Avoid repeated #define's in public headers.
Signed-off-by: Keir Fraser <keir@xensource.com>

Remove non-ISO attributes from public headers.
Signed-off-by: Keir Fraser <keir@xensource.com>

Register the portio handler only once.
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

Refactor guest exception injection code.

- Exceptions get reflected to the guest by default, instead of crashing
  the domain.
- Reduce code duplication and improve maintainability.

Signed-off-by: Asit Mallick <asit.k.mallick@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

Improve the IDE HD geometry auto detection algorithm.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Asit K Mallick <asit.k.mallick@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

Get rid of references to qemu-dm.debug.

Another issue found with building the guest outside the source tree.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

I found several build issues, mostly related to building outside of the
source tree. Below/attached adjustments for these.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

Other than the native kernels, the xen guests install without attempting
to use distribution provided mechanisms, thus e.g. preventing the
auto-generation of an initrd. Below/attached a suggestion to adjust that.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

Other than native x86-64 code, the XEN guest *does* use 4k mappings for
the contiguous kernel mapping of (physical) memory. Thus the code in
change_page_attr needs to work like on i386 (where large pages are used
only conditionally) rather than like native x86-64. Patch below/attached.
Not doing so triggered the BUG_ON during load of intel-agp on machines
with i915 chipset.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

Some exports were misssing/mis-qualified. Adjustment below/attached.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

Other than native i386 code, the page fault handler in the guest for
whatever reason ran a stack_trace even though another one is run from
die(). Since screen space is limited, duplicating (perhaps long) stack
traces seems quite wasteful. Patch removing this duplication below/attached.
Signed-off-by: Jan Beulich <JBeulich@novell.com>

This patch fixes few typos in hg-cheatsheet.txt
Signed-off-by: Nguyen Anh Quynh <aquynh@gmail.com>

Manual merge.

Rewrite xenbus_path and cleanup whitespace.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

Fix Xen 'make install' not to rebuild console.o if called
from sudo environemnt, for example.

Manual merge.

Remove incorrect check for terminating nul character.
The terminating nul character is not stored in the store but
added by read_reply.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

.

Remove BitKeeper files from the repository.
Signed-off-by: Keir Fraser <keir@xensource.com>

Manual merge

Removed "bk root" call from fbsdxensetup script

The script assumes the current directory to be the top of the xen tree.
It makes a rudimentary check to this effect.

Signed-off-by: James Bulpin <james@xensource.com>

mkelf32 explicit-size types are based on defs in inttypes.h.
Signed-off-by: Keir Fraser <keir@xensource.com>

manual merge

Fix to xen-clone, minor top-level Makefile tidy
Signed-off-by: ian.pratt@cl.cam.ac.uk

Device model path cleanup. Remove bochsrc which is no longer needed.
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

manual merge

Add the Mercurial for Xen cheatsheet into the repo.
Signed-off-by: ian@xensource.com

use hg changeset info for xen banner
Signed-off-by: Chris Wright <chrisw@osdl.org>

Formally remove support for linux 2.4 dom0. domU should still work.
Signed-off-by: ian@xensource.com

Update xen-clone script for hg. Still missing revision selection functionality.
Signed-off-by: ian@xensource.com

Add a .hgignore file to the repository. NB: hg takes regexps not wildcards
Signed-off-by: ian@xensource.com

bitkeeper revision 1.1782 (42c3a8e76Lioy1FVEJFTaTkbYVBy7Q)

[PATCH] Use copy_from_user when accessing guest_pt[]

Use copy_from_user when accessing guest_pt[]

Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1781 (42c3a8d63EmXV0wA3jHItrRQ0fKLsg)

[PATCH] Support VMX guests with 512M/1G memory.

Support VMX guests with 512M/1G memory.

Signed-off-by: Edwin Zhai <edwin.zhai@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>
diff -r f6e5bd774f51 -r 13e02f197018 tools/ioemu/configure

bitkeeper revision 1.1780 (42c3a8c4tqEOpGy663gWO35k5QwZTw)

[PATCH] Infrastructure for interrupt handling.

