Home
last modified time | relevance | path

Searched refs:LoadUnaligned16 (Results 1 – 16 of 16) sorted by relevance

/external/libgav1/libgav1/src/dsp/x86/
Dintrapred_smooth_sse4.cc168 pixels[1] = LoadUnaligned16(left); in LoadSmoothPixels4()
196 const __m128i y_weights = LoadUnaligned16(weight_array + 12); in LoadSmoothWeights4()
323 pixels[2] = LoadUnaligned16(left); in LoadSmoothPixels8()
325 pixels[2] = LoadUnaligned16(left); in LoadSmoothPixels8()
328 pixels[6] = LoadUnaligned16(left + 16); in LoadSmoothPixels8()
346 __m128i loaded_weights = LoadUnaligned16(&weight_array[offset]); in LoadSmoothWeights8()
364 loaded_weights = LoadUnaligned16(weight_array + 12); in LoadSmoothWeights8()
371 const __m128i weight_lo = LoadUnaligned16(weight_array + 28); in LoadSmoothWeights8()
376 const __m128i weight_hi = LoadUnaligned16(weight_array + 44); in LoadSmoothWeights8()
788 const __m128i weights = LoadUnaligned16(kSmoothWeights + 12); in SmoothHorizontal16x4_SSE4_1()
[all …]
Dintrapred_sse4.cc173 const __m128i vals = LoadUnaligned16(ref); in DcSum16_SSE4_1()
180 const __m128i vals1 = LoadUnaligned16(ref); in DcSum32_SSE4_1()
181 const __m128i vals2 = LoadUnaligned16(static_cast<const uint8_t*>(ref) + 16); in DcSum32_SSE4_1()
191 const __m128i vals1 = LoadUnaligned16(ref_ptr); in DcSum64_SSE4_1()
192 const __m128i vals2 = LoadUnaligned16(ref_ptr + 16); in DcSum64_SSE4_1()
193 const __m128i vals3 = LoadUnaligned16(ref_ptr + 32); in DcSum64_SSE4_1()
194 const __m128i vals4 = LoadUnaligned16(ref_ptr + 48); in DcSum64_SSE4_1()
440 LoadUnaligned16(static_cast<const uint8_t*>(column) + y); in ColStore32_SSE4_1()
469 LoadUnaligned16(static_cast<const uint8_t*>(column) + y); in ColStore64_SSE4_1()
796 const __m128i left = LoadUnaligned16(left_column); in Paeth4x16_SSE4_1()
[all …]
Dintrapred_cfl_sse4.cc43 __m128i ac_q3 = LoadUnaligned16(input); in CflPredictUnclipped()
245 const __m128i samples = LoadUnaligned16(luma_ptr); in CflSubsampler444_8xH_SSE4_1()
313 samples0 = LoadUnaligned16(c_arr); in CflSubsampler444_SSE4_1()
314 samples1 = LoadUnaligned16(c_arr + 8); in CflSubsampler444_SSE4_1()
317 __m128i samples01 = LoadUnaligned16(src); in CflSubsampler444_SSE4_1()
341 samples2 = LoadUnaligned16(c_arr); in CflSubsampler444_SSE4_1()
342 samples3 = LoadUnaligned16(c_arr + 8); in CflSubsampler444_SSE4_1()
345 __m128i samples23 = LoadUnaligned16(src + 16); in CflSubsampler444_SSE4_1()
392 __m128i samples = LoadUnaligned16(&luma_ptr[x]); in CflSubsampler444_SSE4_1()
607 const __m128i final_fill = LoadUnaligned16(luma_ptr - kCflLumaBufferStride); in CflSubsampler420Impl_8xH_SSE4_1()
[all …]
Dcdef_sse4.cc410 const __m128i division_table[2] = {LoadUnaligned16(kCdefDivisionTable), in CdefDirection_SSE4_1()
411 LoadUnaligned16(kCdefDivisionTable + 4)}; in CdefDirection_SSE4_1()
417 LoadUnaligned16(kCdefDivisionTableOddPadded), in CdefDirection_SSE4_1()
