const auto out = reinterpret_cast<__m256i*>(output);
for (IndexType i = 0; i < kNumChunks; ++i) {
const __m256i words0 = _mm256_srai_epi16(_mm256_packs_epi32(
-
- #if defined(__MINGW32__) || defined(__MINGW64__)
- // HACK: Use _mm256_loadu_si256() instead of _mm256_load_si256. Because the binary
- // compiled with g++ in MSYS2 crashes here because the output memory is not aligned
- // even though alignas is specified.
- _mm256_loadu_si256
- #else
- _mm256_load_si256
- #endif
-
- (&in[i * 4 + 0]),
-
- #if defined(__MINGW32__) || defined(__MINGW64__)
- _mm256_loadu_si256
- #else
- _mm256_load_si256
- #endif
-
- (&in[i * 4 + 1])), kWeightScaleBits);
+ _mm256_loadA_si256(&in[i * 4 + 0]),
+ _mm256_loadA_si256(&in[i * 4 + 1])), kWeightScaleBits);
const __m256i words1 = _mm256_srai_epi16(_mm256_packs_epi32(
-
- #if defined(__MINGW32__) || defined(__MINGW64__)
- _mm256_loadu_si256
- #else
- _mm256_load_si256
- #endif
-
- (&in[i * 4 + 2]),
-
- #if defined(__MINGW32__) || defined(__MINGW64__)
- _mm256_loadu_si256
- #else
- _mm256_load_si256
- #endif
-
- (&in[i * 4 + 3])), kWeightScaleBits);
-
- #if defined(__MINGW32__) || defined(__MINGW64__)
- _mm256_storeu_si256
- #else
- _mm256_store_si256
- #endif
-
- (&out[i], _mm256_permutevar8x32_epi32(_mm256_max_epi8(
+ _mm256_loadA_si256(&in[i * 4 + 2]),
+ _mm256_loadA_si256(&in[i * 4 + 3])), kWeightScaleBits);
+ _mm256_storeA_si256(&out[i], _mm256_permutevar8x32_epi32(_mm256_max_epi8(
_mm256_packs_epi16(words0, words1), kZero), kOffsets));
}
constexpr IndexType kStart = kNumChunks * kSimdWidth;