From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
To: Shaohua Li <shaohua.li@intel.com>
Cc: kosaki.motohiro@jp.fujitsu.com, Mel Gorman <mel@csn.ul.ie>,
Simon Kirby <sim@hostway.ca>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
Dave Hansen <dave@linux.vnet.ibm.com>
Subject: Re: Free memory never fully used, swapping
Date: Fri, 26 Nov 2010 11:31:42 +0900 (JST) [thread overview]
Message-ID: <20101126110244.B6DC.A69D9226@jp.fujitsu.com> (raw)
In-Reply-To: <1290736844.12777.10.camel@sli10-conroe>
> record the order seems not sufficient. in balance_pgdat(), the for look
> exit only when:
> priority <0 or sc.nr_reclaimed >= SWAP_CLUSTER_MAX.
> but we do if (sc.nr_reclaimed < SWAP_CLUSTER_MAX)
> order = sc.order = 0;
> this means before we set order to 0, we already reclaimed a lot of
> pages, so I thought we need set order to 0 earlier before there are
> enough free pages. below is a debug patch.
>
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index d31d7ce..ee5d2ed 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -2117,6 +2117,26 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
> }
> #endif
>
> +static int all_zone_enough_free_pages(pg_data_t *pgdat)
> +{
> + int i;
> +
> + for (i = 0; i < pgdat->nr_zones; i++) {
> + struct zone *zone = pgdat->node_zones + i;
> +
> + if (!populated_zone(zone))
> + continue;
> +
> + if (zone->all_unreclaimable)
> + continue;
> +
> + if (!zone_watermark_ok(zone, 0, high_wmark_pages(zone) * 8,
> + 0, 0))
> + return 0;
> + }
> + return 1;
> +}
> +
> /* is kswapd sleeping prematurely? */
> static int sleeping_prematurely(pg_data_t *pgdat, int order, long remaining)
> {
> @@ -2355,7 +2375,8 @@ out:
> * back to sleep. High-order users can still perform direct
> * reclaim if they wish.
> */
> - if (sc.nr_reclaimed < SWAP_CLUSTER_MAX)
> + if (sc.nr_reclaimed < SWAP_CLUSTER_MAX ||
> + (order > 0 && all_zone_enough_free_pages(pgdat)))
> order = sc.order = 0;
Ummm. this doesn't work. this place is processed every 32 pages reclaimed.
(see below code and comment). Theresore your patch break high order reclaim
logic.
/*
* We do this so kswapd doesn't build up large priorities for
* example when it is freeing in parallel with allocators. It
* matches the direct reclaim path behaviour in terms of impact
* on zone->*_priority.
*/
if (sc.nr_reclaimed >= SWAP_CLUSTER_MAX)
break;
next prev parent reply other threads:[~2010-11-26 2:31 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-15 19:52 Free memory never fully used, swapping Simon Kirby
2010-11-22 23:44 ` Andrew Morton
2010-11-23 1:34 ` Simon Kirby
2010-11-23 8:35 ` Dave Hansen
2010-11-24 8:46 ` Simon Kirby
2010-11-25 1:07 ` Shaohua Li
2010-11-25 9:03 ` Simon Kirby
2010-11-25 10:18 ` KOSAKI Motohiro
2010-11-25 17:13 ` Simon Kirby
2010-11-26 0:33 ` KOSAKI Motohiro
2010-11-25 10:51 ` KOSAKI Motohiro
2010-11-25 16:15 ` Mel Gorman
2010-11-26 2:00 ` Shaohua Li
2010-11-26 2:31 ` KOSAKI Motohiro [this message]
2010-11-26 2:40 ` Shaohua Li
2010-11-26 9:18 ` KOSAKI Motohiro
2010-11-29 1:03 ` Shaohua Li
2010-11-29 1:13 ` KOSAKI Motohiro
2010-11-26 0:07 ` KOSAKI Motohiro
2010-11-25 16:12 ` Mel Gorman
2010-11-26 1:05 ` Shaohua Li
2010-11-26 1:25 ` Mel Gorman
2010-11-26 2:05 ` Shaohua Li
2010-11-26 11:03 ` KOSAKI Motohiro
2010-11-26 11:11 ` Mel Gorman
2010-11-30 6:31 ` KOSAKI Motohiro
2010-11-30 10:41 ` Mel Gorman
2010-11-30 11:19 ` KOSAKI Motohiro
2010-11-30 8:22 ` Simon Kirby
2010-11-29 9:31 ` KOSAKI Motohiro
2010-11-23 10:04 ` Mel Gorman
2010-11-24 6:43 ` Simon Kirby
2010-11-24 9:27 ` Mel Gorman
2010-11-24 19:17 ` Simon Kirby
2010-11-25 1:18 ` KOSAKI Motohiro
2010-11-26 15:48 ` Christoph Lameter
2010-11-30 0:25 ` KOSAKI Motohiro
2010-11-30 19:10 ` Christoph Lameter
2010-12-01 10:17 ` KOSAKI Motohiro
2010-12-01 15:29 ` Christoph Lameter
2010-12-02 2:44 ` KOSAKI Motohiro
2010-12-02 14:39 ` Christoph Lameter
2010-11-30 9:13 ` Simon Kirby
2010-11-30 19:13 ` Christoph Lameter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20101126110244.B6DC.A69D9226@jp.fujitsu.com \
--to=kosaki.motohiro@jp.fujitsu.com \
--cc=dave@linux.vnet.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mel@csn.ul.ie \
--cc=shaohua.li@intel.com \
--cc=sim@hostway.ca \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).