Digital PDFs

EK-454AA-MG-1

May 1992

326 pages

Original

11MB

Document:	KA675/KA680/KA690 CPU System Maintenance (Final Draft)
Order Number:	EK-454AA-MG
Revision:	1
Pages:	326
Original Filename:

OCR Text

KA675/KA680/KA690 CPU
System Maintenance
Order Number EK-454AA-MG-001

Digital Equipment Corporation
Maynard. Massachusetts

First Printing, May 1992

The information in this document is subject to change without notice and should not be
construed as a commitment by Digital Equipment Corporation.
Digital Equipment Corporation assumes no responsibility for any elTOl'S that may appear in
this document.
The software, if any, descn"bed in this document is furnished under a license and may be used
or copied only in accordance with the terms of such license. No responsibility is assumed
for the use or reliability of software or equipment that is not supplied by Digital Equipment
Corporation or its affiliated companies.
Restricted Rights: Use, duplication. or disclosure by the U.s. Government is subject to
restrictions as set forth in subparagraph (c)(l)(ii) oftbe Rights in Technical Data and Computer
Software clause at DFARS 252.227-7013.
Copyright e Digital Equipment Corporation, 1992. All Rights Reserved.
The Reader's Comments form at the end of this document requests your critical evaluation to
assist in preparing future documentation.
The fonowing are trademarks of Digital Equipment Corporation: CompacTape, ex, DDCMP,

DEC, DECconnect, DECctirect, DECnet,· DECscan, DECserlel', DECUS. DECwindows,
DELNI, DEMl'R, DESQA, DESTA, DSRVB. DSSI, IVAX, KDA, KLESI, MicroVAX. MSCP,

Q-bus, Q22-bus, RA. RQDX, RRD4O, SDI, TbinWire. TK, TMSCP, TQK5O, TQK70, TSV05.

TU, ULTRIX, UNIBUS, VAX. VAX. 4000, VAX. DOCUMENT, VAXcluster, VAXELN. VAXlab,
VAX.server, VAXsimPLUS, VMS, VT, and the DIGITAL loge.

FCC NOTICE: The equipment described in this manual generates, uses, and may emit radio
frequency energy. The equipment has been type tested. and found to comply with the limits for

a Class A computing device pursaant to Subpart J of Part 15 of FCC Rules, which are designed
to provide reasonable protection against such radio frequency interference when operated in
a commercial environment. Operation of this equipment in a residential area may cause
interference, in which case the user at his own expense may be required to take measures to
correct the interference.
'

MlrSl678

This dcc:u.ment was prepared using VAX DOCUMENT, Version 2.0.

Contents
Preface'

Chapter 1 System Maintenance Strategy
1.1
1.2
1.3
1.4

Service Delivery Methodology ........................
Product Service Tools and Utilities ....................
Information Services ...............................
Field Feedback. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1-1
1-2
1-4
1~

Chapter 2 CPU System Overview
2.1 CPU Module Features ............................. . 2-1
2.2 MS690 Memory Modules ........................... . 2-5
BA440 Enclosure Components ....................... " 2-6
2.3
H3604 Console Module ........................... . 2-6
2.3.1
System Control Panel (SCP) ...........•..•........ 2-12
2.3.2
BA440 Backplane ............................... . 2-14
2.3.3
Power Supply . . . . . . . . . . . . . . . . . . . ............... . 2-15
2.3.4
System Airflow. . . . . ............................ . 2-17
2.3.5

Chapter 3 System Setup and Configuration
3.1 CPU and Memory Module Order ..................... .
Installing Add-On MS690 Memory Modules .......... .
3.1.1
3.2 General Module Order for Q-Bus Options .............. .
3.3 Recommended Module Order of Q-bus Options .......... .
3.4 Mass Storage Options (Internal) ..................... .
3.5 System Expansion ................................ .
Mass Storage Expanders ......................... .
3.5.1
Q-Bus Expanders ............................... .
3.5.2
Control Power Bus for Expanders .................. .
3.5.3

3-1
3-2
3-4
3-5
3-6

3-8
3-8
3-8
3-9
iii

3.5.4
Adding Options to the System Enclosure. . . . . . . . . . . . .. 3-10
3.6
DSSI VAXclusters ................................. 3-13
3.6.1
DSSI VAXcluster Configuration Rules. . . . . . . . . . . . . . .. 3-15
3.7
Finnware Commands and Utilities Used in System
Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3-20
3.7.1
Examining System Configuration. . . . . . . . . . . . . . . . . . .. 3-20
3.7.2
Using the CONFIGURE Command to Detennine CSR
Addresses for Q-Bus Modules. . . . . . . . . . . . . . . . . . . . . .. 3-22
3.7.3
Setting and Examining Parameters for DSSI Devices . . .. 3-24
3.7.3.1
DSSI Device Parameters. . . . . . . . . . . . . . . . . . . . . . .. 3--24
How VMS Uses the DSSI Device Parameters ........ 3-25
3.7.3.2
3.7.3.3
Entering the DUP Driver Utility from Console Mode .. 3-31
Entering the DUP Driver Utility from VMS ......... 3-32
3.7.3.4
Setting Allocation Class . . . . . . . . . . . . . . . . . . . . . . . .. 3-33
3.7.3.5
3.7.3.6
Setting Unit Number . . . . . . . . . . . . . . . . . . . . . . . . . .. 3-34
3.7.3.7
Setting Node Name ........... . . . . . . . . . . . . . . . .. 3-37
3.7.3.8
Setting System ID . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3--37
3.7.3.9
Exiting the DUP Driver Utility . . . . . . . . . . . . . . . . . .. 3--38
Write-Protecting an RF35 ISE . . . . . . . . . . . . . . . . . . . . .. 3-40
3.7.4
3.7.4.1
Software Write-Protect for RF-Series ISEs. . . . . . . . . .. 3-40
3.7.4.2
Hardware Write-Protect For RF35 ISEs. . . . . . . . . . . .. 3-41
Setting System Parameters: Boot Defaults, Bootfiags, Halt
3.7.5
and Restart Action. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8-44
3.7.5.1
Setting the Boot Default ........................ 3-44
3.7.5.2
Setting Boot Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 3-46
3.7.5.3
Setting the Halt Action ......................... 3-47

Chapter 4

System Initialization and Acceptance Testing
(Normal Operation)

4.1
Basic Initialization Flow ........................... .
4.2
Power-On Self-Tests (POST) ......................... .
4.2.1
Power-Up Tests for Kernel ........................ .
4.2.2
Power-Up Tests for Q-Bus Options .................. .
4.2.3
Power-Up Tests for Mass Storage Devices ............ .
4.3 CPU ROM-Based Diagnostics ....................... .
Diagnostic Tests ................................ .
4.3.1

4-1
4-4
4-4

4-7
4-8

4-8
4-9

4.3.2
Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.4 Basic Acceptance Test Procedure. . . . . . . . . . . . . . . . . . . . ..
4.5 Machine State on PowereUp . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6 Main Memory Layout and State ......................
4.6.1
Reserved Main Memory. . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.1.1
PFN Bitmap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.1.2
Scatter/Gather map . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.1.3
Firmware "Scratch Memory" .....................
4.6.2
Contents of Main Memory . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.3
Memory Controller Registers . . . . . . . . . . . . . . . . . . . . . ..
4.6.4
On-Chip Cache. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.5
Translation Buffer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.6.6
Halt-Protected Space . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.7 Operating System Bootstrap .........................
Preparing for the Bootstrap. . . . . . . . . . . . . . . . . . . . . . ..
4.7.1
4.7.2
Primary Bootstrap Procedures (VMB) .•••. '" • . . • . . . • ••
4.7.3
Device Dependent Secondary Bootstrap Procedures. . . . ..
4.7.3.1
Disk and Tape Bootstrap Procedure ........ ~ . . . . . ..
4.7.3.2
PROM Bootstrap Procedure. . . . . . . . . . . . . . . . . . . . ..
4.7.3.3
MOP Ethernet Functions and Network Bootstrap
Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.7.3.4
Network "Listening" . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.8 Operating System Restart . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.8.1
Locating the RPB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

Chapter 5

4-10
4-15
4-20
4-20
4-21
4-21
4-22
4-22
4-22
4-23
4-23
4-23
4-23
4-23
4-24
4-26
4-30
4-30
4-31
4-32
4-33
4-39
4-40

System Troub!eshooting and Diagnostics

5.1 Basic Troubleshooting Flow. . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Product Fault Management and Symptom-Directed Diagnosis
5.2.1
General Exception and Interrupt Handling . . . . . . . . . . . .
5.2.2
VMS Error Handling. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.3
VMS Error Logging and Event Log Entry Format. . . . . . .
5.2.4
VMS Event Record Translation. ... . . . . . . . . . . . . . . . . ..
5.2.5
Interpreting CPU Faults Using ANALYZE/ERROR . . . . ..
5.2.6
Interpreting Memory Faults Using ANALYZE/ERROR ...
5.2.6.1
Un correctable ECC Errors. . . . . . . . . . . . . . . . . . . . . ..

5-1
5-4
5-4
&-5

5-7
5-15
5-16
5-18
5-19

5.2.6.2
5.2.7
5.2.8

Correctable ECC Errors. . . . . . . . . . . . . . . . . . . . . . . ..
Interpreting System Bus Faults Using ANALYZE/ERROR
Interpreting DMA <=> Host Transaction Faults Using
ANALYZEIERROR ...............................
5.2.9
VAXsimPLUS and System-Initiated Call Logging (SICL)
Support ............. . . . . . . . . . . . . . . . . . . . . . . . . ..
5.2.9.1
Converting the SICL Service Request MEL File ......
5.2.9.2
VAXsimPLUS Installation Tips. . . . . . . . . . . . . . . . . ..
5.2.9.3
VAXsimPLUS Post-Installation Tips. . . . . . . . . . . . . ..
5.2.10 . Repair Data for Returning FRUs . . . . . . . . . . . . . . . . . . ..
5.3
Interpreting Power-On Self Test (POST) and ROM-Based
Diagnostic (RBD) Failures . . . . . . . . . . . . . . . . . . . . . . . . . ..
5.3.1
FE Utility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5.3.2
Overriding Halt Protection. . . . . . . . . . . . . . . . . . . . . . . ..
5.3.3
Isolating Memory Failures . . . . . . . . . . . . . . . . . . . . . . . ..
5.4 Testing DSSI Storage Devices ........................
5.5
Using MOP Ethernet Functions to Isolate Failures . . . . . . ..
5.6
Interpreting User Environmental Test Package (UETP) VMS
Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5.6.1
Interpreting UETP Output ........................
5.6.1.1
UETP Log Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5.6.1.2
Possible UETP Errors. . . . . . . . . . . . . . . . . . . . . . . . ..
5.7 Using Loopback Tests to Isolate Failures. . . . . . . . . . . . . . ..
5.7.1
Testing the Console Port ..........................
5.7.2
Embedded nSSI Loopback Testing. . . . . . . . . . . . . . . . . ..
5.7.3
Embedded Ethernet Loopback Testing. . . . . . . . . . . . . . ..
5.7.4
Q-Bus Option Loopback Testing .....................

Chapter 6
6.1
6.2
6.3
6.4

5-22
5-27
5-29
5-31
5-36
5-37
5-38
5-40
5-40
5-53
5-54
5-54
5-57
5-59
5-62
~2

5-62
5-63
5-64
~6

5-67
5-68
~9

FEPROM Firmware Update

Preparing the Processor for an FEPROM Update. . . . . . . . .
Updating Firmware Via Ethernet .....................
Updating Firmware Via Tape. . . . . . . . . . . . . . . . . . . . . . . . .
FEPROM Update Error Messages. . . . . . . . . . . . . . . . . . . . .

6-2
6-3
6-6
6-9

Appendix A

KA675/KA680/KA690 Firmware Commands

A.I Console I/O Mode Control Characters ................. . A-I
Command Syntax ............................... . A-2
A.I.I
Address Specifiers .............................. . A-3
A. 1.2
Symbolic Addresses ............................. . A-3
A. 1.3
A.I.4
Console Numeric Expression Radix Specifiers. . . . . . . . .. A-6
A. 1.5
Console Command Qualifiers. . . . . . . . . . . . . . . . . . . . . .. A-6
A. 1.6
Console Command Keywords . . . . . . . . . . . . . . . . . . . . . .. A-S
A.2 Console Commands ................................ A-9
A.2.1
BOOT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-IO
A.2.2
CONFIGURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-II
A.2.3
CONTINUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-13
A..2.4
DEPOSIT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-IS
A2.5
EXAMINE ................ ~ .................... A-14
A.2.6
FIND. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-15
A.2.7
HALT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-16
A.2.8
HELP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-16
A.2.9
INITIALIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-18
A.2.I0 MOVE ........................................ A-19
A.2.I1 ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-20
A.2.12 REPE.M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-22
A.2.13 SEARCH........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. A-22
A.2.14 SET.......................................... A-24
A.2.I5 SHOW ........................................ A-28
A.2.16 START........................................ A-32
A.2.I7 TEST......................................... A-82
A.2.18 UNJAM ....................................... A-83
A.2.19 X-Binary Load and Unload ....................... A-83
A.2.20 ! (Comment) .................................... A-85

vii

Appendix B
B.l
B.2
B.3
B.4
B.5
B.6

Address Assignments

KA6751KA.6801KA690 General Local Address Space Map
B-1
KA6751KA.6801KA690 Detailed Local Address Space Map . .. B-2
External, Internal Processor Registers. . . . . . . . . . . . . . . . .. B-8
Global Q22-bus Address Space Map
B-8
Processor Registers ......
B-9
IPR Address Space Decoding ......
B-16
0

Appendix C

•

••

•

•••••••••

•

••

ROM Partitioning

e.l Firmware EPRO~1 Layout, . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.1.l
System Identification Registers. . . . . . . . . . . . . . . . . . . ..
C.l.l.l
PR$_SID (IPR 62) . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
C.1.1.2
SIE (20040004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
C.1.2
Call-Back Entry Points ...........................
C.1.2.l
CP$GETCHARJU . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
C.1.2.2
CP$MSG_OUT_NOLF_R4 ....
C.1.2.3
CP$READ_WTH_PRMPT_R4 .. 0.................
C.1.3
Boot Information Pointers. . . . . . . . . . . . . . . . . . . . . . . ..
0

Appendix D
Dol
D.2
D.3

• • • • • • • • • • • • • • • • • •'

C-5
C-6
C-7
C-7

Data Structures and Memory Layout

Halt Dispatch State Machine. . . . . . . . . . . . . . . . . . . . . . . ..
RPB ..................
VMB Argument List..... . . . .... . . . .. . .. ..... .. ... ..
0

•

Appendix E

Configurable Machine State

Appendix F

NVRAM Partitioning

•

••

F.l SSC RAM Layout ..........
F.1.l
Public Data Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . .
F.1.2
Console Program MailBox (CPMBX) .................
F.l.3
Firmware Stack. ... .. .. . .... . . . . . . . ... . .. ... .. ..
F.l.4
Diagnostic State. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
F.l.5
USER Area ....................................
0

viii

C-::.
C-3
C-3
C-4
C-5

•

D-l
D--5
D-9

F-l
F-l
F-2
F-3
F-3
F-4

Appendix G

MOP Counters

Appendix H

Programming the KFQSA Adapter

Appendix I

Error Messages

1.1
1.2
1.3
1.4

Machine Check Register Dump. . . . . . . . . . . . . . . . . . . . . . .
Halt Code Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
VMB Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Console Error Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appendix J

I-I
I-I
1-3
1-4

Related Documents

Glossary

Index
Examples
SHOW DSSI Display (Embedded DSSI). . . . . . . . . . . . . . . ..
SHOW UQSSP Display CKFQSA-Based DSSI) ............
3-3 Accessing the DUP Driver Utility From Console Mode
(Embedded DSSi). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
3-4 Accessing the DUP Driver Utility From Console Mode
(KFQSA-Based DSSI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
3-5 Accessing the DUP Driver Utility From VMS ............
3-6 Setting Allocation Class for a Specified Device ...........
3-7 Setting a Unit Number for a Specified Device . . . . . . . . . . ..
3-8 Changing a Node Name for a Specified Device ............
3-9 Changing a System ID for a Specified Device ............
3-10 Exiting the DUP Driver Utility for a Specified Device. . . . ..
3-11 SHOW DSSI Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
3-12 SHOW UQSSP Display (KF'QSA-Based DSSI) . . . . . . . . . . ..

3-1
3-2

3-30
3-31

3-32
3-32
3-33
3-34
3-35
3-37
3-38
3-39
3-39
3-40

Setting Hardware Write-Protection Through Firmware . . . ..
Setting Hardware Write-Protection Through VMS . . . . . . . ..
Language Selection Menu ....... . . . . . . . . . . . . . . . . . . . .
N onnal Diagnostic Countdown. . . . . . . . . . . . . . . . . . . . . . . .
Successful Power-Up to List of Bootable Devices. . . . . . . . . .
4-4 Test 9E. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5-1 Error Log Entry Indicating CPU Error. . . . . . . . . . . . . . . ..
5-2 SHOW ERROR Display Using VMS. . . . . . . . . . . . . . . . . . ..
5-3 Error Log Entry Indicating Uncorrectable ECC Error . . . . ..
5-4 SHOW MEMORY Display Under VMS .................
5-5 U sing .AJ.~ALYZE/SYSTE~1 to Check the Physical Address in
Memory for a Replaced Page . . . . . . . . . . . . . . . . . . . . . . . ..
5-6 Error Log Entry Indicating Correctable ECC Error. . . . . . ..
5-7 Error Log Entry Indicating Q-Bus Error. . . . . . . . . . . . . . ..
5-8 Error Log Entry Indicating Polled Error . . . . . . . . . . . . . . ..
5-9 Device Attention Entry .............................
5-10 SICL Service Request with Appended MEL File... . . . . . ..
5-11 Sample Output with Errors. . . . . . . . . . . . . . . . . . . . . . . . ..
5-12 FE Utility Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5-13 Running DRVTST .................................
5-14 Running DRVEXR .................................
6-1 FEPROM Update Via Ethernet . . . . . . . . . . . . . . . . . . . . . . .
6-2 FEPROM Update Via Tape ..........................

3-13
3-14
4-1
4-2
4-3

3-43
3-44
4-3
4-4
4-7
4-12
5-17
5-18
5-20
5-21

5-22
5-25
5-28
5-29
~1

5-41
5-53
5-58
5-59
6-5
6-8

Figures
2-1
2-2
2-3
2-4
2-5

KA6751KA6801KA690 CPU Module Component Side . . . . . . .
KA6751KA6801KA690 Kernel System Functional Diagram ..
KA6751KA6801KA690 CPU Module Block Diagram . . . . . . . .
Ratchet Handles for CPU and Memory Modules . . . . . . . . . .
H3604 Console Module (Front). . . . . . . . . . . . . . . . . . . . . . . .
2-6 H3604 Console Module (Back) . . . . . . . . . . . . . . . . . . . . . . ..
2-7 System Control Panel ..............................
2-8 BA440 Backplane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
2--9 Power Supply. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
2-10 Fans............................................
2-11 Fan Speed Control (FSC) Jumper Location ..............

2-2
2-4

2--5
2-6
2-7
2-10
2-12
2-14
2-15
2-18
2-19

Memory Module Ratchet Handles .................... . 3-3
Storage Configuration Example ...................... . 3-7
3-3 Sample Power Bus Configuration ..................... . 3-9
3-4 VAX 4000 Model 500 Configuration Worksheet .......... . 3-11
3-5 DSSI Cabling for a Generic Two-System DSSI VAXcluster
Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14
3-6 ~o-System DSSI VAXcluster ....................... . 3-18
3-7 Expanded 'l\vo-System DSSI VAXcluster ............... . 3-19
3-8 VMS Operating System Requires Unique Unit Numbers for
DSSI Devices .................................... . 3-27
3-9 Sample DSSI Busses for an Expanded VAX 4000 Model 500
System ......................................... . 3-28
3-10 Attaching a MSCP Unit Number Label to the Device Front
Panel .......................................... . 3-36
4-1 Console Banner .................................. . 4-5
4-2 ~Iemory Layout After Power-Up Diagnostics ............ . 4-21
4-3 Memory Layout prior to VMB Entry .................. . 4-26
4-4 Memory Layout at VMB Exit ........................ . 4-29
4-5 Boot Block Format ................................ . 4-31
4-6 Locating the Restart Parameter Block ................. . 4-40
5-1 Event Log Entry Format ........................... . 5-9
5-2 Machine Check Stack. Frame Subpacket ............... . 5-10
5-3 Processor Register Subpacket ....................... . 5-11
5-4 Memory Subpack.et for ECC Memory Errors ............. . 5-12
5-5 Memory SBE Reduction Subpacket (Correctable Memory
Errors) ......................................... . 5-12
5-6 eRn Entry Subpacket Header; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ;;; 5-13
5-7 Correctable Read Data (eRD) Entry ................... 5-14
5-8 Trigger Flow for the VAXsi'mPLUS MOflitor. . . . . . . . . . . . .. 5-33
5-9 Five-Level VAXsimPLUS Monitor Display ............... 5-85
5-10 H3604 Console Module Fuses ........................ 5-66
6-1 Firmware Update Utility Layout. . . . . . . . . . . . . . . . . . . . . . 6-2
6-2 W4 Jumper Setting for Updating Finnware. . . . . . . . . . . . . . 6-3
C-l KA6751KA6801KA690 FE PROM Layout. . . . . . . . . . . . . . . .. C-2
C-2 SIn: System Identification Register ................... C-3
C-3 SIE: System Identification Extension (20040004) . . . . . . . .. C-4

3-1
3-2

C-8

F-l
F-2
F-8
F-4

KA6751KA6801KA690 sse NVRAM Layout...... .... ....
NVRO (20140400) : Console Program MailBoX (CPMBX) ...
NVRI (20140401) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
NVR2 (20140402) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

F-l
F-2
F-8
F-8

Tables
2-1
2-2

KA6751KA6801KA690 CPU Module Components.. . ....... 2-8
H3604 Console Module Controls and Indicators. . . . . . . . . . . 2-8 ....
2-3 H3604 Console Module (Back) ........................ 2-10
2-4 System Control Panel Controls and Indicators. . . . . . . . . . .. 2-13
~
Hfi~'14 Power Suppiy ~witches, Controls,
and Indicators ......................................... 2-15
3-1 BA440 Module Order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
3-2 Power Requirements e , ' •
3-12
3-3 Boot Devices Supported by the KA6751KA6801KA690 ...... 3-46
3-4 Virtual Memory Bootstrap (VMB) Boot Flags ............ 3-47
3-5 Actions Taken on a Halt ............................ 3-48
4-1 Language Inquiry on Power-Up or Reset. . . . . . . . . . . . . . . . 4-2
4-2 LED Codes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-6
4-3 Scripts Available to Customer Services ................. 4-14
4-4 Signature Field Values. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4-17
4-5 Network Maintenance Operations Summary ............ , 4-35
4-6
Supported MOP Messages .......................... 4-36
4-7 MOP Multicast Addresses and Protocol Specifiers. . . . . . . .. 4-38
5-1 Console Terminal/Console Module Problems ............. 5-3
5-2 Power Supply Status Indicators . . . . . . . . . . . . . . . . . . . . . . . 5-3
5-3 VMS Error Handler Entry Types . . . . . . . . . . . . . . . . . . . . . . 5-8
5-4 Conditions That Trigger VAXsimPLUS Notification and
Updating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5-32
5-5 Five-Level VAXsimPLUS Monitor Screen Displays ........ 5-34
5-6 Machine Check Exception During Executive . . . . . . . . . . . .. 5-42
5-7 Exception During Executive with No Parameters ......... 5-43
5-8 Other Exceptions with Parameters, No Machine Check. . . .. 5-43
5-9 KA6751KA6801KA690 Console Displays As Pointers to FRUs 5-45
5-10 H3604 Console Module Fuses ........................ 5-65
5-11 Loopback Connectors for Common Devices .............. 5-70
0

xii

•••••••••

•

••

A-I Console Symbolic Addresses. . . . . . . . . . . . . . . .......... . A=3
Symbolic Addresses Used in Any Address Space ......... . A-6
Console Radix Specifiers ........................... . A-6
A-4 Console Command Qualifiers ........................ . A-7
A-5 Command Keywords by Type ........................ . A-8
A-6. Console Command Summary ........................ . A-8
B-1 Processor Registers ............................... . B-9
B-2 IPR Address Space Decoding ........................ . B-16
C-l System Identification Register ....................... . C-3
C-2 System Identification Extension ...................... . C-4
C-3 Call-Back Entry Points ............................ . C-5
D-l Firmware State Transition Table ..................... . D-2
D-2 Restart Parameter Block Fields ...................... . D-5
D-3 VMB Argument List ............................... . D-9
A-2
A-3

F-l

F-2

F-2
F-3
0-1 MOP Counter Block. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
H-l Preferred KFQSA Switch Settings. . . . . . . . . . . . . . . . . . . ..

F-3
F-3
0-1
B-1

1-1
1-2
1-3

1-2
1-3
1-4

HALT Messages ....... . . . . . . . . . . . . . . . . . . . . . . . . . . .
VMB Error Messages .......... . . . . . . . . . . . . . . . . . . . .
Console Error Messages ............................

xiii

Preface
This guide describes the procedures and tests used ·to maintain and
troubleshoot VAX 4000 Model 400, 500, and 600 systems, which use the
KA675, KA6S0, and KA690 kernel, respectively.
System

Kernel

VAX 4000 Model 400

KA675

VAX 4000 Model 500

KA680

VAX 4000 Model 600

KA690

Intended Audience
This guide is intended for use by Digital Equipment Corporation Service
personnel and qualified self-maintenance customers.

Warnings, Cautions, Notes
Warnings, cautions, and notes appear throughout this guide. They have
the following meanings:
WABNING

Provides information to prevent personal injury.

CAUTION

Provides information to prevent damage to equipment or software.

NOTE

Provides general information about the current topic.

Conventions
A system prompt and a command in boldface, uppercase type, for example,
»>SHOW OSSI, shows that the user enters the command at the system
prompt.

Chapter 1

System Maintenance Strategy
Any successful maintenance strategy is predicated on the proper
understanding and use of information services, service tools, service
support and escalation procedures, and field feedback. This chapter lists
the various service tools, information services, and service delivery methods
used in system maintenance.

1.1 Service Delivery Methodology
Before beginning any maintenance operation, you should be familiar with
the following:
1. The site agreement
2. Your local and area geography support and escalation procedures
3. Your Digital Services product delivery plan

Service delivery methods are part of the service support and escalation
procedure. When appropriate, remote services should be part of the initial
system installation. Methods of service delivery include the following:
•

Local support

•

Remote call screening

Remete diagnosis and syst.em initiated service requests (using
DSNLink, SICL, MDS01, modem, etc.)

The recommended system h~sta11ation includes:
1. Hardware installation and acceptance testing.

Acceptance testing
(Chapter 4) includes running ROM-based diagnostics and running
MDM to test Q-bus options.

2. Software installation and acceptance testing. For example, using VMS
Factory Installed Software (FIS), and then acceptance testing with
UETP.
3. Installation of the Symptom-Directed Diagnosis (SDD) toolkit
(V~.xsimPLUS a..'ld SPRA..R) and remot.e services tools and equipment
(this includes installing DSNlink, modems, etc., and enabling SICL).
System Maintenance Strategy

1-1

When the installation is complete, the system should be able to dial out
using SICL, and the Digital Service Center should be able to call into
the system. Refer to your remote service delivery strategy.
If your service delivery methodology is not followed, service expenses for
any product could be excessive.

1.2 Product Service Tools and Utilities
This section lists the array of service tools and utilities available for
acceptance testing, diagnosis, and overall serviceability; and it provides
recommendations as for their use.
VMS Error HandlinglLogging
VMS provides recovery from errors, fault handling, and event
logging. The Error Report Formatter (ERF) provides bit-to-text
.translation of the event logs for interpretation.
RECOMMENDED USE: Analysis of error logs is the primary
method of diagnosis and fault isolation. If the system is up, or
the customer allows the service engineer to bring the system up,
this information should be looked at first. Refer to Section 5.2 for
information on Product Fault Management and Symptom-Directed
Diagnosis.
Symptom-Directed Diagnostic (SDD) Tools (VAXsimPLUS)
SDD tools are used primarily for notification of the existence of
errors that have reached a critical threshold. SDD tools must be
installed during system installation or as soon as product support
is provided. snn tools are not bundled with VMS.
RECOMMENDED USE: Used primarily for onsite notification to
the user via mail or to a remote Digital support center via System
Initiated Call Logging (SICL). Refer to Section 5.2.9 for information
on VAXsimPLUS and SICL.
ROM-Based Diagnostics
ROM-based diagnostics have significant advantages:
•

Load time is virtually nonexistent.

•

The boot path is more reliable.

•

Diagnosis is done in a more primitive state.

1-2 KA675/KA680/KA690 CPU System Maintenance

RECOMMENDED USE: The CPU ROM-based diagnostic facility
is the primary means of offiine testing and diagnosis of the
CPU, memory, Ethernet, and DSSI subsystems. The ROM-based
diagnostics are used in the acceptance test procedures (Section 4.4)
when installing a system, adding a memory module, or replacing
the following: CPU module, memory module(s), backplane, nSSI
device, or H3604 console module. Use the ROM-based diagnostic
error messages in Table 5-9 to isolate FRUs.
Firmware Console Commands
Several commands and utilities are needed in configuring a system
and setting and examining system and device parameters. For
example, the CONFIGURE command is used to determine the
proper CSR addresses for modules; the SHOW MEMORY, SHOW
DSSI, and SHOW QBUS commands are used to examine the
configuration and memory error status; and the SET HOST
command is used to access the DUP driver to configure DSS!
parameters.
RECOMMENDED USE: Use console commands to configure
the system and in setting and examining device parameters.
Refer to Section 3.7 for information on firmware commands and
utilities. Appendix A provides information on all available console
commands.
Option LEDs During Power-Up
Many options and modules have LEDs that display pass/fail selftest results.
RECOMMENDED USE: Monitor option and module LEDs during
power-up to see if they pass their sen-tests. Refer to Sections 4.2.2
and 4.2.3 for infonnation on power-up tests for Q-bus and mass
storage devices. For more information on individual options, refer
to your Microsystems Options manual.
Operating System Exercisers (VMS UETP)
The User Environment Test Package (UETP) is a VMS software
package designed to test whether the VMS operating system is
installed correctly.
RECOMMENDED USE: Use UETP as part of acceptance testing to
ensure that VMS is correctly installed. UETP is also used to stress

System Maintenance Strategy

1-3

test the user's environment and configuration by simulating system
operation under heavy loads.
MicroVAX Diagnostic Monitor (MDM)
The loadable diagnostic MDM requires a mmlmum of Release
136 to support VAX 4000 Model 400/500/600 systems. Consult
your MicroVAX Diagnostic Monitor User~ Guide for instructions
on running MDM.
RECOMMENDED USE: MDM is used primarily for testing Q-bus
options.
Loopback Test-c::

Internal and externalloopback tests can be used to isolate a failure
by testing segments of a particular control or data path. The
loopback tests are a subset of the ROM-based diagnostics and MDM
diagnostics.
RECOMMENDED USE: Loopback. tests can be used to isolate
problems with the console port, nSS! adapters, Ethernet controller,
and many common Q-bus options. Refer to Section 5.7 for
instructions on performing loopback tests.
Crash Dumps
For fatal errors, VMS will save the contents of memory to a crash
dump file, e.g. fatal bugchecks.
RECOMMENDED USE: Crash dump file analysis should be
performed by support. Saving a crash dump file for analysis
requires proper system settings. Refer to your VMS documentation
for instructions.

1.3 Information Services
Digital Service engineers may access several information resources,
including advanced database applications, online training courses, and
remote diagnosis tools. A brief description of some of these resources
follows.
Technical Information Management Architecture (TIMA)
TIMA is used by Digital Services to deliver technical and reference
information to its service engineers. One of the main benefits of
TIMA is the pooling of worldwide knowledge and expertise. Both

1-4 KA675/KA6801KA690 CPU System Maintenance

service and customer documentation for VAX 4000 systems are
available on TIMA.
Entry Systems Service Information Kits
Service documentation containing information on enclosures, CPUs,
and options, makes up the Entry Systems Service Information Kit.
The manual you are reading is part of the kit. Refer to your Guide
to Entry Systems Service Information Kits (EK-276A*-MI) for more
information.
Training
Computer Based Instruction (CBI) and lecture lab courses are
available from the Digital training center:
•

VAX 4000 Model 500 System Installation and Troubleshooting
(CBI course, EY-I089E-EO (applicable for VAX 4000 Model 400
1500/600 systems)).

MicroV~v:.. Installation and Troubleshooting (Lecture lab course,
EY-940BE-LO)

Digital Services Product Delivery Plan (Hardware or Software)
The Product Delivery Plan documents Digital Services delivery
commitments. The plan is the communications vehicle used
among the various groups responsible for ensuring consistency
between Digital Services delivery strategies and engineering
product strategies.
Blitzes
Technical updates are "blitzed" to the field using mail and TWA
Storage and Retrieval System (STARS)
Stars is a worldwide database for storing and retrieving technical
information. The STARS databases, which contain more than
150,000 entries, are updated daily.
Using STARS, a service specialist can quickly retrieve the most upto-date technical information via DSNlink or DSIN.
VAX Notes
VAX Notes is a worldwide notes file.
DSNlink
DSNlink software application lets the Digital Services Center
communicate electronically with the customer site. DSNlink serves
as the platform for the delivery of electronic services.

System Maintenance Strategy

1-5

1.4 Field Feedback
Providing the proper feedback to the corporation is essential in closing the
loop on any service call. Consider the following when completing a service
call:
•

Repair tags should be filled out accurately and with as much symptom
information as possible so that repair centers can fix a problem.

•

Call closeout information for Labor Activity Reporting System (LARS)
or Call-Handling and Management Planning (CHAMP) needs to be
accurate.

•

The site maintenance log, whether hardcopy or electronic, should
"",,-~ • .:~~ ~ ",.l....,.., __ ~Al",.
j:lA. U Y .... \.4~ Q.. ...., • .a..,L v"'........ "" .. w

".c
~,,~ 9'\~~".........-".:I ...... """.: ....... ~'",.,""",.!l'\
\01 .... w ..
j:Ji." ...
u. .... """ .... ......
w"""....
~

~V4

~.\,or""".

1-6 KA675/KA680IKA690 CPU System Maintenance

Chapter 2

CPU System Overview
This chapter provides an overview of the components that make up KA675
/KA680/KA690-based systems. These components are listed below:
•

CPU: KA675 (IAOO2-CA), KA680 (IA002-BA), or KA690 (IAOO2-AA)

•

MS690 memory modules

•

BA440 enclosure components
H3604 console module
System control panel (SCP)
BA440 backplanes
Power supply
Fans

CAUTION: Static electricity can damage integrated circuits. Always use a
grounded wrist strap (pN 2~11762-O0) and grounded work surface when
working with the internal parts of a computer system.

2.1 CPU Module Features
The KA6751KA6801KA690 CPUs are quad-height VAX processor modules
that use the Q22-bus and DSS! bus. The CPUs are used in the following
systems:
.
System

CPU

VAX 4000 Model 400

KA67S

VAX 4000 Model 500

KA680

VAX 4000 Model 600

KA690

The CPU module is designed for use in high-speed, real-time applications
and for multiuser, multitasking environments. The KA6751KA6891KA690
employ multiple levels of cache memory to maximize performance. See
Figure 2-1 for a view of the major chips, LEDs, and connectors. See
Figure 2-2 and Figure 2-3 for biock diagrams of the major functions.

CPU System Overview

2-1

The CPU module and MS690 memory modules combine to form the CPU
Imemory subsystem that uses DSSI busses to communicate with mass
storage devices, the Q22-bus to communicate with I/O devices, and the
Ethernet to communicate across the network.
The CPU module.is configured as an arbiter CPU on the Q22-bus, where
it arbitrates bus mastership and fields anyon-board interrupt requests.
Figure 2-1: KA675/KA680/KA690 CPU Module Component Side
Console Connector, J2

RUnLED~/
Diagnostic LEOs

IE-Net ROM

OC541

SGEC

BCache
(Tag
Store)

NVAX

~~
0C542

SHAC

OC246
OC243

NCA

ROMs

B ~

Obit RAMs

OC244

NMC

Irmware

ssc

B-CACHE
(Data Store)

OC527

COBIC

§J
SHAC

-jBackplane Connector, J 1
MLo.oo7693

2-2 KA675/KA680IKA690 CPU System Maintenance

Table 2-1: KA675/KA680/KA690 CPU Module Components
Components

Function

DC246 (NVAX)

Central processor unit. Contains a 64-entry translation
buffer integral floating.point unit, 2·KB virtual instruction
stream cache (VIC), 8·KB physical instruction and data
stream primary cache (P-cache), and backup cache control
and error coITeCtion code (ECC)
KA675: Central processor unit has 16·ns cycle time and
the virtual instrnction stream cache is disabled
KA680: Central processor unit has l4·ns cycle time
KA690: Central processor unit has l2·ns cycle time

Backup cache RAMs

KA675, KASSO: 128-KB backup cache (B-cache)
KA690: 512·KB backup cache (B-cache)

DC243 (NCA)

NDAL to CDAL 110 bus interface chip

DC244 (NMC)

Main memory controller (also provides ECC protection)

DC527 (CQBIC)

Q22-bus interface

DC541 (SGEC)

Ethernet interface

Ethernet Station Address ROM

Provides unique hardware address

DC542 (SHAC)

DSSI interface chips (2)

DC511 (SSC)

System support chip

DC509 (CLK)

Clock

Firmware ROMs

Four resident firmware chips, each 128 K by 8 bits of
FLASH programmable EPROMS for a total of 512 KB.

Obit RAMs

The ECC protected ownership-bit RAMs provide coherency
between backup cache and memory.

Console connector

lOO·pin for connection to the H3604 console modUle (J2)

Backplane cc:cnec+..cr

270=pin for connection to backplane for Q22-bus, DSS!
bus, and memory intermnnect (Jl)

Run LED

This LED indicates the CPU module is receiving power.

Diagnostic LEDs

A hexadecimal value displays on the four diagnostic LEDs.
The values correspond to the decimal value displayed on
the H3604 console module LED.

CPU System Overview

2-3

Figure 2-2:

KA675/KA680/KA690 Kernel System Functional Diagram
Backplane Interconnect

To Mass Storage
Slots

:::I

_CD

..... 0

H3S04

Console
Module

Ribbon
Cable

CPU

~ g Module

To Q22-bus Slots

-:::I

MSS90 Memory Modules
(1 minimuml4 maximum)

MLO.oo7694

2-4 KA675/KA680/KA690 CPU System Maintenance

Figure 2-3: KA6751KA680/KA690 CPU Module Block Diagram

To Console Module

NVAX

To BA440
Disks

CPU

IP-cache

To Console
Module

To BA440

MLO-OO7262

2.2 MS690 Memory Modules
The MS690 memory module is a double-sided, quad-height memory board
that uses a 150-pin, high-density connector to communicate to the CPU
module. MS690 memory modules are ECC protected via the NMC chip on
the CPU module.
The MS690 memories are available in three variations:
•

MS69~BA (L4004-BA) 32 ME memory

•

MS69O-CA (IAOO4-CA) 64 M.t:S memory

•

MS69~DA (L4004-DA) 128 MB memory

KA6751KA6801KA690-based systems allow for any combination of up to four
MS690 memory arrays providing a memory capacity from 32 Mbytes up to
512 Mbytes.
Figure 2-4 shows a sample memory module, which, like the CPU module,
uses ejector handles designed to ensure proper seating of the modules in
the backplane connectors.

CPU System Overview

2-5

Figure 2-4: Ratchet Handles for CPU and Memory Modules

Ejector
Handles

MLo-oo4227

2.3 BA440 Enclosure Components
KA6751KA680!KA690-based systems use the BA440 enclosure. A brief
description of the components that make up the BA440 enclosure follows.
For information on FRU removal and replacement procedures refer to the
BA430 I BA440 Enclosure Maintenance manual.

2.3.1 H3604 Console Module
The H3604 console module covers the five slots dedicated to the CPU
and memory modules (one slot for the KA6751KA.6801KA690, and four
available slots for MS690 memory modules). Switches on the console
module allow you to configure the kernel. The console module also provides

2-6

KA675/KA680/KA690 CPU System Maintenance

the connectors for a serial line console device, an external DSSI bus, and
the Ethernet. See Figures 2-5 and 2-0.
Figure 2-5: H3604 Console Module (Front)

Console Module

~III=I=-r-I,
I
Power-Up

Mode SWitch
Baud Rate
Select Switch

11_ ~

6 /iI Console Jack

-~ ~ ~[II

Ed I ~ Break Enablel
=r~
rnsable SYrltch
l!!J ~
iY Bus Node

LED Display ...J...-.!.-'

I I

"--~'_.'~I

10 Plugs

-:::r-JI Ethernet

1_- =~r
f~"
I

Standard
Ethernet
Connector

i!----!.-!.-

u=====::!6===.JU

ThinWire
Ethernet
Connector

The front of the console module has the components listed in Table 2-2.

CPU System Overview

2-7

Table 2-2: H3604 Console Module Controls and Indicators
ControllIndicator

Function

Power-Up Mode Switch

This three-position rotary switch determines how the
system responds at power-up.
Language Inquiry Mode (in the top position, indicated by
a profile of a face) causes the system to display a language
selection menu at power-up if the console terminal has
multinational character set (MCS) support. Also, if a
default boot device has not been selected, this mode causes
the system to issue a list of bootable devices and prompts
you to select a device from the list. Once a device is
selected, the system autoboots from that device each time
you tum it on.
Hun Mod.e (in the middie POSltlOn, mcilcate<i by an arrow)
is the normal operating setting.
Loopback Test Mode (in the bottom position, indicated by
a T in a circle) causes the system to run loopback. tests on
the console serial line at power-up.

Baud Rate Select switch

The Baud Rate Select switch is used to set the system's
baud rate to match that of the console terminal. The
factory setting is position 5 (9600).

Console serial MMJ mDllector

This console terminal connector provides the RS-423
interface for the mnsole terminal.

LED Display

The LED displays the testing sequence during power-up.

Break EnablelDisable switch

When the switch is down (position 0), breaks are disabled.
When the switch is up (position 1), breaks are enabled.
When breaks are enabled, pressing IBreak I on the console
terminal halts the processor and transfers control to
the console program. Using the console mmmand SET
CONTROLP, you can specify the control character, ICtrLIP L
rather than IBreak I to initiate a break signal.
The Break EnableJDisable switch also mntrols what
happens at power-up. When breaks are disabled (down,
position 0), the system attempts to automatically boot
software at power-up. When breaks are enabled (up,
position 1), the system enters console mode (indicated by
the »>prompt) at power-up.

Using the console mmmand, SET HALT REBOOT or SET
HALT RESTART_REBOOT, you can set your system to
automatically boot software after the system is halted due
to pressing IBreak L

2-8

KA675/KA680/KA690 CPU System Maintenance

Table 2-2 (Cont.): H3604 Console Module Controls and Indicators
ControllIndicator

Function

Two nSSI bus node ID plugs

KA6751KA6801KA690-based systems have two separate
Digital Storage Systems Interconnect (DSSD busses. Two
DBSI bus node ID plugs, one for the internal nSS! bus,
Bus 0, and one for the external bus, Bus 1, identify the bus
nodes of the DSSI adapters, which are part of the CPU.

Two DBS! connectors for Bus 1

Two Wout nBS! connectors, labeled X and y, on the
console module allow you to eApand the system by
connecting additional mass storage devices to the second
DBSI bus. You can also share mass storage devices
with another system by forming a nBS! VAXcluster
configuration.

Ethernet port features

The console module has two Ethernet connectors: a
BNe-type connector for ThinWire Ethernet., and a 15pin mnnector for a standard Ethernet transceiver cable.
The Ethernet connector switch allows you to set the type
of connection. To use the standard transceiver cable
connection, set the switch to the up position. To use the
ThinWIre cable connection, set the switch to the down
position. A green indicator light (LED) for each mnnector
indicates which connection is active.

CPU System Overview

2-9

Figure 2-6: H3604 Console Module (Back)

r--_--lA~t;;;;~mL--- Battery Backup Unit
J 1 = TOY Clock Battery
J5 = H3604 Power
J6 = CPU Interface
W2 = Remote Boot Enable
W4 = FEPROM Write Enable
F j = ThinWire Ethernet ?ower. 0.5 A
PN = 12-09159-00
F2 = -12V Power, 0.062 A
PN = 90-09122-00
F3 = DSSI Terminator Power, 2.0 A
PN = 12-10929-06
F4 = Standard Ethernet Power, 1.5 A
PN = 12-10929-08

M~1

The back of the console module has the components listed in Table 2-3.

Table 2-3: H3604 Console Module (Back)
Component

Function

Battery Backup Unit

Provides battery backup power to the sse RAM.

TOY Clock Battery connector (J 1)

Provides the connection between the battery
backup unit and the sse RAM.

H3604 power connector (J5)

Four-pin power connector to power harness
module.

CPU Interface connector (J6)

100-pin connector to the CPU module.

ThinWlre Ethernet Power Fuse (Fl)

Protects ThinWU'e Ethernet.

-12 V Power Fuse (F2)

Protects console serial line.

2-10 KA675/KA6801KA690 CPU System Maintenance

Table 2-3 (Cont.): H3604 Console Module (Back)
Component

Function

DSSI Terminator Power Fuse (Fa)

Protects against shorts from the accidental
grounding of the DSSI cable power pin.

Standard Ethernet Power Fuse (F4)

Protects Standard Ethernet.

Remote Boot Enable jumper (W2)

Not used

FEPROM Write Enable jumper (W4)

This jumper must be in the write enable position
to update FEPROMs on the CPU module. Refer
to Chapter 6 for procedures on updating ROMs.

-9 V DClAC converter

Ethernet serial transceiver chip (SlA)

TOY clock oscillator

Time of year oscillator. Privides TOY signal for

the TOY cock in the system support chip (sse) on
the CPU module.

CPU System Overview

2-11

2.3.2 System Control Panel (SCP)
The system control panel (SCP) (Figure 2-7) provides the controls to halt
the processor (external halt type) and enter console mode, as well as to
restart the system and return the processor state to power-up and self
tests.
Figure 2-7: System Control Panel

fil
I~I
~

Over Temperature
Warning Indicator

~-I- DC OK Indicator

~-I- Halt Button

&1-

Restart Button

MLo-oo8652

2-12 KA67SIKA680/KA690 CPU System Maintenance

The SCP has the controls and indicators listed in Table 2-4.

Table 2-4: System Control Panel Controls and Indicators
ControllIndicator

Function

Over Temperature Warning indicator

The red Over Temperature Warning indicator flashes
to indicate that the system's internal temperature is
approaching a level that may cause system components to
overheat. In addition to the Bashing Over Temperature
Warning indicator, an audible alarm also provides
warning of a possible over temperature condition. Ii
the components continue to heat, the system will
automatically shut down to prevent components from
being damaged.

DC OK indicator

The green DC OK indicator shows that the power supply
voltages are within the correct operating range.

Halt Button

The Halt button is a two-position button. When you press
the button. the system halts. A red indicator on the Halt
button lights when the button is set to the in position.
Before you can enter console commands, press the Halt
button again to retum it to the out position. When the Halt
button is returned to the out position, the console mode
prompt »> is displayed on the console terminal screen.
Now you can enter console commands. If you inadvertently
press the Halt button, type "c fRetum to continue.

CAUTION: Pressing the Halt button halts the

system regardless of the setting of the Break
Enable I Disable switch on the console module.
Restart Button

The Restart button has a green indicator. When you
press the Restart button, the system retums to a plwerup condition and self-tests are run. If you have specified a
device as the boot device and if the BreaklEnable Disable
switch is set to disable, the system will reboot system
software.

CPU System Overview

2-13

2.3.3 BA440 Backplane
KA6751KA6801KA690-based systems use the 54-19354-01 backplane,
shown in Figure 2-8.
Figure 2-8: BA440 Backplane

Vterm Module

n
n~l ~ n~ ~ ~~SCP
=:~geL ~ rn m~ um~ mmm Connecmr
J

I;j

m I;j
w

Module

1111111

Connectors

1111111
1211109 8 7 6
\.

Q-bus
Option

)

~~:
~
0

Power Supply
Connectors

5/4 3 2 1
CPU
c:::::=.a
Memory

'--v-J

Power
Board
for H3604
ML().OO7695

2-14 KA675/KA680/KA690 CPU System Maintenance

2.3.4 Power Supply
The BA440 enclosure uses the H7874 power supply (Figure 2-9). Table 2-5
describes the power supply components.
Figure 2-9: Power Supply

Power Switch

I---o-~<
AC Present Indicator
00i=° DC OK Indicator 1~~~~~~:a:Qj~F?;t

0uI
I

Fan Failure
Indicator

o~ ~ Over Temperature

· ~~ 11

Condition

Ind~IO'

~powerBus
~

Connectors

~:-=-- Power Cable

Connector

ULo-oo4040

Table 2-5: H7874 Power Supply Switches, Controls,
and Indicators
ControVlndicator

Function

AC Present indicator (orange)

Lights when the Power switch is set to on (1), and the ac
voltage is present at the input of the power supply.

CPU System Overview

2-15

Table 2-5 (Cont.): H7874 Power Supply Switches, Controls,
and Indicators
ControllIndicator

Function

Power switch

The Power switch is used to turn system power on and
off. The off position is indicated by a 0; the on position is
indicated by a 1.
The Power switch also functions as the system
circuit breaker. In the event of a power surge, the breaker
will trip causing the power switch to return to the off position (0). Taming on the system resets the circuit breaker.
If the circuit breaker trips, wait one minute before turning
the system back on.

DC OK indicator (green)

When the DC OK indicator is lit, the voltages are within
the conect operating range. An unlit DC OK indicator
shows a problem with the power supply.

Fan Failure indicator (amber)

The Fan Failure indicator lights if either of the two cooling
fans stops working. The power supply will automatically
shut down the system as a precautionary measure when a
fan rallure is detected.

Over 'Iemperature
(amber)
Power bus connectors

indicator

The Over Temperature indicator lights if the system has
shut down due to an over temperature condition.

Three power bus connectors' allow you to configure a power
bus for systems expanded with a system expander. The
power bus allows you to tum power on and off for the
system through one power supply designated as the main
power supply: this way, one power switch. can control
power for an expanded system.

NOTE: DSSI VAXcluster systems should not

be configured with a power bus. Inadvertently
shutting off a host system and bringing down
the cluster defeats the added reliability of a
DSSI VAXcluster.
MO

2-16

The main out connector sends the power control bus signal
to the expander. One end of a power bus cable is connected
here; the other end is connected to the 81 (secondary in)
connector or the expander power supply.

KA675/KA680/KA690 CPU System Maintenance

Table 2-5 (Cont.): H7874 Power Supply Switches, Controls,
and Indicators
ControllIndicator

Function

The secondary in connector receives the power bus control
signal from the main power supply. In a power bus with
more than one expander, the power bus signal is passed
along using the secondaly in and out connectors.

The secondary out connector sends the signal down the
power bus for oonfigurations of more than one expander.

2.3.5 System Airflow
Two fans are located under the card cage (Figure 2-10). The power supply
monitors the fans. If either fan stops working, the Fan Failure indicator
on the power supply lights, and the system automatically shuts down as a
precautionary measure.

CPU System Overview

2-17

Figure 2-10: Fans

ML().()()4220

Some system managers request that the fans ron at the maximum rate
at all times to take advantage of a potential increase in system reliability.
The system environment must not exceed the limits described in the Site
Preparation manual. Figure 2-11 shows the location of the fan speed control
(FSC) jumper on the bottom of the power supply. Setting the FSC jumper
to disable causes the fans to run at the maximum rate.

2-18

KA675/KA680/KA690 CPU System Maintenance

Figure 2-11: Fan Speed Control (FSC) Jumper Location

•
•

FSC
Enabled
(Factory
Setting)

FSC
Disabled

MLo-oo4204

CPU System Overview

2-19

Chapter 3

System Setup and Configuration
This chapter describes the guidelines for the configuration of a KA675
/KA680/KA690-based system. Configuration issues covered in this chapter
include module order, mass storage configurations, system expansion, DSSI
VAXcluster configurations, and firmware commands and utilities used in
system configuration.

3.1 CPU and Memory Module Order
The five rightmost BA440 backplane slots are dedicated to CPU and
memory modules. The seven slots to the left are for Q-bus modules. See
Table 3-1.
Table 3-1: BA440 Module Order
Slot

Module

1 through 4

Reserved for up to four MS690 memory modules. MS690 modules
are installed from lett to right with no gaps: first memory module
in slot 4, second memory module in slot 3, and so on.

NOTE: Proper placement of memory modules is
necessary for FRU isolation using error logs.
5

CPU module: KA67S (lAOO2-CA), KASSO (lAOO2-BA). KA690
(UOO2-AA)

6 thmugh 12

~usoptions

A system can have up to four memory modules. Memory modules are
available in 32 MB (MS690-BA), 64 MB (MS69O-CA), and 128 MB (MS690DA), and can be used in any combination. The firmware logically configures
the memory modules at power-up.

System Setup and Configuration

3-1

3.1.1 Installing Add-On MS690 Memory Modules
Perform the following steps to install add-on MS690 memory module(s). You
do not set any jumpers or switches on the memory module. The memory
address for the memory module is mapped by the system.
CAUTION: Thrn off the system before installing modules. Installing
modules while this system is powered-up can damage the modules.

1. Two captive screws hold the console module (H3604) in place. To loosen,
both screws should be turned counterclockwise. The console module is
hinged on the left. Swing the assembly open.
NOTE: TIvo cables connect to the H3601. cor..so!e mod::.!e: C :-:obo1:.
cable which connects to the CPU module; and a four-pin power harness
connects to a power harness module (also known as the power board
H3604) which plugs into the backplane. The power harness module is
located directly to the right of the CPU module.

2. Install the module(s) starting with the first empty slot, which is located
on the right side of the power harness module. The power harness
module is located between the CPU module (slot 5) and the first memory
module (slot 4).
The memory module(s) must be installed in adjacent slots with no
empty slots between. Slots 12 through 6 are ~us slots; slot 5 is
the CPU slot; and slots 4 through 1 are the memory module slots.
3. Make sure the ratchet handles on the memory module on the right
side of the module, and the 150-pin connector is facing to the system.
Wearing the antistatic wrist strap, install the memory module in the
:first available memory slot to the right of the CPU. Ensure that the
memory module is vertically aligned. Push the memory module in until
the ratchet bandIes engage with the enclosure frame. Push the ratchet
handles inward toward the rear of the cabinet until the memory module
is firmly seated in the backplane. When the memory module is firmly
seated, the ratchet handles will lock the module in place.
NOTE: The CPU and MS690 memory modules are equipped with
ratchet handles (Figure 3-1) which are shipped in a horizontal position.
The ratchet handles are designed to ensure that the modules are properly
seated in the backplane connectors.

3-2

KA675/KA6801KA690 CPU System Maintenance

Figure 3-1: Memory Module Ratchet Handles

Ejector
Handles

MLQ-008453

4. Close the H3604 console module and lock the 114-turn captive screws.
5. To identify the memory module, place the MS690 option iabei (supplied
in the option kit) in the proper location on the H3604 panel. Indicate
the revision number a..Y}d memory option (BA,CA, or DA).

6. Refer to Chapter 4 for information on initialization and acceptance
testing.

System Setup and Configuration

.3-3.

3.2 General Module Order for Q-Bus Options
The order of the supported Q-bus options in the backplane depends on four
factors:
•

Relative use of devices in the system

•

Expected performance of each device relative to other devices

•

The ability of a device to tolerate delays between bus requests and bus
grants (called delay tolerance or interrupt latency)

•

The tendency of a device to prevent other devices farther from the CPU
from accessing the bus

'l'he supported options arranged by type are:

Communications
CXAI6-AAlAF: I6-line DEC-423 asynchronous controller
CXBI6-AAlAF: 16-line RS-422 asynchronous controller
CXY08-AAlAF: 8-line RS-232C asynchronous controller with modem
DEQRA-CA: Token Ring Network Controller
DESQA-SAlSF: ThinWire Ethernet adapter
DFAOl-AAlAF: 240011200 BPS modem
DIV32-SA1SF: Q-bus ISDN basic rate access interface
DPVll-SAlSF: Q-bus synchronous programmable interface
DRVIW-SAlSF: General purpose 16-bit parallel DMA interface
DRVlJ-SAlSF: Q-bus parallel interface
DSVI1-SAlSF: Q-bus 2-line synchronous
KMVlA-SAlSF: Single-line programmable controller with DMA
General
ADQ32-SAlSF: 32-channel ADC module
ADVll-SAlSF: I6-channel ADC module
AXVII-SAlSF: I6-channel ADC, 2-channel DAC module
DRQ3B-SAlSF: Q-bus parallel 110 interface
DTC05-SA: Digital encoded voice, multifu.ction
IBQOI--SAlSF: DECscanlBITBUS controller
IEQll-SA1SF: Dual-bit DMA serial Q-bus controller
KITHA-AA: Mira AS option
KWVI1-SAlSF: Programmable real-time clock
KXJI1-SA: Q-bus peripheral processor with s-box adapter kit
LPVll--SAlSF: Line printer controller
MRVll: Q-bus, universal socket, 32-Kbyte EPROM
VS30U-GAlG3/G4: Graphics option

3-4 KA6751KA680/KA690 CPU System Maintenance

Mass Storage, Tape, Pedestal Expansions
RF35-AAlAF: 852-Mbyte half-height DSSI integrated storage element
RF73E-AAlAF: 2.0-Gbyte full-height DSSI integrated storage element
RF72E-AAlAF: l.O-Gbyte full-height DSSI integrated storage element
RF71E-AAlAF: 400-Mbyte full-height DSSI integrated storage element
RF3lE-AAlAF: 38l-Mbyte half-height DSSI integrated storage element
RF3lT-AAlAF: 38l-Mbyte full-height DSSI integrated storage element
TF85~A: 2.6-Gbyte DSSI integrated storage element with 5.25-inch
cartridge
TLZ04-JAlJF/GA: 1.2-Gbyte cassette (nAT) tape drive (requires
KZQSA storage adapter)
TK70E-AAlAFtrQK70-SAlSF: 5.25-inch cartridge, 296-Mbyte tape
drive, tape controller
TK50E-AAIAFtrQK5O-SAlSF: 5.25-inch cartridge, 95-Mbyte tape
drive, tape controller
KLESI-SA: Q-bus to LES! adapter
KFQSA-SElSG: DSSI Q-bus adapter
KZQSA-SAlSF: Storage adapter for TLZ04 tape drive and RRD42
compact disc drive
RA81182: Storage array (separate cabinets only)
RA90/92: Storage array (separate cabinets only)
KDA5O-SE/SG: SDI Q-bus adapter
KRQ50-SAlSF: Q-bus controller for RRD40-DC
TU8lE-SAlSB: Magnetic tape (requires KLESI controller)
TSV05-SElSFISHlSJ: Q-bus TS05 magnetic tape controller
B400X: Expansion box with 10 Q-bus slots and up to 4 ISEs
R400X-B9: Expansion box with up to 7 RF-series ISEs
RRD40:
600=Mbyte CDROM table-top drive (requires KRQ50
controller)
RRD42: 600-Mbyte tabletop compact disc drive (requires KZQSA
storage adapter)
RSV20-A: WORM optical drive subsystem (requires KLESI controller)
RWZOl: 594-Mbyte Magneto-Optical Disc (requires KZQSA storage
adapter)
ESE20: Electronic storage element (requires KDA50 controller)

3.3 Recommended Module Order of Q-bus Options
The recommended module order for placement of Q-bus options is provided
in the following list:
MRVll (Placement not critical)
AAVll
ADVll
AXVll

System Setup and Configuration

3-5

KWVll
DRVIJ
KMVlA
DESQA
DPVl1
DIV32
VCB02
DFAOI

CXA16
CXY08

CXB16
LPVll
DRVIW
KRQ50

IEQll
ADQ32
DRQ3B
DSVll
KLESI
IBQOl
TSV05 (M7530 controller)
KDA5O-SE

KFQSA-SE
KZQSA

TQK50
TQK70
M9060-YA

3.4 Mass Storage Options (Internal)
The mass storage shelf of a BA440 enclosure provides four storage cavities
for embedded mass storage options. The right-most storage cavity can
contain a tape drive (TF85, TK-series, or TLZ04); all four storage cavities
can contain an RF-series ISE.
Combinations of dual-disk, single-disk, or tape ISEs can be used to gain
the full complement of seven DSSI devices on the internal DSSI bus (Bus
0).
VAX 4000 Model 400/5001600 systems can support the following
combinations of mass storage options embedded in the sYstem enclosure:
•

One tape drive (TF85, TK-series, or TLZ04) and up to six RF-series
ISEs using the dual-disk RF35.

