filesystem freeze: fix sys_umount induced perpetual freeze

diff -urNp linux-2.6.31-rc7-orig/fs/namespace.c linux-2.6.31-rc7/fs/namespace.c
@@ -1033,6 +1033,8 @@ static int do_umount(struct vfsmount *mn
 	if (retval)
 		return retval;
 
+	vfs_check_frozen(sb, SB_FREEZE_WRITE);
+
 	/*
 	 * Allow userspace to request a mountpoint be expired rather than
 	 * unmounting unconditionally. Unmount only happens if:

diff -urNp linux-2.6.31-rc7-orig/fs/block_dev.c linux-2.6.31-rc7/fs/block_dev.c
@@ -216,8 +216,6 @@ EXPORT_SYMBOL(fsync_bdev);
  * freeze_bdev  --  lock a filesystem and force it into a consistent state
  * @bdev:	blockdevice to lock
  *
- * This takes the block device bd_mount_sem to make sure no new mounts
- * happen on bdev until thaw_bdev() is called.
  * If a superblock is found on this device, we take the s_umount semaphore
  * on it to make sure nobody unmounts until the snapshot creation is done.
  * The reference counter (bd_fsfreeze_count) guarantees that only the last

@@ -240,7 +238,6 @@ struct super_block *freeze_bdev(struct b
 	}
 	bdev->bd_fsfreeze_count++;
 
-	down(&bdev->bd_mount_sem);
 	sb = get_super(bdev);
 	if (sb && !(sb->s_flags & MS_RDONLY)) {
 		sb->s_frozen = SB_FREEZE_WRITE;

@@ -260,7 +257,6 @@ struct super_block *freeze_bdev(struct b
 					"VFS:Filesystem freeze failed\n");
 				sb->s_frozen = SB_UNFROZEN;
 				drop_super(sb);
-				up(&bdev->bd_mount_sem);
 				bdev->bd_fsfreeze_count--;
 				mutex_unlock(&bdev->bd_fsfreeze_mutex);
 				return ERR_PTR(error);

@@ -271,7 +267,7 @@ struct super_block *freeze_bdev(struct b
 	sync_blockdev(bdev);
 	mutex_unlock(&bdev->bd_fsfreeze_mutex);
 
-	return sb;	/* thaw_bdev releases s->s_umount and bd_mount_sem */
+	return sb;	/* thaw_bdev releases s->s_umount */
 }
 EXPORT_SYMBOL(freeze_bdev);
 

@@ -321,7 +317,6 @@ int thaw_bdev(struct block_device *bdev,
 		drop_super(sb);
 	}
 
-	up(&bdev->bd_mount_sem);
 	mutex_unlock(&bdev->bd_fsfreeze_mutex);
 	return 0;
 }

@@ -431,7 +426,6 @@ static void init_once(void *foo)
 
 	memset(bdev, 0, sizeof(*bdev));
 	mutex_init(&bdev->bd_mutex);
-	sema_init(&bdev->bd_mount_sem, 1);
 	INIT_LIST_HEAD(&bdev->bd_inodes);
 	INIT_LIST_HEAD(&bdev->bd_list);
 #ifdef CONFIG_SYSFS

diff -urNp linux-2.6.31-rc7-orig/fs/super.c linux-2.6.31-rc7/fs/super.c
@@ -740,9 +740,14 @@ int get_sb_bdev(struct file_system_type
 	 * will protect the lockfs code from trying to start a snapshot
 	 * while we are mounting
 	 */
-	down(&bdev->bd_mount_sem);
+	mutex_lock(&bdev->bd_fsfreeze_mutex);
+	if (bdev->bd_fsfreeze_count > 0) {
+		mutex_unlock(&bdev->bd_fsfreeze_mutex);
+		error = -EBUSY;
+		goto error_bdev;
+	}
 	s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
-	up(&bdev->bd_mount_sem);
+	mutex_unlock(&bdev->bd_fsfreeze_mutex);
 	if (IS_ERR(s))
 		goto error_s;
 

