Initial commit; kernel source import

fs/notify/Kconfig

@@ -0,0 +1,6 @@
+config FSNOTIFY
+	def_bool n
+
+source "fs/notify/dnotify/Kconfig"
+source "fs/notify/inotify/Kconfig"
+source "fs/notify/fanotify/Kconfig"

fs/notify/Makefile

@@ -0,0 +1,6 @@
+obj-$(CONFIG_FSNOTIFY)		+= fsnotify.o notification.o group.o inode_mark.o \
+				   mark.o vfsmount_mark.o fdinfo.o
+
+obj-y			+= dnotify/
+obj-y			+= inotify/
+obj-y			+= fanotify/

fs/notify/dnotify/Kconfig

@@ -0,0 +1,11 @@
+config DNOTIFY
+	bool "Dnotify support"
+	select FSNOTIFY
+	default y
+	help
+	  Dnotify is a directory-based per-fd file change notification system
+	  that uses signals to communicate events to user-space.  There exist
+	  superior alternatives, but some applications may still rely on
+	  dnotify.
+
+	  If unsure, say Y.

fs/notify/dnotify/Makefile

@@ -0,0 +1 @@
+obj-$(CONFIG_DNOTIFY)		+= dnotify.o

fs/notify/dnotify/dnotify.c

@@ -0,0 +1,411 @@
+/*
+ * Directory notifications for Linux.
+ *
+ * Copyright (C) 2000,2001,2002 Stephen Rothwell
+ *
+ * Copyright (C) 2009 Eric Paris <Red Hat Inc>
+ * dnotify was largly rewritten to use the new fsnotify infrastructure
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+#include <linux/fs.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/dnotify.h>
+#include <linux/init.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/fdtable.h>
+#include <linux/fsnotify_backend.h>
+
+int dir_notify_enable __read_mostly = 1;
+
+static struct kmem_cache *dnotify_struct_cache __read_mostly;
+static struct kmem_cache *dnotify_mark_cache __read_mostly;
+static struct fsnotify_group *dnotify_group __read_mostly;
+static DEFINE_MUTEX(dnotify_mark_mutex);
+
+/*
+ * dnotify will attach one of these to each inode (i_fsnotify_marks) which
+ * is being watched by dnotify.  If multiple userspace applications are watching
+ * the same directory with dnotify their information is chained in dn
+ */
+struct dnotify_mark {
+	struct fsnotify_mark fsn_mark;
+	struct dnotify_struct *dn;
+};
+
+/*
+ * When a process starts or stops watching an inode the set of events which
+ * dnotify cares about for that inode may change.  This function runs the
+ * list of everything receiving dnotify events about this directory and calculates
+ * the set of all those events.  After it updates what dnotify is interested in
+ * it calls the fsnotify function so it can update the set of all events relevant
+ * to this inode.
+ */
+static void dnotify_recalc_inode_mask(struct fsnotify_mark *fsn_mark)
+{
+	__u32 new_mask, old_mask;
+	struct dnotify_struct *dn;
+	struct dnotify_mark *dn_mark  = container_of(fsn_mark,
+						     struct dnotify_mark,
+						     fsn_mark);
+
+	assert_spin_locked(&fsn_mark->lock);
+
+	old_mask = fsn_mark->mask;
+	new_mask = 0;
+	for (dn = dn_mark->dn; dn != NULL; dn = dn->dn_next)
+		new_mask |= (dn->dn_mask & ~FS_DN_MULTISHOT);
+	fsnotify_set_mark_mask_locked(fsn_mark, new_mask);
+
+	if (old_mask == new_mask)
+		return;
+
+	if (fsn_mark->i.inode)
+		fsnotify_recalc_inode_mask(fsn_mark->i.inode);
+}
+
+/*
+ * Mains fsnotify call where events are delivered to dnotify.
+ * Find the dnotify mark on the relevant inode, run the list of dnotify structs
+ * on that mark and determine which of them has expressed interest in receiving
+ * events of this type.  When found send the correct process and signal and
+ * destroy the dnotify struct if it was not registered to receive multiple
+ * events.
+ */
+static int dnotify_handle_event(struct fsnotify_group *group,
+				struct fsnotify_mark *inode_mark,
+				struct fsnotify_mark *vfsmount_mark,
+				struct fsnotify_event *event)
+{
+	struct dnotify_mark *dn_mark;
+	struct inode *to_tell;
+	struct dnotify_struct *dn;
+	struct dnotify_struct **prev;
+	struct fown_struct *fown;
+	__u32 test_mask = event->mask & ~FS_EVENT_ON_CHILD;
+
+	BUG_ON(vfsmount_mark);
+
+	to_tell = event->to_tell;
+
+	dn_mark = container_of(inode_mark, struct dnotify_mark, fsn_mark);
+
+	spin_lock(&inode_mark->lock);
+	prev = &dn_mark->dn;
+	while ((dn = *prev) != NULL) {
+		if ((dn->dn_mask & test_mask) == 0) {
+			prev = &dn->dn_next;
+			continue;
+		}
+		fown = &dn->dn_filp->f_owner;
+		send_sigio(fown, dn->dn_fd, POLL_MSG);
+		if (dn->dn_mask & FS_DN_MULTISHOT)
+			prev = &dn->dn_next;
+		else {
+			*prev = dn->dn_next;
+			kmem_cache_free(dnotify_struct_cache, dn);
+			dnotify_recalc_inode_mask(inode_mark);
+		}
+	}
+
+	spin_unlock(&inode_mark->lock);
+
+	return 0;
+}
+
+/*
+ * Given an inode and mask determine if dnotify would be interested in sending
+ * userspace notification for that pair.
+ */
+static bool dnotify_should_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)
+{
+	/* not a dir, dnotify doesn't care */
+	if (!S_ISDIR(inode->i_mode))
+		return false;
+
+	return true;
+}
+
+static void dnotify_free_mark(struct fsnotify_mark *fsn_mark)
+{
+	struct dnotify_mark *dn_mark = container_of(fsn_mark,
+						    struct dnotify_mark,
+						    fsn_mark);
+
+	BUG_ON(dn_mark->dn);
+
+	kmem_cache_free(dnotify_mark_cache, dn_mark);
+}
+
+static struct fsnotify_ops dnotify_fsnotify_ops = {
+	.handle_event = dnotify_handle_event,
+	.should_send_event = dnotify_should_send_event,
+	.free_group_priv = NULL,
+	.freeing_mark = NULL,
+	.free_event_priv = NULL,
+};
+
+/*
+ * Called every time a file is closed.  Looks first for a dnotify mark on the
+ * inode.  If one is found run all of the ->dn structures attached to that
+ * mark for one relevant to this process closing the file and remove that
+ * dnotify_struct.  If that was the last dnotify_struct also remove the
+ * fsnotify_mark.
+ */
+void dnotify_flush(struct file *filp, fl_owner_t id)
+{
+	struct fsnotify_mark *fsn_mark;
+	struct dnotify_mark *dn_mark;
+	struct dnotify_struct *dn;
+	struct dnotify_struct **prev;
+	struct inode *inode;
+
+	inode = file_inode(filp);
+	if (!S_ISDIR(inode->i_mode))
+		return;
+
+	fsn_mark = fsnotify_find_inode_mark(dnotify_group, inode);
+	if (!fsn_mark)
+		return;
+	dn_mark = container_of(fsn_mark, struct dnotify_mark, fsn_mark);
+
+	mutex_lock(&dnotify_mark_mutex);
+
+	spin_lock(&fsn_mark->lock);
+	prev = &dn_mark->dn;
+	while ((dn = *prev) != NULL) {
+		if ((dn->dn_owner == id) && (dn->dn_filp == filp)) {
+			*prev = dn->dn_next;
+			kmem_cache_free(dnotify_struct_cache, dn);
+			dnotify_recalc_inode_mask(fsn_mark);
+			break;
+		}
+		prev = &dn->dn_next;
+	}
+
+	spin_unlock(&fsn_mark->lock);
+
+	/* nothing else could have found us thanks to the dnotify_mark_mutex */
+	if (dn_mark->dn == NULL)
+		fsnotify_destroy_mark(fsn_mark, dnotify_group);
+
+	mutex_unlock(&dnotify_mark_mutex);
+
+	fsnotify_put_mark(fsn_mark);
+}
+
+/* this conversion is done only at watch creation */
+static __u32 convert_arg(unsigned long arg)
+{
+	__u32 new_mask = FS_EVENT_ON_CHILD;
+
+	if (arg & DN_MULTISHOT)
+		new_mask |= FS_DN_MULTISHOT;
+	if (arg & DN_DELETE)
+		new_mask |= (FS_DELETE | FS_MOVED_FROM);
+	if (arg & DN_MODIFY)
+		new_mask |= FS_MODIFY;
+	if (arg & DN_ACCESS)
+		new_mask |= FS_ACCESS;
+	if (arg & DN_ATTRIB)
+		new_mask |= FS_ATTRIB;
+	if (arg & DN_RENAME)
+		new_mask |= FS_DN_RENAME;
+	if (arg & DN_CREATE)
+		new_mask |= (FS_CREATE | FS_MOVED_TO);
+
+	return new_mask;
+}
+
+/*
+ * If multiple processes watch the same inode with dnotify there is only one
+ * dnotify mark in inode->i_fsnotify_marks but we chain a dnotify_struct
+ * onto that mark.  This function either attaches the new dnotify_struct onto
+ * that list, or it |= the mask onto an existing dnofiy_struct.
+ */
+static int attach_dn(struct dnotify_struct *dn, struct dnotify_mark *dn_mark,
+		     fl_owner_t id, int fd, struct file *filp, __u32 mask)
+{
+	struct dnotify_struct *odn;
+
+	odn = dn_mark->dn;
+	while (odn != NULL) {
+		/* adding more events to existing dnofiy_struct? */
+		if ((odn->dn_owner == id) && (odn->dn_filp == filp)) {
+			odn->dn_fd = fd;
+			odn->dn_mask |= mask;
+			return -EEXIST;
+		}
+		odn = odn->dn_next;
+	}
+
+	dn->dn_mask = mask;
+	dn->dn_fd = fd;
+	dn->dn_filp = filp;
+	dn->dn_owner = id;
+	dn->dn_next = dn_mark->dn;
+	dn_mark->dn = dn;
+
+	return 0;
+}
+
+/*
+ * When a process calls fcntl to attach a dnotify watch to a directory it ends
+ * up here.  Allocate both a mark for fsnotify to add and a dnotify_struct to be
+ * attached to the fsnotify_mark.
+ */
+int fcntl_dirnotify(int fd, struct file *filp, unsigned long arg)
+{
+	struct dnotify_mark *new_dn_mark, *dn_mark;
+	struct fsnotify_mark *new_fsn_mark, *fsn_mark;
+	struct dnotify_struct *dn;
+	struct inode *inode;
+	fl_owner_t id = current->files;
+	struct file *f;
+	int destroy = 0, error = 0;
+	__u32 mask;
+
+	/* we use these to tell if we need to kfree */
+	new_fsn_mark = NULL;
+	dn = NULL;
+
+	if (!dir_notify_enable) {
+		error = -EINVAL;
+		goto out_err;
+	}
+
+	/* a 0 mask means we are explicitly removing the watch */
+	if ((arg & ~DN_MULTISHOT) == 0) {
+		dnotify_flush(filp, id);
+		error = 0;
+		goto out_err;
+	}
+
+	/* dnotify only works on directories */
+	inode = file_inode(filp);
+	if (!S_ISDIR(inode->i_mode)) {
+		error = -ENOTDIR;
+		goto out_err;
+	}
+
+	/* expect most fcntl to add new rather than augment old */
+	dn = kmem_cache_alloc(dnotify_struct_cache, GFP_KERNEL);
+	if (!dn) {
+		error = -ENOMEM;
+		goto out_err;
+	}
+
+	/* new fsnotify mark, we expect most fcntl calls to add a new mark */
+	new_dn_mark = kmem_cache_alloc(dnotify_mark_cache, GFP_KERNEL);
+	if (!new_dn_mark) {
+		error = -ENOMEM;
+		goto out_err;
+	}
+
+	/* convert the userspace DN_* "arg" to the internal FS_* defines in fsnotify */
+	mask = convert_arg(arg);
+
+	/* set up the new_fsn_mark and new_dn_mark */
+	new_fsn_mark = &new_dn_mark->fsn_mark;
+	fsnotify_init_mark(new_fsn_mark, dnotify_free_mark);
+	new_fsn_mark->mask = mask;
+	new_dn_mark->dn = NULL;
+
+	/* this is needed to prevent the fcntl/close race described below */
+	mutex_lock(&dnotify_mark_mutex);
+
+	/* add the new_fsn_mark or find an old one. */
+	fsn_mark = fsnotify_find_inode_mark(dnotify_group, inode);
+	if (fsn_mark) {
+		dn_mark = container_of(fsn_mark, struct dnotify_mark, fsn_mark);
+		spin_lock(&fsn_mark->lock);
+	} else {
+		fsnotify_add_mark(new_fsn_mark, dnotify_group, inode, NULL, 0);
+		spin_lock(&new_fsn_mark->lock);
+		fsn_mark = new_fsn_mark;
+		dn_mark = new_dn_mark;
+		/* we used new_fsn_mark, so don't free it */
+		new_fsn_mark = NULL;
+	}
+
+	rcu_read_lock();
+	f = fcheck(fd);
+	rcu_read_unlock();
+
+	/* if (f != filp) means that we lost a race and another task/thread
+	 * actually closed the fd we are still playing with before we grabbed
+	 * the dnotify_mark_mutex and fsn_mark->lock.  Since closing the fd is the
+	 * only time we clean up the marks we need to get our mark off
+	 * the list. */
+	if (f != filp) {
+		/* if we added ourselves, shoot ourselves, it's possible that
+		 * the flush actually did shoot this fsn_mark.  That's fine too
+		 * since multiple calls to destroy_mark is perfectly safe, if
+		 * we found a dn_mark already attached to the inode, just sod
+		 * off silently as the flush at close time dealt with it.
+		 */
+		if (dn_mark == new_dn_mark)
+			destroy = 1;
+		goto out;
+	}
+
+	error = __f_setown(filp, task_pid(current), PIDTYPE_PID, 0);
+	if (error) {
+		/* if we added, we must shoot */
+		if (dn_mark == new_dn_mark)
+			destroy = 1;
+		goto out;
+	}
+
+	error = attach_dn(dn, dn_mark, id, fd, filp, mask);
+	/* !error means that we attached the dn to the dn_mark, so don't free it */
+	if (!error)
+		dn = NULL;
+	/* -EEXIST means that we didn't add this new dn and used an old one.
+	 * that isn't an error (and the unused dn should be freed) */
+	else if (error == -EEXIST)
+		error = 0;
+
+	dnotify_recalc_inode_mask(fsn_mark);
+out:
+	spin_unlock(&fsn_mark->lock);
+
+	if (destroy)
+		fsnotify_destroy_mark(fsn_mark, dnotify_group);
+
+	mutex_unlock(&dnotify_mark_mutex);
+	fsnotify_put_mark(fsn_mark);
+out_err:
+	if (new_fsn_mark)
+		fsnotify_put_mark(new_fsn_mark);
+	if (dn)
+		kmem_cache_free(dnotify_struct_cache, dn);
+	return error;
+}
+
+static int __init dnotify_init(void)
+{
+	dnotify_struct_cache = KMEM_CACHE(dnotify_struct, SLAB_PANIC);
+	dnotify_mark_cache = KMEM_CACHE(dnotify_mark, SLAB_PANIC);
+
+	dnotify_group = fsnotify_alloc_group(&dnotify_fsnotify_ops);
+	if (IS_ERR(dnotify_group))
+		panic("unable to allocate fsnotify group for dnotify\n");
+	return 0;
+}
+
+module_init(dnotify_init)

fs/notify/fanotify/Kconfig

@@ -0,0 +1,26 @@
+config FANOTIFY
+	bool "Filesystem wide access notification"
+	select FSNOTIFY
+	select ANON_INODES
+	default n
+	---help---
+	   Say Y here to enable fanotify support.  fanotify is a file access
+	   notification system which differs from inotify in that it sends
+	   an open file descriptor to the userspace listener along with
+	   the event.
+
+	   If unsure, say Y.
+
+config FANOTIFY_ACCESS_PERMISSIONS
+	bool "fanotify permissions checking"
+	depends on FANOTIFY
+	depends on SECURITY
+	default n
+	---help---
+	   Say Y here is you want fanotify listeners to be able to make permissions
+	   decisions concerning filesystem events.  This is used by some fanotify
+	   listeners which need to scan files before allowing the system access to
+	   use those files.  This is used by some anti-malware vendors and by some
+	   hierarchical storage managent systems.
+
+	   If unsure, say N.

