Skip to contentSkip to conversation

VFS: Destroy the dentries contributed by a superblock on unmounting

PATCH In review David Howells 2 participants 5 messages Jun 30, 2006 – Jul 7, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/2 VFS: Destroy the dentries contributed by a superblock on unmounting

From: David Howells <[email protected]>

The attached patch destroys all the dentries attached to a superblock in one go by:

(1) Destroying the tree rooted at s_root.

(2) Destroying every entry in the anon list, one at a time.

(3) Each entry in the anon list has its subtree consumed from the leaves inwards.

This reduces the amount of work generic_shutdown_super() does, and avoids iterating through the dentry_unused list.

Note that locking is almost entirely absent in the shrink_dcache_for_umount*() functions added by this patch. This is because:

(1) at the point the filesystem calls generic_shutdown_super(), it is not permitted to further touch the superblock's set of dentries, and nor may it remove aliases from inodes;

(2) the dcache memory shrinker now skips dentries that are being unmounted; and

(3) the superblock no longer has any external references through which the VFS can reach it.

Given these points, the only locking we need to do is when we remove dentries from the unused list and the name hashes, which we do a directory's worth at a time.

We also don't need to guard against reference counts going to zero unexpectedly and removing bits of the tree we're working on as nothing else can call dput().

A cut down version of dentry_iput() has been folded into shrink_dcache_for_umount_subtree() function. Apart from not needing to unlock things, it also doesn't need to check for inotify watches.

In this version of the patch, the complaint about a dentry still being in use has been expanded from a single BUG_ON() and now gives much more information.

---

 fs/dcache.c            |  133 ++++++++++++++++++++++++++++++++++++++++++++++++
 fs/super.c             |   12 ++--
 include/linux/dcache.h |    1 
 3 files changed, 140 insertions(+), 6 deletions(-)
fs/dcache.c +133 -0
index 48b44a7..ae041da 100644
@@ -548,6 +548,139 @@ repeat:
 }
 