Infrastructure for interrupt handling.

- support interruptibility
- handle interrupt window exiting control appropriately
- Add ioapic device models imported from Bochs under LGPL
- generalize the interrupt architecture to support both PIC/APIC

This patch is necessary to fix several bugs on 32 bit VMX and prepares the
ground for adding a local APIC device model in the hypervisor.

Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1779 (42c3a8b3kGpPBNwXb3dn_b8FcmVPgQ)

[PATCH] ACPI support for guest firmware.

ACPI support for guest firmware.

Signed-off-by: Ke Yu <ke.yu@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1778 (42c3a89fKOC3ZLo9T-f0b9tyflqDzg)

[PATCH] Enable vmxassist for 64 bit.

Enable vmxassist for 64 bit.

Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1777 (42c3a88aPiHvcATujFTezqA1w1FMCg)

[PATCH] 64 bit xend cleanups.

64 bit xend cleanups.

Signed-off-by: Xin Li <xin.bi.li@intel.com>
Signed-off-by: Xiofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1776 (42c3a877bezcV9uHoSXuU-GWoyVOKw)

[PATCH] Enable qcow in the device models.

Enable qcow in the device models.

This saves disk space when dealing with large guest images.

Signed-off-by: Nitin Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>
diff -r 5d58883fc7b8 -r 499ac8e8df00 tools/ioemu/block.c

bitkeeper revision 1.1775 (42c3a866BfZr3l3NiVNwr6Z_1Uin_A)

[PATCH] Fix PCI BIOS double initialization.

Fix PCI BIOS double initialization.

Because the qemu device models are already doing the PCI BIOS initialization,
don't do it in rombios. With this change, the PCI network card becomes
functional.

Signed-off-by: Winston Wang <winston.l.wang@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>
diff -r e71f644f492e -r 5d58883fc7b8 tools/firmware/rombios/rombios.c

bitkeeper revision 1.1774 (42c3a855yAuKBfmJfftjhiVDml26pg)

[PATCH] Extend the VMX intercept mechanism to include mmio as well as portio.

Extend the VMX intercept mechanism to include mmio as well as portio.

Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1773 (42c3a841nLib9kdSmthr05jouqiNeg)

[PATCH] Device model cleanup.

Device model cleanup.

- Single config file for xen and device models
  (Shell script doesn't source xmdefconfig anymore)
- No more device-model shell script by default. You can
  have one if needed (qemu-dm.debug)
- All logic in the script device-model is moved to python
- $DISPLAY is passed from xm to xend
- Don't fork vncviewer on dryruns
- Add support for killing device models on domain destroy
- info vmxiopage command added to the monitor
- Refactor shared io page into global and per vcpu state
- Remove the hard coding of IOPACKET_PORT
- move the virtual_platform_def up to domain struct from vcpu
- xm create -n:

(vm
    (name ExampleVMXDomain)
    (memory 128)
    (ssidref -1)
    (image
        (vmx
            (kernel /usr/lib/xen/boot/vmxloader)
            (root '/dev/hda1 ro')
            (vcpus 1)
        )
    )
    (memmap )
    (device_model /tmp/foo)
    (hda /var/images/min-el3-i386.img)
    (hdb )
    (hdc )
    (hdd )
    (cdrom )
    (boot c)
    (fda )
    (fdb )
    (localtime 0)
    (serial )
    (macaddr )
    (stdvga 0)
    (isa 0)
    (nographic 0)
    (vnc 0)
    (sdl 0)
    (display localhost:10.0)
)

Signed-off-by: Edwin Zhai <edwin.zhai@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1772 (42c31843CEATPivypr5LYbRYU08TtQ)

added tag ia64-stable

bitkeeper revision 1.1772 (42c31571ulukoJpvEOsK4akggVt37Q)

null change

bitkeeper revision 1.1771 (42c314a41N8GOY476cnW9IejbgNDJw)

