I've seen it mentioned before that the A5h instruction was a MOV BIT,BIT instruction. I've never witnessed it myself and my own tests on an original Intel 8032 showed that the A5h instruction was a 1-byte, 1-cycle instruction that had the effect of setting the carry bit. I was able to discern no other effect of executing the A5h instruction.

It's entirely possible (even probable) that the effect of the A5h instruction is dependent on the derivitive chip that you are using.

However, assuming the A5h instruction was, in fact, a MOV BIT,BIT instruction, your note that the call to L34AF makes sense does lend credibility to that possibility.

Also, does the rest of the source code show any reference to the 70h bit? What does the L34AF function do? With the MOV BIT,BIT interpretation, the code is saying "Get a byre from external RAM address 2000h and put it in the accumulator. Then replace the highest bit of the accumulator with the contents of bit 70h and call L34AF".

It also must be said that the second set of code seems more probable. If the A5h instruction does nothing (or just sets the carry bit), then the instruction sequence...

MOV dptr,#2000h ;0749 90 20 00
MOVX A,@dptr ;074C E0
DB 0A5h ;074D A5
MOV A,@R1 ;074E E7

.. would make the whole MOVX instruction unnecessary since the MOVX instruction moves the data from XRAM 2000h to the accumulator, and the MOV A,@R1 moves the value of Internal RAM @R1 to the accumulator--thus overwriting the previous move. That would not be logical.

When reverse engineering code it really comes down to going through the code and seeing if it makes sense.

Craig Steiner