pub unsafe fn _mm_movelh_ps(a: __m128, b: __m128) -> __m128
This is supported on x86-64 and target feature sse only.
Expand description

Combine lower half of a and b. The lower half of b occupies the higher half of result.

Intel’s documentation