/external/libjpeg-turbo/simd/arm/ |
D | jidctint-neon.c | 201 int16x4_t row0 = vld1_s16(coef_block + 0 * DCTSIZE); in jsimd_idct_islow_neon() 202 int16x4_t row1 = vld1_s16(coef_block + 1 * DCTSIZE); in jsimd_idct_islow_neon() 203 int16x4_t row2 = vld1_s16(coef_block + 2 * DCTSIZE); in jsimd_idct_islow_neon() 204 int16x4_t row3 = vld1_s16(coef_block + 3 * DCTSIZE); in jsimd_idct_islow_neon() 205 int16x4_t row4 = vld1_s16(coef_block + 4 * DCTSIZE); in jsimd_idct_islow_neon() 206 int16x4_t row5 = vld1_s16(coef_block + 5 * DCTSIZE); in jsimd_idct_islow_neon() 207 int16x4_t row6 = vld1_s16(coef_block + 6 * DCTSIZE); in jsimd_idct_islow_neon() 208 int16x4_t row7 = vld1_s16(coef_block + 7 * DCTSIZE); in jsimd_idct_islow_neon() 211 int16x4_t quant_row0 = vld1_s16(quantptr + 0 * DCTSIZE); in jsimd_idct_islow_neon() 212 int16x4_t quant_row1 = vld1_s16(quantptr + 1 * DCTSIZE); in jsimd_idct_islow_neon() [all …]
|
D | jidctred-neon.c | 103 const int16x4_t consts = vld1_s16(jsimd_idct_2x2_neon_consts); in jsimd_idct_2x2_neon() 230 const int16x4_t consts1 = vld1_s16(jsimd_idct_4x4_neon_consts); in jsimd_idct_4x4_neon() 231 const int16x4_t consts2 = vld1_s16(jsimd_idct_4x4_neon_consts + 4); in jsimd_idct_4x4_neon() 232 const int16x4_t consts3 = vld1_s16(jsimd_idct_4x4_neon_consts + 8); in jsimd_idct_4x4_neon() 254 int16x4_t quant_row1 = vld1_s16(quantptr + 1 * DCTSIZE + 4); in jsimd_idct_4x4_neon() 255 int16x4_t quant_row2 = vld1_s16(quantptr + 2 * DCTSIZE + 4); in jsimd_idct_4x4_neon() 256 int16x4_t quant_row3 = vld1_s16(quantptr + 3 * DCTSIZE + 4); in jsimd_idct_4x4_neon() 257 int16x4_t quant_row5 = vld1_s16(quantptr + 5 * DCTSIZE + 4); in jsimd_idct_4x4_neon() 258 int16x4_t quant_row6 = vld1_s16(quantptr + 6 * DCTSIZE + 4); in jsimd_idct_4x4_neon() 259 int16x4_t quant_row7 = vld1_s16(quantptr + 7 * DCTSIZE + 4); in jsimd_idct_4x4_neon() [all …]
|
D | jidctfst-neon.c | 95 const int16x4_t consts = vld1_s16(jsimd_idct_ifast_neon_consts); in jsimd_idct_ifast_neon() 118 int16x4_t quant_row1 = vld1_s16(quantptr + 1 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 119 int16x4_t quant_row2 = vld1_s16(quantptr + 2 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 120 int16x4_t quant_row3 = vld1_s16(quantptr + 3 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 121 int16x4_t quant_row4 = vld1_s16(quantptr + 4 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 122 int16x4_t quant_row5 = vld1_s16(quantptr + 5 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 123 int16x4_t quant_row6 = vld1_s16(quantptr + 6 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 124 int16x4_t quant_row7 = vld1_s16(quantptr + 7 * DCTSIZE + 4); in jsimd_idct_ifast_neon() 193 int16x4_t quant_row1 = vld1_s16(quantptr + 1 * DCTSIZE); in jsimd_idct_ifast_neon() 194 int16x4_t quant_row2 = vld1_s16(quantptr + 2 * DCTSIZE); in jsimd_idct_ifast_neon() [all …]
|
D | jfdctint-neon.c | 94 const int16x4_t consts1 = vld1_s16(jsimd_fdct_islow_neon_consts); in jsimd_fdct_islow_neon() 95 const int16x4_t consts2 = vld1_s16(jsimd_fdct_islow_neon_consts + 4); in jsimd_fdct_islow_neon() 96 const int16x4_t consts3 = vld1_s16(jsimd_fdct_islow_neon_consts + 8); in jsimd_fdct_islow_neon()
