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

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

Intel’s documentation