From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760177Ab2D0OHR (ORCPT ); Fri, 27 Apr 2012 10:07:17 -0400 Received: from hrndva-omtalb.mail.rr.com ([71.74.56.122]:28542 "EHLO hrndva-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758636Ab2D0OHO (ORCPT ); Fri, 27 Apr 2012 10:07:14 -0400 X-Authority-Analysis: v=2.0 cv=IaEFqBWa c=1 sm=0 a=ZycB6UtQUfgMyuk2+PxD7w==:17 a=XQbtiDEiEegA:10 a=Co6SMZqKBZgA:10 a=5SG0PmZfjMsA:10 a=Q9fys5e9bTEA:10 a=pGHU-0qZd6m6RciWSMoA:9 a=PUjeQqilurYA:10 a=ZycB6UtQUfgMyuk2+PxD7w==:117 X-Cloudmark-Score: 0 X-Originating-IP: 74.67.80.29 Message-ID: <1335535630.28106.209.camel@gandalf.stny.rr.com> Subject: Re: [PATCH v4 1/3] make jump_labels wait while updates are in place From: Steven Rostedt To: Jason Baron Cc: Glauber Costa , cgroups@vger.kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Li Zefan , Tejun Heo , kamezawa.hiroyu@jp.fujitsu.com, linux-mm@kvack.org, devel@openvz.org, Johannes Weiner , Michal Hocko , Ingo Molnar Date: Fri, 27 Apr 2012 10:07:10 -0400 In-Reply-To: <20120427135320.GA13762@redhat.com> References: <1335480667-8301-1-git-send-email-glommer@parallels.com> <1335480667-8301-2-git-send-email-glommer@parallels.com> <20120427004305.GC23877@home.goodmis.org> <20120427135320.GA13762@redhat.com> Content-Type: text/plain; charset="ISO-8859-15" X-Mailer: Evolution 3.2.2-1 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2012-04-27 at 09:53 -0400, Jason Baron wrote: > Right, for x86 which uses stop_machine currently, we guarantee that all > cpus are going to see the updated code, before the inc of key->enabled. > However, other arches (sparc, mips, powerpc, for example), seem to be > using much lighter weight updates, which I hope are ok :) And x86 will soon be removing stop_machine() from its path too. But all archs should perform some kind of memory sync after patching code. Thus the update should be treated as if a memory barrier was added after it, and before the inc. -- Steve