test(int): RF mode/profile matrix integration tests (HIL rung 6)

[Mikefly123]

What

Codifies the RF mode/profile matrix (rung 6 of logs/HIL-REGRESSION-PLAN-fork-migration.md, evidence in logs/HIL-REGRESSION-REPORT-phaseA.md) as repeatable F´ integration tests in PROVESFlightControllerReference/test/int/rf_profile_matrix_test.py:

Test	Coverage	Mark
test_01_tx_profile_sweep	SET_TX_PROFILE P0→P1→P2→P3→P0, ProfileChanged per switch, no ConfigurationFailed/InvalidProfile, NO_OP alive after each	uart_only
test_02_rx_profile_sweep	Same sweep for SET_RX_PROFILE (full RX stop→reconfig→re-arm per switch)	uart_only
test_03_continuous_wave_restore_to_rx	CW on/off, then post-CW RX reconfig + TRANSMIT with no radio error events — complements the issue #207 test by probing the receive chain	uart_only
test_04_post_wake_profile_switch_hammer	Profile switch as first post-wake command (0006 wakeup-race shape), N cycles P2↔P0	uart_only
test_05_post_idle_profile_switch	≥90 s command-idle, then profile switch as first command	uart_only
test_06_two_board_pairing_downlink	Flight→ground RF frames at each TX/RX profile pairing	uart_only, two_board_rf
test_07_two_board_pairing_uplink	Ground→flight RF frame (LastRssi update) at each pairing	uart_only, two_board_rf

All tests are uart_only (profile switches sever any RF GDS link) → they run in integration-uart and are excluded from integration-radio. The two_board_rf tests (new marker in pytest.ini) self-skip unless the USP_GROUND_* env hooks are configured, so CI is unaffected.

Runtime knobs (defaults CI-sane): RF_PROFILE_HAMMER_CYCLES (5; use 100 for a bench hammer), RF_PROFILE_WAKE_IDLE_S (3), RF_PROFILE_LONG_IDLE_S (90), USP_GROUND_DATA_TTY / USP_GROUND_CMD / USP_GROUND_UPLINK_CMD for the two-board setup.

Bench validation (2026-07-23, macOS HIL bench)

Flight v5e on fork-branch build (banner e008def5f20e), ground board on GRC-USP firmware; GDS = fprime_gds.executables.comm on /dev/cu.usbmodem21201 with authenticate-space-data-link framing, seq counters aligned.

• Tests 01–05: PASS (two full module runs; second run 5 passed in 158 s).
• test_06 downlink pairing: PASS — RF bytes observed on the ground data CDC at all four pairings (ground RX profile set via a hil_sender.py wrapper against the GRC GDS).
• test_07 uplink pairing: PASS — flight LastRssi updated at all four pairings (ground TX profile + bypass NO_OP sent through a GDS on the ground data CDC).
• One bench incident during the first test_07 attempt: the flight board's control CDC dropped/re-enumerated right after ground TRANSMIT ENABLED and its uplink stayed dead until an openocd reset run with no host port open; not reproduced on the passing retry (0 serial exceptions). Logged as a watch item — same family as the known device_next CDC OUT-endpoint wedge.

Do not merge — draft for review on top of #439.

🤖 Generated with Claude Code

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: f669298f-e020-43c3-952f-c29dd3e69758

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}