Summary: | Applying cubism filter to a layer makes Krita crash happily | ||
---|---|---|---|
Product: | [Applications] krita | Reporter: | sinozzuke <sinozzuke> |
Component: | General | Assignee: | Halla Rempt <halla> |
Status: | RESOLVED FIXED | ||
Severity: | crash | CC: | cberger, halla, lambdae2 |
Priority: | NOR | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Unlisted Binaries | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: | |||
Attachments: | simple test file, there's nothing special in it really |
Description
sinozzuke
2009-10-19 19:11:50 UTC
I'm afraid I cannot reproduce with 8 bits rgb. Which colorspace/bit depth did you use and what image size? "KoMixColorsOpImpl<KoRgbTraits<unsigned char> >::mixColors" would indicate RGB 8bit. If Gaizka used a different color space, then there is so misassumption in the cubism filter. I received Gaizka's file last night, and unfortunately, I cannot make it crash :-( have you try decresing to 5 the cubes? On Tue, Oct 20, 2009 at 9:37 AM, <boud@valdyas.org> wrote: > https://bugs.kde.org/show_bug.cgi?id=211116 > > > > > > --- Comment #3 from <boud valdyas org> 2009-10-20 09:37:42 --- > I received Gaizka's file last night, and unfortunately, I cannot make it > crash > :-( > > -- > Configure bugmail: https://bugs.kde.org/userprefs.cgi?tab=email > ------- You are receiving this mail because: ------- > You reported the bug. > Yes... And that doesn't matter much. I see that the progress bar is broken, though. Memcheck doesn't give any hints either :-( Perhaps is my instalation o my system configuration? On Tue, Oct 20, 2009 at 12:45 PM, <boud@valdyas.org> wrote: > https://bugs.kde.org/show_bug.cgi?id=211116 > > > > > > --- Comment #6 from <boud valdyas org> 2009-10-20 12:44:58 --- > Memcheck doesn't give any hints either :-( > > -- > Configure bugmail: https://bugs.kde.org/userprefs.cgi?tab=email > ------- You are receiving this mail because: ------- > You reported the bug. > I can reproduce: 1. I start a 8bits rgb 1024x768 white image 2. add a new layer 3. draw something with the brush in the new layer, I just drew a zigzag line 4. open cubism filter dialog and hit "apply and close" with defaults values 5. sometimes it crashes, if doesn't, repeat from step 2 I'm using kubuntu 9.10 beta, kde 4.3.2 and qt 4.5.2 My backtrace doesn't seem the same though (gdb) r Starting program: /home/m4v/dev/kde/install/koffice/trunk/bin/krita [Thread debugging using libthread_db enabled] [Nuevo Thread 0xb25d8b70 (LWP 6589)] X Error: BadAtom (invalid Atom parameter) 5 Major opcode: 20 (X_GetProperty) Resource id: 0x0 Enchant dict for "en_US" 0x8e08338 Enchant dict for "en_US" 0x8e08338 Enchant dict for "en_US" 0x8e08338 [Nuevo Thread 0xb1739b70 (LWP 6590)] [Nuevo Thread 0xb0f38b70 (LWP 6591)] [Thread 0xb0f38b70 (LWP 6591) terminado] [Thread 0xb1739b70 (LWP 6590) terminado] [Nuevo Thread 0xb1739b70 (LWP 6592)] [Nuevo Thread 0xb0f38b70 (LWP 6593)] [Nuevo Thread 0xb05ffb70 (LWP 6594)] [Thread 0xb05ffb70 (LWP 6594) terminado] [Thread 0xb1739b70 (LWP 6592) terminado] [Thread 0xb0f38b70 (LWP 6593) terminado] [Nuevo Thread 0xb0f38b70 (LWP 6595)] [Thread 0xb0f38b70 (LWP 6595) terminado] KisImage(0x8a763b8) QLayout: Attempting to add QLayout "" to QWidget "KritaShape/KisToolBrushoption widget", which already has a layout QCoreApplication::postEvent: Unexpected null receiver QCoreApplication::postEvent: Unexpected null receiver QGridLayout: Multi-cell fromRow greater than toRow [Nuevo Thread 0xb0f38b70 (LWP 6599)] [Thread 0xb0f38b70 (LWP 6599) terminado] [Nuevo Thread 0xb0f38b70 (LWP 6600)] Program received signal SIGSEGV, Segmentation fault. 0x015ca011 in QWidget::setEnabled(bool) () from /usr/lib/libQtGui.so.4 (gdb) bt #0 0x015ca011 in QWidget::setEnabled(bool) () from /usr/lib/libQtGui.so.4 #1 0x00195891 in KisFilterDialog::apply (this=0x91a4470) at /home/m4v/dev/kde/src/koffice/trunk/krita/ui/dialogs/kis_dlg_filter.cpp:128 #2 0x001964cb in KisFilterDialog::qt_metacall (this=0x91a4470, _c=QMetaObject::InvokeMetaMethod, _id=2, _a=0xbfffd6bc) at /home/m4v/dev/kde/build/koffice/trunk/krita/ui/kis_dlg_filter.moc:77 #3 0x0074f263 in QMetaObject::activate(QObject*, int, int, void**) () from /usr/lib/libQtCore.so.4 #4 0x0074fec2 in QMetaObject::activate(QObject*, QMetaObject const*, int, void**) () from /usr/lib/libQtCore.so.4 #5 0x01c00c67 in QAbstractButton::pressed() () from /usr/lib/libQtGui.so.4 #6 0x019290f9 in ?? () from /usr/lib/libQtGui.so.4 #7 0x019292f1 in QAbstractButton::mousePressEvent(QMouseEvent*) () from /usr/lib/libQtGui.so.4 #8 0x015c8ff6 in QWidget::event(QEvent*) () from /usr/lib/libQtGui.so.4 #9 0x019293ee in QAbstractButton::event(QEvent*) () from /usr/lib/libQtGui.so.4 #10 0x019d366d in QPushButton::event(QEvent*) () from /usr/lib/libQtGui.so.4 #11 0x01573f54 in QApplicationPrivate::notify_helper(QObject*, QEvent*) () from /usr/lib/libQtGui.so.4 #12 0x0157c033 in QApplication::notify(QObject*, QEvent*) () from /usr/lib/libQtGui.so.4 #13 0x01293bfa in KApplication::notify(QObject*, QEvent*) () from /usr/lib/libkdeui.so.5 #14 0x007396cb in QCoreApplication::notifyInternal(QObject*, QEvent*) () from /usr/lib/libQtCore.so.4 #15 0x0157af6e in QApplicationPrivate::sendMouseEvent(QWidget*, QMouseEvent*, QWidget*, QWidget*, QWidget**, QPointer<QWidget>&) () from /usr/lib/libQtGui.so.4 #16 0x015e97c0 in ?? () from /usr/lib/libQtGui.so.4 #17 0x015e8c4b in QApplication::x11ProcessEvent(_XEvent*) () from /usr/lib/libQtGui.so.4 #18 0x01616050 in ?? () from /usr/lib/libQtGui.so.4 #19 0x00737c79 in QEventLoop::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/libQtCore.so.4 #20 0x007380ca in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/libQtCore.so.4 #21 0x0073a53f in QCoreApplication::exec() () from /usr/lib/libQtCore.so.4 #22 0x01573dd7 in QApplication::exec() () from /usr/lib/libQtGui.so.4 #23 0x009e6bfd in kdemain (argc=1, argv=0xbffff5b4) at /home/m4v/dev/kde/src/koffice/trunk/krita/main.cc:49 #24 0x0804886b in main (argc=1, argv=0xbffff5b4) at /home/m4v/dev/kde/build/koffice/trunk/krita/krita_dummy.cpp:3 Created attachment 37690 [details]
simple test file, there's nothing special in it really
open and apply cubism in the topmost layer, usually it doesn't crash the first time, but then just undo and try again (which usually success in crashing)
it doesn't work with the "apply cubism again" option in the filter menu, always try with filter dialog
after some more testing, the crash is unrelated to cubism, I got the same backtrace with 3 different filters. I think this happen when you apply the filter while the preview filter mask is still getting updated. And that the bug is more general than just cubism filter, I remember having a similar issue with the blur filter, not too long ago. This part of the backtrace seems to confirm my intuition: #3 0x00007fdc363bda76 in KisCubismFilter::process (this=0xe07650, srcInfo={d = 0x7fdc2a87d190}, dstInfo=<value optimized out>, size=@0x7fdc2a87d1a0, configuration=<value optimized out>, progressUpdater=<value optimized out>) at /home/sinozzuke/kde4/src/koffice/krita/plugins/filters/cubismfilter/kis_cubism_filter.cpp:108 #4 0x00007fdc56a8c079 in KisFilterMask::decorateRect (this=0x6da8590, src=<value optimized out>, dst=@0x7fdc2a87d3d0, rc=@0x7fdc2a87d510) at /home/sinozzuke/kde4/src/koffice/krita/image/kis_filter_mask.cpp:124 #5 0x00007fdc56aae26a in KisMask::apply (this=0x6da8590, projection={d = 0x7fdc2a87d5c0}, rc=@0x7fdc2a87d510) at /home/sinozzuke/kde4/src/koffice/krita/image/kis_mask.cc:114 #6 0x00007fdc56aab2b3 in KisLayer::applyMasks (this=0xc993c0, source=<value optimized out>, destination={d = 0x7fdc2a87d6a0}, requestedRect=<value optimized out>) at /home/sinozzuke/kde4/src/koffice/krita/image/kis_layer.cc:357 #7 0x00007fdc56aab7f5 in KisLayer::updateProjection (this=0xc993c0, rect=<value optimized out>) at /home/sinozzuke/kde4/src/koffice/krita/image/kis_layer.cc:417 A possible fix would be to make sure the filter dialog cancel the preview and then wait for the preview to be done with filter mask. It's confirmed then. Can't reproduce at the moment, but on the other filter do some strange things at the moment. SVN commit 1098752 by rempt: Cubism filter: don't install Still compile it, so it won't bitrot and will be ready for moving to extensions after 2.1, but don't install it anymore since it doesn't fit our purpose. BUG:211116 M +2 -2 CMakeLists.txt WebSVN link: http://websvn.kde.org/?view=rev&revision=1098752 the bug wasn't related to cubism, since I got the same crash with other filters... but I can't reproduce anymore anyway |