Something to keep in mind here and in the future:

openpilot doesn't actually care about the specific gear we're in. It just needs to know whether we should engage or not.

All the Bosch C appear to be gearbox 50. Even the Accord which is CVT: 4df2d9d0197e20e3/0000000e--54aa88db6f

I just realized you moved dbc signals names/locations too - will check that out and report back on Bosch C.

All the Bosch C appear to be gearbox 50. Even the Accord which is CVT: 4df2d9d0197e20e3/0000000e--54aa88db6f

Are you sure that route isn't from a hybrid Accord with an e-CVT? It's possible that I didn't get this distinction quite right, it's based on messages present aligned with Wikipedia research:

two messages: one for traditional CVTs, and another for traditional autos, eCVTs and direct-drive EVs

If it makes sense to store the detected difference in a flag rather than trans type, I can do that. But what really matters is that we're coalescing around two messages, each of which have consistent frequencies and signals.

Sorry, read discord history from the driver and it looks like it was a hybrid, and I have no DMs active with CVT Accord drivers. But did confirm that a CVT CRV is using the same CVT signal you build: 2dc4489d7e1410ca/00000000--8cc91f7e24

I presume Accord would be the same.

But there is an issue with Bosch C:

• canfd cvt share the cvt signal with bosch A/B so can reuse your common gearbox include file
• canfd auto has the same gearbox signal starting position number, but the dbc has individual message bits in different spots, some of it overlapping so probably needs a separate dbc.

edit: seems you refactored the canfd pilot signal too. Guess it was fun endian stuff... I'll merge this into my combined Honda build in discord so there is early testing in case any issues arise.

I'll merge this into my combined Honda build in discord so there is early testing in case any issues arise.

Taking longer than anticipated. vanillagorilla didn't build canfd long control safety tests. Unseen until now since safety coverage ci used to timeout on forks. Will fix (target Wednesday evening) in combined-honda before asking people to test.

You don't have to worry about this one too much. It's mostly good but not quite finished or quite fully tested yet, and I'll just add the CI test routes from your new PRs to the validation notebook in commaai/openpilot#35682.

@adeebshihadeh @sshane This is large change to the most crufty part of the Honda port, and touches ALL Hondas. I'm reasonably confident in my testing, but be on the lookout in case I missed something.

Expect process_replay to see a diff on CP.transmissionType, but not for actual gear positions.

***** results for segment regenB328FF8BA0A|2025-04-08--22-57-45--0 *****
    controlsd
    card
        ref: https://commadataci.blob.core.windows.net/openpilotci/regenB328FF8BA0A|2025-04-08--22-57-45--0_card_a6dd96e4a23b0af5b5bebec07c8ab1d2703f91ab.zst
        new: /tmp/openpilot/openpilot/selfdrive/test/process_replay/fakedata/regenB328FF8BA0A|2025-04-08--22-57-45--0_card_006a2e3fccd8a143b1f960386994f51836f55f10.zst

        carParams.transmissionType: 2
    lagd
***** results for segment regen6170C8C9A35|2025-04-08--22-57-46--0 *****
    controlsd
    card
        ref: https://commadataci.blob.core.windows.net/openpilotci/regen6170C8C9A35|2025-04-08--22-57-46--0_card_a6dd96e4a23b0af5b5bebec07c8ab1d2703f91ab.zst
        new: /tmp/openpilot/openpilot/selfdrive/test/process_replay/fakedata/regen6170C8C9A35|2025-04-08--22-57-46--0_card_006a2e3fccd8a143b1f960386994f51836f55f10.zst

        carParams.transmissionType: 2
    lagd

Looks like one of the fuzzing tests failed: https://github.com/commaai/opendbc/actions/runs/16451100901/job/46496300736

Looks like one of the fuzzing tests failed: https://github.com/commaai/opendbc/actions/runs/16451100901/job/46496300736

#2525, hopefully. I'm not fully clear on what the fuzzer was changing and it doesn't seem to happen every run.

Integra CVT uses the auto message 0x191. Seems to cause no problems, but the label isn't 100% precise.

ad9840558640c31d/0000011e--89f22afa86