tuned-amdgpu/playbook.yml

---
- hosts: localhost
  become: true
  vars:
    # list of source tuned profiles available on Fedora (TODO: should dynamically discover)
    # further modified with AMD GPU power/clock parameters, creating new profiles. eg: 'balanced-amdgpu-VR'
    base_profiles:
      - balanced
      - desktop
      - latency-performance
      - network-latency
      - network-throughput
      - powersave
      - virtual-host
  handlers:
    - name: restart tuned
      ansible.builtin.service:
        name: tuned
        state: restarted
  tasks:
    - name: Gather package facts
      ansible.builtin.package_facts:
        manager: auto
    - name: replace 'power-profiles-daemon' with 'tuned' on Fedora 35+
      dnf: # use with_items since 'dnf' module in Ansible doesn't support 'swap'
        name: "{{ item.name }}"
        state: "{{ item.state }}"
      with_items:
        - {name: 'power-profiles-daemon', state: 'absent'}
        - {name: 'tuned', state: 'present'}
      when:
        - ('power-profiles-daemon' in ansible_facts.packages) or ('tuned' not in ansible_facts.packages)
        - ansible_distribution == 'Fedora'
        - ansible_distribution_major_version|int > 35
      register: fed_ppdtuned_swap
      # 'power-profiles-daemon' was added/conflicts with 'tuned' since F35
      # otherwise, ensure the 'tuned' package is installed
    - name: install tuned
      package:
        name: tuned
        state: present
      when: (fed_ppdtuned_swap is not defined) or ('tuned' not in ansible_facts.packages)
    - name: determine GPU device in drm subsystem
      shell:
        cmd: grep -ls ^connected /sys/class/drm/*/status | grep -o card[0-9] | sort | uniq | tail -1
        executable: /bin/bash
      changed_when: false
      register: card
    - name: find hwmon/max power capability file for {{ card.stdout }}
      find:
        paths: /sys/class/drm/{{ card.stdout }}/device/hwmon
        file_type: file
        recurse: true
        use_regex: true
        patterns:
          - '^power1_cap_max$'
      register: hwmon
    - name: find hwmon/current power limit file for {{ card.stdout }}
      find:
        paths: /sys/class/drm/{{ card.stdout }}/device/hwmon
        file_type: file
        recurse: true
        use_regex: true
        patterns:
          - '^power1_cap$'
      register: powercap_set
    - name: get max power capability for {{ card.stdout }}
      slurp:
        src: "{{ hwmon.files.0.path }}"
      register: power_max_b64
    - name: create custom profile directories
      file:
        state: directory
        path: /etc/tuned/{{ item.1 }}-amdgpu-{{ item.0.key }}
        mode: "0755"
      with_nested:
        - "{{ lookup('dict', amdgpu_profiles) }}"
        - "{{ base_profiles }}"
    - name: template AMDGPU control/reset scripts
      template:
        src: templates/amdgpu-clock.sh.j2
        dest: /etc/tuned/{{ item.1 }}-amdgpu-{{ item.0.key }}/amdgpu-clock.sh
        owner: root
        group: root
        mode: "0755"
      with_nested:
        - "{{ lookup('dict', amdgpu_profiles) }}"
        - "{{ base_profiles }}"
      notify: restart tuned
    - name: template custom tuned profiles
      template:
        src: templates/tuned.conf.j2
        dest: /etc/tuned/{{ item.1 }}-amdgpu-{{ item.0.key }}/tuned.conf
        owner: root
        group: root
        mode: "0644"
      with_nested:
        - "{{ lookup('dict', amdgpu_profiles) }}"
        - "{{ base_profiles }}"
      notify: restart tuned
    - name: ensure tuned is enabled
      service:
        name: tuned
        enabled: true