Merge branch 'audit.b43' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/audit-current

* 'audit.b43' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/audit-current:
  [PATCH] audit: watching subtrees
  [PATCH] new helper - inotify_evict_watch()
  [PATCH] new helper - inotify_clone_watch()
  [PATCH] new helpers - collect_mounts() and release_collected_mounts()
  [PATCH] pass dentry to audit_inode()/audit_inode_child()

fs/dcache.c

@@ -38,7 +38,7 @@ int sysctl_vfs_cache_pressure __read_mostly = 100;
 EXPORT_SYMBOL_GPL(sysctl_vfs_cache_pressure);
 
  __cacheline_aligned_in_smp DEFINE_SPINLOCK(dcache_lock);
-static __cacheline_aligned_in_smp DEFINE_SEQLOCK(rename_lock);
+__cacheline_aligned_in_smp DEFINE_SEQLOCK(rename_lock);
 
 EXPORT_SYMBOL(dcache_lock);
 

fs/debugfs/inode.c

@@ -413,7 +413,7 @@ struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
 	d_move(old_dentry, dentry);
 	fsnotify_move(old_dir->d_inode, new_dir->d_inode, old_name,
 		old_dentry->d_name.name, S_ISDIR(old_dentry->d_inode->i_mode),
-		NULL, old_dentry->d_inode);
+		NULL, old_dentry);
 	fsnotify_oldname_free(old_name);
 	unlock_rename(new_dir, old_dir);
 	dput(dentry);

fs/inotify.c

@@ -666,6 +666,49 @@ s32 inotify_add_watch(struct inotify_handle *ih, struct inotify_watch *watch,
 }
 EXPORT_SYMBOL_GPL(inotify_add_watch);
 
+/**
+ * inotify_clone_watch - put the watch next to existing one
+ * @old: already installed watch
+ * @new: new watch
+ *
+ * Caller must hold the inotify_mutex of inode we are dealing with;
+ * it is expected to remove the old watch before unlocking the inode.
+ */
+s32 inotify_clone_watch(struct inotify_watch *old, struct inotify_watch *new)
+{
+	struct inotify_handle *ih = old->ih;
+	int ret = 0;
+
+	new->mask = old->mask;
+	new->ih = ih;
+
+	mutex_lock(&ih->mutex);
+
+	/* Initialize a new watch */
+	ret = inotify_handle_get_wd(ih, new);
+	if (unlikely(ret))
+		goto out;
+	ret = new->wd;
+
+	get_inotify_handle(ih);
+
+	new->inode = igrab(old->inode);
+
+	list_add(&new->h_list, &ih->watches);
+	list_add(&new->i_list, &old->inode->inotify_watches);
+out:
+	mutex_unlock(&ih->mutex);
+	return ret;
+}
+
+void inotify_evict_watch(struct inotify_watch *watch)
+{
+	get_inotify_watch(watch);
+	mutex_lock(&watch->ih->mutex);
+	inotify_remove_watch_locked(watch->ih, watch);
+	mutex_unlock(&watch->ih->mutex);
+}
+
 /**
  * inotify_rm_wd - remove a watch from an inotify instance
  * @ih: inotify handle

fs/namei.c

@@ -1174,7 +1174,7 @@ static int fastcall do_path_lookup(int dfd, const char *name,
 out:
 	if (unlikely(!retval && !audit_dummy_context() && nd->dentry &&
 				nd->dentry->d_inode))
-		audit_inode(name, nd->dentry->d_inode);
+		audit_inode(name, nd->dentry);
 out_fail:
 	return retval;
 
@@ -1214,7 +1214,7 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
 	retval = path_walk(name, nd);
 	if (unlikely(!retval && !audit_dummy_context() && nd->dentry &&
 				nd->dentry->d_inode))
-		audit_inode(name, nd->dentry->d_inode);
+		audit_inode(name, nd->dentry);
 
 	return retval;
 
@@ -1469,7 +1469,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir)
 		return -ENOENT;
 
 	BUG_ON(victim->d_parent->d_inode != dir);
-	audit_inode_child(victim->d_name.name, victim->d_inode, dir);
+	audit_inode_child(victim->d_name.name, victim, dir);
 
 	error = permission(dir,MAY_WRITE | MAY_EXEC, NULL);
 	if (error)
@@ -1783,7 +1783,7 @@ int open_namei(int dfd, const char *pathname, int flag,
 	 * It already exists.
 	 */
 	mutex_unlock(&dir->d_inode->i_mutex);
-	audit_inode(pathname, path.dentry->d_inode);
+	audit_inode(pathname, path.dentry);
 
 	error = -EEXIST;
 	if (flag & O_EXCL)
@@ -2562,7 +2562,7 @@ int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (!error) {
 		const char *new_name = old_dentry->d_name.name;
 		fsnotify_move(old_dir, new_dir, old_name, new_name, is_dir,
-			      new_dentry->d_inode, old_dentry->d_inode);
+			      new_dentry->d_inode, old_dentry);
 	}
 	fsnotify_oldname_free(old_name);
 

fs/namespace.c

@@ -246,7 +246,7 @@ static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
 			list_add(&mnt->mnt_slave, &old->mnt_slave_list);
 			mnt->mnt_master = old;
 			CLEAR_MNT_SHARED(mnt);
-		} else {
+		} else if (!(flag & CL_PRIVATE)) {
 			if ((flag & CL_PROPAGATION) || IS_MNT_SHARED(old))
 				list_add(&mnt->mnt_share, &old->mnt_share);
 			if (IS_MNT_SLAVE(old))
@@ -746,6 +746,26 @@ struct vfsmount *copy_tree(struct vfsmount *mnt, struct dentry *dentry,
 	return NULL;
 }
 
+struct vfsmount *collect_mounts(struct vfsmount *mnt, struct dentry *dentry)
+{
+	struct vfsmount *tree;
+	down_read(&namespace_sem);
+	tree = copy_tree(mnt, dentry, CL_COPY_ALL | CL_PRIVATE);
+	up_read(&namespace_sem);
+	return tree;
+}
+
+void drop_collected_mounts(struct vfsmount *mnt)
+{
+	LIST_HEAD(umount_list);
+	down_read(&namespace_sem);
+	spin_lock(&vfsmount_lock);
+	umount_tree(mnt, 0, &umount_list);
+	spin_unlock(&vfsmount_lock);
+	up_read(&namespace_sem);
+	release_mounts(&umount_list);
+}
+
 /*
  *  @source_mnt : mount tree to be attached
  *  @nd         : place the mount tree @source_mnt is attached

fs/open.c

@@ -569,7 +569,7 @@ asmlinkage long sys_fchmod(unsigned int fd, mode_t mode)
 	dentry = file->f_path.dentry;
 	inode = dentry->d_inode;
 
-	audit_inode(NULL, inode);
+	audit_inode(NULL, dentry);
 
 	err = -EROFS;
 	if (IS_RDONLY(inode))
@@ -727,7 +727,7 @@ asmlinkage long sys_fchown(unsigned int fd, uid_t user, gid_t group)
 		goto out;
 
 	dentry = file->f_path.dentry;
-	audit_inode(NULL, dentry->d_inode);
+	audit_inode(NULL, dentry);
 	error = chown_common(dentry, user, group);
 	fput(file);
 out:

fs/pnode.h

@@ -22,6 +22,7 @@
 #define CL_COPY_ALL 		0x04
 #define CL_MAKE_SHARED 		0x08
 #define CL_PROPAGATION 		0x10
+#define CL_PRIVATE 		0x20
 
