[Bug tree-optimization/109722] New: ABSU<a> == 0 is not optimized to just a == 0

[Bug tree-optimization/109722] New: ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

            Bug ID: 109722
           Summary: ABSU<a> == 0 is not optimized to just a == 0
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: enhancement
          Priority: P3
         Component: tree-optimization
          Assignee: pinskia at gcc dot gnu.org
          Reporter: pinskia at gcc dot gnu.org
  Target Milestone: ---

Take:
```
int f(signed char a)
{
        int t = a;
        if (t < 0) t = -t;
        a = t;
        return a == 0;
}
```
Right now this gets optimized to just:
```
  _9 = ABSU_EXPR <a_3(D)>;
  _1 = _9 == 0;
  _7 = (int) _1;
```

If we change `signed char` to `unsigned char`, we get:
  _1 = a_2(D) == 0;
  _3 = (int) _1;

As we had expected for the original case too.

The fix is to change in match.pd:
/* Convert ABS_EXPR<x> == 0 or ABS_EXPR<x> != 0 to x == 0 or x != 0.  */
(for op (eq ne)
 (simplify
  (op (abs @0) zerop@1)
  (op @0 @1)))
to:
/* Convert ABS[U]_EXPR<x> == 0 or ABS[U]_EXPR<x> != 0 to x == 0 or x != 0.  */
(for op (abs absu)
 (for eqne (eq ne)
  (simplify
   (eqne (op @0) zerop@1)
   (eqne @0 @1)))

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2023-05-04
     Ever confirmed|0                           |1

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note I accidently found this while creating a patch for PR 101541 and I had
messed up creating ABSU when I didn't need to and it was causing the testsuite
patterns in tree-ssa/vrp20.c  not to match (though it was optimized correctly
still).

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This is the corrected version which surives a bootstrap:
/* Convert ABS[U]_EXPR<x> == 0 or ABS[U]_EXPR<x> != 0 to x == 0 or x != 0.  */
(for op (abs absu)
 (for eqne (eq ne)
  (simplify
   (eqne (op @0) zerop@1)
   (eqne @0 { build_zero_cst (TREE_TYPE (@0)); }))))

I forgot the type of ABSU is always unsigned while its operand can be
different.

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch
                URL|                            |https://gcc.gnu.org/piperma
                   |                            |il/gcc-patches/2023-May/617
                   |                            |494.html

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Patch submitted:
https://gcc.gnu.org/pipermail/gcc-patches/2023-May/617494.html

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Andrew Pinski <pinskia@gcc.gnu.org>:

https://gcc.gnu.org/g:6fe385eac6ff8ecddb6cbdff2c706b27b5137006

commit r14-488-g6fe385eac6ff8ecddb6cbdff2c706b27b5137006
Author: Andrew Pinski <apinski@marvell.com>
Date:   Thu May 4 23:37:51 2023 +0000

    MATCH: Add ABSU<a> == 0 to a == 0 simplification

    There is already an `ABS<a> == 0` to `a == 0` pattern,
    this just extends that to ABSU too.

    OK? Bootstrapped and tested on x86_64-linux-gnu with no regressions.

            PR tree-optimization/109722

    gcc/ChangeLog:

            * match.pd: Extend the `ABS<a> == 0` pattern
            to cover `ABSU<a> == 0` too.

    gcc/testsuite/ChangeLog:

            * gcc.dg/tree-ssa/abs-1.c: New test.

[Bug tree-optimization/109722] ABSU<a> == 0 is not optimized to just a == 0

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109722

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |14.0
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed.