Searched refs:shuffled_v (Results 1 – 2 of 2) sorted by relevance
/external/ruy/ruy/ |
D | kernel_x86.h | 313 __m256i shuffled_v; 317 shuffled_v = intrin_utils::mm256_shuffle_epi8<path>(v, repack_perm); 326 mm_storeu_si16<path>(dst, mm256_extracti128_si256<path>(shuffled_v, 0)); 329 __m128i trailing_packed = mm256_extracti128_si256<path>(shuffled_v, 0); 335 mm_storeu_si32<path>(dst, mm256_extracti128_si256<path>(shuffled_v, 0)); 338 mm_storeu_si32<path>(dst, mm256_extracti128_si256<path>(shuffled_v, 0)); 339 dst[4] = _mm256_extract_epi8(shuffled_v, 16); 342 mm_storeu_si32<path>(dst, mm256_extracti128_si256<path>(shuffled_v, 0)); 344 mm256_extracti128_si256<path>(shuffled_v, 1)); 347 mm_storeu_si32<path>(dst, mm256_extracti128_si256<path>(shuffled_v, 0)); [all …]
|
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
D | avx2_quantization_utils.h | 92 const __m256i shuffled_v = _mm256_shuffle_epi8(v, repack_perm); in CastInt32ToInt16AndStore() local 93 mm_storeu_si64(dst, _mm256_extracti128_si256(shuffled_v, 0)); in CastInt32ToInt16AndStore() 94 mm_storeu_si64(dst + 4, _mm256_extracti128_si256(shuffled_v, 1)); in CastInt32ToInt16AndStore()
|