[PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[Gleb Natapov]

LMSW is present in both group tables. It was marked privileged only in
one of them. Intel analog of VMMCALL is already marked privileged.

Signed-off-by: Gleb Natapov <gleb@redhat.com>
diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index 3d2e115..3af63d2 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -363,9 +363,9 @@ static u32 group_table[] = {
 
 static u32 group2_table[] = {
 	[Group7*8] =
-	SrcNone | ModRM | Priv, 0, 0, SrcNone | ModRM,
+	SrcNone | ModRM | Priv, 0, 0, SrcNone | ModRM | Priv,
 	SrcNone | ModRM | DstMem | Mov, 0,
-	SrcMem16 | ModRM | Mov, 0,
+	SrcMem16 | ModRM | Mov | Priv, 0,
 	[Group9*8] =
 	0, 0, 0, 0, 0, 0, 0, 0,
 };
--
			Gleb.

Re: [PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[Avi Kivity]

On 03/03/2010 05:53 PM, Gleb Natapov wrote:
> LMSW is present in both group tables. It was marked privileged only in
> one of them. Intel analog of VMMCALL is already marked privileged.
>
>    

Both VMCALL and VMMCALL are unprivileged.

-- 
error compiling committee.c: too many arguments to function

Re: [PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[Gleb Natapov]

On Sun, Mar 07, 2010 at 10:45:47AM +0200, Avi Kivity wrote:
> On 03/03/2010 05:53 PM, Gleb Natapov wrote:
> >LMSW is present in both group tables. It was marked privileged only in
> >one of them. Intel analog of VMMCALL is already marked privileged.
> >
> 
> Both VMCALL and VMMCALL are unprivileged.
> 
But _we_ do not allow them in CPL!=0.

--
			Gleb.

Re: [PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[Avi Kivity]

On 03/07/2010 10:46 AM, Gleb Natapov wrote:
> On Sun, Mar 07, 2010 at 10:45:47AM +0200, Avi Kivity wrote:
>    
>> On 03/03/2010 05:53 PM, Gleb Natapov wrote:
>>      
>>> LMSW is present in both group tables. It was marked privileged only in
>>> one of them. Intel analog of VMMCALL is already marked privileged.
>>>
>>>        
>> Both VMCALL and VMMCALL are unprivileged.
>>
>>      
> But _we_ do not allow them in CPL!=0.
>    

Then that should be done outside the emulator.

-- 
error compiling committee.c: too many arguments to function

Re: [PATCH] KVM: x86 emulator mark VMMCALL and LMSW as privileged

[Gleb Natapov]

On Sun, Mar 07, 2010 at 10:52:42AM +0200, Avi Kivity wrote:
> On 03/07/2010 10:46 AM, Gleb Natapov wrote:
> >On Sun, Mar 07, 2010 at 10:45:47AM +0200, Avi Kivity wrote:
> >>On 03/03/2010 05:53 PM, Gleb Natapov wrote:
> >>>LMSW is present in both group tables. It was marked privileged only in
> >>>one of them. Intel analog of VMMCALL is already marked privileged.
> >>>
> >>Both VMCALL and VMMCALL are unprivileged.
> >>
> >But _we_ do not allow them in CPL!=0.
> 
> Then that should be done outside the emulator.
> 
As long as KVM is the only user of the emulator it doesn't matter.

--
			Gleb.