Explanation
The dispatcher of tasks and service routines detected
an error. A hexadecimal reason code in register 15 explains the error:
- Code
- Explanation
- 04
- A completed routine that ran under a service request block (SRB)
held a system lock when it returned to the dispatcher.
- 0C
- A routine that ran under a suspended service request block (SSRB)
held the cross memory local (CML) lock of a failing address space.
The ASCBFAIL bit is on. The ASCBFAIL bit is in the address space control
block (ASCB) of the failing address space.
- 10
- The SRBASCB field of an SRB or an SSRB points to an ASCB that
has an erroneous acronym.
- 14
- A task holds the CML lock of a failing address space. The ASCBFAIL
bit for the failing address space is on.
- 18
- Status could not be saved for the current task due to an error
in the dispatcher.
- 1C
- A task attempted to reenter the dispatcher through the SRB entry
point.
- 20
- A routine that ran under an SRB attempted to reenter the dispatcher
through a task entry point.
- 24
- Status could not be saved for the current SRB due to an error
in the dispatcher.
- 28
- Task could not be dispatched because the CML lock was stolen.
- 2C
- IEA0DS1B was entered not in wait mode.
System action
The system abends the task or SRB and
may write a logrec data set error record.
Operator response
If the system programmer asks for an SVC dump,
set the following SLIP trap:
SLIP SET,ENABLE,COMP=066
System programmer response
Do the following:
- For reason code X'04', go to step 3.
- For any other reason code, go to step 6.
- For reason code X'04', if a logrec data set
error record is not written with the abend, ask the operator to set
a SLIP trap to obtain an SVC dump. Format the dump to see:
- Register 2 through register 8
- Register 10 through register 13
These are the SRB routine's registers at the time it gave
up control.
- Check these registers to see if one looks like a base
register, as follows. Look for a group of registers containing addresses
near the same place in storage. If you find such a group of registers,
look for the register with the lowest address. This is possibly a
base register. Identify the module that is using this register as
its base register.
- Determine which lock was held when the problem occurred.
The current lock held string can be found in two different places:
- Register 14 at the time of error
- In the PSACLHS field of the SVCE D trace entry for abend X'066'
- Search problem reporting data bases for a fix for the
problem. If no fix exists, and an SVC dump has not be obtained, ask
the operator to set a SLIP trap to obtain an SVC dump. Contact the IBM® Support Center. Provide:
- The SVC dump
- The system trace
- The logrec data set error record, if available
Source