-
Notifications
You must be signed in to change notification settings - Fork 33
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Sync: commaai/opendbc:master
into sunnypilot/opendbc:master-new
#75
Conversation
* reimplement * use average wheel speed * fix factor with eEgoRaw * Revert "fix factor with eEgoRaw" This reverts commit 35f3624. * Revert "use average wheel speed" This reverts commit ec4e9c1. * revert to previous * mutation again ugh * mutation you got to be kidding me * nice, this passes test models now --------- Co-authored-by: Shane Smiskol <[email protected]>
* test test_angle_cmd_when_enabled above max angle and set them * nissan is too high * clean up * more
* Hyundai CAN FD: Steering API updates * comments * comments * keep it for now * not used * should use DBC spec * consistent --------- Co-authored-by: Shane Smiskol <[email protected]>
* struct to prevent forgetting something and make more organized * use it * move all over * clean up * rm
Revert "Hyundai CAN FD: ADAS steering API updates (commaai#1873)" This reverts commit 2a75806.
* restart * fix * missing * remove stuff we don't use * huh * fix * fix fix fix * we haven't seen LKAS with angle yet * this is missing from the new message * ... * more explicit * flip * import
…#1890) * Update replay_drive.py * warn * no true * tqdm too
Use gcovr instead of gcov to support exclusion markers. This also replaces lcov, since gcov has far fewer dependencies, can be installed via uv (and looks better if anything).
* curvature safety: limit to ISO max * . * cmt * implement in carcontroller * MISRAAAAAA * use ~average road roll * stash * clean up * compiles * fix safety replay for ford can fd * always init * stash * fix replay * fix that * safety doesn't consider rate limits when clipping max curvature consider: at the max curvature limit and user gas overrides heavily, max curvature limit might drop quicker than the safet rate limit will allow for * don't limit for CAN * do tests (need to clean up) * stash * rm * clean up tests * fix that * smaller * fix capnp error * because we allow tolerance on the error limiting, we need the same type of tolerance on the max accel curvature to not block * fix a test * fix another * consistent naming * debugging * debug * remove * stash * clean up * clean up * minor * and this * space * misra * doesn't matter here * safety mutation test fix * misra false positive? * rename * ah good catch, these are equivalent * comment * clean up * rename like safety * fix * simpler * not needed
CAN FD out of ALLOW_DEBUG
* debug * remove debugging * huh
* fixup * test * bye copilot * rev * fix * fix * test again * revert * need
* Only run tx hooks on msgs in safety_config.tx_msgs. * 'fix' tests * retrigger checks bc timeouts * not hit * formatting * add back --------- Co-authored-by: Shane Smiskol <[email protected]>
* in upstream * fix tests
these are the same
* no tron (2010) * switch to actual ACC message * switch to camera * add logging
add fingerprint
* highlight * not here
…p fail (commaai#1919) add checkout check to prevent opendbc bump fail
* add VDM fault signal * fix
* support support-type * fix * fix * fix
stock longitudinal when stock long too
* add candidate cancel command and test * fixes * this was correct * try spacing out the messages * try adas status to acm to cancel * try a bunch of stuff * send the entire range * try faulted * try two faults * try faster * 50hz * Revert "50hz" This reverts commit babcbdf. * no fault on dash * 50hz * see if "unavailable" also works * try sending just 1 * 100 Hz * Revert "100 Hz" This reverts commit 29ae625. * Revert "try sending just 1" This reverts commit 64908eb. * revert other stuff - forgot counter! * Revert "revert other stuff - forgot counter!" This reverts commit 5ce09b7. * try counter * final thing * 100 hz * Revert "100 hz" (faults) This reverts commit bea3977. * nvm we can't do this since the state management to know when openpilot stops sending cancel is too complex * Revert "nvm we can't do this since the state management to know when openpilot stops sending cancel is too complex" This reverts commit 19fb385. * forward through openpilot * try these two * let's also try this * are we doing this right? * temp fault? * ? * try oscillating?! * that works?! * maybe ACM needs to see VDM status rise before VDM can request cancel with a falling edge * try to cancel immediately (1 frame of available) * no counter * cmts * 3 frames * comments * clean up safety * clean up * cc clean up * typo * comments * clean this up too * and this too
* fix possible gear exception * fix cruise state mismatch * clean up
* Rivian: don't show set speed in UI * carparam * just use speed
…e test speed (commaai#1930) * Replace @parameterized.expand with @pytest.mark.parametrize for faster test collection This change reduces the opendbc/car/tests/ collection time to ~0.4s. * clean up --------- Co-authored-by: Shane Smiskol <[email protected]>
so far only tesla has been seen being able to set to 0
* Update interface.py * Update fingerprints.py * Revert "Update interface.py" This reverts commit a2abbc0.
