linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Frederic Weisbecker <fweisbec@gmail.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>,
	linux-kernel@vger.kernel.org,
	Dipankar Sarma <dipankar@in.ibm.com>,
	Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@elte.hu>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Lai Jiangshan <laijs@cn.fujitsu.com>
Subject: Re: linux-next-20110923: warning kernel/rcutree.c:1833
Date: Sun, 2 Oct 2011 17:28:32 -0700	[thread overview]
Message-ID: <20111003002832.GC2539@linux.vnet.ibm.com> (raw)
In-Reply-To: <20111002225019.GA1835@somewhere>

On Mon, Oct 03, 2011 at 12:50:22AM +0200, Frederic Weisbecker wrote:
> On Fri, Sep 30, 2011 at 12:24:38PM -0700, Paul E. McKenney wrote:
> > @@ -328,11 +326,11 @@ static int rcu_implicit_offline_qs(struct rcu_data *rdp)
> >  		return 1;
> >  	}
> >  
> > -	/* If preemptible RCU, no point in sending reschedule IPI. */
> > -	if (rdp->preemptible)
> > -		return 0;
> > -
> > -	/* The CPU is online, so send it a reschedule IPI. */
> > +	/*
> > +	 * The CPU is online, so send it a reschedule IPI.  This forces
> > +	 * it through the scheduler, and (inefficiently) also handles cases
> > +	 * where idle loops fail to inform RCU about the CPU being idle.
> > +	 */
> 
> If the idle loop forgets to call rcu_idle_enter() before going to
> sleep, I don't know if it's a good idea to try to cure that situation
> by forcing a quiescent state remotely. It may make the thing worse
> because we actually won't notice the lack of call to rcu_idle_enter()
> that the rcu stall detector would otherwise report to us.
> 
> Also I don't think that works. If the task doesn't have
> TIF_RESCHED, it won't go through the scheduler on irq exit.
> smp_send_reschedule() doesn't set the flag. And also scheduler_ipi()
> returns right away if no wake up is pending.
> 
> So, other than resuming the idle loop to sleep again, nothing may happen.
> 
> Or am I missing something?

Hmmm...  Seems like the IPIs aren't helping in any case, then?

I suppose that I could do an smp_call_function_single(), which then
did a set_need_resched()...

But this is a separate issue that I need to deal with.  That said, any
suggestions are welcome!

							Thanx, Paul

> >  	if (rdp->cpu != smp_processor_id())
> >  		smp_send_reschedule(rdp->cpu);
> >  	else


  reply	other threads:[~2011-10-03  0:28 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-09-25  0:24 linux-next-20110923: warning kernel/rcutree.c:1833 Kirill A. Shutemov
2011-09-25  5:08 ` Paul E. McKenney
2011-09-25 11:26   ` Kirill A. Shutemov
2011-09-25 13:06     ` Frederic Weisbecker
2011-09-25 14:19       ` Kirill A. Shutemov
2011-09-25 16:48       ` Paul E. McKenney
2011-09-26  1:04         ` Frederic Weisbecker
2011-09-26  1:10           ` Frederic Weisbecker
2011-09-26  1:26             ` Paul E. McKenney
2011-09-26  1:41               ` Paul E. McKenney
2011-09-26  9:39                 ` Frederic Weisbecker
2011-09-26 22:34                   ` Paul E. McKenney
2011-09-27 12:07                     ` Frederic Weisbecker
2011-09-26  9:42                 ` Frederic Weisbecker
2011-09-26 22:35                   ` Paul E. McKenney
2011-09-26  9:20               ` Frederic Weisbecker
2011-09-26 22:50                 ` Paul E. McKenney
2011-09-27 12:16                   ` Frederic Weisbecker
2011-09-27 18:01                     ` Paul E. McKenney
2011-09-28 12:31                       ` Frederic Weisbecker
2011-09-28 18:40                         ` Paul E. McKenney
2011-09-28 23:46                           ` Frederic Weisbecker
2011-09-29  0:55                             ` Paul E. McKenney
2011-09-29  4:49                               ` Paul E. McKenney
2011-09-29 12:30                               ` Frederic Weisbecker
2011-09-29 17:12                                 ` Paul E. McKenney
2011-09-29 17:19                                   ` Paul E. McKenney
2011-09-29 23:18                                     ` Paul E. McKenney
2011-09-30 13:11                                   ` Frederic Weisbecker
2011-09-30 15:29                                     ` Paul E. McKenney
2011-09-30 19:24                                       ` Paul E. McKenney
2011-10-01  4:34                                         ` Paul E. McKenney
2011-10-01 12:24                                         ` Frederic Weisbecker
2011-10-01 12:28                                           ` Frederic Weisbecker
2011-10-01 16:35                                             ` Paul E. McKenney
2011-10-01 17:07                                           ` Paul E. McKenney
2011-10-02  3:23                                             ` Paul E. McKenney
2011-10-02 11:45                                               ` Frederic Weisbecker
2011-10-02 22:50                                         ` Frederic Weisbecker
2011-10-03  0:28                                           ` Paul E. McKenney [this message]
2011-10-03 12:59                                             ` Frederic Weisbecker
2011-10-03 16:22                                               ` Paul E. McKenney
2011-10-03 17:11                                                 ` Frederic Weisbecker
2011-10-02 23:07                                         ` Frederic Weisbecker
2011-10-03  0:32                                           ` Paul E. McKenney
2011-10-03 13:03                                             ` Frederic Weisbecker
2011-10-03 16:30                                               ` Paul E. McKenney
2011-10-06  0:58                                                 ` Paul E. McKenney
2011-10-06  1:59                                                   ` Paul E. McKenney
2011-10-06 12:11                                                   ` Frederic Weisbecker
2011-10-06 18:44                                                     ` Paul E. McKenney
2011-10-06 23:44                                                       ` Paul E. McKenney
2011-09-26  1:25           ` Paul E. McKenney
2011-09-26  8:48             ` Frederic Weisbecker
2011-09-26  8:49             ` Frederic Weisbecker
2011-09-26 22:30               ` Paul E. McKenney
2011-09-27 11:55                 ` Frederic Weisbecker

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=20111003002832.GC2539@linux.vnet.ibm.com \
    --to=paulmck@linux.vnet.ibm.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=dipankar@in.ibm.com \
    --cc=fweisbec@gmail.com \
    --cc=kirill@shutemov.name \
    --cc=laijs@cn.fujitsu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=tglx@linutronix.de \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).