Arch Linux User Repository

For the beginners in need of assistance, first take a look at this recommended guide https://github.com/korvahannu/arch-nvidia-drivers-installation-guide - maybe it helps to solve the issues your are having with this version of the driver.

Hi,

Making an upgrade of my system, I just discover that builds with kernels 6.12.25-2-lts and 6.14.4-arch1-2 crash during dkms compilation with following message:

==> dkms install --no-depmod nvidia/470.256.02 -k 6.12.25-2-lts

Error! Bad return status for module build on kernel: 6.12.25-2-lts (x86_64)
Consult /var/lib/dkms/nvidia/470.256.02/build/make.log for more information.
==> WARNING: `dkms install --no-depmod nvidia/470.256.02 -k 6.12.25-2-lts' exited 10
==> dkms install --no-depmod nvidia/470.256.02 -k 6.14.4-arch1-2

Error! Bad return status for module build on kernel: 6.14.4-arch1-2 (x86_64)
Consult /var/lib/dkms/nvidia/470.256.02/build/make.log for more information.
==> WARNING: `dkms install --no-depmod nvidia/470.256.02 -k 6.14.4-arch1-2' exited 10

From make log file one can find those messages at the end of the file:

cc1 : certains avertissements sont traités comme des erreurs
make[4]: *** [/usr/lib/modules/6.14.4-arch1-2/build/scripts/Makefile.build:207: nvidia/nv-cray.o] Error 1
nvidia/nv-acpi.c: Dans la fonction « nv_acpi_methods_init »:
nvidia/nv-acpi.c:632:5: erreur: déclaration implicite de la fonction « NV_ACPI_WALK_NAMESPACE »; vouliez-vous utiliser « ACPI_NAMESPACE » ? [-Wimplicit-function-declaration]
  632 |     NV_ACPI_WALK_NAMESPACE(ACPI_TYPE_DEVICE, ACPI_ROOT_OBJECT,
      |     ^~~~~~~~~~~~~~~~~~~~~~
      |     ACPI_NAMESPACE
nvidia/nv-acpi.c:644:22: erreur: déclaration implicite de la fonction « acpi_bus_get_device »; vouliez-vous utiliser « acpi_is_pnp_device » ? [-Wimplicit-function-declaration]
  644 |             retVal = acpi_bus_get_device(nvif_parent_gpu_handle, &device);
      |                      ^~~~~~~~~~~~~~~~~~~
      |                      acpi_is_pnp_device
cc1 : certains avertissements sont traités comme des erreurs

Hi,

Just a heads-up:

The following versions: - glibc-2.41+r48+g5cb575ca9a3d-1
- gcc 15.1.1+r7+gf36ec88aa85a-1
- gcc-libs 15.1.1+r7+gf36ec88aa85a-1

break module builds due to various changes in calls.

@Nishan: I also got a black screen, but only briefly until the Chrome or other Chrome-based apps started. Also, I had some graphical glitches in bspwm and lemonbar, and redshift stopped working. I added this to the kernel parameters in the GRUB config: nvidia-drm.modeset=1 nvidia-drm.fbdev=1

Regenerated the grub config, and now it seems this solved the problem for me. If you are using GRUB, try adding those parameters in the GRUB menu when booting.

I got the info from: https://github.com/korvahannu/arch-nvidia-drivers-installation-guide

Im getting a black screen, but my system is running coz i can blindy do things like log in reboot etc, also disabled the display manager thinking it might be the problem but problem still remains, i used your installation Process again and now i can only see the first journal after grub but still black on system. My gpu is gt710.

Working on my GT 710 with kernel 6.14.1. Thanks Juan and SoftEcpert!

I noticed that Juan provided a patch, so I updated the release to include it.

Again, a heartfelt "thank you", Juan !

Hi @lighterowl and @injiniero; thanks for the patch and, respectively, for trying it and confirming that it works.

@everybody: I pushed the v7.97 of the package; please test it and give feedback.

I can confirm that lighterowl's patch works for my old GeForce GTX 760 with 6.14 kernel. I hope that SoftExpert publish that patch.

Thank you

Doesn't compile out-of-the-box with 6.14, thankfully the fix is simple and amounts to ifdefing out a removed field from a struct, see https://github.com/NVIDIA/open-gpu-kernel-modules/pull/783 for a link to the relevant Linux commit.

Here's the patch I applied locally to build :

diff -u a/nvidia-drm/nvidia-drm-drv.c b/nvidia-drm/nvidia-drm-drv.c
--- a/nvidia-drm/nvidia-drm-drv.c       2024-05-02 16:50:00.000000000 +0200
+++ b/nvidia-drm/nvidia-drm-drv.c       2025-04-09 08:45:50.752546448 +0200
@@ -868,7 +868,9 @@
     .name                   = "nvidia-drm",

     .desc                   = "NVIDIA DRM driver",
+#if LINUX_VERSION_CODE < KERNEL_VERSION(6, 14, 0)
     .date                   = "20160202",
+#endif

 #if defined(NV_DRM_DRIVER_HAS_DEVICE_LIST)
     .device_list            = LIST_HEAD_INIT(nv_drm_driver.device_list),
 .device_list            = LIST_HEAD_INIT(nv_drm_driver.device_list),

Can confirm nvidia-smi, vkcube, and wine+dxvk still work OK with my GT 645M under 6.14.

installing this driver over my nouveau fixed whatever issue i was having when trying to launch dota2 (getting the lack of vulkan error), but i am still having some kind of issues in game now where i freeze and the game will shut down to desktop in 20-30 sec, and one time i got a vulkan error of some sort. Another issue i have found is about half way through a game, so 30+ minutes in, i start to get some kind of screen stuttering that isn't caused by fps or net issues afaik. I'm running the latest Garuda broadwing XFCE with an Asus GTX 660 Ti and an Asus mobo of around the same time period (2011), dual core AMD 3.4ghz (270). I've followed instructions for installing that korvahannu link had. I've did my best to eliminate what i could from too many things running, to removing temperature as a potential issue. I've just begun to learn Linux after being a long time windows user; win7 never had a single crash or freeze in Dota2 for me, but as most probably know by now, win7 is really almost out of support for anything anyone wants, and dota2 was the last straw for me about a month ago when it stopped working as well. i'm gradually trying different video settings in game, although im almost always running it in lowest settings with fps locked at 60 max and 1920x1080 reso but with most anything else turned off that i can. Any ideas about what i can do to make things more reliable? maybe use an older kernel than the latest LTS i have now?