From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 10 Aug 2001 16:05:41 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 10 Aug 2001 16:05:21 -0400 Received: from ns.suse.de ([213.95.15.193]:50187 "HELO Cantor.suse.de") by vger.kernel.org with SMTP id ; Fri, 10 Aug 2001 16:05:18 -0400 To: Ian Stirling Cc: linux-kernel@vger.kernel.org Subject: Re: Network device aliases In-Reply-To: <200108100327.EAA22951@mauve.demon.co.uk.suse.lists.linux.kernel> From: Andi Kleen Date: 10 Aug 2001 22:05:28 +0200 In-Reply-To: Ian Stirling's message of "10 Aug 2001 05:46:42 +0200" Message-ID: User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Ian Stirling writes: > I think I've more or less worked out how network devices are initiated, > and configured, with the help of the htmlised sources, but am not > finding anything on how aliases (eth0:1 ...) work. > Do they have an entire device structure that only differs in name and > IP address? In 2.0 they had. In 2.2+ they are just another entry on a per interface address list. The notion of alias interfaces is just emulated for compatibility, they really do not exist anymore. -Andi