ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-16 10:59:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

change build

Browse Source 
[ROCm/composable_kernel commit: c82b833d8e]
This commit is contained in:
Chao Liu
2019-06-12 10:47:25 -05:00
parent 42e9613bb8
commit 11c6b2ab9a
10 changed files with 89 additions and 56 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
script/cmake-cuda.sh
View File

@@ -12,7 +12,7 @@ cmake
-D CMAKE_CXX_COMPILER=clang++ \
-D CMAKE_BUILD_TYPE=Release \
-D CMAKE_VERBOSE_MAKEFILE:BOOL=ON \
-D DEVICE_BACKEND=CUDA \
-D DEVICE_BACKEND=NVIDIA \
-D BOOST_ROOT="/package/install/boost_1.67.0" \
-D CUDA_COMMON_INCLUDE_DIR="/home/chao/code/test_feature/cuda_common/cuda_10.0_common/inc" \
-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" \

12
script/cmake-hip.sh
View File

@@ -4,16 +4,16 @@ rm -f CMakeCache.txt
rm -f *.cmake
rm -rf CMakeFiles
MY_PROJECT_SOURCE=/home/chao/code/modular_convolution
MY_PROJECT_SOURCE=../../../
MY_PROJECT_INSTALL=../install.dir
cmake \
-D CMAKE_INSTALL_PREFIX=${MY_PROJECT_INSTALL} \
-D CMAKE_BUILD_TYPE=Release \
-D DEVICE_BACKEND="HIP" \
-D HIP_HIPCC_FLAGS="${HIP_HIPCC_FLAGS} -gline-tables-only" \
-D CMAKE_CXX_FLAGS="-gline-tables-only" \
-D CMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc \
-D CMAKE_PREFIX_PATH="/opt/rocm;/home/package/build/mlopen_dep" \
-D DEVICE_BACKEND="AMD" \
-D HIP_HIPCC_FLAGS="${HIP_HIPCC_FLAGS} -gline-tables-only -v" \
-D CMAKE_CXX_FLAGS="-gline-tables-only --amdgpu-target=gfx906" \
-D CMAKE_CXX_COMPILER=/opt/rocm/hip/bin/hipcc \
-D CMAKE_PREFIX_PATH="/opt/rocm" \
-D CMAKE_VERBOSE_MAKEFILE:BOOL=ON \
${MY_PROJECT_SOURCE}

3
script/trace.sh Executable file
View File

@@ -0,0 +1,3 @@
#!/bin/bash
/root/workspace/rocprofiler_pkg/bin/rpl_run.sh --timestamp on -i /root/workspace/rocprofiler_pkg/input.xml -d ./trace ./driver/driver 0 10