fix soft lock up at NFS mount via per-SB LRU-list of unused dentries

[Summary]

Split LRU-list of unused dentries to one per superblock to avoid soft
lock up during NFS mounts and remounting of any filesystem.

Previously I posted here:
http://lkml.org/lkml/2008/3/5/590

[Descriptions]

- background

dentry_unused is a list of dentries which are not referenced.
dentry_unused grows up when references on directories or files are
released. This list can be very long if there is huge free memory.

- the problem

When shrink_dcache_sb() is called, it scans all dentry_unused linearly
under spin_lock(), and if dentry->d_sb is differnt from given
superblock, scan next dentry. This scan costs very much if there are
many entries, and very ineffective if there are many superblocks.

IOW, When we need to shrink unused dentries on one dentry, but scans
unused dentries on all superblocks in the system. For example, we scan
500 dentries to unmount a filesystem, but scans 1,000,000 or more unused
dentries on other superblocks.

In our case , At mounting NFS*, shrink_dcache_sb() is called to shrink
unused dentries on NFS, but scans 100,000,000 unused dentries on
superblocks in the system such as local ext3 filesystems. I hear NFS
mounting took 1 min on some system in use.

* : NFS uses virtual filesystem in rpc layer, so NFS is affected by
this problem.

100,000,000 is possible number on large systems.

Per-superblock LRU of unused dentried can reduce the cost in
reasonable manner.

- How to fix

I found this problem is solved by David Chinner's "Per-superblock
unused dentry LRU lists V3"(1), so I rebase it and add some fix to
reclaim with fairness, which is in Andrew Morton's comments(2).

1) http://lkml.org/lkml/2006/5/25/318
2) http://lkml.org/lkml/2006/5/25/320

Split LRU-list of unused dentries to each superblocks. Then, NFS
mounting will check dentries under a superblock instead of all. But
this spliting will break LRU of dentry-unused. So, I've attempted to
make reclaim unused dentrins with fairness by calculate number of
dentries to scan on this sb based on following way

number of dentries to scan on this sb =
count * (number of dentries on this sb / number of dentries in the machine)

- ToDo
- I have to measuring performance number and do stress tests.

- When unmount occurs during prune_dcache(), scanning on same
superblock, It is unable to reach next superblock because it is gone
away. We restart scannig superblock from first one, it causes
unfairness of reclaim unused dentries on first superblock. But I think
this happens very rarely.

- Test Results

Result on 6GB boxes with excessive unused dentries.

Without patch:

$ cat /proc/sys/fs/dentry-state
10181835 10180203 45 0 0 0
# mount -t nfs 10.124.60.70:/work/kernel-src nfs
real 0m1.830s
user 0m0.001s
sys 0m1.653s

With this patch:
$ cat /proc/sys/fs/dentry-state
10236610 10234751 45 0 0 0
# mount -t nfs 10.124.60.70:/work/kernel-src nfs
real 0m0.106s
user 0m0.002s
sys 0m0.032s

[[e-mail]: fix comments]
Signed-off-by: Kentaro Makita <[e-mail]>
Cc: Neil Brown <[e-mail]>
Cc: Trond Myklebust <[e-mail]>
Cc: David Chinner <[e-mail]>
Cc: "J. Bruce Fields" <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>
Signed-off-by: Linus Torvalds <[e-mail]>

fs/dcache.c

···
 static unsigned int d_hash_mask __read_mostly;
 static unsigned int d_hash_shift __read_mostly;
 static struct hlist_head *dentry_hashtable __read_mostly;
-static LIST_HEAD(dentry_unused);
 
 /* Statistics gathering. */
 struct dentry_stat_t dentry_stat = {
···
 		call_rcu(&dentry->d_u.d_rcu, d_callback);
 }
 
