RBD/COBOL programs abend ASRA on CALL to ELAESAB abend

[alice....@terra.com.br]

Hello Listers,

Anyone running Rational generated COBOL programs under zVSE5.2?

Same program generated under zVSE4.2 runs fine on zVSE5.2, but abends if regenerated under zVSE5.2.

Abend occurs at the following CALL statement at line 3331:

*EXEC CICS LINK PROGRAM("ELAESABD") END-EXEC             
  003329                     MOVE ".........03127   " TO DFHEIV0                
  003330                     MOVE "ELAESABD" TO DFHC0080           
  003331                    CALL "DFHEI1" USING DFHEIV0  DFHC0080               
  003332                    EXIT. 
 
IUI online problem determination says:

Program 'ELAESABD' starts at X'842FB148'.
Program Check PSW ==> 00000000 00000000
REG15= 00000000
 
CEMT I PROG(ELAESABD)
Prog(ELAESABD) Len(0005202) Ass Pro Ena Pri     Ced  
   Res(000) Use(0000000016) Bel Uex Ful 

- LIBDEF search chain verified (there is no old ELAESABD version being picked)
- Rational library is the same (MSHP BACKUP from zVSE4.2 then MSHP INSTALLed on zVSE5.2)
- DFHCSD is the same 
 
Any sugestions are very much appreciated!  
 
Best Regards.
Alice  
   

[Juha Vuori]

Hello Alice,

Did you refresh Rational Cobol Runtime installation in VSE 5.2 after FSU, with the latest PTFs and
recustomizations?
We did recustomize too - I can't remember the exact reason now, but we haven't had any similar problems.
Iirc, there was a bug in one of the customization scripts, which is now corrected in one of the PTFs.


By the way, the VSE Runtime will soon include support for calling generated EGL batch subprograms
from (native) Cobol batch programs. We needed that flexibility in application development, and IBM
was kindly willing to provide that.
Just in case you have ever considered this kind of appl structures..

Regards,
Juha

> _______________________________________________
> VSE-L mailing list
> VS...@lists.lehigh.edu
> https://lists.lehigh.edu/mailman/listinfo/vse-l
>

_______________________________________________
VSE-L mailing list
VS...@lists.lehigh.edu
https://lists.lehigh.edu/mailman/listinfo/vse-l

[Martin Truebner]

Hello Alice,

did you check your compile listing?

The LNKEDT step is of interest here-

Was the LNKEDT able to resolve DFHEI1?

When I was a systemprogrammer (and helping application programmers in
tricky situation) we had a display on the door which said-

No dump, no problem
No compile+link listing with RC < 5 = don't even think about asking

Martin Trübner; everything around "PoOps of z/arch"

[Mick Poil]

Alice,

I don't remember exactly, but I think that some people had issues after re-compiling COBOL programs at 5.2 that had worked OK before. Maybe check COBOL and LE PTF levels in case this is relevant here.

Mike

[alice....@terra.com.br]

  
Thank you all!   :o)
 
There were some EXTRN for ELAnnnnn modules.
 
LINK job generated by RATIONAL has ELALIB=' '
Before migration, LIBDEF proc had PRD2.VAGEN in it, so ELAnnnnn modules were being linked from VisualAGE lib instead of RATIONAL lib!
 
Best Regards,
Alice

lehigh.edu] På vegne av alice....@terra.com.br
Sendt: 1. september 2016 03:58
Til: vs...@lists.lehigh.edu
Emne: RBD/COBOL programs abend ASRA on CALL to ELAESAB abend