Function safe_arch::fused_mul_sub_m256d

#[must_use]pub fn fused_mul_sub_m256d(a: m256d, b: m256d, c: m256d) -> m256d

This is supported with target feature fma only.

Lanewise fused (a * b) - c

let a = m256d::from_array([1.0, 2.0, 3.0, 4.0]);
let b = m256d::from_array([5.0, 6.0, 7.0, 8.0]);
let c = m256d::from_array([1.0, 1.0, 1.0, 1.0]);
let d = fused_mul_sub_m256d(a, b, c).to_array();
assert_eq!(d, [4.0, 11.0, 20.0, 31.0]);

• Intrinsic: _mm256_fmsub_pd
• Assembly: one of
  • vfmsub132pd ymm, ymm, ymm
  • vfmsub213pd ymm, ymm, ymm
  • vfmsub231pd ymm, ymm, ymm