updated READMEs for v4 #52
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: compile | |
| on: | |
| push: | |
| branches: | |
| - v4 | |
| pull_request: | |
| branches: | |
| - v4 | |
| jobs: | |
| # test only compilation succeeds (no execution) | |
| build-test: | |
| name: > | |
| ${{ matrix.os == 'ubuntu-latest' && 'Linux' || matrix.os == 'macos-latest' && 'MacOS' || 'Windows' }} | |
| [${{ matrix.precision }}] | |
| ${{ matrix.omp == 'ON' && 'OMP' || '' }} | |
| ${{ matrix.mpi == 'ON' && 'MPI' || '' }} | |
| ${{ matrix.cuda == 'ON' && 'CUDA' || '' }} | |
| ${{ matrix.hip == 'ON' && 'HIP' || '' }} | |
| ${{ matrix.cuquantum == 'ON' && 'CUQ' || '' }} | |
| runs-on: ${{ matrix.os }} | |
| strategy: | |
| # continue other jobs if any fail | |
| fail-fast: false | |
| # limit concurrent jobs to avoid running out of disk | |
| max-parallel: 64 | |
| # compile QuEST with all combinations of below flags | |
| matrix: | |
| os: [ubuntu-latest, macos-latest, windows-latest] | |
| precision: [1, 2, 4] | |
| omp: [ON, OFF] | |
| mpi: [ON, OFF] | |
| cuda: [ON, OFF] | |
| hip: [ON, OFF] | |
| cuquantum: [ON, OFF] | |
| # disable deprecated API on MSVC, and assign unique compilers, | |
| # so that we can concisely consult e.g. matrix.compiler=='cl' | |
| # instead of hardcoding the operating system label. Note that | |
| # ultimately env.compiler, not matrix.compiler, will be set/used | |
| include: | |
| - os: ubuntu-latest | |
| compiler: g++ | |
| deprecated: ON | |
| - os: macos-latest | |
| compiler: clang++ | |
| deprecated: ON | |
| - os: windows-latest | |
| compiler: cl | |
| deprecated: OFF | |
| exclude: | |
| # cannot simultaneously use CUDA and HIP | |
| - cuda: ON | |
| hip: ON | |
| # cannot use GPU with quad-precision | |
| - cuda: ON | |
| precision: 4 | |
| - hip: ON | |
| precision: 4 | |
| # cannot use cuquantum without CUDA | |
| - cuda: OFF | |
| cuquantum: ON | |
| # cannot use GPU on MacOS | |
| - cuda: ON | |
| os: macos-latest | |
| - hip: ON | |
| os: macos-latest | |
| # cannot use cuquantum on Windows or MacOS | |
| - cuquantum: ON | |
| os: windows-latest | |
| - cuquantum: ON | |
| os: macos-latest | |
| # cannot presently install HIP on Windows CI (times out) | |
| - hip: ON | |
| os: windows-latest | |
| # cannot presently compile HIP + MPI; the linker fails with | |
| # "undefined reference to 'vtable for thrust::system::system_error' | |
| # (see failed attempted solutions in PR #554) | |
| - hip: ON | |
| mpi: ON | |
| # constants | |
| env: | |
| build_dir: "build" | |
| cuda_arch: 70 | |
| hip_arch: gfx801 | |
| rocm_path: /opt/rocm/bin | |
| # perform the job | |
| steps: | |
| # free space for big-chungus ROCm compiler | |
| - name: Free disk space | |
| if: ${{ matrix.hip == 'ON' }} | |
| uses: jlumbroso/free-disk-space@main | |
| with: | |
| tool-cache: false | |
| # download QuEST | |
| - name: Get QuEST | |
| uses: actions/checkout@v4 | |
| # obtain OMP for Clang multithreading | |
| - name: Setup libomp | |
