tests/kernel/timer/timer_api/kernel.timer sporadically (high frequency) fails in CI on qemu-xtensa

[nashif]

***** Booting Zephyr OS zephyr-v1.14.0-1113-g8806967aed2e *****
Running test suite timer_api
===================================================================
starting test - test_timer_duration_period
PASS - test_timer_duration_period
===================================================================
starting test - test_timer_period_0

    Assertion failed at ./tests/kernel/timer/timer_api/src/main.c:167: test_timer_period_0: tdata.expire_cnt == 1 is false

FAIL - test_timer_period_0
===================================================================
starting test - test_timer_expirefn_null
PASS - test_timer_expirefn_null
===================================================================
starting test - test_timer_periodicity
PASS - test_timer_periodicity
===================================================================
starting test - test_timer_status_get
PASS - test_timer_status_get
===================================================================
starting test - test_timer_status_get_anytime
PASS - test_timer_status_get_anytime
===================================================================
starting test - test_timer_status_sync
PASS - test_timer_status_sync
===================================================================
starting test - test_timer_k_define

    Assertion failed at ./tests/kernel/timer/timer_api/src/main.c:394: test_timer_k_define: tdata.expire_cnt == 4 is false

FAIL - test_timer_k_define
===================================================================
starting test - test_timer_user_data
PASS - test_timer_user_data
===================================================================
starting test - test_timer_remaining_get
PASS - test_timer_remaining_get
===================================================================
Test suite timer_api failed.
===================================================================
PROJECT EXECUTION FAILED

[andrewboie]

I don't think this should be high, only seen with QEMU which is known to have timing issues since we don't use icount

In fact, if this is a QEMU timing issue, we should just close this, since that is adequately covered by other bugs

[andyross]

I'll take spurious bugs, regardless. Drop to medium for now, will do another pass with a modified rig to rule out timing on tests like this again.

[andyross]

This had 1400+ iterations for me over the weekend using the tooling in #18656 without a single failure. I buy that this could certainly be more likely to fail on this platform and this test than others, it just doesn't seem like it's any different fundamentally than the timing artifacts we already know about.

I think we should close this. If the failures are so frequent that they're impacting CI reliability, maybe we can come up with some kind of workaround.

[nashif]

agree with @andyross. I can't reproduce this anymore and it does not show up in CI anymore, must have been fixed somehow...