Digital PDFs

EK-179AA-MG-001

October 1988

157 pages

Original

10MB

view download

OCR Version

8.1MB

view download

Document:	KA640 CPU System Maintenance
Order Number:	EK-179AA-MG
Revision:	001
Pages:	157
Original Filename:

OCR Text

KA640 CPU System Maintenance
Order Number EK—-179AA-MG-001

digital equipment corporation
maynard, massachusetts

October 1988

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 in
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.
© Digital Equipment Corporation. 1988. All rights reserved.
Printed in U.S.A.
The READER'S COMMENTS 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

MicroVMS

UNIBUS

DECnet

PDP

VAX

DECUS

P/OS

VAXBI

DECwriter
DELNI
DEQNA

Professional
Q-bus
Rainbow

VAXELN
VAXcluster
VAXstation

RSTS

VMS

RSX

MASSBUS

Work Processor

MicroPDP-11

ThinWire

DESTA
DIBOL

ULTRIX

dlilgliltlall |

ML-S970

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.

Contents
Preface

Chapter 1

KA640 CPU and Memory Subsystem

1.1

Introduction .. ...... .. ... . .. . .. . . ...

1-1

1.2

KA640 Features . ...... ... ... .. . . .. ... ...

1-3

CVAX Chip. ..ot
e
e ee

1-3

1.2.2

Clock Functions . ... ....... ...

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

1.2.3

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

1.2.4

Cache Memory .. ...........
.. ...

1-5

1.2.5

MemoryController..

1.2.6

MicroVAX System Support Functions

1.2.7

1.2.1

.. ... ...

... . ........... PR

1-5

.. ... e

1-5

Resident Firmware

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

1-6

1.2.8

Q22-bus Interface.

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

1-6

1.2.9

KA640 Network Interface

1210

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

1-7

KA640DSSI Interface ...... ... ... . ... . . ... . ... .

1-7

1.3

H3602-SAI/OPanel ....... . ... .. . ... ... ...... ...

MS650-AA Memory Module. . .. ......................

1-10

RF30DiskDrive .. ... ...

..., e

1-12

2.1

Introduction .. ....... ... . . . ...
.

2-1

2.2

General Module Order . ............... e

2-1

Module Order for KA640 Systems . .. ................

2-2

2.2.1

. .. ... .. ...

2.3

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

2-3

2.4

DSSIConfiguration ....... ... ......

241

Changingthe Node Name ...........
... ...........

2-5

242

Changingthe Unit Number . .. ...

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

2-7

Access to RF30 Firmware in VMS Through DUP .......

2-8

243

2.4.4

DSSICabling . .......

2441

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

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

2-9

2-11

2.4.5

Dual-Host Capability . .. ...... ...

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

2-11

24.6

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

2-12

Allocation Class ... .. ... ... ... ... ... . ... ......

2-13

2.4.6.1

2.4.6.2

2.5

Changing the KA640 Node ID .. ..................

2-13

Configuration Worksheet . ............
... ... .........

2-14

Chapter 3

KA640 Firmware

3.1

Introduction ................. e e

3-1

3.2

KA640 Firmware Features. ..........................

3-1

3.3

Halt Entryand Dispatch Code .. .. ............. P

3-2

ExtermalHalts . ... ... .. ... .. .. .. ... ... ..

3-3

Power-UpSequence ..............
..t iuenunnn..

3.5.0.1

Mode Switch SettoTest. . . .. ...

... ...

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

34
34

3.5.0.2

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

3-5

3.5.0.3

Mode Switch SettoNormal ... ... ... ...

.. .......

3—-6

Bootstrap ............
.. ...
...

3—-6

3.7

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

3-8

Locatingthe RPB . . . ... ... ... ... ... .. . ... ......

3-9

3.7.0.1

Console 'OMode . ....... .. .. . ... . . . . . @i,

3-10

3.8.1

Command Syntax.................... e

3-10

3.8.2

Address Specifiers ........ e

3-11

3.8.3

Symbolic Addresses .. ........ ...

e e e e
... ... ... .. .....

3-11

3.8.4

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

3-14

3.8.5

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

3-16

ConsoleCommands ..............
.. .00,

3-18

39
3.9.1

3.9.1.1
392

BOOT . . .

3-18

Supported Boot Devices . .. .. .o v o
CONFIGURE . .... ...

...

3-19

.. ..

3-22

3.9.3

CONTINUE . ...

3-24

3.94

DEPOSIT . ... e

3-25

3.9.5

EXAMINE . ... ...

3-26

396

FIND . ...
.

3-28

3.9.7

HALT . . e

3-29

398

HELP..... ...

. . i,

3-30

3.9.9

INITIALIZE . . . ... . e

3-32

3.9.10

MOVE . . ... e

3-33

3.9.11

NEXT . ... . . e

3-35

3.9.12

REPEAT . . ... . . .. . . e,

3-37

39.13

SEARCH ....... ... ... .. . . . .. ... e

3-38

3914

SET .................. e

e e

3—40

3.9.15

SHOW .. . e

343

3.9.16

START .. ... .. e

347

3.9.17

TEST . ... . .

348

3.9.18

UNJAM ... . .. e

349

3.9.19

X—BinaryLoadand Unload .......................

3-50

3.9.20

!'—Comment ...........
... .. . . ...
..

3-52

Chapter 4

e e e

Troubleshooting and Diagnostics

4.1

Introduction . ..........
... . ...

4-1

4.2

General Procedures . ... .. ...

4-1

4.3

KA640 ROM-Based Diagnostics . ......................

4-2

43.1

DuagnosticTests...........
... ... ... . .. .. ... ...

4-3

4.3.2

Seripts ... ... .. e
e e e
e

4-6

4.3.3

Script Calling Sequence . .. ........................

4-8

434

UserCreated Scripts . ..........
... .. . 0. iii....

4-10

4.3.5

Console Displays ...... .. ...

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

4-14

4.3.6

System Halt Messages . ............. e

4-22

...

4.3.7

Console Error Messages . . ...........
... ..........

4-23

43.8

VMBErrorMessages . .......... ...,

4-24

Acceptance Testing . .........
... ........... e

4-24

4.5

Troubleshooting ......... ... ... .. .. .. .. ... . ... .. ....

4-31

451

FEUtility. . ...

... ..

4-31

4.5.2

Isolating Memory Failures . .. ......................

4-34

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

4-37

Loopback Tests...........
.. .. .. . . ...

4-38

4.5.3

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

4-39

4.6.1

ModuleSelf-Tests.........

Testingthe Console Port ...... .. ...

.. . . . .. ...

4-40

4.8

RF30 Troubleshooting and Diagnostics .................

441

...

481

DRVIST. ... ... . .e

4-43

482

DRVEXR . ... ... ... . ...

443

483

HISTRY ...

... e
e

4-45

484

ERASE ... ... . . ..

446

485

PARAMS . ... ...

4-47

4.8.5.1

EXIT ..
e
e e e

448

4.8.5.2

HELP . ... ... e

448

4.8.5.3

SET .
e

4-48

4854

SHOW . .. e

4-49

4.8.5.5

STATUS .. ...

... ... ...... e ...

449

4.8.5.6

WRITE . . ... e

4-49

DiagnosticErrorCodes .........
... ... ... .. ..
.....

4-50

4.9

Appendix A

Address Assignments

General Local Address Space Map ... ..................

A-1

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

A-2

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

A—-6

Global Q22-Bus Address Space Map ...................

A-8

Appendix B

Related Documentation

Index

Examples
2-1

Changinga DSSINodeName ........................

2-6

2-2

Changinga DSSIUnit Number .. .....................

2-7

3-1

Language SelectionMenu ..........c.cceeuueeeee....

3-6

4-1

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

4-12

4-2

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

4-13

4-3

Console Display(NoErrors). ........
... ...,

4-14

4-4

SampleOutputwithErrors .........
... ... .. ..........

4-14

4-5

FE Utility Example .. .........

4-31

.. ...

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

4-6

Isolating Bad Memory Using T 9C ... ... .. e

4-36

4~7

9C—Conditions for Determining a Memory

FRU . . . ...
.. ..

4-37

Figures

1-1

KA640CPUModule. . ... ... ... . ...

1-2

H3602-SAI/OPanel ........ ... . . . .. ...

1-9

1-3

MS650-AA Memory Module. . . ..... ... . ... ... ... ...

1-11

2~-1

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

2-10

2-2

RF30O0CP . ... . e

2-11

2-3

BA213 Configuration Worksheet ......................
KA640CPUModule LEDs . .. ... ... ... ... .. ..

2-16
4-17

4-1

Tables

1-1

RF30 Specifications . .................. J

1-12

2-1

DSSIDisk Drive Order .. ... ... ... . . . ...

2-2

RF30 DIP Switch Settings . ... ..........
... ... ...

2-5

2-3

Changing the KA640 Node ID ... ... ... ... ... .. .. ....

2-14

2—4

Power and Bus Loads for KA640 Options .. .............

2-15

3-1

Actions TakenonaHalt ..... ... ... .. ... ... ... ...

3-3

3-2

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

3-5

3-3

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

...

3-12

3—4

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

3-14

3-5

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

3-15

3-6

Command Keywords by Type . .. ... .. ... ... .. ... .....

3-16

3-7

Console Command Summary .. ...........c..cueenm....

3-16

3-8

VMBBootFlags........

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

3-19

3-9

Boot Devices Supported by the KA640-AA . ... ..........

3-20

4-1

Testand Utility Numbers ... ... ... . ... . ... ... ... ...

4-2

Scripts Available to Field Service.

4-7

.. ... ...

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

4-3

Commonly Used Field Service Scripts . .. ... ............

4-8

4—4

Values Saved, Machine Check Exception During Executive ..

4-16

' 4-5

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

4-16

4-6

KA640 Console Displaysand FRUs .. ..... ...

.....

4-18

4-7

System Halt Messages .. .........
.. ... .. ... . ... ...

4-22

4-8
4-9

Console Error MeSsages . . . . . .o oo, 4-23
VMB Error MeSSages . . - -« o oo vv e e 4-24

...

vii

4-10
4-11

Hardware Error Summary Register . . . .................
KAB40 FUuses . ... .ciii ittt i e et e it iieennns

4-12 Loopback Connectors for Q22-Bus Devices. ... ...........
4-13 DRVEXR MeSSages . .. ...vummeeieeeaeiaannaanns

4-14 DRVEXR Messages . .............. e
4-15 HISTRY Messages . .. .............. e
4-16 ERASE MeSSages . . . .o oottt it ieeeiaaeaaannns
4-17 RF30 Diagnostic Error Codes . . . . .. .. ... ... ...,

VAX Memory Space . ... ......c..oouumnnnnnannnenens

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

VAX Memory Space .. ... ....cuitmeantiinennnnnnnn.
VAX Input/Output Space ....... e [

e e e
RKAGAD IPRS . . . oot
.
... .....
.......
...
.
Space
Memory
Q22-Bus
Q22-Bus I/0O Space with BBS7 Asserted ................

Viii

4-33

4-38

4-41

4-43
4-44
4-45
447
4-50
A-1

A-2
A-2

A-3
A-6
A-8

A-8

Preface
This guide describes the base system, configuration, ROM-based
diagnostics, and troubleshooting procedures for systems containing the
KA640 CPU.

Intended Audience
This guide is intended for use by DIGITAL Field Service personnel and
qualified self-maintenance customers.

‘Organization
This guide has four chapters and two appendixes, as follows:

Chapter 1 describes the KA640/MS650 CPU and memory subsystem, and

the RF30 disk drive.

Chapter 2 contains system configuration guidelines, and provides a table
listing current, power, and bus loads for supported options. It also describes
the DIGITAL Small Storage Interconnect (DSSI) bus interface cabling
between the CPU, the CPU I/O panel, the operator console panel (OCP),
and the RF30 dlSk drives.
Chapter 3 describes the firmware that resides in ROM on the KA640, and
provides a list of console error messages and their meaning.

Chapter 4 describes the KA640 diagnostics, including an error message and
FRU cross-reference table. It also describes diagnostics that reside on the
RF30.
Appendix A lists the KA640 address space.
Appendix B is a list of related documentation. It contains the order numbers
for all manuals mentioned in this manual.

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

They have

WARNING

Provides information to prevent personal injury.

CAUTION

Provides information to prevent damage to equipment or software.

NOTE

Provides general information about the current topic.

Chapter

KA640 CPU and Memory Subsystem
1.1 Introduction
This chapter describes the KA640 CPU (Figure 1-1). The KA640 is
a quad- helght VAX processor module for the Q22-bus (extended LSI-11
bus). It is designed for use in high-speed, real-time applications and
for multiuser, mu]tztaskmg environments. The KA640 employs a cache
memory to maximize performance.

There are two variants: the KA640—AA which runs multiuser software;
and the KA640-BA, which runs single-user software.

The KA640 is used in two systems, the MicroVAX 3300 and the MicroVAX
3400.
The MicroVAX 3300 is housed in a BA215 enclosure.
The
MicroVAX 3400 is housed in a BA213 enclosure. Refer to BA215 Enclosure
Maintenance and BA213 Enclosure Maintenance for a detailed description
of each enclosure.

CAUTION: Static electricity can damage integrated circuits. Always use a
grounded wrist strap (part no. 29-11762-00) and grounded work surface
when working with the internal parts of a computer system.

The KA640 CPU module and MS650 memory modules combine to form a

VAX CPU and memory subsystem that uses the Q22-bus to communicate

with I/O devices. The KA640 and MS650 modules mount in standard Q22bus backplane slots that implement the Q22-busin the AB rows and the CD
interconnect in the CD rows. The KA640 can support up to three MS650
modules, if enough Q22/CD slots are available.

The KA640 communicates with the console device through the H3602-SA
CPU I/O panel, which also contains configuration switches and an LED
display. The H3602-SA is described in Section 1.3.

KA640 CPU and Memory Subsystem

1-1

The KA640-AA module number (M7624) stamped on the handle varies
slightly, depending on the vendor used for the RAM chips:
M7624-AL

KA640-AA, Hitachi chips

M7624-AF

KA640-AA, Toshiba chips

M7624-BF

KA640-BA, Toshiba chips

Figure 1-1:

KA640 CPU Module

M7624-BL

KA640-BA, Hitachi chips

F3 (FUSE)

50-PIN DSSI

(+5 V DSSI

(TOP)

S0-PIN MEM

(BOTTOM)

TERMINATION)

40-PIN CONNECTOR

_gps

TO H3602-SA

DCOK

ALY

[
*

n_—

F2 (FUSE)

sovmm

DSSI

F1 (FUSE)

(+5 V TO

REMOTE

wane o

PANEL)

(+12 V TO ETHERNET)

LANCE

LOW

l '

HIGH

]

4 MB MEM

38 1MX12ZIPS

)
40

CMCTL

SSC

CCLK

CFU

CFPA

[

|
casic

[

N
MLO-0D0 1280

i-2

KAB40 CPU System Maintenance

1.2 KA640 Features
The major features of the KA640 CPU are listed below.

The VAX central processor, which is-implemented in a single VLSI chip
called the CVAX. It achieves a 100 nanosecond (ns) microcycle and a
200 ns bus cycle at an operating frequency of 20 megahertz (MHz).
It 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 4-Mbyte, 400 ns, 39 bit-wide array (32-bit data and 7-bit ECC)
implemented with 1 Mbit dynamic RAMs in zig-zag in-line packages

(ZIPs).
*

A console port compatible with the VAX processor whose baud rate can
be set through an external switch on the H3602-SA.

A set of processor clock registers that support:
—

A VAX standard time-of-year (TOY) clock with support for battery
backup. (Batteries are located in the H3602-SA..)

—

An interval 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 on the
H3602-SA.

128 Kbytes of 16 bit-wide ROM.

A Q22-bus interface.

A DSSI bus interface.
*

An Ethernet interface.

1.2.1 CVAX Chip
The CVAX chip contains all general purpose registers (GPRs) visible to the
VAX processor, several system registers such as MSER, CADR, SCBB, the
cache memory (1 Kbyte), and all memory management hardware, including
a 28-entry translation buffer.
'

KA640 CPU and Memory Subsystem

1-3

The CVAX chip supports the MicroVAX chip subset of the VAX instruction
set and data types, pius the following string instructions:

CMPC3

The CVAX chip provides the following subset of the VAX data types:
Byte
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 reset signal for the CPU, the floating point accelerator,
and the main memory controller

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

1.2.3 Floating Point Accelerator
The floating point accelerator is implemented by a chip called the CFPA.
The CFPA chip contains approximately 60,000 transistors in a 68-pin
CERQUAD surface mount package.
It executes the VAX f , d_, and
g_floating point instructions (except for CLRx, MOVx, and TSTx), and
accelerates the execution of MULL, DIVL, and EMUL integer instructions.
The CFPA 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 chip.

1-4

KA640 CPU System Maintenance

1.2.4 Cache Memory
The KA640 module incorporates a cache memory to maximize CPU
performance The cacheis implemented within the CVAX chip. The cache
is a 1-Kbyte, two-way associative, wnte-through cache memory, with a 100
nanosecond (ns) cycle time.

1.2.5 Memory Controller
- The main memory controller is implemented by a VLSI chip called the
CMCTL. The CMCTL contains approximately 25,000 transistors in a 132pin CERQUAD surface mount package. It supports ECC (error correction
code) memory, with a 400 ns cycle time for longword read transfers and
a 600 ns cycle time for quadword transfers. It has 200 ns cycle time for
unmasked longword writes and a 500 ns cycle time for masked longword
writes.

The maximum amount of main memory supported by KA640 systems is 28
Mbytes. This memory resides on the KA640 module (4 Mbytes) and on one
to three MS650-AA 8-Mbyte memory modules, depending on the system
configuration. The MS650 modules communicate with the KA640 through
the MS650 memory interconnect, which utilizes the CD interconnect and 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 millisecond (ms) interrupts

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

backup
*

An IORESET register

KA640 CPU and Memory Subsystem

1-5

Programmable CDAL bus timeout

Two programmable timers

A register for controlling the diagnostic LEDs

1.2.7 Resident Firmware
The resident firmware consists of 128 Kbytes of 16 bit-wide ROM, located
on two 27512 EPROMs. The firmware gains control when the processor
halts, and contains programs that provide the following services:

Board initialization

Power-up self-testing of the KA640 and MS650 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 firmware is described in detail in Chapter 3.

