I’m not entirely sure if I fully understood what you were trying to point out, but a positive Pi-Under-Pressure result only counts if the test actually finishes, the report is displayed, and everything in that report is green - success.A question for cmd0s: At higher overclocks using Pi-Under-Pressure, while I did sometimes get segmentation faults or the Pi rebooting itself, I never had it catch and report any errors. Was I just lucky (or unlucky)? What kinds of errors would you think I should have seen from PI-Under-Pressure?
It can happen that instability caused by overclocking allows the test to seemingly reach the end, but the report never gets displayed. That’s why it’s important to wait for the full completion and the final summary.
Interestingly, based on my own experience, a 5-minute test is usually enough to catch errors. Longer runs are mostly useful for confirmation or for identifying thermal-related issues rather than logic or stability errors.
If you have a case where Pi-Under-Pressure reports success, but you still see errors in system logs, please let me know where exactly you’re seeing those errors. That would help me extend the application to monitor a wider range of signals and conditions.
Statistics: Posted by cmd0s — Sun Jan 11, 2026 9:11 pm