X-Git-Url: https://git.sesse.net/?p=stockfish;a=blobdiff_plain;f=src%2Fnnue%2Flayers%2Fclipped_relu.h;h=7ade598f4d21788d3725a180f4828b7e405cdf8b;hp=7e5fcf4ac59399ac0a2310763fd9e30da689d9bf;hb=651ec3b31ee68db50f38ccd8fcdedbd6673cd9ed;hpb=27b593a94477a821f80a041320683f805114d4a3

diff --git a/src/nnue/layers/clipped_relu.h b/src/nnue/layers/clipped_relu.h
index 7e5fcf4a..7ade598f 100644
--- a/src/nnue/layers/clipped_relu.h
+++ b/src/nnue/layers/clipped_relu.h
@@ -74,13 +74,50 @@ namespace Eval::NNUE::Layers {
       const auto out = reinterpret_cast<__m256i*>(output);
       for (IndexType i = 0; i < kNumChunks; ++i) {
         const __m256i words0 = _mm256_srai_epi16(_mm256_packs_epi32(
-          _mm256_load_si256(&in[i * 4 + 0]),
-          _mm256_load_si256(&in[i * 4 + 1])), kWeightScaleBits);
+
+  #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);
         const __m256i words1 = _mm256_srai_epi16(_mm256_packs_epi32(
-          _mm256_load_si256(&in[i * 4 + 2]),
-          _mm256_load_si256(&in[i * 4 + 3])), kWeightScaleBits);
-        _mm256_store_si256(
-            &out[i], _mm256_permutevar8x32_epi32(_mm256_max_epi8(
+
+  #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_packs_epi16(words0, words1), kZero), kOffsets));
       }
       constexpr IndexType kStart = kNumChunks * kSimdWidth;