From: David Miller Subject: Re: ip_gre headroom allocation Date: Thu, 11 Jun 2009 03:09:22 -0700 (PDT) Message-ID: <20090611.030922.91415233.davem@davemloft.net> References: <4A30BECA.8030205@iki.fi> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org, herbert@gondor.apana.org.au, kuznet@ms2.inr.ac.ru To: timo.teras@iki.fi Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:40130 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756983AbZFKKJT convert rfc822-to-8bit (ORCPT ); Thu, 11 Jun 2009 06:09:19 -0400 In-Reply-To: <4A30BECA.8030205@iki.fi> Sender: netdev-owner@vger.kernel.org List-ID: =46rom: Timo Ter=E4s Date: Thu, 11 Jun 2009 11:22:34 +0300 > I'm not entirely sure about this, but shouldn't we do something like > the below to make ip_gre allocation enough headroom for paths where > e.g. xfrm transformations take place (as a speed optimization)? This should be handled by the code one level up that allocates the SKB. And as routes get stacked, the MTUs and headroom values get adjusted up to the root. At least that's how it's supposed to work :-) Otherwise we'd need a similar hack in ipip.c too.