/external/libaom/libaom/av1/encoder/x86/ |
D | highbd_fwd_txfm_avx2.c | 79 in[0] = _mm256_add_epi32(in[0], rounding); in col_txfm_8x8_rounding() 80 in[1] = _mm256_add_epi32(in[1], rounding); in col_txfm_8x8_rounding() 81 in[2] = _mm256_add_epi32(in[2], rounding); in col_txfm_8x8_rounding() 82 in[3] = _mm256_add_epi32(in[3], rounding); in col_txfm_8x8_rounding() 83 in[4] = _mm256_add_epi32(in[4], rounding); in col_txfm_8x8_rounding() 84 in[5] = _mm256_add_epi32(in[5], rounding); in col_txfm_8x8_rounding() 85 in[6] = _mm256_add_epi32(in[6], rounding); in col_txfm_8x8_rounding() 86 in[7] = _mm256_add_epi32(in[7], rounding); in col_txfm_8x8_rounding() 189 in[stride * i] = _mm256_add_epi32(in[stride * i], round); in av1_round_shift_32_8xn_avx2() 220 x = _mm256_add_epi32(x, y); in av1_half_btf_avx2() [all …]
|
D | av1_fwd_txfm_avx2.h | 27 __m256i temp0 = _mm256_add_epi32(in0_w0, in1_w1); in btf_32_avx2_type0() 28 temp0 = _mm256_add_epi32(temp0, _r); in btf_32_avx2_type0() 33 temp1 = _mm256_add_epi32(temp1, _r); in btf_32_avx2_type0() 46 __m256i temp0 = _mm256_add_epi32(in0_w0, in1_w1); in btf_32_avx2_type1() 47 temp0 = _mm256_add_epi32(temp0, _r); in btf_32_avx2_type1() 52 temp1 = _mm256_add_epi32(temp1, _r); in btf_32_avx2_type1() 66 __m256i temp0 = _mm256_add_epi32(in0_w0, in1_w1); in btf_32_avx2_type0_new() 67 temp0 = _mm256_add_epi32(temp0, _r); in btf_32_avx2_type0_new() 72 temp1 = _mm256_add_epi32(temp1, _r); in btf_32_avx2_type0_new() 86 __m256i temp0 = _mm256_add_epi32(in0_w0, in1_w1); in btf_32_avx2_type1_new() [all …]
|
D | corner_match_avx2.c | 54 sumsq2_vec = _mm256_add_epi32(sumsq2_vec, _mm256_madd_epi16(v2_1, v2_1)); in compute_cross_correlation_avx2() 57 cross_vec = _mm256_add_epi32(cross_vec, _mm256_madd_epi16(v1_1, v2_1)); in compute_cross_correlation_avx2() 62 sum_vec = _mm256_add_epi32(sum_vec, sum_vec1); in compute_cross_correlation_avx2() 68 temp1 = _mm256_add_epi32(unp_low, unp_hig); in compute_cross_correlation_avx2()
|
D | pickrst_avx2.c | 26 const __m256i r0 = _mm256_add_epi32(dst0, d0); in acc_stat_avx2() 527 _mm256_srai_epi32(_mm256_add_epi32(v0, rounding), shift); in av1_lowbd_pixel_proj_error_avx2() 529 _mm256_srai_epi32(_mm256_add_epi32(v1, rounding), shift); in av1_lowbd_pixel_proj_error_avx2() 533 sum32 = _mm256_add_epi32(sum32, err0); in av1_lowbd_pixel_proj_error_avx2() 572 _mm256_srai_epi32(_mm256_add_epi32(v0, rounding), shift); in av1_lowbd_pixel_proj_error_avx2() 574 _mm256_srai_epi32(_mm256_add_epi32(v1, rounding), shift); in av1_lowbd_pixel_proj_error_avx2() 578 sum32 = _mm256_add_epi32(sum32, err0); in av1_lowbd_pixel_proj_error_avx2() 604 sum32 = _mm256_add_epi32(sum32, err0); in av1_lowbd_pixel_proj_error_avx2() 676 const __m256i vl = _mm256_add_epi32(v0l, v1l); in av1_highbd_pixel_proj_error_avx2() 677 const __m256i vh = _mm256_add_epi32(v0h, v1h); in av1_highbd_pixel_proj_error_avx2() [all …]
|
/external/libvpx/libvpx/vpx_dsp/x86/ |
D | fwd_dct32x32_impl_avx2.h | 383 _mm256_add_epi32(s2_20_2, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 385 _mm256_add_epi32(s2_20_3, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 387 _mm256_add_epi32(s2_21_2, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 389 _mm256_add_epi32(s2_21_3, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 391 _mm256_add_epi32(s2_22_2, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 393 _mm256_add_epi32(s2_22_3, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 395 _mm256_add_epi32(s2_23_2, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 397 _mm256_add_epi32(s2_23_3, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 399 _mm256_add_epi32(s2_24_2, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() 401 _mm256_add_epi32(s2_24_3, k__DCT_CONST_ROUNDING); in FDCT32x32_2D_AVX2() [all …]
|
D | sad4d_avx2.c | 57 sums[0] = _mm256_add_epi32(sums[0], r[0]); in vpx_sad32x32x4d_avx2() 58 sums[1] = _mm256_add_epi32(sums[1], r[1]); in vpx_sad32x32x4d_avx2() 59 sums[2] = _mm256_add_epi32(sums[2], r[2]); in vpx_sad32x32x4d_avx2() 60 sums[3] = _mm256_add_epi32(sums[3], r[3]); in vpx_sad32x32x4d_avx2() 113 sums[0] = _mm256_add_epi32(sums[0], r_lo[0]); in vpx_sad64x64x4d_avx2() 114 sums[1] = _mm256_add_epi32(sums[1], r_lo[1]); in vpx_sad64x64x4d_avx2() 115 sums[2] = _mm256_add_epi32(sums[2], r_lo[2]); in vpx_sad64x64x4d_avx2() 116 sums[3] = _mm256_add_epi32(sums[3], r_lo[3]); in vpx_sad64x64x4d_avx2() 117 sums[0] = _mm256_add_epi32(sums[0], r_hi[0]); in vpx_sad64x64x4d_avx2() 118 sums[1] = _mm256_add_epi32(sums[1], r_hi[1]); in vpx_sad64x64x4d_avx2() [all …]
|
D | avg_intrin_avx2.c | 29 __m256i b0 = _mm256_add_epi32(a0, a1); in highbd_hadamard_col8_avx2() 31 __m256i b2 = _mm256_add_epi32(a2, a3); in highbd_hadamard_col8_avx2() 33 __m256i b4 = _mm256_add_epi32(a4, a5); in highbd_hadamard_col8_avx2() 35 __m256i b6 = _mm256_add_epi32(a6, a7); in highbd_hadamard_col8_avx2() 38 a0 = _mm256_add_epi32(b0, b2); in highbd_hadamard_col8_avx2() 39 a1 = _mm256_add_epi32(b1, b3); in highbd_hadamard_col8_avx2() 42 a4 = _mm256_add_epi32(b4, b6); in highbd_hadamard_col8_avx2() 43 a5 = _mm256_add_epi32(b5, b7); in highbd_hadamard_col8_avx2() 48 b0 = _mm256_add_epi32(a0, a4); in highbd_hadamard_col8_avx2() 49 b7 = _mm256_add_epi32(a1, a5); in highbd_hadamard_col8_avx2() [all …]
|
D | sad_avx2.c | 30 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 35 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 62 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 67 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 112 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 118 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 150 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 156 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \
|
/external/flac/libFLAC/ |
