Re: [PATCH] mm: Set N_CPU to node_states during boot

[Andrew Morton]

On Mon, 14 Oct 2013 14:04:56 -0600 Toshi Kani <toshi.kani@xxxxxx> wrote:

> After a system booted, N_CPU is not set to any node as has_cpu
> shows an empty line.
> 
>   # cat /sys/devices/system/node/has_cpu
>   (show-empty-line)
> 
> setup_vmstat() registers its CPU notifier callback,
> vmstat_cpuup_callback(), which marks N_CPU to a node when
> a CPU is put into online.  However, setup_vmstat() is called
> after all CPUs are launched in the boot sequence.
> 
> Change setup_vmstat() to mark N_CPU to the nodes with online
> CPUs at boot.
> 
> ...
>
> --- a/mm/vmstat.c
> +++ b/mm/vmstat.c
> @@ -1276,8 +1276,10 @@ static int __init setup_vmstat(void)
>  
>  	register_cpu_notifier(&vmstat_notifier);
>  
> -	for_each_online_cpu(cpu)
> +	for_each_online_cpu(cpu) {
>  		start_cpu_timer(cpu);
> +		node_set_state(cpu_to_node(cpu), N_CPU);
> +	}
>  #endif
>  #ifdef CONFIG_PROC_FS
>  	proc_create("buddyinfo", S_IRUGO, NULL, &fragmentation_file_operations);

This seems a bit hacky.  Would it not be better to register
vmstat_notifier() before bringing up CPUs?


And this patch might be racy as well - what happens if a CPU comes up
and goes down again before setup_vmstat() is called?

(Where does N_CPU get cleared?  It doesn't, afaict.  Should we clear it
if a node's final CPU goes offline?)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/