https://bugs.kde.org/show_bug.cgi?id=523606
--- Comment #2 from rogeriolucas <[email protected]> --- (In reply to TraceyC from comment #1) > Thanks for the bug report. I'm not able to reproduce this on Plasma built > from git-master, on a laptop with a fingerprint reader connected to an > external monitor (and only the monitor enabled). Multiple attempts. > > Since this is a race condition, I may just not have hit it. I'll leave this > open so the maintainers can look into it Thanks for taking the time to test this. The window is very small: the outputs are only gone for as long as the display link takes to come back (about half a second on my machine), and the fingerprint match has to complete inside it. I only ever hit it because I touch the reader to wake the machine, so the finger is already on the sensor during the gap - and even then it took several days of normal use for the timing to line up once. Reproducing it also depends on the outputs actually disappearing on wake: here it is a DisplayPort monitor renegotiating its link; if the compositor keeps a screen alive throughout, the window never opens. The dropped success can be observed deterministically, without racing the outputs: run `kscreenlocker_greet --testing` with a copy of the lock screen theme whose onSucceeded handler has been removed. That reproduces the same situation as having no view: PAM reports success ("Success from non-interactive authenticator" in the log) and the greeter stays locked. A fix is up for review at https://invent.kde.org/plasma/kscreenlocker/-/merge_requests/352 - it moves the unlock into the greeter, so it no longer depends on a view being alive. -- You are receiving this mail because: You are watching all bug changes.
