Skip to contentSkip to conversation

set accessed bit for VMCB segment selectors

PATCH RESEND In review Andre Przywara 2 participants 2 messages Feb 20, 2009 – Feb 23, 2009

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH RESEND 1/1 set accessed bit for VMCB segment selectors

Avi, could you apply this patch to kvm.git and linux-2.6.29-rcx? It fixes cross vendor migration with multithreaded guest applications. Also the two cross-vendor migration patches (in this very place) from Amit from October are not in upstream Linux, could you push (or pull-request) them?
Commit-IDs: 397248be81d79a3e03871b05d136e742a737b640 dc2a503e2f2fbab2dca788d49bb7057604db11b4

Thanks,
Andre.

-----------------
In the segment descriptor _cache_ the accessed bit is always set (although it can be cleared in the descriptor itself). Since Intel checks for this condition on a VMENTRY, set this bit in the AMD path to enable cross vendor migration.

---
 arch/x86/kvm/svm.c |   23 +++++++++++++++++++++--
 1 files changed, 21 insertions(+), 2 deletions(-)
arch/x86/kvm/svm.c +21 -2
index 22e88a4..dc5d88f 100644
@@ -796,20 +796,39 @@ static void svm_get_segment(struct kvm_vcpu *vcpu,
 	var->db = (s->attrib >> SVM_SELECTOR_DB_SHIFT) & 1;
 	var->g = (s->attrib >> SVM_SELECTOR_G_SHIFT) & 1;
 
+	switch (seg) {
 	/*
 	 * SVM always stores 0 for the 'G' bit in the CS selector in
 	 * the VMCB on a VMEXIT. This hurts cross-vendor migration:
 	 * Intel's VMENTRY has a check on the 'G' bit.
 	 */
-	if (seg == VCPU_SREG_CS)
+	case VCPU_SREG_CS:
 		var->g = s->limit > 0xfffff;
+		break;
 
 	/*
 	 * Work around a bug where the busy flag in the tr selector
 	 * isn't exposed
 	 */
-	if (seg == VCPU_SREG_TR)
+	case VCPU_SREG_TR:
 		var->type |= 0x2;
+		break;
+
+	/*
+	* The accessed bit must always be set in the segment
+	* descriptor cache, although it can be cleared in the
+	* descriptor, the cached bit always remains at 1. Since
+	* Intel has a check on this, set it here to support
+	* cross-vendor migration.
+	*/
+	case VCPU_SREG_DS:
+	case VCPU_SREG_ES:
+	case VCPU_SREG_FS:
+	case VCPU_SREG_GS:
+		if (!var->unusable)
+			var->type |= 0x1;
+		break;
+	}
 
 	var->unusable = !var->present;
 }
Signature · 1 line
1.5.2.2
#1 lore raw
Quoted 8 lines · Andre Przywara

Avi, could you apply this patch to kvm.git and linux-2.6.29-rcx? It fixes cross vendor migration with multithreaded guest applications. Also the two cross-vendor migration patches (in this very place) from Amit from October are not in upstream Linux, could you push (or pull-request) them?
Commit-IDs: 397248be81d79a3e03871b05d136e742a737b640 dc2a503e2f2fbab2dca788d49bb7057604db11b4

Sorry, I thought I already applied it for some reason. I now applied a different version of the patch (which you posted earlier) with better formatting. Thanks.

I won't push them into -rc since they don't fix regressions. They'll go into 2.6.30 and then into -stable.

Signature · 1 line
error compiling committee.c: too many arguments to function