VM66839: Z/VM 7.4 CP: HOT FIX - FEATURE 01 FIX 01

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• This APAR contains service for z/VM version 7 release 4 Feature
  Pack 1 Fix Pack 1
  
  
  INCORRECT OUTPUT FROM DIAG26C SUBCODE 20
  ---------------------------------------------------------------
  ------
  Diagnose 26c subcode 20 for response formats 0000 0001 thru
  0000 0005 can return VSWITCH information structures with an
  incorrect structure counter.
  
  The diagnose is provided by the invoker with a response
  buffer. If the response can only partially fit in the response
  buffer then return code 16 is returned along with the size
  of the required buffer. In the identified issue, the last
  VSWITCH structure is the same size as the response buffer.
  A good return code is provided but the response data
  contains an incorrect value.
  
  ACTIVE SERVER ON VSWITCH WITH HIGH RX DISCARD RATE
  ---------------------------------------------------------------
  ------
  CP QUERY VSWITCH DETAILS indicated a steady increase in RX
  Packet Discard count for one user with no clear explanation.
  
  After implementing RedHat OCP on zVM the customer noticed
  significant increases in the number of RX Packets Discarded for
  the server responsible for managing several other dependent
  servers. A review of the dropped packets collected via TRSOURCE
  type LAN trace revealed that these were overflow packets. Most
  packets were being discarded in sets of (approximately) 1000
  packets. The timing of these events lined up with startup times
  for one or more dependent servers. This normally suggests the
  presence of a resource constraint affecting the receiver, but
  monitor data revealed no significant problem in that area.
  
  UNEXPECTED FORCE ARCH REQUIRED FOR LGR
  ---------------------------------------------------------------
  ------
  Customers applying VM66819 APAR across an SSI would eventually
  need to issue FORCE ARCH relocations to move virtual machines
  from systems without the APAR applied back to systems that do
  have VM66819 applied. The problem was caused by the changing
  of an internal system value (SYSMAXD) between processors.
  
  VARIOUS ABENDS SYSTEM ITERATOR SERIALIZATION ISSUE
  ---------------------------------------------------------------
  ------
  Abends: SIT002 HRM002 PRG005 MPC009 ASR091 MCW002 HTT001
  In the iterator, there was an underlying assumption about how
  to check the table level using STEs that was not correct.
  This error can cause various abends. We now obtain the table
  level info from elsewhere. This affects all configurations.
  
  PRG004 ABENDS
  ---------------------------------------------------------------
  ------
  In rare situations, when a guest does not respond to IUCV
  *MONITOR sample notifications, a PRG004 failure may occur.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of z/VM                            *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  ****************************************************************
  * RECOMMENDATION: APPLY PTF                                    *
  ****************************************************************
  This is the APAR for z/VM 7.4 Fix Pack 0101. It contains
  fixes for the following five problems:
  
  Problem 1:
  UNEXPECTED FORCE ARCH REQUIRED FOR LGR
  
  Customers applying APAR VM66819 across an SSI would
  eventually need to issue FORCE ARCH relocations to
  move virtual machines from systems without the APAR
  applied back to systems that do have VM66819 applied.
  The problem was caused by the changing of an internal
  system value (SYSMAXD) between processors.
  
  -----
  
  Problem 2:
  PRG004 ABENDS
  
  In rare situations, when a guest does not respond to
  IUCV *MONITOR sample notifications, a PRG004 failure
  may occur.
  
  -----
  
  Problem 3:
  VARIOUS ABENDS SYSTEM ITERATOR SERIALIZATION ISSUE
  
  ABENDS: SIT002 HRM002 PRG005 MPC009 ASR091 MCW002 HTT001
  
  In the iterator, there was an underlying assumption about
  how to check the table level using STEs that was not
  correct. This error can cause various abends. We now
  obtain the table level info from elsewhere. This affects
  all configurations.
  
  -----
  
  Problem 4:
  ACTIVE SERVER ON VSWITCH WITH HIGH RX DISCARD RATE
  
  CP QUERY VSWITCH DETAILS indicated a steady increase in
  RX Packet Discard count for one user with no clear
  explanation.
  
