Status of phycore_mpc5554 with latest RTEMS and RSB built tools

Sebastian Huber sebastian.huber at embedded-brains.de
Fri Feb 21 12:52:53 UTC 2014


On 2014-02-21 13:44, Peter Dufault wrote:
> On Feb 21, 2014, at 07:16 , Sebastian Huber<sebastian.huber at embedded-brains.de>  wrote:
>
>>> >>I forgot to ask the important question: What happens when you increase the number of tasks to e.g. 8?  SP02 still didn't run, when I broke in with the debugger it was always in the idle task.
>> >
>> >It is the same problem: blocking on output.  As a quick hack you can add
>> >
>> >CPU_CFLAGS += -DTESTS_USE_PRINTK
>> >
>> >to the *.cfg file of your BSP.
> I don't understand.  I'd expect the test to run once there were enough tasks.  It sounds like a real bug.

Yes, its a bug in the test if you ask me.  A lot of tests assume that output is 
non-blocking.  They use this to determine which task executes when and in which 
sequence.

>
> You say that the PREEMPT task is blocking on "puts( "PREEMPT - rtems_task_delete - deleting self" );" and never gets to the rtems_task_delete() that follows that line, and I guess you're saying all the other tasks are also blocking on their respective "puts()" calls, and that's why nothing happens when I increase the number of tasks.

If you increase the task count, then only the first problem spot is skipped.

>
> I assume the default termios mode for the interrupt driven console is line buffered.  Why isn't the output eventually finishing?  Why doesn't increasing the number of tasks to 8 let things run through?

You hit another bug in the test.

-- 
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.



More information about the devel mailing list