Boot failure: Plasma login service fails & core dumps

Hello. I’m yet another user of a legacy NVIDIA card (Zotac 1080) that’s having boot problems. For the moment, all I can provide are pictures of the journal, so please bear with me.

This is the second time I’m having this problem. The first time, I attempted to upgrade from Fedora 42 to 43, and all sorts of havoc broke loose, so I just clean reinstalled Fedora 44 from a live USB. Things start off okay; I do a few post-installation steps, get the proper 580 drivers. And now, after a whole 16 hours of the system working, it magically doesn’t. (Yes, it was most certainly something I did, but I couldn’t even begin to nail down what it is specifically on my own.)

Here are the pictures of the very blatant and obvious problem area.

I’m not about to do a clean reinstall for a second time, so any advice would be greatly appreciated. However, there’s no critical personal files on the drive, so you’re welcome to provide any and all troubleshooting steps I may need to take, without fear of personal data loss.

When you said the proper nvidia 580 drivers are installed were you meaning the 580xx version from the rpmfusion repo?
If so which repo and which specific version?
dnf list --installed \*nvidia\* will answer that question.
The akmod-nvidia-580xx package exists in 3 different repos for f44. Installation from only the first and last listed repos below are usable

$ dnf list --available akmod-nvidia-580xx  --repo rpmfusion-nonfree --refresh
Updating and loading repositories:
 RPM Fusion for Fedora 44 - Nonfree                                                                                                   100% |   8.1 KiB/s |   5.7 KiB |  00m01s
Repositories loaded.
Available packages (available for reinstall, available for upgrade)
akmod-nvidia-580xx.x86_64 3:580.142-2.fc44 rpmfusion-nonfree


$ dnf list --available akmod-nvidia-580xx  --repo rpmfusion-nonfree-updates --refresh
Updating and loading repositories:
 RPM Fusion for Fedora 44 - Nonfree - Updates                                                                                         100% |   8.7 KiB/s |   5.6 KiB |  00m01s
Repositories loaded.
Available packages (available for reinstall, available for upgrade)
akmod-nvidia-580xx.x86_64 3:580.159.04-1.fc44 rpmfusion-nonfree-updates


$ dnf list --available akmod-nvidia-580xx  --repo rpmfusion-nonfree-updates-testing --refresh
Updating and loading repositories:
 RPM Fusion for Fedora 44 - Nonfree - Test Updates                                                                                    100% |  40.5 KiB/s |   6.1 KiB |  00m00s
Repositories loaded.
Available packages (available for reinstall, available for upgrade)
akmod-nvidia-580xx.x86_64 3:580.173.02-1.fc44 rpmfusion-nonfree-updates-testing

Also I see that you may have plasma 6.5.0. There have been updates to plasma, and the most recent version is 6.7.1 which may solve the issues.

Please run sudo dnf upgrade --refresh from that terminal so you have the latest software then reboot and potentially your problem may have already been fixed.

First reboot after doing so failed. Second succeeded, but I have no idea if this is a guarantee. Below were the errors on the first reboot attempt. Why this didn’t happen the second time is beyond me, but I’m going to make absolutely certain it’s not happening anymore once I’ve had the chance to do a few other things.




Checking for updates no longer works, and is stuck here.

https://discussion.fedoraproject.org/t/discover-reports-updates-but-is-stuck-at-fetching-updates

Discover is broken

After updating from the command line, I rebooted again, and it seems to be okay. However it still concerns me that the first reboot immediately following the initial upgrade still failed. I’ve marked this as solved, but will return to this thread if this crops up again.

I would imagine that the updated version of the 580xx drivers had not fully rebuilt and installed when you rebooted, they were built during the first boot so the second boot worked. The messages in yellow say that you had a version mismatch which then went away, and that only happens with a successful compilation.

You could check to see if you have any *.failed files in /var/cache/akmods/nvidia for the times of the reboots if you want to confirm either way.

Are you now using the latest version of the nvidia driver?
Please post the output of modinfo -l nvidia && modinfo -F version nvidia

The latest version currently available is

$ dnf list --available akmod-nvidia-580xx --repo rpmfusion-nonfree-updates-testing
Updating and loading repositories:
Repositories loaded.
Available packages (available for reinstall, available for upgrade)
akmod-nvidia-580xx.x86_64 3:580.173.02-1.fc44 rpmfusion-nonfree-updates-testing