Latest update causes constant logout - have to log back in again

Recently loaded F39 and do updates on Friday evenings. All was good until 15 Dec update. There was no issues with the update until I rebooted. I get the log in screen - enter credentials with no issue, and the desktop starts to load (I get a dark screen and the top menu) but never finishes. When ever I click on something like menu I get thrown back to the login screen. Each login does the same thing. I can ssh to the box with no issues. I can vnc to the box with no issues (everything is normal and works great). I reloaded the OS. Everything works great on the initial install, but when I dnf update - it happens again - constant logout - relogin. I spun up a second test box (same hardware) and it does the exact same thing. I have other F39 boxes with different hardware and they updated fine and are working fine.

Any ideas or should I do a bug

Try a dnf clean all then try the upgrade again.

I wonder what you have for a GPU in that machine. I have seen several similar issues when an nvidia GPU is used while trying to use the default nouveau drivers.

If you have an nvidia gpu with the nvidia drivers it is also possible that the drivers may not have been properly updated when the upgrade occurred on the 15th.

If you already have the nvidia drivers and this is happening the fix is usually very easy.

  1. sudo dnf remove kmod-nvidia-$(uname -r)
  2. sudo akmods --force
  3. reboot

I’m seeing the same issue. Going back to earlier kernels doesn’t seem to work. Saw another thread about downgrading the mutter version but haven’t gotten this to work.
It seems like anytime I press the “super” key, I am booted out of my session.
If I open more than one program from my task bar, I get booted out

This thing is pretty old (i5 with a Windoz Vista sticker). I have heard stories about nvidia and Fedora. This thing has the built in:
Intel Corporation 4 Series Chipset Integrated Graphics Controller (rev 03)
Intel Q45/Q43(ELK)

I also have started with a fresh load of F39 on a similar machine and after the initial install did a normal update and it craps out after the reboot. I will try the remove command on the test box I stood up tonight.

Exact same issue with vanilla F39 installed on a old Thinkpad X200.

Similar issues (login gives dark screen or GUI crashes and sends you back to the login screen, but system still responds to ssh) were reported by several Fedora users on Reddit so this looks like a bug that affects multiple users. This is not the first time changes to current software inadvertently break things for people using older hardware – developers don’t normally use old hardware, and vendors may not support it with firmware updates.

Anyone experiencing the issue should check that the kernel is still alive by logging in from another system with ssh and then post hardware details (sudo inxi -Fzxx) or upload a [LHDB probe] (https://linux-hardware.org) (sudo -E hw-probe -all -upload) and post the link. journalctl should have some helpful details if you make the effort to find them among the mass of messages unrelated to your issue.

SSH works and the vnc desktop works (I know its different) everything except physical desktop.

The Link
It said X11 logs are not collected even though I did the command as you listed it.