 static inline void set_mnt_shared(struct vfsmount *mnt)
 {

fs/xattr.c

@@ -267,7 +267,7 @@ sys_fsetxattr(int fd, char __user *name, void __user *value,
 	if (!f)
 		return error;
 	dentry = f->f_path.dentry;
-	audit_inode(NULL, dentry->d_inode);
+	audit_inode(NULL, dentry);
 	error = setxattr(dentry, name, value, size, flags);
 	fput(f);
 	return error;
@@ -349,7 +349,7 @@ sys_fgetxattr(int fd, char __user *name, void __user *value, size_t size)
 	f = fget(fd);
 	if (!f)
 		return error;
-	audit_inode(NULL, f->f_path.dentry->d_inode);
+	audit_inode(NULL, f->f_path.dentry);
 	error = getxattr(f->f_path.dentry, name, value, size);
 	fput(f);
 	return error;
@@ -422,7 +422,7 @@ sys_flistxattr(int fd, char __user *list, size_t size)
 	f = fget(fd);
 	if (!f)
 		return error;
-	audit_inode(NULL, f->f_path.dentry->d_inode);
+	audit_inode(NULL, f->f_path.dentry);
 	error = listxattr(f->f_path.dentry, list, size);
 	fput(f);
 	return error;
@@ -485,7 +485,7 @@ sys_fremovexattr(int fd, char __user *name)
 	if (!f)
 		return error;
 	dentry = f->f_path.dentry;
-	audit_inode(NULL, dentry->d_inode);
+	audit_inode(NULL, dentry);
 	error = removexattr(dentry, name);
 	fput(f);
 	return error;

include/linux/audit.h

@@ -63,6 +63,8 @@
 #define AUDIT_ADD_RULE		1011	/* Add syscall filtering rule */
 #define AUDIT_DEL_RULE		1012	/* Delete syscall filtering rule */
 #define AUDIT_LIST_RULES	1013	/* List syscall filtering rules */
+#define AUDIT_TRIM		1014	/* Trim junk from watched tree */
+#define AUDIT_MAKE_EQUIV	1015	/* Append to watched tree */
 #define AUDIT_TTY_GET		1016	/* Get TTY auditing status */
 #define AUDIT_TTY_SET		1017	/* Set TTY auditing status */
 
@@ -203,6 +205,7 @@
 #define AUDIT_SUCCESS   104	/* exit >= 0; value ignored */
 #define AUDIT_WATCH	105
 #define AUDIT_PERM	106
+#define AUDIT_DIR	107
 
 #define AUDIT_ARG0      200
 #define AUDIT_ARG1      (AUDIT_ARG0+1)
@@ -366,8 +369,8 @@ extern void audit_syscall_entry(int arch,
 extern void audit_syscall_exit(int failed, long return_code);
 extern void __audit_getname(const char *name);
 extern void audit_putname(const char *name);
-extern void __audit_inode(const char *name, const struct inode *inode);
-extern void __audit_inode_child(const char *dname, const struct inode *inode,
+extern void __audit_inode(const char *name, const struct dentry *dentry);
+extern void __audit_inode_child(const char *dname, const struct dentry *dentry,
 				const struct inode *parent);
 extern void __audit_ptrace(struct task_struct *t);
 
@@ -381,15 +384,15 @@ static inline void audit_getname(const char *name)
 	if (unlikely(!audit_dummy_context()))
 		__audit_getname(name);
 }
-static inline void audit_inode(const char *name, const struct inode *inode) {
+static inline void audit_inode(const char *name, const struct dentry *dentry) {
 	if (unlikely(!audit_dummy_context()))
-		__audit_inode(name, inode);
+		__audit_inode(name, dentry);
 }
 static inline void audit_inode_child(const char *dname, 
-				     const struct inode *inode,
+				     const struct dentry *dentry,
 				     const struct inode *parent) {
 	if (unlikely(!audit_dummy_context()))
-		__audit_inode_child(dname, inode, parent);
+		__audit_inode_child(dname, dentry, parent);
 }
 void audit_core_dumps(long signr);
 
@@ -477,9 +480,9 @@ extern int audit_signals;
 #define audit_dummy_context() 1
 #define audit_getname(n) do { ; } while (0)
 #define audit_putname(n) do { ; } while (0)
-#define __audit_inode(n,i) do { ; } while (0)
+#define __audit_inode(n,d) do { ; } while (0)
 #define __audit_inode_child(d,i,p) do { ; } while (0)
-#define audit_inode(n,i) do { ; } while (0)
+#define audit_inode(n,d) do { ; } while (0)
 #define audit_inode_child(d,i,p) do { ; } while (0)
 #define audit_core_dumps(i) do { ; } while (0)
 #define auditsc_get_stamp(c,t,s) do { BUG(); } while (0)

include/linux/dcache.h

@@ -178,6 +178,7 @@ d_iput:		no		no		no       yes
 #define DCACHE_INOTIFY_PARENT_WATCHED	0x0020 /* Parent inode is watched */
 
 extern spinlock_t dcache_lock;
+extern seqlock_t rename_lock;
 
 /**
  * d_drop - drop a dentry

include/linux/fs.h

@@ -1470,6 +1470,8 @@ extern long do_mount(char *, char *, char *, unsigned long, void *);
 extern struct vfsmount *copy_tree(struct vfsmount *, struct dentry *, int);
 extern void mnt_set_mountpoint(struct vfsmount *, struct dentry *,
 				  struct vfsmount *);
+extern struct vfsmount *collect_mounts(struct vfsmount *, struct dentry *);
+extern void drop_collected_mounts(struct vfsmount *);
 
 extern int vfs_statfs(struct dentry *, struct kstatfs *);
 

include/linux/fsnotify.h

@@ -41,8 +41,9 @@ static inline void fsnotify_d_move(struct dentry *entry)
  */
 static inline void fsnotify_move(struct inode *old_dir, struct inode *new_dir,
 				 const char *old_name, const char *new_name,
-				 int isdir, struct inode *target, struct inode *source)
+				 int isdir, struct inode *target, struct dentry *moved)
 {
+	struct inode *source = moved->d_inode;
 	u32 cookie = inotify_get_cookie();
 
 	if (old_dir == new_dir)
@@ -67,7 +68,7 @@ static inline void fsnotify_move(struct inode *old_dir, struct inode *new_dir,
 	if (source) {
 		inotify_inode_queue_event(source, IN_MOVE_SELF, 0, NULL, NULL);
 	}
-	audit_inode_child(new_name, source, new_dir);
+	audit_inode_child(new_name, moved, new_dir);
 }
 
 /*
@@ -98,7 +99,7 @@ static inline void fsnotify_create(struct inode *inode, struct dentry *dentry)
 	inode_dir_notify(inode, DN_CREATE);
 	inotify_inode_queue_event(inode, IN_CREATE, 0, dentry->d_name.name,
 				  dentry->d_inode);
-	audit_inode_child(dentry->d_name.name, dentry->d_inode, inode);
+	audit_inode_child(dentry->d_name.name, dentry, inode);
 }
 
 /*
@@ -109,7 +110,7 @@ static inline void fsnotify_mkdir(struct inode *inode, struct dentry *dentry)
 	inode_dir_notify(inode, DN_CREATE);
 	inotify_inode_queue_event(inode, IN_CREATE | IN_ISDIR, 0, 
 				  dentry->d_name.name, dentry->d_inode);
-	audit_inode_child(dentry->d_name.name, dentry->d_inode, inode);
+	audit_inode_child(dentry->d_name.name, dentry, inode);
 }
 
 /*

include/linux/inotify.h

@@ -120,6 +120,8 @@ extern __s32 inotify_find_update_watch(struct inotify_handle *, struct inode *,
 				       u32);
 extern __s32 inotify_add_watch(struct inotify_handle *, struct inotify_watch *,
 			       struct inode *, __u32);
+extern __s32 inotify_clone_watch(struct inotify_watch *, struct inotify_watch *);
+extern void inotify_evict_watch(struct inotify_watch *);
 extern int inotify_rm_watch(struct inotify_handle *, struct inotify_watch *);
 extern int inotify_rm_wd(struct inotify_handle *, __u32);
 extern void inotify_remove_watch_locked(struct inotify_handle *,

init/Kconfig

@@ -234,6 +234,10 @@ config AUDITSYSCALL
 	  such as SELinux.  To use audit's filesystem watch feature, please
 	  ensure that INOTIFY is configured.
 
