Some processes were leaked after coap request

[wangzhenandy]

Hello，

I Found the processes of coap_channel_sup and coap_responder_sup can't be closed after coap request socket closed.

I test it like:

1. When a client request came, the server creates 4 processes: coap_channel_sup, coap_channel, coap_responder_sup, coap_responder;
2. Process coap_responder was closed first after finished the request;
3. Process coap_channel was closed when the client ack timeout;
4. Processes coap_channel_sup and coap_responder_sup could't closed;
5. The two processed above will not be reused.

The code in file coap_udp_socker.erl is like:
handle_info({terminated, SupPid, ChId}, State=#state{chans=Chans}) ->
Chans2 = dict:erase(ChId, Chans),
exit(SupPid, normal),
{noreply, State#state{chans=Chans2}};

is it right to change it like this?
handle_info({terminated, SupPid, ChId}, State=#state{chans=Chans, pool=PoolPid}) ->
Chans2 = dict:erase(ChId, Chans),
exit(SupPid, normal),
coap_channel_sup_sup:delete_channel(PoolPid, ChId),
{noreply, State#state{chans=Chans2}};

[hejin1026]

exit(SupPid, normal)
but SupPid(coap_channel_sup) can not exit because

If Reason is the atom normal, Pid does not exit. If it is trapping exits, the exit signal is transformed into a message {'EXIT', From, normal} and delivered to its message queue.

so why use exit(SupPid, normal), it not work
exit(SupPid, kill) and coap_channel_sup_sup:delete_channel(PoolPid, ChId) can make SupPid shutdown
is that right?