Finding more information about the error

The summary information for the task in error follows the kernel task summary and provides more information about the failure. It provides a storage report for the task, including registers and PSWs, and any data addressed by the registers.

About this task

The PSW is the program status word that is used by the machine hardware to record the address of the current instruction being executed, the addressing mode, and other control information. Figure 1 is the first part of an example of such a storage report for a program check.

Figure 1. Storage report for a task that has experienced a program check - part 1
==KE: Tasks in Error: Error Data follows.

** Task in Error: Error Data follows.

=KE: Error Number:  00000001

 KERRD 2A7FF4B0 KERNEL ERROR DATA

    0000  F0C3F461 C1D2C5C1 018400C4 00000000  C4C6C8C1 D7D3C9F1 00000000 2998BA00  *0C4/AKEA.d.D....DFHAPLI1.....q..*    2A7FF4B0
    0020  00000000 29390700 00000000 2A7FF100  3BF2F080 00000001 00000004 FFFFFFFF  *............."1..20.............*    2A7FF4D0
    0040  079D1001 80000000 00000000 2A1E7742  00040004 00000000 00000000 00000000  *................................*    2A7FF4F0
    0060  90800000 00000000 00000000 C6F4E2C1  00000000 2AA00740 00000000 0008196E  *............F4SA....... .......>*    2A7FF510
    0080  00000000 29992B78 00000000 2A8411D0  00000000 00000002 00000000 29390988  *.....r.......d.}...............h*    2A7FF530
    00A0  00000000 2A1E7564 00000000 2AA00690  00000000 000C0000 00000000 2AA00048  *................................*    2A7FF550
.
.
.
    02A0  40A63D70 0000000E 00010004 C0000000  00000000 2A1E768A 2917B200 0000049C  * w..........{...................*    2A7FF750
    02C0  00000000 00000000 00000000 00000000                                       *................                *    2A7FF770

   Error Code:  0C4/AKEA    Error Type:  PROGRAM_CHECK           Timestamp:  CA434E0E38DF218F

   Date (GMT)   :  03/10/12         Time (GMT)   :  09:52:18.791922
   Date (LOCAL) :  03/10/12         Time (LOCAL) :  10:52:18.791922

   KE_NUM:  0090  KE_TASK:  00000000_2A7FF100  TCA_ADDR:  29390700  DS_TASK:  3BF2F080  XM_TOKEN:  2930A500  TRAN_NO:  00049

=KE: KTCB Details:

   KTCB_ADDR:    2917B200           KTCB_TYPE:    Q                  KTCB_MODE:    QR                 MVS_TCB_ADDR: 009CE9A8
   ACCUM_TIME:   00000000761B5E9B   STIMER_TIME:  000000007D000000   TIMER_STATE:  C0000003           ESTAE_STATE:  00
   ABEND_999:    00

 Error happened in program DFHGA680 at offset 00000322

 Error happened under the CICS RB.

Procedure

  1. Look first in the dump for the following header, which introduces the error report for the task:
    ==KE: Tasks in Error: Error Data follows.
  2. Next, you will see the kernel error number for the task. The kernel assigns error numbers consecutively, starting from 00000001.
    =KE: Error Number:  00000001
    The error number shows the number of program checks and system abends that occurred for this run of CICS®. Not all of them have necessarily resulted in a system dump.
  3. Optional: Some kernel error data follows. Usually, you do not need to find the format of this data, but if you do, refer to KERRD - Kernel error data.
  4. The next area of interest is the interpretation by the kernel of what went wrong. This information includes the error code, the error type, the name of the program that was running, and the offset in the program.
    • The error code shows the system and user completion codes issued when the abend occurred.
    • The error type shows whether the error was associated with, for example, a program check, a system abend, or an internal request for system recovery.
  5. There is a report of where the system has recorded that the error occurred, and the circumstances of the failure. The information has the following general format:
    Error happened in program PPPPPPPP at offset xxxxxxxx
    The program name (PPPPPPPP) and offset (xxxxxxxx) are determined by searching through the CICS loader's control blocks for a program that owned the abending instruction at the time of the abend. If this search does not find such a program, the report shows the following text:
          PROGRAM PPPPPPPP WAS IN CONTROL, BUT THE PSW WAS ELSEWHERE.
    The reported program name (PPPPPPPP) is the program that owns the current kernel stack entry for the abending task. If this text is shown, it might be possible to locate the failing program using the method described in Using the linkage stack to identify the failing module. The failing program name and offset are also displayed in the section of the report immediately after the contents of the registers are reported. This information has the following format: :
            DATA AT PSW: AAAAAAAA   MODULE: PPPPPPPP   OFFSET: XXXXXXXX
    If the failing program cannot be located, the module name and offset are reported as unknown. Possible reasons why the program cannot be located are as follows:
    • The failure occurred in a z/OS® loaded module.
    • The failing program had been released by the CICS loader before the dump was taken.
    • A wild branch in the failing program caused the PSW to point to storage not occupied by a CICS loaded program.
    The accuracy of the program name and offset reported in a formatted dump that was produced as the result of a program executing a wild branch cannot be guaranteed.
  6. After the interpretation by the kernel of the error, one of the following diagnostic messages is shown:
    • Error happened under the CICS RB

      The error was detected either when CICS code was executing, or when an access method called by CICS was running (for example, VSAM or QSAM). The CICS RB is the CICS request block, an MVS™ control block that records the state of the CICS program.

    • Error did not happen under the CICS RB
      This message can be issued in any of the following circumstances:
      • An error occurs in CICS SVC code.
      • An error occurs in a CICS z/OS Communications Server exit.
      • CICS detects a runaway task during the execution of an MVS service request.
      • An error occurs during the execution of an SVC request that was made by CICS or an access method invoked by CICS.
  7. After either of these messages, data that is likely to be related to the problem is shown. The data shown depends on whether the error happened under the CICS request block.
    • If the error happened under the CICS RB, the error data in the task storage report is based on values in the PSW and the CICS registers at the time the error was detected. Figure 1 shows the storage report for a task that failed when a program check was detected. It illustrates the error data supplied when an error happens under the CICS RB.
    • If the error did not happen under the CICS RB, for example when CICS was calling an MVS service, you get data based on two sets of registers and PSWs. The registers and PSW of the CICS RB at the time of the error constitute one set. The registers and PSW of the RB in which the error occurred constitute the other set. This data will relate, very probably, to the execution of an SVC routine called by CICS. The error might have occurred, however, during an IRB interrupt or in an SRB. You can confirm whether this has happened by checking flags KERNEL_ERROR_IRB and KERNEL_ERROR_SRB_MODE.

    For more information about this data in the task storage report, see The storage addressed by the CICS registers and PSW.



dfhs10y.html | Timestamp icon Last updated: Monday, 5 June 2017