mm: move most file-based accounting to the node
authorMel Gorman <mgorman@techsingularity.net>
Thu, 28 Jul 2016 22:46:20 +0000 (15:46 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 28 Jul 2016 23:07:41 +0000 (16:07 -0700)
There are now a number of accounting oddities such as mapped file pages
being accounted for on the node while the total number of file pages are
accounted on the zone.  This can be coped with to some extent but it's
confusing so this patch moves the relevant file-based accounted.  Due to
throttling logic in the page allocator for reliable OOM detection, it is
still necessary to track dirty and writeback pages on a per-zone basis.

[mgorman@techsingularity.net: fix NR_ZONE_WRITE_PENDING accounting]
Link: http://lkml.kernel.org/r/1468404004-5085-5-git-send-email-mgorman@techsingularity.net
Link: http://lkml.kernel.org/r/1467970510-21195-20-git-send-email-mgorman@techsingularity.net
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
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>
24 files changed:
arch/s390/appldata/appldata_mem.c
arch/tile/mm/pgtable.c
drivers/base/node.c
drivers/staging/android/lowmemorykiller.c
drivers/staging/lustre/lustre/osc/osc_cache.c
fs/fs-writeback.c
fs/fuse/file.c
fs/nfs/internal.h
fs/nfs/write.c
fs/proc/meminfo.c
include/linux/mmzone.h
include/trace/events/writeback.h
mm/filemap.c
mm/huge_memory.c
mm/khugepaged.c
mm/migrate.c
mm/page-writeback.c
mm/page_alloc.c
mm/rmap.c
mm/shmem.c
mm/swap_state.c
mm/util.c
mm/vmscan.c
mm/vmstat.c

index edcf2a7..598df57 100644 (file)
@@ -102,7 +102,7 @@ static void appldata_get_mem_data(void *data)
        mem_data->totalhigh = P2K(val.totalhigh);
        mem_data->freehigh  = P2K(val.freehigh);
        mem_data->bufferram = P2K(val.bufferram);
-       mem_data->cached    = P2K(global_page_state(NR_FILE_PAGES)
+       mem_data->cached    = P2K(global_node_page_state(NR_FILE_PAGES)
                                - val.bufferram);
 
        si_swapinfo(&val);
index c606b0e..7cc6ee7 100644 (file)
@@ -49,16 +49,16 @@ void show_mem(unsigned int filter)
                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),
+              global_node_page_state(NR_FILE_DIRTY),
+              global_node_page_state(NR_WRITEBACK),
+              global_node_page_state(NR_UNSTABLE_NFS),
               global_page_state(NR_FREE_PAGES),
               (global_page_state(NR_SLAB_RECLAIMABLE) +
                global_page_state(NR_SLAB_UNRECLAIMABLE)),
               global_node_page_state(NR_FILE_MAPPED),
               global_page_state(NR_PAGETABLE),
               global_page_state(NR_BOUNCE),
-              global_page_state(NR_FILE_PAGES),
+              global_node_page_state(NR_FILE_PAGES),
               get_nr_swap_pages());
 
        for_each_zone(zone) {
index 6cd9ff4..264cc21 100644 (file)
@@ -118,28 +118,28 @@ static ssize_t node_read_meminfo(struct device *dev,
                       "Node %d ShmemPmdMapped: %8lu kB\n"
 #endif
                        ,
-                      nid, K(sum_zone_node_page_state(nid, NR_FILE_DIRTY)),
-                      nid, K(sum_zone_node_page_state(nid, NR_WRITEBACK)),
-                      nid, K(sum_zone_node_page_state(nid, NR_FILE_PAGES)),
+                      nid, K(node_page_state(pgdat, NR_FILE_DIRTY)),
+                      nid, K(node_page_state(pgdat, NR_WRITEBACK)),
+                      nid, K(node_page_state(pgdat, NR_FILE_PAGES)),
                       nid, K(node_page_state(pgdat, NR_FILE_MAPPED)),
                       nid, K(node_page_state(pgdat, NR_ANON_MAPPED)),
                       nid, K(i.sharedram),
                       nid, sum_zone_node_page_state(nid, NR_KERNEL_STACK) *
                                THREAD_SIZE / 1024,
                       nid, K(sum_zone_node_page_state(nid, NR_PAGETABLE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_UNSTABLE_NFS)),
+                      nid, K(node_page_state(pgdat, NR_UNSTABLE_NFS)),
                       nid, K(sum_zone_node_page_state(nid, NR_BOUNCE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_WRITEBACK_TEMP)),
+                      nid, K(node_page_state(pgdat, NR_WRITEBACK_TEMP)),
                       nid, K(sum_zone_node_page_state(nid, NR_SLAB_RECLAIMABLE) +
                                sum_zone_node_page_state(nid, NR_SLAB_UNRECLAIMABLE)),
                       nid, K(sum_zone_node_page_state(nid, NR_SLAB_RECLAIMABLE)),
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
                       nid, K(sum_zone_node_page_state(nid, NR_SLAB_UNRECLAIMABLE)),
-                      nid, K(sum_zone_node_page_state(nid, NR_ANON_THPS) *
+                      nid, K(node_page_state(pgdat, NR_ANON_THPS) *
                                       HPAGE_PMD_NR),
-                      nid, K(sum_zone_node_page_state(nid, NR_SHMEM_THPS) *
+                      nid, K(node_page_state(pgdat, NR_SHMEM_THPS) *
                                       HPAGE_PMD_NR),
-                      nid, K(sum_zone_node_page_state(nid, NR_SHMEM_PMDMAPPED) *
+                      nid, K(node_page_state(pgdat, NR_SHMEM_PMDMAPPED) *
                                       HPAGE_PMD_NR));
 #else
                       nid, K(sum_zone_node_page_state(nid, NR_SLAB_UNRECLAIMABLE)));
index 93dbcc3..45a1b4e 100644 (file)
@@ -91,8 +91,8 @@ static unsigned long lowmem_scan(struct shrinker *s, struct shrink_control *sc)
        short selected_oom_score_adj;
        int array_size = ARRAY_SIZE(lowmem_adj);
        int other_free = global_page_state(NR_FREE_PAGES) - totalreserve_pages;
-       int other_file = global_page_state(NR_FILE_PAGES) -
-                                               global_page_state(NR_SHMEM) -
+       int other_file = global_node_page_state(NR_FILE_PAGES) -
+                                               global_node_page_state(NR_SHMEM) -
                                                total_swapcache_pages();
 
