Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1724305 > unrolled thread

Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)

Started bySebastian Andrzej Siewior <bigeasy@linutronix.de>
First post2017-08-31 17:00 +0200
Last post2017-09-02 16:00 +0200
Articles 3 — 3 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Possible BT deadlock due to recursive read locking (Was: Re: kernel  BUG at kernel/locking/rtmutex.c:1059) Sebastian Andrzej Siewior <bigeasy@linutronix.de> - 2017-08-31 17:00 +0200
    Re: Possible BT deadlock due to recursive read locking (Was: Re:  kernel BUG at kernel/locking/rtmutex.c:1059) Mart van de Wege <mvdwege@gmail.com> - 2017-09-01 11:00 +0200
      Re: Possible BT deadlock due to recursive read locking (Was: Re:  kernel BUG at kernel/locking/rtmutex.c:1059) Mart van de Wege <mvdwege@mail.com> (by way of Mart van de Wege         <mvdwege@gmail.com>) - 2017-09-02 16:00 +0200

#1724305 — Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)

FromSebastian Andrzej Siewior <bigeasy@linutronix.de>
Date2017-08-31 17:00 +0200
SubjectPossible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)
Message-ID<ukz6O-6TE-25@gated-at.bofh.it>
On 2017-08-30 23:25:53 [+0200], Mart van de Wege wrote:
> Hi,
Hi,

> The issue persists up to v4.11.12-rt10
Does
	CONFIG_RWLOCK_RT_READER_BIASED=y
make it go away?

> Stacktrace:
> 
> Aug 29 17:28:04 localhost kernel: [   46.483810] ------------[ cut here ]------------
> Aug 29 17:28:04 localhost kernel: [   46.483812] kernel BUG at kernel/locking/rtmutex.c:1059!

this is a deadlock on RT however !RT has a hidden problem which not
yelled at by lockdep. We have the following call path:

| hci_send_monitor_ctrl_event()
|   read_lock(&hci_sk_list.lock);
|
|   hci_send_to_channel()
|      read_lock(&hci_sk_list.lock);

So both functions acquire the same read_lock. If a write comes along
between the first read_lock() and second read_lock() then we have a
deadlock because the write_lock() will lock down further readers from
acquiring the read-lock until the writer completed its task.

This recursive locking was introduced in 38ceaa00d02d ("Bluetooth: Add
support for sending MGMT commands and events to monitor").

Sebastian

[toc] | [next] | [standalone]


#1724889 — Re: Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)

FromMart van de Wege <mvdwege@gmail.com>
Date2017-09-01 11:00 +0200
SubjectRe: Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)
Message-ID<ukPXY-1xv-23@gated-at.bofh.it>
In reply to#1724305
On Thu, 31 Aug 2017 16:52:12 +0200
Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:

> On 2017-08-30 23:25:53 [+0200], Mart van de Wege wrote:
> > Hi,  
> Hi,
> 
> > The issue persists up to v4.11.12-rt10  
> Does
> 	CONFIG_RWLOCK_RT_READER_BIASED=y
> make it go away?
> 
Yes, that does make it go away. -rt8 now boots cleanly, and works fine.
Building -rt10 now to be sure.

> > Stacktrace:
> > 
> > Aug 29 17:28:04 localhost kernel: [   46.483810] ------------[ cut
> > here ]------------ Aug 29 17:28:04 localhost kernel: [   46.483812]
> > kernel BUG at kernel/locking/rtmutex.c:1059!  
> 
> this is a deadlock on RT however !RT has a hidden problem which not
> yelled at by lockdep. We have the following call path:
> 
> | hci_send_monitor_ctrl_event()
> |   read_lock(&hci_sk_list.lock);
> |
> |   hci_send_to_channel()
> |      read_lock(&hci_sk_list.lock);
> 
> So both functions acquire the same read_lock. If a write comes along
> between the first read_lock() and second read_lock() then we have a
> deadlock because the write_lock() will lock down further readers from
> acquiring the read-lock until the writer completed its task.
> 
> This recursive locking was introduced in 38ceaa00d02d ("Bluetooth: Add
> support for sending MGMT commands and events to monitor").
> 
> Sebastian

[toc] | [prev] | [next] | [standalone]


#1725509 — Re: Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)

FromMart van de Wege <mvdwege@mail.com> (by way of Mart van de Wege <mvdwege@gmail.com>)
Date2017-09-02 16:00 +0200
SubjectRe: Possible BT deadlock due to recursive read locking (Was: Re: kernel BUG at kernel/locking/rtmutex.c:1059)
Message-ID<ulh7Q-3Ed-9@gated-at.bofh.it>
In reply to#1724889
On Fri, 1 Sep 2017 10:51:07 +0200
Mart van de Wege <mvdwege@gmail.com> wrote:

> On Thu, 31 Aug 2017 16:52:12 +0200
> Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:
> 
> > On 2017-08-30 23:25:53 [+0200], Mart van de Wege wrote:  
> > > Hi,    
> > Hi,
> >   
> > > The issue persists up to v4.11.12-rt10    
> > Does
> > 	CONFIG_RWLOCK_RT_READER_BIASED=y
> > make it go away?
> >   
> Yes, that does make it go away. -rt8 now boots cleanly, and works
> fine. Building -rt10 now to be sure.
> 
Confirm that as of 4.11.12-rt11 the deadlock still goes away when
enabling CONFIG_RWLOCK_RT_READER_BIASED

Mart

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web