All of lore.kernel.org
 help / color / mirror / Atom feed
From: Oleg Nesterov <oleg@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
	Ingo Molnar <mingo@elte.hu>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] sched: RCU-protect __set_task_cpu() in set_task_cpu()
Date: Mon, 6 Jun 2011 18:46:57 +0200	[thread overview]
Message-ID: <20110606164657.GA20752@redhat.com> (raw)
In-Reply-To: <1307351198.2353.7415.camel@twins>

On 06/06, Peter Zijlstra wrote:
>
> You're right, p->pi_lock for wakeups, rq->lock for runnable tasks.

Good, thanks.

Help! I have another question.

	try_to_wake_up:

		raw_spin_lock_irqsave(&p->pi_lock, flags);
		if (!(p->state & state))
			goto out;

		cpu = task_cpu(p);

		if (p->on_rq && ttwu_remote(p, wake_flags))
			goto stat;

This doesn't look a bit confusing, we can't trust "cpu = task_cpu" before
we check ->on_rq. OK, not a problem, this cpu number can only be used in
ttwu_stat(cpu).

But ttwu_stat(cpu) in turn does

	if (cpu != task_cpu(p))
		schedstat_inc(p, se.statistics.nr_wakeups_migrate);

Ignoring the theoretical races with pull_task/etc, how it is possible
that cpu != task_cpu(p) ? Another caller is try_to_wake_up_local(), it
obviously can't trigger this case.

This looks broken to me. Looking at its name, I guess nr_wakeups_migrate
should be incremented if ttwu does set_task_cpu(), correct?

IOW. Don't we need something like the (untested/ucompiled) patch below?
_If_ I am right, I can resend it with the changelog/etc but please feel
free to make another fix.

Oleg.

--- x/kernel/sched.c
+++ x/kernel/sched.c
@@ -2423,13 +2423,14 @@ static void update_avg(u64 *avg, u64 sam
 #endif
 
 static void
-ttwu_stat(struct task_struct *p, int cpu, int wake_flags)
+ttwu_stat(struct task_struct *p, bool migrate, int wake_flags)
 {
 #ifdef CONFIG_SCHEDSTATS
 	struct rq *rq = this_rq();
 
 #ifdef CONFIG_SMP
 	int this_cpu = smp_processor_id();
+	int cpu = task_cpu(p);
 
 	if (cpu == this_cpu) {
 		schedstat_inc(rq, ttwu_local);
@@ -2455,7 +2456,7 @@ ttwu_stat(struct task_struct *p, int cpu
 	if (wake_flags & WF_SYNC)
 		schedstat_inc(p, se.statistics.nr_wakeups_sync);
 
-	if (cpu != task_cpu(p))
+	if (migrate)
 		schedstat_inc(p, se.statistics.nr_wakeups_migrate);
 
 #endif /* CONFIG_SCHEDSTATS */
@@ -2630,6 +2631,7 @@ try_to_wake_up(struct task_struct *p, un
 {
 	unsigned long flags;
 	int cpu, success = 0;
+	bool migrate = false;
 
 	smp_wmb();
 	raw_spin_lock_irqsave(&p->pi_lock, flags);
@@ -2637,7 +2639,6 @@ try_to_wake_up(struct task_struct *p, un
 		goto out;
 
 	success = 1; /* we're going to change ->state */
-	cpu = task_cpu(p);
 
 	if (p->on_rq && ttwu_remote(p, wake_flags))
 		goto stat;
@@ -2674,13 +2675,15 @@ try_to_wake_up(struct task_struct *p, un
 		p->sched_class->task_waking(p);
 
 	cpu = select_task_rq(p, SD_BALANCE_WAKE, wake_flags);
-	if (task_cpu(p) != cpu)
+	if (task_cpu(p) != cpu) {
 		set_task_cpu(p, cpu);
+		migrate = true;
+	}
 #endif /* CONFIG_SMP */
 
 	ttwu_queue(p, cpu);
 stat:
-	ttwu_stat(p, cpu, wake_flags);
+	ttwu_stat(p, migrate, wake_flags);
 out:
 	raw_spin_unlock_irqrestore(&p->pi_lock, flags);
 
@@ -2716,7 +2719,7 @@ static void try_to_wake_up_local(struct 
 		ttwu_activate(rq, p, ENQUEUE_WAKEUP);
 
 	ttwu_do_wakeup(rq, p, 0);
-	ttwu_stat(p, smp_processor_id(), 0);
+	ttwu_stat(p, false, 0);
 out:
 	raw_spin_unlock(&p->pi_lock);
 }



  reply	other threads:[~2011-06-06 17:36 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-31 17:26 [PATCH] sched: RCU-protect __set_task_cpu() in set_task_cpu() Sergey Senozhatsky
2011-05-31 19:45 ` Peter Zijlstra
2011-06-03 15:37 ` Peter Zijlstra
2011-06-03 18:16   ` Sergey Senozhatsky
2011-06-03 22:49   ` Sergey Senozhatsky
2011-06-05 19:12   ` Oleg Nesterov
2011-06-06  9:06     ` Peter Zijlstra
2011-06-06 16:46       ` Oleg Nesterov [this message]
2011-06-07  9:31         ` Peter Zijlstra
2011-06-07 14:03           ` Oleg Nesterov
2011-06-06 13:43     ` Peter Zijlstra
2011-06-07 12:03   ` [tip:sched/urgent] sched: Fix/clarify set_task_cpu() locking rules tip-bot for Peter Zijlstra

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=20110606164657.GA20752@redhat.com \
    --to=oleg@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=peterz@infradead.org \
    --cc=sergey.senozhatsky@gmail.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.