From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263584AbTDXMMF (ORCPT ); Thu, 24 Apr 2003 08:12:05 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263599AbTDXMME (ORCPT ); Thu, 24 Apr 2003 08:12:04 -0400 Received: from 203-236-217-11.rev.nextel.co.kr ([203.236.217.11]:518 "EHLO taonetworks.com") by vger.kernel.org with ESMTP id S263548AbTDXMMC (ORCPT ); Thu, 24 Apr 2003 08:12:02 -0400 Date: Thu, 24 Apr 2003 21:24:09 +0900 From: soyoung@taonetworks.com To: linux-kernel@vger.kernel.org Cc: linux-net@vger.kernel.org Subject: [Q]About using Netlink socket and endian transition.. Message-ID: <20030424122408.GB60541@taonetworks.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org (This mail is CC'ed to linux-net mailing list) Hi, I have a question about netlink socket. It can be a 'stupid' question, but I need help about it. The question is : "Does it matter to change endian in netlink socket communication?" Usually there is no endian-transition code in netlink socket program, but netlink socket is also 'socket', so I think when I want to use netlink communication, I have to use endian transition, but no one did it! :) I'm testing netlink socket for Traffic Control (TC) in MPC8250, which is a Motorola PowerPC based CPU. TC works good in normal Pentium-based PC, but it doesn't look working very well on MPC8250 - Traffics cannot recognize their own filters when running through the linux router. So I wonder whether this problem is related to netlink endian transition. Is there anybody who has tested netlink socket in PPC-based platform? Thanks.