+config AUDIT_TREE
+	def_bool y
+	depends on AUDITSYSCALL && INOTIFY
+
 config IKCONFIG
 	tristate "Kernel .config support"
 	---help---

ipc/mqueue.c

@@ -676,7 +676,7 @@ asmlinkage long sys_mq_open(const char __user *u_name, int oflag, mode_t mode,
 
 	if (oflag & O_CREAT) {
 		if (dentry->d_inode) {	/* entry already exists */
-			audit_inode(name, dentry->d_inode);
+			audit_inode(name, dentry);
 			error = -EEXIST;
 			if (oflag & O_EXCL)
 				goto out;
@@ -689,7 +689,7 @@ asmlinkage long sys_mq_open(const char __user *u_name, int oflag, mode_t mode,
 		error = -ENOENT;
 		if (!dentry->d_inode)
 			goto out;
-		audit_inode(name, dentry->d_inode);
+		audit_inode(name, dentry);
 		filp = do_open(dentry, oflag);
 	}
 
@@ -837,7 +837,7 @@ asmlinkage long sys_mq_timedsend(mqd_t mqdes, const char __user *u_msg_ptr,
 	if (unlikely(filp->f_op != &mqueue_file_operations))
 		goto out_fput;
 	info = MQUEUE_I(inode);
-	audit_inode(NULL, inode);
+	audit_inode(NULL, filp->f_path.dentry);
 
 	if (unlikely(!(filp->f_mode & FMODE_WRITE)))
 		goto out_fput;
@@ -921,7 +921,7 @@ asmlinkage ssize_t sys_mq_timedreceive(mqd_t mqdes, char __user *u_msg_ptr,
 	if (unlikely(filp->f_op != &mqueue_file_operations))
 		goto out_fput;
 	info = MQUEUE_I(inode);
-	audit_inode(NULL, inode);
+	audit_inode(NULL, filp->f_path.dentry);
 
 	if (unlikely(!(filp->f_mode & FMODE_READ)))
 		goto out_fput;

kernel/Makefile

@@ -46,6 +46,7 @@ obj-$(CONFIG_IKCONFIG) += configs.o
 obj-$(CONFIG_STOP_MACHINE) += stop_machine.o
 obj-$(CONFIG_AUDIT) += audit.o auditfilter.o
 obj-$(CONFIG_AUDITSYSCALL) += auditsc.o
+obj-$(CONFIG_AUDIT_TREE) += audit_tree.o
 obj-$(CONFIG_KPROBES) += kprobes.o
 obj-$(CONFIG_SYSFS) += ksysfs.o
 obj-$(CONFIG_DETECT_SOFTLOCKUP) += softlockup.o

kernel/audit.c

@@ -468,6 +468,21 @@ int audit_send_list(void *_dest)
 	return 0;
 }
 
+#ifdef CONFIG_AUDIT_TREE
+static int prune_tree_thread(void *unused)
+{
+	mutex_lock(&audit_cmd_mutex);
+	audit_prune_trees();
+	mutex_unlock(&audit_cmd_mutex);
+	return 0;
+}
+
+void audit_schedule_prune(void)
+{
+	kthread_run(prune_tree_thread, NULL, "audit_prune_tree");
+}
+#endif
+
 struct sk_buff *audit_make_reply(int pid, int seq, int type, int done,
 				 int multi, void *payload, int size)
 {
@@ -540,6 +555,8 @@ static int audit_netlink_ok(struct sk_buff *skb, u16 msg_type)
 	case AUDIT_SIGNAL_INFO:
 	case AUDIT_TTY_GET:
 	case AUDIT_TTY_SET:
+	case AUDIT_TRIM:
+	case AUDIT_MAKE_EQUIV:
 		if (security_netlink_recv(skb, CAP_AUDIT_CONTROL))
 			err = -EPERM;
 		break;
@@ -756,6 +773,76 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 					   uid, seq, data, nlmsg_len(nlh),
 					   loginuid, sid);
 		break;
+	case AUDIT_TRIM:
+		audit_trim_trees();
+		ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE);
+		if (!ab)
+			break;
+		audit_log_format(ab, "auid=%u", loginuid);
+		if (sid) {
+			u32 len;
+			ctx = NULL;
+			if (selinux_sid_to_string(sid, &ctx, &len))
+				audit_log_format(ab, " ssid=%u", sid);
+			else
+				audit_log_format(ab, " subj=%s", ctx);
+			kfree(ctx);
+		}
+		audit_log_format(ab, " op=trim res=1");
+		audit_log_end(ab);
+		break;
+	case AUDIT_MAKE_EQUIV: {
+		void *bufp = data;
+		u32 sizes[2];
+		size_t len = nlmsg_len(nlh);
+		char *old, *new;
+
+		err = -EINVAL;
+		if (len < 2 * sizeof(u32))
+			break;
+		memcpy(sizes, bufp, 2 * sizeof(u32));
+		bufp += 2 * sizeof(u32);
+		len -= 2 * sizeof(u32);
+		old = audit_unpack_string(&bufp, &len, sizes[0]);
+		if (IS_ERR(old)) {
+			err = PTR_ERR(old);
+			break;
+		}
+		new = audit_unpack_string(&bufp, &len, sizes[1]);
+		if (IS_ERR(new)) {
+			err = PTR_ERR(new);
+			kfree(old);
+			break;
+		}
+		/* OK, here comes... */
+		err = audit_tag_tree(old, new);
+
+		ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE);
+		if (!ab) {
+			kfree(old);
+			kfree(new);
+			break;
+		}
+		audit_log_format(ab, "auid=%u", loginuid);
+		if (sid) {
+			u32 len;
+			ctx = NULL;
+			if (selinux_sid_to_string(sid, &ctx, &len))
+				audit_log_format(ab, " ssid=%u", sid);
+			else
+				audit_log_format(ab, " subj=%s", ctx);
+			kfree(ctx);
+		}
+		audit_log_format(ab, " op=make_equiv old=");
+		audit_log_untrustedstring(ab, old);
+		audit_log_format(ab, " new=");
+		audit_log_untrustedstring(ab, new);
+		audit_log_format(ab, " res=%d", !err);
+		audit_log_end(ab);
+		kfree(old);
+		kfree(new);
+		break;
+	}
 	case AUDIT_SIGNAL_INFO:
 		err = selinux_sid_to_string(audit_sig_sid, &ctx, &len);
 		if (err)

kernel/audit.h

@@ -73,6 +73,9 @@ struct audit_field {
 	struct selinux_audit_rule	*se_rule;
 };
 
+struct audit_tree;
+struct audit_chunk;
+
 struct audit_krule {
 	int			vers_ops;
 	u32			flags;
@@ -86,7 +89,8 @@ struct audit_krule {
 	struct audit_field	*arch_f; /* quick access to arch field */
 	struct audit_field	*inode_f; /* quick access to an inode field */
 	struct audit_watch	*watch;	/* associated watch */
-	struct list_head	rlist;	/* entry in audit_watch.rules list */
+	struct audit_tree	*tree;	/* associated watched tree */
+	struct list_head	rlist;	/* entry in audit_{watch,tree}.rules list */
 };
 
 struct audit_entry {
@@ -130,6 +134,34 @@ extern void audit_handle_ievent(struct inotify_watch *, u32, u32, u32,
 				const char *, struct inode *);
 extern int selinux_audit_rule_update(void);
 
+extern struct mutex audit_filter_mutex;
+extern void audit_free_rule_rcu(struct rcu_head *);
+
+#ifdef CONFIG_AUDIT_TREE
+extern struct audit_chunk *audit_tree_lookup(const struct inode *);
+extern void audit_put_chunk(struct audit_chunk *);
+extern int audit_tree_match(struct audit_chunk *, struct audit_tree *);
+extern int audit_make_tree(struct audit_krule *, char *, u32);
+extern int audit_add_tree_rule(struct audit_krule *);
+extern int audit_remove_tree_rule(struct audit_krule *);
+extern void audit_trim_trees(void);
+extern int audit_tag_tree(char *old, char *new);
+extern void audit_schedule_prune(void);
+extern void audit_prune_trees(void);
+extern const char *audit_tree_path(struct audit_tree *);
+extern void audit_put_tree(struct audit_tree *);
+#else
+#define audit_remove_tree_rule(rule) BUG()
+#define audit_add_tree_rule(rule) -EINVAL
+#define audit_make_tree(rule, str, op) -EINVAL
+#define audit_trim_trees() (void)0
+#define audit_put_tree(tree) (void)0
+#define audit_tag_tree(old, new) -EINVAL
+#define audit_tree_path(rule) ""	/* never called */
+#endif
+
+extern char *audit_unpack_string(void **, size_t *, size_t);
+
 #ifdef CONFIG_AUDITSYSCALL
 extern int __audit_signal_info(int sig, struct task_struct *t);
 static inline int audit_signal_info(int sig, struct task_struct *t)

kernel/audit_tree.c

+#include "audit.h"
+#include <linux/inotify.h>
+#include <linux/namei.h>
+#include <linux/mount.h>
+
+struct audit_tree;
+struct audit_chunk;
+
+struct audit_tree {
+	atomic_t count;
+	int goner;
+	struct audit_chunk *root;
+	struct list_head chunks;
+	struct list_head rules;
+	struct list_head list;
+	struct list_head same_root;
+	struct rcu_head head;
+	char pathname[];
+};
+
+struct audit_chunk {
+	struct list_head hash;
+	struct inotify_watch watch;
+	struct list_head trees;		/* with root here */
+	int dead;
+	int count;
+	struct rcu_head head;
+	struct node {
+		struct list_head list;
+		struct audit_tree *owner;
+		unsigned index;		/* index; upper bit indicates 'will prune' */
+	} owners[];
+};
+
+static LIST_HEAD(tree_list);
+static LIST_HEAD(prune_list);
+
+/*
+ * One struct chunk is attached to each inode of interest.
+ * We replace struct chunk on tagging/untagging.
+ * Rules have pointer to struct audit_tree.
+ * Rules have struct list_head rlist forming a list of rules over
+ * the same tree.
+ * References to struct chunk are collected at audit_inode{,_child}()
+ * time and used in AUDIT_TREE rule matching.
+ * These references are dropped at the same time we are calling
+ * audit_free_names(), etc.
+ *
+ * Cyclic lists galore:
+ * tree.chunks anchors chunk.owners[].list			hash_lock
+ * tree.rules anchors rule.rlist				audit_filter_mutex
+ * chunk.trees anchors tree.same_root				hash_lock
+ * chunk.hash is a hash with middle bits of watch.inode as
+ * a hash function.						RCU, hash_lock
+ *
+ * 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.
+ *
+ * node.index allows to get from node.list to containing chunk.
+ * MSB of that sucker is stolen to mark taggings that we might have to
+ * revert - several operations have very unpleasant cleanup logics and
+ * that makes a difference.  Some.
+ */
+
+static struct inotify_handle *rtree_ih;
+
+static struct audit_tree *alloc_tree(const char *s)
+{
+	struct audit_tree *tree;
+
+	tree = kmalloc(sizeof(struct audit_tree) + strlen(s) + 1, GFP_KERNEL);
+	if (tree) {
+		atomic_set(&tree->count, 1);
+		tree->goner = 0;
+		INIT_LIST_HEAD(&tree->chunks);
+		INIT_LIST_HEAD(&tree->rules);
+		INIT_LIST_HEAD(&tree->list);
+		INIT_LIST_HEAD(&tree->same_root);
+		tree->root = NULL;
+		strcpy(tree->pathname, s);
+	}
+	return tree;
+}
+
+static inline void get_tree(struct audit_tree *tree)
+{
+	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);
+}
+
+/* to avoid bringing the entire thing in audit.h */
+const char *audit_tree_path(struct audit_tree *tree)
+{
+	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;
+	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 rcu_head *rcu)
+{
+	struct audit_chunk *chunk = container_of(rcu, struct audit_chunk, head);
+	int i;
+
+	for (i = 0; i < chunk->count; i++) {
+		if (chunk->owners[i].owner)
+			put_tree(chunk->owners[i].owner);
+	}
+	kfree(chunk);
+}
+
+static inline void free_chunk(struct audit_chunk *chunk)
+{
+	call_rcu(&chunk->head, __free_chunk);
+}
+
+void audit_put_chunk(struct audit_chunk *chunk)
+{
+	put_inotify_watch(&chunk->watch);
+}
+
+enum {HASH_SIZE = 128};
+static struct list_head chunk_hash_heads[HASH_SIZE];
+static __cacheline_aligned_in_smp DEFINE_SPINLOCK(hash_lock);
+
+static inline struct list_head *chunk_hash(const struct inode *inode)
+{
+	unsigned long n = (unsigned long)inode / L1_CACHE_BYTES;
+	return chunk_hash_heads + n % HASH_SIZE;
+}
+
+/* hash_lock is held by caller */
+static void insert_hash(struct audit_chunk *chunk)
+{
+	struct list_head *list = chunk_hash(chunk->watch.inode);
+	list_add_rcu(&chunk->hash, list);
+}
+
+/* called under rcu_read_lock */
+struct audit_chunk *audit_tree_lookup(const struct inode *inode)
+{
+	struct list_head *list = chunk_hash(inode);
+	struct list_head *pos;
+
+	list_for_each_rcu(pos, list) {
+		struct audit_chunk *p = container_of(pos, struct audit_chunk, hash);
+		if (p->watch.inode == inode) {
+			get_inotify_watch(&p->watch);
+			return p;
+		}
+	}
+	return NULL;
+}
+
+int audit_tree_match(struct audit_chunk *chunk, struct audit_tree *tree)
+{
+	int n;
+	for (n = 0; n < chunk->count; n++)
+		if (chunk->owners[n].owner == tree)
+			return 1;
+	return 0;
+}
+
+/* tagging and untagging inodes with trees */
+
+static void untag_chunk(struct audit_chunk *chunk, struct node *p)
+{
+	struct audit_chunk *new;
+	struct audit_tree *owner;
+	int size = chunk->count - 1;
+	int i, j;
+
+	mutex_lock(&chunk->watch.inode->inotify_mutex);
+	if (chunk->dead) {
+		mutex_unlock(&chunk->watch.inode->inotify_mutex);
+		return;
+	}
+
+	owner = p->owner;
+
+	if (!size) {
+		chunk->dead = 1;
+		spin_lock(&hash_lock);
+		list_del_init(&chunk->trees);
+		if (owner->root == chunk)
+			owner->root = NULL;
+		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);
+		return;
+	}
+
+	new = alloc_chunk(size);
+	if (!new)
+		goto Fallback;
+	if (inotify_clone_watch(&chunk->watch, &new->watch) < 0) {
+		free_chunk(new);
+		goto Fallback;
+	}
+
+	chunk->dead = 1;
+	spin_lock(&hash_lock);
+	list_replace_init(&chunk->trees, &new->trees);
+	if (owner->root == chunk) {
+		list_del_init(&owner->same_root);
+		owner->root = NULL;
+	}
+
+	for (i = j = 0; i < size; i++, j++) {
+		struct audit_tree *s;
+		if (&chunk->owners[j] == p) {
+			list_del_init(&p->list);
+			i--;
+			continue;
+		}
+		s = chunk->owners[j].owner;
+		new->owners[i].owner = s;
+		new->owners[i].index = chunk->owners[j].index - j + i;
+		if (!s) /* result of earlier fallback */
+			continue;
+		get_tree(s);
+		list_replace_init(&chunk->owners[i].list, &new->owners[j].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);
+	return;
+
+Fallback:
+	// do the best we can
+	spin_lock(&hash_lock);
+	if (owner->root == chunk) {
+		list_del_init(&owner->same_root);
+		owner->root = NULL;
+	}
+	list_del_init(&p->list);
+	p->owner = NULL;
+	put_tree(owner);
+	spin_unlock(&hash_lock);
+	mutex_unlock(&chunk->watch.inode->inotify_mutex);
+}
+
+static int create_chunk(struct inode *inode, struct audit_tree *tree)
+{
+	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) {
+		free_chunk(chunk);
+		return -ENOSPC;
+	}
+
+	mutex_lock(&inode->inotify_mutex);
+	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);
+		return 0;
+	}
+	chunk->owners[0].index = (1U << 31);
+	chunk->owners[0].owner = tree;
+	get_tree(tree);
+	list_add(&chunk->owners[0].list, &tree->chunks);
+	if (!tree->root) {
+		tree->root = chunk;
+		list_add(&tree->same_root, &chunk->trees);
+	}
+	insert_hash(chunk);
+	spin_unlock(&hash_lock);
+	mutex_unlock(&inode->inotify_mutex);
+	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 audit_tree *owner;
+	struct audit_chunk *chunk, *old;
+	struct node *p;
+	int n;
+
+	if (inotify_find_watch(rtree_ih, inode, &watch) < 0)
+		return create_chunk(inode, tree);
+
+	old = container_of(watch, struct audit_chunk, watch);
+
+	/* 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);
+			return 0;
+		}
+	}
+	spin_unlock(&hash_lock);
+
+	chunk = alloc_chunk(old->count + 1);
+	if (!chunk)
+		return -ENOMEM;
+
+	mutex_lock(&inode->inotify_mutex);
+	if (inotify_clone_watch(&old->watch, &chunk->watch) < 0) {
+		mutex_unlock(&inode->inotify_mutex);
+		free_chunk(chunk);
+		return -ENOSPC;
+	}
+	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);
+		return 0;
+	}
+	list_replace_init(&old->trees, &chunk->trees);
+	for (n = 0, p = chunk->owners; n < old->count; n++, p++) {
+		struct audit_tree *s = old->owners[n].owner;
+		p->owner = s;
+		p->index = old->owners[n].index;
+		if (!s) /* result of fallback in untag */
+			continue;
+		get_tree(s);
+		list_replace_init(&old->owners[n].list, &p->list);
+	}
+	p->index = (chunk->count - 1) | (1U<<31);
+	p->owner = tree;
+	get_tree(tree);
+	list_add(&p->list, &tree->chunks);
+	list_replace_rcu(&old->hash, &chunk->hash);
+	list_for_each_entry(owner, &chunk->trees, same_root)
+		owner->root = chunk;
+	old->dead = 1;
+	if (!tree->root) {
+		tree->root = chunk;
+		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);
+	return 0;
+}
+
+static struct audit_chunk *find_chunk(struct node *p)
+{
+	int index = p->index & ~(1U<<31);
+	p -= index;
+	return container_of(p, struct audit_chunk, owners[0]);
+}
+
+static void kill_rules(struct audit_tree *tree)
+{
+	struct audit_krule *rule, *next;
+	struct audit_entry *entry;
+	struct audit_buffer *ab;
+
+	list_for_each_entry_safe(rule, next, &tree->rules, rlist) {
+		entry = container_of(rule, struct audit_entry, rule);
+
+		list_del_init(&rule->rlist);
+		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_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_format(ab, " list=%d res=1", rule->listnr);
+			audit_log_end(ab);
+			rule->tree = NULL;
+			list_del_rcu(&entry->list);
+			call_rcu(&entry->rcu, audit_free_rule_rcu);
+		}
+	}
+}
+
+/*
+ * finish killing struct audit_tree
+ */
+static void prune_one(struct audit_tree *victim)
+{
+	spin_lock(&hash_lock);
+	while (!list_empty(&victim->chunks)) {
+		struct node *p;
+		struct audit_chunk *chunk;
+
