From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCHv2 net-next-2.6 0/5] XFRM,IPv6: Removal of RH2/HAO from IPsec-protected MIPv6 traffic Date: Tue, 28 Sep 2010 14:40:05 -0700 (PDT) Message-ID: <20100928.144005.260072379.davem@davemloft.net> References: <20100928.133849.112575548.davem@davemloft.net> <87r5gdtuxv.fsf@small.ssi.corp> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: eric.dumazet@gmail.com, herbert@gondor.apana.org.au, yoshfuji@linux-ipv6.org, netdev@vger.kernel.org To: arno@natisbad.org Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:58473 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753639Ab0I1Vjp (ORCPT ); Tue, 28 Sep 2010 17:39:45 -0400 In-Reply-To: <87r5gdtuxv.fsf@small.ssi.corp> Sender: netdev-owner@vger.kernel.org List-ID: From: arno@natisbad.org (Arnaud Ebalard) Date: Tue, 28 Sep 2010 23:33:16 +0200 > Before following the (dumb) #ifdef path, I was about to do that but > worried about the penalty of the additional xfrm_state_get/put_afinfo() > calls on each packet I was about to add. Should I just reduce my amount > of coffee or is it a valid concern? Indeed, it is. Even without the concern of afinfo refcounting, this test is very heavy handed for the packet path. Can you make it small enough that it can reasonably be inlined?