name: build

on:
  push:
    paths-ignore:
      - '*.md'

  pull_request:
    paths-ignore:
      - '*.md'

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

jobs:
  build:
    runs-on: ubuntu-18.04
    steps:
      - uses: actions/checkout@v2
      - name: Install LLVM and Clang
        uses: KyleMayes/install-llvm-action@v1
        with:
          version: "14.0"
      - name: Download CUDA header files
        run: |
          cd ${{ github.workspace }}
          wget https://www.dropbox.com/s/r18io0zu3idke5p/cuda-header.tar.gz?dl=1
          tar -xzf 'cuda-header.tar.gz?dl=1'
          cp -r include/* runtime/threadPool/include/
      - name: Download CUDA files used for compiling NVVM IR
        run: |
          cd ${{ github.workspace }}
          wget https://www.dropbox.com/s/4pckqsjnl920gpn/cuda-10.1.tar.gz?dl=1
          tar -xzf 'cuda-10.1.tar.gz?dl=1'
      - name: Build project
        run: |
          mkdir build
          cd build
          cmake .. -DLLVM_CONFIG_PATH=`which llvm-config`
          make -j8
      - name: clone SC evaluation code
        run: |
          cd ${{ github.workspace }}
          git clone https://github.com/drcut/SC_evaluate
      - name: Download data used to verify
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox
          wget https://www.dropbox.com/s/ie2hcxw9lfoghg8/data.tar.gz?dl=1
          tar -xzf 'data.tar.gz?dl=1'
      - name: Execute the AES example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/aes
          clang++ -std=c++11 cuda/aes_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator aes_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator aes_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o aes -fPIC -no-pie -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./aes -i ../../data/aes/1KB.data -k ../../data/aes/key.data -q -v
      - name: Execute the BS example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/bs
          clang++ -std=c++11 cuda/bs_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator bs_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator bs_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o bs -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./bs -x 16777216 -q -t --chunk 512 -v
      - name: Execute the EP example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/ep
          clang++ -std=c++11 cuda/ep_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator ep_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator ep_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o ep -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./ep -q -v
      - name: Execute the FIR example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/fir
          clang++ -std=c++11 cuda/fir_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator fir_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator fir_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o fir -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./fir -q -v
      - name: Execute the GA example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/ga
          clang++ -std=c++11 cuda/ga_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator ga_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator ga_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o ga -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
      - name: Execute the HIST example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/hist
          clang++ -std=c++11 cuda/hist_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator hist_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator hist_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o hist -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./hist -q -v
      - name: Execute the KMeans example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/kmeans
          clang++ -std=c++11 cuda/kmeans_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator kmeans_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator kmeans_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o kmeans -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./kmeans -i ../../data/kmeans/100_34.txt -q -v
      - name: Execute the PR example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/Hetero-cox/src/pr
          clang++ -std=c++11 cuda/pr_cuda_benchmark.cu -I../.. --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_50 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator pr_cuda_benchmark-cuda-nvptx64-nvidia-cuda-sm_50.bc kernel.bc
