[SOLVED] Loading loop after GPU switch. From Nvidia to AMD. | Arch Linux Community | Page 1

wanton pewter Oct 8, 2023, 1:50 PM

#

Hey guys,

I bought a new GPU, in this case a RX 7700 XT.

Setup:
Ryzen 5 3600
MSI x570-A Pro
Sapphire RX 7700XT

Problem: After startup it stays in a loop, that's how I would describe it. It's still possible to switch to a different tty. I get an error about sddm it could not start the display server. So my guess is that it should be because of my GPU switch.

What I did already: Downloaded free driver for amdgpu and added amdgpu (radeon too) in mkinitcpio modules. For sddm I checked my keyboard layout, but everything should be correct. So they(googled) said it could be the layout, but unfortunately it's not.

I have two screenshots, one with the loop and the other is the journal error.

hardy perch Oct 8, 2023, 1:53 PM

#

have you tried reinstalling sddm?

#

I would consider deleting all associated files with sddm and then reinstalling if a simple reinstall of the package doesn't do it, in case there's a config somewhere that's referencing nvidia?

#

and if that doesn't work perhaps wayland as well?

real spruce Oct 8, 2023, 2:00 PM

#

@wanton pewtertry loading the module early perhaps that would work

wanton pewter Oct 8, 2023, 2:00 PM

#

hardy perch have you tried reinstalling sddm?

I will try it. I think I never configured myself for something for sddm.

real spruce Oct 8, 2023, 2:00 PM

#

put amdgpu in modules in /etc/mkinitcpio.conf and do sudo mkinitcpio -P

#

you can do this in the tty

wanton pewter Oct 8, 2023, 2:01 PM

#

real spruce put amdgpu in modules in /etc/mkinitcpio.conf and do `sudo mkinitcpio -P`

Already did that.

#

But I did add Radeon too, I think I will remove that

#

I configured the modprobe files for amdgpu too

hardy perch Oct 8, 2023, 2:02 PM

#

.aw amdgpu

fossil thicketBOT Oct 8, 2023, 2:02 PM

#

hardy perch .aw amdgpu

AMDGPU

https://wiki.archlinux.org/title/AMDGPU

hardy perch Oct 8, 2023, 2:06 PM

#

https://wiki.archlinux.org/title/Kernel_mode_setting#Early_KMS_start
that of any help?
And did you have any nvidia shit potentially?
Like I have this in the top of /etc/default/grub

GRUB_CMDLINE_LINUX_DEFAULT="quiet loglevel=0 nvidia_drm.modeset=1"
#GRUB_CMDLINE_LINUX=""
GRUB_CMDLINE_LINUX="rd.driver.blacklist=nouveau modprobe.blacklist=nouveau rhgb quiet rd.driver.blacklist=nouveau modprobe.blacklist=nouveau"

real spruce Oct 8, 2023, 2:06 PM

#

i misread

wanton pewter Oct 8, 2023, 2:07 PM

#

hardy perch https://wiki.archlinux.org/title/Kernel_mode_setting#Early_KMS_start that of any...

Yeah checked that already. But i0bz mentioned I could load there amdgpu too ?

real spruce Oct 8, 2023, 2:07 PM

#

yes

wanton pewter Oct 8, 2023, 2:08 PM

#

Can you post that command that I can upload some lines ?

real spruce Oct 8, 2023, 2:08 PM

#

real spruce put amdgpu in modules in /etc/mkinitcpio.conf and do `sudo mkinitcpio -P`

.

hardy perch Oct 8, 2023, 2:09 PM

#

what does lspci -v show for your graphics?

wanton pewter Oct 8, 2023, 2:09 PM

#

real spruce .

Yep added it mkinitcpio executed

real spruce Oct 8, 2023, 2:09 PM

#

hardy perch what does `lspci -v` show for your graphics?

