Error when attempting to stop messages with same arbitration ID

[sbourkedev]

This error is returned when attempting to stop messages that have the same arbitration ID

Traceback (most recent call last):
File "/usr/local/lib/python3.5/dist-packages/can/interfaces/socketcan/socketcan.py", line 280, in send_bcm
  return bcm_socket.send(data)
OSError: [Errno 22] Invalid argument

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "simple-nm-sim.py", line 238, in <module>
  sim.force_stop()
File "simple-nm-sim.py", line 179, in force_stop
  self.vin1_task.stop()
File "/usr/local/lib/python3.5/dist-packages/can/bus.py", line 218, in wrapped_stop_method
  original_stop_method()
File "/usr/local/lib/python3.5/dist-packages/can/interfaces/socketcan/socketcan.py", line 365, in stop
  send_bcm(self.bcm_socket, stopframe)
File "/usr/local/lib/python3.5/dist-packages/can/interfaces/socketcan/socketcan.py", line 285, in send_bcm
  raise can.CanError(base + "You are probably referring to a non-existing frame.")
can.CanError: Couldn't send CAN BCM frame. OS Error 22: Invalid argument
You are probably referring to a non-existing frame.

Below is the code snippet where messages are defined and also where messages are being stopped. pm_snapshot_task is stopped first then an error is thrown when attempting to stop vin1_task as it has the same arbitration ID.

self.pm_snapshot_msg = can.Message(
            arbitration_id=0x400,
            data=self.pm_snapshot_data,
            extended_id=False,
            dlc=8)

        self.nm_msg = can.Message(
            arbitration_id=0x51E,
            data=self.nm_data,
            extended_id=False,
            dlc=8)

        self.vs_msg = can.Message(
            arbitration_id=0x11,
            data=self.vs_data,
            extended_id=False,
            dlc=8)

        self.vin1_msg = can.Message(
            arbitration_id=0x400,
            data=self.vin1_data,
            extended_id=False,
            dlc=8)

Stopping the tasks

if self.pm_snapshot_task is not None:
            self.pm_snapshot_task.stop()
            self.pm_snapshot_task = None

        if self.vs_task is not None:
            self.vs_task.stop()
            self.vs_task = None

        if self.vin1_task is not None:
            self.vin1_task.stop()
            self.vin1_task = None

[karlding]

Sorry, just realized this went untriaged all this time! The reason for the behaviour that you're observing is pretty simple.

This is because python-can naively wraps the TX_SETUP BCM opcode when setting up periodic transmissions. The semantics that the kernel driver provides for TX_SETUP are essentially an "upsert", where if the number of frames is equal to the original periodic task, then it updates the existing task, otherwise it creates a new task. As with all abstractions, the Task is a leaky abstraction (for what the kernel is doing).

We could potentially fix this by issuing a TX_READ first before creating a new Task, and seeing if one already exists. If we get EINVAL, then we should be able to proceed with creating a new periodic transmission task, since none exists. Otherwise, we can raise a ValueError or some other Exception.

Alternatively, we can keep this all in userspace, and have python-can keep track of currently active periodic tasks, so we don't have the additional overhead of a socket read/write for each task that's created. I don't think the overhead is too big of a deal, but I'm open to either solution.

Note: I think in order to do what you intended to do, you can create a periodic transmission with an array of 3 elements, or modify your periodic task with the message contents as needed, instead of trying to create 3 separate Tasks.