/external/libjpeg-turbo/simd/arm/ |
D | jdmrgext-neon.c | 89 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), cr)); in jsimd_h2v1_merged_upsample_neon() 91 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), cb)); in jsimd_h2v1_merged_upsample_neon() 109 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(g_sub_y), in jsimd_h2v1_merged_upsample_neon() 112 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(r_sub_y), in jsimd_h2v1_merged_upsample_neon() 115 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(b_sub_y), in jsimd_h2v1_merged_upsample_neon() 118 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(g_sub_y), in jsimd_h2v1_merged_upsample_neon() 121 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(r_sub_y), in jsimd_h2v1_merged_upsample_neon() 124 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(b_sub_y), in jsimd_h2v1_merged_upsample_neon() 168 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), cr)); in jsimd_h2v1_merged_upsample_neon() 170 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), cb)); in jsimd_h2v1_merged_upsample_neon() [all …]
|
D | jdcolext-neon.c | 77 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), in jsimd_ycc_rgb_convert_neon() 80 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), in jsimd_ycc_rgb_convert_neon() 83 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), in jsimd_ycc_rgb_convert_neon() 86 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(neg_128), in jsimd_ycc_rgb_convert_neon() 120 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(r_sub_y_l), in jsimd_ycc_rgb_convert_neon() 123 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(r_sub_y_h), in jsimd_ycc_rgb_convert_neon() 126 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(b_sub_y_l), in jsimd_ycc_rgb_convert_neon() 129 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(b_sub_y_h), in jsimd_ycc_rgb_convert_neon() 132 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(g_sub_y_l), in jsimd_ycc_rgb_convert_neon() 135 vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(g_sub_y_h), in jsimd_ycc_rgb_convert_neon() [all …]
|
/external/libvpx/libvpx/vpx_dsp/arm/ |
D | avg_neon.c | 43 sum = vaddw_u8(sum, d); in vpx_avg_8x8_neon() 96 vec_sum_lo = vaddw_u8(vec_sum_lo, vget_low_u8(vec_row1)); in vpx_int_pro_row_neon() 97 vec_sum_hi = vaddw_u8(vec_sum_hi, vget_high_u8(vec_row1)); in vpx_int_pro_row_neon() 99 vec_sum_lo = vaddw_u8(vec_sum_lo, vget_low_u8(vec_row2)); in vpx_int_pro_row_neon() 100 vec_sum_hi = vaddw_u8(vec_sum_hi, vget_high_u8(vec_row2)); in vpx_int_pro_row_neon() 102 vec_sum_lo = vaddw_u8(vec_sum_lo, vget_low_u8(vec_row3)); in vpx_int_pro_row_neon() 103 vec_sum_hi = vaddw_u8(vec_sum_hi, vget_high_u8(vec_row3)); in vpx_int_pro_row_neon() 105 vec_sum_lo = vaddw_u8(vec_sum_lo, vget_low_u8(vec_row4)); in vpx_int_pro_row_neon() 106 vec_sum_hi = vaddw_u8(vec_sum_hi, vget_high_u8(vec_row4)); in vpx_int_pro_row_neon() 108 vec_sum_lo = vaddw_u8(vec_sum_lo, vget_low_u8(vec_row5)); in vpx_int_pro_row_neon() [all …]