        if (lowmem_adj_size < array_size)
index d1a7d6b..d011135 100644 (file)
@@ -1864,7 +1864,8 @@ void osc_dec_unstable_pages(struct ptlrpc_request *req)
        LASSERT(page_count >= 0);
 
        for (i = 0; i < page_count; i++)
-               dec_zone_page_state(desc->bd_iov[i].kiov_page, NR_UNSTABLE_NFS);
+               dec_node_page_state(desc->bd_iov[i].kiov_page,
+                                                       NR_UNSTABLE_NFS);
 
        atomic_sub(page_count, &cli->cl_cache->ccc_unstable_nr);
        LASSERT(atomic_read(&cli->cl_cache->ccc_unstable_nr) >= 0);
@@ -1898,7 +1899,8 @@ void osc_inc_unstable_pages(struct ptlrpc_request *req)
        LASSERT(page_count >= 0);
 
        for (i = 0; i < page_count; i++)
-               inc_zone_page_state(desc->bd_iov[i].kiov_page, NR_UNSTABLE_NFS);
+               inc_node_page_state(desc->bd_iov[i].kiov_page,
+                                                       NR_UNSTABLE_NFS);
 
        LASSERT(atomic_read(&cli->cl_cache->ccc_unstable_nr) >= 0);
        atomic_add(page_count, &cli->cl_cache->ccc_unstable_nr);
index 6f9c9f6..56c8fda 100644 (file)
@@ -1807,8 +1807,8 @@ static struct wb_writeback_work *get_next_work_item(struct bdi_writeback *wb)
  */
 static unsigned long get_nr_dirty_pages(void)
 {
-       return global_page_state(NR_FILE_DIRTY) +
-               global_page_state(NR_UNSTABLE_NFS) +
+       return global_node_page_state(NR_FILE_DIRTY) +
+               global_node_page_state(NR_UNSTABLE_NFS) +
                get_nr_dirty_inodes();
 }
 
index 9154f86..2382f22 100644 (file)
@@ -1452,7 +1452,7 @@ static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req)
        list_del(&req->writepages_entry);
        for (i = 0; i < req->num_pages; i++) {
                dec_wb_stat(&bdi->wb, WB_WRITEBACK);
-               dec_zone_page_state(req->pages[i], NR_WRITEBACK_TEMP);
+               dec_node_page_state(req->pages[i], NR_WRITEBACK_TEMP);
                wb_writeout_inc(&bdi->wb);
        }
        wake_up(&fi->page_waitq);
@@ -1642,7 +1642,7 @@ static int fuse_writepage_locked(struct page *page)
        req->inode = inode;
 
        inc_wb_stat(&inode_to_bdi(inode)->wb, WB_WRITEBACK);
-       inc_zone_page_state(tmp_page, NR_WRITEBACK_TEMP);
+       inc_node_page_state(tmp_page, NR_WRITEBACK_TEMP);
 
        spin_lock(&fc->lock);
        list_add(&req->writepages_entry, &fi->writepages);
@@ -1756,7 +1756,7 @@ static bool fuse_writepage_in_flight(struct fuse_req *new_req,
                spin_unlock(&fc->lock);
 
                dec_wb_stat(&bdi->wb, WB_WRITEBACK);
-               dec_zone_page_state(page, NR_WRITEBACK_TEMP);
+               dec_node_page_state(page, NR_WRITEBACK_TEMP);
                wb_writeout_inc(&bdi->wb);
                fuse_writepage_free(fc, new_req);
                fuse_request_free(new_req);
@@ -1855,7 +1855,7 @@ static int fuse_writepages_fill(struct page *page,
        req->page_descs[req->num_pages].length = PAGE_SIZE;
 
        inc_wb_stat(&inode_to_bdi(inode)->wb, WB_WRITEBACK);
-       inc_zone_page_state(tmp_page, NR_WRITEBACK_TEMP);
+       inc_node_page_state(tmp_page, NR_WRITEBACK_TEMP);
 
        err = 0;
        if (is_writeback && fuse_writepage_in_flight(req, page)) {
index 5154fa6..5ea04d8 100644 (file)
@@ -623,7 +623,7 @@ void nfs_mark_page_unstable(struct page *page, struct nfs_commit_info *cinfo)
        if (!cinfo->dreq) {
                struct inode *inode = page_file_mapping(page)->host;
 
-               inc_zone_page_state(page, NR_UNSTABLE_NFS);
+               inc_node_page_state(page, NR_UNSTABLE_NFS);
                inc_wb_stat(&inode_to_bdi(inode)->wb, WB_RECLAIMABLE);
                __mark_inode_dirty(inode, I_DIRTY_DATASYNC);
        }
index e1c74d3..593fa21 100644 (file)
@@ -898,7 +898,7 @@ nfs_mark_request_commit(struct nfs_page *req, struct pnfs_layout_segment *lseg,
 static void
 nfs_clear_page_commit(struct page *page)
 {
-       dec_zone_page_state(page, NR_UNSTABLE_NFS);
+       dec_node_page_state(page, NR_UNSTABLE_NFS);
        dec_wb_stat(&inode_to_bdi(page_file_mapping(page)->host)->wb,
                    WB_RECLAIMABLE);
 }
index 40f1087..c1fdcc1 100644 (file)
@@ -40,7 +40,7 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
        si_swapinfo(&i);
        committed = percpu_counter_read_positive(&vm_committed_as);
 
