test.yml

name: 'Test Suite'

on:
  push:
    branches: [master]
  pull_request:
    types: [opened, synchronize, reopened, ready_for_review]
  workflow_dispatch:
  schedule:
    - cron: '0 6 * * 1'  # Weekly Monday 6 AM UTC: refresh coverage cache before 7-day expiry

concurrency:
  # PRs: group by branch (new push cancels old). Push to master: unique per SHA (never cancelled).
  group: ${{ github.workflow }}-${{ github.event_name == 'push' && github.sha || github.ref }}
  cancel-in-progress: ${{ github.event_name != 'push' }}

jobs:
  lint-gate:
    name: Lint Gate
    runs-on: ubuntu-latest
    steps:
      - name: Clone
        uses: actions/checkout@v4

      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.12'

      - name: Initialize MFC
        run: ./mfc.sh init

      - name: Check Formatting
        run: |
          ./mfc.sh format -j "$(nproc)"
          git diff --exit-code || (echo "::error::Code is not formatted. Run './mfc.sh format' locally." && exit 1)

      - name: Spell Check
        run: ./mfc.sh spelling

      - name: Lint Toolchain
        run: ./mfc.sh lint

      - name: Lint Source
        run: python3 toolchain/mfc/lint_source.py

      - name: Lint Docs
        run: python3 toolchain/mfc/lint_docs.py

      - name: Lint Parameter Docs
        run: python3 toolchain/mfc/lint_param_docs.py

  file-changes:
    name: Detect File Changes
    runs-on: 'ubuntu-latest'
    outputs:
      checkall: ${{ steps.changes.outputs.checkall }}
      cases_py: ${{ steps.changes.outputs.cases_py }}
      dep_changed: ${{ steps.dep-check.outputs.dep_changed }}
    steps:
      - name: Clone
        uses: actions/checkout@v4

      - name: Detect Changes
        uses: dorny/paths-filter@v3
        id: changes
        with:
          filters: ".github/file-filter.yml"

      - name: Check for Fortran dependency changes
        id: dep-check
        env:
          GH_TOKEN: ${{ github.token }}
        run: |
          # Detect added/removed use/include statements that change the
          # Fortran dependency graph, which would make the coverage cache stale.
          PR_NUMBER="${{ github.event.pull_request.number }}"
          BEFORE="${{ github.event.before }}"
          AFTER="${{ github.event.after }}"
          if [ "${{ github.event_name }}" = "pull_request" ]; then
            # Default to dep_changed=true if gh pr diff fails (safe fallback).
            DIFF=$(gh pr diff "$PR_NUMBER" 2>/dev/null) || {
              echo "gh pr diff failed — defaulting to dep_changed=true for safety."
              echo "dep_changed=true" >> "$GITHUB_OUTPUT"
              exit 0
            }
          elif [ "${{ github.event_name }}" = "push" ]; then
            DIFF=$(git diff "$BEFORE".."$AFTER" 2>/dev/null) || {
              echo "git diff failed for push event — defaulting to dep_changed=true for safety."
              echo "dep_changed=true" >> "$GITHUB_OUTPUT"
              exit 0
            }
          else
            DIFF=""
          fi
          if echo "$DIFF" | \
             grep -qE '^[+-][[:space:]]*(use[[:space:],]+[a-zA-Z_]|#:include[[:space:]]|include[[:space:]]+['"'"'"])'; then
            echo "dep_changed=true" >> "$GITHUB_OUTPUT"
            echo "Fortran dependency change detected — will rebuild coverage cache."
          else
            echo "dep_changed=false" >> "$GITHUB_OUTPUT"
          fi

  rebuild-cache:
    name: Rebuild Coverage Cache
    needs: [lint-gate, file-changes]
    if: >-
      github.repository == 'MFlowCode/MFC' &&
      (
        (github.event_name == 'pull_request' &&
         (needs.file-changes.outputs.cases_py == 'true' ||
          needs.file-changes.outputs.dep_changed == 'true')) ||
        (github.event_name == 'push' &&
         (needs.file-changes.outputs.cases_py == 'true' ||
          needs.file-changes.outputs.dep_changed == 'true')) ||
        github.event_name == 'workflow_dispatch' ||
        github.event_name == 'schedule'
      )
    timeout-minutes: 240
    runs-on:
      group:  phoenix
      labels: gt
    steps:
      - name: Clone
        uses: actions/checkout@v4
        with:
          ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
          clean: false