D | lpc_intrin_avx2.c | 81 …d_epi16(q10, _mm256_loadu_si256((const __m256i*)(data+i-11))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 82 …d_epi16(q9, _mm256_loadu_si256((const __m256i*)(data+i-10))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 83 …d_epi16(q8, _mm256_loadu_si256((const __m256i*)(data+i-9 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 84 …d_epi16(q7, _mm256_loadu_si256((const __m256i*)(data+i-8 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 85 …d_epi16(q6, _mm256_loadu_si256((const __m256i*)(data+i-7 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 86 …d_epi16(q5, _mm256_loadu_si256((const __m256i*)(data+i-6 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 87 …d_epi16(q4, _mm256_loadu_si256((const __m256i*)(data+i-5 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 88 …d_epi16(q3, _mm256_loadu_si256((const __m256i*)(data+i-4 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 89 …d_epi16(q2, _mm256_loadu_si256((const __m256i*)(data+i-3 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 90 …d_epi16(q1, _mm256_loadu_si256((const __m256i*)(data+i-2 ))); summ = _mm256_add_epi32(summ, mull); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() [all …]
|
/external/libaom/libaom/av1/common/x86/ |
D | selfguided_avx2.c | 53 const __m256i x02 = _mm256_add_epi32(x, x01); in scan_32() 55 const __m256i x04 = _mm256_add_epi32(x02, x03); in scan_32() 59 return _mm256_add_epi32(x04, s02); in scan_32() 114 _mm256_add_epi32(_mm256_add_epi32(sc1, above1), ldiff1); in integral_images() 116 _mm256_add_epi32(_mm256_add_epi32(sc2, above2), ldiff2); in integral_images() 162 _mm256_add_epi32(_mm256_add_epi32(sc1, above1), ldiff1); in integral_images_highbd() 164 _mm256_add_epi32(_mm256_add_epi32(sc2, above2), ldiff2); in integral_images_highbd() 202 _mm256_srl_epi32(_mm256_add_epi32(sum2, rounding_a), shift_a); in compute_p() 204 _mm256_srl_epi32(_mm256_add_epi32(sum1, rounding_b), shift_b); in compute_p() 262 _mm256_srli_epi32(_mm256_add_epi32(_mm256_mullo_epi32(p, s), rnd_z), in calc_ab() [all …]
|
D | highbd_wiener_convolve_avx2.c | 116 const __m256i res_even_sum = _mm256_add_epi32( in av1_highbd_wiener_convolve_add_src_avx2() 117 _mm256_add_epi32(res_0, res_4), _mm256_add_epi32(res_2, res_6)); in av1_highbd_wiener_convolve_add_src_avx2() 119 _mm256_add_epi32(res_even_sum, round_const), conv_params->round_0); in av1_highbd_wiener_convolve_add_src_avx2() 121 const __m256i res_odd_sum = _mm256_add_epi32( in av1_highbd_wiener_convolve_add_src_avx2() 122 _mm256_add_epi32(res_1, res_5), _mm256_add_epi32(res_3, res_7)); in av1_highbd_wiener_convolve_add_src_avx2() 124 _mm256_add_epi32(res_odd_sum, round_const), conv_params->round_0); in av1_highbd_wiener_convolve_add_src_avx2() 201 const __m256i res_even = _mm256_add_epi32( in av1_highbd_wiener_convolve_add_src_avx2() 202 _mm256_add_epi32(res_0, res_2), _mm256_add_epi32(res_4, res_6)); in av1_highbd_wiener_convolve_add_src_avx2() 215 const __m256i res_odd = _mm256_add_epi32( in av1_highbd_wiener_convolve_add_src_avx2() 216 _mm256_add_epi32(res_1, res_3), _mm256_add_epi32(res_5, res_7)); in av1_highbd_wiener_convolve_add_src_avx2() [all …]
|
