mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Silas Boyd-Wickizer <sbw@mit.edu>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@kernel.org>,
	"Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com>,
	Rusty Russell <rusty@rustcorp.com.au>,
	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
	Tejun Heo <tj@kernel.org>
Subject: Re: [RFC patch 1/5] kthread: Implement park/unpark facility
Date: Thu, 14 Jun 2012 13:01:29 -0700	[thread overview]
Message-ID: <20120614200125.GA26852@mit.edu> (raw)
In-Reply-To: <20120613105815.132283147@linutronix.de>

Hello,

I've taken a stab at converting stop_machine to use kthread_park and
kthread_unpark.  I don't think stop_machine can use
smpboot_park_threads (in their current implementation) because, among
other reasons, _cpu_down calls __stop_machine after parking the
stopper threads via smpboot_park_threads.

Silas

Signed-off-by: Silas Boyd-Wickizer <sbw@mit.edu>
---
 kernel/stop_machine.c |   41 ++++++++++++++++++-----------------------
 1 file changed, 18 insertions(+), 23 deletions(-)

diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
index 2f194e9..631f133 100644
--- a/kernel/stop_machine.c
+++ b/kernel/stop_machine.c
@@ -244,7 +244,7 @@ int try_stop_cpus(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg)
 	return ret;
 }
 
-static int cpu_stopper_thread(void *data)
+static __noreturn int cpu_stopper_thread(void *data)
 {
 	struct cpu_stopper *stopper = data;
 	struct cpu_stop_work *work;
@@ -253,10 +253,8 @@ static int cpu_stopper_thread(void *data)
 repeat:
 	set_current_state(TASK_INTERRUPTIBLE);	/* mb paired w/ kthread_stop */
 
-	if (kthread_should_stop()) {
-		__set_current_state(TASK_RUNNING);
-		return 0;
-	}
+	if (kthread_should_park())
+		kthread_parkme();
 
 	work = NULL;
 	spin_lock_irq(&stopper->lock);
@@ -308,23 +306,23 @@ static int __cpuinit cpu_stop_cpu_callback(struct notifier_block *nfb,
 
 	switch (action & ~CPU_TASKS_FROZEN) {
 	case CPU_UP_PREPARE:
-		BUG_ON(stopper->thread || stopper->enabled ||
-		       !list_empty(&stopper->works));
-		p = kthread_create_on_node(cpu_stopper_thread,
-					   stopper,
-					   cpu_to_node(cpu),
-					   "migration/%d", cpu);
-		if (IS_ERR(p))
-			return notifier_from_errno(PTR_ERR(p));
-		get_task_struct(p);
-		kthread_bind(p, cpu);
-		sched_set_stop_task(cpu, p);
-		stopper->thread = p;
+		if (!stopper->thread) {
+			BUG_ON(stopper->enabled ||
+			       !list_empty(&stopper->works));
+			p = kthread_create_on_cpu(cpu_stopper_thread,
+						  stopper,
+						  cpu,
+						  "migration/%d");
+			if (IS_ERR(p))
+				return notifier_from_errno(PTR_ERR(p));
+			get_task_struct(p);
+			stopper->thread = p;
+		}
+		sched_set_stop_task(cpu, stopper->thread);
 		break;
 
 	case CPU_ONLINE:
-		/* strictly unnecessary, as first user will wake it */
-		wake_up_process(stopper->thread);
+		kthread_unpark(stopper->thread);
 		/* mark enabled */
 		spin_lock_irq(&stopper->lock);
 		stopper->enabled = true;
@@ -339,16 +337,13 @@ static int __cpuinit cpu_stop_cpu_callback(struct notifier_block *nfb,
 
 		sched_set_stop_task(cpu, NULL);
 		/* kill the stopper */
-		kthread_stop(stopper->thread);
+		kthread_park(stopper->thread);
 		/* drain remaining works */
 		spin_lock_irq(&stopper->lock);
 		list_for_each_entry(work, &stopper->works, list)
 			cpu_stop_signal_done(work->done, false);
 		stopper->enabled = false;
 		spin_unlock_irq(&stopper->lock);
-		/* release the stopper */
-		put_task_struct(stopper->thread);
-		stopper->thread = NULL;
 		break;
 	}
 #endif

