/external/libaom/libaom/av1/common/x86/ |
D | highbd_inv_txfm_avx2.c | 103 __m256i offset = _mm256_set1_epi32((1 << shift) >> 1); in neg_shift_avx2() 251 __m256i offset = _mm256_set1_epi32((1 << shift) >> 1); in addsub_shift_avx2() 423 const __m256i clamp_lo_out = _mm256_set1_epi32(AOMMAX( in idct32_stage9_avx2() 425 const __m256i clamp_hi_out = _mm256_set1_epi32(AOMMIN( in idct32_stage9_avx2() 466 const __m256i cospi32 = _mm256_set1_epi32(cospi[32]); in idct32_low1_avx2() 467 const __m256i rounding = _mm256_set1_epi32(1 << (bit - 1)); in idct32_low1_avx2() 469 const __m256i clamp_lo = _mm256_set1_epi32(-(1 << (log_range - 1))); in idct32_low1_avx2() 470 const __m256i clamp_hi = _mm256_set1_epi32((1 << (log_range - 1)) - 1); in idct32_low1_avx2() 491 const __m256i clamp_lo_out = _mm256_set1_epi32(AOMMAX( in idct32_low1_avx2() 493 const __m256i clamp_hi_out = _mm256_set1_epi32(AOMMIN( in idct32_low1_avx2() [all …]
|
D | selfguided_avx2.c | 122 ldiff1 = _mm256_set1_epi32( in integral_images() 124 ldiff2 = _mm256_set1_epi32( in integral_images() 170 ldiff1 = _mm256_set1_epi32( in integral_images_highbd() 172 ldiff2 = _mm256_set1_epi32( in integral_images_highbd() 191 return _mm256_set1_epi32((1 << shift) >> 1); in round_for_shift() 208 an = _mm256_max_epi32(_mm256_mullo_epi32(a, _mm256_set1_epi32(n)), bb); in compute_p() 211 an = _mm256_mullo_epi32(sum2, _mm256_set1_epi32(n)); in compute_p() 225 const __m256i s = _mm256_set1_epi32(params->s[radius_idx]); in calc_ab() 227 const __m256i one_over_n = _mm256_set1_epi32(one_by_x[n - 1]); in calc_ab() 264 _mm256_set1_epi32(255)); in calc_ab() [all …]
|
D | highbd_jnt_convolve_avx2.c | 44 const __m256i wt0 = _mm256_set1_epi32(w0); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 45 const __m256i wt1 = _mm256_set1_epi32(w1); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 52 const __m256i offset_const = _mm256_set1_epi32(offset); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 56 const __m256i rounding_const = _mm256_set1_epi32((1 << rounding_shift) >> 1); in av1_highbd_dist_wtd_convolve_2d_copy_avx2() 256 const __m256i wt0 = _mm256_set1_epi32(w0); in av1_highbd_dist_wtd_convolve_2d_avx2() 257 const __m256i wt1 = _mm256_set1_epi32(w1); in av1_highbd_dist_wtd_convolve_2d_avx2() 260 const __m256i round_const_x = _mm256_set1_epi32( in av1_highbd_dist_wtd_convolve_2d_avx2() 264 const __m256i round_const_y = _mm256_set1_epi32( in av1_highbd_dist_wtd_convolve_2d_avx2() 272 const __m256i offset_const = _mm256_set1_epi32(offset); in av1_highbd_dist_wtd_convolve_2d_avx2() 275 const __m256i rounding_const = _mm256_set1_epi32((1 << rounding_shift) >> 1); in av1_highbd_dist_wtd_convolve_2d_avx2() [all …]
|
D | highbd_convolve_2d_avx2.c | 44 const __m256i round_const_x = _mm256_set1_epi32( in av1_highbd_convolve_2d_sr_avx2() 48 const __m256i round_const_y = _mm256_set1_epi32( in av1_highbd_convolve_2d_sr_avx2() 56 const __m256i round_const_bits = _mm256_set1_epi32((1 << bits) >> 1); in av1_highbd_convolve_2d_sr_avx2()
|
D | highbd_wiener_convolve_avx2.c | 87 const __m256i round_const = _mm256_set1_epi32( in av1_highbd_wiener_convolve_add_src_avx2() 171 _mm256_set1_epi32((1 << (conv_params->round_1 - 1)) - in av1_highbd_wiener_convolve_add_src_avx2()
|
