Faster `widening_mul(int16x, int16x) -> int32x` for x86 (AVX2 and SSE2)#6677

Merged

rootjalex merged 4 commits intomainfrom

rootjalex/x86_wmul

Apr 12, 2022

Commits on Apr 5, 2022

add widening_mul using vpmaddwd for AVX2
Alexander Root
committed
add vpmaddwd test
Alexander Root
committed
move pmaddwd check
Alexander Root
committed

Commits on Apr 6, 2022

add widening_mul with pmaddwd for SSE2
Alexander Root
committed