Re: [PATCH 03/13] writeback: per-task rate limit on balance_dirty_pages()

[Wu Fengguang <fengguang.wu@intel.com>]

On Wed, Nov 24, 2010 at 06:23:07PM +0800, Peter Zijlstra wrote:
> On Wed, 2010-11-17 at 12:27 +0800, Wu Fengguang wrote:
> > +       if (unlikely(current->nr_dirtied >= current->nr_dirtied_pause ||
> > +                    bdi->dirty_exceeded)) {
> > +               balance_dirty_pages(mapping, current->nr_dirtied);
> > +               current->nr_dirtied = 0;
> >         } 
> 
> Was it a conscious choice to use
>   current->nr_dirtied = 0
> over 
>   current->nr_dirtied -= current->nr_dirtied_pause
> ?
> 
> The former will cause a drift in pause times due to truncation of the
> excess.

It should be fine in either way, as long as the "truncated" number is
passed to balance_dirty_pages():

+               balance_dirty_pages(mapping, current->nr_dirtied);
+               current->nr_dirtied = 0;

or

+               balance_dirty_pages(mapping, current->nr_dirtied_pause);
+               current->nr_dirtied -= current->nr_dirtied_pause;

Thanks,
Fengguang