| /external/libaom/libaom/aom_dsp/x86/ |
| D | convolve_sse2.h | 79 const __m128i *const res_unsigned, in comp_avg() argument
84 const __m128i data_lo = _mm_unpacklo_epi16(*data_ref_0, *res_unsigned); in comp_avg()
85 const __m128i data_hi = _mm_unpackhi_epi16(*data_ref_0, *res_unsigned); in comp_avg()
95 const __m128i wt_res = _mm_add_epi16(*data_ref_0, *res_unsigned); in comp_avg()
101 static INLINE __m128i convolve_rounding(const __m128i *const res_unsigned, in convolve_rounding() argument
105 const __m128i res_signed = _mm_sub_epi16(*res_unsigned, *offset_const); in convolve_rounding()
112 const __m128i *const res_unsigned, const __m128i *const offset_const, in highbd_convolve_rounding_sse2() argument
114 const __m128i res_signed = _mm_sub_epi32(*res_unsigned, *offset_const); in highbd_convolve_rounding_sse2()
|
| D | convolve_avx2.h | 179 const __m256i res_unsigned = _mm256_add_epi16(res_16b, offset_const); \
185 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); \
198 const __m128i res_0 = _mm256_castsi256_si128(res_unsigned); \
201 const __m128i res_1 = _mm256_extracti128_si256(res_unsigned, 1); \
207 const __m256i res_unsigned = _mm256_add_epi16(res_16b, offset_const); \
214 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); \
228 const __m128i res_0 = _mm256_castsi256_si128(res_unsigned); \
231 const __m128i res_1 = _mm256_extracti128_si256(res_unsigned, 1); \
378 const __m256i *const res_unsigned, in comp_avg() argument
383 const __m256i data_lo = _mm256_unpacklo_epi16(*data_ref_0, *res_unsigned); in comp_avg()
[all …]
|
| D | convolve_sse4_1.h | 35 const __m128i *const res_unsigned, in highbd_comp_avg_sse4_1() argument
42 const __m128i wt1_res = _mm_mullo_epi32(*res_unsigned, *wt1); in highbd_comp_avg_sse4_1()
47 const __m128i wt_res = _mm_add_epi32(*data_ref_0, *res_unsigned); in highbd_comp_avg_sse4_1()
|
| /external/libaom/libaom/av1/common/x86/ |
| D | jnt_convolve_sse2.c | 73 const __m128i res_unsigned = _mm_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_x_sse2() local
80 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_x_sse2()
88 _mm_store_si128((__m128i *)(&dst[0]), res_unsigned); in av1_dist_wtd_convolve_x_sse2()
129 const __m128i res_unsigned = _mm_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_x_sse2() local
137 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_x_sse2()
145 _mm_store_si128((__m128i *)(&dst[i * dst_stride + j]), res_unsigned); in av1_dist_wtd_convolve_x_sse2()
221 __m128i res_unsigned = _mm_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_y_sse2() local
228 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_y_sse2()
237 _mm_store_si128((__m128i *)dst, res_unsigned); in av1_dist_wtd_convolve_y_sse2()
250 res_unsigned = _mm_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_y_sse2()
[all …]
|
| D | jnt_convolve_avx2.c | 96 const __m256i res_unsigned = _mm256_add_epi16(res, offset_const); in av1_dist_wtd_convolve_x_avx2() local
103 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_x_avx2()
123 const __m128i res_0 = _mm256_castsi256_si128(res_unsigned); in av1_dist_wtd_convolve_x_avx2()
126 const __m128i res_1 = _mm256_extracti128_si256(res_unsigned, 1); in av1_dist_wtd_convolve_x_avx2()
151 const __m256i res_unsigned = _mm256_add_epi16(res, offset_const); in av1_dist_wtd_convolve_x_avx2() local
158 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_x_avx2()
178 const __m128i res_0 = _mm256_castsi256_si128(res_unsigned); in av1_dist_wtd_convolve_x_avx2()
181 const __m128i res_1 = _mm256_extracti128_si256(res_unsigned, 1); in av1_dist_wtd_convolve_x_avx2()
718 const __m256i res_unsigned = _mm256_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_2d_avx2() local
724 const __m256i comp_avg_res = comp_avg(&data_ref_0, &res_unsigned, in av1_dist_wtd_convolve_2d_avx2()
[all …]
|
| D | jnt_convolve_ssse3.c | 205 const __m128i res_unsigned = _mm_add_epi16(res_16b, offset_const); in av1_dist_wtd_convolve_2d_ssse3() local
213 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_2d_ssse3()
226 _mm_store_si128((__m128i *)(&dst[i * dst_stride + j]), res_unsigned); in av1_dist_wtd_convolve_2d_ssse3()
|
| D | convolve_2d_sse2.c | 445 const __m128i res_unsigned = _mm_add_epi16(res, offset_const); in av1_dist_wtd_convolve_2d_copy_sse2() local
451 comp_avg(&data_ref_0, &res_unsigned, &wt, use_dist_wtd_comp_avg); in av1_dist_wtd_convolve_2d_copy_sse2()
463 _mm_store_si128((__m128i *)(&dst[j]), res_unsigned); in av1_dist_wtd_convolve_2d_copy_sse2()
|