From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH] ipv4: fix ipsec forward performance regression Date: Thu, 03 Nov 2011 22:43:29 -0400 (EDT) Message-ID: <20111103.224329.2276058062630695656.davem@davemloft.net> References: <20111101185022.b156ea957baf286f1bf00f9c@freescale.com> <20111101.203430.110242050755204948.davem@davemloft.net> <20111103135851.2229247682ec040d653479e1@freescale.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: stable@vger.kernel.org, eric.dumazet@gmail.com, zheng.z.yan@intel.com, netdev@vger.kernel.org To: kim.phillips@freescale.com Return-path: Received: from shards.monkeyblade.net ([198.137.202.13]:49536 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753236Ab1KDCnh (ORCPT ); Thu, 3 Nov 2011 22:43:37 -0400 In-Reply-To: <20111103135851.2229247682ec040d653479e1@freescale.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Kim Phillips Date: Thu, 3 Nov 2011 13:58:51 -0500 > I see b732339 (ipv4: fix ipsec forward performance regression) has > been added to the 3.0- and 3.1-stable trees. > > b732339 increases IPSec fwding performance from 0.2kpps to ~3.5kpps. > > However, adding upstream commit aa1c366 (net: Handle different key > sizes between address families in flow cache) brings performance > back up to 2.6.38 levels, i.e., ~44kpps. > > So can aa1c366 also be queued to the v2.6.39 and v3.0 stable > trees? 1) stable@kernel.org no longer exists and will just go into the bit bucket, it's now stable@vger.kernel.org 2) The patch in question doesn't apply cleanly to v3.0.8 -stable so if it's important to you you'll need to submit a backport. 3) 2.6.39 is no longer a maintained -stable kernel series.