 /*
+ * destroy a single subtree of dentries for unmount
+ * - see the comments on shrink_dcache_for_umount() for a description of the
+ *   locking
+ */
+static void shrink_dcache_for_umount_subtree(struct dentry *dentry)
+{
+	struct dentry *parent;
+
+	BUG_ON(!IS_ROOT(dentry));
+
+	/* detach this root from the system */
+	spin_lock(&dcache_lock);
+	if (!list_empty(&dentry->d_lru)) {
+		dentry_stat.nr_unused--;
+		list_del_init(&dentry->d_lru);
+	}
+	__d_drop(dentry);
+	spin_unlock(&dcache_lock);
+
+	for (;;) {
+		/* descend to the first leaf in the current subtree */
+		while (!list_empty(&dentry->d_subdirs)) {
+			struct dentry *loop;
+
+			/* this is a branch with children - detach all of them
+			 * from the system in one go */
+			spin_lock(&dcache_lock);
+			list_for_each_entry(loop, &dentry->d_subdirs,
+					    d_u.d_child) {
+				if (!list_empty(&loop->d_lru)) {
+					dentry_stat.nr_unused--;
+					list_del_init(&loop->d_lru);
+				}
+
+				__d_drop(loop);
+				cond_resched_lock(&dcache_lock);
+			}
+			spin_unlock(&dcache_lock);
+
+			/* move to the first child */
+			dentry = list_entry(dentry->d_subdirs.next,
+					    struct dentry, d_u.d_child);
+		}
+
+		/* consume the dentries from this leaf up through its parents
+		 * until we find one with children or run out altogether */
+		do {
+			struct inode *inode;
+
+			if (atomic_read(&dentry->d_count) != 0) {
+				printk(KERN_ERR
+				       "BUG: Dentry %p{i=%lx,n=%s}"
+				       " still in use (%d)"
+				       " [unmount of %s %s]\n",
+				       dentry,
+				       dentry->d_inode ?
+				       dentry->d_inode->i_ino : 0UL,
+				       dentry->d_name.name,
+				       atomic_read(&dentry->d_count),
+				       dentry->d_sb->s_type->name,
+				       dentry->d_sb->s_id);
+				BUG();
+			}
+
+			parent = dentry->d_parent;
+			if (parent == dentry)
+				parent = NULL;
+			else
+				atomic_dec(&parent->d_count);
+
+			list_del(&dentry->d_u.d_child);
+			dentry_stat.nr_dentry--;	/* For d_free, below */
+
+			inode = dentry->d_inode;
+			if (inode) {
+#ifdef CONFIG_INOTIFY
+				BUG_ON(!list_empty(&inode->inotify_watches));
+#endif
+				dentry->d_inode = NULL;
+				list_del_init(&dentry->d_alias);
+				if (dentry->d_op && dentry->d_op->d_iput)
+					dentry->d_op->d_iput(dentry, inode);
+				else
+					iput(inode);
+			}
+
+			d_free(dentry);
+
+			/* finished when we fall off the top of the tree,
+			 * otherwise we ascend to the parent and move to the
+			 * next sibling if there is one */
+			if (!parent)
+				return;
+
+			dentry = parent;
+
+		} while (list_empty(&dentry->d_subdirs));
+
+		dentry = list_entry(dentry->d_subdirs.next,
+				    struct dentry, d_u.d_child);
+	}
+}
+
+/*
+ * destroy the dentries attached to a superblock on unmounting
+ * - we don't need to use dentry->d_lock, and only need dcache_lock when
+ *   removing the dentry from the system lists and hashes because:
+ *   - the superblock is detached from all mountings and open files, so the
+ *     dentry trees will not be rearranged by the VFS
+ *   - s_umount is write-locked, so the memory pressure shrinker will ignore
+ *     any dentries belonging to this superblock that it comes across
+ *   - the filesystem itself is no longer permitted to rearrange the dentries
+ *     in this superblock
+ */
+void shrink_dcache_for_umount(struct super_block *sb)
+{
+	struct dentry *dentry;
+
+	if (down_read_trylock(&sb->s_umount))
+		BUG();
+
+	dentry = sb->s_root;
+	sb->s_root = NULL;
+	atomic_dec(&dentry->d_count);
+	shrink_dcache_for_umount_subtree(dentry);
+
+	while (!hlist_empty(&sb->s_anon)) {
+		dentry = hlist_entry(sb->s_anon.first, struct dentry, d_hash);
+		shrink_dcache_for_umount_subtree(dentry);
+	}
+}
+
+/*
  * Search for at least 1 mount point in the dentry's subdirs.
  * We descend to the next level whenever the d_subdirs
  * list is non-empty and continue searching.
fs/super.c +6 -6
index 8a669f6..bd655b1 100644
@@ -222,17 +222,17 @@ static int grab_super(struct super_block
  *	that need destruction out of superblock, call generic_shutdown_super()
  *	and release aforementioned objects.  Note: dentries and inodes _are_
  *	taken care of and do not need specific handling.
+ *
+ *	Upon calling this function, the filesystem may no longer alter or
+ *	rearrange the set of dentries belonging to this super_block, nor may it
+ *	change the attachments of dentries to inodes.
  */
 void generic_shutdown_super(struct super_block *sb)
 {
-	struct dentry *root = sb->s_root;
 	struct super_operations *sop = sb->s_op;
 
-	if (root) {
-		sb->s_root = NULL;
-		shrink_dcache_parent(root);
-		shrink_dcache_sb(sb);
-		dput(root);
+	if (sb->s_root) {
+		shrink_dcache_for_umount(sb);
 		fsync_super(sb);
 		lock_super(sb);
 		sb->s_flags &= ~MS_ACTIVE;
include/linux/dcache.h +1 -0
index 0dd1610..fe8bcd5 100644
@@ -217,6 +217,7 @@ extern struct dentry * d_alloc_anon(stru
 extern struct dentry * d_splice_alias(struct inode *, struct dentry *);
 extern void shrink_dcache_sb(struct super_block *);
 extern void shrink_dcache_parent(struct dentry *);
+extern void shrink_dcache_for_umount(struct super_block *);
 extern int d_invalidate(struct dentry *);
 
 /* only used at mount-time */
#1 lore raw

PATCH 2/2 AUTOFS: Make sure all dentries refs are released before calling kill_anon_super()

From: David Howells <[email protected]>

Make sure all dentries refs are released before calling kill_anon_super() so that the assumption that generic_shutdown_super() can completely destroy the dentry tree for there will be no external references holds true.

What was being done in the put_super() superblock op, is now done in the kill_sb() filesystem op instead, prior to calling kill_anon_super().

This makes the struct autofs_sb_info::root member variable redundant (since sb->s_root is still available), and so that is removed. The calls to shrink_dcache_sb() are also removed since they're also redundant as shrink_dcache_for_umount() will now be called after the cleanup routine.

---

 fs/autofs4/autofs_i.h |    3 +--
 fs/autofs4/init.c     |    2 +-
 fs/autofs4/inode.c    |   22 ++++------------------
 fs/autofs4/waitq.c    |    1 -
 4 files changed, 6 insertions(+), 22 deletions(-)
fs/autofs4/autofs_i.h +1 -2
index d6603d0..47e38f3 100644
@@ -96,7 +96,6 @@ #define AUTOFS_TYPE_OFFSET       0x0004
 
 struct autofs_sb_info {
 	u32 magic;
-	struct dentry *root;
 	int pipefd;
 	struct file *pipe;
 	pid_t oz_pgrp;
@@ -231,4 +230,4 @@ out:
 }
 
 void autofs4_dentry_release(struct dentry *);
-
+extern void autofs4_kill_sb(struct super_block *);
fs/autofs4/init.c +1 -1
index 5d91933..723a1c5 100644
@@ -24,7 +24,7 @@ static struct file_system_type autofs_fs
 	.owner		= THIS_MODULE,
 	.name		= "autofs",
 	.get_sb		= autofs_get_sb,
-	.kill_sb	= kill_anon_super,
+	.kill_sb	= autofs4_kill_sb,
 };
 
 static int __init init_autofs4_fs(void)
fs/autofs4/inode.c +4 -18
index fde78b1..1bf68c5 100644
@@ -95,7 +95,7 @@ void autofs4_free_ino(struct autofs_info
  */
 static void autofs4_force_release(struct autofs_sb_info *sbi)
 {
-	struct dentry *this_parent = sbi->root;
+	struct dentry *this_parent = sbi->sb->s_root;
 	struct list_head *next;
 
 	spin_lock(&dcache_lock);
@@ -126,7 +126,7 @@ resume:
 		spin_lock(&dcache_lock);
 	}
 
-	if (this_parent != sbi->root) {
+	if (this_parent != sbi->sb->s_root) {
 		struct dentry *dentry = this_parent;
 
 		next = this_parent->d_u.d_child.next;
@@ -139,15 +139,9 @@ resume:
 		goto resume;
 	}
 	spin_unlock(&dcache_lock);
-
-	dput(sbi->root);
-	sbi->root = NULL;
-	shrink_dcache_sb(sbi->sb);
-
-	return;
 }
 
-static void autofs4_put_super(struct super_block *sb)
+void autofs4_kill_sb(struct super_block *sb)
 {
 	struct autofs_sb_info *sbi = autofs4_sbi(sb);
 
@@ -162,6 +156,7 @@ static void autofs4_put_super(struct sup
 	kfree(sbi);
 
