Function safe_arch::unpack_hi_m256

#[must_use]pub fn unpack_hi_m256(a: m256, b: m256) -> m256

This is supported with target feature avx only.

Unpack and interleave the high lanes.

let a = m256::from_array([1.0, 2.0, 3.0, 4.0, 20.0, 30.0, 40.0, 50.0]);
let b = m256::from_array([59.0, 61.0, 79.0, 81.5, 90.0, 100.0, 110.0, 51.0]);
let c = unpack_hi_m256(a, b).to_array();
assert_eq!(c, [3.0, 79.0, 4.0, 81.5, 40.0, 110.0, 50.0, 51.0]);