> > BTW do you know that rfc1122 prohibits advertising MSS < 536. 8)
>
> As I read it it says you must be prepared to accept 536 byte MSS whatever
> you argued. It doesn't say you cant ask for a lower MSS.
I also thought this 8). But rereading it week ago, I found that
it requires to advertise as MSS maximal value, which it can accept.
And minimal allowed maximum is 576.
Yes. We are buggy here and draft-ietf-tcpimpl-pmtud-03
explicitly announces us buggy 8)
Apparently, they forgot about ax.25.
> Not on radio links. Trust me. Linux acquired the per route mss originally
> to handle amateur radio links. Some of these folks got 40% throughput
> increases when TCP SACK went in, thats how grotty the links are.
I believe. But do not understand. Seems, AX.25 is reliable protocol, is not it?
Then fragments are delivered reliably and you can only win of the fact,
that frames traversing internet are of 576 bytes.
And in backward direction nobody prohibited you to send as small frames
as possible. No, I do not understand.
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/