On Fri, Apr 12, 2013 at 11:39:51AM +0530, Srivatsa S. Bhat wrote:
> On 04/12/2013 11:07 AM, Ingo Molnar wrote:
> > 
> > * Robin Holt <[email protected]> wrote:
> > 
> >> For the v3.9 release, can we consider my awful patch?
> > 
> > How about trying what I suggested, to make reboot affine to the boot CPU 
> > explicitly, not by shutting down all the other CPUs, but by 
> > set_cpus_allowed() or 
> > so?
> > 
> 
> I agree, that sounds like the right thing to do for 3.9. Of course, it would 
> be
> nice if Shawn could verify that doing that doesn't break his platform due to
> some unknown corner case.
> 
> > That should solve the regression, without the ugly special-casing - while 
> > giving 
> > time to address the hot-unplug performance bottleneck.
> > 
> > Once that is done disable_nonboot_cpus() can be used again for reboot.
> > 
> > (But no strong feelings either way - both solutions are a workaround in a 
> > sense.)


>From 1767003c943325e52ac78cac6fdbaf2ab638888d Mon Sep 17 00:00:00 2001
From: Robin Holt <[email protected]>
Date: Wed, 3 Apr 2013 13:52:00 -0500
Subject: [PATCH] Migrate shutdown/reboot to boot cpu.

We recently noticed that reboot of a 1024 cpu machine takes approx 16
minutes of just stopping the cpus.  The slowdown was tracked to commit
f96972f.

The current implementation does all the work of hot removing the cpus
before halting the system.  We are switching to just migrating to the
boot cpu and then calling continuing with shutdown/reboot.

This also has the effect of not breaking x86's command line parameter for
specifying the reboot cpu.  Note, this code was shamelessly copied from
arch/x86/kernel/reboot.c with bits removed pertaining to the reboot_cpu
command line parameter.

Signed-off-by: Robin Holt <[email protected]>
To: Shawn Guo <[email protected]>
To: Ingo Molnar <[email protected]>
To: Russ Anderson <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: Lai Jiangshan <[email protected]>
Cc: Linus Torvalds <[email protected]>
Cc: Linux Kernel Mailing List <[email protected]>
Cc: Michel Lespinasse <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Cc: "Paul E. McKenney" <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Robin Holt <[email protected]>
Cc: "[email protected]" <[email protected]>
Cc: Tejun Heo <[email protected]>
Cc: the arch/x86 maintainers <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: <[email protected]>
---
 kernel/sys.c | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

diff --git a/kernel/sys.c b/kernel/sys.c
index 0da73cf..4d1047d 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -357,6 +357,19 @@ int unregister_reboot_notifier(struct notifier_block *nb)
 }
 EXPORT_SYMBOL(unregister_reboot_notifier);
 
+void migrate_to_boot_cpu(void)
+{
+       /* The boot cpu is always logical cpu 0 */
+       int reboot_cpu_id = 0;
+
+       /* Make certain the cpu I'm about to reboot on is online */
+       if (!cpu_online(reboot_cpu_id))
+               reboot_cpu_id = smp_processor_id();
+
+       /* Make certain I only run on the appropriate processor */
+       set_cpus_allowed_ptr(current, cpumask_of(reboot_cpu_id));
+}
+
 /**
  *     kernel_restart - reboot the system
  *     @cmd: pointer to buffer containing command to execute for restart
@@ -368,7 +381,7 @@ EXPORT_SYMBOL(unregister_reboot_notifier);
 void kernel_restart(char *cmd)
 {
        kernel_restart_prepare(cmd);
-       disable_nonboot_cpus();
+       migrate_to_boot_cpu();
        syscore_shutdown();
        if (!cmd)
                printk(KERN_EMERG "Restarting system.\n");
@@ -414,7 +427,7 @@ void kernel_power_off(void)
        kernel_shutdown_prepare(SYSTEM_POWER_OFF);
        if (pm_power_off_prepare)
                pm_power_off_prepare();
-       disable_nonboot_cpus();
+       migrate_to_boot_cpu();
        syscore_shutdown();
        printk(KERN_EMERG "Power down.\n");
        kmsg_dump(KMSG_DUMP_POWEROFF);
-- 
1.8.1.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Reply via email to