Re: Chain to subfile turns on *INKL?

If you're really, really stuck, have a look at all the parameter lists.
Make sure the called programs' definitions match the calling programs.
It seems bizarre, but it can happen that mis-mapped parameters can
cause memory corruption and not just around a 'CALL' operation.

Bizarre, but quite understandable when you find the problem. I've seen
this happen a number of times

where:

field B100 in program A (a parm on a call to Prog B) is clobbered after a
call to program B

where:

the programmer "knows" that program B did NOT touch field field B100, so
program B is not suspect and not examined.

What can happen is that the length of field B100 was increased in Prog A
but since the programmer "knew" prog B didn't use that field, Prog B never
got changed.

Let's say field B100 in program A is 100 bytes long, but in prog B it is
still an original 80 bytes. If field B100 is in D spec (in both programs)
and is follow by a data field B20 20 bytes long (in both programs) and
field B20 is changed in program B, then data gets corruupted because the
20 bytes of field B20 are addressing the same bytes in memory as bytes 81
to 100 of field B100 in program A.

The nastiest time I've had with this involved a MODS where I changed the
number occurrances. Mine was harder to find because the failure was on a
customers machine where the customer forgot to make the change to the
number of occurrences in their copy of program B. Drove me nuts trying to
figure out why their application failed and my copy worked fine.

Program B was essentially a USER EXIT program which I didn't write.

Steve






--
This is the RPG programming on the AS400 / iSeries (RPG400-L) mailing list
To post a message email: RPG400-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/rpg400-l
or email: RPG400-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/rpg400-l.