gnome-shell hardlocks under load

Bug #1850615 reported by Michał Sawicz
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
gnome-shell (Ubuntu)
Incomplete
Undecided
Unassigned

Bug Description

With the upgrade to eoan, my GNOME and Ubuntu sessions started locking up hard when under load (i.e. when I start my session and start a bunch of applications).

I couldn't find anything special in gdb, whether it's stuck or not, the stack looks mostly the same:

(gdb) info threads
  Id Target Id Frame
* 1 Thread 0x7f7439f6bcc0 (LWP 26352) "gnome-shell" 0x00007f743f7f5c2f in __GI___poll (fds=0x7ffc1a361ba8, nfds=1, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
  2 Thread 0x7f7439f68700 (LWP 26354) "gmain" 0x00007f743f7f5c2f in __GI___poll (fds=0x556aeb3369e0, nfds=2, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
  3 Thread 0x7f7433fff700 (LWP 26356) "gdbus" 0x00007f743f7f5c2f in __GI___poll (fds=0x7f73fc002070, nfds=4, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
  4 Thread 0x7f74337fe700 (LWP 26357) "dconf worker" 0x00007f743f7f5c2f in __GI___poll (fds=0x556aeb377c50, nfds=1, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
  5 Thread 0x7f7432336700 (LWP 26358) "gnome-s:disk$0" futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x556aeb339d28) at ../sysdeps/unix/sysv/linux/futex-internal.h:80
  6 Thread 0x7f74311a7700 (LWP 26384) "JS Helper" futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x556aeb334468) at ../sysdeps/unix/sysv/linux/futex-internal.h:80
  7 Thread 0x7f7430fa6700 (LWP 26385) "JS Helper" futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x556aeb334468) at ../sysdeps/unix/sysv/linux/futex-internal.h:80
  8 Thread 0x7f7430da5700 (LWP 26386) "JS Helper" futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x556aeb334468) at ../sysdeps/unix/sysv/linux/futex-internal.h:80
  9 Thread 0x7f7430ba4700 (LWP 26387) "JS Helper" futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x556aeb33446c) at ../sysdeps/unix/sysv/linux/futex-internal.h:80
  10 Thread 0x7f74082d4700 (LWP 26696) "threaded-ml" 0x00007f743f7f5c2f in __GI___poll (fds=0x7f73f4002910, nfds=3, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
  11 Thread 0x7f73d3fff700 (LWP 2290) "threaded-ml" 0x00007f743f7f5c2f in __GI___poll (fds=0x7f73cc003130, nfds=3, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29

ProblemType: Bug
DistroRelease: Ubuntu 19.10
Package: gnome-shell 3.34.1-1ubuntu1
ProcVersionSignature: Ubuntu 5.3.0-19.20-generic 5.3.1
Uname: Linux 5.3.0-19-generic x86_64
NonfreeKernelModules: zfs zunicode zavl icp zcommon znvpair
ApportVersion: 2.20.11-0ubuntu8
Architecture: amd64
Date: Wed Oct 30 10:46:53 2019
DisplayManager: gdm3
RelatedPackageVersions: mutter-common 3.34.1-1ubuntu1
SourcePackage: gnome-shell
UpgradeStatus: Upgraded to eoan on 2019-10-28 (1 days ago)
---
ProblemType: Bug
ApportVersion: 2.20.11-0ubuntu8.2
Architecture: amd64
CurrentDesktop: GNOME
DisplayManager: gdm3
DistroRelease: Ubuntu 19.10
NonfreeKernelModules: zfs zunicode zavl icp zcommon znvpair
Package: gnome-shell 3.34.1-1ubuntu1
PackageArchitecture: amd64
ProcVersionSignature: Ubuntu 5.3.0-19.20-generic 5.3.1
RelatedPackageVersions: mutter-common 3.34.1-1ubuntu1
Tags: eoan
Uname: Linux 5.3.0-19-generic x86_64
UpgradeStatus: Upgraded to eoan on 2019-10-28 (9 days ago)
UserGroups: adm cdrom dialout dip docker kvm libvirt lp lpadmin lxd plugdev sambashare sudo vboxusers wireshark
_MarkForUpload: True

Revision history for this message
Michał Sawicz (saviq) wrote :
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This sounds suspicious:
5 Thread 0x7f7432336700 (LWP 26358) "gnome-s:disk$0"

Separately, you seem to have a number of extensions installed:

'<email address hidden>'
'<email address hidden>'
'<email address hidden>'
'<email address hidden>'
'<email address hidden>'
'<email address hidden>'
'cpufreq@konkor'
'<email address hidden>'
'<email address hidden>'

Although I mostly suspect the last one, please uninstall all of them and see if that solves the problem. You will need to uninstall them all and not just disable them because extensions can still interfere when installed and disabled. We need to do this because so many bugs are caused by extensions. Eliminating them very often solves the problem.

Changed in gnome-shell (Ubuntu):
status: New → Incomplete
Revision history for this message
Michał Sawicz (saviq) wrote : Dependencies.txt

apport information

tags: added: apport-collected
description: updated
Revision history for this message
Michał Sawicz (saviq) wrote : GsettingsChanges.txt

apport information

Revision history for this message
Michał Sawicz (saviq) wrote : ProcCpuinfoMinimal.txt

apport information

Revision history for this message
Michał Sawicz (saviq) wrote : ProcEnviron.txt

apport information

Revision history for this message
Michał Sawicz (saviq) wrote : ShellJournal.txt

apport information

Revision history for this message
Michał Sawicz (saviq) wrote : monitors.xml.txt

apport information

Revision history for this message
Michał Sawicz (saviq) wrote :

I just had it locked up a number of times with all the extensions disabled. Trace looked basically the same. It seems to be launching Thunderbird (it takes a while due to calendars…) that's triggering this most often.

Revision history for this message
Michał Sawicz (saviq) wrote :

Here's a gdb log, it hung again with all extensions disabled, under heavier IO and load.

Revision history for this message
Michał Sawicz (saviq) wrote :

I've had this happen a number of times again on a fresh 19.10 install, with no extra extensions enabled.

It seems to happen when an app isn't responding for a handful of seconds and trying to switch between apps via Alt+Tab or between workspaces. A SIGSTOP would probably be enough to simulate the app hanging.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Per comment #2 please remember to uninstall all extensions. Just disabling them is not enough to stop buggy extensions from interfering :(

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Though comment #11 sounds like maybe bug 1845302 which is now fixed. So is this bug still happening?

Revision history for this message
Michał Sawicz (saviq) wrote :

I've not seen this in a while indeed. Marking dupe, then :)

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.