LXRT hard realtime C++

Paolo Mantegazza mantegazza at aero.polimi.it
Mon Sep 30 12:09:05 CEST 2002


Ivan Martinez wrote:
> 
> On Monday 30 September 2002 11:15 am, Suessmilch Bernd wrote:
> > > -----Original Message-----
> > > From: Paolo Mantegazza [mailto:mantegazza at aero.polimi.it]
> > > Sent: Freitag, 27. September 2002 16:39
> > > To: Ivan Martinez
> > > Cc: rtai at rtai.org
> > > Subject: Re: LXRT hard realtime C++
> > >
> > > NO and yes. (NEW)LXRT passes you back to soft mode and satisfy the
> > > syscall, if you do that. You'll remain in the hand of Linux till the
> > > next RTAI call. In any case you'll loose hard real time performances.
> >
> > Is there a way to change this policy (eg. to the one used in
> > early versions of LXRT)? I bear in mind that you wrote about
> > such a possibility some month ago. If I am right, who
> > could it be achieved?
> 
> It looks like both the "old" LXRT and NEWLXRT are available in the latest RTAI
> distribution, so I guess you can choose between them.
> Ivan

Since NEWLXRT is a new born baby I'd like to remark that NEWLXRT is not
just LXRT. You can think so if you like but you'll miss the important
point that under the NEWLXRT naming there is a comprehensive kernel/user
space usage unified on Linux tasks/kthreads only.

Paolo.




More information about the Rtai mailing list