Re: kernel BUG at kernel/rtmutex_common.h:75

[Thomas Gleixner <tglx@linutronix.de>]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 1184 bytes --]

On Fri, 6 Nov 2015, Thomas Gleixner wrote:
> On Wed, 4 Nov 2015, Yimin Deng wrote:
> > It seems that the purpose to call the remove_waiter() is to remove the
> > waiter added by “plist_add(&waiter->list_entry, &lock->wait_list);” in
> > the task_blocks_on_rt_mutex(). But in the scenario above there's no
> > waiter on the lock yet and
> > the waiter has not been added into the wait list of the lock in the
> > task_blocks_on_rt_mutex() due to the failure “-EAGAIN”. So it reported
> > kernel BUG in the rt_mutex_top_waiter().
> > 
> > I modified it as below and the issue seems disappear.
> > - if (unlikely(ret))
> > + if (unlikely(ret && (-EAGAIN != ret)))
> >        remove_waiter(lock, waiter);
> > 
> > Could the scenario above be possible? If so, how to resolve this issue?
> > Thanks!
> 
> Yes it is possible. Nice detective work!
> 
> Your solution is correct, but actually it's not sufficient, because we
> have another possibility to return early without being queued
> (-EDEADLOCK). Find the full solution below.
> 
> Thanks for tracking that down!

Btw, please update to 3.12.48-rt66. It contains quite some bugfixes in
the area of futex/rtmutex.

Thanks,

	tglx