Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate

From	Mel Gorman <mgorman@techsingularity.net>
Newsgroups	linux.kernel
Subject	Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate
Date	2016-07-21 10:20 +0200
Message-ID	<rXhn4-4Uk-19@gated-at.bofh.it> (permalink)
References	<rX1BD-39C-7@gated-at.bofh.it> <rX1BD-39C-5@gated-at.bofh.it> <rXghj-4gp-13@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

On Thu, Jul 21, 2016 at 04:10:50PM +0900, Joonsoo Kim wrote:
> > @@ -1993,6 +1994,32 @@ static bool inactive_list_is_low(struct lruvec *lruvec, bool file)
> >  	inactive = lruvec_lru_size(lruvec, file * LRU_FILE);
> >  	active = lruvec_lru_size(lruvec, file * LRU_FILE + LRU_ACTIVE);
> >  
> > +	/*
> > +	 * For global reclaim on zone-constrained allocations, it is necessary
> > +	 * to check if rotations are required for lowmem to be reclaimed. This
> > +	 * calculates the inactive/active pages available in eligible zones.
> > +	 */
> > +	if (global_reclaim(sc)) {
> > +		struct pglist_data *pgdat = lruvec_pgdat(lruvec);
> > +		int zid;
> > +
> > +		for (zid = sc->reclaim_idx; zid < MAX_NR_ZONES; zid++) {
> 
> Should be changed to "zid = sc->reclaim_idx + 1"
> 

You're right, well spotted!

-- 
Mel Gorman
SUSE Labs

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

[PATCH 5/5] mm: consider per-zone inactive ratio to deactivate Mel Gorman <mgorman@techsingularity.net> - 2016-07-20 17:30 +0200
  Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate Minchan Kim <minchan@kernel.org> - 2016-07-21 07:40 +0200
    Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate Mel Gorman <mgorman@techsingularity.net> - 2016-07-21 10:10 +0200
  Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate Joonsoo Kim <iamjoonsoo.kim@lge.com> - 2016-07-21 09:10 +0200
    Re: [PATCH 5/5] mm: consider per-zone inactive ratio to deactivate Mel Gorman <mgorman@techsingularity.net> - 2016-07-21 10:20 +0200

csiph-web