mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Frederic Weisbecker <fweisbec@gmail.com>
Cc: linux-kernel@vger.kernel.org, mingo@elte.hu,
	laijs@cn.fujitsu.com, dipankar@in.ibm.com,
	akpm@linux-foundation.org, mathieu.desnoyers@polymtl.ca,
	josh@joshtriplett.org, niv@us.ibm.com, tglx@linutronix.de,
	peterz@infradead.org, rostedt@goodmis.org, dhowells@redhat.com,
	edumazet@google.com, darren@dvhart.com, sbw@mit.edu
Subject: Re: [PATCH RFC nohz_full v2 6/7] nohz_full: Add full-system-idle state machine
Date: Mon, 1 Jul 2013 11:10:40 -0700	[thread overview]
Message-ID: <20130701181040.GO3773@linux.vnet.ibm.com> (raw)
In-Reply-To: <20130701163529.GO7246@somewhere.redhat.com>

On Mon, Jul 01, 2013 at 06:35:31PM +0200, Frederic Weisbecker wrote:
> On Fri, Jun 28, 2013 at 01:10:21PM -0700, Paul E. McKenney wrote:
> >  /*
> > + * Unconditionally force exit from full system-idle state.  This is
> > + * invoked when a normal CPU exits idle, but must be called separately
> > + * for the timekeeping CPU (tick_do_timer_cpu).  The reason for this
> > + * is that the timekeeping CPU is permitted to take scheduling-clock
> > + * interrupts while the system is in system-idle state, and of course
> > + * rcu_sysidle_exit() has no way of distinguishing a scheduling-clock
> > + * interrupt from any other type of interrupt.
> > + */
> > +void rcu_sysidle_force_exit(void)
> > +{
> > +	int oldstate = ACCESS_ONCE(full_sysidle_state);
> > +	int newoldstate;
> > +
> > +	/*
> > +	 * Each pass through the following loop attempts to exit full
> > +	 * system-idle state.  If contention proves to be a problem,
> > +	 * a trylock-based contention tree could be used here.
> > +	 */
> > +	while (oldstate > RCU_SYSIDLE_SHORT) {
> > +		newoldstate = cmpxchg(&full_sysidle_state,
> > +				      oldstate, RCU_SYSIDLE_NOT);
> > +		if (oldstate == newoldstate &&
> > +		    oldstate == RCU_SYSIDLE_FULL_NOTED) {
> > +			rcu_kick_nohz_cpu(tick_do_timer_cpu);
> > +			return; /* We cleared it, done! */
> > +		}
> > +		oldstate = newoldstate;
> > +	}
> > +	smp_mb(); /* Order initial oldstate fetch vs. later non-idle work. */
> > +}
> > +
> > +/*
> >   * Invoked to note entry to irq or task transition from idle.  Note that
> >   * usermode execution does -not- count as idle here!  The caller must
> >   * have disabled interrupts.
> > @@ -2474,6 +2506,214 @@ static void rcu_sysidle_exit(struct rcu_dynticks *rdtp, int irq)
> >  	atomic_inc(&rdtp->dynticks_idle);
> >  	smp_mb__after_atomic_inc();
> >  	WARN_ON_ONCE(!(atomic_read(&rdtp->dynticks_idle) & 0x1));
> > +
> > +	/*
> > +	 * If we are the timekeeping CPU, we are permitted to be non-idle
> > +	 * during a system-idle state.  This must be the case, because
> > +	 * the timekeeping CPU has to take scheduling-clock interrupts
> > +	 * during the time that the system is transitioning to full
> > +	 * system-idle state.  This means that the timekeeping CPU must
> > +	 * invoke rcu_sysidle_force_exit() directly if it does anything
> > +	 * more than take a scheduling-clock interrupt.
> > +	 */
> > +	if (smp_processor_id() == tick_do_timer_cpu)
> > +		return;
> > +
> > +	/* Update system-idle state: We are clearly no longer fully idle! */
> > +	rcu_sysidle_force_exit();
> > +}
> > +
> > +/*
> > + * Check to see if the current CPU is idle.  Note that usermode execution
> > + * does not count as idle.  The caller must have disabled interrupts.
> > + */
> > +static void rcu_sysidle_check_cpu(struct rcu_data *rdp, bool *isidle,
> > +				  unsigned long *maxj)
> > +{
> > +	int cur;
> > +	int curnmi;
> > +	unsigned long j;
> > +	struct rcu_dynticks *rdtp = rdp->dynticks;
> > +
> > +	/*
> > +	 * If some other CPU has already reported non-idle, if this is
> > +	 * not the flavor of RCU that tracks sysidle state, or if this
> > +	 * is an offline or the timekeeping CPU, nothing to do.
> > +	 */
> > +	if (!*isidle || rdp->rsp != rcu_sysidle_state ||
> > +	    cpu_is_offline(rdp->cpu) || rdp->cpu == tick_do_timer_cpu)
> > +		return;
> > +	/* WARN_ON_ONCE(smp_processor_id() != tick_do_timer_cpu); */
> > +
> > +	/*
> > +	 * Pick up current idle and NMI-nesting counters, check.  We check
> > +	 * for NMIs using RCU's main ->dynticks counter.  This works because
> > +	 * any time ->dynticks has its low bit set, ->dynticks_idle will
> > +	 * too -- unless the only reason that ->dynticks's low bit is set
> > +	 * is due to an NMI from idle.  Which is exactly the case we need
> > +	 * to account for.
> > +	 */
> > +	cur = atomic_read(&rdtp->dynticks_idle);
> > +	curnmi = atomic_read(&rdtp->dynticks);
> > +	if ((cur & 0x1) || (curnmi & 0x1)) {
> 
> I think you wanted to ignore NMIs this time because they don't read walltime?
> 
> By the way they can still read jiffies, but unlike irq_enter(), nmi_enter()
> don't catch up with missing jiffies update. So the behaviour doesn't change
> compared to !NO_HZ_FULL.

