On Thu, Jul 12, 2012 at 02:01:54PM -0700, Andrew Morton wrote: > On Thu, 12 Jul 2012 11:50:49 +0900 > Minchan Kim <minchan@xxxxxxxxxx> wrote: > > > When hotplug offlining happens on zone A, it starts to mark freed page > > as MIGRATE_ISOLATE type in buddy for preventing further allocation. > > (MIGRATE_ISOLATE is very irony type because it's apparently on buddy > > but we can't allocate them). > > When the memory shortage happens during hotplug offlining, > > current task starts to reclaim, then wake up kswapd. > > Kswapd checks watermark, then go sleep because current zone_watermark_ok_safe > > doesn't consider MIGRATE_ISOLATE freed page count. > > Current task continue to reclaim in direct reclaim path without kswapd's helping. > > The problem is that zone->all_unreclaimable is set by only kswapd > > so that current task would be looping forever like below. > > > > __alloc_pages_slowpath > > restart: > > wake_all_kswapd > > rebalance: > > __alloc_pages_direct_reclaim > > do_try_to_free_pages > > if global_reclaim && !all_unreclaimable > > return 1; /* It means we did did_some_progress */ > > skip __alloc_pages_may_oom > > should_alloc_retry > > goto rebalance; > > > > If we apply KOSAKI's patch[1] which doesn't depends on kswapd > > about setting zone->all_unreclaimable, we can solve this problem > > by killing some task in direct reclaim path. But it doesn't wake up kswapd, still. > > It could be a problem still if other subsystem needs GFP_ATOMIC request. > > So kswapd should consider MIGRATE_ISOLATE when it calculate free pages > > BEFORE going sleep. > > > > This patch counts the number of MIGRATE_ISOLATE page block and > > zone_watermark_ok_safe will consider it if the system has such blocks > > (fortunately, it's very rare so no problem in POV overhead and kswapd is never > > hotpath). > > > > Copy/modify from Mel's quote > > " > > Ideal solution would be "allocating" the pageblock. > > It would keep the free space accounting as it is but historically, > > memory hotplug didn't allocate pages because it would be difficult to > > detect if a pageblock was isolated or if part of some balloon. > > Allocating just full pageblocks would work around this, However, > > it would play very badly with CMA. > > " > > > > [1] http://lkml.org/lkml/2012/6/14/74 > > > > ... > > > > +#ifdef CONFIG_MEMORY_ISOLATION > > +static inline unsigned long nr_zone_isolate_freepages(struct zone *zone) > > +{ > > + unsigned long nr_pages = 0; > > + > > + if (unlikely(zone->nr_pageblock_isolate)) { > > + nr_pages = zone->nr_pageblock_isolate * pageblock_nr_pages; > > + } > > + return nr_pages; > > +} > > That's pretty verbose. I couldn't resist fiddling with it ;) > > > +#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) > > { > > @@ -1633,6 +1655,14 @@ 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 page, > > + * we should consider it. nr_zone_isolate_freepages is never > > + * accurate so kswapd might not sleep although she can. > > + * But it's more desirable for memory hotplug rather than > > + * forever sleep which cause livelock in direct reclaim path. > > + */ > > I had a go at clarifying this comment. Please review the result. > > > + free_pages -= nr_zone_isolate_freepages(z); > > return __zone_watermark_ok(z, order, mark, classzone_idx, alloc_flags, > > free_pages); > > } > > @@ -4397,6 +4427,7 @@ static void __paginginit free_area_init_core(struct pglist_data *pgdat, > > lruvec_init(&zone->lruvec, zone); > > zap_zone_vm_stats(zone); > > zone->flags = 0; > > + zone->nr_pageblock_isolate = 0; > > if (!size) > > continue; > > > > diff --git a/mm/page_isolation.c b/mm/page_isolation.c > > index fb482cf..64abb33 100644 > > --- a/mm/page_isolation.c > > +++ b/mm/page_isolation.c > > @@ -8,6 +8,31 @@ > > #include <linux/memory.h> > > #include "internal.h" > > > > +/* called by holding zone->lock */ > > +static void set_pageblock_isolate(struct zone *zone, struct page *page) > > +{ > > + BUG_ON(page_zone(page) != zone); > > Well. If this is the case then why not eliminate this function's > `zone' argument? > > > + if (get_pageblock_migratetype(page) == MIGRATE_ISOLATE) > > + return; > > + > > + set_pageblock_migratetype(page, MIGRATE_ISOLATE); > > + zone->nr_pageblock_isolate++; > > +} > > + > > +/* called by holding zone->lock */ > > +static void restore_pageblock_isolate(struct zone *zone, struct page *page, > > + int migratetype) > > +{ > > + BUG_ON(page_zone(page) != zone); > > ditto > > > + if (WARN_ON(get_pageblock_migratetype(page) != MIGRATE_ISOLATE)) > > + return; > > + > > + BUG_ON(zone->nr_pageblock_isolate <= 0); > > + set_pageblock_migratetype(page, migratetype); > > + zone->nr_pageblock_isolate--; > > +} > > From: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> > Subject: memory-hotplug-fix-kswapd-looping-forever-problem-fix > > simplify nr_zone_isolate_freepages(), rework zone_watermark_ok_safe() comment, simplify set_pageblock_isolate() and restore_pageblock_isolate(). > > Cc: Aaditya Kumar <aaditya.kumar.30@xxxxxxxxx> > Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> > Cc: Mel Gorman <mgorman@xxxxxxx> > Cc: Michal Hocko <mhocko@xxxxxxx> > Cc: Minchan Kim <minchan@xxxxxxxxxx> > Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> Acked-by: Minchan Kim <minchan@xxxxxxxxxx> Looks better. Thanks, Andrew! -- Kind regards, Minchan Kim -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>