Home
last modified time | relevance | path

Searched refs:dst4 (Results 1 – 25 of 55) sorted by relevance

123

/third_party/flutter/skia/src/core/
DSkMatrix44.cpp767 typedef void (*Map2Procf)(const SkMScalar mat[][4], const float src2[], int count, float dst4[]);
768 typedef void (*Map2Procd)(const SkMScalar mat[][4], const double src2[], int count, double dst4[]);
771 int count, float* SK_RESTRICT dst4) { in map2_if() argument
773 dst4[0] = src2[0]; in map2_if()
774 dst4[1] = src2[1]; in map2_if()
775 dst4[2] = 0; in map2_if()
776 dst4[3] = 1; in map2_if()
778 dst4 += 4; in map2_if()
783 int count, double* SK_RESTRICT dst4) { in map2_id() argument
785 dst4[0] = src2[0]; in map2_id()
[all …]
DSkBlitRow_D32.cpp63 auto dst4 = ( __m128i*)dst; in blit_row_s32_blend() local
66 _mm_storeu_si128(dst4, SkPMLerp_SSE2(_mm_loadu_si128(src4), in blit_row_s32_blend()
67 _mm_loadu_si128(dst4), in blit_row_s32_blend()
70 dst4++; in blit_row_s32_blend()
75 dst = ( SkPMColor*)dst4; in blit_row_s32_blend()
128 auto dst4 = ( __m128i*)dst; in blit_row_s32a_blend() local
131 _mm_storeu_si128(dst4, SkBlendARGB32_SSE2(_mm_loadu_si128(src4), in blit_row_s32a_blend()
132 _mm_loadu_si128(dst4), in blit_row_s32a_blend()
135 dst4++; in blit_row_s32a_blend()
140 dst = ( SkPMColor*)dst4; in blit_row_s32a_blend()
DSk4px.h117 dst4 = fn(Load4(src+4)); in MapSrc() local
119 dst4.store4(dst+4); in MapSrc()
147 dst4 = fn(Load4(dst+4), Load4(src+4)); in MapDstSrc() local
149 dst4.store4(dst+4); in MapDstSrc()
177 dst4 = fn(Load4(dst+4), Load4Alphas(a+4)); in MapDstAlpha() local
179 dst4.store4(dst+4); in MapDstAlpha()
209 dst4 = fn(Load4(dst+4), Load4(src+4), Load4Alphas(a+4)); in MapDstSrcAlpha() local
211 dst4.store4(dst+4); in MapDstSrcAlpha()
/third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/intrinsics/sse/
Dsse_common.h88 static inline void WriteCol1(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol1() argument
107 static inline void WriteCol2(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol2() argument
132 static inline void WriteCol2Opt(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4 in WriteCol2Opt() argument
159 static inline void WriteCol3(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol3() argument
189 static inline void WriteCol4(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol4() argument
208 static inline void WriteCol5(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol5() argument
215 _mm_store_ss(*dst + 4, *dst4); in WriteCol5()
231 static inline void WriteCol6(float **dst, __m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, _… in WriteCol6() argument