Merge bk://xen.bkbits.net/xeno-unstable.bk
into  bkbits.net:/repos/x/xen-ia64/xeno-unstable-ia64.bk

bitkeeper revision 1.1770 (42c2d66cG6YTU0WhG1b97E-InSNWWA)

Ensure global variables required by get_s_time() are initialised before
first use. Rejuggle bootstrap code slightly.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1769 (42c2cac1Ua7yeT9FnpvjJNvdSrd3BA)

Remove ticks_per_usec global variable. cpu_khz suffices.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1726.1.8 (42c1acafTSObuXeDW7_GDxYN5dosRA)

Grant table support changes and fix cmpxchg_user
Signed-off-by: Matt Chapman <matthewc@hp.com>

bitkeeper revision 1.1726.1.7 (42c194ee68Qy_Egi118UCoJdFnkGPQ)

Add bank switch for hyper_rfi hyperprivop
Signed-off-by: Dan Magenheimer <dan.magenheimer@hp.com>

bitkeeper revision 1.1768 (42c18d2259NPELcGV7ohyZNh72ufSw)

Based on the Keir's suggestion yesterday, I fixed the bug in xenlinux.
Now the LTP test cases pass well in domU; I ran LTP in domU along with
an infinite loop of "make clean; make -j4" on dom0 for sanity tests.

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>

bitkeeper revision 1.1767 (42c18c1dNRHfk7ZFajLHHtesQcrKlg)

At the time of debugging the fs segment problem, I noticed this bug (and
was able to test it).

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>

bitkeeper revision 1.1766 (42c17071SPDLo7qPlXAiQzrKjMu6Mw)

Add xenbus driver.
XendDomainInfo.py:
  Connect/Disconnect domains to/from store.
Signed-off-by: Rusty Russel <rusty@rustcorp.com.au> (authored)
Signed-off-by: Mike Wray <mike.wray@hp.com>
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

bitkeeper revision 1.1765 (42c16e51XwZ3-cKPHI29xhO2Hzk_sQ)

Increase size of xen_start_info_union since start_info structure
size is > 512 bytes.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

bitkeeper revision 1.1764 (42c12471iroiktRDv2QDqEDoIQVDNg)

comments

bitkeeper revision 1.1763 (42c11c79p9Z_K0YwZk_N5zJU0rjzDg)

Merge arcadians.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xeno.bk
into arcadians.cl.cam.ac.uk:/auto/anfs/nos1/ach61/bk

bitkeeper revision 1.1760.1.6 (42c10677TIdwirEem_EpnMFWS1MTtw)

Fix build on gcc <= 3.2.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1760.1.5 (42c105a0IGE_9KbvmOfox8TMahmQVw)

Only enable PGE in CR4 if the CPU supports the feature.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1760.1.4 (42c10578DlONZUYkjuzFvNR6idseEQ)

Simplify x86_32 boot code by removing bogus P6 check (really it was a
486 check, and we can assume everyone has CPUID I think).
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1760.1.3 (42c10175tVLAVy7ULar9xcSazDReXw)

[PATCH] [PATCH] VMXassist print e820 table

This patch integrates some of my debug environment back into the main
stream and it prints the e820 map as it is seen by a VMX domain.

        Leendert

Signed-Off-By: Leendert van Doorn <leendert@watson.ibm.com>

bitkeeper revision 1.1760.1.2 (42c10164eZ8JiY0-m8yKOwE_3OopGw)

[PATCH] [PATCH] Improved pending interrupt handling during VMX

This patch improves VMX exit handling while software interrupts are
pending: These should not stored as a VMX entry exception but should
instead just be restarted. This was the cause of the VMRESUME failure.

Leendert

Signed-Off-By: Leendert van Doorn <leendert@watson.ibm.com>

bitkeeper revision 1.1762 (42c0fafdBom2bVfipwX4FXIl_SgKmg)

Merge arcadians.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xeno.bk
into arcadians.cl.cam.ac.uk:/auto/anfs/nos1/ach61/bk

bitkeeper revision 1.1760.1.1 (42c0e8fa-Uxs1rA3BiXdTbynda9HmA)

convert some use of "if (condition) BUG()" to "BUG_ON(condition)" in netback driver.

