IV67682: CRASH WHEN JFS2 WITH NO LOG BECOMES DIRTY AND GETS FORCE-UNMOUNT APPLIES TO AIX 7100-03

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• System may crash when JFS2 with no log becomes dirty and
  gets force-unmounted.
  
  CRASH INFORMATION:
  
  CPU 3 CSA F00000002FF47600 at time of crash, error code
  for LEDs:
  30000000
  
  pvthread+038400 STACK:
  [00009514].simple_lock+000014 ()
  [00188ED4]bmInodeFlush@AF123_69+000214 (??, ??, ??, ??,
  ??, ??, ??, ??)
  [001851B8]bmInodeFlush+000038 (??)
  [0037B7CC]pagerUnbind@AF66_41+00012C (??, ??, ??, ??)
  [0037A8AC]pagerUnbind+00002C (??, ??, ??)
  [0017B49C]iRecycle+00019C (??, ??, ??)
  [00177D54]iUnmount+000314 (??, ??, ??)
  [003DC4D4]umount_fs_forced+000274 (??, ??)
  [003DD480]j2_umount+000180 (??, ??, ??)
  [005A6958]vfs_unmount+000098 (??, ??, ??)
  [006549DC]kunmount+0000DC (??, ??, ??)
  [00655228]uvmount+000208 (??, ??)
  [F1000000C03AB75C]my_uvmount+000428 (0000005000000050,
  0000000100000001)
  [00003888]mfspurr_sc_flih01+0000E4 ()
  [kdb_get_virtual_memory] no real storage @ 2FF21E58
  [10005C08]10005C08 ()
  [kdb_read_mem] no real storage @ FFFFFFFFFFF63C0
  

Local fix

• N/A
  

Problem summary

• System may crash when JFS2 with no log becomes dirty and
  gets force-unmounted.
  
  CRASH INFORMATION:
  
  CPU 3 CSA F00000002FF47600 at time of crash, error code
  for LEDs:
  30000000
  
  pvthread+038400 STACK:
   00009514 .simple_lock+000014 ()
   00188ED4 bmInodeFlush@AF123_69+000214 (??, ??, ??, ??,
  ??, ??, ??, ??)
   001851B8 bmInodeFlush+000038 (??)
   0037B7CC pagerUnbind@AF66_41+00012C (??, ??, ??, ??)
   0037A8AC pagerUnbind+00002C (??, ??, ??)
   0017B49C iRecycle+00019C (??, ??, ??)
   00177D54 iUnmount+000314 (??, ??, ??)
   003DC4D4 umount_fs_forced+000274 (??, ??)
   003DD480 j2_umount+000180 (??, ??, ??)
   005A6958 vfs_unmount+000098 (??, ??, ??)
   006549DC kunmount+0000DC (??, ??, ??)
   00655228 uvmount+000208 (??, ??)
   F1000000C03AB75C my_uvmount+000428 (0000005000000050,
  0000000100000001)
   00003888 mfspurr_sc_flih01+0000E4 ()
   kdb_get_virtual_memory  no real storage @ 2FF21E58
   10005C08 10005C08 ()
   kdb_read_mem  no real storage @ FFFFFFFFFFF63C0
  

Problem conclusion

• pb_lsn indicates that the buf is on the log sync list.
  If it isn't on the list then there is no need for the lock.
  pb_lsn should never be set if there is no log.
  
  Hence moving lock on log under pb_lsn check to avoid crash.
  

Temporary fix

Comments

• 6100-08 - use AIX APAR IV69965
  6100-09 - use AIX APAR IV67666
  7100-02 - use AIX APAR IV70215
  7100-03 - use AIX APAR IV67682
  

APAR Information

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

  IV52940

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

Fix information

• Fixed component name

  AIX V7.1

• Fixed component ID

  5765H4000

Applicable component levels

• R710 PSY U865837

     UP15/05/19 I 1000

PTF to Fileset Mapping

• U865837 bos.mp64 7.1.3.45