Interesting DRAM Defects in Dynamic PET

[daver2]

>>> Of course the statement I made about putting my memory check program into ROM was very silly.

I didn't want to say anything at the time !

Think of self modifying code as self modifying hardware !

Yes, copying from ROM to RAM would be the answer in this case. Or you do what BASIC does with the 'wedge' and copies a small amount of code (in your case the self modifying subroutine) into RAM and then you call that. The bulk of the code can still be executed out of ROM in this case. This would save RAM...

Dave

 

[Hugo Holden]

Yes, copying from ROM to RAM would be the answer in this case. Or you do what BASIC does with the 'wedge' and copies a small amount of code (in your case the self modifying subroutine) into RAM and then you call that. The bulk of the code can still be executed out of ROM in this case. This would save RAM...

Dave

That simplifies another problem. I found that I had to assemble the program once first to find the address of the byte to modify as it was a fair way into the program about 10 or so bytes and I'm not familiar enough yet to count up bytes for each previous line of code, without seeing it assembled, but if the line of code to be modified has been moved to a RAM address, its easy to figure out from the few bytes of that, what address the target byte is living at from the previous examples of that assembled line of code. And it only uses a tiny amount of RAM vs moving the whole program.

Is it better to use a JSR and and RTS to get back, or just use a JMP in each direction ?