diff -urNp linux-2.6.31-rc7-orig/include/linux/fs.h linux-2.6.31-rc7/include/linux/fs.h
@@ -640,7 +640,6 @@ struct block_device {
 	struct super_block *	bd_super;
 	int			bd_openers;
 	struct mutex		bd_mutex;	/* open/close mutex */
-	struct semaphore	bd_mount_sem;
 	struct list_head	bd_inodes;
 	void *			bd_holder;
 	int			bd_holders;

diff -urNp linux-2.6.31-rc7-orig/fs/block_dev.c linux-2.6.31-rc7/fs/block_dev.c
@@ -230,43 +230,54 @@ struct super_block *freeze_bdev(struct b
 	int error = 0;
 
 	mutex_lock(&bdev->bd_fsfreeze_mutex);
-	if (bdev->bd_fsfreeze_count > 0) {
-		bdev->bd_fsfreeze_count++;
+	if (++bdev->bd_fsfreeze_count > 1) {
+		/*
+		 * We don't even need to grab a reference - the first call
+		 * to freeze_bdev grab an active reference and only the last
+		 * thaw_bdev drops it.
+		 */
 		sb = get_super(bdev);
+		drop_super(sb);
 		mutex_unlock(&bdev->bd_fsfreeze_mutex);
 		return sb;
 	}
-	bdev->bd_fsfreeze_count++;
 
-	sb = get_super(bdev);
-	if (sb && !(sb->s_flags & MS_RDONLY)) {
-		sb->s_frozen = SB_FREEZE_WRITE;
-		smp_wmb();
-
-		sync_filesystem(sb);
-
-		sb->s_frozen = SB_FREEZE_TRANS;
-		smp_wmb();
-
-		sync_blockdev(sb->s_bdev);
-
-		if (sb->s_op->freeze_fs) {
-			error = sb->s_op->freeze_fs(sb);
-			if (error) {
-				printk(KERN_ERR
-					"VFS:Filesystem freeze failed\n");
-				sb->s_frozen = SB_UNFROZEN;
-				drop_super(sb);
-				bdev->bd_fsfreeze_count--;
-				mutex_unlock(&bdev->bd_fsfreeze_mutex);
-				return ERR_PTR(error);
-			}
+	sb = get_active_super(bdev);
+	if (!sb)
+		goto out;
+	if (sb->s_flags & MS_RDONLY) {
+		deactivate_locked_super(sb);
+		mutex_unlock(&bdev->bd_fsfreeze_mutex);
+		return sb;
+	}
+
+	sb->s_frozen = SB_FREEZE_WRITE;
+	smp_wmb();
+
+	sync_filesystem(sb);
+
+	sb->s_frozen = SB_FREEZE_TRANS;
+	smp_wmb();
+
+	sync_blockdev(sb->s_bdev);
+
+	if (sb->s_op->freeze_fs) {
+		error = sb->s_op->freeze_fs(sb);
+		if (error) {
+			printk(KERN_ERR
+				"VFS:Filesystem freeze failed\n");
+			sb->s_frozen = SB_UNFROZEN;
+			deactivate_locked_super(sb);
+			bdev->bd_fsfreeze_count--;
+			mutex_unlock(&bdev->bd_fsfreeze_mutex);
+			return ERR_PTR(error);
 		}
 	}
+	up_write(&sb->s_umount);
 
+ out:
 	sync_blockdev(bdev);
 	mutex_unlock(&bdev->bd_fsfreeze_mutex);
-
 	return sb;	/* thaw_bdev releases s->s_umount */
 }
 EXPORT_SYMBOL(freeze_bdev);

