Re: [RFC patch 2/5] smpboot: Provide infrastructure for percpu hotplug threads

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Wed, Jun 13, 2012 at 12:17:45PM -0700, Paul E. McKenney wrote:
> On Wed, Jun 13, 2012 at 09:02:55PM +0200, Thomas Gleixner wrote:
> > On Wed, 13 Jun 2012, Paul E. McKenney wrote:
> > > On Wed, Jun 13, 2012 at 11:00:54AM -0000, Thomas Gleixner wrote:
> > > >  	/* Now call notifier in preparation. */
> > > >  	cpu_notify(CPU_ONLINE | mod, hcpu);
> > > > +	smpboot_unpark_threads(cpu);
> > > 
> > > OK, RCU must use the lower-level interfaces, given that one of
> > > then CPU_ONLINE notifiers might invoke synchronize_rcu().
> > 
> > We can start the threads before the notifiers. There is no
> > restriction.
> 
> Sounds very good in both cases!

Just for reference, here is what I am using.


Anticipatory changes in advance of Thomas's next patchset

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

diff --git a/kernel/cpu.c b/kernel/cpu.c
index 6ab5d82..510ba5e 100644
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -372,10 +372,10 @@ static int __cpuinit _cpu_up(unsigned int cpu, int tasks_frozen)
 	if (ret != 0)
 		goto out_notify;
 	BUG_ON(!cpu_online(cpu));
+	smpboot_unpark_threads(cpu);
 
 	/* Now call notifier in preparation. */
 	cpu_notify(CPU_ONLINE | mod, hcpu);
-	smpboot_unpark_threads(cpu);
 
 out_notify:
 	if (ret != 0)
diff --git a/kernel/smpboot.c b/kernel/smpboot.c
index d7a3659..7d54e7c 100644
--- a/kernel/smpboot.c
+++ b/kernel/smpboot.c
@@ -201,7 +201,7 @@ void smpboot_park_threads(unsigned int cpu)
 	struct smp_hotplug_thread *cur;
 
 	mutex_lock(&smpboot_threads_lock);
-	list_for_each_entry(cur, &hotplug_threads, list)
+	list_for_each_entry_reverse(cur, &hotplug_threads, list)
 		smpboot_park_thread(cur, cpu);
 	mutex_unlock(&smpboot_threads_lock);
 }