|
/external/libhevc/common/arm/ |
D | ihevc_weighted_pred_neon_intr.c | 149 pi2_src_val1 = vld1_s16((int16_t *)pi2_src); in ihevc_weighted_pred_uni_neonintr() 152 pi2_src_val2 = vld1_s16((int16_t *)pi2_src_tmp); in ihevc_weighted_pred_uni_neonintr() 296 pi2_src_val1 = vld1_s16((int16_t *)pi2_src); in ihevc_weighted_pred_chroma_uni_neonintr() 299 pi2_src_val2 = vld1_s16((int16_t *)pi2_src_tmp); in ihevc_weighted_pred_chroma_uni_neonintr() 449 pi2_src1_val1 = vld1_s16((int16_t *)pi2_src1); in ihevc_weighted_pred_bi_neonintr() 453 pi2_src2_val1 = vld1_s16((int16_t *)pi2_src2); in ihevc_weighted_pred_bi_neonintr() 457 pi2_src1_val2 = vld1_s16((int16_t *)pi2_src_tmp1); in ihevc_weighted_pred_bi_neonintr() 460 pi2_src2_val2 = vld1_s16((int16_t *)pi2_src_tmp2); in ihevc_weighted_pred_bi_neonintr() 635 pi2_src1_val1 = vld1_s16((int16_t *)pi2_src1); in ihevc_weighted_pred_chroma_bi_neonintr() 639 pi2_src2_val1 = vld1_s16((int16_t *)pi2_src2); in ihevc_weighted_pred_chroma_bi_neonintr() [all …]
|
D | ihevc_quant_iquant_ssd_neon_intr.c | 161 s0 = vld1_s16(pi2_coeffs + j); in ihevc_quant_iquant_ssd_flat_scale_mat_neon() 162 s1 = vld1_s16(pi2_coeffs + j + (src_strd)); in ihevc_quant_iquant_ssd_flat_scale_mat_neon() 163 s2 = vld1_s16(pi2_coeffs + j + (2 * src_strd)); in ihevc_quant_iquant_ssd_flat_scale_mat_neon() 164 s3 = vld1_s16(pi2_coeffs + j + (3 * src_strd)); in ihevc_quant_iquant_ssd_flat_scale_mat_neon() 447 s0 = vld1_s16(pi2_coeffs + j); in ihevc_q_iq_ssd_flat_scale_mat_var_rnd_fact_neon() 448 s1 = vld1_s16(pi2_coeffs + j + (src_strd)); in ihevc_q_iq_ssd_flat_scale_mat_var_rnd_fact_neon() 449 s2 = vld1_s16(pi2_coeffs + j + (2 * src_strd)); in ihevc_q_iq_ssd_flat_scale_mat_var_rnd_fact_neon() 450 s3 = vld1_s16(pi2_coeffs + j + (3 * src_strd)); in ihevc_q_iq_ssd_flat_scale_mat_var_rnd_fact_neon()
|
/external/libvpx/libvpx/vpx_dsp/arm/ |
D | fdct_neon.c | 25 int16x4_t input_0 = vshl_n_s16(vld1_s16(input + 0 * stride), 4); in vpx_fdct4x4_neon() 26 int16x4_t input_1 = vshl_n_s16(vld1_s16(input + 1 * stride), 4); in vpx_fdct4x4_neon() 27 int16x4_t input_2 = vshl_n_s16(vld1_s16(input + 2 * stride), 4); in vpx_fdct4x4_neon() 28 int16x4_t input_3 = vshl_n_s16(vld1_s16(input + 3 * stride), 4); in vpx_fdct4x4_neon()
|
D | fdct_partial_neon.c | 32 a0 = vld1_s16(input); in vpx_fdct4x4_1_neon() 34 a1 = vld1_s16(input); in vpx_fdct4x4_1_neon() 36 a2 = vld1_s16(input); in vpx_fdct4x4_1_neon() 38 a3 = vld1_s16(input); in vpx_fdct4x4_1_neon()
|
D | sum_squares_neon.c | 24 s[0] = vld1_s16(src + 0 * stride); in vpx_sum_squares_2d_i16_neon() 25 s[1] = vld1_s16(src + 1 * stride); in vpx_sum_squares_2d_i16_neon() 26 s[2] = vld1_s16(src + 2 * stride); in vpx_sum_squares_2d_i16_neon() 27 s[3] = vld1_s16(src + 3 * stride); in vpx_sum_squares_2d_i16_neon()
|
D | idct16x16_add_neon.c | 600 ind[0] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 602 ind[1] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 604 ind[2] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 606 ind[3] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 608 ind[4] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 610 ind[5] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 612 ind[6] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2() 614 ind[7] = vld1_s16(input); in vpx_idct16x16_10_add_half1d_pass2()
