* zoom in on the error message
Yep, of course it’s Nvidia.
Every time I have seen a funky black screen with text against my will Nvidia was involved.
Nice! That looks like a fun one.
Ikr. So fun
if your updates included a new kernel try installing the kernel headers for the new kernel.
then if it’s still not working reinstall the nvidia driver.
i used to daily drive debian with proprietary nvidia drivers and it would break with every kernel upgrade
from memory, so almost certainly incorrect, the fix was simply something like
sudo apt install linux-headers-`uname -r`Are you running a proprietary video driver? It might be worthwhile to disable it in case it became incompatible perhaps after a kernel upgrade.
Did you perform a graphical login prior?
Yes to both. What should i do
First, you might try booting an older kernel to see if that runs for you. Your bootloader such as grub might help you pick an old one.
The older kernels are actually combinations of kernel + initial ramdisk that contains the version of your graphics drivers that were being used at that time. It could be a way to test the hypothesis.
Same error on older kernel
Hmm, interesting. That tells us that it’s not actually a problem with your graphics driver or kernel version, and given that it was working on this version before, I would think some aspect of Xorg configuration, your graphics hardware has an issue, or your installation in general has been corrupted when it tried to upgrade.
You might try to detect corruption by using a tool like
debsumsto check for any obviously corrupted files.What’s the state of your debian packages I wonder… does something like
apt-get updateorapt-get checkhighlight any problems with the state of installed packages that could point to a failed upgrade?I think its because / is full. Some packages cant update. Is there a way to combine them without gui as i am disabled and cant use a mouse? I know u cant edit partitions booted
A full root will absolutely kill your system.
You have unlocked a new achievement: the software hoarder!
If you can boot into terminal session, e.g. by pressing Ctrl + Alt + F2, you can try:
sudo apt clean sudo apt -f install sudo apt clean sudo apt dist-upgrade sudo apt cleanIf sudo apt -f install doesn’t work properly, you can create an apt-cache folder on, e.g. your home partition, assuming this is the one with sufficient amounts of free storage.
sudo apt clean sudo mkdir /home/apt-cache sudo nano /etc/fstabIn the fstab you specify where this directory shall be mounted:
/home/apt-cache /var/cache/apt/archives none bind 0 0Now you copy the files in place and mount the partition:
sudo cp -r /var/cache/apt/archives/* /home/apt-cache sudo mount -aNou you should be able to run the fix-installation and update commands without the errors:
sudo apt -f install sudo apt dist-upgradeI booted into i3!
This may seem like an obvious question, but are there files you can remove or perhaps move to another drive or USB stick temporarily to make enough space to get through your updates? You should be able to do those while rootfs is full.
We can certainly delete or copy files using the terminal.
Are you sure the root is full and not readonly due to other errors? Why do you believe root is full?
duf / lists 0gb as available. Idk what i should move
What should i do
Avoid Nvidia like the plague.
Wait months to buy a amd gpu as im on disablity? Got it
It’s good to see that the same problems from Knoppix in 1998 still persist into 2024.
It’s become my standard procedure to do a full backup before a major version upgrade of Linux nowadays as a result
Xserver has failed to start.
Gotta love xorg
Knoppix
now there’s a name I haven’t heard in ages…
Looks like it wants to remove the cause but not the symptom.
I’m gonna be honest: I don’t know what you’re trying to say.
Rocky Horror
Ohhhhhhh
Good movie. Thanks!
I’m not knocking it, but I feel like you really wanted to use this image somewhere. 😁👍
Look on top of op’s monitor
Oh snap!
Haha
Yes thats a toy
Looks like Eddie Izzard
Looks like Tim Curry.
Did you solve it? Recently there was a problem with graphics thing and downgrading mess from 1.24 to 1.23 helped me. It was in arch with AMD graphics, but some people said Nvidia ones also had the problem.
Yes. Apt-cache is mounted on /home which allowed me to update fully
It’s a known and fixed problem. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1062932
The last post in that thread was a week ago, if OP just updated then this problem shouldn’t have happened.
I expect that they’re running into this instead: https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1955941.html
That was on the kernel 6.1.0-18, I had it too, fixed several days ago, but in OP picture the kernel is 6.1.0-17, that one wasn’t affected.
I would be looking for ways to revert that update. Either using a pre-existing timeshift shapshot or maybe apt’s built in reversion capability. (Which I’m not familiar with, sorry.) Hopefully someone with real skills will chime in, good luck.
You may want to go back to driver 525… im not running debian at the moment, nor do i use mvidia graphics, but this page seems to say that 525 is the newest supported version as of feb 2
You bought an Nvidia graphics card.
I know, shame on me. Its like i wasnt on linux when i bought this computer
Yeah. Why would anyone expect one of the most popular video cards in the world to work in Linux. Those idiots.
This would implies the package management system to know the inner working of nvidia drivers so it can warn users when updates would break it. How one’s implement it?
The distro should be testing the driver within their update system before including it in their repo. The package should include metadata about system settings that could potentially break the system.
Distro makers have test machines to test their packages before pushing them out, but they can only cover popular and common hardware. Expecting them to test it on all graphic card permutations is too much considering most distros are released for free and maintained by volunteers, especially since nvidia cards aren’t cheap anymore.
isn’t acceptable
Have you contacted whoever you pay for your support contract?
As long as you keep relying on this cliche excuse, Linux is never going to be treated as a serious desktop operating system.
Distros are trying to create usable, friendly systems. They failed at that if their distros are this fragile. That’s what “unacceptable” means in this context. You can’t just throw your middle finger up in the air at the user when their system fails by saying “you didn’t pay for it” and scurry off giggling. Yet Linux advocates keep pushing Linux on inexperienced users, saying that it’s the solution to everything; that it’s so easy their grandma uses it.
I don’t use Linux as anything more than a toy for this very reason. I’ll start taking it seriously when its advocates do.
Linux is never going to be treated as a serious desktop operating system
Oh noes!
You can’t just throw your middle finger up in the air at the user when their system fails by saying “you didn’t pay for it” and scurry off giggling.
I disagree.
saying that it’s the solution to everything
LOL
… And your reply shows exactly why no one should take anything you say seriously. You’re just trolling.
Pretty much sums up every thing
MommysBabygiwl ?? On the third line. Is this canon to debian lore or did OP change it to that?
that’s the hostname
I had same problem on Arch based distros, it’s Nvidia problem. Try booting with LTS kernel.
That’s the novideo graphics card for you
libglxserver_nvidia crashed, either because of a bug or a hardware issue.
If you haven’t already, try to update your Nvidia drivers. Assuming you used the APT repository, this can be done by hitting ctrl+alt+f3 and logging in through the command line. Then you van apt update+apt upgrade and see if the Nvidia driver gets updated.
If it doesn’t, there’s this thread: https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1956817.html
It suggests that installing nvidia-vulkan-icd fixes a crash very similar to yours. You may also need to install libnvidia-glvkspirv and nvidia-vulkan-common manually, though I think those are installed automatically alongside the driver update.
If this fixes the issue, it appears Debian didn’t add the right dependencies to their packages.
