Home
last modified time | relevance | path

Searched refs:vld1q_u16 (Results 1 – 25 of 53) sorted by relevance

123

/external/libvpx/libvpx/vpx_dsp/arm/
Dhighbd_vpx_convolve_avg_neon.c50 s0 = vld1q_u16(src); in vpx_highbd_convolve_avg_neon()
51 d0 = vld1q_u16(dst); in vpx_highbd_convolve_avg_neon()
53 s1 = vld1q_u16(src); in vpx_highbd_convolve_avg_neon()
54 d1 = vld1q_u16(dst + dst_stride); in vpx_highbd_convolve_avg_neon()
69 s0l = vld1q_u16(src); in vpx_highbd_convolve_avg_neon()
70 s0h = vld1q_u16(src + 8); in vpx_highbd_convolve_avg_neon()
71 d0l = vld1q_u16(dst); in vpx_highbd_convolve_avg_neon()
72 d0h = vld1q_u16(dst + 8); in vpx_highbd_convolve_avg_neon()
74 s1l = vld1q_u16(src); in vpx_highbd_convolve_avg_neon()
75 s1h = vld1q_u16(src + 8); in vpx_highbd_convolve_avg_neon()
[all …]
Dhighbd_intrapred_neon.c83 const uint16x8_t ref_u16 = vld1q_u16(ref); in dc_sum_8()
101 const uint16x8_t above_u16 = vld1q_u16(above); in vpx_highbd_dc_predictor_8x8_neon()
102 const uint16x8_t left_u16 = vld1q_u16(left); in vpx_highbd_dc_predictor_8x8_neon()
292 const uint16x8_t ABCDEFGH = vld1q_u16(above); in vpx_highbd_d45_predictor_4x4_neon()
293 const uint16x8_t BCDEFGH0 = vld1q_u16(above + 1); in vpx_highbd_d45_predictor_4x4_neon()
294 const uint16x8_t CDEFGH00 = vld1q_u16(above + 2); in vpx_highbd_d45_predictor_4x4_neon()
324 const uint16x8_t A0 = vld1q_u16(above); in vpx_highbd_d45_predictor_8x8_neon()
326 const uint16x8_t A1 = vld1q_u16(above + 1); in vpx_highbd_d45_predictor_8x8_neon()
327 const uint16x8_t A2 = vld1q_u16(above + 2); in vpx_highbd_d45_predictor_8x8_neon()
358 const uint16x8_t A0_0 = vld1q_u16(above); in vpx_highbd_d45_predictor_16x16_neon()
[all …]
Dhighbd_idct32x32_add_neon.c21 const uint16x8_t a0 = vld1q_u16(*dest); in highbd_idct32x32_1_add_pos_kernel()
22 const uint16x8_t a1 = vld1q_u16(*dest + 8); in highbd_idct32x32_1_add_pos_kernel()
23 const uint16x8_t a2 = vld1q_u16(*dest + 16); in highbd_idct32x32_1_add_pos_kernel()
24 const uint16x8_t a3 = vld1q_u16(*dest + 24); in highbd_idct32x32_1_add_pos_kernel()
43 const uint16x8_t a0 = vld1q_u16(*dest); in highbd_idct32x32_1_add_neg_kernel()
44 const uint16x8_t a1 = vld1q_u16(*dest + 8); in highbd_idct32x32_1_add_neg_kernel()
45 const uint16x8_t a2 = vld1q_u16(*dest + 16); in highbd_idct32x32_1_add_neg_kernel()
46 const uint16x8_t a3 = vld1q_u16(*dest + 24); in highbd_idct32x32_1_add_neg_kernel()
Dhighbd_vpx_convolve8_neon.c35 *s0 = vld1q_u16(s); in load_8x4()
37 *s1 = vld1q_u16(s); in load_8x4()
39 *s2 = vld1q_u16(s); in load_8x4()
41 *s3 = vld1q_u16(s); in load_8x4()
553 d0 = vrhaddq_u16(d0, vld1q_u16(d + 0 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
