Function _mm256_insertf128_ps
#[target_feature(enable = "avx")]
pub const fn _mm256_insertf128_ps(a: __m256, b: __m128, IMM1: i32) -> __m256
Copies a to result, then inserts 128 bits (composed of 4 packed
single-precision (32-bit) floating-point elements) from b into result
at the location specified by imm8.