No Subject

[jfox@xxxxxxxxxx (Jeff Fox)]

Dear Raul,

>   It is not meaningless.  You are just not understanding what is
>   going on.  000 001 002 003 becomes 6AA 6AB 6A8 6A9.  You are
>   counting in patterns, you need to count in numbers and look at the
>   address for the pattern.  Check out 6AA 6AB 6A8 6A9 and you should
>   see the first instructions that are executed in the boot code.
>
>I don't understand what you're getting at here.

I think you do.  Your earlier info in this mail tells me you now
have the idea.  The P21 sees it as 0,1,2,3 but it appears at
AA AB A8 A9 in the rom.  If you look at Chuck's source for
boot code and look at the object code in the boot room you
can firgure it out.  It gets confusing because if you xor with AAAAA
twice it is invisible.  It can give you a headache remembering
which way you viewing things, patterns, numbers, cross compiling,
boot code vs dram access etc.

Chuck used to really confuse people when he would talk about this
stuff in public!

Yes, you are right it is this way because it cuts the transistor count
by almost two, and it makes some things work much faster.  It is a bit
of a pain for cross development tool writers.  It mostly goes away
however when you are dealing with code ON the chip.  The P21Forth
version of Chuck's assembler is much cleaner than cross compiler
version.  You will also notice that all the opcodes are XORed with
AAAAA in the two versions. (cross development is more complicated when
the machines do not have the same bus logic)

>Hmm... what is that likely to mean in terms of availability?

If F21 is submitted next week, 9 weeks to first  prototype back.
Then some redesign will be needed and a second prototype could go
back in at the end of July if all goes well.  So we would be
done with prototyping in Oct if things go well.  So  volume
production would not be available till year end.

Jeff Fox