Re: [PATCH] Export CPU Scheduler Tunables via DebugFS

[Andreas Mohr <andi@rhlx01.fht-esslingen.de>]

Hi,

On Sun, Sep 03, 2006 at 08:28:02AM -0500, Brandon Philips wrote:
> This patch exports the CPU scheduler tunables via DebugFS.


> +int min_timeslice;
> +#define MIN_TIMESLICE		(min_timeslice)
> +int def_timeslice;
> +#define DEF_TIMESLICE		(def_timeslice)
> +int on_runqueue_weight;
> +#define ON_RUNQUEUE_WEIGHT	(on_runqueue_weight)
> +int child_penalty;
> +#define CHILD_PENALTY		(child_penalty)
> +int parent_penalty;
> +#define PARENT_PENALTY		(parent_penalty)
> +int exit_weight;
> +#define EXIT_WEIGHT		(exit_weight)
> +int prio_bonus_ratio;
> +#define PRIO_BONUS_RATIO	(prio_bonus_ratio)
> +int max_bonus;
> +#define MAX_BONUS		(max_bonus)
> +int interactive_delta;
> +#define INTERACTIVE_DELTA	(interactive_delta)
> +int max_sleep_avg;
> +#define MAX_SLEEP_AVG		(max_sleep_avg)
> +int starvation_limit;
> +#define STARVATION_LIMIT	(starvation_limit)

__read_mostly? Except for those variables which are being tweaked
quite often by live scheduler code... (probably none)

This only concerns the case of CONFIG_DEBUGFS_SCHED being activated,
but it may still be nice to have, I think, since scheduler stuff
should rather be fast than slow.

But since it is a debugging-only option omitting __read_mostly may
make sense...

Andreas Mohr

-- 
VGER BF report: H 0.00129204
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/