NFS: Don't hold the inode lock across fsync()

Commits are no longer required to be serialised.

Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>

fs/nfs/file.c

@@ -277,11 +277,9 @@ nfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 		ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
 		if (ret != 0)
 			break;
-		inode_lock(inode);
 		ret = nfs_file_fsync_commit(file, start, end, datasync);
 		if (!ret)
 			ret = pnfs_sync_inode(inode, !!datasync);
-		inode_unlock(inode);
 		/*
 		 * If nfs_file_fsync_commit detected a server reboot, then
 		 * resend all dirty pages that might have been covered by