https://bugs.kde.org/show_bug.cgi?id=369014
--- Comment #16 from Ryokun <renaudpi...@free.fr> --- Hello, Clever thing to ask but unfortunately the issue doesn't seem to be with Sierra. I explain myself: I currently own 3 macs really different the one from the other. An iMac 2009 Core2 Duo on Mac OS 10.11, a MacBook Air 2013 (2 cores) on Mac OS 10.12 and an iMac 2015 5K (retina) Core-i5 (4 cores) on Mac OS 10.12. I made a test just now and here are the results: - iMac 2015 (Mac os 10.12): freeze 2 times on 3 - MacBook Air (Mac os 10.12): No freezes !!!!! Doesn't matter how many times I try, it doesn't hang at all. - iMac 2009 (Mac OS 10.11): freeze 1 time on 5 I don't get a bug report each time but I had one just now on the 2009 iMac and it was exactly the same I posted yesterday (that one was from the iMac 2015). I'm myself a developer (though I don't know anything about QT) but when I look at the stacktrace it shows that it's calling an Import/export method that seems to destroy an object (widget here, so I suppose that's an UI element) that has already been freed. So my educated guess would be that another thread is destroying the object BEFORE it and so the main thread can't destroy it when asked. The computers that doesn't show this behavior surely destroy the object in the main thread before the background thread is able to do it (why doesn't the background thread crash Krita ? Maybe there's a check in it to destroy the object only if it isn't NULL or maybe the program doesn't catch this type of bug when on a background thread, I can't be sure without debugging in detail). This type of behavior is really typical of a thread problem so my money's on it. But I don't want to assume too many things too because I never seen 1 line of code from Krita. I really would like to help but my hands are tied for the next 3 to 5 days so I won't be any help before that. After that, if you're ok with it, I'll gladly take a look at the code (even if, again, I don't know anything at QT. I somewhat can manage C++ so it shouldn't be too problematic...). Thanks. -- You are receiving this mail because: You are watching all bug changes.
