))
++
++#define ef_vi_fifo2_init(f, cap) \
++ do{ ef_assert(EF_VI_IS_POW2((cap) + 1)); \
++ (f)->fifo_rd_i = (f)->fifo_wr_i = 0u; \
++ (f)->fifo_mask = (cap); \
++ }while(0)
++
++#define ef_vi_fifo2_is_empty(f) ((f)->fifo_rd_i == (f)->fifo_wr_i)
++#define ef_vi_fifo2_capacity(f) ((f)->fifo_mask)
++#define ef_vi_fifo2_buf_size(f) ((f)->fifo_mask + 1u)
++#define ef_vi_fifo2_end(f) ((f)->fifo + ef_vi_fifo2_buf_size(f))
++#define ef_vi_fifo2_peek(f) ((f)->fifo[(f)->fifo_rd_i])
++#define ef_vi_fifo2_poke(f) ((f)->fifo[(f)->fifo_wr_i])
++#define ef_vi_fifo2_num(f) EF_VI_FIFO2_M((f),(f)->fifo_wr_i-(f)->fifo_rd_i)
++
++#define ef_vi_fifo2_wr_prev(f) \
++ do{ (f)->fifo_wr_i = EF_VI_FIFO2_M((f), (f)->fifo_wr_i - 1u); }while(0)
++#define ef_vi_fifo2_wr_next(f) \
++ do{ (f)->fifo_wr_i = EF_VI_FIFO2_M((f), (f)->fifo_wr_i + 1u); }while(0)
++#define ef_vi_fifo2_rd_adv(f, n) \
++ do{ (f)->fifo_rd_i = EF_VI_FIFO2_M((f), (f)->fifo_rd_i + (n)); }while(0)
++#define ef_vi_fifo2_rd_prev(f) \
++ do{ (f)->fifo_rd_i = EF_VI_FIFO2_M((f), (f)->fifo_rd_i - 1u); }while(0)
++#define ef_vi_fifo2_rd_next(f) \
++ do{ (f)->fifo_rd_i = EF_VI_FIFO2_M((f), (f)->fifo_rd_i + 1u); }while(0)
++
++#define ef_vi_fifo2_put(f, v) \
++ do{ ef_vi_fifo2_poke(f) = (v); ef_vi_fifo2_wr_next(f); }while(0)
++#define ef_vi_fifo2_get(f, pv) \
++ do{ *(pv) = ef_vi_fifo2_peek(f); ef_vi_fifo2_rd_next(f); }while(0)
++
++
++/* *********************************************************************
++ * Eventq handling
++ */
++
++typedef union {
++ uint64_t u64;
++ struct {
++ uint32_t a;
++ uint32_t b;
++ } opaque;
++} ef_vi_event;
++
++
++#define EF_VI_EVENT_OFFSET(q, i) \
++ (((q)->evq_state->evq_ptr - (i) * sizeof(ef_vi_event)) & (q)->evq_mask)
++
++#define EF_VI_EVENT_PTR(q, i) \
++ ((ef_vi_event*) ((q)->evq_base + EF_VI_EVENT_OFFSET((q), (i))))
++
++/* *********************************************************************
++ * Miscellaneous goodies
++ */
++#ifdef NDEBUG
++# define EF_VI_DEBUG(x)
++#else
++# define EF_VI_DEBUG(x) x
++#endif
++
++#define EF_VI_ROUND_UP(i, align) (((i)+(align)-1u) & ~((align)-1u))
++#define EF_VI_ALIGN_FWD(p, align) (((p)+(align)-1u) & ~((align)-1u))
++#define EF_VI_ALIGN_BACK(p, align) ((p) & ~((align)-1u))
++#define EF_VI_PTR_ALIGN_BACK(p, align) \
++ ((char*)EF_VI_ALIGN_BACK(((intptr_t)(p)), ((intptr_t)(align))))
++#define EF_VI_IS_POW2(x) ((x) && ! ((x) & ((x) - 1)))
++
++
++/* ********************************************************************
++ */
++
++extern void falcon_vi_init(ef_vi*, void* vvis ) EF_VI_HF;
++extern void ef_eventq_state_init(ef_vi* evq) EF_VI_HF;
++extern void __ef_init(void) EF_VI_HF;
++
++
++#endif /* __CI_EF_VI_INTERNAL_H__ */
++
+diff --git a/drivers/xen/sfc_netfront/etherfabric/ef_vi.h b/drivers/xen/sfc_netfront/etherfabric/ef_vi.h
+new file mode 100644
+index 0000000..44c91a5
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/etherfabric/ef_vi.h
+@@ -0,0 +1,665 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \brief Virtual Interface
++ * \date 2007/05/16
++ */
++
++#ifndef __EFAB_EF_VI_H__
++#define __EFAB_EF_VI_H__
++
++
++/**********************************************************************
++ * Primitive types ****************************************************
++ **********************************************************************/
++
++/* We standardise on the types from stdint.h and synthesise these types
++ * for compilers/platforms that don't provide them */
++
++# include <linux/types.h>
++# define EF_VI_ALIGN(x) __attribute__ ((aligned (x)))
++# define ef_vi_inline static inline
++
++
++
++/**********************************************************************
++ * Types **************************************************************
++ **********************************************************************/
++
++typedef uint32_t ef_eventq_ptr;
++
++typedef uint64_t ef_addr;
++typedef char* ef_vi_ioaddr_t;
++
++/**********************************************************************
++ * ef_event ***********************************************************
++ **********************************************************************/
++
++/*! \i_ef_vi A DMA request identifier.
++**
++** This is an integer token specified by the transport and associated
++** with a DMA request. It is returned to the VI user with DMA completion
++** events. It is typically used to identify the buffer associated with
++** the transfer.
++*/
++typedef int ef_request_id;
++
++typedef union {
++ uint64_t u64[1];
++ uint32_t u32[2];
++} ef_vi_qword;
++
++typedef ef_vi_qword ef_hw_event;
++
++#define EF_REQUEST_ID_BITS 16u
++#define EF_REQUEST_ID_MASK ((1u << EF_REQUEST_ID_BITS) - 1u)
++
++/*! \i_ef_event An [ef_event] is a token that identifies something that
++** has happened. Examples include packets received, packets transmitted
++** and errors.
++*/
++typedef union {
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ } generic;
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ /*ef_request_id request_id :EF_REQUEST_ID_BITS;*/
++ unsigned q_id :16;
++ unsigned len :16;
++ unsigned flags :16;
++ } rx;
++ struct { /* This *must* have same layout as [rx]. */
++ ef_hw_event ev;
++ unsigned type :16;
++ /*ef_request_id request_id :EF_REQUEST_ID_BITS;*/
++ unsigned q_id :16;
++ unsigned len :16;
++ unsigned flags :16;
++ unsigned subtype :16;
++ } rx_discard;
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ /*ef_request_id request_id :EF_REQUEST_ID_BITS;*/
++ unsigned q_id :16;
++ } tx;
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ /*ef_request_id request_id :EF_REQUEST_ID_BITS;*/
++ unsigned q_id :16;
++ unsigned subtype :16;
++ } tx_error;
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ unsigned q_id :16;
++ } rx_no_desc_trunc;
++ struct {
++ ef_hw_event ev;
++ unsigned type :16;
++ unsigned data;
++ } sw;
++} ef_event;
++
++
++#define EF_EVENT_TYPE(e) ((e).generic.type)
++enum {
++ /** Good data was received. */
++ EF_EVENT_TYPE_RX,
++ /** Packets have been sent. */
++ EF_EVENT_TYPE_TX,
++ /** Data received and buffer consumed, but something is wrong. */
++ EF_EVENT_TYPE_RX_DISCARD,
++ /** Transmit of packet failed. */
++ EF_EVENT_TYPE_TX_ERROR,
++ /** Received packet was truncated due to lack of descriptors. */
++ EF_EVENT_TYPE_RX_NO_DESC_TRUNC,
++ /** Software generated event. */
++ EF_EVENT_TYPE_SW,
++ /** Event queue overflow. */
++ EF_EVENT_TYPE_OFLOW,
++};
++
++#define EF_EVENT_RX_BYTES(e) ((e).rx.len)
++#define EF_EVENT_RX_Q_ID(e) ((e).rx.q_id)
++#define EF_EVENT_RX_CONT(e) ((e).rx.flags & EF_EVENT_FLAG_CONT)
++#define EF_EVENT_RX_SOP(e) ((e).rx.flags & EF_EVENT_FLAG_SOP)
++#define EF_EVENT_RX_ISCSI_OKAY(e) ((e).rx.flags & EF_EVENT_FLAG_ISCSI_OK)
++#define EF_EVENT_FLAG_SOP 0x1
++#define EF_EVENT_FLAG_CONT 0x2
++#define EF_EVENT_FLAG_ISCSI_OK 0x4
++
++#define EF_EVENT_TX_Q_ID(e) ((e).tx.q_id)
++
++#define EF_EVENT_RX_DISCARD_Q_ID(e) ((e).rx_discard.q_id)
++#define EF_EVENT_RX_DISCARD_LEN(e) ((e).rx_discard.len)
++#define EF_EVENT_RX_DISCARD_TYPE(e) ((e).rx_discard.subtype)
++enum {
++ EF_EVENT_RX_DISCARD_CSUM_BAD,
++ EF_EVENT_RX_DISCARD_CRC_BAD,
++ EF_EVENT_RX_DISCARD_TRUNC,
++ EF_EVENT_RX_DISCARD_RIGHTS,
++ EF_EVENT_RX_DISCARD_OTHER,
++};
++
++#define EF_EVENT_TX_ERROR_Q_ID(e) ((e).tx_error.q_id)
++#define EF_EVENT_TX_ERROR_TYPE(e) ((e).tx_error.subtype)
++enum {
++ EF_EVENT_TX_ERROR_RIGHTS,
++ EF_EVENT_TX_ERROR_OFLOW,
++ EF_EVENT_TX_ERROR_2BIG,
++ EF_EVENT_TX_ERROR_BUS,
++};
++
++#define EF_EVENT_RX_NO_DESC_TRUNC_Q_ID(e) ((e).rx_no_desc_trunc.q_id)
++
++#define EF_EVENT_SW_DATA_MASK 0xffff
++#define EF_EVENT_SW_DATA(e) ((e).sw.data)
++
++#define EF_EVENT_FMT "[ev:%x:%08x:%08x]"
++#define EF_EVENT_PRI_ARG(e) (unsigned) (e).generic.type, \
++ (unsigned) (e).generic.ev.u32[1], \
++ (unsigned) (e).generic.ev.u32[0]
++
++#define EF_GET_HW_EV(e) ((e).generic.ev)
++#define EF_GET_HW_EV_PTR(e) (&(e).generic.ev)
++#define EF_GET_HW_EV_U64(e) ((e).generic.ev.u64[0])
++
++
++/* ***************** */
++
++/*! Used by netif shared state. Must use types of explicit size. */
++typedef struct {
++ uint16_t rx_last_desc_ptr; /* for RX duplicates */
++ uint8_t bad_sop; /* bad SOP detected */
++ uint8_t frag_num; /* next fragment #, 0=>SOP */
++} ef_rx_dup_state_t;
++
++
++/* Max number of ports on any SF NIC. */
++#define EFAB_DMAQS_PER_EVQ_MAX 32
++
++typedef struct {
++ ef_eventq_ptr evq_ptr;
++ int32_t trashed;
++ ef_rx_dup_state_t rx_dup_state[EFAB_DMAQS_PER_EVQ_MAX];
++} ef_eventq_state;
++
++
++/*! \i_ef_base [ef_iovec] is similar the standard [struct iovec]. An
++** array of these is used to designate a scatter/gather list of I/O
++** buffers.
++*/
++typedef struct {
++ ef_addr iov_base EF_VI_ALIGN(8);
++ unsigned iov_len;
++} ef_iovec;
++
++/* Falcon constants */
++#define TX_EV_DESC_PTR_LBN 0
++
++/**********************************************************************
++ * ef_iobufset ********************************************************
++ **********************************************************************/
++
++/*! \i_ef_bufs An [ef_iobufset] is a collection of buffers to be used
++** with the NIC.
++*/
++typedef struct ef_iobufset {
++ unsigned magic;
++ unsigned bufs_mmap_bytes;
++ unsigned bufs_handle;
++ int bufs_ptr_off;
++ ef_addr bufs_addr;
++ unsigned bufs_size; /* size rounded to pow2 */
++ int bufs_num;
++ int faultonaccess;
++} ef_iobufset;
++
++
++/**********************************************************************
++ * ef_vi **************************************************************
++ **********************************************************************/
++
++enum ef_vi_flags {
++ EF_VI_RX_SCATTER = 0x1,
++ EF_VI_ISCSI_RX_HDIG = 0x2,
++ EF_VI_ISCSI_TX_HDIG = 0x4,
++ EF_VI_ISCSI_RX_DDIG = 0x8,
++ EF_VI_ISCSI_TX_DDIG = 0x10,
++ EF_VI_TX_PHYS_ADDR = 0x20,
++ EF_VI_RX_PHYS_ADDR = 0x40,
++ EF_VI_TX_IP_CSUM_DIS = 0x80,
++ EF_VI_TX_TCPUDP_CSUM_DIS= 0x100,
++ EF_VI_TX_TCPUDP_ONLY = 0x200,
++ /* Flags in range 0xXXXX0000 are for internal use. */
++};
++
++typedef struct {
++ uint32_t added;
++ uint32_t removed;
++} ef_vi_txq_state;
++
++typedef struct {
++ uint32_t added;
++ uint32_t removed;
++} ef_vi_rxq_state;
++
++typedef struct {
++ uint32_t mask;
++ void* doorbell;
++ void* descriptors;
++ uint16_t* ids;
++ unsigned misalign_mask;
++} ef_vi_txq;
++
++typedef struct {
++ uint32_t mask;
++ void* doorbell;
++ void* descriptors;
++ uint16_t* ids;
++} ef_vi_rxq;
++
++typedef struct {
++ ef_eventq_state evq;
++ ef_vi_txq_state txq;
++ ef_vi_rxq_state rxq;
++ /* Followed by request id fifos. */
++} ef_vi_state;
++
++/*! \i_ef_vi A virtual interface.
++**
++** An [ef_vi] represents a virtual interface on a specific NIC. A
++** virtual interface is a collection of an event queue and two DMA queues
++** used to pass Ethernet frames between the transport implementation and
++** the network.
++*/
++typedef struct ef_vi {
++ unsigned magic;
++
++ unsigned vi_resource_id;
++ unsigned vi_resource_handle_hack;
++ unsigned vi_i;
++
++ char* vi_mem_mmap_ptr;
++ int vi_mem_mmap_bytes;
++ char* vi_io_mmap_ptr;
++ int vi_io_mmap_bytes;
++
++ ef_eventq_state* evq_state;
++ char* evq_base;
++ unsigned evq_mask;
++ ef_vi_ioaddr_t evq_timer_reg;
++
++ ef_vi_txq vi_txq;
++ ef_vi_rxq vi_rxq;
++ ef_vi_state* ep_state;
++ enum ef_vi_flags vi_flags;
++} ef_vi;
++
++
++enum ef_vi_arch {
++ EF_VI_ARCH_FALCON,
++};
++
++
++struct ef_vi_nic_type {
++ unsigned char arch;
++ char variant;
++ unsigned char revision;
++};
++
++
++/* This structure is opaque to the client & used to pass mapping data
++ * from the resource manager to the ef_vi lib. for ef_vi_init().
++ */
++struct vi_mappings {
++ uint32_t signature;
++# define VI_MAPPING_VERSION 0x02 /*Byte: Increment me if struct altered*/
++# define VI_MAPPING_SIGNATURE (0xBA1150 + VI_MAPPING_VERSION)
++
++ struct ef_vi_nic_type nic_type;
++
++ int vi_instance;
++
++ unsigned evq_bytes;
++ char* evq_base;
++ ef_vi_ioaddr_t evq_timer_reg;
++
++ unsigned rx_queue_capacity;
++ ef_vi_ioaddr_t rx_dma_ef1;
++ char* rx_dma_falcon;
++ ef_vi_ioaddr_t rx_bell;
++
++ unsigned tx_queue_capacity;
++ ef_vi_ioaddr_t tx_dma_ef1;
++ char* tx_dma_falcon;
++ ef_vi_ioaddr_t tx_bell;
++};
++/* This is used by clients to allocate a suitably sized buffer for the
++ * resource manager to fill & ef_vi_init() to use. */
++#define VI_MAPPINGS_SIZE (sizeof(struct vi_mappings))
++
++
++/**********************************************************************
++ * ef_config **********************************************************
++ **********************************************************************/
++
++struct ef_config_t {
++ int log; /* debug logging level */
++};
++
++extern struct ef_config_t ef_config;
++
++
++/**********************************************************************
++ * ef_vi **************************************************************
++ **********************************************************************/
++
++/* Initialise [data_area] with information required to initialise an ef_vi.
++ * In the following, an unused param should be set to NULL. Note the case
++ * marked (*) of [iobuf_mmap] for falcon/driver; for normal driver this
++ * must be NULL.
++ *
++ * \param data_area [in,out] required, must ref at least VI_MAPPINGS_SIZE
++ * bytes
++ * \param evq_capacity [in] number of events in event queue. Specify 0 for
++ * no event queue.
++ * \param rxq_capacity [in] number of descriptors in RX DMA queue. Specify
++ * 0 for no RX queue.
++ * \param txq_capacity [in] number of descriptors in TX DMA queue. Specify
++ * 0 for no TX queue.
++ * \param mmap_info [in] mem-map info for resource
++ * \param io_mmap [in] ef1, required
++ * falcon, required
++ * \param iobuf_mmap [in] ef1, UL: unused
++ * falcon, UL: required
++ */
++extern void ef_vi_init_mapping_vi(void* data_area, struct ef_vi_nic_type,
++ unsigned rxq_capacity,
++ unsigned txq_capacity, int instance,
++ void* io_mmap, void* iobuf_mmap_rx,
++ void* iobuf_mmap_tx, enum ef_vi_flags);
++
++
++extern void ef_vi_init_mapping_evq(void* data_area, struct ef_vi_nic_type,
++ int instance, unsigned evq_bytes,
++ void* base, void* timer_reg);
++
++ef_vi_inline unsigned ef_vi_resource_id(ef_vi* vi)
++{
++ return vi->vi_resource_id;
++}
++
++ef_vi_inline enum ef_vi_flags ef_vi_flags(ef_vi* vi)
++{
++ return vi->vi_flags;
++}
++
++
++/**********************************************************************
++ * Receive interface **************************************************
++ **********************************************************************/
++
++/*! \i_ef_vi Returns the amount of space in the RX descriptor ring.
++**
++** \return the amount of space in the queue.
++*/
++ef_vi_inline int ef_vi_receive_space(ef_vi* vi)
++{
++ ef_vi_rxq_state* qs = &vi->ep_state->rxq;
++ return vi->vi_rxq.mask - (qs->added - qs->removed);
++}
++
++
++/*! \i_ef_vi Returns the fill level of the RX descriptor ring.
++**
++** \return the fill level of the queue.
++*/
++ef_vi_inline int ef_vi_receive_fill_level(ef_vi* vi)
++{
++ ef_vi_rxq_state* qs = &vi->ep_state->rxq;
++ return qs->added - qs->removed;
++}
++
++
++ef_vi_inline int ef_vi_receive_capacity(ef_vi* vi)
++{
++ return vi->vi_rxq.mask;
++}
++
++/*! \i_ef_vi Complete a receive operation.
++**
++** When a receive completion event is received, it should be passed to
++** this function. The request-id for the buffer that the packet was
++** delivered to is returned.
++**
++** After this function returns, more space may be available in the
++** receive queue.
++*/
++extern ef_request_id ef_vi_receive_done(const ef_vi*, const ef_event*);
++
++/*! \i_ef_vi Return request ID indicated by a receive event
++ */
++ef_vi_inline ef_request_id ef_vi_receive_request_id(const ef_vi* vi,
++ const ef_event* ef_ev)
++{
++ const ef_vi_qword* ev = EF_GET_HW_EV_PTR(*ef_ev);
++ return ev->u32[0] & vi->vi_rxq.mask;
++}
++
++
++/*! \i_ef_vi Form a receive descriptor.
++**
++** If \c initial_rx_bytes is zero use a reception size at least as large
++** as an MTU.
++*/
++extern int ef_vi_receive_init(ef_vi* vi, ef_addr addr, ef_request_id dma_id,
++ int intial_rx_bytes);
++
++/*! \i_ef_vi Submit initialised receive descriptors to the NIC. */
++extern void ef_vi_receive_push(ef_vi* vi);
++
++/*! \i_ef_vi Post a buffer on the receive queue.
++**
++** \return 0 on success, or -EAGAIN if the receive queue is full
++*/
++extern int ef_vi_receive_post(ef_vi*, ef_addr addr,
++ ef_request_id dma_id);
++
++/**********************************************************************
++ * Transmit interface *************************************************
++ **********************************************************************/
++
++/*! \i_ef_vi Return the amount of space (in descriptors) in the transmit
++** queue.
++**
++** \return the amount of space in the queue (in descriptors)
++*/
++ef_vi_inline int ef_vi_transmit_space(ef_vi* vi)
++{
++ ef_vi_txq_state* qs = &vi->ep_state->txq;
++ return vi->vi_txq.mask - (qs->added - qs->removed);
++}
++
++
++/*! \i_ef_vi Returns the fill level of the TX descriptor ring.
++**
++** \return the fill level of the queue.
++*/
++ef_vi_inline int ef_vi_transmit_fill_level(ef_vi* vi)
++{
++ ef_vi_txq_state* qs = &vi->ep_state->txq;
++ return qs->added - qs->removed;
++}
++
++
++/*! \i_ef_vi Returns the total capacity of the TX descriptor ring.
++**
++** \return the capacity of the queue.
++*/
++ef_vi_inline int ef_vi_transmit_capacity(ef_vi* vi)
++{
++ return vi->vi_txq.mask;
++}
++
++
++/*! \i_ef_vi Transmit a packet.
++**
++** \param bytes must be greater than ETH_ZLEN.
++** \return -EAGAIN if the transmit queue is full, or 0 on success
++*/
++extern int ef_vi_transmit(ef_vi*, ef_addr, int bytes, ef_request_id dma_id);
++
++/*! \i_ef_vi Transmit a packet using a gather list.
++**
++** \param iov_len must be greater than zero
++** \param iov the first must be non-zero in length (but others need not)
++**
++** \return -EAGAIN if the queue is full, or 0 on success
++*/
++extern int ef_vi_transmitv(ef_vi*, const ef_iovec* iov, int iov_len,
++ ef_request_id dma_id);
++
++/*! \i_ef_vi Initialise a DMA request.
++**
++** \return -EAGAIN if the queue is full, or 0 on success
++*/
++extern int ef_vi_transmit_init(ef_vi*, ef_addr, int bytes,
++ ef_request_id dma_id);
++
++/*! \i_ef_vi Initialise a DMA request.
++**
++** \return -EAGAIN if the queue is full, or 0 on success
++*/
++extern int ef_vi_transmitv_init(ef_vi*, const ef_iovec*, int iov_len,
++ ef_request_id dma_id);
++
++/*! \i_ef_vi Submit DMA requests to the NIC.
++**
++** The DMA requests must have been initialised using
++** ef_vi_transmit_init() or ef_vi_transmitv_init().
++*/
++extern void ef_vi_transmit_push(ef_vi*);
++
++
++/*! \i_ef_vi Maximum number of transmit completions per transmit event. */
++#define EF_VI_TRANSMIT_BATCH 64
++
++/*! \i_ef_vi Determine the set of [ef_request_id]s for each DMA request
++** which has been completed by a given transmit completion
++** event.
++**
++** \param ids must point to an array of length EF_VI_TRANSMIT_BATCH
++** \return the number of valid [ef_request_id]s (can be zero)
++*/
++extern int ef_vi_transmit_unbundle(ef_vi* ep, const ef_event*,
++ ef_request_id* ids);
++
++
++/*! \i_ef_event Returns true if ef_eventq_poll() will return event(s). */
++extern int ef_eventq_has_event(ef_vi* vi);
++
++/*! \i_ef_event Returns true if there are quite a few events in the event
++** queue.
++**
++** This looks ahead in the event queue, so has the property that it will
++** not ping-pong a cache-line when it is called concurrently with events
++** being delivered.
++*/
++extern int ef_eventq_has_many_events(ef_vi* evq, int look_ahead);
++
++/*! Type of function to handle unknown events arriving on event queue
++** Return CI_TRUE iff the event has been handled.
++*/
++typedef int/*bool*/ ef_event_handler_fn(void* priv, ef_vi* evq, ef_event* ev);
++
++/*! Standard poll exception routine */
++extern int/*bool*/ ef_eventq_poll_exception(void* priv, ef_vi* evq,
++ ef_event* ev);
++
++/*! \i_ef_event Retrieve events from the event queue, handle RX/TX events
++** and pass any others to an exception handler function
++**
++** \return The number of events retrieved.
++*/
++extern int ef_eventq_poll_evs(ef_vi* evq, ef_event* evs, int evs_len,
++ ef_event_handler_fn *exception, void *expt_priv);
++
++/*! \i_ef_event Retrieve events from the event queue.
++**
++** \return The number of events retrieved.
++*/
++ef_vi_inline int ef_eventq_poll(ef_vi* evq, ef_event* evs, int evs_len)
++{
++ return ef_eventq_poll_evs(evq, evs, evs_len,
++ &ef_eventq_poll_exception, (void*)0);
++}
++
++/*! \i_ef_event Returns the capacity of an event queue. */
++ef_vi_inline int ef_eventq_capacity(ef_vi* vi)
++{
++ return (vi->evq_mask + 1u) / sizeof(ef_hw_event);
++}
++
++/* Returns the instance ID of [vi] */
++ef_vi_inline unsigned ef_vi_instance(ef_vi* vi)
++{ return vi->vi_i; }
++
++
++/**********************************************************************
++ * Initialisation *****************************************************
++ **********************************************************************/
++
++/*! Return size of state buffer of an initialised VI. */
++extern int ef_vi_state_bytes(ef_vi*);
++
++/*! Return size of buffer needed for VI state given sizes of RX and TX
++** DMA queues. Queue sizes must be legal sizes (power of 2), or 0 (no
++** queue).
++*/
++extern int ef_vi_calc_state_bytes(int rxq_size, int txq_size);
++
++/*! Initialise [ef_vi] from the provided resources. [vvis] must have been
++** created by ef_make_vi_data() & remains owned by the caller.
++*/
++extern void ef_vi_init(ef_vi*, void* vi_info, ef_vi_state* state,
++ ef_eventq_state* evq_state, enum ef_vi_flags);
++
++extern void ef_vi_state_init(ef_vi*);
++extern void ef_eventq_state_init(ef_vi*);
++
++/*! Convert an efhw device arch to ef_vi_arch, or returns -1 if not
++** recognised.
++*/
++extern int ef_vi_arch_from_efhw_arch(int efhw_arch);
++
++
++#endif /* __EFAB_EF_VI_H__ */
+diff --git a/drivers/xen/sfc_netfront/falcon_event.c b/drivers/xen/sfc_netfront/falcon_event.c
+new file mode 100644
+index 0000000..3d8c9c5
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/falcon_event.c
+@@ -0,0 +1,346 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Routine to poll event queues.
++ * \date 2003/03/04
++ */
++
++/*! \cidoxg_lib_ef */
++#include "ef_vi_internal.h"
++
++/* Be worried about this on byteswapped machines */
++/* Due to crazy chipsets, we see the event words being written in
++** arbitrary order (bug4539). So test for presence of event must ensure
++** that both halves have changed from the null.
++*/
++# define EF_VI_IS_EVENT(evp) \
++ ( (((evp)->opaque.a != (uint32_t)-1) && \
++ ((evp)->opaque.b != (uint32_t)-1)) )
++
++
++#ifdef NDEBUG
++# define IS_DEBUG 0
++#else
++# define IS_DEBUG 1
++#endif
++
++
++/*! Check for RX events with inconsistent SOP/CONT
++**
++** Returns true if this event should be discarded
++*/
++ef_vi_inline int ef_eventq_is_rx_sop_cont_bad_efab(ef_vi* vi,
++ const ef_vi_qword* ev)
++{
++ ef_rx_dup_state_t* rx_dup_state;
++ uint8_t* bad_sop;
++
++ unsigned label = QWORD_GET_U(RX_EV_Q_LABEL, *ev);
++ unsigned sop = QWORD_TEST_BIT(RX_SOP, *ev);
++
++ ef_assert(vi);
++ ef_assert_lt(label, EFAB_DMAQS_PER_EVQ_MAX);
++
++ rx_dup_state = &vi->evq_state->rx_dup_state[label];
++ bad_sop = &rx_dup_state->bad_sop;
++
++ if( ! ((vi->vi_flags & EF_VI_BUG5692_WORKAROUND) || IS_DEBUG) ) {
++ *bad_sop = (*bad_sop && !sop);
++ }
++ else {
++ unsigned cont = QWORD_TEST_BIT(RX_JUMBO_CONT, *ev);
++ uint8_t *frag_num = &rx_dup_state->frag_num;
++
++ /* bad_sop should latch till the next sop */
++ *bad_sop = (*bad_sop && !sop) || ( !!sop != (*frag_num==0) );
++
++ /* we do not check the number of bytes relative to the
++ * fragment number and size of the user rx buffer here
++ * because we don't know the size of the user rx
++ * buffer - we probably should perform this check in
++ * the nearest code calling this though.
++ */
++ *frag_num = cont ? (*frag_num + 1) : 0;
++ }
++
++ return *bad_sop;
++}
++
++
++ef_vi_inline int falcon_rx_check_dup(ef_vi* evq, ef_event* ev_out,
++ const ef_vi_qword* ev)
++{
++ unsigned q_id = QWORD_GET_U(RX_EV_Q_LABEL, *ev);
++ unsigned desc_ptr = QWORD_GET_U(RX_EV_DESC_PTR, *ev);
++ ef_rx_dup_state_t* rx_dup_state = &evq->evq_state->rx_dup_state[q_id];
++
++ if(likely( desc_ptr != rx_dup_state->rx_last_desc_ptr )) {
++ rx_dup_state->rx_last_desc_ptr = desc_ptr;
++ return 0;
++ }
++
++ rx_dup_state->rx_last_desc_ptr = desc_ptr;
++ rx_dup_state->bad_sop = 1;
++#ifndef NDEBUG
++ rx_dup_state->frag_num = 0;
++#endif
++ BUG_ON(!QWORD_TEST_BIT(RX_EV_FRM_TRUNC, *ev));
++ BUG_ON( QWORD_TEST_BIT(RX_EV_PKT_OK, *ev));
++ BUG_ON(!QWORD_GET_U(RX_EV_BYTE_CNT, *ev) == 0);
++ ev_out->rx_no_desc_trunc.type = EF_EVENT_TYPE_RX_NO_DESC_TRUNC;
++ ev_out->rx_no_desc_trunc.q_id = q_id;
++ return 1;
++}
++
++
++ef_vi_inline void falcon_rx_event(ef_event* ev_out, const ef_vi_qword* ev)
++{
++ if(likely( QWORD_TEST_BIT(RX_EV_PKT_OK, *ev) )) {
++ ev_out->rx.type = EF_EVENT_TYPE_RX;
++ ev_out->rx.q_id = QWORD_GET_U(RX_EV_Q_LABEL, *ev);
++ ev_out->rx.len = QWORD_GET_U(RX_EV_BYTE_CNT, *ev);
++ if( QWORD_TEST_BIT(RX_SOP, *ev) )
++ ev_out->rx.flags = EF_EVENT_FLAG_SOP;
++ else
++ ev_out->rx.flags = 0;
++ if( QWORD_TEST_BIT(RX_JUMBO_CONT, *ev) )
++ ev_out->rx.flags |= EF_EVENT_FLAG_CONT;
++ if( QWORD_TEST_BIT(RX_iSCSI_PKT_OK, *ev) )
++ ev_out->rx.flags |= EF_EVENT_FLAG_ISCSI_OK;
++ }
++ else {
++ ev_out->rx_discard.type = EF_EVENT_TYPE_RX_DISCARD;
++ ev_out->rx_discard.q_id = QWORD_GET_U(RX_EV_Q_LABEL, *ev);
++ ev_out->rx_discard.len = QWORD_GET_U(RX_EV_BYTE_CNT, *ev);
++#if 1 /* hack for ptloop compatability: ?? TODO purge */
++ if( QWORD_TEST_BIT(RX_SOP, *ev) )
++ ev_out->rx_discard.flags = EF_EVENT_FLAG_SOP;
++ else
++ ev_out->rx_discard.flags = 0;
++ if( QWORD_TEST_BIT(RX_JUMBO_CONT, *ev) )
++ ev_out->rx_discard.flags |= EF_EVENT_FLAG_CONT;
++ if( QWORD_TEST_BIT(RX_iSCSI_PKT_OK, *ev) )
++ ev_out->rx_discard.flags |= EF_EVENT_FLAG_ISCSI_OK;
++#endif
++ /* Order matters here: more fundamental errors first. */
++ if( QWORD_TEST_BIT(RX_EV_BUF_OWNER_ID_ERR, *ev) )
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_RIGHTS;
++ else if( QWORD_TEST_BIT(RX_EV_FRM_TRUNC, *ev) )
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_TRUNC;
++ else if( QWORD_TEST_BIT(RX_EV_ETH_CRC_ERR, *ev) )
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_CRC_BAD;
++ else if( QWORD_TEST_BIT(RX_EV_IP_HDR_CHKSUM_ERR, *ev) )
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_CSUM_BAD;
++ else if( QWORD_TEST_BIT(RX_EV_TCP_UDP_CHKSUM_ERR, *ev) )
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_CSUM_BAD;
++ else
++ ev_out->rx_discard.subtype =
++ EF_EVENT_RX_DISCARD_OTHER;
++ }
++}
++
++
++ef_vi_inline void falcon_tx_event(ef_event* ev_out, const ef_vi_qword* ev)
++{
++ /* Danger danger! No matter what we ask for wrt batching, we
++ ** will get a batched event every 16 descriptors, and we also
++ ** get dma-queue-empty events. i.e. Duplicates are expected.
++ **
++ ** In addition, if it's been requested in the descriptor, we
++ ** get an event per descriptor. (We don't currently request
++ ** this).
++ */
++ if(likely( QWORD_TEST_BIT(TX_EV_COMP, *ev) )) {
++ ev_out->tx.type = EF_EVENT_TYPE_TX;
++ ev_out->tx.q_id = QWORD_GET_U(TX_EV_Q_LABEL, *ev);
++ }
++ else {
++ ev_out->tx_error.type = EF_EVENT_TYPE_TX_ERROR;
++ ev_out->tx_error.q_id = QWORD_GET_U(TX_EV_Q_LABEL, *ev);
++ if(likely( QWORD_TEST_BIT(TX_EV_BUF_OWNER_ID_ERR, *ev) ))
++ ev_out->tx_error.subtype = EF_EVENT_TX_ERROR_RIGHTS;
++ else if(likely( QWORD_TEST_BIT(TX_EV_WQ_FF_FULL, *ev) ))
++ ev_out->tx_error.subtype = EF_EVENT_TX_ERROR_OFLOW;
++ else if(likely( QWORD_TEST_BIT(TX_EV_PKT_TOO_BIG, *ev) ))
++ ev_out->tx_error.subtype = EF_EVENT_TX_ERROR_2BIG;
++ else if(likely( QWORD_TEST_BIT(TX_EV_PKT_ERR, *ev) ))
++ ev_out->tx_error.subtype = EF_EVENT_TX_ERROR_BUS;
++ }
++}
++
++
++static void mark_bad(ef_event* ev)
++{
++ ev->generic.ev.u64[0] &=~ ((uint64_t) 1u << RX_EV_PKT_OK_LBN);
++}
++
++
++int ef_eventq_poll_evs(ef_vi* evq, ef_event* evs, int evs_len,
++ ef_event_handler_fn *exception, void *expt_priv)
++{
++ int evs_len_orig = evs_len;
++
++ EF_VI_CHECK_EVENT_Q(evq);
++ ef_assert(evs);
++ ef_assert_gt(evs_len, 0);
++
++ if(unlikely( EF_VI_IS_EVENT(EF_VI_EVENT_PTR(evq, 1)) ))
++ goto overflow;
++
++ do {
++ { /* Read the event out of the ring, then fiddle with
++ * copied version. Reason is that the ring is
++ * likely to get pushed out of cache by another
++ * event being delivered by hardware. */
++ ef_vi_event* ev = EF_VI_EVENT_PTR(evq, 0);
++ if( ! EF_VI_IS_EVENT(ev) )
++ break;
++ evs->generic.ev.u64[0] = cpu_to_le64 (ev->u64);
++ evq->evq_state->evq_ptr += sizeof(ef_vi_event);
++ ev->u64 = (uint64_t)(int64_t) -1;
++ }
++
++ /* Ugly: Exploit the fact that event code lies in top
++ * bits of event. */
++ ef_assert_ge(EV_CODE_LBN, 32u);
++ switch( evs->generic.ev.u32[1] >> (EV_CODE_LBN - 32u) ) {
++ case RX_IP_EV_DECODE:
++ /* Look for duplicate desc_ptr: it signals
++ * that a jumbo frame was truncated because we
++ * ran out of descriptors. */
++ if(unlikely( falcon_rx_check_dup
++ (evq, evs, &evs->generic.ev) )) {
++ --evs_len;
++ ++evs;
++ break;
++ }
++ else {
++ /* Cope with FalconA1 bugs where RX
++ * gives inconsistent RX events Mark
++ * events as bad until SOP becomes
++ * consistent again
++ * ef_eventq_is_rx_sop_cont_bad() has
++ * side effects - order is important
++ */
++ if(unlikely
++ (ef_eventq_is_rx_sop_cont_bad_efab
++ (evq, &evs->generic.ev) )) {
++ mark_bad(evs);
++ }
++ }
++ falcon_rx_event(evs, &evs->generic.ev);
++ --evs_len;
++ ++evs;
++ break;
++
++ case TX_IP_EV_DECODE:
++ falcon_tx_event(evs, &evs->generic.ev);
++ --evs_len;
++ ++evs;
++ break;
++
++ default:
++ break;
++ }
++ } while( evs_len );
++
++ return evs_len_orig - evs_len;
++
++
++ overflow:
++ evs->generic.type = EF_EVENT_TYPE_OFLOW;
++ evs->generic.ev.u64[0] = (uint64_t)((int64_t)-1);
++ return 1;
++}
++
++
++int/*bool*/ ef_eventq_poll_exception(void* priv, ef_vi* evq, ef_event* ev)
++{
++ int /*bool*/ handled = 0;
++
++ switch( ev->generic.ev.u32[1] >> (EV_CODE_LBN - 32u) ) {
++ case DRIVER_EV_DECODE:
++ if( QWORD_GET_U(DRIVER_EV_SUB_CODE, ev->generic.ev) ==
++ EVQ_INIT_DONE_EV_DECODE )
++ /* EVQ initialised event: ignore. */
++ handled = 1;
++ break;
++ }
++ return handled;
++}
++
++
++void ef_eventq_iterate(ef_vi* vi,
++ void (*fn)(void* arg, ef_vi*, int rel_pos,
++ int abs_pos, void* event),
++ void* arg, int stop_at_end)
++{
++ int i, size_evs = (vi->evq_mask + 1) / sizeof(ef_vi_event);
++
++ for( i = 0; i < size_evs; ++i ) {
++ ef_vi_event* e = EF_VI_EVENT_PTR(vi, -i);
++ if( EF_VI_IS_EVENT(e) )
++ fn(arg, vi, i,
++ EF_VI_EVENT_OFFSET(vi, -i) / sizeof(ef_vi_event),
++ e);
++ else if( stop_at_end )
++ break;
++ }
++}
++
++
++int ef_eventq_has_event(ef_vi* vi)
++{
++ return EF_VI_IS_EVENT(EF_VI_EVENT_PTR(vi, 0));
++}
++
++
++int ef_eventq_has_many_events(ef_vi* vi, int look_ahead)
++{
++ ef_assert_ge(look_ahead, 0);
++ return EF_VI_IS_EVENT(EF_VI_EVENT_PTR(vi, -look_ahead));
++}
++
++
++int ef_eventq_has_rx_event(ef_vi* vi)
++{
++ ef_vi_event* ev;
++ int i, n_evs = 0;
++
++ for( i = 0; EF_VI_IS_EVENT(EF_VI_EVENT_PTR(vi, i)); --i ) {
++ ev = EF_VI_EVENT_PTR(vi, i);
++ if( EFVI_FALCON_EVENT_CODE(ev) == EF_EVENT_TYPE_RX ) n_evs++;
++ }
++ return n_evs;
++}
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netfront/falcon_vi.c b/drivers/xen/sfc_netfront/falcon_vi.c
+new file mode 100644
+index 0000000..27c0653
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/falcon_vi.c
+@@ -0,0 +1,465 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr, stg
++ * \brief Falcon-specific VI
++ * \date 2006/11/30
++ */
++
++#include "ef_vi_internal.h"
++
++
++#define EFVI_FALCON_DMA_TX_FRAG 1
++
++
++/* TX descriptor for both physical and virtual packet transfers */
++typedef union {
++ uint32_t dword[2];
++} ef_vi_falcon_dma_tx_buf_desc;
++typedef ef_vi_falcon_dma_tx_buf_desc ef_vi_falcon_dma_tx_phys_desc;
++
++
++/* RX descriptor for physical addressed transfers */
++typedef union {
++ uint32_t dword[2];
++} ef_vi_falcon_dma_rx_phys_desc;
++
++
++/* RX descriptor for virtual packet transfers */
++typedef struct {
++ uint32_t dword[1];
++} ef_vi_falcon_dma_rx_buf_desc;
++
++/* Buffer table index */
++typedef uint32_t ef_vi_buffer_addr_t;
++
++ef_vi_inline int64_t dma_addr_to_u46(int64_t src_dma_addr)
++{
++ return (src_dma_addr & __FALCON_MASK(46, int64_t));
++}
++
++/*! Setup a physical address based descriptor with a specified length */
++ef_vi_inline void
++__falcon_dma_rx_calc_ip_phys(ef_vi_dma_addr_t dest_pa,
++ ef_vi_falcon_dma_rx_phys_desc *desc,
++ int bytes)
++{
++ int region = 0; /* TODO fixme */
++ int64_t dest = dma_addr_to_u46(dest_pa); /* lower 46 bits */
++
++ DWCHCK(__DW2(RX_KER_BUF_SIZE_LBN), RX_KER_BUF_SIZE_WIDTH);
++ DWCHCK(__DW2(RX_KER_BUF_REGION_LBN),RX_KER_BUF_REGION_WIDTH);
++
++ LWCHK(RX_KER_BUF_ADR_LBN, RX_KER_BUF_ADR_WIDTH);
++
++ RANGECHCK(bytes, RX_KER_BUF_SIZE_WIDTH);
++ RANGECHCK(region, RX_KER_BUF_REGION_WIDTH);
++
++ ef_assert(desc);
++
++ desc->dword[1] = ((bytes << __DW2(RX_KER_BUF_SIZE_LBN)) |
++ (region << __DW2(RX_KER_BUF_REGION_LBN)) |
++ (HIGH(dest,
++ RX_KER_BUF_ADR_LBN,
++ RX_KER_BUF_ADR_WIDTH)));
++
++ desc->dword[0] = LOW(dest,
++ RX_KER_BUF_ADR_LBN,
++ RX_KER_BUF_ADR_WIDTH);
++}
++
++/*! Setup a virtual buffer descriptor for an IPMODE transfer */
++ef_vi_inline void
++__falcon_dma_tx_calc_ip_buf(unsigned buf_id, unsigned buf_ofs, unsigned bytes,
++ int port, int frag,
++ ef_vi_falcon_dma_tx_buf_desc *desc)
++{
++ DWCHCK(__DW2(TX_USR_PORT_LBN), TX_USR_PORT_WIDTH);
++ DWCHCK(__DW2(TX_USR_CONT_LBN), TX_USR_CONT_WIDTH);
++ DWCHCK(__DW2(TX_USR_BYTE_CNT_LBN), TX_USR_BYTE_CNT_WIDTH);
++ LWCHK(RX_KER_BUF_ADR_LBN, RX_KER_BUF_ADR_WIDTH);
++ DWCHCK(TX_USR_BYTE_OFS_LBN, TX_USR_BYTE_OFS_WIDTH);
++
++ RANGECHCK(bytes, TX_USR_BYTE_CNT_WIDTH);
++ RANGECHCK(port, TX_USR_PORT_WIDTH);
++ RANGECHCK(frag, TX_USR_CONT_WIDTH);
++ RANGECHCK(buf_id, TX_USR_BUF_ID_WIDTH);
++ RANGECHCK(buf_ofs, TX_USR_BYTE_OFS_WIDTH);
++
++ ef_assert(desc);
++
++ desc->dword[1] = ((port << __DW2(TX_USR_PORT_LBN)) |
++ (frag << __DW2(TX_USR_CONT_LBN)) |
++ (bytes << __DW2(TX_USR_BYTE_CNT_LBN)) |
++ (HIGH(buf_id,
++ TX_USR_BUF_ID_LBN,
++ TX_USR_BUF_ID_WIDTH)));
++
++ desc->dword[0] = ((LOW(buf_id,
++ TX_USR_BUF_ID_LBN,
++ (TX_USR_BUF_ID_WIDTH))) |
++ (buf_ofs << TX_USR_BYTE_OFS_LBN));
++}
++
++ef_vi_inline void
++falcon_dma_tx_calc_ip_buf_4k(unsigned buf_vaddr, unsigned bytes,
++ int port, int frag,
++ ef_vi_falcon_dma_tx_buf_desc *desc)
++{
++ /* TODO FIXME [buf_vaddr] consists of the buffer index in the
++ ** high bits, and an offset in the low bits. Assumptions
++ ** permate the code that these can be rolled into one 32bit
++ ** value, so this is currently preserved for Falcon. But we
++ ** should change to support 8K pages
++ */
++ unsigned buf_id = EFVI_FALCON_BUFFER_4K_PAGE(buf_vaddr);
++ unsigned buf_ofs = EFVI_FALCON_BUFFER_4K_OFF(buf_vaddr);
++
++ __falcon_dma_tx_calc_ip_buf( buf_id, buf_ofs, bytes, port, frag, desc);
++}
++
++ef_vi_inline void
++falcon_dma_tx_calc_ip_buf(unsigned buf_vaddr, unsigned bytes, int port,
++ int frag, ef_vi_falcon_dma_tx_buf_desc *desc)
++{
++ falcon_dma_tx_calc_ip_buf_4k(buf_vaddr, bytes, port, frag, desc);
++}
++
++/*! Setup a virtual buffer based descriptor */
++ef_vi_inline void
++__falcon_dma_rx_calc_ip_buf(unsigned buf_id, unsigned buf_ofs,
++ ef_vi_falcon_dma_rx_buf_desc *desc)
++{
++ /* check alignment of buffer offset and pack */
++ ef_assert((buf_ofs & 0x1) == 0);
++
++ buf_ofs >>= 1;
++
++ DWCHCK(RX_USR_2BYTE_OFS_LBN, RX_USR_2BYTE_OFS_WIDTH);
++ DWCHCK(RX_USR_BUF_ID_LBN, RX_USR_BUF_ID_WIDTH);
++
++ RANGECHCK(buf_ofs, RX_USR_2BYTE_OFS_WIDTH);
++ RANGECHCK(buf_id, RX_USR_BUF_ID_WIDTH);
++
++ ef_assert(desc);
++
++ desc->dword[0] = ((buf_ofs << RX_USR_2BYTE_OFS_LBN) |
++ (buf_id << RX_USR_BUF_ID_LBN));
++}
++
++ef_vi_inline void
++falcon_dma_rx_calc_ip_buf_4k(unsigned buf_vaddr,
++ ef_vi_falcon_dma_rx_buf_desc *desc)
++{
++ /* TODO FIXME [buf_vaddr] consists of the buffer index in the
++ ** high bits, and an offset in the low bits. Assumptions
++ ** permeate the code that these can be rolled into one 32bit
++ ** value, so this is currently preserved for Falcon. But we
++ ** should change to support 8K pages
++ */
++ unsigned buf_id = EFVI_FALCON_BUFFER_4K_PAGE(buf_vaddr);
++ unsigned buf_ofs = EFVI_FALCON_BUFFER_4K_OFF(buf_vaddr);
++
++ __falcon_dma_rx_calc_ip_buf(buf_id, buf_ofs, desc);
++}
++
++ef_vi_inline void
++falcon_dma_rx_calc_ip_buf(unsigned buf_vaddr,
++ ef_vi_falcon_dma_rx_buf_desc *desc)
++{
++ falcon_dma_rx_calc_ip_buf_4k(buf_vaddr, desc);
++}
++
++
++ef_vi_inline ef_vi_dma_addr_t ef_physaddr(ef_addr efaddr)
++{
++ return (ef_vi_dma_addr_t) efaddr;
++}
++
++
++/*! Convert between an ef_addr and a buffer table index
++** Assert that this was not a physical address
++*/
++ef_vi_inline ef_vi_buffer_addr_t ef_bufaddr(ef_addr efaddr)
++{
++ ef_assert(efaddr < ((uint64_t)1 << 32) );
++
++ return (ef_vi_buffer_addr_t) efaddr;
++}
++
++
++/*! Setup an physical address based descriptor for an IPMODE transfer */
++ef_vi_inline void
++falcon_dma_tx_calc_ip_phys(ef_vi_dma_addr_t src_dma_addr, unsigned bytes,
++ int port, int frag,
++ ef_vi_falcon_dma_tx_phys_desc *desc)
++{
++
++ int region = 0; /* FIXME */
++ int64_t src = dma_addr_to_u46(src_dma_addr); /* lower 46 bits */
++
++ DWCHCK(__DW2(TX_KER_PORT_LBN), TX_KER_PORT_WIDTH);
++ DWCHCK(__DW2(TX_KER_CONT_LBN), TX_KER_CONT_WIDTH);
++ DWCHCK(__DW2(TX_KER_BYTE_CNT_LBN), TX_KER_BYTE_CNT_WIDTH);
++ DWCHCK(__DW2(TX_KER_BUF_REGION_LBN),TX_KER_BUF_REGION_WIDTH);
++
++ LWCHK(TX_KER_BUF_ADR_LBN, TX_KER_BUF_ADR_WIDTH);
++
++ RANGECHCK(port, TX_KER_PORT_WIDTH);
++ RANGECHCK(frag, TX_KER_CONT_WIDTH);
++ RANGECHCK(bytes, TX_KER_BYTE_CNT_WIDTH);
++ RANGECHCK(region, TX_KER_BUF_REGION_WIDTH);
++
++ desc->dword[1] = ((port << __DW2(TX_KER_PORT_LBN)) |
++ (frag << __DW2(TX_KER_CONT_LBN)) |
++ (bytes << __DW2(TX_KER_BYTE_CNT_LBN)) |
++ (region << __DW2(TX_KER_BUF_REGION_LBN)) |
++ (HIGH(src,
++ TX_KER_BUF_ADR_LBN,
++ TX_KER_BUF_ADR_WIDTH)));
++
++ ef_assert_equal(TX_KER_BUF_ADR_LBN, 0);
++ desc->dword[0] = (uint32_t) src_dma_addr;
++}
++
++
++void falcon_vi_init(ef_vi* vi, void* vvis)
++{
++ struct vi_mappings *vm = (struct vi_mappings*)vvis;
++ uint16_t* ids;
++
++ ef_assert(vi);
++ ef_assert(vvis);
++ ef_assert_equal(vm->signature, VI_MAPPING_SIGNATURE);
++ ef_assert_equal(vm->nic_type.arch, EF_VI_ARCH_FALCON);
++
++ /* Initialise masks to zero, so that ef_vi_state_init() will
++ ** not do any harm when we don't have DMA queues. */
++ vi->vi_rxq.mask = vi->vi_txq.mask = 0;
++
++ /* Used for BUG5391_WORKAROUND. */
++ vi->vi_txq.misalign_mask = 0;
++
++ /* Initialise doorbell addresses to a distinctive small value
++ ** which will cause a segfault, to trap doorbell pushes to VIs
++ ** without DMA queues. */
++ vi->vi_rxq.doorbell = vi->vi_txq.doorbell = (ef_vi_ioaddr_t)0xdb;
++
++ ids = (uint16_t*) (vi->ep_state + 1);
++
++ if( vm->tx_queue_capacity ) {
++ vi->vi_txq.mask = vm->tx_queue_capacity - 1;
++ vi->vi_txq.doorbell = vm->tx_bell + 12;
++ vi->vi_txq.descriptors = vm->tx_dma_falcon;
++ vi->vi_txq.ids = ids;
++ ids += vi->vi_txq.mask + 1;
++ /* Check that the id fifo fits in the space allocated. */
++ ef_assert_le((char*) (vi->vi_txq.ids + vm->tx_queue_capacity),
++ (char*) vi->ep_state
++ + ef_vi_calc_state_bytes(vm->rx_queue_capacity,
++ vm->tx_queue_capacity));
++ }
++ if( vm->rx_queue_capacity ) {
++ vi->vi_rxq.mask = vm->rx_queue_capacity - 1;
++ vi->vi_rxq.doorbell = vm->rx_bell + 12;
++ vi->vi_rxq.descriptors = vm->rx_dma_falcon;
++ vi->vi_rxq.ids = ids;
++ /* Check that the id fifo fits in the space allocated. */
++ ef_assert_le((char*) (vi->vi_rxq.ids + vm->rx_queue_capacity),
++ (char*) vi->ep_state
++ + ef_vi_calc_state_bytes(vm->rx_queue_capacity,
++ vm->tx_queue_capacity));
++ }
++
++ if( vm->nic_type.variant == 'A' ) {
++ vi->vi_txq.misalign_mask = 15; /* BUG5391_WORKAROUND */
++ vi->vi_flags |= EF_VI_BUG5692_WORKAROUND;
++ }
++}
++
++
++int ef_vi_transmitv_init(ef_vi* vi, const ef_iovec* iov, int iov_len,
++ ef_request_id dma_id)
++{
++ ef_vi_txq* q = &vi->vi_txq;
++ ef_vi_txq_state* qs = &vi->ep_state->txq;
++ ef_vi_falcon_dma_tx_buf_desc* dp;
++ unsigned len, dma_len, di;
++ unsigned added_save = qs->added;
++ ef_addr dma_addr;
++ unsigned last_len = 0;
++
++ ef_assert(iov_len > 0);
++ ef_assert(iov);
++ ef_assert_equal((dma_id & EF_REQUEST_ID_MASK), dma_id);
++ ef_assert_nequal(dma_id, 0xffff);
++
++ dma_addr = iov->iov_base;
++ len = iov->iov_len;
++
++ if( vi->vi_flags & EF_VI_ISCSI_TX_DDIG ) {
++ /* Last 4 bytes of placeholder for digest must be
++ * removed for h/w */
++ ef_assert(len > 4);
++ last_len = iov[iov_len - 1].iov_len;
++ if( last_len <= 4 ) {
++ ef_assert(iov_len > 1);
++ --iov_len;
++ last_len = iov[iov_len - 1].iov_len - (4 - last_len);
++ }
++ else {
++ last_len = iov[iov_len - 1].iov_len - 4;
++ }
++ if( iov_len == 1 )
++ len = last_len;
++ }
++
++ while( 1 ) {
++ if( qs->added - qs->removed >= q->mask ) {
++ qs->added = added_save;
++ return -EAGAIN;
++ }
++
++ dma_len = (~((unsigned) dma_addr) & 0xfff) + 1;
++ if( dma_len > len ) dma_len = len;
++ { /* BUG5391_WORKAROUND */
++ unsigned misalign =
++ (unsigned) dma_addr & q->misalign_mask;
++ if( misalign && dma_len + misalign > 512 )
++ dma_len = 512 - misalign;
++ }
++
++ di = qs->added++ & q->mask;
++ dp = (ef_vi_falcon_dma_tx_buf_desc*) q->descriptors + di;
++ if( vi->vi_flags & EF_VI_TX_PHYS_ADDR )
++ falcon_dma_tx_calc_ip_phys
++ (ef_physaddr(dma_addr), dma_len, /*port*/ 0,
++ (iov_len == 1 && dma_len == len) ? 0 :
++ EFVI_FALCON_DMA_TX_FRAG, dp);
++ else
++ falcon_dma_tx_calc_ip_buf
++ (ef_bufaddr(dma_addr), dma_len, /*port*/ 0,
++ (iov_len == 1 && dma_len == len) ? 0 :
++ EFVI_FALCON_DMA_TX_FRAG, dp);
++
++ dma_addr += dma_len;
++ len -= dma_len;
++
++ if( len == 0 ) {
++ if( --iov_len == 0 ) break;
++ ++iov;
++ dma_addr = iov->iov_base;
++ len = iov->iov_len;
++ if( (vi->vi_flags & EF_VI_ISCSI_TX_DDIG) &&
++ (iov_len == 1) )
++ len = last_len;
++ }
++ }
++
++ q->ids[di] = (uint16_t) dma_id;
++ return 0;
++}
++
++
++void ef_vi_transmit_push(ef_vi* vi)
++{
++ ef_vi_wiob();
++ writel((vi->ep_state->txq.added & vi->vi_txq.mask) <<
++ __DW4(TX_DESC_WPTR_LBN),
++ vi->vi_txq.doorbell);
++}
++
++
++/*! The value of initial_rx_bytes is used to set RX_KER_BUF_SIZE in an initial
++** receive descriptor here if physical addressing is being used. A value of
++** zero represents 16384 bytes. This is okay, because caller must provide a
++** buffer than is > MTU, and mac should filter anything bigger than that.
++*/
++int ef_vi_receive_init(ef_vi* vi, ef_addr addr, ef_request_id dma_id,
++ int initial_rx_bytes)
++{
++ ef_vi_rxq* q = &vi->vi_rxq;
++ ef_vi_rxq_state* qs = &vi->ep_state->rxq;
++ unsigned di;
++
++ if( ef_vi_receive_space(vi) ) {
++ di = qs->added++ & q->mask;
++ ef_assert_equal(q->ids[di], 0xffff);
++ q->ids[di] = (uint16_t) dma_id;
++
++ if( ! (vi->vi_flags & EF_VI_RX_PHYS_ADDR) ) {
++ ef_vi_falcon_dma_rx_buf_desc* dp;
++ dp = (ef_vi_falcon_dma_rx_buf_desc*)
++ q->descriptors + di;
++ falcon_dma_rx_calc_ip_buf(ef_bufaddr(addr), dp);
++ }
++ else {
++ ef_vi_falcon_dma_rx_phys_desc* dp;
++ dp = (ef_vi_falcon_dma_rx_phys_desc*)
++ q->descriptors + di;
++ __falcon_dma_rx_calc_ip_phys(addr, dp,
++ initial_rx_bytes);
++ }
++
++ return 0;
++ }
++
++ return -EAGAIN;
++}
++
++
++void ef_vi_receive_push(ef_vi* vi)
++{
++ ef_vi_wiob();
++ writel ((vi->ep_state->rxq.added & vi->vi_rxq.mask) <<
++ __DW4(RX_DESC_WPTR_LBN),
++ vi->vi_rxq.doorbell);
++}
++
++
++ef_request_id ef_vi_receive_done(const ef_vi* vi, const ef_event* ef_ev)
++{
++ const ef_vi_qword* ev = EF_GET_HW_EV_PTR(*ef_ev);
++ unsigned di = ev->u32[0] & vi->vi_rxq.mask;
++ ef_request_id rq_id;
++
++ ef_assert(EF_EVENT_TYPE(*ef_ev) == EF_EVENT_TYPE_RX ||
++ EF_EVENT_TYPE(*ef_ev) == EF_EVENT_TYPE_RX_DISCARD);
++
++ /* Detect spurious / duplicate RX events. We may need to modify this
++ ** code so that we are robust if they happen. */
++ ef_assert_equal(di, vi->ep_state->rxq.removed & vi->vi_rxq.mask);
++
++ /* We only support 1 port: so events should be in order. */
++ ef_assert(vi->vi_rxq.ids[di] != 0xffff);
++
++ rq_id = vi->vi_rxq.ids[di];
++ vi->vi_rxq.ids[di] = 0xffff;
++ ++vi->ep_state->rxq.removed;
++ return rq_id;
++}
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netfront/pt_tx.c b/drivers/xen/sfc_netfront/pt_tx.c
+new file mode 100644
+index 0000000..bdc1f88
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/pt_tx.c
+@@ -0,0 +1,91 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Packet-mode transmit interface.
++ * \date 2003/04/02
++ */
++
++/*! \cidoxg_lib_ef */
++#include "ef_vi_internal.h"
++
++
++int ef_vi_transmit_init(ef_vi* vi, ef_addr base, int len, ef_request_id dma_id)
++{
++ ef_iovec iov = { base, len };
++ return ef_vi_transmitv_init(vi, &iov, 1, dma_id);
++}
++
++
++int ef_vi_transmit(ef_vi* vi, ef_addr base, int len, ef_request_id dma_id)
++{
++ ef_iovec iov = { base, len };
++ int rc = ef_vi_transmitv_init(vi, &iov, 1, dma_id);
++ if( rc == 0 ) ef_vi_transmit_push(vi);
++ return rc;
++}
++
++
++int ef_vi_transmitv(ef_vi* vi, const ef_iovec* iov, int iov_len,
++ ef_request_id dma_id)
++{
++ int rc = ef_vi_transmitv_init(vi, iov, iov_len, dma_id);
++ if( rc == 0 ) ef_vi_transmit_push(vi);
++ return rc;
++}
++
++
++int ef_vi_transmit_unbundle(ef_vi* vi, const ef_event* __ev,
++ ef_request_id* ids)
++{
++ ef_request_id* ids_in = ids;
++ ef_vi_txq* q = &vi->vi_txq;
++ ef_vi_txq_state* qs = &vi->ep_state->txq;
++ const ef_vi_qword* ev = EF_GET_HW_EV_PTR(*__ev);
++ unsigned i, stop = (ev->u32[0] + 1) & q->mask;
++
++ ef_assert(EF_EVENT_TYPE(*__ev) == EF_EVENT_TYPE_TX ||
++ EF_EVENT_TYPE(*__ev) == EF_EVENT_TYPE_TX_ERROR);
++
++ /* Shouldn't be batching more than 64 descriptors, and should not go
++ ** backwards. */
++ ef_assert_le((((ev->u32[0] + 1) - qs->removed) & q->mask), 64);
++ /* Should not complete more than we've posted. */
++ ef_assert_le((((ev->u32[0] + 1) - qs->removed) & q->mask),
++ qs->added - qs->removed);
++
++ for( i = qs->removed & q->mask; i != stop; i = ++qs->removed & q->mask )
++ if( q->ids[i] != 0xffff ) {
++ *ids++ = q->ids[i];
++ q->ids[i] = 0xffff;
++ }
++
++ ef_assert_le(ids - ids_in, EF_VI_TRANSMIT_BATCH);
++
++ return (int) (ids - ids_in);
++}
++
++/*! \cidoxg_end */
+diff --git a/drivers/xen/sfc_netfront/sysdep.h b/drivers/xen/sfc_netfront/sysdep.h
+new file mode 100644
+index 0000000..c5eca8c
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/sysdep.h
+@@ -0,0 +1,184 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author stg
++ * \brief System dependent support for ef vi lib
++ * \date 2007/05/10
++ */
++
++/*! \cidoxg_include_ci_ul */
++#ifndef __CI_CIUL_SYSDEP_LINUX_H__
++#define __CI_CIUL_SYSDEP_LINUX_H__
++
++/**********************************************************************
++ * Kernel version compatability
++ */
++
++#if defined(__GNUC__)
++
++/* Linux kernel doesn't have stdint.h or [u]intptr_t. */
++# if !defined(LINUX_VERSION_CODE)
++# include <linux/version.h>
++# endif
++# include <asm/io.h>
++
++/* In Linux 2.6.24, linux/types.h has uintptr_t */
++# if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
++# if BITS_PER_LONG == 32
++ typedef __u32 uintptr_t;
++# else
++ typedef __u64 uintptr_t;
++# endif
++# endif
++
++/* But even 2.6.24 doesn't define intptr_t */
++# if BITS_PER_LONG == 32
++ typedef __s32 intptr_t;
++# else
++ typedef __s64 intptr_t;
++# endif
++
++# if defined(__ia64__)
++# define EF_VI_PRIx64 "lx"
++# else
++# define EF_VI_PRIx64 "llx"
++# endif
++
++# define EF_VI_HF __attribute__((visibility("hidden")))
++# define EF_VI_HV __attribute__((visibility("hidden")))
++
++# if defined(__i386__) || defined(__x86_64__) /* GCC x86/x64 */
++ typedef unsigned long long ef_vi_dma_addr_t;
++# if __GNUC__ >= 3 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 96)
++# define ef_vi_wiob() __asm__ __volatile__ ("sfence")
++# else
++# define ef_vi_wiob() __asm__ __volatile__ (".byte 0x0F, 0xAE, 0xF8")
++# endif
++
++# endif
++#endif
++
++#ifdef EFX_NOT_UPSTREAM
++
++/* Stuff for architectures/compilers not officially supported */
++
++#if !defined(__GNUC__)
++# if defined(__PPC__) /* GCC, PPC */
++ typedef unsigned long ef_vi_dma_addr_t;
++# define ef_vi_wiob() wmb()
++
++# ifdef __powerpc64__
++# ifdef CONFIG_SMP
++# define CI_SMP_SYNC "\n eieio \n" /* memory cache sync */
++# define CI_SMP_ISYNC "\n isync \n" /* instr cache sync */
++# else
++# define CI_SMP_SYNC
++# define CI_SMP_ISYNC
++# endif
++# else /* for ppc32 systems */
++# ifdef CONFIG_SMP
++# define CI_SMP_SYNC "\n eieio \n"
++# define CI_SMP_ISYNC "\n sync \n"
++# else
++# define CI_SMP_SYNC
++# define CI_SMP_ISYNC
++# endif
++# endif
++
++# elif defined(__ia64__) /* GCC, IA64 */
++ typedef unsigned long ef_vi_dma_addr_t;
++# define ef_vi_wiob() __asm__ __volatile__("mf.a": : :"memory")
++
++# else
++# error Unknown processor - GNU C
++# endif
++
++#elif defined(__PGI)
++# error PGI not supported
++
++#elif defined(__INTEL_COMPILER)
++
++/* Intel compilers v7 claim to be very gcc compatible. */
++# if __INTEL_COMPILER >= 700
++# if __GNUC__ >= 3 || (__GNUC__ == 2 && __GNUC_MINOR__ > 91)
++# define EF_VI_LIKELY(t) __builtin_expect((t), 1)
++# define EF_VI_UNLIKELY(t) __builtin_expect((t), 0)
++# endif
++
++# if __GNUC__ >= 3 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 96)
++# define ef_vi_wiob() __asm__ __volatile__ ("sfence")
++# else
++# define ef_vi_wiob() __asm__ __volatile__ (".byte 0x0F, 0xAE, 0xF8")
++# endif
++
++# else
++# error Old Intel compiler not supported.
++# endif
++
++#else
++# error Unknown compiler.
++#endif
++
++#endif
++
++
++# include <linux/errno.h>
++
++
++/**********************************************************************
++ * Extracting bit fields.
++ */
++
++#define _QWORD_GET_LOW(f, v) \
++ (((v).u32[0] >> (f##_LBN)) & ((1u << f##_WIDTH) - 1u))
++#define _QWORD_GET_HIGH(f, v) \
++ (((v).u32[1] >> (f##_LBN - 32u)) & ((1u << f##_WIDTH) - 1u))
++#define _QWORD_GET_ANY(f, v) \
++ (((v).u64[0] >> f##_LBN) & (((uint64_t) 1u << f##_WIDTH) - 1u))
++
++#define QWORD_GET(f, v) \
++ ((f##_LBN + f##_WIDTH) <= 32u \
++ ? _QWORD_GET_LOW(f, (v)) \
++ : ((f##_LBN >= 32u) ? _QWORD_GET_HIGH(f, (v)) : _QWORD_GET_ANY(f, (v))))
++
++#define QWORD_GET_U(f, v) ((unsigned) QWORD_GET(f, (v)))
++
++#define _QWORD_TEST_BIT_LOW(f, v) ((v).u32[0] & (1u << (f##_LBN)))
++#define _QWORD_TEST_BIT_HIGH(f, v) ((v).u32[1] & (1u << (f##_LBN - 32u)))
++
++#define QWORD_TEST_BIT(f, v) \
++ (f##_LBN < 32 ? _QWORD_TEST_BIT_LOW(f, (v)) : _QWORD_TEST_BIT_HIGH(f, (v)))
++
++
++
++
++#ifndef DECLSPEC_NORETURN
++/* normally defined on Windows to expand to a declaration that the
++ function will not return */
++# define DECLSPEC_NORETURN
++#endif
++
++#endif /* __CI_CIUL_SYSDEP_LINUX_H__ */
+diff --git a/drivers/xen/sfc_netfront/vi_init.c b/drivers/xen/sfc_netfront/vi_init.c
+new file mode 100644
+index 0000000..4e7e19b
+--- /dev/null
++++ b/drivers/xen/sfc_netfront/vi_init.c
+@@ -0,0 +1,183 @@
++/****************************************************************************
++ * Copyright 2002-2005: Level 5 Networks Inc.
++ * Copyright 2005-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications
++ * <linux-xen-drivers at solarflare.com>
++ * <onload-dev at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * \author djr
++ * \brief Initialisation of VIs.
++ * \date 2007/06/08
++ */
++
++#include "ef_vi_internal.h"
++
++#define EF_VI_STATE_BYTES(rxq_sz, txq_sz) \
++ (sizeof(ef_vi_state) + (rxq_sz) * sizeof(uint16_t) \
++ + (txq_sz) * sizeof(uint16_t))
++
++int ef_vi_calc_state_bytes(int rxq_sz, int txq_sz)
++{
++ ef_assert(rxq_sz == 0 || EF_VI_IS_POW2(rxq_sz));
++ ef_assert(txq_sz == 0 || EF_VI_IS_POW2(txq_sz));
++
++ return EF_VI_STATE_BYTES(rxq_sz, txq_sz);
++}
++
++
++int ef_vi_state_bytes(ef_vi* vi)
++{
++ int rxq_sz = 0, txq_sz = 0;
++ if( ef_vi_receive_capacity(vi) )
++ rxq_sz = ef_vi_receive_capacity(vi) + 1;
++ if( ef_vi_transmit_capacity(vi) )
++ txq_sz = ef_vi_transmit_capacity(vi) + 1;
++
++ ef_assert(rxq_sz == 0 || EF_VI_IS_POW2(rxq_sz));
++ ef_assert(txq_sz == 0 || EF_VI_IS_POW2(txq_sz));
++
++ return EF_VI_STATE_BYTES(rxq_sz, txq_sz);
++}
++
++
++void ef_eventq_state_init(ef_vi* evq)
++{
++ int j;
++
++ for (j = 0; j<EFAB_DMAQS_PER_EVQ_MAX; j++) {
++ ef_rx_dup_state_t *rx_dup_state =
++ &evq->evq_state->rx_dup_state[j];
++ rx_dup_state->bad_sop = 0;
++ rx_dup_state->rx_last_desc_ptr = -1;
++ rx_dup_state->frag_num = 0;
++ }
++
++ evq->evq_state->evq_ptr = 0;
++}
++
++
++void ef_vi_state_init(ef_vi* vi)
++{
++ ef_vi_state* state = vi->ep_state;
++ unsigned i;
++
++ state->txq.added = state->txq.removed = 0;
++ state->rxq.added = state->rxq.removed = 0;
++
++ if( vi->vi_rxq.mask )
++ for( i = 0; i <= vi->vi_rxq.mask; ++i )
++ vi->vi_rxq.ids[i] = (uint16_t) -1;
++ if( vi->vi_txq.mask )
++ for( i = 0; i <= vi->vi_txq.mask; ++i )
++ vi->vi_txq.ids[i] = (uint16_t) -1;
++}
++
++
++void ef_vi_init_mapping_evq(void* data_area, struct ef_vi_nic_type nic_type,
++ int instance, unsigned evq_bytes, void* base,
++ void* timer_reg)
++{
++ struct vi_mappings* vm = (struct vi_mappings*) data_area;
++
++ vm->signature = VI_MAPPING_SIGNATURE;
++ vm->vi_instance = instance;
++ vm->nic_type = nic_type;
++ vm->evq_bytes = evq_bytes;
++ vm->evq_base = base;
++ vm->evq_timer_reg = timer_reg;
++}
++
++
++void ef_vi_init(ef_vi* vi, void* vvis, ef_vi_state* state,
++ ef_eventq_state* evq_state, enum ef_vi_flags vi_flags)
++{
++ struct vi_mappings* vm = (struct vi_mappings*) vvis;
++
++ vi->vi_i = vm->vi_instance;
++ vi->ep_state = state;
++ vi->vi_flags = vi_flags;
++
++ switch( vm->nic_type.arch ) {
++ case EF_VI_ARCH_FALCON:
++ falcon_vi_init(vi, vvis);
++ break;
++ default:
++ /* ?? TODO: We should return an error code. */
++ ef_assert(0);
++ break;
++ }
++
++ if( vm->evq_bytes ) {
++ vi->evq_state = evq_state;
++ vi->evq_mask = vm->evq_bytes - 1u;
++ vi->evq_base = vm->evq_base;
++ vi->evq_timer_reg = vm->evq_timer_reg;
++ }
++
++ EF_VI_MAGIC_SET(vi, EF_VI);
++}
++
++
++/* Initialise [data_area] with information required to initialise an ef_vi.
++ * In the following, an unused param should be set to NULL. Note the case
++ * marked (*) of [iobuf_mmap] for falcon/driver; for the normal driver this
++ * must be NULL.
++ *
++ * \param data_area [in,out] required, must ref at least VI_MAPPING_SIZE
++ * bytes
++ * \param io_mmap [in] ef1, required
++ * falcon, required
++ * \param iobuf_mmap [in] ef1, unused
++ * falcon, required
++ */
++void ef_vi_init_mapping_vi(void* data_area, struct ef_vi_nic_type nic_type,
++ unsigned rxq_capacity, unsigned txq_capacity,
++ int instance, void* io_mmap,
++ void* iobuf_mmap_rx, void* iobuf_mmap_tx,
++ enum ef_vi_flags vi_flags)
++{
++ struct vi_mappings* vm = (struct vi_mappings*) data_area;
++ int rx_desc_bytes, rxq_bytes;
++
++ ef_assert(rxq_capacity > 0 || txq_capacity > 0);
++ ef_assert(vm);
++ ef_assert(io_mmap);
++ ef_assert(iobuf_mmap_rx || iobuf_mmap_tx);
++
++ vm->signature = VI_MAPPING_SIGNATURE;
++ vm->vi_instance = instance;
++ vm->nic_type = nic_type;
++
++ rx_desc_bytes = (vi_flags & EF_VI_RX_PHYS_ADDR) ? 8 : 4;
++ rxq_bytes = rxq_capacity * rx_desc_bytes;
++ rxq_bytes = (rxq_bytes + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1);
++
++ if( iobuf_mmap_rx == iobuf_mmap_tx )
++ iobuf_mmap_tx = (char*) iobuf_mmap_rx + rxq_bytes;
++
++ vm->rx_queue_capacity = rxq_capacity;
++ vm->rx_dma_falcon = iobuf_mmap_rx;
++ vm->rx_bell = (char*) io_mmap + (RX_DESC_UPD_REG_KER_OFST & 4095);
++ vm->tx_queue_capacity = txq_capacity;
++ vm->tx_dma_falcon = iobuf_mmap_tx;
++ vm->tx_bell = (char*) io_mmap + (TX_DESC_UPD_REG_KER_OFST & 4095);
++}
+diff --git a/drivers/xen/sfc_netutil/Makefile b/drivers/xen/sfc_netutil/Makefile
+new file mode 100644
+index 0000000..3fce370
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/Makefile
+@@ -0,0 +1,11 @@
++EXTRA_CFLAGS += -Idrivers/xen/sfc_netutil
++EXTRA_CFLAGS += -Werror
++
++ifdef GGOV
++EXTRA_CFLAGS += -fprofile-arcs -ftest-coverage -DEFX_GCOV
++endif
++
++obj-$(CONFIG_XEN_NETDEV_ACCEL_SFC_UTIL) := sfc_netutil.o
++
++sfc_netutil-objs := accel_cuckoo_hash.o accel_msg_iface.o accel_util.o
++
+diff --git a/drivers/xen/sfc_netutil/accel_cuckoo_hash.c b/drivers/xen/sfc_netutil/accel_cuckoo_hash.c
+new file mode 100644
+index 0000000..d738968
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_cuckoo_hash.c
+@@ -0,0 +1,651 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/types.h> /* needed for linux/random.h */
++#include <linux/random.h>
++
++#include "accel_cuckoo_hash.h"
++#include "accel_util.h"
++
++static inline int cuckoo_hash_key_compare(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key1,
++ cuckoo_hash_key *key2)
++{
++ return !memcmp(key1, key2, hashtab->key_length);
++}
++
++
++static inline void cuckoo_hash_key_set(cuckoo_hash_key *key1,
++ cuckoo_hash_key *key2)
++{
++ *key1 = *key2;
++}
++
++
++/*
++ * Sets hash function parameters. Chooses "a" to be odd, 0 < a < 2^w
++ * where w is the length of the key
++ */
++static void set_hash_parameters(cuckoo_hash_table *hashtab)
++{
++ again:
++ hashtab->a0 = hashtab->a1 = 0;
++
++ /* Make sure random */
++ get_random_bytes(&hashtab->a0, hashtab->key_length);
++ get_random_bytes(&hashtab->a1, hashtab->key_length);
++
++ /* Make sure odd */
++ hashtab->a0 |= 1;
++ hashtab->a1 |= 1;
++
++ /* Being different is good */
++ if (hashtab->a0 != hashtab->a1)
++ return;
++
++ goto again;
++}
++
++int cuckoo_hash_init(cuckoo_hash_table *hashtab, unsigned length_bits,
++ unsigned key_length)
++{
++ char *table_mem;
++ unsigned length = 1 << length_bits;
++
++ BUG_ON(length_bits >= sizeof(unsigned) * 8);
++ BUG_ON(key_length > sizeof(cuckoo_hash_key));
++
++ table_mem = kmalloc(sizeof(cuckoo_hash_entry) * 2 * length, GFP_KERNEL);
++
++ if (table_mem == NULL)
++ return -ENOMEM;
++
++ hashtab->length = length;
++ hashtab->length_bits = length_bits;
++ hashtab->key_length = key_length;
++ hashtab->entries = 0;
++
++ hashtab->table0 = (cuckoo_hash_entry *)table_mem;
++ hashtab->table1 = (cuckoo_hash_entry *)
++ (table_mem + length * sizeof(cuckoo_hash_entry));
++
++ set_hash_parameters(hashtab);
++
++ /* Zero the table */
++ memset(hashtab->table0, 0, length * 2 * sizeof(cuckoo_hash_entry));
++
++ return 0;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_init);
++
++void cuckoo_hash_destroy(cuckoo_hash_table *hashtab)
++{
++ if (hashtab->table0 != NULL)
++ kfree(hashtab->table0);
++}
++
++EXPORT_SYMBOL_GPL(cuckoo_hash_destroy);
++
++/*
++ * This computes sizeof(cuckoo_hash) bits of hash, not all will be
++ * necessarily used, but the hash function throws away any that
++ * aren't
++ */
++static inline void cuckoo_compute_hash_helper(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *a,
++ cuckoo_hash_key *x,
++ cuckoo_hash *result)
++{
++ u64 multiply_result = 0, a_temp, x_temp;
++ u32 carry = 0;
++ u32 *a_words;
++ u32 *x_words;
++ int i;
++
++ /*
++ * As the mod and div operations in the function effectively
++ * reduce and shift the bits of the product down to just the
++ * third word, we need only compute that and return it as a
++ * result.
++ *
++ * Do enough long multiplication to get the word we need
++ */
++
++ /* This assumes things about the sizes of the key and hash */
++ BUG_ON(hashtab->key_length % sizeof(u32) != 0);
++ BUG_ON(sizeof(cuckoo_hash) != sizeof(u32));
++
++ a_words = (u32 *)a;
++ x_words = (u32 *)x;
++
++ for (i = 0; i < hashtab->key_length / sizeof(u32); i++) {
++ a_temp = a_words[i];
++ x_temp = x_words[i];
++
++ multiply_result = (a_temp * x_temp) + carry;
++ carry = (multiply_result >> 32) & 0xffffffff;
++ }
++
++ *result = multiply_result & 0xffffffff;
++}
++
++
++/*
++ * Want to implement (ax mod 2^w) div 2^(w-q) for odd a, 0 < a < 2^w;
++ * w is the length of the key, q is the length of the hash, I think.
++ * See http://www.it-c.dk/people/pagh/papers/cuckoo-jour.pdf
++ */
++static cuckoo_hash cuckoo_compute_hash(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_key *a)
++{
++ unsigned q = hashtab->length_bits;
++ unsigned shift = 32 - q;
++ unsigned mask = ((1 << q) - 1) << shift;
++ cuckoo_hash hash;
++
++ cuckoo_compute_hash_helper(hashtab, a, key, &hash);
++
++ /*
++ * Take the top few bits to get the right length for this
++ * hash table
++ */
++ hash = (hash & mask) >> shift;
++
++ BUG_ON(hash >= hashtab->length);
++
++ return hash;
++}
++
++
++static int cuckoo_hash_lookup0(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value *value)
++{
++ cuckoo_hash hash = cuckoo_compute_hash(hashtab, key, &hashtab->a0);
++
++ if ((hashtab->table0[hash].state == CUCKOO_HASH_STATE_OCCUPIED)
++ && cuckoo_hash_key_compare(hashtab, &(hashtab->table0[hash].key),
++ key)) {
++ *value = hashtab->table0[hash].value;
++ return 1;
++ }
++
++ return 0;
++}
++
++static int cuckoo_hash_lookup1(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value *value)
++{
++ cuckoo_hash hash = cuckoo_compute_hash(hashtab, key, &hashtab->a1);
++
++ if ((hashtab->table1[hash].state == CUCKOO_HASH_STATE_OCCUPIED)
++ && cuckoo_hash_key_compare(hashtab, &(hashtab->table1[hash].key),
++ key)) {
++ *value = hashtab->table1[hash].value;
++ return 1;
++ }
++
++ return 0;
++}
++
++
++int cuckoo_hash_lookup(cuckoo_hash_table *hashtab, cuckoo_hash_key *key,
++ cuckoo_hash_value *value)
++{
++ return cuckoo_hash_lookup0(hashtab, key, value)
++ || cuckoo_hash_lookup1(hashtab, key, value);
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_lookup);
++
++
++/* Transfer any active entries from "old_table" into hashtab */
++static int cuckoo_hash_transfer_entries(cuckoo_hash_table *hashtab,
++ cuckoo_hash_entry *old_table,
++ unsigned capacity)
++{
++ int i, rc;
++ cuckoo_hash_entry *entry;
++
++ hashtab->entries = 0;
++
++ for (i = 0; i < capacity; i++) {
++ entry = &old_table[i];
++ if (entry->state == CUCKOO_HASH_STATE_OCCUPIED) {
++ rc = cuckoo_hash_add(hashtab, &(entry->key),
++ entry->value, 0);
++ if (rc != 0) {
++ return rc;
++ }
++ }
++ }
++
++ return 0;
++}
++
++
++int cuckoo_hash_rehash(cuckoo_hash_table *hashtab)
++{
++ cuckoo_hash_entry *new_table;
++ cuckoo_hash_table old_hashtab;
++ int resize = 0, rc, rehash_count;
++
++ /*
++ * Store old tables so we can access the existing values and
++ * copy across
++ */
++ memcpy(&old_hashtab, hashtab, sizeof(cuckoo_hash_table));
++
++ /* resize if hashtable is more than half full */
++ if (old_hashtab.entries > old_hashtab.length &&
++ old_hashtab.length_bits < 32)
++ resize = 1;
++
++ resize:
++ if (resize) {
++ new_table = kmalloc(sizeof(cuckoo_hash_entry) * 4 * hashtab->length,
++ GFP_ATOMIC);
++ if (new_table == NULL) {
++ rc = -ENOMEM;
++ goto err;
++ }
++
++ hashtab->length = 2 * hashtab->length;
++ hashtab->length_bits++;
++ } else {
++ new_table = kmalloc(sizeof(cuckoo_hash_entry) * 2 * hashtab->length,
++ GFP_ATOMIC);
++ if (new_table == NULL) {
++ rc = -ENOMEM;
++ goto err;
++ }
++ }
++
++ /*
++ * Point hashtab to new memory region so we can try to
++ * construct new table
++ */
++ hashtab->table0 = new_table;
++ hashtab->table1 = (cuckoo_hash_entry *)
++ ((char *)new_table + hashtab->length * sizeof(cuckoo_hash_entry));
++
++ rehash_count = 0;
++
++ again:
++ /* Zero the new tables */
++ memset(new_table, 0, hashtab->length * 2 * sizeof(cuckoo_hash_entry));
++
++ /* Choose new parameters for the hash functions */
++ set_hash_parameters(hashtab);
++
++ /*
++ * Multiply old_table_length by 2 as the length refers to each
++ * table, and there are two of them. This assumes that they
++ * are arranged sequentially in memory, so assert it
++ */
++ BUG_ON(((char *)old_hashtab.table1) !=
++ ((char *)old_hashtab.table0 + old_hashtab.length
++ * sizeof(cuckoo_hash_entry)));
++ rc = cuckoo_hash_transfer_entries(hashtab, old_hashtab.table0,
++ old_hashtab.length * 2);
++ if (rc < 0) {
++ /* Problem */
++ if (rc == -ENOSPC) {
++ ++rehash_count;
++ if (rehash_count < CUCKOO_HASH_MAX_LOOP) {
++ /*
++ * Wanted to rehash, but rather than
++ * recurse we can just do it here
++ */
++ goto again;
++ } else {
++ /*
++ * Didn't manage to rehash, so let's
++ * go up a size (if we haven't already
++ * and there's space)
++ */
++ if (!resize && hashtab->length_bits < 32) {
++ resize = 1;
++ kfree(new_table);
++ goto resize;
++ }
++ else
++ goto err;
++ }
++ }
++ else
++ goto err;
++ }
++
++ /* Success, I think. Free up the old table */
++ kfree(old_hashtab.table0);
++
++ /* We should have put all the entries from old table in the new one */
++ BUG_ON(hashtab->entries != old_hashtab.entries);
++
++ return 0;
++ err:
++ EPRINTK("%s: Rehash failed, giving up\n", __FUNCTION__);
++ /* Some other error, give up, at least restore table to how it was */
++ memcpy(hashtab, &old_hashtab, sizeof(cuckoo_hash_table));
++ if (new_table)
++ kfree(new_table);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_rehash);
++
++
++static int
++cuckoo_hash_insert_or_displace(cuckoo_hash_entry *table, unsigned hash,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value value,
++ cuckoo_hash_key *displaced_key,
++ cuckoo_hash_value *displaced_value)
++{
++ if (table[hash].state == CUCKOO_HASH_STATE_VACANT) {
++ cuckoo_hash_key_set(&(table[hash].key), key);
++ table[hash].value = value;
++ table[hash].state = CUCKOO_HASH_STATE_OCCUPIED;
++
++ return 1;
++ } else {
++ cuckoo_hash_key_set(displaced_key, &(table[hash].key));
++ *displaced_value = table[hash].value;
++ cuckoo_hash_key_set(&(table[hash].key), key);
++ table[hash].value = value;
++
++ return 0;
++ }
++}
++
++
++int cuckoo_hash_add(cuckoo_hash_table *hashtab, cuckoo_hash_key *key,
++ cuckoo_hash_value value, int can_rehash)
++{
++ cuckoo_hash hash0, hash1;
++ int i, rc;
++ cuckoo_hash_key key1, key2;
++
++ cuckoo_hash_key_set(&key1, key);
++
++ again:
++ i = 0;
++ do {
++ hash0 = cuckoo_compute_hash(hashtab, &key1, &hashtab->a0);
++ if (cuckoo_hash_insert_or_displace(hashtab->table0, hash0,
++ &key1, value, &key2,
++ &value)) {
++ /* Success */
++ hashtab->entries++;
++ return 0;
++ }
++
++ hash1 = cuckoo_compute_hash(hashtab, &key2, &hashtab->a1);
++ if (cuckoo_hash_insert_or_displace(hashtab->table1, hash1,
++ &key2, value, &key1,
++ &value)) {
++ /* Success */
++ hashtab->entries++;
++ return 0;
++ }
++ } while (++i < CUCKOO_HASH_MAX_LOOP);
++
++ if (can_rehash) {
++ if ((rc = cuckoo_hash_rehash(hashtab)) < 0) {
++ /*
++ * Give up - this will drop whichever
++ * key/value pair we have currently displaced
++ * on the floor
++ */
++ return rc;
++ }
++ goto again;
++ }
++
++ EPRINTK("%s: failed hash add\n", __FUNCTION__);
++ /*
++ * Couldn't do it - bad as we've now removed some random thing
++ * from the table, and will just drop it on the floor. Better
++ * would be to somehow revert the table to the state it was in
++ * at the start
++ */
++ return -ENOSPC;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_add);
++
++
++int cuckoo_hash_add_check(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key, cuckoo_hash_value value,
++ int can_rehash)
++{
++ int stored_value;
++
++ if (cuckoo_hash_lookup(hashtab, key, &stored_value))
++ return -EBUSY;
++
++ return cuckoo_hash_add(hashtab, key, value, can_rehash);
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_add_check);
++
++
++int cuckoo_hash_remove(cuckoo_hash_table *hashtab, cuckoo_hash_key *key)
++{
++ cuckoo_hash hash;
++
++ hash = cuckoo_compute_hash(hashtab, key, &hashtab->a0);
++ if ((hashtab->table0[hash].state == CUCKOO_HASH_STATE_OCCUPIED) &&
++ cuckoo_hash_key_compare(hashtab, &(hashtab->table0[hash].key),
++ key)) {
++ hashtab->table0[hash].state = CUCKOO_HASH_STATE_VACANT;
++ hashtab->entries--;
++ return 0;
++ }
++
++ hash = cuckoo_compute_hash(hashtab, key, &hashtab->a1);
++ if ((hashtab->table1[hash].state == CUCKOO_HASH_STATE_OCCUPIED) &&
++ cuckoo_hash_key_compare(hashtab, &(hashtab->table1[hash].key),
++ key)) {
++ hashtab->table1[hash].state = CUCKOO_HASH_STATE_VACANT;
++ hashtab->entries--;
++ return 0;
++ }
++
++ return -EINVAL;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_remove);
++
++
++int cuckoo_hash_update(cuckoo_hash_table *hashtab, cuckoo_hash_key *key,
++ cuckoo_hash_value value)
++{
++ cuckoo_hash hash;
++
++ hash = cuckoo_compute_hash(hashtab, key, &hashtab->a0);
++ if ((hashtab->table0[hash].state == CUCKOO_HASH_STATE_OCCUPIED) &&
++ cuckoo_hash_key_compare(hashtab, &(hashtab->table0[hash].key),
++ key)) {
++ hashtab->table0[hash].value = value;
++ return 0;
++ }
++
++ hash = cuckoo_compute_hash(hashtab, key, &hashtab->a1);
++ if ((hashtab->table1[hash].state == CUCKOO_HASH_STATE_OCCUPIED) &&
++ cuckoo_hash_key_compare(hashtab, &(hashtab->table1[hash].key),
++ key)) {
++ hashtab->table1[hash].value = value;
++ return 0;
++ }
++
++ return -EINVAL;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_update);
++
++
++void cuckoo_hash_iterate_reset(cuckoo_hash_table *hashtab)
++{
++ hashtab->iterate_index = 0;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_iterate_reset);
++
++
++int cuckoo_hash_iterate(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key, cuckoo_hash_value *value)
++{
++ unsigned index;
++
++ while (hashtab->iterate_index < hashtab->length) {
++ index = hashtab->iterate_index;
++ ++hashtab->iterate_index;
++ if (hashtab->table0[index].state == CUCKOO_HASH_STATE_OCCUPIED) {
++ *key = hashtab->table0[index].key;
++ *value = hashtab->table0[index].value;
++ return 0;
++ }
++ }
++
++ while (hashtab->iterate_index >= hashtab->length &&
++ hashtab->iterate_index < hashtab->length * 2) {
++ index = hashtab->iterate_index - hashtab->length;
++ ++hashtab->iterate_index;
++ if (hashtab->table1[index].state == CUCKOO_HASH_STATE_OCCUPIED) {
++ *key = hashtab->table1[index].key;
++ *value = hashtab->table1[index].value;
++ return 0;
++ }
++ }
++
++ return -ENOSPC;
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_iterate);
++
++
++#if 0
++void cuckoo_hash_valid(cuckoo_hash_table *hashtab)
++{
++ int i, entry_count = 0;
++
++ for (i=0; i < hashtab->length; i++) {
++ EPRINTK_ON(hashtab->table0[i].state != CUCKOO_HASH_STATE_VACANT &&
++ hashtab->table0[i].state != CUCKOO_HASH_STATE_OCCUPIED);
++ if (hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ entry_count++;
++ EPRINTK_ON(hashtab->table1[i].state != CUCKOO_HASH_STATE_VACANT &&
++ hashtab->table1[i].state != CUCKOO_HASH_STATE_OCCUPIED);
++ if (hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ entry_count++;
++ }
++
++ if (entry_count != hashtab->entries) {
++ EPRINTK("%s: bad count\n", __FUNCTION__);
++ cuckoo_hash_dump(hashtab);
++ return;
++ }
++
++ for (i=0; i< hashtab->length; i++) {
++ if (hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ if (i != cuckoo_compute_hash(hashtab,
++ &hashtab->table0[i].key,
++ &hashtab->a0)) {
++ EPRINTK("%s: Bad key table 0 index %d\n",
++ __FUNCTION__, i);
++ cuckoo_hash_dump(hashtab);
++ return;
++ }
++ if (hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ if (i != cuckoo_compute_hash(hashtab,
++ &hashtab->table1[i].key,
++ &hashtab->a1)) {
++ EPRINTK("%s: Bad key table 1 index %d\n",
++ __FUNCTION__, i);
++ cuckoo_hash_dump(hashtab);
++ return;
++ }
++ }
++
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_valid);
++
++
++void cuckoo_hash_dump(cuckoo_hash_table *hashtab)
++{
++ int i, entry_count;
++
++ entry_count = 0;
++ for (i=0; i < hashtab->length; i++) {
++ EPRINTK_ON(hashtab->table0[i].state != CUCKOO_HASH_STATE_VACANT &&
++ hashtab->table0[i].state != CUCKOO_HASH_STATE_OCCUPIED);
++ if (hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ entry_count++;
++ EPRINTK_ON(hashtab->table1[i].state != CUCKOO_HASH_STATE_VACANT &&
++ hashtab->table1[i].state != CUCKOO_HASH_STATE_OCCUPIED);
++ if (hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ entry_count++;
++ }
++
++ EPRINTK("======================\n");
++ EPRINTK("Cuckoo hash table dump\n");
++ EPRINTK("======================\n");
++ EPRINTK("length: %d; length_bits: %d; key_length: %d\n", hashtab->length,
++ hashtab->length_bits, hashtab->key_length);
++ EPRINTK("Recorded entries: %d\n", hashtab->entries);
++ EPRINTK("Counted entries: %d\n", entry_count);
++ EPRINTK("a0: %llx; a1: %llx\n", hashtab->a0, hashtab->a1);
++ EPRINTK("-----------------------------------------\n");
++ EPRINTK("Index Occupied Key Value Index0 Index1\n");
++ EPRINTK("-----------------------------------------\n");
++ for (i=0; i< hashtab->length; i++) {
++ if (hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ EPRINTK("%d %d %llx %d %d %d\n", i,
++ hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED,
++ hashtab->table0[i].key, hashtab->table0[i].value,
++ cuckoo_compute_hash(hashtab, &hashtab->table0[i].key,
++ &hashtab->a0),
++ cuckoo_compute_hash(hashtab, &hashtab->table0[i].key,
++ &hashtab->a1));
++ else
++ EPRINTK("%d %d - - - -\n", i,
++ hashtab->table0[i].state == CUCKOO_HASH_STATE_OCCUPIED);
++
++ }
++ EPRINTK("-----------------------------------------\n");
++ EPRINTK("Index Occupied Key Value Index0 Index1\n");
++ EPRINTK("-----------------------------------------\n");
++ for (i=0; i< hashtab->length; i++) {
++ if (hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED)
++ EPRINTK("%d %d %llx %d %d %d\n", i,
++ hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED,
++ hashtab->table1[i].key, hashtab->table1[i].value,
++ cuckoo_compute_hash(hashtab, &hashtab->table1[i].key,
++ &hashtab->a0),
++ cuckoo_compute_hash(hashtab, &hashtab->table1[i].key,
++ &hashtab->a1));
++ else
++ EPRINTK("%d %d - - - -\n", i,
++ hashtab->table1[i].state == CUCKOO_HASH_STATE_OCCUPIED);
++ }
++ EPRINTK("======================\n");
++}
++EXPORT_SYMBOL_GPL(cuckoo_hash_dump);
++#endif
+diff --git a/drivers/xen/sfc_netutil/accel_cuckoo_hash.h b/drivers/xen/sfc_netutil/accel_cuckoo_hash.h
+new file mode 100644
+index 0000000..83518f9
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_cuckoo_hash.h
+@@ -0,0 +1,227 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++/*
++ * A cuckoo hash table consists of two sub tables. Each entry can
++ * hash to a position in each table. If, on entry, its position is
++ * found to be occupied, the existing element is moved to it's other
++ * location. This recurses until success or a loop is found. If a
++ * loop is found the table is rehashed.
++ *
++ * See http://www.it-c.dk/people/pagh/papers/cuckoo-jour.pdf
++ */
++
++#ifndef NET_ACCEL_CUCKOO_HASH_H
++#define NET_ACCEL_CUCKOO_HASH_H
++
++/*! Type used for hash table keys of ip pairs */
++typedef struct {
++ u32 local_ip;
++ //u32 remote_ip;
++ u16 local_port;
++ //u16 remote_port;
++ /* Technically only 1 bit, but use 16 to make key a round
++ number size */
++ u16 proto;
++} cuckoo_hash_ip_key;
++
++/*! Type used for hash table keys of mac addresses */
++typedef u64 cuckoo_hash_mac_key;
++
++/*! This type is designed to be large enough to hold all supported key
++ * sizes to avoid having to malloc storage for them.
++ */
++typedef u64 cuckoo_hash_key;
++
++/*! Type used for the values stored in the hash table */
++typedef int cuckoo_hash_value;
++
++/*! Type used for the hash used to index the table */
++typedef u32 cuckoo_hash;
++
++/*! How long to spend displacing values when adding before giving up
++ * and rehashing */
++#define CUCKOO_HASH_MAX_LOOP (hashtab->length)
++
++/*! State of hash table entry */
++typedef enum {
++ CUCKOO_HASH_STATE_VACANT = 0,
++ CUCKOO_HASH_STATE_OCCUPIED
++} cuckoo_hash_state;
++
++/*! An entry in the hash table */
++typedef struct {
++ cuckoo_hash_state state;
++ cuckoo_hash_key key;
++ cuckoo_hash_value value;
++} cuckoo_hash_entry;
++
++/*! A cuckoo hash table */
++typedef struct {
++ /*! The length of each table (NB. there are two tables of this
++ * length) */
++ unsigned length;
++ /*! The length of each table in bits */
++ unsigned length_bits;
++ /*! The length of the key in bytes */
++ unsigned key_length;
++ /*! The number of entries currently stored in the table */
++ unsigned entries;
++ /*! Index into table used by cuckoo_hash_iterate */
++ unsigned iterate_index;
++
++ /* parameter of hash functions */
++ /*! The "a" parameter of the first hash function */
++ cuckoo_hash_key a0;
++ /*! The "a" parameter of the second hash function */
++ cuckoo_hash_key a1;
++
++ /*! The first table */
++ cuckoo_hash_entry *table0;
++ /*! The second table */
++ cuckoo_hash_entry *table1;
++} cuckoo_hash_table;
++
++/*! Initialise the cuckoo has table
++ *
++ * \param hashtab A pointer to an unitialised hash table structure
++ * \param length_bits The number of elements in each table equals
++ * 2**length_bits
++ * \param key_length The length of the key in bytes
++ *
++ * \return 0 on success, -ENOMEM if it couldn't allocate the tables
++ */
++extern
++int cuckoo_hash_init(cuckoo_hash_table *hashtab, unsigned length_bits,
++ unsigned key_length);
++
++
++/*! Destroy a hash table
++ *
++ * \param hashtab A hash table that has previously been passed to a
++ * successful call of cuckoo_hash_init()
++ */
++extern
++void cuckoo_hash_destroy(cuckoo_hash_table *hashtab);
++
++
++/*! Lookup an entry in the hash table
++ *
++ * \param hashtab The hash table in which to look.
++ * \param key Pointer to a mac address to use as the key
++ * \param value On exit set to the value stored if key was present
++ *
++ * \return 0 if not present in the table, non-zero if it is (and value
++ * is set accordingly)
++ */
++extern
++int cuckoo_hash_lookup(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value *value);
++
++/*! Add an entry to the hash table. Key must not be a duplicate of
++ * anything already in the table. If this is a risk, see
++ * cuckoo_hash_add_check
++ *
++ * \param hashtab The hash table to add the entry to
++ * \param key Pointer to a mac address to use as a key
++ * \param value The value to store
++ * \param can_rehash Flag to allow the add function to rehash the
++ * table if necessary
++ *
++ * \return 0 on success, non-zero on failure. -ENOSPC means it just
++ * couldn't find anywhere to put it - this is bad and probably means
++ * an entry has been dropped on the floor (but the entry you just
++ * tried to add may now be included)
++ */
++extern
++int cuckoo_hash_add(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value value,
++ int can_rehash);
++
++/*! Same as cuckoo_hash_add but first checks to ensure entry is not
++ * already there
++ * \return -EBUSY if already there
++ */
++
++extern
++int cuckoo_hash_add_check(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key,
++ cuckoo_hash_value value,
++ int can_rehash);
++/*! Remove an entry from the table
++ *
++ * \param hashtab The hash table to remove the entry from
++ * \param key The key that was used to previously add the entry
++ *
++ * \return 0 on success, -EINVAL if the entry couldn't be found
++ */
++extern
++int cuckoo_hash_remove(cuckoo_hash_table *hashtab, cuckoo_hash_key *key);
++
++
++/*! Helper for those using mac addresses to convert to a key for the
++ * hash table
++ */
++static inline cuckoo_hash_mac_key cuckoo_mac_to_key(const u8 *mac)
++{
++ return (cuckoo_hash_mac_key)(mac[0])
++ | (cuckoo_hash_mac_key)(mac[1]) << 8
++ | (cuckoo_hash_mac_key)(mac[2]) << 16
++ | (cuckoo_hash_mac_key)(mac[3]) << 24
++ | (cuckoo_hash_mac_key)(mac[4]) << 32
++ | (cuckoo_hash_mac_key)(mac[5]) << 40;
++}
++
++
++/*! Update an entry already in the hash table to take a new value
++ *
++ * \param hashtab The hash table to add the entry to
++ * \param key Pointer to a mac address to use as a key
++ * \param value The value to store
++ *
++ * \return 0 on success, non-zero on failure.
++ */
++int cuckoo_hash_update(cuckoo_hash_table *hashtab, cuckoo_hash_key *key,
++ cuckoo_hash_value value);
++
++
++/*! Go through the hash table and return all used entries (one per call)
++ *
++ * \param hashtab The hash table to iterate over
++ * \param key Pointer to a key to take the returned key
++ * \param value Pointer to a value to take the returned value
++ *
++ * \return 0 on success (key, value set), non-zero on failure.
++ */
++int cuckoo_hash_iterate(cuckoo_hash_table *hashtab,
++ cuckoo_hash_key *key, cuckoo_hash_value *value);
++void cuckoo_hash_iterate_reset(cuckoo_hash_table *hashtab);
++
++/* debug, not compiled by default */
++void cuckoo_hash_valid(cuckoo_hash_table *hashtab);
++void cuckoo_hash_dump(cuckoo_hash_table *hashtab);
++
++#endif /* NET_ACCEL_CUCKOO_HASH_H */
+diff --git a/drivers/xen/sfc_netutil/accel_msg_iface.c b/drivers/xen/sfc_netutil/accel_msg_iface.c
+new file mode 100644
+index 0000000..0353358
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_msg_iface.c
+@@ -0,0 +1,301 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <xen/evtchn.h>
++
++#include "accel_util.h"
++#include "accel_msg_iface.h"
++
++#define NET_ACCEL_MSG_Q_SIZE (1024)
++#define NET_ACCEL_MSG_Q_MASK (NET_ACCEL_MSG_Q_SIZE - 1)
++
++#ifdef NDEBUG
++#define NET_ACCEL_CHECK_MAGIC(_p, _errval)
++#define NET_ACCEL_SHOW_QUEUE(_t, _q, _id)
++#else
++#define NET_ACCEL_CHECK_MAGIC(_p, _errval) \
++ if (_p->magic != NET_ACCEL_MSG_MAGIC) { \
++ printk(KERN_ERR "%s: passed invalid shared page %p!\n", \
++ __FUNCTION__, _p); \
++ return _errval; \
++ }
++#define NET_ACCEL_SHOW_QUEUE(_t, _q, _id) \
++ printk(_t ": queue %d write %x read %x base %x limit %x\n", \
++ _id, _q->write, _q->read, _q->base, _q->limit);
++#endif
++
++/*
++ * We've been passed at least 2 pages. 1 control page and 1 or more
++ * data pages.
++ */
++int net_accel_msg_init_page(void *mem, int len, int up)
++{
++ struct net_accel_shared_page *shared_page =
++ (struct net_accel_shared_page*)mem;
++
++ if ((unsigned long)shared_page & NET_ACCEL_MSG_Q_MASK)
++ return -EINVAL;
++
++ shared_page->magic = NET_ACCEL_MSG_MAGIC;
++
++ shared_page->aflags = 0;
++
++ shared_page->net_dev_up = up;
++
++ return 0;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_init_page);
++
++
++void net_accel_msg_init_queue(sh_msg_fifo2 *queue,
++ struct net_accel_msg_queue *indices,
++ struct net_accel_msg *base, int size)
++{
++ queue->fifo = base;
++ spin_lock_init(&queue->lock);
++ sh_fifo2_init(queue, size-1, &indices->read, &indices->write);
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_init_queue);
++
++
++static inline int _net_accel_msg_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ struct net_accel_msg *msg,
++ int is_reply)
++{
++ int rc = 0;
++ NET_ACCEL_CHECK_MAGIC(sp, -EINVAL);
++ rmb();
++ if (is_reply) {
++ EPRINTK_ON(sh_fifo2_is_full(queue));
++ sh_fifo2_put(queue, *msg);
++ } else {
++ if (sh_fifo2_not_half_full(queue)) {
++ sh_fifo2_put(queue, *msg);
++ } else {
++ rc = -ENOSPC;
++ }
++ }
++ wmb();
++ return rc;
++}
++
++/* Notify after a batch of messages have been sent */
++void net_accel_msg_notify(int irq)
++{
++ notify_remote_via_irq(irq);
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_notify);
++
++/*
++ * Send a message on the specified FIFO. Returns 0 on success, -errno
++ * on failure. The message in msg is copied to the current slot of the
++ * FIFO.
++ */
++int net_accel_msg_send(struct net_accel_shared_page *sp, sh_msg_fifo2 *q,
++ struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ int rc;
++ net_accel_msg_lock_queue(q, &flags);
++ rc = _net_accel_msg_send(sp, q, msg, 0);
++ net_accel_msg_unlock_queue(q, &flags);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_send);
++
++
++/* As net_accel_msg_send but also posts a notification to the far end. */
++int net_accel_msg_send_notify(struct net_accel_shared_page *sp, int irq,
++ sh_msg_fifo2 *q, struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ int rc;
++ net_accel_msg_lock_queue(q, &flags);
++ rc = _net_accel_msg_send(sp, q, msg, 0);
++ net_accel_msg_unlock_queue(q, &flags);
++ if (rc >= 0)
++ notify_remote_via_irq(irq);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_send_notify);
++
++
++int net_accel_msg_reply(struct net_accel_shared_page *sp, sh_msg_fifo2 *q,
++ struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ int rc;
++ net_accel_msg_lock_queue(q, &flags);
++ rc = _net_accel_msg_send(sp, q, msg, 1);
++ net_accel_msg_unlock_queue(q, &flags);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_reply);
++
++
++/* As net_accel_msg_send but also posts a notification to the far end. */
++int net_accel_msg_reply_notify(struct net_accel_shared_page *sp, int irq,
++ sh_msg_fifo2 *q, struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ int rc;
++ net_accel_msg_lock_queue(q, &flags);
++ rc = _net_accel_msg_send(sp, q, msg, 1);
++ net_accel_msg_unlock_queue(q, &flags);
++ if (rc >= 0)
++ notify_remote_via_irq(irq);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_reply_notify);
++
++
++/*
++ * Look at a received message, if any, so a decision can be made about
++ * whether to read it now or not. Cookie is a bit of debug which is
++ * set here and checked when passed to net_accel_msg_recv_next()
++ */
++int net_accel_msg_peek(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ struct net_accel_msg *msg, int *cookie)
++{
++ unsigned long flags;
++ int rc = 0;
++ NET_ACCEL_CHECK_MAGIC(sp, -EINVAL);
++ net_accel_msg_lock_queue(queue, &flags);
++ rmb();
++ if (sh_fifo2_is_empty(queue)) {
++ rc = -ENOENT;
++ } else {
++ *msg = sh_fifo2_peek(queue);
++ *cookie = *(queue->fifo_rd_i);
++ }
++ net_accel_msg_unlock_queue(queue, &flags);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_peek);
++
++
++/*
++ * Move the queue onto the next element, used after finished with a
++ * peeked msg
++ */
++int net_accel_msg_recv_next(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue, int cookie)
++{
++ unsigned long flags;
++ NET_ACCEL_CHECK_MAGIC(sp, -EINVAL);
++ net_accel_msg_lock_queue(queue, &flags);
++ rmb();
++ /* Mustn't be empty */
++ BUG_ON(sh_fifo2_is_empty(queue));
++ /*
++ * Check cookie matches, i.e. we're advancing over the same message
++ * as was got using peek
++ */
++ BUG_ON(cookie != *(queue->fifo_rd_i));
++ sh_fifo2_rd_next(queue);
++ wmb();
++ net_accel_msg_unlock_queue(queue, &flags);
++ return 0;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_recv_next);
++
++
++/*
++ * Receive a message on the specified FIFO. Returns 0 on success,
++ * -errno on failure.
++ */
++int net_accel_msg_recv(struct net_accel_shared_page *sp, sh_msg_fifo2 *queue,
++ struct net_accel_msg *msg)
++{
++ unsigned long flags;
++ int rc = 0;
++ NET_ACCEL_CHECK_MAGIC(sp, -EINVAL);
++ net_accel_msg_lock_queue(queue, &flags);
++ rmb();
++ if (sh_fifo2_is_empty(queue)) {
++ rc = -ENOENT;
++ } else {
++ sh_fifo2_get(queue, msg);
++ }
++ wmb();
++ net_accel_msg_unlock_queue(queue, &flags);
++ return rc;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_recv);
++
++
++/*
++ * Start sending a message without copying. returns a pointer to a message
++ * that will be filled out in place. The queue is locked until the message
++ * is sent.
++ */
++struct net_accel_msg *net_accel_msg_start_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue, unsigned long *flags)
++{
++ struct net_accel_msg *msg;
++ NET_ACCEL_CHECK_MAGIC(sp, NULL);
++ net_accel_msg_lock_queue(queue, flags);
++ rmb();
++ if (sh_fifo2_not_half_full(queue)) {
++ msg = sh_fifo2_pokep(queue);
++ } else {
++ net_accel_msg_unlock_queue(queue, flags);
++ msg = NULL;
++ }
++ return msg;
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_start_send);
++
++
++static inline void _msg_complete(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags)
++{
++ sh_fifo2_wr_next(queue);
++ net_accel_msg_unlock_queue(queue, flags);
++}
++
++/*
++ * Complete the sending of a message started with net_accel_msg_start_send. The
++ * message is implicit since the queue was locked by _start
++ */
++void net_accel_msg_complete_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags)
++{
++ _msg_complete(sp, queue, flags);
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_complete_send);
++
++/* As net_accel_msg_complete_send but does the notify. */
++void net_accel_msg_complete_send_notify(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags, int irq)
++{
++ _msg_complete(sp, queue, flags);
++ notify_remote_via_irq(irq);
++}
++EXPORT_SYMBOL_GPL(net_accel_msg_complete_send_notify);
+diff --git a/drivers/xen/sfc_netutil/accel_msg_iface.h b/drivers/xen/sfc_netutil/accel_msg_iface.h
+new file mode 100644
+index 0000000..0033378
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_msg_iface.h
+@@ -0,0 +1,414 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NET_ACCEL_MSG_IFACE_H
++#define NET_ACCEL_MSG_IFACE_H
++
++#include <linux/ip.h>
++#include <linux/tcp.h>
++#include <linux/udp.h>
++#include <linux/in.h>
++#include <linux/netdevice.h>
++#include <linux/etherdevice.h>
++
++#include "accel_shared_fifo.h"
++
++#define NET_ACCEL_MSG_MAGIC (0x85465479)
++
++/*! We talk version 0.010 of the interdomain protocol */
++#define NET_ACCEL_MSG_VERSION (0x00001000)
++
++/*! Shared memory portion of inter-domain FIFO */
++struct net_accel_msg_queue {
++ u32 read;
++ u32 write;
++};
++
++
++/*
++ * The aflags in the following structure is used as follows:
++ *
++ * - each bit is set when one of the corresponding variables is
++ * changed by either end.
++ *
++ * - the end that has made the change then forwards an IRQ to the
++ * other
++ *
++ * - the IRQ handler deals with these bits either on the fast path, or
++ * for less common changes, by jumping onto the slow path.
++ *
++ * - once it has seen a change, it clears the relevant bit.
++ *
++ * aflags is accessed atomically using clear_bit, test_bit,
++ * test_and_set_bit etc
++ */
++
++/*
++ * The following used to signify to the other domain when the queue
++ * they want to use is full, and when it is no longer full. Could be
++ * compressed to use fewer bits but done this way for simplicity and
++ * clarity
++ */
++
++/* "dom0->domU queue" is full */
++#define NET_ACCEL_MSG_AFLAGS_QUEUE0FULL 0x1
++#define NET_ACCEL_MSG_AFLAGS_QUEUE0FULL_B 0
++/* "dom0->domU queue" is not full */
++#define NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL 0x2
++#define NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL_B 1
++/* "domU->dom0 queue" is full */
++#define NET_ACCEL_MSG_AFLAGS_QUEUEUFULL 0x4
++#define NET_ACCEL_MSG_AFLAGS_QUEUEUFULL_B 2
++/* "domU->dom0 queue" is not full */
++#define NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL 0x8
++#define NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL_B 3
++/* dom0 -> domU net_dev up/down events */
++#define NET_ACCEL_MSG_AFLAGS_NETUPDOWN 0x10
++#define NET_ACCEL_MSG_AFLAGS_NETUPDOWN_B 4
++
++/*
++ * Masks used to test if there are any messages for domU and dom0
++ * respectively
++ */
++#define NET_ACCEL_MSG_AFLAGS_TO_DOMU_MASK \
++ (NET_ACCEL_MSG_AFLAGS_QUEUE0FULL | \
++ NET_ACCEL_MSG_AFLAGS_QUEUEUNOTFULL | \
++ NET_ACCEL_MSG_AFLAGS_NETUPDOWN)
++#define NET_ACCEL_MSG_AFLAGS_TO_DOM0_MASK \
++ (NET_ACCEL_MSG_AFLAGS_QUEUE0NOTFULL | \
++ NET_ACCEL_MSG_AFLAGS_QUEUEUFULL)
++
++/*! The shared data structure used for inter-VM communication. */
++struct net_accel_shared_page {
++ /*! Sanity check */
++ u32 magic;
++ /*! Used by host/Dom0 */
++ struct net_accel_msg_queue queue0;
++ /*! Used by guest/DomU */
++ struct net_accel_msg_queue queue1;
++ /*! Atomic flags, used to communicate simple state changes */
++ u32 aflags;
++ /*! State of net_dev used for acceleration */
++ u32 net_dev_up;
++};
++
++
++enum net_accel_hw_type {
++ /*! Not a virtualisable NIC: use slow path. */
++ NET_ACCEL_MSG_HWTYPE_NONE = 0,
++ /*! NIC is Falcon-based */
++ NET_ACCEL_MSG_HWTYPE_FALCON_A = 1,
++ NET_ACCEL_MSG_HWTYPE_FALCON_B = 2,
++};
++
++/*! The maximum number of pages used by an event queue. */
++#define EF_HW_FALCON_EVQ_PAGES 8
++
++struct net_accel_hw_falcon_b {
++ /* VI */
++ /*! Grant for Tx DMA Q */
++ u32 txdmaq_gnt;
++ /*! Grant for Rx DMA Q */
++ u32 rxdmaq_gnt;
++ /*! Machine frame number for Tx/Rx doorbell page */
++ u32 doorbell_mfn;
++ /*! Grant for Tx/Rx doorbell page */
++ u32 doorbell_gnt;
++
++ /* Event Q */
++ /*! Grants for the pages of the EVQ */
++ u32 evq_mem_gnts[EF_HW_FALCON_EVQ_PAGES];
++ u32 evq_offs;
++ /*! log2(pages in event Q) */
++ u32 evq_order;
++ /*! Capacity in events */
++ u32 evq_capacity;
++ /*! Eventq pointer register physical address */
++ u32 evq_rptr;
++ /*! Interface instance */
++ u32 instance;
++ /*! Capacity of RX queue */
++ u32 rx_capacity;
++ /*! Capacity of TX queue */
++ u32 tx_capacity;
++
++ /* NIC */
++ s32 nic_arch;
++ s32 nic_revision;
++ u8 nic_variant;
++};
++
++struct net_accel_hw_falcon_a {
++ struct net_accel_hw_falcon_b common;
++ u32 evq_rptr_gnt;
++};
++
++
++/*! Description of the hardware that the DomU is being given. */
++struct net_accel_msg_hw {
++ u32 type; /*!< Hardware type */
++ union {
++ struct net_accel_hw_falcon_a falcon_a;
++ struct net_accel_hw_falcon_b falcon_b;
++ } resources;
++};
++
++/*! Start-of-day handshake message. Dom0 fills in its version and
++ * sends, DomU checks, inserts its version and replies
++ */
++struct net_accel_msg_hello {
++ /*! Sender's version (set by each side in turn) */
++ u32 version;
++ /*! max pages allocated/allowed for buffers */
++ u32 max_pages;
++};
++
++/*! Maximum number of page requests that can fit in a message. */
++#define NET_ACCEL_MSG_MAX_PAGE_REQ (8)
++
++/*! Request for NIC buffers. DomU fils out pages and grants (and
++ * optionally) reqid, dom0 fills out buf and sends reply
++ */
++struct net_accel_msg_map_buffers {
++ u32 reqid; /*!< Optional request ID */
++ u32 pages; /*!< Number of pages to map */
++ u32 grants[NET_ACCEL_MSG_MAX_PAGE_REQ]; /*!< Grant ids to map */
++ u32 buf; /*!< NIC buffer address of pages obtained */
++};
++
++/*! Notification of a change to local mac address, used to filter
++ locally destined packets off the fast path */
++struct net_accel_msg_localmac {
++ u32 flags; /*!< Should this be added or removed? */
++ u8 mac[ETH_ALEN]; /*!< The mac address to filter onto slow path */
++};
++
++struct net_accel_msg_fastpath {
++ u32 flags; /*!< Should this be added or removed? */
++ u8 mac[ETH_ALEN];/*!< The mac address to filter onto fast path */
++ u16 port; /*!< The port of the connection */
++ u32 ip; /*!< The IP address of the connection */
++ u8 proto; /*!< The protocol of connection (TCP/UDP) */
++};
++
++/*! Values for struct ef_msg_localmac/fastpath.flags */
++#define NET_ACCEL_MSG_ADD 0x1
++#define NET_ACCEL_MSG_REMOVE 0x2
++
++/*! Overall message structure */
++struct net_accel_msg {
++ /*! ID specifying type of messge */
++ u32 id;
++ union {
++ /*! handshake */
++ struct net_accel_msg_hello hello;
++ /*! hardware description */
++ struct net_accel_msg_hw hw;
++ /*! buffer map request */
++ struct net_accel_msg_map_buffers mapbufs;
++ /*! mac address of a local interface */
++ struct net_accel_msg_localmac localmac;
++ /*! address of a new fastpath connection */
++ struct net_accel_msg_fastpath fastpath;
++ /*! make the message a fixed size */
++ u8 pad[128 - sizeof(u32)];
++ } u;
++};
++
++
++#define NET_ACCEL_MSG_HW_TO_MSG(_u) container_of(_u, struct net_accel_msg, u.hw)
++
++/*! Inter-domain message FIFO */
++typedef struct {
++ struct net_accel_msg *fifo;
++ u32 fifo_mask;
++ u32 *fifo_rd_i;
++ u32 *fifo_wr_i;
++ spinlock_t lock;
++ u32 is_locked; /* Debug flag */
++} sh_msg_fifo2;
++
++
++#define NET_ACCEL_MSG_OFFSET_MASK PAGE_MASK
++
++/* Modifiers */
++#define NET_ACCEL_MSG_REPLY (0x80000000)
++#define NET_ACCEL_MSG_ERROR (0x40000000)
++
++/* Dom0 -> DomU and reply. Handshake/version check. */
++#define NET_ACCEL_MSG_HELLO (0x00000001)
++/* Dom0 -> DomU : hardware setup (VI info.) */
++#define NET_ACCEL_MSG_SETHW (0x00000002)
++/*
++ * Dom0 -> DomU. Notification of a local mac to add/remove from slow
++ * path filter
++ */
++#define NET_ACCEL_MSG_LOCALMAC (0x00000003)
++/*
++ * DomU -> Dom0 and reply. Request for buffer table entries for
++ * preallocated pages.
++ */
++#define NET_ACCEL_MSG_MAPBUF (0x00000004)
++/*
++ * Dom0 -> DomU. Notification of a local mac to add/remove from fast
++ * path filter
++ */
++#define NET_ACCEL_MSG_FASTPATH (0x00000005)
++
++/*! Initialise a message and set the type
++ * \param message : the message
++ * \param code : the message type
++ */
++static inline void net_accel_msg_init(struct net_accel_msg *msg, int code) {
++ msg->id = (u32)code;
++}
++
++/*! initialise a shared page structure
++ * \param shared_page : mapped memory in which the structure resides
++ * \param len : size of the message FIFO area that follows
++ * \param up : initial up/down state of netdev
++ * \return 0 or an error code
++ */
++extern int net_accel_msg_init_page(void *shared_page, int len, int up);
++
++/*! initialise a message queue
++ * \param queue : the message FIFO to initialise
++ * \param indices : the read and write indices in shared memory
++ * \param base : the start of the memory area for the FIFO
++ * \param size : the size of the FIFO in bytes
++ */
++extern void net_accel_msg_init_queue(sh_msg_fifo2 *queue,
++ struct net_accel_msg_queue *indices,
++ struct net_accel_msg *base, int size);
++
++/* Notify after a batch of messages have been sent */
++extern void net_accel_msg_notify(int irq);
++
++/*! Send a message on the specified FIFO. The message is copied to the
++ * current slot of the FIFO.
++ * \param sp : pointer to shared page
++ * \param q : pointer to message FIFO to use
++ * \param msg : pointer to message
++ * \return 0 on success, -errno on
++ */
++extern int net_accel_msg_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *q,
++ struct net_accel_msg *msg);
++extern int net_accel_msg_reply(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *q,
++ struct net_accel_msg *msg);
++
++/*! As net_accel_msg_send but also posts a notification to the far end. */
++extern int net_accel_msg_send_notify(struct net_accel_shared_page *sp,
++ int irq, sh_msg_fifo2 *q,
++ struct net_accel_msg *msg);
++/*! As net_accel_msg_send but also posts a notification to the far end. */
++extern int net_accel_msg_reply_notify(struct net_accel_shared_page *sp,
++ int irq, sh_msg_fifo2 *q,
++ struct net_accel_msg *msg);
++
++/*! Receive a message on the specified FIFO. Returns 0 on success,
++ * -errno on failure.
++ */
++extern int net_accel_msg_recv(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *q,
++ struct net_accel_msg *msg);
++
++/*! Look at a received message, if any, so a decision can be made
++ * about whether to read it now or not. Cookie is a bit of debug
++ * which is set here and checked when passed to
++ * net_accel_msg_recv_next()
++ */
++extern int net_accel_msg_peek(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ struct net_accel_msg *msg, int *cookie);
++/*! Move the queue onto the next element, used after finished with a
++ * peeked msg
++ */
++extern int net_accel_msg_recv_next(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue, int cookie);
++
++/*! Start sending a message without copying. returns a pointer to a
++ * message that will be filled out in place. The queue is locked
++ * until the message is sent.
++ */
++extern
++struct net_accel_msg *net_accel_msg_start_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags);
++
++
++/*! Complete the sending of a message started with
++ * net_accel_msg_start_send. The message is implicit since the queue
++ * was locked by _start
++ */
++extern void net_accel_msg_complete_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags);
++
++/*! As net_accel_msg_complete_send but does the notify. */
++extern void net_accel_msg_complete_send_notify(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue,
++ unsigned long *flags, int irq);
++
++/*! Lock the queue so that multiple "_locked" functions can be called
++ * without the queue being modified by others
++ */
++static inline
++void net_accel_msg_lock_queue(sh_msg_fifo2 *queue, unsigned long *flags)
++{
++ spin_lock_irqsave(&queue->lock, (*flags));
++ rmb();
++ BUG_ON(queue->is_locked);
++ queue->is_locked = 1;
++}
++
++/*! Unlock the queue */
++static inline
++void net_accel_msg_unlock_queue(sh_msg_fifo2 *queue, unsigned long *flags)
++{
++ BUG_ON(!queue->is_locked);
++ queue->is_locked = 0;
++ wmb();
++ spin_unlock_irqrestore(&queue->lock, (*flags));
++}
++
++/*! Give up without sending a message that was started with
++ * net_accel_msg_start_send()
++ */
++static inline
++void net_accel_msg_abort_send(struct net_accel_shared_page *sp,
++ sh_msg_fifo2 *queue, unsigned long *flags)
++{
++ net_accel_msg_unlock_queue(queue, flags);
++}
++
++/*! Test the queue to ensure there is sufficient space */
++static inline
++int net_accel_msg_check_space(sh_msg_fifo2 *queue, unsigned space)
++{
++ return sh_fifo2_space(queue) >= space;
++}
++
++#endif /* NET_ACCEL_MSG_IFACE_H */
+diff --git a/drivers/xen/sfc_netutil/accel_shared_fifo.h b/drivers/xen/sfc_netutil/accel_shared_fifo.h
+new file mode 100644
+index 0000000..a55608a
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_shared_fifo.h
+@@ -0,0 +1,127 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NET_ACCEL_SHARED_FIFO_H
++#define NET_ACCEL_SHARED_FIFO_H
++
++/*
++ * This is based on fifo.h, but handles sharing between address spaces
++ * that don't trust each other, by splitting out the read and write
++ * indices. This costs at least one pointer indirection more than the
++ * vanilla version per access.
++ */
++
++typedef struct {
++ char* fifo;
++ unsigned fifo_mask;
++ unsigned *fifo_rd_i;
++ unsigned *fifo_wr_i;
++} sh_byte_fifo2;
++
++#define SH_FIFO2_M(f, x) ((x) & ((f)->fifo_mask))
++
++static inline unsigned log2_ge(unsigned long n, unsigned min_order) {
++ unsigned order = min_order;
++ while((1ul << order) < n) ++order;
++ return order;
++}
++
++static inline unsigned long pow2(unsigned order) {
++ return (1ul << order);
++}
++
++#define is_pow2(x) (pow2(log2_ge((x), 0)) == (x))
++
++#define sh_fifo2_valid(f) ((f) && (f)->fifo && (f)->fifo_mask > 0 && \
++ is_pow2((f)->fifo_mask+1u))
++
++#define sh_fifo2_init(f, cap, _rptr, _wptr) \
++ do { \
++ BUG_ON(!is_pow2((cap) + 1)); \
++ (f)->fifo_rd_i = _rptr; \
++ (f)->fifo_wr_i = _wptr; \
++ *(f)->fifo_rd_i = *(f)->fifo_wr_i = 0u; \
++ (f)->fifo_mask = (cap); \
++ } while(0)
++
++#define sh_fifo2_num(f) SH_FIFO2_M((f),*(f)->fifo_wr_i - *(f)->fifo_rd_i)
++#define sh_fifo2_space(f) SH_FIFO2_M((f),*(f)->fifo_rd_i - *(f)->fifo_wr_i-1u)
++#define sh_fifo2_is_empty(f) (sh_fifo2_num(f)==0)
++#define sh_fifo2_not_empty(f) (sh_fifo2_num(f)!=0)
++#define sh_fifo2_is_full(f) (sh_fifo2_space(f)==0u)
++#define sh_fifo2_not_full(f) (sh_fifo2_space(f)!=0u)
++#define sh_fifo2_buf_size(f) ((f)->fifo_mask + 1u)
++#define sh_fifo2_capacity(f) ((f)->fifo_mask)
++#define sh_fifo2_end(f) ((f)->fifo + sh_fifo2_buf_size(f))
++#define sh_fifo2_not_half_full(f) (sh_fifo2_space(f) > (sh_fifo2_capacity(f) >> 1))
++
++#define sh_fifo2_peek(f) ((f)->fifo[SH_FIFO2_M((f), *(f)->fifo_rd_i)])
++#define sh_fifo2_peekp(f) ((f)->fifo + SH_FIFO2_M((f), *(f)->fifo_rd_i))
++#define sh_fifo2_poke(f) ((f)->fifo[SH_FIFO2_M((f), *(f)->fifo_wr_i)])
++#define sh_fifo2_pokep(f) ((f)->fifo + SH_FIFO2_M((f), *(f)->fifo_wr_i))
++#define sh_fifo2_peek_i(f,i) ((f)->fifo[SH_FIFO2_M((f), *(f)->fifo_rd_i+(i))])
++#define sh_fifo2_poke_i(f,i) ((f)->fifo[SH_FIFO2_M((f), *(f)->fifo_wr_i+(i))])
++
++#define sh_fifo2_rd_next(f) \
++ do {*(f)->fifo_rd_i = *(f)->fifo_rd_i + 1u;} while(0)
++#define sh_fifo2_wr_next(f) \
++ do {*(f)->fifo_wr_i = *(f)->fifo_wr_i + 1u;} while(0)
++#define sh_fifo2_rd_adv(f, n) \
++ do {*(f)->fifo_rd_i = *(f)->fifo_rd_i + (n);} while(0)
++#define sh_fifo2_wr_adv(f, n) \
++ do {*(f)->fifo_wr_i = *(f)->fifo_wr_i + (n);} while(0)
++
++#define sh_fifo2_put(f, v) \
++ do {sh_fifo2_poke(f) = (v); wmb(); sh_fifo2_wr_next(f);} while(0)
++
++#define sh_fifo2_get(f, pv) \
++ do {*(pv) = sh_fifo2_peek(f); mb(); sh_fifo2_rd_next(f);} while(0)
++
++static inline unsigned sh_fifo2_contig_num(sh_byte_fifo2 *f)
++{
++ unsigned fifo_wr_i = SH_FIFO2_M(f, *f->fifo_wr_i);
++ unsigned fifo_rd_i = SH_FIFO2_M(f, *f->fifo_rd_i);
++
++ return (fifo_wr_i >= fifo_rd_i)
++ ? fifo_wr_i - fifo_rd_i
++ : f->fifo_mask + 1u - *(f)->fifo_rd_i;
++}
++
++static inline unsigned sh_fifo2_contig_space(sh_byte_fifo2 *f)
++{
++ unsigned fifo_wr_i = SH_FIFO2_M(f, *f->fifo_wr_i);
++ unsigned fifo_rd_i = SH_FIFO2_M(f, *f->fifo_rd_i);
++
++ return (fifo_rd_i > fifo_wr_i)
++ ? fifo_rd_i - fifo_wr_i - 1
++ : (f->fifo_mask + 1u - fifo_wr_i
++ /*
++ * The last byte can't be used if the read pointer
++ * is at zero.
