#include "audit.h"
-#include <linux/inotify.h>
+#include <linux/fsnotify_backend.h>
#include <linux/namei.h>
#include <linux/mount.h>
+#include <linux/kthread.h>
+#include <linux/slab.h>
struct audit_tree;
struct audit_chunk;
struct audit_chunk {
struct list_head hash;
- struct inotify_watch watch;
+ struct fsnotify_mark mark;
struct list_head trees; /* with root here */
int dead;
int count;
* tree is refcounted; one reference for "some rules on rules_list refer to
* it", one for each chunk with pointer to it.
*
- * chunk is refcounted by embedded inotify_watch + .refs (non-zero refcount
+ * chunk is refcounted by embedded fsnotify_mark + .refs (non-zero refcount
* of watch contributes 1 to .refs).
*
* node.index allows to get from node.list to containing chunk.
* that makes a difference. Some.
*/
-static struct inotify_handle *rtree_ih;
+static struct fsnotify_group *audit_tree_group;
static struct audit_tree *alloc_tree(const char *s)
{
atomic_inc(&tree->count);
}
-static void __put_tree(struct rcu_head *rcu)
-{
- struct audit_tree *tree = container_of(rcu, struct audit_tree, head);
- kfree(tree);
-}
-
static inline void put_tree(struct audit_tree *tree)
{
if (atomic_dec_and_test(&tree->count))
- call_rcu(&tree->head, __put_tree);
+ kfree_rcu(tree, head);
}
/* to avoid bringing the entire thing in audit.h */
return tree->pathname;
}
-static struct audit_chunk *alloc_chunk(int count)
-{
- struct audit_chunk *chunk;
- size_t size;
- int i;
-
- size = offsetof(struct audit_chunk, owners) + count * sizeof(struct node);
- chunk = kzalloc(size, GFP_KERNEL);
- if (!chunk)
- return NULL;
-
- INIT_LIST_HEAD(&chunk->hash);
- INIT_LIST_HEAD(&chunk->trees);
- chunk->count = count;
- atomic_long_set(&chunk->refs, 1);
- for (i = 0; i < count; i++) {
- INIT_LIST_HEAD(&chunk->owners[i].list);
- chunk->owners[i].index = i;
- }
- inotify_init_watch(&chunk->watch);
- return chunk;
-}
-
static void free_chunk(struct audit_chunk *chunk)
{
int i;
audit_put_chunk(chunk);
}
+static void audit_tree_destroy_watch(struct fsnotify_mark *entry)
+{
+ struct audit_chunk *chunk = container_of(entry, struct audit_chunk, mark);
+ call_rcu(&chunk->head, __put_chunk);
+}
+
+static struct audit_chunk *alloc_chunk(int count)
+{
+ struct audit_chunk *chunk;
+ size_t size;
+ int i;
+
+ size = offsetof(struct audit_chunk, owners) + count * sizeof(struct node);
+ chunk = kzalloc(size, GFP_KERNEL);
+ if (!chunk)
+ return NULL;
+
+ INIT_LIST_HEAD(&chunk->hash);
+ INIT_LIST_HEAD(&chunk->trees);
+ chunk->count = count;
+ atomic_long_set(&chunk->refs, 1);
+ for (i = 0; i < count; i++) {
+ INIT_LIST_HEAD(&chunk->owners[i].list);
+ chunk->owners[i].index = i;
+ }
+ fsnotify_init_mark(&chunk->mark, audit_tree_destroy_watch);
+ return chunk;
+}
+
enum {HASH_SIZE = 128};
static struct list_head chunk_hash_heads[HASH_SIZE];
static __cacheline_aligned_in_smp DEFINE_SPINLOCK(hash_lock);
return chunk_hash_heads + n % HASH_SIZE;
}
-/* hash_lock is held by caller */
+/* hash_lock & entry->lock is held by caller */
static void insert_hash(struct audit_chunk *chunk)
{
- struct list_head *list = chunk_hash(chunk->watch.inode);
+ struct fsnotify_mark *entry = &chunk->mark;
+ struct list_head *list;
+
+ if (!entry->i.inode)
+ return;
+ list = chunk_hash(entry->i.inode);
list_add_rcu(&chunk->hash, list);
}
struct audit_chunk *p;
list_for_each_entry_rcu(p, list, hash) {
- if (p->watch.inode == inode) {
+ /* mark.inode may have gone NULL, but who cares? */
+ if (p->mark.i.inode == inode) {
atomic_long_inc(&p->refs);
return p;
}
static void untag_chunk(struct node *p)
{
struct audit_chunk *chunk = find_chunk(p);
- struct audit_chunk *new;
+ struct fsnotify_mark *entry = &chunk->mark;
+ struct audit_chunk *new = NULL;
struct audit_tree *owner;
int size = chunk->count - 1;
int i, j;
- if (!pin_inotify_watch(&chunk->watch)) {
- /*
- * Filesystem is shutting down; all watches are getting
- * evicted, just take it off the node list for this
- * tree and let the eviction logics take care of the
- * rest.
- */
- owner = p->owner;
- if (owner->root == chunk) {
- list_del_init(&owner->same_root);
- owner->root = NULL;
- }
- list_del_init(&p->list);
- p->owner = NULL;
- put_tree(owner);
- return;
- }
+ fsnotify_get_mark(entry);
spin_unlock(&hash_lock);
- /*
- * pin_inotify_watch() succeeded, so the watch won't go away
- * from under us.
- */
- mutex_lock(&chunk->watch.inode->inotify_mutex);
- if (chunk->dead) {
- mutex_unlock(&chunk->watch.inode->inotify_mutex);
+ if (size)
+ new = alloc_chunk(size);
+
+ spin_lock(&entry->lock);
+ if (chunk->dead || !entry->i.inode) {
+ spin_unlock(&entry->lock);
+ if (new)
+ free_chunk(new);
goto out;
}
list_del_init(&p->list);
list_del_rcu(&chunk->hash);
spin_unlock(&hash_lock);
- inotify_evict_watch(&chunk->watch);
- mutex_unlock(&chunk->watch.inode->inotify_mutex);
- put_inotify_watch(&chunk->watch);
+ spin_unlock(&entry->lock);
+ fsnotify_destroy_mark(entry);
+ fsnotify_put_mark(entry);
goto out;
}
- new = alloc_chunk(size);
if (!new)
goto Fallback;
- if (inotify_clone_watch(&chunk->watch, &new->watch) < 0) {
+
+ fsnotify_duplicate_mark(&new->mark, entry);
+ if (fsnotify_add_mark(&new->mark, new->mark.group, new->mark.i.inode, NULL, 1)) {
free_chunk(new);
goto Fallback;
}
owner->root = NULL;
}
- for (i = j = 0; i < size; i++, j++) {
+ for (i = j = 0; j <= size; i++, j++) {
struct audit_tree *s;
if (&chunk->owners[j] == p) {
list_del_init(&p->list);
if (!s) /* result of earlier fallback */
continue;
get_tree(s);
- list_replace_init(&chunk->owners[i].list, &new->owners[j].list);
+ list_replace_init(&chunk->owners[j].list, &new->owners[i].list);
}
list_replace_rcu(&chunk->hash, &new->hash);
list_for_each_entry(owner, &new->trees, same_root)
owner->root = new;
spin_unlock(&hash_lock);
- inotify_evict_watch(&chunk->watch);
- mutex_unlock(&chunk->watch.inode->inotify_mutex);
- put_inotify_watch(&chunk->watch);
+ spin_unlock(&entry->lock);
+ fsnotify_destroy_mark(entry);
+ fsnotify_put_mark(entry);
goto out;
Fallback:
p->owner = NULL;
put_tree(owner);
spin_unlock(&hash_lock);
- mutex_unlock(&chunk->watch.inode->inotify_mutex);
+ spin_unlock(&entry->lock);
out:
- unpin_inotify_watch(&chunk->watch);
+ fsnotify_put_mark(entry);
spin_lock(&hash_lock);
}
static int create_chunk(struct inode *inode, struct audit_tree *tree)
{
+ struct fsnotify_mark *entry;
struct audit_chunk *chunk = alloc_chunk(1);
if (!chunk)
return -ENOMEM;
- if (inotify_add_watch(rtree_ih, &chunk->watch, inode, IN_IGNORED | IN_DELETE_SELF) < 0) {
+ entry = &chunk->mark;
+ if (fsnotify_add_mark(entry, audit_tree_group, inode, NULL, 0)) {
free_chunk(chunk);
return -ENOSPC;
}
- mutex_lock(&inode->inotify_mutex);
+ spin_lock(&entry->lock);
spin_lock(&hash_lock);
if (tree->goner) {
spin_unlock(&hash_lock);
chunk->dead = 1;
- inotify_evict_watch(&chunk->watch);
- mutex_unlock(&inode->inotify_mutex);
- put_inotify_watch(&chunk->watch);
+ spin_unlock(&entry->lock);
+ fsnotify_destroy_mark(entry);
