The 390.154 nvidia driver no longer works with Linux kernel 6.0+. While there is a patch to at least get the driver to compile for this kernel, it doesn’t actually work for my GeForce GT 630M. Xorg fails to start, telling me:
[ 752.678] (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA GPU at PCI:1:0:0. Please
[ 752.678] (EE) NVIDIA(GPU-0): check your system's kernel log for additional error
[ 752.678] (EE) NVIDIA(GPU-0): messages and refer to Chapter 8: Common Problems in the
[ 752.678] (EE) NVIDIA(GPU-0): README for additional information.
[ 752.678] (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA graphics device!
[ 752.678] (EE) NVIDIA(0): Failing initialization of X screen 0
[ 752.679] (II) UnloadModule: "nvidia"
And in dmesg output i found:
[ 108.021040] NVRM: failed to copy vbios to system memory.
[ 108.021321] NVRM: RmInitAdapter failed! (0x30:0xffff:663)
[ 108.021387] NVRM: rm_init_adapter failed for device bearing minor number 0
Thanks, but the problem isn’t with the “nvidia” package. I’m using the “nvidia-390xx” AUR package, The package you’re suggesting provides the driver version 520.56.06, which doesn’t support GeForce 630M. I’m limited to 390 versions, and that needs to be updated for the latest kernels.
This is because the nvidia drivers don’t suit the 6.0 kernel for now. You can install the linux-headers and nvidia-390xx-dkms driver until an official driver update comes out. Or use a kernel-lts along with the nvidia-lts driver.
Thanks for the suggestions, but if you follow the discussion on the archlinux forum i linked to, you’ll see that nvidia-390xx-dkms just doesn’t work with kernels 6.0 and up. I’m unsure of the exact details, but the patch that enables that kernel module to compile does so by removing some ACPI requirements of the driver that are no longer provided by the kernel. The driver module loads, but then fails when it’s actually used.
So, yeah, i’ve rolled back to an older kernel for now. My question is, how long will i have to wait for an updated driver? :) The support pages say Nvidia will support this GPU with latest kernels until the end of this year, so i hope a fix is coming soon!
Hi guys, any update on this issue ? Is NVIDIA fixing this compatibility issue between kernel 6.0+ and nvidia-390xx ?
Just wanted to know if you can submit an inciden or bug report so that we linux users can follow and get updates regarding the issue solving process technically.
Also, just wanted to remind you kindly that an open source driver for linux seems to be the only viable way forward, otherwise compatibility will always break whenever a kernel interface gets updated. The old binary driver that’s supposed to be stable and work will simply get obsolete and broken. That is exactly the current situation with the binary blob. It just wont work with linux 6.0+.
ATI/AMD has already switched to the open source model for linux, long ago. and nothing bad happened, on the contrary, support, compatibility and even sales have increased for them, not only for desktop usage, also for servers and crypto mining.
To NVIDIA, all I want to say to them is that too much damage has been made already. Damage for the end users, but also for the people that is supposed to get their hardware working and simply cannot because of this ridiculous blockade.
But also that we have no hard feelings. There is always an opportunity to open your eyes and grow. Don’t let ignorance and fear prevent you from advancing and taking the right path forward.
Thanks for the work done so far. Hopefully things will improve in the future.
Hi All
Man glad to hear you’ve made it.
I’m struggling with that. I’ve already made the downgrade of my kernel to the same as yours. I installed the 390.157-6 and everything else. but when I run mkinitcpio -P I get for errors: Module not found: nvidia, nvidia_modeset, nvidia_uvm, nvidia_drm