/external/libaom/libaom/aom_dsp/simd/ |
D | v256_intrinsics_x86.h | 143 _mm256_permute4x64_epi64(b, _MM_SHUFFLE(3, 1, 2, 0)), in v256_ziplo_8() 144 _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0))); in v256_ziplo_8() 149 _mm256_permute4x64_epi64(b, _MM_SHUFFLE(3, 1, 2, 0)), in v256_ziphi_8() 150 _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0))); in v256_ziphi_8() 155 _mm256_permute4x64_epi64(b, _MM_SHUFFLE(3, 1, 2, 0)), in v256_ziplo_16() 156 _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0))); in v256_ziplo_16() 161 _mm256_permute4x64_epi64(b, _MM_SHUFFLE(3, 1, 2, 0)), in v256_ziphi_16() 162 _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0))); in v256_ziphi_16() 167 _mm256_permute4x64_epi64(b, _MM_SHUFFLE(3, 1, 2, 0)), in v256_ziplo_32() 168 _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0))); in v256_ziplo_32() [all …]
|
/external/tensorflow/third_party/eigen3/unsupported/Eigen/CXX11/src/FixedPoint/ |
D | TypeCastingAVX512.h | 43 Packet8i ab_int16_low = _mm256_permute4x64_epi64( 47 Packet8i ab_int16_high = _mm256_permute4x64_epi64( 74 Packet8i ab_int16_low = _mm256_permute4x64_epi64( 78 Packet8i cd_int16_low = _mm256_permute4x64_epi64( 82 Packet8i ab_int16_high = _mm256_permute4x64_epi64( 86 Packet8i cd_int16_high = _mm256_permute4x64_epi64( 90 Packet8i abcd_int8_low = _mm256_permute4x64_epi64( 93 _mm256_permute4x64_epi64(_mm256_packs_epi16(ab_int16_high, cd_int16_high),
|
/external/libaom/libaom/av1/common/x86/ |
D | reconinter_avx2.c | 59 _mm256_castsi256_si128(_mm256_permute4x64_epi64(m8, 0xd8)); in av1_build_compound_diffwtd_mask_avx2() 104 _mm256_permute4x64_epi64(_mm256_packus_epi16(m16AL, m16BL), 0xd8); in av1_build_compound_diffwtd_mask_avx2() 126 _mm256_permute4x64_epi64(_mm256_packus_epi16(m16L, m16H), 0xd8); in av1_build_compound_diffwtd_mask_avx2() 197 _mm256_castsi256_si128(_mm256_permute4x64_epi64(m8, 0xd8))); in build_compound_diffwtd_mask_d16_avx2() 216 yy_storeu_256(mask, _mm256_permute4x64_epi64(m8, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() 233 yy_storeu_256(mask, _mm256_permute4x64_epi64(m8, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() 250 yy_storeu_256(mask, _mm256_permute4x64_epi64(m8, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() 276 yy_storeu_256(mask, _mm256_permute4x64_epi64(m8AB, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() 277 yy_storeu_256(mask + 32, _mm256_permute4x64_epi64(m8CD, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() 321 yy_storeu_256(mask, _mm256_permute4x64_epi64(m8AB, 0xd8)); in build_compound_diffwtd_mask_d16_avx2() [all …]
|
D | cfl_avx2.c | 134 top = _mm256_permute4x64_epi64(top, _MM_SHUFFLE(3, 1, 2, 0)); in cfl_luma_subsampling_444_lbd_avx2() 181 hsum = _mm256_permute4x64_epi64(hsum, _MM_SHUFFLE(3, 1, 2, 0)); in cfl_luma_subsampling_420_hbd_avx2() 214 hsum = _mm256_permute4x64_epi64(hsum, _MM_SHUFFLE(3, 1, 2, 0)); in cfl_luma_subsampling_422_hbd_avx2() 268 res = _mm256_permute4x64_epi64(res, _MM_SHUFFLE(3, 1, 2, 0)); in cfl_predict_lbd_avx2() 387 a = _mm256_permute4x64_epi64(a, _MM_SHUFFLE(3, 1, 2, 0)); in fill_sum_epi32()
|
D | av1_inv_txfm_avx2.h | 50 _mm256_permute4x64_epi64(_mm256_packus_epi16(u, u), 168)); in write_recon_w16_avx2()
|
D | selfguided_avx2.c | 706 const __m256i tmp2 = _mm256_permute4x64_epi64(tmp, 0xd8); in av1_apply_selfguided_restoration_avx2() 715 const __m256i tmp2 = _mm256_permute4x64_epi64(tmp, 0xd8); in av1_apply_selfguided_restoration_avx2() 719 _mm256_castsi256_si128(_mm256_permute4x64_epi64(res, 0xd8)); in av1_apply_selfguided_restoration_avx2()
|
D | convolve_avx2.c | 360 res_8b = _mm256_permute4x64_epi64(res_8b, 216); in av1_convolve_x_sr_avx2() 432 res_8b = _mm256_permute4x64_epi64(res_8b, 216); in av1_convolve_x_sr_avx2()
