testing new SMP scheduler

[Joel Sherrill]

On 6/13/2013 4:53 PM, Gedare Bloom wrote:
> What do you mean? I would expect that selecting this scheduler when
> SMP is not enabled will issue an error in confdefs.h, although I see
> the current logic is to fall back to the default (priority) scheduler.
configure with --enable-smp. That builds with SMP support.

But around line 632 of confdefs.h, it has

#if defined(RTEMS_SMP) && defined(CONFIGURE_SMP_APPLICATION)
   #define CONFIGURE_SCHEDULER_SIMPLE_SMP
#else
   #define CONFIGURE_SCHEDULER_PRIORITY
#endif

So even with --enable-smp, only the smptests get the SMP scheduler.
The uniprocessor tests use the known good scheduler. You have
to hack that conditional to use the other scheduler.

FWIW I know it is better to tests SMP priority schedulers against
all uniprocessor tests. But what should the uniprocessor scheduler
default default to?

The old simple scheduler passed all the uniprocessor tests. I am
asking if this one does also.

--joel

> On Thu, Jun 13, 2013 at 5:30 PM, Joel Sherrill
> <joel.sherrill at oarcorp.com> wrote:
>> Hi
>>
>> Out of curiosity, did you hack confdefs.h to force this as the
>> scheduler in uniprocessor configurations to ensure it does
>> the correct things in that situation?
>>
>> --
>> Joel Sherrill, Ph.D.             Director of Research & Development
>> joel.sherrill at OARcorp.com        On-Line Applications Research
>> Ask me about RTEMS: a free RTOS  Huntsville AL 35805
>> Support Available                (256) 722-9985
>>
>> _______________________________________________
>> rtems-devel mailing list
>> rtems-devel at rtems.org
>> http://www.rtems.org/mailman/listinfo/rtems-devel


-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel.sherrill at OARcorp.com        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
Support Available                (256) 722-9985