1 files changed, 261 insertions, 66 deletions

diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c
index 559bc1e9926d..4546da4a54f9 100644
--- a/fs/notify/fanotify/fanotify_user.c
+++ b/fs/notify/fanotify/fanotify_user.c
@@ -59,7 +59,7 @@ static int fanotify_max_queued_events __read_mostly;
 static long ft_zero = 0;
 static long ft_int_max = INT_MAX;
 
-struct ctl_table fanotify_table[] = {
+static struct ctl_table fanotify_table[] = {
 	{
 		.procname	= "max_user_groups",
 		.data	= &init_user_ns.ucount_max[UCOUNT_FANOTIFY_GROUPS],
@@ -88,6 +88,13 @@ struct ctl_table fanotify_table[] = {
 	},
 	{ }
 };
+
+static void __init fanotify_sysctls_init(void)
+{
+	register_sysctl("fs/fanotify", fanotify_table);
+}
+#else
+#define fanotify_sysctls_init() do { } while (0)
 #endif /* CONFIG_SYSCTL */
 
 /*
@@ -129,12 +136,28 @@ static int fanotify_fid_info_len(int fh_len, int name_len)
 		       FANOTIFY_EVENT_ALIGN);
 }
 
+/* FAN_RENAME may have one or two dir+name info records */
+static int fanotify_dir_name_info_len(struct fanotify_event *event)
+{
+	struct fanotify_info *info = fanotify_event_info(event);
+	int dir_fh_len = fanotify_event_dir_fh_len(event);
+	int dir2_fh_len = fanotify_event_dir2_fh_len(event);
+	int info_len = 0;
+
+	if (dir_fh_len)
+		info_len += fanotify_fid_info_len(dir_fh_len,
+						  info->name_len);
+	if (dir2_fh_len)
+		info_len += fanotify_fid_info_len(dir2_fh_len,
+						  info->name2_len);
+
+	return info_len;
+}
+
 static size_t fanotify_event_len(unsigned int info_mode,
 				 struct fanotify_event *event)
 {
 	size_t event_len = FAN_EVENT_METADATA_LEN;
-	struct fanotify_info *info;
-	int dir_fh_len;
 	int fh_len;
 	int dot_len = 0;
 
@@ -144,11 +167,8 @@ static size_t fanotify_event_len(unsigned int info_mode,
 	if (fanotify_is_error_event(event->mask))
 		event_len += FANOTIFY_ERROR_INFO_LEN;
 
-	info = fanotify_event_info(event);
-
-	if (fanotify_event_has_dir_fh(event)) {
-		dir_fh_len = fanotify_event_dir_fh_len(event);
-		event_len += fanotify_fid_info_len(dir_fh_len, info->name_len);
+	if (fanotify_event_has_any_dir_fh(event)) {
+		event_len += fanotify_dir_name_info_len(event);
 	} else if ((info_mode & FAN_REPORT_NAME) &&
 		   (event->mask & FAN_ONDIR)) {
 		/*
@@ -229,7 +249,7 @@ out:
 	return event;
 }
 
-static int create_fd(struct fsnotify_group *group, struct path *path,
+static int create_fd(struct fsnotify_group *group, const struct path *path,
 		     struct file **file)
 {
 	int client_fd;
@@ -244,7 +264,7 @@ static int create_fd(struct fsnotify_group *group, struct path *path,
 	 * originally opened O_WRONLY.
 	 */
 	new_file = dentry_open(path,
-			       group->fanotify_data.f_flags | FMODE_NONOTIFY,
+			       group->fanotify_data.f_flags | __FMODE_NONOTIFY,
 			       current_cred());
 	if (IS_ERR(new_file)) {
 		/*
@@ -332,11 +352,10 @@ static int process_access_response(struct fsnotify_group *group,
 static size_t copy_error_info_to_user(struct fanotify_event *event,
 				      char __user *buf, int count)
 {
-	struct fanotify_event_info_error info;
+	struct fanotify_event_info_error info = { };
 	struct fanotify_error_event *fee = FANOTIFY_EE(event);
 
