PH08267: ABEND0C6 IN DFHNCMN AFTER A BATCH PROGRAM CALLED THE NAMED COUNTER SERVER

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A batch program called the Named Counter Server directly via the
  CALL DFHNCTR interface and very intermittently an abend0C6
  occurs in the DFHNCMN+84A9 at PTF level UI54059.
  .
  In the batch job log :
  .
  AXMSC0031I Connection to server DFHNC.name1 has been opened.
  Abend0C6 occurs here *
  AXMSC0041I Connection to server DFHNC.name1 has been closed.
  From a SLIP COMP=0C6 dump, the systrace shows that the Named
  Counter Server had been called by the batch address space
  via PC  2501 issued from the batch asid.
  .
  The PIC6 happens at 26A084A9 = DFHNCMN+84A9 label AXMPGFRE and
  was issued because the instruction's address is
  odd (it ends in ...9).  This is trapped abend in DFHNCMN because
  an invalid R1 was detected.
  RTM2WA SUMMARY
                                --------------
  +001C  Completion code                    840C6000
  +008C  Abending program name/SVRB address 00000000 00000000
  +0094  Abending program addr              00000000
  
         GPRs at time of error
       0-3  D95E5000  00000000  26A04E30  26A38120
       4-7  26A1AC68  00000000  26A1AC08  26A1AC28
       8-11 26A08466  A6A04EAC  26A1A930  00000000
      12-15 26A38010  26A38058  A6A01528  26A04E30
  
  +06D8  PSW at time of error: 07841000 80000000 00000000 26A084A9
  +0084  Instruction Length Code: 0004 interruption Code:0006
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: Abend S0C6 followed by S0C1 in          *
  *                      DFHNCNM.                                *
  ****************************************************************
  Batch Named Counter Services work was being issued from
  multiple jobs. As the jobs disconnected from the NCS server,
  requests interleaved within the AXM disconnect logic in
  AXMXMEND. This allowed a purge of an XMSAV area which is in the
  process of being reused for a new request. This resulted in
  two threads being active against the same XMSAV, with
  unpredictable consequences such as the 0C6 and 0C1 abends
  observed by the customer.
  
  Keywords: abendS0C6 abendS0C1
  

Problem conclusion

• AXMXM has been changed to clear SCDET pointer (XMSAVDET) before
  the  area can be reused. This addresses the small timing
  loophole where code in XMBEGDSC can attempt to purge an XMSAV
  area which is in the process of being reused for a new request.
    In addition, AXMLF has been changed to reset the next
  available byte pointer to its initial value within the reset
  loop logic that resets LIFO storage stack to its initial state.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

  UI63215 UI63216 UI63217

Modules/Macros

• AXMLF    AXMXM
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R000 PSY UI63216

     UP19/05/25 P F905

• R100 PSY UI63215

     UP19/05/25 P F905

• R200 PSY UI63217

     UP19/05/30 P F905

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.