Created attachment 31832 [details]
presentation which crashes Open Office

Reproducible with final version but not with current internal version m142 anymore.

Closed. Thanks for your help.

It's still crashing with my test case in ooo-build from HEAD. Shreyas has
rebuilt vcl with debug symbols to try and narrow it down.

Off hand, can you say what the patch in m142 was? I can back port that to my
ooo-build (and to downstream distro) if I can isolate which patch it was.

AfC
Bangalore

Oh, yes. Thanks for insisting on this. Sometimes it needs a little time to crash
(build 8978 m142).

Reproducible. Reassigned. I will send a stack.

Sorry, should have pointed out that the bug kicks when the thumbnails are
drawing, either in the left hand pane of Normal view or when you switch to Slide
Sorter and the tiles draw. In Normal view, slide down a bit...

AfC

Philipp - sounds like some native gtk+ issue - havn't had a chance to look yet
but ...

worth doublechecking that it's not https://bugs.freedesktop.org/show_bug.cgi?id=5125

Created attachment 31942 [details]
All threads backtrace

Created attachment 31943 [details]
A better (?) stack trace. Note top call in Thread 1 to GetPaintObjectViewContact()

The ID of the error report is rxm6r8.

Thanks for the hints afcowie, I will reassign this issue to our slidesorter god

Have not been able to reproduce yet.

This bug may be related to issue 58769 and may be due to lifetime problems of
the drawing layer contact objects of the slide previews.

The attached bug document does contain linked graphics.  There is an internal
bug (127723) about a crash of the VacationPhotos example document.  The loading
of the linked bitmaps executes Reschedule() in a loop and leads so to problems
with asynchronous calls executed at the wrong time.

This crash is likely not caused by the the Reschedule() problem mentioned above
but may be due to life time problems fixed for the same bug.

Back to QA.

->WG: Can you please try to reproduce this crash on the CWS impress78?

I can not reproduce it.

I just tried it with OpenOffice 2.0.1 and it still crashes...

AfC
Sydney

So the issue should be reopened then.

Submitter and QA can reproduce the crash in OOo 2.0.1 and in the internal build
m142. It seems to be fixed in CWS impress78 that has target 2.0.2.
If the issue is fixed after 2.0.1 its status should be "fixed" or "duplicate"
instead of "WFM".

I checked this again on a m152 and had no problems. Impress78 was intergated in
m152 so from this build on there should not problems anymore. Do you have a
crash when opening the bugdoc in m152 or later? Thanks in advance.

I set the state to FIXED because neither I nor the QA can reproduce this crash
after the integration of cws impress78.  Most likely candidates for the fix are
issue 58769 and (internal) issue 127723.

Please verify.

re-open issue and reassign to wg@openoffice.org

reassign to wg@openoffice.org

reset resolution to FIXED

So rather than throwing the ball at each other, why don't you assign it to me
for verification?

Incidentally, OpenOffice terms like  m152 and Impress78 mean absolutely nothing
to normal users. Shit, I've built ooo from source and I can only barely follow.
I assume all that means that you expect the issue resolved (since you can't dup
it any more) and assuming these branches make it into OpenOffice 2.0.2 (how DOES
one know whether or not that is/will happen?) then that is what needs checking.

So I will look forward to OOo 2.0.2 and  REOPEN if it's still a crasher;
otherwise, thanks and well done.

AfC

cannot reproduce in OOo 2.0.2 anymore, so I'm closing

afcowie: please reopen, if you still see the crash

OpenOffice 2.0.2 just became available to us. After installing it I tried
opening that presentation. Sure enuogh, scroll down the slide thumbnailer a few
tiles and BANG down she goes. Still.

How do I reopen a bug here?

I will attach the current version of the presentation on the off chance that the
earlier one wasn't allowing you to duplicate.

AfC
Sydney

Created attachment 35277 [details]
Final version of the presentation that crashes oo.