apport-gtk crashed with SIGSEGV in gtk_widget_translate_coordinates()

[Eliah Kagan]

On an Xubuntu Artful amd64 system with linux-generic 4.11.0.12.12, libgtk-3-0 3.22.17-0ubuntu1, and apport-gtk 2.20.6-0ubuntu4, I viewed the details on a kernel crash report prepared by apport-gtk titled "NMI watchdog: BUG: soft lockup - CPU#0 stuck for 24s! [kworker/0:0:2661]" but decided not to report that bug. This is a VMware Workstation 12.5.7 virtual machine that I believe incurs slowdowns from the host from time to time. After I successfully quit the Apport dialog box that would allow me to report the bug, an unexpected smaller window appeared with no contents and "Ubuntu" (without the quotes) as its title. I waited several minutes but no window contents appeared.

This all actually occurred twice at a time -- apparently there were two kernel soft locks at different times, and Apport offered to report bugs relating to both events. I clicked through to view the details on both. I quit both dialog boxes and got two smaller windows with no contents; these windows never changed. I did not take a screenshot of the windows, but I did use xwininfo to obtain the window ID of one of them and then the xprop command to obtain information about it, following the technique described at https://unix.stackexchange.com/a/84981. The output of "xprop -id 0x42011b1" is attached as xprop.txt.

Then I closed the two windows by clicking the X. My window manager told me one of them was not responding and asked if I wanted to force it to close, which I did. After this, at least one of the two instances of apport-gtk apparently segfaulted; another apport window, which appeared to work correctly, eventually came up to help me report the crash as a bug, which I am doing now. I don't know for sure that the crash is actually from when I forced the window to close, or from before that; it could even have happened before the two mysterious windows were created. Furthermore, I don't know if the window I ran xprop to get information about came from the same instance of apport-gtk that crashed, or from the other instance.

I don't know if this bug is in apport or some GTK+ package such as libgtk-3-0. It resembles bug 1539995, which is considered to be a GTK+ bug rather than an apport bug and, as of this writing, appears (along with its duplicates) not known to affect any releases except Xenial and Yakkety. (Perhaps the forthcoming symbolic retrace will shed light on the similarity.)

ProblemType: Crash
DistroRelease: Ubuntu 17.10
Package: apport-gtk 2.20.6-0ubuntu4
ProcVersionSignature: Ubuntu 4.11.0-12.18-generic 4.11.12
Uname: Linux 4.11.0-12-generic x86_64
ApportVersion: 2.20.6-0ubuntu4
Architecture: amd64
Date: Tue Aug 1 02:34:38 2017
ExecutablePath: /usr/share/apport/apport-gtk
InstallationDate: Installed on 2017-07-31 (0 days ago)
InstallationMedia: Xubuntu 17.10 "Artful Aardvark" - Alpha amd64 (20170731)
InterpreterPath: /usr/bin/python3.6
PackageArchitecture: all
ProcCmdline: /usr/bin/python3 /usr/share/apport/apport-gtk
ProcEnviron:
 SHELL=/bin/bash
 LANG=en_US.UTF-8
 LANGUAGE=en_US
 PATH=(custom, no user)
SegvAnalysis:
 Segfault happened at: 0x7f789c06ea78 <gtk_widget_translate_coordinates+168>: mov (%r14),%rax
 PC (0x7f789c06ea78) ok
 source "(%r14)" (0x00000000) not located in a known VMA region (needed readable region)!
 destination "%rax" ok
SegvReason: reading NULL VMA
Signal: 11
SourcePackage: apport
StacktraceTop:
 gtk_widget_translate_coordinates () from /usr/lib/x86_64-linux-gnu/libgtk-3.so.0
 ?? () from /usr/lib/x86_64-linux-gnu/libgtk-3.so.0
 ?? () from /usr/lib/x86_64-linux-gnu/libgtk-3.so.0
 ?? () from /usr/lib/x86_64-linux-gnu/libgtk-3.so.0
 gtk_event_controller_handle_event () from /usr/lib/x86_64-linux-gnu/libgtk-3.so.0
Title: apport-gtk crashed with SIGSEGV in gtk_widget_translate_coordinates()
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups:

[Apport retracing service (apport)]

Thank you for taking the time to report this crash and helping to make this software better. This particular crash has already been reported and is a duplicate of bug #1539995, so is being marked as such. Please look at the other bug report to see if there is any missing information that you can provide, or to see if there is a workaround for the bug. Additionally, any further discussion regarding the bug should occur in the other report. Please continue to report any other bugs you may find.

[Eliah Kagan (degeneracypressure)]

This bug report is for the bug in apport or GTK+, *not* for the kernel bug (if any) that triggered it. However, in case the kernel oops that triggered the apport-gtk instance that crashed is relevant, here's the output of dmesg.