From owner-freebsd-bugs  Wed Oct  2 02:12:52 1996
Return-Path: owner-bugs
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id CAA23196
          for bugs-outgoing; Wed, 2 Oct 1996 02:12:52 -0700 (PDT)
Received: from root.com (implode.root.com [198.145.90.17])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id CAA23185
          for <bugs@FreeBSD.org>; Wed, 2 Oct 1996 02:12:46 -0700 (PDT)
Received: from localhost (localhost [127.0.0.1]) by root.com (8.7.5/8.6.5) with SMTP id CAA06758; Wed, 2 Oct 1996 02:13:12 -0700 (PDT)
Message-Id: <199610020913.CAA06758@root.com>
X-Authentication-Warning: implode.root.com: Host localhost [127.0.0.1] didn't use HELO protocol
To: Rowan Crowe <rowan@ozramp.net.au>
cc: bugs@FreeBSD.org
Subject: Re: Possible low level TCP bug 
In-reply-to: Your message of "Wed, 02 Oct 1996 19:00:08 +1000."
             <Pine.LNX.3.91.961002185805.32199A-100000@ozramp> 
From: David Greenman <dg@root.com>
Reply-To: dg@root.com
Date: Wed, 02 Oct 1996 02:13:12 -0700
Sender: owner-bugs@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

>Here's another capture for you, this time coming from a remote IRC 
>server, being routed to a SLIP link. My FreeBSD box is not in the path, 
>but the remote end (dropbear.aussie.net) runs FreeBSD.
>
>
>Wed Oct 02 18:54:43 1996 - sl0 sent:
>IP: len 140 203.20.52.13->203.30.11.132 ihl 20 ttl 63 tos 8 prot TCP
>TCP: 6668->1026 Seq x7c07420 Ack xbe4b006e ACK PSH Wnd 8576 Data 100
>0000  :aussie.sydney.oz.org 332 rowan #newbies :learn irc language fro
>0040  m http://newbies.aussie.net/..:aussi
>
>Wed Oct 02 18:54:44 1996 - sl0 recv:
>IP: len 40 203.30.11.132->203.20.52.13 ihl 20 ttl 60 prot TCP
>TCP: 1026->6668 Seq xbe4b006e Ack x7c07484 ACK Wnd 3996
>
>Wed Oct 02 18:54:44 1996 - sl0 sent:
>IP: len 88 203.20.52.13->203.30.11.132 ihl 20 ttl 63 tos 8 prot TCP
>TCP: 6668->1026 Seq x7c07484 Ack xbe4b006e ACK PSH Wnd 8576 Data 48
>0000  e.sydney.oz.org 333 rowan #newbies Z 844233027..
>
>
>Something important to note, is that the split occurred at the same 
>number of characters (about 100).

   This is likely caused by IRC doing short writes to the socket (the fact
that PUSH is set is a good indicator of this). I don't know why it would be
doing that, but it's not a bug in FreeBSD. In my previous response, I was
only concerned with the long delays you were seeing - not in the multiple
small packets.

-DG

David Greenman
Core-team/Principal Architect, The FreeBSD Project