linux/fs/hfsplus
Chao Yu 312d272dc4 hfsplus: fix to avoid false alarm of circular locking
[ Upstream commit be4edd1642 ]

Syzbot report potential ABBA deadlock as below:

loop0: detected capacity change from 0 to 1024
======================================================
WARNING: possible circular locking dependency detected
6.9.0-syzkaller-10323-g8f6a15f095a6 #0 Not tainted
------------------------------------------------------
syz-executor171/5344 is trying to acquire lock:
ffff88807cb980b0 (&tree->tree_lock){+.+.}-{3:3}, at: hfsplus_file_truncate+0x811/0xb50 fs/hfsplus/extents.c:595

but task is already holding lock:
ffff88807a930108 (&HFSPLUS_I(inode)->extents_lock){+.+.}-{3:3}, at: hfsplus_file_truncate+0x2da/0xb50 fs/hfsplus/extents.c:576

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (&HFSPLUS_I(inode)->extents_lock){+.+.}-{3:3}:
       lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
       __mutex_lock_common kernel/locking/mutex.c:608 [inline]
       __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
       hfsplus_file_extend+0x21b/0x1b70 fs/hfsplus/extents.c:457
       hfsplus_bmap_reserve+0x105/0x4e0 fs/hfsplus/btree.c:358
       hfsplus_rename_cat+0x1d0/0x1050 fs/hfsplus/catalog.c:456
       hfsplus_rename+0x12e/0x1c0 fs/hfsplus/dir.c:552
       vfs_rename+0xbdb/0xf00 fs/namei.c:4887
       do_renameat2+0xd94/0x13f0 fs/namei.c:5044
       __do_sys_rename fs/namei.c:5091 [inline]
       __se_sys_rename fs/namei.c:5089 [inline]
       __x64_sys_rename+0x86/0xa0 fs/namei.c:5089
       do_syscall_x64 arch/x86/entry/common.c:52 [inline]
       do_syscall_64+0xf5/0x240 arch/x86/entry/common.c:83
       entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #0 (&tree->tree_lock){+.+.}-{3:3}:
       check_prev_add kernel/locking/lockdep.c:3134 [inline]
       check_prevs_add kernel/locking/lockdep.c:3253 [inline]
       validate_chain+0x18cb/0x58e0 kernel/locking/lockdep.c:3869
       __lock_acquire+0x1346/0x1fd0 kernel/locking/lockdep.c:5137
       lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
       __mutex_lock_common kernel/locking/mutex.c:608 [inline]
       __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
       hfsplus_file_truncate+0x811/0xb50 fs/hfsplus/extents.c:595
       hfsplus_setattr+0x1ce/0x280 fs/hfsplus/inode.c:265
       notify_change+0xb9d/0xe70 fs/attr.c:497
       do_truncate+0x220/0x310 fs/open.c:65
       handle_truncate fs/namei.c:3308 [inline]
       do_open fs/namei.c:3654 [inline]
       path_openat+0x2a3d/0x3280 fs/namei.c:3807
       do_filp_open+0x235/0x490 fs/namei.c:3834
       do_sys_openat2+0x13e/0x1d0 fs/open.c:1406
       do_sys_open fs/open.c:1421 [inline]
       __do_sys_creat fs/open.c:1497 [inline]
       __se_sys_creat fs/open.c:1491 [inline]
       __x64_sys_creat+0x123/0x170 fs/open.c:1491
       do_syscall_x64 arch/x86/entry/common.c:52 [inline]
       do_syscall_64+0xf5/0x240 arch/x86/entry/common.c:83
       entry_SYSCALL_64_after_hwframe+0x77/0x7f

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(&HFSPLUS_I(inode)->extents_lock);
                               lock(&tree->tree_lock);
                               lock(&HFSPLUS_I(inode)->extents_lock);
  lock(&tree->tree_lock);

This is a false alarm as tree_lock mutex are different, one is
from sbi->cat_tree, and another is from sbi->ext_tree:

Thread A			Thread B
- hfsplus_rename
 - hfsplus_rename_cat
  - hfs_find_init
   - mutext_lock(cat_tree->tree_lock)
				- hfsplus_setattr
				 - hfsplus_file_truncate
				  - mutex_lock(hip->extents_lock)
				  - hfs_find_init
				   - mutext_lock(ext_tree->tree_lock)
  - hfs_bmap_reserve
   - hfsplus_file_extend
    - mutex_lock(hip->extents_lock)

So, let's call mutex_lock_nested for tree_lock mutex lock, and pass
correct lock class for it.

