pub fn _mm_maskz_mov_ps(k: u8, a: __m128) -> __m128
Move packed single-precision (32-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Intel’s documentation