D | wiener_convolve_avx2.c | 124 const __m256i res_even_sum = _mm256_add_epi32( in av1_wiener_convolve_add_src_avx2() 125 _mm256_add_epi32(res_0, res_4), _mm256_add_epi32(res_2, res_6)); in av1_wiener_convolve_add_src_avx2() 126 const __m256i res_odd_sum = _mm256_add_epi32( in av1_wiener_convolve_add_src_avx2() 127 _mm256_add_epi32(res_1, res_5), _mm256_add_epi32(res_3, res_7)); in av1_wiener_convolve_add_src_avx2() 130 _mm256_add_epi32(res_even_sum, round_const), conv_params->round_0); in av1_wiener_convolve_add_src_avx2() 132 _mm256_add_epi32(res_odd_sum, round_const), conv_params->round_0); in av1_wiener_convolve_add_src_avx2() 207 const __m256i res_even = _mm256_add_epi32( in av1_wiener_convolve_add_src_avx2() 208 _mm256_add_epi32(res_0, res_2), _mm256_add_epi32(res_4, res_6)); in av1_wiener_convolve_add_src_avx2() 221 const __m256i res_odd = _mm256_add_epi32( in av1_wiener_convolve_add_src_avx2() 222 _mm256_add_epi32(res_1, res_3), _mm256_add_epi32(res_5, res_7)); in av1_wiener_convolve_add_src_avx2() [all …]
|
D | highbd_inv_txfm_avx2.c | 56 x0 = _mm256_add_epi32(res0, x0); in highbd_get_recon_16x8_avx2() 57 x1 = _mm256_add_epi32(res1, x1); in highbd_get_recon_16x8_avx2() 79 x0 = _mm256_add_epi32(res, x0); in highbd_get_recon_8x8_avx2() 104 __m256i a0 = _mm256_add_epi32(offset, in0); in neg_shift_avx2() 205 x = _mm256_add_epi32(x, *rounding); in half_btf_0_avx2() 217 x = _mm256_add_epi32(x, y); in half_btf_avx2() 218 x = _mm256_add_epi32(x, *rounding); in half_btf_avx2() 226 __m256i a0 = _mm256_add_epi32(in0, in1); in addsub_avx2() 240 __m256i a0 = _mm256_add_epi32(in0, in1); in addsub_no_clamp_avx2() 252 __m256i in0_w_offset = _mm256_add_epi32(in0, offset); in addsub_shift_avx2() [all …]
|
D | highbd_jnt_convolve_avx2.c | 79 _mm256_add_epi32(res_32b_lo, offset_const); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 87 _mm256_add_epi32(res_32b_hi, offset_const); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 138 _mm256_add_epi32(res_32b, offset_const); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 183 _mm256_add_epi32(res_32b_lo, offset_const); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 191 _mm256_add_epi32(res_32b_hi, offset_const); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 304 res_even = _mm256_sra_epi32(_mm256_add_epi32(res_even, round_const_x), in av1_highbd_dist_wtd_convolve_2d_avx2() 314 res_odd = _mm256_sra_epi32(_mm256_add_epi32(res_odd, round_const_x), in av1_highbd_dist_wtd_convolve_2d_avx2() 356 _mm256_add_epi32(res_a, round_const_y), round_shift_y); in av1_highbd_dist_wtd_convolve_2d_avx2() 359 _mm256_add_epi32(res_a_round, offset_const); in av1_highbd_dist_wtd_convolve_2d_avx2() 403 _mm256_add_epi32(res_b, round_const_y), round_shift_y); in av1_highbd_dist_wtd_convolve_2d_avx2() [all …]
|
/external/libaom/libaom/aom_dsp/x86/ |