  After implementing RedHat OCP on zVM the customer noticed
  significant increases in the number of RX Packets
  Discarded for the server responsible for managing several
  other dependent servers. A review of the dropped packets
  collected via TRSOURCE type LAN trace revealed that these
  were overflow packets. Most packets were being discarded
  in sets of (approximately) 1000 packets. The timing of
  these events lined up with startup times for one or more
  dependent servers. This normally suggests the presence of
  a resource constraint affecting the receiver, but monitor
  data revealed no significant problem in that area.
  
  -----
  
  Problem 5:
  INCORRECT OUTPUT FROM DIAG26C SUBCODE 20
  
  Diagnose 26c subcode 20 for response formats 0000 0001
  thru 0000 0005 can return VSWITCH information structures
  with an incorrect structure counter.
  
  The diagnose is provided by the invoker with a response
  buffer. If the response can only partially fit in the
  response buffer then return code 16 is returned along
  with the size of the required buffer. In the identified
  issue, the last VSWITCH structure is the same size as the
  response buffer. A good return code is provided but the
  response data contains an incorrect value.
  

Problem conclusion

• Problem 1:
  UNEXPECTED FORCE ARCH REQUIRED FOR LGR
  
  The setting of a default architectural value in system
  initialization was reset to a previous value. This
  prevents virtual machine relocation from failing due to
  architecture incompatibility checks.
  
  -----
  
  Problem 2:
  PRG004 ABENDS
  
  A missing parameter was added to a subroutine call.
  
  -----
  
  Problem 3:
  VARIOUS ABENDS SYSTEM ITERATOR SERIALIZATION ISSUE
  
  The iterator was updated to check the HDPBK for table
  level information instead of checking the first STE in
  the ST for the table level information. This avoids
  making the wrong decision if serialization is held on the
  first STE.
  
  -----
  
  Problem 4:
  ACTIVE SERVER ON VSWITCH WITH HIGH RX DISCARD RATE
  
  Additional conditions are established to determine
  whether VSWITCH should discard packets waiting to be
  delivered to a guest:
  (1) Was at least one packet accepted into the guest input
      buffers recently (within the last 2 seconds)? If not,
      the queue can be discarded because this server is not
      able to process the incoming stream of data.
  (2) Is the queue size still within a more relaxed
      threshold (four times the normal threshold)?
      If not, the queue can be discarded because this
      server is not able to keep up with the inbound flow.
  If both answers are "yes" then VSWITCH logic will keep
  the pending queue and attempt to deliver it when the
  guest is open for more input data.
  
  -----
  
  Problem 5:
  INCORRECT OUTPUT FROM DIAG26C SUBCODE 20
  
  To resolve the problem, the improper return code and
  possible hanging pointer were addressed. In the event
  that the lengths are equivalent, as described in the
  Summary, then the routine will ensure that a portion in
  the data structure for the output buffer is cleared. This
  will prevent infinite looping in the event that the
  caller does attempt to process the data returned.
  Furthermore, upon equivalence, the return code will
  indicate to the caller that a larger buffer size is
  necessary, instead of indicating that the call was
  successful with no issues (as the routine normally does
  under non-equivalent conditions).
  

Temporary fix

• *********
  * HIPER *
  *********
  FOR RELEASE VM/ESA CP/ESA R740 :
  PREREQ: VM66829
  CO-REQ: NONE
  IF-REQ: NONE
  

Comments

• ×**** PE25/05/27 PTF IN ERROR. SEE APAR VM66844  FOR DESCRIPTION
  ×**** PE25/04/30 FIX IN ERROR. SEE APAR VM66844  FOR DESCRIPTION
  ×**** PE25/04/30 FIX IN ERROR. SEE APAR VM66844  FOR DESCRIPTION
  

APAR Information

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

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

  UM90524

Modules/Macros

• HCPCPE   HCPHDI   HCPHDPBK HCPHDT   HCPHDU   HCPIQR   HCPMNS
  HCPNIDBK HCPSEGTE HCPSWD   HCPZCP   7VMCPR40
  

Fix information

• Fixed component name

  VM CP CP

• Fixed component ID

  568411202

Applicable component levels

• R740 PSY UM90524

     UP25/05/02 P 2502 ¢

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.