From owner-freebsd-net  Fri Nov 30 11:50:19 2001
Delivered-To: freebsd-net@freebsd.org
Received: from prism.flugsvamp.com (cb58709-a.mdsn1.wi.home.com [24.17.241.9])
	by hub.freebsd.org (Postfix) with ESMTP id E7BE037B416
	for <net@freebsd.org>; Fri, 30 Nov 2001 11:50:06 -0800 (PST)
Received: (from jlemon@localhost)
	by prism.flugsvamp.com (8.11.0/8.11.0) id fAUJl3Z85729;
	Fri, 30 Nov 2001 13:47:03 -0600 (CST)
	(envelope-from jlemon)
Date: Fri, 30 Nov 2001 13:47:03 -0600 (CST)
From: Jonathan Lemon <jlemon@flugsvamp.com>
Message-Id: <200111301947.fAUJl3Z85729@prism.flugsvamp.com>
To: jayanth@yahoo-inc.com, net@freebsd.org
Subject: Re: FreeBSD performing worse than Linux?
X-Newsgroups: local.mail.freebsd-net
In-Reply-To: <local.mail.freebsd-net/20011130100515.C63426@yahoo-inc.com>
References: <local.mail.freebsd-net/20011128153817.T61580@monorchid.lemis.com>
	<local.mail.freebsd-net/15364.38174.938500.946169@caddis.yogotech.com>
	<local.mail.freebsd-net/20011129004234.A16101@exuma.irbs.com>
	<local.mail.freebsd-net/200111300527.fAU5R0s11199@c527597-a.cstvl1.sfba.home.com>
Organization: 
Cc: 
Sender: owner-freebsd-net@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-net.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-net>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-net>
X-Loop: FreeBSD.org

In article <local.mail.freebsd-net/20011130100515.C63426@yahoo-inc.com> you write:
>Hmm, this is what has broken fast retransmit.
>I should probably replace 
>	if (tp->t_dupacks < tcprexmtthresh)
>                        tp->t_dupacks = 0;
>
>	with
>	if (tcp_do_newreno && (tp->t_dupacks < tcprexmtthresh))
>			tp->t_dupacks = 0;

I don't think that's it.  The original (pre-NewReno) code had:

                /*
                 * If the congestion window was inflated to account
                 * for the other side's cached packets, retract it.
                 */
                if (tp->t_dupacks >= tcprexmtthresh &&
                    tp->snd_cwnd > tp->snd_ssthresh)
                        tp->snd_cwnd = tp->snd_ssthresh;
                tp->t_dupacks = 0;

So t_dupacks was always getting set to 0 on this codepath.  In the
existing code, the 'if' statement you point out above is only relevant
if newreno is turned on.  However, the pathologies happen regardless
of the setting of the newreno flag.
-- 
Jonathan

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message