Signed-off-by: Vincent Hanquez <vincent@xensource.com>

bitkeeper revision 1.1761 (42c0704c_K5N6pcA--cz5L-FX_z0zw)

Merge arcadians.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xeno.bk
into arcadians.cl.cam.ac.uk:/auto/anfs/nos1/ach61/bk

bitkeeper revision 1.1726.3.1 (42c07007i9Bkz3ggP_PXIksWaGD1Eg)

add framework for debugging processes

bitkeeper revision 1.1760 (42c05ebeLIfrneiw1jaZMwle-z9usw)

Check set_gdt() bounds before copy_from_user.
Signed-off-by: Chris Wright <chrisw@osdl.org>

bitkeeper revision 1.1759 (42c035d96-RzgfZ63O07ihY0xpew2g)

This patch adapts the Java Policy Processor to the default ssid change of
0xfffffff... to 0.

Signed-off by Ray Valdez <rvaldez@us.ibm.com>

bitkeeper revision 1.1758 (42bff8419M7X5aOGlHhSQueJtvoYaA)

Fix out of source tree build bugs:
        - add (srctree)/include/asm-xen to include list
        - fix include2/asm link to reference asm-XENARCH
        - fix install target for public headeres
This fix enables the complication of several configurations from the same (read-only) source tree.

Signed-off-by: Shahar Frank <shaharf@qumranet.com>

bitkeeper revision 1.1757 (42be9ad5TTWQjsCBju5noyzBrTMyjA)

Attached is the patch that changes the default ssid from 0xffffffff to 0
as discussed in previous emails.

Signed-off-by: Reiner Sailer <sailer@us.ibm.com>
Signed-off-by: Stefan Berger <stefanb@us.ibm.com>

bitkeeper revision 1.1756 (42bd28eb7ZP0AhHHKYuqhX2z3XItTQ)

Allow setting of ACM security policy at build time.
Signed-off-by: Nguyen Anh Quynh <aquynh@gmail.com>

bitkeeper revision 1.1755 (42bd1fcfzi0ZazGyTao_Ff-m4vjOqQ)

.del-plan9a.out.h~cdc6aa24ef3ba635:
  Delete: tools/libxc/plan9a.out.h

bitkeeper revision 1.1754 (42bc7207le_rYvV9kj8p-Dztt3T_eA)

Fix signedness issues, found by Jerone Young.

bitkeeper revision 1.1753 (42bbe67ePvjh9K0djYf6xRw2rGNreQ)

Properly remove old plan9 builder from xc header fiel and from xend.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1752 (42bbe5ba5hLPfcw1f77_F22fq6v6Yw)

The set of patches below add support for loading plan9's a.out
format using the linux builder.  This is considerably simpler
than having a seperate builder, shares more code and should
be easier to maintain.  Hopefully the original plan9 builder
can go away in the future.

I've been able to test this manually with vm-tools but am
still having some problems (unrelated I think) with xm.

Signed-off-by: Tim Newsham <newsham@lava.net>

bitkeeper revision 1.1751 (42bbe480z9Fp_L5Tc500W8c8CL3g9A)

Rationalise x86 CRn guest state into a ctrlreg array in the per-vcpu
context structure. Most noticeably this means the pt_base field has
gone away -- replaced by ctrlreg[3] (CR3). VCPU_guest_stts is also
gone -- it was never arch-independent anyway.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1750 (42bbd4dcGK7sWi0DELA8n8Ap3cfa0w)

Merge arcadians.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xen-unstable.bk
into arcadians.cl.cam.ac.uk:/local/scratch-2/vh249/unstable.bk

bitkeeper revision 1.1748.1.3 (42bbd472VxKIpvOzuAFDfVAi9gXwDA)

With this patch, now x86_64 domU is fully up!