@@ -280,43 +291,44 @@ EXPORT_SYMBOL(freeze_bdev);
  */
 int thaw_bdev(struct block_device *bdev, struct super_block *sb)
 {
-	int error = 0;
+	int error = -EINVAL;
 
 	mutex_lock(&bdev->bd_fsfreeze_mutex);
-	if (!bdev->bd_fsfreeze_count) {
-		mutex_unlock(&bdev->bd_fsfreeze_mutex);
-		return -EINVAL;
-	}
+	if (!bdev->bd_fsfreeze_count)
+		goto out_unlock;
 
-	bdev->bd_fsfreeze_count--;
-	if (bdev->bd_fsfreeze_count > 0) {
-		if (sb)
-			drop_super(sb);
-		mutex_unlock(&bdev->bd_fsfreeze_mutex);
-		return 0;
-	}
-
-	if (sb) {
-		BUG_ON(sb->s_bdev != bdev);
-		if (!(sb->s_flags & MS_RDONLY)) {
-			if (sb->s_op->unfreeze_fs) {
-				error = sb->s_op->unfreeze_fs(sb);
-				if (error) {
-					printk(KERN_ERR
-						"VFS:Filesystem thaw failed\n");
-					sb->s_frozen = SB_FREEZE_TRANS;
-					bdev->bd_fsfreeze_count++;
-					mutex_unlock(&bdev->bd_fsfreeze_mutex);
-					return error;
-				}
-			}
-			sb->s_frozen = SB_UNFROZEN;
-			smp_wmb();
-			wake_up(&sb->s_wait_unfrozen);
+	error = 0;
+	if (--bdev->bd_fsfreeze_count > 0)
+		goto out_unlock;
+
+	if (!sb)
+		goto out_unlock;
+
+	BUG_ON(sb->s_bdev != bdev);
+	down_write(&sb->s_umount);
+	if (sb->s_flags & MS_RDONLY)
+		goto out_deactivate;
+
+	if (sb->s_op->unfreeze_fs) {
+		error = sb->s_op->unfreeze_fs(sb);
+		if (error) {
+			printk(KERN_ERR
+				"VFS:Filesystem thaw failed\n");
+			sb->s_frozen = SB_FREEZE_TRANS;
+			bdev->bd_fsfreeze_count++;
+			mutex_unlock(&bdev->bd_fsfreeze_mutex);
+			return error;
 		}
-		drop_super(sb);
 	}
 
+	sb->s_frozen = SB_UNFROZEN;
+	smp_wmb();
+	wake_up(&sb->s_wait_unfrozen);
+
+out_deactivate:
+	if (sb)
+		deactivate_locked_super(sb);
+out_unlock:
 	mutex_unlock(&bdev->bd_fsfreeze_mutex);
 	return 0;
 }

64 lines not on this page · raw message

diff -urNp linux-2.6.31-rc7-orig/include/linux/fs.h linux-2.6.31-rc7/include/linux/fs.h
@@ -2320,6 +2320,7 @@ extern void get_filesystem(struct file_s
 extern void put_filesystem(struct file_system_type *fs);
 extern struct file_system_type *get_fs_type(const char *name);
 extern struct super_block *get_super(struct block_device *);
+extern struct super_block *get_active_super(struct block_device *bdev);
 extern struct super_block *user_get_super(dev_t);
 extern void drop_super(struct super_block *sb);
 

diff -urNp linux-2.6.31-rc7-orig/fs/namespace.c linux-2.6.31-rc7/fs/namespace.c
@@ -1086,6 +1086,14 @@ static int do_umount(struct vfsmount *mn
 		return retval;
 	}
 
+	if (sb->s_bdev != NULL) {
+		mutex_lock(&sb->s_bdev->bd_fsfreeze_mutex);
+		if (sb->s_frozen != SB_UNFROZEN) {
+			mutex_unlock(&sb->s_bdev->bd_fsfreeze_mutex);
+			return -EBUSY;
+		}
+	}
+
 	down_write(&namespace_sem);
 	spin_lock(&vfsmount_lock);
 	event++;

@@ -1104,6 +1112,10 @@ static int do_umount(struct vfsmount *mn
 		security_sb_umount_busy(mnt);
 	up_write(&namespace_sem);
 	release_mounts(&umount_list);
+
+	if (sb->s_bdev != NULL)
+		mutex_unlock(&sb->s_bdev->bd_fsfreeze_mutex);
+
 	return retval;
 }
 

diff -urNp linux-2.6.31-rc7-orig/fs/block_dev.c linux-2.6.31-rc7/fs/block_dev.c
@@ -334,6 +334,21 @@ out_unlock:
 }
 EXPORT_SYMBOL(thaw_bdev);
 
+/**
+ *
+ */
+int isfrozen_bdev(struct block_device *bdev, struct super_block *sb)
+{
+	int ret;
+
+	mutex_lock(&bdev->bd_fsfreeze_mutex);
+	ret = sb->s_frozen > SB_UNFROZEN ? 1:0;
+	mutex_unlock(&bdev->bd_fsfreeze_mutex);
+
+	return ret;
+}
+EXPORT_SYMBOL(isfrozen_bdev);
+
 static int blkdev_writepage(struct page *page, struct writeback_control *wbc)
 {
 	return block_write_full_page(page, blkdev_get_block, wbc);

diff -urNp linux-2.6.31-rc7-orig/fs/ioctl.c linux-2.6.31-rc7/fs/ioctl.c
@@ -536,6 +536,20 @@ static int ioctl_fsthaw(struct file *fil
 	return thaw_bdev(sb->s_bdev, sb);
 }
 
+static int ioctl_isfrozen(struct file *filp)
+{
+	struct super_block *sb = filp->f_path.dentry->d_inode->i_sb;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	/* If a blockdevice-backed filesystem isn't specified, return EINVAL. */
+	if (sb->s_bdev == NULL)
+		return -EINVAL;
+
+	return isfrozen_bdev(sb->s_bdev, sb);
+}
+
 /*
  * When you add any new common ioctls to the switches above and below
  * please update compat_sys_ioctl() too.

@@ -586,6 +600,12 @@ int do_vfs_ioctl(struct file *filp, unsi
 		error = ioctl_fsthaw(filp);
 		break;
 
+	case FIISFROZEN:
+		error = ioctl_isfrozen(filp);
+		if (error >= 0)
+			return put_user(error, (int __user *)arg);
+		break;
+
 	case FS_IOC_FIEMAP:
 		return ioctl_fiemap(filp, arg);
 

diff -urNp linux-2.6.31-rc7-orig/include/linux/fs.h linux-2.6.31-rc7/include/linux/fs.h
@@ -306,6 +306,7 @@ struct inodes_stat_t {
 #define FIGETBSZ   _IO(0x00,2)	/* get the block size used for bmap */
 #define FIFREEZE	_IOWR('X', 119, int)	/* Freeze */
 #define FITHAW		_IOWR('X', 120, int)	/* Thaw */
+#define FIISFROZEN	_IOWR('X', 121, int)	/* Is frozen? */
 
 #define	FS_IOC_GETFLAGS			_IOR('f', 1, long)
 #define	FS_IOC_SETFLAGS			_IOW('f', 2, long)

@@ -1955,6 +1956,7 @@ extern int sync_blockdev(struct block_de
 extern struct super_block *freeze_bdev(struct block_device *);
 extern void emergency_thaw_all(void);
 extern int thaw_bdev(struct block_device *bdev, struct super_block *sb);
+extern int isfrozen_bdev(struct block_device *bdev, struct super_block *sb);
 extern int fsync_bdev(struct block_device *);
 #else
 static inline void bd_forget(struct inode *inode) {}

@@ -1970,6 +1972,12 @@ static inline int thaw_bdev(struct block
 {
 	return 0;
 }
+
+extern int isfrozen_bdev(struct block_device *bdev, struct super_block *sb)
+{
+	return 0;
+}
+
 #endif
 extern int sync_filesystem(struct super_block *);
 extern const struct file_operations def_blk_fops;