+ fsnotify_put_mark(entry);
return 0;
}
chunk->owners[0].index = (1U << 31);
}
insert_hash(chunk);
spin_unlock(&hash_lock);
- mutex_unlock(&inode->inotify_mutex);
+ spin_unlock(&entry->lock);
return 0;
}
/* the first tagged inode becomes root of tree */
static int tag_chunk(struct inode *inode, struct audit_tree *tree)
{
- struct inotify_watch *watch;
+ struct fsnotify_mark *old_entry, *chunk_entry;
struct audit_tree *owner;
struct audit_chunk *chunk, *old;
struct node *p;
int n;
- if (inotify_find_watch(rtree_ih, inode, &watch) < 0)
+ old_entry = fsnotify_find_inode_mark(audit_tree_group, inode);
+ if (!old_entry)
return create_chunk(inode, tree);
- old = container_of(watch, struct audit_chunk, watch);
+ old = container_of(old_entry, struct audit_chunk, mark);
/* are we already there? */
spin_lock(&hash_lock);
for (n = 0; n < old->count; n++) {
if (old->owners[n].owner == tree) {
spin_unlock(&hash_lock);
- put_inotify_watch(watch);
+ fsnotify_put_mark(old_entry);
return 0;
}
}
spin_unlock(&hash_lock);
chunk = alloc_chunk(old->count + 1);
- if (!chunk)
+ if (!chunk) {
+ fsnotify_put_mark(old_entry);
return -ENOMEM;
+ }
+
+ chunk_entry = &chunk->mark;
+
+ spin_lock(&old_entry->lock);
+ if (!old_entry->i.inode) {
+ /* old_entry is being shot, lets just lie */
+ spin_unlock(&old_entry->lock);
+ fsnotify_put_mark(old_entry);
+ free_chunk(chunk);
+ return -ENOENT;
+ }
- mutex_lock(&inode->inotify_mutex);
- if (inotify_clone_watch(&old->watch, &chunk->watch) < 0) {
- mutex_unlock(&inode->inotify_mutex);
+ fsnotify_duplicate_mark(chunk_entry, old_entry);
+ if (fsnotify_add_mark(chunk_entry, chunk_entry->group, chunk_entry->i.inode, NULL, 1)) {
+ spin_unlock(&old_entry->lock);
free_chunk(chunk);
+ fsnotify_put_mark(old_entry);
return -ENOSPC;
}
+
+ /* even though we hold old_entry->lock, this is safe since chunk_entry->lock could NEVER have been grabbed before */
+ spin_lock(&chunk_entry->lock);
spin_lock(&hash_lock);
+
+ /* we now hold old_entry->lock, chunk_entry->lock, and hash_lock */
if (tree->goner) {
spin_unlock(&hash_lock);
chunk->dead = 1;
- inotify_evict_watch(&chunk->watch);
- mutex_unlock(&inode->inotify_mutex);
- put_inotify_watch(&chunk->watch);
+ spin_unlock(&chunk_entry->lock);
+ spin_unlock(&old_entry->lock);
+
+ fsnotify_destroy_mark(chunk_entry);
+
+ fsnotify_put_mark(chunk_entry);
+ fsnotify_put_mark(old_entry);
return 0;
}
list_replace_init(&old->trees, &chunk->trees);
list_add(&tree->same_root, &chunk->trees);
}
spin_unlock(&hash_lock);
- inotify_evict_watch(&old->watch);
- mutex_unlock(&inode->inotify_mutex);
- put_inotify_watch(&old->watch);
+ spin_unlock(&chunk_entry->lock);
+ spin_unlock(&old_entry->lock);
+ fsnotify_destroy_mark(old_entry);
+ fsnotify_put_mark(old_entry); /* pair to fsnotify_find mark_entry */
+ fsnotify_put_mark(old_entry); /* and kill it */
return 0;
}
if (rule->tree) {
/* not a half-baked one */
ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE);
