Re: possible deadlock in __generic_file_fsync

From: syzbot
Date: Sat Oct 20 2018 - 12:13:10 EST

Next message: Vinod: "Re: [[PATCH] 8/9] DMA-UART-Driver-for-AST2500"
Previous message: Paul E. McKenney: "Interrupts, smp_load_acquire(), smp_store_release(), etc."
In reply to: syzbot: "Re: possible deadlock in __generic_file_fsync"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

syzbot has found a reproducer for the following crash on:

HEAD commit: 270b77a0f30e Merge tag 'drm-fixes-2018-10-20-1' of git://a..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=146f4ad9400000
kernel config: https://syzkaller.appspot.com/x/.config?x=b3f55cb3dfcc6c33
dashboard link: https://syzkaller.appspot.com/bug?extid=5cd33f0e6abe2bb3e397
compiler: gcc (GCC) 8.0.1 20180413 (experimental)
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1436fc45400000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=11058e2d400000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+5cd33f0e6abe2bb3e397@xxxxxxxxxxxxxxxxxxxxxxxxx

syz-executor388 (5339) used greatest stack depth: 15944 bytes left
syz-executor388 (5337) used greatest stack depth: 15800 bytes left

======================================================
WARNING: possible circular locking dependency detected
4.19.0-rc8+ #293 Not tainted
------------------------------------------------------
kworker/0:1/14 is trying to acquire lock:
000000008e61a3a9 (&sb->s_type->i_mutex_key#10){+.+.}, at: inode_lock include/linux/fs.h:738 [inline]
000000008e61a3a9 (&sb->s_type->i_mutex_key#10){+.+.}, at: __generic_file_fsync+0xb5/0x200 fs/libfs.c:981

but task is already holding lock:
00000000160c39d9 ((work_completion)(&dio->complete_work)){+.+.}, at: process_one_work+0xb9a/0x1b90 kernel/workqueue.c:2128

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 ((work_completion)(&dio->complete_work)){+.+.}:
process_one_work+0xc0a/0x1b90 kernel/workqueue.c:2129
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413

-> #1 ((wq_completion)"dio/%s"sb->s_id){+.+.}:
flush_workqueue+0x30a/0x1e10 kernel/workqueue.c:2655
drain_workqueue+0x2a9/0x640 kernel/workqueue.c:2820
destroy_workqueue+0xc6/0x9c0 kernel/workqueue.c:4155
sb_init_dio_done_wq+0x74/0x90 fs/direct-io.c:634
do_blockdev_direct_IO+0x12ea/0x9d70 fs/direct-io.c:1283
__blockdev_direct_IO+0x9d/0xc6 fs/direct-io.c:1417
ext4_direct_IO_write fs/ext4/inode.c:3743 [inline]
ext4_direct_IO+0xae8/0x2230 fs/ext4/inode.c:3870
generic_file_direct_write+0x275/0x4b0 mm/filemap.c:3042
__generic_file_write_iter+0x2ff/0x630 mm/filemap.c:3221
ext4_file_write_iter+0x390/0x1420 fs/ext4/file.c:266
call_write_iter include/linux/fs.h:1808 [inline]
aio_write+0x3b1/0x610 fs/aio.c:1561
io_submit_one+0xaa1/0xf80 fs/aio.c:1835
__do_sys_io_submit fs/aio.c:1916 [inline]
__se_sys_io_submit fs/aio.c:1887 [inline]
__x64_sys_io_submit+0x1b7/0x580 fs/aio.c:1887
do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&sb->s_type->i_mutex_key#10){+.+.}:
lock_acquire+0x1ed/0x520 kernel/locking/lockdep.c:3900
down_write+0x8a/0x130 kernel/locking/rwsem.c:70
inode_lock include/linux/fs.h:738 [inline]
__generic_file_fsync+0xb5/0x200 fs/libfs.c:981
ext4_sync_file+0xa4f/0x1510 fs/ext4/fsync.c:120
vfs_fsync_range+0x140/0x220 fs/sync.c:197
generic_write_sync include/linux/fs.h:2732 [inline]
dio_complete+0x75c/0x9e0 fs/direct-io.c:329
dio_aio_complete_work+0x20/0x30 fs/direct-io.c:341
process_one_work+0xc90/0x1b90 kernel/workqueue.c:2153
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413

other info that might help us debug this:

