Montavista's "Open Source Real-Time Linux Project" again

Thomas Gleixner tglx at linutronix.de
Fri Oct 15 01:21:29 CEST 2004


On Thu, 2004-10-14 at 20:13, Karim Yaghmour wrote:
> Except that RTAI was designed with hard-rt in mind, and if people did
> want to go ahead and show that it is deterministic, they can. That's
> the difference.

Yeah, I know. I doubt that. 

> > As for Montavista announce, it has two sides, the marketing side, which
> > is not RTAI project's business, and the technical side which is rather
> > good news. It shows some interest for Ingo Molnar's work, and make it
> > more likely to be included (even if partially) in Linux.
> > 
> > After all, applications using LXRT or Fusion use Linux calls, this is
> > what these technologies are all about. Such applications hence benefit
> > from any Linux responsiveness improvement.
> 
> There is no dispute of whether latency is good or not, it certainly is
> good. But there's a difference between reducing latency, and claiming
> to somehow provide hard-rt in Linux.

The missing bits are a few heartbeats away. You know, that there are in
Kernel modifications around, which are at least as deterministic as
RTAI. Modifing the kernel itself to do irq deferring and priviledged
preemption of self contained irqs and userspace tasks with a restricted
functionality is not that magic.

Responding to the marketing "blah" with the same amount of "blah" does
not change anything. Neither way of self-adulation is doing any good.

Provide a static codepath analysis for all execution paths and I might
buy your arguments.

tglx






More information about the Rtai mailing list