CI - Free-threading and Thread Sanitizer (nightly) #49
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: CI - Free-threading and Thread Sanitizer (nightly) | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.ref }} | |
| cancel-in-progress: true | |
| on: | |
| schedule: | |
| - cron: "0 5 * * *" # Daily at 05:00 UTC == 00:00 EST == 21:00 PST | |
| workflow_dispatch: # allows triggering the workflow run manually | |
| pull_request: # Automatically trigger on pull requests affecting this file | |
| branches: | |
| - main | |
| paths: | |
| - '**/workflows/tsan.yaml' | |
| - '**/workflows/tsan-suppressions*.txt' | |
| permissions: {} | |
| env: | |
| UV_DEFAULT_INDEX: "https://us-python.pkg.dev/ml-oss-artifacts-published/pypi-mirror/simple" | |
| PIP_INDEX_URL: "https://us-python.pkg.dev/ml-oss-artifacts-published/pypi-mirror/simple" | |
| jobs: | |
| tsan: | |
| runs-on: linux-x86-n4-64 | |
| container: | |
| image: index.docker.io/library/ubuntu@sha256:b359f1067efa76f37863778f7b6d0e8d911e3ee8efa807ad01fbf5dc1ef9006b # ratchet:ubuntu:24.04 | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| include: | |
| # Disabled. 3.13 has unfixed races in CPython itself; we get better signal from 3.14. | |
| # - name-prefix: "with 3.13" | |
| # python-version: "3.13" | |
| # github_branch: "3.13" | |
| # requirements_lock_name: "requirements_lock_3_13_ft" | |
| - name-prefix: "with 3.14" | |
| python-version: "3.14" | |
| github_branch: "3.14" | |
| requirements_lock_name: "requirements_lock_3_14_ft" | |
| defaults: | |
| run: | |
| shell: bash -l {0} | |
| steps: | |
| # Install git before actions/checkout as otherwise it will download the code with the GitHub | |
| # REST API and therefore any subsequent git commands will fail. | |
| - name: Install clang 18 | |
| env: | |
| DEBIAN_FRONTEND: noninteractive | |
| run: | | |
| apt update | |
| apt install -q -y clang-18 libstdc++-14-dev build-essential libssl-dev \ | |
| zlib1g-dev libbz2-dev libreadline-dev libsqlite3-dev curl git \ | |
| libncursesw5-dev xz-utils tk-dev libxml2-dev libxmlsec1-dev \ | |
| libffi-dev liblzma-dev file zip | |
| - uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0 | |
| with: | |
| path: openarchx | |
| persist-credentials: false | |
| - uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0 | |
| with: | |
| repository: numpy/numpy | |
| path: numpy | |
| submodules: true | |
| persist-credentials: false | |
| - name: Get year & week number | |
| id: get-date | |
| run: echo "date=$(/bin/date "+%Y-%U")" >> $GITHUB_OUTPUT | |
| shell: bash -l {0} | |
| - name: Restore cached TSAN CPython ${{ matrix.python-version }} | |
| id: cache-cpython-tsan-restore | |
| uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0 | |
| with: | |
| path: | | |
| ./python-tsan.tgz | |
| key: ${{ runner.os }}-cpython-tsan-${{ matrix.python-version }}-${{ steps.get-date.outputs.date }} | |
| - uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0 | |
| if: steps.cache-cpython-tsan-restore.outputs.cache-hit != 'true' | |
