Bug 164482 - Writer: Moving arrow handles crashes writer
Summary: Writer: Moving arrow handles crashes writer
Status: RESOLVED WORKSFORME
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Writer (show other bugs)
Version:
(earliest affected)
24.8.4.2 release
Hardware: All All
: medium normal
Assignee: Not Assigned
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2024-12-27 14:45 UTC by BDF
Modified: 2025-07-09 06:37 UTC (History)
2 users (show)

See Also:
Crash report or crash signature:


Attachments
arrow handle crash video example (3.92 MB, video/webm)
2024-12-27 14:50 UTC, BDF
Details
arrow crash test file (10.43 KB, application/vnd.oasis.opendocument.text)
2024-12-27 14:52 UTC, BDF
Details

Note You need to log in before you can comment on or make changes to this bug.
Description BDF 2024-12-27 14:45:41 UTC
Description:
When I include an arrow and adjust it's direction by clicking on it's handle and dragging it around with the mouse, writer very often crashes.
Sometimes it takes a bit of time, but when you have an arrow that crashes writer you can crash writer repeatedly. 

Steps to Reproduce:
1. Include an arrow in Writer (included via the UI for Arrow Styles at the bottom)
2. Grab a handle of the arrow and move it around until writer crashes

Actual Results:
LibreOffice crashes

Expected Results:
LibreOffice does not crash


Reproducible: Always


User Profile Reset: No

Additional Info:
In the video I used the arrows and it's handles to position them over images (maybe this has an effect). But maybe it's just easier to reproduce it that way as I could cause the same crash with a new writer file where I just included an arrow and moved it around.
HOWEVER (as there will be at least once comment with "works for me"), it sometimes does work fine without any problem and I simply can't get LibreOffice to crash. So to reproduce the bug you probably have to include multiple arrows and move each of their handles around until one of them makes LibreOffice crash.

SYSTEM INFORMATION
-vVersion: 24.8.4.2 (X86_64) / LibreOffice Community
- Build ID: bb3cfa12c7b1bf994ecc5649a80400d06cd71002
- CPU threads: 12; OS: Linux 6.8; UI render: default; VCL: gtk3
- Locale: de-AT (de_AT.UTF-8); UI: de-DE
- Flatpak
- Calc: threaded
Comment 1 BogdanB 2024-12-27 14:50:28 UTC Comment hidden (obsolete)
Comment 2 BDF 2024-12-27 14:50:30 UTC
Created attachment 198286 [details]
arrow handle crash video example

A video where the action and the consecutive crash was recorded.

If you have an arrow that repeatedly crashes writer, you can move it around once very shortly. After that you can move the handle around much longer without writer crashing. (Part after the second example)
Comment 3 BogdanB 2024-12-27 14:51:07 UTC
Also to be certain the reported issue is not
related to corruption in the user profile, could you please reset your
Libreoffice profile (<https://wiki.documentfoundation.org/UserProfile>) and
re-test?
Comment 4 BDF 2024-12-27 14:52:00 UTC
Created attachment 198287 [details]
arrow crash test file

A test file with which writer crashes reproducibly.

I grab the right handle of the arrow and just move it around until writer crashes.
Comment 5 BogdanB 2024-12-27 14:53:07 UTC
I have seen your video, I can NOT reproduce, using a similar image and a similar arrow.
Comment 6 BDF 2024-12-27 14:55:46 UTC
(In reply to BogdanB from comment #1)

With my bugreports you always have to wait a few minutes until I uploaded all my files and wrote some documentation as I try to be as detailed as possible and providing images or videos or the issue and also providing a sample file.

In this case I provided a sample file, but it shouldn't be needed after all as I could also case the crash with a brand new writer file.
Comment 7 BDF 2024-12-27 15:02:33 UTC
(In reply to BogdanB from comment #3)
> Also to be certain the reported issue is not
> related to corruption in the user profile, could you please reset your
> Libreoffice profile (<https://wiki.documentfoundation.org/UserProfile>) and
> re-test?

I just tried to reproduce this in the safe mode.
It seems to be less common*, but it's still possible and I get the same crash.

* on "less common": As said, if the bug appears or not does seem to be random. Sometimes it works just fine and then the exact same arrow (that has been working before just fine) causes writer to crash.
Comment 8 Julien Nabet 2024-12-27 16:02:27 UTC
On pc Debian x86-64 with master sources updated today or with LO Debian package 24.8.4.2, both with gtk3 rendering, I don't reproduce this.
(I tested about a dozen of times)

Would it be possible you retrieve a backtrace? (see https://wiki.documentfoundation.org/QA/BugReport/Debug_Information#GNU/Linux:_How_to_get_a_backtrace)
Comment 9 BDF 2024-12-27 16:37:21 UTC
Is there any more information that I could provide?