D | wiener_convolve_avx2.c | 85 const __m256i round_const = _mm256_set1_epi32( in av1_wiener_convolve_add_src_avx2() 177 _mm256_set1_epi32((1 << (conv_params->round_1 - 1)) - in av1_wiener_convolve_add_src_avx2()
|
D | convolve_2d_avx2.c | 43 const __m256i sum_round_v = _mm256_set1_epi32( in av1_convolve_2d_sr_avx2() 47 const __m256i round_const_v = _mm256_set1_epi32( in av1_convolve_2d_sr_avx2()
|
D | av1_inv_txfm_avx2.c | 68 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct16_new_avx2() 140 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct16_low8_new_avx2() 311 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in iadst16_new_avx2() 372 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in iadst16_low8_new_avx2() 408 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in iadst16_low1_new_avx2() 628 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct32_low8_new_avx2() 686 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct32_low16_new_avx2() 753 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct32_new_avx2() 1198 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct64_low8_new_avx2() 1319 const __m256i _r = _mm256_set1_epi32(1 << (INV_COS_BIT - 1)); in idct64_low16_new_avx2() [all …]
|
D | cfl_avx2.c | 437 _mm256_add_epi32(fill, _mm256_set1_epi32(round_offset)), num_pel_log2); in subtract_average_avx2()
|
D | jnt_convolve_avx2.c | 203 _mm256_set1_epi32((1 << conv_params->round_1) >> 1); in av1_dist_wtd_convolve_y_avx2() 626 const __m256i round_const_v = _mm256_set1_epi32( in av1_dist_wtd_convolve_2d_avx2()
|
/external/flac/libFLAC/ |
D | lpc_intrin_avx2.c | 65 q0 = _mm256_set1_epi32(0xffff & qlp_coeff[0 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 66 q1 = _mm256_set1_epi32(0xffff & qlp_coeff[1 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 67 q2 = _mm256_set1_epi32(0xffff & qlp_coeff[2 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 68 q3 = _mm256_set1_epi32(0xffff & qlp_coeff[3 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 69 q4 = _mm256_set1_epi32(0xffff & qlp_coeff[4 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 70 q5 = _mm256_set1_epi32(0xffff & qlp_coeff[5 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 71 q6 = _mm256_set1_epi32(0xffff & qlp_coeff[6 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 72 q7 = _mm256_set1_epi32(0xffff & qlp_coeff[7 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 73 q8 = _mm256_set1_epi32(0xffff & qlp_coeff[8 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() 74 q9 = _mm256_set1_epi32(0xffff & qlp_coeff[9 ]); in FLAC__lpc_compute_residual_from_qlp_coefficients_16_intrin_avx2() [all …]
|
/external/libaom/libaom/av1/encoder/x86/ |
D | highbd_fwd_txfm_avx2.c | 77 const __m256i rounding = _mm256_set1_epi32(1 << (shift - 1)); in col_txfm_8x8_rounding() 187 __m256i round = _mm256_set1_epi32(1 << (bit - 1)); in av1_round_shift_32_8xn_avx2() 227 const __m256i ww0 = _mm256_set1_epi32(w0); \ 228 const __m256i ww1 = _mm256_set1_epi32(w1); \ 259 const __m256i cospi32 = _mm256_set1_epi32(cospi[32]); in av1_fdct8_avx2() 260 const __m256i cospim32 = _mm256_set1_epi32(-cospi[32]); in av1_fdct8_avx2() 261 const __m256i cospi48 = _mm256_set1_epi32(cospi[48]); in av1_fdct8_avx2() 262 const __m256i cospi16 = _mm256_set1_epi32(cospi[16]); in av1_fdct8_avx2() 263 const __m256i cospi56 = _mm256_set1_epi32(cospi[56]); in av1_fdct8_avx2() 264 const __m256i cospi8 = _mm256_set1_epi32(cospi[8]); in av1_fdct8_avx2() [all …]
