Fedora Account System
Red Hat Associate
Red Hat Customer
When logging on my test computer this morning, I see the login screen for a split second, then it transitions to black. Usually, it transitions from black to the login screen, so this seems to happen backwards now oddly. After this happens, all I am left is a cursor. I effectively had to drop to TTY and run startplasma-wayland to get to my desktop.
Proposed as a Blocker for 45-beta by Fedora user farchord using the blocker tracking app because: When booting Fedora KDE, the screen completely goes black, making it impossible for a normal user to get to their desktop. Side note: The system isn't frozen, you can drop to tty and run startplasma-wayland to get to your desktop.
Okay Seems this is intermittent and doesn't always happen. I'll try to record it, but I tried 4-5 reboots in a row, some being from a cold boot and it doesn't happen... And I was able to make it happen twice in a row, maybe I got lucky?
Created attachment 2155808 [details] system log
I also see the crash in drkonqi but honestly not too sure how to make it spit out the information you need.
Steve so far for me on my Thinkpad I cant repo the issue.
Huh. This is the failure mode openQA was seeing yesterday, that seemed to go away after I forced a regen of the f46 buildroot. Note your journal has a crash in it: Aug 27 08:25:58 steve-desktop systemd-coredump[2587]: Process 2374 (plasma-login-gr) of user 965 dumped core. I saw this in openQA too.
Yep it's incredibly random. Machine info: Ryzen 9 9900x3d 128gb ram Radeon 9070xt Let me know if you need more
Created attachment 2155846 [details] Full backtrace Full backtrace, hopefully it helps
Ok so far one of my Thinkpads which is Intel is going good. My main AMD system is looking good. Since its sporadic will keep on it. Adam didn’t Kamil find some issue with the Ryzen 9 cards awhile back?
I don't think this is that. I don't know if this is as simple as "works on hardware X / doesn't work on hardware Y". I think it's more "sometimes it works, sometimes it doesn't" and we don't know the key factor yet...
someone added this comment at the bloccker status discussion: https://forge.fedoraproject.org/quality/blocker-review/issues/2153#issuecomment-1351239
AGREED Delayed Decision (Punt) Discussed at the 2026-08-31 (blocker / freeze exception) review meeting: Though this bug has a lot of ticket votes, the situation seems unclear. Neal and Derek can't reproduce it. openQA only hit it briefly then stopped hitting it after a buildroot regeneration was forced. It is not happening on openQA compose tests either. Two other folks besides Steve claim to have hit it, but it is not clear that the cases are the same. https://meetbot-raw.fedoraproject.org//blocker-review_matrix_fedoraproject-org/2026-08-31/f45-blocker-review.2026-08-31-16.01.log.txt
So can folks who think they're hitting this please provide more details? Let's see: 1) Is your system fully up-to-date with current Fedora 45? 2) Does the problem always happen, or only sometimes? 3) Can you please attach the journal messages from a failed boot (aside from Steve, who already did)?
To answer my share of the questions regarding the logs I posted: 1) Yep, or it was at the time 2) Very seldomly. It happened twice in a row (Two fresh reboots in a row) and then I tried to replicate it, and never succeeded in repeating it since.
I have faced this only one time, and after a distro-sync (my f45 came from a f44 upgrade) and a reboot all was just fine and never hit the bug again. I'm trying to reproduce it in the new Beta 1.2 KDE iso but until now nothing buggy happens.
Rejected as a Beta blocker. See https://forge.fedoraproject.org/quality/blocker-review/issues/2153