xen.git
19 years ago[XENSTAT] Ignore domains that are being created/destroyed during get_info call.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 09:36:32 +0000 (10:36 +0100)]
[XENSTAT] Ignore domains that are being created/destroyed during get_info call.

Currently, xenstat_get_node exits with a failure and returns NULL if a
VM is deleted between the time the list of VMs is obtained and the time
it attempts to collect information about the VMs. This patch modifies xenstat
to ignore domains that transition during the call.

Signed-off-by: Simon Graham <Simon.Graham@stratus.com>
19 years ago[HVM][VMX] Cleanup vmx vmexit handler.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 09:29:58 +0000 (10:29 +0100)]
[HVM][VMX] Cleanup vmx vmexit handler.

Current code doesn't handle vmexits caused by triple fault, INIT, SIPI,
etc, which are listed in IA32 Architecture System Programming Guide 2,
Appendix I VMX Basic Exit Reasons, and calls __hvm_bug() when not
handling such cases, actually domain crash should be good enough.

Also various coding style cleanups.

Signed-off-by: Xin Li <xin.b.li@intel.com>
19 years ago[XEND] Daemonise vncviewer when it is spawned from xend.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 09:21:51 +0000 (10:21 +0100)]
[XEND] Daemonise vncviewer when it is spawned from xend.

There are various reasons it should be daemonized, but the particular
problem we hit was that YaST called "xm create" and waited on output on
stdout/stderr; xm then spawned vncviewer (which never closed its
inherited stdout and stderr); xm then would exit, but YaST still had
open file descriptors, and therefore waited forever.  It would be
possible to work around in YaST, but it seemed cleaner to daemonize
vncviewer.

Signed-off-by: Charles Coffing <ccoffing@novell.com>
19 years ago[PCI] back: New virtual pci backend: slot
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 09:17:42 +0000 (10:17 +0100)]
[PCI] back: New virtual pci backend: slot

This backend use a slot per pci device.  Contrary to vpci two functions from
one slot appear as two slots. This is useful on ia64 because it doesn't scan
functions > 0 if function 0 does not exist.

Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
19 years ago[XEND] Fix fallback and error path for domain destruction by domid,
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 09:03:30 +0000 (10:03 +0100)]
[XEND] Fix fallback and error path for domain destruction by domid,
or of non-existent domain.

I saw the following error message. (A domain of domain name XXX does
not exist.)

# xm list
Name                              ID Mem(MiB) VCPUs State  Time(s)
Domain-0                           0     1019     2 r-----    29.8
# xm destroy XXX
Error: an integer is required

This patch fixes it to print the following error message.

# xm list
Name                              ID Mem(MiB) VCPUs State  Time(s)
Domain-0                           0     1019     2 r-----    35.8
# xm destroy XXX
Error: the domain 'XXX' does not exist.

Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
19 years ago[HVM][VMX] Move hvm interface init_hypercall_page initialization into vmx_setup_hvm_f...
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:59:47 +0000 (09:59 +0100)]
[HVM][VMX] Move hvm interface init_hypercall_page initialization into vmx_setup_hvm_funcs.
Signed-off-by: Xin Li <xin.b.li@intel.com>
19 years ago[LIBXC] Do not gzclose() after close() on a gzdopen()'ed fd.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:57:24 +0000 (09:57 +0100)]
[LIBXC] Do not gzclose() after close() on a gzdopen()'ed fd.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
19 years ago[HVMLOADER] Only install hvmloader if it is built.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:51:08 +0000 (09:51 +0100)]
[HVMLOADER] Only install hvmloader if it is built.
This reverts a broken change in changeset 10948 that broke
the ia64 build.
Signed-off-by: Al Stone <ahs3@fc.hp.com>
19 years agoAdd RH-ish localhost to xend-relocation-hosts-allow in xend-config.sxp
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:33:38 +0000 (09:33 +0100)]
Add RH-ish localhost to xend-relocation-hosts-allow in xend-config.sxp
Signed-off-by: Jim Dykman <dykman@us.ibm.com>
19 years ago[PYGRUB] Flesh out some of pygrub's functionality as was originally
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:30:11 +0000 (09:30 +0100)]
[PYGRUB] Flesh out some of pygrub's functionality as was originally
intended. Changes include:
 * Addition of basic command line mode much like grub's so that you can
   boot things without having them specified in the config file
 * Edit/append mode for modifying kernel command lines, etc
 * Fix handling of case where the grub config didn't have a default
   specified

Signed-off-by: Jeremy Katz <katzj@redhat.com>
19 years ago[HVM] Fix hvm boot when acpi=1 is specified.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:25:46 +0000 (09:25 +0100)]
[HVM] Fix hvm boot when acpi=1 is specified.
Signed-off-by: Winston Wang <winston.l.wang@intel.com>
19 years ago[XEN] Remove redundant redeclaration of 'machine_restart'.
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Aug 2006 08:19:32 +0000 (09:19 +0100)]
[XEN] Remove redundant redeclaration of 'machine_restart'.
Fixes compile warning with gcc 3.4.2.
Signed-off-by: Hollis Blanchard <hollis@us.ibm.com>
19 years ago[qemu patches] Update patches for changesets 10957:08a11694b109 - 10959:06e2a95d3395.
chris@kneesaa.uk.xensource.com [Mon, 7 Aug 2006 17:28:50 +0000 (18:28 +0100)]
[qemu patches] Update patches for changesets 10957:08a11694b109 - 10959:06e2a95d3395.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Update ioemu to qemu 0.8.2.
chris@kneesaa.uk.xensource.com [Mon, 7 Aug 2006 17:27:27 +0000 (18:27 +0100)]
[qemu] Update ioemu to qemu 0.8.2.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Update ioemu to qemu 0.8.2. (.CVS files)
chris@kneesaa.uk.xensource.com [Mon, 7 Aug 2006 17:26:20 +0000 (18:26 +0100)]
[qemu] Update ioemu to qemu 0.8.2. (.CVS files)

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Update ioemu to qemu 0.8.2.
chris@kneesaa.uk.xensource.com [Mon, 7 Aug 2006 17:25:30 +0000 (18:25 +0100)]
[qemu] Update ioemu to qemu 0.8.2.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years agoLimit blktap static link to just libaio.
jchesterfield@wicket.uk.xensource.com [Mon, 7 Aug 2006 17:11:59 +0000 (18:11 +0100)]
Limit blktap static link to just libaio.

19 years ago[LINUX] Fix a couple of benign diffs against vanilla spinlock.h.
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Aug 2006 15:07:20 +0000 (16:07 +0100)]
[LINUX] Fix a couple of benign diffs against vanilla spinlock.h.
From: Jan Beulich
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years agoProvides hooks for perfc print/reset/gather.
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Aug 2006 14:55:39 +0000 (15:55 +0100)]
Provides hooks for perfc print/reset/gather.

These arch-neutral hooks allow architectures to report hardware counters.

Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
19 years agoChange DOM0_PERFCCONTROL: remove array limit.
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Aug 2006 14:53:06 +0000 (15:53 +0100)]
Change DOM0_PERFCCONTROL: remove array limit.

Descriptors and values are passed by two distinct buffers.

Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
19 years ago[XEN] Allow users to specify mce/nomce boot parameters.
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Aug 2006 14:42:25 +0000 (15:42 +0100)]
[XEN] Allow users to specify mce/nomce boot parameters.
Signed-off-by: Arun Sharma <arun@sharma-home.net>
19 years ago[XEN] Clean up shutdown handling and ignore opt_noreboot if dom0
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Aug 2006 14:35:06 +0000 (15:35 +0100)]
[XEN] Clean up shutdown handling and ignore opt_noreboot if dom0
shuts down cleanly. The option is intended only to retain information
on the local console in case of a crash.

Based on a patch from Muli Ben-Yehuda <muli@il.ibm.com>

Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[xend] Fix reboot for hvm domains with cdrom devices.
chris@kneesaa.uk.xensource.com [Mon, 7 Aug 2006 10:59:31 +0000 (11:59 +0100)]
[xend] Fix reboot for hvm domains with cdrom devices.
The device-type was not added to the device details which caused
cdrom devices to appear as harddisks after a reboot.

