SMP: Next steps

[Sebastian Huber]

On 2014-04-16 16:23, Sebastian Huber wrote:
> On 2014-04-16 16:17, Gedare Bloom wrote:
>>> >  o support for a priority boost to fix priority inheritance across scheduler
>>> >boundaries,
>>> >
>> It would also be great if this might fix
>> https://www.rtems.org/bugzilla/show_bug.cgi?id=2124
>>
>> The bug may apply to SMP scheduling, but I have not thought that hard about it.
>>
>
> It is not related to this bug.  What I want to address is described in "The
> limits of priority inheritance", page 125, Björn B. Brandenburg, Scheduling and
> Locking in Multiprocessor Real-Time Operating Systems, 2011.
>

After some more research I think that priority boosting is a very bad solution. 
  I will now try to implement Migratory Priority Inheritance:

https://www.mpi-sws.org/~bbb/papers/pdf/rtlws12.pdf

This will be a bit difficult though, since now a thread may be temporarily 
assigned to multiple schedulers.

-- 
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.