From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH] vlan: propogate MTU changes Date: Tue, 07 Oct 2008 01:04:33 +0200 Message-ID: <48EA9981.1020109@trash.net> References: <20081006173024.2741cc01@speedy> <48EA369F.3090306@trash.net> <20081006195446.1dc5a372@speedy> <48EA9223.8090700@trash.net> <48EA964A.6060503@hp.com> <48EA98F0.40302@trash.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Cc: Stephen Hemminger , "David S. Miller" , netdev@vger.kernel.org To: Rick Jones Return-path: Received: from stinky.trash.net ([213.144.137.162]:56883 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753805AbYJFXEk (ORCPT ); Mon, 6 Oct 2008 19:04:40 -0400 In-Reply-To: <48EA98F0.40302@trash.net> Sender: netdev-owner@vger.kernel.org List-ID: Patrick McHardy wrote: > Rick Jones wrote: >> If physical interface MTUs are going to be bouncing around and VLANs >> get their MTUs changed then perhaps a VLAN needs both a desired and >> actual MTU setting. The VLAN's interface would then be the minimum of >> the desired and actual MTU. I suppose it isn't too unlike having both >> an administrative (desired) and operational (actual) interface state. > > Thats assuming that the VLAN device is actually restricted by the > ethernet device settings. I don't know if its always not the case, > but I'm pretty sure it usually isn't. Which means there's no real > need for an operational state wrt. MTUs. Actually it would be useful to have this kind of information on the *ethernet* device to specify the upper limit.