TheRock/.github/workflows/test_rocm_wheels.yml

# Copyright Advanced Micro Devices, Inc.
# SPDX-License-Identifier: MIT

name: Test ROCm Wheels

on:
  workflow_dispatch:
    inputs:
      amdgpu_family:
        description: GPU family to test (e.g., gfx94X-dcgpu, gfx110X-all)
        required: true
        type: string
        default: "gfx94X-dcgpu"
      test_runs_on:
        description: Runner label to use. The selected runner should have a GPU supported by amdgpu_family
        required: true
        type: string
        default: "linux-gfx942-1gpu-ccs-csp-ossci-rocm"
      package_index_url:
        description: >-
          Full Python package index URL for use with --index-url.
        type: string
        default: "https://rocm.nightlies.amd.com/whl-multi-arch/"
      package_find_links_url:
        description: Full Python package index URL for use with --find-links (usually instead of package_index_url)
        type: string
      python_version:
        required: true
        type: string
        default: "3.12"
      rocm_version:
        description: ROCm version to pip install (e.g. "7.10.0a20251124")
        required: true
        type: string
      kpack_split:
        description: "'true' if this is a kpack-split flat build."
        type: string
        default: "false"
      container_image_name:
        description: >-
          Short, human-readable label for the container image (e.g. 'ubuntu24.04',
          'ubi10', 'native'). Use 'native' (with an empty container_image_url) for
          native runners like Windows.
        type: string
        default: "ubuntu24.04"
      container_image_url:
        description: >-
          Container image reference (e.g. 'ghcr.io/rocm/no_rocm_image_ubuntu24_04@sha256:...').
          Only applied when test_runs_on targets Linux. Leave empty to run directly
          on the runner (required for Windows and native Linux runners).
        type: string
        default: "ghcr.io/rocm/no_rocm_image_ubuntu24_04@sha256:405945a40deaff9db90b9839c0f41d4cba4a383c1a7459b28627047bf6302a26"

  workflow_call:
    inputs:
      amdgpu_family:
        required: true
        type: string
      test_runs_on:
        required: true
        type: string
      package_index_url:
        description: >-
          Full Python package index URL for use with --index-url.
        type: string
      package_find_links_url:
        type: string
      python_version:
        required: true
        type: string
      rocm_version:
        required: true
        type: string
      container_image_name:
        description: >-
          Short, human-readable label for the container image (e.g. 'ubuntu24.04',
          'ubi10', 'native'). Callers must pass this explicitly so this workflow
          stays image-agnostic. Use 'native' (with an empty container_image_url)
          for native runners like Windows.
        required: true
        type: string
      container_image_url:
        description: >-
          Container image reference. Only applied when test_runs_on targets Linux.
          Pass an empty string to run directly on the runner (required for Windows
          and native Linux runners).
        required: true
        type: string
      repository:
        description: "Repository to checkout. Defaults to github.repository."
        type: string
        default: ""
      ref:
        description: "Branch, tag, or SHA to checkout. Defaults to the triggering ref."
        type: string
        default: ""

permissions:
  contents: read

run-name: Test ROCm Wheels (${{ inputs.amdgpu_family }}, py${{ inputs.python_version }}, ${{ inputs.rocm_version }}, ${{ inputs.test_runs_on }})

jobs:
  test_wheels:
    name: Test ROCm wheels | ${{ inputs.amdgpu_family }} | py${{ inputs.python_version }} | ${{ inputs.container_image_name }}
    if: ${{ inputs.test_runs_on != '' }}
    runs-on: ${{ inputs.test_runs_on }}
    timeout-minutes: 30
    container:
      image: ${{ contains(inputs.test_runs_on, 'linux') && inputs.container_image_url != '' && inputs.container_image_url || null }}
      options: --ipc host
        --group-add video
        --device /dev/kfd
        --device /dev/dri
        --group-add 992
        --group-add 110
        --env-file /etc/podinfo/gha-gpu-isolation-settings
        --user 0:0 # Running as root, by recommendation of GitHub: https://docs.github.com/en/actions/reference/workflows-and-actions/dockerfile-support#user
    defaults:
      run:
        shell: bash
    env:
      VENV_DIR: ${{ github.workspace }}/.venv

    steps:
      - name: Checkout
        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
        with:
          repository: ${{ inputs.repository || github.repository }}
          ref: ${{ inputs.ref }}

      # actions/setup-python only has prebuilt binaries for Ubuntu
      - name: Set up Python
        if: ${{ inputs.container_image_url == '' || contains(inputs.container_image_name, 'ubuntu') }}
        uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
        with:
          python-version: ${{ inputs.python_version }}

      - name: Compute ROCm device extras
        id: device_extras
        run: |
          python build_tools/github_actions/expand_amdgpu_families.py \
            --amdgpu-families "${{ inputs.amdgpu_family }}" \
            --output-mode=device-extras

      # Key off install_libraries success (not test_libraries) so devel install/test still run
      - name: Install rocm[libraries]
        id: install_libraries
        run: |
          pkg="rocm[libraries]"
          if [ -n "${{ steps.device_extras.outputs.device_extras }}" ]; then
            pkg="${pkg%]},${{ steps.device_extras.outputs.device_extras }}]"
          fi
          python build_tools/setup_venv.py ${VENV_DIR} \
            --packages "${pkg}==${{ inputs.rocm_version }}" \
            --index-url=${{ inputs.package_index_url }} \
            --find-links=${{ inputs.package_find_links_url }} \
            --activate-in-future-github-actions-steps

      - name: Show installed packages (libraries)
        if: ${{ !cancelled() && steps.install_libraries.outcome == 'success' }}
        run: |
          pip freeze

      - name: Test rocm[libraries]
        id: test_libraries
        if: ${{ !cancelled() && steps.install_libraries.outcome == 'success' }}
        timeout-minutes: 5
        run: |
          rocm-sdk test

      - name: Install rocm[libraries,devel]
        id: install_devel
        if: ${{ !cancelled() && steps.install_libraries.outcome == 'success' }}
        run: |
          pkg="rocm[libraries,devel]"
          if [ -n "${{ steps.device_extras.outputs.device_extras }}" ]; then
            pkg="${pkg%]},${{ steps.device_extras.outputs.device_extras }}]"
          fi
          python build_tools/setup_venv.py ${VENV_DIR} \
            --packages "${pkg}==${{ inputs.rocm_version }}" \
            --index-url=${{ inputs.package_index_url }} \
            --find-links=${{ inputs.package_find_links_url }}

      - name: Show installed packages (libraries,devel)
        if: ${{ !cancelled() && steps.install_devel.outcome == 'success' }}
        run: |
          pip freeze

      - name: Test rocm[libraries,devel]
        if: ${{ !cancelled() && steps.install_devel.outcome == 'success' }}
        timeout-minutes: 5
        run: |
          rocm-sdk test