      - name: Rebuild Cache via SLURM
        run: bash .github/scripts/submit-slurm-job.sh .github/workflows/common/rebuild-cache.sh cpu none phoenix

      - name: Print Logs
        if:   always()
        run:  cat rebuild-cache-cpu-none.out

      - name: Upload Cache Artifact
        uses: actions/upload-artifact@v4
        with:
          name: coverage-cache
          path: toolchain/mfc/test/test_coverage_cache.json.gz
          retention-days: 1

      - name: Save Coverage Cache
        uses: actions/cache/save@v4
        with:
          path: toolchain/mfc/test/test_coverage_cache.json.gz
          key: coverage-cache-${{ github.event.pull_request.number || 'master' }}-${{ hashFiles('toolchain/mfc/test/cases.py') }}-${{ github.sha }}
        continue-on-error: true

  github:
    name: Github
    needs: [lint-gate, file-changes, rebuild-cache]
    if: >-
      !cancelled() &&
      needs.lint-gate.result == 'success' &&
      needs.file-changes.result == 'success' &&
      (needs.rebuild-cache.result == 'success' || needs.rebuild-cache.result == 'skipped') &&
      needs.file-changes.outputs.checkall == 'true'
    strategy:
      matrix:
        os:    ['ubuntu', 'macos']
        mpi:   ['mpi']
        precision: ['']
        debug: ['debug', 'no-debug']
        intel: [true, false]
        exclude:
          - os:    macos
            intel: true

        include:
          - os:    ubuntu
            mpi:   no-mpi
            precision: single
            debug: no-debug
            intel: false

      fail-fast: false
    continue-on-error: true
    runs-on: ${{ matrix.os }}-latest

    steps:
      - name: Clone
        uses: actions/checkout@v4

      - name: Fetch master for coverage diff
        run: |
          git fetch origin master:master --depth=1
          git fetch --deepen=200
        continue-on-error: true

      - name: Download Coverage Cache (from rebuild in this run)
        id: cache-artifact
        uses: actions/download-artifact@v4
        with:
          name: coverage-cache
          path: toolchain/mfc/test
        continue-on-error: true

      - name: Restore Coverage Cache (from previous run)
        if: steps.cache-artifact.outcome != 'success'
        id: cache-restore
        uses: actions/cache/restore@v4
        with:
          path: toolchain/mfc/test/test_coverage_cache.json.gz
          key: coverage-cache-${{ github.event.pull_request.number || 'master' }}-${{ hashFiles('toolchain/mfc/test/cases.py') }}-${{ github.sha }}
          restore-keys: |
            coverage-cache-${{ github.event.pull_request.number || 'master' }}-
            coverage-cache-master-
        continue-on-error: true

      - name: Coverage Cache Status
        run: |
          if [ "${{ steps.cache-artifact.outcome }}" = "success" ]; then
            echo "Coverage cache: loaded from rebuild artifact (this run)"
          elif [ "${{ steps.cache-restore.outputs.cache-hit }}" = "true" ]; then
            echo "Coverage cache: restored from actions/cache (previous run)"
          elif [ -f toolchain/mfc/test/test_coverage_cache.json.gz ]; then
            echo "Coverage cache: using committed fallback in repo"
          else
            echo "Coverage cache: none available — full test suite will run"
          fi

