WAIT 064: Wait state 064 — error during nucleus initialization (NIP)

The system stopped during IPL because NIP hit an error before recovery services were ready. The reason code in the PSW tells you what kind.

What happened

During IPL the system entered a wait state and the HMC shows a wait state code of 064. The console may show messages just before it, for example IEA304W for a program check. The system did not finish initializing.

What the operator records (illustrative)Illustrative example

PSW: 000A0000 80000000 00000000 00009064
Wait state code: 064   Reason code: 009
IEA304W  ... (program check details issued before the wait)

What it means

The nucleus initialization program found a software or hardware error and tried to pass it to the recovery termination manager (RTM). Because RTM is not ready that early in IPL, NIP traps the request and loads a wait state instead. The three hex digits in front of the wait code in the PSW are the reason code. Common ones:

Reason	Meaning
001	A module branched to abnormal-end processing.
003	RESTART was pressed instead of the TOD key.
004	A machine check — for example an I/O interrupt from a device not defined in the IOCP, or a resource changing state during IPL.
005	An SVC was issued in the wrong mode. Often a symptom of an earlier abend — e.g. while reading PARMLIB, loading a wrong module, or I/O errors on the sysres volume.
006	A paging I/O error.
009	A program check; message IEA304W has the details.

IBM lists further reason codes (002, 007, 008, 00A–010) in MVS System Codes. This is a restartable wait state.

Typical causes

A PARMLIB change with an error picked up by this IPL.
A wrong or mismatched load module in the system libraries after maintenance.
An I/O configuration that does not match the hardware (wrong IODF/IOCDS).
I/O errors on the system residence volume.
A warm or quick start after a failed cold start (reason 009).

Symptoms

IPL stops; the HMC shows a disabled wait with code 064.
Console messages up to the failing point, then nothing.
No address spaces beyond the early system ones.

Where to look

The wait state PSW on the HMC — record it exactly.
The last console messages before the wait.
IPL information and recent changes: PARMLIB members, load parameter, IODF, maintenance.
A stand-alone dump, if the system programmer asks for one; IPCS IPLDATA STATUS shows the last initialization routine that ran.

How to diagnose

Record the PSW and decode the reason code.
Collect console messages leading up to the wait.
For 005, IBM's guidance is to check whether the last SVC was SVC 13 (abend); in a stand-alone dump SYSTRACE ALL then FIND *SVC PREV locates it.
For 009, read IEA304W and the program-check old PSW; in a dump, the LCCA holds the registers (CBFORMAT LCCAx).
Search IBM's problem databases with the wait code, reason code and module name.

How to fix

Undo or correct whatever changed: fall back to the previous PARMLIB member or load parameter, use the correct IODF/IOCDS, or restore the previous system libraries. For reason 003, simply re-IPL. For 009 after a failed cold start, IBM's operator guidance is to restart and reply CLPA to IEA101A. If no known fix exists, open a case with IBM with the PSW, messages and dump.

Wait state 064 — error during nucleus initialization (NIP)