fs/notify/fanotify/Makefile

@@ -0,0 +1 @@
+obj-$(CONFIG_FANOTIFY)		+= fanotify.o fanotify_user.o

fs/notify/fanotify/fanotify.c

@@ -0,0 +1,235 @@
+#include <linux/fanotify.h>
+#include <linux/fdtable.h>
+#include <linux/fsnotify_backend.h>
+#include <linux/init.h>
+#include <linux/jiffies.h>
+#include <linux/kernel.h> /* UINT_MAX */
+#include <linux/mount.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+
+static bool should_merge(struct fsnotify_event *old, struct fsnotify_event *new)
+{
+	pr_debug("%s: old=%p new=%p\n", __func__, old, new);
+
+	if (old->to_tell == new->to_tell &&
+	    old->data_type == new->data_type &&
+	    old->tgid == new->tgid) {
+		switch (old->data_type) {
+		case (FSNOTIFY_EVENT_PATH):
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+			/* dont merge two permission events */
+			if ((old->mask & FAN_ALL_PERM_EVENTS) &&
+			    (new->mask & FAN_ALL_PERM_EVENTS))
+				return false;
+#endif
+			if ((old->path.mnt == new->path.mnt) &&
+			    (old->path.dentry == new->path.dentry))
+				return true;
+			break;
+		case (FSNOTIFY_EVENT_NONE):
+			return true;
+		default:
+			BUG();
+		};
+	}
+	return false;
+}
+
+/* and the list better be locked by something too! */
+static struct fsnotify_event *fanotify_merge(struct list_head *list,
+					     struct fsnotify_event *event)
+{
+	struct fsnotify_event_holder *test_holder;
+	struct fsnotify_event *test_event = NULL;
+	struct fsnotify_event *new_event;
+
+	pr_debug("%s: list=%p event=%p\n", __func__, list, event);
+
+
+	list_for_each_entry_reverse(test_holder, list, event_list) {
+		if (should_merge(test_holder->event, event)) {
+			test_event = test_holder->event;
+			break;
+		}
+	}
+
+	if (!test_event)
+		return NULL;
+
+	fsnotify_get_event(test_event);
+
+	/* if they are exactly the same we are done */
+	if (test_event->mask == event->mask)
+		return test_event;
+
+	/*
+	 * if the refcnt == 2 this is the only queue
+	 * for this event and so we can update the mask
+	 * in place.
+	 */
+	if (atomic_read(&test_event->refcnt) == 2) {
+		test_event->mask |= event->mask;
+		return test_event;
+	}
+
+	new_event = fsnotify_clone_event(test_event);
+
+	/* done with test_event */
+	fsnotify_put_event(test_event);
+
+	/* couldn't allocate memory, merge was not possible */
+	if (unlikely(!new_event))
+		return ERR_PTR(-ENOMEM);
+
+	/* build new event and replace it on the list */
+	new_event->mask = (test_event->mask | event->mask);
+	fsnotify_replace_event(test_holder, new_event);
+
+	/* we hold a reference on new_event from clone_event */
+	return new_event;
+}
+
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+static int fanotify_get_response_from_access(struct fsnotify_group *group,
+					     struct fsnotify_event *event)
+{
+	int ret;
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	wait_event(group->fanotify_data.access_waitq, event->response ||
+				atomic_read(&group->fanotify_data.bypass_perm));
+
+	if (!event->response) /* bypass_perm set */
+		return 0;
+
+	/* userspace responded, convert to something usable */
+	spin_lock(&event->lock);
+	switch (event->response) {
+	case FAN_ALLOW:
+		ret = 0;
+		break;
+	case FAN_DENY:
+	default:
+		ret = -EPERM;
+	}
+	event->response = 0;
+	spin_unlock(&event->lock);
+
+	pr_debug("%s: group=%p event=%p about to return ret=%d\n", __func__,
+		 group, event, ret);
+	
+	return ret;
+}
+#endif
+
+static int fanotify_handle_event(struct fsnotify_group *group,
+				 struct fsnotify_mark *inode_mark,
+				 struct fsnotify_mark *fanotify_mark,
+				 struct fsnotify_event *event)
+{
+	int ret = 0;
+	struct fsnotify_event *notify_event = NULL;
+
+	BUILD_BUG_ON(FAN_ACCESS != FS_ACCESS);
+	BUILD_BUG_ON(FAN_MODIFY != FS_MODIFY);
+	BUILD_BUG_ON(FAN_CLOSE_NOWRITE != FS_CLOSE_NOWRITE);
+	BUILD_BUG_ON(FAN_CLOSE_WRITE != FS_CLOSE_WRITE);
+	BUILD_BUG_ON(FAN_OPEN != FS_OPEN);
+	BUILD_BUG_ON(FAN_EVENT_ON_CHILD != FS_EVENT_ON_CHILD);
+	BUILD_BUG_ON(FAN_Q_OVERFLOW != FS_Q_OVERFLOW);
+	BUILD_BUG_ON(FAN_OPEN_PERM != FS_OPEN_PERM);
+	BUILD_BUG_ON(FAN_ACCESS_PERM != FS_ACCESS_PERM);
+	BUILD_BUG_ON(FAN_ONDIR != FS_ISDIR);
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	notify_event = fsnotify_add_notify_event(group, event, NULL, fanotify_merge);
+	if (IS_ERR(notify_event))
+		return PTR_ERR(notify_event);
+
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	if (event->mask & FAN_ALL_PERM_EVENTS) {
+		/* if we merged we need to wait on the new event */
+		if (notify_event)
+			event = notify_event;
+		ret = fanotify_get_response_from_access(group, event);
+	}
+#endif
+
+	if (notify_event)
+		fsnotify_put_event(notify_event);
+
+	return ret;
+}
+
+static bool fanotify_should_send_event(struct fsnotify_group *group,
+				       struct inode *to_tell,
+				       struct fsnotify_mark *inode_mark,
+				       struct fsnotify_mark *vfsmnt_mark,
+				       __u32 event_mask, void *data, int data_type)
+{
+	__u32 marks_mask, marks_ignored_mask;
+	struct path *path = data;
+
+	pr_debug("%s: group=%p to_tell=%p inode_mark=%p vfsmnt_mark=%p "
+		 "mask=%x data=%p data_type=%d\n", __func__, group, to_tell,
+		 inode_mark, vfsmnt_mark, event_mask, data, data_type);
+
+	/* if we don't have enough info to send an event to userspace say no */
+	if (data_type != FSNOTIFY_EVENT_PATH)
+		return false;
+
+	/* sorry, fanotify only gives a damn about files and dirs */
+	if (!S_ISREG(path->dentry->d_inode->i_mode) &&
+	    !S_ISDIR(path->dentry->d_inode->i_mode))
+		return false;
+
+	if (inode_mark && vfsmnt_mark) {
+		marks_mask = (vfsmnt_mark->mask | inode_mark->mask);
+		marks_ignored_mask = (vfsmnt_mark->ignored_mask | inode_mark->ignored_mask);
+	} else if (inode_mark) {
+		/*
+		 * if the event is for a child and this inode doesn't care about
+		 * events on the child, don't send it!
+		 */
+		if ((event_mask & FS_EVENT_ON_CHILD) &&
+		    !(inode_mark->mask & FS_EVENT_ON_CHILD))
+			return false;
+		marks_mask = inode_mark->mask;
+		marks_ignored_mask = inode_mark->ignored_mask;
+	} else if (vfsmnt_mark) {
+		marks_mask = vfsmnt_mark->mask;
+		marks_ignored_mask = vfsmnt_mark->ignored_mask;
+	} else {
+		BUG();
+	}
+
+	if (S_ISDIR(path->dentry->d_inode->i_mode) &&
+	    (marks_ignored_mask & FS_ISDIR))
+		return false;
+
+	if (event_mask & marks_mask & ~marks_ignored_mask)
+		return true;
+
+	return false;
+}
+
+static void fanotify_free_group_priv(struct fsnotify_group *group)
+{
+	struct user_struct *user;
+
+	user = group->fanotify_data.user;
+	atomic_dec(&user->fanotify_listeners);
+	free_uid(user);
+}
+
+const struct fsnotify_ops fanotify_fsnotify_ops = {
+	.handle_event = fanotify_handle_event,
+	.should_send_event = fanotify_should_send_event,
+	.free_group_priv = fanotify_free_group_priv,
+	.free_event_priv = NULL,
+	.freeing_mark = NULL,
+};

fs/notify/fanotify/fanotify_user.c

@@ -0,0 +1,891 @@
+#include <linux/fanotify.h>
+#include <linux/fcntl.h>
+#include <linux/file.h>
+#include <linux/fs.h>
+#include <linux/anon_inodes.h>
+#include <linux/fsnotify_backend.h>
+#include <linux/init.h>
+#include <linux/mount.h>
+#include <linux/namei.h>
+#include <linux/poll.h>
+#include <linux/security.h>
+#include <linux/syscalls.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/uaccess.h>
+#include <linux/compat.h>
+
+#include <asm/ioctls.h>
+
+#include "../../mount.h"
+#include "../fdinfo.h"
+
+#define FANOTIFY_DEFAULT_MAX_EVENTS	16384
+#define FANOTIFY_DEFAULT_MAX_MARKS	8192
+#define FANOTIFY_DEFAULT_MAX_LISTENERS	128
+
+extern const struct fsnotify_ops fanotify_fsnotify_ops;
+
+static struct kmem_cache *fanotify_mark_cache __read_mostly;
+static struct kmem_cache *fanotify_response_event_cache __read_mostly;
+
+struct fanotify_response_event {
+	struct list_head list;
+	__s32 fd;
+	struct fsnotify_event *event;
+};
+
+/*
+ * Get an fsnotify notification event if one exists and is small
+ * enough to fit in "count". Return an error pointer if the count
+ * is not large enough.
+ *
+ * Called with the group->notification_mutex held.
+ */
+static struct fsnotify_event *get_one_event(struct fsnotify_group *group,
+					    size_t count)
+{
+	BUG_ON(!mutex_is_locked(&group->notification_mutex));
+
+	pr_debug("%s: group=%p count=%zd\n", __func__, group, count);
+
+	if (fsnotify_notify_queue_is_empty(group))
+		return NULL;
+
+	if (FAN_EVENT_METADATA_LEN > count)
+		return ERR_PTR(-EINVAL);
+
+	/* held the notification_mutex the whole time, so this is the
+	 * same event we peeked above */
+	return fsnotify_remove_notify_event(group);
+}
+
+static int create_fd(struct fsnotify_group *group,
+			struct fsnotify_event *event,
+			struct file **file)
+{
+	int client_fd;
+	struct file *new_file;
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	client_fd = get_unused_fd();
+	if (client_fd < 0)
+		return client_fd;
+
+	if (event->data_type != FSNOTIFY_EVENT_PATH) {
+		WARN_ON(1);
+		put_unused_fd(client_fd);
+		return -EINVAL;
+	}
+
+	/*
+	 * we need a new file handle for the userspace program so it can read even if it was
+	 * originally opened O_WRONLY.
+	 */
+	/* it's possible this event was an overflow event.  in that case dentry and mnt
+	 * are NULL;  That's fine, just don't call dentry open */
+	if (event->path.dentry && event->path.mnt)
+		new_file = dentry_open(&event->path,
+				       group->fanotify_data.f_flags | FMODE_NONOTIFY,
+				       current_cred());
+	else
+		new_file = ERR_PTR(-EOVERFLOW);
+	if (IS_ERR(new_file)) {
+		/*
+		 * we still send an event even if we can't open the file.  this
+		 * can happen when say tasks are gone and we try to open their
+		 * /proc files or we try to open a WRONLY file like in sysfs
+		 * we just send the errno to userspace since there isn't much
+		 * else we can do.
+		 */
+		put_unused_fd(client_fd);
+		client_fd = PTR_ERR(new_file);
+	} else {
+		*file = new_file;
+	}
+
+	return client_fd;
+}
+
+static int fill_event_metadata(struct fsnotify_group *group,
+				   struct fanotify_event_metadata *metadata,
+				   struct fsnotify_event *event,
+				   struct file **file)
+{
+	int ret = 0;
+
+	pr_debug("%s: group=%p metadata=%p event=%p\n", __func__,
+		 group, metadata, event);
+
+	*file = NULL;
+	metadata->event_len = FAN_EVENT_METADATA_LEN;
+	metadata->metadata_len = FAN_EVENT_METADATA_LEN;
+	metadata->vers = FANOTIFY_METADATA_VERSION;
+	metadata->reserved = 0;
+	metadata->mask = event->mask & FAN_ALL_OUTGOING_EVENTS;
+	metadata->pid = pid_vnr(event->tgid);
+	if (unlikely(event->mask & FAN_Q_OVERFLOW))
+		metadata->fd = FAN_NOFD;
+	else {
+		metadata->fd = create_fd(group, event, file);
+		if (metadata->fd < 0)
+			ret = metadata->fd;
+	}
+
+	return ret;
+}
+
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+static struct fanotify_response_event *dequeue_re(struct fsnotify_group *group,
+						  __s32 fd)
+{
+	struct fanotify_response_event *re, *return_re = NULL;
+
+	mutex_lock(&group->fanotify_data.access_mutex);
+	list_for_each_entry(re, &group->fanotify_data.access_list, list) {
+		if (re->fd != fd)
+			continue;
+
+		list_del_init(&re->list);
+		return_re = re;
+		break;
+	}
+	mutex_unlock(&group->fanotify_data.access_mutex);
+
+	pr_debug("%s: found return_re=%p\n", __func__, return_re);
+
+	return return_re;
+}
+
+static int process_access_response(struct fsnotify_group *group,
+				   struct fanotify_response *response_struct)
+{
+	struct fanotify_response_event *re;
+	__s32 fd = response_struct->fd;
+	__u32 response = response_struct->response;
+
+	pr_debug("%s: group=%p fd=%d response=%d\n", __func__, group,
+		 fd, response);
+	/*
+	 * make sure the response is valid, if invalid we do nothing and either
+	 * userspace can send a valid response or we will clean it up after the
+	 * timeout
+	 */
+	switch (response) {
+	case FAN_ALLOW:
+	case FAN_DENY:
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (fd < 0)
+		return -EINVAL;
+
+	re = dequeue_re(group, fd);
+	if (!re)
+		return -ENOENT;
+
+	re->event->response = response;
+
+	wake_up(&group->fanotify_data.access_waitq);
+
+	kmem_cache_free(fanotify_response_event_cache, re);
+
+	return 0;
+}
+
+static int prepare_for_access_response(struct fsnotify_group *group,
+				       struct fsnotify_event *event,
+				       __s32 fd)
+{
+	struct fanotify_response_event *re;
+
+	if (!(event->mask & FAN_ALL_PERM_EVENTS))
+		return 0;
+
+	re = kmem_cache_alloc(fanotify_response_event_cache, GFP_KERNEL);
+	if (!re)
+		return -ENOMEM;
+
+	re->event = event;
+	re->fd = fd;
+
+	mutex_lock(&group->fanotify_data.access_mutex);
+
+	if (atomic_read(&group->fanotify_data.bypass_perm)) {
+		mutex_unlock(&group->fanotify_data.access_mutex);
+		kmem_cache_free(fanotify_response_event_cache, re);
+		event->response = FAN_ALLOW;
+		return 0;
+	}
+		
+	list_add_tail(&re->list, &group->fanotify_data.access_list);
+	mutex_unlock(&group->fanotify_data.access_mutex);
+
+	return 0;
+}
+
+#else
+static int prepare_for_access_response(struct fsnotify_group *group,
+				       struct fsnotify_event *event,
+				       __s32 fd)
+{
+	return 0;
+}
+
+#endif
+
+static ssize_t copy_event_to_user(struct fsnotify_group *group,
+				  struct fsnotify_event *event,
+				  char __user *buf)
+{
+	struct fanotify_event_metadata fanotify_event_metadata;
+	struct file *f;
+	int fd, ret;
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	ret = fill_event_metadata(group, &fanotify_event_metadata, event, &f);
+	if (ret < 0)
+		goto out;
+
+	fd = fanotify_event_metadata.fd;
+	ret = -EFAULT;
+	if (copy_to_user(buf, &fanotify_event_metadata,
+			 fanotify_event_metadata.event_len))
+		goto out_close_fd;
+
+	ret = prepare_for_access_response(group, event, fd);
+	if (ret)
+		goto out_close_fd;
+
+	if (fd != FAN_NOFD)
+		fd_install(fd, f);
+	return fanotify_event_metadata.event_len;
+
+out_close_fd:
+	if (fd != FAN_NOFD) {
+		put_unused_fd(fd);
+		fput(f);
+	}
+out:
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	if (event->mask & FAN_ALL_PERM_EVENTS) {
+		event->response = FAN_DENY;
+		wake_up(&group->fanotify_data.access_waitq);
+	}
+#endif
+	return ret;
+}
+
+/* intofiy userspace file descriptor functions */
+static unsigned int fanotify_poll(struct file *file, poll_table *wait)
+{
+	struct fsnotify_group *group = file->private_data;
+	int ret = 0;
+
+	poll_wait(file, &group->notification_waitq, wait);
+	mutex_lock(&group->notification_mutex);
+	if (!fsnotify_notify_queue_is_empty(group))
+		ret = POLLIN | POLLRDNORM;
+	mutex_unlock(&group->notification_mutex);
+
+	return ret;
+}
+
+static ssize_t fanotify_read(struct file *file, char __user *buf,
+			     size_t count, loff_t *pos)
+{
+	struct fsnotify_group *group;
+	struct fsnotify_event *kevent;
+	char __user *start;
+	int ret;
+	DEFINE_WAIT(wait);
+
+	start = buf;
+	group = file->private_data;
+
+	pr_debug("%s: group=%p\n", __func__, group);
+
+	while (1) {
+		prepare_to_wait(&group->notification_waitq, &wait, TASK_INTERRUPTIBLE);
+
+		mutex_lock(&group->notification_mutex);
+		kevent = get_one_event(group, count);
+		mutex_unlock(&group->notification_mutex);
+
+		if (kevent) {
+			ret = PTR_ERR(kevent);
+			if (IS_ERR(kevent))
+				break;
+			ret = copy_event_to_user(group, kevent, buf);
+			fsnotify_put_event(kevent);
+			if (ret < 0)
+				break;
+			buf += ret;
+			count -= ret;
+			continue;
+		}
+
+		ret = -EAGAIN;
+		if (file->f_flags & O_NONBLOCK)
+			break;
+		ret = -ERESTARTSYS;
+		if (signal_pending(current))
+			break;
+
+		if (start != buf)
+			break;
+
+		schedule();
+	}
+
+	finish_wait(&group->notification_waitq, &wait);
+	if (start != buf && ret != -EFAULT)
+		ret = buf - start;
+	return ret;
+}
+
+static ssize_t fanotify_write(struct file *file, const char __user *buf, size_t count, loff_t *pos)
+{
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	struct fanotify_response response = { .fd = -1, .response = -1 };
+	struct fsnotify_group *group;
+	int ret;
+
+	group = file->private_data;
+
+	if (count > sizeof(response))
+		count = sizeof(response);
+
+	pr_debug("%s: group=%p count=%zu\n", __func__, group, count);
+
+	if (copy_from_user(&response, buf, count))
+		return -EFAULT;
+
+	ret = process_access_response(group, &response);
+	if (ret < 0)
+		count = ret;
+
+	return count;
+#else
+	return -EINVAL;
+#endif
+}
+
+static int fanotify_release(struct inode *ignored, struct file *file)
+{
+	struct fsnotify_group *group = file->private_data;
+
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	struct fanotify_response_event *re, *lre;
+
+	mutex_lock(&group->fanotify_data.access_mutex);
+
+	atomic_inc(&group->fanotify_data.bypass_perm);
+
+	list_for_each_entry_safe(re, lre, &group->fanotify_data.access_list, list) {
+		pr_debug("%s: found group=%p re=%p event=%p\n", __func__, group,
+			 re, re->event);
+
+		list_del_init(&re->list);
+		re->event->response = FAN_ALLOW;
+
+		kmem_cache_free(fanotify_response_event_cache, re);
+	}
+	mutex_unlock(&group->fanotify_data.access_mutex);
+
+	wake_up(&group->fanotify_data.access_waitq);
+#endif
+
+	if (file->f_flags & FASYNC)
+		fsnotify_fasync(-1, file, 0);
+
+	/* matches the fanotify_init->fsnotify_alloc_group */
+	fsnotify_destroy_group(group);
+
+	return 0;
+}
+
+static long fanotify_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+	struct fsnotify_group *group;
+	struct fsnotify_event_holder *holder;
+	void __user *p;
+	int ret = -ENOTTY;
+	size_t send_len = 0;
+
+	group = file->private_data;
+
+	p = (void __user *) arg;
+
+	switch (cmd) {
+	case FIONREAD:
+		mutex_lock(&group->notification_mutex);
+		list_for_each_entry(holder, &group->notification_list, event_list)
+			send_len += FAN_EVENT_METADATA_LEN;
+		mutex_unlock(&group->notification_mutex);
+		ret = put_user(send_len, (int __user *) p);
+		break;
+	}
+
+	return ret;
+}
+
+static const struct file_operations fanotify_fops = {
+	.show_fdinfo	= fanotify_show_fdinfo,
+	.poll		= fanotify_poll,
+	.read		= fanotify_read,
+	.write		= fanotify_write,
+	.fasync		= NULL,
+	.release	= fanotify_release,
+	.unlocked_ioctl	= fanotify_ioctl,
+	.compat_ioctl	= fanotify_ioctl,
+	.llseek		= noop_llseek,
+};
+
+static void fanotify_free_mark(struct fsnotify_mark *fsn_mark)
+{
+	kmem_cache_free(fanotify_mark_cache, fsn_mark);
+}
+
+static int fanotify_find_path(int dfd, const char __user *filename,
+			      struct path *path, unsigned int flags)
+{
+	int ret;
+
+	pr_debug("%s: dfd=%d filename=%p flags=%x\n", __func__,
+		 dfd, filename, flags);
+
+	if (filename == NULL) {
+		struct fd f = fdget(dfd);
+
+		ret = -EBADF;
+		if (!f.file)
+			goto out;
+
+		ret = -ENOTDIR;
+		if ((flags & FAN_MARK_ONLYDIR) &&
+		    !(S_ISDIR(file_inode(f.file)->i_mode))) {
+			fdput(f);
+			goto out;
+		}
+
+		*path = f.file->f_path;
+		path_get(path);
+		fdput(f);
+	} else {
+		unsigned int lookup_flags = 0;
+
+		if (!(flags & FAN_MARK_DONT_FOLLOW))
+			lookup_flags |= LOOKUP_FOLLOW;
+		if (flags & FAN_MARK_ONLYDIR)
+			lookup_flags |= LOOKUP_DIRECTORY;
+
+		ret = user_path_at(dfd, filename, lookup_flags, path);
+		if (ret)
+			goto out;
+	}
+
+	/* you can only watch an inode if you have read permissions on it */
+	ret = inode_permission(path->dentry->d_inode, MAY_READ);
+	if (ret)
+		path_put(path);
+out:
+	return ret;
+}
+
+static __u32 fanotify_mark_remove_from_mask(struct fsnotify_mark *fsn_mark,
+					    __u32 mask,
+					    unsigned int flags,
+					    int *destroy)
+{
+	__u32 oldmask;
+
+	spin_lock(&fsn_mark->lock);
+	if (!(flags & FAN_MARK_IGNORED_MASK)) {
+		oldmask = fsn_mark->mask;
+		fsnotify_set_mark_mask_locked(fsn_mark, (oldmask & ~mask));
+	} else {
+		oldmask = fsn_mark->ignored_mask;
+		fsnotify_set_mark_ignored_mask_locked(fsn_mark, (oldmask & ~mask));
+	}
+	spin_unlock(&fsn_mark->lock);
+
+	*destroy = !(oldmask & ~mask);
+
+	return mask & oldmask;
+}
+
+static int fanotify_remove_vfsmount_mark(struct fsnotify_group *group,
+					 struct vfsmount *mnt, __u32 mask,
+					 unsigned int flags)
+{
+	struct fsnotify_mark *fsn_mark = NULL;
+	__u32 removed;
+	int destroy_mark;
+
+	fsn_mark = fsnotify_find_vfsmount_mark(group, mnt);
+	if (!fsn_mark)
+		return -ENOENT;
+
+	removed = fanotify_mark_remove_from_mask(fsn_mark, mask, flags,
+						 &destroy_mark);
+	if (destroy_mark)
+		fsnotify_destroy_mark(fsn_mark, group);
+
+	fsnotify_put_mark(fsn_mark);
+	if (removed & real_mount(mnt)->mnt_fsnotify_mask)
+		fsnotify_recalc_vfsmount_mask(mnt);
+
+	return 0;
+}
+
+static int fanotify_remove_inode_mark(struct fsnotify_group *group,
+				      struct inode *inode, __u32 mask,
+				      unsigned int flags)
+{
+	struct fsnotify_mark *fsn_mark = NULL;
+	__u32 removed;
+	int destroy_mark;
+
+	fsn_mark = fsnotify_find_inode_mark(group, inode);
+	if (!fsn_mark)
+		return -ENOENT;
+
+	removed = fanotify_mark_remove_from_mask(fsn_mark, mask, flags,
+						 &destroy_mark);
+	if (destroy_mark)
+		fsnotify_destroy_mark(fsn_mark, group);
+	/* matches the fsnotify_find_inode_mark() */
+	fsnotify_put_mark(fsn_mark);
+	if (removed & inode->i_fsnotify_mask)
+		fsnotify_recalc_inode_mask(inode);
+
+	return 0;
+}
+
+static __u32 fanotify_mark_add_to_mask(struct fsnotify_mark *fsn_mark,
+				       __u32 mask,
+				       unsigned int flags)
+{
+	__u32 oldmask = -1;
+
+	spin_lock(&fsn_mark->lock);
+	if (!(flags & FAN_MARK_IGNORED_MASK)) {
+		oldmask = fsn_mark->mask;
+		fsnotify_set_mark_mask_locked(fsn_mark, (oldmask | mask));
+	} else {
+		__u32 tmask = fsn_mark->ignored_mask | mask;
+		fsnotify_set_mark_ignored_mask_locked(fsn_mark, tmask);
+		if (flags & FAN_MARK_IGNORED_SURV_MODIFY)
+			fsn_mark->flags |= FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY;
+	}
+
+	if (!(flags & FAN_MARK_ONDIR)) {
+		__u32 tmask = fsn_mark->ignored_mask | FAN_ONDIR;
+		fsnotify_set_mark_ignored_mask_locked(fsn_mark, tmask);
+	}
+
+	spin_unlock(&fsn_mark->lock);
+
+	return mask & ~oldmask;
+}
+
+static int fanotify_add_vfsmount_mark(struct fsnotify_group *group,
+				      struct vfsmount *mnt, __u32 mask,
+				      unsigned int flags)
+{
+	struct fsnotify_mark *fsn_mark;
+	__u32 added;
+	int ret = 0;
+
+	fsn_mark = fsnotify_find_vfsmount_mark(group, mnt);
+	if (!fsn_mark) {
+		if (atomic_read(&group->num_marks) > group->fanotify_data.max_marks)
+			return -ENOSPC;
+
+		fsn_mark = kmem_cache_alloc(fanotify_mark_cache, GFP_KERNEL);
+		if (!fsn_mark)
+			return -ENOMEM;
+
+		fsnotify_init_mark(fsn_mark, fanotify_free_mark);
+		ret = fsnotify_add_mark(fsn_mark, group, NULL, mnt, 0);
+		if (ret)
+			goto err;
+	}
+	added = fanotify_mark_add_to_mask(fsn_mark, mask, flags);
+
+	if (added & ~real_mount(mnt)->mnt_fsnotify_mask)
+		fsnotify_recalc_vfsmount_mask(mnt);
+err:
+	fsnotify_put_mark(fsn_mark);
+	return ret;
+}
+
+static int fanotify_add_inode_mark(struct fsnotify_group *group,
+				   struct inode *inode, __u32 mask,
+				   unsigned int flags)
+{
+	struct fsnotify_mark *fsn_mark;
+	__u32 added;
+	int ret = 0;
+
+	pr_debug("%s: group=%p inode=%p\n", __func__, group, inode);
+
+	/*
+	 * If some other task has this inode open for write we should not add
+	 * an ignored mark, unless that ignored mark is supposed to survive
+	 * modification changes anyway.
+	 */
+	if ((flags & FAN_MARK_IGNORED_MASK) &&
+	    !(flags & FAN_MARK_IGNORED_SURV_MODIFY) &&
+	    (atomic_read(&inode->i_writecount) > 0))
+		return 0;
+
+	fsn_mark = fsnotify_find_inode_mark(group, inode);
+	if (!fsn_mark) {
+		if (atomic_read(&group->num_marks) > group->fanotify_data.max_marks)
+			return -ENOSPC;
+
+		fsn_mark = kmem_cache_alloc(fanotify_mark_cache, GFP_KERNEL);
+		if (!fsn_mark)
+			return -ENOMEM;
+
+		fsnotify_init_mark(fsn_mark, fanotify_free_mark);
+		ret = fsnotify_add_mark(fsn_mark, group, inode, NULL, 0);
+		if (ret)
+			goto err;
+	}
+	added = fanotify_mark_add_to_mask(fsn_mark, mask, flags);
+
+	if (added & ~inode->i_fsnotify_mask)
+		fsnotify_recalc_inode_mask(inode);
+err:
+	fsnotify_put_mark(fsn_mark);
+	return ret;
+}
+
+/* fanotify syscalls */
+SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags)
+{
+	struct fsnotify_group *group;
+	int f_flags, fd;
+	struct user_struct *user;
+
+	pr_debug("%s: flags=%d event_f_flags=%d\n",
+		__func__, flags, event_f_flags);
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (flags & ~FAN_ALL_INIT_FLAGS)
+		return -EINVAL;
+
+	user = get_current_user();
+	if (atomic_read(&user->fanotify_listeners) > FANOTIFY_DEFAULT_MAX_LISTENERS) {
+		free_uid(user);
+		return -EMFILE;
+	}
+
+	f_flags = O_RDWR | FMODE_NONOTIFY;
+	if (flags & FAN_CLOEXEC)
+		f_flags |= O_CLOEXEC;
+	if (flags & FAN_NONBLOCK)
+		f_flags |= O_NONBLOCK;
+
+	/* fsnotify_alloc_group takes a ref.  Dropped in fanotify_release */
+	group = fsnotify_alloc_group(&fanotify_fsnotify_ops);
+	if (IS_ERR(group)) {
+		free_uid(user);
+		return PTR_ERR(group);
+	}
+
+	group->fanotify_data.user = user;
+	atomic_inc(&user->fanotify_listeners);
+
+	group->fanotify_data.f_flags = event_f_flags;
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	mutex_init(&group->fanotify_data.access_mutex);
+	init_waitqueue_head(&group->fanotify_data.access_waitq);
+	INIT_LIST_HEAD(&group->fanotify_data.access_list);
+	atomic_set(&group->fanotify_data.bypass_perm, 0);
+#endif
+	switch (flags & FAN_ALL_CLASS_BITS) {
+	case FAN_CLASS_NOTIF:
+		group->priority = FS_PRIO_0;
+		break;
+	case FAN_CLASS_CONTENT:
+		group->priority = FS_PRIO_1;
+		break;
+	case FAN_CLASS_PRE_CONTENT:
+		group->priority = FS_PRIO_2;
+		break;
+	default:
+		fd = -EINVAL;
+		goto out_destroy_group;
+	}
+
+	if (flags & FAN_UNLIMITED_QUEUE) {
+		fd = -EPERM;
+		if (!capable(CAP_SYS_ADMIN))
+			goto out_destroy_group;
+		group->max_events = UINT_MAX;
+	} else {
+		group->max_events = FANOTIFY_DEFAULT_MAX_EVENTS;
+	}
+
+	if (flags & FAN_UNLIMITED_MARKS) {
+		fd = -EPERM;
+		if (!capable(CAP_SYS_ADMIN))
+			goto out_destroy_group;
+		group->fanotify_data.max_marks = UINT_MAX;
+	} else {
+		group->fanotify_data.max_marks = FANOTIFY_DEFAULT_MAX_MARKS;
+	}
+
+	fd = anon_inode_getfd("[fanotify]", &fanotify_fops, group, f_flags);
+	if (fd < 0)
+		goto out_destroy_group;
+
+	return fd;
+
+out_destroy_group:
+	fsnotify_destroy_group(group);
+	return fd;
+}
+
+SYSCALL_DEFINE5(fanotify_mark, int, fanotify_fd, unsigned int, flags,
+			      __u64, mask, int, dfd,
+			      const char  __user *, pathname)
+{
+	struct inode *inode = NULL;
+	struct vfsmount *mnt = NULL;
+	struct fsnotify_group *group;
+	struct fd f;
+	struct path path;
+	int ret;
+
+	pr_debug("%s: fanotify_fd=%d flags=%x dfd=%d pathname=%p mask=%llx\n",
+		 __func__, fanotify_fd, flags, dfd, pathname, mask);
+
+	/* we only use the lower 32 bits as of right now. */
+	if (mask & ((__u64)0xffffffff << 32))
+		return -EINVAL;
+
+	if (flags & ~FAN_ALL_MARK_FLAGS)
+		return -EINVAL;
+	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE | FAN_MARK_FLUSH)) {
+	case FAN_MARK_ADD:		/* fallthrough */
+	case FAN_MARK_REMOVE:
+		if (!mask)
+			return -EINVAL;
+	case FAN_MARK_FLUSH:
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (mask & FAN_ONDIR) {
+		flags |= FAN_MARK_ONDIR;
+		mask &= ~FAN_ONDIR;
+	}
+
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	if (mask & ~(FAN_ALL_EVENTS | FAN_ALL_PERM_EVENTS | FAN_EVENT_ON_CHILD))
+#else
+	if (mask & ~(FAN_ALL_EVENTS | FAN_EVENT_ON_CHILD))
+#endif
+		return -EINVAL;
+
+	f = fdget(fanotify_fd);
+	if (unlikely(!f.file))
+		return -EBADF;
+
+	/* verify that this is indeed an fanotify instance */
+	ret = -EINVAL;
+	if (unlikely(f.file->f_op != &fanotify_fops))
+		goto fput_and_out;
+	group = f.file->private_data;
+
+	/*
+	 * group->priority == FS_PRIO_0 == FAN_CLASS_NOTIF.  These are not
+	 * allowed to set permissions events.
+	 */
+	ret = -EINVAL;
+	if (mask & FAN_ALL_PERM_EVENTS &&
+	    group->priority == FS_PRIO_0)
+		goto fput_and_out;
+
+	ret = fanotify_find_path(dfd, pathname, &path, flags);
+	if (ret)
+		goto fput_and_out;
+
+	/* inode held in place by reference to path; group by fget on fd */
+	if (!(flags & FAN_MARK_MOUNT))
+		inode = path.dentry->d_inode;
+	else
+		mnt = path.mnt;
+
+	/* create/update an inode mark */
+	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE | FAN_MARK_FLUSH)) {
+	case FAN_MARK_ADD:
+		if (flags & FAN_MARK_MOUNT)
+			ret = fanotify_add_vfsmount_mark(group, mnt, mask, flags);
+		else
+			ret = fanotify_add_inode_mark(group, inode, mask, flags);
+		break;
+	case FAN_MARK_REMOVE:
+		if (flags & FAN_MARK_MOUNT)
+			ret = fanotify_remove_vfsmount_mark(group, mnt, mask, flags);
+		else
+			ret = fanotify_remove_inode_mark(group, inode, mask, flags);
+		break;
+	case FAN_MARK_FLUSH:
+		if (flags & FAN_MARK_MOUNT)
+			fsnotify_clear_vfsmount_marks_by_group(group);
+		else
+			fsnotify_clear_inode_marks_by_group(group);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	path_put(&path);
+fput_and_out:
+	fdput(f);
+	return ret;
+}
+
+#ifdef CONFIG_COMPAT
+COMPAT_SYSCALL_DEFINE6(fanotify_mark,
+				int, fanotify_fd, unsigned int, flags,
+				__u32, mask0, __u32, mask1, int, dfd,
+				const char  __user *, pathname)
+{
+	return sys_fanotify_mark(fanotify_fd, flags,
+#ifdef __BIG_ENDIAN
+				((__u64)mask1 << 32) | mask0,
+#else
+				((__u64)mask0 << 32) | mask1,
+#endif
+				 dfd, pathname);
+}
+#endif
+
+/*
+ * fanotify_user_setup - Our initialization function.  Note that we cannot return
+ * error because we have compiled-in VFS hooks.  So an (unlikely) failure here
+ * must result in panic().
+ */
+static int __init fanotify_user_setup(void)
+{
+	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
+	fanotify_response_event_cache = KMEM_CACHE(fanotify_response_event,
+						   SLAB_PANIC);
+
+	return 0;
+}
+device_initcall(fanotify_user_setup);