•

No tape drive and up to seven RF'-series ISEs using.the dual-disk RF35.

3-0 KA675/KA680/KA690 CPU System Maintenance

NOTE: The RF35, which has dual-disk capability, can be ordered with a
single disk.

Figure 3-2 shows an example of a mass storage configuration consisting of
a TF85 tape drive, two RF35s, and two RF73s.
Figure 3-2: Storage Configuration Example
ISE3

ISE2

ISE 1 and 0

Tape Drive

ML~7696

Rules for Numbering Storage Devices
Use the rules below for numbering (bus node ID and MSCP unit numbers)
storage devices:

•

For each DSSI bus, do not duplicate bus node ID numbers for your
storage devices/adapter. For Bus 0, you can have only one storage
device identified as bus node 0, one storage device as 1, and so on;
for Bus 1, you can have only one storage device identified as bus node
0, one storage device as 1, 3-'ld so on.

•

The previous rule also applies to DSSI VAXcluster configurations, all
DSSI bus node numbers for storage devices 8.a."'ld DSSI adapters must
be unique in a shared nSSI bus.

•

By convention, the RF-series ISEs are numbered in increasing order
from right to left beginning with zero.

•

nSSI adapters use the highest available bus nodes. The next highest
available bus node (usually five) is reserved for the TF-series tape drive.

•

When more than one nSSI bus is being used and the system is using a
nonzero allocation class, you need to assign new MSCP unit numbers
for devices on all but one of the DSSI busses since the unit numbers

System Setup and Configuration

3-7

for all nSSI devices connected to a system's associated nSSI busses
must be unique. Refer to Section 3.7.3 for more information on setting
parameters for nSSI devices.
NOTE: If you change the bus node ID plugs, power-down the system, change
the plugs and then power-up the system.

3.5 System Expansion
The mass storage and Q-bus capacity of VAX 4000 Model 500 systems can
be increased using the following expanders.

•

The R400X mass storage expander provides space for up to seven
additional RF-series ISEs or up to six RF-series ISEs and a tape drive
(TF85 or TLZ04). Using R400X expanders, you can fill both nSSI busses
for a total of 14 nSSI mass storage devices.
NOTE: Using the dual-disk RF35, the R400X can accommodate up to

13ISEs.
•

The R215F expander provides space for up to three RF-series ISEs.
NOTE: Using the dual-disk RF35, you can increase the number of
[SEs-up to seven [SEs per nSS[ bus.

•

The SFl00 storage array pedestal provides space for a TF857 magazine
tape subsystem and one SFxx storage array building block.

•

The SF200 storage array subsystem provides space for up to two TF857
magazine tape subsystems and up to six SFxx storage array building
blocks.

3.5.2 Q-Bus Expanders
•

The B400X expander provides 10 additional usable Q-bus slots for a
system total of 17 usable Q-bus slots. The B400X also has space for up
to four additional RF-series ISEs or up to three ISEs and a tape drive
(TF85, TK70, or TLZ04).
NOTE: Using the dual-disk RF35, the B400X can accommodate up to

seven [SEse

3-8 KA675/KA680/KA690 CPU System Maintenance

•

The B213F expander also provides 10 additional usable Q-bus slots and
provides space for up to three RF-series ISEs or up to three ISEs and
a TK-series tape drive.

NOTE: Installing a B213F or R215F on a VAX 4000 system requires the
H401O-AA expander cable kit.

3.5.3 Control Power Bus for Expanders
The three power bus connectors on the H7874 power supply allow you
to configure a power bus for systems expanded with R400X and B400X
expanders. The power bus allows you to tum power on and off for" one or
more expanders through the power supply designated as the main power
supply (Figure 3-3).
NOTE: nBSI VAXcluster systems should not be configured with a power
bus. Inadvertently bringing down the cluster defeats the added reliability
of a DSSI VAXcluster.

Figure 3-3: Sample Power Bus Configuration
System

Expander 1

Expander 2

1~lsl
III I

I~I

ML()'()04041

System Setup and Configuration

3-9

3.5.4 Adding Options to the System Enclosure
To determine what options you can add to the system enclosure, you must
list the options currently installed and their power requirements on the
VAX. 4000 Model 500 Configuration Worksheet, provided in Figure 3-4.
The worksheet in Figure 3-4 is for the BA440 enclosure. All backplane
slots and mass storage devices are powered by the H7874 power supply.
Use the worksheets as follows:
1. In the Module column, list all options and mass storage devices
currently installed in your system. The CPU module has already been

entered.
Use the label on the cover panel of each slot to identify the module
installed in that slot.
2. List each embedded storage device.
3. List the options and mass storage devices you wish to add to your
system.
4. If the system includes a TK70 tape drive, list the TQK70 controller last.

5. Fill in the power requirements for each module and each mass storage
device. The power requirements for the more common options are listed
in Table 3-2.
6. Add each column and make sure the totals do not exceed the specified
limit.

3-10 KA675/KA680/KA690 CPU System Maintenance

-Figure 3-4: VAX 4000 ModelSOO Configuration Worksheet

Module

Slot

Current (Amps) 1
..s Vdc +12 Vdc +3.3 Vdc 12 Vdc

Power

Bus Load

(Watts)

53.8

4.0

1.0

I 584.0W ! 31

MEM 1
MEM2
MEM3
MEM4
CPU 2

L4002-nA 3

4.8

1.6

3.2

0.0

Q-bus 1
Q-bus2
Q-bus3
Q-bus4

0-0055
Q-bus6
O-bus 7

Mass Storage:
1

2
3
4

Total these columns:
I. • • • _ ....
"'"
Mus
...... s ..
casd.

II ~nn
.0 A
__ ._A I ~nA
_._ .. I! 1.S.OA I, ~L

1. Total output power from +3.3 Vdc and +5 Vdc must not exceed 330 watts.
2. Power requirements in this line include CPU module, H3604 console module,
and backplane terminator combined.
3. KA690(L4002-AA), KA680(L4002-BA), or KA675(L4002-CA)
ML().()05361

System Setup and Configuration

3-11

Table 3-2: Power Requirements
Current (Amps)
Max

Power
Max

J!:~s

Watts

DC
0.5

Option

Module

+5V

+12 V

AAVll-sA

AlOO9-PA

2.10

0.00

10.50

2.5

ADQ3Z-SA

A030

4.45

0.00

22.25

2.5

0.5

ADVll-SA.

AlOOS-PA

2.00

0.00

10.00

2.3

0.5

AXVll-8A

A~A

2.00

0.00

10.00

1.2

0.3

CXAl6-AA

M3l18-YA

1.60

0.20

10.40

3.0

0.5

CXBI6-AA

M3118-'YB

2.00

0.00

10.00

3.3

0.5

CX'Y08-M.

M3119-YA

l.64

0.395

12.94

3.0

0.5

DESQA-SA

M3127-PA

2.40

0.22

14.M

3.3

0.5

DEQRA-cA

M7533-.\B

4.0

1.0

21.20

2.2

0.5

DFA01-AA

M3121-PA

1.97

0.04

10.30

3.0

1.0

D1V32-SA.

M'7S7l-PA

5.5

0.00

35.4

3.5

1.0

DPV11-SA.

M802O-PA

1.20

0.30

9.60

1.0

DRQ3B-SA

M1658-PA

4.50

0.00

22.60

2.0

0.5

DRV1J-SA.

M80(9-PA

1.80

0.00

9.00

2.0

1.0

DRVlW-sA

M'7651-PA

1.80

0.00

9.00

2.0

1.0

DSVll-SA.

M3108

5.43

0.69

35.43

3.9

1.0

DTC05-SA

M3136

4.0

0.0

15.80

3.6

0.76

1.70

o.so

14.50

IBQ01-SA

M312S-PA

5.00

0.30

28.60

4.6

1.0

JEQll-&.

M8634-PA

3.50

0.00

11.50

2.0

1.0

KA6'76-CA

L4OO2-CA

4080

1.60

53.8

4.0

1.0

H3604 1

~BA

L4OO2-BA

4.10

1.60

53.8

4.0

1.0

KA&8O-AA

IAOO2-AA

4.80

1.60

53.8

4.0

1.0

KDA5O-SE

M'71M

6.93

0.00

34.65

3.0

0.5

M7165

6.51

0.03

33.21

KFQSA-&VSE

M1'769

s.so

0.00

27.50

4.4

0.5

KLES1-SNSF

M7740-PA

4.00

0.00

20.00

0.5

1.0

KRQ5O-SAJSF

M16S2

2.70

0.00

13.50

2.7

1.0

KWVll-SA

M4002-PA

2.20

0.013

11.166

1.0

0.3

XXJl1-sF

M'7616

6.0

0.4

46.8

2.0

1.0

EZQSA-SA. .

M5916

5.4

0.0

27.0

4.4

0.5

LPVl1-SA.

M808&-PA

2.80

0.00

14.00

1.8

0.5

M9404-PA

M9404

0.00

0.0

M9405-PA

M9405

0.00

0.0

MRVlI-D

MBS18

0.00

8.00

3.0

0.5

1.s02

lAlso include -12 Vdc @0.25A, 3 W.
2Value is for the unpopulated module only.

3-12

KA675/KA680/KA690 CPU System Maintenance

Table 3-2 (Cont.): Power Requirements
Current (Amps)
Max

Power
Max

J!~s
AC

Option

Module

+5V

+12 V

Watts

MS690-BA

L4004-BA

5.03

0.00

26.5

MS69O-CA

L4004-CA

4.2

0.00

21.0

MS69O-DA

L4OO4-DA

6.4

0.00

32.0

BF31E-AA1AF

1.2

2.21

32.52

NJA

BF3IF-AAJAF

1.2

2.21

32.52

NJA

NlA

BF31T-AAIAF

1.71

0.85

13.7

NJA

NlA

RF35E-AAIAF

0.11

2.29

31.1

NlA

RF3S2-AAlAF

1.69

5.10

33.0

NJA

RF71E-AAIAF

1.25

1.64

25.93

NJA

IfF'12E-AAIAF

1.20

1.15

27.00

NlA

RF73E-AAIAF

1.20

1.75

2'7.00

NJA

TFSSE-J'AlJF

1.50

2.40

36.30

NlA

NJA

TKSOE-AA

1.50

2.40

36.30

NlA

NJA

TK70E-AAlAF

1.50

2.40

36.30

NlA

NJA

1.5

2.4

36.3

NlA

NJA

14.5

2.8

0.5

M7S46

2.9

0.00

TQK70-S1JSF

M7S59

3.50

0.00

17.50

4.3

0.5

TSVOS-SA

M7S3O
)(7615

6.50

0.00

32.50

1.5

1.0

4.60

0.10

24.2

3.5

1.0

VCB02-B

K716i-OO
M7169

8.8S

0.41

49.89

3.S

1.0

VCB02-C

(2) H'716&-OO

12.0

0.41

66.64

3.5

1.0

VCB02-A

K7169

3.6 DSSi VAXciusters
A nSS! VAXcluster configuration is one in which up to three systems can
access the same nSSI devices. Some failures of B..L"1Y system can be tolerated,
in which case the remaining system(s) continues to access all available nSSI
devices and assure continued processing.

System Setup and Configuration

3-13

Figure 3-5: DSSI cabling for a Generic Two-System DSSI VAXcluster
Configuration
SYSO

RFxx

SYS1

RFxx

DSSI

MLO-003295

If one of the CPU modules fails, all satellite nodes booted through that
CPU module lose connections to the system disk. However, the DSSI

VAXcluster enables each satellite node to know that the system disk is still
available through a different path-that of the functioning CPU module. A
connection through that CPU is then established, and the satellite nodes
are able to continue operation. The entire cluster will run slower, since
one CPU is now serving the satellite nodes of both systems. Processing can
continue, however, until Digital Services can repair the problem.
A DSSI VAXc1uster system cannot recover from the following conditions:
•

System disk failure, which can be caused by such factors as a power
supply failure in the enclosure containing the disk.

•

DSSI cabling failure, which must be repaired to continue operation.

3-14 KA6751KA680/KA690 CPU System Maintenance

3.6.1 DSSI VAXcluster Configuration Rules
1. An Ethernet (NI)IFDDI is required on all CPU nodes.
2. A DECnet license is required (At least one full function license).
3. At least one common (primary) DSSI bus is required to connect with a
system disk containing system critical files.
4. VAXcluster and VMS license is required.

5. A maximum of eight nodes per DSS! bus:
a. Each adapter or ISE (disk/tape) counts as one node.
b. A DSSI bus is a collection of all DSSI cable/path segments (inside
and outside of cabinets) between two end terminators.
c.

Each node must have a unique bus node ID number (0-7).

6. A maximu..TD. of three CPUs1adapters per nSSI bus is supported.
7. The nSSI bus MUST be terminated at both ends.
8. The nSSI bus MUST have a common ground between all elements
(CPU, disks). The ground offset is a function of the total DSS! bus
length (terminator to terminator).
Use a voltmeter to make sure the ground offset voltage between any
two enclosures does not exceed one of the limits listed below.
Allowable Ground Offset Voltage
Total Bus Length

AC (rms)

Up to 20 meters (65 feet)

200 millivolts

70 millivolts

20 to 25 meters (65 to 82 feet)
(Computer room)

40 millivolts

14 millivolts

Total bus length includes all DSSI cable lenghts, internal and external.
Refer to the DSSI VAXcluster Installation and Troubleshooting manual
for instructions on calculating internal cable lengths.
To measure the ground offset voltage, .connect the voltmeter leads to
bare (unpainted) metal on each enclosure.
NOTE: The ground offset voltage may vary over time, if equipment is
added to the system or plugged into the power outlets. Therefore, this

System Setup and Configuration

3-15

measurement does not guarantee that the voltage will remain within
acceptable limits.
9. Maximum single cable length is 15 m (50 ft) between connectors.
10. Disconnecting the DSSI cables is NOT allowed while bus is operational.
11. Number of nSSI busses per CPU:
CPU

DSSlBusses

KA640

1 Embedded DSSI Adapter (EDA), 2 KFQSA on Q-bus

KA660

1 Embedded DSSI Adapter (EDA), 2 KFQSA 011 Q-bus

KA670

2 Embedded DSSI Adapters (EDAs), 2 KFQSA on Q-bus

KA675

2 Embedded DSSI Adapters (EDAs), 2 KFQSA on Q-bus

KA680

2 Embedded DSSI Adapters (EDAs), 2 KFQSA on Q-bus

KA690

2 Embedded DSSI Adapters (EDAs), 2 KFQSA on Q-bus

KA6501KA655

2 KFQSAs 011 Q-bus

KA630

2 KFQSAs on Q-bus

6xxx

6 KFMSAs per system

9000

6 KFMSAs per XMI
12 KFMSAs per system

12. The minimum VMS revision for DSSI VAXcluster of more than two
nodes with:
a. VAX 4000 Model 400 is VMS 5.5
b. VAX 4000 Model 500 is VMS 5.5
c.

VAX 4000 Model 600 is VMS 5.5

13. These rules apply to Digital supplied hardware.
Third party
devices may not conform to nSSI electrical specification requirements.
Therefore, bus length, ground offset, basic noise margining, and warm
swap characteristics are at risk when using third party devices.
14. Like adapters should be connected together whenever possible.
15. Like CPUs should be connected together whenever possible.

3-16

KA67SIKA680/KA690 CPU System Maintenance

For more information on DSSI VAXcluster configurations, refer to the
nSSI VAXcluster InstaUation and Troubleshooting manual. Figure 3-6 and
Figure 3-7 show two popular DSSI VAXcluster configurations using a VAX
4000 system.

System Setup and Configuration

3-17

Figure 3-6: Two-System DSSI VAXcluster

System A

Ststem B

DSSI Cables " - - - - - - - - - - - " ,
Shared DSSI Busses and Devices

I DSSI Terminator locations
System A

System B

DSSI
Adapter 1

KAnn

DSSI
Adapter 0

--_.

......

SHAC
Bus Node 6

DSSI
Adapter 0
DSSI
Adapter1
SHAC
Bus Node 7

m1 DSSI Bus Nodes for Storage Devices in System B

OSSI Bus Nodes for Storage Devices in System A
ML00008312

3-18 KA675/KA680/KA690 CPU System Maintenance

Figure 3-7: Expanded Two-System DSSI VAXcluster
Expander

System A

System B

I DSSI Terminator Locations
~--~---r---T----~--~--~--~SHAC

Bus Node 7
System B
DSSI
KAnn
AdapterO ""-..1

~
SHAC
Bus Node 7

System A

System B

1m DSSI Bus Nodes for Storage Devices in Expander

DSSI Bus Nodes for Storage Devices in System A and B
MLO-OO8663

System Setup and Configuration

3-19

3.7 Firmware Commands and Utilities Used in
System Configuration
Several commands and utilities are needed to configure a system.. This
section covers commands for examining and setting system parameters,
DSSI parameters, and module addresses. For a complete listing of firmware
commands, refer to Appendix A

3.7.1 Examining System Configuration
Several variations of the SHOW command provide a display of options and
key configuration information.
:

SHOW DEV1CE - Lists devices (il.lasS storage, Ethernet, and Q-bus) In
the system. (The SHOW DEVICE command combines the information
displayed using the SHOW command with nSSI, UQSSP, SCSI, and
Ethernet.)

•

SHOW nSSI - Lists all nSSI devices and their associated nSSI
parameters for embedded nSSI adapters.

•

SHOW ETHERNET - Lists the hardware Ethernet address for each
Ethernet adapter.

•

SHOW QBUS - Lists all Q-bus devices and their I/O addresses in hex,
the address as it would appear in the Q-Bus 110 space in octal, and well
as the word data read in hex.

•

SHOW SCSI - Lists all SCSI devices in the system.

•

SHOW UQSSP adapters.

•

Lists all nSS! devices for KFQSA-based nSSI

SHOW MEMORY - Lists main memory configuration for each memory
board.

Sample displays of each of the above commands are provided below.
»>SBON DEVICE

DSSI Bus 0 Node 0 (CLYDE)
-DIAO (RF73)
DSSI Bus 0 Node 1 (BONNIE)
-DIAl (RF73)
DSSI Bus 0 Node 5 (TFDR1)
-MIAS

(TF85)

DSSI Bus 0 Node 6 (*)
DSSI Bus 1 Node 7 (*)
UQSSP Disk Controller 0 (772lS0)
-DUA2 0 (RF31)
UQSSP Disk Controller 1 (760334)

3-20 KA675/KA680/KA690 CPU System Maintenance

-DUB21 (RF31)
OQSSP Disk Controller 2 (760340)
-DUC22 (RF31)
OQSSP Disk Controller 3 (760322)
-DOD2.3 (RF31)
OQSSP Tape Controller 0 (774500)
-MUAO (TK70)
SCSI Adaptor 0 (761400), SCSI ID 7
-MKAO (DEC TLZ04 1991(c)DEC)
Ethernet Adapter
-EZAO (08-00-2B-06-10-42)
»>SHOW DSSI
DSSI Bus 0 Node 0
-DIAO (RF73)
DSSI Bus 0 Node 1
-DIAl (RF73)
DSSI Bus 0 Node 5
-MIAS (TF85)
DSSI Bus o Node 6
DSSI Bus 1 Node 7
»>

(CLYDE)
(BONNIE)
(TFDR1)
(* )
(* )

»>SHOW ETHERNET
Ethernet Adapter
-EZAO (08-00-2B-OB-29-14)
»>SHOW OQSSP
UQSSP Disk Controller 0
-DOA20 (RF31)
OQSSP Disk Controller 1
-DUB21 (RF31)
OQSSP Disk Controller 2
-DOC22 (RF31)
OQSSP Disk Controller 3
-DOD23 (RF31)
UQSSP Tape Controller 0
-MUAO (TK70)

(772150)
(760334)
(760340)
(760322)
(774500)

»SROW gaus
Scan of Q-bus I/O Space
-20001920 (774440) = FF08 DELQA/DESQA
-20001922 (714442) - FFOO
-20001924 (714444) - FF2B
-20001926 (714446) - FF08
-20001928 (714450) = FFD7
-2000192A (774452) = FF41
-2000192C (774454) - 0000
-2000192E (714456) = 1030
-20001F40 (717500) - 0020 IPCR

System Setup and Configuration

3-21

Scan of Q-bus Memory Space

»>SHOW SCSI
. SCSI Adapter 0 (761300), SCSI ID 7
-MKASOO (DEC TLZ04 1991 (c) DEC)

»>SHOW MEMORY
Memory 0: 00000000 to 01FFFFFF, 32 Mbytes, 0 bad pages
Total of 32 Mbytes, 0 bad pages, 112 reserved pages

3.7.2 Using the CONFIGURE Command to Determine CSR
Addresses for Q-Bus Modules
Each Q-bus module in a system must use a uniaue device address and
interrupt vector. The device' address is also known as the control and status
register (CSR) address. Most modules have switches or jumpers for setting
the CSR address values. The value of a floating address depends on what
other modules are housed in the system. The CONFIGURE command is
used to determine what the proper CSR addresses should be for the given
configuration. You can than configure the Q-bus modules according to this
information.
NOTE: These recommended values simplify the use of the MDM diagnostic
package and are compatible with VMS device drivers. YOu can select
nonstandard addresses, but they require a special setup for use with VMS
drivers and MDM. See the Micro VAX Diagnostic Monitor User's Guide for
information about the CONNECT and IGNORE commands, which are used
to set up MDM for testing nonstandard configurations.

Determine CSR address values for a module as follows:
1.

Use the SHOW QBUS firmware command to get a listing of the Q-bus
modules currently in the system.

Determine the correct values for the module using the CONFIGURE
firmware command command. The CONFIG utility eliminates the need
to boot the VMS operating system to determine CSRs and interrupt
vectors. Enter the CONFIGURE command, then HELP for the list
of supported devices. Enter the device and number of devices for all
existing modules in the system, as well as for those devices you are
adding.
»>CONFlGORE
Enter device configuration, HELP, or EXIT
Device, Number? help
Devices:

3-22

KA675/KA6801KA690 CPU System Maintenance

Devices:
LPVl1
KXJ11
TSV05
RLVl2
DELQA
DMVl1
RQC25
RRD50
RV20
KFQSA-TAPE
CXA16
CXB16
QPSS
LNV2l
ADV1lD
KWVllC
VSV2l
DRQ3B
IOVllO
IAV1lA
IGQ11
DESNA
KZQSA
KWV32
Device, Number?
Numbers:
1 to 255, default is 1
Device,Nurnber? cxa16,1
Device,Number? desqa,1
Device,Nurnber? tqk70
Oevice,Number? qza
Oevice,Number? kfqsa-disk
Device,Number? exit

DLV11J
RXV21
DEQNA
KFQSA-DISK
KMVl1
CXY08
DSVl1
AAVl10
IBQ01
IAVl1B
DIV32
M7577

DZQ11
DRV11W·
DESQA
TQK50
IEQl1
VCBOl
ADV1lC
VCB02
IDV11A
MIRA
KIV32
LNV24

DZV11
DRV1lB
RQDX3
TQK70
DRQll
QVSS
AAVllC
QDSS
IDV1lB
ADQ32
OTCN5
M7576

Address/Vector Assignments
-774440/120 DESQA
-772150/154 KFQSA-OISK
-774500/260 TQK70
-760440/300 CXA16
-761300/310 KZQSA

NOTE: Of the devices listed in the CONFIG display, not all are
supported on the VAX 4000 Model 500 systems. See Section 3.2 for
supported options.

The LPVll-SA has two sets of CSR address and interrupt vectors. To
determine the correct values for an LPVll-SA, enter LPVll,2 at the
DEVICE prompt for one LPVll-SA or enter LPV11,4 for two LPVl1-SA
modules.
3. See the Microsystems Options manual for switch and CSR and interrupt
vector jumper settings for supported options.
NOTE: The CSR address for KFQSA storage adapter is programmed

using firmware commands.

Refer to the Appendix H for using the

SET I HOST I UQSSP I MAINT command to access the Diagnostic Utility

Program (DUP) driver utility to configure the CSRs for the KFQSA
module.

System Setup and Configuration

3-23

3.7.3 Setting and Examining Parameters for DSSI Devices
Two types of nSSI storage adapters are available for VAX 4000 systems: an
embedded nSSI adapter, which is part of the CPU, and the KFQSA adapter.
The KA6751KA6801KA690 CPU has two embedded nSS! adapters: Bus 0
and Bus 1.
Each adapter provides a separate nSS! bus that can support up to eight
nodes, where the adapter and each nSS! storage devices count as one node,
hence each nSS! adapter can support up to seven nSS! storage devices (six
DSSI storage devices for a two-system nSSI VAXcluster; five DSSI storage
devices for a three-system DSSI VAXcluster configuration). The adapters
make a connection between the CPU and the requested device on their
resnective nSSI bus. Each DSSI device has its own controller and SeTVp.,.
that contain the intelligence and logic necessary to control data transfers
over the nSSI bus.
3.7.3.1 DSSI Device Parameters

Six principal parameters are associated with each DSSI device:
•

Bus Node ID

•

ALLCLASS

•
•

UNITNUM
FORCEUNI

•

NODENAME

•

SYSTEMID

NOTE: Each of the above parameters, with the exception of the bus node
ID, are programmed and examined using the console-based Diagnostic
and Utility Program (DUP) driver utility. The bus node ID is physically
determined by the numbered bus node ID plug that inserts into the device's
front panel.

A brief description of each parameter follows:
The bus node ID parameter is provided by the bus node ID plug on
the device's front panel. Each DSSI bus can support up to eight nodes,
0-7. Each nSSI adapter and each device count as a node. Hence,
in a single-system configuration, a nSSI bus can support up to seven
devices, bus nodes 0-6 (with node 7 reserved for the adapter); in a twosystem DSSI VAX.cluster configuration, up to six devices, 0-5 (with nodes
6 and 7 reserved for the adapters); in a three-system DSSI VAXcluster

3-24 KA675/KA6801KA690 CPU System Maintenance

configuration, up to five devices, 0-4 (with nodes 5, 6, and 7 reserved for
the adapters).
The ALLCLASS parameter determines the device allocation class. The
allocation class is a numeric value from 0 to 255 that is used by the VMS
operating system to derive a path-independent name for multiple access
paths to the same device. The ALLCLASS firmware parameter corresponds
to the VMS SYSGEN parameter ALLOCLASS.
DSSI devices are shipped from the factory with a default allocation class of
zero. Each device to be served to a cluster must have a nonzero
allocation class that matches the allocation class of the system.
Refer to the VMS VAXcluster manual for rules on specifying allocation class
values.
The UNITNUM parameter determines the unit number of the device. By
default, the device unit number is supplied by the bus node ID plug on the
device's front panel. Systems with multiple nSSI busses, as described
later in this section, require that the default values be replaced
with unique unit numbers. To set unit numbers and override the default
values, you use the console-based DUP driver utility to supply values to
the UNITNUM parameter and to set a value of zero to device parameter
FORCEUNI.
The FORCEUNI parameter controls the use of UNITNUM to override the
default device unit number supplied by the bus node ID plug. When
FORCEUNI is set to a value of 0, the operating system uses the value
assigned to the UNITNUM parameter; when FORCEUNI is set to a value
of 1, the operating system uses the value supplied by the bus node ID plug.
The NODENAME parameter allows each device to have an alphanumeric
node name of up to eight characters. DSSI devices are shipped from the
factory with a unique identifier, such as R7CZZC, R7ALUC, and so on. You
can provide your own node name.
The SYSTEMID parameter provides a number that uniquely identifies the
device to the operating system. This parameter is modified when replacing
a device using warmswapping procedures.
3.7.3.2 How VMS Uses the DSSI Device Parameters

This section describes how the operating system uses the parameters to
form unique identifiers for each device. Configurations that require you to
assign new unit numbers for devices are also described
With an allocation class of zero, the operating system can use the default
parameter values to provide each device with a unique device name. The
operating system uses the node name along with the device logical name
in the following manner:

System Setup and Configuration

3-25

NODENAME$DlAu
where:
NODENAME is a unique node name and u is the unit number.
With a nonzero allocation class, the operating system relies on unit number
values to create a unique device name. The operating system uses the
allocation class along with the device logical name in the following manner:
$ALLCLASS$DlAu
where:
ALLCLASS is the allocation class for the system and devices, and u is a
unique .unit number.
U sing mass storage expanders, you can fill multiple nSSI busses: busses 0
and 1 supplied by the CPU module, and a third and fourth nSS! bus using
two KFQSA adapters. Each bus can have up to seven DSS! devices (bus
nodes 0-6). When more than one bus is being used, and your system is
using a nonzero allocation class, you need to assign new unit numbers for
devices on all but one of the DSSI busses, since the unit numbers for all
DSSI storage devices connected to a system's associated DSSI busses must
be unique.
Figure 3-8 illustrates the need to program unit numbers for a system using
more than one nSSI bus and a nonzero allocation class. In the case of the
nonzero allocation class, the operating system sees three of the ISEs as
having duplicate device names, which is an error, as all unit numbers must
be unique.

3-26

KA6751KA680/KA690 CPU System Maintenance

Figure 3-8: VMS Operating System Requires Unique Unit Numbers for
DSSI Devices

Allocation Class:O

Nonzero Allocation Class
(Example: ALLCLASS=l)

R7BUCC$OIAO

$1$DIAO ....
c(----,

R7CZZC$OIA 1

$1 $OIA 1 ....Cf------+i--.,j

R7ALUC$OIA2

$1$OIA2

R7EB3C$OIA3

$1$OIA3

TFOR1$MIA5

$1$MIA5

R7IDFC$DIAO

$1$D1AO

R7IBZC$OIA 1

$1$OIA1

R7IKJC$OIA2

$1$OIA2
$1$OIA3 ..._ _ _ _ _ _ _....J

R7I03C$DIA3

*oUPUcate 0

I *Duplicato 1
* Dupl'icate 2
*Duplicate 3

R7XA4C$OIA4

$1$OIA4

R7QIYC$OIA5

$1$OIA5

R70A4C$D1A6

$1$OIA6

* Nonzero allocation class examples with an asterisk Indicate duplicate device names.
For one of the 0551 busses. tho unit numbers need to be reprogrammed to avoid this error.

MLO-OO7176

NOTE: Digital recommends configuring systems to have unique unit
numbers even for standalone systems using an allocation class of zero. This
practice wiU avoid problems with duplicate device names if the system is
later configured in a cluster.

The following instructions describe how to change DSSI parameters, using
the DUP driver utility. In the example procedures, the allocation class will
be set to 1, the devices for Bus 0 (in the VAX. 4000) will be assigned new unit
numbers (to avoid the problem of duplicate unit numbers), and the system
disk will be assigned a new node name. To examine DSSI parameters from
the VMS operating system, refer to Section 3.7.3.4.

System Setup and Configuration

3-27

Figure 3-9 shows sample DSSI busses and bus node IDs for an expanded
VAX 4000 Model 500 system.
Figure 3-9:

Sample DSSI Busses for an Expanded VAX 4000 Model 500
System

System

_BusO

Expander

DSSICabie

~Bus1

I Dssi Terminator Locations
1.

Enter the console mode.
The procedure for programming parameters for nSSI devices from
console mode requires that you issue commands to those devices at
the console prompt (»». You may enter these commands in either
uppercase or lowercase letters. Unless otherwise instructed, enter each
command, then press Return.
Enter console mode as follows:

a. Set the Break EnableIDisable switch on the system console module
to the enable position (up, position 1).
b. Set the Power switch for each unit (each system in a DSSI
VAXcluster configuration, and any expanders for expanded systems)
to on (1).
3-28

KA675/KA680/KA690 CPU System Maintenance

Wait for the system to display the console prompt (»».
2. To display the DSS! devices on embedded DSSI busses, enter SHOW
nSSI at the console prompt. To display the DSSI devices on KFQSAbased DSSI busses, enter SHOW UQSSP.
The firmware displays two lines of information for each device. For
embedded DSSI, the firmware displays the following:
•

The first line contains the bus number, node number, and node
name.

•

The second line contains the device name and Wlit number followed
by the device type in parentheses.

For embedded DSSI, the device name consists of the letters DIAu or
DIBu (MIAu or Mmu for the TF85 tape drive)-devices on bus 0 are
listed as DIA, devices on bus 1 are listed as DIB-and u is a unique
unit number. The embedded DSSI adapter for each bus is identified by
an asterisk (*).
The embedded nSSI display for Example 3-1 shows a system with four
DSSI devices (unit numbers 0--8) and an R400X expander with seven
nSS! devices (unit numbers 0--6).'
,

System Setup and Configuration

3-29

Example 3-1: SHOW DSSI Display (Embedded DSSI)
»>SHOK DSSl

OSSI Bus 0 Node 0
-DIAO (RF3l)
OSSI Bus 0 Node 1
-DIAl (RF3l)
OSSI Bus 0 Node 2
-OIA2 (RF3l)
OSSI Bus 0 Node 5
-MIAS (TF8S)
OSSI Bus 0 Node 6

(R7ALOC)
(R7EB3C)
(R7EB22)
(TFDR1)
(*)

OSSI Bus 1 Node 0 (SNEEZY)
-jjrEiJ

{:t\F3:i)

DSSI Bus 1 Node 1
-DIB1 (RF31)
DSSI Bus 1 Node 2
-OIB2 (RF3l)
DSSI Bus 1 Node 3
-DIB3 (RF31)
DSSI Bus 1 Node 4
-DIB4 (RF31)
DSSI Bus 1 Node 5
-OIBS (RF31)
DSSI Bus 1 Node 6
-DIB6 (RF3l)
OSSI Bus 1 Node 7

(DOPEY)
(SLEEPY)
(GRUMPY)
(BASHFUL)
(HAPPY)
(DOC)
(*)

»>

For KFQSA-based DSSI, the firmware displays the following:
•

The first line contains the UQSSP disk controller number and device
node name.

•

The second line contains the device name and unit number followed by
the device type in parentheses.
For KFQSA-based DSSI, the device name consists of the letters DUcu,
where c is the controller letter, and u is a unique unit number.

3-30 KA6751KA680/KA690 CPU System Maintenance

Example 3-2 shows a sample KFQSA-based DSSI bus.
Example 3-2:

SHOW UQSSP Display (KFQSA-Based DSSI)

»>SBOW UQSSP
UQSSP Disk Controller 0 (112150)
-DUAO (RF31)

UQSSP Disk Controller 1 (160334)
-DUB1

(RF31) .

UQSSP Disk Controller 2 (160340)
-DUC2 (RF31)

UQSSP Disk Controller 3 (160322)
-DUD3 (RF31)

UQSSP Tape Controller 0 (714500)
(TK70)

-MUAO

For the examples in this section, each device will be assigned an allocation
class of 1, and the system disk will be given a new node name. Also,
devices DIAO, DIAl, and DIA2; and DUAO, DUB1, DUC2, and DUD3 will
be assigned new unit numbers.
NOTE: The DUP server examples throughout this section are for RF-series
[SESe The displays for the TF85 tape drive differ slightly from the RF-series
displays.
3.7.3.3 Entering the DUP Driver Utility from Console Mode

To examine and change DSSI parameters, you must first activate the DUP
driver utility by setting host to the specific device for which you want to
modify or examine parameters.
Use the following command for embedded DSSI:
SET HOST/DOP /DSSI/BUS : <bus_number> <node_number> PARAMS

where:
<bus_number> is the DSS! bus nnmber (0 or 1), and <node_number> is the
bus node ID (0-6) for the device on the bus.
Use the following command for KFQSA-based DSSI:
SET HOST/DUP/OQSSP/DISK <controller_number> PARAMS

where:
<controller_number> is the controller number (provided by the SHOW
UQSSP display) for the device on the bus.

System Setup and Configuration

3-31

In Example 3-3, SET HOST /DOP lOSS I /BOS: 1 0 PARAMS is entered to start the
DUP server for the ISE at node 0 of embedded DSSI bus 1. In Example 3-4,
SET HOST/DUP/OQSSP/DISK 0 PARAMS is entered to start the DUP server for
the ISE at controller 0 of a KFQSA-based DSSI bus.
Example 3-3:

Accessing the DUP Driver Utility From Console Mode
(Embedded DSSI)

»>SET BOST/DOP/DSSI/BOS:l 0 PARAMS
Starting DUP server •••
Copyright (e) 1991 Digital Equipment Corporation
PARAMS>

Example 3-4:

Accessing the DUP Driver Utility From Console Mode
(KFQSA-Based DSSI)

»>SET BOST/DOP/OQSSP/DISK 0 PARAMS
Starting CUP server •••
Copyright (e) 1991 Digital Equipment Corporation
PARAMS>

3.7.3A Entering the DUP Driver Utility from VMS
To examine and change DSSI parameters, you must first access the DUP
driver utility by setting host to the specific device for which you want to
modify or examine parameters.

To access the DUP driver from VMS:
a. Connect to the Diagnostic and Utility Program (DUP) and load its driver
using the VMS System Generation Utility (SYSGEN) as shown below:
$ MeR SYSGEN
SYSGEN> CORNECT/NOADAPTER FYAO
SYSGEN> EXIT
$

b. Access the DUP driver by setting host to the specific device you want
to write protect. Use the following command:
SET HOST /DOP ISERVER-MSCP$DUP /TASK==PARAMS <node_name>

3-32 KA675/KA6801KA690 CPU System Maintenance

where:
<node_name> is the device node name (the node name, in parenthesis,
is listed using the VMS DeL command SHOW DEVICE Dl).
In Example 3-5, SET HOST/DUP/SERVER-MSCP$DOP/TASK=PARAMS R35F3C is
entered to start the DUP server for the ISE with a nodename of R35F3C.
Example 3-5: Accessing the CUP Driver Utility From VMS
$ HCR SYSGEN

SYSGEN> CONNECT !NOADAPTER FYAO
SYSGEN> EXIT
$ SET HOST/DOP/SERVER=MSCP$DOP/TASK=PARAMS R35F3C

Starting DUP server •••
Copyright (c) 1992 Digital Equipment Corporation
PARAMS>

3.7.3.5 Setting Allocation Class

After entering the DUP driver utility for a specified device, you can examine
and set the allocation class for the device as follows:
NOTE: The ALLCLASS parameter should only be set through console mode.

Setting the ALLCLASS parameter from VMS is not recommended.
1. At the PARAMS> prompt, enter SHOW ALLCLASS to check the allocation
class of the ISE to which you are currently connected.

2. Enter SET ALLCLASS 1 (or enter the allocation class you desire).
3. Enter SHOW ALLCLASS to verify the new allocation class.

System Setup and Configuration

3-33

Example 3-6 shows the steps for examining and changing the allocation
class for a specified device. In the example, the allocation class is changed
from an allocation class of 0 to an allocation class of 1.
Example 3-6: Setting Allocation Class for a Specified Device
PARAMS>SBOW ALLCLASS

Parameter

Current

Default

ALLCLASS

Type

Byte

Radix

Dec

PARAMS>SET ALLCLASS 1
PARAMS>SBOW ALLCLASS

Parameter

Current

ALLCLASS

Default
1

Type

Byte

Radix

Dec

3.7.3.6 Setting Unit Number

After entering the DUP driver utility for a specified device, you can examine
and set the unit number for the device as follows:
1. At the PARAMS> prompt, enter SHOW UNITNOM to check the unit number
of the ISE to which you are currently connected.

2. Enter SET UNITNUM 10 (or enter the unit number you desire).
3. Enter SET FORCEUNI 0 to override the default unit number value
supplied by the bus node ID plug.
4. Enter SHOW ONITNUM to verify the new unit number.
5. Enter SHOW FORCEONI to verify that the current value for the FORCEONI
parameter is o.
Example 3-7 shows the steps for changing the unit number of a
specified device from unit number 0 to unit number 10.
6. Label the device with its unit number, using the unit number labels
shipped with your system. Figure 3-10 shows where to affix a unit
number label on the device front panel.

3-34 KA675/KA680/KA690 CPU System Maintenance

Example 3-7: Setting a Unit Number for a SpecHied Device
PARAMS>SHOW ONITNOM
Current
Parameter

Default

Type

----------------------0 ----------------0 -------Word
UNITNUM

Radix

Dec

PARAMS>SET ONITNOM 10
PARAMS>SET FORCEtJNI 0
PARAMS>SHOW ONITNOM
Parameter

Current

Default

Type

-------- -------------------------------0 -------10
Word
ONITNUM

PARAMS>SHOW FORCEONI
Current
Parameter

Default

Type

--------- ----------------0 ----------------1 -------Boolean
FORCEUNI

Radix

Dec

Radix

0/1

System Setup and Configuration

3-35

Figure 3-10:

Attaching a MSCP Unit Number Label to the Device Front
Panel
RF30170-Series ISE

RF351SE

Attach Unit

Number Label ----;...;--~)

TF85
Attach Unit

Number Labels

MLO~7178

3-36 KA675/KA6801KA690 CPU System Maintenance

3.7.3.7 Setting Node Name

After entering the DUP driver utility for a specified device, you can examine
and set the node name for the device as follows:
1. At the PARAMS> prompt, enter SHOW NODENAME to check the node name of
the ISE to which you are currently connected.
2. Enter SET NODENAME SYSDSK (or enter the desired alphanumeric node
name of up to eight characters).
3. Enter SHOW NODENAME to verify the new node name.
Example 3-8 shows the steps for changing the node name of a specified
device from the factory-supplied name to SYSDSK
Example 3-8: Changing a Node Name for a Specified Device
PARAMS>SBOW HODENJUm

Parameter
NODENAME

Current
R7CZZC

Default
RF31

Type

Radix

String

Ascii

Type

Radix

String

Ascii

PARAMS>SET HODENAME SYSDSK
PARAMS>SBOW HODENJUm

Parameter
NODENAME

Current
SYSDSK

Default
RF31

3.7.3.8 Setting System 10
NOTE: This parameter is modified only when warm swapping a device.
All parameters for the replacement device should be programmed to match
those of the original device. Refer to the nSS! Warm. Swapping Guide for
BA400-Senes Enclosures and KFQSA Adapters.

After entering the DUP driver utility fOT ill specified device, you can examine
and set the system ID for the device as follows:
1. At the PARAMS> prompt, enter SHOW SYSTEMID to check the system ID of

the device to which you are currently connected.
2. Enter SET SYSTEMID System ID (enter the desired serial number-based
system ID).

System Setup and Configuration

3-37

3. Enter SHOW SYSTEMID to verify the new system ID.
Example 3-9 shows the steps for changing the system ID of a specified
device from the factory-supplied system ID to 1402193310841 (the system
ID for the replacement device is programmed to match that of the origina}).
Example 3-9: Changing a System ID for a Specified Device
PARAMS>SBOW SYSTEHID
Parameter
Current

Default

Type

---------------------_\
-----------------------SYSTEMID
0402193310841
0000000000000 Quadword
P~wlS>Si:4 S"'.i'~~&j

Radix
Hex

j" .. O"l;33l0&·n

PARAMS>SBOW SYSTEMID
Parameter

Current

Default

Type

---------------------------------------------SYSTEMID
0000000000000 Quadword
1402193310841

Radix
Hex

3.7.3.9 Exiting the DUP Driver Utility

After you have completed setting and examining DSSI device parameters,
enter the WRITE command at the PARAMS> prompt to save the device
parameters you have changed using the SET command. The changes are
recorded to nonvolatile memory.
IT you have changed the allocation class or node name of a device, the DUP
driver utility will ask. you to initialize the controller. Answer Yes (Y) to
allow the changes to be recorded and to exit the DUP driver utility.
IT you have not changed the allocation class or node name, enter the EXIT
command at the PARAMS> prompt to exit the DUP driver utility for the
specified device. Example 3-10 shows the procedure for saving parameter
changes. In the example, the controller is initialized.

3-38 KA675/KA680/KA690 CPU System Maintenance

Example 3-10:

ExHlng the DUP Driver UlilHy for a Specified Device

PARAMS>WRITE
Changes require controller initialization, ok? [Y/(N}] Y
Stopping DUP server •..
»>

NOTE: You must repeat the procedures in this section for each device for
which you want to change parameters.

Example 3-11 shows the DSSI busses for the embedded DSSI adapters after
the unit numbers for the disk devices on bus 0 have been changed from 0,
1, and 2 to 10, 11, and 12 (by adding 10 to the bus node ID number, the
unit number's least significant digit will still correspond to the number on
the bus node ID plug). Note that the bus 0 device names are now DIA10,
DIA11, and DIA12.
Example 3-11: SHOW DSS! Display
»>SBOW DSSI
DSSI Bus 0 Node 0 (SYSDSK)

-DIA1O (RF3l)
DSSI Bus 0 Node 1 (R7EB3C)
-DIAll (RF3l)
DSSI Bus 0 Node 2 (R7EB22)
-DIA12 (RF3l)
DSSI Bus 0 Node 5 (TFDRl)
-MIAS (TFS5)
DSSI Bus 0 Node 6 (*)
DSSI Bus 1 Node 0 (SNEEZY)
-DIBO (RF3l)
DSSI Bus 1 Node 1 (DOPEY)
-DIB1 (RF3l)
DSSI Bus 1 Node 2 (SLEEPY)

-DIB2

(RF31)

DSSI Bus 1 Node 3 (GRUMPY)
-01B3 (RF3l)
DSS1 Bus 1 Node 4 (BASHFUL)
-01B4 (RF3l)
OSSI Bus 1 Node 5 (HAPPY)
-OIB5 (RF3l)

Example 3-11 (continued on next page)

System Setup and Configuration

3-39

Example 3-11 (Cont.): SHOW DSSI Display
DSSI Bus 1 Node 6 (DOC)
-DIB6 (RF31)
DSSI Bus 1 Node 7 (*)

»>

Example 3-12 shows the sample KFQSA-based DSSI bus after the unit
numbers have been changed from 0, 1, 2, and 3 to 20, 21, 22, and 23. Note
that the device names are now DUA20, DUB21, DUC22, and DUD23.
Example 3-12: SHOW UQSSP Display (KFQSA-Based DSSI)
.>.>.>SiiOii uQgSP

UQSSP Disk Controller 0
-DUA2 0 (RF31)
UQSSP Disk Controller 1
-DUB21 (RF31)
UQSSP Disk Controller 2
-DUC22 (RF31)
UQSSP Disk Controller 3
-DUD23 (RF31)
UQSSP Tape Controller 0
-MUAO (TK70)

(772150)
(760334)
(760340)
(760322)
(774500)

3.7.4 Write-Protecting an RF35 ISE
You may want to write-protect an ISE containing sensitive data you do not
want changed or accidentally erased.
The system disk (the ISE containing system software) and ISEs containing
work areas for users should be write-enabled, the normal operating setting.
For the RF35 ISE, which has no Write-Protect button, you set writeprotection through VMS commands or through finnware commands in
console mode.
3.7.4.1 Software Write-Protect for RF-Series ISEs

Since the RF35 does not have a Write-Protect button, the software writeprotect is the primary method for write-protecting an RF35.
The software write-protect is available through VMS using the MOUNT
utility with the INOWRITE qualifier.

To software write-protect an ISE, enter the following DCL command from
the VMS operating system.

3-40 KA6751KA680/KA690 CPU System Maintenance

MOUNT <device_name> <volume_label>/SYSTEM/NOWRlTE

where:
<device_name> is the device name, as shown using the VMS DCL command
SHOW DEVICE DI, and <volume_label> is the volume label for the device.
For example,
$ HOUNT $l$OIAl OMEGA/SYSTEM/NOWRITE

will software write-protect device $l$DIAl.
Dismounting, and then remounting the device (without using the
INOWRITE qualifier), will write-enable the device.
Use the VMS DeL command SHOW DEVICE DI to check the protection
status of the drive. A write-protected drive will show a device status of
"Mounted wrtlck". Refer to your VMS documentation for more information
on using the MOUNT Utility.
CAUTION: When you dismount tr"en mount the device again, it will no
longer be write-protected.
3.7.4.2 Hardware Write-Protect For RF35 ISEs

The hardware write-protect provides a more permanent write-protection
than the software write-protect in that, once you hardware write-protect
an RF35, it remains write-protected, regardless of the availability of
the operating system or if the system is powered-down. In addition, a
hardware write-protect cannot be removed using the MOUNT command.
The hardware write-protect simply provides the same degree of writeprotection available to RF-series ISEs that have a Write-Protect button.
You should consider hardware write-protecting an RF35 in the following
situations:
•

If you want to write-protect an RF35 ISE when the VMS operating

system is not available, such as before running the MicroVAX
Diagnostic Monitor (MDM).
•

If you want to ensure that an RF35 remains write-protected, since the
hardware write-protect cannot be removed using the VMS command
MOUNT and will remain in effect even if the operating system is
brought down.

You can hardware write-protect an RF35 from VMS or through firmware
commands entered at the console prompt (»». Use the following
instructions:

System Setup and Configuration

3-41

1. Access the Diagnostic and Utility Program (DUP) driver for the device
you want to write-protect.

•

To access the DUP driver from console mode:
a.

Enter console mode by pressing the Halt Button or powering up
the system with the Break EnablelDisable switch set to enable
(up, position 1).
CAUTION: Haltl.ng your system without following the shutdown
procedure described in your system software manuals may result
in loss of data.

b. Access the DUP driver by setting host to the specific device you
want to write protect.
Use the following command for embedded DSSI:
SET HOST/DOP /DSSI/BOS : <bus_number> <node_number> PARAMS

where:
<bus_number> is the DSSI bus number (0 or 1), and <Ilode_
number> is the bus node ID (0-6) for the device on the bus
(bus number and node number are listed in the SHOW DSSI
display).
Use the following command for KFQSA-based DSSI:
SET HOST/DOP/OQSSP/DISK <controller_number> PARAMS

where:
<controller_number> is the controller number Oisted in the
SHOW UQSSP display) for the device on the bus.

•

To access the DUP driver from VMS:
a.

Connect to the Diagnostic and Utility Program (DUP) and load
its driver using the VMS System Generation Utility (SYSGEN)
as shown below:
$ HCR SYSGEN

SYSGEN> CONNECT /N0ADAP'lER FYAO
SYSGEN> EXIT
$

b. Access the DUP driver by setting host to the specific device you
want to write protect. Use the following command:
SET HOST /DOP / SERVER-MSCP$DOP /TASK=PARAMS <node_name>

3-42 KA675/KA6801KA690 CPU System Maintenance

where:
<node_name> is the device node name (the node name, in
parenthesis, is listed in the SHOW DEVICE DI display).
2. At the PARAMS> prompt, enter SET WRT_PROT 1 to write-protect the ISE
to which you are currently connected.
NOTE: To verify that you have set host to the intended drive, you can
enter the command LOCATE at the PARAMS> prompt. The LOCATE
command causes the drive's Fault indicator to blink momentarily.

3. Enter SHOWWRT_PROT to verify the WRT_PROT parameter is set to 1.

4. After you have completed setting and examining the WRT_PROT device
parameter, enter the WRITE command at the PARAMS> prompt to save
the device parameter. The change is recorded to nonvolatile memory.
5. Enter the EXIT command at the PARAMS> prompt to exit the DUP driver
utility for the specified device.
Example 3-13 provides an example of setting a hardware write-protect
through firmware; Example 3-14 provides an example of setting a hardware
write-protect through VMS.
Example 3-13: Setting Hardware Write-Protection Through Flnnware
»>SET EOST/DOP/DSSI/BOS:O 1 PARAMS
Starting DOP server •••
Copyright (c) 1992 Digital Equipment Corporation

PARAMS>SET WRT_PROT 1
PARAMS>WRlTE
PARAMS>SBOW WRT_PRO'l'
Parameter

Current

WRT PROT
PARA..~S>EX!T

Type

Default
1

Boolean

Radix
0/1

Exiting ••.
Stopping DOP server ..•

»>

System Setup and Configuration

3-43

Example 3-14: Setting Hardware Write-Protection Through VMS
$ MCR SYSGEN

SYSGEN> CONNECT /NOADAPTER FYAO
SYSGEN> EXIT
$ SET HOST/DW /SERVER=MSCP$DOP /TASX=PARltMS R35F3C

Starting DUP server •.•
Copyright (c) 1992 Digital Equipment Corporation

PARAMS>SET 1mT PROT 1
PARAMS>WlUTE PARAMS>SHOW WR'.r_PROT
Parameter

Current

Default

Type

Radix

WRT PROT

n'1

PARAMS>EXIT
Exiting ..•
Stopping DUP server •••
$

To remove the hardware write-protection, repeat the above procedure, only
set the WRT_PROT value to O.
You can verify that the device is write-protected while running VMS-when
you issue the VMS DCL command SHOW DEVICE DI, a write-protected
drive will show a device status of "Mounted wrtlck". If you issue the VMS
command SHOW DEVICFJFULL, a write-protected drive will be listed as
"software write-locked".
NOTE: You cannot remove hardware write-protection using the VMS
MOUNT utility.

3.7.5 Setting System Parameters: Boot Defaults, Bootflags,
Halt and Restart Action
Several firmware commands are used to set and examine system
parameters.

3.7.5.1 Setting the Boot DefauH
To direct the system to boot automatically from a specific device or to change
the setting of the default boot device, put the system into console mode and
at the »> prompt, enter "SET BOOT device-name". For example,
»>SET BOOT EZAO

sets the system default boot device to be the Ethernet controller.
Once you have selected a boot device, the system autoboots from that device
each time you tum it on (provided the Break EnablelDisable switch is set to

3-44 KA675/KA6801KA690 CPU System Maintenance

disable or that a halt action of REBOOT or RESTART_REBOOT has been
defined}.
U sing "SET BOOT deuice-name,deuice-name,deuice-name", you can also
specify a string of default boot devices (up to 32 characters, with devices
separated by commas and no spaces) for which the system will check for
boatable software. The system checks th e devices in the order specified and
boots from the first one that contains bootable software. For example,
»>SET BOOT DUAO, DIAO , MIAS, EZAO

directs the system to use nUAO, DIAO, MIA5, and EZAO as the default
boot devices. When the system autoboots, or if the BOOT command is used
without specifying a device, the system will boot from the first default boot
device that contains bootable software.
NOTE: If included in a string of boot deuices, the Ethernet deuice, EZAO,
should only be placed as the last deuice of the string. The system will
continuously attempt to boot from EZAO..

Refer to Appendix A for examples.
Supported Boot Devices
Table 3-3 lists the boot devices supported by the CPU. The table correlates
the boot device names expected"in a BOOT command with the corresponding
supported devices. The device name used for the bootstrap operation is one
of three:

•

EZAO, if no default boot device has been specified

•

The default boot device specified at initial power-up or through SET
BOOT

Name explicitly specified in a BOOT command line

Boot device names consist of a device code of at least two letters (A through
Z) in length, followed by a single-character controller letter (A through Z),
and ending in a device unit number (0 through 16,383).

System Setup and Configuration

3-45

Table 3-3: Boot Devices Supported by the KA675/KA680IKA690
Boot Name Controller Type

Device 1'ype(s)

[node$]DImu
DUcu

RF:xx
RF:xx
RAxx

Disk
On-board DSSI
KFQSA DSSI
KDA50 MSCP
RDXS MSCP

RDxx

Compact Disc
r_ .. ~ ... ~TnVA..

~_-"~""'IIiI"'~"';""'..iIroA"_

V'7nQA Q~T
.........
"'"" ...... ..... """"' __
~

P.P.!)~=:

DUcu

KRQ50 MSCP

RRD40

Tape
[node$]MImu
MUcu

MKAu

On-board DSSI

TF85

TQK50MSCP
TQK70MSCP

TK50
TK70

KLESI

TU8lE

KZQSASCSI

TLZ04

Network
EZAO

On-board Ethernet

XQcu

DESQA

PRAu
PRBO

MRVll

PROM
Customer EPROM space -

NOTE: For diskless and tapeless systems that boot software over the
network, select only the Ethernet adapter. All other boot devices are
inappropriate.

3.7.5.2 Setting Boot Flags
The Virtual Memory Boot (VMB) action is qualified by the value passed to
it in R5. R5 contains boot flags that specify conditions of the bootstrap.
The firmware passes to VMB either the R5 value specified in the BOOT
command or the default boot fiag value specified with a SET BFLG
command. The VMB boot flags are listed in Table 34.

3-46 KA675/KA680IKA690 CPU System Maintenance

Refer to Appendix A for examples.

Table 3-4: Virtual Memory Bootstrap (VMB) Boot Flags
Bit

Name

Description

RPB$V_CONY

RPB$V_DEBUG

RPB$V_INIBPT

RPB$V_BBLDCK

RPB$V_DIAG

Conversational boot. At various points in the system boot
procedure, the bootstrap code solicits parameters and other
input from the console terminal.
Debug. If this flag is set, VMS maps the code for the XDELTA
debugger into the system page tables of the running system.
Initial breakpoint.
If RPB$V-DEBUG is set, the VMS
operating system executes a BPT instruction in module INIT
immediately after enabling mapping.
Secondalj' bootstrap from bootblock. When set, VMB reads
logical block number 0 of the boot device and tests it for
conformance with the bootblock format. If in conformance, the
block is executed to continue the bootstrap. No attempt is made
to perform a Files-ll bootstrap.
Diagnostic bootstrap. When set, the load image requested is

RPB$V..BOOBPT

RPB$VJIEADER

RPB$VJIALT

31:28

RPB$V_TOPSYS

[syso.sYSMAINT]DIAGBOOT.EXE.

Bootstrap breakpoint. When set, a breakpoint instruction is
executed in VMB and control is transferred to XDELTA before
booting.
.
Image header. When set, VMB transfers control to the address
specified by the file's image header. When not se~ VMB
transfers control to the first location of the load image.
File name solicit. When set, VMB prompts the operator
for the name of the application image file. A maximum 39
character file specification is allocated at RPB$TJ'IIE. Only
16 characters are utilized in both tape boot and network MOP
V3 booting.
Halt before transfer. When se~ VMB halts before transfening
control to the application image.
This field can be any value from 0 through F. This flag changes
the top-level directozy name for system disks with multiple
operating systems. For example. if TOPSYS is 1. the top-level
directory name is (SYSl...J. This does not apply to network
bootstraps.