* Update fingerprints.py * 2020 confirmed! * should be fine --------- Co-authored-by: Jason Young <[email protected]>
* parse vin * Update opendbc/car/vin.py
* TODO: cant pass down due to circular imports * revert * screw it, we can split vin.py and vin_query.py later. this is nicer * Revert "screw it, we can split vin.py and vin_query.py later. this is nicer" This reverts commit 9ddab33.
* add match_fw_to_car_fuzzy * remove comment * fix dbc * format * don't remove... * parse_vin function * test caught bug * safe string accessing --------- Co-authored-by: Shane Smiskol <[email protected]>
* this is what we want * rm old * and now counter * need this * toyota * subaru's good * Tesla and Chrysler * Ford * do some find and replace * do the rest * clean up * remove this * rm * not broken but bad * this was wrong!
* set FWD_BUS_LOOKUP to most common * this is always assumed to be a number, test models tests mismatches * clean up * and this
only what we've triggered
…aster-new # Conflicts: # opendbc/car/hyundai/hyundaicanfd.py # opendbc/car/rivian/carcontroller.py # opendbc/safety/tests/common.py # opendbc/safety/tests/test_chrysler.py # opendbc/safety/tests/test_hyundai_canfd.py
* Fix get_params_for_docs call to use correct parameters Updated the function call to match the required arguments by removing the redundant 'model' parameter. This ensures compatibility with the function definition and prevents potential errors. * Add _tx_acc_state_msg method and standardize PT_BUS usage Introduce the _tx_acc_state_msg method in Hyundai safety tests to handle ACC state messaging. Replace hardcoded bus values with the self.PT_BUS variable for consistency and maintainability across SCC_CONTROL messages. * Add MADS heartbeat engaged state validation test Introduce a new test function `test_heartbeat_engaged_mads_check` to verify the behavior of MADS heartbeat engagement logic under varying conditions. This ensures correct state transitions and validates robustness against mismatched heartbeat signals. * Increase timeout for safety mutation tests to 60 minutes Extended the timeout for the entire safety mutation tests workflow and its individual steps from 20 and 10 minutes respectively to 60 minutes. This change ensures that longer-running tests complete successfully without premature termination.
Reviewer's Guide by SourceryThe pull request synchronizes changes from Sequence diagram for Rivian ACC cancelsequenceDiagram
participant User
participant CarController
participant VDM_AdasSts
User->>CarController: Cancels ACC
activate CarController
CarController->>VDM_AdasSts: Sets VDM_AdasInterfaceStatus to 1 (available) for 5 frames, then 0 (unavailable)
activate VDM_AdasSts
VDM_AdasSts-->>CarController: Sends VDM_AdasSts message
deactivate VDM_AdasSts
loop for 5 frames
CarController->>VDM_AdasSts: Sets VDM_AdasInterfaceStatus to 0
activate VDM_AdasSts
VDM_AdasSts-->>CarController: Sends VDM_AdasSts message
deactivate VDM_AdasSts
end
deactivate CarController
Sequence diagram for Ford curvature limitssequenceDiagram
participant CarController
participant EPS
CarController->>CarController: Calculates desired curvature
CarController->>CarController: Applies rate limits and curvature error limits
CarController->>CarController: Checks if CANFD flag is set
alt CANFD flag is set
CarController->>CarController: Limits curvature based on max lateral acceleration
end
CarController->>EPS: Sends steering command with limited curvature
EPS-->>CarController: Acknowledges steering command
Updated class diagram for AngleSteeringLimitsclassDiagram
class AngleSteeringLimits {
float STEER_ANGLE_MAX
tuple[list[float], list[float]] ANGLE_RATE_LIMIT_UP
tuple[list[float], list[float]] ANGLE_RATE_LIMIT_DOWN
}
Updated class diagram for RivianPlatformConfigclassDiagram
class RivianPlatformConfig {
DbcDict dbc_dict
set[WMI] wmis
set[ModelLine] lines
set[ModelYear] years
}
Updated class diagram for VinclassDiagram
class Vin {
str vin
str wmi
str vds
str vis
}
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We've reviewed this pull request using the Sourcery rules engine.
No description provided.