tests/kernel/sched/schedule_api: Assertion failed for test_slice_scheduling

[cinlyooi-intel]

Describe the bug
'''
Assertion failed at src/test_slice_scheduling.c:51: thread_tslice: ((tdelta >= expected_slice_min) && (tdelta <= expected_slice_max) && ((int)p1 == thread_idx)) is false
'''

To Reproduce
Steps to reproduce the behavior:

1. mkdir build; cd build
2. cmake -DBOARD=nrf51_pca10028 -DARCH=arm
   cmake -DBOARD=nrf52_pca10040 -DARCH=arm
3. make
4. make run

Screenshots or console output

===================================================================
starting test - test_slice_scheduling
AB
A
Assertion failed at zephyr.git/tests/kernel/sched/schedule_api/src/test_slice_scheduling.c:51: thread_tslice: ((tdelta >= expected_slice_min) && (tdelta <= expected_slice_max) && ((int)p1 == thread_idx)) is false

B
Assertion failed at zephyr.git/tests/kernel/sched/schedule_api/src/test_slice_scheduling.c:51: thread_tslice: ((tdelta >= expected_slice_min) && (tdelta <= expected_slice_max) && ((int)p1 == thread_idx)) is false

 Assertion failed at zephyr.git/tests/kernel/sched/schedule_api/src/test_slice_scheduling.c:51: thread_tslice: ((tdelta >= expected_slice_min) && (tdelta <= expected_slice_max) && ((int)p1 == thread_idx)) is false

FAIL - test_slice_scheduling
======================================================

Environment (please complete the following information):

• OS: Fedora 29
• Toolchain : zephyr
• Commit SHA or Version used: dd5ff39

[ghost]

@cinlyooi-intel Are you running this test on real hardware?

[andyross]

And is this reliable? I was looking at exactly this test on a nrf52840_pca10056 ~2-3 weeks ago and it was running great for me. If it's failing, even spuriously, it's likely to be a recent regression.

[andyross]

Read more carefully, this is a duplicate of #11721, which got fixed by #13916 last week but is not included in the SHA referenced in the report above.