3.7.5.3 Setting the Halt Action

The user-defined halt action feature allows users to determine what action
should be taken on error halts and at power-up. The halt action is defined
using the SET HALT command and overrides the setting of the Break
Enable1Disable switch.
Table 3-5 summarizes the action taken on all halt conditions (excluding
external halts). The user-defined halt is used when the OIS Mailbox halt
action field is 0 and on power-up ifbreaks are enabled. Refer to ADnendix A
•for an example of the SET HALT command.

System Setup and Configuration

3-47

For external halts caused by pressing the Halt button on the SCP or
pressinglsREAKVCTRL-P (ifbreaks are enabled), the firmware enters console
mode.
NOTE: Using the console command SET CONTROLP, you can specify the
control character, ICtrvpL rather than ISreakl to initiate a break signal.

Table 3-5: Actions Taken on a Halt
ResetJ
Power-Up
or Halt

Break
Enable
Switch

UserOIS
Defined
Mailbox
Halt ACtiOD BaltAction Action(s)

'!'

n, .,

.. ,_'v
2,4-

F
F

1
0

0
0

1
2
3

..
x

F
F
F
F

x
x
x

0
0
0
0

F
F
F

x
x
x

1
2
3

x
x

"Jj.;""~,,e"";~

............. af"\la

-"~-"'''''''':'''-:.I'''J

;.~

........ -..., ... ~

Diagnostics, if success boot, if
either fail console
Diagnostics, if success boot, if
either fail console
Console
Restart, if this fails boot, if
that fails console
Restart, if it fails console
Boot, if it fails console
Console
Restart, if this fails boot, if
that fails console
Restart, if it fails console
Boot, if it fails console
Console

"T" indicates that the condition is true.

"F" indicates that the condition is false.
"X," indicates that the condition is "don't care".

Halt Action 0 =DEFAULT
Halt Amon 1 =RESTART
Halt Action 2 = REBOOT
Halt Action 3 = HALT
Halt Action 4- = RESTART_REBOOT

3-48 KA675!KA6801KA690 CPU System Maintenance

Chapter 4

System Initialization and Acceptance
Testing (Normal Operation)
This chapter describes the system initialization, testing, and bootstrap
processes that occur at power-up. In addition, the acceptance test procedure
to be performed when installing a system or whenever adding or replacing
FRUs is described.

4.1 Basic Initialization Flow
On power-up, the firmware identifies the console device, optionally performs
a language inquiry, and runs the diagnostics.
Power-up actions differ, depending on the state of the Power-Up Mode
switch on the console module. The mode switch has three settings: loopback
test, language inquiry, and run. The differences are described below.
The firmware waits for power to stabilize by monitoring SCR<15>(POK).
Once power is stable, the firmware verifies that the console battery
backup RAM (BBU RAM) is valid (backup battery is charged) by checking
SSCCR<31>{BLO). Ifit is invalid or zero (battery is discharged), BBU RAM
is initialized.
After the battery check, the firmware tries to determine the type of terminal
attached to the console seriallineo It uses this information to determine if
multinational support is appropriate.
Power-Up Mode Switch Set to Test

Use the test position on the H3604 to verify a proper connection between
the CPU and the console terminal. During the test, the firmware toggles
between' the active and passive states. Refer to Chapter 5 for instructions
on performing loopback tests.
Power-Up Mode Switch Set to Language Inquiry
If the Power-Up Mode switch is set to language inquiry mode, or the

firmware detects that the contents of BBU RAM are invalid, the firmware
prompts you for the language to be used for displaying the following system
messages (if the console terminal supports the multinational character set).

System Initialization and Acceptance Testing (Normal Operation)

4-1

Loading system software.
Failure.
Restarting system software.
Performing normal system tests.
Tests completed.
Normal operation not possible.
Bootfile.
Memory configuration error.
No default boot device has been specified.
Available devices.
Device?
Retrying network bootstrap.

The language selection menu appears under the conditions listed in
Table 4-1. The oosition of the Break EnablelDisable switch has no effect
on these conditions. The firmware will not prompt for a language if the
console terminal, such as the VT100, does not support the multinational
character set (MeS).

Table 4-1: Language Inquiry on Power·Up or Reset
Mode

Language Not
Previously Setl

Language
Previously Set

Language Inquiry
Run

Prompt2
Prompt

Prompt
No Prompt

1 Action if contents of BBU RAM invalid same as Language Not Previously Set.
2Prompt = Language selection menu displayed.

The language selection menu is shown in Example 4-1. If no response is
received within 30 seconds, the firmware defaults to English (5).

4-2 KA675/KA680/KA690 CPU System Maintenance

Example 4-1: Language Selection Menu
KA6nn-A Vn.n VMB n.n
1) Dansk
Deutsch (Deutschland/Osterreich)
Deutsch (Schweiz)
English (United Kingdom)
English {United States/Canada}
6} Espanol
7) Fran~ais (Canada)
8) Fran~ais (France/Belgique)
9) Fran9ais (Suisse)
10) Italiano
11) Nederlands
12) Norsk
13) Portugues
14) Suomi
15) Svenska
(1 •• 15) :

2}
3}
4)
5)

NOTE: The information contained within the parentheses indicates the
specific keyboard variant.

In addition, the console may prompt you for a default boot device following
a successful diagnostic countdown.
After the language inquiry, the firmware continues as if on a normal powerup.
Power-Up Mode Switch set to Run
The console displays the language selection menu if the Power-Up Mode
switch is set to run mode and the contents of BBU RAM are invalid or a
language has not yet been selected. The next step in the power-up sequence
is to execute the bulk of ROM-based diagnostics. In addition to message
text, a countdown is displayed in Example 4-2.

System Initialization and Acceptance Testing (Normal Operation)

4-3

Example 4-2: Nannal Diagnostic Countdown
KA6nn-A Vn.n VMB n.n
Performing normal system tests.
66 .• 65 •. 64 •. 63 .• 62 •• 61 .. 60 •• 59 •. 58 •. 57 •• 56 •• 55 •. 54 •• 53 •• 52 •• 51 •.
50 •• 49 •• 48 •. 47 •• 46 •• 45 •. 44 •• 43 •• 42 •. 41 •• 40 •. 39 •• 38 •• 37 •. 36 •• 35 ••
34 •• 33 •. 32 .. 31. .30 .• 29 .• 28 .. 27 •. 26 .. 25 .. 24 .• 23 •• 22 .. 21. .20 •. 19 ••
18 .• 17 •. 16 .. 15 .• 14 .• 13 •. 12 .. 11 .• 10 •. 09 .• 08 .• 07 •• 06 •. 05 .. 04 .• 03 ..

Tests completed.

The console uses the saved console language if the mode switch is set to
run mode and the contents of BBU RAM are valid.

4.2 power-On 5elf-lests (POST)
Power-on self-tests provide core testing of the system kernel. The CPU,
memory, DSSI bus, and Q-bus are tested, certain registers are flushed,
and data structures are set up to initialize and set the system to a known
state for the operating system.
.

4.2.1 Power-Up Tests for Kernel
In a nonmanufacturing environment where the intended console device
is the serial line unit (SLU), the console program perfonns the following
actions at power-up:
1. Checks for POK

2. Establishes SLU as console device.
3. Prints banner message.
The banner message contains the processor name, the version of the
firmware, and the version of VMB. The letter code in the firmware
version indicates if the firmware is pre-field test, field test, or official
release. The first digit indicates the major release number and the
trailing digit indicates the minor release number (Figure 4-1).

4-4 KA6751KA680/KA690 CPU System Maintenance

Figure 4-1: Console Banner
KA6nn-A V n.n. VMS n.n

II :

minor release of VMS
major release of VMS

minor release of firmware

major release of firmware

L
4.

type of release: X - engineering release
T - field test release
V - volume release

processor type
MLo-ooB459

Displays language inquiry menu on consoie It console supports
multinational character set (MCS) and any of the following are true:
•

Battery is dead.

•

Power-Up Mode switch is set to language inquiry mode.

•

Contents of SSC RAM are invalid.

5. Calls the diagnostic executive (DE) with Test Code =O.
a. DE determines environment is nonmanufacturing from H3604.
b. DE executes script Al (Tests CPU, Floating Point Accelerator (FPA),
and memory).
While the diagnostics are nmning: the LEDs on the H3604 display
a hexadecimal test code ranging from F to 3 before booting the
operating system, and 2 to 0 while booting the operating system.
A different countdown appears on the console terminal. Refer to
Table 5-9 for a complete explanation of the power-up test display.
Table 4-2 lists the LED codes and the associated actions performed
at power-up. Example 4-3 shows a successful power-up to a list of
bootable devices.
c.

DE passes control back to the console program.

6. Issues end message and »> prompt.

System Initialization and Acceptance Testing (Normal Operation)

4-5

Table 4-2: LED Codes
LED ValueActioDS
F

Initial state on power-up, no code has executed

Entered ROM space, some instructions have executed

Waiting for power to stabilize (POI{)

SSC RAM, SSC registers, and ROM checksum tests

O-bit memory, interval timer, and virtual mode tests

FPA tests

Backup cache, primary cache, and memory tests

NMC, NCA, memory, and 110 interaction tests

CQBIC (Q22-bus) tests

Console loopback tests

~CDSSlm~mtests

SGEC Ethernet mbsystem tests

"Console 110" mode

Control passed to VMB

Control passed to semndary bootstrap

"Program 110" mode, control passed to operating system

4-6 KA675/KA680/KA690 CPU System Maintenance

Example 4-3: Successful Power-Up to List of Bootable Devices
KA6nn-A Vn.n VMB n.n
Performing normal system tests.
66 .• 65 .. 64 •. 63 •• 62 •• 61 •. 60 •• 59 .• 58 •• 57 •• 56 .. 55 •• 54 •• 53 •• 52 •• 51 ••
50. . 4 9. . 4 8. • 47. • 4 6. • 4 5. . 4 4 • • 4 3. • 42. • 41 • • 4 o. . 3 9. • 3 8 • • 37 . • 3 6. • 35. •
34 •• 33 •• 32 •• 31 .• 30 •• 29 •• 28 .• 27 •. 26 .• 25 .• 24 •• 23 •• 22 •• 21 .• 20 •• 19 ••
18 .. 17 •• 16 •• 15 .. 14 .• 13 •• 12 .• 11. .10 • • 09 .• 08 .• 07 •• 06 •• 05 •• 04 • • 03 ••

Tests completed.
Loading system software.
No default boot device has been specified.
Available devices.

-DIAO (RF73)
-DIAl (RF73)
-MIAS (TF85)
-EZAO (08-00-2S-06-10-42)
Device? [EZAO]:

4.2.2 Power-Up Tests for Q...Bus Options
Module self-tests run when you power up the system. A module self-test
can detect hard or repeatable errors, but usually not intermittent errors.
Module LEDs display pass/fail test results:
•

A pass by a module self-test does not guarantee that the module is good,
because the test usually checks only the controller logic.

•

A fan by a module self-test is accurate, because the test does not require
any other part of the system to be working.

The following modules do not have LED self-test indicators:
DFAOI
DPVii
DRQ3B

KLESI

LPVll
TSV05
The following modules have one green LED, which indicates that the
module is receiving +5 and + 12 Vdc and has passed self-tests:
CXA16
CXB16

CXYOS

System Initialization and Acceptance Testing (Normal Operation)

4-7

4.2.3 Power-Up Tests for Mass Storage Devices
An RF-series ISE may fail either during initial power-up or during normal
operation. In both cases, the failure is indicated by the lighting of the red
fault LED on the drive's front panel. The ISE also has a red fault LED, but
it is not visible from the outside of the system enclosure.
If the drive is unable to execute the Power-On Self-Test (POST) successfully,
the red fault LED remains lit and the ready LED does not come on, or both
LEDs remain on.

POST is also used to handle two types of error conditions in the drive:

•

Controller errors are caused by the hardware associated with the
(;uii.ti"vlllC?:i- fWlCtiuu uf tbe \!rive ~cdu.l~ .
cc~t=cnar crrcr iz fc.~! to
the operation of the drive, since the controller cannot establish a logical
connection to the host. The red fault LED lights. If this occurs, replace
the drive module.
.;AJ.

•

4.3 CPU ROM-Based Diagnostics
The KA6751KA6801KA690 ROM-based diagnostic facility is the primary
diagnostic tool for troubleshooting and testing of the CPU, memory,
Ethernet, and DSSI subsystems. ROM-based diagnostics have significant
advantages:
•

Load time is virtually nonexistent.

•

The boot path is more reliable.

•

Diagnosis is done in a more primitive state.

The ROM-based diagnostics can detect failures in field-replaceable units
(FRUs) other than the CPU module. For example, they can isolate one of
up to four· memory modules as FRUs. (Table 5-9 lists the FRUs indicated
by ROM-based diagnostic error messages.)
The diagnostics run automatically on power-up_ While the diagnostics are
running, the LED on the H3604 displays a hexadecimal number; while
booting the operating system, 2 through 0 display.
The ROM-based diagnostics are a collection of individual tests with
parameters that you can specify. A data structure called a script points

4-8

KA675/KA680/KA690 CPU System Maintenance

to the tests (see Section 4.3.2). There are several field and manufacturing

scripts.
A program called the diagnostic executive determines which of the available
scripts to invoke. The script sequence varies if the system is in the
manufacturing environment. The diagnostic executive interprets the script
to determine what tests to run, the correct order to run the tests, and the
correct parameters to use for each test.

The diagnostic executive also controls tests so that errors can be detected
and reported. It ensures that when the tests are run, the machine is left
in a consistent and well-defined state.

4.3.1 Diagnostic Tests
Example 4-4 shows a list of the ROM-based tests and utilities. To get this
listing, enter T 9E at the console prompt (T is the abbreviation of TEST).
The column headings have the following meanings:
NOTE: Base addresses shown in this document may not be the same as
the addresses you see when you run T 9E. Run T 9E to get a list of actual
addresses. See Example 4-4.

•

Test is the test number or utility code.

•

Address is the base address of where the test or utility starts in ROM.
If a test fails, entering T FE displays diagnostic state to the console.
You can subtract the base address of the failing test from the last_
exception-pc to find the index into the failing test's diagnostic listing.

•

Name is a brief description of the test or utility.

•

Parameters shows the parameters for each diagnostic test or utility.
These parameters are encoded in ROM and are provided by the
diagnostic executive. Tests accept up to 10 parameters. Tne asterisks
(*) represent parameters that are used by the tests but that you cannot
specify individually. These parameters are displayed in error messages,
each one preceded by identifiers PI through PIO.

Parameters that you can specify are written out, as shown in the following
examples:
30
54

2005C33C
20055181

Memory Init Bitmap *** mark Hard SBEs ******

Virtual Mod;

*********-

For example, the virtual mode test contains several parameters, but you
cannot specify any that appear in the table as asterisks. To run this test
individually, enter:

System Initialization and Acceptance Testing (Normal Operation)

.4-9

»>T 54

The MEM_bitmap test, for example, accepts 10 parameters, but you can
only specify mark_hard_SBEs because the rest are asterisks. To map out
solid, single-bit ECC memory errors, type:
»>T 30 0 0 0 1

Even though you cannot change the first three parameters, you need to
enter either zeros (0) or ones (1) as placeholders. Zeros are more common
and are shown in this example. The zeros are placeholders for parameters
1 through 3, which allows the program to parse the command line correctly.
The diagnostic executive then provides the proper value for the test.
Vnn 0,"+0,. 1 ~n,. "0;:1,.0;:1,",,0+0,. A......+n ;"1'1;,.0+0
+'ho
n"t
cznl;1'1
--=-"",'"" t'ho;:l+
""' ....
..... """'" +oC!t
-" .... -"- C!'hn"ll'1
-,"' ....... _.,'"""" ,,"on
..
""
6OOoW_

.., _ _ _ _ . . . . . . . .

... a

r--~"'~-'"""'"

''IIf

"" ......

~.-..'"

,"'.~

'.~.'"

..w_"

-..

--'-~'1

single-bit as well as multi-bit ECC memory errors. You then terminate the
command line by pressing IRETURN I. You do not need to specify parameters 5
through 10; placeholders are needed only for parameters that precede the
user-definable parameter.
For the most part tests and scripts can be run without any special setup. If
a test or script is run interactively without an intervening power up, such as
after a system crash or shutdown, enter the UNJAM and INIT commands
before running the tests or script. This will ensure that the CPU is in a
well known state. If the commands are not entered, misleading errors may
occur.
Other considerations to be aware of when running individual tests or scripts
interactively:
•

When using the TEST or REPEAT TEST commands, you must specify
a test number, test code or script number following the TEST command
before pressing IRElURN I.

•

The memory bitmap and Q-bus scatter-gather map are created in main
memory and the memory tests are run with these data structures left
intact. Therefore, the upper portion of memory should not be accessed
to avoid corrup~g these data structures. The location of the maps are
displayed using the SHOW MEMORYIFULL command.

4.3.2 Scripts
Most of the tests shown by utility 9E are arranged into scripts. A script
is a data structure that points to various tests and defines the order in
which they are run. Scripts should be thought of as diagnostic tablesthese tables do not contain the actual diagnostic tests themselves, instead
scripts simply define what tests or scripts should be run, the order that the
tests or scripts should be run, and any input parameters to be parsed by
the Diagnostic Executive.

4-10 KA675/KA680/KA690 CPU System Maintenance

Different scripts can run the same set of tests, but in a different order and
lor with different parameters and flags. A script also contains the following
information:
•

The parameters and flags that need to be passed to the test.

•

Where the tests can be run from. For example, certain tests can be run
only from the FEPROM. Other tests are program-independent code,
and can be run from FEPROM or main memory to enhance execution
speed.

•

What is to be shown, if anything, on the console.

•

What is to be shown, if anything, in the LED display.

•

What action to take on errors (halt, repeat, continue).

The power-up script runs every time the system is powered on. You can
also invoke the power-up script at any time by entering T o.
Additional scripts are included in the ROMs for use in manufacturing and
engineering environments. Customer Services personnel can run these
scripts and tests individually, using the T command. When doing so, note
that certain tests may be dependent upon a state set up from a previous
test. For this reason, use the UNJAM and INITIALIZE commands before
running an individual test. You do not need these commands on system
power-up because the system power-up leaves the machine in a defined
state.
Customer Services Engineers (CSE) with a detailed knowledge of the
system hardware and firmware can also create their own scripts by using
the 9F User Script Utility. Table 4-3 lists the scripts available to Customer
Services.

System Initialization and Acceptance Testing (Normal Operation)

4-11

Example 4-4: Test 9E
»>'l 9E

Test

• Address

30
31
32
33
34
35
37
3F
40
41

42
46
47
48
49
4A
4B
4C
40
4E
4F
Sl
52
53
54
55
56
58
59
5C
SF
60
63
80
81
82
83
84
85
86
87
90
91
99
9A
9B
9C
90
9E

20053EOO
20054E14
20063A20
200641BC
20064CBO
20064E4C
2005B730
200 67AEC
2006868C
2006443C
20062608
20056S0C
200SA3CC
2006782C
20063F48
20061878
2006342C
20063138
20061EOC
20062AC8
200616F8
20061CEO
20062814
200SA88C
200SABCC
200SAE9C
2005A4A2
200SB052
200SFF38
200607B4
2005F080
2005FSE8
2005E36C
20050067
2005B504
20065280
2005B236
2005B3FB
200 577FA
20058EB4
20056A34
20056EFO
200SAOF8
2005AB4A
2005AAEO
20065048
20050080
20064ECC
200SB7FA
2005E138
2005B208

Name

Parameters

SCB
De executive
Memory Inlt BItmap ••• mark_Hard_SBEs ••••••
Memory-Setup CSRs ••••••••• *
NMC registers
••••••••• *
NMc:powerup
SSC ROM
•
B Cache diag mode bypass_test_mask ••••• **.*
Cache_W:Memory
bypass_test_mask •••••••••
Mem FOM Addr shorts ••• cont on err ••••••
Memory Count-pages First boaru Last bd Soft_errs_allowed ••• ***.
Board Reset Chk for Interrupts •••••
P CaChe-diag mode bypass test mask •••• *** ••
Memory Refresh
start
end-lncr cont on err tlme seconds * •• **
Memory- Addr short s start-add end add * cont- on err pat 2 pat3
Memory-roM ••• cOnt on err *..... - start add end add add Incr cont_on_err ••••••
Memory-Eec SBEs
Memory-Byte Errors start-add end-add add-Incr cont_on_err
Memory-EcC Loglc
start-add end-add add-incr cont_on_err
Memory-Address
start-add end-add add-incr cont_on_err ••••••
Memory-Byte
start-add end-add add-lncr cont_on_err
Memory:oata
start:add end:add add:lncr cont_on_err
FPA
•••••••
SSC Prog timers
which timer wait time us •••
SSC-TOY Clock
repeat_test_2SOms_ea Tolerance
Virtual-Mode
•••••••• *
Interval Timer
•••• *
SHAC LPBCK
**** •• *.
SHAC-RESET
dssi bus port number time_secs
SGEC-LPBCK ASSIST time-secs·*SHACshac-number *** •• *.
SGEC
loopbaCk_type no_raIn_tests
SSC Console SLU
start BAUD end BAUD .*****
QOSS any
lnput-csr self test rO self test r1
CQBle memory
bypass test mask .**.*....
Qbus MsCP
!~ csr-·*··*·
Qbus-OELQA
device num addr **.*
QZA lntlpbck1
controller-number * ••••• **
QZA-Intlpbck2
controller-number ••••••• **
QZA-memory
incr test pattern controller number *******
QZA-OMA
Controller number main mem buf .*******
QZA-EXTLPBCK
controller:number .* •• CQBlC registers
CQBIc:Powerup
*.
Flush Ena Caches
dis flush virtual dIs flush backup dls_flush_prlmary
INTERACTION
pass_count disable_device ***.
Inlt memory 16MB
*
LIst-CPU registers *
~t;~~~ra:nostlcs :XPDd_err_ms g get_mode inlt_LEOs clr-p s _ cnt

Example 4-4 (continued on next page)

4-12

KA675/KA680/KA690 CPU System Maintenance

Example 4-4 (Cont.): Test 9E
9F
Cl
C2
cs
C6
DO
02
DA
DB
DC
DO
DE
OF

20060D4C
200566£0
200568B6
2005E25A
20056624
20067400
200 65AlC
200684B4
200661BO
200643£0
2006691C
20066404
20065DFO

******.*.*
Create_AO_Script
SSC RAM Data
SSC-RAM-Data Addr
SSC-regIsters
ssc-powerup
* .... *****
V Cache diaq mode bypass test mask *********
O-Bit dIaq mode
bypass-test-mask *********
pa Flush cache
********** Speed
print_speed *********
NO_Memory-present fIr*****_*
B Cache Data debuq start add end add add incr *******
S-cache-faq Debug start-add end-add add-incr *******
O:BIT_DEBUGstart:add end:add add:incr seq_incr ******

Scripts
f
Description
AO
Al
A3
A4
AS
A6
Ai

A8
A9
»>

User defined scripts
Powerup tests, Functional Verify, continue on error, numeric countdown
Functional Verify, stop on error, test f announcements
Loop on A3 Functional Verify
Address shorts test, run fastest way possible
Memory tests, mark only multiple bit errors
Memory tests
Memory acceptance tests, mark single and multi-bit errors, call A7
Memory tests, stop on error

System Initialization and Acceptance Testing (Normal Operation)

4-13

Table 4-3: Scripts Available to Customer Services
Enter with
TEST
Script! Command

AO
Al

AO
Al,O

A'j,AI:S

AE,AD

Description
Runs user-defined script. Enter T 9F to create.
Primary power-up script; builds memory bitmap; marks hard
single-bit errors and multi-bit errors. Continues on error.
Runs address shorts test from RAM; invokes tests 3F and 48;
runs test 48 the fastest way possible using fast mode and
running cached from RAM.
Memory test script; initializes memory bitmap and marks only
multiple bit errol'S.
I\'iemory teSt portion in voiteci by scrip~ A5. Ael u..IJl:I i.L.~ W~~"J:J
tests without rebuilding and reinitializing the bitmap. Run
script AS once before running script A7 separately to allow
mapping out of both single-bit and double-bit main memory
ECC errors.
Memory acceptance. Running script AS with script A7 tests
main memory more extensively. It enables hard single-bit and
multibit main memory ECC errors to be marked bad in the
bitmap. Invokes script A7 when it has completed its tests.
Memory tests. Halts and reports the first error. Does not reset
the bitmap or busmap. It is a quick way to specify which test
caused a failure when a hard error is present.
Console program. Runs memory tests, marks bitmap, resets
busmap, and resets caches. Calls script AE.
Console program. Resets :memory CSRs and resets caches. Also
called by the INn' command.
Console program. Resets busmap and resets eaches.

lScripts AD, AE, and AF exist primarily for console program; error displays and progress
messages are suppressed (not recommended for CSE use).

In most cases, the service engineer needs only the scripts shown below for
effective troubleshooting and acceptance testing.

4-14 KA675/KA680IKA690 CPU System Maintenance

Scripts
i
Description

AD
Al
A3
A4
AS
A6
A7
AS
A9

User defined scripts
Powerup tests, Functional Verify, continue on error, numeric
countdown
Functional Verify, stop on error, test i announcements
Loop on A3 Functional Verify
Address shorts test, run fastest way possible
Memory tests, mark only multiple bit errors
Memory tests
Memory acceptance tests, mark single and multi-bit errors,
call A7
Memory tests, stop on error

»>

4.4 Basic Acceptance Test Procedure
Penorm the acceptance testing procedure listed below, after installing a
system, or whenever adding or replacing the following:
CPU module
MS690 memory module
Backplane
DSSI device
H3604 console module
1. While monitoring the test display on the console terminal, run five
error·free passes of the power·up scripts by entering the following
command:
»>R T 0

If you can not monitor the console terminal during this step, use the
following command.
»>T A4

Script A4 "ill halt on an error so that the error message will not scroll
off the screen.
Press IClRLIC I to terminate the scripts. Refer to Chapter 5 if failures
occur.
2. Double·check the memory configuration, since test 31 can check for only
a few invalid configurations. For example, test 31 cannot report that a
memory board is missing from the configuration, since it has no way of
knowing if the board should be there or not.

_~ystem Initialization and Acceptance Testing (Normal Operation)

4-15

To check the memory configuration and to ensure there are no bad
pages, enter the following command line:
»>SROW MEMORY /FULL
Memory 0: 00000000 to 01FFFFFF, 32 Mbytes, 0 bad pages
Total of 32 Mbytes, 0 bad pages, 112 reserved pages
Memory Bitmap
-01FF2000 to 01FF3FFF, 16 pages
Console Scratch Area
-OlFF4000 to OlFF7FFF, 32 pages
Q-bus Map
-OlFF8000 to OlFFFFFF, 64 pages
Scan of Bad Pages
»>

Memories 0 through 3 are the MS690 memory modules. The Q22-bus
map always spans the top 32 Kbytes of good memory. The memory
bitmap always spans two pages (1 Kbyte) for each 4 Mbytes of memory
configured Each bit within the memory bit map represents a page of
memory.
Use utility 9C to examjne the contents of configuration registers
MEMCON 0-7 to verify the memory configuration:
»~

SBR-07FB8000
SLR-00002021 SAVPC-20047FS8 SAVPSL-20047FS8 BCETSTS-OOOOoooo
SCBB-200S3EOO
POBR-80000000
POLR-00100A80
P1BR-0080aooo BCETIDX-OOOOoooo
P1LR-00600000
SID-13000202
TODR-OOOOOooo
ICCS-OOoooooo BCEDSTS-00000700
ECR-OOOOOOCA MAP£N-OOOOOooo BDMTR-20084 000 BOMKR-0000007C BCEDIDX-00000010
TIRO-0112BD68 TNIRO-OOOOoooo TIVRO-00000078 BCEDECC-OOOooooo
TCRO-OOOOooos
TCRl-00000001
TIRI-0117BFA9 TNIRI-OOOOOOOF TIVRl-0000007C NEDATHI-OOOOOooo
RXCS-OOoooooo
RXDB-OOOOOOOD
TXCS-OOOooooo
TXDB-00000030 NEDATLO-OOOOOooo
SCR-OOOODOOO
DSER-OOOOOooo OBEAR-OOOOOOOF
OEAR-OOOOOooo
CESR-OOOOOooo
QBMBR-07FF8000
BDR-3CFD08AB DLEDR-OOOOOooc SSCCR-OOD5SS70 CMCDSR-0000CI08
CSEAR1-OOOOOooo CSEAR2-OOOOOoco CIOEAR1-OOOOOooo
CBTCR-00004000 IPCRO-OOOO
PCSTS-FFFFF800 PCADR-FFFFFFF8 PCCTL-FFFFFE13
ICSR-OOOOOOOl CIOEAR2-00000300
VMAR-000007EO
CNEAR-OOOOOooo
CCTL-00000007 BC£TAG-OOCOOooo
NESTS-OOOOoooo CEFSTS-00019200 N£OADR-EOOSBFD8 NEOCMO-8000FF04 NEICMO-OOOOOooo
DSSI 1-03 (BUS 1)
POBBR 1-03060022
PMCSR 1-00000000
SSHMA 1-00008A20
-PSR 1-00000000
PESR-I-OOOOOOOO
PFAR-I-OOOOOOOO
PPR-I-OOOOOOOO
DSSI 2-02 (BUS 0)
PQBBR-2-03060022
PMCSR-2-00000000
SSHMA-2-0000CA20
-PSR 2-00000000
PESR-2-00000000
PFAR-2-00000000
PPR-2-D0000000
3-00004030- 4-00004050
S-8039FFOO
6-B3EOFOOO -7-00000000
NICSRO-IFFF0003
NICSR9-04E204£2 10-00040000 11-00000000 12-00000000 13-00000000 lS-OOOOFFFF
NISA-08-00-2B-26-AS-S3
MEAR-184060l0 ADD-210l8040
MESR-00006000
MEMCON 0:3; 0-80000005, 1-84000005, 2-00000007~ 3-00000007
MMCDSR-01111000
MEMCON:4:7; 4-00000007, 5-00000007, 6-00000007, 7-00000007
HOAMR-OOOOOOOO

»>

To identify registers and register bit fields, see the KA675/ KA680
/ KA690 CPU Technical Manual.

4-16

KA675/KA680/KA690 CPU System Maintenance

Examine MEMCON 0-7 to verify the memory configuration. Each pair
of MEMCONs maps one MS690 memory module as follows:
MEMCON0-1
MEMCON2-3
MEMCON4-5
MEMCON6-7

First MS690; slot 4, closest to CPU
Second MS69O; slot 3
Third MS690; slot 2
Fourth MS690; slot 1, farthest from CPU

Verify the following:
•

The bank enable bit «31» in both MEMCONs for each memory
module is set to (8xxx xxxh), which indicates that the base address
for the hanks contained on the module is valid.

•

MEMCON bits <2:1> are the signature field and contain the
following value, in relation to the size of the array.

Table 4-4: Signature Field Values
MCSRO-15
<2:1>

Hex
EqurY'

CGnfiguration

Unassigned

RAM size 1 Mbit

RAM size 4 Mbits

Bank no response

•

MEMCON hits <28:24> indicate the base address for each memory
bank. The first valid bank starts at o. The memory subsystem can
mix the different sized memory modules (32 MB, 64 MB, and 128
MB). The largest sized memory module will be configured first, no
matter where it is in the system. A.4ter all modules of the largest
size are configured, the next iargest size wili be configured.

•

MEMCONs display 0000 0007 if no memory module is present;
there should be no gaps in the memory configuration.

3. Check the Q22-bus and the Q22-bus logic in the KA6751KA6801KA690
CQBIC chip and the configuration of the Q22-bus, as follows:

System Initialization and Acceptance Testing (Normal Operation)

4-17

»>SHOW QBOS
Scan of Q-bus I/O Space
-200000DC (760334)=0000 RQDX3/KDASO/RRDSO/RQC25/KFQSA-DISK
-200000DE (760336)=OAAO
-20001468 (772150)=0000 RQDX3/KDA50/RRD50/RQC25/KFQSA-DISK
-2000146A (772152)=OAAO
-20001920 (774440)=FF08 DESQA
-20001922 (774442)=FFOO
-20001924 (774444)=FF2B
-20001926 (774446)=FF09
-20001928 (774450)=FFA3
-2000192A (774452)=FF96
-2000192C (774454)-0050
-2000192E (774456)=1030
-20001942 (774502)=OBCO
-20001F40 (777500)=0020 IPCR
Scan of

Q-~us

Memory Space

»>

The columns are described below. The examples listed are from the last
line of the example above.
First column = the VAX. 110 address of the eSR, in hex (20001F40).
Second column =the Q22-bus address of the eSR, in octal (777500).
Third column = the data, contained at the eSR address, in hex
(0020).

Fourth column = the speculated device name (IPCR, the epu
interprocessor communications register).
Additional lines for the device are displayed if more than one eSR
exists.

The last line, Scan of Q-bus Memory Space, displays memory residing
on the Q22-bus, if present. VAX memory mapped by the Q22-bus map
is not displayed under SHOW QBUS, but is displayed using SHOW
MEMORYIFULL.
If the system contains an MSCP orTMSCP controller, run test 81. This
test performs the following functions:
Performs step one of the UQ port initialization sequence
Performs the SA wraparound test
Checks the Q22-bus interrupt logic
If you do not specify the eSR address, the test searches for and runs
on the first MSCP device by default. To test the first TMSCP device,
you must specify the first parameter:
»>'1 81 20001940

4-18

KA67SIKA680/KA690 CPU System Maintenance

You can specify other addresses if you have multiple MSCP or TMSCP
devices. This action may be useful to isolate a problem with a controller,
the CPU module, or the backplane. Use the VAX I/O address provided
by the SHOW QBUS command to determine the CSR value. If you do
not specify a value, the MSCP device at address 20001468 is tested by
default.
4. Check that all UQSSP, MSCP, TMSCP, and Ethernet controllers and
devices are visible by typing the following command line:
»>SHOW DEVICE

DSSI Bus 0 Node 0 (ALPHA)
-DIAO (RF72)
DSSI Bus 0 Node 1 (BETA)
-DIAl (RF72)
DSSI Bus 0 Node 2 (GAMMA)
-DIA2 (RF72)
DSSI Bus 0 Node S (ZETA)
-MIAS (TFSS)
DSSI Bus 0 Node 6 (*)
DSSI Bus 1 Node 7 (*)
Ethernet Adapter

-EZAO (OS-OO-2B-OS-ES-6E)
Ethernet Adapter 0 (774440)

-XQAO (OS-OO-2B-06-16-F2)

In the example, the console displays the node numbers of disk and tape
ISEs it recognizes. The line below each node name and number is the
logical device name DIAO, DIAl, DIA2, and MlA5 in this case.
The two lines marked by an asterisk (*) are for the embedded DSSI
adapters. DSSI node names and node numbers must be unique.
The next two lines show the logical name and station address for the
embedded Ethernet adapter. The last two lines refer to a DESQA
Ethernet controller, its Q22=bus CSR address, its logical name (XQAO),
and its station address.
5. Run one pass of the DSSI internal drive tests (DRVTST and DRVEXR)
using the Diagnostic Utility Protocol (DUP) driver as described in
Section 5.4.
6. If the above steps have completed successfully and you have time to
test the Q-bus ootions. load MDM (minimum release of MDM 136 is
required for VAX-4000 Model 500 systems). Run the system tests from
System Initialization and Acceptance Testing (Normal Operation)

4-19

the Main Menu. If they run successfully, the system has gone through
its basic checkout and the operating system software can be loaded.
7. Bring up the operating system.
8. Bringing up VMS completes the installation procedures. Run the VMS
User Environment Test Package (UETP) to test that VMS is correctly
installed. Refer to the VAX 3520, 3540 VMS Installation and Operations
(ZKS166) manual for instructions on running UETP.

4.5 Machine State on Power-Up
This section describes the state of the kernel after a power-up halt.
The descriptions in this section assume the system has just powered-up
and the power-up diagnostics have successfully completed. The state of
the machine is not defined if individual diagnostics are run or for any
other halts other than a power-up halt (SAVPSL<13:8>(RESTART_CODE)
= 3). Refer to Appendix E for a description of the normal state of CPU
configurable bits following completion of power-up tests.

4.6 Main Memory Layout and State
Main memory is tested and initialized by the firmware on power-up.
Figure 4-2 is a diagram of how main memory is partitioned after
diagnostics.

4-20 KA675/KA680!KA690 CPU System Maintenance

Figure 4-2: Memory Layout After Power-Up Diagnostics

o
Available system memory
(pages potentially good or bad)
PFN bitmap

OMR base
~

PFN bitmap
(always on page boundary and
size in pages n = (# of MB )/2)

n pages

Firmware ·scratch memory·
(always 16 KB)

32 pages

--1
________________________
022-Bus Scatter/Gather Map
(always on 32 KB boundary)

64 pages

~.~~i

Potential -bad· memory
Top of Memory

4.6.1 Reserved Main Memory
In order to build the scatter/gather map and the bitmap, the firmware
attempts to find a physically contiguous page-aligned 176-KB block of
memory at the highest possible address that has no multiple bit errors.
Single-bit errors are tolerated in this section.
Of the 176 KB, the upper 32 KB is dedicated to the Q22-bus scatter/gather
map~ as shown in Figure 4-2. Of the iower portion, up to 128 Kb at the
bottom of the block is allocated to the Page Frame Number (PFN) bitmap.
The size of the PFN bitmap is dependent on the extent of physical memory,
each bit in the bitmap maps one page (512 bytes) of memory. The remainder
of the block between the bitmap and scatter/gather map (minimally 16 KB)
is allocated for the firmware.
.
4.6.1.1 PFN Bitmap

The PFN bitmap is a data structure that indicates which pages in memory
are deemed usable by operating systems. The bitmap is built by the
diagnostics as a side effect of the memory tests on power-up. The bitmap
always starts on a page boundary. The bitmap requires 1 KB for every 4
MB of main memory, hence, a 8 MB system requires 2 KB, 16 MB requires
4 KB, 32 MB requires 8 KB, and a 64 MB requires 16 lUS. The bitmap

System Initialization and Acceptance Testing (Normal Operation)

4-21

does not map itself or anything above it. There may be memory above the
bitmap which has both good and bad pages.
Each bit in the PFN bitmap corresponds to a page in main memory. There
is a one to one correspondence between a page frame number (origin 0) and
a bit index in the bitmap. A one in the bitmap indicates that the page is
"good" and can be used. A zero indicates that the page is "bad" and should
not be used. By default, a page is flagged "bad", if a multiple bit error occurs
when referencing the page. Single bit errors, regardless of frequency, will
not cause a page to be flagged "bad".
The PFN bitmap is protected by a checksum stored in the NVRAM. The
checksum is a simple byte wide, two's complement checksum. The sum of
~J1 hyt.p~ in t.'hp hitm~p ~nn thp hitmAp ~hP.ek~llm ~hollld result in zero.
4.6.1.2 Scatter/Gather map

On power-up, the scatter/gather map is initialized by the firmware to map
to the first 4 MB of of main memory. Main memory pages will not be
mapped, if there is a corresponding page in Q22-bus memory, or if the
page is marked bad by the PFN bitmap.
On a processor halt other than power-up, the contents of the scatter/gather
map is undefined, and is dependent on operating system usage.
Operating systems should not move the location of the scatter/gather map,
and should access the map only on aligned longwords through the local 110
space of 20088000 to 2008FFFC, inclusive. The Q22-bus map base register,
(QMBR) is set up by the firmware to point to this area, and should not be
changed by software.
4.6.1.3 Firmware "Scratch Memory"

This section of memory is reserved for the firmware. However, it is only
used after successful execution of the memory diagnostics and initialization
of the PFN bitmap and scatter/gather map. This memory is primarily used
for diagnostic purposes.

4.6.2 Contents of Main Memory
The contents of main memory are undefined after the diagnostics have run.
Typically, nonzero test patterns will be left in memory.
The diagnostics will "scrub" all of main memory, so that no power-up
induced errors remain in the memory system. On the KA6751KA680
/KA690 memory subsystem, the state of the ECC bits and the data bits
are undefined on initial power-up. This can result in single and multiple
bit errors if the locations are read before written, because the ECC bits are
not in agreement with their corresponding data bits. An aligned 10ngword

4-22

KA67SIKA680/KA690 CPU System Maintenance

write to every location (done by diagnostics) eliminates all power-up induced
errors.

4.6.3 Memory Controller Registers
The CPU firmware assigns bank numbers to the MEMCONn registers in
ascending order, without attempting to disable physical banks that contain
errors. High order unused banks are set to zero. Error loggers should
capture the following bits from each MEMCONn register:
MEMCONn <31> (bank enable bit). As the firmware always assigns
banks in ascending order, knowing which banks are enabled is sufficient
information to derive the bank numbers. MEMCONn <1:0> (bank usage).
This field determines the size of the banks on the particular memory board.
Additional information should be captured from the NMCDSR, MOAMR,
MSER, and MEAR as needed.

4.6.4 On-Chip cache
The CPU on-chip cache is tested during the power-up diagnostics, flushed,
and then turned on. The cache is also turned on by the BOOT and the INIT
command.

4.6.5 Translation Buffer
The CPU translation buffer is tested by diagnostics on power-up, but not
used by the firmware because it runs in physical mode. The translation
buffer can be invalidated by using PR$_TBIA, IPR 57.

4.6.6 Halt-Protected Space
On the KA6751KA6801KA690 halt-protected space spans the 512-KB
FEPROM from 20040000 to 2007FFFF.
The firmware always runs in halt-protected space. When passing control
to the bootstrap, the firmware exits the halt-protected space, so ifhalts are
enabled, and the halt line is asserted, the processor will then halt before
booting.

4.7 Operating System Bootstrap
Bootstrapping is the process by which an operating system loads and
assumes control of the system. The KA6751KA6801KA690 support bootstrap
of the VAXNMS and VAXELN operating systems. Additionally, the KA675
1KA680JKA690 will boot MDM diagnostics and any user application image
which conforms to the boot formats described herein;

System Initialization and Acceptance Testing (Normal Operation)

4-23

On the KA6751KA6801KA690 a bootstrap occurs whenever a BOOT
command is issued at the console or whenever the processor halts and the
conditions specified in the Table 3-5 for automatic bootstrap are satisfied.

4.7.1 Preparing for the Bootstrap
Prior to dispatching to the primary bootstrap (VMB), the firmware
initializes the system to a known state. The initialization sequence follows:
l.

Check the console program mailbox "bootstrap in progress" bit
(CPMBX<2>(BIP». If it is set, bootstrap fails.

2. If this is an automatic bootstrap, display the message "Loading system
softwm:-e." on the console terminal.
3. Set CPMBX<2>(BIP).
4. Validate the Page Frame Number (PFN) bitmap.
checksum is invalid, then:

If PFN bitmap

a. .Perform an UNJAM.
b. Perform an INIT .
c.

Retest memory and rebuild PFN bitmap.

5. Validate the boot device name. If none exists, supply a list of available
devices and prompt user for a device. If no device is entered within 30
seconds, use EZAO.
6. Write a form of this BOOT request including the active boot flags and
boot device on the console, for example "(BOOTIR5:0 nUAO)".
7. Initialize the Q22-bus scatter/gather map.

a. Set IPCR<8>(AUX_HLT).
h. Clear IPCR<5>(LMEAE).
c.

Perform an UNJAM .

d. Perform an INIT .
e. If an arbiter, map all vacant Q22-bus pages to the corresponding
page in local memory and validate each entry if that page is "good".
f.

Set IPCR<5>(LMEAE).

8. Search for a 128 KB contiguous block of good memory as defined by the
PFN bitmap. If 128 KB can not be found, the bootstrap fails.

4-24 KA675/KA680/KA690 CPU System Maintenance

9. Initialize the general purpose registers as follows:
RO
R2

R3
R4
R5
RIO
R11
AP

SP
PC
Rl, R6, R7, RS,

Address of descriptor of boot device name; 0 if none specified
Length of PFN bitmap in bytes
Address of PFN bitmap
Time-of-day of bootstrap from PR$_TODR
Boot flags
Halt PC value
Halt PSL value (without halt code and map enable)
Halt code
Base of 128-Kbyte good memory block + 512
Base of 128-Kbyte good memory block + 512

R9,FP

10. Copy the VMB image from FEPROM to local memory beginning at the
base of the 128 KB good memory block + 512.
11. Exit from the firmware to memory resident VMB.

On entry to VMB the processor is running at IPL 31 on the interrupt stack
with memory management disabled. Also, local memory is partitioned as
shown in Figure 4-3.

SYStem Initialization and Acceptance Testing (Normal Operation)

4-25

Figure 4-3:

Memory Layout prior to VMS Entry

o
Potential -bad· memory
Base
Reserved for RPB, initial stack

Base+S12(SP,PC)

1--------------. .
VMBimage

256 pages for VMS
128 KB block of
·good· memory
, ...... "'". ... 1l", ... M,

Balance of 128 KB block
to be used for SCB, stack,
and the secondary bootstrap.

'r"-::r- -_·;:'"'--1

Unused memory
PFN bitmap
PFN ~itmap
(always on page boundary and
size in pages n = (# of MB )/2)
Firmware ·scratch memory(always 16 KB)
QMR base
Q22-Bus Scatter/Gather Map
(always on 32 KB boundary)

n pages

-1
-1

32 pages

64 pages

Potential -bad- memory
Top of Memory

4.7.2 Primary Bootstrap Procedures (VMS)
Virtual Memory Boot (VMB) is the primary bootstrap for booting VAX.
processors. On the KA6751KA6801KA690 module, VMB is resident in the
firmware and is copied into main memory before control is transferred to
it. VMB then loads the secondary bootstrap image and transfers control to
it.

4-26

KA6751KA680/KA690 CPU System Maintenance

In certain cases, such as VAXELN, VMB actually loads the operating
system directly. However, for the purpose of this discussion "secondary
bootstrap" refers to any VMB loadable image.
VMB inherits a well defined environment and is responsible for further
initialization. The following summarizes the operation of VMB.
1. Initialize a two page 8CB on the first page boundary above VMB.

2. Allocate a three page stack above the SCB.

3. Initialize the Restart Parameter Block (RPB).
4. Initialize the secondary bootstrap argument list.
5. If not a PROM boot, locate a minimum of three consecutive valid QMRs.
6. Write "2" to the diagnostic LEDs and display 112.." on the console to
huncate that VMB is searching for the device.
7. Optionally, solicit from the console a "Bootfile: tt name.
8. Write the name of the bOot device from which VMB will attempt to boot
on the console, for example, "_DUAOtl.

9. Copy the secondary bootstrap from the boot device into local memory
above the stack. If this fails, the bootstrap fails.
10. Write "1'" to the diagnostic LEDs and display "1.." on the console to
indicate that VMB has found the secondary bootstrap image on the
boot device and has loaded the image into local memory.
11. Clear CPMBX<2>(BIP) and CPMBX<3>(RIP).

12. Write "0" to the diagnostic LEDs and display "0.." on the console to
indicate that VMB is now transferring control to the loaded image.
13. Tr~wo:lsfer contrel to the loaded image v.ith the follovr.Jlg register usage.
R5
RIO

R11
AP

Transfer address in secondary bootstrap image
Base adcL~ss of secondary boots+...,p memo1"Y
Base address of RPB
Base address of secondary boot parameter block
Base address of secondary boot parameter block

If the bootstrap operation fails, VMB relinquishes control to the console by
halting with a HALT instruction. VMB makes no assumptions about the
location of Q22-bus memory. However, VMB searches through the Q22-bus
Map Registers (QMRs) for the first QMR marked "valid", VMB requires
minimally 3 and maximally 129 contiguous "valid" maps to complete a
bootstrap operation. If the search exhausts all map registers or there are
fewer than ~ the required number of Itvalid maps, a bootstrap cannot be
lt

System Initialization and Acceptance Testing (Normal Operation)

4-27

performed. It is recommended that a suitable block of Q22-bus memory
address space be available (unmapped to other devices) for proper operation.
After a successful bootstrap operation, control is passed to the secondary
bootstrap image with the memory layout as shown in Figure 4-4.

4-28

KA6751KA6801KAS90 CPU System Maintenance

Figure 4-4: Memory Layout at VMB Exit

o
Potential -bad- memory
Base
Reserved for RPB. initial stack
Base+512(SP ,PC)
VMBimage
Next page
SCB (2 pages)
Nextpage+1024
Stack (3 pages)
Next page+2560
I

Secondary bootstrap image
(potentially exceeds block)

-----------------

256 pages for VMS
128 KB block of
-good- memory
(page aligned)

.-J

Unused memory
PFN bitmap
PFN bitmap
(always on page boundary and
size in pages n = (# of MB )/2)
Firmware ·scratch memory(always 16 KB)

OMRbase

n pages

----1

32 pages

I
·
64~9~
.----------------11 . . .
022-Bus Scatter/Gather Map
(always on 32 KB boundary)

1----"

Potential ·bad- memory
Top of Memory

In the event that an operating system has an extraordinarily large
secondary bootstrap which overflows the 128 KB of "good" memory, VMB
loads the remainder of the image in memory above the "good" block.

System Initialization and Acceptance Testing (Normal Operation)

4-29

However, if there are not enough contiguous "good" pages above the block
to load the remainder of the image, the bootstrap fails.

4.7.3 Device Dependent Secondary Bootstrap Procedures
The following sections describe the various device dependent boot
procedures.
4.7.3.1 Disk and Tape Bootstrap Procedure

The disk and tape bootstrap supports Files-11 lookup (supporting only the
ODS level 2 file structure) or the boot block mechanism (used in PROM
boot also). Of the standard DEC operating systems VMS and ELN use the
Files-11 bootstrap procedure and Ultrix-32 uses the boot block mechanism.

VMB first attempts a Files-11 lookup, unless the RPB$V_BBLOCK boot
flag is set. If VMB determines that the designated boot disk is a Files-11
volume, it searches the volume for the designated boot program, usually
[SYSO.SYSEXE]SYSBOOT.EXE. However, VMB can request a diagnostic
image or prompt the user for an alternate file specification. If the boot
image can't be found, VMB fails.
If the volume is not a Files-11 volume or the RPB$V_BBLOCK boot flag
was set, the boot block mechanism proceeds as follows:

1. Read logical block 0 of the selected boot device (this is the boot block).
2. Validate that the contents of the boot block conform to the boot block
format (see below).
3. Use the boot block to find and read in the secondary bootstrap.
4. Transfer control to the secondary bootstrap image, just as for. a Files-11
boot.

The format of the boot block must confonn to that shown in Figure 4-5.

4-30 KA675/KA680/KA690 CPU System Maintenance

Figure 4-5: Boot Block Format
24 23

BB-O:

16 15

any value

low LBN

high LBN

(The next segment is also used as a PROM ·signature block. -)

CHK

18 (Hex)

any value, most likely 0
size in blocks of the image
load offset
offset into image to start
sum of the previous three longwords
Where:

1) the 18 (hex) indicates this is a VAX instruction set
2) 18 (hex) + ak· = the one's complement if ·CHKML~7

4.7.3.2 PROM Bootstrap Procedure

The PROM bootstrap uses a variant of the boot block mechanism. v~m
searches for a valid PROM "signature block" , the second segment of the
boot block defined in Figure 4-5. IfPRAO is the selected "device then VMB
searches through Q22-bus memory on 16 KB boundaries. If the selected
"device" is PRBO, VMB checks the top 4096 byte block of the FEPROM.
lt

At each boundary, VMB :
1. Validates the readability of that Q22-bus memory page.

2. If readable, check to see if it contains a valid PROM signature block.

mam memory
and VMB will transfer control to that image at the offset specified in the
PROM bootbiock. If not, the next page win be tested.

If verification passes, the PROM image will be copied into

System Initialization and Acceptance Testing (Normal Operation)

4-31

Note that it is not necessary that the boot image actually reside in PROM.
Any boot image in Q22-bus memory space with a valid signature block on
a 16 KB boundary is a candidate. Indeed, auxiliary bootstrap assumes that
the image is in shared memory.
The PROM image is copied into main memory in 127 page "chunks" until
the entire PROM is moved. All destination pages beyond the primary 128
KB block are verified to make sure they are marked good in the PFN bitmap.
The PROM must be copied contiguously and if all required pages cannot fit
into the memory immediately following the VMB image, the boot fails.
4.7.3.3 MOP Ethernet Functions and Network Bootstrap Procedure

Whenever a network bootstrap is selected on KA6751KA6801KA690, the
\'1vIB code makes continuous ai.i:.empi.s l.o booi. from. t.he llei.wo.rk. \~m use::)
the DNA Maintenance Operations Protocol (MOP) as the transport protocol
for network bootstraps and other network operations. Once a network boot
has been invoked, VMB turns on the designated network link and repeats
load attempt, until either a successful boot occurs, a fatal controller error
occurs, or VMB is halted from the operator console.
The KA6751KA6801KA690 support the load of a standard operating system,
a diagnostic image, or a user-designated program via network bootstraps.
The default image is the standard operating system, however, a user may
select an alternate image by setting either the RPB$V_DIAG bit or in the
RPB$V_BOLICT bit in the boot ilag longword RS. Note that the RPB$V_
SOLICT bit has precedence over the RPB$V_DIAG bit. Hence, ifboth bits
are set, then the solicited file is requested.
NOTE: VMB accepts a maximum 39 characters for a file specification for
solicited boots. However, MOP V3 only supports a 16-character file name. If
the network server is running VMS, the following defaults apply to the file
specification: the directory MOM$LOAD:, and the extension .SYS. Therefore,
the file specification need only consist of the filename if the default directory
and extension attributes are used.

The KA6751KA6801KA.690 VMB uses the MOP program load sequence
for bootstrapping the module and the MOP "dumplload" protocol type for
load related message exchanges. The types of MOP message used in the
exchange are listed in Table 4-5 and Table 4-6.
VMB, the requester, starts by sending a RE~PROGRAM message to the
MOP 'dumplload' multicast address. It then waits for a response in the
fonn of a VOLUNTEER message from another node on the network, the
MOP server. If a response is received, then the destination address is
changed from the multicast address to the node address of the server and

4-32

KA675/KA680/KA690 CPU System Maintenance

the same RE(LPROGRAM message is retransmitted to the server as an
Acknowledge.

Next, VMB begins sending RE(LMEM_LOAD messages to the server. The
server responds with either:
•

MEM_LOAD message, while there is still more to load.

•

MEM_LOAD_w_XFER, nit is the end of the image.

•

PARAM_LOAD_w_XFER, if it is·.the end of the image and operating
system parameters are required.

The ''load number" field in the load messages is used to synchronize the load
sequence. At the beginning of the exchange, both the requester and server
initialize the load number. The requester only increments the load number
if a load packet has been successfully received and loaded. This forms the
Acknowledge to each exchange. The server will resend a packet with a
specific load number, until it sees the load number incremented. The final
Acknowledge is sent by the requester and has a load number equivalent to
the load number of the appropriate LOAD_w_XFER message + l.
Because the request for load assistance is a MOP "must transact" operation,
the network bootstrap continues indefinitely until a volunteer is found.
The RE(LPROGRAM message is sent out in bursts of eight at four second
intervals, the first four in MOP Version four IEEE 802.3 format and the
last four in MOP Version 3 Ethernet format. The backoff period between
bursts doubles each cycle from an initial value of four seconds, to eight
seconds,... up to a maximum of five minutes. However, to reduce the
likelihood of many nodes posting requests in lock-step, a random "jitter"
is applied to the backoff period. The actual backoff time is computed as
(.75+('5*RND(x)))*BACKOFF, where O<=x<1.
4.7.3.4 Network "Listening"
Vlhile the CPU module is waiting for a load volunteer du..?"i..,ng bootstrap,

it ''listens'' on the network for other maintenance messages directed to the
node and periodically identifies itself at the end of each 8- to 12-minute
interval before a bootstrap retry. In particular, this '1istener" supplements
the Maintenance Operation Protocol (MOP) functions of the VMB load
requester typically found in bootstrap firmware and supports.
•

A remote console server that generates COUNTERS messages in
response to RE(LCOUNTERS messages, unsolicited SYSTEM_ID
messages every 8 to 12 minutes, and solicited SYSTEM_ID messages
in response to REQUEST_ID messages, as well as recognition of BOOT
messages.

System Initialization and Acceptance Testing (Normal Operation)

4-33

•

A loopback server that responds to Ethernet loopback messages by
echoing the message to the requester.

•

An IEEE 802.2 responder that replies to both XID and TEST messages.

During network bootstrap operation, the KA6751KA6801KA690 complies
with the requirements defined in the I'NI Node Architecture Specification"
for a primitive node. The firmware listens only to MOP "LoadlDump",
MOP "Remote Console", Ethernet "Loopback Assistance", and IEEE 802.3
. XID/I'EST messages (listed in Table 4-7) directed to the Ethernet physical
address of the node. All other Ethernet protocols are filtered by the network
device driver.
The MOP functions and message types, which are supported by the KA675
iKAb8UlKA6~O, are summarized in :rabies 4-6 ana 4-/.

4-34 KA67S/KA680/KA690 CPU System Maintenance

Table 4-5: Network Maintenance Operations Summary
Function

Role

Transmit

Receive

MOP Ethernet and IEEE 802.3 Messages l
Requester

Dump

Server
Requester

Load

RE(LPROGRAM2 to solicit

VOLUNTEER

REQ...ME~LOAD to solicit & ACK

MEM_LOAD

MEM_LOAD_w..XFER

P~LOAD_w_XFER

COUNTERS

in response to

RE(LCOUNTERS

SYSTEM_ID3

in response to

REQUEST_ID

Server

Console

Requester
Server

BOOT

Loopback

Requester
Server

LOOPED.J)ATA4 in response to

LOOP_DATA

lAll unsolicited messages are sent in Ethernet (MOP V3) and IEEE 802.2 (MOP V4), until
the MOP version of the server is known. All solicited messages are sent in the format used
for the reqt..lest.
2The initial RE(LPROGRAM message is sent to the dumpload multicast address. If an
assistance VOLUNTEER message is received, then the responders address is used as the
destination to repeat the RE'LPROGRA.'\! message and for all S'.lbseq'.lent REQ..ME..'~LLOAD
messages.
3SYSTEM_ID messages are sent out every 8 to 12 minutes to the remote console multicast
address and, on receipt of a REQUEST_ID message, they are sent to the initiator.
4LOOPED_DATA messages are sent out in response to LOOP_DATA messages. These
messages are actually in Ethernet LOOP TEST format, not in MOP format, and when sent in
Ethernet frames, omit the additional length field (padding is disabled).

System Initialization and Acceptance Testing (Normal Operation)

4-35

Table 4-5 (Cont.): Network Maintenance Operations Summary
Function

Role

Receive

Transmit
IEEE 802.3 MessagesS

ExchangeID Requester

Server

Test

XID_RSP

in response to

XID_CMD

TEST_RSP

in response to

TEST_CMD

Requester
Server

SIEEE 802.2 support ofXID and TEST is limited to Class 1 operations.

Table 4-6:

Supported MOP Messages

Message Type

Message Fields
DUMPILOAD

IlEM..LOAD_w,..XFER Code

Load. •

Load addr

Imqedata

Xferaddr

...........

NOlle

a .........

Code
02

Load •

Load addr

...........

Imapdata
dd-_.

Code

Derice
2SLQA
49SGEC

Format

OlV3

swm 3

04V4

O2S,.

C-l1 1
C-128 2
HCn]
>00 Leu
00 Nom

FFOS
FEMaint

1MOP V3.0 only.

2Mop x4.0 only.
3Soft;ware ID field is load from the string stored in the 40 byte field, RPB$T.J'n..E, of the RPB
on a solicited boot.

