nvidia driver crash ubuntu 22

Asked by Aditya Mishra

after a driver update the system stopped recognizing nvidia gpu. it used to work fine with nvidia driver 510.

ProblemType: Bug
DistroRelease: Ubuntu 22.04
Package: nvidia-driver-515 515.48.07-0ubuntu0.22.04.2
ProcVersionSignature: Ubuntu 5.15.0-33.34-generic 5.15.30
Uname: Linux 5.15.0-33-generic x86_64
ApportVersion: 2.20.11-0ubuntu82.1
Architecture: amd64
CasperMD5CheckResult: pass
CurrentDesktop: ubuntu:GNOME
Date: Wed Jul 13 18:00:05 2022
InstallationDate: Installed on 2022-05-25 (49 days ago)
InstallationMedia: Ubuntu 22.04 LTS "Jammy Jellyfish" - Release amd64 (20220419)
SourcePackage: nvidia-graphics-drivers-515
UpgradeStatus: No upgrade log present (probably fresh install)

Question information

Language:
English Edit question
Status:
Solved
For:
Ubuntu nvidia-graphics-drivers-515 Edit question
Assignee:
No assignee Edit question
Solved by:
Aditya Mishra
Solved:
Last query:
Last reply:

This question was originally filed as bug #1981575.

Revision history for this message
Aaron Rainbolt (arraybolt3) said :
#1

Thank you for taking the time to report this issue and helping to make Ubuntu better. Examining the information you have given us, this does not appear to be a bug report so we are closing it and converting it to a question in the support tracker. We understand the difficulties you are facing, but it is better to raise problems you are having in the support tracker at https://answers.launchpad.net/ubuntu if you are uncertain if they are bugs. You can also find help with your problem in the support forum of your local Ubuntu community http://loco.ubuntu.com/ or asking at https://askubuntu.com or https://ubuntuforums.org. For help on reporting bugs, see https://help.ubuntu.com/community/ReportingBugs.

Also, if you'd like live chat support, you can ask for help on the #ubuntu IRC chatroom here: https://web.libera.chat/?nick=webchat?#ubuntu Or, if you prefer using an IRC client like HexChat, Quassel, WeeChat, or the like, you can connect to libera.chat and join the #ubuntu channel.

Revision history for this message
Manfred Hampl (m-hampl) said :
#2

"nvidia driver crash"
Crash in which sense? No display at all, wrong resolution, wrong colors ...?

"after a driver update"
Which update was that? Did it start automatically or did you manually initiate that update?

Your system seems to be missing the update to the newest version of the kernel (there is already 5.15.0-41 available)

Revision history for this message
Aditya Mishra (aditya-r-m) said (last edit ):
#3

Thank you for the response.

0. For more context, the problem happened in two of my systems simultaneously

Desktop : Nvidia GeForce RTX 3060 (LHR)
Laptop : Nvidia GeForce GTX 1650 (Dual GPU)

On my laptop I use extended screen.
Both dual booted with Windows 10.

1. Driver crash in the following sense:

a. System startup time increased from ~8 seconds to ~12 seconds.
b. nvidia-smi & docker-tensorflow-gpu stopped working due to "no running driver detected" error.
c. On my laptop the extended screen stopped working (not detected in settings->displays).

I assumed the above because of some defaulting to nouveau drivers?

2. The issue happened after a manual "sudo apt update && sudo apt upgrade". I tend to avoid using GUI software updater due to some other issues (sometimes "Additional Drivers" fails with empty error dialog).

3. I'm not sure how to upgrade Kernel, will it be through GUI? (full-upgrade shows nothing for me).

4. Current Resolution

Desktop : downloaded the following binary from Nvidia website & installed : NVIDIA-Linux-x86_64-515.57.run. Worked.

Laptop : The above installation failed with some error regarding running X server.
- The GPU in laptop Ubuntu is not that useful anyway for me & startup time is not that bad. So reverted completely to nouveau driver.
- This caused complete white screen on extended monitor connected through HDMI. I disabled nvidia card completely from an askubuntu answer (https://askubuntu.com/questions/757177/disable-nvidia-optimus-graphics-card).
- The extended monitor stopped working from laptop HDMI port, but USB-c port on my laptop worked.

Revision history for this message
Manfred Hampl (m-hampl) said :
#4

"how to upgrade Kernel"

The commands

sudo apt update && sudo apt full-upgrade

should load all updates, including the kernel packages. Maybe you did already install the new version, but did not yet reboot.

What is the output of

apt policy linux-image-generic

Revision history for this message
Aditya Mishra (aditya-r-m) said :
#5

I think that makes sense.

$ sudo apt policy linux-image-generic
linux-image-generic:
  Installed: (none)
  Candidate: 5.15.0.41.43
  Version table:
     5.15.0.41.43 500
        500 http://security.ubuntu.com/ubuntu jammy-security/main amd64 Packages
     5.15.0.25.27 500
        500 http://in.archive.ubuntu.com/ubuntu jammy/main amd64 Packages

Also, the current resolution works for me for the most part, so we can close this thread.

One minor question: "Additional Drivers" GUI shows nvidia 515 selected even though I installed it via binary & actual version probably differs (515.48.07ubuntu vs 515.57). Would you recommend waiting for Ubuntu recommended driver to catch up & doing "ubuntu-drivers autoinstall"? Or do you think the kernel upgrade would have solved the problem?

Thanks a lot for the support in any case though :)