II07747: MVS/ESA AS A V=R GUEST ENTERS DISABLED SPIN LOOP DUE TO VM/ESA PUTTING A LOGICAL PROCESSOR IN A STOPPED STATE.

APAR status

• INTRAN

Error description

• MVS/ESA system receives MSGIEE331A processor (x) is in an
  excessive disabled spin loop waiting for CPU in stopped state.
  This message can be seen if the MVS system is running as a
  VM guest and the VM system has stopped the logical processor
  due to a soft abend on the VM system.  There may or may not be
  ABEND071 logrec records, The operator may be prompted to
  reply to MSGIEE331A or it may go directly into ACR processing
  regardless of the EXSPATxx parmlib member parameters for
  excessive spin conditions.  The ACR may not work whether
  entered automatically or by reply.
       VM console logs will have a message along these lines:
  HCPABNxxxxE A system soft abend yyyyyy has occurred on CPU 000x
  and the virtual CPU has been put into a stopped state.
      This has been seen on MVS/ESA R422 and R430 systems and
  VM/ESA R110 and R120 systems.
  VERIFICATION PROCEDURES:
  From a STANDALONE DUMP:
      Do IPCS STATUS CPU REGS:
      On the MVS system the stopped processor will be in a NO
  WORK WAIT, or it may look like it is in valid code.  The non-
  stopped processors may be in IEAVERI and/or ISNDWTOR.
      If, for example, there are 3 logical processors and
  processor 2 is stopped, find the SVT by doing IPCS CBF CVT
  then get the address at CVT + x'364' CVTSVT and do
  IPCS CBF aaaaaaaa. STR(SVT) to format the SVT.  At SVT + x'1A0'
  is SVTSPCP.  It will look like 42420000.  Each byte represents
  a processor, the first byte is for CP00, since it has 42 there,
  it is spinning for CP02.  The second byte is for CP01, it is
  also spinning for CP02.  If CP01 were stopped, you would see
  SVTSPCP as 41004100, indicating both CP00 and CP02 were spinning
  for CP01.
      If you do not know what the MSGIEE331A message was but
  suspect this problem do IPCS VERBX TRACE and look from the
  bottom up for an 'ss' trace entry.  The second entry to the
  right of the psw is the psaeparm which points to the MSSF
  buffer.  Browse storage at this address and go to the
  address pointed to by the first word in this buffer.  It
  should point to an area of storage that contains the MSGIEE331A
  message contents that will show which processor was stopped.
      Then do IPCS VERBX TRACE and find the last entry for the
  stopped processor.  The logical processor number is at the
  far left of the trace entry, the physical processor number is
  the farthest right digit in the timestamp at the far right of
  the trace entry.  If you see the physical processor number
  used again after the logical processor has stopped, this
  indicates you are dealing with a logical processor that has
  been stopped by VM and not a hardware problem where a physical
  processor stopped.
     Check the VM CONSOLE LOG for the time period preceeding the
  spin and look for indications of a soft abend that caused VM
  to stop the logical processor.  If you find this is the case
  please contact VM Level 2 support for assistance.
  (compid 568411202)  If this is not your problem, or you need
  further assistance with the disabled spin debugging, please
  contact MVS Supervisor Component Level 2 support for assistance
  (compid 5752SC1C5).
  

Local fix

• The VM support personnel can issue a CP command to start
  the logically stopped processor under VM.  This may be too late
  for MVS, however, depending on how far along ACR processing
  has proceeded, if it looks like the ACR processing completed
  successfully, you may still be able to configure the logical
  processor online to MVS after it has been started under VM.
  

Problem summary

Problem conclusion

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:

Fix information

Applicable component levels