syzbot


possible deadlock in chown_common

Status: upstream: reported C repro on 2022/10/05 17:28
Reported-by: syzbot+197c713b38a631a8ba6c@syzkaller.appspotmail.com
First crash: 125d, last: 16d
similar bugs (1):
Kernel Title Repro Cause bisect Fix bisect Count Last Reported Patched Status
upstream possible deadlock in chown_common C error 2 12d 121d 0/24 upstream: reported C repro on 2022/10/10 03:09

Sample crash report:
REISERFS (device loop0): journal params: device loop0, size 512, journal first block 18, max trans len 256, max batch 225, max commit age 30, max trans age 30
REISERFS (device loop0): checking transaction log (loop0)
REISERFS (device loop0): Using rupasov hash to sort names
REISERFS (device loop0): Created .reiserfs_priv - reserved for xattr storage.
======================================================
WARNING: possible circular locking dependency detected
4.14.295-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor380/7979 is trying to acquire lock:
 (&type->i_mutex_dir_key#7){+.+.}, at: [<ffffffff818644e2>] inode_lock include/linux/fs.h:719 [inline]
 (&type->i_mutex_dir_key#7){+.+.}, at: [<ffffffff818644e2>] chown_common+0x272/0x4b0 fs/open.c:628

but task is already holding lock:
 (sb_writers#10){.+.+}, at: [<ffffffff818e083a>] sb_start_write include/linux/fs.h:1551 [inline]
 (sb_writers#10){.+.+}, at: [<ffffffff818e083a>] mnt_want_write+0x3a/0xb0 fs/namespace.c:386

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #2 (sb_writers#10){.+.+}:
       percpu_down_read_preempt_disable include/linux/percpu-rwsem.h:36 [inline]
       percpu_down_read include/linux/percpu-rwsem.h:59 [inline]
       __sb_start_write+0x64/0x260 fs/super.c:1342
       sb_start_write include/linux/fs.h:1551 [inline]
       mnt_want_write_file+0xfd/0x3b0 fs/namespace.c:497
       reiserfs_ioctl+0x18e/0x8b0 fs/reiserfs/ioctl.c:110
       vfs_ioctl fs/ioctl.c:46 [inline]
       file_ioctl fs/ioctl.c:500 [inline]
       do_vfs_ioctl+0x75a/0xff0 fs/ioctl.c:684
       SYSC_ioctl fs/ioctl.c:701 [inline]
       SyS_ioctl+0x7f/0xb0 fs/ioctl.c:692
       do_syscall_64+0x1d5/0x640 arch/x86/entry/common.c:292
       entry_SYSCALL_64_after_hwframe+0x46/0xbb

-> #1 (&sbi->lock){+.+.}:
       __mutex_lock_common kernel/locking/mutex.c:756 [inline]
       __mutex_lock+0xc4/0x1310 kernel/locking/mutex.c:893
       reiserfs_write_lock+0x75/0xf0 fs/reiserfs/lock.c:27
       reiserfs_lookup+0x130/0x400 fs/reiserfs/namei.c:363
       lookup_real fs/namei.c:1555 [inline]
       __lookup_hash fs/namei.c:1575 [inline]
       __lookup_hash+0x1bb/0x270 fs/namei.c:1563
       lookup_one_len+0x279/0x3a0 fs/namei.c:2539
       reiserfs_lookup_privroot+0x92/0x270 fs/reiserfs/xattr.c:970
       reiserfs_fill_super+0x1d12/0x2990 fs/reiserfs/super.c:2187
       mount_bdev+0x2b3/0x360 fs/super.c:1134
       mount_fs+0x92/0x2a0 fs/super.c:1237
       vfs_kern_mount.part.0+0x5b/0x470 fs/namespace.c:1046
       vfs_kern_mount fs/namespace.c:1036 [inline]
       do_new_mount fs/namespace.c:2572 [inline]
       do_mount+0xe65/0x2a30 fs/namespace.c:2905
       SYSC_mount fs/namespace.c:3121 [inline]
       SyS_mount+0xa8/0x120 fs/namespace.c:3098
       do_syscall_64+0x1d5/0x640 arch/x86/entry/common.c:292
       entry_SYSCALL_64_after_hwframe+0x46/0xbb

