Re: [PATCH 1/3 v2] wait: introduce wait_event_exclusive_cmd

From: Peter Zijlstra
Date: Wed Apr 29 2015 - 05:31:27 EST


On Wed, Apr 29, 2015 at 10:48:53AM +0800, Yuanhan Liu wrote:
> It's just a variant of wait_event_cmd(), with exclusive flag being set.
>
> For cases like RAID5, which puts many processes to sleep until 1/4
> resources are free, a wake_up wakes up all processes to run, but
> there is one process being able to get the resource as it's protected
> by a spin lock. That ends up introducing heavy lock contentions, and
> hurts performance badly.
>
> Here introduce wait_event_exclusive_cmd to relieve the lock contention
> naturally by letting wake_up just wake up one process.
>
> Cc: Ingo Molnar <mingo@xxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Signed-off-by: Yuanhan Liu <yuanhan.liu@xxxxxxxxxxxxxxx>

Thanks!

Assuming you want to route this through Neil's MD tree:

Acked-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>

> ---
> include/linux/wait.h | 13 +++++++++++++
> 1 file changed, 13 insertions(+)
>
> diff --git a/include/linux/wait.h b/include/linux/wait.h
> index 2db8334..db78c72 100644
> --- a/include/linux/wait.h
> +++ b/include/linux/wait.h
> @@ -358,6 +358,19 @@ do { \
> __ret; \
> })
>
> +#define __wait_event_exclusive_cmd(wq, condition, cmd1, cmd2) \
> + (void)___wait_event(wq, condition, TASK_UNINTERRUPTIBLE, 1, 0, \
> + cmd1; schedule(); cmd2)
> +/*
> + * Just like wait_event_cmd(), except it sets exclusive flag
> + */
> +#define wait_event_exclusive_cmd(wq, condition, cmd1, cmd2) \
> +do { \
> + if (condition) \
> + break; \
> + __wait_event_exclusive_cmd(wq, condition, cmd1, cmd2); \
> +} while (0)
> +
> #define __wait_event_cmd(wq, condition, cmd1, cmd2) \
> (void)___wait_event(wq, condition, TASK_UNINTERRUPTIBLE, 0, 0, \
> cmd1; schedule(); cmd2)
> --
> 1.9.0
>
--
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/