[Bug target/95082] New: LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[Bug target/95082] New: LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[wschmidt at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

            Bug ID: 95082
           Summary: LE implementations of vec_cnttz_lsbb and
                    vec_cntlz_lsbb are wrong
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: wschmidt at gcc dot gnu.org
  Target Milestone: ---

For little endian, we need to swap vctzlsbb and vclzlsbb, but today we generate
the BE instruction in all cases.

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[wschmidt at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

Bill Schmidt <wschmidt at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-code
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2020-05-12
     Ever confirmed|0                           |1
   Target Milestone|---                         |11.0
                 CC|                            |segher at gcc dot gnu.org

--- Comment #1 from Bill Schmidt <wschmidt at gcc dot gnu.org> ---
Confirmed.

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[pc at us dot ibm.com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

Paul Clarke <pc at us dot ibm.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pc at us dot ibm.com

--- Comment #2 from Paul Clarke <pc at us dot ibm.com> ---
This is a dup of bug 95070.
(I am unable to mark it as such.)

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[segher at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

--- Comment #3 from Segher Boessenkool <segher at gcc dot gnu.org> ---
*** Bug 95070 has been marked as a duplicate of this bug. ***

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[segher at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

--- Comment #4 from Segher Boessenkool <segher at gcc dot gnu.org> ---
But this PR has more info, so I'll close the older one as dup, instead.

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.0                        |11.2

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 11.1 has been released, retargeting bugs to GCC 11.2.

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.2                        |---

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[wschmidt at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

--- Comment #6 from Bill Schmidt <wschmidt at gcc dot gnu.org> ---
*** Bug 103981 has been marked as a duplicate of this bug. ***

[Bug target/95082] [11/12] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by William Schmidt <wschmidt@gcc.gnu.org>:

https://gcc.gnu.org/g:3f30f2d1dbb3228b8468b26239fe60c2974ce2ac

commit r12-7029-g3f30f2d1dbb3228b8468b26239fe60c2974ce2ac
Author: Bill Schmidt <wschmidt@linux.ibm.com>
Date:   Wed Feb 2 21:24:22 2022 -0600

    rs6000: Fix LE code gen for vec_cnt[lt]z_lsbb [PR95082]

    These built-ins were misimplemented as always having big-endian semantics.

    2022-01-18  Bill Schmidt  <wschmidt@linux.ibm.com>

    gcc/
            PR target/95082
            * config/rs6000/rs6000-builtin.cc (rs6000_expand_builtin): Handle
            endianness for vclzlsbb and vctzlsbb.
            * config/rs6000/rs6000-builtins.def (VCLZLSBB_V16QI): Change
            default pattern and indicate a different pattern will be used for
            big endian.
            (VCLZLSBB_V4SI): Likewise.
            (VCLZLSBB_V8HI): Likewise.
            (VCTZLSBB_V16QI): Likewise.
            (VCTZLSBB_V4SI): Likewise.
            (VCTZLSBB_V8HI): Likewise.

    gcc/testsuite/
            PR target/95082
            * gcc.target/powerpc/vsu/vec-cntlz-lsbb-0.c: Restrict to -mbig.
            * gcc.target/powerpc/vsu/vec-cntlz-lsbb-1.c: Likewise.
            * gcc.target/powerpc/vsu/vec-cntlz-lsbb-3.c: New.
            * gcc.target/powerpc/vsu/vec-cntlz-lsbb-4.c: New.
            * gcc.target/powerpc/vsu/vec-cnttz-lsbb-0.c: Restrict to -mbig.
            * gcc.target/powerpc/vsu/vec-cnttz-lsbb-1.c: Likewise.
            * gcc.target/powerpc/vsu/vec-cnttz-lsbb-3.c: New.
            * gcc.target/powerpc/vsu/vec-cnttz-lsbb-4.c: New.

[Bug target/95082] [11] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[wschmidt at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

Bill Schmidt <wschmidt at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |12.0
            Summary|[11/12] LE implementations  |[11] LE implementations of
                   |of vec_cnttz_lsbb and       |vec_cnttz_lsbb and
                   |vec_cntlz_lsbb are wrong    |vec_cntlz_lsbb are wrong
      Known to fail|                            |11.0

--- Comment #8 from Bill Schmidt <wschmidt at gcc dot gnu.org> ---
This still needs a backport to 11, but I won't be able to work on that.

[Bug target/95082] [11] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[segher at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

--- Comment #9 from Segher Boessenkool <segher at gcc dot gnu.org> ---
It still needs fixing on trunk as well, as discussed.

[Bug target/95082] LE implementations of vec_cnttz_lsbb and vec_cntlz_lsbb are wrong

[wschmidt at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95082

Bill Schmidt <wschmidt at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|11.0                        |
            Summary|[11] LE implementations of  |LE implementations of
                   |vec_cnttz_lsbb and          |vec_cnttz_lsbb and
                   |vec_cntlz_lsbb are wrong    |vec_cntlz_lsbb are wrong

--- Comment #10 from Bill Schmidt <wschmidt at gcc dot gnu.org> ---
>From some private discussion today, this goes back quite some ways, and we're
simply going to document this as a change in behavior.  The old behavior was
wrong and usage of these built-ins is not very prevalent.  We'll make a point
of this change in the release notes, rather than do any backports.