https://bugs.kde.org/show_bug.cgi?id=428014
--- Comment #12 from Dmitry Kazakov <dimul...@gmail.com> --- Hi, Deif Lou! What happens in this crash is the following: 1) The GUI thread calls KisGeneratorLayer::requestUpdateJobsWithStroke(), which calls resetCacheWithoutUpdate(), which, in turn, calls m_d->paintDevice->clear(). 2) At the same time, there is a generator stroke running in the background and writes the result of the filter to the same device. It results in two simultaneous write accesses to the paint device, which is not allowed by the data manager. As far as I remember, this situation should have been guarded by `m_d->updateCookie`, I'm not sure why it doesn't work here. -- You are receiving this mail because: You are watching all bug changes.