From: Alex Brett <Alex.Brett@xensource.com>
Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[XEN] Add Xen-attached event channels, which will be used
kfraser@localhost.localdomain [Fri, 4 Aug 2006 19:34:44 +0000 (20:34 +0100)]
[XEN] Add Xen-attached event channels, which will be used
by HVM for the ioreq_packet port.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[HVMLOADER] HVM loader initialises hypercall shim and uses
kfraser@localhost.localdomain [Fri, 4 Aug 2006 19:30:12 +0000 (20:30 +0100)]
[HVMLOADER] HVM loader initialises hypercall shim and uses
it to interrogate Xen version information. Also add support
for HVM hypercall execution on 64-bit host.

Signed-off-by: Steven Smith <ssmith@xensource.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[NET] front: Check for received packets in network_open0
ack@localhost.localdomain [Fri, 4 Aug 2006 15:07:58 +0000 (16:07 +0100)]
[NET] front: Check for received packets in network_open0

Because the backend brings up the interface long before the frontend
has booted up, it is possible that by the time we get here we already
have packets queued up for processing.

If we don't process them here, we may delay them more than what is
necessary.  Worse yet, it is possible to miss the notification
interrupt from the backend in such a way that we never get another
one until we bring the interface down and up.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
19 years ago[qemu patches] Update patches for changeset 10945:baa6d08e62a0.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 10:36:07 +0000 (11:36 +0100)]
[qemu patches] Update patches for changeset 10945:baa6d08e62a0.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Add missing semicolon.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 10:35:51 +0000 (11:35 +0100)]
[qemu] Add missing semicolon.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10943:1de1bb6a51c5.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 10:34:01 +0000 (11:34 +0100)]
[qemu patches] Update patches for changeset 10943:1de1bb6a51c5.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Update acpi timer to not use a qemu timer.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 10:33:41 +0000 (11:33 +0100)]
[qemu] Update acpi timer to not use a qemu timer.
Compute the acpi timer's value when it is accessed instead of using
a qemu timer to keep it uptodate.

From: Wang, Winston L <winston.l.wang@intel.com>
Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[hvm firmware] Update hvm acpi dsdt table.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 10:03:17 +0000 (11:03 +0100)]
[hvm firmware] Update hvm acpi dsdt table.
1)Code style clean up
2)Mininor fix for pci bus 0, dev 4 apic IRQ number
3)Add support for HCT "unreported IO and Memory" test by reporting low
640 MB mem as motherboard resource

Signed-off-by: Winston Wang <winston.l.wang@intel.com>
19 years ago[qemu patches] Refresh patches.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:57:34 +0000 (10:57 +0100)]
[qemu patches] Refresh patches.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10939:5980fb28b050.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:57:13 +0000 (10:57 +0100)]
[qemu patches] Update patches for changeset 10939:5980fb28b050.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Free page_array when it's no longer used.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:56:53 +0000 (10:56 +0100)]
[qemu] Free page_array when it's no longer used.
Seems no special reason for keeping page_array, free it
to avoid memroy leak in Qemu.

Signed-off-by : Zhang xiantao<xiantao.zhang@intel.com>

19 years ago[qemu patches] Refresh patches.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:53:11 +0000 (10:53 +0100)]
[qemu patches] Refresh patches.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10936:a0b3cf802d99.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:52:49 +0000 (10:52 +0100)]
[qemu patches] Update patches for changeset 10936:a0b3cf802d99.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Fix qemu memory access beyond 3G @VTi side.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:52:25 +0000 (10:52 +0100)]
[qemu] Fix qemu memory access beyond 3G @VTi side.

Signed-off-by: Zhang xiantao <xiantao.zhang@intel.com>
19 years ago[qemu patches] Refresh patches.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:43:02 +0000 (10:43 +0100)]
[qemu patches] Refresh patches.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10933:386990d004b8.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:42:02 +0000 (10:42 +0100)]
[qemu patches] Update patches for changeset 10933:386990d004b8.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Add switch to control whether acpi bridge is enabled.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:41:27 +0000 (10:41 +0100)]
[qemu] Add switch to control whether acpi bridge is enabled.

