| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 499 _sum0 = vfmacc_vf_f32m1(_sum0, r6[2], _k62, vl); in conv7x7s2_pack1ton_rvv()
500 _sum1 = vfmacc_vf_f32m1(_sum1, r6[4], _k62, vl); in conv7x7s2_pack1ton_rvv()
501 _sum2 = vfmacc_vf_f32m1(_sum2, r6[6], _k62, vl); in conv7x7s2_pack1ton_rvv()
502 _sum3 = vfmacc_vf_f32m1(_sum3, r6[8], _k62, vl); in conv7x7s2_pack1ton_rvv()
503 _sum4 = vfmacc_vf_f32m1(_sum4, r6[10], _k62, vl); in conv7x7s2_pack1ton_rvv()
504 _sum5 = vfmacc_vf_f32m1(_sum5, r6[12], _k62, vl); in conv7x7s2_pack1ton_rvv()
505 _sum6 = vfmacc_vf_f32m1(_sum6, r6[14], _k62, vl); in conv7x7s2_pack1ton_rvv()
818 _sum0 = vfmacc_vf_f32m1(_sum0, r6[2], _k62, vl); in conv7x7s2_pack1ton_rvv()
819 _sum1 = vfmacc_vf_f32m1(_sum1, r6[4], _k62, vl); in conv7x7s2_pack1ton_rvv()
820 _sum2 = vfmacc_vf_f32m1(_sum2, r6[6], _k62, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 499 _sum0 = vfmacc_vf_f16m1(_sum0, r6[2], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
500 _sum1 = vfmacc_vf_f16m1(_sum1, r6[4], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
501 _sum2 = vfmacc_vf_f16m1(_sum2, r6[6], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
502 _sum3 = vfmacc_vf_f16m1(_sum3, r6[8], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
503 _sum4 = vfmacc_vf_f16m1(_sum4, r6[10], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
504 _sum5 = vfmacc_vf_f16m1(_sum5, r6[12], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
505 _sum6 = vfmacc_vf_f16m1(_sum6, r6[14], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
818 _sum0 = vfmacc_vf_f16m1(_sum0, r6[2], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
819 _sum1 = vfmacc_vf_f16m1(_sum1, r6[4], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
820 _sum2 = vfmacc_vf_f16m1(_sum2, r6[6], _k62, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 408 v4f32 _k62 = (v4f32)__msa_ld_w(kptr + 4 * 2, 0); in conv7x7s2_pack1to4_msa() local
442 _sum0 = __msa_fmadd_w(_sum0, _r62, _k62); in conv7x7s2_pack1to4_msa()
443 _sum1 = __msa_fmadd_w(_sum1, _r64, _k62); in conv7x7s2_pack1to4_msa()
444 _sum2 = __msa_fmadd_w(_sum2, _r66, _k62); in conv7x7s2_pack1to4_msa()
445 _sum3 = __msa_fmadd_w(_sum3, _r68, _k62); in conv7x7s2_pack1to4_msa()
610 v4f32 _k62 = (v4f32)__msa_ld_w(kptr + 4 * 2, 0); in conv7x7s2_pack1to4_msa() local
623 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r6, 2), _k62); 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 | 313 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
324 _sum00 = _mm256_fmadd_ps(_r202, _k62, _sum00); in conv3x3s1_pack8_avx()
342 _sum10 = _mm256_fmadd_ps(_r212, _k62, _sum10); in conv3x3s1_pack8_avx()
626 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
637 _sum0 = _mm256_fmadd_ps(_r202, _k62, _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 | 313 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
324 _sum00 = _mm256_fmadd_ps(_r202, _k62, _sum00); in conv3x3s1_pack8_avx()
342 _sum10 = _mm256_fmadd_ps(_r212, _k62, _sum10); in conv3x3s1_pack8_avx()
626 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
637 _sum0 = _mm256_fmadd_ps(_r202, _k62, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 313 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
324 _sum00 = _mm256_fmadd_ps(_r202, _k62, _sum00); in conv3x3s1_pack8_avx()
342 _sum10 = _mm256_fmadd_ps(_r212, _k62, _sum10); in conv3x3s1_pack8_avx()
626 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
637 _sum0 = _mm256_fmadd_ps(_r202, _k62, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 313 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
324 _sum00 = _mm256_comp_fmadd_ps(_r202, _k62, _sum00); in conv3x3s1_pack8_avx()
342 _sum10 = _mm256_comp_fmadd_ps(_r212, _k62, _sum10); in conv3x3s1_pack8_avx()
626 __m256 _k62 = _mm256_loadu_ps(kptr + 16); in conv3x3s1_pack8_avx() local
637 _sum0 = _mm256_comp_fmadd_ps(_r202, _k62, _sum0); in conv3x3s1_pack8_avx()
|