We are running LTP there now.

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Xin B Li <xin.b.li@intel.com>

bitkeeper revision 1.1748.1.2 (42bbd22bGsDJd8zBp_EBCg97pPrPqA)

With the patch attached, we can bring up x86_64 domU to the point where
it tries to mount the root fs (see the log below). We are looking at
problems around blkif_connect at this point.

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
Signed-off-by: Xin B Li <xin.b.li@intel.com>

bitkeeper revision 1.1748.1.1 (42bbd1c5CE48yPys_hWh-39fCz3QGw)

Patch below is needed to make my system work stable in PAE mode.
Havn't seen problems without PAE, not sure whenever thats just
pure luck or whenever there is a bug in my PAE xenlinux kernel.
To me it looks like a generic bug though.

I've actually trapped into problems with unpin only:  A process
exits, somewhere in exit_mm() the page tables are unpinned,
shortly thereafter the mappings are cleared.  While doing so the
kernel oopses in zap_pte_range(), on page table write access.
Probably due to some stale tlb entry where the page is still
tagged read-only.

bitkeeper revision 1.1749 (42bbc434_aK0LSM1BC4FiLM_Xam_YA)

Merge arcadians.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xen-unstable.bk
into arcadians.cl.cam.ac.uk:/local/scratch-2/vh249/unstable.bk

bitkeeper revision 1.1748 (42bbc14djIjScNrLp20eDhlxUjQl4g)

xenstored_transaction.c, xenstored_core.h, xenstored_core.c, TODO:
  Update TODO list
  Wrap opendir in talloc so it gets cleaned up on OOM.
  Remove last call to system by open-coding "cp -al" to create
  transaction.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

bitkeeper revision 1.1747 (42bbc01fY31MfvOqdz4Mq_h_2D4Usg)

Merge firebug.cl.cam.ac.uk:/auto/groups/xeno-xenod/BK/xen-unstable.bk
into firebug.cl.cam.ac.uk:/local/scratch/cl349/xen-unstable.bk

bitkeeper revision 1.1742.1.1 (42bbc016sc__D7n9oSLMG75cKXdixQ)

XendDomainInfo.py: Fix imports.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>

bitkeeper revision 1.1744.1.1 (42bb2b36zVlhMosXY3bEaiG-q9siMg)

fix tools compilation on x86_64 adding 2 fields that will be unused

Signed-off-by: Vincent Hanquez <vincent@xensource.com>

bitkeeper revision 1.1726.1.6 (42bb0675eG3PXvB-XpEs0EYcdos-EA)

ia64: Add fast reflection for most clock ticks
Also, some additional debug output for various situations.
Signed-off-by: Dan Magenheimer <dan.magenheimer@hp.com>

bitkeeper revision 1.1746 (42badf41asjaKoYentyODb8CdFEydg)

Cset exclude: arun.sharma@intel.com[kaf24]|ChangeSet|20050623091344|24134

bitkeeper revision 1.1745 (42badeb1hE5PDZxZYF5DYBsU0Jya9w)

Fix ACM so that it can be built with NULL policy.
Signed-off-by: Nguyen Anh Quynh <aquynh@gmail.com>
Signed-off-by: Stefan Berger <stefanb@us.ibm.com>

bitkeeper revision 1.1744 (42baac12VHiYuszqWxXm5lsjjqZq4g)

This patch fixes some wrong comment in STE policy code.

Signed-off-by: Nguyen Anh Quynh <aquynh@gmail.com>

bitkeeper revision 1.1743 (42baabd1aq7F8kKY4ptBZ39ebtEqOw)

null_set_binary_policy() should return ACM_OK instead of -1. This
patch also makes few cleanups.

