Terminal (and a few other apps) freezes

raphix · 2025-06-11 13:11:04

My laptop has some trouble with some apps (but not all of them): usually when I swicth between some tags (I use AwesomeWM), but also sometimes while the app is focused, there is a freeze. By that I mean the app is unresponsive, and if I focus another app and get back to the freezed one, all is blank (black, or white with striped) inside the window border.

Some apps, such as Firefox, never freeze. Opera does freeze, and my terminal (alacritty, wezterm, at least).

I thought it was related to suspend-wake up (it seems that it's worse after a wake up), but it also happen after a reboot (but clearly reboots improve the behavior).

I also thought it was due to "high pid" number (because when I kill the freezed app, I usually see high pid numbers, above 1 million). However, it sometimes happen when the pid is rather low.

It may also look like a video driver issue, but I don't know how to debug that. I have an HP Elitebook x360 G8, with an Intel chip. I have changed a few things recently by installing some Wine/playonlinux things (ie, lib32), but I think I had reverted to the previous config (I may have missed an offending package, but can't find it).

I am not an expert, but launching a term and running "sudo dmesg" have not yield lines that would seem suspect.

Thank you all for your ideas.

seth · 2025-06-11 20:16:08

Is it purely visual or does the process ignore input, eg. does ctrl+o on opera (still) open a file dilaog (even if that's maybe black as well)
Is the process SIGSTOPped? (in"ps aux" shows up with a "T" state) and can you "pikill SIGCONT opera"?

raphix · 2025-06-13 10:18:19

I have not had another opera freeze to test the Ctrl+o. However, with Alacritty, the process appears as other working terminals:

$ ps aux | grep -i alac
raph      400351  0.0  0.5 1010996 190232 ?      Ssl  11:08   0:03 alacritty -o font.size=18 --class mutt -e tmux new mutt -F ~/.muttrc
raph      404070  0.1  0.6 1002732 221712 ?      Ssl  11:16   0:04 alacritty -o font.size=18
raph      410285  0.0  0.5 1011068 194996 ?      Ssl  11:27   0:01 alacritty -o font.size=18
raph      476407  0.0  0.3 1091468 98792 ?       Ssl  11:45   0:00 alacritty -o font.size=18
raph      484214  0.0  0.0   6480  4152 pts/5    S+   12:04   0:00 grep --color=auto --exclude-dir=.bzr --exclude-dir=CVS --exclude-dir=.git --exclude-dir=.hg --exclude-dir=.svn --exclude-dir=.idea --exclude-dir=.tox --exclude-dir=.venv --exclude-dir=venv -i alac

Terminal with PID 404070 is not responsive, others are.

The freezed app looks like this: https://i.imgur.com/Vi4FWsu.png
(before switching to it, I was on a "white background" app, so mostly white. If I switch from a dark background window, the freezed app appears as dark)

So, I think the app has not only visual problems… but I am not sure.

seth · 2025-06-13 12:50:06

Looks like the digital equivalent of a Jackson Pollock painting - I'm somewhat inclined to make it my wallpaper

So, I think the app has not only visual problems… but I am not sure.

Since it's a terminal, likely running a shell, should be easy enough to figure, enter sth. like "notify-send itlives" - the process isn't stopped nor dormant, have you tried to resize the window?
What will hopefully make alacritty resize the GL context, cause this looks like you lost that - itr: do you get GPU resets in your journal/dmesg when this happens?

raphix · 2025-06-18 21:20:40

I have tried your idea of resizing the window: it resizes, but the content of the terminal does not comes back. I just get a smaller Pollock ;-)
It also does not work to send a command, I don't get a notification from the freezed term.
So, it looks more than just a "lost GL context".

As mentioned above, it looks like it's related to power management: it appears more often (only?) when the laptop is unplugged (on battery), and more ofter after a suspend2ram/wake up than on a freshly boot system.

In the journal or dmesg, I don't get suspicious lines. I got one related to rtkit, I have a regular inoffensive bluetooth message (bluetoothd[672]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down), but nothing clear.

seth · 2025-06-19 15:51:37

So, it looks more than just a "lost GL context".

Resizing the window would typically resize the context and usually force a regeneration. Hence the test.
(Though intel might be taking some shortcuts here and also X11 on the modesetting driver would render everything via glamor by default anyway)
Maybe we should take a look at the Xorg log, https://wiki.archlinux.org/title/Xorg#General …

raphix · 2025-06-23 13:38:46

Well, the log has a lot of Logitech noise (I use a logitech usb mouse), and a probably unrelated keymap error.

Here is the Xorg log after a freeze that just happened, while plugged.

