Function core::arch::x86::_mm_blend_ps

pub unsafe fn _mm_blend_ps<const IMM4: i32>(a: __m128, b: __m128) -> __m128

This is supported on x86 and target feature sse4.1 only.

使用掩码 IMM4 从 a 和 b 混合包装的单精度 (32-bit) 浮点元素

Intel’s documentation