4-36 KA675/KA680/KA690 CPU System Maintenance

Table 4-6 (Cont.):

Supported MOP Messages

Message Type

Message Fields
DUMPJLOAD

VOWNTEER

Load'

Prmtyp

Prm len
1·16

02
03

1·16

1-06

05
06
00 End

OA
08

1-06

Prm val
Target name 1
Target addr 1

Xferaddr
aa-aa-aa-aa

Host name/
Hostaddr
Hoattime 12
Hoattime

Code
03

REMOTE CONSOLE
REQUEST_ID

SYSTDcm

Code
05

Code

Bsrvd

Becpt,

Im-nD

Rsrvd

Recpt'
Im-nn

or
00-00

~COUNTERS

Code
09

Becpt,
nn-nn

COUNTERS

Code

Becpt.
nn·nn

Info type
01-00 Ven:ion
02-00 FwlctioDS
07-00 HW addr
64-00 Derice
90-01 DataliDk
91-01 BWr me

Info len
03
02
06
01
01
02

IDf'oYalue
()4..00.()()

QO..S9
ee ee ee ee ee !Ie
25 or 49
01
Q6.04

Counter block

1MOP VS.O only.
2MOP x4.0 only.

System Initialization and Acceptance Testing (Normal Operation)

4-37

Table 4-6 (Cont.):

Supported MOP Messages

Message Type

Message Fields

REMOTE CONSOLE
BOOT 4

Code
06

VerificatioD

.,.,,-.,.,,-TV·.,.,-.,.,·.,.,,·.,.,,·

Procesr

00 Sys

Control
xx

DeY m

sw m 3

C·17

Script m 2
C·l28

(see

.,."

REQJ'ROGRAM)

LOOPBACK
LOOP..DATA

LOOPEDJ)ATA

Skpcnt.

Skpcnt

Skipped bytes

Funct.ion

bb-_

QO..02 Forward data

Skipped bytes

Function

bb-_.

00-01 Reply

Fortrard addr
ee ee ee ee ee ee

Data
dd.-•••

Data
dd.••••

IID-DD

IEEE 802.2
XID_ClIDiRSP

Form
81

Claaa
01

!be window size (K)
00

Optioaal clata.

2MOP x4.0 only.
3So~ware m field is load from the string stored in the 40 byte field, RPB$T....FILE, of the RPB
on a solicited boot.
4.A BOOT message is not verified, because in this context, a boot is already in progress.
However, a received BOOT message will cause the boot backofT timer to be reset to it's
minimum value.

Table 4-7: MOP Multicast Addresses and Protocol Specifiers
Function

Address

Protocol - Owner

DumJlt'Load

AB-CJO.OO-Ol-OO-OO

~2B

60-01

Dilital

Remote CoDlOle

AB-CJO.QO..02..QO.OO

~2B

60-02

Dilital

Loopback AImt.aDce

CF~oo2

08-00-2B

90-00

Dilital

IMOP V4.0 only.
2Not used.

4-38 KA675/KA680/KA690 CPU System Maintenance

4.8 Operating System Restart
An operating' system restart is the process of bringing up the operating
system from a known initialization state following a processor halt. This
procedure is often called restart or warmstart, and should not be confused
with a processor restart which results in firmware entry.

On the KA6751KA6801KA690 a restart occurs, if the conditions specified in
Table 3-5 are satisfied.
To restart a halted operating system, the firmware searches system memory
for the Restart Parameter Block (RPB), a data structure constructed for
this purpose by VMB. (Refer to Table D-2 in Appendix D for a detailed
description of this data structure.) If a valid RPB is found, the firmware
passes control to the operating system at an address specified in the RPB.

The firmware keeps a "restart in progress" (RIP) flag in CPMBX which it
uses to avoid repeated attempts to restart a failing operating system. An
additional "restart in progress" flag is maintained by the operating system
in the RPB.
The firmware uses the following algorithm to restart the operating system:
1. Check CPMBX<3>(RIP). If it is set, restart fails.
2. Print the message "Restarting system software." on the console
terminal.

3. Set CPMBX<3>(RIP).
4. Search for a valid RPB. If none is found, restart fails.

5. CheCk the operating system RPB$L_RSTRTFLG<O>(RIP) flag. If it is
set, restart fails.
6. Write "0" on the diagnostic LEDs.
7. Dispatch to the restart address, RPB$L_RESTART, with :
Physical address of the RPB plus 512
SP
AP

Halt mele

PSL

041FOOOO

PR$.-MAPEN

If the restart is successful, the operating system must clear CPMBX<3>(RIP).

If restart fails, the firmware prints "Restart failure." on the system console.

System Initialization and Acceptance Testing (Normal Operation)

4-39

4.8.1 Locating the RPB
The RPB is a page-aligned control block which can be identified by the first
three longwords. The format of the RPB "signature" is shown below: (Refer
to Table D-2 in Appendix D for a complete description of the RPB.)
Figure 4-6:

RPB: +00

Locating the Restan Parameter Block

physical address of the RPB
physical address of the restart routine

+08

checksum of first 31 longwords of restart routine
ML~

The firmware uses the following algorithm to find a valid RPB:
1. Search for a page of memory that contains its address in the first
longword. If none is found, the search for a valid RPB has failed.

2. Read the second longword in the page (the physical address of the
restart routine). If it is not a valid physical address, or if it is zero,
return to step 1. The check for zero is necessary to ensure that a page
of zeros does not pass the test for a valid RPB.
3. Calculate the 32 bit twos-complement sum (ignoring overflows) of the
first 31 longwords of the restart routine. If the sum does not match the
third longword of the RPB, return to step 1.
4. A valid RPB has been found.

4-40 KA675/KA680/KA690 CPU System Maintenance

Chapter 5

System. Troubleshooting and
Diagnostics
This chapter provides troubleshooting information for the two primary
diagnostic methods: online, interpreting error logs to isolate the FRU; and
offline, interpreting ROM-based diagnostic messages to isolate the FRU.
In addition, the chapter provides information on testing DSSI storage
devices, using MOP Ethernet functions to isolate errors, and interpreting
UETP failures.

The chapter concludes with a section on running loopback tests to test the
console port, embedded Ethernet ports, Embedded DSSI busses, and Q-bus
modules.

5.1 Basic Troubleshooting Flow
Before troubleshooting any system problem, check the site maintenance
log for the system's service history. Be sure to ask the system manager the
following questions:
•

Has the system been used before and did it work correctly?

•

Have changes (changes to hardware, updates to firmware or software)
been made to the system recently?

•

What is the state of the system-is it online or offiine?

If the system is omine and you are not able to bring it up, use the ofiline
diagnostic tools, such as RBDs, MDM, and LEDs.
If the system is online, use the online diagnostic tools, such as error
logs, crash dumps, UETp, and other log files.
Four common problems occur when you make a change to the system:
1. Incorrect cabling

2. Module configuration errors (incorrect CSR addresses and interrupt
vectors)

System Troubleshooting and Diagnostics

5-1

3. Incorrect grant continuity
4.

Incorrect bus node ID plugs

In addition, check the following:
•

If you have received error notification using VAXsimPLUS, check the
mail messages and error logs as described in Section 5.2.

•

If the operating system fails to boot (or appears to fai1), check. the
console terminal screen for an error message. If the terminal displays
an error message, see Section 5.3.

•

Check the LEDs on the device you suspect is bad. If no errors are
indicated by the device LEDs. run the ROM-based dirumostics described
in this c h a p t e r . '
-

•

If the system boots successfully, but a device seems to fail or an intermittent failure occurs, check the error log ([SYSERR]ERRLOG.SYS) as
described in Section 5.2.

•

For fatal errors, check that the crash dump file exists for further
analysis ([SYSEXE]SYSDUMP.DMP).

•

Check other log files, such as OPERATOR.LOG, OPCOM.LOG,
SETH OST. LOG, etc. Many of these can be found in the [SYSMGR]
account. SETHOST.LOG is useful in comparing the console output with
event logs and crash dumps in order to see what the system was doing
at the time of the error.
Use the following command to create SETHOST.LOG files, then log into
the system account.
S sri' HOST/LOG 0

After logging out this file will reside in the [SYSMGR] account.
If the system is failing in the boot or start-up phase, it may be useful
to include the command SET VERIFY in the front of various start-up
.COM files to obtain a trace of the start-up commands and procedures.

When troubleshooting, note the status of cables and connectors before you
perform each step. Label cables before you disconnect them. This step
saves you time and prevents you from introducing new problems.
Most communications modules use floating CSR addresses and interrupt
vectors. If you remove a module from the system, you may have to change
the addresses and vectors of other modules.
If you change the system configuration, run the CONFIGURE utility at the
console YO prompt (»» to determine the eSR addresses and interrupt

5-2

KA675/KA680/KA690 CPU System Maintenance

vectors recommended by Digital. These recommended values simplify
the use of the MDM diagnostic package and are compatible with VMS
device drivers. You can select nonstandard addresses, but they require
a special setup for use with VMS drivers and MDM. See the Micro VAX
Diagnostic Monitor User's Guide for information about the CONNECT and
IGNORE commands, which are used to set up MDM for testing nonstandard
configurations.
In addition, see Table 5-1 and Table 5-2 for possible problems and power
supply status indicators.

Table 5-1: Console Terminal/Console Module Problems
Problem

First Steps

No Console Message

Check the Power switch on both the console terminal and the
system. If the terminal has a DC OK LED, be sure it is lit.
Check the cabling to the console terminal.
Check the terminal setup.
Check the power supply status indicators. See Table 5-2.
Check fuse F2 on the console model. See Section 5.7.

H3604 Display Off

Check the CPU module LEDs and the H3604 cabling.

H3604 Displays Error

See Table 5-9 to determine error status.

Table 5-2: Power Supply Status Indicators
AC
Present

DC OK

1emp

Fan
Failure

Probable Cause

Off

System not plugged in. Be source not
present, or system circuit breaker tripped.

Off

Ovel'CUITent or overwltage protection
circuits activated.

Off

Excessive ambient temp; air vents blocked

Off

OtT

Failure of one or both system fans

Off

OtT

Normal operation

Over

System Troubleshooting and Diagnostics

5-3

5.2 Product Fault Management and
Symptom-Directed Diagnosis
This section describes how errors are handled by the microcode and
software, how the errors are logged, and how, through the SymptomDirected Diagnosis (SDD) tool, VAXsimPLUS, errors are brought to the
attention of the user. This section also provides the service theory used to
interpret error logs to isolate the FRU. Interpreting error logs to isolate the
FRU is the primary method of diagnosis.

5.2.1 General Exception and Interrupt Handling
This section describes the first step of error notification: the errors are
first handled by the microcode and then are dispatched to the "'!vIS error
handler.
The kernel uses the NVAX core chipset: NVAX CPU, NVAX. Memory
Controller (NMC), and NDAL to CDAL adapter (NCA).
Internal errors within the NVAX CPU result in machine check exceptions,
through System Control Block (SCB) vector 004, or soft error interrupts at
Interrupt Priority Level (IPL) 1A, SCB vector 054 hex.
External errors to the NVAX CPU, which are detected by the NMC or NDAL
to CDAL adapter (NCA), usually result in these chips posting an error
condition to the NVAX CPU. The NVAX CPU will then generate a machine
check exception through SCB vector 004, hard error interrupt, IPL ID,
through SCB vector 060 (hex), or a soft error interrupt through SCB vector
054.
External errors to the NMC and NCA, which are detected by chips on the
CDAL busses for transactions which originated by the NVAX CPU, are
typically signaled back to the NCA adapter. The NCA adapter will post
an error signal back to the NVAX CPU which generates a machine check
or high level interrupt.
In the case of Direct Memory Access (DMA) transactions where the NCA or
NMC detects the error, the errors are typically signaled back to the CDAL-

Bus device, but not posted to the NVAX CPU. In these cases the CDALBus device typically posts a device level interrupt to the NVAX. CPU via
the NCA. In almost all cases, error state is latched by the NMC and NCA
Although these errors won't result in a machine check exception or high
level interrupt (i.e. results in device level IPL 14-17 versus error level IPL
l.A, lD), the VMS machine check handler has a polling routine which will
search for this state at one second intervals. This will result in the host
logging a polled error entry.

5-4 KA675/KA6801KA690 CPU System Maintenance

These conditions cover all of the cases which will eventually be handled
by the VMS error handler. The VMS error handler will generate entries
that correspond to the machine check exception, hard or soft error interrupt
type, or polled error. .

5.2.2 VMS Error Handling
Upon detection of a machine check exception, hard error interrupt, soft
error interrupt or polled error, VMS will perform the following actions:
•

Snapshot the state of the kernel.

•

In most entry points, disable the caches.

•

If it is a machine check and if the machine check is recoverable,
determine if instruction retry is possible.
Instruction retry is possible if one of the following conditions is true:
1. If PCSTS <10>PTE_ER = 0:

Check that (ISTATE2 <07>VR = 1) or (PSL <27> FPD = 1)
Otherwise crash the system or process depending on PSL <25:24>
Current Mode.
2. If PCSTS <10>PTE_ER = 1:
Check that (ISTATE2 <07>VR = 1) and (PSL <27>FPD = 0) and
(PCSTS <09>P1'E_ER_WR =0)
Otherwise crash the system.
ISTATE2 is a longword in the machine check stack frame at offset
(SP)+24; PSL is a longword in the machine check stack frame at offset
(SP)+32; VR is the VAX Restart flag; and FPD is the First Part Done
flag.
•

Check to see if the threshold has been exceeded for various errors
(typically the L.~reshold is exceeded if 3 errors occur within a 10 '""'j"ute
interval).

•

If the threshold has been exceeded for a particular type of cache error,
mark a flag that will signify that this resource is to be disabled (the
cache will be disabled in most, but not all, cases).

•

Update the SYSTAT software register with results of errorlfault
handling.

•

For memory uncorrectable Error Correction Code (ECC) errors:

If machine check, mark page bad and attempt to replace p~e.
System Troubleshooting and Diagnostics

5-5

•

Fill in MEMCON software register with memory configuration and
error status for use in FRU isolation.

For memory single-bit correctable ECC errors:
Fill in Corrected Read Data (CRD) entry FOOTPRINT with set,
bank, and syndrome information for use in FRU isolation.
Update the CRD entry for time, address range, and count;
fill the MEMCON software register with memory configuration
information.
Scrub memory location for first occurrence of error within a
particular footprint. If second or more occurrence within a footprint,
mark page bad in hopes that page win be replaced later. Disable
soft error logging for 10 minutes if threshold is exceeded.
Signify that CRD buffer be logged for the following events: system
shutdown (operator shutdown or crash), hard single-cell address
within footprint, multiple addresses within footprint, memory
uncorrectable ECC error, or CRD buffer full.

•

For ownership memory correctable ECC error, scrub location.

•
•

Log error.
Crash process or system, dependent upon PSL (Current Mode) with a
fatal bugcheck for the following situations:
Retry is not possible.
Memory page could not be replaced for uncorrectable ECC memory
error.
Uncorrectable tag store ECC errors present in writeback cache.
Uncorrectable data store ECC errors present in writeback cache for
locations marked as OWNED.
Most INT60 errors.
Threshold is exceeded (except for cache errors).

A few other errors of the sort considered nonrecoverable are
present.
•

Disable cache(s) permanently if error threshold is exceeded.

•

Flush and re-enable those caches which have been marked as good.

•

Clear the error flags.

5-6 KA675/KA680/KA690 CPU System Maintenance

•

Perfonn Return from Exception or Interrupt (REI) to recover and
restart or continue the instruction stream for the following situations:
Most INT54 errors.
Those INT60 and INT54 errors which result in bad ECC written to a
memory location. (These errors can provide clues that the problem
is not memory related.)
Machine check conditions where instruction retry is possible.
Memory uncorrectable ECC error where page replacement is
possible and instruction retry is possible.
Threshold exceeded (for cache errors only).
Return from Subroutine (RSB) and return from all polled errors.

NOTE: The results of the VMS error handler may be preserved within the
operating system session (for example disabling a cache) but not across
reboots.
7

A.lthough the system can .recover with cache disabled7 the system
performance will be degraded, since access time increases as available cache
decreases.

5.2.3 VMS Error Logging and Event Log Entry Format
The VMS error handler for the kernel can generate six different entry
types, as shown in Table 5-3. All error entry types with the exception
of correctable ECC memory errors, are logged immediately.

System Troubleshooting and Diagnostics

5-7

Table 5-3: VMS Error Handler Entry Types
VMS Entry Type

Code

Description

EMB$C~C

(002.)

Machine Check Exception

SCB Vector 4, IPL 1F
EMB$C_SE

(006.)

Soft Error Interrupt
Correctable ECC Memoty Error

SCB Vector 54, IPL 1A
EMB$C_INT54

(026.)

Soft Error Interrupt

SCB Vector 54, IPL 1A
EMB$C_INT60

(027.)

Hard Error Interrupt 60
seB Vector 60, IPL lD

EMBSC..,.POLLED

(044.)

Polled Errors
No exception or intenu pt generated by
hardware.

EMB$C_BUGCHECK

Fatal bugcheck
Bugcheck 'tYPes:
MACHINECHK

ASYNCWRTER
BADMCKCOD
INCONSTATE
UNXlNTEXC

Each entry consists of a VMS header, a packet header, and one or more
subpackets (Figure 5-1). Entries can be of variable length based on the
number of subpackets within the entry. The FLAGS software register in
the packet header shows which subpackets are included within a given
entTy.
Refer to Section 5.2.4 for actual examples of the error and event logs
described throughout this section.

!HI KA675/KA6801KA690 CPU System Maintenance

Figure 5-1: Event Log Entry Format
00

VMS Header
Packet Revision
Packet
Header

SYSTAT
Subpacket Valid Flags
Subpacket 1

Subpacketn
MLO-OO7263

Machine check exception entries contain, at a minimum, a Machine Check
Stack Frame subpacket (Figure 5-2).

System Troubleshooting and Diagnostics

5-9

Figure &-2: Machine Check Stack Frame Subpacket
31

2423

16 15

0807

00000018 (hex) byte count (not including this longword, PC or PSL)
AST
LVL

xxxxxx

Machine
Check Code

xxxxxxxx

4. ISTATE 1

INT. SYS register

SAVEPC register

12.

VA register

16.

a register

20.

MoRN

CPUJD

Opcode

xx xx xx xx

xxxxxxxx

24.ISTATE2

28.

PSL

32.
MLO-OO7264

INT54, INT60, Polled, and some Machine Check entries contain a processor
Register subpacket (Figure 5-3), which consists of some 40 plus hardware
registers.

5-10

KA67SIKA680/KA690 CPU System Maintenance

Figure 5-3: Processor Register Subpacket
00

BPCR

(IPRD4)

MMEAOR

(IPRE8)

PAMOOE

(IPRE7)

MMEPTE

(IPRE9)

VMAR

(IPR 00)

96.

TBAOR

(IPREC)

100.

MMESTS

(IPREA)

12.

PCADR

(IPRF2)

104.

92.

PCSCR

(IPR7C)

16.

BCEDIOX

(IPRA7)

108.

ICSR

(IPRo3)

20.

BCEDECC

(IPRA8)

112.

ECR

(IPR 70)

24.

BCETIOX

(IPRA4)

116.

TBSTS

(IPR ED)

2S.

BCETAG

(IPRA5)

120.

PCCTL

(IPRF8)

32.

MEAR

(2101.8040)

124.

PCSTS

(IPR F4)

36.

MOAMR

(2101.804C)

128.

CCTL

(IPRAO)

40.

CSEAR1

(2102.0008)

132.

BCEoSTS

(IPRA6)

44.

CSEAR2

(2102.000c)

136.

BCETSTS

(IPRA3)

48.

CIOEAR1

(2102.0010)

140.

MESR

(2101.8044)

52.

CIOEAR2

(2102.0014)

144.

MMCoSR

(2101.8048)

56.

CNEAR

(2102.0018)

148.

CESR

(2102.0000)

60.

CEFDAR

(IPRAB)

152.

CMCoSR

(2102.0004)

64.

NEOADR

(IPR BO)

156.

CEFSTS

(IPRAC)

68.

NEDATHI

(IPR 84)

160.

NESTS

(IPRAE)

72.

NEDATLO

(IPR 86)

164.

NEOCMO

(IPRB2)

76.

QBEAR

(2008.0008)

168.

NEICMO

(IPRBa)

SO.

DEAR

(2008.000c)

172.

....

, ..1\1\

_ _ _ _IJ_S_E_n_,_CC8_·C004
_ _,_ _..1B488A ••
CBreR (2014.0020)
•
...

I.
on ,_000
I~
is:4..n\, I, 176.
, .PC ....
...

NOTE: The byte count, although part of the stack frame, is not included in
the error log entry itself.

Bugcheck entries generated by the VMS kernel error handler include the
first 23 registers from the processor Register subpacket along with the Time
of Day Register (TODR) and other software context state.
Un correctable ECC memory error entries include a Memory subpacket
(Figure 5-4). The memory subpacket consists of MEMCON, which is a

System Troubleshooting and Diagnostics

5-11

software register containing the memory configuration and error status
used for FRU isolation, and MEMCONn, the hardware register that
matched the error address in MEAR.
Figure 5-4: Memory Subpacket for ECC Memory Errors
31

00
MEMCON

MEMCONn (one longword from 2101.8000 - 2101.801C)

4.
M~7266

Correctable Memory Error entries have a Memory (Single-Bit Error) SBE
Reduction subpacket (Figure 5-5). This subpacket, unlike all others, is
of variable length. It consists solely of software registers from state
maintained by the error handler, as well as hardware state transformed
into a more usable format.
Figure 5-5: Memory SBE Reduction Subpacket (Correctable Memory
Errors)
31

Memory SSE Reduction Subpacket

CRD Entry Subpacket Header
CRD Entry #1
CRD Entry #2

CRD Entry n
Max n = 16
MLOOO7267

The VMS error handler maintains a Correctable Read Data (CRD) buffer
internally within memory that is flushed asynchronously for high-level
events to the error log file. The CRD buffer and resultant error log entry
are maintained and organized as follows.

5-12

KA6751KA680/KA690 CPU System Maintenance

•

Each entry has a subpacket header (Figure 5-6) consisting of LOGGING
REASON, PAGE MAPour CNT, MEMCON,. VALID ENTRY CNT,
and CURRENT ENTRY. MEMCON contains memory configuration .
information, but no error status as is done for the Memory subpacket.

Figure 5-6: eRD Entry Subpacket Header
00

31
Logging Reason

Page Mapout CNT

MEMCON

Valid Entry CNT

12.

Current Entry

16.
MLO-OO7268

•

Following the subpacket header are 1 to 16 fixed-length Memory CRD
Entries (Figure 5-7). The number of Memory eRD entries is shown in
VALID ENTRY CNT. The entry which caused the report to be generated
is in CURRENT ENTRY

System Troubleshooting and Diagnostics

5-13

Figure 5-7: Correctable Read Data (CRD) Entry
00

31
Footprint

Status

CRO CNT

Pages Marked Bad CNT

12.

First Event

16.

Last Event

24.

Lowest Address

32.

Highest Address

36.

Each Memory CRD Entry represents one unique DRAM within the
memory subsystem. A unique set, bank, and syndrome are stored in
footprint to construct a unique ID for the DRAM.
Rather than logging an error for each occurrence of a single symbol
correctable ECC memory error, the VMS error handler maintains the
CRD buffer-it creates a Memory CRD Entry for new footprints and
updates an existing Memory CRD Entry for errors that occur within
the range specified by the ID in FOOTPRINT. This reduces the amount
of data logged overall without losing important information~rrors are
logged per unique failure mode rather than on a per error basis.
Each Memory CRD entry consists of a FOOTPRINT, STATUS, CRD
CNT, PAGE MAPOUT CNT, FIRST EVENT, LAST EVENT, LOWEST
ADDRESS and mGHEST ADDRESS.
FIRST EVENT, LAST EVENT, LOWEST ADDRESS and HIGHEST
ADDRESS are updated to show the range of time and addresses of
errors which have occurred for a DRAM. CRD CNT is simply the total
count per footprint. PAGE MAPOUT CNT is the number of pages that
have been marked bad for a particular DRAM.

5-14 KA675/KA680/KA690 CPU System Maintenance

· STATUS contains a record of the failure mode -status of a particular
DRAM over time. This in turn determines whether or not the CRD
buffer is logged. For the first occurrence of an error within a particular .
DRAM, the memory location will be scrubbed (corrected read data is
read, then written back to the memory location) and CRD CN'!' will be
set to 1. Since most memory single-bit errors are transient due to alpha
particles, logging of the eRD buffer will not be done immediately for
the first occurrence of an error within a DRAM. The CRD buffer will,
however, be logged at the time of system shutdown (operator or crash
induced), or when a more severe memory subsystem error occurs.
If the FOOTPRINTIDRAM experiences another error (CRD CNT > 1),
VMS will set HARD SINGLE ADDRESS or MULTIPLE ADDRESSES
along with SCRUBBED in STATUS. Scrubbing is no longer performed;
instead, pages are marked bad. In this case, VMS will log the CRD
buffer immediately. The CRD Buffer will also be logged immediately
if PAGE MAPOUT THRESHOLD EXCEEDED is set in SYSTAT as a
result of pages being marked bad. The threshold is reached if more
than one page per Mbyte of system memory is marked bad. .
NOTE: CURRENT ENTRY will be zero in the Memory SBE Reduction
subpacket header if the CRD buffer was logged, not as a result of
a HARD SINGLE ADDRESS or MULTIPLE ADDRESSES error in
STATUS, but as a result of a memory unco"ectable ECC error shown as
RELATED ERROR, or as a result of CRD BUFFER FULL or SYSTEM
SHUTDOWN, all of which are shown under LOGGING REASON.

5.2.4 VMS Event Record Translation
The kernel error log entries are translated from binary to ASCII using the
ANALYZE/ERROR command. To invoke the error log utility, enter the DCL
comma.lld ANALYZE/ERROR_LOG.

Format:
~~ALYZE_ERROR_LOG [!qu.alifier(s)] [file-spec] [,...]

Example:
$ ANALYmERROR_LOG/INCLUDE=(CPU,MEMORY>/SINCE=TODAY
The error log utility translates the entry into the traditional three-column
format. The first column shows the register mnemonics, the second column
depicts the data in hex, and the last column shows the actual English
translations.

System Troubleshooting and Diagnostics

5-15

As in the above example, the VMS error handler also provides support for
the /INCLUDE qualifier, such that CPU and MEMORY error entries can
be selectively translated.
Since most kernel errors are bounded to either the processor module/system
board or memory modules, the individual error flags and fields are not
covered by the service theory. Although these flags are generally not
required to diagnose a system to the FRU (Field Replaceable Unit), this
information can be useful for component isolation.
ERF bit to text translation highlights all error flags that are set, and other
significant state-these are displayed in capital letters in the third column.
Otherwise, nothing is shown in the translation column. The translation
rules also have qualifiers such that if the setting of an error flag causes
other registers to be latched, the other registers will be translated as well.
For example, if a memory ECC error occurs, the syndrome and error address
fields will be latched as well. If such a field is valid, the translation will
be shown (e.g. MEMORY ERROR ADDRESS); otherwise, no translation is
provided.

5.2.5 Interpreting CPU Faults Using ANALVZE/ERROR
If the following three conditions are satisfied, the most likely FRU is the
CPU module. Example 5-1 shows an abbreviated error log with numbers
to highlight the key registers.
•

No memory subpacket is listed in the third column of the FLAGS
register.

•

CESR register bit <09>, CP2 10 Error, is equal to zero in the KA675
1KA6801KA690 Register Subpacket.

6) DSER register bits <07>, Q22 Bus NXM:, <05>, Q22 Bus Device Parity

Error, or <02>, Q-22 Bus No Grant, are equal to zero in the KA675
1KA6801KA690 Register Subpacket.

The FLAGS register is located in the packet header, which immediately
follows the system identification header; the CESR and DSER registers are
listed under the KA6751KA6801KA690 Register Subpacket.
CPU errors will increment a VMS global counter, which can be viewed using
the DCL command SHOW ERROR, as shown in Example 5-2.
To determine if any resources have been disabled, for example, if cache has
been disabled for the duration of the VMS session, examine the flags for
the SYSTAT register in the packet header.
In Example 5-1, a translation buffer data parity error latched in the TBSTS
register caused a machine check exception error.

5-16

KA675/KA680/KA690 CPU System Maintenance

Example 5-1: Error Log Entry Indicating CPU Error
VAX/VMS

SYSTEM ERROR REPORT

COMPILED 14-JAN-199Z 18:55:52
PAGE
1.

••• *.************************** ENTRY

LOGGED ON:

ERROR SEQUENCE 11.
DATE/TIME 27-SEP~1991 14:40:10.85
SYSTEM UPTIME: 0 DAYS 00:12:12
SCS NODE: OMEGAl
MACHINE CHECK KA680-A CPU FW REV. 2.
REVISION
00000000
SYSTAT
00000001

SID 13000202
SYS_TYPE 01390601
VAX/VMS V5. 5-1

CONSOLE FW REV. 3.9
ATTEMPTING RECOVERY

FLAGS

00000003
machine check stack frame
KA680 subpacket
..

STACK FRAME SUBPACKET
80050000
CHECK FAULT CODE - OS(x)
Current AST level - 4 (X)
ASYNCHRONOUS HARDWARE ERROR

~ACHINE

PSL

04140001
c-bit
executing on interrupt stack
PSL previous mode - kernel
PSL current mode - kernel
first part done set

KA680

ECCB0024

TBSTS

80000103
LOCK SET
TRANSLATION BUFFER DATA PARITY ERROR
em lat.ch invalid.
s5-c~~nd

- lO(X)

valid Ibex specifier ref. error stored

CESR

00000000 •

DSER

00000000 •

IPCRO

00000020
LOCAL MEMORY EXTERNAL ACCESS ENABLED

System Troubleshooting and Diagnostics

5-17

NOTE: Ownership (O-bit) memory correctable or fatal errors (MESR <04>
or MESR <03> of the processor Register Subpacket set equal to 1) are
processor module errors, NOT memory errors.

Example 5-2: SHOW ERROR Display Using VMS
$ SHOW ERROR
Device

Error Count

CPU

MEMORY

1
1
1
1

PABO:
PAAO:
PTAO:
RTA2:

5.2.6 Interpreting Memory Faults USing ANALYZE/ERROR
If"memory subpacket" or "memory sbe reduction subpacket" is listed in the
third column of the FLAGS register, there is a problem with one or more
of the memory modules, CPU module, or backplane.
•

The "memory subpacket" message indicates an uncorrectable ECC
error. Refer to Section 5.2.6.1 for instructions in isolating uncorrectable
ECC error problems.

•

The "memory she reduction subpacket" message indicates correctable
ECC errors. Refer to Section 5.2.6.2 for instructions in isolating
correctable ECC error problems.

NOTE: The memory fault interpretation procedures work only if the memory

modules have been properly installed and configured. For example, memory
modules should start in backplane slot 4 (next to the processor module in
slot 5) and proceed to slot 1 with no gaps.
NOTE: Although the VMS error handler has built in features to aid Services
in memory repair, good judgment is needed by the Service Engineer. It is
essential to understand that in many, if not most cases, correctable ECC
errors are transient in nature. No amount of repair will fix them, as
generally there is nothing to be fixed.

Memory modules can represent a great expense to the Corporation when they
are sent back to Repair with no errors. If one disagrees with the strategy

5-18 KA675/KA680/KA690 CPU System Maintenance

in this section or has questions or suggestions, please contact Corporate
Support.
5.2.6.1 Uncorrectable ECC Errors
Refer to Example 5-3, which provides an abbreviated error log for

uncorrectable ECC errors.
For uncorrectable ECC errors, a memory subpacket will be logged as
indicated by "memory subpacket" listed in the third column of the FLAGS
software register CO). Also, the hardware register MESR <11> (f) of the
processor Register Subpacket will be set equal to 1, and MEAR will latch
the error address (0).
Examine the MEMCON software register (0) under the memory subpacket.
The MEMCON register provides memory configuration information and a
MEMORY ERROR STATUS buffer Ce) that points to the memory moduleCs)
that is the most likely FRU.
Replace the indicated memory module. In Example 5-3 the most likely
FRU is indicated as memory module #2, slot 3.
The VMS error handler will mark each page bad and attempt page
replacement, indicated in SYSTAT (0). The DCL command SHOW
MEMORY (Example 5-4) will also indicate the result of VMS page
replacement.
Un correctable memory errors will increment the VMS global counter, which
can be viewed using the DCL command SHOW ERROR.
NOTE: If register MESR <11> was set equal to 1, but MESR <19:12>
syndrome equals 07, no memory subpacket will be logged as a result of
incorrect check bits written to memory because of an NDAL bus parity
error detected by the NMC. In short, this indicates a problem with the CPU
module, not memory. There should be a previous entry with MESR <22>,
NDAL Data Parity Error set equal to 1.
NOTE: One type of uncorrectable ECC error, that due to a "disown write",
will result in a eRD entry like those for correctable ECC errors. The
FOOTPRINT longword for this entry contains the message "Uncorrectable
ECC errors due to disown write". The failing module should be replaced for
this error.

System Troubleshooting and Diagnostics

5-19

Example 5-3: Error Log Entry-Indicating Uncorrectable ECC Error
VAX/VMS

SYSTEM ERROR REPORT

******************************* ENTRY

ERROR SEQUENCE 2.
DATE/TIME 4-OCT-1991 09:14:29.86
SYSTEM UPTIME: 0 DAYS 00:01:39
SCS NODE: OMEGAl
INT54 ERROR KA680-A
REVISION
SYSTAT

CPU FW REV. 2.
00000000
00000601

COMPILED 6-NOV-1991 10:16:49
PAGE 25.
13. *******************************
LOGGED ON:
SID 13000202
SYS_TYPE 01390601
VAXIVMS V5. 5-1

CONSOLE FW REV. 3.9
ATTEMPTING RECOVERY
PAGE MARKED BAD
PAGE REPLACED

fLAGS

00000006

memory subpacket •
I<A680 subpacket
KA680 REGISTER SUBPACKET
SPCR

ECC80000

MESR

80006800
ONCORRECTABLE MEMORY tcc ERROR
ERROR SUMMARY
MEMORY ERROR SYNDROME - 06 (X)

MEAR

02FFDCOO

nCRO

00000020

•

main memory error address - OBFF7000 Cit
ndal commander 1d - OOlX)

LOCAL MEMORY EXTERNAL ACCESS ENABLED
MEMORY SUBPACKET
MEMCON

0357E53F

•

MEMORY CONFIGURATION:
sets enabled - 00111111
Hs690-BA MEMORY MODULE • 1.32MB SLOT 4
MS690-BA MEMORY MODULE I 2. 32MB SLOT 3
MS690-DA MEMORY MODULE I 3. 128MB SLOT 2
_total memory - 192MB

Example 5-3 (continued on next page)

5-20 KA675/KA680!KA690 CPU System Maintenance

Example 5-3 (Cont.): Error Log Entry Indicating Uncorrectable ECC
Error

MEMORY ERROR STATUS:
MEMORY MODULE 12 SLOT 3
Bank = 00 (X)
Set .. 03 (X)
MEMCON3

88000003
64 bit mode
Base address valid
RAM size = 1MB
base address - OB(X)

Example 5-4: SHOW MEMORY Display Under VMS
$ SltON MEMOR%

System Memory Resources on 21-FEB-1992 05:58:52.58
Physical Memory Usage (pages):
Main Memory (128.00Mb)
Sad pages

Total
262144

Free
224527

In Use
28759

Modified
8858

Total

Dynamic
1

I/O Errors

Static
0
Swapped
0

Slot Usage (slots):
Process Entry Slots
Balance Set Slots

'rotal
360
324

Free
347
313

Resident
13
11

Fixed-Size Pool Areas (packets) :
Small Packet (SRP) List
1/0 Request Packet (IRP) List
Large Packet (LRP) List

Total
3061
2263
87

Free
2724
2070
61

In Use
343
193
26

Size
128
176
1856

Total
1037824
1468416

Free
. 503920
561584

In Use
533904
906832

Largest
473184
560624

Paging File Usage (pages) :
DISKSVMSOS4-0:[SYSO.SYSEXE]PAGEFILE.SYS

Free
300000

Reservable
266070

Total
300000

Dynamic Memory Usage (bytes):
Nonpaged DynamiC Memory
Paged Dynamic Memory

Of the physical pages in use, 24120 pages are permanently allocated to VMS.
$

Using the VMS command ANALYZE/SYSTEM, you can associate a page
that had been replaced (Bad Pages in SHOW MEMORY display) with the
physical address in memory.
In Example 5-5, 5ftb8 (under the Page Frame Number (PFN) column) is
identified as the single page that has been replaced. The command EVAL
5ftb8 * 200 converts the PFN to a physical page address. The result is
Obft7000, which is the MEAR address translated in Example 5-3. (Bits
<8:0> of the addresses may differ since the page address from EVAL always
shows bits <8:0> as o.

System Troubleshooting and Diagnostics

5-21

Example 5-5: Using ANALYZE/SYSTEM to Check the Physical Address
In Memory for a Replaced Page
S ADLTZE/SYS'r!:K

VAX/VMS System analyzer
SDA> SHOW PFH /B1t1)
Bad page list
Count:
Lolimlt:
High limit:

1
-1
1073741824

PFN

PTE ADDRESS

BAK

OOOSFFB8

00000000

REFCNT

FLINK

BLINK

00000000 00000000

TYPE

STATE

20 PROCESS

02 BADLIST

SDA> ~ SffbS * 200
Hex - OBFF7000
Decimal - 201289728
SDA> EXI:T
S

5.2.6.2 Correctable ECC Errors
Refer to Example 5-6, which provides an error log showing correctable ECC
errors.
For correctable ECC errors, a Single-Bit Error (SBE) Memory Subpacket
will be logged as indicated by "memory she reduction subpacket" listed in
the third column of the FLAGS software register (0).
The Memory SBE Reduction Subpacket header contains a CURRENT
ENTRY register (0) that displays the number of the Memory CRD Entry
that caused the error notification. If CURRENT ENTRY > 0, examine
which bits are set in the STATUS register (0) for this entry-GENERATE
REPORT should be set.
NOTE: If CURRENT ENTRY = 0, then the entry was logged for something
other than a single-bit memory correctable error Footprint. You will need to
examine all of the Memory CRD Entries and Footprints to try to determine
the likely FRU.

Check for the following:
•

SCRUBBED (e)-If SCRUBBED is the only hit set in the STATUS
register, memory modules should NOT generally be replaced.
The kernel performs memory scrubbing of DRAM memory cells that
may flip due to transient alpha particles. Scrubbing simply reads the
corrected data and writes it back to the memory location. Returning
memory modules that only have SCRUBBED set in STATUS will cost

5-22

KA675!KA680/KA690 CPU System Maintenance

the corporation money, since the repair centers will generally not find
.
a problem.
•

HARD SINGLE ADDRESS (e)-If the second occurrence of an error
within a footprint is at the same address (LOWEST ADDRESS =
HIGHEST ADDRESS (0», then HARD SINGLE ADDRESS will be set
in STATUS along with SCRUBBED. Scrubbing will not be tried after
the first occurrence of any error within a particular footprint. The page
will be marked bad by VMS.
Unlike uncorrectable ECC errors, the error handling code cannot
indicate if the page has been replaced. To get some idea, use DCL
command, SHOW MEMORY. If the page mapout threshold has not
been reached ("PAGE MAPOUT THRESHOLD EXCEEDED" is not set
in SYSTAT packet header register (0», the system should be restarted
at a convenient time to allow the power-up self-test and ROM-based
diagnostics to map out these pages. This can be done by entering
TEST 0 at the console prompt, running an extended script TEST AS,
or by powering down then powering up the system. In all cases, the
diagnostic code will mark the page bad for hard single address errors,
as well as any uncorrectable ECC error by default.
If there are many locations affected by hard single-cell errors, on the
order of one or more pages per MB of system memory, the memory
module should be replaced. The console command SHOW MEMORY
will indicate the number of bad pages per module. For example, if the
system contains 64 MB of main memory and there are 64 or more bad
pages, the affected memory should be replaced.
NOTE: Under VMS, the page mapout threshold is calculated
automatically. If npAGE MAPOUT THRESHOLD EXCEEDED" is set
in SYSTAT (0), the failing memory module should be replaced.

In cases of a new memory module used for rep~;r or as pa..~ of system
installation, one may elect to replace the module rather than having
diagnostics map them out, even if the threshold has not been reached
for hard single-address errors.
•

MULTIPLE ADDRESSES (O)-If the second occurrence of an error
within a footprint is at a different address (LOWEST ADDRESS not
equal to HIGHEST ADDRESS (0), MULTIPLE ADDRESSES will be
set in STATUS along with SCRUBBED. Scrubbing will not be attempted
for this situation. In most cases, the failing memory module should be
replaced regardless of the page mapout threshold.

If CRD BUFFER FULL is set in LOGGING REASON (e) Oocated in the
subpacket header) or PAGE MAPOUT THRESHOLD EXCEEDED is set in

System Troubleshooting and Diagnostics

5-23

SYSTAT (8), the failing memory module should be replaced regardless of
any thresholds.
For all cases (except when SCRUBBED is the only flag set in STATUS)
isolate the offending memory by examining the translation in FOOTPRINT
called MEMORY ERROR STATUS (.): The memory module is identified
by its backplane position. In Example 5-6, memory module #3, slot 2, is
identified as the failing module.
The Memory SBE Reduction Subpacket header translates the MEMCON
register (.) for memory subsystem configuration information.
Unlike uncorrectable memory and CPU errors, the VMS global counter,
as shown by the DeL command SHOW ERROR, is not incremented for
correctable ECC errors uniess it resuits in an error log entry for reasons
other than system shutdown.
NOTE: If footprints are being generated for more than one memory module,
especiaUy if they all have the same bit in error, the processor module,
backplane, or other component may be the cause.
NOTE: One type of uncorrectable ECC error, that due to a "disown write",

wiU result in a CRD entry like those for correctable ECC errors. The
FOOTPRINT longword for this entry contains the message 'Vncorrectable
ECC errors due to disown write". The failing module should be replaced for
this error.

5-24

KA675/KA6801KA690 CPU System Maintenance

Example 5-6:

Error Log Entry Indicating Correctable ECC Error

VAX/VMS

SYSTEM ERROR REPORT

******************************* ENTRY
ERROR SEQUENCE 2.
DATE/TIME 27-SEP-1991 09:51:13.98
SYSTEM UPTIME: 0 DAYS 00:05:06
SCS NODE: OMEGA1

CORRECTABLE MEMORY ERROR KA680-A
REVISION
00000000
SYSTAT
00000040

FLAGS

COMPILED 21-NOV-1991 16:55:58
PAGE
1.
1. ***.*******;*******************

LOGGED ON:

SID 13001401
SYS_TYPE 01390601
VAX/VMS VS. 5-1

CPO FW REV. 1.

CONSOLE FW REV. 3. 9

MEMORY SOFT ERROR LOGGING DISABLED

00000008

memory sbe reduction subpacket

MEMORY SBE REDUCTION SUBPACKET
LOGGING REASON

00000001

NORMAL REPORT

PAGE MAPOUT CNT 00000003
0357E53:

•

MEMORY CONFIGURATION:
sets enabled - 00111111
MS690-BA MEMORY MODOLE • 1.32MB SLOT 4
MS690-BA MEMORY MODOLE • 2. 32MB SLOT 3
MS690-0A MEMORY MODOLE • 3. 128MB SLOT 2
_total memory - 192MB
VALID ENTRY CNT 00000003
3.

CURRENT ENTRY

00000003
3• •

MEMORY CRe ENTRY 1.
FOOTPRINT

000C0373
MEMORY ERROR STATUS:
MEMORY MODULE 12 SLOT 3
-set - 3.
-bank - O.

icc SYNDROME - 7l(X)
_CORRECTED DATA BIT - O.
STATUS

00000010

CRe CNT

00000001

scrubbed

•

PAGE MAPOUT CNT 00000000
O.

FIRST EVENT

OD3£26EO
0094F438

LAST EVENT

OD3E26EO
0094F438

27-SEP-1991 09:50:13.07
27-SEP-1991 09:50:13.07
LOWEST ADDRESS OBFF4000
HIGHEST ADDRESS OBFF4000

Example 5-6 (continued on next page)

System Troubleshooting and Diagnostics

5-25

Example 5-6 (Cont.):

Error Log "Entry Indicating Correctable ECC Error

MEMORY CRD ENTRY 2.
FOOTPRINT

0000001C
MEMORY ERROR STATUS:
MEMORY MODOLE 11 SLOT 4
-set - O.
-bank - o.
icc SYNDROME - IC (X)
_CORRECTED DATA BIT - 4.

STATOS

00000019
PAGE MARKED BAD
HARD SINGLE ADDRESS •
scrubbed

CRD CNT

00000002
2.

PAGE KAPooT eNT 00000001
1.

FIRST EVENT

OFFFlBAO
0094F438

lAST EVEN'!'

OFFFlBAO
0094F438

27-SEP-1991 09:50:17.69
27-SEP-1991 09:50:17.69
LOWEST ADDRESS 0057FD44.
HIGHEST ADDRES3 0057FD44
MEMORY CRD ENTRY 3.
FOOTPRINT

00000500
MEMORY ERROR STATUS: •
MEMORY MODOLE 13 SLOT 2
-set - 5.
-bank - o.
icc SYNDROME - 00 (X)
_CORRECTED DATA BIT - 15.

STATOS

00000055
PAGE MARKED BAD
MULTIPLE ADDRESSES
scrubbed
GENERATE REPORT . .

CRD CNT

00000003
3.

PAGE MAPOOT CNT 00000002
2.
FIRST EVENT

122F1BOO
0094F438

lAST EVENT

122FIBOO
0094F438

27-SEP-1991 09:50:21.36
27-SEP-1991 09:50:21.36
LOWEST ADDRESS OeC72140
HIGHEST ADDRESS 08E43B28
ANAL/ERRIOUT-CRD CRD.ZPD

•

5-26 KA67S/KA6801KA690 CPU System Maintenance

NOTE: Ownership (O-bit) memory correctable or fatal errors (MESR <04>
or MESR <03> of the processor Register Subpacket set equal to 1) are
processor module errors, NOT memory errors.

5.2.7 Interpreting System Bus Faults Using
ANALVZE/ERROR
If hardware register CESR <09> (0) and/or CQBIC hardware register
DSER <07>, <05>, or <02> (f) is set equal to 1, there may be a problem
with the Q-bus or Q-bus option.
When CESR <09> is set equal to 1, examine the hardware register
CIOEAR2 (0) to determine the address of the offending option.
Example 5-7 provides an error log showing a faulty Q-bus option. The
CIOEAR2 error register indicates the first UQSSP controller as the
offending address.

System Troubleshooting and Diagnostics

5-27

Example 5-7: Error Log Entry -Indicating Q-Bus Error
VAX/VMS

SYSTEM ERROR REPORT

******************************* ENTRY

ERROR SEQUENCE 1852.
DATE/TIME 20-NOV-1991 14:26:11.14
SYSTEM UPTIME: 12 DAYS 20:04:19
SCS NODE:
MACHINE CHECK KA680-A CPU FW REV' 2.
REVISION
00000000
SYSTAT
00000001

75.

COMPILED 20-NOV-1991 14:28:13
PAGE
1.
****._************.***** •• *.***

LOGGED ON:

SID 13000202
SYS_TYPE 01410601
VAXIVMS V5.S-1

CONSOLE FW REV. 4.1
ATTEMPTING RECOVERY

FLAGS

00000003
machine check stack frame
KA680 subpacket

STACK FRAME SUBPACKET
80060000

PSL

03COOOOO
PSL previous mode - user
PSL current mode - user
first part done set

KA6BO REGISTER SUBPACKET
BPeR

ECC80024

CESR

80000200

DSER

CIOEAR2

00000080

00001468

CP2 IO ERROR
ERROR SUMMARY

0-22 BUS NXM

ep2 IO error address - 20001468
NDAL commander 1d (cp2 transac) - O(X)

IPCRO

00000020
LOCAL MEMORY EXTERNAL ACCESS ENABLED

ANAL/ERR/OUT-QBOS QBUS.ZPD

5-28 KA675/KA6801KA690 CPU System Maintenance

5.2.8 Interpreting DMA *> Host Transaction Faults Using
ANALVZE/ERROR
Some kernel errors may result in two or mor-e entries being logged. If
the SHAC DSSI adapters or the SGEC Ethernet controller or other CDAL
device (residing on the processor module) encounter host main memory
uncorrectable ECC errors, main memory NXMs or CDAL parity errors or
timeouts, more than one entry results. Usually there will be one Polled
Error entry logged by the host, and one or more Device Attention and other
assorted entries logged by the device drivers.
In these cases the processor module or one of the four memory modules

are the most likely cause of the errors. Therefore, it is essential to analyze
Polled Error entries, since a polled entry usually represents the source of
the error versus other entries, which are simply aftereffects of the original
error.
Example 5-8 provides an abbreviated error log for a polled error.
Example 5=9 provides an example of a device attention entry.
Example 5-8: Error Log Entry Indicating Polled Error
V A X / VMS

SYSTEM ERROR REPORT

***********.******************* ENTRY
ERROR SEQUENCE 15.
DATE/TIME 17-FEB-1992 05:22:00.90
SYSTEM UPTIME: 0 CAYS 00:27:48
SCS NOOE:

POLU:O tRROR KA680-A CPO FW REV. 2.
REVISION
00000000
SYSTAT
00000001

COMPILED 17-FEB-1992 05:32:21
PAGE
l.
2. *******************************
LOGGED ON:
SID 13000202
SYS_TYPE 01430701
VAX/VMS VS.S-1

CONSOLE FW REV. 4.3
ATTEMPTING RECOVERY

FLAGS

00000006
memory subpaclcet

KA6S0 subpaclcet
KA680 REGISTER SUBPACKET

Example 5-8 (continued on next page)

System Troubleshooting and Diagnostics

5-29

Example 5-8 (Cont.):

Error Log Entry Indicating Polled Error

BPCR

ECC80024

MESR

8001B800
UNCORRECTABLE MEMORY ECC ERROR
ERROR SUMMARY
MEMORY ERROR SYNDROME - 18 (Xl

MEAR

50000410
main memory error address - 00001040
ndal commander id - OS(X)

IPCRO

00000020
LOCAL MEMORY EXTERNAL ACCESS ENABLED

MEMORY SUBPACKET
MEMCON

OOS7ES3F
MEMORY CONFIGURATION:
sets enabled - 00111111
MS690-BA MEMORY MODULE t 1.32MB SLOT 4
MS690-BA MEMORY MODULE f 2. 32MB SLOT 3
MS690-DA MEMORY MODULE t 3. 128MB SLOT 2
_total memory - 192MB
MEMORY ERROR STATUS:
MEMORY MODULE 13 SLOT 2
Bank - OO(X)
Set • 00 (X)

MEMCONO

80000003
64 bit mode
Base address valid
RAM size • 1MB
base address - OO(X)

ANALlERRIOUT-TBl TB1.ZPO

5-30 KA67SIKA680/KA690 CPU System Maintenance

Example 5-9: Device Attention Entry
VAX/VMS

SYSTEM ERROR REPORT

******************************* ENTRY

ERROR SEQUENCE 15.
DATE/TIME 17-FEB-1992 05:22:00.90
SYSTEM UPTIME: 0 DAYS 00:27:48
SCS NODE:
DEVICE ATTENTION

KA680-A

COMPILED 17-FEB-1992 05:32:21
PAGE
1.
2. ***************** •• ************
LOGGED ON:
SID 13000202
SYS_TYPE 01430701
VAX/VMS VS. 5-1

CPU FW REV. 2.

CONSOLE FW REV. 4.3

OSSI SUB-SYSTEM, PABO: - PORT WILL BE RE-STARTED
PORT TIMEOUT, DRIVER RESETTING PORT
CNF

03060022
MAINTENANCE ID - 0022(X)
FIRMWARE REVISION - 06 (X)
HARDWARE REVISION - 03 (X)

PMCSR
PSR

00000000
80010000
MAINTENANCE ERROR
SHARED HOST MEMORY ERROR

PFA.~

40001044

PESR

00010000

PPR

00000000

APPROX HOST ADDR 40001044(X}
CPDAL BUS ERROR
NODE 10.
O. BYTE INTERNAL BUFFER
16. NODES MAXIMUM
44. RETRIES REMAINING
SO. RETRIES ALLOWABLE

OC450000
SHARABLE

AVAILABLE
ERROR LOGGING
CAPABLE OF INPUT
CAPABLE OF OUTPUT
C01C

ONLINE
7. ERRORS THIS ONIT

5.2.9 VAXsimPLUS and System-Initiated Call Logging (SICl)
Support
Symptom-Directed Diagnostic (SDD) toolkit support for KA6751KA680
/KA690 kernels is provided in version 2.0 of the toolkit. If version 2.0 is not
available, you should install the previous version, as it provides support for
many existing options.

System Troubleshooting and Diagnostics

5-31

VAX 4000 systems use Symptom.. Directed Diagnosis tools primarily for
notification. The VAX System Integrity Monitor Plus (VAXsimPLUS)
interactive reporting tool triggers notification for high-level events recorded
in SYSTAT and LOGGING REASON.
The VAXsimPLUS monitor simply parses for a handful of SYSTAT flags
and LOGGING reason codes. The VAXsimPLUS monitor display is updated
and triggering occurs if the threshold has been reached. Some flags have
a threshold of one; for example, SYSTAT <08> ERROR THRESHOLD
EXCEEDED will trigger VAXsimPLUS upon the first occurrence, since at
least three errors would have already occurred and been handled by VMS.
All lower level errors will ultimately set one of the conditions shown in
Table 5-4. VAXsimPLUS will examine the conditions within a 24-hour
period-thresholds are typically one or two flags or logging reason codes
within that period.
Table 5-4 lists the conditions that will trigger VAXsimPLUS notification
and updating. Figure &-8 shows the flow for the VAXsimPLUS monitor
trigger (for decision blocks with only one branch, the alternative is treated
as an ignore condition). The entries ultimately are classified as either hard
or soft. Errors that require corrective maintenance are classified as hard;
while errors potentially requiring corrective maintenance are classified as
soft.

Table 5-4: Conditions That Trigger VAXslmPLUS Notification and
Updating
Condition

Description

SYSTAT <00> = 1

"Attempting recovery"

SYSTAT <00> =0

"Full recovery or retry not possible"

SY~AT <08>

"Error threshold exceeded"

SY~AT <09> = 1

"Page marked bad for uncorrectable ECC error in main
memory"

SYSTAT <11> = 1

"Page mapout threshold for single bit ECC errors in main
memory exceeded"

LOGGING REASON <3:0> = 1

"Memory CRD buffer fulr

LOGGING REASON <3:0> =2

"Generate report as a result of hard single address or
multiple address DRAM memory fault"

LOGGING REASON <3:0> =0,
3,5-F

'1llegal LOGGING REASON"

KA6751KA680/KA690 CPU System Maintenance

Figure 5-8: Trigger Flow for the VAXsimPLUS Monitor

MLo-oo8656

VAXsimPLUS triggering notifies the customer and Services using three
message types: HARD, SOFl', and SICL Service Request. Each message
contains the single STARS article theory number, as well as the SYSTAT
or LOGGING REASON state. In addition, the SICL Service Request will
have a Merged Error Log (MEL) datafile appended. Both hard and soft
triggers will generate SICL Service Request messages.

System Troubleshooting and Diagnostics

5-33

Figure 5-9 shows the five VAXsim~LUS monitor screen displays. Table 5-5
provides a brief explanation of the fiye levels of screen displays.

Table 5-5: Five-Level VAXslmPLUS Monitor Screen Displays
Level

Explanation

1. System

The system level screen provides one box for each system being
analyzed (in Figure ~9 a single system is being analyzed). As
with each screen level, the number of reported errors is displayed
in the box. The boxes blink when the hard error thresholds are
reached; the boxes are highlighted when the soft error thresholds
are reached.

2. Subsystem

The subsystem level screen provides separate boxes for the kernel
and node information. Other boxes that may be displayed are bus,
disk., tape, etc.

3. Unit

The unit level screen provides a box for the kernel. If the
subsystem has more than one unit or device with errors, those
will be displayed as well.

4. Error Class

The error class level screen provides a box for both hard and soft
errol'S.

5. Error Detail

Two error detail level screens (hard and soft) provide the number
of reported errors along with a brief error description.

5-34 KA675lKASSO/KA690 CPU System Maintenance

Figure ~9: Five-Level VAXsimPLUS Monitor .DJsplay

1
(Systems)

AB1X

c:;:J

AB1X

I
Node Info

3
AB1X Kernel

AB1X Kernel AB1X$Kemel (NVAX4000)

AB1 X$Kernel

Soft

Hard
1

AB1X Kernel AB1X$Kemel (NVAX4000) Soft

Count: Explanation

2: Attempting Recovery

System Troubleshooting and Diagnostics

5-35

Once notification occurs, the service engineer should examine the error log
file (after using the ANALYZE/ERROR command) or read the appended
Merged Error Log (MEL) file in the SICL service request message. (The
MEL file is encrypted, refer to Section 5.2.9.1 for instructions converting
these files.) U sing the theory of interpretation provided in the previous
sections, you can manually interpret the error logs.
.NOTE: The interpretation theory provided in this manual is also a STARS

article and can be accessed via the Decoder Kit.
reproduces in full, Section 5.2 of this manual).

(Theory 30B01.xxx

In summary, a service engineer should use VAXsimPLUS notification as
follows:
1.

Make sure all four message types are sent to the Field and System
accounts.

2. Log into the Field or System account.
3. Read mail Oook for the SICL service request message with its append~
MEL file).
4.

Convert the encrypted MEL file and use the theory provided in -this
manual to interpret the error log file.

5.2.9.1 Convening the SICL Service Request MEL File

Use the following procedure to convert the encrypted MEL file that is
appended to the SICL service request message (MEL files can be converted
on site or at a support center). Example 5-10 shows a sample SICL service
request message and appended MEL file.
1. Extract the SICL mail message from mail.
2. Edit the extracted file to obtain the appended MEL file. The MEL file
is the encrypted code that appears between the rows of asterisks and
includes the words ·SICL" and "end."
3. Convert the encrypted code to a binary file using the VAXsimPLUS
decode command file as follows:

$ MCR SDD$EXE:FMGR$SICL_DECODE [MEL filename] [binary
filename]
4. Use the ANALYZE/ERROR command to produce an error log entry.

$ ANALYZEfERROR [binary filename]

5-36 KA675/KA680/KA690 CPU System Maintenance

Example 5-10:

SICL Service Request with Appended MEL File

From: AB1X::SDDSMANAGER
·VAXsimELOS Message- lS-APR-1992 10:29:21.05
To: SYSTEM
CC:
Subj: SOD T2.0 Service Request - Analysis: [30B01.200]
*****.*****************.*.******.*********************---**-*-*--*-*-*._._-.*---

VAXsimPLUS Notification Message
VAXsimPLUS has detected that the following device needs attention:
AB1XSKERNEL (NVAX4000)
AB1X
I<Al36H1S20
VAX 4000-600

DEVICE:
NODE:
SYSTEM SERIAL NUMBER:
SYSTEM TnE:

VAXslmPLOS Diagnosis Information
Attn:

Field Service

Device:
Count:

AB1X$KE~~EL

Theory:

(30501.200]

(NVAX4000)

Evidence:

Urgent action required - AB1X$KERNEL Hard error(s):
SYSTAT 9 & 1 - Page Marked Bad For Uncorrectable ECC Error In
Main Memory

*--.-._.*--*--*---*-*--*_.-.-_._-_.-.... _----* ... _------_.*._*-.-.. -._--_._.-._.

