| /external/libgav1/libgav1/src/dsp/x86/ |
| D | loop_restoration_avx2.cc | 1104 __m128i* const row_sq5) { in SumHorizontalLo() argument
1107 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontalLo()
1111 __m256i* const row_sq5) { in SumHorizontalLo() argument
1114 *row_sq5 = _mm256_add_epi32(sum04, *row_sq3); in SumHorizontalLo()
1118 __m128i* const row_sq5) { in SumHorizontalHi() argument
1121 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontalHi()
1125 __m256i* const row_sq5) { in SumHorizontalHi() argument
1128 *row_sq5 = _mm256_add_epi32(sum04, *row_sq3); in SumHorizontalHi()
1268 __m256i row3[2], row5[2], row_sq3[2], row_sq5[2]; in BoxSum() local
1278 SumHorizontal(sq + 0, &row_sq3[0], &row_sq3[1], &row_sq5[0], &row_sq5[1]); in BoxSum()
[all …]
|
| D | loop_restoration_sse4.cc | 939 __m128i* const row_sq5) { in SumHorizontalLo() argument
942 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontalLo()
946 __m128i* const row_sq5) { in SumHorizontalHi() argument
949 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontalHi()
1061 __m128i row3[2], row5[2], row_sq3[2], row_sq5[2]; in BoxSum() local
1071 SumHorizontal(sq + 0, &row_sq3[0], &row_sq3[1], &row_sq5[0], &row_sq5[1]); in BoxSum()
1073 StoreAligned32U32(square_sum5 + 0, row_sq5); in BoxSum()
1074 SumHorizontal(sq + 1, &row_sq3[0], &row_sq3[1], &row_sq5[0], &row_sq5[1]); in BoxSum()
1076 StoreAligned32U32(square_sum5 + 8, row_sq5); in BoxSum()
|
| D | loop_restoration_10bit_avx2.cc | 1005 __m128i* const row_sq5) { in SumHorizontal32() argument
1008 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontal32()
1012 __m256i* const row_sq5) { in SumHorizontal32() argument
1015 *row_sq5 = _mm256_add_epi32(sum04, *row_sq3); in SumHorizontal32()
1165 __m256i s[2], row3[2], row5[2], row_sq3[2], row_sq5[2]; in BoxSum() local
1182 SumHorizontal32(sq + 0, &row_sq3[0], &row_sq3[1], &row_sq5[0], in BoxSum()
1183 &row_sq5[1]); in BoxSum()
1185 StoreAligned64(square_sum5 + 0, row_sq5); in BoxSum()
1186 SumHorizontal32(sq + 4, &row_sq3[0], &row_sq3[1], &row_sq5[0], in BoxSum()
1187 &row_sq5[1]); in BoxSum()
[all …]
|
| D | loop_restoration_10bit_sse4.cc | 823 __m128i* const row_sq5) { in SumHorizontal32() argument
826 *row_sq5 = _mm_add_epi32(sum04, *row_sq3); in SumHorizontal32()
907 __m128i row3[2], row5[2], row_sq3[2], row_sq5[2]; in BoxSum() local
919 SumHorizontal32(sq + 0, &row_sq3[0], &row_sq3[1], &row_sq5[0], in BoxSum()
920 &row_sq5[1]); in BoxSum()
922 StoreAligned32U32(square_sum5 + 0, row_sq5); in BoxSum()
923 SumHorizontal32(sq + 2, &row_sq3[0], &row_sq3[1], &row_sq5[0], in BoxSum()
924 &row_sq5[1]); in BoxSum()
926 StoreAligned32U32(square_sum5 + 8, row_sq5); in BoxSum()
|
| /external/libgav1/libgav1/src/dsp/arm/ |
| D | loop_restoration_neon.cc | 827 uint32x4_t* const row_sq5) { in SumHorizontal() argument
831 *row_sq5 = vaddq_u32(sum04, *row_sq3); in SumHorizontal()
835 uint32x4x2_t* const row_sq5) { in SumHorizontal() argument
838 SumHorizontal(low, &row_sq3->val[0], &row_sq5->val[0]); in SumHorizontal()
839 SumHorizontal(high, &row_sq3->val[1], &row_sq5->val[1]); in SumHorizontal()
844 uint32x4x2_t* const row_sq3, uint32x4x2_t* const row_sq5) { in SumHorizontal() argument
846 SumHorizontal(sq, row_sq3, row_sq5); in SumHorizontal()
851 uint32x4x2_t* const row_sq3, uint32x4x2_t* const row_sq5) { in SumHorizontal() argument
855 return SumHorizontal(s, sq, row3, row5, row_sq3, row_sq5); in SumHorizontal()
919 uint32x4x2_t row_sq3, row_sq5; in BoxSum() local
[all …]
|