[Bug tree-optimization/111795] New: OMP SIMD inbranch call vectorization missing for AVX512

[Bug tree-optimization/111795] New: OMP SIMD inbranch call vectorization missing for AVX512

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111795

            Bug ID: 111795
           Summary: OMP SIMD inbranch call vectorization missing for
                    AVX512
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: rguenth at gcc dot gnu.org
  Target Milestone: ---

We don't currently handle the following

int x[1024];

#pragma omp declare simd simdlen(16) inbranch
__attribute__((noinline)) int
foo (int a, int b)
{
  return a + b;
}

void __attribute__((noipa))
bar (void)
{
#pragma omp simd
  for (int i = 0; i < 1024; i++)
    {
      if (x[i] < 10)
        x[i] = foo (x[i], x[i]);
    }
}

[Bug tree-optimization/111795] OMP SIMD inbranch call vectorization missing for AVX512

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111795

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
           Keywords|                            |missed-optimization
   Last reconfirmed|                            |2023-10-13
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1
             Target|                            |x86_64-*-*

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Mine.

[Bug tree-optimization/111795] OMP SIMD inbranch call vectorization missing for AVX512

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111795

--- Comment #2 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:3179ad72f67f31824c444ef30ef171ad7495d274

commit r14-4629-g3179ad72f67f31824c444ef30ef171ad7495d274
Author: Richard Biener <rguenther@suse.de>
Date:   Fri Oct 13 12:32:51 2023 +0200

    OMP SIMD inbranch call vectorization for AVX512 style masks

    The following teaches vectorizable_simd_clone_call to handle
    integer mode masks.  The tricky bit is to second-guess the
    number of lanes represented by a single mask argument - the following
    uses simdlen and the number of mask arguments to calculate that,
    assuming ABIs have them uniform.

    Similar to the VOIDmode handling there's a restriction on not
    supporting splitting/merging of incoming vector masks to
    more/less SIMD call arguments.

            PR tree-optimization/111795
            * tree-vect-stmts.cc (vectorizable_simd_clone_call): Handle
            integer mode mask arguments.

            * gcc.target/i386/vect-simd-clone-avx512-1.c: New testcase.
            * gcc.target/i386/vect-simd-clone-avx512-2.c: Likewise.
            * gcc.target/i386/vect-simd-clone-avx512-3.c: Likewise.

[Bug tree-optimization/111795] OMP SIMD inbranch call vectorization missing for AVX512

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111795

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed with similar restrictions as for non-AVX512 for now.