[coreboot] [RFC] port bochs-bios to gcc

[Jun Koi]

On 2/25/08, Jun Koi <junkoi2004 at gmail.com> wrote:
> On 2/25/08, Jun Koi <junkoi2004 at gmail.com> wrote:
>  > On 2/25/08, Jun Koi <junkoi2004 at gmail.com> wrote:
>  >  > Hi Kevin,
>  >  >
>  >  >
>  >  >  On 2/22/08, Kevin O'Connor <kevin at koconnor.net> wrote:
>  >  >  > FYI.  I'm interested in reviving the ADLO effort.  I sent the
>  >  >  >  following email to the bochs developers mailing list.
>  >  >  >
>  >  >
>  >  >
>  >
>  >
>  > I looked at the code, and found a lot of tricks there. Interesting! I
>  >  just wonder where the __start symbol is defined? (which you use it to
>  >  fix the jump at set_entry32)
>  >
>
>
> Hmm somehow I missed it: that is the _start in romlayout.S. But then
>  the question is: why the _start code jump back to itself at start
>  up???
>

Ah, now I see what happens. The -fwhole-program --combine option
really confuses me!

An idea: the BIOS code is not only used by Bochs, but others as well
(qemu and Xen come first in my mind). So how about maintaining it
separately from Bochs code? That makes patches to be merged quickly
than it should be, and new versions should be released more quickly,
too.

Thanks,
J