fs/notify/fdinfo.c

@@ -0,0 +1,179 @@
+#include <linux/file.h>
+#include <linux/fs.h>
+#include <linux/fsnotify_backend.h>
+#include <linux/idr.h>
+#include <linux/init.h>
+#include <linux/inotify.h>
+#include <linux/fanotify.h>
+#include <linux/kernel.h>
+#include <linux/namei.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/seq_file.h>
+#include <linux/proc_fs.h>
+#include <linux/exportfs.h>
+
+#include "inotify/inotify.h"
+#include "../fs/mount.h"
+
+#if defined(CONFIG_PROC_FS)
+
+#if defined(CONFIG_INOTIFY_USER) || defined(CONFIG_FANOTIFY)
+
+static int show_fdinfo(struct seq_file *m, struct file *f,
+		       int (*show)(struct seq_file *m, struct fsnotify_mark *mark))
+{
+	struct fsnotify_group *group = f->private_data;
+	struct fsnotify_mark *mark;
+	int ret = 0;
+
+	mutex_lock(&group->mark_mutex);
+	list_for_each_entry(mark, &group->marks_list, g_list) {
+		ret = show(m, mark);
+		if (ret)
+			break;
+	}
+	mutex_unlock(&group->mark_mutex);
+	return ret;
+}
+
+#if defined(CONFIG_EXPORTFS)
+static int show_mark_fhandle(struct seq_file *m, struct inode *inode)
+{
+	struct {
+		struct file_handle handle;
+		u8 pad[64];
+	} f;
+	int size, ret, i;
+
+	f.handle.handle_bytes = sizeof(f.pad);
+	size = f.handle.handle_bytes >> 2;
+
+	ret = exportfs_encode_inode_fh(inode, (struct fid *)f.handle.f_handle, &size, 0);
+	if ((ret == 255) || (ret == -ENOSPC)) {
+		WARN_ONCE(1, "Can't encode file handler for inotify: %d\n", ret);
+		return 0;
+	}
+
+	f.handle.handle_type = ret;
+	f.handle.handle_bytes = size * sizeof(u32);
+
+	ret = seq_printf(m, "fhandle-bytes:%x fhandle-type:%x f_handle:",
+			 f.handle.handle_bytes, f.handle.handle_type);
+
+	for (i = 0; i < f.handle.handle_bytes; i++)
+		ret |= seq_printf(m, "%02x", (int)f.handle.f_handle[i]);
+
+	return ret;
+}
+#else
+static int show_mark_fhandle(struct seq_file *m, struct inode *inode)
+{
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_INOTIFY_USER
+
+static int inotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark)
+{
+	struct inotify_inode_mark *inode_mark;
+	struct inode *inode;
+	int ret = 0;
+
+	if (!(mark->flags & (FSNOTIFY_MARK_FLAG_ALIVE | FSNOTIFY_MARK_FLAG_INODE)))
+		return 0;
+
+	inode_mark = container_of(mark, struct inotify_inode_mark, fsn_mark);
+	inode = igrab(mark->i.inode);
+	if (inode) {
+		ret = seq_printf(m, "inotify wd:%x ino:%lx sdev:%x "
+				 "mask:%x ignored_mask:%x ",
+				 inode_mark->wd, inode->i_ino,
+				 inode->i_sb->s_dev,
+				 mark->mask, mark->ignored_mask);
+		ret |= show_mark_fhandle(m, inode);
+		ret |= seq_putc(m, '\n');
+		iput(inode);
+	}
+
+	return ret;
+}
+
+int inotify_show_fdinfo(struct seq_file *m, struct file *f)
+{
+	return show_fdinfo(m, f, inotify_fdinfo);
+}
+
+#endif /* CONFIG_INOTIFY_USER */
+
+#ifdef CONFIG_FANOTIFY
+
+static int fanotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark)
+{
+	unsigned int mflags = 0;
+	struct inode *inode;
+	int ret = 0;
+
+	if (!(mark->flags & FSNOTIFY_MARK_FLAG_ALIVE))
+		return 0;
+
+	if (mark->flags & FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY)
+		mflags |= FAN_MARK_IGNORED_SURV_MODIFY;
+
+	if (mark->flags & FSNOTIFY_MARK_FLAG_INODE) {
+		inode = igrab(mark->i.inode);
+		if (!inode)
+			goto out;
+		ret = seq_printf(m, "fanotify ino:%lx sdev:%x "
+				 "mflags:%x mask:%x ignored_mask:%x ",
+				 inode->i_ino, inode->i_sb->s_dev,
+				 mflags, mark->mask, mark->ignored_mask);
+		ret |= show_mark_fhandle(m, inode);
+		ret |= seq_putc(m, '\n');
+		iput(inode);
+	} else if (mark->flags & FSNOTIFY_MARK_FLAG_VFSMOUNT) {
+		struct mount *mnt = real_mount(mark->m.mnt);
+
+		ret = seq_printf(m, "fanotify mnt_id:%x mflags:%x mask:%x "
+				 "ignored_mask:%x\n", mnt->mnt_id, mflags,
+				 mark->mask, mark->ignored_mask);
+	}
+out:
+	return ret;
+}
+
+int fanotify_show_fdinfo(struct seq_file *m, struct file *f)
+{
+	struct fsnotify_group *group = f->private_data;
+	unsigned int flags = 0;
+
+	switch (group->priority) {
+	case FS_PRIO_0:
+		flags |= FAN_CLASS_NOTIF;
+		break;
+	case FS_PRIO_1:
+		flags |= FAN_CLASS_CONTENT;
+		break;
+	case FS_PRIO_2:
+		flags |= FAN_CLASS_PRE_CONTENT;
+		break;
+	}
+
+	if (group->max_events == UINT_MAX)
+		flags |= FAN_UNLIMITED_QUEUE;
+
+	if (group->fanotify_data.max_marks == UINT_MAX)
+		flags |= FAN_UNLIMITED_MARKS;
+
+	seq_printf(m, "fanotify flags:%x event-flags:%x\n",
+		   flags, group->fanotify_data.f_flags);
+
+	return show_fdinfo(m, f, fanotify_fdinfo);
+}
+
+#endif /* CONFIG_FANOTIFY */
+
+#endif /* CONFIG_INOTIFY_USER || CONFIG_FANOTIFY */
+
+#endif /* CONFIG_PROC_FS */