554 d1 = vrhaddq_u16(d1, vld1q_u16(d + 1 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
555 d2 = vrhaddq_u16(d2, vld1q_u16(d + 2 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
556 d3 = vrhaddq_u16(d3, vld1q_u16(d + 3 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
557 d4 = vrhaddq_u16(d4, vld1q_u16(d + 4 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
558 d5 = vrhaddq_u16(d5, vld1q_u16(d + 5 * dst_stride)); in vpx_highbd_convolve8_avg_horiz_neon()
[all …]
Dhighbd_loopfilter_neon.c348 *p3 = vld1q_u16(s); in load_8x8()
350 *p2 = vld1q_u16(s); in load_8x8()
352 *p1 = vld1q_u16(s); in load_8x8()
354 *p0 = vld1q_u16(s); in load_8x8()
356 *q0 = vld1q_u16(s); in load_8x8()
358 *q1 = vld1q_u16(s); in load_8x8()
360 *q2 = vld1q_u16(s); in load_8x8()
362 *q3 = vld1q_u16(s); in load_8x8()
372 *s0 = vld1q_u16(s); in load_8x16()
374 *s1 = vld1q_u16(s); in load_8x16()
[all …]
Dhighbd_vpx_convolve_copy_neon.c40 vst1q_u16(dst, vld1q_u16(src)); in vpx_highbd_convolve_copy_neon()
43 vst1q_u16(dst, vld1q_u16(src)); in vpx_highbd_convolve_copy_neon()
Dhighbd_idct_neon.h107 d0 = vld1q_u16(dst); in highbd_add8x8()
109 d1 = vld1q_u16(dst); in highbd_add8x8()
111 d2 = vld1q_u16(dst); in highbd_add8x8()
113 d3 = vld1q_u16(dst); in highbd_add8x8()
115 d4 = vld1q_u16(dst); in highbd_add8x8()
117 d5 = vld1q_u16(dst); in highbd_add8x8()
119 d6 = vld1q_u16(dst); in highbd_add8x8()
121 d7 = vld1q_u16(dst); in highbd_add8x8()
/external/libaom/libaom/aom_dsp/arm/
Dsse_neon.c281 q2 = vld1q_u16(a); in aom_highbd_sse_neon()
282 q3 = vld1q_u16(b); in aom_highbd_sse_neon()
292 q2 = vld1q_u16(a); in aom_highbd_sse_neon()
293 q3 = vld1q_u16(b); in aom_highbd_sse_neon()
297 q2 = vld1q_u16(a + 8); in aom_highbd_sse_neon()
298 q3 = vld1q_u16(b + 8); in aom_highbd_sse_neon()
308 q2 = vld1q_u16(a); in aom_highbd_sse_neon()
309 q3 = vld1q_u16(b); in aom_highbd_sse_neon()
313 q2 = vld1q_u16(a + 8); in aom_highbd_sse_neon()
314 q3 = vld1q_u16(b + 8); in aom_highbd_sse_neon()
[all …]
/external/libgav1/libgav1/src/dsp/arm/
Dintra_edge_neon.cc339 uint16x8_t src_0 = vld1q_u16(dst_buffer); in IntraEdgeFilter_NEON()
346 const uint16x8_t src_1 = vld1q_u16(dst_buffer + i); in IntraEdgeFilter_NEON()
347 const uint16x8_t src_2 = vld1q_u16(dst_buffer + i + 1); in IntraEdgeFilter_NEON()
353 src_0 = vld1q_u16(dst_buffer + i + 7); in IntraEdgeFilter_NEON()
361 const uint16x8_t src_1 = vld1q_u16(dst_buffer + i); in IntraEdgeFilter_NEON()
362 const uint16x8_t src_2 = vld1q_u16(dst_buffer + i + 1); in IntraEdgeFilter_NEON()
366 const uint16x8_t mask = vld1q_u16(kRemainderMask[remainder]); in IntraEdgeFilter_NEON()
400 uint16x8_t src_0 = vld1q_u16(dst_buffer - 1); in IntraEdgeFilter_NEON()
401 uint16x8_t src_1 = vld1q_u16(dst_buffer); in IntraEdgeFilter_NEON()
