Function `_mm512_mask_blend_ph`

#[target_feature(enable = "avx512fp16")]
pub const fn _mm512_mask_blend_ph(k: __mmask32, a: __m512h, b: __m512h) -> __m512h

Blend packed half-precision (16-bit) floating-point elements from a and b using control mask k, and store the results in dst.