You are not logged in.
Hello,
on my Arch Linux machine with an NVIDIA RTX 2060 card and an integrated GPU in the Intel processor, I installed NVIDIA proprietary drivers and set it by following https://wiki.archlinux.org/title/NVIDIA
The NVIDIA modules (nvidia nvidia_modeset nvidia_uvm nvidia_drm) are loaded correctly and I set nvidia-drm.modeset=1 as kernel parameter.
I removed bumblebee and optimus-manager packages because blacklisted NVIDIA modules.
Despite this, when I run 'glxinfo | grep -i "opengl renderer"', I get 'OpenGL renderer string: Mesa Intel(R) UHD Graphics 630 (CFL GT2)' (that will shown also in Settings -> About).
How can I make OpenGL to refer to NVIDIA and not to the integrated GPU?
Offline
Hello,
on my Arch Linux machine with an NVIDIA RTX 2060 card and an integrated GPU in the Intel processor, I installed NVIDIA proprietary drivers and set it by following https://wiki.archlinux.org/title/NVIDIA
That article covers a variety of different cards, drivers, configuration methods etc. What exactly did you do?
CLI Paste | How To Ask Questions
Arch Linux | x86_64 | GPT | EFI boot | refind | stub loader | systemd | LVM2 on LUKS
Lenovo x270 | Intel(R) Core(TM) i5-7200U CPU @ 2.50GHz | Intel Wireless 8265/8275 | US keyboard w/ Euro | 512G NVMe INTEL SSDPEKKF512G7L
Offline
D3vil0p3r wrote:Hello,
on my Arch Linux machine with an NVIDIA RTX 2060 card and an integrated GPU in the Intel processor, I installed NVIDIA proprietary drivers and set it by following https://wiki.archlinux.org/title/NVIDIAThat article covers a variety of different cards, drivers, configuration methods etc. What exactly did you do?
In particular, I installed the following packages: nvidia, nvidia-dkms, nvidia-settings, nvidia-utils.
Then I installed NVIDIA related applications as: bumblebee, cuda, disper, envycontrol, gwe, nvflash, nvidia-prime, nvidia-xrun, nvtop, optimus-manager, optimus-manager-qt.
I added nvidia nvidia_modeset nvidia_uvm nvidia_drm in MODULES variable in /etc/mkinitcpio.conf and added the kernel parameter nvidia-drm.modeset=1 in /etc/default/grub.
Since bumblebee and optimus-manager blacklist NVIDIA modules, then I uninstalled them.
In this manner, I was able to make working the dual screen (that at the beginning didnt work).
I am on GNOME 43 with Wayland and I'm using LightDM as Login Manager. On the wiki there is only a dedicated part of GDM with NVIDIA drivers but no with LightDM.
Offline
Which of the Available methods are you trying to configure?
For example, I'm not a nvidia person, but don't you need something like
prime-run glxinfo | grep "OpenGL renderer"if you're using PRIME?
Compare, for example, nvidia-xrun and so on.
As I understand it, you have to pick a method and configure for that specific option, whereas you seem to be installing everything and anything and hoping it will somehow work.
CLI Paste | How To Ask Questions
Arch Linux | x86_64 | GPT | EFI boot | refind | stub loader | systemd | LVM2 on LUKS
Lenovo x270 | Intel(R) Core(TM) i5-7200U CPU @ 2.50GHz | Intel Wireless 8265/8275 | US keyboard w/ Euro | 512G NVMe INTEL SSDPEKKF512G7L
Offline
If you want to first of all make sure your nvidia card is used, I would configure your system to use only the nvidia card. And after that switch to whatever method you like.
Here is the link: https://wiki.archlinux.org/title/NVIDIA … phics_only
On the other hand., it only describes the configuration for Xorg.
The only method I found that mentions wayland is this one: https://wiki.archlinux.org/title/NVIDIA … Vidia-eXec.
Maybe you can try this one.
Offline
@cfr @1uk thank you for your clear answers.
If I run prime-run, I get:
prime-run glxinfo | grep "OpenGL renderer"
OpenGL renderer string: NVIDIA GeForce RTX 2060/PCIe/SSE2So, does it mean that if I want to run an application by NVIDIA GPU, I need to use "prime-run <command>"?
About https://wiki.archlinux.org/title/NVIDIA … phics_only as you reported, it seems to be related only to Xorg. Currently I'm on Wayland. Indeed "nvidia-xrun" seems to not work on Wayland.
I installed NVidia-eXec and when I run "nvx ps" I get:
25250 Xwayland
24871 gnome-shellSo I guess the GPU is already running but OpenGL is still using the integrated Intel GPU.
By using NVidia eXec should I expect to have OpenGL renderer string as NVIDIA GPU?
Offline