[PATCH v2 3/3] hfs: add lock nesting notation to hfs_find_init

[Viacheslav Dubeyko]

> On Jun 30, 2021, at 8:07 PM, Desmond Cheong Zhi Xi <desmondcheongzx at gmail.com> wrote:
> 
> Syzbot reports a possible recursive lock:
> https://syzkaller.appspot.com/bug?id=f007ef1d7a31a469e3be7aeb0fde0769b18585db
> 
> This happens due to missing lock nesting information. From the logs,
> we see that a call to hfs_fill_super is made to mount the hfs
> filesystem. While searching for the root inode, the lock on the
> catalog btree is grabbed. Then, when the parent of the root isn't
> found, a call to __hfs_bnode_create is made to create the parent of
> the root. This eventually leads to a call to hfs_ext_read_extent which
> grabs a lock on the extents btree.
> 
> Since the order of locking is catalog btree -> extents btree, this
> lock hierarchy does not lead to a deadlock.
> 
> To tell lockdep that this locking is safe, we add nesting notation to
> distinguish between catalog btrees, extents btrees, and attributes
> btrees (for HFS+). This has already been done in hfsplus.
> 
> Reported-and-tested-by: syzbot+b718ec84a87b7e73ade4 at syzkaller.appspotmail.com
> Signed-off-by: Desmond Cheong Zhi Xi <desmondcheongzx at gmail.com>
> Reviewed-by: Viacheslav Dubeyko <slava at dubeyko.com>
> ---
> fs/hfs/bfind.c | 14 +++++++++++++-
> fs/hfs/btree.h |  7 +++++++
> 2 files changed, 20 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/hfs/bfind.c b/fs/hfs/bfind.c
> index 4af318fbda77..ef9498a6e88a 100644
> --- a/fs/hfs/bfind.c
> +++ b/fs/hfs/bfind.c
> @@ -25,7 +25,19 @@ int hfs_find_init(struct hfs_btree *tree, struct hfs_find_data *fd)
> 	fd->key = ptr + tree->max_key_len + 2;
> 	hfs_dbg(BNODE_REFS, "find_init: %d (%p)\n",
> 		tree->cnid, __builtin_return_address(0));
> -	mutex_lock(&tree->tree_lock);
> +	switch (tree->cnid) {
> +	case HFS_CAT_CNID:
> +		mutex_lock_nested(&tree->tree_lock, CATALOG_BTREE_MUTEX);
> +		break;
> +	case HFS_EXT_CNID:
> +		mutex_lock_nested(&tree->tree_lock, EXTENTS_BTREE_MUTEX);
> +		break;
> +	case HFS_ATTR_CNID:
> +		mutex_lock_nested(&tree->tree_lock, ATTR_BTREE_MUTEX);
> +		break;
> +	default:
> +		return -EINVAL;
> +	}
> 	return 0;
> }
> 
> diff --git a/fs/hfs/btree.h b/fs/hfs/btree.h
> index 4ba45caf5939..0e6baee93245 100644
> --- a/fs/hfs/btree.h
> +++ b/fs/hfs/btree.h
> @@ -13,6 +13,13 @@ typedef int (*btree_keycmp)(const btree_key *, const btree_key *);
> 
> #define NODE_HASH_SIZE  256
> 
> +/* B-tree mutex nested subclasses */
> +enum hfs_btree_mutex_classes {
> +	CATALOG_BTREE_MUTEX,
> +	EXTENTS_BTREE_MUTEX,
> +	ATTR_BTREE_MUTEX,
> +};
> +
> /* A HFS BTree held in memory */
> struct hfs_btree {
> 	struct super_block *sb;
> -- 
> 2.25.1
> 

Looks good.

Reviewed-by: Viacheslav Dubeyko <slava at dubeyko.com>

Thanks,
Slava.