From owner-freebsd-hackers  Thu Jan  2 11:36:46 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id LAA26000
          for hackers-outgoing; Thu, 2 Jan 1997 11:36:46 -0800 (PST)
Received: from whistle.com (s205m131.whistle.com [207.76.205.131])
          by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id LAA25995
          for <hackers@freebsd.org>; Thu, 2 Jan 1997 11:36:44 -0800 (PST)
Received: (from smap@localhost) by whistle.com (8.7.5/8.6.12) id LAA15424; Thu, 2 Jan 1997 11:36:09 -0800 (PST)
Received: from bubba.whistle.com(207.76.205.7) by whistle.com via smap (V1.3)
	id sma015422; Thu Jan  2 11:36:06 1997
Received: (from archie@localhost) by bubba.whistle.com (8.7.5/8.6.12) id LAA26380; Thu, 2 Jan 1997 11:36:06 -0800 (PST)
From: Archie Cobbs <archie@whistle.com>
Message-Id: <199701021936.LAA26380@bubba.whistle.com>
Subject: Re: divert code not thread/smp compatible
In-Reply-To: <32C858F6.41C67EA6@whistle.com> from Julian Elischer at "Dec 30, 96 04:06:14 pm"
To: julian@whistle.com (Julian Elischer)
Date: Thu, 2 Jan 1997 11:36:06 -0800 (PST)
Cc: proff@iq.org, hackers@freebsd.org
X-Mailer: ELM [version 2.4ME+ PL25 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


> > /*
> >  * ip_input() and ip_output() set this secret value before calling us to
> >  * let us know which divert port to divert a packet to; this is done so
> >  * we can use the existing prototype for struct protosw's pr_input().
> >  * This is stored in host order.
> >  */
> > u_short ip_divert_port;
> > 
> > /*
> >  * We set this value to a non-zero port number when we want the call to
> >  * ip_fw_chk() in ip_input() or ip_output() to ignore ``divert <port>''
> >  * chain entries. This is stored in host order.
> >  */
> > u_short ip_divert_ignore;
> > 
> > Is this an acceptable trick in the FreeBSD kernel, passing parameters
> > with global variables?
> > 
> > -Julian <proff@iq.org>

My only excuse is that at the time I started writing the divert code,
it wasn't planned on being checked in to the main branch, but instead
was going to be a patch to be applied locally... so I tried to keep
the patch as small as possible... and as a result committed some
non-esthetically pleasing programming in the process... :-)

Any suggestions on the best way to properly threadify this?

Thanks,
-Archie

___________________________________________________________________________
Archie Cobbs   *   Whistle Communications, Inc.  *   http://www.whistle.com