tweaking

driver/CMakeLists.txt

@@ -24,9 +24,9 @@ elseif(DEVICE_BACKEND STREQUAL "NVIDIA")
     set(CONV_BWD_DATA_SOURCE src/conv_bwd_data_driver.cu)
 endif()
 
-add_executable(conv ${CONV_SOURCE}) 
-add_executable(col2im ${COL2IM_SOURCE}) 
-add_executable(conv_bwd_data ${CONV_BWD_DATA_SOURCE}) 
-target_link_libraries(conv PRIVATE host)
-target_link_libraries(col2im PRIVATE host)
-target_link_libraries(conv_bwd_data PRIVATE host)
+add_executable(conv_driver ${CONV_SOURCE}) 
+add_executable(col2im_driver ${COL2IM_SOURCE}) 
+add_executable(conv_bwd_data_driver ${CONV_BWD_DATA_SOURCE}) 
+target_link_libraries(conv_driver PRIVATE host)
+target_link_libraries(col2im_driver PRIVATE host)
+target_link_libraries(conv_bwd_data_driver PRIVATE host)

driver/include/device_convolution_backward_data_implicit_gemm_v2r1_nchw_kcyx_nkhw.hpp

@@ -51,7 +51,7 @@ void device_convolution_backward_data_implicit_gemm_v2r1_nchw_kcyx_nkhw(InDesc i
     wei_kcyx_device_buf.ToDevice(wei_kcyx.mData.data());
     out_nkhw_device_buf.ToDevice(out_nkhw.mData.data());
 
-#if 1
+#if 0
     // BlockSize = 256, each thread hold 64 data
     constexpr index_t BlockSize = 256;
 
@@ -80,6 +80,36 @@ void device_convolution_backward_data_implicit_gemm_v2r1_nchw_kcyx_nkhw(InDesc i
     constexpr index_t GemmBBlockCopySrcDataPerRead_GemmN  = 1;
     constexpr index_t GemmBBlockCopyDstDataPerWrite_GemmN = 1;
 
+    constexpr index_t GemmCThreadCopyDstDataPerWrite_GemmN1 = 1;
+#elif 1
+    // BlockSize = 256, each thread hold 64 data
+    constexpr index_t BlockSize = 256;
+
+    constexpr index_t GemmMPerBlock              = 128;
+    constexpr index_t GemmNPerBlock              = 128;
+    constexpr index_t GemmKPerBlock              = 8;
+    constexpr index_t GemmMPerThreadSubC         = 4;
+    constexpr index_t GemmNPerThreadSubC         = 4;
+    constexpr index_t GemmMLevel0Cluster         = 4;
+    constexpr index_t GemmNLevel0Cluster         = 4;
+    constexpr index_t GemmMLevel1Cluster         = 4;
+    constexpr index_t GemmNLevel1Cluster         = 4;
+    constexpr index_t GemmKPerThreadLoop         = 1;
+    constexpr index_t GemmThreadGemmDataPerReadM = 4;
+    constexpr index_t GemmThreadGemmDataPerReadN = 4;
+
+    using GemmABlockCopyThreadSliceLengths_GemmK_GemmM   = Sequence<1, 4>;
+    using GemmABlockCopyThreadClusterLengths_GemmK_GemmM = Sequence<8, 32>;
+
+    constexpr index_t GemmABlockCopySrcDataPerRead_GemmM  = 1;
+    constexpr index_t GemmABlockCopyDstDataPerWrite_GemmM = 1;
+
+    using GemmBBlockCopyThreadSliceLengths_GemmK_GemmN   = Sequence<1, 4>;
+    using GemmBBlockCopyThreadClusterLengths_GemmK_GemmN = Sequence<8, 32>;
+
+    constexpr index_t GemmBBlockCopySrcDataPerRead_GemmN  = 1;
+    constexpr index_t GemmBBlockCopyDstDataPerWrite_GemmN = 1;
+
     constexpr index_t GemmCThreadCopyDstDataPerWrite_GemmN1 = 1;
 #elif 1
     // BlockSize = 256, each thread hold 64 data

driver/src/conv_bwd_data_driver.cpp

@@ -34,6 +34,21 @@ int main(int argc, char* argv[])
     using ConvStrides   = Sequence<1, 1>;
     using ConvDilations = Sequence<1, 1>;
 
+    using LeftPads  = Sequence<0, 0>;
+    using RightPads = Sequence<0, 0>;
+#elif 1
+    // 3x3, 28x28
+    constexpr index_t N  = 64;
+    constexpr index_t C  = 256;
+    constexpr index_t HI = 28;
+    constexpr index_t WI = 28;
+    constexpr index_t K  = 256;
+    constexpr index_t Y  = 3;
+    constexpr index_t X  = 3;
+
+    using ConvStrides   = Sequence<1, 1>;
+    using ConvDilations = Sequence<1, 1>;
+
     using LeftPads  = Sequence<0, 0>;
     using RightPads = Sequence<0, 0>;
 #elif 0
@@ -109,21 +124,6 @@ int main(int argc, char* argv[])
     using ConvStrides   = Sequence<1, 1>;
     using ConvDilations = Sequence<1, 1>;
 
