From mboxrd@z Thu Jan  1 00:00:00 1970
From: Pablo Neira Ayuso <pablo@netfilter.org>
Subject: Re: [PATCH] iptables: change large file support handling
Date: Wed, 4 Oct 2017 16:46:06 +0200
Message-ID: <20171004144606.GA22714@salvia>
References: <201709281059.45061.jbe@pengutronix.de>
 <20171004141458.GA28637@salvia>
 <201710041639.23481.jbe@pengutronix.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: netfilter-devel@vger.kernel.org
To: Juergen Borleis <jbe@pengutronix.de>
Return-path: <netfilter-devel-owner@vger.kernel.org>
Received: from ganesha.gnumonks.org ([213.95.27.120]:50696 "EHLO
        ganesha.gnumonks.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1752442AbdJDOqL (ORCPT
        <rfc822;netfilter-devel@vger.kernel.org>);
        Wed, 4 Oct 2017 10:46:11 -0400
Content-Disposition: inline
In-Reply-To: <201710041639.23481.jbe@pengutronix.de>
Sender: netfilter-devel-owner@vger.kernel.org
List-ID: <netfilter-devel.vger.kernel.org>

On Wed, Oct 04, 2017 at 04:39:23PM +0200, Juergen Borleis wrote:
> Hi Pablo,
> 
> On Wednesday 04 October 2017 16:14:58 Pablo Neira Ayuso wrote:
> > > The configure script handles the "--*-largefile" parameter badly. It
> > > defaults to large file support if not given and always disables
> > > largefile support if given (and it doesn't matter if 'enable' or
> > > 'disable' was used)
> >
> > It seems the default behaviour for a while has been to enable
> > largefile by default.
> >
> > Then, someone came in to add this --disable-largefile configure time
> > option, see:
> >
> > commit 967cb7106f0f61cd7b8fbb10bc2451a3f7372a43
> > Author: Karl Hiramoto <karl@hiramoto.org>
> > Date:   Mon May 10 17:50:41 2010 +0200
> >
> >     iptables: optionally disable largefile support
> >
> > So I would just fix --enable-* --disable-*.
> >
> > Would you follow up on this? Let me know, thanks.
> 
> With my change the default is still "enabled" and you can disable it on 
> demand.

I'm going to remove this from your log message: "It defaults to large
file support if not given", when I read this I thought this was the
problem.

> To have still support for "disable on demand" makes sense for some 
> small 32 bit embedded systems.

Sure, this knob is useful, not questioning this, and it needs to be
fixed.

Thanks.