pub unsafe fn _mm_blendv_ps(a: __m128, b: __m128, mask: __m128) -> __m128
Blend packed single-precision (32-bit) floating-point elements from a and b using mask
a
b
mask
Intel’s documentation