Re: [PATCH v2 5/5] KVM: MMU: retry #PF for softmmu

[Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>]

On 11/15/2010 05:59 PM, Avi Kivity wrote:
> On 11/15/2010 11:55 AM, Xiao Guangrong wrote:
>> >>  >   Or another thread may have mmap()ed something else over the
>> >>  >   same address.
>> >>
>> >>  The mmap virtual address is also visible for other threads since the
>> >>  threads
>> >>  have the same page table, so i think this case is the same as above?
>> >
>> >  Again, don't we install the wrong spte in this case?
>> >
>>
>> I think it doesn't corrupts spte since we will walk guest page table
>> again
>> and map it to shadow pages when we retry #PF.
> 
> Well, you're right, we don't use any gfn/pfn info from the async page
> fault.
> 
> However, we're still not modelling the cpu accurately.  For example we
> will set dirty and accessed bits, or inject a page fault if the gpte
> turns out to be not present.
> 

Yes, i missed this, will cook it. Thanks.