407 const uint16x8_t src_2 = vld1q_u16(dst_buffer + i); in IntraEdgeFilter_NEON()
[all …]
Dintrapred_cfl_neon.cc602 samples = vld1q_u16(src); in CflSubsampler444_8xH_NEON()
682 samples[0] = vld1q_u16(src); in CflSubsampler444_WxH_NEON()
684 (max_luma_width >= 16) ? vld1q_u16(src + 8) : LastRowResult(samples[0]); in CflSubsampler444_WxH_NEON()
687 samples[2] = (max_luma_width >= 24) ? vld1q_u16(src + 16) in CflSubsampler444_WxH_NEON()
689 samples[3] = (max_luma_width == 32) ? vld1q_u16(src + 24) in CflSubsampler444_WxH_NEON()
786 const uint16x8_t samples_row0 = vld1q_u16(src); in CflSubsampler420_4xH_NEON()
788 const uint16x8_t samples_row1 = vld1q_u16(src); in CflSubsampler420_4xH_NEON()
792 const uint16x8_t samples_row2 = vld1q_u16(src); in CflSubsampler420_4xH_NEON()
794 const uint16x8_t samples_row3 = vld1q_u16(src); in CflSubsampler420_4xH_NEON()
800 const uint16x8_t samples_row4 = vld1q_u16(src); in CflSubsampler420_4xH_NEON()
[all …]
Dloop_restoration_neon.cc1167 s5[0] = vld1q_u16(sum5[0]); in BoxFilterPreProcess5Lo()
1168 s5[1] = vld1q_u16(sum5[1]); in BoxFilterPreProcess5Lo()
1169 s5[2] = vld1q_u16(sum5[2]); in BoxFilterPreProcess5Lo()
1200 s5[0][0] = vld1q_u16(sum5[0] + x); in BoxFilterPreProcess5()
1201 s5[0][1] = vld1q_u16(sum5[1] + x); in BoxFilterPreProcess5()
1202 s5[0][2] = vld1q_u16(sum5[2] + x); in BoxFilterPreProcess5()
1221 s5[1][0] = vld1q_u16(sum5[0] + x + 8); in BoxFilterPreProcess5()
1222 s5[1][1] = vld1q_u16(sum5[1] + x + 8); in BoxFilterPreProcess5()
1223 s5[1][2] = vld1q_u16(sum5[2] + x + 8); in BoxFilterPreProcess5()
1244 s5[0] = vld1q_u16(sum5[0]); in BoxFilterPreProcess5LastRowLo()
[all …]
Dintrapred_neon.cc741 const uint16x8_t val_0 = vld1q_u16(buf); in LoadAndAdd16()
742 const uint16x8_t val_1 = vld1q_u16(buf + 8); in LoadAndAdd16()
748 const uint16x8_t val_0 = vld1q_u16(buf); in LoadAndAdd32()
749 const uint16x8_t val_1 = vld1q_u16(buf + 8); in LoadAndAdd32()
750 const uint16x8_t val_2 = vld1q_u16(buf + 16); in LoadAndAdd32()
751 const uint16x8_t val_3 = vld1q_u16(buf + 24); in LoadAndAdd32()
757 const uint16x8_t val_0 = vld1q_u16(buf); in LoadAndAdd64()
758 const uint16x8_t val_1 = vld1q_u16(buf + 8); in LoadAndAdd64()
759 const uint16x8_t val_2 = vld1q_u16(buf + 16); in LoadAndAdd64()
760 const uint16x8_t val_3 = vld1q_u16(buf + 24); in LoadAndAdd64()
[all …]
Dcdef_neon.cc427 output[0] = vld1q_u16(src + y_0 * stride + x_0); in LoadDirection()
428 output[1] = vld1q_u16(src - y_0 * stride - x_0); in LoadDirection()
429 output[2] = vld1q_u16(src + y_1 * stride + x_1); in LoadDirection()
430 output[3] = vld1q_u16(src - y_1 * stride - x_1); in LoadDirection()
512 pixel = vld1q_u16(src); in CdefFilter_NEON()