240 _mm_store_ss(*dst + 4, *dst4); in WriteCol6()
241 *dst4 = _mm_shuffle_ps(*dst4, *dst4, _MM_SHUFFLE(0, 3, 2, 1)); in WriteCol6()
[all …]
DMatMul_Sse.c41 __m128 dst4 = _mm_setzero_ps(); in MatrixMultiplyWinograd() local
57 dst4 = _mm_add_ps(dst4, tmp4); in MatrixMultiplyWinograd()
64 _mm_storeu_ps(matrix_c + 12, dst4); in MatrixMultiplyWinograd()
117 …__m128 dst1 = _mm_setzero_ps(), dst2 = _mm_setzero_ps(), dst3 = _mm_setzero_ps(), dst4 = _mm_setze… in MatmulFloatSse64Opt() local
127 dst3 = _mm_add_ps(dst3, tmp3), dst4 = _mm_add_ps(dst4, tmp4); in MatmulFloatSse64Opt()
136 DoBiasBlock8(bias_d, &dst1, &dst2, &dst3, &dst4, &dst5, &dst6, &dst7, &dst8); in MatmulFloatSse64Opt()
140 ActBlock8(&dst1, &dst2, &dst3, &dst4, &dst5, &dst6, &dst7, &dst8, act_type); in MatmulFloatSse64Opt()
146 _mm_storeu_ps(dst, dst3), _mm_storeu_ps(dst + 4, dst4); in MatmulFloatSse64Opt()
153 _mm_storeu_ps(c + 8, dst3), _mm_storeu_ps(c + 12, dst4); in MatmulFloatSse64Opt()
161 WriteCol1(&dst, &dst1, &dst2, &dst3, &dst4, &dst5, &dst6, &dst7, &dst8, stride, 1, r); in MatmulFloatSse64Opt()
[all …]
DTiledC4MatMulFp32.c20 …atic inline void TiledC4MatmulFp32_Transfer(__m128 *dst1, __m128 *dst2, __m128 *dst3, __m128 *dst4, in TiledC4MatmulFp32_Transfer() argument
26 *dst4 = _mm_add_ps(*dst4, _mm_mul_ps(weight, _mm_set_ps1(v4))); in TiledC4MatmulFp32_Transfer()
57 __m128 dst4 = _mm_mul_ps(weight_data[0], _mm_set_ps1(MS_F32X4_GETI(src4, 0))); in TiledC4MatmulFp32() local
59 TiledC4MatmulFp32_Transfer(&dst1, &dst2, &dst3, &dst4, weight_data[j], MS_F32X4_GETI(src1, j), in TiledC4MatmulFp32()
84 dst4 = _mm_add_ps(dst4, _mm_mul_ps(weight_data[0], _mm_set_ps1(MS_F32X4_GETI(src4, 0)))); in TiledC4MatmulFp32()
86 … TiledC4MatmulFp32_Transfer(&dst1, &dst2, &dst3, &dst4, weight_data[1], MS_F32X4_GETI(src1, 1), in TiledC4MatmulFp32()
93 … TiledC4MatmulFp32_Transfer(&dst1, &dst2, &dst3, &dst4, weight_data[2], MS_F32X4_GETI(src1, 2), in TiledC4MatmulFp32()
101 dst4 = _mm_add_ps(dst4, _mm_mul_ps(weight_data[3], _mm_set_ps1(MS_F32X4_GETI(src4, 3)))); in TiledC4MatmulFp32()
128 dst4 = _mm_add_ps(dst4, _mm_mul_ps(weight_data[0], _mm_set_ps1(MS_F32X4_GETI(src4, 0)))); in TiledC4MatmulFp32()
130 TiledC4MatmulFp32_Transfer(&dst1, &dst2, &dst3, &dst4, weight_data[1], MS_F32X4_GETI(src1, 1), in TiledC4MatmulFp32()
[all …]
DConvDwFp32Row_sse.c34 __m128 dst4 = _mm_loadu_ps(output_ptr + 12); in ConvDwFp32Row() local
46 dst4 = MS_MLAQ_F32(dst4, w4, in4); in ConvDwFp32Row()
50 _mm_storeu_ps(output_ptr + 12, dst4); in ConvDwFp32Row()
/third_party/skia/src/core/
DSkBlitRow_D32.cpp62 auto dst4 = ( __m128i*)dst; in blit_row_s32_blend() local
65 _mm_storeu_si128(dst4, SkPMLerp_SSE2(_mm_loadu_si128(src4), in blit_row_s32_blend()
66 _mm_loadu_si128(dst4), in blit_row_s32_blend()
69 dst4++; in blit_row_s32_blend()
74 dst = ( SkPMColor*)dst4; in blit_row_s32_blend()
127 auto dst4 = ( __m128i*)dst; in blit_row_s32a_blend() local
130 _mm_storeu_si128(dst4, SkBlendARGB32_SSE2(_mm_loadu_si128(src4), in blit_row_s32a_blend()
131 _mm_loadu_si128(dst4), in blit_row_s32a_blend()
134 dst4++; in blit_row_s32a_blend()
139 dst = ( SkPMColor*)dst4; in blit_row_s32a_blend()
DSk4px.h117 dst4 = fn(Load4(src+4)); in MapSrc() local
119 dst4.store4(dst+4); in MapSrc()
147 dst4 = fn(Load4(dst+4), Load4(src+4)); in MapDstSrc() local
149 dst4.store4(dst+4); in MapDstSrc()
177 dst4 = fn(Load4(dst+4), Load4Alphas(a+4)); in MapDstAlpha() local
179 dst4.store4(dst+4); in MapDstAlpha()
209 dst4 = fn(Load4(dst+4), Load4(src+4), Load4Alphas(a+4)); in MapDstSrcAlpha() local
211 dst4.store4(dst+4); in MapDstSrcAlpha()
/third_party/ffmpeg/libavcodec/mips/
Dhevc_mc_uniw_msa.c138 v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7, offset_vec; in hevc_uniwgt_copy_6w_msa() local
158 ILVRL_B2_SH(zero, src2, dst4, dst5); in hevc_uniwgt_copy_6w_msa()
162 SLLI_4V(dst4, dst5, dst6, dst7, 6); in hevc_uniwgt_copy_6w_msa()
167 HEVC_UNIW_RND_CLIP4_MAX_SATU_H(dst4, dst5, dst6, dst7, weight_vec, in hevc_uniwgt_copy_6w_msa()
168 offset_vec, rnd_vec, dst4, dst5, dst6, in hevc_uniwgt_copy_6w_msa()
171 PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3); in hevc_uniwgt_copy_6w_msa()
200 v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7, offset_vec; in hevc_uniwgt_copy_8w_msa() local
238 ILVRL_B2_SH(zero, src2, dst4, dst5); in hevc_uniwgt_copy_8w_msa()
240 SLLI_2V(dst4, dst5, 6); in hevc_uniwgt_copy_8w_msa()
244 HEVC_UNIW_RND_CLIP2_MAX_SATU_H(dst4, dst5, weight_vec, offset_vec, in hevc_uniwgt_copy_8w_msa()
[all …]
Dhevc_mc_bi_msa.c143 v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in hevc_bi_copy_6w_msa() local
158 ILVRL_B2_SH(zero, src2, dst4, dst5); in hevc_bi_copy_6w_msa()
161 SLLI_4V(dst4, dst5, dst6, dst7, 6); in hevc_bi_copy_6w_msa()
164 HEVC_BI_RND_CLIP4_MAX_SATU(in4, in5, in6, in7, dst4, dst5, dst6, dst7, in hevc_bi_copy_6w_msa()
165 7, dst4, dst5, dst6, dst7); in hevc_bi_copy_6w_msa()
167 PCKEV_B2_UB(dst5, dst4, dst7, dst6, out2, out3); in hevc_bi_copy_6w_msa()
194 v8i16 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in hevc_bi_copy_8w_msa() local
226 ILVRL_B2_SH(zero, src2, dst4, dst5); in hevc_bi_copy_8w_msa()
229 SLLI_2V(dst4, dst5, 6); in hevc_bi_copy_8w_msa()
232 HEVC_BI_RND_CLIP2_MAX_SATU(in4, in5, dst4, dst5, 7, dst4, dst5); in hevc_bi_copy_8w_msa()
[all …]
Dhevcdsp_msa.c567 v8i16 filt0, filt1, filt2, filt3, dst0, dst1, dst2, dst3, dst4, dst5; in hevc_hz_8t_12w_msa() local
596 dst4 = const_vec; in hevc_hz_8t_12w_msa()
603 DPADD_SB2_SH(vec4, vec5, filt0, filt0, dst4, dst5); in hevc_hz_8t_12w_msa()
609 DPADD_SB2_SH(vec4, vec5, filt1, filt1, dst4, dst5); in hevc_hz_8t_12w_msa()
615 DPADD_SB2_SH(vec4, vec5, filt2, filt2, dst4, dst5); in hevc_hz_8t_12w_msa()
621 DPADD_SB2_SH(vec4, vec5, filt3, filt3, dst4, dst5); in hevc_hz_8t_12w_msa()
623 res0 = __msa_copy_s_d((v2i64) dst4, 0); in hevc_hz_8t_12w_msa()
624 res1 = __msa_copy_s_d((v2i64) dst4, 1); in hevc_hz_8t_12w_msa()
699 v8i16 dst0, dst1, dst2, dst3, dst4, dst5; in hevc_hz_8t_24w_msa() local
729 dst4 = const_vec; in hevc_hz_8t_24w_msa()
[all …]
Dvc1dsp_msa.c145 v4i32 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in ff_vc1_inv_trans_4x8_msa() local
204 LD_SW8(dest, linesize, dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); in ff_vc1_inv_trans_4x8_msa()
206 zero_m, dst4, zero_m, dst5, zero_m, dst6, zero_m, dst7, in ff_vc1_inv_trans_4x8_msa()
207 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); in ff_vc1_inv_trans_4x8_msa()
210 ILVR_H4_SW(zero_m, dst4, zero_m, dst5, zero_m, dst6, zero_m, dst7, in ff_vc1_inv_trans_4x8_msa()
211 dst4, dst5, dst6, dst7); in ff_vc1_inv_trans_4x8_msa()
222 ADD4(in_r4, dst4, in_r5, dst5, in_r6, dst6, in_r7, dst7, in ff_vc1_inv_trans_4x8_msa()
235 v4i32 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in ff_vc1_inv_trans_8x4_msa() local
283 dst4, dst5, dst6, dst7); in ff_vc1_inv_trans_8x4_msa()
306 ADD4(in4, dst4, in5, dst5, in6, dst6, in7, dst7, in4, in5, in6, in7); in ff_vc1_inv_trans_8x4_msa()
Dhevc_mc_biw_msa.c244 v8i16 dst0, dst1, dst2, dst3, dst4, dst5; in hevc_biwgt_copy_8w_msa() local
277 ILVRL_B2_SH(zero, src2, dst4, dst5); in hevc_biwgt_copy_8w_msa()
280 SLLI_2V(dst4, dst5, 6); in hevc_biwgt_copy_8w_msa()
284 HEVC_BIW_RND_CLIP2_MAX_SATU(dst4, dst5, in4, in5, weight_vec, rnd_vec, in hevc_biwgt_copy_8w_msa()
285 offset_vec, dst4, dst5); in hevc_biwgt_copy_8w_msa()
286 PCKEV_B3_UB(dst1, dst0, dst3, dst2, dst5, dst4, out0, out1, out2); in hevc_biwgt_copy_8w_msa()
332 v8i16 dst0, dst1, dst2, dst3, dst4, dst5; in hevc_biwgt_copy_12w_msa() local
356 ILVR_B2_SH(zero, src0, zero, src1, dst4, dst5); in hevc_biwgt_copy_12w_msa()
358 dst4 <<= 6; in hevc_biwgt_copy_12w_msa()
363 HEVC_BIW_RND_CLIP2_MAX_SATU(dst4, dst5, in4, in5, weight_vec, rnd_vec, in hevc_biwgt_copy_12w_msa()
[all …]
Dhevc_lpf_sao_msa.c43 v16u8 dst0, dst1, dst2, dst3, dst4, dst5; in hevc_loopfilter_luma_hor_msa() local
175 dst4 = (v16u8) (temp2 + (v8i16) q1_src); in hevc_loopfilter_luma_hor_msa()
184 dst4 = __msa_bmz_v(dst4, (v16u8) q1_src, (v16u8) q_is_pcm_vec); in hevc_loopfilter_luma_hor_msa()
189 dst2 = (v16u8) __msa_pckev_b((v16i8) dst5, (v16i8) dst4); in hevc_loopfilter_luma_hor_msa()
192 PCKEV_B2_UB(p1_src, p2_src, q0_src, p0_src, dst3, dst4); in hevc_loopfilter_luma_hor_msa()
196 dst1 = __msa_bmz_v(dst1, dst4, (v16u8) cmp3); in hevc_loopfilter_luma_hor_msa()
277 dst4 = (v16u8) __msa_bmz_v((v16u8) delta2, (v16u8) q1_src, in hevc_loopfilter_luma_hor_msa()
280 PCKEV_B2_UB(dst2, dst1, dst4, dst3, dst0, dst1); in hevc_loopfilter_luma_hor_msa()
333 dst4 = (v16u8) (temp2 + (v8i16) q1_src); in hevc_loopfilter_luma_hor_msa()
342 dst4 = __msa_bmz_v(dst4, (v16u8) q1_src, (v16u8) q_is_pcm_vec); in hevc_loopfilter_luma_hor_msa()
[all …]
Dhevc_idct_msa.c773 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in hevc_addblk_16x16_msa() local
778 LD_UB4(temp_dst, stride, dst4, dst5, dst6, dst7); in hevc_addblk_16x16_msa()
785 UNPCK_UB_SH(dst4, dst_r0, dst_l0); in hevc_addblk_16x16_msa()
800 LD_UB4(temp_dst, stride, dst4, dst5, dst6, dst7); in hevc_addblk_16x16_msa()
815 UNPCK_UB_SH(dst4, dst_r0, dst_l0); in hevc_addblk_16x16_msa()
840 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in hevc_addblk_32x32_msa() local
845 LD_UB2(temp_dst, 16, dst4, dst5); in hevc_addblk_32x32_msa()
854 UNPCK_UB_SH(dst4, dst_r0, dst_l0); in hevc_addblk_32x32_msa()
869 LD_UB2(temp_dst, 16, dst4, dst5); in hevc_addblk_32x32_msa()
887 UNPCK_UB_SH(dst4, dst_r0, dst_l0); in hevc_addblk_32x32_msa()
[all …]
Dh264idct_msa.c122 v16i8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in avc_idct8_addblk_msa() local
227 LD_SB8(dst, dst_stride, dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); in avc_idct8_addblk_msa()
230 ILVR_B4_SH(zeros, dst4, zeros, dst5, zeros, dst6, zeros, dst7, in avc_idct8_addblk_msa()
246 v16i8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; in avc_idct8_dc_addblk_msa() local
256 LD_SB8(dst, dst_stride, dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); in avc_idct8_dc_addblk_msa()
259 ILVR_B4_SH(zeros, dst4, zeros, dst5, zeros, dst6, zeros, dst7, in avc_idct8_dc_addblk_msa()
/third_party/mesa3d/src/mesa/main/
Dimage.c506 GLfloat (*dst4)[4] = (GLfloat (*)[4]) (useTemp ? tempBuffer : dst); in _mesa_convert_colors() local
511 dst4[i][RCOMP] = UBYTE_TO_FLOAT(src1[i][RCOMP]); in _mesa_convert_colors()
512 dst4[i][GCOMP] = UBYTE_TO_FLOAT(src1[i][GCOMP]); in _mesa_convert_colors()
513 dst4[i][BCOMP] = UBYTE_TO_FLOAT(src1[i][BCOMP]); in _mesa_convert_colors()
514 dst4[i][ACOMP] = UBYTE_TO_FLOAT(src1[i][ACOMP]); in _mesa_convert_colors()
539 GLfloat (*dst4)[4] = (GLfloat (*)[4]) (useTemp ? tempBuffer : dst); in _mesa_convert_colors() local