++ */
++ - (fifo_rd_i==0));
++}
++
++
++#endif /* NET_ACCEL_SHARED_FIFO_H */
+diff --git a/drivers/xen/sfc_netutil/accel_util.c b/drivers/xen/sfc_netutil/accel_util.c
+new file mode 100644
+index 0000000..3293079
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_util.c
+@@ -0,0 +1,333 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#include <linux/if_ether.h>
++#include <asm/io.h>
++#include <asm/pgtable.h>
++#include <asm/hypercall.h>
++#include <xen/xenbus.h>
++#include <xen/driver_util.h>
++#include <xen/gnttab.h>
++
++#include "accel_util.h"
++
++#ifdef EFX_GCOV
++#include "gcov.h"
++
++static int __init net_accel_init(void)
++{
++ gcov_provider_init(THIS_MODULE);
++ return 0;
++}
++module_init(net_accel_init);
++
++static void __exit net_accel_exit(void)
++{
++ gcov_provider_fini(THIS_MODULE);
++}
++module_exit(net_accel_exit);
++#endif
++
++/* Shutdown remote domain that is misbehaving */
++int net_accel_shutdown_remote(int domain)
++{
++ struct sched_remote_shutdown sched_shutdown = {
++ .domain_id = domain,
++ .reason = SHUTDOWN_crash
++ };
++
++ EPRINTK("Crashing domain %d\n", domain);
++
++ return HYPERVISOR_sched_op(SCHEDOP_remote_shutdown, &sched_shutdown);
++}
++EXPORT_SYMBOL(net_accel_shutdown_remote);
++
++
++/* Based on xenbus_backend_client.c:xenbus_map_ring() */
++static int net_accel_map_grant(struct xenbus_device *dev, int gnt_ref,
++ grant_handle_t *handle, void *vaddr,
++ u64 *dev_bus_addr, unsigned flags)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)vaddr, flags,
++ gnt_ref, dev->otherend_id);
++
++ BUG_ON(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1));
++
++ if (op.status != GNTST_okay) {
++ xenbus_dev_error
++ (dev, op.status,
++ "failed mapping in shared page %d from domain %d\n",
++ gnt_ref, dev->otherend_id);
++ } else {
++ *handle = op.handle;
++ if (dev_bus_addr)
++ *dev_bus_addr = op.dev_bus_addr;
++ }
++
++ return op.status;
++}
++
++
++/* Based on xenbus_backend_client.c:xenbus_unmap_ring() */
++static int net_accel_unmap_grant(struct xenbus_device *dev,
++ grant_handle_t handle,
++ void *vaddr, u64 dev_bus_addr,
++ unsigned flags)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)vaddr, flags, handle);
++
++ if (dev_bus_addr)
++ op.dev_bus_addr = dev_bus_addr;
++
++ BUG_ON(HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1));
++
++ if (op.status != GNTST_okay)
++ xenbus_dev_error(dev, op.status,
++ "failed unmapping page at handle %d error %d\n",
++ handle, op.status);
++
++ return op.status;
++}
++
++
++int net_accel_map_device_page(struct xenbus_device *dev,
++ int gnt_ref, grant_handle_t *handle,
++ u64 *dev_bus_addr)
++{
++ return net_accel_map_grant(dev, gnt_ref, handle, 0, dev_bus_addr,
++ GNTMAP_device_map);
++}
++EXPORT_SYMBOL_GPL(net_accel_map_device_page);
++
++
++int net_accel_unmap_device_page(struct xenbus_device *dev,
++ grant_handle_t handle, u64 dev_bus_addr)
++{
++ return net_accel_unmap_grant(dev, handle, 0, dev_bus_addr,
++ GNTMAP_device_map);
++}
++EXPORT_SYMBOL_GPL(net_accel_unmap_device_page);
++
++
++struct net_accel_valloc_grant_mapping {
++ struct vm_struct *vm;
++ int pages;
++ grant_handle_t grant_handles[0];
++};
++
++/* Map a series of grants into a contiguous virtual area */
++static void *net_accel_map_grants_valloc(struct xenbus_device *dev,
++ unsigned *grants, int npages,
++ unsigned flags, void **priv)
++{
++ struct net_accel_valloc_grant_mapping *map;
++ struct vm_struct *vm;
++ void *addr;
++ int i, j, rc;
++
++ vm = alloc_vm_area(PAGE_SIZE * npages);
++ if (vm == NULL) {
++ EPRINTK("No memory from alloc_vm_area.\n");
++ return NULL;
++ }
++ /*
++ * Get a structure in which we will record all the info needed
++ * to undo the mapping.
++ */
++ map = kzalloc(sizeof(struct net_accel_valloc_grant_mapping) +
++ npages * sizeof(grant_handle_t), GFP_KERNEL);
++ if (map == NULL) {
++ EPRINTK("No memory for net_accel_valloc_grant_mapping\n");
++ free_vm_area(vm);
++ return NULL;
++ }
++ map->vm = vm;
++ map->pages = npages;
++
++ /* Do the actual mapping */
++ addr = vm->addr;
++ for (i = 0; i < npages; i++) {
++ rc = net_accel_map_grant(dev, grants[i], map->grant_handles + i,
++ addr, NULL, flags);
++ if (rc != 0)
++ goto undo;
++ addr = (void*)((unsigned long)addr + PAGE_SIZE);
++ }
++
++ if (priv)
++ *priv = (void *)map;
++ else
++ kfree(map);
++
++ return vm->addr;
++
++ undo:
++ EPRINTK("Aborting contig map due to single map failure %d (%d of %d)\n",
++ rc, i+1, npages);
++ for (j = 0; j < i; j++) {
++ addr = (void*)((unsigned long)vm->addr + (j * PAGE_SIZE));
++ net_accel_unmap_grant(dev, map->grant_handles[j], addr, 0,
++ flags);
++ }
++ free_vm_area(vm);
++ kfree(map);
++ return NULL;
++}
++
++/* Undo the result of the mapping */
++static void net_accel_unmap_grants_vfree(struct xenbus_device *dev,
++ unsigned flags, void *priv)
++{
++ struct net_accel_valloc_grant_mapping *map =
++ (struct net_accel_valloc_grant_mapping *)priv;
++
++ void *addr = map->vm->addr;
++ int npages = map->pages;
++ int i;
++
++ for (i = 0; i < npages; i++) {
++ net_accel_unmap_grant(dev, map->grant_handles[i], addr, 0,
++ flags);
++ addr = (void*)((unsigned long)addr + PAGE_SIZE);
++ }
++ free_vm_area(map->vm);
++ kfree(map);
++}
++
++
++void *net_accel_map_grants_contig(struct xenbus_device *dev,
++ unsigned *grants, int npages,
++ void **priv)
++{
++ return net_accel_map_grants_valloc(dev, grants, npages,
++ GNTMAP_host_map, priv);
++}
++EXPORT_SYMBOL(net_accel_map_grants_contig);
++
++
++void net_accel_unmap_grants_contig(struct xenbus_device *dev,
++ void *priv)
++{
++ net_accel_unmap_grants_vfree(dev, GNTMAP_host_map, priv);
++}
++EXPORT_SYMBOL(net_accel_unmap_grants_contig);
++
++
++void *net_accel_map_iomem_page(struct xenbus_device *dev, int gnt_ref,
++ void **priv)
++{
++ return net_accel_map_grants_valloc(dev, &gnt_ref, 1,
++ GNTMAP_host_map, priv);
++}
++EXPORT_SYMBOL(net_accel_map_iomem_page);
++
++
++void net_accel_unmap_iomem_page(struct xenbus_device *dev, void *priv)
++{
++ net_accel_unmap_grants_vfree(dev, GNTMAP_host_map, priv);
++}
++EXPORT_SYMBOL(net_accel_unmap_iomem_page);
++
++
++int net_accel_grant_page(struct xenbus_device *dev, unsigned long mfn,
++ int is_iomem)
++{
++ int err = gnttab_grant_foreign_access(dev->otherend_id, mfn,
++ is_iomem ? GTF_PCD : 0);
++ if (err < 0)
++ xenbus_dev_error(dev, err, "failed granting access to page\n");
++ return err;
++}
++EXPORT_SYMBOL_GPL(net_accel_grant_page);
++
++
++int net_accel_ungrant_page(grant_ref_t gntref)
++{
++ if (unlikely(gnttab_query_foreign_access(gntref) != 0)) {
++ EPRINTK("%s: remote domain still using grant %d\n", __FUNCTION__,
++ gntref);
++ return -EBUSY;
++ }
++
++ gnttab_end_foreign_access(gntref, 0);
++ return 0;
++}
++EXPORT_SYMBOL_GPL(net_accel_ungrant_page);
++
++
++int net_accel_xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
++{
++ char *s, *e, *macstr;
++ int i;
++
++ macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
++ if (IS_ERR(macstr))
++ return PTR_ERR(macstr);
++
++ for (i = 0; i < ETH_ALEN; i++) {
++ mac[i] = simple_strtoul(s, &e, 16);
++ if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
++ kfree(macstr);
++ return -ENOENT;
++ }
++ s = e+1;
++ }
++
++ kfree(macstr);
++ return 0;
++}
++EXPORT_SYMBOL_GPL(net_accel_xen_net_read_mac);
++
++
++void net_accel_update_state(struct xenbus_device *dev, int state)
++{
++ struct xenbus_transaction tr;
++ int err;
++
++ DPRINTK("%s: setting accelstate to %s\n", __FUNCTION__,
++ xenbus_strstate(state));
++
++ if (xenbus_exists(XBT_NIL, dev->nodename, "")) {
++ VPRINTK("%s: nodename %s\n", __FUNCTION__, dev->nodename);
++ again:
++ err = xenbus_transaction_start(&tr);
++ if (err == 0)
++ err = xenbus_printf(tr, dev->nodename, "accelstate",
++ "%d", state);
++ if (err != 0) {
++ xenbus_transaction_end(tr, 1);
++ } else {
++ err = xenbus_transaction_end(tr, 0);
++ if (err == -EAGAIN)
++ goto again;
++ }
++ }
++}
++EXPORT_SYMBOL_GPL(net_accel_update_state);
++
++MODULE_LICENSE("GPL");
+diff --git a/drivers/xen/sfc_netutil/accel_util.h b/drivers/xen/sfc_netutil/accel_util.h
+new file mode 100644
+index 0000000..66f96d8
+--- /dev/null
++++ b/drivers/xen/sfc_netutil/accel_util.h
+@@ -0,0 +1,124 @@
++/****************************************************************************
++ * Solarflare driver for Xen network acceleration
++ *
++ * Copyright 2006-2008: Solarflare Communications Inc,
++ * 9501 Jeronimo Road, Suite 250,
++ * Irvine, CA 92618, USA
++ *
++ * Maintained by Solarflare Communications <linux-xen-drivers at solarflare.com>
++ *
++ * This program is free software; you can redistribute it and/or modify it
++ * under the terms of the GNU General Public License version 2 as published
++ * by the Free Software Foundation, incorporated herein by reference.
++ *
++ * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
++ ****************************************************************************
++ */
++
++#ifndef NETBACK_ACCEL_UTIL_H
++#define NETBACK_ACCEL_UTIL_H
++
++#ifdef DPRINTK
++#undef DPRINTK
++#endif
++
++#define FILE_LEAF strrchr(__FILE__, '/') ? strrchr(__FILE__, '/') + 1 : __FILE__
++
++#if 1
++#define VPRINTK(_f, _a...)
++#else
++#define VPRINTK(_f, _a...) \
++ printk("(file=%s, line=%d) " _f, \
++ FILE_LEAF , __LINE__ , ## _a )
++#endif
++
++#if 1
++#define DPRINTK(_f, _a...)
++#else
++#define DPRINTK(_f, _a...) \
++ printk("(file=%s, line=%d) " _f, \
++ FILE_LEAF , __LINE__ , ## _a )
++#endif
++
++#define EPRINTK(_f, _a...) \
++ printk("(file=%s, line=%d) " _f, \
++ FILE_LEAF , __LINE__ , ## _a )
++
++#define EPRINTK_ON(exp) \
++ do { \
++ if (exp) \
++ EPRINTK("%s at %s:%d\n", #exp, __FILE__, __LINE__); \
++ } while(0)
++
++#define DPRINTK_ON(exp) \
++ do { \
++ if (exp) \
++ DPRINTK("%s at %s:%d\n", #exp, __FILE__, __LINE__); \
++ } while(0)
++
++#include <xen/xenbus.h>
++
++/*! Map a set of pages from another domain
++ * \param dev The xenbus device context
++ * \param priv The private data returned by the mapping function
++ */
++extern
++void *net_accel_map_grants_contig(struct xenbus_device *dev,
++ unsigned *grants, int npages,
++ void **priv);
++
++/*! Unmap a set of pages mapped using net_accel_map_grants_contig.
++ * \param dev The xenbus device context
++ * \param priv The private data returned by the mapping function
++ */
++extern
++void net_accel_unmap_grants_contig(struct xenbus_device *dev, void *priv);
++
++/*! Read the MAC address of a device from xenstore */
++extern
++int net_accel_xen_net_read_mac(struct xenbus_device *dev, u8 mac[]);
++
++/*! Update the accelstate field for a device in xenstore */
++extern
++void net_accel_update_state(struct xenbus_device *dev, int state);
++
++/* These four map/unmap functions are based on
++ * xenbus_backend_client.c:xenbus_map_ring(). However, they are not
++ * used for ring buffers, instead just to map pages between domains,
++ * or to map a page so that it is accessible by a device
++ */
++extern
++int net_accel_map_device_page(struct xenbus_device *dev,
++ int gnt_ref, grant_handle_t *handle,
++ u64 *dev_bus_addr);
++extern
++int net_accel_unmap_device_page(struct xenbus_device *dev,
++ grant_handle_t handle, u64 dev_bus_addr);
++extern
++void *net_accel_map_iomem_page(struct xenbus_device *dev, int gnt_ref,
++ void **priv);
++extern
++void net_accel_unmap_iomem_page(struct xenbus_device *dev, void *priv);
++
++/*! Grrant a page to remote domain */
++extern
++int net_accel_grant_page(struct xenbus_device *dev, unsigned long mfn,
++ int is_iomem);
++/*! Undo a net_accel_grant_page */
++extern
++int net_accel_ungrant_page(grant_ref_t gntref);
++
++
++/*! Shutdown remote domain that is misbehaving */
++extern
++int net_accel_shutdown_remote(int domain);
++
++
++#endif
+diff --git a/drivers/xen/tpmback/Makefile b/drivers/xen/tpmback/Makefile
+new file mode 100644
+index 0000000..d5865c4
+--- /dev/null
++++ b/drivers/xen/tpmback/Makefile
+@@ -0,0 +1,4 @@
++
++obj-$(CONFIG_XEN_TPMDEV_BACKEND) += tpmbk.o
++
++tpmbk-y += tpmback.o interface.o xenbus.o
+diff --git a/drivers/xen/tpmback/common.h b/drivers/xen/tpmback/common.h
+new file mode 100644
+index 0000000..a944a39
+--- /dev/null
++++ b/drivers/xen/tpmback/common.h
+@@ -0,0 +1,85 @@
++/******************************************************************************
++ * drivers/xen/tpmback/common.h
++ */
++
++#ifndef __TPM__BACKEND__COMMON_H__
++#define __TPM__BACKEND__COMMON_H__
++
++#include <linux/version.h>
++#include <linux/module.h>
++#include <linux/interrupt.h>
++#include <linux/slab.h>
++#include <xen/evtchn.h>
++#include <xen/driver_util.h>
++#include <xen/interface/grant_table.h>
++#include <xen/interface/io/tpmif.h>
++#include <asm/io.h>
++#include <asm/pgalloc.h>
++
++#define DPRINTK(_f, _a...) \
++ pr_debug("(file=%s, line=%d) " _f, \
++ __FILE__ , __LINE__ , ## _a )
++
++struct backend_info;
++
++typedef struct tpmif_st {
++ struct list_head tpmif_list;
++ /* Unique identifier for this interface. */
++ domid_t domid;
++ unsigned int handle;
++
++ /* Physical parameters of the comms window. */
++ unsigned int irq;
++
++ /* The shared rings and indexes. */
++ tpmif_tx_interface_t *tx;
++ struct vm_struct *tx_area;
++
++ /* Miscellaneous private stuff. */
++ enum { DISCONNECTED, DISCONNECTING, CONNECTED } status;
++ int active;
++
++ struct tpmif_st *hash_next;
++ struct list_head list; /* scheduling list */
++ atomic_t refcnt;
++
++ struct backend_info *bi;
++
++ grant_handle_t shmem_handle;
++ grant_ref_t shmem_ref;
++ struct page **mmap_pages;
++
++ char devname[20];
++} tpmif_t;
++
++void tpmif_disconnect_complete(tpmif_t * tpmif);
++tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi);
++int tpmif_interface_init(void);
++void tpmif_interface_exit(void);
++void tpmif_schedule_work(tpmif_t * tpmif);
++void tpmif_deschedule_work(tpmif_t * tpmif);
++int tpmif_xenbus_init(void);
++void tpmif_xenbus_exit(void);
++int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn);
++irqreturn_t tpmif_be_int(int irq, void *dev_id);
++
++long int tpmback_get_instance(struct backend_info *bi);
++
++int vtpm_release_packets(tpmif_t * tpmif, int send_msgs);
++
++
++#define tpmif_get(_b) (atomic_inc(&(_b)->refcnt))
++#define tpmif_put(_b) \
++ do { \
++ if (atomic_dec_and_test(&(_b)->refcnt)) \
++ tpmif_disconnect_complete(_b); \
++ } while (0)
++
++extern int num_frontends;
++
++static inline unsigned long idx_to_kaddr(tpmif_t *t, unsigned int idx)
++{
++ return (unsigned long)pfn_to_kaddr(page_to_pfn(t->mmap_pages[idx]));
++}
++
++#endif /* __TPMIF__BACKEND__COMMON_H__ */
+diff --git a/drivers/xen/tpmback/interface.c b/drivers/xen/tpmback/interface.c
+new file mode 100644
+index 0000000..d1146ca
+--- /dev/null
++++ b/drivers/xen/tpmback/interface.c
+@@ -0,0 +1,170 @@
++ /*****************************************************************************
++ * drivers/xen/tpmback/interface.c
++ *
++ * Vritual TPM interface management.
++ *
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ *
++ * This code has been derived from drivers/xen/netback/interface.c
++ * Copyright (c) 2004, Keir Fraser
++ */
++
++#include "common.h"
++#include <linux/err.h>
++#include <xen/balloon.h>
++#include <xen/gnttab.h>
++
++static struct kmem_cache *tpmif_cachep;
++int num_frontends = 0;
++
++LIST_HEAD(tpmif_list);
++
++static tpmif_t *alloc_tpmif(domid_t domid, struct backend_info *bi)
++{
++ tpmif_t *tpmif;
++
++ tpmif = kmem_cache_alloc(tpmif_cachep, GFP_KERNEL);
++ if (tpmif == NULL)
++ goto out_of_memory;
++
++ memset(tpmif, 0, sizeof (*tpmif));
++ tpmif->domid = domid;
++ tpmif->status = DISCONNECTED;
++ tpmif->bi = bi;
++ snprintf(tpmif->devname, sizeof(tpmif->devname), "tpmif%d", domid);
++ atomic_set(&tpmif->refcnt, 1);
++
++ tpmif->mmap_pages = alloc_empty_pages_and_pagevec(TPMIF_TX_RING_SIZE);
++ if (tpmif->mmap_pages == NULL)
++ goto out_of_memory;
++
++ list_add(&tpmif->tpmif_list, &tpmif_list);
++ num_frontends++;
++
++ return tpmif;
++
++ out_of_memory:
++ if (tpmif != NULL)
++ kmem_cache_free(tpmif_cachep, tpmif);
++ printk("%s: out of memory\n", __FUNCTION__);
++ return ERR_PTR(-ENOMEM);
++}
++
++static void free_tpmif(tpmif_t * tpmif)
++{
++ num_frontends--;
++ list_del(&tpmif->tpmif_list);
++ free_empty_pages_and_pagevec(tpmif->mmap_pages, TPMIF_TX_RING_SIZE);
++ kmem_cache_free(tpmif_cachep, tpmif);
++}
++
++tpmif_t *tpmif_find(domid_t domid, struct backend_info *bi)
++{
++ tpmif_t *tpmif;
++
++ list_for_each_entry(tpmif, &tpmif_list, tpmif_list) {
++ if (tpmif->bi == bi) {
++ if (tpmif->domid == domid) {
++ tpmif_get(tpmif);
++ return tpmif;
++ } else {
++ return ERR_PTR(-EEXIST);
++ }
++ }
++ }
++
++ return alloc_tpmif(domid, bi);
++}
++
++static int map_frontend_page(tpmif_t *tpmif, unsigned long shared_page)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)tpmif->tx_area->addr,
++ GNTMAP_host_map, shared_page, tpmif->domid);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return op.status;
++ }
++
++ tpmif->shmem_ref = shared_page;
++ tpmif->shmem_handle = op.handle;
++
++ return 0;
++}
++
++static void unmap_frontend_page(tpmif_t *tpmif)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)tpmif->tx_area->addr,
++ GNTMAP_host_map, tpmif->shmem_handle);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++}
++
++int tpmif_map(tpmif_t *tpmif, unsigned long shared_page, unsigned int evtchn)
++{
++ int err;
++
++ if (tpmif->irq)
++ return 0;
++
++ if ((tpmif->tx_area = alloc_vm_area(PAGE_SIZE)) == NULL)
++ return -ENOMEM;
++
++ err = map_frontend_page(tpmif, shared_page);
++ if (err) {
++ free_vm_area(tpmif->tx_area);
++ return err;
++ }
++
++ tpmif->tx = (tpmif_tx_interface_t *)tpmif->tx_area->addr;
++ memset(tpmif->tx, 0, PAGE_SIZE);
++
++ err = bind_interdomain_evtchn_to_irqhandler(
++ tpmif->domid, evtchn, tpmif_be_int, 0, tpmif->devname, tpmif);
++ if (err < 0) {
++ unmap_frontend_page(tpmif);
++ free_vm_area(tpmif->tx_area);
++ return err;
++ }
++ tpmif->irq = err;
++
++ tpmif->shmem_ref = shared_page;
++ tpmif->active = 1;
++
++ return 0;
++}
++
++void tpmif_disconnect_complete(tpmif_t *tpmif)
++{
++ if (tpmif->irq)
++ unbind_from_irqhandler(tpmif->irq, tpmif);
++
++ if (tpmif->tx) {
++ unmap_frontend_page(tpmif);
++ free_vm_area(tpmif->tx_area);
++ }
++
++ free_tpmif(tpmif);
++}
++
++int __init tpmif_interface_init(void)
++{
++ tpmif_cachep = kmem_cache_create("tpmif_cache", sizeof (tpmif_t),
++ 0, 0, NULL);
++ return tpmif_cachep ? 0 : -ENOMEM;
++}
++
++void tpmif_interface_exit(void)
++{
++ kmem_cache_destroy(tpmif_cachep);
++}
+diff --git a/drivers/xen/tpmback/tpmback.c b/drivers/xen/tpmback/tpmback.c
+new file mode 100644
+index 0000000..62b8ba2
+--- /dev/null
++++ b/drivers/xen/tpmback/tpmback.c
+@@ -0,0 +1,952 @@
++/******************************************************************************
++ * drivers/xen/tpmback/tpmback.c
++ *
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from drivers/xen/netback/netback.c
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ */
++
++#include "common.h"
++#include <xen/evtchn.h>
++
++#include <linux/types.h>
++#include <linux/list.h>
++#include <linux/miscdevice.h>
++#include <linux/poll.h>
++#include <asm/uaccess.h>
++#include <xen/xenbus.h>
++#include <xen/interface/grant_table.h>
++#include <xen/gnttab.h>
++
++/* local data structures */
++struct data_exchange {
++ struct list_head pending_pak;
++ struct list_head current_pak;
++ unsigned int copied_so_far;
++ u8 has_opener:1;
++ u8 aborted:1;
++ rwlock_t pak_lock; // protects all of the previous fields
++ wait_queue_head_t wait_queue;
++};
++
++struct vtpm_resp_hdr {
++ uint32_t instance_no;
++ uint16_t tag_no;
++ uint32_t len_no;
++ uint32_t ordinal_no;
++} __attribute__ ((packed));
++
++struct packet {
++ struct list_head next;
++ unsigned int data_len;
++ u8 *data_buffer;
++ tpmif_t *tpmif;
++ u32 tpm_instance;
++ u8 req_tag;
++ u32 last_read;
++ u8 flags;
++ struct timer_list processing_timer;
++};
++
++enum {
++ PACKET_FLAG_DISCARD_RESPONSE = 1,
++};
++
++/* local variables */
++static struct data_exchange dataex;
++
++/* local function prototypes */
++static int _packet_write(struct packet *pak,
++ const char *data, size_t size, int userbuffer);
++static void processing_timeout(unsigned long ptr);
++static int packet_read_shmem(struct packet *pak,
++ tpmif_t * tpmif,
++ u32 offset,
++ char *buffer, int isuserbuffer, u32 left);
++static int vtpm_queue_packet(struct packet *pak);
++
++/***************************************************************
++ Buffer copying fo user and kernel space buffes.
++***************************************************************/
++static inline int copy_from_buffer(void *to,
++ const void *from, unsigned long size,
++ int isuserbuffer)
++{
++ if (isuserbuffer) {
++ if (copy_from_user(to, (void __user *)from, size))
++ return -EFAULT;
++ } else {
++ memcpy(to, from, size);
++ }
++ return 0;
++}
++
++static inline int copy_to_buffer(void *to,
++ const void *from, unsigned long size,
++ int isuserbuffer)
++{
++ if (isuserbuffer) {
++ if (copy_to_user((void __user *)to, from, size))
++ return -EFAULT;
++ } else {
++ memcpy(to, from, size);
++ }
++ return 0;
++}
++
++
++static void dataex_init(struct data_exchange *dataex)
++{
++ INIT_LIST_HEAD(&dataex->pending_pak);
++ INIT_LIST_HEAD(&dataex->current_pak);
++ dataex->has_opener = 0;
++ rwlock_init(&dataex->pak_lock);
++ init_waitqueue_head(&dataex->wait_queue);
++}
++
++/***************************************************************
++ Packet-related functions
++***************************************************************/
++
++static struct packet *packet_find_instance(struct list_head *head,
++ u32 tpm_instance)
++{
++ struct packet *pak;
++ struct list_head *p;
++
++ /*
++ * traverse the list of packets and return the first
++ * one with the given instance number
++ */
++ list_for_each(p, head) {
++ pak = list_entry(p, struct packet, next);
++
++ if (pak->tpm_instance == tpm_instance) {
++ return pak;
++ }
++ }
++ return NULL;
++}
++
++static struct packet *packet_find_packet(struct list_head *head, void *packet)
++{
++ struct packet *pak;
++ struct list_head *p;
++
++ /*
++ * traverse the list of packets and return the first
++ * one with the given instance number
++ */
++ list_for_each(p, head) {
++ pak = list_entry(p, struct packet, next);
++
++ if (pak == packet) {
++ return pak;
++ }
++ }
++ return NULL;
++}
++
++static struct packet *packet_alloc(tpmif_t * tpmif,
++ u32 size, u8 req_tag, u8 flags)
++{
++ struct packet *pak = NULL;
++ pak = kzalloc(sizeof (struct packet), GFP_ATOMIC);
++ if (NULL != pak) {
++ if (tpmif) {
++ pak->tpmif = tpmif;
++ pak->tpm_instance = tpmback_get_instance(tpmif->bi);
++ tpmif_get(tpmif);
++ }
++ pak->data_len = size;
++ pak->req_tag = req_tag;
++ pak->last_read = 0;
++ pak->flags = flags;
++
++ /*
++ * cannot do tpmif_get(tpmif); bad things happen
++ * on the last tpmif_put()
++ */
++ init_timer(&pak->processing_timer);
++ pak->processing_timer.function = processing_timeout;
++ pak->processing_timer.data = (unsigned long)pak;
++ }
++ return pak;
++}
++
++static void inline packet_reset(struct packet *pak)
++{
++ pak->last_read = 0;
++}
++
++static void packet_free(struct packet *pak)
++{
++ if (timer_pending(&pak->processing_timer)) {
++ BUG();
++ }
++
++ if (pak->tpmif)
++ tpmif_put(pak->tpmif);
++ kfree(pak->data_buffer);
++ /*
++ * cannot do tpmif_put(pak->tpmif); bad things happen
++ * on the last tpmif_put()
++ */
++ kfree(pak);
++}
++
++
++/*
++ * Write data to the shared memory and send it to the FE.
++ */
++static int packet_write(struct packet *pak,
++ const char *data, size_t size, int isuserbuffer)
++{
++ int rc = 0;
++
++ if (0 != (pak->flags & PACKET_FLAG_DISCARD_RESPONSE)) {
++ /* Don't send a respone to this packet. Just acknowledge it. */
++ rc = size;
++ } else {
++ rc = _packet_write(pak, data, size, isuserbuffer);
++ }
++
++ return rc;
++}
++
++int _packet_write(struct packet *pak,
++ const char *data, size_t size, int isuserbuffer)
++{
++ /*
++ * Write into the shared memory pages directly
++ * and send it to the front end.
++ */
++ tpmif_t *tpmif = pak->tpmif;
++ grant_handle_t handle;
++ int rc = 0;
++ unsigned int i = 0;
++ unsigned int offset = 0;
++
++ if (tpmif == NULL) {
++ return -EFAULT;
++ }
++
++ if (tpmif->status == DISCONNECTED) {
++ return size;
++ }
++
++ while (offset < size && i < TPMIF_TX_RING_SIZE) {
++ unsigned int tocopy;
++ struct gnttab_map_grant_ref map_op;
++ struct gnttab_unmap_grant_ref unmap_op;
++ tpmif_tx_request_t *tx;
++
++ tx = &tpmif->tx->ring[i].req;
++
++ if (0 == tx->addr) {
++ DPRINTK("ERROR: Buffer for outgoing packet NULL?! i=%d\n", i);
++ return 0;
++ }
++
++ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, tx->ref, tpmif->domid);
++
++ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &map_op, 1))) {
++ BUG();
++ }
++
++ handle = map_op.handle;
++
++ if (map_op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return 0;
++ }
++
++ tocopy = min_t(size_t, size - offset, PAGE_SIZE);
++
++ if (copy_from_buffer((void *)(idx_to_kaddr(tpmif, i) |
++ (tx->addr & ~PAGE_MASK)),
++ &data[offset], tocopy, isuserbuffer)) {
++ tpmif_put(tpmif);
++ return -EFAULT;
++ }
++ tx->size = tocopy;
++
++ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, handle);
++
++ if (unlikely
++ (HYPERVISOR_grant_table_op
++ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
++ BUG();
++ }
++
++ offset += tocopy;
++ i++;
++ }
++
++ rc = offset;
++ DPRINTK("Notifying frontend via irq %d\n", tpmif->irq);
++ notify_remote_via_irq(tpmif->irq);
++
++ return rc;
++}
++
++/*
++ * Read data from the shared memory and copy it directly into the
++ * provided buffer. Advance the read_last indicator which tells
++ * how many bytes have already been read.
++ */
++static int packet_read(struct packet *pak, size_t numbytes,
++ char *buffer, size_t buffersize, int isuserbuffer)
++{
++ tpmif_t *tpmif = pak->tpmif;
++
++ /*
++ * Read 'numbytes' of data from the buffer. The first 4
++ * bytes are the instance number in network byte order,
++ * after that come the data from the shared memory buffer.
++ */
++ u32 to_copy;
++ u32 offset = 0;
++ u32 room_left = buffersize;
++
++ if (pak->last_read < 4) {
++ /*
++ * copy the instance number into the buffer
++ */
++ u32 instance_no = htonl(pak->tpm_instance);
++ u32 last_read = pak->last_read;
++
++ to_copy = min_t(size_t, 4 - last_read, numbytes);
++
++ if (copy_to_buffer(&buffer[0],
++ &(((u8 *) & instance_no)[last_read]),
++ to_copy, isuserbuffer)) {
++ return -EFAULT;
++ }
++
++ pak->last_read += to_copy;
++ offset += to_copy;
++ room_left -= to_copy;
++ }
++
++ /*
++ * If the packet has a data buffer appended, read from it...
++ */
++
++ if (room_left > 0) {
++ if (pak->data_buffer) {
++ u32 to_copy = min_t(u32, pak->data_len - offset, room_left);
++ u32 last_read = pak->last_read - 4;
++
++ if (copy_to_buffer(&buffer[offset],
++ &pak->data_buffer[last_read],
++ to_copy, isuserbuffer)) {
++ return -EFAULT;
++ }
++ pak->last_read += to_copy;
++ offset += to_copy;
++ } else {
++ offset = packet_read_shmem(pak,
++ tpmif,
++ offset,
++ buffer,
++ isuserbuffer, room_left);
++ }
++ }
++ return offset;
++}
++
++static int packet_read_shmem(struct packet *pak,
++ tpmif_t * tpmif,
++ u32 offset, char *buffer, int isuserbuffer,
++ u32 room_left)
++{
++ u32 last_read = pak->last_read - 4;
++ u32 i = (last_read / PAGE_SIZE);
++ u32 pg_offset = last_read & (PAGE_SIZE - 1);
++ u32 to_copy;
++ grant_handle_t handle;
++
++ tpmif_tx_request_t *tx;
++
++ tx = &tpmif->tx->ring[0].req;
++ /*
++ * Start copying data at the page with index 'index'
++ * and within that page at offset 'offset'.
++ * Copy a maximum of 'room_left' bytes.
++ */
++ to_copy = min_t(u32, PAGE_SIZE - pg_offset, room_left);
++ while (to_copy > 0) {
++ void *src;
++ struct gnttab_map_grant_ref map_op;
++ struct gnttab_unmap_grant_ref unmap_op;
++
++ tx = &tpmif->tx->ring[i].req;
++
++ gnttab_set_map_op(&map_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, tx->ref, tpmif->domid);
++
++ if (unlikely(HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref,
++ &map_op, 1))) {
++ BUG();
++ }
++
++ if (map_op.status) {
++ DPRINTK(" Grant table operation failure !\n");
++ return -EFAULT;
++ }
++
++ handle = map_op.handle;
++
++ if (to_copy > tx->size) {
++ /*
++ * User requests more than what's available
++ */
++ to_copy = min_t(u32, tx->size, to_copy);
++ }
++
++ DPRINTK("Copying from mapped memory at %08lx\n",
++ (unsigned long)(idx_to_kaddr(tpmif, i) |
++ (tx->addr & ~PAGE_MASK)));
++
++ src = (void *)(idx_to_kaddr(tpmif, i) |
++ ((tx->addr & ~PAGE_MASK) + pg_offset));
++ if (copy_to_buffer(&buffer[offset],
++ src, to_copy, isuserbuffer)) {
++ return -EFAULT;
++ }
++
++ DPRINTK("Data from TPM-FE of domain %d are %d %d %d %d\n",
++ tpmif->domid, buffer[offset], buffer[offset + 1],
++ buffer[offset + 2], buffer[offset + 3]);
++
++ gnttab_set_unmap_op(&unmap_op, idx_to_kaddr(tpmif, i),
++ GNTMAP_host_map, handle);
++
++ if (unlikely
++ (HYPERVISOR_grant_table_op
++ (GNTTABOP_unmap_grant_ref, &unmap_op, 1))) {
++ BUG();
++ }
++
++ offset += to_copy;
++ pg_offset = 0;
++ last_read += to_copy;
++ room_left -= to_copy;
++
++ to_copy = min_t(u32, PAGE_SIZE, room_left);
++ i++;
++ } /* while (to_copy > 0) */
++ /*
++ * Adjust the last_read pointer
++ */
++ pak->last_read = last_read + 4;
++ return offset;
++}
++
++/* ============================================================
++ * The file layer for reading data from this device
++ * ============================================================
++ */
++static int vtpm_op_open(struct inode *inode, struct file *f)
++{
++ int rc = 0;
++ unsigned long flags;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ if (dataex.has_opener == 0) {
++ dataex.has_opener = 1;
++ } else {
++ rc = -EPERM;
++ }
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return rc;
++}
++
++static ssize_t vtpm_op_read(struct file *file,
++ char __user * data, size_t size, loff_t * offset)
++{
++ int ret_size = -ENODATA;
++ struct packet *pak = NULL;
++ unsigned long flags;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ if (dataex.aborted) {
++ dataex.aborted = 0;
++ dataex.copied_so_far = 0;
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return -EIO;
++ }
++
++ if (list_empty(&dataex.pending_pak)) {
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ wait_event_interruptible(dataex.wait_queue,
++ !list_empty(&dataex.pending_pak));
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ dataex.copied_so_far = 0;
++ }
++
++ if (!list_empty(&dataex.pending_pak)) {
++ unsigned int left;
++
++ pak = list_entry(dataex.pending_pak.next, struct packet, next);
++ left = pak->data_len - dataex.copied_so_far;
++ list_del(&pak->next);
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++
++ DPRINTK("size given by app: %zu, available: %u\n", size, left);
++
++ ret_size = min_t(size_t, size, left);
++
++ ret_size = packet_read(pak, ret_size, data, size, 1);
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++
++ if (ret_size < 0) {
++ del_singleshot_timer_sync(&pak->processing_timer);
++ packet_free(pak);
++ dataex.copied_so_far = 0;
++ } else {
++ DPRINTK("Copied %d bytes to user buffer\n", ret_size);
++
++ dataex.copied_so_far += ret_size;
++ if (dataex.copied_so_far >= pak->data_len + 4) {
++ DPRINTK("All data from this packet given to app.\n");
++ /* All data given to app */
++
++ del_singleshot_timer_sync(&pak->
++ processing_timer);
++ list_add_tail(&pak->next, &dataex.current_pak);
++ /*
++ * The more fontends that are handled at the same time,
++ * the more time we give the TPM to process the request.
++ */
++ mod_timer(&pak->processing_timer,
++ jiffies + (num_frontends * 60 * HZ));
++ dataex.copied_so_far = 0;
++ } else {
++ list_add(&pak->next, &dataex.pending_pak);
++ }
++ }
++ }
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++
++ DPRINTK("Returning result from read to app: %d\n", ret_size);
++
++ return ret_size;
++}
++
++/*
++ * Write operation - only works after a previous read operation!
++ */
++static ssize_t vtpm_op_write(struct file *file,
++ const char __user * data, size_t size,
++ loff_t * offset)
++{
++ struct packet *pak;
++ int rc = 0;
++ unsigned int off = 4;
++ unsigned long flags;
++ struct vtpm_resp_hdr vrh;
++
++ /*
++ * Minimum required packet size is:
++ * 4 bytes for instance number
++ * 2 bytes for tag
++ * 4 bytes for paramSize
++ * 4 bytes for the ordinal
++ * sum: 14 bytes
++ */
++ if (size < sizeof (vrh))
++ return -EFAULT;
++
++ if (copy_from_user(&vrh, data, sizeof (vrh)))
++ return -EFAULT;
++
++ /* malformed packet? */
++ if ((off + ntohl(vrh.len_no)) != size)
++ return -EFAULT;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ pak = packet_find_instance(&dataex.current_pak,
++ ntohl(vrh.instance_no));
++
++ if (pak == NULL) {
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ DPRINTK(KERN_ALERT "No associated packet! (inst=%d)\n",
++ ntohl(vrh.instance_no));
++ return -EFAULT;
++ }
++
++ del_singleshot_timer_sync(&pak->processing_timer);
++ list_del(&pak->next);
++
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++
++ /*
++ * The first 'offset' bytes must be the instance number - skip them.
++ */
++ size -= off;
++
++ rc = packet_write(pak, &data[off], size, 1);
++
++ if (rc > 0) {
++ /* I neglected the first 4 bytes */
++ rc += off;
++ }
++ packet_free(pak);
++ return rc;
++}
++
++static int vtpm_op_release(struct inode *inode, struct file *file)
++{
++ unsigned long flags;
++
++ vtpm_release_packets(NULL, 1);
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ dataex.has_opener = 0;
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return 0;
++}
++
++static unsigned int vtpm_op_poll(struct file *file,
++ struct poll_table_struct *pts)
++{
++ unsigned int flags = POLLOUT | POLLWRNORM;
++
++ poll_wait(file, &dataex.wait_queue, pts);
++ if (!list_empty(&dataex.pending_pak)) {
++ flags |= POLLIN | POLLRDNORM;
++ }
++ return flags;
++}
++
++static const struct file_operations vtpm_ops = {
++ .owner = THIS_MODULE,
++ .llseek = no_llseek,
++ .open = vtpm_op_open,
++ .read = vtpm_op_read,
++ .write = vtpm_op_write,
++ .release = vtpm_op_release,
++ .poll = vtpm_op_poll,
++};
++
++static struct miscdevice vtpms_miscdevice = {
++ .minor = 225,
++ .name = "vtpm",
++ .fops = &vtpm_ops,
++};
++
++/***************************************************************
++ Utility functions
++***************************************************************/
++
++static int tpm_send_fail_message(struct packet *pak, u8 req_tag)
++{
++ int rc;
++ static const unsigned char tpm_error_message_fail[] = {
++ 0x00, 0x00,
++ 0x00, 0x00, 0x00, 0x0a,
++ 0x00, 0x00, 0x00, 0x09 /* TPM_FAIL */
++ };
++ unsigned char buffer[sizeof (tpm_error_message_fail)];
++
++ memcpy(buffer, tpm_error_message_fail,
++ sizeof (tpm_error_message_fail));
++ /*
++ * Insert the right response tag depending on the given tag
++ * All response tags are '+3' to the request tag.
++ */
++ buffer[1] = req_tag + 3;
++
++ /*
++ * Write the data to shared memory and notify the front-end
++ */
++ rc = packet_write(pak, buffer, sizeof (buffer), 0);
++
++ return rc;
++}
++
++static int _vtpm_release_packets(struct list_head *head,
++ tpmif_t * tpmif, int send_msgs)
++{
++ int aborted = 0;
++ int c = 0;
++ struct packet *pak;
++ struct list_head *pos, *tmp;
++
++ list_for_each_safe(pos, tmp, head) {
++ pak = list_entry(pos, struct packet, next);
++ c += 1;
++
++ if (tpmif == NULL || pak->tpmif == tpmif) {
++ int can_send = 0;
++
++ del_singleshot_timer_sync(&pak->processing_timer);
++ list_del(&pak->next);
++
++ if (pak->tpmif && pak->tpmif->status == CONNECTED) {
++ can_send = 1;
++ }
++
++ if (send_msgs && can_send) {
++ tpm_send_fail_message(pak, pak->req_tag);
++ }
++ packet_free(pak);
++ if (c == 1)
++ aborted = 1;
++ }
++ }
++ return aborted;
++}
++
++int vtpm_release_packets(tpmif_t * tpmif, int send_msgs)
++{
++ unsigned long flags;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++
++ dataex.aborted = _vtpm_release_packets(&dataex.pending_pak,
++ tpmif,
++ send_msgs);
++ _vtpm_release_packets(&dataex.current_pak, tpmif, send_msgs);
++
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++ return 0;
++}
++
++static int vtpm_queue_packet(struct packet *pak)
++{
++ int rc = 0;
++
++ if (dataex.has_opener) {
++ unsigned long flags;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ list_add_tail(&pak->next, &dataex.pending_pak);
++ /* give the TPM some time to pick up the request */
++ mod_timer(&pak->processing_timer, jiffies + (30 * HZ));
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++
++ wake_up_interruptible(&dataex.wait_queue);
++ } else {
++ rc = -EFAULT;
++ }
++ return rc;
++}
++
++static int vtpm_receive(tpmif_t * tpmif, u32 size)
++{
++ int rc = 0;
++ unsigned char buffer[10];
++ __be32 *native_size;
++ struct packet *pak = packet_alloc(tpmif, size, 0, 0);
++
++ if (!pak)
++ return -ENOMEM;
++ /*
++ * Read 10 bytes from the received buffer to test its
++ * content for validity.
++ */
++ if (sizeof (buffer) != packet_read(pak,
++ sizeof (buffer), buffer,
++ sizeof (buffer), 0)) {
++ goto failexit;
++ }
++ /*
++ * Reset the packet read pointer so we can read all its
++ * contents again.
++ */
++ packet_reset(pak);
++
++ native_size = (__force __be32 *) (&buffer[4 + 2]);
++ /*
++ * Verify that the size of the packet is correct
++ * as indicated and that there's actually someone reading packets.
++ * The minimum size of the packet is '10' for tag, size indicator
++ * and ordinal.
++ */
++ if (size < 10 ||
++ be32_to_cpu(*native_size) != size ||
++ 0 == dataex.has_opener || tpmif->status != CONNECTED) {
++ rc = -EINVAL;
++ goto failexit;
++ } else {
++ rc = vtpm_queue_packet(pak);
++ if (rc < 0)
++ goto failexit;
++ }
++ return 0;
++
++ failexit:
++ if (pak) {
++ tpm_send_fail_message(pak, buffer[4 + 1]);
++ packet_free(pak);
++ }
++ return rc;
++}
++
++/*
++ * Timeout function that gets invoked when a packet has not been processed
++ * during the timeout period.
++ * The packet must be on a list when this function is invoked. This
++ * also means that once its taken off a list, the timer must be
++ * destroyed as well.
++ */
++static void processing_timeout(unsigned long ptr)
++{
++ struct packet *pak = (struct packet *)ptr;
++ unsigned long flags;
++
++ write_lock_irqsave(&dataex.pak_lock, flags);
++ /*
++ * The packet needs to be searched whether it
++ * is still on the list.
++ */
++ if (pak == packet_find_packet(&dataex.pending_pak, pak) ||
++ pak == packet_find_packet(&dataex.current_pak, pak)) {
++ if ((pak->flags & PACKET_FLAG_DISCARD_RESPONSE) == 0) {
++ tpm_send_fail_message(pak, pak->req_tag);
++ }
++ /* discard future responses */
++ pak->flags |= PACKET_FLAG_DISCARD_RESPONSE;
++ }
++
++ write_unlock_irqrestore(&dataex.pak_lock, flags);
++}
++
++static void tpm_tx_action(unsigned long unused);
++static DECLARE_TASKLET(tpm_tx_tasklet, tpm_tx_action, 0);
++
++static struct list_head tpm_schedule_list;
++static spinlock_t tpm_schedule_list_lock;
++
++static inline void maybe_schedule_tx_action(void)
++{
++ smp_mb();
++ tasklet_schedule(&tpm_tx_tasklet);
++}
++
++static inline int __on_tpm_schedule_list(tpmif_t * tpmif)
++{
++ return tpmif->list.next != NULL;
++}
++
++static void remove_from_tpm_schedule_list(tpmif_t * tpmif)
++{
++ spin_lock_irq(&tpm_schedule_list_lock);
++ if (likely(__on_tpm_schedule_list(tpmif))) {
++ list_del(&tpmif->list);
++ tpmif->list.next = NULL;
++ tpmif_put(tpmif);
++ }
++ spin_unlock_irq(&tpm_schedule_list_lock);
++}
++
++static void add_to_tpm_schedule_list_tail(tpmif_t * tpmif)
++{
++ if (__on_tpm_schedule_list(tpmif))
++ return;
++
++ spin_lock_irq(&tpm_schedule_list_lock);
++ if (!__on_tpm_schedule_list(tpmif) && tpmif->active) {
++ list_add_tail(&tpmif->list, &tpm_schedule_list);
++ tpmif_get(tpmif);
++ }
++ spin_unlock_irq(&tpm_schedule_list_lock);
++}
++
++void tpmif_schedule_work(tpmif_t * tpmif)
++{
++ add_to_tpm_schedule_list_tail(tpmif);
++ maybe_schedule_tx_action();
++}
++
++void tpmif_deschedule_work(tpmif_t * tpmif)
++{
++ remove_from_tpm_schedule_list(tpmif);
++}
++
++static void tpm_tx_action(unsigned long unused)
++{
++ struct list_head *ent;
++ tpmif_t *tpmif;
++ tpmif_tx_request_t *tx;
++
++ DPRINTK("%s: Getting data from front-end(s)!\n", __FUNCTION__);
++
++ while (!list_empty(&tpm_schedule_list)) {
++ /* Get a tpmif from the list with work to do. */
++ ent = tpm_schedule_list.next;
++ tpmif = list_entry(ent, tpmif_t, list);
++ tpmif_get(tpmif);
++ remove_from_tpm_schedule_list(tpmif);
++
++ tx = &tpmif->tx->ring[0].req;
++
++ /* pass it up */
++ vtpm_receive(tpmif, tx->size);
++
++ tpmif_put(tpmif);
++ }
++}
++
++irqreturn_t tpmif_be_int(int irq, void *dev_id)
++{
++ tpmif_t *tpmif = (tpmif_t *) dev_id;
++
++ add_to_tpm_schedule_list_tail(tpmif);
++ maybe_schedule_tx_action();
++ return IRQ_HANDLED;
++}
++
++static int __init tpmback_init(void)
++{
++ int rc;
++
++ if ((rc = misc_register(&vtpms_miscdevice)) != 0) {
++ printk(KERN_ALERT
++ "Could not register misc device for TPM BE.\n");
++ return rc;
++ }
++
++ dataex_init(&dataex);
++
++ spin_lock_init(&tpm_schedule_list_lock);
++ INIT_LIST_HEAD(&tpm_schedule_list);
++
++ rc = tpmif_interface_init();
++ if (!rc) {
++ rc = tpmif_xenbus_init();
++ if (rc)
++ tpmif_interface_exit();
++ }
++ if (rc) {
++ misc_deregister(&vtpms_miscdevice);
++ return rc;
++ }
++
++ printk(KERN_ALERT "Successfully initialized TPM backend driver.\n");
++
++ return 0;
++}
++module_init(tpmback_init);
++
++static void __exit tpmback_exit(void)
++{
++ vtpm_release_packets(NULL, 0);
++ tpmif_xenbus_exit();
++ tpmif_interface_exit();
++ misc_deregister(&vtpms_miscdevice);
++}
++module_exit(tpmback_exit)
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/tpmback/xenbus.c b/drivers/xen/tpmback/xenbus.c
+new file mode 100644
+index 0000000..e023b18
+--- /dev/null
++++ b/drivers/xen/tpmback/xenbus.c
+@@ -0,0 +1,288 @@
++/* Xenbus code for tpmif backend
++ Copyright (C) 2005 IBM Corporation
++ Copyright (C) 2005 Rusty Russell <rusty at rustcorp.com.au>
++
++ This program is free software; you can redistribute it and/or modify
++ it under the terms of the GNU General Public License as published by
++ the Free Software Foundation; either version 2 of the License, or
++ (at your option) any later version.
++
++ This program is distributed in the hope that it will be useful,
++ but WITHOUT ANY WARRANTY; without even the implied warranty of
++ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ GNU General Public License for more details.
++
++ You should have received a copy of the GNU General Public License
++ along with this program; if not, write to the Free Software
++ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++*/
++#include <stdarg.h>
++#include <linux/module.h>
++#include <xen/xenbus.h>
++#include "common.h"
++
++struct backend_info
++{
++ struct xenbus_device *dev;
++
++ /* our communications channel */
++ tpmif_t *tpmif;
++
++ long int frontend_id;
++ long int instance; // instance of TPM
++ u8 is_instance_set;// whether instance number has been set
++
++ /* watch front end for changes */
++ struct xenbus_watch backend_watch;
++};
++
++static void maybe_connect(struct backend_info *be);
++static void connect(struct backend_info *be);
++static int connect_ring(struct backend_info *be);
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len);
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state);
++
++long int tpmback_get_instance(struct backend_info *bi)
++{
++ long int res = -1;
++ if (bi && bi->is_instance_set)
++ res = bi->instance;
++ return res;
++}
++
++static int tpmback_remove(struct xenbus_device *dev)
++{
++ struct backend_info *be = dev->dev.driver_data;
++
++ if (!be) return 0;
++
++ if (be->backend_watch.node) {
++ unregister_xenbus_watch(&be->backend_watch);
++ kfree(be->backend_watch.node);
++ be->backend_watch.node = NULL;
++ }
++ if (be->tpmif) {
++ be->tpmif->bi = NULL;
++ vtpm_release_packets(be->tpmif, 0);
++ tpmif_put(be->tpmif);
++ be->tpmif = NULL;
++ }
++ kfree(be);
++ dev->dev.driver_data = NULL;
++ return 0;
++}
++
++static int tpmback_probe(struct xenbus_device *dev,
++ const struct xenbus_device_id *id)
++{
++ int err;
++ struct backend_info *be = kzalloc(sizeof(struct backend_info),
++ GFP_KERNEL);
++
++ if (!be) {
++ xenbus_dev_fatal(dev, -ENOMEM,
++ "allocating backend structure");
++ return -ENOMEM;
++ }
++
++ be->is_instance_set = 0;
++ be->dev = dev;
++ dev->dev.driver_data = be;
++
++ err = xenbus_watch_path2(dev, dev->nodename,
++ "instance", &be->backend_watch,
++ backend_changed);
++ if (err) {
++ goto fail;
++ }
++
++ err = xenbus_switch_state(dev, XenbusStateInitWait);
++ if (err) {
++ goto fail;
++ }
++ return 0;
++fail:
++ tpmback_remove(dev);
++ return err;
++}
++
++
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ int err;
++ long instance;
++ struct backend_info *be
++ = container_of(watch, struct backend_info, backend_watch);
++ struct xenbus_device *dev = be->dev;
++
++ err = xenbus_scanf(XBT_NIL, dev->nodename,
++ "instance","%li", &instance);
++ if (XENBUS_EXIST_ERR(err)) {
++ return;
++ }
++
++ if (err != 1) {
++ xenbus_dev_fatal(dev, err, "reading instance");
++ return;
++ }
++
++ if (be->is_instance_set == 0) {
++ be->instance = instance;
++ be->is_instance_set = 1;
++ }
++}
++
++
++static void frontend_changed(struct xenbus_device *dev,
++ enum xenbus_state frontend_state)
++{
++ struct backend_info *be = dev->dev.driver_data;
++ int err;
++
++ switch (frontend_state) {
++ case XenbusStateInitialising:
++ case XenbusStateInitialised:
++ break;
++
++ case XenbusStateConnected:
++ err = connect_ring(be);
++ if (err) {
++ return;
++ }
++ maybe_connect(be);
++ break;
++
++ case XenbusStateClosing:
++ be->instance = -1;
++ xenbus_switch_state(dev, XenbusStateClosing);
++ break;
++
++ case XenbusStateUnknown: /* keep it here */
++ case XenbusStateClosed:
++ xenbus_switch_state(dev, XenbusStateClosed);
++ device_unregister(&be->dev->dev);
++ tpmback_remove(dev);
++ break;
++
++ default:
++ xenbus_dev_fatal(dev, -EINVAL,
++ "saw state %d at frontend",
++ frontend_state);
++ break;
++ }
++}
++
++
++
++static void maybe_connect(struct backend_info *be)
++{
++ if (be->tpmif == NULL || be->tpmif->status == CONNECTED)
++ return;
++
++ connect(be);
++}
++
++
++static void connect(struct backend_info *be)
++{
++ struct xenbus_transaction xbt;
++ int err;
++ struct xenbus_device *dev = be->dev;
++ unsigned long ready = 1;
++
++again:
++ err = xenbus_transaction_start(&xbt);
++ if (err) {
++ xenbus_dev_fatal(be->dev, err, "starting transaction");
++ return;
++ }
++
++ err = xenbus_printf(xbt, be->dev->nodename,
++ "ready", "%lu", ready);
++ if (err) {
++ xenbus_dev_fatal(be->dev, err, "writing 'ready'");
++ goto abort;
++ }
++
++ err = xenbus_transaction_end(xbt, 0);
++ if (err == -EAGAIN)
++ goto again;
++ if (err)
++ xenbus_dev_fatal(be->dev, err, "end of transaction");
++
++ err = xenbus_switch_state(dev, XenbusStateConnected);
++ if (!err)
++ be->tpmif->status = CONNECTED;
++ return;
++abort:
++ xenbus_transaction_end(xbt, 1);
++}
++
++
++static int connect_ring(struct backend_info *be)
++{
++ struct xenbus_device *dev = be->dev;
++ unsigned long ring_ref;
++ unsigned int evtchn;
++ int err;
++
++ err = xenbus_gather(XBT_NIL, dev->otherend,
++ "ring-ref", "%lu", &ring_ref,
++ "event-channel", "%u", &evtchn, NULL);
++ if (err) {
++ xenbus_dev_error(dev, err,
++ "reading %s/ring-ref and event-channel",
++ dev->otherend);
++ return err;
++ }
++
++ if (!be->tpmif) {
++ be->tpmif = tpmif_find(dev->otherend_id, be);
++ if (IS_ERR(be->tpmif)) {
++ err = PTR_ERR(be->tpmif);
++ be->tpmif = NULL;
++ xenbus_dev_fatal(dev,err,"creating vtpm interface");
++ return err;
++ }
++ }
++
++ if (be->tpmif != NULL) {
++ err = tpmif_map(be->tpmif, ring_ref, evtchn);
++ if (err) {
++ xenbus_dev_error(dev, err,
++ "mapping shared-frame %lu port %u",
++ ring_ref, evtchn);
++ return err;
++ }
++ }
++ return 0;
++}
++
++
++static const struct xenbus_device_id tpmback_ids[] = {
++ { "vtpm" },
++ { "" }
++};
++
++
++static struct xenbus_driver tpmback = {
++ .name = "vtpm",
++ .ids = tpmback_ids,
++ .probe = tpmback_probe,
++ .remove = tpmback_remove,
++ .otherend_changed = frontend_changed,
++};
++
++
++int tpmif_xenbus_init(void)
++{
++ return xenbus_register_backend(&tpmback);
++}
++
++void tpmif_xenbus_exit(void)
++{
++ xenbus_unregister_driver(&tpmback);
++}
+diff --git a/drivers/xen/util.c b/drivers/xen/util.c
+new file mode 100644
+index 0000000..48dadee
+--- /dev/null
++++ b/drivers/xen/util.c
+@@ -0,0 +1,20 @@
++#include <linux/err.h>
++#include <linux/module.h>
++#include <xen/driver_util.h>
++
++struct class *get_xen_class(void)
++{
++ static struct class *xen_class;
++
++ if (xen_class)
++ return xen_class;
++
++ xen_class = class_create(THIS_MODULE, "xen");
++ if (IS_ERR(xen_class)) {
++ printk("Failed to create xen sysfs class.\n");
++ xen_class = NULL;
++ }
++
++ return xen_class;
++}
++EXPORT_SYMBOL_GPL(get_xen_class);
+diff --git a/drivers/xen/xenbus/Makefile b/drivers/xen/xenbus/Makefile
+index 5571f5b..d7091dc 100644
+--- a/drivers/xen/xenbus/Makefile
++++ b/drivers/xen/xenbus/Makefile
+@@ -1,7 +1,9 @@
+-obj-y += xenbus.o
++obj-y += xenbus_client.o xenbus_comms.o xenbus_xs.o xenbus_probe.o
++obj-$(CONFIG_XEN_BACKEND) += xenbus_be.o
+
+-xenbus-objs =
+-xenbus-objs += xenbus_client.o
+-xenbus-objs += xenbus_comms.o
+-xenbus-objs += xenbus_xs.o
+-xenbus-objs += xenbus_probe.o
++xenbus_be-objs =
++xenbus_be-objs += xenbus_backend_client.o
++
++xenbus-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o
++obj-y += $(xenbus-y) $(xenbus-m)
++obj-$(CONFIG_XEN_XENBUS_DEV) += xenbus_dev.o
+diff --git a/drivers/xen/xenbus/xenbus_backend_client.c b/drivers/xen/xenbus/xenbus_backend_client.c
+new file mode 100644
+index 0000000..9b6fbc9
+--- /dev/null
++++ b/drivers/xen/xenbus/xenbus_backend_client.c
+@@ -0,0 +1,147 @@
++/******************************************************************************
++ * Backend-client-facing interface for the Xenbus driver. In other words, the
++ * interface between the Xenbus and the device-specific code in the backend
++ * driver.
++ *
++ * Copyright (C) 2005-2006 XenSource Ltd
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/err.h>
++#include <xen/gnttab.h>
++#include <xen/xenbus.h>
++#include <xen/driver_util.h>
++
++/* Based on Rusty Russell's skeleton driver's map_page */
++struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev, int gnt_ref)
++{
++ struct gnttab_map_grant_ref op;
++ struct vm_struct *area;
++
++ area = alloc_vm_area(PAGE_SIZE);
++ if (!area)
++ return ERR_PTR(-ENOMEM);
++
++ gnttab_set_map_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
++ gnt_ref, dev->otherend_id);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status != GNTST_okay) {
++ free_vm_area(area);
++ xenbus_dev_fatal(dev, op.status,
++ "mapping in shared page %d from domain %d",
++ gnt_ref, dev->otherend_id);
++ BUG_ON(!IS_ERR(ERR_PTR(op.status)));
++ return ERR_PTR(op.status);
++ }
++
++ /* Stuff the handle in an unused field */
++ area->phys_addr = (unsigned long)op.handle;
++
++ return area;
++}
++EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc);
++
++
++int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
++ grant_handle_t *handle, void *vaddr)
++{
++ struct gnttab_map_grant_ref op;
++
++ gnttab_set_map_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
++ gnt_ref, dev->otherend_id);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status != GNTST_okay) {
++ xenbus_dev_fatal(dev, op.status,
++ "mapping in shared page %d from domain %d",
++ gnt_ref, dev->otherend_id);
++ } else
++ *handle = op.handle;
++
++ return op.status;
++}
++EXPORT_SYMBOL_GPL(xenbus_map_ring);
++
++
++/* Based on Rusty Russell's skeleton driver's unmap_page */
++int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *area)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)area->addr, GNTMAP_host_map,
++ (grant_handle_t)area->phys_addr);
++
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status == GNTST_okay)
++ free_vm_area(area);
++ else
++ xenbus_dev_error(dev, op.status,
++ "unmapping page at handle %d error %d",
++ (int16_t)area->phys_addr, op.status);
++
++ return op.status;
++}
++EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree);
++
++
++int xenbus_unmap_ring(struct xenbus_device *dev,
++ grant_handle_t handle, void *vaddr)
++{
++ struct gnttab_unmap_grant_ref op;
++
++ gnttab_set_unmap_op(&op, (unsigned long)vaddr, GNTMAP_host_map,
++ handle);
++ if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
++ BUG();
++
++ if (op.status != GNTST_okay)
++ xenbus_dev_error(dev, op.status,
++ "unmapping page at handle %d error %d",
++ handle, op.status);
++
++ return op.status;
++}
++EXPORT_SYMBOL_GPL(xenbus_unmap_ring);
++
++int xenbus_dev_is_online(struct xenbus_device *dev)
++{
++ int rc, val;
++
++ rc = xenbus_scanf(XBT_NIL, dev->nodename, "online", "%d", &val);
++ if (rc != 1)
++ val = 0; /* no online node present */
++
++ return val;
++}
++EXPORT_SYMBOL_GPL(xenbus_dev_is_online);
++
++MODULE_LICENSE("Dual BSD/GPL");
+diff --git a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c
+index 0f86b0f..621ffc8 100644
+--- a/drivers/xen/xenbus/xenbus_client.c
++++ b/drivers/xen/xenbus/xenbus_client.c
+@@ -30,6 +30,12 @@
+ * IN THE SOFTWARE.
+ */
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++#include <linux/slab.h>
++#include <xen/evtchn.h>
++#include <xen/gnttab.h>
++#include <xen/driver_util.h>
++#else
+ #include <linux/types.h>
+ #include <linux/vmalloc.h>
+ #include <asm/xen/hypervisor.h>
+@@ -37,8 +43,13 @@
+ #include <xen/interface/event_channel.h>
+ #include <xen/events.h>
+ #include <xen/grant_table.h>
++#endif
+ #include <xen/xenbus.h>
+
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
+ const char *xenbus_strstate(enum xenbus_state state)
+ {
+ static const char *const name[] = {
+@@ -91,6 +102,26 @@ int xenbus_watch_path(struct xenbus_device *dev, const char *path,
+ EXPORT_SYMBOL_GPL(xenbus_watch_path);
+
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
++ const char *path2, struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int))
++{
++ int err;
++ char *state = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", path, path2);
++ if (!state) {
++ xenbus_dev_fatal(dev, -ENOMEM, "allocating path for watch");
++ return -ENOMEM;
++ }
++ err = xenbus_watch_path(dev, state, watch, callback);
++
++ if (err)
++ kfree(state);
++ return err;
++}
++EXPORT_SYMBOL_GPL(xenbus_watch_path2);
++#else
+ /**
+ * xenbus_watch_pathfmt - register a watch on a sprintf-formatted path
+ * @dev: xenbus device
+@@ -131,6 +162,7 @@ int xenbus_watch_pathfmt(struct xenbus_device *dev,
+ return err;
+ }
+ EXPORT_SYMBOL_GPL(xenbus_watch_pathfmt);
++#endif
+
+
+ /**
+@@ -201,13 +233,12 @@ static char *error_path(struct xenbus_device *dev)
+ }
+
+
+-static void xenbus_va_dev_error(struct xenbus_device *dev, int err,
+- const char *fmt, va_list ap)
++static void _dev_error(struct xenbus_device *dev, int err,
++ const char *fmt, va_list ap)
+ {
+ int ret;
+ unsigned int len;
+- char *printf_buffer = NULL;
+- char *path_buffer = NULL;
++ char *printf_buffer = NULL, *path_buffer = NULL;
+
+ #define PRINTF_BUFFER_SIZE 4096
+ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL);
+@@ -224,20 +255,24 @@ static void xenbus_va_dev_error(struct xenbus_device *dev, int err,
+ path_buffer = error_path(dev);
+
+ if (path_buffer == NULL) {
+- dev_err(&dev->dev, "failed to write error node for %s (%s)\n",
+- dev->nodename, printf_buffer);
++ dev_err(&dev->dev,
++ "xenbus: failed to write error node for %s (%s)\n",
++ dev->nodename, printf_buffer);
+ goto fail;
+ }
+
+ if (xenbus_write(XBT_NIL, path_buffer, "error", printf_buffer) != 0) {
+- dev_err(&dev->dev, "failed to write error node for %s (%s)\n",
+- dev->nodename, printf_buffer);
++ dev_err(&dev->dev,
++ "xenbus: failed to write error node for %s (%s)\n",
++ dev->nodename, printf_buffer);
+ goto fail;
+ }
+
+ fail:
+- kfree(printf_buffer);
+- kfree(path_buffer);
++ if (printf_buffer)
++ kfree(printf_buffer);
++ if (path_buffer)
++ kfree(path_buffer);
+ }
+
+
+@@ -250,16 +285,18 @@ fail:
+ * Report the given negative errno into the store, along with the given
+ * formatted message.
+ */
+-void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt, ...)
++void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
++ ...)
+ {
+ va_list ap;
+
+ va_start(ap, fmt);
+- xenbus_va_dev_error(dev, err, fmt, ap);
++ _dev_error(dev, err, fmt, ap);
+ va_end(ap);
+ }
+ EXPORT_SYMBOL_GPL(xenbus_dev_error);
+
++
+ /**
+ * xenbus_dev_fatal
+ * @dev: xenbus device
+@@ -270,24 +307,25 @@ EXPORT_SYMBOL_GPL(xenbus_dev_error);
+ * xenbus_switch_state(dev, NULL, XenbusStateClosing) to schedule an orderly
+ * closedown of this driver and its peer.
+ */
+-
+-void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt, ...)
++void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
++ ...)
+ {
+ va_list ap;
+
+ va_start(ap, fmt);
+- xenbus_va_dev_error(dev, err, fmt, ap);
++ _dev_error(dev, err, fmt, ap);
+ va_end(ap);
+
+ xenbus_switch_state(dev, XenbusStateClosing);
+ }
+ EXPORT_SYMBOL_GPL(xenbus_dev_fatal);
+
++
+ /**
+ * xenbus_grant_ring
+ * @dev: xenbus device
+ * @ring_mfn: mfn of ring to grant
+-
++ *
+ * Grant access to the given @ring_mfn to the peer of the given device. Return
+ * 0 on success, or -errno on error. On error, the device will switch to
+ * XenbusStateClosing, and the error will be saved in the store.
+@@ -313,7 +351,7 @@ int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port)
+ struct evtchn_alloc_unbound alloc_unbound;
+ int err;
+
+- alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.dom = DOMID_SELF;
+ alloc_unbound.remote_dom = dev->otherend_id;
+
+ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
+@@ -328,6 +366,7 @@ int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port)
+ EXPORT_SYMBOL_GPL(xenbus_alloc_evtchn);
+
+
++#if 0 /* !defined(CONFIG_XEN) && !defined(MODULE) */
+ /**
+ * Bind to an existing interdomain event channel in another domain. Returns 0
+ * on success and stores the local port in *port. On error, returns -errno,
+@@ -353,6 +392,7 @@ int xenbus_bind_evtchn(struct xenbus_device *dev, int remote_port, int *port)
+ return err;
+ }
+ EXPORT_SYMBOL_GPL(xenbus_bind_evtchn);
++#endif
+
+
+ /**
+@@ -374,6 +414,7 @@ int xenbus_free_evtchn(struct xenbus_device *dev, int port)
+ EXPORT_SYMBOL_GPL(xenbus_free_evtchn);
+
+
++#if 0 /* !defined(CONFIG_XEN) && !defined(MODULE) */
+ /**
+ * xenbus_map_ring_valloc
+ * @dev: xenbus device
+@@ -548,6 +589,7 @@ int xenbus_unmap_ring(struct xenbus_device *dev,
+ return op.status;
+ }
+ EXPORT_SYMBOL_GPL(xenbus_unmap_ring);
++#endif
+
+
+ /**
+diff --git a/drivers/xen/xenbus/xenbus_comms.c b/drivers/xen/xenbus/xenbus_comms.c
+index 6efbe3f..9b14602 100644
+--- a/drivers/xen/xenbus/xenbus_comms.c
++++ b/drivers/xen/xenbus/xenbus_comms.c
+@@ -35,13 +35,25 @@
+ #include <linux/sched.h>
+ #include <linux/err.h>
+ #include <xen/xenbus.h>
++#if defined(CONFIG_XEN) || defined(MODULE)
++#include <xen/evtchn.h>
++#include <asm/hypervisor.h>
++#else
+ #include <asm/xen/hypervisor.h>
+ #include <xen/events.h>
+ #include <xen/page.h>
++#endif
++
+ #include "xenbus_comms.h"
+
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
+ static int xenbus_irq;
+
++extern void xenbus_probe(struct work_struct *);
++extern int xenstored_ready;
+ static DECLARE_WORK(probe_work, xenbus_probe);
+
+ static DECLARE_WAIT_QUEUE_HEAD(xb_waitq);
+@@ -219,7 +231,11 @@ int xb_init_comms(void)
+ if (xenbus_irq)
+ unbind_from_irqhandler(xenbus_irq, &xb_waitq);
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++ err = bind_caller_port_to_irqhandler(
++#else
+ err = bind_evtchn_to_irqhandler(
++#endif
+ xen_store_evtchn, wake_waiting,
+ 0, "xenbus", &xb_waitq);
+ if (err <= 0) {
+diff --git a/drivers/xen/xenbus/xenbus_dev.c b/drivers/xen/xenbus/xenbus_dev.c
+new file mode 100644
+index 0000000..2db77f0
+--- /dev/null
++++ b/drivers/xen/xenbus/xenbus_dev.c
+@@ -0,0 +1,408 @@
++/*
++ * xenbus_dev.c
++ *
++ * Driver giving user-space access to the kernel's xenbus connection
++ * to xenstore.
++ *
++ * Copyright (c) 2005, Christian Limpach
++ * Copyright (c) 2005, Rusty Russell, IBM Corporation
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#include <linux/kernel.h>
++#include <linux/errno.h>
++#include <linux/uio.h>
++#include <linux/notifier.h>
++#include <linux/wait.h>
++#include <linux/fs.h>
++#include <linux/poll.h>
++#include <linux/mutex.h>
++
++#include "xenbus_comms.h"
++
++#include <asm/uaccess.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <asm/hypervisor.h>
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++struct xenbus_dev_transaction {
++ struct list_head list;
++ struct xenbus_transaction handle;
++};
++
++struct read_buffer {
++ struct list_head list;
++ unsigned int cons;
++ unsigned int len;
++ char msg[];
++};
++
++struct xenbus_dev_data {
++ /* In-progress transaction. */
++ struct list_head transactions;
++
++ /* Active watches. */
++ struct list_head watches;
++
++ /* Partial request. */
++ unsigned int len;
++ union {
++ struct xsd_sockmsg msg;
++ char buffer[PAGE_SIZE];
++ } u;
++
++ /* Response queue. */
++ struct list_head read_buffers;
++ wait_queue_head_t read_waitq;
++
++ struct mutex reply_mutex;
++};
++
++static struct proc_dir_entry *xenbus_dev_intf;
++
++static ssize_t xenbus_dev_read(struct file *filp,
++ char __user *ubuf,
++ size_t len, loff_t *ppos)
++{
++ struct xenbus_dev_data *u = filp->private_data;
++ struct read_buffer *rb;
++ int i, ret;
++
++ mutex_lock(&u->reply_mutex);
++ while (list_empty(&u->read_buffers)) {
++ mutex_unlock(&u->reply_mutex);
++ ret = wait_event_interruptible(u->read_waitq,
++ !list_empty(&u->read_buffers));
++ if (ret)
++ return ret;
++ mutex_lock(&u->reply_mutex);
++ }
++
++ rb = list_entry(u->read_buffers.next, struct read_buffer, list);
++ for (i = 0; i < len;) {
++ put_user(rb->msg[rb->cons], ubuf + i);
++ i++;
++ rb->cons++;
++ if (rb->cons == rb->len) {
++ list_del(&rb->list);
++ kfree(rb);
++ if (list_empty(&u->read_buffers))
++ break;
++ rb = list_entry(u->read_buffers.next,
++ struct read_buffer, list);
++ }
++ }
++ mutex_unlock(&u->reply_mutex);
++
++ return i;
++}
++
++static void queue_reply(struct xenbus_dev_data *u,
++ char *data, unsigned int len)
++{
++ struct read_buffer *rb;
++
++ if (len == 0)
++ return;
++
++ rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
++ BUG_ON(rb == NULL);
++
++ rb->cons = 0;
++ rb->len = len;
++
++ memcpy(rb->msg, data, len);
++
++ list_add_tail(&rb->list, &u->read_buffers);
++
++ wake_up(&u->read_waitq);
++}
++
++struct watch_adapter
++{
++ struct list_head list;
++ struct xenbus_watch watch;
++ struct xenbus_dev_data *dev_data;
++ char *token;
++};
++
++static void free_watch_adapter (struct watch_adapter *watch)
++{
++ kfree(watch->watch.node);
++ kfree(watch->token);
++ kfree(watch);
++}
++
++static void watch_fired(struct xenbus_watch *watch,
++ const char **vec,
++ unsigned int len)
++{
++ struct watch_adapter *adap =
++ container_of(watch, struct watch_adapter, watch);
++ struct xsd_sockmsg hdr;
++ const char *path, *token;
++ int path_len, tok_len, body_len, data_len = 0;
++
++ path = vec[XS_WATCH_PATH];
++ token = adap->token;
++
++ path_len = strlen(path) + 1;
++ tok_len = strlen(token) + 1;
++ if (len > 2)
++ data_len = vec[len] - vec[2] + 1;
++ body_len = path_len + tok_len + data_len;
++
++ hdr.type = XS_WATCH_EVENT;
++ hdr.len = body_len;
++
++ mutex_lock(&adap->dev_data->reply_mutex);
++ queue_reply(adap->dev_data, (char *)&hdr, sizeof(hdr));
++ queue_reply(adap->dev_data, (char *)path, path_len);
++ queue_reply(adap->dev_data, (char *)token, tok_len);
++ if (len > 2)
++ queue_reply(adap->dev_data, (char *)vec[2], data_len);
++ mutex_unlock(&adap->dev_data->reply_mutex);
++}
++
++static LIST_HEAD(watch_list);
++
++static ssize_t xenbus_dev_write(struct file *filp,
++ const char __user *ubuf,
++ size_t len, loff_t *ppos)
++{
++ struct xenbus_dev_data *u = filp->private_data;
++ struct xenbus_dev_transaction *trans = NULL;
++ uint32_t msg_type;
++ void *reply;
++ char *path, *token;
++ struct watch_adapter *watch, *tmp_watch;
++ int err, rc = len;
++
++ if ((len + u->len) > sizeof(u->u.buffer)) {
++ rc = -EINVAL;
++ goto out;
++ }
++
++ if (copy_from_user(u->u.buffer + u->len, ubuf, len) != 0) {
++ rc = -EFAULT;
++ goto out;
++ }
++
++ u->len += len;
++ if ((u->len < sizeof(u->u.msg)) ||
++ (u->len < (sizeof(u->u.msg) + u->u.msg.len)))
++ return rc;
++
++ msg_type = u->u.msg.type;
++
++ switch (msg_type) {
++ case XS_TRANSACTION_START:
++ case XS_TRANSACTION_END:
++ case XS_DIRECTORY:
++ case XS_READ:
++ case XS_GET_PERMS:
++ case XS_RELEASE:
++ case XS_GET_DOMAIN_PATH:
++ case XS_WRITE:
++ case XS_MKDIR:
++ case XS_RM:
++ case XS_SET_PERMS:
++ if (msg_type == XS_TRANSACTION_START) {
++ trans = kmalloc(sizeof(*trans), GFP_KERNEL);
++ if (!trans) {
++ rc = -ENOMEM;
++ goto out;
++ }
++ }
++
++ reply = xenbus_dev_request_and_reply(&u->u.msg);
++ if (IS_ERR(reply)) {
++ kfree(trans);
++ rc = PTR_ERR(reply);
++ goto out;
++ }
++
++ if (msg_type == XS_TRANSACTION_START) {
++ trans->handle.id = simple_strtoul(reply, NULL, 0);
++ list_add(&trans->list, &u->transactions);
++ } else if (msg_type == XS_TRANSACTION_END) {
++ list_for_each_entry(trans, &u->transactions, list)
++ if (trans->handle.id == u->u.msg.tx_id)
++ break;
++ BUG_ON(&trans->list == &u->transactions);
++ list_del(&trans->list);
++ kfree(trans);
++ }
++ mutex_lock(&u->reply_mutex);
++ queue_reply(u, (char *)&u->u.msg, sizeof(u->u.msg));
++ queue_reply(u, (char *)reply, u->u.msg.len);
++ mutex_unlock(&u->reply_mutex);
++ kfree(reply);
++ break;
++
++ case XS_WATCH:
++ case XS_UNWATCH: {
++ static const char *XS_RESP = "OK";
++ struct xsd_sockmsg hdr;
++
++ path = u->u.buffer + sizeof(u->u.msg);
++ token = memchr(path, 0, u->u.msg.len);
++ if (token == NULL) {
++ rc = -EILSEQ;
++ goto out;
++ }
++ token++;
++
++ if (msg_type == XS_WATCH) {
++ watch = kzalloc(sizeof(*watch), GFP_KERNEL);
++ watch->watch.node = kmalloc(strlen(path)+1,
++ GFP_KERNEL);
++ strcpy((char *)watch->watch.node, path);
++ watch->watch.callback = watch_fired;
++ watch->token = kmalloc(strlen(token)+1, GFP_KERNEL);
++ strcpy(watch->token, token);
++ watch->dev_data = u;
++
++ err = register_xenbus_watch(&watch->watch);
++ if (err) {
++ free_watch_adapter(watch);
++ rc = err;
++ goto out;
++ }
++
++ list_add(&watch->list, &u->watches);
++ } else {
++ list_for_each_entry_safe(watch, tmp_watch,
++ &u->watches, list) {
++ if (!strcmp(watch->token, token) &&
++ !strcmp(watch->watch.node, path))
++ {
++ unregister_xenbus_watch(&watch->watch);
++ list_del(&watch->list);
++ free_watch_adapter(watch);
++ break;
++ }
++ }
++ }
++
++ hdr.type = msg_type;
++ hdr.len = strlen(XS_RESP) + 1;
++ mutex_lock(&u->reply_mutex);
++ queue_reply(u, (char *)&hdr, sizeof(hdr));
++ queue_reply(u, (char *)XS_RESP, hdr.len);
++ mutex_unlock(&u->reply_mutex);
++ break;
++ }
++
++ default:
++ rc = -EINVAL;
++ break;
++ }
++
++ out:
++ u->len = 0;
++ return rc;
++}
++
++static int xenbus_dev_open(struct inode *inode, struct file *filp)
++{
++ struct xenbus_dev_data *u;
++
++ if (xen_store_evtchn == 0)
++ return -ENOENT;
++
++ nonseekable_open(inode, filp);
++
++ u = kzalloc(sizeof(*u), GFP_KERNEL);
++ if (u == NULL)
++ return -ENOMEM;
++
++ INIT_LIST_HEAD(&u->transactions);
++ INIT_LIST_HEAD(&u->watches);
++ INIT_LIST_HEAD(&u->read_buffers);
++ init_waitqueue_head(&u->read_waitq);
++
++ mutex_init(&u->reply_mutex);
++
++ filp->private_data = u;
++
++ return 0;
++}
++
++static int xenbus_dev_release(struct inode *inode, struct file *filp)
++{
++ struct xenbus_dev_data *u = filp->private_data;
++ struct xenbus_dev_transaction *trans, *tmp;
++ struct watch_adapter *watch, *tmp_watch;
++
++ list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
++ xenbus_transaction_end(trans->handle, 1);
++ list_del(&trans->list);
++ kfree(trans);
++ }
++
++ list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
++ unregister_xenbus_watch(&watch->watch);
++ list_del(&watch->list);
++ free_watch_adapter(watch);
++ }
++
++ kfree(u);
++
++ return 0;
++}
++
++static unsigned int xenbus_dev_poll(struct file *file, poll_table *wait)
++{
++ struct xenbus_dev_data *u = file->private_data;
++
++ poll_wait(file, &u->read_waitq, wait);
++ if (!list_empty(&u->read_buffers))
++ return POLLIN | POLLRDNORM;
++ return 0;
++}
++
++static const struct file_operations xenbus_dev_file_ops = {
++ .read = xenbus_dev_read,
++ .write = xenbus_dev_write,
++ .open = xenbus_dev_open,
++ .release = xenbus_dev_release,
++ .poll = xenbus_dev_poll,
++};
++
++int xenbus_dev_init(void)
++{
++ xenbus_dev_intf = create_xen_proc_entry("xenbus", 0400);
++ if (xenbus_dev_intf)
++ xenbus_dev_intf->proc_fops = &xenbus_dev_file_ops;
++
++ return 0;
++}
+diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c
+index 57ceb53..db28d44 100644
+--- a/drivers/xen/xenbus/xenbus_probe.c
++++ b/drivers/xen/xenbus/xenbus_probe.c
+@@ -4,6 +4,7 @@
+ * Copyright (C) 2005 Rusty Russell, IBM Corporation
+ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
+ * Copyright (C) 2005, 2006 XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+@@ -32,7 +33,7 @@
+
+ #define DPRINTK(fmt, args...) \
+ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
+- __func__, __LINE__, ##args)
++ __FUNCTION__, __LINE__, ##args)
+
+ #include <linux/kernel.h>
+ #include <linux/err.h>
+@@ -41,24 +42,40 @@
+ #include <linux/fcntl.h>
+ #include <linux/mm.h>
+ #include <linux/notifier.h>
+-#include <linux/kthread.h>
+ #include <linux/mutex.h>
+ #include <linux/io.h>
+
+ #include <asm/page.h>
+ #include <asm/pgtable.h>
++#if defined(CONFIG_XEN) || defined(MODULE)
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <xen/evtchn.h>
++#include <xen/features.h>
++#ifdef MODULE
++#include <xen/hvm.h>
++#endif
++#else
+ #include <asm/xen/hypervisor.h>
+ #include <xen/xenbus.h>
+ #include <xen/events.h>
+ #include <xen/page.h>
++#endif
+
+ #include "xenbus_comms.h"
+ #include "xenbus_probe.h"
+
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
+ int xen_store_evtchn;
+ struct xenstore_domain_interface *xen_store_interface;
+ static unsigned long xen_store_mfn;
+
++extern struct mutex xenwatch_mutex;
++
+ static BLOCKING_NOTIFIER_HEAD(xenstore_chain);
+
+ static void wait_for_devices(struct xenbus_driver *xendrv);
+@@ -88,16 +105,6 @@ int xenbus_match(struct device *_dev, struct device_driver *_drv)
+ return match_device(drv->ids, to_xenbus_device(_dev)) != NULL;
+ }
+
+-static int xenbus_uevent(struct device *_dev, struct kobj_uevent_env *env)
+-{
+- struct xenbus_device *dev = to_xenbus_device(_dev);
+-
+- if (add_uevent_var(env, "MODALIAS=xen:%s", dev->devicetype))
+- return -ENOMEM;
+-
+- return 0;
+-}
+-
+ /* device/<type>/<id> => <type>-<id> */
+ static int frontend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
+ {
+@@ -166,6 +173,27 @@ static int read_backend_details(struct xenbus_device *xendev)
+ return read_otherend_details(xendev, "backend-id", "backend");
+ }
+
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
++static int xenbus_uevent_frontend(struct device *dev, struct kobj_uevent_env *env)
++{
++ struct xenbus_device *xdev;
++
++ if (dev == NULL)
++ return -ENODEV;
++ xdev = to_xenbus_device(dev);
++ if (xdev == NULL)
++ return -ENODEV;
++
++ /* stuff we want to pass to /sbin/hotplug */
++#if defined(CONFIG_XEN) || defined(MODULE)
++ add_uevent_var(env, "XENBUS_TYPE=%s", xdev->devicetype);
++ add_uevent_var(env, "XENBUS_PATH=%s", xdev->nodename);
++#endif
++ add_uevent_var(env, "MODALIAS=xen:%s", xdev->devicetype);
++
++ return 0;
++}
++#endif
+
+ /* Bus type for frontend drivers. */
+ static struct xen_bus_type xenbus_frontend = {
+@@ -173,14 +201,22 @@ static struct xen_bus_type xenbus_frontend = {
+ .levels = 2, /* device/type/<id> */
+ .get_bus_id = frontend_bus_id,
+ .probe = xenbus_probe_frontend,
++ .error = -ENODEV,
+ .bus = {
+ .name = "xen",
+ .match = xenbus_match,
+- .uevent = xenbus_uevent,
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
+ .probe = xenbus_dev_probe,
+ .remove = xenbus_dev_remove,
+ .shutdown = xenbus_dev_shutdown,
++ .uevent = xenbus_uevent_frontend,
++#endif
+ },
++#if defined(CONFIG_XEN) || defined(MODULE)
++ .dev = {
++ .bus_id = "xen",
++ },
++#endif
+ };
+
+ static void otherend_changed(struct xenbus_watch *watch,
+@@ -196,17 +232,17 @@ static void otherend_changed(struct xenbus_watch *watch,
+ if (!dev->otherend ||
+ strncmp(dev->otherend, vec[XS_WATCH_PATH],
+ strlen(dev->otherend))) {
+- dev_dbg(&dev->dev, "Ignoring watch at %s\n",
+- vec[XS_WATCH_PATH]);
++ dev_dbg(&dev->dev, "Ignoring watch at %s", vec[XS_WATCH_PATH]);
+ return;
+ }
+
+ state = xenbus_read_driver_state(dev->otherend);
+
+- dev_dbg(&dev->dev, "state is %d, (%s), %s, %s\n",
++ dev_dbg(&dev->dev, "state is %d (%s), %s, %s",
+ state, xenbus_strstate(state), dev->otherend_watch.node,
+ vec[XS_WATCH_PATH]);
+
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,16)
+ /*
+ * Ignore xenbus transitions during shutdown. This prevents us doing
+ * work that can fail e.g., when the rootfs is gone.
+@@ -220,6 +256,7 @@ static void otherend_changed(struct xenbus_watch *watch,
+ xenbus_frontend_closed(dev);
+ return;
+ }
++#endif
+
+ if (drv->otherend_changed)
+ drv->otherend_changed(dev, state);
+@@ -239,8 +276,13 @@ static int talk_to_otherend(struct xenbus_device *dev)
+
+ static int watch_otherend(struct xenbus_device *dev)
+ {
++#if defined(CONFIG_XEN) || defined(MODULE)
++ return xenbus_watch_path2(dev, dev->otherend, "state",
++ &dev->otherend_watch, otherend_changed);
++#else
+ return xenbus_watch_pathfmt(dev, &dev->otherend_watch, otherend_changed,
+ "%s/%s", dev->otherend, "state");
++#endif
+ }
+
+
+@@ -266,8 +308,9 @@ int xenbus_dev_probe(struct device *_dev)
+
+ err = talk_to_otherend(dev);
+ if (err) {
+- dev_warn(&dev->dev, "talk_to_otherend on %s failed.\n",
+- dev->nodename);
++ dev_warn(&dev->dev,
++ "xenbus_probe: talk_to_otherend on %s failed.\n",
++ dev->nodename);
+ return err;
+ }
+
+@@ -277,8 +320,9 @@ int xenbus_dev_probe(struct device *_dev)
+
+ err = watch_otherend(dev);
+ if (err) {
+- dev_warn(&dev->dev, "watch_otherend on %s failed.\n",
+- dev->nodename);
++ dev_warn(&dev->dev,
++ "xenbus_probe: watch_otherend on %s failed.\n",
++ dev->nodename);
+ return err;
+ }
+
+@@ -313,17 +357,20 @@ static void xenbus_dev_shutdown(struct device *_dev)
+
+ DPRINTK("%s", dev->nodename);
+
++ if (is_initial_xendomain())
++ return;
++
+ get_device(&dev->dev);
+ if (dev->state != XenbusStateConnected) {
+- printk(KERN_INFO "%s: %s: %s != Connected, skipping\n", __func__,
+- dev->nodename, xenbus_strstate(dev->state));
++ dev_info(&dev->dev, "%s: %s: %s != Connected, skipping\n", __FUNCTION__,
++ dev->nodename, xenbus_strstate(dev->state));
+ goto out;
+ }
+ xenbus_switch_state(dev, XenbusStateClosing);
+ timeout = wait_for_completion_timeout(&dev->down, timeout);
+ if (!timeout)
+- printk(KERN_INFO "%s: %s timeout closing device\n",
+- __func__, dev->nodename);
++ dev_info(&dev->dev, "%s: %s timeout closing device\n",
++ __FUNCTION__, dev->nodename);
+ out:
+ put_device(&dev->dev);
+ }
+@@ -333,12 +380,29 @@ int xenbus_register_driver_common(struct xenbus_driver *drv,
+ struct module *owner,
+ const char *mod_name)
+ {
++ int ret;
++
++ if (bus->error)
++ return bus->error;
++
+ drv->driver.name = drv->name;
+ drv->driver.bus = &bus->bus;
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,10)
+ drv->driver.owner = owner;
++#endif
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,21)
+ drv->driver.mod_name = mod_name;
++#endif
++#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16)
++ drv->driver.probe = xenbus_dev_probe;
++ drv->driver.remove = xenbus_dev_remove;
++ drv->driver.shutdown = xenbus_dev_shutdown;
++#endif
+
+- return driver_register(&drv->driver);
++ mutex_lock(&xenwatch_mutex);
++ ret = driver_register(&drv->driver);
++ mutex_unlock(&xenwatch_mutex);
++ return ret;
+ }
+
+ int __xenbus_register_frontend(struct xenbus_driver *drv,
+@@ -436,21 +500,30 @@ static void xenbus_dev_release(struct device *dev)
+ }
+
+ static ssize_t xendev_show_nodename(struct device *dev,
+- struct device_attribute *attr, char *buf)
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
++ struct device_attribute *attr,
++#endif
++ char *buf)
+ {
+ return sprintf(buf, "%s\n", to_xenbus_device(dev)->nodename);
+ }
+ DEVICE_ATTR(nodename, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_nodename, NULL);
+
+ static ssize_t xendev_show_devtype(struct device *dev,
+- struct device_attribute *attr, char *buf)
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
++ struct device_attribute *attr,
++#endif
++ char *buf)
+ {
+ return sprintf(buf, "%s\n", to_xenbus_device(dev)->devicetype);
+ }
+ DEVICE_ATTR(devtype, S_IRUSR | S_IRGRP | S_IROTH, xendev_show_devtype, NULL);
+
+ static ssize_t xendev_show_modalias(struct device *dev,
+- struct device_attribute *attr, char *buf)
++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,13)
++ struct device_attribute *attr,
++#endif
++ char *buf)
+ {
+ return sprintf(buf, "xen:%s\n", to_xenbus_device(dev)->devicetype);
+ }
+@@ -467,6 +540,9 @@ int xenbus_probe_node(struct xen_bus_type *bus,
+
+ enum xenbus_state state = xenbus_read_driver_state(nodename);
+
++ if (bus->error)
++ return bus->error;
++
+ if (state != XenbusStateInitialising) {
+ /* Device is not new, so ignore it. This can happen if a
+ device is going away after switching to Closed. */
+@@ -491,6 +567,9 @@ int xenbus_probe_node(struct xen_bus_type *bus,
+ xendev->devicetype = tmpstring;
+ init_completion(&xendev->down);
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++ xendev->dev.parent = &bus->dev;
++#endif
+ xendev->dev.bus = &bus->bus;
+ xendev->dev.release = xenbus_dev_release;
+
+@@ -571,6 +650,9 @@ int xenbus_probe_devices(struct xen_bus_type *bus)
+ char **dir;
+ unsigned int i, dir_n;
+
++ if (bus->error)
++ return bus->error;
++
+ dir = xenbus_directory(XBT_NIL, bus->root, "", &dir_n);
+ if (IS_ERR(dir))
+ return PTR_ERR(dir);
+@@ -614,7 +696,7 @@ void xenbus_dev_changed(const char *node, struct xen_bus_type *bus)
+ char type[BUS_ID_SIZE];
+ const char *p, *root;
+
+- if (char_count(node, '/') < 2)
++ if (bus->error || char_count(node, '/') < 2)
+ return;
+
+ exists = xenbus_exists(XBT_NIL, node, "");
+@@ -748,7 +830,8 @@ void xenbus_suspend(void)
+ {
+ DPRINTK("");
+
+- bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev);
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev);
+ xenbus_backend_suspend(suspend_dev);
+ xs_suspend();
+ }
+@@ -758,7 +841,8 @@ void xenbus_resume(void)
+ {
+ xb_init_comms();
+ xs_resume();
+- bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev);
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev);
+ xenbus_backend_resume(resume_dev);
+ }
+ EXPORT_SYMBOL_GPL(xenbus_resume);
+@@ -766,7 +850,8 @@ EXPORT_SYMBOL_GPL(xenbus_resume);
+ void xenbus_suspend_cancel(void)
+ {
+ xs_suspend_cancel();
+- bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev);
++ if (!xenbus_frontend.error)
++ bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev);
+ xenbus_backend_resume(suspend_cancel_dev);
+ }
+ EXPORT_SYMBOL_GPL(xenbus_suspend_cancel);
+@@ -794,6 +879,7 @@ void unregister_xenstore_notifier(struct notifier_block *nb)
+ }
+ EXPORT_SYMBOL_GPL(unregister_xenstore_notifier);
+
++
+ void xenbus_probe(struct work_struct *unused)
+ {
+ BUG_ON((xenstored_ready <= 0));
+@@ -807,63 +893,191 @@ void xenbus_probe(struct work_struct *unused)
+ blocking_notifier_call_chain(&xenstore_chain, 0, NULL);
+ }
+
++
++#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
++static struct file_operations xsd_kva_fops;
++static struct proc_dir_entry *xsd_kva_intf;
++static struct proc_dir_entry *xsd_port_intf;
++
++static int xsd_kva_mmap(struct file *file, struct vm_area_struct *vma)
++{
++ size_t size = vma->vm_end - vma->vm_start;
++
++ if ((size > PAGE_SIZE) || (vma->vm_pgoff != 0))
++ return -EINVAL;
++
++ if (remap_pfn_range(vma, vma->vm_start, mfn_to_pfn(xen_store_mfn),
++ size, vma->vm_page_prot))
++ return -EAGAIN;
++
++ return 0;
++}
++
++static int xsd_kva_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len;
++
++ len = sprintf(page, "0x%p", xen_store_interface);
++ *eof = 1;
++ return len;
++}
++
++static int xsd_port_read(char *page, char **start, off_t off,
++ int count, int *eof, void *data)
++{
++ int len;
++
++ len = sprintf(page, "%d", xen_store_evtchn);
++ *eof = 1;
++ return len;
++}
++#endif
++
++#ifndef MODULE
+ static int __init xenbus_probe_init(void)
++#else
++static int __devinit xenbus_probe_init(void)
++#endif
+ {
+ int err = 0;
++#if defined(CONFIG_XEN) || defined(MODULE)
++ unsigned long page = 0;
++#endif
+
+ DPRINTK("");
+
+- err = -ENODEV;
+ if (!is_running_on_xen())
+- goto out_error;
++ return -ENODEV;
+
+ /* Register ourselves with the kernel bus subsystem */
+- err = bus_register(&xenbus_frontend.bus);
+- if (err)
+- goto out_error;
+-
+- err = xenbus_backend_bus_register();
+- if (err)
+- goto out_unreg_front;
++ xenbus_frontend.error = bus_register(&xenbus_frontend.bus);
++ if (xenbus_frontend.error)
++ printk(KERN_WARNING
++ "XENBUS: Error registering frontend bus: %i\n",
++ xenbus_frontend.error);
++ xenbus_backend_bus_register();
+
+ /*
+ * Domain0 doesn't have a store_evtchn or store_mfn yet.
+ */
+ if (is_initial_xendomain()) {
++#if defined(CONFIG_XEN) || defined(MODULE)
++ struct evtchn_alloc_unbound alloc_unbound;
++
++ /* Allocate page. */
++ page = get_zeroed_page(GFP_KERNEL);
++ if (!page)
++ return -ENOMEM;
++
++ xen_store_mfn = xen_start_info->store_mfn =
++ pfn_to_mfn(virt_to_phys((void *)page) >>
++ PAGE_SHIFT);
++
++ /* Next allocate a local port which xenstored can bind to */
++ alloc_unbound.dom = DOMID_SELF;
++ alloc_unbound.remote_dom = 0;
++
++ err = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
++ &alloc_unbound);
++ if (err == -ENOSYS)
++ goto err;
++ BUG_ON(err);
++ xen_store_evtchn = xen_start_info->store_evtchn =
++ alloc_unbound.port;
++
++#if defined(CONFIG_PROC_FS) && defined(CONFIG_XEN_PRIVILEGED_GUEST)
++ /* And finally publish the above info in /proc/xen */
++ xsd_kva_intf = create_xen_proc_entry("xsd_kva", 0600);
++ if (xsd_kva_intf) {
++ memcpy(&xsd_kva_fops, xsd_kva_intf->proc_fops,
++ sizeof(xsd_kva_fops));
++ xsd_kva_fops.mmap = xsd_kva_mmap;
++ xsd_kva_intf->proc_fops = &xsd_kva_fops;
++ xsd_kva_intf->read_proc = xsd_kva_read;
++ }
++ xsd_port_intf = create_xen_proc_entry("xsd_port", 0400);
++ if (xsd_port_intf)
++ xsd_port_intf->read_proc = xsd_port_read;
++#endif
++#else
+ /* dom0 not yet supported */
++#endif
++ xen_store_interface = mfn_to_virt(xen_store_mfn);
+ } else {
+ xenstored_ready = 1;
++#ifndef MODULE
+ xen_store_evtchn = xen_start_info->store_evtchn;
+ xen_store_mfn = xen_start_info->store_mfn;
++ xen_store_interface = mfn_to_virt(xen_store_mfn);
++#else
++ xen_store_evtchn = hvm_get_parameter(HVM_PARAM_STORE_EVTCHN);
++ xen_store_mfn = hvm_get_parameter(HVM_PARAM_STORE_PFN);
++ xen_store_interface = ioremap(xen_store_mfn << PAGE_SHIFT,
++ PAGE_SIZE);
++#endif
+ }
+- xen_store_interface = mfn_to_virt(xen_store_mfn);
++
++
++#if defined(CONFIG_XEN) || defined(MODULE)
++ xenbus_dev_init();
++#endif
+
+ /* Initialize the interface to xenstore. */
+ err = xs_init();
+ if (err) {
+ printk(KERN_WARNING
+ "XENBUS: Error initializing xenstore comms: %i\n", err);
+- goto out_unreg_back;
++ goto err;
+ }
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++ /* Register ourselves with the kernel device subsystem */
++ if (!xenbus_frontend.error) {
++ xenbus_frontend.error = device_register(&xenbus_frontend.dev);
++ if (xenbus_frontend.error) {
++ bus_unregister(&xenbus_frontend.bus);
++ printk(KERN_WARNING
++ "XENBUS: Error registering frontend device: %i\n",
++ xenbus_frontend.error);
++ }
++ }
++#endif
++ xenbus_backend_device_register();
++
+ if (!is_initial_xendomain())
+ xenbus_probe(NULL);
+
+ return 0;
+
+- out_unreg_back:
+- xenbus_backend_bus_unregister();
++ err:
++#if defined(CONFIG_XEN) || defined(MODULE)
++ if (page)
++ free_page(page);
++#endif
+
+- out_unreg_front:
+- bus_unregister(&xenbus_frontend.bus);
++ /*
++ * Do not unregister the xenbus front/backend buses here. The buses
++ * must exist because front/backend drivers will use them when they are
++ * registered.
++ */
+
+- out_error:
+ return err;
+ }
+
++#ifndef MODULE
+ postcore_initcall(xenbus_probe_init);
+-
++#ifdef CONFIG_XEN
++MODULE_LICENSE("Dual BSD/GPL");
++#else
+ MODULE_LICENSE("GPL");
++#endif
++#else
++int __devinit xenbus_init(void)
++{
++ return xenbus_probe_init();
++}
++#endif
+
+ static int is_disconnected_device(struct device *dev, void *data)
+ {
+@@ -883,12 +1097,14 @@ static int is_disconnected_device(struct device *dev, void *data)
+ return 0;
+
+ xendrv = to_xenbus_driver(dev->driver);
+- return (xendev->state != XenbusStateConnected ||
++ return (xendev->state < XenbusStateConnected ||
+ (xendrv->is_ready && !xendrv->is_ready(xendev)));
+ }
+
+ static int exists_disconnected_device(struct device_driver *drv)
+ {
++ if (xenbus_frontend.error)
++ return xenbus_frontend.error;
+ return bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
+ is_disconnected_device);
+ }
+@@ -897,6 +1113,7 @@ static int print_device_status(struct device *dev, void *data)
+ {
+ struct xenbus_device *xendev = to_xenbus_device(dev);
+ struct device_driver *drv = data;
++ struct xenbus_driver *xendrv;
+
+ /* Is this operation limited to a particular driver? */
+ if (drv && (dev->driver != drv))
+@@ -906,12 +1123,23 @@ static int print_device_status(struct device *dev, void *data)
+ /* Information only: is this too noisy? */
+ printk(KERN_INFO "XENBUS: Device with no driver: %s\n",
+ xendev->nodename);
+- } else if (xendev->state != XenbusStateConnected) {
++ return 0;
++ }
++
++ if (xendev->state < XenbusStateConnected) {
++ enum xenbus_state rstate = XenbusStateUnknown;
++ if (xendev->otherend)
++ rstate = xenbus_read_driver_state(xendev->otherend);
+ printk(KERN_WARNING "XENBUS: Timeout connecting "
+- "to device: %s (state %d)\n",
+- xendev->nodename, xendev->state);
++ "to device: %s (local state %d, remote state %d)\n",
++ xendev->nodename, xendev->state, rstate);
+ }
+
++ xendrv = to_xenbus_driver(dev->driver);
++ if (xendrv->is_ready && !xendrv->is_ready(xendev))
++ printk(KERN_WARNING "XENBUS: Device not ready: %s\n",
++ xendev->nodename);
++
+ return 0;
+ }
+
+@@ -919,7 +1147,7 @@ static int print_device_status(struct device *dev, void *data)
+ static int ready_to_wait_for_devices;
+
+ /*
+- * On a 10 second timeout, wait for all devices currently configured. We need
++ * On a 5-minute timeout, wait for all devices currently configured. We need
+ * to do this to guarantee that the filesystems and / or network devices
+ * needed for boot are available, before we can allow the boot to proceed.
+ *
+@@ -934,18 +1162,30 @@ static int ready_to_wait_for_devices;
+ */
+ static void wait_for_devices(struct xenbus_driver *xendrv)
+ {
+- unsigned long timeout = jiffies + 10*HZ;
++ unsigned long start = jiffies;
+ struct device_driver *drv = xendrv ? &xendrv->driver : NULL;
++ unsigned int seconds_waited = 0;
+
+ if (!ready_to_wait_for_devices || !is_running_on_xen())
+ return;
+
+ while (exists_disconnected_device(drv)) {
+- if (time_after(jiffies, timeout))
+- break;
++ if (time_after(jiffies, start + (seconds_waited+5)*HZ)) {
++ if (!seconds_waited)
++ printk(KERN_WARNING "XENBUS: Waiting for "
++ "devices to initialise: ");
++ seconds_waited += 5;
++ printk("%us...", 300 - seconds_waited);
++ if (seconds_waited == 300)
++ break;
++ }
++
+ schedule_timeout_interruptible(HZ/10);
+ }
+
++ if (seconds_waited)
++ printk("\n");
++
+ bus_for_each_dev(&xenbus_frontend.bus, NULL, drv,
+ print_device_status);
+ }
+@@ -953,10 +1193,18 @@ static void wait_for_devices(struct xenbus_driver *xendrv)
+ #ifndef MODULE
+ static int __init boot_wait_for_devices(void)
+ {
+- ready_to_wait_for_devices = 1;
+- wait_for_devices(NULL);
++ if (!xenbus_frontend.error) {
++ ready_to_wait_for_devices = 1;
++ wait_for_devices(NULL);
++ }
+ return 0;
+ }
+
+ late_initcall(boot_wait_for_devices);
+ #endif
++
++int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *))
++{
++ return bus_for_each_dev(&xenbus_frontend.bus, NULL, arg, fn);
++}
++EXPORT_SYMBOL_GPL(xenbus_for_each_frontend);
+diff --git a/drivers/xen/xenbus/xenbus_probe.h b/drivers/xen/xenbus/xenbus_probe.h
+index e09b194..1d1ec74 100644
+--- a/drivers/xen/xenbus/xenbus_probe.h
++++ b/drivers/xen/xenbus/xenbus_probe.h
+@@ -34,27 +34,31 @@
+ #ifndef _XENBUS_PROBE_H
+ #define _XENBUS_PROBE_H
+
+-#ifdef CONFIG_XEN_BACKEND
++#if defined(CONFIG_XEN_BACKEND) || defined(CONFIG_XEN_BACKEND_MODULE)
+ extern void xenbus_backend_suspend(int (*fn)(struct device *, void *));
+ extern void xenbus_backend_resume(int (*fn)(struct device *, void *));
+ extern void xenbus_backend_probe_and_watch(void);
+-extern int xenbus_backend_bus_register(void);
+-extern void xenbus_backend_bus_unregister(void);
++extern void xenbus_backend_bus_register(void);
++extern void xenbus_backend_device_register(void);
+ #else
+ static inline void xenbus_backend_suspend(int (*fn)(struct device *, void *)) {}
+ static inline void xenbus_backend_resume(int (*fn)(struct device *, void *)) {}
+ static inline void xenbus_backend_probe_and_watch(void) {}
+-static inline int xenbus_backend_bus_register(void) { return 0; }
+-static inline void xenbus_backend_bus_unregister(void) {}
++static inline void xenbus_backend_bus_register(void) {}
++static inline void xenbus_backend_device_register(void) {}
+ #endif
+
+ struct xen_bus_type
+ {
+ char *root;
++ int error;
+ unsigned int levels;
+ int (*get_bus_id)(char bus_id[BUS_ID_SIZE], const char *nodename);
+ int (*probe)(const char *type, const char *dir);
+ struct bus_type bus;
++#if defined(CONFIG_XEN) || defined(MODULE)
++ struct device dev;
++#endif
+ };
+
+ extern int xenbus_match(struct device *_dev, struct device_driver *_drv);
+@@ -72,3 +76,4 @@ extern int xenbus_probe_devices(struct xen_bus_type *bus);
+ extern void xenbus_dev_changed(const char *node, struct xen_bus_type *bus);
+
+ #endif
++
+diff --git a/drivers/xen/xenbus/xenbus_probe_backend.c b/drivers/xen/xenbus/xenbus_probe_backend.c
+new file mode 100644
+index 0000000..9ab55af
+--- /dev/null
++++ b/drivers/xen/xenbus/xenbus_probe_backend.c
+@@ -0,0 +1,279 @@
++/******************************************************************************
++ * Talks to Xen Store to figure out what devices we have (backend half).
++ *
++ * Copyright (C) 2005 Rusty Russell, IBM Corporation
++ * Copyright (C) 2005 Mike Wray, Hewlett-Packard
++ * Copyright (C) 2005, 2006 XenSource Ltd
++ * Copyright (C) 2007 Solarflare Communications, Inc.
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#define DPRINTK(fmt, args...) \
++ pr_debug("xenbus_probe (%s:%d) " fmt ".\n", \
++ __FUNCTION__, __LINE__, ##args)
++
++#include <linux/kernel.h>
++#include <linux/err.h>
++#include <linux/string.h>
++#include <linux/ctype.h>
++#include <linux/fcntl.h>
++#include <linux/mm.h>
++#include <linux/notifier.h>
++
++#include <asm/io.h>
++#include <asm/page.h>
++#include <asm/maddr.h>
++#include <asm/pgtable.h>
++#include <asm/hypervisor.h>
++#include <xen/xenbus.h>
++#include <xen/xen_proc.h>
++#include <xen/evtchn.h>
++#include <xen/features.h>
++
++#include "xenbus_comms.h"
++#include "xenbus_probe.h"
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++static int xenbus_uevent_backend(struct device *dev, struct kobj_uevent_env *env);
++static int xenbus_probe_backend(const char *type, const char *domid);
++
++extern int read_otherend_details(struct xenbus_device *xendev,
++ char *id_node, char *path_node);
++
++static int read_frontend_details(struct xenbus_device *xendev)
++{
++ return read_otherend_details(xendev, "frontend-id", "frontend");
++}
++
++/* backend/<type>/<fe-uuid>/<id> => <type>-<fe-domid>-<id> */
++static int backend_bus_id(char bus_id[BUS_ID_SIZE], const char *nodename)
++{
++ int domid, err;
++ const char *devid, *type, *frontend;
++ unsigned int typelen;
++
++ type = strchr(nodename, '/');
++ if (!type)
++ return -EINVAL;
++ type++;
++ typelen = strcspn(type, "/");
++ if (!typelen || type[typelen] != '/')
++ return -EINVAL;
++
++ devid = strrchr(nodename, '/') + 1;
++
++ err = xenbus_gather(XBT_NIL, nodename, "frontend-id", "%i", &domid,
++ "frontend", NULL, &frontend,
++ NULL);
++ if (err)
++ return err;
++ if (strlen(frontend) == 0)
++ err = -ERANGE;
++ if (!err && !xenbus_exists(XBT_NIL, frontend, ""))
++ err = -ENOENT;
++ kfree(frontend);
++
++ if (err)
++ return err;
++
++ if (snprintf(bus_id, BUS_ID_SIZE,
++ "%.*s-%i-%s", typelen, type, domid, devid) >= BUS_ID_SIZE)
++ return -ENOSPC;
++ return 0;
++}
++
++static struct xen_bus_type xenbus_backend = {
++ .root = "backend",
++ .levels = 3, /* backend/type/<frontend>/<id> */
++ .get_bus_id = backend_bus_id,
++ .probe = xenbus_probe_backend,
++ .error = -ENODEV,
++ .bus = {
++ .name = "xen-backend",
++ .match = xenbus_match,
++ .probe = xenbus_dev_probe,
++ .remove = xenbus_dev_remove,
++// .shutdown = xenbus_dev_shutdown,
++ .uevent = xenbus_uevent_backend,
++ },
++ .dev = {
++ .bus_id = "xen-backend",
++ },
++};
++
++static int xenbus_uevent_backend(struct device *dev, struct kobj_uevent_env *env)
++{
++ struct xenbus_device *xdev;
++ struct xenbus_driver *drv;
++
++ DPRINTK("");
++
++ if (dev == NULL)
++ return -ENODEV;
++
++ xdev = to_xenbus_device(dev);
++ if (xdev == NULL)
++ return -ENODEV;
++
++ /* stuff we want to pass to /sbin/hotplug */
++ add_uevent_var(env, "XENBUS_TYPE=%s", xdev->devicetype);
++
++ add_uevent_var(env, "XENBUS_PATH=%s", xdev->nodename);
++
++ add_uevent_var(env, "XENBUS_BASE_PATH=%s", xenbus_backend.root);
++
++ if (dev->driver) {
++ drv = to_xenbus_driver(dev->driver);
++ if (drv && drv->uevent)
++ return drv->uevent(xdev, env);
++ }
++
++ return 0;
++}
++
++int __xenbus_register_backend(struct xenbus_driver *drv,
++ struct module *owner, const char *mod_name)
++{
++ drv->read_otherend_details = read_frontend_details;
++
++ return xenbus_register_driver_common(drv, &xenbus_backend,
++ owner, mod_name);
++}
++EXPORT_SYMBOL_GPL(__xenbus_register_backend);
++
++/* backend/<typename>/<frontend-uuid>/<name> */
++static int xenbus_probe_backend_unit(const char *dir,
++ const char *type,
++ const char *name)
++{
++ char *nodename;
++ int err;
++
++ nodename = kasprintf(GFP_KERNEL, "%s/%s", dir, name);
++ if (!nodename)
++ return -ENOMEM;
++
++ DPRINTK("%s\n", nodename);
++
++ err = xenbus_probe_node(&xenbus_backend, type, nodename);
++ kfree(nodename);
++ return err;
++}
++
++/* backend/<typename>/<frontend-domid> */
++static int xenbus_probe_backend(const char *type, const char *domid)
++{
++ char *nodename;
++ int err = 0;
++ char **dir;
++ unsigned int i, dir_n = 0;
++
++ DPRINTK("");
++
++ nodename = kasprintf(GFP_KERNEL, "%s/%s/%s", xenbus_backend.root, type, domid);
++ if (!nodename)
++ return -ENOMEM;
++
++ dir = xenbus_directory(XBT_NIL, nodename, "", &dir_n);
++ if (IS_ERR(dir)) {
++ kfree(nodename);
++ return PTR_ERR(dir);
++ }
++
++ for (i = 0; i < dir_n; i++) {
++ err = xenbus_probe_backend_unit(nodename, type, dir[i]);
++ if (err)
++ break;
++ }
++ kfree(dir);
++ kfree(nodename);
++ return err;
++}
++
++static void backend_changed(struct xenbus_watch *watch,
++ const char **vec, unsigned int len)
++{
++ DPRINTK("");
++
++ xenbus_dev_changed(vec[XS_WATCH_PATH], &xenbus_backend);
++}
++
++static struct xenbus_watch be_watch = {
++ .node = "backend",
++ .callback = backend_changed,
++};
++
++void xenbus_backend_suspend(int (*fn)(struct device *, void *))
++{
++ DPRINTK("");
++ if (!xenbus_backend.error)
++ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
++}
++
++void xenbus_backend_resume(int (*fn)(struct device *, void *))
++{
++ DPRINTK("");
++ if (!xenbus_backend.error)
++ bus_for_each_dev(&xenbus_backend.bus, NULL, NULL, fn);
++}
++
++void xenbus_backend_probe_and_watch(void)
++{
++ xenbus_probe_devices(&xenbus_backend);
++ register_xenbus_watch(&be_watch);
++}
++
++void xenbus_backend_bus_register(void)
++{
++ xenbus_backend.error = bus_register(&xenbus_backend.bus);
++ if (xenbus_backend.error)
++ printk(KERN_WARNING
++ "XENBUS: Error registering backend bus: %i\n",
++ xenbus_backend.error);
++}
++
++void xenbus_backend_device_register(void)
++{
++ if (xenbus_backend.error)
++ return;
++
++ xenbus_backend.error = device_register(&xenbus_backend.dev);
++ if (xenbus_backend.error) {
++ bus_unregister(&xenbus_backend.bus);
++ printk(KERN_WARNING
++ "XENBUS: Error registering backend device: %i\n",
++ xenbus_backend.error);
++ }
++}
++
++int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *))
++{
++ return bus_for_each_dev(&xenbus_backend.bus, NULL, arg, fn);
++}
++EXPORT_SYMBOL_GPL(xenbus_for_each_backend);
+diff --git a/drivers/xen/xenbus/xenbus_xs.c b/drivers/xen/xenbus/xenbus_xs.c
+index 227d53b..31dae64 100644
+--- a/drivers/xen/xenbus/xenbus_xs.c
++++ b/drivers/xen/xenbus/xenbus_xs.c
+@@ -47,6 +47,14 @@
+ #include <xen/xenbus.h>
+ #include "xenbus_comms.h"
+
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#ifndef PF_NOFREEZE /* Old kernel (pre-2.6.6). */
++#define PF_NOFREEZE 0
++#endif
++
+ struct xs_stored_msg {
+ struct list_head list;
+
+@@ -108,7 +116,7 @@ static DEFINE_SPINLOCK(watch_events_lock);
+ * carrying out work.
+ */
+ static pid_t xenwatch_pid;
+-static DEFINE_MUTEX(xenwatch_mutex);
++/* static */ DEFINE_MUTEX(xenwatch_mutex);
+ static DECLARE_WAIT_QUEUE_HEAD(watch_events_waitq);
+
+ static int get_error(const char *errorstring)
+@@ -177,7 +185,7 @@ void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg)
+
+ mutex_unlock(&xs_state.request_mutex);
+
+- if ((msg->type == XS_TRANSACTION_END) ||
++ if ((req_msg.type == XS_TRANSACTION_END) ||
+ ((req_msg.type == XS_TRANSACTION_START) &&
+ (msg->type == XS_ERROR)))
+ up_read(&xs_state.transaction_mutex);
+@@ -283,9 +291,9 @@ static char *join(const char *dir, const char *name)
+ char *buffer;
+
+ if (strlen(name) == 0)
+- buffer = kasprintf(GFP_KERNEL, "%s", dir);
++ buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s", dir);
+ else
+- buffer = kasprintf(GFP_KERNEL, "%s/%s", dir, name);
++ buffer = kasprintf(GFP_NOIO | __GFP_HIGH, "%s/%s", dir, name);
+ return (!buffer) ? ERR_PTR(-ENOMEM) : buffer;
+ }
+
+@@ -294,10 +302,10 @@ static char **split(char *strings, unsigned int len, unsigned int *num)
+ char *p, **ret;
+
+ /* Count the strings. */
+- *num = count_strings(strings, len);
++ *num = count_strings(strings, len) + 1;
+
+ /* Transfer to one big alloc for easy freeing. */
+- ret = kmalloc(*num * sizeof(char *) + len, GFP_KERNEL);
++ ret = kmalloc(*num * sizeof(char *) + len, GFP_NOIO | __GFP_HIGH);
+ if (!ret) {
+ kfree(strings);
+ return ERR_PTR(-ENOMEM);
+@@ -308,6 +316,7 @@ static char **split(char *strings, unsigned int len, unsigned int *num)
+ strings = (char *)&ret[*num];
+ for (p = strings, *num = 0; p < strings + len; p += strlen(p) + 1)
+ ret[(*num)++] = p;
++ ret[*num] = strings + len;
+
+ return ret;
+ }
+@@ -498,7 +507,7 @@ int xenbus_printf(struct xenbus_transaction t,
+ #define PRINTF_BUFFER_SIZE 4096
+ char *printf_buffer;
+
+- printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_KERNEL);
++ printf_buffer = kmalloc(PRINTF_BUFFER_SIZE, GFP_NOIO | __GFP_HIGH);
+ if (printf_buffer == NULL)
+ return -ENOMEM;
+
+@@ -621,6 +630,10 @@ void unregister_xenbus_watch(struct xenbus_watch *watch)
+ char token[sizeof(watch) * 2 + 1];
+ int err;
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++ BUG_ON(watch->flags & XBWF_new_thread);
++#endif
++
+ sprintf(token, "%lX", (long)watch);
+
+ down_read(&xs_state.watch_mutex);
+@@ -693,11 +706,32 @@ void xs_suspend_cancel(void)
+ up_write(&xs_state.transaction_mutex);
+ }
+
++#if defined(CONFIG_XEN) || defined(MODULE)
++static int xenwatch_handle_callback(void *data)
++{
++ struct xs_stored_msg *msg = data;
++
++ msg->u.watch.handle->callback(msg->u.watch.handle,
++ (const char **)msg->u.watch.vec,
++ msg->u.watch.vec_size);
++
++ kfree(msg->u.watch.vec);
++ kfree(msg);
++
++ /* Kill this kthread if we were spawned just for this callback. */
++ if (current->pid != xenwatch_pid)
++ do_exit(0);
++
++ return 0;
++}
++#endif
++
+ static int xenwatch_thread(void *unused)
+ {
+ struct list_head *ent;
+ struct xs_stored_msg *msg;
+
++ current->flags |= PF_NOFREEZE;
+ for (;;) {
+ wait_event_interruptible(watch_events_waitq,
+ !list_empty(&watch_events));
+@@ -713,17 +747,39 @@ static int xenwatch_thread(void *unused)
+ list_del(ent);
+ spin_unlock(&watch_events_lock);
+
+- if (ent != &watch_events) {
+- msg = list_entry(ent, struct xs_stored_msg, list);
+- msg->u.watch.handle->callback(
+- msg->u.watch.handle,
+- (const char **)msg->u.watch.vec,
+- msg->u.watch.vec_size);
+- kfree(msg->u.watch.vec);
+- kfree(msg);
++ if (ent == &watch_events) {
++ mutex_unlock(&xenwatch_mutex);
++ continue;
+ }
+
++ msg = list_entry(ent, struct xs_stored_msg, list);
++
++#if defined(CONFIG_XEN) || defined(MODULE)
++ /*
++ * Unlock the mutex before running an XBWF_new_thread
++ * handler. kthread_run can block which can deadlock
++ * against unregister_xenbus_watch() if we need to
++ * unregister other watches in order to make
++ * progress. This can occur on resume before the swap
++ * device is attached.
++ */
++ if (msg->u.watch.handle->flags & XBWF_new_thread) {
++ mutex_unlock(&xenwatch_mutex);
++ kthread_run(xenwatch_handle_callback,
++ msg, "xenwatch_cb");
++ } else {
++ xenwatch_handle_callback(msg);
++ mutex_unlock(&xenwatch_mutex);
++ }
++#else
++ msg->u.watch.handle->callback(
++ msg->u.watch.handle,
++ (const char **)msg->u.watch.vec,
++ msg->u.watch.vec_size);
+ mutex_unlock(&xenwatch_mutex);
++ kfree(msg->u.watch.vec);
++ kfree(msg);
++#endif
+ }
+
+ return 0;
+@@ -751,7 +807,7 @@ static int process_msg(void)
+ }
+
+
+- msg = kmalloc(sizeof(*msg), GFP_KERNEL);
++ msg = kmalloc(sizeof(*msg), GFP_NOIO | __GFP_HIGH);
+ if (msg == NULL) {
+ err = -ENOMEM;
+ goto out;
+@@ -763,7 +819,7 @@ static int process_msg(void)
+ goto out;
+ }
+
+- body = kmalloc(msg->hdr.len + 1, GFP_KERNEL);
++ body = kmalloc(msg->hdr.len + 1, GFP_NOIO | __GFP_HIGH);
+ if (body == NULL) {
+ kfree(msg);
+ err = -ENOMEM;
+@@ -817,6 +873,7 @@ static int xenbus_thread(void *unused)
+ {
+ int err;
+
++ current->flags |= PF_NOFREEZE;
+ for (;;) {
+ err = process_msg();
+ if (err)
+diff --git a/drivers/xen/xenoprof/xenoprofile.c b/drivers/xen/xenoprof/xenoprofile.c
+new file mode 100644
+index 0000000..913d528
+--- /dev/null
++++ b/drivers/xen/xenoprof/xenoprofile.c
+@@ -0,0 +1,546 @@
++/**
++ * @file xenoprofile.c
++ *
++ * @remark Copyright 2002 OProfile authors
++ * @remark Read the file COPYING
++ *
++ * @author John Levon <levon at movementarian.org>
++ *
++ * Modified by Aravind Menon and Jose Renato Santos for Xen
++ * These modifications are:
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ *
++ * Separated out arch-generic part
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ */
++
++#include <linux/init.h>
++#include <linux/notifier.h>
++#include <linux/smp.h>
++#include <linux/oprofile.h>
++#include <linux/sysdev.h>
++#include <linux/slab.h>
++#include <linux/interrupt.h>
++#include <linux/vmalloc.h>
++#include <asm/pgtable.h>
++#include <xen/evtchn.h>
++#include <xen/xenoprof.h>
++#include <xen/driver_util.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/xenoprof.h>
++#include "../../../drivers/oprofile/cpu_buffer.h"
++#include "../../../drivers/oprofile/event_buffer.h"
++
++#define MAX_XENOPROF_SAMPLES 16
++
++/* sample buffers shared with Xen */
++xenoprof_buf_t * xenoprof_buf[MAX_VIRT_CPUS];
++/* Shared buffer area */
++struct xenoprof_shared_buffer shared_buffer;
++
++/* Passive sample buffers shared with Xen */
++xenoprof_buf_t *p_xenoprof_buf[MAX_OPROF_DOMAINS][MAX_VIRT_CPUS];
++/* Passive shared buffer area */
++struct xenoprof_shared_buffer p_shared_buffer[MAX_OPROF_DOMAINS];
++
++static int xenoprof_start(void);
++static void xenoprof_stop(void);
++
++static int xenoprof_enabled = 0;
++static int xenoprof_is_primary = 0;
++static int active_defined;
++
++extern unsigned long backtrace_depth;
++
++/* 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 on XENOPROF_init command */
++char cpu_type[XENOPROF_CPU_TYPE_SIZE];
++
++#ifdef CONFIG_PM
++
++static int xenoprof_suspend(struct sys_device * dev, pm_message_t state)
++{
++ if (xenoprof_enabled == 1)
++ xenoprof_stop();
++ return 0;
++}
++
++
++static int xenoprof_resume(struct sys_device * dev)
++{
++ if (xenoprof_enabled == 1)
++ xenoprof_start();
++ return 0;
++}
++
++
++static struct sysdev_class oprofile_sysclass = {
++ .name = "oprofile",
++ .resume = xenoprof_resume,
++ .suspend = xenoprof_suspend
++};
++
++
++static struct sys_device device_oprofile = {
++ .id = 0,
++ .cls = &oprofile_sysclass,
++};
++
++
++static int __init init_driverfs(void)
++{
++ int error;
++ if (!(error = sysdev_class_register(&oprofile_sysclass)))
++ error = sysdev_register(&device_oprofile);
++ return error;
++}
++
++
++static void exit_driverfs(void)
++{
++ sysdev_unregister(&device_oprofile);
++ sysdev_class_unregister(&oprofile_sysclass);
++}
++
++#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];
++
++/* Check whether the given entry is an escape code */
++static int xenoprof_is_escape(xenoprof_buf_t * buf, int tail)
++{
++ return (buf->event_log[tail].eip == XENOPROF_ESCAPE_CODE);
++}
++
++/* Get the event at the given entry */
++static uint8_t xenoprof_get_event(xenoprof_buf_t * buf, int tail)
++{
++ return (buf->event_log[tail].event);
++}
++
++static void xenoprof_add_pc(xenoprof_buf_t *buf, int is_passive)
++{
++ int head, tail, size;
++ int tracing = 0;
++
++ head = buf->event_head;
++ tail = buf->event_tail;
++ size = buf->event_size;
++
++ while (tail != head) {
++ if (xenoprof_is_escape(buf, tail) &&
++ xenoprof_get_event(buf, tail) == XENOPROF_TRACE_BEGIN) {
++ tracing=1;
++ oprofile_add_pc(ESCAPE_CODE, buf->event_log[tail].mode,
++ CPU_TRACE_BEGIN);
++ if (!is_passive)
++ oprofile_samples++;
++ else
++ p_oprofile_samples++;
++
++ } else {
++ oprofile_add_pc(buf->event_log[tail].eip,
++ buf->event_log[tail].mode,
++ buf->event_log[tail].event);
++ if (!tracing) {
++ if (!is_passive)
++ oprofile_samples++;
++ else
++ p_oprofile_samples++;
++ }
++
++ }
++ tail++;
++ if(tail==size)
++ tail=0;
++ }
++ buf->event_tail = tail;
++}
++
++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);
++}
++
++static irqreturn_t
++xenoprof_ovf_interrupt(int irq, void * dev_id)
++{
++ struct xenoprof_buf * buf;
++ static unsigned long flag;
++
++ buf = xenoprof_buf[smp_processor_id()];
++
++ xenoprof_add_pc(buf, 0);
++
++ if (xenoprof_is_primary && !test_and_set_bit(0, &flag)) {
++ xenoprof_handle_passive();
++ smp_mb__before_clear_bit();
++ clear_bit(0, &flag);
++ }
++
++ return IRQ_HANDLED;
++}
++
++
++static void unbind_virq(void)
++{
++ unsigned int i;
++
++ for_each_online_cpu(i) {
++ if (ovf_irq[i] >= 0) {
++ unbind_from_irqhandler(ovf_irq[i], NULL);
++ ovf_irq[i] = -1;
++ }
++ }
++}
++
++
++static int bind_virq(void)
++{
++ unsigned int i;
++ int result;
++
++ for_each_online_cpu(i) {
++ result = bind_virq_to_irqhandler(VIRQ_XENOPROF,
++ i,
++ xenoprof_ovf_interrupt,
++ IRQF_DISABLED|IRQF_NOBALANCING,
++ "xenoprof",
++ NULL);
++
++ if (result < 0) {
++ unbind_virq();
++ return result;
++ }
++
++ ovf_irq[i] = result;
++ }
++
++ return 0;
++}
++
++
++static void unmap_passive_list(void)
++{
++ int i;
++ for (i = 0; i < pdomains; i++)
++ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
++ pdomains = 0;
++}
++
++
++static int map_xenoprof_buffer(int max_samples)
++{
++ struct xenoprof_get_buffer get_buffer;
++ struct xenoprof_buf *buf;
++ int ret, i;
++
++ if ( shared_buffer.buffer )
++ return 0;
++
++ get_buffer.max_samples = max_samples;
++ ret = xenoprof_arch_map_shared_buffer(&get_buffer, &shared_buffer);
++ if (ret)
++ return ret;
++ nbuf = get_buffer.nbuf;
++
++ for (i=0; i< nbuf; i++) {
++ buf = (struct xenoprof_buf*)
++ &shared_buffer.buffer[i * get_buffer.bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ xenoprof_buf[buf->vcpu_id] = buf;
++ }
++
++ return 0;
++}
++
++
++static int xenoprof_setup(void)
++{
++ int ret;
++
++ if ( (ret = map_xenoprof_buffer(MAX_XENOPROF_SAMPLES)) )
++ return ret;
++
++ if ( (ret = bind_virq()) )
++ return ret;
++
++ if (xenoprof_is_primary) {
++ /* Define dom0 as an active domain if not done yet */
++ if (!active_defined) {
++ domid_t domid;
++ ret = HYPERVISOR_xenoprof_op(
++ XENOPROF_reset_active_list, NULL);
++ if (ret)
++ goto err;
++ domid = 0;
++ ret = HYPERVISOR_xenoprof_op(
++ XENOPROF_set_active, &domid);
++ if (ret)
++ goto err;
++ active_defined = 1;
++ }
++
++ if (backtrace_depth > 0) {
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_backtrace,
++ &backtrace_depth);
++ if (ret)
++ backtrace_depth = 0;
++ }
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reserve_counters, NULL);
++ if (ret)
++ goto err;
++
++ xenoprof_arch_counter();
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_setup_events, NULL);
++ if (ret)
++ goto err;
++ }
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_enable_virq, NULL);
++ if (ret)
++ goto err;
++
++ xenoprof_enabled = 1;
++ return 0;
++ err:
++ unbind_virq();
++ return ret;
++}
++
++
++static void xenoprof_shutdown(void)
++{
++ xenoprof_enabled = 0;
++
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_disable_virq, NULL));
++
++ if (xenoprof_is_primary) {
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_release_counters,
++ NULL));
++ active_defined = 0;
++ }
++
++ unbind_virq();
++
++ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
++ if (xenoprof_is_primary)
++ unmap_passive_list();
++}
++
++
++static int xenoprof_start(void)
++{
++ int ret = 0;
++
++ if (xenoprof_is_primary)
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_start, NULL);
++ if (!ret)
++ xenoprof_arch_start();
++ return ret;
++}
++
++
++static void xenoprof_stop(void)
++{
++ if (xenoprof_is_primary)
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_stop, NULL));
++ xenoprof_arch_stop();
++}
++
++
++static int xenoprof_set_active(int * active_domains,
++ unsigned int adomains)
++{
++ int ret = 0;
++ int i;
++ int set_dom0 = 0;
++ domid_t domid;
++
++ if (!xenoprof_is_primary)
++ return 0;
++
++ if (adomains > MAX_OPROF_DOMAINS)
++ return -E2BIG;
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list, NULL);
++ if (ret)
++ return ret;
++
++ for (i=0; i<adomains; i++) {
++ domid = active_domains[i];
++ if (domid != active_domains[i]) {
++ ret = -EINVAL;
++ goto out;
++ }
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
++ if (ret)
++ goto out;
++ if (active_domains[i] == 0)
++ set_dom0 = 1;
++ }
++ /* dom0 must always be active but may not be in the list */
++ if (!set_dom0) {
++ domid = 0;
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_set_active, &domid);
++ }
++
++out:
++ if (ret)
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_reset_active_list,
++ NULL));
++ active_defined = !ret;
++ return ret;
++}
++
++static int xenoprof_set_passive(int * p_domains,
++ unsigned int pdoms)
++{
++ int ret;
++ unsigned int i, j;
++ struct xenoprof_buf *buf;
++
++ if (!xenoprof_is_primary)
++ return 0;
++
++ if (pdoms > MAX_OPROF_DOMAINS)
++ return -E2BIG;
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_reset_passive_list, NULL);
++ if (ret)
++ return ret;
++ unmap_passive_list();
++
++ for (i = 0; i < pdoms; i++) {
++ passive_domains[i].domain_id = p_domains[i];
++ passive_domains[i].max_samples = 2048;
++ ret = xenoprof_arch_set_passive(&passive_domains[i],
++ &p_shared_buffer[i]);
++ if (ret)
++ goto out;
++ for (j = 0; j < passive_domains[i].nbuf; j++) {
++ buf = (struct xenoprof_buf *)
++ &p_shared_buffer[i].buffer[
++ j * passive_domains[i].bufsize];
++ BUG_ON(buf->vcpu_id >= MAX_VIRT_CPUS);
++ p_xenoprof_buf[i][buf->vcpu_id] = buf;
++ }
++ }
++
++ pdomains = pdoms;
++ return 0;
++
++out:
++ for (j = 0; j < i; j++)
++ xenoprof_arch_unmap_shared_buffer(&p_shared_buffer[i]);
++
++ return ret;
++}
++
++
++/* The dummy backtrace function to keep oprofile happy
++ * The real backtrace is done in xen
++ */
++static void xenoprof_dummy_backtrace(struct pt_regs * const regs,
++ unsigned int depth)
++{
++ /* this should never be called */
++ BUG();
++ return;
++}
++
++
++
++struct oprofile_operations xenoprof_ops = {
++#ifdef HAVE_XENOPROF_CREATE_FILES
++ .create_files = xenoprof_create_files,
++#endif
++ .set_active = xenoprof_set_active,
++ .set_passive = xenoprof_set_passive,
++ .setup = xenoprof_setup,
++ .shutdown = xenoprof_shutdown,
++ .start = xenoprof_start,
++ .stop = xenoprof_stop,
++ .backtrace = xenoprof_dummy_backtrace
++};
++
++
++/* in order to get driverfs right */
++static int using_xenoprof;
++
++int __init xenoprofile_init(struct oprofile_operations * ops)
++{
++ struct xenoprof_init init;
++ unsigned int i;
++ int ret;
++
++ ret = HYPERVISOR_xenoprof_op(XENOPROF_init, &init);
++ if (!ret) {
++ xenoprof_arch_init_counter(&init);
++ xenoprof_is_primary = init.is_primary;
++
++ /* 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;
++
++ init_driverfs();
++ using_xenoprof = 1;
++ *ops = xenoprof_ops;
++
++ for (i=0; i<NR_CPUS; i++)
++ ovf_irq[i] = -1;
++
++ active_defined = 0;
++ }
++
++ printk(KERN_INFO "%s: ret %d, events %d, xenoprof_is_primary %d\n",
++ __func__, ret, init.num_events, xenoprof_is_primary);
++ return ret;
++}
++
++
++void xenoprofile_exit(void)
++{
++ if (using_xenoprof)
++ exit_driverfs();
++
++ xenoprof_arch_unmap_shared_buffer(&shared_buffer);
++ if (xenoprof_is_primary) {
++ unmap_passive_list();
++ WARN_ON(HYPERVISOR_xenoprof_op(XENOPROF_shutdown, NULL));
++ }
++}
+diff --git a/fs/Kconfig b/fs/Kconfig
+index 2694648..6027e00 100644
+--- a/fs/Kconfig
++++ b/fs/Kconfig
+@@ -1007,6 +1007,7 @@ config HUGETLBFS
+ bool "HugeTLB file system support"
+ depends on X86 || IA64 || PPC64 || SPARC64 || (SUPERH && MMU) || \
+ (S390 && 64BIT) || 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 --git a/fs/aio.c b/fs/aio.c
+index 0fb3117..9ea3bf3 100644
+--- a/fs/aio.c
++++ b/fs/aio.c
+@@ -36,6 +36,11 @@
+ #include <asm/uaccess.h>
+ #include <asm/mmu_context.h>
+
++#ifdef CONFIG_EPOLL
++#include <linux/poll.h>
++#include <linux/anon_inodes.h>
++#endif
++
+ #if DEBUG > 1
+ #define dprintk printk
+ #else
+@@ -1012,6 +1017,11 @@ put_rq:
+ if (waitqueue_active(&ctx->wait))
+ wake_up(&ctx->wait);
+
++#ifdef CONFIG_EPOLL
++ if (ctx->file && waitqueue_active(&ctx->poll_wait))
++ wake_up(&ctx->poll_wait);
++#endif
++
+ spin_unlock_irqrestore(&ctx->ctx_lock, flags);
+ return ret;
+ }
+@@ -1019,6 +1029,8 @@ put_rq:
+ /* aio_read_evt
+ * Pull an event off of the ioctx's event ring. Returns the number of
+ * events fetched (0 or 1 ;-)
++ * If ent parameter is 0, just returns the number of events that would
++ * be fetched.