Chain exists of:
&sb->s_type->i_mutex_key#10 --> (wq_completion)"dio/%s"sb->s_id --> (work_completion)(&dio->complete_work)

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock((work_completion)(&dio->complete_work));
lock((wq_completion)"dio/%s"sb->s_id);
lock((work_completion)(&dio->complete_work));
lock(&sb->s_type->i_mutex_key#10);

*** DEADLOCK ***

2 locks held by kworker/0:1/14:
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: __write_once_size include/linux/compiler.h:215 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: arch_atomic64_set arch/x86/include/asm/atomic64_64.h:34 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: atomic64_set include/asm-generic/atomic-instrumented.h:40 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: atomic_long_set include/asm-generic/atomic-long.h:59 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: set_work_data kernel/workqueue.c:617 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: set_work_pool_and_clear_pending kernel/workqueue.c:644 [inline]
#0: 000000006dbdba7d ((wq_completion)"dio/%s"sb->s_id){+.+.}, at: process_one_work+0xb43/0x1b90 kernel/workqueue.c:2124
#1: 00000000160c39d9 ((work_completion)(&dio->complete_work)){+.+.}, at: process_one_work+0xb9a/0x1b90 kernel/workqueue.c:2128

stack backtrace:
CPU: 0 PID: 14 Comm: kworker/0:1 Not tainted 4.19.0-rc8+ #293
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: dio/sda1 dio_aio_complete_work
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x1c4/0x2b4 lib/dump_stack.c:113
print_circular_bug.isra.33.cold.54+0x1bd/0x27d kernel/locking/lockdep.c:1221
check_prev_add kernel/locking/lockdep.c:1861 [inline]
check_prevs_add kernel/locking/lockdep.c:1974 [inline]
validate_chain kernel/locking/lockdep.c:2415 [inline]
__lock_acquire+0x33e4/0x4ec0 kernel/locking/lockdep.c:3411
lock_acquire+0x1ed/0x520 kernel/locking/lockdep.c:3900
down_write+0x8a/0x130 kernel/locking/rwsem.c:70
inode_lock include/linux/fs.h:738 [inline]
__generic_file_fsync+0xb5/0x200 fs/libfs.c:981
ext4_sync_file+0xa4f/0x1510 fs/ext4/fsync.c:120
vfs_fsync_range+0x140/0x220 fs/sync.c:197
generic_write_sync include/linux/fs.h:2732 [inline]
dio_complete+0x75c/0x9e0 fs/direct-io.c:329
dio_aio_complete_work+0x20/0x30 fs/direct-io.c:341
process_one_work+0xc90/0x1b90 kernel/workqueue.c:2153
worker_thread+0x17f/0x1390 kernel/workqueue.c:2296
kthread+0x35a/0x420 kernel/kthread.c:246
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:413
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.Ay7bpq/1/bus PID: 14 Comm: kworker/0:1
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.l9bGzq/7/bus PID: 2682 Comm: kworker/0:2
syz-executor388 (5488) used greatest stack depth: 15128 bytes left
syz-executor388 (5560) used greatest stack depth: 14328 bytes left
syz-executor388 (5630) used greatest stack depth: 12872 bytes left
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.l9bGzq/18/bus PID: 2682 Comm: kworker/0:2
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.Ay7bpq/19/bus PID: 2682 Comm: kworker/0:2
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.Ay7bpq/22/bus PID: 5540 Comm: kworker/0:4
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.ygzABq/30/bus PID: 14 Comm: kworker/0:1
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.aGaJtq/29/bus PID: 5 Comm: kworker/0:0
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.qLFhwq/30/bus PID: 5540 Comm: kworker/0:4
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.Ay7bpq/35/bus PID: 5540 Comm: kworker/0:4
syz-executor388 (6261) used greatest stack depth: 12504 bytes left
Page cache invalidation failure on direct I/O. Possible data corruption due to collision with buffered I/O!
File: /root/syzkaller.aGaJtq/49/bus PID: 14 Comm: kworker/0:1
syz-executor388 (7306) used greatest stack depth: 12488 bytes left

Next message: Vinod: "Re: [[PATCH] 8/9] DMA-UART-Driver-for-AST2500"
Previous message: Paul E. McKenney: "Interrupts, smp_load_acquire(), smp_store_release(), etc."
In reply to: syzbot: "Re: possible deadlock in __generic_file_fsync"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]