Re: [PATCH] rcu: more careful check for the last non-dyntick-idle CPU

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

On Tue, Mar 30, 2010 at 06:40:36PM +0800, Lai Jiangshan wrote:
> Warning: I just did build test.
> 
> Subject: [PATCH] rcu: more careful check for the last non-dyntick-idle CPU
> 
> If a CPU is not in nohz_cpu_mask, it does not means this
> CPU is a non-dyntick-idle CPU, because this CPU may be offline or not exist.

Good catch, queued, thank you!

							Thanx, Paul

> Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
> ---
> diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h
> index 4befb64..6086a7d 100644
> --- a/kernel/rcutree_plugin.h
> +++ b/kernel/rcutree_plugin.h
> @@ -1026,7 +1026,7 @@ int rcu_needs_cpu(int cpu)
> 
>  	/* Don't bother unless we are the last non-dyntick-idle CPU. */
>  	for_each_cpu_not(thatcpu, nohz_cpu_mask)
> -		if (thatcpu != cpu) {
> +		if (cpu_online(thatcpu) && thatcpu != cpu) {
>  			per_cpu(rcu_dyntick_drain, cpu) = 0;
>  			per_cpu(rcu_dyntick_holdoff, cpu) = jiffies - 1;
>  			return rcu_needs_cpu_quick_check(cpu);