* [Xenomai-help] ensuring code running in kernel modules has realtime priority
@ 2012-01-18 7:59 RT
2012-01-18 10:52 ` Gilles Chanteperdrix
0 siblings, 1 reply; 3+ messages in thread
From: RT @ 2012-01-18 7:59 UTC (permalink / raw)
To: Xenomai-help
Hi all,
we have an interesting situation. We are using a module to do communication
over a network (ethercat master module) to some remote IO.
The ethercat module is loaded against the kernel in the normal manner.
Now we possibly have overlooked that even though the call to the ethercat
module (ioctl call) from our real-time application takes place in a xenomai
real time task, the code of the ethercatmodule may be interrupted by other
processes running on the machine. Due to the fact that the modules code is
not running in any real-time prioritized space.
If this is indeed the case, what would be the best or easiest manner to
ensure that the module does not get interrupted, ie. to encapsulate it into
the xenomai real-time priority sphere?
Hope someone can point us in the right direction here. Its pretty urgent
because the system is already running live and we are intermittently
encountering this problem.
Kind regards,
Roland.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [Xenomai-help] ensuring code running in kernel modules has realtime priority
2012-01-18 7:59 [Xenomai-help] ensuring code running in kernel modules has realtime priority RT
@ 2012-01-18 10:52 ` Gilles Chanteperdrix
2012-01-20 15:31 ` Benner Bernd (MH/EES)
0 siblings, 1 reply; 3+ messages in thread
From: Gilles Chanteperdrix @ 2012-01-18 10:52 UTC (permalink / raw)
To: rolandtollenaar; +Cc: RT, Xenomai-help
On 01/18/2012 08:59 AM, RT wrote:
> Hi all,
>
> we have an interesting situation. We are using a module to do communication
> over a network (ethercat master module) to some remote IO.
>
> The ethercat module is loaded against the kernel in the normal manner.
>
> Now we possibly have overlooked that even though the call to the ethercat
> module (ioctl call) from our real-time application takes place in a xenomai
> real time task, the code of the ethercatmodule may be interrupted by other
> processes running on the machine. Due to the fact that the modules code is
> not running in any real-time prioritized space.
>
> If this is indeed the case, what would be the best or easiest manner to
> ensure that the module does not get interrupted, ie. to encapsulate it into
> the xenomai real-time priority sphere?
>
> Hope someone can point us in the right direction here. Its pretty urgent
> because the system is already running live and we are intermittently
> encountering this problem.
If I understand correctly, what you need is to rewrite the ethercat
driver as a real-time driver using the RTDM skin.
--
Gilles.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [Xenomai-help] ensuring code running in kernel modules has realtime priority
2012-01-18 10:52 ` Gilles Chanteperdrix
@ 2012-01-20 15:31 ` Benner Bernd (MH/EES)
0 siblings, 0 replies; 3+ messages in thread
From: Benner Bernd (MH/EES) @ 2012-01-20 15:31 UTC (permalink / raw)
To: rolandtollenaar@domain.hid; +Cc: Xenomai-help@domain.hid
Hi Roland,
i think that there is a commercial ethercat master with a xenomai rtdm interface available
from koenig ( https://koenig-pa.com )
We are using a open source linux ethercat master ( http://www.etherlab.org/de/ethercat/index.php )
from IGH.
I have developed a xenomai rtdm realtime interface for this igh ethercat master. All sources of this interface are now part of the development tree of the IGH- ethercat Master and can be downloaded from the IGH- website.
My way was to do all initializations and slow non realtime stuff ( SDO etc.) with the linux device interface and attach the rtdm driver to the running master and port only the "send","receive" and "process domain" ioctrl to rtdm. But this will only work if your master will not use the linux network layer.
The IGH master works, because special patched network drivers are part of the project and there is no other linux interfacing in the real time path . ( The Igh Master is designed for RTAI )
Mit freundlichen Grüßen / Best Regards
Bernd Benner
MH/SGT3
moehwald
Bosch Group
Moehwald GmbH
Michelinstraße 21
66424 Homburg
GERMANY
www.moehwald.de
Tel.: +49 (0)6841 707-229
b.benner@domain.hid
Moehwald GmbH, Sitz/Registered Office: Homburg/Saar, Zentrales Handelsregister/Registered Court: Saarbrücken HRB 2812
Geschäftsführung/Managing Directors: Christian Artmann, Wolfgang Geenen
-----Ursprüngliche Nachricht-----
Von: xenomai-help-bounces@domain.hid [mailto:xenomai-help-bounces@domain.hid] Im Auftrag von Gilles Chanteperdrix
Gesendet: Mittwoch, 18. Januar 2012 11:52
An: rolandtollenaar@domain.hid
Cc: RT; Xenomai-help@domain.hid
Betreff: Re: [Xenomai-help] ensuring code running in kernel modules has realtime priority
On 01/18/2012 08:59 AM, RT wrote:
> Hi all,
>
> we have an interesting situation. We are using a module to do
> communication over a network (ethercat master module) to some remote IO.
>
> The ethercat module is loaded against the kernel in the normal manner.
>
> Now we possibly have overlooked that even though the call to the
> ethercat module (ioctl call) from our real-time application takes
> place in a xenomai real time task, the code of the ethercatmodule may
> be interrupted by other processes running on the machine. Due to the
> fact that the modules code is not running in any real-time prioritized space.
>
> If this is indeed the case, what would be the best or easiest manner
> to ensure that the module does not get interrupted, ie. to encapsulate
> it into the xenomai real-time priority sphere?
>
> Hope someone can point us in the right direction here. Its pretty
> urgent because the system is already running live and we are
> intermittently encountering this problem.
If I understand correctly, what you need is to rewrite the ethercat driver as a real-time driver using the RTDM skin.
--
Gilles.
_______________________________________________
Xenomai-help mailing list
Xenomai-help@domain.hid
https://mail.gna.org/listinfo/xenomai-help
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2012-01-20 15:31 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-01-18 7:59 [Xenomai-help] ensuring code running in kernel modules has realtime priority RT
2012-01-18 10:52 ` Gilles Chanteperdrix
2012-01-20 15:31 ` Benner Bernd (MH/EES)
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.