| /third_party/python/Modules/_blake2/impl/ |
| D | blake2s-round.h | 44 #define G1(row1,row2,row3,row4,buf) \ argument
45 row1 = _mm_add_epi32( _mm_add_epi32( row1, buf), row2 ); \
49 row2 = _mm_xor_si128( row2, row3 ); \
50 row2 = _mm_roti_epi32(row2, -12);
52 #define G2(row1,row2,row3,row4,buf) \ argument
53 row1 = _mm_add_epi32( _mm_add_epi32( row1, buf), row2 ); \
57 row2 = _mm_xor_si128( row2, row3 ); \
58 row2 = _mm_roti_epi32(row2, -7);
60 #define DIAGONALIZE(row1,row2,row3,row4) \ argument
63 row2 = _mm_shuffle_epi32( row2, _MM_SHUFFLE(0,3,2,1) );
[all …]
|
| /third_party/flutter/skia/third_party/externals/sdl/src/render/ |
| D | SDL_yuv_sw.c | 111 unsigned short *row2; in Color16DitherYV12Mod1X() local
120 row2 = row1 + cols + mod; in Color16DitherYV12Mod1X()
152 *row2++ = (unsigned short) (rgb_2_pix[L + cr_r] | in Color16DitherYV12Mod1X()
157 *row2++ = (unsigned short) (rgb_2_pix[L + cr_r] | in Color16DitherYV12Mod1X()
170 row2 += mod; in Color16DitherYV12Mod1X()
182 unsigned char *row2; in Color24DitherYV12Mod1X() local
191 row2 = row1 + cols * 3 + mod * 3; in Color24DitherYV12Mod1X()
230 *row2++ = (value) & 0xFF; in Color24DitherYV12Mod1X()
231 *row2++ = (value >> 8) & 0xFF; in Color24DitherYV12Mod1X()
232 *row2++ = (value >> 16) & 0xFF; in Color24DitherYV12Mod1X()
[all …]
|
| D | SDL_yuv_mmx.c | 85 Uint32 *row2; in ColorRGBDitherYV12MMX1X() local
90 row2 = (Uint32 *)out+cols+mod; /* start of second row */ in ColorRGBDitherYV12MMX1X()
235 "r"(row1),"r"(cols),"r"(row2),"m"(x),"m"(y),"m"(mod), in ColorRGBDitherYV12MMX1X()
248 Uint16 *row2; in Color565DitherYV12MMX1X() local
253 row2 = (Uint16 *)out+cols+mod; /* start of second row */ in Color565DitherYV12MMX1X()
420 "r"(row1),"r"(cols),"r"(row2),"m"(x),"m"(y),"m"(mod), in Color565DitherYV12MMX1X()
|
| /third_party/libjpeg-turbo/simd/powerpc/ |
| D | jquanti-altivec.c | 119 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_quantize_altivec() local
138 row2 = vec_ld(32, workspace); in jsimd_quantize_altivec()
148 row2s = vec_sra(row2, pw_word_bit_m1); in jsimd_quantize_altivec()
156 row2 = vec_xor(row2, row2s); in jsimd_quantize_altivec()
164 row2 = vec_sub(row2, row2s); in jsimd_quantize_altivec()
182 row2 = vec_add(row2, corr2); in jsimd_quantize_altivec()
200 MULTIPLY(row2, recip2, row2); in jsimd_quantize_altivec()
218 MULTIPLY(row2, scale2, row2); in jsimd_quantize_altivec()
227 row2 = vec_xor(row2, row2s); in jsimd_quantize_altivec()
235 row2 = vec_sub(row2, row2s); in jsimd_quantize_altivec()
[all …]
|
| D | jfdctfst-altivec.c | 92 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_fdct_ifast_altivec() local
111 row2 = vec_ld(32, data); in jsimd_fdct_ifast_altivec()
139 tmp2 = vec_add(row2, row5); in jsimd_fdct_ifast_altivec()
140 tmp5 = vec_sub(row2, row5); in jsimd_fdct_ifast_altivec()
|
| D | jfdctint-altivec.c | 184 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_fdct_islow_altivec() local
215 row2 = vec_ld(32, data); in jsimd_fdct_islow_altivec()
243 tmp2 = vec_add(row2, row5); in jsimd_fdct_islow_altivec()
244 tmp5 = vec_sub(row2, row5); in jsimd_fdct_islow_altivec()
|