-       cached = global_page_state(NR_FILE_PAGES) -
+       cached = global_node_page_state(NR_FILE_PAGES) -
                        total_swapcache_pages() - i.bufferram;
        if (cached < 0)
                cached = 0;
@@ -138,8 +138,8 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
 #endif
                K(i.totalswap),
                K(i.freeswap),
-               K(global_page_state(NR_FILE_DIRTY)),
-               K(global_page_state(NR_WRITEBACK)),
+               K(global_node_page_state(NR_FILE_DIRTY)),
+               K(global_node_page_state(NR_WRITEBACK)),
                K(global_node_page_state(NR_ANON_MAPPED)),
                K(global_node_page_state(NR_FILE_MAPPED)),
                K(i.sharedram),
@@ -152,9 +152,9 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
 #ifdef CONFIG_QUICKLIST
                K(quicklist_total_size()),
 #endif
-               K(global_page_state(NR_UNSTABLE_NFS)),
+               K(global_node_page_state(NR_UNSTABLE_NFS)),
                K(global_page_state(NR_BOUNCE)),
-               K(global_page_state(NR_WRITEBACK_TEMP)),
+               K(global_node_page_state(NR_WRITEBACK_TEMP)),
                K(vm_commit_limit()),
                K(committed),
                (unsigned long)VMALLOC_TOTAL >> 10,
@@ -164,9 +164,9 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
                , atomic_long_read(&num_poisoned_pages) << (PAGE_SHIFT - 10)
 #endif
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
-               , K(global_page_state(NR_ANON_THPS) * HPAGE_PMD_NR)
-               , K(global_page_state(NR_SHMEM_THPS) * HPAGE_PMD_NR)
-               , K(global_page_state(NR_SHMEM_PMDMAPPED) * HPAGE_PMD_NR)
+               , K(global_node_page_state(NR_ANON_THPS) * HPAGE_PMD_NR)
+               , K(global_node_page_state(NR_SHMEM_THPS) * HPAGE_PMD_NR)
+               , K(global_node_page_state(NR_SHMEM_PMDMAPPED) * HPAGE_PMD_NR)
 #endif
 #ifdef CONFIG_CMA
                , K(totalcma_pages)
index 2d4a880..acd4665 100644 (file)
@@ -114,21 +114,16 @@ enum zone_stat_item {
        NR_ZONE_LRU_BASE, /* Used only for compaction and reclaim retry */
        NR_ZONE_LRU_ANON = NR_ZONE_LRU_BASE,
        NR_ZONE_LRU_FILE,
+       NR_ZONE_WRITE_PENDING,  /* Count of dirty, writeback and unstable pages */
        NR_MLOCK,               /* mlock()ed pages found and moved off LRU */
-       NR_FILE_PAGES,
-       NR_FILE_DIRTY,
-       NR_WRITEBACK,
        NR_SLAB_RECLAIMABLE,
        NR_SLAB_UNRECLAIMABLE,
        NR_PAGETABLE,           /* used for pagetables */
        NR_KERNEL_STACK,
        /* Second 128 byte cacheline */
-       NR_UNSTABLE_NFS,        /* NFS unstable pages */
        NR_BOUNCE,
        NR_VMSCAN_WRITE,
        NR_VMSCAN_IMMEDIATE,    /* Prioritise for reclaim when writeback ends */
-       NR_WRITEBACK_TEMP,      /* Writeback using temporary buffers */
-       NR_SHMEM,               /* shmem pages (included tmpfs/GEM pages) */
        NR_DIRTIED,             /* page dirtyings since bootup */
        NR_WRITTEN,             /* page writings since bootup */
 #if IS_ENABLED(CONFIG_ZSMALLOC)
@@ -142,9 +137,6 @@ enum zone_stat_item {
        NUMA_LOCAL,             /* allocation from local node */
        NUMA_OTHER,             /* allocation from other node */
 #endif
-       NR_ANON_THPS,
-       NR_SHMEM_THPS,
-       NR_SHMEM_PMDMAPPED,
        NR_FREE_CMA_PAGES,
        NR_VM_ZONE_STAT_ITEMS };
 
@@ -164,6 +156,15 @@ enum node_stat_item {
        NR_ANON_MAPPED, /* Mapped anonymous pages */
        NR_FILE_MAPPED, /* pagecache pages mapped into pagetables.
                           only modified from process context */
+       NR_FILE_PAGES,
+       NR_FILE_DIRTY,
+       NR_WRITEBACK,
+       NR_WRITEBACK_TEMP,      /* Writeback using temporary buffers */
+       NR_SHMEM,               /* shmem pages (included tmpfs/GEM pages) */
+       NR_SHMEM_THPS,
+       NR_SHMEM_PMDMAPPED,
+       NR_ANON_THPS,
+       NR_UNSTABLE_NFS,        /* NFS unstable pages */
        NR_VM_NODE_STAT_ITEMS
 };
 
index 531f581..ad20f2d 100644 (file)
@@ -412,9 +412,9 @@ TRACE_EVENT(global_dirty_state,
        ),
 
        TP_fast_assign(
-               __entry->nr_dirty       = global_page_state(NR_FILE_DIRTY);
-               __entry->nr_writeback   = global_page_state(NR_WRITEBACK);
-               __entry->nr_unstable    = global_page_state(NR_UNSTABLE_NFS);
+               __entry->nr_dirty       = global_node_page_state(NR_FILE_DIRTY);
+               __entry->nr_writeback   = global_node_page_state(NR_WRITEBACK);
+               __entry->nr_unstable    = global_node_page_state(NR_UNSTABLE_NFS);
                __entry->nr_dirtied     = global_page_state(NR_DIRTIED);
                __entry->nr_written     = global_page_state(NR_WRITTEN);
                __entry->background_thresh = background_thresh;
index 7ec50bd..c5f5e46 100644 (file)
@@ -218,11 +218,11 @@ void __delete_from_page_cache(struct page *page, void *shadow)
 
