Works without problems.
The only way that you could run into problems would be, if the LE initialization
would step up the save area chain and check somehow, if the caller of the main
module (at entry CEESTART) is indeed the operating system. But this is not done.
At our site, we have some sort of test driver which builds a sort of test driver
environment (@M.: using TRAP2, you know) and then calls main programs
written in ASSEMBLER, C or PL/1 to test them. No problem at all.
Am 20.07.2011 15:58, schrieb K.W.:
> I apologize if this is a bit O.T., but the audience on this list seems to be
> a little more focused than IBM-MAIN for this kind of question....
> I was curious about what would happen in the following scenario:
> Suppose you have a LE enabled "main" program written in C, that is normally
> invoked as a job step via "EXEC PGM=". The program accepts parameters in
> PARM=, but you can also set LE options for it (as with other LE programs) by
> putting them in the PARM before the first slash, e.g.: EXEC
> PGM=CLEPROG,PARM='HEAP(12M)/arg1 arg2".
> Now, suppose I have a small non-LE assembler program (call it "ASMXCTL"),
> that does a "XCTL EP=CLEPROG", passing the original R1 (PARM=)......
> What would happen if I did this? -
> EXEC PGM=ASMXCTL,PARM='HEAP(12M)/arg1 arg2'
> Does the "main" initialization of "CLEPROG" process the LE options as
> Of course, I will have to try it I guess, but I was curious if one of the
> experts here knew what will happen and why...