 	DPRINTK("shutting down");
+	kill_anon_super(sb);
 }
 
 static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt)
@@ -188,7 +183,6 @@ static int autofs4_show_options(struct s
 }
 
 static struct super_operations autofs4_sops = {
-	.put_super	= autofs4_put_super,
 	.statfs		= simple_statfs,
 	.show_options	= autofs4_show_options,
 };
@@ -314,7 +308,6 @@ int autofs4_fill_super(struct super_bloc
 
 	s->s_fs_info = sbi;
 	sbi->magic = AUTOFS_SBI_MAGIC;
-	sbi->root = NULL;
 	sbi->pipefd = -1;
 	sbi->catatonic = 0;
 	sbi->exp_timeout = 0;
@@ -396,13 +389,6 @@ int autofs4_fill_super(struct super_bloc
 	sbi->pipefd = pipefd;
 
 	/*
-	 * Take a reference to the root dentry so we get a chance to
-	 * clean up the dentry tree on umount.
-	 * See autofs4_force_release.
-	 */
-	sbi->root = dget(root);
-
-	/*
 	 * Success! Install the root dentry now to indicate completion.
 	 */
 	s->s_root = root;
fs/autofs4/waitq.c +0 -1
index ce103e7..c0a6c8d 100644
@@ -45,7 +45,6 @@ void autofs4_catatonic_mode(struct autof
 		fput(sbi->pipe);	/* Close the pipe */
 		sbi->pipe = NULL;
 	}
-	shrink_dcache_sb(sbi->sb);
 }
 
 static int autofs4_write(struct file *file, const void *addr, int bytes)
maintainer
#2 lore raw

Re: [PATCH 2/2] AUTOFS: Make sure all dentries refs are released before calling kill_anon_super()

Quoted 5 lines · David Howells

From: David Howells <[email protected]>

Make sure all dentries refs are released before calling kill_anon_super() so that the assumption that generic_shutdown_super() can completely destroy the dentry tree for there will be no external references holds true.

Thanks for doing this David.
I'll try and give it some testing this weekend.

Quoted 138 lines

What was being done in the put_super() superblock op, is now done in the kill_sb() filesystem op instead, prior to calling kill_anon_super().

This makes the struct autofs_sb_info::root member variable redundant (since sb->s_root is still available), and so that is removed. The calls to shrink_dcache_sb() are also removed since they're also redundant as shrink_dcache_for_umount() will now be called after the cleanup routine.

---

 fs/autofs4/autofs_i.h |    3 +--
 fs/autofs4/init.c     |    2 +-
 fs/autofs4/inode.c    |   22 ++++------------------
 fs/autofs4/waitq.c    |    1 -
 4 files changed, 6 insertions(+), 22 deletions(-)
fs/autofs4/autofs_i.h +1 -2
index d6603d0..47e38f3 100644
@@ -96,7 +96,6 @@ #define AUTOFS_TYPE_OFFSET       0x0004
 
 struct autofs_sb_info {
 	u32 magic;
-	struct dentry *root;
 	int pipefd;
 	struct file *pipe;
 	pid_t oz_pgrp;
@@ -231,4 +230,4 @@ out:
 }
 
 void autofs4_dentry_release(struct dentry *);
-
+extern void autofs4_kill_sb(struct super_block *);
fs/autofs4/init.c +1 -1
index 5d91933..723a1c5 100644
@@ -24,7 +24,7 @@ static struct file_system_type autofs_fs
 	.owner		= THIS_MODULE,
 	.name		= "autofs",
 	.get_sb		= autofs_get_sb,
-	.kill_sb	= kill_anon_super,
+	.kill_sb	= autofs4_kill_sb,
 };
 
 static int __init init_autofs4_fs(void)
fs/autofs4/inode.c +4 -18
index fde78b1..1bf68c5 100644
@@ -95,7 +95,7 @@ void autofs4_free_ino(struct autofs_info
  */
 static void autofs4_force_release(struct autofs_sb_info *sbi)
 {
-	struct dentry *this_parent = sbi->root;
+	struct dentry *this_parent = sbi->sb->s_root;
 	struct list_head *next;
 
 	spin_lock(&dcache_lock);
@@ -126,7 +126,7 @@ resume:
 		spin_lock(&dcache_lock);
 	}
 
