https://bugs.kde.org/show_bug.cgi?id=517499
--- Comment #19 from Aaron Tinio <[email protected]> --- Created attachment 195400 --> https://bugs.kde.org/attachment.cgi?id=195400&action=edit Controlled undock/redock reproduction, 2026-08-23 (Plasma/KWin 6.7.4) Controlled undock/redock reproduction attached. At 17:44:47 I undocked. The live state became eDP-1 only, priority 1. The panel was temporarily on screen -1; the workaround moved it to internal screen 0. The delayed sample was stable. At 17:45:16 I redocked. The first DRM-triggered sample already showed DP-2 and eDP-1 both enabled at priority 1. Plasma ordered the internal display as screen 0 and the external display as screen 1, with the panel still on internal screen 0. At 17:45:18 the workaround moved it to external screen 1; the 17:45:24 delayed sample was stable. The kwinoutputconfig topology hash was unchanged across the cycle. So this captures reconnect restoring an already-persisted duplicate-priority topology into live state and reproducing the wrong panel placement; it does not capture the original creation of the persisted tie. The attachment contains eight correlated JSONL records. I removed the boot ID and relabeled the two correlations. It contains no EDID, serial, or output UUID. -- You are receiving this mail because: You are watching all bug changes.
