Macro safe_arch::shuffle_m256

macro_rules! shuffle_m256 {
    ($a:expr, $b:expr, $z:expr, $o:expr, $t:expr, $e:expr) => { ... };
}

This is supported with target feature avx only.

Shuffles the f32 lanes around.

• args are 0, 1, 2, 3 for which lane to use in the lower or upper half.
• the same pattern is used for the four low lanes and the four high lanes.
• a low, a low, b low, b low, a high, a high, b high, b high

let a = m256::from_array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0]);
let b = m256::from_array([9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0]);
//
let c = shuffle_m256!(a, b, 1, 3, 2, 0).to_array();
assert_eq!(c, [2.0, 4.0, 11.0, 9.0, 6.0, 8.0, 15.0, 13.0]);