Re: [PATCH] Fix longstanding load balancing bug in the scheduler.

[Siddha, Suresh B]

On Wed, Sep 06, 2006 at 04:38:33PM -0700, Christoph Lameter wrote:
> +int cpu_of(struct rq *rq)
> +{
> +#ifdef CONFIG_SMP
> +	return rq->cpu;
> +#else
> +	return 0;
> +#endif
> +}

Is there a reason why this is not made inline?

>  static struct sched_group *
>  find_busiest_group(struct sched_domain *sd, int this_cpu,
> -		   unsigned long *imbalance, enum idle_type idle, int
> *sd_idle)
> +		   unsigned long *imbalance, enum idle_type idle, int
> *sd_idle,
> +		   cpumask_t *cpus)
>  {
>  	struct sched_group *busiest = NULL, *this = NULL, *group =
> sd->groups;
>  	unsigned long max_load, avg_load, total_load, this_load,
> total_pwr;
> @@ -2243,7 +2254,12 @@
>  		sum_weighted_load = sum_nr_running = avg_load = 0;
>  
>  		for_each_cpu_mask(i, group->cpumask) {
> -			struct rq *rq = cpu_rq(i);
> +			struct rq *rq;
> +
> +			if (!cpu_isset(i, *cpus))
> +				continue;

In normal conditions can we make this "cpus" argument NULL and only set/use it
when we run into pinned condition? That will atleast avoid unnecessary memory
test bit operations in the normal case.

thanks,
suresh
-
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/