 	info.hdr.info_type = FAN_EVENT_INFO_TYPE_ERROR;
-	info.hdr.pad = 0;
 	info.hdr.len = FANOTIFY_ERROR_INFO_LEN;
 
 	if (WARN_ON(count < info.hdr.len))
@@ -380,6 +399,8 @@ static int copy_fid_info_to_user(__kernel_fsid_t *fsid, struct fanotify_fh *fh,
 			return -EFAULT;
 		break;
 	case FAN_EVENT_INFO_TYPE_DFID_NAME:
+	case FAN_EVENT_INFO_TYPE_OLD_DFID_NAME:
+	case FAN_EVENT_INFO_TYPE_NEW_DFID_NAME:
 		if (WARN_ON_ONCE(!name || !name_len))
 			return -EFAULT;
 		break;
@@ -479,11 +500,19 @@ static int copy_info_records_to_user(struct fanotify_event *event,
 	unsigned int pidfd_mode = info_mode & FAN_REPORT_PIDFD;
 
 	/*
-	 * Event info records order is as follows: dir fid + name, child fid.
+	 * Event info records order is as follows:
+	 * 1. dir fid + name
+	 * 2. (optional) new dir fid + new name
+	 * 3. (optional) child fid
 	 */
 	if (fanotify_event_has_dir_fh(event)) {
 		info_type = info->name_len ? FAN_EVENT_INFO_TYPE_DFID_NAME :
 					     FAN_EVENT_INFO_TYPE_DFID;
+
+		/* FAN_RENAME uses special info types */
+		if (event->mask & FAN_RENAME)
+			info_type = FAN_EVENT_INFO_TYPE_OLD_DFID_NAME;
+
 		ret = copy_fid_info_to_user(fanotify_event_fsid(event),
 					    fanotify_info_dir_fh(info),
 					    info_type,
@@ -497,6 +526,22 @@ static int copy_info_records_to_user(struct fanotify_event *event,
 		total_bytes += ret;
 	}
 
+	/* New dir fid+name may be reported in addition to old dir fid+name */
+	if (fanotify_event_has_dir2_fh(event)) {
+		info_type = FAN_EVENT_INFO_TYPE_NEW_DFID_NAME;
+		ret = copy_fid_info_to_user(fanotify_event_fsid(event),
+					    fanotify_info_dir2_fh(info),
+					    info_type,
+					    fanotify_info_name2(info),
+					    info->name2_len, buf, count);
+		if (ret < 0)
+			return ret;
+
+		buf += ret;
+		count -= ret;
+		total_bytes += ret;
+	}
+
 	if (fanotify_event_has_object_fh(event)) {
 		const char *dot = NULL;
 		int dot_len = 0;
@@ -574,7 +619,7 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group,
 				  char __user *buf, size_t count)
 {
 	struct fanotify_event_metadata metadata;
-	struct path *path = fanotify_event_path(event);
+	const struct path *path = fanotify_event_path(event);
 	struct fanotify_info *info = fanotify_event_info(event);
 	unsigned int info_mode = FAN_GROUP_FLAG(group, FANOTIFY_INFO_MODES);
 	unsigned int pidfd_mode = info_mode & FAN_REPORT_PIDFD;
@@ -656,9 +701,6 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group,
 	if (fanotify_is_perm_event(event->mask))
 		FANOTIFY_PERM(event)->fd = fd;
 
-	if (f)
-		fd_install(fd, f);
-
 	if (info_mode) {
 		ret = copy_info_records_to_user(event, info, info_mode, pidfd,
 						buf, count);
@@ -666,6 +708,9 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group,
 			goto out_close_fd;
 	}
 
+	if (f)
+		fd_install(fd, f);
+
 	return metadata.event_len;
 
 out_close_fd:
@@ -958,27 +1003,28 @@ static __u32 fanotify_mark_remove_from_mask(struct fsnotify_mark *fsn_mark,
 					    __u32 mask, unsigned int flags,
 					    __u32 umask, int *destroy)
 {
-	__u32 oldmask = 0;
+	__u32 oldmask, newmask;
 
