Messages in this thread Patch in this message | | | Date | Thu, 14 May 2009 16:40:10 +0900 | From | Minchan Kim <> | Subject | Re: [PATCH] Prevent shrinking of active anon lru list in case of no swap space |
| |
On Thu, 14 May 2009 16:19:52 +0900 (JST) KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> wrote:
> > > > Now shrink_active_list is called several places. > > But if we don't have a swap space, we can't reclaim anon pages. > > So, we don't need deactivating anon pages in anon lru list. > > > > Signed-off-by: Minchan Kim <minchan.kim@gmail.com> > > Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> > > Cc: Johannes Weiner <hannes@cmpxchg.org> > > Cc: Rik van Riel <riel@redhat.com> > > hm, The analysis seems right. but > > In general branch should put into non frequency path. > why caller modification as following patch is wrong?
Yes. You're right. Firstly I though that. But there are a few place where inactve_anon_is_low is called. In addition to it, we always have to call with (nr_swap_pages <= 0) with it. So alone inactive_anon_is_low is lost function's meaning.
I think it's meaningless. But We can improve one function wrapper.
/* I can't be remind proper function name. Can you suggest me ? */ if (do_we_have_to_shrink_list(zone, &sc))
int do_we_have_to_shrink_list(...) { if (inactive_anon_is_low(zone, &sc) && (nr_swap_pages <= 0)) }
What do you think ?
> > /* > * Do some background aging of the anon list, to give > * pages a chance to be referenced before reclaiming. > */ > - if (inactive_anon_is_low(zone, &sc)) > + if (inactive_anon_is_low(zone, &sc) && (nr_swap_pages <= 0)) > shrink_active_list(SWAP_CLUSTER_MAX, zone, > &sc, priority, 0); > > > > > > --- > > mm/vmscan.c | 6 ++++++ > > 1 files changed, 6 insertions(+), 0 deletions(-) > > > > diff --git a/mm/vmscan.c b/mm/vmscan.c > > index 2f9d555..e4d71f4 100644 > > --- a/mm/vmscan.c > > +++ b/mm/vmscan.c > > @@ -1238,6 +1238,12 @@ static void shrink_active_list(unsigned long nr_pages, struct zone *zone, > > enum lru_list lru; > > struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); > > > > + /* > > + * we can't shrink anon list in case of no swap space. > > + */ > > + if (file == 0 && nr_swap_pages <= 0) > > + return; > > + > > > > lru_add_drain(); > > spin_lock_irq(&zone->lru_lock); > > pgmoved = sc->isolate_pages(nr_pages, &l_hold, &pgscanned, sc->order, > > -- > > 1.5.4.3 > > > > > > -- > > Kinds Regards > > Minchan Kim > > >
-- Kinds Regards Minchan Kim
| |