Merge branch 'for-linus' of git://git.infradead.org/users/eparis/notify

* 'for-linus' of git://git.infradead.org/users/eparis/notify:
  fsnotify: drop two useless bools in the fnsotify main loop
  fsnotify: fix list walk order
  fanotify: Return EPERM when a process is not privileged
  fanotify: resize pid and reorder structure
  fanotify: drop duplicate pr_debug statement
  fanotify: flush outstanding perm requests on group destroy
  fsnotify: fix ignored mask handling between inode and vfsmount marks
  fanotify: add MAINTAINERS entry
  fsnotify: reset used_inode and used_vfsmount on each pass
  fanotify: do not dereference inode_mark when it is unset

6 files changed, 74 insertions, 46 deletions

diff --git a/MAINTAINERS b/MAINTAINERS
index 4523d6cb4c4..c36f5d76e1a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -2296,6 +2296,12 @@ S:	Maintained
 F:	Documentation/hwmon/f71805f
 F:	drivers/hwmon/f71805f.c
 
+FANOTIFY
+M:	Eric Paris <eparis@redhat.com>
+S:	Maintained
+F:	fs/notify/fanotify/
+F:	include/linux/fanotify.h
+
 FARSYNC SYNCHRONOUS DRIVER
 M:	Kevin Curtis <kevin.curtis@farsite.co.uk>
 W:	http://www.farsite.co.uk/
diff --git a/fs/notify/fanotify/fanotify.c b/fs/notify/fanotify/fanotify.c
index 756566fe844..85366c78cc3 100644
--- a/fs/notify/fanotify/fanotify.c
+++ b/fs/notify/fanotify/fanotify.c
@@ -165,9 +165,6 @@ static bool fanotify_should_send_event(struct fsnotify_group *group,
 		 "mask=%x data=%p data_type=%d\n", __func__, group, to_tell,
 		 inode_mark, vfsmnt_mark, event_mask, data, data_type);
 
-	pr_debug("%s: group=%p vfsmount_mark=%p inode_mark=%p mask=%x\n",
-		 __func__, group, vfsmnt_mark, inode_mark, event_mask);
-
 	/* sorry, fanotify only gives a damn about files and dirs */
 	if (!S_ISREG(to_tell->i_mode) &&
 	    !S_ISDIR(to_tell->i_mode))
diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c
index 032b837fcd1..5ed8e58d7bf 100644
--- a/fs/notify/fanotify/fanotify_user.c
+++ b/fs/notify/fanotify/fanotify_user.c
@@ -195,6 +195,14 @@ static int prepare_for_access_response(struct fsnotify_group *group,
 	re->fd = fd;
 
 	mutex_lock(&group->fanotify_data.access_mutex);
+
+	if (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);
 
@@ -364,9 +372,28 @@ static ssize_t fanotify_write(struct file *file, const char __user *buf, size_t
 static int fanotify_release(struct inode *ignored, struct file *file)
 {
 	struct fsnotify_group *group = file->private_data;
+	struct fanotify_response_event *re, *lre;
 
 	pr_debug("%s: file=%p group=%p\n", __func__, file, group);
 
+#ifdef CONFIG_FANOTIFY_ACCESS_PERMISSIONS
+	mutex_lock(&group->fanotify_data.access_mutex);
+
+	group->fanotify_data.bypass_perm = true;
+
+	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
 	/* matches the fanotify_init->fsnotify_alloc_group */
 	fsnotify_put_group(group);
 
@@ -614,7 +641,7 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags)
 		__func__, flags, event_f_flags);
 
 	if (!capable(CAP_SYS_ADMIN))
-		return -EACCES;
+		return -EPERM;
 
 	if (flags & ~FAN_ALL_INIT_FLAGS)
 		return -EINVAL;
diff --git a/fs/notify/fsnotify.c b/fs/notify/fsnotify.c
index 3970392b272..36802420d69 100644
--- a/fs/notify/fsnotify.c
+++ b/fs/notify/fsnotify.c
@@ -148,13 +148,14 @@ static int send_to_group(struct inode *to_tell, struct vfsmount *mnt,
 			 const unsigned char *file_name,
 			 struct fsnotify_event **event)
 {
-	struct fsnotify_group *group = inode_mark->group;
-	__u32 inode_test_mask = (mask & ~FS_EVENT_ON_CHILD);
-	__u32 vfsmount_test_mask = (mask & ~FS_EVENT_ON_CHILD);
+	struct fsnotify_group *group = NULL;
+	__u32 inode_test_mask = 0;
+	__u32 vfsmount_test_mask = 0;
 
-	pr_debug("%s: group=%p to_tell=%p mnt=%p mark=%p mask=%x data=%p"
-		 " data_is=%d cookie=%d event=%p\n", __func__, group, to_tell,
-		 mnt, inode_mark, mask, data, data_is, cookie, *event);
+	if (unlikely(!inode_mark && !vfsmount_mark)) {
+		BUG();
+		return 0;
+	}
 
