#include <linux/fsnotify.h>
#include <linux/slab.h>
#include <linux/init.h>
-#include <linux/smp_lock.h>
#include <linux/hash.h>
#include <linux/cache.h>
#include <linux/module.h>
#include <linux/seqlock.h>
#include <linux/swap.h>
#include <linux/bootmem.h>
+#include "internal.h"
int sysctl_vfs_cache_pressure __read_mostly = 100;
EXPORT_SYMBOL(dcache_lock);
-static kmem_cache_t *dentry_cache __read_mostly;
+static struct kmem_cache *dentry_cache __read_mostly;
#define DNAME_INLINE_LEN (sizeof(struct dentry)-offsetof(struct dentry,d_iname))
.age_limit = 45,
};
-static void d_callback(struct rcu_head *head)
+static void __d_free(struct dentry *dentry)
{
- struct dentry * dentry = container_of(head, struct dentry, d_u.d_rcu);
-
if (dname_external(dentry))
kfree(dentry->d_name.name);
kmem_cache_free(dentry_cache, dentry);
}
+static void d_callback(struct rcu_head *head)
+{
+ struct dentry * dentry = container_of(head, struct dentry, d_u.d_rcu);
+ __d_free(dentry);
+}
+
/*
* no dcache_lock, please. The caller must decrement dentry_stat.nr_dentry
* inside dcache_lock.
{
if (dentry->d_op && dentry->d_op->d_release)
dentry->d_op->d_release(dentry);
- call_rcu(&dentry->d_u.d_rcu, d_callback);
+ /* if dentry was never inserted into hash, immediate free is OK */
+ if (dentry->d_hash.pprev == NULL)
+ __d_free(dentry);
+ else
+ call_rcu(&dentry->d_u.d_rcu, d_callback);
}
/*
}
}
+/**
+ * d_kill - kill dentry and return parent
+ * @dentry: dentry to kill
+ *
+ * Called with dcache_lock and d_lock, releases both. The dentry must
+ * already be unhashed and removed from the LRU.
+ *
+ * If this is the root of the dentry tree, return NULL.
+ */
+static struct dentry *d_kill(struct dentry *dentry)
+{
+ struct dentry *parent;
+
+ list_del(&dentry->d_u.d_child);
+ dentry_stat.nr_dentry--; /* For d_free, below */
+ /*drops the locks, at that point nobody can reach this dentry */
+ dentry_iput(dentry);
+ parent = dentry->d_parent;
+ d_free(dentry);
+ return dentry == parent ? NULL : parent;
+}
+
/*
* This is dput
*
unhash_it:
__d_drop(dentry);
-
-kill_it: {
- struct dentry *parent;
-
- /* If dentry was on d_lru list
- * delete it from there
- */
- if (!list_empty(&dentry->d_lru)) {
- list_del(&dentry->d_lru);
- dentry_stat.nr_unused--;
- }
- list_del(&dentry->d_u.d_child);
- dentry_stat.nr_dentry--; /* For d_free, below */
- /*drops the locks, at that point nobody can reach this dentry */
- dentry_iput(dentry);
- parent = dentry->d_parent;
- d_free(dentry);
- if (dentry == parent)
- return;
- dentry = parent;
- goto repeat;
+kill_it:
+ /* If dentry was on d_lru list
+ * delete it from there
+ */
+ if (!list_empty(&dentry->d_lru)) {
+ list_del(&dentry->d_lru);
+ dentry_stat.nr_unused--;
}
+ dentry = d_kill(dentry);
+ if (dentry)
+ goto repeat;
}
/**
* it can be unhashed only if it has no children, or if it is the root
* of a filesystem.
*
- * If the inode has a DCACHE_DISCONNECTED alias, then prefer
+ * If the inode has an IS_ROOT, DCACHE_DISCONNECTED alias, then prefer
* any other hashed alias over that one unless @want_discon is set,
- * in which case only return a DCACHE_DISCONNECTED alias.
+ * in which case only return an IS_ROOT, DCACHE_DISCONNECTED alias.
*/
static struct dentry * __d_find_alias(struct inode *inode, int want_discon)
prefetch(next);
alias = list_entry(tmp, struct dentry, d_alias);
if (S_ISDIR(inode->i_mode) || !d_unhashed(alias)) {
- if (alias->d_flags & DCACHE_DISCONNECTED)
+ if (IS_ROOT(alias) &&
+ (alias->d_flags & DCACHE_DISCONNECTED))
discon_alias = alias;
else if (!want_discon) {
__dget_locked(alias);
* Throw away a dentry - free the inode, dput the parent. This requires that
* the LRU list has already been removed.
