nvidia-driver-515 (515.76-0ubuntu0.22.04.1): blank screen after boot, can't log in

[fprietog]

System: Ubuntu 22.04.1 LTS (x86-64)
Kernel: 5.15.0-50-generic
NVIDIA card: NVIDIA Corporation GA102 [GeForce RTX 3080 Lite Hash Rate] (rev a1)

Problem summary:
Just updated "nvidia-driver-515" package from "515.65.01-0ubuntu0.22.04.1" to newest "515.76-0ubuntu0.22.04.1". It results in a blank screen after boot and before showing graphical login (gdm) so can't log in.
Seems that keyboard doesn't work in this situation so I can't change to a text session. Instead, I used a remote ssh session to get some info: journalctl.txt, dmesg.txt, lspci.txt and nvidia-smi.txt files attached.
Card is connected to monitor using HDMI output.

Notes:
- Previous "nvidia-driver-515" version, "515.65.01-0ubuntu0.22.04.1", worked as expected. Unfortunately I can't install that version again because it seems to be deleted from the repository.
- Previous "nvidia-driver-510" version, "510.85.02-0ubuntu0.22.04.1", works as expected.

[fprietog (fprietog)]

I think these journalctl.txt lines shows the problem:

oct 15 23:39:55 fpgmsi kernel: NVRM: GPU at PCI:0000:01:00: GPU-c1ee789d-58fe-8ea8-1720-dc7a7f9ee100
oct 15 23:39:55 fpgmsi kernel: NVRM: Xid (PCI:0000:01:00): 62, pid='<unknown>', name=<unknown>, 0000(0000) 00000000 00000000

In this page (https://docs.nvidia.com/deploy/xid-errors/index.html#topic_4) it shows that xid-error 62 is: "Failure: Internal micro-controller halt (newer drivers)"

[fprietog (fprietog)]

This bug is probably fixed in newer driver version 520.56.06:

* Fixed a regression in 515.76 that caused blank screens and hangs when starting an X server on RTX 30 series GPUs in some configurations where the boot display is connected via HDMI.

https://www.nvidia.com/Download/driverResults.aspx/193764/en-us/

Hope it'll be updated in Ubuntu repos soon to ensure smooth upgrading to Kinetic...

[fprietog (fprietog)]

The packages "nvidia-graphics-drivers-520 - 520.56.06-0ubuntu0.22.04.1" from PPA "https://launchpad.net/~oem-solutions-group/+archive/ubuntu/nvidia-driver-staging" solves this bug (tested).

[Simon Chopin (schopin)]

This also affects Kinetic systems (well, at least mine :) )

[Ubuntu QA Website (ubuntuqa)]

This bug has been reported on the Ubuntu ISO testing tracker.

A list of all reports related to this bug can be found here:
https://iso.qa.ubuntu.com/qatracker/reports/bugs/1993041

[fprietog (fprietog)]

@schopin
What's your graphics card model? Theoretically this bug happens only with RTX 30 Series GPU connected via HDMI. If it's connected using DP it should not happen.
Also you can install the 520.56 packages for Kinetic from this PPA:
https://launchpad.net/~oem-solutions-group/+archive/ubuntu/nvidia-driver-staging
That PPA is maintained by the same person that maintain the Ubuntu packages so it's totally trustable.

[Rik Mills (rikmills)]

FYI, upstream reports where NVidia confirm the issue:

https://forums.developer.nvidia.com/t/bug-report-black-x11-screen-and-partial-lockup-when-upgraded-to-515-76-and-dual-rtx3060/228912/23

https://forums.developer.nvidia.com/t/515-76-nvidia-drivers/229132/14

[Launchpad Janitor (janitor)]

This bug was fixed in the package nvidia-graphics-drivers-515 - 515.76+really.515.65.01-0ubuntu1

---------------
nvidia-graphics-drivers-515 (515.76+really.515.65.01-0ubuntu1) kinetic; urgency=medium

  * Revert to the 515.65.01 version (LP: #1993041).
  * Revert "debian/templates/dkms_nvidia.conf.in: drop buildfix_kernel_6.0.patch"
  * debian/rules:
    Make sure that debian_version variable includes the full version, so that
    we depend on the correct libraries in case of release rollbacks.

 -- Alberto Milone <email address hidden> Wed, 19 Oct 2022 16:18:24 +0000

[Jonas Gamao (yamiyukisenpai)]

Will 22.04 get 520?

[fprietog (fprietog)]

@yamiyukisenpai: it's a matter of time but you have it in the proposed repos so you can easily install it now:

https://launchpad.net/ubuntu/+source/nvidia-graphics-drivers-520

[Andy Whitcroft (apw)]

The verification of the Stable Release Update for nvidia-graphics-drivers-515 has completed successfully and the package is now being released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regressions.