> John N. Power wrote:
> suspect that you could do something like:
> > abc   EQU  0x100
>> ...
>> goto   abc
>>
>> and it would probably be accepted by MPASM. After all,
>> abc would appear in the symbol table (see the listing
>> file, and it will be there).

> Yes, but then, "abc" isn't a "label", it's still just a symbol
> holding a value used as an address in the goto instruction...

> And, even if this "works", it would be much better to actualy
> label the relevant *source line* with the "label" abc, of course.

> Jan-Erik.

Yes, it would be better. In the "old days", when addresses were
just numbers, it was useful at times to define assembler
variables with EQU or SET, perform arithmetic on them (possibly
in conditional assembly mode), and then use them as addresses
in instructions. Very bad, but useful.
Two things changed this: relocatable assemblers for microprocessors
and segmented addresses. Both of these defined addresses as
more than just a number. In the former case, addresses could be
local or external; in the latter, far or near, with a segment and
offset. With this additional information, the assembler could perform
type checking on addresses, and now variables were no longer
addresses.
In absolute mode, entries in the symbol table can be made with
EQU, SET, by beginning the name in column 1, or by following the
name with a colon. Once the item is in the symbol table, there
is no way to tell how it got there. The assembler uses symbols in
the table if it needs to do so to satisfy a reference. You can see where
this leads. If you really believe in structured programming, this
leads to bad habits; if not, then it has a lot of possibilities.

John Power

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body