indicator-application-service crashed with SIGSEGV in g_strcmp0()

[David Nielsen]

Binary package hint: indicator-application

Not sure what happened, hopefully apport will explain the crash.

ProblemType: Crash
DistroRelease: Ubuntu 10.10
Package: indicator-application 0.2.7-0ubuntu1
ProcVersionSignature: Ubuntu 2.6.35-20.29-generic 2.6.35.4
Uname: Linux 2.6.35-20-generic i686
Architecture: i386
Date: Mon Sep 13 21:25:26 2010
EcryptfsInUse: Yes
ExecutablePath: /usr/lib/indicator-application/indicator-application-service
InstallationMedia: Ubuntu 10.04.1 LTS "Lucid Lynx" - Release i386 (20100816.1)
ProcCmdline: /usr/lib/indicator-application/indicator-application-service
ProcEnviron:
 SHELL=/bin/bash
 LANG=da_DK.utf8
SegvAnalysis:
 Segfault happened at: 0xbd14ab: cmp %cl,(%edx)
 PC (0x00bd14ab) ok
 source "%cl" ok
 destination "(%edx)" (0x616e6174) not located in a known VMA region (needed writable region)!
SegvReason: writing unknown VMA
Signal: 11
SourcePackage: indicator-application
StacktraceTop:
 ?? () from /lib/libc.so.6
 g_strcmp0 () from /lib/libglib-2.0.so.0
 ?? ()
 ?? ()
 ?? ()
Title: indicator-application-service crashed with SIGSEGV in g_strcmp0()
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare

[Apport retracing service (apport)]

StacktraceTop:
 __strcmp_ssse3 () at ../sysdeps/i386/i686/multiarch/strcmp-ssse3.S:77
 g_strcmp0 () from /lib/libglib-2.0.so.0
 find_application (appstore=<value optimized out>,
 application_service_appstore_application_add (
 _notification_watcher_server_register_status_notifier_item (

[Sebastien Bacher (seb128)]

the bug seems to have some duplicate, could some dxteam member work on it?

[Ted Gould (ted)]

Unfortunately there doesn't seem to be enough information in the stack traces to figure out what is going on here. Could someone who's having it please try to get a Valgrind log of their indicator-application-service: https://wiki.ubuntu.com/Valgrind Thank you!

[David Barth (dbarth)]

Hmm, maybe some more checks around arguments processed by find_application in application-service-appstore.c
It sounds odd for app->dbus_name to be null, but if it's the case, the function should bail out without crashing.

[Aurélien Gâteau (agateau)]

We'd like to figure out what's causing this bug for you, but we haven't heard back from you in a while. Could you please provide the requested information? Thanks!

[MaxNegro (maxnegro)]

Aurélien Gâteau: could you please argument a little more on "requested information"? I'd like to provide a valgrind log if needed but I have no clue at how I should be intended to exec this "service" under valgrind (/usr/lib/indicator-application/indicator-application-service is not something I usually start from shell).

Anyway, this is quite hard to reproduce.

[Michael Terry (mterry)]

MaxNegro, you can start the service under valgrind with the following:

killall indicator-application-service
valgrind /usr/lib/indicator-application/indicator-application-service

But I tried it and couldn't get particularly egregious errors. Nor have I been able to reliably reproduce the crash. From the traces, it appears to be an Application object that has been free'd but not removed from the list of applications. Not sure how that could happen with recent code (I've hit this with 0.3.1-0ubuntu2).

[Kate Stewart (kate.stewart)]

Given the dups and the number of folks affecting this should probably be high. Have marked it for Natty updates. Is it still an issue with Oneiric?

[Sebastien Bacher (seb128)]

closing the bug didn't get new comments or duplicates for quite some time