RPG Trigger Programs -- IBM "sample program" Error

["Steve Brazzell" <steve@xxxxxxxxxxxx>]

Just thought I'd pass this along for any of you who have written or plan to
write RPG trigger programs.

In the "DB2 for OS/400 Database Programming" manuals, from V3R2 through
V4R3, there is a very small yet dangerous error in the RPG sample trigger
program.  I discovered this yesterday while assisting my client in a move
from a CISC V3R2 to a RISC V4R4 machine.

The *ENTRY PLIST in these manuals is shown this way:



******************************************************************
            * SET UP THE ENTRY PARAMETER LIST.

******************************************************************
           C           *ENTRY    PLIST
           C           PARM1     PARM           PARM1
           C           PARM2     PARM           PARM2

******************************************************************


This is wrong, but may or may not show up as an intermittent problem
(depending on what is in memory where and when).

The corrected sample program first shows up in the V4R4 manual, which lists
the same trigger program, but with the entry parms shown like this:



******************************************************************
            * SET UP THE ENTRY PARAMETER LIST.

******************************************************************
           C           *ENTRY    PLIST
           C                     PARM           PARM1
           C                     PARM           PARM2

******************************************************************

My discovery of this came when we began getting MCH3601 (Pointer not
initialized, or some such wording) errors when files with a trigger program
were modified.  After pulling out all of my hair (actually, I do have a few
strands left), I discovered this difference, changed the code to that shown
in the latter example, and the problem went away.  I currently had a call in
to IBM and, after my discovery, actually got one of the RPG Compiler guys on
the phone who said that, indeed, the earlier sample was incorrect and prone
to causing problems, but no one caught it until V4R4 docs were being
produced.  It also turns out, according to the IBM people I had on the
phone, that this change was never *noted* as such to alert customers to the
potential problem.

So, if you have any trigger programs that you based on the sample, make sure
they look like the bottom example, not the top, regardless of the version of
the OS you're on.

Also, I'd be interested in an explanation of exactly why the first example
would cause a problem.  The "PARM1" and "PARM2" in the result field get
moved to the corresponding data structures in the factor 1 column (in this
case, themselves) upon program invocation.  I'm still not clear why this
would cause a problem as the structures are the same size, by definition.
Anyone want to tackle this one??

Steve Brazzell
Innovatum, Inc.



+---
| 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
+---