Fix c/s 17104
PATCH 1/1 Fix c/s 17104
Hi Keir,
Changeset 17104 does not look correct to me. svm_set_interrupt_shadow() seems to be a copy & paste bug, because the intr_shadow parameter is unused.
I think, what you intended to do is the attached patch.
That's all what I found so far just from reading the diff - but there may be more. Please double-check.
Signature · 12 lines
AMD Saxony, Dresden, Germany Operating System Research Center Legal Information: AMD Saxony Limited Liability Company & Co. KG Sitz (Geschäftsanschrift): Wilschdorfer Landstr. 101, 01109 Dresden, Deutschland Registergericht Dresden: HRA 4896 vertretungsberechtigter Komplementär: AMD Saxony LLC (Sitz Wilmington, Delaware, USA) Geschäftsführer der AMD Saxony LLC: Dr. Hans-R. Deppe, Thomas McCoy
xen/arch/x86/hvm/svm/svm.c +1 -1
@@ -375,7 +375,7 @@ static void svm_set_interrupt_shadow(strstatic void svm_set_interrupt_shadow (struct vcpu *v,unsigned int intr_shadow) {struct vmcb_struct *vmcb = v->arch.hvm_svm.vmcb; - vmcb->interrupt_shadow = !!(vmcb->interrupt_shadow & + vmcb->interrupt_shadow = !!(intr_shadow & (HVM_INTR_SHADOW_MOV_SS |HVM_INTR_SHADOW_STI )); }
_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xensource.com/xen-devel
Nice one. I didn't have a good coding afternoon on Friday. :-(
In general the mmio (and other) emulation changes have had very little testing on AMD, so please do test and let me know if anything else has broken.
-- Keir
Quoted 12 lines · Christoph Egger
Hi Keir,
Changeset 17104 does not look correct to me.
svm_set_interrupt_shadow()seems to be a copy & paste bug, because theintr_shadowparameter is unused.
I think, what you intended to do is the attached patch.
That's all what I found so far just from reading the diff - but there may be more. Please double-check.Signed-off-by Christoph Egger <[email protected]>
Quoted 5 lines · On Monday 25 February 2008 10:03:18 Keir Fraser wrote:
Nice one. I didn't have a good coding afternoon on Friday. :-(
In general the mmio (and other) emulation changes have had very little testing on AMD, so please do test and let me know if anything else has broken.
Yes, there are MMIO problems. We tracked them down to changeset 17090.
Christoph
Quoted 14 lines
-- Keir
Quoted 10 lines · Christoph Egger
Hi Keir,
Changeset 17104 does not look correct to me.
svm_set_interrupt_shadow()seems to be a copy & paste bug, because theintr_shadowparameter is unused.
I think, what you intended to do is the attached patch. That's all what I found so far just from reading the diff - but there may be more. Please double-check.Signed-off-by Christoph Egger <[email protected]>
Signature · 12 lines
AMD Saxony, Dresden, Germany Operating System Research Center Legal Information: AMD Saxony Limited Liability Company & Co. KG Sitz (Geschäftsanschrift): Wilschdorfer Landstr. 101, 01109 Dresden, Deutschland Registergericht Dresden: HRA 4896 vertretungsberechtigter Komplementär: AMD Saxony LLC (Sitz Wilmington, Delaware, USA) Geschäftsführer der AMD Saxony LLC: Dr. Hans-R. Deppe, Thomas McCoy
Yes, Intel's most recently mentioned bug (1180: low-res and low-colour WinXP display) is also due to the mmio changes. At least it goes away when I revert changesets 17105, 17104, 17102, 17100, 17098, 17090, in that order.
-- Keir
Quoted 28 lines · Christoph Egger
Quoted 5 lines · On Monday 25 February 2008 10:03:18 Keir Fraser wrote:
Nice one. I didn't have a good coding afternoon on Friday. :-(
In general the mmio (and other) emulation changes have had very little testing on AMD, so please do test and let me know if anything else has broken.
Yes, there are MMIO problems. We tracked them down to changeset 17090.
Christoph
Quoted 14 lines
-- Keir
Quoted 10 lines · Christoph Egger
Hi Keir,
Changeset 17104 does not look correct to me.
svm_set_interrupt_shadow()seems to be a copy & paste bug, because theintr_shadowparameter is unused.
I think, what you intended to do is the attached patch. That's all what I found so far just from reading the diff - but there may be more. Please double-check.Signed-off-by Christoph Egger <[email protected]>