/external/libaom/libaom/av1/encoder/x86/ |
D | encodetxb_sse4.c | 39 const __m128i coeffA = xx_loadu_128(cf); in av1_txb_init_levels_sse4_1() 40 const __m128i coeffB = xx_loadu_128(cf + 4); in av1_txb_init_levels_sse4_1() 52 const __m128i coeffA = xx_loadu_128(cf); in av1_txb_init_levels_sse4_1() 53 const __m128i coeffB = xx_loadu_128(cf + 4); in av1_txb_init_levels_sse4_1() 66 const __m128i coeffA = xx_loadu_128(cf); in av1_txb_init_levels_sse4_1() 67 const __m128i coeffB = xx_loadu_128(cf + 4); in av1_txb_init_levels_sse4_1() 68 const __m128i coeffC = xx_loadu_128(cf + 8); in av1_txb_init_levels_sse4_1() 69 const __m128i coeffD = xx_loadu_128(cf + 12); in av1_txb_init_levels_sse4_1()
|
D | pickrst_sse4.c | 22 const __m128i s = _mm_shuffle_epi8(xx_loadu_128(src), *shuffle); in acc_stat_sse41() 26 const __m128i dst0 = xx_loadu_128(dst); in acc_stat_sse41() 27 const __m128i dst1 = xx_loadu_128(dst + 4); in acc_stat_sse41() 88 const __m128i shuffle = xx_loadu_128(g_shuffle_stats_data); in compute_stats_win7_opt_sse4_1() 133 const __m128i s0l = xx_loadu_128(dgd); in acc_stat_highbd_sse41() 134 const __m128i s0h = xx_loadu_128(dgd + 4); in acc_stat_highbd_sse41() 159 const __m128i rll = _mm_add_epi64(xx_loadu_128(dst), dll); in acc_stat_highbd_sse41() 161 const __m128i rlh = _mm_add_epi64(xx_loadu_128(dst + 2), dlh); in acc_stat_highbd_sse41() 163 const __m128i rhl = _mm_add_epi64(xx_loadu_128(dst + 4), dhl); in acc_stat_highbd_sse41() 165 const __m128i rhh = _mm_add_epi64(xx_loadu_128(dst + 6), dhh); in acc_stat_highbd_sse41() [all …]
|
D | pickrst_avx2.c | 23 const __m128i s = _mm_shuffle_epi8(xx_loadu_128(src), *shuffle); in acc_stat_avx2() 85 const __m128i shuffle = xx_loadu_128(g_shuffle_stats_data); in compute_stats_win7_opt_avx2() 440 const __m128i shuffle = xx_loadu_128(g_shuffle_stats_data); in compute_stats_win5_opt_avx2() 517 const __m256i d0 = _mm256_cvtepu8_epi16(xx_loadu_128(dat + j)); in av1_lowbd_pixel_proj_error_avx2() 518 const __m256i s0 = _mm256_cvtepu8_epi16(xx_loadu_128(src + j)); in av1_lowbd_pixel_proj_error_avx2() 569 const __m256i d0 = _mm256_cvtepu8_epi16(xx_loadu_128(dat + j)); in av1_lowbd_pixel_proj_error_avx2() 570 const __m256i s0 = _mm256_cvtepu8_epi16(xx_loadu_128(src + j)); in av1_lowbd_pixel_proj_error_avx2() 608 const __m256i d0 = _mm256_cvtepu8_epi16(xx_loadu_128(dat + j)); in av1_lowbd_pixel_proj_error_avx2() 609 const __m256i s0 = _mm256_cvtepu8_epi16(xx_loadu_128(src + j)); in av1_lowbd_pixel_proj_error_avx2()
|
/external/libaom/libaom/av1/common/x86/ |
D | av1_convolve_horiz_rs_sse4.c | 63 const __m128i fil0_16 = xx_loadu_128(x_filter0); in av1_convolve_horiz_rs_sse4_1() 64 const __m128i fil1_16 = xx_loadu_128(x_filter1); in av1_convolve_horiz_rs_sse4_1() 65 const __m128i fil2_16 = xx_loadu_128(x_filter2); in av1_convolve_horiz_rs_sse4_1() 66 const __m128i fil3_16 = xx_loadu_128(x_filter3); in av1_convolve_horiz_rs_sse4_1() 173 const __m128i fil0_16 = xx_loadu_128(x_filter0); in av1_highbd_convolve_horiz_rs_sse4_1() 174 const __m128i fil1_16 = xx_loadu_128(x_filter1); in av1_highbd_convolve_horiz_rs_sse4_1() 175 const __m128i fil2_16 = xx_loadu_128(x_filter2); in av1_highbd_convolve_horiz_rs_sse4_1() 176 const __m128i fil3_16 = xx_loadu_128(x_filter3); in av1_highbd_convolve_horiz_rs_sse4_1() 192 const __m128i src0_16 = xx_loadu_128(src_x0); in av1_highbd_convolve_horiz_rs_sse4_1() 193 const __m128i src1_16 = xx_loadu_128(src_x1); in av1_highbd_convolve_horiz_rs_sse4_1() [all …]
|
