You are not logged in.
Fast track note:
This did it for me, it's post #9 here
https://bbs.archlinux.org/viewtopic.php … 9#p2297539
---
Hello Arch Friends,
I have gotten myself new hardware (Motherboard, CPU, DDR5) and ran into a login obstacle with SDDM. It is black screen with a cursor.
My display setup has to be abominable unfortunately (at this time and I hope it's only temporary - I'm fighting some stupid BIOS issues with MSI because I can't get the POST, BIOS and volume decryption visuals to be displayed on my discrete graphics card, it's showing these ONLY on the motherboard's I/O backpanel!)
Anyway, I have uploaded my journalctl here -> https://pastes.io/GwuafHe4
From what I can read, the Linux display component (don't know what it's called) created the HDMI display and there's a conflict here since my connected cables are
HDMI from Motherboards Backpanel -> Monitor 1 // This one is coming from an AMD Ryzen 7 9850 X3D. I used to have an Intel i7 6700K but I never connected any video signal cables to the motherboard's backpanel with the Intel one.
DP-1 from GPU -> Monitor 1
DP-2 from GPU -> Monitor 2
And I have not installed any AMD Integrated Graphics drivers yet nor do I intend to use my motherboard's backpanel for my monitor's video signals.
How to fix?
I tried restarting SDDM and unlocking loginctl session. Disk space is not an issue. These were per the Troubleshooting part of SDDM's Wiki (3.1).
Last edited by Big Scorpio (2026-05-09 14:32:56)
Offline
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: failed to acquire GL context to resolve capabilities, using defaults..
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: Adding view for "HDMI-2" QRect(0,0 2560x1440)
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: QRhiGles2: Failed to create temporary context
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: QRhiGles2: Failed to create context
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: Failed to create RHI (backend 2)
May 06 00:19:41 Nexus sddm-greeter-qt6[1038]: Failed to initialize graphics backend for OpenGL.Please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General
Does it work if you only boot the mulit-user.target (2nd link below), get a cup of coffee and then manually progress to the graphical.target?
Offline
Here are my Xorg logs -> https://pastes.io/pDUFlIEq
I crammed them together into a single paste but separated them by
=== Xorg.0.log ===
=== Xorg.1.log ===
=== Xorg.2.log ===Does it work if you only boot the mulit-user.target (2nd link below), get a cup of coffee and then manually progress to the graphical.target?
Same results.
sudo systemctl set-default multi-user.target
reboot
...
sudo systemctl isolate graphical.targetOffline
[ 72.853] (==) Log file: "/var/log/Xorg.0.log", Time: Thu May 7 00:33:52 2026
[ 1189.359] (==) Log file: "/var/log/Xorg.1.log", Time: Tue Mar 24 22:20:52 2026
[ 12329.394] (==) Log file: "/var/log/Xorg.2.log", Time: Mon Oct 13 16:01:31 2025[ 72.866] (--) PCI: (1@0:0:0) 10de:1b06:10de:120f rev 161, Mem @ 0xdd000000/16777216, 0xf800000000/268435456, 0xf810000000/33554432, I/O @ 0x0000f000/128, BIOS @ 0x????????/524288
[ 72.866] (--) PCI:*(120@0:0:0) 1002:13c0:1462:7e49 rev 230, Mem @ 0xf820000000/268435456, 0xdec00000/2097152, 0xdf100000/524288, I/O @ 0x0000e000/256
…
[ 73.068] (II) modeset(0): EDID for output HDMI-2
[ 73.068] (II) modeset(0): Manufacturer: GSM Model: 778a Serial#: 62879
