| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 456 _sum0 = vfmacc_vf_f32m1(_sum0, r5[5], _k55, vl); in conv7x7s2_pack1ton_rvv()
457 _sum1 = vfmacc_vf_f32m1(_sum1, r5[7], _k55, vl); in conv7x7s2_pack1ton_rvv()
458 _sum2 = vfmacc_vf_f32m1(_sum2, r5[9], _k55, vl); in conv7x7s2_pack1ton_rvv()
459 _sum3 = vfmacc_vf_f32m1(_sum3, r5[11], _k55, vl); in conv7x7s2_pack1ton_rvv()
460 _sum4 = vfmacc_vf_f32m1(_sum4, r5[13], _k55, vl); in conv7x7s2_pack1ton_rvv()
461 _sum5 = vfmacc_vf_f32m1(_sum5, r5[15], _k55, vl); in conv7x7s2_pack1ton_rvv()
462 _sum6 = vfmacc_vf_f32m1(_sum6, r5[17], _k55, vl); in conv7x7s2_pack1ton_rvv()
463 _sum7 = vfmacc_vf_f32m1(_sum7, r5[19], _k55, vl); in conv7x7s2_pack1ton_rvv()
791 _sum0 = vfmacc_vf_f32m1(_sum0, r5[5], _k55, vl); in conv7x7s2_pack1ton_rvv()
792 _sum1 = vfmacc_vf_f32m1(_sum1, r5[7], _k55, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 456 _sum0 = vfmacc_vf_f16m1(_sum0, r5[5], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
457 _sum1 = vfmacc_vf_f16m1(_sum1, r5[7], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
458 _sum2 = vfmacc_vf_f16m1(_sum2, r5[9], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
459 _sum3 = vfmacc_vf_f16m1(_sum3, r5[11], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
460 _sum4 = vfmacc_vf_f16m1(_sum4, r5[13], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
461 _sum5 = vfmacc_vf_f16m1(_sum5, r5[15], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
462 _sum6 = vfmacc_vf_f16m1(_sum6, r5[17], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
463 _sum7 = vfmacc_vf_f16m1(_sum7, r5[19], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
791 _sum0 = vfmacc_vf_f16m1(_sum0, r5[5], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
792 _sum1 = vfmacc_vf_f16m1(_sum1, r5[7], _k55, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 354 v4f32 _k55 = (v4f32)__msa_ld_w(kptr + 4 * 5, 0); in conv7x7s2_pack1to4_msa() local
397 _sum0 = __msa_fmadd_w(_sum0, _r55, _k55); in conv7x7s2_pack1to4_msa()
398 _sum1 = __msa_fmadd_w(_sum1, _r57, _k55); in conv7x7s2_pack1to4_msa()
399 _sum2 = __msa_fmadd_w(_sum2, _r59, _k55); in conv7x7s2_pack1to4_msa()
400 _sum3 = __msa_fmadd_w(_sum3, _r5b, _k55); in conv7x7s2_pack1to4_msa()
592 v4f32 _k55 = (v4f32)__msa_ld_w(kptr + 4 * 5, 0); in conv7x7s2_pack1to4_msa() local
605 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r5n, 1), _k55); 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 | 269 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
280 _sum01 = _mm256_fmadd_ps(_r125, _k55, _sum01); in conv3x3s1_pack8_avx()
298 _sum11 = _mm256_fmadd_ps(_r135, _k55, _sum11); in conv3x3s1_pack8_avx()
600 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
611 _sum1 = _mm256_fmadd_ps(_r125, _k55, _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 | 269 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
280 _sum01 = _mm256_fmadd_ps(_r125, _k55, _sum01); in conv3x3s1_pack8_avx()
298 _sum11 = _mm256_fmadd_ps(_r135, _k55, _sum11); in conv3x3s1_pack8_avx()
600 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
611 _sum1 = _mm256_fmadd_ps(_r125, _k55, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 269 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
280 _sum01 = _mm256_fmadd_ps(_r125, _k55, _sum01); in conv3x3s1_pack8_avx()
298 _sum11 = _mm256_fmadd_ps(_r135, _k55, _sum11); in conv3x3s1_pack8_avx()
600 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
611 _sum1 = _mm256_fmadd_ps(_r125, _k55, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 269 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
280 _sum01 = _mm256_comp_fmadd_ps(_r125, _k55, _sum01); in conv3x3s1_pack8_avx()
298 _sum11 = _mm256_comp_fmadd_ps(_r135, _k55, _sum11); in conv3x3s1_pack8_avx()
600 __m256 _k55 = _mm256_loadu_ps(kptr + 40); in conv3x3s1_pack8_avx() local
611 _sum1 = _mm256_comp_fmadd_ps(_r125, _k55, _sum1); in conv3x3s1_pack8_avx()
|