 	/* umask bits cannot be removed by user */
 	mask &= ~umask;
 	spin_lock(&fsn_mark->lock);
-	if (!(flags & FAN_MARK_IGNORED_MASK)) {
-		oldmask = fsn_mark->mask;
+	oldmask = fsnotify_calc_mask(fsn_mark);
+	if (!(flags & FANOTIFY_MARK_IGNORE_BITS)) {
 		fsn_mark->mask &= ~mask;
 	} else {
-		fsn_mark->ignored_mask &= ~mask;
+		fsn_mark->ignore_mask &= ~mask;
 	}
+	newmask = fsnotify_calc_mask(fsn_mark);
 	/*
 	 * We need to keep the mark around even if remaining mask cannot
 	 * result in any events (e.g. mask == FAN_ONDIR) to support incremenal
 	 * changes to the mask.
 	 * Destroy mark when only umask bits remain.
 	 */
-	*destroy = !((fsn_mark->mask | fsn_mark->ignored_mask) & ~umask);
+	*destroy = !((fsn_mark->mask | fsn_mark->ignore_mask) & ~umask);
 	spin_unlock(&fsn_mark->lock);
 
-	return mask & oldmask;
+	return oldmask & ~newmask;
 }
 
 static int fanotify_remove_mark(struct fsnotify_group *group,
@@ -989,10 +1035,10 @@ static int fanotify_remove_mark(struct fsnotify_group *group,
 	__u32 removed;
 	int destroy_mark;
 
-	mutex_lock(&group->mark_mutex);
+	fsnotify_group_lock(group);
 	fsn_mark = fsnotify_find_mark(connp, group);
 	if (!fsn_mark) {
-		mutex_unlock(&group->mark_mutex);
+		fsnotify_group_unlock(group);
 		return -ENOENT;
 	}
 
@@ -1002,7 +1048,7 @@ static int fanotify_remove_mark(struct fsnotify_group *group,
 		fsnotify_recalc_mask(fsn_mark->connector);
 	if (destroy_mark)
 		fsnotify_detach_mark(fsn_mark);
-	mutex_unlock(&group->mark_mutex);
+	fsnotify_group_unlock(group);
 	if (destroy_mark)
 		fsnotify_free_mark(fsn_mark);
 
@@ -1035,29 +1081,72 @@ static int fanotify_remove_inode_mark(struct fsnotify_group *group,
 				    flags, umask);
 }
 
-static __u32 fanotify_mark_add_to_mask(struct fsnotify_mark *fsn_mark,
-				       __u32 mask,
-				       unsigned int flags)
+static bool fanotify_mark_update_flags(struct fsnotify_mark *fsn_mark,
+				       unsigned int fan_flags)
+{
+	bool want_iref = !(fan_flags & FAN_MARK_EVICTABLE);
+	unsigned int ignore = fan_flags & FANOTIFY_MARK_IGNORE_BITS;
+	bool recalc = false;
+
+	/*
+	 * When using FAN_MARK_IGNORE for the first time, mark starts using
+	 * independent event flags in ignore mask.  After that, trying to
+	 * update the ignore mask with the old FAN_MARK_IGNORED_MASK API
+	 * will result in EEXIST error.
+	 */
+	if (ignore == FAN_MARK_IGNORE)
+		fsn_mark->flags |= FSNOTIFY_MARK_FLAG_HAS_IGNORE_FLAGS;
+
+	/*
+	 * Setting FAN_MARK_IGNORED_SURV_MODIFY for the first time may lead to
+	 * the removal of the FS_MODIFY bit in calculated mask if it was set
+	 * because of an ignore mask that is now going to survive FS_MODIFY.
+	 */
+	if (ignore && (fan_flags & FAN_MARK_IGNORED_SURV_MODIFY) &&
+	    !(fsn_mark->flags & FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY)) {
+		fsn_mark->flags |= FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY;
+		if (!(fsn_mark->mask & FS_MODIFY))
+			recalc = true;
+	}
+
+	if (fsn_mark->connector->type != FSNOTIFY_OBJ_TYPE_INODE ||
+	    want_iref == !(fsn_mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF))
+		return recalc;
+
+	/*
+	 * NO_IREF may be removed from a mark, but not added.
+	 * When removed, fsnotify_recalc_mask() will take the inode ref.
+	 */
+	WARN_ON_ONCE(!want_iref);
+	fsn_mark->flags &= ~FSNOTIFY_MARK_FLAG_NO_IREF;
+
+	return true;
+}
+
+static bool fanotify_mark_add_to_mask(struct fsnotify_mark *fsn_mark,
+				      __u32 mask, unsigned int fan_flags)
 {
-	__u32 oldmask = -1;
+	bool recalc;
 
