You are not logged in.
My booth screens often freeze for few seconds since the last upgrade and I don't know why. My 2nd external screen also completely freezes until I restart it with xrandr, but if I change its brightness or gamma with xrandr it changes, only the rendering stops.
The freezing is more intense when I'm playing 3D games and even more with one wine app that uses Vulkan.
WM: i3
compositor: picom
GPUs:
NVIDIA GeForce GTX 1650 Mobile / Max-Q
Intel CoffeeLake-H GT2 [UHD Graphics 630]
Also this is what I found in dmesg https://pastebin.com/nNe2Y6fn it doesn't seem right.
I noticed that my Intel GPU usage drops to 0 when it freezes.
Last edited by Wolfyxon (2023-02-23 18:03:55)
Offline
if I change its brightness or gamma with xrandr it changes, only the rendering stops.
compositor: picom
Next time it freezes, kill picom and see whether that "unfreezes" the output.
Other than that please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General
Offline
if I change its brightness or gamma with xrandr it changes, only the rendering stops.
compositor: picom
Next time it freezes, kill picom and see whether that "unfreezes" the output.
Other than that please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General
I tried already, picom isn't causing it.
Here's my xorg log https://pastebin.com/UNeTR2yM (Asus is my external monitor)
Last edited by Wolfyxon (2023-02-23 07:20:25)
Offline
I think this might be an issue with my GPU drivers but not sure
Offline
[ 6.111] (==) Using config file: "/etc/X11/xorg.conf"
Get rid of that file, it's nvidia-settings generated pointless cruft.
It's a hybrid system, the ASUS VY279 is attached to the nvidia GPU, so you're at https://wiki.archlinux.org/title/PRIME#Reverse_PRIME and that is what you want the config file to look like.
Enable https://wiki.archlinux.org/title/NVIDIA … de_setting in doubt make sure to DISable https://wiki.archlinux.org/title/PRIME# … ronization and if the issues remain, post an updated xorg log and the entire system journal
sudo journalctl -b | curl -F 'file=@-' 0x0.st
Offline
until I restart it with xrandr
This also works if I change the mode. For now I made a script that changes the mode to lower then goes back.
#!/bin/bash
pkill -f polybar
xrandr --output HDMI-1-0 --mode 1680x1050
sleep 2
xrandr --output HDMI-1-0 --auto --left-of eDP-1
~/.local/bin/polybar & disown
Offline
[ 6.111] (==) Using config file: "/etc/X11/xorg.conf"
Get rid of that file, it's nvidia-settings generated pointless cruft.
It's a hybrid system, the ASUS VY279 is attached to the nvidia GPU, so you're at https://wiki.archlinux.org/title/PRIME#Reverse_PRIME and that is what you want the config file to look like.
Enable https://wiki.archlinux.org/title/NVIDIA … de_setting in doubt make sure to DISable https://wiki.archlinux.org/title/PRIME# … ronization and if the issues remain, post an updated xorg log and the entire system journalsudo journalctl -b | curl -F 'file=@-' 0x0.st
Thanks, I'll let you know if anything changes.
Last edited by Wolfyxon (2023-02-23 08:19:36)
Offline
Still nothing.
Journal: http://0x0.st/HsWM.txt
Xorg log: https://pastebin.com/p8yVGCGh
xorg config
Section "ServerLayout"
Identifier "layout"
Screen 0 "intel"
Inactive "nvidia"
Option "AllowNVIDIAGPUScreens"
EndSection
Section "Device"
Identifier "nvidia"
Driver "nvidia"
EndSection
Section "Screen"
Identifier "nvidia"
Device "nvidia"
EndSection
Section "Device"
Identifier "intel"
Driver "modesetting"
BusID "PCI:0:2:0"
EndSection
Section "Screen"
Identifier "intel"
Device "intel"
Last edited by Wolfyxon (2023-02-23 08:41:45)
Offline
nvidia_drm.modset=1
"modeset"
Ftr, that's not a valid xorg config (misses EndSection, but probably a cnp error when posting)
~/.local/bin/polybar & disown
Why are you restarting polybar around that?
You'll proabably also get away w/ only juggling the refresh rate?
When the output is "frozen", does that extend to the cursor as well?
(Ie. how exactly do you determine the frozen-ness, given the gamma changes are still applied?)
Offline
nvidia_drm.modset=1
"modeset"
Ftr, that's not a valid xorg config (misses EndSection, but probably a cnp error when posting)~/.local/bin/polybar & disown
Why are you restarting polybar around that?
You'll proabably also get away w/ only juggling the refresh rate?When the output is "frozen", does that extend to the cursor as well?
(Ie. how exactly do you determine the frozen-ness, given the gamma changes are still applied?)
