Re: [coreboot] r593 -

[ron minnich]

On Thu, Feb 14, 2008 at 1:10 PM, Stefan Reinauer  wrote:
> Peter Stuge wrote:
> You are right. I think it should stay SuperIO.

I'll try to put it back tomorrow.
> > > {
> > >      ramsize = "128";
> > >      constructor = "i440bx_constructors";
> > > domainid = "0x8086, 0x7190";
> > >
> > > };
> >
> > Again, does this identify a new domain created by this device, or
> > does it identify this device within the containing domain? Can we
> > escape this ambiguity?
>
> I don't even think this should be in the dts at all. The driver code
>  should know which pci ids to bind to. And we are doing this already:

This identifies the device id of this device. That's got nothing to do
with the containing domain. It's the type of device. The containing
domain is an instance of this type.

Any suggestions for lessening confusion? I don't know what to do here.
> > > .ops = &i440bxemulation_pcidomainops},
> > > {.id = {.type = DEVICE_ID_PCI,
> > >            .u = {.pci = {.vendor = 0x8086,.device = 0x7190}}},
> > These ids are repeated four times, that is at least two times too
> > many. Yes, this device id is both a PCI device and a domain, but
> > still.
>
> Fully agreed.

I'll look at it tomorrow and see what I can do. Until then, ideas welcome.

We have 6 weeks until Denver, it would be nice to lock this set of issues down.

ron