-	if (this_parent != sbi->root) {
+	if (this_parent != sbi->sb->s_root) {
 		struct dentry *dentry = this_parent;
 
 		next = this_parent->d_u.d_child.next;
@@ -139,15 +139,9 @@ resume:
 		goto resume;
 	}
 	spin_unlock(&dcache_lock);
-
-	dput(sbi->root);
-	sbi->root = NULL;
-	shrink_dcache_sb(sbi->sb);
-
-	return;
 }
 
-static void autofs4_put_super(struct super_block *sb)
+void autofs4_kill_sb(struct super_block *sb)
 {
 	struct autofs_sb_info *sbi = autofs4_sbi(sb);
 
@@ -162,6 +156,7 @@ static void autofs4_put_super(struct sup
 	kfree(sbi);
 
 	DPRINTK("shutting down");
+	kill_anon_super(sb);
 }
 
 static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt)
@@ -188,7 +183,6 @@ static int autofs4_show_options(struct s
 }
 
 static struct super_operations autofs4_sops = {
-	.put_super	= autofs4_put_super,
 	.statfs		= simple_statfs,
 	.show_options	= autofs4_show_options,
 };
@@ -314,7 +308,6 @@ int autofs4_fill_super(struct super_bloc
 
 	s->s_fs_info = sbi;
 	sbi->magic = AUTOFS_SBI_MAGIC;
-	sbi->root = NULL;
 	sbi->pipefd = -1;
 	sbi->catatonic = 0;
 	sbi->exp_timeout = 0;
@@ -396,13 +389,6 @@ int autofs4_fill_super(struct super_bloc
 	sbi->pipefd = pipefd;
 
 	/*
-	 * Take a reference to the root dentry so we get a chance to
-	 * clean up the dentry tree on umount.
-	 * See autofs4_force_release.
-	 */
-	sbi->root = dget(root);
-
-	/*
 	 * Success! Install the root dentry now to indicate completion.
 	 */
 	s->s_root = root;
fs/autofs4/waitq.c +0 -1
index ce103e7..c0a6c8d 100644
@@ -45,7 +45,6 @@ void autofs4_catatonic_mode(struct autof
 		fput(sbi->pipe);	/* Close the pipe */
 		sbi->pipe = NULL;
 	}
-	shrink_dcache_sb(sbi->sb);
 }
 
 static int autofs4_write(struct file *file, const void *addr, int bytes)
maintainer Acked-by on v1 2/2
#3 lore raw

Re: [PATCH 2/2] AUTOFS: Make sure all dentries refs are released before calling kill_anon_super()

In reply to David Howells

Thanks for you patience.
Carried out a number of basic autofs tests without problem.

So in so far as autofs4 is concerned:

Quoted 9 lines · David Howells

From: David Howells <[email protected]>

Make sure all dentries refs are released before calling kill_anon_super() so that the assumption that generic_shutdown_super() can completely destroy the dentry tree for there will be no external references holds true.

What was being done in the put_super() superblock op, is now done in the kill_sb() filesystem op instead, prior to calling kill_anon_super().

Quoted 6 lines

This makes the struct autofs_sb_info::root member variable redundant (since sb->s_root is still available), and so that is removed. The calls to shrink_dcache_sb() are also removed since they're also redundant as shrink_dcache_for_umount() will now be called after the cleanup routine.

Quoted 128 lines

---

 fs/autofs4/autofs_i.h |    3 +--
 fs/autofs4/init.c     |    2 +-
 fs/autofs4/inode.c    |   22 ++++------------------
 fs/autofs4/waitq.c    |    1 -
 4 files changed, 6 insertions(+), 22 deletions(-)
fs/autofs4/autofs_i.h +1 -2
index d6603d0..47e38f3 100644
@@ -96,7 +96,6 @@ #define AUTOFS_TYPE_OFFSET       0x0004
 
