https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87218
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target Milestone|--- |10.0
Status|NEW |RESOLVED
Known to fail| |9.5.0
Resolution|--- |FIXED
Known to work| |10.1.0, 13.2.1, 14.0
--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Seems to be good now. Trunk at -O2 takes 2.5s for me. -O1 is slower at 18s
with
dead store elim1 : 8.44 ( 47%) 0.00 ( 0%) 8.44 ( 46%)
1991k ( 1%)
combiner : 6.97 ( 39%) 0.00 ( 0%) 6.96 ( 38%)
16M ( 12%)
with -O0 we take 0.5s
Quite likely we optimize this too well now to trigger the original behavior
(but I do remember some cfgleanup compile-time fixes).
-O2 times
11.4: 9s
10.5: 12.7s
9.5: 464s
so fixed for GCC 10.