- audit_log_format(ab, "op=remove rule dir=");
+ audit_log_format(ab, "op=");
+ audit_log_string(ab, "remove rule");
+ audit_log_format(ab, " dir=");
audit_log_untrustedstring(ab, rule->tree->pathname);
- if (rule->filterkey) {
- audit_log_format(ab, " key=");
- audit_log_untrustedstring(ab, rule->filterkey);
- } else
- audit_log_format(ab, " key=(null)");
+ audit_log_key(ab, rule->filterkey);
audit_log_format(ab, " list=%d res=1", rule->listnr);
audit_log_end(ab);
rule->tree = NULL;
}
}
+static void audit_schedule_prune(void);
+
/* called with audit_filter_mutex */
int audit_remove_tree_rule(struct audit_krule *rule)
{
return 0;
}
+static int compare_root(struct vfsmount *mnt, void *arg)
+{
+ return mnt->mnt_root->d_inode == arg;
+}
+
void audit_trim_trees(void)
{
struct list_head cursor;
struct path path;
struct vfsmount *root_mnt;
struct node *node;
- struct list_head list;
int err;
tree = container_of(cursor.next, struct audit_tree, list);
if (err)
goto skip_it;
- root_mnt = collect_mounts(path.mnt, path.dentry);
+ root_mnt = collect_mounts(&path);
path_put(&path);
if (!root_mnt)
goto skip_it;
- list_add_tail(&list, &root_mnt->mnt_list);
spin_lock(&hash_lock);
list_for_each_entry(node, &tree->chunks, list) {
struct audit_chunk *chunk = find_chunk(node);
- struct inode *inode = chunk->watch.inode;
- struct vfsmount *mnt;
+ /* this could be NULL if the watch is dying else where... */
+ struct inode *inode = chunk->mark.i.inode;
node->index |= 1U<<31;
- list_for_each_entry(mnt, &list, mnt_list) {
- if (mnt->mnt_root->d_inode == inode) {
- node->index &= ~(1U<<31);
- break;
- }
- }
+ if (iterate_mounts(compare_root, inode, root_mnt))
+ node->index &= ~(1U<<31);
}
spin_unlock(&hash_lock);
trim_marked(tree);
put_tree(tree);
- list_del_init(&list);
drop_collected_mounts(root_mnt);
skip_it:
mutex_lock(&audit_filter_mutex);
mutex_unlock(&audit_filter_mutex);
}
-static int is_under(struct vfsmount *mnt, struct dentry *dentry,
- struct path *path)
-{
- if (mnt != path->mnt) {
- for (;;) {
- if (mnt->mnt_parent == mnt)
- return 0;
- if (mnt->mnt_parent == path->mnt)
- break;
- mnt = mnt->mnt_parent;
- }
- dentry = mnt->mnt_mountpoint;
- }
- return is_subdir(dentry, path->dentry);
-}
-
int audit_make_tree(struct audit_krule *rule, char *pathname, u32 op)
{
put_tree(tree);
}
+static int tag_mount(struct vfsmount *mnt, void *arg)
+{
+ return tag_chunk(mnt->mnt_root->d_inode, arg);
+}
+
/* called with audit_filter_mutex */
int audit_add_tree_rule(struct audit_krule *rule)
{
struct audit_tree *seed = rule->tree, *tree;
struct path path;
- struct vfsmount *mnt, *p;
- struct list_head list;
+ struct vfsmount *mnt;
int err;
list_for_each_entry(tree, &tree_list, list) {
err = kern_path(tree->pathname, 0, &path);
if (err)
goto Err;
- mnt = collect_mounts(path.mnt, path.dentry);
+ mnt = collect_mounts(&path);
path_put(&path);
if (!mnt) {
err = -ENOMEM;
goto Err;
}
- list_add_tail(&list, &mnt->mnt_list);
