Subject: Bootloader development - the saga continues
To: None <news-innominate.list.netbsd.port-arm32@innominate.de>
From: Reinoud Zandijk <imago@kabel065011.kabel.utwente.nl>
List: port-arm32
Date: 12/06/2000 13:07:56
Hi all,

you can imagine how suprised i was (and pleased i must say when i saw the
result) when i tried to boot an old NetBSD kernel with the old bootloader
in 800x600 : it gave the same problems !!! ... that means that either both
bootloaders have the same bug concerning screenaddresses or more likely :
it isnt supported by the vidcconsole code.... a problem less to worry
about...

Apart from the redisch colour until a virtual console swap i can conclude
that the video part is (nearly) finished. Even X works fine....

Rests still the debug code and the in kernel debugger.... I'm now trying
to find out what debug option makes it crash ... but one of them does.
I'll build a test-kernel today with just the DDB in it ... Does anyone
know how to get into DDB to see if my symbol table is correct ? (yeah RTFM
! )

I hope i can conclude that the following stands for the new bootloader :
	- works on RiscOS 3.5/3.6/4.0x + NCOS 1.xx
	- works on RiscPC, A7000 (?), NC
	- works with and without VRAM
	- works on ARM7 (+ARM6?) + StrongARM
	- starts up in current screen mode
	- automatically adds VRAM emulation megs
	- supports all screenmodes that have a size of a multiple of 4k
	- all currently used arguments work (exept dmunge)
	- it needs some refinements : an automatic boot trugh (working on)
	- it *fails* to run debug kernels (*snif*)
	- .......

Any comments / results / challenges?

Cheers,
Reinoud