Searched refs:output0_tm3_4 (Results 1 – 5 of 5) sorted by relevance
/dports/graphics/waifu2x-ncnn-vulkan/waifu2x-ncnn-vulkan-20210521/src/ncnn/src/layer/arm/ |
H A D | convolution_3x3.h | 4015 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 7); in conv3x3s1_winograd64_neon4() local 4027 float32x4_t _output0_tm3_4567 = vld1q_f32(output0_tm3_4); in conv3x3s1_winograd64_neon4() 4090 output0_tm3_4 += out0_tm.w * tiles * 2 * 4; in conv3x3s1_winograd64_neon4() 4325 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon4() 4335 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon4() 7438 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 56); in conv3x3s1_winograd64_neon5() 7669 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon5() 7679 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon5()
|
/dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/arm/ |
H A D | convolution_3x3.h | 4015 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 7); in conv3x3s1_winograd64_neon4() local 4027 float32x4_t _output0_tm3_4567 = vld1q_f32(output0_tm3_4); in conv3x3s1_winograd64_neon4() 4090 output0_tm3_4 += out0_tm.w * tiles * 2 * 4; in conv3x3s1_winograd64_neon4() 4325 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon4() 4335 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon4() 7438 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 56); in conv3x3s1_winograd64_neon5() 7669 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon5() 7679 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon5()
|
/dports/misc/ncnn/ncnn-20211208/src/layer/arm/ |
H A D | convolution_3x3.h | 4015 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 7); in conv3x3s1_winograd64_neon4() local 4027 float32x4_t _output0_tm3_4567 = vld1q_f32(output0_tm3_4); in conv3x3s1_winograd64_neon4() 4090 output0_tm3_4 += out0_tm.w * tiles * 2 * 4; in conv3x3s1_winograd64_neon4() 4325 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon4() 4335 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon4() 7438 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 56); in conv3x3s1_winograd64_neon5() 7669 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon5() 7679 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon5()
|
/dports/graphics/vapoursynth-waifu2x-ncnn-vulkan/vapoursynth-waifu2x-ncnn-vulkan-r4/deps/ncnn/src/layer/arm/ |
H A D | convolution_3x3.h | 4015 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 7); in conv3x3s1_winograd64_neon4() local 4027 float32x4_t _output0_tm3_4567 = vld1q_f32(output0_tm3_4); in conv3x3s1_winograd64_neon4() 4090 output0_tm3_4 += out0_tm.w * tiles * 2 * 4; in conv3x3s1_winograd64_neon4() 4325 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon4() 4335 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon4() 7438 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 56); in conv3x3s1_winograd64_neon5() 7669 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon5() 7679 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon5()
|
/dports/graphics/realsr-ncnn-vulkan/realsr-ncnn-vulkan-20210210/src/ncnn/src/layer/arm/ |
H A D | convolution_3x3.h | 4015 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 7); in conv3x3s1_winograd64_neon4() local 4027 float32x4_t _output0_tm3_4567 = vld1q_f32(output0_tm3_4); in conv3x3s1_winograd64_neon4() 4090 output0_tm3_4 += out0_tm.w * tiles * 2 * 4; in conv3x3s1_winograd64_neon4() 4325 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon4() 4335 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon4() 7438 const float* output0_tm3_4 = out0_tm.row(i * w_tm / 8 + j + tiles * 56); in conv3x3s1_winograd64_neon5() 7669 "=r"(output0_tm3_4) // %9 in conv3x3s1_winograd64_neon5() 7679 "9"(output0_tm3_4), in conv3x3s1_winograd64_neon5()
|