 struct autofs_sb_info {
 	u32 magic;
-	struct dentry *root;
 	int pipefd;
 	struct file *pipe;
 	pid_t oz_pgrp;
@@ -231,4 +230,4 @@ out:
 }
 
 void autofs4_dentry_release(struct dentry *);
-
+extern void autofs4_kill_sb(struct super_block *);
fs/autofs4/init.c +1 -1
index 5d91933..723a1c5 100644
@@ -24,7 +24,7 @@ static struct file_system_type autofs_fs
 	.owner		= THIS_MODULE,
 	.name		= "autofs",
 	.get_sb		= autofs_get_sb,
-	.kill_sb	= kill_anon_super,
+	.kill_sb	= autofs4_kill_sb,
 };
 
 static int __init init_autofs4_fs(void)
fs/autofs4/inode.c +4 -18
index fde78b1..1bf68c5 100644
@@ -95,7 +95,7 @@ void autofs4_free_ino(struct autofs_info
  */
 static void autofs4_force_release(struct autofs_sb_info *sbi)
 {
-	struct dentry *this_parent = sbi->root;
+	struct dentry *this_parent = sbi->sb->s_root;
 	struct list_head *next;
 
 	spin_lock(&dcache_lock);
@@ -126,7 +126,7 @@ resume:
 		spin_lock(&dcache_lock);
 	}
 
-	if (this_parent != sbi->root) {
+	if (this_parent != sbi->sb->s_root) {
 		struct dentry *dentry = this_parent;
 
 		next = this_parent->d_u.d_child.next;
@@ -139,15 +139,9 @@ resume:
 		goto resume;
 	}
 	spin_unlock(&dcache_lock);
-
-	dput(sbi->root);
-	sbi->root = NULL;
-	shrink_dcache_sb(sbi->sb);
-
-	return;
 }
 
-static void autofs4_put_super(struct super_block *sb)
+void autofs4_kill_sb(struct super_block *sb)
 {
 	struct autofs_sb_info *sbi = autofs4_sbi(sb);
 
@@ -162,6 +156,7 @@ static void autofs4_put_super(struct sup
 	kfree(sbi);
 
 	DPRINTK("shutting down");
+	kill_anon_super(sb);
 }
 
 static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt)
@@ -188,7 +183,6 @@ static int autofs4_show_options(struct s
 }
 
 static struct super_operations autofs4_sops = {
-	.put_super	= autofs4_put_super,
 	.statfs		= simple_statfs,
 	.show_options	= autofs4_show_options,
 };
@@ -314,7 +308,6 @@ int autofs4_fill_super(struct super_bloc
 
 	s->s_fs_info = sbi;
 	sbi->magic = AUTOFS_SBI_MAGIC;
-	sbi->root = NULL;
 	sbi->pipefd = -1;
 	sbi->catatonic = 0;
 	sbi->exp_timeout = 0;
@@ -396,13 +389,6 @@ int autofs4_fill_super(struct super_bloc
 	sbi->pipefd = pipefd;
 
 	/*
-	 * Take a reference to the root dentry so we get a chance to
-	 * clean up the dentry tree on umount.
-	 * See autofs4_force_release.
-	 */
-	sbi->root = dget(root);
-
-	/*
 	 * Success! Install the root dentry now to indicate completion.
 	 */
 	s->s_root = root;
fs/autofs4/waitq.c +0 -1
index ce103e7..c0a6c8d 100644
@@ -45,7 +45,6 @@ void autofs4_catatonic_mode(struct autof
 		fput(sbi->pipe);	/* Close the pipe */
 		sbi->pipe = NULL;
 	}
-	shrink_dcache_sb(sbi->sb);
 }
 
 static int autofs4_write(struct file *file, const void *addr, int bytes)
maintainer Acked-by on v1 1/2
#4 lore raw
In reply to David Howells

Thanks for you patience.
Carried out a number of basic autofs tests without problem.

So in so far as autofs4 is concerned:

Quoted 39 lines · David Howells

From: David Howells <[email protected]>

The attached patch destroys all the dentries attached to a superblock in one go by:

(1) Destroying the tree rooted at s_root.

(2) Destroying every entry in the anon list, one at a time.

(3) Each entry in the anon list has its subtree consumed from the leaves inwards.

This reduces the amount of work generic_shutdown_super() does, and avoids iterating through the dentry_unused list.

Note that locking is almost entirely absent in the shrink_dcache_for_umount*() functions added by this patch. This is because:

(1) at the point the filesystem calls generic_shutdown_super(), it is not permitted to further touch the superblock's set of dentries, and nor may it remove aliases from inodes;

(2) the dcache memory shrinker now skips dentries that are being unmounted; and

(3) the superblock no longer has any external references through which the VFS can reach it.

Given these points, the only locking we need to do is when we remove dentries from the unused list and the name hashes, which we do a directory's worth at a time.

We also don't need to guard against reference counts going to zero unexpectedly and removing bits of the tree we're working on as nothing else can call dput().

A cut down version of dentry_iput() has been folded into shrink_dcache_for_umount_subtree() function. Apart from not needing to unlock things, it also doesn't need to check for inotify watches.

Quoted 5 lines

In this version of the patch, the complaint about a dentry still being in use has been expanded from a single BUG_ON() and now gives much more information.

Quoted 191 lines

---

 fs/dcache.c            |  133 ++++++++++++++++++++++++++++++++++++++++++++++++
 fs/super.c             |   12 ++--
 include/linux/dcache.h |    1 
 3 files changed, 140 insertions(+), 6 deletions(-)
fs/dcache.c +133 -0
index 48b44a7..ae041da 100644
@@ -548,6 +548,139 @@ repeat:
 }
 