|
/external/libhevc/encoder/arm/ |
D | ihevce_scan_coeffs_neon.c | 193 quant0 = vld1_s16(pi2_temp_quant_coeff + 0 * trans_size); in ihevce_scan_coeffs_neon() 194 quant1 = vld1_s16(pi2_temp_quant_coeff + 1 * trans_size); in ihevce_scan_coeffs_neon() 195 quant2 = vld1_s16(pi2_temp_quant_coeff + 2 * trans_size); in ihevce_scan_coeffs_neon() 196 quant3 = vld1_s16(pi2_temp_quant_coeff + 3 * trans_size); in ihevce_scan_coeffs_neon() 385 quant0 = vld1_s16(pi2_temp_quant_coeff + 0 * trans_size); in ihevce_scan_coeffs_neon() 386 quant1 = vld1_s16(pi2_temp_quant_coeff + 1 * trans_size); in ihevce_scan_coeffs_neon() 387 quant2 = vld1_s16(pi2_temp_quant_coeff + 2 * trans_size); in ihevce_scan_coeffs_neon() 388 quant3 = vld1_s16(pi2_temp_quant_coeff + 3 * trans_size); in ihevce_scan_coeffs_neon()
|
/external/webrtc/modules/audio_coding/codecs/isac/fix/source/ |
D | transform_neon.c | 162 int16x4_t tmpi = vld1_s16(kSinTab1); in PostShiftAndSeparateNeon() 164 int16x4_t tmpr = vld1_s16(kSinTab2); in PostShiftAndSeparateNeon() 166 int16x4_t inre_0 = vld1_s16(inre1); in PostShiftAndSeparateNeon() 168 int16x4_t inre_1 = vld1_s16(inre2); in PostShiftAndSeparateNeon() 170 int16x4_t inim_0 = vld1_s16(inim1); in PostShiftAndSeparateNeon() 172 int16x4_t inim_1 = vld1_s16(inim2); in PostShiftAndSeparateNeon() 268 int16x4_t tmpi = vld1_s16(kSinTab1); in TransformAndFindMaxNeon() 270 int16x4_t tmpr = vld1_s16(kSinTab2); in TransformAndFindMaxNeon() 272 int16x4_t inre_0 = vld1_s16(inre1); in TransformAndFindMaxNeon() 274 int16x4_t inre_1 = vld1_s16(inre2); in TransformAndFindMaxNeon() [all …]
|
D | filters_neon.c | 41 x0_v = vld1_s16(x_start); in WebRtcIsacfix_AutocorrNeon() 86 int16x4_t x2_v = vld1_s16(x_start); in WebRtcIsacfix_AutocorrNeon() 87 int16x4_t y2_v = vld1_s16(y_start); in WebRtcIsacfix_AutocorrNeon()
|
D | pitch_estimator_c.c | 71 int16x4_t int_16x4_x = vld1_s16(&x[n]); in WebRtcIsacfix_PCorr2Q32() 72 int16x4_t int_16x4_in = vld1_s16(&inptr[n]); in WebRtcIsacfix_PCorr2Q32()
|
D | entropy_coding_neon.c | 103 vshll_n_s16(vld1_s16(&matrix0[matrix0_index]), 15); in WebRtcIsacfix_MatrixProduct1Neon() 148 vshll_n_s16(vld1_s16(&matrix0[matrix0_index]), 15); in WebRtcIsacfix_MatrixProduct1Neon()
|
/external/libaom/libaom/av1/common/arm/ |
D | mem_neon.h | 111 *s0 = vld1_s16(s); in load_s16_4x8() 113 *s1 = vld1_s16(s); in load_s16_4x8() 115 *s2 = vld1_s16(s); in load_s16_4x8() 117 *s3 = vld1_s16(s); in load_s16_4x8() 119 *s4 = vld1_s16(s); in load_s16_4x8() 121 *s5 = vld1_s16(s); in load_s16_4x8() 123 *s6 = vld1_s16(s); in load_s16_4x8() 125 *s7 = vld1_s16(s); in load_s16_4x8() 131 *s0 = vld1_s16(s); in load_s16_4x4() 133 *s1 = vld1_s16(s); in load_s16_4x4() [all …]
|
/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | shortidct4x4llm_neon.c | 34 d2 = vld1_s16(input); in vp8_short_idct4x4llm_neon() 35 d3 = vld1_s16(input + 4); in vp8_short_idct4x4llm_neon() 36 d4 = vld1_s16(input + 8); in vp8_short_idct4x4llm_neon() 37 d5 = vld1_s16(input + 12); in vp8_short_idct4x4llm_neon()
|
/external/libvpx/libvpx/vp8/encoder/arm/neon/ |
