Bug 525649 - ABRT should analyze back-traces better to catch more duplicates
Summary: ABRT should analyze back-traces better to catch more duplicates
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: abrt
Version: 12
Hardware: All
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Karel Klíč
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
: 524416 541442 543517 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2009-09-25 08:27 UTC by Martin Stransky
Modified: 2018-04-11 07:40 UTC (History)
11 users (show)

Fixed In Version: abrt-1.1.14-1
Clone Of:
Environment:
Last Closed: 2010-12-04 07:38:50 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
sample backtrace from gtk application (7.26 KB, text/plain)
2009-09-25 13:53 UTC, Tomáš Bžatek
no flags Details

Description Martin Stransky 2009-09-25 08:27:55 UTC
Description of problem:

ABRT fails to find duplicates sometimes...

Some ideas:

- only 5-6 functions from stack may be used for hash
- http://live.gnome.org/Bugsquad/TriageGuide/FindingDuplicates

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:

Comment 1 Tomáš Bžatek 2009-09-25 13:53:33 UTC
Created attachment 362678 [details]
sample backtrace from gtk application

I'd like to raise importance of this issue. Looks like abrt takes into the account all frames found in the stracktrace. This is wrong as there may be infinite number of ways to get to the crash in glib-based applications (using glib mainloop) as event handling is (probably) done in idle.

For example, look at the attached backtrace. It goes from gtk_main() to g_main_loop_run(), processing some X events (xlib), ending with calling exit handlers (frame #16), which involves a lot of other things before abort () is actually called. The real cause here lies in between frames #20 and #18. Now, this is maybe wrong example due to exit handlers involved in the backtrace, but imagine that most of X protocol stack is async. Trace will be always different, chance to correctly finding duplicate is very low.

This makes ABRT nearly unusable for glib and gtk applications, we're receiving a lot of duplicates for the same thing. For now, blacklisting apps linking to libglib-2.0.so might be a solution until this is fixed.

Comment 2 Tomáš Bžatek 2009-09-29 11:57:14 UTC
Typical example of differences in backtraces with random glib and gtk stuff: https://bugzilla.redhat.com/attachment.cgi?id=362493 vs. https://bugzilla.redhat.com/attachment.cgi?id=362963

Comment 3 Denys Vlasenko 2009-10-06 15:49:48 UTC
from 524416:

Description of problem:
Bug 522853 and bug 522854 are IMHO duplicates, but abrt didn't figured it out.

Comment 4 Denys Vlasenko 2009-10-06 15:50:03 UTC
*** Bug 524416 has been marked as a duplicate of this bug. ***

Comment 5 Martin Stransky 2009-11-13 12:27:07 UTC
Please don't steal my bug. This one is about better strack-trace analyze (select propper thread, print relevant info to bugzilla and so on). Not about duplicates.

Comment 6 Denys Vlasenko 2009-11-13 12:31:55 UTC
For me as a abrt developer, "analyze backtrace better" is too vague. "Analyze backtrace better" is what way?

I changed the title so that it actually says what should be achieved according to your own bug descripton.

Did I misunderstand you?

Comment 7 Bug Zapper 2009-11-16 12:54:28 UTC
This bug appears to have been reported against 'rawhide' during the Fedora 12 development cycle.
Changing version to '12'.

More information and reason for this action is here:
http://fedoraproject.org/wiki/BugZappers/HouseKeeping

Comment 8 Christopher Beland 2010-01-04 15:50:06 UTC
*** Bug 543517 has been marked as a duplicate of this bug. ***

Comment 9 Dave Malcolm 2010-01-04 16:18:48 UTC
Carrying forward info from bug 543517 (just marked as a dup of this one), another specific example of where two backtraces that ought to be marked as duplicates aren't being matched as such:

> For example, bug 543286 shows a segfault inside qt's python bindings.
> Bug 543317 shows the same segfault, but with (presumably) richer debuginfo.

It could be argued that we're seeing different underlying reasons for the lack of duplicate-matching, of course.

Comment 10 Denys Vlasenko 2010-02-05 16:37:57 UTC
*** Bug 541442 has been marked as a duplicate of this bug. ***

Comment 11 Denys Vlasenko 2010-02-05 16:38:58 UTC
From dup 541442:

Description of problem:
I have developed to my jetpack script
(http://git.fedorahosted.org/git/?p=triage.git;a=blob_plain;f=jetpack/bugzillaBugTriage.js;hb=jetpack
look around function parseBacktrace) which parses attached backtrace and
provides to the comment box most relevant part (if available). See bug 541173
comment 3 for an example. 

I guess it could be useful if abrt provided something similar automagically.
Maybe shorter? How many frames are most useful?

And no, I don't think should REPLACE attaching full all-threads backtrace, just
SUPPLEMENT it. The attached backtrace should remain, IMHO.

Comment 12 Bug Zapper 2010-11-04 09:50:13 UTC
This message is a reminder that Fedora 12 is nearing its end of life.
Approximately 30 (thirty) days from now Fedora will stop maintaining
and issuing updates for Fedora 12.  It is Fedora's policy to close all
bug reports from releases that are no longer maintained.  At that time
this bug will be closed as WONTFIX if it remains open with a Fedora 
'version' of '12'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version prior to Fedora 12's end of life.

Bug Reporter: Thank you for reporting this issue and we are sorry that 
we may not be able to fix it before Fedora 12 is end of life.  If you 
would still like to see this bug fixed and are able to reproduce it 
against a later version of Fedora please change the 'version' of this 
bug to the applicable version.  If you are unable to change the version, 
please add a comment here and someone will do it for you.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events.  Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

The process we are following is described here: 
http://fedoraproject.org/wiki/BugZappers/HouseKeeping

Comment 13 Karel Klíč 2010-11-04 12:19:37 UTC
ABRT's backtrace analysis is probably good enough now, esp. in the latest-git (next version).

Comment 14 Bug Zapper 2010-12-04 07:38:50 UTC
Fedora 12 changed to end-of-life (EOL) status on 2010-12-02. Fedora 12 is 
no longer maintained, which means that it will not receive any further 
security or bug fix updates. As a result we are closing this bug.

If you can reproduce this bug against a currently maintained version of 
Fedora please feel free to reopen this bug against that version.

Thank you for reporting this bug and we are sorry it could not be fixed.

Comment 15 Jiri Moskovcak 2010-12-06 11:34:50 UTC
Fixed in abrt-1.1.14-1


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