Hello!
> problem. But in 2.3.x everything is fine.... I am at a loss.
Funnier and funnier...
TCP really behaves differently in 2.3. But the difference looks not
enough essential to achieve such turbo-effect. 8)
Probably, it is problem with your provider. Maybe, it shapes
your ppp link "smartly"? Look, 2.3 does not delay ACKs
at connection startup, 2.2 does and loss starts exactly,
when your machine delay ACK first time (third packet).
I cannot imagine, why it has such strange effect.
Seems, this www.slashdot.org is linux-2.2 too (at least, it has FACK)
and it should not misbehave.
All this is really strange.
Listen, please, go to tcp_input.c, find line
/* Two full frames received or... */
if (((tp->rcv_nxt - tp->rcv_wup) >= tp->rcv_mss * MAX_DELAY_ACK) ||
replace it with
if (((tp->rcv_nxt - tp->rcv_wup) >= tp->rcv_mss) ||
and look, what will occur. If it will start to work well,
problem is _surely_ outside your machine. Either at provider
or at slashdot.
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/
This archive was generated by hypermail 2b29 : Sat Jul 15 2000 - 21:00:09 EST