Open philipandag opened 1 day ago
I have re-run the suite increasing the timeouts in the Power On
significantly (10s and 15s) and the whole suite has passed. Solving the issue is just a matter of choosing a less overshot sleep durations.
cpu-cores-count.robot_log.zip
Looks related/similar: https://github.com/Dasharo/open-source-firmware-validation/issues/578
Device
MSI z690 ddr5
RTE version
-
OSFV version
540-reset-to-defaults-restore-serial
Affected component(s) or functionality
dasharo-compatibility/cpu-core-count.robot: CCC test case. Most probably any other case that reboots the platform multiple times too.
Brief summary
The platform sometimes simply does not power on after the keyword
How reproducible
~50% on single tests. The whole CCC suite will surely experience at least one fail
How to reproduce
Run the CCC test suite
Expected behavior
The platform should always power on after the
Power On
KW as multiple tests rely on thatActual behavior
It sometimes simply does not power on. In the failed test cases the platform is turned off after the
Power On
keyword, which was determined with the lack of video output on PiKVM. Powering it on manually usingosfv_cli rte pwr on
after thePower On
keyword fails to do so results in the platform booting normally and the tests continuing.Link to screenshots or logs
Logs from two runs of the CCC test: cpu-cores-count.robot_log.zip
Additional context
I have no clue why it wouldn't work looking at the implementation of the keyword in msi-z690-common. Maybe the sleep times are just too short?
Solutions you've tried
No response