        /* hugetlb pages do not participate in page cache accounting. */
        if (!PageHuge(page))
-               __mod_zone_page_state(page_zone(page), NR_FILE_PAGES, -nr);
+               __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, -nr);
        if (PageSwapBacked(page)) {
-               __mod_zone_page_state(page_zone(page), NR_SHMEM, -nr);
+               __mod_node_page_state(page_pgdat(page), NR_SHMEM, -nr);
                if (PageTransHuge(page))
-                       __dec_zone_page_state(page, NR_SHMEM_THPS);
+                       __dec_node_page_state(page, NR_SHMEM_THPS);
        } else {
                VM_BUG_ON_PAGE(PageTransHuge(page) && !PageHuge(page), page);
        }
@@ -568,9 +568,9 @@ int replace_page_cache_page(struct page *old, struct page *new, gfp_t gfp_mask)
                 * hugetlb pages do not participate in page cache accounting.
                 */
                if (!PageHuge(new))
-                       __inc_zone_page_state(new, NR_FILE_PAGES);
+                       __inc_node_page_state(new, NR_FILE_PAGES);
                if (PageSwapBacked(new))
-                       __inc_zone_page_state(new, NR_SHMEM);
+                       __inc_node_page_state(new, NR_SHMEM);
                spin_unlock_irqrestore(&mapping->tree_lock, flags);
                mem_cgroup_migrate(old, new);
                radix_tree_preload_end();
@@ -677,7 +677,7 @@ static int __add_to_page_cache_locked(struct page *page,
 
        /* hugetlb pages do not participate in page cache accounting. */
        if (!huge)
-               __inc_zone_page_state(page, NR_FILE_PAGES);
+               __inc_node_page_state(page, NR_FILE_PAGES);
        spin_unlock_irq(&mapping->tree_lock);
        if (!huge)
                mem_cgroup_commit_charge(page, memcg, false, false);
index 481fb01..121a7f8 100644 (file)
@@ -1586,7 +1586,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
 
        if (atomic_add_negative(-1, compound_mapcount_ptr(page))) {
                /* Last compound_mapcount is gone. */
-               __dec_zone_page_state(page, NR_ANON_THPS);
+               __dec_node_page_state(page, NR_ANON_THPS);
                if (TestClearPageDoubleMap(page)) {
                        /* No need in mapcount reference anymore */
                        for (i = 0; i < HPAGE_PMD_NR; i++)
@@ -2061,7 +2061,7 @@ int split_huge_page_to_list(struct page *page, struct list_head *list)
                        list_del(page_deferred_list(head));
                }
                if (mapping)
-                       __dec_zone_page_state(page, NR_SHMEM_THPS);
+                       __dec_node_page_state(page, NR_SHMEM_THPS);
                spin_unlock(&pgdata->split_queue_lock);
                __split_huge_page(page, list, flags);
                ret = 0;
index 374237b..d03b14a 100644 (file)
@@ -1483,10 +1483,10 @@ tree_unlocked:
                }
 
                local_irq_save(flags);
-               __inc_zone_page_state(new_page, NR_SHMEM_THPS);
+               __inc_node_page_state(new_page, NR_SHMEM_THPS);
                if (nr_none) {
-                       __mod_zone_page_state(zone, NR_FILE_PAGES, nr_none);
-                       __mod_zone_page_state(zone, NR_SHMEM, nr_none);
+                       __mod_node_page_state(zone->zone_pgdat, NR_FILE_PAGES, nr_none);
+                       __mod_node_page_state(zone->zone_pgdat, NR_SHMEM, nr_none);
                }
                local_irq_restore(flags);
 
index fba770c..ed02682 100644 (file)
@@ -505,15 +505,17 @@ int migrate_page_move_mapping(struct address_space *mapping,
         * are mapped to swap space.
         */
        if (newzone != oldzone) {
-               __dec_zone_state(oldzone, NR_FILE_PAGES);
-               __inc_zone_state(newzone, NR_FILE_PAGES);
+               __dec_node_state(oldzone->zone_pgdat, NR_FILE_PAGES);
+               __inc_node_state(newzone->zone_pgdat, NR_FILE_PAGES);
                if (PageSwapBacked(page) && !PageSwapCache(page)) {
-                       __dec_zone_state(oldzone, NR_SHMEM);
-                       __inc_zone_state(newzone, NR_SHMEM);
+                       __dec_node_state(oldzone->zone_pgdat, NR_SHMEM);
+                       __inc_node_state(newzone->zone_pgdat, NR_SHMEM);
                }
                if (dirty && mapping_cap_account_dirty(mapping)) {
-                       __dec_zone_state(oldzone, NR_FILE_DIRTY);
-                       __inc_zone_state(newzone, NR_FILE_DIRTY);
+                       __dec_node_state(oldzone->zone_pgdat, NR_FILE_DIRTY);
+                       __dec_zone_state(oldzone, NR_ZONE_WRITE_PENDING);
+                       __inc_node_state(newzone->zone_pgdat, NR_FILE_DIRTY);
+                       __inc_zone_state(newzone, NR_ZONE_WRITE_PENDING);
                }
        }
        local_irq_enable();
index f7c0fb9..f97591d 100644 (file)
@@ -498,20 +498,12 @@ static unsigned long node_dirty_limit(struct pglist_data *pgdat)
  */
 bool node_dirty_ok(struct pglist_data *pgdat)
 {
-       int z;
        unsigned long limit = node_dirty_limit(pgdat);
        unsigned long nr_pages = 0;
 