| /third_party/flutter/skia/third_party/externals/libjpeg-turbo/simd/powerpc/ |
| D | jquanti-altivec.c | 119 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_quantize_altivec() local
138 row2 = vec_ld(32, workspace); in jsimd_quantize_altivec()
148 row2s = vec_sra(row2, pw_word_bit_m1); in jsimd_quantize_altivec()
156 row2 = vec_xor(row2, row2s); in jsimd_quantize_altivec()
164 row2 = vec_sub(row2, row2s); in jsimd_quantize_altivec()
182 row2 = vec_add(row2, corr2); in jsimd_quantize_altivec()
200 MULTIPLY(row2, recip2, row2); in jsimd_quantize_altivec()
218 MULTIPLY(row2, scale2, row2); in jsimd_quantize_altivec()
227 row2 = vec_xor(row2, row2s); in jsimd_quantize_altivec()
235 row2 = vec_sub(row2, row2s); in jsimd_quantize_altivec()
[all …]
|
| D | jfdctfst-altivec.c | 92 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_fdct_ifast_altivec() local
111 row2 = vec_ld(32, data); in jsimd_fdct_ifast_altivec()
139 tmp2 = vec_add(row2, row5); in jsimd_fdct_ifast_altivec()
140 tmp5 = vec_sub(row2, row5); in jsimd_fdct_ifast_altivec()
|
| D | jfdctint-altivec.c | 184 __vector short row0, row1, row2, row3, row4, row5, row6, row7, in jsimd_fdct_islow_altivec() local
215 row2 = vec_ld(32, data); in jsimd_fdct_islow_altivec()
243 tmp2 = vec_add(row2, row5); in jsimd_fdct_islow_altivec()
244 tmp5 = vec_sub(row2, row5); in jsimd_fdct_islow_altivec()
|
| /third_party/skia/third_party/externals/libjpeg-turbo/simd/arm/ |
| D | jquanti-neon.c | 61 int16x8_t row2 = in jsimd_convsamp_neon() local
76 vst1q_s16(workspace + 2 * DCTSIZE, row2); in jsimd_convsamp_neon()
107 int16x8_t row2 = vld1q_s16(workspace + (i + 2) * DCTSIZE); in jsimd_quantize_neon() local
126 int16x8_t sign_row2 = vshrq_n_s16(row2, 15); in jsimd_quantize_neon()
131 uint16x8_t abs_row2 = vreinterpretq_u16_s16(vabsq_s16(row2)); in jsimd_quantize_neon()
159 row2 = vcombine_s16(vshrn_n_s32(row2_l, 16), vshrn_n_s32(row2_h, 16)); in jsimd_quantize_neon()
169 row2 = vreinterpretq_s16_u16(vshlq_u16(vreinterpretq_u16_s16(row2), in jsimd_quantize_neon()
179 row2 = veorq_s16(row2, sign_row2); in jsimd_quantize_neon()
180 row2 = vsubq_s16(row2, sign_row2); in jsimd_quantize_neon()
187 vst1q_s16(out_ptr + (i + 2) * DCTSIZE, row2); in jsimd_quantize_neon()
|
| D | jidctint-neon.c | 99 int16x4_t row2,
118 int16x4_t row2,
203 int16x4_t row2 = vld1_s16(coef_block + 2 * DCTSIZE); in jsimd_idct_islow_neon() local
228 bitmap = vorr_s16(bitmap, row2); in jsimd_idct_islow_neon()
239 jsimd_idct_islow_pass1_sparse(row0, row1, row2, row3, quant_row0, in jsimd_idct_islow_neon()
244 jsimd_idct_islow_pass1_regular(row0, row1, row2, row3, row4, row5, in jsimd_idct_islow_neon()
256 row2 = vld1_s16(coef_block + 2 * DCTSIZE + 4); in jsimd_idct_islow_neon()
279 bitmap = vorr_s16(bitmap, row2); in jsimd_idct_islow_neon()
299 jsimd_idct_islow_pass1_sparse(row0, row1, row2, row3, quant_row0, in jsimd_idct_islow_neon()
304 jsimd_idct_islow_pass1_regular(row0, row1, row2, row3, row4, row5, in jsimd_idct_islow_neon()
[all …]
|
| D | jidctred-neon.c | 204 int16x8_t row2 = vld1q_s16(coef_block + 2 * DCTSIZE); in jsimd_idct_4x4_neon() local
