IV88438: CRASH IN JFS2 SNAPSHOT CODE IN SECOWIOWAIT() APPLIES TO AIX 7100-03

APAR status

• Closed as program error.

Error description

• A rare race condition can cause a crash with stack
  similar to:
  (4)> f
  pvthread+042700 STACK:
  [00308324]seCOWIOWait+000404 (F1000A003AE20000,
  F1000A0EE01CC090 [??])
  [0030DCCC]seCOW+000B4C (??, ??)
  [0030CC68]seCOD+0000A8 (??, ??, ??, ??)
  [0032C03C]txFreeMap+00029C (??, ??, ??, ??)
  [0035EFFC]xtFreeMap+00021C (??, ??, ??)
  [0032A420]txUpdateMap+000240 (??, ??)
  [0032D4D0]txCommit+0006D0 (??, ??, ??, ??)
  [0034EE50]j2_remove+000470 (??, ??, ??, ??)
  [0066DE78]vnop_remove+000438 (??, ??, ??, ??)
  [006C0A60]kunlinkat+000460 (FFFFFFFEFFFFFFFE,
  0000000111CAED11,
     0000000000000000, 0000000000000000)
  [00003938]syscall+000230 ()
  [kdb_get_virtual_memory] no real storage @
  FFFFFFFFFFFB840
  [100047098]0000000100047098 ()
  [kdb_read_mem] no real storage @ FFFFFFFFFFF8D60
  
  ==============================
  with crash occurring in seCOWIOWait function.  A second
  thread
  will also be seen named snapshot where the snapshot
  command
  is being run to eithor delete a snapshot or create a new
  snapshot.
  
  The snapshot command deletes a snapshot resource needed
  by
  the thread that crashes.
  

Local fix

Problem summary

• A rare race condition can cause a crash with stack
  similar to:
  
  (4)> f
  pvthread+042700 STACK:
   00308324 seCOWIOWait+000404 (F1000A003AE20000,
  F1000A0EE01CC090  ?? )
   0030DCCC seCOW+000B4C (??, ??)
   0030CC68 seCOD+0000A8 (??, ??, ??, ??)
   0032C03C txFreeMap+00029C (??, ??, ??, ??)
   0035EFFC xtFreeMap+00021C (??, ??, ??)
   0032A420 txUpdateMap+000240 (??, ??)
   0032D4D0 txCommit+0006D0 (??, ??, ??, ??)
   0034EE50 j2_remove+000470 (??, ??, ??, ??)
   0066DE78 vnop_remove+000438 (??, ??, ??, ??)
   006C0A60 kunlinkat+000460 (FFFFFFFEFFFFFFFE,
  0000000111CAED11,
     0000000000000000, 0000000000000000)
   00003938 syscall+000230 ()
   kdb_get_virtual_memory  no real storage @
  FFFFFFFFFFFB840
   100047098 0000000100047098 ()
   kdb_read_mem  no real storage @ FFFFFFFFFFF8D60
  
  ==============================
  with crash occurring in seCOWIOWait function.  A second
  thread will also be seen named snapshot where the snapshot
  command is being run to eithor delete a snapshot or create
  a new snapshot.
  
  The snapshot command deletes a snapshot resource needed
  by the thread that crashes.
  

Problem conclusion

• Added the check to avoid sync issue between snapshot delete
  command and file remove which will avoid system crash.
  

Temporary fix

Comments

APAR Information

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

  IV79798

• 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 U872346

     UP17/01/19 I 1000