Q: interleave logic?

[David Woodhouse]

ebiederman at lnxi.com said:
>  What do we have nasty half incomplete interleave logic all throughout
> the cfi code?

> It looks like it could be done more cleanly with a wrapper around the
> probe code, and a wrapper around the mapping driver. 

The chip driver needs to know about the interleave to some extent - it's 
got to be able to deal with errors on individual chips in the interleave. 
(How well we actually do that at the moment is a different question, of 
course.)

I thought the macros that were introduced (by Nico?) were roughly the right 
thing. If you have better suggestions, though, I'm interested.

--
dwmw2