Created attachment 61206 [details] mac os x crash report I run LibreOffice 3.5.2 on Mac OS X 64 bit. In some case LibreOffice crashes, when adding a frame on cells. Please find attached the automatic generated chrash report from mac os x.
@Thorsten: Could you maybe try this once with memcheck on Mac. Maybe a memory v-table problem when the called virtual method is no longer accessible because the destructor has been called already.
Thank you very much for your bug report! According to the crash log, this is the same iusse as bug 49942; cf. also https://bugs.freedesktop.org/show_bug.cgi?id=51791#c9 where I argue that this kind of crash can happen when adding borders to cells, just like in the present bug report. @ phartzsch: This crash is/was probably caused by the activation of some accessibility stuff in Mac OS X. So, to solve the problem for you (and to help us finding a fix for it), please check the following: 1) Do you have any accesibility features enabled? Apple’s accessibility features like “VoiceOver” or “Enable access for assistive devices”, which get enabled in “System Preferences > Universal Access” (in German: “Bedienungshilfen”), are known to cause many crashes in LibreOffice (see bug 47368). So please try to disable any accesibility features. 2) Do you have installed any window management/user interface utilities/apps for MacOS X like * Moom (see bug 42014) * Cinch (see bug 51791) * RightZoom (see bug 51686) * ShiftIt (see bug 52147) * BetterTouchTool * BetterSnapTool ... All these and many similar utilities rely heavily on accessibility issues and therefore cause LibreOffice to crash. So please check if you have installed any utility of this kind and try to disable it (or to add LibreOffice to the list of excluded applications for the utility, if there is such a thing). Thank you very much for trying this out -- and please report the results here, too! *** This bug has been marked as a duplicate of bug 49942 ***
(In reply to comment #1) > @Thorsten: Could you maybe try this once with memcheck on Mac. Maybe a memory > v-table problem when the called virtual method is no longer accessible because > the destructor has been called already. @Thorsten: If you do this -- which would be very much appreciated! --, please report the results in bug 49942, which is almost certainly the same issue (cf. the stack traces). Thank you very much!
(In reply to comment #3) > > @Thorsten: Could you maybe try this once with memcheck on Mac. Maybe a memory > > v-table problem when the called virtual method is no longer accessible because > > the destructor has been called already. > > @Thorsten: > If you do this -- which would be very much appreciated! --, please report the > results in bug 49942, which is almost certainly the same issue (cf. the stack > traces). Thank you very much! Sorry, I forgot to mention: to reproduce this issue, you probably need to install Cinch, which seems to trigger this bug, cf. bug 49942 and bug 51791.
@Roman Cinch was the trouble maker! I have disabled Cinch an LO will not crash anymore when I add a frame to cells etc. Thank you very much!!!
@phartzsch: > Cinch was the trouble maker! I have disabled Cinch an LO will not crash anymore > when I add a frame to cells etc. Thank you for your answer! Your result confirms that this is really the same issue as bug 49942 ... which is important to know for us, of course.
(In reply to comment #6) > @phartzsch: > > > Cinch was the trouble maker! I have disabled Cinch an LO will not crash anymore > > when I add a frame to cells etc. > > Thank you for your answer! Your result confirms that this is really the same > issue as bug 49942 ... which is important to know for us, of course. Hi Roman, how do you proceed now? Will this bug be fixed or have I try to deactivate/deinstall Cinch on my Mac OS X when run LO? Do you know an alternative to Cinch?
@Thorsten Behrens: (In reply to comment #3) > > @Thorsten: Could you maybe try this once with memcheck on Mac. Maybe a memory > > v-table problem when the called virtual method is no longer accessible because > > the destructor has been called already. > > @Thorsten: > If you do this -- which would be very much appreciated! --, please report the > results in bug 49942, which is almost certainly the same issue (cf. the stack > traces). Thank you very much! Good news -- this is no longer necessary! Bug 49942 was fixed by Tor’s patch for bug 47368, too, and therefore is no longer reproducible in LibO >= 3.6.2.1. So you don’t need to do any memcheck for this issue. But there are probably other issues when LibreOffice is used while Cinch is running (or some other window management utilities). I will take the liberty to inform you about them ;-)
@phartzsch: (In reply to comment #7) > how do you proceed now? Will this bug be fixed or have I try to > deactivate/deinstall Cinch on my Mac OS X when run LO? Do you know an > alternative to Cinch? Please see my survey results in bug 49942: https://bugs.freedesktop.org/show_bug.cgi?id=49942#c15 In short words, you can upgrade to LibreOffice 3.6.2 (a pre-relase is already available from http://www.libreoffice.org/download/pre-releases/ ) and this bug will be gone. But, as I state at the end of my comment, there is at least one issue with Cinch and LibreOffice which is still reproducible (I will do some further testing). So, if you do not need Cinch very much, it may be save to deactivate it, at least temporarily, when you are working mainly with LibreOffice for some hours. About alternatives to Cinch -- I am not an expert here, but the problem is that other Mac window management utilities like Moom, RightZoom, BetterSnapTool etc. all use Apple’s accessibility features, too, just like Cinch, so that they suffer from the same stability issues as Cinch. At the moment, I hope we have fixed the most important of these issues now, but I can not guarantee that there are not left some more ... :-( So, if you really like Cinch, just stay with it and report any crashes/hangs you still experience with LibreOffice while running Cinch -- it would help us very very much if for each such crash/hang you would immediately test if you can reproduce it without Cinch, too (then this is probably a normal LibreOffice bug), or only with Cinch running (then this bug would be related to Cinch/Mac accessibility features, so please inform me about this bug by adding my address: bugs@eikota.de to the CC list of the bug report). Thank you very much!