-> #0 (&type->i_mutex_dir_key#7){+.+.}:
       lock_acquire+0x170/0x3f0 kernel/locking/lockdep.c:3998
       down_write+0x34/0x90 kernel/locking/rwsem.c:54
       inode_lock include/linux/fs.h:719 [inline]
       chown_common+0x272/0x4b0 fs/open.c:628
       SYSC_fchownat fs/open.c:661 [inline]
       SyS_fchownat+0xf6/0x190 fs/open.c:641
       do_syscall_64+0x1d5/0x640 arch/x86/entry/common.c:292
       entry_SYSCALL_64_after_hwframe+0x46/0xbb

other info that might help us debug this:

Chain exists of:
  &type->i_mutex_dir_key#7 --> &sbi->lock --> sb_writers#10

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(sb_writers#10);
                               lock(&sbi->lock);
                               lock(sb_writers#10);
  lock(&type->i_mutex_dir_key#7);

 *** DEADLOCK ***

1 lock held by syz-executor380/7979:
 #0:  (sb_writers#10){.+.+}, at: [<ffffffff818e083a>] sb_start_write include/linux/fs.h:1551 [inline]
 #0:  (sb_writers#10){.+.+}, at: [<ffffffff818e083a>] mnt_want_write+0x3a/0xb0 fs/namespace.c:386

stack backtrace:
CPU: 1 PID: 7979 Comm: syz-executor380 Not tainted 4.14.295-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/22/2022
Call Trace:
 __dump_stack lib/dump_stack.c:17 [inline]
 dump_stack+0x1b2/0x281 lib/dump_stack.c:58
 print_circular_bug.constprop.0.cold+0x2d7/0x41e kernel/locking/lockdep.c:1258
 check_prev_add kernel/locking/lockdep.c:1905 [inline]
 check_prevs_add kernel/locking/lockdep.c:2022 [inline]
 validate_chain kernel/locking/lockdep.c:2464 [inline]
 __lock_acquire+0x2e0e/0x3f20 kernel/locking/lockdep.c:3491
 lock_acquire+0x170/0x3f0 kernel/locking/lockdep.c:3998
 down_write+0x34/0x90 kernel/locking/rwsem.c:54
 inode_lock include/linux/fs.h:719 [inline]
 chown_common+0x272/0x4b0 fs/open.c:628
 SYSC_fchownat fs/open.c:661 [inline]
 SyS_fchownat+0xf6/0x190 fs/open.c:641
 do_syscall_64+0x1d5/0x640 arch/x86/entry/common.c:292
 entry_SYSCALL_64_after_hwframe+0x46/0xbb
RIP: 0033:0x7f1edcaba8a9
RSP: 002b:00007ffe73f1cb08 EFLAGS: 00000246 ORIG_RAX: 000000000000005e
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f1edcaba8a9
RDX: 0000000000000000 RSI: 000000000000ee01 RDI: 0000000020000080
RBP: 0000000000000000 R08: 00007f1edcb28ec0 R09: 00007f1edcb28ec0
R10: 0000000000000000 R11: 0000000000000246 R12: 00007ffe73f1cb30
R13: 0000000000000000 R14: 431bde82d7b634db R15: 0000000000000000

Fix bisection attempts:
Manager Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets Title
ci2-linux-4-14 2023/01/23 00:38 linux-4.14.y 97205fccccdc 267e3bb1 .config console log report syz C
ci2-linux-4-14 2022/11/05 03:52 linux-4.14.y a901bb6c7db7 267e3bb1 .config console log report syz C
* Struck through repros no longer work on HEAD.
Crashes (1):
Manager Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets Title
ci2-linux-4-14 2022/10/05 17:27 linux-4.14.y 9d5c0b3a8e1a 267e3bb1 .config console log report syz C [disk image] [vmlinux] [mounted in repro] possible deadlock in chown_common
* Struck through repros no longer work on HEAD.