>  For HVCOMMON, I would think that SDUMP should be turning on
>the common storage bit in the dump record prefix, and IPCS should
>be respecting that and allowing you to access the storage via any
>ASID.

Big thanks (again) to Jim Mulder for looking at this. SDUMP does indeed 
everything it should and the command *should* have worked. Jim was able to 
reproduce my problem, though. It was caused by the ADCD system parmlib setup 
(done by IBM). ADCD systems are delivered with this parmlib concat:
USER.PARMLIB     
ADCD.Z113.PARMLIB
SYS1.PARMLIB     

SYS1.PARMLIB is SMP/E maintained (or what is considered SMP/E maintained with 
ADCD systems), ADCD.*.PARMLIB is what comes with each ADCD release, and 
USER.PARMLIB is intended for the ADCD user to make changes (I had limited my 
changes to user.parmlib so far). At some point in the past  the ADCD people 
apparenly felt compelled to copy each and every parmlib member from 
SYS1.PARMLIB to ADCD.*.PARMLIB, but they never bothered to update 
ADCD.*.PARMLIB when an update was done in sys1.parmlib. Hence the exclusive 
mixture of crap as far as IPCS (and ctrace) goes. While the ADCD system comes 
with a very good setup to get into IPCS even for dumps from older releases 
(they correctly allocate old needed data sets and even kept sys1.parmlib from 
those releases around), on the current system very old and obsolete parmlib 
members were used for IPCS analysis.

After I deleted everything IPCS related from ADCD.*.PARMLIB, the TCPIPCS 
command worked as expected, and the error messages are all gone, as I am  now 
using my current release sys1.parmlib members.

I guess my next step will be parmlib cleanup. I found an IEALODxx member (I 
didn't know that there ever existed such a thing) with this comment in it:
THIS MEMBER CONTAINS A DUMMY ENTRY,    IEFBR14.  ITS PURPOSE IS TO STOP THE   
MESSAGE, 'IEA301I IEALOD00 NOT FOUND IN SYS1.PARMLIB' FROM COMING OUT ON THE   
CONSOLE AT IPL TIME. 

I also found an empty TSOKEY00 (in sys1.parmlib, no less!) with this comment:
THIS MEMBER CONTAINS THE TSO/VTAM TIME-SHARING PARAMETERS.  THIS MEMBER IS 
SPECIFIED BY THE   MBR KEYWORD OF THE TSO PROCEDURE USED TO START THE TERMINAL 
CONTROL ADDRESS SPACE (TCAS) OR BY THE MEMBER OPERAND OF THE START PARAMETER    
  PASSED WHEN TCAS IS STARTED.  

So all you out there running on an ADCD system, beware! 

Barbara

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to