[Xen-changelog] [xen-unstable] xen: allow HVM save/restore from different cpus

[Xen patchbot-unstable <patchbot-unstable@xxxxxxxxxxxxxxxxxxx>]

# HG changeset patch
# User Keir Fraser <keir.fraser@xxxxxxxxxx>
# Date 1278674555 -3600
# Node ID 8b2453940dac4d08ee2e9e04965ab286af4f1afe
# Parent  8afa0bb43c45b2ca08c62ca8ff5efd79ad224da2
xen: allow HVM save/restore from different cpus

The hardware CPUID-levelling features level the feature flags but
don't change the CPU family/model/stepping.  Relax the HVM restore
check on family/model/stepping to printk but not veto the load, so
that VMs can be migrated between machines that have been
CPUID-levelled.

Signed-off-by: Tim Deegan <Tim.Deegan@xxxxxxxxxx>
---
 xen/arch/x86/hvm/save.c |    9 +++++----
 1 files changed, 5 insertions(+), 4 deletions(-)

diff -r 8afa0bb43c45 -r 8b2453940dac xen/arch/x86/hvm/save.c
--- a/xen/arch/x86/hvm/save.c   Fri Jul 09 12:22:00 2010 +0100
+++ b/xen/arch/x86/hvm/save.c   Fri Jul 09 12:22:35 2010 +0100
@@ -55,10 +55,11 @@ int arch_hvm_load(struct domain *d, stru
     }
 
     cpuid(1, &eax, &ebx, &ecx, &edx);
-    /* TODO: need to define how big a difference is acceptable? */
-    if ( hdr->cpuid != eax )
-        gdprintk(XENLOG_WARNING, "HVM restore: saved CPUID (%#"PRIx32") "
-               "does not match host (%#"PRIx32").\n", hdr->cpuid, eax);
+    /* CPUs ought to match but with feature-masking they might not */
+    if ( (hdr->cpuid & ~0x0fUL) != (eax & ~0x0fUL) )
+        gdprintk(XENLOG_INFO, "HVM restore (%u): VM saved on one CPU "
+                 "(%#"PRIx32") and restored on another (%#"PRIx32").\n", 
+                 d->domain_id, hdr->cpuid, eax);
 
     /* Restore guest's preferred TSC frequency. */
     if ( hdr->gtsc_khz )

_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog