Message-ID: <20100721130435.GH16031@cmpxchg.org>
Date:	Wed, 21 Jul 2010 15:04:35 +0200
From:	Johannes Weiner <hannes@...xchg.org>
To:	Mel Gorman <mel@....ul.ie>
Cc:	linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
	linux-mm@...ck.org, Dave Chinner <david@...morbit.com>,
	Chris Mason <chris.mason@...cle.com>,
	Nick Piggin <npiggin@...e.de>, Rik van Riel <riel@...hat.com>,
	Christoph Hellwig <hch@...radead.org>,
	Wu Fengguang <fengguang.wu@...el.com>,
	KAMEZAWA Hiroyuki <kamezawa.hiroyu@...fujitsu.com>,
	KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Andrea Arcangeli <aarcange@...hat.com>
Subject: Re: [PATCH 4/8] vmscan: Do not writeback filesystem pages in direct
 reclaim

On Wed, Jul 21, 2010 at 12:52:50PM +0100, Mel Gorman wrote:
> On Wed, Jul 21, 2010 at 12:02:18AM +0200, Johannes Weiner wrote:
> > What
> > I had in mind is the attached patch.  It is not tested and hacked up
> > rather quickly due to time constraints, sorry, but you should get the
> > idea.  I hope I did not miss anything fundamental.
> > 
> > Note that since only kswapd enters pageout() anymore, everything
> > depending on PAGEOUT_IO_SYNC in there is moot, since there are no sync
> > cycles for kswapd.  Just to mitigate the WTF-count on the patch :-)
> > 
> 
> Anon page writeback can enter pageout. See
> 
> static inline bool reclaim_can_writeback(struct scan_control *sc,
>                                         struct page *page)
> {
>         return !page_is_file_cache(page) || current_is_kswapd();
> }
> 
> So the logic still applies.

Yeah, I noticed it only after looking at it again this morning.  My
bad, it got a bit late when I wrote it.

> > @@ -643,12 +639,14 @@ static noinline_for_stack void free_page
> >   * shrink_page_list() returns the number of reclaimed pages
> >   */
> >  static unsigned long shrink_page_list(struct list_head *page_list,
> > -					struct scan_control *sc,
> > -					enum pageout_io sync_writeback)
> > +				      struct scan_control *sc,
> > +				      enum pageout_io sync_writeback,
> > +				      int *dirty_seen)
> >  {
> >  	LIST_HEAD(ret_pages);
> >  	LIST_HEAD(free_pages);
> >  	int pgactivate = 0;
> > +	unsigned long nr_dirty = 0;
> >  	unsigned long nr_reclaimed = 0;
> >  
> >  	cond_resched();
> > @@ -657,7 +655,7 @@ static unsigned long shrink_page_list(st
> >  		enum page_references references;
> >  		struct address_space *mapping;
> >  		struct page *page;
> > -		int may_enter_fs;
> > +		int may_pageout;
> >  
> >  		cond_resched();
> >  
> > @@ -681,10 +679,15 @@ static unsigned long shrink_page_list(st
> >  		if (page_mapped(page) || PageSwapCache(page))
> >  			sc->nr_scanned++;
> >  
> > -		may_enter_fs = (sc->gfp_mask & __GFP_FS) ||
> > +		/*
> > +		 * To prevent stack overflows, only kswapd can enter
> > +		 * the filesystem.  Swap IO is always fine (for now).
> > +		 */
> > +		may_pageout = current_is_kswapd() ||
> >  			(PageSwapCache(page) && (sc->gfp_mask & __GFP_IO));
> >  
> 
> We lost the __GFP_FS check and it's vaguely possible kswapd could call the
> allocator with GFP_NOFS. While you check it before wait_on_page_writeback it
> needs to be checked before calling pageout(). I toyed around with
> creating a may_pageout that took everything into account but I couldn't
> convince myself there was no holes or serious change in functionality.

Yeah, I checked balance_pgdat(), saw GFP_KERNEL and went for it.  But
it's probably better to keep such dependencies out.

> Ok, is this closer to what you had in mind?

IMHO this is (almost) ready to get merged, so I am including the
nitpicking comments :-)

> ==== CUT HERE ====
> [PATCH] vmscan: Do not writeback filesystem pages in direct reclaim
> 
> When memory is under enough pressure, a process may enter direct
> reclaim to free pages in the same manner kswapd does. If a dirty page is
> encountered during the scan, this page is written to backing storage using
> mapping->writepage. This can result in very deep call stacks, particularly
> if the target storage or filesystem are complex. It has already been observed
> on XFS that the stack overflows but the problem is not XFS-specific.
> 
> This patch prevents direct reclaim writing back filesystem pages by checking
> if current is kswapd or the page is anonymous before writing back.  If the
> dirty pages cannot be written back, they are placed back on the LRU lists
> for either background writing by the BDI threads or kswapd. If in direct
> lumpy reclaim and dirty pages are encountered, the process will stall for
> the background flusher before trying to reclaim the pages again.
> 
> As the call-chain for writing anonymous pages is not expected to be deep
> and they are not cleaned by flusher threads, anonymous pages are still
> written back in direct reclaim.
> 
> Signed-off-by: Mel Gorman <mel@....ul.ie>
> Acked-by: Rik van Riel <riel@...hat.com>
> 
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 6587155..e3a5816 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c

[...]

Does factoring pageout() still make sense in this patch?  It does not
introduce a second callsite.

> @@ -639,18 +645,25 @@ static noinline_for_stack void free_page_list(struct list_head *free_pages)
>  	pagevec_free(&freed_pvec);
>  }
>  
> +/* Direct lumpy reclaim waits up to 5 seconds for background cleaning */
> +#define MAX_SWAP_CLEAN_WAIT 50

That's placed a bit randomly now that shrink_page_list() doesn't use
it anymore.  I moved it just above shrink_inactive_list() but maybe it
would be better at the file's head?

>  /*
>   * shrink_page_list() returns the number of reclaimed pages
>   */
>  static unsigned long shrink_page_list(struct list_head *page_list,
>  					struct scan_control *sc,
> -					enum pageout_io sync_writeback)
> +					enum pageout_io sync_writeback,
> +					unsigned long *nr_still_dirty)
>  {
> -	LIST_HEAD(ret_pages);
>  	LIST_HEAD(free_pages);
> -	int pgactivate = 0;
> +	LIST_HEAD(putback_pages);
> +	LIST_HEAD(dirty_pages);
> +	int pgactivate;
> +	unsigned long nr_dirty = 0;
>  	unsigned long nr_reclaimed = 0;
>  
> +	pgactivate = 0;

Spurious change?

>  	cond_resched();
>  
>  	while (!list_empty(page_list)) {
> @@ -741,7 +754,18 @@ static unsigned long shrink_page_list(struct list_head *page_list,
>  			}
>  		}
>  
> -		if (PageDirty(page)) {
> +		if (PageDirty(page))  {

Ha!

> +			/*
> +			 * Only kswapd can writeback filesystem pages to
> +			 * avoid risk of stack overflow
> +			 */
> +			if (page_is_file_cache(page) && !current_is_kswapd()) {
> +				list_add(&page->lru, &dirty_pages);
> +				unlock_page(page);
> +				nr_dirty++;
> +				goto keep_dirty;
> +			}

I don't understand why you keep the extra dirty list.  Couldn't this
just be `goto keep_locked'?

>  			if (references == PAGEREF_RECLAIM_CLEAN)
>  				goto keep_locked;
>  			if (!may_enter_fs)
> @@ -852,13 +876,19 @@ activate_locked:
>  keep_locked:
>  		unlock_page(page);
>  keep:
> -		list_add(&page->lru, &ret_pages);
> +		list_add(&page->lru, &putback_pages);
> +keep_dirty:
>  		VM_BUG_ON(PageLRU(page) || PageUnevictable(page));
>  	}
>  
>  	free_page_list(&free_pages);
>  
> -	list_splice(&ret_pages, page_list);
> +	if (nr_dirty) {
> +		*nr_still_dirty = nr_dirty;

You either have to set *nr_still_dirty unconditionally or
(re)initialize the variable in shrink_inactive_list().

> +		list_splice(&dirty_pages, page_list);
> +	}
> +	list_splice(&putback_pages, page_list);

When we retry those pages, the dirty ones come last on the list.  Was
this maybe the intention behind collecting dirties separately?

> @@ -1245,6 +1275,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct zone *zone,
>  	unsigned long nr_active;
>  	unsigned long nr_anon;
>  	unsigned long nr_file;
> +	unsigned long nr_dirty;
>  
>  	while (unlikely(too_many_isolated(zone, file, sc))) {
>  		congestion_wait(BLK_RW_ASYNC, HZ/10);
> @@ -1293,26 +1324,34 @@ shrink_inactive_list(unsigned long nr_to_scan, struct zone *zone,
>  
>  	spin_unlock_irq(&zone->lru_lock);
>  
> -	nr_reclaimed = shrink_page_list(&page_list, sc, PAGEOUT_IO_ASYNC);
> +	nr_reclaimed = shrink_page_list(&page_list, sc, PAGEOUT_IO_ASYNC,
> +								&nr_dirty);
>  
>  	/*
> -	 * If we are direct reclaiming for contiguous pages and we do
> +	 * If specific pages are needed such as with direct reclaiming
> +	 * for contiguous pages or for memory containers and we do
>  	 * not reclaim everything in the list, try again and wait
> -	 * for IO to complete. This will stall high-order allocations
> -	 * but that should be acceptable to the caller
> +	 * for IO to complete. This will stall callers that require
> +	 * specific pages but it should be acceptable to the caller
>  	 */
> -	if (nr_reclaimed < nr_taken && !current_is_kswapd() &&
> -			sc->lumpy_reclaim_mode) {
> -		congestion_wait(BLK_RW_ASYNC, HZ/10);
> +	if (sc->may_writepage && !current_is_kswapd() &&
> +			(sc->lumpy_reclaim_mode || sc->mem_cgroup)) {
> +		int dirty_retry = MAX_SWAP_CLEAN_WAIT;
>  
> -		/*
> -		 * The attempt at page out may have made some
> -		 * of the pages active, mark them inactive again.
> -		 */
> -		nr_active = clear_active_flags(&page_list, NULL);
> -		count_vm_events(PGDEACTIVATE, nr_active);
> +		while (nr_reclaimed < nr_taken && nr_dirty && dirty_retry--) {
> +			wakeup_flusher_threads(laptop_mode ? 0 : nr_dirty);

Yup, minding laptop_mode (together with may_writepage).  Agreed.

> +			congestion_wait(BLK_RW_ASYNC, HZ/10);
>  
> -		nr_reclaimed += shrink_page_list(&page_list, sc, PAGEOUT_IO_SYNC);
> +			/*
> +			 * The attempt at page out may have made some
> +			 * of the pages active, mark them inactive again.
> +			 */
> +			nr_active = clear_active_flags(&page_list, NULL);
> +			count_vm_events(PGDEACTIVATE, nr_active);
> +	
> +			nr_reclaimed += shrink_page_list(&page_list, sc,
> +						PAGEOUT_IO_SYNC, &nr_dirty);
> +		}
>  	}
>  
>  	local_irq_disable();

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

Powered by blists - more mailing lists