 	spin_lock(&fsn_mark->lock);
-	if (!(flags & FAN_MARK_IGNORED_MASK)) {
-		oldmask = fsn_mark->mask;
+	if (!(fan_flags & FANOTIFY_MARK_IGNORE_BITS))
 		fsn_mark->mask |= mask;
-	} else {
-		fsn_mark->ignored_mask |= mask;
-		if (flags & FAN_MARK_IGNORED_SURV_MODIFY)
-			fsn_mark->flags |= FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY;
-	}
+	else
+		fsn_mark->ignore_mask |= mask;
+
+	recalc = fsnotify_calc_mask(fsn_mark) &
+		~fsnotify_conn_mask(fsn_mark->connector);
+
+	recalc |= fanotify_mark_update_flags(fsn_mark, fan_flags);
 	spin_unlock(&fsn_mark->lock);
 
-	return mask & ~oldmask;
+	return recalc;
 }
 
 static struct fsnotify_mark *fanotify_add_new_mark(struct fsnotify_group *group,
 						   fsnotify_connp_t *connp,
-						   unsigned int type,
+						   unsigned int obj_type,
+						   unsigned int fan_flags,
 						   __kernel_fsid_t *fsid)
 {
 	struct ucounts *ucounts = group->fanotify_data.ucounts;
@@ -1080,7 +1169,10 @@ static struct fsnotify_mark *fanotify_add_new_mark(struct fsnotify_group *group,
 	}
 
 	fsnotify_init_mark(mark, group);
-	ret = fsnotify_add_mark_locked(mark, connp, type, 0, fsid);
+	if (fan_flags & FAN_MARK_EVICTABLE)
+		mark->flags |= FSNOTIFY_MARK_FLAG_NO_IREF;
+
+	ret = fsnotify_add_mark_locked(mark, connp, obj_type, 0, fsid);
 	if (ret) {
 		fsnotify_put_mark(mark);
 		goto out_dec_ucounts;
@@ -1104,41 +1196,81 @@ static int fanotify_group_init_error_pool(struct fsnotify_group *group)
 					 sizeof(struct fanotify_error_event));
 }
 
+static int fanotify_may_update_existing_mark(struct fsnotify_mark *fsn_mark,
+					      unsigned int fan_flags)
+{
+	/*
+	 * Non evictable mark cannot be downgraded to evictable mark.
+	 */
+	if (fan_flags & FAN_MARK_EVICTABLE &&
+	    !(fsn_mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF))
+		return -EEXIST;
+
+	/*
+	 * New ignore mask semantics cannot be downgraded to old semantics.
+	 */
+	if (fan_flags & FAN_MARK_IGNORED_MASK &&
+	    fsn_mark->flags & FSNOTIFY_MARK_FLAG_HAS_IGNORE_FLAGS)
+		return -EEXIST;
+
+	/*
+	 * An ignore mask that survives modify could never be downgraded to not
+	 * survive modify.  With new FAN_MARK_IGNORE semantics we make that rule
+	 * explicit and return an error when trying to update the ignore mask
+	 * without the original FAN_MARK_IGNORED_SURV_MODIFY value.
+	 */
+	if (fan_flags & FAN_MARK_IGNORE &&
+	    !(fan_flags & FAN_MARK_IGNORED_SURV_MODIFY) &&
+	    fsn_mark->flags & FSNOTIFY_MARK_FLAG_IGNORED_SURV_MODIFY)
+		return -EEXIST;
+
+	return 0;
+}
+
 static int fanotify_add_mark(struct fsnotify_group *group,
-			     fsnotify_connp_t *connp, unsigned int type,
-			     __u32 mask, unsigned int flags,
+			     fsnotify_connp_t *connp, unsigned int obj_type,
+			     __u32 mask, unsigned int fan_flags,
 			     __kernel_fsid_t *fsid)
 {
 	struct fsnotify_mark *fsn_mark;
-	__u32 added;
+	bool recalc;
 	int ret = 0;
 