D | selfguided_sse4.c | 134 const __m128i tl = xx_loadu_128(ii - (r + 1) - (r + 1) * stride); in boxsum_from_ii() 135 const __m128i tr = xx_loadu_128(ii + (r + 0) - (r + 1) * stride); in boxsum_from_ii() 136 const __m128i bl = xx_loadu_128(ii - (r + 1) + r * stride); in boxsum_from_ii() 137 const __m128i br = xx_loadu_128(ii + (r + 0) + r * stride); in boxsum_from_ii() 260 const __m128i xtl = xx_loadu_128(buf - 1 - stride); in cross_sum() 261 const __m128i xt = xx_loadu_128(buf - stride); in cross_sum() 262 const __m128i xtr = xx_loadu_128(buf + 1 - stride); in cross_sum() 263 const __m128i xl = xx_loadu_128(buf - 1); in cross_sum() 264 const __m128i x = xx_loadu_128(buf); in cross_sum() 265 const __m128i xr = xx_loadu_128(buf + 1); in cross_sum() [all …]
|
D | selfguided_avx2.c | 30 return _mm256_cvtepu16_epi32(xx_loadu_128(p)); in yy256_load_extend_16_32() 339 xx_loadu_128(dgd_real + ((i * dgd_stride + j) << highbd)); in final_filter() 517 xx_loadu_128(dgd_real + ((i * dgd_stride + j) << highbd)); in final_filter_fast() 534 xx_loadu_128(dgd_real + ((i * dgd_stride + j) << highbd)); in final_filter_fast() 662 src_0 = xx_loadu_128(CONVERT_TO_SHORTPTR(dat8ij)); in av1_apply_selfguided_restoration_avx2() 663 src_1 = xx_loadu_128(CONVERT_TO_SHORTPTR(dat8ij + 8)); in av1_apply_selfguided_restoration_avx2() 667 src_0 = xx_loadu_128(dat8ij); in av1_apply_selfguided_restoration_avx2()
|
D | highbd_wiener_convolve_avx2.c | 62 const __m128i coeffs_x = _mm_add_epi16(xx_loadu_128(filter_x), offset); in av1_highbd_wiener_convolve_add_src_avx2() 145 const __m128i coeffs_y = _mm_add_epi16(xx_loadu_128(filter_y), offset); in av1_highbd_wiener_convolve_add_src_avx2()
|
D | wiener_convolve_avx2.c | 102 const __m128i coeffs_y = _mm_add_epi16(xx_loadu_128(filter_y), offset_0); in av1_wiener_convolve_add_src_avx2()
|
/external/libaom/libaom/aom_dsp/x86/ |
D | blend_a64_mask_sse4.c | 83 const __m128i v_m0_b = xx_loadu_128(mask + c); in blend_a64_mask_w16n_sse4_1() 108 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_w4_sse4_1() 135 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_w8_sse4_1() 139 const __m128i v_r_b = xx_loadu_128(mask); in blend_a64_mask_sx_w8_sse4_1() 161 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_w16n_sse4_1() 168 const __m128i v_r0_b = xx_loadu_128(mask + 2 * c); in blend_a64_mask_sx_w16n_sse4_1() 169 const __m128i v_r1_b = xx_loadu_128(mask + 2 * c + 16); in blend_a64_mask_sx_w16n_sse4_1() 252 const __m128i v_ra_b = xx_loadu_128(mask + c); in blend_a64_mask_sy_w16n_sse4_1() 253 const __m128i v_rb_b = xx_loadu_128(mask + c + mask_stride); in blend_a64_mask_sy_w16n_sse4_1() 277 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_sy_w4_sse4_1() [all …]
|
D | jnt_sad_ssse3.c | 101 __m128i x = xx_loadu_128(a); in aom_sad16xh_sse2() 102 __m128i y = xx_loadu_128(b); in aom_sad16xh_sse2() 126 __m128i x = xx_loadu_128(a + j * 16); in aom_sad32xh_sse2() 127 __m128i y = xx_loadu_128(b + j * 16); in aom_sad32xh_sse2() 152 __m128i x = xx_loadu_128(a + j * 16); in aom_sad64xh_sse2() 153 __m128i y = xx_loadu_128(b + j * 16); in aom_sad64xh_sse2() 178 __m128i x = xx_loadu_128(a + j * 16); in aom_sad128xh_sse2() 179 __m128i y = xx_loadu_128(b + j * 16); in aom_sad128xh_sse2()
|
D | jnt_variance_ssse3.c | 67 __m128i p0 = xx_loadu_128(ref); in aom_dist_wtd_comp_avg_pred_ssse3() 68 __m128i p1 = xx_loadu_128(pred); in aom_dist_wtd_comp_avg_pred_ssse3() 86 __m128i p1 = xx_loadu_128(pred); in aom_dist_wtd_comp_avg_pred_ssse3() 108 __m128i p1 = xx_loadu_128(pred); in aom_dist_wtd_comp_avg_pred_ssse3() 141 __m128i p0 = xx_loadu_128(comp_pred); in aom_dist_wtd_comp_avg_upsampled_pred_ssse3() 142 __m128i p1 = xx_loadu_128(pred); in aom_dist_wtd_comp_avg_upsampled_pred_ssse3()
