2 Replies Latest reply on Dec 16, 2011 8:41 PM by Ivanigorovich-Oracle

    Pstack from core file with no usable info

      One of our applications is coring and the customer is sending a pstack from the core file that looks like:

      core 'core' of 16879: /path/to/executable
      ----------------------------- lwp# 1 --------------
      ffffffff7a50711c ???????? (6, 0, ffffffff7dfffff90, 0, 0, 0)
      ---------------------------- lwp# 2 --------------------------------------
      ** zombie (exited, not detached, not yet joined) **
      ------------------------------ lwp# 4 -----------------------------------
      ffffffff620db950 ???????? (0, 0, 0, 0, ffffffff70f01ac0, ffffffff6224a240)
      ----------------------------- lwp# 11 ------------------------------------------
      ffffffff620d7410 ???????? (ffffffff63aec0a0, ffffffff6caec080, ffffffff5b1fbba0, 4, 16df10, ffffffff6eb0be00)

      The customer says they are generating the pstack from the core on the same computer and OS it was generated on.

      Could this be an indication of a corrupted stack?

      An incomplete core file?

      Can mdb be used to get additional info from the core?

      What else can I have the customer do with the core file so they can provide more information?

      I cannot get access to the actual core file.

      The core is generated from a system running Solaris 10, Studio 11 C++ code.