+ * FIXME: make this use cmpxchg.
+ * TODO: make the ringbuffer user mmap()able (requires FIXME).
+ */
+@@ -1041,13 +1053,18 @@ static int aio_read_evt(struct kioctx *ioctx, struct io_event *ent)
+
+ head = ring->head % info->nr;
+ if (head != ring->tail) {
+- struct io_event *evp = aio_ring_event(info, head, KM_USER1);
+- *ent = *evp;
+- head = (head + 1) % info->nr;
+- smp_mb(); /* finish reading the event before updatng the head */
+- ring->head = head;
+- ret = 1;
+- put_aio_ring_event(evp, KM_USER1);
++ if (ent) { /* event requested */
++ struct io_event *evp =
++ aio_ring_event(info, head, KM_USER1);
++ *ent = *evp;
++ head = (head + 1) % info->nr;
++ /* finish reading the event before updatng the head */
++ smp_mb();
++ ring->head = head;
++ ret = 1;
++ put_aio_ring_event(evp, KM_USER1);
++ } else /* only need to know availability */
++ ret = 1;
+ }
+ spin_unlock(&info->ring_lock);
+
+@@ -1237,6 +1254,14 @@ static void io_destroy(struct kioctx *ioctx)
+
+ aio_cancel_all(ioctx);
+ wait_for_all_aios(ioctx);
++#ifdef CONFIG_EPOLL
++ /* forget the poll file, but it's up to the user to close it */
++ if (ioctx->file) {
++ fput(ioctx->file);
++ ioctx->file->private_data = 0;
++ ioctx->file = 0;
++ }
++#endif
+
+ /*
+ * Wake up any waiters. The setting of ctx->dead must be seen
+@@ -1247,6 +1272,70 @@ static void io_destroy(struct kioctx *ioctx)
+ put_ioctx(ioctx); /* once for the lookup */
+ }
+
++#ifdef CONFIG_EPOLL
++
++static int aio_queue_fd_close(struct inode *inode, struct file *file)
++{
++ struct kioctx *ioctx = file->private_data;
++ if (ioctx) {
++ file->private_data = 0;
++ spin_lock_irq(&ioctx->ctx_lock);
++ ioctx->file = 0;
++ spin_unlock_irq(&ioctx->ctx_lock);
++ fput(file);
++ }
++ return 0;
++}
++
++static unsigned int aio_queue_fd_poll(struct file *file, poll_table *wait)
++{ unsigned int pollflags = 0;
++ struct kioctx *ioctx = file->private_data;
++
++ if (ioctx) {
++
++ spin_lock_irq(&ioctx->ctx_lock);
++ /* Insert inside our poll wait queue */
++ poll_wait(file, &ioctx->poll_wait, wait);
++
++ /* Check our condition */
++ if (aio_read_evt(ioctx, 0))
++ pollflags = POLLIN | POLLRDNORM;
++ spin_unlock_irq(&ioctx->ctx_lock);
++ }
++
++ return pollflags;
++}
++
++static const struct file_operations aioq_fops = {
++ .release = aio_queue_fd_close,
++ .poll = aio_queue_fd_poll
++};
++
++/* make_aio_fd:
++ * Create a file descriptor that can be used to poll the event queue.
++ * Based on the excellent epoll code.
++ */
++
++static int make_aio_fd(struct kioctx *ioctx)
++{
++ int fd;
++ struct file *file;
++
++ fd = anon_inode_getfd("[aioq]", &aioq_fops, ioctx);
++ if (fd < 0)
++ return fd;
++
++ /* associate the file with the IO context */
++ file = fget(fd);
++ if (!file)
++ return -EBADF;
++ file->private_data = ioctx;
++ ioctx->file = file;
++ init_waitqueue_head(&ioctx->poll_wait);
++ return fd;
++}
++#endif
++
+ /* sys_io_setup:
+ * Create an aio_context capable of receiving at least nr_events.
+ * ctxp must not point to an aio_context that already exists, and
+@@ -1259,18 +1348,30 @@ static void io_destroy(struct kioctx *ioctx)
+ * resources are available. May fail with -EFAULT if an invalid
+ * pointer is passed for ctxp. Will fail with -ENOSYS if not
+ * implemented.
++ *
++ * To request a selectable fd, the user context has to be initialized
++ * to 1, instead of 0, and the return value is the fd.
++ * This keeps the system call compatible, since a non-zero value
++ * was not allowed so far.
+ */
+ asmlinkage long sys_io_setup(unsigned nr_events, aio_context_t __user *ctxp)
+ {
+ struct kioctx *ioctx = NULL;
+ unsigned long ctx;
+ long ret;
++ int make_fd = 0;
+
+ ret = get_user(ctx, ctxp);
+ if (unlikely(ret))
+ goto out;
+
+ ret = -EINVAL;
++#ifdef CONFIG_EPOLL
++ if (ctx == 1) {
++ make_fd = 1;
++ ctx = 0;
++ }
++#endif
+ if (unlikely(ctx || nr_events == 0)) {
+ pr_debug("EINVAL: io_setup: ctx %lu nr_events %u\n",
+ ctx, nr_events);
+@@ -1281,8 +1382,12 @@ asmlinkage long sys_io_setup(unsigned nr_events, aio_context_t __user *ctxp)
+ ret = PTR_ERR(ioctx);
+ if (!IS_ERR(ioctx)) {
+ ret = put_user(ioctx->user_id, ctxp);
+- if (!ret)
+- return 0;
++#ifdef CONFIG_EPOLL
++ if (make_fd && ret >= 0)
++ ret = make_aio_fd(ioctx);
++#endif
++ if (ret >= 0)
++ return ret;
+
+ get_ioctx(ioctx); /* io_destroy() expects us to hold a ref */
+ io_destroy(ioctx);
+diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c
+index 97dba0d..5dc788d 100644
+--- a/fs/compat_ioctl.c
++++ b/fs/compat_ioctl.c
+@@ -114,6 +114,13 @@
+ #include <asm/fbio.h>
+ #endif
+
++#ifdef CONFIG_XEN
++#include <xen/interface/xen.h>
++#include <xen/public/evtchn.h>
++#include <xen/public/privcmd.h>
++#include <xen/compat_ioctl.h>
++#endif
++
+ static int do_ioctl32_pointer(unsigned int fd, unsigned int cmd,
+ unsigned long arg, struct file *f)
+ {
+@@ -2833,6 +2840,18 @@ IGNORE_IOCTL(FBIOGETCMAP32)
+ IGNORE_IOCTL(FBIOSCURSOR32)
+ IGNORE_IOCTL(FBIOGCURSOR32)
+ #endif
++
++#ifdef CONFIG_XEN
++HANDLE_IOCTL(IOCTL_PRIVCMD_MMAP_32, privcmd_ioctl_32)
++HANDLE_IOCTL(IOCTL_PRIVCMD_MMAPBATCH_32, privcmd_ioctl_32)
++COMPATIBLE_IOCTL(IOCTL_PRIVCMD_HYPERCALL)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_VIRQ)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_INTERDOMAIN)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_BIND_UNBOUND_PORT)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_UNBIND)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_NOTIFY)
++COMPATIBLE_IOCTL(IOCTL_EVTCHN_RESET)
++#endif
+ };
+
+ #define IOCTL_HASHSIZE 256
+diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
+index 98e0e86..31c5882 100644
+--- a/fs/xfs/linux-2.6/xfs_buf.c
++++ b/fs/xfs/linux-2.6/xfs_buf.c
+@@ -187,7 +187,7 @@ free_address(
+ {
+ a_list_t *aentry;
+
+-#ifdef CONFIG_XEN
++#if defined(CONFIG_XEN) || defined(CONFIG_PARAVIRT_XEN)
+ /*
+ * Xen needs to be able to make sure it can get an exclusive
+ * RO mapping of pages it wants to turn into a pagetable. If
+diff --git a/include/acpi/processor.h b/include/acpi/processor.h
+index 06ebb6e..9637c32 100644
+--- a/include/acpi/processor.h
++++ b/include/acpi/processor.h
+@@ -17,6 +17,12 @@
+ #define ACPI_PROCESSOR_MAX_THROTTLE 250 /* 25% */
+ #define ACPI_PROCESSOR_MAX_DUTY_WIDTH 4
+
++#ifdef CONFIG_XEN
++#define NR_ACPI_CPUS (NR_CPUS < 256 ? 256 : NR_CPUS)
++#else
++#define NR_ACPI_CPUS NR_CPUS
++#endif /* CONFIG_XEN */
++
+ #define ACPI_PDC_REVISION_ID 0x1
+
+ #define ACPI_PSD_REV0_REVISION 0 /* Support for _PSD as in ACPI 3.0 */
+@@ -42,6 +48,17 @@
+
+ struct acpi_processor_cx;
+
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++struct acpi_csd_package {
++ acpi_integer num_entries;
++ acpi_integer revision;
++ acpi_integer domain;
++ acpi_integer coord_type;
++ acpi_integer num_processors;
++ acpi_integer index;
++} __attribute__ ((packed));
++#endif
++
+ struct acpi_power_register {
+ u8 descriptor;
+ u16 length;
+@@ -74,6 +91,12 @@ struct acpi_processor_cx {
+ u32 power;
+ u32 usage;
+ u64 time;
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++ /* Require raw information for external control logic */
++ struct acpi_power_register reg;
++ u32 csd_count;
++ struct acpi_csd_package *domain_info;
++#endif
+ struct acpi_processor_cx_policy promotion;
+ struct acpi_processor_cx_policy demotion;
+ char desc[ACPI_CX_DESC_LEN];
+@@ -304,6 +327,9 @@ static inline void acpi_processor_ppc_exit(void)
+ {
+ return;
+ }
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++int acpi_processor_ppc_has_changed(struct acpi_processor *pr);
++#else
+ static inline int acpi_processor_ppc_has_changed(struct acpi_processor *pr)
+ {
+ static unsigned int printout = 1;
+@@ -316,6 +342,7 @@ static inline int acpi_processor_ppc_has_changed(struct acpi_processor *pr)
+ }
+ return 0;
+ }
++#endif /* CONFIG_PROCESSOR_EXTERNAL_CONTROL */
+ #endif /* CONFIG_CPU_FREQ */
+
+ /* in processor_throttling.c */
+@@ -352,4 +379,79 @@ static inline void acpi_thermal_cpufreq_exit(void)
+ }
+ #endif
+
++/*
++ * Following are interfaces geared to external processor PM control
++ * logic like a VMM
++ */
++/* Events notified to external control logic */
++#define PROCESSOR_PM_INIT 1
++#define PROCESSOR_PM_CHANGE 2
++#define PROCESSOR_HOTPLUG 3
++
++/* Objects for the PM events */
++#define PM_TYPE_IDLE 0
++#define PM_TYPE_PERF 1
++#define PM_TYPE_THR 2
++#define PM_TYPE_MAX 3
++
++/* Processor hotplug events */
++#define HOTPLUG_TYPE_ADD 0
++#define HOTPLUG_TYPE_REMOVE 1
++
++#ifdef CONFIG_PROCESSOR_EXTERNAL_CONTROL
++struct processor_extcntl_ops {
++ /* Transfer processor PM events to external control logic */
++ int (*pm_ops[PM_TYPE_MAX])(struct acpi_processor *pr, int event);
++ /* Notify physical processor status to external control logic */
++ int (*hotplug)(struct acpi_processor *pr, int event);
++};
++extern const struct processor_extcntl_ops *processor_extcntl_ops;
++
++static inline int processor_cntl_external(void)
++{
++ return (processor_extcntl_ops != NULL);
++}
++
++static inline int processor_pm_external(void)
++{
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_IDLE] != NULL);
++}
++
++static inline int processor_pmperf_external(void)
++{
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_PERF] != NULL);
++}
++
++static inline int processor_pmthr_external(void)
++{
++ return processor_cntl_external() &&
++ (processor_extcntl_ops->pm_ops[PM_TYPE_THR] != NULL);
++}
++
++extern int processor_notify_external(struct acpi_processor *pr,
++ int event, int type);
++extern void processor_extcntl_init(void);
++extern int processor_extcntl_prepare(struct acpi_processor *pr);
++extern int acpi_processor_get_performance_info(struct acpi_processor *pr);
++extern int acpi_processor_get_psd(struct acpi_processor *pr);
++void arch_acpi_processor_init_extcntl(const struct processor_extcntl_ops **);
++#else
++static inline int processor_cntl_external(void) {return 0;}
++static inline int processor_pm_external(void) {return 0;}
++static inline int processor_pmperf_external(void) {return 0;}
++static inline int processor_pmthr_external(void) {return 0;}
++static inline int processor_notify_external(struct acpi_processor *pr,
++ int event, int type)
++{
++ return 0;
++}
++static inline void processor_extcntl_init(void) {}
++static inline int processor_extcntl_prepare(struct acpi_processor *pr)
++{
++ return 0;
++}
++#endif /* CONFIG_PROCESSOR_EXTERNAL_CONTROL */
++
+ #endif
+diff --git a/include/asm-generic/pci.h b/include/asm-generic/pci.h
+index c36a77d..daed1c5 100644
+--- a/include/asm-generic/pci.h
++++ b/include/asm-generic/pci.h
+@@ -43,7 +43,9 @@ pcibios_select_root(struct pci_dev *pdev, struct resource *res)
+ return root;
+ }
+
++#ifndef pcibios_scan_all_fns
+ #define pcibios_scan_all_fns(a, b) 0
++#endif
+
+ #ifndef HAVE_ARCH_PCI_GET_LEGACY_IDE_IRQ
+ static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel)
+diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
+index 44ef329..7ce4039 100644
+--- a/include/asm-generic/pgtable.h
++++ b/include/asm-generic/pgtable.h
+@@ -99,6 +99,10 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addres
+ }
+ #endif
+
++#ifndef arch_change_pte_range
++#define arch_change_pte_range(mm, pmd, addr, end, newprot, dirty_accountable) 0
++#endif
++
+ #ifndef __HAVE_ARCH_PTE_SAME
+ #define pte_same(A,B) (pte_val(A) == pte_val(B))
+ #endif
+diff --git a/include/asm-x86/acpi.h b/include/asm-x86/acpi.h
+index 14411c9..7cf48f9 100644
+--- a/include/asm-x86/acpi.h
++++ b/include/asm-x86/acpi.h
+@@ -29,6 +29,10 @@
+ #include <asm/processor.h>
+ #include <asm/mmu.h>
+
++#ifdef CONFIG_XEN
++#include <xen/interface/platform.h>
++#endif
++
+ #define COMPILER_DEPENDENT_INT64 long long
+ #define COMPILER_DEPENDENT_UINT64 unsigned long long
+
+@@ -123,6 +127,27 @@ extern unsigned long acpi_wakeup_address;
+ /* early initialization routine */
+ extern void acpi_reserve_bootmem(void);
+
++#ifdef CONFIG_XEN
++static inline int acpi_notify_hypervisor_state(u8 sleep_state,
++ u32 pm1a_cnt_val,
++ u32 pm1b_cnt_val)
++{
++ struct xen_platform_op op = {
++ .cmd = XENPF_enter_acpi_sleep,
++ .interface_version = XENPF_INTERFACE_VERSION,
++ .u = {
++ .enter_acpi_sleep = {
++ .pm1a_cnt_val = pm1a_cnt_val,
++ .pm1b_cnt_val = pm1b_cnt_val,
++ .sleep_state = sleep_state,
++ },
++ },
++ };
++
++ return HYPERVISOR_platform_op(&op);
++}
++#endif /* CONFIG_XEN */
++
+ /*
+ * Check if the CPU can handle C2 and deeper
+ */
+@@ -153,7 +178,9 @@ static inline void disable_acpi(void) { }
+
+ #endif /* !CONFIG_ACPI */
+
++#ifndef CONFIG_XEN
+ #define ARCH_HAS_POWER_INIT 1
++#endif
+
+ struct bootnode;
+
+diff --git a/include/asm-x86/apic.h b/include/asm-x86/apic.h
+index be9639a..708f512 100644
+--- a/include/asm-x86/apic.h
++++ b/include/asm-x86/apic.h
+@@ -3,12 +3,16 @@
+
+ #include <linux/pm.h>
+ #include <linux/delay.h>
++#ifndef CONFIG_XEN
+ #include <asm/fixmap.h>
++#endif
+ #include <asm/apicdef.h>
+ #include <asm/processor.h>
+ #include <asm/system.h>
+
++#ifndef CONFIG_XEN
+ #define ARCH_APICTIMER_STOPS_ON_C3 1
++#endif
+
+ #define Dprintk(x...)
+
+@@ -36,6 +40,7 @@ extern void generic_apic_probe(void);
+ #ifdef CONFIG_X86_LOCAL_APIC
+
+ extern int apic_verbosity;
++#ifndef CONFIG_XEN
+ extern int timer_over_8254;
+ extern int local_apic_timer_c2_ok;
+ extern int local_apic_timer_disabled;
+@@ -101,6 +106,7 @@ static inline void ack_APIC_irq(void)
+ /* Docs say use 0 for future compatibility */
+ apic_write_around(APIC_EOI, 0);
+ }
++#endif
+
+ extern int lapic_get_maxlvt(void);
+ extern void clear_local_APIC(void);
+diff --git a/include/asm-x86/apicdef.h b/include/asm-x86/apicdef.h
+index 6b9008c..2774dff 100644
+--- a/include/asm-x86/apicdef.h
++++ b/include/asm-x86/apicdef.h
+@@ -1,6 +1,8 @@
+ #ifndef _ASM_X86_APICDEF_H
+ #define _ASM_X86_APICDEF_H
+
++#ifndef CONFIG_XEN
++
+ /*
+ * Constants for various Intel APICs. (local APIC, IOAPIC, etc.)
+ *
+@@ -129,6 +131,16 @@
+
+ #define APIC_BASE (fix_to_virt(FIX_APIC_BASE))
+
++#else /* CONFIG_XEN */
++
++enum {
++ APIC_DEST_ALLBUT = 0x1,
++ APIC_DEST_SELF,
++ APIC_DEST_ALLINC
++};
++
++#endif /* CONFIG_XEN */
++
+ #ifdef CONFIG_X86_32
+ # define MAX_IO_APICS 64
+ #else
+@@ -136,6 +148,8 @@
+ # define MAX_LOCAL_APIC 32768
+ #endif
+
++#ifndef CONFIG_XEN
++
+ /*
+ * All x86-64 systems are xAPIC compatible.
+ * In the following, "apicid" is a physical APIC ID.
+@@ -406,6 +420,8 @@ struct local_apic {
+
+ #undef u32
+
++#endif /* CONFIG_XEN */
++
+ #ifdef CONFIG_X86_32
+ #define BAD_APICID 0xFFu
+ #else
+diff --git a/include/asm-x86/dma-mapping.h b/include/asm-x86/dma-mapping.h
+index a1a4dc7..4e0b4e7 100644
+--- a/include/asm-x86/dma-mapping.h
++++ b/include/asm-x86/dma-mapping.h
+@@ -187,8 +187,13 @@ static inline dma_addr_t dma_map_page(struct device *dev, struct page *page,
+ int direction)
+ {
+ BUG_ON(!valid_dma_direction(direction));
++#ifndef CONFIG_XEN
+ return dma_ops->map_single(dev, page_to_phys(page)+offset,
+ size, direction);
++#else
++ return dma_ops->map_single(dev, page_to_pseudophys(page)+offset,
++ size, direction);
++#endif
+ }
+
+ static inline void dma_unmap_page(struct device *dev, dma_addr_t addr,
+diff --git a/include/asm-x86/genapic_64.h b/include/asm-x86/genapic_64.h
+index 1de931b..33e49a6 100644
+--- a/include/asm-x86/genapic_64.h
++++ b/include/asm-x86/genapic_64.h
+@@ -35,6 +35,7 @@ extern struct genapic apic_flat;
+ extern struct genapic apic_physflat;
+ extern int acpi_madt_oem_check(char *, char *);
+
++#ifndef CONFIG_XEN
+ enum uv_system_type {UV_NONE, UV_LEGACY_APIC, UV_X2APIC, UV_NON_UNIQUE_APIC};
+ extern enum uv_system_type get_uv_system_type(void);
+ extern int is_uv_system(void);
+@@ -43,5 +44,9 @@ extern struct genapic apic_x2apic_uv_x;
+ DECLARE_PER_CPU(int, x2apic_extra_bits);
+ extern void uv_cpu_init(void);
+ extern int uv_wakeup_secondary(int phys_apicid, unsigned int start_rip);
++#else
++#define is_uv_system() 0
++#define uv_cpu_init() ((void)0)
++#endif
+
+ #endif
+diff --git a/include/asm-x86/i8253.h b/include/asm-x86/i8253.h
+index b51c048..e006e18 100644
+--- a/include/asm-x86/i8253.h
++++ b/include/asm-x86/i8253.h
+@@ -8,10 +8,14 @@
+
+ extern spinlock_t i8253_lock;
+
++#ifdef CONFIG_GENERIC_CLOCKEVENTS
++
+ extern struct clock_event_device *global_clock_event;
+
+ extern void setup_pit_timer(void);
+
++#endif
++
+ #define inb_pit inb_p
+ #define outb_pit outb_p
+
+diff --git a/include/asm-x86/ipi.h b/include/asm-x86/ipi.h
+index ecc80f3..0191fc7 100644
+--- a/include/asm-x86/ipi.h
++++ b/include/asm-x86/ipi.h
+@@ -1,6 +1,8 @@
+ #ifndef __ASM_IPI_H
+ #define __ASM_IPI_H
+
++#ifndef CONFIG_XEN
++
+ /*
+ * Copyright 2004 James Cleverdon, IBM.
+ * Subject to the GNU Public License, v.2
+@@ -128,4 +130,6 @@ static inline void send_IPI_mask_sequence(cpumask_t mask, int vector)
+ local_irq_restore(flags);
+ }
+
++#endif /* CONFIG_XEN */
++
+ #endif /* __ASM_IPI_H */
+diff --git a/include/asm-x86/irq_32.h b/include/asm-x86/irq_32.h
+index 0b79f31..4c8b1a3 100644
+--- a/include/asm-x86/irq_32.h
++++ b/include/asm-x86/irq_32.h
+@@ -20,7 +20,7 @@ static inline int irq_canonicalize(int irq)
+ return ((irq == 2) ? 9 : irq);
+ }
+
+-#ifdef CONFIG_X86_LOCAL_APIC
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
+ # define ARCH_HAS_NMI_WATCHDOG /* See include/linux/nmi.h */
+ #endif
+
+diff --git a/include/asm-x86/kexec.h b/include/asm-x86/kexec.h
+index 8f855a1..a1e9ab7 100644
+--- a/include/asm-x86/kexec.h
++++ b/include/asm-x86/kexec.h
+@@ -164,6 +164,19 @@ relocate_kernel(unsigned long indirection_page,
+ unsigned long start_address) ATTRIB_NORET;
+ #endif
+
++/* Under Xen we need to work with machine addresses. These macros give the
++ * machine address of a certain page to the generic kexec code instead of
++ * the pseudo physical address which would be given by the default macros.
++ */
++
++#ifdef CONFIG_XEN
++#define KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) pfn_to_mfn(page_to_pfn(page))
++#define kexec_pfn_to_page(pfn) pfn_to_page(mfn_to_pfn(pfn))
++#define kexec_virt_to_phys(addr) virt_to_machine(addr)
++#define kexec_phys_to_virt(addr) phys_to_virt(machine_to_phys(addr))
++#endif
++
+ #endif /* __ASSEMBLY__ */
+
+ #endif /* _KEXEC_H */
+diff --git a/include/asm-x86/mach-xen/asm/agp.h b/include/asm-x86/mach-xen/asm/agp.h
+new file mode 100644
+index 0000000..c0bb022
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/agp.h
+@@ -0,0 +1,43 @@
++#ifndef _ASM_X86_AGP_H
++#define _ASM_X86_AGP_H
++
++#include <asm/pgtable.h>
++#include <asm/cacheflush.h>
++#include <asm/system.h>
++
++/*
++ * Functions to keep the agpgart mappings coherent with the MMU. The
++ * GART gives the CPU a physical alias of pages in memory. The alias
++ * region is mapped uncacheable. Make sure there are no conflicting
++ * mappings with different cachability attributes for the same
++ * page. This avoids data corruption on some CPUs.
++ */
++
++#define map_page_into_agp(page) ( \
++ xen_create_contiguous_region((unsigned long)page_address(page), 0, 32) \
++ ?: set_pages_uc(page, 1))
++#define unmap_page_from_agp(page) ( \
++ xen_destroy_contiguous_region((unsigned long)page_address(page), 0), \
++ /* only a fallback: xen_destroy_contiguous_region uses PAGE_KERNEL */ \
++ set_pages_wb(page, 1))
++
++/*
++ * Could use CLFLUSH here if the cpu supports it. But then it would
++ * need to be called for each cacheline of the whole page so it may
++ * not be worth it. Would need a page for it.
++ */
++#define flush_agp_cache() wbinvd()
++
++/* Convert a physical address to an address suitable for the GART. */
++#define phys_to_gart(x) phys_to_machine(x)
++#define gart_to_phys(x) machine_to_phys(x)
++
++/* GATT allocation. Returns/accepts GATT kernel virtual address. */
++#define alloc_gatt_pages(order) ({ \
++ char *_t; dma_addr_t _d; \
++ _t = dma_alloc_coherent(NULL,PAGE_SIZE<<(order),&_d,GFP_KERNEL); \
++ _t; })
++#define free_gatt_pages(table, order) \
++ dma_free_coherent(NULL,PAGE_SIZE<<(order),(table),virt_to_bus(table))
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/desc.h b/include/asm-x86/mach-xen/asm/desc.h
+new file mode 100644
+index 0000000..21f56ab
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/desc.h
+@@ -0,0 +1,405 @@
++#ifndef _ASM_DESC_H_
++#define _ASM_DESC_H_
++
++#ifndef __ASSEMBLY__
++#include <asm/desc_defs.h>
++#include <asm/ldt.h>
++#include <asm/mmu.h>
++#include <linux/smp.h>
++
++static inline void fill_ldt(struct desc_struct *desc,
++ const struct user_desc *info)
++{
++ desc->limit0 = info->limit & 0x0ffff;
++ desc->base0 = info->base_addr & 0x0000ffff;
++
++ desc->base1 = (info->base_addr & 0x00ff0000) >> 16;
++ desc->type = (info->read_exec_only ^ 1) << 1;
++ desc->type |= info->contents << 2;
++ desc->s = 1;
++ desc->dpl = 0x3;
++ desc->p = info->seg_not_present ^ 1;
++ desc->limit = (info->limit & 0xf0000) >> 16;
++ desc->avl = info->useable;
++ desc->d = info->seg_32bit;
++ desc->g = info->limit_in_pages;
++ desc->base2 = (info->base_addr & 0xff000000) >> 24;
++}
++
++#ifndef CONFIG_X86_NO_IDT
++extern struct desc_ptr idt_descr;
++extern gate_desc idt_table[];
++#endif
++
++#ifdef CONFIG_X86_64
++extern struct desc_struct cpu_gdt_table[GDT_ENTRIES];
++extern struct desc_ptr cpu_gdt_descr[];
++/* the cpu gdt accessor */
++#define get_cpu_gdt_table(x) ((struct desc_struct *)cpu_gdt_descr[x].address)
++
++static inline void pack_gate(gate_desc *gate, unsigned type, unsigned long func,
++ unsigned dpl, unsigned ist, unsigned seg)
++{
++ gate->offset_low = PTR_LOW(func);
++ gate->segment = __KERNEL_CS;
++ gate->ist = ist;
++ gate->p = 1;
++ gate->dpl = dpl;
++ gate->zero0 = 0;
++ gate->zero1 = 0;
++ gate->type = type;
++ gate->offset_middle = PTR_MIDDLE(func);
++ gate->offset_high = PTR_HIGH(func);
++}
++
++#else
++struct gdt_page {
++ struct desc_struct gdt[GDT_ENTRIES];
++} __attribute__((aligned(PAGE_SIZE)));
++DECLARE_PER_CPU(struct gdt_page, gdt_page);
++
++static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
++{
++ return per_cpu(gdt_page, cpu).gdt;
++}
++
++static inline void pack_gate(gate_desc *gate, unsigned char type,
++ unsigned long base, unsigned dpl, unsigned flags,
++ unsigned short seg)
++{
++ gate->a = (seg << 16) | (base & 0xffff);
++ gate->b = (base & 0xffff0000) |
++ (((0x80 | type | (dpl << 5)) & 0xff) << 8);
++}
++
++#endif
++
++static inline int desc_empty(const void *ptr)
++{
++ const u32 *desc = ptr;
++ return !(desc[0] | desc[1]);
++}
++
++#ifndef CONFIG_XEN
++#define load_TR_desc() native_load_tr_desc()
++#define load_gdt(dtr) native_load_gdt(dtr)
++#define load_idt(dtr) native_load_idt(dtr)
++#define load_tr(tr) asm volatile("ltr %0"::"m" (tr))
++#define load_ldt(ldt) asm volatile("lldt %0"::"m" (ldt))
++
++#define store_gdt(dtr) native_store_gdt(dtr)
++#define store_idt(dtr) native_store_idt(dtr)
++#define store_tr(tr) (tr = native_store_tr())
++#define store_ldt(ldt) asm("sldt %0":"=m" (ldt))
++
++#define load_TLS(t, cpu) native_load_tls(t, cpu)
++#define set_ldt native_set_ldt
++
++#define write_ldt_entry(dt, entry, desc) \
++ native_write_ldt_entry(dt, entry, desc)
++#define write_gdt_entry(dt, entry, desc, type) \
++ native_write_gdt_entry(dt, entry, desc, type)
++#define write_idt_entry(dt, entry, g) \
++ native_write_idt_entry(dt, entry, g)
++
++static inline void native_write_idt_entry(gate_desc *idt, int entry,
++ const gate_desc *gate)
++{
++ memcpy(&idt[entry], gate, sizeof(*gate));
++}
++
++static inline void native_write_ldt_entry(struct desc_struct *ldt, int entry,
++ const void *desc)
++{
++ memcpy(&ldt[entry], desc, 8);
++}
++
++static inline void native_write_gdt_entry(struct desc_struct *gdt, int entry,
++ const void *desc, int type)
++{
++ unsigned int size;
++ switch (type) {
++ case DESC_TSS:
++ size = sizeof(tss_desc);
++ break;
++ case DESC_LDT:
++ size = sizeof(ldt_desc);
++ break;
++ default:
++ size = sizeof(struct desc_struct);
++ break;
++ }
++ memcpy(&gdt[entry], desc, size);
++}
++#endif
++
++static inline void pack_descriptor(struct desc_struct *desc, unsigned long base,
++ unsigned long limit, unsigned char type,
++ unsigned char flags)
++{
++ desc->a = ((base & 0xffff) << 16) | (limit & 0xffff);
++ desc->b = (base & 0xff000000) | ((base & 0xff0000) >> 16) |
++ (limit & 0x000f0000) | ((type & 0xff) << 8) |
++ ((flags & 0xf) << 20);
++ desc->p = 1;
++}
++
++
++#ifndef CONFIG_XEN
++static inline void set_tssldt_descriptor(void *d, unsigned long addr,
++ unsigned type, unsigned size)
++{
++#ifdef CONFIG_X86_64
++ struct ldttss_desc64 *desc = d;
++ memset(desc, 0, sizeof(*desc));
++ desc->limit0 = size & 0xFFFF;
++ desc->base0 = PTR_LOW(addr);
++ desc->base1 = PTR_MIDDLE(addr) & 0xFF;
++ desc->type = type;
++ desc->p = 1;
++ desc->limit1 = (size >> 16) & 0xF;
++ desc->base2 = (PTR_MIDDLE(addr) >> 8) & 0xFF;
++ desc->base3 = PTR_HIGH(addr);
++#else
++ pack_descriptor((struct desc_struct *)d, addr, size, 0x80 | type, 0);
++#endif
++}
++
++static inline void __set_tss_desc(unsigned cpu, unsigned int entry, void *addr)
++{
++ struct desc_struct *d = get_cpu_gdt_table(cpu);
++ tss_desc tss;
++
++ /*
++ * sizeof(unsigned long) coming from an extra "long" at the end
++ * of the iobitmap. See tss_struct definition in processor.h
++ *
++ * -1? seg base+limit should be pointing to the address of the
++ * last valid byte
++ */
++ set_tssldt_descriptor(&tss, (unsigned long)addr, DESC_TSS,
++ IO_BITMAP_OFFSET + IO_BITMAP_BYTES +
++ sizeof(unsigned long) - 1);
++ write_gdt_entry(d, entry, &tss, DESC_TSS);
++}
++
++#define set_tss_desc(cpu, addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr)
++
++static inline 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();
++ ldt_desc ldt;
++
++ set_tssldt_descriptor(&ldt, (unsigned long)addr, DESC_LDT,
++ entries * LDT_ENTRY_SIZE - 1);
++ write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_LDT,
++ &ldt, DESC_LDT);
++ asm volatile("lldt %w0"::"q" (GDT_ENTRY_LDT*8));
++ }
++}
++
++static inline void native_load_tr_desc(void)
++{
++ asm volatile("ltr %w0"::"q" (GDT_ENTRY_TSS*8));
++}
++
++static inline void native_load_gdt(const struct desc_ptr *dtr)
++{
++ asm volatile("lgdt %0"::"m" (*dtr));
++}
++
++static inline void native_load_idt(const struct desc_ptr *dtr)
++{
++ asm volatile("lidt %0"::"m" (*dtr));
++}
++
++static inline void native_store_gdt(struct desc_ptr *dtr)
++{
++ asm volatile("sgdt %0":"=m" (*dtr));
++}
++
++static inline void native_store_idt(struct desc_ptr *dtr)
++{
++ asm volatile("sidt %0":"=m" (*dtr));
++}
++
++static inline unsigned long native_store_tr(void)
++{
++ unsigned long tr;
++ asm volatile("str %0":"=r" (tr));
++ return tr;
++}
++
++static inline void native_load_tls(struct thread_struct *t, unsigned int cpu)
++{
++ unsigned int i;
++ struct desc_struct *gdt = get_cpu_gdt_table(cpu);
++
++ for (i = 0; i < GDT_ENTRY_TLS_ENTRIES; i++)
++ gdt[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i];
++}
++#else
++#define load_TLS(t, cpu) xen_load_tls(t, cpu)
++#define set_ldt xen_set_ldt
++
++extern int write_ldt_entry(struct desc_struct *ldt, int entry,
++ const void *desc);
++extern int write_gdt_entry(struct desc_struct *gdt, int entry,
++ const void *desc, int type);
++
++static inline void xen_load_tls(struct thread_struct *t, unsigned int cpu)
++{
++ unsigned int i;
++ struct desc_struct *gdt = get_cpu_gdt_table(cpu) + GDT_ENTRY_TLS_MIN;
++
++ for (i = 0; i < GDT_ENTRY_TLS_ENTRIES; i++)
++ if (HYPERVISOR_update_descriptor(virt_to_machine(&gdt[i]),
++ *(u64 *)&t->tls_array[i]))
++ BUG();
++}
++#endif
++
++#define _LDT_empty(info) \
++ ((info)->base_addr == 0 && \
++ (info)->limit == 0 && \
++ (info)->contents == 0 && \
++ (info)->read_exec_only == 1 && \
++ (info)->seg_32bit == 0 && \
++ (info)->limit_in_pages == 0 && \
++ (info)->seg_not_present == 1 && \
++ (info)->useable == 0)
++
++#ifdef CONFIG_X86_64
++#define LDT_empty(info) (_LDT_empty(info) && ((info)->lm == 0))
++#else
++#define LDT_empty(info) (_LDT_empty(info))
++#endif
++
++static inline void clear_LDT(void)
++{
++ set_ldt(NULL, 0);
++}
++
++/*
++ * load one particular LDT into the current CPU
++ */
++static inline void load_LDT_nolock(mm_context_t *pc)
++{
++ set_ldt(pc->ldt, pc->size);
++}
++
++static inline void load_LDT(mm_context_t *pc)
++{
++ preempt_disable();
++ load_LDT_nolock(pc);
++ preempt_enable();
++}
++
++static inline unsigned long get_desc_base(const struct desc_struct *desc)
++{
++ return desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24);
++}
++
++static inline unsigned long get_desc_limit(const struct desc_struct *desc)
++{
++ return desc->limit0 | (desc->limit << 16);
++}
++
++#ifndef CONFIG_X86_NO_IDT
++static inline void _set_gate(int gate, unsigned type, void *addr,
++ unsigned dpl, unsigned ist, unsigned seg)
++{
++ gate_desc s;
++ pack_gate(&s, type, (unsigned long)addr, dpl, ist, seg);
++ /*
++ * does not need to be atomic because it is only done once at
++ * setup time
++ */
++ write_idt_entry(idt_table, gate, &s);
++}
++
++/*
++ * This needs to use 'idt_table' rather than 'idt', and
++ * thus use the _nonmapped_ version of the IDT, as the
++ * Pentium F0 0F bugfix can have resulted in the mapped
++ * IDT being write-protected.
++ */
++static inline void set_intr_gate(unsigned int n, void *addr)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_INTERRUPT, addr, 0, 0, __KERNEL_CS);
++}
++
++/*
++ * This routine sets up an interrupt gate at directory privilege level 3.
++ */
++static inline void set_system_intr_gate(unsigned int n, void *addr)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_INTERRUPT, addr, 0x3, 0, __KERNEL_CS);
++}
++
++static inline void set_trap_gate(unsigned int n, void *addr)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_TRAP, addr, 0, 0, __KERNEL_CS);
++}
++
++static inline void set_system_gate(unsigned int n, void *addr)
++{
++ BUG_ON((unsigned)n > 0xFF);
++#ifdef CONFIG_X86_32
++ _set_gate(n, GATE_TRAP, addr, 0x3, 0, __KERNEL_CS);
++#else
++ _set_gate(n, GATE_INTERRUPT, addr, 0x3, 0, __KERNEL_CS);
++#endif
++}
++
++static inline void set_task_gate(unsigned int n, unsigned int gdt_entry)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_TASK, (void *)0, 0, 0, (gdt_entry<<3));
++}
++
++static inline void set_intr_gate_ist(int n, void *addr, unsigned ist)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_INTERRUPT, addr, 0, ist, __KERNEL_CS);
++}
++
++static inline void set_system_gate_ist(int n, void *addr, unsigned ist)
++{
++ BUG_ON((unsigned)n > 0xFF);
++ _set_gate(n, GATE_INTERRUPT, addr, 0x3, ist, __KERNEL_CS);
++}
++#endif
++
++#else
++/*
++ * 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 --git a/include/asm-x86/mach-xen/asm/dma-mapping.h b/include/asm-x86/mach-xen/asm/dma-mapping.h
+new file mode 100644
+index 0000000..98c2db9
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/dma-mapping.h
+@@ -0,0 +1,17 @@
++#ifndef _ASM_DMA_MAPPING_H_
++
++#include "../../dma-mapping.h"
++
++static inline int
++address_needs_mapping(struct device *hwdev, dma_addr_t addr)
++{
++ dma_addr_t mask = 0xffffffff;
++ /* If the device has a mask, use it, otherwise default to 32 bits */
++ if (hwdev && hwdev->dma_mask)
++ mask = *hwdev->dma_mask;
++ return (addr & ~mask) != 0;
++}
++
++extern int range_straddles_page_boundary(paddr_t p, size_t size);
++
++#endif /* _ASM_DMA_MAPPING_H_ */
+diff --git a/include/asm-x86/mach-xen/asm/e820.h b/include/asm-x86/mach-xen/asm/e820.h
+new file mode 100644
+index 0000000..baea144
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/e820.h
+@@ -0,0 +1,39 @@
++#ifndef __ASM_E820_H
++#define __ASM_E820_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
++#define E820_NVS 4
++
++#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 {
++ __u32 nr_map;
++ struct e820entry map[E820MAX];
++};
++#endif /* __ASSEMBLY__ */
++
++#define ISA_START_ADDRESS 0xa0000
++#define ISA_END_ADDRESS 0x100000
++
++#define BIOS_BEGIN 0x000a0000
++#define BIOS_END 0x00100000
++
++#ifdef __KERNEL__
++#ifdef CONFIG_X86_32
++# include "../../e820_32.h"
++#else
++# include "e820_64.h"
++#endif
++#endif /* __KERNEL__ */
++
++#endif /* __ASM_E820_H */
+diff --git a/include/asm-x86/mach-xen/asm/e820_64.h b/include/asm-x86/mach-xen/asm/e820_64.h
+new file mode 100644
+index 0000000..504fc26
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/e820_64.h
+@@ -0,0 +1,56 @@
++/*
++ * 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/ioport.h>
++
++#ifndef __ASSEMBLY__
++extern unsigned long find_e820_area(unsigned long start, unsigned long end,
++ unsigned long size, unsigned long align);
++extern unsigned long find_e820_area_size(unsigned long start,
++ unsigned long *sizep,
++ unsigned long align);
++extern void add_memory_region(unsigned long start, unsigned long size,
++ int type);
++extern void update_memory_range(u64 start, u64 size, unsigned old_type,
++ unsigned new_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_mark_nosave_regions(void);
++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 int e820_any_non_reserved(unsigned long start, unsigned long end);
++extern int is_memory_any_valid(unsigned long start, unsigned long end);
++extern int e820_all_non_reserved(unsigned long start, unsigned long end);
++extern int is_memory_all_valid(unsigned long start, unsigned long end);
++extern unsigned long e820_hole_size(unsigned long start, unsigned long end);
++
++extern void e820_setup_gap(struct e820entry *e820, int nr_map);
++extern void e820_register_active_regions(int nid, unsigned long start_pfn,
++ unsigned long end_pfn);
++
++extern void finish_e820_parsing(void);
++
++extern struct e820map e820;
++extern void update_e820(void);
++
++extern void reserve_early(unsigned long start, unsigned long end, char *name);
++extern void free_early(unsigned long start, unsigned long end);
++extern void early_res_to_bootmem(unsigned long start, unsigned long end);
++
++#endif/*!__ASSEMBLY__*/
++
++#endif/*__E820_HEADER*/
+diff --git a/include/asm-x86/mach-xen/asm/fixmap.h b/include/asm-x86/mach-xen/asm/fixmap.h
+new file mode 100644
+index 0000000..5bd2069
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/fixmap.h
+@@ -0,0 +1,13 @@
++#ifndef _ASM_FIXMAP_H
++#define _ASM_FIXMAP_H
++
++#ifdef CONFIG_X86_32
++# include "fixmap_32.h"
++#else
++# include "fixmap_64.h"
++#endif
++
++#define clear_fixmap(idx) \
++ __set_fixmap(idx, 0, __pgprot(0))
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/fixmap_32.h b/include/asm-x86/mach-xen/asm/fixmap_32.h
+new file mode 100644
+index 0000000..ac4ed3f
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/fixmap_32.h
+@@ -0,0 +1,170 @@
++/*
++ * fixmap.h: compile-time virtual memory allocation
++ *
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file "COPYING" in the main directory of this archive
++ * for more details.
++ *
++ * Copyright (C) 1998 Ingo Molnar
++ *
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
++ */
++
++#ifndef _ASM_FIXMAP_32_H
++#define _ASM_FIXMAP_32_H
++
++/* used by vmalloc.c, vsyscall.lds.S.
++ *
++ * Leave one empty page between vmalloc'ed areas and
++ * the start of the fixmap.
++ */
++extern unsigned long __FIXADDR_TOP;
++#define FIXADDR_USER_START __fix_to_virt(FIX_VDSO)
++#define FIXADDR_USER_END __fix_to_virt(FIX_VDSO - 1)
++
++#ifndef __ASSEMBLY__
++#include <linux/kernel.h>
++#include <asm/acpi.h>
++#include <asm/apicdef.h>
++#include <asm/page.h>
++#ifdef CONFIG_HIGHMEM
++#include <linux/threads.h>
++#include <asm/kmap_types.h>
++#endif
++
++/*
++ * Here we define all the compile-time 'special' virtual
++ * addresses. The point is to have a constant address at
++ * compile time, but to set the physical address only
++ * in the boot process. We allocate these special addresses
++ * from the end of virtual memory (0xfffff000) backwards.
++ * Also this lets us do fail-safe vmalloc(), we
++ * can guarantee that these special addresses and
++ * vmalloc()-ed addresses never overlap.
++ *
++ * these 'compile-time allocated' memory buffers are
++ * fixed-size 4k pages. (or larger if used with an increment
++ * highger than 1) use fixmap_set(idx,phys) to associate
++ * physical memory with fixmap indices.
++ *
++ * TLB entries of such buffers will not be flushed across
++ * task switches.
++ */
++enum fixed_addresses {
++ FIX_HOLE,
++ FIX_VDSO,
++ FIX_DBGP_BASE,
++ FIX_EARLYCON_MEM_BASE,
++#ifndef CONFIG_XEN
++#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
++#endif
++#ifdef CONFIG_X86_VISWS_APIC
++ FIX_CO_CPU, /* Cobalt timer */
++ FIX_CO_APIC, /* Cobalt APIC Redirection Table */
++ FIX_LI_PCIA, /* Lithium PCI Bridge A */
++ FIX_LI_PCIB, /* Lithium PCI Bridge B */
++#endif
++#ifdef CONFIG_X86_F00F_BUG
++ FIX_F00F_IDT, /* Virtual mapping for IDT */
++#endif
++#ifdef CONFIG_X86_CYCLONE_TIMER
++ FIX_CYCLONE_TIMER, /*cyclone timer register*/
++#endif
++#ifdef CONFIG_HIGHMEM
++ FIX_KMAP_BEGIN, /* reserved pte's for temporary kernel mappings */
++ FIX_KMAP_END = FIX_KMAP_BEGIN+(KM_TYPE_NR*NR_CPUS)-1,
++#endif
++#ifdef CONFIG_ACPI
++ FIX_ACPI_BEGIN,
++ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
++#endif
++#ifdef CONFIG_PCI_MMCONFIG
++ FIX_PCIE_MCFG,
++#endif
++#ifdef CONFIG_PARAVIRT
++ FIX_PARAVIRT_BOOTMAP,
++#endif
++ FIX_SHARED_INFO,
++#define NR_FIX_ISAMAPS 256
++ FIX_ISAMAP_END,
++ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
++ __end_of_permanent_fixed_addresses,
++ /*
++ * 256 temporary boot-time mappings, used by early_ioremap(),
++ * before ioremap() is functional.
++ *
++ * We round it up to the next 512 pages boundary so that we
++ * can have a single pgd entry and a single pte table:
++ */
++#define NR_FIX_BTMAPS 64
++#define FIX_BTMAPS_NESTING 4
++ FIX_BTMAP_END = __end_of_permanent_fixed_addresses + 512 -
++ (__end_of_permanent_fixed_addresses & 511),
++ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS*FIX_BTMAPS_NESTING - 1,
++ FIX_WP_TEST,
++#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT
++ FIX_OHCI1394_BASE,
++#endif
++ __end_of_fixed_addresses
++};
++
++extern void __set_fixmap(enum fixed_addresses idx,
++ maddr_t phys, pgprot_t flags);
++extern void reserve_top_address(unsigned long reserve);
++
++#define set_fixmap(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL)
++/*
++ * Some hardware wants to get fixmapped without caching.
++ */
++#define set_fixmap_nocache(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
++
++#define FIXADDR_TOP ((unsigned long)__FIXADDR_TOP)
++
++#define __FIXADDR_SIZE (__end_of_permanent_fixed_addresses << PAGE_SHIFT)
++#define __FIXADDR_BOOT_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
++#define FIXADDR_START (FIXADDR_TOP - __FIXADDR_SIZE)
++#define FIXADDR_BOOT_START (FIXADDR_TOP - __FIXADDR_BOOT_SIZE)
++
++#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
++#define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT)
++
++extern void __this_fixmap_does_not_exist(void);
++
++/*
++ * 'index to address' translation. If anyone tries to use the idx
++ * directly without tranlation, we catch the bug with a NULL-deference
++ * kernel oops. Illegal ranges of incoming indices are caught too.
++ */
++static __always_inline unsigned long fix_to_virt(const unsigned int idx)
++{
++ /*
++ * this branch gets completely eliminated after inlining,
++ * except when someone tries to use fixaddr indices in an
++ * illegal way. (such as mixing up address types or using
++ * out-of-range indices).
++ *
++ * If it doesn't get removed, the linker will complain
++ * loudly with a reasonably clear error message..
++ */
++ if (idx >= __end_of_fixed_addresses)
++ __this_fixmap_does_not_exist();
++
++ return __fix_to_virt(idx);
++}
++
++static inline unsigned long virt_to_fix(const unsigned long vaddr)
++{
++ BUG_ON(vaddr >= FIXADDR_TOP || vaddr < FIXADDR_START);
++ return __virt_to_fix(vaddr);
++}
++
++#endif /* !__ASSEMBLY__ */
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/fixmap_64.h b/include/asm-x86/mach-xen/asm/fixmap_64.h
+new file mode 100644
+index 0000000..b275f57
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/fixmap_64.h
+@@ -0,0 +1,126 @@
++/*
++ * fixmap.h: compile-time virtual memory allocation
++ *
++ * This file is subject to the terms and conditions of the GNU General Public
++ * License. See the file "COPYING" in the main directory of this archive
++ * for more details.
++ *
++ * Copyright (C) 1998 Ingo Molnar
++ */
++
++#ifndef _ASM_FIXMAP_64_H
++#define _ASM_FIXMAP_64_H
++
++#include <linux/kernel.h>
++#include <asm/page.h>
++#include <asm/vsyscall.h>
++#include <asm/efi.h>
++#include <asm/acpi.h>
++
++/*
++ * Here we define all the compile-time 'special' virtual
++ * addresses. The point is to have a constant address at
++ * compile time, but to set the physical address only
++ * in the boot process.
++ *
++ * These 'compile-time allocated' memory buffers are
++ * fixed-size 4k pages (or larger if used with an increment
++ * higher than 1). Use set_fixmap(idx,phys) to associate
++ * physical memory with fixmap indices.
++ *
++ * TLB entries of such buffers will not be flushed across
++ * task switches.
++ */
++
++enum fixed_addresses {
++ VSYSCALL_LAST_PAGE,
++ VSYSCALL_FIRST_PAGE = VSYSCALL_LAST_PAGE
++ + ((VSYSCALL_END-VSYSCALL_START) >> PAGE_SHIFT) - 1,
++ VSYSCALL_HPET,
++ FIX_DBGP_BASE,
++ FIX_EARLYCON_MEM_BASE,
++ FIX_HPET_BASE,
++#ifndef CONFIG_XEN
++ FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */
++ FIX_IO_APIC_BASE_0,
++ FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS - 1,
++#endif
++#ifdef CONFIG_EFI
++ FIX_EFI_IO_MAP_LAST_PAGE,
++ FIX_EFI_IO_MAP_FIRST_PAGE = FIX_EFI_IO_MAP_LAST_PAGE
++ + MAX_EFI_IO_PAGES - 1,
++#endif
++#ifdef CONFIG_ACPI
++ FIX_ACPI_BEGIN,
++ FIX_ACPI_END = FIX_ACPI_BEGIN + FIX_ACPI_PAGES - 1,
++#endif
++ FIX_SHARED_INFO,
++#define NR_FIX_ISAMAPS 256
++ FIX_ISAMAP_END,
++ FIX_ISAMAP_BEGIN = FIX_ISAMAP_END + NR_FIX_ISAMAPS - 1,
++ __end_of_permanent_fixed_addresses,
++ /*
++ * 256 temporary boot-time mappings, used by early_ioremap(),
++ * before ioremap() is functional.
++ *
++ * We round it up to the next 512 pages boundary so that we
++ * can have a single pgd entry and a single pte table:
++ */
++#define NR_FIX_BTMAPS 64
++#define FIX_BTMAPS_NESTING 4
++ FIX_BTMAP_END =
++ __end_of_permanent_fixed_addresses + 512 -
++ (__end_of_permanent_fixed_addresses & 511),
++ FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS*FIX_BTMAPS_NESTING - 1,
++#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT
++ FIX_OHCI1394_BASE,
++#endif
++ __end_of_fixed_addresses
++};
++
++extern void __set_fixmap(enum fixed_addresses idx,
++ unsigned long phys, pgprot_t flags);
++
++#define set_fixmap(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL)
++/*
++ * Some hardware wants to get fixmapped without caching.
++ */
++#define set_fixmap_nocache(idx, phys) \
++ __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE)
++
++#define FIXADDR_TOP (VSYSCALL_END-PAGE_SIZE)
++#define FIXADDR_SIZE (__end_of_fixed_addresses << PAGE_SHIFT)
++#define FIXADDR_START (FIXADDR_TOP - FIXADDR_SIZE)
++
++/* Only covers 32bit vsyscalls currently. Need another set for 64bit. */
++#define FIXADDR_USER_START ((unsigned long)VSYSCALL32_VSYSCALL)
++#define FIXADDR_USER_END (FIXADDR_USER_START + PAGE_SIZE)
++
++#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
++
++extern void __this_fixmap_does_not_exist(void);
++
++/*
++ * 'index to address' translation. If anyone tries to use the idx
++ * directly without translation, we catch the bug with a NULL-deference
++ * kernel oops. Illegal ranges of incoming indices are caught too.
++ */
++static __always_inline unsigned long fix_to_virt(const unsigned int idx)
++{
++ /*
++ * this branch gets completely eliminated after inlining,
++ * except when someone tries to use fixaddr indices in an
++ * illegal way. (such as mixing up address types or using
++ * out-of-range indices).
++ *
++ * If it doesn't get removed, the linker will complain
++ * loudly with a reasonably clear error message..
++ */
++ if (idx >= __end_of_fixed_addresses)
++ __this_fixmap_does_not_exist();
++
++ return __fix_to_virt(idx);
++}
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/gnttab_dma.h b/include/asm-x86/mach-xen/asm/gnttab_dma.h
+new file mode 100644
+index 0000000..fd7197c
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/gnttab_dma.h
+@@ -0,0 +1,41 @@
++/*
++ * Copyright (c) 2007 Herbert Xu <herbert at gondor.apana.org.au>
++ * Copyright (c) 2007 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ */
++
++#ifndef _ASM_I386_GNTTAB_DMA_H
++#define _ASM_I386_GNTTAB_DMA_H
++
++static inline int gnttab_dma_local_pfn(struct page *page)
++{
++ /* Has it become a local MFN? */
++ return pfn_valid(mfn_to_local_pfn(pfn_to_mfn(page_to_pfn(page))));
++}
++
++static inline maddr_t gnttab_dma_map_page(struct page *page)
++{
++ __gnttab_dma_map_page(page);
++ return ((maddr_t)pfn_to_mfn(page_to_pfn(page)) << PAGE_SHIFT);
++}
++
++static inline void gnttab_dma_unmap_page(maddr_t maddr)
++{
++ __gnttab_dma_unmap_page(virt_to_page(bus_to_virt(maddr)));
++}
++
++#endif /* _ASM_I386_GNTTAB_DMA_H */
+diff --git a/include/asm-x86/mach-xen/asm/highmem.h b/include/asm-x86/mach-xen/asm/highmem.h
+new file mode 100644
+index 0000000..628faa1
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/highmem.h
+@@ -0,0 +1,79 @@
++/*
++ * 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.
++ */
++/*
++ * 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 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 *kmap_high(struct page *page);
++extern void kunmap_high(struct page *page);
++
++void *kmap(struct page *page);
++void kunmap(struct page *page);
++void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot);
++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 kmap_atomic_pte(page, type) \
++ kmap_atomic_prot(page, type, \
++ PagePinned(page) ? PAGE_KERNEL_RO : kmap_prot)
++
++#define flush_cache_kmaps() do { } while (0)
++
++#endif /* __KERNEL__ */
++
++#endif /* _ASM_HIGHMEM_H */
+diff --git a/include/asm-x86/mach-xen/asm/hw_irq.h b/include/asm-x86/mach-xen/asm/hw_irq.h
+new file mode 100644
+index 0000000..bf02539
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hw_irq.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "hw_irq_32.h"
++#else
++# include "hw_irq_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/hw_irq_32.h b/include/asm-x86/mach-xen/asm/hw_irq_32.h
+new file mode 100644
+index 0000000..6df4a1e
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hw_irq_32.h
+@@ -0,0 +1,61 @@
++#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>
++
++/*
++ * 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 void (*const interrupt[NR_IRQS])(void);
++
++#ifdef CONFIG_SMP
++void reschedule_interrupt(void);
++void invalidate_interrupt(void);
++void call_function_interrupt(void);
++#endif
++
++#ifdef CONFIG_X86_LOCAL_APIC
++void apic_timer_interrupt(void);
++void error_interrupt(void);
++void spurious_interrupt(void);
++void thermal_interrupt(void);
++#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 send_IPI_self(int vector);
++void init_VISWS_APIC_irqs(void);
++void setup_IO_APIC(void);
++void disable_IO_APIC(void);
++#define print_IO_APIC()
++int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
++void send_IPI(int dest, int vector);
++void setup_ioapic_dest(void);
++
++extern unsigned long io_apic_irqs;
++
++#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
++
++#endif /* _ASM_HW_IRQ_H */
+diff --git a/include/asm-x86/mach-xen/asm/hw_irq_64.h b/include/asm-x86/mach-xen/asm/hw_irq_64.h
+new file mode 100644
+index 0000000..a95722e
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hw_irq_64.h
+@@ -0,0 +1,171 @@
++#ifndef _ASM_HW_IRQ_H
++#define _ASM_HW_IRQ_H
++
++/*
++ * linux/include/asm/hw_irq.h
++ *
++ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
++ *
++ * moved some of the old arch/i386/kernel/irq.h to here. VY
++ *
++ * IRQ/IPI changes taken from work by Thomas Radke
++ * <tomsoft at informatik.tu-chemnitz.de>
++ *
++ * hacked by Andi Kleen for x86-64.
++ */
++
++#ifndef __ASSEMBLY__
++#include <asm/atomic.h>
++#include <asm/irq.h>
++#include <linux/profile.h>
++#include <linux/smp.h>
++#include <linux/percpu.h>
++#endif
++
++/*
++ * IDT vectors usable for external interrupt sources start
++ * at 0x20:
++ */
++#define FIRST_EXTERNAL_VECTOR 0x20
++
++#define IA32_SYSCALL_VECTOR 0x80
++
++#ifndef CONFIG_XEN
++
++/* Reserve the lowest usable priority level 0x20 - 0x2f for triggering
++ * cleanup after irq migration.
++ */
++#define IRQ_MOVE_CLEANUP_VECTOR FIRST_EXTERNAL_VECTOR
++
++/*
++ * Vectors 0x30-0x3f are used for ISA interrupts.
++ */
++#define IRQ0_VECTOR (FIRST_EXTERNAL_VECTOR + 0x10)
++#define IRQ1_VECTOR (IRQ0_VECTOR + 1)
++#define IRQ2_VECTOR (IRQ0_VECTOR + 2)
++#define IRQ3_VECTOR (IRQ0_VECTOR + 3)
++#define IRQ4_VECTOR (IRQ0_VECTOR + 4)
++#define IRQ5_VECTOR (IRQ0_VECTOR + 5)
++#define IRQ6_VECTOR (IRQ0_VECTOR + 6)
++#define IRQ7_VECTOR (IRQ0_VECTOR + 7)
++#define IRQ8_VECTOR (IRQ0_VECTOR + 8)
++#define IRQ9_VECTOR (IRQ0_VECTOR + 9)
++#define IRQ10_VECTOR (IRQ0_VECTOR + 10)
++#define IRQ11_VECTOR (IRQ0_VECTOR + 11)
++#define IRQ12_VECTOR (IRQ0_VECTOR + 12)
++#define IRQ13_VECTOR (IRQ0_VECTOR + 13)
++#define IRQ14_VECTOR (IRQ0_VECTOR + 14)
++#define IRQ15_VECTOR (IRQ0_VECTOR + 15)
++
++/*
++ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
++ *
++ * some of the following vectors are 'rare', they are merged
++ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
++ * TLB, reschedule and local APIC vectors are performance-critical.
++ */
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
++#define RESCHEDULE_VECTOR 0xfd
++#define CALL_FUNCTION_VECTOR 0xfc
++/* fb free - please don't readd KDB here because it's useless
++ (hint - think what a NMI bit does to a vector) */
++#define THERMAL_APIC_VECTOR 0xfa
++#define THRESHOLD_APIC_VECTOR 0xf9
++/* f8 free */
++#define INVALIDATE_TLB_VECTOR_END 0xf7
++#define INVALIDATE_TLB_VECTOR_START 0xf0 /* f0-f7 used for TLB flush */
++
++#define NUM_INVALIDATE_TLB_VECTORS 8
++
++/*
++ * Local APIC timer IRQ vector is on a different priority level,
++ * to work around the 'lost local interrupt if more than 2 IRQ
++ * sources per level' errata.
++ */
++#define LOCAL_TIMER_VECTOR 0xef
++
++/*
++ * First APIC vector available to drivers: (vectors 0x30-0xee)
++ * we start at 0x41 to spread out vectors evenly between priority
++ * levels. (0x80 is the syscall vector)
++ */
++#define FIRST_DEVICE_VECTOR (IRQ15_VECTOR + 2)
++#define FIRST_SYSTEM_VECTOR 0xef /* duplicated in irq.h */
++
++#endif
++
++#ifndef __ASSEMBLY__
++
++/* Interrupt handlers registered during init_IRQ */
++void apic_timer_interrupt(void);
++void spurious_interrupt(void);
++void error_interrupt(void);
++void reschedule_interrupt(void);
++void call_function_interrupt(void);
++void irq_move_cleanup_interrupt(void);
++void invalidate_interrupt0(void);
++void invalidate_interrupt1(void);
++void invalidate_interrupt2(void);
++void invalidate_interrupt3(void);
++void invalidate_interrupt4(void);
++void invalidate_interrupt5(void);
++void invalidate_interrupt6(void);
++void invalidate_interrupt7(void);
++void thermal_interrupt(void);
++void threshold_interrupt(void);
++void i8254_timer_resume(void);
++
++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,
++ * time.c, io_apic.c and smp.c
++ *
++ * Interrupt entry/exit code at both C and assembly level
++ */
++
++extern void disable_8259A_irq(unsigned int irq);
++extern void enable_8259A_irq(unsigned int irq);
++extern int i8259A_irq_pending(unsigned int irq);
++extern void make_8259A_irq(unsigned int irq);
++extern void init_8259A(int aeoi);
++extern void send_IPI_self(int vector);
++extern void init_VISWS_APIC_irqs(void);
++extern void setup_IO_APIC(void);
++extern void enable_IO_APIC(void);
++extern void disable_IO_APIC(void);
++#define print_IO_APIC()
++extern int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn);
++extern void send_IPI(int dest, int vector);
++extern void setup_ioapic_dest(void);
++extern void native_init_IRQ(void);
++
++extern unsigned long io_apic_irqs;
++
++#define IO_APIC_IRQ(x) (((x) >= 16) || ((1<<(x)) & io_apic_irqs))
++
++#include <asm/ptrace.h>
++
++#define IRQ_NAME2(nr) nr##_interrupt(void)
++#define IRQ_NAME(nr) IRQ_NAME2(IRQ##nr)
++
++/*
++ * SMP has a few special interrupts for IPI messages
++ */
++
++#define BUILD_IRQ(nr) \
++ asmlinkage void IRQ_NAME(nr); \
++ asm("\n.p2align\n" \
++ "IRQ" #nr "_interrupt:\n\t" \
++ "push $~(" #nr ") ; " \
++ "jmp common_interrupt");
++
++#define platform_legacy_irq(irq) ((irq) < 16)
++
++#endif
++
++#endif /* _ASM_HW_IRQ_H */
+diff --git a/include/asm-x86/mach-xen/asm/hypercall.h b/include/asm-x86/mach-xen/asm/hypercall.h
+new file mode 100644
+index 0000000..c5fd1d7
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hypercall.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "hypercall_32.h"
++#else
++# include "hypercall_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/hypercall_32.h b/include/asm-x86/mach-xen/asm/hypercall_32.h
+new file mode 100644
+index 0000000..db53073
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hypercall_32.h
+@@ -0,0 +1,416 @@
++/******************************************************************************
++ * hypercall.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __HYPERCALL_H__
++#define __HYPERCALL_H__
++
++#include <linux/string.h> /* memcpy() */
++#include <linux/stringify.h>
++
++#ifndef __HYPERVISOR_H__
++# error "please don't include this file directly"
++#endif
++
++#ifdef CONFIG_XEN
++#define HYPERCALL_STR(name) \
++ "call hypercall_page + ("__stringify(__HYPERVISOR_##name)" * 32)"
++#else
++#define HYPERCALL_STR(name) \
++ "mov hypercall_stubs,%%eax; " \
++ "add $("__stringify(__HYPERVISOR_##name)" * 32),%%eax; "\
++ "call *%%eax"
++#endif
++
++#define _hypercall0(type, name) \
++({ \
++ type __res; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res) \
++ : \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall1(type, name, a1) \
++({ \
++ type __res; \
++ long __ign1; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1) \
++ : "1" ((long)(a1)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall2(type, name, a1, a2) \
++({ \
++ type __res; \
++ long __ign1, __ign2; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2) \
++ : "1" ((long)(a1)), "2" ((long)(a2)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall3(type, name, a1, a2, a3) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall4(type, name, a1, a2, a3, a4) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3, __ign4; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3), "=S" (__ign4) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)), "4" ((long)(a4)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3, __ign4, __ign5; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=b" (__ign1), "=c" (__ign2), \
++ "=d" (__ign3), "=S" (__ign4), "=D" (__ign5) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)), "4" ((long)(a4)), \
++ "5" ((long)(a5)) \
++ : "memory" ); \
++ __res; \
++})
++
++static inline int __must_check
++HYPERVISOR_set_trap_table(
++ const trap_info_t *table)
++{
++ return _hypercall1(int, set_trap_table, table);
++}
++
++static inline int __must_check
++HYPERVISOR_mmu_update(
++ mmu_update_t *req, unsigned int count, unsigned int *success_count,
++ domid_t domid)
++{
++ return _hypercall4(int, mmu_update, req, count, success_count, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_mmuext_op(
++ struct mmuext_op *op, unsigned int count, unsigned int *success_count,
++ domid_t domid)
++{
++ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_set_gdt(
++ unsigned long *frame_list, unsigned int entries)
++{
++ return _hypercall2(int, set_gdt, frame_list, entries);
++}
++
++static inline int __must_check
++HYPERVISOR_stack_switch(
++ unsigned long ss, unsigned long esp)
++{
++ return _hypercall2(int, stack_switch, ss, esp);
++}
++
++static inline int __must_check
++HYPERVISOR_set_callbacks(
++ unsigned long event_selector, unsigned long event_address,
++ unsigned long failsafe_selector, unsigned long failsafe_address)
++{
++ return _hypercall4(int, set_callbacks,
++ event_selector, event_address,
++ failsafe_selector, failsafe_address);
++}
++
++static inline int
++HYPERVISOR_fpu_taskswitch(
++ int set)
++{
++ return _hypercall1(int, fpu_taskswitch, set);
++}
++
++static inline int __must_check
++HYPERVISOR_sched_op_compat(
++ int cmd, unsigned long arg)
++{
++ return _hypercall2(int, sched_op_compat, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_sched_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, sched_op, cmd, arg);
++}
++
++static inline long __must_check
++HYPERVISOR_set_timer_op(
++ u64 timeout)
++{
++ unsigned long timeout_hi = (unsigned long)(timeout>>32);
++ unsigned long timeout_lo = (unsigned long)timeout;
++ return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
++}
++
++static inline int __must_check
++HYPERVISOR_platform_op(
++ struct xen_platform_op *platform_op)
++{
++ platform_op->interface_version = XENPF_INTERFACE_VERSION;
++ return _hypercall1(int, platform_op, platform_op);
++}
++
++static inline int __must_check
++HYPERVISOR_set_debugreg(
++ unsigned int reg, unsigned long value)
++{
++ return _hypercall2(int, set_debugreg, reg, value);
++}
++
++static inline unsigned long __must_check
++HYPERVISOR_get_debugreg(
++ unsigned int reg)
++{
++ return _hypercall1(unsigned long, get_debugreg, reg);
++}
++
++static inline int __must_check
++HYPERVISOR_update_descriptor(
++ u64 ma, u64 desc)
++{
++ return _hypercall4(int, update_descriptor, ma, ma>>32, desc, desc>>32);
++}
++
++static inline int __must_check
++HYPERVISOR_memory_op(
++ unsigned int cmd, void *arg)
++{
++ return _hypercall2(int, memory_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_multicall(
++ multicall_entry_t *call_list, unsigned int nr_calls)
++{
++ return _hypercall2(int, multicall, call_list, nr_calls);
++}
++
++static inline int __must_check
++HYPERVISOR_update_va_mapping(
++ unsigned long va, pte_t new_val, unsigned long flags)
++{
++ unsigned long pte_hi = 0;
++#ifdef CONFIG_X86_PAE
++ pte_hi = new_val.pte_high;
++#endif
++ return _hypercall4(int, update_va_mapping, va,
++ new_val.pte_low, pte_hi, flags);
++}
++
++static inline int __must_check
++HYPERVISOR_event_channel_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, event_channel_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct evtchn_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, event_channel_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
++
++ return rc;
++}
++
++static inline int __must_check
++HYPERVISOR_acm_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, acm_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_xen_version(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, xen_version, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_console_io(
++ int cmd, unsigned int count, char *str)
++{
++ return _hypercall3(int, console_io, cmd, count, str);
++}
++
++static inline int __must_check
++HYPERVISOR_physdev_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, physdev_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct physdev_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, physdev_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
++
++ return rc;
++}
++
++static inline int __must_check
++HYPERVISOR_grant_table_op(
++ unsigned int cmd, void *uop, unsigned int count)
++{
++ return _hypercall3(int, grant_table_op, cmd, uop, count);
++}
++
++static inline int __must_check
++HYPERVISOR_update_va_mapping_otherdomain(
++ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
++{
++ unsigned long pte_hi = 0;
++#ifdef CONFIG_X86_PAE
++ pte_hi = new_val.pte_high;
++#endif
++ return _hypercall5(int, update_va_mapping_otherdomain, va,
++ new_val.pte_low, pte_hi, flags, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_vm_assist(
++ unsigned int cmd, unsigned int type)
++{
++ return _hypercall2(int, vm_assist, cmd, type);
++}
++
++static inline int __must_check
++HYPERVISOR_vcpu_op(
++ int cmd, unsigned int vcpuid, void *extra_args)
++{
++ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
++}
++
++static inline int __must_check
++HYPERVISOR_suspend(
++ unsigned long srec)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = SHUTDOWN_suspend
++ };
++
++ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
++ &sched_shutdown, srec);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
++ SHUTDOWN_suspend, srec);
++#endif
++
++ return rc;
++}
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++static inline int
++HYPERVISOR_nmi_op(
++ unsigned long op, void *arg)
++{
++ return _hypercall2(int, nmi_op, op, arg);
++}
++#endif
++
++#ifndef CONFIG_XEN
++static inline unsigned long __must_check
++HYPERVISOR_hvm_op(
++ int op, void *arg)
++{
++ return _hypercall2(unsigned long, hvm_op, op, arg);
++}
++#endif
++
++static inline int __must_check
++HYPERVISOR_callback_op(
++ int cmd, const void *arg)
++{
++ return _hypercall2(int, callback_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_xenoprof_op(
++ int op, void *arg)
++{
++ return _hypercall2(int, xenoprof_op, op, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_kexec_op(
++ unsigned long op, void *args)
++{
++ return _hypercall2(int, kexec_op, op, args);
++}
++
++
++
++#endif /* __HYPERCALL_H__ */
+diff --git a/include/asm-x86/mach-xen/asm/hypercall_64.h b/include/asm-x86/mach-xen/asm/hypercall_64.h
+new file mode 100644
+index 0000000..651b449
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hypercall_64.h
+@@ -0,0 +1,415 @@
++/******************************************************************************
++ * hypercall.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * 64-bit updates:
++ * Benjamin Liu <benjamin.liu at intel.com>
++ * Jun Nakajima <jun.nakajima at intel.com>
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __HYPERCALL_H__
++#define __HYPERCALL_H__
++
++#include <linux/string.h> /* memcpy() */
++#include <linux/stringify.h>
++
++#ifndef __HYPERVISOR_H__
++# error "please don't include this file directly"
++#endif
++
++#ifdef CONFIG_XEN
++#define HYPERCALL_STR(name) \
++ "call hypercall_page + ("__stringify(__HYPERVISOR_##name)" * 32)"
++#else
++#define HYPERCALL_STR(name) \
++ "mov $("__stringify(__HYPERVISOR_##name)" * 32),%%eax; "\
++ "add hypercall_stubs(%%rip),%%rax; " \
++ "call *%%rax"
++#endif
++
++#define _hypercall0(type, name) \
++({ \
++ type __res; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res) \
++ : \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall1(type, name, a1) \
++({ \
++ type __res; \
++ long __ign1; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1) \
++ : "1" ((long)(a1)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall2(type, name, a1, a2) \
++({ \
++ type __res; \
++ long __ign1, __ign2; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2) \
++ : "1" ((long)(a1)), "2" ((long)(a2)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall3(type, name, a1, a2, a3) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall4(type, name, a1, a2, a3, a4) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ register long __arg4 asm("r10") = (long)(a4); \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3), "+r" (__arg4) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
++})
++
++#define _hypercall5(type, name, a1, a2, a3, a4, a5) \
++({ \
++ type __res; \
++ long __ign1, __ign2, __ign3; \
++ register long __arg4 asm("r10") = (long)(a4); \
++ register long __arg5 asm("r8") = (long)(a5); \
++ asm volatile ( \
++ HYPERCALL_STR(name) \
++ : "=a" (__res), "=D" (__ign1), "=S" (__ign2), \
++ "=d" (__ign3), "+r" (__arg4), "+r" (__arg5) \
++ : "1" ((long)(a1)), "2" ((long)(a2)), \
++ "3" ((long)(a3)) \
++ : "memory" ); \
++ __res; \
++})
++
++static inline int __must_check
++HYPERVISOR_set_trap_table(
++ const trap_info_t *table)
++{
++ return _hypercall1(int, set_trap_table, table);
++}
++
++static inline int __must_check
++HYPERVISOR_mmu_update(
++ mmu_update_t *req, unsigned int count, unsigned int *success_count,
++ domid_t domid)
++{
++ return _hypercall4(int, mmu_update, req, count, success_count, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_mmuext_op(
++ struct mmuext_op *op, unsigned int count, unsigned int *success_count,
++ domid_t domid)
++{
++ return _hypercall4(int, mmuext_op, op, count, success_count, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_set_gdt(
++ unsigned long *frame_list, unsigned int entries)
++{
++ return _hypercall2(int, set_gdt, frame_list, entries);
++}
++
++static inline int __must_check
++HYPERVISOR_stack_switch(
++ unsigned long ss, unsigned long esp)
++{
++ return _hypercall2(int, stack_switch, ss, esp);
++}
++
++static inline int __must_check
++HYPERVISOR_set_callbacks(
++ unsigned long event_address, unsigned long failsafe_address,
++ unsigned long syscall_address)
++{
++ return _hypercall3(int, set_callbacks,
++ event_address, failsafe_address, syscall_address);
++}
++
++static inline int
++HYPERVISOR_fpu_taskswitch(
++ int set)
++{
++ return _hypercall1(int, fpu_taskswitch, set);
++}
++
++static inline int __must_check
++HYPERVISOR_sched_op_compat(
++ int cmd, unsigned long arg)
++{
++ return _hypercall2(int, sched_op_compat, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_sched_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, sched_op, cmd, arg);
++}
++
++static inline long __must_check
++HYPERVISOR_set_timer_op(
++ u64 timeout)
++{
++ return _hypercall1(long, set_timer_op, timeout);
++}
++
++static inline int __must_check
++HYPERVISOR_platform_op(
++ struct xen_platform_op *platform_op)
++{
++ platform_op->interface_version = XENPF_INTERFACE_VERSION;
++ return _hypercall1(int, platform_op, platform_op);
++}
++
++static inline int __must_check
++HYPERVISOR_set_debugreg(
++ unsigned int reg, unsigned long value)
++{
++ return _hypercall2(int, set_debugreg, reg, value);
++}
++
++static inline unsigned long __must_check
++HYPERVISOR_get_debugreg(
++ unsigned int reg)
++{
++ return _hypercall1(unsigned long, get_debugreg, reg);
++}
++
++static inline int __must_check
++HYPERVISOR_update_descriptor(
++ unsigned long ma, unsigned long word)
++{
++ return _hypercall2(int, update_descriptor, ma, word);
++}
++
++static inline int __must_check
++HYPERVISOR_memory_op(
++ unsigned int cmd, void *arg)
++{
++ return _hypercall2(int, memory_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_multicall(
++ multicall_entry_t *call_list, unsigned int nr_calls)
++{
++ return _hypercall2(int, multicall, call_list, nr_calls);
++}
++
++static inline int __must_check
++HYPERVISOR_update_va_mapping(
++ unsigned long va, pte_t new_val, unsigned long flags)
++{
++ return _hypercall3(int, update_va_mapping, va, new_val.pte, flags);
++}
++
++static inline int __must_check
++HYPERVISOR_event_channel_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, event_channel_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct evtchn_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, event_channel_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
++
++ return rc;
++}
++
++static inline int __must_check
++HYPERVISOR_acm_op(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, acm_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_xen_version(
++ int cmd, void *arg)
++{
++ return _hypercall2(int, xen_version, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_console_io(
++ int cmd, unsigned int count, char *str)
++{
++ return _hypercall3(int, console_io, cmd, count, str);
++}
++
++static inline int __must_check
++HYPERVISOR_physdev_op(
++ int cmd, void *arg)
++{
++ int rc = _hypercall2(int, physdev_op, cmd, arg);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (unlikely(rc == -ENOSYS)) {
++ struct physdev_op op;
++ op.cmd = cmd;
++ memcpy(&op.u, arg, sizeof(op.u));
++ rc = _hypercall1(int, physdev_op_compat, &op);
++ memcpy(arg, &op.u, sizeof(op.u));
++ }
++#endif
++
++ return rc;
++}
++
++static inline int __must_check
++HYPERVISOR_grant_table_op(
++ unsigned int cmd, void *uop, unsigned int count)
++{
++ return _hypercall3(int, grant_table_op, cmd, uop, count);
++}
++
++static inline int __must_check
++HYPERVISOR_update_va_mapping_otherdomain(
++ unsigned long va, pte_t new_val, unsigned long flags, domid_t domid)
++{
++ return _hypercall4(int, update_va_mapping_otherdomain, va,
++ new_val.pte, flags, domid);
++}
++
++static inline int __must_check
++HYPERVISOR_vm_assist(
++ unsigned int cmd, unsigned int type)
++{
++ return _hypercall2(int, vm_assist, cmd, type);
++}
++
++static inline int __must_check
++HYPERVISOR_vcpu_op(
++ int cmd, unsigned int vcpuid, void *extra_args)
++{
++ return _hypercall3(int, vcpu_op, cmd, vcpuid, extra_args);
++}
++
++static inline int __must_check
++HYPERVISOR_set_segment_base(
++ int reg, unsigned long value)
++{
++ return _hypercall2(int, set_segment_base, reg, value);
++}
++
++static inline int __must_check
++HYPERVISOR_suspend(
++ unsigned long srec)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = SHUTDOWN_suspend
++ };
++
++ int rc = _hypercall3(int, sched_op, SCHEDOP_shutdown,
++ &sched_shutdown, srec);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = _hypercall3(int, sched_op_compat, SCHEDOP_shutdown,
++ SHUTDOWN_suspend, srec);
++#endif
++
++ return rc;
++}
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++static inline int
++HYPERVISOR_nmi_op(
++ unsigned long op, void *arg)
++{
++ return _hypercall2(int, nmi_op, op, arg);
++}
++#endif
++
++#ifndef CONFIG_XEN
++static inline unsigned long __must_check
++HYPERVISOR_hvm_op(
++ int op, void *arg)
++{
++ return _hypercall2(unsigned long, hvm_op, op, arg);
++}
++#endif
++
++static inline int __must_check
++HYPERVISOR_callback_op(
++ int cmd, const void *arg)
++{
++ return _hypercall2(int, callback_op, cmd, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_xenoprof_op(
++ int op, void *arg)
++{
++ return _hypercall2(int, xenoprof_op, op, arg);
++}
++
++static inline int __must_check
++HYPERVISOR_kexec_op(
++ unsigned long op, void *args)
++{
++ return _hypercall2(int, kexec_op, op, args);
++}
++
++#endif /* __HYPERCALL_H__ */
+diff --git a/include/asm-x86/mach-xen/asm/hypervisor.h b/include/asm-x86/mach-xen/asm/hypervisor.h
+new file mode 100644
+index 0000000..bfe1379
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/hypervisor.h
+@@ -0,0 +1,250 @@
++/******************************************************************************
++ * hypervisor.h
++ *
++ * Linux-specific hypervisor handling.
++ *
++ * Copyright (c) 2002-2004, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __HYPERVISOR_H__
++#define __HYPERVISOR_H__
++
++#include <linux/types.h>
++#include <linux/kernel.h>
++#include <linux/version.h>
++#include <linux/errno.h>
++#include <xen/interface/xen.h>
++#include <xen/interface/platform.h>
++#include <xen/interface/event_channel.h>
++#include <xen/interface/physdev.h>
++#include <xen/interface/sched.h>
++#include <xen/interface/nmi.h>
++#include <asm/ptrace.h>
++#include <asm/page.h>
++
++extern shared_info_t *HYPERVISOR_shared_info;
++
++#define vcpu_info(cpu) (HYPERVISOR_shared_info->vcpu_info + (cpu))
++#ifdef CONFIG_SMP
++#define current_vcpu_info() vcpu_info(smp_processor_id())
++#else
++#define current_vcpu_info() vcpu_info(0)
++#endif
++
++#ifdef CONFIG_X86_32
++extern unsigned long hypervisor_virt_start;
++#endif
++
++/* arch/xen/i386/kernel/setup.c */
++extern start_info_t *xen_start_info;
++#ifdef CONFIG_XEN_PRIVILEGED_GUEST
++#define is_initial_xendomain() (xen_start_info->flags & SIF_INITDOMAIN)
++#else
++#define is_initial_xendomain() 0
++#endif
++
++/* arch/xen/kernel/evtchn.c */
++/* Force a proper event-channel callback from Xen. */
++void force_evtchn_callback(void);
++
++/* arch/xen/kernel/process.c */
++void xen_cpu_idle (void);
++
++/* arch/xen/i386/kernel/hypervisor.c */
++void do_hypervisor_callback(struct pt_regs *regs);
++
++/* arch/xen/i386/mm/hypervisor.c */
++/*
++ * NB. ptr values should be PHYSICAL, not MACHINE. 'vals' should be already
++ * be MACHINE addresses.
++ */
++
++void xen_pt_switch(unsigned long ptr);
++void xen_new_user_pt(unsigned long ptr); /* x86_64 only */
++void xen_load_gs(unsigned int selector); /* x86_64 only */
++void xen_tlb_flush(void);
++void xen_invlpg(unsigned long ptr);
++
++void xen_l1_entry_update(pte_t *ptr, pte_t val);
++void xen_l2_entry_update(pmd_t *ptr, pmd_t val);
++void xen_l3_entry_update(pud_t *ptr, pud_t val); /* x86_64/PAE */
++void xen_l4_entry_update(pgd_t *ptr, pgd_t val); /* x86_64 only */
++void xen_pgd_pin(unsigned long ptr);
++void xen_pgd_unpin(unsigned long ptr);
++
++void xen_set_ldt(const void *ptr, unsigned int ents);
++
++#ifdef CONFIG_SMP
++#include <linux/cpumask.h>
++void xen_tlb_flush_all(void);
++void xen_invlpg_all(unsigned long ptr);
++void xen_tlb_flush_mask(cpumask_t *mask);
++void xen_invlpg_mask(cpumask_t *mask, unsigned long ptr);
++#endif
++
++/* Returns zero on success else negative errno. */
++int xen_create_contiguous_region(
++ unsigned long vstart, unsigned int order, unsigned int address_bits);
++void xen_destroy_contiguous_region(
++ unsigned long vstart, unsigned int order);
++
++struct page;
++
++int xen_limit_pages_to_max_mfn(
++ struct page *pages, unsigned int order, unsigned int address_bits);
++
++/* Turn jiffies into Xen system time. */
++u64 jiffies_to_st(unsigned long jiffies);
++
++#ifdef CONFIG_XEN_SCRUB_PAGES
++void scrub_pages(void *, unsigned int);
++#else
++#define scrub_pages(_p,_n) ((void)0)
++#endif
++
++#include <xen/hypercall.h>
++
++#if defined(CONFIG_X86_64)
++#define MULTI_UVMFLAGS_INDEX 2
++#define MULTI_UVMDOMID_INDEX 3
++#else
++#define MULTI_UVMFLAGS_INDEX 3
++#define MULTI_UVMDOMID_INDEX 4
++#endif
++
++#ifdef CONFIG_XEN
++#define is_running_on_xen() 1
++#else
++extern char *hypercall_stubs;
++#define is_running_on_xen() (!!hypercall_stubs)
++#endif
++
++static inline int
++HYPERVISOR_yield(
++ void)
++{
++ int rc = HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
++#endif
++
++ return rc;
++}
++
++static inline int
++HYPERVISOR_block(
++ void)
++{
++ int rc = HYPERVISOR_sched_op(SCHEDOP_block, NULL);
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_block, 0);
++#endif
++
++ return rc;
++}
++
++static inline void __noreturn
++HYPERVISOR_shutdown(
++ unsigned int reason)
++{
++ struct sched_shutdown sched_shutdown = {
++ .reason = reason
++ };
++
++ VOID(HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown));
++#if CONFIG_XEN_COMPAT <= 0x030002
++ VOID(HYPERVISOR_sched_op_compat(SCHEDOP_shutdown, reason));
++#endif
++ /* Don't recurse needlessly. */
++ BUG_ON(reason != SHUTDOWN_crash);
++ for(;;);
++}
++
++static inline int __must_check
++HYPERVISOR_poll(
++ evtchn_port_t *ports, unsigned int nr_ports, u64 timeout)
++{
++ int rc;
++ struct sched_poll sched_poll = {
++ .nr_ports = nr_ports,
++ .timeout = jiffies_to_st(timeout)
++ };
++ set_xen_guest_handle(sched_poll.ports, ports);
++
++ rc = HYPERVISOR_sched_op(SCHEDOP_poll, &sched_poll);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (rc == -ENOSYS)
++ rc = HYPERVISOR_sched_op_compat(SCHEDOP_yield, 0);
++#endif
++
++ return rc;
++}
++
++#ifdef CONFIG_XEN
++
++static inline void
++MULTI_update_va_mapping(
++ multicall_entry_t *mcl, unsigned long va,
++ pte_t new_val, unsigned long flags)
++{
++ mcl->op = __HYPERVISOR_update_va_mapping;
++ mcl->args[0] = va;
++#if defined(CONFIG_X86_64)
++ mcl->args[1] = new_val.pte;
++#elif defined(CONFIG_X86_PAE)
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = new_val.pte_high;
++#else
++ mcl->args[1] = new_val.pte_low;
++ mcl->args[2] = 0;
++#endif
++ mcl->args[MULTI_UVMFLAGS_INDEX] = flags;
++}
++
++static inline void
++MULTI_grant_table_op(multicall_entry_t *mcl, unsigned int cmd,
++ void *uop, unsigned int count)
++{
++ mcl->op = __HYPERVISOR_grant_table_op;
++ mcl->args[0] = cmd;
++ mcl->args[1] = (unsigned long)uop;
++ mcl->args[2] = count;
++}
++
++#else /* !defined(CONFIG_XEN) */
++
++/* Multicalls not supported for HVM guests. */
++#define MULTI_update_va_mapping(a,b,c,d) ((void)0)
++#define MULTI_grant_table_op(a,b,c,d) ((void)0)
++
++#endif
++
++#endif /* __HYPERVISOR_H__ */
+diff --git a/include/asm-x86/mach-xen/asm/io.h b/include/asm-x86/mach-xen/asm/io.h
+new file mode 100644
+index 0000000..f0ff0c1
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/io.h
+@@ -0,0 +1,22 @@
++#ifndef _ASM_X86_IO_H
++#define _ASM_X86_IO_H
++
++#define ARCH_HAS_IOREMAP_WC
++
++#ifdef CONFIG_X86_32
++# include "io_32.h"
++#else
++# include "io_64.h"
++#endif
++
++extern void *xlate_dev_mem_ptr(unsigned long phys);
++extern void unxlate_dev_mem_ptr(unsigned long phys, void *addr);
++
++extern void map_devmem(unsigned long pfn, unsigned long len, pgprot_t);
++extern void unmap_devmem(unsigned long pfn, unsigned long len, pgprot_t);
++
++extern int ioremap_change_attr(unsigned long vaddr, unsigned long size,
++ unsigned long prot_val);
++extern void __iomem *ioremap_wc(unsigned long offset, unsigned long size);
++
++#endif /* _ASM_X86_IO_H */
+diff --git a/include/asm-x86/mach-xen/asm/io_32.h b/include/asm-x86/mach-xen/asm/io_32.h
+new file mode 100644
+index 0000000..d86f211
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/io_32.h
+@@ -0,0 +1,354 @@
++#ifndef _ASM_IO_H
++#define _ASM_IO_H
++
++#include <linux/string.h>
++#include <linux/compiler.h>
++
++/*
++ * This file contains the definitions for the x86 IO instructions
++ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
++ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
++ * versions of the single-IO instructions (inb_p/inw_p/..).
++ *
++ * This file is not meant to be obfuscating: it's just complicated
++ * to (a) handle it all in a way that makes gcc able to optimize it
++ * as well as possible and (b) trying to avoid writing the same thing
++ * over and over again with slight variations and possibly making a
++ * mistake somewhere.
++ */
++
++/*
++ * Thanks to James van Artsdalen for a better timing-fix than
++ * the two short jumps: using outb's to a nonexistent port seems
++ * to guarantee better timings even on fast machines.
++ *
++ * On the other hand, I'd like to be sure of a non-existent port:
++ * I feel a bit unsafe about using 0x80 (should be safe, though)
++ *
++ * Linus
++ */
++
++ /*
++ * Bit simplified and optimized by Jan Hubicka
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
++ *
++ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
++ * isa_read[wl] and isa_write[wl] fixed
++ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
++ */
++
++#define IO_SPACE_LIMIT 0xffff
++
++#define XQUAD_PORTIO_BASE 0xfe400000
++#define XQUAD_PORTIO_QUAD 0x40000 /* 256k per quad. */
++
++#ifdef __KERNEL__
++
++#include <asm-generic/iomap.h>
++
++#include <linux/vmalloc.h>
++#include <asm/fixmap.h>
++
++/*
++ * Convert a virtual cached pointer to an uncached pointer
++ */
++#define xlate_dev_kmem_ptr(p) p
++
++/**
++ * virt_to_phys - map virtual addresses to physical
++ * @address: address to remap
++ *
++ * The returned physical address is the physical (CPU) mapping for
++ * the memory address given. It is only valid to use this function on
++ * addresses directly mapped or allocated via kmalloc.
++ *
++ * This function does not give bus mappings for DMA transfers. In
++ * almost all conceivable cases a device driver should not be using
++ * this function
++ */
++
++static inline unsigned long virt_to_phys(volatile void *address)
++{
++ return __pa(address);
++}
++
++/**
++ * phys_to_virt - map physical address to virtual
++ * @address: address to remap
++ *
++ * The returned virtual address is a current CPU mapping for
++ * the memory address given. It is only valid to use this function on
++ * addresses that have a kernel mapping
++ *
++ * This function does not handle bus mappings for DMA transfers. In
++ * almost all conceivable cases a device driver should not be using
++ * this function
++ */
++
++static inline void *phys_to_virt(unsigned long address)
++{
++ return __va(address);
++}
++
++/*
++ * Change "struct page" to physical address.
++ */
++#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
++#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
++#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
++
++#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
++ (unsigned long) bio_offset((bio)))
++#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
++
++#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
++ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
++ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
++ bvec_to_pseudophys((vec2))))
++
++/**
++ * ioremap - map bus memory into CPU space
++ * @offset: bus address of the memory
++ * @size: size of the resource to map
++ *
++ * ioremap performs a platform specific sequence of operations to
++ * make bus memory CPU accessible via the readb/readw/readl/writeb/
++ * writew/writel functions and the other mmio helpers. The returned
++ * address is not guaranteed to be usable directly as a virtual
++ * address.
++ *
++ * If the area you are trying to map is a PCI BAR you should have a
++ * look at pci_iomap().
++ */
++extern void __iomem *ioremap_nocache(resource_size_t offset, unsigned long size);
++extern void __iomem *ioremap_cache(resource_size_t offset, unsigned long size);
++
++/*
++ * The default ioremap() behavior is non-cached:
++ */
++static inline void __iomem *ioremap(resource_size_t offset, unsigned long size)
++{
++ return ioremap_nocache(offset, size);
++}
++
++extern void iounmap(volatile void __iomem *addr);
++
++/*
++ * early_ioremap() and early_iounmap() are for temporary early boot-time
++ * mappings, before the real ioremap() is functional.
++ * A boot-time mapping is currently limited to at most 16 pages.
++ */
++extern void early_ioremap_init(void);
++extern void early_ioremap_clear(void);
++extern void early_ioremap_reset(void);
++extern void *early_ioremap(unsigned long offset, unsigned long size);
++extern void early_iounmap(void *addr, unsigned long size);
++extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys);
++
++/*
++ * ISA I/O bus memory addresses are 1:1 with the physical address.
++ */
++#define isa_virt_to_bus(_x) ({ BUG(); virt_to_bus(_x); })
++#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
++#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
++
++/*
++ * However PCI ones are not necessarily 1:1 and therefore these interfaces
++ * are forbidden in portable PCI drivers.
++ *
++ * Allow them on x86 for legacy drivers, though.
++ */
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
++
++/*
++ * readX/writeX() are used to access memory mapped devices. On some
++ * architectures the memory mapped IO stuff needs to be accessed
++ * differently. On the x86 architecture, we just read/write the
++ * memory location directly.
++ */
++
++static inline unsigned char readb(const volatile void __iomem *addr)
++{
++ return *(volatile unsigned char __force *)addr;
++}
++
++static inline unsigned short readw(const volatile void __iomem *addr)
++{
++ return *(volatile unsigned short __force *)addr;
++}
++
++static inline unsigned int readl(const volatile void __iomem *addr)
++{
++ return *(volatile unsigned int __force *) addr;
++}
++
++#define readb_relaxed(addr) readb(addr)
++#define readw_relaxed(addr) readw(addr)
++#define readl_relaxed(addr) readl(addr)
++#define __raw_readb readb
++#define __raw_readw readw
++#define __raw_readl readl
++
++static inline void writeb(unsigned char b, volatile void __iomem *addr)
++{
++ *(volatile unsigned char __force *)addr = b;
++}
++
++static inline void writew(unsigned short b, volatile void __iomem *addr)
++{
++ *(volatile unsigned short __force *)addr = b;
++}
++
++static inline void writel(unsigned int b, volatile void __iomem *addr)
++{
++ *(volatile unsigned int __force *)addr = b;
++}
++#define __raw_writeb writeb
++#define __raw_writew writew
++#define __raw_writel writel
++
++#define mmiowb()
++
++static inline void
++memset_io(volatile void __iomem *addr, unsigned char val, int count)
++{
++ memset((void __force *)addr, val, count);
++}
++
++static inline void
++memcpy_fromio(void *dst, const volatile void __iomem *src, int count)
++{
++ __memcpy(dst, (const void __force *)src, count);
++}
++
++static inline void
++memcpy_toio(volatile void __iomem *dst, const void *src, int count)
++{
++ __memcpy((void __force *)dst, src, count);
++}
++
++/*
++ * ISA space is 'always mapped' on a typical x86 system, no need to
++ * explicitly ioremap() it. The fact that the ISA IO space is mapped
++ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
++ * are physical addresses. The following constant pointer can be
++ * used as the IO-area pointer (it can be iounmapped as well, so the
++ * analogy with PCI is quite large):
++ */
++#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
++
++/*
++ * Cache management
++ *
++ * This needed for two cases
++ * 1. Out of order aware processors
++ * 2. Accidentally out of order processors (PPro errata #51)
++ */
++
++#if defined(CONFIG_X86_OOSTORE) || defined(CONFIG_X86_PPRO_FENCE)
++
++static inline void flush_write_buffers(void)
++{
++ asm volatile("lock; addl $0,0(%%esp)": : :"memory");
++}
++
++#else
++
++#define flush_write_buffers() do { } while (0)
++
++#endif
++
++#endif /* __KERNEL__ */
++
++extern void xen_io_delay(void);
++#define native_io_delay xen_io_delay
++
++extern int io_delay_type;
++extern void io_delay_init(void);
++
++static inline void slow_down_io(void)
++{
++ native_io_delay();
++#ifdef REALLY_SLOW_IO
++ native_io_delay();
++ native_io_delay();
++ native_io_delay();
++#endif
++}
++
++#define __BUILDIO(bwl, bw, type) \
++static inline void out##bwl(unsigned type value, int port) \
++{ \
++ out##bwl##_local(value, port); \
++} \
++ \
++static inline unsigned type in##bwl(int port) \
++{ \
++ return in##bwl##_local(port); \
++}
++
++#define BUILDIO(bwl, bw, type) \
++static inline void out##bwl##_local(unsigned type value, int port) \
++{ \
++ asm volatile("out" #bwl " %" #bw "0, %w1" \
++ : : "a"(value), "Nd"(port)); \
++} \
++ \
++static inline unsigned type in##bwl##_local(int port) \
++{ \
++ unsigned type value; \
++ asm volatile("in" #bwl " %w1, %" #bw "0" \
++ : "=a"(value) : "Nd"(port)); \
++ return value; \
++} \
++ \
++static inline void out##bwl##_local_p(unsigned type value, int port) \
++{ \
++ out##bwl##_local(value, port); \
++ slow_down_io(); \
++} \
++ \
++static inline unsigned type in##bwl##_local_p(int port) \
++{ \
++ unsigned type value = in##bwl##_local(port); \
++ slow_down_io(); \
++ return value; \
++} \
++ \
++__BUILDIO(bwl, bw, type) \
++ \
++static inline void out##bwl##_p(unsigned type value, int port) \
++{ \
++ out##bwl(value, port); \
++ slow_down_io(); \
++} \
++ \
++static inline unsigned type in##bwl##_p(int port) \
++{ \
++ unsigned type value = in##bwl(port); \
++ slow_down_io(); \
++ return value; \
++} \
++ \
++static inline void outs##bwl(int port, const void *addr, unsigned long count) \
++{ \
++ asm volatile("rep; outs" #bwl \
++ : "+S"(addr), "+c"(count) : "d"(port)); \
++} \
++ \
++static inline void ins##bwl(int port, void *addr, unsigned long count) \
++{ \
++ asm volatile("rep; ins" #bwl \
++ : "+D"(addr), "+c"(count) : "d"(port)); \
++}
++
++BUILDIO(b, b, char)
++BUILDIO(w, w, short)
++BUILDIO(l, , int)
++
++/* We will be supplying our own /dev/mem implementation */
++#define ARCH_HAS_DEV_MEM
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/io_64.h b/include/asm-x86/mach-xen/asm/io_64.h
+new file mode 100644
+index 0000000..cf1e1d1
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/io_64.h
+@@ -0,0 +1,340 @@
++#ifndef _ASM_IO_H
++#define _ASM_IO_H
++
++#include <asm/fixmap.h>
++
++/*
++ * This file contains the definitions for the x86 IO instructions
++ * inb/inw/inl/outb/outw/outl and the "string versions" of the same
++ * (insb/insw/insl/outsb/outsw/outsl). You can also use "pausing"
++ * versions of the single-IO instructions (inb_p/inw_p/..).
++ *
++ * This file is not meant to be obfuscating: it's just complicated
++ * to (a) handle it all in a way that makes gcc able to optimize it
++ * as well as possible and (b) trying to avoid writing the same thing
++ * over and over again with slight variations and possibly making a
++ * mistake somewhere.
++ */
++
++/*
++ * Thanks to James van Artsdalen for a better timing-fix than
++ * the two short jumps: using outb's to a nonexistent port seems
++ * to guarantee better timings even on fast machines.
++ *
++ * On the other hand, I'd like to be sure of a non-existent port:
++ * I feel a bit unsafe about using 0x80 (should be safe, though)
++ *
++ * Linus
++ */
++
++ /*
++ * Bit simplified and optimized by Jan Hubicka
++ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999.
++ *
++ * isa_memset_io, isa_memcpy_fromio, isa_memcpy_toio added,
++ * isa_read[wl] and isa_write[wl] fixed
++ * - Arnaldo Carvalho de Melo <acme at conectiva.com.br>
++ */
++
++extern void xen_io_delay(void);
++#define native_io_delay xen_io_delay
++
++extern int io_delay_type;
++extern void io_delay_init(void);
++
++#if defined(CONFIG_PARAVIRT)
++#include <asm/paravirt.h>
++#else
++
++static inline void slow_down_io(void)
++{
++ native_io_delay();
++#ifdef REALLY_SLOW_IO
++ native_io_delay();
++ native_io_delay();
++ native_io_delay();
++#endif
++}
++#endif
++
++/*
++ * Talk about misusing macros..
++ */
++#define __OUT1(s, x) \
++static inline void out##s(unsigned x value, unsigned short port) {
++
++#define __OUT2(s, s1, s2) \
++asm volatile ("out" #s " %" s1 "0,%" s2 "1"
++
++#ifndef REALLY_SLOW_IO
++#define REALLY_SLOW_IO
++#define UNSET_REALLY_SLOW_IO
++#endif
++
++#define __OUT(s, s1, x) \
++ __OUT1(s, x) __OUT2(s, s1, "w") : : "a" (value), "Nd" (port)); \
++ } \
++ __OUT1(s##_p, x) __OUT2(s, s1, "w") : : "a" (value), "Nd" (port)); \
++ slow_down_io(); \
++}
++
++#define __IN1(s) \
++static inline RETURN_TYPE in##s(unsigned short port) \
++{ \
++ RETURN_TYPE _v;
++
++#define __IN2(s, s1, s2) \
++ asm volatile ("in" #s " %" s2 "1,%" s1 "0"
++
++#define __IN(s, s1, i...) \
++ __IN1(s) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); \
++ return _v; \
++ } \
++ __IN1(s##_p) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); \
++ slow_down_io(); \
++ return _v; }
++
++#ifdef UNSET_REALLY_SLOW_IO
++#undef REALLY_SLOW_IO
++#endif
++
++#define __INS(s) \
++static inline void ins##s(unsigned short port, void *addr, \
++ unsigned long count) \
++{ \
++ asm volatile ("rep ; ins" #s \
++ : "=D" (addr), "=c" (count) \
++ : "d" (port), "0" (addr), "1" (count)); \
++}
++
++#define __OUTS(s) \
++static inline void outs##s(unsigned short port, const void *addr, \
++ unsigned long count) \
++{ \
++ asm volatile ("rep ; outs" #s \
++ : "=S" (addr), "=c" (count) \
++ : "d" (port), "0" (addr), "1" (count)); \
++}
++
++#define RETURN_TYPE unsigned char
++__IN(b, "")
++#undef RETURN_TYPE
++#define RETURN_TYPE unsigned short
++__IN(w, "")
++#undef RETURN_TYPE
++#define RETURN_TYPE unsigned int
++__IN(l, "")
++#undef RETURN_TYPE
++
++__OUT(b, "b", char)
++__OUT(w, "w", short)
++__OUT(l, , int)
++
++__INS(b)
++__INS(w)
++__INS(l)
++
++__OUTS(b)
++__OUTS(w)
++__OUTS(l)
++
++#define IO_SPACE_LIMIT 0xffff
++
++#if defined(__KERNEL__) && defined(__x86_64__)
++
++#include <linux/vmalloc.h>
++
++#ifndef __i386__
++/*
++ * Change virtual addresses to physical addresses and vv.
++ * These are pretty trivial
++ */
++static inline unsigned long virt_to_phys(volatile void *address)
++{
++ return __pa(address);
++}
++
++static inline void *phys_to_virt(unsigned long address)
++{
++ return __va(address);
++}
++
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
++#endif
++
++/*
++ * Change "struct page" to physical address.
++ */
++#define page_to_pseudophys(page) ((dma_addr_t)page_to_pfn(page) << PAGE_SHIFT)
++#define page_to_phys(page) (phys_to_machine(page_to_pseudophys(page)))
++#define page_to_bus(page) (phys_to_machine(page_to_pseudophys(page)))
++
++#define bio_to_pseudophys(bio) (page_to_pseudophys(bio_page((bio))) + \
++ (unsigned long) bio_offset((bio)))
++#define bvec_to_pseudophys(bv) (page_to_pseudophys((bv)->bv_page) + \
++ (unsigned long) (bv)->bv_offset)
++
++#define BIOVEC_PHYS_MERGEABLE(vec1, vec2) \
++ (((bvec_to_phys((vec1)) + (vec1)->bv_len) == bvec_to_phys((vec2))) && \
++ ((bvec_to_pseudophys((vec1)) + (vec1)->bv_len) == \
++ bvec_to_pseudophys((vec2))))
++
++#include <asm-generic/iomap.h>
++
++extern void early_ioremap_init(void);
++extern void early_ioremap_clear(void);
++extern void early_ioremap_reset(void);
++extern void *early_ioremap(unsigned long addr, unsigned long size);
++extern void early_iounmap(void *addr, unsigned long size);
++
++/*
++ * This one maps high address device memory and turns off caching for that area.
++ * it's useful if some control registers are in such an area and write combining
++ * or read caching is not desirable:
++ */
++extern void __iomem *ioremap_nocache(resource_size_t offset, unsigned long size);
++extern void __iomem *ioremap_cache(resource_size_t offset, unsigned long size);
++
++/*
++ * The default ioremap() behavior is non-cached:
++ */
++static inline void __iomem *ioremap(resource_size_t offset, unsigned long size)
++{
++ return ioremap_nocache(offset, size);
++}
++
++extern void iounmap(volatile void __iomem *addr);
++
++extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys);
++
++/*
++ * ISA I/O bus memory addresses are 1:1 with the physical address.
++ */
++
++#define isa_virt_to_bus(_x) ({ BUG(); virt_to_bus(_x); })
++#define isa_page_to_bus(_x) isa_page_to_bus_is_UNSUPPORTED->x
++#define isa_bus_to_virt(_x) (void *)(__fix_to_virt(FIX_ISAMAP_BEGIN) + (_x))
++
++/*
++ * However PCI ones are not necessarily 1:1 and therefore these interfaces
++ * are forbidden in portable PCI drivers.
++ *
++ * Allow them on x86 for legacy drivers, though.
++ */
++#define virt_to_bus(_x) phys_to_machine(__pa(_x))
++#define bus_to_virt(_x) __va(machine_to_phys(_x))
++
++/*
++ * readX/writeX() are used to access memory mapped devices. On some
++ * architectures the memory mapped IO stuff needs to be accessed
++ * differently. On the x86 architecture, we just read/write the
++ * memory location directly.
++ */
++
++static inline __u8 __readb(const volatile void __iomem *addr)
++{
++ return *(__force volatile __u8 *)addr;
++}
++
++static inline __u16 __readw(const volatile void __iomem *addr)
++{
++ return *(__force volatile __u16 *)addr;
++}
++
++static __always_inline __u32 __readl(const volatile void __iomem *addr)
++{
++ return *(__force volatile __u32 *)addr;
++}
++
++static inline __u64 __readq(const volatile void __iomem *addr)
++{
++ return *(__force volatile __u64 *)addr;
++}
++
++#define readb(x) __readb(x)
++#define readw(x) __readw(x)
++#define readl(x) __readl(x)
++#define readq(x) __readq(x)
++#define readb_relaxed(a) readb(a)
++#define readw_relaxed(a) readw(a)
++#define readl_relaxed(a) readl(a)
++#define readq_relaxed(a) readq(a)
++#define __raw_readb readb
++#define __raw_readw readw
++#define __raw_readl readl
++#define __raw_readq readq
++
++#define mmiowb()
++
++static inline void __writel(__u32 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u32 *)addr = b;
++}
++
++static inline void __writeq(__u64 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u64 *)addr = b;
++}
++
++static inline void __writeb(__u8 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u8 *)addr = b;
++}
++
++static inline void __writew(__u16 b, volatile void __iomem *addr)
++{
++ *(__force volatile __u16 *)addr = b;
++}
++
++#define writeq(val, addr) __writeq((val), (addr))
++#define writel(val, addr) __writel((val), (addr))
++#define writew(val, addr) __writew((val), (addr))
++#define writeb(val, addr) __writeb((val), (addr))
++#define __raw_writeb writeb
++#define __raw_writew writew
++#define __raw_writel writel
++#define __raw_writeq writeq
++
++void __memcpy_fromio(void *, unsigned long, unsigned);
++void __memcpy_toio(unsigned long, const void *, unsigned);
++
++static inline void memcpy_fromio(void *to, const volatile void __iomem *from,
++ unsigned len)
++{
++ __memcpy_fromio(to, (unsigned long)from, len);
++}
++
++static inline void memcpy_toio(volatile void __iomem *to, const void *from,
++ unsigned len)
++{
++ __memcpy_toio((unsigned long)to, from, len);
++}
++
++void memset_io(volatile void __iomem *a, int b, size_t c);
++
++/*
++ * ISA space is 'always mapped' on a typical x86 system, no need to
++ * explicitly ioremap() it. The fact that the ISA IO space is mapped
++ * to PAGE_OFFSET is pure coincidence - it does not mean ISA values
++ * are physical addresses. The following constant pointer can be
++ * used as the IO-area pointer (it can be iounmapped as well, so the
++ * analogy with PCI is quite large):
++ */
++#define __ISA_IO_base ((char __iomem *)(fix_to_virt(FIX_ISAMAP_BEGIN)))
++
++#define flush_write_buffers()
++
++extern int iommu_bio_merge;
++#define BIO_VMERGE_BOUNDARY iommu_bio_merge
++
++/*
++ * Convert a virtual cached pointer to an uncached pointer
++ */
++#define xlate_dev_kmem_ptr(p) p
++
++#endif /* __KERNEL__ */
++
++#define ARCH_HAS_DEV_MEM
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/irq.h b/include/asm-x86/mach-xen/asm/irq.h
+new file mode 100644
+index 0000000..b4f2c3e
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/irq.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "../../irq_32.h"
++#else
++# include "irq_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/irq_64.h b/include/asm-x86/mach-xen/asm/irq_64.h
+new file mode 100644
+index 0000000..da62355
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/irq_64.h
+@@ -0,0 +1,38 @@
++#ifndef _ASM_IRQ_H
++#define _ASM_IRQ_H
++
++/*
++ * linux/include/asm/irq.h
++ *
++ * (C) 1992, 1993 Linus Torvalds, (C) 1997 Ingo Molnar
++ *
++ * IRQ/IPI changes taken from work by Thomas Radke
++ * <tomsoft at informatik.tu-chemnitz.de>
++ */
++
++#include <linux/sched.h>
++/* include comes from machine specific directory */
++#include "irq_vectors.h"
++#include <asm/thread_info.h>
++
++static inline int irq_canonicalize(int irq)
++{
++ return ((irq == 2) ? 9 : irq);
++}
++
++#ifndef CONFIG_XEN
++#define ARCH_HAS_NMI_WATCHDOG /* See include/linux/nmi.h */
++#endif
++
++#define KDB_VECTOR 0xf9
++
++# define irq_ctx_init(cpu) do { } while (0)
++
++#ifdef CONFIG_HOTPLUG_CPU
++#include <linux/cpumask.h>
++extern void fixup_irqs(cpumask_t map);
++#endif
++
++#define __ARCH_HAS_DO_SOFTIRQ 1
++
++#endif /* _ASM_IRQ_H */
+diff --git a/include/asm-x86/mach-xen/asm/irqflags.h b/include/asm-x86/mach-xen/asm/irqflags.h
+new file mode 100644
+index 0000000..e403958
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/irqflags.h
+@@ -0,0 +1,250 @@
++#ifndef _X86_IRQFLAGS_H_
++#define _X86_IRQFLAGS_H_
++
++#include <asm/processor-flags.h>
++
++#ifndef __ASSEMBLY__
++/*
++ * The use of 'barrier' in the following reflects their use as local-lock
++ * operations. Reentrancy must be prevented (e.g., __cli()) /before/ following
++ * critical operations are executed. All critical operations must complete
++ * /before/ reentrancy is permitted (e.g., __sti()). Alpha architecture also
++ * includes these barriers, for example.
++ */
++
++#define xen_save_fl(void) (current_vcpu_info()->evtchn_upcall_mask)
++
++#define xen_restore_fl(f) \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ if ((_vcpu->evtchn_upcall_mask = (f)) == 0) { \
++ barrier(); /* unmask then check (avoid races) */\
++ if (unlikely(_vcpu->evtchn_upcall_pending)) \
++ force_evtchn_callback(); \
++ } \
++} while (0)
++
++#define xen_irq_disable() \
++do { \
++ current_vcpu_info()->evtchn_upcall_mask = 1; \
++ barrier(); \
++} while (0)
++
++#define xen_irq_enable() \
++do { \
++ vcpu_info_t *_vcpu; \
++ barrier(); \
++ _vcpu = current_vcpu_info(); \
++ _vcpu->evtchn_upcall_mask = 0; \
++ barrier(); /* unmask then check (avoid races) */ \
++ if (unlikely(_vcpu->evtchn_upcall_pending)) \
++ force_evtchn_callback(); \
++} while (0)
++
++void xen_safe_halt(void);
++
++void xen_halt(void);
++#endif
++
++#ifndef __ASSEMBLY__
++
++#define __raw_local_save_flags(void) xen_save_fl()
++
++#define raw_local_irq_restore(flags) xen_restore_fl(flags)
++
++#define raw_local_irq_disable() xen_irq_disable()
++
++#define raw_local_irq_enable() xen_irq_enable()
++
++/*
++ * Used in the idle loop; sti takes one instruction cycle
++ * to complete:
++ */
++static inline void raw_safe_halt(void)
++{
++ xen_safe_halt();
++}
++
++/*
++ * Used when interrupts are already enabled or to
++ * shutdown the processor:
++ */
++static inline void halt(void)
++{
++ xen_halt();
++}
++
++/*
++ * For spinlocks, etc:
++ */
++#define __raw_local_irq_save() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_local_irq_disable(); \
++ \
++ flags; \
++})
++#else
++
++/* Offsets into shared_info_t. */
++#define evtchn_upcall_pending /* 0 */
++#define evtchn_upcall_mask 1
++
++#define sizeof_vcpu_shift 6
++
++#ifdef CONFIG_X86_64
++# define __REG_si %rsi
++# define __CPU_num %gs:pda_cpunumber
++#else
++# define __REG_si %esi
++# define __CPU_num TI_cpu(%ebp)
++#endif
++
++#ifdef CONFIG_SMP
++#define GET_VCPU_INFO movl __CPU_num,%esi ; \
++ shl $sizeof_vcpu_shift,%esi ; \
++ add HYPERVISOR_shared_info,__REG_si
++#else
++#define GET_VCPU_INFO mov HYPERVISOR_shared_info,__REG_si
++#endif
++
++#define __DISABLE_INTERRUPTS movb $1,evtchn_upcall_mask(__REG_si)
++#define __ENABLE_INTERRUPTS movb $0,evtchn_upcall_mask(__REG_si)
++#define __TEST_PENDING testb $0xFF,evtchn_upcall_pending(__REG_si)
++#define DISABLE_INTERRUPTS(clb) GET_VCPU_INFO ; \
++ __DISABLE_INTERRUPTS
++#define ENABLE_INTERRUPTS(clb) GET_VCPU_INFO ; \
++ __ENABLE_INTERRUPTS
++
++#ifndef CONFIG_X86_64
++#define INTERRUPT_RETURN iret
++#define ENABLE_INTERRUPTS_SYSCALL_RET __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 ; \
++ TRACE_IRQS_OFF ; \
++sysexit_ecrit: /**** END OF SYSEXIT CRITICAL REGION ****/ ; \
++ push %esp ; \
++ call evtchn_do_upcall ; \
++ add $4,%esp ; \
++ jmp ret_from_intr
++#endif
++
++
++#endif /* __ASSEMBLY__ */
++
++#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);
++}
++
++#define raw_irqs_disabled() \
++({ \
++ unsigned long flags = __raw_local_save_flags(); \
++ \
++ raw_irqs_disabled_flags(flags); \
++})
++
++/*
++ * makes the traced hardirq state match with the machine state
++ *
++ * should be a rarely used function, only in places where its
++ * otherwise impossible to know the irq state, like in traps.
++ */
++static inline void trace_hardirqs_fixup_flags(unsigned long flags)
++{
++ if (raw_irqs_disabled_flags(flags))
++ trace_hardirqs_off();
++ else
++ trace_hardirqs_on();
++}
++
++#define trace_hardirqs_fixup() \
++ trace_hardirqs_fixup_flags(__raw_local_save_flags())
++
++#else
++
++#ifdef CONFIG_X86_64
++/*
++ * Currently paravirt can't handle swapgs nicely when we
++ * don't have a stack we can rely on (such as a user space
++ * stack). So we either find a way around these or just fault
++ * and emulate if a guest tries to call swapgs directly.
++ *
++ * Either way, this is a good way to document that we don't
++ * have a reliable stack. x86_64 only.
++ */
++#define SWAPGS_UNSAFE_STACK swapgs
++#define ARCH_TRACE_IRQS_ON call trace_hardirqs_on_thunk
++#define ARCH_TRACE_IRQS_OFF call trace_hardirqs_off_thunk
++#define ARCH_LOCKDEP_SYS_EXIT call lockdep_sys_exit_thunk
++#define ARCH_LOCKDEP_SYS_EXIT_IRQ \
++ TRACE_IRQS_ON; \
++ ENABLE_INTERRUPTS(CLBR_NONE); \
++ SAVE_REST; \
++ LOCKDEP_SYS_EXIT; \
++ RESTORE_REST; \
++ __DISABLE_INTERRUPTS; \
++ TRACE_IRQS_OFF;
++
++#else
++#define ARCH_TRACE_IRQS_ON \
++ pushl %eax; \
++ pushl %ecx; \
++ pushl %edx; \
++ call trace_hardirqs_on; \
++ popl %edx; \
++ popl %ecx; \
++ popl %eax;
++
++#define ARCH_TRACE_IRQS_OFF \
++ pushl %eax; \
++ pushl %ecx; \
++ pushl %edx; \
++ call trace_hardirqs_off; \
++ popl %edx; \
++ popl %ecx; \
++ popl %eax;
++
++#define ARCH_LOCKDEP_SYS_EXIT \
++ pushl %eax; \
++ pushl %ecx; \
++ pushl %edx; \
++ call lockdep_sys_exit; \
++ popl %edx; \
++ popl %ecx; \
++ popl %eax;
++
++#define ARCH_LOCKDEP_SYS_EXIT_IRQ
++#endif
++
++#ifdef CONFIG_TRACE_IRQFLAGS
++# define TRACE_IRQS_ON ARCH_TRACE_IRQS_ON
++# define TRACE_IRQS_OFF ARCH_TRACE_IRQS_OFF
++#else
++# define TRACE_IRQS_ON
++# define TRACE_IRQS_OFF
++#endif
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define LOCKDEP_SYS_EXIT ARCH_LOCKDEP_SYS_EXIT
++# define LOCKDEP_SYS_EXIT_IRQ ARCH_LOCKDEP_SYS_EXIT_IRQ
++# else
++# define LOCKDEP_SYS_EXIT
++# define LOCKDEP_SYS_EXIT_IRQ
++# endif
++
++#endif /* __ASSEMBLY__ */
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/maddr.h b/include/asm-x86/mach-xen/asm/maddr.h
+new file mode 100644
+index 0000000..934054c
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/maddr.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "maddr_32.h"
++#else
++# include "maddr_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/maddr_32.h b/include/asm-x86/mach-xen/asm/maddr_32.h
+new file mode 100644
+index 0000000..ed30dac
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/maddr_32.h
+@@ -0,0 +1,179 @@
++#ifndef _I386_MADDR_H
++#define _I386_MADDR_H
++
++#include <asm/bug.h>
++#include <xen/features.h>
++#include <xen/interface/xen.h>
++
++/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
++#define INVALID_P2M_ENTRY (~0UL)
++#define FOREIGN_FRAME_BIT (1UL<<31)
++#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
++
++/* Definitions for machine and pseudophysical addresses. */
++#ifdef CONFIG_X86_PAE
++typedef unsigned long long paddr_t;
++typedef unsigned long long maddr_t;
++#else
++typedef unsigned long paddr_t;
++typedef unsigned long maddr_t;
++#endif
++
++#ifdef CONFIG_XEN
++
++extern unsigned long *phys_to_machine_mapping;
++extern unsigned long max_mapnr;
++
++#undef machine_to_phys_mapping
++extern unsigned long *machine_to_phys_mapping;
++extern unsigned int machine_to_phys_order;
++
++static inline unsigned long pfn_to_mfn(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return pfn;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
++}
++
++static inline int phys_to_machine_mapping_valid(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 1;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
++}
++
++static inline unsigned long mfn_to_pfn(unsigned long mfn)
++{
++ unsigned long pfn;
++
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return mfn;
++
++ if (unlikely((mfn >> machine_to_phys_order) != 0))
++ return max_mapnr;
++
++ /* The array access can fail (e.g., device space beyond end of RAM). */
++ asm (
++ "1: movl %1,%0\n"
++ "2:\n"
++ ".section .fixup,\"ax\"\n"
++ "3: movl %2,%0\n"
++ " jmp 2b\n"
++ ".previous\n"
++ ".section __ex_table,\"a\"\n"
++ " .align 4\n"
++ " .long 1b,3b\n"
++ ".previous"
++ : "=r" (pfn)
++ : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) );
++
++ return pfn;
++}
++
++/*
++ * We detect special mappings in one of two ways:
++ * 1. If the MFN is an I/O page then Xen will set the m2p entry
++ * to be outside our maximum possible pseudophys range.
++ * 2. If the MFN belongs to a different domain then we will certainly
++ * not have MFN in our p2m table. Conversely, if the page is ours,
++ * then we'll have p2m(m2p(MFN))==MFN.
++ * If we detect a special mapping then it doesn't have a 'struct page'.
++ * We force !pfn_valid() by returning an out-of-range pointer.
++ *
++ * NB. These checks require that, for any MFN that is not in our reservation,
++ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
++ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
++ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
++ *
++ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
++ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
++ * require. In all the cases we care about, the FOREIGN_FRAME bit is
++ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
++ */
++static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
++{
++ unsigned long pfn = mfn_to_pfn(mfn);
++ if ((pfn < max_mapnr)
++ && !xen_feature(XENFEAT_auto_translated_physmap)
++ && (phys_to_machine_mapping[pfn] != mfn))
++ return max_mapnr; /* force !pfn_valid() */
++ return pfn;
++}
++
++static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
++{
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
++ return;
++ }
++ phys_to_machine_mapping[pfn] = mfn;
++}
++
++static inline maddr_t phys_to_machine(paddr_t phys)
++{
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
++ return machine;
++}
++
++static inline paddr_t machine_to_phys(maddr_t machine)
++{
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
++ return phys;
++}
++
++#ifdef CONFIG_X86_PAE
++static inline paddr_t pte_phys_to_machine(paddr_t phys)
++{
++ /*
++ * In PAE mode, the NX bit needs to be dealt with in the value
++ * passed to pfn_to_mfn(). On x86_64, we need to mask it off,
++ * but for i386 the conversion to ulong for the argument will
++ * clip it off.
++ */
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
++ return machine;
++}
++
++static inline paddr_t pte_machine_to_phys(maddr_t machine)
++{
++ /*
++ * In PAE mode, the NX bit needs to be dealt with in the value
++ * passed to mfn_to_pfn(). On x86_64, we need to mask it off,
++ * but for i386 the conversion to ulong for the argument will
++ * clip it off.
++ */
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
++ return phys;
++}
++#else
++#define pte_phys_to_machine phys_to_machine
++#define pte_machine_to_phys machine_to_phys
++#endif
++
++#else /* !CONFIG_XEN */
++
++#define pfn_to_mfn(pfn) (pfn)
++#define mfn_to_pfn(mfn) (mfn)
++#define mfn_to_local_pfn(mfn) (mfn)
++#define set_phys_to_machine(pfn, mfn) ((void)0)
++#define phys_to_machine_mapping_valid(pfn) (1)
++#define phys_to_machine(phys) ((maddr_t)(phys))
++#define machine_to_phys(mach) ((paddr_t)(mach))
++#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
++#define __pte_ma(x) __pte(x)
++
++#endif /* !CONFIG_XEN */
++
++/* VIRT <-> MACHINE conversion */
++#define virt_to_machine(v) (phys_to_machine(__pa(v)))
++#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
++#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
++
++#endif /* _I386_MADDR_H */
+diff --git a/include/asm-x86/mach-xen/asm/maddr_64.h b/include/asm-x86/mach-xen/asm/maddr_64.h
+new file mode 100644
+index 0000000..cc97cc5
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/maddr_64.h
+@@ -0,0 +1,160 @@
++#ifndef _X86_64_MADDR_H
++#define _X86_64_MADDR_H
++
++#include <asm/bug.h>
++#include <xen/features.h>
++#include <xen/interface/xen.h>
++
++/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
++#define INVALID_P2M_ENTRY (~0UL)
++#define FOREIGN_FRAME_BIT (1UL<<63)
++#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
++
++/* Definitions for machine and pseudophysical addresses. */
++typedef unsigned long paddr_t;
++typedef unsigned long maddr_t;
++
++#ifdef CONFIG_XEN
++
++extern unsigned long *phys_to_machine_mapping;
++extern unsigned long max_mapnr;
++
++#undef machine_to_phys_mapping
++extern unsigned long *machine_to_phys_mapping;
++extern unsigned int machine_to_phys_order;
++
++static inline unsigned long pfn_to_mfn(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return pfn;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return phys_to_machine_mapping[pfn] & ~FOREIGN_FRAME_BIT;
++}
++
++static inline int phys_to_machine_mapping_valid(unsigned long pfn)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return 1;
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ return (phys_to_machine_mapping[pfn] != INVALID_P2M_ENTRY);
++}
++
++static inline unsigned long mfn_to_pfn(unsigned long mfn)
++{
++ unsigned long pfn;
++
++ if (xen_feature(XENFEAT_auto_translated_physmap))
++ return mfn;
++
++ if (unlikely((mfn >> machine_to_phys_order) != 0))
++ return max_mapnr;
++
++ /* The array access can fail (e.g., device space beyond end of RAM). */
++ asm (
++ "1: movq %1,%0\n"
++ "2:\n"
++ ".section .fixup,\"ax\"\n"
++ "3: movq %2,%0\n"
++ " jmp 2b\n"
++ ".previous\n"
++ ".section __ex_table,\"a\"\n"
++ " .align 8\n"
++ " .quad 1b,3b\n"
++ ".previous"
++ : "=r" (pfn)
++ : "m" (machine_to_phys_mapping[mfn]), "m" (max_mapnr) );
++
++ return pfn;
++}
++
++/*
++ * We detect special mappings in one of two ways:
++ * 1. If the MFN is an I/O page then Xen will set the m2p entry
++ * to be outside our maximum possible pseudophys range.
