RFC:new superio proposal

[Peter Stuge]

On Tue, May 13, 2003 at 08:11:46AM -0600, ron minnich wrote:
[..snip..]

> The remaining question is how should the superio command look in 
> freebios2? 
> 
> superio sis/950 "com1=115200,8n1 lpt=1 com2=9600"

The way to go! As long as keywords are documented just like in the kernel,
this is excellent! Just the fact that people may recognize this approach
from the kernel is IMHO reason enough.


> superio sis/950 "com1baud=115200 lpt=1 com1chars=8n1"

Unneccessarily lengthy and 8n1 is related to 115200 so no need to separate
them.


> superio sis/950 ((com1 115200 8n1) (lpt 1))

Eww, lisp! Hehe. :)


> So,  my questions:
> 
> 1. Does this new scheme look workable. If not, what needs to change?

I think it looks very good and quite scalable. IANA LinuxBIOS expert though,
so I might be overlooking things that are missing or unworkable.


> 2. What should the 'struct configuration' be? does keyword/value work? 

I think so. Then leave value parsing up to each module, and maybe create a
couple of common primitives that can be used. How much is left to each
superio also depends on the diversity of keyword/value pairs.


> 3. what should the superio command look like? 

See above.


//Peter