|
D | av1_fwd_txfm2d_avx2.c | 26 const __m256i _r = _mm256_set1_epi32(1 << (cos_bit - 1)); in fdct16x16_new_avx2() 119 const __m256i _r = _mm256_set1_epi32(1 << (cos_bit - 1)); in fdct16x32_new_avx2() 310 const __m256i _r = _mm256_set1_epi32(1 << (cos_bit - 1)); in fdct16x64_new_avx2() 718 const __m256i _r = _mm256_set1_epi32(1 << (cos_bit - 1)); in av1_fdct32_new_avx2() 871 const __m256i _r = _mm256_set1_epi32(1 << (cos_bit - 1)); in av1_fdct64_new_avx2() 873 __m256i cospi_m32 = _mm256_set1_epi32(-cospi[32]); in av1_fdct64_new_avx2() 874 __m256i cospi_p32 = _mm256_set1_epi32(cospi[32]); in av1_fdct64_new_avx2() 875 __m256i cospi_m16 = _mm256_set1_epi32(-cospi[16]); in av1_fdct64_new_avx2() 876 __m256i cospi_p48 = _mm256_set1_epi32(cospi[48]); in av1_fdct64_new_avx2() 877 __m256i cospi_m48 = _mm256_set1_epi32(-cospi[48]); in av1_fdct64_new_avx2() [all …]
|
D | av1_fwd_txfm_avx2.h | 23 const __m256i ww0 = _mm256_set1_epi32(w0); in btf_32_avx2_type0() 24 const __m256i ww1 = _mm256_set1_epi32(w1); in btf_32_avx2_type0() 42 const __m256i ww0 = _mm256_set1_epi32(w0); in btf_32_avx2_type1() 43 const __m256i ww1 = _mm256_set1_epi32(w1); in btf_32_avx2_type1()
|
D | pickrst_avx2.c | 182 _mm256_set1_epi32(*((uint32_t *)(dgd_ijk + l))); in acc_stat_highbd_win7_one_line_avx2() 286 _mm256_set1_epi32(*((uint32_t *)(dgd_ijk + l))); in acc_stat_highbd_win5_one_line_avx2() 489 return _mm256_set1_epi32( in pair_set_epi16() 499 const __m256i rounding = _mm256_set1_epi32(1 << (shift - 1)); in av1_lowbd_pixel_proj_error_avx2() 631 const __m256i rounding = _mm256_set1_epi32(1 << (shift - 1)); in av1_highbd_pixel_proj_error_avx2() 637 const __m256i xq0 = _mm256_set1_epi32(xq[0]); in av1_highbd_pixel_proj_error_avx2() 638 const __m256i xq1 = _mm256_set1_epi32(xq[1]); in av1_highbd_pixel_proj_error_avx2() 723 const __m256i xq_active = _mm256_set1_epi32(xq_on); in av1_highbd_pixel_proj_error_avx2() 725 _mm256_set1_epi32(-xq_on * (1 << SGRPROJ_RST_BITS)); in av1_highbd_pixel_proj_error_avx2()
|
D | wedge_utils_avx2.c | 158 const __m256i v_neg_w = _mm256_set1_epi32(0xffff0001); in av1_wedge_compute_delta_squares_avx2()
|
/external/libaom/libaom/aom_dsp/x86/ |
D | obmc_sad_avx2.c | 35 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in obmc_sad_w4_avx2() 76 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in obmc_sad_w8n_avx2() 158 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in hbd_obmc_sad_w4_avx2() 201 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in hbd_obmc_sad_w8n_avx2()
|
D | highbd_convolve_avx2.c | 136 const __m256i round_const_bits = _mm256_set1_epi32((1 << bits) >> 1); in av1_highbd_convolve_y_sr_avx2() 282 _mm256_set1_epi32(((1 << conv_params->round_0) >> 1)); in av1_highbd_convolve_x_sr_avx2() 287 const __m256i round_const_bits = _mm256_set1_epi32((1 << bits) >> 1); in av1_highbd_convolve_x_sr_avx2() 445 const __m256i p0 = _mm256_set1_epi32(0x03020100); in pack_filters() 446 const __m256i p1 = _mm256_set1_epi32(0x07060504); in pack_filters() 447 const __m256i p2 = _mm256_set1_epi32(0x0b0a0908); in pack_filters() 448 const __m256i p3 = _mm256_set1_epi32(0x0f0e0d0c); in pack_filters() 487 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in filter_8x1_pixels() 569 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in aom_highbd_filter_block1d4_h4_avx2() 640 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in aom_highbd_filter_block1d8_h4_avx2() [all …]