 	/* clear ignored on inode modification */
 	if (mask & FS_MODIFY) {
@@ -168,18 +169,29 @@ static int send_to_group(struct inode *to_tell, struct vfsmount *mnt,
 
 	/* 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 mnt=%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, mnt, 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;
 
@@ -207,13 +219,12 @@ static int send_to_group(struct inode *to_tell, struct vfsmount *mnt,
 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, *vfsmount_node;
+	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 vfsmount *mnt;
 	int idx, ret = 0;
-	bool used_inode = false, used_vfsmount = false;
 	/* global tests shouldn't care about events on child only the specific event */
 	__u32 test_mask = (mask & ~FS_EVENT_ON_CHILD);
 
@@ -238,57 +249,50 @@ int fsnotify(struct inode *to_tell, __u32 mask, void *data, int data_is,
 	    (test_mask & to_tell->i_fsnotify_mask))
 		inode_node = srcu_dereference(to_tell->i_fsnotify_marks.first,
 					      &fsnotify_mark_srcu);
-	else
-		inode_node = NULL;
 
-	if (mnt) {
-		if ((mask & FS_MODIFY) ||
-		    (test_mask & mnt->mnt_fsnotify_mask))
-			vfsmount_node = srcu_dereference(mnt->mnt_fsnotify_marks.first,
-							 &fsnotify_mark_srcu);
-		else
-			vfsmount_node = NULL;
-	} else {
-		mnt = NULL;
-		vfsmount_node = NULL;
+	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;
-		} else
-			inode_group = (void *)-1;
+		}
 
 		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;
-		} else
-			vfsmount_group = (void *)-1;
+		}
 
-		if (inode_group < vfsmount_group) {
+		if (inode_group > vfsmount_group) {
 			/* handle inode */
 			send_to_group(to_tell, NULL, inode_mark, NULL, mask, data,
 				      data_is, cookie, file_name, &event);
-			used_inode = true;
-		} else if (vfsmount_group < inode_group) {
+			/* we didn't use the vfsmount_mark */
+			vfsmount_group = NULL;
+		} else if (vfsmount_group > inode_group) {
 			send_to_group(to_tell, mnt, NULL, vfsmount_mark, mask, data,
 				      data_is, cookie, file_name, &event);
-			used_vfsmount = true;
+			inode_group = NULL;
 		} else {
 			send_to_group(to_tell, mnt, inode_mark, vfsmount_mark,
 				      mask, data, data_is, cookie, file_name,
 				      &event);
-			used_vfsmount = true;
-			used_inode = true;
 		}
 
-		if (used_inode)
+		if (inode_group)
 			inode_node = srcu_dereference(inode_node->next,
 						      &fsnotify_mark_srcu);
-		if (used_vfsmount)
+		if (vfsmount_group)
 			vfsmount_node = srcu_dereference(vfsmount_node->next,
 							 &fsnotify_mark_srcu);
 	}
diff --git a/include/linux/fanotify.h b/include/linux/fanotify.h
index f0949a57ca9..63531a6b4d2 100644
--- a/include/linux/fanotify.h
+++ b/include/linux/fanotify.h
@@ -65,14 +65,14 @@
 				 FAN_ALL_PERM_EVENTS |\
 				 FAN_Q_OVERFLOW)
 
-#define FANOTIFY_METADATA_VERSION	1
+#define FANOTIFY_METADATA_VERSION	2
 
 struct fanotify_event_metadata {
 	__u32 event_len;
 	__u32 vers;
-	__s32 fd;
 	__u64 mask;
-	__s64 pid;
+	__s32 fd;
+	__s32 pid;
 } __attribute__ ((packed));
 
 struct fanotify_response {
@@ -95,11 +95,4 @@ struct fanotify_response {
 				(long)(meta)->event_len >= (long)FAN_EVENT_METADATA_LEN && \
 				(long)(meta)->event_len <= (long)(len))
 
-#ifdef __KERNEL__
-
-struct fanotify_wait {
-	struct fsnotify_event *event;
-	__s32 fd;
-};
-#endif /* __KERNEL__ */
 #endif /* _LINUX_FANOTIFY_H */
diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h
index ed36fb57c42..e40190d1687 100644
--- a/include/linux/fsnotify_backend.h
+++ b/include/linux/fsnotify_backend.h
@@ -156,6 +156,7 @@ struct fsnotify_group {
 			struct mutex access_mutex;
 			struct list_head access_list;
 			wait_queue_head_t access_waitq;
+			bool bypass_perm; /* protected by access_mutex */
 #endif /* CONFIG_FANOTIFY_ACCESS_PERMISSIONS */
 			int f_flags;
 		} fanotify_data;