get_tree(tree);
- list_for_each_entry(p, &list, mnt_list) {
- err = tag_chunk(p->mnt_root->d_inode, tree);
- if (err)
- break;
- }
-
- list_del(&list);
+ err = iterate_mounts(tag_mount, tree, mnt);
drop_collected_mounts(mnt);
if (!err) {
{
struct list_head cursor, barrier;
int failed = 0;
- struct path path;
+ struct path path1, path2;
struct vfsmount *tagged;
- struct list_head list;
- struct vfsmount *mnt;
- struct dentry *dentry;
int err;
- err = kern_path(new, 0, &path);
+ err = kern_path(new, 0, &path2);
if (err)
return err;
- tagged = collect_mounts(path.mnt, path.dentry);
- path_put(&path);
+ tagged = collect_mounts(&path2);
+ path_put(&path2);
if (!tagged)
return -ENOMEM;
- err = kern_path(old, 0, &path);
+ err = kern_path(old, 0, &path1);
if (err) {
drop_collected_mounts(tagged);
return err;
}
- mnt = mntget(path.mnt);
- dentry = dget(path.dentry);
- path_put(&path);
-
- if (dentry == tagged->mnt_root && dentry == mnt->mnt_root)
- follow_up(&mnt, &dentry);
-
- list_add_tail(&list, &tagged->mnt_list);
mutex_lock(&audit_filter_mutex);
list_add(&barrier, &tree_list);
while (cursor.next != &tree_list) {
struct audit_tree *tree;
- struct vfsmount *p;
+ int good_one = 0;
tree = container_of(cursor.next, struct audit_tree, list);
get_tree(tree);
list_add(&cursor, &tree->list);
mutex_unlock(&audit_filter_mutex);
- err = kern_path(tree->pathname, 0, &path);
- if (err) {
- put_tree(tree);
- mutex_lock(&audit_filter_mutex);
- continue;
+ err = kern_path(tree->pathname, 0, &path2);
+ if (!err) {
+ good_one = path_is_under(&path1, &path2);
+ path_put(&path2);
}
- spin_lock(&vfsmount_lock);
- if (!is_under(mnt, dentry, &path)) {
- spin_unlock(&vfsmount_lock);
- path_put(&path);
+ if (!good_one) {
put_tree(tree);
mutex_lock(&audit_filter_mutex);
continue;
}
- spin_unlock(&vfsmount_lock);
- path_put(&path);
-
- list_for_each_entry(p, &list, mnt_list) {
- failed = tag_chunk(p->mnt_root->d_inode, tree);
- if (failed)
- break;
- }
+ failed = iterate_mounts(tag_mount, tree, tagged);
if (failed) {
put_tree(tree);
mutex_lock(&audit_filter_mutex);
}
list_del(&barrier);
list_del(&cursor);
- list_del(&list);
mutex_unlock(&audit_filter_mutex);
- dput(dentry);
- mntput(mnt);
+ path_put(&path1);
drop_collected_mounts(tagged);
return failed;
}
/*
* That gets run when evict_chunk() ends up needing to kill audit_tree.
- * Runs from a separate thread, with audit_cmd_mutex held.
+ * Runs from a separate thread.
*/
-void audit_prune_trees(void)
+static int prune_tree_thread(void *unused)
{
+ mutex_lock(&audit_cmd_mutex);
mutex_lock(&audit_filter_mutex);
while (!list_empty(&prune_list)) {
}
mutex_unlock(&audit_filter_mutex);
+ mutex_unlock(&audit_cmd_mutex);
+ return 0;
+}
+
+static void audit_schedule_prune(void)
+{
+ kthread_run(prune_tree_thread, NULL, "audit_prune_tree");
+}
+
+/*
+ * ... and that one is done if evict_chunk() decides to delay until the end
+ * of syscall. Runs synchronously.