-static void dentry_lru_remove(struct dentry *dentry)
-{
-	if (!list_empty(&dentry->d_lru)) {
-		list_del_init(&dentry->d_lru);
-		dentry_stat.nr_unused--;
-	}
-}
-
 /*
  * Release the dentry's inode, using the filesystem
  * d_iput() operation if defined.
···
 	} else {
 		spin_unlock(&dentry->d_lock);
 		spin_unlock(&dcache_lock);
+	}
+}
+
+/*
+ * dentry_lru_(add|add_tail|del|del_init) must be called with dcache_lock held.
+ */
+static void dentry_lru_add(struct dentry *dentry)
+{
+	list_add(&dentry->d_lru, &dentry->d_sb->s_dentry_lru);
+	dentry->d_sb->s_nr_dentry_unused++;
+	dentry_stat.nr_unused++;
+}
+
+static void dentry_lru_add_tail(struct dentry *dentry)
+{
+	list_add_tail(&dentry->d_lru, &dentry->d_sb->s_dentry_lru);
+	dentry->d_sb->s_nr_dentry_unused++;
+	dentry_stat.nr_unused++;
+}
+
+static void dentry_lru_del(struct dentry *dentry)
+{
+	if (!list_empty(&dentry->d_lru)) {
+		list_del(&dentry->d_lru);
+		dentry->d_sb->s_nr_dentry_unused--;
+		dentry_stat.nr_unused--;
+	}
+}
+
+static void dentry_lru_del_init(struct dentry *dentry)
+{
+	if (likely(!list_empty(&dentry->d_lru))) {
+		list_del_init(&dentry->d_lru);
+		dentry->d_sb->s_nr_dentry_unused--;
+		dentry_stat.nr_unused--;
 	}
 }
 
···
 		goto kill_it;
   	if (list_empty(&dentry->d_lru)) {
   		dentry->d_flags |= DCACHE_REFERENCED;
-  		list_add(&dentry->d_lru, &dentry_unused);
-  		dentry_stat.nr_unused++;
+		dentry_lru_add(dentry);
   	}
  	spin_unlock(&dentry->d_lock);
 	spin_unlock(&dcache_lock);
···
 unhash_it:
 	__d_drop(dentry);
 kill_it:
-	dentry_lru_remove(dentry);
+	/* if dentry was on the d_lru list delete it from there */
+	dentry_lru_del(dentry);
 	dentry = d_kill(dentry);
 	if (dentry)
 		goto repeat;
···
 static inline struct dentry * __dget_locked(struct dentry *dentry)
 {
 	atomic_inc(&dentry->d_count);
-	dentry_lru_remove(dentry);
+	dentry_lru_del_init(dentry);
 	return dentry;
 }
 
···
 
 		if (dentry->d_op && dentry->d_op->d_delete)
 			dentry->d_op->d_delete(dentry);
-		dentry_lru_remove(dentry);
+		dentry_lru_del_init(dentry);
 		__d_drop(dentry);
 		dentry = d_kill(dentry);
 		spin_lock(&dcache_lock);
 	}
 }
 
-/**
- * prune_dcache - shrink the dcache
- * @count: number of entries to try and free
- * @sb: if given, ignore dentries for other superblocks
- *         which are being unmounted.
- *
- * Shrink the dcache. This is done when we need
- * more memory, or simply when we need to unmount
- * something (at which point we need to unuse
- * all dentries).
- *
- * This function may fail to free any resources if
- * all the dentries are in use.
+/*
+ * Shrink the dentry LRU on a given superblock.
+ * @sb   : superblock to shrink dentry LRU.
+ * @count: If count is NULL, we prune all dentries on superblock.
+ * @flags: If flags is non-zero, we need to do special processing based on
+ * which flags are set. This means we don't need to maintain multiple
+ * similar copies of this loop.
  */
- 
-static void prune_dcache(int count, struct super_block *sb)
+static void __shrink_dcache_sb(struct super_block *sb, int *count, int flags)
 {
+	LIST_HEAD(referenced);
+	LIST_HEAD(tmp);
+	struct dentry *dentry;
+	int cnt = 0;
+
+	BUG_ON(!sb);
+	BUG_ON((flags & DCACHE_REFERENCED) && count == NULL);
 	spin_lock(&dcache_lock);
-	for (; count ; count--) {
-		struct dentry *dentry;
-		struct list_head *tmp;
-		struct rw_semaphore *s_umount;
+	if (count != NULL)
+		/* called from prune_dcache() and shrink_dcache_parent() */
+		cnt = *count;
+restart:
+	if (count == NULL)
+		list_splice_init(&sb->s_dentry_lru, &tmp);
+	else {
+		while (!list_empty(&sb->s_dentry_lru)) {
+			dentry = list_entry(sb->s_dentry_lru.prev,
+					struct dentry, d_lru);
+			BUG_ON(dentry->d_sb != sb);
 
