https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118776
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jakub at gcc dot gnu.org
--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
#include <x86intrin.h>
__v32qi a[4];
__m256i
foo (__m256i b, __m256i w, __mmask32 m)
{
return _mm256_mask_max_epu8 (w, m, ((__m256i *) &a)[2], b);
}
unfortunately doesn't reproduce it, wonder what keeps the MEM in the insn.
