[syzbot] possible deadlock in md_open

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 372b2891c15a Add linux-next specific files for 20210824
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=124dca75300000
kernel config: https://syzkaller.appspot.com/x/.config?x=18ce42745c8b0dd6
dashboard link: https://syzkaller.appspot.com/bug?extid=fadc0aaf497e6a493b9f
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=13b10d05300000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=13fa60fe300000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+fadc0a...@syzkaller.appspotmail.com

======================================================
WARNING: possible circular locking dependency detected
5.14.0-rc7-next-20210824-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor325/6558 is trying to acquire lock:
ffff888013f90230 (&mddev->open_mutex){+.+.}-{3:3}, at: md_open+0xfd/0x2e0 drivers/md/md.c:7815

but task is already holding lock:
ffff888077ce3118 (&disk->open_mutex){+.+.}-{3:3}, at: blkdev_get_by_dev.part.0+0x9b/0xb60 fs/block_dev.c:1227

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&disk->open_mutex){+.+.}-{3:3}:
__mutex_lock_common kernel/locking/mutex.c:596 [inline]
__mutex_lock+0x131/0x12f0 kernel/locking/mutex.c:729
bd_register_pending_holders+0x2c/0x470 block/holder.c:160
device_add_disk+0x75e/0xfd0 block/genhd.c:505
add_disk include/linux/genhd.h:221 [inline]
md_alloc+0x91d/0x1150 drivers/md/md.c:5707
md_probe+0x69/0x70 drivers/md/md.c:5738
blk_request_module+0x111/0x1d0 block/genhd.c:667
blkdev_get_no_open+0x178/0x1e0 fs/block_dev.c:1150
blkdev_get_by_dev.part.0+0x22/0xb60 fs/block_dev.c:1214
blkdev_get_by_dev+0x6b/0x80 fs/block_dev.c:1267
swsusp_check+0x4d/0x270 kernel/power/swap.c:1525
software_resume.part.0+0x102/0x1f0 kernel/power/hibernate.c:977
software_resume kernel/power/hibernate.c:86 [inline]
resume_store+0x161/0x190 kernel/power/hibernate.c:1179
kobj_attr_store+0x50/0x80 lib/kobject.c:856
sysfs_kf_write+0x110/0x160 fs/sysfs/file.c:139
kernfs_fop_write_iter+0x342/0x500 fs/kernfs/file.c:296
call_write_iter include/linux/fs.h:2163 [inline]
new_sync_write+0x429/0x660 fs/read_write.c:511
vfs_write+0x7cf/0xae0 fs/read_write.c:598
ksys_write+0x12d/0x250 fs/read_write.c:651
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae

-> #0 (&mddev->open_mutex){+.+.}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3051 [inline]
check_prevs_add kernel/locking/lockdep.c:3174 [inline]
validate_chain kernel/locking/lockdep.c:3789 [inline]
__lock_acquire+0x2a07/0x54a0 kernel/locking/lockdep.c:5015
lock_acquire kernel/locking/lockdep.c:5625 [inline]
lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5590
__mutex_lock_common kernel/locking/mutex.c:596 [inline]
__mutex_lock+0x131/0x12f0 kernel/locking/mutex.c:729
md_open+0xfd/0x2e0 drivers/md/md.c:7815
blkdev_get_whole+0x99/0x2a0 fs/block_dev.c:1079
blkdev_get_by_dev.part.0+0x354/0xb60 fs/block_dev.c:1234
blkdev_get_by_dev+0x6b/0x80 fs/block_dev.c:1267
swsusp_check+0x4d/0x270 kernel/power/swap.c:1525
software_resume.part.0+0x102/0x1f0 kernel/power/hibernate.c:977
software_resume kernel/power/hibernate.c:86 [inline]
resume_store+0x161/0x190 kernel/power/hibernate.c:1179
kobj_attr_store+0x50/0x80 lib/kobject.c:856
sysfs_kf_write+0x110/0x160 fs/sysfs/file.c:139
kernfs_fop_write_iter+0x342/0x500 fs/kernfs/file.c:296
call_write_iter include/linux/fs.h:2163 [inline]
new_sync_write+0x429/0x660 fs/read_write.c:511
vfs_write+0x7cf/0xae0 fs/read_write.c:598
ksys_write+0x12d/0x250 fs/read_write.c:651
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae

other info that might help us debug this:

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&disk->open_mutex);
lock(&mddev->open_mutex);
lock(&disk->open_mutex);
lock(&mddev->open_mutex);

