Bug 570655 - [abrt] crash in thunderbird-3.0.3-1.fc12: Process /usr/lib64/thunderbird-3.0/thunderbird-bin was killed by signal 11 (SIGSEGV)
Summary: [abrt] crash in thunderbird-3.0.3-1.fc12: Process /usr/lib64/thunderbird-3.0/...
Keywords:
Status: CLOSED DUPLICATE of bug 570391
Alias: None
Product: Fedora
Classification: Fedora
Component: thunderbird
Version: 12
Hardware: x86_64
OS: Linux
low
high
Target Milestone: ---
Assignee: Jan Horak
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard: abrt_hash:25eaf0bbbccc45b476f9aa494c9...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2010-03-04 22:47 UTC by Dave Allan
Modified: 2018-04-11 12:30 UTC (History)
5 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2010-03-26 15:19:49 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
log of updates and crashes (2.45 KB, text/plain)
2010-03-08 20:40 UTC, Dave Allan
no flags Details

Description Dave Allan 2010-03-04 22:47:50 UTC
abrt 1.0.7 detected a crash.

architecture: x86_64
cmdline: /usr/lib64/thunderbird-3.0/thunderbird-bin
comment: Opened or, I think in one case, simply received and viewed in pane, new mail.
component: thunderbird
executable: /usr/lib64/thunderbird-3.0/thunderbird-bin
kernel: 2.6.31.12-174.2.22.fc12.x86_64
package: thunderbird-3.0.3-1.fc12
rating: 4
reason: Process /usr/lib64/thunderbird-3.0/thunderbird-bin was killed by signal 11 (SIGSEGV)
release: Fedora release 12 (Constantine)
How to reproduce: Thunderbird crashed twice today, I believe when I've opened ordinary messages (ordinary == not particularly large or containing any content other than text).  I have no idea if the backtrace contains sensitive data, so I've deleted it entirely.  I may be able to provide it privately.

Comment 1 Chris Campbell 2010-03-05 15:58:29 UTC
Well, without a back-trace there is simply no way for us to determine the actual cause of the crash. You can actually open the back-trace and read through it. They usually don't show anything which might be considered private, but it should be easy enough to read it.

Closing this report as is. If you decide to attach the back-trace, you can reopen it at that time.



-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 2 Dave Allan 2010-03-08 16:12:02 UTC
As I said before, I cannot validate that the backtrace does not contain confidential information.  

Let me restate the problem:  

This crash occurred while using my corporate email account.  I work for a corporation that has a strict policy about revealing confidential information.  The backtrace is 1444 lines long, and it is filled with all kinds of values, some human readable strings, some not.  Suggesting that it's feasible to review it is unrealistic.  I'm a C programmer; I read backtraces all day long; there's no way that such a data set can be vetted in a reasonable amount of time, even by someone familiar with the code.

As I said in the original BZ, Thunderbird segfaulted twice in one day without any unusual input.  That's fairly concerning to me at least.  I offered to provide the backtrace privately, but you ignored that offer and closed the bug.  Typically, that's what the NEEDINFO flag is used for.

Since Thunderbird just crashed again, I'm reopening and taking the bug.  I will find someone to whom I can provide the backtrace.  Please do not close it again.  I will update it as I have further information.

Comment 3 Matěj Cepl 2010-03-08 18:04:46 UTC
(In reply to comment #2)
> This crash occurred while using my corporate email account.  I work for a
> corporation that has a strict policy about revealing confidential information. 

??? Aren't we working for the same corporation? If you mark the attachment as private, no-one outside of private_comment group (and volunteer bugzappers are not included) can see it.

> As I said in the original BZ, Thunderbird segfaulted twice in one day without
> any unusual input.  That's fairly concerning to me at least.  I offered to
> provide the backtrace privately, but you ignored that offer and closed the bug.
>  Typically, that's what the NEEDINFO flag is used for.

That must be some kind of misunderstanding. Sorry for that.

> Since Thunderbird just crashed again, I'm reopening and taking the bug.  I will
> find someone to whom I can provide the backtrace.  Please do not close it
> again.  I will update it as I have further information.    

As I said, mark it as private and I will take a look.

Comment 4 Dave Allan 2010-03-08 19:29:17 UTC
(In reply to comment #3)

Hi Matej, thanks for the info.  We do indeed work for the same company--I just wasn't aware that everybody who could see private attachments on Fedora bugs was at RH.  I have two separate backtraces; I'll attach both.

Comment 8 Dave Allan 2010-03-08 19:55:52 UTC
I thought this segfault might be related to a folder I had with >27k messages in it, so I deleted all the messages in that folder.  The most recent segfault happened after deleting those messages.

Comment 9 Dave Allan 2010-03-08 20:40:36 UTC
Created attachment 398617 [details]
log of updates and crashes

Statistically speaking, it looks like this crash is new with 3.0.3-1

Comment 12 Dave Allan 2010-03-12 20:23:16 UTC
Matej or Jan, do any of those traces shed any light on what's going on?  This must be some corner case that my particular setup is triggering, or the world would be screaming as it's fairly painful--I'm seeing several crashes a day.

Comment 13 Matěj Cepl 2010-03-26 15:07:26 UTC
I found this on the upstream bugzilla https://bugzilla.mozilla.org/show_bug.cgi?id=411147 but it is marked as "Windows only".

All backtraces are identical in ending in nsQueryInterface::operator() (with slightly different paths leading to the same point), except for the attachment 398605 [details], which is in nsSSLThread::requestRecvMsgPeek

Comment 14 Matěj Cepl 2010-03-26 15:19:49 UTC

*** This bug has been marked as a duplicate of bug 570391 ***

Comment 15 Wayne Mery (:wsmwk) 2010-03-26 15:53:50 UTC
(In reply to comment #13)
> I found this on the upstream bugzilla
> https://bugzilla.mozilla.org/show_bug.cgi?id=411147 but it is marked as
> "Windows only".
> 
> All backtraces are identical in ending in nsQueryInterface::operator() (with
> slightly different paths leading to the same point), except for the attachment
> 398605, which is in nsSSLThread::requestRecvMsgPeek    

Note: bug 411147 might not be "windows-only" - my comment "windows only per crash-stats" only means I didn't see any linux crashes, which isn't unusual because we tend to not get many linux crashes feeding to crash-stats


Note You need to log in before you can comment on or make changes to this bug.