D | sad4d_avx2.c | 47 sum_ref0 = _mm256_add_epi32(sum_ref0, ref0_reg); in aom_sad32x32x4d_avx2() 48 sum_ref1 = _mm256_add_epi32(sum_ref1, ref1_reg); in aom_sad32x32x4d_avx2() 49 sum_ref2 = _mm256_add_epi32(sum_ref2, ref2_reg); in aom_sad32x32x4d_avx2() 50 sum_ref3 = _mm256_add_epi32(sum_ref3, ref3_reg); in aom_sad32x32x4d_avx2() 75 sum_mlow = _mm256_add_epi32(sum_mlow, sum_mhigh); in aom_sad32x32x4d_avx2() 128 sum_ref0 = _mm256_add_epi32(sum_ref0, ref0_reg); in aom_sad64x64x4d_avx2() 129 sum_ref1 = _mm256_add_epi32(sum_ref1, ref1_reg); in aom_sad64x64x4d_avx2() 130 sum_ref2 = _mm256_add_epi32(sum_ref2, ref2_reg); in aom_sad64x64x4d_avx2() 131 sum_ref3 = _mm256_add_epi32(sum_ref3, ref3_reg); in aom_sad64x64x4d_avx2() 132 sum_ref0 = _mm256_add_epi32(sum_ref0, ref0next_reg); in aom_sad64x64x4d_avx2() [all …]
|
D | sad_avx2.c | 33 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 38 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 66 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 71 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 119 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 125 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \ 158 _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); \ 164 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); \
|
D | obmc_variance_avx2.c | 58 _mm256_add_epi32(_mm256_add_epi32(v_diff0_d, v_bias_d), v_sign_d); in obmc_variance_w8n() 123 _mm256_add_epi32(_mm256_add_epi32(v_diff0_d, v_bias_d), v_sign0_d); in obmc_variance_w16n() 125 _mm256_add_epi32(_mm256_add_epi32(v_diff1_d, v_bias_d), v_sign1_d); in obmc_variance_w16n() 130 const __m256i v_rdiff1_d = _mm256_add_epi32(v_rdiff0_d, v_rdiff2_d); in obmc_variance_w16n() 134 v_sum_d = _mm256_add_epi32(v_sum_d, v_rdiff1_d); in obmc_variance_w16n() 135 v_sse_d = _mm256_add_epi32(v_sse_d, v_sqrdiff_d); in obmc_variance_w16n()
|
D | obmc_sad_avx2.c | 55 const __m256i v_tmp_d = _mm256_add_epi32(v_absdiff_d, v_bias_d); in obmc_sad_w4_avx2() 58 v_sad_d = _mm256_add_epi32(v_sad_d, v_rad_d); in obmc_sad_w4_avx2() 96 const __m256i v_tmp_d = _mm256_add_epi32(v_absdiff0_d, v_bias_d); in obmc_sad_w8n_avx2() 99 v_sad_d = _mm256_add_epi32(v_sad_d, v_rad0_d); in obmc_sad_w8n_avx2() 178 const __m256i v_tmp_d = _mm256_add_epi32(v_absdiff_d, v_bias_d); in hbd_obmc_sad_w4_avx2() 181 v_sad_d = _mm256_add_epi32(v_sad_d, v_rad_d); in hbd_obmc_sad_w4_avx2() 222 const __m256i v_tmp_d = _mm256_add_epi32(v_absdiff0_d, v_bias_d); in hbd_obmc_sad_w8n_avx2() 225 v_sad_d = _mm256_add_epi32(v_sad_d, v_rad0_d); in hbd_obmc_sad_w8n_avx2()
|
D | convolve_avx2.h | 92 _mm256_sra_epi32(_mm256_add_epi32(res_a, sum_round_v), sum_shift_v); \ 94 _mm256_sra_epi32(_mm256_add_epi32(res_b, sum_round_v), sum_shift_v); \ 97 _mm256_add_epi32(res_a, round_const_v), round_shift_v); \ 99 _mm256_add_epi32(res_b, round_const_v), round_shift_v); \ 172 _mm256_add_epi32(res_a, round_const_v), round_shift_v); \ 177 _mm256_add_epi32(res_b, round_const_v), round_shift_v); \ 324 const __m256i res = _mm256_add_epi32(_mm256_add_epi32(res_0, res_1), in convolve() 325 _mm256_add_epi32(res_2, res_3)); in convolve() 335 const __m256i res = _mm256_add_epi32(res_1, res_2); in convolve_4tap() 369 d = _mm256_add_epi32(d, *res); in add_store_aligned_256() [all …]
|