| if: ${{ matrix.compiler == 'clang++' && matrix.omp == 'ON' }} | |
| run: > | |
| brew install libomp; | |
| echo "OpenMP_ROOT=$(brew --prefix)/opt/libomp" >> $GITHUB_ENV | |
| # obtain MPI for distribution | |
| - name: Setup MPI | |
| if: ${{ matrix.mpi == 'ON' }} | |
| uses: mpi4py/setup-mpi@v1 | |
| # obtain CUDA for GPU acceleration (cuQuantum needs cuBLAS) | |
| - name: Setup CUDA (non-MSVC) | |
| if: ${{ matrix.cuda == 'ON' && matrix.compiler != 'cl' }} | |
| uses: Jimver/[email protected] | |
| with: | |
| method: network | |
| sub-packages: '["nvcc", "cudart", "thrust"]' | |
| non-cuda-sub-packages: ${{ matrix.cuquantum == 'ON' && '["libcublas"]' || '[]' }} | |
| # bug in Nsight for MSVC breaks above action for windows-2022, so we revert to script | |
| - name: Setup CUDA (MSVC) | |
| if: ${{ matrix.cuda == 'ON' && matrix.compiler == 'cl' }} | |
| shell: pwsh | |
| run: .github/scripts/setup_cuda.ps1 | |
| env: | |
| INPUT_CUDA_VERSION: 12.5.0 | |
| # obtain cuQuantum on linux | |
| - name: Setup cuQuantum | |
| if: ${{ matrix.cuquantum == 'ON' }} | |
| run: > | |
| wget https://developer.download.nvidia.com/compute/cuquantum/redist/cuquantum/linux-x86_64/cuquantum-linux-x86_64-24.08.0.5_cuda12-archive.tar.xz; | |
| tar -xvf cuquantum-linux-x86_64-24.08.0.5_cuda12-archive.tar.xz; | |
| echo "CUQUANTUM_ROOT=cuquantum-linux-x86_64-24.08.0.5_cuda12-archive" >> $GITHUB_ENV | |
| # obtain ROCm for HIP acceleration on Linux | |
| - name: Install ROCm | |
| if: ${{ matrix.hip == 'ON' }} | |
| run: | | |
| sudo apt install "linux-headers-$(uname -r)" "linux-modules-extra-$(uname -r)" | |
| sudo apt install python3-setuptools python3-wheel | |
| sudo usermod -a -G render,video $USER | |
| wget https://repo.radeon.com/amdgpu-install/6.3.3/ubuntu/noble/amdgpu-install_6.3.60303-1_all.deb | |
| sudo apt install ./amdgpu-install_6.3.60303-1_all.deb | |
| sudo apt update | |
| sudo apt install amdgpu-dkms rocm | |
| echo "${{ env.rocm_path }}" >> $GITHUB_PATH | |
| # invoke cmake, disabling LTO (it duplicates symbols with CUDA + MPI) | |
| - name: Configure CMake | |
| run: > | |
| cmake -B ${{ env.build_dir }} | |
| -DCMAKE_VERBOSE_MAKEFILE=ON | |
| -DBUILD_EXAMPLES=ON | |
| -DENABLE_TESTING=ON | |
| -DFLOAT_PRECISION=${{ matrix.precision }} | |
| -DENABLE_DEPRECATED_API=${{ matrix.deprecated }} | |
| -DENABLE_MULTITHREADING=${{ matrix.omp }} | |
| -DENABLE_DISTRIBUTION=${{ matrix.mpi }} | |
| -DENABLE_CUDA=${{ matrix.cuda }} | |
| -DENABLE_HIP=${{ matrix.hip }} | |
| -DENABLE_CUQUANTUM=${{ matrix.cuquantum }} | |
| -DCMAKE_CUDA_ARCHITECTURES=${{ env.cuda_arch }} | |
| -DCMAKE_HIP_ARCHITECTURES=${{ env.hip_arch }} | |
| -DCMAKE_CXX_COMPILER=${{ matrix.compiler }} | |
| -DCMAKE_CXX_FLAGS=${{ matrix.mpi == 'ON' && matrix.cuda == 'ON' && '-fno-lto' || '' }} | |
| # force 'Release' build (needed by MSVC to enable optimisations) | |
| - name: Compile | |
| run: cmake --build ${{ env.build_dir }} --config Release | |
| # TODO: | |
| # - ARM compilation (on paid linux and windows runners) | |
| # - CUDA-aware MPI?!?!?! |