|
D | loopfilter_neon.c | 178 *sum = vaddw_u8(*sum, add0); in filter_update_8() 179 *sum = vaddw_u8(*sum, add1); in filter_update_8() 191 *sum0 = vaddw_u8(*sum0, vget_low_u8(add0)); in filter_update_16() 192 *sum1 = vaddw_u8(*sum1, vget_high_u8(add0)); in filter_update_16() 193 *sum0 = vaddw_u8(*sum0, vget_low_u8(add1)); in filter_update_16() 194 *sum1 = vaddw_u8(*sum1, vget_high_u8(add1)); in filter_update_16() 241 sum = vaddw_u8(sum, p3); // 3*p3 in calc_7_tap_filter_8() 242 sum = vaddw_u8(sum, p2); // 3*p3+p2 in calc_7_tap_filter_8() 243 sum = vaddw_u8(sum, p2); // 3*p3+2*p2 in calc_7_tap_filter_8() 244 sum = vaddw_u8(sum, p1); // 3*p3+2*p2+p1 in calc_7_tap_filter_8() [all …]
|
D | idct4x4_add_neon.c | 48 sum[0] = vaddw_u8(vreinterpretq_u16_s16(a[0]), s); in vpx_idct4x4_16_add_neon() 49 sum[1] = vaddw_u8(vreinterpretq_u16_s16(a[1]), vreinterpret_u8_u32(s32)); in vpx_idct4x4_16_add_neon()
|
D | idct4x4_1_add_neon.c | 25 a = vaddw_u8(vreinterpretq_u16_s16(res), vreinterpret_u8_u32(*d)); in idct4x4_1_add_kernel()
|
D | idct32x32_add_neon.c | 60 q0 = vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(q0), d[0])); in store_combine_results() 61 q1 = vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(q1), d[1])); in store_combine_results() 62 q2 = vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(q2), d[2])); in store_combine_results() 63 q3 = vreinterpretq_s16_u16(vaddw_u8(vreinterpretq_u16_s16(q3), d[3])); in store_combine_results()
|
/external/libaom/libaom/aom_dsp/arm/ |
D | loopfilter_neon.c | 232 out = vaddw_u8(out, *p1q1); in lpf_14_neon() 233 out = vaddw_u8(out, *p0q0); in lpf_14_neon() 235 out = vaddw_u8(out, q0p0); in lpf_14_neon() 236 out_pq1 = vaddw_u8(out, *p3q3); in lpf_14_neon() 237 out_pq2 = vaddw_u8(out_pq1, *p3q3); in lpf_14_neon() 238 out_pq2 = vaddw_u8(out_pq2, *p2q2); in lpf_14_neon() 239 out_pq1 = vaddw_u8(out_pq1, *p1q1); in lpf_14_neon() 240 out_pq1 = vaddw_u8(out_pq1, q1p1); in lpf_14_neon() 242 out_pq0 = vaddw_u8(out, *p0q0); in lpf_14_neon() 243 out_pq0 = vaddw_u8(out_pq0, q1p1); in lpf_14_neon() [all …]
|
D | avg_neon.c | 43 sum = vaddw_u8(sum, e); in aom_avg_8x8_neon()
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | cdef_neon.cc | 96 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[2], 6)); in AddPartial_D0_D4() 103 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[3], 5)); in AddPartial_D0_D4() 105 *partial_hi = vaddw_u8(*partial_hi, vext_u8(v_src[3], v_zero, 5)); in AddPartial_D0_D4() 108 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[4], 4)); in AddPartial_D0_D4() 110 *partial_hi = vaddw_u8(*partial_hi, vext_u8(v_src[4], v_zero, 4)); in AddPartial_D0_D4() 113 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[5], 3)); in AddPartial_D0_D4() 115 *partial_hi = vaddw_u8(*partial_hi, vext_u8(v_src[5], v_zero, 3)); in AddPartial_D0_D4() 118 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[6], 2)); in AddPartial_D0_D4() 120 *partial_hi = vaddw_u8(*partial_hi, vext_u8(v_src[6], v_zero, 2)); in AddPartial_D0_D4() 123 *partial_lo = vaddw_u8(*partial_lo, vext_u8(v_zero, v_src[7], 1)); in AddPartial_D0_D4() [all …]