D | sum_squares_avx2.c | 43 const __m256i v_sum_01_d = _mm256_add_epi32(v_sq_0_d, v_sq_1_d); in aom_sum_squares_2d_i16_nxn_avx2() 44 const __m256i v_sum_23_d = _mm256_add_epi32(v_sq_2_d, v_sq_3_d); in aom_sum_squares_2d_i16_nxn_avx2() 45 const __m256i v_sum_0123_d = _mm256_add_epi32(v_sum_01_d, v_sum_23_d); in aom_sum_squares_2d_i16_nxn_avx2() 47 v_acc_d = _mm256_add_epi32(v_acc_d, v_sum_0123_d); in aom_sum_squares_2d_i16_nxn_avx2()
|
D | sad_highbd_avx2.c | 26 u = _mm256_add_epi32(u, *v); in get_sad_from_mm256_epi32() 104 s0 = _mm256_add_epi32(s0, s1); in aom_highbd_sad16x8_avx2() 105 r0 = _mm256_add_epi32(r0, r1); in aom_highbd_sad16x8_avx2() 106 sum0 = _mm256_add_epi32(s0, r0); in aom_highbd_sad16x8_avx2() 150 sum = _mm256_add_epi32(sum, s0); in aom_highbd_sad16x16_avx2() 151 sum = _mm256_add_epi32(sum, s1); in aom_highbd_sad16x16_avx2() 205 r0 = _mm256_add_epi32(r0, r1); in sad32x4() 206 *sad_acc = _mm256_add_epi32(*sad_acc, r0); in sad32x4() 332 r[0] = _mm256_add_epi32(r[0], r[1]); in sad64x2() 333 r[0] = _mm256_add_epi32(r[0], r[2]); in sad64x2() [all …]
|
D | highbd_variance_avx2.c | 38 v_sse_d = _mm256_add_epi32(v_sse_d, v_sqrdiff); in aom_highbd_calc8x8var_avx2() 44 __m256i v_sum0 = _mm256_add_epi32(v_sum00, v_sum01); in aom_highbd_calc8x8var_avx2() 47 __m256i v_d_lh = _mm256_add_epi32(v_d_l, v_d_h); in aom_highbd_calc8x8var_avx2() 68 v_sse_d = _mm256_add_epi32(v_sse_d, v_sqrdiff); in aom_highbd_calc16x16var_avx2() 75 __m256i v_d_lh = _mm256_add_epi32(v_d_l, v_d_h); in aom_highbd_calc16x16var_avx2()
|
D | sse_avx2.c | 31 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d00_w, v_d00_w)); in sse_w32_avx2() 32 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d01_w, v_d01_w)); in sse_w32_avx2() 84 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d_w, v_d_w)); in sse_w4x4_avx2() 95 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d_w, v_d_w)); in sse_w8x2_avx2() 139 _mm256_add_epi32(_mm256_madd_epi16(v_asub, v_asub), in aom_sse_avx2() 141 sum = _mm256_add_epi32(sum, temp); in aom_sse_avx2() 219 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d_w, v_d_w)); in highbd_sse_w16_avx2() 238 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d_w, v_d_w)); in highbd_sse_w4x4_avx2() 247 *sum = _mm256_add_epi32(*sum, _mm256_madd_epi16(v_d_w, v_d_w)); in highbd_sse_w8x2_avx2()
|
D | sad_impl_avx2.c | 29 sum = _mm256_add_epi32(sum, _mm256_add_epi32(s1, s2)); in sad32x32() 34 sum = _mm256_add_epi32(sum, _mm256_srli_si256(sum, 8)); in sad32x32() 182 sum_sad = _mm256_add_epi32(sum_sad, _mm256_add_epi32(sad1_reg, sad2_reg)); in sad_w64_avg_avx2() 188 sum_sad = _mm256_add_epi32(sum_sad, sum_sad_h); in sad_w64_avg_avx2()
|
/external/libvpx/libvpx/vp9/encoder/x86/ |
D | vp9_error_avx2.c | 44 dqcoeff_256 = _mm256_add_epi32(dqcoeff_256, dqcoeff_hi); in vp9_block_error_avx2() 45 coeff_256 = _mm256_add_epi32(coeff_256, coeff_hi); in vp9_block_error_avx2() 72 dqcoeff_0 = _mm256_add_epi32(dqcoeff_0, dqcoeff_1); in vp9_block_error_avx2() 74 coeff_0 = _mm256_add_epi32(coeff_0, coeff_1); in vp9_block_error_avx2() 129 const __m256i error = _mm256_add_epi32(error_lo, error_hi); in vp9_block_error_fp_avx2()
|