-	mutex_lock(&group->mark_mutex);
+	fsnotify_group_lock(group);
 	fsn_mark = fsnotify_find_mark(connp, group);
 	if (!fsn_mark) {
-		fsn_mark = fanotify_add_new_mark(group, connp, type, fsid);
+		fsn_mark = fanotify_add_new_mark(group, connp, obj_type,
+						 fan_flags, fsid);
 		if (IS_ERR(fsn_mark)) {
-			mutex_unlock(&group->mark_mutex);
+			fsnotify_group_unlock(group);
 			return PTR_ERR(fsn_mark);
 		}
 	}
 
 	/*
+	 * Check if requested mark flags conflict with an existing mark flags.
+	 */
+	ret = fanotify_may_update_existing_mark(fsn_mark, fan_flags);
+	if (ret)
+		goto out;
+
+	/*
 	 * Error events are pre-allocated per group, only if strictly
 	 * needed (i.e. FAN_FS_ERROR was requested).
 	 */
-	if (!(flags & FAN_MARK_IGNORED_MASK) && (mask & FAN_FS_ERROR)) {
+	if (!(fan_flags & FANOTIFY_MARK_IGNORE_BITS) &&
+	    (mask & FAN_FS_ERROR)) {
 		ret = fanotify_group_init_error_pool(group);
 		if (ret)
 			goto out;
 	}
 
-	added = fanotify_mark_add_to_mask(fsn_mark, mask, flags);
-	if (added & ~fsnotify_conn_mask(fsn_mark->connector))
+	recalc = fanotify_mark_add_to_mask(fsn_mark, mask, fan_flags);
+	if (recalc)
 		fsnotify_recalc_mask(fsn_mark->connector);
 
 out:
-	mutex_unlock(&group->mark_mutex);
+	fsnotify_group_unlock(group);
 
 	fsnotify_put_mark(fsn_mark);
 	return ret;
@@ -1168,10 +1300,10 @@ static int fanotify_add_inode_mark(struct fsnotify_group *group,
 
 	/*
 	 * 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
+	 * an ignore mask, unless that ignore mask is supposed to survive
 	 * modification changes anyway.
 	 */
-	if ((flags & FAN_MARK_IGNORED_MASK) &&
+	if ((flags & FANOTIFY_MARK_IGNORE_BITS) &&
 	    !(flags & FAN_MARK_IGNORED_SURV_MODIFY) &&
 	    inode_is_open_for_write(inode))
 		return 0;
@@ -1275,14 +1407,24 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags)
 	if ((fid_mode & FAN_REPORT_NAME) && !(fid_mode & FAN_REPORT_DIR_FID))
 		return -EINVAL;
 
-	f_flags = O_RDWR | FMODE_NONOTIFY;
+	/*
+	 * FAN_REPORT_TARGET_FID requires FAN_REPORT_NAME and FAN_REPORT_FID
+	 * and is used as an indication to report both dir and child fid on all
+	 * dirent events.
+	 */
+	if ((fid_mode & FAN_REPORT_TARGET_FID) &&
+	    (!(fid_mode & FAN_REPORT_NAME) || !(fid_mode & FAN_REPORT_FID)))
+		return -EINVAL;
+
+	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_user_group(&fanotify_fsnotify_ops);
+	group = fsnotify_alloc_group(&fanotify_fsnotify_ops,
+				     FSNOTIFY_GROUP_USER | FSNOTIFY_GROUP_NOFS);
 	if (IS_ERR(group)) {
 		return PTR_ERR(group);
 	}
@@ -1410,8 +1552,16 @@ static int fanotify_test_fid(struct dentry *dentry)
 	return 0;
 }
 
