conda-cpp-tests.yaml

on:
  workflow_call:
    inputs:
      build_type:
        required: true
        type: string
      branch:
        type: string
      date:
        type: string
      sha:
        type: string
      repo:
        type: string
      script:
        type: string
        default: "ci/test_cpp.sh"
      matrix_filter:
        type: string
        default: "."
      container-options:
        required: false
        type: string
        default: "-e _NOOP"

defaults:
  run:
    shell: bash

permissions:
  actions: read
  checks: none
  contents: read
  deployments: none
  discussions: none
  id-token: write
  issues: none
  packages: read
  pages: none
  pull-requests: read
  repository-projects: none
  security-events: none
  statuses: none

jobs:
  compute-matrix:
    runs-on: ubuntu-latest
    env:
      BUILD_TYPE: ${{ inputs.build_type }}
    outputs:
      MATRIX: ${{ steps.compute-matrix.outputs.MATRIX }}
    steps:
      - name: Validate Test Type
        run: |
          if [[ "$BUILD_TYPE" != "pull-request" ]] && [[ "$BUILD_TYPE" != "nightly" ]] && [[ "$BUILD_TYPE" != "branch" ]]; then
              echo "Invalid build type! Must be one of 'nightly', 'pull-request', or 'branch'."
              exit 1
          fi
      - name: Compute C++ Test Matrix
        id: compute-matrix
        run: |
          set -eo pipefail

          # please keep the matrices sorted in ascending order by the following:
          #
          #     [ARCH, PY_VER, CUDA_VER, LINUX_VER, GPU, DRIVER, DEPENDENCIES]
          #
          export MATRICES="
            pull-request:
              # amd64
              - { ARCH: 'amd64', PY_VER: '3.10', CUDA_VER: '11.8.0', LINUX_VER: 'rockylinux8', GPU: 'l4',   DRIVER: 'earliest', DEPENDENCIES: 'oldest' }
              - { ARCH: 'amd64', PY_VER: '3.12', CUDA_VER: '12.8.0', LINUX_VER: 'ubuntu24.04', GPU: 'h100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
              # arm64
              # - { ARCH: 'arm64', PY_VER: '3.11', CUDA_VER: '12.0.1', LINUX_VER: 'ubuntu20.04', GPU: 'a100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
            nightly:
              # amd64
              - { ARCH: 'amd64', PY_VER: '3.10', CUDA_VER: '11.4.3', LINUX_VER: 'rockylinux8', GPU: 'l4',   DRIVER: 'earliest', DEPENDENCIES: 'oldest' }
              - { ARCH: 'amd64', PY_VER: '3.10', CUDA_VER: '11.8.0', LINUX_VER: 'ubuntu20.04', GPU: 'h100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
              - { ARCH: 'amd64', PY_VER: '3.11', CUDA_VER: '12.0.1', LINUX_VER: 'rockylinux8', GPU: 'l4',   DRIVER: 'latest',   DEPENDENCIES: 'latest' }
              - { ARCH: 'amd64', PY_VER: '3.12', CUDA_VER: '12.8.0', LINUX_VER: 'ubuntu24.04', GPU: 'h100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
              # arm64
              # - { ARCH: 'arm64', PY_VER: '3.10', CUDA_VER: '12.2.2', LINUX_VER: 'ubuntu22.04', GPU: 'a100', DRIVER: 'latest',   DEPENDENCIES: 'oldest' }
              # - { ARCH: 'arm64', PY_VER: '3.11', CUDA_VER: '11.8.0', LINUX_VER: 'ubuntu20.04', GPU: 'a100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
              - { ARCH: 'arm64', PY_VER: '3.12', CUDA_VER: '12.8.0', LINUX_VER: 'rockylinux8', GPU: 'a100', DRIVER: 'latest',   DEPENDENCIES: 'latest' }
          "

          # Use the nightly matrix for branch tests
          MATRIX_TYPE="${BUILD_TYPE}"
          if [[ "${MATRIX_TYPE}" == "branch" ]]; then
            MATRIX_TYPE="nightly"
          fi
          export MATRIX_TYPE
          TEST_MATRIX=$(yq -n 'env(MATRICES) | .[strenv(MATRIX_TYPE)]')
          export TEST_MATRIX

