| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/intrinsics/avx/ |
| D | DeconvMatMulAvx.c | 24 __m256 res7 = _mm256_setzero_ps(); in Deconv4X8AvxKernel() local
37 res7 = _mm256_fmadd_ps(tmp2, w0, res7); in Deconv4X8AvxKernel()
43 _mm256_storeu_ps(dst + C16NUM, res7); in Deconv4X8AvxKernel()
52 __m256 res7 = _mm256_setzero_ps(); in Deconv4X16AvxKernel() local
70 res7 = _mm256_fmadd_ps(tmp2, w0, res7); in Deconv4X16AvxKernel()
78 _mm256_storeu_ps(dst + C16NUM, res7); in Deconv4X16AvxKernel()
94 __m256 res7 = _mm256_setzero_ps(); in Deconv4X24AvxKernel() local
114 res7 = _mm256_fmadd_ps(tmp, w0, res7); in Deconv4X24AvxKernel()
127 _mm256_storeu_ps(dst + C16NUM, res7); in Deconv4X24AvxKernel()
|
| /third_party/ffmpeg/libavcodec/mips/ |
| D | h264idct_msa.c | 116 v8i16 res0, res1, res2, res3, res4, res5, res6, res7; in avc_idct8_addblk_msa() local
157 res0, res1, res2, res3, res4, res5, res6, res7); in avc_idct8_addblk_msa()
158 TRANSPOSE8x8_SH_SH(res0, res1, res2, res3, res4, res5, res6, res7, in avc_idct8_addblk_msa()
159 res0, res1, res2, res3, res4, res5, res6, res7); in avc_idct8_addblk_msa()
167 UNPCK_SH_SW(res7, tmp7_r, tmp7_l); in avc_idct8_addblk_msa()
226 res4, res5, res6, res7); in avc_idct8_addblk_msa()
234 ADD4(res4, tmp4, res5, tmp5, res6, tmp6, res7, tmp7, in avc_idct8_addblk_msa()
235 res4, res5, res6, res7); in avc_idct8_addblk_msa()
236 CLIP_SH8_0_255(res0, res1, res2, res3, res4, res5, res6, res7); in avc_idct8_addblk_msa()
237 PCKEV_B4_SB(res1, res0, res3, res2, res5, res4, res7, res6, in avc_idct8_addblk_msa()
|
| D | vp9_idct_msa.c | 712 v8i16 res0, res1, res2, res3, res4, res5, res6, res7; in vp9_iadst8x8_colcol_addblk_msa() local
771 res7 = (v8i16) __msa_ilvr_b((v16i8) zero, (v16i8) dst7); in vp9_iadst8x8_colcol_addblk_msa()
772 res7 += out7; in vp9_iadst8x8_colcol_addblk_msa()
773 CLIP_SH_0_255(res7); in vp9_iadst8x8_colcol_addblk_msa()
774 res7 = (v8i16) __msa_pckev_b((v16i8) res7, (v16i8) res7); in vp9_iadst8x8_colcol_addblk_msa()
775 ST_D1(res7, 0, dst + 7 * dst_stride); in vp9_iadst8x8_colcol_addblk_msa()
1176 v8i16 vec, res0, res1, res2, res3, res4, res5, res6, res7; in vp9_idct16x16_1_add_msa() local
1191 UNPCK_UB_SH(dst3, res3, res7); in vp9_idct16x16_1_add_msa()
1194 ADD4(res4, vec, res5, vec, res6, vec, res7, vec, res4, res5, res6, in vp9_idct16x16_1_add_msa()
1195 res7); in vp9_idct16x16_1_add_msa()
[all …]
|
| D | h264qpel_msa.c | 750 v8i16 res0, res1, res2, res3, res4, res5, res6, res7; in ff_put_h264_qpel16_mc10_msa() local
788 HADD_SB4_SH(vec0, vec3, vec6, vec9, res4, res5, res6, res7); in ff_put_h264_qpel16_mc10_msa()
790 minus5b, res4, res5, res6, res7); in ff_put_h264_qpel16_mc10_msa()
792 plus20b, res4, res5, res6, res7); in ff_put_h264_qpel16_mc10_msa()
796 SRARI_H4_SH(res4, res5, res6, res7, 5); in ff_put_h264_qpel16_mc10_msa()
798 SAT_SH4_SH(res4, res5, res6, res7, 7); in ff_put_h264_qpel16_mc10_msa()
800 PCKEV_B2_SB(res5, res4, res7, res6, dst2, dst3); in ff_put_h264_qpel16_mc10_msa()
818 v8i16 res0, res1, res2, res3, res4, res5, res6, res7; in ff_put_h264_qpel16_mc30_msa() local
856 HADD_SB4_SH(vec0, vec3, vec6, vec9, res4, res5, res6, res7); in ff_put_h264_qpel16_mc30_msa()
858 minus5b, res4, res5, res6, res7); in ff_put_h264_qpel16_mc30_msa()
[all …]
|
| D | h264chroma_msa.c | 215 v8u16 res0, res1, res2, res3, res4, res5, res6, res7; in avc_chroma_hz_8x8_msa() local
231 coeff_vec, res4, res5, res6, res7); in avc_chroma_hz_8x8_msa()
233 SLLI_4V(res4, res5, res6, res7, 3); in avc_chroma_hz_8x8_msa()
235 SRARI_H4_UH(res4, res5, res6, res7, 6); in avc_chroma_hz_8x8_msa()
237 SAT_UH4_UH(res4, res5, res6, res7, 7); in avc_chroma_hz_8x8_msa()
239 PCKEV_B2_UB(res5, res4, res7, res6, out2, out3); in avc_chroma_hz_8x8_msa()
491 v8u16 res0, res1, res2, res3, res4, res5, res6, res7; in avc_chroma_vt_8x8_msa() local
506 coeff_vec, res4, res5, res6, res7); in avc_chroma_vt_8x8_msa()
508 SLLI_4V(res4, res5, res6, res7, 3); in avc_chroma_vt_8x8_msa()
510 SRARI_H4_UH(res4, res5, res6, res7, 6); in avc_chroma_vt_8x8_msa()
[all …]
|
| D | hpeldsp_msa.c | 503 v16u8 res0, res1, res2, res3, res4, res5, res6, res7; in common_vt_bil_and_aver_dst_16w_msa() local
515 res4, res5, res6, res7); in common_vt_bil_and_aver_dst_16w_msa()
520 AVER_UB4_UB(dst4, res4, dst5, res5, dst6, res6, dst7, res7, in common_vt_bil_and_aver_dst_16w_msa()
521 res4, res5, res6, res7); in common_vt_bil_and_aver_dst_16w_msa()
522 ST_UB8(res0, res1, res2, res3, res4, res5, res6, res7, dst, dst_stride); in common_vt_bil_and_aver_dst_16w_msa()
|
| D | hevcpred_msa.c | 603 v8i16 res0, res1, res2, res3, res4, res5, res6, res7; in hevc_intra_pred_plane_8x8_msa() local
625 res4, res5, res6, res7); in hevc_intra_pred_plane_8x8_msa()
635 res7 += tmp2; in hevc_intra_pred_plane_8x8_msa()
652 res7 += 8 * tmp1; in hevc_intra_pred_plane_8x8_msa()
655 SRARI_H4_SH(res4, res5, res6, res7, 4); in hevc_intra_pred_plane_8x8_msa()
656 PCKEV_B4_SB(res1, res0, res3, res2, res5, res4, res7, res6, in hevc_intra_pred_plane_8x8_msa()
|
| D | vp9_mc_msa.c | 3020 v8u16 res0, res1, res2, res3, res4, res5, res6, res7, filt; in ff_avg_bilin_16h_msa() local
3039 res6, res7); in ff_avg_bilin_16h_msa()
3041 SRARI_H4_UH(res4, res5, res6, res7, 7); in ff_avg_bilin_16h_msa()
3049 PCKEV_AVG_ST_UB(res7, res6, dst3, dst); in ff_avg_bilin_16h_msa()
3064 res5, res6, res7); in ff_avg_bilin_16h_msa()
3066 SRARI_H4_UH(res4, res5, res6, res7, 7); in ff_avg_bilin_16h_msa()
3074 PCKEV_AVG_ST_UB(res7, res6, dst3, dst); in ff_avg_bilin_16h_msa()
3088 v8u16 res0, res1, res2, res3, res4, res5, res6, res7, filt; in ff_avg_bilin_32h_msa() local
3115 res4, res5, res6, res7); in ff_avg_bilin_32h_msa()
3117 SRARI_H4_UH(res4, res5, res6, res7, 7); in ff_avg_bilin_32h_msa()
[all …]
|
| /third_party/skia/third_party/externals/swiftshader/third_party/subzero/crosstest/ |
| D | test_vector_ops_ll.ll | 96 %res7 = zext <8 x i1> %res7_i1 to <8 x i16>
97 ret <8 x i16> %res7
153 %res7 = zext <16 x i1> %res7_i1 to <16 x i8>
154 ret <16 x i8> %res7
257 %res7 = insertelement <8 x i16> %vec, i16 %elt, i32 7
258 ret <8 x i16> %res7
312 %res7 = insertelement <16 x i8> %vec, i8 %elt, i32 7
313 ret <16 x i8> %res7
441 %res7 = zext i1 %res7_i1 to i64
442 ret i64 %res7
[all …]
|
| /third_party/skia/third_party/externals/spirv-cross/shaders-hlsl-no-opt/asm/frag/ |
| D | subgroup-arithmetic-cast.invalid.nofxc.sm60.asm.frag | 47 ;%res7 = OpGroupNonUniformUMin %uint %uint_3 ExclusiveScan %u
59 ;OpStore %FragColor %res7
|
| /third_party/skia/third_party/externals/spirv-cross/shaders-msl-no-opt/asm/frag/ |
| D | subgroup-arithmetic-cast.msl21.asm.frag | 47 ;%res7 = OpGroupNonUniformUMin %uint %uint_3 ExclusiveScan %u
59 ;OpStore %FragColor %res7
|
| /third_party/skia/third_party/externals/spirv-cross/shaders-no-opt/asm/frag/ |
| D | subgroup-arithmetic-cast.nocompat.vk.asm.frag | 47 %res7 = OpGroupNonUniformUMin %uint %uint_3 ExclusiveScan %u
59 OpStore %FragColor %res7
|
| /third_party/mesa3d/src/amd/compiler/tests/ |
| D | test_optimizer_postRA.cpp | 352 auto res7 = bld.vop2_e64(aco_opcode::v_sub_f32, bld.def(v1, reg_v2), b, Operand(tmp7, reg_v2)); variable
353 res7.instr->vop3().abs[0] = true;
354 writeout(7, Operand(res7, reg_v2));
|
| D | test_optimizer.cpp | 1021 auto res7 = bld.vop2_e64(aco_opcode::v_sub_f32, bld.def(v1), b, tmp7); variable
1022 res7.instr->vop3().abs[0] = true;
1023 writeout(7, res7);
|
| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/fp32/ |
| D | matmul_fp32.c | 729 __m128 res7 = _mm_shuffle_ps(hi2, hi0, _MM_SHUFFLE(3, 2, 1, 0)); in RowMajor2Col6Major() local
741 _mm_storeu_ps(dst_c + 28, res7); in RowMajor2Col6Major()
|