mm, vmscan: move LRU lists to node
authorMel Gorman <mgorman@techsingularity.net>
Thu, 28 Jul 2016 22:45:31 +0000 (15:45 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 28 Jul 2016 23:07:41 +0000 (16:07 -0700)
This moves the LRU lists from the zone to the node and related data such
as counters, tracing, congestion tracking and writeback tracking.

Unfortunately, due to reclaim and compaction retry logic, it is
necessary to account for the number of LRU pages on both zone and node
logic.  Most reclaim logic is based on the node counters but the retry
logic uses the zone counters which do not distinguish inactive and
active sizes.  It would be possible to leave the LRU counters on a
per-zone basis but it's a heavier calculation across multiple cache
lines that is much more frequent than the retry checks.

Other than the LRU counters, this is mostly a mechanical patch but note
that it introduces a number of anomalies.  For example, the scans are
per-zone but using per-node counters.  We also mark a node as congested
when a zone is congested.  This causes weird problems that are fixed
later but is easier to review.

In the event that there is excessive overhead on 32-bit systems due to
the nodes being on LRU then there are two potential solutions

1. Long-term isolation of highmem pages when reclaim is lowmem

   When pages are skipped, they are immediately added back onto the LRU
   list. If lowmem reclaim persisted for long periods of time, the same
   highmem pages get continually scanned. The idea would be that lowmem
   keeps those pages on a separate list until a reclaim for highmem pages
   arrives that splices the highmem pages back onto the LRU. It potentially
   could be implemented similar to the UNEVICTABLE list.

   That would reduce the skip rate with the potential corner case is that
   highmem pages have to be scanned and reclaimed to free lowmem slab pages.

2. Linear scan lowmem pages if the initial LRU shrink fails

   This will break LRU ordering but may be preferable and faster during
   memory pressure than skipping LRU pages.

Link: http://lkml.kernel.org/r/1467970510-21195-4-git-send-email-mgorman@techsingularity.net
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Rik van Riel <riel@surriel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
29 files changed:
arch/tile/mm/pgtable.c
drivers/base/node.c
drivers/staging/android/lowmemorykiller.c
include/linux/backing-dev.h
include/linux/memcontrol.h
include/linux/mm_inline.h
include/linux/mmzone.h
include/linux/swap.h
include/linux/vm_event_item.h
include/linux/vmstat.h
include/trace/events/vmscan.h
kernel/power/snapshot.c
mm/backing-dev.c
mm/compaction.c
mm/huge_memory.c
mm/internal.h
mm/khugepaged.c
mm/memcontrol.c
mm/memory-failure.c
mm/memory_hotplug.c
mm/mempolicy.c
mm/migrate.c
mm/mlock.c
mm/page-writeback.c
mm/page_alloc.c
mm/swap.c
mm/vmscan.c
mm/vmstat.c
mm/workingset.c

index c4d5bf8..9e38921 100644 (file)
@@ -45,10 +45,10 @@ void show_mem(unsigned int filter)
        struct zone *zone;
 
        pr_err("Active:%lu inactive:%lu dirty:%lu writeback:%lu unstable:%lu free:%lu\n slab:%lu mapped:%lu pagetables:%lu bounce:%lu pagecache:%lu swap:%lu\n",
-              (global_page_state(NR_ACTIVE_ANON) +
-               global_page_state(NR_ACTIVE_FILE)),
-              (global_page_state(NR_INACTIVE_ANON) +
-               global_page_state(NR_INACTIVE_FILE)),
+              (global_node_page_state(NR_ACTIVE_ANON) +
+               global_node_page_state(NR_ACTIVE_FILE)),
+              (global_node_page_state(NR_INACTIVE_ANON) +
+               global_node_page_state(NR_INACTIVE_FILE)),
               global_page_state(NR_FILE_DIRTY),
               global_page_state(NR_WRITEBACK),
               global_page_state(NR_UNSTABLE_NFS),
index 0a1b643..d4698f0 100644 (file)
@@ -56,6 +56,7 @@ static ssize_t node_read_meminfo(struct device *dev,
 {
        int n;
        int nid = dev->id;
+       struct pglist_data *pgdat = NODE_DATA(nid);
        struct sysinfo i;
 
        si_meminfo_node(&i, nid);
@@ -74,15 +75,15 @@ static ssize_t node_read_meminfo(struct device *dev,
                       nid, K(i.totalram),
                       nid, K(i.freeram),
                       nid, K(i.totalram - i.freeram),
-                      nid, K(sum_zone_node_page_state(nid, NR_ACTIVE_ANON) +
-                               sum_zone_node_page_state(nid, NR_ACTIVE_FILE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_INACTIVE_ANON) +
-                               sum_zone_node_page_state(nid, NR_INACTIVE_FILE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_ACTIVE_ANON)),
-                      nid, K(sum_zone_node_page_state(nid, NR_INACTIVE_ANON)),
-                      nid, K(sum_zone_node_page_state(nid, NR_ACTIVE_FILE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_INACTIVE_FILE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_UNEVICTABLE)),
+                      nid, K(node_page_state(pgdat, NR_ACTIVE_ANON) +
+                               node_page_state(pgdat, NR_ACTIVE_FILE)),
+                      nid, K(node_page_state(pgdat, NR_INACTIVE_ANON) +
+                               node_page_state(pgdat, NR_INACTIVE_FILE)),
+                      nid, K(node_page_state(pgdat, NR_ACTIVE_ANON)),
+                      nid, K(node_page_state(pgdat, NR_INACTIVE_ANON)),
+                      nid, K(node_page_state(pgdat, NR_ACTIVE_FILE)),
+                      nid, K(node_page_state(pgdat, NR_INACTIVE_FILE)),
+                      nid, K(node_page_state(pgdat, NR_UNEVICTABLE)),
                       nid, K(sum_zone_node_page_state(nid, NR_MLOCK)));
 
 #ifdef CONFIG_HIGHMEM
index 24d2745..93dbcc3 100644 (file)
@@ -72,10 +72,10 @@ static unsigned long lowmem_deathpending_timeout;
 static unsigned long lowmem_count(struct shrinker *s,
                                  struct shrink_control *sc)
 {
-       return global_page_state(NR_ACTIVE_ANON) +
-               global_page_state(NR_ACTIVE_FILE) +
-               global_page_state(NR_INACTIVE_ANON) +
-               global_page_state(NR_INACTIVE_FILE);
+       return global_node_page_state(NR_ACTIVE_ANON) +
+               global_node_page_state(NR_ACTIVE_FILE) +
+               global_node_page_state(NR_INACTIVE_ANON) +
+               global_node_page_state(NR_INACTIVE_FILE);
 }
 
 static unsigned long lowmem_scan(struct shrinker *s, struct shrink_control *sc)
index c82794f..491a917 100644 (file)
@@ -197,7 +197,7 @@ static inline int wb_congested(struct bdi_writeback *wb, int cong_bits)
 }
 
 long congestion_wait(int sync, long timeout);
-long wait_iff_congested(struct zone *zone, int sync, long timeout);
+long wait_iff_congested(struct pglist_data *pgdat, int sync, long timeout);
 int pdflush_proc_obsolete(struct ctl_table *table, int write,
                void __user *buffer, size_t *lenp, loff_t *ppos);
 
index 1c4df44..6d2321c 100644 (file)
@@ -339,7 +339,7 @@ static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
        struct lruvec *lruvec;
 
        if (mem_cgroup_disabled()) {
-               lruvec = &zone->lruvec;
+               lruvec = zone_lruvec(zone);
                goto out;
        }
 
@@ -348,15 +348,15 @@ static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
 out:
        /*
         * Since a node can be onlined after the mem_cgroup was created,
-        * we have to be prepared to initialize lruvec->zone here;
+        * we have to be prepared to initialize lruvec->pgdat here;
         * and if offlined then reonlined, we need to reinitialize it.
         */
-       if (unlikely(lruvec->zone != zone))
-               lruvec->zone = zone;
+       if (unlikely(lruvec->pgdat != zone->zone_pgdat))
+               lruvec->pgdat = zone->zone_pgdat;
        return lruvec;
 }
 
-struct lruvec *mem_cgroup_page_lruvec(struct page *, struct zone *);
+struct lruvec *mem_cgroup_page_lruvec(struct page *, struct pglist_data *);
 
 bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
 struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
@@ -437,7 +437,7 @@ static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
 int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
 
 void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
-               int nr_pages);
+               enum zone_type zid, int nr_pages);
 
 unsigned long mem_cgroup_node_nr_lru_pages(struct mem_cgroup *memcg,
                                           int nid, unsigned int lru_mask);
@@ -612,13 +612,13 @@ static inline void mem_cgroup_migrate(struct page *old, struct page *new)
 static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
                                                    struct mem_cgroup *memcg)
 {
-       return &zone->lruvec;
+       return zone_lruvec(zone);
 }
 
 static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
-                                                   struct zone *zone)
+                                                   struct pglist_data *pgdat)
 {
-       return &zone->lruvec;
+       return &pgdat->lruvec;
 }
 
 static inline bool mm_match_cgroup(struct mm_struct *mm,
index 5bd29ba..9aadcc7 100644 (file)
@@ -23,25 +23,32 @@ static inline int page_is_file_cache(struct page *page)
 }
 
 static __always_inline void __update_lru_size(struct lruvec *lruvec,
-                               enum lru_list lru, int nr_pages)
+                               enum lru_list lru, enum zone_type zid,
+                               int nr_pages)
 {
-       __mod_zone_page_state(lruvec_zone(lruvec), NR_LRU_BASE + lru, nr_pages);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
+
+       __mod_node_page_state(pgdat, NR_LRU_BASE + lru, nr_pages);
+       __mod_zone_page_state(&pgdat->node_zones[zid],
+               NR_ZONE_LRU_BASE + !!is_file_lru(lru),
+               nr_pages);
 }
 
 static __always_inline void update_lru_size(struct lruvec *lruvec,
-                               enum lru_list lru, int nr_pages)
+                               enum lru_list lru, enum zone_type zid,
+                               int nr_pages)
 {
 #ifdef CONFIG_MEMCG
-       mem_cgroup_update_lru_size(lruvec, lru, nr_pages);
+       mem_cgroup_update_lru_size(lruvec, lru, zid, nr_pages);
 #else
-       __update_lru_size(lruvec, lru, nr_pages);
+       __update_lru_size(lruvec, lru, zid, nr_pages);
 #endif
 }
 
 static __always_inline void add_page_to_lru_list(struct page *page,
                                struct lruvec *lruvec, enum lru_list lru)
 {
-       update_lru_size(lruvec, lru, hpage_nr_pages(page));
+       update_lru_size(lruvec, lru, page_zonenum(page), hpage_nr_pages(page));
        list_add(&page->lru, &lruvec->lists[lru]);
 }
 