+ */
+void audit_kill_trees(struct list_head *list)
+{
+ mutex_lock(&audit_cmd_mutex);
+ mutex_lock(&audit_filter_mutex);
+
+ while (!list_empty(list)) {
+ struct audit_tree *victim;
+
+ victim = list_entry(list->next, struct audit_tree, list);
+ kill_rules(victim);
+ list_del_init(&victim->list);
+
+ mutex_unlock(&audit_filter_mutex);
+
+ prune_one(victim);
+
+ mutex_lock(&audit_filter_mutex);
+ }
+
+ mutex_unlock(&audit_filter_mutex);
+ mutex_unlock(&audit_cmd_mutex);
}
/*
* Here comes the stuff asynchronous to auditctl operations
*/
-/* inode->inotify_mutex is locked */
static void evict_chunk(struct audit_chunk *chunk)
{
struct audit_tree *owner;
+ struct list_head *postponed = audit_killed_trees();
+ int need_prune = 0;
int n;
if (chunk->dead)
owner->root = NULL;
list_del_init(&owner->same_root);
spin_unlock(&hash_lock);
- kill_rules(owner);
- list_move(&owner->list, &prune_list);
- audit_schedule_prune();
+ if (!postponed) {
+ kill_rules(owner);
+ list_move(&owner->list, &prune_list);
+ need_prune = 1;
+ } else {
+ list_move(&owner->list, postponed);
+ }
spin_lock(&hash_lock);
}
list_del_rcu(&chunk->hash);
for (n = 0; n < chunk->count; n++)
list_del_init(&chunk->owners[n].list);
spin_unlock(&hash_lock);
+ if (need_prune)
+ audit_schedule_prune();
mutex_unlock(&audit_filter_mutex);
}
-static void handle_event(struct inotify_watch *watch, u32 wd, u32 mask,
- u32 cookie, const char *dname, struct inode *inode)
+static int audit_tree_handle_event(struct fsnotify_group *group,
+ struct fsnotify_mark *inode_mark,
+ struct fsnotify_mark *vfsmonut_mark,
+ struct fsnotify_event *event)
+{
+ BUG();
+ return -EOPNOTSUPP;
+}
+
+static void audit_tree_freeing_mark(struct fsnotify_mark *entry, struct fsnotify_group *group)
{
- struct audit_chunk *chunk = container_of(watch, struct audit_chunk, watch);
+ struct audit_chunk *chunk = container_of(entry, struct audit_chunk, mark);
- if (mask & IN_IGNORED) {
- evict_chunk(chunk);
- put_inotify_watch(watch);
- }
+ evict_chunk(chunk);
+ fsnotify_put_mark(entry);
}
-static void destroy_watch(struct inotify_watch *watch)
+static bool audit_tree_send_event(struct fsnotify_group *group, struct inode *inode,
+ struct fsnotify_mark *inode_mark,
+ struct fsnotify_mark *vfsmount_mark,
+ __u32 mask, void *data, int data_type)
{
- struct audit_chunk *chunk = container_of(watch, struct audit_chunk, watch);
- call_rcu(&chunk->head, __put_chunk);
+ return false;
}
-static const struct inotify_operations rtree_inotify_ops = {
- .handle_event = handle_event,
- .destroy_watch = destroy_watch,
+static const struct fsnotify_ops audit_tree_ops = {
+ .handle_event = audit_tree_handle_event,
+ .should_send_event = audit_tree_send_event,
+ .free_group_priv = NULL,
+ .free_event_priv = NULL,
+ .freeing_mark = audit_tree_freeing_mark,
};
static int __init audit_tree_init(void)
{
int i;
- rtree_ih = inotify_init(&rtree_inotify_ops);
- if (IS_ERR(rtree_ih))
- audit_panic("cannot initialize inotify handle for rectree watches");
+ audit_tree_group = fsnotify_alloc_group(&audit_tree_ops);
+ if (IS_ERR(audit_tree_group))
+ audit_panic("cannot initialize fsnotify group for rectree watches");
for (i = 0; i < HASH_SIZE; i++)
INIT_LIST_HEAD(&chunk_hash_heads[i]);
Code Review / linux-2.6.git / blobdiff