On 12/17/2010 11:46 PM, Martin Trübner wrote:
While Tony gave a perfect answer-....it might be useless for a newbee.
Here is my attempt: there is no point in determining the coverage (or
non-coverage) - use baseless approach all the time and you are safe.
I believe the concern is that fields in a control block mapped by a DSECT will
sometimes be referenced with different instructions than fields in the CSECT.
For example, you can't use LARL to get the address of a field in a control block
mapped by a DSECT. That instruction is used only to point to fields in your
CSECT.
We have an inner macro that is used by every macro-based service we author. It
processes macro parameters. It allows them to be passed as a name or as a
sublist consisting of a name and a technique. If technique is not present, the
macro makes assumptions about how to handle the parameter (based on syntax,
&SYSALVL,etc.) and is almost always right. If technique is present, it
explicitly specifies the instruction that should be used to access the
parameter. For example KWD=(NAME,LG) says to use LOAD GRANDE when accessing the
field called NAME.
--
Edward E Jaffe
Phoenix Software International, Inc
831 Parkview Drive North
El Segundo, CA 90245
310-338-0400 x318
[email protected]
http://www.phoenixsoftware.com/