._.**______

"__ ._t***
SDDSPROFILE
is defined to be NONE, no Customer Profile included in message " *
••••• _ •• ___ *_*_* __ *_* ____ ** ____ *_ ••••• ___ * •••• _ ••• *.****. ___
SIC!.
134
AS24U)3S\@(
&0\
@

M @( $ 0 O-t 0
@ 034N-2U-,2 7

M
M

!\F»
M (H
M-A(
M

FO@4
end

('I,UP

to 0- IS SIX \A
-yp ---

\\
(S+<]P ,12

RO R.P

\\31!03
F 6

ICA-O

-«

0
@

(PO

'@U@. '

P P

@ G!: :G+Y*S

,13

?YfP (%

5.2.9.2 VAXslmPLUS Installation Tips

When installing VAXsimPLUS s the system will prompt you for infonnation.
You will need to know the serial number and system model number for the
system on which you are installing VAXsimPLUS. The serial number is
located on the front of the chassis at the bottom and to the left (the front
door must be open). The system model number is attached to the outside
of the door.
Also, if the· system does not have dialout capability, you should answer
no when asked if you want to enable SICL-if you enter yes, the system
will attempt to send mail via DSNLink resulting in error messages.
After VAXsimPLUS is installed you can activate SICL and customize
the VAXsimPLUS mailing lists so that SlCL messages are sent to an

System Troubleshooting and Diagnostics

&-37

appropriate destination(s) on site~, This way, SICL messages are received
onsite without incurring error messages regarding remote link failures .
. 5.2.9.3 VAXslmPLUS Post-Installation Tips

Once VAXsimPLUS is installed, you can set up mailing lists to direct
VAXsimPLUS messages to the appropriate destinations. If the system has
no dailout capability, SICL messages should be directed to the System and
lor Field account-this is good practice for systems with dialout and service
center support as well.

In the example that follows, the four types of mailing lists are displayed
and System and Field accounts are added to all four mailing lists using
VAXSIMIFAULT_MANAGER commands.
NOTE: The commands can be abbreviated.

DSN%SICL appears under the SICL mailing list ifyou enabled SICL during
installation.

5-38 KA675/KA680/KA690 CPU System Maintenance

$ VAXSIK/FAOL'l' SHOW MAIL

-- FSE mailing list -FIELD
COSTOMER mailing list -SYSTEM
-- MONITOR mailing list is empty -SICL mailing list --

DSN%SICL
$ VAXSIM/FAUL'l' ADD SYS'l'EK ALL

$ VAXSIM/FAUL'l' ADD FIELD ALL
$ VAXSIM/FAtJL'l' SHOW MAIL

-- FSE mailing list -FIELD
SYSTEM
CUSTOMER mailing list -FIELD
SYSTEM
-- MONITOR mailing list -FIELD
SYSTEM
SICL mailing list -DSN%SICL
FIELD
SYSTEM

To activate SICL after installation, use the following command:

$ VAXSIMlFALlLT SET SICL ON
VAXsimPLUS customer notification messages should display a phone
number for the customer to call in the event the system needs service. Use
the following commands to examine and set the phone number parameter:
$ VAXSIM/FAtJLT SHOW PARAMETER

(SET parameter)
PHONE_NUMBER
COPY
SICL
SYSTEM_INFO

(Parameter settings)
Customer Service Phone Number is unknown
Automatic copying is OFF
System Initiated Call Logging is ON

System info for AB1X
Serial n~~Der
K-~136H1520
System type
VAX 4000-600

System Troubleshooting and Diagnostics

5-39

$ VAXSIH/FAOLT SET PHONE l-aoO-DIGITAL

Finally, the VAXSIMPLUSIMERGE command is useful in examining how a
device is functioning in a cluster. The merge command collects the messages
that are being sent to the other CPUs in the cluster.

5.2.10 Repair Data for Returning FRUs
When sending back an FRU for repair, include as much of the error log
information as possible. If one or more error flags are set in a particular
entry, record the mnemonic(s) of the register(s), the hex data, and error
flag translation(s) on the repair tag. If an error address is valid, include
the mnemonic, hex data, and translation on the repair tag as well. For
memory and cache errors, include the syndrome and corrected-bitlbitin-error information, along with the register mnemonic and hex data.
Other registers which should be recorded for any entry type are SYSTAT,
MEMCON and FOOTPRINT.

5.3 Interpreting Power-On Self Test (POST) and
ROM-Based Diagnostic (RBD) Failures
If any of the tests fail, the test code displays on the console LED and, if
specified in the firmware script, a diagnostic console printout displays in
the format shown in Example 5-11.

5-40 KA675/KA680/KA690 CPU System Maintenance

Example 5-11: Sample Output with Errors

8888
" 0.
?40 2 06 FF 0000 0010 00

; SUBTEST_40_06,

•

D~_Memory_count_pages.LIS

Pi-OOOOOOOl P2-00000004 P3-FFFFFFFF P4-00000000 P5-00000004
P6-00010000 P7-00000004 PS~OOOOOOOO P9-00000000 P10-00000000
rO a 01FF4000 rl-00000004 r2-00000003 r3=FFFFFFFF r4-00000070
rS-OOOOOOOO r6-0nOOOooo r7-00000000 rS-OOOOOOOO EPC-OOOOOOOO
SCBB-20053COO
TODR-9FEBF5E9
ECR-0000008A
SCR-OOOODOOO
DSER-OOOOOOOO
QBtAR-OOOOOOO:
DEAR-OOOOOOOO
QBHBR=OlFF8000
BDR-B9F80SAF
SSCCR 00D55570
IPCRO-OOOO
CESR-OOOOOOOO
CMCDSR-OOOOC308 CSEAR1-OOOOOOOO
CSEAR2 a OOOOOOOO
CIOEARl-OOOOOOOO CI0£AR2-10000000 CNEAR-OOOOOOOO
MAPEN-OOOOOOOO
PCSTS-FFFFFSOO
PCADR-FFFFFFF8
PCCTLEFFFFFEOO
ICSR-OOOOOOOl
VMAR-000007EO
VTAG-0004008D
VDATA-AC31024E
CCTL-00000007 BCETSTS-OOOOOOOO BCETIDX-OOOOOOOO
BCETAG-OOOOOOOO
BCEDSTS-00000700 BCEDIDX-00000008 CEFSTS-00000200 BCEDECC-oOOOOOOO
CEFADR-00000008
NESTS-OOOOOOOO NEOADR-E005C9ES
NEOCMD-SOOOFF04
NEICMD-OOOOOOOO NtDATHI-OOOOOOOO NtDATLO-OOOOOOOO
MOAMR-OOOOOOOO
MMCDSRaOllllOCO
MEAR-08406010
ADO-2101B040
MESR-00080000
MEHCON 0:7; 0-80000003, i-810000~
2-00000007, 3-00000007, 4-00000007, 5-00000007, 6-00000007, 7.00000007
Normal operat1on not poss1ble.
»>
3

Several lines are printed in the error display. The first line has eight column
headings:

•

Test identifies the diagnostic test, test 140 in Example 5-11. Using
Table 5-9, you can use the test number to point to possible problems in
field replaceable units (FRUs).

•

Severity is the severity level of a test failure, as dictated by the script..
Failure of a severity level 2 test causes the display of this error printout
and halts an autoboot. An error of severity level 1 causes a display of
the first line of the error printout but does not interrupt an autoboot.
Most tests have a severity level of 2.

•

Error is two hex digits identifying, usually within 10 instructions,
where in the diagnostic the error occurred. This field is also called
the subtestlog.

e De_error (diagnostic executive error) signals the diagnostic's state and

any illegal behavior. This field indicates a condition that the diagnostic
expects on detecting a failure. FE or EF in this field means that an
unexpected exception or interrupt was detected. FF indicates an error
as a result of normal testing, such as a miscompare. The possible codes
are:

System Troubleshooting and Diagnostics

5-41

Error Code

Description·' .

Normal error exit from diagnostic

Unanticipated interrupt

Interrupt in cleanup routine

Interrupt in interrupt handler

Script requirements not met

No such diagnostic

Unanticipated exception in executive

•

vector identifies the SCB vector through which the unexpected exception
or interrupt trapped, when the de_error field detects an unexpected
exception or interrupt (FE or EF).

•

Count is four hex digits. It shows the number of previous errors that
have occurred (16 in Example 5-11).

•

Loop_8ubtest_log is an additional log generated out of the current test
specified by the current test number and subtestlog. Usually these logs
occur in common subroutines called from a diagnostic test.

e ASCII messages contain unique symbols that are terminated by the

comma in the ASCII field. These symbols identify the most recent
subtestlog entry in the listing file. The characters to the right of
the comma give the name of the listing file that contains the failed
diagnostic.

Lines 2 and 3 of the error printout are parameters 1 through 10. When
the diagnostics are running normally, these parameters are the same
parameters listed in Example 4-4.
When an unexpected machine check exception or other type of exception
occurs during the executive (de_error is EF), the stack is saved in the
parameters on lines 2 and 3, as listed in Table 5-6, Table 5-7, and
Table 5-8.

Table 5-6: Machine Check exception During Executive
Parameter Value

PI
P2

5-42

Contents of stack pointer, points to vector in P2
Vector =004, machine check

KA675/KA680/KA690 CPU System Maintenance

Table 5-6 (cant.): Machine Check Exception During executive
Parameter Value
P3

P7
P8

Machine check code
Contents of VA register
Contents of VIBA register
ICes register bit <6> and SISR <15:0>
Internal state information
Contents of shift. count (SC) register

PlO

PSL

P4
P5

Table 5-7: Exception During Executive with No Parameters
Parameter Value
PI
P2

Contents of stack pointer, points to vector in P2
Vector::: nnn (000-3FC). 200-3FC ::: Q-bus

PSL

P5
P6
P7
P8
P9
PlO

Contents of stack
Contents of stack
Contents of stack
Contents of stack
Contents of stack
Contents of stack

Table 5-8: Other Exceptions with Parameters, No Machine Check
Parameter Value
PI
P2

Contents of~..ack pointer, points to vector in P2
Vector = lUlI1 (20, 24, 34, 40, 44, 48, 4C, CB)
Optional paramete~ could be more than one LW (20, 24, CB)

Fa
P4
P5

PC
PSL

P6
P7
P8
P9
PIO

Contents of stack
Contents of stack
Contents of stack
Contents of stack
Contents of stack

Lines 4 and 5 of the error printout are General Purpose Registers (GPRs)
RO through R8 and the error program counter.

System Troubleshooting and Diagnostics

5-43

In general, the machine check exeeptions can provide a clue to the cause
of the problem. Machine check codes 01-05, 08-10, 13, OA, OB, OC, and
OD are probably due to CPU fault. Machine check codes 11 and 12 could
be a memory problem or a CPU problem. In the case of exceptions with
or without parameters (Table 5-7 and Table 5-8) the vector can provide a
clue to the fault.
When returning a module for repair, record the first line of the error
printout and the version of the ROMs on the module repair tag.
Table 5-9 lists the hex LED display, the default action on errors, and the
most likely unit that needs replacing.
The Default on Error column refers to the action taken by the diagnostic
executive under the following circumstances:
•

The diagnostic executive detects an unexpected exception or interrupt.

•

A test fails and that failure is reported to the diagnostic executive.

The Default on Error column does not refer to the action taken by the
memory tests. The diagnostic executive either halts the script or continues
execution at the next test in the script.
Most memory tests have a continue on error parameter Oabeled cont_
on_error). If you explicitly set cont_on_error, using parameter 4 in a
memory test, the test marks bad pages in the bitmap and continues without
notifying the diagnostic executive of the error. In this case, a halt on
error does not occur even if you specify halt on error in the diagnostic
executive (by answering Yes to Stop script on error'? in Utility 9F), since
the memory test does not notify the diagnostic executive that an error has
occurred.
Table 5-9 shows the various LED values and console terminal displays as
they point to problems in field-replaceable units (FRUs).

5-44 KA67SIKA680/KA690 CPU System Maintenance

Table 5-9: KA675/KA680/KA690 Console ~isplays As Pointers to
FRUs
OnEr- Normal Default Ac- On Error

rorHex: Console tion on Er- Console Display
Display ror
Test Description
LED

FROl

Power-Up Tests (Script AI)
F
E
D
C
B
9

8
8
B
B

B
B

8
8
B
9
9
9
9
B
6
7
7
C

"'V
C
C

None
None
None
66

64
63

62
61
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44

Loop
Loop
Loop
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Halt
Cont
Halt
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cant
Cont
Cont

None
None
None
?9D
?42
?35
?33
?32
?oo
?D2
?DF
?DC
?31
?30
?46
?35
?DE
?DD
?DA
154
160
?91
190
?C6
?52
?52
153

Power up
Wait for power

5.1
5.1

Utility
Check for interrupts
B_Cache diag..mode
NMC_powerop
NMC_registers
V_Cache_diag..mode

1
1.4
1
1
1.2
1
1

O~bit_Dia&..mode

O-bit_debug
No_memory_present
Memory_setup_CSRs
Memory_InitJ3itmap
P _cache_diag..mode
B_cache_diag..mode
B_CaChe_tag..debug
B_Cache_data_debug
PB...Fl~cache

VirtuaLMode
SSC_Console_SLU
CQBIC_powerup
CQBIC_registers
SSC-POwerup
SSC_Prog..timers
SSC_Prog..timers
SSC_TOY_Clock

1
1,2.3
1.2.3
2.1
1
1
1
1
1
1

1.6
1.4.3
1,4.3

1,6
i
1
7,1

1Field-replaceable unit key:

1 = KA6751KA6801KA690
2 = MS690
3 = Backplane
4 = Q22~us device
5 = System power supply
6 = H3604 console module
7 = Battery

System Troubleshooting and Diagnostics

5-45

Table 5-9 (COnt.): KA67SIKA6801KA690
Pointers to FRUs
OnEr· Normal DefaultAc- On Error

Console Displays

ror Hex Console tion on Er- Console DisDisplay ror
play
Test Description

LED

FRtJ1

Power-Up Tests (Script Al)
C
C
C
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
9
8
9
C
7

42
41
40

39
38
37
36

35
34
33
32

31
30
29
28
1:1
26
25

23
22
21
20
19
18
17
16
15

Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cent
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cont

?Cl
?84
?C5
?55
?49
?4F
?4E
?4B
?4A
14C
13F
13F
148
148
148
148
148
148
?48
148
?48
140
?47
140

?4O
?31
?C2

?8O

SSC_RAMJ)ata
SSC_ROM
SSC_registers

IntervaLTimer
Memory..FJ)M
Memory_Data
Memory_Byte
Memory_Byte..Errors
MemoryJ:CC_SBEs
MemoryJ:CC_Logic
Mem...FD~address_shorts
Mem...FD~address_shorts

Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_address_shorts
Memory_Adress
Memory_Refresh
Memory_count-P8ges
Memory_count_pages
Cache_W_memory
SSC_RAM..Data..Addr
CQBIC.-memory

1Field-replaceable unit key:

1 = KA6751KA68OIKA690
2 =MS690
3 = Backplane
4 = Q22-bus device
5 = System power supply
6 = H3604 console module
7 = Battery

5-46 KA675/KA680/KA690 CPU System Maintenance

1
1
1
1
1
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2,1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3
2,1,3
2,1,3
2, 1,3
2, 1,3
1,2
1
1,2

Table 5-9 (Cont.): KA675/KA6801KA690
Pointers to FRUs

COnsole Displays As

On Er- Normal Default Ac- On Error
ror Hex Console tion on Er- Console DisLED
Display ror
play
Test Description

FRtJ1

Power-Up Tests (Script AI)
9

12
11

5
5
8
7
7
7
7
B

10
9
8
7

6
5
4

Cant
Cont
Cant
Cont
Cont
Cont
Cont
Cont
Cont
Cont
Cant
Cont

?37
?51

Cache_w_memory
FPA

?5F

SGEe

?5C
?5C

BHAC (Bus 1)
BHAC (Bus 0)

?9A
?83
?84

INTERACTION

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt.
Halt
Halt

185

QZAJoopbackl
QZAJoopback2
QZAJIlemory

?86
?DB
?41

QZAJ)MA
Speed.
Board_Reset

19D
?42

Utility

1.2
1
1.6

1.3
1.6
1,6
4
4
4
4
1
1.4

Script AS
C
B
9
8

9D
42

B
B
B
8

35
33

DF
DC

8
B

~rorJnterrupts

?35

B_Cache_di~mode

?33

NMC_powerop
NMC_registers
V_Cache_di~mode
O-bit_DiagJnode
O_bit_debug
NO,Jnemory-presnt
MemuQ_5et"llp_CSRs
Memory_lnit-Bitmap
p _cache_di~mode

?32

?oo
?D2

?DF
?DC

131
?30
?46

1
1.4
1
1
1.2
1

1
1
1,2,3

2,1
1

1Field-replaceable unit key:

1 = KA6751KA6801KA690
2=MS690
3 = Backplane
4= Q22-bus device
5 = System power supply
6 = H3604 console module
7 Battery

System Troubleshooting and Diagnostics

5-47

" Table 5-9 (Cont.): KA675/KA680IKA690
Pointers to FRUs

Console

Displays

OnEr- Normal Default Ac-"Ou Error
rorHex Console tion on Er- Console DisDisplay ror
play
LED
Test Description

FRtJ1

Script AS
9
9
9
9
B
6
7
7
C
C
C
C
C
C
C
B
8
8
8
8
8
8
8
8
8
8
8
8

35
DE
DD
DA
54
60

91
90
C6
52
52
53
C1
34
C5
55
49

4F
4E
o4B

4A
4C
3F
SF
48
48
48
48

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt

?35
?DE
?DD
?DA
?54
?60
191
?90
?C6
?52
?52
?53
?C1
?34
?C5
?55
149
?4F'
?4E
?4B
?o4A
?4C
?SF
?3F
?48
148
?48
148

B_cache_di~mode

B_Cache_tag".debug
B_Cbace_data_debug
PB....Flush_cache
VutuaLMode
SSC_Console_SLU
CQBIC_powerup
CQBIC_registers
SSC-POwerup
SSC..ProLtimers
SSC...ProLtimers
SSC_TOY_Clock
SSC_RAMJ)ata
sse_ROM
SSC..registers
IntervaCTimer
MemoryJ'DM
MemoryJ)ata
Memory~yte

Memory~yte..Errors

Memory..BCC_SBEs
Memory..BCC_Logic
MemoryYDM".Addr_shorts
Memory_FDM...Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory...Addr_shorts
Memory...Addr_shorts

1Field-replaceable unit key:

1 =KA6751KA68OIKA69O
2 =MS690
S =Backplane
4 =Q22-bus device
5 =System power supply
6 =H3604 console module
7 =Battery

5-48 KA675/KA6801KA690 CPU System Maintenance

1
1
1
1
1
1,6
1,4,3
1,4,3
1,6
1
1
7,1
1
1
1
1

2,1,3
~ 1,3
2,1, S
2,1, S
2.1. S
2.1, S
2,1. S
2,1, S
2,1, S
2,1, S
2,1. S
2,1, S

Table 5-9 (Cont.): KA675/KA680IKA690
Painters to FRUs

Console Displays As

OnEr- Normal Default Ac- On Error
rorHex Console tion on Er- Console DisDisplay ror
play
Test Description
LED

FRlJl

ScriptA3
8
8
8
8
8
8
8
8
8
9

C
7
9
A
4
5
5
8
7

7
7

7
B
C

48
48
48
48
48
4D
47
40
40
37
C2

80
37
51
5F

5C
5C
9A

83
8485
86
DB
41

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt

?48
?48
148
?48
148
?4D
?47
?4O
?4O
?37
1C2
180
137
151
15F
?5C
15C
?SA
183
?84?85
?86
?DB
?41

Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Address
Memory_Refresh
Memory_count-pages
Memory_count-pages
Cacbe_W_memory
SSC_RAMJ)ata_Addr
CQBIC_memory
Cache_w_memory
SGEC
SHAC
SHAC
INTERACl'ION
QZA..LPBCKI
QZAJ,.PBCK2
QZA..memory
QZA_DMA
Speed
Board_Reset

2,1,3
2,1,3
2,1,3
2,1,3
2,1,3
2,1,3
2, I, 3
2,1,3
2,1,3
1,2
1
1,2
1,2
1
1,6
1,3
1,6
1,2,3,4
4
4
4
4
1
1,4

System Troubleshooting and Diagnostics

5-49

FPA

ScriptA4
1Field.replaceable unit key:

1 = KA6751KA6801KA690
2=MS690
3 = Backplane
4 = Q22-bus device
5 = System power supply
6 = H3604 console module
7 = Battery

Table 5-9 (Cont.): KA675/KA680IKA690
Pointers to FRUs

Console Displays As

On Er- Normal DefaultAc- On Error
ror Hex Console tion on Er- Console DisLED
Display ror
play
Test Description

FRtJ1

ScriptA4
Invoke script A3 (Loop on A3)

ScriptA5
8
8
8
8
8
8
8
8
8
8

SF
3F
48
48
48
48
48
48
48
48

Cont
Cont
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt

?3F
?3F
?48
?48
?48
?48
?48
?48
?48
?48

MemJ'DM..Addr_Shorts
MemYDM....Addr_Shorts
Memory_Addr_shorts
MemoIY..Addr_shorts
MemolY_Addr_shorts
MemolY_Addr_shorts
Memoty..Addr_shorts
Memoty_Addr_shorts
Memoty_Addr_shorts
Memoty_Addr_shorts

2, 1, 3
2, 1, 3
2, 1, 3
2, 1,3

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt

130
14F

MemoryJnit..Bitmap
MemoryJ)ata
Memory..Byte
Memory..Address
MemozyJ;CC_Logic
Memory_Byte..,Errors
Memory_ECC_SBEs
MemJ'DM..,Addr_Shorts
Mem...Addr_Shorts
Mem-Addr_Shorts

2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3

2,1,3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1, 3
2, 1,3

ScriptA6
8
8
8
8
8
8
8
8
8
8

30
4F
4E
4D
4C
4B
4A
3F
48
48

?4E
?4D
?4C
?4B
?4A
?3F
?48
?48

1Field-replaceable unit key:

1 = KA6751KA6801KA690
2 =MS690
3 = Backplane
4 = Q22-bus device
5 =System power supply
6 =H3604 console module
7 = Battery

5-50 KA67SIKA680lKA690 CPU System Maintenance

2,1,3

Table 5-9 (Cont.): KA675/KA680IKA690
Pointers to FRUs

COnsole Displays As

OnEr- Normal Default Ac- On Error
rorHex Console tion on Er- Console Display
Display ror
Test Description
LED

FRUl

ScriptA6
8
8
8
8
8
8
8
8
7

48
48
48
48
48
48
47
40

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt

148
148
148
148
148
148
147
140

Halt
Halt
Halt

131

Halt
Halt
Halt
Halt
Halt.
Halt
Halt

14F

180

Mem_Addr_Shorts
Mem_Addr_Shorts
MeIILAddr_Shorts
Mem_Addr_Shorts
Mem_Addr_Shorts
Mem.-Addr_Shorts
Memory_Refresh
Memory_count-pages
CQBIC_memory

2,1,3
2, 1,3

Memory_Setup_CSRs
Memory_InitJUtmap
Memory_FDM

2,1,3
2,1,3
2,1,3

Memory_Data
Memory_Byte
Memory_Address
MemoryJ:CC_Logic
Memory_B".-..e..Errors
Memory_ECC_SBEs

2.1,3
2,1,3
2,1,3
2,1,3
2,1,3
2, 1,3
2,1,3
2~ 1.3

2,1,3
2,1,3
2,1,3
2,1,3
2,1,3
2,1,3
2,1,3

Script AS
31
8
30
8
49
8
Invoke script A7.

130
149

ScriptA7
8
8
8
8
8
8
8
8

4F
4E
4D
4C
4B
4A
3F
48

Halt

?4E
14D
?4C
?4B
14A
?3F
?48

Mem..FD~ddr_shorts

Memo!y_Addr_~norts

IField-replaceable unit key:

1 = KA6751KA6801KA690
2=MS690
3 = Backplane
4 = Q22-bus device
5 = System power supply
6 = H3604 console module
7 = Battery

System Troubleshooting and Diagnostics

5-51

Table 5-9 (Cont.): KA675/KA6801KA690
Pointers to FRUs

Console Displays

OnEr- Normal Default Ac- On Error
rorHex Console tion on Er- Console DisLED
Display ror
play
Test Description

FRtJ1

ScriptA7
8
8
8
8
8
8
8
8
8
7
C

48
48
48
48
48
48
48
47
40
80

Halt
Halt
Halt
Halt
Halt
Halt
Halt
Halt
Cent
Cent
Halt

148
?48
?48
148
148
148
148
147
?40
?80
?41

Memory.-A,ddr_shorts
Memory.-A,ddr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Addr_shorts
Memory_Refresh
Memory_count-pages
CQBIC_memory
Boarc:l.Reset

2,1,3
2, 1,3
2,1,3
2, 1,3
2, 1,3
2, 1,3
2,1,3
2,1,3
2, 1,3
2,1,3
2, 1,3

Halt
Halt
Halt
Halt
Halt
Halt

?4F

MemoryJ)ata

?4E

Memory~yte

?4D
?4C
?4B
?4A

Memory.-A,ddress
MemoryJCC_Logic
Memory..Byte-Errors
MemoryJCC_SBEs

2, 1,3
2, 1,3
2, 1,3
2, 1,3
2, 1,3

Halt
Halt
Cent
Cont

141)
?47
?40
?41

Memory_Refresh
Memory_count..JNlges
Board_Reset

ScriptA9
8
8
8
8
8
8

4D
4C
4B
4A
Invoke script A5.
8
4D
47
8
40
8
41
C
End of script.

Memory~ddress

1Field-replaceable unit key:

1 = KA675iKA6801KA690
2=MS690
3 = Backplane
4 = Q22-bus device
5 = System power supply
6 = H3604 console module
7 = Battery

5-52 KA675/KA680/KA690 CPU System Maintenance

2,1,3
2,1,3
2, 1,3
2,1,3
2,1,3

5.3.1 FE Utility
In addition to the diagnostic console display and' the LED code, the FE
utility dumps diagnostic state to the console (Example 5-12). This state
indicates the major and minor test code of the test that failed, the 10
parameters associated with the test, and the hardware error summary
register.

Running the FE utility is useful if the message Normal operation not
possible is displayed after the tests have completed and there is no other
error indication, or if you need more information than what is provided in
the error display.
Example 5-12:

FE Utility Example

»>'r !"E
Lenqch-00008000, Checksum-OOOO, Busmap-07FF8000
number-OO, Subtest-OO, Loop Subtest-OO, Error type-OO
Error vector-OOOO. Severity-02. Last excep~ion PC-OOOOOOOO
Total-error coun~-OOOO, Led display-Og, Console display-9E, save mchk code-OO
parameter 1=00000000 2-00000000 3-00000000 4=00000000 5-00000000parame~er-6-00000000 7-0001E9FC 8&0001£££5 9-0001£C72 10-00000000
previous error-OOOOOOOO, 00000000, 00000000, 00000000
Flags-FFFF COSO 443£
BCache_Disable-06 KA680, 128KB BC, 14.0 ns
Return_stack-201406A8, Subtest-pc-200SB22S, Timeout-00030D40
Interrupted test number - 48, Subtest1og-04, Loop_Subtestlog-OO, Error_type-FF
Bi~map-07FECOOO,

Tes~

»>

The most useful fields displayed above are as follows:
•

Error_vector, which is the SCB vector tbro1lgh which the unexpected
intenupt or exception trapped if de_error equals FE or EF.

•

Total_error_count. Four hex digits showing the number of previous '
errors that have occurred.

•

Parameters 1 t.ltrough 10. Valid only if the test halts on e!"!'Or.

•

Previous_error. Contains the history of the last four errors. Each
long-word contains four bytes of h&formation. From left to right these
are the de_error, subtest_log, test, and subtest number (OO=FF in the
de_error).

•

Save machine check code (save_mchk_code). Valid only if the test halts
on error. This field has the same format as the hardware error summary
register.

System Troubleshooting and Diagnostics

5-53

5.3.2 Overriding Halt Protection
The ROM-based diagnostics ron in halt-protected space. When you want to
halt diagnostic execution, if the diagnostic program hangs during execution
or if the runtime of the diagnostic program is so long you want to suspend
it, enter the following commands:
»>E 20140010
!Examine the SSCCR
P 20140010
00055570
»>D * 00D05570
!Clear halt-protected space
»>T 0
!Tests can now be halted

This state is in effect only until the first break or a restart.

5.3.3 Isolating Memory Failures
This section describes procedures for isolating memory subsystem failures,
particularly when the system contains more than one MS690 memory
module.
1. SHOW MEMORYIFULL

Use the SHOW MEMORYIFULL command to examine failures detected
by the memory tests. Use this command if test 40 fails, which indicates
that pages have been marked bad in the bitmap.
You can also use SHOW MEMORYIFULL after terminating a script
tha't is taking an unusually long time to run. After terminating the
script, enter SHOW MEMORYIFULL to see if the tests have marked
any pages bad up to that point. The following is an example using this
command.
»>SHOW MEMORY/FOLL
Memory 0: 00000000 to 01FFFFFF, 32MB, 37 bad pages
Total of 32MB, 37 bad pages, 112 reserved pages
Memory Bitmap
-01FF2000 to 01FF3FFF, 16 pages
Console Scratch Area
-01FF4000 to 01FF7FFF, 32 pages
Qbus Map
-01FF8000 to 01FFFFFF, 64 pages

5-54

KA6751KA6801KA690 CPU System Maintenance

Scan of Bad Pages
-OOOOCOOO to OOOOCFFF,
-OOOOEOOO to OOOOEFFF,
-00724200 to 007247FF,
-00724AOO to 007251FF,
-00725400 to 00725BFF,
-00726400 to 00726DFF,
-00727400 to 00727DFF,

8 pages
8 pages
3 pages
4 pages
4 pages
5 pages
5 pages

»>

2. T A9
»>T (mamoxy test) starting board number ending board number
adr_incr
--

Script A9 runs only the memory tests and halts on the first error
detected. Unlike the power-up script, it does not continue. Since the
script does not rerun the test, it detects memory-related failures that
are not hard errors. You should then run the individual test that failed
on each memory module one MS690 module at a time. You can input
parameters 1 and 2 of tests 40, 47, 48, and 4A through 4F as the
starting and ending address for testing. It is easier, however, to input
the memory module numbers 1-4. For example, if test 4F fails, test
the second memory module as follows:
»>'1' 4F 2 2

You should run this test for each memory module; if a failure is detected
on· MS690 number 2, for example, and there are four memory modules
in the system, continue testing the rest of the modules to isolate the
FRU using the process of elimination.
You can also specify the address increment. For example, to test the
third memory module on each page boundary, type:
»>T 4F 3 3 200

By default, the memory tests increment by 1 Mbyte, testing one
longword in each 1 Mbyte block. If an error is detected, the tests st~"'t
testing on a page boundary. Test 48 (address shorts test) is an exception:
it checks every location in memory since it can do so in a reasonable
amount of time. Other tests, such as 4F (floating ones and zeros test),
can take up to one hour, depending on the amount of memory, if each
location were to be tested. If you do specify an address increment, do
not input less than 200 (testing on a page boundary), since almost all
hard memory failures span at least one page. For normal servicing,
do not specify the address increment, since it adds unnecessary repair
time; most memory subsystem failures can be found using the default
nQ'PQTno+o,,"

1:' ................""""" ...

System Troubleshooting and Diagnostics

5-55

All memory tests, except for 40, save the MMCDSR, MESR, MEAR in
parameters 7, 8, and 9, respectively.

3. T 9C
The utility 9C is useful if test 31 or some other memory test failed
because memory was not configured correctly. Refer to Section 4.4 to
see an example of the test 9C output.
To help in isolating an FRU, examine registers MEMCON 0-7 by
entering T 9C at the console I/O mode prompt.
4. T40
Although the SHOW MEMORYIFULL command displays pages that
are marked bad by the memory test and is easier to interpret than
test 40, there is one instance in which test 40 reports information that
SHOW MEMORYIFULL does not report. You can use test 40 as an
alternative to running script A9 to detect soft memory errors. Specify
the third parameter in test 40 (see Table 5-9) to be the threshold for
soft errors. To allow zero errors, enter the following:
»>T 40 1 4 0

This command tests the memory on four memory modules. Use it after
running memoty tests individually or within a script. If test 40 fails
with subtestlog =6, examine R5-RS to determine how many errors have
been detected.
.
Additional Troubleshooting SUggestions

If more than one memory module is failing, the CPU module, or backplane,
as well as other MS690 modules may be the cause of failure.
Always check the seating of the module before replacing it. If the seating
appears to be improper, rerun the tests.
If you are rotating MS690 modules to verify that a particular memory
module is causing the failure, be aware that a module may fail in a different
way when in a different slot. Be sure that you map out both solid singlebit and multibit ECC failures as shown in step 2 of acceptance testing
(Section 4.4), since in one slot a board may fail most frequently with multibit
ECC failures, and in another slot with single-bit ECC failures.
Be sure to put the modules back in their original positions when you are
finished.
.
If memory errors are found in the operating system error log, use the CPU
ROM-based diagnostics to verify if it is an MS690 problem or if it is related

5-56

KA675/KA6801KA690 CPU System Maintenance

to the CPU or backplane. Follow steps 1-3 of Section 4.4 and step 4 above
to aid in isolating the failure.

5.4 Testing DSSI Storage Devices
A DSSI storage device (ISE) may fail either during initial power-up or
during normal operation. In both cases, the failure is indicated by the
lighting of the red Fault indicator on the drive's front panel.

If the drive is unable to execute the Power-On Self Test (POST) successfully,
the red Fault indicator remains lit and the RunlReady indicator does not
come on, or both indicators remain on.
POST is also used to handle two types of error conditions in the drive:

•

Controller errors are caused by the hardware associated with the
controller function of the drive module. A controller error is fatal to
the operation of the drive, since the controller cannot establish a logical
connection to the host. The red Fault indicator lights. If this occurs,
replace the drive module.

•

Drive errors are caused by the hardware associated with the drive
control function of the drive module. These errors are not fatal to the
drive, since the drive can establish a logical connection and report the
error to the host. Both indicators go out for about 1 second, then the
red Fault indicator lights. In this case, run either DRVTST, DRVEXR,
or PARAMS (described in drive's service documentation) to determine
the error code.

Three configuration errors also commonly occur:
•

More than one node with the same bus node ID number

•

Identical node names

Identical ~'ffSCP urJ.t numbers

The first error cannot be detected by software. Use the SHOW DSSI
command to display the second and third types of errors. This command
lists each device connected to the DSSI bus by node name and unit number.
If the ISE is connected to ~ts front panel, you must install a bus node ID plug
in the corresponding SOCKet on the front panel. If the ISE is not connected
to its front panel, it reads the bus node ID from the three-switch DIP switch
on the side of the drive. DSSI storage devices contain the following local
programs:

System Troubleshooting and Diagnostics

5-57

DIRECT
DRVTST
DRVEXR
HISTRY
ERASE

VERIFY

DKUTIL
PARAMS

A directory, in DUP specified format, of available local programs
A comprehensive drive functionality verification test
A utility that exercises the ISE
A utility that saves information retained by the drive, including the internal
elTOrlog
A utility that erases all user data from the disk
A utility that is used to determine the amount of "margin" remaining in ondisk structures.
A utility that displays disk structures and disk data.
A utility that allows you to look at or change drive status, history, parameters,
and the internal error log

Use the SET HOSTIDUP command (described in Section 3.7.3.3) to access
the local programs listed above. Example 5-13 provides an abbreviated
example of running DRVTST for an ISE (Bus node 2 on Bus 0).
CAUTION: When running internal drive tests, always use the default (0
= No) in responding to the -Write / read anywere on medium?- prompt.
Answering yes could destroy data.

Example 5-13: Running DRVTST
»>SZ'1' IIOS'%/DOP/DSSI/BtJS:O 2 DRV'fS'f
S~arting oop server •••
Copyright (C) 1992 Digital Equipment Corporation
Write/read anywhere on medium? [l-Yes/ (O-No») !Returnl
5 minutes to complete.
GAMMA::MSCPSDUP 17-MAY-1991 12:51:20 ORVTST CPU- 0 00:00:09.29 PI-160
GAHMA::~~CPSOUP
17-MAY-1991 12:51:40 DRVTST cpo- 0 00:00:18.75 PI-332
GAMMA::MSCPSOUP 17-MAY-1991 12:52:00 ORVTST CPU- 0 00:00:28.40 PI-S03

GAHMA::MSCPSDUP
Test passed.

17-MAY-1991 12:55:42 ORVIS!

CPU-

0 00:02:13.41 PI-2388

Stopping OUP server •.•
»>

Example 5-14 provides an abbreviated example of running DRVEXR for an
ISE (Bus node 2 on Bus 0).
CAUTION: When running internal drive tests, always use the default (0
= No) in responding to the «Write/read anywere on medium?" prompt.
Answering yes could destroy data.

5-58

KA675/KA680/KA690 CPU System Maintenance

Example 5-14:
»>SE~

Running DRVEXR

BOSf/DOP/DSSI/BUS:O 2 DRVEXR

Starting DUP server •••
Copyright eCl 1992 Digital Equipment Corporation
Write/read anywhere on medium? [1-Yes/(0-No)] I~uml
Test time in minutes? [(10)-100] I~uml
Number of sectors to transfer at a time? [0 - 50J 5
Compare after each transfer? [l-Yes/{O-NO)}: IRswrnl
Test the DBN area? [2-DBN only/(l-DBN and LBN)/O-LBN only]: IRerum I
10 minutes to complete.
GAMMA::MSCPSDOP 17-MAY-1991 13:02:40 DRVEXR CPU- 0 00:00 25.37 PI-liSS
GAMMA::MSCPSDOP 17-MAY-1991 13:03:00 DRVEXR CPU- 0 00:00 29.53 PI-2503
GAMMA::MSCPSDOP 17-MAY-1991 13:03:20 DRVEXR CPU- 0 00:00 33.89 PI-3835

GAHMA::MSCPSDUP 17-MAY-1991 13:12:24 DRVEXR
13332 operations completed.
33240 LBN blocks (512 bytes) read.
o LaN blocks (512 bytes) written.
33420 DBN blocks (512 bytes) read.
o CBN blocks (512 bytes) written.
o bytes 1n error (soft).
o uncorrectable tCC errors.
complete.

CPU-

0 00:02:24.19 PI-40028

Stopping CUP server •••
»>

Refer to the RF-Series Integrated Storage Element Service Guide for
instructions on running these programs.

5,.5 USing MOP Ethernet Functions to Isolate
Failures
The console requester can receive LOOPED_DATA messages from the
server by sending out a LOOP_DATA message using NCP to set this up.
An example follows.
Identify the Ethernet adapter address for the system under test (system 1)
and attempt to boot over the network.
···system 1 (system under test)···
»>SBOW E~

Ethernet Adapter
-£ZAO (OB-OO-2B-28-18-2C)
»>~ BDO
(BOOT/R5:2 EZAO)
2 ••
-£ZAO
Retrying network bootstrap.

System Troubleshooting and Diagnostics

5-59

Unless the system is able to boot, the "Retrying network bootstrap" message
will display every 8-12 minutes. - -..
Identify the system's Ethernet circuit and circuit state, enter the SHOW
KNOWN CIRCUITS command from the system conducting the test (system
2).
***system 2 (system conducting test)***
S ICCR. !fCP
. NCP>SBOW Dam aRCOJ:~S

Known Circuit Volatile Summary as of l4-NOV-1991 16:01:53
Circuit

State

ISA-O

Loopback
Name

Adjacent
Routing Node
2S .1023 (LAR2S)

NCP>S~ c~

%SA-O S~ OFF
NCP>S~ c:IllCU%~ %SA-O SEltV%CE EDBLED
NCP>SE'r cnC07~ %SA-O S'lATE OR
NCP>LOOP CIRCO%~ %SA-O PBDXCAL ADDRESS OS-00-2B-2S-1S-2C
wrl'iI UltOES
NCP>EXt'l

H the loopback message was received successfully, the NCP prompt will
reappear with no messages.
The following two examples show how to perform the Loopback. Assist
Function using another node on the network as an assistant (system 3)
and the system under test as the destination. Both assistant and system
under test are attempting to boot from the network. We will also need the
physical address of the assistant node.
***system .3 (loopback assistant)***
»>SBOIf~

Ethernet Adapter
-EZAO (08-00-28-1E-76-9E)
»>b ezaO
(BOOT/R5:2 EZAO)
2 ••
-EZAO
Retrying network bootstrap.
***system 2***
NCP>LOOl' cacor: %SA-O PBD%aL ADDRESS OS-oo-2b-28-18-2C ASS%STAIIr.l' PJlYS%CAL
ADDIlESS 08-00-2&-11:-76-9& 1f%D MDCED ccmrr 20 LEHG'tII 200 JIZIolI I'ULL
NCP>

Instead of using the physical address, you could use the assistant node's
area address. When using the area address, system 3 is running VMS.
_··system 3···
SJICll !lCP
NCP>SBOW ROOZ

5-60

KA6751KA680/KA690 CPU System Maintenance

Node Volatile Summary as of 2i-FEB-1992 21:04:11
Executor node - 25.900 (KLATCH)
State
Ident ificat ion
Active links

- on
- DECnet-VAX V5.4-1,
s
2

VMS V5.4-2

NCP>SBOW Dam LIliES ~S'UCS

Known Line Volatile Characteristics as of 2i-FEB-1992 11:20:50
Line - ISA-O
Receive buffers
Controller
Protocol
Service timer
Hardware address
Device buffer size

- 6

•
-

normal
Ethernet
4000
08-00-28-1E-76-9£
1498

NCP>Sft CIR.CtJrr lSA-O SDD OFF
NCP>Sft cncun lSA-O SERnCE EDBLED
NCl?>Sft cncun lSA-O SDD OIl
NCP>EXI'f
$

***system 2***
$ XCR. lIa

NCP>LOOP ClRCUX'f %SA-O PBYS%CAL ADDRESS 08-00-28-2S-1S-2C ASS%~ NODE 25.900

wrm KIXED cotJ!I'f 20 I.EH'G'fB 200 HELP FtJI.L
NCl?>EXI'f
$

NOTE: The kernel's Ethernet buffer is 1024 bytes deep for the LOOP
{unctions and will not support the maximum 1500-byte transfer length.

In order to verify that the address is reaching this node, a remote node can
examjne the status of the periodic SYSTEM_IDs sent by the KA6751KA680
!KA690 Ethernet server. The SYSTEM_ID is sent every 8-l2 minutes using
NCP as in the following example:
$ MeR. lIa

NCP>Sft KCDOIoE COJIF%(;URA~Jl ClRC'Ol:'f lSA-O scavEILUtIICZ EDBt.J:D
NCP>SBOW JiODU'LE COHF%G1nlA%OJl DOiDi CDlCO'Z'fS S'fUlJS '1'0 CJIEa. LIS
NCP>EXI:T
$ 'mE ftBEJI.. LIS

Circuit name
Surveillance flaq
Elapsed time
Physical address
Time of last report
Maintenance version
Function list
Hardware address
Device type

- ISA-O
- enabled
- 00:09:37
- 08-00-2B-28-18-2C
- 21-Feb 11:50:34
- V4.0.0
- Loop, Multi-block loader, Boot, Data link counters
- 08-00-2B-2S-1S-2C

-ISA

Depending on your network, the file used to receive the output from the
SHOW MODULE CONFIGURATOR command may contain many entries,

System Troubleshooting and Diagnostics

5-61

most of which do not apply to tbe"system you are testing. It is helpful to
use an editor to search the file fot-the Ethernet hardware address of the
system under test. Existence of the hardware address verifies that you are
able to receive the address from the system under test.

5.6 Interpreting User Environmental Test Package
(UETP) VMS Failures
When UETP encounters an error, it reacts like a user program. It either
returns an error message and continues, or it reports a fatal error and
terminates the image or phase. In either case, UETP assumes the hardware
is operating properly and it does not attempt to diagnose the error.
If the cause of an error is not readily apparent, use the following methods
to diagnose the error:

•

VMS En-or Log Utility-Run the Error Log Utility to obtain a detailed
report of hardware and system errors. Error log reports provide
information about the state of the hardware device and 110 request
at the time of each error. For information about running the Error Log
Utility, refer to the VMS Error Log Utility Manual and Section 5.2 of
this manual.

•

Diagnostic facilities-Use the diagnostic facilities to test exhaustively
a device or medium to isolate the source of the error.

5.6.1 Interpreting UETP Output
You can monitor the progress ofUETP tests at the terminal from which they
were started. This terminal always displays status information, such as
messages that announce the beginning and end of each phase and messages
that signal an error.
The tests send other types of output to various log files, depending on how
you started the tests. The log files contain output generated by the test
procedures. Even if UETP completes successfully, with no errors displayed
at the terminal, it is good practice to check these log files for errors.
Furthermore, when errors are displayed at the terminal, check the log files
for more information about their origin and nature.
5.6.1.1 U ETP Log Files

UETP stores all information generated by all UETP tests and phases from
its current run in one or more UETP.LOG files, and it stores the information
from the previous run in one or more OLDUETP.LOG files. If a run
of UETP involves multiple passes, there will be one UETP.LOG or one
OLDUETP.LOG file for each pass.

5-62

KA675/KA680/KA690 CPU System Maintenance

At the beginning of a run, UETP deletes all OLDUETP.LOG files, and
renames any UETP.LOG files to OLDUETP.LOG~ Then UETP creates a
new UETP.LOG file and stores the information from the current pass in the
new file. Subsequent passes of UETP create higher versions of UETP.LOG.
Thus, at the end of a run of UETP that involves multiple passes, there is
one UETP.LOG file for each pass. In producing the files UETP.LOG and
OLDUETP.LOG, UETP provides the output from the two most recent runs.
If the run involves multiple passes, UETP.LOG contains information from
all the passes. However, only information from the latest run is stored
in this file. Information from the previous run is stored in a file named
OLDUETP.LOG. Using these two files, UETP provides the output from its
tests and phases from the two most recent runs.
The cluster test creates a NETSERVER.LOG file in SYS$TEST for each
pass on each system included in the run. If the test is unable to
report errors (for example, if the connection to another node is lost), the
NETSERVER.LOG file on that node contains the result of the test run
on that node. UETP does not purge or delete NETSERVER.LOG files;
therefore, you must delete them occasionally to recover disk space.
If a UETP run does not complete normally, SYS$TEST might contain other
log files. Ordinarily these log files are concatenated and placed within
UETP.LOG. You can use any log files that appear on the system disk for
error checking, but you must delete these log files before you run any new
tests. You may delete these log files yourself or rerun the entire UETp,
which checks for old UETP.LOG files and deletes them.
5.6.1.2 Possible UETP Errors

This section is intended to help you identify problems you might encounter
running UETP.
The following are the most common failures encountered while running
UETP:
•

Wrong quotas, privileges, or account

•

UETINITOI failure

•

Ethernet device allocated or in use by another application

•

Insufficient disk space

•

Incorrect VAXcluster setup

•

Problems during the load test

•

DECnet-VAX. error

•

Lack of default access for the FAL object

System Troubleshooting and Diagnostics

5-03

•

Errors logged but not displayed

•

No PCB or swap slots

•
•

Hangs
Bug checks and machine checks

For more information refer to the VAX 3520, 3540 VMS Installation and
Operations (ZKS166) manual.

5.7 Using Loopback Tests to Isolate Failures
You can use external loopback tests to isolate problems with the console
port, DSSI adapters (SHAC chips), Ethernet controller (SGEC chip), and
many common Q-bus options.
If one or more of these tests fail, check that the DC power and pica fuses on
the H3604 are OK. There are four pico fuses located on the back. ofH3604
console module. One fuse (F3) is on the outside, the other three are on the
component side. If a fuse is bad, replace the fuse-not the B3604.

Table 5-10 lists symptoms associated with faulty fuses. Figure 5-10 shows
the location of the H3604 fuses.

5-64 KA675/KA6801KA690 CPU System Maintenance

Table 5-10: H3604 Console Module Fuses_"__
Fuse

Part Number

Symptom

F1 (+12 V, 112 A)

12~9159-00

ThinWire Ethernet LED on H3604 is not lit.
Ethernet external loopback test 5F fails if the
Ethernet connector switch is set to ThinWire.

F2 (-12 V, 1116 A)

90-09122-00

No console display

F3 (+5 V, 2 A)

12-10929-06

LEDs on both DSSI terminators (Bus 1) on the H3604
console module are not lit; the DSSI terminator for
Bus 0 is lit.

SHOW DSSI or SHOW DEVICE commands show
DSSI bus 0, but console displays message indicating
that DSSI bus 1 terminators are missing or not
functioning.
DSSI SHAC (Bus 1) test 5C fails (oountdown number
11).
F4 (+12 V. 1.5 A)

12-10929-08

The LED on the loopback connector (12-22196-02) for
standard Ethernet is not lit.
External loopback test 5F for the standard Ethernet
passes, however.

System Troubleshooting and Diagnostics

5-65

Figure 5-10:

H3604 Console Module Fuses

IIL__~Jb;;;~It--Battery Backup Unit
=
=

J 1 TOY Clock Battery
J5 H3604 Power
J6 CPU Interface
W2 = Remote Boot Enable
W4 = FEPROM Write Enable

F1 = ThinWire Ethernet Power, 0.5 A
PN = 12-09159-00
F2 = -12V Power, 0.062 A
PN = 90-09122-00
F3 essl Terminator Power, 2.0 A
PN 12-10929-06
F4 Standard Ethernet Power, 1.5 A
PN = 12-10929-08

=
=

..
ML0-006351

5.7.1 Testing the Console Port
10 test the console port at power-up, set the Power-Up Mode switch on
the H3604 console module to the Loop Back Test Mode position (bottom)
and install an H3l03 loopback connector into the MMJ of the H3604. The
H3l03 connects the console port transmit and receive lines. At power-up,
the SLU_EXT_LOOPBACK test then runs a continuous loopback test.
While the test is running, the LED display on the H3604 console module
should alternate between 6 and 3. A value of 6 latched in the display
indicates a test failure. If the test fails, one of the following parts is faulty:
the KA6751KA6801KA690, the H3604, or the cabling.

To test out to the end of the console terminal cable:
1. Plug the :MM:J end of the console terminal cable into the H3604.
2. Disconnect the other end of the cable from the terminal.
3. Place an H8572 adapter into the disconnected end of the cable.
4. Connect the H3l03 to the H8572.

5-66

KA675/KA680/KA690 CPU System Maintenance

5. Cycle power and observe the LED.

5.7.2 Embedded DSSI Loopback Testing
NOTE: Loopback tests do not test for termination power. Use the following
procedure to check termination power:

Remove the erternal DSS] cable and terminate Busses 0 and 1. Check the
terminator LEDs to see if termination power is present.
•

No termination power at Bus 0 indicates a possible problem with the
internal cable (PN 17-02502-01) that connects DSS] Bus 0 from the
backplane.

•

No termination power at Bus 1 indicates a possible problem with the
Pico fuse (F3, PN 12-1092~6) on the H3604 console module or the
power harness module (PN 54-19789-01) for the console module. Refer
to Table 5-10 for symptoms of bad fuses.)

Power for DSS] Bus 0 is supplied by the Vterm regulator module, which
plugs into the BA440 backplane. There are no fuses on this module.
Test 56 tests both SHAC chips (the DSSI adapters). This test can be used
to check both SHAC chips, the internal DSSI (Bus 0) connectivity, external
DSSI cables, and the H3604 DSSI bus interconnect. Complete the following
procedures before running test 56.
1. Make sure the system is powered down, then connect DSSI Bus 0 to
DSSI Bus 1 with a standard external DSSI cable (BC2IM-09). Place a
DSSI terminator on the remaining DSSI connector for Bus 1. It is not
critical which Bus 1 connector is used in connecting the cable.
NOTE: The DSS] bus must be terminated for the tests to execute
successfully.

2. Remove all DSSI bus node ID plugs from storage devices on Bus O.
3. Install bus node ID plugs on the console module (H3604) so that Bus 0
and Bus 1 do not have the same bus node ID. For example, assign bus
node ill 6 to Bus 0 and bus node ID 7 to Bus 1.
4. Power up the system. Note that the red Fault indicator on the ISE front
panels will remain lit. This is normal when the bus node ID plugs have
been removed.

5. Run test 56. When tests have successfully completed, the .cons ole
prompt is displayed.

System Troubleshooting and Diagnostics

5-67

»>T 56
»>

This loopback test is useful for isolating DSSI problems. A list of FRUs in
order of probability follows:
1. The external BC2IM-09 cable
2. The Vterm dual regulator module (PN 54-20404-01)

3. The internal cable that connects DSSI Bus 0 from the backplane to the
edge of the enclosure (PN 17-02502-01)
4. The internal cable that connects the CPU to the H3604 (PN 17-0235301)
5. The 2.0 A Pico fuse (F3) on the H3604 (PN 12-10929-06)
6. The KA6751KA6801KA690 module
Test 58 is a SHAC and ISE reset and can be used to verify that ISEs can be
accessed on the DSSI storage bus. Test 58 causes data packets to be passed
between the ISEs and the adapters, verifying that the ISEs are accessible.
Enter T 58 and specify DSSI bus (0 or 1) and the DSSI node ID of the ISE
to be tested.
»>T 58 0 S

In the example above, Bus 0 node 5 was tested. (Each ISE has to be tested
separately.)

5.7.3 Embedded Ethernet Loopback Testing
NOTE: Before running Ethernet loopback tests, check that the problem is
not due to a missing terminator on a Thin Wire T-connector. Also, refer to
Table 5-10 to check for symptoms of a bad fuse.
Test SF is the intemalloopback test for SGEC (Ethernet controller).

»>T SF

For an external SGEC loopback, enter "1".
»>T SF 1

Before running test 5F on the ThinWire Ethe~et port, connect an H8223
T-connector with two H8225 terminators.
Before running test 5F on the standard Ethernet port, you must have a
12-2219fHl2Ioopback connector installed.

5-68 KA67SIKA680/KA690 CPU System Maintenance

NOTE: Make sure the Ethernet Connector Switch is set for the correct
Ethernet port.

T 59 polls other nodes on Ethernet to verify SGEC functionality. The
Ethernet cable must be connected to a functioning Ethernet. A series of
MOP messages are generated; look for response messages from other nodes.
»>T 59
Reply received from node: AA-OO-04-00-FC-64
Total responses: 1
Reply received from node: AA-OO-04-00-47-i6
Total responses: 2
Reply received from node: 08-00-2B-15-48-70
Total responses: 3

Reply received from node: AA-OO-04-00-17-14
Total responses: 25
»>

5.7.4 Q-Bus Option Loopback Testing
Module self-tests run when you power up the system. A module self-test
can detect hard or repeatable errors, but usually not intermittent errors.
A pass by a module self-test "does not guarantee that the module is good,
because the test usually checks only the controller logic.
Table 5-11 lists loopback connectors for common devices. Refer to the
Microsystems Options manual for a description of specific module self-tests.

System Troubleshooting and Diagnostics

5-69

Table 5-11: Loopback Conn~tors for Common Devices
Device

Module Loopback

CXAl61CXB16
CXY08
DIV32
DPVll
DRQB3
DRV1W
DZQl1

H3l03 + H8572 1
H3046 (SO-pin)
H3072
12-15336-10 or H325

Cable Loopback
H3l97 (25-pin)
H329 (12-27351-{)1)
17~148l-{)1 (from port 1 to port2)

70-24767~1

12-15336-10 or H325

H329 (12-27351--01)

Ethernet2
IBQOl

IEQll
KA61lll1H3604
KFQSA

KMVlA
KZQSA
LPVll

IBQ01-TA
17-01988-01
H3103
DSSI terminators
H3255
12--30552--01
12-15336-11

H3l03 + H8572
H3251

1Use the appropriate cable to collJleCt transmit-to-receive lines. H3l0 1 and Hal03 are doubleended cable connectors.
2For ThinW'ue, use H8223-00 plus two H8225-00 terminators. For standard Etherne~ use
12-22196--02.

5-70 KA6751KA680/KA690 CPU System Maintenance

Chapter 6

FEPROM Firmware Update
KA6751KA6801KA690 firmware is located on four chips, each 128 K by 8
bits of FLASH programmable EPROMs, for a total of 512 Kbytes of ROM.
(A FLASH EPROM (FEPROM) is a programmable read-only memory that
uses electrical (bulk) erasure rather than ultraviolet erasure.)

FEPROMs provide nonvolatile storage of the CPU power-up diagnostics,
console interface, and operating system primary bootstrap (VMB). An
advantage of this technology is that the entire image in the FEPROMs
may be erased, reprogrammed, and verified in place without removing the
CPU module or replacing components.
A slight disadvantage to the FEPROM technology is that the entire part
must be erased before reprogramming. Hence, there is a small "window
of vulnerability" when the CPU has inoperable firmware. Normally, this
window is less than 30 seconds. Nonetheless, an update should be allowed
to execute undisturbed.

Firmware updates are provided through a package called the Firmware
Update Utility. A Firmware Update Utility contains a bootable image,
which can be booted from tape or Ethernet, that performs the FEPROM
update. Firmware update packages, like software, are distributed through
Digital's SSB. Service engineers are notified of updates through a service
blitz or Engineering Change Order (ECO)lField Change Order (FCO)
notification.
NOTE: The NVAX CPU chip has an area called the Patchable Control Store
(peS); which can be used to update the microcode for the CPU chip.

Updates to the PCS require a new version of the firmware.
A Firmware Update Utility· image, consists of two parts, the update
program and the new firmware, as shown in Figure 6-1. The update

program uniformly programs, erases, reprograms, and verifies the entire
FEPROM.

FEPROM Rrmware Update

6-1

Figure 6-1: Firmware Update Utility Layout

Update Program

New Firmware Image
MLC>OO7271

Once the update has completed successfully, normal operation of the system
may continue. The operator may then either halt or reset the system and
reboot the operating system.

6.1 Preparing the Processor for an FEPROM Update
Complete the following steps to prepare the processor for an FEPROM
update:
1. The system manager should perform. operating system shutdown.
2. Enter console mode by pressing the Halt button twice-in to halt the
system, and out to enter console mode (»». If the Break Enable
!Disable switch on the console module is set to enable (indicated by
1), you can halt the system by pressing the IBreak I key on the console
terminal.

3. In order to update the firmware, jumper W4 on the inside of the
H3604 console module must be in the "write enable mode," as shown in
Figure 6-2. (Write enable is the factory setting.)

To access the jumper you must open the H3604 console module by
unlocking the two half-tum screws that hold it closed.

6-2 KA675/KA680/KA690 CPU System Maintenance

Figure 6-2: W4 Jumper Setting for Updating Flnnware

M~7697

6.2 Updating Firmware Via Ethernet
To update firmware via the EL~ernet, the "client" system (the t$l1"get
system to be updated) and the "server" system (the system that serves
boot requests) must be on the same Ethernet segment. The Maintenance
Operation Protocol (MOP) is the transport used to copy the network image.
Use the following procedure to update firmware via the Ethernet: .1. Be sure H3604 jumper W4 is in the correct ("write enable mode")

position (Section 6.1).
2. Enable the server system's NCP circuit using the following VMS
commands:
$ MeR NCP

NCP>SET CIRCUIT <<::ireuit> STATE OFF

FEPROM Rrmware Update

6-3

NCP>SET CIRCUIT <circuit>, SERVICE ENABLED
NCP>SET CIRCUIT <circuit~-STATE ON

Where <circuit> is the system Ethernet circuit. Use the SHOW
KNOWN CIRCUITS command to find the name of the circuit.
NOTE: The SET CIRCUIT STATE OFF command will bring down
the system ~ network.

