/third_party/python/Modules/_blake2/impl/ |
D | blake2s-round.h | 44 #define G1(row1,row2,row3,row4,buf) \ argument 48 row3 = _mm_add_epi32( row3, row4 ); \ 49 row2 = _mm_xor_si128( row2, row3 ); \ 52 #define G2(row1,row2,row3,row4,buf) \ argument 56 row3 = _mm_add_epi32( row3, row4 ); \ 57 row2 = _mm_xor_si128( row2, row3 ); \ 60 #define DIAGONALIZE(row1,row2,row3,row4) \ argument 62 row3 = _mm_shuffle_epi32( row3, _MM_SHUFFLE(1,0,3,2) ); \ 65 #define UNDIAGONALIZE(row1,row2,row3,row4) \ argument 67 row3 = _mm_shuffle_epi32( row3, _MM_SHUFFLE(1,0,3,2) ); \ [all …]
|
D | blake2s.c | 268 __m128i row1, row2, row3, row4; in blake2s_compress() local 306 row3 = _mm_setr_epi32( 0x6A09E667, 0xBB67AE85, 0x3C6EF372, 0xA54FF53A ); in blake2s_compress() 318 STOREU( &S->h[0], _mm_xor_si128( ff0, _mm_xor_si128( row1, row3 ) ) ); in blake2s_compress()
|
/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 139 row3 = vec_ld(48, workspace); in jsimd_quantize_altivec() 149 row3s = vec_sra(row3, pw_word_bit_m1); in jsimd_quantize_altivec() 157 row3 = vec_xor(row3, row3s); in jsimd_quantize_altivec() 165 row3 = vec_sub(row3, row3s); in jsimd_quantize_altivec() 183 row3 = vec_add(row3, corr3); in jsimd_quantize_altivec() 201 MULTIPLY(row3, recip3, row3); in jsimd_quantize_altivec() 219 MULTIPLY(row3, scale3, row3); in jsimd_quantize_altivec() 228 row3 = vec_xor(row3, row3s); in jsimd_quantize_altivec() 236 row3 = vec_sub(row3, row3s); 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 112 row3 = vec_ld(48, data); in jsimd_fdct_ifast_altivec() 141 tmp3 = vec_add(row3, row4); in jsimd_fdct_ifast_altivec() 142 tmp4 = vec_sub(row3, row4); 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 216 row3 = vec_ld(48, data); in jsimd_fdct_islow_altivec() 245 tmp3 = vec_add(row3, row4); in jsimd_fdct_islow_altivec() 246 tmp4 = vec_sub(row3, row4); 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 139 row3 = vec_ld(48, workspace); in jsimd_quantize_altivec() 149 row3s = vec_sra(row3, pw_word_bit_m1); in jsimd_quantize_altivec() 157 row3 = vec_xor(row3, row3s); in jsimd_quantize_altivec() 165 row3 = vec_sub(row3, row3s); in jsimd_quantize_altivec() 183 row3 = vec_add(row3, corr3); in jsimd_quantize_altivec() 201 MULTIPLY(row3, recip3, row3); in jsimd_quantize_altivec() 219 MULTIPLY(row3, scale3, row3); in jsimd_quantize_altivec() 228 row3 = vec_xor(row3, row3s); in jsimd_quantize_altivec() 236 row3 = vec_sub(row3, row3s); 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 112 row3 = vec_ld(48, data); in jsimd_fdct_ifast_altivec() 141 tmp3 = vec_add(row3, row4); in jsimd_fdct_ifast_altivec() 142 tmp4 = vec_sub(row3, row4); 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 216 row3 = vec_ld(48, data); in jsimd_fdct_islow_altivec() 245 tmp3 = vec_add(row3, row4); in jsimd_fdct_islow_altivec() 246 tmp4 = vec_sub(row3, row4); in jsimd_fdct_islow_altivec()
|
/third_party/skia/third_party/externals/libjpeg-turbo/simd/arm/ |
D | jquanti-neon.c | 63 int16x8_t row3 = in jsimd_convsamp_neon() local 77 vst1q_s16(workspace + 3 * DCTSIZE, row3); in jsimd_convsamp_neon() 108 int16x8_t row3 = vld1q_s16(workspace + (i + 3) * DCTSIZE); in jsimd_quantize_neon() local 127 int16x8_t sign_row3 = vshrq_n_s16(row3, 15); in jsimd_quantize_neon() 132 uint16x8_t abs_row3 = vreinterpretq_u16_s16(vabsq_s16(row3)); in jsimd_quantize_neon() 160 row3 = vcombine_s16(vshrn_n_s32(row3_l, 16), vshrn_n_s32(row3_h, 16)); in jsimd_quantize_neon() 171 row3 = vreinterpretq_s16_u16(vshlq_u16(vreinterpretq_u16_s16(row3), in jsimd_quantize_neon() 181 row3 = veorq_s16(row3, sign_row3); in jsimd_quantize_neon() 182 row3 = vsubq_s16(row3, sign_row3); in jsimd_quantize_neon() 188 vst1q_s16(out_ptr + (i + 3) * DCTSIZE, row3); in jsimd_quantize_neon()
