TL/UCP: select exact allgather radix schedules #4137
Workflow file for this run
This file contains hidden or 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: Linter-NVIDIA | |
| on: [push, pull_request] | |
| concurrency: | |
| group: lint-cuda-${{ github.event.pull_request.number || github.ref }} | |
| cancel-in-progress: true | |
| env: | |
| OPEN_UCX_LINK: https://github.com/openucx/ucx | |
| OPEN_UCX_BRANCH: master | |
| HPCX_LINK: https://content.mellanox.com/hpc/hpc-x/v2.25.1_cuda13/hpcx-v2.25.1-gcc-doca_ofed-ubuntu24.04-cuda13-x86_64.tbz | |
| CLANG_VER: 17 | |
| MLNX_OFED_VER: 24.10-4.1.4.0 | |
| CUDA_VER: 13-1 | |
| LIBRARY_PATH: /tmp/ucx/install/lib | |
| LD_LIBRARY_PATH: /tmp/ucx/install/lib | |
| jobs: | |
| clang-tidy: | |
| runs-on: ubuntu-24.04 | |
| name: clang-tidy (CUDA) | |
| steps: | |
| - name: Install dependencies | |
| run: | | |
| sudo apt-get update | |
| sudo apt-get install -y --no-install-recommends wget gpg | |
| # Setup LLVM repository | |
| sudo mkdir -p /etc/apt/keyrings | |
| wget -O - https://apt.llvm.org/llvm-snapshot.gpg.key | sudo gpg --dearmor -o /etc/apt/keyrings/llvm.gpg | |
| echo "deb [signed-by=/etc/apt/keyrings/llvm.gpg] http://apt.llvm.org/noble/ llvm-toolchain-noble-${CLANG_VER} main" | sudo tee /etc/apt/sources.list.d/llvm.list | |
| sudo apt-get update | |
| sudo apt-get install -y --no-install-recommends clang-tidy-${CLANG_VER} bear clang-${CLANG_VER} clang++-${CLANG_VER} | |
| - name: Cache MLNX_OFED tarball | |
| id: cache-ofed | |
| uses: actions/cache@v4 | |
| with: | |
| path: MLNX_OFED_LINUX-${{ env.MLNX_OFED_VER }}-ubuntu24.04-x86_64.tgz | |
| key: mlnx-ofed-${{ env.MLNX_OFED_VER }}-ubuntu24.04 | |
| - name: Download MLNX_OFED | |
| if: steps.cache-ofed.outputs.cache-hit != 'true' | |
| run: wget --no-verbose http://content.mellanox.com/ofed/MLNX_OFED-${MLNX_OFED_VER}/MLNX_OFED_LINUX-${MLNX_OFED_VER}-ubuntu24.04-x86_64.tgz | |
| - name: Install extra rdma dependencies | |
| run: | | |
| sudo tar -xzf MLNX_OFED_LINUX-${MLNX_OFED_VER}-ubuntu24.04-x86_64.tgz | |
| sudo chmod -R a+rwx MLNX_OFED_LINUX-${MLNX_OFED_VER}-ubuntu24.04-x86_64 | |
| sudo MLNX_OFED_LINUX-${MLNX_OFED_VER}-ubuntu24.04-x86_64/mlnxofedinstall --skip-unsupported-devices-check --user-space-only --without-fw-update --force --basic -v | |
| - name: Install extra cuda dependencies | |
| run: | | |
| wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/x86_64/cuda-keyring_1.1-1_all.deb | |
| sudo dpkg -i cuda-keyring_1.1-1_all.deb | |
| sudo apt-get update | |
| sudo apt-get install -y --no-install-recommends cuda-cudart-dev-${CUDA_VER} cuda-nvcc-${CUDA_VER} cuda-nvml-dev-${CUDA_VER} | |
| - name: Resolve UCX HEAD commit | |
| id: ucx-sha | |
| run: echo "sha=$(git ls-remote ${OPEN_UCX_LINK} refs/heads/${OPEN_UCX_BRANCH} | cut -f1)" >> $GITHUB_OUTPUT | |
| - name: Cache UCX | |
| id: cache-ucx | |
| uses: actions/cache@v4 | |
| with: | |
| path: /tmp/ucx/install | |
| key: ucx-nvidia-${{ steps.ucx-sha.outputs.sha }}-${{ hashFiles('.github/workflows/clang-tidy-nvidia.yaml') }} | |
| - name: Get UCX | |
| if: steps.cache-ucx.outputs.cache-hit != 'true' | |
| run: | | |
| rm -rf /tmp/ucx | |
| git clone --depth 1 ${OPEN_UCX_LINK} -b ${OPEN_UCX_BRANCH} /tmp/ucx | |
| - name: Build UCX | |
| if: steps.cache-ucx.outputs.cache-hit != 'true' | |
| run: | | |
| cd /tmp/ucx && ./autogen.sh | |
| CC=clang-${CLANG_VER} CXX=clang++-${CLANG_VER} ./contrib/configure-release --without-java --without-go --disable-numa --prefix $PWD/install | |
| make -j install | |
| - name: Cache HPCX | |
| id: cache-hpcx | |
| uses: actions/cache@v4 | |
| with: | |
| path: /tmp/hpcx | |
| key: hpcx-v2.25.1-${{ hashFiles('.github/workflows/clang-tidy-nvidia.yaml') }} | |
| - name: Download HPCX | |
| if: steps.cache-hpcx.outputs.cache-hit != 'true' | |
| run: | | |
| cd /tmp | |
| wget --no-verbose ${HPCX_LINK} | |
| tar xjf hpcx-v2.25.1-gcc-doca_ofed-ubuntu24.04-cuda13-x86_64.tbz | |
| mv hpcx-v2.25.1-gcc-doca_ofed-ubuntu24.04-cuda13-x86_64 hpcx | |
| - uses: actions/checkout@v6 | |
| - name: Build UCC | |
| run: | | |
| ./autogen.sh | |
| CC=clang-${CLANG_VER} CXX=clang++-${CLANG_VER} ./configure --with-tls=ucp,mlx5,cuda,self,sharp --with-sharp=/tmp/hpcx/sharp --prefix=/tmp/ucc/install --with-ucx=/tmp/ucx/install --with-cuda=/usr/local/cuda --with-nvcc-gencode="-gencode=arch=compute_80,code=sm_80" --enable-assert | |
| bear --output /tmp/compile_commands.json -- make -j | |
| - name: Run clang-tidy | |
| run: | | |
| echo "Workspace: ${GITHUB_WORKSPACE}" | |
| cd ${GITHUB_WORKSPACE} | |
| run-clang-tidy-${CLANG_VER} -p /tmp/ -header-filter='^(?!.*(${GITHUB_WORKSPACE}/src/components/ec/cuda/kernel/.*|${GITHUB_WORKSPACE}/src/components/mc/cuda/kernel/.*)).*$' "^(?!.*\.cu$).*$" 2>&1 | tee /tmp/clang_tidy.log | |
| nerrors=$(grep 'error:' /tmp/clang_tidy.log | wc -l) | |
| if [ $nerrors -ne 0 ]; then | |
| exit 125; | |
| fi |