Arch Linux User Repository

Installing this package requires building kernel module, which requires linux-headers to make successful build. Add linux-headers to dependency?

@ogarcia

same here for some reason its just happens with the gdm display manager but if i initialize with sddm display manager and select the same desktop with gnome/wayland it will work fine i dont know why its happening

@ogarcia. exactly the same issue here ! Same workaround. Very strange indeed...

Edit : problem solved (just workaround ?) by adding the "nvidia_drm.modeset=1" option to my grub. (Note: I use x11 and not wayland)

Hi! I don't know if this has happened to anyone but with the latest driver update a second monitor has appeared (connected to None-1-1) which makes me that the desktop is displayed twice as wide as normal and as I approach the mouse pointer to the corners it moves, ie, my resolution is 1680x1050 but I have a desktop of 3360x1050.

I have this card:

[   822.146] (II) NVIDIA(0): NVIDIA GPU NVIDIA GeForce GTX 660 (GK106) at PCI:15:0:0
[   822.146] (II) NVIDIA(0):     (GPU-0)
[   822.146] (--) NVIDIA(0): Memory: 2097152 kBytes
[   822.146] (--) NVIDIA(0): VideoBIOS: 80.06.58.00.09
[   822.146] (II) NVIDIA(0): Detected PCI Express Link width: 16X

I have solved it with the GNOME monitor configuration telling it to mirror the real monitor but it is something strange and it has never happened to me before. Any idea? Has it happened to anyone else?

@caiotbc, I have a similar hardware setup, with a similar problem, almost 6 months ago with windows giving me error 43, then I migrated to an arch VM, and these last months even the arch VM won't boot anymore, I was using already linux-lts510, but as the kernel versions keep getting updated and not many people are testing on this card the problems keep piling, I suggest trying some older kernels to see if any of them work for you. But as you have said, there are not many users of this hardware around, specially people working with GPU Passthrough. But this is a kernel problem and arch kernel and dkms are not exactly easy to bisect and report as regressions upstream, but if you have the time and know-how to report it, I would be glad to go back to VM's.

@benjarobin, I believe the nvidia enabled some support for it: https://nvidia.custhelp.com/app/answers/detail/a_id/5173/~/geforce-gpu-passthrough-for-windows-virtual-machine-%28beta%29

@caiotbc Just for information, GPU passthrough is (was?) blocked by Nvidia on Windows. It is only allowed with Quadro cards. So maybe this is the reason...

@Sinyria and @SoftExpert sorry for replying late, just got around to trying it again. So, it seems the culprit is something to do with OVMF GPU Passthrough after all. Yesterday when I tried installing it on the bare metal arch install and it failed I incorrectly assumed it was the same error I had described on the comments below, however investigating further revealed it was some sort of conflict with the VFIO drivers loading before the nvidia ones. After messing around in mkinipcio.conf I have now successfully installed the nvidia-470xx-dkms on bare metal, it is running fine. It still refuses to install inside the arch virtual machine, I guess maybe I should seek help on the various GPU passthrough forums, or maybe just give up getting this old GPU to work on this rather specific configuration. I tried both of your suggestions inside the VM but it did not change the errors.

In any case, thank you all for your help, I would appreciate if you guys had further suggestions but I understand that it is almost 100% certainly not a fault of this package and the error has something to do with the virtual machine setup, sorry for the confusion!

@caiotbc: thanks indeed for the various logs.

I found

Aug 04 20:04:16 caio-arch-x470 kernel: NVRM: GPU 0000:06:00.0: RmInitAdapter failed! (0x22:0x56:667)
Aug 04 20:04:16 caio-arch-x470 kernel: NVRM: GPU 0000:06:00.0: rm_init_adapter failed, device minor number 0

in your log, and there is this fix from here 1 :

where they changed the grub cmdline from

GRUB_CMDLINE_LINUX_DEFAULT="quiet"

to

GRUB_CMDLINE_LINUX_DEFAULT="quiet rcutree.rcu_idle_gp_delay=1" 

in

/etc/default/grub

and that supposedly fixed it.

would be cool if you could try it and let us know how it goes. =)

best of luck, Siny

@caiotbc: Thanks for the detailed set of information!

From what I could see:

• nvidia module for the kernel 6.4.8 was built (which means the kernel headers were indeed present at the time of compilation)

• at boot time, the system tries to load the nvidia kernel module, but there is some issue (udev-worker)

Aug 04 20:04:07 caio-arch-x470 (udev-worker)[255]: nvidia: Process '/usr/bin/bash -c 'if [ ! -c /dev/nvidiactl ]; then /usr/bin/mknod -Z -m 666 /dev/nvidiactl c $(grep nvidia-frontend /proc/devices | cut -d \  -f 1) 255; fi'' failed with exit code 1.

• X server loads the nvidia module, tries to configure it, but the nvidia module refuses to accept the video card:

[    12.541] (II) NVIDIA GLX Module  470.199.02  Thu May 11 11:49:03 UTC 2023
[    12.541] (II) NVIDIA: The X server supports PRIME Render Offload.
[    15.623] (EE) NVIDIA(GPU-0): The NVIDIA GPU at PCI:6:0:0 is not supported by the 470.199.02
[    15.623] (EE) NVIDIA(GPU-0):     NVIDIA driver.
[    15.623] (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA graphics device!
[    15.623] (EE) NVIDIA(0): Failing initialization of X screen

• nvidia device is indeed at the PCI location refused by the driver

06:00.0 VGA compatible controller [0300]: NVIDIA Corporation GK104 [GeForce GTX 680] [10de:1180] (rev a1) (prog-if 00 [VGA controller])
        Subsystem: eVga.com. Corp. GK104 [GeForce GTX 680] [3842:2682]
        Physical Slot: 0-6
        Flags: bus master, fast devsel, latency 0, IRQ 60
        Memory at 80000000 (32-bit, non-prefetchable) [size=16M]
        Memory at 706000000000 (64-bit, prefetchable) [size=128M]
        Memory at 706008000000 (64-bit, prefetchable) [size=32M]
        I/O ports at c000 [size=128]
        Expansion ROM at <ignored> [disabled]
        Capabilities: <access denied>
        Kernel driver in use: nouveau
        Kernel modules: nouveau

At this point, I would be curious to experiment with the package nvidia-prime, since I saw that Prime offloading is supported.

There is also optimus-manager, which is supposed to help for configurations like yours.

Also, could you share the contents of your xorg.conf file ? The trace from Xorg.0.log is surprisingly quiet about the parameters read from xorg.conf which makes me believe there is no such file and, perhaps, you might need one.

Since my config is really different from yours, I cannot experiment or copy elements from it, and we need to take it slowly to find the real reason why the driver rejects your card.

Edit: corrected some markdown and added reference to optimus-manager.