| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 263 _sum0 = vfmacc_vf_f32m1(_sum0, r2[6], _k26, vl); in conv7x7s2_pack1ton_rvv()
264 _sum1 = vfmacc_vf_f32m1(_sum1, r2[8], _k26, vl); in conv7x7s2_pack1ton_rvv()
265 _sum2 = vfmacc_vf_f32m1(_sum2, r2[10], _k26, vl); in conv7x7s2_pack1ton_rvv()
266 _sum3 = vfmacc_vf_f32m1(_sum3, r2[12], _k26, vl); in conv7x7s2_pack1ton_rvv()
267 _sum4 = vfmacc_vf_f32m1(_sum4, r2[14], _k26, vl); in conv7x7s2_pack1ton_rvv()
268 _sum5 = vfmacc_vf_f32m1(_sum5, r2[16], _k26, vl); in conv7x7s2_pack1ton_rvv()
269 _sum6 = vfmacc_vf_f32m1(_sum6, r2[18], _k26, vl); in conv7x7s2_pack1ton_rvv()
270 _sum7 = vfmacc_vf_f32m1(_sum7, r2[20], _k26, vl); in conv7x7s2_pack1ton_rvv()
678 _sum0 = vfmacc_vf_f32m1(_sum0, r2[6], _k26, vl); in conv7x7s2_pack1ton_rvv()
679 _sum1 = vfmacc_vf_f32m1(_sum1, r2[8], _k26, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 263 _sum0 = vfmacc_vf_f16m1(_sum0, r2[6], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
264 _sum1 = vfmacc_vf_f16m1(_sum1, r2[8], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
265 _sum2 = vfmacc_vf_f16m1(_sum2, r2[10], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
266 _sum3 = vfmacc_vf_f16m1(_sum3, r2[12], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
267 _sum4 = vfmacc_vf_f16m1(_sum4, r2[14], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
268 _sum5 = vfmacc_vf_f16m1(_sum5, r2[16], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
269 _sum6 = vfmacc_vf_f16m1(_sum6, r2[18], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
270 _sum7 = vfmacc_vf_f16m1(_sum7, r2[20], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
678 _sum0 = vfmacc_vf_f16m1(_sum0, r2[6], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
679 _sum1 = vfmacc_vf_f16m1(_sum1, r2[8], _k26, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 184 v4f32 _k26 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
230 _sum0 = __msa_fmadd_w(_sum0, _r26, _k26); in conv7x7s2_pack1to4_msa()
231 _sum1 = __msa_fmadd_w(_sum1, _r28, _k26); in conv7x7s2_pack1to4_msa()
232 _sum2 = __msa_fmadd_w(_sum2, _r2a, _k26); in conv7x7s2_pack1to4_msa()
233 _sum3 = __msa_fmadd_w(_sum3, _r2c, _k26); in conv7x7s2_pack1to4_msa()
530 v4f32 _k26 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
543 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r2n, 2), _k26); 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 | 147 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
158 _sum00 = _mm256_fmadd_ps(_r026, _k26, _sum00); in conv3x3s1_pack8_avx()
176 _sum10 = _mm256_fmadd_ps(_r036, _k26, _sum10); in conv3x3s1_pack8_avx()
514 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
525 _sum0 = _mm256_fmadd_ps(_r026, _k26, _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 | 147 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
158 _sum00 = _mm256_fmadd_ps(_r026, _k26, _sum00); in conv3x3s1_pack8_avx()
176 _sum10 = _mm256_fmadd_ps(_r036, _k26, _sum10); in conv3x3s1_pack8_avx()
514 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
525 _sum0 = _mm256_fmadd_ps(_r026, _k26, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 147 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
158 _sum00 = _mm256_fmadd_ps(_r026, _k26, _sum00); in conv3x3s1_pack8_avx()
176 _sum10 = _mm256_fmadd_ps(_r036, _k26, _sum10); in conv3x3s1_pack8_avx()
514 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
525 _sum0 = _mm256_fmadd_ps(_r026, _k26, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 147 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
158 _sum00 = _mm256_comp_fmadd_ps(_r026, _k26, _sum00); in conv3x3s1_pack8_avx()
176 _sum10 = _mm256_comp_fmadd_ps(_r036, _k26, _sum10); in conv3x3s1_pack8_avx()
514 __m256 _k26 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
525 _sum0 = _mm256_comp_fmadd_ps(_r026, _k26, _sum0); in conv3x3s1_pack8_avx()
|