++ * 2. If the MFN belongs to a different domain then we will certainly
++ * not have MFN in our p2m table. Conversely, if the page is ours,
++ * then we'll have p2m(m2p(MFN))==MFN.
++ * If we detect a special mapping then it doesn't have a 'struct page'.
++ * We force !pfn_valid() by returning an out-of-range pointer.
++ *
++ * NB. These checks require that, for any MFN that is not in our reservation,
++ * there is no PFN such that p2m(PFN) == MFN. Otherwise we can get confused if
++ * we are foreign-mapping the MFN, and the other domain as m2p(MFN) == PFN.
++ * Yikes! Various places must poke in INVALID_P2M_ENTRY for safety.
++ *
++ * NB2. When deliberately mapping foreign pages into the p2m table, you *must*
++ * use FOREIGN_FRAME(). This will cause pte_pfn() to choke on it, as we
++ * require. In all the cases we care about, the FOREIGN_FRAME bit is
++ * masked (e.g., pfn_to_mfn()) so behaviour there is correct.
++ */
++static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
++{
++ unsigned long pfn = mfn_to_pfn(mfn);
++ if ((pfn < max_mapnr)
++ && !xen_feature(XENFEAT_auto_translated_physmap)
++ && (phys_to_machine_mapping[pfn] != mfn))
++ return max_mapnr; /* force !pfn_valid() */
++ return pfn;
++}
++
++static inline void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
++{
++ BUG_ON(max_mapnr && pfn >= max_mapnr);
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
++ return;
++ }
++ phys_to_machine_mapping[pfn] = mfn;
++}
++
++static inline maddr_t phys_to_machine(paddr_t phys)
++{
++ maddr_t machine = pfn_to_mfn(phys >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PAGE_MASK);
++ return machine;
++}
++
++static inline paddr_t machine_to_phys(maddr_t machine)
++{
++ paddr_t phys = mfn_to_pfn(machine >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PAGE_MASK);
++ return phys;
++}
++
++static inline paddr_t pte_phys_to_machine(paddr_t phys)
++{
++ maddr_t machine;
++ machine = pfn_to_mfn((phys & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
++ machine = (machine << PAGE_SHIFT) | (phys & ~PHYSICAL_PAGE_MASK);
++ return machine;
++}
++
++static inline paddr_t pte_machine_to_phys(maddr_t machine)
++{
++ paddr_t phys;
++ phys = mfn_to_pfn((machine & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT);
++ phys = (phys << PAGE_SHIFT) | (machine & ~PHYSICAL_PAGE_MASK);
++ return phys;
++}
++
++#else /* !CONFIG_XEN */
++
++#define pfn_to_mfn(pfn) (pfn)
++#define mfn_to_pfn(mfn) (mfn)
++#define mfn_to_local_pfn(mfn) (mfn)
++#define set_phys_to_machine(pfn, mfn) ((void)0)
++#define phys_to_machine_mapping_valid(pfn) (1)
++#define phys_to_machine(phys) ((maddr_t)(phys))
++#define machine_to_phys(mach) ((paddr_t)(mach))
++#define pfn_pte_ma(pfn, prot) pfn_pte(pfn, prot)
++#define __pte_ma(x) __pte(x)
++
++#endif /* !CONFIG_XEN */
++
++/* VIRT <-> MACHINE conversion */
++#define virt_to_machine(v) (phys_to_machine(__pa(v)))
++#define virt_to_mfn(v) (pfn_to_mfn(__pa(v) >> PAGE_SHIFT))
++#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
++
++#endif /* _X86_64_MADDR_H */
++
+diff --git a/include/asm-x86/mach-xen/asm/mmu_context.h b/include/asm-x86/mach-xen/asm/mmu_context.h
+new file mode 100644
+index 0000000..6598450
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/mmu_context.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "mmu_context_32.h"
++#else
++# include "mmu_context_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/mmu_context_32.h b/include/asm-x86/mach-xen/asm/mmu_context_32.h
+new file mode 100644
+index 0000000..832716f
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/mmu_context_32.h
+@@ -0,0 +1,116 @@
++#ifndef __I386_SCHED_H
++#define __I386_SCHED_H
++
++#include <asm/desc.h>
++#include <asm/atomic.h>
++#include <asm/pgalloc.h>
++#include <asm/tlbflush.h>
++
++void arch_exit_mmap(struct mm_struct *mm);
++void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm);
++
++void mm_pin(struct mm_struct *mm);
++void mm_unpin(struct mm_struct *mm);
++void mm_pin_all(void);
++
++static inline void xen_activate_mm(struct mm_struct *prev,
++ struct mm_struct *next)
++{
++ if (!PagePinned(virt_to_page(next->pgd)))
++ mm_pin(next);
++}
++
++/*
++ * Used for LDT copy/destruction.
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
++void destroy_context(struct mm_struct *mm);
++
++
++static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
++{
++#if 0 /* XEN: no lazy tlb */
++ unsigned cpu = smp_processor_id();
++ if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK)
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_LAZY;
++#endif
++}
++
++#define prepare_arch_switch(next) __prepare_arch_switch()
++
++static inline void __prepare_arch_switch(void)
++{
++ /*
++ * Save away %gs. No need to save %fs, as it was saved on the
++ * stack on entry. No need to save %es and %ds, as those are
++ * always kernel segments while inside the kernel.
++ */
++ asm volatile ( "mov %%gs,%0"
++ : "=m" (current->thread.gs));
++ asm volatile ( "movl %0,%%gs"
++ : : "r" (0) );
++}
++
++static inline void switch_mm(struct mm_struct *prev,
++ struct mm_struct *next,
++ struct task_struct *tsk)
++{
++ int cpu = smp_processor_id();
++ struct mmuext_op _op[2], *op = _op;
++
++ if (likely(prev != next)) {
++ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
++ !PagePinned(virt_to_page(next->pgd)));
++
++ /* stop flush ipis for the previous mm */
++ cpu_clear(cpu, prev->cpu_vm_mask);
++#if 0 /* XEN: no lazy tlb */
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
++ per_cpu(cpu_tlbstate, cpu).active_mm = next;
++#endif
++ cpu_set(cpu, next->cpu_vm_mask);
++
++ /* Re-load page tables: load_cr3(next->pgd) */
++ op->cmd = MMUEXT_NEW_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
++ op++;
++
++ /*
++ * load the LDT, if the LDT is different:
++ */
++ if (unlikely(prev->context.ldt != next->context.ldt)) {
++ /* load_LDT_nolock(&next->context, cpu) */
++ op->cmd = MMUEXT_SET_LDT;
++ op->arg1.linear_addr = (unsigned long)next->context.ldt;
++ op->arg2.nr_ents = next->context.size;
++ op++;
++ }
++
++ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
++ }
++#if 0 /* XEN: no lazy tlb */
++ else {
++ per_cpu(cpu_tlbstate, cpu).state = TLBSTATE_OK;
++ BUG_ON(per_cpu(cpu_tlbstate, cpu).active_mm != next);
++
++ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
++ /* We were in lazy tlb mode and leave_mm disabled
++ * tlb flush IPI delivery. We must reload %cr3.
++ */
++ load_cr3(next->pgd);
++ load_LDT_nolock(&next->context);
++ }
++ }
++#endif
++}
++
++#define deactivate_mm(tsk, mm) \
++ asm("movl %0,%%gs": :"r" (0));
++
++#define activate_mm(prev, next) \
++do { \
++ xen_activate_mm(prev, next); \
++ switch_mm((prev), (next), NULL); \
++} while (0)
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/mmu_context_64.h b/include/asm-x86/mach-xen/asm/mmu_context_64.h
+new file mode 100644
+index 0000000..e311015
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/mmu_context_64.h
+@@ -0,0 +1,134 @@
++#ifndef __X86_64_MMU_CONTEXT_H
++#define __X86_64_MMU_CONTEXT_H
++
++#include <asm/desc.h>
++#include <asm/atomic.h>
++#include <asm/pgalloc.h>
++#include <asm/page.h>
++#include <asm/pda.h>
++#include <asm/pgtable.h>
++#include <asm/tlbflush.h>
++
++void arch_exit_mmap(struct mm_struct *mm);
++void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm);
++
++/*
++ * possibly do the LDT unload here?
++ */
++int init_new_context(struct task_struct *tsk, struct mm_struct *mm);
++void destroy_context(struct mm_struct *mm);
++
++static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
++{
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ if (read_pda(mmu_state) == TLBSTATE_OK)
++ write_pda(mmu_state, TLBSTATE_LAZY);
++#endif
++}
++
++#define prepare_arch_switch(next) __prepare_arch_switch()
++
++static inline void __prepare_arch_switch(void)
++{
++ /*
++ * Save away %es, %ds, %fs and %gs. Must happen before reload
++ * of cr3/ldt (i.e., not in __switch_to).
++ */
++ __asm__ __volatile__ (
++ "mov %%es,%0 ; mov %%ds,%1 ; mov %%fs,%2 ; mov %%gs,%3"
++ : "=m" (current->thread.es),
++ "=m" (current->thread.ds),
++ "=m" (current->thread.fsindex),
++ "=m" (current->thread.gsindex) );
++
++ if (current->thread.ds)
++ __asm__ __volatile__ ( "movl %0,%%ds" : : "r" (0) );
++
++ if (current->thread.es)
++ __asm__ __volatile__ ( "movl %0,%%es" : : "r" (0) );
++
++ if (current->thread.fsindex) {
++ __asm__ __volatile__ ( "movl %0,%%fs" : : "r" (0) );
++ current->thread.fs = 0;
++ }
++
++ if (current->thread.gsindex) {
++ load_gs_index(0);
++ current->thread.gs = 0;
++ }
++}
++
++extern void mm_pin(struct mm_struct *mm);
++extern void mm_unpin(struct mm_struct *mm);
++void mm_pin_all(void);
++
++static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
++ struct task_struct *tsk)
++{
++ unsigned cpu = smp_processor_id();
++ struct mmuext_op _op[3], *op = _op;
++
++ if (likely(prev != next)) {
++ BUG_ON(!xen_feature(XENFEAT_writable_page_tables) &&
++ !PagePinned(virt_to_page(next->pgd)));
++
++ /* stop flush ipis for the previous mm */
++ cpu_clear(cpu, prev->cpu_vm_mask);
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ write_pda(mmu_state, TLBSTATE_OK);
++ write_pda(active_mm, next);
++#endif
++ cpu_set(cpu, next->cpu_vm_mask);
++
++ /* load_cr3(next->pgd) */
++ op->cmd = MMUEXT_NEW_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(next->pgd) >> PAGE_SHIFT);
++ op++;
++
++ /* xen_new_user_pt(__pa(__user_pgd(next->pgd))) */
++ op->cmd = MMUEXT_NEW_USER_BASEPTR;
++ op->arg1.mfn = pfn_to_mfn(__pa(__user_pgd(next->pgd)) >> PAGE_SHIFT);
++ op++;
++
++ if (unlikely(next->context.ldt != prev->context.ldt)) {
++ /* load_LDT_nolock(&next->context) */
++ op->cmd = MMUEXT_SET_LDT;
++ op->arg1.linear_addr = (unsigned long)next->context.ldt;
++ op->arg2.nr_ents = next->context.size;
++ op++;
++ }
++
++ BUG_ON(HYPERVISOR_mmuext_op(_op, op-_op, NULL, DOMID_SELF));
++ }
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++ else {
++ write_pda(mmu_state, TLBSTATE_OK);
++ if (read_pda(active_mm) != next)
++ BUG();
++ if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) {
++ /* We were in lazy tlb mode and leave_mm disabled
++ * tlb flush IPI delivery. We must reload CR3
++ * to make sure to use no freed page tables.
++ */
++ load_cr3(next->pgd);
++ xen_new_user_pt(__pa(__user_pgd(next->pgd)));
++ load_LDT_nolock(&next->context);
++ }
++ }
++#endif
++}
++
++#define deactivate_mm(tsk, mm) \
++do { \
++ load_gs_index(0); \
++ asm volatile("movl %0,%%fs"::"r"(0)); \
++} while (0)
++
++static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
++{
++ if (!PagePinned(virt_to_page(next->pgd)))
++ mm_pin(next);
++ switch_mm(prev, next, NULL);
++}
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/page.h b/include/asm-x86/mach-xen/asm/page.h
+new file mode 100644
+index 0000000..71e3d8d
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/page.h
+@@ -0,0 +1,237 @@
++#ifndef _ASM_X86_PAGE_H
++#define _ASM_X86_PAGE_H
++
++#include <linux/const.h>
++
++/* PAGE_SHIFT determines the page size */
++#define PAGE_SHIFT 12
++#define PAGE_SIZE (_AC(1,UL) << PAGE_SHIFT)
++#define PAGE_MASK (~(PAGE_SIZE-1))
++
++#ifdef __KERNEL__
++
++/*
++ * Need to repeat this here in order to not include pgtable.h (which in turn
++ * depends on definitions made here), but to be able to use the symbolics
++ * below. The preprocessor will warn if the two definitions aren't identical.
++ */
++#define _PAGE_BIT_PRESENT 0
++#define _PAGE_PRESENT (_AC(1, L)<<_PAGE_BIT_PRESENT)
++#define _PAGE_BIT_IO 9
++#define _PAGE_IO (_AC(1, L)<<_PAGE_BIT_IO)
++
++#define __PHYSICAL_MASK ((phys_addr_t)(1ULL << __PHYSICAL_MASK_SHIFT) - 1)
++#define __VIRTUAL_MASK ((1UL << __VIRTUAL_MASK_SHIFT) - 1)
++
++/* Cast PAGE_MASK to a signed type so that it is sign-extended if
++ virtual addresses are 32-bits but physical addresses are larger
++ (ie, 32-bit PAE). */
++#define PHYSICAL_PAGE_MASK (((signed long)PAGE_MASK) & __PHYSICAL_MASK)
++
++/* PTE_MASK extracts the PFN from a (pte|pmd|pud|pgd)val_t */
++#define PTE_MASK ((pteval_t)PHYSICAL_PAGE_MASK)
++
++#define PMD_PAGE_SIZE (_AC(1, UL) << PMD_SHIFT)
++#define PMD_PAGE_MASK (~(PMD_PAGE_SIZE-1))
++
++#define HPAGE_SHIFT PMD_SHIFT
++#define HPAGE_SIZE (_AC(1,UL) << HPAGE_SHIFT)
++#define HPAGE_MASK (~(HPAGE_SIZE - 1))
++#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
++
++/* to align the pointer to the (next) page boundary */
++#define PAGE_ALIGN(addr) (((addr)+PAGE_SIZE-1)&PAGE_MASK)
++
++#ifndef __ASSEMBLY__
++#include <linux/types.h>
++#endif
++
++#ifdef CONFIG_X86_64
++#include <asm/page_64.h>
++#else
++#include <asm/page_32.h>
++#endif /* CONFIG_X86_64 */
++
++#define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET)
++
++#define VM_DATA_DEFAULT_FLAGS \
++ (((current->personality & READ_IMPLIES_EXEC) ? VM_EXEC : 0 ) | \
++ VM_READ | VM_WRITE | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC)
++
++
++#ifndef __ASSEMBLY__
++
++extern int page_is_ram(unsigned long pagenr);
++extern int devmem_is_allowed(unsigned long pagenr);
++
++extern unsigned long max_pfn_mapped;
++
++struct page;
++
++static inline void clear_user_page(void *page, unsigned long vaddr,
++ struct page *pg)
++{
++ clear_page(page);
++}
++
++static inline void copy_user_page(void *to, void *from, unsigned long vaddr,
++ struct page *topage)
++{
++ copy_page(to, from);
++}
++
++#define __alloc_zeroed_user_highpage(movableflags, vma, vaddr) \
++ alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO | movableflags, vma, vaddr)
++#define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE
++
++typedef struct { pgprotval_t pgprot; } pgprot_t;
++
++#define pgprot_val(x) ((x).pgprot)
++#define __pgprot(x) ((pgprot_t) { (x) } )
++
++#include <asm/maddr.h>
++
++typedef struct { pgdval_t pgd; } pgd_t;
++
++#define __pgd_ma(x) ((pgd_t) { (x) } )
++static inline pgd_t xen_make_pgd(pgdval_t val)
++{
++ if (val & _PAGE_PRESENT)
++ val = pte_phys_to_machine(val);
++ return (pgd_t) { val };
++}
++
++#define __pgd_val(x) ((x).pgd)
++static inline pgdval_t xen_pgd_val(pgd_t pgd)
++{
++ pgdval_t ret = __pgd_val(pgd);
++#if PAGETABLE_LEVELS == 2 && CONFIG_XEN_COMPAT <= 0x030002
++ if (ret)
++ ret = machine_to_phys(ret) | _PAGE_PRESENT;
++#else
++ if (ret & _PAGE_PRESENT)
++ ret = pte_machine_to_phys(ret);
++#endif
++ return ret;
++}
++
++#if PAGETABLE_LEVELS >= 3
++#if PAGETABLE_LEVELS == 4
++typedef struct { pudval_t pud; } pud_t;
++
++#define __pud_ma(x) ((pud_t) { (x) } )
++static inline pud_t xen_make_pud(pudval_t val)
++{
++ if (val & _PAGE_PRESENT)
++ val = pte_phys_to_machine(val);
++ return (pud_t) { val };
++}
++
++#define __pud_val(x) ((x).pud)
++static inline pudval_t xen_pud_val(pud_t pud)
++{
++ pudval_t ret = __pud_val(pud);
++ if (ret & _PAGE_PRESENT)
++ ret = pte_machine_to_phys(ret);
++ return ret;
++}
++#else /* PAGETABLE_LEVELS == 3 */
++#include <asm-generic/pgtable-nopud.h>
++
++#define __pud_val(x) __pgd_val((x).pgd)
++static inline pudval_t xen_pud_val(pud_t pud)
++{
++ return xen_pgd_val(pud.pgd);
++}
++#endif /* PAGETABLE_LEVELS == 4 */
++
++typedef struct { pmdval_t pmd; } pmd_t;
++
++#define __pmd_ma(x) ((pmd_t) { (x) } )
++static inline pmd_t xen_make_pmd(pmdval_t val)
++{
++ if (val & _PAGE_PRESENT)
++ val = pte_phys_to_machine(val);
++ return (pmd_t) { val };
++}
++
++#define __pmd_val(x) ((x).pmd)
++static inline pmdval_t xen_pmd_val(pmd_t pmd)
++{
++ pmdval_t ret = __pmd_val(pmd);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret)
++ ret = pte_machine_to_phys(ret) | _PAGE_PRESENT;
++#else
++ if (ret & _PAGE_PRESENT)
++ ret = pte_machine_to_phys(ret);
++#endif
++ return ret;
++}
++#else /* PAGETABLE_LEVELS == 2 */
++#include <asm-generic/pgtable-nopmd.h>
++
++#define __pmd_ma(x) ((pmd_t) { .pud.pgd = __pgd_ma(x) } )
++#define __pmd_val(x) __pgd_val((x).pud.pgd)
++static inline pmdval_t xen_pmd_val(pmd_t pmd)
++{
++ return xen_pgd_val(pmd.pud.pgd);
++}
++#endif /* PAGETABLE_LEVELS >= 3 */
++
++#define __pte_ma(x) ((pte_t) { .pte = (x) } )
++static inline pte_t xen_make_pte(unsigned long long val)
++{
++ if ((val & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT)
++ val = pte_phys_to_machine(val);
++ return (pte_t) { .pte = val };
++}
++
++#define __pte_val(x) ((x).pte)
++static inline pteval_t xen_pte_val(pte_t pte)
++{
++ pteval_t ret = __pte_val(pte);
++ if ((pte.pte_low & (_PAGE_PRESENT|_PAGE_IO)) == _PAGE_PRESENT)
++ ret = pte_machine_to_phys(ret);
++ return ret;
++}
++
++#define pgd_val(x) xen_pgd_val(x)
++#define __pgd(x) xen_make_pgd(x)
++
++#ifndef __PAGETABLE_PUD_FOLDED
++#define pud_val(x) xen_pud_val(x)
++#define __pud(x) xen_make_pud(x)
++#endif
++
++#ifndef __PAGETABLE_PMD_FOLDED
++#define pmd_val(x) xen_pmd_val(x)
++#define __pmd(x) xen_make_pmd(x)
++#endif
++
++#define pte_val(x) xen_pte_val(x)
++#define __pte(x) xen_make_pte(x)
++
++#define __pa(x) __phys_addr((unsigned long)(x))
++/* __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(__phys_reloc_hide((unsigned long)(x)))
++
++#define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET))
++
++#define __boot_va(x) __va(x)
++#define __boot_pa(x) __pa(x)
++
++#define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
++#define pfn_to_kaddr(pfn) __va((pfn) << PAGE_SHIFT)
++#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
++
++#endif /* __ASSEMBLY__ */
++
++#include <asm-generic/memory_model.h>
++#include <asm-generic/page.h>
++
++#define __HAVE_ARCH_GATE_AREA 1
++
++#endif /* __KERNEL__ */
++#endif /* _ASM_X86_PAGE_H */
+diff --git a/include/asm-x86/mach-xen/asm/page_64.h b/include/asm-x86/mach-xen/asm/page_64.h
+new file mode 100644
+index 0000000..461271e
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/page_64.h
+@@ -0,0 +1,101 @@
++#ifndef _X86_64_PAGE_H
++#define _X86_64_PAGE_H
++
++#define PAGETABLE_LEVELS 4
++
++#define THREAD_ORDER 1
++#define THREAD_SIZE (PAGE_SIZE << THREAD_ORDER)
++#define CURRENT_MASK (~(THREAD_SIZE - 1))
++
++#define EXCEPTION_STACK_ORDER 0
++#define EXCEPTION_STKSZ (PAGE_SIZE << EXCEPTION_STACK_ORDER)
++
++#define DEBUG_STACK_ORDER (EXCEPTION_STACK_ORDER + 1)
++#define DEBUG_STKSZ (PAGE_SIZE << DEBUG_STACK_ORDER)
++
++#define IRQSTACK_ORDER 2
++#define IRQSTACKSIZE (PAGE_SIZE << IRQSTACK_ORDER)
++
++#define STACKFAULT_STACK 1
++#define DOUBLEFAULT_STACK 2
++#define NMI_STACK 3
++#define DEBUG_STACK 4
++#define MCE_STACK 5
++#define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
++
++#define PUD_PAGE_SIZE (_AC(1, UL) << PUD_SHIFT)
++#define PUD_PAGE_MASK (~(PUD_PAGE_SIZE-1))
++
++#define __PAGE_OFFSET _AC(0xffff880000000000, UL)
++
++#define __PHYSICAL_START CONFIG_PHYSICAL_START
++#define __KERNEL_ALIGN 0x200000
++
++/*
++ * Make sure kernel is aligned to 2MB address. Catching it at compile
++ * time is better. Change your config file and compile the kernel
++ * for a 2MB aligned address (CONFIG_PHYSICAL_START)
++ */
++#if (CONFIG_PHYSICAL_START % __KERNEL_ALIGN) != 0
++#error "CONFIG_PHYSICAL_START must be a multiple of 2MB"
++#endif
++
++#define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START)
++#define __START_KERNEL_map _AC(0xffffffff80000000, UL)
++
++#if CONFIG_XEN_COMPAT <= 0x030002
++#undef LOAD_OFFSET
++#define LOAD_OFFSET 0
++#endif
++
++/* See Documentation/x86_64/mm.txt for a description of the memory map. */
++#define __PHYSICAL_MASK_SHIFT 46
++#define __VIRTUAL_MASK_SHIFT 48
++
++/*
++ * Kernel image size is limited to 512 MB (see level2_kernel_pgt in
++ * arch/x86/kernel/head_64.S), and it is mapped here:
++ */
++#define KERNEL_IMAGE_SIZE (512 * 1024 * 1024)
++#define KERNEL_IMAGE_START _AC(0xffffffff80000000, UL)
++
++#ifndef __ASSEMBLY__
++void clear_page(void *page);
++void copy_page(void *to, void *from);
++
++extern unsigned long end_pfn;
++
++static inline unsigned long __phys_addr(unsigned long x)
++{
++ return x - (x >= __START_KERNEL_map ? __START_KERNEL_map : __PAGE_OFFSET);
++}
++
++#define __phys_reloc_hide(x) (x)
++
++/*
++ * These are used to make use of C type-checking..
++ */
++typedef unsigned long pteval_t;
++typedef unsigned long pmdval_t;
++typedef unsigned long pudval_t;
++typedef unsigned long pgdval_t;
++typedef unsigned long pgprotval_t;
++typedef unsigned long phys_addr_t;
++
++typedef struct page *pgtable_t;
++
++typedef union { pteval_t pte; unsigned int pte_low; } pte_t;
++
++#define vmemmap ((struct page *)VMEMMAP_START)
++
++extern unsigned long init_memory_mapping(unsigned long start,
++ unsigned long end);
++
++#endif /* !__ASSEMBLY__ */
++
++#ifdef CONFIG_FLATMEM
++#define pfn_valid(pfn) ((pfn) < max_mapnr)
++#endif
++
++
++#endif /* _X86_64_PAGE_H */
+diff --git a/include/asm-x86/mach-xen/asm/pci.h b/include/asm-x86/mach-xen/asm/pci.h
+new file mode 100644
+index 0000000..5e1f7f5
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pci.h
+@@ -0,0 +1,118 @@
++#ifndef __x86_PCI_H
++#define __x86_PCI_H
++
++#include <linux/mm.h> /* for struct page */
++#include <linux/types.h>
++#include <linux/slab.h>
++#include <linux/string.h>
++#include <asm/scatterlist.h>
++#include <asm/io.h>
++
++#ifdef __KERNEL__
++
++struct pci_sysdata {
++ int domain; /* PCI domain */
++ int node; /* NUMA node */
++#ifdef CONFIG_X86_64
++ void *iommu; /* IOMMU private data */
++#endif
++#ifdef CONFIG_XEN_PCIDEV_FRONTEND
++ struct pcifront_device *pdev;
++#endif
++};
++
++/* scan a bus after allocating a pci_sysdata for it */
++extern struct pci_bus *pci_scan_bus_on_node(int busno, struct pci_ops *ops,
++ int node);
++extern struct pci_bus *pci_scan_bus_with_sysdata(int busno);
++
++static inline int pci_domain_nr(struct pci_bus *bus)
++{
++ struct pci_sysdata *sd = bus->sysdata;
++ return sd->domain;
++}
++
++static inline int pci_proc_domain(struct pci_bus *bus)
++{
++ return pci_domain_nr(bus);
++}
++
++extern void pci_iommu_alloc(void);
++
++/* 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
++
++#include <asm/hypervisor.h>
++#define pcibios_scan_all_fns(a, b) (!is_initial_xendomain())
++
++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);
++
++
++#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);
++
++
++#ifdef CONFIG_PCI
++extern void early_quirks(void);
++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;
++}
++#else
++static inline void early_quirks(void) { }
++#endif
++
++#endif /* __KERNEL__ */
++
++#ifdef CONFIG_X86_32
++# include "pci_32.h"
++#else
++# include "pci_64.h"
++#endif
++
++/* 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_NUMA
++/* Returns the node based on pci bus */
++static inline int __pcibus_to_node(struct pci_bus *bus)
++{
++ struct pci_sysdata *sd = bus->sysdata;
++
++ return sd->node;
++}
++
++static inline cpumask_t __pcibus_to_cpumask(struct pci_bus *bus)
++{
++ return node_to_cpumask(__pcibus_to_node(bus));
++}
++#endif
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/pci_32.h b/include/asm-x86/mach-xen/asm/pci_32.h
+new file mode 100644
+index 0000000..0b931eb
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pci_32.h
+@@ -0,0 +1,54 @@
++#ifndef __i386_PCI_H
++#define __i386_PCI_H
++
++
++#ifdef __KERNEL__
++
++/* Dynamic DMA mapping stuff.
++ * i386 has everything mapped statically.
++ */
++
++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
++
++
++#endif /* __KERNEL__ */
++
++
++#endif /* __i386_PCI_H */
+diff --git a/include/asm-x86/mach-xen/asm/pci_64.h b/include/asm-x86/mach-xen/asm/pci_64.h
+new file mode 100644
+index 0000000..cf97fdd
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pci_64.h
+@@ -0,0 +1,82 @@
++#ifndef __x8664_PCI_H
++#define __x8664_PCI_H
++
++#ifdef __KERNEL__
++
++#ifdef CONFIG_CALGARY_IOMMU
++static inline void *pci_iommu(struct pci_bus *bus)
++{
++ struct pci_sysdata *sd = bus->sysdata;
++ return sd->iommu;
++}
++
++static inline void set_pci_iommu(struct pci_bus *bus, void *val)
++{
++ struct pci_sysdata *sd = bus->sysdata;
++ sd->iommu = val;
++}
++#endif /* CONFIG_CALGARY_IOMMU */
++
++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);
++
++extern void dma32_reserve_bootmem(void);
++
++/* 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_GART_IOMMU) || defined(CONFIG_CALGARY_IOMMU)
++
++#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 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
++
++#endif /* __KERNEL__ */
++
++#endif /* __x8664_PCI_H */
+diff --git a/include/asm-x86/mach-xen/asm/pgalloc.h b/include/asm-x86/mach-xen/asm/pgalloc.h
+new file mode 100644
+index 0000000..5107718
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgalloc.h
+@@ -0,0 +1,149 @@
++#ifndef _ASM_X86_PGALLOC_H
++#define _ASM_X86_PGALLOC_H
++
++#include <linux/threads.h>
++#include <linux/mm.h> /* for struct page */
++#include <linux/pagemap.h>
++
++#include <asm/io.h> /* for phys_to_virt and page_to_pseudophys */
++
++static inline void paravirt_alloc_pte(struct mm_struct *mm, unsigned long pfn) {}
++static inline void paravirt_alloc_pmd(struct mm_struct *mm, unsigned long pfn) {}
++static inline void paravirt_alloc_pmd_clone(unsigned long pfn, unsigned long clonepfn,
++ unsigned long start, unsigned long count) {}
++static inline void paravirt_alloc_pud(struct mm_struct *mm, unsigned long pfn) {}
++static inline void paravirt_release_pte(unsigned long pfn) {}
++static inline void paravirt_release_pmd(unsigned long pfn) {}
++static inline void paravirt_release_pud(unsigned long pfn) {}
++
++#ifdef CONFIG_X86_64
++void early_make_page_readonly(void *va, unsigned int feature);
++pmd_t *early_get_pmd(unsigned long va);
++#define make_lowmem_page_readonly make_page_readonly
++#define make_lowmem_page_writable make_page_writable
++#endif
++
++/*
++ * Allocate and free page tables.
++ */
++extern pgd_t *pgd_alloc(struct mm_struct *);
++extern void pgd_free(struct mm_struct *mm, pgd_t *pgd);
++
++extern pte_t *pte_alloc_one_kernel(struct mm_struct *, unsigned long);
++extern pgtable_t pte_alloc_one(struct mm_struct *, unsigned long);
++
++/* Should really implement gc for free page table pages. This could be
++ done with a reference count in struct page. */
++
++static inline void pte_free_kernel(struct mm_struct *mm, pte_t *pte)
++{
++ BUG_ON((unsigned long)pte & (PAGE_SIZE-1));
++ make_lowmem_page_writable(pte, XENFEAT_writable_page_tables);
++ free_page((unsigned long)pte);
++}
++
++extern void __pte_free(pgtable_t);
++static inline void pte_free(struct mm_struct *mm, struct page *pte)
++{
++ __pte_free(pte);
++}
++
++extern void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte);
++
++static inline void pmd_populate_kernel(struct mm_struct *mm,
++ pmd_t *pmd, pte_t *pte)
++{
++ paravirt_alloc_pte(mm, __pa(pte) >> PAGE_SHIFT);
++ set_pmd(pmd, __pmd(__pa(pte) | _PAGE_TABLE));
++}
++
++static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd,
++ struct page *pte)
++{
++ unsigned long pfn = page_to_pfn(pte);
++
++ paravirt_alloc_pte(mm, pfn);
++ if (PagePinned(virt_to_page(mm->pgd))) {
++ if (!PageHighMem(pte))
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)__va(pfn << PAGE_SHIFT),
++ pfn_pte(pfn, PAGE_KERNEL_RO), 0));
++#ifndef CONFIG_X86_64
++ else if (!TestSetPagePinned(pte))
++ kmap_flush_unused();
++#endif
++ set_pmd(pmd, __pmd(((pmdval_t)pfn << PAGE_SHIFT) | _PAGE_TABLE));
++ } else
++ *pmd = __pmd(((pmdval_t)pfn << PAGE_SHIFT) | _PAGE_TABLE);
++}
++
++#define pmd_pgtable(pmd) pmd_page(pmd)
++
++#if PAGETABLE_LEVELS > 2
++extern pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr);
++extern void __pmd_free(pgtable_t);
++
++static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd)
++{
++ BUG_ON((unsigned long)pmd & (PAGE_SIZE-1));
++ __pmd_free(virt_to_page(pmd));
++}
++
++extern void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd);
++
++#ifdef CONFIG_X86_PAE
++extern void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd);
++#else /* !CONFIG_X86_PAE */
++static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
++{
++ paravirt_alloc_pmd(mm, __pa(pmd) >> PAGE_SHIFT);
++ if (unlikely(PagePinned(virt_to_page((mm)->pgd)))) {
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pmd,
++ pfn_pte(virt_to_phys(pmd)>>PAGE_SHIFT,
++ PAGE_KERNEL_RO), 0));
++ set_pud(pud, __pud(_PAGE_TABLE | __pa(pmd)));
++ } else
++ *pud = __pud(_PAGE_TABLE | __pa(pmd));
++}
++#endif /* CONFIG_X86_PAE */
++
++#if PAGETABLE_LEVELS > 3
++#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
++
++/*
++ * We need to use the batch mode here, but pgd_pupulate() won't be
++ * be called frequently.
++ */
++static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
++{
++ paravirt_alloc_pud(mm, __pa(pud) >> PAGE_SHIFT);
++ if (unlikely(PagePinned(virt_to_page((mm)->pgd)))) {
++ BUG_ON(HYPERVISOR_update_va_mapping(
++ (unsigned long)pud,
++ pfn_pte(virt_to_phys(pud)>>PAGE_SHIFT,
++ PAGE_KERNEL_RO), 0));
++ set_pgd(pgd, __pgd(_PAGE_TABLE | __pa(pud)));
++ set_pgd(__user_pgd(pgd), __pgd(_PAGE_TABLE | __pa(pud)));
++ } else {
++ *(pgd) = __pgd(_PAGE_TABLE | __pa(pud));
++ *__user_pgd(pgd) = *(pgd);
++ }
++}
++
++static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
++{
++ return (pud_t *)pmd_alloc_one(mm, addr);
++}
++
++static inline void pud_free(struct mm_struct *mm, pud_t *pud)
++{
++ BUG_ON((unsigned long)pud & (PAGE_SIZE-1));
++ __pmd_free(virt_to_page(pud));
++}
++
++extern void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pud);
++#endif /* PAGETABLE_LEVELS > 3 */
++#endif /* PAGETABLE_LEVELS > 2 */
++
++#endif /* _ASM_X86_PGALLOC_H */
+diff --git a/include/asm-x86/mach-xen/asm/pgtable-3level-defs.h b/include/asm-x86/mach-xen/asm/pgtable-3level-defs.h
+new file mode 100644
+index 0000000..5bd9cd7
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgtable-3level-defs.h
+@@ -0,0 +1,24 @@
++#ifndef _I386_PGTABLE_3LEVEL_DEFS_H
++#define _I386_PGTABLE_3LEVEL_DEFS_H
++
++#define 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
++
++/*
++ * 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 --git a/include/asm-x86/mach-xen/asm/pgtable-3level.h b/include/asm-x86/mach-xen/asm/pgtable-3level.h
+new file mode 100644
+index 0000000..b1eb2b8
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgtable-3level.h
+@@ -0,0 +1,168 @@
++#ifndef _I386_PGTABLE_3LEVEL_H
++#define _I386_PGTABLE_3LEVEL_H
++
++/*
++ * Intel Physical Address Extension (PAE) Mode - three-level page
++ * tables on PPro+ CPUs.
++ *
++ * Copyright (C) 1999 Ingo Molnar <mingo at redhat.com>
++ */
++
++#define pte_ERROR(e) \
++ printk("%s:%d: bad pte %p(%016Lx pfn %08lx).\n", \
++ __FILE__, __LINE__, &(e), __pte_val(e), pte_pfn(e))
++#define pmd_ERROR(e) \
++ printk("%s:%d: bad pmd %p(%016Lx pfn %08Lx).\n", \
++ __FILE__, __LINE__, &(e), __pmd_val(e), \
++ (pmd_val(e) & PTE_MASK) >> PAGE_SHIFT)
++#define pgd_ERROR(e) \
++ printk("%s:%d: bad pgd %p(%016Lx pfn %08Lx).\n", \
++ __FILE__, __LINE__, &(e), __pgd_val(e), \
++ (pgd_val(e) & PTE_MASK) >> PAGE_SHIFT)
++
++static inline int pud_none(pud_t pud)
++{
++ return __pud_val(pud) == 0;
++
++}
++static inline int pud_bad(pud_t pud)
++{
++ return (__pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER)) != 0;
++}
++
++static inline int pud_present(pud_t pud)
++{
++ return __pud_val(pud) & _PAGE_PRESENT;
++}
++
++/* Rules for using set_pte: the pte being assigned *must* be
++ * either not present or in a state where the hardware will
++ * not attempt to update the pte. In places where this is
++ * not possible, use pte_get_and_clear to obtain the old pte
++ * value and then use set_pte to update it. -ben
++ */
++
++static inline void xen_set_pte(pte_t *ptep, pte_t pte)
++{
++ ptep->pte_high = pte.pte_high;
++ smp_wmb();
++ ptep->pte_low = pte.pte_low;
++}
++
++static inline void xen_set_pte_atomic(pte_t *ptep, pte_t pte)
++{
++ set_64bit((unsigned long long *)(ptep), __pte_val(pte));
++}
++
++static inline void xen_set_pmd(pmd_t *pmdp, pmd_t pmd)
++{
++ xen_l2_entry_update(pmdp, pmd);
++}
++
++static inline void xen_set_pud(pud_t *pudp, pud_t pud)
++{
++ xen_l3_entry_update(pudp, pud);
++}
++
++/*
++ * For PTEs and PDEs, we must clear the P-bit first when clearing a page table
++ * entry, so clear the bottom half first and enforce ordering with a compiler
++ * barrier.
++ */
++static inline void __xen_pte_clear(pte_t *ptep)
++{
++ ptep->pte_low = 0;
++ smp_wmb();
++ ptep->pte_high = 0;
++}
++
++static inline void xen_pmd_clear(pmd_t *pmd)
++{
++ xen_l2_entry_update(pmd, __pmd(0));
++}
++
++static inline void pud_clear(pud_t *pudp)
++{
++ pgdval_t pgd;
++
++ set_pud(pudp, __pud(0));
++
++ /*
++ * According to Intel App note "TLBs, Paging-Structure Caches,
++ * and Their Invalidation", April 2007, document 317080-001,
++ * section 8.1: in PAE mode we explicitly have to flush the
++ * TLB via cr3 if the top-level pgd is changed...
++ *
++ * Make sure the pud entry we're updating is within the
++ * current pgd to avoid unnecessary TLB flushes.
++ */
++ pgd = read_cr3();
++ if (__pa(pudp) >= pgd && __pa(pudp) <
++ (pgd + sizeof(pgd_t)*PTRS_PER_PGD))
++ xen_tlb_flush();
++}
++
++#define pud_page(pud) ((struct page *) __va(pud_val(pud) & PTE_MASK))
++
++#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PTE_MASK))
++
++
++/* Find an entry in the second-level page table.. */
++#define pmd_offset(pud, address) ((pmd_t *)pud_page(*(pud)) + \
++ pmd_index(address))
++
++#ifdef CONFIG_SMP
++static inline pte_t xen_ptep_get_and_clear(pte_t *ptep, pte_t res)
++{
++ uint64_t val = __pte_val(res);
++ if (__cmpxchg64(ptep, val, 0) != val) {
++ /* xchg acts as a barrier before the setting of the high bits */
++ res.pte_low = xchg(&ptep->pte_low, 0);
++ res.pte_high = ptep->pte_high;
++ ptep->pte_high = 0;
++ }
++ return res;
++}
++#else
++#define xen_ptep_get_and_clear(xp, pte) xen_local_ptep_get_and_clear(xp, pte)
++#endif
++
++#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;
++}
++
++#define pte_page(x) pfn_to_page(pte_pfn(x))
++
++static inline int pte_none(pte_t pte)
++{
++ return !(pte.pte_low | pte.pte_high);
++}
++
++#define __pte_mfn(_pte) (((_pte).pte_low >> PAGE_SHIFT) | \
++ ((_pte).pte_high << (32-PAGE_SHIFT)))
++#define pte_mfn(_pte) ((_pte).pte_low & _PAGE_PRESENT ? \
++ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
++#define pte_pfn(_pte) ((_pte).pte_low & _PAGE_IO ? max_mapnr : \
++ (_pte).pte_low & _PAGE_PRESENT ? \
++ mfn_to_local_pfn(__pte_mfn(_pte)) : \
++ __pte_mfn(_pte))
++
++/*
++ * Bits 0, 6 and 7 are taken in the low part of the pte,
++ * put the 32 bits of offset into the high part.
++ */
++#define pte_to_pgoff(pte) ((pte).pte_high)
++#define pgoff_to_pte(off) \
++ ((pte_t) { { .pte_low = _PAGE_FILE, .pte_high = (off) } })
++#define PTE_FILE_MAX_BITS 32
++
++/* Encode and de-code a swap entry */
++#define __swp_type(x) (((x).val) & 0x1f)
++#define __swp_offset(x) ((x).val >> 5)
++#define __swp_entry(type, offset) ((swp_entry_t){(type) | (offset) << 5})
++#define __pte_to_swp_entry(pte) ((swp_entry_t){ (pte).pte_high })
++#define __swp_entry_to_pte(x) ((pte_t){ { .pte_high = (x).val } })
++
++#endif /* _I386_PGTABLE_3LEVEL_H */
+diff --git a/include/asm-x86/mach-xen/asm/pgtable.h b/include/asm-x86/mach-xen/asm/pgtable.h
+new file mode 100644
+index 0000000..a9ff073
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgtable.h
+@@ -0,0 +1,568 @@
++#ifndef _ASM_X86_PGTABLE_H
++#define _ASM_X86_PGTABLE_H
++
++#define FIRST_USER_ADDRESS 0
++
++#define _PAGE_BIT_PRESENT 0 /* is present */
++#define _PAGE_BIT_RW 1 /* writeable */
++#define _PAGE_BIT_USER 2 /* userspace addressable */
++#define _PAGE_BIT_PWT 3 /* page write through */
++#define _PAGE_BIT_PCD 4 /* page cache disabled */
++#define _PAGE_BIT_ACCESSED 5 /* was accessed (raised by CPU) */
++#define _PAGE_BIT_DIRTY 6 /* was written to (raised by CPU) */
++#define _PAGE_BIT_FILE 6
++#define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */
++#define _PAGE_BIT_PAT 7 /* on 4KB pages */
++#define _PAGE_BIT_GLOBAL 8 /* Global TLB entry PPro+ */
++#define _PAGE_BIT_IO 9 /* Mapped page is I/O or foreign and
++ * has no associated page struct. */
++#define _PAGE_BIT_UNUSED2 10 /* available for programmer */
++#define _PAGE_BIT_UNUSED3 11
++#define _PAGE_BIT_PAT_LARGE 12 /* On 2MB or 1GB pages */
++#define _PAGE_BIT_NX 63 /* No execute: only valid after cpuid check */
++
++/*
++ * Note: we use _AC(1, L) instead of _AC(1, UL) so that we get a
++ * sign-extended value on 32-bit with all 1's in the upper word,
++ * which preserves the upper pte values on 64-bit ptes:
++ */
++#define _PAGE_PRESENT (_AC(1, L)<<_PAGE_BIT_PRESENT)
++#define _PAGE_RW (_AC(1, L)<<_PAGE_BIT_RW)
++#define _PAGE_USER (_AC(1, L)<<_PAGE_BIT_USER)
++#define _PAGE_PWT (_AC(1, L)<<_PAGE_BIT_PWT)
++#define _PAGE_PCD (_AC(1, L)<<_PAGE_BIT_PCD)
++#define _PAGE_ACCESSED (_AC(1, L)<<_PAGE_BIT_ACCESSED)
++#define _PAGE_DIRTY (_AC(1, L)<<_PAGE_BIT_DIRTY)
++#define _PAGE_PSE (_AC(1, L)<<_PAGE_BIT_PSE) /* 2MB page */
++#define _PAGE_GLOBAL (_AC(1, L)<<_PAGE_BIT_GLOBAL) /* Global TLB entry */
++#define _PAGE_IO (_AC(1, L)<<_PAGE_BIT_IO)
++#define _PAGE_UNUSED2 (_AC(1, L)<<_PAGE_BIT_UNUSED2)
++#define _PAGE_UNUSED3 (_AC(1, L)<<_PAGE_BIT_UNUSED3)
++#define _PAGE_PAT (_AC(1, L)<<_PAGE_BIT_PAT)
++#define _PAGE_PAT_LARGE (_AC(1, L)<<_PAGE_BIT_PAT_LARGE)
++
++#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
++#define _PAGE_NX (_AC(1, ULL) << _PAGE_BIT_NX)
++#else
++#define _PAGE_NX 0
++#endif
++
++/* If _PAGE_PRESENT is clear, we use these: */
++#define _PAGE_FILE _PAGE_DIRTY /* nonlinear file mapping,
++ * saved PTE; unset:swap */
++#define _PAGE_PROTNONE _PAGE_PSE /* if the user mapped it with PROT_NONE;
++ pte_present gives true */
++
++#ifndef __ASSEMBLY__
++#if defined(CONFIG_X86_64) && CONFIG_XEN_COMPAT <= 0x030002
++extern unsigned int __kernel_page_user;
++#else
++#define __kernel_page_user 0
++#endif
++#endif
++
++#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | \
++ _PAGE_ACCESSED | _PAGE_DIRTY)
++#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | \
++ _PAGE_DIRTY | __kernel_page_user)
++
++/* Set of bits not changed in pte_modify */
++#define _PAGE_CHG_MASK (PTE_MASK | _PAGE_CACHE_MASK | _PAGE_IO | \
++ _PAGE_ACCESSED | _PAGE_DIRTY)
++
++/*
++ * PAT settings are part of the hypervisor interface, which sets the
++ * MSR to 0x050100070406 (i.e. WB, WT, UC-, UC, WC, WP [, UC, UC]).
++ */
++#define _PAGE_CACHE_MASK (_PAGE_PCD | _PAGE_PWT | _PAGE_PAT)
++#define _PAGE_CACHE_WB (0)
++#define _PAGE_CACHE_WT (_PAGE_PWT)
++#define _PAGE_CACHE_WC (_PAGE_PAT)
++#define _PAGE_CACHE_WP (_PAGE_PAT | _PAGE_PWT)
++#define _PAGE_CACHE_UC_MINUS (_PAGE_PCD)
++#define _PAGE_CACHE_UC (_PAGE_PCD | _PAGE_PWT)
++
++#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
++#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | \
++ _PAGE_ACCESSED | _PAGE_NX)
++
++#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | \
++ _PAGE_USER | _PAGE_ACCESSED)
++#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \
++ _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \
++ _PAGE_ACCESSED)
++#define PAGE_COPY PAGE_COPY_NOEXEC
++#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | \
++ _PAGE_ACCESSED | _PAGE_NX)
++#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \
++ _PAGE_ACCESSED)
++
++#ifdef CONFIG_X86_32
++#define _PAGE_KERNEL_EXEC \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED)
++#define _PAGE_KERNEL (_PAGE_KERNEL_EXEC | _PAGE_NX)
++
++#ifndef __ASSEMBLY__
++extern pteval_t __PAGE_KERNEL, __PAGE_KERNEL_EXEC;
++#endif /* __ASSEMBLY__ */
++#else
++#define __PAGE_KERNEL_EXEC \
++ (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | __kernel_page_user)
++#define __PAGE_KERNEL (__PAGE_KERNEL_EXEC | _PAGE_NX)
++#endif
++
++#define __PAGE_KERNEL_RO (__PAGE_KERNEL & ~_PAGE_RW)
++#define __PAGE_KERNEL_RX (__PAGE_KERNEL_EXEC & ~_PAGE_RW)
++#define __PAGE_KERNEL_EXEC_NOCACHE (__PAGE_KERNEL_EXEC | _PAGE_PCD | _PAGE_PWT)
++#define __PAGE_KERNEL_WC (__PAGE_KERNEL | _PAGE_CACHE_WC)
++#define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_PCD | _PAGE_PWT)
++#define __PAGE_KERNEL_UC_MINUS (__PAGE_KERNEL | _PAGE_PCD)
++#define __PAGE_KERNEL_VSYSCALL (__PAGE_KERNEL_RX | _PAGE_USER)
++#define __PAGE_KERNEL_VSYSCALL_NOCACHE (__PAGE_KERNEL_VSYSCALL | _PAGE_PCD | _PAGE_PWT)
++#define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE)
++#define __PAGE_KERNEL_LARGE_EXEC (__PAGE_KERNEL_EXEC | _PAGE_PSE)
++
++/*
++ * We don't support GLOBAL page in xenolinux64
++ */
++#define MAKE_GLOBAL(x) __pgprot((x))
++
++#define PAGE_KERNEL MAKE_GLOBAL(__PAGE_KERNEL)
++#define PAGE_KERNEL_RO MAKE_GLOBAL(__PAGE_KERNEL_RO)
++#define PAGE_KERNEL_EXEC MAKE_GLOBAL(__PAGE_KERNEL_EXEC)
++#define PAGE_KERNEL_RX MAKE_GLOBAL(__PAGE_KERNEL_RX)
++#define PAGE_KERNEL_WC MAKE_GLOBAL(__PAGE_KERNEL_WC)
++#define PAGE_KERNEL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_NOCACHE)
++#define PAGE_KERNEL_UC_MINUS MAKE_GLOBAL(__PAGE_KERNEL_UC_MINUS)
++#define PAGE_KERNEL_EXEC_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_EXEC_NOCACHE)
++#define PAGE_KERNEL_LARGE MAKE_GLOBAL(__PAGE_KERNEL_LARGE)
++#define PAGE_KERNEL_LARGE_EXEC MAKE_GLOBAL(__PAGE_KERNEL_LARGE_EXEC)
++#define PAGE_KERNEL_VSYSCALL MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL)
++#define PAGE_KERNEL_VSYSCALL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_VSYSCALL_NOCACHE)
++
++/* xwr */
++#define __P000 PAGE_NONE
++#define __P001 PAGE_READONLY
++#define __P010 PAGE_COPY
++#define __P011 PAGE_COPY
++#define __P100 PAGE_READONLY_EXEC
++#define __P101 PAGE_READONLY_EXEC
++#define __P110 PAGE_COPY_EXEC
++#define __P111 PAGE_COPY_EXEC
++
++#define __S000 PAGE_NONE
++#define __S001 PAGE_READONLY
++#define __S010 PAGE_SHARED
++#define __S011 PAGE_SHARED
++#define __S100 PAGE_READONLY_EXEC
++#define __S101 PAGE_READONLY_EXEC
++#define __S110 PAGE_SHARED_EXEC
++#define __S111 PAGE_SHARED_EXEC
++
++#ifndef __ASSEMBLY__
++
++/*
++ * ZERO_PAGE is a global shared page that is always zero: used
++ * for zero-mapped memory areas etc..
++ */
++extern unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)];
++#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
++
++extern spinlock_t pgd_lock;
++extern struct list_head pgd_list;
++
++/*
++ * The following only work if pte_present() is true.
++ * Undefined behaviour if not..
++ */
++static inline int pte_dirty(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_DIRTY;
++}
++
++static inline int pte_young(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_ACCESSED;
++}
++
++static inline int pte_write(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_RW;
++}
++
++static inline int pte_file(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_FILE;
++}
++
++static inline int pte_huge(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_PSE;
++}
++
++static inline int pte_global(pte_t pte)
++{
++ return __pte_val(pte) & _PAGE_GLOBAL;
++}
++
++static inline int pte_exec(pte_t pte)
++{
++ return !(__pte_val(pte) & _PAGE_NX);
++}
++
++static inline int pte_special(pte_t pte)
++{
++ return 0;
++}
++
++static inline int pmd_large(pmd_t pte)
++{
++ return (__pmd_val(pte) & (_PAGE_PSE | _PAGE_PRESENT)) ==
++ (_PAGE_PSE | _PAGE_PRESENT);
++}
++
++static inline pte_t pte_mkclean(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_DIRTY);
++}
++
++static inline pte_t pte_mkold(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_ACCESSED);
++}
++
++static inline pte_t pte_wrprotect(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_RW);
++}
++
++static inline pte_t pte_mkexec(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_NX);
++}
++
++static inline pte_t pte_mkdirty(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) | _PAGE_DIRTY);
++}
++
++static inline pte_t pte_mkyoung(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) | _PAGE_ACCESSED);
++}
++
++static inline pte_t pte_mkwrite(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) | _PAGE_RW);
++}
++
++static inline pte_t pte_mkhuge(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) | _PAGE_PSE);
++}
++
++static inline pte_t pte_clrhuge(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_PSE);
++}
++
++static inline pte_t pte_mkglobal(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) | _PAGE_GLOBAL);
++}
++
++static inline pte_t pte_clrglobal(pte_t pte)
++{
++ return __pte_ma(__pte_val(pte) & ~(pteval_t)_PAGE_GLOBAL);
++}
++
++static inline pte_t pte_mkspecial(pte_t pte)
++{
++ return pte;
++}
++
++extern pteval_t __supported_pte_mask;
++
++static inline pte_t pfn_pte(unsigned long page_nr, pgprot_t pgprot)
++{
++ return __pte((((phys_addr_t)page_nr << PAGE_SHIFT) |
++ pgprot_val(pgprot)) & __supported_pte_mask);
++}
++
++static inline pte_t pfn_pte_ma(unsigned long page_nr, pgprot_t pgprot)
++{
++ return __pte_ma((((phys_addr_t)page_nr << PAGE_SHIFT) |
++ pgprot_val(pgprot)) & __supported_pte_mask);
++}
++
++static inline pmd_t pfn_pmd(unsigned long page_nr, pgprot_t pgprot)
++{
++ return __pmd((((phys_addr_t)page_nr << PAGE_SHIFT) |
++ pgprot_val(pgprot)) & __supported_pte_mask);
++}
++
++static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
++{
++ pteval_t val = pte_val(pte);
++
++ val &= _PAGE_CHG_MASK;
++ val |= pgprot_val(newprot) & (~_PAGE_CHG_MASK) & __supported_pte_mask;
++
++ return __pte(val);
++}
++
++/* mprotect needs to preserve PAT bits when updating vm_page_prot */
++#define pgprot_modify pgprot_modify
++static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot)
++{
++ pgprotval_t preservebits = pgprot_val(oldprot) & _PAGE_CHG_MASK;
++ pgprotval_t addbits = pgprot_val(newprot);
++ return __pgprot(preservebits | addbits);
++}
++
++#define pte_pgprot(x) __pgprot(__pte_val(x) & ~PTE_MASK)
++
++#define canon_pgprot(p) __pgprot(pgprot_val(p) & __supported_pte_mask)
++
++#ifndef __ASSEMBLY__
++#define __HAVE_PHYS_MEM_ACCESS_PROT
++struct file;
++pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
++ unsigned long size, pgprot_t vma_prot);
++int phys_mem_access_prot_allowed(struct file *file, unsigned long pfn,
++ unsigned long size, pgprot_t *vma_prot);
++#endif
++
++#define set_pte(ptep, pte) xen_set_pte(ptep, pte)
++#define set_pte_at(mm, addr, ptep, pte) xen_set_pte_at(mm, addr, ptep, pte)
++
++#define set_pte_atomic(ptep, pte) \
++ xen_set_pte_atomic(ptep, pte)
++
++#define set_pmd(pmdp, pmd) xen_set_pmd(pmdp, pmd)
++
++#ifndef __PAGETABLE_PUD_FOLDED
++#define set_pgd(pgdp, pgd) xen_set_pgd(pgdp, pgd)
++#define pgd_clear(pgd) xen_pgd_clear(pgd)
++#endif
++
++#ifndef set_pud
++# define set_pud(pudp, pud) xen_set_pud(pudp, pud)
++#endif
++
++#ifndef __PAGETABLE_PMD_FOLDED
++#define pud_clear(pud) xen_pud_clear(pud)
++#endif
++
++#define pte_clear(mm, addr, ptep) xen_pte_clear(mm, addr, ptep)
++#define pmd_clear(pmd) xen_pmd_clear(pmd)
++
++#define pte_update(mm, addr, ptep) do { } while (0)
++#define pte_update_defer(mm, addr, ptep) do { } while (0)
++
++#endif /* __ASSEMBLY__ */
++
++#ifdef CONFIG_X86_32
++# include "pgtable_32.h"
++#else
++# include "pgtable_64.h"
++#endif
++
++#define KERNEL_PGD_BOUNDARY pgd_index(PAGE_OFFSET)
++#define KERNEL_PGD_PTRS (PTRS_PER_PGD - KERNEL_PGD_BOUNDARY)
++
++#ifndef __ASSEMBLY__
++
++enum {
++ PG_LEVEL_NONE,
++ PG_LEVEL_4K,
++ PG_LEVEL_2M,
++ PG_LEVEL_1G,
++};
++
++/*
++ * Helper function that returns the kernel pagetable entry controlling
++ * the virtual address 'address'. NULL means no pagetable entry present.
++ * NOTE: the return type is pte_t but if the pmd is PSE then we return it
++ * as a pte too.
++ */
++extern pte_t *lookup_address(unsigned long address, unsigned int *level);
++
++/* local pte updates need not use xchg for locking */
++static inline pte_t xen_local_ptep_get_and_clear(pte_t *ptep, pte_t res)
++{
++ xen_set_pte(ptep, __pte(0));
++ return res;
++}
++
++static inline void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
++ pte_t *ptep , pte_t pte)
++{
++ if ((mm != current->mm && mm != &init_mm) ||
++ HYPERVISOR_update_va_mapping(addr, pte, 0))
++ xen_set_pte(ptep, pte);
++}
++
++static inline void xen_pte_clear(struct mm_struct *mm, unsigned long addr,
++ pte_t *ptep)
++{
++ if ((mm != current->mm && mm != &init_mm)
++ || HYPERVISOR_update_va_mapping(addr, __pte(0), 0))
++ __xen_pte_clear(ptep);
++}
++
++#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
++
++/*
++ * 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
++extern int ptep_set_access_flags(struct vm_area_struct *vma,
++ unsigned long address, pte_t *ptep,
++ pte_t entry, int dirty);
++
++#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
++extern int ptep_test_and_clear_young(struct vm_area_struct *vma,
++ unsigned long addr, pte_t *ptep);
++
++#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
++extern int ptep_clear_flush_young(struct vm_area_struct *vma,
++ unsigned long address, pte_t *ptep);
++
++#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
++#define ptep_clear_flush(vma, addr, ptep) \
++({ \
++ pte_t *__ptep = (ptep); \
++ pte_t __res = *__ptep; \
++ if (!pte_none(__res) && \
++ ((vma)->vm_mm != current->mm || \
++ HYPERVISOR_update_va_mapping(addr, __pte(0), \
++ (unsigned long)(vma)->vm_mm->cpu_vm_mask.bits| \
++ UVMF_INVLPG|UVMF_MULTI))) { \
++ __xen_pte_clear(__ptep); \
++ flush_tlb_page(vma, addr); \
++ } \
++ __res; \
++})
++
++#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)
++{
++ pte_t pte = *ptep;
++ if (!pte_none(pte)
++ && (mm != &init_mm
++ || HYPERVISOR_update_va_mapping(addr, __pte(0), 0))) {
++ pte = xen_ptep_get_and_clear(ptep, pte);
++ pte_update(mm, addr, ptep);
++ }
++ return pte;
++}
++
++#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
++#define ptep_get_and_clear_full(mm, addr, ptep, full) \
++ ((full) ? ({ \
++ pte_t *__ptep = (ptep); \
++ pte_t __res = *__ptep; \
++ if (!PagePinned(virt_to_page((mm)->pgd))) \
++ __xen_pte_clear(__ptep); \
++ else if (!pte_none(__res)) \
++ xen_l1_entry_update(__ptep, __pte(0)); \
++ __res; \
++ }) : \
++ ptep_get_and_clear(mm, addr, ptep))
++
++pte_t xen_ptep_get_and_clear_full(struct vm_area_struct *, unsigned long, pte_t *, int);
++
++#define __HAVE_ARCH_PTEP_SET_WRPROTECT
++static inline void ptep_set_wrprotect(struct mm_struct *mm,
++ unsigned long addr, pte_t *ptep)
++{
++ pte_t pte = *ptep;
++ if (pte_write(pte))
++ set_pte_at(mm, addr, ptep, pte_wrprotect(pte));
++}
++
++/*
++ * clone_pgd_range(pgd_t *dst, pgd_t *src, int count);
++ *
++ * dst - pointer to pgd range anwhere on a pgd page
++ * src - ""
++ * count - the number of pgds to copy.
++ *
++ * dst and src can be on the same page, but the range must not overlap,
++ * and must not cross a page boundary.
++ */
++static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count)
++{
++ memcpy(dst, src, count * sizeof(pgd_t));
++}
++
++#define arch_change_pte_range(mm, pmd, addr, end, newprot, dirty_accountable) \
++ xen_change_pte_range(mm, pmd, addr, end, newprot, dirty_accountable)
++
++#define arbitrary_virt_to_machine(va) \
++({ \
++ unsigned int __lvl; \
++ pte_t *__ptep = lookup_address((unsigned long)(va), &__lvl); \
++ BUG_ON(!__ptep || __lvl != PG_LEVEL_4K || !pte_present(*__ptep));\
++ (((maddr_t)pte_mfn(*__ptep) << PAGE_SHIFT) \
++ | ((unsigned long)(va) & (PAGE_SIZE - 1))); \
++})
++
++#include <asm-generic/pgtable.h>
++
++#include <xen/features.h>
++void make_page_readonly(void *va, unsigned int feature);
++void make_page_writable(void *va, unsigned int feature);
++void make_pages_readonly(void *va, unsigned int nr, unsigned int feature);
++void make_pages_writable(void *va, unsigned int nr, unsigned int feature);
++
++struct vm_area_struct;
++
++int direct_remap_pfn_range(struct vm_area_struct *vma,
++ unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
++int direct_kernel_remap_pfn_range(unsigned long address,
++ unsigned long mfn,
++ unsigned long size,
++ pgprot_t prot,
++ domid_t domid);
++int create_lookup_pte_addr(struct mm_struct *mm,
++ unsigned long address,
++ uint64_t *ptep);
++int touch_pte_range(struct mm_struct *mm,
++ unsigned long address,
++ unsigned long size);
++
++int xen_change_pte_range(struct mm_struct *mm, pmd_t *pmd,
++ unsigned long addr, unsigned long end, pgprot_t newprot,
++ int dirty_accountable);
++
++#endif /* __ASSEMBLY__ */
++
++#endif /* _ASM_X86_PGTABLE_H */
+diff --git a/include/asm-x86/mach-xen/asm/pgtable_32.h b/include/asm-x86/mach-xen/asm/pgtable_32.h
+new file mode 100644
+index 0000000..db6416c
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgtable_32.h
+@@ -0,0 +1,220 @@
++#ifndef _I386_PGTABLE_H
++#define _I386_PGTABLE_H
++
++#include <asm/hypervisor.h>
++
++/*
++ * The Linux memory management assumes a three-level page table setup. On
++ * the i386, we use that, but "fold" the mid level into the top-level page
++ * table, so that we physically have the same two-level page table as the
++ * i386 mmu expects.
++ *
++ * This file contains the functions and defines necessary to modify and use
++ * the i386 page table tree.
++ */
++#ifndef __ASSEMBLY__
++#include <asm/processor.h>
++#include <asm/fixmap.h>
++#include <linux/threads.h>
++
++#include <linux/bitops.h>
++#include <linux/slab.h>
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/sched.h>
++
++struct vm_area_struct;
++
++extern pgd_t *swapper_pg_dir;
++
++static inline void pgtable_cache_init(void) { }
++static inline void check_pgt_cache(void) { }
++void paging_init(void);
++
++
++/*
++ * The Linux x86 paging architecture is 'compile-time dual-mode', it
++ * implements both the traditional 2-level x86 page tables and the
++ * newer 3-level PAE-mode page tables.
++ */
++#ifdef CONFIG_X86_PAE
++# include <asm/pgtable-3level-defs.h>
++# define PMD_SIZE (1UL << PMD_SHIFT)
++# define PMD_MASK (~(PMD_SIZE - 1))
++#else
++# include <asm/pgtable-2level-defs.h>
++#endif
++
++#define PGDIR_SIZE (1UL << PGDIR_SHIFT)
++#define PGDIR_MASK (~(PGDIR_SIZE - 1))
++
++/* Just any arbitrary offset to the start of the vmalloc VM area: the
++ * current 8MB value just means that there will be a 8MB "hole" after the
++ * physical memory until the kernel virtual memory starts. That means that
++ * any out-of-bounds memory accesses will hopefully be caught.
++ * The vmalloc() routines leaves a hole of 4kB between each vmalloced
++ * area for the same reason. ;)
++ */
++#define VMALLOC_OFFSET (8 * 1024 * 1024)
++#define VMALLOC_START (((unsigned long)high_memory + 2 * VMALLOC_OFFSET - 1) \
++ & ~(VMALLOC_OFFSET - 1))
++#ifdef CONFIG_X86_PAE
++#define LAST_PKMAP 512
++#else
++#define LAST_PKMAP 1024
++#endif
++
++#define PKMAP_BASE ((FIXADDR_BOOT_START - PAGE_SIZE * (LAST_PKMAP + 1)) \
++ & PMD_MASK)
++
++#ifdef CONFIG_HIGHMEM
++# define VMALLOC_END (PKMAP_BASE - 2 * PAGE_SIZE)
++#else
++# define VMALLOC_END (FIXADDR_START - 2 * PAGE_SIZE)
++#endif
++
++/*
++ * Define this if things work differently on an i386 and an i486:
++ * it will (on an i486) warn about kernel memory accesses that are
++ * done without a 'access_ok(VERIFY_WRITE,..)'
++ */
++#undef TEST_ACCESS_OK
++
++/* The boot page tables (all created as a single array) */
++extern unsigned long pg0[];
++
++#define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE))
++
++/* To avoid harmful races, pmd_none(x) should check only the lower when PAE */
++#define pmd_none(x) (!(unsigned long)__pmd_val(x))
++#if CONFIG_XEN_COMPAT <= 0x030002
++/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
++ can temporarily clear it. */
++#define pmd_present(x) (__pmd_val(x))
++#define pmd_bad(x) ((__pmd_val(x) & (~PTE_MASK & ~_PAGE_USER & ~_PAGE_PRESENT)) != (_KERNPG_TABLE & ~_PAGE_PRESENT))
++#else
++#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
++#define pmd_bad(x) ((__pmd_val(x) & (~PTE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE)
++#endif
++
++
++#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
++
++#ifdef CONFIG_X86_PAE
++# include <asm/pgtable-3level.h>
++#else
++# include <asm/pgtable-2level.h>
++#endif
++
++/*
++ * Macro to mark a page protection value as "uncacheable".
++ * On processors which do not support it, this is a no-op.
++ */
++#define pgprot_noncached(prot) \
++ ((boot_cpu_data.x86 > 3) \
++ ? (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) \
++ : (prot))
++
++/*
++ * Conversion functions: convert a page and protection to a page entry,
++ * and a page entry and page directory to the page they refer to.
++ */
++#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot))
++
++/*
++ * the pgd page can be thought of an array like this: pgd_t[PTRS_PER_PGD]
++ *
++ * this macro returns the index of the entry in the pgd page which would
++ * control the given virtual address
++ */
++#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD - 1))
++#define pgd_index_k(addr) pgd_index((addr))
++
++/*
++ * pgd_offset() returns a (pgd_t *)
++ * pgd_index() is used get the offset into the pgd page's array of pgd_t's;
++ */
++#define pgd_offset(mm, address) ((mm)->pgd + pgd_index((address)))
++
++/*
++ * a shortcut which implies the use of the kernel's pgd, instead
++ * of a process's
++ */
++#define pgd_offset_k(address) pgd_offset(&init_mm, (address))
++
++static inline int pud_large(pud_t pud) { return 0; }
++
++/*
++ * the pmd page can be thought of an array like this: pmd_t[PTRS_PER_PMD]
++ *
++ * this macro returns the index of the entry in the pmd page which would
++ * control the given virtual address
++ */
++#define pmd_index(address) \
++ (((address) >> PMD_SHIFT) & (PTRS_PER_PMD - 1))
++
++/*
++ * the pte page can be thought of an array like this: pte_t[PTRS_PER_PTE]
++ *
++ * this macro returns the index of the entry in the pte page which would
++ * control the given virtual address
++ */
++#define pte_index(address) \
++ (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
++#define pte_offset_kernel(dir, address) \
++ ((pte_t *)pmd_page_vaddr(*(dir)) + pte_index((address)))
++
++#define pmd_page(pmd) (pfn_to_page(pmd_val((pmd)) >> PAGE_SHIFT))
++
++#define pmd_page_vaddr(pmd) \
++ ((unsigned long)__va(pmd_val((pmd)) & PTE_MASK))
++
++#if defined(CONFIG_HIGHPTE)
++#define pte_offset_map(dir, address) \
++ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)), KM_PTE0) + \
++ pte_index((address)))
++#define pte_offset_map_nested(dir, address) \
++ ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)), KM_PTE1) + \
++ pte_index((address)))
++#define pte_unmap(pte) kunmap_atomic((pte), KM_PTE0)
++#define pte_unmap_nested(pte) kunmap_atomic((pte), KM_PTE1)
++#else
++#define pte_offset_map(dir, address) \
++ ((pte_t *)page_address(pmd_page(*(dir))) + pte_index((address)))
++#define pte_offset_map_nested(dir, address) pte_offset_map((dir), (address))
++#define pte_unmap(pte) do { } while (0)
++#define pte_unmap_nested(pte) do { } while (0)
++#endif
++
++/* Clear a kernel PTE and flush it from the TLB */
++#define kpte_clear_flush(ptep, vaddr) \
++do { \
++ if (HYPERVISOR_update_va_mapping(vaddr, __pte(0), UVMF_INVLPG)) \
++ BUG(); \
++} while (0)
++
++/*
++ * The i386 doesn't have any external MMU info: the kernel page
++ * tables contain all the necessary information.
++ */
++#define update_mmu_cache(vma, address, pte) do { } while (0)
++
++void make_lowmem_page_readonly(void *va, unsigned int feature);
++void make_lowmem_page_writable(void *va, unsigned int feature);
++
++#endif /* !__ASSEMBLY__ */
++
++/*
++ * kern_addr_valid() is (1) for FLATMEM and (0) for
++ * SPARSEMEM and DISCONTIGMEM
++ */
++#ifdef CONFIG_FLATMEM
++#define kern_addr_valid(addr) (1)
++#else
++#define kern_addr_valid(kaddr) (0)
++#endif
++
++#define io_remap_pfn_range(vma, from, pfn, size, prot) \
++ direct_remap_pfn_range(vma, from, pfn, size, prot, DOMID_IO)
++
++#endif /* _I386_PGTABLE_H */
+diff --git a/include/asm-x86/mach-xen/asm/pgtable_64.h b/include/asm-x86/mach-xen/asm/pgtable_64.h
+new file mode 100644
+index 0000000..c2509f4
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/pgtable_64.h
+@@ -0,0 +1,304 @@
++#ifndef _X86_64_PGTABLE_H
++#define _X86_64_PGTABLE_H
++
++#include <linux/const.h>
++#ifndef __ASSEMBLY__
++
++/*
++ * This file contains the functions and defines necessary to modify and use
++ * the x86-64 page table tree.
++ */
++#include <asm/processor.h>
++#include <linux/bitops.h>
++#include <linux/threads.h>
++#include <linux/sched.h>
++#include <asm/pda.h>
++#ifdef CONFIG_XEN
++#include <asm/hypervisor.h>
++
++extern pud_t level3_user_pgt[512];
++
++extern void xen_init_pt(void);
++#endif
++
++extern pud_t level3_kernel_pgt[512];
++extern pud_t level3_ident_pgt[512];
++extern pmd_t level2_kernel_pgt[512];
++extern pgd_t init_level4_pgt[];
++
++#define swapper_pg_dir init_level4_pgt
++
++extern void paging_init(void);
++
++#endif /* !__ASSEMBLY__ */
++
++#define SHARED_KERNEL_PMD 0
++
++/*
++ * PGDIR_SHIFT determines what a top-level page table entry can map
++ */
++#define PGDIR_SHIFT 39
++#define PTRS_PER_PGD 512
++
++/*
++ * 3rd level page
++ */
++#define PUD_SHIFT 30
++#define PTRS_PER_PUD 512
++
++/*
++ * 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
++
++#ifndef __ASSEMBLY__
++
++#define pte_ERROR(e) \
++ printk("%s:%d: bad pte %p(%016lx pfn %010lx).\n", \
++ __FILE__, __LINE__, &(e), __pte_val(e), pte_pfn(e))
++#define pmd_ERROR(e) \
++ printk("%s:%d: bad pmd %p(%016lx pfn %010lx).\n", \
++ __FILE__, __LINE__, &(e), __pmd_val(e), pmd_pfn(e))
++#define pud_ERROR(e) \
++ printk("%s:%d: bad pud %p(%016lx pfn %010lx).\n", \
++ __FILE__, __LINE__, &(e), __pud_val(e), \
++ (pud_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++#define pgd_ERROR(e) \
++ printk("%s:%d: bad pgd %p(%016lx pfn %010lx).\n", \
++ __FILE__, __LINE__, &(e), __pgd_val(e), \
++ (pgd_val(e) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++
++#define pgd_none(x) (!__pgd_val(x))
++#define pud_none(x) (!__pud_val(x))
++
++struct mm_struct;
++
++#define __xen_pte_clear(ptep) xen_set_pte(ptep, __pte(0))
++
++static inline void xen_set_pte(pte_t *ptep, pte_t pte)
++{
++ *ptep = pte;
++}
++
++static inline void xen_set_pte_atomic(pte_t *ptep, pte_t pte)
++{
++ xen_set_pte(ptep, pte);
++}
++
++#ifdef CONFIG_SMP
++static inline pte_t xen_ptep_get_and_clear(pte_t *xp, pte_t ret)
++{
++ return __pte_ma(xchg(&xp->pte, 0));
++}
++#else
++#define xen_ptep_get_and_clear(xp, pte) xen_local_ptep_get_and_clear(xp, pte)
++#endif
++
++static inline void xen_set_pmd(pmd_t *pmdp, pmd_t pmd)
++{
++ xen_l2_entry_update(pmdp, pmd);
++}
++
++static inline void xen_pmd_clear(pmd_t *pmd)
++{
++ xen_set_pmd(pmd, xen_make_pmd(0));
++}
++
++static inline void xen_set_pud(pud_t *pudp, pud_t pud)
++{
++ xen_l3_entry_update(pudp, pud);
++}
++
++static inline void xen_pud_clear(pud_t *pud)
++{
++ xen_set_pud(pud, xen_make_pud(0));
++}
++
++#define __user_pgd(pgd) ((pgd) + PTRS_PER_PGD)
++
++static inline void xen_set_pgd(pgd_t *pgdp, pgd_t pgd)
++{
++ xen_l4_entry_update(pgdp, pgd);
++}
++
++static inline void xen_pgd_clear(pgd_t *pgd)
++{
++ xen_set_pgd(pgd, xen_make_pgd(0));
++ xen_set_pgd(__user_pgd(pgd), xen_make_pgd(0));
++}
++
++#define pte_same(a, b) ((a).pte == (b).pte)
++
++#endif /* !__ASSEMBLY__ */
++
++#define PMD_SIZE (_AC(1, UL) << PMD_SHIFT)
++#define PMD_MASK (~(PMD_SIZE - 1))
++#define PUD_SIZE (_AC(1, UL) << PUD_SHIFT)
++#define PUD_MASK (~(PUD_SIZE - 1))
++#define PGDIR_SIZE (_AC(1, UL) << PGDIR_SHIFT)
++#define PGDIR_MASK (~(PGDIR_SIZE - 1))
++
++
++#define MAXMEM _AC(0x00003fffffffffff, UL)
++#define VMALLOC_START _AC(0xffffc20000000000, UL)
++#define VMALLOC_END _AC(0xffffe1ffffffffff, UL)
++#define VMEMMAP_START _AC(0xffffe20000000000, UL)
++#define MODULES_VADDR _AC(0xffffffffa0000000, UL)
++#define MODULES_END _AC(0xfffffffffff00000, UL)
++#define MODULES_LEN (MODULES_END - MODULES_VADDR)
++
++#ifndef __ASSEMBLY__
++
++static inline int pgd_bad(pgd_t pgd)
++{
++ return (__pgd_val(pgd) & ~(PTE_MASK | _PAGE_USER)) != _KERNPG_TABLE;
++}
++
++static inline int pud_bad(pud_t pud)
++{
++ return (__pud_val(pud) & ~(PTE_MASK | _PAGE_USER)) != _KERNPG_TABLE;
++}
++
++static inline int pmd_bad(pmd_t pmd)
++{
++ return (__pmd_val(pmd) & ~(PTE_MASK | _PAGE_USER)) != _KERNPG_TABLE;
++}
++
++#define pte_none(x) (!(x).pte)
++#define pte_present(x) ((x).pte & (_PAGE_PRESENT | _PAGE_PROTNONE))
++
++#define pages_to_mb(x) ((x) >> (20 - PAGE_SHIFT)) /* FIXME: is this right? */
++
++#define __pte_mfn(_pte) (((_pte).pte & PTE_MASK) >> PAGE_SHIFT)
++#define pte_mfn(_pte) ((_pte).pte & _PAGE_PRESENT ? \
++ __pte_mfn(_pte) : pfn_to_mfn(__pte_mfn(_pte)))
++#define pte_pfn(_pte) ((_pte).pte & _PAGE_IO ? max_mapnr : \
++ (_pte).pte & _PAGE_PRESENT ? \
++ mfn_to_local_pfn(__pte_mfn(_pte)) : \
++ __pte_mfn(_pte))
++
++#define pte_page(x) pfn_to_page(pte_pfn((x)))
++
++/*
++ * Macro to mark a page protection value as "uncacheable".
++ */
++#define pgprot_noncached(prot) \
++ (__pgprot(pgprot_val((prot)) | _PAGE_PCD | _PAGE_PWT))
++
++/*
++ * Conversion functions: convert a page and protection to a page entry,
++ * and a page entry and page directory to the page they refer to.
++ */
++
++/*
++ * Level 4 access.
++ */
++#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, address) ((mm)->pgd + pgd_index((address)))
++#define pgd_offset_k(address) (init_level4_pgt + pgd_index((address)))
++#define pgd_present(pgd) (__pgd_val(pgd) & _PAGE_PRESENT)
++static inline int pgd_large(pgd_t pgd) { return 0; }
++#define mk_kernel_pgd(address) __pgd((address) | _KERNPG_TABLE)
++
++/* 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_vaddr(*(pgd)) + pud_index((address)))
++#define pud_present(pud) (__pud_val(pud) & _PAGE_PRESENT)
++
++static inline int pud_large(pud_t pte)
++{
++ return (__pud_val(pte) & (_PAGE_PSE | _PAGE_PRESENT)) ==
++ (_PAGE_PSE | _PAGE_PRESENT);
++}
++
++/* PMD - Level 2 access */
++#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_vaddr(*(dir)) + \
++ pmd_index(address))
++#define pmd_none(x) (!__pmd_val(x))
++#if CONFIG_XEN_COMPAT <= 0x030002
++/* pmd_present doesn't just test the _PAGE_PRESENT bit since wr.p.t.
++ can temporarily clear it. */
++#define pmd_present(x) (__pmd_val(x))
++#else
++#define pmd_present(x) (__pmd_val(x) & _PAGE_PRESENT)
++#endif
++#define pfn_pmd(nr, prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val((prot))))
++#define pmd_pfn(x) ((pmd_val((x)) & __PHYSICAL_MASK) >> PAGE_SHIFT)
++
++#define pte_to_pgoff(pte) ((__pte_val(pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT)
++#define pgoff_to_pte(off) ((pte_t) { .pte = ((off) << PAGE_SHIFT) | \
++ _PAGE_FILE })
++#define PTE_FILE_MAX_BITS __PHYSICAL_MASK_SHIFT
++
++/* PTE - Level 1 access. */
++
++/* page, protection -> pte */
++#define mk_pte(page, pgprot) pfn_pte(page_to_pfn((page)), (pgprot))
++
++#define pte_index(address) (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
++#define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_vaddr(*(dir)) + \
++ pte_index((address)))
++
++/* x86-64 always has all page tables mapped. */
++#define pte_offset_map(dir, address) pte_offset_kernel((dir), (address))
++#define pte_offset_map_nested(dir, address) pte_offset_kernel((dir), (address))
++#define pte_unmap(pte) /* NOP */
++#define pte_unmap_nested(pte) /* NOP */
++
++#define update_mmu_cache(vma, address, pte) do { } while (0)
++
++extern int direct_gbpages;
++
++/* Encode and de-code a swap entry */
++#define __swp_type(x) (((x).val >> 1) & 0x3f)
++#define __swp_offset(x) ((x).val >> 8)
++#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | \
++ ((offset) << 8) })
++#define __pte_to_swp_entry(pte) ((swp_entry_t) { __pte_val(pte) })
++#define __swp_entry_to_pte(x) ((pte_t) { .pte = (x).val })
++
++extern int kern_addr_valid(unsigned long addr);
++extern void cleanup_highmap(void);
++
++#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \
++ direct_remap_pfn_range(vma, vaddr, pfn, size, prot, DOMID_IO)
++
++#define HAVE_ARCH_UNMAPPED_AREA
++#define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN
++
++#define pgtable_cache_init() do { } while (0)
++#define check_pgt_cache() do { } while (0)
++
++#define PAGE_AGP PAGE_KERNEL_NOCACHE
++#define HAVE_PAGE_AGP 1
++
++/* fs/proc/kcore.c */
++#define kc_vaddr_to_offset(v) ((v) & __VIRTUAL_MASK)
++#define kc_offset_to_vaddr(o) \
++ (((o) & (1UL << (__VIRTUAL_MASK_SHIFT - 1))) \
++ ? ((o) | ~__VIRTUAL_MASK) \
++ : (o))
++
++#define __HAVE_ARCH_PTE_SAME
++#endif /* !__ASSEMBLY__ */
++
++#endif /* _X86_64_PGTABLE_H */
+diff --git a/include/asm-x86/mach-xen/asm/processor.h b/include/asm-x86/mach-xen/asm/processor.h
+new file mode 100644
+index 0000000..78ce1e5
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/processor.h
+@@ -0,0 +1,869 @@
++#ifndef __ASM_X86_PROCESSOR_H
++#define __ASM_X86_PROCESSOR_H
++
++#include <asm/processor-flags.h>
++
++/* Forward declaration, a strange C thing */
++struct task_struct;
++struct mm_struct;
++
++#include <asm/vm86.h>
++#include <asm/math_emu.h>
++#include <asm/segment.h>
++#include <asm/types.h>
++#include <asm/sigcontext.h>
++#include <asm/current.h>
++#include <asm/cpufeature.h>
++#include <asm/system.h>
++#include <asm/page.h>
++#include <asm/percpu.h>
++#include <asm/msr.h>
++#include <asm/desc_defs.h>
++#include <asm/nops.h>
++
++#include <linux/personality.h>
++#include <linux/cpumask.h>
++#include <linux/cache.h>
++#include <linux/threads.h>
++#include <linux/init.h>
++#include <xen/interface/physdev.h>
++
++/*
++ * Default implementation of macro that returns current
++ * instruction pointer ("program counter").
++ */
++static inline void *current_text_addr(void)
++{
++ void *pc;
++
++ asm volatile("mov $1f, %0; 1:":"=r" (pc));
++
++ return pc;
++}
++
++#ifdef CONFIG_X86_VSMP
++# define ARCH_MIN_TASKALIGN (1 << INTERNODE_CACHE_SHIFT)
++# define ARCH_MIN_MMSTRUCT_ALIGN (1 << INTERNODE_CACHE_SHIFT)
++#else
++# define ARCH_MIN_TASKALIGN 16
++# define ARCH_MIN_MMSTRUCT_ALIGN 0
++#endif
++
++/*
++ * CPU type and hardware bug flags. Kept separately for each CPU.
++ * Members of this structure are referenced in head.S, so think twice
++ * before touching them. [mj]
++ */
++
++struct cpuinfo_x86 {
++ __u8 x86; /* CPU family */
++ __u8 x86_vendor; /* CPU vendor */
++ __u8 x86_model;
++ __u8 x86_mask;
++#ifdef CONFIG_X86_32
++ char wp_works_ok; /* It doesn't on 386's */
++
++ /* Problems on some 486Dx4's and old 386's: */
++ char hlt_works_ok;
++ char hard_math;
++ char rfu;
++ char fdiv_bug;
++ char f00f_bug;
++ char coma_bug;
++ char pad0;
++#else
++ /* Number of 4K pages in DTLB/ITLB combined(in pages): */
++ int x86_tlbsize;
++ __u8 x86_virt_bits;
++ __u8 x86_phys_bits;
++ /* CPUID returned core id bits: */
++ __u8 x86_coreid_bits;
++ /* Max extended CPUID function supported: */
++ __u32 extended_cpuid_level;
++#endif
++ /* Maximum supported CPUID level, -1=no CPUID: */
++ int cpuid_level;
++ __u32 x86_capability[NCAPINTS];
++ char x86_vendor_id[16];
++ char x86_model_id[64];
++ /* in KB - valid for CPUS which support this call: */
++ int x86_cache_size;
++ int x86_cache_alignment; /* In bytes */
++ int x86_power;
++ unsigned long loops_per_jiffy;
++#ifdef CONFIG_SMP
++ /* cpus sharing the last level cache: */
++ cpumask_t llc_shared_map;
++#endif
++ /* cpuid returned max cores value: */
++ u16 x86_max_cores;
++ u16 apicid;
++ u16 initial_apicid;
++ u16 x86_clflush_size;
++#ifdef CONFIG_SMP
++ /* number of cores as seen by the OS: */
++ u16 booted_cores;
++ /* Physical processor id: */
++ u16 phys_proc_id;
++ /* Core id: */
++ u16 cpu_core_id;
++ /* Index into per_cpu list: */
++ u16 cpu_index;
++#endif
++} __attribute__((__aligned__(SMP_CACHE_BYTES)));
++
++#define X86_VENDOR_INTEL 0
++#define X86_VENDOR_CYRIX 1
++#define X86_VENDOR_AMD 2
++#define X86_VENDOR_UMC 3
++#define X86_VENDOR_CENTAUR 5
++#define X86_VENDOR_TRANSMETA 7
++#define X86_VENDOR_NSC 8
++#define X86_VENDOR_NUM 9
++
++#define X86_VENDOR_UNKNOWN 0xff
++
++/*
++ * capabilities of CPUs
++ */
++extern struct cpuinfo_x86 boot_cpu_data;
++extern struct cpuinfo_x86 new_cpu_data;
++
++extern __u32 cleared_cpu_caps[NCAPINTS];
++
++#ifdef CONFIG_SMP
++DECLARE_PER_CPU(struct cpuinfo_x86, cpu_info);
++#define cpu_data(cpu) per_cpu(cpu_info, cpu)
++#define current_cpu_data cpu_data(smp_processor_id())
++#else
++#define cpu_data(cpu) boot_cpu_data
++#define current_cpu_data boot_cpu_data
++#endif
++
++static inline int hlt_works(int cpu)
++{
++#ifdef CONFIG_X86_32
++ return cpu_data(cpu).hlt_works_ok;
++#else
++ return 1;
++#endif
++}
++
++#define cache_line_size() (boot_cpu_data.x86_cache_alignment)
++
++extern void cpu_detect(struct cpuinfo_x86 *c);
++
++extern void identify_cpu(struct cpuinfo_x86 *);
++extern void identify_boot_cpu(void);
++extern void identify_secondary_cpu(struct cpuinfo_x86 *);
++extern void print_cpu_info(struct cpuinfo_x86 *);
++extern void init_scattered_cpuid_features(struct cpuinfo_x86 *c);
++extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
++extern unsigned short num_cache_leaves;
++
++#if defined(CONFIG_X86_HT) || defined(CONFIG_X86_64)
++extern void detect_ht(struct cpuinfo_x86 *c);
++#else
++static inline void detect_ht(struct cpuinfo_x86 *c) {}
++#endif
++
++static inline void xen_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" (*eax), "2" (*ecx));
++}
++
++static inline void load_cr3(pgd_t *pgdir)
++{
++ write_cr3(__pa(pgdir));
++}
++
++#ifndef CONFIG_X86_NO_TSS
++#ifdef CONFIG_X86_32
++/* This is the TSS defined by the hardware. */
++struct x86_hw_tss {
++ unsigned short back_link, __blh;
++ unsigned long sp0;
++ unsigned short ss0, __ss0h;
++ unsigned long sp1;
++ /* ss1 caches MSR_IA32_SYSENTER_CS: */
++ unsigned short ss1, __ss1h;
++ unsigned long sp2;
++ unsigned short ss2, __ss2h;
++ unsigned long __cr3;
++ unsigned long ip;
++ unsigned long flags;
++ unsigned long ax;
++ unsigned long cx;
++ unsigned long dx;
++ unsigned long bx;
++ unsigned long sp;
++ unsigned long bp;
++ unsigned long si;
++ unsigned long di;
++ unsigned short es, __esh;
++ unsigned short cs, __csh;
++ unsigned short ss, __ssh;
++ unsigned short ds, __dsh;
++ unsigned short fs, __fsh;
++ unsigned short gs, __gsh;
++ unsigned short ldt, __ldth;
++ unsigned short trace;
++ unsigned short io_bitmap_base;
++
++} __attribute__((packed));
++extern struct tss_struct doublefault_tss;
++#else
++struct x86_hw_tss {
++ u32 reserved1;
++ u64 sp0;
++ u64 sp1;
++ u64 sp2;
++ u64 reserved2;
++ u64 ist[7];
++ u32 reserved3;
++ u32 reserved4;
++ u16 reserved5;
++ u16 io_bitmap_base;
++
++} __attribute__((packed)) ____cacheline_aligned;
++#endif
++#endif /* CONFIG_X86_NO_TSS */
++
++/*
++ * IO-bitmap sizes:
++ */
++#define IO_BITMAP_BITS 65536
++#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8)
++#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long))
++#define IO_BITMAP_OFFSET offsetof(struct tss_struct, io_bitmap)
++#define INVALID_IO_BITMAP_OFFSET 0x8000
++#define INVALID_IO_BITMAP_OFFSET_LAZY 0x9000
++
++#ifndef CONFIG_X86_NO_TSS
++struct tss_struct {
++ /*
++ * The hardware state:
++ */
++ struct x86_hw_tss x86_tss;
++
++ /*
++ * The extra 1 is there because the CPU will access an
++ * additional byte beyond the end of the IO permission
++ * bitmap. The extra byte must be all 1 bits, and must
++ * be within the limit.
++ */
++ unsigned long io_bitmap[IO_BITMAP_LONGS + 1];
++ /*
++ * Cache the current maximum and the last task that used the bitmap:
++ */
++ unsigned long io_bitmap_max;
++ struct thread_struct *io_bitmap_owner;
++
++ /*
++ * Pad the TSS to be cacheline-aligned (size is 0x100):
++ */
++ unsigned long __cacheline_filler[35];
++ /*
++ * .. and then another 0x100 bytes for the emergency kernel stack:
++ */
++ unsigned long stack[64];
++
++} __attribute__((packed));
++
++DECLARE_PER_CPU(struct tss_struct, init_tss);
++
++/*
++ * Save the original ist values for checking stack pointers during debugging
++ */
++struct orig_ist {
++ unsigned long ist[7];
++};
++#endif /* CONFIG_X86_NO_TSS */
++
++#define MXCSR_DEFAULT 0x1f80
++
++struct i387_fsave_struct {
++ u32 cwd; /* FPU Control Word */
++ u32 swd; /* FPU Status Word */
++ u32 twd; /* FPU Tag Word */
++ u32 fip; /* FPU IP Offset */
++ u32 fcs; /* FPU IP Selector */
++ u32 foo; /* FPU Operand Pointer Offset */
++ u32 fos; /* FPU Operand Pointer Selector */
++
++ /* 8*10 bytes for each FP-reg = 80 bytes: */
++ u32 st_space[20];
++
++ /* Software status information [not touched by FSAVE ]: */
++ u32 status;
++};
++
++struct i387_fxsave_struct {
++ u16 cwd; /* Control Word */
++ u16 swd; /* Status Word */
++ u16 twd; /* Tag Word */
++ u16 fop; /* Last Instruction Opcode */
++ union {
++ struct {
++ u64 rip; /* Instruction Pointer */
++ u64 rdp; /* Data Pointer */
++ };
++ struct {
++ u32 fip; /* FPU IP Offset */
++ u32 fcs; /* FPU IP Selector */
++ u32 foo; /* FPU Operand Offset */
++ u32 fos; /* FPU Operand Selector */
++ };
++ };
++ u32 mxcsr; /* MXCSR Register State */
++ u32 mxcsr_mask; /* MXCSR Mask */
++
++ /* 8*16 bytes for each FP-reg = 128 bytes: */
++ u32 st_space[32];
++
++ /* 16*16 bytes for each XMM-reg = 256 bytes: */
++ u32 xmm_space[64];
++
++ u32 padding[24];
++
++} __attribute__((aligned(16)));
++
++struct i387_soft_struct {
++ u32 cwd;
++ u32 swd;
++ u32 twd;
++ u32 fip;
++ u32 fcs;
++ u32 foo;
++ u32 fos;
++ /* 8*10 bytes for each FP-reg = 80 bytes: */
++ u32 st_space[20];
++ u8 ftop;
++ u8 changed;
++ u8 lookahead;
++ u8 no_update;
++ u8 rm;
++ u8 alimit;
++ struct info *info;
++ u32 entry_eip;
++};
++
++union thread_xstate {
++ struct i387_fsave_struct fsave;
++ struct i387_fxsave_struct fxsave;
++ struct i387_soft_struct soft;
++};
++
++#if defined(CONFIG_X86_64) && !defined(CONFIG_X86_NO_TSS)
++DECLARE_PER_CPU(struct orig_ist, orig_ist);
++#endif
++
++extern void print_cpu_info(struct cpuinfo_x86 *);
++extern unsigned int xstate_size;
++extern void free_thread_xstate(struct task_struct *);
++extern struct kmem_cache *task_xstate_cachep;
++extern void init_scattered_cpuid_features(struct cpuinfo_x86 *c);
++extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c);
++extern unsigned short num_cache_leaves;
++
++struct thread_struct {
++ /* Cached TLS descriptors: */
++ struct desc_struct tls_array[GDT_ENTRY_TLS_ENTRIES];
++ unsigned long sp0;
++ unsigned long sp;
++#ifdef CONFIG_X86_32
++ unsigned long sysenter_cs;
++#else
++ unsigned long usersp; /* Copy from PDA */
++ unsigned short es;
++ unsigned short ds;
++ unsigned short fsindex;
++ unsigned short gsindex;
++#endif
++ unsigned long ip;
++ unsigned long fs;
++ unsigned long gs;
++ /* Hardware debugging registers: */
++ unsigned long debugreg0;
++ unsigned long debugreg1;
++ unsigned long debugreg2;
++ unsigned long debugreg3;
++ unsigned long debugreg6;
++ unsigned long debugreg7;
++ /* Fault info: */
++ unsigned long cr2;
++ unsigned long trap_no;
++ unsigned long error_code;
++ /* floating point and extended processor state */
++ union thread_xstate *xstate;
++#ifdef CONFIG_X86_32
++ /* Virtual 86 mode info */
++ struct vm86_struct __user *vm86_info;
++ unsigned long screen_bitmap;
++ unsigned long v86flags, v86mask, saved_sp0;
++ unsigned int saved_fs, saved_gs;
++#endif
++ /* IO permissions: */
++ unsigned long *io_bitmap_ptr;
++ unsigned long iopl;
++ /* Max allowed port in the bitmap, in bytes: */
++ unsigned io_bitmap_max;
++/* MSR_IA32_DEBUGCTLMSR value to switch in if TIF_DEBUGCTLMSR is set. */
++ unsigned long debugctlmsr;
++/* Debug Store - if not 0 points to a DS Save Area configuration;
++ * goes into MSR_IA32_DS_AREA */
++ unsigned long ds_area_msr;
++};
++
++static inline unsigned long xen_get_debugreg(int regno)
++{
++ return HYPERVISOR_get_debugreg(regno);
++}
++
++static inline void xen_set_debugreg(int regno, unsigned long value)
++{
++ WARN_ON(HYPERVISOR_set_debugreg(regno, value));
++}
++
++/*
++ * Set IOPL bits in EFLAGS from given mask
++ */
++static inline void xen_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;
++ WARN_ON(HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl));
++}
++
++#ifndef CONFIG_X86_NO_TSS
++static inline void
++native_load_sp0(struct tss_struct *tss, struct thread_struct *thread)
++{
++ tss->x86_tss.sp0 = thread->sp0;
++#ifdef CONFIG_X86_32
++ /* Only happens when SEP is enabled, no need to test "SEP"arately: */
++ if (unlikely(tss->x86_tss.ss1 != thread->sysenter_cs)) {
++ tss->x86_tss.ss1 = thread->sysenter_cs;
++ wrmsr(MSR_IA32_SYSENTER_CS, thread->sysenter_cs, 0);
++ }
++#endif
++}
++#else
++#define xen_load_sp0(tss, thread) do { \
++ if (HYPERVISOR_stack_switch(__KERNEL_DS, (thread)->sp0)) \
++ BUG(); \
++} while (0)
++#endif
++
++#define __cpuid xen_cpuid
++#define paravirt_enabled() 0
++
++/*
++ * These special macros can be used to get or set a debugging register
++ */
++#define get_debugreg(var, register) \
++ (var) = xen_get_debugreg(register)
++#define set_debugreg(value, register) \
++ xen_set_debugreg(register, value)
++
++#define load_sp0 xen_load_sp0
++
++#define set_iopl_mask xen_set_iopl_mask
++
++/*
++ * Save the cr4 feature set we're using (ie
++ * Pentium 4MB enable and PPro Global page
++ * enable), so that any CPU's that boot up
++ * after us can get the correct flags.
++ */
++extern unsigned long mmu_cr4_features;
++
++static inline void set_in_cr4(unsigned long mask)
++{
++ unsigned cr4;
++
++ mmu_cr4_features |= mask;
++ cr4 = read_cr4();
++ cr4 |= mask;
++ write_cr4(cr4);
++}
++
++static inline void clear_in_cr4(unsigned long mask)
++{
++ unsigned cr4;
++
++ mmu_cr4_features &= ~mask;
++ cr4 = read_cr4();
++ cr4 &= ~mask;
++ write_cr4(cr4);
++}
++
++struct microcode_header {
++ unsigned int hdrver;
++ unsigned int rev;
++ unsigned int date;
++ unsigned int sig;
++ unsigned int cksum;
++ unsigned int ldrver;
++ unsigned int pf;
++ unsigned int datasize;
++ unsigned int totalsize;
++ unsigned int reserved[3];
++};
++
++struct microcode {
++ struct microcode_header hdr;
++ unsigned int bits[0];
++};
++
++typedef struct microcode microcode_t;
++typedef struct microcode_header microcode_header_t;
++
++/* microcode format is extended from prescott processors */
++struct extended_signature {
++ unsigned int sig;
++ unsigned int pf;
++ unsigned int cksum;
++};
++
++struct extended_sigtable {
++ unsigned int count;
++ unsigned int cksum;
++ unsigned int reserved[3];
++ struct extended_signature sigs[0];
++};
++
++typedef struct {
++ unsigned long seg;
++} mm_segment_t;
++
++
++/*
++ * create a kernel thread without removing it from tasklists
++ */
++extern int kernel_thread(int (*fn)(void *), void *arg, unsigned long flags);
++
++/* Free all resources held by a thread. */
++extern void release_thread(struct task_struct *);
++
++/* Prepare to copy thread state - unlazy all lazy state */
++extern void prepare_to_copy(struct task_struct *tsk);
++
++unsigned long get_wchan(struct task_struct *p);
++
++/*
++ * 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(unsigned int op, int count,
++ unsigned int *eax, unsigned int *ebx,
++ unsigned int *ecx, unsigned 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;
++}
++
++/* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */
++static inline void rep_nop(void)
++{
++ asm volatile("rep; nop" ::: "memory");
++}
++
++static inline void cpu_relax(void)
++{
++ rep_nop();
++}
++
++/* Stop speculative execution: */
++static inline void sync_core(void)
++{
++ int tmp;
++
++ asm volatile("cpuid" : "=a" (tmp) : "0" (1)
++ : "ebx", "ecx", "edx", "memory");
++}
++
++static inline void __monitor(const void *eax, unsigned long ecx,
++ unsigned long edx)
++{
++ /* "monitor %eax, %ecx, %edx;" */
++ asm volatile(".byte 0x0f, 0x01, 0xc8;"
++ :: "a" (eax), "c" (ecx), "d"(edx));
++}
++
++static inline void __mwait(unsigned long eax, unsigned long ecx)
++{
++ /* "mwait %eax, %ecx;" */
++ asm volatile(".byte 0x0f, 0x01, 0xc9;"
++ :: "a" (eax), "c" (ecx));
++}
++
++static inline void __sti_mwait(unsigned long eax, unsigned long ecx)
++{
++ trace_hardirqs_on();
++ /* "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);
++
++extern int force_mwait;
++
++extern void select_idle_routine(const struct cpuinfo_x86 *c);
++
++extern unsigned long boot_option_idle_override;
++
++extern void enable_sep_cpu(void);
++extern int sysenter_setup(void);
++
++/* Defined in head.S */
++extern struct desc_ptr early_gdt_descr;
++
++extern void cpu_set_gdt(int);
++extern void switch_to_new_gdt(void);
++extern void cpu_init(void);
++extern void init_gdt(int cpu);
++
++static inline void update_debugctlmsr(unsigned long debugctlmsr)
++{
++#ifndef CONFIG_X86_DEBUGCTLMSR
++ if (boot_cpu_data.x86 < 6)
++ return;
++#endif
++ wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr);
++}
++
++/*
++ * from system description table in BIOS. Mostly for MCA use, but
++ * others may find it useful:
++ */
++extern unsigned int machine_id;
++extern unsigned int machine_submodel_id;
++extern unsigned int BIOS_revision;
++
++/* Boot loader type from the setup header: */
++extern int bootloader_type;
++
++extern char ignore_fpu_irq;
++
++#define HAVE_ARCH_PICK_MMAP_LAYOUT 1
++#define ARCH_HAS_PREFETCHW
++#define ARCH_HAS_SPINLOCK_PREFETCH
++
++#ifdef CONFIG_X86_32
++# define BASE_PREFETCH ASM_NOP4
++# define ARCH_HAS_PREFETCH
++#else
++# define BASE_PREFETCH "prefetcht0 (%1)"
++#endif
++
++/*
++ * Prefetch instructions for Pentium III (+) and AMD Athlon (+)
++ *
++ * It's not worth to care about 3dnow prefetches for the K6
++ * because they are microcoded there and very slow.
++ */
++static inline void prefetch(const void *x)
++{
++ alternative_input(BASE_PREFETCH,
++ "prefetchnta (%1)",
++ X86_FEATURE_XMM,
++ "r" (x));
++}
++
++/*
++ * 3dnow prefetch to get an exclusive cache line.
++ * Useful for spinlocks to avoid one state transition in the
++ * cache coherency protocol:
++ */
++static inline void prefetchw(const void *x)
++{
++ alternative_input(BASE_PREFETCH,
++ "prefetchw (%1)",
++ X86_FEATURE_3DNOW,
++ "r" (x));
++}
++
++static inline void spin_lock_prefetch(const void *x)
++{
++ prefetchw(x);
++}
++
++#ifdef CONFIG_X86_32
++/*
++ * User space process size: 3GB (default).
++ */
++#define TASK_SIZE PAGE_OFFSET
++#define STACK_TOP TASK_SIZE
++#define STACK_TOP_MAX STACK_TOP
++
++#define INIT_THREAD { \
++ .sp0 = sizeof(init_stack) + (long)&init_stack, \
++ .vm86_info = NULL, \
++ .sysenter_cs = __KERNEL_CS, \
++ .io_bitmap_ptr = NULL, \
++ .fs = __KERNEL_PERCPU, \
++}
++
++/*
++ * Note that the .io_bitmap member must be extra-big. This is because
++ * the CPU will access an additional byte beyond the end of the IO
++ * permission bitmap. The extra byte must be all 1 bits, and must
++ * be within the limit.
++ */
++#define INIT_TSS { \
++ .x86_tss = { \
++ .sp0 = sizeof(init_stack) + (long)&init_stack, \
++ .ss0 = __KERNEL_DS, \
++ .ss1 = __KERNEL_CS, \
++ .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \
++ }, \
++ .io_bitmap = { [0 ... IO_BITMAP_LONGS] = ~0 }, \
++}
++
++extern unsigned long thread_saved_pc(struct task_struct *tsk);
++
++#define THREAD_SIZE_LONGS (THREAD_SIZE/sizeof(unsigned long))
++#define KSTK_TOP(info) \
++({ \
++ unsigned long *__ptr = (unsigned long *)(info); \
++ (unsigned long)(&__ptr[THREAD_SIZE_LONGS]); \
++})
++
++/*
++ * The below -8 is to reserve 8 bytes on top of the ring0 stack.
++ * This is necessary to guarantee that the entire "struct pt_regs"
++ * is accessable even if the CPU haven't stored the SS/ESP registers
++ * on the stack (interrupt gate does not save these registers
++ * when switching to the same priv ring).
++ * Therefore beware: accessing the ss/esp fields of the
++ * "struct pt_regs" is possible, but they may contain the
++ * completely wrong values.
++ */
++#define task_pt_regs(task) \
++({ \
++ struct pt_regs *__regs__; \
++ __regs__ = (struct pt_regs *)(KSTK_TOP(task_stack_page(task))-8); \
++ __regs__ - 1; \
++})
++
++#define KSTK_ESP(task) (task_pt_regs(task)->sp)
++
++#else
++/*
++ * User space process size. 47bits minus one guard page.
++ */
++#define TASK_SIZE64 ((1UL << 47) - PAGE_SIZE)
++
++/* This decides where the kernel will search for a free chunk of vm
++ * space during mmap's.
++ */
++#define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? \
++ 0xc0000000 : 0xFFFFe000)
++
++#define TASK_SIZE (test_thread_flag(TIF_IA32) ? \
++ IA32_PAGE_OFFSET : TASK_SIZE64)
++#define TASK_SIZE_OF(child) ((test_tsk_thread_flag(child, TIF_IA32)) ? \
++ IA32_PAGE_OFFSET : TASK_SIZE64)
++
++#define STACK_TOP TASK_SIZE
++#define STACK_TOP_MAX TASK_SIZE64
++
++#define INIT_THREAD { \
++ .sp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++}
++
++#define INIT_TSS { \
++ .x86_tss.sp0 = (unsigned long)&init_stack + sizeof(init_stack) \
++}
++
++/*
++ * Return saved PC of a blocked thread.
++ * What is this good for? it will be always the scheduler or ret_from_fork.
++ */
++#define thread_saved_pc(t) (*(unsigned long *)((t)->thread.sp - 8))
++
++#define task_pt_regs(tsk) ((struct pt_regs *)(tsk)->thread.sp0 - 1)
++#define KSTK_ESP(tsk) -1 /* sorry. doesn't work for syscall. */
++#endif /* CONFIG_X86_64 */
++
++extern void start_thread(struct pt_regs *regs, unsigned long new_ip,
++ unsigned long new_sp);
++
++/*
++ * This decides where the kernel will search for a free chunk of vm
++ * space during mmap's.
++ */
++#define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3))
++
++#define KSTK_EIP(task) (task_pt_regs(task)->ip)
++
++/* Get/set a process' ability to use the timestamp counter instruction */
++#define GET_TSC_CTL(adr) get_tsc_mode((adr))
++#define SET_TSC_CTL(val) set_tsc_mode((val))
++
++extern int get_tsc_mode(unsigned long adr);
++extern int set_tsc_mode(unsigned int val);
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/segment.h b/include/asm-x86/mach-xen/asm/segment.h
+new file mode 100644
+index 0000000..f86a0d2
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/segment.h
+@@ -0,0 +1,205 @@
++#ifndef _ASM_X86_SEGMENT_H_
++#define _ASM_X86_SEGMENT_H_
++
++/* 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)
++
++#define GDT_ENTRY_BOOT_TSS (GDT_ENTRY_BOOT_CS + 2)
++#define __BOOT_TSS (GDT_ENTRY_BOOT_TSS * 8)
++
++#ifdef CONFIG_X86_32
++/*
++ * 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 - per-cpu [ offset to per-cpu data area ]
++ * 28 - unused
++ * 29 - unused
++ * 30 - unused
++ * 31 - TSS for double fault handler
++ */
++#define GDT_ENTRY_TLS_MIN 6
++#define GDT_ENTRY_TLS_MAX (GDT_ENTRY_TLS_MIN + GDT_ENTRY_TLS_ENTRIES - 1)
++
++#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 GDT_ENTRY_KERNEL_DS (GDT_ENTRY_KERNEL_BASE + 1)
++#define __KERNEL_DS (GDT_ENTRY_KERNEL_DS * 8)
++
++#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_PERCPU (GDT_ENTRY_KERNEL_BASE + 15)
++#ifdef CONFIG_SMP
++#define __KERNEL_PERCPU (GDT_ENTRY_PERCPU * 8)
++#else
++#define __KERNEL_PERCPU 0
++#endif
++
++#define GDT_ENTRY_DOUBLEFAULT_TSS 31
++
++/*
++ * The GDT has 32 entries
++ */
++#define GDT_ENTRIES 32
++
++/* 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 */
++
++/* Bottom two bits of selector give the ring privilege level */
++#define SEGMENT_RPL_MASK 0x3
++/* Bit 2 is table indicator (LDT/GDT) */
++#define SEGMENT_TI_MASK 0x4
++
++/* User mode is privilege level 3 */
++#define USER_RPL 0x3
++/* LDT segment has TI set, GDT has it cleared */
++#define SEGMENT_LDT 0x4
++#define SEGMENT_GDT 0x0
++
++/*
++ * Matching rules for certain types of segments.
++ */
++
++/* Matches only __KERNEL_CS, ignoring PnP / USER / APM segments */
++#define SEGMENT_IS_KERNEL_CODE(x) (((x) & ~3) == GDT_ENTRY_KERNEL_CS * 8 \
++ || ((x) & ~3) == (FLAT_KERNEL_CS & ~3))
++
++/* Matches __KERNEL_CS and __USER_CS (they must be 2 entries apart) */
++#define SEGMENT_IS_FLAT_CODE(x) (((x) & ~0x13) == GDT_ENTRY_KERNEL_CS * 8 \
++ || ((x) & ~3) == (FLAT_KERNEL_CS & ~3) \
++ || ((x) & ~3) == (FLAT_USER_CS & ~3))
++
++/* Matches PNP_CS32 and PNP_CS16 (they must be consecutive) */
++#define SEGMENT_IS_PNP_CODE(x) (((x) & 0xf4) == GDT_ENTRY_PNPBIOS_BASE * 8)
++
++#define get_kernel_rpl() (xen_feature(XENFEAT_supervisor_mode_kernel)?0:1)
++
++#else
++#include <asm/cache.h>
++
++#define __KERNEL_CS 0x10
++#define __KERNEL_DS 0x18
++
++#define __KERNEL32_CS 0x08
++
++/*
++ * we cannot use the same code segment descriptor for user and kernel
++ * -- not even in the long flat mode, because of different DPL /kkeil
++ * The segment offset needs to contain a RPL. Grr. -AK
++ * GDT layout to get 64bit syscall right (sysret hardcodes gdt offsets)
++ */
++
++#define __USER32_CS 0x23 /* 4*8+3 */
++#define __USER_DS 0x2b /* 5*8+3 */
++#define __USER_CS 0x33 /* 6*8+3 */
++#define __USER32_DS __USER_DS
++
++#define GDT_ENTRY_TSS 8 /* needs two entries */
++#define GDT_ENTRY_LDT 10 /* needs two entries */
++#define GDT_ENTRY_TLS_MIN 12
++#define GDT_ENTRY_TLS_MAX 14
++
++#define GDT_ENTRY_PER_CPU 15 /* Abused to load per CPU data from limit */
++#define __PER_CPU_SEG (GDT_ENTRY_PER_CPU * 8 + 3)
++
++/* TLS indexes for 64bit - hardcoded in arch_prctl */
++#define FS_TLS 0
++#define GS_TLS 1
++
++#define GS_TLS_SEL ((GDT_ENTRY_TLS_MIN+GS_TLS)*8 + 3)
++#define FS_TLS_SEL ((GDT_ENTRY_TLS_MIN+FS_TLS)*8 + 3)
++
++#define GDT_ENTRIES 16
++
++#endif
++
++/* User mode is privilege level 3 */
++#define USER_RPL 0x3
++/* LDT segment has TI set, GDT has it cleared */
++#define SEGMENT_LDT 0x4
++#define SEGMENT_GDT 0x0
++
++/* Bottom two bits of selector give the ring privilege level */
++#define SEGMENT_RPL_MASK 0x3
++/* Bit 2 is table indicator (LDT/GDT) */
++#define SEGMENT_TI_MASK 0x4
++
++#define IDT_ENTRIES 256
++#define NUM_EXCEPTION_VECTORS 32
++#define GDT_SIZE (GDT_ENTRIES * 8)
++#define GDT_ENTRY_TLS_ENTRIES 3
++#define TLS_SIZE (GDT_ENTRY_TLS_ENTRIES * 8)
++
++#ifdef __KERNEL__
++#ifndef __ASSEMBLY__
++extern const char early_idt_handlers[NUM_EXCEPTION_VECTORS][10];
++#endif
++#endif
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/smp.h b/include/asm-x86/mach-xen/asm/smp.h
+new file mode 100644
+index 0000000..35a33cc
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/smp.h
+@@ -0,0 +1,227 @@
++#ifndef _ASM_X86_SMP_H_
++#define _ASM_X86_SMP_H_
++#ifndef __ASSEMBLY__
++#include <linux/cpumask.h>
++#include <linux/init.h>
++#include <asm/percpu.h>
++
++/*
++ * We need the APIC definitions automatically as part of 'smp.h'
++ */
++#ifdef CONFIG_X86_LOCAL_APIC
++# include <asm/mpspec.h>
++# include <asm/apic.h>
++# ifdef CONFIG_X86_IO_APIC
++# include <asm/io_apic.h>
++# endif
++#endif
++#include <asm/pda.h>
++#include <linux/thread_info.h>
++
++#define cpu_callout_map cpu_possible_map
++extern cpumask_t cpu_initialized;
++#define cpu_callin_map cpu_possible_map
++
++extern void (*mtrr_hook)(void);
++extern void zap_low_mappings(void);
++
++extern int smp_num_siblings;
++extern unsigned int num_processors;
++extern cpumask_t cpu_initialized;
++
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
++extern u16 x86_cpu_to_apicid_init[];
++extern u16 x86_bios_cpu_apicid_init[];
++extern void *x86_cpu_to_apicid_early_ptr;
++extern void *x86_bios_cpu_apicid_early_ptr;
++#else
++#define x86_cpu_to_apicid_early_ptr NULL
++#define x86_bios_cpu_apicid_early_ptr NULL
++#endif
++
++DECLARE_PER_CPU(cpumask_t, cpu_sibling_map);
++DECLARE_PER_CPU(cpumask_t, cpu_core_map);
++DECLARE_PER_CPU(u16, cpu_llc_id);
++DECLARE_PER_CPU(u16, x86_cpu_to_apicid);
++DECLARE_PER_CPU(u16, x86_bios_cpu_apicid);
++
++#ifdef CONFIG_SMP
++
++#ifndef CONFIG_XEN
++
++/* Static state in head.S used to set up a CPU */
++extern struct {
++ void *sp;
++ unsigned short ss;
++} stack_start;
++
++struct smp_ops {
++ void (*smp_prepare_boot_cpu)(void);
++ void (*smp_prepare_cpus)(unsigned max_cpus);
++ int (*cpu_up)(unsigned cpu);
++ void (*smp_cpus_done)(unsigned max_cpus);
++
++ void (*smp_send_stop)(void);
++ void (*smp_send_reschedule)(int cpu);
++ int (*smp_call_function_mask)(cpumask_t mask,
++ void (*func)(void *info), void *info,
++ int wait);
++};
++
++/* Globals due to paravirt */
++extern void set_cpu_sibling_map(int cpu);
++
++#ifndef CONFIG_PARAVIRT
++#define startup_ipi_hook(phys_apicid, start_eip, start_esp) do { } while (0)
++#endif
++extern struct smp_ops smp_ops;
++
++static inline void smp_send_stop(void)
++{
++ smp_ops.smp_send_stop();
++}
++
++static inline void smp_prepare_boot_cpu(void)
++{
++ smp_ops.smp_prepare_boot_cpu();
++}
++
++static inline void smp_prepare_cpus(unsigned int max_cpus)
++{
++ smp_ops.smp_prepare_cpus(max_cpus);
++}
++
++static inline void smp_cpus_done(unsigned int max_cpus)
++{
++ smp_ops.smp_cpus_done(max_cpus);
++}
++
++static inline int __cpu_up(unsigned int cpu)
++{
++ return smp_ops.cpu_up(cpu);
++}
++
++static inline void smp_send_reschedule(int cpu)
++{
++ smp_ops.smp_send_reschedule(cpu);
++}
++
++static inline int smp_call_function_mask(cpumask_t mask,
++ void (*func) (void *info), void *info,
++ int wait)
++{
++ return smp_ops.smp_call_function_mask(mask, func, info, wait);
++}
++
++void native_smp_prepare_boot_cpu(void);
++void native_smp_prepare_cpus(unsigned int max_cpus);
++void native_smp_cpus_done(unsigned int max_cpus);
++int native_cpu_up(unsigned int cpunum);
++
++#else /* CONFIG_XEN */
++
++void xen_smp_send_stop(void);
++void xen_smp_send_reschedule(int cpu);
++int xen_smp_call_function_mask(cpumask_t mask,
++ void (*func) (void *info), void *info,
++ int wait);
++
++#define smp_send_stop xen_smp_send_stop
++#define smp_send_reschedule xen_smp_send_reschedule
++#define smp_call_function_mask xen_smp_call_function_mask
++
++extern void prefill_possible_map(void);
++
++#endif /* CONFIG_XEN */
++
++extern int __cpu_disable(void);
++extern void __cpu_die(unsigned int cpu);
++
++extern void prefill_possible_map(void);
++
++void smp_store_cpu_info(int id);
++#define cpu_physical_id(cpu) per_cpu(x86_cpu_to_apicid, cpu)
++
++/* We don't mark CPUs online until __cpu_up(), so we need another measure */
++static inline int num_booting_cpus(void)
++{
++ return cpus_weight(cpu_callout_map);
++}
++#endif /* CONFIG_SMP */
++
++extern unsigned disabled_cpus __cpuinitdata;
++
++#ifdef CONFIG_X86_32_SMP
++/*
++ * This function is needed by all SMP systems. It must _always_ be valid
++ * from the initial startup. We map APIC_BASE very early in page_setup(),
++ * so this is correct in the x86 case.
++ */
++DECLARE_PER_CPU(int, cpu_number);
++#define raw_smp_processor_id() (x86_read_percpu(cpu_number))
++#define safe_smp_processor_id() smp_processor_id()
++
++#elif defined(CONFIG_X86_64_SMP)
++#define raw_smp_processor_id() read_pda(cpunumber)
++
++#define stack_smp_processor_id() \
++({ \
++ struct thread_info *ti; \
++ __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \
++ ti->cpu; \
++})
++#define safe_smp_processor_id() smp_processor_id()
++
++#else /* !CONFIG_X86_32_SMP && !CONFIG_X86_64_SMP */
++#define cpu_physical_id(cpu) boot_cpu_physical_apicid
++#define safe_smp_processor_id() 0
++#define stack_smp_processor_id() 0
++#endif
++
++#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_XEN)
++
++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(*(u32 *)(APIC_BASE + APIC_LDR));
++}
++
++#ifndef CONFIG_X86_64
++static inline unsigned int read_apic_id(void)
++{
++ return *(u32 *)(APIC_BASE + APIC_ID);
++}
++#else
++extern unsigned int read_apic_id(void);
++#endif
++
++
++# ifdef APIC_DEFINITION
++extern int hard_smp_processor_id(void);
++# else
++# include <mach_apicdef.h>
++static inline int hard_smp_processor_id(void)
++{
++ /* we don't want to mark this access volatile - bad code generation */
++ return GET_APIC_ID(read_apic_id());
++}
++# endif /* APIC_DEFINITION */
++
++#else /* CONFIG_X86_LOCAL_APIC */
++
++# ifndef CONFIG_SMP
++# define hard_smp_processor_id() 0
++# endif
++
++#endif /* CONFIG_X86_LOCAL_APIC */
++
++#ifdef CONFIG_HOTPLUG_CPU
++extern void cpu_exit_clear(void);
++extern void cpu_uninit(void);
++#endif
++
++extern void smp_alloc_memory(void);
++extern void lock_ipi_call_lock(void);
++extern void unlock_ipi_call_lock(void);
++#endif /* __ASSEMBLY__ */
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/spinlock.h b/include/asm-x86/mach-xen/asm/spinlock.h
+new file mode 100644
+index 0000000..90b0a0f
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/spinlock.h
+@@ -0,0 +1,317 @@
++#ifndef _X86_SPINLOCK_H_
++#define _X86_SPINLOCK_H_
++
++#include <asm/atomic.h>
++#include <asm/rwlock.h>
++#include <asm/page.h>
++#include <asm/processor.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.
++ *
++ * These are fair FIFO ticket locks, which are currently limited to 256
++ * CPUs.
++ *
++ * (the type definitions are in asm/spinlock_types.h)
++ */
++
++#ifdef CONFIG_X86_32
++# define LOCK_PTR_REG "a"
++# define REG_PTR_MODE "k"
++#else
++# define LOCK_PTR_REG "D"
++# define REG_PTR_MODE "q"
++#endif
++
++#if defined(CONFIG_X86_32) && \
++ (defined(CONFIG_X86_OOSTORE) || defined(CONFIG_X86_PPRO_FENCE))
++/*
++ * On PPro SMP or if we are using OOSTORE, we use a locked operation to unlock
++ * (PPro errata 66, 92)
++ */
++# define UNLOCK_LOCK_PREFIX LOCK_PREFIX
++#else
++# define UNLOCK_LOCK_PREFIX
++#endif
++
++int xen_spinlock_init(unsigned int cpu);
++void xen_spinlock_cleanup(unsigned int cpu);
++extern int xen_spin_wait(raw_spinlock_t *, unsigned int token);
++extern int xen_spin_wait_flags(raw_spinlock_t *, unsigned int *token,
++ unsigned int flags);
++extern void xen_spin_kick(raw_spinlock_t *, unsigned int token);
++
++/*
++ * Ticket locks are conceptually two parts, one indicating the current head of
++ * the queue, and the other indicating the current tail. The lock is acquired
++ * by atomically noting the tail and incrementing it by one (thus adding
++ * ourself to the queue and noting our position), then waiting until the head
++ * becomes equal to the the initial value of the tail.
++ *
++ * We use an xadd covering *both* parts of the lock, to increment the tail and
++ * also load the position of the head, which takes care of memory ordering
++ * issues and should be optimal for the uncontended case. Note the tail must be
++ * in the high part, because a wide xadd increment of the low part would carry
++ * up and contaminate the high part.
++ *
++ * With fewer than 2^8 possible CPUs, we can use x86's partial registers to
++ * save some instructions and make the code more elegant. There really isn't
++ * much between them in performance though, especially as locks are out of line.
++ */
++#if (NR_CPUS < 256)
++#define TICKET_SHIFT 8
++#define __raw_spin_lock_preamble \
++ asm(LOCK_PREFIX "xaddw %w0, %1" \
++ : "=r" (token), "+m" (lock->slock) \
++ : "0" (0x0100) \
++ : "memory", "cc")
++#define __raw_spin_lock_body \
++ asm("1:\t" \
++ "cmpb %h0, %b0\n\t" \
++ "je 2f\n\t" \
++ "decl %1\n\t" \
++ "jz 2f\n\t" \
++ "rep ; nop\n\t" \
++ "movb %2, %b0\n\t" \
++ /* don't need lfence here, because loads are in-order */ \
++ "jmp 1b\n" \
++ "2:" \
++ : "+Q" (token), "+g" (count) \
++ : "m" (lock->slock) \
++ : "memory", "cc")
++
++
++static __always_inline int __raw_spin_trylock(raw_spinlock_t *lock)
++{
++ int tmp, new;
++
++ asm("movzwl %2, %0\n\t"
++ "cmpb %h0, %b0\n\t"
++ "leal 0x100(%" REG_PTR_MODE "0), %1\n\t"
++ "jne 1f\n\t"
++ LOCK_PREFIX "cmpxchgw %w1, %2\n\t"
++ "1:\t"
++ "sete %b1\n\t"
++ "movzbl %b1, %0\n\t"
++ : "=&a" (tmp), "=&q" (new), "+m" (lock->slock)
++ :
++ : "memory", "cc");
++
++ return tmp;
++}
++
++static __always_inline void __raw_spin_unlock(raw_spinlock_t *lock)
++{
++ unsigned int token;
++ unsigned char kick;
++
++ asm(UNLOCK_LOCK_PREFIX "incb %2\n\t"
++ "movzwl %2, %0\n\t"
++ "cmpb %h0, %b0\n\t"
++ "setne %1"
++ : "=&Q" (token), "=qm" (kick), "+m" (lock->slock)
++ :
++ : "memory", "cc");
++ if (kick)
++ xen_spin_kick(lock, token);
++}
++#else
++#define TICKET_SHIFT 16
++#define __raw_spin_lock_preamble \
++ asm(LOCK_PREFIX "xaddl %0, %1" \
++ : "=r" (token), "+m" (lock->slock) \
++ : "0" (0x00010000) \
++ : "memory", "cc")
++#define __raw_spin_lock_body \
++ do { \
++ unsigned int tmp; \
++ asm("shldl $16, %0, %2\n" \
++ "1:\t" \
++ "cmpw %w2, %w0\n\t" \
++ "je 2f\n\t" \
++ "decl %1\n\t" \
++ "jz 2f\n\t" \
++ "rep ; nop\n\t" \
++ "movw %3, %w0\n\t" \
++ /* don't need lfence here, because loads are in-order */ \
++ "jmp 1b\n" \
++ "2:" \
++ : "+r" (token), "+g" (count), "=&g" (tmp) \
++ : "m" (lock->slock) \
++ : "memory", "cc"); \
++ } while (0)
++
++static __always_inline int __raw_spin_trylock(raw_spinlock_t *lock)
++{
++ int tmp;
++ int new;
++
++ asm("movl %2, %0\n\t"
++ "movl %0, %1\n\t"
++ "roll $16, %0\n\t"
++ "cmpl %0, %1\n\t"
++ "leal 0x00010000(%" REG_PTR_MODE "0), %1\n\t"
++ "jne 1f\n\t"
++ LOCK_PREFIX "cmpxchgl %1, %2\n"
++ "1:\t"
++ "sete %b1\n\t"
++ "movzbl %b1, %0\n\t"
++ : "=&a" (tmp), "=&q" (new), "+m" (lock->slock)
++ :
++ : "memory", "cc");
++
++ return tmp;
++}
++
++static __always_inline void __raw_spin_unlock(raw_spinlock_t *lock)
++{
++ unsigned int token, tmp;
++ unsigned char kick;
++
++ asm(UNLOCK_LOCK_PREFIX "incw %2\n\t"
++ "movl %2, %0\n\t"
++ "shldl $16, %0, %3\n\t"
++ "cmpw %w3, %w0\n\t"
++ "setne %1"
++ : "=&r" (token), "=qm" (kick), "+m" (lock->slock), "=&r" (tmp)
++ :
++ : "memory", "cc");
++ if (kick)
++ xen_spin_kick(lock, token);
++}
++#endif
++
++static inline int __raw_spin_is_locked(raw_spinlock_t *lock)
++{
++ int tmp = ACCESS_ONCE(lock->slock);
++
++ return !!(((tmp >> TICKET_SHIFT) ^ tmp) & ((1 << TICKET_SHIFT) - 1));
++}
++
++static inline int __raw_spin_is_contended(raw_spinlock_t *lock)
++{
++ int tmp = ACCESS_ONCE(lock->slock);
++
++ return (((tmp >> TICKET_SHIFT) - tmp) & ((1 << TICKET_SHIFT) - 1)) > 1;
++}
++
++static __always_inline void __raw_spin_lock(raw_spinlock_t *lock)
++{
++ unsigned int token, count;
++
++ __raw_spin_lock_preamble;
++ do {
++ count = 1 << 10;
++ __raw_spin_lock_body;
++ } while (!count && !xen_spin_wait(lock, token));
++}
++
++static __always_inline void __raw_spin_lock_flags(raw_spinlock_t *lock,
++ unsigned long flags)
++{
++ unsigned int token, count;
++
++ __raw_spin_lock_preamble;
++ do {
++ count = 1 << 10;
++ __raw_spin_lock_body;
++ } while (!count && !xen_spin_wait_flags(lock, &token, flags));
++}
++
++static inline void __raw_spin_unlock_wait(raw_spinlock_t *lock)
++{
++ while (__raw_spin_is_locked(lock))
++ cpu_relax();
++}
++
++/*
++ * 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.
++ */
++
++/**
++ * read_can_lock - would read_trylock() succeed?
++ * @lock: the rwlock in question.
++ */
++static inline int __raw_read_can_lock(raw_rwlock_t *lock)
++{
++ return (int)(lock)->lock > 0;
++}
++
++/**
++ * write_can_lock - would write_trylock() succeed?
++ * @lock: the rwlock in question.
++ */
++static inline int __raw_write_can_lock(raw_rwlock_t *lock)
++{
++ return (lock)->lock == RW_LOCK_BIAS;
++}
++
++static inline void __raw_read_lock(raw_rwlock_t *rw)
++{
++ asm volatile(LOCK_PREFIX " subl $1,(%0)\n\t"
++ "jns 1f\n"
++ "call __read_lock_failed\n\t"
++ "1:\n"
++ ::LOCK_PTR_REG (rw) : "memory");
++}
++
++static inline void __raw_write_lock(raw_rwlock_t *rw)
++{
++ asm volatile(LOCK_PREFIX " subl %1,(%0)\n\t"
++ "jz 1f\n"
++ "call __write_lock_failed\n\t"
++ "1:\n"
++ ::LOCK_PTR_REG (rw), "i" (RW_LOCK_BIAS) : "memory");
++}
++
++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 %1, %0"
++ : "+m" (rw->lock) : "i" (RW_LOCK_BIAS) : "memory");
++}
++
++#define _raw_spin_relax(lock) cpu_relax()
++#define _raw_read_relax(lock) cpu_relax()
++#define _raw_write_relax(lock) cpu_relax()
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/swiotlb.h b/include/asm-x86/mach-xen/asm/swiotlb.h
+new file mode 100644
+index 0000000..515b4cc
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/swiotlb.h
+@@ -0,0 +1,8 @@
++#ifndef _ASM_SWIOTLB_H
++
++#include "../../swiotlb.h"
++
++dma_addr_t swiotlb_map_single_phys(struct device *, phys_addr_t, size_t size,
++ int dir);
++
++#endif /* _ASM_SWIOTLB_H */
+diff --git a/include/asm-x86/mach-xen/asm/synch_bitops.h b/include/asm-x86/mach-xen/asm/synch_bitops.h
+new file mode 100644
+index 0000000..be5f59b
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/synch_bitops.h
+@@ -0,0 +1,126 @@
++#ifndef __XEN_SYNCH_BITOPS_H__
++#define __XEN_SYNCH_BITOPS_H__
++
++/*
++ * Copyright 1992, Linus Torvalds.
