I’m running a laptop with a 3060 mobile and an AMD cpu. My default arch kernel, after updating to 6.15, does not recognise my dGPU. However, the LTS kernel works just fine, using all the 3 drivers stated here: proprietary, open, and DKMS. To make it clear: LTS is fine, most recent drivers, no big, but the main Arch Kernel is screwed. I installed from the main repo, this is the stable release. After some troubleshooting, the failure to mount seems to be the most likely option.
I have added a nouveau blacklist, set the DRM mode to 1 and modprobe.nvidia to 1 in the efi loader config files for the main kernel. Still nothing.
lastly, the kernel buffer (dmesg) displays this:
[ 215.428029] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[ 215.428030] NVRM: This can occur when another driver was loaded and
NVRM: obtained ownership of the NVIDIA device(s).
[ 215.428032] NVRM: Try unloading the conflicting kernel module (and/or
NVRM: reconfigure your kernel without the conflicting
NVRM: driver(s)), then try loading the NVIDIA kernel module
NVRM: again.
[ 215.428033] NVRM: No NVIDIA devices probed. (edited)
[22:13]
[16.578320] NVRM: GPU 0000:01:00.0 is already bound to NovaCore.
I only have a single driver installed, plus the nouveau in the kernel. this does not make sense.
if you guys can save the main arch kernel, my blessings will be forever in your favour. I’m a newbie, this don’t make sense to me but it might to you.
Godspeed.
That is pretty much what I have, I have a Legion 5 Ryzen/RTX3070 and have no issues (though the 6.15 kernel only came down this morning for me, but I rebooted OK).
- Are you running grub or systemd boot?
- What is in
/etc/kernel/cmdline
(if systemd-boot). - What is in your grub (can’t remember the file name!!) (if grub)
- Have you tried uninstalling, rebooting and reinstalling the kernel?
- is your EFI full? See below…
As always at work in the land of windoze
What I do for nVidia, exclude from the initramfs:
Create /etc/dracut.conf.d/nvidia.conf
# Omit the nvidia driver from the ramdisk, to avoid needing to regenerate
# the ramdisk on updates.
omit_drivers+=" nvidia nvidia-drm nvidia-modeset nvidia-uvm "
Rebuild your mkinitcpio and reboot. If it doesn’t work remove the file and rebuild again.
So, run sudo reinstall-kernels
to rebuild.
systemd.
nvme_load=YES nowatchdog rw root=UUID=d42cb41c-560f-499c-9d92-521ecab5e19c resume=UUID=6c052a18-0d4d-4880-8a3a-6d1789e63012
this is the whole cmdline file.
i have not tried a full reinstall of the kernel: i do not know how.
i will attempt this solution in a couple hours. i’m just using the fallback kernel for the work today since LTS is treating me nicely.
Okay, i solved the problem by downgrading the kernel entirely. apparently it’s an issue that’s popped up around that kernel release. no idea what the selection criteria for doom is, though.
thank you for your help, @xircon
No problem here either, Lenovo Legion with a 3070 TI and AMD cpu
[2025-06-12T10:10:49+0200] [ALPM-SCRIPTLET] Installing kernel 6.15.2-arch1-1
[2025-06-12T10:09:44+0200] [ALPM] upgraded nvidia-dkms (575.57.08-1 -> 575.57.08-3)
No issues here with a MSI crosshair 15 B12UGSZ (RTX 3070 Ti), Acer Predator G9-793-71SJ (GTX 1070) and a old HP pavilion notebook.
tl;dr, update for utils dropped that fixed my issue, i didn’t know and didn’t receive the package.
@MichelN Yeah, i upgraded to .15.2 from 14.10 and upgraded nvidia utils in the process, fixed everything. i’m convinced it was just a small window of time with that kernel, before nvidia utils were updated for it, as according to https://bbs.archlinux.org/viewtopic.php?pid=2245579 (which is my exact issue) updating nvidia utils and updating initramfs fixed the problem. i never updated the utils (beyond what updated with the full system alongside the kernel initially) which is probably my issue. anyway, all resolved now, and nothing to worry about. to michel: i pinged you here because i don’t want to flood the page with replies by replying to your post too.
So a timing issue, I had one a few months back, spent hours trying to sort it, it magically fixed itself… after I had updated again.
This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.