Hello Ed,
I agree that base-less is somehow misleading, but today I had the problem to
explain to my managers what I am doing, and why this is good for the ASSEMBLER
developers, and the managers always nead a kind of keyword, and "base-less
coding" works very well as a keyword for the managers - no matter if they
understand what it is or not - in fact, it is very hard to explain it to
non-techies.
Maybe "base-less with respect to the code area" could be a valid extension - of
course, there will always be base registers for the data areas in the near and
not-so-near future.
Kind regards
Bernd
Am 05.04.2013 22:37, schrieb E.J.:
> On 4/5/2013 1:11 PM, P.F.wrote:
>> That works for the CALL only if the parameters are not in a dynamic
>> area in a DSECT somewhere.
>
> Of course! And, if they are you use MF=E instead of MF=I format.
>
>> For the GETMAIN, note that the "L" for the literal loaded into R15
>> requires a base register able to address the LTORG area.
>
> Literals, constants, and working storage DSECTs require base register
> coverage. This is why I don't like the term 'base-less' to describe code
> that uses relative branches.
>
> --
> E.J
>
|