RTEMS Posix lock failure

[Sebastian Huber]

Hello,

On 01/03/2014 11:32 PM, John Harwell wrote:
> Hello,
>
> In my code, I have determined that after a small number of pthread 
> mutex lock/unlocks, suddenly
> RTEMS enters Internal_error_Occurred() from _CORE_mutex_Seize(). In 
> order for this to happen, I have
> determined that the variable _Thread_Dispatch_disable_level must be > 
> 0. As I understand it, RTEMS should
> take care of that when entering/leaving ISRs, but somehow it is not 
> being properly decremented ( I have observed
> it to go to 2).

what is the value of the _ISR_Nest_level in this case?  You cannot 
obtain a mutex in interrupt context.

>
> If in gdb I clear _Thread_Dispatch_disable_level upon the call to 
> _CORE_mutex_Seize() right before before my code
> had been locking up, it continues on and runs smoothly, and all 
> subsequent lock/unlock calls work fine, implying
> that it is some sort of one-time issue.
>
> Any ideas on why this is happening?

You probably try to obtain a mutex in interrupt context.  It can be also 
a general memory corruption or a broken interrupt/context switch code.

>
> I am using RTEMS 4.10.
>
> Thanks,
>


-- 
Sebastian Huber, embedded brains GmbH

Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone   : +49 89 189 47 41-16
Fax     : +49 89 189 47 41-09
E-Mail  : sebastian.huber at embedded-brains.de
PGP     : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.