Function haccumulate

Source/astcenc_vecmathlib_avx2_8.h:868–875

* @brief Accumulate lane-wise sums for a vector, folded 4-wide. * * This is invariant with 4-wide implementations. */

 * This is invariant with 4-wide implementations.
 */
ASTCENC_SIMD_INLINE void haccumulate(vfloat4& accum, vfloat8 a)
{
	vfloat4 lo(_mm256_extractf128_ps(a.m, 0));
	haccumulate(accum, lo);

	vfloat4 hi(_mm256_extractf128_ps(a.m, 1));
	haccumulate(accum, hi);
}

/**
 * @brief Accumulate lane-wise sums for a vector.