gpartedbin crashed with signal 5 in Glib::exception_handlers_invoke()

[Vipul]

Binary package hint: gparted

Well i need to format a new sdcard and the gparted crashed.

ProblemType: Crash
DistroRelease: Ubuntu 10.10
Package: gparted 0.5.2-2ubuntu3
ProcVersionSignature: Ubuntu 2.6.35-10.15-generic-pae 2.6.35-rc5
Uname: Linux 2.6.35-10-generic-pae i686
NonfreeKernelModules: nvidia
Architecture: i386
Date: Sat Jul 24 15:33:43 2010
ExecutablePath: /usr/sbin/gpartedbin
ProcCmdline: /usr/sbin/gpartedbin
ProcEnviron:
 PATH=(custom, no user)
 LANG=en_IN
 SHELL=/bin/bash
Signal: 5
SourcePackage: gparted
StacktraceTop:
 Glib::exception_handlers_invoke() ()
 ?? () from /usr/lib/libglibmm-2.4.so.1
 ?? () from /lib/libglib-2.0.so.0
 start_thread () from /lib/libpthread.so.0
 clone () from /lib/libc.so.6
Title: gpartedbin crashed with signal 5 in Glib::exception_handlers_invoke()
UserGroups:

[Apport retracing service (apport)]

StacktraceTop:
 Glib::exception_handlers_invoke ()
 ?? () from /usr/lib/libglibmm-2.4.so.1
 g_thread_create_proxy (data=0x84d0650)
 start_thread () from /lib/libpthread.so.0
 clone () at ../sysdeps/unix/sysv/linux/i386/clone.S:130

[Nicolas DERIVE (kalon33)]

Mark it "confirmed" as it affects 7 people and has 3 dupes

[john mall (mantikor)]

Nice to see some improvemnet here. I appreciate your work!

[Frédéric Stemmelin (frederic-stemmelin)]

Hello, here is what i have done to provoc this crash. I have installed the Ubuntu alternate 10.10 beta amd version on a gpt partition sheme sata disk from a usb key /dev/sdc generated with usb-creator, but i didnt install grub on the MBR, i choose to install grub on a second usb key, /dev/sdd that i plugged in just for that. On reboot (with both usb keys plugged in) i choose the partial upgrade and after reboot i installed gparted, who crashes at every program launch. I did these steps because i dont want to overwrite the /dev/sda MBR used by my main Ubuntu studio 10.04 installation.

Hope this helps you to find the bug.

[Id2ndR (id2ndr)]

In my case, this happen just after I created an unformatted partition with gparted (when it scanned the result after the operation was successfully done).

[Mylan Connolly (mr-mylanman)]

I am using the Live CD, and just booted it up to create a few partitions on my hard drive before installing Windows XP, and gparted crashed before it could fully load. The hard drive had a corrupted NTFS partition on it, and the rest was free space.

[Tilo Körner (x-pinguin)]

I started gparted from the GNOME menu and while it was searching my devices for partitions, it crashed and apport presented me a report in the GNOME panel.

Here please find some technical info about my Ubuntu installation:
Release: 10.10 beta AMD64, installed from Alternate ISO image on USB thumb drive.
package version: gparted 0.6.2-1ubuntu1 from http://ftp.ticklers.org/archive.ubuntu.org/ubuntu/ maverick/main amd64 Packages

Running from the terminal makes the following happen:
~$ gksudo gparted
======================
libparted : 2.3
======================

glibmm-ERROR **:
unhandled exception (type std::exception) in signal handler:
what: basic_string::_S_create

aborting...

Figuring out that partitionmanager uses libparted, too, I installed partitionmanager with all necessary KDE libraries to see what happens upon starting that programme.

~$ partitionmanager
Connecting to deprecated signal QDBusConnectionInterface::serviceOwnerChanged(QString,QString,QString)
QMetaObject::invokeMethod: No such method KUniqueApplication::loadCommandLineOptionsForNewInstance()
kbuildsycoca4 running...
kbuildsycoca4(29253) KBuildMimeTypeFactory::createEntry: Missing <comment> field in "application/x-msdownload.xml"
kbuildsycoca4(29253) KBuildMimeTypeFactory::createEntry: Missing <comment> field in "text/x-component.xml"
kbuildsycoca4(29253) KBuildMimeTypeFactory::createEntry: Missing <comment> field in "application/x-note.xml"
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry Categories in "/usr/share/applications/gcalctool.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry Categories in "/usr/share/applications/im-switch.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry Categories in "/home/tilo/.local/share/applications/stellarium.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-png.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-chm.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-gif.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-txt.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-application.desktop" is not compliant with XDG standard (missing trailing semicolon).
kbuildsycoca4(29253) KConfigGroup::readXdgListEntry: List entry MimeType in "/home/tilo/.local/share/applications/wine-extension-a...

[William Bouterse (bill-bouterse)]

First boot of live-desktop 10.10 amd 64 daily build downloaded and burned 9/14/10
Attempted to install with btfs on / and /boot as ext4
Stalled out at formatting /boot

Had to reboot and decided to bring up gparted via teminal;

root@ubuntu:/home/ubuntu# gparted
======================
libparted : 2.3
======================

glibmm-ERROR **:
unhandled exception (type std::exception) in signal handler:
what: basic_string::_S_create

aborting...

So this has come similarities to #8
An unsuccessful partition format
The clincher is gparted via terminal came up fine on first boot before any partition work done

[joopbraak (joopbraak)]

It just crashes while scanning.

I just upgraded from Lucid to Maverick beta.

