| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 406 vfloat32m1_t _k50 = vle32_v_f32m1(kptr, vl); in conv7x7s2_pack1ton_rvv() local
416 _sum0 = vfmacc_vf_f32m1(_sum0, r5[0], _k50, vl); in conv7x7s2_pack1ton_rvv()
417 _sum1 = vfmacc_vf_f32m1(_sum1, r5[2], _k50, vl); in conv7x7s2_pack1ton_rvv()
418 _sum2 = vfmacc_vf_f32m1(_sum2, r5[4], _k50, vl); in conv7x7s2_pack1ton_rvv()
419 _sum3 = vfmacc_vf_f32m1(_sum3, r5[6], _k50, vl); in conv7x7s2_pack1ton_rvv()
420 _sum4 = vfmacc_vf_f32m1(_sum4, r5[8], _k50, vl); in conv7x7s2_pack1ton_rvv()
421 _sum5 = vfmacc_vf_f32m1(_sum5, r5[10], _k50, vl); in conv7x7s2_pack1ton_rvv()
422 _sum6 = vfmacc_vf_f32m1(_sum6, r5[12], _k50, vl); in conv7x7s2_pack1ton_rvv()
761 vfloat32m1_t _k50 = vle32_v_f32m1(kptr, vl); in conv7x7s2_pack1ton_rvv() local
771 _sum0 = vfmacc_vf_f32m1(_sum0, r5[0], _k50, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 406 vfloat16m1_t _k50 = vle16_v_f16m1(kptr, vl); in conv7x7s2_pack1ton_fp16sa_rvv() local
416 _sum0 = vfmacc_vf_f16m1(_sum0, r5[0], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
417 _sum1 = vfmacc_vf_f16m1(_sum1, r5[2], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
418 _sum2 = vfmacc_vf_f16m1(_sum2, r5[4], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
419 _sum3 = vfmacc_vf_f16m1(_sum3, r5[6], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
420 _sum4 = vfmacc_vf_f16m1(_sum4, r5[8], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
421 _sum5 = vfmacc_vf_f16m1(_sum5, r5[10], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
761 vfloat16m1_t _k50 = vle16_v_f16m1(kptr, vl); in conv7x7s2_pack1ton_fp16sa_rvv() local
771 _sum0 = vfmacc_vf_f16m1(_sum0, r5[0], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
772 _sum1 = vfmacc_vf_f16m1(_sum1, r5[2], _k50, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 349 v4f32 _k50 = (v4f32)__msa_ld_w(kptr, 0); in conv7x7s2_pack1to4_msa() local
377 _sum0 = __msa_fmadd_w(_sum0, _r50, _k50); in conv7x7s2_pack1to4_msa()
378 _sum1 = __msa_fmadd_w(_sum1, _r52, _k50); in conv7x7s2_pack1to4_msa()
379 _sum2 = __msa_fmadd_w(_sum2, _r54, _k50); in conv7x7s2_pack1to4_msa()
380 _sum3 = __msa_fmadd_w(_sum3, _r56, _k50); in conv7x7s2_pack1to4_msa()
587 v4f32 _k50 = (v4f32)__msa_ld_w(kptr, 0); in conv7x7s2_pack1to4_msa() local
600 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r5, 0), _k50); 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 | 264 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
275 _sum00 = _mm256_fmadd_ps(_r120, _k50, _sum00); in conv3x3s1_pack8_avx()
293 _sum10 = _mm256_fmadd_ps(_r130, _k50, _sum10); in conv3x3s1_pack8_avx()
595 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
606 _sum0 = _mm256_fmadd_ps(_r120, _k50, _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 | 264 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
275 _sum00 = _mm256_fmadd_ps(_r120, _k50, _sum00); in conv3x3s1_pack8_avx()
293 _sum10 = _mm256_fmadd_ps(_r130, _k50, _sum10); in conv3x3s1_pack8_avx()
595 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
606 _sum0 = _mm256_fmadd_ps(_r120, _k50, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 264 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
275 _sum00 = _mm256_fmadd_ps(_r120, _k50, _sum00); in conv3x3s1_pack8_avx()
293 _sum10 = _mm256_fmadd_ps(_r130, _k50, _sum10); in conv3x3s1_pack8_avx()
595 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
606 _sum0 = _mm256_fmadd_ps(_r120, _k50, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 264 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
275 _sum00 = _mm256_comp_fmadd_ps(_r120, _k50, _sum00); in conv3x3s1_pack8_avx()
293 _sum10 = _mm256_comp_fmadd_ps(_r130, _k50, _sum10); in conv3x3s1_pack8_avx()
595 __m256 _k50 = _mm256_loadu_ps(kptr); in conv3x3s1_pack8_avx() local
606 _sum0 = _mm256_comp_fmadd_ps(_r120, _k50, _sum0); in conv3x3s1_pack8_avx()
|