Signed-off-by: Nguyen Anh Quynh <aquynh@gmail.com>

bitkeeper revision 1.1742 (42ba8bdbUMsXF10VHPltV1P5J9z0Ow)

added tag ia64-stable

bitkeeper revision 1.1742 (42ba8b7fiNx5Cu0J91l9HCIo_TwvXQ)

Merge firebug.cl.cam.ac.uk:/local/scratch/kaf24/xen-unstable.bk
into firebug.cl.cam.ac.uk:/local/scratch/kaf24/xeno-unstable-ia64.bk

bitkeeper revision 1.1741 (42ba8952xTiiA50azru9j4tKvXIVDA)

[PATCH] vmx-gdbserver-paging.patch

gdbserver should not try to convert guest physical to machine physical if
paging is not enabled.

Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1740 (42ba88cewc_4MMbiCLInZ8gEcgqwhQ)

[PATCH] vmx-pit-reset2.patch

A VMX guest can reprogram the same PIT channel twice. We should not create
two ac_timers to handle this case.

Signed-off-by: Edwin Zhai <edwin.zhai@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1739 (42ba886dkLcSLF35FYJLZkKeANRJYg)

Trivial cleanup in netfront driver.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1738 (42ba87barFA27QtqNiI4v7D2u57koA)

This is a patch to the README file to provide the missing
info on building from the source tree. It's by no means
clean, complete or even necessarily correct - just a first
draft to get the doc/instructions going.

All cleanups would be appreciated.

Signed-off-by: Nivedita Singhvi <niv@us.ibm.com>
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1737 (42ba86724fSOFpp6L107qJDWXNKxug)

Attached is the patch with
        1. Cleanup of code & correct error handling for MMIO
instructions.
        2. Decoding of 64bit instructions for MMIO.

Signed-Off-By: Chengyuan Li <chengyuan.li@intel.com>
Signed-Off-By: Yunhong Jiang <hunhong.jiang@intel.com>
Signed-Off-By: Jun Nakajima <jun.nakajima@intel.com>
Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com>

bitkeeper revision 1.1736 (42ba85d8Vh1WXA4F1eQpFRpsTyq1xg)

[PATCH] sysenter-msr.patch

Handle MSR reads/writes to sysenter related MSRs.

Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Chengyuan Li <chengyuan.li@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

bitkeeper revision 1.1735 (42ba857eIL9j0maam3-Dv-gI3hOJUw)

Don't zero the vcpu flags in domain builder.
Signed-off-by: Scott Parish <srparish@us.ibm.com>

bitkeeper revision 1.1734 (42ba83ba7bRweVDldhJ4nYZQomQ8Lg)

Fix the x86/64 build. My change to make PAGE_SIZE a signed quantity
also made it int rather than long, which turns out to break stuff.
Debugged by Xin Li at Intel.
Signed-off-by: Keir Fraser <keir@xensource.com>

bitkeeper revision 1.1733 (42ba827d4k5gz1Dasgd0gorMu-SSmQ)

The 32-bit x86 LTP exposed bugs with LDT handling with x86_64 Xen and
XenLinux:
- fill the code for arbitrary_virt_to_machine(XenLinux)
- set 64-bit value for the base address for LDT (Xen)
- fix a bug (64-bit cleanup) in map_ldt_shadow_page (Xen)

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>

bitkeeper revision 1.1732 (42ba7fc5JvRaJiT14LMkB8SUdohfeA)

The attached patch now allows x86_64 xenlinux to run 32-bit x86 binaries
as the native x86_64 Linux does. I checked LTP using 32-bit binaries,
and got the same results on the native x86_64 Linux. At this point, 'int
0x80' is used for system calls, as it's unlikely that the fast systems
(sysenter or syscall) are significantly faster under the current
trampoline mechanism. However, it should be easy to replace
vsyscall-int80 with the fast system calls if one wants to do so (look at
USE_INT80 in the patch).

I found bugs with LDT handling in x86_64 Xen/XenLinux (exposed by 32-bit
LTP testcases fork05 and modify_ldt02), and the bugs have been fixed by
a separate patch in the next email:=20
[PATCH] [x86_64] Fixing LDT handling with x86_64 Xen

Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>