Re: [PATCH 07/24] KVM: x86 emulator: fix 0f 01 /5 emulation

[Avi Kivity <avi@redhat.com>]

On 03/09/2010 04:33 PM, Gleb Natapov wrote:
> On Tue, Mar 09, 2010 at 04:27:39PM +0200, Avi Kivity wrote:
>    
>> On 03/09/2010 04:09 PM, Gleb Natapov wrote:
>>      
>>> It is undefined and should generate #UD.
>>>
>>> Signed-off-by: Gleb Natapov<gleb@redhat.com>
>>> ---
>>>   arch/x86/kvm/emulate.c |    3 +++
>>>   1 files changed, 3 insertions(+), 0 deletions(-)
>>>
>>> diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
>>> index 2df510b..1a32b78 100644
>>> --- a/arch/x86/kvm/emulate.c
>>> +++ b/arch/x86/kvm/emulate.c
>>> @@ -2486,6 +2486,9 @@ twobyte_insn:
>>>   				    (c->src.val&   0x0f), ctxt->vcpu);
>>>   			c->dst.type = OP_NONE;
>>>   			break;
>>> +		case 5: /* not defined */
>>> +			kvm_queue_exception(ctxt->vcpu, UD_VECTOR);
>>> +			goto done;
>>>   		case 7: /* invlpg*/
>>>   			emulate_invlpg(ctxt->vcpu, memop);
>>>   			/* Disable writeback. */
>>>        
>> Why is this needed?  We can only get here if the guest tricks us
>> (otherwise the #UD would go back to the guest, or rather, we'd trap
>> it to see if it's a hypercall instruction, but not pass it on to the
>> emulator).
>>
>>      
> For completes. A lot of code we added recently is there only because guest
> can trick us to enter emulator. Unfortunately we have to take suck tricks
> into account. Without this patch if emulator gets here it will report failed
> emulation.
>
>    

Okay.

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