[Bug tree-optimization/118267] Suboptimal code for bool bitfield tests

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118267

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
                 CC|                            |aoliva at gcc dot gnu.org
            Version|unknown                     |15.0
           Keywords|                            |missed-optimization
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2025-01-07

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

  <bb 2> [local count: 1073741824]:
  _3 = rt_2(D)->cfg1;
  if (_3 != 0)
    goto <bb 5>; [50.00%]
  else
    goto <bb 3>; [50.00%]

  <bb 3> [local count: 536870912]:
  _6 = rt_2(D)->cfg2;
  if (_6 != 0)
    goto <bb 5>; [50.00%]
  else
    goto <bb 4>; [50.00%]

  <bb 4> [local count: 268435456]:
  _7 = rt_2(D)->cfg3;

  <bb 5> [local count: 1073741824]:
  # iftmp.0_5 = PHI <1(2), _7(4), 1(3)>
  return iftmp.0_5;

Maybe it's confused about the value-replacements in the PHI (the constant 1)
or it's confused by the lack of a compare in BB 4 (it's implicitly _7 != 0).