On 25 Dec 2010 at 0:30, Eric Smith wrote:
Does anyone have information on undocumented 8080
instructions?
There's plenty of information out there about undocumented Z-80 and
8085 instructions, but I haven't found anything on the 8080.
Some of the non-Intel 8080 and 8080-compatible parts were independent
designs, rather than licensed second-sources using Intel-provided
masks, so those might have different behavior for the undocumented
opcodes.
I recall investigating this after Intel the undocumented 8085
instructions. To the best of my recollection, the undocumented 8080
opcodes are, by and large, no-ops or redundancies (i.e. they do the
same as a close-valued opcode probably from incomplete PLA decoding).
My test was pretty simple--plug the opcode, followed by two 00 bytes
(in case the opcode might be interpreted as a 3-byte instruction,
execute, check the flags and registers and location 0000 to see if
anything has changed, lather, rinse, repeat. No startling new
"triple precision divide" instructions.
I'm certain that if I had discovered some fabulous new instruction, I
would have certainly documented it.
--Chuck