pub unsafe fn _mm256_movemask_ps(a: __m256) -> i32
Sets each bit of the returned mask based on the most significant bit of the corresponding packed single-precision (32-bit) floating-point element in a.
a
Intel’s documentation