|
/external/libaom/libaom/aom_dsp/x86/ |
D | adaptive_quantize_avx2.c | 26 *zbin = _mm256_permute4x64_epi64(*zbin, 0x54); in load_b_values_avx2() 29 *round = _mm256_permute4x64_epi64(*round, 0x54); in load_b_values_avx2() 31 *quant = _mm256_permute4x64_epi64(*quant, 0x54); in load_b_values_avx2() 34 *dequant = _mm256_permute4x64_epi64(*dequant, 0x54); in load_b_values_avx2() 36 *shift = _mm256_permute4x64_epi64(*shift, 0x54); in load_b_values_avx2() 69 _mm256_permute4x64_epi64(_mm256_packs_epi32(cmp_mask0, cmp_mask1), 0xd8); in update_mask0_avx2() 138 cmp_mask = _mm256_permute4x64_epi64(temp0, 0xd8); in aom_quantize_b_adaptive_avx2() 171 cmp_mask = _mm256_permute4x64_epi64(temp0, 0xd8); in aom_quantize_b_adaptive_avx2()
|
D | highbd_adaptive_quantize_avx2.c | 57 _mm256_permute4x64_epi64(_mm256_packs_epi32(cmp_mask0, cmp_mask1), 0xd8); in highbd_update_mask0_avx2() 154 cmp_mask = _mm256_permute4x64_epi64(_mm256_packs_epi32(temp0, temp1), 0xd8); in aom_highbd_quantize_b_adaptive_avx2() 195 cmp_mask = _mm256_permute4x64_epi64(_mm256_packs_epi32(temp0, temp1), 0xd8); in aom_highbd_quantize_b_adaptive_avx2() 336 cmp_mask = _mm256_permute4x64_epi64(_mm256_packs_epi32(temp0, temp1), 0xd8); in aom_highbd_quantize_b_32x32_adaptive_avx2() 379 cmp_mask = _mm256_permute4x64_epi64(_mm256_packs_epi32(temp0, temp1), 0xd8); in aom_highbd_quantize_b_32x32_adaptive_avx2()
|
D | masked_sad_intrin_avx2.c | 63 res = _mm256_permute4x64_epi64(res, 0xd8); in masked_sad32xh_avx2() 116 res = _mm256_permute4x64_epi64(res, 0xd8); in masked_sad16xh_avx2()
|
D | blend_a64_mask_avx2.c | 45 res = _mm256_permute4x64_epi64(res, 0xd8); in blend_a64_d16_mask_w16_avx2() 78 res = _mm256_permute4x64_epi64(res, 0xd8); in blend_a64_d16_mask_w32_avx2() 410 const __m256i v_s0_s_b = _mm256_permute4x64_epi64(v_s0_b, 0xd8); in blend_16_u8_avx2() 411 const __m256i v_s1_s_b = _mm256_permute4x64_epi64(v_s1_b, 0xd8); in blend_16_u8_avx2() 419 const __m256i v_res = _mm256_permute4x64_epi64(v_res_b, 0xd8); in blend_16_u8_avx2() 500 _mm256_permute4x64_epi64(_mm256_packus_epi16(v_m0l_w, v_m0h_w), 0xd8); in blend_a64_mask_sx_sy_w32n_avx2() 625 _mm256_permute4x64_epi64(_mm256_unpacklo_epi64(v_al_b, v_ah_b), 0xd8); in blend_a64_mask_sx_w32n_avx2()
|
/external/libvpx/libvpx/vp9/encoder/x86/ |
D | vp9_quantize_avx2.c | 38 const __m256i _coeff256 = _mm256_permute4x64_epi64(*coeff256, 0xd8); in scan_eob_256() 78 round256 = _mm256_permute4x64_epi64(round256, 0x54); in vp9_quantize_fp_avx2() 81 quant256 = _mm256_permute4x64_epi64(quant256, 0x54); in vp9_quantize_fp_avx2() 84 dequant256 = _mm256_permute4x64_epi64(dequant256, 0x54); in vp9_quantize_fp_avx2()
|
/external/libaom/libaom/av1/encoder/x86/ |
D | encodetxb_avx2.c | 48 const __m256i res = _mm256_permute4x64_epi64(res_, 0xd8); in av1_txb_init_levels_avx2() 65 const __m256i res_ = _mm256_permute4x64_epi64(absABCD, 0xd8); in av1_txb_init_levels_avx2() 92 const __m256i res_ = _mm256_permute4x64_epi64(absABCD, 0xd8); in av1_txb_init_levels_avx2() 113 const __m256i res_ = _mm256_permute4x64_epi64(absABCD, 0xd8); in av1_txb_init_levels_avx2()
|
D | av1_quantize_avx2.c | 24 *c = _mm256_permute4x64_epi64(*c, 0xD8); in read_coeff() 186 round256 = _mm256_permute4x64_epi64(round256, 0x54); in av1_quantize_lp_avx2() 189 quant256 = _mm256_permute4x64_epi64(quant256, 0x54); in av1_quantize_lp_avx2() 192 dequant256 = _mm256_permute4x64_epi64(dequant256, 0x54); in av1_quantize_lp_avx2()
|
D | error_intrin_avx2.c | 26 *c = _mm256_permute4x64_epi64(y, 0xD8); in read_coeff()
|