Signed-off-by: Winston Wang <winston.l.wang@intel.com>
19 years ago[qemu patches] Update patches for changeset 10931:3ea0a6847198.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:29:18 +0000 (10:29 +0100)]
[qemu patches] Update patches for changeset 10931:3ea0a6847198.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Make sure disk writes really made it to disk before we report I/O
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:28:51 +0000 (10:28 +0100)]
[qemu] Make sure disk writes really made it to disk before we report I/O
completion to the guest domain.  The DMA_MULTI_THREAD functionality
from the qemu-dm IDE emulation should make the performance overhead
of synchronous writes bearable, or at least comparable to native
hardware.

Signed-off-by: Rik van Riel <riel@redhat.com>
19 years ago[qemu patches] Update patches for changeset 10929:4dd640c6ce57.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:20:33 +0000 (10:20 +0100)]
[qemu patches] Update patches for changeset 10929:4dd640c6ce57.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Fix memset arguments in qemu audio code.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 09:19:53 +0000 (10:19 +0100)]
[qemu] Fix memset arguments in qemu audio code.
The arguments for the calls to memset are reversed in audio.c.

Signed-off-by: Charles Coffing <ccoffing@novell.com>
19 years ago[HVM] Wnable sound card support for qemu.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 08:31:06 +0000 (09:31 +0100)]
[HVM] Wnable sound card support for qemu.
Need a sound card to pass some benchmark check.

Signed-off-by: Edwin Zhai <edwin.zhai@intel.com>
19 years ago[qemu patches] Update patches for changeset 10926:e47aad0605a3.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 08:28:34 +0000 (09:28 +0100)]
[qemu patches] Update patches for changeset 10926:e47aad0605a3.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Always allow --disable-sdl since there's always vnc support.
chris@kneesaa.uk.xensource.com [Fri, 4 Aug 2006 08:27:52 +0000 (09:27 +0100)]
[qemu] Always allow --disable-sdl since there's always vnc support.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[blktap] Add dependencies to Makefiles and simplify to avoid needless rebuilds.
chris@kneesaa.uk.xensource.com [Thu, 3 Aug 2006 18:45:14 +0000 (19:45 +0100)]
[blktap] Add dependencies to Makefiles and simplify to avoid needless rebuilds.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10923:155385a02d.
chris@kneesaa.uk.xensource.com [Thu, 3 Aug 2006 18:19:27 +0000 (19:19 +0100)]
[qemu patches] Update patches for changeset 10923:155385a02d.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Write the port number where the vnc server is listening to xenstore.
chris@kneesaa.uk.xensource.com [Thu, 3 Aug 2006 18:14:41 +0000 (19:14 +0100)]
[qemu] Write the port number where the vnc server is listening to xenstore.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu patches] Update patches for changeset 10921:4c2fab8f8c34.
chris@kneesaa.uk.xensource.com [Thu, 3 Aug 2006 17:34:03 +0000 (18:34 +0100)]
[qemu patches] Update patches for changeset 10921:4c2fab8f8c34.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[qemu] Use xenstore to configure ioemu block devices.
chris@kneesaa.uk.xensource.com [Thu, 3 Aug 2006 17:28:29 +0000 (18:28 +0100)]
[qemu] Use xenstore to configure ioemu block devices.
- read ioemu block device config from xenstore
- don't require the ioemu: prefix on block devices any longer
- allow change of media associated with cdrom drives
- replace cdrom= option by :cdrom suffix on regular block device config:
  'file:/root/mytest.iso,hdc:cdrom,r'
- don't create default cdrom drive anymore - to create default empty
  cdrom drive use: ',hdc:cdrom,r'

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[HVM] Make copy_{to,from}_guest work for HVM domains.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 14:22:25 +0000 (15:22 +0100)]
[HVM] Make copy_{to,from}_guest work for HVM domains.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years agoChange default scheduler in user's manual (from sedf to credit).
ack@localhost.localdomain [Thu, 3 Aug 2006 14:05:54 +0000 (15:05 +0100)]
Change default scheduler in user's manual (from sedf to credit).
Signed-off-by: Atsushi SAKAI <sakaia@jp.fujitsu.com>
19 years ago[HVM] Tidy up e820 parsing in Xen.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 14:02:34 +0000 (15:02 +0100)]
[HVM] Tidy up e820 parsing in Xen.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years agoremove '=' between command line name/argument.
vhanquez@gwig.uk.xensource.com [Thu, 3 Aug 2006 13:44:05 +0000 (13:44 +0000)]
remove '=' between command line name/argument.
ease parsing without getopt.

