Re: [patch 2.5.31] transparent PCI-to-PCI bridges

[Benjamin Herrenschmidt]

>I agree. There is absolutely no reason to artificially limit the "bus" 
>structure to only three resoruces (and with hardcoded behaviour at that).
>
> .../...
>
>In short: if anything, we may at some point make the number of resources
>_larger_, not smaller.

What about removing that hard coded 4 where ever it lives, and either
do a #define (that could eventually be overriden by arch) or do a
dynamically allocated array ? If you are ok with that, I'll put that
on my list of things to do when I get back to moving the pmac drivers
to 2.5

Ben.