How to move to Nvidia 470xx drivers with Kepler GPU?

Hello, I’ve been putting off updating my system for a while now since Nvidia dropped its support of its Kepler-based GPUs as doing sudo pacman -Syu would update my system with the linux-514-495.44-5 package which is incompatible with my GPU as it only allowed me to use the tty after a reboot when I was on unstable. I switched to testing as a temporary measure but the update has been available for some time now.

Here is the output of inxi -G:

Graphics:  Device-1: NVIDIA GK106 [GeForce GTX 650 Ti Boost] driver: nvidia v: 470.82.00
           Display: x11 server: X.Org 1.20.13 driver: loaded: nvidia resolution: 1: 1600x1200~85Hz 2: 1680x1050~60Hz
           3: 1024x1280~60Hz
           OpenGL: renderer: NVIDIA GeForce GTX 650 Ti BOOST/PCIe/SSE2 v: 4.6.0 NVIDIA 470.82.00

I’m looking to move to linux514-nvidia-470xx or linux515-nvidia-470xx if I need a kernel upgrade without causing any problems. The option for video-nvidia-470xx is not available in the settings menu under Hardware Configuration so I’m unable to do it that way. I was told that nouveau was an option although I feel like it doesn’t fit my needs.

The Nvidia related packages I have atm that shows with pacman -Q | grep nvidia are:

lib32-nvidia-utils 470.82.00-1
linux514-nvidia 470.82.00-1
mhwd-nvidia 470.82.00-1
mhwd-nvidia-390xx 390.144-1
mhwd-nvidia-470xx 470.86-2
nvidia-utils 470.82.00-1
opencl-nvidia 470.82.00-1

I also am using BTRFS and have made a manual snapshot using Timeshift just in case there’s an issue.

My questions are:

  1. Is what I’m doing the right way of going about this?
  2. Is there any reason for me to move to 5.15? I dual boot with Windows 10 (sadly) and the new NTFS seems like a nice change.
  3. How should I approach this situation in regards to moving to a different driver package?
  4. If I should do it through the Hardware Configuration menu, should I wait for the 470xx option to become available?

I see it here:

[tmo@mj ~]$ mhwd -l
> 0000:01:00.0 (0302:10de:1c02) Display controller nVidia Corporation:
                  NAME               VERSION          FREEDRIVER           TYPE
video-hybrid-intel-nvidia-prime            2021.11.04               false            PCI
video-hybrid-intel-nvidia-470xx-prime            2021.11.04               false            PCI
video-hybrid-intel-nvidia-390xx-bumblebee            2021.11.04               false            PCI
          video-nvidia            2021.11.04               false            PCI
    video-nvidia-470xx            2021.11.04               false            PCI
    video-nvidia-390xx            2021.11.04               false            PCI
           video-linux            2018.05.04                true            PCI

and in the Settings’ Hardware Configuration, which is essentially the same (kcmshell5 msm_mhwd), and it’s in both unstable and testing, so I don’t know what you are talking about, maybe it’s too late and I should go to bed :sleeping:. Maybe you didn’t downgrade properly?

Open a terminal and first make sure you are up to date

sudo pacman -Syyu

then check the output of mhwd

mhwd -l

and also what is currently installed as I see you already have the 470 driver you want to install

mhwd -li

But the problem I’m starting to see is that since driver 470 went from video-nvidia to video-nvidia-470xx and driver 495 replaced driver 470 as video-nvidia, you have messed up you mhwd and have conflicting packages.

Give the result of all the command I gave now. I’m pretty sure you didn’t properly downgrade when going back to Testing after this major driver change.

if you need to go nouveau
in TTY or Terminal

sudo mhwd -a pci free 0300

if you need to tests others or futures version of 470 series
you go in Testing

if you need to have others kernels ( there is always nvidia drivers by version kernel , 390 series , 470 series or nvidia series and for hybrid-drivers ( laptops )

Sorry for the late reply as I’ve been away.

I’ve done the sudo pacman -Syyu as you asked.

I’ve also done mhwd -l which gave me this output:

> 0000:26:00.0 (0300:10de:11c2) Display controller nVidia Corporation:
                  NAME               VERSION          FREEDRIVER           TYPE
    video-nvidia-470xx            2021.11.04               false            PCI
    video-nvidia-390xx            2021.11.04               false            PCI
           video-linux            2018.05.04                true            PCI
     video-modesetting            2020.01.13                true            PCI
            video-vesa            2017.03.12                true            PCI

And mhwd -li outputs this:

> Installed PCI configs:
                  NAME               VERSION          FREEDRIVER           TYPE
          video-nvidia            2021.11.04               false            PCI

Warning: No installed USB configs!

Since the preferred driver now appears, I tried to install it through the settings menu but I get this error:

Error: config 'video-nvidia-470xx' conflicts with config(s): video-nvidia

Done ...

Should I approach this differently?

remove video-nvidia and install video-nvidia-470xx, it should do the trick (what you should have done on Unstable branch initially).

sudo mhwd -r pci video-nvidia
sudo mhwd -i pci video-nvidia-470xx

but still you have a messed up mhwd database mixed up, it may catch up later with system updates.

mhwd-nvidia 470.82.00-1
mhwd-nvidia-470xx 470.86-2

Maybe reinstalling mhwd-database can fix it?


Okay there is an error that I get from using the command sudo mhwd -r pci video-nvidia regarding dependencies and can’t continue after this:

error: failed to prepare transaction (could not satisfy dependencies) 
:: removing nvidia-utils breaks dependency 'nvidia-utils' required by cuda
Error: pacman failed!
Error: script failed!

What should I do in this case? Should I remove the packages that I intend to replace manually?

As the messages says, CUDA requires what you want to remove, so remove CUDA, resolve the nvidia remove/install, then reinstall CUDA only when everything is sorted out.

1 Like

Okay. I managed to get the drivers installed and then CUDA afterwards and my computer boots as normal now. Thank you very much for your help and time again!

On a side note, it’s in your best interests to NOT use the unstable branch if you can’t solve such simple dependency conflicts.

Those that use Unstable need to have the skills to get themselves out of trouble when they move their system to this branch. They are the Manjaro users who are most likely to need to use such skills. Using the unstable branch is usually safe but - in rare cases - may cause issues with your system!

Manjaro specific packages such as kernels, kernel modules and Manjaro applications enter the repo on unstable branch and it is those packages which are considered unstable when they enter.

from the wiki Switching Branches - Manjaro

1 Like

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.