Re: [Etherboot-developers] [COMMIT] 5.1.2+ boot from disk.

[<ebi...@ln...>]

ke...@us... (Ken Yap) writes:

> >I think I'm seeing the same problem.  I'm getting total garbage in the
> >E820 map.  I added a config option IGNORE_E820_MAP that forces use of the
> >"fake" memory map used when the e820 calls fail, which gets it to the
> >point of "Loading 10.1.0.1:boot-8139too.nbi ...(NBI)" at which point it
> >locks up and requires a hard reset.
> >
> >Ideas, toolchain test, etc. welcome.
> 
> Hmm, suspicion falls on the e820 BIOS call shim in pcbios.S. I should
> mention that I didn't actually go all the way and boot a kernel with my
> test, only another copy of Etherboot. Also it was on an old Pentium
> which may be using the e801 call.

I don't think the e820 BIOS fails completely.  As it worked
on the one system I tested it on.  But it is a very recent dual P4
Xeon system.  But there are a lot of cases not convered with that one
datapoint. 

Ken that gives us at least there is one data point the NBI loader
works.

Ken which toolchain gcc/binutils.  If we can collect some datapoints
we may have something to go on.

Eric