Re: [PATCH 2/4] tty: Hold tty_ldisc_lock() during tty_reopen()

From: Tetsuo Handa
Date: Fri Aug 31 2018 - 07:18:14 EST


On 2018/08/31 15:51, Jiri Slaby wrote:
> On 08/29/2018, 05:19 PM, Tetsuo Handa wrote:
>> On 2018/08/29 11:23, Dmitry Safonov wrote:
>>> tty_ldisc_reinit() doesn't race with neither tty_ldisc_hangup()
>>> nor set_ldisc() nor tty_ldisc_release() as they use tty lock.
>>> But it races with anyone who expects line discipline to be the same
>>> after hoding read semaphore in tty_ldisc_ref().
>>>
>>> We've seen the following crash on v4.9.108 stable:
>>>
>>> BUG: unable to handle kernel paging request at 0000000000002260
>>> IP: [..] n_tty_receive_buf_common+0x5f/0x86d
>>> Workqueue: events_unbound flush_to_ldisc
>>> Call Trace:
>>> [..] n_tty_receive_buf2
>>> [..] tty_ldisc_receive_buf
>>> [..] flush_to_ldisc
>>> [..] process_one_work
>>> [..] worker_thread
>>> [..] kthread
>>> [..] ret_from_fork
>>>
>>> I think, tty_ldisc_reinit() should be called with ldisc_sem hold for
>>> writing, which will protect any reader against line discipline changes.
>>>
>>> Note: I failed to reproduce the described crash, so obiviously can't
>>> guarantee that this is the place where line discipline was switched.
>>
>> This will be same with a report at
>> https://syzkaller.appspot.com/bug?id=f08670354701fa64cc0dd3c0128a491bdb16adcc .
>>
>> syzbot is now testing a patch from Jiri Slaby.
>
> Yes, my patch passed, so could you add:
> Reported-by: syzbot+3aa9784721dfb90e984d@xxxxxxxxxxxxxxxxxxxxxxxxx
>
> (not adding tested-by as this particular patch was not tested, but
> shoiuld work the same way.)
>
> thanks,
>

Tested with all 4 patches applied using syzbot-provided reproducer and
my simplified reproducer. No crashes and no lockdep warnings.
Also, noisy messages like

pts pts4033: tty_release: tty->count(10529) != (#fd's(7) + #kopen's(0))

are gone. Very nice. Thank you.