@@ -49,7 +56,7 @@ static __always_inline void del_page_from_lru_list(struct page *page,
                                struct lruvec *lruvec, enum lru_list lru)
 {
        list_del(&page->lru);
-       update_lru_size(lruvec, lru, -hpage_nr_pages(page));
+       update_lru_size(lruvec, lru, page_zonenum(page), -hpage_nr_pages(page));
 }
 
 /**
index cfa8701..d4f5cac 100644 (file)
@@ -111,12 +111,9 @@ enum zone_stat_item {
        /* First 128 byte cacheline (assuming 64 bit words) */
        NR_FREE_PAGES,
        NR_ALLOC_BATCH,
-       NR_LRU_BASE,
-       NR_INACTIVE_ANON = NR_LRU_BASE, /* must match order of LRU_[IN]ACTIVE */
-       NR_ACTIVE_ANON,         /*  "     "     "   "       "         */
-       NR_INACTIVE_FILE,       /*  "     "     "   "       "         */
-       NR_ACTIVE_FILE,         /*  "     "     "   "       "         */
-       NR_UNEVICTABLE,         /*  "     "     "   "       "         */
+       NR_ZONE_LRU_BASE, /* Used only for compaction and reclaim retry */
+       NR_ZONE_LRU_ANON = NR_ZONE_LRU_BASE,
+       NR_ZONE_LRU_FILE,
        NR_MLOCK,               /* mlock()ed pages found and moved off LRU */
        NR_ANON_PAGES,  /* Mapped anonymous pages */
        NR_FILE_MAPPED, /* pagecache pages mapped into pagetables.
@@ -134,12 +131,9 @@ enum zone_stat_item {
        NR_VMSCAN_WRITE,
        NR_VMSCAN_IMMEDIATE,    /* Prioritise for reclaim when writeback ends */
        NR_WRITEBACK_TEMP,      /* Writeback using temporary buffers */
-       NR_ISOLATED_ANON,       /* Temporary isolated pages from anon lru */
-       NR_ISOLATED_FILE,       /* Temporary isolated pages from file lru */
        NR_SHMEM,               /* shmem pages (included tmpfs/GEM pages) */
        NR_DIRTIED,             /* page dirtyings since bootup */
        NR_WRITTEN,             /* page writings since bootup */
-       NR_PAGES_SCANNED,       /* pages scanned since last reclaim */
 #if IS_ENABLED(CONFIG_ZSMALLOC)
        NR_ZSPAGES,             /* allocated in zsmalloc */
 #endif
@@ -161,6 +155,15 @@ enum zone_stat_item {
        NR_VM_ZONE_STAT_ITEMS };
 
 enum node_stat_item {
+       NR_LRU_BASE,
+       NR_INACTIVE_ANON = NR_LRU_BASE, /* must match order of LRU_[IN]ACTIVE */
+       NR_ACTIVE_ANON,         /*  "     "     "   "       "         */
+       NR_INACTIVE_FILE,       /*  "     "     "   "       "         */
+       NR_ACTIVE_FILE,         /*  "     "     "   "       "         */
+       NR_UNEVICTABLE,         /*  "     "     "   "       "         */
+       NR_ISOLATED_ANON,       /* Temporary isolated pages from anon lru */
+       NR_ISOLATED_FILE,       /* Temporary isolated pages from file lru */
+       NR_PAGES_SCANNED,       /* pages scanned since last reclaim */
        NR_VM_NODE_STAT_ITEMS
 };
 
@@ -219,7 +222,7 @@ struct lruvec {
        /* Evictions & activations on the inactive file list */
        atomic_long_t                   inactive_age;
 #ifdef CONFIG_MEMCG
-       struct zone                     *zone;
+       struct pglist_data *pgdat;
 #endif
 };
 
@@ -357,13 +360,6 @@ struct zone {
 #ifdef CONFIG_NUMA
        int node;
 #endif
-
-       /*
-        * The target ratio of ACTIVE_ANON to INACTIVE_ANON pages on
-        * this zone's LRU.  Maintained by the pageout code.
-        */
-       unsigned int inactive_ratio;
-
        struct pglist_data      *zone_pgdat;
        struct per_cpu_pageset __percpu *pageset;
 
@@ -495,9 +491,6 @@ struct zone {
 
        /* Write-intensive fields used by page reclaim */
 
-       /* Fields commonly accessed by the page reclaim scanner */
-       struct lruvec           lruvec;
-
        /*
         * When free pages are below this point, additional steps are taken
         * when reading the number of free pages to avoid per-cpu counter
@@ -537,17 +530,20 @@ struct zone {
 
 enum zone_flags {
        ZONE_RECLAIM_LOCKED,            /* prevents concurrent reclaim */
-       ZONE_CONGESTED,                 /* zone has many dirty pages backed by
+       ZONE_FAIR_DEPLETED,             /* fair zone policy batch depleted */
+};
+
+enum pgdat_flags {
+       PGDAT_CONGESTED,                /* pgdat has many dirty pages backed by
                                         * a congested BDI
                                         */
-       ZONE_DIRTY,                     /* reclaim scanning has recently found
+       PGDAT_DIRTY,                    /* reclaim scanning has recently found
                                         * many dirty file pages at the tail
                                         * of the LRU.
                                         */
-       ZONE_WRITEBACK,                 /* reclaim scanning has recently found
+       PGDAT_WRITEBACK,                /* reclaim scanning has recently found
                                         * many pages under writeback
                                         */
-       ZONE_FAIR_DEPLETED,             /* fair zone policy batch depleted */
 };
 
 static inline unsigned long zone_end_pfn(const struct zone *zone)
@@ -707,6 +703,19 @@ typedef struct pglist_data {
        unsigned long split_queue_len;
 #endif
 
+       /* Fields commonly accessed by the page reclaim scanner */
+       struct lruvec           lruvec;
+
+       /*
+        * The target ratio of ACTIVE_ANON to INACTIVE_ANON pages on
+        * this node's LRU.  Maintained by the pageout code.
+        */
+       unsigned int inactive_ratio;
+
+       unsigned long           flags;
+
+       ZONE_PADDING(_pad2_)
+
        /* Per-node vmstats */
        struct per_cpu_nodestat __percpu *per_cpu_nodestats;
        atomic_long_t           vm_stat[NR_VM_NODE_STAT_ITEMS];
@@ -728,6 +737,11 @@ static inline spinlock_t *zone_lru_lock(struct zone *zone)
        return &zone->zone_pgdat->lru_lock;
 }
 
+static inline struct lruvec *zone_lruvec(struct zone *zone)
+{
+       return &zone->zone_pgdat->lruvec;
+}
+
 static inline unsigned long pgdat_end_pfn(pg_data_t *pgdat)
 {
        return pgdat->node_start_pfn + pgdat->node_spanned_pages;
@@ -779,12 +793,12 @@ extern int init_currently_empty_zone(struct zone *zone, unsigned long start_pfn,
 
 extern void lruvec_init(struct lruvec *lruvec);
 
-static inline struct zone *lruvec_zone(struct lruvec *lruvec)
+static inline struct pglist_data *lruvec_pgdat(struct lruvec *lruvec)
 {
 #ifdef CONFIG_MEMCG
-       return lruvec->zone;
+       return lruvec->pgdat;
 #else
-       return container_of(lruvec, struct zone, lruvec);
+       return container_of(lruvec, struct pglist_data, lruvec);
 #endif
 }
 
index 0af2bb2..c82f916 100644 (file)
@@ -317,6 +317,7 @@ extern void lru_cache_add_active_or_unevictable(struct page *page,
 
 /* linux/mm/vmscan.c */
 extern unsigned long zone_reclaimable_pages(struct zone *zone);
+extern unsigned long pgdat_reclaimable_pages(struct pglist_data *pgdat);
 extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
                                        gfp_t gfp_mask, nodemask_t *mask);
 extern int __isolate_lru_page(struct page *page, isolate_mode_t mode);
index 4260417..1798ff5 100644 (file)
@@ -26,11 +26,11 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT,
                PGFREE, PGACTIVATE, PGDEACTIVATE,
                PGFAULT, PGMAJFAULT,
                PGLAZYFREED,
-               FOR_ALL_ZONES(PGREFILL),
-               FOR_ALL_ZONES(PGSTEAL_KSWAPD),
-               FOR_ALL_ZONES(PGSTEAL_DIRECT),
-               FOR_ALL_ZONES(PGSCAN_KSWAPD),
-               FOR_ALL_ZONES(PGSCAN_DIRECT),
+               PGREFILL,
+               PGSTEAL_KSWAPD,
+               PGSTEAL_DIRECT,
+               PGSCAN_KSWAPD,
+               PGSCAN_DIRECT,
                PGSCAN_DIRECT_THROTTLE,
 #ifdef CONFIG_NUMA
                PGSCAN_ZONE_RECLAIM_FAILED,
index d1744aa..fee321c 100644 (file)
@@ -178,6 +178,23 @@ static inline unsigned long zone_page_state_snapshot(struct zone *zone,
        return x;
 }
 
+static inline unsigned long node_page_state_snapshot(pg_data_t *pgdat,
+                                       enum node_stat_item item)
+{
+       long x = atomic_long_read(&pgdat->vm_stat[item]);
+
+#ifdef CONFIG_SMP
+       int cpu;
+       for_each_online_cpu(cpu)
+               x += per_cpu_ptr(pgdat->per_cpu_nodestats, cpu)->vm_node_stat_diff[item];
+
+       if (x < 0)
+               x = 0;
+#endif
+       return x;
+}
+
+
 #ifdef CONFIG_NUMA
 extern unsigned long sum_zone_node_page_state(int node,
                                                enum zone_stat_item item);
index 0101ef3..897f1aa 100644 (file)
@@ -352,15 +352,14 @@ TRACE_EVENT(mm_vmscan_writepage,
 
 TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
 
-       TP_PROTO(struct zone *zone,
+       TP_PROTO(int nid,
                unsigned long nr_scanned, unsigned long nr_reclaimed,
                int priority, int file),
 
-       TP_ARGS(zone, nr_scanned, nr_reclaimed, priority, file),
+       TP_ARGS(nid, nr_scanned, nr_reclaimed, priority, file),
 
        TP_STRUCT__entry(
                __field(int, nid)
-               __field(int, zid)
                __field(unsigned long, nr_scanned)
                __field(unsigned long, nr_reclaimed)
                __field(int, priority)
@@ -368,16 +367,15 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
        ),
 
        TP_fast_assign(
-               __entry->nid = zone_to_nid(zone);
-               __entry->zid = zone_idx(zone);
+               __entry->nid = nid;
                __entry->nr_scanned = nr_scanned;
                __entry->nr_reclaimed = nr_reclaimed;
                __entry->priority = priority;
                __entry->reclaim_flags = trace_shrink_flags(file);
        ),
 
-       TP_printk("nid=%d zid=%d nr_scanned=%ld nr_reclaimed=%ld priority=%d flags=%s",
-               __entry->nid, __entry->zid,
+       TP_printk("nid=%d nr_scanned=%ld nr_reclaimed=%ld priority=%d flags=%s",
+               __entry->nid,
                __entry->nr_scanned, __entry->nr_reclaimed,
                __entry->priority,
                show_reclaim_flags(__entry->reclaim_flags))
index d90df92..9a0178c 100644 (file)
@@ -1627,11 +1627,11 @@ static unsigned long minimum_image_size(unsigned long saveable)
        unsigned long size;
 
        size = global_page_state(NR_SLAB_RECLAIMABLE)
-               + global_page_state(NR_ACTIVE_ANON)
-               + global_page_state(NR_INACTIVE_ANON)
-               + global_page_state(NR_ACTIVE_FILE)
-               + global_page_state(NR_INACTIVE_FILE)
-               - global_page_state(NR_FILE_MAPPED);
+               + global_node_page_state(NR_ACTIVE_ANON)
+               + global_node_page_state(NR_INACTIVE_ANON)
+               + global_node_page_state(NR_ACTIVE_FILE)
+               + global_node_page_state(NR_INACTIVE_FILE)
+               - global_node_page_state(NR_FILE_MAPPED);
 
        return saveable <= size ? 0 : saveable - size;
 }
index ed173b8..efe2377 100644 (file)
@@ -947,24 +947,24 @@ long congestion_wait(int sync, long timeout)
 EXPORT_SYMBOL(congestion_wait);
 
 /**
- * wait_iff_congested - Conditionally wait for a backing_dev to become uncongested or a zone to complete writes
- * @zone: A zone to check if it is heavily congested
+ * wait_iff_congested - Conditionally wait for a backing_dev to become uncongested or a pgdat to complete writes
+ * @pgdat: A pgdat to check if it is heavily congested
  * @sync: SYNC or ASYNC IO
  * @timeout: timeout in jiffies
  *
  * In the event of a congested backing_dev (any backing_dev) and the given
- * @zone has experienced recent congestion, this waits for up to @timeout
+ * @pgdat has experienced recent congestion, this waits for up to @timeout
  * jiffies for either a BDI to exit congestion of the given @sync queue
  * or a write to complete.
  *
- * In the absence of zone congestion, cond_resched() is called to yield
+ * In the absence of pgdat congestion, cond_resched() is called to yield
  * the processor if necessary but otherwise does not sleep.
  *
  * The return value is 0 if the sleep is for the full timeout. Otherwise,
  * it is the number of jiffies that were still remaining when the function
  * returned. return_value == timeout implies the function did not sleep.
  */
-long wait_iff_congested(struct zone *zone, int sync, long timeout)
+long wait_iff_congested(struct pglist_data *pgdat, int sync, long timeout)
 {
        long ret;
        unsigned long start = jiffies;
@@ -973,12 +973,13 @@ long wait_iff_congested(struct zone *zone, int sync, long timeout)
 
        /*
         * If there is no congestion, or heavy congestion is not being
-        * encountered in the current zone, yield if necessary instead
+        * encountered in the current pgdat, yield if necessary instead
         * of sleeping on the congestion queue
         */
        if (atomic_read(&nr_wb_congested[sync]) == 0 ||
-           !test_bit(ZONE_CONGESTED, &zone->flags)) {
+           !test_bit(PGDAT_CONGESTED, &pgdat->flags)) {
                cond_resched();
+
                /* In case we scheduled, work out time remaining */
                ret = timeout - (jiffies - start);
                if (ret < 0)
index 5c65fad..e5995f3 100644 (file)
@@ -646,8 +646,8 @@ static void acct_isolated(struct zone *zone, struct compact_control *cc)
        list_for_each_entry(page, &cc->migratepages, lru)
                count[!!page_is_file_cache(page)]++;
 
-       mod_zone_page_state(zone, NR_ISOLATED_ANON, count[0]);
-       mod_zone_page_state(zone, NR_ISOLATED_FILE, count[1]);
+       mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_ANON, count[0]);
+       mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE, count[1]);
 }
 
 /* Similar to reclaim, but different enough that they don't share logic */
@@ -655,12 +655,12 @@ static bool too_many_isolated(struct zone *zone)
 {
        unsigned long active, inactive, isolated;
 
-       inactive = zone_page_state(zone, NR_INACTIVE_FILE) +
-                                       zone_page_state(zone, NR_INACTIVE_ANON);
-       active = zone_page_state(zone, NR_ACTIVE_FILE) +
-                                       zone_page_state(zone, NR_ACTIVE_ANON);
-       isolated = zone_page_state(zone, NR_ISOLATED_FILE) +
-                                       zone_page_state(zone, NR_ISOLATED_ANON);
+       inactive = node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE) +
+                       node_page_state(zone->zone_pgdat, NR_INACTIVE_ANON);
+       active = node_page_state(zone->zone_pgdat, NR_ACTIVE_FILE) +
+                       node_page_state(zone->zone_pgdat, NR_ACTIVE_ANON);
+       isolated = node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE) +
+                       node_page_state(zone->zone_pgdat, NR_ISOLATED_ANON);
 
        return isolated > (inactive + active) / 2;
 }
@@ -856,7 +856,7 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
                        }
                }
 
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
 
                /* Try isolate the page */
                if (__isolate_lru_page(page, isolate_mode) != 0)
index 99578b6..481fb01 100644 (file)
@@ -1818,7 +1818,7 @@ static void __split_huge_page(struct page *page, struct list_head *list,
        pgoff_t end = -1;
        int i;
 
-       lruvec = mem_cgroup_page_lruvec(head, zone);
+       lruvec = mem_cgroup_page_lruvec(head, zone->zone_pgdat);
 
        /* complete memcg works before add pages to LRU */
        mem_cgroup_split_huge_fixup(head);
index 9b6a6c4..2f80d03 100644 (file)
@@ -78,7 +78,7 @@ extern unsigned long highest_memmap_pfn;
  */
 extern int isolate_lru_page(struct page *page);
 extern void putback_lru_page(struct page *page);
-extern bool zone_reclaimable(struct zone *zone);
+extern bool pgdat_reclaimable(struct pglist_data *pgdat);
 
 /*
  * in mm/rmap.c:
index 7dbee69..374237b 100644 (file)
@@ -480,7 +480,7 @@ void __khugepaged_exit(struct mm_struct *mm)
 static void release_pte_page(struct page *page)
 {
        /* 0 stands for page_is_file_cache(page) == false */
-       dec_zone_page_state(page, NR_ISOLATED_ANON + 0);
+       dec_node_page_state(page, NR_ISOLATED_ANON + 0);
        unlock_page(page);
        putback_lru_page(page);
 }
@@ -576,7 +576,7 @@ static int __collapse_huge_page_isolate(struct vm_area_struct *vma,
                        goto out;
                }
                /* 0 stands for page_is_file_cache(page) == false */
-               inc_zone_page_state(page, NR_ISOLATED_ANON + 0);
+               inc_node_page_state(page, NR_ISOLATED_ANON + 0);
                VM_BUG_ON_PAGE(!PageLocked(page), page);
                VM_BUG_ON_PAGE(PageLRU(page), page);
 
index 9b70f9c..50c86ad 100644 (file)
@@ -943,14 +943,14 @@ static void invalidate_reclaim_iterators(struct mem_cgroup *dead_memcg)
  * and putback protocol: the LRU lock must be held, and the page must
  * either be PageLRU() or the caller must have isolated/allocated it.
  */
-struct lruvec *mem_cgroup_page_lruvec(struct page *page, struct zone *zone)
+struct lruvec *mem_cgroup_page_lruvec(struct page *page, struct pglist_data *pgdat)
 {
        struct mem_cgroup_per_zone *mz;
        struct mem_cgroup *memcg;
        struct lruvec *lruvec;
 
        if (mem_cgroup_disabled()) {
-               lruvec = &zone->lruvec;
+               lruvec = &pgdat->lruvec;
                goto out;
        }
 
@@ -970,8 +970,8 @@ out:
         * we have to be prepared to initialize lruvec->zone here;
         * and if offlined then reonlined, we need to reinitialize it.
         */
-       if (unlikely(lruvec->zone != zone))
-               lruvec->zone = zone;
+       if (unlikely(lruvec->pgdat != pgdat))
+               lruvec->pgdat = pgdat;
        return lruvec;
 }
 
@@ -979,6 +979,7 @@ out:
  * mem_cgroup_update_lru_size - account for adding or removing an lru page
  * @lruvec: mem_cgroup per zone lru vector
  * @lru: index of lru list the page is sitting on
+ * @zid: Zone ID of the zone pages have been added to
  * @nr_pages: positive when adding or negative when removing
  *
  * This function must be called under lru_lock, just before a page is added
@@ -986,14 +987,14 @@ out:
  * so as to allow it to check that lru_size 0 is consistent with list_empty).
  */
 void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
-                               int nr_pages)
+                               enum zone_type zid, int nr_pages)
 {
        struct mem_cgroup_per_zone *mz;
        unsigned long *lru_size;
        long size;
        bool empty;
 
-       __update_lru_size(lruvec, lru, nr_pages);
+       __update_lru_size(lruvec, lru, zid, nr_pages);
 
        if (mem_cgroup_disabled())
                return;
@@ -2069,7 +2070,7 @@ static void lock_page_lru(struct page *page, int *isolated)
        if (PageLRU(page)) {
                struct lruvec *lruvec;
 
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
                ClearPageLRU(page);
                del_page_from_lru_list(page, lruvec, page_lru(page));
                *isolated = 1;
@@ -2084,7 +2085,7 @@ static void unlock_page_lru(struct page *page, int isolated)
        if (isolated) {
                struct lruvec *lruvec;
 