|
D | intra_edge_neon.cc | 160 const uint16x8_t sum_123_lo = vaddw_u8( in IntraEdgeFilter_NEON() 167 vaddw_u8(vaddl_u8(vget_high_u8(src_1), vget_high_u8(src_2)), in IntraEdgeFilter_NEON() 190 const uint16x8_t sum_123_lo = vaddw_u8( in IntraEdgeFilter_NEON() 197 vaddw_u8(vaddl_u8(vget_high_u8(src_1), vget_high_u8(src_2)), in IntraEdgeFilter_NEON()
|
D | intrapred_neon.cc | 175 return Sum(vaddw_u8(vpaddlq_u8(val_1), val)); in DcSum_NEON() 196 return Sum(vaddw_u8(vpaddlq_u8(val_1), val_0)); in DcSum_NEON() 198 return Sum(vaddw_u8(LoadAndAdd32(ref_1_u8), val_0)); in DcSum_NEON() 208 return Sum(vaddw_u8(vpaddlq_u8(val_0), val_1)); in DcSum_NEON() 211 return Sum(vaddw_u8(vpaddlq_u8(val_0), val_1)); in DcSum_NEON() 232 return Sum(vaddw_u8(sum_0, val_1)); in DcSum_NEON()
|
D | loop_filter_neon.cc | 294 uint16x8_t sum = vaddw_u8(p2q2_double, p2q2); in Filter6() 313 sum = vaddw_u8(sum, q0p0); in Filter6() 530 uint16x8_t sum = vaddw_u8(vaddl_u8(p3q3, p3q3), p3q3); in Filter8() 549 sum = vaddw_u8(sum, q0p0); in Filter8() 779 sum = vaddw_u8(sum, q0p0); in Filter14()
|
D | inverse_transform_neon.cc | 1836 vaddw_u8(vreinterpretq_u16_s16(vcombine_s16(a, a)), frame_data); in IdentityColumnStoreToFrame() 1865 const uint16x8_t b = vaddw_u8(vreinterpretq_u16_s16(a), frame_data); in IdentityColumnStoreToFrame() 1882 const uint16x8_t b = vaddw_u8(vreinterpretq_u16_s16(a), frame_data); in IdentityColumnStoreToFrame() 1912 vaddw_u8(vreinterpretq_u16_s16(vcombine_s16(a, a)), frame_data); in Identity4RowColumnStoreToFrame() 1932 const uint16x8_t b = vaddw_u8(vreinterpretq_u16_s16(a), frame_data); in Identity4RowColumnStoreToFrame() 2177 const uint16x8_t b = vaddw_u8(vreinterpretq_u16_s16(residual), frame_data); in Wht4_NEON() 2291 vaddw_u8(vreinterpretq_u16_s16(vcombine_s16(a, a)), frame_data); in StoreToFrameWithRound() 2303 const uint16x8_t b = vaddw_u8(vreinterpretq_u16_s16(a), frame_data); in StoreToFrameWithRound() 2322 vaddw_u8(vreinterpretq_u16_s16(a), vget_low_u8(frame_data)); in StoreToFrameWithRound() 2324 vaddw_u8(vreinterpretq_u16_s16(a_hi), vget_high_u8(frame_data)); in StoreToFrameWithRound()
|
D | loop_restoration_neon.cc | 674 return vaddw_u8(sum, src[2]); in Sum3W_16() 679 return vaddw_u8(sum, vget_low_u8(src[2])); in Sum3WLo16() 684 return vaddw_u8(sum, vget_high_u8(src[2])); in Sum3WHi16() 691 return vaddw_u8(sum, vget_low_u8(src[4])); in Sum5WLo16() 698 return vaddw_u8(sum, vget_high_u8(src[4])); in Sum5WHi16() 774 return vaddw_u8(sum0123, s[4]); in Sum5Horizontal() 822 *row3 = vaddw_u8(sum12, s[3]); in SumHorizontal() 862 return vaddw_u8(sum3, in Sum343() 886 return vaddw_u8(sum5, in Sum565() 1101 *sum_ma343 = vaddw_u8( in Store343_444() [all …]
|
