UAE Debugging with VASM lists

[Tigerskunk]

I generate a list with VASM's -L option, and get some output of my code looking like this:

F14:0002       	add.l 	#44,a0
               S01:0000253E:  41 E8 00 2C

In my understanding, $0000253E is the address of the code in RAM.

If I stop the execution of the program, and do a mem dump with

m 0000253E 2

I get

0000253E 0000 0000 0000 0000 0000 0000 D5C0 0000  ................
0000254E D5D0 0000 D5E0 0000 D5F8 0000 0000 4000  ..............@.

..as an output here. Which looks nothing like the mem area shown above in the code list.

I am not sure why this is happening, I think I have used these VASM lists addresses successfully in the MAME debugger when I created some NeoGeo stuff, but here the addresses don't seem to be correct. Or am i doing something completely wrong here? It's been a while since I used these things.

Thanks for any explanation.

[roondar]

I'd say the problem has to do with the fact that the Amiga doesn't really do fixed addresses.

Presuming you link to the standard Amiga Hunk format, you have no guarantees where the OS will load your data - addresses in Amiga OS are never fixed, your code/data could technically end up anywhere in RAM. If you're the one doing the loading yourself through for instance a track loader, code will naturally end up wherever you tell it to. But that still won't necessarily be the addresses reported by VASM unless you make it load to those addresses yourself.

[Tigerskunk]

Originally Posted by roondar

I'd say the problem has to do with the fact that the Amiga doesn't really do fixed addresses.

Presuming you link to the standard Amiga Hunk format, you have no guarantees where the OS will load your data - addresses in Amiga OS are never fixed, your code/data could technically end up anywhere in RAM. If you're the one doing the loading yourself through for instance a track loader, code will naturally end up wherever you tell it to. But that still won't necessarily be the addresses reported by VASM unless you make it load to those addresses yourself.

Hmm, okay. That's it also probably why it's working on NeoGeo and not on the Amiga...

[phx]

Originally Posted by Steril707

F14:0002           add.l     #44,a0
               S01:0000253E:  41 E8 00 2C

In my understanding, $0000253E is the address of the code in RAM.

Depends whether S01 is a relocatible section or absolute code. Did you use an ORG directive somewhere in front of it? Otherwise it is just the section offset, which always start with zero.

See roondar for more information.