$ tail -f ~/.local/share/xorg/Xorg.0.log | ts                                                                                                 15:31 250623

juin 23 15:31:40 [253837.672] (II) libinput: Logitech USB Receiver: Step value 0 was provided, libinput Fallback acceleration function is used.
juin 23 15:31:40 [253837.672] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:14.0/usb3/3-5/3-5.1/3-5.1.3/3-5.1.3:1.3/0003:046D:C548.009D/input/input165/event30"
juin 23 15:31:40 [253837.672] (II) XINPUT: Adding extended input device "Logitech USB Receiver" (type: TOUCHPAD, id 30)
juin 23 15:31:40 [253837.674] (**) Option "AccelerationScheme" "none"
juin 23 15:31:40 [253837.674] (**) Logitech USB Receiver: (accel) selected scheme none/0
juin 23 15:31:40 [253837.674] (**) Logitech USB Receiver: (accel) acceleration factor: 2.000
juin 23 15:31:40 [253837.674] (**) Logitech USB Receiver: (accel) acceleration threshold: 4
juin 23 15:31:40 [253837.676] (II) event30 - Logitech USB Receiver: is tagged by udev as: Touchpad
juin 23 15:31:40 [253837.677] (II) event30 - Logitech USB Receiver: device is a touchpad
juin 23 15:31:40 [254523.939] (EE) Error loading keymap /tmp/server-0.xkm

Repeating the command 3 minutes later, and not just after freeze, I see the same messages…

$ tail -n30 -f ~/.local/share/xorg/Xorg.0.log | ts                                                                                            15:34 250623

juin 23 15:34:20 [253837.535] (**) Option "xkb_model" "pc105"
juin 23 15:34:20 [253837.535] (**) Option "xkb_layout" "fr"
juin 23 15:34:20 [253837.535] (**) Option "xkb_variant" "oss"
juin 23 15:34:20 [253837.535] (**) Option "xkb_options" "caps:escape"
juin 23 15:34:20 [253837.536] (II) event22 - Ortholinear Keyboards The Preonic Keyboard: is tagged by udev as: Keyboard
juin 23 15:34:20 [253837.537] (II) event22 - Ortholinear Keyboards The Preonic Keyboard: device is a keyboard
juin 23 15:34:20 [253837.618] (II) config/udev: Adding input device Logitech USB Receiver (/dev/input/event30)
juin 23 15:34:20 [253837.618] (**) Logitech USB Receiver: Applying InputClass "libinput touchpad catchall"
juin 23 15:34:20 [253837.618] (**) Logitech USB Receiver: Applying InputClass "touchpad"
juin 23 15:34:20 [253837.618] (II) Using input driver 'libinput' for 'Logitech USB Receiver'
juin 23 15:34:20 [253837.669] (II) systemd-logind: got fd for /dev/input/event30 13:94 fd 97 paused 0
juin 23 15:34:20 [253837.669] (**) Logitech USB Receiver: always reports core events
juin 23 15:34:20 [253837.669] (**) Option "Device" "/dev/input/event30"
juin 23 15:34:20 [253837.671] (II) event30 - Logitech USB Receiver: is tagged by udev as: Touchpad
juin 23 15:34:20 [253837.672] (II) event30 - Logitech USB Receiver: device is a touchpad
juin 23 15:34:20 [253837.672] (II) event30 - Logitech USB Receiver: device removed
juin 23 15:34:20 [253837.672] (**) Option "Tapping" "on"
juin 23 15:34:20 [253837.672] (**) Option "TappingButtonMap" "lmr"
juin 23 15:34:20 [253837.672] (II) libinput: Logitech USB Receiver: Step value 0 was provided, libinput Fallback acceleration function is used.
juin 23 15:34:20 [253837.672] (II) libinput: Logitech USB Receiver: Step value 0 was provided, libinput Fallback acceleration function is used.
juin 23 15:34:20 [253837.672] (II) libinput: Logitech USB Receiver: Step value 0 was provided, libinput Fallback acceleration function is used.
juin 23 15:34:20 [253837.672] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:14.0/usb3/3-5/3-5.1/3-5.1.3/3-5.1.3:1.3/0003:046D:C548.009D/input/input165/event30"
juin 23 15:34:20 [253837.672] (II) XINPUT: Adding extended input device "Logitech USB Receiver" (type: TOUCHPAD, id 30)
juin 23 15:34:20 [253837.674] (**) Option "AccelerationScheme" "none"
juin 23 15:34:20 [253837.674] (**) Logitech USB Receiver: (accel) selected scheme none/0
juin 23 15:34:20 [253837.674] (**) Logitech USB Receiver: (accel) acceleration factor: 2.000
juin 23 15:34:20 [253837.674] (**) Logitech USB Receiver: (accel) acceleration threshold: 4
juin 23 15:34:20 [253837.676] (II) event30 - Logitech USB Receiver: is tagged by udev as: Touchpad
juin 23 15:34:20 [253837.677] (II) event30 - Logitech USB Receiver: device is a touchpad
juin 23 15:34:20 [254523.939] (EE) Error loading keymap /tmp/server-0.xkm

BTW, I discovered and narrowed that my laptop currently refuses to suspend when I press the power button AND it's plugged, while it goes to suspend state without problem when unplugged. As mentioned above, the power management seems to have issue on this laptop.

