There are no modules in your mkinitcpio.conf.
Read up on nvidia arch wiki and see what you have to do for your gpu.
On my system I have “crc32c nvidia nvidia_drm nvidia_uvm nvidia_modeset”
I also have “nvidia_drm.modeset=1” in my grub, but again, you should read up on this, it might differ from what I have depending on what nvidia gpu you have.
Not sure uninstalling anything would do any good here, I think it has to do with your computer not loading the nvidia drivers.
thanks bedna - I have read that arch thing on gpus, could not really understand a lot of of it though. Some of it I could comprehend.
hmmm, well that’s not good…I sorta thought all of the above would fix that…I’ve pretty much followed all the advice and tips to the letter, which to me should have meant the OS loading the driver as instructed. Seems there is still something blocking this after all 
Could this be an answer? (not that I can understand it 
) [SOLVED] Nvidia modules not being recognized by mkinitcpio / Newbie Corner / Arch Linux Forums
…wrong place
I repeat, read the nvidia arch wiki and do what you need to.
For example, are you saying you do not need/want to do anything under point 1.3? (with sub points)
https://wiki.archlinux.org/title/NVIDIA#DRM_kernel_mode_setting
I am looking at this one… Configure Graphics Cards - Manjaro
Then read what I keep recommending instead. xD
doing that now
Seems to be only for a single card…mine is a dual card system. But I’ll keep reading…
See, what I don’t get is the grep report that that page suggests right at the top shows the intel Xe Graphics card Kernel driver in use = i915 and the RTX2060 Kernel driver in use = nvidia
Which, to me, means they are being used, but maybe in the “linux speak” world it means something different??..however you write above there is no module loaded…all very confusing. I found that link above by google searching for the term “no module loaded”
I think that’s best because your statement makes no sense to me.
hmmm, well it means exactly what it means, sorry it makes no sense to you! That wiki, to me, seems geared toward single gpu…makes no mention of a dual gpu system…so are you saying I need to run those processes twice to do both gpus?
Whereas the Manjaro GPU guide does indeed make mention and has instructions for dual gpu systems.
Sorry, still makes no sense to me.
Do you have dual nvidia cards or what on earth are you talking about?
If you are talking about the PRIME function that makes it so you can use igpu for light work and egpu for heavy work. 2 things.
The problem you have seems that you are running on the open source drivers rather than the closed source.
If you are not willing to do what is suggested then there is nothing to do. Manjaros wiki clearly does not cover this and the situation demand bigger intervention, and I THINK what you need to do is in the arch wiki.
Good luck!
tbh, the ENTIRE thread has been about using the intel iris xE GPU for day to day, and the nvidia GPU for gaming and other video intensive projects that also involve audio.
Sorry you seemed to have missed that entirely when recommending that wiki. Which as you say has nothing to do with dual cards.
To your point 2. - Again - read above ^.
I’d hazard a guess that others are doing exactly this, so it is possible. Perhaps not under Manjaro (or Arch) but perhaps another distro may do the job? Even back to windows perhaps, although writing that gives me shivers!
And by the by, I have also read that although Manjaro is based on Arch, they run different kernels, so perhaps a lot of Arch stuff is not applicable??
I agree, me for example is doing it, I have internal amd gpu and an nvidia pci card I use with prime-run, but you refuse to do what I recommend and keep asking strange questions instead.
So, before this gets out of hand, I just wish you good luck and leave. <3
were you installing at one point the nvidia drivers from other sources than manjaro?
post output from:
pacman -Qm
pacman -Qs nvidia
were you adding some kernel parameters?
head /etc/default/grub
I am going to re-install and start from scratch again…as far as I can tell I have followed every step. The drivers are installed, the gpu’s are both seen, but the system STILL hangs on to using the generic drive installed at OS load-in.
Your help is well appreciated and I am sorry you got frustrated in the end (understandable!) Maybe by restarting I might discover what step (perhaps even a minor one) I either missed or f@#$ed up!
Cheers
if you are going to reinstall you might as well try this before you do.
sudo nano /etc//etc/mkinitcpio.conf and add nvidia nvidia_drm nvidia_uvm nvidia_modeset to the MODULES= section
sudo nano /etc/default/grub and add nvidia_drm.modeset=1 somewhere between the " in GRUB_CMDLINE_LINUX_DEFAULT=
sudo mkinitcpio -P
and then
sudo update-grub
Reboot and see if any different.
This is exactly what was suggested in the link I posted.
^^ Thanks bedna - when I get to it, that is what I will do…just had to re-install windows temporarily as the drive somehow got shizzled after this foray into linux…needed the chkdsk and repair!
Back to Manjaro now, and have completed the update, got my bluetooth working, changed over to pipewire and now to finally set the graphics hahaha!
Still a bit leery of running the suggested kernel update though, last time it broke the bluetooth…but that’s another issue
I’m going back to the start of the thread and working my way back through. I guess it’s still recommended to just allow Manjaro Hardware manager do the automatic thing, yeah?
When I run that commnad line the editor opens with this message-
[ Directory '/etc//etc' does not exist ] ```That was obv a typo.
sudo nano /etc/mkinitcpio.conf
haha, I thought it might have been - I did try it again with a single / but didn’t remove the second etc…
Just to make sure I am doing it correctly…
It shows MODULES=" " Should those entries go between the " " or directly under it vis:
MODULES " "
nvidia nvidia_drm nvidia_uvm nvidia_modeset
Sorry for the question, first time editing this sorta stuff on Linux…
All g - did some research, seems it goes next to the =
Righto - did all of that, now to reboot. Assuming it goes ok, what commands do I run again to check?
I’m going to mark it as solved, as at least the system is now running on the nVidia card. The Intel Iris xe graphics, while referenced in lspci grep, still make no appearance in any other info generating commands, inxi and glxinfo
I’ve tried the glxi renderer string thing and both (normal and prime-run) both come up with nVidia. As bedna commented, it can be a bit “flaky” - prime run that is.
prime-run isnt being flaky.
its doing what it should.
It cant possibly switch you to nvidia if you are already on it.
This could be due to the port you are using to connect your display, some quirk of the acpi table, or a BIOS setting (some have the ability to disable the iGPU), etc.
If none of those simple fixes work then you may be stuck using nvidia only or looking at configuring reverse prime if there ends up being no way to explicitly set the correct gpu hierarchy (like using an xorg file).
But none of that is the fault of prime-run.
prime-run is just a script that does this:
__NV_PRIME_RENDER_OFFLOAD=1 __VK_LAYER_NV_optimus=NVIDIA_only __GLX_VENDOR_LIBRARY_NAME=nvidia "$@"