@Edwin:

Thank you very much, that's a great deal of information you gave,
that's definitely worth being considered.

> Polling for state-change is never a good idea (you might miss stuff if 
> your cpu is busy doing something else)
> This is why interrupts where invented I think.

You are right, it's neither very elegant nor efficient to poll
input signals, it's much more a quick and dirty hack, that I prefer
at the beginning of my project, as I do not want to elaborate
on hardware interfacing too much right now. But I guess, your solution
will be the right way to go at a later time.

I was working on a Tiny Tiger system that seems to be polling it's
input pins (timer interrupt driven) periodically (it's done in a
proprietary device driver) and the results were
quite satisfying in my opinion, so I guessed, it could be a good
enough solution to do it in a similar manner on a fox board (which is
much more  powerful I think).

---------------------------------------------------------------------

Anyway, I still wonder what might be the problem installing the timer1
interrupt. The interrupt (3) is not used on my system but enabling it
keeps hanging my system. I even disabled the serial driver in the
kernel, but that did not correct the problem.
Any further ideas?

Kind regards
Dietmar Dreyer



Reply via email to