*
+ * If prune_parents is true, try to prune ancestors as well.
+ *
* Called with dcache_lock, drops it and then regains.
* Called with dentry->d_lock held, drops it.
*/
-static void prune_one_dentry(struct dentry * dentry)
+static void prune_one_dentry(struct dentry * dentry, int prune_parents)
{
- struct dentry * parent;
-
__d_drop(dentry);
- list_del(&dentry->d_u.d_child);
- dentry_stat.nr_dentry--; /* For d_free, below */
- dentry_iput(dentry);
- parent = dentry->d_parent;
- d_free(dentry);
- if (parent != dentry)
- dput(parent);
+ dentry = d_kill(dentry);
+ if (!prune_parents) {
+ dput(dentry);
+ spin_lock(&dcache_lock);
+ return;
+ }
+
+ /*
+ * Prune ancestors. Locking is simpler than in dput(),
+ * because dcache_lock needs to be taken anyway.
+ */
spin_lock(&dcache_lock);
+ while (dentry) {
+ if (!atomic_dec_and_lock(&dentry->d_count, &dentry->d_lock))
+ return;
+
+ if (dentry->d_op && dentry->d_op->d_delete)
+ dentry->d_op->d_delete(dentry);
+ if (!list_empty(&dentry->d_lru)) {
+ list_del(&dentry->d_lru);
+ dentry_stat.nr_unused--;
+ }
+ __d_drop(dentry);
+ dentry = d_kill(dentry);
+ spin_lock(&dcache_lock);
+ }
}
/**
* @count: number of entries to try and free
* @sb: if given, ignore dentries for other superblocks
* which are being unmounted.
+ * @prune_parents: if true, try to prune ancestors as well in one go
*
* Shrink the dcache. This is done when we need
* more memory, or simply when we need to unmount
* all the dentries are in use.
*/
-static void prune_dcache(int count, struct super_block *sb)
+static void prune_dcache(int count, struct super_block *sb, int prune_parents)
{
spin_lock(&dcache_lock);
for (; count ; count--) {
* without taking the s_umount lock (I already hold it).
*/
if (sb && dentry->d_sb == sb) {
- prune_one_dentry(dentry);
+ prune_one_dentry(dentry, prune_parents);
continue;
}
/*
s_umount = &dentry->d_sb->s_umount;
if (down_read_trylock(s_umount)) {
if (dentry->d_sb->s_root != NULL) {
- prune_one_dentry(dentry);
+ prune_one_dentry(dentry, prune_parents);
up_read(s_umount);
continue;
}
up_read(s_umount);
}
spin_unlock(&dentry->d_lock);
- /* Cannot remove the first dentry, and it isn't appropriate
- * to move it to the head of the list, so give up, and try
- * later
+ /*
+ * Insert dentry at the head of the list as inserting at the
+ * tail leads to a cycle.
*/
- break;
+ list_add(&dentry->d_lru, &dentry_unused);
+ dentry_stat.nr_unused++;
}
spin_unlock(&dcache_lock);
}
spin_unlock(&dentry->d_lock);
continue;
}
- prune_one_dentry(dentry);
+ prune_one_dentry(dentry, 1);
cond_resched_lock(&dcache_lock);
goto repeat;
}
spin_unlock(&dcache_lock);
}
+/*
+ * 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;
+ unsigned detached = 0;
+
+ 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);
+ detached++;
+
+ inode = dentry->d_inode;
+ if (inode) {
+ 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)
+ goto out;
+
+ dentry = parent;
+
+ } while (list_empty(&dentry->d_subdirs));
+
+ dentry = list_entry(dentry->d_subdirs.next,
+ struct dentry, d_u.d_child);
+ }
+out:
+ /* several dentries were freed, need to correct nr_dentry */
+ spin_lock(&dcache_lock);
+ dentry_stat.nr_dentry -= detached;
+ spin_unlock(&dcache_lock);
+}
+
+/*
+ * 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
int found;
while ((found = select_parent(parent)) != 0)
- prune_dcache(found, parent->d_sb);
+ prune_dcache(found, parent->d_sb, 1);
}
/*
if (nr) {
if (!(gfp_mask & __GFP_FS))
return -1;
- prune_dcache(nr, NULL);
+ prune_dcache(nr, NULL, 1);
}
return (dentry_stat.nr_unused / 100) * sysctl_vfs_cache_pressure;
}
+static struct shrinker dcache_shrinker = {
+ .shrink = shrink_dcache_memory,
+ .seeks = DEFAULT_SEEKS,
+};
+
/**
* d_alloc - allocate a dcache entry
* @parent: parent of entry to allocate
{
struct dentry *new = NULL;
- if (inode) {
+ if (inode && S_ISDIR(inode->i_mode)) {
spin_lock(&dcache_lock);
new = __d_find_alias(inode, 1);
if (new) {
* deleted it.
