https://bugs.kde.org/show_bug.cgi?id=388228
--- Comment #9 from Albert Astals Cid <aa...@kde.org> --- (In reply to Tobias Deiminger from comment #7) > Thanks again. I'll try to fix the problem as it occurs in code from a patch > I did. > > > Looks like a corrupted vtable. > > Nope, dangling pointer... The __cxa_pure_virtual call is a consecutive fault > of that. Your trace shows the annotation object got deleted during undo of > AddAnnotationCommand. > > ==30208== Address 0x27a2b9f0 is 0 bytes inside a block of size 16 free'd > ==30208== at 0x4C3123B: operator delete(void*) (in > /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) > ==30208== by 0x1EC658E2: > Okular::AddAnnotationCommand::~AddAnnotationCommand() > (documentcommands.cpp:72) > ==30208== by 0x1EC65928: > Okular::AddAnnotationCommand::~AddAnnotationCommand() > (documentcommands.cpp:74) > ==30208== by 0x71FC861: QUndoStack::push(QUndoCommand*) (in > /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.9.1) > ==30208== by 0x1EC49A24: Okular::Document::addPageAnnotation(int, > Okular::Annotation*) (document.cpp:3230) > [...] > > MouseAnnotation did not notice the deletion and continued to track the now > deleted annotation. On next access (in MouseAnnotation::cursor) we can crash > with various kinds of errors, depending on what has happened in the meantime > to the freed memory. Is this problem still existing? Can you reproduce it? I wonder why it works for me and I can't get valgrind to complain at all :S -- You are receiving this mail because: You are the assignee for the bug.