Signed-off-by: Vincent Hanquez <vincent@xensource.com>
19 years ago[XEN] Allow add_to_physmap to be applied to DOMID_SELF.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 13:20:45 +0000 (14:20 +0100)]
[XEN] Allow add_to_physmap to be applied to DOMID_SELF.
Also sanitise handling of existing mappings of Xen heap
frames.
Signed-off-by: Steven Smith <ssmith@xensource.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[HVM] Remove unused ioreq state IORESP_HOOK.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 13:02:29 +0000 (14:02 +0100)]
[HVM] Remove unused ioreq state IORESP_HOOK.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years ago[HVM][VMX] Get rid of unused 'error' local variable.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 13:01:09 +0000 (14:01 +0100)]
[HVM][VMX] Get rid of unused 'error' local variable.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years ago[HVM] Declare vlapic_ipi() as static. Used only once.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 12:55:41 +0000 (13:55 +0100)]
[HVM] Declare vlapic_ipi() as static. Used only once.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years agoMissing file from previous changeset.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 12:54:05 +0000 (13:54 +0100)]
Missing file from previous changeset.

19 years ago[HVM] Add a concept of HVM parameters to the hypervisor.
kfraser@localhost.localdomain [Thu, 3 Aug 2006 12:53:33 +0000 (13:53 +0100)]
[HVM] Add a concept of HVM parameters to the hypervisor.

Each HVM domain has a space of HVM parameters associated with it,
and these can be manipulated via a new hvm_op hypercall. This means
that the hypervisor no longer needs to parse the hvm_info table, so
remove that code.

Signed-off-by: Steven Smith <ssmith@xensource.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years agoClarify the use of vcpu-set. Closes bug #686.
emellor@leeni.uk.xensource.com [Wed, 2 Aug 2006 16:42:38 +0000 (17:42 +0100)]
Clarify the use of vcpu-set.  Closes bug #686.

Signed-off-by Harry Butterworth <butterwo@uk.ibm.com>

19 years ago[HVM][SVM] Fix PIO breakage (fixed for VMX in previous changeset).
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:47:04 +0000 (15:47 +0100)]
[HVM][SVM] Fix PIO breakage (fixed for VMX in previous changeset).
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[HVM][VMX] Fix dead loop in PIO handling, introduced in c/s 10894.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:42:44 +0000 (15:42 +0100)]
[HVM][VMX] Fix dead loop in PIO handling, introduced in c/s 10894.
RIP update is missed for PIO handling.
Signed-off-by: Xin Li <xin.b.li@intel.com>
19 years ago[PCI] Transparent virtualization for pcifront and pciback.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:07:21 +0000 (15:07 +0100)]
[PCI] Transparent virtualization for pcifront and pciback.
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
19 years ago[XEN] Allow quick building of individual files.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:06:29 +0000 (15:06 +0100)]
[XEN] Allow quick building of individual files.

For the purpose of quick compile testing of (namely smaller) changes,
provide a
means to build individual pieces underneath the xen dir (could equally
be applied
to tools, and the top level Makefile could also be made aware of this
logic to
make things even more consistent). At once, add a way to only
pre-process files
in case some of the more complicated macros need change and hence
inspection
of the results.

Signed-off-by: Jan Beulich <jbeulich@novell.com>
19 years ago[LINUX] Allows userspace tools on a domU to set up a watch on the xenstore.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:01:59 +0000 (15:01 +0100)]
[LINUX] Allows userspace tools on a domU to set up a watch on the xenstore.
It does this by intercepting XS_WATCH requests written
to /proc/xen/xenbus and then re-submitting the request to the
in-kernel xenstore interface, in
linux-2.6-xen-sparse/drivers/xen/xenbus/xenbus_xs.c.  When a callback
occurs, an in-kernel function is invoked, which then reconstructs a
response in the format expected by userspace, and sends this response
through /proc/xen/xenbus.

