Re: [colorforth] Reverse engineering the BIOS

[Mark Slicker <maslicke@xxxxxxxxxxx>]

On Sun, 7 Mar 2004, Albert van der Horst wrote:

> On Sat, 6 Mar 2004, Mark Slicker wrote:
>
> > I am just curious what you expect to find. The two sources of
> > incompatibility of colorForth are the floppy and video. The video is
>
> Exactly that.
>
> > hopeless unless you are prepared to write a driver for the card. For the
>   ^^^^^^^^
> I don't think so. From the perspective of a Forther who is used to write
> his own, it may look so. From my perspective (with maybe more experience
> in reverse engineering than in Forth) it isn't.
>

I wish you luck. However, Chuck uses the bios to initialize the video. If
that is not functioning I don't know what good it is to disasemble the
bios. Besides this, each video card has a different bios, a different way
of configuring itself.

> > floppy perhaps a look at the bios will provide insight. However in
>                                                   ^^^^^^
> A reverse engineer is not necessarily after insight about how it works.
> Insight about how you can call it is sufficient.

Insight is required in applying the knowledge of a specific case more
generally.

>
> > this case you are not interested in whole bios, just the part that deals
> > with floppy. This is just something to consider, you may find the
> > complexity of the bios undermines a complete understanding.
>
> I am pretty sure I must set my dogs (the disassembler) after the whole
> BIOS. You are right that it helps that I want to deal with one or two
> tiny aspects. If not the complexity, then the amount of code in the BIOS
> makes complete understanding impossible. But that is not needed to
> attain our very limited goals. Interestingly a 2 Mbyte BIOS doesn't
> scare me dogs, no more than a 32 kbyte BIOS. They'll chew a little longer.
>
> Look at it this way. Why would you love to have a nice 3000 page
> description of the BIOS? To look in the index and see what chapters
> you can skip...

I think the interrupt vectors are a kind of index. I believe Terry
Loveall improved Chuck's floppy driver by examining a disassembly of the
right ISRs. I don't exactly how he did it.

Mark

---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com