tbh this problem sounds a lot like what I’m experiencing, except that I’m running MUCH older hardware, and definitely running an AMD graphics card (I think back even when they still used to call themselves ATI lol). Most times I can barely get 1-10 minutes of use in before it just… locks up, and then I have to power cycle over and over and over again - sometimes I get a day or so, but eventually that locks up, sticking me back into the loop of powercycling.
I have a journalctl from today, but journalctl --list-boots does not appear to reflect honestly - journalctl -k -b -1 appears to pull the boot that journalctl --list-boots lists as -3, rather than -1.
In any case, I have, er, a boot from today (20 July, in my neck of the woods), kernel logs, saved to a text file.
I moved your post here since your GPU is AMD and the other thread is about an NVIDIA GPU.
You said you have older hardware so we need to know what it is.
Please install inxi if needed then post the output of inxi -Fzx here inside the </> Preformatted text tags available on the tool bar above.
How is it locking?
mouse and keyboard do not respond
running app just halts
With 1 you may be able to still connect with a ping or ssh into the system. That usually indicates it may be graphics related.
With 2 you still may be able to connect remotely, and kill the running app
With 3 the system is totally stuck with no way to connect and will likely only respond to a full hard power off.
Please give us the info requested and we can dig further.
Happened again. I tried booting from a different kernel (specifically, kernel-5.18.7-200.fc36.x86_64 vs kernel-5.18.11-200.fc36.x86_64), and samey samey. Froze right up. I do have a journalctl output from an earlier boot (using the 5.18.11-200 kernel).
Switching to a different TTY did not work - Ctrl + Alt + F3 (or indeed, any F-key) was not responsive.
I have been dealing with a similar problem on F36 with AMD integrated graphics. It has also lead to crashes… throwing me out to the login screen. I am lucky to have not lost any work
The types of freezes have mostly been display, Sound continues to play and it seems inputs are taken.
I have not tried to ssh as I don’t have anything to ssh with but at least to my eyes it feels like display.
Then there are crashes where I am thrown out to the login screen. I think the shell had crashed but I do not know how to check that/
Have you tried logging in with xorg instead of wayland to see if that gives a different response. If it is truly just graphics (mouse + keyboard) that are not responding then it may be nothing more than a wayland vs xorg issue.
That can be done by selecting ‘gnome with xorg’ with the gear icon in the lower right corner when entering your password.
A cursory look and it seems it still persists on xorg, I am getting worse freezes though on xorg my mouse is free to move around (though the rest of the shell and apps are unresponsive) in this instance wayland actually is better as the freezes are usually shorter.
I can do some more testing if need be (though there may be a gap in my response time)
that DOES look like the issue I’m having, except that one of his conditions is explicitly “The system must have been suspended (put to sleep) once, then resumed” - and I have DEFINITELY gotten severe system lockups anywhere from one to ten minutes after a complete reboot (Press and hold the power button until the system shuts off, wait five to ten seconds, press the power button again to turn the system back on).
However, the error messages in his logs are incontrovertibly similar to the ones I see in mine…
I have faced a very similar issue with Ryzen 5 + integrated radeon vega 8 . It is generally worse on Xorg than Wayland. I found an workaround that prevents the random lock-ups. Install Corectrl and set cpu and gpu governor to powersave.
I have not experienced anything in this regard, and when I have its usually from something else. I am just thankful that it has seemed to have gone away and the risk of me losing work has gone down drastically.
I actually had plenty of freezes between my last comment and this one, even after multiple kernel updates - so I decided to give this a try after one of my journalctl -k -b -1 commands FINALLY revealed that the GPU was seizing up, so I tried that command… but it does not appear to have taken. I just rebooted and it shows the radeon driver is still loaded, even after doing that little grubby command.
Both per that command, as well as inxi -Fzx. I tried running the grubby command again, I’ll try another reboot and I’ll check my grub kernel options before I pick one to see if I see those options in the line.