You are not logged in.
Last tested kernel version: 6.15.9-arch1-1 (core/linux package)
DE and Windowing system: Xfce4 on X11
GPU: AMD Radeon RX 580 2048SP 8GB
(* RX 580 2048SP and RX 580 are different.)
Screen: 1600x900 60Hz connected to GPU's HDMI port
Symptom:
On panic, the DRM panic screen "transitions" in with chunks rendered beginning from the top left. However, as seen here, the rendered panic screen turns out to be broken gibberish with chunks of the original desktop display remaining.
There is no time when I can 'salvage' a usable screen: from the moment the rendering starts, it's been producing gibberish.
Is this a fixable issue, or does the problem lie in the kernel implementation itself?
Last edited by nphuracm (2025-08-10 12:00:54)
Offline
That is modern art, you cretin!![]()
There's a mouse cursor in the center of the monitor - does that still move?
Are we sure this is actually a kernel panic? Can you switch the VT? Or reboot w/ sysrq+REISUB, https://wiki.archlinux.org/title/Keyboa … el_(SysRq) ?
Please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General and if you can reboot or switch the VT, also your complete system journal for the boot:
sudo journalctl -b | curl -F 'file=@-' 0x0.st("… -b -1 …" if you had to reboot)
Offline
That is modern art, you cretin!
There's a mouse cursor in the center of the monitor - does that still move?
Are we sure this is actually a kernel panic? Can you switch the VT? Or reboot w/ sysrq+REISUB, https://wiki.archlinux.org/title/Keyboa … el_(SysRq) ?Please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General and if you can reboot or switch the VT, also your complete system journal for the boot:
sudo journalctl -b | curl -F 'file=@-' 0x0.st("… -b -1 …" if you had to reboot)
I am very sure that this is, indeed, a kernel panic - this happens after an Alt-SysRq-C.
(The mouse cursor does not move.)
Offline
Does the same happen from the multi-user.target? (2nd link below, no GUI started)
If not, please post your xorg log and test the behavior w/o the https://wiki.archlinux.org/title/Xfwm#Composite_manager
Offline
That does not happen in multi-user.target.
Xorg logs: https://0x0.st/8F3X.0.log
Issue is reproducible even after disabling the compositor.
Offline
radeon.si_support=0 amdgpu.si_support=1 radeon.cik_support=0 amdgpu.cik_support=1You've a polaris chip anyway…
amd_iommu=on iommu=ptWhy are you enforcing any IOMMU setttings?
Are you passing through any devices to a VM?
[ 29.578] (II) Applying OutputClass "AMDgpu" to /dev/dri/card1Remove all xf86-video-* packages and add "nvidia_drm.modeset=1" to the kernel parameters to also block the simpledrm device.
Offline
I'll be honest, the SI/CIK and IOMMU settings are actually leftover stuff from the past I didn't bother to take care of...
I removed the two xf86-video packages and started the system with nvidia_drm.modeset=1. In the desktop session, Alt-SysRq-C still leads to gibberish, with or without the compositor.
Last edited by nphuracm (2025-08-11 06:53:25)
Offline
[ 40.043] (II) AMDGPU(0): Output HDMI-A-0 using initial mode 1600x900 +0+0
[ 40.042] (II) AMDGPU(0): Modeline "1600x900"x60.0 108.00 1600 1624 1704 1800 900 901 904 1000 +hsync +vsync (60.0 kHz eP)
[ 40.042] (II) AMDGPU(0): Modeline "1920x1080"x60.0 148.50 1920 2008 2052 2200 1080 1084 1089 1125 +hsync +vsync (67.5 kHz e)
[ 40.042] (II) AMDGPU(0): Modeline "1920x1080"x59.9 148.35 1920 2008 2052 2200 1080 1084 1089 1125 +hsync +vsync (67.4 kHz e)
On the console:
# fbset -iDo you run the framebuffer console at the same resolution as X11 ?
Last edited by seth (2025-08-11 13:14:10)
Offline
I suppose you can say so.
mode "1600x900"
geometry 1600 900 1600 900 32
timings 0 0 0 0 0 0 0
rgba 8/16,8/8,8/0,0/0
endmode
Frame buffer device information:
Name : amdgpudrmfb
Address : (nil)
Size : 5763072
Type : PACKED PIXELS
Visual : TRUECOLOR
XPanStep : 1
YPanStep : 1
YWrapStep : 0
LineLength : 6400
Accelerator : NoOffline
Recap
That does not happen in multi-user.target.
DE and Windowing system: Xfce4 on X11
Issue is reproducible even after disabling the compositor.
My gut feeling is that this is more related to the TTY switch than X11 or the "panic", if this is still a thing (sorry for the looooooong delay, btw)
1. are you running X11 on TTY1?
2. What if you don't (but eg. TTY2)
3. if not, what if
3.a: you make sure to have switched to and eg. logged in on TTY1 before triggering the panic
3.b: you do (run X11 on TTY1)?
have you ever tried to make sure that you've switched to
Offline