Function _mm256_mask_multishift_epi64_epi8
#[target_feature(enable = "avx512vbmi", enable = "avx512vl")]
pub fn _mm256_mask_multishift_epi64_epi8(src: __m256i, k: __mmask32, a: __m256i, b: __m256i) -> __m256iFor each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using writemask k (elements are copied from src when the corresponding mask bit is not set).