Bug 1266786

Summary: [gdb] libreoffice: Program received signal SIGSEGV, Segmentation fault.
Product: [Fedora] Fedora Reporter: Christian Stadelmann <fedora>
Component: libreofficeAssignee: Caolan McNamara <caolanm>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 22CC: caolanm, dtardon, erack, ltinkl, mstahl, sbergman
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-09-28 07:41:27 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
gdb log file none

Description Christian Stadelmann 2015-09-27 22:13:14 UTC 
Created attachment 1077788 [details]
gdb log file

Description of problem:
I ran LibreOffice with the debugger since I ran into an issue (https://bugs.documentfoundation.org/show_bug.cgi?id=85076). I followed these steps: https://wiki.documentfoundation.org/Valgrind#GNU.2FLinux:_How_to_get_a_backtrace

Version-Release number of selected component (if applicable):
libreoffice-4.4.5.2-6.fc22.x86_64

How reproducible:
not reliably, don't know the details yet.

Steps to Reproduce:
1. Open LibreOffice
2. Open a document (only tested Writer so far)
3. Open the menu bar by mouse click

Actual results:
With the debugger attached LibreOffice crashes.

Expected results:
LibreOffice should not crash.

Additional info:
See attached gdb debug log. I'll attach a valgrind log when I have more time.
I don't know whether this issue is specific to changes made by fedora package maintainers. Since there is quite a lot of downstream patches I decided to post it here.
Comment 1 Stephan Bergmann 2015-09-28 07:41:27 UTC 
The various threads' backtraces indicate that there is a JVM instantiated in process, and the backtrace for the thread that generates a SIGSEGV looks like it runs JIT'ed Java code.  It routinely happens in the JVM that SIGSEGV is generated, handled by the JVM's signal handler, and translated into a java.lang.NullPointerException.  Therefore, when you have gdb attached to an soffice.bin process in which a JVM has been instantiated, it routinely happens that harmless SIGSEGV are reported.  ("Open the menu bar" confirms this, as code in LibreOffice extensions written in Java may be run then.)