[ 73.068] (II) modeset(0): Year: 2024 Week: 12
[ 73.068] (II) modeset(0): EDID Version: 1.3
[ 73.068] (II) modeset(0): Digital Display Input
…
[ 73.114] (II) NVIDIA(G0): Validated MetaModes:
[ 73.114] (II) NVIDIA(G0): "NULL"
[ 73.115] (II) NVIDIA(G0): Virtual screen size determined to be 640 x 480
[ 73.117] (WW) NVIDIA(G0): Cannot find size of first mode for LG Electronics LG ULTRAGEAR
[ 73.117] (WW) NVIDIA(G0): (DFP-4); cannot compute DPI from LG Electronics LG
[ 73.117] (WW) NVIDIA(G0): ULTRAGEAR (DFP-4)'s EDID.You're running on the AMD GPU, there's only one output attached to the nvidia GPU and there seems a problem w/ its EDID, but
[ 72.866] (II) LoadModule: "glx"
[ 72.867] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[ 72.870] (II) Module glx: vendor="X.Org Foundation"
[ 72.870] (==) Assigned the driver to the xf86ConfigLayout
[ 72.894] (II) Loading sub module "glamoregl"
[ 72.894] (II) LoadModule: "glamoregl"
[ 72.894] (II) Loading /usr/lib/xorg/modules/libglamoregl.so
[ 72.896] (II) Module glamoregl: vendor="X.Org Foundation"
[ 73.066] (II) modeset(0): glamor: Using OpenGL 4.6 context.
[ 73.066] (II) modeset(0): glamor X acceleration enabled on AMD Ryzen 7 9850X3D 8-Core Processor (radeonsi, raphael_mendocino, ACO, DRM 3.64, 6.18.21-1-lts)
[ 73.066] (II) modeset(0): glamor initialized
[ 73.069] (II) Loading sub module "glxserver_nvidia"
[ 73.069] (II) LoadModule: "glxserver_nvidia"
[ 73.069] (II) Loading /usr/lib/nvidia/xorg/libglxserver_nvidia.so
[ 73.086] (II) Module glxserver_nvidia: vendor="NVIDIA Corporation"
[ 73.086] (II) NVIDIA GLX Module 580.142 Tue Mar 3 19:04:44 UTC 2026
[ 73.140] (II) Initializing extension GLX
[ 73.141] (II) Initializing extension GLX
[ 73.141] (II) Indirect GLX disabled.
[ 73.144] (II) AIGLX: Loaded and initialized radeonsi
[ 73.144] (II) GLX: Initialized DRI2 GL provider for screen 0
[ 73.144] (II) Initializing extension NV-GLX
[ 242.316] (II) AIGLX: Suspending AIGLX clients for VT switchGL looks ok?
printenvCan you startx and run glxgears/glxifo -B/eglinfo -B ?
Offline
Currently I have my 2nd monitor connected but shut off as the extra display strains my uncooled (broken liquid cooling) GTX 1080 Ti significantly. I'm waiting for my AMD Radeon 9070 XT to arrive.
Printenv
LC_ADDRESS=en_AG.UTF-8
LC_NAME=en_AG.UTF-8
LC_MONETARY=pt_PT.UTF-8
LC_PAPER=en_AG.UTF-8
LANG=en_US.UTF-8
LC_IDENTIFICATION=en_AG.UTF-8
TERM=linux
LC_TELEPHONE=en_AG.UTF-8
LC_MESSAGES=en_AG.UTF-8
LC_MEASUREMENT=en_AG.UTF-8
LC_CTYPE=en_AG.UTF-8
LC_TIME=en_AG.UTF-8
LC_COLLATE=en_AG.UTF-8
PATH=/usr/local/sbin:/usr/local/bin:/usr/bin
LC_NUMERIC=de_DE.UTF-8
MAIL=/var/mail/root
LOGNAME=root
USER=root
HOME=/root
SHELL=/usr/bin/bash
SUDO_COMMAND=/usr/bin/printenv
SUDO_USER=motherfucker_big_scorpio
SUDO_UID=1000
SUDO_GID=1000
SUDO_HOME=/home/motherfucker_big_scorpio
SUDO_TTY=/dev/tty4
QT_QPA_PLATFORMTHEME=wayland;xcb
__GLX_VENDOR_LIBRARY_NAME=nvidia
__GL_VRR_ALLOWED=1
__GL_NGX_ALLOWED=1
__GL_SYNC_TO_VBLANK=0
LIBVA_DRIVER_NAME=nvidia
QSG_RENDERER_LOOP=threaded
VDPAU_DRIVER=nvidia
PROTON_ENABLE_WAYLAND=1
PROTON_ENABLE_HDR=1glxgears, glxinfo and eglinfo -> https://pastes.io/O5nYvXuD
Reckon I have to install AMD Display drivers for Linux and for my to-be-awaited AMD 9070 XT + the AMD integrated GPU (?) and then (to avoid potential conflicts, that don't have to be) uninstall the nvidia and intel display drivers?
