The Strange Birth and Long Life of Unix - IEEE Spectrum

[Ori Idan]

On Wed, Dec 7, 2011 at 2:14 PM, Scott Allen <mlxxxp-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org> wrote:

> On 6 December 2011 16:23, James Knott <james.knott-bJEeYj9oJeDQT0dZR+AlfA at public.gmane.org> wrote:
> > Other computers had separate I/O instructions and signals on the bus.
>  Since
> > the PDP-11 didn't have those, it had to use memory mapped I/O.  However,
> you
> > could generally use memory mapped I/O on any CPU.
>
> True; unless there are timing or caching problems, memory mapped I/O
> just appears as regular memory to a CPU.
>
> When I designed an I/O board for my 8080 based system (back around
> 1978) I made all the I/O registers both I/O and memory mapped at the
> same time. In some cases, you could write more efficient assembler if
> you could choose between using a memory or I/O instruction to access a
> register.
>

I think the AVR processor from ATMEL can see the I/O as memory space but
can still access it with special instructions.

-- 
Ori Idan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gtalug.org/pipermail/legacy/attachments/20111207/ad1fc5c7/attachment.html>