[VPID] Invalidate VPID mapping on INVLPG? [PATCH]

[VPID] Invalidate VPID mapping on INVLPG? [PATCH]

[Gianluca Guida]

[-- Attachment #1: Type: text/plain, Size: 219 bytes --]

Hello,

In processors supporting VPID I think we should invalidate the VPID 
mapping in the TLB after a shadow invlpg (in case EPT is turned off).

Signed-off-by: Gianluca Guida <gianluca.guida@eu.citrix.com>

Gianluca

[-- Attachment #2: vpid_invlpg.patch --]
[-- Type: text/x-patch, Size: 1043 bytes --]

diff -r 483d006cc607 xen/arch/x86/hvm/vmx/vmx.c
--- a/xen/arch/x86/hvm/vmx/vmx.c	Fri Apr 25 13:46:27 2008 +0100
+++ b/xen/arch/x86/hvm/vmx/vmx.c	Wed Apr 30 15:07:48 2008 +0100
@@ -1368,7 +1368,8 @@ static void vmx_invlpg_intercept(unsigne
 {
     struct vcpu *curr = current;
     HVMTRACE_2D(INVLPG, curr, /*invlpga=*/ 0, vaddr);
-    paging_invlpg(curr, vaddr);
+    if ( paging_invlpg(curr, vaddr) )
+        vpid_sync_vcpu_gva(curr, vaddr);
 }
 
 #define CASE_SET_REG(REG, reg)      \
diff -r 483d006cc607 xen/include/asm-x86/hvm/vmx/vmx.h
--- a/xen/include/asm-x86/hvm/vmx/vmx.h	Fri Apr 25 13:46:27 2008 +0100
+++ b/xen/include/asm-x86/hvm/vmx/vmx.h	Wed Apr 30 15:07:48 2008 +0100
@@ -301,6 +301,12 @@ static inline void ept_sync_all(void)
 
 void ept_sync_domain(struct domain *d);
 
+static inline void vpid_sync_vcpu_gva(struct vcpu *v, unsigned long gva)
+{
+    if ( cpu_has_vmx_vpid )
+        __invvpid(0, v->arch.hvm_vmx.vpid, (u64)gva);
+}
+
 static inline void vpid_sync_vcpu_all(struct vcpu *v)
 {
     if ( cpu_has_vmx_vpid )

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

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

RE: [VPID] Invalidate VPID mapping on INVLPG? [PATCH]

[Li, Xin B]

thanks, this is the right way to go.
-Xin

>-----Original Message-----
>From: xen-devel-bounces@lists.xensource.com 
>[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of 
>Gianluca Guida
>Sent: 2008年4月30日 22:26
>To: xen-devel@lists.xensource.com
>Subject: [Xen-devel] [VPID] Invalidate VPID mapping on INVLPG? [PATCH]
>
>Hello,
>
>In processors supporting VPID I think we should invalidate the VPID 
>mapping in the TLB after a shadow invlpg (in case EPT is turned off).
>
>Signed-off-by: Gianluca Guida <gianluca.guida@eu.citrix.com>
>
>Gianluca
>