From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH 3/8] [NETFILTER]: rename NF_ARP to NFPROTO_ARP and assign a non-clashing value Date: Wed, 09 Apr 2008 15:42:08 +0200 Message-ID: <47FCC7B0.6080205@trash.net> References: <5130e28c2130c57a9a07ae21c552fe7db519473c.1207668694.git.jengelh@computergmbh.de> <2a75525ff0d60f2e71df2c9c3d040d3a76451713.1207668694.git.jengelh@computergmbh.de> <47FCBC24.4020209@trash.net> <47FCC0D2.7060306@trash.net> <47FCC2F4.4030901@trash.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Cc: netfilter-devel@vger.kernel.org To: Jan Engelhardt Return-path: Received: from stinky.trash.net ([213.144.137.162]:46894 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752734AbYDINmL (ORCPT ); Wed, 9 Apr 2008 09:42:11 -0400 In-Reply-To: Sender: netfilter-devel-owner@vger.kernel.org List-ID: Jan Engelhardt wrote: > On Wednesday 2008-04-09 15:21, Patrick McHardy wrote: >> The decoupling of netfilter supported protocols from AF values >> makes sense. >> >> However I think _MAX definitions actually having the value MAX + 1 >> is pretty poor style. When you see an array dimensioned as >> [XYZ_MAX] you always have to check whether it is really the maximum >> (and thus a bug) or maximum + 1. So I'd prefer to have MAX really >> be the maximum and use max + 1 for arrays etc. >> > In this case, I'd just follow AF_ suit. > Especially since NAME_MAX=256, PATH_MAX=4096 for example, Thats a bad example. > _MAX is more often the total size rather than the last element. > (The more even since loops use for (; x < MAX; ) > rather than for (; x<= MAX;) as can probably be seen > in a lot of userspace code.) Which is in my opinion a sign of poor coding and leads to off-by-ones. Please don't redefine the meaning of maximum.