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
                VM_BUG_ON_PAGE(PageLRU(page), page);
                SetPageLRU(page);
                add_page_to_lru_list(page, lruvec, page_lru(page));
index 2fcca6b..11de752 100644 (file)
@@ -1663,7 +1663,7 @@ static int __soft_offline_page(struct page *page, int flags)
        put_hwpoison_page(page);
        if (!ret) {
                LIST_HEAD(pagelist);
-               inc_zone_page_state(page, NR_ISOLATED_ANON +
+               inc_node_page_state(page, NR_ISOLATED_ANON +
                                        page_is_file_cache(page));
                list_add(&page->lru, &pagelist);
                ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL,
@@ -1671,7 +1671,7 @@ static int __soft_offline_page(struct page *page, int flags)
                if (ret) {
                        if (!list_empty(&pagelist)) {
                                list_del(&page->lru);
-                               dec_zone_page_state(page, NR_ISOLATED_ANON +
+                               dec_node_page_state(page, NR_ISOLATED_ANON +
                                                page_is_file_cache(page));
                                putback_lru_page(page);
                        }
index 82d0b98..c527836 100644 (file)
@@ -1586,7 +1586,7 @@ do_migrate_range(unsigned long start_pfn, unsigned long end_pfn)
                        put_page(page);
                        list_add_tail(&page->lru, &source);
                        move_pages--;
-                       inc_zone_page_state(page, NR_ISOLATED_ANON +
+                       inc_node_page_state(page, NR_ISOLATED_ANON +
                                            page_is_file_cache(page));
 
                } else {
index 53e40d3..d8c4e38 100644 (file)
@@ -962,7 +962,7 @@ static void migrate_page_add(struct page *page, struct list_head *pagelist,
        if ((flags & MPOL_MF_MOVE_ALL) || page_mapcount(page) == 1) {
                if (!isolate_lru_page(page)) {
                        list_add_tail(&page->lru, pagelist);
-                       inc_zone_page_state(page, NR_ISOLATED_ANON +
+                       inc_node_page_state(page, NR_ISOLATED_ANON +
                                            page_is_file_cache(page));
                }
        }
index 2232f69..3033dae 100644 (file)
@@ -168,7 +168,7 @@ void putback_movable_pages(struct list_head *l)
                        continue;
                }
                list_del(&page->lru);
-               dec_zone_page_state(page, NR_ISOLATED_ANON +
+               dec_node_page_state(page, NR_ISOLATED_ANON +
                                page_is_file_cache(page));
                /*
                 * We isolated non-lru movable page so here we can use
@@ -1119,7 +1119,7 @@ out:
                 * restored.
                 */
                list_del(&page->lru);
-               dec_zone_page_state(page, NR_ISOLATED_ANON +
+               dec_node_page_state(page, NR_ISOLATED_ANON +
                                page_is_file_cache(page));
        }
 
@@ -1460,7 +1460,7 @@ static int do_move_page_to_node_array(struct mm_struct *mm,
                err = isolate_lru_page(page);
                if (!err) {
                        list_add_tail(&page->lru, &pagelist);
-                       inc_zone_page_state(page, NR_ISOLATED_ANON +
+                       inc_node_page_state(page, NR_ISOLATED_ANON +
                                            page_is_file_cache(page));
                }
 put_and_set:
@@ -1726,15 +1726,16 @@ static bool migrate_balanced_pgdat(struct pglist_data *pgdat,
                                   unsigned long nr_migrate_pages)
 {
        int z;
+
+       if (!pgdat_reclaimable(pgdat))
+               return false;
+
        for (z = pgdat->nr_zones - 1; z >= 0; z--) {
                struct zone *zone = pgdat->node_zones + z;
 
                if (!populated_zone(zone))
                        continue;
 
-               if (!zone_reclaimable(zone))
-                       continue;
-
                /* Avoid waking kswapd by allocating pages_to_migrate pages. */
                if (!zone_watermark_ok(zone, 0,
                                       high_wmark_pages(zone) +
@@ -1828,7 +1829,7 @@ static int numamigrate_isolate_page(pg_data_t *pgdat, struct page *page)
        }
 
        page_lru = page_is_file_cache(page);
-       mod_zone_page_state(page_zone(page), NR_ISOLATED_ANON + page_lru,
+       mod_node_page_state(page_pgdat(page), NR_ISOLATED_ANON + page_lru,
                                hpage_nr_pages(page));
 
        /*
@@ -1886,7 +1887,7 @@ int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma,
        if (nr_remaining) {
                if (!list_empty(&migratepages)) {
                        list_del(&page->lru);
-                       dec_zone_page_state(page, NR_ISOLATED_ANON +
+                       dec_node_page_state(page, NR_ISOLATED_ANON +
                                        page_is_file_cache(page));
                        putback_lru_page(page);
                }
@@ -1979,7 +1980,7 @@ fail_putback:
                /* Retake the callers reference and putback on LRU */
                get_page(page);
                putback_lru_page(page);
-               mod_zone_page_state(page_zone(page),
+               mod_node_page_state(page_pgdat(page),
                         NR_ISOLATED_ANON + page_lru, -HPAGE_PMD_NR);
 
                goto out_unlock;
@@ -2030,7 +2031,7 @@ fail_putback:
        count_vm_events(PGMIGRATE_SUCCESS, HPAGE_PMD_NR);
        count_vm_numa_events(NUMA_PAGE_MIGRATE, HPAGE_PMD_NR);
 
-       mod_zone_page_state(page_zone(page),
+       mod_node_page_state(page_pgdat(page),
                        NR_ISOLATED_ANON + page_lru,
                        -HPAGE_PMD_NR);
        return isolated;
index 997f630..14645be 100644 (file)
@@ -103,7 +103,7 @@ static bool __munlock_isolate_lru_page(struct page *page, bool getpage)
        if (PageLRU(page)) {
                struct lruvec *lruvec;
 
-               lruvec = mem_cgroup_page_lruvec(page, page_zone(page));
+               lruvec = mem_cgroup_page_lruvec(page, page_pgdat(page));
                if (getpage)
                        get_page(page);
                ClearPageLRU(page);
index d578d2a..0ada2b2 100644 (file)
@@ -285,8 +285,8 @@ static unsigned long zone_dirtyable_memory(struct zone *zone)
         */
        nr_pages -= min(nr_pages, zone->totalreserve_pages);
 
-       nr_pages += zone_page_state(zone, NR_INACTIVE_FILE);
-       nr_pages += zone_page_state(zone, NR_ACTIVE_FILE);
+       nr_pages += node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE);
+       nr_pages += node_page_state(zone->zone_pgdat, NR_ACTIVE_FILE);
 
        return nr_pages;
 }
@@ -348,8 +348,8 @@ static unsigned long global_dirtyable_memory(void)
         */
        x -= min(x, totalreserve_pages);
 
-       x += global_page_state(NR_INACTIVE_FILE);
-       x += global_page_state(NR_ACTIVE_FILE);
+       x += global_node_page_state(NR_INACTIVE_FILE);
+       x += global_node_page_state(NR_ACTIVE_FILE);
 
        if (!vm_highmem_is_dirtyable)
                x -= highmem_dirtyable_memory(x);
index 5760c62..35e2d0f 100644 (file)
@@ -1078,9 +1078,9 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 
        spin_lock(&zone->lock);
        isolated_pageblocks = has_isolate_pageblock(zone);
-       nr_scanned = zone_page_state(zone, NR_PAGES_SCANNED);
+       nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED);
        if (nr_scanned)
-               __mod_zone_page_state(zone, NR_PAGES_SCANNED, -nr_scanned);
+               __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned);
 
        while (count) {
                struct page *page;
@@ -1135,9 +1135,9 @@ static void free_one_page(struct zone *zone,
 {
        unsigned long nr_scanned;
        spin_lock(&zone->lock);
-       nr_scanned = zone_page_state(zone, NR_PAGES_SCANNED);
+       nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED);
        if (nr_scanned)
-               __mod_zone_page_state(zone, NR_PAGES_SCANNED, -nr_scanned);
+               __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned);
 
        if (unlikely(has_isolate_pageblock(zone) ||
                is_migrate_isolate(migratetype))) {
@@ -4288,6 +4288,7 @@ void show_free_areas(unsigned int filter)
        unsigned long free_pcp = 0;
        int cpu;
        struct zone *zone;
+       pg_data_t *pgdat;
 
        for_each_populated_zone(zone) {
                if (skip_free_areas_node(filter, zone_to_nid(zone)))
@@ -4306,13 +4307,13 @@ void show_free_areas(unsigned int filter)
                " anon_thp: %lu shmem_thp: %lu shmem_pmdmapped: %lu\n"
 #endif
                " free:%lu free_pcp:%lu free_cma:%lu\n",
-               global_page_state(NR_ACTIVE_ANON),
-               global_page_state(NR_INACTIVE_ANON),
-               global_page_state(NR_ISOLATED_ANON),
-               global_page_state(NR_ACTIVE_FILE),
-               global_page_state(NR_INACTIVE_FILE),
-               global_page_state(NR_ISOLATED_FILE),
-               global_page_state(NR_UNEVICTABLE),
+               global_node_page_state(NR_ACTIVE_ANON),
+               global_node_page_state(NR_INACTIVE_ANON),
+               global_node_page_state(NR_ISOLATED_ANON),
+               global_node_page_state(NR_ACTIVE_FILE),
+               global_node_page_state(NR_INACTIVE_FILE),
+               global_node_page_state(NR_ISOLATED_FILE),
+               global_node_page_state(NR_UNEVICTABLE),
                global_page_state(NR_FILE_DIRTY),
                global_page_state(NR_WRITEBACK),
                global_page_state(NR_UNSTABLE_NFS),
@@ -4331,6 +4332,28 @@ void show_free_areas(unsigned int filter)
                free_pcp,
                global_page_state(NR_FREE_CMA_PAGES));
 
