You are not logged in.
Hello,
First hardware and software:
Hardware:
Nvidia 3060 Max Q - GPU
AMD 6800H CPU - IGPU - Radeon 680M
Software:
Archlinux installed using arch-install
Envycontrol
The problem:
I installed Envycontrol using - yay and it seemed that it went well.
I tried using Envycontrol to switch from the nvidia graphics card to the amd one by doing ```$ sudo envycontrol -s integrated``` - message - "Switching to integrated mode / Operation completed successfully / Please reboot your computer for changes to take effect!"
After rebooting by doing ```$ sudo reboot now``` Archlinux will start booting and when its about to enter desktop manager the screen basically turns off (or "goes black") but the laptop stays on.
I guess it's a driver issue but I am a complete newbie and I cannot figure out how to fix it exactly.
I've gone through every single post that would pop up on google and I would try every single fix there but to no avail.
To specify - I go back to the nvidia card by using arch-chroot from an archIsoFlash and writing ```# envycontrol -s nvidia``` and in return I am greeted to a similar message like the one above.
Here's the journalctl from a ```$ sudo envycontrol -s integrated``` boot - https://pastebin.com/MQUVSGMp
Again I am very new to Linux, this might as well be my very first Linux experience. I tried fixing this issue on my own by trying to understand most of what's written in the wiki regarding this issue.
I'll try to provide any files/logs necessary.
I've been trying to fix the issue for around 7 hours now.
EDIT 1: I cannot choose which GPU to use in the bios. If it's of any help my laptop is an Asus A15 2022.
EDIT 2: I'll be back in a couple of hours because sleep.
Last edited by Ivogada (2023-05-24 01:09:23)
Offline
1. bbs predates markdown, https://bbs.archlinux.org/help.php#bbcode
2. That's only a fragment of the jorunal, copied out of the pager
Please post your complete system journal for the boot:
sudo journalctl -b | curl -F 'f:1=<-' ix.io
3. "Switching to integrated mode / Operation completed successfully / Please reboot your computer for changes to take effect!"
"--reset" this - obviously… There's also a good chance your system is eligible for https://wiki.archlinux.org/title/PRIME#NVIDIA
4. Yo're running iwd and wpa_supplicant (likely from NM) concurrently. Disable iwd. If you want to use it as backend, see https://wiki.archlinux.org/title/Networ … Fi_backend
Offline
Hello again,
I ran the command -
$ sudo journalctl -b -1 | curl -F 'f:1=<-' ix.io
with -1 so it would include the boot with the integrated GPU
Link for paste: http://ix.io/4wCi
-I did this after following the wiki on https://wiki.archlinux.org/title/PRIME#NVIDIA
Last edited by Ivogada (2023-05-24 17:16:48)
Offline
Also at the moment I use the nvidia driver from
$ sudo pacman -S nvidia
but I tried switching to xf86-video-nouveau after I removed the nvidia driver from before that by doing
$ sudo pacman -Rs nvidia
and the screen freezed during boot - so I switched back to the nvidia driver.
I plan on trying to install the proprietary AMD drivers only and see how that goes since right now I only have mesa for the ATI one.
Here is a list with all the packages I have https://pastebin.com/Rrc7y6DY
I assume I'm missing something important
Last edited by Ivogada (2023-05-24 20:35:29)
Offline
Using config file: "/etc/X11/xorg.conf"
Get rid of that (and envycontrol if you didn't get rid of that already), try again. Also heed the point regarding conflicting network managers they can lead to all sorts of weirdness, including a broken GUI/Xorg session.
Installing the prop amdgpu driver will do nothing for the current issue at hand as that has no relation to the OpenGL implementation which is the only thing that would get swapped. Most issues here are due to people following old advice and configuring things they do not need, breaking what auto-detection would setup correctly by default.
You might also want to remove xf86-video-amdgpu, it has come up with some trouble recently especially in a PRIME context.
Last edited by V1del (2023-05-24 20:48:17)
Online
Hello,
I just got rid of envycontrol and the conf file (it was generated by envycontrol i believe)
I disabled iwd by doing
$ sudo systemctl disable iwd
I also removed the xf86-video-amdgpu package
The reason I had envycontrol was to switch easily between GPUs because when I'm at home I only run it as it is right now - Only on the Nvidia Card (I assume it's only using it right now) because I am constantly plugged in, but when I'm outside I want to safe battery and I find the iGPU to be powerful enough on its own for any load outside.
I was wondering if there are any other viable alternatives to envycontrol except changing between them myself.
Offline
There's also a good chance your system is eligible for https://wiki.archlinux.org/title/PRIME#NVIDIA
And unless that optimus-manager clone also powered down the GPU otherwise, it would not have saved any extra battery.
(I assume it's only using it right now)
Don't assume, post your xorg log - but it'd not be the default behavior.
Offline
Properly configured and working PRIME support does both of those desires automatically and better than envycontrol. You only need to run applications you want on the nvidia card with the prime-run command, everything else renders on the integrated GPU by default and the nvidia GPU is powered off when not invoked via prime-run.
Online
Xorg log(s) as requested
xorg.0.log https://pastebin.com/iGbpzW2X
xorg.1.log https://pastebin.com/AJfHpUJZ
Last edited by Ivogada (2023-05-24 22:18:40)
Offline
You're indeed running on the nvidia GPU, please post your /etc/mkinitcpio.conf
Offline
mkinitcpio.conf as requested
MODULES=()
BINARIES=()
FILES=()
HOOKS=(base systemd autodetect keyboard sd-vconsole modconf block filesystems fsck)
Offline
Add "amdgpu" to the MODULES and rebuild the initramfs.
See whether that changes the order in the xorg log (and yo run on the amd chip)
Otherwise we'll have to tell the server which GPU to use.
Offline
I added amdgpu to the Modules (I'm also wondering if that's where you add other kernel modules)
I'm unsure on what to look for in the xorg log exactly (I think I saw that the open source nvidia driver was being used)
p1 - https://pastebin.com/LebtAX7R
p2 - https://pastebin.com/NSt7Wg94
But I for sure know that either both GPUs are being used or just the Nvidia one since I have an external monitor connected and the port is connected straight to the Nvidia GPU and doesn't work unless its on.
EDIT: I forgot to mention that I didn't know to rebuild the intramfs so I rebooted instead and it seemed to work since a new log files had appeared.
Last edited by Ivogada (2023-05-25 05:43:22)
Offline
But I for sure know that either both GPUs are being used or just the Nvidia one since I have an external monitor connected and the port is connected straight to the Nvidia GPU and doesn't work unless its on.
If you want to use an output attached that chip you cannot power it down, no matter what.
And it'll probably be auto-invoked for reverse-prime when there's an output attached.
I forgot to mention that I didn't know to rebuild the intramfs so I rebooted instead
https://wiki.archlinux.org/title/Mkinit … generation
Editing the config will have achieved nothing because it doesn't do anything.
Edit: and the nvidia GPU is still the primary one and you're running the X11 server on it
Last edited by seth (2023-05-25 06:27:00)
Offline