Function _mm256_mask_blend_ps 

pub fn _mm256_mask_blend_ps(k: __mmask8, a: __m256, b: __m256) -> __m256

Available on x86 and target feature avx512f and target feature avx512vl only.

Blend packed single-precision (32-bit) floating-point elements from a and b using control mask k, and store the results in dst.

Intel’s documentation