fs/notify/fdinfo.h

@@ -0,0 +1,27 @@
+#ifndef __FSNOTIFY_FDINFO_H__
+#define __FSNOTIFY_FDINFO_H__
+
+#include <linux/errno.h>
+#include <linux/proc_fs.h>
+
+struct seq_file;
+struct file;
+
+#ifdef CONFIG_PROC_FS
+
+#ifdef CONFIG_INOTIFY_USER
+extern int inotify_show_fdinfo(struct seq_file *m, struct file *f);
+#endif
+
+#ifdef CONFIG_FANOTIFY
+extern int fanotify_show_fdinfo(struct seq_file *m, struct file *f);
+#endif
+
+#else /* CONFIG_PROC_FS */
+
+#define inotify_show_fdinfo	NULL
+#define fanotify_show_fdinfo	NULL
+
+#endif /* CONFIG_PROC_FS */
+
+#endif /* __FSNOTIFY_FDINFO_H__ */

fs/notify/fsnotify.c

@@ -0,0 +1,311 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/dcache.h>
+#include <linux/fs.h>
+#include <linux/gfp.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/mount.h>
+#include <linux/srcu.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+#include "../mount.h"
+
+/*
+ * Clear all of the marks on an inode when it is being evicted from core
+ */
+void __fsnotify_inode_delete(struct inode *inode)
+{
+	fsnotify_clear_marks_by_inode(inode);
+}
+EXPORT_SYMBOL_GPL(__fsnotify_inode_delete);
+
+void __fsnotify_vfsmount_delete(struct vfsmount *mnt)
+{
+	fsnotify_clear_marks_by_mount(mnt);
+}
+
+/*
+ * Given an inode, first check if we care what happens to our children.  Inotify
+ * and dnotify both tell their parents about events.  If we care about any event
+ * on a child we run all of our children and set a dentry flag saying that the
+ * parent cares.  Thus when an event happens on a child it can quickly tell if
+ * if there is a need to find a parent and send the event to the parent.
+ */
+void __fsnotify_update_child_dentry_flags(struct inode *inode)
+{
+	struct dentry *alias;
+	int watched;
+
+	if (!S_ISDIR(inode->i_mode))
+		return;
+
+	/* determine if the children should tell inode about their events */
+	watched = fsnotify_inode_watches_children(inode);
+
+	spin_lock(&inode->i_lock);
+	/* run all of the dentries associated with this inode.  Since this is a
+	 * directory, there damn well better only be one item on this list */
+	hlist_for_each_entry(alias, &inode->i_dentry, d_u.d_alias) {
+		struct dentry *child;
+
+		/* run all of the children of the original inode and fix their
+		 * d_flags to indicate parental interest (their parent is the
+		 * original inode) */
+		spin_lock(&alias->d_lock);
+		list_for_each_entry(child, &alias->d_subdirs, d_child) {
+			if (!child->d_inode)
+				continue;
+
+			spin_lock_nested(&child->d_lock, DENTRY_D_LOCK_NESTED);
+			if (watched)
+				child->d_flags |= DCACHE_FSNOTIFY_PARENT_WATCHED;
+			else
+				child->d_flags &= ~DCACHE_FSNOTIFY_PARENT_WATCHED;
+			spin_unlock(&child->d_lock);
+		}
+		spin_unlock(&alias->d_lock);
+	}
+	spin_unlock(&inode->i_lock);
+}
+
+/* Notify this dentry's parent about a child's events. */
+int __fsnotify_parent(struct path *path, struct dentry *dentry, __u32 mask)
+{
+	struct dentry *parent;
+	struct inode *p_inode;
+	int ret = 0;
+
+	if (!dentry)
+		dentry = path->dentry;
+
+	if (!(dentry->d_flags & DCACHE_FSNOTIFY_PARENT_WATCHED))
+		return 0;
+
+	parent = dget_parent(dentry);
+	p_inode = parent->d_inode;
+
+	if (unlikely(!fsnotify_inode_watches_children(p_inode)))
+		__fsnotify_update_child_dentry_flags(p_inode);
+	else if (p_inode->i_fsnotify_mask & mask) {
+		/* we are notifying a parent so come up with the new mask which
+		 * specifies these are events which came from a child. */
+		mask |= FS_EVENT_ON_CHILD;
+
+		if (path)
+			ret = fsnotify(p_inode, mask, path, FSNOTIFY_EVENT_PATH,
+				       dentry->d_name.name, 0);
+		else
+			ret = fsnotify(p_inode, mask, dentry->d_inode, FSNOTIFY_EVENT_INODE,
+				       dentry->d_name.name, 0);
+	}
+
+	dput(parent);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(__fsnotify_parent);
+
+static int send_to_group(struct inode *to_tell,
+			 struct fsnotify_mark *inode_mark,
+			 struct fsnotify_mark *vfsmount_mark,
+			 __u32 mask, void *data,
+			 int data_is, u32 cookie,
+			 const unsigned char *file_name,
+			 struct fsnotify_event **event)
+{
+	struct fsnotify_group *group = NULL;
+	__u32 inode_test_mask = 0;
+	__u32 vfsmount_test_mask = 0;
+
+	if (unlikely(!inode_mark && !vfsmount_mark)) {
+		BUG();
+		return 0;
+	}
+
+	/* clear ignored on inode modification */
+	if (mask & FS_MODIFY) {
+		if (inode_mark &&
+		    !(inode_mark->flags & FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY))
+			inode_mark->ignored_mask = 0;
+		if (vfsmount_mark &&
+		    !(vfsmount_mark->flags & FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY))
+			vfsmount_mark->ignored_mask = 0;
+	}
+
+	/* does the inode mark tell us to do something? */
+	if (inode_mark) {
+		group = inode_mark->group;
+		inode_test_mask = (mask & ~FS_EVENT_ON_CHILD);
+		inode_test_mask &= inode_mark->mask;
+		inode_test_mask &= ~inode_mark->ignored_mask;
+	}
+
+	/* does the vfsmount_mark tell us to do something? */
+	if (vfsmount_mark) {
+		vfsmount_test_mask = (mask & ~FS_EVENT_ON_CHILD);
+		group = vfsmount_mark->group;
+		vfsmount_test_mask &= vfsmount_mark->mask;
+		vfsmount_test_mask &= ~vfsmount_mark->ignored_mask;
+		if (inode_mark)
+			vfsmount_test_mask &= ~inode_mark->ignored_mask;
+	}
+
+	pr_debug("%s: group=%p to_tell=%p mask=%x inode_mark=%p"
+		 " inode_test_mask=%x vfsmount_mark=%p vfsmount_test_mask=%x"
+		 " data=%p data_is=%d cookie=%d event=%p\n",
+		 __func__, group, to_tell, mask, inode_mark,
+		 inode_test_mask, vfsmount_mark, vfsmount_test_mask, data,
+		 data_is, cookie, *event);
+
+	if (!inode_test_mask && !vfsmount_test_mask)
+		return 0;
+
+	if (group->ops->should_send_event(group, to_tell, inode_mark,
+					  vfsmount_mark, mask, data,
+					  data_is) == false)
+		return 0;
+
+	if (!*event) {
+		*event = fsnotify_create_event(to_tell, mask, data,
+						data_is, file_name,
+						cookie, GFP_KERNEL);
+		if (!*event)
+			return -ENOMEM;
+	}
+	return group->ops->handle_event(group, inode_mark, vfsmount_mark, *event);
+}
+
+/*
+ * This is the main call to fsnotify.  The VFS calls into hook specific functions
+ * in linux/fsnotify.h.  Those functions then in turn call here.  Here will call
+ * out to all of the registered fsnotify_group.  Those groups can then use the
+ * notification event in whatever means they feel necessary.
+ */
+int fsnotify(struct inode *to_tell, __u32 mask, void *data, int data_is,
+	     const unsigned char *file_name, u32 cookie)
+{
+	struct hlist_node *inode_node = NULL, *vfsmount_node = NULL;
+	struct fsnotify_mark *inode_mark = NULL, *vfsmount_mark = NULL;
+	struct fsnotify_group *inode_group, *vfsmount_group;
+	struct fsnotify_event *event = NULL;
+	struct mount *mnt;
+	int idx, ret = 0;
+	/* global tests shouldn't care about events on child only the specific event */
+	__u32 test_mask = (mask & ~FS_EVENT_ON_CHILD);
+
+	if (data_is == FSNOTIFY_EVENT_PATH)
+		mnt = real_mount(((struct path *)data)->mnt);
+	else
+		mnt = NULL;
+
+	/*
+	 * if this is a modify event we may need to clear the ignored masks
+	 * otherwise return if neither the inode nor the vfsmount care about
+	 * this type of event.
+	 */
+	if (!(mask & FS_MODIFY) &&
+	    !(test_mask & to_tell->i_fsnotify_mask) &&
+	    !(mnt && test_mask & mnt->mnt_fsnotify_mask))
+		return 0;
+
+	idx = srcu_read_lock(&fsnotify_mark_srcu);
+
+	if ((mask & FS_MODIFY) ||
+	    (test_mask & to_tell->i_fsnotify_mask))
+		inode_node = srcu_dereference(to_tell->i_fsnotify_marks.first,
+					      &fsnotify_mark_srcu);
+
+	if (mnt && ((mask & FS_MODIFY) ||
+		    (test_mask & mnt->mnt_fsnotify_mask))) {
+		vfsmount_node = srcu_dereference(mnt->mnt_fsnotify_marks.first,
+						 &fsnotify_mark_srcu);
+		inode_node = srcu_dereference(to_tell->i_fsnotify_marks.first,
+					      &fsnotify_mark_srcu);
+	}
+
+	while (inode_node || vfsmount_node) {
+		inode_group = vfsmount_group = NULL;
+
+		if (inode_node) {
+			inode_mark = hlist_entry(srcu_dereference(inode_node, &fsnotify_mark_srcu),
+						 struct fsnotify_mark, i.i_list);
+			inode_group = inode_mark->group;
+		}
+
+		if (vfsmount_node) {
+			vfsmount_mark = hlist_entry(srcu_dereference(vfsmount_node, &fsnotify_mark_srcu),
+							struct fsnotify_mark, m.m_list);
+			vfsmount_group = vfsmount_mark->group;
+		}
+
+		if (inode_group > vfsmount_group) {
+			/* handle inode */
+			ret = send_to_group(to_tell, inode_mark, NULL, mask, data,
+					    data_is, cookie, file_name, &event);
+			/* we didn't use the vfsmount_mark */
+			vfsmount_group = NULL;
+		} else if (vfsmount_group > inode_group) {
+			ret = send_to_group(to_tell, NULL, vfsmount_mark, mask, data,
+					    data_is, cookie, file_name, &event);
+			inode_group = NULL;
+		} else {
+			ret = send_to_group(to_tell, inode_mark, vfsmount_mark,
+					    mask, data, data_is, cookie, file_name,
+					    &event);
+		}
+
+		if (ret && (mask & ALL_FSNOTIFY_PERM_EVENTS))
+			goto out;
+
+		if (inode_group)
+			inode_node = srcu_dereference(inode_node->next,
+						      &fsnotify_mark_srcu);
+		if (vfsmount_group)
+			vfsmount_node = srcu_dereference(vfsmount_node->next,
+							 &fsnotify_mark_srcu);
+	}
+	ret = 0;
+out:
+	srcu_read_unlock(&fsnotify_mark_srcu, idx);
+	/*
+	 * fsnotify_create_event() took a reference so the event can't be cleaned
+	 * up while we are still trying to add it to lists, drop that one.
+	 */
+	if (event)
+		fsnotify_put_event(event);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(fsnotify);
+
+static __init int fsnotify_init(void)
+{
+	int ret;
+
+	BUG_ON(hweight32(ALL_FSNOTIFY_EVENTS) != 23);
+
+	ret = init_srcu_struct(&fsnotify_mark_srcu);
+	if (ret)
+		panic("initializing fsnotify_mark_srcu");
+
+	return 0;
+}
+core_initcall(fsnotify_init);

fs/notify/fsnotify.h

@@ -0,0 +1,47 @@
+#ifndef __FS_NOTIFY_FSNOTIFY_H_
+#define __FS_NOTIFY_FSNOTIFY_H_
+
+#include <linux/list.h>
+#include <linux/fsnotify.h>
+#include <linux/srcu.h>
+#include <linux/types.h>
+
+/* destroy all events sitting in this groups notification queue */
+extern void fsnotify_flush_notify(struct fsnotify_group *group);
+
+/* protects reads of inode and vfsmount marks list */
+extern struct srcu_struct fsnotify_mark_srcu;
+
+extern void fsnotify_set_inode_mark_mask_locked(struct fsnotify_mark *fsn_mark,
+						__u32 mask);
+/* add a mark to an inode */
+extern int fsnotify_add_inode_mark(struct fsnotify_mark *mark,
+				   struct fsnotify_group *group, struct inode *inode,
+				   int allow_dups);
+/* add a mark to a vfsmount */
+extern int fsnotify_add_vfsmount_mark(struct fsnotify_mark *mark,
+				      struct fsnotify_group *group, struct vfsmount *mnt,
+				      int allow_dups);
+
+/* final kfree of a group */
+extern void fsnotify_final_destroy_group(struct fsnotify_group *group);
+
+/* vfsmount specific destruction of a mark */
+extern void fsnotify_destroy_vfsmount_mark(struct fsnotify_mark *mark);
+/* inode specific destruction of a mark */
+extern void fsnotify_destroy_inode_mark(struct fsnotify_mark *mark);
+/* run the list of all marks associated with inode and flag them to be freed */
+extern void fsnotify_clear_marks_by_inode(struct inode *inode);
+/* run the list of all marks associated with vfsmount and flag them to be freed */
+extern void fsnotify_clear_marks_by_mount(struct vfsmount *mnt);
+/*
+ * update the dentry->d_flags of all of inode's children to indicate if inode cares
+ * about events that happen to its children.
+ */
+extern void __fsnotify_update_child_dentry_flags(struct inode *inode);
+
+/* allocate and destroy and event holder to attach events to notification/access queues */
+extern struct fsnotify_event_holder *fsnotify_alloc_event_holder(void);
+extern void fsnotify_destroy_event_holder(struct fsnotify_event_holder *holder);
+
+#endif	/* __FS_NOTIFY_FSNOTIFY_H_ */

fs/notify/group.c

@@ -0,0 +1,111 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/srcu.h>
+#include <linux/rculist.h>
+#include <linux/wait.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+
+#include <linux/atomic.h>
+
+/*
+ * Final freeing of a group
+ */
+void fsnotify_final_destroy_group(struct fsnotify_group *group)
+{
+	if (group->ops->free_group_priv)
+		group->ops->free_group_priv(group);
+
+	kfree(group);
+}
+
+/*
+ * Trying to get rid of a group. Remove all marks, flush all events and release
+ * the group reference.
+ * Note that another thread calling fsnotify_clear_marks_by_group() may still
+ * hold a ref to the group.
+ */
+void fsnotify_destroy_group(struct fsnotify_group *group)
+{
+	/* clear all inode marks for this group */
+	fsnotify_clear_marks_by_group(group);
+
+	synchronize_srcu(&fsnotify_mark_srcu);
+
+	/* clear the notification queue of all events */
+	fsnotify_flush_notify(group);
+
+	fsnotify_put_group(group);
+}
+
+/*
+ * Get reference to a group.
+ */
+void fsnotify_get_group(struct fsnotify_group *group)
+{
+	atomic_inc(&group->refcnt);
+}
+
+/*
+ * Drop a reference to a group.  Free it if it's through.
+ */
+void fsnotify_put_group(struct fsnotify_group *group)
+{
+	if (atomic_dec_and_test(&group->refcnt))
+		fsnotify_final_destroy_group(group);
+}
+
+/*
+ * Create a new fsnotify_group and hold a reference for the group returned.
+ */
+struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops)
+{
+	struct fsnotify_group *group;
+
+	group = kzalloc(sizeof(struct fsnotify_group), GFP_KERNEL);
+	if (!group)
+		return ERR_PTR(-ENOMEM);
+
+	/* set to 0 when there a no external references to this group */
+	atomic_set(&group->refcnt, 1);
+	atomic_set(&group->num_marks, 0);
+
+	mutex_init(&group->notification_mutex);
+	INIT_LIST_HEAD(&group->notification_list);
+	init_waitqueue_head(&group->notification_waitq);
+	group->max_events = UINT_MAX;
+
+	mutex_init(&group->mark_mutex);
+	INIT_LIST_HEAD(&group->marks_list);
+
+	group->ops = ops;
+
+	return group;
+}
+
+int fsnotify_fasync(int fd, struct file *file, int on)
+{
+	struct fsnotify_group *group = file->private_data;
+
+	return fasync_helper(fd, file, on, &group->fsn_fa) >= 0 ? 0 : -EIO;
+}