      - name: Setup MacOS
        if:   matrix.os == 'macos'
        run:  |
          brew update
          brew upgrade
          brew install coreutils python fftw hdf5 gcc@15 boost open-mpi lapack
          echo "FC=gfortran-15" >> $GITHUB_ENV
          echo "BOOST_INCLUDE=/opt/homebrew/include/" >> $GITHUB_ENV

      - name: Setup Ubuntu
        if:   matrix.os == 'ubuntu' && matrix.intel == false
        run: |
           sudo apt update -y
           sudo apt install -y cmake gcc g++ python3 python3-dev hdf5-tools \
                    libfftw3-dev libhdf5-dev openmpi-bin libopenmpi-dev \
                    libblas-dev liblapack-dev

      - name: Setup Ubuntu (Intel)
        if:   matrix.os == 'ubuntu' && matrix.intel == true
        run: |
          wget https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB
          sudo apt-key add GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB
          sudo add-apt-repository "deb https://apt.repos.intel.com/oneapi all main"
          sudo apt-get update
          sudo apt-get install -y intel-oneapi-compiler-fortran intel-oneapi-mpi intel-oneapi-mpi-devel
          # Export only new/changed env vars from setvars.sh.
          # `printenv >> $GITHUB_ENV` dumps all vars including shell internals
          # with special characters that corrupt GITHUB_ENV parsing.
          printenv | sort > /tmp/env_before
          source /opt/intel/oneapi/setvars.sh
          printenv | sort > /tmp/env_after
          diff /tmp/env_before /tmp/env_after | grep '^>' | sed 's/^> //' >> $GITHUB_ENV

      - name: Build
        run:  |
          /bin/bash mfc.sh test -v --dry-run -j $(nproc) --${{ matrix.debug }} --${{ matrix.mpi }} $PRECISION $TEST_ALL
        env:
          TEST_ALL: ${{ matrix.mpi == 'mpi' && '--test-all' || '' }}
          PRECISION: ${{ matrix.precision != '' && format('--{0}', matrix.precision) || '' }}

      - name: Test
        run:  |
          /bin/bash mfc.sh test -v --max-attempts 3 -j $(nproc) $ONLY_CHANGES $TEST_ALL $TEST_PCT
        env:
          TEST_ALL: ${{ matrix.mpi == 'mpi' && '--test-all' || '' }}
          TEST_PCT: ${{ matrix.debug == 'debug' && '-% 20' || '' }}
          ONLY_CHANGES: ${{ github.event_name == 'pull_request' && '--only-changes' || '' }}

