Re: [sched/get_online_cpus] INFO: task swapper/0:1 blocked for more than 120 seconds.

[Peter Zijlstra <peterz@infradead.org>]

On Mon, Nov 11, 2013 at 03:47:11PM +0800, Michael wang wrote:
> Hi, Fengguang
> 
> On 11/10/2013 06:16 PM, Fengguang Wu wrote:
> > Greetings,
> > 
> > I got the below dmesg and the first bad commit is
> 
> I guess this will disappear when '!CONFIG_RCU_BOOST'...
> 
> AFAIK, if the rsp was in boost mode, we count on smpboot-thread
> 'rcu_cpu_thread_spec' to finish the callback, which will be
> parked before do sync-rcu inside _cpu_down(), if that was true,
> then the sync will never finish...
> 
> May be some brainless fix like this?
> 
> 
> 
> diff --git a/kernel/cpu.c b/kernel/cpu.c
> index 63aa50d..aa24338 100644
> --- a/kernel/cpu.c
> +++ b/kernel/cpu.c
> @@ -306,7 +306,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
>                                 __func__, cpu);
>                 goto out_release;
>         }
> -       smpboot_park_threads(cpu);
>  
>         /*
>          * By now we've cleared cpu_active_mask, wait for all preempt-disabled
> @@ -321,6 +320,8 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
>  #endif
>         synchronize_rcu();
>  
> +       smpboot_park_threads(cpu);
> +
>         /*
>          * So now all preempt/rcu users must observe !cpu_active().
>          */

Good thinking.. Wu did this cure stuff?