-       for (z = 0; z < MAX_NR_ZONES; z++) {
-               struct zone *zone = pgdat->node_zones + z;
-
-               if (!populated_zone(zone))
-                       continue;
-
-               nr_pages += zone_page_state(zone, NR_FILE_DIRTY);
-               nr_pages += zone_page_state(zone, NR_UNSTABLE_NFS);
-               nr_pages += zone_page_state(zone, NR_WRITEBACK);
-       }
+       nr_pages += node_page_state(pgdat, NR_FILE_DIRTY);
+       nr_pages += node_page_state(pgdat, NR_UNSTABLE_NFS);
+       nr_pages += node_page_state(pgdat, NR_WRITEBACK);
 
        return nr_pages <= limit;
 }
@@ -1601,10 +1593,10 @@ static void balance_dirty_pages(struct address_space *mapping,
                 * written to the server's write cache, but has not yet
                 * been flushed to permanent storage.
                 */
-               nr_reclaimable = global_page_state(NR_FILE_DIRTY) +
-                                       global_page_state(NR_UNSTABLE_NFS);
+               nr_reclaimable = global_node_page_state(NR_FILE_DIRTY) +
+                                       global_node_page_state(NR_UNSTABLE_NFS);
                gdtc->avail = global_dirtyable_memory();
-               gdtc->dirty = nr_reclaimable + global_page_state(NR_WRITEBACK);
+               gdtc->dirty = nr_reclaimable + global_node_page_state(NR_WRITEBACK);
 
                domain_dirty_limits(gdtc);
 
@@ -1941,8 +1933,8 @@ bool wb_over_bg_thresh(struct bdi_writeback *wb)
         * as we're trying to decide whether to put more under writeback.
         */
        gdtc->avail = global_dirtyable_memory();
-       gdtc->dirty = global_page_state(NR_FILE_DIRTY) +
-                     global_page_state(NR_UNSTABLE_NFS);
+       gdtc->dirty = global_node_page_state(NR_FILE_DIRTY) +
+                     global_node_page_state(NR_UNSTABLE_NFS);
        domain_dirty_limits(gdtc);
 
        if (gdtc->dirty > gdtc->bg_thresh)
@@ -1986,8 +1978,8 @@ void throttle_vm_writeout(gfp_t gfp_mask)
                  */
                 dirty_thresh += dirty_thresh / 10;      /* wheeee... */
 
-                if (global_page_state(NR_UNSTABLE_NFS) +
-                       global_page_state(NR_WRITEBACK) <= dirty_thresh)
+                if (global_node_page_state(NR_UNSTABLE_NFS) +
+                       global_node_page_state(NR_WRITEBACK) <= dirty_thresh)
                                break;
                 congestion_wait(BLK_RW_ASYNC, HZ/10);
 
@@ -2015,8 +2007,8 @@ int dirty_writeback_centisecs_handler(struct ctl_table *table, int write,
 void laptop_mode_timer_fn(unsigned long data)
 {
        struct request_queue *q = (struct request_queue *)data;
-       int nr_pages = global_page_state(NR_FILE_DIRTY) +
-               global_page_state(NR_UNSTABLE_NFS);
+       int nr_pages = global_node_page_state(NR_FILE_DIRTY) +
+               global_node_page_state(NR_UNSTABLE_NFS);
        struct bdi_writeback *wb;
 
        /*
@@ -2467,7 +2459,8 @@ void account_page_dirtied(struct page *page, struct address_space *mapping)
                wb = inode_to_wb(inode);
 
                mem_cgroup_inc_page_stat(page, MEM_CGROUP_STAT_DIRTY);
-               __inc_zone_page_state(page, NR_FILE_DIRTY);
+               __inc_node_page_state(page, NR_FILE_DIRTY);
+               __inc_zone_page_state(page, NR_ZONE_WRITE_PENDING);
                __inc_zone_page_state(page, NR_DIRTIED);
                __inc_wb_stat(wb, WB_RECLAIMABLE);
                __inc_wb_stat(wb, WB_DIRTIED);
@@ -2488,7 +2481,8 @@ void account_page_cleaned(struct page *page, struct address_space *mapping,
 {
        if (mapping_cap_account_dirty(mapping)) {
                mem_cgroup_dec_page_stat(page, MEM_CGROUP_STAT_DIRTY);
-               dec_zone_page_state(page, NR_FILE_DIRTY);
+               dec_node_page_state(page, NR_FILE_DIRTY);
+               dec_zone_page_state(page, NR_ZONE_WRITE_PENDING);
                dec_wb_stat(wb, WB_RECLAIMABLE);
                task_io_account_cancelled_write(PAGE_SIZE);
        }
@@ -2744,7 +2738,8 @@ int clear_page_dirty_for_io(struct page *page)
                wb = unlocked_inode_to_wb_begin(inode, &locked);
                if (TestClearPageDirty(page)) {
                        mem_cgroup_dec_page_stat(page, MEM_CGROUP_STAT_DIRTY);
-                       dec_zone_page_state(page, NR_FILE_DIRTY);
+                       dec_node_page_state(page, NR_FILE_DIRTY);
+                       dec_zone_page_state(page, NR_ZONE_WRITE_PENDING);
                        dec_wb_stat(wb, WB_RECLAIMABLE);
                        ret = 1;
                }
@@ -2790,7 +2785,8 @@ int test_clear_page_writeback(struct page *page)
        }
        if (ret) {
                mem_cgroup_dec_page_stat(page, MEM_CGROUP_STAT_WRITEBACK);
-               dec_zone_page_state(page, NR_WRITEBACK);
+               dec_node_page_state(page, NR_WRITEBACK);
+               dec_zone_page_state(page, NR_ZONE_WRITE_PENDING);
                inc_zone_page_state(page, NR_WRITTEN);
        }
        unlock_page_memcg(page);
@@ -2844,7 +2840,8 @@ int __test_set_page_writeback(struct page *page, bool keep_write)
        }
        if (!ret) {
                mem_cgroup_inc_page_stat(page, MEM_CGROUP_STAT_WRITEBACK);
-               inc_zone_page_state(page, NR_WRITEBACK);
+               inc_node_page_state(page, NR_WRITEBACK);
+               inc_zone_page_state(page, NR_ZONE_WRITE_PENDING);
        }
        unlock_page_memcg(page);
        return ret;
index c11935b..0f92e04 100644 (file)
@@ -3492,14 +3492,12 @@ should_reclaim_retry(gfp_t gfp_mask, unsigned order,
                         * prevent from pre mature OOM
                         */
                        if (!did_some_progress) {
-                               unsigned long writeback;
-                               unsigned long dirty;
+                               unsigned long write_pending;
 