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. It 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. It has a 500 ns cycle time
for longword read transfers and an 800 ns cycle time for quadword read
transfers. It has a 400 ns cycle time for unmasked longword writes and
a 600 ns cycle time for masked longword writes. The Q22-bus interface
contains the following:

A 16-entry map cache for the 8,192-entry, scatter/gather map that
residesin main memory, used for translating 22-bit Q22-bus addresses
into 26-bit main memory addresses

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

The Q22-bus interface handles programmed and power-up resets, and CPU
halts (deassertion of DCOK).
The KA640-AA module contains 240 ohm termination for the Q22-bus.

1-6

KA640 CPU System Maintenance

1.2.9 KA640 Network Interface
The KA640 features an on-board network interface implemented through
a LANCE chip, a 32K x 8 bit-wide static ROM, and two 32K x 8 bit
static RAMs. This interface allows the KA640 to be connected to either
a ThinWire or standard Ethernet cable, through the H3602—SA /O panel.
The network interface includes four registers for control and status
reporting, a 24-word transmit silo, and a 24-word receive silo. It also
includes a word-wide 64-Kbyte buffer (two 32K x 8 static RAM chips). The

DMA controller reads control information and writes status information
to and from main memory. It also transfers data (one word per memory
reference) between main memory and either the transmit
or receive silo.
The DMA controller can perform up to eight masked longword references
before giving up the CDAL bus. Each reference takes 600 ns and contains
either a byte or word of data. The minimum time between bus requests is
8 usec.

1.2.10 KA640 DSSI Interface
The KA640 contains an SII chip, a DXX chip, and four 32K x 8-bit static
RAMs that implement the DIGITAL Small Storage Interconnect (DSSI) bus
interface. The DSSI interface allows the KA640 to transmit packets of data
to, and receive packets of data from, up to seven other DSSI devices (RFseries disk drives or a second KA640 module). The DSSI bus improves
system performance for two reasons:
e

It is faster than the Q22-bus.

It relieves the Q22-bus of disk traffic, allowing more bandwidth for
Q22-bus devices.

The physical characteristics of the DSSI bus are as follows:
4 Mbytes per second bandwidth
Distributed arbitration
Synchronous operation
Parity checking
|
Six meter total bus length (includes internal and external cabling)
Single-ended bus transceivers
Maximum of eight nodes (KA640 counts as one)
Eight data lines
One parity line
Eight control lines

KA640 CPU and Memory Subsystem

1-7

Refer to the following sections for more information about the DSSI bus
and disk drives:
Section 2.4

Setting and changing DSSI node names, addresses and unit numbers, dual
"~

host configuration rules.

Section 3.9.14

Console SET HOST command.

Section 4.4

DSSI drive acceptance testing.

Section 4.8

RF30 drive resident diagnostics and local programs.

1.3 H3602-SA I/0 Panel
The H3602-SA (Figure 1-2) contains the console serial line connector,
console baud rate switch, two Ethernet connectors and LEDs, hex LED
display, and power-up mode switch. The switches are read by the firmware
when the processor halts. For this reason changing the baud rate on the
H3602-SA does not take effect until the next power-up or system reset. (By
contrast, on the KA630, the switches are hardwired into the hardware.)
The switches are also read when the power-up mode switch is in the

test position. The H3602-SA has the following switches, connectors, and
indicators:

Baud rate select switch.

Power-up mode select switch.

Halt enable/disable switch from the console keyboard (grReéak] key or
[€7auP), depending on the state of SSCCR <15>. Break is the default.
If this switch is set to the enable position, the system does not autoboot
on power-up. It enters console I/0 mode and displays the >>> prompt.

Ethernet connector select.
The H3602-SA has two connectors for
Ethernet cable: a 15-conductor connector for standard Ethernet cable,

and a male BNC connector for a ThinWire Ethernet coaxial cable. The
H3602—-SA contains a switch to select the Ethernet connector, and LEDs
to indicate the selected connector and valid +12 vdc for that connector.
e

1-8

Hex LED display, which provides a countdown of the system power-up
self tests. See Table 4—6 for the meaning of this display.

KA640 CPU System Maintenance

Figure 1-2:

H3602-SA I/O Panel

FRONT

REAR

BAUD RATE

v
HEX DISPLAY

)
|
BREAK
ENABLE

ceLECT

i
POWER-UP

MODE

cONSOLE

/ CONNECTOR

=
A

ETHERNET

CONNECTOR
SELECT

CABLE TO

-~
=

MLC-001283

KA640 CPU and Memory Subsystem

1-9

1.4 MS650-AA Memory Module
The MS650-AA memory module is a quad-height, Q22-bus module
(Figure 1-3). The MS650-AA is an 8-Mbyte, 400 ns, 39 bit-wide array

(32-bit data and 7-bit ECC) implemented with 256-Kbyte dynamic RAMs

in zig-zag in-line packages (ZIPs).

The KA640 and MS650-AA memory modules are connected through the CD
rows of backplane slots 1 through 4, and through a 50-conductor cable. The
part number of this cable varies depending on the number of connectors,
as follows:

Number of

Connectors

CPU/Memory

Configuration

Part Number

KA640 + 2 MS650-AA modules

17-01898-01

KA640 + 3 MS650-AA modules

17-01898-02!

KA640 + 3 MS650-AA modules

17-01898-03

l1Recommended cable. Use five-connector cable only if this cable is not available.

The cable is keyed so that it is installed in the correct connector on the
KA640 (the connector next to the module). The DSSI cable is attached to
the connector “piggy backed” to the memory connector.

1-10

KA640 CPU System Maintenance

Figure 1-3: MS650-AA Memory Module

]

KA640 CPU and Memory Subsystem

1-11

1.5 RF30 Disk Drive
The RF30 is a half-height, 13.3-cm (5.25-in) fixed-disk drive for BA200series enclosures. Table 1-1 lists the specifications for the RF30 drive.

Table 1-1:

RF30 Specifications

Specifications
Average seek time

22 milliseconds

Average rotational latency

8.33 milliseconds

Average access time

30.33 milliseconds

Peak transfer rate

12 Mbits/second

User capacity

150 Mbytes

User capacity (blocks)
Width

293,040
14.60 cm (5.75 in)

Depth

20.45 cm (8.25 1n)

Height

4.40 cm (1.75 in)

Form factor

Standard 5.25-in footprint

Power requirements

+5Vde, 1.10 A
+12 Vde, 0.80 A

Power consumption

15.1 W

VMS support

Version 5.0-2A and later

ULTRIX-32 support

VAXELN support

Version 3.0 and later

Version 3.2 and later

MicroVAX Diagnostic Monitor

Revision 2.3 and later

support

The RF30 disk drive is based on the DIGITAL Small Storage Interconnect
(DSSI) architecture.

DSSI supports up to seven storage devices, daisy-

chained to the host system through the KA640 CPU or a host adapter
module.
The disk drive controller is built into the RF30 drive, rather than being a
separate module. This feature enables many drive functions to be handled
without host-system or adapter intervention, resulting in improved /O
performance and throughput rates.

DSSI node ID switches are located on the electronics controller module.
Set these switches to assign a unique node ID number to each drive on the
DSSI bus. Refer to Table 2-2 for the correct DIP switch settings.
The RF30 disk drive contains a Ready indicator and a fault indicator.

The Ready indicator displays the activity status of the drive. It lights
on power-up. After successful completion of the power-up diagnostics, the

1-12

KA640 CPU System Maintenance

indicator goes out, until the media heads are on the requested cylinder and
the drive is read/write ready.

The Fault indicator lights at power-up. After successful completion of the
power-up diagnostics, this indicator goes out. If the Fault indicator lights

again after going out, a read/write safety error or a drive error condition
has occurred.

KA640 CPU and Memory Subsystem

1-13

Chapter 2

Configuration

2.1 Introduction
This chapter describes the guidelines for changing the configuration of a
KA640 system, and for configuring a multihost 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/0 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 accessing the bus

Configuration

2-1

2.2.1 Module Order for KA640 Systems
Observe the following rules about module order:
e

Install the KA640 CPU in slot 1.

Install MS650 memory modules in slots 2, 3, and 4.

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.
Here is the recommended module order in a KA640 system:

KA640
MS650
AAV11-SA
ADVI11-SA

AXV11-SA
KWV11-SA
TSV05-SA
DELQA-SA
DPV11-SA
KMV1A-SA, -SB, -SC
DFAO1
CXYO08-AA
CXB16-M
CXAl16-M
LPV11-SA

DRV1IW-5A
IEQ11-SA
ADQ32-M
DRQ3B-SA
IBQO1-SA
KLESI-SA
TQK50-SA
TQK70-SA
M9060-YA

2-2

KA640 CPU System Maintenance

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
d2pends 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 /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:
>>> config

Enter

device

Device,

configuration,

HELP,

or EXIT

Number? help

Devices:

LPV11

KXJ11

DLV11J

D2zQ1l1

DZV11l

RLV21

TSV05

RXV21

DRV11W

DRV11B

DpPV1l

DMV11

DELQA

DEQNA

RQDX3

KDASO

RRDS0

RV20

RQC25

DFA01

KXOOXX-DISK

TQKSO

TQK70

TUS1E

KOOXX-TAPE

EKMV1l

IEQ11

DHQ11

DHV11

CXAl6

CXB16

cXYO08

VCBO2

QDSS

DRV11J

DRQ3B

vsv2l

IBQO1

IDVilia

IAV11A

IAV11B

MIRA

IDV11B

"ADQ32

IDV11C

IDV11D

DTCO04

DESQA

IGQ11

See the description of the CONFIGURE command in Chapter 3
(Section 3.9.2) for an example of obtaining the correct CSR addresses

and interrupt vectors using this command.

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.
2.

See Microsystems Options for switch and CSR and interrupt vector
jumper settings for supported options.

Configuration

2-3

2.4 DSSI Configuration
Each device must have a unique DIGITAL Small Storage Interconnect
(DSSI) node ID. The RF30 receives its node ID from a plug on the operator
control panel (OCP) on the front panel. By convention, DSSI drives are
mounted in the BA213 or BA215 enclosures from right to left, as listed in

Table 2-1.

Table 2-1:
Device

DSSI Disk Drive Order
Position

Node ID!

BA213 enclosure
First

Right side

Second

Center
Left side

Third

BA215 enclosure?
First

Right side

Second

Center

1KA640 node ID = 7
2BA215 OCP has three drive plugs. but only two drives. The third plug is blank.

If the cable between the RF30 and the OCP is disconnected, the RF30 reads
the node ID from three DIP switches on its electronics controller module

(ECM).

NOTE: Pressing the system reset button on the front of a BA213 or BA215
power supply has no effect on the RF30 drives. You must perform a power
cycle.

The node ID switches are located behind the 50-pin connector on the ECM.
Switch 1 (the MSB) is nearest to the connector. Switch 3 (the LSB) is
farthest from the connector. Refer to the RF30 section in Microsystems
Options for an illustration and further information. Table 2-2 lists the
switch settings for the eight possible node addresses.

2-4

KA640 CPU System Maintenance

Table 2-2:

RF30 DIP Switch Settings

Node ID

Down

Up
Down

Down

The VMS operating system creates DSSI disk device names according to
the following scheme:
nodename $ DIA unit number. For example, SUSANSDIA3

You can use the device name for booting, as follows:
>>> BOOT

SUSANSDIA3

You can access local programs in the RF30 through the MicroVAX
Diagnostic Monitor (MDM), or through the VMS operating system (version
5.0) and console YO mode SET HOST/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 dialog. Section 2.4.3 describes the procedure for accessing DUP
through the VMS operating system. Section 3.9.14 describes the console
/O mode SET HOST/DUP command.
,

2.4.1 Changihg the Node Name
Each RF30 drive 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 drive 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 RF30
drive at DSSI node address 1 is changed from R3YBNE to DATADISK.
See Section 4.8.5 for further information about the PARAMS local program.

Configuration

2-5

Example 2—-1: Changing a DSSI Node Name
>>> sho dssi

(MDC)

DSSI Node 0
-DIAO

(RF30)

DSSI Node 1 (R3YBNE)
-DIAl (RF30)

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

(*)

DSSI Node 7
>>>

>>> set host/dup/dssi 1
server.

Starting DUP

1988

DRVEXR V1.0
DRVTST V1.0
HISTRY V1.0
ERASE
V1.0
PARAMS V1.0
DIRECT V1.0

D
D
D
D
D
D

ngztal Equipment Corporation
5-NOV-1988 15:33:06
5-NOV-1988 15:33:06
5-NOV-1988 15:33:06
5-NOV-1988 15:33:06
5-NOV-1988 15:33:06
5-NOV-1988 15:33:06

End of directory
Task Name? params

1988

Digital Equipment Corporation

PARAMS> sho nodename
Parameter

Current

NODENAME

Default

R3YBNE

Type

RF30

String

Radix
Ascii

PARAMS> set nodename datadisk
PARAMS> write

!This command writes the change
'to EEPROM.

Changes require controller initialization,
Stopping DUP

ok?

[Y/(N)]

server...

>>> sho dssa

DSSI Node 0
-DIAC

DSSI Node 1
-DIAl

(MDC)

(RF30)

(DATADISK)

(RF30)

!'The node name has changed from
-~

IR3YBNE to DATADISK.

DSSI Node 7 (*)

2-6

KA640 CPU System Maintenance

2.4.2 Changing the Unit Number
By default, the RF30 disk drive assigns the disk’s unit number to the same
value as the DSSI node address for that drive. This occurs whether the
DSSI node address is determined from the OCP unit ID plugs or from the
three DIP switches on the RF30 controller module.
RF30 drives conform to the DIGITAL Storage Architecture (DSA). Each
drive can be assigned a unit number from 0 to 16,383 (decimal). The unit
number does not have to 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 RF30 drive at DSSI node address
2 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 further information about the PARAMS local program.
Example 2-2:

Changing a DSSI Unit Number

>>> sho dssi

DSSI Node 0

-DIAQ0

(MDC)

(RF30)

DSSI Node 1

-DIAl1

(R3QJNE)

(RF30)

DSSI Node

'The unit number for this drive will be
'changed from 1 to 50 (DIAl to DIASO0).

(*)

>>>

>>> set host/dup/dssi 1
Starting DUP

serxver...

1988

Digital Equipment Corporation

DRVEXR V1.0

5-NOV-15988

15: 33 :06

DRVIST V1.0

5-NOV-1988

15: 33 :06

BISTRY V1.0

5-NOV-1988

15: 33 :06

ERASE

V1.0

5-NOV-1988

15: 33 :06

PARAMS V1.0

5-NOV-1988 15: 33 :06

DIRECT V1.0

5-NOV-1988 15: 33: 06

End of directory

Example 2-2 Cont’d. on next page

Configuration

2-7

Example 2-2 (Cont.):

Changing a DSSI Unit Number

Task Name? params

Digital Equipment Corporation

PARAMS> sho unitnum
Parameter

Current

Default

Type

Radix

Default

Type

Radix

PARAMS> sho forceuni
Parameter

Current

FORCEUNI

Boolean

0/1

PARAMS> set unitnum 50
PARAMS> set

forceuni

PARAMS> write

'This command writes the changes to EEPROM.

PARAMS> ex

..
Exiting.
Task Name?

- Stopping DUP

sexver...

>>>

>>>sho dssi

DSSI Node
-DIAO

(MDC)

(RF30)

DSSI Node 1
-DIAS0O

(R3QJNE)

(RF30)

DSSI Node 7

'The unit number has changed
tand the node ID remains at

(*)

2.4.3 Access to RF30 Firmware in VMS Through DUP
You can also access the RF30 firmware utilities from the VMS operating
- system as well as through the console commands described
in Section 4.8.

NOTE: Access the RF30 firmware through the VMS operating system to look
up or to view parameter settings, but not to change them. To change RF30

parameter settings, enter the RF30 firmware through the console 1/0O mode
SET HOST /DUP command.

2-8

KA640 CPU System Maintenance

Load the FYDRIVER using the following commands in SYSGEN:
$§ MCR SYSGEN
SYSGEN> LOAD FYDRIVER/NOADAPTER

SYSGEN> CONNECT FYAO/NOADAPTER
SYSGEN> EXIT

You can then access the RF30 firmware utilities using the following VMS
command:
$ SET HOST/DUP/SERVER=MSCPS$DUP/TASK=PARAMS nodename

2.4.4 DSSI Cabling
A 50-conductor ribbon cable connects the RF30 drive to the DSSI bus
(Figure 2-1). A separate 5-conductor cable carries +5 Vdc and +12 Vdc
to the drive from the enclosure power supply.
A 2-conductor cable connects the fifth pin on the RF30 power connector
to the operator control panel (OCP, Figure 2-2). In the BA213 enclosure,
one of these (two) cables is for an RF30 connected to the right side power
supply, and the other is for an RF30 connected to the left side supply.

These cables carry the ACOK signal (same as POK) to the RF30. The OCP
delays this signal to one RF30 for each power supply to stagger the start-up
of one of two possible devices attached to each supply. This delay prevents
excessive current draw at power-up.

The BA215 enclosure has only one

power supply, but implements this signal delay in the same way.
The 50-conductor DSSI ribbon cable connects to a 50-conductor round cable
that is routed through the bottom of the mass storage area to the DSSI
connector on the KA640.

CAUTION: When removing or installing new drives, be sure to connect the
rightmost connector of the DSSI ribbon cable to the round cable connected

to the KA640. Do not “T” the bus by connecting the round connector to any

of the ribbon cable’s center connectors.

Configuration

2-9

Figure 2-1:

DSSI Cabling, BA213 Enclosure

DSS! 8US

TERMINATION

ERMINATION

16 Re3

*ACOK’ SIGNAL,

|l POWER.UP

QUfEt

SUPPLY TO

|| STAGGERS RF30

FROM POWER

)

FROM POWER

SUPPLY TO

FROM

i
o

%{
!

_SUPPLY

>—"TORF30

BACKPLANE ]

CABLE
KAB40 DSS!I

CONNECTOR
MLO -00:282

2-10

KAB40 CPU System Maintenance

Figure 2-2:

RF30 OCP

FRONT

TO POK LEAD
POWER

RESERVED FOR

SUPPLY\F
UTURE USE

10-PIN

10-PIN TO

:{%%\&!b &M—J}- DRIVE-SELECT
L

TO RF1

::]

|
7

Lt =1C

17l

TO RF3

PLUGS

5] @ QQ

BACKPLANE

AAE

DRIVE FAULTS

/(RED)-

wRITE-PROTECT

_— BUTTONS

. . n‘—/ READY BUTTONS

]

1__2_1

= "'5“""‘
RESTART/RUN

|~ DCOK (GREEN)

had
CPU HALT

MLO-00128

2.4.4.1 DSSI Bus Termination and Length

The DSSI bus must be terminated at both ends.
The KA640 module
terminates the DSSI bus at one end. A 50-conductor Honda connector on
the left side of the media faceplate terminates the bus at the other end.
This connector can be removed if you need to expand the bus.
The DSSI bus has a maximum length of 6 m (19.8 ft), including internal
and external cabling.