216 int16x8_t bitmap = vorrq_s16(row1, row2); in jsimd_idct_4x4_neon()
243 row2 = dcval; in jsimd_idct_4x4_neon()
264 int16x4_t z2 = vmul_s16(vget_high_s16(row2), quant_row2); in jsimd_idct_4x4_neon()
296 row2 = vcombine_s16(dcval, vrshrn_n_s32(vsubq_s32(tmp12, tmp0), in jsimd_idct_4x4_neon()
317 int16x4_t z2 = vmul_s16(vget_low_s16(row2), quant_row2); in jsimd_idct_4x4_neon()
349 row2 = vcombine_s16(vrshrn_n_s32(vsubq_s32(tmp12, tmp0), in jsimd_idct_4x4_neon()
364 int16x8_t z2 = vmulq_s16(row2, quant_row2); in jsimd_idct_4x4_neon()
414 row2 = vcombine_s16(vrshrn_n_s32(vsubq_s32(tmp12_l, tmp0_l), in jsimd_idct_4x4_neon()
422 int16x8x2_t row_23 = vtrnq_s16(row2, row3); in jsimd_idct_4x4_neon()
|
| D | jfdctfst-neon.c | 154 int16x8_t row2 = vreinterpretq_s16_s32(rows_26.val[0]); in jsimd_fdct_ifast_neon() local
167 tmp2 = vaddq_s16(row2, row5); in jsimd_fdct_ifast_neon()
168 tmp5 = vsubq_s16(row2, row5); in jsimd_fdct_ifast_neon()
182 row2 = vaddq_s16(tmp13, z1); /* phase 5 */ in jsimd_fdct_ifast_neon()
208 vst1q_s16(data + 2 * DCTSIZE, row2); in jsimd_fdct_ifast_neon()
|
| D | jidctfst-neon.c | 71 int16x8_t row2 = vld1q_s16(coef_block + 2 * DCTSIZE); in jsimd_idct_ifast_neon() local
84 int16x8_t bitmap = vorrq_s16(row1, row2); in jsimd_idct_ifast_neon()
103 row2 = dcval; in jsimd_idct_ifast_neon()
128 int16x4_t tmp1 = vmul_s16(vget_high_s16(row2), quant_row2); in jsimd_idct_ifast_neon()
180 row2 = vcombine_s16(dcval, vadd_s16(tmp2, tmp5)); in jsimd_idct_ifast_neon()
203 int16x4_t tmp1 = vmul_s16(vget_low_s16(row2), quant_row2); in jsimd_idct_ifast_neon()
255 row2 = vcombine_s16(vadd_s16(tmp2, tmp5), dcval); in jsimd_idct_ifast_neon()
273 int16x8_t tmp1 = vmulq_s16(row2, quant_row2); in jsimd_idct_ifast_neon()
325 row2 = vaddq_s16(tmp2, tmp5); in jsimd_idct_ifast_neon()
333 int16x8x2_t rows_23 = vtrnq_s16(row2, row3); in jsimd_idct_ifast_neon()
|
| /third_party/skia/third_party/externals/libjpeg-turbo/simd/arm/aarch32/ |
| D | jchuff-neon.c | 70 int16x8_t row2 = vld1q_dup_s16(block + 12); in jsimd_huff_encode_one_block_neon() local
71 row2 = vld1q_lane_s16(block + 19, row2, 1); in jsimd_huff_encode_one_block_neon()
72 row2 = vld1q_lane_s16(block + 26, row2, 2); in jsimd_huff_encode_one_block_neon()
73 row2 = vld1q_lane_s16(block + 33, row2, 3); in jsimd_huff_encode_one_block_neon()
74 row2 = vld1q_lane_s16(block + 40, row2, 4); in jsimd_huff_encode_one_block_neon()
75 row2 = vld1q_lane_s16(block + 48, row2, 5); in jsimd_huff_encode_one_block_neon()
76 row2 = vld1q_lane_s16(block + 41, row2, 6); in jsimd_huff_encode_one_block_neon()
77 row2 = vld1q_lane_s16(block + 34, row2, 7); in jsimd_huff_encode_one_block_neon()
90 int16x8_t abs_row2 = vabsq_s16(row2); in jsimd_huff_encode_one_block_neon()
120 vshlq_u16(vreinterpretq_u16_s16(vshrq_n_s16(row2, 15)), in jsimd_huff_encode_one_block_neon()
|
| /third_party/skia/third_party/externals/swiftshader/src/Pipeline/ |
| D | ShaderCore.cpp | 472 void transpose4x4(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3) in transpose4x4() argument
475 Int2 tmp1 = UnpackHigh(row2, row3); in transpose4x4()
477 Int2 tmp3 = UnpackLow(row2, row3); in transpose4x4()
