https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95784
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Component|tree-optimization |target
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I dont see an issue here. Gcc is avoiding one cmov by doing a sign extend from
bit 0 (sal/sar) and an and instruction.
Cmov on x86 processors is a bit weird and sometimes slower than using
arthematic instructions.