linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Rik van Riel <riel@surriel.com>
From: Rik van Riel <riel@redhat.com>
To: Johannes Weiner <hannes@saeurebad.de>
Cc: linux-mm@kvack.org
Subject: [RFC][PATCH -mm] vmscan: fix swapout on sequential IO
Date: Wed, 23 Jul 2008 14:41:15 -0400	[thread overview]
Message-ID: <20080723144115.72803eb8@bree.surriel.com> (raw)

Only force the scanning of every LRU list if we could not easily
find a page to evict.  This preserves the balancing done by
get_scan_ratio(), while ensuring the VM will make progress
when there is serious memory pressure.

Signed-off-by: Rik van Riel <riel@redhat.com>
---
This should fix the "dd if=/dev/sda of=/dev/null causes swapout"
problem that has been seen with the new split LRU VM.

 mm/vmscan.c |   17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

Index: linux-2.6.26-rc8-mm1/mm/vmscan.c
===================================================================
--- linux-2.6.26-rc8-mm1.orig/mm/vmscan.c	2008-07-23 14:19:09.000000000 -0400
+++ linux-2.6.26-rc8-mm1/mm/vmscan.c	2008-07-23 14:36:26.000000000 -0400
@@ -1447,24 +1447,27 @@ static unsigned long shrink_zone(int pri
 	unsigned long nr_to_scan;
 	unsigned long nr_reclaimed = 0;
 	unsigned long percent[2];	/* anon @ 0; file @ 1 */
+	unsigned int force_scan = 0;
 	enum lru_list l;
 
+	/*
+         * If we do not immediately find pages to evict, put some
+         * pressure on every LRU list to guarantee progress.
+         */
+	if (priority < DEF_PRIORITY - 2)
+		force_scan = 1;
+
 	get_scan_ratio(zone, sc, percent);
 
 	for_each_evictable_lru(l) {
 		if (scan_global_lru(sc)) {
 			int file = is_file_lru(l);
-			int scan;
-			/*
-			 * Add one to nr_to_scan just to make sure that the
-			 * kernel will slowly sift through each list.
-			 */
-			scan = zone_page_state(zone, NR_LRU_BASE + l);
+			int scan = zone_page_state(zone, NR_LRU_BASE + l);
 			if (priority) {
 				scan >>= priority;
 				scan = (scan * percent[file]) / 100;
 			}
-			zone->lru[l].nr_scan += scan + 1;
+			zone->lru[l].nr_scan += scan + force_scan;
 			nr[l] = zone->lru[l].nr_scan;
 			if (nr[l] >= sc->swap_cluster_max)
 				zone->lru[l].nr_scan = 0;

--
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>

             reply	other threads:[~2008-07-23 18:41 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-07-23 18:41 Rik van Riel, Rik van Riel [this message]
2008-07-23 18:48 ` Johannes Weiner
2008-07-23 19:29   ` Rik van Riel
2008-07-23 19:38     ` Johannes Weiner
2008-07-24  6:26 ` KOSAKI Motohiro
2008-07-24 12:20   ` Rik van Riel

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=20080723144115.72803eb8@bree.surriel.com \
    --to=riel@surriel.com \
    --cc=hannes@saeurebad.de \
    --cc=linux-mm@kvack.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox