Operation Manual

SMI handler messages
SMI handler is special system code that provides system monitoring functions.
When the SMI handler detects a microprocessor error, the SMI handler message
identifies the affected processor using a ProcID number or an APICID number. The
following table provides a cross reference between the physical microprocessor
number, the ProcID and the APICID.
Physical Microprocessor
Number
1
ProcID APICID
1 01 0000 or 0001
2 02 0002 or 0003
3 03 0004 or 0005
4 04 0006 or 0007
1
See “Microprocessor board connectors, jumpers and LEDs” on page 68 to find the
physical microprocessor numbering
Following is an example of an SMI handler message:
---------------------------------
Date/Time: 2002/05/08 03:35:30
DMI Type: 00
Source: SMI Hdlr
Error Code: 00150700 PERR: Slave signaled parity error
Error Code:
Error Data: Bus = 02, Slot = 01, VendID = 1014, DevID = 0302
Error Data:
---------------------------------
This message indicates that a PCI Bus 02 parity error occurred at the recorded
time. All of the SMI handler message codes are listed in the following table.
Note: The message codes in the following table do not include leading or trailing
zeros.
Note: See Chapter 7, “Parts listing, Types 8861, 8862,” on page 139 to determine which components are
replaceable by the customer and which components must be replaced by an IBM field service technician.
SMI message code Suggested action
1500 (Single bit error occurred: begin
scrub)
v Information only; run memory diagnostic programs when
multiple entries accumulate.
1501 (Multi-bit error occurred: forcing NMI)
v Run memory diagnostic programs; then, replace indicated
memory DIMM.
1502 (SERR: Addr. or Spec. Cyc. DPE)
1. Check Operating System event log.
2. Run adapter diagnostics, if available.
3. Adapter.
4. PCI switch card assembly.
5. I/O board.
106 IBM xSeries 365 Types 8861, 8862: Hardware Maintenance Manual and Troubleshooting Guide