481 row2 = UnpackLow(tmp0, tmp1); in transpose4x4()
485 void transpose4x3(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3) in transpose4x3() argument
488 Int2 tmp1 = UnpackHigh(row2, row3); in transpose4x3()
490 Int2 tmp3 = UnpackLow(row2, row3); in transpose4x3()
494 row2 = UnpackLow(tmp0, tmp1); in transpose4x3()
497 void transpose4x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3) in transpose4x4() argument
500 Float4 tmp1 = UnpackLow(row2, row3); in transpose4x4()
[all …]
|
| D | ShaderCore.hpp | 214 void transpose4x4(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3);
215 void transpose4x3(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3);
216 void transpose4x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
217 void transpose4x3(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
218 void transpose4x2(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
219 void transpose4x1(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
220 void transpose2x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
221 void transpose4xN(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3, int N);
|
| /third_party/mindspore/tests/ut/cpp/dataset/ |
| D | c_api_dataset_tfrecord_test.cc | 147 std::unordered_map<std::string, mindspore::MSTensor> row2; in TEST_F() local
148 ASSERT_OK(iter2->GetNextRow(&row2)); in TEST_F()
153 while (row1.size() != 0 && row2.size() != 0) { in TEST_F()
159 auto scalars2 = row2["scalars"]; in TEST_F()
167 ASSERT_OK(iter2->GetNextRow(&row2)); in TEST_F()
427 std::unordered_map<std::string, mindspore::MSTensor> row2; in TEST_F() local
428 ASSERT_OK(iter2->GetNextRow(&row2)); in TEST_F()
437 while (row2.size() != 0) { in TEST_F()
439 ASSERT_OK(iter2->GetNextRow(&row2)); in TEST_F()
|
| /third_party/skia/third_party/externals/libjpeg-turbo/simd/arm/aarch64/ |
| D | jchuff-neon.c | 125 int16x8_t row2 = in jsimd_huff_encode_one_block_neon() local
144 row2 = in jsimd_huff_encode_one_block_neon()
146 4), row2, 0); in jsimd_huff_encode_one_block_neon()
147 row2 = in jsimd_huff_encode_one_block_neon()
149 0), row2, 5); in jsimd_huff_encode_one_block_neon()
163 int16x8_t abs_row2 = vabsq_s16(row2); in jsimd_huff_encode_one_block_neon()
176 vreinterpretq_u16_s16(veorq_s16(abs_row2, vshrq_n_s16(row2, 15))); in jsimd_huff_encode_one_block_neon()
|
| /third_party/ffmpeg/libavcodec/ |
| D | a64multienc.c | 104 uint8_t row1, row2; in render_charset() local
139 row1 = 0; row2 = 0; in render_charset()
152 row2 <<= 2; in render_charset()
159 row2 |= 3-(index2[pix] & 3); in render_charset()
161 row2 |= 3-(index1[pix] & 3); in render_charset()
171 if (INTERLACED) charset[y+0x800] = row2; in render_charset()
|
| D | ivi_dsp.c | 669 int i, row2, row4, row8; in ff_ivi_col_slant8() local
672 row2 = pitch << 1; in ff_ivi_col_slant8()
680 out[0], out[pitch], out[row2], out[row2 + pitch], out[row4], in ff_ivi_col_slant8()
681 out[row4 + pitch], out[row4 + row2], out[row8 - pitch], in ff_ivi_col_slant8()
684 out[0] = out[pitch] = out[row2] = out[row2 + pitch] = out[row4] = in ff_ivi_col_slant8()