-static int fanotify_events_supported(struct path *path, __u64 mask)
+static int fanotify_events_supported(struct fsnotify_group *group,
+				     const struct path *path, __u64 mask,
+				     unsigned int flags)
 {
+	unsigned int mark_type = flags & FANOTIFY_MARK_TYPE_BITS;
+	/* Strict validation of events in non-dir inode mask with v5.17+ APIs */
+	bool strict_dir_events = FAN_GROUP_FLAG(group, FAN_REPORT_TARGET_FID) ||
+				 (mask & FAN_RENAME) ||
+				 (flags & FAN_MARK_IGNORE);
+
 	/*
 	 * Some filesystems such as 'proc' acquire unusual locks when opening
 	 * files. For them fanotify permission events have high chances of
@@ -1423,6 +1573,16 @@ static int fanotify_events_supported(struct path *path, __u64 mask)
 	if (mask & FANOTIFY_PERM_EVENTS &&
 	    path->mnt->mnt_sb->s_type->fs_flags & FS_DISALLOW_NOTIFY_PERM)
 		return -EINVAL;
+
+	/*
+	 * We shouldn't have allowed setting dirent events and the directory
+	 * flags FAN_ONDIR and FAN_EVENT_ON_CHILD in mask of non-dir inode,
+	 * but because we always allowed it, error only when using new APIs.
+	 */
+	if (strict_dir_events && mark_type == FAN_MARK_INODE &&
+	    !d_is_dir(path->dentry) && (mask & FANOTIFY_DIRONLY_EVENT_BITS))
+		return -ENOTDIR;
+
 	return 0;
 }
 
@@ -1437,7 +1597,8 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	__kernel_fsid_t __fsid, *fsid = NULL;
 	u32 valid_mask = FANOTIFY_EVENTS | FANOTIFY_EVENT_FLAGS;
 	unsigned int mark_type = flags & FANOTIFY_MARK_TYPE_BITS;
-	bool ignored = flags & FAN_MARK_IGNORED_MASK;
+	unsigned int mark_cmd = flags & FANOTIFY_MARK_CMD_BITS;
+	unsigned int ignore = flags & FANOTIFY_MARK_IGNORE_BITS;
 	unsigned int obj_type, fid_mode;
 	u32 umask = 0;
 	int ret;
@@ -1466,7 +1627,7 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		return -EINVAL;
 	}
 
-	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE | FAN_MARK_FLUSH)) {
+	switch (mark_cmd) {
 	case FAN_MARK_ADD:
 	case FAN_MARK_REMOVE:
 		if (!mask)
@@ -1486,9 +1647,19 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	if (mask & ~valid_mask)
 		return -EINVAL;
 
-	/* Event flags (ONDIR, ON_CHILD) are meaningless in ignored mask */
-	if (ignored)
+
+	/* We don't allow FAN_MARK_IGNORE & FAN_MARK_IGNORED_MASK together */
+	if (ignore == (FAN_MARK_IGNORE | FAN_MARK_IGNORED_MASK))
+		return -EINVAL;
+
+	/*
+	 * Event flags (FAN_ONDIR, FAN_EVENT_ON_CHILD) have no effect with
+	 * FAN_MARK_IGNORED_MASK.
+	 */
+	if (ignore == FAN_MARK_IGNORED_MASK) {
 		mask &= ~FANOTIFY_EVENT_FLAGS;
+		umask = FANOTIFY_EVENT_FLAGS;
+	}
 
 	f = fdget(fanotify_fd);
 	if (unlikely(!f.file))
@@ -1525,6 +1696,14 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		goto fput_and_out;
 
 	/*
+	 * Evictable is only relevant for inode marks, because only inode object
+	 * can be evicted on memory pressure.
+	 */
+	if (flags & FAN_MARK_EVICTABLE &&
+	     mark_type != FAN_MARK_INODE)
+		goto fput_and_out;
+
+	/*
 	 * Events that do not carry enough information to report
 	 * event->fd require a group that supports reporting fid.  Those
 	 * events are not supported on a mount mark, because they do not
@@ -1536,7 +1715,15 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	    (!fid_mode || mark_type == FAN_MARK_MOUNT))
 		goto fput_and_out;
 
