AW: AW: [En-Nut-Discussion] irqstack.h in cvs

[Oliver Schulz]

Hi Matthias,

Sorry for the late answer, I had to do some work...

Well, you're right, the latency is increased by the jump to the new function
by 3 cycles. The Nut/OS IRQ dispatcher is still used.

Cheers,
Oliver.

-----Ursprüngliche Nachricht-----
Von: mringwal at inf.ethz.ch [mailto:mringwal at inf.ethz.ch] 
Gesendet: Samstag, 8. Oktober 2005 19:21
An: Ethernut User Chat (English)
Cc: Oliver Schulz
Betreff: Re: AW: [En-Nut-Discussion] irqstack.h in cvs

Hi Oliver

one short follow up question on this.
by replacing a macro with a function call, the latency is increased.
do you have an idea on how much ?
or is this negectable, as so far the Nut/OS IRQ dispatcher was used anyway?

cheers,
  matthias

On 08.10.2005, at 15:39, Oliver Schulz wrote:

> Hi William,
>
> Let me explain my changes.
> The main intention for the change is to move the whole push/pop code 
> out of the macro definition to a separate function. This results in 
> smaller program code, b/c in the former code, for _each_ used 
> interrupt the whole push/pop code was generated by the compiler.

_______________________________________________
En-Nut-Discussion mailing list
En-Nut-Discussion at egnite.de
http://www.egnite.de/mailman/listinfo.cgi/en-nut-discussion