Re: Opdesc question

["David Morris" <dmorris@xxxxxxxxxxxxx>]

Hans,

Glad it should work.  Changing the input parameter would be the ideal 
fix, but this is used by every program we have.   This would have a very 
big impact on our system and the timing is not right.  I will try backing off 
our most recent PTFs and if that fails, I will try versioning this procedure.  
Fortunately we only have this particular situation in a few places.

Would you have an APAR number?  It is impossible to search on VARYING 
and OPDESC/*VARSIZE don't come up with any relevant hits.

Thanks,

David Morris


>>> boldt@ca.ibm.com 12/29/99 07:36AM >>>


David wrote:
>A couple of years ago, I created some procedures that used operational
>descriptors to retrieve the length of *VARSIZE parameters. These
parameters
>were also defined as CONST.  Since VARYING length strings became
available,
>I have been using VARYING in most cases.  I also found out that passing a
>varying field to a procedure that received a *VARSIZE CONST parameter
worked
>just fine... at least I thought it was OK.  I had a program receive an
error that has not
>been modified for some time.  The error was caused because the operational
>descriptor for the field said the length received was 32768, where the
length of the
>varying string passed was 1.  The *VARSIZE field was defined as 2048, and
the
>caller was defined as 10 VARYING.  The problem appears to be caused when
the
>field being passed is the return value of another procedure.  Built-in and
stand
>alone fields do not appear to have any problems.
>
>Here is a snippet of code that represents the problem situation:
>
>D Proc1       PR          10a  VARYING
>
>D Proc 2      PR
>D   Input                         2048a  OPTIONS(*VARSIZE) CONST
>
>C                     CALLP  Proc2(Proc1)
>
>Should this work?  The manual does not say otherwise, and as far as I can
tell
>it used to work.  We did apply some RPG PTF a few days ago, but they do
not
>mention either OPDESC, or VARYING.  It is possible that this has never
worked
>and CEEDOD has been returning a valid value by chance.

Hi David!  Yes, that should work.  We are aware of some
problems in this area and we're already working on an
APAR fix.

But probably the better approach would be to define the
parameter "Input" as VARYING CONST.

Cheers!  Hans

Hans Boldt, ILE RPG Development, IBM Toronto Lab, boldt@ca.ibm.com 


+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---