fs/notify/inode_mark.c

@@ -0,0 +1,321 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/spinlock.h>
+
+#include <linux/atomic.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+
+#include "../internal.h"
+
+/*
+ * Recalculate the mask of events relevant to a given inode locked.
+ */
+static void fsnotify_recalc_inode_mask_locked(struct inode *inode)
+{
+	struct fsnotify_mark *mark;
+	__u32 new_mask = 0;
+
+	assert_spin_locked(&inode->i_lock);
+
+	hlist_for_each_entry(mark, &inode->i_fsnotify_marks, i.i_list)
+		new_mask |= mark->mask;
+	inode->i_fsnotify_mask = new_mask;
+}
+
+/*
+ * Recalculate the inode->i_fsnotify_mask, or the mask of all FS_* event types
+ * any notifier is interested in hearing for this inode.
+ */
+void fsnotify_recalc_inode_mask(struct inode *inode)
+{
+	spin_lock(&inode->i_lock);
+	fsnotify_recalc_inode_mask_locked(inode);
+	spin_unlock(&inode->i_lock);
+
+	__fsnotify_update_child_dentry_flags(inode);
+}
+
+void fsnotify_destroy_inode_mark(struct fsnotify_mark *mark)
+{
+	struct inode *inode = mark->i.inode;
+
+	BUG_ON(!mutex_is_locked(&mark->group->mark_mutex));
+	assert_spin_locked(&mark->lock);
+
+	spin_lock(&inode->i_lock);
+
+	hlist_del_init_rcu(&mark->i.i_list);
+	mark->i.inode = NULL;
+
+	/*
+	 * this mark is now off the inode->i_fsnotify_marks list and we
+	 * hold the inode->i_lock, so this is the perfect time to update the
+	 * inode->i_fsnotify_mask
+	 */
+	fsnotify_recalc_inode_mask_locked(inode);
+
+	spin_unlock(&inode->i_lock);
+}
+
+/*
+ * Given an inode, destroy all of the marks associated with that inode.
+ */
+void fsnotify_clear_marks_by_inode(struct inode *inode)
+{
+	struct fsnotify_mark *mark, *lmark;
+	struct hlist_node *n;
+	LIST_HEAD(free_list);
+
+	spin_lock(&inode->i_lock);
+	hlist_for_each_entry_safe(mark, n, &inode->i_fsnotify_marks, i.i_list) {
+		list_add(&mark->i.free_i_list, &free_list);
+		hlist_del_init_rcu(&mark->i.i_list);
+		fsnotify_get_mark(mark);
+	}
+	spin_unlock(&inode->i_lock);
+
+	list_for_each_entry_safe(mark, lmark, &free_list, i.free_i_list) {
+		struct fsnotify_group *group;
+
+		spin_lock(&mark->lock);
+		fsnotify_get_group(mark->group);
+		group = mark->group;
+		spin_unlock(&mark->lock);
+
+		fsnotify_destroy_mark(mark, group);
+		fsnotify_put_mark(mark);
+		fsnotify_put_group(group);
+	}
+}
+
+/*
+ * Given a group clear all of the inode marks associated with that group.
+ */
+void fsnotify_clear_inode_marks_by_group(struct fsnotify_group *group)
+{
+	fsnotify_clear_marks_by_group_flags(group, FSNOTIFY_MARK_FLAG_INODE);
+}
+
+/*
+ * given a group and inode, find the mark associated with that combination.
+ * if found take a reference to that mark and return it, else return NULL
+ */
+static struct fsnotify_mark *fsnotify_find_inode_mark_locked(
+		struct fsnotify_group *group,
+		struct inode *inode)
+{
+	struct fsnotify_mark *mark;
+
+	assert_spin_locked(&inode->i_lock);
+
+	hlist_for_each_entry(mark, &inode->i_fsnotify_marks, i.i_list) {
+		if (mark->group == group) {
+			fsnotify_get_mark(mark);
+			return mark;
+		}
+	}
+	return NULL;
+}
+
+/*
+ * given a group and inode, find the mark associated with that combination.
+ * if found take a reference to that mark and return it, else return NULL
+ */
+struct fsnotify_mark *fsnotify_find_inode_mark(struct fsnotify_group *group,
+					       struct inode *inode)
+{
+	struct fsnotify_mark *mark;
+
+	spin_lock(&inode->i_lock);
+	mark = fsnotify_find_inode_mark_locked(group, inode);
+	spin_unlock(&inode->i_lock);
+
+	return mark;
+}
+
+/*
+ * If we are setting a mark mask on an inode mark we should pin the inode
+ * in memory.
+ */
+void fsnotify_set_inode_mark_mask_locked(struct fsnotify_mark *mark,
+					 __u32 mask)
+{
+	struct inode *inode;
+
+	assert_spin_locked(&mark->lock);
+
+	if (mask &&
+	    mark->i.inode &&
+	    !(mark->flags & FSNOTIFY_MARK_FLAG_OBJECT_PINNED)) {
+		mark->flags |= FSNOTIFY_MARK_FLAG_OBJECT_PINNED;
+		inode = igrab(mark->i.inode);
+		/*
+		 * we shouldn't be able to get here if the inode wasn't
+		 * already safely held in memory.  But bug in case it
+		 * ever is wrong.
+		 */
+		BUG_ON(!inode);
+	}
+}
+
+/*
+ * Attach an initialized mark to a given inode.
+ * These marks may be used for the fsnotify backend to determine which
+ * event types should be delivered to which group and for which inodes.  These
+ * marks are ordered according to priority, highest number first, and then by
+ * the group's location in memory.
+ */
+int fsnotify_add_inode_mark(struct fsnotify_mark *mark,
+			    struct fsnotify_group *group, struct inode *inode,
+			    int allow_dups)
+{
+	struct fsnotify_mark *lmark, *last = NULL;
+	int ret = 0;
+
+	mark->flags |= FSNOTIFY_MARK_FLAG_INODE;
+
+	BUG_ON(!mutex_is_locked(&group->mark_mutex));
+	assert_spin_locked(&mark->lock);
+
+	spin_lock(&inode->i_lock);
+
+	mark->i.inode = inode;
+
+	/* is mark the first mark? */
+	if (hlist_empty(&inode->i_fsnotify_marks)) {
+		hlist_add_head_rcu(&mark->i.i_list, &inode->i_fsnotify_marks);
+		goto out;
+	}
+
+	/* should mark be in the middle of the current list? */
+	hlist_for_each_entry(lmark, &inode->i_fsnotify_marks, i.i_list) {
+		last = lmark;
+
+		if ((lmark->group == group) && !allow_dups) {
+			ret = -EEXIST;
+			goto out;
+		}
+
+		if (mark->group->priority < lmark->group->priority)
+			continue;
+
+		if ((mark->group->priority == lmark->group->priority) &&
+		    (mark->group < lmark->group))
+			continue;
+
+		hlist_add_before_rcu(&mark->i.i_list, &lmark->i.i_list);
+		goto out;
+	}
+
+	BUG_ON(last == NULL);
+	/* mark should be the last entry.  last is the current last entry */
+	hlist_add_after_rcu(&last->i.i_list, &mark->i.i_list);
+out:
+	fsnotify_recalc_inode_mask_locked(inode);
+	spin_unlock(&inode->i_lock);
+
+	return ret;
+}
+
+/**
+ * fsnotify_unmount_inodes - an sb is unmounting.  handle any watched inodes.
+ * @list: list of inodes being unmounted (sb->s_inodes)
+ *
+ * Called during unmount with no locks held, so needs to be safe against
+ * concurrent modifiers. We temporarily drop inode_sb_list_lock and CAN block.
+ */
+void fsnotify_unmount_inodes(struct list_head *list)
+{
+	struct inode *inode, *next_i, *need_iput = NULL;
+
+	spin_lock(&inode_sb_list_lock);
+	list_for_each_entry_safe(inode, next_i, list, i_sb_list) {
+		struct inode *need_iput_tmp;
+
+		/*
+		 * We cannot __iget() an inode in state I_FREEING,
+		 * I_WILL_FREE, or I_NEW which is fine because by that point
+		 * the inode cannot have any associated watches.
+		 */
+		spin_lock(&inode->i_lock);
+		if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) {
+			spin_unlock(&inode->i_lock);
+			continue;
+		}
+
+		/*
+		 * If i_count is zero, the inode cannot have any watches and
+		 * doing an __iget/iput with MS_ACTIVE clear would actually
+		 * evict all inodes with zero i_count from icache which is
+		 * unnecessarily violent and may in fact be illegal to do.
+		 */
+		if (!atomic_read(&inode->i_count)) {
+			spin_unlock(&inode->i_lock);
+			continue;
+		}
+
+		need_iput_tmp = need_iput;
+		need_iput = NULL;
+
+		/* In case fsnotify_inode_delete() drops a reference. */
+		if (inode != need_iput_tmp)
+			__iget(inode);
+		else
+			need_iput_tmp = NULL;
+		spin_unlock(&inode->i_lock);
+
+		/* In case the dropping of a reference would nuke next_i. */
+		if ((&next_i->i_sb_list != list) &&
+		    atomic_read(&next_i->i_count)) {
+			spin_lock(&next_i->i_lock);
+			if (!(next_i->i_state & (I_FREEING | I_WILL_FREE))) {
+				__iget(next_i);
+				need_iput = next_i;
+			}
+			spin_unlock(&next_i->i_lock);
+		}
+
+		/*
+		 * We can safely drop inode_sb_list_lock here because we hold
+		 * references on both inode and next_i.  Also no new inodes
+		 * will be added since the umount has begun.
+		 */
+		spin_unlock(&inode_sb_list_lock);
+
+		if (need_iput_tmp)
+			iput(need_iput_tmp);
+
+		/* for each watch, send FS_UNMOUNT and then remove it */
+		fsnotify(inode, FS_UNMOUNT, inode, FSNOTIFY_EVENT_INODE, NULL, 0);
+
+		fsnotify_inode_delete(inode);
+
+		iput(inode);
+
+		spin_lock(&inode_sb_list_lock);
+	}
+	spin_unlock(&inode_sb_list_lock);
+}

fs/notify/inotify/Kconfig

@@ -0,0 +1,17 @@
+config INOTIFY_USER
+	bool "Inotify support for userspace"
+	select ANON_INODES
+	select FSNOTIFY
+	default y
+	---help---
+	  Say Y here to enable inotify support for userspace, including the
+	  associated system calls.  Inotify allows monitoring of both files and
+	  directories via a single open fd.  Events are read from the file
+	  descriptor, which is also select()- and poll()-able.
+	  Inotify fixes numerous shortcomings in dnotify and introduces several
+	  new features including multiple file events, one-shot support, and
+	  unmount notification.
+
+	  For more information, see <file:Documentation/filesystems/inotify.txt>
+
+	  If unsure, say Y.

fs/notify/inotify/Makefile

@@ -0,0 +1 @@
+obj-$(CONFIG_INOTIFY_USER)	+= inotify_fsnotify.o inotify_user.o

fs/notify/inotify/inotify.h

@@ -0,0 +1,21 @@
+#include <linux/fsnotify_backend.h>
+#include <linux/inotify.h>
+#include <linux/slab.h> /* struct kmem_cache */
+
+extern struct kmem_cache *event_priv_cachep;
+
+struct inotify_event_private_data {
+	struct fsnotify_event_private_data fsnotify_event_priv_data;
+	int wd;
+};
+
+struct inotify_inode_mark {
+	struct fsnotify_mark fsn_mark;
+	int wd;
+};
+
+extern void inotify_ignored_and_remove_idr(struct fsnotify_mark *fsn_mark,
+					   struct fsnotify_group *group);
+extern void inotify_free_event_priv(struct fsnotify_event_private_data *event_priv);
+
+extern const struct fsnotify_ops inotify_fsnotify_ops;

fs/notify/inotify/inotify_fsnotify.c

@@ -0,0 +1,223 @@
+/*
+ * fs/inotify_user.c - inotify support for userspace
+ *
+ * Authors:
+ *	John McCutchan	<ttb@tentacle.dhs.org>
+ *	Robert Love	<rml@novell.com>
+ *
+ * Copyright (C) 2005 John McCutchan
+ * Copyright 2006 Hewlett-Packard Development Company, L.P.
+ *
+ * Copyright (C) 2009 Eric Paris <Red Hat Inc>
+ * inotify was largely rewriten to make use of the fsnotify infrastructure
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/dcache.h> /* d_unlinked */
+#include <linux/fs.h> /* struct inode */
+#include <linux/fsnotify_backend.h>
+#include <linux/inotify.h>
+#include <linux/path.h> /* struct path */
+#include <linux/slab.h> /* kmem_* */
+#include <linux/types.h>
+#include <linux/sched.h>
+
+#include "inotify.h"
+
+/*
+ * Check if 2 events contain the same information.  We do not compare private data
+ * but at this moment that isn't a problem for any know fsnotify listeners.
+ */
+static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new)
+{
+	if ((old->mask == new->mask) &&
+	    (old->to_tell == new->to_tell) &&
+	    (old->data_type == new->data_type) &&
+	    (old->name_len == new->name_len)) {
+		switch (old->data_type) {
+		case (FSNOTIFY_EVENT_INODE):
+			/* remember, after old was put on the wait_q we aren't
+			 * allowed to look at the inode any more, only thing
+			 * left to check was if the file_name is the same */
+			if (!old->name_len ||
+			    !strcmp(old->file_name, new->file_name))
+				return true;
+			break;
+		case (FSNOTIFY_EVENT_PATH):
+			if ((old->path.mnt == new->path.mnt) &&
+			    (old->path.dentry == new->path.dentry))
+				return true;
+			break;
+		case (FSNOTIFY_EVENT_NONE):
+			if (old->mask & FS_Q_OVERFLOW)
+				return true;
+			else if (old->mask & FS_IN_IGNORED)
+				return false;
+			return true;
+		};
+	}
+	return false;
+}
+
+static struct fsnotify_event *inotify_merge(struct list_head *list,
+					    struct fsnotify_event *event)
+{
+	struct fsnotify_event_holder *last_holder;
+	struct fsnotify_event *last_event;
+
+	/* and the list better be locked by something too */
+	spin_lock(&event->lock);
+
+	last_holder = list_entry(list->prev, struct fsnotify_event_holder, event_list);
+	last_event = last_holder->event;
+	if (event_compare(last_event, event))
+		fsnotify_get_event(last_event);
+	else
+		last_event = NULL;
+
+	spin_unlock(&event->lock);
+
+	return last_event;
+}
+
+static int inotify_handle_event(struct fsnotify_group *group,
+				struct fsnotify_mark *inode_mark,
+				struct fsnotify_mark *vfsmount_mark,
+				struct fsnotify_event *event)
+{
+	struct inotify_inode_mark *i_mark;
+	struct inode *to_tell;
+	struct inotify_event_private_data *event_priv;
+	struct fsnotify_event_private_data *fsn_event_priv;
+	struct fsnotify_event *added_event;
+	int wd, ret = 0;
+
+	BUG_ON(vfsmount_mark);
+
+	pr_debug("%s: group=%p event=%p to_tell=%p mask=%x\n", __func__, group,
+		 event, event->to_tell, event->mask);
+
+	to_tell = event->to_tell;
+
+	i_mark = container_of(inode_mark, struct inotify_inode_mark,
+			      fsn_mark);
+	wd = i_mark->wd;
+
+	event_priv = kmem_cache_alloc(event_priv_cachep, GFP_KERNEL);
+	if (unlikely(!event_priv))
+		return -ENOMEM;
+
+	fsn_event_priv = &event_priv->fsnotify_event_priv_data;
+
+	fsnotify_get_group(group);
+	fsn_event_priv->group = group;
+	event_priv->wd = wd;
+
+	added_event = fsnotify_add_notify_event(group, event, fsn_event_priv, inotify_merge);
+	if (added_event) {
+		inotify_free_event_priv(fsn_event_priv);
+		if (!IS_ERR(added_event))
+			fsnotify_put_event(added_event);
+		else
+			ret = PTR_ERR(added_event);
+	}
+
+	if (inode_mark->mask & IN_ONESHOT)
+		fsnotify_destroy_mark(inode_mark, group);
+
+	return ret;
+}
+
+static void inotify_freeing_mark(struct fsnotify_mark *fsn_mark, struct fsnotify_group *group)
+{
+	inotify_ignored_and_remove_idr(fsn_mark, group);
+}
+
+static bool inotify_should_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)
+{
+	if ((inode_mark->mask & FS_EXCL_UNLINK) &&
+	    (data_type == FSNOTIFY_EVENT_PATH)) {
+		struct path *path = data;
+
+		if (d_unlinked(path->dentry))
+			return false;
+	}
+
+	return true;
+}
+
+/*
+ * This is NEVER supposed to be called.  Inotify marks should either have been
+ * removed from the idr when the watch was removed or in the
+ * fsnotify_destroy_mark_by_group() call when the inotify instance was being
+ * torn down.  This is only called if the idr is about to be freed but there
+ * are still marks in it.
+ */
+static int idr_callback(int id, void *p, void *data)
+{
+	struct fsnotify_mark *fsn_mark;
+	struct inotify_inode_mark *i_mark;
+	static bool warned = false;
+
+	if (warned)
+		return 0;
+
+	warned = true;
+	fsn_mark = p;
+	i_mark = container_of(fsn_mark, struct inotify_inode_mark, fsn_mark);
+
+	WARN(1, "inotify closing but id=%d for fsn_mark=%p in group=%p still in "
+		"idr.  Probably leaking memory\n", id, p, data);
+
+	/*
+	 * I'm taking the liberty of assuming that the mark in question is a
+	 * valid address and I'm dereferencing it.  This might help to figure
+	 * out why we got here and the panic is no worse than the original
+	 * BUG() that was here.
+	 */
+	if (fsn_mark)
+		printk(KERN_WARNING "fsn_mark->group=%p inode=%p wd=%d\n",
+			fsn_mark->group, fsn_mark->i.inode, i_mark->wd);
+	return 0;
+}
+
+static void inotify_free_group_priv(struct fsnotify_group *group)
+{
+	/* ideally the idr is empty and we won't hit the BUG in the callback */
+	idr_for_each(&group->inotify_data.idr, idr_callback, group);
+	idr_destroy(&group->inotify_data.idr);
+	atomic_dec(&group->inotify_data.user->inotify_devs);
+	free_uid(group->inotify_data.user);
+}
+
+void inotify_free_event_priv(struct fsnotify_event_private_data *fsn_event_priv)
+{
+	struct inotify_event_private_data *event_priv;
+
+
+	event_priv = container_of(fsn_event_priv, struct inotify_event_private_data,
+				  fsnotify_event_priv_data);
+
+	fsnotify_put_group(fsn_event_priv->group);
+	kmem_cache_free(event_priv_cachep, event_priv);
+}
+
+const struct fsnotify_ops inotify_fsnotify_ops = {
+	.handle_event = inotify_handle_event,
+	.should_send_event = inotify_should_send_event,
+	.free_group_priv = inotify_free_group_priv,
+	.free_event_priv = inotify_free_event_priv,
+	.freeing_mark = inotify_freeing_mark,
+};