In a dual-host system,
termination.

the second

KA640 module

provides the bus

2.4.5 Dual-Host Capability
A DSSI disk drive such as the RF30 has a multihost capability built into the
firmware, which allows the drive to maintain connections with more than
one DSSI adapter. Since the KA640 CPU has a built-in DSSI adapter, more

than one KA640 CPU can be connected to the same DSSI bus, allowing each
KA640 to access all other drives on the bus.

The primary application for such a configuration is a VAXcluster system
using Ethernet as the interconnect medium between the boot and the

Configuration

2-11

satellite members.
described below.

This configuration improves system availability, as

Two KA640 systems are connected through an external DSSI cable
(BC21M). Each KA640 system is a boot member for a number of satellite
nodes. The system disk resides in the first enclosure, and serves as the
system disk for both KA640 systems. The KA640 in each enclosure has
equal access to the system disk, and to any other DSSI disk in either
enclosure.

If one of the KA640 modules fails, all satellite nodes booted through that
KA640 module lose connections to the system disk. However, the multihost
capability enables each satellite node to know that the system disk is still
available through a different path—that of the remaining good KA640
module. A connection through that KA640 is then established, and the
satellite nodes are able to continue operation.

Thus, even if one KA640 module fails, the satellites booted through it
are able to continue operation. The entire cluster will run in a degraded
condition, since one KA640 is now serving the satellite nodes of both
KA640s. 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.
Disk failure can be caused by such factors as a power supply failure in
the enclosure containing the disk.

DSSI cabling failure. If a failure in one of the DSSI cables renders
“access to the disks 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 Dual-Host Configuration

Dual-host systems have the followihg configuration limitations:
e

A maximum of two systems can be connected, because of cabling and
enclosure limitations.

The DSSI bus supports eight devices or adapters. Since a dual-host
system has two KA640 modules, and each has a connection to the DSSI

2-12

KA640 CPU System Maintenance

bus, a maximum of six DSSI devices can be attached to the bus. Two
variants are possible:

—

Two BA213 enclosures, containing two KA640 CPUs, and six DSSI
devices (three in each enclosure). This configuration uses all eight
possible DSSI devices.

—

Two BA215 enclosures, containing two KA640 CPUs, and four DSSI
devices (two in each enclosure). This configuration uses six of eight
possible DSSI devices.

Set DSSI node IDs as follows:
—

The first (or only) KA640 is 7.

—

The second KA640 in a dual-host system is 6.
explains how to change the KA640 node ID.

—

The remaining devices in a dual-host system are 0-5.

Section 2.4.6.2

2.4.6.1 Allocation Class

When a KA640 system containing RF-series drives is configured in a cluster,
either as a boot node or a satellite node, you must assign the allocation class
in VMS SYSGEN and for the RF-series drive to matching nonzero values.
To change the allocation class of the RF-series drive, use the following
commands:
>>>

SET

HOST/DUP/DSSI

Starting DUP
PARAMS>

<DSSI

node number> PARAMS

server..

SET ALLCLASS

<allocation

class

value>

PARAMS> WRITE

Changes

require

Stopping DUP

controller initialization,

ok?

[Y/N]

server..

>>>

2.4.6.2 Changing the KA640 Node ID

The KA640 node address is configured by three jumpers. Table 2-3 lists
the jumper positions and node IDs. Figure 1-1 shows the location of the
jumpers.

Configuration

2-13

Table 2-3:

Changing the KA640 Node ID

Node ID

Out

Out
Out

Out

Out
Out

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

For the BA215 enclosure, use the top half of the BA213 enclosure
worksheet, and allow for two disk drives instead of one.

Table 2—4 lists power values for supported devices.
configuration, follow these steps:

To check a system

List all the devices to be installed in the system.

Fill in the information from Table 2—4 for each device.

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 (M9060YA) 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 an
error mode and shuts down the system.

2-14

KAB40 CPU System Maintenance

Table 2—4:

Power and Bus Loads for KA640 Options
Current

(Amps)

Option

Module

+5V

+12V

Power

Bus Loads

Watts

DC
0.5

AAV11-SA

A1009-PA

1.8

0.0

9.0

2.1

ADV11-SA

A1008-PA

3.2

0.0

16.0

2.3

0.5

AXV11-SA

A026—PA

2.0

0.0

10.0

1.2

0.3

CXAl6-AA/~-AF

M3118-YA

1.6

0.20

10.4

3.0

0.5

CXBl16-AA/-AF

M3118-YB

2.0

0.0

10.0

3.0

0.5

CXY08-AA/-AF

M3119-YA

1.64

0.395

12.94

3.0

0.5

DELQA-SA

M7516-PA

2.7

0.5

19.5

2.2

0.5

DFAQ1-AA/-AF

M3121-PA

1.97

0.40

14.7

3.0

1.0

DPV11-SA

M8020-PA

1.2

0.30

9.6

1.0

DRQ3B-SA

M7658-PA

4.5

0.0

22.5

2.0

1.0

DRV1J-SA

M804S-PA

1.8

0.0

9.0

2.0

1.0

DRV1IW-SA

M7651-PA

1.8

0.0

9.0

2.0

1.0

DSV11-SA

M3108-PA

543

0.69

38.0

3.6

1.0

DZQ11-SA

M3106-PA

1.0

0.36

9.3

0.5

IBQO1-SA

M3125-PA

5.0

0.0

25.0

4.6

IEQ11-SA

M8634-PA

3.5

0.0

175

2.0

1.0

KA640-AA

%117624-—AA/— 6.0

32.88

3.5

1.0

0.24

KLESI-SA

M7740-PA

3.0

0.0

150

2.3

KMV1A-SA

M7500-PA

2.6

0.2

154

3.0

1.0

KWV1i-SA

M4002-PA

2.2

0.13

11.15

1.0

0.3
0.5

LPV1i-SA

M8086-PA

1.6

0.0

8.0

1.8

M9Y060-YA

5.3

0.0

26.5

0.0

MS650-AA

M7621-A

2.7

0.0

13.5

0.0

0.0
-

RF30

1.10

0.80

15.1

TK50E-EA

1.35

2.4

35.6

TK70E-EA

1.5

2.4

36.3

TQKS0

M7546

0.0

14.5

2.8

0.5

TQK70-SA

M7559

3.5

0.0

17.5

4.3

0.5

TSV05-SA

M7196

6.5

0.0

32.5

3.0

1.0

Configuration

2-15

Figure 2-3:

BA213 Configuration Worksheet

RIGHT POWER SUPPLY

SLOT

MODULE

Current {Amps)

Power

+5 Vdc

+12 Vdc

(Watts)

33.0 A

7.6 A

2300w

Current (Amps)

Power

Nibjwin

6
MASS STORAGE:
TK Drive

FIXED DISK
To1tal these columns:
Must not exceed:

LEFT POWER SUPPLY
SLOT

MODULE

+5 Vdc

+12 Vdc

{Watts)

33.0A

7.6 A

2300 W

8
S
10
11

MASS STORAGE:
FIXED DISK(S)

Tota! these columns:
Must not exceed:

MLO-00128S

2-16

KA640 CPU System Maintenance

Chapter 3

KA640 Firmware
3.1 Introduction
This chapter describes the KA640 firmware, which gains control of the
processor whenever the KA640 performs a processor halt. A processor halt
transfers control to the firmware. The processor does not actually stop
executing instructions.

3.2 KA640 Firmware Features
The firmware is located in two 64-Kbyte EPROMS on the KA640. The
firmware address range is 20040000 to 2007FFFF, inclusive (20040000—
2005FFFF in halt-protected space and 20060000—2007FFFF in haltunprotected space) in the KA640 local I/O space. The firmware displays
diagnostic progress and error reports on the KA640 LEDs and on the console
terminal. It provides the following features:

°*

Automatic or manual restart or bootstrap of customer application
images at power-up, reset, or conditionally after processor halts.

(Restartin this context is not the same as restarting or resetting the

hardware.)
*

Automatic or manual

bootstrap of an

operating system

following

processor halts.
*

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 as the system console.

Multilingual support.
several languages.

The firmware can issue system messages in

The processor must be functioning.at a level able to execute instructions
from the console program ROM for the console program to operate.

KAB640 Firmware

3-1

The firmware consists of the following major functional areas:
Halt entry and dispatch code
Bootstrap
Console I/O mode
Diagnostics
The halt entry and dispatch code, bootstrap, and console I/O mode are
described in this chapter. Diagnostics are described in Chapter 4.

3.3 Halt Entry and Dispatch Code
The processor enters the halt entry code at physical address 20040000
whenever a halt occurs. The halt entry code saves 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, it occurs 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

General purpose registers

PRS$_SAVPSL

Saved processor status longword register

PRS_SCBB

System control block base register

DLEDR

Diagnostic LED register

SSCCR

SSC configuration register

ADxMAT

SSC address match register
SSC address mask register

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.
SSCR

SSC configuration register

ADxMAT

SSC address match register

ADxMSK

SSC address mask register

CBTCR

TIVRx

CDAL bus timeout control register

‘

SSC timer interrupt vector registers

The console command interpreter does not modify actual processor
registers. Instead it saves the processor registers in console memory when
it enters the halt entry code, then directs all references to the processor
registers to the corresponding saved values, not to the registers themselves.
When the processor reenters program mode, the saved registers are
restored and any changes become operative only then.
References to

3-2

KA640 CPU System Maintenance

processor memory are handled normally. 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
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 H3602—SA panel. Table 3-1 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.
Actions Taken on a Hait

Breaks Enabled

on H3602-SA

Power-up Halt! Halt Action?
B
B
O BRE B

O B B

W OO
M

Table 3—-1:

Action
Diagnostics, halt
Halt

Diagnostics, bootstrap, halt

Restart, bootstrap, halt
Restart, halt

Bootstrap, halt
Halt

1Power-up halt: PR$_SAVPSL<13:08>=3

Halt action: PR$_CPMBX<01:00>
3T = condition js true, F = condition is false, X = does not matter

3.4 External Halts
Several conditions can trigger an external halt, and different actions are
taken depending on the condition. The conditions are listed below.
e

The break enable switch is set to enable, and you press (8REak] on the
system console terminal.

Assertion of the BHALT line on the Q-bus.

Deassertion of DCOK. A halt is delivered if the processor is not running

out of halt-protected space, and the BHALT ENB bit is set. The system
restart switch deasserts DCOK. DCOK may also be deasserted by the
- DELQA sanity timer, or any other Q22-bus module that chooses to
implement the Q22-bus restart/reboot protocol.

The KA640 cannot detect the deassertion of DCOK when in console /O
mode, so no action is taken. More important, however, the deassertion of
DCOK destroys system state without notifying the firmware.

KAB40 Firmware

3-3

CAUTION: Do not press the Restart button while in console I/0O mode.
Doing so will destroy system state without notifying the firmware.

The action taken by the halt dispatch code on a console [8Reak] or Q22-bus
BHALT is the same: the firmware enters console I/O mode if halts are
enabled.
The halt dispatch code distinguishes between DCOK deasserted and
BHALT by assuming that BHALT must be asserted for at least 10 msec,
and that DCOK is 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. The firmware
keeps a halt-in-progress flag to tell if it is halting because of stepping out
into halt-unprotected space. This flag is cleared on power-up.

3.5 Power-Up Sequence
On power-up, the firmware performs several unique actions.
It runs
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 H3602-SA (Figure 1-2). The mode switch has three settings: test,
language inquiry, and normal. The differences are described in Sections
3.5.0.1 through 3.5.0.3.
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 (battery is discharged), then the
IPT tests and initializes the NVRAM.
After the battery check, the firmware tries to determine the type of terminal
attached to the console serial line. If the terminal is a known type, it is
treated as the system console.

3.5.0.1 Mode Switch Set to Test
Use the test position on the H3602—-SA to verify that the connection between
the KA640 and the console terminalis good.

To test the console terminal, insert the H3103 loopback connector into
the H3602—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.

KA640 CPU System Maintenance

During the test, the firmware toggles between the active and passive states.
During the active state (3 seconds), the LED is set to 6. The firmware reads
the baud rate and mode switch, then transmits and receives a character
sequence.

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

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 2 normal power-up.
3.5.0.2 Mode Switch Set to Language Inquiry

If the H3602-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.

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

Table 3-2: Language lnquiry on Power-Up or Reset
Mode

Language Not

Previously Set!

Language

Previously Set

Language Inquiry

Prompt?

Prompt

Normal

Prompt

No Prompt

1 Action if contents of NVRAM invalid same as Language Not Previously Set.

2«Prompt” = Language selection menu displayed.

The language selection menu is shown in Example 3-1. If no response 1is
received within 30 seconds, the firmware defaults to English.

KA640 Firmware

3-5

Example 3-1: Language Selection Menu
1)

Dansk

2) Deutsch

(Deutschland/Osterreich)

Deutsch

(Schweiz)

English

(United Kingdom)

English

(United States/Canada)

Espanol

Francais

(Canada)

'8)

Frangais

(France/Belgique)

Francais

(Suisse)

10)

Italiano

11)

Nederlands

12)

Norsk

13)

Portugues

14)

Suomi

15)

Svenska

(1..15):

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

See

After the language inquiry, the firmware continues as if on a normal powerup.

3.5.0.3 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 KA640 supports bootstrap of VAX/VMS, ULTRIX-32, VAXELN, and
MDM diagnostics.

The firmware initializes the system to a known state before dispatching to
the primary bootstrap (VMB), as follows:

Checks CPMBX<2>(BIP), bootstrap in progress. If it is set, bootstrap

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

3-6

in the selected

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

KA640 CPU System Maintenance

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 ESAQ.

o
o

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.

Searchés 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

Length of PFN bitmap in bytes

Address of PFN bitmap

Time-of-day of bootstrap from PR$_TODR

Boot flags

R10

Halt PC value

R11

Halt PSL value (without halt code and mapenable)

Halt code

Base of 128-Kbyte good memory block + 512

R1, R6, R7,R8,

R9, FP

10.

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

base of the 128 Kbytes of good memory block + 512.

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

Virtual

Memory Bootstrap (VMB) is the

primary bootstrap for VAX

processors. The KA640 VMB resides in the firmware, and is copied into
main memory before control is transferred to it. VMB then loads the
secondary bootstrap image and transfers control to it.

KAB40 Firmware

3-7

3.7 Operating System Restart
An operating system restart is the process of bringing up the operating
system from a known initialization state following a processor halt.
A 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 a 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.
The firmware restarts the operating system in the following sequence:
Checks CPMBX<3>(RIP). If it is set, restart fails.
2.

Prints the message Restarting system software.
|
terminal.

on the console

Sets CPMBX<3>(RIP).
Searches for a valid RPB. If none 1s found, restart fails.

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 Failure. on the console terminal.

3-8

KA640 CPU System Maintenance

3.7.0.1 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 address of the restart routine
+08 (third longword)
= checksumpf first 31 longwords of restart routine
The firmware finds a valid RPB as follows:

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

Reads the second longword in the page (the physical address of the
restart routine). If it is not a valid physical address, or if it is zero,
returns to step 1. The check for zero is necessary to ensure that a page

of zeros does not pass the test for a valid RPB.

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 been found.

KAB40 Firmware

3-9

3.8 Console I/0 Mode
In console I/0 mode several characters have special meaning:
Also <CR>. The carriage return ends a command line. No action is taken on a
command until after it is terminated by a carriage return. A null line terminated
by a carriage return is treated as a valid, null command. No action is taken, and
the console prompts for input. Carriage return is echoed as carriage return, line
feed (<CR><LF>).
When you press the
key, the console deletes the previously typed
character. The resulting display differs, depending on whether the console is
a video or a hard-copy terminal.

For hard-copy terminals, the console echoes a backslash (\), followed by the
character being deleted. 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, surrounding them with backslashes. For example:
EXAMI:E[RUBOUT | RUBOUT INE<CR>
The console echoes: EXAMI;:E\ E;\NE<CR>
The console sees the cornmand 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 line is ignored.

Echoes ~U<CR>, and deletes the entire line. Entered but otherwise ignored if
typed on an empty line.

Stops output to the console terminal until

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.

[CTRUC]

Echoes ~"C<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 [CTRUC] is entered.

Echoes 2O when disabling output, not echoed when it reenables output. Output is
reenabled if the console prints an error message. or if it prompts for a command
from the terminal. Output is also enabled by entering console /O mode, by

pressing the

key, and by pressing [CTRUC].

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

3-10

KAB640 CPU System Maintenance

You can abbreviate a command by entering only as many characters as are
required to make the command unique. Most commands can be recognized
from their first character. See Table 3—6.

The console treats two or more consecutive spaces and tabs as a single
space. Leading and trailing spaces and tabs are ignored. You can place
command qualifiers after the command keyword or after any symbol or
number in the command.

All numbers (addresses, data, counts) are hexadecimal (hex), but symbolic
register names contain decimal register numbers. The hex digits are 0
through 9 and A through F. You can use uppercase and lowercase letters in
hex numbers (A through F) and commands.
The following are qualifier and argument conventions:
{]

an optional gqualifier or argument

a required qualifier or argument

3.8.2 Address Specifiers
Several commands take an address or addresses as arguments. An address
defines the address space, and the offset into that space. The console
supports six address spaces:

Physical memory
Virtual memory
Protected memory

General purpose registers (GPR)
Internal processor registers (IPR)
|
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 addresses. A symbolic reference
defines the address space, and the offset into that space. Table 3-3 lists
symbolic references supported by the console, grouped according to address
space. You do not have to use an address space qualifier when using a
svmbolic address.

KA640 Firmware

3-11

Table 3-3:
Symbol

Consoie Symbolic Addresses
Address

Symbol

Address

GPR Address Space (/G)
RO

R10

R11

R12

R13

R14

R15

1)3)

PSL

IPR Address Space (/I)
prd_ksp

pr$_esp

prd_ssp

prd_usp

prd_isp

pr$_pObr

prs_pOlr

pré_plbr

pr$_plir

pr$_sbr

0oC

pr$_sir

pr$_pcbb

pr$_scbb

pr$_ipl

pr$_astlv

prd_sirr

prd_sisr

prd_icer

pr$_nicr

prd_icr

prd_todr

pr$_rxcs

pr$_rxdb

- prd_txcs

pr$_txdb

pr$_tbdr

pr$_cadr

pr3_mcesr

pr$_mser

prd_savpc

2A
37

pr$_savpsl

prd_ioreset

pr$_mapen

pr$_tbia

prd_tbis

prd_sid

pr$_tbchk

3-12

KA640 CPU System Maintenance

Table 3-3 (Cont.):
Symbol

Console Symbolic Addresses

Address

Symbol

Address

Physical Memory (/P)
gbio

20000000

gbmem

30000000

gbmbr

20080010

rom

20040000

cacr

20084000

bdr

20084004

dscr

20080000

dser

20080004

2008000C

dmear

20080008

dsear

ipcr0

20001£40

ipcrl

2000142

ipcr2

2000144

ipcr3

20001{46

- ssc_ram

- 20140400

ssc_cr

20140010

ssc_cdal

20140020

ssc_dledr

20140030

ssc_adOmat

20140130

ssc_adOmsk

20140134

ssc_adlmat

20140140

ssc_adlmsk

20140144

ssc_ter0

20140100

ssc_tir0

20140104

ssc_tnir0

20140108

ssc_tivr0

2014010c¢

ssc_terl

20140110

ssc_tirl

20140114

ssc_tnirl

20140118

ssc_tivrl

2014011c

memesr0

20080100

memcesrl

20080104

memcsr2

20080108

memecesr3

2008010c¢

memcsr4

20080110

mermcsrd

20080114

2008011c
20080124

memcsro

0080118

memcesr7

memcsr8

20080120

memecsr9

memcsrl0

20080128

memcesrill

2008012c¢

memecsrl2

20080130

memecsrl3

20080134

memcsrl4

20080138

memecesrls

2008013c¢

memcsrl6

20080140

memcsrl7

20080144

nisarom

20084200

nirdp

20084400

nirap

20084404

nibuf

20120000

msi_sbb

20084600

msi_scl

20084604

msi_sc2

20084608

msi_csr

2008460C

msi_id

20084610

msi_slesr

20084614

msi_destat

20084618

msi1_dstmo

2008461C

msi_data

20084620

msi_dmectrl

20084624

msi_cmlote

20084628

msi_dmaddrl

2008462C

msi_dmaddrh

20084630

msi_dmabyte

20084634

msi_stlp

20084638

msi_ltlp

2008463C

msi_ilp

20084640

msi_dsctrf

20084644

msi_cstat

20084648

msi_dstat

2008464C

msi_comm

20084650

msi_dictrl]

20084654

KA640 Firmware

3-13

Table 3-3 (Cont.):

Console Symbolic Addresses

Address

Symbol

Address

Physical Memory (/P)
msi_clock

20084658

msi_bhdiag

2008465C

msi_sidiag

20084660

msi_dmdiag

20084664

msi_mcdiag

20084668

msi_ram

20100000

Table 3—4 lists symbolic addresses that can be used in any address space.

Table 3—4:
Symbol
x

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
bvte, 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 3-5 lists and
describes the data control and address space control qualifiers. Command
specific qualifiers are described in the command descriptions.

3-14

KA640 CPU System Maintenance

Table 3-5:

Console Command Qualifiers

Qualifier

Description

Data Control
/B

The data size is byte.

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, use the complement. On reads, ignore ECC errors.

Address Space Control
/G

General purpose register
always longword.

Internal processor register (IPR) address space. Accessible only by the MTPR
and MFPR instructions. The data size is always longword.

(GPR) address space, RO-R15.

The data size is

2 9

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
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.

The data size is always

Access to console private memory is allowed. 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.

KAB640 Firmware

3-15

3.8.5 Console Command Keywords
Table 3-6 lists command keywords by type. Table 3-7 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.
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 a new command or
parameter is added in an updated version of the firmware. For this reason
you should not use abbreviations in programs.

Table 3-6:

Command Keywords by Type

Processor Control

Data Transfer

Console Control

BOOT

EXAMINE

CONFIGURE

CONTINUE

DEPOSIT

FIND

HALT

MOVE

REPEAT

INITIALIZE

SET

SHOW

START

TEST

UNJAM

Table 3—-7:

Console Command Summary

Command

Qualifiers

BOOT

/R5:{bitmap} /{bitmap!

{device_name]

CONFIGURE

CONTINUE

DEPOSIT

/B/W/LIQ

{address}

{data} {data]

faddress]

IGANP MU
/N:{count} /STEP:{size}

- Argument

Other(s)
-

/WRONG

EXAMINE

/B/W /L /Q

IGANPM/U

/N:{count} /STEP:{size}
/WRONG/INSTRUCTION

FIND

/MEM /RPB

HALT

HELP

INITIALIZE

3-16

KA640 CPU System Maintenance

Table 3-7 (Cont.):
Command
MOVE

Console Command Summary

Qualifiers

Argument

Other(s)

/B/W/L/Q

{src_address|

{dest_address!}

N P /G

/N:{count} /STEP:{size}
/WRONG

[count]

REPEAT

{command}
{start_address|

{pattern} [mask]

B/W/ML/Q
N /P /O

/N:{count} /STEP:{size}
/WRONG/NOT

SET BFLAG

{bitmap}

SET BOOT

f{device_string}

/DUP {/DSSI n /TUQSSP}

{node} n

{task]

SET HOST

{/DISK n

/TAPE n csr_address}
/MAINTENANCE /UQSSP

{controller_number}

{/SERVICE n csr_address)

SET LANGUAGE

{language_type}

SHOW BFLAG

SHOW BOOT

SHOW DEVICE

SHOW DSSI

SHOW ETHERNET -

SHOW LANGUAGE -

SHOW MEMORY

/FULL

SHOW QBUS

SHOW RLV12

SHOW UQSSP

SHOW VERSION

START

{address}

TEST

UNJAM

{test_number}
-

[test_argument]
-

{address}

{count!}

KA640 Firmware

3-17

3.9 Console Commands
This section describes the console I’/0 mode commands.
commands at the console I/O mode prompt >>>.

Enter the

3.9.1 BOOT
The BOOT command initializes the processor and transfers execution to
VMB. VMB attempts to boot the operating system from the specified device,
or 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, then 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 on-board Ethernet
port, ESAQ.
Qualifiers:

Command specific:
/R5:{bitmap!

A 32-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 command to display the longword. Table 3-8 lists the
supported R5 boot flags.
/{bitmap}

Same as /R5:{bitmap}

[device_name]

A character string of up to 39 characters. Longer strings cause a VAL TOO
BIG error message. Apart from length, the console makes no attempt to
interpret or validate the device name. The console converts the string to
uppercase, then passes VMB 2 string descriptor to this device name in RO.
Table 3-9 lists the boot devices supported by the KA640-AA.

3-18

KAB640 CPU System Maintenance

Table 3-8: VMB Boot Flags
Bit

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 RPBS$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 over
the network is |SYS0.SYSMAINTIDIAGBOOT.EXE.

RPB$V_BOOBPT

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

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.

-RPB$V_SOLICT

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_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 TOPSYSis 1, the top-level
directory name is [SYS1...1.

3.9.1.1 Supported Boot Devices

Table 3-9 lists the boot devices supported by the KA640-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 through
Z) in length, followed by a single character controller letter (A through Z),
and ending in a device unit number (0-16,383).

DSSI devices names may also include a node prefix, consisting of either
a node number (0-7) or a node name (a string of up to eight characters),
ending in a dollar sign ($).

KAB40 Firmware 3-19

Table 3-9:

Boot Devices Supported by the KA640-AA

Boot Name Controller Type

Device Type(s)

{node$IDIAn
DUcn

On-board DSSI
RQDX3 MSCP

RF30, RF71
RD52, RD53, RD54, RX33, RX50

KDA50 MSCP

RA70, RA80, RA81, RA82, RA90

Disk

DLen

KLESI

RC25

RLV21

RLO1, RLO2

Tape
[node$IMIcn

On-board DSSI

TF70

MUen

TQK50 MSCP

TK50

TQK70 MSCP

TK70

KLESI

TUS1E

ESAQ

On-board Ethernet

XQcn

DEQNA

DELQA

MRV11

Network

PROM

PRAO

3-20

KAB640 CPU System Maintenance

Examples:
>>>

show boot

o)
>>> show bflag

ESAQ
>>> b

Boot using default boot

(BOOT/R5:0 ESAQ)

flags

and device.

2..
-ESAQ0

>>> b xqal

! Boot

from XQAO0 using default boot flags.

(BOOT/R5:0 XQAO0)

2..
-XQA0

>>> b/10

! Boot using supplied boot flag (4)

(BOOT/R5:10 ESAQ0)

and default device.

2..
-ESA0

>>> boot /xr5:220 xgal

Boot using supplied boot

(BOOT/R5:220 XQAO0)

(5 and 9)

flags

and device.

2..
-XQA0

KA640 Firmware

3-21

3.9.2 CONFIGURE
The CONFIGURE command invokes an interactive mode that permits
you to enter Q22-bus device names, then generates a table of Q22-bus
/O page device CSR addresses and interrupt vectors. CONFIGURE is
similar to the VMS SYSGEN CONFIG utility. This command simplifies
field coniiguration 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:

Enter CONFIGURE at the console I/O prompt.

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
assignments.

obtain

the

CSR

address

and

interrupt

wvector

The devices listed in the HELP display are not necessarily supported by
the KA640-AA CPU.

Format:

CONFIGURE

3-22

KA640 CPU System Maintenance

Example:
>>>

configure

Enter device configuration,

HELP,

or EXIT

Device, Number? help
Devices:

LpPV1l
RLV21
DMV11
RQC25
KOOXX-TAPE
CXBl6 QPSS

KXJ11
TSVOS
DELQA
KXXXX~-DISK
KMV1l
CXYO08
DSV11

DLV11lJ
RXV21
DEQNA
TQOKS50
IEQ11
VCBO1
ADV11C

DZQ11
DRV11W
RODX3
TOK70
DHQ11
QVsSS
AAV11C

DZV1l
DRV11B
KDASO
TUS1E
DHV1l
ILNV11
AXV11C

DFAO1l
DPV11
RRD50
RV20
CXAl6
ILNV21
KWVv11iC

ADV11D

AAV11D

vCBO02

QDSS

DRV11J

DRQO3B

vsv2l

IBQO1

IDV11Aa

IDV11B

IDV1icC

IDV11D

IAV11A

IAV11E

MIRA

ADQ32

DTCO4

DESNA

IGR11
Numbers:

1 to 255,

default

Pevice, Number?

rgdx3,
2

Device, Number?

dhvll

Device, Number?

qgdss

Device, Number? tgqk50
Device, Number? tqgk70
Device, Number?

exit

Address/Vector Assignments
-772150/154 RQDX3

~-760334/300 RQDX3
-774500/260 TQKSO0
-760444/304

TQK70

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

KA640 Firmware

3-23

3.9.3 CONTINUE
The CONTINUE command causes the processor to begin instruction
execution at the address currently contained in the PC. It does not perform
a processor initialization. The console enters program YO mode.
Format:

CONTINUE

Example:
>>>

3-24

continue

KA640 CPU System Maintenance

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, /Q, /N:{count}, /STEP:{size}, WRONG

Address space control: /G, /1, /P, /V, /U
Arguments:
{address}

A longword address that specifies the first location into which data is deposited.

{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 address can be an actual address or a symbolic address.

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 0 0
>>> D/V/L/N:3

Deposit

at virtual memory address

>>> D/N:8 RO FFFFFFFF

Loads

>>> D/N:200

Starting at previous address,

bytes.

1234

! Clear first 512 bytes of physical memory.
5

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

GPRs

into four longwords starting
RO

Deposit

the first

memory.

1234.

through R8 with

in the first

-1.

clear 513

longword of

17 pages in physical

KAB40 Firmware 3-25

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, /P, /V, /U

Command specific:
/AINSTRUCTION
-

Disassembles and displays the VAX Macro-32 instruction at the specified
address.

Arguments:
|address]

A longword address that specifies the first location to be examined. The
address can be an actual or a symbolic address. If no address is specified,
+ 1s assumed.

3-26

KA640 CPU System Maintenance

Examples:
>>>

G
>>>

G 0000000E
M 00000000

00006’6’)83

>>>

00000000

Wy
o

>>>

041F0000

Examine the SP.

Examine the PSL.

! Examine PSL another way.
!

Examine R4 through R9.

ex pr$ scbb

Examine the SCBB,

00000011 2004A000

(decimal).

e r4/n:5
00000004

00000000

00000005

00000000

00000006

00000000

00000007

00000000

00000008

00000000

00000009 801D9000

e/p O

11 BRB

ex /ins/n:5 20040019

Examine local memory O.

Examine

1lst byte of ROM.

20040019

Disassemble from branch.

I~#20140000, £#20140000

20040019

DO MOVL

20040024

D2 MCOML

@#20140030,8#20140502

2004002F

D2 MCOML

S~“#0E, @#20140030
RO, @#201404B2

20040036

7D MOVQ

2004003D

DO MOVL

I*$201404B2,R1

20040044

DB MEFPR

S~#2A,B~44 (R1)

DB MFPR

S~*42B,B"48
(R1)

e/ins
20040048

IPR 17

00000000

ex /ins 20040000
20040000

>>>

041F0000

e/m

00000000
>>>

Examine the PC.

00000200

>>> ex psl
>>>

0000000F FFFEFFFFC

Look at next instruction.

>>>

KAB40 Firmware

3~27

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

Format:
FIND [qualifier-list]

Qualifiers:

Command specific:
/MEMORY

Searches memory for a page-aligned block of good memory, 128 Kbytes in length.
The search looks only at memory that is deemed usable by the bitmap. This
command leaves the contents of memory unchanged.

/RPB

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.

Examples:
>>> ex

Check the

SP.

0000000E 00000000

>>> find /mem

Look

for a wvalid 128 Kbyte.

>>>

Note

where

G 0000000E 00000200

>>> f£ind /xpb

?22C FND ERR 00C00004

was

found.

Check for valid RPB.

None to be

>>>

3-28

KA640 CPU System Maintenance

found here.

3.9.7 HALT
The HALT command has no effect.
other VAX consoles.

It is included for compatibility with

Format:

HALT
Example:
>>>

halt

Pretend to halt.

>>>

KA640 Firmware

3-29

3.9.8 HELP
The HELP command provides information about command syntax and
usage.

Format:

HELP

Example:
>>> help

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

UPPERCASE

denotes a keyword that you must type 1in

denotes

an OR condition

[]

denotes

optional parameters

< >

denotes a field that must be filled in
with a syntactically correct vaiue

Valid qualifiers:

/B /W /L /Q /INSTRUCTION

/G /I /V /P /M
/STEP: /N: /NOT
/WRONG /U

Valid commands:

DEPOSIT

[qualifiers]

EXAMINE

[qualifiers]

[address]

MOVE

[qualifiers]

[datum

[datum]]

[qualifiers]

[mask]

SET BFLAG <BOOT_ FLAGS>

SET BOOT <BOOT_DEVICE>
SET HOST/DUP/DSSI <NODE_ NUMBER>

SET BOST/DUP/UQSSP </DISK

[task]

/TAPE> <CONTROLLER NUMBER>

SET HOST/DUP/UQSSP <PHYSICAL
CSR _ADDRESS> [task]
SET HOST/MAINTENANCE/UQSSP/SERVICE <CONTROLLER NUMBER>

[task]
([task]

SET HOST/MAINTENANCE/UQSSP <PHYSICAL_CSR;ADDRESS> [task]
SET LANGUAGE <LANGUAGE NUMBER>

3-30

KA640 CPU System Maintenance

SHOW BFLAG
SHOW BOOT

SHOW DEVICE
SHOW DSS1I

SHOW ETHERNET
SHOW LANGUAGE

SHOW MEMORY

[/FULL]

SHOW QBUS

SHOW RLV12
SHOW UQSSP
SHOW VERSION
HALT
INITIALIZE

UNJAM
CONTINUE

START <ADDRESS>
REPEAT

X <ADDRESS> <COUNT>
FIND

[/MEMORY or /RPB]

TEST

[test _code

BOOT

[/R5:<BOOT_FLAGS> or /<BOOT_FLAGS>]

[count]

[parameters]]

[boot_device]

CONF IGURE
HELP
>>>

KA640 Firmware

3-31

3.9.9 INITIALIZE

The INITIALIZE command performs a processor initialization.
Format:

INITIALIZE
The following registers are initialized:
Register

State at Initialization

PSL

041F0000

IPL

ASTLVL

SISR

ICCS

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

RXCS

TXCS

MAPEN

CVAX cache

Disabled, all entries invalid

Instruction buffer

Unaffected

Console previous reference

Longword, physical, address O

‘TODR

Unaffected

Main memory

Unaffected

General registers

Unaffected

Halt code

Unaffected

Bootstrap-in-progress flag

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

>>>

3-32

KA640 CPU System Maintenance

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 actually performs byte, word, longword, and
quadword reads and writes as needed in the process of moving the data.
Moves are supported only for the physical and virtual address spaces.
Format:

MOVE [qualifier-list] {src_address} {dest_address}
Qualifiers:

Data control: /B, /W, /L, /W, /N:{count}, /STEP:{size}, /WRONG

Address space control: v, /4, P
Arguments:

A longword address that specifies the first locétion of the source data to be

{src_address}

copied.

{dest_address!

A longword address that specifies the destination of the first byte of data.
These addresses may be an actual address or a symbolic address.
address is specified, + is assumed.

If no

Examples:

L B

e B o)

>>> ex/n:4

00000000 00000000
00000004

00000000

00000008

00000000

Observe destination.

0000000C 00000000

00000010 00000000

KA640 Firmware

3-33

>>> ex/n:4

200

Observe source data.

Move the data.

Observe moved data.
|

P 00000200 58DDO0520

P 00000204 585E04C1
P 00000208 OOFFS8FEB
P 0000020C 5208A8D0
P+ 00000210 540CASDE
>>> mov/n:4 200

- >>> ex/n:4 0
00000000 58DDO0520
00000004

585E04C1

00000008 OOFFS8FEB

0000000C 5208A8D0
00000010 540CAS8DE

3-34

KA640 CPU System Maintenance

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/O mode.

Format:
NEXT {count}

The console implements the

NEXT command using the trace trap enable

and trace pending bits in the PSL, and the trace pending vector in the SCB.

The following restrictions apply:

If memory management is enabled, the NEXT command works only 1f

~ the first page in SSC RAMis mappedin SO (system) space.

(QOverhead associated with the NEXT command affects execution time

of an instruction.

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

Unpredictable results occur if the macro instruction being stepped over
modifies either the SCBB or the trace trap entry. This means that you
cannot use the NEXT command in conjunction with other debuggers.

Arguments:
{count}

A value representing the number of macro instructions to execute.

Examples:
>>>

ex pc

G 0000000F

00000200

>>> next

PC = 00000202
>>> next

= 00000213
>0>

KAB640 Firmware

3-35

00000000

01 NOP

00000001

01 NOP

00000002

00000003

01 NOP

NOP

00000004

01 NOP

00000005

01 NOP

00000006

NOP

00000007

NOP

Wiy

igioigigyiygtlogrd

>>> ex /ins /n:10 O

00000008

11 BRB

0000000A

00000002

NOP

0000000B

01 NOP

0000000C

0000000D

00 HALT

HALT

0000000E

HALT

0000000F

BALT

00000010

HALT

00000011

BALT

dep pc 0
n

00000001

01 NOP

00000002

01 NOP

00000003

01 NOP

00000004

NOP

00000005

01 NOP

nbS
00000006

3-36

01 NOP

00000007

01 NOP

00000008

00000002

01 NOP

00000003

01l NOP

BRB

00000002

KA640 CPU System Maintenance

3.9.12 REPEAT
The REPEAT command repeatedly displays and executes the specified
command. Press
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:

VHHHMKHKMMMH
M K H
M HMHH

>>>

repeat ex pr$_todr

Watch the clock.

0000001RB SAFE78CE
0000001B SAFE78D1
0000001B SAFE78FD
0000001B

SAFE7900

0000001B SAFE7903
0000001B

SAFE7907

0000001B SAFE790A
0000001B SAFE790D
0000001B S5AFE7910
0000001B SAFE793C

0000001B S5AFE793F
0000001B

SAFE7942

0000001B SAFE7946
0000001B SAFE7949

0000001B SAFE794C
0000001B SAFE794F
0000001B 5~C

KA640 Firmware

3-37

3.9.13 SEARCH
The SEARCH command finds all occurrences of a pattern and reports the
addresses where the pattern was found. If the /NOT qualifier is present,
the command reports all addresses in which the pattern did not match.
Format:

SEARCH [qualifier_list] {address} {pattern} [mask]

SEARCH accepts an optional mask that indicates bits to be ignored (don't

care bits). For example, to ignore bit 0 in the comparison, specify a mask
of 1. The mask, if not present, defaults to 0.
A match occurs if (pattern AND mask complement) = (data AND mask
complement), 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
Present

False
True
False

No report
No report
Report address

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

Qualifiers:

Data control: /B, /W, /L, /Q, /N:{count}, /STEP:{size}, WRONG
/
Address space control: /P, /V, /U
Command specific:
/NOT

3-38

Inverts the sense of the match.

KA640 CPU System Maintenance

Arguments:
Istart_address! A longword address that specifies the first location subject to the search. This
address can be an actual address or a symbolic address.
specified, + is assumed.

{pattern}

The target data.

imask]

A mask of the bits desired in the comparison.

If no address is

Examples:

>>>

search /w/step:1/n:£f£££f 20040000
!

Find all two-byte

200403C7 FE1l1

ROM that

"branch to

(brb assembles to FEl1l)

20040ECB FE11

d/n:10000

d/1 555 aaaaaaaa

>>>

search/p/b/not/n:££££f 0

v gy

>>>
>>>

vy
o

>>>

gty

>>>

oy
N

>>>

fell

20040002 FE11

sequences

in the

could be interpreted as a
self”

(10$:

brb

108%)

00000555 AA
00000556 aa

00000557 aaA
00000558 aa

search /w/step:1/n:£££f 20040000 fell
20040002 FE11

200403C7 FE11
20040ECB FE1l1

dep 1000 87654321

/1
search /p/b/n:£f£f£ff 0 1 fe

00001000

00001001 43
00001002

00001003

search /p/b/n:f£fff/not
00001000

00001001

00001002

00001003

KA640 Firmware

3-39

3.9.14 SET
The SET command sets the parameter to the value you specify.
Format:
SET {parameter} {value}
Parameters:
BFLAG

Set the default R5 boot flags. The value must be a hex number of up to 8 digits.
See Table 3-8 under the BOOT command description 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.
Note the hierarchy of the SET HOST qualifiers below.

DUP—Use the DUP driver to execute local programs of a device on either the
DSSI bus or the Q22-bus.

/DSSI node—Attach to the DSSI node. A node is a name up to 8 characters
in length or a number from 0 to 7.
/OQSSP—Attach to the UQSSP device specified using one of the following
methods:
/DISK n—Specifies the disk controller number, where n is a number
from O to 255. The resulting fixed address for n=0 is 20001468 and
the floating rank for n>0 is 26.
/TAPE n—Specifies the tape controller number, where n is a number
from O to 255. The resulting fixed address for n=0 is 20001940 and
the floating rank for n>0 is 30.
csr_address—Specifies the Q22-bus I/0 page CSR address for the
device.

/MAINTENANCE—Examines and modifies DSSI controller module configuration values. Does not accept a task value.
UQSSP—

/SERVICE n—Specifies service for DSSI controller module n where n
is a value from 0 to 3. (The resulting fixed address of 2 DSSI controller

module in maintenance mode is 20001910+4*n.)
/esr_address—Specifies the Q22-bus YO page CSR address for the
DSSI controller module.
|
"

LANGUAGE

Sets console language and keyboard type. If the current console terminal
does not support the DIGITAL Muitinational 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.

340

KAB40 CPU System Maintenance

Examples:
! Sets boot flags 5 and 9 (See boot £flag
! table in the BOOT command description.)

>>> set bflag 220
>>> set boot dual

>>> set host/dup/dssi 0

DSSI Node

DRVEXR V1.0
DRVTST V1.0

HISTRY V1.0

ERASE

V1.0

PARAMS V1.0
DIRECT V1.0

server...
(SUSAN)
DooboovY

Starting DUP

25-APR-1988

10: 01:35

25-APR-1988

10: 01:35

25-APR-1988 10: 01:35

25-APR-1988 10: 01:35
25-APR-1988 10: 01:35
25-APR-1988

10: 01:35

Digital Equipment Corporation

Remote Node

Path Block

DGS_S

DGS_R

MSGS S

MSGS R

FF8120D4

KAREN

RFX V101

FF8121D8

WILMA

RFX V101

FF8120DC

BETTY

RFX V101

FF8122E0

DSSI1

VMS V5.0

816

3045

FF8124E4

VMB BOOT

Internal Path

FF811lECC

PARAMS> exit

Exiting...
Task Name?
Stopping DUP

server...

KA640 Firmware

3-41

>>> set host/dup/dssi 0 params
Starting DUP

serxrver...

DSSI Node 0

(SUSAN)

Digital Equipment Corporation

PARAMS> show node
Parameterxr

Current

Default

Type

RF30

String

Default

Type

PARAMS> show allclass
Parametex

Current

PARAMS> exit
..
Exiting.
Stopping DUP

server...

>>>

342

KA640 CPU System Maintenance

Byte

3.9.15 SHOW
The SHOW command displays the console parameter you specify.
Format:

SHOW {parameter}
Parameters:
BFLAG

Displays the default RS boot flags.

BOOT

Displays the default boot device.

DEVICE

DSSI

Displays all devices displayed by the SHOW DSSI, SHOW ETHERNET, and

SHOW UQSSP commands.

Displays the status of all nodes that can be found on the DSSI bus. For each

node on the DSSI bus, the firmware displays the node number, the node name,
and the boot name and type of the device, if available. The command does not
indicate if the device contains a bootable image.

The node that issues the command is listed with a node name of * (asterisk).
The device information is obtained from the media type field of the MSCP
command GET UNIT STATUS. If a node is not running or is not capable of
running an MSCP server, then no device information is displayed.
ETHERNET

Displays hardware Ethernet address for all Ethernet adapters that can be
found, both on-board and on the Q22-bus. Displays as blank if no Ethernet
adapter is present.

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

D:splays 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 reports the addresses of bad pages, as defined
by the bitmap.

QBUS

Displays all Q22-bus I/O addresses that respond to an aligned word read, and

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

address in the VAX I/O space in hex, the address as it would appear in the
Q22-bus 170 space in octal, and the word data that was read in hex.

This command may take several minutes to complete.
Press
to
terminate the command.
During execution, the command disables the
scatter/gather map.

KA640 Firmware 3-43

RLV12

Displays all RL0O1 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
does not indicate if the device contains a bootable image.

This information is obtained from the media type field of the MSCP command GET UNIT STATUS. The console does not display device information if a node
is not running (or cannot run) an MSCP server.

VERSION

Displays the current firmware version.

Qualifiers: Listed in the parameter descriptions above.
Examples:
>>> show bflag

00000220
>>> show boot
' XQAO0
>>> show device

DSSI Node

-DIAO

DSSI Node

-DIA1l

DSSI

(KAREN)

(WILMA)

(RF30)

DSSI Node

-DIAS

(SUSAN)

(RF30)

DSSI Node

-DIA4

(RF30)

(BETTY)

(RF30)
Node

(*)

UQSSP Disk Controller 0
-DUA4

(772150)

(RDS3)

-DUAS

(RXS50)

-DUA6

(RX50)

UQSSP

Tape

-MUAO

(TKSO)

Controllexr

(774500)

Ethernet Adapter

-ESA0

(AA-00-03-01-2E-3F)

3-44

KA640 CPU System Maintenance

>>> show dssi

DSSI Node 0
-DIAO0

DSSI Node 1

-DIA1

(WILMA)

(BETTY)

(RF30)

.DSSI Node
>>>

(RF30)

DSSI Node

-DIAS

(KAREN)

(RF30)

DSSI Node

-DIA4

(SUSAN)

(RF30)

(*)

show ether

Ethernet Adapter

-ESA0
>>>

(AA-00-03-01-2E-3F)

show lang

English
>>>

(United States/Canada)

show memory

Memory O:

00000000 to OO3FFFFF,

4MB,

0 bad pages

Memory

00400000 to

OOBFFFEF,

'8MB, 0 bad pages

Memory 2:

00C00000 to

O13FFFFF,

8MB,

0 bad pages

Total

of 20MB,

0 bad pages,

106 reserved pages

>>> show memory/full
Memory

00000000 to

OO3FFFFF,

4MB,

0 bad pages

Memory

00400000 to

OOBFFFFF,

8MB,

00C00000 to O13FFFFF,

8MB,

0 bad pages

Memory 2:
Total

of 20MB,

0 bad pages,

106

bad pages

reserved pages

Memory Bitmap
-013F2C00 to Ol13F3FFF,
Console

pages

Scratch Area

-013F4000

013F7FFF,

Qbus Map

-013F8000 to O013FFFFF,
Scan

64 pages

of Bad Pages

KA640 Firmware

3-45

>>>

show gbus

Scan of Qbus

I/0 Space

-20001468

(772150)

4000

-2000146A

(772152)

= 0B4O

-20001940

(774500)

= 0000

-20001942

(774502)

= OBCO

-20001F40

(777500)

0020

Scan
>>>

RQEDX3/KDAS0/RRD50/RQC25/X-DISK

(260)

TQKS50/TQK70/TUS1E/RV20/X-TAPE

(004)

IPCR

of Qbus Memory Space
show ugssp

UQSSP

Disk Controller 0

-DUA4

(RD53)

~-DUAS

(RX50)

-DUA6

(RX50)

UQSSP

Tape Controller 0

-MUAO

(TKS50)

>>>

(154)

(772150)

(774500)

show version

KA640-A V4.0,

VMB 2.4

>>>

3-46

KAB40 CPU System Maintenance

3.9.16 START
The START command starts instruction execution at the address you
~ specify. If nc 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])

The address at which to begin execution. This address is loaded into the user’s
PC.

Examples:
>>>

start

1000

KAB40 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-;iigit 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

41..40..39..38..37..36..35..34..33..32..31..30..29..28..27..26..

25..24..23..22..21..20..19..18..17..16..15..14..13..12..11..10..
09..08..07..06..05..04..03..
>>>

348

KA640 CPU System Maintenance

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

Format:
UNJAM

Examples:
>>> unjam
>>>

KA640 Firmware

349

3.9.19 X—Binary Load and Unload
The X command is for use by automatic systems communicating with the
_
console.

' The X command loads or unloads (that is, writes to memory, or reads from
memory) the specified number of data bytes through the console serial line
(regardless of console type) starting at the specified address.

Format:

X {address}] {count] CR {line_checksum} {data} {data_checksum]}
If bit 31 of the count is clear, data is received by the console and deposited
into memory. If bit 31 is set, data is read from memory and sent by the
console. The remaining bits in the count are a positive number indicating
the number of bytes to load or unload.
The console accepts the command upon receiving the carriage return.
The next byte the console receives is the command checksum, which is
not echoed. The command checksum is verified by adding all command
characters, 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.

3-50

KA640 CPU System Maintenance

If the data checksum is incorrect on a load, or if memory or line errors occur
during the transmission of data, the entire transmission is completed, then
the console issues an error message. If an error occurs during loading, the
contents of the memory being loaded are unpredictable.
The console represses echo while it is receiving the data string and
checksums.

The console terminates all flow control when it receives the carriage return

at the end of the command line in order to avoid treating flow control
characters from the terminal as valid command line checksums.
You can control the console serial line during a binary unload using control
characters ([cTRic], [CTRUS], [€TRUO], and so on). You cannot control the console
serial line during a binary load, since all received characters are valid
binary data.
The console has the following timing requirements:
*

It must receive data being loaded with a binary load command at a rate
of at least one byte every 60 seconds.

It must receive the command checksum that precedes the data within

60 seconds of the carriage return that terminates the command line.

It must receive the data checksum within 60 seconds of the last data

byte.

If any of these timing requirements are not met, then the console aborts
the transmission by issuing an error message and returning to the console
prompt.

The entire command, including the checksum, can be sent to the console
as a single burst of characters at the specified character rate of the console

serial line Tbhe cousole is able to receive at least 4 Kbytes of data in a
single X command.
|

KAB640 Firmware

3-51

3.9.20 '—Comment
The comment character (an exclamation point) is used to document
command sequences. It can appear anywhere on the command line. All
characters following the comment character are ignored.
Format:

Examples
>>>

The console

ignores

this

line.

>>>

3-52

KA640 CPU System Maintenance

Chapter 4

Troubleshooting and Diagnostics

4.1 Introduction
This chapter contains a description of KA640 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.
address and vector values for most options.

Microsystems Options lists

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. Nonstandard addresses can be selected, 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 to save time
and prevent you from introducing new problems.

If the system fails (or appears to fail) to boot the operating system, 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-based diagnostics described in this chapter. In addition, check the
following connections:

If no message appears, make sure the console terminal and the system
are 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 on.

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 H3602-SA. If the display is off, check
the CPU module LEDs and the CPU cabling. If a hex error message
appears on the H3602—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 KA640 ROM-Based Diagnostics
The KA640 ROM-based diagnostic facility, rather than the MicroVAX Diagnostic Monitor (MDM), is the primary diagnostic tool for troubleshooting
and testing of the CPU, memory, Ethernet, and DSSI subsystems. ROMbased diagnostics have significant advantages:
e

Load time is virtually nonexistent.

The boot path
is more reliable.

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

4-2

KA640 CPU System Maintenance

The ROM-based diagnostics can indicate several different FRUs, not just
the CPU module. For example, they can isolate one of up to three memory
modules as FRUs. (Table 4-6 lists the FRUs indicated by ROM-based
diagnostic error messages.)

The diagnostics run automatically on power-up. While the diagnostics are
running, the LEDs on the H3602-SA display a hexadecimal countdown of
the tests from F to 3 (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 manufactuning
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 KA640 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 also 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:

Test is the test code or utility code.

Address is the test or utility’s base address in ROM. This address varies.
The addresses shown are examples only. 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.

Parameters shows the parameters for each diagnostic test or utility.
The asterisks (*) represent
Tests accept up to ten parameters.
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.

Troubleshooting and Diagnostics

4-3

Table 4-1:
Test

Test and Utility Numbers

Address

Name

Parameters

2004C96B

SSC RAM

C2
C5

2004CB32
2004CCA2

SSC RAM ALL
SSC regs

*
*

kK Kk K
sk

Cé6

2004CD9C

SSC_powerup

2004CE60

CBTCR timeout

2004CF1C

ROM logic test

2004CFE4

CMCTL_powerup

33
32

2004D02C

CMCTL regs

MEMCSRO_addr *x##*swe

2004D150

CQBIC_powerup

2004D1E2

CQBIC regs

2004D23B

CQBIC-memory

A
SR

60
61

.2004D63D
2004D98A

Console serial
Console QVSS

start_baud end_baud ***¥**
mark_not_present ***

2004DCCC

QDSS self-test
CFPA

input_csr selftest_r0 selftest_r] ##****
e

2004DA38
2004DE33

Console QDSS

mark_not_present selftest_r0 selftest_rl *****

2004E01F

Prog timer

which_timer wait_ time_us ***

2004E2EC

TOY clock

repeat_count_250ms_ea *#**

2004E557

. Virtual mode

2004E834

interval timer

2004300
2004ECO05

SII_ext_loopbck
SII_initiator

ok
e

2005109A

DSSI reset

port_no time_secs *

20051484

VAX CMCTLCDAL

dont_report_memory_bad repeat_count *

2005159C

SII_memory

incr test_pattern *#¥%%

20051954

SII_registers

20051A9C

NI_memory

incr data_pattern ***

41
42

20052768
200527EC

Board reset
Check-for_intrs

200528E8

Cache_memory

addr_incr sty

20052C3C

Cache_mem_cqbic

start_addr end_addr addr_incr #*%%

20053C6D

MEM__bitmap

*#% mark_Hard_SBEsg ###%%#

20053D69

MEM_data

start_add end_add add_incr cont_on_err *¥*##*

20053F2E

20054050

MEM_byte

start_add end_add add_incr cont_on_err ******

200541F9

MEM_ECC_error

start_add end_add add_incr cont_on_err **¥¥#x

4-4

KA640 CPU System Maintenance

MEM_address

start_add end_add add_incr cont_on_err S#*s %

Table 4-1 (Cont.):

Test and Utility Numbers

Test

Address

Name

4B
4A
49
48

20054595
20054779
20054995
20054FCE

MEM_maskd_errs
MEM_correction
MEM_FDM_logic
MEM_addr_shrts

47
40

2005540A
200555AC

MEM_refresh
MEM_count_errs

200557BD

20055FCC

List CPU regs

Utilities

S9E

200560C6

List diags

200560C6

Create script

81.

200567E4

MSCP-QBUS test

IP_csr *¥*¥**

200569AB

DELQA

device_num_addr ****

- Parameters
start_add end_add add_incr cont_on_err ******
start_add end_add add_incr cont_on_err ¥*****
*%% cont_on_err *¥¥rxx
start_add end_add * cont_on_err patl pat2

3q kKK

start end incr cont_on_err time_seconds *¥***
First_board Last_board ******* Soft_errs_
allowed

Expnd_err_msg get_mode init_LEDs cir_ps_
cnt

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

2004ES557

Virtual mode

XKL XX

20053CéD

MEM bitmap

**x* mark Hard SBEs *****%

The virtual mode test on the first line contains several parameters, but you
cannot specify any of them. To run this test individually, enter:
>>> T

The MEM_bitmap test on the second line accepts ten parameters, but you
can specify only the fourth one. To mark pages bad in the bitmap for singlebit or multi-bit errors, enter a 1 in the fourth parameter field:
>>> T

You must enter a value of either 0 (zero) or 1 (one) for the first three
parameters. (0 is used in this example.) The values have no effect on
the test; they are simply place holders for the first three parameters. You
do not have to specify a value for parameters that follow the user-defined
parameter.

Troubleshooting and Diagnostics

4-5

4.3.2 Scripts
Most of the tests shown by utility 9K 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, continue).

The power-up script runs every time the system is powered on. You can
also invoke the power-up script at any time by entering T 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 KA640 hardware
and firmware can also create their own scripts, by using the 9F utility. (See
Section 4.3.4.)

KAB640 CPU System Maintenance

Table 4-2 lists the scripts.

Table 4-2:

Script!

Scripts Available to Field Service

Enter with
TEST
Command

Description

Soft script created by de_test9f. Enter T SF to create..

Al, AA,

Common section of power-up script. Scripts AA, AB, and AC

'AB, AC,
0,3

invoke this script at power-up.

A7, A8

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.
A8

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.
A9

AA 0

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

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

Console program.

Does not invoke any

Runs memory tests, marks bitmap, resets

busmap and resets caches. Calls script AE.

AE, AD

Console program. Resets memory CSRs and resets caches.

Console program. Resets busmap and resets caches.

BA, 2, AA
,

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

BC,AA,AC,0,3

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

1Scripts A2-A6, BO-B3, and B3 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 at this time.
Scripts BD, BE, BF, B4, and B6-B9 are not used.

Troubleshooting and Diagnostics

4-7

In most cases, Field Service needs only the commands 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 powerup.

Primarily runs the memory tests; halts upon first hard or soft error.
Memory acceptance script; marks hard multi-bit and single-bit ECC errors in

the bitmap. Script A8 calls script A7 when this command is entered. Script
A7 contains the memory tests that will continue on error.
AT

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.

DE determines that the environment is nonmanufacturing from
H3602-SA. (Manufacturing sets a jumper on the H3602-SA for
testing.)

DE selects script sequence for console SLU.

DE executes Script BA.

- —
d.

Script BA directs DE to execute test.
are off.)

(Console announcements

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:

4-8

KA640 CPU System Maintenance

Battery is dead.
H3602-SA switch set to language inquiry.

Contents of SSC NVRAM are invalid.
7.

Calls DE with Test Code = 3
a.

DE executes Script AC.

Script AC directs DE to execute scripts BC and Al.
— Script BC directs DE to execute tests.

(Console announcements

are on.)

— Script Al directs DE to execute tests. (Console announcements
are on.)

b.
8.

DE passes control back to CP.

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 0 at the console prompt (>>> T 0):

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

DE determines environment is nonmanufacturing from H3602-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 announcements
are on.)

— Script Al directs DE to execute tests. (Console announcements
are on.)
C.

DE passes control back to the CP.

CP prints 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 Al) are run both times.

Troubleshooting and Diagnostics

4-9

4.3.4 User Created 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, in the 128-Kbyte mass storage interface (MSI) RAM in the
SII chip, 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 MSI RAM, and a larger script can be built in MSI
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

4-10

KAB40 CPU System Maintenance

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 = MSI RAM, 2 = Main Memory,
and 3 = Fastest Possible. Choose number 3 to select the fastest possible
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
Example 4-1 shows how to build and run a user script.

The utility displays the test name after you enter the test number, 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 A0<CR>
to run the new script.

You cannot review or edit a script you have created.

Troubleshooting and Diagnostics

4-11

Examplé 4-1: Creating a Script with Utility 9F
>>>T

SP=20140604

Create script in ?[0=SSC,1=SII RAM,62=RAM]
:
Script starts at 2011FCO00

1024 bytes left

Next test number :51
>>Run from ?[0=ROM, 1=SII RAM, 2=RAM, 3=fastest possible]
CFPA
CFPA

>>Repeat?

[O=no,l=on error,2=forever,>2=count<fFF]

CFPA

>>Error severity

CEPA

>>Console error report?

[O=none,l=full]

CFPA

>>Stop script on error?

[0=NO,1=YES]

CFPA

>>LED on entry

CFPA

>>Console on entxy

1017 bytes
Next

[0,1,2,3]

(0):

(2):

(1):

(05):

(51):

left

test number :52

Prog timer

>>Run from ? [0=ROM, 1=SII_RAM,
2=RAM, 3=fastest possible]

Prog timer

>>Repeat?

Prog timer >>Error

[0=no,l=on error,2=forever,>2=count<fFF]

severity

[0,1,2,3]

(2):

Prog timer

>>Console error report?

[O=none,l=full]

Prog timer

>>Stop script on error?

[0=NO,1=YES]

Prog timer

>>LED on entry

Prog timer

>>Console on entry

Prog timer

>> which timer

Prog timer

>> wait_time
us

1002 bytes

left

test number

(0):

(1):

(05):

(52):

00000000
:

00000001

2(00000000)

- FFFFFFFF ?(0000000RA)

>>>T AQ

51..52..
>>>

Example 4-2 shows the script building procedure to follow 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 AO at the next Next test number: prompt.
Press <CR> at the next Next test number: prompt.

Enter T AO to begin running the script repeatedly.
Press

4-12

to stop the test.

KA640 CPU System Maintenance

(0):

The above 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.
Example 4-2:
>>>

Listing and Repeating Tests with Utility 9F

SpP=20140604
Create

script

Script

starts at 20140758

bytes

test

?2[0=SSC,1=SII_RAM,
2=RAM]

left
number

Test

Address

Name

Parameters

2004D587

SSC RAM

2004D74E

SSC RaM ALL

'C5

2004DS8BE

SSC regs

20056166

MEM Refresh

start

20056308

MEM Count Errs

First board Last board Soft errsallowed **x#*x=2x

20056519

List

20056D28

Utilities

Expnd
err msg get mode

20056DFC

List

20056E22

Create

20057540

MSCP-QBUS test

IP csx

20057707

DELQA

device
num addr

bytes

CPU regs

end

incr

script

test number

*****+*

Count Errs>>Console

Count Errs>>Stop

MEM Count Errs>>LED
Count Errs>>

error

script

report?

Next
>>>

(2):

[O=none,l=full]

on error?

on entry

[0=NO, 1=YES]

(0):

(1):

(04):

First board

MEM Count Exrs>> Last board
Count Exrs>>

(0):

[0=no,l=on error, 2=foreverx,>2=count<tF]

MEM Count Errs>>Console on entry
:

(40):

00000001

Soft_ errs_allowed

00000004

00000000

?(00000004)

FFEFFFFF

left

test number

bytes

****

from 2[0=ROM, 1=SII_RAM, 3=fastest possible]

Errs>>Repeat?

MEM

clr_ ps_cnt

:40

MEM

S5 bytes

LEDs

ERXKKS

ME&ZCount:?:rs>>Error severity 2 {0,1,2,3]

init

left

MEM Count

MEM

***=x*

diags

MEM Count Erxrs>>Run

MEM

cont_on err time seconds

tA0 ~

script

left

test number

T AC

40..40..40..40..40..40..40..40..40..40..40..40..40..40..40..4vu..
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
>>>

>>>

test

number

T A0

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..

40..40..40..40..40..40..40..40..40..40..40..40..40..40..40..40..
>>>

Troubleshooting and Diagnostics

4-13

4.3.5 Console Displays
Example 4-3 shows a typical console display during execution of the ROMbased 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).
Example 4-3:

Console Display (No Errors)

KA640-A V4.1,

VMB 2.5

Performing Normal System Tests

41..40..39..38..37..36..35..34..33..32..31..30..29..28..27..26..
25..24..23..22..21..20..19..18..17..16..15..14..13..12..11..10..
09..08..07..06..05..04..03..
Tests

completed

>>>

The first line contains the firmware revision (V4.1 in this example) and the
virtual memory bootstrap (VMB) revision (V2.5 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 44.

Example 44: Sample Output with Errors
2?46 2

07 FE 10

0002

P1=002F0000 P2=00000000 P3=00000000 P4=00FF0000

P5=00000000

P6=00000000 P7=00000000 P8=00000000 PS=00FF0000

P10=00000000

r0=00000000 rl1=00010000 r2=55555555 r3=00000080
r5=00000080 r6=01EF0000

xr7=20080144 r8=00010000

Tests completed

4-14

KA640 CPU System Maintenance

r4=AAAAAAAA
ERF=20140770

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

Severity

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

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.

Troubleshooting and Diagnostics

4-15

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 4-4:

Values

Saved,

Machine

Check

Executive
Value

Parameter
P1

Contents of SP, points to vector value in P2

P2
P3

Vector = 04, vector of exception 04—FC, 00 = Q-bus
Address of PC pointing to failed instruction, P9

Byte count = 10

PS5

Machine check code

Most recent virtual address

" Internal state information 1
Internal state information 2

Exception

PC, points to failing instruction

P9
P10

PSL

Table 4-5:

Values Saved, Exception During Executive

Parameter

Value

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

Contents of stack

P10

Contents of stack

4-16

KA640 CPU System Maintenance

During

Lines 4 and 5 of the error printout are general registers RO 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.

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 Action on Error column refers to the action taken by the
diagnostic executive under the following circumstances:
*

The diagnostic executive detects an unexpected exception or interrupt.

A test fails and that failure is reported to the diagnostic executive.

The Default on Error column does not refer to the action taken by the
memory tests. The diagnostic executive either halts the script or continues
execution at the next test in the script.
Most memory tests have a continue on error parameter (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 diagnostic
executive that an error has occurred.

Figure 4-1 shows the LEDs on the KA640 CPU. They correspond to the
hex display on the H3602-SA.
Figure 4-1:

KA640 CPU Module LEDs

—

GREEN
DCOK LED
RED LEDs

VALUE ON

VALUE OFF

0
wm O-001286

Troubleshooting and Diagnostics

4-17

Table 4-6:
Hex

LED

KA640 Console Displays and FRUs

Normal Error

Console Console

Display Display

Defaulton

Error

Description

FRU

6,1,4,5-2

Initial Power-Up Tests

None

Loopontest

Power-up

None

Loopontest

WAIT_POK

None

Loopon self

Entering IPT

None

Loopontest

SLU_EXT LOOPBACK3

7,1

Continue
Continue

Utilities
Check_for_intrs
SSC_power-up

1
1
1

CONSOLE_SERIAL

Script BA

C
B
C

None
None
None

29D
242
?C6

None

760

Continue
Continue

End of script

Script AC
Invoke script BC.
Invoke script Al.

End of script.

Script AA
Invoke script BA.
Invoke script BC.
Invoke script Al.

End of script.

1In the case of multiple FRUs, refer to Section 4.5.2 for further information.

2If a problem recurs with the same FRU, check that the tolerance for system power supply +5
Vdc, +12 Vdc, and AC ripple are within specification.

3This test runs only if the power-up mode switch on the H3602-SA is set to TEST mode. See
Section 4.6.1.
FRU key:

1 = KA640, 2 = MS650, 3 = memory interconnect cable
4 = Q22-bus device, 5 = Q22/CD backplane, 6 = system power supply
7 = H3602-SA 1/0 panel

4-18

KA640 CPU System Maintenance

Table 4-6 (Cont.): KA640 Console Displays and FRUs
Hex
LED

Normal Error
Console Console
Display Display

Defaulton
Error
Description

FRU

Script BC
7

791

Continue

CQBIC_power-up

?90

Continue

CQBIC_registers

1
1

Continue

CMCTL_power-up

- 232

733 .

Continue

CMCTL _registers

?5B

Continue

SII_registers

731

Continue

CMCTL _setup_CSRs

1,2,3,5

749

Continue

MEMORY_FDM_logic

2,1.3,5

End of script

‘

Script Al
C

752

Continue

PROG_TIMER_0O

752

Continue

PROG_TIMER_1

753

Continue

TOY_CLOCK

?2C1

Continue

SSC_RAM

734

Continue

ROM_logic

7C5

Continue

SSC_registers

257

Halt

SII_memory

2C2

continue

SSC_RAM_addr_shorts

735

Continue

INTERVAL_TIMER

751

Continue

CFPA

2C7

Continue

246

Continue

= CBTCR_timeout
CACHE_DIAG_MODE

1
1

230

Halt

MEMORY_bitmap

1,2,3,3

24F

Continue

MEMORY_data

2,135

24E

Continue

MEMORY_byte

2.1,3,5

24D

Continue

MEMORY_addr

2,1.3,5

FRU key:
1 = KA640, 2 = MS650. 3 = memory interconnect cable
4 = Q22-bus device, 5 = Q22/CD backplane, 6 = system power supply
7 = H3602-SA 1/0 panel

Troubleshooting and Diagnostics

4-19

Table 4-6 (Cont.):
Hex

LED

KA640 Console Displays and FRUs

Normal Error

Console Console

Display Display

Default on

Error

Description

FRU

Continue
Continue
Continue
Continue
Continue

MEMORY_ECC_error
MEMORY_masked_errors
MEMORY _correction
MEMORY_address_shorts
MEMORY _refresh

2,1,3,5

Script Al

24C

8
8
8
8

17
16
15
14

74B
24A
748
247

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

240

Continue

MEMORY_count_badpages

244

Continue

CACHE1_MEMORY

1,2,3,5

780

Continue

CQBIC_MEMORY

1,2,4,3,5

B
7

10
09

754
245

Continue
Continue

VIRTUAL_MODE
CACHE_MEM_CQBIC

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

75A

Continue

CVAX_CMCTL drivers

5
5
4

07
06
05

725C
?5D
725E

Continue
Continue
Continue

Continue

SII_initiator
SII_target
NI_memory

NI_functional

1
1
1

7241

Continue

KA640_RESET

1,4

?5F

End of script.

Script A9

8
8
8
8
8
8
8
8

4F
4E
4D
4C
4B
4A
48
47

?24F
4E
24D
24C
?4B
24A
7248
247

Halit
Halt
Halt
Halt
Halt
Halt
Continue
Continue

MEMORY_data
MEMORY_byte
MEMORY_addr
MEMORY_ECC_error
MEMORY_masked_errors
MEMORY_correction
MEMORY_address_shorts
MEMORY_refresh

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

240

Continue

MEMORY_count_badpages

2,1,3,5

FRU key:

1 = KA640, 2 = MS650, 3 = memory interconnect cable
4 = Q22-bus device, 5 = Q22/CD backplane, 6 = system power supply
7 = H3602-SA /O panel

4-20

KA640 CPU System Maintenance

Table 4-6 (Cont.):
Hex

LED

KA640 Console Displays and FRUs

Normal Error
Console Console

Display Display

Defaulton
Error
Description

FRU

Continue

1,4

Script A9
C

241

KA640_RESET

End of script.

Script A8
9

231

Halt

CMCTL_setup_CSRs

1,2,3,5

249

Halt

MEMORY_FDM_logic

2,1,3,5

730

Halt

MEMORY_bitmap

1,2,3,5

Invoke script A7.

End of script.

Script A7
8

24F

Halt

MEMORY_data

2,1,3,5

Halt

MEMORY_byte

2,1,3,5

24D

Halt

MEMORY_addr

2,1,3,5

24C

Halt

MEMORY_ECC_error

2,1,3,5

?24B

Halt

MEMORY_masked_errors

2,1,3,5

24A

Halt

MEMORY_correction

2,1,3,5

7248

Halt

MEMORY_address_shorts

2,1,3,5

247

Halt

MEMORY_refresh

2,1,3,5

240

Cont

MEMORY_count_bad pages

2,1, 3,5

780

Cont

CQBIC_memory

741

Halt

KA640_RESET

1,2,4,3,5

1, 4

End of script.
FRU key:
1 = KA640, 2 = MS650, 3 = memory interconnect cable

4 = Q22-bus device, 5 = Q22/CD backplane, 6 = system power supply
7 = H3602-SA VO panel

Troubleshooting and Diagnostics

4-21

4.3.6 System Halt Messages
Table 4-7 lists messages that may appear on the console terminal when a
system error occurs.

Table 4-7:

System Halt Messages

Code

Message

Explanation

202

EXT HLT

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

204

ISP ERR

Caused by attempt to push interrupt or exception state onto
the interrupt stack when the interrupt stack was mapped NO
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

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

208

SCB ERR2

The vector had bits <1:0> = 2.

?0A

CHM FR ISTK

A change mode instruction was executed when PSL<IS> was

set.

?0B

CHM TO ISTK

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

?20C

SCB RD ERR

A hard memory error occurred during a processor read of an

MCHK AV

An access violation or an invalid translation occurred during

?11

KSP AV

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

212

DBL ERR2

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

213

DBL ERR3

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

?19

PSL EXC5

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

?1A

PSL EXCé6

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

exception or interrupt vector.

- 710

machine check exception processing.

exception.

?1B

PSL EXC5

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

?1D

PSL EXC5

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

21E

PSL EXC5

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

?1F

PSL EXC5

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

4-22

KA640 CPU System Maintenance

4.3.7 Console Error Messages
Table 4-8 lists messages issued in response to an error or to a console
command that was entered incorrectly.

Table 4-8:

Console Error Messages

Code Message

Explanation

720

CORRPTN

The console data base was corrupted. The console simulates
a power-up sequence and rebuilds its data base.

721

ILL REF

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.

722

ILL CMD

The command string cannot be parsed.

223

INV DGT

A number has an invalid digit.

724

LTL

The command was too large for the console to buffer.
message is sent only after the console receives the
the end of the command.

The
at

225

ILL ADR

The specified address is not in the address space.

726

VAL TOO LRG

The specified value does not fit in the destination.

227

SW CONF

298

Switch conflict.
For example,
specifies two different data sizes.

UNK SW

The switch is not recognized.

229

UNK SYM

The EXAMINE
recognized.

CHKSM

An X command has an incorrect command or data checksum.

DEPOSIT

an EXAMINE command

symbolic

address

not

If the data checksum is incorrect, this message is issued, and
is not abbreviated to “Illegal command.”

HLTED

The operator entered a HALT command.

22C

FND ERR

A FIND command failed either to find the RPB or 64 Kbytes
of good memory.

722D

TMOUT

Data failled to arrive in the expected time during an X
command.

MEM ERR

Memory error or machine check occurred.

22F

UNXINT

An unexpected interrupt or exception occurred.

730

UNIMPLEMENTED

Unimplemented function.

7231

QUAL NOVAL

Qualifier does not take a value.
Ambiguous qualifier.

232

QUAL AMBG

233

QUAL REQ VAL

Qualifier requires a value.

734

QUAL OVERF

Too many qualifiers.

735

ARG OVERF

Too many arguments.

736

AMBG CMD

Ambiguous command.

237

INSUF ARG

Too few arguments.

Troubleshooting and Diagnostics

4-23

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
Number

VMB Error Messages
Mnemonic

Interpretation

240

NOSUCHDEV

No bootable devices found

?41

DEVASSIGN

Device is not present

242

NOSUCHFILE

Program image not found

243

FILESTRUCT

Invalid boot device file structure

244

BADCHKSUM

Bad checksum on header file

245

BADFILEHDR

Bad file header

246

BADDIRECTORY

Bad directory file

247

FILNOTCNTG

Invalid program image format
Premature end-of-file encountered

748

ENDOFFILE

749

BADFILENAME

Bad file name given

?24A

BUFFEROVF

Program image does not fit in available memory

?24B

CTRLERR

Boot device 1/0O error

24C

DEVINACT

Failed to initialize boot device

24D

DEVOFFLINE

Device is off line

?4E

MEMERR

Memory initialization error

Unexpected SCB exception or machine check

?24F

SCBINT

?50

SCB2NDINT

Unexpected exception after starting program image

751

NOROM

No valid ROM image found

252

NOSUCHNODE

No response from load server

233

INSFMAPREG

Insufficient Q-bus mapping registers due to invalid
memory configuration, bad memory, or because Q-bus
map was not relocated to main memory

7254

RETRY

No devices bootable, retrying

4.4 Acceptance Testing

Perform the acceptance testing procedure listed 'below, after installing a
system or whenever replacing the following:

KA640 module
MS650 module
|
Memory data interconnect cable
Backplane
DSSI drive
H3602-SA

4-24

KA640 CPU System Maintenance

Run five error-free passes of the power-up scripts by entering the

following command:
>>> RTO

Press

to terminate the scripts.

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

>>>
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 Al, which invokes CPU and memory
tests without resetting the bitmap to mark only solid multi-bit ECC
errors in main memory. This command gives you a quick memory check,
since most tests run on a 256-Kbyte boundary.

Perform the next two steps for a more granular test of memory.
>>> T A8

>>> R T A7

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[cTRuc] 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 of the
memory diagnostics test memory on a page boundary.

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
reportedin test 40. Such a failure indicates that pages in memory have
been marked badin the bitmap because of solid single-bit and/or multibit ECC errors. The error printout does not display the 20 longwords,
since it is a sevenity level 1 error.

Check the memory configuration again, since test 31 can check for only
a few invalid configurations. For example, test 31 cannot report that a
memory board is missing from the configuration, since it has no way of
knowing if the board should be there or not.

Troubleshooting and Diagnostics

4-25

To check the memory configuration, enter the following command line:
>>>SHOW MEMORY/FULL

Memory 0:

0000000 to

Memory

00400000

Total of 12MB,

OO3FFFFF,

to OOBFFFFF,

0 bad pages,

102

4MB,

0 bad pages

8MB,

O bad pages

reserved pages

Memory Bitmap

-00BF3400

OOBF3FFF,

6 pages

- Console Scratch Area
-00BF4000

OOBF7FFF,

32 pages

-00BF8000 to

OOBFFFFF,

64 pages

Qbus Map

Scan of Bad Pages
>>>

Memory 0 is the KA640 CPU. Memories 1 through 3 are the MS650
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
TOY

9C
=018200B8

ICCs =00000000

TCRO =00000000

TIRO =00000000

TNIR0=00000000

TCR1 =00000000

TIR1 =00000000

TNIR1=00090000

TIVR1=0000007C

RXCS =00000000

RXDB =0000000D

TXCS =00000000

TXDB =00000030

CADR =0000000C

TIVRO0=00000078

CACR =FFB40080

MSER =00000000

BDR

=FFFFFFDO

DLEDR=0000000C

SSCCRrR=00D45077

CBTCR=00000004

SCR

=0000C000

DSER =00000000

QBEAR=0000000F

DEAR =00000000

OBMBR=007F8000

IPCRn=0020

MEMFRU 1 MEMCSR0=80000015

1=00000015

2=00000015

MEM | FRU 2

MEMCSR 4—80400016

5=80800016

6=00000016

7=00000016

MEM FRU

MEMCSR
. 8=00000000

9=00000000

10=00000000

11=00000000

MEMCSR12=00000000

13=00000000

14=00000000

15=00000000

MEMCSR16=8094000F

17=00000000

MEM FRU 4

3=00000015

One memory bank is enabled for each 4 Mby‘tes of memory.
MEMCSRs map modules as follows:
MEMCSR 0

KA640 CPU

MEMCSR 4-7

First MS650 memory module

MEMCSR 8-11

Second MS650 memory module

MEMCSR 12-15

Third MS650 memory module

4-26

KA640 CPU System Maintenance

The

Verify the following:

For the KA640 CPU module, the bank enable bit (<31>) in
MEMCSR_O is set, indicating the memory bank on the KA640
is enabled. MEMCSR1-3<31> should always be clear, indicating
banks are not enabled. MEMCSRs <6:0> should equal 15 hex for
MEMCSR 0-3.
If MS650-AA modules are installed, the bank enable bits are set
in the first two MEMCSRs and cleared in the last two MEMCSRs.
MEMCSRs <6:0> should equal 16 hex for all four MEMCSRs. See
the values for MEMCSR 4-7 in the example.
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 8-11 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 4 and 5 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 KA640 CQBIC chip,
and the configuration of the Q22-bus, as follows:
>>>

show gbus

Scan of Qbus I/0O Space

-20000120

(760440)

= 0080

-20000122

(760442)

= FO081

(300)

DHQl1l/DHV11l/CXAl6/CXB16/CXY08

-=20000124

(760444)

= DD18

-20000126

(760446)

= 0200

-20000128

(760450)

= 0000

-2000012a

(760452)

= 0000

-2000012C

(760454)

= 8000

-2000012E

(760456)

-20001920

(774440)

= FF08

-20001922

(120)

DELQA/DEQNA

(774442)

= FFOO

-20001924

(774444)

= FF2B

-20001926

(774446)

= FFO06

-20001928

(774450)

= FF16

-20001922

(774452)

= FFF2

-2000192C

(774454)

-2000192E

(774456)

= 1030

-20001940

(774500)

= 0000

~-20001942

(774502)

-20001F40

(777500)

= 0020

Scan

0000

O0OFS8

(260)

TQKS0/TQK70/TU81E/RV20/K-TAPE

(004)

IPCR

O0BCO

of Qbus Memory Space

>>>

Troubleshooting and Diagnostics

4-27

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 (20001F40).

Second column = the Q22-bus address of the CSR, in octal (777500).
Third column = the data, contained at the CSR address, in hex
(0020).

Fourth column = the device vector in octal, according to the fixed
or floating Q22-bus and UNIBUS algorithm (004).
Fifth column = the device name(IPCR, the KA640 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:
>>> T

20001940

You can specify other addresses if you have multiple MSCP or TMSCP
devices in the first parameter. This action may be useful to isolate
a problem with a controller, the KA640, 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 1is tested by default.

Check that all UQSSP, MSCP, TMSCP, and Ethernet controllers and
devices are visible by typing the following command line:

4-28

KA640 CPU System Maintenance

>>>

show device

DSSI Node 0
-DIAO0

(R3YRME)

(RF30)

DSSI Node 1

-DIA1l

(R3VBNC)

(RF30)

DSSI Node 7

(*)

UQSSP

Tape

-MUAO

(TK70)

Controller

(774500)

Ethernet Adapter

-ESA0

(08-00-2B-08-E8-6E)

Ethernet Adapter 0 (774440)
~XQOAO

(08-00-2B-06-16-F2)

In the above example, the console displays the remote DSSI node names
and node numbers of two RF30 controllers it recognizes. The lines
below each node name and number are the logical unit numbers of any
attached devices, DIAO and DIA1 in this case.
DSSI Node 7 (*) is the KA640 DSSI adapter. In most cases, the KA640
is the local DSSI node shown by the asterisk and has a node number
of 7. DSSI node names, node numbers, and unit numbers should be
unique.

The UQSSP (TQK70) tape controller and its CSR address are also
shown. The line below this display shows a TK70 connected.
The next two lines show the logical name and station address for the
KA640 Ethernet adapter.
The last two lines refer to DELQA and DEQNA controllers, the Q22-bus

CSR address, logical name (XQAO), and the station address.

Test the DSSI subsystem using the KA640 ROM-based Diagnostics and
Utilities Protocol (DUP) facility. This facility allows you to connect to
the DUP server in the RF drive controller. Here are some examples:

- >>> set host/dup/dssi 7
Starting DUP

server...

Stopping DUP

server.
..

In this example, a DUP connection was made with DSSI node 7, the
KA640. The DUP server times out, since no local programs exist and
no response packet was received.

Troubleshooting and Diagnostics 4-29

>>> set host/dup/dssi 1
Starting DUP

server...

DSSI Node 1

(R3VBNC)

DRVEXR V1.0

D 21-FEB-1988 21:27:54

DRVTST V1.0

D 21-FEB-1988 21:27:54

HISTRY V1.0

D 21-FEB-1988 21:27:54

V1.0

D 21-FEB-1988 21:27:54

PARAMS V1.0

ERASE

D 21-FEB-1988 21:27:54

DIRECT V1.0

D 21-FEB-1988

21:27:54

End of directory
Task Name? drvtst

Write/read anywhere on medium?
5 minutes

[l=Yes/ (0=No)]:

<CR>

for test to complete.

Compare failed on head
Compare failed on head

1 track 1091.
O track
529.

Task Name? drvexr

Write/read anywhere on medium?
Test time in minutes?

[l-Yes/(O—No)]

<CR>

[(10)-100]:

10 minutes for test to complete.
R3VEBNC: :MSCPSDUP 21-FEB-1988 21: 37:35 DRVEXR CPU=00: 00:01.88 PI=43
R3VBNC: :MSCPSDUP 21-FEB-1988 21: 37:38 DRVEXR CPU=00: 00:03.38 PI=79
Compare failed on head

1 track 1091.

R3VBNC: :MSCP$DUP 21-FEB-1988 21: 37:40 DRVEXR CPU=00: 00:04.97 PI=116
~C
>>>

In the above example, the local programs DRVTST and DRVEXR are
run on drive 1. Do not enter 1 in response to the question Write/read
anywhere on medium?. Doing so will destroy data on the disk. Enter
<CR>, which uses the default, allowing reads and writes to the DBNs

only.
or
displays a message as shown in the DRVEXR
example above (the lines beginning with R3VBNC::). In the example,

has been pressed twice, to show the difference in the time and in
the value of the progress indicator (PI).
Press

to terminate the program.

Use the local programs PARAMS (Section 4.8.5) and HISTRY
(Section 4.8.3) to determine the cause of errors displayed during
DRVTST or DRVEXR. DRVTST should run successfully for one pass
on each drive. Field Service can refer to the RF30 Disk Drive Service
Manual for details about the DUP local programs and corrective action.
If there are one or more DELQA modules in the system, use test 82 to

invoke the Ethernet option’s self-test and receive status from the host
firmware. Test 82 is useful for acceptance testing if you cannot access
the system enclosure to see the DELQA LEDs.

4-30

KA640 CPU System Maintenance

After the above steps have completed successfully, load MDM and run
the system tests from the Main Menu. If they run successfully, 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-based
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. Run the FE utility
if the message, Normal operation not possible, is displayed after the tests
have completed and there is no other error indication, or if you need more
information than what is provided in the error display.
The FE utility dumps diagnostic state to the console (Example 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 hardware error summary
register.

Example 4-5:
>>>

FE Utility Example

bitmap=00BF3400,

length=0C00,

checksum=007E

busmap=00BF8000
return stack=201406A8
subtest

pc=2004F4C4

timeout=00000001,

error=0B,

deerrorvector=18,

de_error=FE

severitycode=02,

previouserror=FEOB5D5D,

00000000,

total_errorcount=0001

00000000, 00000000,

00000000

last_exception pc=20050807
flags=01FFFD7F,

test flags=20

highest severity=00
leddisplay=05
consoledisplay=5D

save_mchkcode=80,

saveerrflags=000000

param1=00000100

2=00000100

3=000000F7

4=00000000

5=00000001

param 6=00000004

7=20050527

8=00000000

9=20140698

10=200521F4

Troubleshooting and Diagnostics

4-31

The most useful fields displayed above are as follows:

| De__error__vector, which is the SCB vector through which the unexpected
interrupt or exception trapped if de_error equals FE or EF.

Total_error_count. Four hex digits showing the number of previous
errors that have occurred.
-

Previous_error. Contains the history of the last four errors. Each
longword contains four bytes of information. From left to right these
are the de_error, subtest_log, and test number (copied in both bytes).

Save machine check code (save_mchk_code). Valid only if the test halts
on error. This field has the same format as the hardware error summary
register.
Save error flags (save_err_flags). Valid only if the test halts on error.
This field has the same format as the hardware error summary register.
Parameters 1 through 10. Valid only if the test halts on error. The
parameters have the same format as the hardware error summary
register.

EF in the previous_error field indicates that an unexpected exception has
occurred. If any of the tests that announce to the console fail, and the
error code is EF, examine the last longword of the error printout. The last
longword is the hardware error summary register and contains the machine
check code (<31:24>) and KA640 error status bits (<23:0>). Table 4-10 lists
the status bits.

4-32

KA640 CPU System Maintenance

Table 4-10: Hardware Error Summary Register

Bit

Machine check code

‘Machine check code

MSER <6>

Description

; CDAL data parity error

MSER <5>

; Mchn chck CDAL parity error

MSER <4>

; Machine check cache parity

MSER <1>

; Cache data parity error

MSER <0>

; Cache tag parity error

Unused

MEMCSR16 <31>

; Uncorrectable ECC error

MEMCSR16 <30>

: Two or more uncorrectable errors

MEMCSR16 <29>

; Correctable single-bit error

MEMCSR16 <25>

; Page address bits 25:22 of

: Location that caused error.

MEMCSR16 <24>

MEMCSR16 <23>

; These four bits point to the

MEMCSR16 <22>

; failing 4-Mbyte bank of memory.

MEMCSR16 <8>

: DMA read/write error.

MEMCSR16 <7>

; CDAL parity error on write.

CBCTR <31>

CBCTR <30>

DSER <7>

; Q22-bus NXM.

DSER <5>

; Q22-bus parity error.

DSER <4>

; Read main memory error.

DSER <3>

; Lost error.

DSER <2>

; No grant timeout.

IPCRn <15>

; DMA Q22-bus memory error.

;: CDAL bus timeout.

: CPU read/write bus timeout.

Troubleshooting and Diagnostics

4-33

4.5.2 Isolating Memory Failures

" This section describes procedures for isolating memory subsystem failures,

particularly when the system contains more than one MS650 memory

module.

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 [CTRUC] to terminate
the script after the completion of the current test.
on the
KA650 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.
2.

TAS
>>> T

[memory test]

starting
board number ending board number

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 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: >>>
T 4F 2 2

If a failure is detected for a second of three MS650 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 boundary, type:
>>> T 4F

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

4-34

KA640 CPU System Maintenance

location in memory since it can do so in a reasonable amount of time. Other
tests, such as 4F (floating ones and zeros test) can take up to one hour,
depending on the amount of memory, if each location were to be tested. If
you do specify an address increment, do not input less than 200 (testing
on a page boundary), since almost all hard memory failures span at least
one page. For normal servicing, do not specify the address increment, since
it adds unnecessary repair time; most memory subsystem failures can be
found using the default 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 I/O mode prompt (Example 4-6). Utility
9C is also useful for examining the error registers MSER, CACR, DSER,
and MEMCSRI16, upon a fatal system crash or similar event.
4.

T40

Although the SHOW/MEMORY command displays pages that aremarked 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 alternative
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 soft errors.
To allow 0 (zero) errors, enter the following:
>>>T 40140

This command tests the memory on the CPU and the three 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 9C

>>>T

TOY

=00034283

ICCs =00000000

TCRO =00000000

TIRO =00000000

TNIR0=00000000

TIVR0=00000078

TCR1 =00000000

TIR1 =00000000

TNIR1=00000000

TIVR1=0000007C

TXCS =00000000

TXDB =00000030

RXCS

=00000000

RXDB =0000000D

MSER

=00000000

CADR =0000000C

BDR

=FFFFFFDO

DLEDR=0000000C

SSCCR=00D45033

CBTCR=00000004

SCR

=0000C000

DSER =00000000

QBEAR=00000000

DEAR =00000000

QBMBR=00BF8000

MEM FRU

IPCRn=0000

MEMCSR 0=80000015

1=00000015

2=00000015

MEM FRU 2

MEMCSR 4=80400016

5=80800016

6=00000016

7=00000016

8=00000000

10=00000000

11=00000000

13=00000000 14=00000000

15=00000000

MEM FRU 3 MEMCSR 8=00000000
MEM FRU 4 MEMCSR12=00000000

MEMCSR16=8094000F
Ethernet

SII

SA =

3=00000015

17=00002000

08-00-2B-08-E8-6E

NICSR0=0004

MSIDRO =0l1FF

MSIDR1l

=0002

MSIDR2

=0000

MSICSR =0010

MSIIDR =8007

MSITR

=0000

MSITLP

=0000

MSIILP =0000

MSIDSCR=80FF

MSIDSSR=8500

MSIDCR =0008

>>>

In this case, 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, indicating 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, which is the
first MS650-AA module (KA640-AA is the first 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
Number) match.

MEMCSRI16,

since

bits

<25:22>
|

(Bank

The Bank Enable bit <31> in MEMCSR_5 is set, indicating that the

bank number is valid.

KA640 CPU System Maintenance

Example 4-7:

9C—Conditions for Determining a Memory FRU
3

MEMCSR16 = 8094000F Hex = 1000 0000 1001 0100 0000 0000 0000 1111
It 1l

5 = 80800016 Hex = 1000
MEMCSR

0000 1000 0000 OOOO 0000 00C1 0110
A

bit 31

set

25:22 match

4.5.3 Additional Troubleshooting Suggestions
Note the following additional suggestions when diagnosing a possible

memory failure.

