| /external/libaom/libaom/aom_dsp/x86/ |
| D | aom_subpixel_8t_intrin_sse2.c | 36 const __m128i tmp_0 = _mm_unpacklo_epi32(filtersReg, filtersReg); in aom_filter_block1d16_h4_sse2() local
40 secondFilters = _mm_unpackhi_epi64(tmp_0, tmp_0); // coeffs 2 3 2 3 2 3 2 3 in aom_filter_block1d16_h4_sse2()
121 __m128i tmp_0, tmp_1; in aom_filter_block1d16_v4_sse2() local
171 tmp_0 = _mm_madd_epi16(resReg23_lo_1, secondFilters); in aom_filter_block1d16_v4_sse2()
173 resReg23_lo = _mm_packs_epi32(tmp_0, tmp_1); in aom_filter_block1d16_v4_sse2()
175 tmp_0 = _mm_madd_epi16(resReg34_lo_1, secondFilters); in aom_filter_block1d16_v4_sse2()
177 resReg34_lo = _mm_packs_epi32(tmp_0, tmp_1); in aom_filter_block1d16_v4_sse2()
181 tmp_0 = _mm_madd_epi16(resReg45_lo_1, thirdFilters); in aom_filter_block1d16_v4_sse2()
183 resReg45_lo = _mm_packs_epi32(tmp_0, tmp_1); in aom_filter_block1d16_v4_sse2()
187 tmp_0 = _mm_madd_epi16(resReg56_lo_1, thirdFilters); in aom_filter_block1d16_v4_sse2()
[all …]
|
| D | highbd_convolve_sse2.c | 120 const __m128i tmp_0 = _mm_unpacklo_epi32(filtersReg, filtersReg); in aom_highbd_filter_block1d4_h4_sse2() local
124 secondFilters = _mm_unpackhi_epi64(tmp_0, tmp_0); // coeffs 2 3 2 3 2 3 2 3 in aom_highbd_filter_block1d4_h4_sse2()
281 const __m128i tmp_0 = _mm_unpacklo_epi32(filtersReg, filtersReg); in aom_highbd_filter_block1d8_h4_sse2() local
285 secondFilters = _mm_unpackhi_epi64(tmp_0, tmp_0); // coeffs 2 3 2 3 2 3 2 3 in aom_highbd_filter_block1d8_h4_sse2()
|
| D | masked_variance_intrin_ssse3.c | 946 const __m128i tmp_0 = _mm_unpacklo_epi32(prod, sign); in highbd_masked_variance() local
948 sum_sq = _mm_add_epi64(sum_sq, _mm_add_epi64(tmp_0, tmp_1)); in highbd_masked_variance()
|
| /external/aac/libAACenc/src/ |
| D | chaosmeasure.cpp | 138 FIXP_DBL tmp_0 = left_0_div2 + (right_0 >> 1); in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast() local
144 if (tmp_0 < center_0) { in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast()
146 tmp_0 = schur_div(tmp_0 << leadingBits, center_0 << leadingBits, 8); in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast()
147 tmp_0 = fMult(tmp_0, tmp_0); in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast()
149 tmp_0 = (FIXP_DBL)MAXVAL_DBL; in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast()
151 chaosMeasure[j + 0] = tmp_0; in FDKaacEnc_FDKaacEnc_CalculateChaosMeasurePeakFast()
|
| /external/libaom/libaom/av1/common/x86/ |
| D | highbd_wiener_convolve_ssse3.c | 52 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_highbd_wiener_convolve_add_src_ssse3() local
57 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_highbd_wiener_convolve_add_src_ssse3()
59 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_highbd_wiener_convolve_add_src_ssse3()
120 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_highbd_wiener_convolve_add_src_ssse3() local
125 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_highbd_wiener_convolve_add_src_ssse3()
127 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_highbd_wiener_convolve_add_src_ssse3()
|
| D | wiener_convolve_sse2.c | 51 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_wiener_convolve_add_src_sse2() local
56 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_wiener_convolve_add_src_sse2()
58 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_wiener_convolve_add_src_sse2()
118 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_wiener_convolve_add_src_sse2() local
