pub unsafe fn _mm_moveldup_ps(a: __m128) -> __m128
Duplicate even-indexed single-precision (32-bit) floating-point elements from a.
a
Intel’s documentation