*/
-/**
- * d_move - move a dentry
+/*
+ * d_move_locked - move a dentry
* @dentry: entry to move
* @target: new dentry
*
* Update the dcache to reflect the move of a file name. Negative
* dcache entries should not be moved in this way.
*/
-
-void d_move(struct dentry * dentry, struct dentry * target)
+static void d_move_locked(struct dentry * dentry, struct dentry * target)
{
struct hlist_head *list;
if (!dentry->d_inode)
printk(KERN_WARNING "VFS: moving negative dcache entry\n");
- spin_lock(&dcache_lock);
write_seqlock(&rename_lock);
/*
* XXXX: do we really need to take target->d_lock?
fsnotify_d_move(dentry);
spin_unlock(&dentry->d_lock);
write_sequnlock(&rename_lock);
+}
+
+/**
+ * d_move - move a dentry
+ * @dentry: entry to move
+ * @target: new dentry
+ *
+ * Update the dcache to reflect the move of a file name. Negative
+ * dcache entries should not be moved in this way.
+ */
+
+void d_move(struct dentry * dentry, struct dentry * target)
+{
+ spin_lock(&dcache_lock);
+ d_move_locked(dentry, target);
+ spin_unlock(&dcache_lock);
+}
+
+/*
+ * Helper that returns 1 if p1 is a parent of p2, else 0
+ */
+static int d_isparent(struct dentry *p1, struct dentry *p2)
+{
+ struct dentry *p;
+
+ for (p = p2; p->d_parent != p; p = p->d_parent) {
+ if (p->d_parent == p1)
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * This helper attempts to cope with remotely renamed directories
+ *
+ * It assumes that the caller is already holding
+ * dentry->d_parent->d_inode->i_mutex and the dcache_lock
+ *
+ * Note: If ever the locking in lock_rename() changes, then please
+ * remember to update this too...
+ *
+ * On return, dcache_lock will have been unlocked.
+ */
+static struct dentry *__d_unalias(struct dentry *dentry, struct dentry *alias)
+{
+ struct mutex *m1 = NULL, *m2 = NULL;
+ struct dentry *ret;
+
+ /* If alias and dentry share a parent, then no extra locks required */
+ if (alias->d_parent == dentry->d_parent)
+ goto out_unalias;
+
+ /* Check for loops */
+ ret = ERR_PTR(-ELOOP);
+ if (d_isparent(alias, dentry))
+ goto out_err;
+
+ /* See lock_rename() */
+ ret = ERR_PTR(-EBUSY);
+ if (!mutex_trylock(&dentry->d_sb->s_vfs_rename_mutex))
+ goto out_err;
+ m1 = &dentry->d_sb->s_vfs_rename_mutex;
+ if (!mutex_trylock(&alias->d_parent->d_inode->i_mutex))
+ goto out_err;
+ m2 = &alias->d_parent->d_inode->i_mutex;
+out_unalias:
+ d_move_locked(alias, dentry);
+ ret = alias;
+out_err:
spin_unlock(&dcache_lock);
+ if (m2)
+ mutex_unlock(m2);
+ if (m1)
+ mutex_unlock(m1);
+ return ret;
}
/*
*/
struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
{
- struct dentry *alias, *actual;
+ struct dentry *actual;
BUG_ON(!d_unhashed(dentry));
goto found_lock;
}
- /* See if a disconnected directory already exists as an anonymous root
- * that we should splice into the tree instead */
- if (S_ISDIR(inode->i_mode) && (alias = __d_find_alias(inode, 1))) {
- spin_lock(&alias->d_lock);
-
- /* Is this a mountpoint that we could splice into our tree? */
- if (IS_ROOT(alias))
- goto connect_mountpoint;
-
- if (alias->d_name.len == dentry->d_name.len &&
- alias->d_parent == dentry->d_parent &&
- memcmp(alias->d_name.name,
- dentry->d_name.name,
- dentry->d_name.len) == 0)
- goto replace_with_alias;
-
- spin_unlock(&alias->d_lock);
-
- /* Doh! Seem to be aliasing directories for some reason... */
- dput(alias);
+ if (S_ISDIR(inode->i_mode)) {
+ struct dentry *alias;
+
+ /* Does an aliased dentry already exist? */
+ alias = __d_find_alias(inode, 0);
