You could write a REXX exec that uses the EVALUATE command to validate 
eye-catchers in your control blocks, then use the RUNCHAIN command to 
execute the exec.  Or you could write the whole validation logic in REXX, 
including the chaining part.  If you use RUNCHAIN, each time your exec 
gets driven, the IPCS symbol "X" has the address of the current chain 
element.

something like:

/* REXX  cbval */
ADDRESS "IPCS"

"EVAL X+8 LEN(4) CHAR REXX(STORAGE(EYEC))"

if EYEC <> "CBID" THEN DO
  "NOTE 'Current Control block is invalid' asis "
  return 8
end
return 0


RUNCHAIN ADDR( ) LINK(4) EXEC((cbval))"

(I usually use the second approach, so I don't recall; the runchain 
operands 100%.  Just keep in mind that the EXEC requires the double pair 
of parens.


===============================================
Wayne Driscoll
OMEGAMON DB2 L3 Support/Development
wdrisco(AT)us.ibm.com
===============================================



From:   Dave Day <[email protected]>
To:     [email protected], 
Date:   09/17/2012 12:23 PM
Subject:        [IBM-MAIN] IPCS help needed
Sent by:        IBM Mainframe Discussion List <[email protected]>



     Looking at an SVC dump of my application.  Dump was produced by my 
FRR when the SRB that was running timed out.  05B abend.  I set a timer 
when the SRB initially kicks off.

      The logic in the SRB is processing some data areas, working its 
way from top to bottom in a linked/chained list of these same sized data 
areas.

      The list in the dump has thousands of entries.  One of two 
possibilities.  Either the linked list has an invalid chain ptr, and the 
code is running in circles.  Or, the list is good, and I just needed to 
give the code a longer period of time to run.

     My question.  Is there a way within IPCS I can describe the data 
area containing the chain pointer, and then run some kind of a command 
that would run the list to determine if the list is good or not?

     I'm in the IPCS manuals right now, but if its there, its not 
jumping out at me.  Any help would be appreciated.

      --Dave Day

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



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

Reply via email to