4.X implementations of TCP, initial sequence numbers, and windows

[Richard Wolski]

Hello everyone.

I have a BSD implementation question regarding the initial sequence
number and the advertised window.  In some of the BSD code that we have
(vendors will remain nameless to protect the innocent) the following
piece of code appears:

if(win > 0 && SEQ_GT(tp->rcv_nxt+win, tp->rcv_adv))
	tp->rcv_adv = tp->rcv_nxt + win;

What we notice is that another vendor's implementation of TCP chooses an
initial sending sequence number with the sign bit set (sometimes) and that
the rcv_adv field in the tcpcb always remains 0.  I think the following thing
is happening.  SEQ_GT is defined as:

#define SEQ_GT(a,b)  ((int)((a)-(b)) > 0)

When expanded in the above code, (a) has the sign bit set, (b) is zero, so the
test fails and rcv_adv never gets set properly.  This manifests itself
as unusually large windows where one would not expect them.

My first question:  Am I reading this right?  I checked the code to set the
iss on the sending side, and it periodically increments a global variable
which eventually results in a negative number (when viewed as an int).

But wait, there's more...

We looked further at the sender's code for setting iss and saw the following
statements:

#ifdef TCP_COMPAT_42
	if((int)tcp_iss < 0)
		tcp_iss = 0;			/* XXX */
#endif

This makes me believe that the problem is somehow fixed at 4.3.  Is that
true?  We looked at yet another vendor's implementation which was supposed
to be 4.3, and nothing seems to be different. 

Any thoughts that you might have in this matter will be gratefully appreciated,
and I apologize if I am asking for the answer to a question that everybody
except me is privy to.

Rich Wolski
rwolski at lll-lcc.llnl.gov		inter-net
(415)423-8594				bell-net
P.O. Box 808, L-60			mail-net
Livermore, CA  94538