From: "Arobase" <salle.arob...@ville-rochefort.fr>
Newsgroups: comp.os.cpm
Subject: Z80 to 8080 translation
Date: Fri, 14 Jun 2002 12:55:19 +0200
Organization: Salle multimédia
Lines: 143
Message-ID: <aecha9$rv$1@wanadoo.fr>
Reply-To: "Arobase" <salle.arob...@ville-rochefort.fr>
NNTP-Posting-Host: apoitiers-102-2-1-132.abo.wanadoo.fr
X-Trace: wanadoo.fr 1024051337 895 193.253.213.132 (14 Jun 2002 10:42:17 GMT)
X-Complaints-To: abuse@wanadoo.fr
NNTP-Posting-Date: 14 Jun 2002 10:42:17 GMT
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2600.0000
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000

Z80OPS.txt
----------

"Request For Comments"

Sorry, group, but I  am working  on a  Z-80 problem,  and cannot
find my doc! (Maybe I should finally do some house cleaning?)

I need to translate in 8080 a program  using the  following Z-80
mnemonics:

1) DJNZ label
2) LDIR
3) LDDR
4) CPDR
5) RLC  B
6) SBC  HL,DE
7) SBC  DE,HL
8) LD   BC,(label)
9) LD   DE,(label)
10) LD   (label),DE

The only thing I found so  far is  a French  translation of  the
"Z-80 CPU Technical Manual".  My favorite  books by  Alan Miller
are "somewhere" buried in my stuff.

Here is my first draft at solving the problem.

1) DJNZ label
decrement B-reg
jump "label" if not zero
so, this must be:
DCR  B
JNZ  label

2) LDIR
(DE) <-- (HL)   ("symbolic operation" in Z-80 Manual)
DE <-- DE+1
HL <-- HL+1
BC <-- BC-1
Repeat until BC=0
so, this must be:
Loop1:
MOV  D,H
MOV  E,L
INX  D
INX  H
DCX  B
JNZ  Loop1
But when is a byte moved? Something is missing.

3) LDDR
(DE) <-- (HL)
DE <-- DE-1
HL <-- HL-1
BC <-- BC-1
Repeat until BC=0
so, this must be:
Loop1:
MOV  D,H
MOV  E,L
DCX  D
DCX  H
DCX  B
JNZ  Loop1
But, again, I don't see any byte moved. (Normally, in  the 8080,
HL is used to "point" to a byte. So, there must be  a "MOV  A,M"
and "MOV M,A" somewhere. I think that the "(DE) <-- (HL)"
must be replaced by "A <-- (HL)" and "(DE) <-- A". What do
you think? And there is also the problem of the number of bytes
to move...)

4) CPDR
A <-- (HL)
HL <-- HL-1
BC <-- BC-1
Repeat until A = (HL) or BC = 0
so, this must be:
Loop1:
MOV  A,M
DCX  H
DCX  B
???
since BC has just been decremented, we can jump if BC = 0
JZ   Loop1
and, to compare A with (HL), we could use:
LXI  label
CMP  M
JC   Loop1

5) RLC  B
rotate left circular B-reg
???

6) SBC  HL,DE
subtract with Carry
A <-- A-s-CY    ("s is any of r,n") ???

7) SBC  DE,HL
idem

8) LD   BC,(label)
B-reg <-- (label+1)     (or "HIGH (label)")
C-reg <-- (label)       (or "LOW  (label)")
so, this must be:
LXI  H,label
MOV  C,M
INX  H
MOV  B,M

9) LD   DE,(label)
D-reg <-- (label+1)
E-reg <-- (label)
so, this must be:
LXI  H,label
MOV  E,M
INX  H
MOV  D,M

10) LD   (label),DE
(label+1) <-- D-reg
(label)   <-- E-reg
so, this must be:
LXI  H,label
MOV  M,E
INX  M
MOV  M,D

Of course, DJNZ and LDIR are  pretty "standards"  (in fact,  the
BIOS   of    CP/M   Plus    has   a    function   providing    a
hardware-independent  LDIR...),  but  I  am concerned  about the
accuracy of translation of the other opcodes.

Relevent comments appreciated.

Yours Sincerely,
"French Luser"


EOF