tests: subsys: canbus: isotp: implementation: fails on nucleo_g474re

[ABOSTM]

Describe the bug
tests/subsys/canbus/isotp/implementation/ fails on nucleo_g474re.

This test is failed on main branch.
But even when reverting to original introduction of FDCAN on nucleo_g474re,
#31061 with sha1 3b2fd1c
it is still failed

To Reproduce
Steps to reproduce the behavior:

1. west build -p auto -b nucleo_g474re tests/subsys/canbus/isotp/implementation/
2. west flash
3. See error

Expected behavior
test passed

Logs and console output

*** Booting Zephyr OS build v2.6.0-rc2-39-g94e8c2eb81f3  ***
Running test suite isotp
===================================================================
START - test_bind_unbind
 PASS - test_bind_unbind in 0.905 seconds
===================================================================
START - test_send_receive_net_sf
 PASS - test_send_receive_net_sf in 0.6 seconds
===================================================================
START - test_send_receive_net_blocks
 PASS - test_send_receive_net_blocks in 1.4 seconds
===================================================================
START - test_send_receive_net_single_blocks

E: Reception of next FC has timed out
    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:296: test_send_receive_net_single_blocks: (ret not equal to 0)
recv returned -14

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:61: send_complette_cb: (error_nr not equal to ISOTP_N_OK)
Sending failed (-2)
 FAIL - test_send_receive_net_single_blocks in 1.98 seconds
===================================================================
START - test_send_receive_sf

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:97: get_sf: (ret not equal to 1)
recv returned -14
 FAIL - test_send_receive_sf in 1.14 seconds
===================================================================
START - test_send_receive_blocks
E: No buffer for FF left

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:268: test_send_receive_blocks: (ret not equal to 0)
Binding failed (-11)
 FAIL - test_send_receive_blocks in 0.18 seconds
===================================================================
START - test_send_receive_single_block
E: No buffer for FF left

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:320: test_send_receive_single_block: (ret not equal to 0)
Binding failed (-11)
 FAIL - test_send_receive_single_block in 0.18 seconds
===================================================================
START - test_buffer_allocation
E: No buffer for FF left

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:393: test_buffer_allocation: (ret not equal to 0)
Binding failed (-11)
 FAIL - test_buffer_allocation in 0.18 seconds
===================================================================
START - test_buffer_allocation_wait
E: No buffer for FF left

    Assertion failed at WEST_TOPDIR/zephyr/tests/subsys/canbus/isotp/implementation/src/main.c:409: test_buffer_allocation_wait: (ret not equal to 0)
Binding failed (-11)
 FAIL - test_buffer_allocation_wait in 0.18 seconds
===================================================================
Test suite isotp failed.
===================================================================
PROJECT EXECUTION FAILED

Environment (please complete the following information):

• OS: Windows
• Toolchain gnuarmemb
• Commit SHA 94e8c2e

[ABOSTM]

@alexanderwachter Is there something I miss to pass the test, any loopback wire, transceiver, ...?
(I am not CAN expert)

[str4t0m]

@ABOSTM no external fixture is required to run the test.
I ran the test locally with success on g0 and g4. As CAN_FD_MODE is enabled per default, #35738 may have fixed this.
Would you mind running the test on the latest main?

Edit: Ran the test with a few modifications, but the test passes also as is.

[alexanderwachter]

We have issues with that tests ATM. There are race conditions in the testcase (context is reused before the work queue item was processed).
We are working on a fix.

[ABOSTM]

@str4t0m I still reproduce the issue with latest main
@alexanderwachter ok thanks

[github-actions]

This issue has been marked as stale because it has been open (more than) 60 days with no activity. Remove the stale label or add a comment saying that you would like to have the label removed otherwise this issue will automatically be closed in 14 days. Note, that you can always re-open a closed issue at any time.

[ABOSTM]

@alexanderwachter,
Issue is still reproduced with zephyr master.
do you have some progress with the fix you are talking about ? any reference for the fix ?

[erwango]

@alexanderwachter gentle ping.

[alexanderwachter]

I think this error is due to a re-use of the tx context before the the workqueue has finished. Needs a fix in either the tests, or better prevent such a situation in the stack somehow. @legoabram you have been hunting this bug lately. Any news?

[legoabram]

At the very least, I can confirm that this is the case. There are also some issues related to corrupting the work queue stack by reinitializing the work item before it's completely done. I haven't had to time to work on a fix, but I think the cleanest fix is to modify the library so that anything in the context is fully deallocated or uninitialized before notifying the user that the operation is complete.

[github-actions]

This issue has been marked as stale because it has been open (more than) 60 days with no activity. Remove the stale label or add a comment saying that you would like to have the label removed otherwise this issue will automatically be closed in 14 days. Note, that you can always re-open a closed issue at any time.