-		cond_resched_lock(&dcache_lock);
-
-		tmp = dentry_unused.prev;
-		if (sb) {
-			/* Try to find a dentry for this sb, but don't try
-			 * too hard, if they aren't near the tail they will
-			 * be moved down again soon
+			spin_lock(&dentry->d_lock);
+			/*
+			 * If we are honouring the DCACHE_REFERENCED flag and
+			 * the dentry has this flag set, don't free it. Clear
+			 * the flag and put it back on the LRU.
 			 */
-			int skip = count;
-			while (skip && tmp != &dentry_unused &&
-			    list_entry(tmp, struct dentry, d_lru)->d_sb != sb) {
-				skip--;
-				tmp = tmp->prev;
+			if ((flags & DCACHE_REFERENCED)
+				&& (dentry->d_flags & DCACHE_REFERENCED)) {
+				dentry->d_flags &= ~DCACHE_REFERENCED;
+				list_move_tail(&dentry->d_lru, &referenced);
+				spin_unlock(&dentry->d_lock);
+			} else {
+				list_move_tail(&dentry->d_lru, &tmp);
+				spin_unlock(&dentry->d_lock);
+				cnt--;
+				if (!cnt)
+					break;
 			}
 		}
-		if (tmp == &dentry_unused)
-			break;
-		list_del_init(tmp);
-		prefetch(dentry_unused.prev);
- 		dentry_stat.nr_unused--;
-		dentry = list_entry(tmp, struct dentry, d_lru);
-
- 		spin_lock(&dentry->d_lock);
+	}
+	while (!list_empty(&tmp)) {
+		dentry = list_entry(tmp.prev, struct dentry, d_lru);
+		dentry_lru_del_init(dentry);
+		spin_lock(&dentry->d_lock);
 		/*
 		 * We found an inuse dentry which was not removed from
-		 * dentry_unused because of laziness during lookup.  Do not free
-		 * it - just keep it off the dentry_unused list.
+		 * the LRU because of laziness during lookup.  Do not free
+		 * it - just keep it off the LRU list.
 		 */
- 		if (atomic_read(&dentry->d_count)) {
- 			spin_unlock(&dentry->d_lock);
+		if (atomic_read(&dentry->d_count)) {
+			spin_unlock(&dentry->d_lock);
 			continue;
 		}
-		/* If the dentry was recently referenced, don't free it. */
-		if (dentry->d_flags & DCACHE_REFERENCED) {
-			dentry->d_flags &= ~DCACHE_REFERENCED;
- 			list_add(&dentry->d_lru, &dentry_unused);
- 			dentry_stat.nr_unused++;
- 			spin_unlock(&dentry->d_lock);
-			continue;
-		}
-		/*
-		 * If the dentry is not DCACHED_REFERENCED, it is time
-		 * to remove it from the dcache, provided the super block is
-		 * NULL (which means we are trying to reclaim memory)
-		 * or this dentry belongs to the same super block that
-		 * we want to shrink.
-		 */
-		/*
-		 * If this dentry is for "my" filesystem, then I can prune it
-		 * without taking the s_umount lock (I already hold it).
-		 */
-		if (sb && dentry->d_sb == sb) {
-			prune_one_dentry(dentry);
-			continue;
-		}
-		/*
-		 * ...otherwise we need to be sure this filesystem isn't being
-		 * unmounted, otherwise we could race with
-		 * generic_shutdown_super(), and end up holding a reference to
-		 * an inode while the filesystem is unmounted.
-		 * So we try to get s_umount, and make sure s_root isn't NULL.
-		 * (Take a local copy of s_umount to avoid a use-after-free of
-		 * `dentry').
-		 */
-		s_umount = &dentry->d_sb->s_umount;
-		if (down_read_trylock(s_umount)) {
-			if (dentry->d_sb->s_root != NULL) {
-				prune_one_dentry(dentry);
-				up_read(s_umount);
-				continue;
-			}
-			up_read(s_umount);
-		}
-		spin_unlock(&dentry->d_lock);
-		/*
-		 * Insert dentry at the head of the list as inserting at the
-		 * tail leads to a cycle.
-		 */
- 		list_add(&dentry->d_lru, &dentry_unused);
-		dentry_stat.nr_unused++;
+		prune_one_dentry(dentry);
+		/* dentry->d_lock was dropped in prune_one_dentry() */
+		cond_resched_lock(&dcache_lock);
 	}
+	if (count == NULL && !list_empty(&sb->s_dentry_lru))
+		goto restart;
+	if (count != NULL)
+		*count = cnt;
+	if (!list_empty(&referenced))
+		list_splice(&referenced, &sb->s_dentry_lru);
 	spin_unlock(&dcache_lock);
 }
 
-/*
- * Shrink the dcache for the specified super block.
- * This allows us to unmount a device without disturbing
- * the dcache for the other devices.
+/**
+ * prune_dcache - shrink the dcache
+ * @count: number of entries to try to free
  *
- * This implementation makes just two traversals of the
- * unused list.  On the first pass we move the selected
- * dentries to the most recent end, and on the second
- * pass we free them.  The second pass must restart after
- * each dput(), but since the target dentries are all at
- * the end, it's really just a single traversal.
+ * Shrink the dcache. This is done when we need more memory, or simply when we
+ * need to unmount something (at which point we need to unuse all dentries).
+ *
+ * This function may fail to free any resources if all the dentries are in use.
  */
+static void prune_dcache(int count)
+{
+	struct super_block *sb;
+	int w_count;
+	int unused = dentry_stat.nr_unused;
+	int prune_ratio;
+	int pruned;
+
+	if (unused == 0 || count == 0)
+		return;
+	spin_lock(&dcache_lock);
+restart:
+	if (count >= unused)
+		prune_ratio = 1;
+	else
+		prune_ratio = unused / count;
+	spin_lock(&sb_lock);
+	list_for_each_entry(sb, &super_blocks, s_list) {
+		if (sb->s_nr_dentry_unused == 0)
+			continue;
+		sb->s_count++;
+		/* Now, we reclaim unused dentrins with fairness.
+		 * We reclaim them same percentage from each superblock.
+		 * We calculate number of dentries to scan on this sb
+		 * as follows, but the implementation is arranged to avoid
+		 * overflows:
+		 * number of dentries to scan on this sb =
+		 * count * (number of dentries on this sb /
+		 * number of dentries in the machine)
+		 */
+		spin_unlock(&sb_lock);
+		if (prune_ratio != 1)
+			w_count = (sb->s_nr_dentry_unused / prune_ratio) + 1;
+		else
+			w_count = sb->s_nr_dentry_unused;
+		pruned = w_count;
+		/*
+		 * We need to be sure this filesystem isn't being unmounted,
+		 * otherwise we could race with generic_shutdown_super(), and
+		 * end up holding a reference to an inode while the filesystem
+		 * is unmounted.  So we try to get s_umount, and make sure
+		 * s_root isn't NULL.
+		 */
+		if (down_read_trylock(&sb->s_umount)) {
+			if ((sb->s_root != NULL) &&
+			    (!list_empty(&sb->s_dentry_lru))) {
+				spin_unlock(&dcache_lock);
+				__shrink_dcache_sb(sb, &w_count,
+						DCACHE_REFERENCED);
+				pruned -= w_count;
+				spin_lock(&dcache_lock);
+			}
+			up_read(&sb->s_umount);
+		}
+		spin_lock(&sb_lock);
+		count -= pruned;
+		/*
+		 * restart only when sb is no longer on the list and
+		 * we have more work to do.
+		 */
+		if (__put_super_and_need_restart(sb) && count > 0) {
+			spin_unlock(&sb_lock);
+			goto restart;
+		}
+	}
+	spin_unlock(&sb_lock);
+	spin_unlock(&dcache_lock);
+}
 
