[Bug middle-end/101909] New: 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[Bug middle-end/101909] New: 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

            Bug ID: 101909
           Summary: 73% regression on tfft benchmark for -O2
                    -ftree-loop-vectorize compared to -O2 on zen hardware
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hubicka at gcc dot gnu.org
  Target Milestone: ---

https://lnt.opensuse.org/db_default/v4/CPP/latest_runs_report?younger_in_days=14&older_in_days=0&all_changes=on&min_percentage_change=0.001&revisions=e54acea9e5a821448af97c66e94a1e4c4f977d5d%2Ce87209a1269622017bf3d98bf71502dcb0f893aa%2C73474527aaa24d9236aca074c5494a07f40ce058&include_user_branches=on

This regression seems to be limited to zenith tester and does not reproduce on
benzen

[Bug tree-optimization/101909] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|middle-end                  |tree-optimization
           Keywords|                            |missed-optimization
             Blocks|                            |53947
             Target|                            |x86_64-*-*
                 CC|                            |rguenth at gcc dot gnu.org

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
So that's znver1 (split AVX IIRC) compared to znver2?


Referenced Bugs:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53947
[Bug 53947] [meta-bug] vectorizer missed-optimizations

[Bug tree-optimization/101909] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[hubicka at ucw dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

--- Comment #2 from Jan Hubicka <hubicka at ucw dot cz> ---
> So that's znver1 (split AVX IIRC) compared to znver2?
Martin will know how to decode machine names.  I am never sure.
It is with generic, so split AVX does not make difference.

Honza

[Bug tree-optimization/101909] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |marxin at gcc dot gnu.org

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
(In reply to Jan Hubicka from comment #2)
> > So that's znver1 (split AVX IIRC) compared to znver2?
> Martin will know how to decode machine names.  I am never sure.
> It is with generic, so split AVX does not make difference.
> 
> Honza

zenith:
https://lnt.opensuse.org/db_default/v4/SPEC/machine/376

benzen:
https://lnt.opensuse.org/db_default/v4/SPEC/machine/393

[Bug tree-optimization/101909] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[rguenther at suse dot de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

--- Comment #4 from rguenther at suse dot de <rguenther at suse dot de> ---
On Mon, 16 Aug 2021, marxin at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909
> 
> Martin Liška <marxin at gcc dot gnu.org> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>                  CC|                            |marxin at gcc dot gnu.org
> 
> --- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
> (In reply to Jan Hubicka from comment #2)
> > > So that's znver1 (split AVX IIRC) compared to znver2?
> > Martin will know how to decode machine names.  I am never sure.
> > It is with generic, so split AVX does not make difference.
> > 
> > Honza
> 
> zenith:
> https://lnt.opensuse.org/db_default/v4/SPEC/machine/376
> 
> benzen:
> https://lnt.opensuse.org/db_default/v4/SPEC/machine/393

So indeed zenith is znver1 and benzen znver2.  But yeah, with
generic -O2 we don't use AVX.

[Bug tree-optimization/101909] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[crazylht at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

--- Comment #5 from Hongtao.liu <crazylht at gmail dot com> ---
Just note there is no regression for tfft benchmark comparing -O2
-ftree-loop-vectorize to -O2 on CLX.

[Bug tree-optimization/101909] [12 Regression] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.0
            Summary|73% regression on tfft      |[12 Regression] 73%
                   |benchmark for -O2           |regression on tfft
                   |-ftree-loop-vectorize       |benchmark for -O2
                   |compared to -O2 on zen      |-ftree-loop-vectorize
                   |hardware                    |compared to -O2 on zen
                   |                            |hardware

[Bug tree-optimization/101909] [12/13 Regression] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.0                        |12.2

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 12.1 is being released, retargeting bugs to GCC 12.2.

[Bug tree-optimization/101909] [12/13/14 Regression] 73% regression on tfft benchmark for -O2 -ftree-loop-vectorize compared to -O2 on zen hardware

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101909

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.3                        |12.4

--- Comment #8 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 12.3 is being released, retargeting bugs to GCC 12.4.