+       for_each_online_pgdat(pgdat) {
+               printk("Node %d"
+                       " active_anon:%lukB"
+                       " inactive_anon:%lukB"
+                       " active_file:%lukB"
+                       " inactive_file:%lukB"
+                       " unevictable:%lukB"
+                       " isolated(anon):%lukB"
+                       " isolated(file):%lukB"
+                       " all_unreclaimable? %s"
+                       "\n",
+                       pgdat->node_id,
+                       K(node_page_state(pgdat, NR_ACTIVE_ANON)),
+                       K(node_page_state(pgdat, NR_INACTIVE_ANON)),
+                       K(node_page_state(pgdat, NR_ACTIVE_FILE)),
+                       K(node_page_state(pgdat, NR_INACTIVE_FILE)),
+                       K(node_page_state(pgdat, NR_UNEVICTABLE)),
+                       K(node_page_state(pgdat, NR_ISOLATED_ANON)),
+                       K(node_page_state(pgdat, NR_ISOLATED_FILE)),
+                       !pgdat_reclaimable(pgdat) ? "yes" : "no");
+       }
+
        for_each_populated_zone(zone) {
                int i;
 
@@ -4347,13 +4370,6 @@ void show_free_areas(unsigned int filter)
                        " min:%lukB"
                        " low:%lukB"
                        " high:%lukB"
-                       " active_anon:%lukB"
-                       " inactive_anon:%lukB"
-                       " active_file:%lukB"
-                       " inactive_file:%lukB"
-                       " unevictable:%lukB"
-                       " isolated(anon):%lukB"
-                       " isolated(file):%lukB"
                        " present:%lukB"
                        " managed:%lukB"
                        " mlocked:%lukB"
@@ -4376,21 +4392,13 @@ void show_free_areas(unsigned int filter)
                        " local_pcp:%ukB"
                        " free_cma:%lukB"
                        " writeback_tmp:%lukB"
-                       " pages_scanned:%lu"
-                       " all_unreclaimable? %s"
+                       " node_pages_scanned:%lu"
                        "\n",
                        zone->name,
                        K(zone_page_state(zone, NR_FREE_PAGES)),
                        K(min_wmark_pages(zone)),
                        K(low_wmark_pages(zone)),
                        K(high_wmark_pages(zone)),
-                       K(zone_page_state(zone, NR_ACTIVE_ANON)),
-                       K(zone_page_state(zone, NR_INACTIVE_ANON)),
-                       K(zone_page_state(zone, NR_ACTIVE_FILE)),
-                       K(zone_page_state(zone, NR_INACTIVE_FILE)),
-                       K(zone_page_state(zone, NR_UNEVICTABLE)),
-                       K(zone_page_state(zone, NR_ISOLATED_ANON)),
-                       K(zone_page_state(zone, NR_ISOLATED_FILE)),
                        K(zone->present_pages),
                        K(zone->managed_pages),
                        K(zone_page_state(zone, NR_MLOCK)),
@@ -4415,9 +4423,7 @@ void show_free_areas(unsigned int filter)
                        K(this_cpu_read(zone->pageset->pcp.count)),
                        K(zone_page_state(zone, NR_FREE_CMA_PAGES)),
                        K(zone_page_state(zone, NR_WRITEBACK_TEMP)),
-                       K(zone_page_state(zone, NR_PAGES_SCANNED)),
-                       (!zone_reclaimable(zone) ? "yes" : "no")
-                       );
+                       K(node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED)));
                printk("lowmem_reserve[]:");
                for (i = 0; i < MAX_NR_ZONES; i++)
                        printk(" %ld", zone->lowmem_reserve[i]);
@@ -5967,7 +5973,7 @@ static void __paginginit free_area_init_core(struct pglist_data *pgdat)
                /* For bootup, initialized properly in watermark setup */
                mod_zone_page_state(zone, NR_ALLOC_BATCH, zone->managed_pages);
 
-               lruvec_init(&zone->lruvec);
+               lruvec_init(zone_lruvec(zone));
                if (!size)
                        continue;
 
index bf37e5c..77af473 100644 (file)
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -63,7 +63,7 @@ static void __page_cache_release(struct page *page)
                unsigned long flags;
 
                spin_lock_irqsave(zone_lru_lock(zone), flags);
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
                VM_BUG_ON_PAGE(!PageLRU(page), page);
                __ClearPageLRU(page);
                del_page_from_lru_list(page, lruvec, page_off_lru(page));
@@ -194,7 +194,7 @@ static void pagevec_lru_move_fn(struct pagevec *pvec,
                        spin_lock_irqsave(zone_lru_lock(zone), flags);
                }
 
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
                (*move_fn)(page, lruvec, arg);
        }
        if (zone)
@@ -319,7 +319,7 @@ void activate_page(struct page *page)
 
        page = compound_head(page);
        spin_lock_irq(zone_lru_lock(zone));
-       __activate_page(page, mem_cgroup_page_lruvec(page, zone), NULL);
+       __activate_page(page, mem_cgroup_page_lruvec(page, zone->zone_pgdat), NULL);
        spin_unlock_irq(zone_lru_lock(zone));
 }
 #endif
@@ -445,16 +445,16 @@ void lru_cache_add(struct page *page)
  */
 void add_page_to_unevictable_list(struct page *page)
 {
-       struct zone *zone = page_zone(page);
+       struct pglist_data *pgdat = page_pgdat(page);
        struct lruvec *lruvec;
 
-       spin_lock_irq(zone_lru_lock(zone));
-       lruvec = mem_cgroup_page_lruvec(page, zone);
+       spin_lock_irq(&pgdat->lru_lock);
+       lruvec = mem_cgroup_page_lruvec(page, pgdat);
        ClearPageActive(page);
        SetPageUnevictable(page);
        SetPageLRU(page);
        add_page_to_lru_list(page, lruvec, LRU_UNEVICTABLE);
-       spin_unlock_irq(zone_lru_lock(zone));
+       spin_unlock_irq(&pgdat->lru_lock);
 }
 
 /**
@@ -730,7 +730,7 @@ void release_pages(struct page **pages, int nr, bool cold)
 {
        int i;
        LIST_HEAD(pages_to_free);
-       struct zone *zone = NULL;
+       struct pglist_data *locked_pgdat = NULL;
        struct lruvec *lruvec;
        unsigned long uninitialized_var(flags);
        unsigned int uninitialized_var(lock_batch);
@@ -741,11 +741,11 @@ void release_pages(struct page **pages, int nr, bool cold)
                /*
                 * Make sure the IRQ-safe lock-holding time does not get
                 * excessive with a continuous string of pages from the
-                * same zone. The lock is held only if zone != NULL.
+                * same pgdat. The lock is held only if pgdat != NULL.
                 */
-               if (zone && ++lock_batch == SWAP_CLUSTER_MAX) {
-                       spin_unlock_irqrestore(zone_lru_lock(zone), flags);
-                       zone = NULL;
+               if (locked_pgdat && ++lock_batch == SWAP_CLUSTER_MAX) {
+                       spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags);
+                       locked_pgdat = NULL;
                }
 
                if (is_huge_zero_page(page)) {
@@ -758,27 +758,27 @@ void release_pages(struct page **pages, int nr, bool cold)
                        continue;
 
                if (PageCompound(page)) {
-                       if (zone) {
-                               spin_unlock_irqrestore(zone_lru_lock(zone), flags);
-                               zone = NULL;
+                       if (locked_pgdat) {
+                               spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags);
+                               locked_pgdat = NULL;
                        }
                        __put_compound_page(page);
                        continue;
                }
 
                if (PageLRU(page)) {
-                       struct zone *pagezone = page_zone(page);
+                       struct pglist_data *pgdat = page_pgdat(page);
 
-                       if (pagezone != zone) {
-                               if (zone)
-                                       spin_unlock_irqrestore(zone_lru_lock(zone),
+                       if (pgdat != locked_pgdat) {
+                               if (locked_pgdat)
+                                       spin_unlock_irqrestore(&locked_pgdat->lru_lock,
                                                                        flags);
                                lock_batch = 0;
-                               zone = pagezone;
-                               spin_lock_irqsave(zone_lru_lock(zone), flags);
+                               locked_pgdat = pgdat;
+                               spin_lock_irqsave(&locked_pgdat->lru_lock, flags);
                        }
 
-                       lruvec = mem_cgroup_page_lruvec(page, zone);
+                       lruvec = mem_cgroup_page_lruvec(page, locked_pgdat);
                        VM_BUG_ON_PAGE(!PageLRU(page), page);
                        __ClearPageLRU(page);
                        del_page_from_lru_list(page, lruvec, page_off_lru(page));
@@ -789,8 +789,8 @@ void release_pages(struct page **pages, int nr, bool cold)
 
                list_add(&page->lru, &pages_to_free);
        }
-       if (zone)
-               spin_unlock_irqrestore(zone_lru_lock(zone), flags);
+       if (locked_pgdat)
+               spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags);
 
        mem_cgroup_uncharge_list(&pages_to_free);
        free_hot_cold_page_list(&pages_to_free, cold);
@@ -826,7 +826,7 @@ void lru_add_page_tail(struct page *page, struct page *page_tail,
        VM_BUG_ON_PAGE(PageCompound(page_tail), page);
        VM_BUG_ON_PAGE(PageLRU(page_tail), page);
        VM_BUG_ON(NR_CPUS != 1 &&
-                 !spin_is_locked(zone_lru_lock(lruvec_zone(lruvec))));
+                 !spin_is_locked(&lruvec_pgdat(lruvec)->lru_lock));
 
        if (!list)
                SetPageLRU(page_tail);
index e7ffcd2..86a523a 100644 (file)
@@ -191,26 +191,42 @@ static bool sane_reclaim(struct scan_control *sc)
 }
 #endif
 
+/*
+ * This misses isolated pages which are not accounted for to save counters.
+ * As the data only determines if reclaim or compaction continues, it is
+ * not expected that isolated pages will be a dominating factor.
+ */
 unsigned long zone_reclaimable_pages(struct zone *zone)
 {
        unsigned long nr;
 