+		p = list_entry(victim->chunks.next, struct node, list);
+		chunk = find_chunk(p);
+		get_inotify_watch(&chunk->watch);
+		spin_unlock(&hash_lock);
+
+		untag_chunk(chunk, p);
+
+		put_inotify_watch(&chunk->watch);
+		spin_lock(&hash_lock);
+	}
+	spin_unlock(&hash_lock);
+	put_tree(victim);
+}
+
+/* trim the uncommitted chunks from tree */
+
+static void trim_marked(struct audit_tree *tree)
+{
+	struct list_head *p, *q;
+	spin_lock(&hash_lock);
+	if (tree->goner) {
+		spin_unlock(&hash_lock);
+		return;
+	}
+	/* reorder */
+	for (p = tree->chunks.next; p != &tree->chunks; p = q) {
+		struct node *node = list_entry(p, struct node, list);
+		q = p->next;
+		if (node->index & (1U<<31)) {
+			list_del_init(p);
+			list_add(p, &tree->chunks);
+		}
+	}
+
+	while (!list_empty(&tree->chunks)) {
+		struct node *node;
+		struct audit_chunk *chunk;
+
+		node = list_entry(tree->chunks.next, struct node, list);
+
+		/* have we run out of marked? */
+		if (!(node->index & (1U<<31)))
+			break;
+
+		chunk = find_chunk(node);
+		get_inotify_watch(&chunk->watch);
+		spin_unlock(&hash_lock);
+
+		untag_chunk(chunk, node);
+
+		put_inotify_watch(&chunk->watch);
+		spin_lock(&hash_lock);
+	}
+	if (!tree->root && !tree->goner) {
+		tree->goner = 1;
+		spin_unlock(&hash_lock);
+		mutex_lock(&audit_filter_mutex);
+		kill_rules(tree);
+		list_del_init(&tree->list);
+		mutex_unlock(&audit_filter_mutex);
+		prune_one(tree);
+	} else {
+		spin_unlock(&hash_lock);
+	}
+}
+
+/* called with audit_filter_mutex */
+int audit_remove_tree_rule(struct audit_krule *rule)
+{
+	struct audit_tree *tree;
+	tree = rule->tree;
+	if (tree) {
+		spin_lock(&hash_lock);
+		list_del_init(&rule->rlist);
+		if (list_empty(&tree->rules) && !tree->goner) {
+			tree->root = NULL;
+			list_del_init(&tree->same_root);
+			tree->goner = 1;
+			list_move(&tree->list, &prune_list);
+			rule->tree = NULL;
+			spin_unlock(&hash_lock);
+			audit_schedule_prune();
+			return 1;
+		}
+		rule->tree = NULL;
+		spin_unlock(&hash_lock);
+		return 1;
+	}
+	return 0;
+}
+
+void audit_trim_trees(void)
+{
+	struct list_head cursor;
+
+	mutex_lock(&audit_filter_mutex);
+	list_add(&cursor, &tree_list);
+	while (cursor.next != &tree_list) {
+		struct audit_tree *tree;
+		struct nameidata nd;
+		struct vfsmount *root_mnt;
+		struct node *node;
+		struct list_head list;
+		int err;
+
+		tree = container_of(cursor.next, struct audit_tree, list);
+		get_tree(tree);
+		list_del(&cursor);
+		list_add(&cursor, &tree->list);
+		mutex_unlock(&audit_filter_mutex);
+
+		err = path_lookup(tree->pathname, 0, &nd);
+		if (err)
+			goto skip_it;
+
+		root_mnt = collect_mounts(nd.mnt, nd.dentry);
+		path_release(&nd);
+		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;
+			node->index |= 1U<<31;
+			list_for_each_entry(mnt, &list, mnt_list) {
+				if (mnt->mnt_root->d_inode == inode) {
+					node->index &= ~(1U<<31);
+					break;
+				}
+			}
+		}
+		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);
+	}
+	list_del(&cursor);
+	mutex_unlock(&audit_filter_mutex);
+}
+
+static int is_under(struct vfsmount *mnt, struct dentry *dentry,
+		    struct nameidata *nd)
+{
+	if (mnt != nd->mnt) {
+		for (;;) {
+			if (mnt->mnt_parent == mnt)
+				return 0;
+			if (mnt->mnt_parent == nd->mnt)
+					break;
+			mnt = mnt->mnt_parent;
+		}
+		dentry = mnt->mnt_mountpoint;
+	}
+	return is_subdir(dentry, nd->dentry);
+}
+
+int audit_make_tree(struct audit_krule *rule, char *pathname, u32 op)
+{
+
+	if (pathname[0] != '/' ||
+	    rule->listnr != AUDIT_FILTER_EXIT ||
+	    op & ~AUDIT_EQUAL ||
+	    rule->inode_f || rule->watch || rule->tree)
+		return -EINVAL;
+	rule->tree = alloc_tree(pathname);
+	if (!rule->tree)
+		return -ENOMEM;
+	return 0;
+}
+
+void audit_put_tree(struct audit_tree *tree)
+{
+	put_tree(tree);
+}
+
+/* called with audit_filter_mutex */
+int audit_add_tree_rule(struct audit_krule *rule)
+{
+	struct audit_tree *seed = rule->tree, *tree;
+	struct nameidata nd;
+	struct vfsmount *mnt, *p;
+	struct list_head list;
+	int err;
+
+	list_for_each_entry(tree, &tree_list, list) {
+		if (!strcmp(seed->pathname, tree->pathname)) {
+			put_tree(seed);
+			rule->tree = tree;
+			list_add(&rule->rlist, &tree->rules);
+			return 0;
+		}
+	}
+	tree = seed;
+	list_add(&tree->list, &tree_list);
+	list_add(&rule->rlist, &tree->rules);
+	/* do not set rule->tree yet */
+	mutex_unlock(&audit_filter_mutex);
+
+	err = path_lookup(tree->pathname, 0, &nd);
+	if (err)
+		goto Err;
+	mnt = collect_mounts(nd.mnt, nd.dentry);
+	path_release(&nd);
+	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);
+	drop_collected_mounts(mnt);
+
+	if (!err) {
+		struct node *node;
+		spin_lock(&hash_lock);
+		list_for_each_entry(node, &tree->chunks, list)
+			node->index &= ~(1U<<31);
+		spin_unlock(&hash_lock);
+	} else {
+		trim_marked(tree);
+		goto Err;
+	}
+
+	mutex_lock(&audit_filter_mutex);
+	if (list_empty(&rule->rlist)) {
+		put_tree(tree);
+		return -ENOENT;
+	}
+	rule->tree = tree;
+	put_tree(tree);
+
+	return 0;
+Err:
+	mutex_lock(&audit_filter_mutex);
+	list_del_init(&tree->list);
+	list_del_init(&tree->rules);
+	put_tree(tree);
+	return err;
+}
+
+int audit_tag_tree(char *old, char *new)
+{
+	struct list_head cursor, barrier;
+	int failed = 0;
+	struct nameidata nd;
+	struct vfsmount *tagged;
+	struct list_head list;
+	struct vfsmount *mnt;
+	struct dentry *dentry;
+	int err;
+
+	err = path_lookup(new, 0, &nd);
+	if (err)
+		return err;
+	tagged = collect_mounts(nd.mnt, nd.dentry);
+	path_release(&nd);
+	if (!tagged)
+		return -ENOMEM;
+
+	err = path_lookup(old, 0, &nd);
+	if (err) {
+		drop_collected_mounts(tagged);
+		return err;
+	}
+	mnt = mntget(nd.mnt);
+	dentry = dget(nd.dentry);
+	path_release(&nd);
+
+	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);
+	list_add(&cursor, &barrier);
+
+	while (cursor.next != &tree_list) {
+		struct audit_tree *tree;
+		struct vfsmount *p;
+
+		tree = container_of(cursor.next, struct audit_tree, list);
+		get_tree(tree);
+		list_del(&cursor);
+		list_add(&cursor, &tree->list);
+		mutex_unlock(&audit_filter_mutex);
+
+		err = path_lookup(tree->pathname, 0, &nd);
+		if (err) {
+			put_tree(tree);
+			mutex_lock(&audit_filter_mutex);
+			continue;
+		}
+
+		spin_lock(&vfsmount_lock);
+		if (!is_under(mnt, dentry, &nd)) {
+			spin_unlock(&vfsmount_lock);
+			path_release(&nd);
+			put_tree(tree);
+			mutex_lock(&audit_filter_mutex);
+			continue;
+		}
+		spin_unlock(&vfsmount_lock);
+		path_release(&nd);
+
+		list_for_each_entry(p, &list, mnt_list) {
+			failed = tag_chunk(p->mnt_root->d_inode, tree);
+			if (failed)
+				break;
+		}
+
+		if (failed) {
+			put_tree(tree);
+			mutex_lock(&audit_filter_mutex);
+			break;
+		}
+
+		mutex_lock(&audit_filter_mutex);
+		spin_lock(&hash_lock);
+		if (!tree->goner) {
+			list_del(&tree->list);
+			list_add(&tree->list, &tree_list);
+		}
+		spin_unlock(&hash_lock);
+		put_tree(tree);
+	}
+
+	while (barrier.prev != &tree_list) {
+		struct audit_tree *tree;
+
+		tree = container_of(barrier.prev, struct audit_tree, list);
+		get_tree(tree);
+		list_del(&tree->list);
+		list_add(&tree->list, &barrier);
+		mutex_unlock(&audit_filter_mutex);
+
+		if (!failed) {
+			struct node *node;
+			spin_lock(&hash_lock);
+			list_for_each_entry(node, &tree->chunks, list)
+				node->index &= ~(1U<<31);
+			spin_unlock(&hash_lock);
+		} else {
+			trim_marked(tree);
+		}
+
+		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);
+	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.
+ */
+void audit_prune_trees(void)
+{
+	mutex_lock(&audit_filter_mutex);
+
+	while (!list_empty(&prune_list)) {
+		struct audit_tree *victim;
+
+		victim = list_entry(prune_list.next, struct audit_tree, list);
+		list_del_init(&victim->list);
+
+		mutex_unlock(&audit_filter_mutex);
+
+		prune_one(victim);
+
+		mutex_lock(&audit_filter_mutex);
+	}
+
+	mutex_unlock(&audit_filter_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;
+	int n;
+
+	if (chunk->dead)
+		return;
+
+	chunk->dead = 1;
+	mutex_lock(&audit_filter_mutex);
+	spin_lock(&hash_lock);
+	while (!list_empty(&chunk->trees)) {
+		owner = list_entry(chunk->trees.next,
+				   struct audit_tree, same_root);
+		owner->goner = 1;
+		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();
+		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);
+	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)
+{
+	struct audit_chunk *chunk = container_of(watch, struct audit_chunk, watch);
+
+	if (mask & IN_IGNORED) {
+		evict_chunk(chunk);
+		put_inotify_watch(watch);
+	}
+}
+
+static void destroy_watch(struct inotify_watch *watch)
+{
+	struct audit_chunk *chunk = container_of(watch, struct audit_chunk, watch);
+	free_chunk(chunk);
+}
+
+static const struct inotify_operations rtree_inotify_ops = {
+	.handle_event	= handle_event,
+	.destroy_watch	= destroy_watch,
+};
+
+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");
+
+	for (i = 0; i < HASH_SIZE; i++)
+		INIT_LIST_HEAD(&chunk_hash_heads[i]);
+
+	return 0;
+}
+__initcall(audit_tree_init);

