Function safe_arch::sub_m256d

#[must_use]pub fn sub_m256d(a: m256d, b: m256d) -> m256d

This is supported with target feature avx only.

Lanewise a - b with f64 lanes.

let a = m256d::from_array([1.0, 2.0, 3.0, 4.0]);
let b = m256d::from_array([5.0, 60.0, 712.0, 8.5]);
let c = sub_m256d(a, b).to_array();
assert_eq!(c, [-4.0, -58.0, -709.0, -4.5]);