|
D | blend_sse4.h | 89 const __m128i v_s0_b = xx_loadu_128(src0); in blend_16_u8() 90 const __m128i v_s1_b = xx_loadu_128(src1); in blend_16_u8() 123 const __m128i v_s0_w = xx_loadu_128(src0); in blend_8_b10() 124 const __m128i v_s1_w = xx_loadu_128(src1); in blend_8_b10() 163 const __m128i v_s0_w = xx_loadu_128(src0); in blend_8_b12() 164 const __m128i v_s1_w = xx_loadu_128(src1); in blend_8_b12()
|
D | blend_mask_sse4.h | 50 const __m128i s0 = xx_loadu_128(src0); in blend_a64_d16_mask_w8_sse41() 51 const __m128i s1 = xx_loadu_128(src1); in blend_a64_d16_mask_w8_sse41() 135 const __m128i m_i0 = xx_loadu_128(mask); in aom_lowbd_blend_a64_d16_mask_subw1_subh1_w8_sse4_1() 136 const __m128i m_i1 = xx_loadu_128(mask + mask_stride); in aom_lowbd_blend_a64_d16_mask_subw1_subh1_w8_sse4_1() 182 const __m128i m_i0 = xx_loadu_128(mask); in aom_lowbd_blend_a64_d16_mask_subw1_subh0_w8_sse4_1()
|
D | sse_sse4.c | 42 const __m128i v_a0 = xx_loadu_128(a); in sse_w16_sse4_1() 43 const __m128i v_b0 = xx_loadu_128(b); in sse_w16_sse4_1() 195 const __m128i v_a_w = xx_loadu_128(a); in highbd_sse_w8_sse4_1() 196 const __m128i v_b_w = xx_loadu_128(b); in highbd_sse_w8_sse4_1()
|
D | blend_a64_mask_avx2.c | 89 const __m128i m = xx_loadu_128(mask); in lowbd_blend_a64_d16_mask_subw0_subh0_w16_avx2() 240 const __m128i m_i00 = xx_loadu_128(mask + j); in lowbd_blend_a64_d16_mask_subw0_subh1_w16_avx2() 241 const __m128i m_i10 = xx_loadu_128(mask + mask_stride + j); in lowbd_blend_a64_d16_mask_subw0_subh1_w16_avx2() 408 const __m256i v_s0_b = _mm256_castsi128_si256(xx_loadu_128(src0)); in blend_16_u8_avx2() 409 const __m256i v_s1_b = _mm256_castsi128_si256(xx_loadu_128(src1)); in blend_16_u8_avx2() 519 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_sy_avx2() 548 const __m128i v_ra_b = xx_loadu_128(mask); in blend_a64_mask_sx_sy_avx2() 549 const __m128i v_rb_b = xx_loadu_128(mask + mask_stride); in blend_a64_mask_sx_sy_avx2() 644 const __m128i v_shuffle_b = xx_loadu_128(g_blend_a64_mask_shuffle); in blend_a64_mask_sx_avx2() 669 const __m128i v_r_b = xx_loadu_128(mask); in blend_a64_mask_sx_avx2() [all …]
|
D | sse_avx2.c | 126 const __m128i v_a0 = xx_loadu_128(a); in aom_sse_avx2() 127 const __m128i v_a1 = xx_loadu_128(a + a_stride); in aom_sse_avx2() 128 const __m128i v_b0 = xx_loadu_128(b); in aom_sse_avx2() 129 const __m128i v_b1 = xx_loadu_128(b + b_stride); in aom_sse_avx2()
|
D | synonyms.h | 45 static INLINE __m128i xx_loadu_128(const void *a) { in xx_loadu_128() function
|
D | variance_sse2.c | 34 const __m128i v = xx_loadu_128(src); in aom_get_mb_ss_sse2() 546 xx_storeu_128(comp_pred, xx_loadu_128(ref)); in aom_upsampled_pred_sse2() 626 __m128i s0 = xx_loadu_128(comp_pred); in aom_comp_avg_upsampled_pred_sse2() 627 __m128i p0 = xx_loadu_128(pred); in aom_comp_avg_upsampled_pred_sse2()
|
D | highbd_variance_sse2.c | 780 __m128i p0 = xx_loadu_128(ref); in aom_highbd_dist_wtd_comp_avg_pred_sse2() 781 __m128i p1 = xx_loadu_128(pred); in aom_highbd_dist_wtd_comp_avg_pred_sse2() 798 __m128i p1 = xx_loadu_128(pred); in aom_highbd_dist_wtd_comp_avg_pred_sse2() 834 __m128i p0 = xx_loadu_128(comp_pred16); in aom_highbd_dist_wtd_comp_avg_upsampled_pred_sse2() 835 __m128i p1 = xx_loadu_128(pred); in aom_highbd_dist_wtd_comp_avg_upsampled_pred_sse2()
|