-                               writeback = zone_page_state_snapshot(zone,
-                                                                    NR_WRITEBACK);
-                               dirty = zone_page_state_snapshot(zone, NR_FILE_DIRTY);
+                               write_pending = zone_page_state_snapshot(zone,
+                                                       NR_ZONE_WRITE_PENDING);
 
-                               if (2*(writeback + dirty) > reclaimable) {
+                               if (2 * write_pending > reclaimable) {
                                        congestion_wait(BLK_RW_ASYNC, HZ/10);
                                        return true;
                                }
@@ -4175,7 +4173,7 @@ EXPORT_SYMBOL_GPL(si_mem_available);
 void si_meminfo(struct sysinfo *val)
 {
        val->totalram = totalram_pages;
-       val->sharedram = global_page_state(NR_SHMEM);
+       val->sharedram = global_node_page_state(NR_SHMEM);
        val->freeram = global_page_state(NR_FREE_PAGES);
        val->bufferram = nr_blockdev_pages();
        val->totalhigh = totalhigh_pages;
@@ -4197,7 +4195,7 @@ void si_meminfo_node(struct sysinfo *val, int nid)
        for (zone_type = 0; zone_type < MAX_NR_ZONES; zone_type++)
                managed_pages += pgdat->node_zones[zone_type].managed_pages;
        val->totalram = managed_pages;
-       val->sharedram = sum_zone_node_page_state(nid, NR_SHMEM);
+       val->sharedram = node_page_state(pgdat, NR_SHMEM);
        val->freeram = sum_zone_node_page_state(nid, NR_FREE_PAGES);
 #ifdef CONFIG_HIGHMEM
        for (zone_type = 0; zone_type < MAX_NR_ZONES; zone_type++) {
@@ -4296,9 +4294,6 @@ void show_free_areas(unsigned int filter)
                " unevictable:%lu dirty:%lu writeback:%lu unstable:%lu\n"
                " slab_reclaimable:%lu slab_unreclaimable:%lu\n"
                " mapped:%lu shmem:%lu pagetables:%lu bounce:%lu\n"
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-               " anon_thp: %lu shmem_thp: %lu shmem_pmdmapped: %lu\n"
-#endif
                " free:%lu free_pcp:%lu free_cma:%lu\n",
                global_node_page_state(NR_ACTIVE_ANON),
                global_node_page_state(NR_INACTIVE_ANON),
@@ -4307,20 +4302,15 @@ void show_free_areas(unsigned int filter)
                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),
+               global_node_page_state(NR_FILE_DIRTY),
+               global_node_page_state(NR_WRITEBACK),
+               global_node_page_state(NR_UNSTABLE_NFS),
                global_page_state(NR_SLAB_RECLAIMABLE),
                global_page_state(NR_SLAB_UNRECLAIMABLE),
                global_node_page_state(NR_FILE_MAPPED),
-               global_page_state(NR_SHMEM),
+               global_node_page_state(NR_SHMEM),
                global_page_state(NR_PAGETABLE),
                global_page_state(NR_BOUNCE),
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-               global_page_state(NR_ANON_THPS) * HPAGE_PMD_NR,
-               global_page_state(NR_SHMEM_THPS) * HPAGE_PMD_NR,
-               global_page_state(NR_SHMEM_PMDMAPPED) * HPAGE_PMD_NR,
-#endif
                global_page_state(NR_FREE_PAGES),
                free_pcp,
                global_page_state(NR_FREE_CMA_PAGES));
@@ -4335,6 +4325,16 @@ void show_free_areas(unsigned int filter)
                        " isolated(anon):%lukB"
                        " isolated(file):%lukB"
                        " mapped:%lukB"
+                       " dirty:%lukB"
+                       " writeback:%lukB"
+                       " shmem:%lukB"
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+                       " shmem_thp: %lukB"
+                       " shmem_pmdmapped: %lukB"
+                       " anon_thp: %lukB"
+#endif
+                       " writeback_tmp:%lukB"
+                       " unstable:%lukB"
                        " all_unreclaimable? %s"
                        "\n",
                        pgdat->node_id,
@@ -4346,6 +4346,17 @@ void show_free_areas(unsigned int filter)
                        K(node_page_state(pgdat, NR_ISOLATED_ANON)),
                        K(node_page_state(pgdat, NR_ISOLATED_FILE)),
                        K(node_page_state(pgdat, NR_FILE_MAPPED)),
+                       K(node_page_state(pgdat, NR_FILE_DIRTY)),
+                       K(node_page_state(pgdat, NR_WRITEBACK)),
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+                       K(node_page_state(pgdat, NR_SHMEM_THPS) * HPAGE_PMD_NR),
+                       K(node_page_state(pgdat, NR_SHMEM_PMDMAPPED)
+                                       * HPAGE_PMD_NR),
+                       K(node_page_state(pgdat, NR_ANON_THPS) * HPAGE_PMD_NR),
+#endif
+                       K(node_page_state(pgdat, NR_SHMEM)),
+                       K(node_page_state(pgdat, NR_WRITEBACK_TEMP)),
+                       K(node_page_state(pgdat, NR_UNSTABLE_NFS)),
                        !pgdat_reclaimable(pgdat) ? "yes" : "no");
        }
 
