Re: [coreboot] [PATCH] put coreboot table in high tables area

[Stefan Reinauer]

On 18.03.2009 12:12 Uhr, Peter Stuge wrote:
> Stefan Reinauer wrote:
> > +++ i386/coreboot.c	(working copy)
>
> ..
> > case CB_TAG_FORWARD:
> > 	return cb_parse_header((void *)(unsigned long)((struct cb_forward *)rec)->forward, len, info);
> > 	continue;
>
> Are these semantics correct? Will this recurse? Should it also
> continue parsing after a forward tag?
>
> I guess we should have decided on these things before committing
> but.. :)

My word is code decides, not the comitee ;-)
> If the intent is to never have more than one forward tag and never
> have anything but the forward tag in one table if there is a forward
> tag at all I think we should at least document it but ideally codify
> it, so that we don't end up with a situation where someone tries to
> "misuse" the structures, or fail to parse them?

Unless someone else is going to create a Firmware that produces a
coreboot table in a different way than it is right now, we might want to
revisit. In the current scenario this would merely be academic.

Stefan