  self:
    name: "${{ matrix.cluster_name }} (${{ matrix.device }}${{ matrix.interface != 'none' && format('-{0}', matrix.interface) || '' }}${{ matrix.shard != '' && format(' [{0}]', matrix.shard) || '' }})"
    needs: [lint-gate, file-changes, rebuild-cache]
    if: >-
      !cancelled() &&
      needs.lint-gate.result == 'success' &&
      needs.file-changes.result == 'success' &&
      (needs.rebuild-cache.result == 'success' || needs.rebuild-cache.result == 'skipped') &&
      github.repository == 'MFlowCode/MFC' &&
      needs.file-changes.outputs.checkall == 'true' &&
      github.event.pull_request.draft != true
    # Frontier CCE compiler is periodically broken by toolchain updates (e.g.
    # cpe/25.03 introduced an IPA SIGSEGV in CCE 19.0.0). Allow Frontier to
    # fail without blocking PR merges; Phoenix remains a hard gate.
    continue-on-error: ${{ matrix.runner == 'frontier' }}
    timeout-minutes: 480
    strategy:
      matrix:
        include:
          # Phoenix (GT) — build+test combined in SLURM job
          - runner:       'gt'
            cluster:      'phoenix'
            cluster_name: 'Georgia Tech | Phoenix'
            device: 'gpu'
            interface: 'acc'
          - runner:       'gt'
            cluster:      'phoenix'
            cluster_name: 'Georgia Tech | Phoenix'
            device: 'gpu'
            interface: 'omp'
          - runner:       'gt'
            cluster:      'phoenix'
            cluster_name: 'Georgia Tech | Phoenix'
            device: 'cpu'
            interface: 'none'
          # Frontier (ORNL) — CCE
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'acc'
            shard: '1/2'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'acc'
            shard: '2/2'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'omp'
            shard: '1/2'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'omp'
            shard: '2/2'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'cpu'
            interface: 'none'
          # Frontier AMD — build on login node, GPU tests sharded for batch partition
          - runner:       'frontier'
            cluster:      'frontier_amd'
            cluster_name: 'Oak Ridge | Frontier (AMD)'
            device: 'gpu'
            interface: 'omp'
            shard: '1/2'
          - runner:       'frontier'
            cluster:      'frontier_amd'
            cluster_name: 'Oak Ridge | Frontier (AMD)'
            device: 'gpu'
            interface: 'omp'
            shard: '2/2'
          - runner:       'frontier'
            cluster:      'frontier_amd'
            cluster_name: 'Oak Ridge | Frontier (AMD)'
            device: 'cpu'
            interface: 'none'
    runs-on:
      group:  phoenix
      labels: ${{ matrix.runner }}
    env:
      NODE_OPTIONS: ${{ matrix.cluster == 'phoenix' && '--max-old-space-size=2048' || '' }}
    steps:
      - name: Clone
        uses: actions/checkout@v4
        with:
          # clean: false preserves .slurm_job_id files across reruns so
          # submit-slurm-job.sh can detect and cancel stale SLURM jobs on retry.
          clean: false

      - name: Clean stale output files
        run:  rm -f *.out

      - name: Download Coverage Cache (from rebuild in this run)
        id: cache-artifact
        uses: actions/download-artifact@v4
        with:
          name: coverage-cache
          path: toolchain/mfc/test
        continue-on-error: true

      - name: Restore Coverage Cache (from previous run)
        if: steps.cache-artifact.outcome != 'success'
        id: cache-restore
        uses: actions/cache/restore@v4
        with:
          path: toolchain/mfc/test/test_coverage_cache.json.gz
          key: coverage-cache-${{ github.event.pull_request.number || 'master' }}-${{ hashFiles('toolchain/mfc/test/cases.py') }}-${{ github.sha }}
          restore-keys: |
            coverage-cache-${{ github.event.pull_request.number || 'master' }}-
            coverage-cache-master-
        continue-on-error: true

      - name: Coverage Cache Status
        run: |
          if [ "${{ steps.cache-artifact.outcome }}" = "success" ]; then
            echo "Coverage cache: loaded from rebuild artifact (this run)"
          elif [ "${{ steps.cache-restore.outputs.cache-hit }}" = "true" ]; then
            echo "Coverage cache: restored from actions/cache (previous run)"
          elif [ -f toolchain/mfc/test/test_coverage_cache.json.gz ]; then
            echo "Coverage cache: using committed fallback in repo"
          else
            echo "Coverage cache: none available — full test suite will run"
          fi

      - name: Build (login node)
        if:   matrix.cluster != 'phoenix'
        timeout-minutes: 60
        run:  bash .github/workflows/${{ matrix.cluster }}/build.sh ${{ matrix.device }} ${{ matrix.interface }}

      - name: Test
        run:  bash .github/scripts/submit-slurm-job.sh .github/workflows/common/test.sh ${{ matrix.device }} ${{ matrix.interface }} ${{ matrix.cluster }} ${{ matrix.shard }}

      - name: Cancel SLURM Jobs
        if: cancelled()
        run: |
          find . -name "*.slurm_job_id" | while read -r f; do
            job_id=$(cat "$f")
            echo "Cancelling SLURM job $job_id"
            scancel "$job_id" 2>/dev/null || true
          done