You are right, I missed this when ripping out NMI handling.  Will fix!

> > +		*isidle = 0; /* We are not idle! */
> > +		return;
> > +	}
> > +	smp_mb(); /* Read counters before timestamps. */
> > +
> > +	/* Pick up timestamps. */
> > +	j = ACCESS_ONCE(rdtp->dynticks_idle_jiffies);
> > +	/* If this CPU entered idle more recently, update maxj timestamp. */
> > +	if (ULONG_CMP_LT(*maxj, j))
> > +		*maxj = j;
> 
> So I'm a bit confused with the ordering so I'm probably going to ask a silly question.
> 
> What makes sure that we are not reading a stale value of rdtp->dynticks_idle
> in the following scenario:
> 
>     CPU 0                          CPU 1
>     
>                                    //CPU 1 idle
>                                    //rdtp(1)->dynticks_idle == 0
> 
> sysidle_check_cpu(CPU 1) {
>     rdtp(1)->dynticks_idle == 0
> }
> cmpxchg(full_sysidle_state, 
>         ...RCU_SYSIDLE_SHORT)
>                                    rcu_irq_exit() {

rcu_irq_enter(), right?

>                                          rdtp(1)->dynticks_idle = 1
>                                          smp_mb()
>                                          rcu_sysidle_force_exit() {
>                                             full_sysidle_state == RCU_SYSIDLE_SHORT
>                                             // no cmpxchg
>                                             smp_mb()
>                                    ...
> 
> [1]
> sysidle_check_cpu(CPU 1) {
>     rdtp(1)->dynticks_idle == 0
> }
> 
> cmpxchg(RCU_SYSIDLE_FULL, ...)

You know, I had an RCU_SYSIDLE_LONG state for this purpose, but later
convinced myself that I didn't need it.  :-/

Time to go put it back in, and thank you for your careful review!

							Thanx, Paul

> [2]
> sysidle_check_cpu(CPU 1) {
>     rdtp(1)->dynticks_idle == 0
> }
> 
> cmpxchg(RCU_SYSIDLE_FULL_NOTED, ...)
> 
> 
> I mean in [1] and [2] I can't see something in the ordering that guarantees that we see
> the new value rdtp(1)->dynticks_idle == 1.
> 


  reply	other threads:[~2013-07-01 18:10 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-28 20:09 [PATCH RFC nohz_full 0/7] v2 Provide infrastructure for full-system idle Paul E. McKenney
2013-06-28 20:10 ` [PATCH RFC nohz_full v2 1/7] nohz_full: Add Kconfig parameter for scalable detection of all-idle state Paul E. McKenney
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 2/7] nohz_full: Add rcu_dyntick data " Paul E. McKenney
2013-07-01 15:31     ` Josh Triplett
2013-07-01 15:52       ` Paul E. McKenney
2013-07-01 18:16         ` Josh Triplett
2013-07-01 18:23           ` Paul E. McKenney
2013-07-01 18:34             ` Josh Triplett
2013-07-01 19:16               ` Paul E. McKenney
2013-07-02  5:10                 ` Mike Galbraith
2013-07-02  5:58                   ` Paul E. McKenney
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 3/7] nohz_full: Add per-CPU idle-state tracking Paul E. McKenney
2013-07-01 15:33     ` Josh Triplett
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 4/7] nohz_full: Add full-system idle states and variables Paul E. McKenney
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 5/7] nohz_full: Add full-system-idle arguments to API Paul E. McKenney
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 6/7] nohz_full: Add full-system-idle state machine Paul E. McKenney
2013-07-01 16:35     ` Frederic Weisbecker
2013-07-01 18:10       ` Paul E. McKenney [this message]
2013-07-01 20:55         ` Frederic Weisbecker
2013-07-01 16:53     ` Frederic Weisbecker
2013-07-01 18:17       ` Paul E. McKenney
2013-07-01 21:38     ` Frederic Weisbecker
2013-07-01 22:51       ` Paul E. McKenney
2013-06-28 20:10   ` [PATCH RFC nohz_full v2 7/7] nohz_full: Force RCU's grace-period kthreads onto timekeeping CPU Paul E. McKenney
2013-07-01 15:19 ` [PATCH RFC nohz_full 0/7] v2 Provide infrastructure for full-system idle Andi Kleen
2013-07-01 16:03   ` Paul E. McKenney
2013-07-01 16:19     ` Andi Kleen
2013-07-01 19:19       ` Paul E. McKenney
2013-07-01 19:43 ` Christoph Lameter
2013-07-01 19:56   ` Paul E. McKenney
2013-07-01 20:24     ` Christoph Lameter
2013-07-01 20:43       ` Thomas Gleixner

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=20130701181040.GO3773@linux.vnet.ibm.com \
    --to=paulmck@linux.vnet.ibm.com \
    --cc=akpm@linux-foundation.org \
    --cc=darren@dvhart.com \
    --cc=dhowells@redhat.com \
    --cc=dipankar@in.ibm.com \
    --cc=edumazet@google.com \
    --cc=fweisbec@gmail.com \
    --cc=josh@joshtriplett.org \
    --cc=laijs@cn.fujitsu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mathieu.desnoyers@polymtl.ca \
    --cc=mingo@elte.hu \
    --cc=niv@us.ibm.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=sbw@mit.edu \
    --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

Powered by JetHome