Bug 270002 - DrKonqi fails to recognize debug symbol installation until it is reinitialized (even when reloaded).
Summary: DrKonqi fails to recognize debug symbol installation until it is reinitialize...
Status: RESOLVED WORKSFORME
Alias: None
Product: drkonqi
Classification: Applications
Component: backtraceparsing (show other bugs)
Version: 5.27.9
Platform: openSUSE Linux
: NOR major
Target Milestone: ---
Assignee: Plasma Bugs List
URL: https://imgur.com/nv69Khr
Keywords:
Depends on:
Blocks:
 
Reported: 2011-04-03 14:52 UTC by Christoph Wickert
Modified: 2023-12-05 13:59 UTC (History)
6 users (show)

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report:


Attachments
useless kcrash (83 bytes, text/plain)
2011-04-03 14:52 UTC, Christoph Wickert
Details
Screenshot to prove drkonqui wrong (122.52 KB, image/png)
2011-04-03 14:53 UTC, Christoph Wickert
Details
Affected .KCrash №2. (5.83 KB, text/vnd.kde.kcrash-report)
2023-10-29 22:00 UTC, Roke Julian Lockhart Beedell
Details
Evidence. (2.72 MB, video/mp4)
2023-10-29 22:18 UTC, Roke Julian Lockhart Beedell
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Christoph Wickert 2011-04-03 14:52:55 UTC
Created attachment 58540 [details]
useless kcrash

Version:           unspecified (using KDE 4.6.1) 
OS:                Linux

After the KDE 4.5.1 update for F14, drkonqi no longer works because it fails to
recognize installed debugsymbols.

Version-Release number of selected component (if applicable):
kdebase-runtime-4.6.1-1.fc14.x86_64

Reproducible: Always

Steps to Reproduce:
1. Have a KDE application crash and try to use drkonqui to submit a crash
report.
2. drkonqui will complain about issing debug symbols
3. let drkonqui install them

Actual Results:  
drkonqui still claims the debuginfo is not installed. Clicking 'Install' again
will open a konsole window that tells you there is nothing to install. Even
clicking 'Reload' doesn't help, you are stuck in an endless loop.

Expected Results:  
drkonqui should see the installed debuginfo packages and continue with the
backtrace.

The backtrace drkonqui generates is indeed useless, see attachment.
Comment 1 Christoph Wickert 2011-04-03 14:53:34 UTC
Created attachment 58541 [details]
Screenshot to prove drkonqui wrong
Comment 2 Christoph Wickert 2011-04-03 14:54:04 UTC
Downstream bug report in Fedora at https://bugzilla.redhat.com/show_bug.cgi?id=693190
Comment 3 Dario Andres 2011-04-03 15:07:28 UTC
If DrKonqi says that there are still some missing debug information symbols is because the new generated backtrace is still lacking some information. You can enable the "Show backtrace" checkbox to see the full backtrace. 
May be there is some function call missing the proper debug information symbol but there is not "debug" package for it. (so the script will install only the avaliable packages).

- Can you check the "Show backtrace" option and paste here the full backtrace ?
Thanks
Comment 4 Christoph Wickert 2011-04-03 16:01:25 UTC
If you have an idea how to trigger a crash I can try but I don't think it will make a big difference. The kcrash said the backtrace was *completely* unusable, it was rated 0 of 3 stars. This means that it's not some library missing debugsymbols but something is fundamentally wrong.

drkonqui installed 37 debuginfo packages, more then 560 MB. I reported crashes of nepoumukservicestub in the past already and it never complained.

Please tell me how to force a crash and I'll give you the backtrace.
Comment 5 Dario Andres 2011-04-03 16:10:46 UTC
There are some crash test included in drkonqi, but I think those need to be manually compiled. So let's try another crash:
- Close all your Konqueror windows
- Run "konqueror www.kde.org" + Return  (on a shell)
- When the Konqueror window appear, start pressing (and releasing) F7 repeatedly (and quickly) until Konqueror crashes

However, judging by the screenshot, may be the nepomukservicestub was compiled without any symbol information, so that doesn't allow gdb to even retrieve a backtrace (I guess it could be a "one line" backtrace). I'm not sure if that is a Fedora problem or not

Regards
Comment 6 George Kiagiadakis 2011-04-03 16:16:37 UTC
Dario, the backtrace was attached on the description. This is most likely the known problem where gdb doesn't output anything for some strange reason. It has been reported many times but I have never been able to reproduce it and I have no idea what might be causing it.
Comment 7 Christoph Wickert 2011-04-03 19:27:10 UTC
I have now tried again twice by crashing konqueror:
1st attempt was with debuginfo installed from previuos nepomuk crash. Backtrace was complete and rated three stars.
2nd try was after I installed all debuginfo. Nevertheless the backtrace was rated 1 star. I was able to install the missing debuginfo from drkonqi and a full backtrace was generated automatically.

I have to admit this is strange. You still want me to submit the kcrash reports?
Comment 8 Harald Sitter 2019-12-14 17:33:09 UTC
Is this still reproducible?
There is some strange confusion in the earlier comments here. The original file attached shows an untracable process (chances are this was because the process or some of its libraries had been updated since it was started rendering the process untracable; this is going to be properly detected in Plasma 5.18).
In other comments more than one package were installed when clicking install debug symbols.
These two points don't really go along. When the backtrace is broken drkonqi can't install anything.

If it is still applicable I'd love some sample backtraces and ideally also some debug output, which you should be able to obtain by adding QT_LOGGING_RULES=org.kde.drkonqi=true to /etc/environment and logging out and back in again. After that drkonqi debug output should be added to ~/.xsession-errors.
Comment 9 Bug Janitor Service 2019-12-29 04:33:07 UTC
Dear Bug Submitter,

This bug has been in NEEDSINFO status with no change for at least
15 days. Please provide the requested information as soon as
possible and set the bug status as REPORTED. Due to regular bug
tracker maintenance, if the bug is still in NEEDSINFO status with
no change in 30 days the bug will be closed as RESOLVED > WORKSFORME
due to lack of needed information.

For more information about our bug triaging procedures please read the
wiki located here:
https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging

If you have already provided the requested information, please
mark the bug as REPORTED so that the KDE team knows that the bug is
ready to be confirmed.

Thank you for helping us make KDE software even better for everyone!
Comment 10 Bug Janitor Service 2020-01-13 04:33:10 UTC
This bug has been in NEEDSINFO status with no change for at least
30 days. The bug is now closed as RESOLVED > WORKSFORME
due to lack of needed information.

For more information about our bug triaging procedures please read the
wiki located here:
https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging

Thank you for helping us make KDE software even better for everyone!
Comment 11 Roke Julian Lockhart Beedell 2023-10-29 22:00:38 UTC
Created attachment 162706 [details]
Affected .KCrash №2.

Just got this again on cpe:/o:opensuse:tumbleweed:20231026 using https://download.opensuse.org/repositories/openSUSE:/Factory/standard/x86_64/drkonqi5-5.27.9-1.1.x86_64.rpm. Shall submit recording as proof subsequently.
Comment 12 Roke Julian Lockhart Beedell 2023-10-29 22:18:30 UTC
Created attachment 162707 [details]
Evidence.

(In reply to `{third: "Beedell", first: "Roke"}`{.JSON5} from comment #11)
> Created attachment 162706 [details]
> Affected .KCrash №2.
> 
> Just got this again on cpe:/o:opensuse:tumbleweed:20231026 using
> https://download.opensuse.org/repositories/openSUSE:/Factory/standard/x86_64/
> drkonqi5-5.27.9-1.1.x86_64.rpm. Shall submit recording as proof subsequently.

Potentially higher-quality version available at https://imgur.com/a/HRwRDhC.

Apologies for forgetting that I still had my webcam enabled. I can't be bothered to reupload without it. You'll have to ignore me dropping a pen in the background.
Comment 13 Roke Julian Lockhart Beedell 2023-10-29 22:28:39 UTC
(In reply to `{third: "Beedell", first: "Roke"}`{.JSON5} from comment #12)
> Created attachment 162707 [details]
> Evidence.
> 
> (In reply to `{third: "Beedell", first: "Roke"}`{.JSON5} from comment #11)
> > Created attachment 162706 [details]
> > Affected .KCrash №2.
> > 
> > Just got this again on cpe:/o:opensuse:tumbleweed:20231026 using
> > https://download.opensuse.org/repositories/openSUSE:/Factory/standard/x86_64/
> > drkonqi5-5.27.9-1.1.x86_64.rpm. Shall submit recording as proof subsequently.
> 
> Potentially higher-quality version available at https://imgur.com/a/HRwRDhC.
> 
> Apologies for forgetting that I still had my webcam enabled. I can't be
> bothered to reupload without it. You'll have to ignore me dropping a pen in
> the background.

Luckily, the issue I was attempting to report was constantly reproducible, so it was easy to test in this case too. Due to this, I've discovered that the issue occurs only the first time that DrKonqi downloads a debuginfo package. Subsequent identical crashes activate the debug symbols correctly. For evidence, see
Comment 14 Roke Julian Lockhart Beedell 2023-10-29 22:29:09 UTC
> For evidence, see https://bugs.kde.org/show_bug.cgi?id=476306#c0
Comment 15 Harald Sitter 2023-12-05 03:53:25 UTC
Please don't necro ancient bug reports.

The reason this happened to you is because you did a live update without re-login, removing files out from under the process and thereby making the process untraceable. Note this line from your kcrash file:

                #5  0x00007f6b574eeb01 n/a (/usr/lib64/libkwin.so.5.27.8 (deleted) + 0x2eeb01)