fs/notify/inotify/inotify_user.c

@@ -0,0 +1,852 @@
+/*
+ * fs/inotify_user.c - inotify support for userspace
+ *
+ * Authors:
+ *	John McCutchan	<ttb@tentacle.dhs.org>
+ *	Robert Love	<rml@novell.com>
+ *
+ * Copyright (C) 2005 John McCutchan
+ * Copyright 2006 Hewlett-Packard Development Company, L.P.
+ *
+ * Copyright (C) 2009 Eric Paris <Red Hat Inc>
+ * inotify was largely rewriten to make use of the fsnotify infrastructure
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/file.h>
+#include <linux/fs.h> /* struct inode */
+#include <linux/fsnotify_backend.h>
+#include <linux/idr.h>
+#include <linux/init.h> /* module_init */
+#include <linux/inotify.h>
+#include <linux/kernel.h> /* roundup() */
+#include <linux/namei.h> /* LOOKUP_FOLLOW */
+#include <linux/sched.h> /* struct user */
+#include <linux/slab.h> /* struct kmem_cache */
+#include <linux/syscalls.h>
+#include <linux/types.h>
+#include <linux/anon_inodes.h>
+#include <linux/uaccess.h>
+#include <linux/poll.h>
+#include <linux/wait.h>
+
+#include "inotify.h"
+#include "../fdinfo.h"
+
+#include <asm/ioctls.h>
+
+/* these are configurable via /proc/sys/fs/inotify/ */
+static int inotify_max_user_instances __read_mostly;
+static int inotify_max_queued_events __read_mostly;
+static int inotify_max_user_watches __read_mostly;
+
+static struct kmem_cache *inotify_inode_mark_cachep __read_mostly;
+struct kmem_cache *event_priv_cachep __read_mostly;
+
+#ifdef CONFIG_SYSCTL
+
+#include <linux/sysctl.h>
+
+static int zero;
+
+ctl_table inotify_table[] = {
+	{
+		.procname	= "max_user_instances",
+		.data		= &inotify_max_user_instances,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &zero,
+	},
+	{
+		.procname	= "max_user_watches",
+		.data		= &inotify_max_user_watches,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &zero,
+	},
+	{
+		.procname	= "max_queued_events",
+		.data		= &inotify_max_queued_events,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &zero
+	},
+	{ }
+};
+#endif /* CONFIG_SYSCTL */
+
+static inline __u32 inotify_arg_to_mask(u32 arg)
+{
+	__u32 mask;
+
+	/*
+	 * everything should accept their own ignored, cares about children,
+	 * and should receive events when the inode is unmounted
+	 */
+	mask = (FS_IN_IGNORED | FS_EVENT_ON_CHILD | FS_UNMOUNT);
+
+	/* mask off the flags used to open the fd */
+	mask |= (arg & (IN_ALL_EVENTS | IN_ONESHOT | IN_EXCL_UNLINK));
+
+	return mask;
+}
+
+static inline u32 inotify_mask_to_arg(__u32 mask)
+{
+	return mask & (IN_ALL_EVENTS | IN_ISDIR | IN_UNMOUNT | IN_IGNORED |
+		       IN_Q_OVERFLOW);
+}
+
+/* intofiy userspace file descriptor functions */
+static unsigned int inotify_poll(struct file *file, poll_table *wait)
+{
+	struct fsnotify_group *group = file->private_data;
+	int ret = 0;
+
+	poll_wait(file, &group->notification_waitq, wait);
+	mutex_lock(&group->notification_mutex);
+	if (!fsnotify_notify_queue_is_empty(group))
+		ret = POLLIN | POLLRDNORM;
+	mutex_unlock(&group->notification_mutex);
+
+	return ret;
+}
+
+/*
+ * Get an inotify_kernel_event if one exists and is small
+ * enough to fit in "count". Return an error pointer if
+ * not large enough.
+ *
+ * Called with the group->notification_mutex held.
+ */
+static struct fsnotify_event *get_one_event(struct fsnotify_group *group,
+					    size_t count)
+{
+	size_t event_size = sizeof(struct inotify_event);
+	struct fsnotify_event *event;
+
+	if (fsnotify_notify_queue_is_empty(group))
+		return NULL;
+
+	event = fsnotify_peek_notify_event(group);
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	if (event->name_len)
+		event_size += roundup(event->name_len + 1, event_size);
+
+	if (event_size > count)
+		return ERR_PTR(-EINVAL);
+
+	/* held the notification_mutex the whole time, so this is the
+	 * same event we peeked above */
+	fsnotify_remove_notify_event(group);
+
+	return event;
+}
+
+/*
+ * Copy an event to user space, returning how much we copied.
+ *
+ * We already checked that the event size is smaller than the
+ * buffer we had in "get_one_event()" above.
+ */
+static ssize_t copy_event_to_user(struct fsnotify_group *group,
+				  struct fsnotify_event *event,
+				  char __user *buf)
+{
+	struct inotify_event inotify_event;
+	struct fsnotify_event_private_data *fsn_priv;
+	struct inotify_event_private_data *priv;
+	size_t event_size = sizeof(struct inotify_event);
+	size_t name_len = 0;
+
+	pr_debug("%s: group=%p event=%p\n", __func__, group, event);
+
+	/* we get the inotify watch descriptor from the event private data */
+	spin_lock(&event->lock);
+	fsn_priv = fsnotify_remove_priv_from_event(group, event);
+	spin_unlock(&event->lock);
+
+	if (!fsn_priv)
+		inotify_event.wd = -1;
+	else {
+		priv = container_of(fsn_priv, struct inotify_event_private_data,
+				    fsnotify_event_priv_data);
+		inotify_event.wd = priv->wd;
+		inotify_free_event_priv(fsn_priv);
+	}
+
+	/*
+	 * round up event->name_len so it is a multiple of event_size
+	 * plus an extra byte for the terminating '\0'.
+	 */
+	if (event->name_len)
+		name_len = roundup(event->name_len + 1, event_size);
+	inotify_event.len = name_len;
+
+	inotify_event.mask = inotify_mask_to_arg(event->mask);
+	inotify_event.cookie = event->sync_cookie;
+
+	/* send the main event */
+	if (copy_to_user(buf, &inotify_event, event_size))
+		return -EFAULT;
+
+	buf += event_size;
+
+	/*
+	 * fsnotify only stores the pathname, so here we have to send the pathname
+	 * and then pad that pathname out to a multiple of sizeof(inotify_event)
+	 * with zeros.  I get my zeros from the nul_inotify_event.
+	 */
+	if (name_len) {
+		unsigned int len_to_zero = name_len - event->name_len;
+		/* copy the path name */
+		if (copy_to_user(buf, event->file_name, event->name_len))
+			return -EFAULT;
+		buf += event->name_len;
+
+		/* fill userspace with 0's */
+		if (clear_user(buf, len_to_zero))
+			return -EFAULT;
+		buf += len_to_zero;
+		event_size += name_len;
+	}
+
+	return event_size;
+}
+
+static ssize_t inotify_read(struct file *file, char __user *buf,
+			    size_t count, loff_t *pos)
+{
+	struct fsnotify_group *group;
+	struct fsnotify_event *kevent;
+	char __user *start;
+	int ret;
+	DEFINE_WAIT(wait);
+
+	start = buf;
+	group = file->private_data;
+
+	while (1) {
+		prepare_to_wait(&group->notification_waitq, &wait, TASK_INTERRUPTIBLE);
+
+		mutex_lock(&group->notification_mutex);
+		kevent = get_one_event(group, count);
+		mutex_unlock(&group->notification_mutex);
+
+		pr_debug("%s: group=%p kevent=%p\n", __func__, group, kevent);
+
+		if (kevent) {
+			ret = PTR_ERR(kevent);
+			if (IS_ERR(kevent))
+				break;
+			ret = copy_event_to_user(group, kevent, buf);
+			fsnotify_put_event(kevent);
+			if (ret < 0)
+				break;
+			buf += ret;
+			count -= ret;
+			continue;
+		}
+
+		ret = -EAGAIN;
+		if (file->f_flags & O_NONBLOCK)
+			break;
+		ret = -ERESTARTSYS;
+		if (signal_pending(current))
+			break;
+
+		if (start != buf)
+			break;
+
+		schedule();
+	}
+
+	finish_wait(&group->notification_waitq, &wait);
+	if (start != buf && ret != -EFAULT)
+		ret = buf - start;
+	return ret;
+}
+
+static int inotify_release(struct inode *ignored, struct file *file)
+{
+	struct fsnotify_group *group = file->private_data;
+
+	pr_debug("%s: group=%p\n", __func__, group);
+
+	/* free this group, matching get was inotify_init->fsnotify_obtain_group */
+	fsnotify_destroy_group(group);
+
+	return 0;
+}
+
+static long inotify_ioctl(struct file *file, unsigned int cmd,
+			  unsigned long arg)
+{
+	struct fsnotify_group *group;
+	struct fsnotify_event_holder *holder;
+	struct fsnotify_event *event;
+	void __user *p;
+	int ret = -ENOTTY;
+	size_t send_len = 0;
+
+	group = file->private_data;
+	p = (void __user *) arg;
+
+	pr_debug("%s: group=%p cmd=%u\n", __func__, group, cmd);
+
+	switch (cmd) {
+	case FIONREAD:
+		mutex_lock(&group->notification_mutex);
+		list_for_each_entry(holder, &group->notification_list, event_list) {
+			event = holder->event;
+			send_len += sizeof(struct inotify_event);
+			if (event->name_len)
+				send_len += roundup(event->name_len + 1,
+						sizeof(struct inotify_event));
+		}
+		mutex_unlock(&group->notification_mutex);
+		ret = put_user(send_len, (int __user *) p);
+		break;
+	}
+
+	return ret;
+}
+
+static const struct file_operations inotify_fops = {
+	.show_fdinfo	= inotify_show_fdinfo,
+	.poll		= inotify_poll,
+	.read		= inotify_read,
+	.fasync		= fsnotify_fasync,
+	.release	= inotify_release,
+	.unlocked_ioctl	= inotify_ioctl,
+	.compat_ioctl	= inotify_ioctl,
+	.llseek		= noop_llseek,
+};
+
+
+/*
+ * find_inode - resolve a user-given path to a specific inode
+ */
+static int inotify_find_inode(const char __user *dirname, struct path *path, unsigned flags)
+{
+	int error;
+
+	error = user_path_at(AT_FDCWD, dirname, flags, path);
+	if (error)
+		return error;
+	/* you can only watch an inode if you have read permissions on it */
+	error = inode_permission(path->dentry->d_inode, MAY_READ);
+	if (error)
+		path_put(path);
+	return error;
+}
+
+static int inotify_add_to_idr(struct idr *idr, spinlock_t *idr_lock,
+			      struct inotify_inode_mark *i_mark)
+{
+	int ret;
+
+	idr_preload(GFP_KERNEL);
+	spin_lock(idr_lock);
+
+	ret = idr_alloc_cyclic(idr, i_mark, 1, 0, GFP_NOWAIT);
+	if (ret >= 0) {
+		/* we added the mark to the idr, take a reference */
+		i_mark->wd = ret;
+		fsnotify_get_mark(&i_mark->fsn_mark);
+	}
+
+	spin_unlock(idr_lock);
+	idr_preload_end();
+	return ret < 0 ? ret : 0;
+}
+
+static struct inotify_inode_mark *inotify_idr_find_locked(struct fsnotify_group *group,
+								int wd)
+{
+	struct idr *idr = &group->inotify_data.idr;
+	spinlock_t *idr_lock = &group->inotify_data.idr_lock;
+	struct inotify_inode_mark *i_mark;
+
+	assert_spin_locked(idr_lock);
+
+	i_mark = idr_find(idr, wd);
+	if (i_mark) {
+		struct fsnotify_mark *fsn_mark = &i_mark->fsn_mark;
+
+		fsnotify_get_mark(fsn_mark);
+		/* One ref for being in the idr, one ref we just took */
+		BUG_ON(atomic_read(&fsn_mark->refcnt) < 2);
+	}
+
+	return i_mark;
+}
+
+static struct inotify_inode_mark *inotify_idr_find(struct fsnotify_group *group,
+							 int wd)
+{
+	struct inotify_inode_mark *i_mark;
+	spinlock_t *idr_lock = &group->inotify_data.idr_lock;
+
+	spin_lock(idr_lock);
+	i_mark = inotify_idr_find_locked(group, wd);
+	spin_unlock(idr_lock);
+
+	return i_mark;
+}
+
+static void do_inotify_remove_from_idr(struct fsnotify_group *group,
+				       struct inotify_inode_mark *i_mark)
+{
+	struct idr *idr = &group->inotify_data.idr;
+	spinlock_t *idr_lock = &group->inotify_data.idr_lock;
+	int wd = i_mark->wd;
+
+	assert_spin_locked(idr_lock);
+
+	idr_remove(idr, wd);
+
+	/* removed from the idr, drop that ref */
+	fsnotify_put_mark(&i_mark->fsn_mark);
+}
+
+/*
+ * Remove the mark from the idr (if present) and drop the reference
+ * on the mark because it was in the idr.
+ */
+static void inotify_remove_from_idr(struct fsnotify_group *group,
+				    struct inotify_inode_mark *i_mark)
+{
+	spinlock_t *idr_lock = &group->inotify_data.idr_lock;
+	struct inotify_inode_mark *found_i_mark = NULL;
+	int wd;
+
+	spin_lock(idr_lock);
+	wd = i_mark->wd;
+
+	/*
+	 * does this i_mark think it is in the idr?  we shouldn't get called
+	 * if it wasn't....
+	 */
+	if (wd == -1) {
+		WARN_ONCE(1, "%s: i_mark=%p i_mark->wd=%d i_mark->group=%p"
+			" i_mark->inode=%p\n", __func__, i_mark, i_mark->wd,
+			i_mark->fsn_mark.group, i_mark->fsn_mark.i.inode);
+		goto out;
+	}
+
+	/* Lets look in the idr to see if we find it */
+	found_i_mark = inotify_idr_find_locked(group, wd);
+	if (unlikely(!found_i_mark)) {
+		WARN_ONCE(1, "%s: i_mark=%p i_mark->wd=%d i_mark->group=%p"
+			" i_mark->inode=%p\n", __func__, i_mark, i_mark->wd,
+			i_mark->fsn_mark.group, i_mark->fsn_mark.i.inode);
+		goto out;
+	}
+
+	/*
+	 * We found an mark in the idr at the right wd, but it's
+	 * not the mark we were told to remove.  eparis seriously
+	 * fucked up somewhere.
+	 */
+	if (unlikely(found_i_mark != i_mark)) {
+		WARN_ONCE(1, "%s: i_mark=%p i_mark->wd=%d i_mark->group=%p "
+			"mark->inode=%p found_i_mark=%p found_i_mark->wd=%d "
+			"found_i_mark->group=%p found_i_mark->inode=%p\n",
+			__func__, i_mark, i_mark->wd, i_mark->fsn_mark.group,
+			i_mark->fsn_mark.i.inode, found_i_mark, found_i_mark->wd,
+			found_i_mark->fsn_mark.group,
+			found_i_mark->fsn_mark.i.inode);
+		goto out;
+	}
+
+	/*
+	 * One ref for being in the idr
+	 * one ref held by the caller trying to kill us
+	 * one ref grabbed by inotify_idr_find
+	 */
+	if (unlikely(atomic_read(&i_mark->fsn_mark.refcnt) < 3)) {
+		printk(KERN_ERR "%s: i_mark=%p i_mark->wd=%d i_mark->group=%p"
+			" i_mark->inode=%p\n", __func__, i_mark, i_mark->wd,
+			i_mark->fsn_mark.group, i_mark->fsn_mark.i.inode);
+		/* we can't really recover with bad ref cnting.. */
+		BUG();
+	}
+
+	do_inotify_remove_from_idr(group, i_mark);
+out:
+	/* match the ref taken by inotify_idr_find_locked() */
+	if (found_i_mark)
+		fsnotify_put_mark(&found_i_mark->fsn_mark);
+	i_mark->wd = -1;
+	spin_unlock(idr_lock);
+}
+
+/*
+ * Send IN_IGNORED for this wd, remove this wd from the idr.
+ */
+void inotify_ignored_and_remove_idr(struct fsnotify_mark *fsn_mark,
+				    struct fsnotify_group *group)
+{
+	struct inotify_inode_mark *i_mark;
+	struct fsnotify_event *ignored_event, *notify_event;
+	struct inotify_event_private_data *event_priv;
+	struct fsnotify_event_private_data *fsn_event_priv;
+	int ret;
+
+	i_mark = container_of(fsn_mark, struct inotify_inode_mark, fsn_mark);
+
+	ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL,
+					      FSNOTIFY_EVENT_NONE, NULL, 0,
+					      GFP_NOFS);
+	if (!ignored_event)
+		goto skip_send_ignore;
+
+	event_priv = kmem_cache_alloc(event_priv_cachep, GFP_NOFS);
+	if (unlikely(!event_priv))
+		goto skip_send_ignore;
+
+	fsn_event_priv = &event_priv->fsnotify_event_priv_data;
+
+	fsnotify_get_group(group);
+	fsn_event_priv->group = group;
+	event_priv->wd = i_mark->wd;
+
+	notify_event = fsnotify_add_notify_event(group, ignored_event, fsn_event_priv, NULL);
+	if (notify_event) {
+		if (IS_ERR(notify_event))
+			ret = PTR_ERR(notify_event);
+		else
+			fsnotify_put_event(notify_event);
+		inotify_free_event_priv(fsn_event_priv);
+	}
+
+skip_send_ignore:
+	/* matches the reference taken when the event was created */
+	if (ignored_event)
+		fsnotify_put_event(ignored_event);
+
+	/* remove this mark from the idr */
+	inotify_remove_from_idr(group, i_mark);
+
+	atomic_dec(&group->inotify_data.user->inotify_watches);
+}
+
+/* ding dong the mark is dead */
+static void inotify_free_mark(struct fsnotify_mark *fsn_mark)
+{
+	struct inotify_inode_mark *i_mark;
+
+	i_mark = container_of(fsn_mark, struct inotify_inode_mark, fsn_mark);
+
+	kmem_cache_free(inotify_inode_mark_cachep, i_mark);
+}
+
+static int inotify_update_existing_watch(struct fsnotify_group *group,
+					 struct inode *inode,
+					 u32 arg)
+{
+	struct fsnotify_mark *fsn_mark;
+	struct inotify_inode_mark *i_mark;
+	__u32 old_mask, new_mask;
+	__u32 mask;
+	int add = (arg & IN_MASK_ADD);
+	int ret;
+
+	mask = inotify_arg_to_mask(arg);
+
+	fsn_mark = fsnotify_find_inode_mark(group, inode);
+	if (!fsn_mark)
+		return -ENOENT;
+
+	i_mark = container_of(fsn_mark, struct inotify_inode_mark, fsn_mark);
+
+	spin_lock(&fsn_mark->lock);
+
+	old_mask = fsn_mark->mask;
+	if (add)
+		fsnotify_set_mark_mask_locked(fsn_mark, (fsn_mark->mask | mask));
+	else
+		fsnotify_set_mark_mask_locked(fsn_mark, mask);
+	new_mask = fsn_mark->mask;
+
+	spin_unlock(&fsn_mark->lock);
+
+	if (old_mask != new_mask) {
+		/* more bits in old than in new? */
+		int dropped = (old_mask & ~new_mask);
+		/* more bits in this fsn_mark than the inode's mask? */
+		int do_inode = (new_mask & ~inode->i_fsnotify_mask);
+
+		/* update the inode with this new fsn_mark */
+		if (dropped || do_inode)
+			fsnotify_recalc_inode_mask(inode);
+
+	}
+
+	/* return the wd */
+	ret = i_mark->wd;
+
+	/* match the get from fsnotify_find_mark() */
+	fsnotify_put_mark(fsn_mark);
+
+	return ret;
+}
+
+static int inotify_new_watch(struct fsnotify_group *group,
+			     struct inode *inode,
+			     u32 arg)
+{
+	struct inotify_inode_mark *tmp_i_mark;
+	__u32 mask;
+	int ret;
+	struct idr *idr = &group->inotify_data.idr;
+	spinlock_t *idr_lock = &group->inotify_data.idr_lock;
+
+	mask = inotify_arg_to_mask(arg);
+
+	tmp_i_mark = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL);
+	if (unlikely(!tmp_i_mark))
+		return -ENOMEM;
+
+	fsnotify_init_mark(&tmp_i_mark->fsn_mark, inotify_free_mark);
+	tmp_i_mark->fsn_mark.mask = mask;
+	tmp_i_mark->wd = -1;
+
+	ret = -ENOSPC;
+	if (atomic_read(&group->inotify_data.user->inotify_watches) >= inotify_max_user_watches)
+		goto out_err;
+
+	ret = inotify_add_to_idr(idr, idr_lock, tmp_i_mark);
+	if (ret)
+		goto out_err;
+
+	/* we are on the idr, now get on the inode */
+	ret = fsnotify_add_mark(&tmp_i_mark->fsn_mark, group, inode, NULL, 0);
+	if (ret) {
+		/* we failed to get on the inode, get off the idr */
+		inotify_remove_from_idr(group, tmp_i_mark);
+		goto out_err;
+	}
+
+	/* increment the number of watches the user has */
+	atomic_inc(&group->inotify_data.user->inotify_watches);
+
+	/* return the watch descriptor for this new mark */
+	ret = tmp_i_mark->wd;
+
+out_err:
+	/* match the ref from fsnotify_init_mark() */
+	fsnotify_put_mark(&tmp_i_mark->fsn_mark);
+
+	return ret;
+}
+
+static int inotify_update_watch(struct fsnotify_group *group, struct inode *inode, u32 arg)
+{
+	int ret = 0;
+
+retry:
+	/* try to update and existing watch with the new arg */
+	ret = inotify_update_existing_watch(group, inode, arg);
+	/* no mark present, try to add a new one */
+	if (ret == -ENOENT)
+		ret = inotify_new_watch(group, inode, arg);
+	/*
+	 * inotify_new_watch could race with another thread which did an
+	 * inotify_new_watch between the update_existing and the add watch
+	 * here, go back and try to update an existing mark again.
+	 */
+	if (ret == -EEXIST)
+		goto retry;
+
+	return ret;
+}
+
+static struct fsnotify_group *inotify_new_group(unsigned int max_events)
+{
+	struct fsnotify_group *group;
+
+	group = fsnotify_alloc_group(&inotify_fsnotify_ops);
+	if (IS_ERR(group))
+		return group;
+
+	group->max_events = max_events;
+
+	spin_lock_init(&group->inotify_data.idr_lock);
+	idr_init(&group->inotify_data.idr);
+	group->inotify_data.user = get_current_user();
+
+	if (atomic_inc_return(&group->inotify_data.user->inotify_devs) >
+	    inotify_max_user_instances) {
+		fsnotify_destroy_group(group);
+		return ERR_PTR(-EMFILE);
+	}
+
+	return group;
+}
+
+
+/* inotify syscalls */
+SYSCALL_DEFINE1(inotify_init1, int, flags)
+{
+	struct fsnotify_group *group;
+	int ret;
+
+	/* Check the IN_* constants for consistency.  */
+	BUILD_BUG_ON(IN_CLOEXEC != O_CLOEXEC);
+	BUILD_BUG_ON(IN_NONBLOCK != O_NONBLOCK);
+
+	if (flags & ~(IN_CLOEXEC | IN_NONBLOCK))
+		return -EINVAL;
+
+	/* fsnotify_obtain_group took a reference to group, we put this when we kill the file in the end */
+	group = inotify_new_group(inotify_max_queued_events);
+	if (IS_ERR(group))
+		return PTR_ERR(group);
+
+	ret = anon_inode_getfd("inotify", &inotify_fops, group,
+				  O_RDONLY | flags);
+	if (ret < 0)
+		fsnotify_destroy_group(group);
+
+	return ret;
+}
+
+SYSCALL_DEFINE0(inotify_init)
+{
+	return sys_inotify_init1(0);
+}
+
+SYSCALL_DEFINE3(inotify_add_watch, int, fd, const char __user *, pathname,
+		u32, mask)
+{
+	struct fsnotify_group *group;
+	struct inode *inode;
+	struct path path;
+	struct fd f;
+	int ret;
+	unsigned flags = 0;
+
+	/* don't allow invalid bits: we don't want flags set */
+	if (unlikely(!(mask & ALL_INOTIFY_BITS)))
+		return -EINVAL;
+
+	f = fdget(fd);
+	if (unlikely(!f.file))
+		return -EBADF;
+
+	/* verify that this is indeed an inotify instance */
+	if (unlikely(f.file->f_op != &inotify_fops)) {
+		ret = -EINVAL;
+		goto fput_and_out;
+	}
+
+	if (!(mask & IN_DONT_FOLLOW))
+		flags |= LOOKUP_FOLLOW;
+	if (mask & IN_ONLYDIR)
+		flags |= LOOKUP_DIRECTORY;
+
+	ret = inotify_find_inode(pathname, &path, flags);
+	if (ret)
+		goto fput_and_out;
+
+	/* inode held in place by reference to path; group by fget on fd */
+	inode = path.dentry->d_inode;
+	group = f.file->private_data;
+
+	/* create/update an inode mark */
+	ret = inotify_update_watch(group, inode, mask);
+	path_put(&path);
+fput_and_out:
+	fdput(f);
+	return ret;
+}
+
+SYSCALL_DEFINE2(inotify_rm_watch, int, fd, __s32, wd)
+{
+	struct fsnotify_group *group;
+	struct inotify_inode_mark *i_mark;
+	struct fd f;
+	int ret = 0;
+
+	f = fdget(fd);
+	if (unlikely(!f.file))
+		return -EBADF;
+
+	/* verify that this is indeed an inotify instance */
+	ret = -EINVAL;
+	if (unlikely(f.file->f_op != &inotify_fops))
+		goto out;
+
+	group = f.file->private_data;
+
+	ret = -EINVAL;
+	i_mark = inotify_idr_find(group, wd);
+	if (unlikely(!i_mark))
+		goto out;
+
+	ret = 0;
+
+	fsnotify_destroy_mark(&i_mark->fsn_mark, group);
+
+	/* match ref taken by inotify_idr_find */
+	fsnotify_put_mark(&i_mark->fsn_mark);
+
+out:
+	fdput(f);
+	return ret;
+}
+
+/*
+ * inotify_user_setup - Our initialization function.  Note that we cannot return
+ * error because we have compiled-in VFS hooks.  So an (unlikely) failure here
+ * must result in panic().
+ */
+static int __init inotify_user_setup(void)
+{
+	BUILD_BUG_ON(IN_ACCESS != FS_ACCESS);
+	BUILD_BUG_ON(IN_MODIFY != FS_MODIFY);
+	BUILD_BUG_ON(IN_ATTRIB != FS_ATTRIB);
+	BUILD_BUG_ON(IN_CLOSE_WRITE != FS_CLOSE_WRITE);
+	BUILD_BUG_ON(IN_CLOSE_NOWRITE != FS_CLOSE_NOWRITE);
+	BUILD_BUG_ON(IN_OPEN != FS_OPEN);
+	BUILD_BUG_ON(IN_MOVED_FROM != FS_MOVED_FROM);
+	BUILD_BUG_ON(IN_MOVED_TO != FS_MOVED_TO);
+	BUILD_BUG_ON(IN_CREATE != FS_CREATE);
+	BUILD_BUG_ON(IN_DELETE != FS_DELETE);
+	BUILD_BUG_ON(IN_DELETE_SELF != FS_DELETE_SELF);
+	BUILD_BUG_ON(IN_MOVE_SELF != FS_MOVE_SELF);
+	BUILD_BUG_ON(IN_UNMOUNT != FS_UNMOUNT);
+	BUILD_BUG_ON(IN_Q_OVERFLOW != FS_Q_OVERFLOW);
+	BUILD_BUG_ON(IN_IGNORED != FS_IN_IGNORED);
+	BUILD_BUG_ON(IN_EXCL_UNLINK != FS_EXCL_UNLINK);
+	BUILD_BUG_ON(IN_ISDIR != FS_ISDIR);
+	BUILD_BUG_ON(IN_ONESHOT != FS_IN_ONESHOT);
+
+	BUG_ON(hweight32(ALL_INOTIFY_BITS) != 21);
+
+	inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark, SLAB_PANIC);
+	event_priv_cachep = KMEM_CACHE(inotify_event_private_data, SLAB_PANIC);
+
+	inotify_max_queued_events = 16384;
+	inotify_max_user_instances = 128;
+	inotify_max_user_watches = 8192;
+
+	return 0;
+}
+module_init(inotify_user_setup);