123 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_wiener_convolve_add_src_sse2()
125 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_wiener_convolve_add_src_sse2()
|
| D | jnt_convolve_ssse3.c | 62 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_dist_wtd_convolve_2d_ssse3() local
67 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_ssse3()
69 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_ssse3()
130 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_dist_wtd_convolve_2d_ssse3() local
135 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_ssse3()
137 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_ssse3()
|
| D | highbd_convolve_2d_sse4.c | 215 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_highbd_dist_wtd_convolve_2d_sse4_1() local
220 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_highbd_dist_wtd_convolve_2d_sse4_1()
222 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_highbd_dist_wtd_convolve_2d_sse4_1()
282 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_highbd_dist_wtd_convolve_2d_sse4_1() local
287 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_highbd_dist_wtd_convolve_2d_sse4_1()
289 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_highbd_dist_wtd_convolve_2d_sse4_1()
|
| D | convolve_2d_sse2.c | 52 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_convolve_2d_sr_sse2() local
57 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_convolve_2d_sr_sse2()
59 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_convolve_2d_sr_sse2()
118 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_convolve_2d_sr_sse2() local
123 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_convolve_2d_sr_sse2()
125 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_convolve_2d_sr_sse2()
|
| D | warp_plane_sse4.c | 273 const __m128i tmp_0 = _mm_loadl_epi64( in prepare_horizontal_filter_coeff() local
291 const __m128i tmp_8 = _mm_unpacklo_epi16(tmp_0, tmp_2); in prepare_horizontal_filter_coeff()
321 const __m128i tmp_0 = in prepare_horizontal_filter_coeff_alpha0() local
326 tmp_0, _mm_loadu_si128((__m128i *)shuffle_alpha0_mask01)); in prepare_horizontal_filter_coeff_alpha0()
329 tmp_0, _mm_loadu_si128((__m128i *)shuffle_alpha0_mask23)); in prepare_horizontal_filter_coeff_alpha0()
332 tmp_0, _mm_loadu_si128((__m128i *)shuffle_alpha0_mask45)); in prepare_horizontal_filter_coeff_alpha0()
335 tmp_0, _mm_loadu_si128((__m128i *)shuffle_alpha0_mask67)); in prepare_horizontal_filter_coeff_alpha0()
459 const __m128i tmp_0 = _mm_loadu_si128( in prepare_vertical_filter_coeffs() local
468 const __m128i tmp_8 = _mm_unpacklo_epi32(tmp_0, tmp_2); in prepare_vertical_filter_coeffs()
470 const __m128i tmp_12 = _mm_unpackhi_epi32(tmp_0, tmp_2); in prepare_vertical_filter_coeffs()
[all …]
|
| D | highbd_warp_plane_sse4.c | 38 const __m128i tmp_0 = _mm_loadu_si128( in highbd_prepare_horizontal_filter_coeff() local
48 const __m128i tmp_8 = _mm_unpacklo_epi32(tmp_0, tmp_2); in highbd_prepare_horizontal_filter_coeff()
52 const __m128i tmp_12 = _mm_unpackhi_epi32(tmp_0, tmp_2); in highbd_prepare_horizontal_filter_coeff()
89 const __m128i tmp_0 = _mm_loadu_si128( in highbd_prepare_horizontal_filter_coeff_alpha0() local
93 tmp_0, _mm_loadu_si128((__m128i *)highbd_shuffle_alpha0_mask0)); in highbd_prepare_horizontal_filter_coeff_alpha0()
95 tmp_0, _mm_loadu_si128((__m128i *)highbd_shuffle_alpha0_mask1)); in highbd_prepare_horizontal_filter_coeff_alpha0()
