Bug 33407 - FILEOPEN Crash on look last page.
Summary: FILEOPEN Crash on look last page.
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Writer (show other bugs)
(earliest affected)
3.3.0 RC4
Hardware: x86 (IA32) Windows (All)
: medium critical
Assignee: Don't use this account, use tml@iki.fi
: 34785 (view as bug list)
Depends on:
Reported: 2011-01-24 06:09 UTC by Dmitry Lujetsky
Modified: 2011-03-06 21:45 UTC (History)
2 users (show)

See Also:
Crash report or crash signature:


Note You need to log in before you can comment on or make changes to this bug.
Description Dmitry Lujetsky 2011-01-24 06:09:34 UTC
attachment have file with error.
Crash on look last page.
Comment 1 Noel Power 2011-01-24 07:46:09 UTC
think you forgot to attach the file :-), can you do that please
Comment 2 Dmitry Lujetsky 2011-01-24 07:57:13 UTC
This archive with error document.

Please open and browse.
Comment 3 tester8 2011-01-24 08:24:50 UTC
Not reproduced with LO RC4 Ubuntu 10.04 x86.
Comment 4 Noel Power 2011-01-24 08:34:01 UTC
yup doesn't crash on linux for me but crashes just fine on windows, tor can you investigate further?
Comment 5 Don't use this account, use tml@iki.fi 2011-02-07 05:40:56 UTC
Please attach the document to this bug, don't provide a link to a Russian language webpage.
Comment 6 Dmitry Lujetsky 2011-02-07 05:53:31 UTC
I am can not attach the document.

"The file you are trying to attach is 2859 kilobytes (KB) in size. Non-patch attachments cannot be more than 1000 KB.
We recommend that you store your attachment elsewhere and then insert the URL to it in a comment, or in the URL field for this bug.
Alternately, if your attachment is an image, you could convert it to a compressible format like JPG or PNG and try again."

Please try this link
Comment 7 Don't use this account, use tml@iki.fi 2011-02-07 06:19:43 UTC
Thanks, got the file.
Comment 8 Don't use this account, use tml@iki.fi 2011-02-07 06:43:02 UTC
For me, in a partial debugging build of 3.3, it crashes right when opening the document, before showing it at all. The backtrace at the crash is:

>	swmi.dll!SwNodeIndex::SwNodeIndex(const SwNodeIndex & rIdx={...}, long nDiff=0x00000000)  Line 84 + 0x6 bytes	C++
 	swmi.dll!SwPosition::SwPosition(const SwPosition & rPos={...})  Line 101 + 0x15 bytes	C++
 	swmi.dll!SwPaM::SwPaM(const SwPosition & rPos={...}, SwPaM * pRing=0x00000000)  Line 405 + 0x2d bytes	C++
Comment 9 Don't use this account, use tml@iki.fi 2011-02-07 06:59:42 UTC
When using also a mswordmi.dll built with debugging, I get lots of assertion failures from sw/source/core/txtnode/ndhints.cxx, line 389: "HintsCheck: Portion inconsistency". (I mean lots like maybe a hundred, I kept clicking my mouse buttont as fast as I could for maybe ten seconds, at this same assertion message window.)

Although, infuriatingly, the assertion message continues, "This can be temporarily ok during undo operations." Oh how I hate this kind of wishy-washy assertions. If I click "No" on them all (to ignore the assertion), eventually the start of the document does show up. Weird, does building mswordmi.dll with debugging change its behaviour?

Anyway, if I then scroll towards the end of the document, I get a "Unhandled exception at 0x7c812afb in soffice.bin: Microsoft C++ exception: com::sun::star::loader::CannotActivateFactoryException"

The stack trace at this point is:
>	swmi.dll!SwRects::_resize(unsigned int n=0x00000000)  Line 48 + 0x3d bytes	C++
 	swmi.dll!SwRects::Remove(unsigned short nP=0x0000, unsigned short nL=0x0001)  Line 48 + 0x105 bytes	C++
 	swmi.dll!SwRegionRects::operator-=(const SwRect & rRect={...})  Line 151	C++
 	swmi.dll!SwLayoutFrm::RefreshLaySubsidiary(const SwPageFrm * pPage=0x0145d758, const SwRect & rRect={...})  Line 5983 + 0xf bytes	C++
 	swmi.dll!SwFlyFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * __formal=0x00000000)  Line 3746	C++
 	swmi.dll!SwFlyCntPortion::Paint(const SwTxtPaintInfo & rInf={...})  Line 269	C++
 	swmi.dll!SwTxtPainter::DrawTextLine(const SwRect & rPaint={...}, SwSaveClip & rClip={...}, const unsigned char bUnderSz=0x00)  Line 423	C++
 	swmi.dll!SwTxtFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * __formal=0x00000000)  Line 734	C++
 	swmi.dll!SwLayoutFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * __formal=0x00000000)  Line 3313	C++
 	swmi.dll!SwBodyFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * __formal=0x00000000)  Line 206	C++
 	swmi.dll!SwLayoutFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * __formal=0x00000000)  Line 3313	C++
 	swmi.dll!SwRootFrm::Paint(const SwRect & rRect={...}, const SwPrtOptions * pPrintData=0x00000000)  Line 2993	C++
 	swmi.dll!ViewShell::Paint(const Rectangle & rRect={...})  Line 1719	C++
 	swmi.dll!SwCrsrShell::Paint(const Rectangle & rRect={...})  Line 1184	C++
 	swmi.dll!SwEditWin::Paint(const Rectangle & rRect={...})  Line 590	C++

I think some other recent bug also involved the SV_IMPL_VARARR( SwRects, SwRect ) call in swregion.cxx?

Cedric, any clue from this?
Comment 10 Cédric Bosdonnat 2011-02-28 02:32:04 UTC
Working in 3.3.1. This was probably related to the EMF-related crash on windows builds.
Comment 11 Rainer Bielefeld Retired 2011-03-06 21:45:51 UTC
*** Bug 34785 has been marked as a duplicate of this bug. ***