Offline
QT_QPA_PLATFORMTHEME=wayland;xcb
__GLX_VENDOR_LIBRARY_NAME=nvidia
__GL_VRR_ALLOWED=1
__GL_NGX_ALLOWED=1
__GL_SYNC_TO_VBLANK=0
LIBVA_DRIVER_NAME=nvidia
PROTON_ENABLE_HDR=1Where are you setting those?
a) you're driving stuff towards wayland, but SDDM (defaults to and) runs on X11
b) more importantly you're directing GL to use the nvidia libs, but you're running on the amd gpu and you're also not setting up DRI_PRIME
What was the idea here?
Offline
Where are you setting those?
/etc/environment
a) you're driving stuff towards wayland, but SDDM (defaults to and) runs on X11
That's right, I'm sticking to Wayland deliberately. Which one of those Display Managers should I better use for Wayland if not SDDM? Because I don't see any data, that points which of these graphical https://wiki.archlinux.org/title/Display_manager were designed with Wayland being primary in mind.
b) more importantly you're directing GL to use the nvidia libs, but you're running on the amd gpu and you're also not setting up DRI_PRIME
What was the idea here?
The AMD GPU is a temporary solution until I get my hands on my AMD 9070 XT (currently with my new motherboard the nVidia display connectors (be it DP or HDMI) does not display POST, BIOS and decryption video signals from my GTX 1080 Ti)
I see this DRI_PRIME for the first time and had to ask an AI LLM to explain it to me. So according to the AI LLM it sets GPU priority for rendering OpenGL and Vulkan, where 0 usually means integrated GPU and 1 means discrete GPU with the value 0 being the default value.
If I understand it right seth: DRI_PRIME should be 1 if GL and Vulkan should be prioritized on the discrete GPU (currently my GTX 1080 Ti, soon to be my AMD 9070 XT).
Then when I get my new AMD GPU i would have to replace all "nvidia" values with their AMD equivalent. According to the AI LLM I would have to perform
# Force discrete AMD GPU (if hybrid setup)
export DRI_PRIME=1
# Disable VSync
export vblank_mode=0
# Enable Variable Refresh Rate
export AMD_VRR=1
# Enable FreeSync optimizations (frame buffer compression)
export AMD_FBC=1
# Optional: RADV Vulkan driver tuning. 'noemutranslate' reduces overhead, 'ngg' enables next-gen geometry shaders
export RADV_PERFTEST=noemutranslate,ngg,gs,cs
# Optional: Force the radeonsi driver (usually auto-detected)
export MESA_LOADER_DRIVER_OVERRIDE=radeonsi
# Unset NVIDIA specific values
unset __GLX_VENDOR_LIBRARY_NAME
unset __GL_VRR_ALLOWED
unset __GL_NGX_ALLOWED
unset __GL_SYNC_TO_VBLANK
unset LIBVA_DRIVER_NAME
# Test the GPU selection
glxinfo | grep "OpenGL renderer"But adjust these in /etc/environment then for system-wide persistence. Am I on the right track with that?
Offline
https://wiki.archlinux.org/title/SDDM#Wayland
I see this DRI_PRIME for the first time and had to ask an AI LLM to explain it to me.
