MSI GP72M 7REX freezes (sometimes when using apt)

[gabriele2000]

How did you upgrade to 21.10? (Fresh install / Upgrade) Upgraded from 21.04 to 21.10 (a lot of manual labor though, still thanks to popupgrade...)

Issue/Bug Description: https://drive.google.com/file/d/1rqbcV-ATMoy6L1pLbLZOa678kwesk8bk/view?usp=sharing (also enjoy the music, as proof that the system works underneath and enjoy the "command spam" that I did while waiting)

This video is the definitive proof, since I cannot even start to describe how frustrating is it, how strange it is and how impossible is to get some sort of logging of what's causing it (though I guess it's gnome-related)

Steps to reproduce (if you know): I found that sometimes it happens if I use "apt"... something like sudo apt upgrade && sudo apt full-upgrade takes forever to process, as you can see (it's stuck between the two commands) and I think it's also related to the issue... though the freeze is also present outside these commands...

Expected behavior: Smooth as butter, just like one month ago

Other Notes: For once, I don't blame Nvidia, because with the Intel GPU the random freezes happens too...

[jabbermacy]

do fresh install; also, what do you mean 'manual labor'? My upgrade took >1 hour but flawless

[gabriele2000]

do fresh install

Not gonna happen, and I don't even want to waste time by specifying everything, over and over again

manual labor

I installed it a month ago or even more... before the beta so I had to do some "manual stuff" such as manual repository change... Again, that's NOT the problem, maybe it's a software-bug, maybe it's a hardware-bug, still I don't know what to search... dmesg doesn't give me any useful input. I read online that this kind of problem could be releated to SSD stall, but it's very strange since both are well functioning, S.M.A.R.T. values are good

[jacobgkau]

It looks like you are using a lot of non-default GNOME extensions. It looks like you may have COSMIC Workspaces disabled (I say "may" because you still have vertical workspaces), it also looks like something is removing your Applications menu, and you have a lot of extra items in the top right of the screen.

If you create a new user account and leave the extensions as default on that one, are you able to replicate the freezes?

The point where your apt update is stuck at would be the point where the system is reading the package lists. What are the hardware specifications of your system?

how strange it is and how impossible is to get some sort of logging of what's causing it

If you open a terminal and run sudo journalctl --follow, do you see any new log lines come up during or after the freeze?

Not gonna happen, and I don't even want to waste time by specifying everything, over and over again

You have various third-party PPAs added, and packages from those can also decrease the stability of your system. Especially when using beta projects (I see at least one beta PPA for something other than Pop), you should always be prepared to start over if one of the components causes problems. Does this issue happen if you boot from a live disk or Recovery mode and use it for a while?

[gabriele2000]

I'll update this comment while I test things...

It looks like you are using a lot of non-default GNOME extensions. It looks like you may have COSMIC Workspaces disabled (I say "may" because you still have vertical workspaces), it also looks like something is removing your Applications menu, and you have a lot of extra items in the top right of the screen.

What are the hardware specifications of your system?

do you see any new log lines come up during or after the freeze

Error.zip

[jabbermacy]

man you have a LOT of packages

[gabriele2000]

man you have a LOT of packages

Never gave me this issue, or any issue at all

[SUPERCILEX]

I've had the same issue, not sure if it's still happening though. Are there any other logs other than journalctl I can look for if it happens again?

[gabriele2000]

logs.zip

Some more info, just to be sure... @jacobgkau

[TS-CUBED]

These freezes don't seem to be limited to a specific device (so the new title may be misleading). Looking at @gabrielle2000 logs, the only thing out of the ordinary in there seem to be the gnome-shell errors that are present in other reports on random freezes. These sometimes seem i/o related, sometimes graphics related (gnome-shell animations), but I always find the gnome-shell/cosmic-dock error messages in the journal afterwards and with exactly the right timestamps.

So these may be related with pop-os/cosmic-dock#119

The freezes are a lot less common when I disable the cosmic-dock extension.

I've seen this on 4 different devices (ZBook 14 G1, ZBook 15 G2, Zbook 17 G2, and a VMware virtual machine), with different graphics drivers: Intel, Nouveau, Nvidia 390, 418 and 470, and with the VMware driver.

The long freezes always have something like this:

Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d194dda70 StButton.show-apps:first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d194dec10 Gjs_ui_iconGrid_BaseIcon.overview-icon:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19504690 StBoxLayout:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19505830 StBin:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19649790 Gjs_cosmic-dock_system76_com_appIcons_DockShowAppsIcon.dash-item-container] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d194dda70 StButton.show-apps:first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d194dec10 Gjs_ui_iconGrid_BaseIcon.overview-icon:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19504690 StBoxLayout:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19505830 StBin:insensitive first-child last-child] which is not in the stage.
Nov 15 13:29:45 z-book-17 gnome-shell[3778]: st_widget_get_theme_node called on the widget [0x555d19654a60 StIcon.show-apps-icon:insensitive first-child last-child] which is not in the stage.

while the shorted freezes that occur less frequently when the cosmic-dock extension is disabled (disabled extensio, not merely switching the dock off in the settings!) have some of these:

Nov 22 10:52:44 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaWindowGroup is on because it needs an allocation.
Nov 22 10:52:44 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaWindowActorX11 is on because it needs an allocation.
Nov 22 10:52:44 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaSurfaceActorX11 is on because it needs an allocation.
Nov 22 10:52:44 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaWindowActorX11 is on because it needs an allocation.
Nov 22 10:52:44 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaSurfaceActorX11 is on because it needs an allocation.
Nov 22 10:52:48 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaWindowGroup is on because it needs an allocation.
Nov 22 10:52:48 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaWindowActorX11 is on because it needs an allocation.
Nov 22 10:52:48 z-book-14 gnome-shell[2816]: Can't update stage views actor MetaSurfaceActorX11 is on because it needs an allocation.

[SUPERCILEX]

This started happening for me again, and a reboot seems to fix it. That makes me think it has something to do with suspending.

[gabriele2000]

It seems that when disconnected from the internet, the problem won't happen, as far as I've seen... The whole PC is just more responsive overall.