Deadlock using @SuperStep with inEventQueue test

[rherrmannr]

1. open inEventQueue.sct
2. add @SuperStep
3. raise event f at least once
4. raise event e -> Deadlock

[tkutz]

There is a cycle in the statechart, so the interpreter runs forever, because in superstep semantic the event e is not cleared directly:

Cycle detection is hard as we could also have guard expressions involved, like this one in state L:

But at least we should make the simulation cancelable, and not run in the UI thread directly.