You are not logged in.
Hello!
This is my first time installing Arch onto a computer. I went through the installation guide and everything booted normally with the text interface. I then tried to install the nvidia drivers, and this is when things took a turn.
I went through the wiki page on nvidia drivers and looked up which driver I should be using. Since I currently have a 3060, I ran
pacman -S nvidia
Then, I removed kms from the HOOKS array in /etc/mkinitcpio.conf and regenerated the initramfs. Next, because I have a Wayland configuration, I went to check my DRM kernel mode setting. However, when I ran
cat /sys/module/nvidia_drm/parameters/modeset
I was told that no such file exists. This didn't seem great, but I tried to reboot the system and see if the file would generate upon reboot. Unfortunately, upon reboot, after the grub bootloader screen appeared and I chose Arch linux, after some initial text referring to ram and my nvme, my screen went black.
I tried to access tty1 by using the keyboard shortcuts, but nothing registered and no terminal screen would come up. I ended up having to chroot using the boot media to get the journalctl as well as the dmesg outputs. I'm at my wits end, and I don't know what I'm doing wrong.
Offline
The journal from the archiso won't help much.
Boot the multi-user.target (2nd link below) in doubt along "nomodeset" and post the journal from there.
Or post the last journal of the installed system from the iso, https://wiki.archlinux.org/title/System … al_to_view
Offline
Thank you for the help!
Offline
That's the tailing 1000 lines only, but nvidia gets loaded and the multi-user.target is reached.
Try to add "nvidia_drm.modeset=1" to the https://wiki.archlinux.org/title/Kernel_parameters -- it'll block the simpledrm device.
Then post the *complete* journal of that boot (if it still doesn't work)
Offline
So sorry!
I did as you asked and added "nvidia_drm.modeset=1" to the linux line, and the screen remained black after booting up.
Here is the full journalctl of that boot
Offline
status quo ante, there're no other framebuffer devices, no errors.
Did you try "nomodeset" and does that work?
Use instead "nvidia.NVreg_EnableGpuFirmware=0 nvidia_drm.fbdev=0 nvidia_drm.modeset=1"
https://wiki.archlinux.org/title/NVIDIA … P_firmware
https://wiki.archlinux.org/title/NVIDIA#fbdev
Offline
The second line worked! Thank you so much--my desktop booted into tty1.
However, when I tried to install kde plasma and launch from console with
sudo pacman -S plasma
/usr/lib/plasma-dbus-run-session-if-needed /usr/bin/startplasma-wayland
,
I was met with the same black screen again.
Here is the journalctl from that boot. Should I make a new forum post about this problem?
Last edited by alicep (2025-05-17 14:16:56)
Offline
May 17 09:50:35 Archie kwin_wayland[821]: kwin_core: Applying output config failed!
https://wiki.archlinux.org/title/KDE#Un … lution_set ?
The output names change w/ the driver (no, idk who's brilliant idea that was… )
May 17 09:50:36 Archie org_kde_powerdevil[1004]: org.kde.powerdevil: [DDCutilDisplay]: ddca_open_display2 -13
May 17 09:50:36 Archie org_kde_powerdevil[1004]: org.kde.powerdevil: [DDCutilDisplay]: "LG ULTRAWIDE" retrying to set DDC/CI brightness in 1000 milliseconds - attempt no. 1
May 17 09:50:37 Archie org_kde_powerdevil[1004]: org.kde.powerdevil: [DDCutilDisplay]: ddca_open_display2 -13
May 17 09:50:37 Archie org_kde_powerdevil[1004]: org.kde.powerdevil: [DDCutilDisplay]: "LG ULTRAWIDE" retrying to set DDC/CI brightness in 2000 milliseconds - attempt no. 2
powerdevil seems to have issues w/ the backlight.
Start easy and try to "startx" (you'll need "xorg-xinit" "xorg-server" and "xterm" (and twm if you want the WM to start up) or weston/sway or so (ie. basic stuff just to see whether you can get X11 or wayland to operate the framebuffer now)
Offline
I installed sway, but when I ran sway, I got the same black screen where I can't even get into a terminal screen. I also tried to download xorg-xinit, xorg-server, and xterm to run startx, but I got the same black screen. I also tried to find where a possible log file would be for sway, but I couldn't.
I'm not sure if the journalctl would be helpful in this, but here it is
Offline
Please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General
For X11: Entirely black screen or black screen and mouse cursor?
Can you from there switch to another VT (ctrl+alt+f3) and check "DISPLAY=:0 xrandr -q"?
