Skip to content

add AMD GPU oversubscription check #8894

add AMD GPU oversubscription check

add AMD GPU oversubscription check #8894

Workflow file for this run

name: CI
on:
pull_request:
push:
branches: main
schedule:
- cron: '5 0 * * *'
# Cancel in progress CI runs when a new run targeting the same PR or branch/tag is triggered.
# https://stackoverflow.com/questions/66335225/how-to-cancel-previous-runs-in-the-pr-when-you-push-new-commitsupdate-the-curre
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
jobs:
flake8:
name: Flake8
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
-
uses: actions/setup-python@v5
with:
# matches compat target in setup.py
python-version: '3.8'
- name: Flake8 test
run: |
python3 -m venv myenv
source myenv/bin/activate
python -m pip install wheel
python -m pip install flake8 pep8-naming flake8-quotes flake8-bugbear
python -m flake8 --show-source --statistics "$(basename $GITHUB_REPOSITORY)" test examples setup.py doc/conf.py bin/ && echo "Flake8 found no errors."
mypy:
name: Mypy
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
-
uses: actions/setup-python@v5
with:
python-version: '3.x'
- name: Install
run: |
. .ci-support/install.sh
- name: Run mypy
run: |
MINIFORGE_INSTALL_DIR=.miniforge3
. "$MINIFORGE_INSTALL_DIR/bin/activate" testing
python -m pip install types-psutil types-PyYAML
./run-mypy.sh
pylint:
name: Pylint
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
-
uses: actions/setup-python@v5
with:
python-version: '3.x'
- name: Install
run: |
. .ci-support/install.sh
- name: Run Pylint
run: |
MINIFORGE_INSTALL_DIR=.miniforge3
. "$MINIFORGE_INSTALL_DIR/bin/activate" testing
./run-pylint.sh
pydocstyle:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
-
uses: actions/setup-python@v5
with:
python-version: '3.x'
- name: Run Pydocstyle
run: |
python3 -m venv myenv
source myenv/bin/activate
python -m pip install wheel
python -m pip install pydocstyle
python -m pydocstyle "$(basename $GITHUB_REPOSITORY)" && echo "pydocstyle found no errors."
pytest:
name: Pytest
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install
run: |
. .ci-support/install.sh
- name: Run tests
run: |
MINIFORGE_INSTALL_DIR=.miniforge3
. "$MINIFORGE_INSTALL_DIR/bin/activate" testing
cd test
python -m pip install pytest
python -m pytest --cov=mirgecom --durations=0 --tb=native --junitxml=pytest.xml --doctest-modules -rxsw . ../doc/*.rst ../doc/*/*.rst
examples:
name: Examples
runs-on: ${{ matrix.os }}
strategy:
matrix:
os: [ubuntu-latest, porter]
steps:
- uses: actions/checkout@v4
- name: Install
run: |
. .ci-support/install.sh
- name: Run examples
run: |
set -x
MINIFORGE_INSTALL_DIR=.miniforge3
. "$MINIFORGE_INSTALL_DIR/bin/activate" testing
export XDG_CACHE_HOME=/tmp
mamba install vtk # needed for the accuracy comparison
# Temporary workaround for https://github.com/conda-forge/openvino-feedstock/pull/73
[[ $(hostname) == "porter" ]] && export PYOPENCL_TEST="port:nv" && unset XDG_CACHE_HOME
# This is only possible because actions run sequentially on porter
[[ $(hostname) == "porter" ]] && rm -rf /tmp/githubrunner/pocl-scratch && rm -rf /tmp/githubrunner/xdg-scratch
scripts/run-integrated-tests.sh --examples
doc:
name: Documentation
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install
run: |
. .ci-support/install.sh
- name: Build docs
run: |
set -x
MINIFORGE_INSTALL_DIR=.miniforge3
. "$MINIFORGE_INSTALL_DIR/bin/activate" testing
sudo apt-get update
sudo apt-get install texlive-latex-extra latexmk
mamba install sphinx graphviz 'docutils>=0.16'
# Work around "Not enough memory to run on this device." errors in CI:
mamba uninstall pocl
pip install sphinx-math-dollar sphinx-copybutton furo
cd doc
make html SPHINXOPTS="-W --keep-going -n"
make latexpdf SPHINXOPTS="-W --keep-going -n"
emirge:
name: Emirge installation
runs-on: ${{ matrix.os }}
strategy:
matrix:
os: [ubuntu-latest, macos-latest]
steps:
- uses: actions/checkout@v4
- name: Install emirge
run: |
[[ $(uname) == Linux ]] && sudo apt-get update && sudo apt-get install -y openmpi-bin libopenmpi-dev
[[ $(uname) == Darwin ]] && ( brew upgrade || true; brew install mpich )
cd ..
git clone https://github.com/illinois-ceesd/emirge
cd emirge
cp -a ../mirgecom .
./install.sh --skip-clone
- name: Run simple mirgecom test
run: |
cd ..
source emirge/config/activate_env.sh
cd mirgecom/examples
python -m mpi4py ./pulse.py
production:
name: Production testing
runs-on: ${{ matrix.os }}
strategy:
matrix:
os: [ubuntu-latest, macos-latest, porter]
steps:
- uses: actions/checkout@v4
with:
fetch-depth: '0'
- name: Prepare production environment
run: |
set -x
[[ $(uname) == Linux ]] && [[ $(hostname) != "porter" ]] && sudo apt-get update && sudo apt-get install -y openmpi-bin libopenmpi-dev
[[ $(uname) == Darwin ]] && ( brew upgrade || true; brew install mpich )
# This is only possible because actions run sequentially on porter
[[ $(hostname) == "porter" ]] && rm -rf /tmp/githubrunner/pocl-scratch && rm -rf /tmp/githubrunner/xdg-scratch
MIRGEDIR=$(pwd)
cat scripts/production-testing-env.sh
. scripts/production-testing-env.sh
cd ..
date
printf "Removing stale install ..."
rm -rf emirge.prod emirge.y1
printf "done.\n"
date
git clone https://github.com/illinois-ceesd/emirge emirge.prod
cd emirge.prod
. ../mirgecom/scripts/install-mirge-from-source.sh ${MIRGEDIR}/..
- name: Run production test
run: |
source ../config/activate_env.sh
# Temporary workaround for https://github.com/conda-forge/openvino-feedstock/pull/73
scripts/run-integrated-tests.sh --production