++ * Heavily modified to provide guaranteed strong synchronisation
++ * when communicating with Xen or other guest OSes running on other CPUs.
++ */
++
++#ifdef HAVE_XEN_PLATFORM_COMPAT_H
++#include <xen/platform-compat.h>
++#endif
++
++#define ADDR (*(volatile long *) addr)
++
++static __inline__ void synch_set_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btsl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
++}
++
++static __inline__ void synch_clear_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btrl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
++}
++
++static __inline__ void synch_change_bit(int nr, volatile void * addr)
++{
++ __asm__ __volatile__ (
++ "lock btcl %1,%0"
++ : "+m" (ADDR) : "Ir" (nr) : "memory" );
++}
++
++static __inline__ int synch_test_and_set_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++ __asm__ __volatile__ (
++ "lock btsl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
++}
++
++static __inline__ int synch_test_and_clear_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++ __asm__ __volatile__ (
++ "lock btrl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
++}
++
++static __inline__ int synch_test_and_change_bit(int nr, volatile void * addr)
++{
++ int oldbit;
++
++ __asm__ __volatile__ (
++ "lock btcl %2,%1\n\tsbbl %0,%0"
++ : "=r" (oldbit), "+m" (ADDR) : "Ir" (nr) : "memory");
++ return oldbit;
++}
++
++struct __synch_xchg_dummy { unsigned long a[100]; };
++#define __synch_xg(x) ((struct __synch_xchg_dummy *)(x))
++
++#define synch_cmpxchg(ptr, old, new) \
++((__typeof__(*(ptr)))__synch_cmpxchg((ptr),\
++ (unsigned long)(old), \
++ (unsigned long)(new), \
++ sizeof(*(ptr))))
++
++static inline unsigned long __synch_cmpxchg(volatile void *ptr,
++ unsigned long old,
++ unsigned long new, int size)
++{
++ unsigned long prev;
++ switch (size) {
++ case 1:
++ __asm__ __volatile__("lock; cmpxchgb %b1,%2"
++ : "=a"(prev)
++ : "q"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++ case 2:
++ __asm__ __volatile__("lock; cmpxchgw %w1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#ifdef CONFIG_X86_64
++ case 4:
++ __asm__ __volatile__("lock; cmpxchgl %k1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++ case 8:
++ __asm__ __volatile__("lock; cmpxchgq %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#else
++ case 4:
++ __asm__ __volatile__("lock; cmpxchgl %1,%2"
++ : "=a"(prev)
++ : "r"(new), "m"(*__synch_xg(ptr)),
++ "0"(old)
++ : "memory");
++ return prev;
++#endif
++ }
++ return old;
++}
++
++#define synch_test_bit test_bit
++
++#define synch_cmpxchg_subword synch_cmpxchg
++
++#endif /* __XEN_SYNCH_BITOPS_H__ */
+diff --git a/include/asm-x86/mach-xen/asm/system.h b/include/asm-x86/mach-xen/asm/system.h
+new file mode 100644
+index 0000000..209c2fe
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/system.h
+@@ -0,0 +1,406 @@
++#ifndef _ASM_X86_SYSTEM_H_
++#define _ASM_X86_SYSTEM_H_
++
++#include <asm/asm.h>
++#include <asm/segment.h>
++#include <asm/cpufeature.h>
++#include <asm/cmpxchg.h>
++#include <asm/nops.h>
++#include <asm/hypervisor.h>
++
++#include <linux/kernel.h>
++#include <linux/irqflags.h>
++
++/* entries in ARCH_DLINFO: */
++#ifdef CONFIG_IA32_EMULATION
++# define AT_VECTOR_SIZE_ARCH 2
++#else
++# define AT_VECTOR_SIZE_ARCH 1
++#endif
++
++#ifdef CONFIG_X86_32
++
++struct task_struct; /* one of the stranger aspects of C forward declarations */
++struct task_struct *__switch_to(struct task_struct *prev,
++ struct task_struct *next);
++
++/*
++ * Saving eflags is important. It switches not only IOPL between tasks,
++ * it also protects other tasks from NT leaking through sysenter etc.
++ */
++#define switch_to(prev, next, last) \
++do { \
++ /* \
++ * Context-switching clobbers all registers, so we clobber \
++ * them explicitly, via unused output variables. \
++ * (EAX and EBP is not listed because EBP is saved/restored \
++ * explicitly for wchan access and EAX is the return value of \
++ * __switch_to()) \
++ */ \
++ unsigned long ebx, ecx, edx, esi, edi; \
++ \
++ asm volatile("pushfl\n\t" /* save flags */ \
++ "pushl %%ebp\n\t" /* save EBP */ \
++ "movl %%esp,%[prev_sp]\n\t" /* save ESP */ \
++ "movl %[next_sp],%%esp\n\t" /* restore ESP */ \
++ "movl $1f,%[prev_ip]\n\t" /* save EIP */ \
++ "pushl %[next_ip]\n\t" /* restore EIP */ \
++ "jmp __switch_to\n" /* regparm call */ \
++ "1:\t" \
++ "popl %%ebp\n\t" /* restore EBP */ \
++ "popfl\n" /* restore flags */ \
++ \
++ /* output parameters */ \
++ : [prev_sp] "=m" (prev->thread.sp), \
++ [prev_ip] "=m" (prev->thread.ip), \
++ "=a" (last), \
++ \
++ /* clobbered output registers: */ \
++ "=b" (ebx), "=c" (ecx), "=d" (edx), \
++ "=S" (esi), "=D" (edi) \
++ \
++ /* input parameters: */ \
++ : [next_sp] "m" (next->thread.sp), \
++ [next_ip] "m" (next->thread.ip), \
++ \
++ /* regparm parameters for __switch_to(): */ \
++ [prev] "a" (prev), \
++ [next] "d" (next)); \
++} while (0)
++
++/*
++ * disable hlt during certain critical i/o operations
++ */
++#define HAVE_DISABLE_HLT
++#else
++#define __SAVE(reg, offset) "movq %%" #reg ",(14-" #offset ")*8(%%rsp)\n\t"
++#define __RESTORE(reg, offset) "movq (14-" #offset ")*8(%%rsp),%%" #reg "\n\t"
++
++/* frame pointer must be last for get_wchan */
++#define SAVE_CONTEXT "pushf ; pushq %%rbp ; movq %%rsi,%%rbp\n\t"
++#define RESTORE_CONTEXT "movq %%rbp,%%rsi ; popq %%rbp ; popf\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 */ \
++ "movq %P[threadrsp](%[next]),%%rsp\n\t" /* restore RSP */ \
++ "call __switch_to\n\t" \
++ ".globl thread_return\n" \
++ "thread_return:\n\t" \
++ "movq %%gs:%P[pda_pcurrent],%%rsi\n\t" \
++ "movq %P[thread_info](%%rsi),%%r8\n\t" \
++ LOCK_PREFIX "btr %[tif_fork],%P[ti_flags](%%r8)\n\t" \
++ "movq %%rax,%%rdi\n\t" \
++ "jc ret_from_fork\n\t" \
++ RESTORE_CONTEXT \
++ : "=a" (last) \
++ : [next] "S" (next), [prev] "D" (prev), \
++ [threadrsp] "i" (offsetof(struct task_struct, thread.sp)), \
++ [ti_flags] "i" (offsetof(struct thread_info, flags)), \
++ [tif_fork] "i" (TIF_FORK), \
++ [thread_info] "i" (offsetof(struct task_struct, stack)), \
++ [pda_pcurrent] "i" (offsetof(struct x8664_pda, pcurrent)) \
++ : "memory", "cc" __EXTRA_CLOBBER)
++#endif
++
++#ifdef __KERNEL__
++#define _set_base(addr, base) do { unsigned long __pr; \
++__asm__ __volatile__ ("movw %%dx,%1\n\t" \
++ "rorl $16,%%edx\n\t" \
++ "movb %%dl,%2\n\t" \
++ "movb %%dh,%3" \
++ :"=&d" (__pr) \
++ :"m" (*((addr)+2)), \
++ "m" (*((addr)+4)), \
++ "m" (*((addr)+7)), \
++ "0" (base) \
++ ); } while (0)
++
++#define _set_limit(addr, limit) do { unsigned long __lr; \
++__asm__ __volatile__ ("movw %%dx,%1\n\t" \
++ "rorl $16,%%edx\n\t" \
++ "movb %2,%%dh\n\t" \
++ "andb $0xf0,%%dh\n\t" \
++ "orb %%dh,%%dl\n\t" \
++ "movb %%dl,%2" \
++ :"=&d" (__lr) \
++ :"m" (*(addr)), \
++ "m" (*((addr)+6)), \
++ "0" (limit) \
++ ); } while (0)
++
++#define set_base(ldt, base) _set_base(((char *)&(ldt)) , (base))
++#define set_limit(ldt, limit) _set_limit(((char *)&(ldt)) , ((limit)-1))
++
++extern void load_gs_index(unsigned);
++
++/*
++ * Load a segment. Fall back on loading the zero
++ * segment if something goes wrong..
++ */
++#define loadsegment(seg, value) \
++ asm volatile("\n" \
++ "1:\t" \
++ "movl %k0,%%" #seg "\n" \
++ "2:\n" \
++ ".section .fixup,\"ax\"\n" \
++ "3:\t" \
++ "movl %k1, %%" #seg "\n\t" \
++ "jmp 2b\n" \
++ ".previous\n" \
++ _ASM_EXTABLE(1b,3b) \
++ : :"r" (value), "r" (0))
++
++
++/*
++ * Save a segment register away
++ */
++#define savesegment(seg, value) \
++ asm volatile("mov %%" #seg ",%0":"=rm" (value))
++
++static inline unsigned long get_limit(unsigned long segment)
++{
++ unsigned long __limit;
++ asm("lsll %1,%0" : "=r" (__limit) : "r" (segment));
++ return __limit + 1;
++}
++
++static inline void xen_clts(void)
++{
++ HYPERVISOR_fpu_taskswitch(0);
++}
++
++static inline void xen_stts(void)
++{
++ HYPERVISOR_fpu_taskswitch(1);
++}
++
++/*
++ * Volatile isn't enough to prevent the compiler from reordering the
++ * read/write functions for the control registers and messing everything up.
++ * A memory clobber would solve the problem, but would prevent reordering of
++ * all loads stores around it, which can hurt performance. Solution is to
++ * use a variable and mimic reads and writes to it to enforce serialization
++ */
++static unsigned long __force_order;
++
++static inline unsigned long xen_read_cr0(void)
++{
++ unsigned long val;
++ asm volatile("mov %%cr0,%0\n\t" : "=r" (val), "=m" (__force_order));
++ return val;
++}
++
++static inline void xen_write_cr0(unsigned long val)
++{
++ asm volatile("mov %0,%%cr0": : "r" (val), "m" (__force_order));
++}
++
++#define xen_read_cr2() (current_vcpu_info()->arch.cr2)
++#define xen_write_cr2(val) ((void)(current_vcpu_info()->arch.cr2 = (val)))
++
++static inline unsigned long xen_read_cr3(void)
++{
++ unsigned long val;
++ asm volatile("mov %%cr3,%0\n\t" : "=r" (val), "=m" (__force_order));
++#ifdef CONFIG_X86_32
++ return mfn_to_pfn(xen_cr3_to_pfn(val)) << PAGE_SHIFT;
++#else
++ return machine_to_phys(val);
++#endif
++}
++
++static inline void xen_write_cr3(unsigned long val)
++{
++#ifdef CONFIG_X86_32
++ val = xen_pfn_to_cr3(pfn_to_mfn(val >> PAGE_SHIFT));
++#else
++ val = phys_to_machine(val);
++#endif
++ asm volatile("mov %0,%%cr3": : "r" (val), "m" (__force_order));
++}
++
++static inline unsigned long xen_read_cr4(void)
++{
++ unsigned long val;
++ asm volatile("mov %%cr4,%0\n\t" : "=r" (val), "=m" (__force_order));
++ return val;
++}
++
++#define xen_read_cr4_safe() xen_read_cr4()
++
++static inline void xen_write_cr4(unsigned long val)
++{
++ asm volatile("mov %0,%%cr4": : "r" (val), "m" (__force_order));
++}
++
++#ifdef CONFIG_X86_64
++static inline unsigned long xen_read_cr8(void)
++{
++ return 0;
++}
++
++static inline void xen_write_cr8(unsigned long val)
++{
++ BUG_ON(val);
++}
++#endif
++
++static inline void xen_wbinvd(void)
++{
++ asm volatile("wbinvd": : :"memory");
++}
++
++#define read_cr0() (xen_read_cr0())
++#define write_cr0(x) (xen_write_cr0(x))
++#define read_cr2() (xen_read_cr2())
++#define write_cr2(x) (xen_write_cr2(x))
++#define read_cr3() (xen_read_cr3())
++#define write_cr3(x) (xen_write_cr3(x))
++#define read_cr4() (xen_read_cr4())
++#define read_cr4_safe() (xen_read_cr4_safe())
++#define write_cr4(x) (xen_write_cr4(x))
++#define wbinvd() (xen_wbinvd())
++#ifdef CONFIG_X86_64
++#define read_cr8() (xen_read_cr8())
++#define write_cr8(x) (xen_write_cr8(x))
++#endif
++
++/* Clear the 'TS' bit */
++#define clts() (xen_clts())
++#define stts() (xen_stts())
++
++#endif /* __KERNEL__ */
++
++static inline void clflush(volatile void *__p)
++{
++ asm volatile("clflush %0" : "+m" (*(volatile char __force *)__p));
++}
++
++#define nop() asm volatile ("nop")
++
++void disable_hlt(void);
++void enable_hlt(void);
++
++extern int es7000_plat;
++void cpu_idle_wait(void);
++
++extern unsigned long arch_align_stack(unsigned long sp);
++extern void free_init_pages(char *what, unsigned long begin, unsigned long end);
++
++void default_idle(void);
++
++/*
++ * Force strict CPU ordering.
++ * And yes, this is required on UP too when we're talking
++ * to devices.
++ */
++#ifdef CONFIG_X86_32
++/*
++ * Some non-Intel clones support out of order store. wmb() ceases to be a
++ * nop for these.
++ */
++#define mb() alternative("lock; addl $0,0(%%esp)", "mfence", X86_FEATURE_XMM2)
++#define rmb() alternative("lock; addl $0,0(%%esp)", "lfence", X86_FEATURE_XMM2)
++#define wmb() alternative("lock; addl $0,0(%%esp)", "sfence", X86_FEATURE_XMM)
++#else
++#define mb() asm volatile("mfence":::"memory")
++#define rmb() asm volatile("lfence":::"memory")
++#define wmb() asm volatile("sfence" ::: "memory")
++#endif
++
++/**
++ * read_barrier_depends - Flush all pending reads that subsequents reads
++ * depend on.
++ *
++ * No data-dependent reads from memory-like regions are ever reordered
++ * over this barrier. All reads preceding this primitive are guaranteed
++ * to access memory (but not necessarily other CPUs' caches) before any
++ * reads following this primitive that depend on the data return by
++ * any of the preceding reads. This primitive is much lighter weight than
++ * rmb() on most CPUs, and is never heavier weight than is
++ * rmb().
++ *
++ * These ordering constraints are respected by both the local CPU
++ * and the compiler.
++ *
++ * Ordering is not guaranteed by anything other than these primitives,
++ * not even by data dependencies. See the documentation for
++ * memory_barrier() for examples and URLs to more information.
++ *
++ * For example, the following code would force ordering (the initial
++ * value of "a" is zero, "b" is one, and "p" is "&a"):
++ *
++ * <programlisting>
++ * CPU 0 CPU 1
++ *
++ * b = 2;
++ * memory_barrier();
++ * p = &b; q = p;
++ * read_barrier_depends();
++ * d = *q;
++ * </programlisting>
++ *
++ * because the read of "*q" depends on the read of "p" and these
++ * two reads are separated by a read_barrier_depends(). However,
++ * the following code, with the same initial values for "a" and "b":
++ *
++ * <programlisting>
++ * CPU 0 CPU 1
++ *
++ * a = 2;
++ * memory_barrier();
++ * b = 3; y = b;
++ * read_barrier_depends();
++ * x = a;
++ * </programlisting>
++ *
++ * does not enforce ordering, since there is no data dependency between
++ * the read of "a" and the read of "b". Therefore, on some CPUs, such
++ * as Alpha, "y" could be set to 3 and "x" to 0. Use rmb()
++ * in cases like this where there are no data dependencies.
++ **/
++
++#define read_barrier_depends() do { } while (0)
++
++#ifdef CONFIG_SMP
++#define smp_mb() mb()
++#ifdef CONFIG_X86_PPRO_FENCE
++# define smp_rmb() rmb()
++#else
++# define smp_rmb() barrier()
++#endif
++#ifdef CONFIG_X86_OOSTORE
++# define smp_wmb() wmb()
++#else
++# define smp_wmb() barrier()
++#endif
++#define smp_read_barrier_depends() read_barrier_depends()
++#define set_mb(var, value) do { (void)xchg(&var, value); } while (0)
++#else
++#define smp_mb() barrier()
++#define smp_rmb() barrier()
++#define smp_wmb() barrier()
++#define smp_read_barrier_depends() do { } while (0)
++#define set_mb(var, value) do { var = value; barrier(); } while (0)
++#endif
++
++/*
++ * Stop RDTSC speculation. This is needed when you need to use RDTSC
++ * (or get_cycles or vread that possibly accesses the TSC) in a defined
++ * code region.
++ *
++ * (Could use an alternative three way for this if there was one.)
++ */
++static inline void rdtsc_barrier(void)
++{
++ alternative(ASM_NOP3, "mfence", X86_FEATURE_MFENCE_RDTSC);
++ alternative(ASM_NOP3, "lfence", X86_FEATURE_LFENCE_RDTSC);
++}
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/system_64.h b/include/asm-x86/mach-xen/asm/system_64.h
+new file mode 100644
+index 0000000..f00e552
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/system_64.h
+@@ -0,0 +1,20 @@
++#ifndef __ASM_SYSTEM_H
++#define __ASM_SYSTEM_H
++
++#include <asm/segment.h>
++#include <asm/cmpxchg.h>
++
++
++static inline unsigned long read_cr8(void)
++{
++ return 0;
++}
++
++static inline void write_cr8(unsigned long val)
++{
++ BUG_ON(val);
++}
++
++#include <linux/irqflags.h>
++
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/tlbflush.h b/include/asm-x86/mach-xen/asm/tlbflush.h
+new file mode 100644
+index 0000000..1426179
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/tlbflush.h
+@@ -0,0 +1,105 @@
++#ifndef _ASM_X86_TLBFLUSH_H
++#define _ASM_X86_TLBFLUSH_H
++
++#include <linux/mm.h>
++#include <linux/sched.h>
++
++#include <asm/processor.h>
++#include <asm/system.h>
++
++#define __flush_tlb() xen_tlb_flush()
++#define __flush_tlb_global() xen_tlb_flush()
++#define __flush_tlb_single(addr) xen_invlpg(addr)
++#define __flush_tlb_all() xen_tlb_flush()
++#define __flush_tlb_one(addr) xen_invlpg(addr)
++
++#ifdef CONFIG_X86_32
++# define TLB_FLUSH_ALL 0xffffffff
++#else
++# define TLB_FLUSH_ALL -1ULL
++#endif
++
++/*
++ * TLB flushing:
++ *
++ * - flush_tlb() flushes the current mm struct TLBs
++ * - flush_tlb_all() flushes all processes TLBs
++ * - flush_tlb_mm(mm) flushes the specified mm context TLB's
++ * - flush_tlb_page(vma, vmaddr) flushes one page
++ * - flush_tlb_range(vma, start, end) flushes a range of pages
++ * - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
++ *
++ * ..but the i386 has somewhat limited tlb flushing capabilities,
++ * and page-granular flushes are available only on i486 and up.
++ *
++ * x86-64 can only flush individual pages or full VMs. For a range flush
++ * we always do the full VM. Might be worth trying if for a small
++ * range a few INVLPGs in a row are a win.
++ */
++
++#ifndef CONFIG_SMP
++
++#define flush_tlb() __flush_tlb()
++#define flush_tlb_all() __flush_tlb_all()
++#define local_flush_tlb() __flush_tlb()
++
++static inline void flush_tlb_mm(struct mm_struct *mm)
++{
++ if (mm == current->active_mm)
++ __flush_tlb();
++}
++
++static inline void flush_tlb_page(struct vm_area_struct *vma,
++ unsigned long addr)
++{
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb_one(addr);
++}
++
++static inline void flush_tlb_range(struct vm_area_struct *vma,
++ unsigned long start, unsigned long end)
++{
++ if (vma->vm_mm == current->active_mm)
++ __flush_tlb();
++}
++
++#else /* SMP */
++
++#include <asm/smp.h>
++
++#define local_flush_tlb() __flush_tlb()
++
++#define flush_tlb_all xen_tlb_flush_all
++#define flush_tlb_current_task() xen_tlb_flush_mask(¤t->mm->cpu_vm_mask)
++#define flush_tlb_mm(mm) xen_tlb_flush_mask(&(mm)->cpu_vm_mask)
++#define flush_tlb_page(vma, va) xen_invlpg_mask(&(vma)->vm_mm->cpu_vm_mask, va)
++
++#define flush_tlb() flush_tlb_current_task()
++
++static inline void flush_tlb_range(struct vm_area_struct *vma,
++ unsigned long start, unsigned long end)
++{
++ flush_tlb_mm(vma->vm_mm);
++}
++
++#define TLBSTATE_OK 1
++#define TLBSTATE_LAZY 2
++
++#ifdef CONFIG_X86_32
++struct tlb_state {
++ struct mm_struct *active_mm;
++ int state;
++ char __cacheline_padding[L1_CACHE_BYTES-8];
++};
++DECLARE_PER_CPU(struct tlb_state, cpu_tlbstate);
++#endif
++
++#endif /* SMP */
++
++static inline void flush_tlb_kernel_range(unsigned long start,
++ unsigned long end)
++{
++ flush_tlb_all();
++}
++
++#endif /* _ASM_X86_TLBFLUSH_H */
+diff --git a/include/asm-x86/mach-xen/asm/vga.h b/include/asm-x86/mach-xen/asm/vga.h
+new file mode 100644
+index 0000000..6e4c2fe
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/vga.h
+@@ -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 --git a/include/asm-x86/mach-xen/asm/xenoprof.h b/include/asm-x86/mach-xen/asm/xenoprof.h
+new file mode 100644
+index 0000000..2733e00
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/xenoprof.h
+@@ -0,0 +1,48 @@
++/******************************************************************************
++ * asm-i386/mach-xen/asm/xenoprof.h
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
++#ifndef __ASM_XENOPROF_H__
++#define __ASM_XENOPROF_H__
++#ifdef CONFIG_XEN
++
++struct super_block;
++struct dentry;
++int xenoprof_create_files(struct super_block * sb, struct dentry * root);
++#define HAVE_XENOPROF_CREATE_FILES
++
++struct xenoprof_init;
++void xenoprof_arch_init_counter(struct xenoprof_init *init);
++void xenoprof_arch_counter(void);
++void xenoprof_arch_start(void);
++void xenoprof_arch_stop(void);
++
++struct xenoprof_arch_shared_buffer {
++ /* nothing */
++};
++struct xenoprof_shared_buffer;
++void xenoprof_arch_unmap_shared_buffer(struct xenoprof_shared_buffer* sbuf);
++struct xenoprof_get_buffer;
++int xenoprof_arch_map_shared_buffer(struct xenoprof_get_buffer* get_buffer, struct xenoprof_shared_buffer* sbuf);
++struct xenoprof_passive;
++int xenoprof_arch_set_passive(struct xenoprof_passive* pdomain, struct xenoprof_shared_buffer* sbuf);
++
++#endif /* CONFIG_XEN */
++#endif /* __ASM_XENOPROF_H__ */
+diff --git a/include/asm-x86/mach-xen/asm/xor.h b/include/asm-x86/mach-xen/asm/xor.h
+new file mode 100644
+index 0000000..55b4951
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/xor.h
+@@ -0,0 +1,5 @@
++#ifdef CONFIG_X86_32
++# include "../../xor_32.h"
++#else
++# include "xor_64.h"
++#endif
+diff --git a/include/asm-x86/mach-xen/asm/xor_64.h b/include/asm-x86/mach-xen/asm/xor_64.h
+new file mode 100644
+index 0000000..4437a22
+--- /dev/null
++++ b/include/asm-x86/mach-xen/asm/xor_64.h
+@@ -0,0 +1,332 @@
++/*
++ * x86-64 changes / gcc fixes from Andi Kleen.
++ * Copyright 2002 Andi Kleen, SuSE Labs.
++ *
++ * This hasn't been optimized for the hammer yet, but there are likely
++ * no advantages to be gotten from x86-64 here anyways.
++ */
++
++typedef struct {
++ unsigned long a, b;
++} __attribute__((aligned(16))) xmm_store_t;
++
++/* Doesn't use gcc to save the XMM registers, because there is no easy way to
++ tell it to do a clts before the register saving. */
++#define XMMS_SAVE \
++do { \
++ preempt_disable(); \
++ if (!(current_thread_info()->status & TS_USEDFPU)) \
++ clts(); \
++ asm volatile( \
++ "movups %%xmm0,(%1) ;\n\t" \
++ "movups %%xmm1,0x10(%1) ;\n\t" \
++ "movups %%xmm2,0x20(%1) ;\n\t" \
++ "movups %%xmm3,0x30(%1) ;\n\t" \
++ : "=&r" (cr0) \
++ : "r" (xmm_save) \
++ : "memory"); \
++} while (0)
++
++#define XMMS_RESTORE \
++do { \
++ asm volatile( \
++ "sfence ;\n\t" \
++ "movups (%1),%%xmm0 ;\n\t" \
++ "movups 0x10(%1),%%xmm1 ;\n\t" \
++ "movups 0x20(%1),%%xmm2 ;\n\t" \
++ "movups 0x30(%1),%%xmm3 ;\n\t" \
++ : \
++ : "r" (cr0), "r" (xmm_save) \
++ : "memory"); \
++ if (!(current_thread_info()->status & TS_USEDFPU)) \
++ stts(); \
++ preempt_enable(); \
++} while (0)
++
++#define OFFS(x) "16*("#x")"
++#define PF_OFFS(x) "256+16*("#x")"
++#define PF0(x) " prefetchnta "PF_OFFS(x)"(%[p1]) ;\n"
++#define LD(x, y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n"
++#define ST(x, y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n"
++#define PF1(x) " prefetchnta "PF_OFFS(x)"(%[p2]) ;\n"
++#define PF2(x) " prefetchnta "PF_OFFS(x)"(%[p3]) ;\n"
++#define PF3(x) " prefetchnta "PF_OFFS(x)"(%[p4]) ;\n"
++#define PF4(x) " prefetchnta "PF_OFFS(x)"(%[p5]) ;\n"
++#define PF5(x) " prefetchnta "PF_OFFS(x)"(%[p6]) ;\n"
++#define XO1(x, y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n"
++#define XO2(x, y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n"
++#define XO3(x, y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n"
++#define XO4(x, y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n"
++#define XO5(x, y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n"
++
++
++static void
++xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
++{
++ unsigned int lines = bytes >> 8;
++ unsigned long cr0;
++ xmm_store_t xmm_save[4];
++
++ XMMS_SAVE;
++
++ asm volatile(
++#undef BLOCK
++#define BLOCK(i) \
++ LD(i, 0) \
++ LD(i + 1, 1) \
++ PF1(i) \
++ PF1(i + 2) \
++ LD(i + 2, 2) \
++ LD(i + 3, 3) \
++ PF0(i + 4) \
++ PF0(i + 6) \
++ XO1(i, 0) \
++ XO1(i + 1, 1) \
++ XO1(i + 2, 2) \
++ XO1(i + 3, 3) \
++ ST(i, 0) \
++ ST(i + 1, 1) \
++ ST(i + 2, 2) \
++ ST(i + 3, 3) \
++
++
++ PF0(0)
++ PF0(2)
++
++ " .align 32 ;\n"
++ " 1: ;\n"
++
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
++
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [p1] "+r" (p1), [p2] "+r" (p2), [cnt] "+r" (lines)
++ : [inc] "r" (256UL)
++ : "memory");
++
++ XMMS_RESTORE;
++}
++
++static void
++xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3)
++{
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
++
++ XMMS_SAVE;
++
++ asm volatile(
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i + 2) \
++ LD(i, 0) \
++ LD(i + 1, 1) \
++ LD(i + 2, 2) \
++ LD(i + 3, 3) \
++ PF2(i) \
++ PF2(i + 2) \
++ PF0(i + 4) \
++ PF0(i + 6) \
++ XO1(i, 0) \
++ XO1(i + 1, 1) \
++ XO1(i + 2, 2) \
++ XO1(i + 3, 3) \
++ XO2(i, 0) \
++ XO2(i + 1, 1) \
++ XO2(i + 2, 2) \
++ XO2(i + 3, 3) \
++ ST(i, 0) \
++ ST(i + 1, 1) \
++ ST(i + 2, 2) \
++ ST(i + 3, 3) \
++
++
++ PF0(0)
++ PF0(2)
++
++ " .align 32 ;\n"
++ " 1: ;\n"
++
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
++
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+r" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3)
++ : [inc] "r" (256UL)
++ : "memory");
++ XMMS_RESTORE;
++}
++
++static void
++xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3, unsigned long *p4)
++{
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
++
++ XMMS_SAVE;
++
++ asm volatile(
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i + 2) \
++ LD(i, 0) \
++ LD(i + 1, 1) \
++ LD(i + 2, 2) \
++ LD(i + 3, 3) \
++ PF2(i) \
++ PF2(i + 2) \
++ XO1(i, 0) \
++ XO1(i + 1, 1) \
++ XO1(i + 2, 2) \
++ XO1(i + 3, 3) \
++ PF3(i) \
++ PF3(i + 2) \
++ PF0(i + 4) \
++ PF0(i + 6) \
++ XO2(i, 0) \
++ XO2(i + 1, 1) \
++ XO2(i + 2, 2) \
++ XO2(i + 3, 3) \
++ XO3(i, 0) \
++ XO3(i + 1, 1) \
++ XO3(i + 2, 2) \
++ XO3(i + 3, 3) \
++ ST(i, 0) \
++ ST(i + 1, 1) \
++ ST(i + 2, 2) \
++ ST(i + 3, 3) \
++
++
++ PF0(0)
++ PF0(2)
++
++ " .align 32 ;\n"
++ " 1: ;\n"
++
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
++
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " addq %[inc], %[p4] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+c" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4)
++ : [inc] "r" (256UL)
++ : "memory" );
++
++ XMMS_RESTORE;
++}
++
++static void
++xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
++ unsigned long *p3, unsigned long *p4, unsigned long *p5)
++{
++ unsigned int lines = bytes >> 8;
++ xmm_store_t xmm_save[4];
++ unsigned long cr0;
++
++ XMMS_SAVE;
++
++ asm volatile(
++#undef BLOCK
++#define BLOCK(i) \
++ PF1(i) \
++ PF1(i + 2) \
++ LD(i, 0) \
++ LD(i + 1, 1) \
++ LD(i + 2, 2) \
++ LD(i + 3, 3) \
++ PF2(i) \
++ PF2(i + 2) \
++ XO1(i, 0) \
++ XO1(i + 1, 1) \
++ XO1(i + 2, 2) \
++ XO1(i + 3, 3) \
++ PF3(i) \
++ PF3(i + 2) \
++ XO2(i, 0) \
++ XO2(i + 1, 1) \
++ XO2(i + 2, 2) \
++ XO2(i + 3, 3) \
++ PF4(i) \
++ PF4(i + 2) \
++ PF0(i + 4) \
++ PF0(i + 6) \
++ XO3(i, 0) \
++ XO3(i + 1, 1) \
++ XO3(i + 2, 2) \
++ XO3(i + 3, 3) \
++ XO4(i, 0) \
++ XO4(i + 1, 1) \
++ XO4(i + 2, 2) \
++ XO4(i + 3, 3) \
++ ST(i, 0) \
++ ST(i + 1, 1) \
++ ST(i + 2, 2) \
++ ST(i + 3, 3) \
++
++
++ PF0(0)
++ PF0(2)
++
++ " .align 32 ;\n"
++ " 1: ;\n"
++
++ BLOCK(0)
++ BLOCK(4)
++ BLOCK(8)
++ BLOCK(12)
++
++ " addq %[inc], %[p1] ;\n"
++ " addq %[inc], %[p2] ;\n"
++ " addq %[inc], %[p3] ;\n"
++ " addq %[inc], %[p4] ;\n"
++ " addq %[inc], %[p5] ;\n"
++ " decl %[cnt] ; jnz 1b"
++ : [cnt] "+c" (lines),
++ [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4),
++ [p5] "+r" (p5)
++ : [inc] "r" (256UL)
++ : "memory");
++
++ XMMS_RESTORE;
++}
++
++static struct xor_block_template xor_block_sse = {
++ .name = "generic_sse",
++ .do_2 = xor_sse_2,
++ .do_3 = xor_sse_3,
++ .do_4 = xor_sse_4,
++ .do_5 = xor_sse_5,
++};
++
++#undef XOR_TRY_TEMPLATES
++#define XOR_TRY_TEMPLATES \
++do { \
++ xor_speed(&xor_block_sse); \
++} while (0)
++
++/* We force the use of the SSE xor block because it can write around L2.
++ We may also be able to load into the L1 only depending on how the cpu
++ deals with a load to a line that is being prefetched. */
++#define XOR_SELECT_TEMPLATE(FASTEST) (&xor_block_sse)
+diff --git a/include/asm-x86/mach-xen/irq_vectors.h b/include/asm-x86/mach-xen/irq_vectors.h
+new file mode 100644
+index 0000000..09e5292
+--- /dev/null
++++ b/include/asm-x86/mach-xen/irq_vectors.h
+@@ -0,0 +1,126 @@
++/*
++ * This file should contain #defines for all of the interrupt vector
++ * numbers used by this architecture.
++ *
++ * In addition, there are some standard defines:
++ *
++ * FIRST_EXTERNAL_VECTOR:
++ * The first free place for external interrupts
++ *
++ * SYSCALL_VECTOR:
++ * The IRQ vector a syscall makes the user to kernel transition
++ * under.
++ *
++ * TIMER_IRQ:
++ * The IRQ number the timer interrupt comes in at.
++ *
++ * NR_IRQS:
++ * The total number of interrupt vectors (including all the
++ * architecture specific interrupts) needed.
++ *
++ */
++#ifndef _ASM_IRQ_VECTORS_H
++#define _ASM_IRQ_VECTORS_H
++
++/*
++ * IDT vectors usable for external interrupt sources start
++ * at 0x20:
++ */
++#define FIRST_EXTERNAL_VECTOR 0x20
++
++#define SYSCALL_VECTOR 0x80
++
++/*
++ * Vectors 0x20-0x2f are used for ISA interrupts.
++ */
++
++#ifndef CONFIG_XEN
++/*
++ * Special IRQ vectors used by the SMP architecture, 0xf0-0xff
++ *
++ * some of the following vectors are 'rare', they are merged
++ * into a single vector (CALL_FUNCTION_VECTOR) to save vector space.
++ * TLB, reschedule and local APIC vectors are performance-critical.
++ *
++ * Vectors 0xf0-0xfa are free (reserved for future Linux use).
++ */
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
++#define INVALIDATE_TLB_VECTOR 0xfd
++#define RESCHEDULE_VECTOR 0xfc
++#define CALL_FUNCTION_VECTOR 0xfb
++
++#define THERMAL_APIC_VECTOR 0xf0
++/*
++ * Local APIC timer IRQ vector is on a different priority level,
++ * to work around the 'lost local interrupt if more than 2 IRQ
++ * sources per level' errata.
++ */
++#define LOCAL_TIMER_VECTOR 0xef
++
++#define SPURIOUS_APIC_VECTOR 0xff
++#define ERROR_APIC_VECTOR 0xfe
++
++/*
++ * First APIC vector available to drivers: (vectors 0x30-0xee)
++ * we start at 0x31 to spread out vectors evenly between priority
++ * levels. (0x80 is the syscall vector)
++ */
++#define FIRST_DEVICE_VECTOR 0x31
++#define FIRST_SYSTEM_VECTOR 0xef
++
++/*
++ * 16 8259A IRQ's, 208 potential APIC interrupt sources.
++ * Right now the APIC is mostly only used for SMP.
++ * 256 vectors is an architectural limit. (we can have
++ * more than 256 devices theoretically, but they will
++ * have to use shared interrupts)
++ * Since vectors 0x00-0x1f are used/reserved for the CPU,
++ * the usable vector space is 0x20-0xff (224 vectors)
++ */
++#endif
++
++#define RESCHEDULE_VECTOR 0
++#define CALL_FUNCTION_VECTOR 1
++#define SPIN_UNLOCK_VECTOR 2
++#define NR_IPIS 3
++
++/*
++ * The maximum number of vectors supported by i386 processors
++ * is limited to 256. For processors other than i386, NR_VECTORS
++ * should be changed accordingly.
++ */
++#define NR_VECTORS 256
++
++#define FPU_IRQ 13
++
++#define FIRST_VM86_IRQ 3
++#define LAST_VM86_IRQ 15
++#define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15)
++
++/*
++ * The flat IRQ space is divided into two regions:
++ * 1. A one-to-one mapping of real physical IRQs. This space is only used
++ * if we have physical device-access privilege. This region is at the
++ * start of the IRQ space so that existing device drivers do not need
++ * to be modified to translate physical IRQ numbers into our IRQ space.
++ * 3. A dynamic mapping of inter-domain and Xen-sourced virtual IRQs. These
++ * are bound using the provided bind/unbind functions.
++ */
++
++#define PIRQ_BASE 0
++#define NR_PIRQS 256
++
++#define DYNIRQ_BASE (PIRQ_BASE + NR_PIRQS)
++#define NR_DYNIRQS 256
++
++#define NR_IRQS (NR_PIRQS + NR_DYNIRQS)
++#define NR_IRQ_VECTORS NR_IRQS
++
++#define pirq_to_irq(_x) ((_x) + PIRQ_BASE)
++#define irq_to_pirq(_x) ((_x) - PIRQ_BASE)
++
++#define dynirq_to_irq(_x) ((_x) + DYNIRQ_BASE)
++#define irq_to_dynirq(_x) ((_x) - DYNIRQ_BASE)
++
++#endif /* _ASM_IRQ_VECTORS_H */
+diff --git a/include/asm-x86/mach-xen/mach_apic.h b/include/asm-x86/mach-xen/mach_apic.h
+new file mode 100644
+index 0000000..945298e
+--- /dev/null
++++ b/include/asm-x86/mach-xen/mach_apic.h
+@@ -0,0 +1,56 @@
++#ifndef __ASM_MACH_APIC_H
++#define __ASM_MACH_APIC_H
++
++#include <linux/cpumask.h>
++
++#ifdef CONFIG_X86_64
++
++#include <asm/genapic.h>
++#define INT_DELIVERY_MODE (genapic->int_delivery_mode)
++#define INT_DEST_MODE (genapic->int_dest_mode)
++#define TARGET_CPUS (genapic->target_cpus())
++#define apic_id_registered (genapic->apic_id_registered)
++#define init_apic_ldr (genapic->init_apic_ldr)
++#define cpu_mask_to_apicid (genapic->cpu_mask_to_apicid)
++#define phys_pkg_id (genapic->phys_pkg_id)
++#define vector_allocation_domain (genapic->vector_allocation_domain)
++extern void setup_apic_routing(void);
++
++#else
++
++#ifdef CONFIG_SMP
++#define TARGET_CPUS cpu_online_map
++#else
++#define TARGET_CPUS cpumask_of_cpu(0)
++#endif
++
++#define INT_DELIVERY_MODE dest_LowestPrio
++#define INT_DEST_MODE 1
++
++static inline u32 phys_pkg_id(u32 cpuid_apic, int index_msb)
++{
++ return cpuid_apic;
++}
++
++static inline void setup_apic_routing(void)
++{
++}
++
++static inline int multi_timer_check(int apic, int irq)
++{
++ return 0;
++}
++
++static inline int apicid_to_node(int logical_apicid)
++{
++ return 0;
++}
++
++static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask)
++{
++ return cpus_addr(cpumask)[0];
++}
++
++#endif /* CONFIG_X86_64 */
++
++#endif /* __ASM_MACH_APIC_H */
+diff --git a/include/asm-x86/mach-xen/mach_time.h b/include/asm-x86/mach-xen/mach_time.h
+new file mode 100644
+index 0000000..5db45bd
+--- /dev/null
++++ b/include/asm-x86/mach-xen/mach_time.h
+@@ -0,0 +1,2 @@
++#include "../mc146818rtc_32.h"
++#include "../mach-default/mach_time.h"
+diff --git a/include/asm-x86/mach-xen/mach_timer.h b/include/asm-x86/mach-xen/mach_timer.h
+new file mode 100644
+index 0000000..ae08800
+--- /dev/null
++++ b/include/asm-x86/mach-xen/mach_timer.h
+@@ -0,0 +1 @@
++#include "../mach-default/mach_timer.h"
+diff --git a/include/asm-x86/mach-xen/mach_traps.h b/include/asm-x86/mach-xen/mach_traps.h
+new file mode 100644
+index 0000000..8daed2c
+--- /dev/null
++++ b/include/asm-x86/mach-xen/mach_traps.h
+@@ -0,0 +1,33 @@
++/*
++ * include/asm-xen/asm-i386/mach-xen/mach_traps.h
++ *
++ * Machine specific NMI handling for Xen
++ */
++#ifndef _MACH_TRAPS_H
++#define _MACH_TRAPS_H
++
++#include <linux/bitops.h>
++#include <xen/interface/nmi.h>
++
++static inline void clear_mem_error(unsigned char reason) {}
++static inline void clear_io_check_error(unsigned char reason) {}
++
++static inline unsigned char get_nmi_reason(void)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ unsigned char reason = 0;
++
++ /* construct a value which looks like it came from
++ * port 0x61.
++ */
++ if (test_bit(_XEN_NMIREASON_io_error, &s->arch.nmi_reason))
++ reason |= 0x40;
++ if (test_bit(_XEN_NMIREASON_parity_error, &s->arch.nmi_reason))
++ reason |= 0x80;
++
++ return reason;
++}
++
++static inline void reassert_nmi(void) {}
++
++#endif /* !_MACH_TRAPS_H */
+diff --git a/include/asm-x86/mach-xen/setup_arch.h b/include/asm-x86/mach-xen/setup_arch.h
+new file mode 100644
+index 0000000..ad19957
+--- /dev/null
++++ b/include/asm-x86/mach-xen/setup_arch.h
+@@ -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 --git a/include/asm-x86/mach-xen/setup_arch_post.h b/include/asm-x86/mach-xen/setup_arch_post.h
+new file mode 100644
+index 0000000..7a24a1d
+--- /dev/null
++++ b/include/asm-x86/mach-xen/setup_arch_post.h
+@@ -0,0 +1,59 @@
++/**
++ * machine_specific_* - Hooks for machine specific setup.
++ *
++ * Description:
++ * This is included late in kernel/setup.c so that it can make
++ * use of all of the static functions.
++ **/
++
++#include <xen/interface/callback.h>
++
++extern void hypervisor_callback(void);
++extern void failsafe_callback(void);
++extern void nmi(void);
++
++static void __init machine_specific_arch_setup(void)
++{
++ int ret;
++ static struct callback_register __initdata event = {
++ .type = CALLBACKTYPE_event,
++ .address = (unsigned long) hypervisor_callback,
++ };
++ static struct callback_register __initdata failsafe = {
++ .type = CALLBACKTYPE_failsafe,
++ .address = (unsigned long)failsafe_callback,
++ };
++ static struct callback_register __initdata syscall = {
++ .type = CALLBACKTYPE_syscall,
++ .address = (unsigned long)system_call,
++ };
++ static struct callback_register __initdata nmi_cb = {
++ .type = CALLBACKTYPE_nmi,
++ .address = (unsigned long)nmi,
++ };
++
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &event);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &failsafe);
++ if (ret == 0)
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &syscall);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS)
++ ret = HYPERVISOR_set_callbacks(
++ event.address,
++ failsafe.address,
++ syscall.address);
++#endif
++ BUG_ON(ret);
++
++ ret = HYPERVISOR_callback_op(CALLBACKOP_register, &nmi_cb);
++#if CONFIG_XEN_COMPAT <= 0x030002
++ if (ret == -ENOSYS) {
++ static struct xennmi_callback __initdata cb = {
++ .handler_address = (unsigned long)nmi
++ };
++
++ HYPERVISOR_nmi_op(XENNMI_register_callback, &cb);
++ }
++#endif
++}
+diff --git a/include/asm-x86/mach-xen/setup_arch_pre.h b/include/asm-x86/mach-xen/setup_arch_pre.h
+new file mode 100644
+index 0000000..b18df68
+--- /dev/null
++++ b/include/asm-x86/mach-xen/setup_arch_pre.h
+@@ -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 --git a/include/asm-x86/mmu.h b/include/asm-x86/mmu.h
+index 00e8867..e9048a2 100644
+--- a/include/asm-x86/mmu.h
++++ b/include/asm-x86/mmu.h
+@@ -16,11 +16,14 @@ typedef struct {
+ rwlock_t ldtlock;
+ #endif
+ int size;
++#ifdef CONFIG_XEN
++ unsigned has_foreign_mappings:1;
++#endif
+ struct mutex lock;
+ void *vdso;
+ } mm_context_t;
+
+-#ifdef CONFIG_SMP
++#if defined(CONFIG_SMP) && !defined(CONFIG_XEN)
+ void leave_mm(int cpu);
+ #else
+ static inline void leave_mm(int cpu)
+diff --git a/include/asm-x86/nmi.h b/include/asm-x86/nmi.h
+index 1e36302..9264c4a 100644
+--- a/include/asm-x86/nmi.h
++++ b/include/asm-x86/nmi.h
+@@ -5,8 +5,6 @@
+ #include <asm/irq.h>
+ #include <asm/io.h>
+
+-#ifdef ARCH_HAS_NMI_WATCHDOG
+-
+ /**
+ * do_nmi_callback
+ *
+@@ -44,9 +42,10 @@ extern void nmi_watchdog_default(void);
+ #define nmi_watchdog_default() do {} while (0)
+ #endif
+
++#ifdef ARCH_HAS_NMI_WATCHDOG
++
+ extern int check_nmi_watchdog(void);
+ extern int nmi_watchdog_enabled;
+-extern int unknown_nmi_panic;
+ extern int avail_to_resrv_perfctr_nmi_bit(unsigned int);
+ extern int avail_to_resrv_perfctr_nmi(unsigned int);
+ extern int reserve_perfctr_nmi(unsigned int);
+@@ -78,8 +77,6 @@ extern int unknown_nmi_panic;
+ void __trigger_all_cpu_backtrace(void);
+ #define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace()
+
+-#endif
+-
+ void lapic_watchdog_stop(void);
+ int lapic_watchdog_init(unsigned nmi_hz);
+ int lapic_wd_event(unsigned nmi_hz);
+@@ -87,7 +84,12 @@ unsigned lapic_adjust_nmi_hz(unsigned hz);
+ int lapic_watchdog_ok(void);
+ void disable_lapic_nmi_watchdog(void);
+ void enable_lapic_nmi_watchdog(void);
++
++#endif
++
+ void stop_nmi(void);
+ void restart_nmi(void);
+
++extern int unknown_nmi_panic;
++
+ #endif
+diff --git a/include/asm-x86/ptrace.h b/include/asm-x86/ptrace.h
+index 9f922b0..5422958 100644
+--- a/include/asm-x86/ptrace.h
++++ b/include/asm-x86/ptrace.h
+@@ -219,7 +219,9 @@ extern void user_enable_single_step(struct task_struct *);
+ extern void user_disable_single_step(struct task_struct *);
+
+ extern void user_enable_block_step(struct task_struct *);
+-#ifdef CONFIG_X86_DEBUGCTLMSR
++#if defined(CONFIG_XEN)
++#define arch_has_block_step() (0)
++#elif defined(CONFIG_X86_DEBUGCTLMSR)
+ #define arch_has_block_step() (1)
+ #else
+ #define arch_has_block_step() (boot_cpu_data.x86 >= 6)
+diff --git a/include/asm-x86/scatterlist.h b/include/asm-x86/scatterlist.h
+index c043206..386df09 100644
+--- a/include/asm-x86/scatterlist.h
++++ b/include/asm-x86/scatterlist.h
+@@ -24,7 +24,7 @@ struct scatterlist {
+ * returns.
+ */
+ #define sg_dma_address(sg) ((sg)->dma_address)
+-#ifdef CONFIG_X86_32
++#if defined(CONFIG_X86_32) && !defined(CONFIG_XEN)
+ # define sg_dma_len(sg) ((sg)->length)
+ #else
+ # define sg_dma_len(sg) ((sg)->dma_length)
+diff --git a/include/asm-x86/thread_info_32.h b/include/asm-x86/thread_info_32.h
+index b633882..83f743d 100644
+--- a/include/asm-x86/thread_info_32.h
++++ b/include/asm-x86/thread_info_32.h
+@@ -141,6 +141,9 @@ static inline struct thread_info *current_thread_info(void)
+ #define TIF_DEBUGCTLMSR 22 /* uses thread_struct.debugctlmsr */
+ #define TIF_DS_AREA_MSR 23 /* uses thread_struct.ds_area_msr */
+ #define TIF_BTS_TRACE_TS 24 /* record scheduling event timestamps */
++#ifdef CONFIG_XEN
++#define TIF_CSTAR 31 /* cstar-based syscall (special handling) */
++#endif
+
+ #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE)
+ #define _TIF_SIGPENDING (1 << TIF_SIGPENDING)
+@@ -159,6 +162,7 @@ static inline struct thread_info *current_thread_info(void)
+ #define _TIF_DEBUGCTLMSR (1 << TIF_DEBUGCTLMSR)
+ #define _TIF_DS_AREA_MSR (1 << TIF_DS_AREA_MSR)
+ #define _TIF_BTS_TRACE_TS (1 << TIF_BTS_TRACE_TS)
++#define _TIF_CSTAR (1 << TIF_CSTAR)
+
+ /* work to do on interrupt/exception return */
+ #define _TIF_WORK_MASK \
+@@ -168,9 +172,15 @@ static inline struct thread_info *current_thread_info(void)
+ #define _TIF_ALLWORK_MASK (0x0000FFFF & ~_TIF_SECCOMP)
+
+ /* flags to check in __switch_to() */
++#ifndef CONFIG_XEN
+ #define _TIF_WORK_CTXSW \
+ (_TIF_IO_BITMAP | _TIF_NOTSC | _TIF_DEBUGCTLMSR | \
+ _TIF_DS_AREA_MSR | _TIF_BTS_TRACE_TS)
++#else
++#define _TIF_WORK_CTXSW \
++ (_TIF_NOTSC /*todo | _TIF_DEBUGCTLMSR | \
++ _TIF_DS_AREA_MSR | _TIF_BTS_TRACE_TS*/)
++#endif
+ #define _TIF_WORK_CTXSW_PREV _TIF_WORK_CTXSW
+ #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW | _TIF_DEBUG)
+
+diff --git a/include/asm-x86/thread_info_64.h b/include/asm-x86/thread_info_64.h
+index cb69f70..0e84166 100644
+--- a/include/asm-x86/thread_info_64.h
++++ b/include/asm-x86/thread_info_64.h
+@@ -157,8 +157,13 @@ static inline struct thread_info *stack_thread_info(void)
+ (_TIF_SIGPENDING|_TIF_SINGLESTEP|_TIF_MCE_NOTIFY|_TIF_HRTICK_RESCHED)
+
+ /* flags to check in __switch_to() */
++#ifndef CONFIG_XEN
+ #define _TIF_WORK_CTXSW \
+ (_TIF_IO_BITMAP|_TIF_DEBUGCTLMSR|_TIF_DS_AREA_MSR|_TIF_BTS_TRACE_TS|_TIF_NOTSC)
++#else
++#define _TIF_WORK_CTXSW \
++ (/*todo _TIF_DEBUGCTLMSR|_TIF_DS_AREA_MSR|_TIF_BTS_TRACE_TS|*/_TIF_NOTSC)
++#endif
+ #define _TIF_WORK_CTXSW_PREV _TIF_WORK_CTXSW
+ #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW|_TIF_DEBUG)
+
+diff --git a/include/asm-x86/time.h b/include/asm-x86/time.h
+index bce72d7..4d419e5 100644
+--- a/include/asm-x86/time.h
++++ b/include/asm-x86/time.h
+@@ -56,4 +56,10 @@ static inline int native_set_wallclock(unsigned long nowtime)
+
+ #endif /* CONFIG_PARAVIRT */
+
++#ifdef CONFIG_XEN
++extern int xen_independent_wallclock(void);
++extern unsigned long xen_read_persistent_clock(void);
++extern int xen_update_persistent_clock(void);
++#endif
++
+ #endif
+diff --git a/include/asm-x86/xen/interface.h b/include/asm-x86/xen/interface.h
+index 6227000..bb1c79c 100644
+--- a/include/asm-x86/xen/interface.h
++++ b/include/asm-x86/xen/interface.h
+@@ -10,17 +10,17 @@
+ #define __XEN_PUBLIC_ARCH_X86_32_H__
+
+ #ifdef __XEN__
+-#define __DEFINE_GUEST_HANDLE(name, type) \
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
+ typedef struct { type *p; } __guest_handle_ ## name
+ #else
+-#define __DEFINE_GUEST_HANDLE(name, type) \
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
+ typedef type * __guest_handle_ ## name
+ #endif
+
+-#define DEFINE_GUEST_HANDLE_STRUCT(name) \
+- __DEFINE_GUEST_HANDLE(name, struct name)
+-#define DEFINE_GUEST_HANDLE(name) __DEFINE_GUEST_HANDLE(name, name)
+-#define GUEST_HANDLE(name) __guest_handle_ ## name
++#define DEFINE_XEN_GUEST_HANDLE_STRUCT(name) \
++ __DEFINE_XEN_GUEST_HANDLE(name, struct name)
++#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
++#define XEN_GUEST_HANDLE(name) __guest_handle_ ## name
+
+ #ifdef __XEN__
+ #if defined(__i386__)
+@@ -47,14 +47,8 @@
+ #endif
+
+ #ifndef __ASSEMBLY__
+-/* Guest handles for primitive C types. */
+-__DEFINE_GUEST_HANDLE(uchar, unsigned char);
+-__DEFINE_GUEST_HANDLE(uint, unsigned int);
+-__DEFINE_GUEST_HANDLE(ulong, unsigned long);
+-DEFINE_GUEST_HANDLE(char);
+-DEFINE_GUEST_HANDLE(int);
+-DEFINE_GUEST_HANDLE(long);
+-DEFINE_GUEST_HANDLE(void);
++typedef unsigned long xen_pfn_t;
++typedef unsigned long xen_ulong_t;
+ #endif
+
+ /*
+@@ -129,7 +123,7 @@ struct trap_info {
+ uint16_t cs; /* code selector */
+ unsigned long address; /* code offset */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(trap_info);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(trap_info);
+
+ struct cpu_user_regs {
+ uint32_t ebx;
+@@ -153,7 +147,7 @@ struct cpu_user_regs {
+ uint16_t fs, _pad4;
+ uint16_t gs, _pad5;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(cpu_user_regs);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(cpu_user_regs);
+
+ typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */
+
+@@ -181,7 +175,7 @@ struct vcpu_guest_context {
+ unsigned long failsafe_callback_eip;
+ unsigned long vm_assist; /* VMASST_TYPE_* bitmap */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(vcpu_guest_context);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(vcpu_guest_context);
+
+ struct arch_shared_info {
+ unsigned long max_pfn; /* max pfn that appears in table */
+@@ -199,6 +193,7 @@ struct xen_callback {
+ unsigned long cs;
+ unsigned long eip;
+ };
++typedef struct xen_callback xen_callback_t;
+ #endif /* !__ASSEMBLY__ */
+
+ /*
+diff --git a/include/linux/aio.h b/include/linux/aio.h
+index b51ddd2..089a221 100644
+--- a/include/linux/aio.h
++++ b/include/linux/aio.h
+@@ -200,6 +200,11 @@ struct kioctx {
+ struct aio_ring_info ring_info;
+
+ struct delayed_work wq;
++#ifdef CONFIG_EPOLL
++ // poll integration
++ wait_queue_head_t poll_wait;
++ struct file *file;
++#endif
+ };
+
+ /* prototypes */
+diff --git a/include/linux/console.h b/include/linux/console.h
+index a4f27fb..dae822e 100644
+--- a/include/linux/console.h
++++ b/include/linux/console.h
+@@ -63,6 +63,7 @@ extern const struct consw dummy_con; /* dummy console buffer */
+ extern const struct consw vga_con; /* VGA text console */
+ extern const struct consw newport_con; /* SGI Newport console */
+ extern const struct consw prom_con; /* SPARC PROM console */
++extern int console_use_vt;
+
+ int con_is_bound(const struct consw *csw);
+ int register_con_driver(const struct consw *csw, int first, int last);
+diff --git a/include/linux/elfnote.h b/include/linux/elfnote.h
+index 278e3ef..9f9816a 100644
+--- a/include/linux/elfnote.h
++++ b/include/linux/elfnote.h
+@@ -52,7 +52,7 @@
+ 4484:.balign 4 ; \
+ .popsection ;
+
+-#define ELFNOTE(name, type, desc) \
++#define ELFNOTE(name, type, desc...) \
+ ELFNOTE_START(name, type, "") \
+ desc ; \
+ ELFNOTE_END
+diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
+index f1fc747..e2384c9 100644
+--- a/include/linux/interrupt.h
++++ b/include/linux/interrupt.h
+@@ -213,6 +213,12 @@ static inline int disable_irq_wake(unsigned int irq)
+ }
+ #endif /* CONFIG_GENERIC_HARDIRQS */
+
++#ifdef CONFIG_HAVE_IRQ_IGNORE_UNHANDLED
++int irq_ignore_unhandled(unsigned int irq);
++#else
++#define irq_ignore_unhandled(irq) 0
++#endif
++
+ #ifndef __ARCH_SET_SOFTIRQ_PENDING
+ #define set_softirq_pending(x) (local_softirq_pending() = (x))
+ #define or_softirq_pending(x) (local_softirq_pending() |= (x))
+diff --git a/include/linux/kexec.h b/include/linux/kexec.h
+index 3265968..d27cf0b 100644
+--- a/include/linux/kexec.h
++++ b/include/linux/kexec.h
+@@ -46,6 +46,13 @@
+ KEXEC_CORE_NOTE_NAME_BYTES + \
+ KEXEC_CORE_NOTE_DESC_BYTES )
+
++#ifndef KEXEC_ARCH_HAS_PAGE_MACROS
++#define kexec_page_to_pfn(page) page_to_pfn(page)
++#define kexec_pfn_to_page(pfn) pfn_to_page(pfn)
++#define kexec_virt_to_phys(addr) virt_to_phys(addr)
++#define kexec_phys_to_virt(addr) phys_to_virt(addr)
++#endif
++
+ /*
+ * This structure is used to hold the arguments that are used when loading
+ * kernel binaries.
+@@ -106,6 +113,12 @@ struct kimage {
+ extern NORET_TYPE void machine_kexec(struct kimage *image) ATTRIB_NORET;
+ extern int machine_kexec_prepare(struct kimage *image);
+ extern void machine_kexec_cleanup(struct kimage *image);
++#ifdef CONFIG_XEN
++extern int xen_machine_kexec_load(struct kimage *image);
++extern void xen_machine_kexec_unload(struct kimage *image);
++extern void xen_machine_kexec_setup_resources(void);
++extern void xen_machine_kexec_register_resources(struct resource *res);
++#endif
+ extern asmlinkage long sys_kexec_load(unsigned long entry,
+ unsigned long nr_segments,
+ struct kexec_segment __user *segments,
+@@ -156,6 +169,10 @@ extern struct kimage *kexec_crash_image;
+ #define kexec_flush_icache_page(page)
+ #endif
+
++#ifndef kexec_flush_icache_page
++#define kexec_flush_icache_page(page)
++#endif
++
+ #define KEXEC_ON_CRASH 0x00000001
+ #define KEXEC_ARCH_MASK 0xffff0000
+
+diff --git a/include/linux/mm.h b/include/linux/mm.h
+index 586a943..467a744 100644
+--- a/include/linux/mm.h
++++ b/include/linux/mm.h
+@@ -100,6 +100,9 @@ extern unsigned int kobjsize(const void *objp);
+ #define VM_DONTEXPAND 0x00040000 /* Cannot expand with mremap() */
+ #define VM_RESERVED 0x00080000 /* Count as reserved_vm like IO */
+ #define VM_ACCOUNT 0x00100000 /* Is a VM accounted object */
++#ifdef CONFIG_XEN
++#define VM_FOREIGN 0x00200000 /* Has pages belonging to another VM */
++#endif
+ #define VM_HUGETLB 0x00400000 /* Huge TLB Page VM */
+ #define VM_NONLINEAR 0x00800000 /* Is non-linear (remap_file_pages) */
+ #define VM_MAPPED_COPY 0x01000000 /* T if mapped copy of data (nommu mmap) */
+@@ -171,6 +174,12 @@ struct vm_operations_struct {
+ /* notification that a previously read-only page is about to become
+ * writable, if an error is returned it will cause a SIGBUS */
+ int (*page_mkwrite)(struct vm_area_struct *vma, struct page *page);
++#ifdef CONFIG_XEN
++ /* Area-specific function for clearing the PTE at @ptep. Returns the
++ * original value of @ptep. */
++ pte_t (*zap_pte)(struct vm_area_struct *vma,
++ unsigned long addr, pte_t *ptep, int is_fullmm);
++#endif
+ #ifdef CONFIG_NUMA
+ /*
+ * set_policy() op must add a reference to any non-NULL @new mempolicy
+diff --git a/include/linux/oprofile.h b/include/linux/oprofile.h
+index 041bb31..4ea26cd 100644
+--- a/include/linux/oprofile.h
++++ b/include/linux/oprofile.h
+@@ -16,6 +16,9 @@
+ #include <linux/types.h>
+ #include <linux/spinlock.h>
+ #include <asm/atomic.h>
++#ifdef CONFIG_XEN
++#include <xen/interface/xenoprof.h>
++#endif
+
+ /* Each escaped entry is prefixed by ESCAPE_CODE
+ * then one of the following codes, then the
+@@ -28,7 +31,7 @@
+ #define CPU_SWITCH_CODE 2
+ #define COOKIE_SWITCH_CODE 3
+ #define KERNEL_ENTER_SWITCH_CODE 4
+-#define KERNEL_EXIT_SWITCH_CODE 5
++#define USER_ENTER_SWITCH_CODE 5
+ #define MODULE_LOADED_CODE 6
+ #define CTX_TGID_CODE 7
+ #define TRACE_BEGIN_CODE 8
+@@ -36,6 +39,7 @@
+ #define XEN_ENTER_SWITCH_CODE 10
+ #define SPU_PROFILING_CODE 11
+ #define SPU_CTX_SWITCH_CODE 12
++#define DOMAIN_SWITCH_CODE 13
+
+ struct super_block;
+ struct dentry;
+@@ -47,6 +51,12 @@ struct oprofile_operations {
+ /* create any necessary configuration files in the oprofile fs.
+ * Optional. */
+ int (*create_files)(struct super_block * sb, struct dentry * root);
++#ifdef CONFIG_XEN
++ /* setup active domains with Xen */
++ int (*set_active)(int *active_domains, unsigned int adomains);
++ /* setup passive domains with Xen */
++ int (*set_passive)(int *passive_domains, unsigned int pdomains);
++#endif
+ /* Do any necessary interrupt setup. Optional. */
+ int (*setup)(void);
+ /* Do any necessary interrupt shutdown. Optional. */
+@@ -113,6 +123,8 @@ void oprofile_add_pc(unsigned long pc, int is_kernel, unsigned long event);
+ /* add a backtrace entry, to be called from the ->backtrace callback */
+ void oprofile_add_trace(unsigned long eip);
+
++/* add a domain switch entry */
++int oprofile_add_domain_switch(int32_t domain_id);
+
+ /**
+ * Create a file of the given name as a child of the given root, with
+diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
+index f31debf..ced09c8 100644
+--- a/include/linux/page-flags.h
++++ b/include/linux/page-flags.h
+@@ -96,6 +96,13 @@ enum pageflags {
+ #ifdef CONFIG_IA64_UNCACHED_ALLOCATOR
+ PG_uncached, /* Page has been mapped as uncached */
+ #endif
++#ifdef CONFIG_XEN
++ PG_foreign, /* Page is owned by foreign allocator. */
++ PG_pinned, /* Cannot alias with PG_owner_priv_1 since
++ * bad_page() checks include this bit.
++ * Should not use PG_arch_1 as that may have
++ * a different purpose elsewhere. */
++#endif
+ __NR_PAGEFLAGS
+ };
+
+@@ -156,7 +163,11 @@ PAGEFLAG(LRU, lru) __CLEARPAGEFLAG(LRU, lru)
+ PAGEFLAG(Active, active) __CLEARPAGEFLAG(Active, active)
+ __PAGEFLAG(Slab, slab)
+ PAGEFLAG(Checked, owner_priv_1) /* Used by some filesystems */
++#ifdef CONFIG_PARAVIRT_XEN
+ PAGEFLAG(Pinned, owner_priv_1) TESTSCFLAG(Pinned, owner_priv_1) /* Xen */
++#elif defined(CONFIG_XEN)
++PAGEFLAG(Pinned, pinned) TESTSCFLAG(Pinned, pinned)
++#endif
+ PAGEFLAG(Reserved, reserved) __CLEARPAGEFLAG(Reserved, reserved)
+ PAGEFLAG(Private, private) __CLEARPAGEFLAG(Private, private)
+ __SETPAGEFLAG(Private, private)
+@@ -243,6 +254,26 @@ static inline void SetPageUptodate(struct page *page)
+
+ CLEARPAGEFLAG(Uptodate, uptodate)
+
++#ifdef CONFIG_XEN
++TESTPAGEFLAG(Foreign, foreign)
++static inline void SetPageForeign(struct page *page,
++ void (*dtor)(struct page *))
++{
++ BUG_ON(!dtor);
++ set_bit(PG_foreign, &page->flags);
++ page->index = (long)dtor;
++}
++static inline void ClearPageForeign(struct page *page)
++{
++ clear_bit(PG_foreign, &page->flags);
++ page->index = 0;
++}
++static inline void PageForeignDestructor(struct page *page)
++{
++ ((void (*)(struct page *))page->index)(page);
++}
++#endif
++
+ extern void cancel_dirty_page(struct page *page, unsigned int account_size);
+
+ int test_clear_page_writeback(struct page *page);
+@@ -307,8 +338,16 @@ static inline void __ClearPageTail(struct page *page)
+
+ #endif /* !PAGEFLAGS_EXTENDED */
+
++#if !defined(CONFIG_XEN)
++# define PAGE_FLAGS_XEN 0
++#elif defined(CONFIG_X86)
++# define PAGE_FLAGS_XEN ((1 << PG_pinned) | (1 << PG_foreign))
++#else
++# define PAGE_FLAGS_XEN (1 << PG_foreign)
++#endif
++
+ #define PAGE_FLAGS (1 << PG_lru | 1 << PG_private | 1 << PG_locked | \
+- 1 << PG_buddy | 1 << PG_writeback | \
++ 1 << PG_buddy | 1 << PG_writeback | PAGE_FLAGS_XEN | \
+ 1 << PG_slab | 1 << PG_swapcache | 1 << PG_active)
+
+ /*
+diff --git a/include/linux/pci.h b/include/linux/pci.h
+index d18b1dd..66fe2fa 100644
+--- a/include/linux/pci.h
++++ b/include/linux/pci.h
+@@ -182,6 +182,9 @@ struct pci_dev {
+ * directly, use the values stored here. They might be different!
+ */
+ unsigned int irq;
++#ifdef CONFIG_XEN
++ unsigned int irq_old;
++#endif
+ struct resource resource[DEVICE_COUNT_RESOURCE]; /* I/O and memory regions + expansion ROMs */
+
+ /* These fields are used by common fixups */
+@@ -606,6 +609,9 @@ int pcie_set_readrq(struct pci_dev *dev, int rq);
+ void pci_update_resource(struct pci_dev *dev, struct resource *res, int resno);
+ int __must_check pci_assign_resource(struct pci_dev *dev, int i);
+ int pci_select_bars(struct pci_dev *dev, unsigned long flags);
++#ifdef CONFIG_XEN
++void pci_restore_bars(struct pci_dev *);
++#endif
+
+ /* ROM control related routines */
+ void __iomem __must_check *pci_map_rom(struct pci_dev *pdev, size_t *size);
+@@ -725,6 +731,10 @@ static inline void msi_remove_pci_irq_vectors(struct pci_dev *dev)
+
+ static inline void pci_restore_msi_state(struct pci_dev *dev)
+ { }
++#ifdef CONFIG_XEN
++#define register_msi_get_owner(func) 0
++#define unregister_msi_get_owner(func) 0
++#endif
+ #else
+ extern int pci_enable_msi(struct pci_dev *dev);
+ extern void pci_msi_shutdown(struct pci_dev *dev);
+@@ -735,6 +745,10 @@ extern void pci_msix_shutdown(struct pci_dev *dev);
+ extern void pci_disable_msix(struct pci_dev *dev);
+ extern void msi_remove_pci_irq_vectors(struct pci_dev *dev);
+ extern void pci_restore_msi_state(struct pci_dev *dev);
++#ifdef CONFIG_XEN
++extern int register_msi_get_owner(int (*func)(struct pci_dev *dev));
++extern int unregister_msi_get_owner(int (*func)(struct pci_dev *dev));
++#endif
+ #endif
+
+ #ifdef CONFIG_HT_IRQ
+diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
+index 299ec4b..a1f244c 100644
+--- a/include/linux/skbuff.h
++++ b/include/linux/skbuff.h
+@@ -217,6 +217,8 @@ typedef unsigned char *sk_buff_data_t;
+ * @local_df: allow local fragmentation
+ * @cloned: Head may be cloned (check refcnt to be sure)
+ * @nohdr: Payload reference only, must not modify header
++ * @proto_data_valid: Protocol data validated since arriving at localhost
++ * @proto_csum_blank: Protocol csum must be added before leaving localhost
+ * @pkt_type: Packet class
+ * @fclone: skbuff clone status
+ * @ip_summed: Driver fed us an IP checksum
+@@ -317,7 +319,11 @@ struct sk_buff {
+ #ifdef CONFIG_IPV6_NDISC_NODETYPE
+ __u8 ndisc_nodetype:2;
+ #endif
+- /* 14 bit hole */
++#ifdef CONFIG_XEN
++ __u8 proto_data_valid:1,
++ proto_csum_blank:1;
++#endif
++ /* 12 bit hole */
+
+ #ifdef CONFIG_NET_DMA
+ dma_cookie_t dma_cookie;
+@@ -1710,5 +1716,12 @@ static inline void skb_forward_csum(struct sk_buff *skb)
+ }
+
+ bool skb_partial_csum_set(struct sk_buff *skb, u16 start, u16 off);
++
++#if defined(CONFIG_XEN) || defined(CONFIG_PARAVIRT_XEN)
++int skb_checksum_setup(struct sk_buff *skb);
++#else
++static inline int skb_checksum_setup(struct sk_buff *skb) { return 0; }
++#endif
++
+ #endif /* __KERNEL__ */
+ #endif /* _LINUX_SKBUFF_H */
+diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h
+index 24141b4..6285c77 100644
+--- a/include/linux/sysctl.h
++++ b/include/linux/sysctl.h
+@@ -69,6 +69,7 @@ enum
+ CTL_BUS=8, /* Busses */
+ CTL_ABI=9, /* Binary emulation */
+ CTL_CPU=10, /* CPU stuff (speed scaling, etc) */
++ CTL_XEN=123, /* Xen info and control */
+ CTL_ARLAN=254, /* arlan wireless driver */
+ CTL_S390DBF=5677, /* s390 debug */
+ CTL_SUNRPC=7249, /* sunrpc debug */
+diff --git a/include/linux/vermagic.h b/include/linux/vermagic.h
+index 79b9837..ee66b45 100644
+--- a/include/linux/vermagic.h
++++ b/include/linux/vermagic.h
+@@ -22,6 +22,11 @@
+ #else
+ #define MODULE_VERMAGIC_MODVERSIONS ""
+ #endif
++#ifdef CONFIG_XEN
++#define MODULE_VERMAGIC_XEN "Xen "
++#else
++#define MODULE_VERMAGIC_XEN
++#endif
+ #ifndef MODULE_ARCH_VERMAGIC
+ #define MODULE_ARCH_VERMAGIC ""
+ #endif
+@@ -30,5 +35,5 @@
+ UTS_RELEASE " " \
+ MODULE_VERMAGIC_SMP MODULE_VERMAGIC_PREEMPT \
+ MODULE_VERMAGIC_MODULE_UNLOAD MODULE_VERMAGIC_MODVERSIONS \
+- MODULE_ARCH_VERMAGIC
++ MODULE_VERMAGIC_XEN MODULE_ARCH_VERMAGIC
+
+diff --git a/include/xen/balloon.h b/include/xen/balloon.h
+index fe43b0f..0c784d1 100644
+--- a/include/xen/balloon.h
++++ b/include/xen/balloon.h
+@@ -36,7 +36,7 @@
+
+ #include <linux/spinlock.h>
+
+-#if 0
++#ifndef CONFIG_PARAVIRT_XEN
+ /*
+ * Inform the balloon driver that it should allow some slop for device-driver
+ * memory activities.
+diff --git a/include/xen/blkif.h b/include/xen/blkif.h
+new file mode 100644
+index 0000000..0ea115d
+--- /dev/null
++++ b/include/xen/blkif.h
+@@ -0,0 +1,123 @@
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_BLKIF_H__
++#define __XEN_BLKIF_H__
++
++#include <xen/interface/io/ring.h>
++#include <xen/interface/io/blkif.h>
++#include <xen/interface/io/protocols.h>
++
++/* Not a real protocol. Used to generate ring structs which contain
++ * the elements common to all protocols only. This way we get a
++ * compiler-checkable way to use common struct elements, so we can
++ * avoid using switch(protocol) in a number of places. */
++struct blkif_common_request {
++ char dummy;
++};
++struct blkif_common_response {
++ char dummy;
++};
++
++/* i386 protocol version */
++#pragma pack(push, 4)
++struct blkif_x86_32_request {
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t id; /* private guest value, echoed in resp */
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++struct blkif_x86_32_response {
++ uint64_t id; /* copied from request */
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
++};
++typedef struct blkif_x86_32_request blkif_x86_32_request_t;
++typedef struct blkif_x86_32_response blkif_x86_32_response_t;
++#pragma pack(pop)
++
++/* x86_64 protocol version */
++struct blkif_x86_64_request {
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t __attribute__((__aligned__(8))) id;
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++};
++struct blkif_x86_64_response {
++ uint64_t __attribute__((__aligned__(8))) id;
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
++};
++typedef struct blkif_x86_64_request blkif_x86_64_request_t;
++typedef struct blkif_x86_64_response blkif_x86_64_response_t;
++
++DEFINE_RING_TYPES(blkif_common, struct blkif_common_request, struct blkif_common_response);
++DEFINE_RING_TYPES(blkif_x86_32, struct blkif_x86_32_request, struct blkif_x86_32_response);
++DEFINE_RING_TYPES(blkif_x86_64, struct blkif_x86_64_request, struct blkif_x86_64_response);
++
++union blkif_back_rings {
++ blkif_back_ring_t native;
++ blkif_common_back_ring_t common;
++ blkif_x86_32_back_ring_t x86_32;
++ blkif_x86_64_back_ring_t x86_64;
++};
++typedef union blkif_back_rings blkif_back_rings_t;
++
++enum blkif_protocol {
++ BLKIF_PROTOCOL_NATIVE = 1,
++ BLKIF_PROTOCOL_X86_32 = 2,
++ BLKIF_PROTOCOL_X86_64 = 3,
++};
++
++static void inline blkif_get_x86_32_req(blkif_request_t *dst, blkif_x86_32_request_t *src)
++{
++ int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ dst->operation = src->operation;
++ dst->nr_segments = src->nr_segments;
++ dst->handle = src->handle;
++ dst->id = src->id;
++ dst->sector_number = src->sector_number;
++ barrier();
++ if (n > dst->nr_segments)
++ n = dst->nr_segments;
++ for (i = 0; i < n; i++)
++ dst->seg[i] = src->seg[i];
++}
++
++static void inline blkif_get_x86_64_req(blkif_request_t *dst, blkif_x86_64_request_t *src)
++{
++ int i, n = BLKIF_MAX_SEGMENTS_PER_REQUEST;
++ dst->operation = src->operation;
++ dst->nr_segments = src->nr_segments;
++ dst->handle = src->handle;
++ dst->id = src->id;
++ dst->sector_number = src->sector_number;
++ barrier();
++ if (n > dst->nr_segments)
++ n = dst->nr_segments;
++ for (i = 0; i < n; i++)
++ dst->seg[i] = src->seg[i];
++}
++
++#endif /* __XEN_BLKIF_H__ */
+diff --git a/include/xen/compat_ioctl.h b/include/xen/compat_ioctl.h
+new file mode 100644
+index 0000000..0393ebf
+--- /dev/null
++++ b/include/xen/compat_ioctl.h
+@@ -0,0 +1,45 @@
++/*
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License as
++ * published by the Free Software Foundation; either version 2 of the
++ * License, or (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
++ *
++ * Copyright IBM Corp. 2007
++ *
++ * Authors: Jimi Xenidis <jimix at watson.ibm.com>
++ * Hollis Blanchard <hollisb at us.ibm.com>
++ */
++
++#ifndef __LINUX_XEN_COMPAT_H__
++#define __LINUX_XEN_COMPAT_H__
++
++#include <linux/compat.h>
++
++extern int privcmd_ioctl_32(int fd, unsigned int cmd, unsigned long arg);
++struct privcmd_mmap_32 {
++ int num;
++ domid_t dom;
++ compat_uptr_t entry;
++};
++
++struct privcmd_mmapbatch_32 {
++ int num; /* number of pages to populate */
++ domid_t dom; /* target domain */
++ __u64 addr; /* virtual address */
++ compat_uptr_t arr; /* array of mfns - top nibble set on err */
++};
++#define IOCTL_PRIVCMD_MMAP_32 \
++ _IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap_32))
++#define IOCTL_PRIVCMD_MMAPBATCH_32 \
++ _IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch_32))
++
++#endif /* __LINUX_XEN_COMPAT_H__ */
+diff --git a/include/xen/cpu_hotplug.h b/include/xen/cpu_hotplug.h
+new file mode 100644
+index 0000000..1d73010
+--- /dev/null
++++ b/include/xen/cpu_hotplug.h
+@@ -0,0 +1,41 @@
++#ifndef __XEN_CPU_HOTPLUG_H__
++#define __XEN_CPU_HOTPLUG_H__
++
++#include <linux/kernel.h>
++#include <linux/cpumask.h>
++
++#if defined(CONFIG_X86) && defined(CONFIG_SMP)
++extern cpumask_t cpu_initialized_map;
++#endif
++
++#if defined(CONFIG_HOTPLUG_CPU)
++
++int cpu_up_check(unsigned int cpu);
++void init_xenbus_allowed_cpumask(void);
++int smp_suspend(void);
++void smp_resume(void);
++
++void cpu_bringup(void);
++
++#else /* !defined(CONFIG_HOTPLUG_CPU) */
++
++#define cpu_up_check(cpu) (0)
++#define init_xenbus_allowed_cpumask() ((void)0)
++
++static inline int smp_suspend(void)
++{
++ if (num_online_cpus() > 1) {
++ printk(KERN_WARNING "Can't suspend SMP guests "
++ "without CONFIG_HOTPLUG_CPU\n");
++ return -EOPNOTSUPP;
++ }
++ return 0;
++}
++
++static inline void smp_resume(void)
++{
++}
++
++#endif /* !defined(CONFIG_HOTPLUG_CPU) */
++
++#endif /* __XEN_CPU_HOTPLUG_H__ */
+diff --git a/include/xen/driver_util.h b/include/xen/driver_util.h
+new file mode 100644
+index 0000000..30a6f0e
+--- /dev/null
++++ b/include/xen/driver_util.h
+@@ -0,0 +1,10 @@
++
++#ifndef __ASM_XEN_DRIVER_UTIL_H__
++#define __ASM_XEN_DRIVER_UTIL_H__
++
++#include <linux/vmalloc.h>
++#include <linux/device.h>
++
++extern struct class *get_xen_class(void);
++
++#endif /* __ASM_XEN_DRIVER_UTIL_H__ */
+diff --git a/include/xen/evtchn.h b/include/xen/evtchn.h
+new file mode 100644
+index 0000000..350943d
+--- /dev/null
++++ b/include/xen/evtchn.h
+@@ -0,0 +1,178 @@
++/******************************************************************************
++ * evtchn.h
++ *
++ * Communication via Xen event channels.
++ * Also definitions for the device that demuxes notifications to userspace.
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __ASM_EVTCHN_H__
++#define __ASM_EVTCHN_H__
++
++#include <linux/interrupt.h>
++#include <asm/hypervisor.h>
++#include <asm/ptrace.h>
++#include <asm/synch_bitops.h>
++#include <xen/interface/event_channel.h>
++#include <linux/smp.h>
++
++/*
++ * LOW-LEVEL DEFINITIONS
++ */
++
++/*
++ * Dynamically bind an event source to an IRQ-like callback handler.
++ * On some platforms this may not be implemented via the Linux IRQ subsystem.
++ * The IRQ argument passed to the callback handler is the same as returned
++ * from the bind call. It may not correspond to a Linux IRQ number.
++ * Returns IRQ or negative errno.
++ */
++int bind_caller_port_to_irqhandler(
++ unsigned int caller_port,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_listening_port_to_irqhandler(
++ unsigned int remote_domain,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_interdomain_evtchn_to_irqhandler(
++ unsigned int remote_domain,
++ unsigned int remote_port,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_virq_to_irqhandler(
++ unsigned int virq,
++ unsigned int cpu,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++int bind_ipi_to_irqhandler(
++ unsigned int ipi,
++ unsigned int cpu,
++ irq_handler_t handler,
++ unsigned long irqflags,
++ const char *devname,
++ void *dev_id);
++
++/*
++ * Common unbind function for all event sources. Takes IRQ to unbind from.
++ * Automatically closes the underlying event channel (except for bindings
++ * made with bind_caller_port_to_irqhandler()).
++ */
++void unbind_from_irqhandler(unsigned int irq, void *dev_id);
++
++void irq_resume(void);
++
++/* Entry point for notifications into Linux subsystems. */
++asmlinkage void evtchn_do_upcall(struct pt_regs *regs);
++
++/* Entry point for notifications into the userland character device. */
++void evtchn_device_upcall(int port);
++
++void mask_evtchn(int port);
++void disable_all_local_evtchn(void);
++void unmask_evtchn(int port);
++
++#ifdef CONFIG_SMP
++void rebind_evtchn_to_cpu(int port, unsigned int cpu);
++#else
++#define rebind_evtchn_to_cpu(port, cpu) ((void)0)
++#endif
++
++static inline int test_and_set_evtchn_mask(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ return synch_test_and_set_bit(port, s->evtchn_mask);
++}
++
++static inline void clear_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ synch_clear_bit(port, s->evtchn_pending);
++}
++
++static inline void set_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ synch_set_bit(port, s->evtchn_pending);
++}
++
++static inline int test_evtchn(int port)
++{
++ shared_info_t *s = HYPERVISOR_shared_info;
++ return synch_test_bit(port, s->evtchn_pending);
++}
++
++static inline void notify_remote_via_evtchn(int port)
++{
++ struct evtchn_send send = { .port = port };
++ VOID(HYPERVISOR_event_channel_op(EVTCHNOP_send, &send));
++}
++
++/* Clear an irq's pending state, in preparation for polling on it. */
++void xen_clear_irq_pending(int irq);
++
++/* Set an irq's pending state, to avoid blocking on it. */
++void xen_set_irq_pending(int irq);
++
++/* Test an irq's pending state. */
++int xen_test_irq_pending(int irq);
++
++/* Poll waiting for an irq to become pending. In the usual case, the
++ irq will be disabled so it won't deliver an interrupt. */
++void xen_poll_irq(int irq);
++
++/*
++ * Use these to access the event channel underlying the IRQ handle returned
++ * by bind_*_to_irqhandler().
++ */
++void notify_remote_via_irq(int irq);
++int irq_to_evtchn_port(int irq);
++
++#define PIRQ_SET_MAPPING 0x0
++#define PIRQ_CLEAR_MAPPING 0x1
++#define PIRQ_GET_MAPPING 0x3
++int pirq_mapstatus(int pirq, int action);
++int set_pirq_hw_action(int pirq, int (*action)(int pirq, int action));
++int clear_pirq_hw_action(int pirq);
++
++#define PIRQ_STARTUP 1
++#define PIRQ_SHUTDOWN 2
++#define PIRQ_ENABLE 3
++#define PIRQ_DISABLE 4
++#define PIRQ_END 5
++#define PIRQ_ACK 6
++
++#endif /* __ASM_EVTCHN_H__ */
+diff --git a/include/xen/features.h b/include/xen/features.h
+index 27292d4..6c89605 100644
+--- a/include/xen/features.h
++++ b/include/xen/features.h
+@@ -10,6 +10,7 @@
+ #define __XEN_FEATURES_H__
+
+ #include <xen/interface/features.h>
++#include <xen/interface/version.h>
+
+ void xen_setup_features(void);
+
+@@ -20,4 +21,4 @@ static inline int xen_feature(int flag)
+ return xen_features[flag];
+ }
+
+-#endif /* __ASM_XEN_FEATURES_H__ */
++#endif /* __XEN_FEATURES_H__ */
+diff --git a/include/xen/firmware.h b/include/xen/firmware.h
+new file mode 100644
+index 0000000..905a9bc
+--- /dev/null
++++ b/include/xen/firmware.h
+@@ -0,0 +1,10 @@
++#ifndef __XEN_FIRMWARE_H__
++#define __XEN_FIRMWARE_H__
++
++#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE)
++void copy_edd(void);
++#endif
++
++void copy_edid(void);
++
++#endif /* __XEN_FIRMWARE_H__ */
+diff --git a/include/xen/gnttab.h b/include/xen/gnttab.h
+new file mode 100644
+index 0000000..8f13122
+--- /dev/null
++++ b/include/xen/gnttab.h
+@@ -0,0 +1,163 @@
++/******************************************************************************
++ * gnttab.h
++ *
++ * Two sets of functionality:
++ * 1. Granting foreign access to our memory reservation.
++ * 2. Accessing others' memory reservations via grant references.
++ * (i.e., mechanisms for both sender and recipient of grant references)
++ *
++ * Copyright (c) 2004-2005, K A Fraser
++ * Copyright (c) 2005, Christopher Clark
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __ASM_GNTTAB_H__
++#define __ASM_GNTTAB_H__
++
++#include <asm/hypervisor.h>
++#include <asm/maddr.h> /* maddr_t */
++#include <linux/mm.h>
++#include <xen/interface/grant_table.h>
++#include <xen/features.h>
++
++struct gnttab_free_callback {
++ struct gnttab_free_callback *next;
++ void (*fn)(void *);
++ void *arg;
++ u16 count;
++};
++
++int gnttab_grant_foreign_access(domid_t domid, unsigned long frame,
++ int flags);
++
++/*
++ * End access through the given grant reference, iff the grant entry is no
++ * longer in use. Return 1 if the grant entry was freed, 0 if it is still in
++ * use.
++ */
++int gnttab_end_foreign_access_ref(grant_ref_t ref);
++
++/*
++ * Eventually end access through the given grant reference, and once that
++ * access has been ended, free the given page too. Access will be ended
++ * immediately iff the grant entry is not in use, otherwise it will happen
++ * some time later. page may be 0, in which case no freeing will occur.
++ */
++void gnttab_end_foreign_access(grant_ref_t ref, unsigned long page);
++
++int gnttab_grant_foreign_transfer(domid_t domid, unsigned long pfn);
++
++unsigned long gnttab_end_foreign_transfer_ref(grant_ref_t ref);
++unsigned long gnttab_end_foreign_transfer(grant_ref_t ref);
++
++int gnttab_query_foreign_access(grant_ref_t ref);
++
++/*
++ * operations on reserved batches of grant references
++ */
++int gnttab_alloc_grant_references(u16 count, grant_ref_t *pprivate_head);
++
++void gnttab_free_grant_reference(grant_ref_t ref);
++
++void gnttab_free_grant_references(grant_ref_t head);
++
++int gnttab_empty_grant_references(const grant_ref_t *pprivate_head);
++
++int gnttab_claim_grant_reference(grant_ref_t *pprivate_head);
++
++void gnttab_release_grant_reference(grant_ref_t *private_head,
++ grant_ref_t release);
++
++void gnttab_request_free_callback(struct gnttab_free_callback *callback,
++ void (*fn)(void *), void *arg, u16 count);
++void gnttab_cancel_free_callback(struct gnttab_free_callback *callback);
++
++void gnttab_grant_foreign_access_ref(grant_ref_t ref, domid_t domid,
++ unsigned long frame, int flags);
++
++void gnttab_grant_foreign_transfer_ref(grant_ref_t, domid_t domid,
++ unsigned long pfn);
++
++int gnttab_copy_grant_page(grant_ref_t ref, struct page **pagep);
++void __gnttab_dma_map_page(struct page *page);
++static inline void __gnttab_dma_unmap_page(struct page *page)
++{
++}
++
++void gnttab_reset_grant_page(struct page *page);
++
++int gnttab_suspend(void);
++int gnttab_resume(void);
++
++void *arch_gnttab_alloc_shared(unsigned long *frames);
++
++static inline void
++gnttab_set_map_op(struct gnttab_map_grant_ref *map, maddr_t addr,
++ uint32_t flags, grant_ref_t ref, domid_t domid)
++{
++ if (flags & GNTMAP_contains_pte)
++ map->host_addr = addr;
++ else if (xen_feature(XENFEAT_auto_translated_physmap))
++ map->host_addr = __pa(addr);
++ else
++ map->host_addr = addr;
++
++ map->flags = flags;
++ map->ref = ref;
++ map->dom = domid;
++}
++
++static inline void
++gnttab_set_unmap_op(struct gnttab_unmap_grant_ref *unmap, maddr_t addr,
++ uint32_t flags, grant_handle_t handle)
++{
++ if (flags & GNTMAP_contains_pte)
++ unmap->host_addr = addr;
++ else if (xen_feature(XENFEAT_auto_translated_physmap))
++ unmap->host_addr = __pa(addr);
++ else
++ unmap->host_addr = addr;
++
++ unmap->handle = handle;
++ unmap->dev_bus_addr = 0;
++}
++
++static inline void
++gnttab_set_replace_op(struct gnttab_unmap_and_replace *unmap, maddr_t addr,
++ maddr_t new_addr, grant_handle_t handle)
++{
++ if (xen_feature(XENFEAT_auto_translated_physmap)) {
++ unmap->host_addr = __pa(addr);
++ unmap->new_addr = __pa(new_addr);
++ } else {
++ unmap->host_addr = addr;
++ unmap->new_addr = new_addr;
++ }
++
++ unmap->handle = handle;
++}
++
++#endif /* __ASM_GNTTAB_H__ */
+diff --git a/include/xen/hvm.h b/include/xen/hvm.h
+new file mode 100644
+index 0000000..4ea8887
+--- /dev/null
++++ b/include/xen/hvm.h
+@@ -0,0 +1,23 @@
++/* Simple wrappers around HVM functions */
++#ifndef XEN_HVM_H__
++#define XEN_HVM_H__
++
++#include <xen/interface/hvm/params.h>
++
++static inline unsigned long hvm_get_parameter(int idx)
++{
++ struct xen_hvm_param xhv;
++ int r;
++
++ xhv.domid = DOMID_SELF;
++ xhv.index = idx;
++ r = HYPERVISOR_hvm_op(HVMOP_get_param, &xhv);
++ if (r < 0) {
++ printk(KERN_ERR "cannot get hvm parameter %d: %d.\n",
++ idx, r);
++ return 0;
++ }
++ return xhv.value;
++}
++
++#endif /* XEN_HVM_H__ */
+diff --git a/include/xen/hypercall.h b/include/xen/hypercall.h
+new file mode 100644
+index 0000000..62071ea
+--- /dev/null
++++ b/include/xen/hypercall.h
+@@ -0,0 +1,30 @@
++#ifndef __XEN_HYPERCALL_H__
++#define __XEN_HYPERCALL_H__
++
++#include <asm/hypercall.h>
++
++static inline int __must_check
++HYPERVISOR_multicall_check(
++ multicall_entry_t *call_list, unsigned int nr_calls,
++ const unsigned long *rc_list)
++{
++ int rc = HYPERVISOR_multicall(call_list, nr_calls);
++
++ if (unlikely(rc < 0))
++ return rc;
++ BUG_ON(rc);
++ BUG_ON((int)nr_calls < 0);
++
++ for ( ; nr_calls > 0; --nr_calls, ++call_list)
++ if (unlikely(call_list->result != (rc_list ? *rc_list++ : 0)))
++ return nr_calls;
++
++ return 0;
++}
++
++/* A construct to ignore the return value of hypercall wrappers in a few
++ * exceptional cases (simply casting the function result to void doesn't
++ * avoid the compiler warning): */
++#define VOID(expr) ((void)((expr)?:0))
++
++#endif /* __XEN_HYPERCALL_H__ */
+diff --git a/include/xen/hypervisor_sysfs.h b/include/xen/hypervisor_sysfs.h
+new file mode 100644
+index 0000000..fed48a1
+--- /dev/null
++++ b/include/xen/hypervisor_sysfs.h
+@@ -0,0 +1,30 @@
++/*
++ * copyright (c) 2006 IBM Corporation
++ * Authored by: Mike D. Day <ncmike at us.ibm.com>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2 as
++ * published by the Free Software Foundation.
++ */
++
++#ifndef _HYP_SYSFS_H_
++#define _HYP_SYSFS_H_
++
++#include <linux/kobject.h>
++#include <linux/sysfs.h>
++
++#define HYPERVISOR_ATTR_RO(_name) \
++static struct hyp_sysfs_attr _name##_attr = __ATTR_RO(_name)
++
++#define HYPERVISOR_ATTR_RW(_name) \
++static struct hyp_sysfs_attr _name##_attr = \
++ __ATTR(_name, 0644, _name##_show, _name##_store)
++
++struct hyp_sysfs_attr {
++ struct attribute attr;
++ ssize_t (*show)(struct hyp_sysfs_attr *, char *);
++ ssize_t (*store)(struct hyp_sysfs_attr *, const char *, size_t);
++ void *hyp_attr_data;
++};
++
++#endif /* _HYP_SYSFS_H_ */
+diff --git a/include/xen/interface/COPYING b/include/xen/interface/COPYING
+new file mode 100644
+index 0000000..ffc6d61
+--- /dev/null
++++ b/include/xen/interface/COPYING
+@@ -0,0 +1,38 @@
++XEN NOTICE
++==========
++
++This copyright applies to all files within this subdirectory and its
++subdirectories:
++ include/public/*.h
++ include/public/hvm/*.h
++ include/public/io/*.h
++
++The intention is that these files can be freely copied into the source
++tree of an operating system when porting that OS to run on Xen. Doing
++so does *not* cause the OS to become subject to the terms of the GPL.
++
++All other files in the Xen source distribution are covered by version
++2 of the GNU General Public License except where explicitly stated
++otherwise within individual source files.
++
++ -- Keir Fraser (on behalf of the Xen team)
++
++=====================================================================
++
++Permission is hereby granted, free of charge, to any person obtaining a copy
++of this software and associated documentation files (the "Software"), to
++deal in the Software without restriction, including without limitation the
++rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++sell copies of the Software, and to permit persons to whom the Software is
++furnished to do so, subject to the following conditions:
++
++The above copyright notice and this permission notice shall be included in
++all copies or substantial portions of the Software.
++
++THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++DEALINGS IN THE SOFTWARE.
+diff --git a/include/xen/interface/acm.h b/include/xen/interface/acm.h
+new file mode 100644
+index 0000000..ef62da0
+--- /dev/null
++++ b/include/xen/interface/acm.h
+@@ -0,0 +1,228 @@
++/*
++ * acm.h: Xen access control module interface defintions
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005, International Business Machines Corporation.
++ */
++
++#ifndef _XEN_PUBLIC_ACM_H
++#define _XEN_PUBLIC_ACM_H
++
++#include "xen.h"
++
++/* if ACM_DEBUG defined, all hooks should
++ * print a short trace message (comment it out
++ * when not in testing mode )
++ */
++/* #define ACM_DEBUG */
++
++#ifdef ACM_DEBUG
++# define printkd(fmt, args...) printk(fmt,## args)
++#else
++# define printkd(fmt, args...)
++#endif
++
++/* default ssid reference value if not supplied */
++#define ACM_DEFAULT_SSID 0x0
++#define ACM_DEFAULT_LOCAL_SSID 0x0
++
++/* Internal ACM ERROR types */
++#define ACM_OK 0
++#define ACM_UNDEF -1
++#define ACM_INIT_SSID_ERROR -2
++#define ACM_INIT_SOID_ERROR -3
++#define ACM_ERROR -4
++
++/* External ACCESS DECISIONS */
++#define ACM_ACCESS_PERMITTED 0
++#define ACM_ACCESS_DENIED -111
++#define ACM_NULL_POINTER_ERROR -200
++
++/*
++ Error codes reported in when trying to test for a new policy
++ These error codes are reported in an array of tuples where
++ each error code is followed by a parameter describing the error
++ more closely, such as a domain id.
++*/
++#define ACM_EVTCHN_SHARING_VIOLATION 0x100
++#define ACM_GNTTAB_SHARING_VIOLATION 0x101
++#define ACM_DOMAIN_LOOKUP 0x102
++#define ACM_CHWALL_CONFLICT 0x103
++#define ACM_SSIDREF_IN_USE 0x104
++
++
++/* primary policy in lower 4 bits */
++#define ACM_NULL_POLICY 0
++#define ACM_CHINESE_WALL_POLICY 1
++#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2
++#define ACM_POLICY_UNDEFINED 15
++
++/* combinations have secondary policy component in higher 4bit */
++#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \
++ ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY)
++
++/* policy: */
++#define ACM_POLICY_NAME(X) \
++ ((X) == (ACM_NULL_POLICY)) ? "NULL" : \
++ ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \
++ ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \
++ ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \
++ "UNDEFINED"
++
++/* the following policy versions must be increased
++ * whenever the interpretation of the related
++ * policy's data structure changes
++ */
++#define ACM_POLICY_VERSION 3
++#define ACM_CHWALL_VERSION 1
++#define ACM_STE_VERSION 1
++
++/* defines a ssid reference used by xen */
++typedef uint32_t ssidref_t;
++
++/* hooks that are known to domains */
++#define ACMHOOK_none 0
++#define ACMHOOK_sharing 1
++
++/* -------security policy relevant type definitions-------- */
++
++/* type identifier; compares to "equal" or "not equal" */
++typedef uint16_t domaintype_t;
++
++/* CHINESE WALL POLICY DATA STRUCTURES
++ *
++ * current accumulated conflict type set:
++ * When a domain is started and has a type that is in
++ * a conflict set, the conflicting types are incremented in
++ * the aggregate set. When a domain is destroyed, the
++ * conflicting types to its type are decremented.
++ * If a domain has multiple types, this procedure works over
++ * all those types.
++ *
++ * conflict_aggregate_set[i] holds the number of
++ * running domains that have a conflict with type i.
++ *
++ * running_types[i] holds the number of running domains
++ * that include type i in their ssidref-referenced type set
++ *
++ * conflict_sets[i][j] is "0" if type j has no conflict
++ * with type i and is "1" otherwise.
++ */
++/* high-16 = version, low-16 = check magic */
++#define ACM_MAGIC 0x0001debc
++
++/* each offset in bytes from start of the struct they
++ * are part of */
++
++/* V3 of the policy buffer aded a version structure */
++struct acm_policy_version
++{
++ uint32_t major;
++ uint32_t minor;
++};
++
++
++/* each buffer consists of all policy information for
++ * the respective policy given in the policy code
++ *
++ * acm_policy_buffer, acm_chwall_policy_buffer,
++ * and acm_ste_policy_buffer need to stay 32-bit aligned
++ * because we create binary policies also with external
++ * tools that assume packed representations (e.g. the java tool)
++ */
++struct acm_policy_buffer {
++ uint32_t policy_version; /* ACM_POLICY_VERSION */
++ uint32_t magic;
++ uint32_t len;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_buffer_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_buffer_offset;
++ struct acm_policy_version xml_pol_version; /* add in V3 */
++};
++
++
++struct acm_policy_reference_buffer {
++ uint32_t len;
++};
++
++struct acm_chwall_policy_buffer {
++ uint32_t policy_version; /* ACM_CHWALL_VERSION */
++ uint32_t policy_code;
++ uint32_t chwall_max_types;
++ uint32_t chwall_max_ssidrefs;
++ uint32_t chwall_max_conflictsets;
++ uint32_t chwall_ssid_offset;
++ uint32_t chwall_conflict_sets_offset;
++ uint32_t chwall_running_types_offset;
++ uint32_t chwall_conflict_aggregate_offset;
++};
++
++struct acm_ste_policy_buffer {
++ uint32_t policy_version; /* ACM_STE_VERSION */
++ uint32_t policy_code;
++ uint32_t ste_max_types;
++ uint32_t ste_max_ssidrefs;
++ uint32_t ste_ssid_offset;
++};
++
++struct acm_stats_buffer {
++ uint32_t magic;
++ uint32_t len;
++ uint32_t primary_policy_code;
++ uint32_t primary_stats_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_stats_offset;
++};
++
++struct acm_ste_stats_buffer {
++ uint32_t ec_eval_count;
++ uint32_t gt_eval_count;
++ uint32_t ec_denied_count;
++ uint32_t gt_denied_count;
++ uint32_t ec_cachehit_count;
++ uint32_t gt_cachehit_count;
++};
++
++struct acm_ssid_buffer {
++ uint32_t len;
++ ssidref_t ssidref;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_max_types;
++ uint32_t primary_types_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_max_types;
++ uint32_t secondary_types_offset;
++};
++
++#endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/acm_ops.h b/include/xen/interface/acm_ops.h
+new file mode 100644
+index 0000000..27a8872
+--- /dev/null
++++ b/include/xen/interface/acm_ops.h
+@@ -0,0 +1,159 @@
++/*
++ * acm_ops.h: Xen access control module hypervisor commands
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005,2006 International Business Machines Corporation.
++ */
++
++#ifndef __XEN_PUBLIC_ACM_OPS_H__
++#define __XEN_PUBLIC_ACM_OPS_H__
++
++#include "xen.h"
++#include "acm.h"
++
++/*
++ * Make sure you increment the interface version whenever you modify this file!
++ * This makes sure that old versions of acm tools will stop working in a
++ * well-defined way (rather than crashing the machine, for instance).
++ */
++#define ACM_INTERFACE_VERSION 0xAAAA000A
++
++/************************************************************************/
++
++/*
++ * Prototype for this hypercall is:
++ * int acm_op(int cmd, void *args)
++ * @cmd == ACMOP_??? (access control module operation).
++ * @args == Operation-specific extra arguments (NULL if none).
++ */
++
++
++#define ACMOP_setpolicy 1
++struct acm_setpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pushcache;
++ uint32_t pushcache_size;
++};
++
++
++#define ACMOP_getpolicy 2
++struct acm_getpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
++
++
++#define ACMOP_dumpstats 3
++struct acm_dumpstats {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
++
++
++#define ACMOP_getssid 4
++#define ACM_GETBY_ssidref 1
++#define ACM_GETBY_domainid 2
++struct acm_getssid {
++ /* IN */
++ uint32_t get_ssid_by; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id;
++ XEN_GUEST_HANDLE_64(void) ssidbuf;
++ uint32_t ssidbuf_size;
++};
++
++#define ACMOP_getdecision 5
++struct acm_getdecision {
++ /* IN */
++ uint32_t get_decision_by1; /* ACM_GETBY_* */
++ uint32_t get_decision_by2; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id1;
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id2;
++ uint32_t hook;
++ /* OUT */
++ uint32_t acm_decision;
++};
++
++
++#define ACMOP_chgpolicy 6
++struct acm_change_policy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) policy_pushcache;
++ uint32_t policy_pushcache_size;
++ XEN_GUEST_HANDLE_64(void) del_array;
++ uint32_t delarray_size;
++ XEN_GUEST_HANDLE_64(void) chg_array;
++ uint32_t chgarray_size;
++ /* OUT */
++ /* array with error code */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
++
++#define ACMOP_relabeldoms 7
++struct acm_relabel_doms {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) relabel_map;
++ uint32_t relabel_map_size;
++ /* OUT */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
++
++/* future interface to Xen */
++struct xen_acmctl {
++ uint32_t cmd;
++ uint32_t interface_version;
++ union {
++ struct acm_setpolicy setpolicy;
++ struct acm_getpolicy getpolicy;
++ struct acm_dumpstats dumpstats;
++ struct acm_getssid getssid;
++ struct acm_getdecision getdecision;
++ struct acm_change_policy change_policy;
++ struct acm_relabel_doms relabel_doms;
++ } u;
++};
++
++typedef struct xen_acmctl xen_acmctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t);
++
++#endif /* __XEN_PUBLIC_ACM_OPS_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/arch-x86/cpuid.h b/include/xen/interface/arch-x86/cpuid.h
+new file mode 100644
+index 0000000..34fd72d
+--- /dev/null
++++ b/include/xen/interface/arch-x86/cpuid.h
+@@ -0,0 +1,68 @@
++/******************************************************************************
++ * arch-x86/cpuid.h
++ *
++ * CPUID interface to Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2007 Citrix Systems, Inc.
++ *
++ * Authors:
++ * Keir Fraser <keir.fraser at citrix.com>
++ */
++
++#ifndef __XEN_PUBLIC_ARCH_X86_CPUID_H__
++#define __XEN_PUBLIC_ARCH_X86_CPUID_H__
++
++/* Xen identification leaves start at 0x40000000. */
++#define XEN_CPUID_FIRST_LEAF 0x40000000
++#define XEN_CPUID_LEAF(i) (XEN_CPUID_FIRST_LEAF + (i))
++
++/*
++ * Leaf 1 (0x40000000)
++ * EAX: Largest Xen-information leaf. All leaves up to an including @EAX
++ * are supported by the Xen host.
++ * EBX-EDX: "XenVMMXenVMM" signature, allowing positive identification
++ * of a Xen host.
++ */
++#define XEN_CPUID_SIGNATURE_EBX 0x566e6558 /* "XenV" */
++#define XEN_CPUID_SIGNATURE_ECX 0x65584d4d /* "MMXe" */
++#define XEN_CPUID_SIGNATURE_EDX 0x4d4d566e /* "nVMM" */
++
++/*
++ * Leaf 2 (0x40000001)
++ * EAX[31:16]: Xen major version.
++ * EAX[15: 0]: Xen minor version.
++ * EBX-EDX: Reserved (currently all zeroes).
++ */
++
++/*
++ * Leaf 3 (0x40000002)
++ * EAX: Number of hypercall transfer pages. This register is always guaranteed
++ * to specify one hypercall page.
++ * EBX: Base address of Xen-specific MSRs.
++ * ECX: Features 1. Unused bits are set to zero.
++ * EDX: Features 2. Unused bits are set to zero.
++ */
++
++/* Does the host support MMU_PT_UPDATE_PRESERVE_AD for this guest? */
++#define _XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD 0
++#define XEN_CPUID_FEAT1_MMU_PT_UPDATE_PRESERVE_AD (1u<<0)
++
++#endif /* __XEN_PUBLIC_ARCH_X86_CPUID_H__ */
+diff --git a/include/xen/interface/arch-x86/hvm/save.h b/include/xen/interface/arch-x86/hvm/save.h
+new file mode 100644
+index 0000000..73b3909
+--- /dev/null
++++ b/include/xen/interface/arch-x86/hvm/save.h
+@@ -0,0 +1,429 @@
++/*
++ * Structure definitions for HVM state that is held by Xen and must
++ * be saved along with the domain's memory and device-model state.
++ *
++ * Copyright (c) 2007 XenSource Ltd.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_SAVE_X86_H__
++#define __XEN_PUBLIC_HVM_SAVE_X86_H__
++
++/*
++ * Save/restore header: general info about the save file.
++ */
++
++#define HVM_FILE_MAGIC 0x54381286
++#define HVM_FILE_VERSION 0x00000001
++
++struct hvm_save_header {
++ uint32_t magic; /* Must be HVM_FILE_MAGIC */
++ uint32_t version; /* File format version */
++ uint64_t changeset; /* Version of Xen that saved this file */
++ uint32_t cpuid; /* CPUID[0x01][%eax] on the saving machine */
++ uint32_t pad0;
++};
++
++DECLARE_HVM_SAVE_TYPE(HEADER, 1, struct hvm_save_header);
++
++
++/*
++ * Processor
++ */
++
++struct hvm_hw_cpu {
++ uint8_t fpu_regs[512];
++
++ uint64_t rax;
++ uint64_t rbx;
++ uint64_t rcx;
++ uint64_t rdx;
++ uint64_t rbp;
++ uint64_t rsi;
++ uint64_t rdi;
++ uint64_t rsp;
++ uint64_t r8;
++ uint64_t r9;
++ uint64_t r10;
++ uint64_t r11;
++ uint64_t r12;
++ uint64_t r13;
++ uint64_t r14;
++ uint64_t r15;
++
++ uint64_t rip;
++ uint64_t rflags;
++
++ uint64_t cr0;
++ uint64_t cr2;
++ uint64_t cr3;
++ uint64_t cr4;
++
++ uint64_t dr0;
++ uint64_t dr1;
++ uint64_t dr2;
++ uint64_t dr3;
++ uint64_t dr6;
++ uint64_t dr7;
++
++ uint32_t cs_sel;
++ uint32_t ds_sel;
++ uint32_t es_sel;
++ uint32_t fs_sel;
++ uint32_t gs_sel;
++ uint32_t ss_sel;
++ uint32_t tr_sel;
++ uint32_t ldtr_sel;
++
++ uint32_t cs_limit;
++ uint32_t ds_limit;
++ uint32_t es_limit;
++ uint32_t fs_limit;
++ uint32_t gs_limit;
++ uint32_t ss_limit;
++ uint32_t tr_limit;
++ uint32_t ldtr_limit;
++ uint32_t idtr_limit;
++ uint32_t gdtr_limit;
++
++ uint64_t cs_base;
++ uint64_t ds_base;
++ uint64_t es_base;
++ uint64_t fs_base;
++ uint64_t gs_base;
++ uint64_t ss_base;
++ uint64_t tr_base;
++ uint64_t ldtr_base;
++ uint64_t idtr_base;
++ uint64_t gdtr_base;
++
++ uint32_t cs_arbytes;
++ uint32_t ds_arbytes;
++ uint32_t es_arbytes;
++ uint32_t fs_arbytes;
++ uint32_t gs_arbytes;
++ uint32_t ss_arbytes;
++ uint32_t tr_arbytes;
++ uint32_t ldtr_arbytes;
++
++ uint32_t sysenter_cs;
++ uint32_t padding0;
++
++ uint64_t sysenter_esp;
++ uint64_t sysenter_eip;
++
++ /* msr for em64t */
++ uint64_t shadow_gs;
++
++ /* msr content saved/restored. */
++ uint64_t msr_flags;
++ uint64_t msr_lstar;
++ uint64_t msr_star;
++ uint64_t msr_cstar;
++ uint64_t msr_syscall_mask;
++ uint64_t msr_efer;
++
++ /* guest's idea of what rdtsc() would return */
++ uint64_t tsc;
++
++ /* pending event, if any */
++ union {
++ uint32_t pending_event;
++ struct {
++ uint8_t pending_vector:8;
++ uint8_t pending_type:3;
++ uint8_t pending_error_valid:1;
++ uint32_t pending_reserved:19;
++ uint8_t pending_valid:1;
++ };
++ };
++ /* error code for pending event */
++ uint32_t error_code;
++};
++
++DECLARE_HVM_SAVE_TYPE(CPU, 2, struct hvm_hw_cpu);
++
++
++/*
++ * PIC
++ */
++
++struct hvm_hw_vpic {
++ /* IR line bitmasks. */
++ uint8_t irr;
++ uint8_t imr;
++ uint8_t isr;
++
++ /* Line IRx maps to IRQ irq_base+x */
++ uint8_t irq_base;
++
++ /*
++ * Where are we in ICW2-4 initialisation (0 means no init in progress)?
++ * Bits 0-1 (=x): Next write at A=1 sets ICW(x+1).
++ * Bit 2: ICW1.IC4 (1 == ICW4 included in init sequence)
++ * Bit 3: ICW1.SNGL (0 == ICW3 included in init sequence)
++ */
++ uint8_t init_state:4;
++
++ /* IR line with highest priority. */
++ uint8_t priority_add:4;
++
++ /* Reads from A=0 obtain ISR or IRR? */
++ uint8_t readsel_isr:1;
++
++ /* Reads perform a polling read? */
++ uint8_t poll:1;
++
++ /* Automatically clear IRQs from the ISR during INTA? */
++ uint8_t auto_eoi:1;
++
++ /* Automatically rotate IRQ priorities during AEOI? */
++ uint8_t rotate_on_auto_eoi:1;
++
++ /* Exclude slave inputs when considering in-service IRQs? */
++ uint8_t special_fully_nested_mode:1;
++
++ /* Special mask mode excludes masked IRs from AEOI and priority checks. */
++ uint8_t special_mask_mode:1;
++
++ /* Is this a master PIC or slave PIC? (NB. This is not programmable.) */
++ uint8_t is_master:1;
++
++ /* Edge/trigger selection. */
++ uint8_t elcr;
++
++ /* Virtual INT output. */
++ uint8_t int_output;
++};
++
++DECLARE_HVM_SAVE_TYPE(PIC, 3, struct hvm_hw_vpic);
++
++
++/*
++ * IO-APIC
++ */
++
++#ifdef __ia64__
++#define VIOAPIC_IS_IOSAPIC 1
++#define VIOAPIC_NUM_PINS 24
++#else
++#define VIOAPIC_NUM_PINS 48 /* 16 ISA IRQs, 32 non-legacy PCI IRQS. */
++#endif
++
++struct hvm_hw_vioapic {
++ uint64_t base_address;
++ uint32_t ioregsel;
++ uint32_t id;
++ union vioapic_redir_entry
++ {
++ uint64_t bits;
++ struct {
++ uint8_t vector;
++ uint8_t delivery_mode:3;
++ uint8_t dest_mode:1;
++ uint8_t delivery_status:1;
++ uint8_t polarity:1;
++ uint8_t remote_irr:1;
++ uint8_t trig_mode:1;
++ uint8_t mask:1;
++ uint8_t reserve:7;
++#if !VIOAPIC_IS_IOSAPIC
++ uint8_t reserved[4];
++ uint8_t dest_id;
++#else
++ uint8_t reserved[3];
++ uint16_t dest_id;
++#endif
++ } fields;
++ } redirtbl[VIOAPIC_NUM_PINS];
++};
++
++DECLARE_HVM_SAVE_TYPE(IOAPIC, 4, struct hvm_hw_vioapic);
++
++
++/*
++ * LAPIC
++ */
++
++struct hvm_hw_lapic {
++ uint64_t apic_base_msr;
++ uint32_t disabled; /* VLAPIC_xx_DISABLED */
++ uint32_t timer_divisor;
++};
++
++DECLARE_HVM_SAVE_TYPE(LAPIC, 5, struct hvm_hw_lapic);
++
++struct hvm_hw_lapic_regs {
++ uint8_t data[1024];
++};
++
++DECLARE_HVM_SAVE_TYPE(LAPIC_REGS, 6, struct hvm_hw_lapic_regs);
++
++
++/*
++ * IRQs
++ */
++
++struct hvm_hw_pci_irqs {
++ /*
++ * Virtual interrupt wires for a single PCI bus.
++ * Indexed by: device*4 + INTx#.
++ */
++ union {
++ DECLARE_BITMAP(i, 32*4);
++ uint64_t pad[2];
++ };
++};
++
++DECLARE_HVM_SAVE_TYPE(PCI_IRQ, 7, struct hvm_hw_pci_irqs);
++
++struct hvm_hw_isa_irqs {
++ /*
++ * Virtual interrupt wires for ISA devices.
++ * Indexed by ISA IRQ (assumes no ISA-device IRQ sharing).
++ */
++ union {
++ DECLARE_BITMAP(i, 16);
++ uint64_t pad[1];
++ };
++};
++
++DECLARE_HVM_SAVE_TYPE(ISA_IRQ, 8, struct hvm_hw_isa_irqs);
++
++struct hvm_hw_pci_link {
++ /*
++ * PCI-ISA interrupt router.
++ * Each PCI <device:INTx#> is 'wire-ORed' into one of four links using
++ * the traditional 'barber's pole' mapping ((device + INTx#) & 3).
++ * The router provides a programmable mapping from each link to a GSI.
++ */
++ uint8_t route[4];
++ uint8_t pad0[4];
++};
++
++DECLARE_HVM_SAVE_TYPE(PCI_LINK, 9, struct hvm_hw_pci_link);
++
++/*
++ * PIT
++ */
++
++struct hvm_hw_pit {
++ struct hvm_hw_pit_channel {
++ uint32_t count; /* can be 65536 */
++ uint16_t latched_count;
++ uint8_t count_latched;
++ uint8_t status_latched;
++ uint8_t status;
++ uint8_t read_state;
++ uint8_t write_state;
++ uint8_t write_latch;
++ uint8_t rw_mode;
++ uint8_t mode;
++ uint8_t bcd; /* not supported */
++ uint8_t gate; /* timer start */
++ } channels[3]; /* 3 x 16 bytes */
++ uint32_t speaker_data_on;
++ uint32_t pad0;
++};
++
++DECLARE_HVM_SAVE_TYPE(PIT, 10, struct hvm_hw_pit);
++
++
++/*
++ * RTC
++ */
++
++#define RTC_CMOS_SIZE 14
++struct hvm_hw_rtc {
++ /* CMOS bytes */
++ uint8_t cmos_data[RTC_CMOS_SIZE];
++ /* Index register for 2-part operations */
++ uint8_t cmos_index;
++ uint8_t pad0;
++};
++
++DECLARE_HVM_SAVE_TYPE(RTC, 11, struct hvm_hw_rtc);
++
++
++/*
++ * HPET
++ */
++
++#define HPET_TIMER_NUM 3 /* 3 timers supported now */
++struct hvm_hw_hpet {
++ /* Memory-mapped, software visible registers */
++ uint64_t capability; /* capabilities */
++ uint64_t res0; /* reserved */
++ uint64_t config; /* configuration */
++ uint64_t res1; /* reserved */
++ uint64_t isr; /* interrupt status reg */
++ uint64_t res2[25]; /* reserved */
++ uint64_t mc64; /* main counter */
++ uint64_t res3; /* reserved */
++ struct { /* timers */
++ uint64_t config; /* configuration/cap */
++ uint64_t cmp; /* comparator */
++ uint64_t fsb; /* FSB route, not supported now */
++ uint64_t res4; /* reserved */
++ } timers[HPET_TIMER_NUM];
++ uint64_t res5[4*(24-HPET_TIMER_NUM)]; /* reserved, up to 0x3ff */
++
++ /* Hidden register state */
++ uint64_t period[HPET_TIMER_NUM]; /* Last value written to comparator */
++};
++
++DECLARE_HVM_SAVE_TYPE(HPET, 12, struct hvm_hw_hpet);
++
++
++/*
++ * PM timer
++ */
++
++struct hvm_hw_pmtimer {
++ uint32_t tmr_val; /* PM_TMR_BLK.TMR_VAL: 32bit free-running counter */
++ uint16_t pm1a_sts; /* PM1a_EVT_BLK.PM1a_STS: status register */
++ uint16_t pm1a_en; /* PM1a_EVT_BLK.PM1a_EN: enable register */
++};
++
++DECLARE_HVM_SAVE_TYPE(PMTIMER, 13, struct hvm_hw_pmtimer);
++
++/*
++ * MTRR MSRs
++ */
++
++struct hvm_hw_mtrr {
++#define MTRR_VCNT 8
++#define NUM_FIXED_MSR 11
++ uint64_t msr_pat_cr;
++ /* mtrr physbase & physmask msr pair*/
++ uint64_t msr_mtrr_var[MTRR_VCNT*2];
++ uint64_t msr_mtrr_fixed[NUM_FIXED_MSR];
++ uint64_t msr_mtrr_cap;
++ uint64_t msr_mtrr_def_type;
++};
++
++DECLARE_HVM_SAVE_TYPE(MTRR, 14, struct hvm_hw_mtrr);
++
++/*
++ * Largest type-code in use
++ */
++#define HVM_SAVE_CODE_MAX 14
++
++#endif /* __XEN_PUBLIC_HVM_SAVE_X86_H__ */
+diff --git a/include/xen/interface/arch-x86/xen-x86_32.h b/include/xen/interface/arch-x86/xen-x86_32.h
+new file mode 100644
+index 0000000..7cb6a01
+--- /dev/null
++++ b/include/xen/interface/arch-x86/xen-x86_32.h
+@@ -0,0 +1,180 @@
++/******************************************************************************
++ * xen-x86_32.h
++ *
++ * Guest OS interface to x86 32-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2007, K A Fraser
++ */
++
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__
++
++/*
++ * Hypercall interface:
++ * Input: %ebx, %ecx, %edx, %esi, %edi (arguments 1-5)
++ * Output: %eax
++ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
++ * call hypercall_page + hypercall-number * 32
++ * Clobbered: Argument registers (e.g., 2-arg hypercall clobbers %ebx,%ecx)
++ */
++
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
++/*
++ * Legacy hypercall interface:
++ * As above, except the entry sequence to the hypervisor is:
++ * mov $hypercall-number*32,%eax ; int $0x82
++ */
++#define TRAP_INSTR "int $0x82"
++#endif
++
++/*
++ * These flat segments are in the Xen-private section of every GDT. Since these
++ * are also present in the initial GDT, many OSes will be able to avoid
++ * installing their own GDT.
++ */
++#define FLAT_RING1_CS 0xe019 /* GDT index 259 */
++#define FLAT_RING1_DS 0xe021 /* GDT index 260 */
++#define FLAT_RING1_SS 0xe021 /* GDT index 260 */
++#define FLAT_RING3_CS 0xe02b /* GDT index 261 */
++#define FLAT_RING3_DS 0xe033 /* GDT index 262 */
++#define FLAT_RING3_SS 0xe033 /* GDT index 262 */
++
++#define FLAT_KERNEL_CS FLAT_RING1_CS
++#define FLAT_KERNEL_DS FLAT_RING1_DS
++#define FLAT_KERNEL_SS FLAT_RING1_SS
++#define FLAT_USER_CS FLAT_RING3_CS
++#define FLAT_USER_DS FLAT_RING3_DS
++#define FLAT_USER_SS FLAT_RING3_SS
++
++#define __HYPERVISOR_VIRT_START_PAE 0xF5800000
++#define __MACH2PHYS_VIRT_START_PAE 0xF5800000
++#define __MACH2PHYS_VIRT_END_PAE 0xF6800000
++#define HYPERVISOR_VIRT_START_PAE \
++ mk_unsigned_long(__HYPERVISOR_VIRT_START_PAE)
++#define MACH2PHYS_VIRT_START_PAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_START_PAE)
++#define MACH2PHYS_VIRT_END_PAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_END_PAE)
++
++/* Non-PAE bounds are obsolete. */
++#define __HYPERVISOR_VIRT_START_NONPAE 0xFC000000
++#define __MACH2PHYS_VIRT_START_NONPAE 0xFC000000
++#define __MACH2PHYS_VIRT_END_NONPAE 0xFC400000
++#define HYPERVISOR_VIRT_START_NONPAE \
++ mk_unsigned_long(__HYPERVISOR_VIRT_START_NONPAE)
++#define MACH2PHYS_VIRT_START_NONPAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_START_NONPAE)
++#define MACH2PHYS_VIRT_END_NONPAE \
++ mk_unsigned_long(__MACH2PHYS_VIRT_END_NONPAE)
++
++#define __HYPERVISOR_VIRT_START __HYPERVISOR_VIRT_START_PAE
++#define __MACH2PHYS_VIRT_START __MACH2PHYS_VIRT_START_PAE
++#define __MACH2PHYS_VIRT_END __MACH2PHYS_VIRT_END_PAE
++
++#ifndef HYPERVISOR_VIRT_START
++#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
++#endif
++
++#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
++#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
++#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>2)
++#ifndef machine_to_phys_mapping
++#define machine_to_phys_mapping ((unsigned long *)MACH2PHYS_VIRT_START)
++#endif
++
++/* 32-/64-bit invariability for control interfaces (domctl/sysctl). */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++#undef ___DEFINE_XEN_GUEST_HANDLE
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { type *p; } \
++ __guest_handle_ ## name; \
++ typedef struct { union { type *p; uint64_aligned_t q; }; } \
++ __guest_handle_64_ ## name
++#undef set_xen_guest_handle
++#define set_xen_guest_handle(hnd, val) \
++ do { if ( sizeof(hnd) == 8 ) *(uint64_t *)&(hnd) = 0; \
++ (hnd).p = val; \
++ } while ( 0 )
++#define uint64_aligned_t uint64_t __attribute__((aligned(8)))
++#define __XEN_GUEST_HANDLE_64(name) __guest_handle_64_ ## name
++#define XEN_GUEST_HANDLE_64(name) __XEN_GUEST_HANDLE_64(name)
++#endif
++
++#ifndef __ASSEMBLY__
++
++struct cpu_user_regs {
++ uint32_t ebx;
++ uint32_t ecx;
++ uint32_t edx;
++ uint32_t esi;
++ uint32_t edi;
++ uint32_t ebp;
++ uint32_t eax;
++ uint16_t error_code; /* private */
++ uint16_t entry_vector; /* private */
++ uint32_t eip;
++ uint16_t cs;
++ uint8_t saved_upcall_mask;
++ uint8_t _pad0;
++ uint32_t eflags; /* eflags.IF == !saved_upcall_mask */
++ uint32_t esp;
++ uint16_t ss, _pad1;
++ uint16_t es, _pad2;
++ uint16_t ds, _pad3;
++ uint16_t fs, _pad4;
++ uint16_t gs, _pad5;
++};
++typedef struct cpu_user_regs cpu_user_regs_t;
++DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
++
++/*
++ * Page-directory addresses above 4GB do not fit into architectural %cr3.
++ * When accessing %cr3, or equivalent field in vcpu_guest_context, guests
++ * must use the following accessor macros to pack/unpack valid MFNs.
++ */
++#define xen_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20))
++#define xen_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20))
++
++struct arch_vcpu_info {
++ unsigned long cr2;
++ unsigned long pad[5]; /* sizeof(vcpu_info_t) == 64 */
++};
++typedef struct arch_vcpu_info arch_vcpu_info_t;
++
++struct xen_callback {
++ unsigned long cs;
++ unsigned long eip;
++};
++typedef struct xen_callback xen_callback_t;
++
++#endif /* !__ASSEMBLY__ */
++
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_32_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/arch-x86/xen-x86_64.h b/include/xen/interface/arch-x86/xen-x86_64.h
+new file mode 100644
+index 0000000..1e54cf9
+--- /dev/null
++++ b/include/xen/interface/arch-x86/xen-x86_64.h
+@@ -0,0 +1,212 @@
++/******************************************************************************
++ * xen-x86_64.h
++ *
++ * Guest OS interface to x86 64-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
++
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__
++
++/*
++ * Hypercall interface:
++ * Input: %rdi, %rsi, %rdx, %r10, %r8 (arguments 1-5)
++ * Output: %rax
++ * Access is via hypercall page (set up by guest loader or via a Xen MSR):
++ * call hypercall_page + hypercall-number * 32
++ * Clobbered: argument registers (e.g., 2-arg hypercall clobbers %rdi,%rsi)
++ */
++
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
++/*
++ * Legacy hypercall interface:
++ * As above, except the entry sequence to the hypervisor is:
++ * mov $hypercall-number*32,%eax ; syscall
++ * Clobbered: %rcx, %r11, argument registers (as above)
++ */
++#define TRAP_INSTR "syscall"
++#endif
++
++/*
++ * 64-bit segment selectors
++ * These flat segments are in the Xen-private section of every GDT. Since these
++ * are also present in the initial GDT, many OSes will be able to avoid
++ * installing their own GDT.
++ */
++
++#define FLAT_RING3_CS32 0xe023 /* GDT index 260 */
++#define FLAT_RING3_CS64 0xe033 /* GDT index 261 */
++#define FLAT_RING3_DS32 0xe02b /* GDT index 262 */
++#define FLAT_RING3_DS64 0x0000 /* NULL selector */
++#define FLAT_RING3_SS32 0xe02b /* GDT index 262 */
++#define FLAT_RING3_SS64 0xe02b /* GDT index 262 */
++
++#define FLAT_KERNEL_DS64 FLAT_RING3_DS64
++#define FLAT_KERNEL_DS32 FLAT_RING3_DS32
++#define FLAT_KERNEL_DS FLAT_KERNEL_DS64
++#define FLAT_KERNEL_CS64 FLAT_RING3_CS64
++#define FLAT_KERNEL_CS32 FLAT_RING3_CS32
++#define FLAT_KERNEL_CS FLAT_KERNEL_CS64
++#define FLAT_KERNEL_SS64 FLAT_RING3_SS64
++#define FLAT_KERNEL_SS32 FLAT_RING3_SS32
++#define FLAT_KERNEL_SS FLAT_KERNEL_SS64
++
++#define FLAT_USER_DS64 FLAT_RING3_DS64
++#define FLAT_USER_DS32 FLAT_RING3_DS32
++#define FLAT_USER_DS FLAT_USER_DS64
++#define FLAT_USER_CS64 FLAT_RING3_CS64
++#define FLAT_USER_CS32 FLAT_RING3_CS32
++#define FLAT_USER_CS FLAT_USER_CS64
++#define FLAT_USER_SS64 FLAT_RING3_SS64
++#define FLAT_USER_SS32 FLAT_RING3_SS32
++#define FLAT_USER_SS FLAT_USER_SS64
++
++#define __HYPERVISOR_VIRT_START 0xFFFF800000000000
++#define __HYPERVISOR_VIRT_END 0xFFFF880000000000
++#define __MACH2PHYS_VIRT_START 0xFFFF800000000000
++#define __MACH2PHYS_VIRT_END 0xFFFF804000000000
++
++#ifndef HYPERVISOR_VIRT_START
++#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
++#define HYPERVISOR_VIRT_END mk_unsigned_long(__HYPERVISOR_VIRT_END)
++#endif
++
++#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
++#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
++#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3)
++#ifndef machine_to_phys_mapping
++#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
++#endif
++
++/*
++ * int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base)
++ * @which == SEGBASE_* ; @base == 64-bit base address
++ * Returns 0 on success.
++ */
++#define SEGBASE_FS 0
++#define SEGBASE_GS_USER 1
++#define SEGBASE_GS_KERNEL 2
++#define SEGBASE_GS_USER_SEL 3 /* Set user %gs specified in base[15:0] */
++
++/*
++ * int HYPERVISOR_iret(void)
++ * All arguments are on the kernel stack, in the following format.
++ * Never returns if successful. Current kernel context is lost.
++ * The saved CS is mapped as follows:
++ * RING0 -> RING3 kernel mode.
++ * RING1 -> RING3 kernel mode.
++ * RING2 -> RING3 kernel mode.
++ * RING3 -> RING3 user mode.
++ * However RING0 indicates that the guest kernel should return to iteself
++ * directly with
++ * orb $3,1*8(%rsp)
++ * iretq
++ * If flags contains VGCF_in_syscall:
++ * Restore RAX, RIP, RFLAGS, RSP.
++ * Discard R11, RCX, CS, SS.
++ * Otherwise:
++ * Restore RAX, R11, RCX, CS:RIP, RFLAGS, SS:RSP.
++ * All other registers are saved on hypercall entry and restored to user.
