Re: [patch] export percpu cpuacct cgroup stats

[Li Zefan <lizf@cn.fujitsu.com>]

(I guess it's more appropriate to ask Ingo to pick up this patch)

Ken Chen wrote:
> This patch export per-cpu CPU cycle usage for a given cpuacct cgroup.
> There is a need for a user space monitor daemon to track group CPU
> usage on per-cpu base.  It is also useful for monitor CFS load
> balancer behavior by tracking per CPU group usage.
> 
> 
> Signed-off-by: Ken Chen <kenchen@google.com>
> 
> diff --git a/kernel/sched.c b/kernel/sched.c
> index b7480fb..cd78948 100644
> --- a/kernel/sched.c
> +++ b/kernel/sched.c
> @@ -9345,12 +9345,34 @@ out:
>  	return err;
>  }
> 
> +static int cpuacct_percpu_seq_read(struct cgroup *cgroup, struct cftype *cft,
> +				   struct seq_file *m)
> +{
> +	struct cpuacct *ca = cgroup_ca(cgroup);
> +	u64 percpu;
> +	int i;
> +
> +	for_each_possible_cpu(i) {
> +		spin_lock_irq(&cpu_rq(i)->lock);
> +		percpu = *percpu_ptr(ca->cpuusage, i);
> +		spin_unlock_irq(&cpu_rq(i)->lock);
> +		seq_printf(m, "%lld ", percpu);

Should be %llu. And if we don't cast percpu to unsigned long long, will it trigger
compile warning in some archs?

> +	}
> +	seq_printf(m, "\n");
> +	return 0;
> +}
> +
>  static struct cftype files[] = {
>  	{
>  		.name = "usage",
>  		.read_u64 = cpuusage_read,
>  		.write_u64 = cpuusage_write,
>  	},
> +	{
> +		.name = "percpu",

IMO percpu_usage is a better name.

> +		.read_seq_string = cpuacct_percpu_seq_read,
> +	},
> +
>  };
>