          hostTranslator pr_cuda_benchmark-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o pr -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   cuda/main.cc host.o kernel.o *.cc  ../common/benchmark/*.cc ../common/command_line_option/*.cc  ../common/time_measurement/*.cc -I../.. -lpthread -lc -lx86Runtime -lthreadPool
          ./pr -i ../../data/pr/8192.data -q -v
      - name: Execute the lavaMD example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/rodinia-cox/lavaMD
          clang++ kernel/kernel_gpu_cuda_wrapper.cu --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_61 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator kernel_gpu_cuda_wrapper-cuda-nvptx64-nvidia-cuda-sm_61.bc kernel.bc
          hostTranslator kernel_gpu_cuda_wrapper-host-x86_64-unknown-linux-gnu.bc host.bc
          llc --relocation-model=pic --filetype=obj  kernel.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          g++ -o lavaMD -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool   main.c host.o kernel.o util/timer/timer.c util/num/num.c -lpthread -lc -lx86Runtime -lthreadPool -pthread
          ./lavaMD -boxes1d 10
      - name: Execute the dwt2d example
        run: |
          cd ${{ github.workspace }}/SC_evaluate/rodinia-cox/dwt2d
          clang++ -I. -I/include -fno-strict-aliasing dwt_cuda/fdwt53.cu dwt_cuda/fdwt97.cu  dwt_cuda/common.cu  dwt_cuda/rdwt97.cu  dwt_cuda/rdwt53.cu components.cu dwt.cu main.cu -c --cuda-path=${{ github.workspace }}/cuda-10.1  --cuda-gpu-arch=sm_61 -L${{ github.workspace }}/cuda-10.1/lib64 -lcudart_static -ldl -lrt -pthread -save-temps -v  || true
          export LD_LIBRARY_PATH=${{ github.workspace }}/build/runtime:${{ github.workspace }}/build/runtime/threadPool:$LD_LIBRARY_PATH
          export PATH=${{ github.workspace }}/build/compilation:$PATH
          kernelTranslator common-cuda-nvptx64-nvidia-cuda-sm_61.bc common.bc
          kernelTranslator components-cuda-nvptx64-nvidia-cuda-sm_61.bc components.bc
          kernelTranslator fdwt53-cuda-nvptx64-nvidia-cuda-sm_61.bc fdwt53.bc
          kernelTranslator dwt-cuda-nvptx64-nvidia-cuda-sm_61.bc dwt.bc
          kernelTranslator fdwt97-cuda-nvptx64-nvidia-cuda-sm_61.bc fdwt97.bc
          kernelTranslator rdwt97-cuda-nvptx64-nvidia-cuda-sm_61.bc rdwt97.bc
          kernelTranslator rdwt53-cuda-nvptx64-nvidia-cuda-sm_61.bc rdwt53.bc
          hostTranslator main-host-x86_64-unknown-linux-gnu.bc host.bc
          hostTranslator common-host-x86_64-unknown-linux-gnu.bc common_host.bc
          hostTranslator components-host-x86_64-unknown-linux-gnu.bc components_host.bc
          hostTranslator dwt-host-x86_64-unknown-linux-gnu.bc dwt_host.bc
          hostTranslator fdwt53-host-x86_64-unknown-linux-gnu.bc fdwt53_host.bc
          hostTranslator fdwt97-host-x86_64-unknown-linux-gnu.bc fdwt97_host.bc
          hostTranslator rdwt53-host-x86_64-unknown-linux-gnu.bc rdwt53_host.bc
          hostTranslator rdwt97-host-x86_64-unknown-linux-gnu.bc rdwt97_host.bc
          llc --relocation-model=pic --filetype=obj  common.bc
          llc --relocation-model=pic --filetype=obj  components.bc
          llc --relocation-model=pic --filetype=obj  fdwt53.bc
          llc --relocation-model=pic --filetype=obj  dwt.bc
          llc --relocation-model=pic --filetype=obj  host.bc
          llc --relocation-model=pic --filetype=obj  common_host.bc
          llc --relocation-model=pic --filetype=obj  components_host.bc
          llc --relocation-model=pic --filetype=obj  fdwt53_host.bc
          llc --relocation-model=pic --filetype=obj  dwt_host.bc
          llc --relocation-model=pic --filetype=obj  fdwt97_host.bc
          llc --relocation-model=pic --filetype=obj  rdwt97_host.bc
          llc --relocation-model=pic --filetype=obj  rdwt53_host.bc
          llc --relocation-model=pic --filetype=obj  fdwt97.bc
          llc --relocation-model=pic --filetype=obj  rdwt97.bc
          llc --relocation-model=pic --filetype=obj  rdwt53.bc
          g++ -o dwt2d -fPIC -no-pie -I${{ github.workspace }}/runtime/threadPool/include -L${{ github.workspace }}/build/runtime   -L${{ github.workspace }}/build/runtime/threadPool common.o components.o dwt.o fdwt53.o fdwt97.o rdwt97.o rdwt53.o host.o common_host.o components_host.o dwt_host.o fdwt53_host.o fdwt97_host.o rdwt97_host.o rdwt53_host.o -lpthread -lc -lx86Runtime -lthreadPool -pthread
          ./dwt2d 192.bmp -d 192x192 -f -5 -l 3