2

My desktop computer hangs after installing kernel 5.19.16. Booting into 5.19.15 works fine.

It freezes when I get to enter the password to unlock the encrypted hard drive (LUKS), which is the first thing that happens after selecting the kernel in GRUB.

I have no log in /var/log/boot*, because we never get to where machine can write to this file (encrypted).

The machine specs:

CPU: AMD Ryzen 9 7950x MB: ASUS ROG Strix x670E-E GPU: Rtx 4090 (Gigabytes)

The same kernel boots fine on my frame.work laptop (intel core i7 11th gen)

UPDATE: I have the same problem with Fedora 37 live, booting from usb: it starts correctly when monitor is connected to the integrated AMD graphics, but not when connected to Nvidia.

Alireza
  • 251

3 Answers3

0

I updated today and my computer only stays on the start up screen of ASUS after choosing 5.19.16.201 in the boot menu. Choosing 5.19.15.201 instead works like before.

0

Well, I had the same thing happen to me on 5.19.16.201 and then on 6.05 too when I updated yesterday. I smelt a rat here and did some digging and found this to be the solution!

https://fedoraproject.org/wiki/GRUB_2#Reinstalling_GRUB

In my case at least, the grub config seems to have been broken somehow and these steps fixed it instantly. This can be caused by editing the wrong config files for grub / overwriting the right ones etc.

This thread illustrates the mistakes one might make to get to such behaviour.

https://ask.fedoraproject.org/t/f36-boot-stick-at-loading-linux-version-after-every-kernel-update/26383

(In a nutshell, remove the config files mentioned in the first link and reinstall all grub-related packages.)

P.S: Make sure GRUB_ENABLE_BLSCFG=false is set in your grub config or else the grub menu entries/text acts weird / disappears.

0

Ok, I was able to resolve this differently than other proposed answers which did not work for me.

In my case, the culprit was Linux Firmware Minimization which removed the collective gpu firmware but never installed amd-gpu-firmware. This bug claims they backported a fix to FC35 and FC36, but that obviously didn't work for me.

Following commands resolved the issue and I was able to boot into latest kernel with no issues:

$ sudo dnf install amd-gpu-firmware
$ sudo dnf reinstall kernel*

Note: I have an AMD gpu so I installed the amd gpu firmware, but OP might be able to resolve the issue they're having by explicitly installing nvidia-gpu-firmware & a kernel reinstall.

Thanks to Kai from comment 18 on above-linked bug report for giving me the necessary info to fix my install.