On 10.09.14 12:21, Paolo Bonzini wrote: > Il 10/09/2014 12:09, Alexander Graf ha scritto: >> Fair enough. >> >> As far as moving "platform bus" logic into sysbus, I'd really like to >> hold back and see what this whole thing ends up getting used for first. >> >> So for now, I'd definitely prefer to keep "platform bus" logic and >> "sysbus" logic separate. If we realize that every user only ever uses >> the dynamic sysbus creation in conjunction with our "platform bus" >> implementation, we can merge them. > > I agree. As you pointed out, we have two usecases: > > 1) arbitrary dynamic sysbus devices, because you're playing with board > design or because you're working on a virtualized platform > > 2) pluggable components in a fixed board design (e.g. CCSR) > > The only thing they share is FDT creation. The other part, which is > assigning the interrupts and memory regions, is different: case (1) has > it driven by command line or simply bottom-to-top; case (2) has it > driven by an implementation of a spec. > > It's not even clear to me that E500 CCSR devices should be sysbus, in > fact...
The problem if you continue that thought process is that we'd end up with 500 different buses and 500 different uart boilerplate devices just to fit into the respective buses ;). Otherwise I agree. Alex