97 tmp_0, _mm_loadu_si128((__m128i *)highbd_shuffle_alpha0_mask2)); in highbd_prepare_horizontal_filter_coeff_alpha0()
99 tmp_0, _mm_loadu_si128((__m128i *)highbd_shuffle_alpha0_mask3)); in highbd_prepare_horizontal_filter_coeff_alpha0()
456 const __m128i tmp_0 = _mm_loadu_si128( in av1_highbd_warp_affine_sse4_1() local
469 const __m128i tmp_8 = _mm_unpacklo_epi32(tmp_0, tmp_2); in av1_highbd_warp_affine_sse4_1()
[all …]
|
| D | jnt_convolve_sse2.c | 431 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_x, coeffs_x); in av1_dist_wtd_convolve_2d_sse2() local
436 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_sse2()
438 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_sse2()
514 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in av1_dist_wtd_convolve_2d_sse2() local
519 const __m128i coeff_01 = _mm_unpacklo_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_sse2()
521 const __m128i coeff_23 = _mm_unpackhi_epi64(tmp_0, tmp_0); in av1_dist_wtd_convolve_2d_sse2()
|
| D | convolve_sse2.c | 28 const __m128i tmp_0 = _mm_unpacklo_epi32(coeffs_y, coeffs_y); in prepare_coeffs() local
32 coeffs[0] = _mm_unpacklo_epi64(tmp_0, tmp_0); // coeffs 0 1 0 1 0 1 0 1 in prepare_coeffs()
33 coeffs[1] = _mm_unpackhi_epi64(tmp_0, tmp_0); // coeffs 2 3 2 3 2 3 2 3 in prepare_coeffs()
|
| /external/libvpx/libvpx/vpx_dsp/x86/ |
| D | post_proc_sse2.c | 38 __m128i tmp_0, tmp_1; in vpx_mbpost_proc_down_sse2() local
52 tmp_0 = _mm_mullo_epi16(sum, s); in vpx_mbpost_proc_down_sse2()
55 sumsq_0 = _mm_unpacklo_epi16(tmp_0, tmp_1); in vpx_mbpost_proc_down_sse2()
56 sumsq_1 = _mm_unpackhi_epi16(tmp_0, tmp_1); in vpx_mbpost_proc_down_sse2()
|
| D | vpx_subpixel_8t_intrin_avx2.c | 358 __m256i tmp_0, tmp_1; in vpx_filter_block1d16_h4_avx2() local
386 tmp_0 = _mm256_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d16_h4_avx2()
388 dst_first = _mm256_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d16_h4_avx2()
397 tmp_0 = _mm256_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d16_h4_avx2()
399 dst_second = _mm256_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d16_h4_avx2()
423 tmp_0 = _mm256_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d16_h4_avx2()
425 dst_first = _mm256_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d16_h4_avx2()
569 __m256i tmp_0, tmp_1; in vpx_filter_block1d8_h4_avx2() local
595 tmp_0 = _mm256_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d8_h4_avx2()
597 dst_reg = _mm256_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d8_h4_avx2()
[all …]
|
| D | vpx_subpixel_4t_intrin_sse2.c | 448 __m128i tmp_0, tmp_1; in vpx_filter_block1d4_h4_sse2() local
480 tmp_0 = _mm_unpacklo_epi32(src_reg, src_reg_shift_1); in vpx_filter_block1d4_h4_sse2()
484 tmp_0 = _mm_madd_epi16(tmp_0, kernel_reg_23); in vpx_filter_block1d4_h4_sse2()
488 dst_first = _mm_add_epi32(tmp_0, tmp_1); in vpx_filter_block1d4_h4_sse2()
800 __m128i tmp_0, tmp_1; in vpx_highbd_filter_block1d8_h4_sse2() local
825 tmp_0 = _mm_srli_si128(src_reg, 4); in vpx_highbd_filter_block1d8_h4_sse2()
827 src_reg_shift_2 = _mm_unpacklo_epi64(tmp_0, tmp_1); in vpx_highbd_filter_block1d8_h4_sse2()
832 tmp_0 = _mm_srli_si128(src_reg, 2); in vpx_highbd_filter_block1d8_h4_sse2()
834 src_reg_shift_1 = _mm_unpacklo_epi64(tmp_0, tmp_1); in vpx_highbd_filter_block1d8_h4_sse2()
836 tmp_0 = _mm_srli_si128(src_reg, 6); in vpx_highbd_filter_block1d8_h4_sse2()