Offline
Here is the Xorg log . As for X11, it's an entirely black screen, and I can't switch to another VT to check.
Offline
Looks unspectacular
[ 6.357] (--) NVIDIA(GPU-0): LG Electronics LG ULTRAWIDE (DFP-0): connected
[ 6.357] (--) NVIDIA(GPU-0): LG Electronics LG ULTRAWIDE (DFP-0): Internal TMDS
[ 6.357] (--) NVIDIA(GPU-0): LG Electronics LG ULTRAWIDE (DFP-0): 600.0 MHz maximum pixel clock
…
[ 6.367] (II) NVIDIA(0): Virtual screen size determined to be 3440 x 1440
Do you have another monitor (FullHD@60 Hz standard office stuff)?
Can you ssh into the system?
You could also try to schedule the xrandr dump:
(sleep 5; DISPLAY=:0 xrandr --verbose > ~/xrandr.txt; printf '\a') & startx
This will startx, wait 5s, dump xrandr -q into a file and ideally then beep (not sure whether the bell still works at this point, but just get a cup of coffee before shutting down X11)
Offline
I do have another monitor--it isn't connected to the pc. However, I can try and see if that monitor gives a better result than this monitor.
I can also ssh into the system. Next, I tried the xrandr command, but the file was empty, and I didn't hear a beep.
Lastly, I don't know if this is related, but when I have turned off my computer for a longer time, I get the completely black screen again. It is only after I reboot it that I'm greeted with the regular sign in text.
Offline
You do have xrandr installed?
Try the other monitor and also to ssh into the system and to obtain the xrandr information from there interactively.
Offline
I do have xrandr installed, and when I ssh into the system, xrandr returns can't open display.
I tried the other monitor, running xrandr again, and it gives the same response.
Offline
xrandr returns can't open display
1. Is an X11 server running ("ps aux | grep -i xorg")
2. Did you export the DISPLAY?
DISPLAY=:0 xrandr
and this has to be the same user that started the X11 server, otherwise you'll run into xauth issue.
The other monitor behaves the same as the ultrawide one?
Edit: you might wanna revert "nvidia_drm.fbdev=0" resp. set "nvidia_drm.fbdev=1" (though it's default) in case the GSP caused the main problem and deactivating fbdev now causes the subsequent one.
Last edited by seth (2025-05-18 15:57:17)
Offline
There is an X11 server running, but when I tried to export the display, the terminal says can't open display, exactly the same as before. I also made sure to run it as the same user that started the X11 server.
And yeah, the other monitor behaves exactly the same as the ultrawide.
When I reverted the kernel paramaters so that nvidia_drm.fbdev =1, I get the same black screen where I can't get into another VT.
Offline
the terminal says can't open display
What is the *exact* output you get?
Offline
I get
Can't open display :0
Offline
Can you run "xauth list"?
Does "xhost +" allow you to access the display afterwards?
Though typically that just means there's no X11 server listening on that display - what doesn't match the xorg log you posted (the server ran unobstructed and cleanly terminated after 33s)
stat /tmp/.X11-unix/X0
Offline
Of course, I ran "xauth list", which returned nothing.
Then, I ran "xhost +", which gave me
xhost: unable to open display ""
Then, when I tried to stat,
/tmp/.X11-unix/X0
such a file didn't exist, so I ran
stat /tmp/.X11-unix
which gave me
File: /tmp/.X11-unix/
Size: 40 Blocks: 0 IO Block: 4096 directory
Device: 0,32 Inode: 2 Links: 2
Access: (1777/drwxrwxrwt) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2025-05-18 13:41:52.477434834 -0400
Modify: 2025-05-18 12:09:28.361753651 -0400
Change: 2025-05-18 12:09:28.361753651 -0400
Birth: 2025-05-18 12:09:28.361753651 -0400
Offline
ls -l /proc/$(pidof Xorg)/fd
ls -l /tmp/.X11-unix/
Offline
For
ls -l /proc/$(pidof Xorg)/fd
I got
ls: cannot access '/proc//fd': No such file or directory
Then, for
ls -l /tmp/.X11-unix/
I got
total 0
It's just odd because when I run, "ps aux | grep -i xorg", I get
alice 1150 0.0 0.0 6700 4092 pts/0 S+ 17:21 0:00 grep --color=auto -i xorg
Offline
ps aux | grep -i xorg
alice 1150 0.0 0.0 6700 4092 pts/0 S+ 17:21 0:00 grep --color=auto -i xorg
ps aux only listing the grep command itself would indicate no running X server.
Offline
Thank you for letting me know! I didn't look at it closely, and that is my mistake.
Offline