From: Andrew Morton <akpm@linux-foundation.org>
To: Rik van Riel <riel@redhat.com>
Cc: torvalds@linux-foundation.org, kamezawa.hiroyu@jp.fujitsu.com,
kosaki.motohiro@jp.fujitsu.com, linux-kernel@vger.kernel.org,
linux-mm@kvack.org, gene.heskett@gmail.com
Subject: Re: [PATCH] vmscan: evict streaming IO first
Date: Mon, 1 Dec 2008 12:24:38 -0800 [thread overview]
Message-ID: <20081201122438.16828a87.akpm@linux-foundation.org> (raw)
In-Reply-To: <20081117190642.3aabd3ff@bree.surriel.com>
On Mon, 17 Nov 2008 19:06:42 -0500
Rik van Riel <riel@redhat.com> wrote:
> Count the insertion of new pages in the statistics used to drive the
> pageout scanning code. This should help the kernel quickly evict
> streaming file IO.
>
> We count on the fact that new file pages start on the inactive file
> LRU and new anonymous pages start on the active anon list. This
> means streaming file IO will increment the recent scanned file
> statistic, while leaving the recent rotated file statistic alone,
> driving pageout scanning to the file LRUs.
>
> Pageout activity does its own list manipulation.
>
> Signed-off-by: Rik van Riel <riel@redhat.com>
> ---
> mm/swap.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> On Mon, 17 Nov 2008 08:22:13 -0800 (PST)
> Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> > .. or how about just considering the act of adding a new page to the LRU
> > to be a "scan" event? IOW, "scanning" is not necessarily just an act of
> > the VM looking for pages to free, but would be a more general "activity"
> > meter.
>
> Linus, this should implement your idea.
>
> Gene, does this patch resolve the problem for you?
Has Gene had a chance to confirm this yet?
> Index: linux-2.6.28-rc5/mm/swap.c
> ===================================================================
> --- linux-2.6.28-rc5.orig/mm/swap.c 2008-11-16 17:47:13.000000000 -0500
> +++ linux-2.6.28-rc5/mm/swap.c 2008-11-17 18:58:32.000000000 -0500
> @@ -445,6 +445,7 @@ void ____pagevec_lru_add(struct pagevec
> for (i = 0; i < pagevec_count(pvec); i++) {
> struct page *page = pvec->pages[i];
> struct zone *pagezone = page_zone(page);
> + int file;
>
> if (pagezone != zone) {
> if (zone)
> @@ -456,8 +457,12 @@ void ____pagevec_lru_add(struct pagevec
> VM_BUG_ON(PageUnevictable(page));
> VM_BUG_ON(PageLRU(page));
> SetPageLRU(page);
> - if (is_active_lru(lru))
> + file = is_file_lru(lru);
> + zone->recent_scanned[file]++;
> + if (is_active_lru(lru)) {
> SetPageActive(page);
> + zone->recent_rotated[file]++;
> + }
> add_page_to_lru_list(zone, page, lru);
> }
> if (zone)
Were you not able to reproduce the problem? It looks like it'd be a
pretty simple test case to set up?
WARNING: multiple messages have this Message-ID (diff)
From: Andrew Morton <akpm@linux-foundation.org>
To: Rik van Riel <riel@redhat.com>
Cc: torvalds@linux-foundation.org, kamezawa.hiroyu@jp.fujitsu.com,
kosaki.motohiro@jp.fujitsu.com, linux-kernel@vger.kernel.org,
linux-mm@kvack.org, gene.heskett@gmail.com
Subject: Re: [PATCH] vmscan: evict streaming IO first
Date: Mon, 1 Dec 2008 12:24:38 -0800 [thread overview]
Message-ID: <20081201122438.16828a87.akpm@linux-foundation.org> (raw)
In-Reply-To: <20081117190642.3aabd3ff@bree.surriel.com>
On Mon, 17 Nov 2008 19:06:42 -0500
Rik van Riel <riel@redhat.com> wrote:
> Count the insertion of new pages in the statistics used to drive the
> pageout scanning code. This should help the kernel quickly evict
> streaming file IO.
>
> We count on the fact that new file pages start on the inactive file
> LRU and new anonymous pages start on the active anon list. This
> means streaming file IO will increment the recent scanned file
> statistic, while leaving the recent rotated file statistic alone,
> driving pageout scanning to the file LRUs.
>
> Pageout activity does its own list manipulation.
>
> Signed-off-by: Rik van Riel <riel@redhat.com>
> ---
> mm/swap.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> On Mon, 17 Nov 2008 08:22:13 -0800 (PST)
> Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> > .. or how about just considering the act of adding a new page to the LRU
> > to be a "scan" event? IOW, "scanning" is not necessarily just an act of
> > the VM looking for pages to free, but would be a more general "activity"
> > meter.
>
> Linus, this should implement your idea.
>
> Gene, does this patch resolve the problem for you?
Has Gene had a chance to confirm this yet?
> Index: linux-2.6.28-rc5/mm/swap.c
> ===================================================================
> --- linux-2.6.28-rc5.orig/mm/swap.c 2008-11-16 17:47:13.000000000 -0500
> +++ linux-2.6.28-rc5/mm/swap.c 2008-11-17 18:58:32.000000000 -0500
> @@ -445,6 +445,7 @@ void ____pagevec_lru_add(struct pagevec
> for (i = 0; i < pagevec_count(pvec); i++) {
> struct page *page = pvec->pages[i];
> struct zone *pagezone = page_zone(page);
> + int file;
>
> if (pagezone != zone) {
> if (zone)
> @@ -456,8 +457,12 @@ void ____pagevec_lru_add(struct pagevec
> VM_BUG_ON(PageUnevictable(page));
> VM_BUG_ON(PageLRU(page));
> SetPageLRU(page);
> - if (is_active_lru(lru))
> + file = is_file_lru(lru);
> + zone->recent_scanned[file]++;
> + if (is_active_lru(lru)) {
> SetPageActive(page);
> + zone->recent_rotated[file]++;
> + }
> add_page_to_lru_list(zone, page, lru);
> }
> if (zone)
Were you not able to reproduce the problem? It looks like it'd be a
pretty simple test case to set up?
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2008-12-01 20:25 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-11-15 9:38 [PATCH] mm: evict streaming IO cache first KOSAKI Motohiro
2008-11-15 9:38 ` KOSAKI Motohiro
2008-11-16 0:58 ` Rik van Riel
2008-11-16 0:58 ` Rik van Riel
2008-11-16 5:00 ` Andrew Morton
2008-11-16 5:00 ` Andrew Morton
2008-11-16 18:15 ` Linus Torvalds
2008-11-16 18:15 ` Linus Torvalds
2008-11-16 21:20 ` Rik van Riel
2008-11-16 21:20 ` Rik van Riel
2008-11-16 21:28 ` Hugh Dickins
2008-11-16 21:28 ` Hugh Dickins
2008-11-17 4:47 ` Andrew Morton
2008-11-17 4:47 ` Andrew Morton
2008-11-17 6:19 ` Nick Piggin
2008-11-17 6:19 ` Nick Piggin
2008-11-17 6:30 ` KAMEZAWA Hiroyuki
2008-11-17 6:30 ` KAMEZAWA Hiroyuki
2008-11-17 6:39 ` KOSAKI Motohiro
2008-11-17 6:39 ` KOSAKI Motohiro
2008-11-17 6:54 ` KAMEZAWA Hiroyuki
2008-11-17 6:54 ` KAMEZAWA Hiroyuki
2008-11-17 7:03 ` KOSAKI Motohiro
2008-11-17 7:03 ` KOSAKI Motohiro
2008-11-17 8:22 ` KAMEZAWA Hiroyuki
2008-11-17 8:22 ` KAMEZAWA Hiroyuki
2008-11-17 8:32 ` KOSAKI Motohiro
2008-11-17 8:32 ` KOSAKI Motohiro
2008-11-17 16:22 ` Rik van Riel
2008-11-17 16:22 ` Rik van Riel
2008-11-17 16:22 ` Linus Torvalds
2008-11-17 16:22 ` Linus Torvalds
2008-11-17 16:27 ` Rik van Riel
2008-11-17 16:27 ` Rik van Riel
2008-11-17 16:37 ` Linus Torvalds
2008-11-17 16:37 ` Linus Torvalds
2008-11-17 16:54 ` Rik van Riel
2008-11-17 16:54 ` Rik van Riel
2008-11-17 17:06 ` Linus Torvalds
2008-11-17 17:06 ` Linus Torvalds
2008-11-17 17:16 ` [PATCH] vmscan: fix get_scan_ratio comment Rik van Riel
2008-11-17 17:32 ` KOSAKI Motohiro
2008-11-17 17:32 ` KOSAKI Motohiro
2008-11-17 17:35 ` Linus Torvalds
2008-11-17 17:35 ` Linus Torvalds
2008-11-17 18:53 ` Rik van Riel
2008-11-17 18:53 ` Rik van Riel
2008-11-17 20:23 ` Rik van Riel
2008-11-17 20:23 ` Rik van Riel
2008-11-18 0:06 ` [PATCH] vmscan: evict streaming IO first Rik van Riel
2008-11-18 0:06 ` Rik van Riel
2008-12-01 20:24 ` Andrew Morton [this message]
2008-12-01 20:24 ` Andrew Morton
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=20081201122438.16828a87.akpm@linux-foundation.org \
--to=akpm@linux-foundation.org \
--cc=gene.heskett@gmail.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=riel@redhat.com \
--cc=torvalds@linux-foundation.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.