Frank Kotler wrote:
> ...
>> Good idea, but be aware to find some different interpretations of
>> register-numbers.
>> ie: Herbert uses a functional logic for ordering: eax,edx,..
>> while I use the given opcode-patterns: eax,ecx,edx... to match PUSHAD
> Yeah... the registers *have* real names - 3-bit numbers - which
> *almost*, but not quite, match Herbert's scheme.
> I couldn't resist adding this. Rosario (I think that's his right name)
> just posted this to a.l.a.
aka "to A or not to B" ? :)
I've seen it and it looks a bit similar to my own variant of Terse
just used in tag-file comments for my source-free code yet.
I can only particular show how it looks because of my self defined
character-set and a graphic flow-chart with logic elements like arrows,
overline, math-symbols and it may show up with different colours for
mem-references and altered stack incl. call/int/push/pop/ret.
a bit modified (just for example):
...00 ax=0013 int10 ax=a000 bx=0140 es=ax clr di ax=00c8
;320*200
...12 y:cx=0100
...15 :es[di]=cl inc di dec cx j<>x ;*this x is a U-turn-arrow
here
...1C lea di,bx+di+ff00 cmp di,ax j<y ;*y draws an arrow to 'y:'
...24 ret
I think this is good readable for almost everyone :)
and also show immediate that loops weren't aligned.
(my "<" and friends mean unsigned, for signed I use "<0","s<","-",..)
and because of the CPU don't know decimal figures, everything is HEX
of course !
yeah, tastes differ a lot. And the term Logic seem to have a wide
range of sense ... :)
__
wolfgang
_______________
> org 100h
> section .text
> start:
> ax=13h; int 10h; /* switch to mode 13h
> ax=0A000h; es=ax; /* load es with our "graphics screen" seg
> di^=di; al=0; /* start at upper left [di=0] color 0 [al=0]
> /* draw the lines
> .1: cx=256; /* ++al is next color,
> /* ++di is next position
> .2: [es:di]=al; ++al,di; --cx; !#.2;
> /* after 256 incs, al is color 0 again
> /* al=0; /* so skip this
> di+=_320-256_; di<_320 * 199 + 256 _#.1
> ah=0; int 16h; /* wait for a key
> ax=3; int 10h; /* back to text mode
> ret /* back to dos
>
> Believe it or not, this assembles with Nasm... *after* passing through
> Rosario's "preasm.exe" (I think I've got his code, if anyone actually
> wants to assemble this). He finds this syntax much superior. I think I
> mentioned... tastes differ!
>
> Best,
> Frank
>
|
