I see that in the erl_interface there ERL_LINK and ERL_EXIT are defined as control message types.
I know that I cant monitor a "PID" on a C node, since this is not supported and the C node will get ERL_ERROR when you try to monitor a "PID" originating from it.
But linking to such "PID" appears to do something, but it seems that this is not documented anywhere.
I tried to call erlang:link/1 on a C node's pid, and my C node gets an ERL_LINK message with correct from to fields. Few moments later it also gets an ERL_EXIT message since the Erlang process which tried to link to C node's pid obviously crashed since my C node just logged the ERL_LINK message and did not reply.
Is this supported at all? And what do I need to do on the C node to handle ERL_LINK properly? Is this worth doing at all or is it too much hassle?
If I have to look at the source of Erlang to understand the protocol I am also fine with that, just please point me in the right direction.