++ */
++/* Guest exited in SYSCALL context? Return to guest with SYSRET? */
++#define _VGCF_in_syscall 8
++#define VGCF_in_syscall (1<<_VGCF_in_syscall)
++#define VGCF_IN_SYSCALL VGCF_in_syscall
++
++#ifndef __ASSEMBLY__
++
++struct iret_context {
++ /* Top of stack (%rsp at point of hypercall). */
++ uint64_t rax, r11, rcx, flags, rip, cs, rflags, rsp, ss;
++ /* Bottom of iret stack frame. */
++};
++
++#if defined(__GNUC__) && !defined(__STRICT_ANSI__)
++/* Anonymous union includes both 32- and 64-bit names (e.g., eax/rax). */
++#define __DECL_REG(name) union { \
++ uint64_t r ## name, e ## name; \
++ uint32_t _e ## name; \
++}
++#else
++/* Non-gcc sources must always use the proper 64-bit name (e.g., rax). */
++#define __DECL_REG(name) uint64_t r ## name
++#endif
++
++struct cpu_user_regs {
++ uint64_t r15;
++ uint64_t r14;
++ uint64_t r13;
++ uint64_t r12;
++ __DECL_REG(bp);
++ __DECL_REG(bx);
++ uint64_t r11;
++ uint64_t r10;
++ uint64_t r9;
++ uint64_t r8;
++ __DECL_REG(ax);
++ __DECL_REG(cx);
++ __DECL_REG(dx);
++ __DECL_REG(si);
++ __DECL_REG(di);
++ uint32_t error_code; /* private */
++ uint32_t entry_vector; /* private */
++ __DECL_REG(ip);
++ uint16_t cs, _pad0[1];
++ uint8_t saved_upcall_mask;
++ uint8_t _pad1[3];
++ __DECL_REG(flags); /* rflags.IF == !saved_upcall_mask */
++ __DECL_REG(sp);
++ uint16_t ss, _pad2[3];
++ uint16_t es, _pad3[3];
++ uint16_t ds, _pad4[3];
++ uint16_t fs, _pad5[3]; /* Non-zero => takes precedence over fs_base. */
++ uint16_t gs, _pad6[3]; /* Non-zero => takes precedence over gs_base_usr. */
++};
++typedef struct cpu_user_regs cpu_user_regs_t;
++DEFINE_XEN_GUEST_HANDLE(cpu_user_regs_t);
++
++#undef __DECL_REG
++
++#define xen_pfn_to_cr3(pfn) ((unsigned long)(pfn) << 12)
++#define xen_cr3_to_pfn(cr3) ((unsigned long)(cr3) >> 12)
++
++struct arch_vcpu_info {
++ unsigned long cr2;
++ unsigned long pad; /* sizeof(vcpu_info_t) == 64 */
++};
++typedef struct arch_vcpu_info arch_vcpu_info_t;
++
++typedef unsigned long xen_callback_t;
++
++#endif /* !__ASSEMBLY__ */
++
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_X86_64_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/arch-x86/xen.h b/include/xen/interface/arch-x86/xen.h
+new file mode 100644
+index 0000000..320b470
+--- /dev/null
++++ b/include/xen/interface/arch-x86/xen.h
+@@ -0,0 +1,203 @@
++/******************************************************************************
++ * arch-x86/xen.h
++ *
++ * Guest OS interface to x86 Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
++
++#include "../xen.h"
++
++#ifndef __XEN_PUBLIC_ARCH_X86_XEN_H__
++#define __XEN_PUBLIC_ARCH_X86_XEN_H__
++
++/* Structural guest handles introduced in 0x00030201. */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030201
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef struct { type *p; } __guest_handle_ ## name
++#else
++#define ___DEFINE_XEN_GUEST_HANDLE(name, type) \
++ typedef type * __guest_handle_ ## name
++#endif
++
++#define __DEFINE_XEN_GUEST_HANDLE(name, type) \
++ ___DEFINE_XEN_GUEST_HANDLE(name, type); \
++ ___DEFINE_XEN_GUEST_HANDLE(const_##name, const type)
++#define DEFINE_XEN_GUEST_HANDLE(name) __DEFINE_XEN_GUEST_HANDLE(name, name)
++#define __XEN_GUEST_HANDLE(name) __guest_handle_ ## name
++#define XEN_GUEST_HANDLE(name) __XEN_GUEST_HANDLE(name)
++#define set_xen_guest_handle(hnd, val) do { (hnd).p = val; } while (0)
++#ifdef __XEN_TOOLS__
++#define get_xen_guest_handle(val, hnd) do { val = (hnd).p; } while (0)
++#endif
++
++/* Allow co-existing Linux 2.6.23+ Xen interface definitions. */
++#define DEFINE_XEN_GUEST_HANDLE_STRUCT(name) struct name
++
++#if defined(__i386__)
++#include "xen-x86_32.h"
++#elif defined(__x86_64__)
++#include "xen-x86_64.h"
++#endif
++
++#ifndef __ASSEMBLY__
++typedef unsigned long xen_pfn_t;
++#define PRI_xen_pfn "lx"
++#endif
++
++/*
++ * SEGMENT DESCRIPTOR TABLES
++ */
++/*
++ * A number of GDT entries are reserved by Xen. These are not situated at the
++ * start of the GDT because some stupid OSes export hard-coded selector values
++ * in their ABI. These hard-coded values are always near the start of the GDT,
++ * so Xen places itself out of the way, at the far end of the GDT.
++ */
++#define FIRST_RESERVED_GDT_PAGE 14
++#define FIRST_RESERVED_GDT_BYTE (FIRST_RESERVED_GDT_PAGE * 4096)
++#define FIRST_RESERVED_GDT_ENTRY (FIRST_RESERVED_GDT_BYTE / 8)
++
++/* Maximum number of virtual CPUs in multi-processor guests. */
++#define MAX_VIRT_CPUS 32
++
++#ifndef __ASSEMBLY__
++
++typedef unsigned long xen_ulong_t;
++
++/*
++ * Send an array of these to HYPERVISOR_set_trap_table().
++ * The privilege level specifies which modes may enter a trap via a software
++ * interrupt. On x86/64, since rings 1 and 2 are unavailable, we allocate
++ * privilege levels as follows:
++ * Level == 0: Noone may enter
++ * Level == 1: Kernel may enter
++ * Level == 2: Kernel may enter
++ * Level == 3: Everyone may enter
++ */
++#define TI_GET_DPL(_ti) ((_ti)->flags & 3)
++#define TI_GET_IF(_ti) ((_ti)->flags & 4)
++#define TI_SET_DPL(_ti,_dpl) ((_ti)->flags |= (_dpl))
++#define TI_SET_IF(_ti,_if) ((_ti)->flags |= ((!!(_if))<<2))
++struct trap_info {
++ uint8_t vector; /* exception vector */
++ uint8_t flags; /* 0-3: privilege level; 4: clear event enable? */
++ uint16_t cs; /* code selector */
++ unsigned long address; /* code offset */
++};
++typedef struct trap_info trap_info_t;
++DEFINE_XEN_GUEST_HANDLE(trap_info_t);
++
++typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */
++
++/*
++ * The following is all CPU context. Note that the fpu_ctxt block is filled
++ * in by FXSAVE if the CPU has feature FXSR; otherwise FSAVE is used.
++ */
++struct vcpu_guest_context {
++ /* FPU registers come first so they can be aligned for FXSAVE/FXRSTOR. */
++ struct { char x[512]; } fpu_ctxt; /* User-level FPU registers */
++#define VGCF_I387_VALID (1<<0)
++#define VGCF_IN_KERNEL (1<<2)
++#define _VGCF_i387_valid 0
++#define VGCF_i387_valid (1<<_VGCF_i387_valid)
++#define _VGCF_in_kernel 2
++#define VGCF_in_kernel (1<<_VGCF_in_kernel)
++#define _VGCF_failsafe_disables_events 3
++#define VGCF_failsafe_disables_events (1<<_VGCF_failsafe_disables_events)
++#define _VGCF_syscall_disables_events 4
++#define VGCF_syscall_disables_events (1<<_VGCF_syscall_disables_events)
++#define _VGCF_online 5
++#define VGCF_online (1<<_VGCF_online)
++ unsigned long flags; /* VGCF_* flags */
++ struct cpu_user_regs user_regs; /* User-level CPU registers */
++ struct trap_info trap_ctxt[256]; /* Virtual IDT */
++ unsigned long ldt_base, ldt_ents; /* LDT (linear address, # ents) */
++ unsigned long gdt_frames[16], gdt_ents; /* GDT (machine frames, # ents) */
++ unsigned long kernel_ss, kernel_sp; /* Virtual TSS (only SS1/SP1) */
++ /* NB. User pagetable on x86/64 is placed in ctrlreg[1]. */
++ unsigned long ctrlreg[8]; /* CR0-CR7 (control registers) */
++ unsigned long debugreg[8]; /* DB0-DB7 (debug registers) */
++#ifdef __i386__
++ unsigned long event_callback_cs; /* CS:EIP of event callback */
++ unsigned long event_callback_eip;
++ unsigned long failsafe_callback_cs; /* CS:EIP of failsafe callback */
++ unsigned long failsafe_callback_eip;
++#else
++ unsigned long event_callback_eip;
++ unsigned long failsafe_callback_eip;
++#ifdef __XEN__
++ union {
++ unsigned long syscall_callback_eip;
++ struct {
++ unsigned int event_callback_cs; /* compat CS of event cb */
++ unsigned int failsafe_callback_cs; /* compat CS of failsafe cb */
++ };
++ };
++#else
++ unsigned long syscall_callback_eip;
++#endif
++#endif
++ unsigned long vm_assist; /* VMASST_TYPE_* bitmap */
++#ifdef __x86_64__
++ /* Segment base addresses. */
++ uint64_t fs_base;
++ uint64_t gs_base_kernel;
++ uint64_t gs_base_user;
++#endif
++};
++typedef struct vcpu_guest_context vcpu_guest_context_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_guest_context_t);
++
++struct arch_shared_info {
++ unsigned long max_pfn; /* max pfn that appears in table */
++ /* Frame containing list of mfns containing list of mfns containing p2m. */
++ xen_pfn_t pfn_to_mfn_frame_list_list;
++ unsigned long nmi_reason;
++ uint64_t pad[32];
++};
++typedef struct arch_shared_info arch_shared_info_t;
++
++#endif /* !__ASSEMBLY__ */
++
++/*
++ * Prefix forces emulation of some non-trapping instructions.
++ * Currently only CPUID.
++ */
++#ifdef __ASSEMBLY__
++#define XEN_EMULATE_PREFIX .byte 0x0f,0x0b,0x78,0x65,0x6e ;
++#define XEN_CPUID XEN_EMULATE_PREFIX cpuid
++#else
++#define XEN_EMULATE_PREFIX ".byte 0x0f,0x0b,0x78,0x65,0x6e ; "
++#define XEN_CPUID XEN_EMULATE_PREFIX "cpuid"
++#endif
++
++#endif /* __XEN_PUBLIC_ARCH_X86_XEN_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/arch-x86_32.h b/include/xen/interface/arch-x86_32.h
+new file mode 100644
+index 0000000..45842b2
+--- /dev/null
++++ b/include/xen/interface/arch-x86_32.h
+@@ -0,0 +1,27 @@
++/******************************************************************************
++ * arch-x86_32.h
++ *
++ * Guest OS interface to x86 32-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
++
++#include "arch-x86/xen.h"
+diff --git a/include/xen/interface/arch-x86_64.h b/include/xen/interface/arch-x86_64.h
+new file mode 100644
+index 0000000..fbb2639
+--- /dev/null
++++ b/include/xen/interface/arch-x86_64.h
+@@ -0,0 +1,27 @@
++/******************************************************************************
++ * arch-x86_64.h
++ *
++ * Guest OS interface to x86 64-bit Xen.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2004-2006, K A Fraser
++ */
++
++#include "arch-x86/xen.h"
+diff --git a/include/xen/interface/callback.h b/include/xen/interface/callback.h
+index 4aadcba..f4962f6 100644
+--- a/include/xen/interface/callback.h
++++ b/include/xen/interface/callback.h
+@@ -84,8 +84,10 @@
+ struct callback_register {
+ uint16_t type;
+ uint16_t flags;
+- struct xen_callback address;
++ xen_callback_t address;
+ };
++typedef struct callback_register callback_register_t;
++DEFINE_XEN_GUEST_HANDLE(callback_register_t);
+
+ /*
+ * Unregister a callback.
+@@ -98,5 +100,22 @@ struct callback_unregister {
+ uint16_t type;
+ uint16_t _unused;
+ };
++typedef struct callback_unregister callback_unregister_t;
++DEFINE_XEN_GUEST_HANDLE(callback_unregister_t);
++
++#if __XEN_INTERFACE_VERSION__ < 0x00030207
++#undef CALLBACKTYPE_sysenter
++#define CALLBACKTYPE_sysenter CALLBACKTYPE_sysenter_deprecated
++#endif
+
+ #endif /* __XEN_PUBLIC_CALLBACK_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/dom0_ops.h b/include/xen/interface/dom0_ops.h
+new file mode 100644
+index 0000000..5d2b324
+--- /dev/null
++++ b/include/xen/interface/dom0_ops.h
+@@ -0,0 +1,120 @@
++/******************************************************************************
++ * dom0_ops.h
++ *
++ * Process command requests from domain-0 guest OS.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2002-2003, B Dragovic
++ * Copyright (c) 2002-2006, K Fraser
++ */
++
++#ifndef __XEN_PUBLIC_DOM0_OPS_H__
++#define __XEN_PUBLIC_DOM0_OPS_H__
++
++#include "xen.h"
++#include "platform.h"
++
++#if __XEN_INTERFACE_VERSION__ >= 0x00030204
++#error "dom0_ops.h is a compatibility interface only"
++#endif
++
++#define DOM0_INTERFACE_VERSION XENPF_INTERFACE_VERSION
++
++#define DOM0_SETTIME XENPF_settime
++#define dom0_settime xenpf_settime
++#define dom0_settime_t xenpf_settime_t
++
++#define DOM0_ADD_MEMTYPE XENPF_add_memtype
++#define dom0_add_memtype xenpf_add_memtype
++#define dom0_add_memtype_t xenpf_add_memtype_t
++
++#define DOM0_DEL_MEMTYPE XENPF_del_memtype
++#define dom0_del_memtype xenpf_del_memtype
++#define dom0_del_memtype_t xenpf_del_memtype_t
++
++#define DOM0_READ_MEMTYPE XENPF_read_memtype
++#define dom0_read_memtype xenpf_read_memtype
++#define dom0_read_memtype_t xenpf_read_memtype_t
++
++#define DOM0_MICROCODE XENPF_microcode_update
++#define dom0_microcode xenpf_microcode_update
++#define dom0_microcode_t xenpf_microcode_update_t
++
++#define DOM0_PLATFORM_QUIRK XENPF_platform_quirk
++#define dom0_platform_quirk xenpf_platform_quirk
++#define dom0_platform_quirk_t xenpf_platform_quirk_t
++
++typedef uint64_t cpumap_t;
++
++/* Unsupported legacy operation -- defined for API compatibility. */
++#define DOM0_MSR 15
++struct dom0_msr {
++ /* IN variables. */
++ uint32_t write;
++ cpumap_t cpu_mask;
++ uint32_t msr;
++ uint32_t in1;
++ uint32_t in2;
++ /* OUT variables. */
++ uint32_t out1;
++ uint32_t out2;
++};
++typedef struct dom0_msr dom0_msr_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_msr_t);
++
++/* Unsupported legacy operation -- defined for API compatibility. */
++#define DOM0_PHYSICAL_MEMORY_MAP 40
++struct dom0_memory_map_entry {
++ uint64_t start, end;
++ uint32_t flags; /* reserved */
++ uint8_t is_ram;
++};
++typedef struct dom0_memory_map_entry dom0_memory_map_entry_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_memory_map_entry_t);
++
++struct dom0_op {
++ uint32_t cmd;
++ uint32_t interface_version; /* DOM0_INTERFACE_VERSION */
++ union {
++ struct dom0_msr msr;
++ struct dom0_settime settime;
++ struct dom0_add_memtype add_memtype;
++ struct dom0_del_memtype del_memtype;
++ struct dom0_read_memtype read_memtype;
++ struct dom0_microcode microcode;
++ struct dom0_platform_quirk platform_quirk;
++ struct dom0_memory_map_entry physical_memory_map;
++ uint8_t pad[128];
++ } u;
++};
++typedef struct dom0_op dom0_op_t;
++DEFINE_XEN_GUEST_HANDLE(dom0_op_t);
++
++#endif /* __XEN_PUBLIC_DOM0_OPS_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/domctl.h b/include/xen/interface/domctl.h
+new file mode 100644
+index 0000000..65e2b64
+--- /dev/null
++++ b/include/xen/interface/domctl.h
+@@ -0,0 +1,649 @@
++/******************************************************************************
++ * domctl.h
++ *
++ * Domain management operations. For use by node control stack.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2002-2003, B Dragovic
++ * Copyright (c) 2002-2006, K Fraser
++ */
++
++#ifndef __XEN_PUBLIC_DOMCTL_H__
++#define __XEN_PUBLIC_DOMCTL_H__
++
++#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
++#error "domctl operations are intended for use by node control tools only"
++#endif
++
++#include "xen.h"
++
++#define XEN_DOMCTL_INTERFACE_VERSION 0x00000005
++
++struct xenctl_cpumap {
++ XEN_GUEST_HANDLE_64(uint8) bitmap;
++ uint32_t nr_cpus;
++};
++
++/*
++ * NB. xen_domctl.domain is an IN/OUT parameter for this operation.
++ * If it is specified as zero, an id is auto-allocated and returned.
++ */
++#define XEN_DOMCTL_createdomain 1
++struct xen_domctl_createdomain {
++ /* IN parameters */
++ uint32_t ssidref;
++ xen_domain_handle_t handle;
++ /* Is this an HVM guest (as opposed to a PV guest)? */
++#define _XEN_DOMCTL_CDF_hvm_guest 0
++#define XEN_DOMCTL_CDF_hvm_guest (1U<<_XEN_DOMCTL_CDF_hvm_guest)
++ /* Use hardware-assisted paging if available? */
++#define _XEN_DOMCTL_CDF_hap 1
++#define XEN_DOMCTL_CDF_hap (1U<<_XEN_DOMCTL_CDF_hap)
++ uint32_t flags;
++};
++typedef struct xen_domctl_createdomain xen_domctl_createdomain_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_createdomain_t);
++
++#define XEN_DOMCTL_destroydomain 2
++#define XEN_DOMCTL_pausedomain 3
++#define XEN_DOMCTL_unpausedomain 4
++#define XEN_DOMCTL_resumedomain 27
++
++#define XEN_DOMCTL_getdomaininfo 5
++struct xen_domctl_getdomaininfo {
++ /* OUT variables. */
++ domid_t domain; /* Also echoed in domctl.domain */
++ /* Domain is scheduled to die. */
++#define _XEN_DOMINF_dying 0
++#define XEN_DOMINF_dying (1U<<_XEN_DOMINF_dying)
++ /* Domain is an HVM guest (as opposed to a PV guest). */
++#define _XEN_DOMINF_hvm_guest 1
++#define XEN_DOMINF_hvm_guest (1U<<_XEN_DOMINF_hvm_guest)
++ /* The guest OS has shut down. */
++#define _XEN_DOMINF_shutdown 2
++#define XEN_DOMINF_shutdown (1U<<_XEN_DOMINF_shutdown)
++ /* Currently paused by control software. */
++#define _XEN_DOMINF_paused 3
++#define XEN_DOMINF_paused (1U<<_XEN_DOMINF_paused)
++ /* Currently blocked pending an event. */
++#define _XEN_DOMINF_blocked 4
++#define XEN_DOMINF_blocked (1U<<_XEN_DOMINF_blocked)
++ /* Domain is currently running. */
++#define _XEN_DOMINF_running 5
++#define XEN_DOMINF_running (1U<<_XEN_DOMINF_running)
++ /* Being debugged. */
++#define _XEN_DOMINF_debugged 6
++#define XEN_DOMINF_debugged (1U<<_XEN_DOMINF_debugged)
++ /* CPU to which this domain is bound. */
++#define XEN_DOMINF_cpumask 255
++#define XEN_DOMINF_cpushift 8
++ /* XEN_DOMINF_shutdown guest-supplied code. */
++#define XEN_DOMINF_shutdownmask 255
++#define XEN_DOMINF_shutdownshift 16
++ uint32_t flags; /* XEN_DOMINF_* */
++ uint64_aligned_t tot_pages;
++ uint64_aligned_t max_pages;
++ uint64_aligned_t shared_info_frame; /* GMFN of shared_info struct */
++ uint64_aligned_t cpu_time;
++ uint32_t nr_online_vcpus; /* Number of VCPUs currently online. */
++ uint32_t max_vcpu_id; /* Maximum VCPUID in use by this domain. */
++ uint32_t ssidref;
++ xen_domain_handle_t handle;
++};
++typedef struct xen_domctl_getdomaininfo xen_domctl_getdomaininfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getdomaininfo_t);
++
++
++#define XEN_DOMCTL_getmemlist 6
++struct xen_domctl_getmemlist {
++ /* IN variables. */
++ /* Max entries to write to output buffer. */
++ uint64_aligned_t max_pfns;
++ /* Start index in guest's page list. */
++ uint64_aligned_t start_pfn;
++ XEN_GUEST_HANDLE_64(uint64) buffer;
++ /* OUT variables. */
++ uint64_aligned_t num_pfns;
++};
++typedef struct xen_domctl_getmemlist xen_domctl_getmemlist_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getmemlist_t);
++
++
++#define XEN_DOMCTL_getpageframeinfo 7
++
++#define XEN_DOMCTL_PFINFO_LTAB_SHIFT 28
++#define XEN_DOMCTL_PFINFO_NOTAB (0x0U<<28)
++#define XEN_DOMCTL_PFINFO_L1TAB (0x1U<<28)
++#define XEN_DOMCTL_PFINFO_L2TAB (0x2U<<28)
++#define XEN_DOMCTL_PFINFO_L3TAB (0x3U<<28)
++#define XEN_DOMCTL_PFINFO_L4TAB (0x4U<<28)
++#define XEN_DOMCTL_PFINFO_LTABTYPE_MASK (0x7U<<28)
++#define XEN_DOMCTL_PFINFO_LPINTAB (0x1U<<31)
++#define XEN_DOMCTL_PFINFO_XTAB (0xfU<<28) /* invalid page */
++#define XEN_DOMCTL_PFINFO_LTAB_MASK (0xfU<<28)
++
++struct xen_domctl_getpageframeinfo {
++ /* IN variables. */
++ uint64_aligned_t gmfn; /* GMFN to query */
++ /* OUT variables. */
++ /* Is the page PINNED to a type? */
++ uint32_t type; /* see above type defs */
++};
++typedef struct xen_domctl_getpageframeinfo xen_domctl_getpageframeinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo_t);
++
++
++#define XEN_DOMCTL_getpageframeinfo2 8
++struct xen_domctl_getpageframeinfo2 {
++ /* IN variables. */
++ uint64_aligned_t num;
++ /* IN/OUT variables. */
++ XEN_GUEST_HANDLE_64(uint32) array;
++};
++typedef struct xen_domctl_getpageframeinfo2 xen_domctl_getpageframeinfo2_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getpageframeinfo2_t);
++
++
++/*
++ * Control shadow pagetables operation
++ */
++#define XEN_DOMCTL_shadow_op 10
++
++/* Disable shadow mode. */
++#define XEN_DOMCTL_SHADOW_OP_OFF 0
++
++/* Enable shadow mode (mode contains ORed XEN_DOMCTL_SHADOW_ENABLE_* flags). */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE 32
++
++/* Log-dirty bitmap operations. */
++ /* Return the bitmap and clean internal copy for next round. */
++#define XEN_DOMCTL_SHADOW_OP_CLEAN 11
++ /* Return the bitmap but do not modify internal copy. */
++#define XEN_DOMCTL_SHADOW_OP_PEEK 12
++
++/* Memory allocation accessors. */
++#define XEN_DOMCTL_SHADOW_OP_GET_ALLOCATION 30
++#define XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION 31
++
++/* Legacy enable operations. */
++ /* Equiv. to ENABLE with no mode flags. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_TEST 1
++ /* Equiv. to ENABLE with mode flag ENABLE_LOG_DIRTY. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_LOGDIRTY 2
++ /* Equiv. to ENABLE with mode flags ENABLE_REFCOUNT and ENABLE_TRANSLATE. */
++#define XEN_DOMCTL_SHADOW_OP_ENABLE_TRANSLATE 3
++
++/* Mode flags for XEN_DOMCTL_SHADOW_OP_ENABLE. */
++ /*
++ * Shadow pagetables are refcounted: guest does not use explicit mmu
++ * operations nor write-protect its pagetables.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_REFCOUNT (1 << 1)
++ /*
++ * Log pages in a bitmap as they are dirtied.
++ * Used for live relocation to determine which pages must be re-sent.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_LOG_DIRTY (1 << 2)
++ /*
++ * Automatically translate GPFNs into MFNs.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_TRANSLATE (1 << 3)
++ /*
++ * Xen does not steal virtual address space from the guest.
++ * Requires HVM support.
++ */
++#define XEN_DOMCTL_SHADOW_ENABLE_EXTERNAL (1 << 4)
++
++struct xen_domctl_shadow_op_stats {
++ uint32_t fault_count;
++ uint32_t dirty_count;
++};
++typedef struct xen_domctl_shadow_op_stats xen_domctl_shadow_op_stats_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_stats_t);
++
++struct xen_domctl_shadow_op {
++ /* IN variables. */
++ uint32_t op; /* XEN_DOMCTL_SHADOW_OP_* */
++
++ /* OP_ENABLE */
++ uint32_t mode; /* XEN_DOMCTL_SHADOW_ENABLE_* */
++
++ /* OP_GET_ALLOCATION / OP_SET_ALLOCATION */
++ uint32_t mb; /* Shadow memory allocation in MB */
++
++ /* OP_PEEK / OP_CLEAN */
++ XEN_GUEST_HANDLE_64(uint8) dirty_bitmap;
++ uint64_aligned_t pages; /* Size of buffer. Updated with actual size. */
++ struct xen_domctl_shadow_op_stats stats;
++};
++typedef struct xen_domctl_shadow_op xen_domctl_shadow_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_shadow_op_t);
++
++
++#define XEN_DOMCTL_max_mem 11
++struct xen_domctl_max_mem {
++ /* IN variables. */
++ uint64_aligned_t max_memkb;
++};
++typedef struct xen_domctl_max_mem xen_domctl_max_mem_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_mem_t);
++
++
++#define XEN_DOMCTL_setvcpucontext 12
++#define XEN_DOMCTL_getvcpucontext 13
++struct xen_domctl_vcpucontext {
++ uint32_t vcpu; /* IN */
++ XEN_GUEST_HANDLE_64(vcpu_guest_context_t) ctxt; /* IN/OUT */
++};
++typedef struct xen_domctl_vcpucontext xen_domctl_vcpucontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpucontext_t);
++
++
++#define XEN_DOMCTL_getvcpuinfo 14
++struct xen_domctl_getvcpuinfo {
++ /* IN variables. */
++ uint32_t vcpu;
++ /* OUT variables. */
++ uint8_t online; /* currently online (not hotplugged)? */
++ uint8_t blocked; /* blocked waiting for an event? */
++ uint8_t running; /* currently scheduled on its CPU? */
++ uint64_aligned_t cpu_time; /* total cpu time consumed (ns) */
++ uint32_t cpu; /* current mapping */
++};
++typedef struct xen_domctl_getvcpuinfo xen_domctl_getvcpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_getvcpuinfo_t);
++
++
++/* Get/set which physical cpus a vcpu can execute on. */
++#define XEN_DOMCTL_setvcpuaffinity 9
++#define XEN_DOMCTL_getvcpuaffinity 25
++struct xen_domctl_vcpuaffinity {
++ uint32_t vcpu; /* IN */
++ struct xenctl_cpumap cpumap; /* IN/OUT */
++};
++typedef struct xen_domctl_vcpuaffinity xen_domctl_vcpuaffinity_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_vcpuaffinity_t);
++
++
++#define XEN_DOMCTL_max_vcpus 15
++struct xen_domctl_max_vcpus {
++ uint32_t max; /* maximum number of vcpus */
++};
++typedef struct xen_domctl_max_vcpus xen_domctl_max_vcpus_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_max_vcpus_t);
++
++
++#define XEN_DOMCTL_scheduler_op 16
++/* Scheduler types. */
++#define XEN_SCHEDULER_SEDF 4
++#define XEN_SCHEDULER_CREDIT 5
++/* Set or get info? */
++#define XEN_DOMCTL_SCHEDOP_putinfo 0
++#define XEN_DOMCTL_SCHEDOP_getinfo 1
++struct xen_domctl_scheduler_op {
++ uint32_t sched_id; /* XEN_SCHEDULER_* */
++ uint32_t cmd; /* XEN_DOMCTL_SCHEDOP_* */
++ union {
++ struct xen_domctl_sched_sedf {
++ uint64_aligned_t period;
++ uint64_aligned_t slice;
++ uint64_aligned_t latency;
++ uint32_t extratime;
++ uint32_t weight;
++ } sedf;
++ struct xen_domctl_sched_credit {
++ uint16_t weight;
++ uint16_t cap;
++ } credit;
++ } u;
++};
++typedef struct xen_domctl_scheduler_op xen_domctl_scheduler_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_scheduler_op_t);
++
++
++#define XEN_DOMCTL_setdomainhandle 17
++struct xen_domctl_setdomainhandle {
++ xen_domain_handle_t handle;
++};
++typedef struct xen_domctl_setdomainhandle xen_domctl_setdomainhandle_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdomainhandle_t);
++
++
++#define XEN_DOMCTL_setdebugging 18
++struct xen_domctl_setdebugging {
++ uint8_t enable;
++};
++typedef struct xen_domctl_setdebugging xen_domctl_setdebugging_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_setdebugging_t);
++
++
++#define XEN_DOMCTL_irq_permission 19
++struct xen_domctl_irq_permission {
++ uint8_t pirq;
++ uint8_t allow_access; /* flag to specify enable/disable of IRQ access */
++};
++typedef struct xen_domctl_irq_permission xen_domctl_irq_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_irq_permission_t);
++
++
++#define XEN_DOMCTL_iomem_permission 20
++struct xen_domctl_iomem_permission {
++ uint64_aligned_t first_mfn;/* first page (physical page number) in range */
++ uint64_aligned_t nr_mfns; /* number of pages in range (>0) */
++ uint8_t allow_access; /* allow (!0) or deny (0) access to range? */
++};
++typedef struct xen_domctl_iomem_permission xen_domctl_iomem_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_iomem_permission_t);
++
++
++#define XEN_DOMCTL_ioport_permission 21
++struct xen_domctl_ioport_permission {
++ uint32_t first_port; /* first port int range */
++ uint32_t nr_ports; /* size of port range */
++ uint8_t allow_access; /* allow or deny access to range? */
++};
++typedef struct xen_domctl_ioport_permission xen_domctl_ioport_permission_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_permission_t);
++
++
++#define XEN_DOMCTL_hypercall_init 22
++struct xen_domctl_hypercall_init {
++ uint64_aligned_t gmfn; /* GMFN to be initialised */
++};
++typedef struct xen_domctl_hypercall_init xen_domctl_hypercall_init_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_hypercall_init_t);
++
++
++#define XEN_DOMCTL_arch_setup 23
++#define _XEN_DOMAINSETUP_hvm_guest 0
++#define XEN_DOMAINSETUP_hvm_guest (1UL<<_XEN_DOMAINSETUP_hvm_guest)
++#define _XEN_DOMAINSETUP_query 1 /* Get parameters (for save) */
++#define XEN_DOMAINSETUP_query (1UL<<_XEN_DOMAINSETUP_query)
++#define _XEN_DOMAINSETUP_sioemu_guest 2
++#define XEN_DOMAINSETUP_sioemu_guest (1UL<<_XEN_DOMAINSETUP_sioemu_guest)
++typedef struct xen_domctl_arch_setup {
++ uint64_aligned_t flags; /* XEN_DOMAINSETUP_* */
++#ifdef __ia64__
++ uint64_aligned_t bp; /* mpaddr of boot param area */
++ uint64_aligned_t maxmem; /* Highest memory address for MDT. */
++ uint64_aligned_t xsi_va; /* Xen shared_info area virtual address. */
++ uint32_t hypercall_imm; /* Break imm for Xen hypercalls. */
++ int8_t vhpt_size_log2; /* Log2 of VHPT size. */
++#endif
++} xen_domctl_arch_setup_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_arch_setup_t);
++
++
++#define XEN_DOMCTL_settimeoffset 24
++struct xen_domctl_settimeoffset {
++ int32_t time_offset_seconds; /* applied to domain wallclock time */
++};
++typedef struct xen_domctl_settimeoffset xen_domctl_settimeoffset_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_settimeoffset_t);
++
++
++#define XEN_DOMCTL_gethvmcontext 33
++#define XEN_DOMCTL_sethvmcontext 34
++typedef struct xen_domctl_hvmcontext {
++ uint32_t size; /* IN/OUT: size of buffer / bytes filled */
++ XEN_GUEST_HANDLE_64(uint8) buffer; /* IN/OUT: data, or call
++ * gethvmcontext with NULL
++ * buffer to get size req'd */
++} xen_domctl_hvmcontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_hvmcontext_t);
++
++
++#define XEN_DOMCTL_set_address_size 35
++#define XEN_DOMCTL_get_address_size 36
++typedef struct xen_domctl_address_size {
++ uint32_t size;
++} xen_domctl_address_size_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_address_size_t);
++
++
++#define XEN_DOMCTL_real_mode_area 26
++struct xen_domctl_real_mode_area {
++ uint32_t log; /* log2 of Real Mode Area size */
++};
++typedef struct xen_domctl_real_mode_area xen_domctl_real_mode_area_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_real_mode_area_t);
++
++
++#define XEN_DOMCTL_sendtrigger 28
++#define XEN_DOMCTL_SENDTRIGGER_NMI 0
++#define XEN_DOMCTL_SENDTRIGGER_RESET 1
++#define XEN_DOMCTL_SENDTRIGGER_INIT 2
++struct xen_domctl_sendtrigger {
++ uint32_t trigger; /* IN */
++ uint32_t vcpu; /* IN */
++};
++typedef struct xen_domctl_sendtrigger xen_domctl_sendtrigger_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_sendtrigger_t);
++
++
++/* Assign PCI device to HVM guest. Sets up IOMMU structures. */
++#define XEN_DOMCTL_assign_device 37
++#define XEN_DOMCTL_test_assign_device 45
++#define XEN_DOMCTL_deassign_device 47
++struct xen_domctl_assign_device {
++ uint32_t machine_bdf; /* machine PCI ID of assigned device */
++};
++typedef struct xen_domctl_assign_device xen_domctl_assign_device_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_assign_device_t);
++
++
++/* Pass-through interrupts: bind real irq -> hvm devfn. */
++#define XEN_DOMCTL_bind_pt_irq 38
++#define XEN_DOMCTL_unbind_pt_irq 48
++typedef enum pt_irq_type_e {
++ PT_IRQ_TYPE_PCI,
++ PT_IRQ_TYPE_ISA,
++ PT_IRQ_TYPE_MSI,
++} pt_irq_type_t;
++struct xen_domctl_bind_pt_irq {
++ uint32_t machine_irq;
++ pt_irq_type_t irq_type;
++ uint32_t hvm_domid;
++
++ union {
++ struct {
++ uint8_t isa_irq;
++ } isa;
++ struct {
++ uint8_t bus;
++ uint8_t device;
++ uint8_t intx;
++ } pci;
++ struct {
++ uint8_t gvec;
++ uint32_t gflags;
++ } msi;
++ } u;
++};
++typedef struct xen_domctl_bind_pt_irq xen_domctl_bind_pt_irq_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_bind_pt_irq_t);
++
++
++/* Bind machine I/O address range -> HVM address range. */
++#define XEN_DOMCTL_memory_mapping 39
++#define DPCI_ADD_MAPPING 1
++#define DPCI_REMOVE_MAPPING 0
++struct xen_domctl_memory_mapping {
++ uint64_aligned_t first_gfn; /* first page (hvm guest phys page) in range */
++ uint64_aligned_t first_mfn; /* first page (machine page) in range */
++ uint64_aligned_t nr_mfns; /* number of pages in range (>0) */
++ uint32_t add_mapping; /* add or remove mapping */
++ uint32_t padding; /* padding for 64-bit aligned structure */
++};
++typedef struct xen_domctl_memory_mapping xen_domctl_memory_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_memory_mapping_t);
++
++
++/* Bind machine I/O port range -> HVM I/O port range. */
++#define XEN_DOMCTL_ioport_mapping 40
++struct xen_domctl_ioport_mapping {
++ uint32_t first_gport; /* first guest IO port*/
++ uint32_t first_mport; /* first machine IO port */
++ uint32_t nr_ports; /* size of port range */
++ uint32_t add_mapping; /* add or remove mapping */
++};
++typedef struct xen_domctl_ioport_mapping xen_domctl_ioport_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ioport_mapping_t);
++
++
++/*
++ * Pin caching type of RAM space for x86 HVM domU.
++ */
++#define XEN_DOMCTL_pin_mem_cacheattr 41
++/* Caching types: these happen to be the same as x86 MTRR/PAT type codes. */
++#define XEN_DOMCTL_MEM_CACHEATTR_UC 0
++#define XEN_DOMCTL_MEM_CACHEATTR_WC 1
++#define XEN_DOMCTL_MEM_CACHEATTR_WT 4
++#define XEN_DOMCTL_MEM_CACHEATTR_WP 5
++#define XEN_DOMCTL_MEM_CACHEATTR_WB 6
++#define XEN_DOMCTL_MEM_CACHEATTR_UCM 7
++struct xen_domctl_pin_mem_cacheattr {
++ uint64_aligned_t start, end;
++ unsigned int type; /* XEN_DOMCTL_MEM_CACHEATTR_* */
++};
++typedef struct xen_domctl_pin_mem_cacheattr xen_domctl_pin_mem_cacheattr_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_pin_mem_cacheattr_t);
++
++
++#define XEN_DOMCTL_set_ext_vcpucontext 42
++#define XEN_DOMCTL_get_ext_vcpucontext 43
++struct xen_domctl_ext_vcpucontext {
++ /* IN: VCPU that this call applies to. */
++ uint32_t vcpu;
++ /*
++ * SET: Size of struct (IN)
++ * GET: Size of struct (OUT)
++ */
++ uint32_t size;
++#if defined(__i386__) || defined(__x86_64__)
++ /* SYSCALL from 32-bit mode and SYSENTER callback information. */
++ /* NB. SYSCALL from 64-bit mode is contained in vcpu_guest_context_t */
++ uint64_aligned_t syscall32_callback_eip;
++ uint64_aligned_t sysenter_callback_eip;
++ uint16_t syscall32_callback_cs;
++ uint16_t sysenter_callback_cs;
++ uint8_t syscall32_disables_events;
++ uint8_t sysenter_disables_events;
++#endif
++};
++typedef struct xen_domctl_ext_vcpucontext xen_domctl_ext_vcpucontext_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_ext_vcpucontext_t);
++
++/*
++ * Set optimizaton features for a domain
++ */
++#define XEN_DOMCTL_set_opt_feature 44
++struct xen_domctl_set_opt_feature {
++#if defined(__ia64__)
++ struct xen_ia64_opt_feature optf;
++#else
++ /* Make struct non-empty: do not depend on this field name! */
++ uint64_t dummy;
++#endif
++};
++typedef struct xen_domctl_set_opt_feature xen_domctl_set_opt_feature_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_opt_feature_t);
++
++/*
++ * Set the target domain for a domain
++ */
++#define XEN_DOMCTL_set_target 46
++struct xen_domctl_set_target {
++ domid_t target;
++};
++typedef struct xen_domctl_set_target xen_domctl_set_target_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_set_target_t);
++
++#if defined(__i386__) || defined(__x86_64__)
++# define XEN_CPUID_INPUT_UNUSED 0xFFFFFFFF
++# define XEN_DOMCTL_set_cpuid 49
++struct xen_domctl_cpuid {
++ unsigned int input[2];
++ unsigned int eax;
++ unsigned int ebx;
++ unsigned int ecx;
++ unsigned int edx;
++};
++typedef struct xen_domctl_cpuid xen_domctl_cpuid_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_cpuid_t);
++#endif
++
++struct xen_domctl {
++ uint32_t cmd;
++ uint32_t interface_version; /* XEN_DOMCTL_INTERFACE_VERSION */
++ domid_t domain;
++ union {
++ struct xen_domctl_createdomain createdomain;
++ struct xen_domctl_getdomaininfo getdomaininfo;
++ struct xen_domctl_getmemlist getmemlist;
++ struct xen_domctl_getpageframeinfo getpageframeinfo;
++ struct xen_domctl_getpageframeinfo2 getpageframeinfo2;
++ struct xen_domctl_vcpuaffinity vcpuaffinity;
++ struct xen_domctl_shadow_op shadow_op;
++ struct xen_domctl_max_mem max_mem;
++ struct xen_domctl_vcpucontext vcpucontext;
++ struct xen_domctl_getvcpuinfo getvcpuinfo;
++ struct xen_domctl_max_vcpus max_vcpus;
++ struct xen_domctl_scheduler_op scheduler_op;
++ struct xen_domctl_setdomainhandle setdomainhandle;
++ struct xen_domctl_setdebugging setdebugging;
++ struct xen_domctl_irq_permission irq_permission;
++ struct xen_domctl_iomem_permission iomem_permission;
++ struct xen_domctl_ioport_permission ioport_permission;
++ struct xen_domctl_hypercall_init hypercall_init;
++ struct xen_domctl_arch_setup arch_setup;
++ struct xen_domctl_settimeoffset settimeoffset;
++ struct xen_domctl_real_mode_area real_mode_area;
++ struct xen_domctl_hvmcontext hvmcontext;
++ struct xen_domctl_address_size address_size;
++ struct xen_domctl_sendtrigger sendtrigger;
++ struct xen_domctl_assign_device assign_device;
++ struct xen_domctl_bind_pt_irq bind_pt_irq;
++ struct xen_domctl_memory_mapping memory_mapping;
++ struct xen_domctl_ioport_mapping ioport_mapping;
++ struct xen_domctl_pin_mem_cacheattr pin_mem_cacheattr;
++ struct xen_domctl_ext_vcpucontext ext_vcpucontext;
++ struct xen_domctl_set_opt_feature set_opt_feature;
++ struct xen_domctl_set_target set_target;
++#if defined(__i386__) || defined(__x86_64__)
++ struct xen_domctl_cpuid cpuid;
++#endif
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_domctl xen_domctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_domctl_t);
++
++#endif /* __XEN_PUBLIC_DOMCTL_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/elfnote.h b/include/xen/interface/elfnote.h
+index a64d3df..4215c6e 100644
+--- a/include/xen/interface/elfnote.h
++++ b/include/xen/interface/elfnote.h
+@@ -3,6 +3,24 @@
+ *
+ * Definitions used for the Xen ELF notes.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2006, Ian Campbell, XenSource Ltd.
+ */
+
+@@ -10,7 +28,7 @@
+ #define __XEN_PUBLIC_ELFNOTE_H__
+
+ /*
+- * The notes should live in a SHT_NOTE segment and have "Xen" in the
++ * The notes should live in a PT_NOTE segment and have "Xen" in the
+ * name field.
+ *
+ * Numeric types are either 4 or 8 bytes depending on the content of
+@@ -22,8 +40,6 @@
+
+ /*
+ * NAME=VALUE pair (string).
+- *
+- * LEGACY: FEATURES and PAE
+ */
+ #define XEN_ELFNOTE_INFO 0
+
+@@ -90,7 +106,12 @@
+ #define XEN_ELFNOTE_LOADER 8
+
+ /*
+- * The kernel supports PAE (x86/32 only, string = "yes" or "no").
++ * The kernel supports PAE (x86/32 only, string = "yes", "no" or
++ * "bimodal").
++ *
++ * For compatibility with Xen 3.0.3 and earlier the "bimodal" setting
++ * may be given as "yes,bimodal" which will cause older Xen to treat
++ * this kernel as PAE.
+ *
+ * LEGACY: PAE (n.b. The legacy interface included a provision to
+ * indicate 'extended-cr3' support allowing L3 page tables to be
+@@ -120,6 +141,85 @@
+ */
+ #define XEN_ELFNOTE_BSD_SYMTAB 11
+
++/*
++ * The lowest address the hypervisor hole can begin at (numeric).
++ *
++ * This must not be set higher than HYPERVISOR_VIRT_START. Its presence
++ * also indicates to the hypervisor that the kernel can deal with the
++ * hole starting at a higher address.
++ */
++#define XEN_ELFNOTE_HV_START_LOW 12
++
++/*
++ * List of maddr_t-sized mask/value pairs describing how to recognize
++ * (non-present) L1 page table entries carrying valid MFNs (numeric).
++ */
++#define XEN_ELFNOTE_L1_MFN_VALID 13
++
++/*
++ * Whether or not the guest supports cooperative suspend cancellation.
++ */
++#define XEN_ELFNOTE_SUSPEND_CANCEL 14
++
++/*
++ * The number of the highest elfnote defined.
++ */
++#define XEN_ELFNOTE_MAX XEN_ELFNOTE_SUSPEND_CANCEL
++
++/*
++ * System information exported through crash notes.
++ *
++ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_INFO
++ * note in case of a system crash. This note will contain various
++ * information about the system, see xen/include/xen/elfcore.h.
++ */
++#define XEN_ELFNOTE_CRASH_INFO 0x1000001
++
++/*
++ * System registers exported through crash notes.
++ *
++ * The kexec / kdump code will create one XEN_ELFNOTE_CRASH_REGS
++ * note per cpu in case of a system crash. This note is architecture
++ * specific and will contain registers not saved in the "CORE" note.
++ * See xen/include/xen/elfcore.h for more information.
++ */
++#define XEN_ELFNOTE_CRASH_REGS 0x1000002
++
++
++/*
++ * xen dump-core none note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_NONE
++ * in its dump file to indicate that the file is xen dump-core
++ * file. This note doesn't have any other information.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_NONE 0x2000000
++
++/*
++ * xen dump-core header note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_HEADER
++ * in its dump file.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_HEADER 0x2000001
++
++/*
++ * xen dump-core xen version note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_XEN_VERSION
++ * in its dump file. It contains the xen version obtained via the
++ * XENVER hypercall.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_XEN_VERSION 0x2000002
++
++/*
++ * xen dump-core format version note.
++ * xm dump-core code will create one XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION
++ * in its dump file. It contains a format version identifier.
++ * See tools/libxc/xc_core.h for more information.
++ */
++#define XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION 0x2000003
++
+ #endif /* __XEN_PUBLIC_ELFNOTE_H__ */
+
+ /*
+diff --git a/include/xen/interface/elfstructs.h b/include/xen/interface/elfstructs.h
+new file mode 100644
+index 0000000..77362f3
+--- /dev/null
++++ b/include/xen/interface/elfstructs.h
+@@ -0,0 +1,527 @@
++#ifndef __XEN_PUBLIC_ELFSTRUCTS_H__
++#define __XEN_PUBLIC_ELFSTRUCTS_H__ 1
++/*
++ * Copyright (c) 1995, 1996 Erik Theisen. All rights reserved.
++ *
++ * Redistribution and use in source and binary forms, with or without
++ * modification, are permitted provided that the following conditions
++ * are met:
++ * 1. Redistributions of source code must retain the above copyright
++ * notice, this list of conditions and the following disclaimer.
++ * 2. Redistributions in binary form must reproduce the above copyright
++ * notice, this list of conditions and the following disclaimer in the
++ * documentation and/or other materials provided with the distribution.
++ * 3. The name of the author may not be used to endorse or promote products
++ * derived from this software without specific prior written permission
++ *
++ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
++ * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
++ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
++ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
++ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
++ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
++ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
++ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
++ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
++ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
++ */
++
++typedef uint8_t Elf_Byte;
++
++typedef uint32_t Elf32_Addr; /* Unsigned program address */
++typedef uint32_t Elf32_Off; /* Unsigned file offset */
++typedef int32_t Elf32_Sword; /* Signed large integer */
++typedef uint32_t Elf32_Word; /* Unsigned large integer */
++typedef uint16_t Elf32_Half; /* Unsigned medium integer */
++
++typedef uint64_t Elf64_Addr;
++typedef uint64_t Elf64_Off;
++typedef int32_t Elf64_Shalf;
++
++typedef int32_t Elf64_Sword;
++typedef uint32_t Elf64_Word;
++
++typedef int64_t Elf64_Sxword;
++typedef uint64_t Elf64_Xword;
++
++typedef uint32_t Elf64_Half;
++typedef uint16_t Elf64_Quarter;
++
++/*
++ * e_ident[] identification indexes
++ * See http://www.caldera.com/developers/gabi/2000-07-17/ch4.eheader.html
++ */
++#define EI_MAG0 0 /* file ID */
++#define EI_MAG1 1 /* file ID */
++#define EI_MAG2 2 /* file ID */
++#define EI_MAG3 3 /* file ID */
++#define EI_CLASS 4 /* file class */
++#define EI_DATA 5 /* data encoding */
++#define EI_VERSION 6 /* ELF header version */
++#define EI_OSABI 7 /* OS/ABI ID */
++#define EI_ABIVERSION 8 /* ABI version */
++#define EI_PAD 9 /* start of pad bytes */
++#define EI_NIDENT 16 /* Size of e_ident[] */
++
++/* e_ident[] magic number */
++#define ELFMAG0 0x7f /* e_ident[EI_MAG0] */
++#define ELFMAG1 'E' /* e_ident[EI_MAG1] */
++#define ELFMAG2 'L' /* e_ident[EI_MAG2] */
++#define ELFMAG3 'F' /* e_ident[EI_MAG3] */
++#define ELFMAG "\177ELF" /* magic */
++#define SELFMAG 4 /* size of magic */
++
++/* e_ident[] file class */
++#define ELFCLASSNONE 0 /* invalid */
++#define ELFCLASS32 1 /* 32-bit objs */
++#define ELFCLASS64 2 /* 64-bit objs */
++#define ELFCLASSNUM 3 /* number of classes */
++
++/* e_ident[] data encoding */
++#define ELFDATANONE 0 /* invalid */
++#define ELFDATA2LSB 1 /* Little-Endian */
++#define ELFDATA2MSB 2 /* Big-Endian */
++#define ELFDATANUM 3 /* number of data encode defines */
++
++/* e_ident[] Operating System/ABI */
++#define ELFOSABI_SYSV 0 /* UNIX System V ABI */
++#define ELFOSABI_HPUX 1 /* HP-UX operating system */
++#define ELFOSABI_NETBSD 2 /* NetBSD */
++#define ELFOSABI_LINUX 3 /* GNU/Linux */
++#define ELFOSABI_HURD 4 /* GNU/Hurd */
++#define ELFOSABI_86OPEN 5 /* 86Open common IA32 ABI */
++#define ELFOSABI_SOLARIS 6 /* Solaris */
++#define ELFOSABI_MONTEREY 7 /* Monterey */
++#define ELFOSABI_IRIX 8 /* IRIX */
++#define ELFOSABI_FREEBSD 9 /* FreeBSD */
++#define ELFOSABI_TRU64 10 /* TRU64 UNIX */
++#define ELFOSABI_MODESTO 11 /* Novell Modesto */
++#define ELFOSABI_OPENBSD 12 /* OpenBSD */
++#define ELFOSABI_ARM 97 /* ARM */
++#define ELFOSABI_STANDALONE 255 /* Standalone (embedded) application */
++
++/* e_ident */
++#define IS_ELF(ehdr) ((ehdr).e_ident[EI_MAG0] == ELFMAG0 && \
++ (ehdr).e_ident[EI_MAG1] == ELFMAG1 && \
++ (ehdr).e_ident[EI_MAG2] == ELFMAG2 && \
++ (ehdr).e_ident[EI_MAG3] == ELFMAG3)
++
++/* ELF Header */
++typedef struct elfhdr {
++ unsigned char e_ident[EI_NIDENT]; /* ELF Identification */
++ Elf32_Half e_type; /* object file type */
++ Elf32_Half e_machine; /* machine */
++ Elf32_Word e_version; /* object file version */
++ Elf32_Addr e_entry; /* virtual entry point */
++ Elf32_Off e_phoff; /* program header table offset */
++ Elf32_Off e_shoff; /* section header table offset */
++ Elf32_Word e_flags; /* processor-specific flags */
++ Elf32_Half e_ehsize; /* ELF header size */
++ Elf32_Half e_phentsize; /* program header entry size */
++ Elf32_Half e_phnum; /* number of program header entries */
++ Elf32_Half e_shentsize; /* section header entry size */
++ Elf32_Half e_shnum; /* number of section header entries */
++ Elf32_Half e_shstrndx; /* section header table's "section
++ header string table" entry offset */
++} Elf32_Ehdr;
++
++typedef struct {
++ unsigned char e_ident[EI_NIDENT]; /* Id bytes */
++ Elf64_Quarter e_type; /* file type */
++ Elf64_Quarter e_machine; /* machine type */
++ Elf64_Half e_version; /* version number */
++ Elf64_Addr e_entry; /* entry point */
++ Elf64_Off e_phoff; /* Program hdr offset */
++ Elf64_Off e_shoff; /* Section hdr offset */
++ Elf64_Half e_flags; /* Processor flags */
++ Elf64_Quarter e_ehsize; /* sizeof ehdr */
++ Elf64_Quarter e_phentsize; /* Program header entry size */
++ Elf64_Quarter e_phnum; /* Number of program headers */
++ Elf64_Quarter e_shentsize; /* Section header entry size */
++ Elf64_Quarter e_shnum; /* Number of section headers */
++ Elf64_Quarter e_shstrndx; /* String table index */
++} Elf64_Ehdr;
++
++/* e_type */
++#define ET_NONE 0 /* No file type */
++#define ET_REL 1 /* relocatable file */
++#define ET_EXEC 2 /* executable file */
++#define ET_DYN 3 /* shared object file */
++#define ET_CORE 4 /* core file */
++#define ET_NUM 5 /* number of types */
++#define ET_LOPROC 0xff00 /* reserved range for processor */
++#define ET_HIPROC 0xffff /* specific e_type */
++
++/* e_machine */
++#define EM_NONE 0 /* No Machine */
++#define EM_M32 1 /* AT&T WE 32100 */
++#define EM_SPARC 2 /* SPARC */
++#define EM_386 3 /* Intel 80386 */
++#define EM_68K 4 /* Motorola 68000 */
++#define EM_88K 5 /* Motorola 88000 */
++#define EM_486 6 /* Intel 80486 - unused? */
++#define EM_860 7 /* Intel 80860 */
++#define EM_MIPS 8 /* MIPS R3000 Big-Endian only */
++/*
++ * Don't know if EM_MIPS_RS4_BE,
++ * EM_SPARC64, EM_PARISC,
++ * or EM_PPC are ABI compliant
++ */
++#define EM_MIPS_RS4_BE 10 /* MIPS R4000 Big-Endian */
++#define EM_SPARC64 11 /* SPARC v9 64-bit unoffical */
++#define EM_PARISC 15 /* HPPA */
++#define EM_SPARC32PLUS 18 /* Enhanced instruction set SPARC */
++#define EM_PPC 20 /* PowerPC */
++#define EM_PPC64 21 /* PowerPC 64-bit */
++#define EM_ARM 40 /* Advanced RISC Machines ARM */
++#define EM_ALPHA 41 /* DEC ALPHA */
++#define EM_SPARCV9 43 /* SPARC version 9 */
++#define EM_ALPHA_EXP 0x9026 /* DEC ALPHA */
++#define EM_IA_64 50 /* Intel Merced */
++#define EM_X86_64 62 /* AMD x86-64 architecture */
++#define EM_VAX 75 /* DEC VAX */
++
++/* Version */
++#define EV_NONE 0 /* Invalid */
++#define EV_CURRENT 1 /* Current */
++#define EV_NUM 2 /* number of versions */
++
++/* Section Header */
++typedef struct {
++ Elf32_Word sh_name; /* name - index into section header
++ string table section */
++ Elf32_Word sh_type; /* type */
++ Elf32_Word sh_flags; /* flags */
++ Elf32_Addr sh_addr; /* address */
++ Elf32_Off sh_offset; /* file offset */
++ Elf32_Word sh_size; /* section size */
++ Elf32_Word sh_link; /* section header table index link */
++ Elf32_Word sh_info; /* extra information */
++ Elf32_Word sh_addralign; /* address alignment */
++ Elf32_Word sh_entsize; /* section entry size */
++} Elf32_Shdr;
++
++typedef struct {
++ Elf64_Half sh_name; /* section name */
++ Elf64_Half sh_type; /* section type */
++ Elf64_Xword sh_flags; /* section flags */
++ Elf64_Addr sh_addr; /* virtual address */
++ Elf64_Off sh_offset; /* file offset */
++ Elf64_Xword sh_size; /* section size */
++ Elf64_Half sh_link; /* link to another */
++ Elf64_Half sh_info; /* misc info */
++ Elf64_Xword sh_addralign; /* memory alignment */
++ Elf64_Xword sh_entsize; /* table entry size */
++} Elf64_Shdr;
++
++/* Special Section Indexes */
++#define SHN_UNDEF 0 /* undefined */
++#define SHN_LORESERVE 0xff00 /* lower bounds of reserved indexes */
++#define SHN_LOPROC 0xff00 /* reserved range for processor */
++#define SHN_HIPROC 0xff1f /* specific section indexes */
++#define SHN_ABS 0xfff1 /* absolute value */
++#define SHN_COMMON 0xfff2 /* common symbol */
++#define SHN_HIRESERVE 0xffff /* upper bounds of reserved indexes */
++
++/* sh_type */
++#define SHT_NULL 0 /* inactive */
++#define SHT_PROGBITS 1 /* program defined information */
++#define SHT_SYMTAB 2 /* symbol table section */
++#define SHT_STRTAB 3 /* string table section */
++#define SHT_RELA 4 /* relocation section with addends*/
++#define SHT_HASH 5 /* symbol hash table section */
++#define SHT_DYNAMIC 6 /* dynamic section */
++#define SHT_NOTE 7 /* note section */
++#define SHT_NOBITS 8 /* no space section */
++#define SHT_REL 9 /* relation section without addends */
++#define SHT_SHLIB 10 /* reserved - purpose unknown */
++#define SHT_DYNSYM 11 /* dynamic symbol table section */
++#define SHT_NUM 12 /* number of section types */
++#define SHT_LOPROC 0x70000000 /* reserved range for processor */
++#define SHT_HIPROC 0x7fffffff /* specific section header types */
++#define SHT_LOUSER 0x80000000 /* reserved range for application */
++#define SHT_HIUSER 0xffffffff /* specific indexes */
++
++/* Section names */
++#define ELF_BSS ".bss" /* uninitialized data */
++#define ELF_DATA ".data" /* initialized data */
++#define ELF_DEBUG ".debug" /* debug */
++#define ELF_DYNAMIC ".dynamic" /* dynamic linking information */
++#define ELF_DYNSTR ".dynstr" /* dynamic string table */
++#define ELF_DYNSYM ".dynsym" /* dynamic symbol table */
++#define ELF_FINI ".fini" /* termination code */
++#define ELF_GOT ".got" /* global offset table */
++#define ELF_HASH ".hash" /* symbol hash table */
++#define ELF_INIT ".init" /* initialization code */
++#define ELF_REL_DATA ".rel.data" /* relocation data */
++#define ELF_REL_FINI ".rel.fini" /* relocation termination code */
++#define ELF_REL_INIT ".rel.init" /* relocation initialization code */
++#define ELF_REL_DYN ".rel.dyn" /* relocaltion dynamic link info */
++#define ELF_REL_RODATA ".rel.rodata" /* relocation read-only data */
++#define ELF_REL_TEXT ".rel.text" /* relocation code */
++#define ELF_RODATA ".rodata" /* read-only data */
++#define ELF_SHSTRTAB ".shstrtab" /* section header string table */
++#define ELF_STRTAB ".strtab" /* string table */
++#define ELF_SYMTAB ".symtab" /* symbol table */
++#define ELF_TEXT ".text" /* code */
++
++
++/* Section Attribute Flags - sh_flags */
++#define SHF_WRITE 0x1 /* Writable */
++#define SHF_ALLOC 0x2 /* occupies memory */
++#define SHF_EXECINSTR 0x4 /* executable */
++#define SHF_MASKPROC 0xf0000000 /* reserved bits for processor */
++ /* specific section attributes */
++
++/* Symbol Table Entry */
++typedef struct elf32_sym {
++ Elf32_Word st_name; /* name - index into string table */
++ Elf32_Addr st_value; /* symbol value */
++ Elf32_Word st_size; /* symbol size */
++ unsigned char st_info; /* type and binding */
++ unsigned char st_other; /* 0 - no defined meaning */
++ Elf32_Half st_shndx; /* section header index */
++} Elf32_Sym;
++
++typedef struct {
++ Elf64_Half st_name; /* Symbol name index in str table */
++ Elf_Byte st_info; /* type / binding attrs */
++ Elf_Byte st_other; /* unused */
++ Elf64_Quarter st_shndx; /* section index of symbol */
++ Elf64_Xword st_value; /* value of symbol */
++ Elf64_Xword st_size; /* size of symbol */
++} Elf64_Sym;
++
++/* Symbol table index */
++#define STN_UNDEF 0 /* undefined */
++
++/* Extract symbol info - st_info */
++#define ELF32_ST_BIND(x) ((x) >> 4)
++#define ELF32_ST_TYPE(x) (((unsigned int) x) & 0xf)
++#define ELF32_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
++
++#define ELF64_ST_BIND(x) ((x) >> 4)
++#define ELF64_ST_TYPE(x) (((unsigned int) x) & 0xf)
++#define ELF64_ST_INFO(b,t) (((b) << 4) + ((t) & 0xf))
++
++/* Symbol Binding - ELF32_ST_BIND - st_info */
++#define STB_LOCAL 0 /* Local symbol */
++#define STB_GLOBAL 1 /* Global symbol */
++#define STB_WEAK 2 /* like global - lower precedence */
++#define STB_NUM 3 /* number of symbol bindings */
++#define STB_LOPROC 13 /* reserved range for processor */
++#define STB_HIPROC 15 /* specific symbol bindings */
++
++/* Symbol type - ELF32_ST_TYPE - st_info */
++#define STT_NOTYPE 0 /* not specified */
++#define STT_OBJECT 1 /* data object */
++#define STT_FUNC 2 /* function */
++#define STT_SECTION 3 /* section */
++#define STT_FILE 4 /* file */
++#define STT_NUM 5 /* number of symbol types */
++#define STT_LOPROC 13 /* reserved range for processor */
++#define STT_HIPROC 15 /* specific symbol types */
++
++/* Relocation entry with implicit addend */
++typedef struct {
++ Elf32_Addr r_offset; /* offset of relocation */
++ Elf32_Word r_info; /* symbol table index and type */
++} Elf32_Rel;
++
++/* Relocation entry with explicit addend */
++typedef struct {
++ Elf32_Addr r_offset; /* offset of relocation */
++ Elf32_Word r_info; /* symbol table index and type */
++ Elf32_Sword r_addend;
++} Elf32_Rela;
++
++/* Extract relocation info - r_info */
++#define ELF32_R_SYM(i) ((i) >> 8)
++#define ELF32_R_TYPE(i) ((unsigned char) (i))
++#define ELF32_R_INFO(s,t) (((s) << 8) + (unsigned char)(t))
++
++typedef struct {
++ Elf64_Xword r_offset; /* where to do it */
++ Elf64_Xword r_info; /* index & type of relocation */
++} Elf64_Rel;
++
++typedef struct {
++ Elf64_Xword r_offset; /* where to do it */
++ Elf64_Xword r_info; /* index & type of relocation */
++ Elf64_Sxword r_addend; /* adjustment value */
++} Elf64_Rela;
++
++#define ELF64_R_SYM(info) ((info) >> 32)
++#define ELF64_R_TYPE(info) ((info) & 0xFFFFFFFF)
++#define ELF64_R_INFO(s,t) (((s) << 32) + (u_int32_t)(t))
++
++/* Program Header */
++typedef struct {
++ Elf32_Word p_type; /* segment type */
++ Elf32_Off p_offset; /* segment offset */
++ Elf32_Addr p_vaddr; /* virtual address of segment */
++ Elf32_Addr p_paddr; /* physical address - ignored? */
++ Elf32_Word p_filesz; /* number of bytes in file for seg. */
++ Elf32_Word p_memsz; /* number of bytes in mem. for seg. */
++ Elf32_Word p_flags; /* flags */
++ Elf32_Word p_align; /* memory alignment */
++} Elf32_Phdr;
++
++typedef struct {
++ Elf64_Half p_type; /* entry type */
++ Elf64_Half p_flags; /* flags */
++ Elf64_Off p_offset; /* offset */
++ Elf64_Addr p_vaddr; /* virtual address */
++ Elf64_Addr p_paddr; /* physical address */
++ Elf64_Xword p_filesz; /* file size */
++ Elf64_Xword p_memsz; /* memory size */
++ Elf64_Xword p_align; /* memory & file alignment */
++} Elf64_Phdr;
++
++/* Segment types - p_type */
++#define PT_NULL 0 /* unused */
++#define PT_LOAD 1 /* loadable segment */
++#define PT_DYNAMIC 2 /* dynamic linking section */
++#define PT_INTERP 3 /* the RTLD */
++#define PT_NOTE 4 /* auxiliary information */
++#define PT_SHLIB 5 /* reserved - purpose undefined */
++#define PT_PHDR 6 /* program header */
++#define PT_NUM 7 /* Number of segment types */
++#define PT_LOPROC 0x70000000 /* reserved range for processor */
++#define PT_HIPROC 0x7fffffff /* specific segment types */
++
++/* Segment flags - p_flags */
++#define PF_X 0x1 /* Executable */
++#define PF_W 0x2 /* Writable */
++#define PF_R 0x4 /* Readable */
++#define PF_MASKPROC 0xf0000000 /* reserved bits for processor */
++ /* specific segment flags */
++
++/* Dynamic structure */
++typedef struct {
++ Elf32_Sword d_tag; /* controls meaning of d_val */
++ union {
++ Elf32_Word d_val; /* Multiple meanings - see d_tag */
++ Elf32_Addr d_ptr; /* program virtual address */
++ } d_un;
++} Elf32_Dyn;
++
++typedef struct {
++ Elf64_Xword d_tag; /* controls meaning of d_val */
++ union {
++ Elf64_Addr d_ptr;
++ Elf64_Xword d_val;
++ } d_un;
++} Elf64_Dyn;
++
++/* Dynamic Array Tags - d_tag */
++#define DT_NULL 0 /* marks end of _DYNAMIC array */
++#define DT_NEEDED 1 /* string table offset of needed lib */
++#define DT_PLTRELSZ 2 /* size of relocation entries in PLT */
++#define DT_PLTGOT 3 /* address PLT/GOT */
++#define DT_HASH 4 /* address of symbol hash table */
++#define DT_STRTAB 5 /* address of string table */
++#define DT_SYMTAB 6 /* address of symbol table */
++#define DT_RELA 7 /* address of relocation table */
++#define DT_RELASZ 8 /* size of relocation table */
++#define DT_RELAENT 9 /* size of relocation entry */
++#define DT_STRSZ 10 /* size of string table */
++#define DT_SYMENT 11 /* size of symbol table entry */
++#define DT_INIT 12 /* address of initialization func. */
++#define DT_FINI 13 /* address of termination function */
++#define DT_SONAME 14 /* string table offset of shared obj */
++#define DT_RPATH 15 /* string table offset of library
++ search path */
++#define DT_SYMBOLIC 16 /* start sym search in shared obj. */
++#define DT_REL 17 /* address of rel. tbl. w addends */
++#define DT_RELSZ 18 /* size of DT_REL relocation table */
++#define DT_RELENT 19 /* size of DT_REL relocation entry */
++#define DT_PLTREL 20 /* PLT referenced relocation entry */
++#define DT_DEBUG 21 /* bugger */
++#define DT_TEXTREL 22 /* Allow rel. mod. to unwritable seg */
++#define DT_JMPREL 23 /* add. of PLT's relocation entries */
++#define DT_BIND_NOW 24 /* Bind now regardless of env setting */
++#define DT_NUM 25 /* Number used. */
++#define DT_LOPROC 0x70000000 /* reserved range for processor */
++#define DT_HIPROC 0x7fffffff /* specific dynamic array tags */
++
++/* Standard ELF hashing function */
++unsigned int elf_hash(const unsigned char *name);
++
++/*
++ * Note Definitions
++ */
++typedef struct {
++ Elf32_Word namesz;
++ Elf32_Word descsz;
++ Elf32_Word type;
++} Elf32_Note;
++
++typedef struct {
++ Elf64_Half namesz;
++ Elf64_Half descsz;
++ Elf64_Half type;
++} Elf64_Note;
++
++
++#if defined(ELFSIZE)
++#define CONCAT(x,y) __CONCAT(x,y)
++#define ELFNAME(x) CONCAT(elf,CONCAT(ELFSIZE,CONCAT(_,x)))
++#define ELFNAME2(x,y) CONCAT(x,CONCAT(_elf,CONCAT(ELFSIZE,CONCAT(_,y))))
++#define ELFNAMEEND(x) CONCAT(x,CONCAT(_elf,ELFSIZE))
++#define ELFDEFNNAME(x) CONCAT(ELF,CONCAT(ELFSIZE,CONCAT(_,x)))
++#endif
++
++#if defined(ELFSIZE) && (ELFSIZE == 32)
++#define Elf_Ehdr Elf32_Ehdr
++#define Elf_Phdr Elf32_Phdr
++#define Elf_Shdr Elf32_Shdr
++#define Elf_Sym Elf32_Sym
++#define Elf_Rel Elf32_Rel
++#define Elf_RelA Elf32_Rela
++#define Elf_Dyn Elf32_Dyn
++#define Elf_Word Elf32_Word
++#define Elf_Sword Elf32_Sword
++#define Elf_Addr Elf32_Addr
++#define Elf_Off Elf32_Off
++#define Elf_Nhdr Elf32_Nhdr
++#define Elf_Note Elf32_Note
++
++#define ELF_R_SYM ELF32_R_SYM
++#define ELF_R_TYPE ELF32_R_TYPE
++#define ELF_R_INFO ELF32_R_INFO
++#define ELFCLASS ELFCLASS32
++
++#define ELF_ST_BIND ELF32_ST_BIND
++#define ELF_ST_TYPE ELF32_ST_TYPE
++#define ELF_ST_INFO ELF32_ST_INFO
++
++#define AuxInfo Aux32Info
++#elif defined(ELFSIZE) && (ELFSIZE == 64)
++#define Elf_Ehdr Elf64_Ehdr
++#define Elf_Phdr Elf64_Phdr
++#define Elf_Shdr Elf64_Shdr
++#define Elf_Sym Elf64_Sym
++#define Elf_Rel Elf64_Rel
++#define Elf_RelA Elf64_Rela
++#define Elf_Dyn Elf64_Dyn
++#define Elf_Word Elf64_Word
++#define Elf_Sword Elf64_Sword
++#define Elf_Addr Elf64_Addr
++#define Elf_Off Elf64_Off
++#define Elf_Nhdr Elf64_Nhdr
++#define Elf_Note Elf64_Note
++
++#define ELF_R_SYM ELF64_R_SYM
++#define ELF_R_TYPE ELF64_R_TYPE
++#define ELF_R_INFO ELF64_R_INFO
++#define ELFCLASS ELFCLASS64
++
++#define ELF_ST_BIND ELF64_ST_BIND
++#define ELF_ST_TYPE ELF64_ST_TYPE
++#define ELF_ST_INFO ELF64_ST_INFO
++
++#define AuxInfo Aux64Info
++#endif
++
++#endif /* __XEN_PUBLIC_ELFSTRUCTS_H__ */
+diff --git a/include/xen/interface/event_channel.h b/include/xen/interface/event_channel.h
+index 919b5bd..0c1356b 100644
+--- a/include/xen/interface/event_channel.h
++++ b/include/xen/interface/event_channel.h
+@@ -3,14 +3,39 @@
+ *
+ * Event channels between domains.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2003-2004, K A Fraser.
+ */
+
+ #ifndef __XEN_PUBLIC_EVENT_CHANNEL_H__
+ #define __XEN_PUBLIC_EVENT_CHANNEL_H__
+
++/*
++ * Prototype for this hypercall is:
++ * int event_channel_op(int cmd, void *args)
++ * @cmd == EVTCHNOP_??? (event-channel operation).
++ * @args == Operation-specific extra arguments (NULL if none).
++ */
++
+ typedef uint32_t evtchn_port_t;
+-DEFINE_GUEST_HANDLE(evtchn_port_t);
++DEFINE_XEN_GUEST_HANDLE(evtchn_port_t);
+
+ /*
+ * EVTCHNOP_alloc_unbound: Allocate a port in domain <dom> and mark as
+@@ -20,13 +45,14 @@ DEFINE_GUEST_HANDLE(evtchn_port_t);
+ * 1. If the caller is unprivileged then <dom> must be DOMID_SELF.
+ * 2. <rdom> may be DOMID_SELF, allowing loopback connections.
+ */
+-#define EVTCHNOP_alloc_unbound 6
++#define EVTCHNOP_alloc_unbound 6
+ struct evtchn_alloc_unbound {
+- /* IN parameters */
+- domid_t dom, remote_dom;
+- /* OUT parameters */
+- evtchn_port_t port;
++ /* IN parameters */
++ domid_t dom, remote_dom;
++ /* OUT parameters */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_alloc_unbound evtchn_alloc_unbound_t;
+
+ /*
+ * EVTCHNOP_bind_interdomain: Construct an interdomain event channel between
+@@ -39,29 +65,35 @@ struct evtchn_alloc_unbound {
+ */
+ #define EVTCHNOP_bind_interdomain 0
+ struct evtchn_bind_interdomain {
+- /* IN parameters. */
+- domid_t remote_dom;
+- evtchn_port_t remote_port;
+- /* OUT parameters. */
+- evtchn_port_t local_port;
++ /* IN parameters. */
++ domid_t remote_dom;
++ evtchn_port_t remote_port;
++ /* OUT parameters. */
++ evtchn_port_t local_port;
+ };
++typedef struct evtchn_bind_interdomain evtchn_bind_interdomain_t;
+
+ /*
+ * EVTCHNOP_bind_virq: Bind a local event channel to VIRQ <irq> on specified
+ * vcpu.
+ * NOTES:
+- * 1. A virtual IRQ may be bound to at most one event channel per vcpu.
+- * 2. The allocated event channel is bound to the specified vcpu. The binding
+- * may not be changed.
++ * 1. Virtual IRQs are classified as per-vcpu or global. See the VIRQ list
++ * in xen.h for the classification of each VIRQ.
++ * 2. Global VIRQs must be allocated on VCPU0 but can subsequently be
++ * re-bound via EVTCHNOP_bind_vcpu.
++ * 3. Per-vcpu VIRQs may be bound to at most one event channel per vcpu.
++ * The allocated event channel is bound to the specified vcpu and the
++ * binding cannot be changed.
+ */
+-#define EVTCHNOP_bind_virq 1
++#define EVTCHNOP_bind_virq 1
+ struct evtchn_bind_virq {
+- /* IN parameters. */
+- uint32_t virq;
+- uint32_t vcpu;
+- /* OUT parameters. */
+- evtchn_port_t port;
++ /* IN parameters. */
++ uint32_t virq;
++ uint32_t vcpu;
++ /* OUT parameters. */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_bind_virq evtchn_bind_virq_t;
+
+ /*
+ * EVTCHNOP_bind_pirq: Bind a local event channel to PIRQ <irq>.
+@@ -69,15 +101,16 @@ struct evtchn_bind_virq {
+ * 1. A physical IRQ may be bound to at most one event channel per domain.
+ * 2. Only a sufficiently-privileged domain may bind to a physical IRQ.
+ */
+-#define EVTCHNOP_bind_pirq 2
++#define EVTCHNOP_bind_pirq 2
+ struct evtchn_bind_pirq {
+- /* IN parameters. */
+- uint32_t pirq;
++ /* IN parameters. */
++ uint32_t pirq;
+ #define BIND_PIRQ__WILL_SHARE 1
+- uint32_t flags; /* BIND_PIRQ__* */
+- /* OUT parameters. */
+- evtchn_port_t port;
++ uint32_t flags; /* BIND_PIRQ__* */
++ /* OUT parameters. */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_bind_pirq evtchn_bind_pirq_t;
+
+ /*
+ * EVTCHNOP_bind_ipi: Bind a local event channel to receive events.
+@@ -85,33 +118,36 @@ struct evtchn_bind_pirq {
+ * 1. The allocated event channel is bound to the specified vcpu. The binding
+ * may not be changed.
+ */
+-#define EVTCHNOP_bind_ipi 7
++#define EVTCHNOP_bind_ipi 7
+ struct evtchn_bind_ipi {
+- uint32_t vcpu;
+- /* OUT parameters. */
+- evtchn_port_t port;
++ uint32_t vcpu;
++ /* OUT parameters. */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_bind_ipi evtchn_bind_ipi_t;
+
+ /*
+ * EVTCHNOP_close: Close a local event channel <port>. If the channel is
+ * interdomain then the remote end is placed in the unbound state
+ * (EVTCHNSTAT_unbound), awaiting a new connection.
+ */
+-#define EVTCHNOP_close 3
++#define EVTCHNOP_close 3
+ struct evtchn_close {
+- /* IN parameters. */
+- evtchn_port_t port;
++ /* IN parameters. */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_close evtchn_close_t;
+
+ /*
+ * EVTCHNOP_send: Send an event to the remote end of the channel whose local
+ * endpoint is <port>.
+ */
+-#define EVTCHNOP_send 4
++#define EVTCHNOP_send 4
+ struct evtchn_send {
+- /* IN parameters. */
+- evtchn_port_t port;
++ /* IN parameters. */
++ evtchn_port_t port;
+ };
++typedef struct evtchn_send evtchn_send_t;
+
+ /*
+ * EVTCHNOP_status: Get the current status of the communication channel which
+@@ -121,75 +157,109 @@ struct evtchn_send {
+ * 2. Only a sufficiently-privileged domain may obtain the status of an event
+ * channel for which <dom> is not DOMID_SELF.
+ */
+-#define EVTCHNOP_status 5
++#define EVTCHNOP_status 5
+ struct evtchn_status {
+- /* IN parameters */
+- domid_t dom;
+- evtchn_port_t port;
+- /* OUT parameters */
+-#define EVTCHNSTAT_closed 0 /* Channel is not in use. */
+-#define EVTCHNSTAT_unbound 1 /* Channel is waiting interdom connection.*/
+-#define EVTCHNSTAT_interdomain 2 /* Channel is connected to remote domain. */
+-#define EVTCHNSTAT_pirq 3 /* Channel is bound to a phys IRQ line. */
+-#define EVTCHNSTAT_virq 4 /* Channel is bound to a virtual IRQ line */
+-#define EVTCHNSTAT_ipi 5 /* Channel is bound to a virtual IPI line */
+- uint32_t status;
+- uint32_t vcpu; /* VCPU to which this channel is bound. */
+- union {
+- struct {
+- domid_t dom;
+- } unbound; /* EVTCHNSTAT_unbound */
+- struct {
+- domid_t dom;
+- evtchn_port_t port;
+- } interdomain; /* EVTCHNSTAT_interdomain */
+- uint32_t pirq; /* EVTCHNSTAT_pirq */
+- uint32_t virq; /* EVTCHNSTAT_virq */
+- } u;
++ /* IN parameters */
++ domid_t dom;
++ evtchn_port_t port;
++ /* OUT parameters */
++#define EVTCHNSTAT_closed 0 /* Channel is not in use. */
++#define EVTCHNSTAT_unbound 1 /* Channel is waiting interdom connection.*/
++#define EVTCHNSTAT_interdomain 2 /* Channel is connected to remote domain. */
++#define EVTCHNSTAT_pirq 3 /* Channel is bound to a phys IRQ line. */
++#define EVTCHNSTAT_virq 4 /* Channel is bound to a virtual IRQ line */
++#define EVTCHNSTAT_ipi 5 /* Channel is bound to a virtual IPI line */
++ uint32_t status;
++ uint32_t vcpu; /* VCPU to which this channel is bound. */
++ union {
++ struct {
++ domid_t dom;
++ } unbound; /* EVTCHNSTAT_unbound */
++ struct {
++ domid_t dom;
++ evtchn_port_t port;
++ } interdomain; /* EVTCHNSTAT_interdomain */
++ uint32_t pirq; /* EVTCHNSTAT_pirq */
++ uint32_t virq; /* EVTCHNSTAT_virq */
++ } u;
+ };
++typedef struct evtchn_status evtchn_status_t;
+
+ /*
+ * EVTCHNOP_bind_vcpu: Specify which vcpu a channel should notify when an
+ * event is pending.
+ * NOTES:
+- * 1. IPI- and VIRQ-bound channels always notify the vcpu that initialised
+- * the binding. This binding cannot be changed.
+- * 2. All other channels notify vcpu0 by default. This default is set when
++ * 1. IPI-bound channels always notify the vcpu specified at bind time.
++ * This binding cannot be changed.
++ * 2. Per-VCPU VIRQ channels always notify the vcpu specified at bind time.
++ * This binding cannot be changed.
++ * 3. All other channels notify vcpu0 by default. This default is set when
+ * the channel is allocated (a port that is freed and subsequently reused
+ * has its binding reset to vcpu0).
+ */
+-#define EVTCHNOP_bind_vcpu 8
++#define EVTCHNOP_bind_vcpu 8
+ struct evtchn_bind_vcpu {
+- /* IN parameters. */
+- evtchn_port_t port;
+- uint32_t vcpu;
++ /* IN parameters. */
++ evtchn_port_t port;
++ uint32_t vcpu;
+ };
++typedef struct evtchn_bind_vcpu evtchn_bind_vcpu_t;
+
+ /*
+ * EVTCHNOP_unmask: Unmask the specified local event-channel port and deliver
+ * a notification to the appropriate VCPU if an event is pending.
+ */
+-#define EVTCHNOP_unmask 9
++#define EVTCHNOP_unmask 9
+ struct evtchn_unmask {
+- /* IN parameters. */
+- evtchn_port_t port;
++ /* IN parameters. */
++ evtchn_port_t port;
++};
++typedef struct evtchn_unmask evtchn_unmask_t;
++
++/*
++ * EVTCHNOP_reset: Close all event channels associated with specified domain.
++ * NOTES:
++ * 1. <dom> may be specified as DOMID_SELF.
++ * 2. Only a sufficiently-privileged domain may specify other than DOMID_SELF.
++ */
++#define EVTCHNOP_reset 10
++struct evtchn_reset {
++ /* IN parameters. */
++ domid_t dom;
+ };
++typedef struct evtchn_reset evtchn_reset_t;
+
++/*
++ * Argument to event_channel_op_compat() hypercall. Superceded by new
++ * event_channel_op() hypercall since 0x00030202.
++ */
+ struct evtchn_op {
+- uint32_t cmd; /* EVTCHNOP_* */
+- union {
+- struct evtchn_alloc_unbound alloc_unbound;
+- struct evtchn_bind_interdomain bind_interdomain;
+- struct evtchn_bind_virq bind_virq;
+- struct evtchn_bind_pirq bind_pirq;
+- struct evtchn_bind_ipi bind_ipi;
+- struct evtchn_close close;
+- struct evtchn_send send;
+- struct evtchn_status status;
+- struct evtchn_bind_vcpu bind_vcpu;
+- struct evtchn_unmask unmask;
+- } u;
++ uint32_t cmd; /* EVTCHNOP_* */
++ union {
++ struct evtchn_alloc_unbound alloc_unbound;
++ struct evtchn_bind_interdomain bind_interdomain;
++ struct evtchn_bind_virq bind_virq;
++ struct evtchn_bind_pirq bind_pirq;
++ struct evtchn_bind_ipi bind_ipi;
++ struct evtchn_close close;
++ struct evtchn_send send;
++ struct evtchn_status status;
++ struct evtchn_bind_vcpu bind_vcpu;
++ struct evtchn_unmask unmask;
++ } u;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(evtchn_op);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(evtchn_op);
++typedef struct evtchn_op evtchn_op_t;
++DEFINE_XEN_GUEST_HANDLE(evtchn_op_t);
+
+ #endif /* __XEN_PUBLIC_EVENT_CHANNEL_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/features.h b/include/xen/interface/features.h
+index d73228d..65b6611 100644
+--- a/include/xen/interface/features.h
++++ b/include/xen/interface/features.h
+@@ -3,6 +3,24 @@
+ *
+ * Feature flags, reported by XENVER_get_features.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2006, Keir Fraser <keir at xensource.com>
+ */
+
+@@ -38,6 +56,19 @@
+ */
+ #define XENFEAT_pae_pgdir_above_4gb 4
+
++/* x86: Does this Xen host support the MMU_PT_UPDATE_PRESERVE_AD hypercall? */
++#define XENFEAT_mmu_pt_update_preserve_ad 5
++
+ #define XENFEAT_NR_SUBMAPS 1
+
+ #endif /* __XEN_PUBLIC_FEATURES_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/grant_table.h b/include/xen/interface/grant_table.h
+index 39da93c..d032148 100644
+--- a/include/xen/interface/grant_table.h
++++ b/include/xen/interface/grant_table.h
+@@ -100,6 +100,7 @@ struct grant_entry {
+ */
+ uint32_t frame;
+ };
++typedef struct grant_entry grant_entry_t;
+
+ /*
+ * Type of grant entry.
+@@ -118,6 +119,7 @@ struct grant_entry {
+ * GTF_readonly: Restrict @domid to read-only mappings and accesses. [GST]
+ * GTF_reading: Grant entry is currently mapped for reading by @domid. [XEN]
+ * GTF_writing: Grant entry is currently mapped for writing by @domid. [XEN]
++ * GTF_PAT, GTF_PWT, GTF_PCD: (x86) cache attribute flags for the grant [GST]
+ */
+ #define _GTF_readonly (2)
+ #define GTF_readonly (1U<<_GTF_readonly)
+@@ -125,6 +127,12 @@ struct grant_entry {
+ #define GTF_reading (1U<<_GTF_reading)
+ #define _GTF_writing (4)
+ #define GTF_writing (1U<<_GTF_writing)
++#define _GTF_PWT (5)
++#define GTF_PWT (1U<<_GTF_PWT)
++#define _GTF_PCD (6)
++#define GTF_PCD (1U<<_GTF_PCD)
++#define _GTF_PAT (7)
++#define GTF_PAT (1U<<_GTF_PAT)
+
+ /*
+ * Subflags for GTF_accept_transfer:
+@@ -185,7 +193,9 @@ struct gnttab_map_grant_ref {
+ grant_handle_t handle;
+ uint64_t dev_bus_addr;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_map_grant_ref);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_map_grant_ref);
++typedef struct gnttab_map_grant_ref gnttab_map_grant_ref_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_map_grant_ref_t);
+
+ /*
+ * GNTTABOP_unmap_grant_ref: Destroy one or more grant-reference mappings
+@@ -207,7 +217,9 @@ struct gnttab_unmap_grant_ref {
+ /* OUT parameters. */
+ int16_t status; /* GNTST_* */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_unmap_grant_ref);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_unmap_grant_ref);
++typedef struct gnttab_unmap_grant_ref gnttab_unmap_grant_ref_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_grant_ref_t);
+
+ /*
+ * GNTTABOP_setup_table: Set up a grant table for <dom> comprising at least
+@@ -225,9 +237,11 @@ struct gnttab_setup_table {
+ uint32_t nr_frames;
+ /* OUT parameters. */
+ int16_t status; /* GNTST_* */
+- GUEST_HANDLE(ulong) frame_list;
++ XEN_GUEST_HANDLE(ulong) frame_list;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_setup_table);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_setup_table);
++typedef struct gnttab_setup_table gnttab_setup_table_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_setup_table_t);
+
+ /*
+ * GNTTABOP_dump_table: Dump the contents of the grant table to the
+@@ -240,7 +254,9 @@ struct gnttab_dump_table {
+ /* OUT parameters. */
+ int16_t status; /* GNTST_* */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_dump_table);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_dump_table);
++typedef struct gnttab_dump_table gnttab_dump_table_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_dump_table_t);
+
+ /*
+ * GNTTABOP_transfer_grant_ref: Transfer <frame> to a foreign domain. The
+@@ -253,13 +269,16 @@ DEFINE_GUEST_HANDLE_STRUCT(gnttab_dump_table);
+ #define GNTTABOP_transfer 4
+ struct gnttab_transfer {
+ /* IN parameters. */
+- unsigned long mfn;
++ xen_pfn_t mfn;
+ domid_t domid;
+ grant_ref_t ref;
+ /* OUT parameters. */
+ int16_t status;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_transfer);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_transfer);
++typedef struct gnttab_transfer gnttab_transfer_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_transfer_t);
++
+
+ /*
+ * GNTTABOP_copy: Hypervisor based copy
+@@ -285,22 +304,23 @@ DEFINE_GUEST_HANDLE_STRUCT(gnttab_transfer);
+ #define GNTCOPY_dest_gref (1<<_GNTCOPY_dest_gref)
+
+ #define GNTTABOP_copy 5
+-struct gnttab_copy {
+- /* IN parameters. */
+- struct {
+- union {
+- grant_ref_t ref;
+- unsigned long gmfn;
+- } u;
+- domid_t domid;
+- uint16_t offset;
+- } source, dest;
+- uint16_t len;
+- uint16_t flags; /* GNTCOPY_* */
+- /* OUT parameters. */
+- int16_t status;
+-};
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_copy);
++typedef struct gnttab_copy {
++ /* IN parameters. */
++ struct {
++ union {
++ grant_ref_t ref;
++ xen_pfn_t gmfn;
++ } u;
++ domid_t domid;
++ uint16_t offset;
++ } source, dest;
++ uint16_t len;
++ uint16_t flags; /* GNTCOPY_* */
++ /* OUT parameters. */
++ int16_t status;
++} gnttab_copy_t;
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_copy);
++DEFINE_XEN_GUEST_HANDLE(gnttab_copy_t);
+
+ /*
+ * GNTTABOP_query_size: Query the current and maximum sizes of the shared
+@@ -318,7 +338,33 @@ struct gnttab_query_size {
+ uint32_t max_nr_frames;
+ int16_t status; /* GNTST_* */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(gnttab_query_size);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(gnttab_query_size);
++typedef struct gnttab_query_size gnttab_query_size_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_query_size_t);
++
++/*
++ * GNTTABOP_unmap_and_replace: Destroy one or more grant-reference mappings
++ * tracked by <handle> but atomically replace the page table entry with one
++ * pointing to the machine address under <new_addr>. <new_addr> will be
++ * redirected to the null entry.
++ * NOTES:
++ * 1. The call may fail in an undefined manner if either mapping is not
++ * tracked by <handle>.
++ * 2. After executing a batch of unmaps, it is guaranteed that no stale
++ * mappings will remain in the device or host TLBs.
++ */
++#define GNTTABOP_unmap_and_replace 7
++struct gnttab_unmap_and_replace {
++ /* IN parameters. */
++ uint64_t host_addr;
++ uint64_t new_addr;
++ grant_handle_t handle;
++ /* OUT parameters. */
++ int16_t status; /* GNTST_* */
++};
++typedef struct gnttab_unmap_and_replace gnttab_unmap_and_replace_t;
++DEFINE_XEN_GUEST_HANDLE(gnttab_unmap_and_replace_t);
++
+
+ /*
+ * Bitfield values for update_pin_status.flags.
+@@ -361,7 +407,8 @@ DEFINE_GUEST_HANDLE_STRUCT(gnttab_query_size);
+ #define GNTST_no_device_space (-7) /* Out of space in I/O MMU. */
+ #define GNTST_permission_denied (-8) /* Not enough privilege for operation. */
+ #define GNTST_bad_page (-9) /* Specified page was invalid for op. */
+-#define GNTST_bad_copy_arg (-10) /* copy arguments cross page boundary */
++#define GNTST_bad_copy_arg (-10) /* copy arguments cross page boundary. */
++#define GNTST_address_too_big (-11) /* transfer page address too large. */
+
+ #define GNTTABOP_error_msgs { \
+ "okay", \
+@@ -374,7 +421,18 @@ DEFINE_GUEST_HANDLE_STRUCT(gnttab_query_size);
+ "no spare translation slot in the I/O MMU", \
+ "permission denied", \
+ "bad page", \
+- "copy arguments cross page boundary" \
++ "copy arguments cross page boundary", \
++ "page address size too large" \
+ }
+
+ #endif /* __XEN_PUBLIC_GRANT_TABLE_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/hvm/e820.h b/include/xen/interface/hvm/e820.h
+new file mode 100644
+index 0000000..5bdc227
+--- /dev/null
++++ b/include/xen/interface/hvm/e820.h
+@@ -0,0 +1,34 @@
++
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_E820_H__
++#define __XEN_PUBLIC_HVM_E820_H__
++
++/* E820 location in HVM virtual address space. */
++#define HVM_E820_PAGE 0x00090000
++#define HVM_E820_NR_OFFSET 0x000001E8
++#define HVM_E820_OFFSET 0x000002D0
++
++#define HVM_BELOW_4G_RAM_END 0xF0000000
++#define HVM_BELOW_4G_MMIO_START HVM_BELOW_4G_RAM_END
++#define HVM_BELOW_4G_MMIO_LENGTH ((1ULL << 32) - HVM_BELOW_4G_MMIO_START)
++
++#endif /* __XEN_PUBLIC_HVM_E820_H__ */
+diff --git a/include/xen/interface/hvm/hvm_info_table.h b/include/xen/interface/hvm/hvm_info_table.h
+new file mode 100644
+index 0000000..dfe34db
+--- /dev/null
++++ b/include/xen/interface/hvm/hvm_info_table.h
+@@ -0,0 +1,41 @@
++/******************************************************************************
++ * hvm/hvm_info_table.h
++ *
++ * HVM parameter and information table, written into guest memory map.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
++#define __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__
++
++#define HVM_INFO_PFN 0x09F
++#define HVM_INFO_OFFSET 0x800
++#define HVM_INFO_PADDR ((HVM_INFO_PFN << 12) + HVM_INFO_OFFSET)
++
++struct hvm_info_table {
++ char signature[8]; /* "HVM INFO" */
++ uint32_t length;
++ uint8_t checksum;
++ uint8_t acpi_enabled;
++ uint8_t apic_mode;
++ uint32_t nr_vcpus;
++};
++
++#endif /* __XEN_PUBLIC_HVM_HVM_INFO_TABLE_H__ */
+diff --git a/include/xen/interface/hvm/hvm_op.h b/include/xen/interface/hvm/hvm_op.h
+new file mode 100644
+index 0000000..19f0cb7
+--- /dev/null
++++ b/include/xen/interface/hvm/hvm_op.h
+@@ -0,0 +1,97 @@
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_HVM_OP_H__
++#define __XEN_PUBLIC_HVM_HVM_OP_H__
++
++/* Get/set subcommands: extra argument == pointer to xen_hvm_param struct. */
++#define HVMOP_set_param 0
++#define HVMOP_get_param 1
++struct xen_hvm_param {
++ domid_t domid; /* IN */
++ uint32_t index; /* IN */
++ uint64_t value; /* IN/OUT */
++};
++typedef struct xen_hvm_param xen_hvm_param_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_param_t);
++
++/* Set the logical level of one of a domain's PCI INTx wires. */
++#define HVMOP_set_pci_intx_level 2
++struct xen_hvm_set_pci_intx_level {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* PCI INTx identification in PCI topology (domain:bus:device:intx). */
++ uint8_t domain, bus, device, intx;
++ /* Assertion level (0 = unasserted, 1 = asserted). */
++ uint8_t level;
++};
++typedef struct xen_hvm_set_pci_intx_level xen_hvm_set_pci_intx_level_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_intx_level_t);
++
++/* Set the logical level of one of a domain's ISA IRQ wires. */
++#define HVMOP_set_isa_irq_level 3
++struct xen_hvm_set_isa_irq_level {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* ISA device identification, by ISA IRQ (0-15). */
++ uint8_t isa_irq;
++ /* Assertion level (0 = unasserted, 1 = asserted). */
++ uint8_t level;
++};
++typedef struct xen_hvm_set_isa_irq_level xen_hvm_set_isa_irq_level_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_isa_irq_level_t);
++
++#define HVMOP_set_pci_link_route 4
++struct xen_hvm_set_pci_link_route {
++ /* Domain to be updated. */
++ domid_t domid;
++ /* PCI link identifier (0-3). */
++ uint8_t link;
++ /* ISA IRQ (1-15), or 0 (disable link). */
++ uint8_t isa_irq;
++};
++typedef struct xen_hvm_set_pci_link_route xen_hvm_set_pci_link_route_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_set_pci_link_route_t);
++
++/* Flushes all VCPU TLBs: @arg must be NULL. */
++#define HVMOP_flush_tlbs 5
++
++/* Following tools-only interfaces may change in future. */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++
++/* Track dirty VRAM. */
++#define HVMOP_track_dirty_vram 6
++struct xen_hvm_track_dirty_vram {
++ /* Domain to be tracked. */
++ domid_t domid;
++ /* First pfn to track. */
++ uint64_aligned_t first_pfn;
++ /* Number of pages to track. */
++ uint64_aligned_t nr;
++ /* OUT variable. */
++ /* Dirty bitmap buffer. */
++ XEN_GUEST_HANDLE_64(uint8) dirty_bitmap;
++};
++typedef struct xen_hvm_track_dirty_vram xen_hvm_track_dirty_vram_t;
++DEFINE_XEN_GUEST_HANDLE(xen_hvm_track_dirty_vram_t);
++
++#endif /* defined(__XEN__) || defined(__XEN_TOOLS__) */
++
++#endif /* __XEN_PUBLIC_HVM_HVM_OP_H__ */
+diff --git a/include/xen/interface/hvm/ioreq.h b/include/xen/interface/hvm/ioreq.h
+new file mode 100644
+index 0000000..5b68730
+--- /dev/null
++++ b/include/xen/interface/hvm/ioreq.h
+@@ -0,0 +1,127 @@
++/*
++ * ioreq.h: I/O request definitions for device models
++ * Copyright (c) 2004, Intel Corporation.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef _IOREQ_H_
++#define _IOREQ_H_
++
++#define IOREQ_READ 1
++#define IOREQ_WRITE 0
++
++#define STATE_IOREQ_NONE 0
++#define STATE_IOREQ_READY 1
++#define STATE_IOREQ_INPROCESS 2
++#define STATE_IORESP_READY 3
++
++#define IOREQ_TYPE_PIO 0 /* pio */
++#define IOREQ_TYPE_COPY 1 /* mmio ops */
++#define IOREQ_TYPE_TIMEOFFSET 7
++#define IOREQ_TYPE_INVALIDATE 8 /* mapcache */
++
++/*
++ * VMExit dispatcher should cooperate with instruction decoder to
++ * prepare this structure and notify service OS and DM by sending
++ * virq
++ */
++struct ioreq {
++ uint64_t addr; /* physical address */
++ uint64_t size; /* size in bytes */
++ uint64_t count; /* for rep prefixes */
++ uint64_t data; /* data (or paddr of data) */
++ uint8_t state:4;
++ uint8_t data_is_ptr:1; /* if 1, data above is the guest paddr
++ * of the real data to use. */
++ uint8_t dir:1; /* 1=read, 0=write */
++ uint8_t df:1;
++ uint8_t pad:1;
++ uint8_t type; /* I/O type */
++ uint8_t _pad0[6];
++ uint64_t io_count; /* How many IO done on a vcpu */
++};
++typedef struct ioreq ioreq_t;
++
++struct vcpu_iodata {
++ struct ioreq vp_ioreq;
++ /* Event channel port, used for notifications to/from the device model. */
++ uint32_t vp_eport;
++ uint32_t _pad0;
++};
++typedef struct vcpu_iodata vcpu_iodata_t;
++
++struct shared_iopage {
++ struct vcpu_iodata vcpu_iodata[1];
++};
++typedef struct shared_iopage shared_iopage_t;
++
++struct buf_ioreq {
++ uint8_t type; /* I/O type */
++ uint8_t pad:1;
++ uint8_t dir:1; /* 1=read, 0=write */
++ uint8_t size:2; /* 0=>1, 1=>2, 2=>4, 3=>8. If 8, use two buf_ioreqs */
++ uint32_t addr:20;/* physical address */
++ uint32_t data; /* data */
++};
++typedef struct buf_ioreq buf_ioreq_t;
++
++#define IOREQ_BUFFER_SLOT_NUM 511 /* 8 bytes each, plus 2 4-byte indexes */
++struct buffered_iopage {
++ unsigned int read_pointer;
++ unsigned int write_pointer;
++ buf_ioreq_t buf_ioreq[IOREQ_BUFFER_SLOT_NUM];
++}; /* NB. Size of this structure must be no greater than one page. */
++typedef struct buffered_iopage buffered_iopage_t;
++
++#if defined(__ia64__)
++struct pio_buffer {
++ uint32_t page_offset;
++ uint32_t pointer;
++ uint32_t data_end;
++ uint32_t buf_size;
++ void *opaque;
++};
++
++#define PIO_BUFFER_IDE_PRIMARY 0 /* I/O port = 0x1F0 */
++#define PIO_BUFFER_IDE_SECONDARY 1 /* I/O port = 0x170 */
++#define PIO_BUFFER_ENTRY_NUM 2
++struct buffered_piopage {
++ struct pio_buffer pio[PIO_BUFFER_ENTRY_NUM];
++ uint8_t buffer[1];
++};
++#endif /* defined(__ia64__) */
++
++#define ACPI_PM1A_EVT_BLK_ADDRESS 0x0000000000001f40
++#define ACPI_PM1A_CNT_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x04)
++#define ACPI_PM_TMR_BLK_ADDRESS (ACPI_PM1A_EVT_BLK_ADDRESS + 0x08)
++#define ACPI_GPE0_BLK_ADDRESS (ACPI_PM_TMR_BLK_ADDRESS + 0x20)
++#define ACPI_GPE0_BLK_LEN 0x08
++
++#endif /* _IOREQ_H_ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/hvm/params.h b/include/xen/interface/hvm/params.h
+new file mode 100644
+index 0000000..f222cba
+--- /dev/null
++++ b/include/xen/interface/hvm/params.h
+@@ -0,0 +1,95 @@
++/*
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_PARAMS_H__
++#define __XEN_PUBLIC_HVM_PARAMS_H__
++
++#include "hvm_op.h"
++
++/*
++ * Parameter space for HVMOP_{set,get}_param.
++ */
++
++/*
++ * How should CPU0 event-channel notifications be delivered?
++ * val[63:56] == 0: val[55:0] is a delivery GSI (Global System Interrupt).
++ * val[63:56] == 1: val[55:0] is a delivery PCI INTx line, as follows:
++ * Domain = val[47:32], Bus = val[31:16],
++ * DevFn = val[15: 8], IntX = val[ 1: 0]
++ * If val == 0 then CPU0 event-channel notifications are not delivered.
++ */
++#define HVM_PARAM_CALLBACK_IRQ 0
++
++/*
++ * These are not used by Xen. They are here for convenience of HVM-guest
++ * xenbus implementations.
++ */
++#define HVM_PARAM_STORE_PFN 1
++#define HVM_PARAM_STORE_EVTCHN 2
++
++#define HVM_PARAM_PAE_ENABLED 4
++
++#define HVM_PARAM_IOREQ_PFN 5
++
++#define HVM_PARAM_BUFIOREQ_PFN 6
++
++#ifdef __ia64__
++#define HVM_PARAM_NVRAM_FD 7
++#define HVM_PARAM_VHPT_SIZE 8
++#define HVM_PARAM_BUFPIOREQ_PFN 9
++#endif
++
++/*
++ * Set mode for virtual timers (currently x86 only):
++ * delay_for_missed_ticks (default):
++ * Do not advance a vcpu's time beyond the correct delivery time for
++ * interrupts that have been missed due to preemption. Deliver missed
++ * interrupts when the vcpu is rescheduled and advance the vcpu's virtual
++ * time stepwise for each one.
++ * no_delay_for_missed_ticks:
++ * As above, missed interrupts are delivered, but guest time always tracks
++ * wallclock (i.e., real) time while doing so.
++ * no_missed_ticks_pending:
++ * No missed interrupts are held pending. Instead, to ensure ticks are
++ * delivered at some non-zero rate, if we detect missed ticks then the
++ * internal tick alarm is not disabled if the VCPU is preempted during the
++ * next tick period.
++ * one_missed_tick_pending:
++ * Missed interrupts are collapsed together and delivered as one 'late tick'.
++ * Guest time always tracks wallclock (i.e., real) time.
++ */
++#define HVM_PARAM_TIMER_MODE 10
++#define HVMPTM_delay_for_missed_ticks 0
++#define HVMPTM_no_delay_for_missed_ticks 1
++#define HVMPTM_no_missed_ticks_pending 2
++#define HVMPTM_one_missed_tick_pending 3
++
++/* Boolean: Enable virtual HPET (high-precision event timer)? (x86-only) */
++#define HVM_PARAM_HPET_ENABLED 11
++
++/* Identity-map page directory used by Intel EPT when CR0.PG=0. */
++#define HVM_PARAM_IDENT_PT 12
++
++/* Device Model domain, defaults to 0. */
++#define HVM_PARAM_DM_DOMAIN 13
++
++#define HVM_NR_PARAMS 14
++
++#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
+diff --git a/include/xen/interface/hvm/save.h b/include/xen/interface/hvm/save.h
+new file mode 100644
+index 0000000..d45f0c1
+--- /dev/null
++++ b/include/xen/interface/hvm/save.h
+@@ -0,0 +1,88 @@
++/*
++ * hvm/save.h
++ *
++ * Structure definitions for HVM state that is held by Xen and must
++ * be saved along with the domain's memory and device-model state.
++ *
++ * Copyright (c) 2007 XenSource Ltd.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XEN_PUBLIC_HVM_SAVE_H__
++#define __XEN_PUBLIC_HVM_SAVE_H__
++
++/*
++ * Structures in this header *must* have the same layout in 32bit
++ * and 64bit environments: this means that all fields must be explicitly
++ * sized types and aligned to their sizes, and the structs must be
++ * a multiple of eight bytes long.
++ *
++ * Only the state necessary for saving and restoring (i.e. fields
++ * that are analogous to actual hardware state) should go in this file.
++ * Internal mechanisms should be kept in Xen-private headers.
++ */
++
++#if !defined(__GNUC__) || defined(__STRICT_ANSI__)
++#error "Anonymous structs/unions are a GNU extension."
++#endif
++
++/*
++ * Each entry is preceded by a descriptor giving its type and length
++ */
++struct hvm_save_descriptor {
++ uint16_t typecode; /* Used to demux the various types below */
++ uint16_t instance; /* Further demux within a type */
++ uint32_t length; /* In bytes, *not* including this descriptor */
++};
++
++
++/*
++ * Each entry has a datatype associated with it: for example, the CPU state
++ * is saved as a HVM_SAVE_TYPE(CPU), which has HVM_SAVE_LENGTH(CPU),
++ * and is identified by a descriptor with typecode HVM_SAVE_CODE(CPU).
++ * DECLARE_HVM_SAVE_TYPE binds these things together with some type-system
++ * ugliness.
++ */
++
++#define DECLARE_HVM_SAVE_TYPE(_x, _code, _type) \
++ struct __HVM_SAVE_TYPE_##_x { _type t; char c[_code]; }
++
++#define HVM_SAVE_TYPE(_x) typeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->t)
++#define HVM_SAVE_LENGTH(_x) (sizeof (HVM_SAVE_TYPE(_x)))
++#define HVM_SAVE_CODE(_x) (sizeof (((struct __HVM_SAVE_TYPE_##_x *)(0))->c))
++
++
++/*
++ * The series of save records is teminated by a zero-type, zero-length
++ * descriptor.
++ */
++
++struct hvm_save_end {};
++DECLARE_HVM_SAVE_TYPE(END, 0, struct hvm_save_end);
++
++#if defined(__i386__) || defined(__x86_64__)
++#include "../arch-x86/hvm/save.h"
++#elif defined(__ia64__)
++#include "../arch-ia64/hvm/save.h"
++#else
++#error "unsupported architecture"
++#endif
++
++#endif /* __XEN_PUBLIC_HVM_SAVE_H__ */
+diff --git a/include/xen/interface/hvm/vmx_assist.h b/include/xen/interface/hvm/vmx_assist.h
+new file mode 100644
+index 0000000..4ef17fe
+--- /dev/null
++++ b/include/xen/interface/hvm/vmx_assist.h
+@@ -0,0 +1,122 @@
++/*
++ * vmx_assist.h: Context definitions for the VMXASSIST world switch.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Leendert van Doorn, leendert at watson.ibm.com
++ * Copyright (c) 2005, International Business Machines Corporation.
++ */
++
++#ifndef _VMX_ASSIST_H_
++#define _VMX_ASSIST_H_
++
++#define VMXASSIST_BASE 0xD0000
++#define VMXASSIST_MAGIC 0x17101966
++#define VMXASSIST_MAGIC_OFFSET (VMXASSIST_BASE+8)
++
++#define VMXASSIST_NEW_CONTEXT (VMXASSIST_BASE + 12)
++#define VMXASSIST_OLD_CONTEXT (VMXASSIST_NEW_CONTEXT + 4)
++
++#ifndef __ASSEMBLY__
++
++#define NR_EXCEPTION_HANDLER 32
++#define NR_INTERRUPT_HANDLERS 16
++#define NR_TRAPS (NR_EXCEPTION_HANDLER+NR_INTERRUPT_HANDLERS)
++
++union vmcs_arbytes {
++ struct arbyte_fields {
++ unsigned int seg_type : 4,
++ s : 1,
++ dpl : 2,
++ p : 1,
++ reserved0 : 4,
++ avl : 1,
++ reserved1 : 1,
++ default_ops_size: 1,
++ g : 1,
++ null_bit : 1,
++ reserved2 : 15;
++ } fields;
++ unsigned int bytes;
++};
++
++/*
++ * World switch state
++ */
++struct vmx_assist_context {
++ uint32_t eip; /* execution pointer */
++ uint32_t esp; /* stack pointer */
++ uint32_t eflags; /* flags register */
++ uint32_t cr0;
++ uint32_t cr3; /* page table directory */
++ uint32_t cr4;
++ uint32_t idtr_limit; /* idt */
++ uint32_t idtr_base;
++ uint32_t gdtr_limit; /* gdt */
++ uint32_t gdtr_base;
++ uint32_t cs_sel; /* cs selector */
++ uint32_t cs_limit;
++ uint32_t cs_base;
++ union vmcs_arbytes cs_arbytes;
++ uint32_t ds_sel; /* ds selector */
++ uint32_t ds_limit;
++ uint32_t ds_base;
++ union vmcs_arbytes ds_arbytes;
++ uint32_t es_sel; /* es selector */
++ uint32_t es_limit;
++ uint32_t es_base;
++ union vmcs_arbytes es_arbytes;
++ uint32_t ss_sel; /* ss selector */
++ uint32_t ss_limit;
++ uint32_t ss_base;
++ union vmcs_arbytes ss_arbytes;
++ uint32_t fs_sel; /* fs selector */
++ uint32_t fs_limit;
++ uint32_t fs_base;
++ union vmcs_arbytes fs_arbytes;
++ uint32_t gs_sel; /* gs selector */
++ uint32_t gs_limit;
++ uint32_t gs_base;
++ union vmcs_arbytes gs_arbytes;
++ uint32_t tr_sel; /* task selector */
++ uint32_t tr_limit;
++ uint32_t tr_base;
++ union vmcs_arbytes tr_arbytes;
++ uint32_t ldtr_sel; /* ldtr selector */
++ uint32_t ldtr_limit;
++ uint32_t ldtr_base;
++ union vmcs_arbytes ldtr_arbytes;
++
++ unsigned char rm_irqbase[2];
++};
++typedef struct vmx_assist_context vmx_assist_context_t;
++
++#endif /* __ASSEMBLY__ */
++
++#endif /* _VMX_ASSIST_H_ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/blkif.h b/include/xen/interface/io/blkif.h
+index c2d1fa4..3b291ff 100644
+--- a/include/xen/interface/io/blkif.h
++++ b/include/xen/interface/io/blkif.h
+@@ -3,6 +3,24 @@
+ *
+ * Unified block-device I/O interface for Xen guest OSes.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
+@@ -24,8 +42,10 @@
+ * rsp_event appropriately (e.g., using RING_FINAL_CHECK_FOR_RESPONSES()).
+ */
+
+-typedef uint16_t blkif_vdev_t;
+-typedef uint64_t blkif_sector_t;
++#ifndef blkif_vdev_t
++#define blkif_vdev_t uint16_t
++#endif
++#define blkif_sector_t uint64_t
+
+ /*
+ * REQUEST CODES.
+@@ -34,7 +54,7 @@ typedef uint64_t blkif_sector_t;
+ #define BLKIF_OP_WRITE 1
+ /*
+ * Recognised only if "feature-barrier" is present in backend xenbus info.
+- * The "feature_barrier" node contains a boolean indicating whether barrier
++ * The "feature-barrier" node contains a boolean indicating whether barrier
+ * requests are likely to succeed or fail. Either way, a barrier request
+ * may fail at any time with BLKIF_RSP_EOPNOTSUPP if it is unsupported by
+ * the underlying block-device hardware. The boolean simply indicates whether
+@@ -43,33 +63,50 @@ typedef uint64_t blkif_sector_t;
+ * create the "feature-barrier" node!
+ */
+ #define BLKIF_OP_WRITE_BARRIER 2
++/*
++ * Recognised if "feature-flush-cache" is present in backend xenbus
++ * info. A flush will ask the underlying storage hardware to flush its
++ * non-volatile caches as appropriate. The "feature-flush-cache" node
++ * contains a boolean indicating whether flush requests are likely to
++ * succeed or fail. Either way, a flush request may fail at any time
++ * with BLKIF_RSP_EOPNOTSUPP if it is unsupported by the underlying
++ * block-device hardware. The boolean simply indicates whether or not it
++ * is worthwhile for the frontend to attempt flushes. If a backend does
++ * not recognise BLKIF_OP_WRITE_FLUSH_CACHE, it should *not* create the
++ * "feature-flush-cache" node!
++ */
++#define BLKIF_OP_FLUSH_DISKCACHE 3
+
+ /*
+ * Maximum scatter/gather segments per request.
+- * This is carefully chosen so that sizeof(struct blkif_ring) <= PAGE_SIZE.
++ * This is carefully chosen so that sizeof(blkif_ring_t) <= PAGE_SIZE.
+ * NB. This could be 12 if the ring indexes weren't stored in the same page.
+ */
+ #define BLKIF_MAX_SEGMENTS_PER_REQUEST 11
+
++struct blkif_request_segment {
++ grant_ref_t gref; /* reference to I/O buffer frame */
++ /* @first_sect: first sector in frame to transfer (inclusive). */
++ /* @last_sect: last sector in frame to transfer (inclusive). */
++ uint8_t first_sect, last_sect;
++};
++
+ struct blkif_request {
+- uint8_t operation; /* BLKIF_OP_??? */
+- uint8_t nr_segments; /* number of segments */
+- blkif_vdev_t handle; /* only for read/write requests */
+- uint64_t id; /* private guest value, echoed in resp */
+- blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
+- struct blkif_request_segment {
+- grant_ref_t gref; /* reference to I/O buffer frame */
+- /* @first_sect: first sector in frame to transfer (inclusive). */
+- /* @last_sect: last sector in frame to transfer (inclusive). */
+- uint8_t first_sect, last_sect;
+- } seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
++ uint8_t operation; /* BLKIF_OP_??? */
++ uint8_t nr_segments; /* number of segments */
++ blkif_vdev_t handle; /* only for read/write requests */
++ uint64_t id; /* private guest value, echoed in resp */
++ blkif_sector_t sector_number;/* start sector idx on disk (r/w only) */
++ struct blkif_request_segment seg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
+ };
++typedef struct blkif_request blkif_request_t;
+
+ struct blkif_response {
+- uint64_t id; /* copied from request */
+- uint8_t operation; /* copied from request */
+- int16_t status; /* BLKIF_RSP_??? */
++ uint64_t id; /* copied from request */
++ uint8_t operation; /* copied from request */
++ int16_t status; /* BLKIF_RSP_??? */
+ };
++typedef struct blkif_response blkif_response_t;
+
+ /*
+ * STATUS RETURN CODES.
+@@ -92,3 +129,13 @@ DEFINE_RING_TYPES(blkif, struct blkif_request, struct blkif_response);
+ #define VDISK_READONLY 0x4
+
+ #endif /* __XEN_PUBLIC_IO_BLKIF_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/console.h b/include/xen/interface/io/console.h
+index e563de7..6716145 100644
+--- a/include/xen/interface/io/console.h
++++ b/include/xen/interface/io/console.h
+@@ -3,6 +3,24 @@
+ *
+ * Console I/O interface for Xen guest OSes.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2005, Keir Fraser
+ */
+
+@@ -21,3 +39,13 @@ struct xencons_interface {
+ };
+
+ #endif /* __XEN_PUBLIC_IO_CONSOLE_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/fbif.h b/include/xen/interface/io/fbif.h
+index 5a934dd..3f346fe 100644
+--- a/include/xen/interface/io/fbif.h
++++ b/include/xen/interface/io/fbif.h
+@@ -41,20 +41,39 @@
+ */
+ #define XENFB_TYPE_UPDATE 2
+
+-struct xenfb_update {
+- uint8_t type; /* XENFB_TYPE_UPDATE */
+- int32_t x; /* source x */
+- int32_t y; /* source y */
+- int32_t width; /* rect width */
+- int32_t height; /* rect height */
++struct xenfb_update
++{
++ uint8_t type; /* XENFB_TYPE_UPDATE */
++ int32_t x; /* source x */
++ int32_t y; /* source y */
++ int32_t width; /* rect width */
++ int32_t height; /* rect height */
++};
++
++/*
++ * Framebuffer resize notification event
++ * Capable backend sets feature-resize in xenstore.
++ */
++#define XENFB_TYPE_RESIZE 3
++
++struct xenfb_resize
++{
++ uint8_t type; /* XENFB_TYPE_RESIZE */
++ int32_t width; /* width in pixels */
++ int32_t height; /* height in pixels */
++ int32_t stride; /* stride in bytes */
++ int32_t depth; /* depth in bits */
++ int32_t offset; /* offset of the framebuffer in bytes */
+ };
+
+ #define XENFB_OUT_EVENT_SIZE 40
+
+-union xenfb_out_event {
+- uint8_t type;
+- struct xenfb_update update;
+- char pad[XENFB_OUT_EVENT_SIZE];
++union xenfb_out_event
++{
++ uint8_t type;
++ struct xenfb_update update;
++ struct xenfb_resize resize;
++ char pad[XENFB_OUT_EVENT_SIZE];
+ };
+
+ /* In events (backend -> frontend) */
+@@ -66,9 +85,10 @@ union xenfb_out_event {
+
+ #define XENFB_IN_EVENT_SIZE 40
+
+-union xenfb_in_event {
+- uint8_t type;
+- char pad[XENFB_IN_EVENT_SIZE];
++union xenfb_in_event
++{
++ uint8_t type;
++ char pad[XENFB_IN_EVENT_SIZE];
+ };
+
+ /* shared page */
+@@ -77,43 +97,51 @@ union xenfb_in_event {
+ #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))
++ ((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])
++ (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))
++ ((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 {
+- uint32_t in_cons, in_prod;
+- uint32_t out_cons, out_prod;
+-
+- int32_t width; /* width of the framebuffer (in pixels) */
+- int32_t height; /* height of the framebuffer (in pixels) */
+- uint32_t line_length; /* length of a row of pixels (in bytes) */
+- uint32_t mem_length; /* length of the framebuffer (in bytes) */
+- uint8_t depth; /* 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.
+- */
++ (XENFB_OUT_RING((page))[(idx) % XENFB_OUT_RING_LEN])
++
++struct xenfb_page
++{
++ uint32_t in_cons, in_prod;
++ uint32_t out_cons, out_prod;
++
++ int32_t width; /* the width of the framebuffer (in pixels) */
++ int32_t height; /* the height of the framebuffer (in pixels) */
++ uint32_t line_length; /* the length of a row of pixels (in bytes) */
++ uint32_t mem_length; /* the length of the framebuffer (in bytes) */
++ uint8_t depth; /* the depth of a pixel (in bits) */
++
++ /*
++ * Framebuffer page directory
++ *
++ * Each directory page holds PAGE_SIZE / sizeof(*pd)
++ * framebuffer pages, and can thus map up to PAGE_SIZE *
++ * PAGE_SIZE / sizeof(*pd) bytes. With PAGE_SIZE == 4096 and
++ * sizeof(unsigned long) == 4/8, that's 4 Megs 32 bit and 2 Megs
++ * 64 bit. 256 directories give enough room for a 512 Meg
++ * framebuffer with a max resolution of 12,800x10,240. Should
++ * be enough for a while with room leftover for expansion.
++ */
++#ifndef CONFIG_PARAVIRT_XEN
++ unsigned long pd[256];
++#else
++ /* Two directory pages should be enough for a while. */
+ unsigned long pd[2];
++#endif
+ };
+
+ /*
+- * Wart: xenkbd needs to know resolution. Put it here until a better
+- * solution is found, but don't leak it to the backend.
++ * Wart: xenkbd needs to know default resolution. Put it here until a
++ * better solution is found, but don't leak it to the backend.
+ */
+ #ifdef __KERNEL__
+ #define XENFB_WIDTH 800
+@@ -122,3 +150,13 @@ struct xenfb_page {
+ #endif
+
+ #endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/fsif.h b/include/xen/interface/io/fsif.h
+new file mode 100644
+index 0000000..8abc075
+--- /dev/null
++++ b/include/xen/interface/io/fsif.h
+@@ -0,0 +1,181 @@
++/******************************************************************************
++ * fsif.h
++ *
++ * Interface to FS level split device drivers.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2007, Grzegorz Milos, Sun Microsystems, Inc.
++ */
++
++#ifndef __XEN_PUBLIC_IO_FSIF_H__
++#define __XEN_PUBLIC_IO_FSIF_H__
++
++#include "ring.h"
++#include "../grant_table.h"
++
++#define REQ_FILE_OPEN 1
++#define REQ_FILE_CLOSE 2
++#define REQ_FILE_READ 3
++#define REQ_FILE_WRITE 4
++#define REQ_STAT 5
++#define REQ_FILE_TRUNCATE 6
++#define REQ_REMOVE 7
++#define REQ_RENAME 8
++#define REQ_CREATE 9
++#define REQ_DIR_LIST 10
++#define REQ_CHMOD 11
++#define REQ_FS_SPACE 12
++#define REQ_FILE_SYNC 13
++
++struct fsif_open_request {
++ grant_ref_t gref;
++};
++
++struct fsif_close_request {
++ uint32_t fd;
++};
++
++struct fsif_read_request {
++ uint32_t fd;
++ grant_ref_t gref;
++ uint64_t len;
++ uint64_t offset;
++};
++
++struct fsif_write_request {
++ uint32_t fd;
++ grant_ref_t gref;
++ uint64_t len;
++ uint64_t offset;
++};
++
++struct fsif_stat_request {
++ uint32_t fd;
++ grant_ref_t gref;
++};
++
++/* This structure is a copy of some fields from stat structure, writen to the
++ * granted page. */
++struct fsif_stat_response {
++ int32_t stat_mode;
++ uint32_t stat_uid;
++ uint32_t stat_gid;
++ int32_t pad;
++ int64_t stat_size;
++ int64_t stat_atime;
++ int64_t stat_mtime;
++ int64_t stat_ctime;
++};
++
++struct fsif_truncate_request {
++ uint32_t fd;
++ int32_t pad;
++ int64_t length;
++};
++
++struct fsif_remove_request {
++ grant_ref_t gref;
++};
++
++struct fsif_rename_request {
++ uint16_t old_name_offset;
++ uint16_t new_name_offset;
++ grant_ref_t gref;
++};
++
++struct fsif_create_request {
++ int8_t directory;
++ int8_t pad;
++ int16_t pad2;
++ int32_t mode;
++ grant_ref_t gref;
++};
++
++struct fsif_list_request {
++ uint32_t offset;
++ grant_ref_t gref;
++};
++
++#define NR_FILES_SHIFT 0
++#define NR_FILES_SIZE 16 /* 16 bits for the number of files mask */
++#define NR_FILES_MASK (((1ULL << NR_FILES_SIZE) - 1) << NR_FILES_SHIFT)
++#define ERROR_SIZE 32 /* 32 bits for the error mask */
++#define ERROR_SHIFT (NR_FILES_SIZE + NR_FILES_SHIFT)
++#define ERROR_MASK (((1ULL << ERROR_SIZE) - 1) << ERROR_SHIFT)
++#define HAS_MORE_SHIFT (ERROR_SHIFT + ERROR_SIZE)
++#define HAS_MORE_FLAG (1ULL << HAS_MORE_SHIFT)
++
++struct fsif_chmod_request {
++ uint32_t fd;
++ int32_t mode;
++};
++
++struct fsif_space_request {
++ grant_ref_t gref;
++};
++
++struct fsif_sync_request {
++ uint32_t fd;
++};
++
++
++/* FS operation request */
++struct fsif_request {
++ uint8_t type; /* Type of the request */
++ uint8_t pad;
++ uint16_t id; /* Request ID, copied to the response */
++ uint32_t pad2;
++ union {
++ struct fsif_open_request fopen;
++ struct fsif_close_request fclose;
++ struct fsif_read_request fread;
++ struct fsif_write_request fwrite;
++ struct fsif_stat_request fstat;
++ struct fsif_truncate_request ftruncate;
++ struct fsif_remove_request fremove;
++ struct fsif_rename_request frename;
++ struct fsif_create_request fcreate;
++ struct fsif_list_request flist;
++ struct fsif_chmod_request fchmod;
++ struct fsif_space_request fspace;
++ struct fsif_sync_request fsync;
++ } u;
++};
++typedef struct fsif_request fsif_request_t;
++
++/* FS operation response */
++struct fsif_response {
++ uint16_t id;
++ uint16_t pad1;
++ uint32_t pad2;
++ uint64_t ret_val;
++};
++
++typedef struct fsif_response fsif_response_t;
++
++
++DEFINE_RING_TYPES(fsif, struct fsif_request, struct fsif_response);
++
++#define STATE_INITIALISED "init"
++#define STATE_READY "ready"
++
++
++
++#endif
+diff --git a/include/xen/interface/io/kbdif.h b/include/xen/interface/io/kbdif.h
+index fb97f42..e1d66a5 100644
+--- a/include/xen/interface/io/kbdif.h
++++ b/include/xen/interface/io/kbdif.h
+@@ -45,32 +45,38 @@
+ */
+ #define XENKBD_TYPE_POS 4
+
+-struct xenkbd_motion {
+- uint8_t type; /* XENKBD_TYPE_MOTION */
+- int32_t rel_x; /* relative X motion */
+- int32_t rel_y; /* relative Y motion */
++struct xenkbd_motion
++{
++ uint8_t type; /* XENKBD_TYPE_MOTION */
++ int32_t rel_x; /* relative X motion */
++ int32_t rel_y; /* relative Y motion */
++ int32_t rel_z; /* relative Z motion (wheel) */
+ };
+
+-struct xenkbd_key {
+- uint8_t type; /* XENKBD_TYPE_KEY */
+- uint8_t pressed; /* 1 if pressed; 0 otherwise */
+- uint32_t keycode; /* KEY_* from linux/input.h */
++struct xenkbd_key
++{
++ uint8_t type; /* XENKBD_TYPE_KEY */
++ uint8_t pressed; /* 1 if pressed; 0 otherwise */
++ uint32_t keycode; /* KEY_* from linux/input.h */
+ };
+
+-struct xenkbd_position {
+- uint8_t type; /* XENKBD_TYPE_POS */
+- int32_t abs_x; /* absolute X position (in FB pixels) */
+- int32_t abs_y; /* absolute Y position (in FB pixels) */
++struct xenkbd_position
++{
++ uint8_t type; /* XENKBD_TYPE_POS */
++ int32_t abs_x; /* absolute X position (in FB pixels) */
++ int32_t abs_y; /* absolute Y position (in FB pixels) */
++ int32_t rel_z; /* relative Z motion (wheel) */
+ };
+
+ #define XENKBD_IN_EVENT_SIZE 40
+
+-union xenkbd_in_event {
+- uint8_t type;
+- struct xenkbd_motion motion;
+- struct xenkbd_key key;
+- struct xenkbd_position pos;
+- char pad[XENKBD_IN_EVENT_SIZE];
++union xenkbd_in_event
++{
++ uint8_t type;
++ struct xenkbd_motion motion;
++ struct xenkbd_key key;
++ struct xenkbd_position pos;
++ char pad[XENKBD_IN_EVENT_SIZE];
+ };
+
+ /* Out events (frontend -> backend) */
+@@ -83,9 +89,10 @@ union xenkbd_in_event {
+
+ #define XENKBD_OUT_EVENT_SIZE 40
+
+-union xenkbd_out_event {
+- uint8_t type;
+- char pad[XENKBD_OUT_EVENT_SIZE];
++union xenkbd_out_event
++{
++ uint8_t type;
++ char pad[XENKBD_OUT_EVENT_SIZE];
+ };
+
+ /* shared page */
+@@ -94,21 +101,32 @@ union xenkbd_out_event {
+ #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))
++ ((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])
++ (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))
++ ((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])
++ (XENKBD_OUT_RING((page))[(idx) % XENKBD_OUT_RING_LEN])
+
+-struct xenkbd_page {
+- uint32_t in_cons, in_prod;
+- uint32_t out_cons, out_prod;
++struct xenkbd_page
++{
++ uint32_t in_cons, in_prod;
++ uint32_t out_cons, out_prod;
+ };
+
+ #endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/netif.h b/include/xen/interface/io/netif.h
+index 518481c..14f1564 100644
+--- a/include/xen/interface/io/netif.h
++++ b/include/xen/interface/io/netif.h
+@@ -3,6 +3,24 @@
+ *
+ * Unified network-device I/O interface for Xen guest OSes.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2003-2004, Keir Fraser
+ */
+
+@@ -47,18 +65,21 @@
+ #define _NETTXF_extra_info (3)
+ #define NETTXF_extra_info (1U<<_NETTXF_extra_info)
+
+-struct xen_netif_tx_request {
++struct netif_tx_request {
+ grant_ref_t gref; /* Reference to buffer page */
+ uint16_t offset; /* Offset within buffer page */
+ uint16_t flags; /* NETTXF_* */
+ uint16_t id; /* Echoed in response message. */
+ uint16_t size; /* Packet size in bytes. */
+ };
++typedef struct netif_tx_request netif_tx_request_t;
+
+ /* Types of netif_extra_info descriptors. */
+-#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
+-#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
+-#define XEN_NETIF_EXTRA_TYPE_MAX (2)
++#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
++#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
++#define XEN_NETIF_EXTRA_TYPE_MCAST_ADD (2) /* u.mcast */
++#define XEN_NETIF_EXTRA_TYPE_MCAST_DEL (3) /* u.mcast */
++#define XEN_NETIF_EXTRA_TYPE_MAX (4)
+
+ /* netif_extra_info flags. */
+ #define _XEN_NETIF_EXTRA_FLAG_MORE (0)
+@@ -71,49 +92,68 @@ struct xen_netif_tx_request {
+ * This structure needs to fit within both netif_tx_request and
+ * netif_rx_response for compatibility.
+ */
+-struct xen_netif_extra_info {
+- uint8_t type; /* XEN_NETIF_EXTRA_TYPE_* */
+- uint8_t flags; /* XEN_NETIF_EXTRA_FLAG_* */
+-
+- union {
+- struct {
+- /*
+- * Maximum payload size of each segment. For
+- * example, for TCP this is just the path MSS.
+- */
+- uint16_t size;
+-
+- /*
+- * GSO type. This determines the protocol of
+- * the packet and any extra features required
+- * to segment the packet properly.
+- */
+- uint8_t type; /* XEN_NETIF_GSO_TYPE_* */
+-
+- /* Future expansion. */
+- uint8_t pad;
+-
+- /*
+- * GSO features. This specifies any extra GSO
+- * features required to process this packet,
+- * such as ECN support for TCPv4.
+- */
+- uint16_t features; /* XEN_NETIF_GSO_FEAT_* */
+- } gso;
+-
+- uint16_t pad[3];
+- } u;
++struct netif_extra_info {
++ uint8_t type; /* XEN_NETIF_EXTRA_TYPE_* */
++ uint8_t flags; /* XEN_NETIF_EXTRA_FLAG_* */
++
++ union {
++ /*
++ * XEN_NETIF_EXTRA_TYPE_GSO:
++ */
++ struct {
++ /*
++ * Maximum payload size of each segment. For example, for TCP this
++ * is just the path MSS.
