On 2022/11/16 13:17, Hillf Danton wrote:
> No deadlock was reported for syz-executor145/4505 and syz-executor145/4516.
> Why?
Lockdep should be able to report this real deadlock case. But unfortunately,
>> INFO: task syz-executor145:4505 blocked for more than 143 seconds.
>> Not tainted 6.1.0-rc5-syzkaller-00008-ge01d50cbd6ee #0
>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>> task:syz-executor145 state:D stack:21896 pid:4505 ppid:3645 flags:0x00004002
>>
>> 2 locks held by syz-executor145/4505:
>> #0: ffff88807268e100 (&dev->mutex){....}-{3:3}, at: device_lock include/linux/device.h:835 [inline]
>> #0: ffff88807268e100 (&dev->mutex){....}-{3:3}, at: nfc_unregister_device+0x87/0x290 net/nfc/core.c:1165
>> #1: ffffffff8e787b08 (rfkill_global_mutex){+.+.}-{3:3}, at: rfkill_unregister+0xcb/0x220 net/rfkill/core.c:1130
>>
>> INFO: task syz-executor145:4516 blocked for more than 144 seconds.
>> Not tainted 6.1.0-rc5-syzkaller-00008-ge01d50cbd6ee #0
>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>> task:syz-executor145 state:D stack:23096 pid:4516 ppid:3647 flags:0x00004004
>>
>> 2 locks held by syz-executor145/4516:
>> #0: ffffffff8e787b08 (rfkill_global_mutex){+.+.}-{3:3}, at: rfkill_fop_write+0x1b3/0x790 net/rfkill/core.c:1278
>> #1: ffff88807268e100 (&dev->mutex){....}-{3:3}, at: device_lock include/linux/device.h:835 [inline]
>> #1: ffff88807268e100 (&dev->mutex){....}-{3:3}, at: nfc_dev_down+0x33/0x260 net/nfc/core.c:143
device_initialize() hides dev->mutex from lockdep tests; explained at
https://groups.google.com/g/syzkaller-bugs/c/Uj9LqEUCwac/m/BhdTjWhNAQAJ .