*** DEADLOCK ***

5 locks held by syz-executor325/6558:
#0: ffff88807f6bc460 (sb_writers#6){.+.+}-{0:0}, at: ksys_write+0x12d/0x250 fs/read_write.c:651
#1: ffff88801a6a3488 (&of->mutex){+.+.}-{3:3}, at: kernfs_fop_write_iter+0x287/0x500 fs/kernfs/file.c:287
#2: ffff8881441a6830 (kn->active#90){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x2ab/0x500 fs/kernfs/file.c:288
#3: ffffffff8b84fc68 (system_transition_mutex/1){+.+.}-{3:3}, at: software_resume.part.0+0x19/0x1f0 kernel/power/hibernate.c:932
#4: ffff888077ce3118 (&disk->open_mutex){+.+.}-{3:3}, at: blkdev_get_by_dev.part.0+0x9b/0xb60 fs/block_dev.c:1227

stack backtrace:
CPU: 1 PID: 6558 Comm: syz-executor325 Not tainted 5.14.0-rc7-next-20210824-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
check_noncircular+0x25f/0x2e0 kernel/locking/lockdep.c:2131
check_prev_add kernel/locking/lockdep.c:3051 [inline]
check_prevs_add kernel/locking/lockdep.c:3174 [inline]
validate_chain kernel/locking/lockdep.c:3789 [inline]
__lock_acquire+0x2a07/0x54a0 kernel/locking/lockdep.c:5015
lock_acquire kernel/locking/lockdep.c:5625 [inline]
lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5590
__mutex_lock_common kernel/locking/mutex.c:596 [inline]
__mutex_lock+0x131/0x12f0 kernel/locking/mutex.c:729
md_open+0xfd/0x2e0 drivers/md/md.c:7815
blkdev_get_whole+0x99/0x2a0 fs/block_dev.c:1079
blkdev_get_by_dev.part.0+0x354/0xb60 fs/block_dev.c:1234
blkdev_get_by_dev+0x6b/0x80 fs/block_dev.c:1267
swsusp_check+0x4d/0x270 kernel/power/swap.c:1525
software_resume.part.0+0x102/0x1f0 kernel/power/hibernate.c:977
software_resume kernel/power/hibernate.c:86 [inline]
resume_store+0x161/0x190 kernel/power/hibernate.c:1179
kobj_attr_store+0x50/0x80 lib/kobject.c:856
sysfs_kf_write+0x110/0x160 fs/sysfs/file.c:139
kernfs_fop_write_iter+0x342/0x500 fs/kernfs/file.c:296
call_write_iter include/linux/fs.h:2163 [inline]
new_sync_write+0x429/0x660 fs/read_write.c:511
vfs_write+0x7cf/0xae0 fs/read_write.c:598
ksys_write+0x12d/0x250 fs/read_write.c:651
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x43f0e9
Code: 28 c3 e8 2a 14 00 00 66 2e 0f 1f 84 00 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 c0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ffd59408358 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 0000000000400488 RCX: 000000000043f0e9
RDX: 000000000000fdef RSI: 0000000020000000 RDI: 0000000000000003
RBP: 0000000000402e40 R08: 0000000000000012 R09: 0000000000400488
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000402ed0
R13: 0000000000000000 R14: 00000000004ad018 R15: 0000000000400488


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzk...@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
syzbot can test patches for this issue, for details see:
https://goo.gl/tpsmEJ#testing-patches

[Song Liu]

Hi Christoph,

I looked into this and found the error was triggered by

d62633873590 ("block: support delayed holder registration")

But I am not quite sure how to fix it. Do you have some suggestions on
this?

Thanks,
Song

[syzbot]

Hello,

syzbot has tested the proposed patch but the reproducer is still triggering an issue:
possible deadlock in md_open

======================================================
WARNING: possible circular locking dependency detected

5.14.0-rc7-next-20210827-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.2/9024 is trying to acquire lock:
ffff88807b16c230 (&mddev->open_mutex){+.+.}-{3:3}, at: md_open+0xfd/0x2e0 drivers/md/md.c:7815

but task is already holding lock:

ffff88807ad42918 (&disk->open_mutex){+.+.}-{3:3}, at: blkdev_get_by_dev.part.0+0x9b/0xb60 fs/block_dev.c:1227

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&disk->open_mutex){+.+.}-{3:3}:
__mutex_lock_common kernel/locking/mutex.c:596 [inline]
__mutex_lock+0x131/0x12f0 kernel/locking/mutex.c:729
bd_register_pending_holders+0x2c/0x470 block/holder.c:160

device_add_disk+0x74c/0xfc0 block/genhd.c:470

add_disk include/linux/genhd.h:221 [inline]
md_alloc+0x91d/0x1150 drivers/md/md.c:5707
md_probe+0x69/0x70 drivers/md/md.c:5738

blk_request_module+0x111/0x1d0 block/genhd.c:632

blkdev_get_no_open+0x178/0x1e0 fs/block_dev.c:1150
blkdev_get_by_dev.part.0+0x22/0xb60 fs/block_dev.c:1214
blkdev_get_by_dev+0x6b/0x80 fs/block_dev.c:1267
swsusp_check+0x4d/0x270 kernel/power/swap.c:1525
software_resume.part.0+0x102/0x1f0 kernel/power/hibernate.c:977
software_resume kernel/power/hibernate.c:86 [inline]
resume_store+0x161/0x190 kernel/power/hibernate.c:1179
kobj_attr_store+0x50/0x80 lib/kobject.c:856
sysfs_kf_write+0x110/0x160 fs/sysfs/file.c:139
kernfs_fop_write_iter+0x342/0x500 fs/kernfs/file.c:296
call_write_iter include/linux/fs.h:2163 [inline]

new_sync_write+0x429/0x660 fs/read_write.c:507
vfs_write+0x7cf/0xae0 fs/read_write.c:594
ksys_write+0x12d/0x250 fs/read_write.c:647

do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae

-> #0 (&mddev->open_mutex){+.+.}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3051 [inline]
check_prevs_add kernel/locking/lockdep.c:3174 [inline]
validate_chain kernel/locking/lockdep.c:3789 [inline]
__lock_acquire+0x2a07/0x54a0 kernel/locking/lockdep.c:5015
lock_acquire kernel/locking/lockdep.c:5625 [inline]
lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5590
__mutex_lock_common kernel/locking/mutex.c:596 [inline]
__mutex_lock+0x131/0x12f0 kernel/locking/mutex.c:729
md_open+0xfd/0x2e0 drivers/md/md.c:7815
blkdev_get_whole+0x99/0x2a0 fs/block_dev.c:1079
blkdev_get_by_dev.part.0+0x354/0xb60 fs/block_dev.c:1234
blkdev_get_by_dev+0x6b/0x80 fs/block_dev.c:1267
swsusp_check+0x4d/0x270 kernel/power/swap.c:1525
software_resume.part.0+0x102/0x1f0 kernel/power/hibernate.c:977
software_resume kernel/power/hibernate.c:86 [inline]
resume_store+0x161/0x190 kernel/power/hibernate.c:1179
kobj_attr_store+0x50/0x80 lib/kobject.c:856
sysfs_kf_write+0x110/0x160 fs/sysfs/file.c:139
kernfs_fop_write_iter+0x342/0x500 fs/kernfs/file.c:296
call_write_iter include/linux/fs.h:2163 [inline]

new_sync_write+0x429/0x660 fs/read_write.c:507
vfs_write+0x7cf/0xae0 fs/read_write.c:594
ksys_write+0x12d/0x250 fs/read_write.c:647

do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae

other info that might help us debug this:

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&disk->open_mutex);
lock(&mddev->open_mutex);
lock(&disk->open_mutex);
lock(&mddev->open_mutex);

*** DEADLOCK ***

6 locks held by syz-executor.2/9024:
#0: ffff8880242b6d70 (&f->f_pos_lock){+.+.}-{3:3}, at: __fdget_pos+0xe9/0x100 fs/file.c:990
#1: ffff88807ed6e460 (sb_writers#6){.+.+}-{0:0}, at: ksys_write+0x12d/0x250 fs/read_write.c:647
#2: ffff888023db6488 (&of->mutex){+.+.}-{3:3}, at: kernfs_fop_write_iter+0x287/0x500 fs/kernfs/file.c:287
#3: ffff8881445fd830 (kn->active#326){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x2ab/0x500 fs/kernfs/file.c:288
#4: ffffffff8b84fbe8 (system_transition_mutex/1){+.+.}-{3:3}, at: software_resume.part.0+0x19/0x1f0 kernel/power/hibernate.c:932
#5: ffff88807ad42918 (&disk->open_mutex){+.+.}-{3:3}, at: blkdev_get_by_dev.part.0+0x9b/0xb60 fs/block_dev.c:1227

stack backtrace:
CPU: 0 PID: 9024 Comm: syz-executor.2 Not tainted 5.14.0-rc7-next-20210827-syzkaller #0

new_sync_write+0x429/0x660 fs/read_write.c:507
vfs_write+0x7cf/0xae0 fs/read_write.c:594
ksys_write+0x12d/0x250 fs/read_write.c:647

do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae

RIP: 0033:0x4665e9
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f14cf6ab188 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 000000000056bf80 RCX: 00000000004665e9

RDX: 000000000000fdef RSI: 0000000020000000 RDI: 0000000000000003

RBP: 00000000004bfcc4 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000056bf80
R13: 00007fffd49a28bf R14: 00007f14cf6ab300 R15: 0000000000022000
PM: Image not found (code -5)


Tested on:

commit: 5e63226c Add linux-next specific files for 20210827
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=13505659300000
kernel config: https://syzkaller.appspot.com/x/.config?x=9c03a8fdabc6b3ae

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-and-tested-by: syzbot+fadc0a...@syzkaller.appspotmail.com

Tested on:

commit: b0f50fa9 Merge branch 'for-5.15/io_uring' into for-next
git tree: git://git.kernel.dk/linux-block for-next
kernel config: https://syzkaller.appspot.com/x/.config?x=90efff855ea11a9

dashboard link: https://syzkaller.appspot.com/bug?extid=fadc0aaf497e6a493b9f
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1

patch: https://syzkaller.appspot.com/x/patch.diff?x=1330dfce300000

Note: testing is done by a robot and is best-effort only.