> I connect to the server (that is
> sleeping) and _then_ I always close the socket, so the file descriptor
> gets released, but the newsk in CONNECTING state is still allocated
> awating for a server action. The DoS is that there's no limit in the
> number of these peer-newsk. Do you see my point?
Mama mia! I see. It was big mistake.
I see... but what to do then? We did not remove syn skb and not yet
accepted endpoint on close of connecting socket only to avoid race
resulting in block in accept().
Alexey
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/