| with: | |
| repository: python/cpython | |
| path: cpython | |
| ref: ${{ matrix.github_branch }} | |
| persist-credentials: false | |
| - name: Build TSAN CPython ${{ matrix.python-version }} | |
| if: steps.cache-cpython-tsan-restore.outputs.cache-hit != 'true' | |
| run: | | |
| cd cpython | |
| mkdir ${GITHUB_WORKSPACE}/cpython-tsan | |
| CC=clang-18 CXX=clang++-18 ./configure --prefix ${GITHUB_WORKSPACE}/cpython-tsan --disable-gil --with-thread-sanitizer | |
| make -j64 | |
| make install -j64 | |
| # Check whether free-threading mode is enabled | |
| PYTHON_GIL=0 ${GITHUB_WORKSPACE}/cpython-tsan/bin/python3 -c "import sys; assert not sys._is_gil_enabled()" | |
| # Create archive to be used with bazel as hermetic python: | |
| cd ${GITHUB_WORKSPACE} && tar -czpf python-tsan.tgz cpython-tsan | |
| - name: Save TSAN CPython ${{ matrix.python-version }} | |
| id: cache-cpython-tsan-save | |
| if: steps.cache-cpython-tsan-restore.outputs.cache-hit != 'true' | |
| uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0 | |
| with: | |
| path: | | |
| ./python-tsan.tgz | |
| key: ${{ runner.os }}-cpython-tsan-${{ matrix.python-version }}-${{ steps.get-date.outputs.date }} | |
| - name: Restore cached TSAN Numpy | |
| id: cache-numpy-tsan-restore | |
| uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0 | |
| with: | |
| path: | | |
| ./wheelhouse | |
| key: ${{ runner.os }}-numpy-tsan-${{ matrix.python-version }}-${{ hashFiles('numpy/pyproject.toml') }}-${{ steps.get-date.outputs.date }} | |
| - name: Build TSAN Numpy wheel | |
| if: steps.cache-numpy-tsan-restore.outputs.cache-hit != 'true' | |
| run: | | |
| set -eux | |
| cd numpy | |
| # If we restored cpython from cache, we need to get python interpreter from python-tsan.tgz | |
| if [ ! -d ${GITHUB_WORKSPACE}/cpython-tsan/bin/ ]; then | |
| echo "Extract cpython from python-tsan.tgz" | |
| pushd . | |
| ls ${GITHUB_WORKSPACE}/python-tsan.tgz | |
| cd ${GITHUB_WORKSPACE} && tar -xzf python-tsan.tgz | |
| ls ${GITHUB_WORKSPACE}/cpython-tsan/bin/ | |
| popd | |
| fi | |
| export PATH=${GITHUB_WORKSPACE}/cpython-tsan/bin/:$PATH | |
| python3 -m pip install uv~=0.5.30 | |
| python3 -m uv pip install -r requirements/build_requirements.txt | |
| CC=clang-18 CXX=clang++-18 python3 -m pip wheel --wheel-dir dist -v . --no-build-isolation -Csetup-args=-Db_sanitize=thread -Csetup-args=-Dbuildtype=debugoptimized | |
| # Create simple index and copy the wheel | |
| mkdir -p ${GITHUB_WORKSPACE}/wheelhouse/numpy | |
| numpy_whl_name=($(cd dist && ls numpy*.whl)) | |
| if [ -z "${numpy_whl_name}" ]; then exit 1; fi | |
| echo "Built TSAN Numpy wheel: ${numpy_whl_name}" | |
| cp dist/${numpy_whl_name} ${GITHUB_WORKSPACE}/wheelhouse/numpy | |
| cat << EOF > ${GITHUB_WORKSPACE}/wheelhouse/index.html | |
| <!DOCTYPE html><html><body> | |
| <a href="numpy">numpy></a></br> | |
| </body></html> | |
| EOF | |
| cat << EOF > ${GITHUB_WORKSPACE}/wheelhouse/numpy/index.html | |
| <!DOCTYPE html><html><body> | |
| <a href="${numpy_whl_name}">${numpy_whl_name}</a></br> | |