try this ^ see if it shows your graphics card

hardy perch Oct 8, 2023, 2:10 PM

#

01:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1080] (rev a1) (prog-if 00 [VGA controller])
    Subsystem: ASUSTeK Computer Inc. GP104 [GeForce GTX 1080]
    Flags: bus master, fast devsel, latency 0, IRQ 128
    Memory at de000000 (32-bit, non-prefetchable) [size=16M]
    Memory at c0000000 (64-bit, prefetchable) [size=256M]
    Memory at d0000000 (64-bit, prefetchable) [size=32M]
    I/O ports at e000 [size=128]
    Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
    Capabilities: <access denied>
    Kernel driver in use: nvidia
    Kernel modules: nouveau, nvidia_drm, nvidia

is mine, for example

#

(just to help with what you're looking for)

wanton pewter Oct 8, 2023, 2:12 PM

#

http://ix.io/4It7

#

It's there

#

Amdgpu module should be loaded

hardy perch Oct 8, 2023, 2:13 PM

#

Kernel driver in use: amdgpu

wanton pewter Oct 8, 2023, 2:13 PM

#

Yeah correct

hardy perch Oct 8, 2023, 2:13 PM

#

that's...annoying, because it's right

#

lol

wanton pewter Oct 8, 2023, 2:13 PM

#

That's normal ?

#

Ahh lol

hardy perch Oct 8, 2023, 2:14 PM

#

if it was wrong we could go "ah ha!"

#

systemctl status sddm

wanton pewter Oct 8, 2023, 2:16 PM

#

http://ix.io/4It9

#

That's weird, it's active

#

The thread I read, there it was disconnected

#

Or not active

hardy perch Oct 8, 2023, 2:16 PM

#

but below it says a load of stuff failed

wanton pewter Oct 8, 2023, 2:16 PM

#

Yeah

#

That's why I'm wondering

real spruce Oct 8, 2023, 2:17 PM

#

could you show me your xorg log

#

cat /var/log/Xorg.0.log

wanton pewter Oct 8, 2023, 2:18 PM

#

http://ix.io/4Ita

hardy perch Oct 8, 2023, 2:18 PM

#

and also if it still happens when you manually start is, sudo systemctl start sddm

#

don't run that with ix.io

#

just run it by itself

wanton pewter Oct 8, 2023, 2:19 PM

#

Ahh ok, thx for the hint 👍

real spruce Oct 8, 2023, 2:19 PM

#

wanton pewter http://ix.io/4Ita

weird why was it searching for an nvidia module

real spruce Oct 8, 2023, 2:19 PM

#

hardy perch https://wiki.archlinux.org/title/Kernel_mode_setting#Early_KMS_start that of any...

spot on

wanton pewter Oct 8, 2023, 2:19 PM

#

I was in there and no Nvidia is mentioned there

#

Should I upload grub ?

real spruce Oct 8, 2023, 2:20 PM

#

wanton pewter I was in there and no Nvidia is mentioned there

last part

[    10.566] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[    10.567] (II) Module glx: vendor="X.Org Foundation"
[    10.567]     compiled for 1.21.1.8, module version = 1.0.0
[    10.567]     ABI class: X.Org Server Extension, version 10.0
[    10.567] (II) LoadModule: "nvidia"
[    10.567] (WW) Warning, couldn't open module nvidia
[    10.567] (EE) Failed to load module "nvidia" (module does not exist, 0)
[    10.567] (EE) No drivers available.
[    10.567] (EE) 
Fatal server error:
[    10.567] (EE) no screens found(EE) 
[    10.567] (EE) ```

#

try deleting your xorg.conf

wanton pewter Oct 8, 2023, 2:22 PM

#

And then I should restart ?

real spruce Oct 8, 2023, 2:22 PM

#

try it

wanton pewter Oct 8, 2023, 2:22 PM

#

Restarting now

#

Omg

#

It was that

#

It's running now everything

#

Omg guys I love you ... Thank you so much.

real spruce Oct 8, 2023, 2:24 PM

#

https://cdn.discordapp.com/emojis/1107146507305615410.webp?size=48&name=skulley&quality=lossless

wanton pewter Oct 8, 2023, 2:24 PM

#

Really I appreciate it, just never thought that the xorg conf could so that

real spruce Oct 8, 2023, 2:24 PM

#

oh wait did you switch gpu?
you probably had an xorg config set to use your old gpu

wanton pewter Oct 8, 2023, 2:25 PM

#

From Nvidia to AMD

#

Now my i3 screen etc is appearing and seem to run

#

Ahh

#

Now I understand

#

If you guys visit Germany someday I will give you some beers for free. @hardy perch @real spruce Again big thanks guys 👍

silent coralBOT Oct 8, 2023, 2:28 PM

#

vegancookies, and i0bz received a thank you cookie!

#[SOLVED] Loading loop after GPU switch. From Nvidia to AMD.