-    using LeftPads  = Sequence<0, 0>;
-    using RightPads = Sequence<0, 0>;
-#elif 0
-    // 3x3 filter, 2x2 stride, 35x35 input, 17x17 output
-    constexpr index_t N  = 128;
-    constexpr index_t C  = 1024;
-    constexpr index_t HI = 35;
-    constexpr index_t WI = 35;
-    constexpr index_t K  = 1024;
-    constexpr index_t Y  = 3;
-    constexpr index_t X  = 3;
-
-    using ConvStrides   = Sequence<2, 2>;
-    using ConvDilations = Sequence<1, 1>;
-
     using LeftPads  = Sequence<0, 0>;
     using RightPads = Sequence<0, 0>;
 #elif 0
@@ -156,7 +156,7 @@ int main(int argc, char* argv[])
 
     using LeftPads  = Sequence<3, 0>;
     using RightPads = Sequence<3, 0>;
-#elif 1
+#elif 0
     // 1x7 filter, 0x3 pad, 17x17 input
     constexpr index_t N  = 128;
     constexpr index_t C  = 1024;
@@ -171,6 +171,21 @@ int main(int argc, char* argv[])
 
     using LeftPads  = Sequence<0, 3>;
     using RightPads = Sequence<0, 3>;
+#elif 1
+    // 3x3 filter, 2x2 stride, 35x35 input, 17x17 output
+    constexpr index_t N  = 128;
+    constexpr index_t C  = 1024;
+    constexpr index_t HI = 35;
+    constexpr index_t WI = 35;
+    constexpr index_t K  = 1024;
+    constexpr index_t Y  = 3;
+    constexpr index_t X  = 3;
+
+    using ConvStrides   = Sequence<2, 2>;
+    using ConvDilations = Sequence<1, 1>;
+
+    using LeftPads  = Sequence<0, 0>;
+    using RightPads = Sequence<0, 0>;
 #endif
 
     constexpr auto in_nchw_desc  = make_native_tensor_descriptor_packed(Sequence<N, C, HI, WI>{});
@@ -216,7 +231,7 @@ int main(int argc, char* argv[])
 
 #if 0
     device_convolution_backward_data_implicit_gemm_v1r1_nchw_kcyx_nkhw
-#elif 1
+#elif 0
     device_convolution_backward_data_implicit_gemm_v1r2_nchw_kcyx_nkhw
 #else
     device_convolution_backward_data_implicit_gemm_v2r1_nchw_kcyx_nkhw

driver/src/conv_driver.cpp

@@ -29,7 +29,7 @@ int main(int argc, char* argv[])
 {
     using namespace ck;
 
-#if 0
+#if 1
     // 1x1
     constexpr index_t N  = 256;
     constexpr index_t C  = 1024;

script/cmake-cuda_docker.sh

@@ -15,11 +15,9 @@ cmake
 -D CMAKE_VERBOSE_MAKEFILE:BOOL=ON                                                           \
 -D DEVICE_BACKEND=NVIDIA                                                                    \
 -D CUDA_COMMON_INCLUDE_DIR="/root/NVIDIA_CUDA-10.1_Samples/common/inc"                      \
--D CMAKE_CUDA_FLAGS="-ccbin clang++-6.0 -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep  -Xptxas -v -gencode=arch=compute_60,code=sm_60 -Xptxas -v -gencode=arch=compute_70,code=sm_70" \
+-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61 -Xptxas -v -maxrregcount=128" \
 ${MY_PROJECT_SOURCE}
 
 
 #-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61" \
-#-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61 -Xptxas -v -Xptxas -v -maxrregcount=128" \
-#-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61 -Xptxas -v -gencode=arch=compute_70,code=sm_70" \
-#-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61 -Xptxas -v -gencode=arch=compute_70,code=sm_70 -Xptxas -v -maxrregcount=128" \
+#-D CMAKE_CUDA_FLAGS="-ccbin clang++ -m64 -Xcompiler -fopenmp -lineinfo --source-in-ptx -keep -Xptxas -v -gencode=arch=compute_61,code=sm_61 -Xptxas -v -maxrregcount=128" \

script/extract_asm-cuda.sh

-cuobjdump -xelf sm_60 ./driver/driver && nvdisasm --print-code -g driver.sm_60.cubin > driver.sm_60.asm
-cuobjdump -xelf sm_61 ./driver/driver && nvdisasm --print-code -g driver.sm_61.cubin > driver.sm_61.asm
-cuobjdump -xelf sm_70 ./driver/driver && nvdisasm --print-code -g driver.sm_70.cubin > driver.sm_70.asm
+DRIVER=$1
+ARCH=$2
+cuobjdump -xelf $ARCH ./driver/$DRIVER && nvdisasm --print-code -g $DRIVER.$ARCH.cubin > $DRIVER.$ARCH.asm