| /dports/misc/ncnn/ncnn-20211208/src/layer/riscv/ |
| H A D | convolution_7x7_pack1ton.h | 330 _sum0 = vfmacc_vf_f32m1(_sum0, r3[6], _k36, vl); in conv7x7s2_pack1ton_rvv()
331 _sum1 = vfmacc_vf_f32m1(_sum1, r3[8], _k36, vl); in conv7x7s2_pack1ton_rvv()
332 _sum2 = vfmacc_vf_f32m1(_sum2, r3[10], _k36, vl); in conv7x7s2_pack1ton_rvv()
333 _sum3 = vfmacc_vf_f32m1(_sum3, r3[12], _k36, vl); in conv7x7s2_pack1ton_rvv()
334 _sum4 = vfmacc_vf_f32m1(_sum4, r3[14], _k36, vl); in conv7x7s2_pack1ton_rvv()
335 _sum5 = vfmacc_vf_f32m1(_sum5, r3[16], _k36, vl); in conv7x7s2_pack1ton_rvv()
336 _sum6 = vfmacc_vf_f32m1(_sum6, r3[18], _k36, vl); in conv7x7s2_pack1ton_rvv()
337 _sum7 = vfmacc_vf_f32m1(_sum7, r3[20], _k36, vl); in conv7x7s2_pack1ton_rvv()
717 _sum0 = vfmacc_vf_f32m1(_sum0, r3[6], _k36, vl); in conv7x7s2_pack1ton_rvv()
718 _sum1 = vfmacc_vf_f32m1(_sum1, r3[8], _k36, vl); in conv7x7s2_pack1ton_rvv()
[all …]
|
| H A D | convolution_7x7_pack1ton_fp16s.h | 330 _sum0 = vfmacc_vf_f16m1(_sum0, r3[6], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
331 _sum1 = vfmacc_vf_f16m1(_sum1, r3[8], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
332 _sum2 = vfmacc_vf_f16m1(_sum2, r3[10], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
333 _sum3 = vfmacc_vf_f16m1(_sum3, r3[12], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
334 _sum4 = vfmacc_vf_f16m1(_sum4, r3[14], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
335 _sum5 = vfmacc_vf_f16m1(_sum5, r3[16], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
336 _sum6 = vfmacc_vf_f16m1(_sum6, r3[18], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
337 _sum7 = vfmacc_vf_f16m1(_sum7, r3[20], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
717 _sum0 = vfmacc_vf_f16m1(_sum0, r3[6], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
718 _sum1 = vfmacc_vf_f16m1(_sum1, r3[8], _k36, vl); in conv7x7s2_pack1ton_fp16sa_rvv()
[all …]
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/mips/ |
| H A D | convolution_7x7_pack1to4.h | 241 v4f32 _k36 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
287 _sum0 = __msa_fmadd_w(_sum0, _r36, _k36); in conv7x7s2_pack1to4_msa()
288 _sum1 = __msa_fmadd_w(_sum1, _r38, _k36); in conv7x7s2_pack1to4_msa()
289 _sum2 = __msa_fmadd_w(_sum2, _r3a, _k36); in conv7x7s2_pack1to4_msa()
290 _sum3 = __msa_fmadd_w(_sum3, _r3c, _k36); in conv7x7s2_pack1to4_msa()
551 v4f32 _k36 = (v4f32)__msa_ld_w(kptr + 4 * 6, 0); in conv7x7s2_pack1to4_msa() local
564 _sum0 = __msa_fmadd_w(_sum0, (v4f32)__msa_splati_w(_r3n, 2), _k36); 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 | 194 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
205 _sum00 = _mm256_fmadd_ps(_r106, _k36, _sum00); in conv3x3s1_pack8_avx()
223 _sum10 = _mm256_fmadd_ps(_r116, _k36, _sum10); in conv3x3s1_pack8_avx()
543 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
554 _sum0 = _mm256_fmadd_ps(_r106, _k36, _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 | 194 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
205 _sum00 = _mm256_fmadd_ps(_r106, _k36, _sum00); in conv3x3s1_pack8_avx()
223 _sum10 = _mm256_fmadd_ps(_r116, _k36, _sum10); in conv3x3s1_pack8_avx()
543 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
554 _sum0 = _mm256_fmadd_ps(_r106, _k36, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 194 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
205 _sum00 = _mm256_fmadd_ps(_r106, _k36, _sum00); in conv3x3s1_pack8_avx()
223 _sum10 = _mm256_fmadd_ps(_r116, _k36, _sum10); in conv3x3s1_pack8_avx()
543 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
554 _sum0 = _mm256_fmadd_ps(_r106, _k36, _sum0); in conv3x3s1_pack8_avx()
|
| /dports/misc/ncnn/ncnn-20211208/src/layer/x86/ |
| H A D | convolution_3x3_pack8.h | 194 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
205 _sum00 = _mm256_comp_fmadd_ps(_r106, _k36, _sum00); in conv3x3s1_pack8_avx()
223 _sum10 = _mm256_comp_fmadd_ps(_r116, _k36, _sum10); in conv3x3s1_pack8_avx()
543 __m256 _k36 = _mm256_loadu_ps(kptr + 48); in conv3x3s1_pack8_avx() local
554 _sum0 = _mm256_comp_fmadd_ps(_r106, _k36, _sum0); in conv3x3s1_pack8_avx()
|