Session cleanup happens usually when the application is being destroyed. By not killing the pending jobs quietly, the jobs emit their result() signal which can be connected to a slot that can try to access objects that are currently being destroyed, thus causing weird crashes (e.g. https://bugs.kde.org/show_bug.cgi?id=387177).

We should consider killing jobs quietly during Session teardown as to prevent them from emitting result() and thus triggering unexpected behaviors. It's safe to assume that at this point no one is interested in the result()/finished() of those jobs anyway.