++ */
++ uint16_t size;
++
++ /*
++ * GSO type. This determines the protocol of the packet and any
++ * extra features required to segment the packet properly.
++ */
++ uint8_t type; /* XEN_NETIF_GSO_TYPE_* */
++
++ /* Future expansion. */
++ uint8_t pad;
++
++ /*
++ * GSO features. This specifies any extra GSO features required
++ * to process this packet, such as ECN support for TCPv4.
++ */
++ uint16_t features; /* XEN_NETIF_GSO_FEAT_* */
++ } gso;
++
++ /*
++ * XEN_NETIF_EXTRA_TYPE_MCAST_{ADD,DEL}:
++ * Backend advertises availability via 'feature-multicast-control'
++ * xenbus node containing value '1'.
++ * Frontend requests this feature by advertising
++ * 'request-multicast-control' xenbus node containing value '1'.
++ * If multicast control is requested then multicast flooding is
++ * disabled and the frontend must explicitly register its interest
++ * in multicast groups using dummy transmit requests containing
++ * MCAST_{ADD,DEL} extra-info fragments.
++ */
++ struct {
++ uint8_t addr[6]; /* Address to add/remove. */
++ } mcast;
++
++ uint16_t pad[3];
++ } u;
+ };
++typedef struct netif_extra_info netif_extra_info_t;
+
+-struct xen_netif_tx_response {
+- uint16_t id;
+- int16_t status; /* NETIF_RSP_* */
++struct netif_tx_response {
++ uint16_t id;
++ int16_t status; /* NETIF_RSP_* */
+ };
++typedef struct netif_tx_response netif_tx_response_t;
+
+-struct xen_netif_rx_request {
+- uint16_t id; /* Echoed in response message. */
+- grant_ref_t gref; /* Reference to incoming granted frame */
++struct netif_rx_request {
++ uint16_t id; /* Echoed in response message. */
++ grant_ref_t gref; /* Reference to incoming granted frame */
+ };
++typedef struct netif_rx_request netif_rx_request_t;
+
+ /* Packet data has been validated against protocol checksum. */
+ #define _NETRXF_data_validated (0)
+@@ -131,23 +171,34 @@ struct xen_netif_rx_request {
+ #define _NETRXF_extra_info (3)
+ #define NETRXF_extra_info (1U<<_NETRXF_extra_info)
+
+-struct xen_netif_rx_response {
++struct netif_rx_response {
+ uint16_t id;
+ uint16_t offset; /* Offset in page of start of received packet */
+ uint16_t flags; /* NETRXF_* */
+ int16_t status; /* -ve: BLKIF_RSP_* ; +ve: Rx'ed pkt size. */
+ };
++typedef struct netif_rx_response netif_rx_response_t;
+
+ /*
+ * Generate netif ring structures and types.
+ */
+
++#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H)
++DEFINE_RING_TYPES(netif_tx, struct netif_tx_request, struct netif_tx_response);
++DEFINE_RING_TYPES(netif_rx, struct netif_rx_request, struct netif_rx_response);
++#else
++#define xen_netif_tx_request netif_tx_request
++#define xen_netif_rx_request netif_rx_request
++#define xen_netif_tx_response netif_tx_response
++#define xen_netif_rx_response netif_rx_response
+ DEFINE_RING_TYPES(xen_netif_tx,
+ struct xen_netif_tx_request,
+ struct xen_netif_tx_response);
+ DEFINE_RING_TYPES(xen_netif_rx,
+ struct xen_netif_rx_request,
+ struct xen_netif_rx_response);
++#define xen_netif_extra_info netif_extra_info
++#endif
+
+ #define NETIF_RSP_DROPPED -2
+ #define NETIF_RSP_ERROR -1
+@@ -156,3 +207,13 @@ DEFINE_RING_TYPES(xen_netif_rx,
+ #define NETIF_RSP_NULL 1
+
+ #endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/pciif.h b/include/xen/interface/io/pciif.h
+new file mode 100644
+index 0000000..0a0ffcc
+--- /dev/null
++++ b/include/xen/interface/io/pciif.h
+@@ -0,0 +1,101 @@
++/*
++ * PCI Backend/Frontend Common Data Structures & Macros
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * 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.
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_PCI_COMMON_H__
++#define __XEN_PCI_COMMON_H__
++
++/* Be sure to bump this number if you change this file */
++#define XEN_PCI_MAGIC "7"
++
++/* xen_pci_sharedinfo flags */
++#define _XEN_PCIF_active (0)
++#define XEN_PCIF_active (1<<_XEN_PCI_active)
++
++/* xen_pci_op commands */
++#define XEN_PCI_OP_conf_read (0)
++#define XEN_PCI_OP_conf_write (1)
++#define XEN_PCI_OP_enable_msi (2)
++#define XEN_PCI_OP_disable_msi (3)
++#define XEN_PCI_OP_enable_msix (4)
++#define XEN_PCI_OP_disable_msix (5)
++
++/* xen_pci_op error numbers */
++#define XEN_PCI_ERR_success (0)
++#define XEN_PCI_ERR_dev_not_found (-1)
++#define XEN_PCI_ERR_invalid_offset (-2)
++#define XEN_PCI_ERR_access_denied (-3)
++#define XEN_PCI_ERR_not_implemented (-4)
++/* XEN_PCI_ERR_op_failed - backend failed to complete the operation */
++#define XEN_PCI_ERR_op_failed (-5)
++
++/*
++ * it should be PAGE_SIZE-sizeof(struct xen_pci_op))/sizeof(struct msix_entry))
++ * Should not exceed 128
++ */
++#define SH_INFO_MAX_VEC 128
++
++struct xen_msix_entry {
++ uint16_t vector;
++ uint16_t entry;
++};
++struct xen_pci_op {
++ /* IN: what action to perform: XEN_PCI_OP_* */
++ uint32_t cmd;
++
++ /* OUT: will contain an error number (if any) from errno.h */
++ int32_t err;
++
++ /* IN: which device to touch */
++ uint32_t domain; /* PCI Domain/Segment */
++ uint32_t bus;
++ uint32_t devfn;
++
++ /* IN: which configuration registers to touch */
++ int32_t offset;
++ int32_t size;
++
++ /* IN/OUT: Contains the result after a READ or the value to WRITE */
++ uint32_t value;
++ /* IN: Contains extra infor for this operation */
++ uint32_t info;
++ /*IN: param for msi-x */
++ struct xen_msix_entry msix_entries[SH_INFO_MAX_VEC];
++};
++
++struct xen_pci_sharedinfo {
++ /* flags - XEN_PCIF_* */
++ uint32_t flags;
++ struct xen_pci_op op;
++};
++
++#endif /* __XEN_PCI_COMMON_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/protocols.h b/include/xen/interface/io/protocols.h
+index 01fc8ae..5fbf10c 100644
+--- a/include/xen/interface/io/protocols.h
++++ b/include/xen/interface/io/protocols.h
+@@ -1,10 +1,31 @@
++/******************************************************************************
++ * protocols.h
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
+ #ifndef __XEN_PROTOCOLS_H__
+ #define __XEN_PROTOCOLS_H__
+
+ #define XEN_IO_PROTO_ABI_X86_32 "x86_32-abi"
+ #define XEN_IO_PROTO_ABI_X86_64 "x86_64-abi"
+ #define XEN_IO_PROTO_ABI_IA64 "ia64-abi"
+-#define XEN_IO_PROTO_ABI_POWERPC64 "powerpc64-abi"
+
+ #if defined(__i386__)
+ # define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_32
+@@ -12,8 +33,6 @@
+ # define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_X86_64
+ #elif defined(__ia64__)
+ # define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_IA64
+-#elif defined(__powerpc64__)
+-# define XEN_IO_PROTO_ABI_NATIVE XEN_IO_PROTO_ABI_POWERPC64
+ #else
+ # error arch fixup needed here
+ #endif
+diff --git a/include/xen/interface/io/ring.h b/include/xen/interface/io/ring.h
+index e8cbf43..5f6ed6a 100644
+--- a/include/xen/interface/io/ring.h
++++ b/include/xen/interface/io/ring.h
+@@ -3,16 +3,42 @@
+ *
+ * Shared producer-consumer ring macros.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Tim Deegan and Andrew Warfield November 2004.
+ */
+
+ #ifndef __XEN_PUBLIC_IO_RING_H__
+ #define __XEN_PUBLIC_IO_RING_H__
+
++#include "../xen-compat.h"
++
++#if __XEN_INTERFACE_VERSION__ < 0x00030208
++#define xen_mb() mb()
++#define xen_rmb() rmb()
++#define xen_wmb() wmb()
++#endif
++
+ typedef unsigned int RING_IDX;
+
+ /* Round a 32-bit unsigned constant down to the nearest power of two. */
+-#define __RD2(_x) (((_x) & 0x00000002) ? 0x2 : ((_x) & 0x1))
++#define __RD2(_x) (((_x) & 0x00000002) ? 0x2 : ((_x) & 0x1))
+ #define __RD4(_x) (((_x) & 0x0000000c) ? __RD2((_x)>>2)<<2 : __RD2(_x))
+ #define __RD8(_x) (((_x) & 0x000000f0) ? __RD4((_x)>>4)<<4 : __RD4(_x))
+ #define __RD16(_x) (((_x) & 0x0000ff00) ? __RD8((_x)>>8)<<8 : __RD8(_x))
+@@ -25,73 +51,76 @@ typedef unsigned int RING_IDX;
+ * power of two (so we can mask with (size-1) to loop around).
+ */
+ #define __RING_SIZE(_s, _sz) \
+- (__RD32(((_sz) - (long)&(_s)->ring + (long)(_s)) / sizeof((_s)->ring[0])))
++ (__RD32(((_sz) - (long)(_s)->ring + (long)(_s)) / sizeof((_s)->ring[0])))
+
+ /*
+ * Macros to make the correct C datatypes for a new kind of ring.
+ *
+ * To make a new ring datatype, you need to have two message structures,
+- * let's say struct request, and struct response already defined.
++ * let's say request_t, and response_t already defined.
+ *
+ * In a header where you want the ring datatype declared, you then do:
+ *
+- * DEFINE_RING_TYPES(mytag, struct request, struct response);
++ * DEFINE_RING_TYPES(mytag, request_t, response_t);
+ *
+ * These expand out to give you a set of types, as you can see below.
+ * The most important of these are:
+ *
+- * struct mytag_sring - The shared ring.
+- * struct mytag_front_ring - The 'front' half of the ring.
+- * struct mytag_back_ring - The 'back' half of the ring.
++ * mytag_sring_t - The shared ring.
++ * mytag_front_ring_t - The 'front' half of the ring.
++ * mytag_back_ring_t - The 'back' half of the ring.
+ *
+ * To initialize a ring in your code you need to know the location and size
+ * of the shared memory area (PAGE_SIZE, for instance). To initialise
+ * the front half:
+ *
+- * struct mytag_front_ring front_ring;
+- * SHARED_RING_INIT((struct mytag_sring *)shared_page);
+- * FRONT_RING_INIT(&front_ring, (struct mytag_sring *)shared_page,
+- * PAGE_SIZE);
++ * mytag_front_ring_t front_ring;
++ * SHARED_RING_INIT((mytag_sring_t *)shared_page);
++ * FRONT_RING_INIT(&front_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
+ *
+ * Initializing the back follows similarly (note that only the front
+ * initializes the shared ring):
+ *
+- * struct mytag_back_ring back_ring;
+- * BACK_RING_INIT(&back_ring, (struct mytag_sring *)shared_page,
+- * PAGE_SIZE);
++ * mytag_back_ring_t back_ring;
++ * BACK_RING_INIT(&back_ring, (mytag_sring_t *)shared_page, PAGE_SIZE);
+ */
+
+-#define DEFINE_RING_TYPES(__name, __req_t, __rsp_t) \
+- \
+-/* Shared ring entry */ \
+-union __name##_sring_entry { \
+- __req_t req; \
+- __rsp_t rsp; \
+-}; \
+- \
+-/* Shared ring page */ \
+-struct __name##_sring { \
+- RING_IDX req_prod, req_event; \
+- RING_IDX rsp_prod, rsp_event; \
+- uint8_t pad[48]; \
+- union __name##_sring_entry ring[1]; /* variable-length */ \
+-}; \
+- \
+-/* "Front" end's private variables */ \
+-struct __name##_front_ring { \
+- RING_IDX req_prod_pvt; \
+- RING_IDX rsp_cons; \
+- unsigned int nr_ents; \
+- struct __name##_sring *sring; \
+-}; \
+- \
+-/* "Back" end's private variables */ \
+-struct __name##_back_ring { \
+- RING_IDX rsp_prod_pvt; \
+- RING_IDX req_cons; \
+- unsigned int nr_ents; \
+- struct __name##_sring *sring; \
+-};
++#define DEFINE_RING_TYPES(__name, __req_t, __rsp_t) \
++ \
++/* Shared ring entry */ \
++union __name##_sring_entry { \
++ __req_t req; \
++ __rsp_t rsp; \
++}; \
++ \
++/* Shared ring page */ \
++struct __name##_sring { \
++ RING_IDX req_prod, req_event; \
++ RING_IDX rsp_prod, rsp_event; \
++ uint8_t pad[48]; \
++ union __name##_sring_entry ring[1]; /* variable-length */ \
++}; \
++ \
++/* "Front" end's private variables */ \
++struct __name##_front_ring { \
++ RING_IDX req_prod_pvt; \
++ RING_IDX rsp_cons; \
++ unsigned int nr_ents; \
++ struct __name##_sring *sring; \
++}; \
++ \
++/* "Back" end's private variables */ \
++struct __name##_back_ring { \
++ RING_IDX rsp_prod_pvt; \
++ RING_IDX req_cons; \
++ unsigned int nr_ents; \
++ struct __name##_sring *sring; \
++}; \
++ \
++/* Syntactic sugar */ \
++typedef struct __name##_sring __name##_sring_t; \
++typedef struct __name##_front_ring __name##_front_ring_t; \
++typedef struct __name##_back_ring __name##_back_ring_t
+
+ /*
+ * Macros for manipulating rings.
+@@ -109,86 +138,94 @@ struct __name##_back_ring { \
+ */
+
+ /* Initialising empty rings */
+-#define SHARED_RING_INIT(_s) do { \
+- (_s)->req_prod = (_s)->rsp_prod = 0; \
+- (_s)->req_event = (_s)->rsp_event = 1; \
+- memset((_s)->pad, 0, sizeof((_s)->pad)); \
++#define SHARED_RING_INIT(_s) do { \
++ (_s)->req_prod = (_s)->rsp_prod = 0; \
++ (_s)->req_event = (_s)->rsp_event = 1; \
++ (void)memset((_s)->pad, 0, sizeof((_s)->pad)); \
+ } while(0)
+
+-#define FRONT_RING_INIT(_r, _s, __size) do { \
+- (_r)->req_prod_pvt = 0; \
+- (_r)->rsp_cons = 0; \
+- (_r)->nr_ents = __RING_SIZE(_s, __size); \
+- (_r)->sring = (_s); \
++#define FRONT_RING_INIT(_r, _s, __size) do { \
++ (_r)->req_prod_pvt = 0; \
++ (_r)->rsp_cons = 0; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++ (_r)->sring = (_s); \
+ } while (0)
+
+-#define BACK_RING_INIT(_r, _s, __size) do { \
+- (_r)->rsp_prod_pvt = 0; \
+- (_r)->req_cons = 0; \
+- (_r)->nr_ents = __RING_SIZE(_s, __size); \
+- (_r)->sring = (_s); \
++#define BACK_RING_INIT(_r, _s, __size) do { \
++ (_r)->rsp_prod_pvt = 0; \
++ (_r)->req_cons = 0; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
++ (_r)->sring = (_s); \
+ } while (0)
+
+ /* Initialize to existing shared indexes -- for recovery */
+-#define FRONT_RING_ATTACH(_r, _s, __size) do { \
+- (_r)->sring = (_s); \
+- (_r)->req_prod_pvt = (_s)->req_prod; \
+- (_r)->rsp_cons = (_s)->rsp_prod; \
+- (_r)->nr_ents = __RING_SIZE(_s, __size); \
++#define FRONT_RING_ATTACH(_r, _s, __size) do { \
++ (_r)->sring = (_s); \
++ (_r)->req_prod_pvt = (_s)->req_prod; \
++ (_r)->rsp_cons = (_s)->rsp_prod; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
+ } while (0)
+
+-#define BACK_RING_ATTACH(_r, _s, __size) do { \
+- (_r)->sring = (_s); \
+- (_r)->rsp_prod_pvt = (_s)->rsp_prod; \
+- (_r)->req_cons = (_s)->req_prod; \
+- (_r)->nr_ents = __RING_SIZE(_s, __size); \
++#define BACK_RING_ATTACH(_r, _s, __size) do { \
++ (_r)->sring = (_s); \
++ (_r)->rsp_prod_pvt = (_s)->rsp_prod; \
++ (_r)->req_cons = (_s)->req_prod; \
++ (_r)->nr_ents = __RING_SIZE(_s, __size); \
+ } while (0)
+
+ /* How big is this ring? */
+-#define RING_SIZE(_r) \
++#define RING_SIZE(_r) \
+ ((_r)->nr_ents)
+
+ /* Number of free requests (for use on front side only). */
+-#define RING_FREE_REQUESTS(_r) \
++#define RING_FREE_REQUESTS(_r) \
+ (RING_SIZE(_r) - ((_r)->req_prod_pvt - (_r)->rsp_cons))
+
+ /* Test if there is an empty slot available on the front ring.
+ * (This is only meaningful from the front. )
+ */
+-#define RING_FULL(_r) \
++#define RING_FULL(_r) \
+ (RING_FREE_REQUESTS(_r) == 0)
+
+ /* Test if there are outstanding messages to be processed on a ring. */
+-#define RING_HAS_UNCONSUMED_RESPONSES(_r) \
++#define RING_HAS_UNCONSUMED_RESPONSES(_r) \
+ ((_r)->sring->rsp_prod - (_r)->rsp_cons)
+
+-#define RING_HAS_UNCONSUMED_REQUESTS(_r) \
+- ({ \
+- unsigned int req = (_r)->sring->req_prod - (_r)->req_cons; \
+- unsigned int rsp = RING_SIZE(_r) - \
+- ((_r)->req_cons - (_r)->rsp_prod_pvt); \
+- req < rsp ? req : rsp; \
+- })
++#ifdef __GNUC__
++#define RING_HAS_UNCONSUMED_REQUESTS(_r) ({ \
++ unsigned int req = (_r)->sring->req_prod - (_r)->req_cons; \
++ unsigned int rsp = RING_SIZE(_r) - \
++ ((_r)->req_cons - (_r)->rsp_prod_pvt); \
++ req < rsp ? req : rsp; \
++})
++#else
++/* Same as above, but without the nice GCC ({ ... }) syntax. */
++#define RING_HAS_UNCONSUMED_REQUESTS(_r) \
++ ((((_r)->sring->req_prod - (_r)->req_cons) < \
++ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt))) ? \
++ ((_r)->sring->req_prod - (_r)->req_cons) : \
++ (RING_SIZE(_r) - ((_r)->req_cons - (_r)->rsp_prod_pvt)))
++#endif
+
+ /* Direct access to individual ring elements, by index. */
+-#define RING_GET_REQUEST(_r, _idx) \
++#define RING_GET_REQUEST(_r, _idx) \
+ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].req))
+
+-#define RING_GET_RESPONSE(_r, _idx) \
++#define RING_GET_RESPONSE(_r, _idx) \
+ (&((_r)->sring->ring[((_idx) & (RING_SIZE(_r) - 1))].rsp))
+
+ /* Loop termination condition: Would the specified index overflow the ring? */
+-#define RING_REQUEST_CONS_OVERFLOW(_r, _cons) \
++#define RING_REQUEST_CONS_OVERFLOW(_r, _cons) \
+ (((_cons) - (_r)->rsp_prod_pvt) >= RING_SIZE(_r))
+
+-#define RING_PUSH_REQUESTS(_r) do { \
+- wmb(); /* back sees requests /before/ updated producer index */ \
+- (_r)->sring->req_prod = (_r)->req_prod_pvt; \
++#define RING_PUSH_REQUESTS(_r) do { \
++ xen_wmb(); /* back sees requests /before/ updated producer index */ \
++ (_r)->sring->req_prod = (_r)->req_prod_pvt; \
+ } while (0)
+
+-#define RING_PUSH_RESPONSES(_r) do { \
+- wmb(); /* front sees responses /before/ updated producer index */ \
+- (_r)->sring->rsp_prod = (_r)->rsp_prod_pvt; \
++#define RING_PUSH_RESPONSES(_r) do { \
++ xen_wmb(); /* front sees resps /before/ updated producer index */ \
++ (_r)->sring->rsp_prod = (_r)->rsp_prod_pvt; \
+ } while (0)
+
+ /*
+@@ -221,40 +258,50 @@ struct __name##_back_ring { \
+ * field appropriately.
+ */
+
+-#define RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(_r, _notify) do { \
+- RING_IDX __old = (_r)->sring->req_prod; \
+- RING_IDX __new = (_r)->req_prod_pvt; \
+- wmb(); /* back sees requests /before/ updated producer index */ \
+- (_r)->sring->req_prod = __new; \
+- mb(); /* back sees new requests /before/ we check req_event */ \
+- (_notify) = ((RING_IDX)(__new - (_r)->sring->req_event) < \
+- (RING_IDX)(__new - __old)); \
++#define RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(_r, _notify) do { \
++ RING_IDX __old = (_r)->sring->req_prod; \
++ RING_IDX __new = (_r)->req_prod_pvt; \
++ xen_wmb(); /* back sees requests /before/ updated producer index */ \
++ (_r)->sring->req_prod = __new; \
++ xen_mb(); /* back sees new requests /before/ we check req_event */ \
++ (_notify) = ((RING_IDX)(__new - (_r)->sring->req_event) < \
++ (RING_IDX)(__new - __old)); \
+ } while (0)
+
+-#define RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(_r, _notify) do { \
+- RING_IDX __old = (_r)->sring->rsp_prod; \
+- RING_IDX __new = (_r)->rsp_prod_pvt; \
+- wmb(); /* front sees responses /before/ updated producer index */ \
+- (_r)->sring->rsp_prod = __new; \
+- mb(); /* front sees new responses /before/ we check rsp_event */ \
+- (_notify) = ((RING_IDX)(__new - (_r)->sring->rsp_event) < \
+- (RING_IDX)(__new - __old)); \
++#define RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(_r, _notify) do { \
++ RING_IDX __old = (_r)->sring->rsp_prod; \
++ RING_IDX __new = (_r)->rsp_prod_pvt; \
++ xen_wmb(); /* front sees resps /before/ updated producer index */ \
++ (_r)->sring->rsp_prod = __new; \
++ xen_mb(); /* front sees new resps /before/ we check rsp_event */ \
++ (_notify) = ((RING_IDX)(__new - (_r)->sring->rsp_event) < \
++ (RING_IDX)(__new - __old)); \
+ } while (0)
+
+-#define RING_FINAL_CHECK_FOR_REQUESTS(_r, _work_to_do) do { \
+- (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
+- if (_work_to_do) break; \
+- (_r)->sring->req_event = (_r)->req_cons + 1; \
+- mb(); \
+- (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
++#define RING_FINAL_CHECK_FOR_REQUESTS(_r, _work_to_do) do { \
++ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
++ if (_work_to_do) break; \
++ (_r)->sring->req_event = (_r)->req_cons + 1; \
++ xen_mb(); \
++ (_work_to_do) = RING_HAS_UNCONSUMED_REQUESTS(_r); \
+ } while (0)
+
+-#define RING_FINAL_CHECK_FOR_RESPONSES(_r, _work_to_do) do { \
+- (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
+- if (_work_to_do) break; \
+- (_r)->sring->rsp_event = (_r)->rsp_cons + 1; \
+- mb(); \
+- (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
++#define RING_FINAL_CHECK_FOR_RESPONSES(_r, _work_to_do) do { \
++ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
++ if (_work_to_do) break; \
++ (_r)->sring->rsp_event = (_r)->rsp_cons + 1; \
++ xen_mb(); \
++ (_work_to_do) = RING_HAS_UNCONSUMED_RESPONSES(_r); \
+ } while (0)
+
+ #endif /* __XEN_PUBLIC_IO_RING_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/tpmif.h b/include/xen/interface/io/tpmif.h
+new file mode 100644
+index 0000000..02ccdab
+--- /dev/null
++++ b/include/xen/interface/io/tpmif.h
+@@ -0,0 +1,77 @@
++/******************************************************************************
++ * tpmif.h
++ *
++ * TPM I/O interface for Xen guest OSes.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2005, IBM Corporation
++ *
++ * Author: Stefan Berger, stefanb at us.ibm.com
++ * Grant table support: Mahadevan Gomathisankaran
++ *
++ * This code has been derived from tools/libxc/xen/io/netif.h
++ *
++ * Copyright (c) 2003-2004, Keir Fraser
++ */
++
++#ifndef __XEN_PUBLIC_IO_TPMIF_H__
++#define __XEN_PUBLIC_IO_TPMIF_H__
++
++#include "../grant_table.h"
++
++struct tpmif_tx_request {
++ unsigned long addr; /* Machine address of packet. */
++ grant_ref_t ref; /* grant table access reference */
++ uint16_t unused;
++ uint16_t size; /* Packet size in bytes. */
++};
++typedef struct tpmif_tx_request tpmif_tx_request_t;
++
++/*
++ * The TPMIF_TX_RING_SIZE defines the number of pages the
++ * front-end and backend can exchange (= size of array).
++ */
++typedef uint32_t TPMIF_RING_IDX;
++
++#define TPMIF_TX_RING_SIZE 1
++
++/* This structure must fit in a memory page. */
++
++struct tpmif_ring {
++ struct tpmif_tx_request req;
++};
++typedef struct tpmif_ring tpmif_ring_t;
++
++struct tpmif_tx_interface {
++ struct tpmif_ring ring[TPMIF_TX_RING_SIZE];
++};
++typedef struct tpmif_tx_interface tpmif_tx_interface_t;
++
++#endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/vscsiif.h b/include/xen/interface/io/vscsiif.h
+new file mode 100644
+index 0000000..3ce2914
+--- /dev/null
++++ b/include/xen/interface/io/vscsiif.h
+@@ -0,0 +1,105 @@
++/******************************************************************************
++ * vscsiif.h
++ *
++ * Based on the blkif.h code.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright(c) FUJITSU Limited 2008.
++ */
++
++#ifndef __XEN__PUBLIC_IO_SCSI_H__
++#define __XEN__PUBLIC_IO_SCSI_H__
++
++#include "ring.h"
++#include "../grant_table.h"
++
++/* command between backend and frontend */
++#define VSCSIIF_ACT_SCSI_CDB 1 /* SCSI CDB command */
++#define VSCSIIF_ACT_SCSI_ABORT 2 /* SCSI Device(Lun) Abort*/
++#define VSCSIIF_ACT_SCSI_RESET 3 /* SCSI Device(Lun) Reset*/
++
++
++#define VSCSIIF_BACK_MAX_PENDING_REQS 128
++
++/*
++ * Maximum scatter/gather segments per request.
++ *
++ * Considering balance between allocating al least 16 "vscsiif_request"
++ * structures on one page (4096bytes) and number of scatter gather
++ * needed, we decided to use 26 as a magic number.
++ */
++#define VSCSIIF_SG_TABLESIZE 26
++
++/*
++ * base on linux kernel 2.6.18
++ */
++#define VSCSIIF_MAX_COMMAND_SIZE 16
++#define VSCSIIF_SENSE_BUFFERSIZE 96
++
++
++struct vscsiif_request {
++ uint16_t rqid; /* private guest value, echoed in resp */
++ uint8_t act; /* command between backend and frontend */
++ uint8_t cmd_len;
++
++ uint8_t cmnd[VSCSIIF_MAX_COMMAND_SIZE];
++ uint16_t timeout_per_command; /* The command is issued by twice
++ the value in Backend. */
++ uint16_t channel, id, lun;
++ uint16_t padding;
++ uint8_t sc_data_direction; /* for DMA_TO_DEVICE(1)
++ DMA_FROM_DEVICE(2)
++ DMA_NONE(3) requests */
++ uint8_t nr_segments; /* Number of pieces of scatter-gather */
++
++ struct scsiif_request_segment {
++ grant_ref_t gref;
++ uint16_t offset;
++ uint16_t length;
++ } seg[VSCSIIF_SG_TABLESIZE];
++ uint32_t reserved[3];
++};
++typedef struct vscsiif_request vscsiif_request_t;
++
++struct vscsiif_response {
++ uint16_t rqid;
++ uint8_t padding;
++ uint8_t sense_len;
++ uint8_t sense_buffer[VSCSIIF_SENSE_BUFFERSIZE];
++ int32_t rslt;
++ uint32_t residual_len; /* request bufflen -
++ return the value from physical device */
++ uint32_t reserved[36];
++};
++typedef struct vscsiif_response vscsiif_response_t;
++
++DEFINE_RING_TYPES(vscsiif, struct vscsiif_request, struct vscsiif_response);
++
++
++#endif /*__XEN__PUBLIC_IO_SCSI_H__*/
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/io/xenbus.h b/include/xen/interface/io/xenbus.h
+index 46508c7..4a053df 100644
+--- a/include/xen/interface/io/xenbus.h
++++ b/include/xen/interface/io/xenbus.h
+@@ -3,42 +3,78 @@
+ *
+ * Xenbus protocol details.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (C) 2005 XenSource Ltd.
+ */
+
+ #ifndef _XEN_PUBLIC_IO_XENBUS_H
+ #define _XEN_PUBLIC_IO_XENBUS_H
+
+-/* The state of either end of the Xenbus, i.e. the current communication
+- status of initialisation across the bus. States here imply nothing about
+- the state of the connection between the driver and the kernel's device
+- layers. */
+-enum xenbus_state
+-{
+- XenbusStateUnknown = 0,
+- XenbusStateInitialising = 1,
+- XenbusStateInitWait = 2, /* Finished early
+- initialisation, but waiting
+- for information from the peer
+- or hotplug scripts. */
+- XenbusStateInitialised = 3, /* Initialised and waiting for a
+- connection from the peer. */
+- XenbusStateConnected = 4,
+- XenbusStateClosing = 5, /* The device is being closed
+- due to an error or an unplug
+- event. */
+- XenbusStateClosed = 6
++/*
++ * The state of either end of the Xenbus, i.e. the current communication
++ * status of initialisation across the bus. States here imply nothing about
++ * the state of the connection between the driver and the kernel's device
++ * layers.
++ */
++enum xenbus_state {
++ XenbusStateUnknown = 0,
++
++ XenbusStateInitialising = 1,
++
++ /*
++ * InitWait: Finished early initialisation but waiting for information
++ * from the peer or hotplug scripts.
++ */
++ XenbusStateInitWait = 2,
++
++ /*
++ * Initialised: Waiting for a connection from the peer.
++ */
++ XenbusStateInitialised = 3,
++
++ XenbusStateConnected = 4,
++
++ /*
++ * Closing: The device is being closed due to an error or an unplug event.
++ */
++ XenbusStateClosing = 5,
++
++ XenbusStateClosed = 6,
++
++ /*
++ * Reconfiguring: The device is being reconfigured.
++ */
++ XenbusStateReconfiguring = 7,
+
++ XenbusStateReconfigured = 8
+ };
++typedef enum xenbus_state XenbusState;
+
+ #endif /* _XEN_PUBLIC_IO_XENBUS_H */
+
+ /*
+ * Local variables:
+- * c-file-style: "linux"
+- * indent-tabs-mode: t
+- * c-indent-level: 8
+- * c-basic-offset: 8
+- * tab-width: 8
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
+ * End:
+ */
+diff --git a/include/xen/interface/io/xs_wire.h b/include/xen/interface/io/xs_wire.h
+index 99fcffb..f6a49ab 100644
+--- a/include/xen/interface/io/xs_wire.h
++++ b/include/xen/interface/io/xs_wire.h
+@@ -1,6 +1,25 @@
+ /*
+ * Details of the "wire" protocol between Xen Store Daemon and client
+ * library or guest kernel.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (C) 2005 Rusty Russell IBM Corporation
+ */
+
+@@ -26,7 +45,9 @@ enum xsd_sockmsg_type
+ XS_SET_PERMS,
+ XS_WATCH_EVENT,
+ XS_ERROR,
+- XS_IS_DOMAIN_INTRODUCED
++ XS_IS_DOMAIN_INTRODUCED,
++ XS_RESUME,
++ XS_SET_TARGET
+ };
+
+ #define XS_WRITE_NONE "NONE"
+@@ -40,7 +61,12 @@ struct xsd_errors
+ const char *errstring;
+ };
+ #define XSD_ERROR(x) { x, #x }
+-static struct xsd_errors xsd_errors[] __attribute__((unused)) = {
++/* LINTED: static unused */
++static struct xsd_errors xsd_errors[]
++#if defined(__GNUC__)
++__attribute__((unused))
++#endif
++ = {
+ XSD_ERROR(EINVAL),
+ XSD_ERROR(EACCES),
+ XSD_ERROR(EEXIST),
+@@ -84,4 +110,21 @@ struct xenstore_domain_interface {
+ XENSTORE_RING_IDX rsp_cons, rsp_prod;
+ };
+
++/* Violating this is very bad. See docs/misc/xenstore.txt. */
++#define XENSTORE_PAYLOAD_MAX 4096
++
++/* Violating these just gets you an error back */
++#define XENSTORE_ABS_PATH_MAX 3072
++#define XENSTORE_REL_PATH_MAX 2048
++
+ #endif /* _XS_WIRE_H */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/kexec.h b/include/xen/interface/kexec.h
+new file mode 100644
+index 0000000..101fd40
+--- /dev/null
++++ b/include/xen/interface/kexec.h
+@@ -0,0 +1,171 @@
++/******************************************************************************
++ * kexec.h - Public portion
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Xen port written by:
++ * - Simon 'Horms' Horman <horms at verge.net.au>
++ * - Magnus Damm <magnus at valinux.co.jp>
++ */
++
++#ifndef _XEN_PUBLIC_KEXEC_H
++#define _XEN_PUBLIC_KEXEC_H
++
++
++/* This file describes the Kexec / Kdump hypercall interface for Xen.
++ *
++ * Kexec under vanilla Linux allows a user to reboot the physical machine
++ * into a new user-specified kernel. The Xen port extends this idea
++ * to allow rebooting of the machine from dom0. When kexec for dom0
++ * is used to reboot, both the hypervisor and the domains get replaced
++ * with some other kernel. It is possible to kexec between vanilla
++ * Linux and Xen and back again. Xen to Xen works well too.
++ *
++ * The hypercall interface for kexec can be divided into three main
++ * types of hypercall operations:
++ *
++ * 1) Range information:
++ * This is used by the dom0 kernel to ask the hypervisor about various
++ * address information. This information is needed to allow kexec-tools
++ * to fill in the ELF headers for /proc/vmcore properly.
++ *
++ * 2) Load and unload of images:
++ * There are no big surprises here, the kexec binary from kexec-tools
++ * runs in userspace in dom0. The tool loads/unloads data into the
++ * dom0 kernel such as new kernel, initramfs and hypervisor. When
++ * loaded the dom0 kernel performs a load hypercall operation, and
++ * before releasing all page references the dom0 kernel calls unload.
++ *
++ * 3) Kexec operation:
++ * This is used to start a previously loaded kernel.
++ */
++
++#include "xen.h"
++
++#if defined(__i386__) || defined(__x86_64__)
++#define KEXEC_XEN_NO_PAGES 17
++#endif
++
++/*
++ * Prototype for this hypercall is:
++ * int kexec_op(int cmd, void *args)
++ * @cmd == KEXEC_CMD_...
++ * KEXEC operation to perform
++ * @args == Operation-specific extra arguments (NULL if none).
++ */
++
++/*
++ * Kexec supports two types of operation:
++ * - kexec into a regular kernel, very similar to a standard reboot
++ * - KEXEC_TYPE_DEFAULT is used to specify this type
++ * - kexec into a special "crash kernel", aka kexec-on-panic
++ * - KEXEC_TYPE_CRASH is used to specify this type
++ * - parts of our system may be broken at kexec-on-panic time
++ * - the code should be kept as simple and self-contained as possible
++ */
++
++#define KEXEC_TYPE_DEFAULT 0
++#define KEXEC_TYPE_CRASH 1
++
++
++/* The kexec implementation for Xen allows the user to load two
++ * types of kernels, KEXEC_TYPE_DEFAULT and KEXEC_TYPE_CRASH.
++ * All data needed for a kexec reboot is kept in one xen_kexec_image_t
++ * per "instance". The data mainly consists of machine address lists to pages
++ * together with destination addresses. The data in xen_kexec_image_t
++ * is passed to the "code page" which is one page of code that performs
++ * the final relocations before jumping to the new kernel.
++ */
++
++typedef struct xen_kexec_image {
++#if defined(__i386__) || defined(__x86_64__)
++ unsigned long page_list[KEXEC_XEN_NO_PAGES];
++#endif
++#if defined(__ia64__)
++ unsigned long reboot_code_buffer;
++#endif
++ unsigned long indirection_page;
++ unsigned long start_address;
++} xen_kexec_image_t;
++
++/*
++ * Perform kexec having previously loaded a kexec or kdump kernel
++ * as appropriate.
++ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
++ */
++#define KEXEC_CMD_kexec 0
++typedef struct xen_kexec_exec {
++ int type;
++} xen_kexec_exec_t;
++
++/*
++ * Load/Unload kernel image for kexec or kdump.
++ * type == KEXEC_TYPE_DEFAULT or KEXEC_TYPE_CRASH [in]
++ * image == relocation information for kexec (ignored for unload) [in]
++ */
++#define KEXEC_CMD_kexec_load 1
++#define KEXEC_CMD_kexec_unload 2
++typedef struct xen_kexec_load {
++ int type;
++ xen_kexec_image_t image;
++} xen_kexec_load_t;
++
++#define KEXEC_RANGE_MA_CRASH 0 /* machine address and size of crash area */
++#define KEXEC_RANGE_MA_XEN 1 /* machine address and size of Xen itself */
++#define KEXEC_RANGE_MA_CPU 2 /* machine address and size of a CPU note */
++#define KEXEC_RANGE_MA_XENHEAP 3 /* machine address and size of xenheap
++ * Note that although this is adjacent
++ * to Xen it exists in a separate EFI
++ * region on ia64, and thus needs to be
++ * inserted into iomem_machine separately */
++#define KEXEC_RANGE_MA_BOOT_PARAM 4 /* machine address and size of
++ * the ia64_boot_param */
++#define KEXEC_RANGE_MA_EFI_MEMMAP 5 /* machine address and size of
++ * of the EFI Memory Map */
++#define KEXEC_RANGE_MA_VMCOREINFO 6 /* machine address and size of vmcoreinfo */
++
++extern size_t vmcoreinfo_size_xen;
++extern unsigned long paddr_vmcoreinfo_xen;
++
++/*
++ * Find the address and size of certain memory areas
++ * range == KEXEC_RANGE_... [in]
++ * nr == physical CPU number (starting from 0) if KEXEC_RANGE_MA_CPU [in]
++ * size == number of bytes reserved in window [out]
++ * start == address of the first byte in the window [out]
++ */
++#define KEXEC_CMD_kexec_get_range 3
++typedef struct xen_kexec_range {
++ int range;
++ int nr;
++ unsigned long size;
++ unsigned long start;
++} xen_kexec_range_t;
++
++#endif /* _XEN_PUBLIC_KEXEC_H */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/libelf.h b/include/xen/interface/libelf.h
+new file mode 100644
+index 0000000..d238330
+--- /dev/null
++++ b/include/xen/interface/libelf.h
+@@ -0,0 +1,265 @@
++/******************************************************************************
++ * libelf.h
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ */
++
++#ifndef __XC_LIBELF__
++#define __XC_LIBELF__ 1
++
++#if defined(__i386__) || defined(__x86_64__) || defined(__ia64__)
++#define XEN_ELF_LITTLE_ENDIAN
++#else
++#error define architectural endianness
++#endif
++
++#undef ELFSIZE
++#include "elfnote.h"
++#include "elfstructs.h"
++#include "features.h"
++
++/* ------------------------------------------------------------------------ */
++
++typedef union {
++ Elf32_Ehdr e32;
++ Elf64_Ehdr e64;
++} elf_ehdr;
++
++typedef union {
++ Elf32_Phdr e32;
++ Elf64_Phdr e64;
++} elf_phdr;
++
++typedef union {
++ Elf32_Shdr e32;
++ Elf64_Shdr e64;
++} elf_shdr;
++
++typedef union {
++ Elf32_Sym e32;
++ Elf64_Sym e64;
++} elf_sym;
++
++typedef union {
++ Elf32_Rel e32;
++ Elf64_Rel e64;
++} elf_rel;
++
++typedef union {
++ Elf32_Rela e32;
++ Elf64_Rela e64;
++} elf_rela;
++
++typedef union {
++ Elf32_Note e32;
++ Elf64_Note e64;
++} elf_note;
++
++struct elf_binary {
++ /* elf binary */
++ const char *image;
++ size_t size;
++ char class;
++ char data;
++
++ const elf_ehdr *ehdr;
++ const char *sec_strtab;
++ const elf_shdr *sym_tab;
++ const char *sym_strtab;
++
++ /* loaded to */
++ char *dest;
++ uint64_t pstart;
++ uint64_t pend;
++ uint64_t reloc_offset;
++
++ uint64_t bsd_symtab_pstart;
++ uint64_t bsd_symtab_pend;
++
++#ifndef __XEN__
++ /* misc */
++ FILE *log;
++#endif
++ int verbose;
++};
++
++/* ------------------------------------------------------------------------ */
++/* accessing elf header fields */
++
++#ifdef XEN_ELF_BIG_ENDIAN
++# define NATIVE_ELFDATA ELFDATA2MSB
++#else
++# define NATIVE_ELFDATA ELFDATA2LSB
++#endif
++
++#define elf_32bit(elf) (ELFCLASS32 == (elf)->class)
++#define elf_64bit(elf) (ELFCLASS64 == (elf)->class)
++#define elf_msb(elf) (ELFDATA2MSB == (elf)->data)
++#define elf_lsb(elf) (ELFDATA2LSB == (elf)->data)
++#define elf_swap(elf) (NATIVE_ELFDATA != (elf)->data)
++
++#define elf_uval(elf, str, elem) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? elf_access_unsigned((elf), (str), \
++ offsetof(typeof(*(str)),e64.elem), \
++ sizeof((str)->e64.elem)) \
++ : elf_access_unsigned((elf), (str), \
++ offsetof(typeof(*(str)),e32.elem), \
++ sizeof((str)->e32.elem)))
++
++#define elf_sval(elf, str, elem) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? elf_access_signed((elf), (str), \
++ offsetof(typeof(*(str)),e64.elem), \
++ sizeof((str)->e64.elem)) \
++ : elf_access_signed((elf), (str), \
++ offsetof(typeof(*(str)),e32.elem), \
++ sizeof((str)->e32.elem)))
++
++#define elf_size(elf, str) \
++ ((ELFCLASS64 == (elf)->class) \
++ ? sizeof((str)->e64) : sizeof((str)->e32))
++
++uint64_t elf_access_unsigned(struct elf_binary *elf, const void *ptr,
++ uint64_t offset, size_t size);
++int64_t elf_access_signed(struct elf_binary *elf, const void *ptr,
++ uint64_t offset, size_t size);
++
++uint64_t elf_round_up(struct elf_binary *elf, uint64_t addr);
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_tools.c */
++
++int elf_shdr_count(struct elf_binary *elf);
++int elf_phdr_count(struct elf_binary *elf);
++
++const elf_shdr *elf_shdr_by_name(struct elf_binary *elf, const char *name);
++const elf_shdr *elf_shdr_by_index(struct elf_binary *elf, int index);
++const elf_phdr *elf_phdr_by_index(struct elf_binary *elf, int index);
++
++const char *elf_section_name(struct elf_binary *elf, const elf_shdr * shdr);
++const void *elf_section_start(struct elf_binary *elf, const elf_shdr * shdr);
++const void *elf_section_end(struct elf_binary *elf, const elf_shdr * shdr);
++
++const void *elf_segment_start(struct elf_binary *elf, const elf_phdr * phdr);
++const void *elf_segment_end(struct elf_binary *elf, const elf_phdr * phdr);
++
++const elf_sym *elf_sym_by_name(struct elf_binary *elf, const char *symbol);
++const elf_sym *elf_sym_by_index(struct elf_binary *elf, int index);
++
++const char *elf_note_name(struct elf_binary *elf, const elf_note * note);
++const void *elf_note_desc(struct elf_binary *elf, const elf_note * note);
++uint64_t elf_note_numeric(struct elf_binary *elf, const elf_note * note);
++const elf_note *elf_note_next(struct elf_binary *elf, const elf_note * note);
++
++int elf_is_elfbinary(const void *image);
++int elf_phdr_is_loadable(struct elf_binary *elf, const elf_phdr * phdr);
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_loader.c */
++
++int elf_init(struct elf_binary *elf, const char *image, size_t size);
++#ifdef __XEN__
++void elf_set_verbose(struct elf_binary *elf);
++#else
++void elf_set_logfile(struct elf_binary *elf, FILE * log, int verbose);
++#endif
++
++void elf_parse_binary(struct elf_binary *elf);
++void elf_load_binary(struct elf_binary *elf);
++
++void *elf_get_ptr(struct elf_binary *elf, unsigned long addr);
++uint64_t elf_lookup_addr(struct elf_binary *elf, const char *symbol);
++
++void elf_parse_bsdsyms(struct elf_binary *elf, uint64_t pstart); /* private */
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_relocate.c */
++
++int elf_reloc(struct elf_binary *elf);
++
++/* ------------------------------------------------------------------------ */
++/* xc_libelf_dominfo.c */
++
++#define UNSET_ADDR ((uint64_t)-1)
++
++enum xen_elfnote_type {
++ XEN_ENT_NONE = 0,
++ XEN_ENT_LONG = 1,
++ XEN_ENT_STR = 2
++};
++
++struct xen_elfnote {
++ enum xen_elfnote_type type;
++ const char *name;
++ union {
++ const char *str;
++ uint64_t num;
++ } data;
++};
++
++struct elf_dom_parms {
++ /* raw */
++ const char *guest_info;
++ const void *elf_note_start;
++ const void *elf_note_end;
++ struct xen_elfnote elf_notes[XEN_ELFNOTE_MAX + 1];
++
++ /* parsed */
++ char guest_os[16];
++ char guest_ver[16];
++ char xen_ver[16];
++ char loader[16];
++ int pae;
++ int bsd_symtab;
++ uint64_t virt_base;
++ uint64_t virt_entry;
++ uint64_t virt_hypercall;
++ uint64_t virt_hv_start_low;
++ uint64_t elf_paddr_offset;
++ uint32_t f_supported[XENFEAT_NR_SUBMAPS];
++ uint32_t f_required[XENFEAT_NR_SUBMAPS];
++
++ /* calculated */
++ uint64_t virt_offset;
++ uint64_t virt_kstart;
++ uint64_t virt_kend;
++};
++
++static inline void elf_xen_feature_set(int nr, uint32_t * addr)
++{
++ addr[nr >> 5] |= 1 << (nr & 31);
++}
++static inline int elf_xen_feature_get(int nr, uint32_t * addr)
++{
++ return !!(addr[nr >> 5] & (1 << (nr & 31)));
++}
++
++int elf_xen_parse_features(const char *features,
++ uint32_t *supported,
++ uint32_t *required);
++int elf_xen_parse_note(struct elf_binary *elf,
++ struct elf_dom_parms *parms,
++ const elf_note *note);
++int elf_xen_parse_guest_info(struct elf_binary *elf,
++ struct elf_dom_parms *parms);
++int elf_xen_parse(struct elf_binary *elf,
++ struct elf_dom_parms *parms);
++
++#endif /* __XC_LIBELF__ */
+diff --git a/include/xen/interface/memory.h b/include/xen/interface/memory.h
+index da76846..fe4f490 100644
+--- a/include/xen/interface/memory.h
++++ b/include/xen/interface/memory.h
+@@ -3,6 +3,24 @@
+ *
+ * Memory reservation and information.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
+@@ -10,8 +28,8 @@
+ #define __XEN_PUBLIC_MEMORY_H__
+
+ /*
+- * Increase or decrease the specified domain's memory reservation. Returns a
+- * -ve errcode on failure, or the # extents successfully allocated or freed.
++ * Increase or decrease the specified domain's memory reservation. Returns the
++ * number of extents successfully allocated or freed.
+ * arg == addr of struct xen_memory_reservation.
+ */
+ #define XENMEM_increase_reservation 0
+@@ -29,10 +47,14 @@ struct xen_memory_reservation {
+ * OUT: GMFN bases of extents that were allocated
+ * (NB. This command also updates the mach_to_phys translation table)
+ */
++#ifndef CONFIG_PARAVIRT_XEN
++ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
++#else
+ ulong extent_start;
++#endif
+
+ /* Number of extents, and size/alignment of each (2^extent_order pages). */
+- unsigned long nr_extents;
++ xen_ulong_t nr_extents;
+ unsigned int extent_order;
+
+ /*
+@@ -48,8 +70,52 @@ struct xen_memory_reservation {
+ * Unprivileged domains can specify only DOMID_SELF.
+ */
+ domid_t domid;
++};
++typedef struct xen_memory_reservation xen_memory_reservation_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_reservation_t);
+
++/*
++ * An atomic exchange of memory pages. If return code is zero then
++ * @out.extent_list provides GMFNs of the newly-allocated memory.
++ * Returns zero on complete success, otherwise a negative error code.
++ * On complete success then always @nr_exchanged == @in.nr_extents.
++ * On partial success @nr_exchanged indicates how much work was done.
++ */
++#define XENMEM_exchange 11
++struct xen_memory_exchange {
++ /*
++ * [IN] Details of memory extents to be exchanged (GMFN bases).
++ * Note that @in.address_bits is ignored and unused.
++ */
++ struct xen_memory_reservation in;
++
++ /*
++ * [IN/OUT] Details of new memory extents.
++ * We require that:
++ * 1. @in.domid == @out.domid
++ * 2. @in.nr_extents << @in.extent_order ==
++ * @out.nr_extents << @out.extent_order
++ * 3. @in.extent_start and @out.extent_start lists must not overlap
++ * 4. @out.extent_start lists GPFN bases to be populated
++ * 5. @out.extent_start is overwritten with allocated GMFN bases
++ */
++ struct xen_memory_reservation out;
++
++ /*
++ * [OUT] Number of input extents that were successfully exchanged:
++ * 1. The first @nr_exchanged input extents were successfully
++ * deallocated.
++ * 2. The corresponding first entries in the output extent list correctly
++ * indicate the GMFNs that were successfully exchanged.
++ * 3. All other input and output extents are untouched.
++ * 4. If not all input exents are exchanged then the return code of this
++ * command will be non-zero.
++ * 5. THIS FIELD MUST BE INITIALISED TO ZERO BY THE CALLER!
++ */
++ xen_ulong_t nr_exchanged;
+ };
++typedef struct xen_memory_exchange xen_memory_exchange_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t);
+
+ /*
+ * Returns the maximum machine frame number of mapped RAM in this system.
+@@ -67,6 +133,11 @@ struct xen_memory_reservation {
+ #define XENMEM_maximum_reservation 4
+
+ /*
++ * Returns the maximum GPFN in use by the guest, or -ve errcode on failure.
++ */
++#define XENMEM_maximum_gpfn 14
++
++/*
+ * Returns a list of MFN bases of 2MB extents comprising the machine_to_phys
+ * mapping table. Architectures which do not have a m2p table do not implement
+ * this command.
+@@ -85,7 +156,11 @@ struct xen_machphys_mfn_list {
+ * any large discontiguities in the machine address space, 2MB gaps in
+ * the machphys table will be represented by an MFN base of zero.
+ */
++#ifndef CONFIG_PARAVIRT_XEN
++ XEN_GUEST_HANDLE(xen_pfn_t) extent_start;
++#else
+ ulong extent_start;
++#endif
+
+ /*
+ * Number of extents written to the above array. This will be smaller
+@@ -93,6 +168,22 @@ struct xen_machphys_mfn_list {
+ */
+ unsigned int nr_extents;
+ };
++typedef struct xen_machphys_mfn_list xen_machphys_mfn_list_t;
++DEFINE_XEN_GUEST_HANDLE(xen_machphys_mfn_list_t);
++
++/*
++ * Returns the location in virtual address space of the machine_to_phys
++ * mapping table. Architectures which do not have a m2p table, or which do not
++ * map it by default into guest address space, do not implement this command.
++ * arg == addr of xen_machphys_mapping_t.
++ */
++#define XENMEM_machphys_mapping 12
++struct xen_machphys_mapping {
++ xen_ulong_t v_start, v_end; /* Start and end virtual addresses. */
++ xen_ulong_t max_mfn; /* Maximum MFN that can be looked up. */
++};
++typedef struct xen_machphys_mapping xen_machphys_mapping_t;
++DEFINE_XEN_GUEST_HANDLE(xen_machphys_mapping_t);
+
+ /*
+ * Sets the GPFN at which a particular page appears in the specified guest's
+@@ -110,11 +201,13 @@ struct xen_add_to_physmap {
+ unsigned int space;
+
+ /* Index into source mapping space. */
+- unsigned long idx;
++ xen_ulong_t idx;
+
+ /* GPFN where the source mapping page should appear. */
+- unsigned long gpfn;
++ xen_pfn_t gpfn;
+ };
++typedef struct xen_add_to_physmap xen_add_to_physmap_t;
++DEFINE_XEN_GUEST_HANDLE(xen_add_to_physmap_t);
+
+ /*
+ * Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
+@@ -126,16 +219,80 @@ struct xen_translate_gpfn_list {
+ domid_t domid;
+
+ /* Length of list. */
+- unsigned long nr_gpfns;
++ xen_ulong_t nr_gpfns;
+
+ /* List of GPFNs to translate. */
++#ifndef CONFIG_PARAVIRT_XEN
++ XEN_GUEST_HANDLE(xen_pfn_t) gpfn_list;
++#else
+ ulong gpfn_list;
++#endif
+
+ /*
+ * Output list to contain MFN translations. May be the same as the input
+ * list (in which case each input GPFN is overwritten with the output MFN).
+ */
++#ifndef CONFIG_PARAVIRT_XEN
++ XEN_GUEST_HANDLE(xen_pfn_t) mfn_list;
++#else
+ ulong mfn_list;
++#endif
+ };
++DEFINE_XEN_GUEST_HANDLE_STRUCT(xen_translate_gpfn_list);
++typedef struct xen_translate_gpfn_list xen_translate_gpfn_list_t;
++DEFINE_XEN_GUEST_HANDLE(xen_translate_gpfn_list_t);
++
++/*
++ * Returns the pseudo-physical memory map as it was when the domain
++ * was started (specified by XENMEM_set_memory_map).
++ * arg == addr of xen_memory_map_t.
++ */
++#define XENMEM_memory_map 9
++struct xen_memory_map {
++ /*
++ * On call the number of entries which can be stored in buffer. On
++ * return the number of entries which have been stored in
++ * buffer.
++ */
++ unsigned int nr_entries;
++
++ /*
++ * Entries in the buffer are in the same format as returned by the
++ * BIOS INT 0x15 EAX=0xE820 call.
++ */
++ XEN_GUEST_HANDLE(void) buffer;
++};
++typedef struct xen_memory_map xen_memory_map_t;
++DEFINE_XEN_GUEST_HANDLE(xen_memory_map_t);
++
++/*
++ * Returns the real physical memory map. Passes the same structure as
++ * XENMEM_memory_map.
++ * arg == addr of xen_memory_map_t.
++ */
++#define XENMEM_machine_memory_map 10
++
++/*
++ * Set the pseudo-physical memory map of a domain, as returned by
++ * XENMEM_memory_map.
++ * arg == addr of xen_foreign_memory_map_t.
++ */
++#define XENMEM_set_memory_map 13
++struct xen_foreign_memory_map {
++ domid_t domid;
++ struct xen_memory_map map;
++};
++typedef struct xen_foreign_memory_map xen_foreign_memory_map_t;
++DEFINE_XEN_GUEST_HANDLE(xen_foreign_memory_map_t);
+
+ #endif /* __XEN_PUBLIC_MEMORY_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/nmi.h b/include/xen/interface/nmi.h
+new file mode 100644
+index 0000000..b2b8401
+--- /dev/null
++++ b/include/xen/interface/nmi.h
+@@ -0,0 +1,78 @@
++/******************************************************************************
++ * nmi.h
++ *
++ * NMI callback registration and reason codes.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
++ */
++
++#ifndef __XEN_PUBLIC_NMI_H__
++#define __XEN_PUBLIC_NMI_H__
++
++/*
++ * NMI reason codes:
++ * Currently these are x86-specific, stored in arch_shared_info.nmi_reason.
++ */
++ /* I/O-check error reported via ISA port 0x61, bit 6. */
++#define _XEN_NMIREASON_io_error 0
++#define XEN_NMIREASON_io_error (1UL << _XEN_NMIREASON_io_error)
++ /* Parity error reported via ISA port 0x61, bit 7. */
++#define _XEN_NMIREASON_parity_error 1
++#define XEN_NMIREASON_parity_error (1UL << _XEN_NMIREASON_parity_error)
++ /* Unknown hardware-generated NMI. */
++#define _XEN_NMIREASON_unknown 2
++#define XEN_NMIREASON_unknown (1UL << _XEN_NMIREASON_unknown)
++
++/*
++ * long nmi_op(unsigned int cmd, void *arg)
++ * NB. All ops return zero on success, else a negative error code.
++ */
++
++/*
++ * Register NMI callback for this (calling) VCPU. Currently this only makes
++ * sense for domain 0, vcpu 0. All other callers will be returned EINVAL.
++ * arg == pointer to xennmi_callback structure.
++ */
++#define XENNMI_register_callback 0
++struct xennmi_callback {
++ unsigned long handler_address;
++ unsigned long pad;
++};
++typedef struct xennmi_callback xennmi_callback_t;
++DEFINE_XEN_GUEST_HANDLE(xennmi_callback_t);
++
++/*
++ * Deregister NMI callback for this (calling) VCPU.
++ * arg == NULL.
++ */
++#define XENNMI_unregister_callback 1
++
++#endif /* __XEN_PUBLIC_NMI_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/physdev.h b/include/xen/interface/physdev.h
+index cd69391..7bbc8ad 100644
+--- a/include/xen/interface/physdev.h
++++ b/include/xen/interface/physdev.h
+@@ -24,7 +24,7 @@
+ /*
+ * Prototype for this hypercall is:
+ * int physdev_op(int cmd, void *args)
+- * @cmd == PHYSDEVOP_??? (physdev operation).
++ * @cmd == PHYSDEVOP_??? (physdev operation).
+ * @args == Operation-specific extra arguments (NULL if none).
+ */
+
+@@ -32,114 +32,185 @@
+ * Notify end-of-interrupt (EOI) for the specified IRQ.
+ * @arg == pointer to physdev_eoi structure.
+ */
+-#define PHYSDEVOP_eoi 12
++#define PHYSDEVOP_eoi 12
+ struct physdev_eoi {
+- /* IN */
+- uint32_t irq;
++ /* IN */
++ uint32_t irq;
+ };
++typedef struct physdev_eoi physdev_eoi_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_eoi_t);
+
+ /*
+ * Query the status of an IRQ line.
+ * @arg == pointer to physdev_irq_status_query structure.
+ */
+-#define PHYSDEVOP_irq_status_query 5
++#define PHYSDEVOP_irq_status_query 5
+ struct physdev_irq_status_query {
+- /* IN */
+- uint32_t irq;
+- /* OUT */
+- uint32_t flags; /* XENIRQSTAT_* */
++ /* IN */
++ uint32_t irq;
++ /* OUT */
++ uint32_t flags; /* XENIRQSTAT_* */
+ };
++typedef struct physdev_irq_status_query physdev_irq_status_query_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_irq_status_query_t);
+
+ /* Need to call PHYSDEVOP_eoi when the IRQ has been serviced? */
+-#define _XENIRQSTAT_needs_eoi (0)
+-#define XENIRQSTAT_needs_eoi (1U<<_XENIRQSTAT_needs_eoi)
++#define _XENIRQSTAT_needs_eoi (0)
++#define XENIRQSTAT_needs_eoi (1U<<_XENIRQSTAT_needs_eoi)
+
+ /* IRQ shared by multiple guests? */
+-#define _XENIRQSTAT_shared (1)
+-#define XENIRQSTAT_shared (1U<<_XENIRQSTAT_shared)
++#define _XENIRQSTAT_shared (1)
++#define XENIRQSTAT_shared (1U<<_XENIRQSTAT_shared)
+
+ /*
+ * Set the current VCPU's I/O privilege level.
+ * @arg == pointer to physdev_set_iopl structure.
+ */
+-#define PHYSDEVOP_set_iopl 6
++#define PHYSDEVOP_set_iopl 6
+ struct physdev_set_iopl {
+- /* IN */
+- uint32_t iopl;
++ /* IN */
++ uint32_t iopl;
+ };
++typedef struct physdev_set_iopl physdev_set_iopl_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_set_iopl_t);
+
+ /*
+ * Set the current VCPU's I/O-port permissions bitmap.
+ * @arg == pointer to physdev_set_iobitmap structure.
+ */
+-#define PHYSDEVOP_set_iobitmap 7
++#define PHYSDEVOP_set_iobitmap 7
+ struct physdev_set_iobitmap {
+- /* IN */
+- uint8_t * bitmap;
+- uint32_t nr_ports;
++ /* IN */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030205
++ XEN_GUEST_HANDLE(uint8) bitmap;
++#else
++ uint8_t *bitmap;
++#endif
++ uint32_t nr_ports;
+ };
++typedef struct physdev_set_iobitmap physdev_set_iobitmap_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_set_iobitmap_t);
+
+ /*
+ * Read or write an IO-APIC register.
+ * @arg == pointer to physdev_apic structure.
+ */
+-#define PHYSDEVOP_apic_read 8
+-#define PHYSDEVOP_apic_write 9
++#define PHYSDEVOP_apic_read 8
++#define PHYSDEVOP_apic_write 9
+ struct physdev_apic {
+- /* IN */
+- unsigned long apic_physbase;
+- uint32_t reg;
+- /* IN or OUT */
+- uint32_t value;
++ /* IN */
++ unsigned long apic_physbase;
++ uint32_t reg;
++ /* IN or OUT */
++ uint32_t value;
+ };
++typedef struct physdev_apic physdev_apic_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_apic_t);
+
+ /*
+ * Allocate or free a physical upcall vector for the specified IRQ line.
+ * @arg == pointer to physdev_irq structure.
+ */
+-#define PHYSDEVOP_alloc_irq_vector 10
+-#define PHYSDEVOP_free_irq_vector 11
++#define PHYSDEVOP_alloc_irq_vector 10
++#define PHYSDEVOP_free_irq_vector 11
+ struct physdev_irq {
+- /* IN */
+- uint32_t irq;
+- /* IN or OUT */
+- uint32_t vector;
++ /* IN */
++ uint32_t irq;
++ /* IN or OUT */
++ uint32_t vector;
+ };
++typedef struct physdev_irq physdev_irq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_irq_t);
++
++#define MAP_PIRQ_TYPE_MSI 0x0
++#define MAP_PIRQ_TYPE_GSI 0x1
++#define MAP_PIRQ_TYPE_UNKNOWN 0x2
++
++#define PHYSDEVOP_map_pirq 13
++struct physdev_map_pirq {
++ domid_t domid;
++ /* IN */
++ int type;
++ /* IN */
++ int index;
++ /* IN or OUT */
++ int pirq;
++ /* IN */
++ struct {
++ int bus, devfn, entry_nr;
++ int msi; /* 0 - MSIX 1 - MSI */
++ } msi_info;
++};
++typedef struct physdev_map_pirq physdev_map_pirq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_map_pirq_t);
++
++#define PHYSDEVOP_unmap_pirq 14
++struct physdev_unmap_pirq {
++ domid_t domid;
++ /* IN */
++ int pirq;
++};
++
++typedef struct physdev_unmap_pirq physdev_unmap_pirq_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_unmap_pirq_t);
++
++#define PHYSDEVOP_manage_pci_add 15
++#define PHYSDEVOP_manage_pci_remove 16
++struct physdev_manage_pci {
++ /* IN */
++ uint8_t bus;
++ uint8_t devfn;
++};
++
++typedef struct physdev_manage_pci physdev_manage_pci_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_manage_pci_t);
+
+ /*
+ * Argument to physdev_op_compat() hypercall. Superceded by new physdev_op()
+ * hypercall since 0x00030202.
+ */
+ struct physdev_op {
+- uint32_t cmd;
+- union {
+- struct physdev_irq_status_query irq_status_query;
+- struct physdev_set_iopl set_iopl;
+- struct physdev_set_iobitmap set_iobitmap;
+- struct physdev_apic apic_op;
+- struct physdev_irq irq_op;
+- } u;
++ uint32_t cmd;
++ union {
++ struct physdev_irq_status_query irq_status_query;
++ struct physdev_set_iopl set_iopl;
++ struct physdev_set_iobitmap set_iobitmap;
++ struct physdev_apic apic_op;
++ struct physdev_irq irq_op;
++ } u;
+ };
++typedef struct physdev_op physdev_op_t;
++DEFINE_XEN_GUEST_HANDLE(physdev_op_t);
+
+ /*
+ * Notify that some PIRQ-bound event channels have been unmasked.
+ * ** This command is obsolete since interface version 0x00030202 and is **
+- * ** unsupported by newer versions of Xen. **
++ * ** unsupported by newer versions of Xen. **
+ */
+-#define PHYSDEVOP_IRQ_UNMASK_NOTIFY 4
++#define PHYSDEVOP_IRQ_UNMASK_NOTIFY 4
+
+ /*
+ * These all-capitals physdev operation names are superceded by the new names
+ * (defined above) since interface version 0x00030202.
+ */
+-#define PHYSDEVOP_IRQ_STATUS_QUERY PHYSDEVOP_irq_status_query
+-#define PHYSDEVOP_SET_IOPL PHYSDEVOP_set_iopl
+-#define PHYSDEVOP_SET_IOBITMAP PHYSDEVOP_set_iobitmap
+-#define PHYSDEVOP_APIC_READ PHYSDEVOP_apic_read
+-#define PHYSDEVOP_APIC_WRITE PHYSDEVOP_apic_write
+-#define PHYSDEVOP_ASSIGN_VECTOR PHYSDEVOP_alloc_irq_vector
+-#define PHYSDEVOP_FREE_VECTOR PHYSDEVOP_free_irq_vector
++#define PHYSDEVOP_IRQ_STATUS_QUERY PHYSDEVOP_irq_status_query
++#define PHYSDEVOP_SET_IOPL PHYSDEVOP_set_iopl
++#define PHYSDEVOP_SET_IOBITMAP PHYSDEVOP_set_iobitmap
++#define PHYSDEVOP_APIC_READ PHYSDEVOP_apic_read
++#define PHYSDEVOP_APIC_WRITE PHYSDEVOP_apic_write
++#define PHYSDEVOP_ASSIGN_VECTOR PHYSDEVOP_alloc_irq_vector
++#define PHYSDEVOP_FREE_VECTOR PHYSDEVOP_free_irq_vector
+ #define PHYSDEVOP_IRQ_NEEDS_UNMASK_NOTIFY XENIRQSTAT_needs_eoi
+-#define PHYSDEVOP_IRQ_SHARED XENIRQSTAT_shared
++#define PHYSDEVOP_IRQ_SHARED XENIRQSTAT_shared
+
+ #endif /* __XEN_PUBLIC_PHYSDEV_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/platform.h b/include/xen/interface/platform.h
+new file mode 100644
+index 0000000..3f4fbfb
+--- /dev/null
++++ b/include/xen/interface/platform.h
+@@ -0,0 +1,346 @@
++/******************************************************************************
++ * platform.h
++ *
++ * Hardware platform operations. Intended for use by domain-0 kernel.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2002-2006, K Fraser
++ */
++
++#ifndef __XEN_PUBLIC_PLATFORM_H__
++#define __XEN_PUBLIC_PLATFORM_H__
++
++#include "xen.h"
++
++#define XENPF_INTERFACE_VERSION 0x03000001
++
++/*
++ * Set clock such that it would read <secs,nsecs> after 00:00:00 UTC,
++ * 1 January, 1970 if the current system time was <system_time>.
++ */
++#define XENPF_settime 17
++struct xenpf_settime {
++ /* IN variables. */
++ uint32_t secs;
++ uint32_t nsecs;
++ uint64_t system_time;
++};
++typedef struct xenpf_settime xenpf_settime_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_settime_t);
++
++/*
++ * Request memory range (@mfn, @mfn+ at nr_mfns-1) to have type @type.
++ * On x86, @type is an architecture-defined MTRR memory type.
++ * On success, returns the MTRR that was used (@reg) and a handle that can
++ * be passed to XENPF_DEL_MEMTYPE to accurately tear down the new setting.
++ * (x86-specific).
++ */
++#define XENPF_add_memtype 31
++struct xenpf_add_memtype {
++ /* IN variables. */
++ xen_pfn_t mfn;
++ uint64_t nr_mfns;
++ uint32_t type;
++ /* OUT variables. */
++ uint32_t handle;
++ uint32_t reg;
++};
++typedef struct xenpf_add_memtype xenpf_add_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_add_memtype_t);
++
++/*
++ * Tear down an existing memory-range type. If @handle is remembered then it
++ * should be passed in to accurately tear down the correct setting (in case
++ * of overlapping memory regions with differing types). If it is not known
++ * then @handle should be set to zero. In all cases @reg must be set.
++ * (x86-specific).
++ */
++#define XENPF_del_memtype 32
++struct xenpf_del_memtype {
++ /* IN variables. */
++ uint32_t handle;
++ uint32_t reg;
++};
++typedef struct xenpf_del_memtype xenpf_del_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_del_memtype_t);
++
++/* Read current type of an MTRR (x86-specific). */
++#define XENPF_read_memtype 33
++struct xenpf_read_memtype {
++ /* IN variables. */
++ uint32_t reg;
++ /* OUT variables. */
++ xen_pfn_t mfn;
++ uint64_t nr_mfns;
++ uint32_t type;
++};
++typedef struct xenpf_read_memtype xenpf_read_memtype_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_read_memtype_t);
++
++#define XENPF_microcode_update 35
++struct xenpf_microcode_update {
++ /* IN variables. */
++ XEN_GUEST_HANDLE(void) data; /* Pointer to microcode data */
++ uint32_t length; /* Length of microcode data. */
++};
++typedef struct xenpf_microcode_update xenpf_microcode_update_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_microcode_update_t);
++
++#define XENPF_platform_quirk 39
++#define QUIRK_NOIRQBALANCING 1 /* Do not restrict IO-APIC RTE targets */
++#define QUIRK_IOAPIC_BAD_REGSEL 2 /* IO-APIC REGSEL forgets its value */
++#define QUIRK_IOAPIC_GOOD_REGSEL 3 /* IO-APIC REGSEL behaves properly */
++struct xenpf_platform_quirk {
++ /* IN variables. */
++ uint32_t quirk_id;
++};
++typedef struct xenpf_platform_quirk xenpf_platform_quirk_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_platform_quirk_t);
++
++#define XENPF_firmware_info 50
++#define XEN_FW_DISK_INFO 1 /* from int 13 AH=08/41/48 */
++#define XEN_FW_DISK_MBR_SIGNATURE 2 /* from MBR offset 0x1b8 */
++#define XEN_FW_VBEDDC_INFO 3 /* from int 10 AX=4f15 */
++struct xenpf_firmware_info {
++ /* IN variables. */
++ uint32_t type;
++ uint32_t index;
++ /* OUT variables. */
++ union {
++ struct {
++ /* Int13, Fn48: Check Extensions Present. */
++ uint8_t device; /* %dl: bios device number */
++ uint8_t version; /* %ah: major version */
++ uint16_t interface_support; /* %cx: support bitmap */
++ /* Int13, Fn08: Legacy Get Device Parameters. */
++ uint16_t legacy_max_cylinder; /* %cl[7:6]:%ch: max cyl # */
++ uint8_t legacy_max_head; /* %dh: max head # */
++ uint8_t legacy_sectors_per_track; /* %cl[5:0]: max sector # */
++ /* Int13, Fn41: Get Device Parameters (as filled into %ds:%esi). */
++ /* NB. First uint16_t of buffer must be set to buffer size. */
++ XEN_GUEST_HANDLE(void) edd_params;
++ } disk_info; /* XEN_FW_DISK_INFO */
++ struct {
++ uint8_t device; /* bios device number */
++ uint32_t mbr_signature; /* offset 0x1b8 in mbr */
++ } disk_mbr_signature; /* XEN_FW_DISK_MBR_SIGNATURE */
++ struct {
++ /* Int10, AX=4F15: Get EDID info. */
++ uint8_t capabilities;
++ uint8_t edid_transfer_time;
++ /* must refer to 128-byte buffer */
++ XEN_GUEST_HANDLE(uint8) edid;
++ } vbeddc_info; /* XEN_FW_VBEDDC_INFO */
++ } u;
++};
++typedef struct xenpf_firmware_info xenpf_firmware_info_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_info_t);
++
++#define XENPF_enter_acpi_sleep 51
++struct xenpf_enter_acpi_sleep {
++ /* IN variables */
++ uint16_t pm1a_cnt_val; /* PM1a control value. */
++ uint16_t pm1b_cnt_val; /* PM1b control value. */
++ uint32_t sleep_state; /* Which state to enter (Sn). */
++ uint32_t flags; /* Must be zero. */
++};
++typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t);
++
++#define XENPF_change_freq 52
++struct xenpf_change_freq {
++ /* IN variables */
++ uint32_t flags; /* Must be zero. */
++ uint32_t cpu; /* Physical cpu. */
++ uint64_t freq; /* New frequency (Hz). */
++};
++typedef struct xenpf_change_freq xenpf_change_freq_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_change_freq_t);
++
++/*
++ * Get idle times (nanoseconds since boot) for physical CPUs specified in the
++ * @cpumap_bitmap with range [0.. at cpumap_nr_cpus-1]. The @idletime array is
++ * indexed by CPU number; only entries with the corresponding @cpumap_bitmap
++ * bit set are written to. On return, @cpumap_bitmap is modified so that any
++ * non-existent CPUs are cleared. Such CPUs have their @idletime array entry
++ * cleared.
++ */
++#define XENPF_getidletime 53
++struct xenpf_getidletime {
++ /* IN/OUT variables */
++ /* IN: CPUs to interrogate; OUT: subset of IN which are present */
++ XEN_GUEST_HANDLE(uint8) cpumap_bitmap;
++ /* IN variables */
++ /* Size of cpumap bitmap. */
++ uint32_t cpumap_nr_cpus;
++ /* Must be indexable for every cpu in cpumap_bitmap. */
++ XEN_GUEST_HANDLE(uint64) idletime;
++ /* OUT variables */
++ /* System time when the idletime snapshots were taken. */
++ uint64_t now;
++};
++typedef struct xenpf_getidletime xenpf_getidletime_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_getidletime_t);
++
++#define XENPF_set_processor_pminfo 54
++
++/* ability bits */
++#define XEN_PROCESSOR_PM_CX 1
++#define XEN_PROCESSOR_PM_PX 2
++#define XEN_PROCESSOR_PM_TX 4
++
++/* cmd type */
++#define XEN_PM_CX 0
++#define XEN_PM_PX 1
++#define XEN_PM_TX 2
++
++/* Px sub info type */
++#define XEN_PX_PCT 1
++#define XEN_PX_PSS 2
++#define XEN_PX_PPC 4
++#define XEN_PX_PSD 8
++
++struct xen_power_register {
++ uint32_t space_id;
++ uint32_t bit_width;
++ uint32_t bit_offset;
++ uint32_t access_size;
++ uint64_t address;
++};
++
++struct xen_processor_csd {
++ uint32_t domain; /* domain number of one dependent group */
++ uint32_t coord_type; /* coordination type */
++ uint32_t num; /* number of processors in same domain */
++};
++typedef struct xen_processor_csd xen_processor_csd_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_csd_t);
++
++struct xen_processor_cx {
++ struct xen_power_register reg; /* GAS for Cx trigger register */
++ uint8_t type; /* cstate value, c0: 0, c1: 1, ... */
++ uint32_t latency; /* worst latency (ms) to enter/exit this cstate */
++ uint32_t power; /* average power consumption(mW) */
++ uint32_t dpcnt; /* number of dependency entries */
++ XEN_GUEST_HANDLE(xen_processor_csd_t) dp; /* NULL if no dependency */
++};
++typedef struct xen_processor_cx xen_processor_cx_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_cx_t);
++
++struct xen_processor_flags {
++ uint32_t bm_control:1;
++ uint32_t bm_check:1;
++ uint32_t has_cst:1;
++ uint32_t power_setup_done:1;
++ uint32_t bm_rld_set:1;
++};
++
++struct xen_processor_power {
++ uint32_t count; /* number of C state entries in array below */
++ struct xen_processor_flags flags; /* global flags of this processor */
++ XEN_GUEST_HANDLE(xen_processor_cx_t) states; /* supported c states */
++};
++
++struct xen_pct_register {
++ uint8_t descriptor;
++ uint16_t length;
++ uint8_t space_id;
++ uint8_t bit_width;
++ uint8_t bit_offset;
++ uint8_t reserved;
++ uint64_t address;
++};
++
++struct xen_processor_px {
++ uint64_t core_frequency; /* megahertz */
++ uint64_t power; /* milliWatts */
++ uint64_t transition_latency; /* microseconds */
++ uint64_t bus_master_latency; /* microseconds */
++ uint64_t control; /* control value */
++ uint64_t status; /* success indicator */
++};
++typedef struct xen_processor_px xen_processor_px_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_px_t);
++
++struct xen_psd_package {
++ uint64_t num_entries;
++ uint64_t revision;
++ uint64_t domain;
++ uint64_t coord_type;
++ uint64_t num_processors;
++};
++
++struct xen_processor_performance {
++ uint32_t flags; /* flag for Px sub info type */
++ uint32_t ppc; /* Platform limitation on freq usage */
++ struct xen_pct_register control_register;
++ struct xen_pct_register status_register;
++ uint32_t state_count; /* total available performance states */
++ XEN_GUEST_HANDLE(xen_processor_px_t) states;
++ struct xen_psd_package domain_info;
++ uint32_t shared_type; /* coordination type of this processor */
++};
++typedef struct xen_processor_performance xen_processor_performance_t;
++DEFINE_XEN_GUEST_HANDLE(xen_processor_performance_t);
++
++struct xenpf_set_processor_pminfo {
++ /* IN variables */
++ uint32_t id; /* ACPI CPU ID */
++ uint32_t type; /* {XEN_PM_CX, XEN_PM_PX} */
++ union {
++ struct xen_processor_power power;/* Cx: _CST/_CSD */
++ struct xen_processor_performance perf; /* Px: _PPC/_PCT/_PSS/_PSD */
++ };
++};
++typedef struct xenpf_set_processor_pminfo xenpf_set_processor_pminfo_t;
++DEFINE_XEN_GUEST_HANDLE(xenpf_set_processor_pminfo_t);
++
++struct xen_platform_op {
++ uint32_t cmd;
++ uint32_t interface_version; /* XENPF_INTERFACE_VERSION */
++ union {
++ struct xenpf_settime settime;
++ struct xenpf_add_memtype add_memtype;
++ struct xenpf_del_memtype del_memtype;
++ struct xenpf_read_memtype read_memtype;
++ struct xenpf_microcode_update microcode;
++ struct xenpf_platform_quirk platform_quirk;
++ struct xenpf_firmware_info firmware_info;
++ struct xenpf_enter_acpi_sleep enter_acpi_sleep;
++ struct xenpf_change_freq change_freq;
++ struct xenpf_getidletime getidletime;
++ struct xenpf_set_processor_pminfo set_pminfo;
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_platform_op xen_platform_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_platform_op_t);
++
++#endif /* __XEN_PUBLIC_PLATFORM_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/sched.h b/include/xen/interface/sched.h
+index 5fec575..b1bf4da 100644
+--- a/include/xen/interface/sched.h
++++ b/include/xen/interface/sched.h
+@@ -3,6 +3,24 @@
+ *
+ * Scheduler state interactions
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+
+@@ -13,17 +31,17 @@
+
+ /*
+ * The prototype for this hypercall is:
+- * long sched_op_new(int cmd, void *arg)
++ * long sched_op(int cmd, void *arg)
+ * @cmd == SCHEDOP_??? (scheduler operation).
+ * @arg == Operation-specific extra argument(s), as described below.
+ *
+- * **NOTE**:
+- * Versions of Xen prior to 3.0.2 provide only the following legacy version
++ * Versions of Xen prior to 3.0.2 provided only the following legacy version
+ * of this hypercall, supporting only the commands yield, block and shutdown:
+ * long sched_op(int cmd, unsigned long arg)
+ * @cmd == SCHEDOP_??? (scheduler operation).
+ * @arg == 0 (SCHEDOP_yield and SCHEDOP_block)
+ * == SHUTDOWN_* code (SCHEDOP_shutdown)
++ * This legacy version is available to new guests as sched_op_compat().
+ */
+
+ /*
+@@ -49,7 +67,9 @@
+ struct sched_shutdown {
+ unsigned int reason; /* SHUTDOWN_* */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(sched_shutdown);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(sched_shutdown);
++typedef struct sched_shutdown sched_shutdown_t;
++DEFINE_XEN_GUEST_HANDLE(sched_shutdown_t);
+
+ /*
+ * Poll a set of event-channel ports. Return when one or more are pending. An
+@@ -58,11 +78,27 @@ DEFINE_GUEST_HANDLE_STRUCT(sched_shutdown);
+ */
+ #define SCHEDOP_poll 3
+ struct sched_poll {
+- GUEST_HANDLE(evtchn_port_t) ports;
++ XEN_GUEST_HANDLE(evtchn_port_t) ports;
+ unsigned int nr_ports;
+ uint64_t timeout;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(sched_poll);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(sched_poll);
++typedef struct sched_poll sched_poll_t;
++DEFINE_XEN_GUEST_HANDLE(sched_poll_t);
++
++/*
++ * Declare a shutdown for another domain. The main use of this function is
++ * in interpreting shutdown requests and reasons for fully-virtualized
++ * domains. A para-virtualized domain may use SCHEDOP_shutdown directly.
++ * @arg == pointer to sched_remote_shutdown structure.
++ */
++#define SCHEDOP_remote_shutdown 4
++struct sched_remote_shutdown {
++ domid_t domain_id; /* Remote domain ID */
++ unsigned int reason; /* SHUTDOWN_xxx reason */
++};
++typedef struct sched_remote_shutdown sched_remote_shutdown_t;
++DEFINE_XEN_GUEST_HANDLE(sched_remote_shutdown_t);
+
+ /*
+ * Reason codes for SCHEDOP_shutdown. These may be interpreted by control
+@@ -75,3 +111,13 @@ DEFINE_GUEST_HANDLE_STRUCT(sched_poll);
+ #define SHUTDOWN_crash 3 /* Tell controller we've crashed. */
+
+ #endif /* __XEN_PUBLIC_SCHED_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/sysctl.h b/include/xen/interface/sysctl.h
+new file mode 100644
+index 0000000..76998c9
+--- /dev/null
++++ b/include/xen/interface/sysctl.h
+@@ -0,0 +1,295 @@
++/******************************************************************************
++ * sysctl.h
++ *
++ * System management operations. For use by node control stack.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2002-2006, K Fraser
++ */
++
++#ifndef __XEN_PUBLIC_SYSCTL_H__
++#define __XEN_PUBLIC_SYSCTL_H__
++
++#if !defined(__XEN__) && !defined(__XEN_TOOLS__)
++#error "sysctl operations are intended for use by node control tools only"
++#endif
++
++#include "xen.h"
++#include "domctl.h"
++
++#define XEN_SYSCTL_INTERFACE_VERSION 0x00000006
++
++/*
++ * Read console content from Xen buffer ring.
++ */
++#define XEN_SYSCTL_readconsole 1
++struct xen_sysctl_readconsole {
++ /* IN: Non-zero -> clear after reading. */
++ uint8_t clear;
++ /* IN: Non-zero -> start index specified by @index field. */
++ uint8_t incremental;
++ uint8_t pad0, pad1;
++ /*
++ * IN: Start index for consuming from ring buffer (if @incremental);
++ * OUT: End index after consuming from ring buffer.
++ */
++ uint32_t index;
++ /* IN: Virtual address to write console data. */
++ XEN_GUEST_HANDLE_64(char) buffer;
++ /* IN: Size of buffer; OUT: Bytes written to buffer. */
++ uint32_t count;
++};
++typedef struct xen_sysctl_readconsole xen_sysctl_readconsole_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_readconsole_t);
++
++/* Get trace buffers machine base address */
++#define XEN_SYSCTL_tbuf_op 2
++struct xen_sysctl_tbuf_op {
++ /* IN variables */
++#define XEN_SYSCTL_TBUFOP_get_info 0
++#define XEN_SYSCTL_TBUFOP_set_cpu_mask 1
++#define XEN_SYSCTL_TBUFOP_set_evt_mask 2
++#define XEN_SYSCTL_TBUFOP_set_size 3
++#define XEN_SYSCTL_TBUFOP_enable 4
++#define XEN_SYSCTL_TBUFOP_disable 5
++ uint32_t cmd;
++ /* IN/OUT variables */
++ struct xenctl_cpumap cpu_mask;
++ uint32_t evt_mask;
++ /* OUT variables */
++ uint64_aligned_t buffer_mfn;
++ uint32_t size;
++};
++typedef struct xen_sysctl_tbuf_op xen_sysctl_tbuf_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_tbuf_op_t);
++
++/*
++ * Get physical information about the host machine
++ */
++#define XEN_SYSCTL_physinfo 3
++ /* (x86) The platform supports HVM guests. */
++#define _XEN_SYSCTL_PHYSCAP_hvm 0
++#define XEN_SYSCTL_PHYSCAP_hvm (1u<<_XEN_SYSCTL_PHYSCAP_hvm)
++ /* (x86) The platform supports HVM-guest direct access to I/O devices. */
++#define _XEN_SYSCTL_PHYSCAP_hvm_directio 1
++#define XEN_SYSCTL_PHYSCAP_hvm_directio (1u<<_XEN_SYSCTL_PHYSCAP_hvm_directio)
++struct xen_sysctl_physinfo {
++ uint32_t threads_per_core;
++ uint32_t cores_per_socket;
++ uint32_t nr_cpus;
++ uint32_t nr_nodes;
++ uint32_t cpu_khz;
++ uint64_aligned_t total_pages;
++ uint64_aligned_t free_pages;
++ uint64_aligned_t scrub_pages;
++ uint32_t hw_cap[8];
++
++ /*
++ * IN: maximum addressable entry in the caller-provided cpu_to_node array.
++ * OUT: largest cpu identifier in the system.
++ * If OUT is greater than IN then the cpu_to_node array is truncated!
++ */
++ uint32_t max_cpu_id;
++ /*
++ * If not NULL, this array is filled with node identifier for each cpu.
++ * If a cpu has no node information (e.g., cpu not present) then the
++ * sentinel value ~0u is written.
++ * The size of this array is specified by the caller in @max_cpu_id.
++ * If the actual @max_cpu_id is smaller than the array then the trailing
++ * elements of the array will not be written by the sysctl.
++ */
++ XEN_GUEST_HANDLE_64(uint32) cpu_to_node;
++
++ /* XEN_SYSCTL_PHYSCAP_??? */
++ uint32_t capabilities;
++};
++typedef struct xen_sysctl_physinfo xen_sysctl_physinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_physinfo_t);
++
++/*
++ * Get the ID of the current scheduler.
++ */
++#define XEN_SYSCTL_sched_id 4
++struct xen_sysctl_sched_id {
++ /* OUT variable */
++ uint32_t sched_id;
++};
++typedef struct xen_sysctl_sched_id xen_sysctl_sched_id_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_sched_id_t);
++
++/* Interface for controlling Xen software performance counters. */
++#define XEN_SYSCTL_perfc_op 5
++/* Sub-operations: */
++#define XEN_SYSCTL_PERFCOP_reset 1 /* Reset all counters to zero. */
++#define XEN_SYSCTL_PERFCOP_query 2 /* Get perfctr information. */
++struct xen_sysctl_perfc_desc {
++ char name[80]; /* name of perf counter */
++ uint32_t nr_vals; /* number of values for this counter */
++};
++typedef struct xen_sysctl_perfc_desc xen_sysctl_perfc_desc_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_desc_t);
++typedef uint32_t xen_sysctl_perfc_val_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_val_t);
++
++struct xen_sysctl_perfc_op {
++ /* IN variables. */
++ uint32_t cmd; /* XEN_SYSCTL_PERFCOP_??? */
++ /* OUT variables. */
++ uint32_t nr_counters; /* number of counters description */
++ uint32_t nr_vals; /* number of values */
++ /* counter information (or NULL) */
++ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_desc_t) desc;
++ /* counter values (or NULL) */
++ XEN_GUEST_HANDLE_64(xen_sysctl_perfc_val_t) val;
++};
++typedef struct xen_sysctl_perfc_op xen_sysctl_perfc_op_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_perfc_op_t);
++
++#define XEN_SYSCTL_getdomaininfolist 6
++struct xen_sysctl_getdomaininfolist {
++ /* IN variables. */
++ domid_t first_domain;
++ uint32_t max_domains;
++ XEN_GUEST_HANDLE_64(xen_domctl_getdomaininfo_t) buffer;
++ /* OUT variables. */
++ uint32_t num_domains;
++};
++typedef struct xen_sysctl_getdomaininfolist xen_sysctl_getdomaininfolist_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getdomaininfolist_t);
++
++/* Inject debug keys into Xen. */
++#define XEN_SYSCTL_debug_keys 7
++struct xen_sysctl_debug_keys {
++ /* IN variables. */
++ XEN_GUEST_HANDLE_64(char) keys;
++ uint32_t nr_keys;
++};
++typedef struct xen_sysctl_debug_keys xen_sysctl_debug_keys_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_debug_keys_t);
++
++/* Get physical CPU information. */
++#define XEN_SYSCTL_getcpuinfo 8
++struct xen_sysctl_cpuinfo {
++ uint64_aligned_t idletime;
++};
++typedef struct xen_sysctl_cpuinfo xen_sysctl_cpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_cpuinfo_t);
++struct xen_sysctl_getcpuinfo {
++ /* IN variables. */
++ uint32_t max_cpus;
++ XEN_GUEST_HANDLE_64(xen_sysctl_cpuinfo_t) info;
++ /* OUT variables. */
++ uint32_t nr_cpus;
++};
++typedef struct xen_sysctl_getcpuinfo xen_sysctl_getcpuinfo_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_getcpuinfo_t);
++
++#define XEN_SYSCTL_availheap 9
++struct xen_sysctl_availheap {
++ /* IN variables. */
++ uint32_t min_bitwidth; /* Smallest address width (zero if don't care). */
++ uint32_t max_bitwidth; /* Largest address width (zero if don't care). */
++ int32_t node; /* NUMA node of interest (-1 for all nodes). */
++ /* OUT variables. */
++ uint64_aligned_t avail_bytes;/* Bytes available in the specified region. */
++};
++typedef struct xen_sysctl_availheap xen_sysctl_availheap_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_availheap_t);
++
++#define XEN_SYSCTL_get_pmstat 10
++struct pm_px_val {
++ uint64_aligned_t freq; /* Px core frequency */
++ uint64_aligned_t residency; /* Px residency time */
++ uint64_aligned_t count; /* Px transition count */
++};
++typedef struct pm_px_val pm_px_val_t;
++DEFINE_XEN_GUEST_HANDLE(pm_px_val_t);
++
++struct pm_px_stat {
++ uint8_t total; /* total Px states */
++ uint8_t usable; /* usable Px states */
++ uint8_t last; /* last Px state */
++ uint8_t cur; /* current Px state */
++ XEN_GUEST_HANDLE_64(uint64) trans_pt; /* Px transition table */
++ XEN_GUEST_HANDLE_64(pm_px_val_t) pt;
++};
++typedef struct pm_px_stat pm_px_stat_t;
++DEFINE_XEN_GUEST_HANDLE(pm_px_stat_t);
++
++struct pm_cx_stat {
++ uint32_t nr; /* entry nr in triggers & residencies, including C0 */
++ uint32_t last; /* last Cx state */
++ uint64_aligned_t idle_time; /* idle time from boot */
++ XEN_GUEST_HANDLE_64(uint64) triggers; /* Cx trigger counts */
++ XEN_GUEST_HANDLE_64(uint64) residencies; /* Cx residencies */
++};
++
++struct xen_sysctl_get_pmstat {
++#define PMSTAT_CATEGORY_MASK 0xf0
++#define PMSTAT_PX 0x10
++#define PMSTAT_CX 0x20
++#define PMSTAT_get_max_px (PMSTAT_PX | 0x1)
++#define PMSTAT_get_pxstat (PMSTAT_PX | 0x2)
++#define PMSTAT_reset_pxstat (PMSTAT_PX | 0x3)
++#define PMSTAT_get_max_cx (PMSTAT_CX | 0x1)
++#define PMSTAT_get_cxstat (PMSTAT_CX | 0x2)
++#define PMSTAT_reset_cxstat (PMSTAT_CX | 0x3)
++ uint32_t type;
++ uint32_t cpuid;
++ union {
++ struct pm_px_stat getpx;
++ struct pm_cx_stat getcx;
++ /* other struct for tx, etc */
++ } u;
++};
++typedef struct xen_sysctl_get_pmstat xen_sysctl_get_pmstat_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_get_pmstat_t);
++
++struct xen_sysctl {
++ uint32_t cmd;
++ uint32_t interface_version; /* XEN_SYSCTL_INTERFACE_VERSION */
++ union {
++ struct xen_sysctl_readconsole readconsole;
++ struct xen_sysctl_tbuf_op tbuf_op;
++ struct xen_sysctl_physinfo physinfo;
++ struct xen_sysctl_sched_id sched_id;
++ struct xen_sysctl_perfc_op perfc_op;
++ struct xen_sysctl_getdomaininfolist getdomaininfolist;
++ struct xen_sysctl_debug_keys debug_keys;
++ struct xen_sysctl_getcpuinfo getcpuinfo;
++ struct xen_sysctl_availheap availheap;
++ struct xen_sysctl_get_pmstat get_pmstat;
++ uint8_t pad[128];
++ } u;
++};
++typedef struct xen_sysctl xen_sysctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_sysctl_t);
++
++#endif /* __XEN_PUBLIC_SYSCTL_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/trace.h b/include/xen/interface/trace.h
+new file mode 100644
+index 0000000..21fe15f
+--- /dev/null
++++ b/include/xen/interface/trace.h
+@@ -0,0 +1,167 @@
++/******************************************************************************
++ * include/public/trace.h
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Mark Williamson, (C) 2004 Intel Research Cambridge
++ * Copyright (C) 2005 Bin Ren
++ */
++
++#ifndef __XEN_PUBLIC_TRACE_H__
++#define __XEN_PUBLIC_TRACE_H__
++
++#define TRACE_EXTRA_MAX 7
++#define TRACE_EXTRA_SHIFT 28
++
++/* Trace classes */
++#define TRC_CLS_SHIFT 16
++#define TRC_GEN 0x0001f000 /* General trace */
++#define TRC_SCHED 0x0002f000 /* Xen Scheduler trace */
++#define TRC_DOM0OP 0x0004f000 /* Xen DOM0 operation trace */
++#define TRC_HVM 0x0008f000 /* Xen HVM trace */
++#define TRC_MEM 0x0010f000 /* Xen memory trace */
++#define TRC_PV 0x0020f000 /* Xen PV traces */
++#define TRC_ALL 0x0ffff000
++#define TRC_HD_TO_EVENT(x) ((x)&0x0fffffff)
++#define TRC_HD_CYCLE_FLAG (1UL<<31)
++#define TRC_HD_INCLUDES_CYCLE_COUNT(x) ( !!( (x) & TRC_HD_CYCLE_FLAG ) )
++#define TRC_HD_EXTRA(x) (((x)>>TRACE_EXTRA_SHIFT)&TRACE_EXTRA_MAX)
++
++/* Trace subclasses */
++#define TRC_SUBCLS_SHIFT 12
++
++/* trace subclasses for SVM */
++#define TRC_HVM_ENTRYEXIT 0x00081000 /* VMENTRY and #VMEXIT */
++#define TRC_HVM_HANDLER 0x00082000 /* various HVM handlers */
++
++/* Trace events per class */
++#define TRC_LOST_RECORDS (TRC_GEN + 1)
++#define TRC_TRACE_WRAP_BUFFER (TRC_GEN + 2)
++#define TRC_TRACE_CPU_CHANGE (TRC_GEN + 3)
++
++#define TRC_SCHED_DOM_ADD (TRC_SCHED + 1)
++#define TRC_SCHED_DOM_REM (TRC_SCHED + 2)
++#define TRC_SCHED_SLEEP (TRC_SCHED + 3)
++#define TRC_SCHED_WAKE (TRC_SCHED + 4)
++#define TRC_SCHED_YIELD (TRC_SCHED + 5)
++#define TRC_SCHED_BLOCK (TRC_SCHED + 6)
++#define TRC_SCHED_SHUTDOWN (TRC_SCHED + 7)
++#define TRC_SCHED_CTL (TRC_SCHED + 8)
++#define TRC_SCHED_ADJDOM (TRC_SCHED + 9)
++#define TRC_SCHED_SWITCH (TRC_SCHED + 10)
++#define TRC_SCHED_S_TIMER_FN (TRC_SCHED + 11)
++#define TRC_SCHED_T_TIMER_FN (TRC_SCHED + 12)
++#define TRC_SCHED_DOM_TIMER_FN (TRC_SCHED + 13)
++#define TRC_SCHED_SWITCH_INFPREV (TRC_SCHED + 14)
++#define TRC_SCHED_SWITCH_INFNEXT (TRC_SCHED + 15)
++
++#define TRC_MEM_PAGE_GRANT_MAP (TRC_MEM + 1)
++#define TRC_MEM_PAGE_GRANT_UNMAP (TRC_MEM + 2)
++#define TRC_MEM_PAGE_GRANT_TRANSFER (TRC_MEM + 3)
++
++#define TRC_PV_HYPERCALL (TRC_PV + 1)
++#define TRC_PV_TRAP (TRC_PV + 3)
++#define TRC_PV_PAGE_FAULT (TRC_PV + 4)
++#define TRC_PV_FORCED_INVALID_OP (TRC_PV + 5)
++#define TRC_PV_EMULATE_PRIVOP (TRC_PV + 6)
++#define TRC_PV_EMULATE_4GB (TRC_PV + 7)
++#define TRC_PV_MATH_STATE_RESTORE (TRC_PV + 8)
++#define TRC_PV_PAGING_FIXUP (TRC_PV + 9)
++#define TRC_PV_GDT_LDT_MAPPING_FAULT (TRC_PV + 10)
++#define TRC_PV_PTWR_EMULATION (TRC_PV + 11)
++#define TRC_PV_PTWR_EMULATION_PAE (TRC_PV + 12)
++ /* Indicates that addresses in trace record are 64 bits */
++#define TRC_PV_64_FLAG (0x100)
++
++/* trace events per subclass */
++#define TRC_HVM_VMENTRY (TRC_HVM_ENTRYEXIT + 0x01)
++#define TRC_HVM_VMEXIT (TRC_HVM_ENTRYEXIT + 0x02)
++#define TRC_HVM_VMEXIT64 (TRC_HVM_ENTRYEXIT + 0x03)
++#define TRC_HVM_PF_XEN (TRC_HVM_HANDLER + 0x01)
++#define TRC_HVM_PF_INJECT (TRC_HVM_HANDLER + 0x02)
++#define TRC_HVM_INJ_EXC (TRC_HVM_HANDLER + 0x03)
++#define TRC_HVM_INJ_VIRQ (TRC_HVM_HANDLER + 0x04)
++#define TRC_HVM_REINJ_VIRQ (TRC_HVM_HANDLER + 0x05)
++#define TRC_HVM_IO_READ (TRC_HVM_HANDLER + 0x06)
++#define TRC_HVM_IO_WRITE (TRC_HVM_HANDLER + 0x07)
++#define TRC_HVM_CR_READ (TRC_HVM_HANDLER + 0x08)
++#define TRC_HVM_CR_WRITE (TRC_HVM_HANDLER + 0x09)
++#define TRC_HVM_DR_READ (TRC_HVM_HANDLER + 0x0A)
++#define TRC_HVM_DR_WRITE (TRC_HVM_HANDLER + 0x0B)
++#define TRC_HVM_MSR_READ (TRC_HVM_HANDLER + 0x0C)
++#define TRC_HVM_MSR_WRITE (TRC_HVM_HANDLER + 0x0D)
++#define TRC_HVM_CPUID (TRC_HVM_HANDLER + 0x0E)
++#define TRC_HVM_INTR (TRC_HVM_HANDLER + 0x0F)
++#define TRC_HVM_NMI (TRC_HVM_HANDLER + 0x10)
++#define TRC_HVM_SMI (TRC_HVM_HANDLER + 0x11)
++#define TRC_HVM_VMMCALL (TRC_HVM_HANDLER + 0x12)
++#define TRC_HVM_HLT (TRC_HVM_HANDLER + 0x13)
++#define TRC_HVM_INVLPG (TRC_HVM_HANDLER + 0x14)
++#define TRC_HVM_MCE (TRC_HVM_HANDLER + 0x15)
++#define TRC_HVM_IO_ASSIST (TRC_HVM_HANDLER + 0x16)
++#define TRC_HVM_MMIO_ASSIST (TRC_HVM_HANDLER + 0x17)
++#define TRC_HVM_CLTS (TRC_HVM_HANDLER + 0x18)
++#define TRC_HVM_LMSW (TRC_HVM_HANDLER + 0x19)
++#define TRC_HVM_PF_XEN64 (TRC_HVM_HANDLER + 0x20)
++
++/* This structure represents a single trace buffer record. */
++struct t_rec {
++ uint32_t event:28;
++ uint32_t extra_u32:3; /* # entries in trailing extra_u32[] array */
++ uint32_t cycles_included:1; /* u.cycles or u.no_cycles? */
++ union {
++ struct {
++ uint32_t cycles_lo, cycles_hi; /* cycle counter timestamp */
++ uint32_t extra_u32[7]; /* event data items */
++ } cycles;
++ struct {
++ uint32_t extra_u32[7]; /* event data items */
++ } nocycles;
++ } u;
++};
++
++/*
++ * This structure contains the metadata for a single trace buffer. The head
++ * field, indexes into an array of struct t_rec's.
++ */
++struct t_buf {
++ /* Assume the data buffer size is X. X is generally not a power of 2.
++ * CONS and PROD are incremented modulo (2*X):
++ * 0 <= cons < 2*X
++ * 0 <= prod < 2*X
++ * This is done because addition modulo X breaks at 2^32 when X is not a
++ * power of 2:
++ * (((2^32 - 1) % X) + 1) % X != (2^32) % X
++ */
++ uint32_t cons; /* Offset of next item to be consumed by control tools. */
++ uint32_t prod; /* Offset of next item to be produced by Xen. */
++ /* Records follow immediately after the meta-data header. */
++};
++
++#endif /* __XEN_PUBLIC_TRACE_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/vcpu.h b/include/xen/interface/vcpu.h
+index 87e6f8a..352cd5a 100644
+--- a/include/xen/interface/vcpu.h
++++ b/include/xen/interface/vcpu.h
+@@ -29,9 +29,9 @@
+
+ /*
+ * Prototype for this hypercall is:
+- * int vcpu_op(int cmd, int vcpuid, void *extra_args)
+- * @cmd == VCPUOP_??? (VCPU operation).
+- * @vcpuid == VCPU to operate on.
++ * int vcpu_op(int cmd, int vcpuid, void *extra_args)
++ * @cmd == VCPUOP_??? (VCPU operation).
++ * @vcpuid == VCPU to operate on.
+ * @extra_args == Operation-specific extra arguments (NULL if none).
+ */
+
+@@ -40,52 +40,54 @@
+ * newly-initialised VCPU will not run until it is brought up by VCPUOP_up.
+ *
+ * @extra_arg == pointer to vcpu_guest_context structure containing initial
+- * state for the VCPU.
++ * state for the VCPU.
+ */
+-#define VCPUOP_initialise 0
++#define VCPUOP_initialise 0
+
+ /*
+ * Bring up a VCPU. This makes the VCPU runnable. This operation will fail
+ * if the VCPU has not been initialised (VCPUOP_initialise).
+ */
+-#define VCPUOP_up 1
++#define VCPUOP_up 1
+
+ /*
+ * Bring down a VCPU (i.e., make it non-runnable).
+ * There are a few caveats that callers should observe:
+- * 1. This operation may return, and VCPU_is_up may return false, before the
+- * VCPU stops running (i.e., the command is asynchronous). It is a good
+- * idea to ensure that the VCPU has entered a non-critical loop before
+- * bringing it down. Alternatively, this operation is guaranteed
+- * synchronous if invoked by the VCPU itself.
+- * 2. After a VCPU is initialised, there is currently no way to drop all its
+- * references to domain memory. Even a VCPU that is down still holds
+- * memory references via its pagetable base pointer and GDT. It is good
+- * practise to move a VCPU onto an 'idle' or default page table, LDT and
+- * GDT before bringing it down.
++ * 1. This operation may return, and VCPU_is_up may return false, before the
++ * VCPU stops running (i.e., the command is asynchronous). It is a good
++ * idea to ensure that the VCPU has entered a non-critical loop before
++ * bringing it down. Alternatively, this operation is guaranteed
++ * synchronous if invoked by the VCPU itself.
++ * 2. After a VCPU is initialised, there is currently no way to drop all its
++ * references to domain memory. Even a VCPU that is down still holds
++ * memory references via its pagetable base pointer and GDT. It is good
++ * practise to move a VCPU onto an 'idle' or default page table, LDT and
++ * GDT before bringing it down.
+ */
+-#define VCPUOP_down 2
++#define VCPUOP_down 2
+
+ /* Returns 1 if the given VCPU is up. */
+-#define VCPUOP_is_up 3
++#define VCPUOP_is_up 3
+
+ /*
+ * Return information about the state and running time of a VCPU.
+ * @extra_arg == pointer to vcpu_runstate_info structure.
+ */
+-#define VCPUOP_get_runstate_info 4
++#define VCPUOP_get_runstate_info 4
+ struct vcpu_runstate_info {
+- /* VCPU's current state (RUNSTATE_*). */
+- int state;
+- /* When was current state entered (system time, ns)? */
+- uint64_t state_entry_time;
+- /*
+- * Time spent in each RUNSTATE_* (ns). The sum of these times is
+- * guaranteed not to drift from system time.
+- */
+- uint64_t time[4];
++ /* VCPU's current state (RUNSTATE_*). */
++ int state;
++ /* When was current state entered (system time, ns)? */
++ uint64_t state_entry_time;
++ /*
++ * Time spent in each RUNSTATE_* (ns). The sum of these times is
++ * guaranteed not to drift from system time.
++ */
++ uint64_t time[4];
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(vcpu_runstate_info);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(vcpu_runstate_info);
++typedef struct vcpu_runstate_info vcpu_runstate_info_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_runstate_info_t);
+
+ /* VCPU is currently running on a physical CPU. */
+ #define RUNSTATE_running 0
+@@ -108,47 +110,54 @@ DEFINE_GUEST_HANDLE_STRUCT(vcpu_runstate_info);
+ * Register a shared memory area from which the guest may obtain its own
+ * runstate information without needing to execute a hypercall.
+ * Notes:
+- * 1. The registered address may be virtual or physical, depending on the
+- * platform. The virtual address should be registered on x86 systems.
+- * 2. Only one shared area may be registered per VCPU. The shared area is
+- * updated by the hypervisor each time the VCPU is scheduled. Thus
+- * runstate.state will always be RUNSTATE_running and
+- * runstate.state_entry_time will indicate the system time at which the
+- * VCPU was last scheduled to run.
++ * 1. The registered address may be virtual or physical or guest handle,
++ * depending on the platform. Virtual address or guest handle should be
++ * registered on x86 systems.
++ * 2. Only one shared area may be registered per VCPU. The shared area is
++ * updated by the hypervisor each time the VCPU is scheduled. Thus
++ * runstate.state will always be RUNSTATE_running and
++ * runstate.state_entry_time will indicate the system time at which the
++ * VCPU was last scheduled to run.
+ * @extra_arg == pointer to vcpu_register_runstate_memory_area structure.
+ */
+ #define VCPUOP_register_runstate_memory_area 5
+ struct vcpu_register_runstate_memory_area {
+- union {
+- GUEST_HANDLE(vcpu_runstate_info) h;
+- struct vcpu_runstate_info *v;
+- uint64_t p;
+- } addr;
++ union {
++ XEN_GUEST_HANDLE(vcpu_runstate_info_t) h;
++ struct vcpu_runstate_info *v;
++ uint64_t p;
++ } addr;
+ };
++typedef struct vcpu_register_runstate_memory_area vcpu_register_runstate_memory_area_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_register_runstate_memory_area_t);
+
+ /*
+ * Set or stop a VCPU's periodic timer. Every VCPU has one periodic timer
+ * which can be set via these commands. Periods smaller than one millisecond
+ * may not be supported.
+ */
+-#define VCPUOP_set_periodic_timer 6 /* arg == vcpu_set_periodic_timer_t */
+-#define VCPUOP_stop_periodic_timer 7 /* arg == NULL */
++#define VCPUOP_set_periodic_timer 6 /* arg == vcpu_set_periodic_timer_t */
++#define VCPUOP_stop_periodic_timer 7 /* arg == NULL */
+ struct vcpu_set_periodic_timer {
+- uint64_t period_ns;
++ uint64_t period_ns;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_periodic_timer);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(vcpu_set_periodic_timer);
++typedef struct vcpu_set_periodic_timer vcpu_set_periodic_timer_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_set_periodic_timer_t);
+
+ /*
+ * Set or stop a VCPU's single-shot timer. Every VCPU has one single-shot
+ * timer which can be set via these commands.
+ */
+-#define VCPUOP_set_singleshot_timer 8 /* arg == vcpu_set_singleshot_timer_t */
++#define VCPUOP_set_singleshot_timer 8 /* arg == vcpu_set_singleshot_timer_t */
+ #define VCPUOP_stop_singleshot_timer 9 /* arg == NULL */
+ struct vcpu_set_singleshot_timer {
+- uint64_t timeout_abs_ns;
+- uint32_t flags; /* VCPU_SSHOTTMR_??? */
++ uint64_t timeout_abs_ns; /* Absolute system time value in nanoseconds. */
++ uint32_t flags; /* VCPU_SSHOTTMR_??? */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_singleshot_timer);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(vcpu_set_singleshot_timer);
++typedef struct vcpu_set_singleshot_timer vcpu_set_singleshot_timer_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_set_singleshot_timer_t);
+
+ /* Flags to VCPUOP_set_singleshot_timer. */
+ /* Require the timeout to be in the future (return -ETIME if it's passed). */
+@@ -161,13 +170,48 @@ DEFINE_GUEST_HANDLE_STRUCT(vcpu_set_singleshot_timer);
+ * structure in a convenient place, such as in a per-cpu data area.
+ * The pointer need not be page aligned, but the structure must not
+ * cross a page boundary.
++ *
++ * This may be called only once per vcpu.
+ */
+-#define VCPUOP_register_vcpu_info 10 /* arg == struct vcpu_info */
++#define VCPUOP_register_vcpu_info 10 /* arg == vcpu_register_vcpu_info_t */
+ struct vcpu_register_vcpu_info {
+ uint64_t mfn; /* mfn of page to place vcpu_info */
+ uint32_t offset; /* offset within page */
+ uint32_t rsvd; /* unused */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(vcpu_register_vcpu_info);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(vcpu_register_vcpu_info);
++typedef struct vcpu_register_vcpu_info vcpu_register_vcpu_info_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_register_vcpu_info_t);
++
++/* Send an NMI to the specified VCPU. @extra_arg == NULL. */
++#define VCPUOP_send_nmi 11
++
++/*
++ * Get the physical ID information for a pinned vcpu's underlying physical
++ * processor. The physical ID informmation is architecture-specific.
++ * On x86: id[31:0]=apic_id, id[63:32]=acpi_id, and all values 0xff and
++ * greater are reserved.
++ * This command returns -EINVAL if it is not a valid operation for this VCPU.
++ */
++#define VCPUOP_get_physid 12 /* arg == vcpu_get_physid_t */
++struct vcpu_get_physid {
++ uint64_t phys_id;
++};
++typedef struct vcpu_get_physid vcpu_get_physid_t;
++DEFINE_XEN_GUEST_HANDLE(vcpu_get_physid_t);
++#define xen_vcpu_physid_to_x86_apicid(physid) \
++ ((((uint32_t)(physid)) >= 0xff) ? 0xff : ((uint8_t)(physid)))
++#define xen_vcpu_physid_to_x86_acpiid(physid) \
++ ((((uint32_t)((physid)>>32)) >= 0xff) ? 0xff : ((uint8_t)((physid)>>32)))
+
+ #endif /* __XEN_PUBLIC_VCPU_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/version.h b/include/xen/interface/version.h
+index 453235e..30fe21c 100644
+--- a/include/xen/interface/version.h
++++ b/include/xen/interface/version.h
+@@ -3,6 +3,24 @@
+ *
+ * Xen version, type, and compile information.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2005, Nguyen Anh Quynh <aquynh at gmail.com>
+ * Copyright (c) 2005, Keir Fraser <keir at xensource.com>
+ */
+@@ -10,17 +28,18 @@
+ #ifndef __XEN_PUBLIC_VERSION_H__
+ #define __XEN_PUBLIC_VERSION_H__
+
+-/* NB. All ops return zero on success, except XENVER_version. */
++/* NB. All ops return zero on success, except XENVER_{version,pagesize} */
+
+ /* arg == NULL; returns major:minor (16:16). */
+ #define XENVER_version 0
+
+ /* arg == xen_extraversion_t. */
+ #define XENVER_extraversion 1
++typedef char xen_extraversion_t[16];
+ struct xen_extraversion {
+- char extraversion[16];
++ xen_extraversion_t extraversion;
+ };
+-#define XEN_EXTRAVERSION_LEN (sizeof(struct xen_extraversion))
++#define XEN_EXTRAVERSION_LEN (sizeof(xen_extraversion_t))
+
+ /* arg == xen_compile_info_t. */
+ #define XENVER_compile_info 2
+@@ -30,31 +49,52 @@ struct xen_compile_info {
+ char compile_domain[32];
+ char compile_date[32];
+ };
++typedef struct xen_compile_info xen_compile_info_t;
+
+ #define XENVER_capabilities 3
++typedef char xen_capabilities_info_t[1024];
+ struct xen_capabilities_info {
+- char info[1024];
++ xen_capabilities_info_t info;
+ };
+-#define XEN_CAPABILITIES_INFO_LEN (sizeof(struct xen_capabilities_info))
++#define XEN_CAPABILITIES_INFO_LEN (sizeof(xen_capabilities_info_t))
+
+ #define XENVER_changeset 4
++typedef char xen_changeset_info_t[64];
+ struct xen_changeset_info {
+- char info[64];
++ xen_changeset_info_t info;
+ };
+-#define XEN_CHANGESET_INFO_LEN (sizeof(struct xen_changeset_info))
++#define XEN_CHANGESET_INFO_LEN (sizeof(xen_changeset_info_t))
+
+ #define XENVER_platform_parameters 5
+ struct xen_platform_parameters {
+ unsigned long virt_start;
+ };
++typedef struct xen_platform_parameters xen_platform_parameters_t;
+
+ #define XENVER_get_features 6
+ struct xen_feature_info {
+ unsigned int submap_idx; /* IN: which 32-bit submap to return */
+ uint32_t submap; /* OUT: 32-bit submap */
+ };
++typedef struct xen_feature_info xen_feature_info_t;
+
+ /* Declares the features reported by XENVER_get_features. */
+ #include "features.h"
+
++/* arg == NULL; returns host memory page size. */
++#define XENVER_pagesize 7
++
++/* arg == xen_domain_handle_t. */
++#define XENVER_guest_handle 8
++
+ #endif /* __XEN_PUBLIC_VERSION_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/xen-compat.h b/include/xen/interface/xen-compat.h
+new file mode 100644
+index 0000000..c289499
+--- /dev/null
++++ b/include/xen/interface/xen-compat.h
+@@ -0,0 +1,44 @@
++/******************************************************************************
++ * xen-compat.h
++ *
++ * Guest OS interface to Xen. Compatibility layer.
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (c) 2006, Christian Limpach
++ */
++
++#ifndef __XEN_PUBLIC_XEN_COMPAT_H__
++#define __XEN_PUBLIC_XEN_COMPAT_H__
++
++#define __XEN_LATEST_INTERFACE_VERSION__ 0x00030208
++
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++/* Xen is built with matching headers and implements the latest interface. */
++#define __XEN_INTERFACE_VERSION__ __XEN_LATEST_INTERFACE_VERSION__
++#elif !defined(__XEN_INTERFACE_VERSION__)
++/* Guests which do not specify a version get the legacy interface. */
++#define __XEN_INTERFACE_VERSION__ 0x00000000
++#endif
++
++#if __XEN_INTERFACE_VERSION__ > __XEN_LATEST_INTERFACE_VERSION__
++#error "These header files do not support the requested interface version."
++#endif
++
++#endif /* __XEN_PUBLIC_XEN_COMPAT_H__ */
+diff --git a/include/xen/interface/xen.h b/include/xen/interface/xen.h
+index 819a033..da1346c 100644
+--- a/include/xen/interface/xen.h
++++ b/include/xen/interface/xen.h
+@@ -3,35 +3,70 @@
+ *
+ * Guest OS interface to Xen.
+ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
+ * Copyright (c) 2004, K A Fraser
+ */
+
+ #ifndef __XEN_PUBLIC_XEN_H__
+ #define __XEN_PUBLIC_XEN_H__
+
+-#include <asm/xen/interface.h>
++#include "xen-compat.h"
++#ifdef CONFIG_PARAVIRT_XEN
+ #include <asm/pvclock-abi.h>
++#endif
+
+-/*
+- * XEN "SYSTEM CALLS" (a.k.a. HYPERCALLS).
+- */
++#if defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H)
++#include <asm/xen/interface.h>
++#elif defined(__i386__) || defined(__x86_64__)
++#include "arch-x86/xen.h"
++#elif defined(__ia64__)
++#include "arch-ia64.h"
++#else
++#error "Unsupported architecture"
++#endif
++
++#ifndef __ASSEMBLY__
++/* Guest handles for primitive C types. */
++DEFINE_XEN_GUEST_HANDLE(char);
++__DEFINE_XEN_GUEST_HANDLE(uchar, unsigned char);
++DEFINE_XEN_GUEST_HANDLE(int);
++__DEFINE_XEN_GUEST_HANDLE(uint, unsigned int);
++DEFINE_XEN_GUEST_HANDLE(long);
++__DEFINE_XEN_GUEST_HANDLE(ulong, unsigned long);
++DEFINE_XEN_GUEST_HANDLE(void);
++
++DEFINE_XEN_GUEST_HANDLE(xen_pfn_t);
++#endif
+
+ /*
+- * x86_32: EAX = vector; EBX, ECX, EDX, ESI, EDI = args 1, 2, 3, 4, 5.
+- * EAX = return value
+- * (argument registers may be clobbered on return)
+- * x86_64: RAX = vector; RDI, RSI, RDX, R10, R8, R9 = args 1, 2, 3, 4, 5, 6.
+- * RAX = return value
+- * (argument registers not clobbered on return; RCX, R11 are)
++ * HYPERCALLS
+ */
++
+ #define __HYPERVISOR_set_trap_table 0
+ #define __HYPERVISOR_mmu_update 1
+ #define __HYPERVISOR_set_gdt 2
+ #define __HYPERVISOR_stack_switch 3
+ #define __HYPERVISOR_set_callbacks 4
+ #define __HYPERVISOR_fpu_taskswitch 5
+-#define __HYPERVISOR_sched_op 6
+-#define __HYPERVISOR_dom0_op 7
++#define __HYPERVISOR_sched_op_compat 6 /* compat since 0x00030101 */
++#define __HYPERVISOR_platform_op 7
+ #define __HYPERVISOR_set_debugreg 8
+ #define __HYPERVISOR_get_debugreg 9
+ #define __HYPERVISOR_update_descriptor 10
+@@ -39,10 +74,10 @@
+ #define __HYPERVISOR_multicall 13
+ #define __HYPERVISOR_update_va_mapping 14
+ #define __HYPERVISOR_set_timer_op 15
+-#define __HYPERVISOR_event_channel_op_compat 16
++#define __HYPERVISOR_event_channel_op_compat 16 /* compat since 0x00030202 */
+ #define __HYPERVISOR_xen_version 17
+ #define __HYPERVISOR_console_io 18
+-#define __HYPERVISOR_physdev_op_compat 19
++#define __HYPERVISOR_physdev_op_compat 19 /* compat since 0x00030202 */
+ #define __HYPERVISOR_grant_table_op 20
+ #define __HYPERVISOR_vm_assist 21
+ #define __HYPERVISOR_update_va_mapping_otherdomain 22
+@@ -50,14 +85,17 @@
+ #define __HYPERVISOR_vcpu_op 24
+ #define __HYPERVISOR_set_segment_base 25 /* x86/64 only */
+ #define __HYPERVISOR_mmuext_op 26
+-#define __HYPERVISOR_acm_op 27
++#define __HYPERVISOR_xsm_op 27
+ #define __HYPERVISOR_nmi_op 28
+-#define __HYPERVISOR_sched_op_new 29
++#define __HYPERVISOR_sched_op 29
+ #define __HYPERVISOR_callback_op 30
+ #define __HYPERVISOR_xenoprof_op 31
+ #define __HYPERVISOR_event_channel_op 32
+ #define __HYPERVISOR_physdev_op 33
+ #define __HYPERVISOR_hvm_op 34
++#define __HYPERVISOR_sysctl 35
++#define __HYPERVISOR_domctl 36
++#define __HYPERVISOR_kexec_op 37
+
+ /* Architecture-specific hypercall definitions. */
+ #define __HYPERVISOR_arch_0 48
+@@ -70,15 +108,46 @@
+ #define __HYPERVISOR_arch_7 55
+
+ /*
++ * HYPERCALL COMPATIBILITY.
++ */
++
++/* New sched_op hypercall introduced in 0x00030101. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030101 || (defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H))
++#undef __HYPERVISOR_sched_op
++#define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_compat
++#endif
++
++/* New event-channel and physdev hypercalls introduced in 0x00030202. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030202
++#undef __HYPERVISOR_event_channel_op
++#define __HYPERVISOR_event_channel_op __HYPERVISOR_event_channel_op_compat
++#undef __HYPERVISOR_physdev_op
++#define __HYPERVISOR_physdev_op __HYPERVISOR_physdev_op_compat
++#endif
++
++/* New platform_op hypercall introduced in 0x00030204. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030204 || (defined(CONFIG_PARAVIRT_XEN) && !defined(HAVE_XEN_PLATFORM_COMPAT_H))
++#define __HYPERVISOR_dom0_op __HYPERVISOR_platform_op
++#endif
++
++/*
+ * VIRTUAL INTERRUPTS
+ *
+ * Virtual interrupts that a guest OS may receive from Xen.
++ *
++ * In the side comments, 'V.' denotes a per-VCPU VIRQ while 'G.' denotes a
++ * global VIRQ. The former can be bound once per VCPU and cannot be re-bound.
++ * The latter can be allocated only once per guest: they must initially be
++ * allocated to VCPU0 but can subsequently be re-bound.
+ */
+-#define VIRQ_TIMER 0 /* Timebase update, and/or requested timeout. */
+-#define VIRQ_DEBUG 1 /* Request guest to dump debug info. */
+-#define VIRQ_CONSOLE 2 /* (DOM0) Bytes received on emergency console. */
+-#define VIRQ_DOM_EXC 3 /* (DOM0) Exceptional event for some domain. */
+-#define VIRQ_DEBUGGER 6 /* (DOM0) A domain has paused for debugging. */
++#define VIRQ_TIMER 0 /* V. Timebase update, and/or requested timeout. */
++#define VIRQ_DEBUG 1 /* V. Request guest to dump debug info. */
++#define VIRQ_CONSOLE 2 /* G. (DOM0) Bytes received on emergency console. */
++#define VIRQ_DOM_EXC 3 /* G. (DOM0) Exceptional event for some domain. */
++#define VIRQ_TBUF 4 /* G. (DOM0) Trace buffer has records available. */
++#define VIRQ_DEBUGGER 6 /* G. (DOM0) A domain has paused for debugging. */
++#define VIRQ_XENOPROF 7 /* V. XenOprofile interrupt: new sample available */
++#define VIRQ_CON_RING 8 /* G. (DOM0) Bytes received on console */
+
+ /* Architecture-specific VIRQ definitions. */
+ #define VIRQ_ARCH_0 16
+@@ -91,6 +160,7 @@
+ #define VIRQ_ARCH_7 23
+
+ #define NR_VIRQS 24
++
+ /*
+ * MMU-UPDATE REQUESTS
+ *
+@@ -114,9 +184,14 @@
+ * ptr[:2] -- Machine address within the frame whose mapping to modify.
+ * The frame must belong to the FD, if one is specified.
+ * val -- Value to write into the mapping entry.
++ *
++ * ptr[1:0] == MMU_PT_UPDATE_PRESERVE_AD:
++ * As MMU_NORMAL_PT_UPDATE above, but A/D bits currently in the PTE are ORed
++ * with those in @val.
+ */
+-#define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */
+-#define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */
++#define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */
++#define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */
++#define MMU_PT_UPDATE_PRESERVE_AD 2 /* atomically: *ptr = val | (*ptr&(A|D)) */
+
+ /*
+ * MMU EXTENDED OPERATIONS
+@@ -180,21 +255,27 @@
+
+ #ifndef __ASSEMBLY__
+ struct mmuext_op {
+- unsigned int cmd;
+- union {
+- /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR */
+- unsigned long mfn;
+- /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */
+- unsigned long linear_addr;
+- } arg1;
+- union {
+- /* SET_LDT */
+- unsigned int nr_ents;
+- /* TLB_FLUSH_MULTI, INVLPG_MULTI */
+- void *vcpumask;
+- } arg2;
++ unsigned int cmd;
++ union {
++ /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR */
++ xen_pfn_t mfn;
++ /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */
++ unsigned long linear_addr;
++ } arg1;
++ union {
++ /* SET_LDT */
++ unsigned int nr_ents;
++ /* TLB_FLUSH_MULTI, INVLPG_MULTI */
++#if __XEN_INTERFACE_VERSION__ >= 0x00030205
++ XEN_GUEST_HANDLE(void) vcpumask;
++#else
++ void *vcpumask;
++#endif
++ } arg2;
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(mmuext_op);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(mmuext_op);
++typedef struct mmuext_op mmuext_op_t;
++DEFINE_XEN_GUEST_HANDLE(mmuext_op_t);
+ #endif
+
+ /* These are passed as 'flags' to update_va_mapping. They can be ORed. */
+@@ -219,11 +300,24 @@ DEFINE_GUEST_HANDLE_STRUCT(mmuext_op);
+ */
+ #define VMASST_CMD_enable 0
+ #define VMASST_CMD_disable 1
++
++/* x86/32 guests: simulate full 4GB segment limits. */
+ #define VMASST_TYPE_4gb_segments 0
++
++/* x86/32 guests: trap (vector 15) whenever above vmassist is used. */
+ #define VMASST_TYPE_4gb_segments_notify 1
++
++/*
++ * x86 guests: support writes to bottom-level PTEs.
++ * NB1. Page-directory entries cannot be written.
++ * NB2. Guest must continue to remove all writable mappings of PTEs.
++ */
+ #define VMASST_TYPE_writable_pagetables 2
++
++/* x86/PAE guests: support PDPTs above 4GB. */
+ #define VMASST_TYPE_pae_extended_cr3 3
+-#define MAX_VMASST_TYPE 3
++
++#define MAX_VMASST_TYPE 3
+
+ #ifndef __ASSEMBLY__
+
+@@ -262,7 +356,9 @@ struct mmu_update {
+ uint64_t ptr; /* Machine address of PTE. */
+ uint64_t val; /* New contents of PTE. */
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(mmu_update);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(mmu_update);
++typedef struct mmu_update mmu_update_t;
++DEFINE_XEN_GUEST_HANDLE(mmu_update_t);
+
+ /*
+ * Send an array of these to HYPERVISOR_multicall().
+@@ -270,10 +366,16 @@ DEFINE_GUEST_HANDLE_STRUCT(mmu_update);
+ */
+ struct multicall_entry {
+ unsigned long op;
++#if !defined(CONFIG_PARAVIRT_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H)
++ unsigned long result;
++#else
+ long result;
++#endif
+ unsigned long args[6];
+ };
+-DEFINE_GUEST_HANDLE_STRUCT(multicall_entry);
++DEFINE_XEN_GUEST_HANDLE_STRUCT(multicall_entry);
++typedef struct multicall_entry multicall_entry_t;
++DEFINE_XEN_GUEST_HANDLE(multicall_entry_t);
+
+ /*
+ * Event channel endpoints per domain:
+@@ -282,173 +384,240 @@ DEFINE_GUEST_HANDLE_STRUCT(multicall_entry);
+ #define NR_EVENT_CHANNELS (sizeof(unsigned long) * sizeof(unsigned long) * 64)
+
+ struct vcpu_time_info {
+- /*
+- * Updates to the following values are preceded and followed
+- * by an increment of 'version'. The guest can therefore
+- * detect updates by looking for changes to 'version'. If the
+- * least-significant bit of the version number is set then an
+- * update is in progress and the guest must wait to read a
+- * consistent set of values. The correct way to interact with
+- * the version number is similar to Linux's seqlock: see the
+- * implementations of read_seqbegin/read_seqretry.
+- */
+- uint32_t version;
+- uint32_t pad0;
+- uint64_t tsc_timestamp; /* TSC at last update of time vals. */
+- uint64_t system_time; /* Time, in nanosecs, since boot. */
+- /*
+- * Current system time:
+- * system_time + ((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul
+- * CPU frequency (Hz):
+- * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift
+- */
+- uint32_t tsc_to_system_mul;
+- int8_t tsc_shift;
+- int8_t pad1[3];
++ /*
++ * Updates to the following values are preceded and followed by an
++ * increment of 'version'. The guest can therefore detect updates by
++ * looking for changes to 'version'. If the least-significant bit of
++ * the version number is set then an update is in progress and the guest
++ * must wait to read a consistent set of values.
++ * The correct way to interact with the version number is similar to
++ * Linux's seqlock: see the implementations of read_seqbegin/read_seqretry.
++ */
++ uint32_t version;
++ uint32_t pad0;
++ uint64_t tsc_timestamp; /* TSC at last update of time vals. */
++ uint64_t system_time; /* Time, in nanosecs, since boot. */
++ /*
++ * Current system time:
++ * system_time +
++ * ((((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul) >> 32)
++ * CPU frequency (Hz):
++ * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift
++ */
++ uint32_t tsc_to_system_mul;
++ int8_t tsc_shift;
++ int8_t pad1[3];
+ }; /* 32 bytes */
++typedef struct vcpu_time_info vcpu_time_info_t;
+
+ struct vcpu_info {
+- /*
+- * 'evtchn_upcall_pending' is written non-zero by Xen to indicate
+- * a pending notification for a particular VCPU. It is then cleared
+- * by the guest OS /before/ checking for pending work, thus avoiding
+- * a set-and-check race. Note that the mask is only accessed by Xen
+- * on the CPU that is currently hosting the VCPU. This means that the
+- * pending and mask flags can be updated by the guest without special
+- * synchronisation (i.e., no need for the x86 LOCK prefix).
+- * This may seem suboptimal because if the pending flag is set by
+- * a different CPU then an IPI may be scheduled even when the mask
+- * is set. However, note:
+- * 1. The task of 'interrupt holdoff' is covered by the per-event-
+- * channel mask bits. A 'noisy' event that is continually being
+- * triggered can be masked at source at this very precise
+- * granularity.
+- * 2. The main purpose of the per-VCPU mask is therefore to restrict
+- * reentrant execution: whether for concurrency control, or to
+- * prevent unbounded stack usage. Whatever the purpose, we expect
+- * that the mask will be asserted only for short periods at a time,
+- * and so the likelihood of a 'spurious' IPI is suitably small.
+- * The mask is read before making an event upcall to the guest: a
+- * non-zero mask therefore guarantees that the VCPU will not receive
+- * an upcall activation. The mask is cleared when the VCPU requests
+- * to block: this avoids wakeup-waiting races.
