From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755839Ab1GOSMG (ORCPT ); Fri, 15 Jul 2011 14:12:06 -0400 Received: from e3.ny.us.ibm.com ([32.97.182.143]:37819 "EHLO e3.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754666Ab1GOSME (ORCPT ); Fri, 15 Jul 2011 14:12:04 -0400 Date: Fri, 15 Jul 2011 11:11:47 -0700 From: "Paul E. McKenney" To: Peter Zijlstra Cc: Ed Tomlinson , Steven Rostedt , Sergey Senozhatsky , Ingo Molnar , Thomas Gleixner , Andrew Morton , Dipankar Sarma , linux-kernel@vger.kernel.org Subject: Re: INFO: possible circular locking dependency detected Message-ID: <20110715181147.GF2327@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <20110714144946.GA3354@swordfish.minsk.epam.com> <1310665613.27864.50.camel@gandalf.stny.rr.com> <20110714191809.GF2349@linux.vnet.ibm.com> <201107150705.46248.edt@aei.ca> <1310729362.2586.325.camel@twins> <1310729960.2586.327.camel@twins> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1310729960.2586.327.camel@twins> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jul 15, 2011 at 01:39:20PM +0200, Peter Zijlstra wrote: > On Fri, 2011-07-15 at 13:29 +0200, Peter Zijlstra wrote: > > > > Paul, what appears to be happening here is that some rcu_read_unlock() > > gets interrupted, possibly before calling rcu_read_unlock_special(), > > possibly not if the interrupt is itself the timer interrupt. > > > > Supposing ->rcu_read_unlock_special is set before, any wakeup happening > > from an interrupt hitting __rcu_read_unlock(): > > Hmm, ok not any wakeup from interrupt context because you have that > in_irq() test in there, but if that IRQ doesn't happen to use RCU and > does trigger softirqs and one of that softirq does a wakeup we're still > in the same boat. Agreed. All fallout from adding rcu_read_unlock() while holding rq/pi locks without the needed adjustments. :-( Thanx, Paul > > void __rcu_read_unlock(void) > > { > > struct task_struct *t = current; > > > > barrier(); /* needed if we ever invoke rcu_read_unlock in rcutree.c */ > > --t->rcu_read_lock_nesting; > > barrier(); /* decrement before load of ->rcu_read_unlock_special */ > > if (t->rcu_read_lock_nesting == 0 && > > unlikely(ACCESS_ONCE(t->rcu_read_unlock_special))) > > rcu_read_unlock_special(t); > > #ifdef CONFIG_PROVE_LOCKING > > WARN_ON_ONCE(ACCESS_ONCE(t->rcu_read_lock_nesting) < 0); > > #endif /* #ifdef CONFIG_PROVE_LOCKING */ > > } > > > > After --t->rcu_read_lock_nesting, but before calling > > rcu_read_unlock_special(), will trigger this lock inversion. > > > > The alternative case, ->rcu_read_unlock_special is not set yet, it can > > be set if the interrupt hitting in that same spot above, is the timer > > interrupt, and the wakeup happens either from the softirq ran from the > > hard IRQ tail, or as I suspect here happens, the wakeup of ksoftirqd/#. > > > >