Displaying 6 results from an estimated 6 matches for "tied2".
Did you mean:
tied
2016 Mar 05
2
[VSXFMAMutate] OldFMAReg may be wrongly rewritten
...ction pass, all MutatingFMA instructions are still
handled and copies get inserted. That's fine. No change is required.
4) Scheduler optionally follows our suggestions.
5) After Scheduling, we do a copy cleanup pass that change:
%vreg0<def> = COPY %vreg2<kill>
%vreg0<def, tied2> = MutatingFMA %vreg1, %vreg0<tied0>, %vreg3
To:
%vreg2<def, tie2> = MutatingFMA %vreg1, %vreg2<tied0>, %vreg3
In the same block.
6) 1)~5) are target independent. Now the tied MutatingFMA instruction gets
lowered to one of its real forms.
The only thing left is to...
2016 Mar 16
2
[VSXFMAMutate] OldFMAReg may be wrongly rewritten
...>> still handled and copies get inserted. That's fine. No change is required.
>> 4) Scheduler optionally follows our suggestions.
>> 5) After Scheduling, we do a copy cleanup pass that change:
>> %vreg0<def> = COPY %vreg2<kill>
>> %vreg0<def, tied2> = MutatingFMA %vreg1, %vreg0<tied0>, %vreg3
>> To:
>> %vreg2<def, tie2> = MutatingFMA %vreg1, %vreg2<tied0>, %vreg3
>> In the same block.
>> 6) 1)~5) are target independent. Now the tied MutatingFMA instruction
>> gets lowered to one o...
2013 Jun 27
0
[LLVMdev] Proposal: extended MDString syntax
...:@stat>[TF=5]
%I3<def> = SETHIi <ga:@lstat>[TF=3]
%I3<def> = ADDri %I3<kill>, <ga:@lstat>[TF=4]
%I3<def> = SLLXri %I3<kill>, 12
%I3<def> = ADDri %I3<kill>, <ga:@lstat>[TF=5]
CMPri %I2<kill>, 0, %ICC<imp-def>
%I1<def,tied2> = MOVXCCrr %I3<kill>, %I1<kill,tied0>, 9, %ICC<imp-use,kill>
JMPLrr %I1<kill>, %G0, %O0<kill>, %O1<undef>, %O0<imp-def,dead>, %O1<imp-def,dead>, %ICC<imp-def,dead>, %O6<imp-use>, ...
%O0<def> = ORrr %G0, %I0<kill>
RET 8...
2016 Feb 29
2
[VSXFMAMutate] OldFMAReg may be wrongly rewritten
Ping?
On Mon, Feb 22, 2016 at 1:06 PM Tim Shen <timshen at google.com> wrote:
> On Fri, Feb 19, 2016 at 5:10 PM Tim Shen <timshen at google.com> wrote:
>
>> I wonder if we can fix this by making the transformation simpler, that
>> is, instead of doing:
>>
>
> I wrote a prototype (see attach) for this idea, it actually improves some
> of the test cases
2016 Mar 23
0
[VSXFMAMutate] OldFMAReg may be wrongly rewritten
...andled and copies get inserted. That's fine. No change is required.
>>> 4) Scheduler optionally follows our suggestions.
>>> 5) After Scheduling, we do a copy cleanup pass that change:
>>> %vreg0<def> = COPY %vreg2<kill>
>>> %vreg0<def, tied2> = MutatingFMA %vreg1, %vreg0<tied0>, %vreg3
>>> To:
>>> %vreg2<def, tie2> = MutatingFMA %vreg1, %vreg2<tied0>, %vreg3
>>> In the same block.
>>> 6) 1)~5) are target independent. Now the tied MutatingFMA instruction
>>> g...
2013 Jun 26
6
[LLVMdev] Proposal: extended MDString syntax
On Wed, Jun 26, 2013 at 3:59 PM, Nadav Rotem <nrotem at apple.com> wrote:
>
> On Jun 26, 2013, at 3:51 PM, Chandler Carruth <chandlerc at google.com> wrote:
>
> Can you suggest an alternative solution? Can you describe why you don't
> think metadata is the right container? This alone isn't really helpful at
> moving us toward something that there has been