+ if (alias) {
+ actual = alias;
+ /* Is this an anonymous mountpoint that we could splice
+ * into our tree? */
+ if (IS_ROOT(alias)) {
+ spin_lock(&alias->d_lock);
+ __d_materialise_dentry(dentry, alias);
+ __d_drop(alias);
+ goto found;
+ }
+ /* Nope, but we must(!) avoid directory aliasing */
+ actual = __d_unalias(dentry, alias);
+ if (IS_ERR(actual))
+ dput(alias);
+ goto out_nolock;
+ }
}
/* Add a unique reference */
_d_rehash(actual);
spin_unlock(&actual->d_lock);
spin_unlock(&dcache_lock);
-
+out_nolock:
if (actual == dentry) {
security_d_instantiate(dentry, inode);
return NULL;
iput(inode);
return actual;
- /* Convert the anonymous/root alias into an ordinary dentry */
-connect_mountpoint:
- __d_materialise_dentry(dentry, alias);
-
- /* Replace the candidate dentry with the alias in the tree */
-replace_with_alias:
- __d_drop(alias);
- actual = alias;
- goto found;
-
shouldnt_be_hashed:
spin_unlock(&dcache_lock);
BUG();
struct vfsmount *rootmnt;
struct dentry *root;
+ /*
+ * We have various synthetic filesystems that never get mounted. On
+ * these filesystems dentries are never used for lookup purposes, and
+ * thus don't need to be hashed. They also don't need a name until a
+ * user wants to identify the object in /proc/pid/fd/. The little hack
+ * below allows us to generate a name for these objects on demand:
+ */
+ if (dentry->d_op && dentry->d_op->d_dname)
+ return dentry->d_op->d_dname(dentry, buf, buflen);
+
read_lock(¤t->fs->lock);
rootmnt = mntget(current->fs->rootmnt);
root = dget(current->fs->root);
return res;
}
+/*
+ * Helper function for dentry_operations.d_dname() members
+ */
+char *dynamic_dname(struct dentry *dentry, char *buffer, int buflen,
+ const char *fmt, ...)
+{
+ va_list args;
+ char temp[64];
+ int sz;
+
+ va_start(args, fmt);
+ sz = vsnprintf(temp, sizeof(temp), fmt, args) + 1;
+ va_end(args);
+
+ if (sz > sizeof(temp) || sz > buflen)
+ return ERR_PTR(-ENAMETOOLONG);
+
+ buffer += buflen - sz;
+ return memcpy(buffer, temp, sz);
+}
+
/*
* NOTE! The user-level library version returns a
* character pointer. The kernel system call just
* but it is probably not worth it because of the cache nature
* of the dcache.
*/
- dentry_cache = kmem_cache_create("dentry_cache",
- sizeof(struct dentry),
- 0,
- (SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|
- SLAB_MEM_SPREAD),
- NULL, NULL);
+ dentry_cache = KMEM_CACHE(dentry,
+ SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|SLAB_MEM_SPREAD);
- set_shrinker(DEFAULT_SEEKS, shrink_dcache_memory);
+ register_shrinker(&dcache_shrinker);
/* Hash may have been set up in dcache_init_early */
if (!hashdist)
}
/* SLAB cache for __getname() consumers */
-kmem_cache_t *names_cachep __read_mostly;
+struct kmem_cache *names_cachep __read_mostly;
/* SLAB cache for file structures */
-kmem_cache_t *filp_cachep __read_mostly;
+struct kmem_cache *filp_cachep __read_mostly;
EXPORT_SYMBOL(d_genocide);
-extern void bdev_cache_init(void);
-extern void chrdev_init(void);
-
void __init vfs_caches_init_early(void)
{
dcache_init_early();
mempages -= reserve;
names_cachep = kmem_cache_create("names_cache", PATH_MAX, 0,
- SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL, NULL);
+ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
filp_cachep = kmem_cache_create("filp", sizeof(struct file), 0,
- SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL, NULL);
+ SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
dcache_init(mempages);
inode_init(mempages);