Hi Stephen,

On Fri, 1 Dec 2000, Stephen Borrill wrote:
> > Were these tried with the newest A7000 kernel suplied together with the
> > newest version of the bootloader ?
> 
> No, they weren't. Just tried it now though. The kernel now prints
> "Undefined instruction in kernel. Heavy man!" over and over again.

Hmm... thats wierd... i've had that error myself once only when i tried to
boot a debug kernel ... I guess it has something to do with the symbol
table... i'll have a look at that.. When i start a debug kernel, it also
started printing VERY slowly... but i dont have a clue as to why...

> > > On the NC it doesn't get that far. The error "No DRAM maped in this
> > > program (wierd) (sic) ... Increase Wimpslot?". Tried increasing the
> > > Wimpslot as high as 8MB; still no joy.
<SNIP>

Fixed this one ... oh boy... the selection statement had to be inside the
while offcource ... *sigh* ... just because 0x100FB000 < 0x10100000 it
couldnt't find it (ARGH) Silly me

> No, this is the NC. The A7000 reports 6 blocks.

A 6 blocks ... hmm.. in RiscOS it isnt allways so that your given logical
memory is one continues piece of physical memory... the memory blocks
reported are the physical memory blocks it got... therefore the number can
change... ideally one but most of the time 2, 3 or more ... depends on how
much you've been messing around with the computer before you start
!BtNetBSD.

> file, only two blocks are shown. I guess one is system workspace and the
> other is the wimpslot for BtNetBSD. After getting into the desktop and
> running !BtNetBSD, 5 blocks are reported and what's more, IT BOOTS INTO
> THE KERNEL! OK, it bombs out in the same way as on the A7000, but that
> further than we've ever got before.

Groovy !!! Is this a debug kernel ? or the suplied A7000 kernel ? (no
debug as far as i know) ... hmm... sounds like there is a slight problem
in the loader... a missing page or so ... will look at it !

> Running !BtNetBSD early on an A7000 doesn't work either. It reports only 2
> blocks again. Different error though (Can't use array reference at line
> 2090) 

Ouch... sounds like i havent given enough space to the loader... fixed ...
given it quite some extra space .... hmm.. have to think of another
strategy ...

> > And did you also run the FastBoot file ? or the !BtNetBSD application ?
> > (parameters)
> 
> Makes no difference. The parameters are in the FastBoot file.

Yeah that is to enable the use of a frontend to create the FastBoot file
when it is desireable in time...

Thanks a lot for your testing ... it really shows a lot of small mistakes
and a good improvement = ) ... I'll upload the new versions as soon as i
have build in the alledged StrongARM support .... This is a candidate for
testing too :)

Cheers,
Reinoud