From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kim Phillips Subject: Re: [stable] net: Handle different key sizes between address families in flow cache Date: Fri, 4 Nov 2011 21:29:58 -0500 Message-ID: <20111104212958.5d54152b38050e2a0917d575@freescale.com> References: <20111101185022.b156ea957baf286f1bf00f9c@freescale.com> <20111101.203430.110242050755204948.davem@davemloft.net> <20111103135851.2229247682ec040d653479e1@freescale.com> <20111103.224329.2276058062630695656.davem@davemloft.net> <20111104144659.3e0a30afffffd9dd73cf740e@freescale.com> <20111104212439.GA27324@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Cc: David Miller , , , , , David Ward To: Greg KH Return-path: Received: from ch1ehsobe001.messaging.microsoft.com ([216.32.181.181]:30102 "EHLO ch1outboundpool.messaging.microsoft.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752180Ab1KECa6 (ORCPT ); Fri, 4 Nov 2011 22:30:58 -0400 In-Reply-To: <20111104212439.GA27324@kroah.com> Sender: netdev-owner@vger.kernel.org List-ID: On Fri, 4 Nov 2011 14:24:39 -0700 Greg KH wrote: > On Fri, Nov 04, 2011 at 02:46:59PM -0500, Kim Phillips wrote: > > commit aa1c366e4febc7f5c2b84958a2dd7cd70e28f9d0 upstream. > > > > With the conversion of struct flowi to a union of AF-specific structs, some > > operations on the flow cache need to account for the exact size of the key. > > > > Signed-off-by: David Ward > > Signed-off-by: David S. Miller > > Cc: stable@vger.kernel.org # v3.0.x > > --- > > This patch is the result of a clean cherry-pick onto v3.0.8. > > It restores IPSec fwding performance from ~4kpps back to ~44kpps on > > a P2020DS board. > > Too bad you forgot to build this patch after you applied it: > CC init/main.o > In file included from include/linux/security.h:39:0, > from init/main.c:32: > include/net/flow.h: In function 'flow_key_size': > include/net/flow.h:174:3: error: size of unnamed array is negative > include/net/flow.h:177:3: error: size of unnamed array is negative > > Please be kind to your poor over-worked stable kernel maintainer and do > the decent thing and TEST YOUR PATCH before you ask him to accept it. > > bah, I think it's time for the weekend to start a bit earlier than normal... so sorry I hadn't tested 64-bit. I found the problem - upstream commit aa1c366 depends on upstream commit 728871b: commit 728871bc05afc8ff310b17dba3e57a2472792b13 Author: David Ward Date: Mon Sep 5 16:47:23 2011 +0000 net: Align AF-specific flowi structs to long AF-specific flowi structs are now passed to flow_key_compare, which must also be aligned to a long. Signed-off-by: David Ward Signed-off-by: David S. Miller so, one more time, in stable sign-off area format: Cc: # v3.0.x: 728871b: net: Align AF-specific flowi structs to long Cc: # v3.0.x build tested on 32- and 64-bit powerpc, and ARCH=x86 {i386,x86_64}_defconfig. Kim