On Thu, Jun 9, 2011 at 11:41 AM, Ingo Weinhold <ingo_weinhold@xxxxxx> wrote: > The problem is that getting the looper (Target()) and locking it > (looper->Lock()) is not an atomic operation. In between the looper could be > deleted and another looper could be created with the same address. Hence the > necessity to recheck whether the correct looper has been locked. Understood, will look into that tonight then. Thanks! Regards, Rene