  parent reply	other threads:[~2012-06-14 20:05 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-06-13 11:00 [RFC patch 0/5] Per cpu thread hotplug infrastructure Thomas Gleixner
2012-06-13 11:00 ` [RFC patch 2/5] smpboot: Provide infrastructure for percpu hotplug threads Thomas Gleixner
2012-06-13 18:33   ` Paul E. McKenney
2012-06-13 18:56     ` Thomas Gleixner
2012-06-14  8:08       ` Peter Zijlstra
2012-06-14  8:17         ` Peter Zijlstra
2012-06-15  1:53           ` Tejun Heo
2012-06-15  9:58             ` Peter Zijlstra
2012-06-14  8:37         ` Thomas Gleixner
2012-06-14  8:38           ` Peter Zijlstra
2012-06-13 18:57   ` Paul E. McKenney
2012-06-13 19:02     ` Thomas Gleixner
2012-06-13 19:17       ` Paul E. McKenney
2012-06-13 20:47         ` Paul E. McKenney
2012-06-14  4:51           ` Paul E. McKenney
2012-06-14 11:20             ` Thomas Gleixner
2012-06-14 12:59               ` Paul E. McKenney
2012-06-14 13:01                 ` Peter Zijlstra
2012-06-14 14:47                   ` Paul E. McKenney
2012-06-14 14:56                     ` Peter Zijlstra
2012-06-14 15:07                       ` Thomas Gleixner
2012-06-14 15:45                       ` Paul E. McKenney
2012-06-14 16:20                         ` Thomas Gleixner
2012-06-14 13:32                 ` Thomas Gleixner
2012-06-14 13:47                   ` Thomas Gleixner
2012-06-14 14:12                     ` Thomas Gleixner
2012-06-14 15:01                       ` Paul E. McKenney
2012-06-14 15:08                         ` Thomas Gleixner
2012-06-14 14:50                   ` Paul E. McKenney
2012-06-14 15:02                     ` Thomas Gleixner
2012-06-14 15:38                     ` Paul E. McKenney
2012-06-14 16:19                       ` Thomas Gleixner
2012-06-14 16:48                         ` Paul E. McKenney
2012-06-14 22:40             ` Paul E. McKenney
2012-06-14  8:31   ` Srivatsa S. Bhat
2012-06-14  8:44     ` Thomas Gleixner
2012-06-18  8:47   ` Cyrill Gorcunov
2012-06-18  8:50     ` Thomas Gleixner
2012-06-13 11:00 ` [RFC patch 1/5] kthread: Implement park/unpark facility Thomas Gleixner
2012-06-14  2:32   ` Namhyung Kim
2012-06-14  8:35     ` Thomas Gleixner
2012-06-14  8:59       ` Thomas Gleixner
2012-06-14  8:36   ` Srivatsa S. Bhat
2012-06-14 20:01   ` Silas Boyd-Wickizer [this message]
2012-06-14 20:13     ` Thomas Gleixner
2012-06-14 22:13   ` Silas Boyd-Wickizer
2012-06-15  1:44     ` Tejun Heo
2012-06-13 11:00 ` [RFC patch 3/5] softirq: Use hotplug thread infrastructure Thomas Gleixner
2012-06-13 11:00 ` [RFC patch 5/5] infiniband: ehca: " Thomas Gleixner
2012-06-18  6:30   ` Rusty Russell
2012-06-18  8:08     ` Thomas Gleixner
2012-06-24 10:29       ` Thomas Gleixner
2012-06-13 11:00 ` [RFC patch 4/5] watchdog: " 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=20120614200125.GA26852@mit.edu \
    --to=sbw@mit.edu \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=peterz@infradead.org \
    --cc=rusty@rustcorp.com.au \
    --cc=srivatsa.bhat@linux.vnet.ibm.com \
    --cc=tglx@linutronix.de \
    --cc=tj@kernel.org \
    /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

all inboxes | Powered by JetHome®