I set the flag and Xorg was unable to start. I tried running it manually but it also didn't work.
Log: https://pastebin.com/jRu57jbi (oh I've just fully read it and I see that I just had to run it as root, but it should start automatically anyway)
It started after I reverted the parameter.
When the output is "frozen", does that extend to the cursor as well?
Yes. It becomes completely frozen.
Why are you restarting polybar around that?
So it shows up again on my external monitor, that's a script that runs polybars for each monitor.
Last edited by Wolfyxon (2023-02-23 14:36:11)
Offline
[ 159.588] (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied)
Post your xinitrc and see the last link below, in doubt ensure to re-use the TTY w/ the xserverrc.
Offline
[ 159.588] (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied)
Post your xinitrc and see the last link below, in doubt ensure to re-use the TTY w/ the xserverrc.
Uh seems like I don't even have it
Offline
You don't have what?
How do you start the session? startx/xinit? Some DM?
Offline
You don't have what?
How do you start the session? startx/xinit? Some DM?
Via lightdm
Offline
https://wiki.archlinux.org/title/LightD … TTY_output
But lightdm won't log X11 into your $HOME …
loginctl session-status
ps fax
Offline
Might be related to https://bbs.archlinux.org/viewtopic.php?id=282669
I also see the same errors in journalctl.
(The kernel mentioned here is newer than the one I had before upgrading, I forgot to do so for few months)
Last edited by Wolfyxon (2023-02-23 15:13:21)
Offline
https://wiki.archlinux.org/title/LightD … TTY_output
But lightdm won't log X11 into your $HOME …loginctl session-status ps fax
ps fax:
617 ? SLsl 0:00 /usr/bin/lightdm
264931 tty7 Ssl+ 4:24 \_ /usr/lib/Xorg :0 -seat seat0 -auth /run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch
265329 ? Sl 0:00 \_ lightdm --session-child 14 22
276489 ? Ss 0:01 \_ i3
I don't see Xorg in loginctl session-status
Offline
Might be related to https://bbs.archlinux.org/viewtopic.php?id=282669
"Idling display engine timed out" wasn't in the journal you posted, though?
I don't see Xorg in loginctl session-status
That'd be normal for lightdm, but then /home/wolfyxon/.local/share/xorg/Xorg.0.log becomes irrelevant.
Offline
"Idling display engine timed out" wasn't in the journal you posted, though?
It is
Feb 23 15:39:03 WolfArch kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:0:0:1128
Feb 23 15:39:08 WolfArch kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:0:0:1128
Feb 23 15:39:20 WolfArch kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:0:0:1128
Feb 23 15:39:22 WolfArch kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:0:0:1128
Offline
I was talking about the one you posted:
https://0x0.st/HsWM.txt
The error is probably linked to the modeset=1 parameter.
But yeah, looks the same.
Since https://wiki.archlinux.org/title/PRIME# … ronization depends on modeset=1 and keeps flaring up, I'd look into that direction.
Offline
I was talking about the one you posted:
That's weird. It seems to have appeared.
Hmm I added intel_idle.max_cstate=1 to my kernel params and now it's not freezing randomly (even with that one app I was talking about) and my 2nd monitor is fine. It only freezes when there's a heavy task being processed (hopefully this is normal but it's rare).
I'll do some deeper testing to make sure it's fixed.
Last edited by Wolfyxon (2023-02-23 16:09:09)
Offline
This seems to be fixed now. Thanks for your time.
For those who found this post:
TESTING: So basically first I pressed E when hovering on "Arch Linux" in grub to enter the edit mode, then to the "linux" line I added intel_idle.max_cstate=1 to test the parameter.
APPLYING: It worked so I edited /etc/default/grub and added intel_idle.max_cstate=1 to GRUB_CMDLINE_LINUX_DEFAULT and ran sudo grub-mkconfig -o /boot/grub/grub.cfg to apply it.
Offline
This seems to be fixed now. Thanks for your time.
For those who found this post:
TESTING: So basically first I pressed E when hovering on "Arch Linux" in grub to enter the edit mode, then to the "linux" line I added intel_idle.max_cstate=1 to test the parameter.
APPLYING: It worked so I edited /etc/default/grub and added intel_idle.max_cstate=1 to GRUB_CMDLINE_LINUX_DEFAULT and ran sudo grub-mkconfig -o /boot/grub/grub.cfg to apply it.
It just made the freezing much rarer. It sadly still happens.
Last edited by Wolfyxon (2023-02-23 18:06:15)
Offline
Since https://wiki.archlinux.org/title/PRIME# … ronization depends on modeset=1 and keeps flaring up, I'd look into that direction.
Offline
I tried everything you said with no effects
Forgot to mention that the freeze also stops all input.
Offline