@@ -4368,24 +4379,14 @@ void show_free_areas(unsigned int filter)
                        " present:%lukB"
                        " managed:%lukB"
                        " mlocked:%lukB"
-                       " dirty:%lukB"
-                       " writeback:%lukB"
-                       " shmem:%lukB"
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-                       " shmem_thp: %lukB"
-                       " shmem_pmdmapped: %lukB"
-                       " anon_thp: %lukB"
-#endif
                        " slab_reclaimable:%lukB"
                        " slab_unreclaimable:%lukB"
                        " kernel_stack:%lukB"
                        " pagetables:%lukB"
-                       " unstable:%lukB"
                        " bounce:%lukB"
                        " free_pcp:%lukB"
                        " local_pcp:%ukB"
                        " free_cma:%lukB"
-                       " writeback_tmp:%lukB"
                        " node_pages_scanned:%lu"
                        "\n",
                        zone->name,
@@ -4396,26 +4397,15 @@ void show_free_areas(unsigned int filter)
                        K(zone->present_pages),
                        K(zone->managed_pages),
                        K(zone_page_state(zone, NR_MLOCK)),
-                       K(zone_page_state(zone, NR_FILE_DIRTY)),
-                       K(zone_page_state(zone, NR_WRITEBACK)),
-                       K(zone_page_state(zone, NR_SHMEM)),
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-                       K(zone_page_state(zone, NR_SHMEM_THPS) * HPAGE_PMD_NR),
-                       K(zone_page_state(zone, NR_SHMEM_PMDMAPPED)
-                                       * HPAGE_PMD_NR),
-                       K(zone_page_state(zone, NR_ANON_THPS) * HPAGE_PMD_NR),
-#endif
                        K(zone_page_state(zone, NR_SLAB_RECLAIMABLE)),
                        K(zone_page_state(zone, NR_SLAB_UNRECLAIMABLE)),
                        zone_page_state(zone, NR_KERNEL_STACK) *
                                THREAD_SIZE / 1024,
                        K(zone_page_state(zone, NR_PAGETABLE)),
-                       K(zone_page_state(zone, NR_UNSTABLE_NFS)),
                        K(zone_page_state(zone, NR_BOUNCE)),
                        K(free_pcp),
                        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(node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED)));
                printk("lowmem_reserve[]:");
                for (i = 0; i < MAX_NR_ZONES; i++)
@@ -4458,7 +4448,7 @@ void show_free_areas(unsigned int filter)
 
        hugetlb_show_meminfo();
 
-       printk("%ld total pagecache pages\n", global_page_state(NR_FILE_PAGES));
+       printk("%ld total pagecache pages\n", global_node_page_state(NR_FILE_PAGES));
 
        show_swap_cache_info();
 }
index fdb3b5b..709bc83 100644 (file)
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1213,7 +1213,7 @@ void do_page_add_anon_rmap(struct page *page,
                 * disabled.
                 */
                if (compound)
-                       __inc_zone_page_state(page, NR_ANON_THPS);
+                       __inc_node_page_state(page, NR_ANON_THPS);
                __mod_node_page_state(page_pgdat(page), NR_ANON_MAPPED, nr);
        }
        if (unlikely(PageKsm(page)))
@@ -1251,7 +1251,7 @@ void page_add_new_anon_rmap(struct page *page,
                VM_BUG_ON_PAGE(!PageTransHuge(page), page);
                /* increment count (starts at -1) */
                atomic_set(compound_mapcount_ptr(page), 0);
-               __inc_zone_page_state(page, NR_ANON_THPS);
+               __inc_node_page_state(page, NR_ANON_THPS);
        } else {
                /* Anon THP always mapped first with PMD */
                VM_BUG_ON_PAGE(PageTransCompound(page), page);
@@ -1282,7 +1282,7 @@ void page_add_file_rmap(struct page *page, bool compound)
                if (!atomic_inc_and_test(compound_mapcount_ptr(page)))
                        goto out;
                VM_BUG_ON_PAGE(!PageSwapBacked(page), page);
-               __inc_zone_page_state(page, NR_SHMEM_PMDMAPPED);
+               __inc_node_page_state(page, NR_SHMEM_PMDMAPPED);
        } else {
                if (PageTransCompound(page)) {
                        VM_BUG_ON_PAGE(!PageLocked(page), page);
@@ -1322,7 +1322,7 @@ static void page_remove_file_rmap(struct page *page, bool compound)
                if (!atomic_add_negative(-1, compound_mapcount_ptr(page)))
                        goto out;
                VM_BUG_ON_PAGE(!PageSwapBacked(page), page);
-               __dec_zone_page_state(page, NR_SHMEM_PMDMAPPED);
+               __dec_node_page_state(page, NR_SHMEM_PMDMAPPED);
        } else {
                if (!atomic_add_negative(-1, &page->_mapcount))
                        goto out;
@@ -1356,7 +1356,7 @@ static void page_remove_anon_compound_rmap(struct page *page)
        if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE))
                return;
 
-       __dec_zone_page_state(page, NR_ANON_THPS);
+       __dec_node_page_state(page, NR_ANON_THPS);
 
        if (TestClearPageDoubleMap(page)) {
                /*
index 62e42c7..2ac19a6 100644 (file)
@@ -575,9 +575,9 @@ static int shmem_add_to_page_cache(struct page *page,
        if (!error) {
                mapping->nrpages += nr;
                if (PageTransHuge(page))
-                       __inc_zone_page_state(page, NR_SHMEM_THPS);
-               __mod_zone_page_state(page_zone(page), NR_FILE_PAGES, nr);
-               __mod_zone_page_state(page_zone(page), NR_SHMEM, nr);
+                       __inc_node_page_state(page, NR_SHMEM_THPS);
+               __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, nr);
+               __mod_node_page_state(page_pgdat(page), NR_SHMEM, nr);
                spin_unlock_irq(&mapping->tree_lock);
        } else {
                page->mapping = NULL;
@@ -601,8 +601,8 @@ static void shmem_delete_from_page_cache(struct page *page, void *radswap)
        error = shmem_radix_tree_replace(mapping, page->index, page, radswap);
        page->mapping = NULL;
        mapping->nrpages--;
-       __dec_zone_page_state(page, NR_FILE_PAGES);
-       __dec_zone_page_state(page, NR_SHMEM);
+       __dec_node_page_state(page, NR_FILE_PAGES);
+       __dec_node_page_state(page, NR_SHMEM);
        spin_unlock_irq(&mapping->tree_lock);
        put_page(page);
        BUG_ON(error);
@@ -1493,8 +1493,8 @@ static int shmem_replace_page(struct page **pagep, gfp_t gfp,
        error = shmem_radix_tree_replace(swap_mapping, swap_index, oldpage,
                                                                   newpage);
        if (!error) {
-               __inc_zone_page_state(newpage, NR_FILE_PAGES);
-               __dec_zone_page_state(oldpage, NR_FILE_PAGES);
+               __inc_node_page_state(newpage, NR_FILE_PAGES);
+               __dec_node_page_state(oldpage, NR_FILE_PAGES);
        }
        spin_unlock_irq(&swap_mapping->tree_lock);
 