And the inxi data (don’t judge - it used to work):
System:
Kernel: 6.6.6-200.fc39.x86_64 arch: x86_64 bits: 64 compiler: gcc v: 2.40-13.fc39
Console: pty pts/1 DM: GDM Distro: Fedora release 39 (Thirty Nine)
Machine:
Type: Desktop System: Hewlett-Packard product: HP Compaq dc7900 Ultra-Slim Desktop v: N/A
serial: Chassis: type: 15 serial:
Mobo: Hewlett-Packard model: 3033h serial: BIOS: Hewlett-Packard v: 786G1 v01.27
date: 10/22/2015
CPU:
Info: dual core model: Intel Core2 Duo E8400 bits: 64 type: MCP arch: Penryn rev: A cache:
L1: 128 KiB L2: 6 MiB
Speed (MHz): avg: 1998 min/max: 1998/3000 cores: 1: 1998 2: 1998 bogomips: 11970
Flags: ht lm nx pae sse sse2 sse3 sse4_1 ssse3
Graphics:
Device-1: Intel 4 Series Integrated Graphics vendor: Hewlett-Packard driver: i915 v: kernel
arch: Gen-5 ports: active: VGA-1 empty: DP-1,HDMI-A-1 bus-ID: 00:02.0 chip-ID: 8086:2e12
Display: server: X.org v: 1.20.14 with: Xwayland v: 23.2.2 compositor: gnome-shell driver:
gpu: i915 tty: 139x24
Monitor-1: VGA-1 model: Acer V193 res: 1280x1024 dpi: 86 diag: 490mm (19.3")
API: OpenGL Message: GL data unavailable in console for root.
Use of uninitialized value $val2 in string eq at /usr/bin/inxi line 7464.
Use of uninitialized value $val2 in split at /usr/bin/inxi line 7469.
Use of uninitialized value $val2 in concatenation (.) or string at /usr/bin/inxi line 7472.
Use of uninitialized value $val2 in concatenation (.) or string at /usr/bin/inxi line 7475.
Use of uninitialized value $val2 in concatenation (.) or string at /usr/bin/inxi line 7476.
API: EGL Message:
Audio:
Device-1: Intel 82801JD/DO HD Audio vendor: Hewlett-Packard driver: snd_hda_intel v: kernel
bus-ID: 00:1b.0 chip-ID: 8086:3a6e
API: ALSA v: k6.6.6-200.fc39.x86_64 status: kernel-api
Server-1: JACK v: 1.9.22 status: off
Server-2: PipeWire v: 1.0.0 status: n/a (root, process) with: 1: pipewire-pulse status: active
2: wireplumber status: active 3: pipewire-alsa type: plugin
Network:
Device-1: Intel 82567LM-3 Gigabit Network vendor: Hewlett-Packard driver: e1000e v: kernel
port: 1100 bus-ID: 00:19.0 chip-ID: 8086:10de
IF: enp0s25 state: up speed: 1000 Mbps duplex: full mac:
Drives:
Local Storage: total: 111.79 GiB used: 4.35 GiB (3.9%)
ID-1: /dev/sda vendor: Samsung model: SSD 840 EVO 120GB size: 111.79 GiB speed: 3.0 Gb/s
serial: temp: 29 C
Partition:
ID-1: / size: 110.79 GiB used: 4.09 GiB (3.7%) fs: btrfs dev: /dev/sda3
ID-2: /boot size: 973.4 MiB used: 271.2 MiB (27.9%) fs: ext4 dev: /dev/sda2
ID-3: /home size: 110.79 GiB used: 4.09 GiB (3.7%) fs: btrfs dev: /dev/sda3
Swap:
ID-1: swap-1 type: zram size: 3.72 GiB used: 0 KiB (0.0%) priority: 100 dev: /dev/zram0
Sensors:
System Temperatures: cpu: 43.0 C mobo: N/A
Fan Speeds (rpm): N/A
Info:
Processes: 229 Uptime: 2m Memory: total: 4 GiB available: 3.72 GiB used: 1 GiB (26.9%)
igpu: 32 MiB Init: systemd v: 254 target: graphical (5) default: graphical Compilers: N/A
Packages: pm: rpm pkgs: N/A note: see --rpm Shell: Sudo v: 1.9.14p3 running-in: pty pts/1
inxi: 3.3.31

Some news on my side: I updated my F39 installation this morning using the terminal (ctrl alt F3) on my old Thinkpad X200 and rebooted. I saw some mesa packages being updated and that gave me hope for a possible fix. I can now log on graphically to GNOME wayland and everything seems fine at first glance. I can also open applications like gnome-terminal and firefox without any issues. However, some applications such as nautilus and gnome-software still have some visual artifacts.

FWIW, I have not seen these issues on my latest Thinkpad (current daily driver).

Cool, same here. Things like file explorer (nautilius) are displayed in x-ray mode and can’t be used, but its a step in the right direction.

You may get useful data by comparing journalctl records for the latest boot with one that was looping to the login screen. This will, however, need work to find the right search terms.

You can make the inix output more readable by selecting the text and using the </> button, or byjust adding triple backquote (```) lines before and after the inxi output. Remember that forum posts are searchable and form a record that often helps others who encounter the same problem.

You aren’t using Xorg – Xwayland just translates Xorg to Wayland (and loses certain things in translation due to fundamental differences like cross-window connections – to can see the difference try usiing xeyes).

The latest update to Mesa this morning fixed all the remaining graphical artifact issues that had been plaguing my Thinkpad X200 setup with vanilla F39 for a few days.

Logged in to say the same thing. Its a beautiful thing!

1 Like