CachyOS Kernel 6.13 / Nvidia 570 Driver issues (ISO REL: 241221)

Issue:
After upgrading the system Plex is unable to play video. You can hear the audio but the screen is either black or you can only see small parts of the video.
Example:

Spectacle causes the screen to blank in and out making it impossible to use. if HDR is turned on it seems to stop that however then you have artifacts randomly popping up around the screen rapidly.
I spent part of an evening and all Saturday trying to figure this out. I ended up wiping the drives to go with a fresh install with the latest ISO and even that did not work because it loaded the new drivers and kernel. in the end, the only way that seemed to work to resolve is to downgrade and avoid Kernel 6.13 and Nvidia 570 drivers.

Steps taken:
Load the Kernel manager and select version 6.12.10-2 LongTerm and execute
Reboot and load with the longterm option when the list comes up
Then open the terminal and run:
sudo downgrade nvidia-utils lib32-nvidia-utils opencl-nvidia nvidia-dkms lib32-opencl-nvidia
Pick the 565.77 drivers when asked
If I recall at the end it will ask if you want to ignore them when its done
reboot and test.

I am new to running Linux full time having ditched Winblows 11 around 2 months ago. So this was a bit of an adventure. I hope someone else finds this helpful.

System Specs
MB: Asus Crosshair VIII WIFI
CPU: AMD Ryzen 9 5950X 16-Core Processor
RAM: Gskill 32GB
GPU: RTX 4070 TI
PSU: Corsair 1000w

Hmm strange. I’m also using Plex, both the HTPC and normal one. I assume you use normal one. I don’t have those issues running latest kernel and latest Nvidia drivers. Maybe it’s due to a configuration or because of another package. I’m not using dkms however. Also, did you update to “org.freedesktop.Platfrom.GL.nvidia-570-86-16”? As far as I know, that needs to match the driver you’re currently running.

Howdy. I am honestly not sure what the root cause could be. I just know that when it bumped my kernel up and dropped the 570’s on everything blew up lol. So right now I am locked on kernel 6.12 LTS and driver version 565.77. I have zero plans to apply any updates until I get a chance to figure out how snapper works. That way if something blows up I can revert and not spend a while day and a half trying to fix things. And as for plex, yea just the normal client via flatpak.

You mean the Btrfs Assistant Snapper I guess. It’s fairly simple. In the CachyOS Hello app, go to Apps/Tweaks and click on the “Install Snapper support” if you haven’t done that yet. That will create a root config for Btrfs Assistant. Then open that app, go to Snapper and on the “Browse/Restore” tap you can restore back to a restore point you want. But if you haven’t installed the Snapper support yet, I guess you’ll see an empty list there for now until you update packages. You can also create manual restore points on the “New/Delete” tab so, maybe start with that. You can even try it out of course. Restore one and boot immediately. Then, you will most likely notice that in case you want to update using pacman, that the DB is locked. You’ll have to remove the lock first before you can continue updating. You can do this with:

sudo rm /var/lib/pacman/db.lck

Yea was referring to the BTRFS Assistant. Not sure what is supposed to be empty, are you referring to the subvolume tab where I see my subvolumes? Those are there if so. I just have not tinkered with it yet to get a feel for how it works.

As far as I remember, if you haven’t installed the Snapper support yet, you won’t have any restore points / snapshots or in other words, your system does not automatically create them. So, the list I’m referring to is this one where you could pick one to do a restore:

@uKsu A quick question about Snapper while on the subject. I have the following layout:

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
zram0 254:0 0 31.1G 0 disk [SWAP]
nvme1n1 259:0 0 1.8T 0 disk
├─nvme1n1p1 259:1 0 2G 0 part /boot
└─nvme1n1p2 259:2 0 1.8T 0 part (all of these are subvolumes)
/var/tmp
/var/log
/var/cache
/root
/srv
/
nvme0n1 259:3 0 3.6T 0 disk
└─nvme0n1p1 259:4 0 3.6T 0 part /home

My /home, on the second drive, is not a subvolume. Will snapper pick that up as well? Looking at one of the snapshots I saw /home listed but it shows no files.

Do I maybe need to add a /@home subvolume on the second drive and have that mounted vs how it is now?

Ah okay. Yea I do not have that. I have literally not run any updates or tweaks since installing from the ISO because I was trying to avoid the problems I encountered. Thanks for showing me what you referred to.

No, generally /home is not part of a snapshot and I think for a good reason. Windows does it the same way. It only affects the installation of the OS and other programs while leaving files such as documents alone.

If you want to really include /home too, I would recommend doing backups instead. A restore point or snapshot is not a backup solution. So, instead of using Btrfs Assistant, use an actual backup software.

Then you can just install the Snapper support and see how it creates those snapshots before and after updating your system or create them manually too. Then restores are very straight forward, just hit the restore button while having the snapshot selected and reboot afterwards.

You’re welcome.

1 Like

Sounds good thanks

Which exactly is causing the issue? It seems unlikely that it’s a combination of both.