Commit Graph
8511 Commits
Author SHA1 Message Date
kaf24@firebug.cl.cam.ac.uk 1088bc0f6d Handle migration of x86 VCPUs between physical CPUs.
If required, context_switch() must pull state off of
old CPU.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-07 17:52:43 +01:00
kaf24@firebug.cl.cam.ac.uk 1198a6a99f Reduce locked critical region in __enter_scheduler(),
changing the context switch interface yet again.

domain_runnable() renamed to vcpu_runnable().

Fix stupid bug resulting in bogus value for
vcpu_dirty_cpumask, which caused vcpu_sync_execstate() to
fail sometimes.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-07 16:53:25 +01:00
vhanquez@kneesa.uk.xensource.com 8300a5a508 separate installation of headers from the linux tree into a new guest-header
directory under tools.

Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2006-01-07 01:31:04 +00:00
vhanquez@kneesa.uk.xensource.com 1d36a78348 test return value of kmalloc to prevent crash is the allocation fail.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2006-01-06 22:07:04 +00:00
vhanquez@kneesa.uk.xensource.com fd0de160a1 Get rid of the err variable by directly returning error values.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2006-01-06 22:03:58 +00:00
kaf24@firebug.cl.cam.ac.uk e92cab943b Change the context-switch interface. Get rid of
context_switch_finalise(). Instead provide a back-call
context_switch_done() for situations where arch-specific
context_switch() function does not return to the caller,
or needs to do some parts of state restoration with
interrupts enabled.

Get rid of ugly hack in arch/ia64.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-06 18:14:29 +01:00
kaf24@firebug.cl.cam.ac.uk 59505bb144 Rename per-domain cpumask to more descriptive domain_dirty_cpumask.
Add a new per-vcpu dirty cpumask (vcpu_dirty_cpumask), useful for
state flushing and selective tlb flushing.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-06 17:45:31 +01:00
kaf24@firebug.cl.cam.ac.uk ff3a72d599 Clean up domain shutdown and modification of vcpu
processor affinity.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-06 16:47:25 +01:00
cl349@firebug.cl.cam.ac.uk 04ddfac1d9 Always EXPORT_SYMBOL pm_power_off now that pm_power_off is always defined.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
2006-01-06 16:24:46 +01:00
cl349@firebug.cl.cam.ac.uk cd478e6f7d Move definition of pm_power_off to a more sensible location.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
2006-01-06 12:59:55 +01:00
kaf24@firebug.cl.cam.ac.uk 5becda0913 Fix a couple of bogus dom0_op names:
setdomaininfo -> setvcpucontext
  pincpudomain  -> setvcpuaffinity

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-06 12:53:19 +01:00
kaf24@firebug.cl.cam.ac.uk 7b1872e6f5 Clean up xen-internal representation of per-vcpu
physical cpu affinity. Rename idle_task variables and
macros to idle_domain.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-06 12:25:47 +01:00
kaf24@firebug.cl.cam.ac.uk 7d947ac220 Fix up x86 emulator header docs and ensure callers use
the X86EMUL_MODE enumeration.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-05 12:19:12 +01:00
kaf24@firebug.cl.cam.ac.uk dc88367fa1 The dom0_ops implementation wasn't matching the interface definition,
leading to ill messages when X starts up (and potentially to other
problems). The attached patch fixes that. However, it also points out
that there is a potential disagreement between the dom0_ops interface
and the internal MTRR interface - the former expects an unsigned
register number from the domain, the latter wants a signed one and does
a lookup when it's negative. I would think that the lookup code should
just be ripped out of xen/arch/x86/mtrr/main.c.

Signed-off-by: Jan Beulich <JBeulich@novell.com>
2006-01-04 18:51:38 +01:00
kaf24@firebug.cl.cam.ac.uk c546f7e7bb Fixes to mtrr interface code in linux guest.
Signed-off-by: Jan Beulich <JBeulich@novell.com>
2006-01-04 18:47:11 +01:00
kaf24@firebug.cl.cam.ac.uk 65abb26969 Add IO-APIC interrupt debugging to 'i' debug key.
Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-04 18:37:24 +01:00
kaf24@firebug.cl.cam.ac.uk ff6e4be615 Fix split of duties between close_netdev() and
netif_disconnect_backend() in netif driver.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-04 14:45:52 +01:00
kaf24@firebug.cl.cam.ac.uk b84e890ff3 Add pending status to irq debugkey dump.
Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-04 12:37:27 +01:00
kaf24@firebug.cl.cam.ac.uk 4a4c2bd2d2 Add dubug key for dumping guest-bound physical interrupts
and their current status.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-04 12:18:02 +01:00
kaf24@firebug.cl.cam.ac.uk f890ca8bcd This patch adds TPM_Seal support to the VTSP. On measured platforms
this can be use to better protect secrets in the VTPM System

Signed-off-by: Vinnie Scarlata <vincent.r.scarlata@intel.com>
2006-01-04 11:37:41 +01:00
kaf24@firebug.cl.cam.ac.uk e8166802fe merge 2006-01-03 19:06:14 +01:00
kaf24@firebug.cl.cam.ac.uk 7238a83fdf Marge with xen-ia64-unstable.hg 2006-01-03 19:03:56 +01:00
vhanquez@kneesa.uk.xensource.com 929f0f09a9 convert latest old initializer to C99 initializer
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2006-01-03 16:57:41 +00:00
djm@kirby.fc.hp.com 9f336e8adf Adapt xen-ia64 to changes in xen-unstable 2006-01-03 10:20:05 -06:00
djm@kirby.fc.hp.com e31c95409c Merge 2006-01-03 10:19:20 -06:00
djm@kirby.fc.hp.com fe8281c24d Small bug fixes 2006-01-03 08:59:00 -06:00
emellor@leeni.uk.xensource.com 3db606124b Remove 03_network_attach_detach_multiple_pos.test from xfail list
Signed-off-by: Murillo F. Bernardes <mfb@br.ibm.com>
2006-01-03 14:51:33 +00:00
emellor@leeni.uk.xensource.com c69c2efbcf Vcpu-list's error message shows an unnecessary header.
Signed-off-by: Yoshinori Katase <y_katase@soft.fujitsu.com>
2006-01-03 14:08:32 +00:00
emellor@leeni.uk.xensource.com 1989719e21 These xm info, top, dmesg and log requre no parameter.
But log's error message differs from other 3 command's error messages,
when parameter is given.

