Fastest way to horizontally sum SSE unsigned byte vector

问题

I need to horizontally add a __m128i that is 16 x epi8 values. The XOP instructions would make this trivial, but I don't have those available.

Current method is:

hd = _mm_hadd_epi16(_mm_cvtepi8_epi16(sum), _mm_cvtepi8_epi16(_mm_shuffle_epi8(sum, swap)));
hd = _mm_hadd_epi16(hd, hd);
hd = _mm_hadd_epi16(hd, hd);

Is there a better way with up to SSE4.1?

回答1:

You can do it with SSE2's _mm_sad_epu8 (psadbw), e.g.:

inline uint32_t _mm_sum_epu8(const __m128i v)
{
    __m128i vsum = _mm_sad_epu8(v, _mm_setzero_si128());
    return _mm_extract_epi16(vsum, 0) + _mm_extract_epi16(vsum, 4);
}

来源：https://stackoverflow.com/questions/36998538/fastest-way-to-horizontally-sum-sse-unsigned-byte-vector