[all …]
|
| D | vpx_subpixel_8t_intrin_ssse3.c | 226 __m128i tmp_0, tmp_1; in vpx_filter_block1d16_h4_ssse3() local
249 tmp_0 = _mm_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d16_h4_ssse3()
251 dst_first = _mm_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d16_h4_ssse3()
260 tmp_0 = _mm_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d16_h4_ssse3()
262 dst_second = _mm_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d16_h4_ssse3()
415 __m128i tmp_0, tmp_1; in vpx_filter_block1d8_h4_ssse3() local
438 tmp_0 = _mm_maddubs_epi16(src_reg_shift_0, kernel_reg_23); in vpx_filter_block1d8_h4_ssse3()
440 dst_first = _mm_adds_epi16(tmp_0, tmp_1); in vpx_filter_block1d8_h4_ssse3()
|
| D | convolve_avx2.h | 155 const __m256i tmp_0 = _mm256_madd_epi16(*src_0, *ker_0); in mm256_madd_add_epi32() local
157 return _mm256_add_epi32(tmp_0, tmp_1); in mm256_madd_add_epi32()
|
| /external/webp/src/dsp/ |
| D | enc_sse41.c | 76 __m128i tmp_0, tmp_1, tmp_2, tmp_3; in TTransform_SSE41() local
98 tmp_0 = _mm_cvtepu8_epi16(inAB_0); in TTransform_SSE41()
112 const __m128i a0 = _mm_add_epi16(tmp_0, tmp_2); in TTransform_SSE41()
115 const __m128i a3 = _mm_sub_epi16(tmp_0, tmp_2); in TTransform_SSE41()
126 VP8Transpose_2_4x4_16b(&b0, &b1, &b2, &b3, &tmp_0, &tmp_1, &tmp_2, &tmp_3); in TTransform_SSE41()
136 const __m128i a0 = _mm_add_epi16(tmp_0, tmp_2); in TTransform_SSE41()
139 const __m128i a3 = _mm_sub_epi16(tmp_0, tmp_2); in TTransform_SSE41()
|
| D | enc_sse2.c | 1100 __m128i tmp_0, tmp_1, tmp_2, tmp_3; in TTransform_SSE2() local
1119 tmp_0 = _mm_unpacklo_epi8(inAB_0, zero); in TTransform_SSE2()
1133 const __m128i a0 = _mm_add_epi16(tmp_0, tmp_2); in TTransform_SSE2()
1136 const __m128i a3 = _mm_sub_epi16(tmp_0, tmp_2); in TTransform_SSE2()
1147 VP8Transpose_2_4x4_16b(&b0, &b1, &b2, &b3, &tmp_0, &tmp_1, &tmp_2, &tmp_3); in TTransform_SSE2()
1157 const __m128i a0 = _mm_add_epi16(tmp_0, tmp_2); in TTransform_SSE2()
1160 const __m128i a3 = _mm_sub_epi16(tmp_0, tmp_2); in TTransform_SSE2()
|
| /external/libaom/libaom/av1/encoder/x86/ |
| D | corner_match_sse4.c | 82 __m128i tmp_0 = _mm_unpacklo_epi32(sum1_vec, sum2_vec); in compute_cross_correlation_sse4_1() local
87 __m128i tmp_4 = _mm_unpacklo_epi64(tmp_0, tmp_2); in compute_cross_correlation_sse4_1()
88 __m128i tmp_5 = _mm_unpackhi_epi64(tmp_0, tmp_2); in compute_cross_correlation_sse4_1()
|
| /external/libaom/libaom/av1/common/arm/ |
| D | warp_plane_neon.c | 237 uint8x16_t tmp_0 = vandq_u8(src_1, mask); in horizontal_filter_neon() local
242 tmp_2 = vextq_u8(tmp_0, tmp_0, 1); in horizontal_filter_neon()
245 src_1 = vaddq_u8(tmp_0, tmp_2); in horizontal_filter_neon()
|
| /external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
| D | depthwiseconv_uint8_3x3_filter.h | 3988 int8x16_t tmp_0 = vqtbl4q_s8(input_data, perm_data_0);
3990 work_reg_a = veorq_s8(tmp_0, sign_bit);
4000 tmp_0 = vqtbl4q_s8(input_data, perm_data_2);
4002 work_reg_a = veorq_s8(tmp_0, sign_bit);
4023 int8x16_t tmp_0 = vqtbl4q_s8(input_data, perm_data_0);
4025 work_reg_a = veorq_s8(tmp_0, sign_bit);
|
| D | depthwiseconv_uint8_transitional.h | 1271 int8x16_t tmp_0 = vqtbl4q_s8(input_data, perm_data_0);
1273 work_reg_a = veorq_s8(tmp_0, sign_bit);
1283 tmp_0 = vqtbl4q_s8(input_data, perm_data_2);
1285 work_reg_a = veorq_s8(tmp_0, sign_bit);
1306 int8x16_t tmp_0 = vqtbl4q_s8(input_data, perm_data_0);
1308 work_reg_a = veorq_s8(tmp_0, sign_bit);
|