Skip to contentSkip to conversation

lockfs - vfs bits

PATCH In review Christoph Hellwig 3 participants 4 messages Apr 17, 2004 – Apr 19, 2004 VFS

lore
Share & tools
lore

Christoph Hellwig

author
#0 lore raw

PATCH 1/3 lockfs - vfs bits

These are the generic lockfs bits. Basically it takes the XFS freezing statemachine into the VFS. It's all behind the kernel-doc documented freeze_bdev and thaw_bdev interfaces.

Based on an older patch from Chris Mason.

fs/block_dev.c +1 -0
@@ -251,6 +251,7 @@
 	{
 		memset(bdev, 0, sizeof(*bdev));
 		sema_init(&bdev->bd_sem, 1);
+		sema_init(&bdev->bd_mount_sem, 1);
 		INIT_LIST_HEAD(&bdev->bd_inodes);
 		INIT_LIST_HEAD(&bdev->bd_list);
 		inode_init_once(&ei->vfs_inode);

===== fs/buffer.c 1.235 vs edited =====

fs/buffer.c +71 -0

77 lines not on this page · raw message

===== fs/super.c 1.116 vs edited =====

fs/super.c +8 -0
@@ -77,6 +77,7 @@
 		sema_init(&s->s_dquot.dqio_sem, 1);
 		sema_init(&s->s_dquot.dqonoff_sem, 1);
 		init_rwsem(&s->s_dquot.dqptr_sem);
+		init_waitqueue_head(&s->s_wait_unfrozen);
 		s->s_maxbytes = MAX_NON_LFS;
 		s->dq_op = sb_dquot_ops;
 		s->s_qcop = sb_quotactl_ops;

@@ -621,7 +622,14 @@
 	if (IS_ERR(bdev))
 		return (struct super_block *)bdev;
 
+	/*
+	 * once the super is inserted into the list by sget, s_umount
+	 * will protect the lockfs code from trying to start a snapshot
+	 * while we are mounting
+	 */
+	down(&bdev->bd_mount_sem);
 	s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
+	up(&bdev->bd_mount_sem);
 	if (IS_ERR(s))
 		goto out;
include/linux/buffer_head.h +2 -0
@@ -164,6 +164,8 @@
 wait_queue_head_t *bh_waitq_head(struct buffer_head *bh);
 void wake_up_buffer(struct buffer_head *bh);
 int fsync_bdev(struct block_device *);
+struct super_block *freeze_bdev(struct block_device *);
+void thaw_bdev(struct block_device *, struct super_block *);
 int fsync_super(struct super_block *);
 int fsync_no_super(struct block_device *);
 struct buffer_head *__find_get_block(struct block_device *, sector_t, int);
include/linux/fs.h +16 -0
@@ -345,6 +345,7 @@
 	struct inode *		bd_inode;	/* will die */
 	int			bd_openers;
 	struct semaphore	bd_sem;	/* open/close mutex */
+	struct semaphore	bd_mount_sem;	/* mount mutex */
 	struct list_head	bd_inodes;
 	void *			bd_holder;
 	int			bd_holders;

@@ -749,6 +750,9 @@
 	struct list_head	s_instances;
 	struct quota_info	s_dquot;	/* Diskquota specific options */
 
+	int			s_frozen;
+	wait_queue_head_t	s_wait_unfrozen;
+
 	char s_id[32];				/* Informational name */
 
 	struct kobject           kobj;          /* anchor for sysfs */

@@ -760,6 +764,18 @@
 	 */
 	struct semaphore s_vfs_rename_sem;	/* Kludge */
 };
+
+/*
+ * Snapshotting support.
+ */
+enum {
+	SB_UNFROZEN = 0,
+	SB_FREEZE_WRITE	= 1,
+	SB_FREEZE_TRANS = 2,
+};
+
+#define vfs_check_frozen(sb, level) \
+	wait_event((sb)->s_wait_unfrozen, ((sb)->s_frozen < (level)))
 
 /*
  * Superblock locking.
#1 lore raw
Quoted 4 lines · Christoph Hellwig

These are the generic lockfs bits. Basically it takes the XFS freezing statemachine into the VFS. It's all behind the kernel-doc documented freeze_bdev and thaw_bdev interfaces.

Do we expect to see snapshotting patches for other filesystems arise as a result of this?

author
#2 lore raw
Quoted 8 lines · Andrew Morton
Quoted 4 lines · Christoph Hellwig

These are the generic lockfs bits. Basically it takes the XFS freezing statemachine into the VFS. It's all behind the kernel-doc documented freeze_bdev and thaw_bdev interfaces.

Do we expect to see snapshotting patches for other filesystems arise as a result of this?

Other filesystems already implement the write_super_lockfs and unlockfs methods and should just work. An earlier version of this patch is in SuSE's tree and I think they've tested it with ext3 and reiserfs. Similarly even earlier variants are in the 2.4 vendor trees. Those filesystem don't really use the state machine to avoid starting new transactions, so to get results as reliable as XFS they need soem more work.

#3 lore raw
In reply to Andrew Morton
Quoted 8 lines · Andrew Morton
Quoted 4 lines · Christoph Hellwig

These are the generic lockfs bits. Basically it takes the XFS freezing statemachine into the VFS. It's all behind the kernel-doc documented freeze_bdev and thaw_bdev interfaces.

Do we expect to see snapshotting patches for other filesystems arise as a result of this?

Reiserfs needs this one liner:

reiserfs_write_super_lockfs() is supposed to wait for the transaction to commit.

fs/reiserfs/super.c +1 -1
Index: linux.t/fs/reiserfs/super.c
===================================================================
@@ -88,7 +88,7 @@ static void reiserfs_write_super_lockfs 
     reiserfs_prepare_for_journal(s, SB_BUFFER_WITH_SB(s), 1);
     journal_mark_dirty(&th, s, SB_BUFFER_WITH_SB (s));
     reiserfs_block_writes(&th) ;
-    journal_end(&th, s, 1) ;
+    journal_end_sync(&th, s, 1) ;
   }
   s->s_dirt = 0;
   reiserfs_write_unlock(s);