/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | idct_blk_neon.c | 40 q1u16 = vaddw_u8(vreinterpretq_u16_s16(q2Add.val[i]), in idct_dequant_0_2x_neon() 42 q2u16 = vaddw_u8(vreinterpretq_u16_s16(q2Add.val[i]), in idct_dequant_0_2x_neon() 200 vaddw_u8(vreinterpretq_u16_s16(q2tmp2.val[0]), vreinterpret_u8_s32(d28))); in idct_dequant_full_2x_neon() 202 vaddw_u8(vreinterpretq_u16_s16(q2tmp2.val[1]), vreinterpret_u8_s32(d29))); in idct_dequant_full_2x_neon() 204 vaddw_u8(vreinterpretq_u16_s16(q2tmp3.val[0]), vreinterpret_u8_s32(d30))); in idct_dequant_full_2x_neon() 206 vaddw_u8(vreinterpretq_u16_s16(q2tmp3.val[1]), vreinterpret_u8_s32(d31))); in idct_dequant_full_2x_neon()
|
D | dequant_idct_neon.c | 125 vaddw_u8(vreinterpretq_u16_s16(q1), vreinterpret_u8_s32(d14))); in vp8_dequant_idct_add_neon() 127 vaddw_u8(vreinterpretq_u16_s16(q2), vreinterpret_u8_s32(d15))); in vp8_dequant_idct_add_neon()
|
D | dc_only_idct_add_neon.c | 33 q1u16 = vaddw_u8(qAdd, vreinterpret_u8_u32(d2u32)); in vp8_dc_only_idct_add_neon()
|
D | shortidct4x4llm_neon.c | 112 q1u16 = vaddw_u8(vreinterpretq_u16_s16(q1s16), vreinterpret_u8_u32(d6u32)); in vp8_short_idct4x4llm_neon()
|
/external/zlib/ |
D | adler32_simd.c | 275 v_column_sum_1 = vaddw_u8(v_column_sum_1, vget_low_u8 (bytes1)); in adler32_simd_() 276 v_column_sum_2 = vaddw_u8(v_column_sum_2, vget_high_u8(bytes1)); in adler32_simd_() 277 v_column_sum_3 = vaddw_u8(v_column_sum_3, vget_low_u8 (bytes2)); in adler32_simd_() 278 v_column_sum_4 = vaddw_u8(v_column_sum_4, vget_high_u8(bytes2)); in adler32_simd_()
|
/external/angle/third_party/zlib/ |
D | adler32_simd.c | 275 v_column_sum_1 = vaddw_u8(v_column_sum_1, vget_low_u8 (bytes1)); in adler32_simd_() 276 v_column_sum_2 = vaddw_u8(v_column_sum_2, vget_high_u8(bytes1)); in adler32_simd_() 277 v_column_sum_3 = vaddw_u8(v_column_sum_3, vget_low_u8 (bytes2)); in adler32_simd_() 278 v_column_sum_4 = vaddw_u8(v_column_sum_4, vget_high_u8(bytes2)); in adler32_simd_()
|
/external/libvpx/libvpx/vp9/common/arm/neon/ |
D | vp9_iht4x4_add_neon.c | 70 sum[0] = vaddw_u8(vreinterpretq_u16_s16(a[0]), s[0]); in vp9_iht4x4_16_add_neon() 71 sum[1] = vaddw_u8(vreinterpretq_u16_s16(a[1]), s[1]); in vp9_iht4x4_16_add_neon()
|
/external/XNNPACK/src/qu8-gavgpool/ |
D | 7p7x-minmax-neon-c8.c | 56 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8() 93 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8() 155 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8() 227 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7p7x__neon_c8()
|
D | 7x-minmax-neon-c8.c | 79 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7x__neon_c8() 149 const uint16x8_t vsum016 = vaddw_u8(vsum01, vi6); in xnn_qu8_gavgpool_minmax_ukernel_7x__neon_c8()
|
/external/XNNPACK/src/qu8-avgpool/ |
D | 9x-minmax-neon-c8.c | 129 const uint16x8_t vsum018 = vaddw_u8(vaddl_u8(vi0, vi1), vi8); in xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8() 202 const uint16x8_t vsum018 = vaddw_u8(vaddl_u8(vi0, vi1), vi8); in xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8()
|