LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: paulmck@linux.vnet.ibm.com
Cc: Ed Tomlinson <edt@aei.ca>, Steven Rostedt <rostedt@goodmis.org>,
	Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
	Ingo Molnar <mingo@elte.hu>, Thomas Gleixner <tglx@linutronix.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	Dipankar Sarma <dipankar@in.ibm.com>,
	linux-kernel@vger.kernel.org
Subject: Re: INFO: possible circular locking dependency detected
Date: Fri, 15 Jul 2011 18:11:55 +0200	[thread overview]
Message-ID: <1310746315.2586.370.camel@twins> (raw)
In-Reply-To: <20110715155916.GB2327@linux.vnet.ibm.com>

On Fri, 2011-07-15 at 08:59 -0700, Paul E. McKenney wrote:

> > Because we're in irq_exit(), after decrementing preempt_count, so
> > in_irq() returns false.
> 
> Can we delay decrementing preempt_count so that RCU has some chance
> of actually working?

No, softirqs must be ran with in_irq() being false.

> > No, the *BANG* being that we end up calling rcu_read_unlock_special()
> > while holding scheduler locks, which is BAD(tm).
> 
> Well, it certainly is BAD(tm) if you guys continue to deprive
> rcu_read_unlock_special() of the means of determining whether it is
> being invoked from hardware irq handler context.

hard irq handler isn't really the problem here, its the nested softirq
code that is.

> > >  (Which I believe, perhaps
> > > incorrectly, to be prevented by the fact that all modifications to
> > > ->rcu_read_unlock_special are carried out with irqs disabled on the
> > > corresponding CPU, at least given no RCU_BOOST.)  The check for in_irq()
> > > should prevent the from-irq rcu_read_unlock_special() from attempting
> > > to acquire any locks.
> > 
> > Right, so in_irq() simply checks a few bits in preempt_count, which we
> > just cleared due to being in irq_exit().
> 
> Right.  So how about delaying clearing those bits until after you get
> done messing with the scheduler from hardware irq handler context?

Can't do.

> > But in_irq() isn't sufficient for RCU usage after the hardirq ends, see
> > irq_exit(). Also there's all of softirq to consider, that too can run
> > and not get caught by in_irq().
> 
> Change the rules without adjusting the callers can in fact result in some
> breakage.  ;-)

There's no changing the rules here, this is how its worked for a very
long time indeed. Softirqs can run from the hardirq tail.

> The bit about local_irq_save() and local_irq_restore() invoking the
> scheduler is rather surprising -- is there a raw_ version that avoids
> this?

They don't, they might for -rt, but that's a different story. But
looking at the latest version I have its only local_irq_save_rt() and
friends that do that.

> > > 3.	It is possible that the task is preempted after the
> > > 	--rcu_read_lock_nesting, in which case the task won't be queued.
> > > 	Of course the task might already be queued if there was an
> > > 	earlier preemption during this same RCU read-side critical
> > > 	section, in which case #2 applies.
> > > 
> > > 	In other words, a preemption in __rcu_read_unlock() after the
> > > 	--rcu_read_lock_nesting has no effect on RCU state: either the
> > > 	task was already marked RCU_READ_UNLOCK_BLOCKED, or it wasn't.
> > > 	Either way, rcu_note_context_switch() does not see this task as
> > > 	being in an RCU read-side critical section.
> > > 
> > > So what am I missing here?
> > 
> >  $task			IRQ		SoftIRQ
> > 
> >  rcu_read_lock()
> > 
> >  /* do stuff */
> > 
> >  <preempt> |= UNLOCK_BLOCKED
> > 
> >  rcu_read_unlock()
> >    --t->rcu_read_lock_nesting
> > 
> > 			irq_enter();
> > 			/* do stuff, don't use RCU */
> > 			irq_exit();
> > 			  sub_preempt_count(IRQ_EXIT_OFFSET);
> > 			  invoke_softirq()
> 
> Why can't we exchange the order of the above two so that RCU correctly
> avoids messing with the scheduler if called from hardware interrupt
> context?

Because softirqs != hardirq ? This has been so like forever, can't go
change the semantics of this without risking tons of borkage. Every time
we try to change softirq semantics (we tried with -rt, because softirqs
are a massive pain) everything goes tits up fast.

