Skip to content

build(ci): compile OAE and WIFI configurations in the build matrix - #329

Open
ClutchplateDude wants to merge 1 commit into
developfrom
build/ci-oae-wifi-variants
Open

ClutchplateDude wants to merge 1 commit into
developfrom
build/ci-oae-wifi-variants

Conversation

@ClutchplateDude

Copy link
Copy Markdown
Member

Fixes #325

What

Three changes so the previously compile-dead configurations get built by CI:

  1. oaeboardv1 joins the board matrix (new Build (oaeboardv1) job). All four steppers always enabled — RA/DEC on TMC2209 UART (on-board, MS pins not routed), AZ/ALT on the on-board A4988s — varying INFO_DISPLAY_TYPE (incl. the OAE info display), RA_DEC_OFF_WHEN_PARKED (the fix(mount): Parked status, optional RA/DEC off when parked #323 option), TEST_VERIFY_MODE and DEBUG_LEVEL: 16 solutions.
  2. WIFI matrix variants on esp32: WIFI_ENABLED/WIFI_MODE flags; WIFI-on builds each of the three modes once on a canonical carrier config (no displays, UART drivers, no test-verify) so esp32 grows from 24 to 27 solutions instead of multiplying. Other boards stay at the WIFI-off default. This compiles src/WifiControl.cpp in CI for the first time (gap noted on refactor: Refactor mount calculations, enhance documentation, and add tests #314/fix(wifi): keep FAIL_TO_AP mode reachable and truthful after failover #315).
  3. OAE is derived from BOARD_OAE_V1 in Configuration.hpp (guarded by #ifndef), so the oaeboardv1 environment is self-sufficient. Existing user configs that define OAE themselves are unaffected.

Solution counts: esp32 27 (was 24), oaeboardv1 16 (new), all other boards unchanged.

Verification

  • Ran the constraint solver locally over the full matrix: counts above, sanity-checked (WIFI only on esp32, RA_DEC_OFF_WHEN_PARKED only on OAE, mode consistent with enabled).
  • Built five representative solutions with the real generate_config_file() and PLATFORMIO_BUILD_FLAGS=-DMATRIX_LOCAL_CONFIG=1 (the exact CI invocation), each from its own generated config:
    • OAE with RA_DEC_OFF_WHEN_PARKED=1 ✅
    • OAE baseline (option off) ✅
    • OAE with AZ/ALT A4988 enabled ✅
    • esp32 with WIFI_MODE_ATTEMPT_INFRASTRUCTURE_FAIL_TO_AP ✅ (WifiControl.cpp.o confirmed in the build log)
    • esp32 WIFI-off regression ✅
  • One incidental validation win while testing: feeding an OAE build an esp32 STANDALONE-driver config was correctly rejected by ConfigurationValidation (missing MS pins → -Werror), so the guard rails around the new variants are live.

🤖 Generated with Claude Code

No CI job built oaeboardv1, and no matrix configuration ever set
WIFI_ENABLED, so all OAE-only code paths (OAE geometry, the
RA_DEC_OFF_WHEN_PARKED option from #323, the onboard A4988 AZ/ALT
drivers) and all of WifiControl.cpp were compile-dead in CI.

- Add oaeboardv1 to the board matrix: all four steppers always enabled
  (RA/DEC TMC2209 UART, AZ/ALT A4988), varying INFO_DISPLAY_TYPE,
  RA_DEC_OFF_WHEN_PARKED, TEST_VERIFY_MODE and DEBUG_LEVEL (16
  solutions).
- Add WIFI_ENABLED/WIFI_MODE matrix flags on esp32: WIFI_ENABLED=1
  builds each of the three wifi modes once, on a canonical carrier
  configuration (no displays, UART drivers) so the matrix grows by 3
  solutions instead of multiplying. All other boards stay at the
  WIFI_ENABLED=0 default.
- Derive OAE from BOARD_OAE_V1 in Configuration.hpp when the local
  configuration did not define it, so the oaeboardv1 environment is
  self-sufficient (user-provided OAE configs are unaffected).

Fixes #325

Co-Authored-By: Claude Code <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

CI never compiles several supported configurations (oaeboardv1, RA_DEC_OFF_WHEN_PARKED, WIFI_ENABLED)

1 participant