Ethereal-users: Re: [Ethereal-users] RE: Window scaling and relative sequence numbers

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

Date Index Thread Index Other Months All Mailing Lists
Date Prev Date Next Thread Prev Thread Next
From: ronnie sahlberg <ronniesahlberg@xxxxxxxxx>
Date: Thu, 26 Aug 2004 20:45:43 +1000
Hi,  I have checked in a patch for the window scaling.

If it sees a SYN+ACK with no such option specified it will disable window 
scaling for the SYN packet as well  in case it was specified.

The patch also adds the string " (scaled)" to the end of the window size line
in the dissect pane in case the number has been scaled.



On Fri, 30 Jul 2004 08:00:38 +0200, Markus Rehbach  wrote:
> Hi all,
> 
> RFC1323 states:
> 
>  This option is an offer, not a promise; both sides must send
>  Window Scale options in their SYN segments to enable window
>  scaling in either direction. If window scaling is enabled,
>  then the TCP that sent this option will right-shift its true
>  receive-window values by 'shift.cnt' bits for transmission in
>  SEG.WND.  The value 'shift.cnt' may be zero (offering to scale,
>  while applying a scale factor of 1 to the receive window)
> 
> That's why I'm sure that the missing WS= in the syn/ack-Packet
> will disable the window scaling completely for this session and
> should not been calculated for the 3rd packet below.
> 
> Cheers
> 
> Markus
> ____________________
> 
> 
> 
> On Thu, 29 Jul 2004 11:42:12 -0400, Ian Schorr wrote:
> 
> Turning on relative sequence numbers also allows Ethereal to detect and understand TCP window scaling. Were scaling
> bits set in the SYN segments for the connection? If so, I'm guessing that "747520" is the CORRECT window size, and
> 5840 is the window size before shifting bits (which is not the true size of the advertised window)...
> 
> Ian
> 
> ____________________
> 
> On Jul 29, 2004, at 5:35 AM, Markus Rehbach wrote:
> 
> Hi all,
> 
> there seems to be a problem in 0.10.5 and probably older ones.
> 
> If preference is 'tcp.relative_sequence_numbers: TRUE' there is a wrong
> value shown for the window size (3rd line below: Win=747520 should be 5840).
> 
> 0.070763 a -> b TCP 32985 > imap [SYN] Seq=0 Ack=0 Win=5840 Len=0 MSS=1460 TSV=3609664 TSER=0 WS=7
> 0.071129 b -> a TCP imap > 32985 [SYN, ACK] Seq=0 Ack=1 Win=5840 Len=0 MSS=1460
> 0.071176 a -> b TCP 32985 > imap [ACK] Seq=1 Ack=1 Win=747520 Len=0
> 
> If preference is 'tcp.relative_sequence_numbers: FALSE' it looks good.
> 
> 0.076386 a -> b TCP 32991 > imap [SYN] Seq=4090487846 Ack=0 Win=5840 Len=0 MSS=1460 TSV=4006974 TSER=0 WS=7
> 0.076748 b -> a TCP imap > 32991 [SYN, ACK] Seq=3661596958 Ack=4090487847 Win=5840 Len=0 MSS=1460
> 0.076798 a -> b TCP 32991 > imap [ACK] Seq=4090487847 Ack=3661596959 Win=5840 Len=0
> 
> Problem shows up in ethereal and tethereal (no surprise;-).
> 
> Markus
> 
> _______________________________________________
> Ethereal-users mailing list
> Ethereal-users@xxxxxxxxxxxx
> http://www.ethereal.com/mailman/listinfo/ethereal-users
>