+- */
+- uint8_t evtchn_upcall_pending;
+- uint8_t evtchn_upcall_mask;
+- unsigned long evtchn_pending_sel;
+- struct arch_vcpu_info arch;
+- struct pvclock_vcpu_time_info time;
++ /*
++ * 'evtchn_upcall_pending' is written non-zero by Xen to indicate
++ * a pending notification for a particular VCPU. It is then cleared
++ * by the guest OS /before/ checking for pending work, thus avoiding
++ * a set-and-check race. Note that the mask is only accessed by Xen
++ * on the CPU that is currently hosting the VCPU. This means that the
++ * pending and mask flags can be updated by the guest without special
++ * synchronisation (i.e., no need for the x86 LOCK prefix).
++ * This may seem suboptimal because if the pending flag is set by
++ * a different CPU then an IPI may be scheduled even when the mask
++ * is set. However, note:
++ * 1. The task of 'interrupt holdoff' is covered by the per-event-
++ * channel mask bits. A 'noisy' event that is continually being
++ * triggered can be masked at source at this very precise
++ * granularity.
++ * 2. The main purpose of the per-VCPU mask is therefore to restrict
++ * reentrant execution: whether for concurrency control, or to
++ * prevent unbounded stack usage. Whatever the purpose, we expect
++ * that the mask will be asserted only for short periods at a time,
++ * and so the likelihood of a 'spurious' IPI is suitably small.
++ * The mask is read before making an event upcall to the guest: a
++ * non-zero mask therefore guarantees that the VCPU will not receive
++ * an upcall activation. The mask is cleared when the VCPU requests
++ * to block: this avoids wakeup-waiting races.
++ */
++ uint8_t evtchn_upcall_pending;
++ uint8_t evtchn_upcall_mask;
++ unsigned long evtchn_pending_sel;
++ struct arch_vcpu_info arch;
++#ifdef CONFIG_PARAVIRT_XEN
++ struct pvclock_vcpu_time_info time;
++#else
++ struct vcpu_time_info time;
++#endif
+ }; /* 64 bytes (x86) */
++#ifndef __XEN__
++typedef struct vcpu_info vcpu_info_t;
++#endif
+
+ /*
+ * Xen/kernel shared data -- pointer provided in start_info.
+- * NB. We expect that this struct is smaller than a page.
++ *
++ * This structure is defined to be both smaller than a page, and the
++ * only data on the shared page, but may vary in actual size even within
++ * compatible Xen versions; guests should not rely on the size
++ * of this structure remaining constant.
+ */
+ struct shared_info {
+- struct vcpu_info vcpu_info[MAX_VIRT_CPUS];
+-
+- /*
+- * A domain can create "event channels" on which it can send and receive
+- * asynchronous event notifications. There are three classes of event that
+- * are delivered by this mechanism:
+- * 1. Bi-directional inter- and intra-domain connections. Domains must
+- * arrange out-of-band to set up a connection (usually by allocating
+- * an unbound 'listener' port and avertising that via a storage service
+- * such as xenstore).
+- * 2. Physical interrupts. A domain with suitable hardware-access
+- * privileges can bind an event-channel port to a physical interrupt
+- * source.
+- * 3. Virtual interrupts ('events'). A domain can bind an event-channel
+- * port to a virtual interrupt source, such as the virtual-timer
+- * device or the emergency console.
+- *
+- * Event channels are addressed by a "port index". Each channel is
+- * associated with two bits of information:
+- * 1. PENDING -- notifies the domain that there is a pending notification
+- * to be processed. This bit is cleared by the guest.
+- * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING
+- * will cause an asynchronous upcall to be scheduled. This bit is only
+- * updated by the guest. It is read-only within Xen. If a channel
+- * becomes pending while the channel is masked then the 'edge' is lost
+- * (i.e., when the channel is unmasked, the guest must manually handle
+- * pending notifications as no upcall will be scheduled by Xen).
+- *
+- * To expedite scanning of pending notifications, any 0->1 pending
+- * transition on an unmasked channel causes a corresponding bit in a
+- * per-vcpu selector word to be set. Each bit in the selector covers a
+- * 'C long' in the PENDING bitfield array.
+- */
+- unsigned long evtchn_pending[sizeof(unsigned long) * 8];
+- unsigned long evtchn_mask[sizeof(unsigned long) * 8];
+-
+- /*
+- * Wallclock time: updated only by control software. Guests should base
+- * their gettimeofday() syscall on this wallclock-base value.
+- */
+- struct pvclock_wall_clock wc;
+-
+- struct arch_shared_info arch;
++ struct vcpu_info vcpu_info[MAX_VIRT_CPUS];
++
++ /*
++ * A domain can create "event channels" on which it can send and receive
++ * asynchronous event notifications. There are three classes of event that
++ * are delivered by this mechanism:
++ * 1. Bi-directional inter- and intra-domain connections. Domains must
++ * arrange out-of-band to set up a connection (usually by allocating
++ * an unbound 'listener' port and avertising that via a storage service
++ * such as xenstore).
++ * 2. Physical interrupts. A domain with suitable hardware-access
++ * privileges can bind an event-channel port to a physical interrupt
++ * source.
++ * 3. Virtual interrupts ('events'). A domain can bind an event-channel
++ * port to a virtual interrupt source, such as the virtual-timer
++ * device or the emergency console.
++ *
++ * Event channels are addressed by a "port index". Each channel is
++ * associated with two bits of information:
++ * 1. PENDING -- notifies the domain that there is a pending notification
++ * to be processed. This bit is cleared by the guest.
++ * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING
++ * will cause an asynchronous upcall to be scheduled. This bit is only
++ * updated by the guest. It is read-only within Xen. If a channel
++ * becomes pending while the channel is masked then the 'edge' is lost
++ * (i.e., when the channel is unmasked, the guest must manually handle
++ * pending notifications as no upcall will be scheduled by Xen).
++ *
++ * To expedite scanning of pending notifications, any 0->1 pending
++ * transition on an unmasked channel causes a corresponding bit in a
++ * per-vcpu selector word to be set. Each bit in the selector covers a
++ * 'C long' in the PENDING bitfield array.
++ */
++ unsigned long evtchn_pending[sizeof(unsigned long) * 8];
++ unsigned long evtchn_mask[sizeof(unsigned long) * 8];
++
++ /*
++ * Wallclock time: updated only by control software. Guests should base
++ * their gettimeofday() syscall on this wallclock-base value.
++ */
++#ifdef CONFIG_PARAVIRT_XEN
++ struct pvclock_wall_clock wc;
++#else
++ uint32_t wc_version; /* Version counter: see vcpu_time_info_t. */
++ uint32_t wc_sec; /* Secs 00:00:00 UTC, Jan 1, 1970. */
++ uint32_t wc_nsec; /* Nsecs 00:00:00 UTC, Jan 1, 1970. */
++#endif
++
++ struct arch_shared_info arch;
+
+ };
++#ifndef __XEN__
++typedef struct shared_info shared_info_t;
++#endif
+
+ /*
+- * Start-of-day memory layout for the initial domain (DOM0):
++ * Start-of-day memory layout:
+ * 1. The domain is started within contiguous virtual-memory region.
+- * 2. The contiguous region begins and ends on an aligned 4MB boundary.
+- * 3. The region start corresponds to the load address of the OS image.
+- * If the load address is not 4MB aligned then the address is rounded down.
+- * 4. This the order of bootstrap elements in the initial virtual region:
++ * 2. The contiguous region ends on an aligned 4MB boundary.
++ * 3. This the order of bootstrap elements in the initial virtual region:
+ * a. relocated kernel image
+ * b. initial ram disk [mod_start, mod_len]
+ * c. list of allocated page frames [mfn_list, nr_pages]
+ * d. start_info_t structure [register ESI (x86)]
+ * e. bootstrap page tables [pt_base, CR3 (x86)]
+ * f. bootstrap stack [register ESP (x86)]
+- * 5. Bootstrap elements are packed together, but each is 4kB-aligned.
+- * 6. The initial ram disk may be omitted.
+- * 7. The list of page frames forms a contiguous 'pseudo-physical' memory
++ * 4. Bootstrap elements are packed together, but each is 4kB-aligned.
++ * 5. The initial ram disk may be omitted.
++ * 6. The list of page frames forms a contiguous 'pseudo-physical' memory
+ * layout for the domain. In particular, the bootstrap virtual-memory
+ * region is a 1:1 mapping to the first section of the pseudo-physical map.
+- * 8. All bootstrap elements are mapped read-writable for the guest OS. The
++ * 7. All bootstrap elements are mapped read-writable for the guest OS. The
+ * only exception is the bootstrap page table, which is mapped read-only.
+- * 9. There is guaranteed to be at least 512kB padding after the final
++ * 8. There is guaranteed to be at least 512kB padding after the final
+ * bootstrap element. If necessary, the bootstrap virtual region is
+ * extended by an extra 4MB to ensure this.
+ */
+
+ #define MAX_GUEST_CMDLINE 1024
+ struct start_info {
+- /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */
+- char magic[32]; /* "xen-<version>-<platform>". */
+- unsigned long nr_pages; /* Total pages allocated to this domain. */
+- unsigned long shared_info; /* MACHINE address of shared info struct. */
+- uint32_t flags; /* SIF_xxx flags. */
+- unsigned long store_mfn; /* MACHINE page number of shared page. */
+- uint32_t store_evtchn; /* Event channel for store communication. */
+- union {
+- struct {
+- unsigned long mfn; /* MACHINE page number of console page. */
+- uint32_t evtchn; /* Event channel for console page. */
+- } domU;
+- struct {
+- uint32_t info_off; /* Offset of console_info struct. */
+- uint32_t info_size; /* Size of console_info struct from start.*/
+- } dom0;
+- } console;
+- /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */
+- unsigned long pt_base; /* VIRTUAL address of page directory. */
+- unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */
+- unsigned long mfn_list; /* VIRTUAL address of page-frame list. */
+- unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */
+- unsigned long mod_len; /* Size (bytes) of pre-loaded module. */
+- int8_t cmd_line[MAX_GUEST_CMDLINE];
++ /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */
++ char magic[32]; /* "xen-<version>-<platform>". */
++ unsigned long nr_pages; /* Total pages allocated to this domain. */
++ unsigned long shared_info; /* MACHINE address of shared info struct. */
++ uint32_t flags; /* SIF_xxx flags. */
++ xen_pfn_t store_mfn; /* MACHINE page number of shared page. */
++ uint32_t store_evtchn; /* Event channel for store communication. */
++ union {
++ struct {
++ xen_pfn_t mfn; /* MACHINE page number of console page. */
++ uint32_t evtchn; /* Event channel for console page. */
++ } domU;
++ struct {
++ uint32_t info_off; /* Offset of console_info struct. */
++ uint32_t info_size; /* Size of console_info struct from start.*/
++ } dom0;
++ } console;
++ /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */
++ unsigned long pt_base; /* VIRTUAL address of page directory. */
++ unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */
++ unsigned long mfn_list; /* VIRTUAL address of page-frame list. */
++ unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */
++ unsigned long mod_len; /* Size (bytes) of pre-loaded module. */
++ int8_t cmd_line[MAX_GUEST_CMDLINE];
+ };
++typedef struct start_info start_info_t;
++
++/* New console union for dom0 introduced in 0x00030203. */
++#if __XEN_INTERFACE_VERSION__ < 0x00030203
++#define console_mfn console.domU.mfn
++#define console_evtchn console.domU.evtchn
++#endif
+
+ /* These flags are passed in the 'flags' field of start_info_t. */
+ #define SIF_PRIVILEGED (1<<0) /* Is the domain privileged? */
+ #define SIF_INITDOMAIN (1<<1) /* Is this the initial control domain? */
+-
+-typedef uint64_t cpumap_t;
++#define SIF_PM_MASK (0xFF<<8) /* reserve 1 byte for xen-pm options */
++
++typedef struct dom0_vga_console_info {
++ uint8_t video_type; /* DOM0_VGA_CONSOLE_??? */
++#define XEN_VGATYPE_TEXT_MODE_3 0x03
++#define XEN_VGATYPE_VESA_LFB 0x23
++
++ union {
++ struct {
++ /* Font height, in pixels. */
++ uint16_t font_height;
++ /* Cursor location (column, row). */
++ uint16_t cursor_x, cursor_y;
++ /* Number of rows and columns (dimensions in characters). */
++ uint16_t rows, columns;
++ } text_mode_3;
++
++ struct {
++ /* Width and height, in pixels. */
++ uint16_t width, height;
++ /* Bytes per scan line. */
++ uint16_t bytes_per_line;
++ /* Bits per pixel. */
++ uint16_t bits_per_pixel;
++ /* LFB physical address, and size (in units of 64kB). */
++ uint32_t lfb_base;
++ uint32_t lfb_size;
++ /* RGB mask offsets and sizes, as defined by VBE 1.2+ */
++ uint8_t red_pos, red_size;
++ uint8_t green_pos, green_size;
++ uint8_t blue_pos, blue_size;
++ uint8_t rsvd_pos, rsvd_size;
++#if __XEN_INTERFACE_VERSION__ >= 0x00030206
++ /* VESA capabilities (offset 0xa, VESA command 0x4f00). */
++ uint32_t gbl_caps;
++ /* Mode attributes (offset 0x0, VESA command 0x4f01). */
++ uint16_t mode_attrs;
++#endif
++ } vesa_lfb;
++ } u;
++} dom0_vga_console_info_t;
++#define xen_vga_console_info dom0_vga_console_info
++#define xen_vga_console_info_t dom0_vga_console_info_t
+
+ typedef uint8_t xen_domain_handle_t[16];
+
+@@ -456,6 +625,11 @@ typedef uint8_t xen_domain_handle_t[16];
+ #define __mk_unsigned_long(x) x ## UL
+ #define mk_unsigned_long(x) __mk_unsigned_long(x)
+
++__DEFINE_XEN_GUEST_HANDLE(uint8, uint8_t);
++__DEFINE_XEN_GUEST_HANDLE(uint16, uint16_t);
++__DEFINE_XEN_GUEST_HANDLE(uint32, uint32_t);
++__DEFINE_XEN_GUEST_HANDLE(uint64, uint64_t);
++
+ #else /* __ASSEMBLY__ */
+
+ /* In assembly code we cannot use C numeric constant suffixes. */
+@@ -463,4 +637,24 @@ typedef uint8_t xen_domain_handle_t[16];
+
+ #endif /* !__ASSEMBLY__ */
+
++/* Default definitions for macros used by domctl/sysctl. */
++#if defined(__XEN__) || defined(__XEN_TOOLS__)
++#ifndef uint64_aligned_t
++#define uint64_aligned_t uint64_t
++#endif
++#ifndef XEN_GUEST_HANDLE_64
++#define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name)
++#endif
++#endif
++
+ #endif /* __XEN_PUBLIC_XEN_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/xenoprof.h b/include/xen/interface/xenoprof.h
+new file mode 100644
+index 0000000..183078d
+--- /dev/null
++++ b/include/xen/interface/xenoprof.h
+@@ -0,0 +1,138 @@
++/******************************************************************************
++ * xenoprof.h
++ *
++ * Interface for enabling system wide profiling based on hardware performance
++ * counters
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Copyright (C) 2005 Hewlett-Packard Co.
++ * Written by Aravind Menon & Jose Renato Santos
++ */
++
++#ifndef __XEN_PUBLIC_XENOPROF_H__
++#define __XEN_PUBLIC_XENOPROF_H__
++
++#include "xen.h"
++
++/*
++ * Commands to HYPERVISOR_xenoprof_op().
++ */
++#define XENOPROF_init 0
++#define XENOPROF_reset_active_list 1
++#define XENOPROF_reset_passive_list 2
++#define XENOPROF_set_active 3
++#define XENOPROF_set_passive 4
++#define XENOPROF_reserve_counters 5
++#define XENOPROF_counter 6
++#define XENOPROF_setup_events 7
++#define XENOPROF_enable_virq 8
++#define XENOPROF_start 9
++#define XENOPROF_stop 10
++#define XENOPROF_disable_virq 11
++#define XENOPROF_release_counters 12
++#define XENOPROF_shutdown 13
++#define XENOPROF_get_buffer 14
++#define XENOPROF_set_backtrace 15
++#define XENOPROF_last_op 15
++
++#define MAX_OPROF_EVENTS 32
++#define MAX_OPROF_DOMAINS 25
++#define XENOPROF_CPU_TYPE_SIZE 64
++
++/* Xenoprof performance events (not Xen events) */
++struct event_log {
++ uint64_t eip;
++ uint8_t mode;
++ uint8_t event;
++};
++
++/* PC value that indicates a special code */
++#define XENOPROF_ESCAPE_CODE ~0UL
++/* Transient events for the xenoprof->oprofile cpu buf */
++#define XENOPROF_TRACE_BEGIN 1
++
++/* Xenoprof buffer shared between Xen and domain - 1 per VCPU */
++struct xenoprof_buf {
++ uint32_t event_head;
++ uint32_t event_tail;
++ uint32_t event_size;
++ uint32_t vcpu_id;
++ uint64_t xen_samples;
++ uint64_t kernel_samples;
++ uint64_t user_samples;
++ uint64_t lost_samples;
++ struct event_log event_log[1];
++};
++#ifndef __XEN__
++typedef struct xenoprof_buf xenoprof_buf_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_buf_t);
++#endif
++
++struct xenoprof_init {
++ int32_t num_events;
++ int32_t is_primary;
++ char cpu_type[XENOPROF_CPU_TYPE_SIZE];
++};
++typedef struct xenoprof_init xenoprof_init_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_init_t);
++
++struct xenoprof_get_buffer {
++ int32_t max_samples;
++ int32_t nbuf;
++ int32_t bufsize;
++ uint64_t buf_gmaddr;
++};
++typedef struct xenoprof_get_buffer xenoprof_get_buffer_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_get_buffer_t);
++
++struct xenoprof_counter {
++ uint32_t ind;
++ uint64_t count;
++ uint32_t enabled;
++ uint32_t event;
++ uint32_t hypervisor;
++ uint32_t kernel;
++ uint32_t user;
++ uint64_t unit_mask;
++};
++typedef struct xenoprof_counter xenoprof_counter_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_counter_t);
++
++typedef struct xenoprof_passive {
++ uint16_t domain_id;
++ int32_t max_samples;
++ int32_t nbuf;
++ int32_t bufsize;
++ uint64_t buf_gmaddr;
++} xenoprof_passive_t;
++DEFINE_XEN_GUEST_HANDLE(xenoprof_passive_t);
++
++
++#endif /* __XEN_PUBLIC_XENOPROF_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/xsm/acm.h b/include/xen/interface/xsm/acm.h
+new file mode 100644
+index 0000000..4417c30
+--- /dev/null
++++ b/include/xen/interface/xsm/acm.h
+@@ -0,0 +1,235 @@
++/*
++ * acm.h: Xen access control module interface defintions
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005, International Business Machines Corporation.
++ */
++
++#ifndef _XEN_PUBLIC_ACM_H
++#define _XEN_PUBLIC_ACM_H
++
++#include "../xen.h"
++
++/* if ACM_DEBUG defined, all hooks should
++ * print a short trace message (comment it out
++ * when not in testing mode )
++ */
++/* #define ACM_DEBUG */
++
++#ifdef ACM_DEBUG
++# define printkd(fmt, args...) printk(fmt,## args)
++#else
++# define printkd(fmt, args...)
++#endif
++
++/* default ssid reference value if not supplied */
++#define ACM_DEFAULT_SSID 0x0
++#define ACM_DEFAULT_LOCAL_SSID 0x0
++
++/* Internal ACM ERROR types */
++#define ACM_OK 0
++#define ACM_UNDEF -1
++#define ACM_INIT_SSID_ERROR -2
++#define ACM_INIT_SOID_ERROR -3
++#define ACM_ERROR -4
++
++/* External ACCESS DECISIONS */
++#define ACM_ACCESS_PERMITTED 0
++#define ACM_ACCESS_DENIED -111
++#define ACM_NULL_POINTER_ERROR -200
++
++/*
++ Error codes reported in when trying to test for a new policy
++ These error codes are reported in an array of tuples where
++ each error code is followed by a parameter describing the error
++ more closely, such as a domain id.
++*/
++#define ACM_EVTCHN_SHARING_VIOLATION 0x100
++#define ACM_GNTTAB_SHARING_VIOLATION 0x101
++#define ACM_DOMAIN_LOOKUP 0x102
++#define ACM_CHWALL_CONFLICT 0x103
++#define ACM_SSIDREF_IN_USE 0x104
++
++
++/* primary policy in lower 4 bits */
++#define ACM_NULL_POLICY 0
++#define ACM_CHINESE_WALL_POLICY 1
++#define ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY 2
++#define ACM_POLICY_UNDEFINED 15
++
++/* combinations have secondary policy component in higher 4bit */
++#define ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY \
++ ((ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY << 4) | ACM_CHINESE_WALL_POLICY)
++
++/* policy: */
++#define ACM_POLICY_NAME(X) \
++ ((X) == (ACM_NULL_POLICY)) ? "NULL" : \
++ ((X) == (ACM_CHINESE_WALL_POLICY)) ? "CHINESE WALL" : \
++ ((X) == (ACM_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "SIMPLE TYPE ENFORCEMENT" : \
++ ((X) == (ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)) ? "CHINESE WALL AND SIMPLE TYPE ENFORCEMENT" : \
++ "UNDEFINED"
++
++/* the following policy versions must be increased
++ * whenever the interpretation of the related
++ * policy's data structure changes
++ */
++#define ACM_POLICY_VERSION 4
++#define ACM_CHWALL_VERSION 1
++#define ACM_STE_VERSION 1
++
++/* defines a ssid reference used by xen */
++typedef uint32_t ssidref_t;
++
++/* hooks that are known to domains */
++#define ACMHOOK_none 0
++#define ACMHOOK_sharing 1
++#define ACMHOOK_authorization 2
++#define ACMHOOK_conflictset 3
++
++/* -------security policy relevant type definitions-------- */
++
++/* type identifier; compares to "equal" or "not equal" */
++typedef uint16_t domaintype_t;
++
++/* CHINESE WALL POLICY DATA STRUCTURES
++ *
++ * current accumulated conflict type set:
++ * When a domain is started and has a type that is in
++ * a conflict set, the conflicting types are incremented in
++ * the aggregate set. When a domain is destroyed, the
++ * conflicting types to its type are decremented.
++ * If a domain has multiple types, this procedure works over
++ * all those types.
++ *
++ * conflict_aggregate_set[i] holds the number of
++ * running domains that have a conflict with type i.
++ *
++ * running_types[i] holds the number of running domains
++ * that include type i in their ssidref-referenced type set
++ *
++ * conflict_sets[i][j] is "0" if type j has no conflict
++ * with type i and is "1" otherwise.
++ */
++/* high-16 = version, low-16 = check magic */
++#define ACM_MAGIC 0x0001debc
++
++/* size of the SHA1 hash identifying the XML policy from which the
++ binary policy was created */
++#define ACM_SHA1_HASH_SIZE 20
++
++/* each offset in bytes from start of the struct they
++ * are part of */
++
++/* V3 of the policy buffer aded a version structure */
++struct acm_policy_version
++{
++ uint32_t major;
++ uint32_t minor;
++};
++
++
++/* each buffer consists of all policy information for
++ * the respective policy given in the policy code
++ *
++ * acm_policy_buffer, acm_chwall_policy_buffer,
++ * and acm_ste_policy_buffer need to stay 32-bit aligned
++ * because we create binary policies also with external
++ * tools that assume packed representations (e.g. the java tool)
++ */
++struct acm_policy_buffer {
++ uint32_t magic;
++ uint32_t policy_version; /* ACM_POLICY_VERSION */
++ uint32_t len;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_buffer_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_buffer_offset;
++ struct acm_policy_version xml_pol_version; /* add in V3 */
++ uint8_t xml_policy_hash[ACM_SHA1_HASH_SIZE]; /* added in V4 */
++};
++
++
++struct acm_policy_reference_buffer {
++ uint32_t len;
++};
++
++struct acm_chwall_policy_buffer {
++ uint32_t policy_version; /* ACM_CHWALL_VERSION */
++ uint32_t policy_code;
++ uint32_t chwall_max_types;
++ uint32_t chwall_max_ssidrefs;
++ uint32_t chwall_max_conflictsets;
++ uint32_t chwall_ssid_offset;
++ uint32_t chwall_conflict_sets_offset;
++ uint32_t chwall_running_types_offset;
++ uint32_t chwall_conflict_aggregate_offset;
++};
++
++struct acm_ste_policy_buffer {
++ uint32_t policy_version; /* ACM_STE_VERSION */
++ uint32_t policy_code;
++ uint32_t ste_max_types;
++ uint32_t ste_max_ssidrefs;
++ uint32_t ste_ssid_offset;
++};
++
++struct acm_stats_buffer {
++ uint32_t magic;
++ uint32_t len;
++ uint32_t primary_policy_code;
++ uint32_t primary_stats_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_stats_offset;
++};
++
++struct acm_ste_stats_buffer {
++ uint32_t ec_eval_count;
++ uint32_t gt_eval_count;
++ uint32_t ec_denied_count;
++ uint32_t gt_denied_count;
++ uint32_t ec_cachehit_count;
++ uint32_t gt_cachehit_count;
++};
++
++struct acm_ssid_buffer {
++ uint32_t len;
++ ssidref_t ssidref;
++ uint32_t policy_reference_offset;
++ uint32_t primary_policy_code;
++ uint32_t primary_max_types;
++ uint32_t primary_types_offset;
++ uint32_t secondary_policy_code;
++ uint32_t secondary_max_types;
++ uint32_t secondary_types_offset;
++};
++
++#endif
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/xsm/acm_ops.h b/include/xen/interface/xsm/acm_ops.h
+new file mode 100644
+index 0000000..1fef7a0
+--- /dev/null
++++ b/include/xen/interface/xsm/acm_ops.h
+@@ -0,0 +1,159 @@
++/*
++ * acm_ops.h: Xen access control module hypervisor commands
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this software and associated documentation files (the "Software"), to
++ * deal in the Software without restriction, including without limitation the
++ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
++ * sell copies of the Software, and to permit persons to whom the Software is
++ * furnished to do so, subject to the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
++ * DEALINGS IN THE SOFTWARE.
++ *
++ * Reiner Sailer <sailer at watson.ibm.com>
++ * Copyright (c) 2005,2006 International Business Machines Corporation.
++ */
++
++#ifndef __XEN_PUBLIC_ACM_OPS_H__
++#define __XEN_PUBLIC_ACM_OPS_H__
++
++#include "../xen.h"
++#include "acm.h"
++
++/*
++ * Make sure you increment the interface version whenever you modify this file!
++ * This makes sure that old versions of acm tools will stop working in a
++ * well-defined way (rather than crashing the machine, for instance).
++ */
++#define ACM_INTERFACE_VERSION 0xAAAA000A
++
++/************************************************************************/
++
++/*
++ * Prototype for this hypercall is:
++ * int acm_op(int cmd, void *args)
++ * @cmd == ACMOP_??? (access control module operation).
++ * @args == Operation-specific extra arguments (NULL if none).
++ */
++
++
++#define ACMOP_setpolicy 1
++struct acm_setpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pushcache;
++ uint32_t pushcache_size;
++};
++
++
++#define ACMOP_getpolicy 2
++struct acm_getpolicy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
++
++
++#define ACMOP_dumpstats 3
++struct acm_dumpstats {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) pullcache;
++ uint32_t pullcache_size;
++};
++
++
++#define ACMOP_getssid 4
++#define ACM_GETBY_ssidref 1
++#define ACM_GETBY_domainid 2
++struct acm_getssid {
++ /* IN */
++ uint32_t get_ssid_by; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id;
++ XEN_GUEST_HANDLE_64(void) ssidbuf;
++ uint32_t ssidbuf_size;
++};
++
++#define ACMOP_getdecision 5
++struct acm_getdecision {
++ /* IN */
++ uint32_t get_decision_by1; /* ACM_GETBY_* */
++ uint32_t get_decision_by2; /* ACM_GETBY_* */
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id1;
++ union {
++ domaintype_t domainid;
++ ssidref_t ssidref;
++ } id2;
++ uint32_t hook;
++ /* OUT */
++ uint32_t acm_decision;
++};
++
++
++#define ACMOP_chgpolicy 6
++struct acm_change_policy {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) policy_pushcache;
++ uint32_t policy_pushcache_size;
++ XEN_GUEST_HANDLE_64(void) del_array;
++ uint32_t delarray_size;
++ XEN_GUEST_HANDLE_64(void) chg_array;
++ uint32_t chgarray_size;
++ /* OUT */
++ /* array with error code */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
++
++#define ACMOP_relabeldoms 7
++struct acm_relabel_doms {
++ /* IN */
++ XEN_GUEST_HANDLE_64(void) relabel_map;
++ uint32_t relabel_map_size;
++ /* OUT */
++ XEN_GUEST_HANDLE_64(void) err_array;
++ uint32_t errarray_size;
++};
++
++/* future interface to Xen */
++struct xen_acmctl {
++ uint32_t cmd;
++ uint32_t interface_version;
++ union {
++ struct acm_setpolicy setpolicy;
++ struct acm_getpolicy getpolicy;
++ struct acm_dumpstats dumpstats;
++ struct acm_getssid getssid;
++ struct acm_getdecision getdecision;
++ struct acm_change_policy change_policy;
++ struct acm_relabel_doms relabel_doms;
++ } u;
++};
++
++typedef struct xen_acmctl xen_acmctl_t;
++DEFINE_XEN_GUEST_HANDLE(xen_acmctl_t);
++
++#endif /* __XEN_PUBLIC_ACM_OPS_H__ */
++
++/*
++ * Local variables:
++ * mode: C
++ * c-set-style: "BSD"
++ * c-basic-offset: 4
++ * tab-width: 4
++ * indent-tabs-mode: nil
++ * End:
++ */
+diff --git a/include/xen/interface/xsm/flask_op.h b/include/xen/interface/xsm/flask_op.h
+new file mode 100644
+index 0000000..af20909
+--- /dev/null
++++ b/include/xen/interface/xsm/flask_op.h
+@@ -0,0 +1,43 @@
++/*
++ * This file contains the flask_op hypercall commands and definitions.
++ *
++ * Author: George Coker, <gscoker at alpha.ncsc.mil>
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License version 2,
++ * as published by the Free Software Foundation.
++ */
++
++#ifndef __FLASK_OP_H__
++#define __FLASK_OP_H__
++
++#define FLASK_LOAD 1
++#define FLASK_GETENFORCE 2
++#define FLASK_SETENFORCE 3
++#define FLASK_CONTEXT_TO_SID 4
++#define FLASK_SID_TO_CONTEXT 5
++#define FLASK_ACCESS 6
++#define FLASK_CREATE 7
++#define FLASK_RELABEL 8
++#define FLASK_USER 9
++#define FLASK_POLICYVERS 10
++#define FLASK_GETBOOL 11
++#define FLASK_SETBOOL 12
++#define FLASK_COMMITBOOLS 13
++#define FLASK_MLS 14
++#define FLASK_DISABLE 15
++#define FLASK_GETAVC_THRESHOLD 16
++#define FLASK_SETAVC_THRESHOLD 17
++#define FLASK_AVC_HASHSTATS 18
++#define FLASK_AVC_CACHESTATS 19
++#define FLASK_MEMBER 20
++
++typedef struct flask_op {
++ int cmd;
++ int size;
++ char *buf;
++} flask_op_t;
++
++DEFINE_XEN_GUEST_HANDLE(flask_op_t);
++
++#endif
+diff --git a/include/xen/pcifront.h b/include/xen/pcifront.h
+new file mode 100644
+index 0000000..bb5713f
+--- /dev/null
++++ b/include/xen/pcifront.h
+@@ -0,0 +1,69 @@
++/*
++ * PCI Frontend - arch-dependendent declarations
++ *
++ * Author: Ryan Wilson <hap9 at epoch.ncsc.mil>
++ */
++#ifndef __XEN_ASM_PCIFRONT_H__
++#define __XEN_ASM_PCIFRONT_H__
++
++#include <linux/spinlock.h>
++
++#ifdef __KERNEL__
++
++#ifndef __ia64__
++
++#include <asm/pci.h>
++
++struct pcifront_device;
++struct pci_bus;
++#define pcifront_sd pci_sysdata
++
++static inline struct pcifront_device *
++pcifront_get_pdev(struct pcifront_sd *sd)
++{
++ return sd->pdev;
++}
++
++static inline void pcifront_init_sd(struct pcifront_sd *sd,
++ unsigned int domain, unsigned int bus,
++ struct pcifront_device *pdev)
++{
++ sd->domain = domain;
++ sd->pdev = pdev;
++}
++
++static inline void pcifront_setup_root_resources(struct pci_bus *bus,
++ struct pcifront_sd *sd)
++{
++}
++
++#else /* __ia64__ */
++
++#include <linux/acpi.h>
++#include <asm/pci.h>
++#define pcifront_sd pci_controller
++
++extern void xen_add_resource(struct pci_controller *, unsigned int,
++ unsigned int, struct acpi_resource *);
++extern void xen_pcibios_setup_root_windows(struct pci_bus *,
++ struct pci_controller *);
++
++static inline struct pcifront_device *
++pcifront_get_pdev(struct pcifront_sd *sd)
++{
++ return (struct pcifront_device *)sd->platform_data;
++}
++
++static inline void pcifront_setup_root_resources(struct pci_bus *bus,
++ struct pcifront_sd *sd)
++{
++ xen_pcibios_setup_root_windows(bus, sd);
++}
++
++#endif /* __ia64__ */
++
++extern struct rw_semaphore pci_bus_sem;
++
++#endif /* __KERNEL__ */
++
++#endif /* __XEN_ASM_PCIFRONT_H__ */
+diff --git a/include/xen/public/evtchn.h b/include/xen/public/evtchn.h
+new file mode 100644
+index 0000000..938d4da
+--- /dev/null
++++ b/include/xen/public/evtchn.h
+@@ -0,0 +1,88 @@
++/******************************************************************************
++ * evtchn.h
++ *
++ * Interface to /dev/xen/evtchn.
++ *
++ * Copyright (c) 2003-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __LINUX_PUBLIC_EVTCHN_H__
++#define __LINUX_PUBLIC_EVTCHN_H__
++
++/*
++ * Bind a fresh port to VIRQ @virq.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_VIRQ \
++ _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq))
++struct ioctl_evtchn_bind_virq {
++ unsigned int virq;
++};
++
++/*
++ * Bind a fresh port to remote <@remote_domain, @remote_port>.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_INTERDOMAIN \
++ _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain))
++struct ioctl_evtchn_bind_interdomain {
++ unsigned int remote_domain, remote_port;
++};
++
++/*
++ * Allocate a fresh port for binding to @remote_domain.
++ * Return allocated port.
++ */
++#define IOCTL_EVTCHN_BIND_UNBOUND_PORT \
++ _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port))
++struct ioctl_evtchn_bind_unbound_port {
++ unsigned int remote_domain;
++};
++
++/*
++ * Unbind previously allocated @port.
++ */
++#define IOCTL_EVTCHN_UNBIND \
++ _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind))
++struct ioctl_evtchn_unbind {
++ unsigned int port;
++};
++
++/*
++ * Unbind previously allocated @port.
++ */
++#define IOCTL_EVTCHN_NOTIFY \
++ _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify))
++struct ioctl_evtchn_notify {
++ unsigned int port;
++};
++
++/* Clear and reinitialise the event buffer. Clear error condition. */
++#define IOCTL_EVTCHN_RESET \
++ _IOC(_IOC_NONE, 'E', 5, 0)
++
++#endif /* __LINUX_PUBLIC_EVTCHN_H__ */
+diff --git a/include/xen/public/gntdev.h b/include/xen/public/gntdev.h
+new file mode 100644
+index 0000000..8bd1467
+--- /dev/null
++++ b/include/xen/public/gntdev.h
+@@ -0,0 +1,119 @@
++/******************************************************************************
++ * gntdev.h
++ *
++ * Interface to /dev/xen/gntdev.
++ *
++ * Copyright (c) 2007, D G Murray
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __LINUX_PUBLIC_GNTDEV_H__
++#define __LINUX_PUBLIC_GNTDEV_H__
++
++struct ioctl_gntdev_grant_ref {
++ /* The domain ID of the grant to be mapped. */
++ uint32_t domid;
++ /* The grant reference of the grant to be mapped. */
++ uint32_t ref;
++};
++
++/*
++ * Inserts the grant references into the mapping table of an instance
++ * of gntdev. N.B. This does not perform the mapping, which is deferred
++ * until mmap() is called with @index as the offset.
++ */
++#define IOCTL_GNTDEV_MAP_GRANT_REF \
++_IOC(_IOC_NONE, 'G', 0, sizeof(struct ioctl_gntdev_map_grant_ref))
++struct ioctl_gntdev_map_grant_ref {
++ /* IN parameters */
++ /* The number of grants to be mapped. */
++ uint32_t count;
++ uint32_t pad;
++ /* OUT parameters */
++ /* The offset to be used on a subsequent call to mmap(). */
++ uint64_t index;
++ /* Variable IN parameter. */
++ /* Array of grant references, of size @count. */
++ struct ioctl_gntdev_grant_ref refs[1];
++};
++
++/*
++ * Removes the grant references from the mapping table of an instance of
++ * of gntdev. N.B. munmap() must be called on the relevant virtual address(es)
++ * before this ioctl is called, or an error will result.
++ */
++#define IOCTL_GNTDEV_UNMAP_GRANT_REF \
++_IOC(_IOC_NONE, 'G', 1, sizeof(struct ioctl_gntdev_unmap_grant_ref))
++struct ioctl_gntdev_unmap_grant_ref {
++ /* IN parameters */
++ /* The offset was returned by the corresponding map operation. */
++ uint64_t index;
++ /* The number of pages to be unmapped. */
++ uint32_t count;
++ uint32_t pad;
++};
++
++/*
++ * Returns the offset in the driver's address space that corresponds
++ * to @vaddr. This can be used to perform a munmap(), followed by an
++ * UNMAP_GRANT_REF ioctl, where no state about the offset is retained by
++ * the caller. The number of pages that were allocated at the same time as
++ * @vaddr is returned in @count.
++ *
++ * N.B. Where more than one page has been mapped into a contiguous range, the
++ * supplied @vaddr must correspond to the start of the range; otherwise
++ * an error will result. It is only possible to munmap() the entire
++ * contiguously-allocated range at once, and not any subrange thereof.
++ */
++#define IOCTL_GNTDEV_GET_OFFSET_FOR_VADDR \
++_IOC(_IOC_NONE, 'G', 2, sizeof(struct ioctl_gntdev_get_offset_for_vaddr))
++struct ioctl_gntdev_get_offset_for_vaddr {
++ /* IN parameters */
++ /* The virtual address of the first mapped page in a range. */
++ uint64_t vaddr;
++ /* OUT parameters */
++ /* The offset that was used in the initial mmap() operation. */
++ uint64_t offset;
++ /* The number of pages mapped in the VM area that begins at @vaddr. */
++ uint32_t count;
++ uint32_t pad;
++};
++
++/*
++ * Sets the maximum number of grants that may mapped at once by this gntdev
++ * instance.
++ *
++ * N.B. This must be called before any other ioctl is performed on the device.
++ */
++#define IOCTL_GNTDEV_SET_MAX_GRANTS \
++_IOC(_IOC_NONE, 'G', 3, sizeof(struct ioctl_gntdev_set_max_grants))
++struct ioctl_gntdev_set_max_grants {
++ /* IN parameter */
++ /* The maximum number of grants that may be mapped at once. */
++ uint32_t count;
++};
++
++#endif /* __LINUX_PUBLIC_GNTDEV_H__ */
+diff --git a/include/xen/public/privcmd.h b/include/xen/public/privcmd.h
+new file mode 100644
+index 0000000..d1162ee
+--- /dev/null
++++ b/include/xen/public/privcmd.h
+@@ -0,0 +1,79 @@
++/******************************************************************************
++ * privcmd.h
++ *
++ * Interface to /proc/xen/privcmd.
++ *
++ * Copyright (c) 2003-2005, K A Fraser
++ *
++ * This program is free software; you can redistribute it and/or
++ * modify it under the terms of the GNU General Public License version 2
++ * as published by the Free Software Foundation; or, when distributed
++ * separately from the Linux kernel or incorporated into other
++ * software packages, subject to the following license:
++ *
++ * Permission is hereby granted, free of charge, to any person obtaining a copy
++ * of this source file (the "Software"), to deal in the Software without
++ * restriction, including without limitation the rights to use, copy, modify,
++ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
++ * and to permit persons to whom the Software is furnished to do so, subject to
++ * the following conditions:
++ *
++ * The above copyright notice and this permission notice shall be included in
++ * all copies or substantial portions of the Software.
++ *
++ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
++ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
++ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
++ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
++ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
++ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
++ * IN THE SOFTWARE.
++ */
++
++#ifndef __LINUX_PUBLIC_PRIVCMD_H__
++#define __LINUX_PUBLIC_PRIVCMD_H__
++
++#include <linux/types.h>
++
++#ifndef __user
++#define __user
++#endif
++
++typedef struct privcmd_hypercall
++{
++ __u64 op;
++ __u64 arg[5];
++} privcmd_hypercall_t;
++
++typedef struct privcmd_mmap_entry {
++ __u64 va;
++ __u64 mfn;
++ __u64 npages;
++} privcmd_mmap_entry_t;
++
++typedef struct privcmd_mmap {
++ int num;
++ domid_t dom; /* target domain */
++ privcmd_mmap_entry_t __user *entry;
++} privcmd_mmap_t;
++
++typedef struct privcmd_mmapbatch {
++ int num; /* number of pages to populate */
++ domid_t dom; /* target domain */
++ __u64 addr; /* virtual address */
++ xen_pfn_t __user *arr; /* array of mfns - top nibble set on err */
++} privcmd_mmapbatch_t;
++
++/*
++ * @cmd: IOCTL_PRIVCMD_HYPERCALL
++ * @arg: &privcmd_hypercall_t
++ * Return: Value returned from execution of the specified hypercall.
++ */
++#define IOCTL_PRIVCMD_HYPERCALL \
++ _IOC(_IOC_NONE, 'P', 0, sizeof(privcmd_hypercall_t))
++#define IOCTL_PRIVCMD_MMAP \
++ _IOC(_IOC_NONE, 'P', 2, sizeof(privcmd_mmap_t))
++#define IOCTL_PRIVCMD_MMAPBATCH \
++ _IOC(_IOC_NONE, 'P', 3, sizeof(privcmd_mmapbatch_t))
++
++#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
+diff --git a/include/xen/sysctl.h b/include/xen/sysctl.h
+new file mode 100644
+index 0000000..7fe9250
+--- /dev/null
++++ b/include/xen/sysctl.h
+@@ -0,0 +1,11 @@
++#ifndef _XEN_SYSCTL_H
++#define _XEN_SYSCTL_H
++
++/* CTL_XEN names: */
++enum
++{
++ CTL_XEN_INDEPENDENT_WALLCLOCK=1,
++ CTL_XEN_PERMITTED_CLOCK_JITTER=2,
++};
++
++#endif /* _XEN_SYSCTL_H */
+diff --git a/include/xen/xen_proc.h b/include/xen/xen_proc.h
+new file mode 100644
+index 0000000..44af17c
+--- /dev/null
++++ b/include/xen/xen_proc.h
+@@ -0,0 +1,12 @@
++
++#ifndef __ASM_XEN_PROC_H__
++#define __ASM_XEN_PROC_H__
++
++#include <linux/proc_fs.h>
++
++extern struct proc_dir_entry *create_xen_proc_entry(
++ const char *name, mode_t mode);
++extern void remove_xen_proc_entry(
++ const char *name);
++
++#endif /* __ASM_XEN_PROC_H__ */
+diff --git a/include/xen/xenbus.h b/include/xen/xenbus.h
+index 6369d89..a0de4ba 100644
+--- a/include/xen/xenbus.h
++++ b/include/xen/xenbus.h
+@@ -39,6 +39,7 @@
+ #include <linux/mutex.h>
+ #include <linux/completion.h>
+ #include <linux/init.h>
++#include <linux/err.h>
+ #include <xen/interface/xen.h>
+ #include <xen/interface/grant_table.h>
+ #include <xen/interface/io/xenbus.h>
+@@ -55,8 +56,21 @@ struct xenbus_watch
+ /* Callback (executed in a process context with no locks held). */
+ void (*callback)(struct xenbus_watch *,
+ const char **vec, unsigned int len);
++
++#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H)
++ /* See XBWF_ definitions below. */
++ unsigned long flags;
++#endif
+ };
+
++#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H)
++/*
++ * Execute callback in its own kthread. Useful if the callback is long
++ * running or heavily serialised, to avoid taking out the main xenwatch thread
++ * for a long period of time (or even unwittingly causing a deadlock).
++ */
++#define XBWF_new_thread 1
++#endif
+
+ /* A xenbus device. */
+ struct xenbus_device {
+@@ -83,8 +97,7 @@ struct xenbus_device_id
+
+ /* A xenbus driver. */
+ struct xenbus_driver {
+- char *name;
+- struct module *owner;
++ const char *name;
+ const struct xenbus_device_id *ids;
+ int (*probe)(struct xenbus_device *dev,
+ const struct xenbus_device_id *id);
+@@ -94,7 +107,7 @@ struct xenbus_driver {
+ int (*suspend)(struct xenbus_device *dev);
+ int (*suspend_cancel)(struct xenbus_device *dev);
+ int (*resume)(struct xenbus_device *dev);
+- int (*uevent)(struct xenbus_device *, char **, int, char *, int);
++ int (*uevent)(struct xenbus_device *, struct kobj_uevent_env *);
+ struct device_driver driver;
+ int (*read_otherend_details)(struct xenbus_device *dev);
+ int (*is_ready)(struct xenbus_device *dev);
+@@ -112,7 +125,6 @@ int __must_check __xenbus_register_frontend(struct xenbus_driver *drv,
+ static inline int __must_check
+ xenbus_register_frontend(struct xenbus_driver *drv)
+ {
+- WARN_ON(drv->owner != THIS_MODULE);
+ return __xenbus_register_frontend(drv, THIS_MODULE, KBUILD_MODNAME);
+ }
+
+@@ -122,7 +134,6 @@ int __must_check __xenbus_register_backend(struct xenbus_driver *drv,
+ static inline int __must_check
+ xenbus_register_backend(struct xenbus_driver *drv)
+ {
+- WARN_ON(drv->owner != THIS_MODULE);
+ return __xenbus_register_backend(drv, THIS_MODULE, KBUILD_MODNAME);
+ }
+
+@@ -136,8 +147,6 @@ struct xenbus_transaction
+ /* Nil transaction ID. */
+ #define XBT_NIL ((struct xenbus_transaction) { 0 })
+
+-int __init xenbus_dev_init(void);
+-
+ char **xenbus_directory(struct xenbus_transaction t,
+ const char *dir, const char *node, unsigned int *num);
+ void *xenbus_read(struct xenbus_transaction t,
+@@ -167,7 +176,6 @@ int xenbus_printf(struct xenbus_transaction t,
+ int xenbus_gather(struct xenbus_transaction t, const char *dir, ...);
+
+ /* notifer routines for when the xenstore comes up */
+-extern int xenstored_ready;
+ int register_xenstore_notifier(struct notifier_block *nb);
+ void unregister_xenstore_notifier(struct notifier_block *nb);
+
+@@ -180,12 +188,9 @@ void xs_suspend_cancel(void);
+ /* Used by xenbus_dev to borrow kernel's store connection. */
+ void *xenbus_dev_request_and_reply(struct xsd_sockmsg *msg);
+
+-struct work_struct;
+-
+ /* Prepare for domain suspend: then resume or cancel the suspend. */
+ void xenbus_suspend(void);
+ void xenbus_resume(void);
+-void xenbus_probe(struct work_struct *);
+ void xenbus_suspend_cancel(void);
+
+ #define XENBUS_IS_ERR_READ(str) ({ \
+@@ -198,38 +203,132 @@ void xenbus_suspend_cancel(void);
+
+ #define XENBUS_EXIST_ERR(err) ((err) == -ENOENT || (err) == -ERANGE)
+
++
++/**
++ * Register a watch on the given path, using the given xenbus_watch structure
++ * for storage, and the given callback function as the callback. Return 0 on
++ * success, or -errno on error. On success, the given path will be saved as
++ * watch->node, and remains the caller's to free. On error, watch->node will
++ * be NULL, the device will switch to XenbusStateClosing, and the error will
++ * be saved in the store.
++ */
+ int xenbus_watch_path(struct xenbus_device *dev, const char *path,
+ struct xenbus_watch *watch,
+ void (*callback)(struct xenbus_watch *,
+ const char **, unsigned int));
++
++
++#if defined(CONFIG_XEN) || defined(HAVE_XEN_PLATFORM_COMPAT_H)
++/**
++ * Register a watch on the given path/path2, using the given xenbus_watch
++ * structure for storage, and the given callback function as the callback.
++ * Return 0 on success, or -errno on error. On success, the watched path
++ * (path/path2) will be saved as watch->node, and becomes the caller's to
++ * kfree(). On error, watch->node will be NULL, so the caller has nothing to
++ * free, the device will switch to XenbusStateClosing, and the error will be
++ * saved in the store.
++ */
++int xenbus_watch_path2(struct xenbus_device *dev, const char *path,
++ const char *path2, struct xenbus_watch *watch,
++ void (*callback)(struct xenbus_watch *,
++ const char **, unsigned int));
++#else
+ int xenbus_watch_pathfmt(struct xenbus_device *dev, struct xenbus_watch *watch,
+ void (*callback)(struct xenbus_watch *,
+ const char **, unsigned int),
+ const char *pathfmt, ...)
+ __attribute__ ((format (printf, 4, 5)));
++#endif
+
++/**
++ * Advertise in the store a change of the given driver to the given new_state.
++ * Return 0 on success, or -errno on error. On error, the device will switch
++ * to XenbusStateClosing, and the error will be saved in the store.
++ */
+ int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state new_state);
++
++
++/**
++ * Grant access to the given ring_mfn to the peer of the given device. Return
++ * 0 on success, or -errno on error. On error, the device will switch to
++ * XenbusStateClosing, and the error will be saved in the store.
++ */
+ int xenbus_grant_ring(struct xenbus_device *dev, unsigned long ring_mfn);
+-int xenbus_map_ring_valloc(struct xenbus_device *dev,
+- int gnt_ref, void **vaddr);
++
++
++/**
++ * Map a page of memory into this domain from another domain's grant table.
++ * xenbus_map_ring_valloc allocates a page of virtual address space, maps the
++ * page to that address, and sets *vaddr to that address.
++ * xenbus_map_ring does not allocate the virtual address space (you must do
++ * this yourself!). It only maps in the page to the specified address.
++ * Returns 0 on success, and GNTST_* (see xen/include/interface/grant_table.h)
++ * or -ENOMEM on error. If an error is returned, device will switch to
++ * XenbusStateClosing and the error message will be saved in XenStore.
++ */
++struct vm_struct *xenbus_map_ring_valloc(struct xenbus_device *dev,
++ int gnt_ref);
+ int xenbus_map_ring(struct xenbus_device *dev, int gnt_ref,
+ grant_handle_t *handle, void *vaddr);
+
+-int xenbus_unmap_ring_vfree(struct xenbus_device *dev, void *vaddr);
++
++/**
++ * Unmap a page of memory in this domain that was imported from another domain.
++ * Use xenbus_unmap_ring_vfree if you mapped in your memory with
++ * xenbus_map_ring_valloc (it will free the virtual address space).
++ * Returns 0 on success and returns GNTST_* on error
++ * (see xen/include/interface/grant_table.h).
++ */
++int xenbus_unmap_ring_vfree(struct xenbus_device *dev, struct vm_struct *);
+ int xenbus_unmap_ring(struct xenbus_device *dev,
+ grant_handle_t handle, void *vaddr);
+
++
++/**
++ * Allocate an event channel for the given xenbus_device, assigning the newly
++ * created local port to *port. Return 0 on success, or -errno on error. On
++ * error, the device will switch to XenbusStateClosing, and the error will be
++ * saved in the store.
++ */
+ int xenbus_alloc_evtchn(struct xenbus_device *dev, int *port);
+-int xenbus_bind_evtchn(struct xenbus_device *dev, int remote_port, int *port);
++
++
++/**
++ * Free an existing event channel. Returns 0 on success or -errno on error.
++ */
+ int xenbus_free_evtchn(struct xenbus_device *dev, int port);
+
++
++/**
++ * Return the state of the driver rooted at the given store path, or
++ * XenbusStateUnknown if no state can be read.
++ */
+ enum xenbus_state xenbus_read_driver_state(const char *path);
+
+-void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt, ...);
+-void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt, ...);
++
++/***
++ * Report the given negative errno into the store, along with the given
++ * formatted message.
++ */
++void xenbus_dev_error(struct xenbus_device *dev, int err, const char *fmt,
++ ...);
++
++
++/***
++ * Equivalent to xenbus_dev_error(dev, err, fmt, args), followed by
++ * xenbus_switch_state(dev, NULL, XenbusStateClosing) to schedule an orderly
++ * closedown of this driver and its peer.
++ */
++void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt,
++ ...);
++
++int xenbus_dev_init(void);
+
+ const char *xenbus_strstate(enum xenbus_state state);
+ int xenbus_dev_is_online(struct xenbus_device *dev);
+ int xenbus_frontend_closed(struct xenbus_device *dev);
+
++int xenbus_for_each_backend(void *arg, int (*fn)(struct device *, void *));
++int xenbus_for_each_frontend(void *arg, int (*fn)(struct device *, void *));
++
+ #endif /* _XEN_XENBUS_H */
+diff --git a/include/xen/xencons.h b/include/xen/xencons.h
+new file mode 100644
+index 0000000..c5a55d6
+--- /dev/null
++++ b/include/xen/xencons.h
+@@ -0,0 +1,17 @@
++#ifndef __ASM_XENCONS_H__
++#define __ASM_XENCONS_H__
++
++struct dom0_vga_console_info;
++void dom0_init_screen_info(const struct dom0_vga_console_info *, size_t);
++
++void xencons_force_flush(void);
++void xencons_resume(void);
++
++/* Interrupt work hooks. Receive data, or kick data out. */
++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 --git a/include/xen/xenoprof.h b/include/xen/xenoprof.h
+new file mode 100644
+index 0000000..4c3ab0f
+--- /dev/null
++++ b/include/xen/xenoprof.h
+@@ -0,0 +1,42 @@
++/******************************************************************************
++ * xen/xenoprof.h
++ *
++ * Copyright (c) 2006 Isaku Yamahata <yamahata at valinux co jp>
++ * VA Linux Systems Japan K.K.
++ *
++ * This program is free software; you can redistribute it and/or modify
++ * it under the terms of the GNU General Public License as published by
++ * the Free Software Foundation; either version 2 of the License, or
++ * (at your option) any later version.
++ *
++ * This program is distributed in the hope that it will be useful,
++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++ * GNU General Public License for more details.
++ *
++ * You should have received a copy of the GNU General Public License
++ * along with this program; if not, write to the Free Software
++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
++ *
++ */
++
++#ifndef __XEN_XENOPROF_H__
++#define __XEN_XENOPROF_H__
++#ifdef CONFIG_XEN
++
++#include <asm/xenoprof.h>
++
++struct oprofile_operations;
++int xenoprofile_init(struct oprofile_operations * ops);
++void xenoprofile_exit(void);
++
++struct xenoprof_shared_buffer {
++ char *buffer;
++ struct xenoprof_arch_shared_buffer arch;
++};
++#else
++#define xenoprofile_init(ops) (-ENOSYS)
++#define xenoprofile_exit() do { } while (0)
++
++#endif /* CONFIG_XEN */
++#endif /* __XEN_XENOPROF_H__ */
+diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt
+index 9fdba03..32100ec 100644
+--- a/kernel/Kconfig.preempt
++++ b/kernel/Kconfig.preempt
+@@ -35,6 +35,7 @@ config PREEMPT_VOLUNTARY
+
+ config PREEMPT
+ bool "Preemptible Kernel (Low-Latency Desktop)"
++ depends on !XEN
+ help
+ This option reduces the latency of the kernel by making
+ all kernel code (that is not executing in a critical section)
+diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c
+index c66d3f1..7db878d 100644
+--- a/kernel/irq/spurious.c
++++ b/kernel/irq/spurious.c
+@@ -182,7 +182,7 @@ void note_interrupt(unsigned int irq, struct irq_desc *desc,
+ */
+ if (time_after(jiffies, desc->last_unhandled + HZ/10))
+ desc->irqs_unhandled = 1;
+- else
++ else if (!irq_ignore_unhandled(irq))
+ desc->irqs_unhandled++;
+ desc->last_unhandled = jiffies;
+ if (unlikely(action_ret != IRQ_NONE))
+diff --git a/kernel/kexec.c b/kernel/kexec.c
+index 1c5fcac..d6b9fb6 100644
+--- a/kernel/kexec.c
++++ b/kernel/kexec.c
+@@ -339,13 +339,26 @@ static int kimage_is_destination_range(struct kimage *image,
+ return 0;
+ }
+
+-static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
++static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order, unsigned long limit)
+ {
+ struct page *pages;
+
+ pages = alloc_pages(gfp_mask, order);
+ if (pages) {
+ unsigned int count, i;
++#ifdef CONFIG_XEN
++ int address_bits;
++
++ if (limit == ~0UL)
++ address_bits = BITS_PER_LONG;
++ else
++ address_bits = ilog2(limit);
++
++ if (xen_limit_pages_to_max_mfn(pages, order, address_bits) < 0) {
++ __free_pages(pages, order);
++ return NULL;
++ }
++#endif
+ pages->mapping = NULL;
+ set_page_private(pages, order);
+ count = 1 << order;
+@@ -364,6 +377,9 @@ static void kimage_free_pages(struct page *page)
+ count = 1 << order;
+ for (i = 0; i < count; i++)
+ ClearPageReserved(page + i);
++#ifdef CONFIG_XEN
++ xen_destroy_contiguous_region((unsigned long)page_address(page), order);
++#endif
+ __free_pages(page, order);
+ }
+
+@@ -409,10 +425,10 @@ static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
+ do {
+ unsigned long pfn, epfn, addr, eaddr;
+
+- pages = kimage_alloc_pages(GFP_KERNEL, order);
++ pages = kimage_alloc_pages(GFP_KERNEL, order, KEXEC_CONTROL_MEMORY_LIMIT);
+ if (!pages)
+ break;
+- pfn = page_to_pfn(pages);
++ pfn = kexec_page_to_pfn(pages);
+ epfn = pfn + count;
+ addr = pfn << PAGE_SHIFT;
+ eaddr = epfn << PAGE_SHIFT;
+@@ -446,6 +462,7 @@ static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
+ return pages;
+ }
+
++#ifndef CONFIG_XEN
+ static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
+ unsigned int order)
+ {
+@@ -499,7 +516,7 @@ static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
+ }
+ /* If I don't overlap any segments I have found my hole! */
+ if (i == image->nr_segments) {
+- pages = pfn_to_page(hole_start >> PAGE_SHIFT);
++ pages = kexec_pfn_to_page(hole_start >> PAGE_SHIFT);
+ break;
+ }
+ }
+@@ -526,6 +543,13 @@ struct page *kimage_alloc_control_pages(struct kimage *image,
+
+ return pages;
+ }
++#else /* !CONFIG_XEN */
++struct page *kimage_alloc_control_pages(struct kimage *image,
++ unsigned int order)
++{
++ return kimage_alloc_normal_control_pages(image, order);
++}
++#endif
+
+ static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
+ {
+@@ -541,7 +565,7 @@ static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
+ return -ENOMEM;
+
+ ind_page = page_address(page);
+- *image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
++ *image->entry = kexec_virt_to_phys(ind_page) | IND_INDIRECTION;
+ image->entry = ind_page;
+ image->last_entry = ind_page +
+ ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
+@@ -602,13 +626,13 @@ static int kimage_terminate(struct kimage *image)
+ #define for_each_kimage_entry(image, ptr, entry) \
+ for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \
+ ptr = (entry & IND_INDIRECTION)? \
+- phys_to_virt((entry & PAGE_MASK)): ptr +1)
++ kexec_phys_to_virt((entry & PAGE_MASK)): ptr +1)
+
+ static void kimage_free_entry(kimage_entry_t entry)
+ {
+ struct page *page;
+
+- page = pfn_to_page(entry >> PAGE_SHIFT);
++ page = kexec_pfn_to_page(entry >> PAGE_SHIFT);
+ kimage_free_pages(page);
+ }
+
+@@ -620,6 +644,10 @@ static void kimage_free(struct kimage *image)
+ if (!image)
+ return;
+
++#ifdef CONFIG_XEN
++ xen_machine_kexec_unload(image);
++#endif
++
+ kimage_free_extra_pages(image);
+ for_each_kimage_entry(image, ptr, entry) {
+ if (entry & IND_INDIRECTION) {
+@@ -695,7 +723,7 @@ static struct page *kimage_alloc_page(struct kimage *image,
+ * have a match.
+ */
+ list_for_each_entry(page, &image->dest_pages, lru) {
+- addr = page_to_pfn(page) << PAGE_SHIFT;
++ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
+ if (addr == destination) {
+ list_del(&page->lru);
+ return page;
+@@ -706,16 +734,16 @@ static struct page *kimage_alloc_page(struct kimage *image,
+ kimage_entry_t *old;
+
+ /* Allocate a page, if we run out of memory give up */
+- page = kimage_alloc_pages(gfp_mask, 0);
++ page = kimage_alloc_pages(gfp_mask, 0, KEXEC_SOURCE_MEMORY_LIMIT);
+ if (!page)
+ return NULL;
+ /* If the page cannot be used file it away */
+- if (page_to_pfn(page) >
++ if (kexec_page_to_pfn(page) >
+ (KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
+ list_add(&page->lru, &image->unuseable_pages);
+ continue;
+ }
+- addr = page_to_pfn(page) << PAGE_SHIFT;
++ addr = kexec_page_to_pfn(page) << PAGE_SHIFT;
+
+ /* If it is the destination page we want use it */
+ if (addr == destination)
+@@ -738,7 +766,7 @@ static struct page *kimage_alloc_page(struct kimage *image,
+ struct page *old_page;
+
+ old_addr = *old & PAGE_MASK;
+- old_page = pfn_to_page(old_addr >> PAGE_SHIFT);
++ old_page = kexec_pfn_to_page(old_addr >> PAGE_SHIFT);
+ copy_highpage(page, old_page);
+ *old = addr | (*old & ~PAGE_MASK);
+
+@@ -788,7 +816,7 @@ static int kimage_load_normal_segment(struct kimage *image,
+ result = -ENOMEM;
+ goto out;
+ }
+- result = kimage_add_page(image, page_to_pfn(page)
++ result = kimage_add_page(image, kexec_page_to_pfn(page)
+ << PAGE_SHIFT);
+ if (result < 0)
+ goto out;
+@@ -820,6 +848,7 @@ out:
+ return result;
+ }
+
++#ifndef CONFIG_XEN
+ static int kimage_load_crash_segment(struct kimage *image,
+ struct kexec_segment *segment)
+ {
+@@ -842,7 +871,7 @@ static int kimage_load_crash_segment(struct kimage *image,
+ char *ptr;
+ size_t uchunk, mchunk;
+
+- page = pfn_to_page(maddr >> PAGE_SHIFT);
++ page = kexec_pfn_to_page(maddr >> PAGE_SHIFT);
+ if (!page) {
+ result = -ENOMEM;
+ goto out;
+@@ -891,6 +920,13 @@ static int kimage_load_segment(struct kimage *image,
+
+ return result;
+ }
++#else /* CONFIG_XEN */
++static int kimage_load_segment(struct kimage *image,
++ struct kexec_segment *segment)
++{
++ return kimage_load_normal_segment(image, segment);
++}
++#endif
+
+ /*
+ * Exec Kernel system call: for obvious reasons only root may call it.
+@@ -1001,6 +1037,13 @@ asmlinkage long sys_kexec_load(unsigned long entry, unsigned long nr_segments,
+ if (result)
+ goto out;
+ }
++#ifdef CONFIG_XEN
++ if (image) {
++ result = xen_machine_kexec_load(image);
++ if (result)
++ goto out;
++ }
++#endif
+ /* Install the new kernel, and Uninstall the old */
+ image = xchg(dest_image, image);
+
+@@ -1146,6 +1189,7 @@ static int __init crash_notes_memory_init(void)
+ module_init(crash_notes_memory_init)
+
+
++#ifndef CONFIG_XEN
+ /*
+ * parsing the "crashkernel" commandline
+ *
+@@ -1308,7 +1352,7 @@ int __init parse_crashkernel(char *cmdline,
+
+ return 0;
+ }
+-
++#endif
+
+
+ void crash_save_vmcoreinfo(void)
+diff --git a/kernel/sysctl.c b/kernel/sysctl.c
+index 2911665..5149ded 100644
+--- a/kernel/sysctl.c
++++ b/kernel/sysctl.c
+@@ -645,6 +645,7 @@ static struct ctl_table kern_table[] = {
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
++#ifndef CONFIG_XEN
+ {
+ .procname = "nmi_watchdog",
+ .data = &nmi_watchdog_enabled,
+@@ -653,6 +654,7 @@ static struct ctl_table kern_table[] = {
+ .proc_handler = &proc_nmi_enabled,
+ },
+ #endif
++#endif
+ #if defined(CONFIG_X86)
+ {
+ .ctl_name = KERN_PANIC_ON_NMI,
+@@ -707,7 +709,7 @@ static struct ctl_table kern_table[] = {
+ .proc_handler = &proc_dointvec,
+ },
+ #endif
+-#if defined(CONFIG_ACPI_SLEEP) && defined(CONFIG_X86)
++#if defined(CONFIG_ACPI_SLEEP) && defined(CONFIG_X86) && !defined(CONFIG_ACPI_PV_SLEEP)
+ {
+ .procname = "acpi_video_flags",
+ .data = &acpi_realmode_flags,
+diff --git a/kernel/sysctl_check.c b/kernel/sysctl_check.c
+index c09350d..d8db2ac 100644
+--- a/kernel/sysctl_check.c
++++ b/kernel/sysctl_check.c
+@@ -4,6 +4,7 @@
+ #include <linux/sunrpc/debug.h>
+ #include <linux/string.h>
+ #include <net/ip_vs.h>
++#include <xen/sysctl.h>
+
+ struct trans_ctl_table {
+ int ctl_name;
+@@ -893,6 +894,14 @@ static const struct trans_ctl_table trans_bus_table[] = {
+ {}
+ };
+
++#ifdef CONFIG_XEN
++static const struct trans_ctl_table trans_xen_table[] = {
++ { CTL_XEN_INDEPENDENT_WALLCLOCK, "independent_wallclock" },
++ { CTL_XEN_PERMITTED_CLOCK_JITTER, "permitted_clock_jitter" },
++ {}
++};
++#endif
++
+ static const struct trans_ctl_table trans_arlan_conf_table0[] = {
+ { 1, "spreadingCode" },
+ { 2, "channelNumber" },
+@@ -1228,6 +1237,9 @@ static const struct trans_ctl_table trans_root_table[] = {
+ { CTL_BUS, "bus", trans_bus_table },
+ { CTL_ABI, "abi" },
+ /* CTL_CPU not used */
++#ifdef CONFIG_XEN
++ { CTL_XEN, "xen", trans_xen_table },
++#endif
+ { CTL_ARLAN, "arlan", trans_arlan_table },
+ { CTL_S390DBF, "s390dbf", trans_s390dbf_table },
+ { CTL_SUNRPC, "sunrpc", trans_sunrpc_table },
+diff --git a/lib/swiotlb-xen.c b/lib/swiotlb-xen.c
+new file mode 100644
+index 0000000..0a75e44
+--- /dev/null
++++ b/lib/swiotlb-xen.c
+@@ -0,0 +1,785 @@
++/*
++ * 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>
++ */
++
++#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/iommu-helper.h>
++#include <linux/highmem.h>
++#include <asm/io.h>
++#include <asm/pci.h>
++#include <asm/dma.h>
++#include <asm/uaccess.h>
++#include <xen/gnttab.h>
++#include <xen/interface/memory.h>
++#include <asm/gnttab_dma.h>
++
++int swiotlb;
++
++#define OFFSET(val,align) ((unsigned long)((val) & ( (align) - 1)))
++
++/*
++ * Maximum allowable number of contiguous slabs to map,
++ * must be a power of 2. What is the appropriate value ?
++ * The complexity of {map,unmap}_single is linearly dependent on this value.
++ */
++#define IO_TLB_SEGSIZE 128
++
++/*
++ * log of the size of each IO TLB slab. The number of slabs is command line
++ * controllable.
++ */
++#define IO_TLB_SHIFT 11
++
++int swiotlb_force;
++
++static char *iotlb_virt_start;
++static unsigned long iotlb_nslabs;
++
++/*
++ * 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;
++
++/* 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));
++}
++
++/*
++ * When the IOMMU overflows we return a fallback buffer. This sets the size.
++ */
++static unsigned long io_tlb_overflow = 32*1024;
++
++void *io_tlb_overflow_buffer;
++
++/*
++ * This is a free list describing the number of free entries available from
++ * each index
++ */
++static unsigned int *io_tlb_list;
++static unsigned int io_tlb_index;
++
++/*
++ * 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;
++
++/*
++ * Protect the above data structures in the map and unmap calls
++ */
++static DEFINE_SPINLOCK(io_tlb_lock);
++
++static unsigned int dma_bits;
++static unsigned int __initdata max_dma_bits = 32;
++static int __init
++setup_dma_bits(char *str)
++{
++ max_dma_bits = simple_strtoul(str, NULL, 0);
++ return 0;
++}
++__setup("dma_bits=", setup_dma_bits);
++
++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? */
++
++/*
++ * Statically reserve bounce buffer space and initialize bounce buffer data
++ * structures for the software IO TLB used to implement the PCI DMA API.
++ */
++void __init
++swiotlb_init_with_default_size(size_t default_size)
++{
++ unsigned long i, bytes;
++ int rc;
++
++ 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");
++
++ dma_bits = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT) + PAGE_SHIFT;
++ for (i = 0; i < iotlb_nslabs; i += IO_TLB_SEGSIZE) {
++ do {
++ rc = xen_create_contiguous_region(
++ (unsigned long)iotlb_virt_start + (i << IO_TLB_SHIFT),
++ get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT),
++ dma_bits);
++ } while (rc && dma_bits++ < max_dma_bits);
++ if (rc) {
++ if (i == 0)
++ panic("No suitable physical memory available for SWIOTLB buffer!\n"
++ "Use dom0_mem Xen boot parameter to reserve\n"
++ "some DMA memory (e.g., dom0_mem=-128M).\n");
++ iotlb_nslabs = i;
++ i <<= IO_TLB_SHIFT;
++ free_bootmem(__pa(iotlb_virt_start + i), bytes - i);
++ bytes = i;
++ for (dma_bits = 0; i > 0; i -= IO_TLB_SEGSIZE << IO_TLB_SHIFT) {
++ unsigned int bits = fls64(virt_to_bus(iotlb_virt_start + i - 1));
++
++ if (bits > dma_bits)
++ dma_bits = bits;
++ }
++ break;
++ }
++ }
++
++ /*
++ * 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));
++
++ /*
++ * Get the overflow emergency buffer
++ */
++ io_tlb_overflow_buffer = alloc_bootmem_low(io_tlb_overflow);
++ if (!io_tlb_overflow_buffer)
++ panic("Cannot allocate SWIOTLB overflow buffer!\n");
++
++ do {
++ rc = xen_create_contiguous_region(
++ (unsigned long)io_tlb_overflow_buffer,
++ get_order(io_tlb_overflow),
++ dma_bits);
++ } while (rc && dma_bits++ < max_dma_bits);
++ if (rc)
++ panic("No suitable physical memory available for SWIOTLB overflow buffer!\n");
++
++ iotlb_pfn_start = __pa(iotlb_virt_start) >> PAGE_SHIFT;
++ iotlb_pfn_end = iotlb_pfn_start + (bytes >> PAGE_SHIFT);
++
++ printk(KERN_INFO "Software IO TLB enabled: \n"
++ " Aperture: %lu megabytes\n"
++ " Kernel range: %p - %p\n"
++ " Address size: %u bits\n",
++ bytes >> 20,
++ iotlb_virt_start, iotlb_virt_start + bytes,
++ dma_bits);
++}
++
++void __init
++swiotlb_init(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");
++}
++
++/*
++ * 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) {
++ unsigned long flags;
++
++ if (((bytes = len) + buffer.offset) > PAGE_SIZE)
++ bytes = PAGE_SIZE - buffer.offset;
++ local_irq_save(flags); /* protects KM_BOUNCE_READ */
++ kmp = kmap_atomic(buffer.page, KM_BOUNCE_READ);
++ dev = dma_addr + size - len;
++ host = kmp + buffer.offset;
++ if (dir == DMA_FROM_DEVICE) {
++ if (__copy_to_user_inatomic(host, dev, bytes))
++ /* inaccessible */;
++ } else
++ memcpy(dev, host, bytes);
++ kunmap_atomic(kmp, KM_BOUNCE_READ);
++ local_irq_restore(flags);
++ len -= bytes;
++ buffer.page++;
++ buffer.offset = 0;
++ }
++ } else {
++ char *host = (char *)phys_to_virt(
++ page_to_pseudophys(buffer.page)) + buffer.offset;
++ if (dir == DMA_FROM_DEVICE) {
++ if (__copy_to_user_inatomic(host, dma_addr, size))
++ /* inaccessible */;
++ } else if (dir == DMA_TO_DEVICE)
++ memcpy(dma_addr, host, size);
++ }
++}
++
++/*
++ * Allocates bounce buffer and returns its kernel virtual address.
++ */
++static void *
++map_single(struct device *hwdev, struct phys_addr buffer, size_t size, int dir)
++{
++ unsigned long flags;
++ char *dma_addr;
++ unsigned int nslots, stride, index, wrap;
++ struct phys_addr slot_buf;
++ int i;
++ unsigned long mask;
++ unsigned long offset_slots;
++ unsigned long max_slots;
++
++ mask = dma_get_seg_boundary(hwdev);
++ offset_slots = -IO_TLB_SEGSIZE;
++ max_slots = mask + 1
++ ? ALIGN(mask + 1, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT
++ : 1UL << (BITS_PER_LONG - IO_TLB_SHIFT);
++
++ /*
++ * 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;
++
++ BUG_ON(!nslots);
++
++ /*
++ * 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);
++ index = ALIGN(io_tlb_index, stride);
++ if (index >= iotlb_nslabs)
++ index = 0;
++ wrap = index;
++
++ do {
++ while (iommu_is_span_boundary(index, nslots, offset_slots,
++ max_slots)) {
++ index += stride;
++ if (index >= iotlb_nslabs)
++ index = 0;
++ if (index == wrap)
++ goto not_found;
++ }
++
++ /*
++ * 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;
++
++ 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);
++
++ /*
++ * Update the indices to avoid searching in the next
++ * round.
++ */
++ io_tlb_index = ((index + nslots) < iotlb_nslabs
++ ? (index + nslots) : 0);
++
++ goto found;
++ }
++ index += stride;
++ if (index >= iotlb_nslabs)
++ index = 0;
++ } while (index != wrap);
++
++not_found:
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++ return NULL;
++found:
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++
++ /*
++ * Save away the mapping from the original address to the DMA address.
++ * This is needed when we sync the memory. Then we sync the buffer if
++ * needed.
++ */
++ slot_buf = buffer;
++ for (i = 0; i < nslots; i++) {
++ slot_buf.page += slot_buf.offset >> PAGE_SHIFT;
++ slot_buf.offset &= PAGE_SIZE - 1;
++ io_tlb_orig_addr[index+i] = slot_buf;
++ slot_buf.offset += 1 << IO_TLB_SHIFT;
++ }
++ if ((dir == DMA_TO_DEVICE) || (dir == DMA_BIDIRECTIONAL))
++ __sync_single(buffer, dma_addr, size, DMA_TO_DEVICE);
++
++ return dma_addr;
++}
++
++static struct phys_addr dma_addr_to_phys_addr(char *dma_addr)
++{
++ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
++ struct phys_addr buffer = io_tlb_orig_addr[index];
++ buffer.offset += (long)dma_addr & ((1 << IO_TLB_SHIFT) - 1);
++ buffer.page += buffer.offset >> PAGE_SHIFT;
++ buffer.offset &= PAGE_SIZE - 1;
++ return buffer;
++}
++
++/*
++ * 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;
++ int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT;
++ int index = (dma_addr - iotlb_virt_start) >> IO_TLB_SHIFT;
++ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
++
++ /*
++ * First, sync the memory before unmapping the entry
++ */
++ if ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL))
++ __sync_single(buffer, dma_addr, size, DMA_FROM_DEVICE);
++
++ /*
++ * Return the buffer to the free list by setting the corresponding
++ * entries to indicate the number of contigous entries available.
++ * While returning the entries to the free list, we merge the entries
++ * with slots below and above the pool being returned.
++ */
++ spin_lock_irqsave(&io_tlb_lock, flags);
++ {
++ count = ((index + nslots) < ALIGN(index + 1, IO_TLB_SEGSIZE) ?
++ io_tlb_list[index + nslots] : 0);
++ /*
++ * Step 1: return the slots to the free list, merging the
++ * slots with superceeding slots
++ */
++ for (i = index + nslots - 1; i >= index; i--)
++ io_tlb_list[i] = ++count;
++ /*
++ * Step 2: merge the returned slots with the preceding slots,
++ * if available (non zero)
++ */
++ for (i = index - 1;
++ (OFFSET(i, IO_TLB_SEGSIZE) !=
++ IO_TLB_SEGSIZE -1) && io_tlb_list[i];
++ i--)
++ io_tlb_list[i] = ++count;
++ }
++ spin_unlock_irqrestore(&io_tlb_lock, flags);
++}
++
++static void
++sync_single(struct device *hwdev, char *dma_addr, size_t size, int dir)
++{
++ struct phys_addr buffer = dma_addr_to_phys_addr(dma_addr);
++ BUG_ON((dir != DMA_FROM_DEVICE) && (dir != DMA_TO_DEVICE));
++ __sync_single(buffer, dma_addr, size, dir);
++}
++
++static void
++swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
++{
++ /*
++ * Ran out of IOMMU space for this operation. This is very bad.
++ * Unfortunately the drivers cannot handle this operation properly.
++ * unless they check for pci_dma_mapping_error (most don't)
++ * When the mapping is small enough return a static buffer to limit
++ * the damage, or panic when the transfer is too big.
++ */
++ printk(KERN_ERR "PCI-DMA: Out of SW-IOMMU space for %zu bytes at "
++ "device %s\n", (unsigned long)size, dev ? dev->bus_id : "?");
++
++ if (size > io_tlb_overflow && do_panic) {
++ if (dir == PCI_DMA_FROMDEVICE || dir == PCI_DMA_BIDIRECTIONAL)
++ panic("PCI-DMA: Memory would be corrupted\n");
++ if (dir == PCI_DMA_TODEVICE || dir == PCI_DMA_BIDIRECTIONAL)
++ panic("PCI-DMA: Random memory would be DMAed\n");
++ }
++}
++
++/*
++ * 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.
++ */
++static dma_addr_t
++_swiotlb_map_single(struct device *hwdev, phys_addr_t paddr, size_t size,
++ int dir, struct dma_attrs *attrs)
++{
++ struct page *page = pfn_to_page(paddr >> PAGE_SHIFT);
++ dma_addr_t dev_addr = gnttab_dma_map_page(page) +
++ offset_in_page(paddr);
++ void *map;
++ struct phys_addr buffer;
++
++ BUG_ON(dir == DMA_NONE);
++
++ /*
++ * If the pointer passed in happens to be in the device's DMA window,
++ * we can safely return the device addr and not worry about bounce
++ * buffering it.
++ */
++ if (!range_straddles_page_boundary(paddr, size) &&
++ !address_needs_mapping(hwdev, dev_addr))
++ return dev_addr;
++
++ /*
++ * Oh well, have to allocate and map a bounce buffer.
++ */
++ gnttab_dma_unmap_page(dev_addr);
++ buffer.page = page;
++ buffer.offset = offset_in_page(paddr);
++ 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;
++}
++
++dma_addr_t
++swiotlb_map_single_attrs(struct device *hwdev, void *ptr, size_t size,
++ int dir, struct dma_attrs *attrs)
++{
++ return _swiotlb_map_single(hwdev, virt_to_phys(ptr), size, dir, attrs);
++}
++EXPORT_SYMBOL(swiotlb_map_single_attrs);
++
++dma_addr_t
++swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir)
++{
++ return _swiotlb_map_single(hwdev, virt_to_phys(ptr), size, dir, NULL);
++}
++
++dma_addr_t
++swiotlb_map_single_phys(struct device *hwdev, phys_addr_t paddr, size_t size, int dir)
++{
++ return _swiotlb_map_single(hwdev, paddr, size, dir, NULL);
++}
++
++/*
++ * 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_attrs(struct device *hwdev, dma_addr_t dev_addr,
++ size_t size, int dir, struct dma_attrs *attrs)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ unmap_single(hwdev, bus_to_virt(dev_addr), size, dir);
++ else
++ gnttab_dma_unmap_page(dev_addr);
++}
++EXPORT_SYMBOL(swiotlb_unmap_single_attrs);
++
++void
++swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size,
++ int dir)
++{
++ return swiotlb_unmap_single_attrs(hwdev, dev_addr, size, dir, NULL);
++}
++/*
++ * 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);
++}
++
++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);
++}
++
++void
++swiotlb_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dev_addr,
++ unsigned long offset, size_t size, int dir)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ sync_single(hwdev, bus_to_virt(dev_addr + offset), size, dir);
++}
++
++void
++swiotlb_sync_single_range_for_device(struct device *hwdev, dma_addr_t dev_addr,
++ unsigned long offset, size_t size, int dir)
++{
++ BUG_ON(dir == DMA_NONE);
++ if (in_swiotlb_aperture(dev_addr))
++ sync_single(hwdev, bus_to_virt(dev_addr + offset), size, dir);
++}
++
++void swiotlb_unmap_sg_attrs(struct device *, struct scatterlist *, int, int,
++ struct dma_attrs *);
++/*
++ * 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_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems,
++ int dir, struct dma_attrs *attrs)
++{
++ struct scatterlist *sg;
++ struct phys_addr buffer;
++ dma_addr_t dev_addr;
++ char *map;
++ int i;
++
++ BUG_ON(dir == DMA_NONE);
++
++ for_each_sg(sgl, sg, nelems, i) {
++ dev_addr = gnttab_dma_map_page(sg_page(sg)) + sg->offset;
++
++ if (range_straddles_page_boundary(page_to_pseudophys(sg_page(sg))
++ + sg->offset, sg->length)
++ || address_needs_mapping(hwdev, dev_addr)) {
++ gnttab_dma_unmap_page(dev_addr);
++ buffer.page = sg_page(sg);
++ 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_attrs(hwdev, sgl, i, dir,
++ attrs);
++ sgl[0].dma_length = 0;
++ return 0;
++ }
++ sg->dma_address = virt_to_bus(map);
++ } else
++ sg->dma_address = dev_addr;
++ sg->dma_length = sg->length;
++ }
++ return nelems;
++}
++EXPORT_SYMBOL(swiotlb_map_sg_attrs);
++
++int
++swiotlb_map_sg(struct device *hwdev, struct scatterlist *sgl, int nelems,
++ int dir)
++{
++ return swiotlb_map_sg_attrs(hwdev, sgl, nelems, dir, NULL);
++}
++
++/*
++ * 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_attrs(struct device *hwdev, struct scatterlist *sgl,
++ int nelems, int dir, struct dma_attrs *attrs)
++{
++ struct scatterlist *sg;
++ int i;
++
++ BUG_ON(dir == DMA_NONE);
++
++ for_each_sg(sgl, sg, nelems, i) {
++ if (in_swiotlb_aperture(sg->dma_address))
++ unmap_single(hwdev, bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++ else
++ gnttab_dma_unmap_page(sg->dma_address);
++ }
++}
++EXPORT_SYMBOL(swiotlb_unmap_sg_attrs);
++
++void
++swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sgl, int nelems,
++ int dir)
++{
++ return swiotlb_unmap_sg_attrs(hwdev, sgl, nelems, dir, NULL);
++}
++
++/*
++ * 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 *sgl,
++ int nelems, int dir)
++{
++ struct scatterlist *sg;
++ int i;
++
++ BUG_ON(dir == DMA_NONE);
++
++ for_each_sg(sgl, sg, nelems, i) {
++ if (in_swiotlb_aperture(sg->dma_address))
++ sync_single(hwdev, bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++ }
++}
++
++void
++swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sgl,
++ int nelems, int dir)
++{
++ struct scatterlist *sg;
++ int i;
++
++ BUG_ON(dir == DMA_NONE);
++
++ for_each_sg(sgl, sg, nelems, i) {
++ if (in_swiotlb_aperture(sg->dma_address))
++ sync_single(hwdev, bus_to_virt(sg->dma_address),
++ sg->dma_length, dir);
++ }
++}
++
++int
++swiotlb_dma_mapping_error(dma_addr_t dma_addr)
++{
++ return (dma_addr == virt_to_bus(io_tlb_overflow_buffer));
++}
++
++/*
++ * Return whether the given PCI device DMA address mask can be supported
++ * properly. For example, if your device can only drive the low 24-bits
++ * during PCI bus mastering, then you would pass 0x00ffffff as the mask to
++ * this function.
++ */
++int
++swiotlb_dma_supported (struct device *hwdev, u64 mask)
++{
++ return (mask >= ((1UL << dma_bits) - 1));
++}
++
++EXPORT_SYMBOL(swiotlb_map_single);
++EXPORT_SYMBOL(swiotlb_unmap_single);
++EXPORT_SYMBOL(swiotlb_map_sg);
++EXPORT_SYMBOL(swiotlb_unmap_sg);
++EXPORT_SYMBOL(swiotlb_sync_single_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_single_for_device);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu);
++EXPORT_SYMBOL(swiotlb_sync_sg_for_device);
++EXPORT_SYMBOL(swiotlb_dma_mapping_error);
++EXPORT_SYMBOL(swiotlb_dma_supported);
+diff --git a/mm/memory.c b/mm/memory.c
+index 0755c52..8a2ba84 100644
+--- a/mm/memory.c
++++ b/mm/memory.c
+@@ -467,7 +467,19 @@ struct page *vm_normal_page(struct vm_area_struct *vma, unsigned long addr,
+ }
+ }
+
++#ifndef CONFIG_XEN
+ VM_BUG_ON(!pfn_valid(pfn));
++#else
++#ifdef CONFIG_X86
++ /* XEN: Covers user-space grant mappings (even of local pages). */
++ if (unlikely(vma->vm_flags & VM_FOREIGN))
++ return NULL;
++#endif
++ if (unlikely(!pfn_valid(pfn))) {
++ VM_BUG_ON(!(vma->vm_flags & VM_RESERVED));
++ return NULL;
++ }
++#endif
+
+ /*
+ * NOTE! We still have PageReserved() pages in the page tables.
+@@ -721,8 +733,14 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
+ page->index > details->last_index))
+ continue;
+ }
+- ptent = ptep_get_and_clear_full(mm, addr, pte,
+- tlb->fullmm);
++#ifdef CONFIG_XEN
++ if (unlikely(vma->vm_ops && vma->vm_ops->zap_pte))
++ ptent = vma->vm_ops->zap_pte(vma, addr, pte,
++ tlb->fullmm);
++ else
++#endif
++ ptent = ptep_get_and_clear_full(mm, addr, pte,
++ tlb->fullmm);
+ tlb_remove_tlb_entry(tlb, pte, addr);
+ if (unlikely(!page))
+ continue;
+@@ -955,6 +973,7 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address,
+ tlb_finish_mmu(tlb, address, end);
+ return end;
+ }
++EXPORT_SYMBOL(zap_page_range);
+
+ /*
+ * Do a quick page-table lookup for a single page.
+@@ -1126,6 +1145,26 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
+ continue;
+ }
+
++#ifdef CONFIG_XEN
++ if (vma && (vma->vm_flags & VM_FOREIGN)) {
++ struct page **map = vma->vm_private_data;
++ int offset = (start - vma->vm_start) >> PAGE_SHIFT;
++ if (map[offset] != NULL) {
++ if (pages) {
++ struct page *page = map[offset];
++
++ pages[i] = page;
++ get_page(page);
++ }
++ if (vmas)
++ vmas[i] = vma;
++ i++;
++ start += PAGE_SIZE;
++ len--;
++ continue;
++ }
++ }
++#endif
+ if (!vma || (vma->vm_flags & (VM_IO | VM_PFNMAP))
+ || !(vm_flags & vma->vm_flags))
+ return i ? : -EFAULT;
+diff --git a/mm/mprotect.c b/mm/mprotect.c
+index a5bf31c..e943715 100644
+--- a/mm/mprotect.c
++++ b/mm/mprotect.c
+@@ -93,6 +93,8 @@ static inline void change_pmd_range(struct mm_struct *mm, pud_t *pud,
+ next = pmd_addr_end(addr, end);
+ if (pmd_none_or_clear_bad(pmd))
+ continue;
++ if (arch_change_pte_range(mm, pmd, addr, next, newprot, dirty_accountable))
++ continue;
+ change_pte_range(mm, pmd, addr, next, newprot, dirty_accountable);
+ } while (pmd++, addr = next, addr != end);
+ }
+diff --git a/mm/page_alloc.c b/mm/page_alloc.c
+index 0d520dc..0603a08 100644
+--- a/mm/page_alloc.c
++++ b/mm/page_alloc.c
+@@ -510,6 +510,12 @@ static void __free_pages_ok(struct page *page, unsigned int order)
+ int i;
+ int reserved = 0;
+
++#ifdef CONFIG_XEN
++ if (PageForeign(page)) {
++ PageForeignDestructor(page);
++ return;
++ }
++#endif
+ for (i = 0 ; i < (1 << order) ; ++i)
+ reserved += free_pages_check(page + i);
+ if (reserved)
+@@ -969,6 +975,12 @@ static void free_hot_cold_page(struct page *page, int cold)
+ struct per_cpu_pages *pcp;
+ unsigned long flags;
+
++#ifdef CONFIG_XEN
++ if (PageForeign(page)) {
++ PageForeignDestructor(page);
++ return;
++ }
++#endif
+ if (PageAnon(page))
+ page->mapping = NULL;
+ if (free_pages_check(page))
+diff --git a/net/core/dev.c b/net/core/dev.c
+index fca23a3..5c80161 100644
+--- a/net/core/dev.c
++++ b/net/core/dev.c
+@@ -123,6 +123,12 @@
+
+ #include "net-sysfs.h"
+
++#if defined(CONFIG_XEN) || defined(CONFIG_PARAVIRT_XEN)
++#include <net/ip.h>
++#include <linux/tcp.h>
++#include <linux/udp.h>
++#endif
++
+ /*
+ * The list of packet types we will receive (as opposed to discard)
+ * and the routines to invoke.
+@@ -1608,6 +1614,58 @@ out_kfree_skb:
+ return 0;
+ }
+
++#if defined(CONFIG_XEN) || defined(CONFIG_PARAVIRT_XEN)
++inline int skb_checksum_setup(struct sk_buff *skb)
++{
++ struct iphdr *iph;
++ unsigned char *th;
++ int err = -EPROTO;
++
++#ifdef CONFIG_XEN
++ if (!skb->proto_csum_blank)
++ return 0;
++#endif
++
++ if (skb->protocol != htons(ETH_P_IP))
++ goto out;
++
++ iph = ip_hdr(skb);
++ th = skb_network_header(skb) + 4 * iph->ihl;
++ if (th >= skb_tail_pointer(skb))
++ goto out;
++
++ skb->csum_start = th - skb->head;
++ switch (iph->protocol) {
++ case IPPROTO_TCP:
++ skb->csum_offset = offsetof(struct tcphdr, check);
++ break;
++ case IPPROTO_UDP:
++ skb->csum_offset = offsetof(struct udphdr, check);
++ break;
++ default:
++ if (net_ratelimit())
++ printk(KERN_ERR "Attempting to checksum a non-"
++ "TCP/UDP packet, dropping a protocol"
++ " %d packet", iph->protocol);
++ goto out;
++ }
++
++ if ((th + skb->csum_offset + 2) > skb_tail_pointer(skb))
++ goto out;
++
++#ifdef CONFIG_XEN
++ skb->ip_summed = CHECKSUM_PARTIAL;
++ skb->proto_csum_blank = 0;
++#endif
++
++ err = 0;
++
++out:
++ return err;
++}
++EXPORT_SYMBOL(skb_checksum_setup);
++#endif
++
+ /**
+ * dev_queue_xmit - transmit a buffer
+ * @skb: buffer to transmit
+@@ -1640,6 +1698,12 @@ int dev_queue_xmit(struct sk_buff *skb)
+ struct Qdisc *q;
+ int rc = -ENOMEM;
+
++ /* If a checksum-deferred packet is forwarded to a device that needs a
++ * checksum, correct the pointers and force checksumming.
++ */
++ if (skb_checksum_setup(skb))
++ goto out_kfree_skb;
++
+ /* GSO will handle the following emulations directly. */
+ if (netif_needs_gso(dev, skb))
+ goto gso;
+@@ -2088,6 +2152,19 @@ int netif_receive_skb(struct sk_buff *skb)
+ }
+ #endif
+
++#ifdef CONFIG_XEN
++ switch (skb->ip_summed) {
++ case CHECKSUM_UNNECESSARY:
++ skb->proto_data_valid = 1;
++ break;
++ case CHECKSUM_PARTIAL:
++ /* XXX Implement me. */
++ default:
++ skb->proto_data_valid = 0;
++ break;
++ }
++#endif
++
+ list_for_each_entry_rcu(ptype, &ptype_all, list) {
+ if (!ptype->dev || ptype->dev == skb->dev) {
+ if (pt_prev)
+diff --git a/net/core/skbuff.c b/net/core/skbuff.c
+index 3666216..9128cbb 100644
+--- a/net/core/skbuff.c
++++ b/net/core/skbuff.c
+@@ -478,6 +478,10 @@ static struct sk_buff *__skb_clone(struct sk_buff *n, struct sk_buff *skb)
+ n->hdr_len = skb->nohdr ? skb_headroom(skb) : skb->hdr_len;
+ n->cloned = 1;
+ n->nohdr = 0;
++#ifdef CONFIG_XEN
++ C(proto_data_valid);
++ C(proto_csum_blank);
++#endif
+ n->destructor = NULL;
+ C(iif);
+ C(tail);
+diff --git a/net/ipv4/netfilter/nf_nat_proto_tcp.c b/net/ipv4/netfilter/nf_nat_proto_tcp.c
+index 399e2cf..17445c6 100644
+--- a/net/ipv4/netfilter/nf_nat_proto_tcp.c
++++ b/net/ipv4/netfilter/nf_nat_proto_tcp.c
+@@ -75,6 +75,9 @@ tcp_manip_pkt(struct sk_buff *skb,
+ if (hdrsize < sizeof(*hdr))
+ return true;
+
++ if (skb_checksum_setup(skb))
++ return false;
++
+ inet_proto_csum_replace4(&hdr->check, skb, oldip, newip, 1);
+ inet_proto_csum_replace2(&hdr->check, skb, oldport, newport, 0);
+ return true;
+diff --git a/net/ipv4/netfilter/nf_nat_proto_udp.c b/net/ipv4/netfilter/nf_nat_proto_udp.c
+index 9e61c79..279ccc4 100644
+--- a/net/ipv4/netfilter/nf_nat_proto_udp.c
++++ b/net/ipv4/netfilter/nf_nat_proto_udp.c
+@@ -60,6 +60,10 @@ udp_manip_pkt(struct sk_buff *skb,
+ newport = tuple->dst.u.udp.port;
+ portptr = &hdr->dest;
+ }
++
++ if (skb_checksum_setup(skb))
++ return false;
++
+ if (hdr->check || skb->ip_summed == CHECKSUM_PARTIAL) {
+ inet_proto_csum_replace4(&hdr->check, skb, oldip, newip, 1);
+ inet_proto_csum_replace2(&hdr->check, skb, *portptr, newport,
+diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c
+index 8c3180a..f5cee36 100644
+--- a/net/ipv4/xfrm4_output.c
++++ b/net/ipv4/xfrm4_output.c
+@@ -81,7 +81,7 @@ static int xfrm4_output_finish(struct sk_buff *skb)
+ #endif
+
+ skb->protocol = htons(ETH_P_IP);
+- return xfrm_output(skb);
++ return skb_checksum_setup(skb) ?: xfrm_output(skb);
+ }
+
+ int xfrm4_output(struct sk_buff *skb)
+diff --git a/scripts/Makefile.build b/scripts/Makefile.build
+index 277cfe0..e89ae81 100644
+--- a/scripts/Makefile.build
++++ b/scripts/Makefile.build
+@@ -73,6 +73,20 @@ ifndef obj
+ $(warning kbuild: Makefile.build is included improperly)
+ endif
+
++ifeq ($(CONFIG_XEN),y)
++$(objtree)/scripts/Makefile.xen: $(srctree)/scripts/Makefile.xen.awk $(srctree)/scripts/Makefile.build
++ @echo ' Updating $@'
++ $(if $(shell echo a | $(AWK) '{ print gensub(/a/, "AA", "g"); }'),\
++ ,$(error 'Your awk program does not define gensub. Use gawk or another awk with gensub'))
++ @$(AWK) -f $< $(filter-out $<,$^) >$@
++
++xen-src-single-used-m := $(patsubst $(srctree)/%,%,$(wildcard $(addprefix $(srctree)/,$(single-used-m:.o=-xen.c))))
++xen-single-used-m := $(xen-src-single-used-m:-xen.c=.o)
++single-used-m := $(filter-out $(xen-single-used-m),$(single-used-m))
++
++-include $(objtree)/scripts/Makefile.xen
++endif
++
+ # ===========================================================================
+
+ ifneq ($(strip $(lib-y) $(lib-m) $(lib-n) $(lib-)),)
+diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
+index 8e44023..8b9f10f 100644
+--- a/scripts/Makefile.lib
++++ b/scripts/Makefile.lib
+@@ -17,6 +17,12 @@ obj-m := $(filter-out $(obj-y),$(obj-m))
+
+ lib-y := $(filter-out $(obj-y), $(sort $(lib-y) $(lib-m)))
+
++# Remove objects forcibly disabled
++
++obj-y := $(filter-out $(disabled-obj-y),$(obj-y))
++obj-m := $(filter-out $(disabled-obj-y),$(obj-m))
++lib-y := $(filter-out $(disabled-obj-y),$(lib-y))
++
+
+ # Handle objects in subdirs
+ # ---------------------------------------------------------------------------
+diff --git a/scripts/Makefile.xen.awk b/scripts/Makefile.xen.awk
+new file mode 100644
+index 0000000..1f7cf1e
+--- /dev/null
++++ b/scripts/Makefile.xen.awk
+@@ -0,0 +1,34 @@
++BEGIN {
++ is_rule = 0
++}
++
++/^[[:space:]]*#/ {
++ next
++}
++
++/^[[:space:]]*$/ {
++ if (is_rule)
++ print("")
++ is_rule = 0
++ next
++}
++
++/:[[:space:]]*\$\(src\)\/%\.[cS][[:space:]]/ {
++ line = gensub(/%.([cS])/, "%-xen.\\1", "g", $0)
++ line = gensub(/(single-used-m)/, "xen-\\1", "g", line)
++ print line
++ is_rule = 1
++ next
++}
++
++/^[^\t]$/ {
++ if (is_rule)
++ print("")
++ is_rule = 0
++ next
++}
++
++is_rule {
++ print $0
++ next
++}
Added: dists/trunk/linux-2.6/debian/patches/features/arm/dns323_mac_addr_read.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/arm/dns323_mac_addr_read.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,111 @@
+Matthew Palmer <mpalmer at hezmatt.org>
+Subject: DNS-323: read MAC address from flash
+
+--- a/arch/arm/mach-orion5x/dns323-setup.c 2008-10-17 21:56:01.000000000 +1100
++++ b/arch/arm/mach-orion5x/dns323-setup.c 2008-10-17 23:04:16.000000000 +1100
+@@ -76,15 +76,6 @@
+ subsys_initcall(dns323_pci_init);
+
+ /****************************************************************************
+- * Ethernet
+- */
+-
+-static struct mv643xx_eth_platform_data dns323_eth_data = {
+- .phy_addr = 8,
+- .force_phy_addr = 1,
+-};
+-
+-/****************************************************************************
+ * 8MiB NOR flash (Spansion S29GL064M90TFIR4)
+ *
+ * Layout as used by D-Link:
+@@ -143,6 +134,79 @@
+ };
+
+ /****************************************************************************
++ * Ethernet
++ */
++
++static struct mv643xx_eth_platform_data dns323_eth_data = {
++ .phy_addr = 8,
++ .force_phy_addr = 1,
++};
++
++/* parse_hex_*() taken from ts209-setup.c; should a common copy of these
++ * functions be kept somewhere?
++ */
++static int __init parse_hex_nibble(char n)
++{
++ if (n >= '0' && n <= '9')
++ return n - '0';
++
++ if (n >= 'A' && n <= 'F')
++ return n - 'A' + 10;
++
++ if (n >= 'a' && n <= 'f')
++ return n - 'a' + 10;
++
++ return -1;
++}
++
++static int __init parse_hex_byte(const char *b)
++{
++ int hi;
++ int lo;
++
++ hi = parse_hex_nibble(b[0]);
++ lo = parse_hex_nibble(b[1]);
++
++ if (hi < 0 || lo < 0)
++ return -1;
++
++ return (hi << 4) | lo;
++}
++
++static int __init dns323_read_mac_addr(void)
++{
++ u_int8_t addr[6] = { 0x00, 0x00, 0xDE, 0xAD, 0xBE, 0xEF };
++ int i;
++ char *mac_page;
++
++ /* MAC address is stored as a regular ol' string in /dev/mtdblock4
++ * (0x007d0000-0x00800000) starting at offset 196480 (0x2ff80).
++ */
++ mac_page = ioremap(DNS323_NOR_BOOT_BASE + 0x7d0000 + 196480, 1024);
++
++ for (i = 0; i < 6; i++) {
++ int byte;
++
++ byte = parse_hex_byte(mac_page + (i * 3));
++ if (byte < 0) {
++ iounmap(mac_page);
++ return -1;
++ }
++
++ addr[i] = byte;
++ }
++
++ iounmap(mac_page);
++ printk("DNS323: Found ethernet MAC address: ");
++ for (i = 0; i < 6; i++)
++ printk("%.2x%s", addr[i], (i < 5) ? ":" : ".\n");
++
++ memcpy(dns323_eth_data.mac_addr, addr, 6);
++
++ return 0;
++}
++
++/****************************************************************************
+ * GPIO LEDs (simple - doesn't use hardware blinking support)
+ */
+
+@@ -305,6 +369,9 @@
+ i2c_register_board_info(0, dns323_i2c_devices,
+ ARRAY_SIZE(dns323_i2c_devices));
+
++ if (dns323_read_mac_addr() < 0)
++ printk("ERROR: Failed to read MAC address\n");
++
+ orion5x_eth_init(&dns323_eth_data);
+ }
+
Added: dists/trunk/linux-2.6/debian/patches/features/arm/dns323_revb1.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/arm/dns323_revb1.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,96 @@
+--- a/arch/arm/mach-orion5x/dns323-setup.c 2008-10-23 21:16:57.000000000 +1100
++++ b/arch/arm/mach-orion5x/dns323-setup.c 2008-10-23 21:17:12.000000000 +1100
+@@ -21,6 +21,7 @@
+ #include <linux/gpio_keys.h>
+ #include <linux/input.h>
+ #include <linux/i2c.h>
++#include <linux/ata_platform.h>
+ #include <asm/mach-types.h>
+ #include <asm/gpio.h>
+ #include <asm/mach/arch.h>
+@@ -67,8 +68,18 @@
+
+ static int __init dns323_pci_init(void)
+ {
+- if (machine_is_dns323())
+- pci_common_init(&dns323_pci);
++ u32 dev, rev;
++
++ orion5x_pcie_id(&dev, &rev);
++
++ if (machine_is_dns323()) {
++ if (dev != MV88F5182_DEV_ID) {
++ /* The 5182 doesn't really use it's PCI bus, so
++ * we don't initialise it.
++ */
++ pci_common_init(&dns323_pci);
++ }
++ }
+
+ return 0;
+ }
+@@ -268,6 +279,13 @@
+ .dev = { .platform_data = &dns323_button_data, },
+ };
+
++/*****************************************************************************
++ * SATA
++ ****************************************************************************/
++static struct mv_sata_platform_data dns323_sata_data = {
++ .n_ports = 2,
++};
++
+ /****************************************************************************
+ * General Setup
+ */
+@@ -311,6 +329,11 @@
+
+ static void __init dns323_init(void)
+ {
++ u32 rev, dev;
++
++ /* Whooooooo are we? Who who, who who? */
++ orion5x_pcie_id(&dev, &rev);
++
+ /* Setup basic Orion functions. Need to be called early. */
+ orion5x_init();
+
+@@ -326,11 +349,21 @@
+ orion5x_setup_pcie_wa_win(ORION5X_PCIE_WA_PHYS_BASE,
+ ORION5X_PCIE_WA_SIZE);
+
+- /* set MPP to 0 as D-Link's 2.6.12.6 kernel did */
+- orion5x_write(MPP_0_7_CTRL, 0);
+- orion5x_write(MPP_8_15_CTRL, 0);
+- orion5x_write(MPP_16_19_CTRL, 0);
+- orion5x_write(MPP_DEV_CTRL, 0);
++ if (dev == MV88F5182_DEV_ID) {
++ /* The 5182 has a different MPP map, so we wire it up this way
++ * to let the HDD LEDs do their thing. Values taken directly
++ * from the D-Link kernel.
++ */
++ orion5x_write(MPP_0_7_CTRL, 0x3);
++ orion5x_write(MPP_8_15_CTRL, 0x55550000);
++ orion5x_write(MPP_16_19_CTRL, 0x5555);
++ } else {
++ /* set MPP to 0 as D-Link's 2.6.12.6 kernel did */
++ orion5x_write(MPP_0_7_CTRL, 0);
++ orion5x_write(MPP_8_15_CTRL, 0);
++ orion5x_write(MPP_16_19_CTRL, 0);
++ orion5x_write(MPP_DEV_CTRL, 0);
++ }
+
+ /* Define used GPIO pins
+
+@@ -373,6 +406,11 @@
+ printk("ERROR: Failed to read MAC address\n");
+
+ orion5x_eth_init(&dns323_eth_data);
++ /* The 5182 has it's SATA controller internally, and it needs it's own
++ * little init routine.
++ */
++ if (dev == MV88F5182_DEV_ID)
++ orion5x_sata_init(&dns323_sata_data);
+ }
+
+ /* Warning: D-Link uses a wrong mach-type (=526) in their bootloader */
Added: dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-mem-leak.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-mem-leak.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,79 @@
+From: Sven Wegener <sven.wegener at stealer.net>
+Subject: [PATCH] leds-pca9532: Fix memory leak and properly handle errors
+
+When the registration fails, we need to release the memory we allocated. Also
+we need to save the error from led_classdev_register and propagate it up, else
+we'll return success, even if we failed.
+
+Signed-off-by: Riku Voipio <riku.voipio at iki.fi>
+---
+ drivers/leds/leds-pca9532.c | 22 +++++++++++++---------
+ 1 files changed, 13 insertions(+), 9 deletions(-)
+
+I don't have the hardware. Riku, can you please test if it still works as
+expected.
+
+diff --git a/drivers/leds/leds-pca9532.c b/drivers/leds/leds-pca9532.c
+index 4064d4f..16af817 100644
+--- a/drivers/leds/leds-pca9532.c
++++ b/drivers/leds/leds-pca9532.c
+@@ -204,8 +204,8 @@ static int pca9532_configure(struct i2c_client *client,
+ led->ldev.brightness = LED_OFF;
+ led->ldev.brightness_set = pca9532_set_brightness;
+ led->ldev.blink_set = pca9532_set_blink;
+- if (led_classdev_register(&client->dev,
+- &led->ldev) < 0) {
++ err = led_classdev_register(&client->dev, &led->ldev);
++ if (err < 0) {
+ dev_err(&client->dev,
+ "couldn't register LED %s\n",
+ led->name);
+@@ -263,7 +263,6 @@ exit:
+ }
+
+ return err;
+-
+ }
+
+ static int pca9532_probe(struct i2c_client *client,
+@@ -271,12 +270,16 @@ static int pca9532_probe(struct i2c_client *client,
+ {
+ struct pca9532_data *data = i2c_get_clientdata(client);
+ struct pca9532_platform_data *pca9532_pdata = client->dev.platform_data;
++ int err;
++
++ if (!pca9532_pdata)
++ return -EIO;
+
+ if (!i2c_check_functionality(client->adapter,
+ I2C_FUNC_SMBUS_BYTE_DATA))
+ return -EIO;
+
+- data = kzalloc(sizeof(struct pca9532_data), GFP_KERNEL);
++ data = kzalloc(sizeof(*data), GFP_KERNEL);
+ if (!data)
+ return -ENOMEM;
+
+@@ -285,12 +288,13 @@ static int pca9532_probe(struct i2c_client *client,
+ data->client = client;
+ mutex_init(&data->update_lock);
+
+- if (pca9532_pdata == NULL)
+- return -EIO;
+-
+- pca9532_configure(client, data, pca9532_pdata);
+- return 0;
++ err = pca9532_configure(client, data, pca9532_pdata);
++ if (err) {
++ kfree(data);
++ i2c_set_clientdata(client, NULL);
++ }
+
++ return err;
+ }
+
+ static int pca9532_remove(struct i2c_client *client)
+
+
+--
+"rm -rf" only sounds scary if you don't have backups
Added: dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-workqueque.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/arm/led-pca9532-workqueque.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,199 @@
+From: Riku Voipio <riku.voipio at iki.fi>
+Subject: [PATCH] Move i2c work to a workqueque
+
+Apparently these might be called under atomic context,
+and i2c operations may sleep. BUG found by
+Ross Burton <ross at burtonini.com>
+
+Signed-off-by: Riku Voipio <riku.voipio at iki.fi>
+---
+ drivers/leds/leds-pca9532.c | 51 +++++++++++++++++++++++++++++++++++------
+ include/linux/leds-pca9532.h | 2 +
+ 2 files changed, 45 insertions(+), 8 deletions(-)
+
+diff --git a/drivers/leds/leds-pca9532.c b/drivers/leds/leds-pca9532.c
+index 62b60a0..cb8e517 100644
+--- a/drivers/leds/leds-pca9532.c
++++ b/drivers/leds/leds-pca9532.c
+@@ -16,6 +16,7 @@
+ #include <linux/leds.h>
+ #include <linux/input.h>
+ #include <linux/mutex.h>
++#include <linux/workqueue.h>
+ #include <linux/leds-pca9532.h>
+
+ static const unsigned short normal_i2c[] = { /*0x60,*/ I2C_CLIENT_END};
+@@ -34,6 +35,7 @@ struct pca9532_data {
+ struct pca9532_led leds[16];
+ struct mutex update_lock;
+ struct input_dev *idev;
++ struct work_struct work;
+ u8 pwm[2];
+ u8 psc[2];
+ };
+@@ -63,7 +65,7 @@ static struct i2c_driver pca9532_driver = {
+ * as a compromise we average one pwm to the values requested by all
+ * leds that are not ON/OFF.
+ * */
+-static int pca9532_setpwm(struct i2c_client *client, int pwm, int blink,
++static int pca9532_calcpwm(struct i2c_client *client, int pwm, int blink,
+ enum led_brightness value)
+ {
+ int a = 0, b = 0, i = 0;
+@@ -84,11 +86,17 @@ static int pca9532_setpwm(struct i2c_client *client, int pwm, int blink,
+ b = b/a;
+ if (b > 0xFF)
+ return -EINVAL;
+- mutex_lock(&data->update_lock);
+ data->pwm[pwm] = b;
++ data->psc[pwm] = blink;
++ return 0;
++}
++
++static int pca9532_setpwm(struct i2c_client *client, int pwm)
++{
++ struct pca9532_data *data = i2c_get_clientdata(client);
++ mutex_lock(&data->update_lock);
+ i2c_smbus_write_byte_data(client, PCA9532_REG_PWM(pwm),
+ data->pwm[pwm]);
+- data->psc[pwm] = blink;
+ i2c_smbus_write_byte_data(client, PCA9532_REG_PSC(pwm),
+ data->psc[pwm]);
+ mutex_unlock(&data->update_lock);
+@@ -124,11 +132,11 @@ static void pca9532_set_brightness(struct led_classdev *led_cdev,
+ led->state = PCA9532_ON;
+ else {
+ led->state = PCA9532_PWM0; /* Thecus: hardcode one pwm */
+- err = pca9532_setpwm(led->client, 0, 0, value);
++ err = pca9532_calcpwm(led->client, 0, 0, value);
+ if (err)
+ return; /* XXX: led api doesn't allow error code? */
+ }
+- pca9532_setled(led);
++ schedule_work(&led->work);
+ }
+
+ static int pca9532_set_blink(struct led_classdev *led_cdev,
+@@ -137,6 +145,7 @@ static int pca9532_set_blink(struct led_classdev *led_cdev,
+ struct pca9532_led *led = ldev_to_led(led_cdev);
+ struct i2c_client *client = led->client;
+ int psc;
++ int err = 0;
+
+ if (*delay_on == 0 && *delay_off == 0) {
+ /* led subsystem ask us for a blink rate */
+@@ -148,7 +157,11 @@ static int pca9532_set_blink(struct led_classdev *led_cdev,
+
+ /* Thecus specific: only use PSC/PWM 0 */
+ psc = (*delay_on * 152-1)/1000;
+- return pca9532_setpwm(client, 0, psc, led_cdev->brightness);
++ err = pca9532_calcpwm(client, 0, psc, led_cdev->brightness);
++ if (err)
++ return err;
++ schedule_work(&led->work);
++ return 0;
+ }
+
+ static int pca9532_event(struct input_dev *dev, unsigned int type,
+@@ -165,13 +178,28 @@ static int pca9532_event(struct input_dev *dev, unsigned int type,
+ else
+ data->pwm[1] = 0;
+
+- dev_info(&dev->dev, "setting beep to %d \n", data->pwm[1]);
++ schedule_work(&data->work);
++
++ return 0;
++}
++
++static void pca9532_input_work(struct work_struct *work)
++{
++ struct pca9532_data *data;
++ data = container_of(work, struct pca9532_data, work);
+ mutex_lock(&data->update_lock);
+ i2c_smbus_write_byte_data(data->client, PCA9532_REG_PWM(1),
+ data->pwm[1]);
+ mutex_unlock(&data->update_lock);
++}
+
+- return 0;
++static void pca9532_led_work(struct work_struct *work)
++{
++ struct pca9532_led *led;
++ led = container_of(work, struct pca9532_led, work);
++ if (led->state == PCA9532_PWM0)
++ pca9532_setpwm(led->client, 0);
++ pca9532_setled(led);
+ }
+
+ static int pca9532_configure(struct i2c_client *client,
+@@ -204,6 +232,7 @@ static int pca9532_configure(struct i2c_client *client,
+ led->ldev.brightness = LED_OFF;
+ led->ldev.brightness_set = pca9532_set_brightness;
+ led->ldev.blink_set = pca9532_set_blink;
++ INIT_WORK(&led->work, pca9532_led_work);
+ err = led_classdev_register(&client->dev, &led->ldev);
+ if (err < 0) {
+ dev_err(&client->dev,
+@@ -233,9 +262,11 @@ static int pca9532_configure(struct i2c_client *client,
+ BIT_MASK(SND_TONE);
+ data->idev->event = pca9532_event;
+ input_set_drvdata(data->idev, data);
++ INIT_WORK(&data->work, pca9532_input_work);
+ err = input_register_device(data->idev);
+ if (err) {
+ input_free_device(data->idev);
++ cancel_work_sync(&data->work);
+ data->idev = NULL;
+ goto exit;
+ }
+@@ -252,11 +283,13 @@ exit:
+ break;
+ case PCA9532_TYPE_LED:
+ led_classdev_unregister(&data->leds[i].ldev);
++ cancel_work_sync(&data->leds[i].work);
+ break;
+ case PCA9532_TYPE_N2100_BEEP:
+ if (data->idev != NULL) {
+ input_unregister_device(data->idev);
+ input_free_device(data->idev);
++ cancel_work_sync(&data->work);
+ data->idev = NULL;
+ }
+ break;
+@@ -307,11 +340,13 @@ static int pca9532_remove(struct i2c_client *client)
+ break;
+ case PCA9532_TYPE_LED:
+ led_classdev_unregister(&data->leds[i].ldev);
++ cancel_work_sync(&data->leds[i].work);
+ break;
+ case PCA9532_TYPE_N2100_BEEP:
+ if (data->idev != NULL) {
+ input_unregister_device(data->idev);
+ input_free_device(data->idev);
++ cancel_work_sync(&data->work);
+ data->idev = NULL;
+ }
+ break;
+diff --git a/include/linux/leds-pca9532.h b/include/linux/leds-pca9532.h
+index 81b4207..f158eb1 100644
+--- a/include/linux/leds-pca9532.h
++++ b/include/linux/leds-pca9532.h
+@@ -15,6 +15,7 @@
+ #define __LINUX_PCA9532_H
+
+ #include <linux/leds.h>
++#include <linux/workqueue.h>
+
+ enum pca9532_state {
+ PCA9532_OFF = 0x0,
+@@ -31,6 +32,7 @@ struct pca9532_led {
+ struct i2c_client *client;
+ char *name;
+ struct led_classdev ldev;
++ struct work_struct work;
+ enum pca9532_type type;
+ enum pca9532_state state;
+ };
+--
+1.5.6.5
+
Added: dists/trunk/linux-2.6/debian/patches/features/arm/ts409-export-reset.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/arm/ts409-export-reset.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,28 @@
+From: Martin Michlmayr <tbm at cyrius.com>
+Subject: [PATCH] Orion: Export the reset button of the QNAP TS-409
+
+The reset button on the QNAP TS-409 is available through gpio.
+
+Signed-off-by: Martin Michlmayr <tbm at cyrius.com>
+
+--- a/arch/arm/mach-orion5x/ts409-setup.c 2008-07-31 04:19:08.000000000 +0000
++++ b/arch/arm/mach-orion5x/ts409-setup.c 2008-07-31 04:21:10.000000000 +0000
+@@ -211,10 +211,17 @@
+ * Power button is attached to the PIC microcontroller
+ ****************************************************************************/
+
++#define QNAP_TS409_GPIO_KEY_RESET 14
+ #define QNAP_TS409_GPIO_KEY_MEDIA 15
+
+ static struct gpio_keys_button qnap_ts409_buttons[] = {
+ {
++ .code = KEY_RESTART,
++ .gpio = QNAP_TS409_GPIO_KEY_RESET,
++ .desc = "Reset Button",
++ .active_low = 1,
++ },
++ {
+ .code = KEY_COPY,
+ .gpio = QNAP_TS409_GPIO_KEY_MEDIA,
+ .desc = "USB Copy Button",
+
Added: dists/trunk/linux-2.6/debian/patches/features/sparc/video-sunxvr500-intergraph.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/sparc/video-sunxvr500-intergraph.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,23 @@
+diff -aur a/drivers/video/sunxvr500.c b/drivers/video/sunxvr500.c
+--- a/drivers/video/sunxvr500.c 2008-11-20 18:02:37.000000000 -0500
++++ b/drivers/video/sunxvr500.c 2008-12-01 15:43:04.000000000 -0500
+@@ -396,6 +396,7 @@
+
+ static struct pci_device_id e3d_pci_table[] = {
+ { PCI_DEVICE(PCI_VENDOR_ID_3DLABS, 0x7a0), },
++ { PCI_DEVICE(PCI_VENDOR_ID_IGRAPH, 0x7a0), },
+ { PCI_DEVICE(PCI_VENDOR_ID_3DLABS, 0x7a2), },
+ { .vendor = PCI_VENDOR_ID_3DLABS,
+ .device = PCI_ANY_ID,
+diff -aur a/include/linux/pci_ids.h b/include/linux/pci_ids.h
+--- a/include/linux/pci_ids.h 2008-11-20 18:02:37.000000000 -0500
++++ b/include/linux/pci_ids.h 2008-12-01 15:43:04.000000000 -0500
+@@ -930,6 +930,8 @@
+ #define PCI_DEVICE_ID_SUN_TOMATILLO 0xa801
+ #define PCI_DEVICE_ID_SUN_CASSINI 0xabba
+
++#define PCI_VENDOR_ID_IGRAPH 0x1091
++
+ #define PCI_VENDOR_ID_CMD 0x1095
+ #define PCI_DEVICE_ID_CMD_643 0x0643
+ #define PCI_DEVICE_ID_CMD_646 0x0646
Added: dists/trunk/linux-2.6/debian/patches/features/x86/drivers-hp_ilo-fixup.patch
==============================================================================
--- (empty file)
+++ dists/trunk/linux-2.6/debian/patches/features/x86/drivers-hp_ilo-fixup.patch Sun Dec 28 16:55:48 2008
@@ -0,0 +1,11 @@
+--- linux-source-2.6.26/drivers/misc/hpilo.c~ 2008-08-06 13:22:32.000000000 -0600
++++ linux-source-2.6.26/drivers/misc/hpilo.c 2008-08-06 13:22:42.000000000 -0600
+@@ -688,7 +688,7 @@
+ for (minor = 0 ; minor < MAX_CCB; minor++) {
+ struct device *dev;
+ dev = device_create(ilo_class, &pdev->dev,
+- MKDEV(ilo_major, minor), NULL,
++ MKDEV(ilo_major, minor),
+ "hpilo!d%dccb%d", devnum, minor);
+ if (IS_ERR(dev))
+ dev_err(&pdev->dev, "Could not create files\n");
Modified: dists/trunk/linux-2.6/debian/patches/series/1~experimental.1
==============================================================================
--- dists/trunk/linux-2.6/debian/patches/series/1~experimental.1 (original)
+++ dists/trunk/linux-2.6/debian/patches/series/1~experimental.1 Sun Dec 28 16:55:48 2008
@@ -4,34 +4,36 @@
+ debian/scripts-kconfig-reportoldconfig.patch
+ debian/drivers-ata-ata_piix-postpone-pata.patch
++ debian/drivers-ata-pata_sis-postpone-pata.patch
+ debian/dfsg/drivers-net-tg3-fix-simple.patch
-+ debian/dfsg/drivers-net-bnx2-request_firmware-1.patch
+#+ debian/dfsg/drivers-net-bnx2-request_firmware-1.patch
#+ features/all/drivers-net-acenic-firmwar_request.patch
+ features/all/export-gfs2-locking-symbols.patch
+ features/all/export-unionfs-symbols.patch
-#+ bugfix/powerpc/mv643xx-hotplug-support.patch
-+ bugfix/powerpc/oldworld-boot-fix.patch
-#+ bugfix/powerpc/prep-utah-ide-interrupt.patch
-#+ bugfix/powerpc/serial.patch
+ bugfix/sparc/drivers_net-broken.patch
#+ bugfix/ia64/hardcode-arch-script-output.patch
+ bugfix/mips/disable-advansys.patch
+#+ bugfix/mips/gbefb_fix_cmap_fifo_timeout.patch
+ bugfix/arm/disable-scsi_acard.patch
+ bugfix/arm/disable-ath5k.patch
+ bugfix/arm/disable-r6040.patch
+ features/all/at76.patch
+ bugfix/fix-hifn_795X-divdi3.patch
+ bugfix/powerpc/mm-mol.patch
-#+ bugfix/powerpc/lpar-console.patch
-+ bugfix/all/stable/patch-2.6.27.1
-+ bugfix/all/stable/patch-2.6.27.2
-+ bugfix/all/stable/patch-2.6.27.3
-+ bugfix/all/stable/patch-2.6.27.4
-+ bugfix/all/stable/patch-2.6.27.5
-+ bugfix/all/stable/patch-2.6.27.6
-+ bugfix/all/stable/patch-2.6.27.7
-+ bugfix/all/stable/patch-2.6.27.8
-+ bugfix/all/stable/patch-2.6.27.9
-+ bugfix/all/stable/patch-2.6.27.10
++ bugfix/powerpc/lpar-console.patch
+#+ features/arm/ts409-export-reset.patch
+#+ bugfix/arm/kurobox_ignore_pci.patch
++ bugfix/all/sata_mv-clear_irq.patch
+#+ bugfix/arm/kurobox_powerdown.patch
+#+ bugfix/parisc/add-lockless-fw-funcs.patch
+#+ bugfix/parisc/hijack-jump-to-start_kernel.patch
+#+ bugfix/all/wireless-regulatory-default-EU.patch
+#+ features/arm/dns323_mac_addr_read.patch
+#+ features/arm/dns323_revb1.patch
+#+ bugfix/all/ext3-add-support-for-non-native-signed-unsigned-htr.patch
+#+ bugfix/all/ext4-add-support-for-non-native-signed-unsigned-htr.patch
++ features/arm/led-pca9532-mem-leak.patch
++ features/arm/led-pca9532-workqueque.patch
++ features/sparc/video-sunxvr500-intergraph.patch
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 Sun Dec 28 16:55:48 2008
@@ -1,114 +1,4 @@
-+ features/all/vserver/vs2.3.0.34.11.patch featureset=vserver featureset=xen-vserver
-+ features/all/vserver/bindmount-dev.patch featureset=vserver featureset=xen-vserver
-
-+ features/all/xen/vmlinuz-target.patch featureset=xen
-#+ features/all/xen/xenctrl.patch featureset=xen
-#+ features/all/xen/xenctrl-capabilities.patch featureset=xen
-#+ features/all/xen/xenctrl-privcmd.patch featureset=xen
-#+ features/all/xen/xenctrl-xenbus.patch featureset=xen
-#+ features/all/xen/xenctrl-sys-hypervisor.patch featureset=xen
-
-# m68k patches
-#
-# updated 2008/10/13 srm, based on
-# Source: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
-# BASE 3fa8749e584b55f1180411ab1b51117190bac1e5
-
-# NEXT_PATCHES_START
-
-# --- for 2.6.28 ---
-+ bugfix/m68k/2.6.27/m68k-use-bcd2bin-bin2bcd.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-use-percent-pS-infrastructure.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-put-bss-at-end-of-data-section.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-add-notes-to-init-data.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-reverse-platform-mmu-logic-so-sun3-is-last.diff arch=m68k
-+ bugfix/m68k/2.6.27/amiga-amiserial-no-serial-console-if-modular.diff arch=m68k
-+ bugfix/m68k/2.6.27/amiga-export-key_maps.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-kill-unused-atari_kbd_translate.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-define-rtc_lock.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-add-dma_sync_single_range_for_cpu-device.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-asm-pci-h-needs-asm-generic-pci-dma-compat.h.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k+parisc-hp_sdc_rtc-printf-format-warnings.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-remove-the-broken-Hades-support.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-remove-no-longer-used-pci-code.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/m68k-init_irq_proc-depends-on-CONFIG_PROC_FS.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scsi-needs-nvram.diff arch=m68k
-+ bugfix/m68k/2.6.27/net-rfkill-rfkill-input-c-needs-linux-sched-h.diff arch=m68k
-
-# NEXT_PATCHES_END
-
-# --- pending ---
-+ bugfix/m68k/2.6.27/m68k-replace-linux-68k-by-linux-m68k.diff arch=m68k
-
-# Atari SCC support
-+ bugfix/m68k/2.6.27/atari-scc-readd-atari_SCC_init_done.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc-readd-CONFIG_ATARI_SCC.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc-checkpatch.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc-gs-fix.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc-break_ctl.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-scc-no-serial-console-if-modular.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/130-adbraw.diff arch=m68k
-#133-arch.diff arch=m68k
-+ bugfix/m68k/2.6.27/revert-7557bc66be629d19a402e752673708bfbb8b5e86.diff arch=m68k
-+ bugfix/m68k/2.6.27/134-atari-fat.diff arch=m68k
-+ bugfix/m68k/2.6.27/143-ioext.diff arch=m68k
-+ bugfix/m68k/2.6.27/149-mc68681.diff arch=m68k
-+ bugfix/m68k/2.6.27/478-serial.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-rom-isa.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-io-fixes.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-ethernec.diff arch=m68k
-
-# Atari ARAnyM
-+ bugfix/m68k/2.6.27/atari-aranym.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-aranym-export-nf-api.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-aranym-block-and-console-access.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-aranym-nfcon-no-early_param-if-modular.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-aranym-nf-wrappers.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/atari-ethernat.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-ethernat-warning-fix.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/falconide-fix-ide_release_lock-imbalance.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/zorro-module-device-table.diff arch=m68k
-+ bugfix/m68k/2.6.27/blinux-list-is-subscribers-only.diff arch=m68k
-+ bugfix/m68k/2.6.27/amiga-debug=mem.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-scsi-HOST_C-cleanup.diff arch=m68k
-+ bugfix/m68k/2.6.27/amiga-platform-device.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-platform-device.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-platform-device.diff arch=m68k
-+ bugfix/m68k/2.6.27/amiga-platform-device2.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/atafb-cfb16.diff arch=m68k
-+ bugfix/m68k/2.6.27/atafb-line_length.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-noswap-linux-swap-h-needs-linux-pagemap-h.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-update-defconfig-latest.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-allow-all-kernel-traps-to-be-handled-via-exception-fixups-all.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/drivers-base-Kconfig-spelling-usees.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/m68k-move-includes-to-arch-m68k-include.diff arch=m68k
-
-+ bugfix/m68k/2.6.27/parisc-__kernel_time_t-is-always-long.diff arch=m68k
-
-# --- for 2.6.28 ---
-+ bugfix/m68k/2.6.27/m68k-arch-m68k-mm-kmap-c-introduce-missing-kfree.diff arch=m68k
-# ---
-
-+ bugfix/m68k/2.6.27/atari-ethernec-IRQF_SHARED.diff arch=m68k
-+ bugfix/m68k/2.6.27/atari-ethernec-fixes.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-reinstate-rtc.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-via-cleanup-and-commentry.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-machw_h_cleanup.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-baboon-interrupt-enable-disable.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-mac_esp-asm-fix.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-macfb-annotations-and-compiler-warning-fix.diff arch=m68k
-+ bugfix/m68k/2.6.27/m68k-fpsp040-fix-trigraph-ignored-warning.diff arch=m68k
-+ bugfix/m68k/2.6.27/fbcon-remove-broken-mac-vbl-handler.diff arch=m68k
-+ bugfix/m68k/2.6.27/fbcon-protect-free-irq-by-MACH_IS_ATARI-check.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-esp-fix-for-quadras-with-two-esp-chips.diff arch=m68k
-+ bugfix/m68k/2.6.27/mac-macfb-do-not-overflow-fix-id.diff arch=m68k
+#+ features/all/openvz/openvz.patch featureset=openvz
+#+ features/all/vserver/vs2.3.0.35.patch featureset=vserver
++ features/all/vserver/bindmount-dev.patch featureset=vserver
Modified: dists/trunk/linux-2.6/debian/rules.real
==============================================================================
--- dists/trunk/linux-2.6/debian/rules.real (original)
+++ dists/trunk/linux-2.6/debian/rules.real Sun Dec 28 16:55:48 2008
@@ -291,6 +291,13 @@
cp -a $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/include $(DIR)/arch/$(KERNEL_ARCH)/include || :
cp -a $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/kernel/asm-offsets.s $(DIR)/arch/$(KERNEL_ARCH)/kernel || :
+ifeq ($(ARCH),powerpc)
+ if [ -f $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/lib/crtsavres.o ]; then \
+ mkdir $(DIR)/arch/$(KERNEL_ARCH)/lib; \
+ cp -a $(SOURCE_DIR)/arch/$(KERNEL_ARCH)/lib/crtsavres.o $(DIR)/arch/$(KERNEL_ARCH)/lib; \
+ fi
+endif
+
cd $(SOURCE_DIR); \
find include -mindepth 1 -maxdepth 1 \
! -name config -a ! -name linux -a ! -name 'asm*' \
Modified: dists/trunk/linux-2.6/debian/templates/control.xen-linux-system.in
==============================================================================
--- dists/trunk/linux-2.6/debian/templates/control.xen-linux-system.in (original)
+++ dists/trunk/linux-2.6/debian/templates/control.xen-linux-system.in Sun Dec 28 16:55:48 2008
@@ -3,5 +3,5 @@
Priority: extra
Depends: linux-image- at upstreamversion@@abiname@@localversion@ (= ${binary:Version})
Description: XEN system with Linux @upstreamversion@ image on @class@
- This package depends on the binary Linux image and the correct hypervisor.
+ This package depends on the binary Linux image and hypervisors.
More information about the Kernel-svn-changes
mailing list