So, are we still having a race condition, like when a linger timeout occur at the same time as a pipe is attached ? It seems to me that this can happen in two different threads and so likely to trigger one of the two zmq_assert (!is_terminating() in process_attach or !pipe in ~session_t), unless we can be sure that they can't both happen at the same time.
Fabien _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
