| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 397 _sum0 = vfmacc_vf_f32m1(_sum0, r4[6], _k46, vl); in conv7x7s2_pack1ton_rvv()
398 _sum1 = vfmacc_vf_f32m1(_sum1, r4[8], _k46, vl); in conv7x7s2_pack1ton_rvv()
399 _sum2 = vfmacc_vf_f32m1(_sum2, r4[10], _k46, vl); in conv7x7s2_pack1ton_rvv()
400 _sum3 = vfmacc_vf_f32m1(_sum3, r4[12], _k46, vl); in conv7x7s2_pack1ton_rvv()
401 _sum4 = vfmacc_vf_f32m1(_sum4, r4[14], _k46, vl); in conv7x7s2_pack1ton_rvv()
402 _sum5 = vfmacc_vf_f32m1(_sum5, r4[16], _k46, vl); in conv7x7s2_pack1ton_rvv()
403 _sum6 = vfmacc_vf_f32m1(_sum6, r4[18], _k46, vl); in conv7x7s2_pack1ton_rvv()
404 _sum7 = vfmacc_vf_f32m1(_sum7, r4[20], _k46, vl); in conv7x7s2_pack1ton_rvv()
756 _sum0 = vfmacc_vf_f32m1(_sum0, r4[6], _k46, vl); in conv7x7s2_pack1ton_rvv()
757 _sum1 = vfmacc_vf_f32m1(_sum1, r4[8], _k46, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 397 _sum0 = vfmacc_vf_f16m1(_sum0, r4[6], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
398 _sum1 = vfmacc_vf_f16m1(_sum1, r4[8], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
399 _sum2 = vfmacc_vf_f16m1(_sum2, r4[10], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
400 _sum3 = vfmacc_vf_f16m1(_sum3, r4[12], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
401 _sum4 = vfmacc_vf_f16m1(_sum4, r4[14], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
402 _sum5 = vfmacc_vf_f16m1(_sum5, r4[16], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
403 _sum6 = vfmacc_vf_f16m1(_sum6, r4[18], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
404 _sum7 = vfmacc_vf_f16m1(_sum7, r4[20], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
756 _sum0 = vfmacc_vf_f16m1(_sum0, r4[6], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
757 _sum1 = vfmacc_vf_f16m1(_sum1, r4[8], _k46, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 298 v4f32 _k46 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
344 _sum0 = __msa_fmadd_w(_sum0, _r46, _k46); in conv7x7s2_pack1to4_msa()
345 _sum1 = __msa_fmadd_w(_sum1, _r48, _k46); in conv7x7s2_pack1to4_msa()
346 _sum2 = __msa_fmadd_w(_sum2, _r4a, _k46); in conv7x7s2_pack1to4_msa()
347 _sum3 = __msa_fmadd_w(_sum3, _r4c, _k46); in conv7x7s2_pack1to4_msa()
572 v4f32 _k46 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
585 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r4n, 2), _k46); 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 | 232 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
243 _sum00 = _mm256_fmadd_ps(_r116, _k46, _sum00); in conv3x3s1_pack8_avx()
261 _sum10 = _mm256_fmadd_ps(_r126, _k46, _sum10); in conv3x3s1_pack8_avx()
572 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
583 _sum0 = _mm256_fmadd_ps(_r116, _k46, _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 | 232 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
243 _sum00 = _mm256_fmadd_ps(_r116, _k46, _sum00); in conv3x3s1_pack8_avx()
261 _sum10 = _mm256_fmadd_ps(_r126, _k46, _sum10); in conv3x3s1_pack8_avx()
572 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
583 _sum0 = _mm256_fmadd_ps(_r116, _k46, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 232 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
243 _sum00 = _mm256_fmadd_ps(_r116, _k46, _sum00); in conv3x3s1_pack8_avx()
261 _sum10 = _mm256_fmadd_ps(_r126, _k46, _sum10); in conv3x3s1_pack8_avx()
572 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
583 _sum0 = _mm256_fmadd_ps(_r116, _k46, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 232 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
243 _sum00 = _mm256_comp_fmadd_ps(_r116, _k46, _sum00); in conv3x3s1_pack8_avx()
261 _sum10 = _mm256_comp_fmadd_ps(_r126, _k46, _sum10); in conv3x3s1_pack8_avx()
572 __m256 _k46 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
583 _sum0 = _mm256_comp_fmadd_ps(_r116, _k46, _sum0); in conv3x3s1_pack8_avx()
|