fix(gitea-actions): serve CI images from the Gitea registry

The CI job images only existed under localhost/ in the gitea-runner store,
and the nightly CI prune deletes any image older than 48h that no container
holds. After every idle stretch CI failed in 0-1s pulling
localhost/gitea-ci:latest until the role was re-run and the images rebuilt.

- Build under git.debyl.io/gitbot/..., push after every run, and pull from the
  registry instead of rebuilding when the Containerfile is unchanged.
- Log gitea-runner in via ~/.docker/config.json, which both act_runner (job
  image pulls) and podman read.
- Label the base images io.debyl.ci-base and skip that label in the CI prune;
  its `until` counts from build time, so a re-pulled image would otherwise be
  deleted again the next night.

Workflows pinning `container: image: localhost/gitea-ci-*` must move to the
registry paths.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

ansible/roles/gitea-actions/defaults/main.yml

@@ -22,18 +22,57 @@ act_runner_bin: /usr/local/bin/act_runner
 act_runner_config_dir: /etc/act_runner
 act_runner_work_dir: /var/lib/act_runner
 
-# Job container images (built locally into the gitea-runner rootless image
+# Job container images. tasks/images.yml builds them into the gitea-runner
-# store by tasks/images.yml; never pulled — force_pull is false).
+# rootless store and pushes them to the Gitea container registry.
-gitea_ci_image: localhost/gitea-ci:latest
+#
+# They used to live only under localhost/, and the nightly podman prune
+# (roles/podman: podman_prune_ci_until) deletes any CI-user image older than
+# 48h that no container is using -- so every idle weekend CI failed in 0-1s on
+# `docker pull localhost/gitea-ci:latest` until someone re-ran this role and
+# waited out a full rebuild. With a registry copy, a pruned image is simply
+# re-pulled by the next job (force_pull stays false, so a present image is
+# never re-pulled), and this role pulls instead of rebuilding when the
+# Containerfile has not changed.
+#
+# Workflows that pin `container: image:` must use these registry paths too
+# (esp-mg-tpms, skudak/esp32-stm32-vcu, skudak/esp32-web-interface).
+gitea_ci_registry: git.debyl.io
+# Namespace = the owner of gitea_registry_username / gitea_registry_token (vault).
+gitea_ci_registry_namespace: gitbot
+gitea_ci_image: "{{ gitea_ci_registry }}/{{ gitea_ci_registry_namespace }}/gitea-ci:latest"
 # ESP-IDF firmware image tag tracks the upstream espressif/idf release we build from.
 esp_idf_version: v5.4.1
-gitea_ci_espidf_image: "localhost/gitea-ci-espidf:{{ esp_idf_version }}"
+gitea_ci_espidf_image: "{{ gitea_ci_registry }}/{{ gitea_ci_registry_namespace }}/gitea-ci-espidf:{{ esp_idf_version }}"
 # PlatformIO image for Arduino-framework ESP32 builds (esp32-web-interface).
 # Tag tracks the pre-baked espressif32 platform version; both pins match the
 # hardware-validated local build.
 platformio_core_version: "6.1.19"
 pio_espressif32_version: "7.0.1"
-gitea_ci_platformio_image: "localhost/gitea-ci-platformio:{{ pio_espressif32_version }}"
+gitea_ci_platformio_image: "{{ gitea_ci_registry }}/{{ gitea_ci_registry_namespace }}/gitea-ci-platformio:{{ pio_espressif32_version }}"
+
+# Registry credentials for the gitea-runner user. The Docker-format path is read
+# by both act_runner (to authenticate job image pulls) and podman (as its
+# fallback auth file), so one login covers the runner and this role.
+gitea_ci_registry_authfile: "{{ gitea_runner_home }}/.docker/config.json"
+
+# Label stamped on the CI base images so the nightly prune skips them; must match
+# podman_prune_ci_keep_label in roles/podman/defaults/main.yml. Without it the
+# prune (whose `until` counts from build time, not pull time) would delete a
+# re-pulled image again the next night -- a 7.8 GB ESP-IDF re-download after
+# every idle day. Superseded tags (e.g. after an esp_idf_version bump) are
+# therefore kept too; remove them by hand.
+gitea_ci_keep_label: io.debyl.ci-base
+
+gitea_ci_images:
+  - image: "{{ gitea_ci_image }}"
+    containerfile: Containerfile.ci
+    template: Containerfile.ci
+  - image: "{{ gitea_ci_espidf_image }}"
+    containerfile: Containerfile.espidf
+    template: Containerfile.espidf.j2
+  - image: "{{ gitea_ci_platformio_image }}"
+    containerfile: Containerfile.platformio
+    template: Containerfile.platformio.j2
 
