[coreboot] some words on Multiboot...

Robert Millan rmh at aybabtu.com
Thu Sep 25 15:28:58 CEST 2008

On Thu, Sep 25, 2008 at 02:04:34PM +0200, Stefan Reinauer wrote:
> Robert Millan wrote:
> > On Thu, Sep 25, 2008 at 11:16:39AM +0200, Stefan Reinauer wrote:
> >   
> >> Yes, I'm curious, too. Which real world payloads work now that didn't
> >> work before?
> >>     
> >
> > http://grub.enbug.org/MultibootSystems has a list of noteworthy payloads
> > using Multiboot.  As I mentioned before, this list shouldn't be considered
> > exhaustive, since Multiboot is commonly used for new small projects, in part
> > due to the wide deployment of GRUB as a bootloader.
> >   
> Yes, the list is well known; Thanks for bringing it up again. My
> question aimed more at which of these, except grub invaders, can
> actually work now.

My expectation is that, aside from multi-image payloads, most of them will
either work right away, or only depend on minor features.

> The idea of booting an AROS kernel from the ROM is pretty nice indeed,
> giving the old Amiga "Kickstart" feeling back.

Got any direct link to a kernel image?  I might give it a try.

> So, do you see a simple method to get that "multiple code images" (is
> that multiboot modules?) support in there?

Once coreboot supports loading a single payload that is composed of more
than one image, it's rather simple to fill the module list structure with
details such as number of modules, etc, and include that in the MBI.

Which would convert "multiple code images" into "multiboot modules" (I
guess this answers your other question).

I'm not particularly interested in this use case though.  While I do think
that it could be interesting to support it in coreboot, it's not one of
the situations I mentioned before in which GRUB doesn't deliver any actual

But I don't go as far as you put it before ("this is bootloader stuff").  If
someone is inclined to implement multi-image support, but not the missing
Multiboot bits, I might be able to hack the Multiboot part in.

> I think this would greatly
> increase the number of supported multiboot systems

Well, surely even if the number wouldn't increase a lot, it'd still be so
due to the "greatness" of Xen and the Hurd, which are both very nice
projects ;-)

> (including openbios
> btw :-)

Haven't checked, but if openbios required multi-image support, coreboot
wouldn't be able to support it, which I'm aware it already does.

Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."

More information about the coreboot mailing list