D | vp8_shortwalsh4x4_neon.c | 36 d0s16 = vld1_s16(input); in vp8_short_walsh4x4_neon() 38 d1s16 = vld1_s16(input); in vp8_short_walsh4x4_neon() 40 d2s16 = vld1_s16(input); in vp8_short_walsh4x4_neon() 42 d3s16 = vld1_s16(input); in vp8_short_walsh4x4_neon()
|
D | shortfdct_neon.c | 33 d0s16 = vld1_s16(input); in vp8_short_fdct4x4_neon() 35 d1s16 = vld1_s16(input); in vp8_short_fdct4x4_neon() 37 d2s16 = vld1_s16(input); in vp8_short_fdct4x4_neon() 39 d3s16 = vld1_s16(input); in vp8_short_fdct4x4_neon()
|
/external/libopus/celt/arm/ |
D | pitch_neon_intr.c | 54 const int16x4_t x_s16x4 = vld1_s16(&x[i]); in celt_inner_prod_neon() 55 const int16x4_t y_s16x4 = vld1_s16(&y[i]); in celt_inner_prod_neon() 97 const int16x4_t x_s16x4 = vld1_s16(&x[i]); in dual_inner_prod_neon() 98 const int16x4_t y01_s16x4 = vld1_s16(&y01[i]); in dual_inner_prod_neon() 99 const int16x4_t y02_s16x4 = vld1_s16(&y02[i]); in dual_inner_prod_neon()
|
/external/libaom/libaom/aom_dsp/arm/ |
D | fwd_txfm_neon.c | 23 *input_0 = vshl_n_s16(vld1_s16(input + 0 * stride), 4); in aom_fdct4x4_helper() 24 *input_1 = vshl_n_s16(vld1_s16(input + 1 * stride), 4); in aom_fdct4x4_helper() 25 *input_2 = vshl_n_s16(vld1_s16(input + 2 * stride), 4); in aom_fdct4x4_helper() 26 *input_3 = vshl_n_s16(vld1_s16(input + 3 * stride), 4); in aom_fdct4x4_helper()
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | intrapred_cfl_neon.cc | 49 const int16x4_t b = vld1_s16(a[y]); in BlockSubtract() 285 const int16x4_t luma_row0 = vld1_s16(luma[y]); in CflIntraPredictor4xN_NEON() 286 const int16x4_t luma_row1 = vld1_s16(luma[y + 1]); in CflIntraPredictor4xN_NEON() 550 ssample = vld1_s16(ssrc); in CflSubsampler444_4xH_NEON() 820 vreinterpret_u16_s16(vld1_s16(luma_ptr - kCflLumaBufferStride)); in CflSubsampler420_4xH_NEON() 833 const int16x4_t samples = vld1_s16(luma_ptr); in CflSubsampler420_4xH_NEON() 1122 const int16x4_t luma_row0 = vld1_s16(luma[y]); in CflIntraPredictor4xN_NEON() 1123 const int16x4_t luma_row1 = vld1_s16(luma[y + 1]); in CflIntraPredictor4xN_NEON()
|
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
D | neon_tensor_utils.cc | 2527 prod.val[0] = vmull_s16(vld1_s16(vector + v), vld1_s16(batch_vector)); in NeonVectorBatchVectorCwiseProductAccumulate() 2529 vmull_s16(vld1_s16(vector + v + 4), vld1_s16(batch_vector + 4)); in NeonVectorBatchVectorCwiseProductAccumulate() 2531 vmull_s16(vld1_s16(vector + v + 8), vld1_s16(batch_vector + 8)); in NeonVectorBatchVectorCwiseProductAccumulate() 2533 vmull_s16(vld1_s16(vector + v + 12), vld1_s16(batch_vector + 12)); in NeonVectorBatchVectorCwiseProductAccumulate() 2539 results.val[0] = vld1_s16(result); in NeonVectorBatchVectorCwiseProductAccumulate() 2540 results.val[1] = vld1_s16(result + 4); in NeonVectorBatchVectorCwiseProductAccumulate() 2541 results.val[2] = vld1_s16(result + 8); in NeonVectorBatchVectorCwiseProductAccumulate() 2542 results.val[3] = vld1_s16(result + 12); in NeonVectorBatchVectorCwiseProductAccumulate()
|
/external/libopus/silk/arm/ |
D | biquad_alt_neon_intr.c | 110 …in_s16x4 = vld1_s16( &in[ 2 * k ] ); /*… in silk_biquad_alt_stride2_neon()
|
/external/speex/ |
D | configure.ac | 106 return vmull_s16(vld1_s16(a), vld1_s16(b));
|