kernel/auditfilter.c

@@ -87,7 +87,7 @@ struct list_head audit_filter_list[AUDIT_NR_FILTERS] = {
 #endif
 };
 
-static DEFINE_MUTEX(audit_filter_mutex);
+DEFINE_MUTEX(audit_filter_mutex);
 
 /* Inotify handle */
 extern struct inotify_handle *audit_ih;
@@ -145,7 +145,7 @@ static inline void audit_free_rule(struct audit_entry *e)
 	kfree(e);
 }
 
-static inline void audit_free_rule_rcu(struct rcu_head *head)
+void audit_free_rule_rcu(struct rcu_head *head)
 {
 	struct audit_entry *e = container_of(head, struct audit_entry, rcu);
 	audit_free_rule(e);
@@ -217,7 +217,7 @@ static inline struct audit_entry *audit_init_entry(u32 field_count)
 
 /* Unpack a filter field's string representation from user-space
  * buffer. */
-static char *audit_unpack_string(void **bufp, size_t *remain, size_t len)
+char *audit_unpack_string(void **bufp, size_t *remain, size_t len)
 {
 	char *str;
 
@@ -247,7 +247,7 @@ static inline int audit_to_inode(struct audit_krule *krule,
 				 struct audit_field *f)
 {
 	if (krule->listnr != AUDIT_FILTER_EXIT ||
-	    krule->watch || krule->inode_f)
+	    krule->watch || krule->inode_f || krule->tree)
 		return -EINVAL;
 
 	krule->inode_f = f;
@@ -266,7 +266,7 @@ static int audit_to_watch(struct audit_krule *krule, char *path, int len,
 	if (path[0] != '/' || path[len-1] == '/' ||
 	    krule->listnr != AUDIT_FILTER_EXIT ||
 	    op & ~AUDIT_EQUAL ||
-	    krule->inode_f || krule->watch) /* 1 inode # per rule, for hash */
+	    krule->inode_f || krule->watch || krule->tree)
 		return -EINVAL;
 
 	watch = audit_init_watch(path);
@@ -622,6 +622,17 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data,
 				goto exit_free;
 			}
 			break;
+		case AUDIT_DIR:
+			str = audit_unpack_string(&bufp, &remain, f->val);
+			if (IS_ERR(str))
+				goto exit_free;
+			entry->rule.buflen += f->val;
+
+			err = audit_make_tree(&entry->rule, str, f->op);
+			kfree(str);
+			if (err)
+				goto exit_free;
+			break;
 		case AUDIT_INODE:
 			err = audit_to_inode(&entry->rule, f);
 			if (err)
@@ -668,7 +679,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data,
 }
 
