Re: [PATCH] vmscan: evict use-once pages first (v3)

[Rik van Riel]

Andrew Morton wrote:
> On Wed, 29 Apr 2009 13:14:36 -0400
> Rik van Riel <riel@xxxxxxxxxx> wrote:
>
> > When the file LRU lists are dominated by streaming IO pages,
> > evict those pages first, before considering evicting other
> > pages.
> >
> > This should be safe from deadlocks or performance problems
> > because only three things can happen to an inactive file page:
> > 1) referenced twice and promoted to the active list
> > 2) evicted by the pageout code
> > 3) under IO, after which it will get evicted or promoted
> >
> > The pages freed in this way can either be reused for streaming
> > IO, or allocated for something else. If the pages are used for
> > streaming IO, this pageout pattern continues. Otherwise, we will
> > fall back to the normal pageout pattern.
> >
> > ..
> >
> > +int mem_cgroup_inactive_file_is_low(struct mem_cgroup *memcg)
> > +{
> > +	unsigned long active;
> > +	unsigned long inactive;
> > +
> > +	inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_FILE);
> > +	active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_FILE);
> > +
> > +	return (active > inactive);
> > +}
>
> This function could trivially be made significantly more efficient by
> changing it to do a single pass over all the zones of all the nodes,
> rather than two passes.

How would I do that in a clean way?

The function mem_cgroup_inactive_anon_is_low and
the global versions all do the same.  It would be
nice to make all four of them go fast :)

If there is no standardized infrastructure for
getting multiple statistics yet, I can probably
whip something up.

> >  static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
> >  	struct zone *zone, struct scan_control *sc, int priority)
> >  {
> >  	int file = is_file_lru(lru);
> >  
> > -	if (lru == LRU_ACTIVE_FILE) {
> > +	if (lru == LRU_ACTIVE_FILE && inactive_file_is_low(zone, sc)) {
> >  		shrink_active_list(nr_to_scan, zone, sc, priority, file);
> >  		return 0;
> >  	}
>
> And it does get called rather often.

Same as inactive_anon_is_low.

Optimizing them might make sense if it turns out to
use a significant amount of CPU.

--
All rights reversed.
--
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/