In Lucid, I had build gparted 0.6.2 from sources and then it worked. Since upgrade to Maverick this started happening.

[joopbraak (joopbraak)]

I installed the 0.6.2 deb package from debian and no crash anymore!

So something went wrong from 0.6.2-1 to 0.6.2-1ubuntu1.

Architecture: amd64

[joopbraak (joopbraak)]

Never mind, it still crashes.

[joopbraak (joopbraak)]

Sorry, ignore #14, I accidentally updated again to 0.6.2-1ubuntu1.

The truth is, 0.6.2-1ubuntu1 crashes, 0.6.2-1 (from debian) doesn't.

[aanno (thomas-pasch)]

Same here on amd64, 10.10-beta (maverick):

$ sudo gparted
======================
libparted : 2.3
======================

glibmm-ERROR **:
unhandled exception (type std::exception) in signal handler:
what: basic_string::_S_create

aborting...

[Pedro Nariyoshi (pedro-nariyoshi)]

same on i386, 10.10-beta

[Mariano Draghi (chaghi)]

GParted started to crash after creating an unformatted partition from a live-CD session of Maverick beta. Same issue reported in comment #8 (gparted started to crash just after that, the first time when it rescanned the drive)

The thing is that after installing 10.10 into the hard disk, I still experience the problem. There seems to be something in the partition table that doesn't get along with gparted...

The partition table and the partitions can be listed/manipulated with "fdisk" and with "parted" w/o issues.

[Pedro Nariyoshi (pedro-nariyoshi)]

My case happens when I plug in my girlfriend's iPod and try to open gParted.

(using 10.10-beta i386)

[Markus Dresch (markus-dresch)]

same thing on "Linux 2.6.35-22-generic #33-Ubuntu SMP Sun Sep 19 20:32:27 UTC 2010 x86_64 GNU/Linux"

> parted /dev/sda print
Model: ATA SAMSUNG HD502HJ (scsi)
Disk /dev/sda: 500GB
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number Start End Size Type File system Flags
 1 1049kB 106MB 105MB primary ntfs boot
 2 106MB 83.9GB 83.8GB primary ntfs
 3 83.9GB 168GB 83.8GB primary
 4 168GB 256GB 88.4GB extended
 5 168GB 172GB 4095MB logical linux-swap(v1)
 6 172GB 172GB 511MB logical ext2
 7 172GB 214GB 41.9GB logical btrfs
 8 214GB 256GB 41.9GB logical ext4

partition 3 is unused and there's unpartitioned space after partition 8.

[Curtis Gedak (gedakc)]

Would someone who is experiencing this problem be able to test the attached proposed patch?

Unfortunately my testing with GParted on the Sept. 20, 2010 maverick-desktop-i386.iso image does not invoke the crash problem with the current package gparted-0.6.2-1ubuntu1

Regards,
Curtis Gedak
(Maintainer of GParted)

[Brian Rogers (brian-rogers)]

Doesn't seem to help...

[Curtis Gedak (gedakc)]

Thank you Brian for testing.

Which repository and which version of the source code did you use as a base to apply the patch?

[Brian Rogers (brian-rogers)]

I used Ubuntu Maverick's gparted and added that patch.

[Curtis Gedak (gedakc)]

Thank you for the clarification Brian.

Would you be able to try using the gparted-0.6.2 source code directly from the GParted web site?
http://sourceforge.net/projects/gparted/files/gparted/

The reason I ask is that in comment #12 and comment #13 joopbraak mentioned that compiling from source or using the debian package worked. It is possible that some changes introduced into the Ubuntu package have caused the problem. By testing using the original source code from the official GParted web site we can determine if this is indeed the case.

[Brian Rogers (brian-rogers)]

Compiling from the original source produces a working gparted. Applying the patch 02-btrfs.patch breaks it again.

[Curtis Gedak (gedakc)]

Excellent work Brian! Now we know the problem originates with the 02-btrfs.patch.

[Markus Dresch (markus-dresch)]

the gparted project source works fine.

[Curtis Gedak (gedakc)]

I believe that the btrfs patch originated with the following bug report:
https://bugzilla.gnome.org/show_bug.cgi?id=571170

Luca Bruno did more work on the patch and committed his changes to the gnome git repository for gparted in a separate branch (lucab/btrfs):
http://git.gnome.org/browse/gparted/

The GParted project plans to include the btrfs branch code in version 0.7.0.

[Curtis Gedak (gedakc)]

For background information, the patch included in comment #21 does not fix this problem. The single line change of code has been included in the GParted git repository for the GParted 0.6.3 release planned for tomorrow.

Git commit link:
Use direct assignment for string in class initialization
http://git.gnome.org/browse/gparted/commit/?id=d040e4e5847905f58207e093917b6c3d36e641d7

Development Plans for the Next Release of GParted (0.6.3):
http://gparted-forum.surf4.info/viewtopic.php?id=14252

[RapierTG (rapier)]

This bug affects me also. Gparted dont crash on start scanning manually on all drives ("gparted /dev/sdxx"), only 2 in software raid0.

[Curtis Gedak (gedakc)]

I believe that this problem has been worked and solved in a different bug report.

See Ubuntu launchpad bug #617885
     Bug #617885 - gparted crash at start: glibmm-ERROR **
     https://bugs.launchpad.net/ubuntu/+source/gparted/+bug/617885

[Curtis Gedak (gedakc)]

Thanks go to Seth for providing a patch to resolve this problem.

The enhancements to address this bug report have been included upstream in GParted 0.7.0
which was released on October 29, 2010.