Nvidia390xx Install fails

Support Graphics & Display
1

Hi all, i tried to install the Nvidia proprietary driver, but my system does not boot after that.
Specs: i3 3570k, AsRock Z77 Pro3, GeForce GTS450, Kernel 5.4 LTS, System up-to-date

I Installed with “sudo mhwd -i pci video-nvidia-390xx” and the Installation finished without Errors.
After a reboot, the System hangs after the Grub Screen.
Journalctl showed this running in a loop until getting reset by switch:

Sep 14 01:50:32 crx-wztv kernel: resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000d3fff window]
Sep 14 01:50:32 crx-wztv kernel: caller _nv001015rm+0x1bf/0x1f0 [nvidia] mapping multiple BARs
Sep 14 01:50:33 crx-wztv kernel: NVRM: GPU at PCI:0000:01:00: GPU-5b33e47b-f31e-d8b8-0c46-ecfa280f1297
Sep 14 01:50:33 crx-wztv kernel: NVRM: Xid (PCI:0000:01:00): 79, GPU has fallen off the bus.
Sep 14 01:50:33 crx-wztv kernel: NVRM: GPU at 0000:01:00.0 has fallen off the bus.
Sep 14 01:50:33 crx-wztv kernel: NVRM: A GPU crash dump has been created. If possible, please run
                                 NVRM: nvidia-bug-report.sh as root to collect this data before
                                 NVRM: the NVIDIA kernel module is unloaded.


Sep 14 01:50:37 crx-wztv kernel: NVRM: RmInitAdapter failed! (0x12:0x45:1920)
Sep 14 01:50:37 crx-wztv kernel: NVRM: rm_init_adapter failed for device bearing minor number 0
Sep 14 01:50:37 crx-wztv kernel: resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000d3fff window]
Sep 14 01:50:37 crx-wztv kernel: caller _nv001015rm+0x1bf/0x1f0 [nvidia] mapping multiple BARs
Sep 14 01:50:37 crx-wztv kernel: NVRM: GPU at PCI:0000:01:00: GPU-5b33e47b-f31e-d8b8-0c46-ecfa280f1297
Sep 14 01:50:37 crx-wztv kernel: NVRM: Xid (PCI:0000:01:00): 79, GPU has fallen off the bus.
Sep 14 01:50:37 crx-wztv kernel: NVRM: GPU at 0000:01:00.0 has fallen off the bus.
Sep 14 01:50:37 crx-wztv kernel: NVRM: A GPU crash dump has been created. If possible, please run
                                 NVRM: nvidia-bug-report.sh as root to collect this data before
                                 NVRM: the NVIDIA kernel module is unloaded.

By uninstalling the nvidia drivers with “mhwd -r pci video-nvidia-390xx” and rebooting, the System becomes usable again, but the nouveau drivers are pretty bad :frowning_face:
Should i try to install the Nvidia drivers provided on the nvidia Website?

2

Hello,

Your GPU has support with the 450xx drivers.
Search for GTS 450 here
https://download.nvidia.com/XFree86/Linux-x86_64/450.66/README/supportedchips.html

First, from TTY identify what drivers you have installed so far:
mhwd -li
Remove them and install:
sudo mhwd -i pci video-nvidia-450xx

If you still get a crash, add to /etc/X11/mhwd.d/nvidia.conf at Section "Device" this line:
Option "ConnectToAcpid" "Off"

Then run this command:
sudo mhwd-gpu --setmod nvidia --setxorg /etc/X11/mhwd.d/nvidia.conf

Reboot after that.

3

When searching for GTS 450, the first result is under this section:

Below are the legacy GPUs that are no longer supported in the unified driver. These GPUs will continue to be maintained through the special legacy NVIDIA GPU driver releases.

The 390.xx driver supports the following set of GPUs:

So i think i need to have the 390xx driver, and not the 450xx

1 Like
4

Correct! I missed that. Sign i was and still am tired.
Keep the one you have installed and add
Option "ConnectToAcpid" "Off"
to Section "Device" inside /etc/X11/mhwd.d/nvidia.conf and see if things are better.

5

I editet the nvidia.conf file according to your directions and did a reboot after that, but the error remains the same.
Freezes after Grub Screen and the “GPU has fallen off the bus” message in journalctl in a loop
Uninstalled the 390xx driver again and using nouveau to boot rn.
Graphic Card is working under Windows 10 and did also work under Devuan Beowulf.