> > 
> > 					ttwu();
> > 					  spin_lock_irq(&pi->lock)
> > 					  rcu_read_lock();
> > 					  /* do stuff */
> > 					  rcu_read_unlock();
> > 					    rcu_read_unlock_special()
> > 					      rcu_report_exp_rnp()
> > 					        ttwu()
> > 					          spin_lock_irq(&pi->lock) /* deadlock */
> > 					      
> > 
> >    rcu_read_unlock_special(t);
> > 
> > Ed can simply trigger this 'easy' because invoke_softirq() immediately
> > does a ttwu() of ksoftirqd/# instead of doing the in-place softirq stuff
> > first, but even without that the above happens.
> 
> An easily reproduced bug is certainly a nice change of pace...
> 
> > Something like the below _might_ fix it..
> 
> Maybe, but how does tglx make PREEMPT_RT work in this case?  The problem
> is that PREEMPT_RT allows ksoftirqd to be preempted, and thus allows it
> to be RCU priority boosted.

RT is mostly easier since it doesn't nest as many contexts, softirqs for
example always run in task context, and the only way to run them in a
random tasks' context is through local_bh_enable() and since there's no
local_bh_enable() call in the middle of __rcu_read_unlock() you're
pretty good there.

I know tglx has some softirq changes he hasn't yet shared with me, but
if the patch I send earlier fixes the problem for mainline, I'm fairly
confident I can cook one up for him as well.



  reply	other threads:[~2011-07-15 16:12 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-14 14:49 Sergey Senozhatsky
2011-07-14 16:41 ` Peter Zijlstra
2011-07-14 16:57   ` Paul E. McKenney
2011-07-14 19:16     ` Sergey Senozhatsky
2011-07-14 19:15   ` Sergey Senozhatsky
2011-07-14 19:34     ` Paul E. McKenney
2011-07-14 19:38       ` Dave Jones
2011-07-14 20:33         ` Paul E. McKenney
2011-07-14 19:38       ` Sergey Senozhatsky
2011-07-14 16:58 ` Steven Rostedt
2011-07-14 17:02   ` Steven Rostedt
2011-07-14 17:05     ` Paul E. McKenney
2011-07-14 17:32       ` Steven Rostedt
2011-07-14 17:46         ` Steven Rostedt
2011-07-14 19:18           ` Paul E. McKenney
2011-07-14 19:41             ` Steven Rostedt
2011-07-14 20:33               ` Paul E. McKenney
2011-07-15 11:05             ` Ed Tomlinson
2011-07-15 11:29               ` Peter Zijlstra
2011-07-15 11:35                 ` Ed Tomlinson
2011-07-15 11:39                 ` Peter Zijlstra
2011-07-15 18:11                   ` Paul E. McKenney
2011-07-15 12:42                 ` Paul E. McKenney
2011-07-15 13:07                   ` Peter Zijlstra
2011-07-15 14:36                     ` Paul E. McKenney
2011-07-15 15:04                       ` Peter Zijlstra
2011-07-15 15:59                         ` Paul E. McKenney
2011-07-15 16:11                           ` Peter Zijlstra [this message]
2011-07-15 16:56                             ` Paul E. McKenney
2011-07-15 21:48                               ` Ed Tomlinson
2011-07-15 22:04                                 ` Paul E. McKenney
2011-07-16 19:42                                   ` Ed Tomlinson
2011-07-17  0:02                                     ` Paul E. McKenney
2011-07-17  1:56                                       ` Ed Tomlinson
2011-07-17 14:28                                         ` Paul E. McKenney
2011-07-18 15:15                                           ` Paul E. McKenney
2011-07-18  9:29                                     ` Peter Zijlstra
2011-07-18 15:29                                       ` Paul E. McKenney
2011-07-15 16:55                     ` Steven Rostedt
2011-07-15 17:03                       ` Paul E. McKenney
2011-07-15 17:16                         ` Steven Rostedt
2011-07-15 17:24                           ` Paul E. McKenney
2011-07-15 17:42                             ` Steven Rostedt
2011-07-15 18:33                               ` Paul E. McKenney
  -- strict thread matches above, loose matches on Subject: below --
2011-08-07 16:22 Justin P. Mattock
2011-08-11 20:57 ` Justin P. Mattock
2009-12-06 10:11 Richard Zidlicky
2009-10-10 23:09 John Kacur
2007-02-08 15:03 Pedro M. López
2006-10-16 14:05 alpha @ steudten Engineering
2006-10-16 14:32 ` Nick Piggin
2006-10-16 15:42   ` Randy Dunlap
2006-10-16 15:46     ` Nick Piggin
2006-10-19  6:02   ` Andrew Morton
2006-10-19  6:30     ` Nick Piggin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1310746315.2586.370.camel@twins \
    --to=a.p.zijlstra@chello.nl \
    --cc=akpm@linux-foundation.org \
    --cc=dipankar@in.ibm.com \
    --cc=edt@aei.ca \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=rostedt@goodmis.org \
    --cc=sergey.senozhatsky@gmail.com \
    --cc=tglx@linutronix.de \
    --subject='Re: INFO: possible circular locking dependency detected' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).