Bug 386003 - Don't submit crash reports that lack debug symbols for the crashing stack frame
Summary: Don't submit crash reports that lack debug symbols for the crashing stack frame
Status: RESOLVED WORKSFORME
Alias: None
Product: drkonqi
Classification: Applications
Component: backtraceparsing (show other bugs)
Version: 5.11.1
Platform: Other Linux
: NOR normal
Target Milestone: ---
Assignee: Unassigned bugs mailing-list
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2017-10-20 20:01 UTC by Nate Graham
Modified: 2020-01-30 15:12 UTC (History)
1 user (show)

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


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Nate Graham 2017-10-20 20:01:21 UTC
Drkonqi is pretty good at analyzing backtraces to determine whether they're useful or not. Yet still, we get crash reports with useless backtraces with no symbols for the problematic stack frame, especially in Arch and Arch-based distros.

These crash reports just aren't useful and add to the burden of screening bugs, decreasing the signal-to-noise ratio. This contributes to actionable bugs getting missed or neglected because bug screening resources are VERY scarce--nobody particularly enjoys screening bugs, except for a few weirdos like Christoph and me. :)

It would help if we could make drkonqi a bit pickier here, both to ease bug the screening burden, and also to encourage users to install and distros to provide debug packages.
Comment 1 Harald Sitter 2019-06-14 14:46:29 UTC
Do we have examples of traces that should not be reported?

I imagine this is impossible to pull off reliably. The top frame is not necessarily the important one. It often is, but definitely not always, and sometimes you may even deduce what the top frame is based on the preceding ones.
Comment 2 Nate Graham 2019-06-14 14:52:21 UTC
Hmm, that's true.

Feel free to close this if you don't think it would add any value.
Comment 3 Harald Sitter 2019-06-14 15:11:25 UTC
I'd rather we collect some examples if possible. Maybe we can identify something that allows us to detect their poor value, other than the top frame being garbage.

e.g. If the trace as a whole is scoring poorly we could possibly weigh the top frame a bit more so it actually tips the scale towards not-useful.

Hard to say if that'd help without examples though :)
Comment 4 Harald Sitter 2019-11-05 13:34:47 UTC
Did you see this popping up again?
Comment 5 Bug Janitor Service 2019-11-20 04:33:09 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 6 Bug Janitor Service 2019-12-05 04:33:09 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 7 Nate Graham 2020-01-30 01:06:20 UTC
I mean, we could just blanket-discard reports with no debug symbols. :)
Comment 8 Harald Sitter 2020-01-30 15:12:08 UTC
That is what the rating system is meant to do, which is why I'd like examples of crappy backtraces that got through.
It may be a simple matter of weighting the first frame by .5 more or something. Really hard to say without some data to reproduce in a controlled environment and check what the actual scoring of the lines is though.