lkml.org 
[lkml]   [2010]   [Jul]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
SubjectRe: [PATCH v2 2/2] vmscan: shrink_slab() require number of lru_pages, not page order
Date
> >  	nr_slab_pages0 = zone_page_state(zone, NR_SLAB_RECLAIMABLE);
> > if (nr_slab_pages0 > zone->min_slab_pages) {
> > + unsigned long lru_pages = zone_reclaimable_pages(zone);
> > +
> > /*
> > * shrink_slab() does not currently allow us to determine how
> > * many pages were freed in this zone. So we take the current
> > @@ -2622,7 +2624,7 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
> > * Note that shrink_slab will free memory on all zones and may
> > * take a long time.
> > */
> > - while (shrink_slab(sc.nr_scanned, gfp_mask, order) &&
> > + while (shrink_slab(sc.nr_scanned, gfp_mask, lru_pages) &&
> > (zone_page_state(zone, NR_SLAB_RECLAIMABLE) + nr_pages >
> > nr_slab_pages0))
> > ;
>
> Wouldn't it be better to recalculate zone_reclaimable_pages() each time
> around the loop? For example, shrink_icache_memory()->prune_icache()
> will remove pagecache from an inode if it hits the tail of the list.
> This can change the number of reclaimable pages by squigabytes, but
> this code thinks nothing changed?

Ah, I missed this. incrementa patch is here.

thank you!



From 8f7c70cfb4a25f8292a59564db6c3ff425a69b53 Mon Sep 17 00:00:00 2001
From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Date: Fri, 16 Jul 2010 08:40:01 +0900
Subject: [PATCH] vmscan: recalculate lru_pages on each shrink_slab()

Andrew Morton pointed out shrink_slab() may change number of reclaimable
pages (e.g. shrink_icache_memory()->prune_icache() will remove unmapped
pagecache).

So, we need to recalculate lru_pages on each shrink_slab() calling.
This patch fixes it.

Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
---
mm/vmscan.c | 18 ++++++++++++------
1 files changed, 12 insertions(+), 6 deletions(-)
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 1bf9f72..1da9b14 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2612,8 +2612,6 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)

nr_slab_pages0 = zone_page_state(zone, NR_SLAB_RECLAIMABLE);
if (nr_slab_pages0 > zone->min_slab_pages) {
- unsigned long lru_pages = zone_reclaimable_pages(zone);
-
/*
* shrink_slab() does not currently allow us to determine how
* many pages were freed in this zone. So we take the current
@@ -2624,10 +2622,18 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
* Note that shrink_slab will free memory on all zones and may
* take a long time.
*/
- while (shrink_slab(sc.nr_scanned, gfp_mask, lru_pages) &&
- (zone_page_state(zone, NR_SLAB_RECLAIMABLE) + nr_pages >
- nr_slab_pages0))
- ;
+ for (;;) {
+ unsigned long lru_pages = zone_reclaimable_pages(zone);
+
+ /* No reclaimable slab or very low memroy pressure */
+ if (!shrink_slab(sc.nr_scanned, gfp_mask, lru_pages))
+ break;
+
+ /* Freed enouch memory */
+ nr_slab_pages1 = zone_page_state(zone, NR_SLAB_RECLAIMABLE);
+ if (nr_slab_pages1 + nr_pages <= nr_slab_pages0)
+ break;
+ }

/*
* Update nr_reclaimed by the number of slab pages we
--
1.6.5.2




\
 
 \ /
  Last update: 2010-07-16 03:43    [from the cache]
©2003-2014 Jasper Spaans. Advertise on this site