mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Christoph Lameter <cl@linux-foundation.org>
To: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: akpm@linux-foundation.org, Christoph Lameter <clameter@sgi.com>,
	Christoph Lameter <cl@linux-foundation.org>
Cc: linux-kernel@vger.kernel.org
Cc: linux-fsdevel@vger.kernel.org
Cc: Mel Gorman <mel@skynet.ie>
Cc: andi@firstfloor.org
Cc: Rik van Riel <riel@redhat.com>
Cc: mpm@selenic.com
Cc: Dave Chinner <david@fromorbit.com>
Subject: [patch 06/19] slub: Add KICKABLE to avoid repeated kick() attempts
Date: Mon, 11 Aug 2008 08:06:22 -0700	[thread overview]
Message-ID: <20080811150656.117393595@quilx.com> (raw)
In-Reply-To: <20080811150616.207532384@quilx.com>

[-- Attachment #1: 0006-SLUB-Add-KICKABLE-to-avoid-repeated-kick-attempts.patch --]
[-- Type: text/plain, Size: 3839 bytes --]

Add a flag KICKABLE to be set on slabs with a defragmentation method

Clear the flag if a kick action is not successful in reducing the
number of objects in a slab. This will avoid future attempts to
kick objects out.

The KICKABLE flag is set again when all objects of the slab have been
allocated (Occurs during removal of a slab from the partial lists).

Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <cl@linux-foundation.org>

---
 include/linux/page-flags.h |    2 ++
 mm/slub.c                  |   23 ++++++++++++++++++-----
 2 files changed, 20 insertions(+), 5 deletions(-)

Index: linux-next/mm/slub.c
===================================================================
--- linux-next.orig/mm/slub.c	2008-08-11 07:47:01.482348822 -0700
+++ linux-next/mm/slub.c	2008-08-11 07:50:09.282347574 -0700
@@ -1138,6 +1138,9 @@ static struct page *new_slab(struct kmem
 			SLAB_STORE_USER | SLAB_TRACE))
 		__SetPageSlubDebug(page);
 
+	if (s->kick)
+		__SetPageSlubKickable(page);
+
 	start = page_address(page);
 
 	if (unlikely(s->flags & SLAB_POISON))
@@ -1181,6 +1184,7 @@ static void __free_slab(struct kmem_cach
 		NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE,
 		-pages);
 
+	__ClearPageSlubKickable(page);
 	__ClearPageSlab(page);
 	reset_page_mapcount(page);
 	__free_pages(page, order);
@@ -1391,6 +1395,8 @@ static void unfreeze_slab(struct kmem_ca
 			if (SLABDEBUG && PageSlubDebug(page) &&
 						(s->flags & SLAB_STORE_USER))
 				add_full(n, page);
+			if (s->kick)
+				__SetPageSlubKickable(page);
 		}
 		slab_unlock(page);
 	} else {
@@ -2820,12 +2826,12 @@ static int kmem_cache_vacate(struct page
 	slab_lock(page);
 
 	BUG_ON(!PageSlab(page));	/* Must be s slab page */
-	BUG_ON(!SlabFrozen(page));	/* Slab must have been frozen earlier */
+	BUG_ON(!PageSlubFrozen(page));	/* Slab must have been frozen earlier */
 
 	s = page->slab;
 	objects = page->objects;
 	map = scratch + objects * sizeof(void **);
-	if (!page->inuse || !s->kick)
+	if (!page->inuse || !s->kick || !PageSlubKickable(page))
 		goto out;
 
 	/* Determine used objects */
@@ -2863,6 +2869,9 @@ out:
 	 * Check the result and unfreeze the slab
 	 */
 	leftover = page->inuse;
+	if (leftover)
+		/* Unsuccessful reclaim. Avoid future reclaim attempts. */
+		__ClearPageSlubKickable(page);
 	unfreeze_slab(s, page, leftover > 0);
 	local_irq_restore(flags);
 	return leftover;
@@ -2924,17 +2933,21 @@ static unsigned long __kmem_cache_shrink
 			continue;
 
 		if (page->inuse) {
-			if (page->inuse * 100 >=
+			if (!PageSlubKickable(page) || page->inuse * 100 >=
 					s->defrag_ratio * page->objects) {
 				slab_unlock(page);
-				/* Slab contains enough objects */
+				/*
+				 * Slab contains enough objects
+				 * or we alrady tried reclaim before and
+				 * it failed. Skip this one.
+				 */
 				continue;
 			}
 
 			list_move(&page->lru, &zaplist);
 			if (s->kick) {
 				n->nr_partial--;
-				SetSlabFrozen(page);
+				__SetPageSlubFrozen(page);
 			}
 			slab_unlock(page);
 		} else {
Index: linux-next/include/linux/page-flags.h
===================================================================
--- linux-next.orig/include/linux/page-flags.h	2008-08-11 07:42:31.882358716 -0700
+++ linux-next/include/linux/page-flags.h	2008-08-11 07:47:01.612357581 -0700
@@ -112,6 +112,7 @@ enum pageflags {
 	/* SLUB */
 	PG_slub_frozen = PG_active,
 	PG_slub_debug = PG_error,
+	PG_slub_kickable = PG_dirty,
 };
 
 #ifndef __GENERATING_BOUNDS_H
@@ -182,6 +183,7 @@ __PAGEFLAG(SlobFree, slob_free)
 
 __PAGEFLAG(SlubFrozen, slub_frozen)
 __PAGEFLAG(SlubDebug, slub_debug)
+__PAGEFLAG(SlubKickable, slub_kickable)
 
 /*
  * Only test-and-set exist for PG_writeback.  The unconditional operators are

-- 

  parent reply	other threads:[~2008-08-11 15:16 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-08-11 15:06 [patch 00/19] Slab Fragmentation Reduction V14 Christoph Lameter
2008-08-11 15:06 ` [patch 01/19] slub: Add defrag_ratio field and sysfs support Christoph Lameter
2008-08-13  0:40   ` Greg KH
2008-08-11 15:06 ` [patch 02/19] slub: Replace ctor field with ops field in /sys/slab/* Christoph Lameter
2008-08-11 15:06 ` [patch 03/19] slub: Add get() and kick() methods Christoph Lameter
2008-08-11 15:06 ` [patch 04/19] slub: Sort slab cache list and establish maximum objects for defrag slabs Christoph Lameter
2008-08-11 15:06 ` [patch 05/19] slub: Slab defrag core Christoph Lameter
2008-08-11 15:06 ` Christoph Lameter [this message]
2008-08-11 15:06 ` [patch 07/19] slub: Extend slabinfo to support -D and -F options Christoph Lameter
2008-08-11 15:06 ` [patch 08/19] slub/slabinfo: add defrag statistics Christoph Lameter
2008-08-11 15:06 ` [patch 09/19] slub: Trigger defragmentation from memory reclaim Christoph Lameter
2008-08-11 15:06 ` [patch 10/19] buffer heads: Support slab defrag Christoph Lameter
2008-08-11 15:06 ` [patch 11/19] inodes: Support generic defragmentation Christoph Lameter
2008-08-11 15:06 ` [patch 12/19] Filesystem: Ext2 filesystem defrag Christoph Lameter
2008-08-11 15:06 ` [patch 13/19] Filesystem: Ext3 " Christoph Lameter
2008-08-11 15:06 ` [patch 14/19] Filesystem: Ext4 " Christoph Lameter
2008-08-11 15:06 ` [patch 15/19] Filesystem: XFS slab defragmentation Christoph Lameter
2008-08-12  0:20   ` Dave Chinner
2008-08-12  7:08     ` Pekka Enberg
2008-08-12 18:21       ` Christoph Lameter
2008-08-13  6:38         ` Pekka Enberg
2008-08-11 15:06 ` [patch 16/19] Filesystem: /proc filesystem support for slab defrag Christoph Lameter
2008-08-11 15:06 ` [patch 17/19] Filesystem: Slab defrag: Reiserfs support Christoph Lameter
2008-08-11 15:06 ` [patch 18/19] dentries: Add constructor Christoph Lameter
2008-08-13  7:12   ` Pekka Enberg
2008-08-11 15:06 ` [patch 19/19] dentries: dentry defragmentation Christoph Lameter
2008-08-12  6:50 ` [patch 00/19] Slab Fragmentation Reduction V14 Pekka Enberg
  -- strict thread matches above, loose matches on Subject: below --
2008-05-10  2:21 [patch 00/19] Slab Fragmentation Reduction V13 Christoph Lameter
2008-05-10  2:21 ` [patch 06/19] slub: Add KICKABLE to avoid repeated kick() attempts 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=20080811150656.117393595@quilx.com \
    --to=cl@linux-foundation.org \
    --cc=akpm@linux-foundation.org \
    --cc=clameter@sgi.com \
    --cc=penberg@cs.helsinki.fi \
    /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

all inboxes | Powered by JetHome®