Digital PDFs

EK-306AA-MG-001

2000

196 pages

Original

47MB

view download

OCR Version

19MB

view download

Document:	KA655 CPU System Maintenance
Order Number:	EK-306AA-MG
Revision:	001
Pages:	196
Original Filename:

OCR Text

KA655 CPU
System Maintenance
EK-306AA-MG.001

”

' KA655 CPU System Maintenance

Order Number EK-306AA-MG-001

digital equlpmam corporation
maynard, massachusetis

March 1989

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 errors that may appear

this document.

)

The software, if any, described 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.

The READER'S comms form on the last page of this document requests the user's
critical evaluation to assist in preparing future documentation.

The following are trademarks of Digital Equipment Corporation:

DEC

MicroVAX

DECmate
DECnet
DECUS

DECwriter
DELNI
DEQNA
DESTA

DIBOL
DSSI

MicroPDP-11

UNIBUS

MicroVMS

PDP
P/0OS

Professional
Q-bus
|
Rainbow
RSTS

RSX
RT

VAX
VAXBI

VAXELN

‘

‘VAXcluster
VAXstation
VMS
vT

Work Processor
|

nARTM

ULTRIX
ML~S1123

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 pursuant 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. =
,
o
-

This document was prepared using VAXDOCUMENT, Version 1.1.

Contents
Preface

Chapter1 KA655 CPU and Memory Subsystem
1.1

Introduction .............ciiiiiiininrenennnnn.

1-1

1.2

KA655CPUFeatures..........coviiiiiiinnnnnnnnnn.

1-7

1.2.1

Central Processing Unit (CPU) ...........ccoov.n...

1.2.2

Clock Functions . ......... ..ottt

1-8
1-9

1.2.3

Floating Point Accelerator ..............c0vu...

1-9

1.2.4

Cache Memory .. ...cci ittt

ittt eieneennnns

1-9

125

MemoryController............ ...,

1-10

1.2.6

MicroVAX System Support Functwns et

1-10

127

ResidentFirmware ..............................

1.2.8

Q22-BusInterface .............. e

1-11

e ereeetee
e

1-11

KA655Connectors ........coiiiiiiiinein e

1-12

H3600-SACPUIL/OPanel .........c0oveurmnunnnin.

1-12

1.5

MS650-BA MemoOry . .o v v vv it ittt

ettt et

1-16

Introduction............oiiiiiiiiinnenmnennnnnns.

2-1

GeneralModuleOrder ..............ciiiienn.n.

2-1

Module Order Rules for KA655 Systems ..............

2-2

Recommended Module Order for KA655 Systems .......

2-2

Chapter 2

2.2.1

2.22

Configuration

2.3

Module Configuration...............................

2-3

2.4

DSSI Configuration ................. e, cee..

2.4.1

Changing RF-Series ISE Parameters . . Ceceeebeene e

2-6

242

Changingthe UnitNumber ........................

24.3

Changing the AllocationClass . .....................

2-7
2-9

244

DSSICabhng.;............

*****eee..

2410

DSSI Bus Termination and Length.......... ..

2912

Limitations to Dual-Host Configurations ..............

2-14

Confi
Worksheet
gurat
. ..........cov
ion
viv ...

2-14

'2.44.1 -

2.6

Dual-Host Capability ..........00vviinmnnnnnnnn. 2-13

Chapter 3
31

Introduction . ........coviiiiiiiininnnnnnnnnnennnns

3.2

KAG655 Firmware Features..................., I

3.3

Halt Entry, Exit, and DispatchCode ...................

3-2

ExternalHalts...............0iiiinnnnnnn.

3-3

Power-UpSequence ............ciiiiiiiinnnnnnnnnn

3—4

3561

ModeSwitchSettoTest...........ccovvvvnnnn...

3.5.2

Mode Switch Set to Language Inquiry ................

3-5

3.5.3

Mode SwitchSettoNormal ........................

3-6

Bootstrap ............iiiiiiiiiinn.. e

3-6

3-1

3.6.1

Bootstrap Initialization Sequenae et e

3-7

8362

VMBBootFlags........ovvviiiinnnnneeennnannnn.

3-8

363

SupportedBootDevices ...........covvveinnennnnn.

3-8

364

Autoboot....
.......
it it i.....
iiiiiee...

38-9

3.7

Operating System Restart ............coovvvinunnennn.

3-11

371

RestartSequence ............coiiiiririnennnnnnn.

3-11

372

Locatingthe RPB.............ccotiiiinnnnnnnnn.

3-12

Comsolel/OMode.......coviiiintineeneennnnnnnnnnn.

3-13

38
-

KA655 Firmware

381

Command Syntax.........vvviimnmmnnennnnnnnnn

3-13

382

AddressSpecifiers .................. iereineee...

3-14

383

SymbolicAddresses ...........0iiiii

3-14

3.84

Console Command Qualifiers . ...........o0veuun...

3-17

385

Console Command Keywords . ........covvrvennnnnn.

3-19
Console Commands ............. Ceenieas P 3-21
391
BOOT......... R F
.. 8-21

3.9

392

CONFIGURE . .......c¢0iittiiiinennene e

393

CONTINUE ..........coiiivvnnnnnnn.et

3-25

894

DEPOSIT . .....cii
ittt et
iiiii

3-26

395

EXAMINE ............. Vet e et eerentatasenerasens

3-27

3-23

V 309&6‘ )

FIND . L

3-9?7

HAIIT» N R

3@948

ImLP uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu 8

899

INITIALIZE ...........ccoviuueeeenennsannnnnns.

3-33

3&9;10

MOVE uuuuuIR » I

3—”34

R A A N A I

C e s e s s s e e eee 6 6 0 08

I I IR IRR .. uVa L

RE R R

3"‘29

I )

3—30

8 0 0 8 % 8

3&9&11

NEXT .'"“ LI

3&9012

REPEAT uuuuuuuuuuuuuuuu I I I T S T

3913

SEARCH ................. et e

309014V

SET NN

39.15

SHOW ...

A A I I A

RN RN

T T

3‘“‘31

3—36

3"‘38
3-39

E R A
ER

3’—41

itirtnnnnnnset

344

3«9116

STMT wwwwwwwwwwwwwwwwwwwwwwwww T N
RE

3_47

3.0.17

TEST i

ittt it

3&9018

UNJAM 58 8 L

8 8 8w

I I A A

ittt

tetien

348

3—49

3.9.19

X—Binary Load and Unload .......................

3-50

3920

! (Comment)...............ettt

3-52

Introduction...... et cereenaae cee.

General Procedures .......... F

4.3

KA655 ROM-Based Diagnostics ................c.......

431

DiagnosticTests . ..........cciiiiiiiiiiinennnnn.

4-3

T 2

- 1o o { )1S

4-6

433

ScriptCallingSequence . ..........ovviininnnennn.

4-8

434

CreatingSeripts.........c.iiiiiiiiiiii
..

4-10

Chapter 4
4.1

Troubleshooting and D:agnostlcs

435

ConsoleDisplays ........coiiiiiiiieiinnnnnennnn,

4-14

436

SystemHaltMessages.........ooiiiiiimenneennnn.

437

ConsoleErrorMessages..........covvieveennennnn.

4-24
4-25

438

VMBErrorMessages .........vvttiieneeennennnensn

4-26

AcceptanceTesting......... ...

nnnnn.

4-26

4.5

Troubleshooting....... Ceeeanan Ceetrreneenuns Ceeean

4-31

451

452
4.5.3
46
46.1

FEUtility............... e i
... 481

Isolating Memory Failures........... [P

4-34

Additional Troubleshooting Suggestions........ e

4-37

Loopback Tests..........oiiiiiiiiinniinennnnnnnns

4-38

Testing the Console Pm;t~~ C ettt

ee et

4-38

4.7

Module Self-Tests...... R et i,

4-38

4.8

- RE-Series ISE 'h'oubleshoomng and Dmgnostmcs e

4-40

74.8.1

DRV ST . ..

neeanennnnes

4-42

482

DRVEXR ............ccuu... et

ittt ettt

483

HISTRY........... @

et e
.

4-43
4-44

. 4.8.4

ERASE . ..ttt
ittt ittt e,

4-45

485

PARAMS ...ttt

4-46

4.8.5.1

9.4
o

4-47

4.8.5.2

HELP
... it
ettt et e eeeann.

4-47

4853

SET ¢ttt .

4-47
448

4854

SHOW .

48.5.5

STATUS .
ee

4-48

4.8.5.6

WRITE . ... .. itt
e et et
e

4-48

DiagnosticErrorCodes ...................couvuu....

4-49

Appendix A
Al

A.l1
A.2

e e

Configuring the KFQSA

KFQSAOverview............ciiiiiiiiinnnnnnnnnns

A-1

Dual-Host Configuration .......... P -

A-2

Configuring the KFQSA at Installation .................

A-2

A2.1

EnteringConsoleI’/OMode ...............couu.....

A2.2

Displaying Current Addresses......................

A-5

A.2.3

Running the Configure Utility .................... ..

A-6

Programmingthe KFQSA ................cc0vvvuu...

A-8

Reprogrammingthe KFQSA ......................... A-15
Changing the ISE Allocation Class and Unit Number ...... A-17

A.5

Appendix B

KA655 CPU Address Assignments

B.1

General Local AddressSpace Map .. ............co.....

B-1

B.2

Detailed Local Address Space Map .................... -

B-2

B.3

Intermal ProwaaorRegwtem cateeseencencns “eeseiea..

KA655 VAX StandardIPRs ..........e,

B-9

B.3.1
B32
B.4

KA655UniqueIPRs.............0coiiiiurumnnnnnn. B-10
Global Q22-Bus AddressSpaceMap ................... B-11
N\

Appendix C

Related Documentation

Index
Examples
2-1

ChangingaDSSINodeName .............ccvvuu....

2-6

2-2

Changinga DSSIUnitNumber.......................

2-8

2-3

Changing a DSSI AllocationClass . . ............c0v....

2-9

3-1

Language Selection Menu ........et

3-6

3-2

SelectingaBootDevice .........oiviiitinnenann

3-10

4-1

Creating a Script with Utility 9F . .....................

411

4-2

Listing and Repeating Tests with Utility 9F .............

4-13

4-3

Console Display (NOEITOTS). . .

oo oo oeee e

4-15

4—4

Sample OQutputwithErrors ..........................

4-15

FEUtilityExample ................................

4-32

4-6

Isolating Bad Memory UsingT9C.....................

4-36

4-7

9C—Conditions for Determining a Memory FRU..........

4-37

A-1

KFQSA (M7769) Service Mode Switch Settings ...........

A-2

Entering Console Mode Display ............. et

A-5

A-3

SHOWQBUSDisplay ........covuuuinnininnnnn.

A-5

A—4

Configure Display ........coviiiiennnenennnnnnn.

A-5

A-6
A-7T

A-8
A-9

A-T7
Display for Programming the First KFQSA . ............. A-10
Display for Programming the KFQSAin a Dual-Host

Configuration (Second System) ....................... A-11

SHOWQBUSDisplay .......covviiieeneennnnnnnnn. A-13
SHOWDEVICEDisplay .................... ceeee... A-14
Reprogrammingthe KFQSADisplay . .................. A-16

A-10 Display for Changing Allocation Class and Unit

Number ... A-18

vil

Figu res
. 1-1

KA655 CPU Module (M7625—AA) e

1-2

1-3

KAG655 System-Level Block Diagram, PartI .............

1-5

1-4

KA655 System-Level Block Diagram, Part II Cieeieiiene..

1-86

'1-2

KA655 CPU Functional Block Diagram .................

1-3

1-5

KA655 Pin Orientation ... ..........oooueusoninnnn.,

1-12

1-6

H3600-SACPUIOPanel ................coovnn....

1-13

1-7

MS650-BA Memory Module (M7622-A) ................

1-17

2-1

DSSI Cabling, BA213 Enclosure ................... ..

2-11

2-2

RF-Series ISE Operator Control Panel (OCP) ............

2-12

2-3

BAZ213 Configuration Worksheet ......................

2-16

4-1

KA655CPUModule LEDs............covviinnnnn..

4-18

A-1

KFQSA Module Layout (M7769) .........ccuvvvuunn...

A-3

A-2

Dual-Host Configuration Nodes and Addresses (Example)...

A-9

Tables

1-1

H3600-SA Controls and Connectors .............. ..

1-2

H3600-SA CPU I/O Panel Switches. ................... 1-15

1-14

2-1

DSSIDevice Order..........vuoemmumemnenennnnnnnn.

2-2

RF-Series ISE Switch Settings............ e

2-5

- 2-3

Power and Bus Loads for KA655Options ...............

2-15

3-1

Actions TakenonaHalt .................... e

3-3

3-2

Language Inquiry on Power-UporReset ................

3-5

3-3

Virtual Memory Bootstrap (VMB) Boot Flags ............

3-8

3-4

Boot Devices Supported by the KA655-AA ..............

3-9

3-6

Console Symbolic Addresses..............ccovvvu....

3-15

3-6

Symbolic Addresses Used in Any Address Space ..........

3-17

3-7

Console Command Qualifiers...............ccco0vn....

3-18

3-8

Command KeywordsbyType .........................

3-19

3-9

Console Command Summary ..................... ee..

319

4-1

Test and Utility Numbers .................... eeee..

4-2

Scripts Available to Field Semca ......................

4-7

4-3

Commonly Used Field Service Seripts ..................

4-8

4-4

Values Saved, Machine Check Exception During Executive ..

4-17

4-5

Values Saved, Exception During Executive ..............

4-17

viil

2-5

*"i

KA655 Console Displays and FRUs . ......... e.. 419
System
Hall Message
.
s .......... U e, 424
47
487 Console Error Messages . .. ....ccoviitneeennnnnnnnn. 4-25

- %

4-6

Error Messages . .. ....coiiiiii
e eeeennnnnn,
i

4-26

4-10

Hardware Error Summary Register . .. .................

4-33

4-11

Loopback Connectors for Q22-Bus Devices.

.. ............

4-39

4-12 DRV ST MeSsa
. . ¢ v oo
ge8
i i et
eiteitneeenesnnnnneens

4-42

4-14 HISTRY Me8888eS . .. cvviviiiiiiiiinneeeenneannnnn

4-44

4-13 DRVEXR MeSSBaZES .+« vt vvvereereennneneesnnnnnnnns. 4-43
4-15

ERASE MeS8aEZES . « « vt v o v e e vee e eee e eeeeens

s 4-46

4-16 RF-Series ISE Diagnostic ErrorCodes . . . ...............

4-49

B-1

VAX Memory Space . .......ouvueeeneereennneennnns

B-1

B-2

VAX Input/Output Space ............covierneen...

B-2

B-3

Detailed VAX Memory Space ...........ouvveuununnn..

B-2

Detailed VAX Input/Output Space .....................

B-3

B-5

KAG655 Internal Processor Registers. ...................

B-7

B-6

IPRs Implemented According to Standard VAX Architecture.

B-9

B-7

KA655 Unique IPRs. . . ...e F

B-8

Q22-Bus Memory Space . . .......... ettt e,

B-9

Q22-Bus I/0 Space with BBS7 Asserted ................ B-11

R B-10
B-11

/7 m‘ ¥

Prefabe
This guide describes the base system, configuration guidelines, ROM-based

diagnostics, and troubleshooting procedures for systems containing the
'KA655 central processing unit (CPU).

“Intended Audience
This document is intended only for Digital Field Service personnel and
qualified self-maintenance customers.

Organization
This guide has four chapters and three appendixes.

Chapter 1 describes the KA655 CPU and the MS650-BA memory.

Chapter 2 contains system configuration guidelines and provides a table

listing current, power, and bus loads for supported options.

Chapter 3 describes KA655 diagnostic firmware.
Chapter 4 describes the KA655 diagnostics, including an error message
and FRU cross-reference table. It also describes diagnostics that reside on
RF-series integrated storage elements (ISEs).
Appendix A explains how to configure the KFQSA storage adapter.

Appendix B contains a list of KA655 CPU address assignments.
Appendix C contains a list of related documentation.

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

Provides infomatian to prevent personal injury.

CAUTION

Provides information to prevent damage to equipment or software.

NOTE

Provides general information about the current topic.

Chapter‘ 1

- KA655 CPU and Memory Subsystem

1.1 Introduction
This chapter describes the KA655 central processing unit (CPU) and the
MS650-BA memory.
The KA655 CPU module (M7625-AA), shown in Figure 1-1, is a quadbeight VAX processor for the Q22-bus (extended LSI-11 bus).
It is
designed for use in high-speed, real-time applications and for multiuser,
multitasking environments.
The KA655 employs a cache memory to

maximize performance.

KA655 CPU and Memory Subsystem

1-1

—.
5 3

KAG655 CPU Module (M7625-AA)

Figure 1-1:

MEMORY CONNNECTION

CONSOLE CONNECTION

CONSOLE SERIAL

LINE CONNECTION

EPROM

. ,f |
;

Ly |

VEL

CACHE

v |

CCLK

CMCTL

“CVAX

CFPA

casic

MLO-O023867

1-2

KA655 CPU System Maintenance

Figure 1-2 shows a functmmal block dmgram of the KA655 CPU

Figure 1--2 KA655 CPU Functlonai Block Dmgmm
f’

Ac<16:2>

MY |

Ae<al2 | LEVEL " %
<31:2>%

{ ADDRESS LATCH|

‘

CPUAFPA

SUPPORT

SSCDAL<31:0>

SSCDAL XCVRS
. |

CDAL<31:0>

1.1

- MAIN MEMORY

INTERCONNECT

CONSOLE PANEL

CONTROLBUS

/1 Q22-bus

{ INTERFACE

‘

MLO-002368

KA655 CPU and Memory Subsystem

1-3

The KA655 has two variants:

KA655-AA. Runs multiuser software.

KA655-BA. Runs single-user software.

The KA655 is used in two systems:
.

MicroVAX 3800 (BA213 enclosure)

MicroVAX 3900 (H9644 cabinet)

Refer to BA213 Enclosure Maintenance and H9644 Cabinet Maintenance
for a detailed description of the enclosure and the cabinet.

CAUTION: Static electricity can damage integrated circuits. Always use a
grounded wrist strap (part no. 29-11762) and grounded work surface when
you work with the internal parts of a computer system.
The KA655 CPU module and MS650-BA memory modules combine to form
a VAX CPU and memory subsystem that uses the Q22-bus to communicate
with I/O devices. The KA655 and MS650-BA modules mount in standard
- Q22-bus backplane slots that implement the Q22-busin the AB rows and
the CD interconnect in the CD rows. The KA655 can support up to four
MS650-BA modules, if enough Q22-bus CD slots are available. Flgures 1-3
and 1-4 show the KA655 systemwlevel block diagram.
|

NOTE: The KA655 CPU supports only the MS650~-BA (16 Mbyte) memory
module. The MS650-AA (8 Mbyte) is not supported because of its slower
access speed.
The KA655 CPU communicates with the console device through the H3600SA CPU I/O panel, which contains configuration switches and an LED
display. The H3600-SAis describedin Section 1.4.

1-4

KAB55 CPU System Maintenance

§‘§

Figure 1-3:

KA655 System-Level Block Diagram, Part |

B0-Pay CABLE

cemm—

—

Micro VAX LOCAL MEMORY INTERCONNECT (32-brt dats + 7-bit ECC! .

CPUEPY
FIRST-LEVEL

SECOND-LEVEL

CACME

_CACHE
’

MSE50 BA

MEMORY

‘«

‘fl

MSE650 8A

M5650 BA

MODULE

MEMORY

mar

508

MODULE

MEMORY

MODULE

‘

027 pus

INTEREAC

)

‘

wwwww

CU SRR

i
N

[ S——

'mmmmmm

'»w

MicioVAX LOCAL MEMORY INTERCONNECT { ADDRESS )

[»]

————

¥ ALt BACKPLANE 5LO0TSY 02200 €D
022-bus

—

1 16.8)T DATA . 22 ADDRESS LINES . CONTROL . POWER |

‘

I
!

I
|

{

!
‘

R
B
e

e o A

SERIAL
LINE

/‘

CONTROLLER

!',g
!

N
HI104

CABLE
CONCENTRATOR

REMOTE TERMINAL

LOCAL TERM

MLO-DO2368

KA655 CPU and Memory Subsystem

1-5

Figure 1-4:

KA655 System-Level Block Diagram, Part Il

»”

| ornen

SYSTEMS

;

NE TWORK

OTHER

CONTROLLER

OPTIONS

INTERFACE

022-bus

BA200 SERIES
ENCLOSURE

Y e

w11

wln

F gg,. - ————
"

g =M

KFQSA

BEED

STORAGE

' “ '

TAPE

powen
SUPPLY

CONTROLLER

ADMAPTER

.‘ =

ocox ~3- |

RESET SWITCH

OPERATOR B

CONTROL

PANEL

TERMINATOR

TAPE ORIVE

‘?s: SERIESNG.

1-6

RFSERIES

KA655 CPU System Maintenance

1.2

KA655 CPU Features

The KA655 CPU provides the functionality of the KA650 CPU, but
reduces the cycle time from 90 to 60 nanoseconds (ns), placing the KA655
performance at about 3.8 VAX Units of Performance (VUPs).

The major features of the KA655 CPU are as follows:
]

A VAX central processor with a 33-MHz clock rate that supports the
MicroVAX chip subset of the VAX instruction set and data types, plus
the following string instructions: CMPC3, CMPC5, LOCC, SCANC,
SKPC, and SPANC. The processor also supports full VAX memory
management with demand paging and a 4-Gbyte virtual address space.

A floating point accelerator with the MicroVAX chip subset of the VAX
floating point instruction set and data types.

A two-level cache consisting of a 1-Kbyte, 60-ns, first-level cache and
a 64-Kbyte, 120-ns, second-level cache.
protection on the tag and data stores.

Both caches provide parity

A main memory controller that supports up to 64 Mbytes of 450-ns
ECC memory. The controller resides on the CPU module and supports
up to four MS650-BA memory modules, depending on the system
configuration.

~ A console port compatible with VAX processors. The console port has an
external baud-rate switch located on the CPU I/O panel (H3600-SA).

A set of processor clock registers that support:
—

A VAX standard time-of-year (TOY) clock with support for battery

backup. (The batteries are located on the inside of the H3600-SA.)

—

Aninterval timer with 10-millisecond (ms) interrupts.

—

Two programmable timers, similar in function to the VAX standard
interval timer.

A boot and diagnostic facility with four on-board LEDs. This facility
supports an external 4-bit display and configuration switches located

on the H3600-SA. It also supports 128 Kbytes of 16 bit-wide ROM
containing programs for:
-

Board initialization

Emulation of a subset of the VAX standard console

—

Power-up self-testing of the KA655 and MS650-BA modules

—

Booting from supported Q22-bus devices

TM.,

KA655 CPU and Memory Subsystem

1-7

—

Help utility

- —

" ¢

KFQSA programming utility

A Q22-bus interface that supports up to 16-word block mode transfers

between a Q22-bus DMA device and main memory, and up to 2-word
block mode transfers between the CPU and Q22-bus devices. This
interface contains:

-—

A 16-entry map cache for the scatter-gather map, which resides in
main memory. This 8192-entry map is used for translating 22-bit,
Q22-bus addresses into 26-bit, main memory addresses.

—

Interrupt arbitration logic that recognizes Q22-bus interrupt
requests BR7 through BR4.

—~

An interprocessor communications facility that supports communication between the Q22-bus arbiter and up to three auxiliary processors by means of doorbell interrupts.

1.2.1 Central Processing Unit (CPU)
The central processing unit (CPU) is implemented by the CVAX chip.
The CVAX chip contains approximately 180,000 transistors in an 84-pin
CERQUAD surface mount package. The chip achieves a 60-ns microcycle
~and a 120-ns bus cycle at an operating frequency of 33 MHz. The chip
also supports full VAX memory management and a 4-Gbyte virtual address
space.

. The CVAX chip contains all general purpose registers (GPRs) visible to the
VAX processor; the MSER, CADR, and SCBB system registers; the 1-Kbyte
first-level cache; and all memory management hardware, including a 28entry translation buffer.
The CVAX chip performs the following functions:
*

Fetches all VAX instructions

Executes 181 VAX instructions

Assists in the execution of 21 additional instructions

Passes 70 floating-point instructions to the CFPA60 chip

The remaining 32 VAX instructions (including h_floating and octaword) are
emulated in macrocode.

1-8

KA655 CPU System Maintenance

In addition, the CVAX chip provides the following subset of the VAX data
types:
-

‘Word
Longword
- Quadword
Character string
=~
Variable-length bit field
. Support for the remaining VAX data types can be provided through

macrocode emulation.

1.2.2 Clock Functions
Clock functions are implemented by the CVAX clock chip (CCLK). The
CVAX clock chip
is a 44-pin CERQUAD surface mount chip that contains
approximately 350 transistors. It provides the following functions:
*

Generates two MOS clocks for the CPU, the floating point accelerator,
and the main memory controller

Generates three auxiliary clocks for other TTL logic

Synchronizes the reset signal for the CPU, the floating point accelerator,
and the main memory controller

Synchronizes data ready and data error signals for the CPU, floatifig
point accelerator, and the main memory controller

1.2.3 Floating Point Accelerator
The floating point accelerator is implemented by the CFPA60 chip.
The CFPA60 chip contains approximately 60,000 transistors in a 68-pin
CERQUAD surface mount package. It processes f, d_, and g_floating
format instructions and accelerates the execution of MULL, DIVL, and
EMUL integer instructions. The CFPA60 chip receives opcode information
from the CVAX chip and receives operands directly from memory or from
the CVAX chip. The floating point result is always returned to the CVAX

1.2.4 Cache Memory
‘The KA655 CPU module contains a two-level cache membry to maximize
CPU performance.

The first-level cache is implemented within the CVAX chip. This cache is a

1-Kbyte, two-way associative, write-through cache memory. It has a 60-ns
cycle time.
%’%

KA655 CPU and Memory Subsystem

1-9

1¢*‘é

1.2.5 Memory Controlier
- The main memory controller is implemented by a VLSI chip called the
Lo

The second-level cache is implemented using 16K x 4-bit static RAMs. This
cache is a 64-Kbyte, direct-mapped, write-through cache memory. It has
~a 120-ns cycle time for longword transfers and a 180-ns cycle time for
* quadword transfers.

CMCTL. The CMCTL contains approximately 25,000 transistors in a 132pin CERQUAD surface mount package. It supports up to 64 Mbytes of
450-ns ECC memory.

The ECC memory resides on from one to four 16-Mbyte memory modules
(MS650-BA), depending on the system configuration.
The MS650BA modules communicate with the KA655 through the MS650 memory
interconnect, which utilizes the CD interconnect by means of a 50-pin
ribbon cable.

1.2.6 MicroVAX System Support Functions
System support functions are implemented by the system support chip
(SSC). The SSC contains approximately 83,000 transistors in an 84-pin
CERQUAD surface mount package. The SSC provides console and boot
code support functions; operating system support functions; timers; and

- many extra features, including the following:
¢

Word-wide ROM unpacking

1.Kbyte battery backed-up RAM

Halt arbitration logic

A console serial line

An interval timer with 10-ms interrupts

A VAX standard time-of-year (TOY) clock with support for battery
backup
|

An IORESET register

Programmable CDAL bus timeout

Two programmable timers

e A register for cénfrolling the diagnostic LEDs

1-10

KA655 CPU System Maintenance

1.2.7 Resident Firmware

" The resident firmware consists of 128 i{bytea of 16 bit-wide ROM, located on
~one 27510 EPROM. The firmware gains control when the processor halts.
It contains programs that provide the following services:
*

Board initialization

Power-up self-testing of the KA655 and MS650-BA modules

Emulation of a subset of the VAX standard console (automatic or
manual bootstrap, automatic or manual restart, and a simple command
language for examining or altering the state of the processor)

Booting from supported Q22-bus devices

Multilingual capability

The KA655 firmware is described in detail in Chapter 8.

1.2.8 Q22-Bus Interface
The Q22-bus interface is implemented by the CQBIC chip. The CQBIC chip
contains approximately 40,870 transistors in a 132-pin CERQUAD surface
mount package. The CQBIC chip supports block mode transfers as follows:
¢

Sixteen words on memory write

Four words on memory read

Longword transfers between CPU and Q22-bus

The Q22-bus interface contains the following:
*

A 16-entry map cache for the 8192-entry scatter-gather map that
resides in main memory. The map is used for translating 22-bit, Q22bus addresses into 26-bit, main memory addresses.

Interrupt arbitration logic that recognizes Q22-bus interrupt requests

BR7 through BR4.

The Q22-bus interface handles programmed and power-up resets, and CPU
halts (deassertion of DCOK).
The KA655-AA module contains 240-ohm termination for the Q22-bus.

KA655 CPU and Memory Subsystem

1-11

1.3 KA655 Connectors.

The KA655 CPU module has three connectors:
e J1. For the cable from the internal console SLU connector on the
- H3600—SA CPU IO panel.

,fl

;’/

”,/‘

o
.

J2. For the cable from the configuration and display connector on the
H3600-SA. (Connects the CPU to the three switches and LED display

- on the I/O panel.)

J3. For a cable from the first MS650-BA memory module.

The orientation of cor;nectors J1, J2, and J3 is shown in Figure 1-5.
Figure 1-5:
g

[

IR O

KAG655 Pin Orlentation
1

]

[ E N

;OQ*O*

EX BN ENE ]

Ifl.&fl‘tlt‘fi

Q o’ooa

( ZF R

B EENEENERENNNENYENRNESMNZSYN]

QQt".i‘flmfi'&'fl.'#‘&‘.t&'.

\
10

8421-

MLO-002371

1.4 H3600-SA CPU /O Panel
The H3600-SA CPU I/O panel, shown in Figure 1-6, has a ribbon cable
with two connectors that plug into the KA655 module console SLU and
console control connectors. The H3600-SA fits over backplane slots 1 and
2, covering both the KA655 module and the first of up to four possible
MS650-BA memory modules.
The controls and connectors on the H3600—-SA are shown in Figure 1-6 and
listed in Table 1-1.

1-12

KAB655 CPU System Maintenance

i
S

pa—— .

Flgure 1-6:
e

H3600—SA CPU IIO Panel
POWER UP MODE SWITCH
LANGUAGE INQUIRY

NORMAL OPERATION

(FACTORY SETTING)

(3 LOOPBACK TEST

CABLETO
y

CPU

BREAK ENABLE SWITCH —
{(DOWN=DISABLED)
LED DISPLAY
—
SLU CONNECTOR . %

BAUD RATE
SELECT SWITCH

{8600 1S FACTORY

SETTING)

BATTERY BACKUP
UN!T (BBU)

L!ST OF SWITCH SETTINGS

<71l FOR BAUD RATES
|

0=300

1=600
2=1200
3=2400

MLO-002372

KA655 CPU and Memory Subsystem

1-13

B
.

Table 1-1

H3600~SA Controls and Connectors .

, Outside
Modified modular jack (MMJ)
SLU connector

Power-up mode switch
Hex LED display

Enable/disable switch

Inside
o

Cable to SLU connector on KA655

Battery backup unit (BBU)
Cable to console control connector on KA655

Baud rate select switch

Enable/Disable Switch
Although the KA630, KA650, and KA655 CPU modules all use the H3600—
SA, the function of the enable/disable switch is slightly different in each

When the H3600-SA is connected to a KA630 CPU, the switch enables
and disables halts from the [BREaK] key on the console keyboard and from

the halt button on the system front panel. In MicroVAX II systems, the
switch is referred to as the halt enable switch.

When the H3600-SA is connected to a KA650 or KA655 CPU, the switch

enables and disables halts from the
key on the console keyboard
only. You cannot disable halts initiated from the halt button on the
system front panel. In MicroVAX 3500, 3600, 3800, and 3900 systems,
the switch is referred to as the break enable switch.

You also select the power-up mode and console terminal baud rate using
- switches on the H3600-SA. Release the H3600-SA quarter-turn fasteners
and turn the H3600-SA around (do not disconnect the ribbon cable) to
change the baud rate of the console serial line, or to change the batteries for
the backup unit. The factory setting for the baud rate is 9600. Figure 1-6
shows the possible baud rate settings.

Hexadecimal LED Display
The LEDs display a hexadecimal number for each power-up test and stage
of the bootstrap process. Chapter 4 lists the meanings of these numbers.
Modified Modular Jack SLU Connector (Outside)

- - The
modified modular jack (MMJ) is a six-pin connector for a cable that
connects to the console terminal.

1-14

KA655 CPU System Maintenance

Battery Backup Unit (Inside)
When the system is turned off, the battery backup unit (BBU) pmvxdes |
power to the KA655 t:me-of«year logic (25.6-kHz oscillator, TODR register,
and 1 Kbyte of RAMin the SSC) The 1 Kbyte of RAM stores the code for
the language thatis dmplayedin the console messages. If the BBU fails,
the codeis lost.
Cable

" The cable contains two connectors that plug into the console SLU and

console control connectors on the KA655 module. The 20-conductor end
connects the baud rate select switch, the power-up mode switch, and the
hexadecimal LED display to the console control connector (J2) of the KA655.
TKT 10~conducwr end connects the console serial line to connector J1 on the
655.

Table 1-2 lists the H3600-SA switch functions.
Table 1-2:

H3600-SA CPU /O Panel Switches

Switch

Position

Function

Enable/disable
(two-position toggle)

Dot outside .
circle

Breaks are disabled (factory setting). On
power-up or restart, the system tries to

Dot inside

Breaks are enabled. On power-up or restart,

load software from one of the devices after

completing the power-up diagnostics.

circle
Power-up mode
(three-position
rotary)

the system enters console /O mode after

completing the power-up diagnostics.
Run (factory setting). If the console terminal
supports the DEC multinational character set
(MCS), the system prompts the user for the
console language if the battery backup has
failed and upon initial power-up of a system
containing a new CPU. All start-up diagnostics

Arrow

run.

Face

Language inquiry. If the console terminal

T inside

Test.

supports the DEC MCS, the user is prompted
for the console language on every power-up and
restart. All power-up diagnostics run.

circle
Baud rate select
S

300 to
38,400
|

ROM programs run the wraparound

console serial line unit (SLU) tests.

Sets the baud rate of the console terminal serial
line. The factory setting is 9600 baud. The
baud rate of this switch must match the rate
of the console terminal.

KA655 CPU and Memory Subsystem

1-15

1.5 MS650-BA Memory
. The MS650-BA (M7622-A) is a 16-Mbyte memory module that provides
" memory for the KA655 CPU module. The MS650-BA is a nonintelligent
memory array module controlled by a custom memory controller chip

(CMCTL) on the KA655 CPU module.

The quad-height MS650-BA, shown in Figure 1-7, has a 450 ns, 39 bit-

wide array (32-bit data and 7-bit ECC), implemented with 1 Mbit dynamic
RAMs in surface mounted SOJs.
Ordering Information

MS650-BA
MS650-BF

16-Mbyte module only (M7622-A).
Option installation kit for BA200-series enclosures. Includes MS650-BA, filler
panel assemblies, blank cover, CPU cable, labels, and installation guide.

Diagnostic Support
MicroVAX Diagnostic Monitor

Release 126 (version 3.01)

Self-test

KAG6655 self-test

1-16

KAB55 CPU System Maintenance

Figure 1-7:

MS650-BA Memory Module (M7622-A)

o0 04!

2
I
I

fg{

0
B

= N

N«

» I

= O

O« O«

JC_pC_JC_JoC_

}gn de0L
_JL_0L_JL_
P L:g.—é%-@-.—-}“%g]fi
»

=000

000

Lo
ol 3 oC C
R

| DO
|N

000

B e O e O e

N
N
& N
IN

wC_

I JoC_Jo_]

|
|

IS« B

A+ B A

0o Joo D0 coJoooJon [p

prC

=
=
B« A

B
B

30300300

MLO-002373

KA655 CPU and Memory Subsystem

1-17

it >

A
A

Chapter 2

Configuration
2.1 Introduction

This chapter describes the guidelines for changing the configuration of a
KA655 system, and for configuring a dual-host system.

Before you change the system configuration, you must consider the
following factors:
Module order in the backplane
Module configuration
Mass storage device configuration
If you are adding a device to a system, you must know the capacity of the
system enclosure in the following areas:
Backplane
I/O panel
Power supply
Mass storage devices

2.2 General Module Order
The order of modules 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 gaining access to the bus
|
|
*

Configuration

2-1

2.2.1 Module Order Rules for KA655 Systems

. Obserée/the 'fo%llbwing rules about module order:
*

Install the KA655 CPU in slot 1.

*
-

Install a MS650-BA memory module in slot 2. Install any additional
MS650-BA memory modules in slots 3 through 5.

Do not install dual-height modules in the CD rows.

The Q22-bus does not pass through the CD rows of the backplane in a
BA200-series enclosure. Install all Q22-bus modules in the AB rows. Install
dual-height grant cards in the AB rows only, or single-height grant cards
in the A row only.

2.2.2 Recommmfi&ad Module Order for KA655 Systems
Here is the recommended module order for a KA655 system:

KA655
MS650-BA
AAV11-SA
ADV11-SA
AXV11-SA
KWV11-SA
DRV1J-SA
TSV05-SA
DMV11
LNV21
DEQNA/DELQA/DESQA~SA
DPV11-SA
KMV1A-SA, -SB, -SC
DZQ11-SA
DFA(O1-AB
CXY08-AA

CSF32-M
LPV11-SA
DRVIW-SA

- IEQ11-SA

ADQ32-M
DRQ3B-SA
IBQO1-SA

2-2

KA655 CPU System Maintenance

- TQK50-SA/TQK70-SA
RQDX3-SA
KDAS0-SA
KFQSA-SA
‘MS060-YA

2.3 Module Configuration
Each module in a system must use a unique 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 and interrupt vector values. The value of a floating address
depends on what other modules are housed in the system.

Set CSR addresses and interrupt vectors for a module as follows:
1.

Determine the correct values for the module with the CONFIGURE
command at the console I/O prompt (>>>).
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:
>>>

configure

Enter device configuration,
Device, Number? help

EELP,

or EXIT

Devices:

LPV11l

KXJ11

DLV11J

Dz2Q11

RLV12

DzZV11l

TSV05

RXV21

DMV11l
RRD50

DRV11W

DELQA
RQC25

DRV11B

DEQNA
KFQSA-DISK

DPV11

RV20

DESQA
TQK50

KFQSA-TAPE

RQODX3
TQK70

KDAS50
TUSB1E

IEQ1l

DHQ11

DHV11

CXAl6
LNV21

CXB16
QPSS

KMV1l

CXYO08
DSvV1l

VCBO1
ADV11C

QVSsSSs
AAV1IC

DFAQ1

LNV11
AXVi1lC

KWV11C

ADV11D

AAV11D

DRQ3B

Vsv2l

IBQO1

IDV11aA

IDV11B

IDV11D

IDV11C

IAV11A

IAV11B

MIRA

ADQ32

DESNA

DTCO4

IGQ1l1l

VCBO02

QDSS

DRV11J

Numbers:

1l to 255, default is 1
Device,Number? exit

See the CONFIGURE command in Chapter 3 (Section 3.9.2) for an

example of how to set the correct CSR addresses and interrupt vectors.

Configuration

2-3

‘The LPV11-SA, which is the LPV11 version compatible with the BA200series enclosures, has two sets of CSR address and interrupt vectors.
To determine the correct values for an LPV11-SA, enter LPV11,2 at the
DEVICE prompt for one LPV11-SA, or enter LPV11.4 for two LPV11SA modules.

See Appendix A for instructions on how to configure the KFQSA storage

adapter. Appendix A explains how to do the following:

Set a four-position switchpack on the KFQSA before you install it

Program the CSR addresses for all the system’s DSSI devices into
the EEROM on the KFQSA

Reprogram the EEROM when you add additional DSSI devices

3. See Microsystems Options for CSR and interrupt vector jumper or
switchpack settings for supported options.

2.4 DSSI Configuration
The KA655 CPU and memory subsystem support RF-series integrated
storage elements (ISEs) and the KFQSA storage adapter.

The RF-series ISEs are part of a series of mass storage devices based
on the Digital Storage Architecture (DSA). These devices use the Digital
Storage System Interconnect (DSSI) bus and interface. The term integrated
storage element is used because each DSSI ISE has an on-board intelligent
controller and mass storage control protocol (MSCP) server in addition to
the drive and the control electronics.
|
DSSI supports up to seven ISEs daisy-chained through a single cable to the
KFQSA storage adapter. The KFQSA is a protocol converter that supports
Q-bus protocols to and from the KA655 CPU and DSSI bus protocols to and
from the ISEs. The KFQSA contains the addressing logic required to make
a connection between the host and a requested ISE on the DSSI bus. DSSI
adapters can also be embedded on a CPU module, such as on the KA640.
Each ISE must have a unique node ID. The ISE receives its node ID from a
plug on the operator control panel (OCP) on the front panel of the BA200series enclosure. By convention, DSSI devices are mounted in the BA200- series enclosures from right to left, as listed in Table 2-1.

2-4

KAGB55 CPU System Maintenance

w5y

;*i

e % g
i

Table 2---1

DSSI Device Order

Device

- Position

Node ID!

~ * First

Right side

Third

Leftside =

Second

Center

1KA655 node ID = 7

If the cable between the ISE and the OCPis disconnected, the ISE reads the
node ID from three DIP switches on its electronics contml module (ECM).

The node ID switches are located behind the 50-pin connector on the ECM.
Switch 1 (the MSB)is nearest to the connector. Refer to the RF30 or RF71
sectionin Microsystems Options for more information. Table 2-2 lists the
switch settings for the eight possible node addresses.

Table 2-2:

RF-Series ISE Switch Settlngs
Switch!

DSSI

Node ID

1(MSB)

- 8 (LSB)

Down

3
4

Down

Down
Up

Down

1Up = toward the head disk assembly (HDA); Down = toward the drive module
2Normally reserved for the host adapter

NOTE: Pressing the system reset button on the front of a BA200-series
enclosure has no effect on ISEs. Ifyou change a node ID, you must perform
a powercycle to enable the new node ID to take effect.

The VMS operating system creates DSSI device names accordmg to Dlan,
where a is the controller letter (A, B, C, and so on) and n is the unit number.

Configuration

2-5

‘You can gain access to local programs-in the ISE through the MicroVAX
-

Diagnostic Monitor (MDM) or through the console I’/0O mode SETHOST/DUP command. This command creates a virtual terminal connection
to the storage device and the designated local program using the Diagnostic
and Utilities Protocol (DUP) standard dialogue. Section 3.9.15 describes the
console I/O mode SET/HOST/DUP/UQSSP command and shows an example
of how to set host to the RF71 ISE through the KFQSA storage adapter.

2.4.1 Changing RF-Serles ISE Parameters
Each ISE has a node name that is maintained in EEPROM on board the
controller module. This node name is determined in manufacturing from
an algorithm based on the device serial number. You can change the node
name of the DSSI device to something more meaningful by following the
procedure in Example 2-1. In the example, the node name for the RF71
ISE at DSSI node address 1 is changed from R3YBNE to DATADISK.
See Section 4.8.5 for more information about the PARAMS local program.

Example 2-1:
>>>

sho

Changing a DSSI Node Name

ugssp

UQSSP Disk Controller 0

-DUAO

UQSSP Disk Controller 1
-DUBl1

(772150)

(RF71)
(760334)

(RF71)

!The node name for this drive
!will be changed from R3YBNE
!to DATADISK.

>>>

set

host/dup/ugssp/disk

Starting DUP server...
Copyright

1988

DRVEXR V1.1

DRVTST V1.1

Digital Equipment Corporation
6-MAR-1989 15:33:06
6-MAR-1989 15:33:06

HISTRY V1.0

6-MAR-1989

ERASE

V1.3
PARAMS V1.2

D
D

6-MAR-1989 15:33:06
6-MAR-1989 15:33:06

15:33:06

DIRECT V1.0

6-MAR-1989 15:33:06

End of directory
Task Name? params

wcapyright

1988

Digital Equipmant Cdrporatian -

Example 2-1 Cont’d. on next page

2-6

KA655 CPU System Maintenance

Example 2-1 (Cont.):

Changing a DSSI Node Name

PARAMS> sho nodename
».

Parameter

Current

NODENAME

R3YBNE

PARAMS> set

Type

RF71

String

Radix

Ascii

nodename datadisk

PARAMS> write
Changes

Default

!This command writes the change
!to EEPROM.

require controller initialization,

ok?

([Y/(N)]

Stopping DUP ser@er...

>>>

sho ugssp

UQSSP Disk Controller 0
-DUAO (RF71)

(772150)

UQSSP Disk Controller 1
-DUB1 (RF71)

(760334)

!The node name has changed
!from R3YBNE to DATADISK,

!lalthough the
'unchanged.

display remains

2.4.2 Changing the Unit Number
By default, the ISE assigns the unit number to the same value as the DSSI
- node address for that device. This occurs whether the DSSI node address
is determined from the OCP unit ID plugs or from the three DIP switches
on the ISE controller module.

RF-series ISEs conform to the DIGITAL Storage Architecture (DSA). Each
ISE can be assigned a unit number from 0 to 16,383 (decimal). The unit
number need not be the same as the DSSI node address.
Example 2-2 shows how to change the unit number of a DSSI device. This
example changes the unit number for the RF71 ISE at DSSI node address
1 from 1 to 50 (decimal). You must change two parameters: UNITNUM
and FORCEUNI. Changing these parameters overrides the default, which
assigns the unit number the same value as the node address.

See Section 4.8.5 for more information about the PARAMS local program.

Configuration

2-7

Example 2-2:
>>>

s
sho

Changing a DSSI Unit Number

udsép |

‘UQSSP Disk Controller 0
-DUAQ

(772150)

(RF71)

UQSSP DiskController
-1 (760334) !The unit number for this
-DUB1

)

(RF71)

!drive will be changed from

1 to 50

>>> set

(DUB1 to DUBSO0).

host/dup/ugssp/disk 1

Starting DUP server...

1988 Digital Equipment Corporation

DRVEXR V1.1

6~-MAR-1989 15:33:06

DRVTST V1.1

D
D

6-MAR-1989
6-MAR-1989

HISTRY V1.0

ERASE

V1.3

15:33:06
15:33:06

D 6-MAR-1989 15:33:06

PARAMS V1.2

6-MAR-1989

15:33:06

DIRECT V1.0

D
6-MAR-1989 15:33:06
End of directory
Task Name? params

1988

Digital Equipment Corporation

sho unitnum

Parameter

Current

UNITNUM

Default

Type

Radix

Word

Dec

Type

Radix

PARAMS> sho forceuni
Parameter

Current

FORCEUNI

Default

Boolean

PARAMS>

set unitnum 50

PARAMS>

set

forceuni

PARAMS> write

0/1

!This command writes the changes to EEPROM.

PARAMS> ex

Exiting...
Task Name?

Stopping DUP server...
>>>sho

ugssp

UQSSP Disk Controller 0
-DUAO (RF71)
UQSSP Disk Controller 1
-DUB50 (RF71)

(772150)

(760334)
.

2-8

KAB55 CPU System Maintenance

!The unit number has changed
!from 1 to 50.
The node ID
!remains at 1.

".b

2.4.3 Changing the Allocation Class

If the system‘ is part of a cluster, yéi; must change the default allocation |

class parameter. The ISEs ship with the allocation class set to zero.
Determine the new allocation class for the RF-series ISEs according to the
rules on clustering.

NOTE: In a dual-host configuration, you must assign the same allocation
. class to both host systems and to the RF-series ISEs. This allocation class
must be different from that of other systems or of hierarchical storage

controllers (HSCs) in a cluster.

Change the allocation class and unit number parameters by setting host to
the console-based DUP driver utility. Example 2-3 shows how to change
the allocation class of a DSSI device. This example displays the existing
allocation class, then resets the allocation class to 2.
Example 2-3:
>>>set host

Changing a DSSI Allocation Class
/dup/ugssp/disk 0 params

Starting DUP server...
UQSSP Disk Controller 0
Copyright

(c)

(772150)

1988 Digital Equipment

Corporation

- PARAMS> sho allclass
Parameter

Current

ALLCLASS

Default
1

PARAMS>

set

allclass

PARAMS>

sho allclass

Parameter
ALLCLASS

Type

Radix

Byte

Dec

Type

Radix

Byte

Dec

Current

Default
2

PARAMS> write

Changes

require controller initialization,

ok?

[Y¥/

(N)

]

Stopping DUP server...

Configuration

2-9

204&4 DSSlcab"ng

A 50-conductor ribbon cable connects an RF-series ISE to

the DSSI bus

- (Figure 2-1). A separate five-conductor cable
carries +5 Vdc and +12 Vdc
to the drive from the enclosure power supply.
]

A 10-conductor cable connects the ISE connec
tor to the operator control
~ panel (OCP, Figure 2-2). In the BA213
enclosure, there are two cables
» that connect the power supplies to the OCP; one
cable connects to the right
power supply, and the other connects to the left
power supply.
These cables carry the ACOK signal (same
as POK) to the ISE. The OCP
delays this signal to one ISE for each power suppl
y in order to stagger the

start-up of one of two possible devices attach
ed to each supply. This delay
prevents the ISEs from drawing excessive
current at power-up.

The 50-conductor DSSI ribbon cable connects to a 50-con

ductor round cable

that is routed through the bottom of the mass
storage area to the KFQSA
storage adapter.
|

CAUTION: When removing or installing new
ISEs, be sure to connect the
rightmost connector of the DSSI ribbon cable
to the round cable connected
to the KFQSA. Do not “T” the bus by connecting
the round connector to any
of the ribbon cable’s center connectors.
|
~
o

2-10

KA655 CPU System Maintenance

Figure 2-1:

DSSI Cabling, BA213 Enclosure

}l‘

, /
‘

:,i,

’

‘

;o
.

FROM POWER

SUPPLY TO

;

"\\
N\

RF-SERIES ISE

DSSI BUS

TERMINATION

| | RF-SERIES ISE
TO RF3

) POWER-UP
“

,'p.

FROM POWER

SUPPLY TO
-

RF-SERIES ISE

';' { "‘!v' ‘

| JLJII

\ -

b 8
:“

BACKPLANE

MLO-002374

Configuration

2-11

Figure 2-2:

RF-Series ISE Op_arator Control Panel (OCP) |

TO POK LEAD

LEFT POWER
SUPPLY

10-PIN

RIGHT POWER
SUPPLY

TO BACK PLANE

DRIVE SELECT

/" PLUGS

‘

10-PIN

DRIVE FAULTS

L~ (RED)

WRITE-PROTECT

TO RF1

~~ BUTTONS
ol

O o
m_z

°
10-PIN
TO RF2

_~ READY BUTTONS

)
SYSTEM DC
L~ OK (GREEN)

.
°

[_—"-""—""l

o]
RESTART

MLO-000127

2.4.4.1 DSSI Bus Termination and Length
The DSSI bus must be terminated at both ends. The KFQSA
storage
adapter terminates the DSSI bus at one end. A terminator
connected to
the 50-conductor Honda connector on the left side of the media
faceplate

terminates the bus at the other end. You can remove this termina

need to expand the bus.

tor if you

CAUTION: Canfiect the DSSI bus vuéing cables approved by Digital. Use

approved configurations only.

In a dual-host system, described below, the second KFQSA storage adapter
provides the bus termination.

2-12

KA655 CPU System Maintenance

2.4.5 Dual-Host Capability

An RF-series ISE has dual-host capability built into the firmware, which

allows the device to maintain connections with more than one DSSI adapter.
You can connect more than one KFQSA storage adapter to the same DSSI
bus to allow each KFQSA access to all other devices on the bus.

The primary application for such a configuration is a VAXcluster system

that uses Ethernet as the interconnect medium between the boot node and
. the satellite members. This configuration improves system availability, as
described below.
Two kernel systems are connected through an external DSSI cable
(BC21M). For this discussion, a kernel system includes the following:

KA655 CPU
MS650-BA memory
KFQSA storage adapter
Ethernet adapter
Q22-bus
If this dual configuration is used to boot a number of satellite nodes, the
system disk resides in one of these enclosures and serves as the system disk
for both kernel systems. The KFQSA storage adapter in each enclosure
lzwasl equal access to the system disk and to any other DSSI ISE in either
enclosure.

If one of the kernel systems fails, all satellite nodes booted through that
kernel system lose connections to the system disk. However, the dualhost capability enables each satellite node to know that the system disk is
still available through a different path—that of the remaining operational
kernel system. A connection through that kernel system is then established,
and the satellite nodes are able to continue operation.
Thus, even if one KFQSA adapter or any other component of the kernel
system fails, the satellites booted through that system are able to continue

operation. In this case, the entire cluster will run in a degraded condition,

since one kernel system is now serving the satellite nodes of both systems.
Processing can continue, however, until Field Service can repair the
problem.

A dual-host system cannot recover from the following conditions:

- ®

System disk failure. If there is only one system disk, its failure causes |

the entire cluster to stop functioning until the disk failure is corrected.

ISE failure can be caused by such factors as a power supply failure in
the enclosure containing the ISE.

Configuration

2-13

“*E

DSSI cabling failure. If a failure in one of the DSSI cables renders
access to the ISEs impossible, the cable must be repaired
in order to
continue operation. Since the DSSI bus cabling is not redundant, a
cable failure usually results in a system failure.

2.4.6 Limitations to Dual-Host Configurations
- 'The following limitations apply to dual-host systems:
e

Because of cabling and enclosure limitations, you can connect a

maximum of two systems.

The DSSI bus sfipports eight devices or adapters. Since a dual-host

system has two KFQSA adapters, and each has a connection to the
DSSI bus, you can attach a maximum of six DSSI devices to the bus.

NOTE: You can make dual-host connections to the same type of DSSI
adapters only. For example, KFQSA to KFQSA or KA640 to KA640.

2.5 Configuration Worksheet
This section provides a configuration worksheet of the BA213 enclosure
(Figure 2-3). Use the worksheet to make sure the configuration does not
exceed the system’s limits for expansion space, I/O space, and power.

Table 2-3 lists power values for éupported devices. To check a system
configuration, follow these steps:
1.

List all the devices already installed in the system.

List all the devices you plan to install in the system.

Fill in the information for each device, using the data listed in
Table 2-3.
|

Add up the columns. Make sure the totals are within the limits for the
enclosure.

In a BA213 enclosure, you must install a quad-height load module (M9060—
YA) in one of backplane slots 7 through 12 if the continuous minimum
current drawn on the second power supply is less than 5 amperes. If the
minimum current of 5 amperes is not reached, the power supply enters

error mode and shuts down the system.

2-14

KAB55 CPU System Maintenance

*’%

Table 2-3:
-

|
.

Current (Amps)

Power

Bus Loads

Option

Module

+5V

+12V

Watts

AAV11-SA
ADV11-SA
AXV11-SA

A1009-PA
A1008-PA
A026-PA

1.8
3.2
2.0

0.0
0.0
0.0

9.0
16.0
10.0

2.1
2.3
1.2

0.5
0.5
0.3

CXA16-AA/~AF

CXB16-AA/~AF

CXY08-AA/-AF

DELQA-SA

DFAO1-AA/-AF
DPVI1-SA

DRQ3B-SA
DRV1J-SA
DRVIW-SA
DSV11-SA
DZQ11-SA
IBQO1-SA
IEQ11-SA
KA655-AA
KDA50-Q
KDA50-Q
-

Power and Bus Loads for KA655 Options

KFQSA-SA

KLESI-SA
KMV1A-SA
KWV11-SA
LPV11-SA
M9060-YA
MS650-AA
RF30-SA
RF71E-SA
TK50E-EA
TK70E-EA
TQK50
TQK70-SA
TSV05-SA

M3118-YA

M3118-YB

M3119-YA

M7516-PA
M3121-PA
M8020-PA

M7658-PA
M8049-PA
M7651-PA
M3108-PA
M3106-PA
M3125-PA
M8634-PA
M7625-AA/-BA
M7164
M6165
M7769

- M7740-PA
M7500-PA
M4002-PA
M8086-PA
-~
M7621-A
M7546
M7559
M7196

1.6

0.20

10.4

3.0

2.0

0.5

0.0

10.0

3.0

0.5

1.64

2.7

0.395

12.94

3.0

0.5

1.97

19.5

0.40

2.2

14.7

1.2

3.0

0.30

1.0

9.6

1.0

45
1.8
1.8
5.43
1.0
5.0
3.5
8.7
693
657
5.5

3.0
2.6
2.2
1.6
5.3
2.7
1.10
1.25
1.35
1.5
2.9
8.5
6.5

0.0
0.0
0.0
0.69
0.36
0.0
0.0
0.14
0
003

22.5
9.0
9.0
38.0
9.3
25.0
17.5
21.0
34.6
8321

0.0

27.0

0.0
0.2
0.13
0.0
0.0
0.0
0.80
4.54
2.4
2.4
0.0
0.0
0.0

15.0
15.4
11.15
8.0
26.5
13.5
15.1
26.5
85.6
36.3
14.5
17.5
32.5

0.5

2.0
2.0
2.0
3.6
1.4
4.6
2.0
2.2
3.0
-

1.0
1.0
1.0
1.0
0.5
1.0
1.0
1.0
0.5
-~

3.8

0.5

2.3
3.0
1.0
1.8
0.0
0.0
2.8
4.3
3.0

1.0
1.0
0.3
0.5
0.0
0.0
0.5
0.5
1.0

Configuration

2-15

Fy
-

Figure 2-3: BA213 Configuration Workshwt

...’

RIGHT POWER SUPPLY |
SLOT

MODULE

Current (Amps)

Power

+5 Vdc | +12 Vdc | (Watts)

E WY

ol £9
e

‘\‘;

MASS STOBAGE:
TK Drive

FIXED DISK
Total these columns:
Must not exceed:

33.0 A

2300 W

Current (Amps)

Power

LEFT POWER SUPPLY
|

SLOT

MODULE

+5 Vdc | +12 Vdc | (Watts)

10
11

12
MASS STORAGE:
FIXED DISK(S)

Total these columns:

Must not exceed:

330A |

76A

(2300W
MLO-001288

2-16

KA655 CPU System Maintenance

V Chapter' 3
KAG655 Firmware
- 3.1 Introduction
This chapter describes the KA655 firmware, which gains control of the
processor whenever the KA655 performs a processor halt. A processor
halt transfers control to the firmware; the processor does not actually stop
executing instructions.

3.2 KA655 Firmware Features
The firmware is located in one 128-Kbyte EPROM on the KA655. The
firmware address range (hexadecimal) in the KA655 local /O space is
20040000 to 2007FFFF, inclusive (20040000-2005FFFF is halt-protected
space and 20060000-2007FFFF is halt-unprotected space). The firmware
displays diagnostic progress and error reports on the KA655 LEDs and on
the console terminal.
|
|

The firmware performs the following functions:
*

Automatic or manual bootstrap and restart of an operating system

Aninteractive command language that allows you to examine and alter
the state of the processor.

Diagnostics that test all components on the board and verify that the
module is working correctly.

Support of various terminals and devices, such as the system console.

Multilingual support.
several languages.

following processor halts.

The firmware can issue system messages in

To allow the console program to operate, the processor must be functioning
at a level such that it is able to execute instructions from the console
program ROM.

KAB55 Firmware

3-1

The firmware consists of the following major functional areas:
<

Halt entry, exit, and dispatch code
Bootstrap
- Console 1/0 mode
Diagnostics

The halt entry, exit, and dispatch code; bootstrap; and console I/O mode are
Qeacribed in this chapter. Diagnostics are described in Chapter 4.

3.3 Halt Entry, Exit, and Dispatch Code
Whenever a halt occurs, the processor enters the halt entry code at physical
address 20040000. The halt entry code saves the machine state, then
transfers control to the firmware halt dispatcher.

After a halt, the halt entry code saves the current LED code, then writes
an E to the LEDs. An E on the LEDs indicates that at least several
instructions have been successfully executed, although if the CPU is
functioning properly, the E flashes too quickly to be seen.

The halt entry code saves the following registers. The console intercepts
any direct reference to these registers and redirects it to the saved copies:
RO-R15

'PR$_SAVPSL

General purpose registers

Saved processor status longword register

PR$_SCBB

System control block base register

DLEDR
SSCCR
ADxMCH
ADxMSK

Diagnostic LED register
SSC configuration register
SSC address match registers
SSC address mask registers

The halt entry code unconditionally sets the following registers to fixed
values on any halt to ensure that the console itself can run and to protect
the module from physical damage:
SSCCR
ADxMAT
ADxMSK

SSC configuration register

CBTCR

CDAL bus timeout control register

TIVRx

SSC timer interrupt vector registers

SSC address match registers
SSC address mask registers

‘When the processor exits the firmware and reenters program mode, the
saved registers are restored and any changes become operative only then.
References to processor memory are handled in the normal way. The binary
load and unload command (X, Section 3.9.19) cannot reference the console
memory pages.

After saving the registers, the halt entry code transfers control to the halt
dispatch code. The halt dispatch code determines the cause of the halt

3-2

KA655 CPU System Maintenance

by reading the halt field (PR$_SAVPSL <13:08>), the processor halt action

field (PR$_CPMBX <01:00>), and the break enable switch on the H3600—

SA panel. Table 38-1 lists the actions taken, by sequence. If an action
fails, the next action is taken, with the exception of bootstrap, which is not
attempted after diagnostic failure.

Table 3-1: Actions Taken on a Halt
Breaks Enabled
on H3600-SA

Power-up Halt!

Halt Action?

Halt

Diagnostics, bootstrap, halt
Restart, bootstrap, halt

Action

Diagnostics, halt

Restart, halt

Bootstrap, halt

Halt

1Power-up halt: PR$_SAVPSL<183:08>=3

2Halt action: CPMBX<01:00>
3T = condition is true, F = condition is false, X = does not matter

3.4 External Halts
The following conditions can trigger an external halt, and different actions
are taken depending on the condition:
*

The break enable switch is set to enable, and you press [BREAX] on the
system console terminal.

Assertion of the BHALT line on the Q22-bus, if the SCR<14>(BHALT
_
ENABLE) bit in the CQBIC is set.

Negation of DCOK, if the SCR<7>(DCOK_ACT) bit is set.

The KA655 cannot detect the deassertion of DCOK when in console I/O
mode, 80 no action is taken. More important, however, the deassertion of
DCOK destroys system state without notifying the firmware.

CAUTION: Do not press the Restart button while in console I/0 mode.

'Doing so will destroy the previously

saved system state.

The action taken by the halt dispatch code on a console [BRERR or Q22-bus

BHALbl'li‘ is the same: the firmware enters console /O mode if halts are
enable

KA655 Firmware

3-3

The halt dispatch code distinguishes between DCOK deasserted and

BHALT by assuming that BHALT must be asserted for at least 10 msec,
- .and that DCOKis deasserted for at most 9 usec. To determine if the BHALT
" line is asserted, the firmware steps out into halt-unprotected space after
9 msec. If the processor halts again, the firmware concludes that the halt
was caused by the BHALT and not by the deassertion of DCOK.

3.5 Power-Up Sequence
On power-up, the firmware performs several unique actions. It runms
the initial power-up test (IPT), locates and identifies the console device,
performs a language inquiry, and runs the remaining diagnostics.
Power-up actions differ, depending on the state of the power-up mode switch
on the H3600-SA (Figure 1-6). The mode switch has three settings: test,
language inquiry, and normal. The differences are described in Sections
3.5.1 through 3.5.3.
The IPT waits for power to stabilize by monitoring SCR<5>(POK). Once
power is stable, the IPT verifies that the console private nonvolatile
RAM (NVRAM) is valid (backup battery is charged) by checking

SSCCR<31>(BLO). If it is invalid or zero (batteryis discharged), then the
IPT tests and initializes the NVRAM.

After the battery check, the firmware tries to determine the type of termmal
attached to the console serial line. If the terminalis a known type, it is
treated as the system console.

3.5.1 Mode Switch Set to Test
Use the test position on the H3600-SA to verify a proper connection
between the KA655 and the console terminal:

To test the console terminal port, insert the H3103 loopback connector
into the H3600-SA console connector, and put the switch in the test
position. You must install the loopback connector to run the test.

To test the console cable, install the H8572 connector on the end of the
console cable, and insert the H3103 into the H8572.

During the test, the firmware togglea between the active and passive states:

During the active state (3 seconds) the LEDis set to 6 The firmware

During the passive state (5 seconds), the LED is set to 3.

reads the baud rate and mode swmtch then transmits and receives a
character sequence.

"'xm%\%

3-4

KAGE55 CPU System Maintenance

If at any time the firmware detects an error (parity, framing, overflow, or
no characters), the display hangs at 6. If the configuration switch is moved
from the test position, the firmware continues as if on a normal power-up.

3.5.2 Mode Switch Set to Language Inquiry

If the H3600—-SA mode switch is set to language inquiry, or the firmware

detects that the contents of NVRAM are invalid, the firmware prompts you
for the language to be used for displaying the following system messages:
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

Available devices.
Device?
Retrying network bootstrap.

specified.

The language selection menu appears under the conditions listed in
Tabcl;: 3-2. The position of the break enable switch has no effect on these
conditions.
|
|
v
|

Table 3-2:

Language Inquiry on Power-Up or Reset

Mode

Language Not
Previously Set!

Language
Previously Set

Language Inquiry

Prompt?

Normal

Prompt

No Prompt

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

The language selection menu is shown in Example 8-1. If no response is
received within 30 seconds, the firmware defaults to English.

KA655 Firmware

3-5

Example 3-1:
-

Language Selection Menu

Dansk

2) Deutsch (Deutschland/Osterreich)
3)

Deutsch

(Schweiz)

English

(United Kingdom)

English

, 6)
7)

Espafiol
Fran¢ais
Frangais
Frangais

(United States/Canada)

8)
9)
10)
11)

Italiano

12)

Norsk

Portugues
Suomi
Svenska

14)

(France/Belgique)
(Suisse)
’

Nederlands

13)
15)

(Canada)

(1..15):

In addition, the console may prompt you for a default boot device.
Section 3.6, Example 3-2.

See

After the language inquiry, the firmware continues as if on a normal power-

3.5.3 Mode Switch Set to Normal
The console displays the language selection menu if the mode switch is set
to normal and the contents of NVRAM are invalid.
The console uses the saved console language if the mode switch is set to
normal and the contents of NVRAM are valid.

3.6 Bootstrap
The KA655 supports bootstrap of VAX/VMS, ULTRIX-32, VAXELN, and
MDM diagnostics.
The firmware initializes the system to a known state before dispatching
:outhe primary bootstrap, called the virtual memory bootstrap (VMB), as
follows:

3-6

KA655 CPU System Maintenance

3.6.1 Bootstrap Initialization Sequence

Checks CPMBX&%(BIP), bootstrap in progress. If it is set, boofsirap

fails and the console displays the message Failure.
console language.

in the selected

If this is an automatic bootstrap, prints the message Loading system
software. on the console terminal.

Validates the boot ”device name. If none exists, supplies a list of
available devices and issues a boot device prompt. If you do not specify

a device within 30 seconds, uses ESAO.

o
e

Initializes the Q22-bus scatter-gather map.
Validates the PFN bitmap. If invalid, rebuilds it.

Sets CPMBX<2>(BIP).

Writes a form of this boot request, including active boot flags and boot
device (BOOT/R5:0 ESAOQ, for example), to the console terminal.

Searches for a 128-Kbyte contiguous block of good memory as defined

by the PFN bitmap. If 128 Kbytes cannot be found, the bootstrap fails.
Initializes the general purpose registers:
RO |

Address of descriptor of the boot device name or 0 if none specified

R2
R3
R4
R5
R10

Length of PFN bitmap in bytes
Address of PFN bitmap

Halt PSL value (without halt code and map enable)

AP
SP
PC

Base of 128-Kbyte good memory block + 512

R1, R6, R7, R8,
R9, fi FP

Time-of-day of bootstrap from PR$_TODR
Boot flags
Halt PC value
Halt code
Base of 128-Kbyte good memory block + 512
0

10. Copies the VMB image from EPROM to local memory, beginning at the

base of the 128 Kbytes of good memory block + 512 (decimal).

11. Exits from the firmware to VMB residing in memory.

VMB is the primary bootstrap for VAX processors. VMB loads the secondary

bootstrap image from the appropriate boot device and transfers control to
it.
N

KAB55 Firmware

3-7

Paat

362 VMBBootFlags

The VMB boot flags are listed in Table 3-3.

>
%

Table 3-3:
Bit

Virtual Memory Bootstrap (VMB) Boot Flags

Name

Description

RPB$V_CONV

Conversational boot. At various points in the system boot
procedure, the bootstrap code solicits parameters and other
input from the console terminal.

RPB$V_INIBPT _

Initial breakpoint.
If RPB$V_DEBUG is set, the VMS
operating system executes a BPT instruction in module INIT
immediately after enabling mapping.

RPB$V_BBLOCK

Secondary 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-11 bootstrap.

RPB$V_DIAG

Diagnostic bootstrap. When set, the load image requested is

RPB$V_BOOBPT

Bootstrap breakpoint. When set, a breakpoint instruction is
gfimted in VMB and control is transferred to XDELTA before

RPB$V_HEADER

Image header. When set, VMB transfers control to the address
specified by the file's image header.
When not set, 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. The maximum file
specification size is 17 characters.

RPB$V_SOLICT

[SYS0.SYSMAINTIDIAGBOOT.EXE.

RPB$V_HALT

Halt before transfer. When set, VMB halts before transferring
control to the application image.

31:28

RPB$V_TOPSYS

This field can be any value from 0 through F. This flag changes
the top-level directory name for system disks with multiple
operating systems. For example, if TOPSYS is 1, the top-level

directory name is [SYS1...].

3.6.3 Supported Boot Devices
Table 3—4 lists the boot devices supported by the KA655—-AA CPU. The table
correlates the boot device names expected in a BOOT command with the
corresponding supported devices.
-

Boot device names consist of a device code at least two letters (A t.hrough

- Z) in length, followed by a single character controller letter (A through Z),
and ending in a device unit number (0 through 16,383).

3-8

KA655 CPU System Maintenance

Table 3-4: Boot Devices Supported by the KA655-AA
BootName

Controller Type

- Device Type(s)

Disk

DUcn

KFQSA DSSI
-

RF30, RF71

RQDX3 MSCP

RD52, RD53, RD54, RX33, RX50

KDA50 MSCP

DLen

Tape

RA70, RA80, RA81, RA82, RA90

KLESI

RC25

RLV21

RLO1, RLO2

i
MUen

TQKS50 MSCP

TK50

TQK70 MSCP

TK70

KLESI

TUSBIE

Network
XQcn

DEQNA

DELQA

DESQA

MRV11

PROM
PRAO

3.6.4 Autoboot
IMPORTANT: Unless you specify otherwise, the KA655 default boot device
is the Ethernet adapter, XQmn. See Example 3-2.
*

If the Break Enable/Disable switch is set to disable, the CPU tries to
autoboot an operating system upon successful completion of the powerup self-tests.

The system looks for a previously selected boot device. If you have not
yet selected a boot device, the system issues a list of bootable devices
and prompts you to select a boot device from the list.
v

NOTE: You can also specify a default boot device by typing the SET BOOT
command (Section 3.9.1).

KA655 Firmware

3-9

dé'i

& Ifyou do not type a boot device name within thirty seconds, the system
|

boots from the Ethernet adapter, XQmn.

If you type a boot device name within thirty seconds, this device
becomes the default boot device and the system boots from that device,
as shown in Example 3-2.

7), o 'i

|/ A

Ty NOTE: For diskless and tapeless systems that boot software over the

network, select the Ethernet adapter only. All other boot devices are
inappropriate.

Example 3—2:

Selecting a Boot Device

Performing normal system tests.

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.
Loading system software.

No default boot device has been specified.
Available devices.
-DUAO
-DUB1

-MUAQ
-XQA0

(RF71)
(RF30)

(TK70)
(08-00-2B-09-95-21)

Device?

([XQA0]:

dual

(BOOT/R5:0 DUAO)
200

-DUAO
1“0'*

3-10

KA655 CPU System Maintenance

3.7 Operating System Restart
An gperating system restart is the process of bringing up the opératingv
system from a known initialization state following a processor halt.

o restart occurs under the conditions listed in Table 3-1, earlier in this
chapter.

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. If the firmware finds a valid RPB, it passes control to the
operating system at an address specified in the RPB.
The firmware keeps an RIP (restart-in-progress) flag in CPMBX, which it

uses to avoid repeated attempts to restart a failing operating system. The
operating system maintains an additional RIP flag in the RPB.

3.7.1 Restart Saquence
The firmware restarts the operating system in the following sequence:
1.

Checks CPMBX<3>(RIP). If it is set, restart fails.

Prints this message on the console terminal:

Restarting system softwaré.
3. Sets CPMBX<3>(RIP).

Searches for a valid RPB. If none is found, restart fails.
6.

Checks the operating system RPB$L_RSTRTFLG<0>(RIP) flag. If it is
set, restart fails.

Writes a 0 (zero) to the diagnostic LEDs.

Dispatches to the restart address, RPB$L_RESTART, with :

SP = the physical address of the RPB plus 512
AP = the halt code
PSL = 041F0000

PR$_MAPEN =0

If the restart is successful, the operating system must clear CPMBX<3>(RIP).
If restart fails, the firmware prints this message on the console terminal:
Failure.

KAG655 Firmware

3-11

3.7.2 Locating the RPB

The' RPB is a page-aligned control block

that can be identified by its

signature in the first three longwords:

- +00 (first longword) = physical address
of the RPB
+04 (second longword) = physical addre
ss of the restart routine
+08 (third longword) = checksum of first
31 longwords of restart routine

- The firmware finds a valid RPB as follows:
1.
2.

Searches for a page of memory that conta
ins its address in the first
longword. If none is found, the search for a
valid RPB has failed.
Reads the second longword in the page

restart routine).

(the physical address of the

If it is not a valid physical address,

returns to step 1. The check for zero

or if it is Zero,
is necessary to ensure that a page

of zeros does not pass the test for a valid

RPB,

Calculates the 32-bit two’s-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, returns to step
1.

If the sum matches, a valid RPB has

3-12

KA655 CPU System Maintenance

been found.

‘3.8 Console I/O Mode
- %

In cohsole /O mode several characters have special meamng

.[RUBOUT)

Also <CR>. The carriage return ends a command line. No action is taken on a
command until after itis 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 [RUBOUT] the console deletes the previously typed character. The
resulting display differs, depending on whether the console is a video or a hardcopy terminal.

For hard-copy terminals, the console echoes a backslash (\), followed by the
deletion of the character. If you press additional rubouts, the additional deleted
characters are echoed. If you type a non-rubout character, the console echoes
another backslash, followed by the character typed. The result is to echo the
characters deleted, mrmuudmg them with backslashes. For example:

The console echoes: EXAMI:E\ E;\ NE<CR>
The console sees the command line: EXAMINE<CR>

For video terminals, the previous character is erased and the cursor is restored
to its previous position.

The console does not delete characters past the beginning of a command line. If
~ you press more rubouts than there are characters on the line, the extra rubouts
are ignored. A rubout entered on a blank lineis ignored.

Echoes AU<CR>, and deletes the entire line. Entered but otherwise ignored if
typed on an empty line.
Stops output to the console terminal until [CTRUQ] is typed. Not echoed.

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 [CTRUO]is entered.
Echoes O when disabling output, not echoed when it reenables output. Outputis
reenabled if the console prmw an error message, or if it prompts for a command
from the termmal Output is also enabled by entering console IO mode, by
pressing
the
[BREAK]

3.8.1 Command Syntax
The console accepts commands up to 80 characters long. Longer commands
produce error messages. The chamcter count does not include rubouts,
rubbed-out characters, or the [FET An] at the end of the command.

KA655 Firmware

3-13

,§’i

You can abbreviate a command by enter
required to make the command uniqu

ing only as many characters as are

e. Most commands can be recognized

from their first character. See Table 3-8,

The console treats two or more consecutive space

s and tabs as a singl

e
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
(]

An optional qualifier or argument

)

A required qualifier or argument

argument conventions:

3.8.2 Address Specifiers
Several commands take an address or addre

defines the address space and the offset

‘supports six address spaces:

sses as arguments. An address
into that space.

The console

Physical memory
Virtual memory
Protected memory
General purpose registers (GPRs)
Internal processor registers (IPRs)
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.

3.8.3 Symbolic Addresses
The console supports symbolic references to addre
sses. A symbolic reference
defines the address space and the offset
into that space. Table 3-5 lists
symbolic

references supported by the console, group
ed according to address |
space. You do not have to use an addre
ss space qualifier when using a
symbolic address.
o
|
o

3-14

KA655 CPU System Maintenance

“

Table 3-5:
Symbol

Console Symbolic Addresses
Address

Symbol

Address

GPR Address Space (/G)

RO
R2

R4
R6
R8
R10

0
2

4
6

R11

R12

0oC

R14

R13

- OE

0C
OE

R15

PSL

FP
PC

0D
OF

pr$_esp

IPR Address Space (/)

pr$_ksp

pr$_ssp

00
02

pré_isp

pré$_usp

pr$_pOlr
pré_pllr
pr$_sir
pr$_scbb

- 09
0B
oD
11

pr$_pObr

pré_astlv
pr$_sisr

pré$_nicr
pr$_todr
pr$_rxdb

pr$_txdb
pr$_cadr

pr$_mser

pr$_savpsl
pr$_mapen

pré_this

pré_tbchk

13
15

19
1B
21

23
25

pré_plbr
pr$_sbr
pr$_pcbb
pr$_ipl

0A
0oC
10
12

pré_sirr
pré_icer

14
18

pr$_icr
pré_rxcs

pré$_txcs

pr$_tbdr
pr$_mcesr

pr$_savpce

pr$_sid

2B
38

pr$_ioreset
pr$_tbia

1A
20
22

24
26

37
39

KA655 Firmware

3-15

F e

Table 3—5 (com.): Console Symbotic"' Addrasses

. Symbol

Address

Symbol

Address

Physical Memory (/P)

qbio

20000000

gbmem

30000000

*. gbmbr
rom

20080010
20040000

cacr
dscr

20084000
20080000

20001¥40
20001F44
20140400
20140020
20140130
20140140
20140100
20140108

dsear

20084004
20080004

iper0
iper2
ssc_ram
ssc_cdal
ssc_adOmat
ssc_adlmat
ssc_ter0
ssc_tnir0

20080008

bdr
dser

iperl
iper3
8sc_Cr
ssc_dledr
ssc_adOmsk
ssc_adlmsk
ssc_tir0
ssc_tivr0

20001F42
20001F46
20140010
20140030
20140134
20140144
20140104
2014010C

ssc_terl
ssc_tnirl
memcsr0

20140110
20140118
20080100

gsc_tirl
ssc_tivrl
memesrl

20140114
2014011C
20080104

memcsr4
memcsro
memcsr8
memecsrl0
memecsrl2
memcesrl4
memcsrl6

20080110
20080118
20080120
20080128
20080130
20080138
20080140

memcsrd
memecsr7
memcsrd
memecsrll
memcsrl3d
memesrld
memcesrl?

20080114
2008011C
20080124
2008012C
20080134
2008013C
20080144

dmear

memcsr2

3-16

20080108

KA655 CPU System Maintenance

" memcsr3

2008000C

2008010C

Table 3-6:
M

iw g

Table 3-6 lists symbolic addresses that you can use in any address space.

- Symbol

Symbolic Addresses Used in Any Address Space

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 reference (1 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.

3.8.4 Console Command Qualifiers
You can enter console command qualifiers in any order on the command
line after the command keyword. There are three types of qualifiers: data
control, address space control, and command specific. Table 3-7 lists and
describes the data control and address space control qualifiers. Command
specific qualifiers are listed in the descriptions of individual commands.

KAB55 Firmware

3-17

‘

| Tab!e,3-7*

Qualifier

Console Command Qualiflers

Description

Data Control
/B

The data size is byte.

- /W

The data size is word.

The data size is longword.

The data size is quadword.

/N:{count}

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:(size]

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

Wrong. Used to override or set error bits when referencing main memory. On
writes, uses a value of 3, which forces a double bit error. On reads, ignores
ECC errors.

Address Space Control
/G

General purpow remu%r (GPR) address apaw, RO—-R15

Internal processor register (IPR) address space. Accessible only by the MTPR

Virtual memory address space. All access and protection checking occur. If
access to a program running with the current PSL is not allowed, the console

always longword.

The data size is -

and MFPR instructions. The data size is always longword.

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

addresses. Note that when you examine virtual memory, the address space
and address
in the response is the physical address of the virtual address.

Physical memory address space.

Processor status longword (PSL) address space.
longword.

Access to console private memory is allowed.

3-18

The data size is always

This qualifier also disables

virtual address protection checks. On virtual address writes, the PTE<M>
bit is not set if the /U qualifier is present. This qualifier is not inherited; it
must be respecified on each command.

KA655 CPU System Maintenance

3.8.5 Console Command Keywords

- %

Table 3-8 lists command keywords by type. Table 3-9 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.

Table 3-8: Command Keywords by Type
Processor Control

Data Transfer

BOOT

EXAMINE

CONTINUE
HALT
INITIALIZE
NEXT
START
UNJAM

DEPOSIT

FIND

MOVE
SEARCH
X

REPEAT
SET
SHOW

Console Control
CONFIGURE

TEST
'

Table 3-9: Console Command Summary
Command

Qualifiers

Argument

Other(s)

BOOT

CONFIGURE
CONTINUE

/R5:{bitmap) /{bitmap)

[device_name]

DEPOSIT

/B/WI/LIQ
IGANPMNO
/N:{count) /STEP:{size}
/WRONG

{address}

{data) [(data]]

EXAMINE

/B/W/L/Q

{{address)]

IGAN/PM/U
/N:(count) /STEP:(size)
/WRONG/INSTRUCTION

FIND
HALT

/MEM /RPB
|

INITIALIZE

KAB55 Firmware

3-19

Table 3-9 (Cont.): Console Command Summary
- Command

MOVE

Qualifiers -

Argument

B/W/LRQ

{src_address)}

{dest_address)

—

[{count}]
(command)
{start_address)

{pattern) [{mask}]

(bitmap}
{device_string)

N [P /U
/N:{count} /STEP:(size}
/WRONG

~NEXT

'REPEAT

BW/MLIKQ

NP/
/N:fcount) /STEP:(size)
/fWRONG/NOT

SET BFLAG

SET BOOT

SET HOST

/DUP /UQSSP
(/DISK n
/TAPE n csr_address)

Other(s)

{node) n
[{task]}]
{controller_number)

/MAINTENANCE /UQSSP

{(/SERVICE n csr_address)

SET LANGUAGE

SHOW BFLAG

SHOW BOOT

SHOW DEVICE

(language_type)

- SHOW ETHERNET

SHOW LANGUAGE

SHOW MEMORY

/FULL

SHOW QBUS

SHOW RLV12

SHOW UQSSP

SHOW VERSION

START

TEST

UNJAM

3-20

KA655 CPU System Maintenance

(address)

{test_number}

{{parameters}]
—-—

{address}

{count}

3.9 Console Commands
-

This, section describes thé console /0 mode commands.
commands at the console I/O mode prompt (>>>).

Enter the

3.9.1 BOOT

The BOOT command ifiitializes the processor and transfers execution to

VMB. VMB attempts to boot the operating system from the specified device,

“or fro the default boot device if none is specified. The console qualifies
the bootstrap operation by passing a boot flags bitmap to VMB in R5.

Format:

BOOT [qualifier-list] [device_name]
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, the corresponding default value.

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,

XQAO.

Qualifiers:
Command specific:
/R5:(boot_flags) A 82-bit hex value passed to VMB in R5. The console does not interpret this
value. Use the SET BFLAG command to specify a default boot flags longword.
Use the SHOW BFLAG commanad to display the longword. Table 3-3 lists the
supported R5 boot flags.
/{boot_flags}]

= Same as /R5:(boot_flags)

{device_name]

A character string of up to 17 characters. Longer strings cause a VAL TOO
BIG error message. Apart from checking the length, the console does not
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. Use the SHOW
BOOT command to display the default boot device. The factory default device

gA
the E%Kmat port, XQA0. Table 3—4 lists the boot devices supported by the
A655-,

KA655 Firmware

3-21

*
7
‘

Examples:

>>> show boot

DUAO
>>> show bflag

Boot using default boot flags and device.

>>> b

~_(BOOT/R5:0 DUAO)

f
o

bo2..
|

-DUAOD

>>> bo xqal

(BOOT/R5:0 XQAO0)

& -

4 :’/

Boot using default boot
specified device.

flags and

2..
=XQA0
>>> boot/10

(BOOT/R5:10 DUAO)

Boot using specified boot

default device.

flags and

2*’
-DUAO

>>> boot

/r5:220 xqal

(BOOT/R5:220 XQA0)

Boot using specified boot
device.

-XQA0

3-22

KAB855 CPU System Maintenance

flags and

e
T

3.9.2 CONFIGURE
The CONFIGURE command iuvdkes an interactive mode that permits
-

you 4o 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
with a running operating system. Refer to the example below and use the
CONFIGURE command as follows:

1. Enter CONFIGURE at the console L/O prompt.
2.

Enter HELP at the pevice,Number?

prompt to see a list of devices

whose CSR addresses and interrupt vectors can be determined.
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 KA655—-AA CPU.

Format:

CONFIGURE

Example:

>>> configufe

Enter device configuration,
Device, Number? help

HELP,

or EXIT

Devices:
LPV1l

KXJ11

DLV11J

DZQ11

DzZv1l

RLV12

TSVOS

RXV21

DRV11W

DRV11B

DPV11

DMV11

DELQA

DEQNA

DESQA

RQDX3

RRD50

RV20

RQC25
KFQSA-TAPE

KFQSA-DISK
KMV1l

TQKS50
IEQl1

TQK70
DHQ11

KDAS0
TUSBLE

CXAl6é

CXBlé6

CXY08

VCBO1

QVSS

DFAD1

DHV11
LNV11

LNV21

QPSS

DSV1l1

ADV11C

AAV1IC

AXV11C

KWvl1lC

ADV11D

AAV11D

vCBO2

QDSS

DRQ3B

DRV11J

VSsval

IDV11D

1IAV11A

IBQO1
IAV11B

IDV1lA
MIRA

IDV11B
ADQ32

DTCO4

DESNA

IGQl1l

IDV11C

KAG655 Firmware

3-23

g
[
x

)

gfi

amet
®

Numbers:

1 to 255, default is 1

|
’
,

E;W

.o
|

Device, Number?
Y Device, Number? dhvll
Device, Number? qdss

Device, Number? tqgk50
Device, Number? tgk70
Device, Number? exit

Rddress/Vector Assignments
~772150/154 RQDX3

-760334/300 RQDX3
~774500/260 TQKS0
-760444/304

TQK70

-760500/310 DHV11
-777400/320 QDSS
>>>

3-24

KA655 CPU System Maintenance

*’i

3.9.3 CONTINUE
The CONTINUE command causes the pfocéasdr 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:
>>»> continue

KAG655 Firmware

38-25

4*‘2

I w
-

3.9.4 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: /B, /W, /L, /1Q, /N:{count}, /STEP:(size}, WRONG

Address space control: /G, /1, M, /P, /V, /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,

[{data}]

Additional data to be deposited (as many as can fit on the command line).

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.

Examples:
>>> D/P/B/N:1FF

>>> D/V/L/N:3

>>> D/N:8

1234

RO FFFFFFFF

>>> D/L/P/N:10/ST:200

- >>> D/N:200

3-26

Clear

physical memory.

Deposit

!
!

starting at virtual memory address
1234.

Loads

first

512 bytes

into

four

longwords

GPRs RO through R8 with -1.

Deposit

the

memory.

first

in the

first

17 pages

Starting at

513

longword of

in physical

address,

longwords or 2052 bytes.
-

KAB655 CPU System Maintenance

clear

3.9.5 EXAMINE
The EXAMINE command examines the contents of the memory location or
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 examines 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: /B, /W, /L, /Q, /N:{count), /STEP:{size}, /WRONG

Address space control: /G, /1, M, /P, IV, /U
Command specific:

/INSTRUCTION

qgg:umMflmnanddhmhwmfih&“&Xh&mwmfi2fimfinmfihmtfitheapmfl%m
address.

Arguments:

[{address}]

A longword address that specifies the first location to be exanfimd. The
address cmze an actual or a symbolic address. If no address is specified,
+ is assume

Examples:
>>>

Examine

the

PC.

Examine

the

SP.

Examine the PSL.

Examine PSL another way.

G O000OOOF FFFFFFFC
>>>

G 0000000E 00000200
>>>

ex psl

00000000 041F0000
>>> e/m
M 00000000 041F0000
>>> e r4/n:5
00000004 00000000
QOO

! Examine R4 through R9.

00000005 00000000
00000006 00000000
00000007 00000000
00000008 00000000
00000009 801DS000

KAGE55 Firmware

3-27

‘¥§

>>> ex pr$_scbb

00000011 2004A000

r->>> e/p

P 00000000

20040019

P 20040024
P 2004002F

P 20040036
P 20040030
P 20040044
>

e/ins

20040048

3-28

IPR

17
_

(decimal).
!

11 BRB

>>> ex /ins/n:5 20040019
P

| Examine
the SCBB,

Examine

local memory O.

00000000

>>> ex /ins 20040000
P 20040000

! Examine 1lst byte of ROM.
‘
20040019

! Disassemble from branch.

DO MOVL

I~420140000,@#20140000

D2 MCOML
D2 MCOML

@#20140030,@#20140502
S*#0E,@#20140030

7D MovQ
DO MOVL
DB MFPR
DB MFPR

RO,@#201404B2
I~#201404B2,R1
S*#2A,B"44(R1)
! Look at next instruction.
S*#2B,B~48
(R1)

KA655 CPU System Maintenance

i
S e

3.9.6 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 messagé and preserves the contents of SP. If you do not
specify a qualifier, /RPB is assumed.

Format:
FIND [qualifier-list]
Qualifiers:

Command specific:

/MEMORY fleamhan‘ 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.
Searches all of 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.

/RPB

Examples:
>>> ex

Check the

SP.

G 0000000OE 00000000
>>>

find

>>>

/mem

Look

for

Note

where

a valid

Check

None to be

was

128

Kbytes.

found.

G 000000O0OE 00000200
>>>

find

/rpb

?2C FND ERR 00C00004

for valid RPB.

found here.

>>>

KAB55 Firmware

3-29

'3.9.7 HALT

. The HALT command has no effect. It is included for compatibility with
*

other VAX consoles.

Format:
\:HALT

‘E‘:i:ample:
>>> halt
>>>

3-30

;

Pretend to halt.

KA655 CPU System Maintenance

3.9.8 HELP
"The HELP cammmd prcmdes mformatwn about command ayntax and
usage.

Format:

Example:
>>> help

Following is a‘brief summary of all the commands supported by
the console:

UPPERCASE
|
.

denotes a keyword that you must type
denotes an OR condition

[]

denctes optional parameters

denotes

field that must be

filled in

with a syntactically correct value

Valid qualifiers:

/B /W /L /Q /INSTRUCTION
/G /I /V /P /M
/STEP: /N: /NOT
/WRQNG /U

valid cammanda*

DEPOSIT [<QUALIFIERS>] <ADDRESS> [(DATUM> [4DATUM>)]
EXAMINE

MOVE

[<QUALIFIERS>]

[qualifiers)

[<ADDRESS>)

[mask]

SET BFLAG <BOOT_FLAGS>

SET BOOT <BOOT_DEVICE>([:]

SET
SET
SET
SET

HOST/DUP/UQSSP </DISK /TAPE> <CONTROLLER_NUMBER> [<TASK>]
HOST/DUP/UQSSP <PHYSICAL_CSR_ADDRESS> [<TASK>)
HQST/MAINTENANCE/UQSSP/SERVICE <CONTROLLER_NUMBER> [task]
HOST/MAINTENANCE/UQSSP <PHYSICRL”CSR“aDDRESS> [task]

SET LANGUAGE <LANGUAGE NUMBER>
SHOW BFLAG

SHOW BOOT

SHOW DEVICE

SHOW ETHERNET
SHOW LANGUAGE

SHOW MEMORY
SHOW . QBUS

[/FULL)
|

SHOW RLV12
SHOW UQSSP

SHOW VERSION
BALT

INITIALIZE
UNJAM

KA655 Firmware

3-31

TM
W g

CONTINUE
START <ADDRESS>
.

REPEAXT <COMMAND>

X <ADDRESS> <COUNT>
FIND [/MEMORY or /RPB]

TEST

BOOT

[<TEST_CODE> [<PARAMETERS>]]

[/R5:<BOOT_FLAGS> or /<BOOT FLAGS>]
a—

[count)

CONFIGURE .

)

HELP
>>>

3-32

KA655 CPU System Maintenance

[<BOOT_DEVICE>)

o
W 2

'3.9.9 INITIALIZE
The INITIALIZE command performs a processor initialization. Format:

The following registers are initialized:
Register

State at Initialization

PSL

’

IPL

041F0000

ASTLVL
SISR

4
.

ICCS

Bits <6> and <0> clear; the rest are unpredictable

RXCS

TXCS

MAPEN

Caches

Flushed

Instruction buffer

Unaffected

Console previous reference

Longword, physical, address 0

TODR

Unaffected

-Main memory

General registers

Unaffected

Halt code
Bootstrap-in-progress flag

Unaffected
Unaffected

Internal restart-in-progress flag

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
Example:
>>> init
S>>

KA655 Firmware

3-33

3.9.10 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 /N 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 aciually 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 [qualifla‘b-list] {src_address} {dest_address)
Qualifiers:
Data control: /B, /W, /L, /Q, /N:(count}, /STEP:(size}, WRONG
'

Address space control: /V, /U, /P
Arguments:
{src_addresa)

A longword address that specifies the first location of the source data to be
copied.

{dest_address)

A longvéord*addmaa that #pecifiea 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.

Examples:
>>> ex/n:4 0
P 00000000 00000000
P 00000004 00000000

00000008

Observe destination.

Observe source data.

Move the data.

00000000

P 0000000C 00000000
P 00000010 00000000
>>> ex/n:4 200
P 00000200 58DD0520
P 00000204 585E04C1
P 00000208 OOFF8FBB

P 0000020C 5208A8D0
P

00000210

540CASDE

>>> mov/n:4 200
0

3-34

KAB55 CPU System Maintenance

>>> ex/n:4

P 00000000 58DD0520

Observe moved

data.

P 00000004 585E04C1
P 00000008

OOFF8FBB

- P 0000000C 5208A8D0
P 00000010 540CAS8DE
>>>
|

KAGB55 Firmware

8-35

3.9.11 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
I/0 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 pendin
g vector in the SCB.
The following restrictions apply:
*

If memory management is enabled, the NEXT comma
nd works only if
the first page in SSC RAM is mapped in SO (syste
m) space.

Overhead associated with the NEXT command

of an instruction.
*

affects execution time

The NEXT command elevates the IPL to 31 for
long periods of time
(milliseconds) while single stepping over severa
l commands.

Unpredictable results occur if the macro instruction

being stepped over

modifies either the SCBB or the trace trap entry. This
cannot use the NEXT command in conjunction

means that you

with other debuggers.

Arguments:
{count)

A value representing the number of macro instructions

to execute.

Examples:

>>> dep 1000

50D650D4

>>>

dep

125005D1

>>>

dep

1004

1008 OOFE1l1lF9
>>> ex /instruction /n:5

1000

P 00001000
P 00001002

D4 CLRL
D6 INCL

P 00001004
P 00001007

D1 cMPL
12 BNEQ

S~#05,R0

P 00001009
P 0000100B

00 HALT

dep pr$_scbb

>>>

dep pc

3-36

Set

and the

Single.

it.

200

D6 INCL

List

simple program.

00001002

1000

P 00001002

00001009

>>>
>>>

Create

11 BRB=

>>>

KAB55 CPU System Maintenance

user

PC.

SCBB

*”é
n

. CMPL

S~#05,
RO

BNEQ

00001002

INCL

00001004
00001007

CMPL

S*#05,R0
00001002

00001004
D> n

00001007
n

00001002
00001004
00001007

BNEQ

INCL

...or multiple step

the program.

BNEQ

S*#05,R0
00001002

CMPL

n 7

00001002

00001004-

INCL

CMPL

S~#05,R0

00001007

BNEQ

00001002

00001002
00001004
00001007

INCL

RO
S~#05,R0

BNEQ

00001002

00001009

BRB

00001009

BRB

00001009

CMPL

00001009

v
v
v
A
Vv
A
Vv vy
Voot d g v

>>>

KA655 Firmware

3-37

3.9.12 REPEAT

‘

The '/REPEAT command repeatedly displays and executes the specified

command. Press [CTACX] to stop the command. You can specify any valid
console command except the REPEAT command.
Format:

REPEAT (command]

Arguments:
{command} A valid console command other than REPEAT.

Examples:

0000001B SAFE78CE
0000001B 5AFE78D1

0000001B SAFE78FD
0000001B SAFE7900
0000001B SAFE7903
0000001B 5AFE7907
0000001B SAFE790A
0000001B SAFE790D
0000001B SAFE7910

0000001B SAFE793C
0000001B SAFE793F
0000001B SAFE7942

0000001B SAFE7946
0000001B SAFE7949
0000001B SAFE794C
0000001B SAFE794F
0000001B 5~C

VHHFMH
H o
A HHHHH

>>> repeat ex pr$_todr

3-38

KA655 CPU System Maintenance

Watch the

clock.

;

gy
S

3.9.13 SEARCH
The SEARCH command finds all occurrences of a pattem 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 0.

A match occurs if fpattem 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

True

Report address

Absent
Present

False

No report

True

Present

False

“

No report
Report address

The address is advanced by the size of the pattern (byte, word,
longword,
or quadword), unless overriden by the /STEP qualifier.

Qualifiers:
Data control: /B, /W, /L, /Q, /N:{count), /STEP:{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

{pattern)

[{mask]})

to the search. This

address can be an actual address or a symbolic address. If no address
is
specified, + is assumed.
S

- The target data.

A mask of the bits desired in the comparison.

KAB655 Firmware

3-39

o Examples:
>>>

dep /p/1/n:1000 0 O

Clear some memory.

b>>
>>> dep
>>>
>>>

300

Deposit some search data.

12345678

dep 401 12345678
dep 502 87654321

P
>>>

P
P

P
P
>> >

SN
B
RS

P
>>2

search /n:1000 /st:1 0 12345678
00000300 12345678
00000401 12345678
search /n:1000 0 12345678
00000300 12345678
search /n:1000 /not 0 O
00000300 12345678
00000400 34567800
00000404 00000012
00000500 43210000
00000504 00008765
search /n:1000 /st:1 0 1 FFFFFFFE

>>>

>>2>

Search for all occurrences
of 12345678 on any byte
boundary.
Then try on
longword boundaries.
Search for all non-zero
longwords.

Search for odd-numbered
longwords on any boundary.

00000502
00000503
00000504
00000505

search /n:1000 /b 0 12

Search for all occurrences

00000303 12

of the byte 12.

00000404

search /n:1000 /st:1 /w 0 FE1l1l
Y

>>>

87654321
00876543
00008765
00000087

>>>

[

>>>

>>2>

3-40

KAB655 CPU System Maintenance

Search for all words that
could be interpreted as
a spin (10$: brb 10§).
Note that none were found.

.
s

3.9.14 SET

The SET command sets the parameter to the value you specify.
r
v

Format:

SET (parameter} {value)
Parameters:

' BFLAG

Set the default R5 boot flags. The value must be a hex number of up to eight
digits. See Table 3-3 for a list of the boot flags.

BOOT

Set the default boot device. The value must be a valid device name as specified
in the BOOT command description Section 3.9.1

HOST

Connect to the DUP or MAINTENANCE driver on the selected node or device.

The KA655 DUP driver supports only send data immediate messages and
those devices that support the messages. Note the hierarchy of the SET HOST
qualifiers below.

/DUP—Use the DUP driver to examine or modify parameters of a device on the
the Q22-bus.

lfl?hso%fmAttach to the UQSSP device specified using one of the following
me

/DISK n—Specifies the disk controller number, where n is a number
from 0 to 255. The resulting fixed address for n=0 is 20001468 and
the floating rank for n>0 is 26.
.
'
|
|
ITAPE n—Specifies the tape controller number, where n is 2 number
from 0 to 255. The resulting fixed address for n=0 is 20001940 and
the floating rank for n>0 is 30.

gflrmaddrea&-specifiea the Q22-bus I/O page CSR address for the
evice.

/MAINTENANCE—Ezxamines and modifies the KFQSA EEPROM configuration values. Does not accept a task value.

/UQSSP—
/SERVICE n—Specifies service for KFQSA controlier module n where

n is a value from O to 3. (The resulting fixed address of a KFQSA
controller module in maintenance mode is 20001910+4*n.)
/csr_address—Specifies the Q22-bus I/O page CSR address for the
KFQSA controller module.

LANGUAGE Sets console language and keyboard type. If the current console terminal does

not support the Digital Multinational Character Set (MCS), then this command

has no effect and the console message appears in English. Values are 1 through
15. Refer to Example 3-1 for the languages you can select.

Qualifiers: Listed in the parameter descriptions above.

KA655 Firmware

3-41

Examples:
»>> 7
-

>>> set bflag 220
>>>

>>>

set boot

dual

>2>>

>>>

show gbus

.Scan of Qbus I/O Sapce
-200000DC

-200000DE
-200000E0

(760334)=0000
(760336) =0AA0
(760340)=0000

(300)

RQDX3/KDAS50/RRD50/RQC25/KFQSA-DISK

(304)

RQDX3/KDAS0/RRD50/RQC25/KFQSA-DISK

(760344)=0000

(310) RQODX3/KDAS50/RRDS0/RQC25/KFQSA-DISK

-200000E2
-200000E4

(760342)=0AA0

-200000E6
-200000ES8
-200000EA
~-20001468
-2000146A
-20001920
-20001922
-20001924
-20001926
-20001928
-2000192A
-2000192C

(760346)=0AA0
(760350)=0000

(760352) =0AA0
(772150)=0000
(772152) =0AA0
(774440) -FFO08
(774442)=FF00

(314)

RQDX3/KDAS0/RRD50/RQC25/KFQSA-DISK

(154)

RQDX3/KDA50/RRD50/RQC25/KFQSA-DISK

(120)

DELQA/DEQNA/DESQA

(260)

TQK50/TQK70/TUB1E/RV20/KFQSA-TAPE

(774444)=FF2B
(774446)=FF09

(774450) =FFA3

(774452)=FF96
(774454)=0050

-2000192E

(774456)=1030

-20001940
-20001942
-20001F40

(774500)=0000
(774502)=0BCO

(777500)=(004)

IPCR

>>> set host/maint/ugssp 20001468

UQSSP Controller

(772150)

Enter SET,

SHOW,

CLEAR,

HELP,

Node
0

CSR Address
772150

Model
21

1
4
5

760334
760340
760344

21
21
21

EXIT,

m——— KFQSA =—====—

3-42

KA655 CPU System Maintenance

or QUIT

Show current

configuration.
Print text.

Program the KFQSA.

Do not program the

KFQSA.

<NODE>

0 to 7
760010 to 777774
21 (disk) or 22 (tape)

<CSR_ADDRESS>

<MODEL>

set

/KFQSA

‘

Node

CSR Address

Model

o0
O

? show

‘

device.

QUIT
,

772150

Disable a DSSI device.

o
a DSSI

EXIT

Parameters:

Enable

HELP

CLEAR <NODE>
SHOW

<CSR_ADDRESS> <MODEL>

g o,

SET <NODE>

Set KFQSA DSSI node
number.

o
T
SET <NODE> /KFQSA

? help
Commands:

760334

760340

760344

~===== KFQSA ==w—=—-

exit

Programming the KFQSA...

>>>
>>>set

language

o
5

>>>

KAG655 Firmware

3-43

- 3.9.15 SHOW
The SHOW command dmplaya the console pammeter you specxfy
Format

SHOW {parameter}
Pammeters
BFLAG

prlayn the default R5 boot flags.

BOOT

Displays the default boot device.

DEVICE

Displays all devices in the system.

ETHERNET Displays hardware Ethernet address for all Ethernet adapters that can be
found. Displays as blank if no Ethernet adapter is present.

LANGUAGE Displays console language and keyboard type. Refer to the corresponding SET
LANGUAGE command for the meaning.
MEMORY

Displays main memory configuration board by board.
/FULL —Additionally, displays the normally inaccessible areas of memory, such
as the PFN bitmap pages, the console scratch memory pages, the Q22-bus
scatter-gather map pages. Also repom the addresses of bad pages, as dafitwd
by the bitmap.

QBUS

Displays all Q22-bus /O addresses that respond to an ahgned word mad, and

vector and device name information. For each address, the console dwplaya the

addressin 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 readin hex.

This command may take several minutes to complete.

Press [CTRLC

terminate the command. During execution, the command disables the acattergather map.

RLV12

Displays all RLO1 and RLO2 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
&m not indicate whether the device contains a bootableimage.

VERSION

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 run) an MsCP server.

Displays the mmnt firmwm version.

Qualifiers: Listedin the parameter descriptions above.

3-44

KAB55 CPU System Maintenance

*’%

Examples:
- >>

>>> show bflag
00000220

>>> show boot
DUAO
>>>

>>>

show

device

UQSSP Disk Controller 0
-DUAO (RF71)

(772150)

UQSSP Disk Controller 1

(760334)

-DUB1

(RF71)

UQSSP

Disk Controller 2

-DUC4

(RF71)

UQSSP Tape Controller 0
-MUAO

(760340)

(774440)

(TK70)

Ethernet Adapter

-XQA0

(08-00-2B-0B-82-29)

>>>

show

ethernet

Ethernet Adapter

-XQA0

(08-00-2B-0B-82-29)

>>>

show

English

language

(United States/Canada)

>>>

show memory

Memory 0:
Total

00000000 to OOFFFFFF,
16MB,

0 bad pages,

104

16MB,

0 bad pages

reserved pages

>>>

show memory/full

Memory

Total

00000000

of 16MB,

OOFFFFFF,

0 bad pages,

Memory Bitmap
-00FF3C00 to OOFF3FFF,

104

16MB,

0 bad pages

reserved pages

8 pages

Console Scratch Area

-00FF4000 to OOFF7FFF,

Q-bus Map

-00FFB8000 to OOFFFFFF,
Scan

32 pages

64 pages

of Bad Pages

>>>

KA655 Firmware

3-45

>>>

show Q-bus

Scan ofQ-bus I/O Space

-200000DC

(300) RQDX3/KDA50/RRD50/RQC25/KFQSA-DISK

(760334)

= 0000

-200000DE

(760336)

= OAAO

~200000E0

(760340)

(760342)

0000
OARO

(304)

RQDX3/KDAS0/RRD50/RQC25 /KFQSA-DISK

~200000E2
-20001468
~2000146A
£20001920
-20001922
-20001924
-20001926
-20001928

(772150)
.

0000

(772152)

(154)

RODX3/KDAS50/RRD50/RQC25/KFQSA-DISK

OAAO

(774440)

FFO8

FF00

{120)

DELQA/DEQNA/DESQA

(774442)

(774446)

FF2B
FF09

(260)

TQKS0/TQK70/TU81E/RV20/KFQSA~-TAPE

(004)

IPCR

-2000192A
-2000192C

~2000192E
~20001940

-20001942
-20001F40

(774444)

(774450)

FF00

(774452)
(774454)
(774456)(774500)
(774502)
(777500)

FFEl

Scan of Qbus

8400
1030
0000
0BCO
0020

Memory Space

>>>

show RLV12

>>>

show UQSSP

- UQSSP Disk C@ntfmllér 0
-DUAO

(RF71)

UQSSP Disk Controller 1
-DUBl1

(760340)

(RF71)

UQSSP Tape Controller 0

-MUAS

(760334)

(RF71)

UQSSP Disk Controller 2
-DUC4

(772150)

(774500)

(TK70)

>>>show version

KA655-A V5.3,

VMB 2.7

>>>

3-46

KA655 CPU System Maintenance

é%

3.9.16 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.”
. Format:

START [{address}]

Arguments:
[address)

|
glée address at which to begin execution. This address is loaded into the user’s

Examples:
>>>

start

1000

KA655 Firmware

3-47

3.9.17 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 4 for a detailed explanation of the diagnostics.

‘Format:
TEST [{test_number} [{test_arguments]]]

Arguments:
{test_number)
{test_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:
>>> test

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..

3-48

KA655 CPU System Maintenance

3.9.18 UNJAM

The UNJAM command performs an 1/0 bus reset, by writing a 1 (one) to
IPR 55 (decimal).

quat:

UNJAM
Examples:
>>> unjam
>>>

KAB55 Firmware

3-49

SR
w5

Loy
R

3.9.19 X—Binary Load and Unload

" The X command is for use by automatic systems communicating with the

7 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.

fi'ormat:
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, including the checksum and separating space (but not including
the terminating carriage return, rubouts, or characters deleted by rubout),
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
content of the register is non-zero, 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.
|
D

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 &re unpredictable.

3-50

KA655 CPU System Maintenance

The console represses echo while it is receiving the data string and
chacksums.
|

Theé 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
charactem from the terminal as valid command line checksums.

You can control the console serial line during a binary unload using control
characters ([TTRIX], [CTALS), [€TAUS), and so on). You cannot control the console
serial line during a binary load, since all received characters are valid
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.

{)t 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.

KAG655 Firmware

3-51

13.9.20 | (Comment)

/ The comment character (an exclamatwn pomt) is used to document
" command sequences. It can appear anywhere on the command line. All
characters following the comment character are ignored.
Format

Examples
>>>

The console n.gnoras this line.

>>>

3~-52

KAB655 CPU System Maintenance

Chapter 4
Troubleshooting and Diagnostics

| 4.1 Introduction
This chapter contains a description of KA655 ROM-based diagnostics,
acceptance test procedures, and power-up self-tests for common options.

4.2 General Procedures
Before troubleshooting any system problem, check the site maintenance
guide for the system’s service history.
Ask the system manager two
questions:

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

Have changes been made to the system recently?

Three common problems occur when you make a change to the system:
*

Incorrect cabling

Module configuration errors (incorrect CSR addresses and interrupt
vectors)

Incorrect grant continuity

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. Microsystems Options lists
address and vector values for most options.
If you change the system configuration, run the CONFIGURE utility
at the console I/O prompt (>>>) to determine the CSR addresses and
interrupt 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 MicroVAX
Diagnostic Monitor User’s Guide for information about the CONNECT and
IGNORE commands, which are used to set up MDM for testing nonstandard
configurations.

Troubleshooting and Diagnostics

4-1

When troubleshooting, note the status of cables and connectors before you

perform each step.

Label cables before you disconnect them.

This step

B - saves You time and prevents you from introducing new problems.

If the operating system fails to boot (or appears to fail), check the console

terminal screen for an error message. If the terminal displays an error
message, see Section 4.3. - Check the LEDs on the device you suspect is
. bad. If no errors are indicated by the device LEDs, run the ROM-bas
ed
\
diagnostics described in this chapter.

In addition, check the following connections:

If no message ap;pem, make sure the console terminal and the system

are powered on. Check the on/off 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.

If you cannot get a display of any kind on the console terminal
, try
another terminal.

If the system DC OK LED remains off, check the power supply and
power supply cabling.
|

Check the hex display on the Ini.?:G()O--SAu~ If the display is off, check

the CPU module LEDs and the CPU cabling. If a hex error message
appears on the H3600-SA or the module, see Section 4.3.

If the system boots successfully, but a device seems to fail or an intermittent
failure occurs, check the error log first for a device problem. The failing
device is usually in one of the following areas:

CPU
Memory
Mass storage
- Communications devices

4.3 KA655 ROM-Based Diagnostics
The KA655 ROM-based diagnostic facility, rather than the MicroVAX Di-

~agnostic Monitor (MDM), 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.

4-2

KAB55 CPU System Maintenance

Diagnosis is done in a more primitive state. (MDM requires successful
- loading of the VAXELN operating system.)
|
-

The ROM-based diagnostics can indicate several different FRUs, not just

. the CPU module. For example, they can isolate one of up to four memory

modules as FRUs.

(Table 4-6 lists the FRUs indicated by ROM-based

diagnostic error messages.)

The diagnostics run autbmatica]ly on power-up. While the diagnostics are

running, the LEDs on the H3600-SA display a hexadecimal countdown of

‘the tests from F to 8 (though not in precise reverse order) before booting
the operating system, and 2 to 0 while booting the operating system. A
different countdown appears on the console terminal.

The ROM-based diagnostics are a collection of individual tests with
parameters that you can specify. A data structure called a script points
to the tests (see Section 4.3.2). There are several field and manufacturing
scripts. Qualified Field Service personnel can also create their own scripts

interactively.

A program called the diagnostic executive determines which of the available
scripts to invoke.
The script sequence varies if the KA655 is in a
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
Table 4-1 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 change. The addresses in Table 4-1 are for V5.3, for
other versions, use as examples only.
*

Test is the test code or utility code.

Address is an example of the test or utility’s base address 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 (available
on microfiche).

Name is a brief description of the test or utility.

Troubleshooting and Diagnostics

4-3

Parameters shows the parameters for each diagnostic

Tests accept up to ten parameters.

test or utility.

The asterisks (*) represent

parameters that are used by the tests but that you
cannot specify

individually. These parameters are encoded in ROM and
are provided
by the diagnostic executive.

Table 4-1:

Test and Utility Numbers

Test

Address!
2004BC00

De_SCB

Cé

2004D2F0

SSC_powerup

2004D3B2

CBTCR timeout

Name

- Parameters

2004D46C ©

ROM logic test

2004D534

CMCTL_powerup

2004D57C

2004D6A0

CQBIC_powerup

2004D730

CQBIC regs

CMCTL regs

MEMCSRO_addr ****sxsxs
e
*

2004D7B2

CQBIC-memory

B ARk

2004DDD9

Console serial

start_baud end_baud ******

2004E128

Console QDSS

2004E2A8

QDSS self-test

.input_csr selftest_r0 selftest_r1 ***wx
which_timer wait_ time_us ***

2004E40E

CFPA

2004ESFA

Prog timer

2004E8CO

TOY clock

2004EA6B

Interval timer

2004EAE0

VAX CMCTL CDAL

2004EBF0

cache_mem_cqbic

2004EEDS

Cachel_diag_md

mark_not_present selftest_r0 selftest_r1 **+**

repeat_count_250ms_ea ****

*
dont_report_memory_bad repeat_count *

start_addr end_addr addr_incr ****

addr_incr ******exs

2004F502

List diags

2004F528

MSCP-QBUS test

IP_csr ***xkx»

device_num_addr ****

2004F6EA

DELQA

2004F8C5

SSC RAM

2004FA8C

SSC RAM ALL

*
*

2004FBFS8

SSC regs

2004FCE9

Virtual mode

start_addr end_addr addr_incr **#x%uk
start_addr end_addr addr_incr ***+***

2004FF68

Cache2_memory

200504DC

Cache?2 integrty

20050CF4

Cache_memory

20050D4D

Cachel_cache2

addr inc-'l' e s o e 2 o o ok ok

addr iner ***ssesex

2005107C

20051269

Check_for_intrs

200512AC

MEM_setup_CSRs

200 2 o ek ook ol

. Board reset

1V5.3 addresses; use as examples only.

4-4

KA655 CPU System Maintenance

‘

Ew ]
=

Table 4-1 (Cont.):

Test. Address'!

Test and Utility Numbers

Name

Parameters

- 30

200518CF

MEM_bitmap

*** mark_Hard_SBEs ******

20051D0A

4E -~

20051ECS

MEM_data

MEM_byte

start_add end_add add_incr cont_on_err ##****

20051FDD

MEM_address

2005216F

20052630

MEM_ECC_error

2005280A

MEM_correction

start_add end_add add_incr cont_on_err ******
start_add end_add add_incr cont_on_err ***+*»

start_add end_add add_incr cont_on_err **++++

MEM_maskd_errs

20052A1D

MEM_FDM_logic

20052FEC

- MEM_addr_shrts

20053643

200537CE

start_add end_add add_incr cont_on_err ****+*

start_add end_add add_incr cont_on_err ******
*** cont_on_err ******

start_ add end_add * cont_on_err patl pat2
Rk

MEM_refresh
"

start end incr cont_on_err time_seconds *****

MEM_count_errs

First board Last _board
allowed

20053B14

Utilities

20053C18

List CPU regs

200541D4

Create script

Mad

******* Soft errs_

Expnd_err_msg get_mode init LEDs clr_ps_
cent

1V5.3 addresses; use as examples only.

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

2004FCE9

Virtual mode

% % Kk %k %

200518CF

MEM bitmap

*** mark_hard_SBEs **xx*xx

The virtual mode test on the first line contains several parameters, but you
cannot specify any that appear in the table as asterisks. To run this test
individually, enter:
>>>

The MEM__bitmap test on the second line accepts ten parameters, but you

can specify only mark_hard_SBEs because the rest are asterisks. To map
out solid, single-bit ECC memory errors, type:
|

>>>

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 hold a place 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.
AN

Troubleshooting and Diagnostics

4-5

You enter 1 for parameter 4 to indicate that the test should map out solid,
single-bit as well as multi-bit ECC memory errors. You then terminate the

- command line by pressing [RETURN ‘You do not need to specify parameters 5
through 10; placeholders are needed only for parameters that precede the
user-definable parameter.
|

. 4.3.2 Scripts

' Most of the tests shown by utility OE are arranged into scripts. A script

is a data structure that points to various tests and defines the order in
which they are run. Different scripts can run the same set of tests, but in
a different order and/or 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 EPROM. Other tests are program independent code, and
can be run from EPROM, cache diagnostic space, 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, confinue).~

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 0.

Additional scripts are included in the ROMs for use in manufacturing and
engineering environments. Field Service 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 you should use the UNJAM and INITIALIZE commands,
described in Chapter 3, before running an individual test. You do not
need to use these commands on system power-up, however, because system
power-up leaves the machine in a defined state.
Field Service personnel with a detailed knowledge of the KA655 hardware
and firmware can also create their own scripts by using the 9F utility. (See
Section 4.34.)
=
|
~
|

4-6

KA655 CPU System Maintenance

Table 4-2 lists the scripts:

Table 4-?:‘ Scripts Available to Field "Service
Script!

Command

Description

Soft script created by de_test9f. Also referred to as user script.

Al, AA,

Common section of power-up script.

Enter T 9F to create.

AB, AC,
0,3

A7, A8

Script AC invokes this

script at power-up. This script does not directly invoke any

;

tests, but calls script BD to run the tests.
Memory test portion invoked by script A8. Reruns the memory
tests without rebuilding and reinitializing the bitmap. Run
script A8 once before running script A7 separately to allow
mapping out of both single-bit and double-bit main memory

ECC errors.

AA 0

Memory acceptance. Running script A8 with script A7 tests

main memory more extensively. It enables hard single-bit and
multi-bit 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.

Console SLU. Invokes scripts BA, BC, and Al. Does not invoke

any tests directly.

AC, 3

AE, AD

Power-up. Invokes scripts BC and Al.
tests directly. Invoked at power-up.

Does not invoke any
‘

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 INIT command.

BA, 2, AA

BC,AA,AC,0,3

Called by scripts AA and AC. Provides console announcements.
Invoked at power-up.

BD, Al,AA,AB,
AC, 0,3

Common section of power-up script.
script at power-up.

Console program. Resets busmap and resets caches.

Initial power-up script for console SLU before first console
announcement. Invoked at power-up.

Script Al invokes this

1Scripts A2~A6, BO-B3, and B5 are for manufacturing use. They should not be used by
Field Service, Scripts AB and BB are used to test the QDSS, which is not supported.
Scripts BE, BF, B4, and B6-B9 are not used.

Troubleshooting and Diagnostics

4-7

In most cases, Field Service needs only the écripts shown in Table 4-3 for

. effective troubleshooting and acceptance testing. -

Table 4-3:

Commonly Used Field Service Scripts

Command

Description

Automatically invokes the proper scripts; runs the same tests as during power-

Primarily runs the memory tests; halts upon first hard or soft error.

up.

Memory gcceptance script; marks hard multi-bit and single-bit ECC errors in
the bitmap. Script A8 calls script A7 when this command is entered. Script
AT contains the memory tests that will continue on error.

Can be run by itself; useful when you want to bypass the bitmap test.
Power-up script that can be run by itself. Bypasses the bitmap test.

4.3.3 Script Calling Sequence
Actions at Power-up

In a nonmanufacturing environment where the intended console device is
the serial line unit (SLU), the console program (referred to as CP below)
performs the following actions at power-up:
|
|
1.

Runs the IPT.

Assumes console device is SLU.

Calls the diagnostic executive (DE) with Test Code = 2.
a.

DE determines environment is nonmanufacturing from H3600-SA.
(Manufacturing sets a jumper on the H3600-SA for testing.)

DE selects script sequence for console SLU.
DE executes Script BA.
—

Script BA directs DE to invoke script BD. Script BD directs DE
to execute tests (console announcements are off).

DE passes control back to the CP,

Establishes SLU as console device (whether or not SLU test passed).
Prints banner message.

Displays language inquiry menu on console if console supports MCS
and any of the following are true:
|
AN

4-8

KA655 CPU System Maintenance

Battery is dead.

)

« H8600-SA switeh is set to language inquiry,

7«
7.

Calls DE with Test Code = 3.

?f' /

Contents of SSC NVRAM are invalid.

R ,

DE executes Script AC.

Script AC directs DE to execute vscfipts BC and Al.

—

Script BC directs DE to execute tests (console announce-

ments are on).

—

Script Al directs DE to invoke script BD. Script BD directs

DE to execute tests (console announcements are on).

b.
8.

DE passes control back to CP.

Issues end message and >>> prompt.

Actions After You Enter T 0
In a nonmanufacturing environment where the intended console
device is
the SLU, the console program (CP) performs the following actions
after you
enter T O at the console prompt (>>> T 0):

Calls the diagnostic executive (DE) with Test Code = 0.
a.

DE determines environment is nonmanufacturing from H3600-SA
switch setting.

DE executes script AA.

Script AA directs DE to execute scripts BA, BC, and Al.
—

—

Script BA directs DE to execute tests (console announcements are off).

Script BC directs DE to execute tests (console announce-

ments are on).

—
c.

Script Al directs DE to invoke script BD, which then directs

DE to execute tests (console announcements are on).

DE passes control back to the CP.

Issues end message and >>> prompt.

Note that although the sequence of actions is different in the two
cases
above, the same tests (those in scripts BA, BC, and BD) run both times.
AN

Troubleshooting and Diagnostics

4-9

- 4.3.4 Creating Scripts

| ~ You can create your own script using utility 9F, to control the order in which

tests are run and to select specific parameters and flags for individual tests.
In this way you do not have to use the defaults provided by the hard-wired
scripts.
|

Utility 9F also provides an easy way to see what flags and parameters are
used by the diagnostic executive for each test.
Run test 9F first to build the user script (see Example 4-1). Press <CR> to
use the default parameters or flags, which are shown in parentheses. 9F
prompts you for the following information:
®

Script location. The script can be located in the 1-Kbyte NVRAM in
the SSC, diagnostic RAM (DIAG_RAM), or in main memory. A script
is limited by the size of the data structure that contains it. A larger
script can be developed in main memory than in DIAG_RAM, and a
larger script can be built in DIAG_RAM than in NVRAM.
A script cannot, however, always be located in main memory. For
example, a script that runs memory tests will overwrite the user script,
since the diagnostic executive cannot relocate the user script to another

area. The diagnostic executive notifies you if you have violated this type
of restriction by issuing a script incompatibility message.

Test number

Run environment. This defines where the actual diagnostic test can be
run from. The choices are 0 = ROM, 1 = DIAG_RAM, 2 = Main Memory,
and 3 = Fastest Possible. Choose number 3 to select the fastest possibie
data structure to run from that will not overwrite the test.

Repeat code

Error severity level
Console error report

Script error treatment

LED display

Console display

Parameters

4-10

KA655 CPU System Maintenance

X =
g

Example 4—1 shows how to build and run a user script.

The utility displays the test name after you enter the test mimber, and the

number of bytes remaining after you enter the information for each test.
. When you have finished entering tests, press <CR> at the next Next test

number: prompt to end the script-building session. Then type T AO<CR>

to run the new script.

You cannot review or edit a script you have created.

Example 4-1: Creating a Script with Utility 9F
,

>>>T

SP=20140604
Create

script in ?2[0=SSC, 1=DIAG_RAM, 2=RAM]
starts at 2011FCO00
1024 bytes left

Script
Next
CFPA

CFPA

CFPA
CFPA

test

number

:51

>>Run

from ?[0=ROM,1=DIAG
2=RAM, _RAM,
3=fastest possible]
>>Repeat? [0=no, 1l=on error, 2=forever, >2=count<fFr]
(0):

(0):

>>Error severity ? [0,1,2,3]
(2):
>>Console error report? [0O=none,l=full] (1):
>>Stop script on error? [0=NO,1=YES]
(1):
>>LED on entry (05):
>>Conscle on entry (51):

1017 bytes

left

Next test number :52
Prog timer >>Run

from ? [0=ROM, 1=DIA2=RAM,
G_RAM
3=fastest
, possible]
Prog timer >>Repeat? (0=no,l=on error,2=forever,>2=count<FF] (0):
Prog timer >>Error severity ? [0,1,2,3] (2):

Prog timer >>Console error report? [O=none,l=full] (1):
Prog timer >>Stop script on error? [0=NO,1=YES] (1):
Prog timer >>LED on entry (05):
Prog timer >>Console on entry (52):
Prog timer >> which_timer : 00000000 - 00000001 ?(00000000)
Prog timer >> wait_time_us

1002 bytes
Next

test

00000001

left
number

- FFFFFFFF

? (0000000A)

>>>T AQ

51..52..
5>>

Troubleshooting and Diagnostics

4-11

(0):

~ Example 4-2 shows the facripbbuilding procedure to wfevlloiv'«'r if (a) you are

unsure of the test number to specify, and (b) you want to run one test
repeatedly.
If you are not sure of the test number, enter ? at the Next test number:
prompt to invoke test 9E and display test numbers, test names, and so on.
To run one test repeatedly, enter the following sequence:
1.

Enter the test number (40 in Example 4-2) at the Next test number:
prompt.

Enter AQ at the next Next test number: prompt.
Press <CR> at the next Next test number: prompt.
Enter T A0 to begin running the script repeatedly.

Press [CTAUT] to stop the test.

This sequence is a useful alternative to using the REPEAT console
command to run a test, because REPEAT (test) displays line feeds only;
it does not display the console test announcement.

4-12

KAB655 CPU System Maintenance

L I
.

Example 4-2:
3

Listing and Repeating Tests with Utility 9F

T 9F

SP=2 0140604
Create script

in
Script starts at
Script starts at
24 bytes left
24 bytes left
Next test number
- Test

?[0=SSC,1=DIAG_RAM, 2=RAM]
20140758
20140758

Address

Name

Displays available tests and utility

2004BCO0

De_sSGB

Cé
C7

2004D2F0
2004D3B2

SSC_powerup
CBTCR timeout

2004D46C

33
32
91
90
80
60

2004D534
2004DS7C
2004D6A0
2004D730
2004D7B2
2004DDDY9
2004E128
2004E2A8
2004E40E
2004ESFA
2004E8CO
2004EA6B
2004EAE0
2004EBF0
2004EED8

ROM logic test

€62
63
51
52
53
55
SR
45
46

2004F502

2004F528

2004F6EA
2004F8C5
2004FA8C

C2
C5

2004FBF8

2004FCE9

2004FF68

35
44

200504DC
20050CF4

20050D4D

2005107C

20051269

200512AC

200518CF

Parameters

TRAXEARAR
wkw

CMCTL powerup
*
CMCTL regs
MEMCSRO_addr **»awxwxx
CQBIC_powerup
*»
CQBIC regs
*
CQBIC_memory
REAXNRL
A K
Console serial
start_baud endbaud ***xx»
console QDSS
marknotmpxuaunt selftest_r0 selftest_xl *x*»
(QDSS self-test
input_csr selftest_r0 selftest_rl *RAH kR
CFPA
RAXRR
Prog timer
which_timer wait_time_us *#x
TOY clock
r&paatmpuatm250mqmaa Tolerance %*%
Interval timer
VAX CMCTL CDAL
dont_report_memory_bad repeat_count *
cache_mem_cgbic start_addr end_addr addr_incr#w*»
Cachel_diagmd
addr_incr mxxxxamxs
1Llist diaqs
*
MSCP-QBUS test
IP_csr **xxxx
DELQA
device_num addr ****
SSC RAM

SSC RAM ALL
SSC regs
Virtual mode
cache2_memory

Cach2_integrty
Cache_memory

*
*
THHAR
W

start_addr end_addr addr_incr *wawxsxx
start uddr end|_addr addr _incr whxsaww
addr_incr s*awawwxw

Cachel_Cache2
addr
_incr *xmmsxawx
Board Reset
kAR
Check_for_intrs w»*
MEM_Setup CSRs
*awawwawwx

MEM Bitmap

*** mark_Hard SBEs *%#xwx

Example 4-2 Cont'd. on next page

Troubleshooting and Diagnostics

4-13

20051FDD

MEM Address

2005216F

MEM ECC_Error

20052630

2005280A

MEM Maskd Errs
MEM Correction

start_add end_add add_incr cont_on_erxr
start_add end_add add_iner cont_on_err
start_add end_add add_incr cont_on_err
start_add end add add iner cont_on_err
start_add end add add incr cont_on_err
start_add end add add_incr cont_on_err
#»*w cont_on_err **wwxx

20052A1D MEM FDM _Logic

200352FEC

47
40

20053643
200537CE

MEM_Addr_shrts

20053B14

Utilities

20053C18

List CPU regs
Create script

9F
200541D4
24 bytes left

start_add end_add *

MEM Refresh
MEM_Count_Errs

**xkww
*wxwax
**wxka

*wrxxw
*wwxxx

cont_on_err pat2 pat3

%wxw

start end incr cont_on_err time_seconds *xw#=
First_board Last_board Soft_errs_allowed **xxwwx

Expn_msg
d_er
get_mode
r init_LEDs

clr_ps_cnt

TRR Ak

Next test number: 40
MEM_Count_Errs>>Run’ from 2{0=ROM, 1=DIAG_RA
M, 3=fastest possible]
MEM_Count_Errs>>Repeat ?[0=no, 1=on error,
2=forever, >2acount<FF]
MEMCount_Errs>>Error severity ? (0,1,2,3]
(2):
MEMCount_Errs>>Console error report? [O=none,l
=full)
(1):
MEMCount_Errs>>Stop script on error? [O=no,
1=yes]) (1):
MEMCount_Errs>>LED on entry (04):
MEMCount_Errs>>Console on entry (40):

MEMCount_Errs>> First board

5 bytes left
Next test number
4 bytes left

:AQ0 - script

Next test

(0):
(0):
-

00000001 - 00000004 21

MEMCount_Errs>> Last board : 00000001 - 00000004
2 (00000004)
MEMCount_Errs>> Soft_errs_allowed : 00000000
- FFFFFFEF 22

number

*xxxkx

MEM Data
MEM Byte

20051EC8

20051DOA

“

Example 4-2 (Cont.): Llstlng and “Rapeatmmg Tests with Utmty 9F )

>>>T AO

40..40..40..40..40..40..40..40..40..40..40..40..
40..40..40..40..
40..40..40..40..40..40..40..40..40..40..
40..40:.40..40..40..40..
40..40..40..40..40..40..40..40..40..40..40..40..
40..40..40..40..
~C
>>>

4.3.5 Console Displays
Example 4-3 shows a typical console display during execution of the ROM-

based diagnostics. The numbers on the console display do not refer to actual

test numbers. Refer to Table 4-6 to see the correspondence between the
numbers displayed (listed in the Normal Console Display column) and the
actual tests being run (listed in the Error Console Display column).

4-14

KA655 CPU System Maintenance

Example 4-3:

Console Display (No Errors)

KA655-A V5.3 VMB 2.7

Performing Normal System Tests
40..39..38..37..36..35..34..33..32..31..30..29..28..27..26..25..

24.;23..22..21..20..19.118..17~‘16..15..14..13..12.‘11..10.*09..
08..07..06..05..04..03..

Tests

completed’

>>>

The first line contains the firmware revision (V5.3 in this example) and the

virtual memory bootstrap (VMB) revision (V2.7 in this example).

During execution of the IPT, normal error messages are displayed if the
console terminal is working. Console announcements such as test numbers
and countdown, however, are suppressed. Tests continue to run after the
IPT, up to and including the appropriate console test.

Diagnostic test failures, if specified in the firmware script, produce an error
display in the format shown in Example 4—4.
Example 4-4:

Sample Output with Errors

7?46 2 07 FE 10 0002
|
P1=002F0000 P2=00000000 P3=00000000 P4=00FF0000
P6=00000000 P7=00000000 P8=00000000 P9=00FF0000
r0=00000000 r1=00010000 r2=55555555 r3=00000080
r5=00000080 ré6=01EF0000 r7=20080144 r8=00010000
Tests

P5=00000000
P10=00000000

r4=AAAAAARAA
ERF=20140770

completed

Troubleshooting and Diagnostics

4-15

«,i{’i

The errors are printed in a five-line display. The first line has six fields:

Test m‘*";Sevérity
f

Error

De_error

Vector

Count

Test identifies the diagnostic test.

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 five-line 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 héx digits identifying, usually within 10 instructions,

where in the diagnostic the error occurred.

the subtestlog.

This field is also called

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:

FF—Normal error exit from diagnostic
FE—Unanticipated interrupt
|
FD—Interrupt in cleanup routine
FC—Interrupt in interrupt handler
FB—Script requirements not met
FA—No such diagnostic
EF—Unanticipated exception in executive

Vector identifies the SCB vector (10 in the example above) 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 (two in Example 4-4).

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 that are listed in Table 4-1.

4-16

KA655 CPU System Maintenance

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 Tables 44 and 4-5.

- Table 44:
Parameter
. P1

P2
P3

P4
P5

Values Saved,
Executive
Value

Machine

Exception

Contents of SP, points to vector value in P2
Vector = 04, vector of exception 04~FC, 00 = Q-bus
Address of PC pointing to failed instruction, P9

Byte count = 10 Machine check code

Most recent virtual address

Internal state information 1
Internal state information 2

PC, points to failing instruction

P10

PSL

Table 4-5:

Values Saved, Exception During Executive

Parameter

Value

Pl
P2

P3
P4

During

’

Pé6
P8

Check

- Contents of SP, points to vector value in P2
- Vector = nn, vector of exception 04-FC, 00 = Q-bus

Address of PC pointing to failed instruction, P4
PC, points to instruction following failed instruction

PSL

Pé

Contents of stack

P10

Contents of stack

Lines 4 and 5 of the error printout are general registers R0 through R8 and
the hardware error summary register.

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.
|

Troubleshooting and Diagnostics

4-17

- Table 4-6 lists the hex LED display, the default action on errors, and the |
most likely FRUs. It is divided into IPTs and scripts.

The Default on Error column refers to the action taken by the diagnostic

executive under the following circumstances:

*
L ®

The diagnostic executive detects an unexpected exception or interrupt.
Atest fails and that failure is reported to the diagnosic executive.

'I‘i:xe 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 (labeled cont_on_
error, as shown in test 47 in Example 4-2). 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 diagnosti
c
executive that an error has occurred.

Figure 4-1 shows the LEDs on the KA655 CPU. They correspond to the
hex display on the H3600-SA.
-

Figure 4-1: KA655 CPU Module LEDs
9

* »

LA N

itouoioiflfit

fifln&.fi'n'fiiflfliflbl

fl&lflfltu’

Qflafi

Ofi0'1000‘01“0!#0‘00.10&0

8421

DC OK

DIAGNOSTIC

LED

LEDs

4-18

KA655 CPU System Maintenance

MLO-002375

Table 4-6:

KAGSS Console Displays and FRUs

.
Normal Error
Hex Console Console
LED, Display Display

Dafau.lt on
Error
Description

FRWU?

61,4, 53

" Initial Powar«-Up Tests

None

_Looponself

None

Power-up

Loop on self

WAIT POK

None

Looponself

None

Entering IPT

None

Looponself

SLU_EXT_LOOPBACKY

7,1

Script AA

‘

Invoke script BA.
Invoke script BC.

Invoke script Al.

End of script.

Script BA
C

None

79D

Continue

None

Utilities

742

Continue -

Check_for_intrs

C
6

None
-

None

End of script

2C6

Continue

1?60

Continue

’

SSC_powerup

CONSOLE_SERIAL

*‘

Script AC
Invoke script BC.

Invoke script Al.

End of script.

IFRU key:
1 = KA655
2 = MS650-BA
3 = Memory interconnect cable
4 = Q22-bus device

5 = Q22/CD backplane
6 = System power supply
7 = H3600-SA I/O panel

21:1 the case of multxple FRUs, refer to Section 4.5.2 for further information,

3If a problem reccurs with the same FRU, check that the tolerance for system power supply
+5 Vdc, +12 Vdc, and AC ripple are within specification.

4This test only runs if the power-up mode switch on the H3600~SAis set to TEST
Section 4.6.1.

mode. See

Troubleshooting and Diagnostics

4-19

‘Table 4-6 (Cont.): KA655 Console Displays and FRUs
Hex

LED

"'Normal Error

Console Console

Defaulton

Display Display

Error

Description

FRU!

Script BC
7

791

Continue

290

CQBIC_power-up

Continue

CQBIC_registers

733

Continue

232

CMCTL _power-up

Continue

CMCTL _registers

731

Continue

749

CMCTL_setup_CSRs

1,2,3,5

Continue

7230

MEMORY_FDM_logic

Halt

MEMORY_bitmap

2,1,3,6

End of script.

1,2,3,5

Script Al
Invoke script BD.

Script BD

752

Continue

752

PROG_TIMER_0

Continue

753

PROG_TIMER_1

Continue

?C1

TOY_CLOCK

Continue

SSC_RAM

734

Continue

?2C5

ROM_logic

Continue

755

SSC_registers

Continue

751

Continue

INTERVAL_TIMER

2C7

Continue

1
1

CFPA

CBTCR_timeout

246

Continue

735

CACHE_DIAG_MODE

Continue

Cache2_integrity

SSC_RAM_ALL
Cachel_cache2

?2C2

Continue

243

Continue

IFRU key:
-

1
1

1 = KA655
2=MS650-BA
|
3 = Memory interconnect cable
4 = Q22-bus device
5= Q22/CD backplane
6 = System power supply
7 = H3600-SA /O panel

4-20 KA655 CPU System Maintenance

Table 4-6 (Cont.):
Hex

LED

KA655 Console msplays and FRUs

Normal Error
Console Console

Display Display

Defaulton

Error

Script BD

Dmription

FRU!

Continue

MEMORY_data

2,185

Continue

18 -

74D

MEMORY_byte

Continue

2,185

74C

Continue

74B

Continue

24A

Continue

MEMORY_addr
= MEMORY_ECC_error

MEMORY_masked_errors

MEMORY_correction

748

Continue

247

Continue

MEMORY _refresh

MEMORY_count_errors

740

Continue

244

Continue

736

Continue

MEMORY_address_shorts

CACHE1_MEMORY
Cache2_memory

780

Continue

754

CQBIC_MEMORY

Continue

VIRTUAL_MODE

?2C5

Continue

234

Continue

7245

75A

241

Continue

End of script.

SSC_registers

2,1,8,6
2,185

2,1, 3, 5

2,1,35
2,1, 8, 5
2,1,38,5

2,135
1,2,3,5
1,2,3,5

1,2,4,8,5

1,2,8,5

ROM logic test

Continue

CACHE_MEM_CQBIC

Continue

CVAX CMCTL CDAL

1,2,4,3,5

‘

Board reset

1,4

Script A8
8

731

Halt

749

Halt

230

Halt

Invoke script A7.

CMCTL_setup_CSRs
MEMORY_FDM_logic

2,185

MEMORY_bitmap

1,2,8,5

End of script.

1FRU key:
1 = KA655
2 = MS650-BA
3 = Memory interconnect cable
4 = Q22-bus device
6 = Q22/CD backplane

6 = System power supply
7 = H3600-SA 1/0 panel

Troubleshooting and Diagnostics

4-21

Table 4-6 (Cont.):
Hex
LED

KA655 Console Displays and FRUs

Normal Error
Console Console
Display Display

Defaulton
r
Description

Script A7

FRU?

24F

Halt

24D

MEMORY _byte

Halt

MEMORY _addr

Halt

MEMORY_ECC_error

74B

Halt

24A

Halt

248

Halt

47 -

Halt

240

Cont

780

Cont

741

Halt

End of script.

MEMORY_data

MEMORY_masked_errors

MEMORY_correction

MEMORY_address_shorts
MEMORY _refresh

MEMORY_count_errors
CQBIC_memory

2,1,8,5

2,1,385
2,1,8,5
2,1,38,5
2, 1,3,5

2,185
2, 1, 3, 5

2,1,3,5

2,1,8,5
1,2 4,3,

Board reset

1,4

MEMORY_data

2,135

MEMORY_addr

2,185

Script A9
8

Halt

24E

Halt

724D

Halt

24C

‘Halt

?4B

Halt

24A

Halt

248

Halt

247

Halt

240

Continue

241

Continue

End of script.

MEMORY_byte

MEMORY_ECC_error

MEMORY_masked_errors
MEMORY_correction

MEMORY_address_shorts
MEMORY _refresh

1 = KA655
2 = MS650-BA

3 = Memory interconnect cable

4 = Q22-bus device
5 = Q22/CD backplane
6 = System power supply
7 = H3600-SA 1/O panel

4-22

2,1,3%5

2, 1, 8, 5
2,1,35

2, 1, 3,5
2,138°5

MEMORY_count_bad pages 2, 1, 3, 5

KA655_RESET

1FRU key:

2,1,85

KA655 CPU System Maintenance

1,4

Table 4-6 (Cont.):

KA655 conaole Dl&plays and FRUs

Normal Error

. Hex Console Console
LED

Display Display

Defaulton

Error

Description

FRU!

Script AD
8

None

230

Continue

None

MEM_bitmap

Continue

1,2,8,5

MEM_data

None

Continue

2,1,3,5

MEM_byte

None

74D

Continue

2,138,5

MEMORY_addr

2,1,8,5
2,135

None

24C

Continue

None

MEMORY_ECC_error

74B

Continue

None

MEMORY_masked_errors

Continue

2,1, 8, 5

MEMORY_correction

2,185
2,1, 8, 5

None

248

Continue

None

MEMORY_address_shorts

740

Continue

None

780

MEMORY_count_errors

2,1,8,5

Continue

None

CQBIC_MEMORY

7241

Continue

1,2,4,3.5

Board reset

1, 4

End of script.

Script AE

None

731

Continmfi

%41

Continue

End of script.

CMCTL_setup_CSRx

1,2,8,5

Board reset

1, 4

Script AF
7

None

780

Continue

None

CQBIC_MEMORY

241

Continue

1,2,4,38,5

Board reset

1,4

End of seript.

1FRU key:
1 = KA655
2 = MS650-BA
|
3 = Memory interconnect cable
4 = Q22-bus device
5 = Q22/CD backplane
6 = System power supply
7= H3600-—-SA 10 pmml

Troubleshooting and Diagnostics

4-23

.%“{

4.3.6 System Halt Messages

Table 4-7 lists meSséges that may appear on the console terminal when a
system error occurs,

Table 4-7: _System Halt Messag_as
"Code

Message

Explanation

202

EXT HLT

204

ISP ERR

External halt, caused either by console BREAK condition, or
because Q22-bus BHALT_L or DBR<AUX_HLT> bit was set
while enabled.

Caused by attempt to push interrupt or exception state onto

the interrupt stack when the interrupt stack was mapped

ACCESS or NOT VALID.

205

DBL ERR

A second machine check occurred while the processor was

206

HLT INST

The processor executed a HALT instruction in kernel mode.

attempting to service a normal exception.

207

SCB ERR3

208

SCB ERR2

70A

CHM FR ISTK

The vector had bits <1:0> = 2.
A change mode instruction was executed when PSL<IS> was

?0B

CHM TO ISTK

?20C

SCB RD ERR

The SCB vector for a change mode had bit <0> set.

A hard memory error occurred during a processor read of an

210

MCHK AV

711

KSP AV

An access violation or an invalid translation occurred during
machine check exception processing.

712

DBL ERR2

713

DBL ERR3

Double machine check error. A machine check occurred
during an attempt to service a kernel stack not valid

719

PSL EXC5

PSL <26:24> = 5 on interrupt or exception.

?1B

PSL EXC7

71E

PSL RE15
PSL RE16

PSL <26:24> = 7 on interrupt or exception.

?1F

PSL RE17

The vector had bits <1:0> = 3.

set.

exception or interrupt vector.

’

An access violation or an invalid translation occurred during
invalid kernel stack pointer exception processing.

Double machine check error. A machine check occurred
during an attempt to service a machine check.

exception.

71A

4-24

PSL EXCé6

PSL <26:24> = 6 on interrupt or exception.

PSL <26:24> = 5 on an rei instruction.

PSL <26:24> = 6 on an rei instruction.
PSL <26:24> = 7 on an rei instruction.

KA655 CPU System Maintenance

. Table 4-8 lists messages issued in re‘spcmse to an error or to a console
-

[

P
)

4.3.7 Console Error Messages
command that was entered incorrectly.

Table 4-8:

Console Error Memges

Code Message
7200
221

CORRPTN

The console data base was corrupted. The console simulates

a power-up sequence and rebuilds its data base.

ILL REF
:

!
722

ILL CMD

723

INVDGT

224

LTL

Explanation

The requested reference would violate virtual memory

protection, address is not mapped, or is invalid in the specified

address space, or value is invalid in the specified destination.

The command string cannot be parsed.

A number has an invalid digit.

The command was too large for the console to buffer. The

message is sent only after the console receives the [Retum) at

225

ILL ADR

726

VAL TOO LRG

727

SW CONF

the end of the command.

The specified address is not in the address space.
The specified value does not fit in the destination.

Switch conflict.
For example, an EXAMINE command
specifies two different data sizes.

728

UNK SW

729

UNK SYM

The EXAMINE

?72A

CHKSM
|

An X command has an incorrect command or data checksum,
If the data checksum is incorrect, this message is issued, and
is not abbreviated to “Illegal command.”
The operator entered a HALT command.
A FIND command failed either to find the RPB or 64 Kbytes
of good memory.
.

The switch is not recognized.

or DEPOSIT symbolic address

is not

HLTED

?2C

FND ERR

22D

TMOUT

Data failed to arrive in the expected time during an X

72E

MEM ERR

72F

UNXINT

Memory error or machine check occurred.

730

UNIMPLEMENTED

731

QUAL NOVAL

732

QUAL AMBG

733

Ambiguous qualifier.

QUAL REQ VAL

234

QUAL OVERF

Qualifier requires a value.
Too many qualifiers.

command.

An unexpected interrupt or exception occurred.

Unimplemented function.

Qualifier does not take a value.

735

ARG OVERF

286

Too many arguments.

AMBG CMD

787

INSUF ARG

Ambiguous command.

Too few arguments.

Troubleshooting and Diagnostics

4-25

g‘i

4.3.8 VMB Error Messages
~ If VMB is unable to boot, it returns an error message to the console.
Table 4-9 lists the error messages and their descriptions.

Table 4-9:
"Message
umber
740

VMB Error Messages
'
Mnemonic

Interpretation

741

DEVASSIGN

NOSUCHDEV

No bootable devices found

742

NOSUCHFILE

Program image not found

243

FILESTRUCT

744

BADCHKSUM

Device is not present

Invalid boot device file structure

Bad checksum on header file

245

BADFILEHDR

746

BADDIRECTORY

747

Bad directory file

FILNOTCNTG

748

ENDOFFILE

Invalid program image format

249

BADFILENAME

24A
?4B

BUFFEROVF
CTRLERR

24C

DEVINACT

74D

EVOFFLINE

MEMERR

24F

SCBINT

750

SCB2NDINT

251

NOROM

752

NOSUCHNODE

753

INSFMAPREG

Bad file header

Premature end-of-file encountered
Bad file name given

Program image does not fit in available memory
Boot device I/O error

Failed to initialize boot device

Device is off line
|
Memory initialization error

Unexpected SCB exception or machine check

Unexpected exception after starting program image
No valid ROM image found

No response from load server

Insufficient Q-bus mapping registers due to invalid

memory configuration, bad memory, or because Q-bus
map was not relocated to main memory

4.4 Acceptance Testing
Perform the acceptance testing procedure listed below, after installing a

system or whenever replacing the following:

KA655 module

MS650-BA module
Memory data interconnect cable
Backplane
DSSI device
H3600-SA

4-26

KA655 CPU System Maintenance

Run five error-free passes of the power-up scripts by entering
the
following command:
|
|
.

S RTO
Press [ETALZ] to terminate the scripts.

Make sure no solid single-bit ECC errors are in memory by entering
the following commands:
>>> T 3000 0 1
>>>

T Al

The first command runs test 30, which enables mapping out of solid
single-bit and multi-bit ECC errors in main memory.
The second command runs script A1, which invokes CPU and
memory
tests without resetting the bitmap to mark only solid multi-bi
t ECC
errors in main memory. This command gives you a quick memory
check,
since most tests run on a 256-Kbyte boundarry.

Perform the next two steps for a more thorough test of memory

>>>

»>>>

The first command runs script A8 for one pass. This command
enables
mapping out of solid single-bit ECC as well as multi-bit ECC
errors. It
will also run script A7 for one pass.

The second command runs script A7 repeatedly. This command runs
the memory tests only and does not reset the bitmap. Press[CThlc) after

two passes to terminate the script. This test takes up to

5 minutes per

pass, depending on the amount of memory in the system. Most

memory diagnostics test memory on a page boundary.

of the

If any of the memory tests fail, they mark the bitmap and continue
with no error printout to the console. An exception is test 40 (count
bad |
pages). If any single-bit or multi-bit ECC errors are detected, they
are
reported in test 40. Such a failure indicates that pages in memory
have

been marked bad in the bitmap because of solid single-bit and/or multibit ECC errors. The error printout does not display the 20 longword
s,
since it is a severity level 1 error.

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

Troubleshooting and Diagnostics

4-27

knowing if the board should be there or not.

To check the memory configuration, enter the following command line:

>>>SHOW MEMORY/FULL
Memory O:

0000000 to OOFFFFFF,

16MB,

0 bad pages

Total of 16MB, 0 bad pages, 104 reserved pages
Memory Bitmap

~00FF3000 to OOFF3FFF,

8 pages

Console

Scratch Area
-00FF4000 to OOFF7FFF,

32 pages

Qbus Map
-00FF8000 to OOFFFFFF,

64 pages

Scan of Bad Pages
>>>

Memory 0 is the KA655 CPU. Memories 1 through 4 are the MS650-BA

memory modules. The Q22-bus map always spans the top 32 Kbytes of

good memory. The memory bitmap always spans two pages (1 Kbyte)
per 4 Mbytes of memory configured.

Use utility 9C to compare the contents of configuration registers
MEMCSR 0-15 with the memory installed in the system:
>>>
T 9C
TOY

=00157FA8

ICCS =00000000

TCRO =00000000

TIRO =000207E3

TCR1

TNIR0O=00000000

=00000000

TIR1 =00000000

TIVRO=00000078

RXCS =00000000

TNIR1=00000000

RXDB =0000000D

TIVR1=0000007C

MSER =00000000

TXCS =00000000

TXDB =00000030

CACR =F5B40040
SSCCR=00D45577

CBTCR=00000004

QBEAR=0000000A

DEAR =00000000

BDR

=FFFFFF50

CADR =0000000C
DLEDR=0000000C

SCR

=0000C000

DSER =00000000

QBMBR=00FF8000
MEM FRU

IPCRn=0020

MEMCSR_0=80000017

1=80400017

MEM _FRU 2

2=80800017

MEMCSR 4=00000000

3=80C00017

5=00000000

6=00000000

7=00000000

MEM FRU 3 MEMCSR_8=00000000
MEM _FRU

9=00000000

10=00000000

MEMCSR12=00000000

11=00000000

13=00000000

14=00000000

MEMCSR16=00000044

15=00000000

17=0000003C

One memory bank is enabled for each 4 Mbytes of memory.

-MEMCSRs map modules as follows:
MEMCSR 0-3
MEMCSR 4-7
MEMCSR 8-11

First MS650
memory
-BA
module
Second MS650-BA memory module
Third MS650
memory
-BA
module

4-28

KA655 CPU System Maintenance

The

* The bank enable bit (<3 15) in all four MEMCSRs for each memory |

~. module is set, which indicates that the memory
bank is enabled.
MEMCSRs <6:0> should equal 17 hex.

If a memory board is not present, bits <31:0> are all
zeros for
the corresponding group of four MEMCSRs. See the values
for
MEMCSR 4-15 in the example.

Bits <25:22> should increment by one starting at zero in any
group
of four MEMCSRs whose bit <31> equals 1. In the example above,
bits <25:22> of MEMCSR 0 and 1 increment by one, resulting in an
increment of four in their longwords. If bit <31> equals 0, <25:22>
should equal zero.

Check the Q22-bus and the Q22-bus logic in the KA655 CQBIC
chip,
and the configuration of the Q22-bus, as follows:
>>>

Scan

show

gbus

of Qbus

-200000DC

-200000DE

-20001468
-2000146A

I/0 Space

(760334)=0000

(760336) =0AA0
(772150)=0000
(772152)=0AR0

-20001920
-20001922
-20001924

(774440) -FF08

-20001926

(774446)=FF09

-20001928
-2000192A
-2000192C

-2000192E

(300)

RQDX3/KDA50/RRD50/RQC25/KFQSA-DISK

(1§4)

RQDX3/KDAS50/RRD50/RQC25/KFQSA~DISK

(120)

DELQA/DEQNA/DESQA

(774442)=FF00

(774444)=FF2B
(774450) =FFA3

(774452) =FF96
(774454)=0050
(774456)=1030

-20001940
-20001942

(774500)=0000
(774502) =0BCO

(260)

TQK50/TQK70/TUB1E/RV20/KFQSA-TAPE

-20001F40

(777500)=0020

(004)

IPCR

Scan

of Qbus Memory

Space

>>>

The columns are described below. The examples listed are
from the last
line of the example above.

First column = the VAX I/O address of the CSR, in hex (20001F4
0).
Second column = the Q22-bus address of the CSR, in octal (777500)
.
Third column = the data, contained at the CSR address, in
hex

(0020).

Troubleshooting and Diagnostics

4-29

Fourth colummn = the device vector in octal, according to the fixed
» or floating Q22-bus and UNIBUS algorithm (004).
"
Fifth column = the device name (IPCR, the KA655 interprocessor
communications register).

Additional lines for the device are displayed if more than one CSR

exists.

The last line, Scan of Qbus Memory Space, displays memory residing
on the Q22-bus, if present. VAX memory mapped by the Q-22 bus map
is not displayed.

If the system contains an MSCP or TMSCP 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 Q-22 bus interrupt logic
If you do not specify the CSR 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:
>>>

20001940

~ You can specify other addresses if you have multible MSCP or TMSCP
devices in the first parameter. This action may be useful to isolate
a problem with a controller, the KA655, or the backplane. Use the
VAX 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.
Check that all UQSSP, MSCP, TMSCP, and Ethernet controllers and
devices are visible by typing the following command line:
>>>

show

device

UQSSP Disk Controller 0
-DUAD

(RF71)

UQSSP

Disk Controller

-DUB1

(RF71)

(772150)

(760334)

UQSSP Tape Controller0 (774500)
-MUAO

(TK70)

Ethernet Adapter 0 (774440)
-XQA0 (08-00-2B-0B-82-29)

4-30

KAB55 CPU System Maintenance

In the example, the console displays the node numbers
of two RF7 1
controllers it recognizes. The line below each node name
and number
1s the logical unit number of any attached devices, DUAO
and DUBI in

this case.

, The UQSSP (TQK70) tape controller and its CSR address
are also
- shown. The line below this display shows a TK70 connecte
d.

The last two lines réfer to DELQA and DEQNA controllers, the Q22-bus
CSR address, logical name (XQAO0), and the station address.

After the above steps have completed successfully, load MDM
and run
the system tests from the Main Menu. If they run successfu
lly, the
system has gone through its basic checkout and you
can load the
software.

4.5 Troubleshooting
This section contains suggestions for determining the cause of ROM-ba
sed
diagnostic test failures.

4.5.1 FE Utility
If any of the tests that run after the IPTs and up to the primary

console

test fail, the major test code is displayed on the LEDs. You
may also be

able to acquire more information with the FE utility.

Running the FE utility is useful if the message, Normal operati

possible, is displayed after the tests have completed and

on not

there is no other

error indication, or if you need more information than what is provided
in
the error display.

The FE utility dumps diagnostic state to the console (Examp
le 4-5). This
state indicates the major and minor test code of the test that
failed, the

ten parameters associated with the test, and the hardwar

e error summary

Troubleshooting and Diagnostics

4-31

Example 4-5:
'

>>>

FE Utllity Example

bitmap=00BF3400,
busmap=00BF8000

length=0C00,
]

checksum=007E

return_stack=201406A8

subtest_pc=2004F4C4

timeout=00000001,

error=0B,

de_error_vector=18,

de_error=FE

severity_ code=02,

previous_error=FEOB5DSD,

total_error_count=0001
00000000, 00000000, 00000000

00000000,

last_exception_pc=20050807

flags=01FFFD7F, test_flags=20
highest_severity=00

leddisplay=05

console_display=5D
save_mchk code=80, save_err
flags=000000
param_1=00000100 2=00000100
3=000000F7 4=00000000
param_6=00000004 7=20050527
8=00000000 9=20140698

The most useful fields displayed above are

5=00000001

10=200521F4

as follows:

De_error_vector, which is the SCB vecto
r through which the unexpected
interrupt or exception trapped if de_error
equals FE or EF.

Total_error_count. Four hex digits showi
ng the number of previous
errors that have occurred.

Previous_error.

Contains the history of the last four

errors.

Each

longword contains four bytes of informatio
n. From left to right these
are the de_error, subtest_log, and test numb
er (copied in both bytes).
*

Save machine check code (save_mchk_code

on error. This field has the same forma

). Valid only if the test halts

t as the hardware error summary

Save error flags (save_err_flags). Valid
only if the test halts on error.

This field has the same format as the hardw
*

are eITor summary register.

Parameters 1 through 10.

Valid only if the test halts on error.

The

parameters have the same forma
t as the hardware error summary

EF in the previous_error field indicates

that an unexpected exception

has
occurred. If any of the tests that annou
nce to the console fail, and the
error code is EF, examine the last longw
ord of the error printout. The last

longword is the hardware error summary
register and contains the machine
check code (<31:24>) and KA655 error
status bits (<23:0>). Table 4-10 lists
the status bits.
N

4-32

KA655 CPU System Maintenance

Table 4-10:
Bit
- 81

Hardware Error Summary Register

, Register

- Description

Machine check code

MSER <6>

MSER <b6>

MSER <4>

MSER <1>

MSER <0>

Unused

MEMCSR16 <31>

MEMCSR16 <30>

MEMCSR16 <29>

MEMCSR16 <25>

MEMCSR16 <24>

'MEMCSR16 <23>

MEMCSR16 <22>

MEMCSR16 <8>

MEMCSRI16 <7>

CBCTR <31>

CBCTR <30>

DSER <7>

DSER <5>

DSER <4>

DSER <3>

DSER <2>

IPCRn <15>

CDAL data parity error.
Mchn chck CDAL parity error.
Machine check cache parity.
Cache data parity error.
Cache tag parity error.

Uncorrectable ECC error.

Two or more uncorrectable errors.
Correctable single-bit error.
Page address bits 25:22 of

Location that caused error.
These four bits point to the
failing 4-Mbyte bank of memory.
DMA read/write error.

CDAL parity error on write.
CDAL bus timeout.
CPU read/write bus timeout.

Q22-bus NXM.
Q22-bus parity error.

Read main memory error.
Lost error.
No grant timeout.

DMA Q22-bus memory error.

Troubleshooting and Diagnostics

4-33

- 4.5.2 V!soplaflng Mamory"Fallums‘
, . This section describes procedures for isolating memory éubsystém fa‘ilurés,
S

particularly when the system contains more than one MS650-BA memory
module.

1.
L}

SHO MEMORY/FULL"
. Use the SHOW MEMORY/FULL 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 MEMORY/FULL after terminating a script
that is taking an unusually long time to run. Press (G7Atc] to terminate
the script after the completion of the current test. (CTIZ] on the
KA655 console takes effect only after the entire script completes.) After
terminating the script, enter SHOW MEMORY/FULL to see if the tests
have marked any pages bad up to that point. See Section 4.4 for an
example of this command.
T A9
>>>

[memory 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 MS650-BA 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:
>>>

4F 2

If a failure is detected for a second of three MS650—-BA modules, for
example, repeat this procedure for all memory modules to isolate the
MS650 module that is 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 baundmy,»type;»
|
>>>

200

By default, most memory tests test one longword on a 256-Kbyte
boundary. If an error is detected, the tests start 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.

4-34

KA655 CPU System Maintenance

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 parameter.

All of the memory tests, with the exception of 40, save MEMCSR17 and
MEMCSR16 memory status and error registers in parameters 7 and 8,

respectively.

TOC

The utility 9C is useful if test 31 or some other memory test
failed
because memory was not configured correctly.

To help in isolating an FRU, examine registers MEMCSR 0-15
by
entering T 9C at the console /O mode prompt (Example 4-6).

Utility

9C is also useful for examining the error registers MSER, CACR, DSER,
and MEMCSR16, upon a fatal system crash or similar event.
.

T 40

Although the SHOW/MEMORY 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/MEMORY does not report. You can use test 40 as an alternati
ve
to running script A9 to detect soft memory errors. Specify the
third
parameter in test 40 (see Table 4-1) to be the threshold for

To allow 0 (zero) errors, enter the following:

soft errors.

>>>T 40140

This command tests the memory on four memory modules. Use it after
running memory tests individually or within a script. If test
40 fails

with subtestlog = 6, examine R5-R8 to determine how many errors
have been detected on the CPU memory and the three memory modules
,
respectively.

Troubleshooting and Diagnostics

4-35

Example 4-6:

Isolating Bad Memory Using T oC

2.,

SCB error 3 PC = 424

>>>

TOY

=00157FAS8

TCRO =00000000

ICCS

=00000000

RXCS =00000000

TIRO =000207E3
TIR1 =00000000
RXDB =0000000D

MSER =00000000

CADR =0000000C

BDR

=FFFFFF50

SCR

=0000C000

DLEDR=0000000C

CACR =F5B40040

TCR1 =00000000

QBMBR=00FF8000

DSER =00000000

TNIRO=00000000

TIVRO=00000078

TNIR1=00000000

TIVR1=0000007C

TXCS =00000000

TXDB =00000030

SSCCR=00D45577

CBTCR=00000004

QBEAR=0000000A

DEAR =00000000

IPCRn=0020

MEM_FRU

1 MEMCSR_0=80000017
MEM FRU 2 MEMCSR_4=81000017

MEM FRU 3 MEMfiSRZB*OOOOOOOO
MEM_FRU 4 MEMCSR12=00000000

1=80400017

2=80800017

3=80C000L17

5=81400017

6=81800017

7=81C00017

13=00000000 14=00000000

15=00000000

9=00000000 10=00000000 11=00000000

MEMCSR16=8154000F 17=0000003C
>>>

In Example 4-6, the diagnostics have passed, but the operating system
does not boot. One of the console/VMB error messages is displayed. Run
utility 9C and examine the error registers. Bit 31 in MEMCSR 16 is set,
which indicates an uncorrectable ECC error in memory. If any of bits
<31:29> are set, there was a memory error. Compare the bits <25:22>
against MEMCSR 0-15. If they match and the MEMCSRn <31> is set,
then the board that was experiencing the failure (the memory FRU) is
the MEM_FRU number on the left.

In Example 4-6, the FRU is the second memory FRU because both
conditions are met by MEMCSR_5 in the MEM_FRU 2 row. The
following conditions are shown in Example 4-7:

4-36

MEMCSR_5 matches MEMCSR16,
Number) match.

since bits <25:22> (Bank

The Bank Enable bit <31> in MEMCSR_5 is set, indicating that the
bank number is valid.

KA655 CPU System Maintenance

Example 4-7:
P

9C—Conditions for Determining a Memory FRU
|

MEMCSR
5 = 81400017 Hex = 1000 0001 0100 0000 0000 0000 0001 0111

bit 31 set

25:22 match

| 4.5.3 Additional Troubleshooting Suggestions
Note the following additional suggestions when diagnosing a possible

memory faflure.

If more than one memory module is failing, you should suspect the KA655
module, CPU/memory cable, backplane, or MS650 modules as the cause of
failure.
Always check the seating of the memory cable first before replacing a
MS650 module. If the seating appears to be improper, rerun the tests. Also
remember to leave the middle connector disconnected for a three-connector

cable when the system is configured with only one MS650.

If you are rotating MS650 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 multi-bit 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 error log, use the KA655 ROM-based

diagnostics to see if it is an MS650 problem, or if it is related to the
KA655, CPU/memory interconnect cable, or backplane. Follow steps 1-3 of
Section 4.4 and Section 4.5.2 to aid in isolating the failure.
Use the SHOW QBUS, SHOW DEVICE, and SET HOST/DUP commands
when troubleshooting I/O subsystem problems.

Use the CONFIG command to help with configuration problems or when
installing new options onto the Q-bus. See the command descriptions in

. Chapter 3.

Troubleshooting and Diagnostics

4-37

4.6 Loopback Tests

. You cari use external loopback tests to localize problems with the console.

Check that the dc power and the pico fuse on the KA655 are functioning

correctly. The 1.5 ampere pico fuse (12-10929-08). is located near the
handle on the component ‘side and is numbered F1. The fuse is shown
-in Figure 1-1. If the fuse is bad, the H3600-SA hex LED display will not
light.

4.6.1 Testing the Console Port
To test the console port at power-up, set the power mode switch on the
H3600-SA to the test position, and install an H3103 loopback connector
into the MMP of the H3600. The H3103 connects the console port transmit
and receive lines. At power-up, the SLU_EXT _LOOPBACK IPT then runs
a continuous loopback test.

While the test is running, the LED display on the CPU I/O panel 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 KA655,
the H3600-SA, the cabling, or the CPU module.
To test out to the end of the console terminal cable:

1. Plug the MMJ end of the console terminal cable into the H3600—SA.
2.

Disconnect the other end of the cable from the terminal.

Place an H8572 adapter into the disconnected end of the cable.

Connect the H3103 to the H8572.

4.7 Module Self-Tests
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. The test
- usually does not check the module Q22-bus interface, the line drivers
- and receivers, or the connector pins—all of which have relatively high
failure rates.

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

4-38

KA655 CPU System Maintenance

The following modules do not have LED self-test indicators:
- DFAO1
PPV1l

DRQ3B
DZQ11

“KLESI

LPV11

TSV05

' The following modules have one green LED, which indicates that the
module is receiving +5 and +12 Vdc:
CXAl6
CXB16
CXYO08

Table 4-11 lists ioapback connectors for common KA655 system modules.

Refer to Microsystems Options for a description of specific module self-tests.

Table 4-11:

Loopback Connectors for Q22-Bus Devices

Device

Module Loopback

CXA16/CXB16
CXY08
DELQA

H3046 (50-pin)
12-22196-02

H3103 + H8572!

DPV11

H3259

DSSIZ

DZQ11

Ethernet®

o
.

H3197 (25-pin)
H3260

12-15336-00 or H325

H329 (12-27351-01)

LPV1l

None

KA655/H3600-SA

H3103

KMV1A

Cable Loopback

H3255

None
H3103 + H8572

H3251

1Use the appropriate cable to connect transmit-to-receive lines.
H3101 and H3103 are double-ended cable connectors.

2For DSSI to KFQSA or RF-series connector, use 17-02216-01 plus H3281 loopback.
For connection to end of bus, use DSSI loopback.

3For ThinWire, use H8223-00 plus two H8225-00 terminators.
For standard Ethernet, use 12-22196-02.

Troubleshooting and Diagnostics

4-39

4.8 RF-Series ISE Troubleshooting and
Diagnostic

, . 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 enclosure’s OCP. The ISE also has a red fault LED, but
it is not visible from the outside of the system enclosure.

’

P |
o

- Ifthe 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
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 LED 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 LEDs go out for about 1 second, then the
red fault LED lights. In this case, run either DRVTST, DRVEXR, or
- PARAMS (described in the next sections) to determine the error code. .

Three configuration errors also commonly occur:
More than one node with the same node number
*

Identical node names

Identical unit numbers

The first error cannot be detected by software. Use the SHOW DSSI

command to display the second and third errors. This command lists each
device connected to the DSSI bus by node name and unit number.

If the ISE is connected to the OCP, you must install a unit ID plug in the

corresponding socket on the OCP. If the ISE is not connected to the OCP,
it
reads the unit ID from the three-switch DIP switch on the side of the drive.

4-40

KAB655 CPU System Maintenance

The RF-series ISE contains the following local programs (described in the
|
|
o
RS
~
following sections):

. DIRECT

DRVTST
DRVEXR
HISTRY

ERASE

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

A utility that erases all user data from the disk

A utility that allows you to look at or change drive status, history, and
parameters

A description of each local program follows, including a table showing the
dialogue of each program. The table also indicates the type of messages
contained in the dialogue, although the screen display does not indicate
the message type. Message types are abbreviated as follows:

Q—QuestionI—Information
T—Termination
FE—Fatal error
To access these local programs, use the console SET HOST/DUP command,
which creates a virtual terminal connection to the storage device and the
designated local program using the Diagnostic and Utilities Protocol (DUP)
standard dialogue.
|

" Once the connection is establylished; the local progi'am is in control. When |
the program terminates, control is returned to the KA655 console.

To abort or prematurely terminate a program and return control to the
KA655 console, press

Troubleshooting and Diagnostics

4-41

4.8.1 DRVTST

~ DRVTSTis a comprehamivé functionality test‘“ Errors detected by this test
.

are isolated to the FRU level. The messages are listed in Table 4-12.

Table 4-12: DRVTST Messages
- Message
I
Q
Q

5 minutes to complete.

Test passed.

Or:

Unit is currently in use.!

Operation aborted by user.

FE
FE

xxoxx—Unit diagnostics failed.?
xxxx—Unit read/write test failed.?

1Either the dmm is inoperative, in use by a host, or is currently running another local program.
2Refer to the diagnostic error code list at the end of this chapter.

Answering No to the first question (“Write/read...?”) results in a read-only
test. DRVTST however, writes to a diagnostic area on the disk. Answering
Yes to the first question causes the the second question to be displayed.
Answering No to the second question (“Proceed?”) is the same as answering
No to the first question. Answering Yes to the second question permits write
and read operations anywhere on the medium.

DRVTST resets the ECC error counters, then calls the timed I/O routine.
After the timed I/O routine ends (5 minutes), DRVTST saves the counters
again. It computes the uncorrectable error rate and byte (symbol) error

rate. If either rate is too high, the test fails and the appropriate error code
is displayed.

4-42

KA655 CPU System Maintenance

4.8.2 DRVEXR

~ The DRVEXR local program exercises the ISE. The test is data

transfer

intensive, and indicates the overall integrity
of the device. Table 4~13 lists
the DRVEXR messages.

Table 4-13: DRVEXR Messages
Message
1

Write/read anywhere on the medium? [1=yes/(0=no))]

User data will be corrupted. Proceed? [1=yes/(0=no)]

Test time in minutes? {(10)-100])

ddd minutes to complete.
dddddddd blocks (512 bytes) transferred.

dddddddd bytes in error (soft).
dddddddd uncorrectable ECC errors (recoverable).

I
T

Complete.

Or:
FE

Unit is currently in use.!
- Operation aborted by user.

'FE
" FE

xox—Unit diagnostics failed.2
xxxx—Unit read/write test failed.2

1Either the drive is inoperative, in use by a host, or is currently running another
local program.
2Refer to the diagnostic error code list at the end of this chapter.

Answering No to the first question (“Write/read...?”) results
in a read-only
test. DRVEXR however, writes to a diagnostic area on
the disk. Answering

Yes to the first question causes the second question to be displaye

Answering No to the second question (“Proceed?”) is the same
as answering
No to the first question. Answering Yes to the second questio
n permits write
and read operations anywhere on the medium.
NOTE: If you press the write-protect switch on the OCP (LED
on)

answer Yes to the second question, the drive does not

and you

allow the test to run. B

DRVEXR displays the error message 2006—Unit read |write test
failed. In
this case, the test has not failed, but has been prevented from
running.

Troubleshooting and Diagnostics

4-43

DRVEXR saves the error counters, then calls the timed I/O routine. After
the timed I/O routine ends, DRVEXR saves the counters again. It then
reports the total number of blocks transferred, bits in error, bytes in error,

" "and uncorrectable errors.

DRVEXR uses the same timed I/O routine as DRVTST, with two exceptions.
First, DRVTST always uses a fixed time of five minutes, while you specify
the time of the DRVEXR routine. Second, DRVTST determines whether
the drive is good or bad. DRVEXR reports the data but does not determine
the condition of the drive.

4.8.3 HISTRY
The HISTRY local program displays information about the history of the
ISE. Table 4-14 lists the HISTRY messages.

Table 4-14:
Message

HISTRY Messages

Type

Field Length

47 ASCII characters

4 ASCII characters

Field Meaning
Copyright notice

12 ASCII characters

6 ASCII characters

Product name
Drive serial number

Node name

1 ASCII character

8 ASCII characters

17 ASCII characters

6 ASCII characters

Power-on hours

6 ASCII characters

4 ASCII characters

Power cycles

Allocation class
Firmware revision level

Hardware revision level

Hexadecimal fault code
Complete.

1This displays the last 11 fault codes as information messages.

Refer to the diagnostic error code list at the end of this chapter.

4-44

KA655 CPU System Maintenance

RF71
EN01082
SUSAN

0
RFX V101

RF71 PCB-5/ECO-00
617
21
AQ4F

AQ4F

Al03

AQ4F
A404
AQ4F
A404

AQ4F

A404

AOQ4F
A404
Complete.

If no errors have been logged, no hexadecimal fault codes are displayed.

' 4.8.4 ERASE
The ERASE local program overwrites application data on the drive while
leaving the replacement control table (RCT) intact. This local program is
used if an HDA must be replaced and the customer wants to protect any
confidential or sensitive data.
Use ERASE only if the HDA must be replaced and only after you have
backed up the customer’s data.

Troubleshooting and Diagnostics

4-45

Table 4-15 lists the ERASE messages:

, Table 4<15: ERASE Messages
Message
~

I
Q
Q

Message

6 minutes to complete.

Complete.

Or:

Unit is currently in use.

Operation aborted by user.

xxxx—Unit diagnostics failed.!

xxxx—Operation failed.?

1Refer to the diagnostic error code list at the end of this chapter.
2xxxx = one of the following error codes:

000D : Cannot write the RCT.
000E : Cannot read the RCT.
000F : Cannot find an RBN to revector to.
0010 : The RAM copy of the bad block table is full.

If a failure is detected, the message that indicates the failure is followed
by one or more messages that contain error codes.

4.8.5 PARAMS
The PARAMS local program supports modifications to device parameters
that you may need to change, such as device node name and allocation
class. You invoke it in the same way as the other local programs. However,
you use the following commands to make the modifications you need:
EXIT

Terminates PARAMS program

HELP

Prints a brief list of commands and their syntax

SET

Sets a parameter to a value

SHOW

Displays a parameter or a class of parameters

STATUS

SR
Displays module configuration, history, or current counters, depending on the

- status type chosen
WRITE

4-46

:
Alters the device parameters

KA655 CPU System Maintenance

4.8.5.1 EXIT

~ Use the EXIT command to terminate the PARAMS local program.

4852 HELP

Use the HELP command to display a brief list of available PARAMS
commands, as shown in the following example:
PARAMS> help
- EXIT
-

HELP

SET

{parameter

SHOW {parameter |
/ALL
/CONST
/SERVO
/8CS
/DUP
STATUS [type]
CONFIG
LOGS

value

/class}

/DRIVE
/MSCP

DATALINK

PATHS
WRITE

PARAMS>

4.85.3 SET

Use the SET command to change the value of a given parameter. To
abbreviate, use the first matching parameter without regard to uniqueness.

For example, SET NODE SUSAN sets the NODENAME parameter to SUSAN.
The following parameters are useful to Field Service:
ALLCLASS
FIVEDIME

UNITNUM
FORCEUNI

The controller allocation class. The allocation class should be set to match
that of the host.
True (1) if MSCP should support five connections with ten credits each. False

(0) if MSCP should support seven connections with seven credits each.
The MSCP unit number.

True (1) if the unit number should be taken from the DSSI ID. False (0) if the
UNITNUM value should be used instead.

NODENAME

The controller’s SCS node name.

FORCENAM

True (1) if the SCS node name should be forced to the string RF71x (where x
is a letter from A through H that corresponds to the DSSI bus ID) instead of
using the NODENAME value. False (0) if NODENAME is to be used.

Troubleshooting and Diagnostics

4-47

4.8.5.4 SHOW

Use the SHOW command to display the settings of a parameter
or a class

7 of parameters. It displays the full name of the parameter (8 characters or
less), the current value, the default value, radix and type, and any flags
associated with each parameter.

4.8.5.5 STATUS

Use the STATUS command to display module configuration, history, or

current counters, depending on the type specified. The type is the optional
ASCII string that denotes the type of display desired. If you omit the type,

all available status information is displayed. If present, the type may be
abbreviated. The following types are available.

CONFIG

Displays the module name, node name, power-on hours, power cycles, and

other such configuration information.
applicable.

Unit failures are also displayed, if

LOGS

Displays the last eleven machine and bug checks on the module. The display
includes the processor registers (D0-D7, A0—A7), the time and date of each
failure (if available; otherwise the date 17 November 1858 is displayed), and

DATALINK

Displays the data link counters.

some of the hardware registers.

PATHS

Displays available path information (open virtual circuits) from the point of

view of the controller. The display includes the remote node names, DSSI IDs,
software type and version, and counters for the messages
and datagrams sent
and/or received.

4.8.5.6 WRITE
Use the WRITE command to write the changes made while in PARAMS
to the drive nonvolatile memory. The WRITE command is similar to the
VMS SYSGEN WRITE command. Parameters are not available, but you
must be aware of the system and/or drive requirements and use the WRITE
command accordingly or it may not succeed in writing the changes.
The WRITE command may fail for one of the following reasons:

You altered a parameter that required the unit, and the unit cannot be
acquired (that is, the unit is not available to the host). Changing the
unit number is an example of a parameter that requires the unit.

You altered a parameter that required a controller initialization, and
you replied negatively to the request for reboot. Changing the node
name or the allocation class are examples of parameters that require
controller initialization.
|
|

Initial drive calibrations were in progress on the unit. The use of the

WRITE command is inhibited while these calibrations are running.
Y

4-48

KA655 CPU System Maintenance

4.9 Diagnostic Error Codes
Di%,gnoatizc error codes appear when you are running DRVTST, DRVEXR,
or PARAMS. Most of the error codes indicate a failure of the drive module.
The error codes are listed in Table 4-16. If you see any error code other
than those listed in Table 4-16, replace the module.

Table 4-16: _RF-Series ISE Diagnostic Error Codes
Code

Message

Meaning

2032/A032

Failed
to see FLT go away
’

FLT bit of the spindle control status register was
asserted for one of the following reasons:
1. Reference clock not present
2. Stuck rotor
8. Bad connection between HDA and module

203A/A03A

Can't spinup, ACLOW is
set in WrtFlt

Did not see ACOK signal, which is supplied by the
host system power supply for staggered spin-up.

1314/9314

Front panel is broken

Could be either the module or the operator control
panel or both.

Troubleshooting and Diagnostics

4-49

‘ Appéndii A
Configuring the KFQSA
' This appendix describes the KFQSA storage adapter and explains how to:
*

Configure the KFQSA storage adapter at installation

Enter console I/O mode

Run the Configure utility

Program the EEROM on the KFQSA

Reprogram the EEROM on the KFQSA

Change the ISE'’s allocation class and unit number

A.1 KFQSA Overview
The KFQSA module is a storage adapter that allows Q-bus host systems
that support the KFQSA to communicate with storage peripherals based
on
the Digital Storage Architecture (DSA), using the Digital Storage System
Interconnect (DSSI). In a KA655-based system, one KFQSA module can

connect up to six RF-series integrated storage elements (ISEs) to
the host
system using a single DSSI bus cable.

The KFQSA contains the addressing logic required to make a connecti
on
between the host and a requested ISE on the DSSI bus. Each ISE has
its own controller, which contains the intelligence and logic necessary
to
control data transfers over the DSSI bus. The KFQSA presents
a mass

storage control protocol (MSCP) U/Q port for each ISE.

The EEROM on the KFQSA contains a configuration table. After you install
the KFQSA, you program the EEROM with the CSR address for each
ISE
in the system.

Configuring the KFQSA

A-1

A.1.1 Dual-Host Configuration

. An RFfserim ISE has dfiabhaat capability built into the firmware, which
allows the device to maintain connections with more than one KFQSA
storage adapter. You can connect more than one KFQSA to the same DSSI
bus, which allows each KFQSA to access all other devices on the bus. For
_more information on dual-host capability, see Chapter 2, Section 2.4.5.

A.2 Configuring the KFQSA at Installation
At installation, configure the KFQSA as follows:

CAUTION: Static electricity can damage integrated circuits. Use the
wrist strap and antistatic mat found in the Antistatic Kit (29-26246)
when you work with the internal parts of a computer system.
1.

Check the KFQSA module for the presence of a jumper intended for
manufacturing use only. The location of this jumper is shown in
Figure A-1. Remove the jumper, if present.

Use the four-position DIP switchpack shown in Figure A-1 as follows to
set a temporary CSR address that enables you to access the EEROM:

~a. Set switches 1, 2, 3,' and 4 to reflect a fixed CSR address to allow the

KFQSA to be programmed. Example A-1 shows the correct switch

settings.

A-2

Install the KFQSA adapter module into the backplane according to
the procedures in the appropriate enclosure maintenance manual.

KAB655 CPU System Maintenance

Figure A-1:

KFQSA Module Layout (M7769)
"JUMPER

(FOR MANUFACTURING

USE ONLY)

FOUR-POSITION
SWITCHPACK

LEDs

MLO~001878

Configuring the KFQSA

A-3

g‘fi

Example A-1: KFQSA (M7769) Service Mode Switch Settings
e
1./‘

KfQSA,Fouerasition Switchpack

‘Switches:

S/N
Mode

.
Fx/Fl

MSB

LSB

S/N = Service mode/Normal operating mode
Fx/Fl1 = fixed/floating CSR address
0 =on,

1 = off

A.2.1 Entering Console /0 Mode
After installing the KFQSA, you issue a series of commands to the KA655
system at the console prompt (>>>) in order to program the EEROM on the
KFQSA. You may type these commands in either uppercase or lowercase
letters. Unless otherwise specified, type each command, then press

Enter the console I/O mode as follows:

1. Set the Break Enable/Disable switch on the CPU cover panel to the
‘enable position (up; dot inside circle).
Set the on/off power switch to on (1).
3.

When the power-up self-tests complete, the console prompt appears, as
shown in Example A-2.

A-4

KA655 CPU System Maintenance

Example A-2: | Entarlng chsole Mode Display
Perfiprmlng normal

system tests.

40..39..38..37..36..35..34..33..32. 31.‘30..29*.28.,27,,26.*25**
Oa.‘07..06.m05.,04..03,¢
Tests aumpleted
P>

'A.2.2 Displaying Current Addresses
Type sHow QBUS to display the current Q22~bm addresses (Example A-3).
Note that the KFQSA adapter appears in service mode as KFQSA #0.

Example A-3:
>>>

SHOW QBUS Display

show gbus

Scan of Qbus

I/0 Space

(774420)

0000

(774422)

0ARO

-20001920

(774440)
(774442)
(774444)

-2000192A

(774452)

-2000192C

(774454)

-2000192E

(774456)

-20001940
-20001942
-20001F40

(774500)
(774502)

(777500)

(000)

KFQSA #0

(120)

DELQA/DEQNA/DESQA

0000
0BCO

(260)

TQKS0/TQK70/TU81E/RV20/KFQSA-TAPE

0020

(004)

IPCR

FF08

FF0O0

FF2B

FF09
'FFA3

FF96
B

-20001924

-20001926 (774446)
-20001928 (774450)

~-20001922

-20001910
-20001912

8000
1030

Scan of Qbua Memory Space
>>>

Configuring the KFQSA

A-5

A.2.3 Rurming the Configure Utility

- Now that you have reconfigured the system by installing the KFQSA
storage adapter, you must run the Configure utility to find the correct
address for each device and modulein the system. The Configure utility
uses floating address space rules.

- Run the Configure utility as follows. Refer to Example A—4.

At the console prompt, type CONFIGURE, then type HELP at the
Device, Number? prompt for a list of devices that can be configured.

NOTE: Some of the devices listedin the HELP display are not supported
by the KA655-AA CPU.
2.

For each device in the system, type the device name at the
Device,Number? prompt. If you have more than one of the same
type, type a comma followed by the total number of that device. In
Example A—4, the system contains one KFQSA with six ISEs.
Be sure you list all the devices: those already installed and those you
plan to install.

3. Type ExiT. The Configure utihty displays an address and vector
assignment for each device. Example A—4 showa the address and vector
assignments and the device input.

For all modules except the KFQSA, verify that the CSR addresses are
set correctly by comparing the addresses listedin the SHOW QBUS
command with those listedin the Configure utility display. If necessary,
remove modules from the backplane and reset switches or jumpers to
the addresses in your Configure display, using module removal and
replacement proceduresin BA213 Enclosure Maintenance.

Write down the addresses for the KFQSA devices.

A-6

KA655 CPU System Maintenance

Fy
A

Example A-4:

Configure Display

>>> éonfigure
. Enter device configuration,
Device, Number? help
Devices:

HELP,

or EXIT

LPV1l

KXJ11

bLV11J

DzQ11

DzZV11l

RLV12

TSVOS5

RXV21

DRV11W

DRV11B

DPV11

DELQA
RQC25
KFQSA-TAPE

DEQNA
KFQSA-DISK
KMV1l

DESQA
TQKS50
IEQ11

RQODX3
TQK70
DHQl11

KDAS0
TUS1E
DHV11

- DMV11
RRDS0
RV20

DFA01

CXAl6

CXB1l6

CXY08

VCBO1

QVSS

LNV11

LNV21
KWV11C

QPSS
ADV11D

DSV11
AAV11D

ADV11C
VCB02

AAV11C
QDSS

AXV11C
DRV11J

DRQ3B

VsSv2l

IBQO1

IDV11A

IDV11B

IDV11C

IDV11D

IAV11A,

IAV11B

MIRA

ADQ32

DTCO4

DESNA

IGQ1l1l

Numbers:
1l to 255,

default is 1
Device,Number? kfgsa-disk,
6
Device, Number? desga
Device, Number? tgk70
Device, Number? exit

Address/Vector Assignments
-774440/120 DESQA
-772150/154 KFQSA-DISK
!
~-760334/300 KFQSA-DISK
!
-760340/304 KFQSA-DISK
!
-760344/310 KFQSA-DISK
!
-760350/314 KFQSA-DISK
!

~-760354/320 KFQSA-DISK

Node O (assigned in order,

0 to n)

Node 1
Node 2

Node 3
Node 4
Node 5

-774500/260 TQK70

Configuring the KFQSA

A-7

A3 Programming the KFQSA

Program the configuration table in the EEROM of the KFQSA to include
all ISE& on the DSSI bus, as follows. Refer to Examples A-5 through A-8.

IMPORTANT. Use this procedure for the first or second KFQSA. In dual-

. host configu rations, both the first and the second KFQSA adapters must

" have equal access to the system disk and to any DSSI ISEs. Program the
KFQSA in the first system, then program the KFQSA in the second system.
1.

Determine the DSSI node plug address for each ISE you are configuring.
Start wilith node 0, then continue up through node 5. In Example A4,
nodes 0, 1, 2, 3, 4, and 5 are used; nodes 6 and 7 are saved for IG‘QSA
adapters Reserve node 7 for the first KFQSA and node 6 for the second
KFQSA.

Figure A-2 shows an example of the nodes and addresses in a dual-host

configuration.

At the console prompt on each system, type SET HOST/UQSSP/MAINT/SERV
0 to set host to the KFQSA.

3. Type HELP to display a list of mpported commands.
4,

Program the KFQSA to include each DSSI dmncein the system:
a.

For each ISE: type sET, followed by the node number, the CSR

address (from the list of addresses you obtained from the Configure

utility), and the model number (disk ISE’s are model 21).

For the KFQSA in the second system of a dual-host configuration,

Type sHow to display the configuration table you just programmed.

type SET 6 /KFQSA to set the node to 6 (Example A-6).

Check the display to make sure the addresses are correct.

A-a

Tyg)e EXIT to save the configuration table, or QuiT to delete the

table.

KA655 CPU System Maintenance

Figure A-2:

Dual-Host Configuration Nodes and Addresses (Example)

SYSTEMB

SYSTEM A

(FIRST KFQSA)

(SECOND KFQSA)

INODE 5{NODE 4|NODE 3[«———>|NODE 2|NODE 1{NODE of
DSSI
BUS

—KFQSA

KFQSA

(NODE 7)

(NODE )

NODE
0
1
2
3
4
5

NODE

ADDRESS

0
1
2
3
4
5

772150
760334
760340
760344
760350
760354

KFQSA

ADDRESS
772150
760334
760340
760344
760350
760354

KFQSA

MLO-002414

Configuring the KFQSA

A-9

éi

Example A-5:

Display for Programming the First KFQSA

!0 refers to first KFQSA

!in the system.

UQSSP

Controller

(772150)

. Enter SET, CLEAR, SHOW, HELP, EXIT, or QUIT

‘Node

CSR Address

Model

KFQSA

? help

Commands:

SET <WODE> /KFQSA

!Sets KFQSA DSSI node
'number

SET <NODE> (CSR ADDRESS> <MODEL>

!Enables a DSSI device

CLEAR <NODE>

!Disables

a DSSI

SHOW

!Displays

current

HELP

EXIT

device

!configuration

!Displays

display

this

!Saves the KFQSA program

QUIT

!Does not

save the KFQSA

!program
Parameters:
<NODE>

<CSR &DDRESS»

<MODEL> |

set

772150 21

set

760334

set

21
2 760340 21

set

3 760344

through
7

1760010

777774

!21 (disk) or 22 (tape)

? set

5 760354 21

760350 21

show

Node

CSR Address

762105

760334

760340

760344

760350

21
21

760354

Model

W mm———— KE'QSA
? exit

Programming the KFQSA..u

!Note from the system that
!the KFQSA is

!being programmed.

A-10

KA655 CPU System Maintenance

Example A-6:

Dlsplay for Programming the KFQSA in a Dual-Host

ConflgurMIOn (Second Systam)
e

>>>

set

host/uqasp/maint/servica

!from the

(772150)

Enter SET,

SHOW,

CLEAR,

CSR Address

to the KFQSA in the

consocle

the

second

!system.

UQSSP Controller

Node

refers

!second system of a dual-host
!lconfiguration.
You set host

HELP,

EXIT,

or QUIT

Model

KFQSA

? help
Commands:

SET <NODE>./KFQSA

!Sets KFQSA DSSI node

SET <NODE> <CSR_ADDRESS> <MODEL>
CLEAR <NODE>

SHOW

!configuration

EXIT

QUIT
v

- <NODE>

<CSR_ADDRESS>

set

760334

760340 21
760344

set

? set

121

772150 21

? set

!0 through 7
1760010 to 777774

<MODEL>

set

!Displays this display
!Saves the KFQSA program
!Does not save the KFQSA
!program

Parameters:

!Enables a DSSI device
!Disables a DSSI device

!Displays current

HELP

;

!number

(disk)

or 22

(tape)

760350 21
5 760354 21

6 /KFQSA

? show
Node

CSR Address

762105

760334

760340

760344

760350

21
21

760354
W
————— KFQSA
? exit

Model

Programming the KFQSA...

!Note from the system that
!the KFQSA is being
!programmed.

Configuring the KFQSA

A-11

To allow the new firb‘gmm to take efi‘eét, turn the éjmtem powér off by

" setting the on/off switch to off (0). 6.

Remove the KFQSA from the backplane.

On the KFQSA, set switch 1 on the four-position switchpack to Off (1).

(Figure A-1 shows the location and position of the switchpack.) This
action sets the KFFQSA to the normal operating mode; switches 2, 3,
and 4 are disabled and the DSSI addresses are read from the EEROM.

Reinstall the KFQSA in the backplane.
9.

Confirm that the unit ID plugs on the enclosure’s operator control panel

(OCP) match the node ID’s you just programmed.

10. Power on the system by setting the on/off switch to on (1). Wait for the
self-tests to complete.

11. At the console prompt, type sHOw QBUS to verify that all addresses are
present and correct, as shown in Example A-7.
12. Type sHow DEVICE to verify that all ISEs are displayed correctly, as
shown in Example A-8.
’

A-12

KA655 CPU System Maintenance

Example A-7: SHOW QBUS Display
>>> show gbus

‘Scah of Qbus I/0 Space
-200000DC (760334)=0000

(300) agnxalxfiaso/Ransolnoczs/xrasn~nxsx

-200000DE

(760336) =0AA0

-200000E2
-200000E4
-200000E6
=-200000E8
-200000EA
-200000EC
-200000EE
-20001468
-2000146A
-20001920
-20001922

(760342)=0AA0
(760344)=0000 (310) RODX3/KDAS50/RRD50/RQC25/KFQSA-DISK
(760346)=0AA0
(760350)=0000 (314) RODX3/KDAS0/RRD50/RQC25/KFQSA-DISK
(760352) =0AA0
(760354)=0000 (320) RODX3/KDAS0/RRD50/RQC25/KFQSA-DISK

-200000E0 (760340)=0000 (304)

-20001924

(760356) =0AA0

RODX3/KDA50/RRD50/RQC25/KFQSA-DISK

(772150)=0000

(154)

RODX3/KDAS50/RRD50/RQC25/KFQSA-DISK

(774440) -FFO8

(120)

DELQA/DEQNA/DESQA

(260)

TQKS50/TQK70/TUB1lE/RV20/KFQSA-TAPE

(772152) =0AA0
(774442) =FF00
(774444)=FF2B

-20001926
-20001928
-2000192a

(774446) =FF09
(774450) =FFA3

-2000192C
-2000192E
-20001940
-20001942
-20001F40

(774454)=0050
(774456)=1030
(774500) =0000
(774502) =0BCO

(774452) =FF96

(777500)
= (004) -IPCR

Scan of Qbus Memory Space
>>>

Configuring the KFQSA

A-13

4%
Y

Example A-8: SHOW DEVICE Display
>>> show device
UQSSP Disk Controller 0
-DUAO
, (RF71)

(772150)

UQSSP Disk Controller 1w(760334)

_=~DUB1

(RF71)

)

'VQSSP Disk Controller 2 (760340)
-DUC2

(RF71)

UQSSP Disk Controller 3
-DUD3 (RF71)

(760344)

UQSSP Disk Controller 4
-DUE4 (RF71)

(760350)

UQSSP Disk Controller 5
-DUFS5 (RF71)

(760354)

UQSSP

Tape Controller

(774500)

-MUAQ

(TK70)

Ethernet Adapter 0 (774440)
-XQA0 (08-00-2B-09-A3-96)

A-14

KA655 CPU System Maintenance

§i

A.4 neproér"amming the KFQSA

When you add a new DSSI device to a aystem with at least one RF-aenes |

ISE that has been programmed correctly, you must reprogram each KFQSA
~ on the DSSI bus to include the new device(s) as follows:
1. /Enter console I/O mode, using the procedurein Section A.2.1.

At the console prompt, type sHow DEVICE for a display of all devices
currently in the system. The display includes tape drives and the
Ethernet adapter, as shown in Section A.3, Example A-8.

This display lists the VAX address and port name of the device, such
as DUAO RF71.

Type sHow oBUS for a display of the eight-digit VAX address (hex) for

Find the e).ghtwmgxt VAX address for an ISE attached to the KFQSA
that you are reprogramming. Use the SET HOST command to enter
the KFQSA through an existing port and edit the configuration table,

each device, as shown in Section A.3, Example A-7.

as follows. Refer to Example A-9.

Type SET HOST/UQSSP/MAINT followed by the VAX address.

Use the SET and CLEAR commands to reccmfigure the KFQSA, as

Type sHow to display the new KFQSA configuratmn table setting.

shownin Example A-9.

Type EXIT to save the configuration table, or QuiT to cancel the

reprogramming.

Configuring the KFQSA

A-15

‘i‘i

Example A-9:

Heprogrammlng the KFQSA Display

;f >>> SQE host/ugssp/maint 260014&5
'

UQSSP Controller

Node ~

CSR Address

Model

772150
760334

- 21
21

o2
3

760340
760344

21
21

4
5

760350

760354

0
.1

(772150)

————- KFQSA -=====

? clear 5
? show

Node
0
1
2
3
4
7

~
CSR Address
772150
760334
760340
760344
760350

Model
21
21
21
21
21

—==== KFQSA -====-

? set 5 760354 21
? show
Node
0
1

2
3
4
7

CSR Address
772150
760334

760340
760344
760350
0

Model
21
21
21
21
21

e KFQSA -===—-

? exit
Programming the KFQSA...

!Note from the

system that the

!KFQSA is being programmed.

A-16

KA655 CPU System Maintenance

~ A.5 Changing the ISE Allocation Class and Unit
Number
This section describes how to change the ISE allocation class and unit
number.

If the system is part of a cluster, you must change the default allocation
class parameter. The ISEs ship with the allocation class set to zero.

NOTE: In a dual-host configuration, you must assign the same allocation
class to both host systems and to the RF-series ISEs. This allocation class

must be different from that of other systems or of hierarchical storage
controllers (HSCs) in a cluster.

For most cbnfimnafions, you will not need to change the default unit

numbers.

Change the allocation class and unit number parameters using the consolebased DUP driver utility, as follows. Refer to Example A-10.
1.

Determine the correct allocation class for the RF-series ISEs according
to the rules on clustering.

Enter console I/O mode, using the procedure in Section A.2.1.

At the console prompt, type SET HOST/nUP/UstP)DISk 0 paraMs (0

through 5 for the ISE to which you want to connect) to start the DUP
server.

At the paraMs> prompt, type sHow ALLCLASS to check the current
allocation class.

5. Type SET ALLCLASS 2 to set the new allocation class to 2 (or type the
number you desire).

Type snow uNITNUM to check the unit number.
7.

To change the ISE’s unit number from the default value, type sET
UNITNUM n (where n is the new unit number). For example, type SET
UNITNUM 20 to change the unit number from 0 to 20.

Type SET FORCEUNI 0 to set the forceunit flag to zero in order to use a
non-default value. If you do not change the FORCEUNI parameter, the
drive unit number defaults to the number of the corresponding DSSI
- plug on the operator control panel (OCP).
|
|
Type sHow aLLcLass to check the new allocation class.

10. Type ssow UNITNUM to show the new unit number.
AN

Configuring the KFQSA

A-17

%g‘

11. Type sHow FORCEUNI to show the new forceunit flag values.

12, Type WRITE, then type Y to save the new values into the EEROM, or &
..

to cancel the reprogramming.

13. Type sHow DEVICE to make sure you have programmed the first ISE to
have a unit number of 20. When you boot the operating system, the
display shows the new allocation class and new unit number.

‘Exampw A-10: Display for Changing Allocation Class and Unit Number
>>> set host/dup/e@ssp/diék 0 params
Starting DUP

server...

UQSSP Disk Controller 0 (772150)

(c)

1988 Digital Equipment Corporation

show allclass

Parameter

Current

Default

ALLCLASS

PARAMS>

set

PARAMS>

show unitnum

Parameter

allclass

Type

Radix

Byte

Dec

Current =

Default

= Type
-

UNITNUM

PARAMS>

set

unitnum 20

PARAMS>

set

forceuni

PARAMS>

show allclass

Parameter

PARAMS>

Default
2

UNITNUM

Current

Default

PARAMS> show forceuni

Parameter -

Current

Dec

Type

Radix

Byte

Dec

Type

Radix

Word

Dec

Default
1l

Example A-10 Cont’d. on next page
\

A-18

Word

show unitnum

Parameter

FORCEUNI

Radix

Current

ALLCLASS

KA655 CPU System Maintenance

Type

Radix

Boolean

0/1

§§

Example Aw‘lo (Cont)
»

Display for Changing Allocafion Class and
Number

Unit

PARAMS> write

Changas require controller
initiallzatlon,
Stopping DUP
>>>

show

Controller 0

(RF71)

UQSSP

Disk Controller 1
(RF71)

(760334)

UQSSP Disk Controller 2
(RF71)

(760340)

UQSSP Disk Controller 3
-DUD3 (RF71)

(760344)

UQSSP Disk Controller 4
-DUE4 (RF71)

(760350)

UQSSP Disk Controller 5

(760354)

~DUC2

- UQSSP

-MUAO

(N)

]

(772150)

-DUAQ

-DUFS

[Y/

d&vmae

UQSSP Disk

-DUB1

ok?

server....

(RF71)

Tape Controllar 0
(TK70)

(774500)

Ethernet Adapter 0 (774440)
-XQA0 (08-00-2B-09-A3-96)

Configuring the KFQSA

A-19

Appendix B

KA655 CPU Address Assignments

| B.1 General Local Address Space Map
Table B-1:
Contents

VAX Memory Space
.

Address Range

Local memory space (64 Mbytes)

0000 0000-03FF FFFF

Reserved memory space (64 Mbytes)

0400 0000-07FF FFFF

Reserved memory space (64 Mbytes)

0800 0000-0BFF FFFF

Reserved memory space (64 Mbytes)

0C00 0000-OFFF FFFF

Cache diagnostic space (64 Mbytes)
.
Reserved cache diagnostic space (64 Mbytes)

1000 0000-13FF FFFF

Reserved cache diagnostic space (64 Mbytes)

1400 0000-17FF FFFF
1800 0000-1BFF FFFF

1C00 0000~-1FFF FFFF

KAB55 CPU Address Assignments

B-1

_ Table B-2: VAX Input/Output 'Spaca
- Address Range

Contents

2000 0000-2000 1FFF
2000 20002003 FFFF

_Local Q22-bus I/O space (8 Kbytes)

2004 0000-2005 FFFF

Local ROM space—halt protected space (128 Kbytes)

2006 0000-2007 FFFF

Local ROM space—halt unprotected space (128 Kbytes)

Reserved local I/O space (248 Kbytes)

2008 0000-201F FFFF

Local register I/O space (1.5 Mbytes)

2020 0000-23FF FFFF

Reserved local I/0 space (62.5 Mbytes)

2400 0000-27FF FFFF

2800 0000-2BFF FFFF

Reserved local J/O space (64 Mbytes)

2C08 0000-2FFF FFFF

Reserved local I/O space (64 Mbytes)
Reserved local I/O space (64 Mbytes)

3000 0000-303F FFFF

Local Q22-bus memory space (4 Mbytes)

3040 0000-33FF FFFF

Reserved local I/O space (60 Mbytes)

3400 0000-37FF FFFF

Reserved local I/O space (64 Mbytes)

3800 0000-3BFF FFFF

Cache tag diagnostic space (64 Mbytes)*
Reserved cache tag diagnostic space (64 Mbytes)

3C00 0000-3FFF FFFF

*Not visible during normal opémtion.

‘B.2 Detailed Local Address Space Map
Table B-3:

Detailed VAX Memory Space

Contents

Address Range

Local memory space (up to 64 Mbytes)

0000 0000—03FF FFFF

Q22-bus map-—top 32 Kbytes of main memory

Reserved memory space

0400 0000-OFFF FFFF

Cache diagnostic space

1000 0000-13FF FFFF

Reserved cache diagnostic space

1800 0000-1FFF FFFF

B-2

KA655 CPU System Maintenance

Table B-4:

Dmalmd VAX |nput/0utput Space

Contents

Address R.ange

Local Q22-bus VO Space

2000 0000-2000 1FFF

Reserved Q22-bus I/O space (diagnostics)

2000 0000-2000 0007
2000 0008-2000 O7FF

Q22-bus floating address space

Q22-bus fixed and user I/O space
Q22-bus fixed and reserved I/O space

2000 0800-2000 OFFF
2000 1000~-2000 1F3F

Interprocessor communication register (normal operation)
Interprocessor communication register (reserved)

2000 1F42

Interprocessor commtmication register (reserved)

2000 1F46

Interprocessor communication register (reserved)

2000 1F40
2000 1F44

Reserved Q22-bus I/O space

2000 1F48-2000 1FFF

Reserved Local VO Space

2000 2000-2008 FFFF

Local ROM Space
Local ROM protected space

2004 0000-2007 FFFF
2004 0000-2005 FFFF

MicroVAX system type register (in ROM)

2004 0004

Local ROM unprotected space

2006 0000-2007 FFFF

Local Register I/O Space

2008 0000-201F FFFF

DMA system configuration register

2008 0000

DMA system error register

2008 0004

DMA master error address register

2008 0008

DMA slave error address register

2008 000C

Q22-bus map base register
Reserved local register I/O space

2008 0010

Main memory configuration register 00

2008 0100

2008 0014-2008 OOFF

Main memory configuration register 01
Main memory configuration register 02

2008 0104

2008 0108

Main memory configuration register 03
Main memory configuration register 04

2008 010C

2008 0110

Main memory configuration register 05
Main memory configuration register 06

2008 0114

2008 0118

Main memory configuration register 07

2008 011C

Main memory configuration register 08
Main memory configuration register 09

2008 0120

2008 0124

Main memory configuration register 10

2008 0128

Main memory configuration register 11

‘Main memory configuration register 12

2008 012C
2008 0130

Main memory configuration register 13

2008 0134

Main memory configuration register 14

2008 0138

KA655 CPU Address Assignments

B-3

2%
!

Table B4 (Cont) Detalled VAX Input/Output Spaw
| Conwnw

Address Range

' Main memory configuration register 16

2008 013C

Main memory error status register

2008 0140

Main memory control/diagnostic status register

2008 0144

Reserved local register I/O space

2008 0148-2008 3FFF

Cache control register

2008 4000

Boot and diagnostic register

2008 4004

Reserved local register /O .space

2008 4008-2008 TFFF

Q22-bus map registers

2008 8000-2008 FFFF

Reserved local register I/O space

2009 0000-2013 FFFF

SSC base address register

2014 0000

SSC configuration register

2014 0010

CDAL bus timeout control register

2014 0020

Diagnostic LED register

2014 0030

Reserved local register I/0 space

2014 0034-2014 0068

The following addresses allow those KA655 IPRs that are implemented in the SSC to be

" accessed through the local I/O page. These addresses are documented for diagnostic purposes
only and should not be used by nondiagnostic programs.

Time-of-year register

2014 006C

Console storage receiver status

2014 0070*

Console storage receiver data

2014 0074*

Console storage transmitter status

2014 0078*

Console storage transmitter data

2014 007C*

Console receiver control/status

2014 0080

Console receiver data buffer

2014 0084

~ Console transmitter control/status

2014 0088

Console transmitter data buffer

2014 008C

Reserved local register I/O space

2014 0090-2014 00DB

* These registers are not fully implemented; accesses yield unpredictable results.
I/O bus reset register

Reserved local register 1/O space

2014 00DC

2014 00E0-2014 OOEF .

ROM data register

2014 OOFQ**

Bus timeout counter

2014 O0F4**

B-4

KA655 CPU System Maintenance

‘Table B4 (Cont) Detalwd VAX Input/Output Space
Contents

)

Address Ranga

‘ Interval timer

2014 O0F8**

Rem'md local register /O space

2014 00FC-2014 O0FF

s Theae registers are internal BSC registers used for SSC test purposes only. They should
not be accessed by the CPU.

Local Register 10O Space (Continued)

Timer 0 control register

2014 0100

Timer 0 interval register

2014 0104

Timer 0 next interval register

2014 0108

Timer 0 interrupt vector

2014 010C

Timer 1 control register

2014 0110

Timer 1 interval register

2014 0114

Timer 1 next interval register

2014 0118

Timer 1 interrupt vector

2014 011C

Reserved local register /O space

2014 0120-2014 012F

CACR address decode match register

”

2014 0130

CACR decode mask register
Reserved local register /O space

|
o

2014 0134
-

2014 0138-2014 013F

BDR address decode match register

2014 0140

BDR decode mask register

2014 0144

Reserved local register I/O space

2014 0148-2014 O3FF

Battery backed-up RAM

2014 0400-2014 O7FF

Reserved local register 1/O space

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 0000-37FF FFFF

Cache tag diagnostic space

3800 0000-3BFF FFFF t
3C00 QOOMFFF FFFF

Reserved cache tag diagnostic upace
1 Not visible during noxmal apemtion,

KA655 CPU Address Assignments

B-5

B.3 Internal Processor Registers

.
Each Internal Processor Register (IPR) falls into one of the catego

|
ries listed

below. You must use the MFPR and MTPR privil
eged instructions to access
the IPRs.

1. Implemented by KA655 (in the CVAX
chip) as specified in the VAX
Architecture Reference Manual.
2.

Implemented by KA655 (in the SSC) as specified
in the VAXArchitecture
Reference Manual.

Implemented bfir KA655 (and all designs that use

uniquely.

Not implemented, timed out by the CDAL Bus
4 usec. Read as zero, NOP on write.

the CVAX chip)

Timer (in the SSC) after

Access not allowed; accesses result in a reser

Accessible, but not fully implemented; access
es yield unpredictable

ved operand fault.

results.

Refer to Table B-5 for a listing of each of the
mnemonic, its access type (read or write),

KA655 IPRs, along with its

and its category number.

NOTE: An I following the category number in Table B-5 indica

processor is halted.

B-6

KA655 CPU System Maintenance

tes that the

the negation of DCOK when the

Table B-5:

KA655 Internal Procassor Raglsmrs

Decimal Hex

- Mnemonic

Kernel Stack Pointer

Supervisor Stack Pointer
User Stack Pointer

Interrupt Stack Pointer

7:5

Reserved

PO Base

Type Category

KSP

Executive Stack Pointer

)

r/w

ESP

r'w

SSp

r'w

USP

riw

ISP

r/w

1
5

POBR

PO Length

POLR

P1 Base

P1BR

riw

r/w

P1 Length

P1LR

System Base

riw

SBR

r/w

SLR

riw

System Length

15:14

Reserved

Process Control Block Base

PCBB

riw

System Control Block Base

SCBB

Interrupt Priority Level

riw

IPL

r/w

11
11

AST Level

ASTLVL

Software Interrupt Request

riw

SIRR

Software Intempt Summary

23:22

17:16

SISR

Reserved

Interval Clock CoanStatua

ICCS

Next Interval Count

riw

NICR

Interval Count

ICR

Time-of-Year Clock

TODR

Console Storage Receiver Status

riw

Console Storage Receiver Data

Console Storage Transmit Status
Console Storage Transmit Data

383

Console Receiver Data Buffer

Console Transmit Control/Status

Translation Buffer Disable

Cache Disable

CSRS

riw

CSRD

CSTS

r/w

CSTD

Console Receiver Control/Status =~ RXCS

Console Transmit Data Buffer

RXDB

riw

TXCS

riw

TXDB

TBDR

riw

CADR

riw

Machine Check Error Summary

MCESR

Memory System Error

r/iw

MSER

Reserved

r'w

SAVPC

SAVPSL

41:40

29:28

‘
Console Saved PC

Console Saved PSL

KA655 CPU Address Assignments

B-7

‘Table B-5 (Cont.): KAB55 Internal Processor Registers

. Decimal Hex

Mnemonic

Type Category
r/w

47:44

2F:2C Reserved

SBI Fault/Status

SBIFS

5
5

SBI Silo

SBIS

-50

SBI Silo Comparator

SBISC

riw

51.

SBIMT

r/w

SBI Error

SBIER

r/w

83
54

35
36

SBI Timeout Address
SBI Quadword Clear

SBITA
SBIQC

r
w

5
5
4

SBI Maintenance

I/O Bus Reset

IORESET

Memory Management Enable

MAPEN

TB Invalidate All

TBIA

TB Invalidate Single

TBIS

TB Data

TBDATA

riw

Microprogram Break

MBRK

riw

Performance Monitor Enable

PMR

riw

System Identification

SID

Translation Buffer Check

TBCHK

64:127

40:7F

Reserved

B-8

KA655 CPU System Maintenance

1
6

Q‘i

B.3.1 KA655 VAX Standard IPRs

* IPRs that are implemented as specified in the VAX Architecture Reference
-

Maniual are listed in Table B-6.

operation and use of these registers.

See that manual for details on the

Tai;le B-6: IPRs Implemented According to Standard VAX
Architecture

- Number Hex

Mnemonic

System Base Register

System Length Register

SBR

SLR

Process Control Block Base

System Control Block Base

SCBB

Interrupt Priority Level

IPL

Software Interrupt Request

Software Interrupt Summary

Time-of-Year Clock

Memory Management Enable

S8A

PCBB

SIRR
SISR

TODR

Translation Buffer Invalidate All

Translation Buffer Invalidate Single
System Identification
Translation Buffer Check“

MAPEN
TBIA

TBIS
SID

TBCHK

KA655 CPU Address Assignments

B-9

.. IPRs tHat are implemented on the KA655 but are not contain

"’

ed in, or do not

fully conform to the standards in the VAX Architecture Reference Manual

are listed in Table B-7.

;fi’

_Table B-7: KAG655 Unique IPRs
‘Number Hex

Mnemonic

Interval Clock Control/Status
Console Receiver Control/Status

RXCS

Console Receiver Data Buffer
Console Transmit Control/Status

Console Transmit Data Buffer

Cache Disable

Memory System Error

Console Saved PC

Console Saved PSL

/O Bus Reset

RXDB

TXCS

TXDB
CADR
MSER

SAVPC
SAVPSL

IORESET

B-10

ICCS

KA655 CPU System Maintenance

)

B.4 Global Q22-Bus Address Space Map

‘The addresses and memory‘contents of the Q22-bus
in Table B-8.

Table B-8:
Contents
.

memory space are listed

Q22-Bus Memory Space
|

’

Address Range

Q22-bus memory space (octal)

0000 00001777 7777

The contents and addresses of the Q22-b
us I/O space with BBS7 asserted
are listed in Table B-9.

Table B-9:
Contents

Q22-Bus /O Space with BBS7 Asserted
Address

Q22-bus YO space (Octal)

1776 0000-1777 7777

Reserved Q22-bus I/O space (diagnostics)
Q22-bus floating address space

1776 0000-1776 0007
1776 0010-1776 3777
1776 40001776 7777

Q22-bus fixed and user address space

Q22-bus fixed and reserved address space
- Interprocessor communication register (norma

l operation)

Interprocessor communication register (reserved)

Interprocessor communication register (reserv
ed)
Interprocessor communication register (reserv
ed)
Reserved Q22-bus I/O space

1777 0000-1777 7477
1777 7600

1777 7502
1777 7504

1777 7506
1777 75610-1777 7777

KA655 CPU Address Assignments

B-11

)

é‘i

Appendix C

Related Documentation
" The follofing documents contain information relating to the KA655 CPU.
Document Title °

Order Number

Modules

CXY08 Technical Manual
DEQNA Ethernet User’s Guide

EK-CXY08-TM
EK-DEQNA-UG

DPV11 Synchronous Controller Technical Manual

EK-DPV11-TM

DPV11 Synchronous Controller User’s Guide

EK-DPV11-UG

DRV11-WA General Purpose DMA User’s Guide

EK-DRVWA-UG

DZQ11 Asynchronous Multiplexer User’s Guide

EK-DZQ11-UG

DRV11-J Interface User’s Manual

EK~-DRV1J-UG

DZQ11 Asynchronous Multiplexer Technical Manua!

EK-DZQ11-TM

IEU11-A/TEQ11-A User's Guide

EK-IEUQI-UG

KDA50-Q CPU Module User’s Guide

EK-KDA5Q-UG

KFQSA Installation Guide

KMV11 Programmable Communications Controller User’s Guide

KMV11 Programmable Communications Controller Technical Manual
LSI-11 Analog System User’s Guide

EK-KFQSA-IN
EK-KMV11-UG

EK-KMV11-TM
EK-AXV11-UG

Q-Bus DMA Analog System User’s Guide
RQDX2 Controller Module User’s Guide

EK-AV11D-UG

RQDX3 Controller Module User’s Guide

EK-RQDX3-UG

EK-RQDX2-UG

Disk and Tape Drives
RF30 Integrated Storage Element
RF30 Integrated Storage Element Installation Manual

RF71 Integrated Storage Element User's Guide
TK50 Tape Drive Subsystem User’s Guide

EK-RF30D-UG
EK-RF30D-IN

EK-RF71D-UG

EK-LEP05-UG