D | pickrst_avx2.c | 519 const __m256i flt0_16b = _mm256_permute4x64_epi64( in av1_lowbd_pixel_proj_error_avx2() 523 const __m256i flt1_16b = _mm256_permute4x64_epi64( in av1_lowbd_pixel_proj_error_avx2() 571 const __m256i flt_16b = _mm256_permute4x64_epi64( in av1_lowbd_pixel_proj_error_avx2() 929 _mm256_permute4x64_epi64(_mm256_packs_epi32(vrl, vrh), 0xd8); in av1_highbd_pixel_proj_error_avx2() 1006 _mm256_permute4x64_epi64(_mm256_packs_epi32(vrl, vrh), 0xd8); in av1_highbd_pixel_proj_error_avx2()
|
/external/libvpx/libvpx/vpx_dsp/x86/ |
D | loopfilter_avx2.c | 623 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_p, res_p), 168)); in vpx_lpf_horizontal_16_dual_avx2() 629 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_q, res_q), 168)); in vpx_lpf_horizontal_16_dual_avx2() 637 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_p, res_p), 168)); in vpx_lpf_horizontal_16_dual_avx2() 645 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_q, res_q), 168)); in vpx_lpf_horizontal_16_dual_avx2() 663 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_p, res_p), 168)); in vpx_lpf_horizontal_16_dual_avx2() 669 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_q, res_q), 168)); in vpx_lpf_horizontal_16_dual_avx2() 681 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_p, res_p), 168)); in vpx_lpf_horizontal_16_dual_avx2() 689 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_q, res_q), 168)); in vpx_lpf_horizontal_16_dual_avx2() 707 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_p, res_p), 168)); in vpx_lpf_horizontal_16_dual_avx2() 713 _mm256_permute4x64_epi64(_mm256_packus_epi16(res_q, res_q), 168)); in vpx_lpf_horizontal_16_dual_avx2() [all …]
|
/external/gemmlowp/internal/ |
D | pack_avx.h | 99 __m256i ymm11 = _mm256_permute4x64_epi64(ymm7, 0xd8); in Pack() 100 __m256i ymm12 = _mm256_permute4x64_epi64(ymm8, 0xd8); in Pack() 102 __m256i ymm15 = _mm256_permute4x64_epi64(ymm13, 0xd8); in Pack() 103 __m256i ymm16 = _mm256_permute4x64_epi64(ymm14, 0xd8); in Pack()
|
/external/libgav1/libgav1/src/dsp/x86/ |
D | cdef_avx2.cc | 397 _mm256_castsi256_si128(_mm256_permute4x64_epi64(e, 0x08)); in Cost0Or4_Pair() 432 _mm256_castsi256_si128(_mm256_permute4x64_epi64(e, 0x08)); in CostOdd_Pair() 449 _mm256_castsi256_si128(_mm256_permute4x64_epi64(c, 0x08)); in Cost2And6_Pair()
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 3x8c8-minmax-avx2.c | 200 vacc01x01234567 = _mm256_permute4x64_epi64(vacc01x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_igemm_minmax_ukernel_3x8c8__avx2() 201 vacc22x01234567 = _mm256_permute4x64_epi64(vacc22x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_igemm_minmax_ukernel_3x8c8__avx2()
|
D | 1x8c8-minmax-avx2.c | 133 vacc00x01234567 = _mm256_permute4x64_epi64(vacc00x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_igemm_minmax_ukernel_1x8c8__avx2()
|
/external/XNNPACK/src/qs8-gemm/gen/ |
D | 3x8c8-minmax-avx2.c | 183 vacc01x01234567 = _mm256_permute4x64_epi64(vacc01x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_minmax_ukernel_3x8c8__avx2() 184 vacc22x01234567 = _mm256_permute4x64_epi64(vacc22x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_minmax_ukernel_3x8c8__avx2()
|
D | 3x8c8-xw-minmax-avx2.c | 179 vacc01x01234567 = _mm256_permute4x64_epi64(vacc01x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_xw_minmax_ukernel_3x8c8__avx2() 180 vacc22x01234567 = _mm256_permute4x64_epi64(vacc22x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_xw_minmax_ukernel_3x8c8__avx2()
|
D | 1x8c8-minmax-avx2.c | 120 vacc00x01234567 = _mm256_permute4x64_epi64(vacc00x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_minmax_ukernel_1x8c8__avx2()
|
D | 1x8c8-xw-minmax-avx2.c | 116 vacc00x01234567 = _mm256_permute4x64_epi64(vacc00x01234567, _MM_SHUFFLE(3, 1, 2, 0)); in xnn_qs8_gemm_xw_minmax_ukernel_1x8c8__avx2()
|