I heard that there is something like a traceback (or trackback ?) that can help a lot with bugs, but

a) I have never worked with this and therefore I have no idea how to do this (though I can look into doing it)
b) I have even less of an idea how this could be done under LibreOffice.
Comment 10 Julien Nabet 2024-12-27 16:43:09 UTC
(In reply to BDF from comment #9)
> Is there any more information that I could provide?
> 
> I heard that there is something like a traceback (or trackback ?) that can
> help a lot with bugs, but
> 
> a) I have never worked with this and therefore I have no idea how to do this
> (though I can look into doing it)
> b) I have even less of an idea how this could be done under LibreOffice.

As indicated a backtrace may help to give some hints.
The link provided in my previous comment explains how to do this.
It's called "backtrace" but you may encounter too "stacktrace".
Comment 11 BDF 2024-12-31 22:03:29 UTC
(In reply to Julien Nabet from comment #10)
> As indicated a backtrace may help to give some hints.
> The link provided in my previous comment explains how to do this.
> It's called "backtrace" but you may encounter too "stacktrace".

I'll try to get this done.
However, I have never done this before and I have absolutely no idea how this works or what I need, so this will take some time.
Additionally, the next three weeks will be filled with learning for an exam, so don't expect an answer all too soon.
Comment 12 QA Administrators 2025-01-01 03:11:48 UTC Comment hidden (obsolete)
Comment 13 BogdanB 2025-01-01 06:56:26 UTC
BDF, there is no hurry, when you have the backtrace, you can change the bug status to Unconfirmed.
Comment 14 BDF 2025-06-30 22:46:26 UTC
Thankfully there was no hurry as university kept me busy until the end of last month.
Today I had a whole day to do all the testing and I have good and bad news.


# The bad one news are that even after several hours of trying I wasn't able to setup a debugging environment on my system and run it.

From what I understood the main problem was that I was using the flatpack version [1] of LibreOffice and that the sandboxed nature of flatpacks were messing with gdb and --backtrace.
I tired to solve with this with the help of ChatGPT and tested everything I could, but with no luck. As I had no idea what the problem was, I asked ChatGPT to summarise the problem:

o) gdb and --backtrace Issue: I initially tried to use gdb with the --backtrace option, but received the error: “Can't find the tool 'gdb', --backtrace option will be ignored.” I resolved this by installing the necessary debug symbols (flatpak install org.libreoffice.LibreOffice.Debug) and enabling debuginfod to download the required symbols.

o) Running flatpak run inside gdb: When attempting to run the application directly with gdb, I encountered an issue with the --backtrace flag, as flatpak didn’t seem to pass it properly to the LibreOffice binary. The application also didn’t consistently reproduce the crash, making it difficult to capture a backtrace during a live session.

Because I was testing on a live system, I wasn't to worried breaking or messing up anything so I also installed the package from the ubuntu repo (24.2.7.2 [1]). But there I couldn't get the gdbtrace.log for whatever reason.


# The good news is that the bug seems to no longer be an issue as I can no longer reproduce it. Not on my test live USB as well as on my system I saw it the encountered it the first time. Neither in my uploaded demo file nor in my arrow heavy file that once was the reason I filed this bug report.

As I had pretty much no time for testing in the last months I can't tell you if the "fix" is related to the ubuntu base, the KDEneon changes on top, changes in the Plasma desktop or if maybe a change in LibreOffice did the trick.
I may can recheck on an older system (eg. with plasma 6.1/2/3) but I'm not sure if there would be a great interest in this as a) nobody was able to reproduce the bug and b) who would install an old version of LibreOffice or Ubuntu if it's not an LTS version?


Long story short: As I can no longer reproduce the bug I will close this bug report but will open it again once it pops up again (and then sink even more hours into finding out how to debug with flatpack)

@BogdanB: Thank you for your patience
@Julien Nabet: Thank you for the link to the site explaining backtrace though I was not able to get one

[1] Version: 25.2.4.3 (X86_64) / LibreOffice Community
Build ID: 33e196637044ead23f5c3226cde09b47731f7e27
CPU threads: 12; OS: Linux 6.11; UI render: default; VCL: gtk3
Locale: de-AT (de_AT.UTF-8); UI: de-DE
Flatpak
Calc: threaded

[2] Version: 24.2.7.2 (X86_64) / LibreOffice Community
Build ID: 420(Build:2)
CPU threads: 16; OS: Linux 6.11; UI render: default; VCL: gtk3
Locale: en-US (C.UTF-8); UI: en-US
Ubuntu package version: 4:24.2.7-0ubuntu0.24.04.4
Calc: threaded
Comment 15 Julien Nabet 2025-07-09 06:37:55 UTC
Let's rather put WFM since there's no fix per se.