index c99463a..c8310a3 100644 (file)
@@ -95,7 +95,7 @@ int __add_to_swap_cache(struct page *page, swp_entry_t entry)
                                        entry.val, page);
        if (likely(!error)) {
                address_space->nrpages++;
-               __inc_zone_page_state(page, NR_FILE_PAGES);
+               __inc_node_page_state(page, NR_FILE_PAGES);
                INC_CACHE_INFO(add_total);
        }
        spin_unlock_irq(&address_space->tree_lock);
@@ -147,7 +147,7 @@ void __delete_from_swap_cache(struct page *page)
        set_page_private(page, 0);
        ClearPageSwapCache(page);
        address_space->nrpages--;
-       __dec_zone_page_state(page, NR_FILE_PAGES);
+       __dec_node_page_state(page, NR_FILE_PAGES);
        INC_CACHE_INFO(del_total);
 }
 
index 8d010ef..662cddf 100644 (file)
--- a/mm/util.c
+++ b/mm/util.c
@@ -528,7 +528,7 @@ int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin)
 
        if (sysctl_overcommit_memory == OVERCOMMIT_GUESS) {
                free = global_page_state(NR_FREE_PAGES);
-               free += global_page_state(NR_FILE_PAGES);
+               free += global_node_page_state(NR_FILE_PAGES);
 
                /*
                 * shmem pages shouldn't be counted as free in this
@@ -536,7 +536,7 @@ int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin)
                 * that won't affect the overall amount of available
                 * memory in the system.
                 */
-               free -= global_page_state(NR_SHMEM);
+               free -= global_node_page_state(NR_SHMEM);
 
                free += get_nr_swap_pages();
 
index 90b4665..b797afe 100644 (file)
@@ -3587,11 +3587,11 @@ int sysctl_min_unmapped_ratio = 1;
  */
 int sysctl_min_slab_ratio = 5;
 
-static inline unsigned long zone_unmapped_file_pages(struct zone *zone)
+static inline unsigned long node_unmapped_file_pages(struct pglist_data *pgdat)
 {
-       unsigned long file_mapped = node_page_state(zone->zone_pgdat, NR_FILE_MAPPED);
-       unsigned long file_lru = node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE) +
-               node_page_state(zone->zone_pgdat, NR_ACTIVE_FILE);
+       unsigned long file_mapped = node_page_state(pgdat, NR_FILE_MAPPED);
+       unsigned long file_lru = node_page_state(pgdat, NR_INACTIVE_FILE) +
+               node_page_state(pgdat, NR_ACTIVE_FILE);
 
        /*
         * It's possible for there to be more file mapped pages than
@@ -3610,17 +3610,17 @@ static unsigned long zone_pagecache_reclaimable(struct zone *zone)
        /*
         * If RECLAIM_UNMAP is set, then all file pages are considered
         * potentially reclaimable. Otherwise, we have to worry about
-        * pages like swapcache and zone_unmapped_file_pages() provides
+        * pages like swapcache and node_unmapped_file_pages() provides
         * a better estimate
         */
        if (zone_reclaim_mode & RECLAIM_UNMAP)
-               nr_pagecache_reclaimable = zone_page_state(zone, NR_FILE_PAGES);
+               nr_pagecache_reclaimable = node_page_state(zone->zone_pgdat, NR_FILE_PAGES);
        else
-               nr_pagecache_reclaimable = zone_unmapped_file_pages(zone);
+               nr_pagecache_reclaimable = node_unmapped_file_pages(zone->zone_pgdat);
 
        /* If we can't clean pages, remove dirty pages from consideration */
        if (!(zone_reclaim_mode & RECLAIM_WRITE))
-               delta += zone_page_state(zone, NR_FILE_DIRTY);
+               delta += node_page_state(zone->zone_pgdat, NR_FILE_DIRTY);
 
        /* Watch for any possible underflows due to delta */
        if (unlikely(delta > nr_pagecache_reclaimable))
index 02e7406..4553921 100644 (file)
@@ -924,20 +924,15 @@ const char * const vmstat_text[] = {
        "nr_alloc_batch",
        "nr_zone_anon_lru",
        "nr_zone_file_lru",
+       "nr_zone_write_pending",
        "nr_mlock",
-       "nr_file_pages",
-       "nr_dirty",
-       "nr_writeback",
        "nr_slab_reclaimable",
        "nr_slab_unreclaimable",
        "nr_page_table_pages",
        "nr_kernel_stack",
-       "nr_unstable",
        "nr_bounce",
        "nr_vmscan_write",
        "nr_vmscan_immediate_reclaim",
-       "nr_writeback_temp",
-       "nr_shmem",
        "nr_dirtied",
        "nr_written",
 #if IS_ENABLED(CONFIG_ZSMALLOC)
@@ -951,9 +946,6 @@ const char * const vmstat_text[] = {
        "numa_local",
        "numa_other",
 #endif
-       "nr_anon_transparent_hugepages",
-       "nr_shmem_hugepages",
-       "nr_shmem_pmdmapped",
        "nr_free_cma",
 
        /* Node-based counters */
@@ -970,6 +962,15 @@ const char * const vmstat_text[] = {
        "workingset_nodereclaim",
        "nr_anon_pages",
        "nr_mapped",
+       "nr_file_pages",
+       "nr_dirty",
+       "nr_writeback",
+       "nr_writeback_temp",
+       "nr_shmem",
+       "nr_shmem_hugepages",
+       "nr_shmem_pmdmapped",
+       "nr_anon_transparent_hugepages",
+       "nr_unstable",
 
        /* enum writeback_stat_item counters */
        "nr_dirty_threshold",