From owner-freebsd-net@FreeBSD.ORG  Fri Feb 18 00:37:58 2005
Return-Path: <owner-freebsd-net@FreeBSD.ORG>
Delivered-To: freebsd-net@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP
	id 523AE16A4CE; Fri, 18 Feb 2005 00:37:58 +0000 (GMT)
Received: from mail-yellow.research.att.com (mail-dark.research.att.com
	[192.20.225.112])	by mx1.FreeBSD.org (Postfix) with ESMTP
	id 469FB43D39; Fri, 18 Feb 2005 00:37:57 +0000 (GMT)
	(envelope-from fenner@research.att.com)
Received: from bright.research.att.com (bright.research.att.com
	[135.207.20.189])
	by mail-green.research.att.com (Postfix) with ESMTP id 42D8AA7BCD;
	Thu, 17 Feb 2005 19:37:56 -0500 (EST)
Received: (from fenner@localhost)
	by bright.research.att.com (8.12.11/8.12.10/Submit) id j1I0bu9F001044;
	Thu, 17 Feb 2005 16:37:56 -0800
Message-Id: <200502180037.j1I0bu9F001044@bright.research.att.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
To: rwatson@freebsd.org
Date: Thu, 17 Feb 2005 16:37:56 -0800
From: Bill Fenner <fenner@research.att.com>
Versions: dmail (linux) 2.6d/makemail 2.10
cc: net@freebsd.org
Subject: Re: solisten() question: why do we check for completed connections?
X-BeenThere: freebsd-net@freebsd.org
X-Mailman-Version: 2.1.1
Precedence: list
List-Id: Networking and TCP/IP with FreeBSD <freebsd-net.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-net>,
	<mailto:freebsd-net-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-net>
List-Post: <mailto:freebsd-net@freebsd.org>
List-Help: <mailto:freebsd-net-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-net>,
	<mailto:freebsd-net-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Fri, 18 Feb 2005 00:37:58 -0000


Robert,

  The code in rev 7.1 (4.3BSD) was:

        if (so->so_q == 0) {
                so->so_q = so;
                so->so_q0 = so;
                so->so_options |= SO_ACCEPTCONN;
        }

which looks like it's initializing q0 and q to be circular queues,
which has to only happen once, and also setting SO_ACCEPTCONN; since
nothing ever turns off SO_ACCEPTCONN, this is not something that
has to happen only once.  When the stuff that had to happen only
once got removed, the conditional remained.

So, I'd say that setting SO_ACCEPTCONN is required, but the conditional
is not.  I don't know if that helps your race or not.

  Bill