It was necessary to add some supporting infrastructure to
linux-2.6-xen-sparse/drivers/xen/xenbus/xenbus_dev.c, such as an
additional mutex to protect the response queue and a list of active
watches associated with each connection.

Signed-off-by: Michael LeMay <mdlemay@epoch.ncsc.mil>
19 years ago[TOOLS] Enhance xenstore-ls to optionally display permissions.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 14:00:39 +0000 (15:00 +0100)]
[TOOLS] Enhance xenstore-ls to optionally display permissions.
Also perform an ioctl to determine the current width of the terminal.

Signed-off-by: Michael LeMay <mdlemay@epoch.ncsc.mil>
19 years ago[LINUX] Clarify a confusing error message in xenbus_probe.c.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 13:59:22 +0000 (14:59 +0100)]
[LINUX] Clarify a confusing error message in xenbus_probe.c.

Signed-off-by: Michael LeMay <mdlemay@epoch.ncsc.mil>
19 years ago[TOOLS] Modify xenstore_client.c to include a new utility, xenstore-chmod.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 13:58:37 +0000 (14:58 +0100)]
[TOOLS] Modify xenstore_client.c to include a new utility, xenstore-chmod.
This utility permits developers and administrators to
manually change the permissions on arbitrary locations in XenStore
from the command line.  This is often helpful if you're trying to debug an
application that relies on XenStore and is encountering difficulties
with permissions.

Signed-off-by: Michael LeMay <mdlemay@epoch.ncsc.mil>
19 years ago[TOOLS] lomount: Fix printf formats inside #ifdef DEBUG.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 09:13:30 +0000 (10:13 +0100)]
[TOOLS] lomount: Fix printf formats inside #ifdef DEBUG.

Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
19 years ago[IA64] Updates the ia64 buildconfigs.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 09:08:33 +0000 (10:08 +0100)]
[IA64] Updates the ia64 buildconfigs.
Enables pcifront for all kernels and pciback for -xen and -xen0.

Signed-off-by: Alex Williamson <alex.williamson@hp.com>
19 years ago[HVM] Place all APIC registers into one page in native format.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 09:07:03 +0000 (10:07 +0100)]
[HVM] Place all APIC registers into one page in native format.
With this change we can re-use code at include/asm-x86/apicdef.h,
making the code much cleaner. Also it help for future enhancement.

This patch does not change any logic except the change to
CONTROL_REG_ACCESS_NUM, which should be 0xf for CR8 access.

Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com
19 years ago[HVM] Fix an issue with APIC priority checks.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 09:04:27 +0000 (10:04 +0100)]
[HVM] Fix an issue with APIC priority checks.

Signed-off-by; Yunhong Jiang <yunhong.jiang@intel.com>

19 years ago[VMXASSIST] Fix GDTR accesses when paging mode enabled.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 09:03:04 +0000 (10:03 +0100)]
[VMXASSIST] Fix GDTR accesses when paging mode enabled.

The gdtr information in oldctx is an address for guest, not for
vmxassist. When access descriptor on guest gdt, we need to go through
guest page table if guest enable paging. This error may happen if
guest enable PE/PG in one instruction.

Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>=20
Signed-off-by: Xin Li <xin.b.li@intel.com>
19 years ago[SHADOW] Build p2m mapping according to m2p mapping.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:58:06 +0000 (09:58 +0100)]
[SHADOW] Build p2m mapping according to m2p mapping.

Signed-off-by: Xin Li <xin.b.li@intel.com>
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
19 years agoUpdate patch to reflect changes from 10867:edf7a25fdc43.
chris@kneesaa.uk.xensource.com [Wed, 2 Aug 2006 08:54:32 +0000 (09:54 +0100)]
Update patch to reflect changes from 10867:edf7a25fdc43.

Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
19 years ago[HVM] Initialise full regs structure for PIO requests.
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:52:03 +0000 (09:52 +0100)]
[HVM] Initialise full regs structure for PIO requests.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[HVM][VMX] Clean up vmx hvm interface functions:
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:38:08 +0000 (09:38 +0100)]
[HVM][VMX] Clean up vmx hvm interface functions:
 1) define vmx hvm interface functions static.
 2) setup hvm interface functions only once.

Signed-off-by: Xin Li <xin.b.li@intel.com>
19 years agoThere is an allocation-failure path in sched_credit.c that BUGs. The
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:20:09 +0000 (09:20 +0100)]
There is an allocation-failure path in sched_credit.c that BUGs.  The
attached patch handles this potential failure case more gracefully.

Signed-off-by: Charles Coffing <ccoffing@novell.com>
19 years ago[TOOLS] ppc64 is not a LIB_64_ARCH
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:16:26 +0000 (09:16 +0100)]
[TOOLS] ppc64 is not a LIB_64_ARCH

Signed-off-by: Jimi Xenidis <jimix@watson.ibm.com>
19 years ago[PCI] back: Fix potential infinite loop in pcistub_match_one().
kfraser@localhost.localdomain [Wed, 2 Aug 2006 08:15:26 +0000 (09:15 +0100)]
[PCI] back: Fix potential infinite loop in pcistub_match_one().

The for loop in pcistub_match_one will loop forever if the
dev->bus->self links to itself at the uppermost bridge.  Adding a
check to prevent linking back in on itself prevents this.

Signed-off-by: Jon Mason <jdmason@us.ibm.com>
19 years agoMerge.
kfraser@localhost.localdomain [Tue, 1 Aug 2006 17:08:01 +0000 (18:08 +0100)]
Merge.

19 years ago[LINUX] Fix save/restore (broken by my own GSO tweaks).
kfraser@localhost.localdomain [Tue, 1 Aug 2006 17:06:44 +0000 (18:06 +0100)]
[LINUX] Fix save/restore (broken by my own GSO tweaks).
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years agoSwitch blktapctrl.c over to using standard system daemon(3) call. Current patch was...
jchesterfield@dhcp92.uk.xensource.com [Tue, 1 Aug 2006 16:59:27 +0000 (17:59 +0100)]
Switch blktapctrl.c over to using standard system daemon(3) call. Current patch was not closing all FDs and redirecting output to /dev/null, preventing tapdisk from launching correctly.

19 years ago[TOOLS] Clean up libxenstore.a on 'make clean'.
kfraser@localhost.localdomain [Tue, 1 Aug 2006 16:47:21 +0000 (17:47 +0100)]
[TOOLS] Clean up libxenstore.a on 'make clean'.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[XEN] Fix 32-bit build.
kfraser@localhost.localdomain [Tue, 1 Aug 2006 16:47:01 +0000 (17:47 +0100)]
[XEN] Fix 32-bit build.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[HVM][SVM] Change the calling convention for SVM VMMCALLs so
kfraser@localhost.localdomain [Tue, 1 Aug 2006 16:28:19 +0000 (17:28 +0100)]
[HVM][SVM] Change the calling convention for SVM VMMCALLs so
that they don't conflict with the hypercall calling convention.
Signed-off-by: Steven Smith <ssmith@xensource.com>
19 years ago[XEN] Add hypercall support for HVM guests. This is
kfraser@localhost.localdomain [Tue, 1 Aug 2006 16:18:05 +0000 (17:18 +0100)]
[XEN] Add hypercall support for HVM guests.  This is
fairly useless at the moment, since all of the hypercalls
fail, since copy_from_user doesn't work correctly in HVM
domains.

Signed-off-by: Steven Smith <ssmith@xensource.com>
Add a CPUID hypervisor platform interface at leaf
0x40000000. Allow hypercall transfer page to be filled
in via MSR 0x40000000.

Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[LINUX] Do not early-unpin pagetables that contain foreign mappings.
kfraser@localhost.localdomain [Tue, 1 Aug 2006 14:48:48 +0000 (15:48 +0100)]
[LINUX] Do not early-unpin pagetables that contain foreign mappings.
This fixes a bug whereby foreign pages were freed by the unpin, which
then become owned by the local domain before it destroys its ptes.
It therefore (erroneously) detects the mappings as local and so
updates reference counts, leading to crashes.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[XEN] Hide compile-time information values behind a
kfraser@localhost.localdomain [Tue, 1 Aug 2006 14:28:09 +0000 (15:28 +0100)]
[XEN] Hide compile-time information values behind a
functional interface, so that the values are contained
within a single object file.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years agoRename all the CVS directories to .CVS. We want to keep the references to
emellor@leeni.uk.xensource.com [Tue, 1 Aug 2006 13:34:18 +0000 (14:34 +0100)]
Rename all the CVS directories to .CVS.  We want to keep the references to
upstream QEMU, but they can be confusing when they are called CVS.

Signed-off-by: Ewan Mellor <ewan@xensource.com>
19 years ago[NET] Enable TCPv4 segmentation offload in front/back drivers.
kfraser@localhost.localdomain [Tue, 1 Aug 2006 10:54:45 +0000 (11:54 +0100)]
[NET] Enable TCPv4 segmentation offload in front/back drivers.
Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[ACM] A wxPython GUI tool for sHype/ACM in Xen. It is used
kfraser@localhost.localdomain [Mon, 31 Jul 2006 17:12:53 +0000 (18:12 +0100)]
[ACM] A wxPython GUI tool for sHype/ACM in Xen. It is used
and described in the user guide to demonstrate how to quickly create,
enforce, and test workload protection policies in Xen. The tool is aimed
to improve usability of sHype/ACM for Xen.It is included in a single
file to make it easy to try it out on different platforms (it inherits
the look and feel of the base platform). It is tested on Linux and
Windows. Please read the new user guide section about how to start and
use the tool (section 10.2).

Requirements: python (www.python.org), wxPython (www.wxpython.org or
try 'yum install wxPython').

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

19 years ago[DOCS][ACM] Update the user guide for Xen with a chapter describing the
kfraser@localhost.localdomain [Mon, 31 Jul 2006 17:12:04 +0000 (18:12 +0100)]
[DOCS][ACM] Update the user guide for Xen with a chapter describing the
sHype/ACM security module for Xen. Section 10.2 offers a quick guide
showing how to setup and test the workload protection with sHype/ACM
in Xen.

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

19 years ago[BLKTAP] Properly daemonise the blktap control daemon. Fixes bug #709.
kfraser@localhost.localdomain [Mon, 31 Jul 2006 17:07:25 +0000 (18:07 +0100)]
[BLKTAP] Properly daemonise the blktap control daemon. Fixes bug #709.
Signed-off-by: Harry Butterworth <butterwo@uk.ibm.com>
19 years ago[NET] back: Transmit TSO packets if supported
kfraser@localhost.localdomain [Mon, 31 Jul 2006 16:49:36 +0000 (17:49 +0100)]
[NET] back: Transmit TSO packets if supported

This patch adds TSO transmission support to the backend.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Disable for now, as domU->dom0 direction.

Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[NET] front: Add TSO support
kfraser@localhost.localdomain [Mon, 31 Jul 2006 16:45:22 +0000 (17:45 +0100)]
[NET] front: Add TSO support

This patch adds TCP Segmentation Offload (TSO) support to the
frontend.
It also advertises this fact through xenbus so that the frontend can
detect this and send through TSO requests only if it is supported.

This is done using an extra request slot which is indicated by a flag
in the first slot.  In future checksum offload can be done in the same
way.

Even though only TSO is supported for now the code actually supports
GSO so it can be applied to any other protocol.  The only missing bit
is the detection of host support for a specific GSO protocol.  Once
that is added we can advertise all supported protocols to the guest.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Disable for now, as in domU->dom0 direction.

Signed-off-by: Keir Fraser <keir@xensource.com>
19 years ago[NET] back: Transmit SG packets if supported
kfraser@localhost.localdomain [Mon, 31 Jul 2006 16:42:13 +0000 (17:42 +0100)]
[NET] back: Transmit SG packets if supported

This patch adds scatter-and-gather transmission support to the
backend. This allows the MTU to be raised right now and the potential
for TSO in future.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>