-	if (flags & FAN_MARK_FLUSH) {
+	/*
+	 * FAN_RENAME uses special info type records to report the old and
+	 * new parent+name.  Reporting only old and new parent id is less
+	 * useful and was not implemented.
+	 */
+	if (mask & FAN_RENAME && !(fid_mode & FAN_REPORT_NAME))
+		goto fput_and_out;
+
+	if (mark_cmd == FAN_MARK_FLUSH) {
 		ret = 0;
 		if (mark_type == FAN_MARK_MOUNT)
 			fsnotify_clear_vfsmount_marks_by_group(group);
@@ -1552,8 +1739,8 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	if (ret)
 		goto fput_and_out;
 
-	if (flags & FAN_MARK_ADD) {
-		ret = fanotify_events_supported(&path, mask);
+	if (mark_cmd == FAN_MARK_ADD) {
+		ret = fanotify_events_supported(group, &path, mask, flags);
 		if (ret)
 			goto path_put_and_out;
 	}
@@ -1576,6 +1763,13 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 	else
 		mnt = path.mnt;
 
+	ret = mnt ? -EINVAL : -EISDIR;
+	/* FAN_MARK_IGNORE requires SURV_MODIFY for sb/mount/dir marks */
+	if (mark_cmd == FAN_MARK_ADD && ignore == FAN_MARK_IGNORE &&
+	    (mnt || S_ISDIR(inode->i_mode)) &&
+	    !(flags & FAN_MARK_IGNORED_SURV_MODIFY))
+		goto path_put_and_out;
+
 	/* Mask out FAN_EVENT_ON_CHILD flag for sb/mount/non-dir marks */
 	if (mnt || !S_ISDIR(inode->i_mode)) {
 		mask &= ~FAN_EVENT_ON_CHILD;
@@ -1585,12 +1779,12 @@ static int do_fanotify_mark(int fanotify_fd, unsigned int flags, __u64 mask,
 		 * events with parent/name info for non-directory.
 		 */
 		if ((fid_mode & FAN_REPORT_DIR_FID) &&
-		    (flags & FAN_MARK_ADD) && !ignored)
+		    (flags & FAN_MARK_ADD) && !ignore)
 			mask |= FAN_EVENT_ON_CHILD;
 	}
 
 	/* create/update an inode mark */
-	switch (flags & (FAN_MARK_ADD | FAN_MARK_REMOVE)) {
+	switch (mark_cmd) {
 	case FAN_MARK_ADD:
 		if (mark_type == FAN_MARK_MOUNT)
 			ret = fanotify_add_vfsmount_mark(group, mnt, mask,
@@ -1667,8 +1861,8 @@ static int __init fanotify_user_setup(void)
 				     FANOTIFY_DEFAULT_MAX_USER_MARKS);
 
 	BUILD_BUG_ON(FANOTIFY_INIT_FLAGS & FANOTIFY_INTERNAL_GROUP_FLAGS);
-	BUILD_BUG_ON(HWEIGHT32(FANOTIFY_INIT_FLAGS) != 11);
-	BUILD_BUG_ON(HWEIGHT32(FANOTIFY_MARK_FLAGS) != 9);
+	BUILD_BUG_ON(HWEIGHT32(FANOTIFY_INIT_FLAGS) != 12);
+	BUILD_BUG_ON(HWEIGHT32(FANOTIFY_MARK_FLAGS) != 11);
 
 	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
 					 SLAB_PANIC|SLAB_ACCOUNT);
@@ -1685,6 +1879,7 @@ static int __init fanotify_user_setup(void)
 	init_user_ns.ucount_max[UCOUNT_FANOTIFY_GROUPS] =
 					FANOTIFY_DEFAULT_MAX_GROUPS;
 	init_user_ns.ucount_max[UCOUNT_FANOTIFY_MARKS] = max_marks;
+	fanotify_sysctls_init();
 
 	return 0;
 }