|
D | jidctred-neon.c | 84 int16x8_t row3 = vld1q_s16(coef_block + 3 * DCTSIZE); in jsimd_idct_2x2_neon() local 98 row3 = vmulq_s16(row3, quant_row3); in jsimd_idct_2x2_neon() 115 tmp0_l = vmlal_lane_s16(tmp0_l, vget_low_s16(row3), consts, 2); in jsimd_idct_2x2_neon() 119 tmp0_h = vmlal_lane_s16(tmp0_h, vget_high_s16(row3), consts, 2); in jsimd_idct_2x2_neon() 205 int16x8_t row3 = vld1q_s16(coef_block + 3 * DCTSIZE); in jsimd_idct_4x4_neon() local 217 bitmap = vorrq_s16(bitmap, row3); in jsimd_idct_4x4_neon() 244 row3 = dcval; in jsimd_idct_4x4_neon() 276 z3 = vmul_s16(vget_high_s16(row3), quant_row3); in jsimd_idct_4x4_neon() 292 row3 = vcombine_s16(dcval, vrshrn_n_s32(vsubq_s32(tmp10, tmp2), in jsimd_idct_4x4_neon() 329 z3 = vmul_s16(vget_low_s16(row3), quant_row3); in jsimd_idct_4x4_neon() [all …]
|
D | jidctint-neon.c | 100 int16x4_t row3, 119 int16x4_t row3, 204 int16x4_t row3 = vld1_s16(coef_block + 3 * DCTSIZE); in jsimd_idct_islow_neon() local 227 bitmap = vorr_s16(bitmap, row3); 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() 257 row3 = vld1_s16(coef_block + 3 * DCTSIZE + 4); in jsimd_idct_islow_neon() 278 bitmap = vorr_s16(bitmap, row3); 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 | jfdctfst-neon.c | 155 int16x8_t row3 = vreinterpretq_s16_s32(rows_37.val[0]); in jsimd_fdct_ifast_neon() local 169 tmp3 = vaddq_s16(row3, row4); in jsimd_fdct_ifast_neon() 170 tmp4 = vsubq_s16(row3, row4); in jsimd_fdct_ifast_neon() 202 row3 = vsubq_s16(z13, z2); in jsimd_fdct_ifast_neon() 209 vst1q_s16(data + 3 * DCTSIZE, row3); in jsimd_fdct_ifast_neon()
|
D | jidctfst-neon.c | 72 int16x8_t row3 = vld1q_s16(coef_block + 3 * DCTSIZE); in jsimd_idct_ifast_neon() local 85 bitmap = vorrq_s16(bitmap, row3); in jsimd_idct_ifast_neon() 104 row3 = dcval; in jsimd_idct_ifast_neon() 148 int16x4_t tmp5 = vmul_s16(vget_high_s16(row3), quant_row3); in jsimd_idct_ifast_neon() 183 row3 = vcombine_s16(dcval, vsub_s16(tmp3, tmp4)); in jsimd_idct_ifast_neon() 223 int16x4_t tmp5 = vmul_s16(vget_low_s16(row3), quant_row3); in jsimd_idct_ifast_neon() 258 row3 = vcombine_s16(vsub_s16(tmp3, tmp4), dcval); in jsimd_idct_ifast_neon() 293 int16x8_t tmp5 = vmulq_s16(row3, quant_row3); in jsimd_idct_ifast_neon() 328 row3 = vsubq_s16(tmp3, tmp4); in jsimd_idct_ifast_neon() 333 int16x8x2_t rows_23 = vtrnq_s16(row2, row3); in jsimd_idct_ifast_neon()
|
D | jfdctint-neon.c | 254 int16x8_t row3 = vreinterpretq_s16_s32(rows_37.val[0]); in jsimd_fdct_islow_neon() local 268 tmp3 = vaddq_s16(row3, row4); in jsimd_fdct_islow_neon() 269 tmp4 = vsubq_s16(row3, row4); in jsimd_fdct_islow_neon() 358 row3 = vcombine_s16(vrshrn_n_s32(tmp6_l, DESCALE_P2), in jsimd_fdct_islow_neon() 371 vst1q_s16(data + 3 * DCTSIZE, row3); in jsimd_fdct_islow_neon()
|
/third_party/skia/third_party/externals/libjpeg-turbo/simd/arm/aarch32/ |
D | jchuff-neon.c | 79 int16x8_t row3 = vld1q_dup_s16(block + 27); in jsimd_huff_encode_one_block_neon() local 80 row3 = vld1q_lane_s16(block + 20, row3, 1); in jsimd_huff_encode_one_block_neon() 81 row3 = vld1q_lane_s16(block + 13, row3, 2); in jsimd_huff_encode_one_block_neon() 82 row3 = vld1q_lane_s16(block + 6, row3, 3); in jsimd_huff_encode_one_block_neon() 83 row3 = vld1q_lane_s16(block + 7, row3, 4); in jsimd_huff_encode_one_block_neon() 84 row3 = vld1q_lane_s16(block + 14, row3, 5); in jsimd_huff_encode_one_block_neon() 85 row3 = vld1q_lane_s16(block + 21, row3, 