Boot stuck at multiple points after messing with GPUs

Hello everyone, I’ve been using endeavourOS for a bit more than 5 months now, it’s been a blast, and the forum has had some of the most helpful solutions I’ve seen of my 5 years using arch based linux distributions, sadly, I’m a bit of an overconfidant tinkerer, and I found a way to brick it after it went smoothly for so long.

For context, I have AMD integrated graphics and an NVIDIA 3060ti (hate it, I want to get a an amd GPU, need money first) and for once a game refused to use the nvidia card, I first tried doing it through the BIOS by disabling the integrated graphics, but the NVIDIA card displayed only in a low resolution, thought it was a driver issue so I switched from nvidia open dkms to nouveau, and thus started the issue, my system would start booting, but it’d then display a black screen with an underscore in the corner, ctrl alt f1 showed the log, and it stopped after a different point each reboot

  • reached target graphical interface
  • starting power management daemon

-and another one related to color profiles but which I didn’t note down

most discussions hint that it is a driver install issue, I arch-chrooted (unecessary as I have access to tty but didn’t think of it) and reinstalled nvidia open dkms, and it fixed it, but then I was a bit buzzed after a night out, and decided to install a tool from the AUR to help instead of learning to use bumblebee or prime, the package was prime-switcher, used it to set my gpu to Performance (what I assumed was the nvidia gpu), and the boot issue came back up, I tty’d and used the --uninstall option for prime switcher which is supposed to undo all actions, but nothing changed, I uninstalled it.

I tried many fixes, updated the nvidia drivers, changed from nvidia-open-dkms to nvidia-open, to nouveau, pacman -Syu, yay -Syu, used nvidia-inst which installed nvidia-open-dkms and removed a few packages (bumblebee, bumblebeed, nouveau, vulkan-nouveau, lib32-vulkan-nouveau, lib32-nvidia-utils and a few others), (thought I) reinstalled all packages including AUR ones by doing yay -Sy $(pacman -Qq), but now that I think about it I didn’t specify the --overwrite option, so it might have only reinstalled aur packages, I remember seeing many “package already installed” notices but I didn’t think about it since I was busy, and lastly, I tried disabling the nvidia GPU in BIOS, but that led to the exact same result of hanging at one of the previously mentioned points, oh yes and I ran the eos-nvidia-fix a few times.

It is quite stumping, but then again I am a tinkerer, not an actual power user.

I don’t really know what other information to provide, I do have tty access but no graphical environment to work with, I also have an artix install and install media in case I need to arch-chroot.

I hope we can find a solution together, else I’ll have to reinstall endeavourOS from scratch and migrate all my stuff to it, isn’t that the casual arch experience ey ?

May yall have an excellent day, cheers !

Oh yes, additional information : I use LightDM, set it to autologin and automatically enter a Plasma Wayland environment, thought it could be useful.

In the meantime I will keep researching and trying out solutions, I will update the post shall i find a fix to help another would be tinkerer !

if you have TTY
check in journal log

sudo journalctl -b0 ( -b -1 boot before , etc ... )
sudo systemctl status lightdm.service

you can try

sudo systemctl restart lightdm.service

Well, I had a little suspicion, and checked which desktop managers I had installed, and it turns out sddm was installed and enabled in systemctl, I do not recall doing this, so can’t quite explain it, regardless, since I had issues with it before, I disabled its service and re enabled light dm, managed to boot this time !