f2fs: fix to avoid race in between atomic write and background GC

From: Chao Yu <[email protected]>

Sqlite user Background GC
- move_data_block : move page #1
- f2fs_is_atomic_file
- f2fs_ioc_start_atomic_write
- f2fs_ioc_commit_atomic_write
- commit_inmem_pages
: commit page #1 & set node #2 dirty
- f2fs_submit_page_write
- f2fs_update_data_blkaddr
- set_page_dirty : set node #2 dirty
- f2fs_do_sync_file
- fsync_node_pages
: commit node #1 & node #2, then sudden power-cut

In a race case, we may check FI_ATOMIC_FILE flag before starting atomic write flow, then we will commit meta data before data with reversed order, after a sudden pow-cut, database transaction will be inconsistent.

So we'd better to exclude gc/atomic_write to each other by using lock instead of flag checking.

---
 fs/f2fs/file.c | 3 +++
 1 file changed, 3 insertions(+)

index 4821ea97e080..d3d5ef6596f6 100644
@@ -1716,6 +1716,8 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp)
 
 	inode_lock(inode);
 
+	down_write(&F2FS_I(inode)->dio_rwsem[WRITE]);
+
 	if (f2fs_is_volatile_file(inode))
 		goto err_out;
 

@@ -1734,6 +1736,7 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp)
 		ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 1, false);
 	}
 err_out:
+	up_write(&F2FS_I(inode)->dio_rwsem[WRITE]);
 	inode_unlock(inode);
 	mnt_drop_write_file(filp);
 	return ret;

2.14.1.145.gb3622a4ee