I don't want to divert the convo on this other topic, but perhaps I should check the suspend log? How can I check the differences between the two situations?

Last edited by raphix (2025-06-23 13:39:48)

seth · 2025-06-23 16:26:41

I'll put on a big boy editor and handle the entire log - noise be damned.
These excerpts don't really help a lot to identify pot. problems.

I discovered and narrowed that my laptop currently refuses to suspend when I press the power button AND it's plugged

/etc/systemd/logind.conf doesn't discriminate the power key by the docking status, maybe some session daemon or acpid is handling this?
systemd/logind records these events in the system journal.

raphix · 2025-07-16 14:52:11

Well, sorry for the late reply.

Here is a log from "sudo journalctl -xe | curl -F 'file=@-' 0x0.st" after two operations:

at "juil. 16 16:40:18", I pressed the power button while the laptop is plugged. The laptop did not suspend: the screen goes black for approximately 2-3 seconds and then everything is back to normal (around 16:40:22-25)
at "juil. 16 16:43:04", I pressed the power button while the laptop was not plugged. The laptop did go to sleep, with the fans desactivated.

I don't see what service could be the culprit here, it would help me a lot if you have an idea!
Thanks again!

seth · 2025-07-16 18:45:55

juil. 16 16:40:22 intersect touchegg[710]: libinput error: event2  - Power Button: client bug: event processing lagging behind by 322ms, your system is too slow
juil. 16 16:40:22 intersect touchegg[710]: libinput error: event3  - Power Button: client bug: event processing lagging behind by 316ms, your system is too slow

yells at me which seems to be

juil. 16 16:31:27 intersect touchegg[710]: Compatible device detected:
juil. 16 16:31:27 intersect touchegg[710]:         Name: Logitech USB Receiver
juil. 16 16:31:27 intersect touchegg[710]:         Size: 115.625mm x 74.4167mm
juil. 16 16:31:27 intersect touchegg[710]:         Calculating start_threshold and finish_threshold. You can tune this values in your service file
juil. 16 16:31:27 intersect touchegg[710]:         start_threshold: 29.2979
juil. 16 16:31:27 intersect touchegg[710]:         finish_threshold_horizontal: 409.695
juil. 16 16:31:27 intersect touchegg[710]:         finish_threshold_vertical: 263.681

so try w//o that device and service.

Also please *never* use -x w/ journalctl, it's flat-out spam.

raphix · 2025-07-17 07:53:06

OK, first, thank you very much for finding the offender! I had checked while plugged to a hub, and did not check while just being plugged to a power strip without this USB receiver. I take note for the warning on journalctl.

For future reference, I was able to find this reddit thread providing a solution, to keep using the receiver but prevent it to wake the computer from s2idle: https://www.reddit.com/r/logitech/comme … _on_linux/

$ sudo vim /etc/udev/rules.d/99-logi-bolt-wakeup.rules
# Disable wakeup for Logitech Logi Bolt Receiver
SUBSYSTEM=="usb", ATTR{idVendor}=="046d", ATTR{idProduct}=="c548", ATTR{power/wakeup}="disabled"
# close vim
$ sudo udevadm control --reload-rules
$ sudo udevadm trigger

This part of my problem is solved. The other, related to apps freezing when the laptop is unplugged, may or may not be related. I'll post when I have more tests/logs to find the problem. The main app freezing is alacritty, I have reverted to urxvt for now.

Last edited by raphix (2025-07-17 07:54:04)

Arch Linux

#1 2025-06-11 13:11:04

Terminal (and a few other apps) freezes

#2 2025-06-11 20:16:08

Re: Terminal (and a few other apps) freezes

#3 2025-06-13 10:18:19

Re: Terminal (and a few other apps) freezes

#4 2025-06-13 12:50:06

Re: Terminal (and a few other apps) freezes

#5 2025-06-18 21:20:40

Re: Terminal (and a few other apps) freezes

#6 2025-06-19 15:51:37

Re: Terminal (and a few other apps) freezes

#7 2025-06-23 13:38:46

Re: Terminal (and a few other apps) freezes

#8 2025-06-23 16:26:41

Re: Terminal (and a few other apps) freezes

#9 2025-07-16 14:52:11

Re: Terminal (and a few other apps) freezes

#10 2025-07-16 18:45:55

Re: Terminal (and a few other apps) freezes

#11 2025-07-17 07:53:06

Re: Terminal (and a few other apps) freezes

Board footer