https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70277
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Last reconfirmed|2016-03-17 00:00:00 |2024-12-20
--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Unrelated to the original mentioned problem is we have a missed optimization,
take the reduced testcase in comment #3:
_4 = MEM[(const char * const *)&C.1 + ivtmp.185_189 * 8];
...
if (_4 == 0B)
goto <bb 4>; [0.00%]
else
goto <bb 5>; [100.00%]
That comparison against 0 should have been optimized away as we know none of
elements of C.1 are nullptr:
C.1.0:
.quad .LC2
.quad .LC3
.quad .LC4
Is there a way to mark that as such?
