[Gmecm] Re: Gmecm Digest, Vol 19, Issue 3 - What really is a 'mask'?

[Robin Handley]

> Because each version of Code may put critical tables in different
> locations in the .bin, you need a different mask for each different
> code, and thus the term mask has come to be synonymous with the specific
> program used.

This is the bit which does not marry completely with my previous
understanding, and I realise that my original description was amibiguous in
this respect, but I think clear and concise otherwise.

So, in an attempt to be absolutely clear, is it true that, for a given code
'mask':

- The algorithms in the executable code are identical for all instantiations
(and, in fact, all executable code, aside from lookup vectors, is
identical).
- The calibration (lookup) data values will (obviously) vary between
applications to cope with different engine configurations etc. etc.
- The size and scaling etc. of all the calibration (lookup) data are
identical for all instantiations.
- The vector memory location of any given element (scalar/array/table) of
calibration data may vary between instantiations.

Is the above a necessary and sufficient description?

Robin

P.S. My school teachers used to have a paper 'mask' for marking multiple
choice questions (a piece of paper with holes punched in). Having programmed
in assembler for over 20 years, I'm also familiar with the concept of
masking bits.

----- Original Message -----
From: "William Lucke" <william.lucke at highspeedlink.net>
To: <gmecm at diy-efi.org>
Sent: 03 September 2006 18:58
Subject: [Gmecm] Re: Gmecm Digest, Vol 19, Issue 3


> Conceptually, I look at it hierarchically:
> Hardware
> Code
> Data
>
> Hardware is the ECM... 7730, 7749, etc.
> Code is the mask... $8D, $8F, etc.
> Data is the calibration... ANHT, AUJP, etc.
>
> The code consists of instructions that tell the ECM what to with the
> calibration data. When the ECM executes the same code instructions on
> different data sets, you get different outputs. Thus there can be
> multiple sets of calibration data for each code mask, because, for
> example, the speedometer data would have to be different for each axle
> ratio offered in a RWD car, or the calibration data for a 305 is
> different than that for a 350, even though they are run by the same ECM
> and code mask.
>
> The term "mask" is a little obtuse in that it stems from old school
> computer speak. If you printed out the .bin file you could theoretically
> cut holes in another piece of paper to lay over top of the printout so
> that you only looked at what you were interested in... you could then
> label each hole "0xAAAA - 0xAAAF: data so-and-so" and the piece of paper
> with the holes would become a "mask" for understanding your .bin. This
> is exactly what a definition file does in your PC's memory.
>
> Because each version of Code may put critical tables in different
> locations in the .bin, you need a different mask for each different
> code, and thus the term mask has come to be synonymous with the specific
> program used.
>
>
> Will
>
>
> > Date: Sat, 2 Sep 2006 19:28:18 +0100
> > From: "Robin Handley" <Robin at FuryWorld.fsnet.co.uk>
> > Subject: [Gmecm] What really is a 'mask'?
> >
> > I've never really known what a GM software 'mask' is - never having seen
a
> > clear definition of it. I had been beginning to think, of late, that a
> > 'mask' (e.g. $58, $8D) is a particular version of GM code, which is used
in
> > different applications (and given different 4-latter designations e.g.
> > ANHT/AUJP) by changing the calibration data only. But now I'm not so
sure.
> > Could somebody clarify?
> >
> > Robin
>
> _______________________________________________
> Gmecm mailing list
> Gmecm at diy-efi.org
> Subscribe: http://lists.diy-efi.org/mailman/listinfo/gmecm
> Main WWW page: http://www.diy-efi.org/gmecm
>