libglib2.0-0 (2.38.0-1ubuntu1) causes eclipse/swt to crash

[aanno]

Newer Versions of libglib2.0-0 (2.38.0-1ubuntu1, saucy salamander, aka 13.10) causes eclipse/swt to crash.

This is bug is triggered from an 'official' eclipse install. I haven't tried the eclipse that comes with ubuntu because it is not what a professional java programmer needs.

Observed behaviour:
- Eclipse regulary crashes (see included VM dumps). The bug is normally trigger by hitting a SWT Button (e.g. Search -> File Search -> Cancel Button).

Expected behaviour:
- libglib2.0-0 in saucy should work without any issue against a 'offical' eclipse/swt.

[aanno (thomas-pasch)]

Created attachment 236954
VM crash file (from oracle jdk 1.7.0_45)

Newer Versions of libglib2.0-0 (2.38.0-1ubuntu1, saucy salamander, aka 13.10) causes eclipse/swt to crash.

Observed behaviour:
- Eclipse regulary crashes (see included VM dumps). The bug is normally trigger by hitting a SWT Button (e.g. Search -> File Search -> Cancel Button).

Expected behaviour:
- 'Official' eclipse builds should work with ubuntu saucy

[Martin St. (hurzl)]

Same problem here on Ubuntu Saucy 13.10 with Oracle JDK 1.7.0_45

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Martin Stockhammer (m-stockhammer)]

Same problem with OpenJDK 7u25-2.3.12-4ubuntu3
#
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00007f99d9e4db3c, pid=25544, tid=140298792535872
#
# JRE version: 7.0_25-b30
# Java VM: OpenJDK 64-Bit Server VM (23.7-b01 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C [libgobject-2.0.so.0+0x31b3c] g_type_check_instance_is_a+0x3c
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# If you would like to submit a bug report, please include
# instructions on how to reproduce the bug and visit:
# https://bugs.launchpad.net/ubuntu/+source/openjdk-7/
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

--------------- T H R E A D ---------------

Current thread (0x000000000214d000): JavaThread "main" [_thread_in_native, id=25544, stack(0x00007fff4c032000,0x00007fff4c132000)]

[Martin St. (hurzl)]

Tested it with OpenJDK 7u25-2.3.12-4ubuntu3 and have the same crashes.

#
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00007f99d9e4db3c, pid=25544, tid=140298792535872
#
# JRE version: 7.0_25-b30
# Java VM: OpenJDK 64-Bit Server VM (23.7-b01 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C [libgobject-2.0.so.0+0x31b3c] g_type_check_instance_is_a+0x3c
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# If you would like to submit a bug report, please include
# instructions on how to reproduce the bug and visit:
# https://bugs.launchpad.net/ubuntu/+source/openjdk-7/
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

--------------- T H R E A D ---------------

Current thread (0x000000000214d000): JavaThread "main" [_thread_in_native, id=25544, stack(0x00007fff4c032000,0x00007fff4c132000)]

siginfo:si_signo=SIGSEGV: si_errno=0, si_code=1 (SEGV_MAPERR), si_addr=0x000000080000001a

[Martin St. (hurzl)]

Its mostly but not always the same frame:

hs_err_pid11846.log:# Problematic frame:
hs_err_pid11846.log-# C [libgtk-x11-2.0.so.0+0x24492a] gtk_widget_queue_draw+0x1a
--
hs_err_pid13821.log:# Problematic frame:
hs_err_pid13821.log-# C [libgtk-x11-2.0.so.0+0x24492a] gtk_widget_queue_draw+0x1a
--
hs_err_pid15333.log:# Problematic frame:
hs_err_pid15333.log-# C [libwebkitgtk-1.0.so.0+0x3dfda8] DumpRenderTreeSupportGtk::setComposition(_WebKitWebView*, char const*, int, int)+0x88a8
--
hs_err_pid15474.log:# Problematic frame:
hs_err_pid15474.log-# C [libgobject-2.0.so.0+0x31b3c] g_type_check_instance_is_a+0x3c
--
hs_err_pid18466.log:# Problematic frame:
hs_err_pid18466.log-# C [libgtk-x11-2.0.so.0+0x24492a] gtk_widget_queue_draw+0x1a
--
hs_err_pid21190.log:# Problematic frame:
hs_err_pid21190.log-# C [libgtk-x11-2.0.so.0+0x24492a] gtk_widget_queue_draw+0x1a
--
hs_err_pid24507.log:# Problematic frame:
hs_err_pid24507.log-# C [libgobject-2.0.so.0+0x31b3c] g_type_check_instance_is_a+0x3c
--
hs_err_pid25544.log:# Problematic frame:
hs_err_pid25544.log-# C [libgobject-2.0.so.0+0x31b3c] g_type_check_instance_is_a+0x3c
--
hs_err_pid31708.log:# Problematic frame:
hs_err_pid31708.log-# V [libjvm.so+0x64f81b] JNI_ArgumentPusherVaArg::JNI_ArgumentPusherVaArg(_jmethodID*, __va_list_tag*)+0x1b

[aanno (thomas-pasch)]

There is an associated ubuntu bug report at https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/1245468

[aanno (thomas-pasch)]

https://bugs.launchpad.net/ubuntu/+source/java-common/+bug/1241101 looks like to be related; however, I could not stop the problem with fiddleing with UBUNTU_MENUPROXY, GTK_MODULES, or GTK_IM_MODULE.

