GitHub-Mirror/Star64_linux
2
0
Fork 0
mirror of https://github.com/Fishwaldo/Star64_linux.git synced 2025-06-27 00:51:35 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

mm: fix zone_watermark_ok_safe() accounting of isolated pages

Browse source 
Commit 702d1a6e07 ("memory-hotplug: fix kswapd looping forever
problem") added an isolated pageblocks counter (nr_pageblock_isolate in
struct zone) and used it to adjust free pages counter in
zone_watermark_ok_safe() to prevent kswapd looping forever problem.

Then later, commit 2139cbe627 ("cma: fix counting of isolated pages")
fixed accounting of isolated pages in global free pages counter.  It
made the previous zone_watermark_ok_safe() fix unnecessary and
potentially harmful (cause now isolated pages may be accounted twice
making free pages counter incorrect).

This patch removes the special isolated pageblocks counter altogether
which fixes zone_watermark_ok_safe() free pages check.

Reported-by: Tomasz Stanislawski <t.stanislaws@samsung.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Aaditya Kumar <aaditya.kumar.30@gmail.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Michal Nazarewicz <mina86@mina86.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Bartlomiej Zolnierkiewicz 2013-01-04 15:35:08 -08:00 • committed by Linus Torvalds
parent 358e419f82
commit a458431e17
3 changed files with 2 additions and 59 deletions
Download patch file Download diff file Expand all files Collapse all files

27
mm/page_alloc.c
View file

@ -221,11 +221,6 @@ EXPORT_SYMBOL(nr_online_nodes);
int page_group_by_mobility_disabled __read_mostly;
/*
* NOTE:
* Don't use set_pageblock_migratetype(page, MIGRATE_ISOLATE) directly.
* Instead, use {un}set_pageblock_isolate.
*/
void set_pageblock_migratetype(struct page *page, int migratetype)
{
@ -1655,20 +1650,6 @@ static bool __zone_watermark_ok(struct zone *z, int order, unsigned long mark,
return true;
}
#ifdef CONFIG_MEMORY_ISOLATION
static inline unsigned long nr_zone_isolate_freepages(struct zone *zone)
{
if (unlikely(zone->nr_pageblock_isolate))
return zone->nr_pageblock_isolate * pageblock_nr_pages;
return 0;
}
#else
static inline unsigned long nr_zone_isolate_freepages(struct zone *zone)
{
return 0;
}
#endif
bool zone_watermark_ok(struct zone *z, int order, unsigned long mark,
int classzone_idx, int alloc_flags)
{
@ -1684,14 +1665,6 @@ bool zone_watermark_ok_safe(struct zone *z, int order, unsigned long mark,
if (z->percpu_drift_mark && free_pages < z->percpu_drift_mark)
free_pages = zone_page_state_snapshot(z, NR_FREE_PAGES);
/*
* If the zone has MIGRATE_ISOLATE type free pages, we should consider
* it. nr_zone_isolate_freepages is never accurate so kswapd might not
* sleep although it could do so. But this is more desirable for memory
* hotplug than sleeping which can cause a livelock in the direct
* reclaim path.
*/
free_pages -= nr_zone_isolate_freepages(z);
return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags,
free_pages);
}