3. Copy the file containing the updated code to the MOM$LOAD area on
the server (this procedure may require system privileges). Refer to
the Firmware Update Utility Release Notes for the Ethernet bootable
filename. Use the following command to copy the file:
$ COPY <filename>.SYS lvIOM$LOAD:*.*

Where <filename> is the Ethernet bootable filename provided in the
release notes.
4. On the client system, enter the command BOOT/100 EZ at the console
prompt (»».

The system then prompts you for the name of the file.
NOTE: Do NOT type the ".SYS" suffi:x when entering the Ethernet
bootfile name. The MOP load protocol only supports 15 character
filenames.

5. .After the FEPROM upgrade program is loaded, simply type "Y' at the
prompt to start the FEPROM blast. Example 6-1 provides a console
display of the FEPROM update program.
CAUTION: Once you enter the bootfile name, do not interrupt the
FEPROM blasting program, as this can damage the CPU module. The
program takes several minutes to complete.
NOTE: On systems with a VCB02 terminal, you wiU see an abbreviated
form of the following example.

6-4 KA675/KA680/KA690 CPU System Maintenance

Example 6-1:

FEPROM Update Via Ethernet

***** On Server System *****
S XClt. lIICP
NCP>S~

CIRCCTr lSA-O ~~ OFF
NCP>SE'l' CIllCOU lSA-O SER.VICE EDBLEJ)
NCP>S~ CIllCOU lSA-O Su.n OR
NCP>EXI~

S
S COpy D.680 V41 EZ. SYS MCIC$LQI,I): * . *
S
***** On Client System

»> ~/l00 ZZAO
(BOOT/R5:100 EZAO)
2 ••

Bootfile: E680_V41_EZ

-£ZAO
1 .• 0 ••

FEPROM BLASTING PROGRAM
blasting in V4.1 •••
--CAUTION--EXECUTING THIS PROGRAM WILL CHANGE YOUR CURRENT ROM --Do you really want to continue [YIN] ? : Y
00 NOT ATTEMPT TO INTERRUPT PROGRAM EXECUTION!
DOING SO MAY RESULT IN LOSS OF OPERABLE STATE!
The program will take at most several minutes.
starting uniform_program •••
byte 00070000 has been written with o· s •••
byte 00060000 has been written with 0'5 •••
byte 00050000 has been written with O·s •••
byte 00040000 has been written with 0'5 •••
byte 00030000 has been written with 0' 5 •••
byte 00020000 has been written with O' s •..
~u 0001000(;: riGS baen writtan with O's •••
byte 00000000 has been written with O's •••
starting erase •••
byte 00070000 has been erased •••
byte 00060000 has been erased •••
byte 00050000 has been erased •••
byte 00040000 has been erased •••
byte 00030000 has been erased •••
byte 00020000 has been erased •••
byte 00010000 has been erased •••
byte 00000000 has been erased •••
starting program •••

Example 6-1 (continued on next page)

FEPROM Firmware Update

6-5

Example 6-1 (Cont.):

FEPROM Update VIa Ethernet

byte 00070000 has been reprogrammed •••
byte 00060000 has been reprogrammed •••
byte 00050000 has been reprogrammed •••
byte 00040000 has been reprogrammed •••
byte 00030000 has been reprogrammed •••
byte 00020000 has been reprogrammed •••
byte 00010000 has been reprogrammed •••
byte 00000000 has been reprogrammed •••
FEPROM Programming successful

»>

6. Press the Restart button on the SCP or enter "T 0" at the console prompt
(»».
7. If the customer requires, return jumper W4 on the inside of the H3604
console module to the "write disable mode" setting and close and secure
the console module by locking the half-turn screws.

6.3 Updating Firmware Via Tape
To update firmware via tape, the system must have a TF85, TK70, or TKSO
tape drive.
If you need to make a bootable tape, copy the bootable image file to a tape
as shown in following example. Refer to the release notes for the name of
the file.
$ IRIT MIAS: "VOLOME NAME"
$ MOORT /BLC)CK SIZE ;; 512 mAS: "VOLUME NAME"
$ COPY/COBTIG-<file Dama> mAS:<fil. Dime>
$ DISMOtJR'l' MIAS
$

Use the following procedure to update firmware via tape:
1. Be sure H3604 jumper W4 is in the correct ("write enable mode")
position (Section 6.1).
2. At the console prompt (»», enter the BOOT/lOO command for the tape
device, for example: BOOT/IOO MIAS.
Use the SHOW DEVICE command if you are not sure of the device
name for the tape drive.
The system prompts you for the name of the file. Enter the bootfile
name.

6-6

KA67SIKA680/KA690 CPU System Maintenance

3. After the FEPROM upgrade program is loaded, simply type "Y" at the
prompt to start the FEPROM blast. Example-6-2 provides a console
display of the FEPROM update program.
CAUTION: Once you enter the bootfile name, do not interrupt the
FEPROM blasting program, as this can damage the CPU module. The
program takes several minutes to complete.
NOTE: On systems with a VCB02 terminal, you will see an abbreviated
form (Example 6-2.

4. Press the Restart button on the SCP or enter "T 0" at the console prompt

(»».
5. If the customer requires, return jumper W4 on the inside of the H3604
console module to the "write disable mode" setting and close and secure
the console module by locking the half~turn screws.

FEPROM Rrmware Update

6-7

Example 6-2: FEPROY Update VIa Tape
»> BOOr/100 ~
(BOOI/RS:100 MIAS)
2 ••
Boo~!ile:

X680_V4l_ZZ

-MIAS
1 •• 0 ••
FEPROM BLASTING PROGRAM
blasting in V4.1 •••
--CAUTION--EXECUTING THIS PROGRAM WILL CHANGE YOUR CURRENT ROM --DC' you red ly ..,ant t ¢ e¢~t i nl,!f,'> (Y IN' 1 ? ! T
DO NOT ATTEMPT TO INTERRUPT 'PROGRAM EXECUTION!
DOING SO MAY RESULT IN LOSS OF OPERABLE STATE!
The program will ~ake a~ mos~ several minu~es.
s~artlng

uniform-program •••

byte 00070000 has been wrl~~en wi~h O's •••
byte 00060000 has been wri~ten wi~h 0'5 •••
by~e 00050000 has been wrl~ten wi~h O's •••
byte 00040000 has been wrl~~en wi~h 0' s ••.
by~e 00030000 has been wri~ten with O's .••
by~e 00020000 has been wrl~~en wl~h O's •••
byte 00010000 has been wri~ten with 0' 5 •••
byte 00000000 has been wri t~en with O's •••
s~arting erase ••.
00070000 has been erased •••
byte 00060000 has been erased •••
by~e 00050000 has been erased •••
byte 00040000 has been erased •••
byte 00030000 has been erased •••
byte 00020000 has been erased •••
by~e OQ010000 has been erased •••
byte 00000000 has been erased •••
starting program •••
by~e

byte 00070000 has been reprogrammed •••
byte 00060000 has been reprogrammed ••.
byte 00050000 has been reprogrammed ••.
byte 00040000 has been reprogrammed •••
byte 00030000 has been reprogrammed •••
byte 00020000 has been reprogrammed ••.
byte 00010000 has been reprogrammed •••
byte 00000000 has been reprogrammed ••.
FEPROM Programming successful

»>

6-8 KA6751KA6801KA690 CPU System Maintenance

6.4 FEPROM Update Error Messages
The following is a list of error messages generated by the FEPROM update
program and actions to take if the errors occur.

MESSAGE:
update enable jumper is disconnected
unable to blast ROMs...

ACTION:
Reposition update enable jumper (Section 6.1).

MESSAGE:
ROM programming error-expected byte: xx actual byte: xx
at address: xxxxxxxx

AGrION:
Replace the CPU module.

MESSAGE:
ROM uniform pgming error-expected byte: 00 actual byte: xx
at address: xxxxxxxx

ACTION:
Tum off the system, then tum it on. If you see the banner message
as expected, re-enter console mode and try booting the update program
again. If you do not see the usual banner message, replace the CPU
module.
.

MESSAGE:
ROM erase error-expected byte: ff actual byte: xx:
at address: xxxxxxxx
ACTION:
Replace the CPU module.
Patchable Control Store (PCS) Loading Error Messages

The following is a list of error messages that may appear if there is a
problem with the pes. The pes is loaded as part of the power-up stream
(before ROM-based diagnostics are executed).

MESSAGE:
CPU is not an NVAX
COMMENT:
CPU_TYPE as read in NVAX. SID is not = 19 (decimal), as is should be
for an NVAX processor.

FEPROM Firmware Update

6-9

MESSAGE:

Microcode patch/CPU rev mismatch
COMMENT:
Header in microcode patch does not match MICROCODE_REV as read
in NVAX SID.
MESSAGE:

PCS Diagnostic failed
COMMENT:
Something is wrong with the PCS. Replace the NVAX chip (or CPU
module).
MESSAGE:

Unexpected SIE
COMMENT:
SYS_TYPE as read in the ROM 8IE does not reflect that an NVAX CPU
is present.

6-10

KA675/KA680/KA690 CPU System Maintenance

Appendix A

KA675/KA680/KA690 Firmware
Commands
This appendix provides information on console mode control characters and
firware commands for the CPU module.

A.1 Console 1/0 Mode Control Characters
In console I/O mode, several characters have special meaning:
IRETURN I
Also <CR>. The carriage return ends a command line. No action is taken

IRUBOUT I

on a command until after it is terminated by a carriage return. A null
line terminated by a carriage return is treated as a valid, null command.
No action is taken, and the console prompts for input. Carriage return is
echoed as carriage return, line feed (<CR><LF».
When you press fRUBOOil the console deletes the previously typed
character. The resulting display differs, depending on whether the console
is a video or a hardcopy terminal.
For hardcopy terminals, the console echoes a backslash (\), followed
by the deletion of the charac:ter. If you press additional rubouts, the
additional deleted characters are echoed. If you type a nonrubout
character, the console echoes another backslash, followed by the character
typed. The result is to echo the characters deleted, surrounding them with
backslashes. For example:
EXAMI;E IRUBOUT I RUBOUT INE<CIb
The console echoes: EXAMI;E\ E;\NE<CR>

The console sees the command line: EXAMINE<CR>
For video terminais, the previous character is erased and the cursor is
restored to its previous position.

IClRUA I and F14

The console does not delete characters past the beginning of a command
line. If you press mere rubouts than there are characters on the line, the
extra rubouts are ignored. A rubout entered on a bla:ak line is ignored.
Toggle insertioDloverstrike mode for command line editing. By default, the
console powers up to overstrike mode.

KA6751KA680/KA690 Firmware Commands

A-1

ICTRVSI or up_
arrow (or dow~
arrow)

ICTRUD I and left

Recalls previous command(s). Command recall is only operable if sufficient
memory is available. This func::tion may then be enabled and disabled using
the SET RECALL command.
Move cursor left one position.

arrow

ICTRUE I
ICTRUF I and right

Moves cursor to the end of the line.
Move cursor right one position.

arrow

ICTRUHL

MoVe cursor to the beginning of the line.

backspace, and

F12

ICTRUUI
ICTRUSI
ICTRlJOI
ICTRUR!
IcTRLIC!

ICTRUOI

Echoes AU<CR> and deletes the entire line. Entered but otherwise ignored
if typed on an empty line.
Stops output to the console terminal untillcTRLIQ! is typed. Not ec:boed.
Resumes output to the console terminal. Not echoed..
Echoes <CR><LF>, followed by the current command line. Can be used to
improve the readability of a command line that has been heavily edited.
Echoes AC<CR> and aborts processing of a command. When entered as
part of a command line, deletes the line.
Ignores transmissions to the console terminal until the next ICTRUO I is
entered. Echoes A() when disabling output, not echoed when it re-enables
output. Output is re-enabled if the console prints an error message, or if
it prompts for a command from the terminal Output is also enabled by
entering console I/O mode, by pressing the IBREAKI key, and by pressing

ICTRLICL

A.1.1 Command Syntax
The console accepts commands up to 80 characters long. Longer commands
produce error messages. The character count does not include rubouts,
rubbed-out characters, or the IREllJRN I at the end of the command.
You can abbreviate a command by entering only as many characters as are
required to make the command unique. Most commands can be recognized
from their first character. See Table A-5.
The console treats two or more consecutive spaces and tabs as a single
space. Leading and trailing spaces and tabs are ignored. You can place
command qualifiers after the command keyword or after any symbol or
number in the command.
All numbers (addresses, data, counts) are hexadecimal (hex), but symbolic
register names contain decimal register numbers. The hex digits are 0
through 9 and A through F. You can use uppercase and lowercase letters in
hex numbers (A through F) and commands.
The following symbols are qualifier and argument conventions:

A-2

KA675!KA6801KA690 CPU

[]

An optional qualifier or argument

A required qualifier or argument

A.1.2 Address Specifiers
Several commands take one or more addresses as arguments. An address
defines the address space and the offset into that space. The console
supports five address spaces:
Physical memory
VIrtual memory
General purpose registers (GPRs)
Internal processor registers (!PRs)
The PSL
The address space that the console references is inherited from the previous
console reference, unless you explicitly specify another address space. The
initial address space is physical memory.

A.1.3 Symbolic Addresses
The console supports symbolic references to addresses. A symbolic reference
defines the address space and the offset into that space. Table A-llists
symbolic references supported by the console, grouped according to address
space. You do not have to use an address space qualifier when using a
symbolic address.
Table A-1: Console Symbolic Addresses
Symb
Addr
Symb
Symb
Addr
Addr

Symb

Addr

IG-General Purpose Registers
ao

R12~}

III

!t9

R13(FP)

ll6

RIO

RI4 (sp)

lUI

RIS (PC)

OE
OF

1M-Processor Status Longword
PSI.

Note: All symbolic values in this table are in hexadecimal.

KA6751KA680/KA690 Firmware Commands

A-3

Table A-1 (Cont.): Console S.Ymbolic Addresses
Symb

Addr

Symb

Addr

Symb

Addr

Symb

Adclr

II-Internal Processor Registers
pr$.Pi»

pr$.JICbb

pr$_txCI

pr$_esp

prS_scbb

prSJ,Edb

prS_1Sp

prSjpl

prS_tzl:1

pr$_U5p

prS_altly

prS_tmb

pr$_iap

prS_lilT

prS_1i1l'

1'(

pr$..ioreset

31
38

prS...mcesr

prS..,PObr

prSjCCl

pr$.,.mapen

pr$.,lIOlr

prSJJicr

pr$_tbia

prS,.plbr

OA.

prSjcr

prS_la'YpC

pr$_tbil

pl'$_t.oclr

pI$_laYpll

prS,.pUr

pr$_1br

pr$_alr

pr$_licl

pr$_tbchk

pr$_ea'

pIS_Cd!

pr$..neoactr

At
pr$Jdecc

pr$J)CIlt.lta

pr$J)CIltid:r.

pr$_bcetar

It5

pr$_bcedsta

pr$_bc:edicbc

pr$..bceGecc

ItS

pr$_ceIiadr

pr$,.aeocmcl

prS..nedathi

pr$_aeclatlo

prS_'f1I:W'

pr$_YIaC

prS_yclata

prS_icsr

pr$,.pc:adr

F2
F3

pr$..,pcsta

pr$..]:IUlOde

pr$_ceI'at.I

pr$_tbadr

pr$_DUt.I

prS_tbsta

pr$..bctac

01000000

01400000

prS,.pctar

01800000

prS..neicmcl

pr$_bcftuah

pr$,.pcctl

pr$,.pcclap

01COOOOO

IP-Physica1 (VAX 110 Space)
qbmem

qbio

200000OO

rom

2004000O

acr

2008000O

clMr

ipcrO

2OOOtr40

UC_raJD

20140400

300000OO

qbmbr

20080010

bcir

20084004

2OO8OOCM

qbear

20080008

dear

2OO8OOOC

ipcrl

20001C42

ipcr2

20001144

ipcrS

2OOOtr46

IK_Cf

20140010

uc_cbtcr

20140020

. ._cDectr

20140030

A-4 KA675/KA680IKA690 CPU

Table A-1 (Cont.): Console Symbolic AddreSses
Symb

Addr

Symb

lK_adOmat

20140130

lIC_aciOmsk

20140134

lK_acilmaL

20140140

AC_adlmsk

20140144

1Ie_ta()

20140100

lIIIC_tirO

20140104

&8C_tnUO

20140108

lK_tinO

201401Oc:

ISC_Lcrl

20140110

S&C_tirl

20140114

lSC_tnirl

20140118

lSC_tivrl

2014011c

nicslO

2000500O

mcsrl

20008004

mcar2

20008008

mcsr3

2OOO8OOC

tUcar4

~10

mearS

20008014

Dicar6

20008018

Dicar'7

2000801C

2OOCS02O

Dicar9

20008024

DicarlO

20008028

mcsrll

2OOO802C

Addr

Symb

Addr

Symb

Addr

IP-Physical (VAX 110 Space)

nicar12

20008030

mcsr13

20008034

mcsr14

20008038

mcsr15

2OOO803C

agee_letup

2000800O

1gec_tzpoll

20008004

spe.ftPOll

20008008

aaec..ma

2OOO8OOC

apc_tba

~10

sgec_ltatua

20008014

agec..mocie

20008018

agec_lbr

2OOO801C

20008020

sgec_wdt

20008024

Igec..mCC

20008028

qec_'Yerlo

2OOO802C

IFC_verhi

20008030

sgec-PZ'OC

20008034

spe-bpt

20008038

qec_cmd

2OOO803C

lhac1_1SWC'

20004030

lhac1_uhma

20004044

shacl..,pqbbr

20004048

abac1-PC'

2OOO404c

Ihacl..,PeSl'

20004050

ahac1-Pf&r

20004054

ahacl~

20004058

Ibacl..J111:1C1r

3)Q()406C

lhac1J)CqOc:r 20004080

1hac1.J)Cqlcr 20004084

1hac1..,pdf'qcr 20004090

lhacl...,pmfqcr 20004094

ahac1..]1Ua'

2OOCM098

Ihacl.J1eC1'

abacl~

200040AO

lhacl.,picr

shac1,.pmtcr

2OOO4OAB

lhacl.,Pmtecr 2OClO4OAC

lhac2_uwcr

20004230

lhac2_ubma 20004244

abac2..Pqbbr

20004248

1hac2.,.par

2000424c

shac2,.Pe1r

20004250

1hac2,J6r

20004254

shac2...PJ1l'

2OOO42S8

abac2,.pmcsr

2OOO42fiC

shac2..,pcqOcr 20004280

lbac2.J)Cqlcr

3)()()4284

abac2,..pcq2cr 20004288

1bac2.JXq3cr

lhac2,.pdiqcr 20004290

1hac2...,pmfqcr 20004294

ahac2..,P1ra'

20004298

abac2.J1eC1'

2OOCM29C

3)()()42A4

ahac2.,pmLcr

2OOO42A8

abac2,.pmLecr 2OOO42AC

ahac2..,Pdcr

200042A0

lhac2.,picr

200<M0A4

ahac1-PCq2cr 20004088

Ihacl.JXq3cr 2OOO408C

2OOO409C

Ibac_uwer

20004230

lhaC_ahma

20004244

lbac..Pqbbr

20004248

sbac..,Plr

2OOO42k

shacJeR'

20004-250

!mu..,pfar

20004254

ahac-PF

20004258

lhac-PZZIcsr

2OOO425C

lhacJCqOcr

20004280

Ihac.J)Cqlcr

20004284

abae..PCCf2cr

2(iO(K288

1bac..JlCCl3cr

~-.-o42SC

Ihac..,pdf'qcr

20004290

Ihac.,pmf'qcr

20004294

abac..,Plra'

20004298

ahac.JleCl'

2OOO429C

IhacJldcr

200042A0

abac.,.picr

3)()()42A4

Ihac..,pmLcr

2OOO42A8

ahac.,Pmtecr

2OOO42AC

amccwb

21000110

memconO

21018000

me1llCOl11

21018004

memC0D2

21018008

memc0a3

2101800c

memc0n4

21018010

memc0n5

21018014

memC0D6

21018018

memcoa7

2101801c

memai&8

21018020

memsig9

21018024

memsi,10

21018028

memlign

2101802c

me:Wa'12

21018030

mezui,13

21018034

. memlig14

21018038

memli,15

2101803c

mear

21018040

mser

210180«

l2mCdsr

21018048

mOflJDZ'

2101804C

cear

21031000

ncadsr

21020004

aear1

21020008

caear2

2102000c

cpioeal

210'JD0I0

cpi0u2

21020014

Ddeu

21020018

KA6751KA680/KA690 Rrmware Commands

A-5

Table A-2 lists symbolic addresses ,that you can use in any address space.

Table A-2: Symbolic Addresses Used in Any Address Space
Symbol

Description

•

The location last referenced in an EXAMINE or DEPOSIT command.
The location immediately following the last location referenced in an EXAMINE
or DEPOSIT command. For references to physical or virtual memory spaces, the
location referenced is the last address, plus the size of the last reference (1 for
byte, 2 for word, 4 for longword, 8 for quadword). For other address spaces, the
address is the last address referenced plus one.
The location immediately preceding the last location referenced in an EXAMINE
or DEPOSIT command. For references to physical or virtual memory spaces,
the location referenced is the last address minus the size of this ~feTent;'!P. (l for
byte, 2 for word, 4 for longword, 8 for quadword). For other address spaces, the
address is the last address referenced minus one.
The location addressed by the last location referenced in an EXAMINE or
DEPOSIT command.

A.1.4 Console Numeric Expression Radix Specifiers
By default, the console treats any numeric expression used as an address
or a datum as a hexadecimal integer. The user may override the default
radix by using one of the specifiers listed in Table A-3.

Table A-3: Console Radix Specifiers
Forml

Form 2

BadU

I\b

Binuy

1\0

Octal

I\d

Decimal

Hexadecimal, default

For instance, the value 19 is by default hexadecimal, but it may also be
represented as %bll00l, %031, %d25, and %x19 (or in the alternate form
as "'bll00l, "'031, "'d25, and "'xI9).

A..1.5 Console Command Qualifiers
You can enter console command qualifiers in any order on the command
line after the command keyword. The three types of qualifiers are data
control, address space control, and command specific. Table A-4 lists and

A-6

KA67SIKA6801KA690 CPU

describes the data control and address space control qualifiers. Command
specific qualifiers are listed in the descriptions of inmvidual commands.

Table A-4: Console Command Qualifiers
Qualifier

Description

Data Control
IB
IW
IL
IQ
1N:{count}

/STEP: {size}

/WRONG

The data size is byte.
The data size is word.
The data size is longword.
The data size is quadword.
An unsigned hexadecimal integer that is evaluated into a longword. This
qualifier determines the number of additional operations that are to take place
on EXAMINE, DEPOSIT. MOVE. and SEARCH commands. An error message
appears if the number overflows 32 bits.
Step. Overrides the default increment of the console current reference.
Commands that manipulate memory. such as EXAMINE. DEPOSIT, MOVE,
and SEARCH. normally increment the console current reference by the size
of the data being used.
Wrong. On writes, 3 is used as the value of the ECC bits, which always
generates double bit errors. Ignores ECC errors on main memory reads.

Address Space Control
IG

II
N

IP
1M
IU

General purpose register (GPR) address space, RO-R15. The data size is
always longword.
Internal processor register (IPR) address space. Accessible only by the MTPR
and MFPR instructions. The data size is always longword.
Vll'tual memory address space. All access and protection checking occur. If
access to a program running with the current PSL is not allowed, the console
issues an error message. Deposits to virtual space cause the PTE<M> bit to be
set. If memory mapping is not enabled, virtual addresses are equal to physical
ad~-esses. Note that when you examine vi..~.!a! memory. the address space
and address in the response is the physical address of the y"h-tuU adoL-ess.
Physical memory address space.
Processor status longword (PSL) address space. The data size is always
longword.
Access to console private memory is allowed. This qualifier also disables
virtual address protection checks. On virtual address writes, the PrE<M>
bit is not set if the IU qualifier is present. This qualifier is not inherited; it
must be respecifi.ed on each command.

KA675!KA680/KA690 Firmware Commands

A-7

A.1.6 Console Command Keywords
Table A-5 lists command keywords" by type. Table A-6 lists the parameters,
qualifiers, and arguments for each console command. Parameters, used
with the SET and SHOW commands only, are listed in the first column
along with the command.
You should not use abbreviations in programs. Although it is possible
to abbreviate by using the minimum number of characters required to
. uniquely identify a command or parameter, these abbreviations may
become ambiguous at a later time if an updated version of the firmware
contains new commands or parameters.

Tabie A-5: Command Keywords by Type
Processor Control

Data Transfer

Console Control

BOOT

DEPOSIT
EXAMINE
MOVE

CONFIGURE
FIND
REPEAT

SET
SHOW
TEST

CO:r-,~

HALT
INITlALIZE

START
UNJAM

Table A-6: Console Command Summary
Command

Quali1iers

Argument

BOOT

/RS:(boot..ftap} It-US.,.}

((boo&_dn'ice}L{boot_deYice}]...)

DEPOSIT

IBIW ILIQ-IGII IV IPIMIU
1N:{coan1) 1STEP:{Ue) IWBONG

{addzua}

EXAKINE

IBIW ILIQ-IGII IV IPIMIU
1N:{coUDt} 1STEP:(Ue) IWBONG

[{1IIidrea}]

FIND

IJIEJIIBPB

Other(s)

CONFICUBE
CONTINtJE
(data) [(data)]

IINSTRUCTION

HALT
HELP

INlTIALlZE
MOVE

IBlWfLlQ-IVIPIU
1N:(coant.} 1Sl'EP:{1ize} !WRONG

NEXT
REPEAT

A-8 KA675/KA6801KA690 CPU

(Sl"C_addreu}

[(coant)]
{commaDd}

(dest_add.rea)

Table A~ (Cont.): Console Command Sumtrlary
Command

Qualifiers

Argument

Other(s)

SEAIlCH

IBIWILIQ-NIPIU

{Ital't_addzal}

{pattern} [{mukl]

/N;{count} lS"l'EP:{size} !WRONG
/NOT

SETBFLAG

{bitmap}

SET BOOT

[{boot_de¥ice}[,{boot_de¥iceJJ.-

SET CONTROLP

{OIl}

SET HALT

!halt_action}

SET HOST

IOUP IDSSI JBUS:(OIl}

{node_number}

[{tukl]

SET HOST

IDUP IUQSSP {IDISK ! trAPE }
IDUPIUQSSP

{controller_number}
{car_address}

[{taakl]
[{taakl]

SET HOST

n.lAINTENANCE IUQSSP
ISERVICE
IMAlNTENANCE IUQSSP

{controller_number}
{car_address}

SETLANGUAGE

{languap-type}

SETBECALL

{OIl}

SHOWBFLWG
SHOW BOOT
SHOW CONTROLP SHOWDSSI
SHOW HALT
SHOW LANGUAGE SHOW MEMORY

etJLL

SHOWQBUS
SHOW RECALL
SHOWRLV12
SHOW SCSI
SHOWTRANSLA·
TION

{pDys_addrea}

SHOWUQSSP
SHOWVEBSION

S"'aART

{~-==}

TEST

{test_number}

[{parameters}]

{address}

{count}

UNJAM
X

A.2 Console Commands
This section describes the console I/O mode commands.
commands at the console I/O mode prompt (»».

Enter the

KA675/KA680/KA690 Firmware Commands

A-9

A.2.1 BOOT
The BOOT command initializes the processor and transfers execution to
VMB. VMB attempts to boot the operating system from the specified device
or list of devices, or from the default boot device if none is specified. The
console qualifies the bootstrap operation by passing a boot flags bitmap to
VMB in RS.

Format:
BOOT [qualifier-list] [(boot_device), {boot_device) ,__]
If you do not enter either the qualifier or the device name, the default value
is used. Explicitly stating the boot flags or the boot device overrides, but
does not permanently change, ~'le corresponding default value.
When specifying a list of boot devices (up to 32 characters, with devices
separated by commas and no spaces), the system checks the devices in the
order specified and boots from the first one that contains bootable software.
NOTE: If included in a string of boot devices~ the Ethernet device~ EZAO,
should be placed only as the last device of the string. The system will
continuously attempt to boot from EZAO.

Set the default boot device and boot flags with the SET BOOT and SET
BFLAG commands. If you do not set a default boot device, the processor
times out after 30 seconds and attempts to boot from the Ethernet port,
EZAO.
Qualifiers:
Command specific:
lR5:lboot_flags} A 32-bit hex value passed to VMB in RS. The console does not interpret this
value. Use the SET BFlAG command to specify a default boot ftags longword.
Use the SHOW BFLAG command to display the longword. Table 3-4 lists the
supported R5 boot ftags.
/{boot_flags}
Same as 1R5:{boot_flags}
[device_name] A character string of up to 32 characters. Longer strings cause a VAL TOO
BIG error message. When specifying a list of boot devices, the device names
should be separated by commas and DO spaces. Apart from checking the
length, the console does DOt interpret or validate the device name. The console
converts the string to uppercase, then passes VMB a string descriptor to this
device name in RO. Use the SET BOOT command to specify a default boot
device or list of devices. Use the SHOW BOOT command to display the default
boot device. The factory default device is the Ethernet port, EZAO. Table 3-3
lists the boot devices supported by the KA6751KA6801KA690.

A-1Q

KA675/KA680IKA690 CPU

Examples:
»>SHOW BOOT
DOAO
»>SHOW BFLAG
00000000
»>B
!Boot using default boot flags and device.
(BOOT/R5:0 DUAO)
2 ••

-DOAO
»>80 XQAO !Boot using default boot flags and
(BOOT/R5:0 XQAO)
!specified device.

2 •.
-XQAO
»>BOOT I/O !Boot using specified boot flags and
(BOOT/R5:10 DOAO)
!default device.
2 ••
-DUAO

»>BOOT /RS:220 X~O !Boot using specified boot
(BOOT/R5:220 XQAO)
! flags and device.
2 ••

-XQAO

A.2.2 CONFIGURE
The CONFIGURE command invokes an interactive mode that permits
you to enter Q22-bus device names, then generates a table of Q22-bus
I/O page device CSR addresses and interrupt vectors. CONFIGURE is
similar to the VMS SYSGEN CONFIG utility. This command simplifies
field configuration by providing information that is typically available only
"\\-ita't a r":"YL.'ling operati.llg system. Refer to the example below and use the
CONFIGURE command as follows:
1. Enter CONFIGURE at the console I/O prompt.

2. Enter HELP at the Device, Number? prompt to see a list of devices
whose CSR addresses and interrupt vectors can be determined.
3. Enter the device names and number of devices.
4.

Enter EXIT to obtain the CSR address and interrupt vector
assignments.

The devices listed in the HELP display are not necessarily supported by
the CPU.

KA6751KA6801KA690 Firmware Commands

A-11

Format:
CONFIGURE

Example:
»>COHFIGORE
Enter device configuration, HELP, or EXIT
Device,Number? help
Devices:
LPVll
KXJll
DLVllJ
DZQ1l
RLVl2
TSV05
RXV2l
DRV1lW
DELQA
DMVll
DEQNA
DESQA
RRD50
RQC25
KFQSA-DISK
TQK50
RV20
KFQSA-TAPE
KMVll
!EOll
CXA16
CXB16
CXY08
VCBOl
LNV21
QPSS
DSV1l
ADV11C
KWVllC
ADVllD
VCB02
AAVllD
DRQ3B
VSV21
IBQOl
IOV11A
IDVIID
IAVllA
!AVllB
MIRA
DESNA
IGQll
DIV32
KIV32
KWV32
KZQSA
Numbers:
1 to 255, default is 1
Device,Number? rqdx3,2
Device,Number? dhvll,2
Device,Number? deqna
Device,Number? kfqsa-tape
Device, Number? cxy08
Device,Number? mira
Device,Number? tqk50
Device,Number? tqk10
Device,Number? dhq1l
Device,Number? Inv1l
Device, Number? exit
Address/Vector Assignments
-774440/120 DEQNA
-712150/154 RQDX3
-760334/300 RQDX3
-774500/260 KFQSA-TAPE
-760444/304 TQK50
-760450/310 TQK10
-760500/320 DHV11
-760520/330 DHV11
-760540/340 CXY08
-760560/350 DRQ11
-776200/360 LNV11
-761260/370 MIRA
»>

A-12 KA6751KA6801KA690 CPU

DZVll
DRVIlB
RQDX3
TQK10

DFAOl
DPVll
KDA50
T08lE

DROll

DHVll

QVSS
AAVllC
QDSS
IDV11B
ADQ32
DTCN5

LNV11
AXVllC
DRVllJ
IDVllC
DTC04
DTC05

A.2.3 CONTINUE
The CONTINUE command causes the processor to begin instruction
execution at the address currently contained in the PC. It does not perform
a processor initialization. The console enters program I/O mode.
Format:

CONTINUE
Example:
»>CORrnroE
$
!VMS DCL prompt

A.2.4 DEPOSIT
The DEPOSIT command deposits data into the address specified. If you
do not specify an address space or data size qualifier, the console uses the
last address space and data size used in a DEPOSIT, EXAMINE, MOVE,
or SEARCH command. After processor initialization, the default address
space is physical memory, the default data size is longword, and the default
address is zero. If you specify conflicting address space or data sizes, the
console ignores the command and issues an error message.
Format:
DEPOSIT [qualifier-list] (address) (data} [data.••]
Qualifiers:

Data control: IB, IW, IL, IQ, 1N:{count}, ISTEP:{size}, /WRONG

Address space control: IG, II, 1M, IP, N, /U
Arguments:
{address}

A longword address that specifies the first location into which data is deposited.
The address can be an actual address or a symbolic address.

{data}

The data to be deposited. If the specified data is larger than the deposit data size,
the firmware ignores the command and issues an error response. If the specified
data is smaller than the deposit data size, it is extended on the left: with zeros.

[{data}]

Additional data to be deposited (as many as can fit on the command line).

Examples:
»>O/P/B/N:1FF 0 0

Clear first 512 bytes of
physical memory.

KA6751KA6801KA690 Firmware Commands

A-13

»>DIV/L/N:3 1234 5

- !, Deposit 5 into four longwords
:-- starting at virtual memory address
! 1234.

»>D/N:8 RO iiWliiii

Loads GPRs RO through R8 with -1.

»>D/L/P/N:l0/ST:200 0 8

Deposit 8 in the first lonqword of
the first 17 pages in physical
memory.

»>D/N:200 - 0

Starting at previous address, clear
513 lonqwords or 2052 bytes.

A.2.S EXAMINE
The EXAMINE command examines the contents of the memorv location OT
register specified by the address. If no address is specified, +.is assumed.
The display line consists of a single character address specifier, the physical
address to be examined, and the examined data.

EXAMINE uses the same qualifiers as DEPOSIT. However; the /WRONG
qualifier causes EXAMINE to ignore ECC errors on reads from physical
memory. The EXAMINE command also supports an /INSTRUCTION
qualifier, which will disassemble the instructions at the current address.

Format:
EXAMINE [qualifier-list] [address]
Qualifiers:

Data control: IB, IW, IL, IQ, IN: {count} , ISTEP:{size}, /WRONG
Address space control: IG, II, 1M, IP, N, /U
Command specific:
IlNSTRUCTION

Disassembles and displays the VAX MACR0-32 instruction at the specified
address.

Arguments:
[{address} ]

A longword address that specifies the first location to be examined. The
address can be an actual or a symbolic address. If no address is specified,

+ is assumed.

A-14

KA675/KA6801KA690 CPU

Examples:
»>EX PC

Examine the PC.

G OOOOOOOF FFFFFFFC

»>EX SP

Examine the SP.

G OOOOOOOE 00000200

»>EX PSL
M 00000000 041FOOOO
»>E/M

Examine the PSL.
Examine PSL another way.

M 00000000 041FOOOO

»>E R4/H:5
G 00000004
G 00000005
G 00000006
G 00000007
G 00000008
G 00000009

Examine R4 through R9.
00000000
00000000
00000000
00000000
00000000
801D9000

»>EX PR$ SCBB
I

! Examine the SCBB, IPR 1 7

(decimal) .

00000011 2004AOOO

Examine local memory o.

»>E/P 0
P 00000000 00000000

»>EX /IRS 20040000
P 20040000
11 BRa

20040019

»>EX /IRS!R: 5 20040019
DO MOVL
P 20040019
P 20040024
D2 MCOML
D2 MCOML
P 2004002F
P 20040036
7D MOVQ
P 2004003D
DO MOVL
P 20040044
DB MFPR

! Disassemble from branch.
I Ai20140000,@i20140000
@i20140030,@i20140502
SAiOE,@i20140030
RO,@i201404B2
I Ai201404B2,R1
SAi2A,B A44 (R1)

»>E/mS
P 20040048

! Look at next instruction.
S"'i2B,B A48 (R1)

Examine 1st byte of ROM.

DB MFPR

»>

A.2.6 FIND
The FIND command searches main memory, starting at address zero for
a page-aligned 128-Kbyte segment of good memory, or a restart parameter
block (RPB). If the command finds the segment or RPB, its address plus
512 is left in SP (R14). If it does not find the segment or RPB, the console
issues an error message and preserves the contents of SP. If you do not
specify a qualifier, IRPB is assumed.
Format:

FIND [qualifier-list]

KA675/KA680/KA690 Firmware Commands

A-15

Qualifiers:
Command specific:
!.MEMORY Searches memory for a page-aligned block of good memory, 128 Kbytes in length.
The search looks only at memory that is deemed usable by the bitmap. This
command leaves the contents of memory unchanged.
IRPB
Searches all physical memory for an RPB. The search does not use the bitmap to
qualify which pages are looked at. The command leaves the contents of memory
unchanged.-

Examples:
»>EX SP

Check the SP.

G OOOOOOOE 00000000
»>FIND I!!EM

Look for a valid 128 ~~~~es.
Note where it was found.

»>EX SP
G OOOOOOOE 00000200

»>Fl:ND /RPB
?2C FND ERR 00C00004

Check for valid RPB.
None to be found here.

»>

A.2.7 HALT
The HALT command has no effect. It is included for compatibility with
other VAX consoles.
Format:
HALT

Example:
»>BALT
»>

Pretend to halt.

A.2.B HELP
The HELP command provides information about command syntax and
usage.
Format:

HELP
Example:
»>BELP
Following is a brief summary of all the commands supported by the
console:

A-16

KA675/KA6801KA690 CPU

UPPERCASE
I
[]
<>

denotes a keyword that you must type in
denotes an OR condition
denotes optional parameters
denotes a field specifying a syntactically correct value
denotes one of an inclusive range of integers
denotes that the previous item may be repeated

Valid qualifiers:
/B /W /L /Q /INSTRUCTION
/G /I /V /P /M

/STEP: /N: /NOT
IWRONG /U
Valid commands:
BOOT [[/R5:]<boot_flags>] [<boot_device>]
CONFIGURE
CONTINUE
DEPOSIT [<qualifiers>] <address> <datum> [<datum> ..• ]
EXAMINE [<qualifiers>] [<address>]
FIND [ /MEMORY I /RPB]
HALT
HELP

INITIALIZE
MOVE [<qualifiers>] <address> <address>
NEXT [<count>]
REPEAT <command>
SEARCH [<qualifiers>] <address> <pattern> [<mask>]
SET BFLG <boot_flags>
SET BOOT <boot device>
SET CONTROLl? <0' •• 1 I DISABLED I ENABLED>
SET HALT <0 •. 4 I DEFAULT I RESTART I REBOOT I HALT I RESTART REBOOT>
SET HOST/DUP/DSSI/BUS:<O .• l> <node_number> [<task>]SET HOST/DUP/UQSSP </DISKI/TAPE> <controller number>[<task>]
SET HOST/DUP/UQSSP <physical CSR address> [<task>]
SET HOST/MAINTENANCE/UQSSP/SERVICE <controller number>
SET HOST/MAINTENANCE/UQSSP <physical CSR address>
SET LANGUAGE <1: :15>

SET RECALL <0 •• 1 I DISABLED I ENABLED>
SHOW BFLG
SHOW BOOT
SHOW CONTROLP
SHOW DEVICE
SHOW DSSI
SHOW ETHERNET
SHOW HALT
SHOW LANGUAGE
SHOW MEMORY [/FOLL]
SHOW QBUS
SHOW RECALL
SHOW RLV12
SilOn SCSI

KA6751KA680/KA690 Rrmware Commands

A-17

SHOW TRANSLATION <physical .address>
SHOW UQSSP
SHOW VERSION
START <address>
TEST [<test code> [<parameters>]]
UNJAM
X <address> <count>

»>

A.2.9 INITIALIZE
The INITIALIZE command performs a processor initialization.

Format:
INITIALIZE

The following registers are initialized:
Register

State at Initialization

PSL

041FOOOO

IPL
ASTLVL

1F
4

SISR

ICes
RXCS
TXCS

Bits <6> and <0> clear, the rest are unpredictable
80

MAPEN
Caches
Instruction buffer
Console previous reference

TODR
Main memory
General registers
Halt code
Bootstrap-in-progress flag
Internal restart-in-progress flag

o
Flushed
Unaffected
Longword, physical, address 0
Unaffected
Unaffected
Unaffected
Unaffected
Unaffected
Unaffected

The firmware clears all error status bits and initializes the following:
CDAL bus timer
Address decode and match registers
Programmable timer interrupt vectors
SSCCR

A-18

KA67S/KA6801KA690 CPU

Example:
»>INIT
»>

A.2.10 MOVE
The MOVE command copies the block of memory starting at the source
address to a block beginning at- the destination address. Typically, this
command has an IN qualifier so that more than one datum is transferred.
The destination correctly reflects the contents of the source. regardless of
the overlap between the source and the data.
The MOVE command actually performs byte, word, longword, and
quadword reads and writes as needed in the process of moving the data.
Moves are supported only for the physical and virtual address spaces.
Format:

MOVE [qualifier-list] {src_addressl (dest_addressl
Qualifiers:

Data control: IB, IW, IL, IQ, 1N:{count), ISTEP:{size}, /WRONG
Address space control: IV, fU, IP
Arguments:
{src_address}

A longword address that specifies the first location of the source data to be
copied.

{deat_address}

A longword address that specifies the destination of the first byte of data.
These addresses may be an actual address or a symbolic address. If no
address is specified, + is assumed.

KA6751KA680/KA690 Firmware Commands

A-19

Examples:
»>EX/N:4 0
P 00000000 00000000
P 00000004 00000000
P 00000008 00000000
P OOOOOOOC 00000000
P 00000010 00000000

Observe destination.

»>EX/N:4 200
P 00000200 58000520
P 00000204 585E04Cl
P 00000208 00FF8FBB
P 0000020C 5208A800
P 00000210 540CA80E

Observe source data.

»>MOV/H:4 200 0

Move the data.

»>EX/N:4 0
P 00000000 58000520
P 00000004 585E04Cl
P 00000008 00FF8FBB
P OOOOOOOC 5208A800
P 00000010 540CA80E
»>

Observe moved data.

A.2.11 NEXT
The NEXT command executes the specified number of macro instructions.
If no count is specified, 1 is assumed.
After the last macro instruction is executed, the console reenters console
110 mode.

Format:
NEXT [count}
The console implements the NEXT command, using the trace trap enable
and trace pending bits in the PSL and the trace pending vector in the SCB.
The console enters the "Spacebar Step Mode". In this mode, subsequent
spacebar strokes initiate single steps and a carriage return forces a return
to the console prompt.
The following restrictions apply:
•

If memory management is enabled, the NEXT command works only if
the first page in SSC RAM is mapped in SO (system) space.

•

Overhead associated with the NEXT command affects execution time
of an instruction.

A-20

KA675/KA6801KA690 CPU

•

The NEXT command elevates the IPL to 31· for long periods of time
(milliseconds) while single-stepping over sev~ commands.

•

Unpredictable results occur if the macro instruction being stepped over
modifies either the SCBB or the trace trap entry. This means that you
cannot use the NEXT command in conjunction with other debuggers.

Arguments:
{count}

A value representing the number of macro instructions to execute.

Examples:
»>DEP 1000 SOD6S004
»>DEP 1004 12S00S01
»>DEP 1008 00FE11F9
»>EX /INSTRUCTION IN:S 1000
P 00001000
P 00001002
P 00001004
P 00001007
P 00001009
P 0000100B

04 CLRL

D6 INCL

D1 CMPL
12 BNEQ
11 BRB
00 HALT
»>OEP PR$ SCBB 200
»>OEP PC 1000

»>
»>5
P 00001002
p 00001004
P 00001007
P 00001002
»>N 5
P 00001004
P 00001007
P 00001002
P 00001004
l? 00001007

Create a simple program.

List it.

S"'iOS,RO
00001002
00001009
Set up a user SCBB •••
••• and the pc.

! Single step •••
RO
SPACEBAR
S"'i05,RO
SPACEBAR
01 CMPL
00001002
SPACEBAR
12 BNEQ
RO
CR
06 INCL
••• or multiple step the program.
S"'iOS,RO
01 CMPL
00001002
12 BNEQ
RO
06 INCL
S"'fOS,RO
01 CMPL
00001002
12 BNEQ

D6 INCL

»>N 7
P 00001002
P 00001004

P 00001007
P 00001002
P 00001004
P 00001007
P 00001009

06 INCL
D1 CMPL
12 BNEQ
06 INCL
01 CMPL
12 BNEQ
11 BRB

11 BRB

00001009

S""i05,RO
00001002
RO
S"'f05,RO
00001002
00001009

»>N
P 00001009

»>

KA675!KA680/KA690 Rrmware Commands

A-21

A.2.12 REPEAT
The REPEAT command repeatedly displays and· executes the specified
command. Press IcTRLJC I to stop the command. You can specify any valid
console command except the REPEAT command.
Fonnat:
REPEAT (command)

Arguments:
{command} A valid console command other than REPEAT.

Examples:
»>REPEAT EX PR$ TODR ! Watch the clock.
I 0000001B 5AFE78CE
I 0000001B 5AFE78D1
I 0000001B 5AFE78FD
I 0000001B 5AFE7900
I 0000001B 5AFE7903
I 0000001B 5AFE7907
I 0000001B 5AFE790A
I 0000001B 5AFE790D
I 0000001B SAFE7910
I 0000001B 5AFE793C
I 0000001B SAFE793F
I 0000001B 5AFE7942
I 0000001B SAFE7946
I 0000001B SAFE7949
I 0000001B 5AFE794C
I 0000001B SAFE794F
I 0000001B SAC
»>

A.2..13 SEARCH
The SEARCH command finds all occurrences of a pattern and reports the
addresses where the pattern was found. If the /NOT qualifier is present,
the command reports all addresses in which the pattern did not match.
Format:

SEARCH [qualifier-list] (address} (pattern) [(mask)]

SEARCH accepts an optional mask that indicates bits to be ignored (don't
care bits). For example, to ignore bit 0 in the comparison, specify a mask

of 1. The mask, if not present, defaults to o.

A-22

KA6751KA680/KA690 CPU

A match occurs if (pattern and not mask) = (data.and not mask), where:
Pattern is the target data
Mask is the optional don't care bitmask (which defaults to 0)
Data is the data at the current address
SEARCH reports the address under the following conditions:
!NOT Qualifier

Match Condition

Action

Absent
Absent
Present
Present

'llue
False
'llue
False

Report address
No report
No report
Report address

The address is advanced by the size of the pattern (byte, word, longword,
or quadword), unless overridden by the ISTEP qualifier.
Qualifiers:
Data control: IB, IW, IL, IQ, IN: {count} , ISTEP:{size}, !WRONG
Address space control: /P, N, /U
Command specific:
!NOT

Inverts the sense of the match.

Arguments:
(start_address) A longword address that specifies the first location subject to the search. This
address can be an actual address or a symbolic address. If no address is
specified, + is assumed.
{pattem}
The target data.
[{mask}]
A mask of the bits desired in the comparison.

Examples:
»>DEP /P/L/N:1000 0 0
»>
»>DEP 300 12345678
»>DEP 401 12345678
»>DEP 502 87654321
»>
»>SEARCB 1.:1000 IST:1 0 12345678
P 00000300 12345678
P 00000401 12345678

»>SEARCB 1.:1000 0 12345678
P 00000300 12345678

»>SEARCB /R:1000 IHOT 0 0

! Clear some memory.
! Deposit some search data.

Search for all occurrences
of 12345678 on any byte
boundary. Then try on
longword boundaries.
Search for all non-zero
longwords.

P 00000300 12345678

KA675/KA6801KA690 Rrmware Commands

A-23

P 00000400 34567800
P 00000404 00000012
P 00000500 43210000
P 00000504 00008765

»>SEARCH /N:l000 /ST:l 0 1 iiiiiiiE

Search for odd-numbered
longwords on any boundary.

P 00000502 87654321
P 00000503 00876543

P 00000504 00008765
P 00000505 00000087
»>SEARCH /N:l000 /8 0 12
P 00000303 12

P 00000404 12

»>SEARCH /N:l000 IST:l
»>
»>
»>

Iv 0 FEll

Search for all occurrences
of the byte 12.
Search for all words that
could be interpreted as
a spin (10$: brb 10$).
Note that none were found.

A.2.14 SET
The SET command sets the parameter to the value you specify.
Format:

SET (parameter) (value)
Parameters:
BFLAG

Sets the default R5 boot flags. The value must be a hex number of up
to eight digits. See Table 3-4 for a list of the boot flags.

BOOT

Sets the default boot device. The value must be a valid device name or
list of device names as specified in the BOOT command description in
Section A.2.1.

CON'I'ROLP

Sets Control-P as the console halt condition. instead of a BREAK.
Values of 1 or Enabled set Control-P recognition. Values of 0 or Disabled
set BREAK recognition. In either case, the setting of the Break Enable
!Disable switch.

HALT

Sets the user-defined halt action. Acceptable values are the keywords
"default", "restart", "reboot", "halt", "restart_reboot", or a number in the
range 0 to 4 inclusive.

HOST

Connects to the DUP or MAINTENANCE driver on the selected node
or device. The KA6751KA6801KA690 DUP driver supports only "send
data immediate" messages and those devices that support the messages.
It does not support "send data" or "receive data" messages. Note the
hierarchy of the SET HOST qualifiers below.

A-24 KA675/KA680/KA690 CPU

IDUP-Uses the DUP driver to examine or modify parameters of a
device on either the nSS! bus or on thtrQ22-bus.
IBUS:n-Selects the desired DSS! bus. A value of 0 selects DSSI
bus 0 (internal backplane bus). A value of 1 selects nSS! bus 1
(external console module bus).
IDSSI node-Selects the DSSI node, where "node" is a number
from Oto 7.
IUQSSP-Attacbes to the UQSSP device specified, using one of
the following methods:

/DISK n-Specifies the disk controller number. where n is a
number from 0 to 255. The resulting fixed address for n=O is
20001468 and the floating rank for n>O is 26.
trAPE n-Specifies the tape controller number. where n is a
number from 0 to 255. The resulting fixed address for n=O is
20001940 and the floating rank for n>O is 30.
csr_address-Specifies the Q22-bus 110 page CSR address
for the device.
IMAINTENANCE-Examines and modifies the KFQSA EEPROM
configuration values. Does not accept a task value.
IUQSSPISERVICE n-Specifies service for KFQSA controller module
n where n is a value from 0 to 3. (The resulting fixed
address of a KFQSA controller module in maintenance mode
is 20001910+4*n.)
Icsr_address--Specifies the Q22-bus 110 page CSR address
for the KFQSA controller module.
LANGUAGE

Sets console language and keyboard type. If the current console
terminal does not support the multinational character set (MCS), then
this command bas no effect and the console message appears in English.
Values are 1 through 15. Refer to Example 4-1 for the languages you
can select.

RECALL

Sets command recall state to either &'iABLED (1) or DISABLED (0).

Qualifiers: Listed in the parameter descriptions above.

KA67SIKA680/KA690 Firmware Commands

A-25

Examples:
»>
»>SET BFLAG 220
»>
»>SET BOOT DOAO
»>
»>SET HOST/DOP/DSSI/BOS:O 0
Starting DOP server •••
DSSI Node 0 (SUSAN)
Copyright e 1990 Digital Equipment Corporation
DRVEXR Vl.0 0 5-JUL-1990 15:33:06
DRVTST V1.0 0 5-JUL-1990 15:33:06
HISTRY Vl.O D 5-JOL-1990 15:33:06
ERASE V1.0 0 5-JUL-1990 15:33:06
PARAMS Vl.0 0 5-JOL-1990 15:33:06
DIRECT·V1.0 0 5-JOL-1990 15:33:06
End of directory
Task Name?PARAMS
Copyright C 1990

Digital Equipment Corporation

PARAMS>STAT PATH
10

Path Block

o PB
6 PB
1 PB
4 PB
5 PB
2 PB
3 PB

FF811ECC
FF811FDO
FF8120D4
FF8121D8
FF8122DC
FF8123EO
FF8124.E4

Remote Node

---------------

Internal Path
KFQSA KFX Vl.0
KAREN RFX VlOl
WILMA RFX VIOl
BETTY RFX V10l
DSSII VMS VS.O
VMS BOOT
3

DGS_S

DGS_R

0
0

0
0
0
0

0
0
0
0
0
0

PARAMS>EXIT
Exiting •••
Task Name?
Stopping DUP server •••
»>
»>SET HOST/DOP/DSSI/BOS:O 0 PARAMS
Starting DUP server •••
DSSI Node 0 (SUSAN)
Copyright C 1990 Digital Equipment Corporation

PARAMS>SHOW NODE
Parameter

Current

Default

Type

Radix

NODENAME

SUSAN

RF71

String

Ascii

A-26 KA675/KA680/KA690 CPU

MSGS_S

MSGS_R

------ ------- ------ --------

0
0
0
0
0
14328
61

0
0

0
0
14328
61

PARAMS>SHOW ALLCLASS
Parameter

CUrrent

Default

Type

Radix

ALLCLASS

Byte

Dec

PARAMS>EXI'r
Exiting ..•
Stopping DOP server .•.
»>
»>SE'r HOST!NAINT/UgSSP 20001468
OQSSP Controller (772150)
Enter SET, CLEAR, SHOW, HELP , EXIT, or QUIT
CSR Address
Model
Node
772150
21
o
760334
21
1
760340
21
4
760344
21
5
7
------ KFQSA -----? help
Commands:
set KFQSA DSSI node number
SET <node> /KFQSA
enable a DSSI device
SET <node> <CSR address> <model>
disable a DSSI device
CLEAR <node>
show current configuration
SHOW
print this text
HELP
program the KFQSA
EXIT
don't program the KFQSA
QUIT
Parameters:
o to 7
<node>
760010 to 777774
<CSR_address>
21 (disk) or 22 (tape)
<model>
? set 6 /kfqsa
? show
CSR Address
Model
Node
o
772150
21
1
760334
21
4
760340
21
5
760344
21
6
------ KFQSA -----? exit
Programming the KFgSA .••
»>
»>SE'r LABGtJAGE 5
»>
»>SE'r BAI.'r RESTAR'r
»>

KA67SIKA680/KA690 Armware Commands

A-27

A.2.1S SHOW
The SHOW command displays the console parameter you specify.
Format:

SHOW {parameter}
Parameters:
BFLAG

Displays the default R5 boot flags.

BOOT

Displays the default boot device.

CONTROLP

Shows the current state of Control-P halt recognition, either Enabled
or Disabled.

DEVICE"

Displays all devices in the system.

HALT

Shows the user-defined halt action.

DSSI

Shows the status of all nodes that can be found on the DSSI bus. For
each node on the DSSI bus, the console displays the node number, the
node name, and the boot name and type of the device, if available. The
command does not indicate the "bootability" of the device.
The node that issues the command reports a node name of "*".
The device information is obtained from the media type field of the
MSCP command GET UNIT STATUS. In the case where the node is not
running or is not capable of running an MSCP server, then no device
information is displayed.

ETHERNET

Displays hardware Ethernet address for all Ethernet adapters that can
be found. Displays as blank if no Ethernet adapter is present.

LANGUAGE

Refer to the
Displays console language and keyboard type.
corresponding SET LANGUAGE command for the meaning.

MEMORY

Displays main memory configuration board by board.
IFULL-Additionally, displays the normally inaccessible areas of
memoty, such as the PFN bitmap pages. the console scratch memoty
pages, the Q22-bus scatter-gather map pages. Also reports the
addresses of bad pages, as defined by the bitmap.

A-28

KA675/KA680/KA690 CPU

QBUS

Displays all Q22-bus I/O addresses that respond to an aligned word
read, and speculative device name information. For each address, the
console displays the address in the VAX. I/O space in hex, the address
as it would appear in the Q22-bus I/O space in octal, and the word data
that was read in hex.
This command may take several minutes to complete. Press IcTRLIC I to
terminate the command. During execution, the command disables the
scatter·gather map.

RECALL

Shows the current state of command recall, either ENABLED or
DISABLED.

RLV12

Displays all RLOI and RL02 disks that appear on the Q22-bus.

UQSSP

Displays the status of all disks and tapes that can be found on the
Q22-bus that support the UQSSP protocol. For each such disk or
tape on the Q22-bus, the firmware displays the controller number, the
controller CSR address, and the boot name and type of each device
connected to the controller. The command does not indicate whether
the device contains a bootable image.

This information is obtained from the media type field of the MSCP
command GET UNIT STATUS. The console does not display device
information if a node is not running (or cannot ron) an MSCP server.

SCSI

Shows any SCSI devices in the system (TLZ04 or RRD4O-series.)

TRANSLATION

Shows any virtual addresses that map to the specified physical address.
The firmware uses the cu.rrent values of page table base and length
registers to perform its search; it is assumed that page tables have
been properly built.

VERSION

Displays the current firmware version.