 /*
+ * destroy a single subtree of dentries for unmount
+ * - see the comments on shrink_dcache_for_umount() for a description of the
+ *   locking
+ */
+static void shrink_dcache_for_umount_subtree(struct dentry *dentry)
+{
+	struct dentry *parent;
+
+	BUG_ON(!IS_ROOT(dentry));
+
+	/* detach this root from the system */
+	spin_lock(&dcache_lock);
+	if (!list_empty(&dentry->d_lru)) {
+		dentry_stat.nr_unused--;
+		list_del_init(&dentry->d_lru);
+	}
+	__d_drop(dentry);
+	spin_unlock(&dcache_lock);
+
+	for (;;) {
+		/* descend to the first leaf in the current subtree */
+		while (!list_empty(&dentry->d_subdirs)) {
+			struct dentry *loop;
+
+			/* this is a branch with children - detach all of them
+			 * from the system in one go */
+			spin_lock(&dcache_lock);
+			list_for_each_entry(loop, &dentry->d_subdirs,
+					    d_u.d_child) {
+				if (!list_empty(&loop->d_lru)) {
+					dentry_stat.nr_unused--;
+					list_del_init(&loop->d_lru);
+				}
+
+				__d_drop(loop);
+				cond_resched_lock(&dcache_lock);
+			}
+			spin_unlock(&dcache_lock);
+
+			/* move to the first child */
+			dentry = list_entry(dentry->d_subdirs.next,
+					    struct dentry, d_u.d_child);
+		}
+
+		/* consume the dentries from this leaf up through its parents
+		 * until we find one with children or run out altogether */
+		do {
+			struct inode *inode;
+
+			if (atomic_read(&dentry->d_count) != 0) {
+				printk(KERN_ERR
+				       "BUG: Dentry %p{i=%lx,n=%s}"
+				       " still in use (%d)"
+				       " [unmount of %s %s]\n",
+				       dentry,
+				       dentry->d_inode ?
+				       dentry->d_inode->i_ino : 0UL,
+				       dentry->d_name.name,
+				       atomic_read(&dentry->d_count),
+				       dentry->d_sb->s_type->name,
+				       dentry->d_sb->s_id);
+				BUG();
+			}
+
+			parent = dentry->d_parent;
+			if (parent == dentry)
+				parent = NULL;
+			else
+				atomic_dec(&parent->d_count);
+
+			list_del(&dentry->d_u.d_child);
+			dentry_stat.nr_dentry--;	/* For d_free, below */
+
+			inode = dentry->d_inode;
+			if (inode) {
+#ifdef CONFIG_INOTIFY
+				BUG_ON(!list_empty(&inode->inotify_watches));
+#endif
+				dentry->d_inode = NULL;
+				list_del_init(&dentry->d_alias);
+				if (dentry->d_op && dentry->d_op->d_iput)
+					dentry->d_op->d_iput(dentry, inode);
+				else
+					iput(inode);
+			}
+
+			d_free(dentry);
+
+			/* finished when we fall off the top of the tree,
+			 * otherwise we ascend to the parent and move to the
+			 * next sibling if there is one */
+			if (!parent)
+				return;
+
+			dentry = parent;
+
+		} while (list_empty(&dentry->d_subdirs));
+
+		dentry = list_entry(dentry->d_subdirs.next,
+				    struct dentry, d_u.d_child);
+	}
+}
+
+/*
+ * destroy the dentries attached to a superblock on unmounting
+ * - we don't need to use dentry->d_lock, and only need dcache_lock when
+ *   removing the dentry from the system lists and hashes because:
+ *   - the superblock is detached from all mountings and open files, so the
+ *     dentry trees will not be rearranged by the VFS
+ *   - s_umount is write-locked, so the memory pressure shrinker will ignore
+ *     any dentries belonging to this superblock that it comes across
+ *   - the filesystem itself is no longer permitted to rearrange the dentries
+ *     in this superblock
+ */
+void shrink_dcache_for_umount(struct super_block *sb)
+{
+	struct dentry *dentry;
+
+	if (down_read_trylock(&sb->s_umount))
+		BUG();
+
+	dentry = sb->s_root;
+	sb->s_root = NULL;
+	atomic_dec(&dentry->d_count);
+	shrink_dcache_for_umount_subtree(dentry);
+
+	while (!hlist_empty(&sb->s_anon)) {
+		dentry = hlist_entry(sb->s_anon.first, struct dentry, d_hash);
+		shrink_dcache_for_umount_subtree(dentry);
+	}
+}
+
+/*
  * Search for at least 1 mount point in the dentry's subdirs.
  * We descend to the next level whenever the d_subdirs
  * list is non-empty and continue searching.
fs/super.c +6 -6
index 8a669f6..bd655b1 100644
@@ -222,17 +222,17 @@ static int grab_super(struct super_block
  *	that need destruction out of superblock, call generic_shutdown_super()
  *	and release aforementioned objects.  Note: dentries and inodes _are_
  *	taken care of and do not need specific handling.
+ *
+ *	Upon calling this function, the filesystem may no longer alter or
+ *	rearrange the set of dentries belonging to this super_block, nor may it
+ *	change the attachments of dentries to inodes.
  */
 void generic_shutdown_super(struct super_block *sb)
 {
-	struct dentry *root = sb->s_root;
 	struct super_operations *sop = sb->s_op;
 
-	if (root) {
-		sb->s_root = NULL;
-		shrink_dcache_parent(root);
-		shrink_dcache_sb(sb);
-		dput(root);
+	if (sb->s_root) {
+		shrink_dcache_for_umount(sb);
 		fsync_super(sb);
 		lock_super(sb);
 		sb->s_flags &= ~MS_ACTIVE;
include/linux/dcache.h +1 -0
index 0dd1610..fe8bcd5 100644
@@ -217,6 +217,7 @@ extern struct dentry * d_alloc_anon(stru
 extern struct dentry * d_splice_alias(struct inode *, struct dentry *);
 extern void shrink_dcache_sb(struct super_block *);
 extern void shrink_dcache_parent(struct dentry *);
+extern void shrink_dcache_for_umount(struct super_block *);
 extern int d_invalidate(struct dentry *);
 
 /* only used at mount-time */