418 LoadUnaligned16(kCdefDivisionTableOddPadded + 4)}; in CdefDirection_SSE4_1()
457 output[0] = LoadUnaligned16(src - y_0 * stride - x_0); in LoadDirection()
458 output[1] = LoadUnaligned16(src + y_0 * stride + x_0); in LoadDirection()
459 output[2] = LoadUnaligned16(src - y_1 * stride - x_1); in LoadDirection()
460 output[3] = LoadUnaligned16(src + y_1 * stride + x_1); in LoadDirection()
551 pixel = LoadUnaligned16(src); in CdefFilter_SSE4_1()
Dwarp_sse4.cc112 const __m128i intermediate_0 = LoadUnaligned16(intermediate_result[y + k]); in WriteVerticalFilter()
114 LoadUnaligned16(intermediate_result[y + k + 1]); in WriteVerticalFilter()
179 f = LoadUnaligned16(kWarpedFilters[offset]); in VerticalFilter()
200 f = LoadUnaligned16(kWarpedFilters[offset]); in VerticalFilter()
309 const __m128i src_row_v = LoadUnaligned16(&src_row[ix4 - 7]); in WarpRegion3()
346 const __m128i src_row_v = LoadUnaligned16(&src_row[ix4 - 7]); in WarpRegion4()
Dobmc_sse4.cc161 const __m128i mask_val = LoadUnaligned16(mask + x); in OverlapBlendFromLeft_SSE4_1()
169 const __m128i pred_val = LoadUnaligned16(pred); in OverlapBlendFromLeft_SSE4_1()
170 const __m128i obmc_pred_val = LoadUnaligned16(obmc_pred); in OverlapBlendFromLeft_SSE4_1()
286 const __m128i pred_val = LoadUnaligned16(pred + x); in OverlapBlendFromTop_SSE4_1()
287 const __m128i obmc_pred_val = LoadUnaligned16(obmc_pred + x); in OverlapBlendFromTop_SSE4_1()
Dintra_edge_sse4.cc45 const __m128i edge_lo = LoadUnaligned16(source); in ComputeKernel1Store12()
81 const __m128i edge_lo = LoadUnaligned16(source); in ComputeKernel2Store12()
119 const __m128i edge_lo = LoadUnaligned16(source); in ComputeKernel3Store8()
217 const __m128i data = LoadUnaligned16(temp); in IntraEdgeUpsampler_SSE4_1()
Dinverse_transform_sse4.cc71 x[i] = LoadUnaligned16(&src[i * stride + idx]); in LoadSrc()
72 x[i + 1] = LoadUnaligned16(&src[(i + 1) * stride + idx]); in LoadSrc()
73 x[i + 2] = LoadUnaligned16(&src[(i + 2) * stride + idx]); in LoadSrc()
74 x[i + 3] = LoadUnaligned16(&src[(i + 3) * stride + idx]); in LoadSrc()
288 const __m128i v_src = LoadUnaligned16(&src[i]); in DctDcOnlyColumn()
1070 LoadUnaligned16(kAdst4DcOnlyMultiplier); in Adst4DcOnly()
1606 const __m128i v_src = LoadUnaligned16(&src[i]); in Adst16DcOnlyColumn()
1637 const __m128i v_src = LoadUnaligned16(&src[i * step]); in Identity4_SSE4_1()
1650 const __m128i v_src = LoadUnaligned16(&src[i * step]); in Identity4_SSE4_1()
1717 const __m128i v_src = LoadUnaligned16(&source[row + j]); in Identity4ColumnStoreToFrame()
[all …]
Dcommon_sse4.h138 inline __m128i LoadUnaligned16(const void* a) { in LoadUnaligned16() function
179 return MaskOverreads(LoadUnaligned16(source), over_read_in_bytes); in LoadUnaligned16Msan()
251 return LoadUnaligned16(kMask + n); in MaskHighNBytes()
Dsuper_res_sse4.cc76 LoadUnaligned16(kUpscaleFilter[remainder >> kSuperResExtraBits]); in ComputeSuperRes4()