Qualifiers: Listed in the parameter descriptions above.
Examples:
»>
»>SHOW ~
00000220

»>
»>SHOW BOOT

DOAO
»>SHOW CORTROLP
»>
»>SHOW DEVICE

KA680-A Vn.n VMBn.n

KA675/KA680IKA690 Rrmware Commands

A-29

DSSI Bus 0 Node 0 (R7CZZC)
-DIAO (RF71)
DSSI Bus 0 Node 1 (R7ALUC)
-DIAl (RF71)
DSSI Bus 0 Node 2 (R7EB3C)
-DIA2 (RF71)
DSSI Bus 0 Node 6 (*)
DSSI Bus 1 Node 7 (* )
SCSI Adapter 0 (761300), SCSI ID 7
-DKA100 (DEC TLZ04)
Ethernet Adapter
-EZAO (08-00-2B-OB-29-14)

»>
»>SHOW DSSI
DSSI Bus 0 Node 0 (R7CZZC)
-DIAO (RF71)
DSSI Bus 0 Node 1 (R7ALUC)
-DIAl (RF71)
DSSI Bus 0 Node 2 (R7EB3C)
-DIA2 (RF71)
DSSI Bus 0 Node 6 (* )
DSSI Bus 1 Node 7 (* )

»>
»>SBOW E'l'BERRET
Ethernet Adapter
-EZAO (08-00-2B-OB-29-14)

»>
»>SBOW HALT
restart

»>
»>SBOW LAlfGUAGE
English (United States/Canada)

»>
»>SHOW MEMORY
Memory 0: 00000000 to 01EEEEEE, 32MB, 0 bad pages,
Memory 0: 02000000 to 03FFFFFF, 32MB, 0 bad pages
Total of 64MB, 0 bad pages, 128 reserved pages

»>
»>SHOW MEHORY/FOLL
Memory 0: 00000000 to 01FFFFFF, 32MB, 0 bad pages
Memory 0: 02000000 to 03EEEEEE, 32MB, 0 bad pages
Total of 64MB, 0 bad pages, 128 reserved pages
Memory Bitmap
-00FF3COO to OOFF3FFF, 8 pages
Console Scratch Area
-OOFF4000 to OOFF7FFF, 32 pages

A-30

KA6751KA6801KA690 CPU

Q-bus Map
-OFF8000 to OFFFFFF, 64 pages
Scan of Bad Pages
»>
»>SBOK gaos
Scan of Qbus I/O Space
-20001920 (774440) = FF08 DELQA/DESQA
-20001922 (774442) = FFOO
-20001924 (774444) = FF2B
-20001926 (774446) = FF08
-20001928 (774450) = FFD7
-2000192A (77(452) = FF41
-2000192C (774454) = 0000
-2000192E (774456) = 1030
-20001F40 (777500) = 0020 IPCR
Scan of Qbus Memory Space
»>
»>SBOK RLV12
»>
»>SBOW SCSI

SCSI Adapter 0 (761300), SCSI ID 7
-DKA100 (DEC TLZ04)
»>
»>SBOW TRANSLATION 1000
V 80001000
»>
»>SBOW ogSSP
UQSSP Disk Controller 0 (772150)
-DUAO (RF30)
UQSSP Disk Controller 1 (760334)
-DUB1 (RF30)
UQSSP Disk Controller 2 (760340)
-DOC4 (RF30)
UQSSP Disk Controller 3 (760344)
-DUDS (RF30)
»>
»>
»>SBOW VERSION
KA680-A Vn.n VMBn.n
»>

KA67SIKA680/KA690 Firmware Commands

A-31

A.2.16 START
The START command starts instruction execution at the address you
specify. If no address is given, the current PC is used. If memory mapping
is enabled, macro instructions are executed from virtual memory, and the
address is treated as a virtual address. The START command is equivalent
to a DEPOSIT to PC, followed by a CONTINUE. It does not perform a
processor initialization.
Fonnat:
START [(address)]
Arguments:
[address]

The address at which to begin execution. This address is loaded into the user's

PC.

Example:
»>STAR.T 1000

A.2.17 TEST
The TEST command invokes a diagnostic test program specified by the test
number. If you enter a test number of 0 (zero), all tests allowed to be
executed from the console terminal are executed. The console accepts an
optional list of up to five additional hexadecimal arguments.
Refer to Chapter 5 for a detailed explanation of the diagnostics.
Format:

TEST [(test_number) [(test_arguments)]]
Arguments:
A two-digit hex number specifying the test to be executed.
Up to five additional test arguments. These arguments are accepted,
but they have no meaning to the console.

Example:
»>'l'EST 0

66 •• 65 •. 64 •• 63 •• 62 •• 61 •• 60 .• 59 •• 58 •• 57 •. 56 .• 55 •. 54 •• 53 •• 52 •• 51 ••
50 .. 49 •• 48 •• 47 .• 46 •. 45 .• 44 •• 43 •. 42 •• 41 •• 40 •• 39 •• 38 •• 37 .• 36 •. 35 •.
34 •• 33 •• 32 •• 31 .• 30 •• 29 .• 28 .• 27 •• 26 •• 25 .• 24 .. 23 •• 22 •. 21 .• 20 •• 19 ••
18 •• 17 •. 16 •• 15 •• 14 •• 13 •• 12 .• 11 •• 10 •• 09 •• 08 •• 07 •• 06 •• 05 .• 04 •• 03 ••

A-32 KA6751KA680/KA690 CPU

A.2.18 UNJAM
The UNJAM command performs an I/O bus reset, by writing a 1 (one) to
IPR 55 (decimal).
Format:
UNJAM

Example:
»>ONJ'AM

»>

A.2.19 X-Binary Load and Unload
The X command is for use by automatic systems communicating with the
console.
The X command loads or unloads (that is, writes to memory, or reads from
memory) the specified number of data bytes through the console serial line
(regardless of console type) starting at the specified address.
-Format:
X {address) {count} CR (line_checksum) {data) {data_checksum)
If bit 31 of the count is clear, data is received by the console and deposited
into memory. If bit 31 is set, data is read from memory and sent by the
console. The remaining bits in the count are a positive number indicating
the number of bytes to load or unload.
The console accepts the command upon receiving the carriage return.
The next byte the console receives is the command checksum, which is
not echoed. The command checksum is verified by adding all command
characters, includin.g the checksum and separating space (but not including
the terminating carriage return, rubouts, or cbaractei'S deleted by ,uhout),
into an 8-bit register initially set to zero. If no errors occur, the result is
zero. If the command checksum is correct, the console responds with the
input prompt and either sends data to the requester or prepares to receive
data. If the command checksum is in error, the console responds with an
error message. The intent is to prevent inadvertent operator entry into a
mode where the console is accepting characters from the keyboard as data,
with no escape mechanism possible.
If the command is a load (bit 31 of the count is clear), the console responds
with the input prompt (»», then accepts the specified number of bytes
of data for depositing to memory, and an additional byte of received data
checksum. The data is verified by adding all data characters and the
checksum character into an 8-bit register initially set to zero. If the final

KA6751KA680/KA690 Rrmware Commands

A-33

content of the register is nonzero, the data or checksum are in error, and
the console responds with an error-message.
If the command is a binary unload (bit 31 of the count is set), the console
responds with the input prompt (»», followed by the specified number of
bytes of binary data. As each byte is sent, it is added to a checksum register
initially set to zero. At the end of the transmission, the two's complement
of the low byte of the register is sent.
If the data checksum is incorrect on a load, or if memory or line errors occur
during the transmission of data, the entire transmission is completed, then
the console issues an error message. If an error occurs during loading, the
contents of the memory being loaded are unpredictable.

The console represses echo while it is receiving the data string and
checksums.
The console terminates all flow control when it receives the carriage return
at the end of the command line in order to avoid treating flow control
characters from the terminal as valid command line checksums.
You can control the console serial line during a binary unload using control
characters nCTRl..CI, ICTRUSI, ICTRlJOI, and so on). You cannot control the console
serial line during a binary load, since all received characters are valid
binary data.
The console has the following timing requirements:
•

It must receive data being loaded with a binary load command at a rate
of at least one byte every 60 seconds.

•

It must receive the command checksum that precedes the data within
60 seconds of the carriage return that terminates the command line.

•

It must receive the data checksum within 60 seconds of the last data
byte.

If any of these timing requirements are not met, then the console aborts
the transmission by issuing an error message and returning to the console
prompt.

The entire command, including the checksum, can be sent to the console
as a single burst of characters at the specified character rate of the console
serial line. The console is able to receive at least 4 Kbytes of data in a
single X command.

A-34

KA675/KA6801KA690 CPU

A.2.20 ! (Comment)
The comment character (an exclamation point)· is used to document
command sequences. It can appear anywhere on the command line. All
characters following the comment character are ignored.
Format: !
Example:
»>! The console ignores this line.
»>

KA67SIKA680/KA690 Firmware Commands

A-35

Appendix B

Address Assignments

B.1 KA675/KA680/KA690 General Local Address

Space Map

VAX Memory Space
Address Range
0000 0000

1FFF FFFF

Contents
Local Memory Space (512MB)

VAX I/O Space

------------Addres s Range

-----------------

2000 0000
2000 2000

2008 0000
2020 0000
2400 0000
2008 0000
2C08 0000

2000 1FFF

- 2003 FFFF
- 201F FFFF
- 23FF FFFF
-

27FF FFFF
2BFF FFFF
2FFF FFFF

3000 0000 - 303F FFFF
3040 0000 - 33FF FFFF
3400 0000 - 37FF FFFF
3800 0000 - 3BFF FFFF
3COO 0000 - 3FFF FFFF
E004 0000 - E007 FFFF

Contents
Local Q22-Bus I/O Space (8KB)
Reserved Local I/O Space (248KB)
Local Register I/O Space (1.5MB)
Reserved Local
Reserved Local
Reserved Local
Reserved Local

I/O Space
I/O Space
I/O Space
I/O Space

(62.5MB)
(64MB)
(64MB)

(64MB)

Local Q22-Bus Memory Space (4MB)
Reserved Local I/O Space (60MB)
Reserved Local I/O Space (64MB)
Reserved Local I/O Space (64MB)
Reserved Local I/O Space (64MB)
Local ROM Space

Address Assignments

B-1

B.2 KA675/KA680/KA690 Detailed Local Address

Space Map
Local Memory Space (up to 512MB)
0000 0000 - 1FFF FFFF
Q22-bus Map - top 32KB of Main Memory

VAX I/O Space

------------Local Q22-bus I/O Space
Reserved Q22-bus I/O Space
Q22-bus Floating Address Space
User Reserved Q22-bus I/O Space
Reserved Q22-bus I/O Space
Interprocessor Comm Reg
Reserved Q22-bus I/O Space

2000 0000 2000 0000 2000 0008 2000 0800 2000 1000 2000 1F40
2000 IF44 -

Local Register I/O Space

2000 2000 - 2003 FFFF

Reserved Local Register I/O Space
SHACl SSWCR
Reserved Local Register I/O Space
SHAC1 SSHMA
SHACl PQBBR
SHACl PSR
SHACl PESR
SHAC1 PFAR
SHAC1 PPR
SHAC1 PMCSR
Reserved Local Register I/O Space
SHACl PCQOCR
SHAC1 PCQICR
SHAC1 PCQ2CR
SHAC1 PCQ3CR
SHAC1 PDFQCR
SHAC1 PMFQCR
SHAC1 PSRCR
SHAC1 PECR
SHAC1 PDCR
SHAC1 PICR
SHACl PMTCR
SHAC1 PMTECR
Reserved Local Register I/O Space

B-2 KA675/KA6801KA690 CPU

2000 4000 2000 4030
2000 4034 2000 4044
2000 4048
2000 404C
2000 4050
2000 4054
2000 4058
2000 405C
2000 4060 2000 4080
2000 4084
2000 4088
2000 408C
2000 4090
2000 4094
2000 4098
2000 409C
2000 40AO
2000 40A4
2000 40A8
2000 40AC
2000 40BO -

2000 1FFF
2000 0007
2000 07FF
2000 OFFF
2000 1F3F
2000 IFFF

2000 402F
2000 4043

2000 407F

2000 422F

KA675/KA680/KA690 DETAILED LOCAL ADDRESS SPACE MAP (Cont.)
SHAC2 SSWCR
Reserved Local Register I/O Space
SHAC2 SSBMA
SHAC2 PQBBR
SHAC2 PSR
SHAC2 PESR
SHAC2 PFAR
SHAC2 PPR
SHAC2 PMCSR
Reser~led Local Register I/O Space
SHAC2 PCQOCR
SHAC2 PCQ1CR
SHAC2 PCQ2CR
SHAC2 PCQ3CR
SHAC2 PDFQCR
SHAC2 PMFQCR
SHAC2 PSRCR
SHAC2 PECR
SHAC2 PDCR
SHAC2 PICR
SHAC2 PMTCR
SHAC2 PMTECR
Reserved Local Register I/O Space

·2000 4230
2000 4234 - 2000 4243
2000 4244
2000 4248
2000 424C
2000 4250
2000 4254
2000 4258
2000 425C
2000 4260 - 2000 427F
2000 4280
2000 4284
2000 4288
2000 428C
2000 4290
2000 4294
2000 4298
2000 429C
2000 42AO
2000 42A4
2000 42A8
2000 42AC
2000 42BO - 2000 7FFF

NICSRO - Vector Add, IPL, Sync/Async
NICSRl - Polling Demand Register
NICSR2 - Reserved
NICSR3 - Receiver List Address·
NICSR4 - Transmitter List Address
NICSRS - Status Register
NICSR6 - Command and Mode Register
NICSR7 - System Base Address
NICSR8 - Reserved
NICSR9 - Wat~~dog Timers
NICSRlO- Reserved
NICSRll- Rev Num & Missed Frame. Count
NICSR12- Reserved
NICSR13- Breakpoint Address
NICSR14- Reserved
NICSR15- Diagnostic Mode & Status
Reserved Local Register I/O Space

2000 8000
2000 8004
2000 8008
2000 800C
2000 8010
2000 8014
2000 8018
2000 801C
2000 8020*
2000 8024*
2000 8028*
2000 802C*
2000 8030*
2000 8034*
2000 8038*
2000 803C
2000 8040 - 2003 FFFF

Address Assignments

B-3

Q-22 Bus Local Register I/O Space
DMA System Configuration Register
DMA System Error Register
DMA Master Error Address Register
DMA Slave Error Address Register
Q22-bus Map Base Register
Reserved Local Register I/O Space
Boot and Diagnostic Reg (32 Copies)
Reserved Local Register I/O Space

B-4 KA675/KA6801KA690 CPU

2008 0000 - 201F FFFF
(SCR) 2008 0000
(DSER) 2008 0004
(QBEAR)2008 0008
(DEAR) 2008 OOOC
(QBMBR)2008 0010
2008 0014 - 2008 3FFF
(BDR)2008 4000 - 2008 407C
2008 4080 - 2008 7FFF

KA675/KA680/KA690 DETAILED LOCAL ADDRESS SPACE MAP (Cont.)

Q22-bus Map Registers
CQMRs) 2008 8000 - 2008 FFFF
Reserved Local Register IIO Space
2009 0000 - 2013 FFFF
SSC CSRs

sse Base Address Register
sse Configuration Register

(SSCBR) 2014 0000
(SSCCR) 2014 0010
CP Bus Timeout Control Register (CBTCR) 2014 0020
Diagnostic LED Register
CDLEDR) 2014 0030
Reserved Local Register I/O Space
2014 0034 = 2014 006B
VAX IPRs implemented by NCA
Interval Clock Control Status Reg (ICCS) 2100 0060
Next Interval Count Register
(NICR) 2100 0064
Interval Count Register
(ICR)
2100 0068
NMC CSRs
o-bit Data Registers

(MOORs) 2101 0000 - 2101 7FFF

Main Memory Co~figuration Reg 0 (MEMCONO) 2101 8000
2101 8004
Main Memory Configuration Reg 1
2101 8008
Main Memory Configuration Reg 2
2101 800C
Main Memory Configuration Reg 3
2101 8010
Main Memory Configuration Reg 4
2101 8014
Main Memory Configuration Reg 5
2101 8018
Main Memory Configuration Reg 6
Main Memory Configuration Reg 7 (MEMCON7) 2101 801C
Main Memory Signature Register 0 (MEMSIGO) 2101 8020
2101 8024
Main Memory Signature Register 1
2101 8028
Main Memory Signature Register 2
2101 S02e
Main Memory Signature Register 3
2101 8030
Main Memory Signature Register 4
2101 8034
Main Memory Signature Register 5
2101 8038
Main Memory Signature Register 6
Main Memory Signature Register 7 (MEMSIG7) 2101 803C
2101 8040
Main Memory Error Address Register (MEAR)
2101 8044
Main Memory Error Status Register (MESR)
(MMCDSR) 2101 8048
Main Memory Mode Control and
Diagnostic Register
O-bit Address and Mode Register

(MOAMR) 2101 804C

Address Assignments

B-5

NCA CSRs
Error Status Register
("CESR)
2102 0000
Mode Control and Diagnostic Reg (CMCDSR) 2102 0004
CP1 Slave Error Address Register (CSEAR1) 2102 0008
CP2 Slave Error Address Register (CSEAR2) 2102 OOOC
CP1 IO Error Address Register
(CIOEAR1) 2102 0010
CP2 IO Error Address Register
(CIOEAR2) 2102 0014
NDAL Error Address Register
(CNEAR)
2102 0018
Local UVROM Space
VAX System Type Register (In ROM)
Local UVROM - (Halt Protected)

B-6 KA675/KA6801KA690 CPU

E004 0000 - E007 FFFF
E004 0004
E004 0000 - E007 FFFF

**********************************************************************
The following addresses allow those KA690 In~~rnal Processor
Registers that are implemented in the SSC chip (External, Internal
Processor Registers) to be accessed via the local I/O page. These
addresses are documented for diagnostic purposes only and should
not be used by non-diagnostic programs.
Time Of Year Register

2014 006C

Console Storage Receiver Status
Console Storage Receiver Data
Console Storage Transmitter Status
Console Storage Transmitter Data
Console Receiver Control/Status
Console Receiver Data Buffer
Console Transmitter Control/Status
Console Transmitter Data Buffer
Reserved Local Register I/O Space

2014 0070*
2014 0074*
2014 0078*
2014 007C*
2014 0080
2014 0084
2014 0088
2014 008C
2014 0090 - 2014 OODB

I/O Bus Reset Register
Reserved Local Register I/O Space

2014 aODC
2014 OOEO

Reserved Local Register I/O Space

2014 OOFC - 2014 OOFF

* These registers are not fully implemented, accesses yield
UNPREDICTABLE results.
**********************************************************************
Local Register I/O Space (Cont.)
Timer 0 Control Register
Timer o Interval Register
Timer a Next Interval Register
Timer a Interrupt Vector
Timer 1 Control Register
Timer 1 Interval Register
Timer 1 Next Interval Register
Timer 1 Interrupt Vector
Reserved Local Register I/O Space

2014 0100
2014 0104
2014 0108
2014 010C
2014 0110
2014 0114
2014 0118

2014 011C
2014 0120 - 2014 012F

BDR Address Decode Match Register
BDR Address Decode Mask Register
Reserved Local Register I/O Space

2014 0140
2014 0144
2014 0138 - 2014 03FF

Battery Backed-Up RAM
Reserved Local Register I/O Space

2014 0400 - 2014 07FF
2014 0800 - 201F FFFF

Reserved Local I/O Space

2020 0000 - 2FFF FFFF

Local Q22-bus Memory Space

3000 0000 - 303F FFFF

Reserved Local Register I/O Space

3040 0000 - 3FFF FFFF

Address AsSignments

B-7

8.3 External, Internal P19cessor Registers
Several of the Internal Processor Registers (lPRs) on the KA690 are
implemented in the NCA or SSC chip rather than the CPU chip. These
registers are referred to as External Internal Processor Registers and are
listed below.
IPR #

Abbrev.

===::a==========

====

Time of Year Register

TOY

28
29
30
31

Console Storage Receiver Status
Console Storage Receiver Data
Console Storage Transmitter Status
Console Storage Tra.."smitter Data

CSRS*
CSRD*
CSTS*
CSDB*

32
33
34
35

Console Receiver Control/Status
Console Receiver Data Buffer
Console Transmitter Control/Status
Console Transmitter ~ata Buffer

RXCS
RXDB
TXCS

I/O System Reset Register

IORESET

TXDB

* These registers are not fully implemented, accesses yield
UNPREDICTABLE results.

B.4 Global Q22-bus Address Space Map
Q22-bus Memory Space
Q22-bus Memory Space
Q22-bus I/O Space

(Octal)

0000 0000 - 1777 7777

(BBS7 ~sserted)

Q22-bus I/O Space (Octal)
Reserved Q22-bus I/O Space
Q22-bus Floating Address Space
User Reserved Q22-bus I/O Space
Reserved Q22-bus IIO Space
Interprocessor Comm Reg
Reserved Q22-bus IIO Space

B-8 KA6751KA6801KA690 CPU

1776 0000 1776 0000 1776 0010 1776 4000 1777 0000 1777 7500
1777 7502 -

1777 7777
1776 0007
1776 3777
1776 7777
1777 7477
1777 7777

B.5 Processor Registers
Table B-1: Processor Registers
Number
Register Name

Mnemonic(Dec) (Hex) Type Impl Cat

Kernel Stack Pointer

KSP

Ezecutive Stack Pointer

ESP

SuperYiIor Stack Pointer

SSP

User Stack Pointer
Int.ernapt Stack Pointer

NVAX

1-1

WAX

1-1

WAX

1-1

USP

NVAX

1-1

ISP

•

NVAX

1-1

BeIerIed

PO Sue Resister

POBR

NVAX

1-2

PO LeDstb Bqiater

POLR

NVAX

1-2

PI Base Begis&er

PIBR

NVAX

1-2

PI LeDgU1 !leJiater

P1LR

WAX

1-2

s,atem Bale Regm.er

SBR

WAX

1-2

s,atem LeDath Bqilter

SLR

NVAX

1-2

CPU IdeDtUlcation

CPU1D

NVAX

2-1

BaerYed

Proceu Ccmtz'Ol mock Base

PCBB

NVAX

1-1

Syatca Control Block Base

SCBB

NVAX

1-1

Interrupt Priority !.eYe}1

IPL

NVAX

1-1

ASTLeYel1

ASn.VL

WAX

1-1

Software Interrupt Bequest Begi&ter

SIBR

WAX

1-1

Software IntemJpt Summay ~

SISR

NVAX

1-1

~23

BaerYed

~d.

ElOOOOl.

ElOOOO3C

ElOOOOS8

llnitialized on reset

Address Assignments

B-9

Table B-1 (Cont.): Processor1teglsters
Number
tOAdess

Mnemonic(J)ec) (Hex) Type Impl Cat

InterYal Coanter Cont:ollStatul

ICCS

NCA

2-7

El000060

Next Interval Count

NICR

NCA

3-7

El000064

InIel'YBl Count

ICR

NCA

3-7

El000068

Time o!Year Bqi.ater

TODa

sse

2-3

ElOOOO6C

Conaole Storap BeceiYer Stat.us

CSRS

sse

2-3

El000070

Conaole Stonp ReceiYer Data

CSRD

sse

2-3

El000074

~. Storap TnInaIitter Status

CSTS

J1.W

sse

2-3

El000018

Coaaole Storqe Transmit.ter Data

CSTD

sse

2-3

E100001C

Couole BeI:eiftr CoatrollStawa

RXCS

J1.W

sse

2-3

El000080

CoIIMle BeI:eiftr Data BuB'er

RXDB

sse

2-3

El000084

~. 'l'nuIamitter Cont.ftlVStatUl

TXCS

sse

2-3

El0000S8

Console 'l'ranamitter Data Buffer

TXDB

sse

2-3

El00008C

BaerYed

El000090

Resened

El000094

Reeerved

ElOOOO9C

Resened

ElOOOOAO

Reserved

El0000A4

Machine Check Error Register

MCESR

NVAX

2-1

Console SaYed PC

SAVPC

NVAX

2-1

Console SaYed PSL

SAVPSL

NVAX

2·1

44-01

Reserved

I/O System Reset Rqiste.r

8-10

IORESET

KA675/KA680/KA690 CPU

sse

El0000BO

2-3

ElOOOODC

Table B-1 (Cont.): Processor Registers
Number
~d-

Mnemonic(Dec) (Hex) Type Impl Cat

Memory Management Enahle1t3

MAPEN

NVAX

1-2

Trmalation Buffer Invalidate All3

TBIA

NVAX

1-1

Trmalation Buffer Invalidate Sinsle3

TBIS

NVAX

1-1

BuerYed

El0000EC

Belerved

El0000FO

Syatem ldentiication

NVAX

1-1

Trmalation Buffer Check

TBCHK

NVAX

1-1

IPL 14 Interrupt Acx.5

IAIO."

64:

sse

2-3

El000100

IPL 15 Interrupt A~

IAIO.5

sse

2-3

El000104

IPL 16 Interrupt A~

IAKl6

sse

2-3

El000108

IPL 17 Interrupt A~

IAIO.7

-i3

sse

2-3

E100010C

Clear Write BufferS

CWB

sse

2-3

El000110

Beaerved

69-99

El000114

Baerved for'VM

100

E1000190

RearYedfor'VM

101

El000194

Be.erYed for' VM

102

El000198

ReIerYed

103-121

El00019C

Interrupt Syat.e S&atua Register

INTS'lS

122

'IA

NVAX

2-1

FerfonuDce MoDitoriDl' Facility Co1mL

PMFCNT

123

'lB

NVAX

2-1

Patchable CcIIltrol Store CcIIltrol Register

PCSCR

124

'IC

NVAX

2-1

Eboz CcIIltrol Becia&er

Eell

125

'7D

NVAX

2-1

MbaIt TB Tq FillS

KTBTAG

126

NVAX

2-1

1Initialized on reset
3Change broadcast to vector unit if present

STestability and diaguQsti~ ~c;.e only; not fQr software '!!!!e in normAl operat-in!l

Address Assignments

8-11

Table B-1 (Cont.): Processc:ir.Reglsters
Number
Register Name

Mnemonic(l)ec) (Hex) 1YPe Impl Cat

Mbax TB PrE FillS

M.TBPTE

121

'TF

NVAX

2-1

Cbox Control Begiater

CCTL

160

NVAX

2-5

161

NVAX

2-6

ReaerYecl
Bcache Data ECC

BCDECC

162

NVAX

2-5

Bcache Error Tag Statu.

BCETSTS

163

NVAX

2-5

Bcache Error Tag b2dex

BCETIDX

164

NVAX

2-5

Bcache Error Tag

BCETAG

165

NVAX

2-5

Bcache Error Data Status

BCEDSTS

166

NVAX

2-5

Bcache Error Data Index

BCEDmX

167

NVAX

2-5

Bcache Error ECC

BCEDECC

168

ItS

NVAX

2-5

Raenecl

169

NVAX

2-6

ReaerYecl

170

NVAX

2-6

Fill Error Addreu

CEFADR

111

NVAX

2-5

Fill Error Statu

CEFSTS

172

NVAX

2-6

173

NVAX

2-6

114

NVAX

2-5

175

Itl'

NVAX

2-6

116

NVAX

2-5

111

NVAX

2-6

118

NVAX

2-5

119

NVAX

2-6

180

NVAX

2-5

ReaerYecl
NDAL Error Status

NESTS

Raenecl
NDAL Error Output Address

NEOADR

ReaerYecl
NDAL Error Output Command

NEOCMD

Besened

NDAL Error Data High

NEDA.TBI

5Testability and diagnostic use only; not for soaware use in normal operation

8-12

KA675/KA680/KA690 CPU

lfr°essAd-

Table B-1 (Cont.): Processor Registers
Number
Register Name

Mnemonic(1)ec) (Hex) Type Impl Cat

Reaerved
NDAL Error Data Low

NEDATLO

BeterYed
NDAL Error Input Command

NEICMD

181

182

183

184

185-207 B9

Reserved

NVAX

2-6

NVAX

2-5

NVAX

2-6

NVAX

2-S

NVAX

2-6

VIC Memory Addraa Register

V1dAR

NVAX

2-5

VIC Tag Register

VTAG

2D9

NVAX

2-5

VIC Data Register

VDATA

210

IN{

NVAX

2-5

!box ConUol aDd Status Begister

ICSR

211

NVAX

2-5

!box Branch Prediction Control Regiat,erS

BPCR·

212

NVAX

2-S

213

NVAX

2-6

BeterYed

!box BackDp peS

BPC

214

NVAX

2-S

!box Backup PC with KLOG Un.;mS

BPCUNW

21S

NVAX

2-S

NVAX

2-6

216-223 D8

BeIened

Kbaz PO Bale BepaurS

MPOBR

224

NVAX

2-S

Kbaz PO Length Regil~

MPOLK

22S

NVAX

2-6

Kbaz PI Bale Begisur5

MPlBR

226

JlW

NVAX

2-6

Kbaz PI Length RegisurS

MPlLK

227

NVAX

2-S

MSBIl

?2S

NVAX

2-6

Kbaz SJStem !.eDIth BesistezS

KSLR

229

NVAX

2-5

KbCIIt Memozylianqement EnableS

IOW'EN

230

JlW

NVAX

2-5

Kbaz Phylical Addreu Mode

PAKODE

231

NVAX

2-5

KbCIIt MKE Addn=a

KKEADR

232

NVAX

2-5

Kbaz System Sue

Regiar.er5

~d.

5tJ'estability and diagnostic use only; not for sobare use in normal operation

Address As~ignments

8-13

Table B-1 (Cont.): Processor- Registers
Number
Register Name

Mnemomc(l)ee) <Hex} Type Impl Cat

Mbalt MME PTE Address

MMEPTE

233

NVAX

2-5

MbGX!OIE Status

MMESTS

234

NVAX

2-5

235

NVAX

2-6

RaerTed

:MbaIt TB Parity Addresa

TBADR

Z36

NVAX

2-5

Mbcm TB Parity Status

TBSTS

237

NVAX

2-5

ReIerw:d

238

NVAX

2-6

Resened

239

NVAX

2-6

Baened

240

NVAX

2-6

llaerftcl

241

NVAX

.2-6

242

NVAX

2-5

243

NVAX

2-6

241

NVAX

2-6

BeIened

245

NVAX

2-6

ReIerw:d

246

NVAX

2-6

Resened

247

NVAX

2-6

248

NVAX

2-6

Reserved

249

NVAX

2-6

Reeerved

2SO

NVAX

2-6

ReterYeCi

251

NVAX

2-6

ReterYed

252

NVAX

2-6

Raerved.

2S3

NVAX

2-6

ReserYeCi .

2S4

NVAX

2-6

Raerved.

2S5

NVAX

2-6

MbGX PCIICbe Parity Add.nA

PCADR

Rnened
MbGX Pcacbe Statas

MbGX Pcacbe Control

PCSTS

PCCTL

8-14 KA675/KA680/KA690 CPU

~Ad.

Table B-1 (Cont.): Processor Registers
Number
Register Name
Unimplemented.

Mnem.onic(Dec) (Hex) Type Impl Cat
100-

110 Adaress

OOFFFFFF
01000000-

FFfFIlliii

Type:
R = Read~nly register
RW =Read-write register
W =Write-only register

Impl(emented):

NVAX = Implemented in the NVAX CPU chip
System = Implemented in the system environment
Vector = Implemented in the optional vector unit or its NDAL interface
Cat(egory), class-subclass, where:
class is one of:
1 =Implemented as per DEC standard 032
2 =NVAX-specific implementation which is unique or different from the DEC standard 032
implementation
3 =Not implemented internally; converted to I/O space read or write and passed to system
environment

subclass is one of:
1 =Processed as appropriate by Ebox microcode
2 =Converted to Mbox IPR number and processed via internal IPR command
3 = Processed by internal IPR command, then converted to IIO space read or write and
passed to system environment
4- = If virtual machine option is implemented, processed as in 1, othenrise as in 3
5 =Processed by internal IPR command
6 = May be block decoded; reference causes UNDEFINED behavior
7 =Full interval timer may be implemented in the system environment. Subset ICes is
implemented in NVAX CPU chip
8 = Converted to MFVP MSYNC

Address Assignments

8-15

8.6 IPR Address Space__Decoding
Table B-2: IPR Address Space Decoding
IPRAddress Range
Mnemonic2 (hex)
Contents

IPRGroup

OOOOOOOO.•OOOOOOFFI 256 individual IPRs.

Normal
Bcache Tag

BCTAG

01000000.. 011FFFEOl64k Bcache tag IPRs, each separated by 2O(hex) from the previous

one.
Bache Deallocate

BCFLUSH

01400000 •.015FFFEOl64k Bcaclie tag deallocate IPRs,
each separated by 20(hex) from

PCTAG

01800000••01SOlFE01 256 Pcache tag IPRs, 128 for each

the previous one.
Pcache Tag

Pcache set, each separated by
2O(hex) from the previous one.
Pcache Data Parity

PCDAP

OlCOOOOO ••OICOIFF81 1024 Pcache data parity 1PRs,
512 for each Pcache set, each
separated by 8(bex) from the
previous one.

lUnused fields in the IPR addresses for these groups should be zero. Neither hardware nor
microcode detects and faults on an address in which these bits are non-zero. Although noncontiguous address ranges are shown for these groups, the entire IPR address space maps into
one of these groups. If these fields are non-zero, the operation of the CPU is UNDEFINED.
2The mnemonic is for the first IPR in the block

Processor registers in all groups except the normal group are processed
entirely by the NVAX CPU chip and will never appear on the NDAL. This
is also true for a number of the IPRs in the normal group. IPRs in the
normal group that are not processed by the NVAX CPU chip are converted
into IJO space references and passed to the system environment via a read
or write command on the NDAL.
Each of the 256 possible IPRs in the normal group are of longword length,
so a 1-KB block of 110 space is required to convert each possible IPR to a
unique 110 space longword. This block starts at address E1000000 (hex).
Conversion of an IPR address to an I/O space address in this block is done
by shifting the IPR address left into bits <9:2>, filling bits <1:0> with zeros,
and merging in the base address of the block. This can be expressed by the
equation:

[0 ADDRESS = EIOOOOOO+ (IPR NUMBER * 4)

8-16 KA67S/KA680/KA690 CPU

Appendix C

ROM Partitioning
This section describes ROM partitioning and subroutine entry points that
are public and are guaranteed to be compatible over future versions of
the firmware. An entry point is the address at which any subroutine or
subprogram will start execution .

C.1 Firmware EPROM Layout
The KA6751KA6801KA690 has 512 Kbytes of FEPROM. Unlike previous
Q22-bus based processors, there is no duplicate decoding of the FEPROM
into halt-protected and halt-unprotected spaces. The entire FEPROM is
halt-protected.

ROM Partitioning

C-1

Figure C-1:

KA675IKA6801KA690 FEPROM Layout

20040000

Branch Instruction

20040006

System 10 Extension

20040008

PC$MSG_OUT_NOLF_R4

2004000C

CP$READ_WITH_PRMPT_R4

20040010

Rsvd Mfg L200 Testing

20040014

Def Boot Dev Dscr Ptr

2004001c

Def Boot Flags Ptr
Console, Diagnostic,
and Boot Code
EPROM Checksum
Reserved for Digital

2005F800

4 Pages Reserved
for Customer Use

2005FFFC
MLC>OO7698

The first instruction executed on halts is a branch around the System ID
Extension (SIE) and the callback entry points. This allows these public
data structures to reside in fixed locations in the FEPROM.
The callback area entry points provide a simple interface to the currently
defined console for VMB and secondary bootstraps. This is documented
further in the next section.
The fixed area checksum is the sum of longwords from 20040000 to the
checksum inclusive. This checksum is distinct from the checksum that the
rest of the console uses.
The console, diagnostic and boot code constitute the bulk of the firmware.
This code is field upgradable. The console checksum is from 20044000 to
the checksum inclusive.
The memory between the console checksum and the user area at the end of
the FEPROM is reserved for Digital for future expansion of the firmware.
The contents of this area is set to FF.

C-2

KA675/KA6801KA690 CPU

The last 4096 bytes of FEPROM is reserved for customer use and is not
included in the console checksum. During a PROM bootstrap with PRBO
as the selected boot device, this block is the tested for a PROM "signature
block".

C.1.1 System Identification Registers
The firmware and operating system software reference two registers to
determine the processor on which they are running. The first, the System
Identification register (SID), is a NVAX internal processor register. The
second, the System Identification Extension register (8IE), is a firmware
register located in the FEPROM.
C.1.1.1 PR$_SID (IPR 62)

The SID longword can be read from IPR 62 using the MFPR instruction.
This longword value is processor specific, however, the layout of this register
is shown in Figure C-2. A desciption of each field is provided in Table C-l.
Figure C-2: SID: System identification Register
2423

0807
Reserved

00
Version
Ml.O-OO7699

Table C-1: System Identification Register
Field

Name

RW Description

31:24

CPU_TYPE

I'D

CPU type is the processor specific identification code.

OA:CVAX
OB:R1GEL
13: NVAX
14: SOC

24:8

reserved

I'D

Reserved for future use.

7:0

VERSION

I'D

Version of the microcode.

ROM Partitioning

C-3

C.1.1.2 SIE (20040004)
The System Identification Extension register is an extention to the SID and
is used to further differentiate between hardware configurations. The SID
identifies which CPU and microcode is executing, and the SIE identifies
what module and firmware revision are present. Note, the fields in this
register are dependent on SID<31:24>(CPU_TYPE).

By convention, all VAX 4000 systems implement a longword at physical
location 20040004 in the firmware FEPROM for the SIE. The layout of
the SIE is shown in Figure C-3. A description of each field is provided in
Table C-2.
Figure 0-3:

SIE: System IdentHicatlon Extension (20040004)
2423

16 15

0807

Version

Variant
ML().OO7700

Table C-2: System Identification Extension
Field

Name

RW Description

31:24

SYS_TYPE

This field identifies the type of system for a specific processor.
01: Q22-bus single processor system.

23:16

VERSION

This field indentifies the resident version of the firmware
encoded as two hexadecimal digits. For example, if the
banner displays V5.0, then this field is 50 (hex).

15:8

SYS_SUB_TYPE

This field indenti.6.es the particular system subtype.
01:KA650
02: KA640
03: KA655
04: KA670
05: KA660
06:KA.680

07:KA690
OC: KA675
7:0

VARIANT

This field indentifies the particular system variant.

C-4 KA675/KA680iKA690 CPU

C.1.2 call-Back Entry Points
The firmware provides several entry points that facilitate I/O to the
designated console device. Users of these entry points do not need to be
aware of the console device type, be it a video terminal or workstation.
The primary intent of these routines is to provide a simple console device
to VMB and· secondary bootstraps, before operating systems load their own
terminal drivers.
These are JSB (subroutine as opposed to procedure) entry points located in
fixed locations in the firwmare. These locations branch to code that in tum
calls the appropriate routines.
All of the entry points are designed to run at IPL 31 on the interrupt stack
in physcial mode. Virtual mode is not supported Due to internal finnware
architectural restrictions, users are encouraged to only call into the haltprotected entry points. These entry points are listed in Table C-3.

Table C-3: call-Back Entry Points
20040008

CP$MSG_OUT..,NOLF_&4

2004000C

CP$READ_WTH..PRMPI'_ 20040010
R4

C.1.2.1 CP$GETCHAR_R4

This routine returns the next character entered by the operator in RO. A
timeout interval can be specified. If the timeout interval is zero, no timeout
is generated. If a timeout is specified and if timeout occurs, a value of 18
(CAN) is returned instead of normal input.
Registers RO,R1,R2,R3 and R4 are modified by this routine, aU others are
preserved.

i--------------------------------------------------------------i Usage with timeout:
movl
ttimeout in tenths of second,rO i
jsb
@iCP$GET:CHAR_R4 - ;
cmpb
rO, '''xIS
;
beql
timeout handler
i Input is in RO.

Specify timeout.
call routine.
Check for timeout.
Branch if timeout.

i--------------------------------------------------------------Usage without timeout:

ROM Partitioning

C-5

clrl
rO
jsb
@ICP$GET CHAR R4-; Input is in RO~

Specify no timeout.
Call routine.

i--------------------------------------------------------------C.1.2.2 CP$MSG_OUT_NOLF_R4

This routine outputs a message to the console. The message is specified
either by a message code or a string descriptor. The routine distinguishes
between message codes and descriptors by requiring that any descriptor
be located outside of the first page of memory. Hence, message codes are
restricted to values between 0 and 51lo
Registers RO,Rl,R2,R3 and R4 are modified by this routine, all others are
preserved.

i--------------------------------------------------------------; Usage with message code:
movzbl
jsb

tconsole message coderrO
@'CP$MSG:OOT_NOLF_R4

; Specify message code.
; Call routine.

;--------------------------------------------------------------; Usage with a message descriptor (position dependent).
movaq
jsb

5$,rO
@tCP$MSG_OOT_NOLF_R4

Specify address of desc.
: call routine.

5$:

.ascid

; Message with descriptor.

/This is a message/

i--------------------------------------------------------------; Usage with a message descriptor (position independent) •
pushab
pushl
movl
jsb
clrq

5$
t10$-5$
sp,rO
@tCP$MSG_OUT NOLF R4
(sp)+

5$:
10$:

.ascii

/This is a message/

Generate message desc.
on stack.
Pass desc. addr. in RO.
Call routine.
Purge desc. from stack.

; Message.

;---------------------------------------------------------------

C-6 KA675/KA680!KA690 CPU

C.1.2.3 CP$READ_WTH_PRMPT_R4

This routine outputs a prompt message and then inputs a character string
from the console. When the input is accepted, DELETE, CONTROL-U and
CONTROL-R functions are supported.
As with CP$MSG_OUT_NOLF_R4, either a message code or the address of
a string descriptor is passed in RO to specify the prompt string. A value of
zero results in no prompt. A time-out value in 10-millisecond ticks may be
passed in RI. If RI is zero, the prompt will not timeout.

A descriptor of the input string is returned in RO and Rl. RO contains the
length of the string and RI contains the address. This routine inputs the
string into the console program string buffer and therefore the caller need
not provide an input buffer. Successive calls however destroy the previous
contents of the input buffer.
Registers RO,RI are modified by this routine, all others are preserved.
;~~-~~-~~-~--------------------------------------------- -----

; Usage with a message descriptor (position independent).
pushab
pushl
movl
clrl
jsb
clrq

5$
110$-5$

5$:
10$:

.ascii

sp,rO
rl
@iCP$READ_WTH_PRMPT_R4
(sp) +

/Prompt> /

Generate prompt desc.
on stack.
Pass desc. addr. in RO.
Specify no time-out.
; Call routine.
Purge prompt desc.
Input desc in RO and R)
; Prompt string.

----------------------------------------------------------------

C.1.3 Boot Information Pointers
Two longwords located in FEPROM are used as pointers to the default boot
device descriptor and the default boot flags (Figure C-4), because the actual
location of this data may change in successive versions of the firmware. Any
software that uses these pointers should reference them at the addresses
in halt-protected space.

ROM Partitioning

C-7

Figure C-4: Boot Information Pointers

20040018

Def Boot Dev Dscr Ptr

Class

1
I

I I
Type

Oesc Length

Boot Device String Ptr

2004001 c

Del Boot Flags Ptr

~ ASCIZ Dev Name String

~---------I~NI Boot Flags (longword)
ML()'()o7701

The following macro defines the boot device descriptor format.

;------------------------------------------------------------; Default Boot Device Descriptor

boot device descriptor::
base = •
• = base + dsc$w length
.word
nvr$s_bOOt_device
• - base + dsc$b dtype
• byte
dsc$k_dtype_z

· = base + dsc$b class
. byte

dsc$k_class_z

· - base + dsc$a-POinter
. long
nvr_base + nvr$b_boot device
· - base + dsc$s_dscdefl

i-------------------------------------------------------------

C-8 KA675/KA6801KA690 CPU

Appendix D

Data Structures and Memory Layout
This appendix contains definitions of the key global data structures used
by the CPU firmware.

D.1 Halt Dispatch State Machine
The CPU halt dispatcher determines what actions the firmware will take on
halt entry based on the machine state. The dispatcher is implemented as a
state machine, which uses a single bitmap control word and the transition
(see Table D-1) to process all halts. The transition table is sequentially
searched for matches with the current state and control word If there is a
match, a transition occurs to the next state.
The control word comprises the following information.

•

BaIt Type, used for resolving external halts. Valid only if Halt Code
is 00.

000 : power-up state
001 : halt in progress
010: negation ofQ22-bus DCOK
011 : console BREAK condition detected
100 : Q22-bus BHALT
101 : SGEC BOOT_L asserted (trigger boot)
~

Halt Code, compressed form ofSAVPSL<13:8>(RESTART_CODE).
00 : RESTART_CODE = 2, external halt
01 : RESTART_CODE =3, power-up/reset
10: RESTART_CODE = 6, halt instruction
11 : RESTART_CODE = any other, error halts

•

Mailbox Action, passed by an operating system in CPMBX<l:O>(HALT_
ACTION).

00 : restart, boot, halt
01 : restart, halt
10 : boot, halt
11: halt

Data Structures and Memory Layout

0-1

•

User Action, specified with the SET HALT console command.
000 : default
001 : restart, halt
010 : boot, halt
011 : halt
100 : restart, boot, halt

•

HEN, Break (halt) EnablelDisable switch, BDR<07>

•

ERR, error status

•

TIP, trace in progress

•

DIP, diagnostics in progress

•

BIP, bootstrap in progress CPMBX<2>

•

RIP, restart in progress CPMBX<3>

A transition to a "next state" occurs if a match is found between the control
word and a "current state" entry in the table. The firmware does a linear
search through the table for a match. Therefore, the order of the entries
in the transition table is important. The control longword is reassembled
before each transition from the current machine state. The state machine
transitions are shown in Table D-1.

Table 0-1 : Firmware State Transition Table
Current
State

Nen
State

Mailbs

Balt

Type

Balt
Code

User

HEN·ERR·TIP·DIP·

Action Action BIP·BIP

Perform conditional initialization. 1

ENTRY

->RESET INIT

xxx

x-x-x-x-x-x

ENTRY

->BREAK INIT

011

xxx

x-x-x-x-x-x

ENTRY

->TRACE INIT

xxx

x-O-l-x-x-x

ENTRY

->OTHER INIT

xxx

x-x-x-x-x-x

Perform common initialization. 2
RESET1NIT

->mIT

xxx

x-x-x-x-x-x

1 Perform a unique initialization routine on entty. In particular, power-ups, BREAKs, and
TRACEs require special initialization. Any other halt entry performs a default initialization.
2 After performmg conditional initialization, complete common initialization.

0-2 KA675/KA680!KA690 CPU

Table 0-1 (Cont.): Firmware State Transltlqra Table
Current
State

Mailbs

Halt

State

Type

Code

BREAKINIT

->1NIT

xxx

User HEN·ERR·TIP·DIP·
Action Action DIP·RIP
xx
xxx
x-x-x-x-x-x

TRACEINIT

->lNIT

xxx

x-x-x-x-x-x

OTHERINIT ->INIT

xxx

x-x-x-x-x-x

Check for external halts. 3

INIT

->BOOTSTRAP

010

xxx

O-x-x-x-x-x

INlT

->BOOTSTRAP

101

:xx

xxx

x-x-x-x-x-x

INlT

->HALT

xxx

x-x-x-x-x-x

Check for pending (NEXT) trace. 4

INIT

->TRACE

xxx

x-x-1-x-x-x

TRACE

->EXlT

xxx

x-0-1-x-x-x

TRACE

->HALT

xxx

x-x-x-x-x-x

Check for bootstrap conditions. S

INlT

->BOOTSTRAP

xxx

0-0-0-0-0-0

INlT

->BOOTSTRAP

xxx

010

1-0-0-0-0-0

INlT

->BOOTSTRAP

xxx

100

1-0-0·0-0-0

INIT

->BOOTSTRAP

xxx

x-O-O-O-O-O

3 Halt Gil all er..emal halts, except:

if DCOK (unlikely) and halts are disabled, bootstrap
if SGEC remote trigger, bootstrap

" Unconditionally enter the TRACE state, if the TIP flag is set and the halt was due to a
HALT instruction. From the TRACE state the firmware exits, if TIP is set and ERR is clear;
otherwise it halts.
5 Bootstrap,
if power-up and halts are disabled.
if power-up and halts are enabled and user action is 2 or 4.
if not power-up and mailbox is 2.
if not power-up and mailbox is 0 and user action is 2.
if not power-up and restart failed and mailbox is 0 and user action is 0 or 4.

Data Structures and Memory Layout

D-3

Table 0-1 (Cont.): Firmware State Transition Table
Current
State

Next
State

INIT

M.aiIbx

Halt
Code

User BEN·ERR-TIP·DIP·
Action Action DIP·RIP

->BOOTSTRAP

Halt
Type
xxx

010

x-O-O-O-O-O

INIT

->BOOTSTRAP

xxx

100

x-0-0-O-0-1

INIT

->BOOTSTRAP

xxx

100

x-1-0-0-0-x

INIT

->BOOTSTRAP

xxx

000

0-0-0-0-0-1

RESTART

->BOOTSTRAP

xxx

000

0-1-0-0-0-x

Check for restart conditions. 6

INIT

->RESTART

xxx

x-O-O-O-O-O

INIT

->RESTART

xxx

001

x-O-O-O-O-O

INIT

->RESTART

xxx

100

x-O-O-O-O-O

INIT

->RESTART

xxx

000

0-0-0-0-0-0

Perform common exit processing, if
no errors. 7

BOOTSTRAP ->EXIT

xxx

x-O-x-x-x-x

RESTART

->EXIT

xxx

x-O-x-x-x-x

HALT

->EXlT

xxx

x-O-x-x-x-x

Exception transitions, just halt. 8

INIT

->HALT

xxx

x-x-x-x-x-x

BOOT

->HALT

xxx

x-x-x-x-x-x

REST

->HALT

xxx

x-x-x-x-x-x

HALT

->HALT

xxx

x-x-x-x-x-x

6 Restart the operating system if not power-up and,

if mailbox is 1.
if mailbox is 0 and user action is 1 or 4.
if mailbox is 0 and user action is 0 and halts are disabled.
7 Exit after halts, bootstrap or restart. The exit state transitions to program 110 mode.

8 Guard block that catches all exception conditions. In all cases, just halt.

D-4 KA675/KA6801KA690 CPU

Table 0-1 (Cont.): Firmware State Transition Table
Mailbx

Code

User BEN.ERR·TIP·DIP.
Action Action BIP·RIP

xxx

:xx

xxx

x-x-x-x-x-x

xxx

x-x-x-x-x-x

Current
State

Next
State

Halt

TRACE

->HALT

EXIT

->HALT

'JYpe

Halt

"x" is used in this table to indicate a "don't care" field.

0.2 RPB
VMB typically utilizes the low portion of memory unless there are bad
pages in the first 128Kbytes. The first page in its block is used for the RPB
(Restart Parameter Block), through which it commWlicates to the operating
system. Usually, this is page O.
VMB will initialize the Restart Parameter Block (RPB) as shown in
Table D-2.

Table 0-2: Restart Parameter Block Fields
(Rll)+Field Name

Description

00:

RPB$L..BASE

Physical address of base of RPB.

04:

RPB$L_RESTART

Cleared.

08:

RPB$L_CHKSUM

-1

OC:

RPB$L_RSTRTFLG

Cleared.

10:

RPB$LJiALTPC

RIO on entry to VMB (HALT PC).

10:

RPB$LJL\LTPSL

PR$_SAVPSL on entty to VMB <HALT PSL).

18:

RPB$L_HALTCODE

AP on entty to VMB (HALT CODE).

Data Strudures and Memory Layout

0-5

Table 0-2 (Cont.): Restart Parameter Block Fields
(Rll)+Field Name
IC:

RPB$L~OOTRO

Description
RO on entry to VMB.

NOTE: The field RPB$W_ROUBVEC, which
overlaps the high-order word of RPB$L_
BOOTRO, is set by the boot device drivers to
the 8CB offset (in the second page of the 8CB)
of the interrupt vector for the boot device.

20:

RPB$L~OOTRl

VMB version number. The high-order word of the version
is the major ID and the low-order word is the minor ID.

24:

RPB$L..B00TR2

R2 on entry to VMB.

28:

RPB$L_B00TR3

R3 on entry to VMS.

2C:

RPB$L_B00TR4

R4 on entry to VMB.

NOTE: The 48-bit booting node address is
stored in RPB$L_BOOTR3 and RPB$L_
BOOTR4 for compatibility with ELN VI.I
(this field is only initialized this way when
performing a network boot).
30:

RPB$L_B00TR5

R5 on entry to VMB.

34:

RPB$L_IOVEC

Physical address of boot driver's YO vector of transfer
addresses.

38:

RPB$LJOVECSZ

Size of BOOT QIO routine.

3C:

RPBSLJ'lLLBN

LBN of secondary bootstrap image.

40:

RPB$L_FILSIZ

Size of secondary bootstrap image in blocks.

D-6 KA675/KA680IKA690 CPU

Table 0-2 (Cont.): Restart Parameter Block_Fields
(lUl)+Field Name
44:

RPB$QYFNMAP

(

Description
The PFN bitmap is a array of bits, where each bit has
the value "1" if the corresponding page of memol'Y is
valid, or has the value "0" if the corresponding page
of memory contains a memory error. Through use of
the PFNMAP. the operating system can avoid memory
errors by avoiding mown bad pages altogether. The
memory bitmap is always page-aligned, and describes all
the pages of memory from physical page ##0 to the high
end of memory, but excluding the PFN bitmap itself and
the Q-bus map registers. If the high byte of the bitmap
spans some pages available to the operating system
and some pages of the PFN bitmap itself, the pages
corresponding to the bitmap itself will be marked as bad
pages. The:first longword of the PFNMAP descriptor
contains the Dumber of bytes in the PFNMAP; the second
longword contains the physical address of the bitmap.

Count of "good" pages of physical memol'Y. but not
including the pages allocated to the Q22-bus scatter
19ather map. the console scratch area, and the PFN
bitmap at the top of memory.

50:

RPB$L_SVASPr

54:

RPB$L_CSRPHY

Physical address of CSR for boot device.

58:

RPB$L_CSRVIR

5C:

RPB$L_ADPPHY

Physical address of ADP. (really the address of QMRs Ax800 to look like a UBA adapter).

o.
64:

RPB$W-'lJNIT

Unit number of boot device.

66:

RPB$B_DEVTYP

Device type code of boot device.

67:

RPB$B_SLAv£

Slave number of boot de...i~.

Data Structures and Memory Layout

0-7

Table 0-2 (Cont.): Restart Parameter Block Fields
(Rll)+Field Name

Description

68:

Name of secondazy bootstrap image (defaults to
[SYSO.8YSEXE]sySBOOT.EXE). This field (up to 40
bytes) is overwritten with the input string on a "solicit"
boot.

RPB$TYILE

NOTE: I : For VAX/VMS, the RPB$T_FILE

must contain the root directory string "SYSn. "
on a non-network bootstrap. This string is
parsed by SYSBOOT (SYSBOOT does not
use the high nibble of BOOTRS). 2: The
RPB$T_FILE is overwritten to contain the
boot node name for compatibility with ELN
VI.I (this field is only initialized this way
when performing a network boot).
Array (16 bytes) of adapter types (NDT$_UBO • UNIBUS
).

AO:

RPB$BJaDRPGCNT

Count of header pages.

AI:

RPB$W..BOOTNDT

Boot adapter nexus device type. Used by SYSBOOT
and INIADP (OF SYSLOA) to configure the adapter of
the boot device (changed from a byte to a word field in
Version 12 of VMB).

80:

RPB$L_SCBB

Physical address of seB.

Be:

RPB$L_MEMDSC

Count of pages in physical memory including both good
and bad pages. The high 8 bits of this longword contain
the TR #, which is always 0 for KA6751KA68OIKA690.

co:

RPB$LJdEMDSC+4

PFN of the first page of memolY. This field is always 0
for KA6751KA6801KA690, even if page #0 is a bad page.

NOTE: No other memory descriptors are

used.
Count of "bad" pages of physical memolY.

D-8 KA675/KA6801KA690 CPU

Table D-2 (Cont.): Restart Parameter BlocJ< Fields
(Rll)+Field Name

Description
Boot device controller number biased by 1. In VAXNMS,
this field is used by INIT (in SYS) to construct the boot
device's controller letter. A 0 implies this field has not
been initialized, else if initialized, A=I, B=2, etc. (this
field was added in Version 13 of VMB).

nn:

The rest of the RPB is zeroed.

0.3 VMB Argument List
The VMB code will also initialize an argument list as shown in Table D-a
(the address of the argument list is passed in the AP).

Table Eh1: VMS Argument List
(AP)+ Field Name

Description

04:

VMB$L_FlLECACHE

Quadword filename.

oc:

VMB$L_LOYFN

PFN of first page of physical memory (always 0,
regardless of where 128 Kbytes of "good" memozy starts).

10:

VMB$LJiIJ'FN

PFN of last page of physical memory.

14:

VMB$Q...PFNMAP

Descriptor ofPFN bitmap. First longword contains count
of bytes in bitmap. Second longword contains physic:al
address of bitmap. (Same rules as for RPB$Q...PFNMAP
listed above.)

lC:

VMB$(LUCODE

Quadword.

24:

VMB$B_SYSTEMID

48-bit (actually a quadword is allocated) booting node
address which is initialized when performing a network
boot. This field is copied from the Target System Address
parameter of the parameters message. (The DECnet
HIORD value is added if the field was two bytes.)

2C:

VMB$L_FLAGS

Set as needed.

30:

VMB$L-<!UlIPFN

Cluster interface high PFN.

Data Structures and Memory Layout

0-9

Table D-3 (Cont.): VMB Argument List
<AP>+ Field Name
Description
34:

VMB$'LNODENAME

Boot node name which is initialized when performing a
netwotk. boot. This field is copied from the Target System
Name parameter of the parameters message.

3C:

VMB$Q.jIOSTADDR

Host node address (this value is only initialized when
booting over the network). This field is copied from
the Host System Address parameter of the parameters
message.

44:

VMB$Q..BOSTNAME

Host node name (this value is only initialized when
performing a network boot). This field is copied from
the Host System Name parameter of the parameters
message.

olC:

VMBSQ...TOD

Time of day (this value is only initialized when
performing a network boot). The time of day is copied
from the first eight bytes of the Host System Time
parameter of the parameters message. (The time
differential values are NOT copied.)

54:

VMB$LJPARAM

Pointer to data retrieved from request of the parameter
file.

58:

The rest of the argument list is zeroed.