fs/notify/mark.c

@@ -0,0 +1,381 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+/*
+ * fsnotify inode mark locking/lifetime/and refcnting
+ *
+ * REFCNT:
+ * The mark->refcnt tells how many "things" in the kernel currently are
+ * referencing this object.  The object typically will live inside the kernel
+ * with a refcnt of 2, one for each list it is on (i_list, g_list).  Any task
+ * which can find this object holding the appropriete locks, can take a reference
+ * and the object itself is guaranteed to survive until the reference is dropped.
+ *
+ * LOCKING:
+ * There are 3 spinlocks involved with fsnotify inode marks and they MUST
+ * be taken in order as follows:
+ *
+ * mark->lock
+ * group->mark_lock
+ * inode->i_lock
+ *
+ * mark->lock protects 2 things, mark->group and mark->inode.  You must hold
+ * that lock to dereference either of these things (they could be NULL even with
+ * the lock)
+ *
+ * group->mark_lock protects the marks_list anchored inside a given group
+ * and each mark is hooked via the g_list.  It also sorta protects the
+ * free_g_list, which when used is anchored by a private list on the stack of the
+ * task which held the group->mark_lock.
+ *
+ * inode->i_lock protects the i_fsnotify_marks list anchored inside a
+ * given inode and each mark is hooked via the i_list. (and sorta the
+ * free_i_list)
+ *
+ *
+ * LIFETIME:
+ * Inode marks survive between when they are added to an inode and when their
+ * refcnt==0.
+ *
+ * The inode mark can be cleared for a number of different reasons including:
+ * - The inode is unlinked for the last time.  (fsnotify_inode_remove)
+ * - The inode is being evicted from cache. (fsnotify_inode_delete)
+ * - The fs the inode is on is unmounted.  (fsnotify_inode_delete/fsnotify_unmount_inodes)
+ * - Something explicitly requests that it be removed.  (fsnotify_destroy_mark)
+ * - The fsnotify_group associated with the mark is going away and all such marks
+ *   need to be cleaned up. (fsnotify_clear_marks_by_group)
+ *
+ * Worst case we are given an inode and need to clean up all the marks on that
+ * inode.  We take i_lock and walk the i_fsnotify_marks safely.  For each
+ * mark on the list we take a reference (so the mark can't disappear under us).
+ * We remove that mark form the inode's list of marks and we add this mark to a
+ * private list anchored on the stack using i_free_list;  At this point we no
+ * longer fear anything finding the mark using the inode's list of marks.
+ *
+ * We can safely and locklessly run the private list on the stack of everything
+ * we just unattached from the original inode.  For each mark on the private list
+ * we grab the mark-> and can thus dereference mark->group and mark->inode.  If
+ * we see the group and inode are not NULL we take those locks.  Now holding all
+ * 3 locks we can completely remove the mark from other tasks finding it in the
+ * future.  Remember, 10 things might already be referencing this mark, but they
+ * better be holding a ref.  We drop our reference we took before we unhooked it
+ * from the inode.  When the ref hits 0 we can free the mark.
+ *
+ * Very similarly for freeing by group, except we use free_g_list.
+ *
+ * This has the very interesting property of being able to run concurrently with
+ * any (or all) other directions.
+ */
+
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/srcu.h>
+
+#include <linux/atomic.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+
+struct srcu_struct fsnotify_mark_srcu;
+static DEFINE_SPINLOCK(destroy_lock);
+static LIST_HEAD(destroy_list);
+static DECLARE_WAIT_QUEUE_HEAD(destroy_waitq);
+
+void fsnotify_get_mark(struct fsnotify_mark *mark)
+{
+	atomic_inc(&mark->refcnt);
+}
+
+void fsnotify_put_mark(struct fsnotify_mark *mark)
+{
+	if (atomic_dec_and_test(&mark->refcnt)) {
+		if (mark->group)
+			fsnotify_put_group(mark->group);
+		mark->free_mark(mark);
+	}
+}
+
+/*
+ * Any time a mark is getting freed we end up here.
+ * The caller had better be holding a reference to this mark so we don't actually
+ * do the final put under the mark->lock
+ */
+void fsnotify_destroy_mark_locked(struct fsnotify_mark *mark,
+				  struct fsnotify_group *group)
+{
+	struct inode *inode = NULL;
+
+	BUG_ON(!mutex_is_locked(&group->mark_mutex));
+
+	spin_lock(&mark->lock);
+
+	/* something else already called this function on this mark */
+	if (!(mark->flags & FSNOTIFY_MARK_FLAG_ALIVE)) {
+		spin_unlock(&mark->lock);
+		return;
+	}
+
+	mark->flags &= ~FSNOTIFY_MARK_FLAG_ALIVE;
+
+	if (mark->flags & FSNOTIFY_MARK_FLAG_INODE) {
+		inode = mark->i.inode;
+		fsnotify_destroy_inode_mark(mark);
+	} else if (mark->flags & FSNOTIFY_MARK_FLAG_VFSMOUNT)
+		fsnotify_destroy_vfsmount_mark(mark);
+	else
+		BUG();
+
+	list_del_init(&mark->g_list);
+
+	spin_unlock(&mark->lock);
+
+	if (inode && (mark->flags & FSNOTIFY_MARK_FLAG_OBJECT_PINNED))
+		iput(inode);
+	/* release lock temporarily */
+	mutex_unlock(&group->mark_mutex);
+
+	spin_lock(&destroy_lock);
+	list_add(&mark->destroy_list, &destroy_list);
+	spin_unlock(&destroy_lock);
+	wake_up(&destroy_waitq);
+	/*
+	 * We don't necessarily have a ref on mark from caller so the above destroy
+	 * may have actually freed it, unless this group provides a 'freeing_mark'
+	 * function which must be holding a reference.
+	 */
+
+	/*
+	 * Some groups like to know that marks are being freed.  This is a
+	 * callback to the group function to let it know that this mark
+	 * is being freed.
+	 */
+	if (group->ops->freeing_mark)
+		group->ops->freeing_mark(mark, group);
+
+	/*
+	 * __fsnotify_update_child_dentry_flags(inode);
+	 *
+	 * I really want to call that, but we can't, we have no idea if the inode
+	 * still exists the second we drop the mark->lock.
+	 *
+	 * The next time an event arrive to this inode from one of it's children
+	 * __fsnotify_parent will see that the inode doesn't care about it's
+	 * children and will update all of these flags then.  So really this
+	 * is just a lazy update (and could be a perf win...)
+	 */
+
+	atomic_dec(&group->num_marks);
+
+	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
+}
+
+void fsnotify_destroy_mark(struct fsnotify_mark *mark,
+			   struct fsnotify_group *group)
+{
+	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
+	fsnotify_destroy_mark_locked(mark, group);
+	mutex_unlock(&group->mark_mutex);
+}
+
+void fsnotify_set_mark_mask_locked(struct fsnotify_mark *mark, __u32 mask)
+{
+	assert_spin_locked(&mark->lock);
+
+	mark->mask = mask;
+
+	if (mark->flags & FSNOTIFY_MARK_FLAG_INODE)
+		fsnotify_set_inode_mark_mask_locked(mark, mask);
+}
+
+void fsnotify_set_mark_ignored_mask_locked(struct fsnotify_mark *mark, __u32 mask)
+{
+	assert_spin_locked(&mark->lock);
+
+	mark->ignored_mask = mask;
+}
+
+/*
+ * Attach an initialized mark to a given group and fs object.
+ * These marks may be used for the fsnotify backend to determine which
+ * event types should be delivered to which group.
+ */
+int fsnotify_add_mark_locked(struct fsnotify_mark *mark,
+			     struct fsnotify_group *group, struct inode *inode,
+			     struct vfsmount *mnt, int allow_dups)
+{
+	int ret = 0;
+
+	BUG_ON(inode && mnt);
+	BUG_ON(!inode && !mnt);
+	BUG_ON(!mutex_is_locked(&group->mark_mutex));
+
+	/*
+	 * LOCKING ORDER!!!!
+	 * group->mark_mutex
+	 * mark->lock
+	 * inode->i_lock
+	 */
+	spin_lock(&mark->lock);
+	mark->flags |= FSNOTIFY_MARK_FLAG_ALIVE;
+
+	fsnotify_get_group(group);
+	mark->group = group;
+	list_add(&mark->g_list, &group->marks_list);
+	atomic_inc(&group->num_marks);
+	fsnotify_get_mark(mark); /* for i_list and g_list */
+
+	if (inode) {
+		ret = fsnotify_add_inode_mark(mark, group, inode, allow_dups);
+		if (ret)
+			goto err;
+	} else if (mnt) {
+		ret = fsnotify_add_vfsmount_mark(mark, group, mnt, allow_dups);
+		if (ret)
+			goto err;
+	} else {
+		BUG();
+	}
+
+	/* this will pin the object if appropriate */
+	fsnotify_set_mark_mask_locked(mark, mark->mask);
+	spin_unlock(&mark->lock);
+
+	if (inode)
+		__fsnotify_update_child_dentry_flags(inode);
+
+	return ret;
+err:
+	mark->flags &= ~FSNOTIFY_MARK_FLAG_ALIVE;
+	list_del_init(&mark->g_list);
+	fsnotify_put_group(group);
+	mark->group = NULL;
+	atomic_dec(&group->num_marks);
+
+	spin_unlock(&mark->lock);
+
+	spin_lock(&destroy_lock);
+	list_add(&mark->destroy_list, &destroy_list);
+	spin_unlock(&destroy_lock);
+	wake_up(&destroy_waitq);
+
+	return ret;
+}
+
+int fsnotify_add_mark(struct fsnotify_mark *mark, struct fsnotify_group *group,
+		      struct inode *inode, struct vfsmount *mnt, int allow_dups)
+{
+	int ret;
+	mutex_lock(&group->mark_mutex);
+	ret = fsnotify_add_mark_locked(mark, group, inode, mnt, allow_dups);
+	mutex_unlock(&group->mark_mutex);
+	return ret;
+}
+
+/*
+ * clear any marks in a group in which mark->flags & flags is true
+ */
+void fsnotify_clear_marks_by_group_flags(struct fsnotify_group *group,
+					 unsigned int flags)
+{
+	struct fsnotify_mark *lmark, *mark;
+
+	mutex_lock_nested(&group->mark_mutex, SINGLE_DEPTH_NESTING);
+	list_for_each_entry_safe(mark, lmark, &group->marks_list, g_list) {
+		if (mark->flags & flags) {
+			fsnotify_get_mark(mark);
+			fsnotify_destroy_mark_locked(mark, group);
+			fsnotify_put_mark(mark);
+		}
+	}
+	mutex_unlock(&group->mark_mutex);
+}
+
+/*
+ * Given a group, destroy all of the marks associated with that group.
+ */
+void fsnotify_clear_marks_by_group(struct fsnotify_group *group)
+{
+	fsnotify_clear_marks_by_group_flags(group, (unsigned int)-1);
+}
+
+void fsnotify_duplicate_mark(struct fsnotify_mark *new, struct fsnotify_mark *old)
+{
+	assert_spin_locked(&old->lock);
+	new->i.inode = old->i.inode;
+	new->m.mnt = old->m.mnt;
+	if (old->group)
+		fsnotify_get_group(old->group);
+	new->group = old->group;
+	new->mask = old->mask;
+	new->free_mark = old->free_mark;
+}
+
+/*
+ * Nothing fancy, just initialize lists and locks and counters.
+ */
+void fsnotify_init_mark(struct fsnotify_mark *mark,
+			void (*free_mark)(struct fsnotify_mark *mark))
+{
+	memset(mark, 0, sizeof(*mark));
+	spin_lock_init(&mark->lock);
+	atomic_set(&mark->refcnt, 1);
+	mark->free_mark = free_mark;
+}
+
+static int fsnotify_mark_destroy(void *ignored)
+{
+	struct fsnotify_mark *mark, *next;
+	LIST_HEAD(private_destroy_list);
+
+	for (;;) {
+		spin_lock(&destroy_lock);
+		/* exchange the list head */
+		list_replace_init(&destroy_list, &private_destroy_list);
+		spin_unlock(&destroy_lock);
+
+		synchronize_srcu(&fsnotify_mark_srcu);
+
+		list_for_each_entry_safe(mark, next, &private_destroy_list, destroy_list) {
+			list_del_init(&mark->destroy_list);
+			fsnotify_put_mark(mark);
+		}
+
+		wait_event_interruptible(destroy_waitq, !list_empty(&destroy_list));
+	}
+
+	return 0;
+}
+
+static int __init fsnotify_mark_init(void)
+{
+	struct task_struct *thread;
+
+	thread = kthread_run(fsnotify_mark_destroy, NULL,
+			     "fsnotify_mark");
+	if (IS_ERR(thread))
+		panic("unable to start fsnotify mark destruction thread.");
+
+	return 0;
+}
+device_initcall(fsnotify_mark_init);