Ask the AI whether it has read https://wiki.archlinux.org/title/PRIME
currently with my new motherboard the nVidia display connectors (be it DP or HDMI) does not display POST, BIOS and decryption video signals from my GTX 1080 Ti
Does the nvidia GPU provide output later on (when the OS is booting)?
https://wiki.archlinux.org/title/KDE#Me … -specific) ?
Otherwise, don't set DRI_PRIME, but (in any event) don't set the other variables (notably "__GLX_VENDOR_LIBRARY_NAME=nvidia" and "LIBVA_DRIVER_NAME=nvidia") either to run on the AMD GPU.
If you want to run on the AMD GPU but globally use prime, you'll need to export the variables in https://gitlab.archlinux.org/archlinux/ … type=heads but I'd avoid that since you've output attached to the nvidia GPU
Offline
That seems to have worked (?). I did the part with creating the file "/etc/sddm.conf.d/10-wayland.conf" file and it's contents as in the Wiki
[General]
DisplayServer=wayland
GreeterEnvironment=QT_WAYLAND_SHELL_INTEGRATION=layer-shell
[Wayland]
CompositorCommand=kwin_wayland --drm --no-lockscreen --no-global-shortcuts --locale1Then I have re-enabled this line in my /etc/environment (from https://bbs.archlinux.org/viewtopic.php … 5#p2292035 )
KWIN_DRM_DEVICES="/dev/dri/by-path/pci-0000\:01\:00.0-card:/dev/dri/by-path/pci-0000\:00\:02.0-card"And I regained my graphical desktop environment interface.
Since I was unsure which of these it was, I have disabled the KWIN_DRM_DEVICES environment variable and rebooted. I'm still in.
I disabled that environment variable earlier because I'm expecting my new GPU soon (at the time when I took it out of the environment variable). As of the time of writing, I have it here next to me now. Gonna replace my GTX 1080 Ti now.
Ask the AI whether it has read https://wiki.archlinux.org/title/PRIME
Yikes!
For the record, when it comes to Linux administration, I use AI LLMs only as a last resort or when i need help with composing Linux commands syntax, not to offload my technical expertise or "vibe code".
Does the nvidia GPU provide output later on (when the OS is booting)?
https://wiki.archlinux.org/title/KDE#Me … -specific) ?
Yes, it does.
Card: card0
GPU: 0x10de 0x1b06
---
Card: card0-DP-1
---
Card: card0-DP-2
---
Card: card0-DP-3
---
Card: card0-DVI-D-1
---
Card: card0-HDMI-A-1
---
Card: card1
GPU: 0x1002 0x13c0
---
Card: card1-DP-4
---
Card: card1-DP-5
---
Card: card1-DP-6
---
Card: card1-HDMI-A-2
---
Card: card1-Writeback-1
---On Spydows this wasn't ever a problem.
---
I'll leave out the DRI_PRIME, __GLX_VENDOR_LIBRARY_NAME and LIBVA_DRIVER_NAME (and the other vars, that force a specific GPU or driver to use) as per your recommendation and since I have my AMD GPU here now, I'm taking out the nvidia env vars out too. I will need to reenable and adjust my KWIN_DRM_DEVICES env var next after I attach my AMD GPU to it.
Adjacently my new AMD GPU might run into this same problem too seth. If it's not against the rules or out of context, I'll leave this topic open for now and once I am all good with the AMD GPU showing me the SDDM login screen and post login, I'll close this one and put a fast-tracking note on my initial post.
Dankeschön an dich sincerely for the great support seth. Hopefully I won't have to open another topic to get my AMD GPU started properly (I'll refer to the Arch Wiki of course)
Last edited by Big Scorpio (2026-05-09 14:32:25)
Offline
OK, I got the AMD GPU in now and the desktop environment is working. Now to install its drivers, lib32, libva and so on. I might miss something.
This topic is now finished.
Offline