>
>There are? I see only one:
><quote>
> USS
>See unformatted system service.
></quote>
>

And this very recent APAR describes a problem with a zFS loop in
renaming processing after a BPX1REN ( rename, rename() ) syscall from
Unformatted System Service.   No wonder it caused a loop! <g,d,& r>


  APAR Identifier ...... OA38688      Last Changed ........ 12/04/13
  ZFS LOOP IN RENAME PROCESSING
 
 
  Symptom ...... LP LOOP              Status ........... CLOSED  PER
  Severity ................... 2      Date Closed ......... 12/04/13
  Component .......... 5696EFS00      Duplicate of ........
  Reported Release ......... 3B0      Fixed Release ............ 999
  Component Name ZFS ZSERIES FIL      Special Notice           HIPER
  Current Target Date ..12/06/29      Flags
  SCP ...................               FUNCTIONLOSS
  Platform ............
 
  Status Detail: APARCLOSURE - APAR is being closed.
 
  PE PTF List:
 
  PTF List:
  Release 3B0   : PTF not available yet
  Release 3B0   : Relief is available in the form of: AA38688
  Release 3D0   : PTF not available yet
  Release 3D0   : Relief is available in the form of: BA38688
 
 
  Parent APAR:
  Child APAR list:
 
 
  ERROR DESCRIPTION:
  A loop in zFS rename processing ( ZFSRENAM ) occurred after a
  BPX1REN ( rename, rename() ) syscall from USS.
  The token structure on the rename consists of the source
  directory vnode and the source vnode and the new token structure
  of the target directory vnode and the target vnode. The trace
  has shown the source directory vnode being the same as the
  target vnode which is incorrect and drove the loop.
 
 
  LOCAL FIX:
  None
 
 
  PROBLEM SUMMARY:
  ****************************************************************
  * USERS AFFECTED: zFS Release 11 and zFS Release 13            *
  *                 zFS FMID:HZFS3B0 and FMID:HZFS3D0            *
  ****************************************************************
  * PROBLEM DESCRIPTION: A loop in zFS rename processing         *
  *                      ZFSRENAM  occurred after a              *
  *                      BPX1REN syscall from USS was made.      *
  ****************************************************************
  * RECOMMENDATION: APPLY PTF                                    *
  ****************************************************************
  The vnodes for the operation are passed as input to zFS by the
  USS. The sdevp should not be the same as tdevp.
 
 
  PROBLEM CONCLUSION:
  A validity check of the input vnodes was added. In addition,
  code was added to ctkc_Lock4Vnodes to exit the loop if we are
  in recovery.
 
 
  TEMPORARY FIX:
  *********
  * HIPER *
  *********
 

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN

Reply via email to