# xm log 1
Error: 'xm log' requires 0 argument.

   log                              Print the xend log

Other 3 commands ignore parameter.
I think it might be better symmetricalize.

Signed-off-by: Yoshinori Katase <y_katase@soft.fujitsu.com>
2006-01-03 14:07:43 +00:00
kaf24@firebug.cl.cam.ac.uk 6b38c4a0e1 Move public hvm interfaces into xen/include/public/hvm.
Add new header hvm_info_table.h for defining location and
contents of acpi-style hvm_info_table. Remove duplicate
definition in vmxassist/acpi_madt.c.

Signed-off-by: Keir Fraser <keir@xensource.com>
2006-01-03 14:58:34 +01:00
kaf24@firebug.cl.cam.ac.uk 8f5424c6bb merge 2006-01-03 14:36:56 +01:00
kaf24@firebug.cl.cam.ac.uk 29c3bf78b9 Rename VMX option name lapic to apic, and pass it to xen by
hvm_info_table.  Previous we pass it to xen by guest vcpu context ECX
register, now we unify to use hvm_info_table.

Signed-off-by: Xin Li <xin.b.li@intel.com>
2006-01-03 14:35:45 +01:00
vhanquez@kneesa.uk.xensource.com 28d7365b01 merge 2006-01-03 13:04:59 +00:00
vhanquez@kneesa.uk.xensource.com ee4eb30e9d mixup during reorganization. when SIF_INITDOMAIN it should kcons_write_dom0
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2006-01-03 13:03:35 +00:00
kaf24@firebug.cl.cam.ac.uk 61cf7e808a Fix VMX disk config string parse (allow leading whitespace).
Signed-off-by: Xin Li <xin.b.li@intel.com>
2006-01-02 11:35:29 +01:00
kaf24@firebug.cl.cam.ac.uk cfa0fac2c2 Provide an option to disable ACPI from the VMX config file.
By default ACPI is disabled.

Signed-off-by: Yu Ke <ke.yu@intel.com>
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
2006-01-01 11:29:17 +01:00
vhanquez@kneesa.uk.xensource.com d33de0f496 fix possible memory leak if one of the kmalloc fail.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 20:17:45 +00:00
vhanquez@kneesa.uk.xensource.com 27c7f637f8 remove one pointless macro and make code paths explicit.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 19:35:11 +00:00
vhanquez@kneesa.uk.xensource.com 05a04604e5 reorganize code to avoid duplication
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 15:58:44 +00:00
vhanquez@kneesa.uk.xensource.com 1b296eaa07 Convert old initializer to C99 initializer.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 14:20:01 +00:00
vhanquez@kneesa.uk.xensource.com 80a654a545 directly return -ENOMEM instead of jumping to a return -ENOMEM label
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 14:17:10 +00:00
vhanquez@kneesa.uk.xensource.com 4b6164a042 move error handling out of line to avoid duplicated code.
Signed-off-by: Vincent Hanquez <vincent@xensource.com>
2005-12-31 14:16:13 +00:00
kaf24@firebug.cl.cam.ac.uk 24b0fec3cf Clean up 'q' debug key output.
Signed-off-by: Keir Fraser <keir@xensource.com>
2005-12-31 14:33:00 +01:00
kaf24@firebug.cl.cam.ac.uk 7e127cd35f Allow non-privileged domains restricted access to
I/O memory and physical interrupts, under control
of domain0. Capabilities are maintained as rangesets
in Xen.

Signed-off-by: Ryan Wilson <hap9@epoch.ncsc.mil>
Signed-off-by: Keir Fraser <keir@xensource.com>
2005-12-31 14:15:22 +01:00
kaf24@firebug.cl.cam.ac.uk a8f5fb8508 Store admin-specified ioport capabilities in a rangeset
rather than a bitmap. arch_do_createdomain() can now
fail and the caller will clean up.

Signed-off-by: Keir Fraser <keir@xensource.com>
2005-12-31 13:11:47 +01:00
djm@kirby.fc.hp.com b6a15a2cb8 Enable stack trace on Xen BUG (by Isaku Yamahata) 2005-12-30 23:40:13 -06:00
djm@kirby.fc.hp.com 53879eb47b Plumb through nr_pages -- needed for correct balloon driver function 2005-12-30 16:11:08 -06:00
djm@kirby.fc.hp.com bc1bd4b0d8 Missing initialization in cache sync code (by Anthony Xu) 2005-12-30 12:55:19 -06:00
kaf24@firebug.cl.cam.ac.uk 9be588cacb Fix a missing break in ioemu command-line parsing switch
statement.

Signed-off-by: Xin Li <xin.b.li@intel.com>
2005-12-30 17:12:40 +01:00
kaf24@firebug.cl.cam.ac.uk 96905f367b Fix rdmsrl() for x86_32 -- high 32 bits must be cast to
u64 before shifting left.

Signed-off-by: Avi Kivity <avi@qumranet.com>
2005-12-30 17:11:05 +01:00