pub fn _mm256_mask_mov_epi32(src: __m256i, k: u8, a: __m256i) -> __m256i
Move packed 32-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Intel’s documentation