-       nr = zone_page_state_snapshot(zone, NR_ACTIVE_FILE) +
-            zone_page_state_snapshot(zone, NR_INACTIVE_FILE) +
-            zone_page_state_snapshot(zone, NR_ISOLATED_FILE);
+       nr = zone_page_state_snapshot(zone, NR_ZONE_LRU_FILE);
+       if (get_nr_swap_pages() > 0)
+               nr += zone_page_state_snapshot(zone, NR_ZONE_LRU_ANON);
+
+       return nr;
+}
+
+unsigned long pgdat_reclaimable_pages(struct pglist_data *pgdat)
+{
+       unsigned long nr;
+
+       nr = node_page_state_snapshot(pgdat, NR_ACTIVE_FILE) +
+            node_page_state_snapshot(pgdat, NR_INACTIVE_FILE) +
+            node_page_state_snapshot(pgdat, NR_ISOLATED_FILE);
 
        if (get_nr_swap_pages() > 0)
-               nr += zone_page_state_snapshot(zone, NR_ACTIVE_ANON) +
-                     zone_page_state_snapshot(zone, NR_INACTIVE_ANON) +
-                     zone_page_state_snapshot(zone, NR_ISOLATED_ANON);
+               nr += node_page_state_snapshot(pgdat, NR_ACTIVE_ANON) +
+                     node_page_state_snapshot(pgdat, NR_INACTIVE_ANON) +
+                     node_page_state_snapshot(pgdat, NR_ISOLATED_ANON);
 
        return nr;
 }
 
-bool zone_reclaimable(struct zone *zone)
+bool pgdat_reclaimable(struct pglist_data *pgdat)
 {
-       return zone_page_state_snapshot(zone, NR_PAGES_SCANNED) <
-               zone_reclaimable_pages(zone) * 6;
+       return node_page_state_snapshot(pgdat, NR_PAGES_SCANNED) <
+               pgdat_reclaimable_pages(pgdat) * 6;
 }
 
 unsigned long lruvec_lru_size(struct lruvec *lruvec, enum lru_list lru)
@@ -218,7 +234,7 @@ unsigned long lruvec_lru_size(struct lruvec *lruvec, enum lru_list lru)
        if (!mem_cgroup_disabled())
                return mem_cgroup_get_lru_size(lruvec, lru);
 
-       return zone_page_state(lruvec_zone(lruvec), NR_LRU_BASE + lru);
+       return node_page_state(lruvec_pgdat(lruvec), NR_LRU_BASE + lru);
 }
 
 /*
@@ -877,7 +893,7 @@ static void page_check_dirty_writeback(struct page *page,
  * shrink_page_list() returns the number of reclaimed pages
  */
 static unsigned long shrink_page_list(struct list_head *page_list,
-                                     struct zone *zone,
+                                     struct pglist_data *pgdat,
                                      struct scan_control *sc,
                                      enum ttu_flags ttu_flags,
                                      unsigned long *ret_nr_dirty,
@@ -917,7 +933,6 @@ static unsigned long shrink_page_list(struct list_head *page_list,
                        goto keep;
 
                VM_BUG_ON_PAGE(PageActive(page), page);
-               VM_BUG_ON_PAGE(page_zone(page) != zone, page);
 
                sc->nr_scanned++;
 
@@ -996,7 +1011,7 @@ static unsigned long shrink_page_list(struct list_head *page_list,
                        /* Case 1 above */
                        if (current_is_kswapd() &&
                            PageReclaim(page) &&
-                           test_bit(ZONE_WRITEBACK, &zone->flags)) {
+                           test_bit(PGDAT_WRITEBACK, &pgdat->flags)) {
                                nr_immediate++;
                                goto keep_locked;
 
@@ -1092,7 +1107,7 @@ static unsigned long shrink_page_list(struct list_head *page_list,
                         */
                        if (page_is_file_cache(page) &&
                                        (!current_is_kswapd() ||
-                                        !test_bit(ZONE_DIRTY, &zone->flags))) {
+                                        !test_bit(PGDAT_DIRTY, &pgdat->flags))) {
                                /*
                                 * Immediately reclaim when written back.
                                 * Similar in principal to deactivate_page()
@@ -1266,11 +1281,11 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,
                }
        }
 
-       ret = shrink_page_list(&clean_pages, zone, &sc,
+       ret = shrink_page_list(&clean_pages, zone->zone_pgdat, &sc,
                        TTU_UNMAP|TTU_IGNORE_ACCESS,
                        &dummy1, &dummy2, &dummy3, &dummy4, &dummy5, true);
        list_splice(&clean_pages, page_list);
-       mod_zone_page_state(zone, NR_ISOLATED_FILE, -ret);
+       mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE, -ret);
        return ret;
 }
 
@@ -1375,7 +1390,8 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 {
        struct list_head *src = &lruvec->lists[lru];
        unsigned long nr_taken = 0;
-       unsigned long scan;
+       unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
+       unsigned long scan, nr_pages;
 
        for (scan = 0; scan < nr_to_scan && nr_taken < nr_to_scan &&
                                        !list_empty(src); scan++) {
@@ -1388,7 +1404,9 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 
                switch (__isolate_lru_page(page, mode)) {
                case 0:
-                       nr_taken += hpage_nr_pages(page);
+                       nr_pages = hpage_nr_pages(page);
+                       nr_taken += nr_pages;
+                       nr_zone_taken[page_zonenum(page)] += nr_pages;
                        list_move(&page->lru, dst);
                        break;
 
@@ -1405,6 +1423,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
        *nr_scanned = scan;
        trace_mm_vmscan_lru_isolate(sc->order, nr_to_scan, scan,
                                    nr_taken, mode, is_file_lru(lru));
+       for (scan = 0; scan < MAX_NR_ZONES; scan++) {
+               nr_pages = nr_zone_taken[scan];
+               if (!nr_pages)
+                       continue;
+
+               update_lru_size(lruvec, lru, scan, -nr_pages);
+       }
        return nr_taken;
 }
 
@@ -1445,7 +1470,7 @@ int isolate_lru_page(struct page *page)
                struct lruvec *lruvec;
 
                spin_lock_irq(zone_lru_lock(zone));
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
                if (PageLRU(page)) {
                        int lru = page_lru(page);
                        get_page(page);
@@ -1465,7 +1490,7 @@ int isolate_lru_page(struct page *page)
  * the LRU list will go small and be scanned faster than necessary, leading to
  * unnecessary swapping, thrashing and OOM.
  */
-static int too_many_isolated(struct zone *zone, int file,
+static int too_many_isolated(struct pglist_data *pgdat, int file,
                struct scan_control *sc)
 {
        unsigned long inactive, isolated;
@@ -1477,11 +1502,11 @@ static int too_many_isolated(struct zone *zone, int file,
                return 0;
 
        if (file) {
-               inactive = zone_page_state(zone, NR_INACTIVE_FILE);
-               isolated = zone_page_state(zone, NR_ISOLATED_FILE);
+               inactive = node_page_state(pgdat, NR_INACTIVE_FILE);
+               isolated = node_page_state(pgdat, NR_ISOLATED_FILE);
        } else {
-               inactive = zone_page_state(zone, NR_INACTIVE_ANON);
-               isolated = zone_page_state(zone, NR_ISOLATED_ANON);
+               inactive = node_page_state(pgdat, NR_INACTIVE_ANON);
+               isolated = node_page_state(pgdat, NR_ISOLATED_ANON);
        }
 
        /*
@@ -1499,7 +1524,7 @@ static noinline_for_stack void
 putback_inactive_pages(struct lruvec *lruvec, struct list_head *page_list)
 {
        struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
-       struct zone *zone = lruvec_zone(lruvec);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
        LIST_HEAD(pages_to_free);
 
        /*
@@ -1512,13 +1537,13 @@ putback_inactive_pages(struct lruvec *lruvec, struct list_head *page_list)
                VM_BUG_ON_PAGE(PageLRU(page), page);
                list_del(&page->lru);
                if (unlikely(!page_evictable(page))) {
-                       spin_unlock_irq(zone_lru_lock(zone));
+                       spin_unlock_irq(&pgdat->lru_lock);
                        putback_lru_page(page);
-                       spin_lock_irq(zone_lru_lock(zone));
+                       spin_lock_irq(&pgdat->lru_lock);
                        continue;
                }
 
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, pgdat);
 
                SetPageLRU(page);
                lru = page_lru(page);
@@ -1535,10 +1560,10 @@ putback_inactive_pages(struct lruvec *lruvec, struct list_head *page_list)
                        del_page_from_lru_list(page, lruvec, lru);
 
                        if (unlikely(PageCompound(page))) {
-                               spin_unlock_irq(zone_lru_lock(zone));
+                               spin_unlock_irq(&pgdat->lru_lock);
                                mem_cgroup_uncharge(page);
                                (*get_compound_page_dtor(page))(page);
-                               spin_lock_irq(zone_lru_lock(zone));
+                               spin_lock_irq(&pgdat->lru_lock);
                        } else
                                list_add(&page->lru, &pages_to_free);
                }
@@ -1582,10 +1607,10 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
        unsigned long nr_immediate = 0;
        isolate_mode_t isolate_mode = 0;
        int file = is_file_lru(lru);
-       struct zone *zone = lruvec_zone(lruvec);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
        struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
 
-       while (unlikely(too_many_isolated(zone, file, sc))) {
+       while (unlikely(too_many_isolated(pgdat, file, sc))) {
                congestion_wait(BLK_RW_ASYNC, HZ/10);
 
                /* We are about to die and free our memory. Return now. */
@@ -1600,48 +1625,45 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
        if (!sc->may_writepage)
                isolate_mode |= ISOLATE_CLEAN;
 
-       spin_lock_irq(zone_lru_lock(zone));
+       spin_lock_irq(&pgdat->lru_lock);
 
        nr_taken = isolate_lru_pages(nr_to_scan, lruvec, &page_list,
                                     &nr_scanned, sc, isolate_mode, lru);
 
-       update_lru_size(lruvec, lru, -nr_taken);
-       __mod_zone_page_state(zone, NR_ISOLATED_ANON + file, nr_taken);
+       __mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, nr_taken);
        reclaim_stat->recent_scanned[file] += nr_taken;
 
        if (global_reclaim(sc)) {
-               __mod_zone_page_state(zone, NR_PAGES_SCANNED, nr_scanned);
+               __mod_node_page_state(pgdat, NR_PAGES_SCANNED, nr_scanned);
                if (current_is_kswapd())
-                       __count_zone_vm_events(PGSCAN_KSWAPD, zone, nr_scanned);
+                       __count_vm_events(PGSCAN_KSWAPD, nr_scanned);
                else
-                       __count_zone_vm_events(PGSCAN_DIRECT, zone, nr_scanned);
+                       __count_vm_events(PGSCAN_DIRECT, nr_scanned);
        }
-       spin_unlock_irq(zone_lru_lock(zone));
+       spin_unlock_irq(&pgdat->lru_lock);
 
        if (nr_taken == 0)
                return 0;
 
-       nr_reclaimed = shrink_page_list(&page_list, zone, sc, TTU_UNMAP,
+       nr_reclaimed = shrink_page_list(&page_list, pgdat, sc, TTU_UNMAP,
                                &nr_dirty, &nr_unqueued_dirty, &nr_congested,
                                &nr_writeback, &nr_immediate,
                                false);
 
