From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH 3/4] xfrm: remove policy lock when accessing policy->walk.dead Date: Wed, 31 Mar 2010 13:57:35 -0700 (PDT) Message-ID: <20100331.135735.00480230.davem@davemloft.net> References: <20100331135505.GC14082@gondor.apana.org.au> <1270044768.26743.58.camel@bigi> <20100331141525.GA14331@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: hadi@cyberus.ca, timo.teras@iki.fi, netdev@vger.kernel.org To: herbert@gondor.apana.org.au Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:59177 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932141Ab0CaU5e (ORCPT ); Wed, 31 Mar 2010 16:57:34 -0400 In-Reply-To: <20100331141525.GA14331@gondor.apana.org.au> Sender: netdev-owner@vger.kernel.org List-ID: From: Herbert Xu Date: Wed, 31 Mar 2010 22:15:25 +0800 > Dave can keep or revert this as he likes. Herbert, if it doesn't break anything, I feel the gains are very worthwhile. When we have a breakage report, I will undo this or fix it immediately. Until then, we have to be cognizant of what positives we get out of this. Have you actually tried to monitor the IPSEC netlink socket events when a daemon is running? It's way too painful before Jamal's changes, and we already emit way too many semantically empty events in netlink. Minimization of the noise is a good thing, if it can legally be done, which I believe is the case here. And this applies to pfkey events too, because processing those is even more expensive on the application side.