fs/notify/notification.c

@@ -0,0 +1,464 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+/*
+ * Basic idea behind the notification queue: An fsnotify group (like inotify)
+ * sends the userspace notification about events asynchronously some time after
+ * the event happened.  When inotify gets an event it will need to add that
+ * event to the group notify queue.  Since a single event might need to be on
+ * multiple group's notification queues we can't add the event directly to each
+ * queue and instead add a small "event_holder" to each queue.  This event_holder
+ * has a pointer back to the original event.  Since the majority of events are
+ * going to end up on one, and only one, notification queue we embed one
+ * event_holder into each event.  This means we have a single allocation instead
+ * of always needing two.  If the embedded event_holder is already in use by
+ * another group a new event_holder (from fsnotify_event_holder_cachep) will be
+ * allocated and used.
+ */
+
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/mount.h>
+#include <linux/mutex.h>
+#include <linux/namei.h>
+#include <linux/path.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+#include <linux/atomic.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+
+static struct kmem_cache *fsnotify_event_cachep;
+static struct kmem_cache *fsnotify_event_holder_cachep;
+/*
+ * This is a magic event we send when the q is too full.  Since it doesn't
+ * hold real event information we just keep one system wide and use it any time
+ * it is needed.  It's refcnt is set 1 at kernel init time and will never
+ * get set to 0 so it will never get 'freed'
+ */
+static struct fsnotify_event *q_overflow_event;
+static atomic_t fsnotify_sync_cookie = ATOMIC_INIT(0);
+
+/**
+ * fsnotify_get_cookie - return a unique cookie for use in synchronizing events.
+ * Called from fsnotify_move, which is inlined into filesystem modules.
+ */
+u32 fsnotify_get_cookie(void)
+{
+	return atomic_inc_return(&fsnotify_sync_cookie);
+}
+EXPORT_SYMBOL_GPL(fsnotify_get_cookie);
+
+/* return true if the notify queue is empty, false otherwise */
+bool fsnotify_notify_queue_is_empty(struct fsnotify_group *group)
+{
+	BUG_ON(!mutex_is_locked(&group->notification_mutex));
+	return list_empty(&group->notification_list) ? true : false;
+}
+
+void fsnotify_get_event(struct fsnotify_event *event)
+{
+	atomic_inc(&event->refcnt);
+}
+
+void fsnotify_put_event(struct fsnotify_event *event)
+{
+	if (!event)
+		return;
+
+	if (atomic_dec_and_test(&event->refcnt)) {
+		pr_debug("%s: event=%p\n", __func__, event);
+
+		if (event->data_type == FSNOTIFY_EVENT_PATH)
+			path_put(&event->path);
+
+		BUG_ON(!list_empty(&event->private_data_list));
+
+		kfree(event->file_name);
+		put_pid(event->tgid);
+		kmem_cache_free(fsnotify_event_cachep, event);
+	}
+}
+
+struct fsnotify_event_holder *fsnotify_alloc_event_holder(void)
+{
+	return kmem_cache_alloc(fsnotify_event_holder_cachep, GFP_KERNEL);
+}
+
+void fsnotify_destroy_event_holder(struct fsnotify_event_holder *holder)
+{
+	if (holder)
+		kmem_cache_free(fsnotify_event_holder_cachep, holder);
+}
+
+/*
+ * Find the private data that the group previously attached to this event when
+ * the group added the event to the notification queue (fsnotify_add_notify_event)
+ */
+struct fsnotify_event_private_data *fsnotify_remove_priv_from_event(struct fsnotify_group *group, struct fsnotify_event *event)
+{
+	struct fsnotify_event_private_data *lpriv;
+	struct fsnotify_event_private_data *priv = NULL;
+
+	assert_spin_locked(&event->lock);
+
+	list_for_each_entry(lpriv, &event->private_data_list, event_list) {
+		if (lpriv->group == group) {
+			priv = lpriv;
+			list_del(&priv->event_list);
+			break;
+		}
+	}
+	return priv;
+}
+
+/*
+ * Add an event to the group notification queue.  The group can later pull this
+ * event off the queue to deal with.  If the event is successfully added to the
+ * group's notification queue, a reference is taken on event.
+ */
+struct fsnotify_event *fsnotify_add_notify_event(struct fsnotify_group *group, struct fsnotify_event *event,
+						 struct fsnotify_event_private_data *priv,
+						 struct fsnotify_event *(*merge)(struct list_head *,
+										 struct fsnotify_event *))
+{
+	struct fsnotify_event *return_event = NULL;
+	struct fsnotify_event_holder *holder = NULL;
+	struct list_head *list = &group->notification_list;
+
+	pr_debug("%s: group=%p event=%p priv=%p\n", __func__, group, event, priv);
+
+	/*
+	 * There is one fsnotify_event_holder embedded inside each fsnotify_event.
+	 * Check if we expect to be able to use that holder.  If not alloc a new
+	 * holder.
+	 * For the overflow event it's possible that something will use the in
+	 * event holder before we get the lock so we may need to jump back and
+	 * alloc a new holder, this can't happen for most events...
+	 */
+	if (!list_empty(&event->holder.event_list)) {
+alloc_holder:
+		holder = fsnotify_alloc_event_holder();
+		if (!holder)
+			return ERR_PTR(-ENOMEM);
+	}
+
+	mutex_lock(&group->notification_mutex);
+
+	if (group->q_len >= group->max_events) {
+		event = q_overflow_event;
+
+		/*
+		 * we need to return the overflow event
+		 * which means we need a ref
+		 */
+		fsnotify_get_event(event);
+		return_event = event;
+
+		/* sorry, no private data on the overflow event */
+		priv = NULL;
+	}
+
+	if (!list_empty(list) && merge) {
+		struct fsnotify_event *tmp;
+
+		tmp = merge(list, event);
+		if (tmp) {
+			mutex_unlock(&group->notification_mutex);
+
+			if (return_event)
+				fsnotify_put_event(return_event);
+			if (holder != &event->holder)
+				fsnotify_destroy_event_holder(holder);
+			return tmp;
+		}
+	}
+
+	spin_lock(&event->lock);
+
+	if (list_empty(&event->holder.event_list)) {
+		if (unlikely(holder))
+			fsnotify_destroy_event_holder(holder);
+		holder = &event->holder;
+	} else if (unlikely(!holder)) {
+		/* between the time we checked above and got the lock the in
+		 * event holder was used, go back and get a new one */
+		spin_unlock(&event->lock);
+		mutex_unlock(&group->notification_mutex);
+
+		if (return_event) {
+			fsnotify_put_event(return_event);
+			return_event = NULL;
+		}
+
+		goto alloc_holder;
+	}
+
+	group->q_len++;
+	holder->event = event;
+
+	fsnotify_get_event(event);
+	list_add_tail(&holder->event_list, list);
+	if (priv)
+		list_add_tail(&priv->event_list, &event->private_data_list);
+	spin_unlock(&event->lock);
+	mutex_unlock(&group->notification_mutex);
+
+	wake_up(&group->notification_waitq);
+	kill_fasync(&group->fsn_fa, SIGIO, POLL_IN);
+	return return_event;
+}
+
+/*
+ * Remove and return the first event from the notification list.  There is a
+ * reference held on this event since it was on the list.  It is the responsibility
+ * of the caller to drop this reference.
+ */
+struct fsnotify_event *fsnotify_remove_notify_event(struct fsnotify_group *group)
+{
+	struct fsnotify_event *event;
+	struct fsnotify_event_holder *holder;
+
+	BUG_ON(!mutex_is_locked(&group->notification_mutex));
+
+	pr_debug("%s: group=%p\n", __func__, group);
+
+	holder = list_first_entry(&group->notification_list, struct fsnotify_event_holder, event_list);
+
+	event = holder->event;
+
+	spin_lock(&event->lock);
+	holder->event = NULL;
+	list_del_init(&holder->event_list);
+	spin_unlock(&event->lock);
+
+	/* event == holder means we are referenced through the in event holder */
+	if (holder != &event->holder)
+		fsnotify_destroy_event_holder(holder);
+
+	group->q_len--;
+
+	return event;
+}
+
+/*
+ * This will not remove the event, that must be done with fsnotify_remove_notify_event()
+ */
+struct fsnotify_event *fsnotify_peek_notify_event(struct fsnotify_group *group)
+{
+	struct fsnotify_event *event;
+	struct fsnotify_event_holder *holder;
+
+	BUG_ON(!mutex_is_locked(&group->notification_mutex));
+
+	holder = list_first_entry(&group->notification_list, struct fsnotify_event_holder, event_list);
+	event = holder->event;
+
+	return event;
+}
+
+/*
+ * Called when a group is being torn down to clean up any outstanding
+ * event notifications.
+ */
+void fsnotify_flush_notify(struct fsnotify_group *group)
+{
+	struct fsnotify_event *event;
+	struct fsnotify_event_private_data *priv;
+
+	mutex_lock(&group->notification_mutex);
+	while (!fsnotify_notify_queue_is_empty(group)) {
+		event = fsnotify_remove_notify_event(group);
+		/* if they don't implement free_event_priv they better not have attached any */
+		if (group->ops->free_event_priv) {
+			spin_lock(&event->lock);
+			priv = fsnotify_remove_priv_from_event(group, event);
+			spin_unlock(&event->lock);
+			if (priv)
+				group->ops->free_event_priv(priv);
+		}
+		fsnotify_put_event(event); /* matches fsnotify_add_notify_event */
+	}
+	mutex_unlock(&group->notification_mutex);
+}
+
+static void initialize_event(struct fsnotify_event *event)
+{
+	INIT_LIST_HEAD(&event->holder.event_list);
+	atomic_set(&event->refcnt, 1);
+
+	spin_lock_init(&event->lock);
+
+	INIT_LIST_HEAD(&event->private_data_list);
+}
+
+/*
+ * Caller damn well better be holding whatever mutex is protecting the
+ * old_holder->event_list and the new_event must be a clean event which
+ * cannot be found anywhere else in the kernel.
+ */
+int fsnotify_replace_event(struct fsnotify_event_holder *old_holder,
+			   struct fsnotify_event *new_event)
+{
+	struct fsnotify_event *old_event = old_holder->event;
+	struct fsnotify_event_holder *new_holder = &new_event->holder;
+
+	enum event_spinlock_class {
+		SPINLOCK_OLD,
+		SPINLOCK_NEW,
+	};
+
+	pr_debug("%s: old_event=%p new_event=%p\n", __func__, old_event, new_event);
+
+	/*
+	 * if the new_event's embedded holder is in use someone
+	 * screwed up and didn't give us a clean new event.
+	 */
+	BUG_ON(!list_empty(&new_holder->event_list));
+
+	spin_lock_nested(&old_event->lock, SPINLOCK_OLD);
+	spin_lock_nested(&new_event->lock, SPINLOCK_NEW);
+
+	new_holder->event = new_event;
+	list_replace_init(&old_holder->event_list, &new_holder->event_list);
+
+	spin_unlock(&new_event->lock);
+	spin_unlock(&old_event->lock);
+
+	/* event == holder means we are referenced through the in event holder */
+	if (old_holder != &old_event->holder)
+		fsnotify_destroy_event_holder(old_holder);
+
+	fsnotify_get_event(new_event); /* on the list take reference */
+	fsnotify_put_event(old_event); /* off the list, drop reference */
+
+	return 0;
+}
+
+struct fsnotify_event *fsnotify_clone_event(struct fsnotify_event *old_event)
+{
+	struct fsnotify_event *event;
+
+	event = kmem_cache_alloc(fsnotify_event_cachep, GFP_KERNEL);
+	if (!event)
+		return NULL;
+
+	pr_debug("%s: old_event=%p new_event=%p\n", __func__, old_event, event);
+
+	memcpy(event, old_event, sizeof(*event));
+	initialize_event(event);
+
+	if (event->name_len) {
+		event->file_name = kstrdup(old_event->file_name, GFP_KERNEL);
+		if (!event->file_name) {
+			kmem_cache_free(fsnotify_event_cachep, event);
+			return NULL;
+		}
+	}
+	event->tgid = get_pid(old_event->tgid);
+	if (event->data_type == FSNOTIFY_EVENT_PATH)
+		path_get(&event->path);
+
+	return event;
+}
+
+/*
+ * fsnotify_create_event - Allocate a new event which will be sent to each
+ * group's handle_event function if the group was interested in this
+ * particular event.
+ *
+ * @to_tell the inode which is supposed to receive the event (sometimes a
+ *	parent of the inode to which the event happened.
+ * @mask what actually happened.
+ * @data pointer to the object which was actually affected
+ * @data_type flag indication if the data is a file, path, inode, nothing...
+ * @name the filename, if available
+ */
+struct fsnotify_event *fsnotify_create_event(struct inode *to_tell, __u32 mask, void *data,
+					     int data_type, const unsigned char *name,
+					     u32 cookie, gfp_t gfp)
+{
+	struct fsnotify_event *event;
+
+	event = kmem_cache_zalloc(fsnotify_event_cachep, gfp);
+	if (!event)
+		return NULL;
+
+	pr_debug("%s: event=%p to_tell=%p mask=%x data=%p data_type=%d\n",
+		 __func__, event, to_tell, mask, data, data_type);
+
+	initialize_event(event);
+
+	if (name) {
+		event->file_name = kstrdup(name, gfp);
+		if (!event->file_name) {
+			kmem_cache_free(fsnotify_event_cachep, event);
+			return NULL;
+		}
+		event->name_len = strlen(event->file_name);
+	}
+
+	event->tgid = get_pid(task_tgid(current));
+	event->sync_cookie = cookie;
+	event->to_tell = to_tell;
+	event->data_type = data_type;
+
+	switch (data_type) {
+	case FSNOTIFY_EVENT_PATH: {
+		struct path *path = data;
+		event->path.dentry = path->dentry;
+		event->path.mnt = path->mnt;
+		path_get(&event->path);
+		break;
+	}
+	case FSNOTIFY_EVENT_INODE:
+		event->inode = data;
+		break;
+	case FSNOTIFY_EVENT_NONE:
+		event->inode = NULL;
+		event->path.dentry = NULL;
+		event->path.mnt = NULL;
+		break;
+	default:
+		BUG();
+	}
+
+	event->mask = mask;
+
+	return event;
+}
+
+static __init int fsnotify_notification_init(void)
+{
+	fsnotify_event_cachep = KMEM_CACHE(fsnotify_event, SLAB_PANIC);
+	fsnotify_event_holder_cachep = KMEM_CACHE(fsnotify_event_holder, SLAB_PANIC);
+
+	q_overflow_event = fsnotify_create_event(NULL, FS_Q_OVERFLOW, NULL,
+						 FSNOTIFY_EVENT_NONE, NULL, 0,
+						 GFP_KERNEL);
+	if (!q_overflow_event)
+		panic("unable to allocate fsnotify q_overflow_event\n");
+
+	return 0;
+}
+subsys_initcall(fsnotify_notification_init);

fs/notify/vfsmount_mark.c

@@ -0,0 +1,200 @@
+/*
+ *  Copyright (C) 2008 Red Hat, Inc., Eric Paris <eparis@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mount.h>
+#include <linux/mutex.h>
+#include <linux/spinlock.h>
+
+#include <linux/atomic.h>
+
+#include <linux/fsnotify_backend.h>
+#include "fsnotify.h"
+#include "../mount.h"
+
+void fsnotify_clear_marks_by_mount(struct vfsmount *mnt)
+{
+	struct fsnotify_mark *mark, *lmark;
+	struct hlist_node *n;
+	struct mount *m = real_mount(mnt);
+	LIST_HEAD(free_list);
+
+	spin_lock(&mnt->mnt_root->d_lock);
+	hlist_for_each_entry_safe(mark, n, &m->mnt_fsnotify_marks, m.m_list) {
+		list_add(&mark->m.free_m_list, &free_list);
+		hlist_del_init_rcu(&mark->m.m_list);
+		fsnotify_get_mark(mark);
+	}
+	spin_unlock(&mnt->mnt_root->d_lock);
+
+	list_for_each_entry_safe(mark, lmark, &free_list, m.free_m_list) {
+		struct fsnotify_group *group;
+
+		spin_lock(&mark->lock);
+		fsnotify_get_group(mark->group);
+		group = mark->group;
+		spin_unlock(&mark->lock);
+
+		fsnotify_destroy_mark(mark, group);
+		fsnotify_put_mark(mark);
+		fsnotify_put_group(group);
+	}
+}
+
+void fsnotify_clear_vfsmount_marks_by_group(struct fsnotify_group *group)
+{
+	fsnotify_clear_marks_by_group_flags(group, FSNOTIFY_MARK_FLAG_VFSMOUNT);
+}
+
+/*
+ * Recalculate the mask of events relevant to a given vfsmount locked.
+ */
+static void fsnotify_recalc_vfsmount_mask_locked(struct vfsmount *mnt)
+{
+	struct mount *m = real_mount(mnt);
+	struct fsnotify_mark *mark;
+	__u32 new_mask = 0;
+
+	assert_spin_locked(&mnt->mnt_root->d_lock);
+
+	hlist_for_each_entry(mark, &m->mnt_fsnotify_marks, m.m_list)
+		new_mask |= mark->mask;
+	m->mnt_fsnotify_mask = new_mask;
+}
+
+/*
+ * Recalculate the mnt->mnt_fsnotify_mask, or the mask of all FS_* event types
+ * any notifier is interested in hearing for this mount point
+ */
+void fsnotify_recalc_vfsmount_mask(struct vfsmount *mnt)
+{
+	spin_lock(&mnt->mnt_root->d_lock);
+	fsnotify_recalc_vfsmount_mask_locked(mnt);
+	spin_unlock(&mnt->mnt_root->d_lock);
+}
+
+void fsnotify_destroy_vfsmount_mark(struct fsnotify_mark *mark)
+{
+	struct vfsmount *mnt = mark->m.mnt;
+
+	BUG_ON(!mutex_is_locked(&mark->group->mark_mutex));
+	assert_spin_locked(&mark->lock);
+
+	spin_lock(&mnt->mnt_root->d_lock);
+
+	hlist_del_init_rcu(&mark->m.m_list);
+	mark->m.mnt = NULL;
+
+	fsnotify_recalc_vfsmount_mask_locked(mnt);
+
+	spin_unlock(&mnt->mnt_root->d_lock);
+}
+
+static struct fsnotify_mark *fsnotify_find_vfsmount_mark_locked(struct fsnotify_group *group,
+								struct vfsmount *mnt)
+{
+	struct mount *m = real_mount(mnt);
+	struct fsnotify_mark *mark;
+
+	assert_spin_locked(&mnt->mnt_root->d_lock);
+
+	hlist_for_each_entry(mark, &m->mnt_fsnotify_marks, m.m_list) {
+		if (mark->group == group) {
+			fsnotify_get_mark(mark);
+			return mark;
+		}
+	}
+	return NULL;
+}
+
+/*
+ * given a group and vfsmount, find the mark associated with that combination.
+ * if found take a reference to that mark and return it, else return NULL
+ */
+struct fsnotify_mark *fsnotify_find_vfsmount_mark(struct fsnotify_group *group,
+						  struct vfsmount *mnt)
+{
+	struct fsnotify_mark *mark;
+
+	spin_lock(&mnt->mnt_root->d_lock);
+	mark = fsnotify_find_vfsmount_mark_locked(group, mnt);
+	spin_unlock(&mnt->mnt_root->d_lock);
+
+	return mark;
+}
+
+/*
+ * Attach an initialized mark to a given group and vfsmount.
+ * These marks may be used for the fsnotify backend to determine which
+ * event types should be delivered to which groups.
+ */
+int fsnotify_add_vfsmount_mark(struct fsnotify_mark *mark,
+			       struct fsnotify_group *group, struct vfsmount *mnt,
+			       int allow_dups)
+{
+	struct mount *m = real_mount(mnt);
+	struct fsnotify_mark *lmark, *last = NULL;
+	int ret = 0;
+
+	mark->flags |= FSNOTIFY_MARK_FLAG_VFSMOUNT;
+
+	BUG_ON(!mutex_is_locked(&group->mark_mutex));
+	assert_spin_locked(&mark->lock);
+
+	spin_lock(&mnt->mnt_root->d_lock);
+
+	mark->m.mnt = mnt;
+
+	/* is mark the first mark? */
+	if (hlist_empty(&m->mnt_fsnotify_marks)) {
+		hlist_add_head_rcu(&mark->m.m_list, &m->mnt_fsnotify_marks);
+		goto out;
+	}
+
+	/* should mark be in the middle of the current list? */
+	hlist_for_each_entry(lmark, &m->mnt_fsnotify_marks, m.m_list) {
+		last = lmark;
+
+		if ((lmark->group == group) && !allow_dups) {
+			ret = -EEXIST;
+			goto out;
+		}
+
+		if (mark->group->priority < lmark->group->priority)
+			continue;
+
+		if ((mark->group->priority == lmark->group->priority) &&
+		    (mark->group < lmark->group))
+			continue;
+
+		hlist_add_before_rcu(&mark->m.m_list, &lmark->m.m_list);
+		goto out;
+	}
+
+	BUG_ON(last == NULL);
+	/* mark should be the last entry.  last is the current last entry */
+	hlist_add_after_rcu(&last->m.m_list, &mark->m.m_list);
+out:
+	fsnotify_recalc_vfsmount_mask_locked(mnt);
+	spin_unlock(&mnt->mnt_root->d_lock);
+
+	return ret;
+}