 # Default labels for every runner — map runs-on values to the local CI image.
 # Firmware jobs opt into the ESP-IDF image per-job via `container:` in their workflow.

ansible/roles/gitea-actions/tasks/images.yml

@@ -1,4 +1,8 @@
 ---
+# CI job images: stage each Containerfile, restore the image from the Gitea
+# registry if the nightly prune removed it, rebuild only when the Containerfile
+# changed, then push so the registry always holds what the runner uses.
+# See gitea_ci_images in defaults/main.yml.
 - name: create CI image build directory
   become: true
   become_user: "{{ gitea_runner_user }}"
@@ -8,71 +12,74 @@
     mode: "0755"
   tags: gitea-actions
 
-- name: stage default CI Containerfile
+- name: create gitea-runner registry auth directory
+  become: true
+  become_user: "{{ gitea_runner_user }}"
+  ansible.builtin.file:
+    path: "{{ gitea_ci_registry_authfile | dirname }}"
+    state: directory
+    mode: "0700"
+  tags: gitea-actions
+
+- name: log gitea-runner in to the Gitea container registry
+  become: true
+  become_user: "{{ gitea_runner_user }}"
+  containers.podman.podman_login:
+    registry: "{{ gitea_ci_registry }}"
+    username: "{{ gitea_registry_username }}"
+    password: "{{ gitea_registry_token }}"
+    authfile: "{{ gitea_ci_registry_authfile }}"
+  environment:
+    XDG_RUNTIME_DIR: "/run/user/{{ gitea_runner_uid }}"
+  no_log: true
+  tags: gitea-actions
+
+- name: stage CI Containerfiles
   become: true
   become_user: "{{ gitea_runner_user }}"
   ansible.builtin.template:
-    src: Containerfile.ci
+    src: "{{ item.template }}"
-    dest: "{{ gitea_runner_home }}/ci-images/Containerfile.ci"
+    dest: "{{ gitea_runner_home }}/ci-images/{{ item.containerfile }}"
     mode: "0644"
-  register: ci_containerfile
+  loop: "{{ gitea_ci_images }}"
+  loop_control:
+    label: "{{ item.containerfile }}"
+  register: ci_containerfiles
   tags: gitea-actions
 
-- name: stage ESP-IDF CI Containerfile
+# A missing image here is normal (first push, or a new tag): the build below
-  become: true
+# creates it. Anything already present locally is left untouched.
-  become_user: "{{ gitea_runner_user }}"
+- name: restore CI images from the registry
-  ansible.builtin.template:
-    src: Containerfile.espidf.j2
-    dest: "{{ gitea_runner_home }}/ci-images/Containerfile.espidf"
-    mode: "0644"
-  register: espidf_containerfile
-  tags: gitea-actions
-
-- name: build default CI image ({{ gitea_ci_image }})
   become: true
   become_user: "{{ gitea_runner_user }}"
   containers.podman.podman_image:
-    name: "{{ gitea_ci_image }}"
+    name: "{{ item.image }}"
-    path: "{{ gitea_runner_home }}/ci-images"
+    auth_file: "{{ gitea_ci_registry_authfile }}"
-    build:
-      file: "{{ gitea_runner_home }}/ci-images/Containerfile.ci"
-    force: "{{ ci_containerfile is changed }}"
   environment:
     XDG_RUNTIME_DIR: "/run/user/{{ gitea_runner_uid }}"
+  loop: "{{ gitea_ci_images }}"
+  loop_control:
+    label: "{{ item.image }}"
+  when: not (ci_containerfiles.results | selectattr('item.image', 'equalto', item.image) | first).changed
+  failed_when: false
   tags: gitea-actions
 
