On Tue, Jun 05, 2012 at 08:57:17PM +0300, Gleb Natapov wrote:
> On Tue, Jun 05, 2012 at 07:41:17PM +0200, Alexander Graf wrote:
> > IIRC it was the size field, but I could be wrong. Please take an old
> > (phenom / barcelona is old enough) machine and create some traces to make
> > 100% sure that
On 06/05/2012 08:19 PM, Gleb Natapov wrote:
> On Tue, Jun 05, 2012 at 07:07:39PM +0300, Avi Kivity wrote:
>> > +
>> > + count = (PAGE_SIZE - offset_in_page(linear_addr))/size;
>> > +
>> > + if (count == 0) /* 'in' crosses page boundry */
>> > + return EMULATE_FAIL;
>> > +
>> > + count =
On Tue, Jun 05, 2012 at 08:50:16PM +0300, Gleb Natapov wrote:
> > No, what I'm saying is that the hardware omitted some fields even though
> > they were provided in the spec. IIRC it was the size field, but I could be
> > wrong. Please take an old (phenom / barcelona is old enough) machine and
>
On Tue, Jun 05, 2012 at 07:41:17PM +0200, Alexander Graf wrote:
> IIRC it was the size field, but I could be wrong. Please take an old (phenom
> / barcelona is old enough) machine and create some traces to make 100% sure
> that all fields you're trying to read out are actually provided.
>
Joerg
On Tue, Jun 05, 2012 at 07:41:17PM +0200, Alexander Graf wrote:
>
> On 05.06.2012, at 19:36, Gleb Natapov wrote:
>
> > On Tue, Jun 05, 2012 at 07:27:40PM +0200, Alexander Graf wrote:
> >>
> >> On 24.05.2012, at 12:54, Roedel, Joerg wrote:
> >>
> >>> On Thu, May 24, 2012 at 01:36:51PM +0300, Avi
On 05.06.2012, at 19:36, Gleb Natapov wrote:
> On Tue, Jun 05, 2012 at 07:27:40PM +0200, Alexander Graf wrote:
>>
>> On 24.05.2012, at 12:54, Roedel, Joerg wrote:
>>
>>> On Thu, May 24, 2012 at 01:36:51PM +0300, Avi Kivity wrote:
On 05/24/2012 01:34 PM, Roedel, Joerg wrote:
> On Wed, M
On Tue, Jun 05, 2012 at 07:27:40PM +0200, Alexander Graf wrote:
>
> On 24.05.2012, at 12:54, Roedel, Joerg wrote:
>
> > On Thu, May 24, 2012 at 01:36:51PM +0300, Avi Kivity wrote:
> >> On 05/24/2012 01:34 PM, Roedel, Joerg wrote:
> >>> On Wed, May 23, 2012 at 05:49:31PM +0300, Avi Kivity wrote:
>
On 24.05.2012, at 12:54, Roedel, Joerg wrote:
> On Thu, May 24, 2012 at 01:36:51PM +0300, Avi Kivity wrote:
>> On 05/24/2012 01:34 PM, Roedel, Joerg wrote:
>>> On Wed, May 23, 2012 at 05:49:31PM +0300, Avi Kivity wrote:
On 05/23/2012 05:40 PM, Avi Kivity wrote:
> On 05/23/2012 05:08 PM,
On Tue, Jun 05, 2012 at 07:07:39PM +0300, Avi Kivity wrote:
> > +
> > + count = (PAGE_SIZE - offset_in_page(linear_addr))/size;
> > +
> > + if (count == 0) /* 'in' crosses page boundry */
> > + return EMULATE_FAIL;
> > +
> > + count = min(count, rcx);
> > +
> > + r = __kvm_fast_st
On 05/23/2012 05:08 PM, Gleb Natapov wrote:
> "rep ins" emulation is going through emulator now. This is slow because
> emulator knows how to write back only one datum at a time. This patch
> provides fast path for the instruction in certain conditions. The
> conditions are: DF flag is not set, des
On Thu, May 24, 2012 at 01:36:51PM +0300, Avi Kivity wrote:
> On 05/24/2012 01:34 PM, Roedel, Joerg wrote:
> > On Wed, May 23, 2012 at 05:49:31PM +0300, Avi Kivity wrote:
> >> On 05/23/2012 05:40 PM, Avi Kivity wrote:
> >> > On 05/23/2012 05:08 PM, Gleb Natapov wrote:
> >> > If decode assists are n
On 05/24/2012 01:34 PM, Roedel, Joerg wrote:
> On Wed, May 23, 2012 at 05:49:31PM +0300, Avi Kivity wrote:
>> On 05/23/2012 05:40 PM, Avi Kivity wrote:
>> > On 05/23/2012 05:08 PM, Gleb Natapov wrote:
>> > If decode assists are not available, we still need to emulate, see 15.33.5.
>> >
>>
>> Joer
On Wed, May 23, 2012 at 05:49:31PM +0300, Avi Kivity wrote:
> On 05/23/2012 05:40 PM, Avi Kivity wrote:
> > On 05/23/2012 05:08 PM, Gleb Natapov wrote:
> > If decode assists are not available, we still need to emulate, see 15.33.5.
> >
>
> Joerg, the 2010 version of the manual says that the effec
On 05/23/2012 05:40 PM, Avi Kivity wrote:
> On 05/23/2012 05:08 PM, Gleb Natapov wrote:
>> "rep ins" emulation is going through emulator now. This is slow because
>> emulator knows how to write back only one datum at a time. This patch
>> provides fast path for the instruction in certain conditions
On 05/23/2012 05:08 PM, Gleb Natapov wrote:
> "rep ins" emulation is going through emulator now. This is slow because
> emulator knows how to write back only one datum at a time. This patch
> provides fast path for the instruction in certain conditions. The
> conditions are: DF flag is not set, des
"rep ins" emulation is going through emulator now. This is slow because
emulator knows how to write back only one datum at a time. This patch
provides fast path for the instruction in certain conditions. The
conditions are: DF flag is not set, destination memory is RAM and single
datum does not cro
16 matches
Mail list logo