          MATRIX="$(
            yq -n -o json 'env(TEST_MATRIX)' | \
            jq -c '${{ inputs.matrix_filter }} | if (. | length) > 0 then {include: .} else "Error: Empty matrix\n" | halt_error(1) end'
          )"

          echo "MATRIX=${MATRIX}" | tee --append "${GITHUB_OUTPUT}"
  tests:
    name: ${{ matrix.CUDA_VER }}, ${{ matrix.PY_VER }}, ${{ matrix.ARCH }}, ${{ matrix.LINUX_VER }}, ${{ matrix.GPU }}, ${{ matrix.DRIVER }}-driver, ${{ matrix.DEPENDENCIES }}-deps
    needs: compute-matrix
    strategy:
      fail-fast: false
      matrix: ${{ fromJSON(needs.compute-matrix.outputs.MATRIX) }}
    runs-on: "linux-${{ matrix.ARCH }}-gpu-${{ matrix.GPU }}-${{ matrix.DRIVER }}-1"
    env:
      RAPIDS_ARTIFACTS_DIR: ${{ github.workspace }}/artifacts
      RAPIDS_DEPENDENCIES: ${{ matrix.DEPENDENCIES }}
      RAPIDS_TESTS_DIR: ${{ github.workspace }}/test-results
    container:
      image: rapidsai/ci-conda:cuda${{ matrix.CUDA_VER }}-${{ matrix.LINUX_VER }}-py${{ matrix.PY_VER }}
      options: ${{ inputs.container-options }}
      env:
        RAPIDS_BUILD_TYPE: ${{ inputs.build_type }}
        NVIDIA_VISIBLE_DEVICES: ${{ env.NVIDIA_VISIBLE_DEVICES }}
    steps:
      - uses: aws-actions/configure-aws-credentials@v4
        with:
          role-to-assume: ${{ vars.AWS_ROLE_ARN }}
          aws-region: ${{ vars.AWS_REGION }}
          role-duration-seconds: 43200 # 12h

      - uses: actions/checkout@v4
        with:
          repository: ${{ inputs.repo }}
          ref: ${{ inputs.sha }}
          fetch-depth: 0
      # This has to be AFTER the checkout step. It creates a telemetry-artifacts directory,
      # and the checkout step would destroy it.
      - name: Telemetry setup
        uses: rapidsai/shared-actions/telemetry-dispatch-setup@main
        continue-on-error: true
        if: ${{ vars.TELEMETRY_ENABLED == 'true' && github.run_attempt == '1'}}
        with:
          extra_attributes: "rapids.PACKAGER=conda,rapids.CUDA_VER=${{ matrix.CUDA_VER }},rapids.PY_VER=${{ matrix.PY_VER }},rapids.ARCH=${{ matrix.ARCH }},rapids.LINUX_VER=${{ matrix.LINUX_VER }},rapids.GPU=${{ matrix.GPU }},rapids.DRIVER=${{ matrix.DRIVER }},rapids.DEPENDENCIES=${{ matrix.DEPENDENCIES }}"
      - name: Standardize repository information
        run: |
          echo "RAPIDS_REPOSITORY=${{ inputs.repo || github.repository }}" >> "${GITHUB_ENV}"
          echo "RAPIDS_SHA=$(git rev-parse HEAD)" >> "${GITHUB_ENV}"
          echo "RAPIDS_REF_NAME=${{ inputs.branch || github.ref_name }}" >> "${GITHUB_ENV}"
          echo "RAPIDS_NIGHTLY_DATE=${{ inputs.date }}" >> "${GITHUB_ENV}"

      - name: Setup proxy cache
        uses: nv-gha-runners/setup-proxy-cache@main
        continue-on-error: true
        # Skip the cache on RDS Lab nodes
        if: ${{ matrix.GPU != 'v100' && matrix.GPU != 'a100' }}

      - name: C++ tests
        run: ${{ inputs.script }}
        env:
          GH_TOKEN: ${{ github.token }}
      - name: Generate test report
        uses: test-summary/action@v2.4
        with:
          paths: "${{ env.RAPIDS_TESTS_DIR }}/*.xml"
        if: always()
      - name: Upload additional artifacts
        if: "!cancelled()"
        run: rapids-upload-artifacts-dir cuda${RAPIDS_CUDA_VERSION%%.*}_$(arch)
      - name: Telemetry upload attributes
        uses: rapidsai/shared-actions/telemetry-dispatch-stash-job-artifacts@main
        continue-on-error: true
        if: ${{ vars.TELEMETRY_ENABLED == 'true' && github.run_attempt == '1' }}