public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug tree-optimization/111762] New: `(a | 3) ^ 1` and `(a & ~1) | 2` should produce the same code
@ 2023-10-10 19:25 pinskia at gcc dot gnu.org
  0 siblings, 0 replies; only message in thread
From: pinskia at gcc dot gnu.org @ 2023-10-10 19:25 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111762

            Bug ID: 111762
           Summary: `(a | 3) ^ 1` and `(a & ~1) | 2` should produce the
                    same code
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Keywords: internal-improvement
          Severity: enhancement
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: pinskia at gcc dot gnu.org
  Target Milestone: ---

Take:
```
int f(int in) {
  in = in | 3;
  in = in ^ 1;
  return in;
}

int f1(int in) {
  in = (in & ~(unsigned long)1);
  in = in | 2;
  return in;
}
```

These 2 functions produce the same result but the code generated is different.

I noticed this while writing the patch for PR 111282 and noticing that
`gcc.dg/tree-ssa/and-1.c` testcase now fails (test needs to be fixed).

LLVM/clang looks like canonicalizes to f1. I suspect xor is harder to optimize
when dealing with reassociation so doing f1 is better.

^ permalink raw reply	[flat|nested] only message in thread
only message in thread, other threads:[~2023-10-10 19:25 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-10-10 19:25 [Bug tree-optimization/111762] New: `(a | 3) ^ 1` and `(a & ~1) | 2` should produce the same code pinskia at gcc dot gnu.org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).