Re: [PATCH][mmotm] don't attempt to reclaim anon page in lumpy reclaim when no swap space is available

[KOSAKI Motohiro]

2009/7/4 Minchan Kim <minchan.kim@xxxxxxxxx>:
>
> This patch is based on mmotm 2009-07-02-19-57 reverted
> 'vmscan: don't attempt to reclaim anon page in lumpy reclaim when no swap space is available.'
>
> This verssion is better than old one.
> That's because enough swap space check is done in case of only lumpy reclaim.
> so it can't degrade performance in normal case.
>
> == CUT HERE ==
>
> VM already avoids attempting to reclaim anon pages in various places, But
> it doesn't avoid it for lumpy reclaim.
>
> It shuffles lru list unnecessary so that it is pointless.
>
> Signed-off-by: Minchan Kim <minchan.kim@xxxxxxxxx>
> Cc: Mel Gorman <mel@xxxxxxxxx>
> Cc: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
> Cc: Rik van Riel <riel@xxxxxxxxxx>
> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
> ---
>  mm/vmscan.c |    7 +++++++
>  1 files changed, 7 insertions(+), 0 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 27558aa..977af15 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -930,6 +930,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>                        /* Check that we have not crossed a zone boundary. */
>                        if (unlikely(page_zone_id(cursor_page) != zone_id))
>                                continue;
> +                       /*
> +                        * If we don't have enough swap space, reclaiming of anon page
> +                        * which don't already have a swap slot is pointless.
> +                        */
> +                       if (nr_swap_pages <= 0 && (PageAnon(cursor_page) &&
> +                                               !PageSwapCache(cursor_page)))
> +                               continue;
>                        if (__isolate_lru_page(cursor_page, mode, file) == 0) {
>                                list_move(&cursor_page->lru, dst);
>                                mem_cgroup_del_lru(cursor_page);

okey. this is definitely better. thanks.
    Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
--
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/