-- name: stage PlatformIO CI Containerfile
+- name: build and push CI images
-  become: true
-  become_user: "{{ gitea_runner_user }}"
-  ansible.builtin.template:
-    src: Containerfile.platformio.j2
-    dest: "{{ gitea_runner_home }}/ci-images/Containerfile.platformio"
-    mode: "0644"
-  register: platformio_containerfile
-  tags: gitea-actions
-
-- name: build ESP-IDF CI image ({{ gitea_ci_espidf_image }})
   become: true
   become_user: "{{ gitea_runner_user }}"
   containers.podman.podman_image:
-    name: "{{ gitea_ci_espidf_image }}"
+    name: "{{ item.image }}"
     path: "{{ gitea_runner_home }}/ci-images"
     build:
-      file: "{{ gitea_runner_home }}/ci-images/Containerfile.espidf"
+      file: "{{ gitea_runner_home }}/ci-images/{{ item.containerfile }}"
-    force: "{{ espidf_containerfile is changed }}"
+      # Exempts the image from the nightly CI prune (gitea_ci_keep_label).
-  environment:
+      extra_args: "--label {{ gitea_ci_keep_label }}=true"
-    XDG_RUNTIME_DIR: "/run/user/{{ gitea_runner_uid }}"
+    force: "{{ (ci_containerfiles.results | selectattr('item.image', 'equalto', item.image) | first).changed }}"
-  tags: gitea-actions
+    push: true
-
+    auth_file: "{{ gitea_ci_registry_authfile }}"
-- name: build PlatformIO CI image ({{ gitea_ci_platformio_image }})
-  become: true
-  become_user: "{{ gitea_runner_user }}"
-  containers.podman.podman_image:
-    name: "{{ gitea_ci_platformio_image }}"
-    path: "{{ gitea_runner_home }}/ci-images"
-    build:
-      file: "{{ gitea_runner_home }}/ci-images/Containerfile.platformio"
-    force: "{{ platformio_containerfile is changed }}"
   environment:
     XDG_RUNTIME_DIR: "/run/user/{{ gitea_runner_uid }}"
+  loop: "{{ gitea_ci_images }}"
+  loop_control:
+    label: "{{ item.image }}"
   tags: gitea-actions

ansible/roles/podman/defaults/main.yml

@@ -311,6 +311,9 @@ podman_prune_ci_users:
   - gitea-runner
   - actions-runner
 podman_prune_ci_until: 48h
+# CI base images built by roles/gitea-actions carry this label (gitea_ci_keep_label
+# there -- keep the two in sync) and are skipped by the CI image prune.
+podman_prune_ci_keep_label: io.debyl.ci-base
 
 # Daily rather than weekly: CI turns over many images a day, and a week of that
 # is what let the store reach 113 GB between runs.

ansible/roles/podman/templates/podman-prune.sh.j2

@@ -43,9 +43,10 @@ run() {
      exec podman "$@"' _ "$@"
 }
 
-# prune_user <user> <until> <prune_containers: yes|no>
+# prune_user <user> <until> <prune_containers: yes|no> [image prune filter...]
 prune_user() {
   local u=$1 keep=$2 do_containers=$3
+  shift 3
   local before after img vol con
 
   if ! id "$u" >/dev/null 2>&1; then
@@ -66,7 +67,7 @@ prune_user() {
     con=$(run "$u" container prune -f --filter "until=$keep" 2>&1 | tail -1)
   fi
 
-  img=$(run "$u" image prune -af --filter "until=$keep" 2>&1 | tail -1)
+  img=$(run "$u" image prune -af --filter "until=$keep" "$@" 2>&1 | tail -1)
   vol=$(run "$u" volume prune -f 2>&1 | tail -1)
 
   after=$(run "$u" system df --format '{{ '{{' }}.Size{{ '}}' }}' 2>/dev/null | head -1)
@@ -80,7 +81,10 @@ for u in {{ podman_prune_users | join(' ') }}; do
 done
 
 for u in {{ podman_prune_ci_users | join(' ') }}; do
-  prune_user "$u" "{{ podman_prune_ci_until }}" yes
+  # CI base images (gitea-ci, -espidf, -platformio) carry the keep label: they
+  # are rebuilt or re-pulled from the registry only when missing, so pruning
+  # them just forces a multi-GB re-download on the next job.
+  prune_user "$u" "{{ podman_prune_ci_until }}" yes --filter "label!={{ podman_prune_ci_keep_label }}"
 done
 
 log "status=ok"