On Tue, Jan 20, 2015 at 4:02 PM, Matt Turner wrote:
> On Tue, Jan 20, 2015 at 3:51 PM, Jason Ekstrand
> wrote:
> > Except typeless... We need some sort of assurance that the result of a
> NIR
> > comparison is always 0 or ~0.
>
> Help me understand how this is a different situation from what we
On Tue, Jan 20, 2015 at 3:51 PM, Jason Ekstrand wrote:
> Except typeless... We need some sort of assurance that the result of a NIR
> comparison is always 0 or ~0.
Help me understand how this is a different situation from what we have today?
Let's take for instance a vec2 == vec2 comparison. On
On Tue, Jan 20, 2015 at 3:31 PM, Matt Turner wrote:
> On Tue, Jan 20, 2015 at 3:17 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Tue, Jan 20, 2015 at 3:09 PM, Matt Turner wrote:
> >>
> >> On Tue, Jan 20, 2015 at 2:58 PM, Jason Ekstrand
> >> wrote:
> >> > On Mon, Jan 19, 2015 at 3:31 PM, Matt Turn
On Tue, Jan 20, 2015 at 3:17 PM, Jason Ekstrand wrote:
>
>
> On Tue, Jan 20, 2015 at 3:09 PM, Matt Turner wrote:
>>
>> On Tue, Jan 20, 2015 at 2:58 PM, Jason Ekstrand
>> wrote:
>> > On Mon, Jan 19, 2015 at 3:31 PM, Matt Turner wrote:
>> >>
>> >> For some reason, we occasionally write the flag r
On Tue, Jan 20, 2015 at 3:09 PM, Matt Turner wrote:
> On Tue, Jan 20, 2015 at 2:58 PM, Jason Ekstrand
> wrote:
> > On Mon, Jan 19, 2015 at 3:31 PM, Matt Turner wrote:
> >>
> >> For some reason, we occasionally write the flag register with a MOV.NZ
> >> instruction:
> >>
> >>add(8)
On Tue, Jan 20, 2015 at 2:58 PM, Jason Ekstrand wrote:
> On Mon, Jan 19, 2015 at 3:31 PM, Matt Turner wrote:
>>
>> For some reason, we occasionally write the flag register with a MOV.NZ
>> instruction:
>>
>>add(8) g25<1>F -g6<0,1,0>F g15<8,8,1>F
>>cmp.l.f0(8) g26<
On Mon, Jan 19, 2015 at 3:31 PM, Matt Turner wrote:
> For some reason, we occasionally write the flag register with a MOV.NZ
> instruction:
>
>add(8) g25<1>F -g6<0,1,0>F g15<8,8,1>F
>cmp.l.f0(8) g26<1>D g25<8,8,1>F 0F
>mov.nz.f0(8)null
For some reason, we occasionally write the flag register with a MOV.NZ
instruction:
add(8) g25<1>F -g6<0,1,0>F g15<8,8,1>F
cmp.l.f0(8) g26<1>D g25<8,8,1>F 0F
mov.nz.f0(8)nullg26<8,8,1>D
A MOV.NZ instruction on the result of a CMP is li