On Mon, 2011-05-16 at 11:01 -0500, Jon Elson wrote: 
> Kent A. Reed wrote:
> >
> > As for PREEMPT_RT, I have no idea whether it gives acceptable real-time 
> > performance, has an acceptable API, or is easy/hard to connect to EMC2 
> > code. Perhaps some of the core developers can comment.
> >   
>  From reading the RTAI list, I think the answer is no.  Prempt_RT is not 
> awful, but it seems to
> give as much as 100 us jitter on machines that give 5 us jitter with 
> RTAI.  This might barely be acceptable
> on a servo machine, but would be a big problem on a system with software 
> step generation.

I think it depends on what preempt-rt kernel you are using and what
other drivers you're running in the kernel.
I once tried it and it ran at about 15-20 us jitter. So that's already
usable for slow stepper systems. I don't know how recent rt development
influenced this further. My last tests were 1.5 years ago. There were
a lot of rt improvements in the meantime.

-- 
Greetings Michael.


------------------------------------------------------------------------------
Achieve unprecedented app performance and reliability
What every C/C++ and Fortran developer should know.
Learn how Intel has extended the reach of its next-generation tools
to help boost performance applications - inlcuding clusters.
http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to