0-10 KA6751KA680/KA690 CPU

Appendix E

Configurable Machine State
The KA6751KA6801KA690 CPU module has many control registers that
need to be configured for proper operation of the module. The following list
shows the normal state of all configurable bits in the CPU module as they
are left after the successful completion of power-up ROM diagnostics.
VAX 4000 Models 400,500,600
Configuration registers and writable bits: (* - power up reset state)
NCA;
CMCDSR:

Mode Control and Diagnostic Status Register

(2102 0004)

15:14: CP2 MT Timer Prescaler
11 - 144000 cycles* - needed for COBlC lOms No Grant
timeout
13:12: CPl MT Timer Prescaler
00 - 144 cycles - minimum for passive releases, no
cycle should take longer than this
11:10: NOAL Timeout Prescaler
00 - 3200 cycles* - this is longer than both NCA and
NMC transactions timeouts, preserves timeout
order
9: 080S TRANS enable (formerly COBle PRESENT)

o - 080S TRANS signal disabled* - this is to avoid
08US_TRANS deadlock

8: 102 IO enable
1 - enableci

7: Force wrong CP2 bus parity
o - off* - diagnostic use only
6: Force wrong CPl bus parity
o - off* - diagnostic use only
5:
4:

Force wrong NOAL master parity

o - off* - diagnostic use only

Force wrong NOAL slave parity

o - off* - diagnostic use only

3: Enable prefetch
1 - enable CP bus prefetch on OMA reads
2: Force write buffer hit
o - off* - diagnostic use only

Configurable Machine State

E-1

1: Force CP2 bus owner
o - disabled - diagnostic use only
0: Force CP1 bus owner
o - disabled - diagnostic use only
ICCS:

Interval Clock Control and Status Register

(2100 0060)

NOTE: VMS sets ICCS, NICR to proper values
6: Interrupt enable
o - disabled*
5: Single step
o - off*

4: Transfer
o - disabled*
0: Run - increment every l~s
o - do not increment*
NICR:

Next Interval Count Register

(2100 0064)

31:0 Initial count value for ICR (FFFF08FO* (10ms»
NMC:
MEMCON_0-7: Memory Configuration Registers

(2101 8000 thru 2101 801C)

NOTE: Diagnostics set these registers based on available memory
31: Base Address Valid
o - not valid*
1 - valid
28:24: Base Address (0 on reset)
1MB RAM - all address bits used
4MB RAM - only <28:26> used
2:1 RAM size
00 - 1MB RAM*
01 - 1MB RAM
10 - 4MB RAM
11 - non-existent bank
0: Mode
1 - 64-bit mode
MMCDSR:
31:

Mode Control and Diagnostic Status Register
Fast Diagnostic Mode (FOM)

o - disabled* - diagnostic use only

30: FDM Second pass
o - disabled* - diagnostic use only
29: Diagnostic Checkbit mode
o - disabled* - diagnostic use only
28: QBus on 101
o - OBus on I02*
27: Enable soft error log (NeAL & memory related)
o & disabled* - VMS enables this

E-2 KA675!KA6801KA690 CPU

(2101 8048)

26: Flush BCache
o - don't flush*
24:17: Memory diagnostic check bits
o - meaningful only in diagnostic check mode* (mayor
may not be read as 0)
8:7: NDAL Timeout Scaler
00 - 2600 cycles* - maximum, to preserve timeout order
6: Disable memory error
o - memory errors deteted and corrected*
5:

Refresh interval timer select

o - 328 cycles* (Model 500,600)

1 - 244 cycles

(Model 400)

4:2: Force wrong parity on NDAL transactions
o - off* - diagnostic use only
1: Disable memory refresh
o - memory refreshed*
0: Force refresh
o - normal refresh*
MOAMR:

O-bit Address and Mode Register

(2101 a04C)

16: Ignore o-bit mode
o - o-bits checked*
15: Disable O-bit error
o - o-bit errors detected*
14:6: o-bit segment address (0*) - meaningful only during
O-bit data register access
5:3: O-bit mask (0*) - meaningful only during .o-bit data
register access
2:0: o-bit operation mode
000 - reconstruction mode* - me~ningful only during
o-bit data register access
MOOR:

o-bit Data Registers

(2101 0000 thru 2101 7FFF)

23:12: O-bit field 1 (0*) - used only during Fast Memory test
11:0: o-bit field 0 (0*) - used only during Fast o-bit test
mode
NVAX:

CPUID:

CPU ID Register

(IPR E)

7:0: CPU identifcation - 0 (for single processor config.)
SID:

System Identification Register

(IPR 3E)

NOTE: this register may only be written by microcode
31:24: CPU type - 13hex (NVAX code)
13:8: ~a~cn revision

Configurable Machine State

E-3

7:0: Microcode revision
ICSR:

IBox Control and Status Register (IPR D3)

0: VIC enable
o - disabled* (Model 400)
1 - enabled
(Models 500,600)
ECR:

EBox Control Register

(IPR 70)

13: FBox test enable
o - disabled* - diagnostic use only
7: Interval time mode
1 - full CPU implemented interval timer
5: S3 stall timeout
o - counts cycles wi timeout_enable asserted* (-3 sec)
3: FBox stage 4 bypass
1 - enabled - result from stage 3 passed directly to
rBox output interface (improves rBox latency)
2: 53 external time base timeout
o - disabled* - use internal time base
1: FBox enable
1 - enabled
0: Vector present
o - no* - no vector option available at this time
MHAPEN:
0:

Memory Map Enable Register

(IPR E6)

Memory map enable

o - disabled* - VMS enables this

PAMODE:

Physical Address Mode Register

(IPR E7)

0: Physical address mode
o - 30-bit physical address space*
PCCTL:

PCache Control Register

(IPR FB)

8: PCache Electrical disable
o • PCache enabled*
7:5 MBox performance monitor mode
o - diagnostic use only·
4: PCache error enable
1 • enables PCache error detection
3: Bank select during force hit mode
o • left bank selected if force hit mode enabled- diagnostic use only
2: Force hit
o - disabled· - diagnostic use only
1: I enable
1 ·-enable pcache for IREAD, INVAL, I_CF commands

E-4 KA675!KA680IKA690 CPU

0: D_enable
1 .. enable PCache for lNVAL, D-stream read/write/fUJ
commands
CCTL:

CBox Control Register

(lPR AO)

30: Software ETM
a - disabled* - diagnostic use only
16: Force NDAL parity error
o - off* - diagnostic use only
15:11: Performance monitoring BCache access and hit type
o - configures BCache for performance monitoring* meaningful only during performance monitoring
10: Disable CBox·write packer
o - write packer enabled* - improves write latency
9: Read timeout counter test
o - test disabled* - use external time base for read
timeout counter
8: So ft ware ECC
o - use correct ECC*
7: Disable BCache errors
a - BCache errors detected*
6:

Force Hit

o - disabled* - diagnostic use only

5:4: BCache size
00 - 128 KB* ~ode1s 400,500)
10 - 512 KB ~odel 600)
3:2: Data store speed
00 - 2 cycle read, 3 cycle write· (Model 600)
01 - 3 cycle read, 4 cycle write (Model 500)
10 - 4 cycle read, 5 cycle write (Model 400)
1: Tag store speed
o - 3 cycle read, 3 cycle write* (Model 600)
1 - 4 cycle read, 4 cycle write (Models 400,500)
0: Enable Bcache
1 - enabled
COBlC:
SCR:

System Configuration Register

(2008 0000)

14: Halt enable
1 - BHALT to COBlC HALTIN pin to cause halts
12: Page prefetch disable
1 - map prefetch disabled - historical latency reasons
7: Restart enable
a - QBus restart causes ARB power-up reset*
3:1:

lCR offset address select bits

o - no effect (ACX mode not supported) *

Configurable Machine State

E-5

ICR:

Interprocessor Communication Register

(2000 1F40)

8: AOX Halt
o - no halt (AOX mode not supported)
6: ICR interrupt enable
o - interprocessor interrupts disabled - only
uniprocessor eonfig. allowed
5: Local memory external access enable
o - external access disabled* - VMS will configure map
QBMBR:

Q-Bus Map Base Address Regi$ter (2008 0010)

28:15:

address where 8K QBus mapping registers are located
reconfigures map)

SHAC:
NOTE:
PQBBR:

all SHAC registers are subsequently configured by VMS driver
Port Queue Block Base Register

(2000 4048)

20:0: upper bits of physical address of base of Port Queue
block. Contains HW version, FW version, shared host
memory version and CI port maintenance IC at power-up.
PPR:

Port Parameter Register

31:29: Cluster size.

(2000 4058)

For SHAC value -

28:16: Internal buffer length - 0* (For SHAC value - 1010 hex)
7:0: Port number.
PMCSR:

Same as SHAC's oSSI Io.

Port Maintenance Control and Status Register

(2000 405C)

2: Interrupt enable
o - disabled*
1:

Maintenance timer disable

o - enabled*

SGEC:
NOTE:
NICSRO:

all SGEC registers are susequently configured by VMS driver
Vector Address, IPL, Synch/Asynch Register

31:30: Interrupt priority
00 - 14*
29: Synch/Asynch bus master operating mode
o - asynchronous*
15:0: Interrupt vector - 0003hex*

E-6 KA675!KA680IKA690 CPU

(2000 BOOO)

NICSR6:

Command and Mode Register

(2000 8018)

30: Interrupt enable
o - disabled*
28:25: Burst limit mode
maximum number of longwords transferred in a single
DMA burst. 1*,2,4,8 when NICSR<l9>is clear;
1*,4 when set.
20: Boot message enable mode
o - disabled*
19: Single cycle enable mode
o "" disabled*
11: Start/Stop transmission command
o - SGEC transmission process in stopped state*
10: Start/Stop reception command
o - SGEC reception process in stopped state*
9:8: Operating mode
00 a normal mode*
7: Disable data chaining mode
o - frames too long for current receive buffer will De
transferred to the next buffer(s) in receive list*
6: Force collision mode (internal loopback mode only)
o - no collision·
3: Pass bad frames mode
o - bad frames discarded2:1: Address filtering mode
00 - normal modeS
NICSR7:

System Base Register

(2000 SOlC)

------------------29:0: System base addr.ss - physical starting address of
the VAX system page table (unpredictable after reset)

NICSR9:

Watchdog T1mers Register

(2000 8024)

31:16: Receive w~tChdo; ~!~eou~
o - never timeout·
defaUlt - 1250 • 2 ms
range
72 ps (45) to 100 ms
3

15:0: Transmit watchdog timeout
o - never timeout·
default - 1250 • 2 ms
range - 72 ps (45) to 100 ms
SSC:
SSCBAR:

SSC Base Address Register

(2014 0000)

29:0 20140000 - Base address·

Configurable Machine State

E-7

SSCCR:

SSC Configuration Register

(2014 0010)

27: Interrupt vector disable
o - interrupt vector enabled*
25:24: IPL Level
00 - 14*
23: ROM access time
o - 350 ns*
22:20: ROM size
101 - 256KB
18:16: Halt protected space
101 - 20040000 - 2007FFFF (historical)
15:

Control P enable

o - 20 spaces recognized as break*, not control-p
(historical)

14:12: Terminal UART baud rate
101 - 9600 (historical)
6: Programmable address strobe 1 ready enable (for BDR)
1 - ready asserted after address strobe
5:4: Programmable address strobe 1 enable (for BOR)
11 - read enabled, write enabled
2: Programmable address strobe 0 ready enable
o - no ready after address strobe* - not used
1:0: Programmable address strobe 0 enable
00 - read disabled, write disabled* - not used
RXCS:

Console Receiver Control and Status Register

(2014 0080)

6: Interrupt enable
o - disabled* - polled in console mode
TXCS:

Console Transmitter Control and Status Register (2014 0088)

6: Interrupt enable
o - disabled*
2: Loopback enable
o - disabled* - diagnostic use only
0: Break transmit
o • terminate SPACE condition*
SSCBT:

SSC Bus Time OUt Register

(2014 0020)

23:0: Bus timeout interval - 4000hex (16.384 ms)
range - 1 to FFFFFF (1 ps to 16.77 sec)

ADSOHAT: Programmable Address Strobe 0 Match Register
29:2: Match address
o - disabled* - not used

E-8 KA675/KA6801KA690 CPU

(2014 0130)

ADSOKAS: Programmable Address Strobe 0 Mask Register

(2014 0134)

29:2: Mask address bits - not used
ADSlMAT: Programmable Address Strobe 1 Match Register

(2014 0140)

29:2: Match address - 20084000 (for BDR)
ADS1KAS: Programmable Address Strobe 1 Mask Register

(2014 0144)

29:2: Mask address bits - 7C (for BDR)
T1CR:

Programmable Timer 0 Control Register

(2014 0100)

6: Interrupt enable

o - disabled*

2: STP
o - run after overflow·
0: RON
o - counter not running*

T1CR:

(historical)

Programmable Timer 1 Control Register

(2014 0110)

6: Interrupt enable

o - disabled·
2: STP
o - run after overflow·
0: RON
1 - counter incrementing every microsecond (historical)

TNIR: Programmable Timer Next Interval Registers (2014 0108,
-----------------------------2014 0118)
31:0: Timer next interval count (use 2's complement)
range - O· to 1.2 hours
TOIV:

Programmable Timer 0 Interrupt Vector Register

9:2: Tl:er interrupt vector T1IV:

(2014 OlOC)

78he-~

Programmable Timer 1 Interrupt Vector Registers (2014 OllC)

9:2: Timer interrupt vector - 7Chex

TOY:

Time of Year Register
31:0:

(2014 OOSC)

Number of 10 ms intervals since written

DLEOR:

Diagnostic LED Register

3:0:

Display bits
(historical)

(2014 0030)

o - LEOs on*

Configurable Machine State

E-9

Appendix F

NVRAM Partitioning
This appendiX describes how the CPU firmware partitions the sse 1 KB
battery-hacked-up (BBU) RAM.

F.1 sse RAM Layout
The KA6751KA6801KA690 firmware uses the 1KB of NVRAM on the sse
for storage of firmware specific data structures and other information that
must be preserved across power cycles. This NVRAM resides in the sse
chip starting at address 20140400. The NVRAM should not be used by
the operating systems except as documented below. This NVRAM is not
reflected in the bitmap built by the firmware.
Figure F-1: KA675/KA680IKA690 sse NVRAM Layout

20140400

Public Data Structures
(CPMBX. etc.)
Service Vectors

Firmware Stack

Diagnosbc State

201407FC

Rsvd for Customer Use
MLO-008655

F.1.1 Public Data Structures
The following is a list of the public data structures in NVRAM used by the
console.
Fields that are designated as reserved andlor internal use should not be
written. because there is no protection against such corruption.

NVRAM Partitioning

F-1

F.1.2 Console Program Mail~ox (CPMBX)
The Console Program MailBoX (CPMBX) is a software data structure
located at the beginning of NVRAM (20140400). The CPMBX is used to
pass information between the CPU firmware and diagnostics, VMB, or an
operating system. It consists of three bytes referred to here as NVRO,
NVR1, and NVR2.
Figure F-2:

NVRO (20140400) : Console Program MailBoX (CPMBX)

NVRO

LANGUAGE

RIP

BIP

HLT_ACT

MLQ.008657

Table F-1:
Field

Name

Description

7:4

LANGUAGE

This field specifies the current selected language for displaying halt
and error messages on terminals which support MCS.

RlP

If set, a restart attempt is in progress. This flag must be cleared by
the operating system, if the restart succeeds.

BlP

If set, a bootstrap attempt is in progress. This flag must be cleared
by the operating system if the bootstrap succeeds.

1:0

HLT_AC'f

Processor halt action - this field in conjunction with the conditions specified in Table 3-5 is used to c:cmtrol the automatic restart
/bootstrap procedure. HLT..AC'f is normally written by the Operate
ing system.

o Restart; if that fails, reboot; if that fails, halt.
1 Restart; if that fails, halt.
2 Reboot; if that fails, halt.
a Halt.

F-2

KA675IKA680/KA690 CPU

Figure F-3:
7

NVR1 (20140401)
6

NVR1
MLo.ooa653

Table F-2:
Field

Name

Description

MCS

If set, indicates that the attached terminal supports Multinational
Character Set. If clear. MCS is not supported.

CRT

If set, indicates that the attached terminal is a CRT. If clear,
indicates that the terminal is hardcopy.

Figure F-4:
7

NVR2 (20140402)
6

KEYBOARD

NVR2

MLO~

Table F-3:
Field

Name

Description

7:0

KEYBO~1ID

This field indicates the !l8.tional keybori variant in use.

F=1..3 Firmware Stack
This section contains the stack that is used by all of the firmware, with the
exception of VMB, which has its own built in stack

F.1.4 Diagnostic State
This area is used by the firmware resident diagnostics. This section is not
documented here.

NVRAM Partitioning

F-3

F.1.5 USER Area
The KA6751KA6801KA690 console reserves the last longword (address
201407FC) of the NVRAM for customer use. This location is not tested
by the console firmware. Its value is undefined.

F-4 KA675IKA680IKA690 CPU

Appendix G

MOP Counters
The following counters are kept for the Ethernet boot channel. All counters
are unsigned integers. V4 counters rollover on overflow. All V3 counters
'1atch" at their maximum value to indicate overflow. Unless otherwise
stated, all counters include both normal and multicast traffic. Furthermore,
they include information for all protocol types. Frames received and bytes
received cOWlters do not include frames received with errors. Table G-1
displayes the byte lengths and ordering' of all the counters in both MOP
Version 3.0 and 4.0.

Table G-1: MOP Counter Block

va
Name

Rx.-BYTES

Off

Len

02 4

V4
Off

Len

Description

Time since last zeroed. The time
which has eUapsed, since the counters
were last zeroed. Provides a frame
of reference for the other counters by
indicating the amount of time they cover.
For MOP V3, this time is the number of
seconds. MOP V4 uses the UTe Binary
Relative 'lime format.

Br"- received. The total number
of user data bytes successfully received.
This does not include Ethernet data liDk
headers. This number is the number of
bytes in the Ethernet data field, which
includes any padding or length fields
when they are enabled. These are
bytes from frames that passed hardware
filtering. When the number of frames
received is used to calculate protocol
overhead, the overhead plus bytes
received provides a measurement of the
amount of Ethernet bandwidth (over
time) consumed by frames addressed to
the local system.

MOP Counters

G-1

Table G-1 (COnt.): MOP Counter Block
V3

Off

Name

Len

Off
Len

Tx..BYTES

06 4

Bytes sent.
The total number of
user data bytes successfully transmitted.
This does not include Ethernet data
link headers or data link generated
retransmissions. This number is the
number of bytes in the Ethernet data
field, which includes any padding or
length fields when they are enabled.
When the number of frames sent is
used to calculate protocol overhead, the
overhead plus bytes sent provides a
measurement of the amount of Ethernet
bandwidth (over time) consumed by
frames sent by the local system.

OA 4

Frames received.
The total
number of frames successfully received.
These are frames that passed hardware
filtering. Provides a gross measurement
of incoming Ethernet usage by the local
system. Provides information used to
determine the ratio of the error counters
to successful transmits.

OE 4

Description

Frames sem.

The total number of

frames successfully transmitted.

This
does not include data link generated retransmissions. Provides a gross measurement of outgoing Ethernet usage by
the local system. Provides information
used to determine the ratio of the error
counters to successful transmits.

Rx..MCAST.-BYTES

12 4

G-2 KA675lKA680/KA690 CPU

Multicast bytes received.
The
total number of multicast data bytes
successfully received. This does not
include Ethemet data link headers. This
number is the number of bytes in the
Ethernet data field. In conjunction
with total bytes received, provides a
measurement of the percentage of this
system's receive bandwidth (over time)
that was consumed by multicast frames
addressed to the local system.

Table G-1 (Cont.): MOP Counter Block
V3

Off

Name

Len

Off
Len

Rx.-MCASTJ'B,AMES

16 4

TxJNIT.J>EFFERED

1A4

Description
Multicast frames received. The total number of multicast frames successfully received. In conjunction with total frames received, provides a gross percentage of the Ethernet usage for multicast frames addressed to this system.
FmmHHm,lwWilly~fand

The total number of times that a
frame transmission was deferred on
its first transmission attempt.
In
conr.mction with total frames sent.
measures Ethernet contention with no
collisions.

~TLCOLUSION

1E 4

Frames sem I, siDgie collision. The
total number of times that a frame was
successfully transmitted on the second
attempt after a normal collision on
the first attempt. In c:cnjunction with
total frames sent. measures Ethernet
contention at a level where there are
collisions but the backofI algorithm still
operates efficiently.

22 4

Frames sentI, m.u1tiple COllisiODS.
The total number of times that a frame
was successfully transmitted on the
third or later attempt after normal
In
collisions on previous attempts.
conjunction with total frames sent,
mesS"res Ethernet contention at a level
where there are collisions and the
backoff algorithm no longer operates
efficiently. NO SINGLE l!'B.AIm IS COL"NTBD
IN KOBE THAN ONE OF THE ABOVE '11IBEE

comrnms.

10nly one of these three counters will be incremented for a given frame.

MOP Counters

G-3

Table G-1 (Cont.): MOP Counter Block

Name

Off
Len

Description

26 2

Send failure count.
S
The
total number of times a transmit
attempt failed. Each time the counter
is incremented, a type of failure is
recorded. When Read-counter function
reads the counter, the list of failures
is also read. When the counter is
set to zero, the list of failures is
cleared. In conjunction with. total frames
sent, provides a measure of significant
transmit problems. TxFAIL_BITMAP
contains the possible reasons.

2C 2

Send fallure reason bitmap. S
This bitmap lists the types of transmit
failures that ocx:urred as summarized
below.

o • Excessive collisions.
1 - Carrier detect failed.
2 - Short c:imJit.
3 - Open circuit.
4 - Frame too long.
S - Remote failure to defer.
58

Send failure - Excessive collisions.
Exceeded the maximum number of retransmissions due to collisions. Indicates
an overload condition on the Ethernet.

Send fanure - Carrier check faRed.
The data link did not sense the receive
signal that is required to accompany
the transmission of a frame. Indicates
a failure in either the transmitting or
receiving hardware. Could be caused by
either transceiver, transceiver cable, or
a babbling controller that has been cut

off.

2Va send/receive failures are collapsed into one counter with bitmap indicating which failures
occurred.

G-4 KA67SIKA680/KA690 CPU

Table G-1 (Cont.): MOP Counter Block
V3
Off

Name

Len

V4
Off

Len

Description

68 8

Send failure - Short circuit. S There
is a short somewhere in the local area
network coaxial cable or the transceiver
or controllerltransceiver cable has failed.
This indicates a problem either in local
hardware or global network. The two
can be distinguished by checking to see
if other systems are reporting the same
problem.

Send failure - Open circuit. 3 There
is a break somewhere in the local area
ne..f-work comal cable. This indicates
a problem either in local hardware
. or global network. The two can be
distinguished by checking to see if other
systems are reporting the same problem.

Send failure - Frame too long.
The controller or transceiver cut
off transmission at the maximum size.
This indicates a problem with the local
system. Either it tried to send a frame
that was too long or the hardware CIltoft'
transmission too soon.
S

SeDd failure - Remote failure to
defer. S A remote system began
transmitting after the allowed window
for collisions.
This indicates either
a problem with some other system's
ca.Pfier sense or a weak tnms:m.itter.

SAlways zero.

MOP Counters

G-5

Table G-1 (Cont.): MOP Counter Block

va
Name

RxFAlLJnTMAP

Off

Len

V4
Off

Len

Description

2A 2

Receive failure COlUlt.
I The
total number of frames received with
some data error. Includes only data
frames that passed either physical or
multicast address comparison. This
counter includes failure reasons in the
same way as the send failure counter. In
conjunction with total frames received,
provides a measure of data related
RxFAIL_BITMAP
receive problems.
contains the possible reasons.

2C 2

Receive fanure reason bitmap. 2
This bitmap lists the types of receive
failures that occurred as summarized
below.

o - Block check failure.
1 - Framing error.
2 - Frame too long.

RxFAIL..FRAMING...ERR

Receive failure - Block check error
A frame failed the CRe check. This
indicates several possible failures, such
~ EM!, late collisions, or improperly set
hardware parameters.

Receive tanure - Framing error. The
frame did not contain an integral number of 8 bit bytes. This indicates several
possible failures, such as, EM!, late collisions, or improperly set hardware parameters.

Receive fanure • Frame too long. I
The frame was discarded because it was
outside the Ethernet maximum length
and could not be received. This indicates
that a remote system is sending invalid
length frames.

!Va send/receive failures are collapsed into one counter with bitmap indicating which failures
occurred.
3Always zero.

G-6 KA6751KA680/KA690 CPU

Table G-1 (Cont.): MOP Counter Block

va
Name

D~OVERRUN

Off

Len

V4
Off

Len

Description

AO 8

Um-ecognjzed frame destiDaiiolL The
number of times a frame was discarded
because there was no portal with the
protocol type or multicast address enabled. This includes frames received for
the physic:al address, the broadcast ad.dress, or a multicast address.

AS 8

Data overrun.
The total number
of times the hardware lost an incoming
frame because it was unable to keep
up with the data rate. In conjunction
with total frames received, provides a
measare of hardware resource failuteS.
The problem re1lected in this counter is
.also captured as an event.

32 2

BO 8

System buffer 1UI8vai1ab1e3 The
total number of times no system buffer
was available for an incoming frame. In
coDjunction with total frames received,
provides a measnre of system buffer
related receive problems. The problem
reflected in this counter is also captured
as an event.
This can be any
buffer between the hardware and the
user buffers (those supplied on Receive
requests).
Farther information as
to potential clift"erent buffer pools is
implementation speci1ic.

B8 8

User baffer UDavailable. S The
total !lU!!!her of times DO user bu11'er
was available for an incoming frame
that passed all filtering. These are the
buffers supplied by users on Receive
requests.
In cxmjunction with total
frames received, provides a measure of
user butTer related receive problems.
The problem re1lected in this counter. is
also captured as an event..

3 Always zero.

MOP Counters

G-7

Table G-1 (Cont.): MOP COunter Block
V3

Name

Off

Len

G-8. KA675/KA6801KA690 CPU

Off

Len

Description

co 8

Collision detect check failure. 'llle
approximate number of times that
collision detect was not sensed after a
transmission. If this counter contains a
number roughly equal to the number of
frames sent, either the collision detect
c:ircaitry is not working correctly or the
test signal is not implemented.

Appendix H

Programming the KFQSA Adapter
The KFQSA emulates a UQssP controller for each ISE (Integrated Storage
Element) to which it is connected, and thus presents a separate eSR
address for each emulated controller. The KFQSA must be programmed
with a correct CSR for each ISE on the DSSI bus. Interrupt vectors for the
KFQSA are programmed automatically by the operating system.
Unlike most other Q-bus controllers, the KFQSA CSR addresses are not
set with switches or jumpers. They are contained in nonvolatile memory
on the KFQSA module, in the form of a configuration table. To access the
configuration table, the KFQSA needs to have a valid address already in
the table. This could be preprogrammed at the factory, but then you need
to have an ISE installed on the DSSI bus with the proper bus node ID that
has already been programmed. Another way to a get a valid address is
to use the service switch (SWitch 1 ON = SERVICE mode) on the KFQSA.
Table H-l shows the addresses available. It is easier to do if the switches
are set as shown for the range of addresses from 0774420 - 0774434 in the
upper portion of the table.

Table H-1: Preferred KFQSA Switch Settings
Switch 1

Switch 2

SwitchS

Switch 4

CSR Address (Octal) .

off
off
off
oft'

on
on
off
off

on
off

0774420 (fixed)
0774424 (fixed)

on.

on
on

0774430 (fixed)

off

0774434 (fixed)

Available Fixed and Floating Addresses
on
on
on
on

oft'
on
on
on

on
off
off

on
off
on
off

0760444 (secondary TMSCP)
0774500 (primaty TMSCP)
0760334 (secondary MSCP)
0772150 (primaty MSCP)

The address that the CSR needs to have must be determined before
programming the configuration table. To determine this address, the

Programming the KFQSA Adapter

H-1

system configuration as a whole needs to be looked at, since some devices
are assigned floating addresses;' while others use the fixed addresses.
Floating addresses vary with each type of module and the number of
modules installed in the system. Because of this, any time a module
is installed or removed from the system, the CSR addresses need to be
checked.
To find recommended CSR address values, use the CONFIGURE Utility at
the console prompt (»» as described in Section 3.7.2.
NOTE: The configure command does not look at any of the devices actuaUy
in the system. This means that one console can be' used to determine the
addresses for different systems. All of the devices in the system must be
listed in this utility, in case any of the devices present affect the address
that is being calculated.

In the following example, the system has a TK70, three RF73s connected to
a KFQSA, and a DESQA The utility responds with the CSR addresslvector
assignments for all entered devices.
»>COBFIGtJRE

Enter device configuration, HELP, or EXIT
Device, Number? help
Devices:
Devices:
LPV11
RLVl2
DMVl1
RRD50
RV20

RXJ11
DLV11J
TSVOS
RXV21
DELQA
DEQNA
RQC25
KFQSA-DISK
KFQSA-TAPE
KMV11
CXAl6
CXY08
CXB16
QPSS
DSVl1
LNV21
KWV11C
AAV11D
ADV11D
DRQ3B
VSV21
1B001
IDV11D
!AV11A
IAV11B
IGQ11
DIV32
DESNA
KZQSA
M7577
KWV32
Device, Number?
Numbers:
1 to 255, default is 1
Device, Number? TgE70
Device, Number? KFQSA-DISK,3
Device, Number? DESQA
Device, Number? EXIT

H-2

KA67SIKA6801KA690 CPU

DZQ11
DRV11W
DESQA
TQK50
IEQ11
VCB01
ADV11C
VCB02
IDV11A
MIRA
KIV32
LNV24

DZV11
DRV11B
RQDX3
TQK70
DB011
QVSS
AAV11C
QDSS
IDV11B
ADQ32
DTCN5
M7576

DFAOl
DPV11
KOASO
T081E
DRY11
LNV11
rucY11C
DRV11J
IDV1lC
DTC04
DTCOS
DEQRA

AddIess/Vector Assignments
-774440/120 DESQA
-772150/154 KFQSA-DISK
-760334/300 KFQSA-DISK
-760340/304 KFQSA-DISK
-774500/260 TQK70

»>

After the proper addresses have been determined, the CSR addresses need
to be programmed. To do so, enter the following command at the console
prompt (»»:
SET HOSTIUQSSPIMAINTENANCEISERVICE n
Where:
The Iservice n parameter specifies the controller number of a KFQSA in
SERVICE mode (in the case of multiple KFQSAs), and n is a number in
the range 0 to 3 (from Table H-l):
o is for address 0774420
1 is for address 0774424
2 is for address 0774430
3 is for address 0774434
Entering the SETlHOSTIUQSSPIMAINTENANCElSERVICE n command
displays the current contents of the KFQSA configuration table. For
example, suppose the first address is selected and the configuration table
is currently blank:
»> SET BOST/UQSSP~/SERVXCE 0
OQSSP Controller (774420)
Enter SET, CLEAR, SHOW, HELP, EXIT, or QUIT
Node

CSR Address

- - - - - XFgSA - - -

Model

Type HELP for a quick reference of the available commands.
? help
Commands:
SET <node> \KFQSA
SET <node> <CSR ADDRESS><MODEL>
CLEAR <node>
-

SHOW
HELP

EXIT
QUIT

set KFQSA DSSI node number
enable a DSSI device
disable a DSSI device
show current configuration
print this text
program the KFQSA
don' t proqram the KFQSA

Progral'TlrT1ng the KFQSA Adapter

H-3

Parameters:
<node>

o to 7

760010 to 777774
21 (disk) or 22 (tape)

To add the three RF-series ISEs from the previous example, enter the
following:
? SET 0 772150 21
? SEr 1 760334 21
? SEr 2 760340 21
?

NOTE: Be sure to enter the addresses in the same order they were listed by
the configure utility.

Enter the SHOW command to display what has just been entered:
? SBOIf
Node

CSR Address

Model
21
21
21

772150

760334
760340

, - - KFQSA - - - -

To delete an entry from the table, use the CLEAR command. For example,
to delete the entry for node 2, enter CLEAR 2 at the prompt.

When finished, enter the EXIT command to write the entries to the
configuration table.
? EXI~
programming the RFQSA

»>
After programming the configuration table check that the bus node ID
plugs on the drive front panels correspond to the numbers that have been
programmed into the KFQSA.. Set the KFQSA to NORMAL mode by setting
switch 1 to off (switches 2-4 have no effect when switch 1 is set to om.
Enter SHOW QBUS to verify that the configuration is as desired. You may
need to program DSSI parameters for the ISEs. Refer to Section 3.7.3.1 for
instructions on setting DSSI parameters.

H-4 KA6751KA6801KA690 CPU

Appendix I

Error Messages
The error messages issued by the KA675JKA6801KA690 finnware fall into
three categories: halt code messages, VMB error messages, and console
messages.

1.1 Machine Check Register Dump
Some error conditions, such as machine check., generate an error summary
register dump preceding the error message. For example, examining a
nonexistent memory location results in the following display.
>>>0 Ip/l 7fffffO

Examine non-exl~en~ memory.

MESR-SOIFFOOO
MEAR-11FFFFF9
HMCDSR-01111000 KOAMR-OOOOOooo
CESR-OOOOOooo
CMCCSR-OOOOC10S
CSEARl-OOOOOooo
CSEA!2-OOOOOooo
CIOEARI-OIOFCOOO CI0EAR2-000002CO CNEAR-OOOOOOoo
ICSR-OOOOOOOl
PCSTS-FFFFF800
PCACR-FFFFFFFS
TBSTS-COOOOOEO
TBADR-OOOOOooo
NESTS-OOOOoooo
NEOADR-E014066C
NEOCMD-SOOOFOOS
NEICKD-OOOOOooo
NEDATHI-OOOOOooo NED~LO-OOOOOOoo CEFSTS-0000022A CEFADR-07FFFFFO
BCETSTS-OOOOOooo BCETIOX-OOOOoooo 8CE~G-OOOOoooo
B~-DSTS-OOOOO'oo
BCEDIDX-oOOOOoos BCEDEOC-OOOOoooo CBTCR-00004000
DSER-OOOOOooo
QBEAR-OOOOOOOF
OEAR-OOOOOooo
IPCRO-oOOO
ECR-OOOOOCCA
,7.0 MACHINE CHECK 80060000 00000000 20047ECC 20047&80 20047&89 B0110080
»>

le2 Halt Code Messages
Except on power-up, which is not treated as an error condition, the following
balt messages are issued by the firmware whenever the processor halts
(Table 1-1).
For example, if the processor encounters a HALT instruction while in kernel
mode, the processor halts and the firmware displays the following before
entering console 110 mode.
?06 HLT INST

PC = 800050D3

The number preceding the halt message is the ilalt code.n This number is
obtained from SAVPSL<13:8><RESTART_CODE), IPR 43, which is saved
on any processor restart operation.

Error Messages

1-1

Table 1-1:

HALT Messages

Code

Message

Description

?02

EXTHLT

External halt, caused by either console BREAK
condition, Q22-bus BHALT_L, or DBR<AtJX....HLT> bit
was set while enabled.

Power-up, DO halt message is displayed. However,
the presence of the firmware banner and diagDOstic
countdown indicates this halt reason.

?04

ISPERR

In attempting to push state onto the interrupt
stack during an interrupt or exception, the processor
discovered that the interrupt stack was mapped NO
ACCESS or NOT VALID.

?OS

DBLERR

The processor attempted to report a machine check
to the operating system, and a second mac:bine check

0CCI11'l'ed.

?06

BLTINST

The proc:essor executed a HALT instruction in kernel

mode.
?01

SCBERR3

The SOB wc:tor had bits <1:0> equal to 3.

?OS

SCBERR2

The SOB wc:tor bad bits <1:0> equal to 2.

?OA

CHMFRISTK

A c:baDge mode iDstrac:ticm was executed when
PSLdS> was set.

10B

CHMTOISTK

The SCB wc:tor for a change mode bad bit <0> set.

10C

SCBRDERR

A hard. memory error occu.rred while the processor was
t!ying to read an exception or interrupt vector.

110

MCBKAV

An access violation or an invalid translation oc:carred
during machine c:hec:1.t e.u:eption processiDg.,

111

KSPAV

An access violation or tnmslation not valid occarred
duriDg pTocessiDg of a kemel stack not valid exception.

112

DBLERR2

Double machine check error. A machine check oc:mred
wbile tr,yiDg to service a machine check.

?13

DBLERR3

Double machine check error. A machine check occured
wbile trying to service a kernel stack not valid

exception.

1-2

KA6751KA6801KA690 CPU

Table 1-1 (Cont.):

HALT Messages

Code

Message

Description

?l9

PSLEXC51

PSL<26:24> = 5 on interrupt or exception.

?1A

PSLEXCS1

PSL<26:24> = 6 on interrupt of exception.

?lB

PSLEXC71

PSL<26:24> = 7 on interrupt or exception.

?1D

PSLRE151

PSL<26:24> = 5 on an REI instruction

?lE

PSLREI61

PSL<26:24> = 6 on an REI iDstructicm.

?1F

PSLREI71

PSL<26:24> = 7 an an REI instrnction.

?3F

MICROVERlFYFAILURE Mic:roc:od.e power-up self-test failed.

IFor the last six cases, the VAX an:hitecture does not allow execution on the interrupt stack
while in a mode other than kernel In the first three cases, an intemlpt is attempting to run
on the interrupt stack while DOt in kernel mode.. In the last three ~ an BEl instruction
is attempting to return to a made other than kemel and still run on the interrupt stadt.

1.3 VMB Error Messages
VMB issues the errors listed in Table 1-2.
Table 1-2:

VMB Error Messages

Code

Message

Description

?4O

NOSUCHDEV

No boatable devices found.

141

DEVASSIGN

Device is not present.

142

NOSUCHFILE

Program image net found.

?43

FILESTRUcr

Invalid. boot device file structure.

144

BADCHKStJM

Bad ebecbum Oil headeT file.

?45

BADFILEHDR

Bad file header.

?46

BADmECTORY

Bad directory file.

?47

FlLNOTCNTG

Invalid program image format.

?48

ENDOFFILE

Premature end of file encountered.

?49

BADFILENAME

Bad file name given.

Error Messages

Table 1-2 (Cont.):

VMB ErroF-Messages

Code

Message

Description

?4A

BUFFEROVF

Program image does not fit in available memory.

?4B

CTRLERR

Boot device I/O error.

?4C

DEVINACT

Failed to initialize boot device.

?4D

DEVOFFLINE

Device is ofliiDe.

14E .

MEMERR

Memory iDitialization error.

14F

SCBINT

Unes:pected sea exception or machine check.

?50 .

SCB2NDJNT

Unes:pected ezceptiDn after startiDg program imqe.

?51

NOROM

No valid ROM image founcL

?52

NOSUCHNODE

No response from load server.

?53

lNSFMAPREG

154

BETRY

No devices bootable, retryiDg.

?55

lVDEVNAM

Invalid devh:e name.

?56

DRVERR

Drive elTOl'.

1.4 Console Error Messages
The error messages listed in Table 1-3 are issued in response to a console
command that bas error(s).
Table 1-3: Console Error Messages
Code Message
Description
?61

CORRUPl'lON

The console program database has been corrupted.

162

ILLEGAL REFERENCE

mega! reference. The ~ested reference would violate
virtual memory protection, the address is not mapped,
the reference is invalid in the spec:i1ied address space,
or the value is invalid in the specifiecl destination.

?63

ILLEGAL COMMAND

The command string CSlmOt be parsed.

?64

INVALID DIGIT

A number bas aD invalid digit.

1-4 KA67SJKA680/KA690 CPU

Table 1-3 (Cont.):

Console Error Messages_

Code

Message

Description

?65

LINE TOO LONG

The command was too large far the console to buffer. .
.The message is issued only after receipt of the
terminating carriage return.

?66

ILLEGAL ADRRESS

The address specified falls outside the limits of the
address space.

?67

VALUE TOO LARGE

The value specified. does not fit in the destination.

?68

QUALIFIER CONFUCT

Qualifier conflict, for example, two difi'erent data sizes

are specified for an EXAMINE command.
?69

UNKNOWN QUALIFIER

The switch is unrecognized.

?6A

UNKNOWN SYMBOL

The symbolic address in an EXAMINE or DEPOSIT
command is unrecognized.

?6B

. ?6C

?6D

CHECKSUM

The command or data checksum of an X command
is incorrect. If the data checksum is incorrect, this
message is issued, and is not abbreviated to "D1egal
command".

HALTED

The operator entered a HALT command.

FIND ERROR

A FIND command failed either to find the RPB or 128

KB of good memory.
?6E

TIME OUT

During an X command, data failed to arrive in the time
expected (60 seconds),

?6F

MEMORY EBROR

A machine check occorred with a code indicating a read
or write memory error.

170

UNIMPLEMENTED

Unimplemented functio~

?71

NO VALUE QUALIFIER

Qualifier does DOt take a value.

172

AMBIGUOUS QUALIFIER There were not enough uniq"ole characters to de+~nnine
the qualifier.

173

VALUE QUALIFIER

174

TOO MANY QUALIFIERS Too many qualifiers supplied for this command

175

TOO MANY ARGUMENTS Too many arguments supplied for this command.

?76

AMBIGUOUS COMMAND There were DOt enough unique characters to determine
the command.

QuaWier requiles a value.

Error Messages

1-5

Table 1-3 (Cont.):

Console Error Messages

Code

Message

Description

?77

TOO FEW ARGUMENTS

l:asuflicient arguments supplied for this command.

?78

TYPEAHEAD OVERFLOW The typeahead buffer overflowed.

179

FRAMING ERROR

A framing error was detected on the console seria1line.

?7A

OVERRUN ERROR

An overrun error was detected on the console serial

line.

?7B

SOFl'ERROR

A soft error occu.rred.

?7C

HARD ERROR

A hard error occarred.

?7D

MACHINE CHECK

A machine check occurred.

1-6 KA6751KA6801KA690 CPU

Appendix J

Related Documents
The following documents contain information relating to the lIlaintenance
of systems that use the KA.6751KA6801KA690 CPU module.
Title

Part Number!

Guide to Entry Systems Service Information Kits

EK-K27~

KA6751.KA6801KA690 CPU Technical Manual

EK-KA~TM

VAX 4000 Site P!eparaticm Guide

EK-387A~

BA43OiBA440 Enclosure Maintenance

EK-348A#-MG

BA400-Series Enclosures Storage Devices mstaDation Procedures

EK-BA44A-IN

DSSI Warm Swapping Guide for BA4OO-Series Enclosures and
KFQSA Adapters

EK-457AA-SG

DSSI VAXcluster Installation and Troubleshooting

EK-410AA-MG

Mic:roSystems OptioDS

EK-192At-MG

Mk:roVAX Diagnostic Monitor User's Guide

AA-FM7A1-DN

KFQSA Storage Adapter Installation and User Manual

~-KFQSA-lN

RF.&ries Integrated Storage Element User Guide

EK-RF72D-UG

RF-Series Integrated Storage Element Serlice Guide

EK-RF72D-SV

Related DoaJments

J-1

Glossary

BFLAG

Boot FLAG is the longword supplied in the SET BFLAG and BOOT
IRS: mynmsDds that qualify the bootstrap operation. SHOW BFLAG
displays the cunent value.

BBALT

Q22-bus Halt signal is usually tied to the front panel Halt switch.

BIP

Boot In Progress flag in CPMBX<2>

Bugcbeclt

Software or hardware error fatal to VMSprocessor or system. .

Cache memory

A small, bigh-speed memory placed between slower main memcny and
the plOCeSSOr. A cache increases effective memory transfer rates and
processor speed.
Console Program Mailbox is used to pass information between operating
systems and the firmware.

CSR

Control and stata.s register. A device or exmtroller register that resides
in the processar's I/O space. The CSR initiates device activity and
records its status.

CQBIC

CVAX to Q22-iras interface chip

DCOK

Q22-bus signal mdicating de power is stable. This signal is tied to the
Restart switch on the System Control Panel.

Diagnostic Execntive is a compcment of the ROM-based diagnostics
!'eSpo!!sible far set-up» execution. and clean-up of compon.ent diagnostic
tests.

DNA

Digital Networlt Architectme

DMA

Direct Memoty Access. Ac:cess to the memar,y by an I/O device that does
not require processor intervention.

EPROM

Erasable Programmable Read-Only Memory is used on some products .
to store firmware. Commanly used synonyms are PROM or ROM.
Erasable by usiDg ultraviolet light.

ECC

Error Conect:ian Code. Code that carries out automatic error correction .
by perlbnnhlg an exclusive or operaticm on the transferred data and
applyin.g a conection mask.

Glossary-1

Factory Installed
Software (FlS)

Operating system software that is loaded into a system disk duriDg
manufacture. On site, -the FIS is bootstraped in the system, prompting
a predefined menu of questions on the final con1igaration.

FEPROM

Flash Erasable Programmable Read-Only Memory (FEPROM) is used
on four chips on the KA6751KA68O/KA690 module. FEPROMs use
elect.rical (bulk) erasare rather than ultraviolet erasure.

Firmware

Firmware in this document refers to VAX instruction code residing at
physical address 2004000O all the KA6751KA6801KA690. Functionally
it consists of diagnostics, bootstraps, console, and halt entry/exit code.

FRU

Field-Rep1acable Unit. Any system component that the field engineer
is able to replace on-site.

GPR

General Purpose Registers on the KA6751KA68OIKA690 are the sixteen
standard VAX longword registers RO through R15. The last four
registers, R12 through R15, me also known by their unique mnemoDic:s
AP (Argument Pointer), FP (Frame Pointer), SP (Stack Painter), and
PC (Program Counter), respectively.

Imtia1jzation

The sequence of steps that prepare the system to start. Initialization
oc::curs after a system has been powered up.

IPL

Interrupt Priority Level ranges from 0 to 31 (0 to IF hex).

IPR

Internal Processor Registers on the KA65IKA.68OIKA69O are those
implemented by the processor chip set. These longword registers
are cmly accessible with the instructions MTPR (Move To Processor
Register) and MFPR (Move From Processor Register) and ~
kernel mode privileges. This document uses the prefix "PR$_"' when
referencing these registers.

ISE

Integrated storage element. An intelligent disk drive used em the
Digital Storage Systems Interccmnect.

KA6151KA68OJKA69O NVAX based Q22-bus CPU processor module with cmboard cache, two
DSSI ports, and Ethernet adapter.

LED

Light EmittiDg Diode

Macbi.ue c:hec:k

An operating system action trigered by certain system errors that ean
be fatal to system operation. Once triggered, machine check handler
software analyzes the error, comparing it to predetermined failure
scenarios. Three outcomes are possible: the system continues to run,
the software program is halted, or the system crashes.

MOP

Maintenance Operations Protocol specifies message protocol far
network Joopback assistance, networlt bootstrap, and remote console
fanctions.

Mass Stmage Control Protocol is used in Digital c:1isks and tapes.

Glossary-2

Mjl1isecond (l0e-3 seconds)

NVRAM

Nonvolatile RAM, on the KA6751KA68OIKA690 this is 1 Kb of battery
backed-up RAM on the sse.

Program Counter or R15

PCB

Process Control Block is a data stn1~ pointed to by the PRSJ'CBB
register and contains the c:nrrent precess' hardware context.

PFN

Page Frame Number is an index of a page (512 bytes) of local memory.
A PFN is derived from the bit field <23:09> of a physical address.

PR$_ICCS

Interval Clodt Control and Status, IPR 24

PR$JPL

Interrupt Priority Level, IPR 18

PR$.-MAPEN

Memory Management Mapping Enable, IPR 56

PR$ycBB

Process Control Block Base zegister, lPR 16

PR$_RXCS

ReX)eceive CoDSOle Status, IPR 32

PR$_RXDB

R(X)eceive Data Bafi'er, IPR 33

PR$_SAVISP

SAVed 1ntenupt Stack Pointer, IPR 41

P.R$_SAVPC

SAVed Program CounteJ; IPR 42

PR$_SAVPSL

SAVed Program Status IDDgWord, IPR 43

PR$_SCBB

System Control Block Base register, IPR 17

PR$_SISR

Sohare Interrupt Summ&IY Register, IPR 21

PR$_TODR

Time Of Day Register, JPR 2:1, is commonly referred to as the Time Of

Year register or TOY clock.
T(X)ransmit Console Status, IPR 34
T(X)ransmit Data BuJIer, IPR 35

Processor Status Longword is the VAX extension of the PSW (Processor
Sta.tns Word). The PSW Gower word) contains instroction condition
codes and is accessible by nonprivileged users; however, the upper .
word contains system status information and is accessible by privileged

users.
Q22-bus Map Base :Register found in the CQBIC determines the base
address in local memoIY for the scatterlgather registers.

QDSS

Q22-bus video controller for workstations

GIossary-3

QMR

Q22-bus Map ~r

QNA

Q22-bus Ethernet contmller module

RAM

Random AI:t:I!ss Memory

RIP

Restart In Progress flag in CPMBX<3>

RPB

Restart Parameter Block is a software data structure used as
a comm111lication mecbanism between firmware and the operating
system. Information in this block is used by the firmware to attempt
an operating system (warm) restart.

SCB

System Control Block is a data structure pointed to by PR$_SCBB. It
contains a list of longword exCeption and interrupt vectors.

SGEC

Second Generation Ethernet Chip

SDn

Symptom-Directed Diagnosis. Online analysis of nonfatal system errors

in order to locate potential system fatal errors before they occur.

SBAC

Single Host Adapter Chip

Stack Pointer or R14

SBM

Standard Reference Manual, as in VAX SRM

sse

System Support Chip

'JlS

Microsecond (10e.6 seconds)

VAXclustercon1iga-

A bighly integrated organization of VMS systems that coDllD11Dicate
o~ a bigh-speed commUDications path. VAXc1uster mDJiguratioDS
have all the fa.Dctions of single-node systems, plus the ability to
share CPU resources, queues, and disk storage. Like a singlenode system, the VAXc1uster canfiguration provides a siDgle security
and management environment. Member nodes can share the same
operating environment or serve specialjzed needs.

ration

VD'tual Memory Boot is the portion oftbe firmware dedicated to booting
the operating system.

Glossary-4

Index
A
Acceptance testing, 4-15 to 4-20
Algorithm
to find a valid RPB, 4-40
to restart operating system, 4-39
ALLCLASS, 3-25
setting, 3-33
ANAL~O~ 5-15
interpreting CPU errors using,
5-16
interpreting DMA to host
transaction faults using,
5-29

interpreting memory errors using,
5-18

interpreting system bus faults
using, 5-27
.
ANAL'YZEISYSTEM, 5-21

B
Backplane
description, 2-14
Binary load and unload
command), A-33

Bits
RPB$V_DIAG, 4-32
RPB$V_SOLICT:; 4-32

Boot
flags, 3-47
supported devices, 3-46, H-l
Boot Block Format, 4-30
BOOT command, A-I0
Boot Flags
RPBSV_BBLOCK, 4-30
Bootstrap
conditions, 4-23

Bootstrap (cont'd)
definition at: 4-23
disk and tape, 4-30
failure, 4-24
initialization, 4-24
memory layout, 4-25
memory layout after successful
bootstrap, 4-28
network, 4-32
preparing for, 4-24
primary, 4-26
PROM, 4-31
secondary, 4-27
control passed to, 4-28
Break EnablelDisable switch, 2-8

c
9C utility, 4-16,5-56
Comment command (!), A-35

! (comment command), A-35
Configuration, 3-1
and module order, 3-1
CONFIGURE, 3-22
CONFIGURE command, 3-22, A-ll
Co-~le ::Q""!--g9)ds

address space control qualifiers,
A-7
address specifiers, A-3
binary load and unload 00, A-33
BOOT, A-lO
! (comment), A-3S
CONFIGURE, A-U
CONTINUE, A-l3
data control qualifiers, A-7
DEPOSIT, A-lS
EXAMINE, A-l4
FIND, A-IS

Index-1

Console commands (cont'd)
HALT, A-16
HELP, A-16
INITIALIZE, A-18
keywords, A-8
list of, A-8
MOVE, A-19
NEXT, A-20
quallfier and argument
conventions, A-2
qualifiers, A-6
REPEAT, A-22
SEARCH, A-22
SET, A-24
SHOW, A-28
START, A-32
symbolic addresses, A-3
syntax, A-2
TEST, A-32
UNJAM, A-33
X (binary load and unload), A-33
Console error messages
sample ~ 5-40
Console I/O mode
special c:baracters, A-l
Console module
desription, 2-6 to 2-11
fuses, 2-10
Console part, testing, S-66
CONTINUE cmnmand, A-13
CPU
features, 2-1 to 2-5
location, 3-1

D
DC OK Indicator
function, 2-13
on System Control Panel, 2-13
DEPOSIT command, A-13
Device Dependent Bootstrap
Procedures, 4-30
Diagnostic executive, 4-9
error field, 5-41

IOOe:x-2

Diagnostic tests
list of, 4-9
parameters for, 4-9
Diagnostics
relationship to UETP, 5-62
Diagnostics, DSS! storage devices,
5-57
Diagnostics, RF-series, 4-8
DNA Mamtenance Operations
Protocol (MOP), 4-32
Documents
related, J-l
DSSI parameters, 3-24
DSSI storage device
elTOrs, 5-57
testing, 5-57
DSSI storage device local programs
list of, 5-57

DSSI VAXcluster
capability, 3-13
configuration rules, 3-15
examples ~ 3-17,3-19
DUP driver utility, 3-24, 3-27
entering from console mode, 3-31
entering from VMS, 3-32
exiting, 3-38

E
Entry Point
definition of, C-l
Error during UETP, 5-63
diagnosing, 5-62
Error Log Utility
relationsbip to UETP, 5-62
Error messages
console, sample ~ 5-40
EXAMINE command, A-I4
Expanders
control power bus, 3-9
mass storage, 3-8
Q-bus, 3-8

INIT, 4-.2.4

Fans

Initial power-up test
See IPR
Initialization
following a processor hal~ "4-39
prior to bootstrap, 4-24
INITIALIZE command, A-1S
IPL_31, 4-25
iSYS$TEST logical name, 5-62

Fan Speed Control Disable (FSC),

2-18
location, 2-17
FE utility, 5-53
Fl1es-lllookup, 4-30
FIND command, A-15
Firmware
commands and utilities, 3-20
power-up sequence, 4-1
updating, 6-1
Flags
restart in progress, 4-39
FORCEUNI, 3-25

Fuses
for H3604 console module, 5-64
troubleshooting, 5-04

G
General purpose registers (GPRs)
in error display, 5-43
symbolic addresses for, A-3

H
H3l0S loopback connector, 5-66
H3604 YO panel, 5-66
HSS12 loopback connector, 5-66
Halt
dispatch, D-1

HALT
on bootstrap failure, ~27
Halt actions
summary, 3-48
Halt Button
location, 2-13
HALT command, A-16
Halt protection, override, 5-54
HELP command, A-16

L
Language selection menu
conditions for display of, 4-2
example of, 4-2
messages, list of, 4-1
Local Memory Partitioning, 4-25
Log file generated by UETP

OLDUETP.LOG, 5-63
Loopback connectors
H3l03, 5-66
H8572, 5-66
list ot: 5-69
Loopback tests, 5-64
console port, 5-66
DSSI, 5-67
Etherne~ ~8

Q-bus, 5-69

M
Maintenance strategy, 1-1
field f'eedbac~ 1..0
information services, 14
service delivery, 1-1
service tools and utilities, 1-2

Mass storage
configuration of, ~
rules for numbering, 3-7
Memory
acceptance testing of, 4-16
isolating FRU, 4-17,5-54
modules, 2-5
testing, 5-54

Index.....s

Memory module
desription, 2-5
installing, 3-2
order, 2-5
Module
configuration, 3-5
order, in backplane, 3-1
self-tests, 4-7, 5-69
MOM$LOAD, 4-32
MOP functions, 4-35
MOP program load sequence, 4-32
MOp, functions, 5-59
MOVE command, A-19

PFN bitmap, 4-24
POST
See Power-on self-tests
errors handled by, 5-57
Power supply
desription, 2-15 to 2-17
minimum load, 3-13
Power-on self test
See POST
.Power-on self-tests
description, 4-4
errors handled by, 4-8
kernel, 4-4
mass storage, 4-8

N
Network listening, 4-33
NEXT command, A-20

NODENAME, 3-25
setting, 3-37

NVRAM
CPMBX, F-2
partitioning, F-1

o
OLDUETP.LOG file, 5-62
Operating System
bool:stlap, 4-23
restarting a halted, 4-39

Q-bus,4-7
power-up
machine state, 4-20
memory layout, 4-21
Power-up mode switch
set to language inquiry, 4-1
set to run, 4-3
set to test, 4-1
Power-up sequence, 4-1
Power-up tests, 4-1
PRAO,4-31
Primary Bootstrap, 4-26

Operating System Restart
definition at: 4-39
Options
adding to enclosure, 3-10 to 3-13
Over 'Dmlperature Warning indicator
system, 2-13

Q-bus options, recommended order,
3-5
Q22-bus Memory
and VMS, 4-28

Registers
initializing the general purpose,
4-25
Q22-bus Map Registers, 4-28

Page Frame Number Bitmap, 4-32
Parameters
for diagnosUc tests, 4-9
in error display, 5-42
Pa~leCon~lS~

Error messages, 6-9

Index-4

Related documents, J-1
REPEAT command, A-22
REQ..PROGRAM, 4-33
Restart, 4-39

Restart Button
location, 2-13
Restart parameter block (PRB),
3-47
Restart Parameter Block (RPB)
RIP flag, 4-39
RF-series ISE
diagnostics, 4-8, 5-57
errors, 5-57
RF-series ISE local programs
list of, 5-57
ROM-based diagnostics, 4--8 to 4-10
and memory testing, 5-55
console displays during, 5-40
isolating failures with, 5-44
list of, 4-9
parameters, 4-9
utilities, 4-9
RPB
initialization, D-5
locating, 4-40
RPB Signature Fonnat, 4-40

s
Scripts, 4-10 to 4-11
list of, 4-13
SEARCH command, A-22
Secondary Bootstrap, 4-27
Self-test, for modules, 4-7, 5-69
SET BOOT device name command
use of, 3-44
SET command, A-24
SET HOSTIDUP command, A-24
SHOW command, A-28
SHOW commands, 3-29
SICL messages, ~3
converting appended MEL files,
5-36
Signature Block
PROM, 4--31
START command, A-32
S~licadmre~,A-3

for any address space, A-6
for GPRs, A-3

System c~qtrol panel, 2-12 to 2-13
System hang, 5-64
SYSTEMID, 3-25

setting, 3-37

T
Tape ISE
diagnostics, 5-57
errors, 5-57
Tape ISE local programs
list of, 5-57
Termination power, tests for, 5-6.7
TESTcommand,A-32
Tests, diagnostic
list of, 4-9
parameters for, 4-9
Troubleshooting
procedures, general, 5-2
suggestions, additional, 5-56
UETP,5-63

u
UETINIT01.EXE image, 5-63
UETP
interpreting VMS failures with,
5-62
UETP.LOG file, 5-62
Unit number labels, 3-34
UNITNUM, 3-25
setting, 3-34
'UNJ~\1:, 4-24
UNJAM command, A-33
User Environment Test Package
(UETP)

interpreting output of, 5-62
running multiple passes of, 5-62
typical failures reported by, 5-63
Utilities, diagnostic, 4-9

v
Valid Maps, 4-28

VAXELN
and VMB, 4-27

Index-5

VAXsimPLUS, 5-4, 5-31
. customizing, 5-38
enabling SICL, 5-39
installing, 5-37
Virtual Memory Boot (VMB), 4-27
definition of, 4-26
primary bootstrap, 4-26
secondary bootstrap, 4-30
VMB
boot flags, 3-47

VMS
error handling, 5-5
event record translation, 5-15

w
Wannstart, 4-39
Write-enabling

a storage element, 3-40
an RF-series storage element,
3-40 to 3-44 .
Write-protecting
a storage element, 3-40
an RF -series storage element,
3-40 to 3-44
an RF35 storage element, 3-40 to
3-44

x
X command (binary load and
unload), A-33

Index-6

Reader's Comments

KA675IKA6801KA690 CPU
System Maintenance

EK-454AA-MG-001
Ycmr comments and suggestions help us improve the quality of our publications.

Please rate the manual in the
following categories:

Excellent

Good

Fair

Poor

0 .
0

0
0
0

0
0

Acc:a.rar:y (product works as described)
Completeness (enough information)
Clarity (easy to understand)
Organization (structure of subject matter)
Figures (usefnl)
Examples (useful)
Table of contents (ability to find tDpic)
Index (abi1i~ to find topic)
Page design (overall appearance)
Print quality

0
0

ODD

O.
Cl
[J
DOD

What I like best about this manual:
What I like least about this manual:
Additional comments or suggestions:

I found the following errors in this manual:

Page

Description

For which tasks did you use this manual?

o In..c;t;al1ation

o Maintenance
o Marketing
Cl OperationlUse

o Progzs"'i1ijtlg

o System Management

o '!raining
o Other (please specify)

Namer1itle
Company

. Address
Date

mDmDDmo

NO P()S'QGE
NECESSARY
F MAl.&)

. ~THE
lNTED STATES

BUSINESS REPLY MAIL
FIRST CLASS PERMIT NO. 33 MAYNARD MASS.

POSTAGE WILl. BE PAID BY ADDRESSEE

DIGITAL EQUIPMENT CORPORATION
CORPORATE USER INFORMAnON PRODUCTS
PK03-11D3O
129 PARKERS
-T-R-EET!!I!!!"P
MAYNARD, IIA 01754-9975

1111111111111111.1111111111.1111111111111.1.11111111
Do Not Tear-Fold Bere aJldTape -