Daverage_blend_neon.cc146 const uint16x8_t pred0 = vld1q_u16(prediction_0); in AverageBlend8Row()
147 const uint16x8_t pred1 = vld1q_u16(prediction_1); in AverageBlend8Row()
/external/libjpeg-turbo/simd/arm/
Djquanti-neon.c110 uint16x8_t recip0 = vld1q_u16(recip_ptr + (i + 0) * DCTSIZE); in jsimd_quantize_neon()
111 uint16x8_t recip1 = vld1q_u16(recip_ptr + (i + 1) * DCTSIZE); in jsimd_quantize_neon()
112 uint16x8_t recip2 = vld1q_u16(recip_ptr + (i + 2) * DCTSIZE); in jsimd_quantize_neon()
113 uint16x8_t recip3 = vld1q_u16(recip_ptr + (i + 3) * DCTSIZE); in jsimd_quantize_neon()
114 uint16x8_t corr0 = vld1q_u16(corr_ptr + (i + 0) * DCTSIZE); in jsimd_quantize_neon()
115 uint16x8_t corr1 = vld1q_u16(corr_ptr + (i + 1) * DCTSIZE); in jsimd_quantize_neon()
116 uint16x8_t corr2 = vld1q_u16(corr_ptr + (i + 2) * DCTSIZE); in jsimd_quantize_neon()
117 uint16x8_t corr3 = vld1q_u16(corr_ptr + (i + 3) * DCTSIZE); in jsimd_quantize_neon()
/external/zlib/contrib/optimizations/
Dslide_hash_neon.h28 uint16x8_t m_low = vld1q_u16(hash); in neon_slide_hash_update()
29 uint16x8_t m_high = vld1q_u16(hash + chunk); in neon_slide_hash_update()
/external/libhevc/encoder/arm/
Dihevce_coarse_layer_sad_neon.c493 uint16x8_t mvy_wt = vld1q_u16((U16 *)&gi2_mvy_range[ABS(mvy)][0]); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon()
505 uint16x8_t curr = vld1q_u16((U16 *)pi2_sads_4x4_current + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon()
506 uint16x8_t south = vld1q_u16((U16 *)pi2_sads_4x4_south + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon()
507 uint16x8_t east = vld1q_u16((U16 *)pi2_sads_4x4_east + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon()
512 vld1q_u16((U16 *)&gi2_mvx_range[mvx + MAX_MVX_SUPPORTED_IN_COARSE_LAYER][0]); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon()
654 uint16x8_t mvy_wt = vld1q_u16((U16 *)&gi2_mvy_range[ABS(mvy)][0]); in hme_combine_4x4_sads_and_compute_cost_high_quality_neon()
666 uint16x8_t curr = vld1q_u16((U16 *)pi2_sads_4x4_current + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_quality_neon()
667 uint16x8_t south = vld1q_u16((U16 *)pi2_sads_4x4_south + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_quality_neon()
668 uint16x8_t east = vld1q_u16((U16 *)pi2_sads_4x4_east + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_quality_neon()
672 uint16x8_t mv_wt = vld1q_u16( in hme_combine_4x4_sads_and_compute_cost_high_quality_neon()
/external/webp/src/dsp/
Dyuv_neon.c192 const int16x8_t A = vreinterpretq_s16_u16(vld1q_u16(ref + i)); in SharpYUVUpdateY_NEON()
193 const int16x8_t B = vreinterpretq_s16_u16(vld1q_u16(src + i)); in SharpYUVUpdateY_NEON()
