https://bugs.kde.org/show_bug.cgi?id=523637
--- Comment #2 from [email protected] --- Created attachment 194687 --> https://bugs.kde.org/attachment.cgi?id=194687&action=edit Journalctl output requested TraceyC Here is my journalctl output. I've trimmed it of identifying information. Additionally, an AI agent indicates these important events: The flood start — first occurrence of the loop, around 07:34:09. Tell them: "The flood begins here. There are no preceding kwin errors — it starts abruptly with no warning." The cadence — point out the ~16 second interval between bursts (e.g. 07:34:09, 07:34:25, 07:34:42). This looks like a retry timer, not a rendering loop. "Each burst is 4–8 messages, repeating every ~16 seconds consistently throughout the 2+ hour run." The abrupt end — the log simply stops at 09:50:56 mid-flood with no shutdown message, no kwin crash handler, nothing. "The log ends abruptly here — no clean exit, no crash signal caught. The system required a hard power cycle." Context for timing — mention in the comments (not the log itself) that PackageKit completed a get-updates transaction at 07:31, and the flood started at 07:34. "The ~3 minute gap between PackageKit finishing and the flood starting is consistent across multiple occurrences." -- You are receiving this mail because: You are watching all bug changes.
