New Nvidia Driver GO6

English Install/Boot/Login
1

I did a zypper dup on my tumbleweed system to the latest snapshot (04032023) this morning and updated the nvidia drivers (to the lasted version uploaded on the 6/3) using the nvidia repo.

On restarting my system the mok registration screen was shown and I completed this (am used to this - as on another computer I have kde neon {based on ubuntu} and this iapproach has been used for a while). On rebooting I get a black screen. I restarted my system again and switched off secure boot in the uefi/bios settings and my system started fine. It also boots fine if I opt to use the previous kernel (6.2.0-1) rather than the current version 6.2.1-1.

I am using kde plasma with x11 on a Dell computer.

Have I missed a step?

Thanks very much.

2

Yes. This topic is widely discuseed on the mailing list and in the forum that starting with kernel 6.2 secure boot needs to be disabled (temporary) as the kernel recieved lock patches.

3

Thanks hui.

Apoligies I have just seen a posting on:

Advising that the new Nvidia drivers don’t work.

4

I think you new to OpenSuse.

Nvidia rpm driver works only when Nvidia-rpm build for exact version of kernel, and Nvidia rebuild their Suse rpm driver… like once per 2 week or once per month or even less…
So it safe to say - Nvidia rpm driver never works in reality.

As solution - you have to install Nvidia driver from .run file, it is simple after you do it once.
And reinstal/update Nvidia driver same way every time Kernel updates.

https://en.opensuse.org/SDB:NVIDIA_the_hard_way

Short:

  1. install kernel-devel/sources with g++ I dont remember full list, if driver installation fails read its log it will be obvious what you missing.
  2. Download Nvidia driver Official Drivers | NVIDIA must be .run file
  3. boot with nomodeset and 3 in kernel boot option - or press E on grub boot screen where you select boot options.
  4. install nvidia driver from terminal, by sh <nvidia_file_name>.run

if/when you need/want working CUDA on OpenSuse - download local CUDA rpm installation from Nvidia that few gigs size, and when installing - blacklist kernel module-driver in rpm installer, and install everything else - and CUDA works with latest Nvidia driver.

5

This is utter nonsense and not backed by facts.

1 Like
6

Wrong. NVIDIA does not rebuild anything, it does not build SUSE rpm at all. These packages are built by SUSE.

Stop spreading FUD. NVIDIA packages from SUSE most certainly work, what happens currently is change in openSUSE kernel configuration which is being discussed and hopefully will be fixed.

1 Like
7

this is my experience, at some point I got tired to see “black screens” every time I update kernel and Nviida driver does not update their kernel module
so I just instlaling Nvidia driver from .run file instead every time when kernel updated

8

Which will fail on the current Tumbleweed kernel under the same conditions as drivers from RPM.

9

what?
https://en.opensuse.org/SDB:NVIDIA_drivers

https://download.nvidia.com/opensuse/tumbleweed

Nvidia does not allow “redistribution” of their software, especially drivers.
They build everything by themselves.

10

Which is the exact reason why they are not distributed by SUSE.

11

im on latest opensuse kernel 6.2.1.1 with latest Nvidia linx driver 525.89

image
image1550Ă—565 129 KB

works

12

You did not even understand what this topic is about. Please read the linked ML and bug and try to comprehend the provided information. The issue is with secure boot and the new kernel patches which prevent loading of unsigned modules…

1 Like
13

secure boot - > exact kernel version → driver not build for exact kernel version ->??? logic???

SDB:NVIDIA_the_hard_way link above include instruction to make secure boot works

14

Read…and try to understand…

15

You can install the nvidia run file with the dkms (and install dkms as well) option to rebuild automatically on a kernel update. I don’t use as prefer to use the run file direct… Nor do I run secure boot on this desktop system.

 inxi -SGxxz
System:
  Kernel: 6.2.1-1-default arch: x86_64 bits: 64 compiler: gcc v: 12.2.1
    Desktop: GNOME v: 43.3 tk: GTK v: 3.24.35 wm: gnome-shell dm: GDM
    Distro: openSUSE Tumbleweed 20230307
Graphics:
  Device-1: NVIDIA TU117GLM [Quadro T400 Mobile] driver: nvidia v: 530.30.02
    arch: Turing pcie: speed: 2.5 GT/s lanes: 16 ports: active: none
    off: DP-1,DP-2,DP-3 empty: none bus-ID: 02:00.0 chip-ID: 10de:1fb2
  Device-2: NVIDIA GP104GL [Tesla P4] driver: nvidia v: 530.30.02 pcie:
    speed: 2.5 GT/s lanes: 16 bus-ID: 03:00.0 chip-ID: 10de:1bb3
  Display: x11 server: X.Org v: 21.1.7 with: Xwayland v: 22.1.8
    compositor: gnome-shell driver: X: loaded: nvidia
    gpu: nvidia,nvidia-nvswitch display-ID: :0 screens: 1
  Screen-1: 0 s-res: 3840x2160 s-dpi: 96
  Monitor-1: DP-1 pos: top-left res: 1920x1080 dpi: 94 diag: 598mm (23.53")
  Monitor-2: DP-3 pos: primary,bottom-c res: 1920x1080 dpi: 94
    diag: 598mm (23.53")
  Monitor-3: DP-5 pos: top-right res: 1920x1080 dpi: 94 diag: 598mm (23.53")
  Monitor-4: DVI-D-1-0 size-res: N/A
  API: OpenGL v: 4.6.0 NVIDIA 530.30.02 renderer: NVIDIA T400/PCIe/SSE2
    direct-render: Yes

There are other issues related to the lockdown patches, not just the Nvidia driver…

16

This issue has now been resolved with snapshot 10032023 and the system boots fine now (with the nvidia driver installed & secure boot switched on).

1 Like