 /* Pack a filter field's string representation into data block. */
-static inline size_t audit_pack_string(void **bufp, char *str)
+static inline size_t audit_pack_string(void **bufp, const char *str)
 {
 	size_t len = strlen(str);
 
@@ -747,6 +758,11 @@ static struct audit_rule_data *audit_krule_to_data(struct audit_krule *krule)
 			data->buflen += data->values[i] =
 				audit_pack_string(&bufp, krule->watch->path);
 			break;
+		case AUDIT_DIR:
+			data->buflen += data->values[i] =
+				audit_pack_string(&bufp,
+						  audit_tree_path(krule->tree));
+			break;
 		case AUDIT_FILTERKEY:
 			data->buflen += data->values[i] =
 				audit_pack_string(&bufp, krule->filterkey);
@@ -795,6 +811,11 @@ static int audit_compare_rule(struct audit_krule *a, struct audit_krule *b)
 			if (strcmp(a->watch->path, b->watch->path))
 				return 1;
 			break;
+		case AUDIT_DIR:
+			if (strcmp(audit_tree_path(a->tree),
+				   audit_tree_path(b->tree)))
+				return 1;
+			break;
 		case AUDIT_FILTERKEY:
 			/* both filterkeys exist based on above type compare */
 			if (strcmp(a->filterkey, b->filterkey))
@@ -897,6 +918,14 @@ static struct audit_entry *audit_dupe_rule(struct audit_krule *old,
 	new->inode_f = old->inode_f;
 	new->watch = NULL;
 	new->field_count = old->field_count;
+	/*
+	 * note that we are OK with not refcounting here; audit_match_tree()
+	 * never dereferences tree and we can't get false positives there
+	 * since we'd have to have rule gone from the list *and* removed
+	 * before the chunks found by lookup had been allocated, i.e. before
+	 * the beginning of list scan.
+	 */
+	new->tree = old->tree;
 	memcpy(new->fields, old->fields, sizeof(struct audit_field) * fcount);
 
 	/* deep copy this information, updating the se_rule fields, because
@@ -1217,6 +1246,7 @@ static inline int audit_add_rule(struct audit_entry *entry,
 	struct audit_entry *e;
 	struct audit_field *inode_f = entry->rule.inode_f;
 	struct audit_watch *watch = entry->rule.watch;
+	struct audit_tree *tree = entry->rule.tree;
 	struct nameidata *ndp = NULL, *ndw = NULL;
 	int h, err;
 #ifdef CONFIG_AUDITSYSCALL
@@ -1238,6 +1268,9 @@ static inline int audit_add_rule(struct audit_entry *entry,
 	mutex_unlock(&audit_filter_mutex);
 	if (e) {
 		err = -EEXIST;
+		/* normally audit_add_tree_rule() will free it on failure */
+		if (tree)
+			audit_put_tree(tree);
 		goto error;
 	}
 
@@ -1259,6 +1292,13 @@ static inline int audit_add_rule(struct audit_entry *entry,
 		h = audit_hash_ino((u32)watch->ino);
 		list = &audit_inode_hash[h];
 	}
+	if (tree) {
+		err = audit_add_tree_rule(&entry->rule);
+		if (err) {
+			mutex_unlock(&audit_filter_mutex);
+			goto error;
+		}
+	}
 
 	if (entry->rule.flags & AUDIT_FILTER_PREPEND) {
 		list_add_rcu(&entry->list, list);
@@ -1292,6 +1332,7 @@ static inline int audit_del_rule(struct audit_entry *entry,
 	struct audit_entry  *e;
 	struct audit_field *inode_f = entry->rule.inode_f;
 	struct audit_watch *watch, *tmp_watch = entry->rule.watch;
+	struct audit_tree *tree = entry->rule.tree;
 	LIST_HEAD(inotify_list);
 	int h, ret = 0;
 #ifdef CONFIG_AUDITSYSCALL
@@ -1336,6 +1377,9 @@ static inline int audit_del_rule(struct audit_entry *entry,
 		}
 	}
 
+	if (e->rule.tree)
+		audit_remove_tree_rule(&e->rule);
+
 	list_del_rcu(&e->list);
 	call_rcu(&e->rcu, audit_free_rule_rcu);
 
@@ -1354,6 +1398,8 @@ static inline int audit_del_rule(struct audit_entry *entry,
 out:
 	if (tmp_watch)
 		audit_put_watch(tmp_watch); /* match initial get */
+	if (tree)
+		audit_put_tree(tree);	/* that's the temporary one */
 
 	return ret;
 }
@@ -1737,6 +1783,7 @@ int selinux_audit_rule_update(void)
 {
 	struct audit_entry *entry, *n, *nentry;
 	struct audit_watch *watch;
+	struct audit_tree *tree;
 	int i, err = 0;
 
 	/* audit_filter_mutex synchronizes the writers */
@@ -1748,6 +1795,7 @@ int selinux_audit_rule_update(void)
 				continue;
 
 			watch = entry->rule.watch;
+			tree = entry->rule.tree;
 			nentry = audit_dupe_rule(&entry->rule, watch);
 			if (unlikely(IS_ERR(nentry))) {
 				/* save the first error encountered for the
@@ -1763,7 +1811,9 @@ int selinux_audit_rule_update(void)
 					list_add(&nentry->rule.rlist,
 						 &watch->rules);
 					list_del(&entry->rule.rlist);
-				}
+				} else if (tree)
+					list_replace_init(&entry->rule.rlist,
+						     &nentry->rule.rlist);
 				list_replace_rcu(&entry->list, &nentry->list);
 			}
 			call_rcu(&entry->rcu, audit_free_rule_rcu);

kernel/auditsc.c

@@ -65,6 +65,7 @@
 #include <linux/binfmts.h>
 #include <linux/highmem.h>
 #include <linux/syscalls.h>
+#include <linux/inotify.h>
 
 #include "audit.h"
 
@@ -179,6 +180,11 @@ struct audit_aux_data_pids {
 	int			pid_count;
 };
 
+struct audit_tree_refs {
+	struct audit_tree_refs *next;
+	struct audit_chunk *c[31];
+};
+
 /* The per-task audit context. */
 struct audit_context {
 	int		    dummy;	/* must be the first element */
@@ -211,6 +217,9 @@ struct audit_context {
 	pid_t		    target_pid;
 	u32		    target_sid;
 
+	struct audit_tree_refs *trees, *first_trees;
+	int tree_count;
+
 #if AUDIT_DEBUG
 	int		    put_count;
 	int		    ino_count;
@@ -265,6 +274,117 @@ static int audit_match_perm(struct audit_context *ctx, int mask)
 	}
 }
 
