Endevor

We currently use IRXEXEC in our production exit7 to verify that the package approver is not the same person who created/cast the package. The rexx used to do this is:/* rexx */arg rusr /* create cast approve */upper rusrparse var rusr rusr1 rusr2 rusr3 junk"EXECIO * diskr ndvckapp (finis stem sam."qusr.='none'do i=1 to sam.0qusr=strip(substr(sam.i,2,5))qemail=strip(substr(sam.i,11,50))qusr.qusr=qemailendreply='BAD'if qusr.rusr1=qusr.rusr3 then reply='SAME' /* create approver */if qusr.rusr2=qusr.rusr3 then reply='SAME' /* cast approver */exit replyI am using exit2 to test whether the following is doable with the IRXEXEC approach:1) use DSNREXX to get to a distributed database and pass back data to the Endevor exit that can be stored somewhere in Endevor (CA suggested that we could use the package notes). The DSNREXX part is working OK.2) address ispexec to display a panel for the user to enter data that can update/replace the package notes. I can't make this work (variable pool issue?) and have tried native rexx say/pull while "messy" this does work, but has re-introduced the "2nd time thru" issue. Obviously this would be done in the package exit7 not exit2.The "2nd time thru" issue is that the second time the exit is used in an Endevor session this message is displayed: Invalid or missing EXECBLK parameter passed to the exec load routine.The REXX exec cannot be interpreted.The REXX exec cannot be loaded. I have not been able to determine what is corrupting the link to IRXEXEC in the exit - nor has George Guntzert (CA) who experiences the same problem.Any ideas?  

Have you looked at using IRXJCL instead of IRXEXEC ?   IBM doc says "The IRXJCL interface is easier to use, but it is not as flexible as the IRXEXEC interface.".     IRXJCL works quite well for me - calling Rexx from COBOL and assembler exits. IBM gives this example...                             01 ARGUMENT.
                                03 ARG-SIZE                                       pic 9(2) comp.
                                03 ARG-CHAR                                       pic x(8).                          77 PGM-NAME                                             pic x(8).                               move "HELLO   3" to ARG-CHAR.
                              move 8 to arg-size.
                    *       Call "IRXJCL" in order to execute the REXX procedure
                              move "IRXJCL" to PGM-NAME.
                              CALL PGM-NAME         USING ARGUMENT.
                    *       Display the return code.
                              display "Return code after   call is " RETURN-CODE.  You can add your parameter(s) to ARGUMENT and increase its length.  

I meanwhile tested the REXX invocation via IRXJCL   from a BAL EXIT02
here,   and it indeed works like a charm !Also the second invocation of the REXX is no problem, and even starting
another ISPF panel in the REXX seems to be working fine ……don’t know about the
ISPF variables part though.  What the problem with the original code using IRXEXEC was I still don’t know,
It might be something simple like making sure the dsects for irxexec are
Getmain’ed/Initialized and clean.
But I would expect one needs additional logic to obtain the Rexx
context/environment one is in, and act depending on that.Someone pointed  to two sample programs on the CBT183 tape; REXXTRY and FASTPATH,
Although   they do not exactly contain what is needed here, they give some
Idea what is involved in using irxexec.

Hi,is it a problem which exist furthermore?  We use EXIT7 invoking a Rexx with IRXEXEC and it works.If there is an interest I can post something here...  B.S.