From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Morton Subject: Re: [PATCH] fix boot-time hangs from PREEMPT_RCU and NO_HZ Date: Thu, 28 Feb 2008 21:01:40 -0800 Message-ID: <20080228210140.37abe9cd.akpm@linux-foundation.org> References: <20080229044329.GA20207@linux.vnet.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Steven Rostedt , LKML , Linus Torvalds , Ingo Molnar , Thomas Gleixner , john stultz , jarkao2@gmail.com, netdev@vger.kernel.org To: paulmck@linux.vnet.ibm.com Return-path: Received: from smtp1.linux-foundation.org ([207.189.120.13]:56849 "EHLO smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755982AbYB2FCV (ORCPT ); Fri, 29 Feb 2008 00:02:21 -0500 In-Reply-To: <20080229044329.GA20207@linux.vnet.ibm.com> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 28 Feb 2008 20:43:29 -0800 "Paul E. McKenney" wrote: > This is a repost of http://lkml.org/lkml/2008/2/2/11. It is needed in > 2.6.25 in order to prevent boot-time hangs when running with PREEMPT_RCU > and NO_HZ. > > Preemptible RCU can get stuck if a CPU goes idle and NO_HZ is set. The > idle CPU will not progress the RCU through its grace period, so that > any subsequent synchronize_rcu() invocations can hang until such time > as a process starts running on the idle CPU. If no such process runs, > subsequent synchronize_rcu() invocations will hang indefinitely, in turn > hanging the system. Without this patch, one of Steve Rostedt's boxes > hangs on boot when PREEMPT_RCU and NO_HZ are set. That same box boots > fine with this patch. > > Note: This patch came directly from the -rt patch where it has been tested > for several months. > > Andrew, could you please apply this? I already have it, via git-sched. Presumably Ingo has it lined up for the next sched-related merge.