Function core::arch::x86_64::_mm256_addsub_ps

pub unsafe fn _mm256_addsub_ps(a: __m256, b: __m256) -> __m256

This is supported on x86-64 and target feature avx only.

Alternatively adds and subtracts packed single-precision (32-bit) floating-point elements in a to/from packed elements in b.

Intel's documentation