gnome-settings-daemon crashed with SIGSEGV in __run_exit_handlers()

[Eliah Kagan]

Binary package hint: gnome-settings-daemon

I attempted to shut down an Oneiric i386 system in a unity-2d session (with gnome-settings-daemon 3.0.2-1ubuntu2) using the indicator-session menu. All my application windows, desktop icons, the Unity launcher, and the top panel disappeared, leaving a completely blank desktop, and I anticipated that X11 was going to quit at any moment, but then Apport informed me that gnome-settings-daemon crashed with SIGSEGV in __run_exit_handlers(). The Apport window had no title bar. I tried to report the bug in that state, but was unable to do so because, after resizing the window to get a better look at the stack trace, I was unable to interact with the window's contents (i.e., clicking on buttons and other UI elements had no effect). Since I was also unable to quit Apport, I switched to a virtual console and rebooted the machine.

The semantic similarity (both pertaining to functions run when the application quits) and slight stack trace similarity seems to suggest a possible connection between this bug and bug 788710.

ProblemType: Crash
DistroRelease: Ubuntu 11.10
Package: gnome-settings-daemon 3.0.2-1ubuntu2
ProcVersionSignature: Ubuntu 2.6.39-3.10-generic 2.6.39
Uname: Linux 2.6.39-3-generic i686
Architecture: i386
Date: Fri Jun 10 00:26:03 2011
Disassembly: => 0x4d3fa70: Cannot access memory at address 0x4d3fa70
ExecutablePath: /usr/lib/gnome-settings-daemon/gnome-settings-daemon
InstallationMedia: Ubuntu 11.10 "Oneiric Ocelot" - Alpha i386 (20110604)
ProcCmdline: /usr/lib/gnome-settings-daemon/gnome-settings-daemon
ProcCwd: /home/ek
ProcEnviron:
 LANGUAGE=en_US:en
 PATH=(custom, no user)
 LANG=en_US.utf8
 LC_MESSAGES=en_US.UTF-8
 SHELL=/bin/bash
SegvAnalysis:
 Segfault happened at: 0x4d3fa70: Cannot access memory at address 0x4d3fa70
 PC (0x04d3fa70) not located in a known VMA region (needed executable region)!
SegvReason: executing unknown VMA
Signal: 11
SourcePackage: gnome-settings-daemon
StacktraceTop:
 ?? ()
 __run_exit_handlers (status=0, listp=0x59c324, run_list_atexit=true) at exit.c:72
 __on_exit (func=0, arg=0xbfb4c454) at on_exit.c:28
 __libc_start_main (main=0x804a8a0 <main>, argc=1, ubp_av=0xbfb4c454, init=0x804d680 <__libc_csu_init>, fini=0x804d6e0 <__libc_csu_fini>, rtld_fini=0xec5a50, stack_end=0xbfb4c44c) at libc-start.c:232
 _start ()
Title: gnome-settings-daemon crashed with SIGSEGV in __run_exit_handlers()
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare

[Paul Coenraats (p-coenraats)]

I am completely new to Ubuntu and The above reports are automatically generated so I can be of no real help except to say that it may possible have something to do with the fact that the "system" appears to be taking on the identity of a dual boot system that is running Ubunto 11.04 Server or DESKTOP 11.10. That is what I assume as this is the end of the desktop installation having initially expected it to run without installing it however that all seems to be okay and I am expecting on boot up that I should get a "Dual Boot" menu to select either option......................here goes................regards Paul 12/06/11

[Eliah Kagan (degeneracypressure)]

@Paul Coenraats
You might be able to be of more help than you think. Please read https://help.ubuntu.com/community/ReportingBugs carefully. That should give you a sense of how bug reporting in Ubuntu works and what kind of information is important to provide in bug reports. Then please post here, and describe in detail what happened on your system to make you think you may be experiencing this bug. (Are you saying that you booted a live CD and this bug occurred when you were trying to shut down your computer from the live CD? If the answer is yes, please specify what live CD you booted from, and please still provide as detailed a description as possible of the steps you took leading up to the problem, and of the problem itself. You can explicitly compare with and contrast against my description above, if you wish.) This way, we can figure out whether or not you have enough information to merit marking this bug as confirmed.

You might have a dual-boot system with both Ubuntu 11.04 and Ubuntu 11.10 installed on separate partitions (with the option to choose between them when you boot up your system), but that is not a problem, and gnome-settings-daemon doesn't have anything to do with dual- (or, more generally, multiple-) booting. Your dual-boot configuration could potentially cause latent bugs in grub2 / grub-pc to be triggered (which would be a *good* thing, as it would be an opportunity to report those bugs so they could be fixed), but unless you are actually sharing configuration data between the two systems (which you would have to set up yourself), it's very unlikely that your dual boot configuration would trigger this bug or other bugs in gnome-settings-daemon. However, since you have a dual-boot system, it is important that when providing the information requested above, you clarify whether the events you think may constitute this bug occurred on your Ubuntu 11.04 Natty system or on your Ubuntu 11.10 Oneiric system.

The above post has assumed that you are posting here because you have experienced something that is similar to something described on this page. However, if there was no particular reason why you posted here rather than elsewhere, and you have not experienced any of the problems detailed in this bug's description, then (1) you should make sure the green link at the top of this page doesn't start with "This bug affects you" (if it does, click it and select the option indicating that you are not affected by this bug), (2) feel free to unsubscribe from the bug if it doesn't interest you, or to otherwise edit your subscription or bug mail options (I subscribed you so you would be notified of this post any any other posts or status changes), and (3) feel free to post a question at https://answers.launchpad.net/ubuntu if you require assistance with an unrelated matter in Ubuntu.

[Jean-Baptiste Lallement (jibel)]

Very reproducible, just booting an oneiric system.
Also check bug 788710 but there's no trace.

[Eliah Kagan (degeneracypressure)]

I can see why this is probably bug 788710, as its title is the same as other duplicate bug 796497 and its trace is semantically similar. However, this bug's stack trace is *not* the same as bug 796497's. For example, there is no call to exit() in this bug's stack trace (though there is a call to __on_exit() in the same point in the trace).

This probably is really a duplicate of bug 788710. After all, the real bug is almost certainly in a function that is registered during runtime as an exit handler (hence the missing information about the first step in all these traces), and this trace and all the other traces are consistent with that.

But in case the seemingly different path this crash took does make a difference, I want to make sure this is called to the attention of people more knowledgeable than myself.