[Bug tree-optimization/95527] New: Failure to optimize __builtin_ffs == 0

[Bug tree-optimization/95527] New: Failure to optimize __builtin_ffs == 0

[gabravier at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

            Bug ID: 95527
           Summary: Failure to optimize __builtin_ffs == 0
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gabravier at gmail dot com
  Target Milestone: ---

bool f(int x)
{
    return __builtin_ffs(x) == 0;
}

This can be optimized to `x == 0`. This transformation is done by LLVM, but not
by GCC.

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
           Keywords|                            |easyhack
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2020-06-04

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.  Possibly related clz, ctz, popcount cases (and parity?)

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 48705
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48705&action=edit
gcc11-pr95527.patch

Untested fix for ffs.  Haven't touched others.  I think parity can't be
optimized that way, for ctz/clz we'd need to decide how do we deal with the
undefined at zero fact (vs. e.g. the aarch64 clz/ctz defined already at GIMPLE
to a particular value through *_DEFINED_VALUE_AT_ZERO giving 2).
__builtin_popcount (x) == 0 (and != 0) should be easy, but for that one not
sure if we can optimize much more than that.

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:653ab081391e9e7e38b304f3234323c93693d40c

commit r11-1080-g653ab081391e9e7e38b304f3234323c93693d40c
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Jun 9 08:38:19 2020 +0200

    match.pd: Optimize ffs comparisons against constants [PR95527]

    The following patch implements various optimizations of __builtin_ffs*
    against constants.

    2020-06-09  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/95527
            * match.pd (__builtin_ffs (X) cmp CST): New optimizations.

            * gcc.dg/tree-ssa/pr95527.c: New test.

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

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

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

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
For popcount and clz we already have patterns for comparison of the result with
integers.
For ctz not sure what useful simplifications we can do, perhaps just
ctz(x) == bitsize-1 (and !=) to x == 1 << (bitsize-1) (or !=)?
Ah, maybe also inequalifies, ctz(x) >= 3 to (x & ((1 << 3) - 1)) == 0 ?

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:75f8900159133ce069ef1d2edf3b67c7bc82e305

commit r12-151-g75f8900159133ce069ef1d2edf3b67c7bc82e305
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Apr 27 14:47:54 2021 +0200

    match.pd: Add some __builtin_ctz (x) cmp cst simplifications [PR95527]

    This patch adds some ctz simplifications (e.g. ctz (x) >= 3 can be done by
    testing if the low 3 bits are zero, etc.).

    In addition, I've noticed that in the CLZ case, the
     #ifdef CLZ_DEFINED_VALUE_AT_ZERO don't really work as intended, they
    are evaluated during genmatch and the macro is not defined then
    (but, because of the missing tm.h includes it isn't defined in
    gimple-match.c or generic-match.c either).  And when tm.h is included,
    defaults.h is included which defines a fallback version of that macro.

    For GCC 12, I wonder if it wouldn't be better to say in addition to
__builtin_c[lt]z*
    is always UB at zero that it would be undefined for .C[LT]Z ifn too if it
    has just one operand and use a second operand to be the constant we expect
    at zero.

    2021-04-27  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/95527
            * generic-match-head.c: Include tm.h.
            * gimple-match-head.c: Include tm.h.
            * match.pd (CLZ == INTEGER_CST): Don't use
            #ifdef CLZ_DEFINED_VALUE_AT_ZERO, only test
CLZ_DEFINED_VALUE_AT_ZERO
            if clz == CFN_CLZ.  Add missing val declaration.
            (CTZ cmp CST): New simplifications.

            * gcc.dg/tree-ssa/pr95527-2.c: New test.

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement

[Bug tree-optimization/95527] Failure to optimize __builtin_ffs == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95527

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
   Target Milestone|---                         |11.0
             Status|NEW                         |RESOLVED

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed.