| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 464 _sum0 = vfmacc_vf_f32m1(_sum0, r5[6], _k56, vl); in conv7x7s2_pack1ton_rvv()
465 _sum1 = vfmacc_vf_f32m1(_sum1, r5[8], _k56, vl); in conv7x7s2_pack1ton_rvv()
466 _sum2 = vfmacc_vf_f32m1(_sum2, r5[10], _k56, vl); in conv7x7s2_pack1ton_rvv()
467 _sum3 = vfmacc_vf_f32m1(_sum3, r5[12], _k56, vl); in conv7x7s2_pack1ton_rvv()
468 _sum4 = vfmacc_vf_f32m1(_sum4, r5[14], _k56, vl); in conv7x7s2_pack1ton_rvv()
469 _sum5 = vfmacc_vf_f32m1(_sum5, r5[16], _k56, vl); in conv7x7s2_pack1ton_rvv()
470 _sum6 = vfmacc_vf_f32m1(_sum6, r5[18], _k56, vl); in conv7x7s2_pack1ton_rvv()
471 _sum7 = vfmacc_vf_f32m1(_sum7, r5[20], _k56, vl); in conv7x7s2_pack1ton_rvv()
795 _sum0 = vfmacc_vf_f32m1(_sum0, r5[6], _k56, vl); in conv7x7s2_pack1ton_rvv()
796 _sum1 = vfmacc_vf_f32m1(_sum1, r5[8], _k56, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 464 _sum0 = vfmacc_vf_f16m1(_sum0, r5[6], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
465 _sum1 = vfmacc_vf_f16m1(_sum1, r5[8], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
466 _sum2 = vfmacc_vf_f16m1(_sum2, r5[10], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
467 _sum3 = vfmacc_vf_f16m1(_sum3, r5[12], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
468 _sum4 = vfmacc_vf_f16m1(_sum4, r5[14], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
469 _sum5 = vfmacc_vf_f16m1(_sum5, r5[16], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
470 _sum6 = vfmacc_vf_f16m1(_sum6, r5[18], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
471 _sum7 = vfmacc_vf_f16m1(_sum7, r5[20], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
795 _sum0 = vfmacc_vf_f16m1(_sum0, r5[6], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
796 _sum1 = vfmacc_vf_f16m1(_sum1, r5[8], _k56, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 355 v4f32 _k56 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
401 _sum0 = __msa_fmadd_w(_sum0, _r56, _k56); in conv7x7s2_pack1to4_msa()
402 _sum1 = __msa_fmadd_w(_sum1, _r58, _k56); in conv7x7s2_pack1to4_msa()
403 _sum2 = __msa_fmadd_w(_sum2, _r5a, _k56); in conv7x7s2_pack1to4_msa()
404 _sum3 = __msa_fmadd_w(_sum3, _r5c, _k56); in conv7x7s2_pack1to4_msa()
593 v4f32 _k56 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
606 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r5n, 2), _k56); in conv7x7s2_pack1to4_msa()
|
| /dports/graphics/vapoursynth-waifu2x-ncnn-vulkan/vapoursynth-waifu2x-ncnn-vulkan-r4/deps/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 270 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
281 _sum00 = _mm256_fmadd_ps(_r126, _k56, _sum00); in conv3x3s1_pack8_avx()
299 _sum10 = _mm256_fmadd_ps(_r136, _k56, _sum10); in conv3x3s1_pack8_avx()
601 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
612 _sum0 = _mm256_fmadd_ps(_r126, _k56, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/graphics/waifu2x-ncnn-vulkan/waifu2x-ncnn-vulkan-20210521/src/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 270 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
281 _sum00 = _mm256_fmadd_ps(_r126, _k56, _sum00); in conv3x3s1_pack8_avx()
299 _sum10 = _mm256_fmadd_ps(_r136, _k56, _sum10); in conv3x3s1_pack8_avx()
601 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
612 _sum0 = _mm256_fmadd_ps(_r126, _k56, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 270 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
281 _sum00 = _mm256_fmadd_ps(_r126, _k56, _sum00); in conv3x3s1_pack8_avx()
299 _sum10 = _mm256_fmadd_ps(_r136, _k56, _sum10); in conv3x3s1_pack8_avx()
601 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
612 _sum0 = _mm256_fmadd_ps(_r126, _k56, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 270 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
281 _sum00 = _mm256_comp_fmadd_ps(_r126, _k56, _sum00); in conv3x3s1_pack8_avx()
299 _sum10 = _mm256_comp_fmadd_ps(_r136, _k56, _sum10); in conv3x3s1_pack8_avx()
601 __m256 _k56 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
612 _sum0 = _mm256_comp_fmadd_ps(_r126, _k56, _sum0); in conv3x3s1_pack8_avx()
|