[coreboot] Subtractive Resources
mylesgw at gmail.com
Fri Nov 7 17:59:20 CET 2008
> -----Original Message-----
> From: ron minnich [mailto:rminnich at gmail.com]
> Sent: Friday, November 07, 2008 9:55 AM
> To: Myles Watson
> Cc: Coreboot
> Subject: Re: Subtractive Resources
> On Fri, Nov 7, 2008 at 8:46 AM, Myles Watson <mylesgw at gmail.com> wrote:
> > My understanding is that the amd8111 _has_ a bridge, but the device that
> > causing the problem is lpc, which is not a bridge, nor behind the
> > bridge. So when its resources are read and it is found to have
> > resources, the code tries to descend. It doesn't have a bus, and this
> > fails.
> you are right. I missed it. That's a good catch.
> Do you need those resources added for correct operation?
They get added fine. That's why I wanted to know why that code was there.
> Even if we had a link, does it make sense to descend the link to read
> subtractive resources?
I don't know.
> I like patch 1 but at the same time it feels like maybe we're not
> getting at the right problem. If we're that point in the code, and
> reading links, why are the links not there?
I agree. It does happen when it goes through the domain, which I think is
correct, and has no problems.
> > Sorry I wasn't more clear. I meant that the fix no longer tries to call
> > functions with NULL pointers. I don't think the dts is wrong, but I
> > that there are a lot fewer intermediaries in v3 then there were in v2.
> Well, that I like to hear.
> Marc made the case that things such as superio should not even be
> "under" the lpc in the dts, since they stand "outside" the tree in
> some sense. He argued that we should instead put them at top level.
> There is merit to his argument. This would simplify the lpc code as
I think it's fine to have the lpc define resources since that's who
implements them, but I like the idea of the resource being visible from the
domain (maybe not outside the tree.)
More information about the coreboot