Node 0 is the header node in HFS+ B-trees and should always be allocated. However, if a filesystem image has node 0's bitmap bit unset (e.g., due to corruption or a buggy image generator), hfs_bmap_alloc() will find node 0 as free and attempt to allocate it. This causes a conflict because node 0 already exists as the header node, leading to a WARN_ON(1) in hfs_bnode_create() when the node is found already hashed. This issue can occur with syzkaller-generated HFS+ images or corrupted real-world filesystems. Add a guard in hfs_bmap_alloc() to skip node 0 during allocation, providing defense-in-depth against such corruption. Reported-by: syzbot+1c8ff72d0cd8a50dfeaa@syzkaller.appspotmail.com Link: https://syzkaller.appspot.com/bug?extid=1c8ff72d0cd8a50dfeaa Signed-off-by: Shardul Bankar --- v2: - Keep the node-0 allocation guard as targeted hardening for corrupted images. fs/hfsplus/btree.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/hfsplus/btree.c b/fs/hfsplus/btree.c index 229f25dc7c49..60985f449450 100644 --- a/fs/hfsplus/btree.c +++ b/fs/hfsplus/btree.c @@ -411,6 +411,9 @@ struct hfs_bnode *hfs_bmap_alloc(struct hfs_btree *tree) if (byte != 0xff) { for (m = 0x80, i = 0; i < 8; m >>= 1, i++) { if (!(byte & m)) { + /* Skip node 0 (header node, always allocated) */ + if (idx == 0 && i == 0) + continue; idx += i; data[off] |= m; set_page_dirty(*pagep); -- 2.34.1 When hfs_bnode_create() finds that a node is already hashed (which should not happen in normal operation), it currently returns the existing node without incrementing its reference count. This causes a reference count inconsistency that leads to a kernel panic when the node is later freed in hfs_bnode_put(): kernel BUG at fs/hfsplus/bnode.c:676! BUG_ON(!atomic_read(&node->refcnt)) This scenario can occur when hfs_bmap_alloc() attempts to allocate a node that is already in use (e.g., when node 0's bitmap bit is incorrectly unset), or due to filesystem corruption. Returning an existing node from a create path is not normal operation. Fix this by returning ERR_PTR(-EEXIST) instead of the node when it's already hashed. This properly signals the error condition to callers, which already check for IS_ERR() return values. Reported-by: syzbot+1c8ff72d0cd8a50dfeaa@syzkaller.appspotmail.com Link: https://syzkaller.appspot.com/bug?extid=1c8ff72d0cd8a50dfeaa Fixes: 634725a92938 ("[PATCH] hfs: cleanup HFS+ prints") Signed-off-by: Shardul Bankar --- v2: - Return ERR_PTR(-EEXIST) for already-hashed nodes, per Slava's suggestion. fs/hfsplus/bnode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/hfsplus/bnode.c b/fs/hfsplus/bnode.c index 191661af9677..250a226336ea 100644 --- a/fs/hfsplus/bnode.c +++ b/fs/hfsplus/bnode.c @@ -629,7 +629,7 @@ struct hfs_bnode *hfs_bnode_create(struct hfs_btree *tree, u32 num) if (node) { pr_crit("new node %u already hashed?\n", num); WARN_ON(1); - return node; + return ERR_PTR(-EEXIST); } node = __hfs_bnode_create(tree, num); if (!node) -- 2.34.1