Re: [PATCH][RFC] dirty balancing for cgroups

[Peter Zijlstra]

On Wed, 2008-07-09 at 15:00 +0900, YAMAMOTO Takashi wrote:
> hi,
> 
> the following patch is a simple implementation of
> dirty balancing for cgroups.  any comments?
> 
> it depends on the following fix:
> 	http://lkml.org/lkml/2008/7/8/428
> 
> YAMAMOTO Takashi
> 
> 
> Signed-off-by: YAMAMOTO Takashi <yamamoto@xxxxxxxxxxxxx>
> ---

Yamamoto-san,

> @@ -408,7 +412,11 @@ get_dirty_limits(long *pbackground, long *pdirty, long *pbdi_dirty,
>  
>  		*pbdi_dirty = bdi_dirty;
>  		clip_bdi_dirty_limit(bdi, dirty, pbdi_dirty);
> -		task_dirty_limit(current, pbdi_dirty);
> +		task_dirty = *pbdi_dirty;
> +		task_dirty_limit(current, &task_dirty);
> +		cgroup_dirty = *pbdi_dirty;
> +		memdirtylimitcgroup_dirty_limit(current, &cgroup_dirty);
> +		*pbdi_dirty = min(task_dirty, cgroup_dirty);
>  	}
>  }

I think this is wrong - is basically breaks task dirty throttling within
groups. You'd need a multiplicative operation, something like:

  bdi_dirty = dirty * p(bdi) * p(cgroup) * (1 - p(task))

However then we still have problems... see the next email further down
the thread.

--
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/