The subject line on this report is something that is normally, and
correctly true. Whether you receive a multicast packet or not
depends on what a socket is bound to and whether anyone (on
the machine) has joined the group on that interface. So, it is correct,
and the same on Solaris and Win32, AFAIK, that a UDP socket
may receive multicast packets for a group it has not joined. This is
definitely the intended and required behavior by RFC and also what
the original Deering implementation does.
If the report is saying the sockets are bound to a multicast address
and receiving non-matching multicast packets, that'd be a bug. But
it mentions joining groups, not what the sockets are bound to.
This also could be a JVM bug, or difference. A C program
exercising the problem, or an strace of the running java might
be helpful.
+-DLS
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
