[Grml] Re: boring

Michael Gebetsroither gebi at sbox.tugraz.at
Fri Dec 8 15:49:12 CET 2006


Quoting Mark <27e3kk302 at sneakemail.com>:

> OK, er, I do embedded work, and know what RT means.  I only suggested an
> obvious kernel change as one idea.

There may also a speed increase up to the numbers of CPU's, because  
the new kernel is SMP enabled.
Additionally i think the new gcc also does his part.

> The general claim that RT makes systems slower is hogwash.

1. Why?
2. Hard rt systems have nothing to do with execution speed, right?  
They _have to_ met their realtime constraints otherwise peoples live  
are at risk.

> A good RT
> system will distribute CPU cycles optimally with or without RT
> constraints.

How could a RT System distribute the cpu cycles optimally?
Does a RT system really distribute cpu cycles?

>  Anyway the person asking has a faster system now.  Sure
> that could result from someplace else, who knows.

time to guess ;)

greets,
michael




More information about the Grml mailing list