|
D | intrapred_avx2.c | 1014 a16 = _mm256_set1_epi32(16); in highbd_dr_prediction_z1_4xN_internal_avx2() 1041 _mm256_slli_epi32(_mm256_set1_epi32(x), upsample_above), in highbd_dr_prediction_z1_4xN_internal_avx2() 1042 _mm256_set1_epi32(0x3f)), in highbd_dr_prediction_z1_4xN_internal_avx2() 1047 _mm256_and_si256(_mm256_set1_epi32(x), _mm256_set1_epi32(0x3f)), 1); in highbd_dr_prediction_z1_4xN_internal_avx2() 1097 a16 = _mm256_set1_epi32(16); in highbd_dr_prediction_z1_8xN_internal_avx2() 1099 max_base_x256 = _mm256_set1_epi32(max_base_x); in highbd_dr_prediction_z1_8xN_internal_avx2() 1134 _mm256_slli_epi32(_mm256_set1_epi32(x), upsample_above), in highbd_dr_prediction_z1_8xN_internal_avx2() 1135 _mm256_set1_epi32(0x3f)), in highbd_dr_prediction_z1_8xN_internal_avx2() 1141 _mm256_and_si256(_mm256_set1_epi32(x), _mm256_set1_epi32(0x3f)), 1); in highbd_dr_prediction_z1_8xN_internal_avx2() 1195 a16 = _mm256_set1_epi32(16); in highbd_dr_prediction_z1_16xN_internal_avx2() [all …]
|
D | obmc_variance_avx2.c | 35 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in obmc_variance_w8n() 91 const __m256i v_bias_d = _mm256_set1_epi32((1 << 12) >> 1); in obmc_variance_w16n()
|
D | txfm_common_avx2.h | 24 return _mm256_set1_epi32( in pair_set_w16_epi16() 224 round = _mm256_set1_epi32(1 << (bit - 1)); in av1_round_shift_32_avx2() 251 const __m256i sqrt2 = _mm256_set1_epi32(val); in av1_round_shift_rect_array_32_avx2()
|
D | masked_sad_intrin_avx2.c | 213 _mm256_set1_epi32((1 << AOM_BLEND_A64_ROUND_BITS) >> 1); in highbd_masked_sad8xh_avx2() 270 _mm256_set1_epi32((1 << AOM_BLEND_A64_ROUND_BITS) >> 1); in highbd_masked_sad16xh_avx2()
|
/external/libvpx/libvpx/vpx_dsp/x86/ |
D | highbd_convolve_avx2.c | 277 const __m256i p0 = _mm256_set1_epi32(0x03020100); in pack_filters() 278 const __m256i p1 = _mm256_set1_epi32(0x07060504); in pack_filters() 279 const __m256i p2 = _mm256_set1_epi32(0x0b0a0908); in pack_filters() 280 const __m256i p3 = _mm256_set1_epi32(0x0f0e0d0c); in pack_filters() 308 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in filter_8x1_pixels() 393 const __m256i p = _mm256_set1_epi32(0x09080706); in pack_2t_filter() 442 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in filter_16_2t_pixels() 453 const __m256i rounding = _mm256_set1_epi32(1 << (CONV8_ROUNDING_BITS - 1)); in filter_8x1_2t_pixels() 954 _mm256_set1_epi32(CONV8_ROUNDING_NUM); // Used for rounding in vpx_highbd_filter_block1d4_h4_avx2() 1041 _mm256_set1_epi32(CONV8_ROUNDING_NUM); // Used for rounding in vpx_highbd_filter_block1d8_h4_avx2() [all …]
|
D | vpx_subpixel_8t_intrin_avx2.c | 766 kernel_reg = _mm256_shuffle_epi8(kernel_reg, _mm256_set1_epi32(0x05040302u)); in vpx_filter_block1d4_h4_avx2() 851 kernel_reg = _mm256_shuffle_epi8(kernel_reg, _mm256_set1_epi32(0x05040302u)); in vpx_filter_block1d4_v4_avx2()
|
/external/mesa3d/src/gallium/drivers/swr/rasterizer/common/ |
D | simdlib_256_avx.inl | 294 __m256i ret = _mm256_set1_epi32(0); 358 __m256i ret = _mm256_set1_epi32(0); 630 return _mm256_set1_epi32(i);
|