| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 507 _sum0 = vfmacc_vf_f32m1(_sum0, r6[3], _k63, vl); in conv7x7s2_pack1ton_rvv()
508 _sum1 = vfmacc_vf_f32m1(_sum1, r6[5], _k63, vl); in conv7x7s2_pack1ton_rvv()
509 _sum2 = vfmacc_vf_f32m1(_sum2, r6[7], _k63, vl); in conv7x7s2_pack1ton_rvv()
510 _sum3 = vfmacc_vf_f32m1(_sum3, r6[9], _k63, vl); in conv7x7s2_pack1ton_rvv()
511 _sum4 = vfmacc_vf_f32m1(_sum4, r6[11], _k63, vl); in conv7x7s2_pack1ton_rvv()
512 _sum5 = vfmacc_vf_f32m1(_sum5, r6[13], _k63, vl); in conv7x7s2_pack1ton_rvv()
822 _sum0 = vfmacc_vf_f32m1(_sum0, r6[3], _k63, vl); in conv7x7s2_pack1ton_rvv()
823 _sum1 = vfmacc_vf_f32m1(_sum1, r6[5], _k63, vl); in conv7x7s2_pack1ton_rvv()
824 _sum2 = vfmacc_vf_f32m1(_sum2, r6[7], _k63, vl); in conv7x7s2_pack1ton_rvv()
825 _sum3 = vfmacc_vf_f32m1(_sum3, r6[9], _k63, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 507 _sum0 = vfmacc_vf_f16m1(_sum0, r6[3], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
508 _sum1 = vfmacc_vf_f16m1(_sum1, r6[5], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
509 _sum2 = vfmacc_vf_f16m1(_sum2, r6[7], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
510 _sum3 = vfmacc_vf_f16m1(_sum3, r6[9], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
511 _sum4 = vfmacc_vf_f16m1(_sum4, r6[11], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
512 _sum5 = vfmacc_vf_f16m1(_sum5, r6[13], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
822 _sum0 = vfmacc_vf_f16m1(_sum0, r6[3], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
823 _sum1 = vfmacc_vf_f16m1(_sum1, r6[5], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
824 _sum2 = vfmacc_vf_f16m1(_sum2, r6[7], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
825 _sum3 = vfmacc_vf_f16m1(_sum3, r6[9], _k63, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 409 v4f32 _k63 = (v4f32)__msa_ld_w(kptr + 4 * 3, 0); in conv7x7s2_pack1to4_msa() local
446 _sum0 = __msa_fmadd_w(_sum0, _r63, _k63); in conv7x7s2_pack1to4_msa()
447 _sum1 = __msa_fmadd_w(_sum1, _r65, _k63); in conv7x7s2_pack1to4_msa()
448 _sum2 = __msa_fmadd_w(_sum2, _r67, _k63); in conv7x7s2_pack1to4_msa()
449 _sum3 = __msa_fmadd_w(_sum3, _r69, _k63); in conv7x7s2_pack1to4_msa()
611 v4f32 _k63 = (v4f32)__msa_ld_w(kptr + 4 * 3, 0); in conv7x7s2_pack1to4_msa() local
624 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r6, 3), _k63); 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 | 314 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
325 _sum01 = _mm256_fmadd_ps(_r203, _k63, _sum01); in conv3x3s1_pack8_avx()
343 _sum11 = _mm256_fmadd_ps(_r213, _k63, _sum11); in conv3x3s1_pack8_avx()
627 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
638 _sum1 = _mm256_fmadd_ps(_r203, _k63, _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 | 314 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
325 _sum01 = _mm256_fmadd_ps(_r203, _k63, _sum01); in conv3x3s1_pack8_avx()
343 _sum11 = _mm256_fmadd_ps(_r213, _k63, _sum11); in conv3x3s1_pack8_avx()
627 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
638 _sum1 = _mm256_fmadd_ps(_r203, _k63, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 314 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
325 _sum01 = _mm256_fmadd_ps(_r203, _k63, _sum01); in conv3x3s1_pack8_avx()
343 _sum11 = _mm256_fmadd_ps(_r213, _k63, _sum11); in conv3x3s1_pack8_avx()
627 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
638 _sum1 = _mm256_fmadd_ps(_r203, _k63, _sum1); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 314 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
325 _sum01 = _mm256_comp_fmadd_ps(_r203, _k63, _sum01); in conv3x3s1_pack8_avx()
343 _sum11 = _mm256_comp_fmadd_ps(_r213, _k63, _sum11); in conv3x3s1_pack8_avx()
627 __m256 _k63 = _mm256_loadu_ps(kptr + 24); in conv3x3s1_pack8_avx() local
638 _sum1 = _mm256_comp_fmadd_ps(_r203, _k63, _sum1); in conv3x3s1_pack8_avx()
|