+/*
+ * We keep a linked list of fixed-sized (31 pointer) arrays of audit_chunk *;
+ * ->first_trees points to its beginning, ->trees - to the current end of data.
+ * ->tree_count is the number of free entries in array pointed to by ->trees.
+ * Original condition is (NULL, NULL, 0); as soon as it grows we never revert to NULL,
+ * "empty" becomes (p, p, 31) afterwards.  We don't shrink the list (and seriously,
+ * it's going to remain 1-element for almost any setup) until we free context itself.
+ * References in it _are_ dropped - at the same time we free/drop aux stuff.
+ */
+
+#ifdef CONFIG_AUDIT_TREE
+static int put_tree_ref(struct audit_context *ctx, struct audit_chunk *chunk)
+{
+	struct audit_tree_refs *p = ctx->trees;
+	int left = ctx->tree_count;
+	if (likely(left)) {
+		p->c[--left] = chunk;
+		ctx->tree_count = left;
+		return 1;
+	}
+	if (!p)
+		return 0;
+	p = p->next;
+	if (p) {
+		p->c[30] = chunk;
+		ctx->trees = p;
+		ctx->tree_count = 30;
+		return 1;
+	}
+	return 0;
+}
+
+static int grow_tree_refs(struct audit_context *ctx)
+{
+	struct audit_tree_refs *p = ctx->trees;
+	ctx->trees = kzalloc(sizeof(struct audit_tree_refs), GFP_KERNEL);
+	if (!ctx->trees) {
+		ctx->trees = p;
+		return 0;
+	}
+	if (p)
+		p->next = ctx->trees;
+	else
+		ctx->first_trees = ctx->trees;
+	ctx->tree_count = 31;
+	return 1;
+}
+#endif
+
+static void unroll_tree_refs(struct audit_context *ctx,
+		      struct audit_tree_refs *p, int count)
+{
+#ifdef CONFIG_AUDIT_TREE
+	struct audit_tree_refs *q;
+	int n;
+	if (!p) {
+		/* we started with empty chain */
+		p = ctx->first_trees;
+		count = 31;
+		/* if the very first allocation has failed, nothing to do */
+		if (!p)
+			return;
+	}
+	n = count;
+	for (q = p; q != ctx->trees; q = q->next, n = 31) {
+		while (n--) {
+			audit_put_chunk(q->c[n]);
+			q->c[n] = NULL;
+		}
+	}
+	while (n-- > ctx->tree_count) {
+		audit_put_chunk(q->c[n]);
+		q->c[n] = NULL;
+	}
+	ctx->trees = p;
+	ctx->tree_count = count;
+#endif
+}
+
+static void free_tree_refs(struct audit_context *ctx)
+{
+	struct audit_tree_refs *p, *q;
+	for (p = ctx->first_trees; p; p = q) {
+		q = p->next;
+		kfree(p);
+	}
+}
+
+static int match_tree_refs(struct audit_context *ctx, struct audit_tree *tree)
+{
+#ifdef CONFIG_AUDIT_TREE
+	struct audit_tree_refs *p;
+	int n;
+	if (!tree)
+		return 0;
+	/* full ones */
+	for (p = ctx->first_trees; p != ctx->trees; p = p->next) {
+		for (n = 0; n < 31; n++)
+			if (audit_tree_match(p->c[n], tree))
+				return 1;
+	}
+	/* partial */
+	if (p) {
+		for (n = ctx->tree_count; n < 31; n++)
+			if (audit_tree_match(p->c[n], tree))
+				return 1;
+	}
+#endif
+	return 0;
+}
+
 /* Determine if any context name data matches a rule's watch data */
 /* Compare a task_struct with an audit_rule.  Return 1 on match, 0
  * otherwise. */
@@ -379,6 +499,10 @@ static int audit_filter_rules(struct task_struct *tsk,
 				result = (name->dev == rule->watch->dev &&
 					  name->ino == rule->watch->ino);
 			break;
+		case AUDIT_DIR:
+			if (ctx)
+				result = match_tree_refs(ctx, rule->tree);
+			break;
 		case AUDIT_LOGINUID:
 			result = 0;
 			if (ctx)
@@ -727,6 +851,8 @@ static inline void audit_free_context(struct audit_context *context)
 			       context->name_count, count);
 		}
 		audit_free_names(context);
+		unroll_tree_refs(context, NULL, 0);
+		free_tree_refs(context);
 		audit_free_aux(context);
 		kfree(context->filterkey);
 		kfree(context);
@@ -1270,6 +1396,7 @@ void audit_syscall_exit(int valid, long return_code)
 		tsk->audit_context = new_context;
 	} else {
 		audit_free_names(context);
+		unroll_tree_refs(context, NULL, 0);
 		audit_free_aux(context);
 		context->aux = NULL;
 		context->aux_pids = NULL;
@@ -1281,6 +1408,95 @@ void audit_syscall_exit(int valid, long return_code)
 	}
 }
 
+static inline void handle_one(const struct inode *inode)
+{
+#ifdef CONFIG_AUDIT_TREE
+	struct audit_context *context;
+	struct audit_tree_refs *p;
+	struct audit_chunk *chunk;
+	int count;
+	if (likely(list_empty(&inode->inotify_watches)))
+		return;
+	context = current->audit_context;
+	p = context->trees;
+	count = context->tree_count;
+	rcu_read_lock();
+	chunk = audit_tree_lookup(inode);
+	rcu_read_unlock();
+	if (!chunk)
+		return;
+	if (likely(put_tree_ref(context, chunk)))
+		return;
+	if (unlikely(!grow_tree_refs(context))) {
+		printk(KERN_WARNING "out of memory, audit has lost a tree reference");
+		audit_set_auditable(context);
+		audit_put_chunk(chunk);
+		unroll_tree_refs(context, p, count);
+		return;
+	}
+	put_tree_ref(context, chunk);
+#endif
+}
+
+static void handle_path(const struct dentry *dentry)
+{
+#ifdef CONFIG_AUDIT_TREE
+	struct audit_context *context;
+	struct audit_tree_refs *p;
+	const struct dentry *d, *parent;
+	struct audit_chunk *drop;
+	unsigned long seq;
+	int count;
+
+	context = current->audit_context;
+	p = context->trees;
+	count = context->tree_count;
+retry:
+	drop = NULL;
+	d = dentry;
+	rcu_read_lock();
+	seq = read_seqbegin(&rename_lock);
+	for(;;) {
+		struct inode *inode = d->d_inode;
+		if (inode && unlikely(!list_empty(&inode->inotify_watches))) {
+			struct audit_chunk *chunk;
+			chunk = audit_tree_lookup(inode);
+			if (chunk) {
+				if (unlikely(!put_tree_ref(context, chunk))) {
+					drop = chunk;
+					break;
+				}
+			}
+		}
+		parent = d->d_parent;
+		if (parent == d)
+			break;
+		d = parent;
+	}
+	if (unlikely(read_seqretry(&rename_lock, seq) || drop)) {  /* in this order */
+		rcu_read_unlock();
+		if (!drop) {
+			/* just a race with rename */
+			unroll_tree_refs(context, p, count);
+			goto retry;
+		}
+		audit_put_chunk(drop);
+		if (grow_tree_refs(context)) {
+			/* OK, got more space */
+			unroll_tree_refs(context, p, count);
+			goto retry;
+		}
+		/* too bad */
+		printk(KERN_WARNING
+			"out of memory, audit has lost a tree reference");
+		unroll_tree_refs(context, p, count);
+		audit_set_auditable(context);
+		return;
+	}
+	rcu_read_unlock();
+#endif
+}
+
 /**
  * audit_getname - add a name to the list
  * @name: name to add
@@ -1403,10 +1619,11 @@ static void audit_copy_inode(struct audit_names *name, const struct inode *inode
  *
  * Called from fs/namei.c:path_lookup().
  */
-void __audit_inode(const char *name, const struct inode *inode)
+void __audit_inode(const char *name, const struct dentry *dentry)
 {
 	int idx;
 	struct audit_context *context = current->audit_context;
+	const struct inode *inode = dentry->d_inode;
 
 	if (!context->in_syscall)
 		return;
@@ -1426,6 +1643,7 @@ void __audit_inode(const char *name, const struct inode *inode)
 		idx = context->name_count - 1;
 		context->names[idx].name = NULL;
 	}
+	handle_path(dentry);
 	audit_copy_inode(&context->names[idx], inode);
 }
 
@@ -1443,17 +1661,20 @@ void __audit_inode(const char *name, const struct inode *inode)
  * must be hooked prior, in order to capture the target inode during
  * unsuccessful attempts.
  */
-void __audit_inode_child(const char *dname, const struct inode *inode,
+void __audit_inode_child(const char *dname, const struct dentry *dentry,
 			 const struct inode *parent)
 {
 	int idx;
 	struct audit_context *context = current->audit_context;
 	const char *found_parent = NULL, *found_child = NULL;
+	const struct inode *inode = dentry->d_inode;
 	int dirlen = 0;
 
 	if (!context->in_syscall)
 		return;
 
+	if (inode)
+		handle_one(inode);
 	/* determine matching parent */
 	if (!dname)
 		goto add_names;