Fixes: 31651c6071 ("hfsplus: avoid deadlock on file truncation")
Reported-by: syzbot+6030b3b1b9bf70e538c4@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-fsdevel/000000000000e37a4005ef129563@google.com
Cc: Ernesto A. Fernández <ernesto.mnd.fernandez@gmail.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/r/20240607142304.455441-1-chao@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
2024-08-03 08:59:11 +02:00
..
attributes.c hfsplus: fix crash and filesystem corruption when deleting files 2020-04-10 15:36:20 -07:00
bfind.c hfsplus: fix to avoid false alarm of circular locking 2024-08-03 08:59:11 +02:00
bitmap.c hfsplus: convert kmap() to kmap_local_page() in bitmap.c 2022-09-11 21:55:04 -07:00
bnode.c hfsplus: convert kmap() to kmap_local_page() in bnode.c 2022-09-11 21:55:04 -07:00
brec.c hfsplus: fix BUG on bnode parent update 2018-10-31 08:54:13 -07:00
btree.c hfsplus: convert kmap() to kmap_local_page() in btree.c 2022-09-11 21:55:05 -07:00
catalog.c hfsplus: convert to new timestamp accessors 2023-10-18 14:08:22 +02:00
dir.c hfsplus: convert to ctime accessor functions 2023-07-24 10:30:00 +02:00
extents.c hfsplus: fix to avoid false alarm of circular locking 2024-08-03 08:59:11 +02:00
hfsplus_fs.h hfsplus: fix to avoid false alarm of circular locking 2024-08-03 08:59:11 +02:00
hfsplus_raw.h hfsplus: use struct_group_attr() for memcpy() region 2022-01-20 08:52:54 +02:00
inode.c hfsplus: really remove hfsplus_writepage 2023-12-29 11:58:34 -08:00
ioctl.c hfsplus: convert to fileattr 2021-04-12 15:04:29 +02:00
Kconfig fs: add CONFIG_BUFFER_HEAD 2023-08-02 09:13:09 -06:00
Makefile hfsplus: drop ACL support 2018-08-22 10:52:50 -07:00
options.c hfsplus: fix bug causing custom uid and gid being unable to be assigned with mount 2022-12-11 19:30:20 -08:00
part_tbl.c fs/hfsplus: Use the enum req_op and blk_opf_t types 2022-07-14 12:14:32 -06:00
super.c hfsplus: switch to rcu-delayed unloading of nls and freeing ->s_fs_info 2024-02-25 02:10:31 -05:00
tables.c License cleanup: add SPDX GPL-2.0 license identifier to files with no license 2017-11-02 11:10:55 +01:00
unicode.c treewide: Remove uninitialized_var() usage 2020-07-16 12:35:15 -07:00
wrapper.c fs/hfsplus: use better @opf description 2024-02-12 11:55:13 +01:00
xattr_security.c fs: port xattr to mnt_idmap 2023-01-19 09:24:28 +01:00
xattr_trusted.c fs: port xattr to mnt_idmap 2023-01-19 09:24:28 +01:00
xattr_user.c fs: port xattr to mnt_idmap 2023-01-19 09:24:28 +01:00
xattr.c hfsplus: fix uninit-value in copy_name 2024-07-06 09:19:31 +02:00
xattr.h hfsplus: move hfsplus_xattr_handlers to .rodata 2023-10-09 16:24:19 +02:00