Fedora 43 Issue; Kernel boot delay followed by kick back to login, crashes

Ask Fedora
, ,
1

I have been having a consistent issue where, despite making no changes update to update, that whenever my system takes longer than 3 seconds following login, I get booted back to that login screen within 2 minutes. When I re-enter my credentials, my system goes entirely non-responsive, and I cannot do anything to access the console. I am forced to hard-reset every time.

This issue seems to only abate when I choose the correct kernel to boot through, but that seems to change day by day. Otherwise, I have zero idea what needs doing to fix this.

I suspect it has to do with my motherboard and AMD processor, but despite my research, I can’t find any resources on how to fix these issues.

My system information is as follows. If more info is needed, I will provide as requested.

Operating System: Fedora Linux 43
KDE Plasma Version: 6.5.5
KDE Frameworks Version: 6.22.0
Qt Version: 6.10.1
Kernel Version: 6.18.4-200.fc43.x86_64 (64-bit)
Graphics Platform: Wayland
Processors: 16 × AMD Ryzen 7 9700X 8-Core Processor
Memory: 32 GiB of RAM (31.0 GiB usable)
Graphics Processor: NVIDIA GeForce RTX 5070 Ti
Manufacturer: ASUS

2

You show the 6.18.4 kernel, which has been updated 3 times to the now current 6.18.7 kernel. There have been many other software updates as well.

It is highly recommended that you perform a full update (if not already done) with sudo dnf upgrade --refresh and wait at least 5 minutes after the update completes before rebooting.

Have you installed the nvidia drivers from rpmfusion?

Installing the nvidia driver and/or performing a kernel upgrade requires the 5 minute wait to allow time for the driver to be locally compiled and installed to match the kernel.

5

I deleted the original response to this, because I had not followed the instructions as directed, and figured I would be better served assuming they were correct. I realized I had often not taken five minutes following updating to let the system compile, so I hoped that was the problem. After updating the kernel, I had no issues for the two days since I opened this thread. As I expected, the long-boot issue has returned.

RPM fusion’s drivers are correctly installed and up-to-date.

I do not understand why my system takes twenty to thirty seconds following login to boot, nor why this long wait ALWAYS causes me to be booted back to login before long.

EDIT: I realize that the kernel version I showed was out of date because, at the time, it was the only kernel version that allowed me in. For whatever reason, I get three options and the recovery mode to choose from when booting.

6

The default kernel options include rhgb quiet, which hides potentially important error messages. Pressing the escape key while booting may allow to see messages, or you can run sudo grubby --remove-args="rhgb quiet" --update-kernel=ALL. [https://www.baeldung.com/linux/grub-menu-management]. Read (https://www.baeldung.com/linux/grub-menu-management) for explanations and examples.

7

I think I’m having the same issue as you. I wanted to share what I learned. When it boots you back to the login screen, I am able to switch back to TTY2 (the Desktop Environment) without having to log in again, just with the ctrl+alt+f2 shortcut. Its not preferable granted, because a bunch of little things tend to be broken when I switch back, like bluetooth, but its something.

Edit: and doing it causes you to only shut down tty2 if you select shut down or reboot, you have to go back to tty1 to shut down

8

This helps.

I’ve read some other possible solutions, being that this problem likely results from NVIDIA dependencies for other programs failing to initialize properly. It may be as simple as constantly maintaining the system to be as up-to-date as possible, but seeing as this has also caused many issues for me in this exact manner, I remain unconvinced.

9

As an NVIDIA user I have the exact same issue and it has been popping up more often the last couple of weeks. Despite jumping around TTY to log back in, I found that doing so breaks Bluetooth connectivity so it’s not an ideal workaround. A complete reboot of the system has fixed it most of the time. I’d love to hear of a permanent fix.

10

Not all Nvidea users have this problem, so more details are needed to identify the affected hardware. Using a terminal, collect:

  1. Nvidia model and type of monitor connection from inxi -Gzxx,
  2. run journalctl —no-hostname —no-pager -b N -p 3 (where N is chosen to select a boot where the problem occurred) in a terminal and look for relevant entries from the journal. N can be omitted if you switch to a terminal after the graphics failure without rebooting. By adding |tee <filename> the output will be copied to the named file so you can edit out irrelevant entries and paste the key details into a forum post after restarting.

Since the journal contains massive levels of detail, some work may be needed to find a “filter” that selects the relevant details. Fedora has https://docs.fedoraproject.org/en-US/quick-docs/viewing-logs/ and there are other good introductions on the web hiding in a mass of AI clickbait nonsense.

11

Here’s the inxi output:

Graphics:
  Device-1: NVIDIA GA102 [GeForce RTX 3090] vendor: Micro-Star MSI
    driver: nvidia v: 580.119.02 arch: Ampere pcie: speed: 5 GT/s lanes: 16
    ports: active: DP-1,DP-2 empty: DP-3,HDMI-A-1 bus-ID: 2d:00.0
    chip-ID: 10de:2204
  Device-2: Logitech BRIO 4K Stream Edition
    driver: hid-generic,snd-usb-audio,usbhid,uvcvideo type: USB rev: 3.1
    speed: 5 Gb/s lanes: 1 bus-ID: 2-1:2 chip-ID: 046d:086b
  Display: wayland server: Xwayland v: 24.1.9 compositor: kwin_wayland
    driver: gpu: nv_platform,nvidia,nvidia-nvswitch d-rect: 4480x2520
    display-ID: 0
  Monitor-1: DP-1 pos: top-right model: LG (GoldStar) ULTRAGEAR
    res: 2560x1440 hz: 144 dpi: 109 diag: 685mm (27")
  Monitor-2: DP-2 pos: bottom-l model: Asus VG248 res: 1920x1080 hz: 144
    dpi: 92 diag: 609mm (24")
  API: EGL v: 1.5 platforms: device: 0 drv: nvidia device: 2 drv: swrast
    gbm: drv: nvidia surfaceless: drv: nvidia wayland: drv: nvidia x11:
    drv: nvidia inactive: device-1
  API: OpenGL v: 4.6.0 compat-v: 4.5 vendor: nvidia mesa v: 580.119.02
    glx-v: 1.4 direct-render: yes renderer: NVIDIA GeForce RTX 3090/PCIe/SSE2
    display-ID: :0.0
  API: Vulkan v: 1.4.328 surfaces: N/A device: 0 type: discrete-gpu
    driver: nvidia device-ID: 10de:2204 device: 1 type: cpu
    driver: mesa llvmpipe device-ID: 10005:0000
  Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
    de: kscreen-console,kscreen-doctor gpu: nvidia-settings,nvidia-smi
    wl: wayland-info x11: xdriinfo, xdpyinfo, xprop, xrandr

I haven’t managed to find a good timestamp for when I last experienced this issue, so I’ll try and remember to note down the hour and minute when it happens next time.