544 dst4[i][RCOMP] = USHORT_TO_FLOAT(src2[i][RCOMP]); in _mesa_convert_colors()
545 dst4[i][GCOMP] = USHORT_TO_FLOAT(src2[i][GCOMP]); in _mesa_convert_colors()
546 dst4[i][BCOMP] = USHORT_TO_FLOAT(src2[i][BCOMP]); in _mesa_convert_colors()
547 dst4[i][ACOMP] = USHORT_TO_FLOAT(src2[i][ACOMP]); in _mesa_convert_colors()
/third_party/mesa3d/src/gallium/drivers/softpipe/
Dsp_quad_blend.c157 uint *dst4 = (uint *) dst; in logicop_quad() local
184 res4[j] = ~(src4[j] | dst4[j]); in logicop_quad()
188 res4[j] = ~src4[j] & dst4[j]; in logicop_quad()
196 res4[j] = src4[j] & ~dst4[j]; in logicop_quad()
200 res4[j] = ~dst4[j]; in logicop_quad()
204 res4[j] = dst4[j] ^ src4[j]; in logicop_quad()
208 res4[j] = ~(src4[j] & dst4[j]); in logicop_quad()
212 res4[j] = src4[j] & dst4[j]; in logicop_quad()
216 res4[j] = ~(src4[j] ^ dst4[j]); in logicop_quad()
220 res4[j] = dst4[j]; in logicop_quad()
[all …]
/third_party/skia/third_party/externals/dng_sdk/source/
Ddng_image.cpp382 dng_rect dst4 (dng_rect (splitV, in GetRepeat() local
387 if (dst4.NotEmpty ()) in GetRepeat()
392 temp.fArea = dst4 + (srcArea.TL () - in GetRepeat()
396 temp.fData = buffer.DirtyPixel (dst4.t, in GetRepeat()
397 dst4.l, in GetRepeat()
/third_party/flutter/skia/gm/
Dpoly2poly.cpp113 const int dst4[] = { 0, 0, 96, 0, 64, 96, 0, 64 }; in onDraw() local
114 doDraw(canvas, font, &paint, src4, dst4, 4); in onDraw()
/third_party/skia/gm/
Dpoly2poly.cpp113 const int dst4[] = { 0, 0, 96, 0, 64, 96, 0, 64 }; in onDraw() local
114 doDraw(canvas, font, &paint, src4, dst4, 4); in onDraw()
/third_party/ffmpeg/libavcodec/aarch64/
Dvp9mc_16bpp_neon.S123 .macro extmlal dst1, dst2, dst3, dst4, dst5, dst6, dst7, dst8, src1, src2, src3, src4, src5, src6, …
139 smlal2 \dst4\().4s, v21.8h, v0.h[\offset]
411 .macro convolve8 dst1, dst2, dst3, dst4, src1, src2, src3, src4, src5, src6, src7, src8, src9
415 smull2 \dst4\().4s, \src2\().8h, v0.h[0]
419 smlal2 \dst4\().4s, \src3\().8h, v0.h[1]
423 smlal2 \dst4\().4s, \src4\().8h, v0.h[2]
427 smlal2 \dst4\().4s, \src5\().8h, v0.h[3]
431 smlal2 \dst4\().4s, \src6\().8h, v0.h[4]
435 smlal2 \dst4\().4s, \src7\().8h, v0.h[5]
439 smlal2 \dst4\().4s, \src8\().8h, v0.h[6]
[all …]
/third_party/flutter/skia/samplecode/
DSamplePolyToPoly.cpp144 const int dst4[] = { 0, 0, 96, 0, 64, 96, 0, 64 }; in onDrawContent() local
145 doDraw(canvas, &paint, font, src4, dst4, 4); in onDrawContent()
/third_party/skia/samplecode/
DSamplePolyToPoly.cpp142 const int dst4[] = { 0, 0, 96, 0, 64, 96, 0, 64 }; in onDrawContent() local
143 doDraw(canvas, &paint, font, src4, dst4, 4); in onDrawContent()

123