perf.yml

name: Performance Monitoring
on:
  push:
    branches:
    - master
  workflow_dispatch: {}
  repository_dispatch:
    types: [benchmark-command]

env:
  TI_CI: "1"
  TI_SKIP_VERSION_CHECK: "ON"
  CI_IMAGE_VERSION: '202305261112'

jobs:
  gpu_backends:
    name: Performance Monitoring
    timeout-minutes: 60
    # Disable this workflow on forks
    if: github.repository_owner == 'taichi-dev'
    runs-on: [self-hosted, x64, cuda, linux, benchmark]
    steps:
      - name: Workaround checkout Needed single revision issue
        run: git submodule foreach 'git rev-parse HEAD > /dev/null 2>&1 || rm -rf $PWD' || true

      - uses: actions/checkout@v2
        with:
          submodules: "recursive"
          fetch-depth: '0'

      - name: Prepare Environment
        run: |
          . .github/workflows/scripts/common-utils.sh
          prepare-build-cache
          echo CI_DOCKER_RUN_EXTRA_ARGS="-v $(pwd):/home/dev/taichi" >> $GITHUB_ENV

      - name: Build Taichi Wheel
        run: |
          . .github/workflows/scripts/common-utils.sh

          ci-docker-run-gpu --name taichi-benchmark-build \
            registry.botmaster.tgr/taichi-build-android:${{ env.CI_IMAGE_VERSION }} \
            /home/dev/taichi/.github/workflows/scripts/build.py

        env:
          PY: '3.8'
          PROJECT_NAME: taichi
          TAICHI_CMAKE_ARGS: >-
            -DTI_WITH_OPENGL:BOOL=ON
            -DTI_WITH_CUDA:BOOL=ON
            -DTI_WITH_VULKAN:BOOL=ON
            -DTI_BUILD_TESTS:BOOL=OFF
            -DTI_WITH_C_API=OFF

      - name: Run taichi-benchmark
        id: run-benchmark
        run: |
          . .github/workflows/scripts/common-utils.sh

          MEMFREQ=($(sudo nvidia-smi --query-supported-clocks=mem --format=csv | head -n 2 | tail -n 1))
          GRFREQ=($(sudo nvidia-smi --query-supported-clocks=gr --format=csv | head -n 35 | tail -n 1))

          function restore-freq-pinning {
            sudo nvidia-smi -rmc
            sudo nvidia-smi -rgc
            sudo nvidia-smi -pm 0
          }

          trap restore-freq-pinning EXIT

          sudo nvidia-smi -pm 1
          sudo nvidia-smi -lmc ${MEMFREQ[0]}
          sudo nvidia-smi -lgc ${GRFREQ[0]}
          sleep 0.5

          ci-docker-run-gpu --name taichi-benchmark-run \
            -e GITHUB_TOKEN \
            -e GITHUB_EVENT_ACTION \
            -e BENCHMARK_UPLOAD_TOKEN \
            -v $GITHUB_EVENT_PATH:/github-event.json \
            registry.botmaster.tgr/taichi-test-android:${{ env.CI_IMAGE_VERSION }} \
            /home/dev/taichi/.github/workflows/scripts/unix-perf-mon.sh

        env:
          PY: '3.8'
          BENCHMARK_UPLOAD_TOKEN: ${{ secrets.BENCHMARK_UPLOAD_TOKEN }}
          GITHUB_EVENT_ACTION: ${{ github.event.action }}
          GITHUB_TOKEN: ${{ secrets.GARDENER_PAT }}