| </body></html> | |
| EOF | |
| - name: Save TSAN Numpy wheel | |
| id: cache-numpy-tsan-save | |
| if: steps.cache-numpy-tsan-restore.outputs.cache-hit != 'true' | |
| uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0 | |
| with: | |
| path: | | |
| ./wheelhouse | |
| key: ${{ runner.os }}-numpy-tsan-${{ matrix.python-version }}-${{ hashFiles('numpy/pyproject.toml') }}-${{ steps.get-date.outputs.date }} | |
| - name: Build OpenArchX and run tests | |
| timeout-minutes: 180 | |
| env: | |
| JAX_NUM_GENERATED_CASES: 1 | |
| JAX_ENABLE_X64: true | |
| JAX_SKIP_SLOW_TESTS: true | |
| PY_COLORS: 1 | |
| DEBIAN_FRONTEND: noninteractive | |
| run: | | |
| set -x | |
| cd openarchx | |
| export PYTHON_SHA256=($(sha256sum ${GITHUB_WORKSPACE}/python-tsan.tgz)) | |
| echo "Python sha256: ${PYTHON_SHA256}" | |
| python3 build/build.py build --configure_only \ | |
| --python_version=${{ matrix.python-version }}-ft \ | |
| --bazel_options=--repo_env=HERMETIC_PYTHON_URL="file://${GITHUB_WORKSPACE}/python-tsan.tgz" \ | |
| --bazel_options=--repo_env=HERMETIC_PYTHON_SHA256=${PYTHON_SHA256} \ | |
| --bazel_options=--repo_env=HERMETIC_PYTHON_PREFIX="cpython-tsan/" \ | |
| --bazel_options=--color=yes \ | |
| --bazel_options=--copt=-fsanitize=thread \ | |
| --bazel_options=--linkopt="-fsanitize=thread" \ | |
| --bazel_options=--copt=-g \ | |
| --clang_path=/usr/bin/clang-18 | |
| mkdir -p dist | |
| # Check whether we have numpy wheel or exit with error | |
| ls ${GITHUB_WORKSPACE}/wheelhouse/numpy/*.whl || exit 1 | |
| cp -v ${GITHUB_WORKSPACE}/wheelhouse/numpy/*.whl dist/ | |
| if [ "${{ matrix.python-version }}" == "3.14" ]; then | |
| # Patch build/requirements_lock_3_14_ft.txt to use TSAN instrumented NumPy | |
| sed -i "s|--extra-index-url.*|--extra-index-url file://${GITHUB_WORKSPACE}/wheelhouse/|" build/${{ matrix.requirements_lock_name }}.txt | |
| fi | |
| echo "JAX_NUM_GENERATED_CASES=$JAX_NUM_GENERATED_CASES" | |
| echo "JAX_ENABLE_X64=$JAX_ENABLE_X64" | |
| echo "JAX_SKIP_SLOW_TESTS=$JAX_SKIP_SLOW_TESTS" | |
| # Set symlink to the bazel executable | |
| bazel_exec=($(ls bazel-*)) | |
| ln -s ${bazel_exec} bazel | |
| # Check python version | |
| python_version="${{ matrix.python-version }}" | |
| ./bazel run --@rules_python//python/config_settings:py_freethreaded="yes" @python_${python_version//./_}_host//:python -- -VV | |
| # Check numpy version | |
| ./bazel cquery @pypi_numpy//:* | grep whl | |
| # Build OPENARCHX and run tests | |
| ./bazel test \ | |
| --test_env=JAX_NUM_GENERATED_CASES=$JAX_NUM_GENERATED_CASES \ | |
| --test_env=JAX_ENABLE_X64=$JAX_ENABLE_X64 \ | |
| --test_env=JAX_SKIP_SLOW_TESTS=$JAX_SKIP_SLOW_TESTS \ | |
| --test_env=PYTHON_GIL=0 \ | |
| --test_env=TSAN_OPTIONS=halt_on_error=1,suppressions=$PWD/.github/workflows/tsan-suppressions_${{ matrix.python-version }}.txt \ | |
| --test_env=JAX_TEST_NUM_THREADS=8 \ | |
| --test_output=errors \ | |
| --local_test_jobs=10 \ | |
| --test_timeout=1800 \ | |
| --config=ci_rbe_cache \ | |
| //tests:cpu_tests |