-       spin_lock_irq(zone_lru_lock(zone));
+       spin_lock_irq(&pgdat->lru_lock);
 
        if (global_reclaim(sc)) {
                if (current_is_kswapd())
-                       __count_zone_vm_events(PGSTEAL_KSWAPD, zone,
-                                              nr_reclaimed);
+                       __count_vm_events(PGSTEAL_KSWAPD, nr_reclaimed);
                else
-                       __count_zone_vm_events(PGSTEAL_DIRECT, zone,
-                                              nr_reclaimed);
+                       __count_vm_events(PGSTEAL_DIRECT, nr_reclaimed);
        }
 
        putback_inactive_pages(lruvec, &page_list);
 
-       __mod_zone_page_state(zone, NR_ISOLATED_ANON + file, -nr_taken);
+       __mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, -nr_taken);
 
-       spin_unlock_irq(zone_lru_lock(zone));
+       spin_unlock_irq(&pgdat->lru_lock);
 
        mem_cgroup_uncharge_list(&page_list);
        free_hot_cold_page_list(&page_list, true);
@@ -1661,7 +1683,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
         * are encountered in the nr_immediate check below.
         */
        if (nr_writeback && nr_writeback == nr_taken)
-               set_bit(ZONE_WRITEBACK, &zone->flags);
+               set_bit(PGDAT_WRITEBACK, &pgdat->flags);
 
        /*
         * Legacy memcg will stall in page writeback so avoid forcibly
@@ -1673,16 +1695,16 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
                 * backed by a congested BDI and wait_iff_congested will stall.
                 */
                if (nr_dirty && nr_dirty == nr_congested)
-                       set_bit(ZONE_CONGESTED, &zone->flags);
+                       set_bit(PGDAT_CONGESTED, &pgdat->flags);
 
                /*
                 * If dirty pages are scanned that are not queued for IO, it
                 * implies that flushers are not keeping up. In this case, flag
-                * the zone ZONE_DIRTY and kswapd will start writing pages from
+                * the pgdat PGDAT_DIRTY and kswapd will start writing pages from
                 * reclaim context.
                 */
                if (nr_unqueued_dirty == nr_taken)
-                       set_bit(ZONE_DIRTY, &zone->flags);
+                       set_bit(PGDAT_DIRTY, &pgdat->flags);
 
                /*
                 * If kswapd scans pages marked marked for immediate
@@ -1701,9 +1723,10 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
         */
        if (!sc->hibernation_mode && !current_is_kswapd() &&
            current_may_throttle())
-               wait_iff_congested(zone, BLK_RW_ASYNC, HZ/10);
+               wait_iff_congested(pgdat, BLK_RW_ASYNC, HZ/10);
 
-       trace_mm_vmscan_lru_shrink_inactive(zone, nr_scanned, nr_reclaimed,
+       trace_mm_vmscan_lru_shrink_inactive(pgdat->node_id,
+                       nr_scanned, nr_reclaimed,
                        sc->priority, file);
        return nr_reclaimed;
 }
@@ -1731,20 +1754,20 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
                                     struct list_head *pages_to_free,
                                     enum lru_list lru)
 {
-       struct zone *zone = lruvec_zone(lruvec);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
        unsigned long pgmoved = 0;
        struct page *page;
        int nr_pages;
 
        while (!list_empty(list)) {
                page = lru_to_page(list);
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, pgdat);
 
                VM_BUG_ON_PAGE(PageLRU(page), page);
                SetPageLRU(page);
 
                nr_pages = hpage_nr_pages(page);
-               update_lru_size(lruvec, lru, nr_pages);
+               update_lru_size(lruvec, lru, page_zonenum(page), nr_pages);
                list_move(&page->lru, &lruvec->lists[lru]);
                pgmoved += nr_pages;
 
@@ -1754,10 +1777,10 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
                        del_page_from_lru_list(page, lruvec, lru);
 
                        if (unlikely(PageCompound(page))) {
-                               spin_unlock_irq(zone_lru_lock(zone));
+                               spin_unlock_irq(&pgdat->lru_lock);
                                mem_cgroup_uncharge(page);
                                (*get_compound_page_dtor(page))(page);
-                               spin_lock_irq(zone_lru_lock(zone));
+                               spin_lock_irq(&pgdat->lru_lock);
                        } else
                                list_add(&page->lru, pages_to_free);
                }
@@ -1783,7 +1806,7 @@ static void shrink_active_list(unsigned long nr_to_scan,
        unsigned long nr_rotated = 0;
        isolate_mode_t isolate_mode = 0;
        int file = is_file_lru(lru);
-       struct zone *zone = lruvec_zone(lruvec);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
 
        lru_add_drain();
 
@@ -1792,20 +1815,19 @@ static void shrink_active_list(unsigned long nr_to_scan,
        if (!sc->may_writepage)
                isolate_mode |= ISOLATE_CLEAN;
 
-       spin_lock_irq(zone_lru_lock(zone));
+       spin_lock_irq(&pgdat->lru_lock);
 
        nr_taken = isolate_lru_pages(nr_to_scan, lruvec, &l_hold,
                                     &nr_scanned, sc, isolate_mode, lru);
 
-       update_lru_size(lruvec, lru, -nr_taken);
-       __mod_zone_page_state(zone, NR_ISOLATED_ANON + file, nr_taken);
+       __mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, nr_taken);
        reclaim_stat->recent_scanned[file] += nr_taken;
 
        if (global_reclaim(sc))
-               __mod_zone_page_state(zone, NR_PAGES_SCANNED, nr_scanned);
-       __count_zone_vm_events(PGREFILL, zone, nr_scanned);
+               __mod_node_page_state(pgdat, NR_PAGES_SCANNED, nr_scanned);
+       __count_vm_events(PGREFILL, nr_scanned);
 
-       spin_unlock_irq(zone_lru_lock(zone));
+       spin_unlock_irq(&pgdat->lru_lock);
 
        while (!list_empty(&l_hold)) {
                cond_resched();
@@ -1850,7 +1872,7 @@ static void shrink_active_list(unsigned long nr_to_scan,
        /*
         * Move pages back to the lru list.
         */
-       spin_lock_irq(zone_lru_lock(zone));
+       spin_lock_irq(&pgdat->lru_lock);
        /*
         * Count referenced pages from currently used mappings as rotated,
         * even though only some of them are actually re-activated.  This
@@ -1861,8 +1883,8 @@ static void shrink_active_list(unsigned long nr_to_scan,
 
        move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
        move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
-       __mod_zone_page_state(zone, NR_ISOLATED_ANON + file, -nr_taken);
-       spin_unlock_irq(zone_lru_lock(zone));
+       __mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, -nr_taken);
+       spin_unlock_irq(&pgdat->lru_lock);
 
        mem_cgroup_uncharge_list(&l_hold);
        free_hot_cold_page_list(&l_hold, true);
@@ -1956,7 +1978,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
        struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
        u64 fraction[2];
        u64 denominator = 0;    /* gcc */
-       struct zone *zone = lruvec_zone(lruvec);
+       struct pglist_data *pgdat = lruvec_pgdat(lruvec);
        unsigned long anon_prio, file_prio;
        enum scan_balance scan_balance;
        unsigned long anon, file;
@@ -1977,7 +1999,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
         * well.
         */
        if (current_is_kswapd()) {
-               if (!zone_reclaimable(zone))
+               if (!pgdat_reclaimable(pgdat))
                        force_scan = true;
                if (!mem_cgroup_online(memcg))
                        force_scan = true;
@@ -2023,14 +2045,24 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
         * anon pages.  Try to detect this based on file LRU size.
         */
        if (global_reclaim(sc)) {
-               unsigned long zonefile;
-               unsigned long zonefree;
+               unsigned long pgdatfile;
+               unsigned long pgdatfree;
+               int z;
+               unsigned long total_high_wmark = 0;
 
-               zonefree = zone_page_state(zone, NR_FREE_PAGES);
-               zonefile = zone_page_state(zone, NR_ACTIVE_FILE) +
-                          zone_page_state(zone, NR_INACTIVE_FILE);
+               pgdatfree = sum_zone_node_page_state(pgdat->node_id, NR_FREE_PAGES);
+               pgdatfile = node_page_state(pgdat, NR_ACTIVE_FILE) +
+                          node_page_state(pgdat, NR_INACTIVE_FILE);
+
+               for (z = 0; z < MAX_NR_ZONES; z++) {
+                       struct zone *zone = &pgdat->node_zones[z];
+                       if (!populated_zone(zone))
+                               continue;
+
+                       total_high_wmark += high_wmark_pages(zone);
+               }
 
-               if (unlikely(zonefile + zonefree <= high_wmark_pages(zone))) {
+               if (unlikely(pgdatfile + pgdatfree <= total_high_wmark)) {
                        scan_balance = SCAN_ANON;
                        goto out;
                }
@@ -2077,7 +2109,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
        file  = lruvec_lru_size(lruvec, LRU_ACTIVE_FILE) +
                lruvec_lru_size(lruvec, LRU_INACTIVE_FILE);
 
-       spin_lock_irq(zone_lru_lock(zone));
+       spin_lock_irq(&pgdat->lru_lock);
        if (unlikely(reclaim_stat->recent_scanned[0] > anon / 4)) {
                reclaim_stat->recent_scanned[0] /= 2;
                reclaim_stat->recent_rotated[0] /= 2;
@@ -2098,7 +2130,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
 
        fp = file_prio * (reclaim_stat->recent_scanned[1] + 1);
        fp /= reclaim_stat->recent_rotated[1] + 1;
-       spin_unlock_irq(zone_lru_lock(zone));
+       spin_unlock_irq(&pgdat->lru_lock);
 
        fraction[0] = ap;
        fraction[1] = fp;
@@ -2352,9 +2384,9 @@ static inline bool should_continue_reclaim(struct zone *zone,
         * inactive lists are large enough, continue reclaiming
         */
        pages_for_compaction = (2UL << sc->order);
-       inactive_lru_pages = zone_page_state(zone, NR_INACTIVE_FILE);
+       inactive_lru_pages = node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE);
        if (get_nr_swap_pages() > 0)
-               inactive_lru_pages += zone_page_state(zone, NR_INACTIVE_ANON);
+               inactive_lru_pages += node_page_state(zone->zone_pgdat, NR_INACTIVE_ANON);
        if (sc->nr_reclaimed < pages_for_compaction &&
                        inactive_lru_pages > pages_for_compaction)
                return true;
@@ -2554,7 +2586,7 @@ static void shrink_zones(struct zonelist *zonelist, struct scan_control *sc)
                                continue;
 
