Re: SETSPPFP fails over a V4R4 *PGM

["Phil Hall" <hallp@xxxxxxxx>]

Gene & Leif;

> > SETSPPFP on V4R4 works okay over V4R3-and-earlier QPRCRTPG-made
programs,
> > but gets MCH0601, space class X"00", over V4R4 QPRCRTPG-made
programs.
> >  V4R4 System API Reference still says QPRCRTPG creates 480 bytes in
> > associated space.  Has this changed?  Neeed we switch to terraspace
> > pointers here?
>
>
> Don't know Gene, I have also had programs fail on me when recompiled
with
> V4R4.
> Their V4R3 versions run fine. Curious thing: if you place a BRK point
(.e.g
> BRK "1")
> at the beginning of the program and then STRDBG prog ADDTRC (1), the
program
> works. Wonder how yours would behave?

I had this very same MCH with a old ILE/C program written a year or two
back when I compiled it on V4R2. The weird thing was I could run the
program once and it would work, run it the second time (no changes just
F9 and enter) the program would fail, run it the third time it would
work. Alternate executions would fail. Turned out to be that the C
runtime realloc() function was failing if the pointer was not explicitly
set to null the very first time the call to realloc() was made.

I fixed the code to ensure the pointer was initialized (even though it
was being initialized to null) and the pgm worked every time. The reason
(yes there is one!) I'm telling you this is that the pgm *never* failed
in a debug session. May help you...

--phil

+---
| This is the MI Programmers Mailing List!
| To submit a new message, send your mail to MI400@midrange.com.
| To subscribe to this list send email to MI400-SUB@midrange.com.
| To unsubscribe from this list send email to MI400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: dr2@cssas400.com
+---