https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89929
--- Comment #14 from H.J. Lu <hjl.tools at gmail dot com> --- Since all AVX512BW processors also have AVX512DQ and AVX512VL, we shouldn't optimize a function with just AVX512BW, but without AVX512DQ and AVX512VL. We should add -misa=AVX512-subset to enable a subset of AVX512XX: 1. PTA_AVX512 = PTA_AVX512F | PTA_AVX512CD 2. PTA_AVX512SKYLAKE = PTA_AVX512 | PTA_AVX512VL | PTA_AVX512BW | PTA_AVX512DQ ....