194 const int16x8_t C = vreinterpretq_s16_u16(vld1q_u16(dst + i)); in SharpYUVUpdateY_NEON()
252 const int16x8_t g0 = vreinterpretq_s16_u16(vld1q_u16(best_y + 2 * i + 0)); in SharpYUVFilterRow_NEON()
253 const int16x8_t g1 = vreinterpretq_s16_u16(vld1q_u16(best_y + 2 * i + 8)); in SharpYUVFilterRow_NEON()
/external/libaom/libaom/av1/common/arm/
Dcfl_neon.c20 vsubq_s16(vreinterpretq_s16_u16(vld1q_u16(src + offset)), sub)); in vldsubstq_s16()
24 return vaddq_u16(vld1q_u16(buf), vld1q_u16(buf + offset)); in vldaddq_u16()
156 const uint16x8_t top = vld1q_u16(input); in cfl_luma_subsampling_420_hbd_neon()
157 const uint16x8_t bot = vld1q_u16(input + input_stride); in cfl_luma_subsampling_420_hbd_neon()
163 const uint16x8_t top_1 = vld1q_u16(input + 8); in cfl_luma_subsampling_420_hbd_neon()
164 const uint16x8_t bot_1 = vld1q_u16(input + 8 + input_stride); in cfl_luma_subsampling_420_hbd_neon()
233 const uint16x8_t top = vld1q_u16(input); in cfl_luma_subsampling_444_hbd_neon()
Dreconinter_neon.c50 diff_q = vabdq_u16(vld1q_u16(src0_1), vld1q_u16(src1_1)); in av1_build_compound_diffwtd_mask_d16_neon()
Dselfguided_neon.c930 xtl = vld1q_u16(buf - buf_stride - 1); in cross_sum_inp_u16()
931 xt = vld1q_u16(buf - buf_stride); in cross_sum_inp_u16()
932 xtr = vld1q_u16(buf - buf_stride + 1); in cross_sum_inp_u16()
933 xl = vld1q_u16(buf - 1); in cross_sum_inp_u16()
934 x = vld1q_u16(buf); in cross_sum_inp_u16()
935 xr = vld1q_u16(buf + 1); in cross_sum_inp_u16()
936 xbl = vld1q_u16(buf + buf_stride - 1); in cross_sum_inp_u16()
937 xb = vld1q_u16(buf + buf_stride); in cross_sum_inp_u16()
938 xbr = vld1q_u16(buf + buf_stride + 1); in cross_sum_inp_u16()
983 xtl = vld1q_u16(buf - buf_stride - 1); in cross_sum_fast_even_row_inp16()
[all …]
/external/libvpx/libvpx/vp8/encoder/arm/neon/
Dfastquantizeb_neon.c27 const uint16x8_t zig_zag0 = vld1q_u16(inv_zig_zag), in vp8_fast_quantize_b_neon()
28 zig_zag1 = vld1q_u16(inv_zig_zag + 8); in vp8_fast_quantize_b_neon()
/external/webrtc/modules/audio_processing/aecm/
Daecm_core_neon.cc68 spectrum_v = vld1q_u16(far_spectrum_p); in WebRtcAecm_CalcLinearEnergiesNeon()
145 far_spectrum_v = vld1q_u16(far_spectrum_p); in WebRtcAecm_StoreAdaptiveChannelNeon()
/external/libgav1/libgav1/src/utils/
Dentropy_decoder.cc200 uint16x8_t cdf_vec = vld1q_u16(cdf); in UpdateCdf7To9()
233 uint16x8_t cdf_vec = vld1q_u16(cdf + 2); in UpdateCdf11()
270 uint16x8_t cdf_vec0 = vld1q_u16(cdf); in UpdateCdf13()
271 uint16x8_t cdf_vec1 = vld1q_u16(cdf + 4); in UpdateCdf13()
303 uint16x8_t cdf_vec = vld1q_u16(cdf); in UpdateCdf16()
320 cdf_vec = vld1q_u16(cdf + 8); in UpdateCdf16()
/external/llvm-project/clang/test/CodeGen/arm-mve-intrinsics/
Dcplusplus.cpp108 return vld1q_u16(base); in test_vld1q_u16()

123