97 LoadUnaligned16(kUpscaleFilter[remainder >> kSuperResExtraBits]); in ComputeSuperRes8()
Dmask_blend_sse4.cc67 const __m128i row_vals = LoadUnaligned16(mask); in GetMask8()
74 const __m128i next_row_vals = LoadUnaligned16(mask + stride); in GetMask8()
94 const __m128i row_vals = LoadUnaligned16(mask); in GetInterIntraMask8()
101 const __m128i next_row_vals = LoadUnaligned16(mask + stride); in GetInterIntraMask8()
Dloop_filter_sse4.cc1026 __m128i x0 = LoadUnaligned16(dst - 8 + 0 * stride); in Vertical14()
1027 __m128i x1 = LoadUnaligned16(dst - 8 + 1 * stride); in Vertical14()
1028 __m128i x2 = LoadUnaligned16(dst - 8 + 2 * stride); in Vertical14()
1029 __m128i x3 = LoadUnaligned16(dst - 8 + 3 * stride); in Vertical14()
1549 __m128i x0 = LoadUnaligned16(dst - 3 + 0 * stride); in Vertical6()
1550 __m128i x1 = LoadUnaligned16(dst - 3 + 1 * stride); in Vertical6()
1551 __m128i x2 = LoadUnaligned16(dst - 3 + 2 * stride); in Vertical6()
1552 __m128i x3 = LoadUnaligned16(dst - 3 + 3 * stride); in Vertical6()
1799 __m128i x0 = LoadUnaligned16(dst - 4 + 0 * stride); in Vertical8()
1800 __m128i x1 = LoadUnaligned16(dst - 4 + 1 * stride); in Vertical8()
[all …]
Dmotion_field_projection_sse4.cc95 mvs[0] = LoadUnaligned16(mv_int + 0); in GetPosition()
96 mvs[1] = LoadUnaligned16(mv_int + 4); in GetPosition()
201 const __m128i division_table = LoadUnaligned16(projection_divisions); in MotionFieldProjectionKernel_SSE4_1()
Dloop_restoration_sse4.cc142 const __m128i s0 = LoadUnaligned16(src); in WienerHorizontalTap7()
148 const __m128i s1 = LoadUnaligned16(src + x + 16); in WienerHorizontalTap7()
172 const __m128i s0 = LoadUnaligned16(src); in WienerHorizontalTap5()
178 const __m128i s1 = LoadUnaligned16(src + x + 16); in WienerHorizontalTap5()
201 const __m128i s0 = LoadUnaligned16(src); in WienerHorizontalTap3()
207 const __m128i s1 = LoadUnaligned16(src + x + 16); in WienerHorizontalTap3()
228 const __m128i s = LoadUnaligned16(src + x); in WienerHorizontalTap1()
Dconvolve_sse4.cc109 const __m128i src_long = LoadUnaligned16(src); in SumHorizontalTaps()
1442 const __m128i v_src = LoadUnaligned16(&src[x]); in ConvolveCompoundCopy_SSE4()
1780 const __m128i src_vals = LoadUnaligned16(src); in PrepareSourceVectors()
2091 s[i] = LoadUnaligned16(src_y + i * src_stride); in ConvolveVerticalScale()
/external/libgav1/libgav1/src/utils/
Dentropy_decoder.cc329 inline __m128i LoadUnaligned16(const void* a) { in LoadUnaligned16() function
363 __m128i cdf_vec = LoadUnaligned16(cdf); in UpdateCdf7To9()
390 __m128i cdf_vec = LoadUnaligned16(cdf + 2); in UpdateCdf11()
424 __m128i cdf_vec0 = LoadUnaligned16(cdf); in UpdateCdf13()
425 __m128i cdf_vec1 = LoadUnaligned16(cdf + 4); in UpdateCdf13()
455 __m128i cdf_vec0 = LoadUnaligned16(cdf); in UpdateCdf16()
472 __m128i cdf_vec1 = LoadUnaligned16(cdf + 8); in UpdateCdf16()