8080 DAA opcode

Can anyone point me at a reasonably detailed description of the
workings of the 8080 DAA instruction?

I've written an 8080 simulator but if I follow the details that I have
found for DAA [1] then it doesn't quite work as expected. If I make a
very small change to the details (reading between the lines really)
then I can get BCD add, subtract (using Intel's published method) and
increment to work correctly with DAA.

I'm sure I remember reading that DAA didn't work properly (or as
intended?) on the 8080 so I'm just trying to work out if my simulator
is wrong or not.

Thanks

[1] I'm following this two step process

(1) If the least significant four bits of the accumulator represent a
number greater than 9, or if the auxiliary carry bit is equal to one,
the accumulator is incremented by six. Otherwise, no incrementing
occurs.

(2) If the most significant four bits of the accumulator now represent
a number greater than 9, or if the normal carry bit is equal to one
(*), the most significant four bits of the accumulator are incremented
by six. Otherwise, no incrementing occurs.

(*) The bit I added to get BCD to work is to add "or step 1 caused a
carry from bit 7"
--
Ian

The From address is valid

http://www.geocities.com/SiliconValley/Peaks/3938/z80syntx.htm#DAA

On Sun, 27 May 2007 12:30:16 +0000, IanB wrote:

Can anyone point me at a reasonably detailed description of the
workings of the 8080 DAA instruction?

I've written an 8080 simulator but if I follow the details that I have
found for DAA [1] then it doesn't quite work as expected. If I make a
very small change to the details (reading between the lines really)
then I can get BCD add, subtract (using Intel's published method) and
increment to work correctly with DAA.

I'm sure I remember reading that DAA didn't work properly (or as
intended?) on the 8080 so I'm just trying to work out if my simulator
is wrong or not.

Thanks

[1] I'm following this two step process

(1) If the least significant four bits of the accumulator represent a
number greater than 9, or if the auxiliary carry bit is equal to one,
the accumulator is incremented by six. Otherwise, no incrementing
occurs.

(2) If the most significant four bits of the accumulator now represent
a number greater than 9, or if the normal carry bit is equal to one
(*), the most significant four bits of the accumulator are incremented
by six. Otherwise, no incrementing occurs.

(*) The bit I added to get BCD to work is to add "or step 1 caused a
carry from bit 7"

On Mon, 28 May 2007 21:51:39 -0700, Andy wrote:

http://www.geocities.com/SiliconValley/Peaks/3938/z80syntx.htm#DAA

Thanks for the link. That's for a Z80 though, which is supposed to be
a "fixed" version of the 8080 DAA instruction. That's why I'm trying
to find out just *what* it fixed :-)

Thanks anyway
--
Ian

The From address is valid

There is a good book that has several pages devoted to this very tricky
instruction: "The 8080A BUGBOOK", published by SAMS, ISBN #
0-672-21447-4. This is an excellent book with many practical programming
examples for the 8080A. Do a Google search and you will find low priced
sources for the book. One thing they do say is " Use the decimal adjust
accumulator (DAA) instruction only after an ADD, ADC, or ADI
instruction." A suggested way to use the very weird DAA is:

Octal Mnem.
306 ADI
000 000
047 DAA


Ian

Another SAMS book by the same authors is "8080 8085 Software Design",
ISBN # 0-672-21541-1. It also covers the DAA instruction in some detail.

On Tue, 29 May 2007 21:37:11 -0400, "Ian D"
wrote:

Thanks for the reply and apologies for the delay in replying

There is a good book that has several pages devoted to this very tricky
instruction: "The 8080A BUGBOOK", published by SAMS, ISBN #
0-672-21447-4.

I've actually got a copy of that book, purchased back in the late 70's
or early 80's I seem to recall it was one of the first "useful'
programming books that Tandys (the UK's version of Radio Shack)
stocked.

I'm pretty sure now that I must have some flag logic wrong, and that's
why I can't get DAA to work even in the "half hearted way" described
in the various 8080 references.

I've just got a Z80 emulator working correctly [1] so, with the added
insight that has given me, I'll go back and see what I had wrong with
the 8080.

Thanks again

Ian

[1] It passes some stringent tests I found on the web and also some
(less stringent) comparisons against a working z80 system..
--
Ian

The From address is valid

IanB wrote:
On Tue, 29 May 2007 21:37:11 -0400, "Ian D"
wrote:

Thanks for the reply and apologies for the delay in replying

There is a good book that has several pages devoted to this very tricky
instruction: "The 8080A BUGBOOK", published by SAMS, ISBN #
0-672-21447-4.

I've actually got a copy of that book, purchased back in the late 70's
or early 80's I seem to recall it was one of the first "useful'
programming books that Tandys (the UK's version of Radio Shack)
stocked.

I'm pretty sure now that I must have some flag logic wrong, and that's
why I can't get DAA to work even in the "half hearted way" described
in the various 8080 references.

I've just got a Z80 emulator working correctly [1] so, with the added
insight that has given me, I'll go back and see what I had wrong with
the 8080.

We've been waiting to hear what you found.

Thanks again

Ian

[1] It passes some stringent tests I found on the web and also some
(less stringent) comparisons against a working z80 system..


--
Bill Davidsen
He was a full-time professional cat, not some moonlighting
ferret or weasel. He knew about these things.