PM92253: WMQ Z/OS CHIN HIGH CPU CAUSED BY TIGHT LOOP IN CSQXBPQY.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• The high CPU usage is due to a tight loop in
  CSQXBPQY.
  You can see from the dump that the loop is in
  code which is attempting to run the chain of
  active storage buffers for the current
  dispatcher.
  However, the pointer to the current buffer being
  processed is on the free buffer chain.
  .
  The issue is caused by a buffer being moved to
  the free chain by another TCB while it was the
  current buffer being examined on this dispatcher.
  The result of this is that the CSQXBPQY starts
  processing the free chain rather than the active
  chain, and so never reaches the expected end of
  chain value.
  .
  Once this process is looping, all other channels
  running on the same  dispatcher are unable to run.
  Therefore this issue can effect other other
  workload impact.
  .
  In the SYSTRACE for the CHIN, you will see
  the following:
  Note: There will be many EXT 1005.  In the
        DUMP that was looked at, there were
        936 "EXT  1005".
  
  .
  DSP      00000000_3D6075DE  00000000 477E1338
                              477E23CC  0
           CSQXBPQY UK73196 + x'746' in CSQXBPQY+0
           07850000 80000000
  EXT 1005 00000000_3D607610  00001005           0
                    CSQXBPQY UK73196 + x'778'
        in CSQXBPQY+0
           07851000 80000000                     0
  EXT 1005 00000000_3D6075DA  00001005           0
                    CSQXBPQY UK73196 + x'742'
        in CSQXBPQY+0
           07850000 80000000                     0
  DSP      00000000_3D6075DA  00000000 477E1338
                    477E23CC  0
                    CSQXBPQY UK73196 + x'742' in
        CSQXBPQY+0
           07850000 80000000
   EXT 1005 00000000_3D6075DA  00001005          0
                     CSQXBPQY UK73196 + x'742' in
         CSQXBPQY+0
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of WebSphere MQ for z/OS Version 7 *
  *                 Release 0 Modification 1 and Release 1       *
  *                 Modification 0.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: High CPU caused by a loop in module     *
  *                      CSQXBPQY in the channel initiator       *
  *                      address space.                          *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  Under rare circumstances, when a buffer is being allocated and
  the code is scanning the chain of available buffers, another
  task may manipulate the current entry moving it to from the free
  to the active chain. This prevents the original scan from
  detecting the end of the list and the code to scan for an
  available buffer will loop.
  

Problem conclusion

• Buffer query code in CSQXBPQY will now take a lock on the pool
  of buffers when scanning the chain to prevent an entry from
  moving while being processed.
  010Y
  100Y
  CSQXBPQY
  

Temporary fix

• *********
  * HIPER *
  *********
  

Comments

APAR Information

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

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

  UK96303 UK96304

Modules/Macros

• CSQXBPQY
  

Fix information

• Fixed component name

  WMQ Z/OS V7

• Fixed component ID

  5655R3600

Applicable component levels

• R010 PSY UK96303

     UP13/09/27 P F309 Ž

• R100 PSY UK96304

     UP13/09/27 P F309 Ž

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.