Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Q4 fingerprint #1754

Closed
blue-genie opened this issue Feb 20, 2025 · 6 comments
Closed

Q4 fingerprint #1754

blue-genie opened this issue Feb 20, 2025 · 6 comments

Comments

@blue-genie
Copy link
Contributor

blue-genie commented Feb 20, 2025

Not yet officially supported yet, just getting ready for Hackaton:

It seems that sporadically errors similar to the below one are logged:

[2.972794] isotp_parallel_query.py:146 - : iso-tp query bad response: (1990, None) - 0x7f2231
[2.972962] isotp_parallel_query.py:146 - : iso-tp query bad response: (1990, None) - 0x7f2231
[2.973382] isotp_parallel_query.py:146 - : iso-tp query bad response: (1840, None) - 0x7f2231
[2.973458] isotp_parallel_query.py:146 - : iso-tp query bad response: (1840, None) - 0x7f2231
[2.973865] isotp_parallel_query.py:146 - : iso-tp query bad response: (1988, None) - 0x7f2231
[2.973961] isotp_parallel_query.py:146 - : iso-tp query bad response: (1988, None) - 0x7f2231

Which results in no fingerprinting.

Route: e36b272d5679115f/00000307--2e7636e703

./selfdrive/debug/fingerprint_from_route.py e36b272d5679115f/00000307--2e7636e703/0 returns

found 33 messages. CAN fingerprint:

65: 8, 66: 8, 73: 8, 119: 8, 125: 8, 126: 8, 130: 8, 145: 8, 146: 8, 357: 8, 359: 8, 390: 8, 391: 8, 512: 8, 514: 8, 516: 8, 517: 8, 531: 8, 532: 8, 534: 8, 535: 8, 560: 8, 578: 8, 936: 8, 937: 8, 943: 8, 972: 8, 976: 8, 981: 8, 1045: 8, 1070: 8, 1111: 8, 1200: 8

FW fingerprint:

Traceback (most recent call last):
  File "/DevWork/op/./selfdrive/debug/fingerprint_from_route.py", line 46, in <module>
    get_fingerprint(lr)
  File "/DevWork/op/./selfdrive/debug/fingerprint_from_route.py", line 32, in get_fingerprint
    for f in fw:
TypeError: 'NoneType' object is not iterable

kill openpilot and restart results in successful fingerprinting, route: e36b272d5679115f/00000308--b61d26edc1/0

right before the bad fingerprinting I updated cereal/services.py, started launch_openpulot.py and I noticed that scons had work to do. After bad fingerprinting i killed launch_openpilot.py and reran it, scons had no work to do and it fingerprinted correctly.

@sshane
Copy link
Contributor

sshane commented Feb 20, 2025

That route is only a couple seconds long, so it could still have fingerprinted if you left it. This route is a better example: e36b272d5679115f/00000302--1a2a40e714

@sshane
Copy link
Contributor

sshane commented Feb 20, 2025

Please do not report issues with dirty git diffs/local changes for in-progress work. card crashed due to local changes related to the radar interface

Traceback (most recent call last):
  File "/data/openpilot/openpilot/system/manager/process.py", line 41, in launcher
    mod.main()
  File "/data/openpilot/selfdrive/car/card.py", line 286, in main
    car = Car()
          ^^^^^
  File "/data/openpilot/selfdrive/car/card.py", line 102, in __init__
    self.RI = get_radar_interface(self.CI.CP)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/openpilot/opendbc/car/car_helpers.py", line 160, in get_radar_interface
    return RadarInterface(CP)
           ^^^^^^^^^^^^^^^^^^
  File "/data/openpilot/opendbc/car/ford/radar_interface.py", line 128, in __init__
    elif self.radar == RADAR.STEER_ASSIST_DATA:
                       ^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: type object 'RADAR' has no attribute 'STEER_ASSIST_DATA'

@sshane sshane closed this as not planned Won't fix, can't repro, duplicate, stale Feb 20, 2025
@sshane
Copy link
Contributor

sshane commented Feb 20, 2025

commaai/openpilot#34579 should improve this by showing an alert that something crashed though

@blue-genie
Copy link
Contributor Author

Please do not report issues with dirty git diffs/local changes for in-progress work. card crashed due to local changes related to the radar interface

Traceback (most recent call last):
  File "/data/openpilot/openpilot/system/manager/process.py", line 41, in launcher
    mod.main()
  File "/data/openpilot/selfdrive/car/card.py", line 286, in main
    car = Car()
          ^^^^^
  File "/data/openpilot/selfdrive/car/card.py", line 102, in __init__
    self.RI = get_radar_interface(self.CI.CP)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/openpilot/opendbc/car/car_helpers.py", line 160, in get_radar_interface
    return RadarInterface(CP)
           ^^^^^^^^^^^^^^^^^^
  File "/data/openpilot/opendbc/car/ford/radar_interface.py", line 128, in __init__
    elif self.radar == RADAR.STEER_ASSIST_DATA:
                       ^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: type object 'RADAR' has no attribute 'STEER_ASSIST_DATA'

Yes, I'm aware that other routes aren failing because local changes, and his is why I only mentioned routes 00000307 and 00000308.

I was not aware that once the dash-cam message for unrecognized car is displayed, it can still be fingerprinted.

@sshane
Copy link
Contributor

sshane commented Feb 20, 2025

Those routes had a exception as well, I didn't see any mock routes. Dashcam is also shown if it fingerprints but the dashcam flag is set in carParams

@blue-genie
Copy link
Contributor Author

blue-genie commented Feb 20, 2025

On e36b272d5679115f/00000308--b61d26edc1 I drove for 10 minutes with OP in control, The only difference between 307 and 308 was a reboot.

When it fails to fingerprint I will let it run longer.

Dashcam check/block was removed - so I can test the fingerprinting and a POC of the radar parsing on Q4, similar to Q3

# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants