f2fs: fix fs corruption due to zero inode page
This patch fixes the following scenario.
- f2fs_create/f2fs_mkdir - write_checkpoint
- f2fs_mark_inode_dirty_sync - block_operations
- f2fs_lock_all
- f2fs_sync_inode_meta
- f2fs_unlock_all
- sync_inode_metadata
- f2fs_lock_op
- f2fs_write_inode
- update_inode_page
- get_node_page
return -ENOENT
- new_inode_page
- fill_node_footer
- f2fs_mark_inode_dirty_sync
- ...
- f2fs_unlock_op
- f2fs_inode_synced
- f2fs_lock_all
- do_checkpoint
In this checkpoint, we can get an inode page which contains zeros having valid
node footer only.
Cc: <stable@vger.kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index 2520fa7..0900814 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -316,7 +316,6 @@ int update_inode_page(struct inode *inode)
} else if (err != -ENOENT) {
f2fs_stop_checkpoint(sbi, false);
}
- f2fs_inode_synced(inode);
return 0;
}
ret = update_inode(inode, node_page);
@@ -450,6 +449,7 @@ void handle_failed_inode(struct inode *inode)
* in a panic when flushing dirty inodes in gdirty_list.
*/
update_inode_page(inode);
+ f2fs_inode_synced(inode);
/* don't make bad inode, since it becomes a regular file. */
unlock_new_inode(inode);
