| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 389 _sum0 = vfmacc_vf_f32m1(_sum0, r4[5], _k45, vl); in conv7x7s2_pack1ton_rvv()
390 _sum1 = vfmacc_vf_f32m1(_sum1, r4[7], _k45, vl); in conv7x7s2_pack1ton_rvv()
391 _sum2 = vfmacc_vf_f32m1(_sum2, r4[9], _k45, vl); in conv7x7s2_pack1ton_rvv()
392 _sum3 = vfmacc_vf_f32m1(_sum3, r4[11], _k45, vl); in conv7x7s2_pack1ton_rvv()
393 _sum4 = vfmacc_vf_f32m1(_sum4, r4[13], _k45, vl); in conv7x7s2_pack1ton_rvv()
394 _sum5 = vfmacc_vf_f32m1(_sum5, r4[15], _k45, vl); in conv7x7s2_pack1ton_rvv()
395 _sum6 = vfmacc_vf_f32m1(_sum6, r4[17], _k45, vl); in conv7x7s2_pack1ton_rvv()
396 _sum7 = vfmacc_vf_f32m1(_sum7, r4[19], _k45, vl); in conv7x7s2_pack1ton_rvv()
752 _sum0 = vfmacc_vf_f32m1(_sum0, r4[5], _k45, vl); in conv7x7s2_pack1ton_rvv()
753 _sum1 = vfmacc_vf_f32m1(_sum1, r4[7], _k45, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 389 _sum0 = vfmacc_vf_f16m1(_sum0, r4[5], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
390 _sum1 = vfmacc_vf_f16m1(_sum1, r4[7], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
391 _sum2 = vfmacc_vf_f16m1(_sum2, r4[9], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
392 _sum3 = vfmacc_vf_f16m1(_sum3, r4[11], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
393 _sum4 = vfmacc_vf_f16m1(_sum4, r4[13], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
394 _sum5 = vfmacc_vf_f16m1(_sum5, r4[15], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
395 _sum6 = vfmacc_vf_f16m1(_sum6, r4[17], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
396 _sum7 = vfmacc_vf_f16m1(_sum7, r4[19], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
752 _sum0 = vfmacc_vf_f16m1(_sum0, r4[5], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
753 _sum1 = vfmacc_vf_f16m1(_sum1, r4[7], _k45, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 297 v4f32 _k45 = (v4f32)__msa_ld_w(kptr + 4 * 5, 0); in conv7x7s2_pack1to4_msa() local
340 _sum0 = __msa_fmadd_w(_sum0, _r45, _k45); in conv7x7s2_pack1to4_msa()
341 _sum1 = __msa_fmadd_w(_sum1, _r47, _k45); in conv7x7s2_pack1to4_msa()
342 _sum2 = __msa_fmadd_w(_sum2, _r49, _k45); in conv7x7s2_pack1to4_msa()
343 _sum3 = __msa_fmadd_w(_sum3, _r4b, _k45); in conv7x7s2_pack1to4_msa()
571 v4f32 _k45 = (v4f32)__msa_ld_w(kptr + 4 * 5, 0); in conv7x7s2_pack1to4_msa() local
584 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r4n, 1), _k45); 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 | 231 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
242 _sum01 = _mm256_fmadd_ps(_r115, _k45, _sum01); in conv3x3s1_pack8_avx()
260 _sum11 = _mm256_fmadd_ps(_r125, _k45, _sum11); in conv3x3s1_pack8_avx()
571 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
582 _sum1 = _mm256_fmadd_ps(_r115, _k45, _sum1); 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 | 231 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
242 _sum01 = _mm256_fmadd_ps(_r115, _k45, _sum01); in conv3x3s1_pack8_avx()
260 _sum11 = _mm256_fmadd_ps(_r125, _k45, _sum11); in conv3x3s1_pack8_avx()
571 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
582 _sum1 = _mm256_fmadd_ps(_r115, _k45, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 231 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
242 _sum01 = _mm256_fmadd_ps(_r115, _k45, _sum01); in conv3x3s1_pack8_avx()
260 _sum11 = _mm256_fmadd_ps(_r125, _k45, _sum11); in conv3x3s1_pack8_avx()
571 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
582 _sum1 = _mm256_fmadd_ps(_r115, _k45, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 231 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
242 _sum01 = _mm256_comp_fmadd_ps(_r115, _k45, _sum01); in conv3x3s1_pack8_avx()
260 _sum11 = _mm256_comp_fmadd_ps(_r125, _k45, _sum11); in conv3x3s1_pack8_avx()
571 __m256 _k45 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
582 _sum1 = _mm256_comp_fmadd_ps(_r115, _k45, _sum1); in conv3x3s1_pack8_avx()
|