 /**
  * shrink_dcache_sb - shrink dcache for a superblock
···
  * is used to free the dcache before unmounting a file
  * system
  */
-
 void shrink_dcache_sb(struct super_block * sb)
 {
-	struct list_head *tmp, *next;
-	struct dentry *dentry;
-
-	/*
-	 * Pass one ... move the dentries for the specified
-	 * superblock to the most recent end of the unused list.
-	 */
-	spin_lock(&dcache_lock);
-	list_for_each_prev_safe(tmp, next, &dentry_unused) {
-		dentry = list_entry(tmp, struct dentry, d_lru);
-		if (dentry->d_sb != sb)
-			continue;
-		list_move_tail(tmp, &dentry_unused);
-	}
-
-	/*
-	 * Pass two ... free the dentries for this superblock.
-	 */
-repeat:
-	list_for_each_prev_safe(tmp, next, &dentry_unused) {
-		dentry = list_entry(tmp, struct dentry, d_lru);
-		if (dentry->d_sb != sb)
-			continue;
-		dentry_stat.nr_unused--;
-		list_del_init(tmp);
-		spin_lock(&dentry->d_lock);
-		if (atomic_read(&dentry->d_count)) {
-			spin_unlock(&dentry->d_lock);
-			continue;
-		}
-		prune_one_dentry(dentry);
-		cond_resched_lock(&dcache_lock);
-		goto repeat;
-	}
-	spin_unlock(&dcache_lock);
+	__shrink_dcache_sb(sb, NULL, 0);
 }
 
 /*
···
 
 	/* detach this root from the system */
 	spin_lock(&dcache_lock);
-	dentry_lru_remove(dentry);
+	dentry_lru_del_init(dentry);
 	__d_drop(dentry);
 	spin_unlock(&dcache_lock);
 
···
 			spin_lock(&dcache_lock);
 			list_for_each_entry(loop, &dentry->d_subdirs,
 					    d_u.d_child) {
-				dentry_lru_remove(loop);
+				dentry_lru_del_init(loop);
 				__d_drop(loop);
 				cond_resched_lock(&dcache_lock);
 			}
···
 		struct dentry *dentry = list_entry(tmp, struct dentry, d_u.d_child);
 		next = tmp->next;
 
-		dentry_lru_remove(dentry);
+		dentry_lru_del_init(dentry);
 		/* 
 		 * move only zero ref count dentries to the end 
 		 * of the unused list for prune_dcache
 		 */
 		if (!atomic_read(&dentry->d_count)) {
-			list_add_tail(&dentry->d_lru, &dentry_unused);
-			dentry_stat.nr_unused++;
+			dentry_lru_add_tail(dentry);
 			found++;
 		}
 
···
  
 void shrink_dcache_parent(struct dentry * parent)
 {
+	struct super_block *sb = parent->d_sb;
 	int found;
 
 	while ((found = select_parent(parent)) != 0)
-		prune_dcache(found, parent->d_sb);
+		__shrink_dcache_sb(sb, &found, 0);
 }
 
 /*
···
 	if (nr) {
 		if (!(gfp_mask & __GFP_FS))
 			return -1;
-		prune_dcache(nr, NULL);
+		prune_dcache(nr);
 	}
 	return (dentry_stat.nr_unused / 100) * sysctl_vfs_cache_pressure;
 }
···
  * rcu_read_lock() and rcu_read_unlock() are used to disable preemption while
  * lookup is going on.
  *
- * dentry_unused list is not updated even if lookup finds the required dentry
+ * The dentry unused LRU is not updated even if lookup finds the required dentry
  * in there. It is updated in places such as prune_dcache, shrink_dcache_sb,
  * select_parent and __dget_locked. This laziness saves lookup from dcache_lock
  * acquisition.

fs/super.c

···
 		INIT_LIST_HEAD(&s->s_instances);
 		INIT_HLIST_HEAD(&s->s_anon);
 		INIT_LIST_HEAD(&s->s_inodes);
+		INIT_LIST_HEAD(&s->s_dentry_lru);
 		init_rwsem(&s->s_umount);
 		mutex_init(&s->s_lock);
 		lockdep_set_class(&s->s_umount, &type->s_umount_key);

include/linux/fs.h

···
 extern struct list_head super_blocks;
 extern spinlock_t sb_lock;
 
+#define sb_entry(list)  list_entry((list), struct super_block, s_list)
 #define S_BIAS (1<<30)
 struct super_block {
 	struct list_head	s_list;		/* Keep this first */
···
 	struct list_head	s_more_io;	/* parked for more writeback */
 	struct hlist_head	s_anon;		/* anonymous dentries for (nfs) exporting */
 	struct list_head	s_files;
+	/* s_dentry_lru and s_nr_dentry_unused are protected by dcache_lock */
+	struct list_head	s_dentry_lru;	/* unused dentry lru */
+	int			s_nr_dentry_unused;	/* # of dentry on lru */
 
 	struct block_device	*s_bdev;
 	struct mtd_info		*s_mtd;