[coreboot] domain vs device statictree order

Mart Raudsepp mart.raudsepp at artecdesign.ee
Wed Jan 7 00:58:46 CET 2009


On T, 2009-01-06 at 16:19 -0700, Myles Watson wrote:
> On Tue, Jan 6, 2009 at 4:13 PM, Marc Jones <marcj303 at gmail.com> wrote:
> 
> > I agree it is parents before siblings but I thought it looked more
> > like a problem with the statictree. I must not understand .next.
> >
> > from http://pastebin.ca/1301081
> >
> > dev_apic_0 has a .sibling = &dev_domain_0 and .next = &dev_domain_0_pci_1_0
> >
> > Then later in the last device domain dev_domain_0_pci_f_2 has a .next
> > = &dev_domain_0.
> >
> > I would expect .sibling and .next to point to the &dev_domain_0 and
> > then dev_domain_0 to have .next = &dev_domain_0_pci_1_0
> 
> Yes.  I would too.  It's that way because parents get defined after
> children.  When the tree is being read, the full parent is not read
> (we don't get to the closing brace) until after we read all the
> children.

It seems the order code-wise is the expected one. Domain structure
definitions appear before its devices, and so on, just the next pointers
jump to the end for one of them in the pastebin example. Couldn't
the .next pointers simply be pointing to the entry that is going to be
printed out next as far as domain/device go?

"The rest of the geodelx fixup patch" looks quite invasive to me. Not
regarding quality, but to my own local changes I had done :)
Will take some digesting from the start to get things understood, merged
with other changes and working order for me. But tomorrow is clean-up
and submission time anyways. I will be adding a Flash device and juggle
around some code to be done in appropriate phases for the interaction
with VSA to be working as far as NAND Flash goes.


Mart Raudsepp





More information about the coreboot mailing list