      - name: Compute Log Slug
        if:   always()
        id:   log
        run:  |
          SLUG="test-${{ matrix.device }}-${{ matrix.interface }}"
          SHARD="${{ matrix.shard }}"
          if [ -n "$SHARD" ]; then
            SLUG="${SLUG}-$(echo "$SHARD" | sed 's|/|-of-|')"
          fi
          echo "slug=${SLUG}" >> "$GITHUB_OUTPUT"

      - name: Print Logs
        if:   always()
        run:  cat ${{ steps.log.outputs.slug }}.out

      - name: Archive Logs
        uses: actions/upload-artifact@v4
        if:   matrix.cluster != 'phoenix'
        with:
          name: logs-${{ strategy.job-index }}-${{ steps.log.outputs.slug }}
          path: ${{ steps.log.outputs.slug }}.out

  case-optimization:
    name: "Case Opt | ${{ matrix.cluster_name }} (${{ matrix.device }}-${{ matrix.interface }})"
    if: github.repository == 'MFlowCode/MFC' && needs.file-changes.outputs.checkall == 'true' && github.event.pull_request.draft != true
    needs: [lint-gate, file-changes]
    # Frontier is non-blocking for the same reason as the self job above.
    continue-on-error: ${{ matrix.runner == 'frontier' }}
    timeout-minutes: 480
    strategy:
      matrix:
        include:
          - runner:       'gt'
            cluster:      'phoenix'
            cluster_name: 'Georgia Tech | Phoenix'
            device: 'gpu'
            interface: 'acc'
          - runner:       'gt'
            cluster:      'phoenix'
            cluster_name: 'Georgia Tech | Phoenix'
            device: 'gpu'
            interface: 'omp'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'acc'
          - runner:       'frontier'
            cluster:      'frontier'
            cluster_name: 'Oak Ridge | Frontier'
            device: 'gpu'
            interface: 'omp'
          - runner:       'frontier'
            cluster:      'frontier_amd'
            cluster_name: 'Oak Ridge | Frontier (AMD)'
            device: 'gpu'
            interface: 'omp'
    runs-on:
      group:  phoenix
      labels: ${{ matrix.runner }}
    steps:
      - name: Clone
        uses: actions/checkout@v4
        with:
          clean: false

      - name: Clean stale output files
        run:  rm -f *.out

      - name: Pre-Build (SLURM)
        if:   matrix.cluster == 'phoenix'
        run:  bash .github/scripts/submit-slurm-job.sh .github/scripts/prebuild-case-optimization.sh cpu ${{ matrix.interface }} ${{ matrix.cluster }}

      - name: Pre-Build (login node)
        if:   matrix.cluster != 'phoenix'
        run:  bash .github/scripts/prebuild-case-optimization.sh ${{ matrix.cluster }} ${{ matrix.device }} ${{ matrix.interface }}

      - name: Run Case-Optimization Tests
        run:  bash .github/scripts/submit-slurm-job.sh .github/scripts/run_case_optimization.sh ${{ matrix.device }} ${{ matrix.interface }} ${{ matrix.cluster }}

      - name: Cancel SLURM Jobs
        if: cancelled()
        run: |
          find . -name "*.slurm_job_id" | while read -r f; do
            job_id=$(cat "$f")
            echo "Cancelling SLURM job $job_id"
            scancel "$job_id" 2>/dev/null || true
          done

      - name: Print Logs
        if:   always()
        run:  |
          for f in prebuild-case-optimization-${{ matrix.device }}-${{ matrix.interface }}.out \
                   run-case-optimization-${{ matrix.device }}-${{ matrix.interface }}.out; do
            [ -f "$f" ] && echo "=== $f ===" && cat "$f"
          done

      - name: Archive Logs
        uses: actions/upload-artifact@v4
        if:   always()
        with:
          name: case-opt-${{ strategy.job-index }}-${{ matrix.cluster }}-${{ matrix.interface }}
          path: |
            prebuild-case-optimization-${{ matrix.device }}-${{ matrix.interface }}.out
            run-case-optimization-${{ matrix.device }}-${{ matrix.interface }}.out