https://bugs.kde.org/show_bug.cgi?id=486178
--- Comment #1 from Thiago Macieira <thi...@kde.org> --- I'd say "vncplugin doesn't work at all, ever, under any conditions, if you try to use it" is sufficient for "critical", but ok for "grave" BTW, I tried to fix that function myself. I can apply a workaround to just restore the previous behaviour, but that's papering over the sun with a sieve. I really don't understand what's going on and what is being locked. Is it trying to protect m_stopped? Why not use an atomic instead? Or, better, QThread::isInterruptionRequested()? But on line 549, it accesses m_stopped without a mutex, so I don't know if it needed protecting in the first place. Ditto for m_passwordError: why not an atomic? Or why not use a local variable and only lock to save the state when exiting the function? This code should be rewritten with closer locks, not that one QMutexLocker at the top that is used incorrectly. -- You are receiving this mail because: You are watching all bug changes.