685 out[row4 + pitch] = out[row4 + row2] = out[row8 - pitch] = 0; in ff_ivi_col_slant8()
730 int i, row2; in ff_ivi_col_slant4() local
733 row2 = pitch << 1; in ff_ivi_col_slant4()
739 out[0], out[pitch], out[row2], out[row2 + pitch], in ff_ivi_col_slant4()
742 out[0] = out[pitch] = out[row2] = out[row2 + pitch] = 0; in ff_ivi_col_slant4()
|
| /third_party/ffmpeg/libavcodec/mips/ |
| D | vp8_lpf_msa.c | 380 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in ff_vp8_h_loop_filter16_msa() local
388 LD_UB8(temp_src, pitch, row0, row1, row2, row3, row4, row5, row6, row7); in ff_vp8_h_loop_filter16_msa()
392 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in ff_vp8_h_loop_filter16_msa()
445 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in ff_vp8_h_loop_filter8uv_msa() local
453 LD_UB8(src_u - 4, pitch, row0, row1, row2, row3, row4, row5, row6, row7); in ff_vp8_h_loop_filter8uv_msa()
456 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in ff_vp8_h_loop_filter8uv_msa()
525 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in ff_vp8_h_loop_filter_simple_msa() local
531 LD_UB8(temp_src, pitch, row0, row1, row2, row3, row4, row5, row6, row7); in ff_vp8_h_loop_filter_simple_msa()
535 TRANSPOSE16x4_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in ff_vp8_h_loop_filter_simple_msa()
595 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in ff_vp8_h_loop_filter8uv_inner_msa() local
[all …]
|
| /third_party/skia/tools/ |
| D | image_diff_metric.cpp | 46 const uint8_t* row2 = reinterpret_cast<const uint8_t*>(bm[1].pixmap().addr32(0, y)); in main() local
48 totalDiffs += std::abs((int)row1[i] - (int)row2[i]); in main()
|
| /third_party/mesa3d/src/gallium/drivers/swr/rasterizer/core/ |
| D | format_utils.h | 34 void vTranspose(simd4scalar& row0, simd4scalar& row1, simd4scalar& row2, simd4scalar& row3) in vTranspose() argument
38 simd4scalari row2i = SIMD128::castps_si(row2); in vTranspose()
59 row2 = SIMD128::castsi_ps(row2i); in vTranspose()
64 void vTranspose(simd4scalari& row0, simd4scalari& row1, simd4scalari& row2, simd4scalari& row3) in vTranspose() argument
66 simd4scalari vTemp = row2; in vTranspose()
67 row2 = SIMD128::unpacklo_epi32(row2, row3); in vTranspose()
75 row0 = SIMD128::unpacklo_epi64(row0, row2); in vTranspose()
76 row1 = SIMD128::unpackhi_epi64(row1, row2); in vTranspose()
78 row2 = row3; in vTranspose()
79 row2 = SIMD128::unpacklo_epi64(row2, vTemp); in vTranspose()
|
| /third_party/skia/third_party/externals/swiftshader/src/Shader/ |
| D | ShaderCore.hpp | 87 void transpose4x4(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3);
88 void transpose4x3(Short4 &row0, Short4 &row1, Short4 &row2, Short4 &row3);
89 void transpose4x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
90 void transpose4x3(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
91 void transpose4x2(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
92 void transpose4x1(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
93 void transpose2x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3);
94 void transpose4xN(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3, int N);
|