6); in jsimd_huff_encode_one_block_neon() 86 row3 = vld1q_lane_s16(block + 28, row3, 7); in jsimd_huff_encode_one_block_neon() 91 int16x8_t abs_row3 = vabsq_s16(row3); in jsimd_huff_encode_one_block_neon() 123 vshlq_u16(vreinterpretq_u16_s16(vshrq_n_s16(row3, 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() 482 row3 = UnpackHigh(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() 497 void transpose4x4(Float4 &row0, Float4 &row1, Float4 &row2, Float4 &row3) in transpose4x4() argument 500 Float4 tmp1 = UnpackLow(row2, row3); in transpose4x4() 502 Float4 tmp3 = UnpackHigh(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/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/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 39 simd4scalari row3i = SIMD128::castps_si(row3); in vTranspose() 60 row3 = SIMD128::castsi_ps(row3i); in vTranspose() 64 void vTranspose(simd4scalari& row0, simd4scalari& row1, simd4scalari& row2, simd4scalari& row3) in vTranspose() argument 67 row2 = SIMD128::unpacklo_epi32(row2, row3); in vTranspose() 68 vTemp = SIMD128::unpackhi_epi32(vTemp, row3); in vTranspose() 70 row3 = row0; in vTranspose() 72 row3 = SIMD128::unpackhi_epi32(row3, row1); in vTranspose() 78 row2 = row3; in vTranspose() 80 row3 = SIMD128::unpackhi_epi64(row3, vTemp); in vTranspose()
|
/third_party/flutter/skia/gm/ |
D | hardstop_gradients.cpp | 117 SkScalar row3[] = {0.00f, 0.25f, 1.00f}; in onDraw() local 127 row3, in onDraw()
|
/third_party/skia/gm/ |
D | hardstop_gradients.cpp | 117 SkScalar row3[] = {0.00f, 0.25f, 1.00f}; in onDraw() local 127 row3, in onDraw()
|
/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);
|
/third_party/skia/third_party/externals/d3d12allocator/src/ |
D | Common.h | 186 const vec4& row3, in mat4() 190 _31(row3.x), _32(row3.y), _33(row3.z), _34(row3.w), in mat4()
|
/third_party/flutter/skia/third_party/externals/libwebp/src/dsp/ |
D | dec_msa.c | 364 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter16() local 371 LD_UB8(ptmp, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilter16() 374 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilter16() 446 v16u8 row0, row1, row2, row3, row4, row5, row6, row7; in HFilterVertEdge16i() local 453 LD_UB8(src - 4, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilterVertEdge16i() 456 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilterVertEdge16i() 530 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter8() local 537 LD_UB8(ptmp_src_u, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilter8() 540 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilter8() 595 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter8i() local [all …]
|
/third_party/skia/third_party/externals/libwebp/src/dsp/ |
D | dec_msa.c | 364 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter16() local 371 LD_UB8(ptmp, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilter16() 374 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilter16() 446 v16u8 row0, row1, row2, row3, row4, row5, row6, row7; in HFilterVertEdge16i() local 453 LD_UB8(src - 4, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilterVertEdge16i() 456 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilterVertEdge16i() 530 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter8() local 537 LD_UB8(ptmp_src_u, stride, row0, row1, row2, row3, row4, row5, row6, row7); in HFilter8() 540 TRANSPOSE16x8_UB_UB(row0, row1, row2, row3, row4, row5, row6, row7, in HFilter8() 595 v16u8 row0, row1, row2, row3, row4, row5, row6, row7, row8; in HFilter8i() local [all …]
|