                        if (sc->priority != DEF_PRIORITY &&
-                           !zone_reclaimable(zone))
+                           !pgdat_reclaimable(zone->zone_pgdat))
                                continue;       /* Let kswapd poll it */
 
                        /*
@@ -2692,7 +2724,7 @@ static bool pfmemalloc_watermark_ok(pg_data_t *pgdat)
        for (i = 0; i <= ZONE_NORMAL; i++) {
                zone = &pgdat->node_zones[i];
                if (!populated_zone(zone) ||
-                   zone_reclaimable_pages(zone) == 0)
+                   pgdat_reclaimable_pages(pgdat) == 0)
                        continue;
 
                pfmemalloc_reserve += min_wmark_pages(zone);
@@ -3000,7 +3032,7 @@ static bool pgdat_balanced(pg_data_t *pgdat, int order, int classzone_idx)
                 * DEF_PRIORITY. Effectively, it considers them balanced so
                 * they must be considered balanced here as well!
                 */
-               if (!zone_reclaimable(zone)) {
+               if (!pgdat_reclaimable(zone->zone_pgdat)) {
                        balanced_pages += zone->managed_pages;
                        continue;
                }
@@ -3063,6 +3095,7 @@ static bool kswapd_shrink_zone(struct zone *zone,
 {
        unsigned long balance_gap;
        bool lowmem_pressure;
+       struct pglist_data *pgdat = zone->zone_pgdat;
 
        /* Reclaim above the high watermark. */
        sc->nr_to_reclaim = max(SWAP_CLUSTER_MAX, high_wmark_pages(zone));
@@ -3087,7 +3120,8 @@ static bool kswapd_shrink_zone(struct zone *zone,
 
        shrink_zone(zone, sc, zone_idx(zone) == classzone_idx);
 
-       clear_bit(ZONE_WRITEBACK, &zone->flags);
+       /* TODO: ANOMALY */
+       clear_bit(PGDAT_WRITEBACK, &pgdat->flags);
 
        /*
         * If a zone reaches its high watermark, consider it to be no longer
@@ -3095,10 +3129,10 @@ static bool kswapd_shrink_zone(struct zone *zone,
         * BDIs but as pressure is relieved, speculatively avoid congestion
         * waits.
         */
-       if (zone_reclaimable(zone) &&
+       if (pgdat_reclaimable(zone->zone_pgdat) &&
            zone_balanced(zone, sc->order, false, 0, classzone_idx)) {
-               clear_bit(ZONE_CONGESTED, &zone->flags);
-               clear_bit(ZONE_DIRTY, &zone->flags);
+               clear_bit(PGDAT_CONGESTED, &pgdat->flags);
+               clear_bit(PGDAT_DIRTY, &pgdat->flags);
        }
 
        return sc->nr_scanned >= sc->nr_to_reclaim;
@@ -3157,7 +3191,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
                                continue;
 
                        if (sc.priority != DEF_PRIORITY &&
-                           !zone_reclaimable(zone))
+                           !pgdat_reclaimable(zone->zone_pgdat))
                                continue;
 
                        /*
@@ -3184,9 +3218,11 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
                                /*
                                 * If balanced, clear the dirty and congested
                                 * flags
+                                *
+                                * TODO: ANOMALY
                                 */
-                               clear_bit(ZONE_CONGESTED, &zone->flags);
-                               clear_bit(ZONE_DIRTY, &zone->flags);
+                               clear_bit(PGDAT_CONGESTED, &zone->zone_pgdat->flags);
+                               clear_bit(PGDAT_DIRTY, &zone->zone_pgdat->flags);
                        }
                }
 
@@ -3216,7 +3252,7 @@ static int balance_pgdat(pg_data_t *pgdat, int order, int classzone_idx)
                                continue;
 
                        if (sc.priority != DEF_PRIORITY &&
-                           !zone_reclaimable(zone))
+                           !pgdat_reclaimable(zone->zone_pgdat))
                                continue;
 
                        sc.nr_scanned = 0;
@@ -3612,8 +3648,8 @@ int sysctl_min_slab_ratio = 5;
 static inline unsigned long zone_unmapped_file_pages(struct zone *zone)
 {
        unsigned long file_mapped = zone_page_state(zone, NR_FILE_MAPPED);
-       unsigned long file_lru = zone_page_state(zone, NR_INACTIVE_FILE) +
-               zone_page_state(zone, NR_ACTIVE_FILE);
+       unsigned long file_lru = node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE) +
+               node_page_state(zone->zone_pgdat, NR_ACTIVE_FILE);
 
        /*
         * It's possible for there to be more file mapped pages than
@@ -3716,7 +3752,7 @@ int zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
            zone_page_state(zone, NR_SLAB_RECLAIMABLE) <= zone->min_slab_pages)
                return ZONE_RECLAIM_FULL;
 
-       if (!zone_reclaimable(zone))
+       if (!pgdat_reclaimable(zone->zone_pgdat))
                return ZONE_RECLAIM_FULL;
 
        /*
@@ -3795,7 +3831,7 @@ void check_move_unevictable_pages(struct page **pages, int nr_pages)
                        zone = pagezone;
                        spin_lock_irq(zone_lru_lock(zone));
                }
-               lruvec = mem_cgroup_page_lruvec(page, zone);
+               lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat);
 
                if (!PageLRU(page) || !PageUnevictable(page))
                        continue;
index 3345d39..de0c170 100644 (file)
@@ -936,11 +936,8 @@ const char * const vmstat_text[] = {
        /* enum zone_stat_item countes */
        "nr_free_pages",
        "nr_alloc_batch",
-       "nr_inactive_anon",
-       "nr_active_anon",
-       "nr_inactive_file",
-       "nr_active_file",
-       "nr_unevictable",
+       "nr_zone_anon_lru",
+       "nr_zone_file_lru",
        "nr_mlock",
        "nr_anon_pages",
        "nr_mapped",
@@ -956,12 +953,9 @@ const char * const vmstat_text[] = {
        "nr_vmscan_write",
        "nr_vmscan_immediate_reclaim",
        "nr_writeback_temp",
-       "nr_isolated_anon",
-       "nr_isolated_file",
        "nr_shmem",
        "nr_dirtied",
        "nr_written",
-       "nr_pages_scanned",
 #if IS_ENABLED(CONFIG_ZSMALLOC)
        "nr_zspages",
 #endif
@@ -981,6 +975,16 @@ const char * const vmstat_text[] = {
        "nr_shmem_pmdmapped",
        "nr_free_cma",
 
+       /* Node-based counters */
+       "nr_inactive_anon",
+       "nr_active_anon",
+       "nr_inactive_file",
+       "nr_active_file",
+       "nr_unevictable",
+       "nr_isolated_anon",
+       "nr_isolated_file",
+       "nr_pages_scanned",
+
        /* enum writeback_stat_item counters */
        "nr_dirty_threshold",
        "nr_dirty_background_threshold",
@@ -1002,11 +1006,11 @@ const char * const vmstat_text[] = {
        "pgmajfault",
        "pglazyfreed",
 
-       TEXTS_FOR_ZONES("pgrefill")
-       TEXTS_FOR_ZONES("pgsteal_kswapd")
-       TEXTS_FOR_ZONES("pgsteal_direct")
-       TEXTS_FOR_ZONES("pgscan_kswapd")
-       TEXTS_FOR_ZONES("pgscan_direct")
+       "pgrefill",
+       "pgsteal_kswapd",
+       "pgsteal_direct",
+       "pgscan_kswapd",
+       "pgscan_direct",
        "pgscan_direct_throttle",
 
 #ifdef CONFIG_NUMA
@@ -1434,7 +1438,7 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
                   "\n        min      %lu"
                   "\n        low      %lu"
                   "\n        high     %lu"
-                  "\n        scanned  %lu"
+                  "\n   node_scanned  %lu"
                   "\n        spanned  %lu"
                   "\n        present  %lu"
                   "\n        managed  %lu",
@@ -1442,13 +1446,13 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
                   min_wmark_pages(zone),
                   low_wmark_pages(zone),
                   high_wmark_pages(zone),
-                  zone_page_state(zone, NR_PAGES_SCANNED),
+                  node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED),
                   zone->spanned_pages,
                   zone->present_pages,
                   zone->managed_pages);
 
        for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
-               seq_printf(m, "\n    %-12s %lu", vmstat_text[i],
+               seq_printf(m, "\n      %-12s %lu", vmstat_text[i],
                                zone_page_state(zone, i));
 
        seq_printf(m,
@@ -1478,12 +1482,12 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
 #endif
        }
        seq_printf(m,
-                  "\n  all_unreclaimable: %u"
-                  "\n  start_pfn:         %lu"
-                  "\n  inactive_ratio:    %u",
-                  !zone_reclaimable(zone),
+                  "\n  node_unreclaimable:  %u"
+                  "\n  start_pfn:           %lu"
+                  "\n  node_inactive_ratio: %u",
+                  !pgdat_reclaimable(zone->zone_pgdat),
                   zone->zone_start_pfn,
-                  zone->inactive_ratio);
+                  zone->zone_pgdat->inactive_ratio);
        seq_putc(m, '\n');
 }
 
@@ -1574,7 +1578,6 @@ static int vmstat_show(struct seq_file *m, void *arg)
 {
        unsigned long *l = arg;
        unsigned long off = l - (unsigned long *)m->private;
-
        seq_printf(m, "%s %lu\n", vmstat_text[off], *l);
        return 0;
 }
index 5ffba0c..7820a7e 100644 (file)
@@ -355,8 +355,8 @@ static unsigned long count_shadow_nodes(struct shrinker *shrinker,
                pages = mem_cgroup_node_nr_lru_pages(sc->memcg, sc->nid,
                                                     LRU_ALL_FILE);
        } else {
-               pages = sum_zone_node_page_state(sc->nid, NR_ACTIVE_FILE) +
-                       sum_zone_node_page_state(sc->nid, NR_INACTIVE_FILE);
+               pages = node_page_state(NODE_DATA(sc->nid), NR_ACTIVE_FILE) +
+                       node_page_state(NODE_DATA(sc->nid), NR_INACTIVE_FILE);
        }
 
        /*