Re: [PATCH 30/45] vmscan: lumpy pageout

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

On Wed, Oct 07, 2009 at 06:38:37PM +0800, Nikita Danilov wrote:
> Hello,
> 
>  > When pageout a dirty page, try to piggy back more consecutive dirty
>  > pages (up to 512KB) to improve IO efficiency.
>  >
>  > Only ext3/reiserfs which don't have its own aops->writepages are
>  > supported in this initial version.
>  >
> 
> [...]
> 
>  >  		/*
>  > +		 * only write_cache_pages() supports for_reclaim for now
>  > +		 */
>  > +		if (!mapping->a_ops->writepages) {
>  > +			wbc.range_start = (page->index + 1) << PAGE_CACHE_SHIFT;
>  > +			wbc.nr_to_write = LUMPY_PAGEOUT_PAGES - 1;
>  > +			generic_writepages(mapping, &wbc);
>  > +		}
> 
> This might end up calling ->writepage() on a page_mapped() page,
> which used to be a problem, at least for shmem (see
> BUG_ON(page_mapped(page)) in shmem_writepage()).

Good catch, thanks a lot!

> I tried to do a similar thing in the past, but eventually decided
> do explicitly call ->writepage() in a loop from vmscan.c. This
> way one can also filter out referenced and active pages. See
> http://linuxhacker.ru/~nikita/patches/2.6.14-rc5/05-cluster-pageout.patch,
> it also contains a comment explaining shmem problem.

Glad to know about your experiences :) Interestingly I started with
->writepage() and then switch to ->writepages() because filesystems
behave better with the latter (i.e. less file fragmentation).

I'd like to just ignore the shmem case, by adding a
bdi_cap_writeback_dirty() check. Because clustered writing to swap
device may be a less gain.

Page filtering should also be possible in write_cache_pages().  But
what do you mean by "hard-to-fix races against inode reclamation"?

Thanks,
Fengguang