Sai Enduri
01d16d4139
* Add tests on slurm cluster * Integrate slurm. * Add flags. * Added dynamic selection of runners for tests and cleanup for slurm reservation * Revert "Added dynamic selection of runners for tests and cleanup for slurm reservation" This reverts commit d5350ff6e4f563ddd56ad81e4bc2a393ed55ba00. * Refactor so tests run on both architectures. * continue on error * fail fast false on matrix * remove scancel * skip all single node tests * fix pattern matching for pytest * switch to always skip github job * Update to latest allocation. * Clean up workflows and update docker image. * Updated container image published from PR #1517 * Switch back to TheRock main branch sha. --------- Co-authored-by: arravikum <arravikum@amd.com>
86 lines
2.5 KiB
YAML
86 lines
2.5 KiB
YAML
name: TheRock CI for rccl
|
|
|
|
on:
|
|
push:
|
|
branches:
|
|
- develop
|
|
workflow_dispatch:
|
|
|
|
permissions:
|
|
contents: read
|
|
|
|
concurrency:
|
|
# A PR number if a pull request and otherwise the commit hash. This cancels
|
|
# queued and in-progress runs for the same PR (presubmit) or commit
|
|
# (postsubmit). The workflow name is prepended to avoid conflicts between
|
|
# different workflows.
|
|
group: ${{ github.workflow }}-${{ github.event.number || github.sha }}
|
|
cancel-in-progress: true
|
|
|
|
jobs:
|
|
setup:
|
|
runs-on: ubuntu-24.04
|
|
env:
|
|
# The commit being checked out is the merge commit for a PR. Its first
|
|
# parent will be the tip of the base branch.
|
|
BASE_REF: HEAD^
|
|
outputs:
|
|
enable_therock_ci: ${{ steps.configure.outputs.enable_therock_ci }}
|
|
steps:
|
|
- name: "Checking out repository"
|
|
uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
|
|
with:
|
|
# We need the parent commit to do a diff
|
|
fetch-depth: 2
|
|
|
|
- name: "Configuring CI options"
|
|
id: configure
|
|
run: python .github/scripts/therock_configure_ci.py
|
|
|
|
therock-ci-linux:
|
|
name: TheRock CI Linux
|
|
needs: setup
|
|
if: ${{ needs.setup.outputs.enable_therock_ci == 'true' }}
|
|
permissions:
|
|
contents: read
|
|
id-token: write
|
|
strategy:
|
|
fail-fast: false
|
|
matrix:
|
|
amdgpu_family: [gfx94X-dcgpu, gfx950-dcgpu]
|
|
uses: ./.github/workflows/therock-ci-linux.yml
|
|
secrets: inherit
|
|
with:
|
|
amdgpu_families: ${{ matrix.amdgpu_family }}
|
|
extra_cmake_options: >
|
|
-DTHEROCK_ENABLE_ALL=OFF
|
|
-DTHEROCK_BUILD_TESTING=ON
|
|
-DTHEROCK_BUNDLE_SYSDEPS=ON
|
|
-DTHEROCK_ENABLE_COMM_LIBS=ON
|
|
-DTHEROCK_ENABLE_ROCPROFV3=ON
|
|
-DTHEROCK_USE_EXTERNAL_RCCL=ON
|
|
-DTHEROCK_USE_EXTERNAL_RCCL_TESTS=ON
|
|
-DTHEROCK_RCCL_SOURCE_DIR=./rccl
|
|
-DTHEROCK_RCCL_TESTS_SOURCE_DIR=./rccl-tests
|
|
-DTHEROCK_ENABLE_MPI=ON
|
|
|
|
therock_ci_summary:
|
|
name: TheRock CI Summary
|
|
if: always()
|
|
needs:
|
|
- setup
|
|
- therock-ci-linux
|
|
runs-on: ubuntu-24.04
|
|
steps:
|
|
- name: Output failed jobs
|
|
run: |
|
|
echo '${{ toJson(needs) }}'
|
|
FAILED_JOBS="$(echo '${{ toJson(needs) }}' \
|
|
| jq --raw-output \
|
|
'map_values(select(.result!="success" and .result!="skipped")) | keys | join(",")' \
|
|
)"
|
|
if [[ "${FAILED_JOBS}" != "" ]]; then
|
|
echo "The following jobs failed: ${FAILED_JOBS}"
|
|
exit 1
|
|
fi
|