[stella] About Stella's debugger asm

[Peter Froehlich]

Hi all,

[We should probably take this to the DASM list?]

On May 1, 2008, at 1:12 PM, Stephen Anthony wrote:

> I guess I mean all equates defined in VCS.H vs. those defined by the
> user.  When we break all this down, all that's happening is a mapping
> from address to string (and vice-versa).  In this case, it would be
> much easier for Stella if DASM produced that info in the .sym file.
>
> Perhaps each symbol output to the .sum file could contain its 'type'
> (ie, ROM, RAM, USER-DEFINED, HARDCODED, etc) by use of a set of  
> numbers
> or something.  Granted, I can figure out much of this in Stella  
> itself,
> but it would complicate and slow parsing of the file.  Since DASM must
> know this info internally anyway, it would be great if it was exported
> to the sym file.

Please keep in mind that DASM is a general 8-bit assembler, not an  
Atari 2600 assembler and not a Stella assembler either. Some of these  
things make sense generally, some don't.

So I have no problem if what you want is mark a symbol as defined in  
an include file (and maybe which one) or defined in the main source  
file, that's something that could be useful for other platforms as  
well (heck, even for error messages).

However, including which addresses are RAM and which are ROM etc is  
*way* too 2600-specific for DASM. Also, it seems easy enough to map  
from address to RAM/ROM in Stella, right? You may not know *where*  
the EQU came from, but you *definitely* know what it is and thus  
whether it's a RAM or ROM address.

I hope we can solve those two parts separately in a way that makes  
sense for both projects, the outcome (of splitting things as  
indicated) should still be what you want: better disassembly in  
Stella. Right?

Cheers,
Peter
--
Peter H. Froehlich <><><><><><> http://www.cs.jhu.edu/~phf/
OpenPGP: ABC2 9BCC 1445 86E9 4D59  F532 A8B2 BFAE 342B E9D9