[aanno (thomas-pasch)]

https://bugs.launchpad.net/ubuntu/+source/java-common/+bug/1242632 is marked as duplicate, however I could not stop the problem with another GTK theme (I'm using Kubuntu/KDE, anyway).

[aanno (thomas-pasch)]

CORRECTION FOR MY COMMENT #11 ABOVE:

https://bugs.launchpad.net/ubuntu/+source/java-common/+bug/1242632 is marked as duplicate, AND I COULD VERIFY THAT THE PROBLEM GOES AWAYS IN KUBUNTU/KDE when doing:

System Settings -> Application Themes -> GTK -> Choose GTK2 Theme

Choose 'Radiance' instead of 'oxygen-gtk'

[aanno (thomas-pasch)]

This problem maybe related to the following ubuntu bugs:

https://bugs.launchpad.net/ubuntu/+source/emacs23/+bug/1142213
https://bugs.launchpad.net/ubuntu/+source/cairo/+bug/889045
https://bugs.launchpad.net/ubuntu/+source/gtk2-engines-oxygen/+bug/948178

[aanno (thomas-pasch)]

Using oxgen-gtk2 causes eclipse/swt to crash. Using another the 'Radience' gtk2 theme works without any problem.

Reproducible: Sometimes

Steps to Reproduce:
1. Use an out-of-stock eclipse from the eclipse web site.
2. Start it.
3. Use 'Search -> File Search -> Cancel Button'
Actual Results:
- Eclipse regulary crashes (see included VM dumps). The bug is normally trigger by hitting a SWT Button (e.g. Search -> File Search -> Cancel Button).

Expected Results:
Expected behaviour:
- oxygen-gtk2 theme should work without any issue against a 'offical' eclipse/swt.

[aanno (thomas-pasch)]

Created attachment 83485
java crash dump file (from oracle jdk 1.7.0_45)

[aanno (thomas-pasch)]

Some more details could be found at https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/1245468

[aanno (thomas-pasch)]

Created attachment 83486
java crash dump file (from oracle jdk 1.7.0_45)

[aanno (thomas-pasch)]

Created attachment 83487
java crash dump file (from oracle jdk 1.7.0_45)

[Hugo Pereira Da Costa (hugo-pereira)]

I am closing this bug report
Last few posted crashes are unrelated to oxygen-gtk: they are libgobject.
I guess it is just someone pushing the "submit report" button.
So that it brings no new information
(besides java coredumps also have very little information to start with)
finally,
- none of the oxygen-gtk devevloppers are able to reproduce the crash,
- The original (oxygen-gtk related) has been said to be fixed after some update (and was likely upstream too).

To the others, sorry for the frustration
This should be reported to either
- glib
- your distribution

[aanno (thomas-pasch)]

Dear Hugo,

thank you for your prompt response. Notice that I've reported the bug to (a) the distribution and (b) eclipse. Just to be very explicit: The bug _only_ occurs with the oxygen-gtk2 theme. If I use the 'Radiance' theme, everything is all right. Please also notice that a java crash dump does contain many information that are of significance for C/C++ developers, including a (symbolic) trace to the root problem.

[Hugo Pereira Da Costa (hugo-pereira)]

Well:
- the fact that it happens in oxygen-gtk does not mean it is an oxygen-gtk bug: just means the code path followed by oxygen-gtk triggers a bug that can be anywhere in the chain
- nowhere in the core dump, as far as I can tell, is oxygen-gtk mentionned. (sorry in advance if I missed it).
(nor are there any function calls mentioned, nor line numbers)
I at least, do not know how to use this report to go and actually debug the code.

[Keith Baker (keibak)]

Maybe also related:
https://bugs.launchpad.net/java-common/+bug/1241101
https://bugs.launchpad.net/ubuntu/+source/gtk+2.0/+bug/1205452

[aanno (thomas-pasch)]

I've reported the bug KDE upstream as https://bugs.kde.org/show_bug.cgi?id=327441 . Strangely enough, they changed the status to 'FIXED, CLOSED' even as they not have looked at it. So I'm deleting the upstream link here to signal that the problem is still NOT solved.