If more than one memory module is failing, you should suspect the KA640
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
KA640 or 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 singleand multi-bit ECC failures as shown in step 2 of acceptance testing, since
in one slot a board may fail most frequently with multi-bit 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 KA640 ROM-based
diagnostics to see if it is an MS650 problem, or if it is related to the
KA640, 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.

You can run a DSSI device power-up diagnostic without performing a cold
restart or spinning the disk drives down and back up.

Troubleshooting and Diagnostics

4-37

Type the following at the console program:
>>>T 58 <NODE_NUMBER>

A CI Reset command is issued to the DSSI device, causing the device to
perform its power-up diagnostics.
Parameter-1 is the DSSI node or port number. It must be in the range of
0-7 (0 is the default). Use the default for parameter 2.

You can perform this test repeatedly with the REPEAT command (R T 58
<node_number>). In that case the drive’s self-tests run repeatedly until
you press
to terminate the test.
Once the test has completed successfully, you can examine the DSSI
device’s internal error logs by running the DUP local programs HISTRY and
PARAMS. Refer to Section 4.8.3 and Section 4.8.5 for further information.

4.6 Loopback Tests
You can use external loopback tests to localize problems with the Ethernet,
console, and DSSI subsystems.
|

Check that dc power and pico fuses on the KA640 are functioning correctly.
Three 1.5 A pico fuses (12-10929-08) are located near the handle on the
component side. The fuses are numbered from left to right as F3, F1, and
F2 and are shown in Figure 1-1. Replace the fuse, not the KA640, if a fuse
has gone bad. Table 4-11 lists some symptoms of faulty fuses.

Table 4-11:

KA640 Fuses

Bad Fuse

vSymptom

F1bad (+3V)

H3602-SA hex LED display is off.

F2bad (+12V)

Both Thinwire and standard Ethernet LEDs are off on the H3602-SA.

Ethérnet external loopback test 5F fails (for ThinWire only, since the fuse
protects +12 V supplied to the DESTA on the H3602-SA).

The LED on the loopback connector (12-22196-02) for standard Ethernet is
off; external loopback tests for standard Ethernet pass, however.
Console SLU external loopback test fails.
F3 bad (+53V)

LED on DSSI bus terminator or external loopback connectors is off.

Only local DSSI node (typically node 7 for the KA640) is reported by SHOW
DEVICE or SHOW DSSI commands.

DSSI external loopback test
56 fails.

4-38

KA640 CPU System Maintenance

DSSI! Problems

For DSSI problems, run the SII external loopback test (test 56). To check
the DSSI bus out to the KA640 connector, plug one end of the cable (17—
02216-01) to the H3281 loopback connector and the other end to the KA640
DSSI connector. To test out to the end of the DSSI bus, power down the
system, remove all DSSI devices with the exception of the KA640 from the
bus, and plug the external DSSI loopback connector in place of the DSSI
bus terminator.
Ethernet Problems
For ThinWire Ethernet problems, run the external loopback test (NI test,
number 5F) by entering the following:
>>> T

1<CR>

Set parameter 1 to run this test. Only the external loopback test runs. Be
sure to set the ThinWire/standard Ethernet switch on the H3602-SA to the
ThinWire position. Use two 50-ohm H8225 terminators connected to an
H8223 T-connector.
To test the standard Ethernet connector, use loopback connector 12-22196—
02 in conjunction with MDM.

4.6.1 Testing the Console Port
To test the console port at power-up, set the power mode switch on the
H3602-SA to the test position, and install an H3103 loopback connector
into the MMP
of the H3602. 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 insert 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 KA640,
the H3602-SA, the cabling, or the CPU module.

Troubleshooting and Diagnostics

4--39

To test out to the end of the console terminal cable:

Plug the MMJ end of the console terminal cable into the H3602—-SA.

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.

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

The following modules do not have LED self-test indicators:
DFAOQ1
DPV11
DRQ3B
DZQ11
KLESI
LPV1l
TSV05

The following modules have one green LED, which indicates that the
module 1s receiving +5 and +12 Vdc:
CXA16
CXB16
CXY08

Table 4-12 lists loopback connectors for common KA640 system modules.
Refer to Microsystems Options for a description of specific module seli-tests.

4-40

KA640 CPU System Maintenance

Table 4-12:
Device
CXA16/CXB16

Loopback Connectors for Q22-Bus Devices
Module Loopback

Cable Loopback

H3103 + H8572!

CXY08

H3046 (50-pin)

DELQA

12-22196-02

H3197 (25-pin)

DPV11

H3259

H3260

DSSI?
DzQu
Ethernet?

12-15336-00 or H325
~

H329(12-27351-01)
-

LPVi1

None

KA640/H3602-SA

H3103

KMVi1A

H3255

None

H3103 + H8572
H3251

1Use the appropriate cable to connect transmit-to-receive lines. H3101 and H3103 are doubleended cable connectors.

2For DSSI to KA640 or RF-series connector, use 17-02216-01 plus H3281 loopback.

For

connection to end of bus, use the DSSI loopback connector 12-30702-01.

3For ThinWire, use H8223-00 plus two H8225-00 terminators. For standard Ethernet, use
12-22196-02.

4.8 RF30 Troubleshooting and Diagnostics
An RF30 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 OCP on the enclosure front panel. The RF30 also has a red fault
LED, but it is not visible from the outside of the system enclosure.
If the drive is unable to execute the Power-On Self Test (POST) successfully,
the red fault LED remains lit and the ready LED does not come on, or both
LEDs remain on.

POST is also used to handle two types of error conditions in the drive:
1.

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.

- Troubleshooting and Diagnostics

4—41

Here are three common configuration errors:

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.

“Ifthe RF30 is connected to the OCP, you must install a unit ID plug in the

corresponding socket on the OCP. If the RF30 is not connected to the OCP,
the RF30 reads its unit ID from the three-switch DIP switch on the side of
the drive.

The RF30 contains the following local programs (described in the 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 RF30

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
dialog of each program. The table also indicates the type of messages
~ contained in the dialog, although the screen display will not indicate the
message type. Message types are abbreviated as follows:
Q—Question
I—Information
T—Termination
FE—Fatal error

Access these local programs using 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 dialog.

Once the connection is established, the local program is in control. When
the program terminates, control is returned to the KA640 console. To
abort or prematurely terminate a program and return control to the KA640
console, press [CTRUC] or [CTRUY].

- 4-42

KA640 CPU System Maintenance

4.8.1 DRVTST
DRVTST is a comprehensive functionality test. Errors detected by this test
are isolated to the FRU level. The messages are listed in Table 4-13.
Table 4-13:
Message

Type

DRVEXR Messages

Message

Write/read anywhere on the medium? [1=yes/(0=no)l

User data will be corrupted. Proceed? |1=yes/(0=no)]

5 minutes to complete.

Test passed.

FE
FE

Unit is currently in use.!
Operation aborted by user.

FE
FE

xxxx—Unit diagnostics failed.?
xxxx—Unit read/write test failed.?

1Either the drive is inoperative, in use by a host, or is currently running another local program.

ZRefer to the diagnostic error 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 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 IO routine.

After the timed I/0 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.

4.8.2 DRVEXR
The DRVEXR local program exercises the RF30 disk drive. The test is
data transfer intensive, and indicates the overall integrity of the device.
Table 4-14 lists the DRVEXR messages.

Troubleshooting and Diagnostics

4-43

Table 4-14:
Message

Type

DRVEXR Messages

Message

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).

Complete.

Or:

Unit is currently in use.*

Operation aborted by user.

FE
FE

xxxx—Unit diagnostics failed.?
xxxx—Unit read/write test failed.?

1Either the drive is inoperative, in use by a host, or
is currently running another local program.

ZRefer to the diagnostic error 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 results in the second question being asked.
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.
NOTE: If the write-protect switch on the OCP is pressed in (LED on) and
you answer Yes to the second question, the drive does not allow the test to
run. DRVEXR displays the error message 2006—Unit read [ write test fatled.

In this case, the test has not failed, but has been prevented from running.

4-44

KA640 CPU System Maintenance

DRVEXR saves the error counters, then calls the timed /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 1I/0 routine as DRVTST, with two exceptions.
First, DRVTST always uses a fixed time of five minutes, while you specify
the time of 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
RF30 disk drive. Table 4-15 lists the HISTRY messages.
Table 4-15:

HISTRY Messages

Message
Type

Field Length

Field Meaning

47 ASCII characters

4 ASCII characters

Product name

12 ASCII characters

Drive serial number

6 ASCII characters

Node name

1 ASCIHI character

Allocation class

8 ASCII characters

Firmware revision level

17 ASCII characters

Hardware revision level

6 ASCII characters

Power-on hours

5 ASCII characters

Power cycles

4 ASCII characters

Hexadecimal fault code
Complete

1Displays the last 11 fault codes as informational messages. Refer to the diagnostic error list
at the end of this chapter.

Troubleshooting and Diagnostics

4-45

The following example shows a typical screen display when you run
HISTRY:
Copvright © 1988 Digital Equipment Corporation
RF30

ENC1082
SUSAN

RFX V101

RF30 PCB-5/ECO-00
617
21

AQ4F
AQ4F

2103
AO4F

2404

AQ4F
A404

AO4F

2404
AO4F

2404

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.

446

KA640 CPU System Maintenance

Table 4-16 lists the ERASE messages.

Table 4-16:
Message

ERASE Messages

Message

Write/read anywhere on the medium? |1=yes/(0=no)]

User data will be corrupted. Proceed? [1=yes/(0=no)]

6 minutes to complete.

Complete.

FE
FE

Unit 1s currently in use.
'

Operation aborted by user.

xxxx—Unit diagnostics failed.?

xxxx—Operation failed.?

1Refer to the diagnostic error 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 failureis detected, the message indicating the failure will be followed
by one or more messages containing 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

Displays module configuration, history, or current counters, depending on the
status type chosen
,

WRITE

Alters the device parameters

Troubleshooting and Diagnostics

4-47

4.8.5.1 EXIT

Use the EXIT command to terminate the PARAMS local program.
4.8.5.2 HELP

Use the HELP command to display a bref list of available PARAMS
commands, as shown in the example below.
PARAMS> help
EXIT
HELP

SET

{parameter

SHOW {parameter

/ALL
/SERVO

/JCONST
/SCS

wvalue

/class}

/DRIVE
/MSCP

/DUP
STATUS

[type]

CONF'IG

LOGS

DATALINK

PATHS

WRITE

PARAMS>

4.8.5.3 SET

Use the SET command to change the value of a given parameter. Parameter
- 1s the name or abbreviation of the parameter to be changed. To abbreviate,
use the first matching parameter without regard to uniqueness. Value is
the value assigned to the parameter.
|
For example, SET NODE SUSAN sets the NODENAME parameter to
SUSAN.
The following parameters are useful to Field Service:

ALLCLASS

The controller allocation class. The alldcation class should be set to match

FIVEDIME

True (1) if MSCP should support five connections with ten credits each. False

that of the host.

(0) if MSCP should support seven connections with seven credits each.

UNITNUM

The MSCP unit number.

FORCEUNI1

True (1) if the unit number should be taken from the DSSI ID. False (0) 1f 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 RF30x (where x

448

is a letter from A to H corresponding to the DSSI bus ID) instead of using the
NODENAME value. Faise (0) if NODENAME is to be used.

KA640 CPU System Maintenance

4.8.5.4 SHOW

Use the SHOW command to display the settings of a parameter or a class

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. Type is the optional

ASCII string that denotes the type of display desired. If you omit Type, ail
available status information is displayed. If present, it 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.

Unit failures are also displayed, if

applicable.

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
some of the hardware registers.

DATALINK

Displays the data link counters.

PATHS

stplays 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 in the available state with respect
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.

Troubleshooting and Diagnostics

4-49

4.9 Diagnostic Error Codes
Diagnostic error codes appear when you are running DRVTST, DRVEXR,
or PARAMS. Most of the error codes indicate a failure of the drive module.
The exceptions are listed below. The error codes are listed in Table 4-17. If
you see any error code other than those listed below, replace the module.

Table 4-17:

RF30 Diagnostic Error Codes

Code

Message

Meaning

2032/A032

Failed to see FLT goaway

FLT bit of the spindle control status register was
asserted for one of the following reasons:
1. Reference clock not present
2. Stuck rotor
3. 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

4-50

Could be either the module or the operator control

panel or both.

KA640 CPU System Maintenance

‘

Appendix A

Address Assignments
A.1 General Local Address Space Map
Table A-1 lists the VAX memory space.

Table A-1:

VAX Memory Space

Contents

Address Range

Local memory space (52 Mbytes)

0000 0000-033F FFFF

Reserved memory space (460 Mbytes)

0340 0000-1FFF FFFF

Address Assignments

A-1

Table A-2 lists the VAX input/output memory space.

Table A-2:

VAX Input/Output Space

Contents

Address Range

Local Q22-bus I/O space (8 Kbytes)

2000 0000-2000 1FFF

Reserved local 1/0 space (248 Kbytes)

2000 2000-2003 FFFF

Local ROM space—halt protected space (128 Kbytes)

2004 0000-2005 FFFF

Local ROM space—halt unprotected space (128 Kbytes)

2006 0000-2007 FFFF

Local register /O space (1.5 Mbytes)

2008 0000-201F FFFF

Reserved local I/O space (62.5 Mbytes)

2020 0000-23FF FFFF

Reserved local I/O space (64 Mbytes)

2400 0000-27FF FFFF

Reserved local I/0 space (64 Mbytes)

2800 0000-2BFF FFFF

Reserved local /O space (64 Mbytes)

2C00 0000-2FFF FFFF

Local Q22-bus memory space (4 Mbytes)

3000 0000-303F FFFF

Reserved local VO space (60 Mbytes)

3040 0000-33FF FFFF

Reserved local I/O space (64 Mbytes)

3400 0000-37FF FFFF

Reserved local I/O space (64 Mbytes)

3800 0000-3BFF FFFF

Reserved local /O space (64 Mbytes)

3C00 0000-3FFF FFFF

A.2 Detailed Local Address Space Map
Table A-3 describes the contents of the VAX memory space.

Table A-3:

VAX Memory Space

Contents

Address Range

Local memory space (up to 52 Mbytes)

0000 0000-033F FFFF

Q22-bus map—top 32 Kbytes of main memory
' Reserved memory space

A-2

KA640 CPU System Maintenance

0340 0000-1FFF FFFF

Table A—4 describes the contents of the VAX input/output memory space.

Table A—4:

VAX Input/Output Space

Contents

Address Range

Local Q22-bus /O Space

2000 0000-2000 1FF¥F

Reserved Q22-bus /O space

2000 0000-2000 0007

Q22-bus floating address space

2000 0008-2000 O7FF

User reserved Q22-bus address space

2000 0800-2000 OFFF

Reserved and Q22-bus fixed address space

2000 1000-2000 1F3F

Interprocessor communication register (arbiter)

2000 1F40

Reserved Q22-bus 1/O space

2000 1F42-2000 1FFF

Reserved Local /O Space

2000 2000-2003 FFFF

Local ROM Space

2004 0000-2007 FFFF

Local ROM protected space

2004 0000-2005 FFFF

MicroVAX system type register (in ROM)

2004 0004

Local ROM unprotected space

2006 0000-2007 FFFF

Local Register /O Space

2008 0000-201F FFFF

DMA system configuration register

2008 0000

DMA system error register

2008 0004

Q22-bus error address register

2008 0008

DMA error address register

2008 000C

Q22-bus map base register

2008 0010

Reserved local register /O space

2008 0014-2008 OOFF

Main memory configuration Reg 00

2008 0100

Main memory configuration Reg 01

2008 0104

Main memory configuration Reg 02

2008 0108

Main memory configuration Reg 03

2008 010C

Main memory configuration Reg 04

2008 0110

Main memory configuration Reg 05

2008 0114

Main memory configuration Reg 06

2008 0118

Main memory configuration Reg 07

2008 011C

Main memory configuration Reg 08

2008 0120

Main memory configuration Reg 09

2008 0124

Main memory configuration Reg 10

2008 0128

Main memory configuration Reg 11

2008 012C

Main memory configuration Reg 12

2008 0130

Main memory configuration Reg 13

2008 0134

Main memory configuration Reg 14

2008 0138

Main memory configuration Reg 15

2008 013C

Address Assignments

A-3

Table A-4 (Cont.):

VAX Input/Output Space

Contents

Address Range

Main memory error status register

2008 0140

Main memory control/diagnostic status register

2008 0144

Reserved local register 1/0 space

2008 0148-2008 3FFF

Reserved (1 copy of BDR)

2008 4000

Boot and diagnostic register

2008 4004

Reserved (126 copies of BDR)

2008 4008-2008 41FF

NI Station Address ROM

2008 4200-2008 427C

Reserved (4 copies of NISA ROM)

2008 4280-2008 43FF

NI Register Data Port

2008 4400

NI Register Address Port

2008 4404

64 copies of NIRDP, NIRAP (reserved)

2008 4408-2008 45FF

MSI Diagnostic Register 0

2008 4600

MSI Diagnostic Register 1

2008 4604

MSI Diagnostic Register 2

2008 4608

MSI Control and Status Register

2008 460C

MSI ID Register

2008 4610

Reserved MSI Register

2008 4614

Reserved MSI Register

2008 4618

MSI Timeout Register

2008 461C

Reserved MSI Register

2008 4620

Reserved MSI Register

2008 4624

Reserved MSI Register

2008 4628

Reserved MSI Register

2008 462C

Reserved MSI Register

2008 4630

Reserved MSI Register

2008 4634

Reserved MSI Register

2008 4638

MSI Long Target List Pointer

2008 463C

MSI Initiator List Pointer

2008 4640

MSI DSSI Control Register

2008 4644

MSI DSSI Status Register

2008 4648

Reserved MSI Register

2008 464C

Reserved MSI Register

2008 4650

MSI Diagnostic Control Register

2008 4654

MSI Clock Control Register

2008 4658

MSI Internal State Register 0

2008 465C

MSI Internal State Register 1

2008 4660

MSI Internal State Register 2

2008 4664

MSI Internal State Register 3

2008 4668

Reserved MSI Register

2008 466C

A-4

KA640 CPU System Maintenance

Table A4 (Cont.):

VAX Input/Output Space

Contents

Address Range

Reserved MSI Register

2008 4670

Reserved MSI Register

2008 4674

Reserved MSI Register

2008 4678

Reserved MSI Register

2008 467C

Reserved (4 copies of MSI reg block)

2008 4680-2008 47FF

Reserved Local Register I/0 Space

2008 4800-2008 7FFF

Q22-bus map registers

2008 8000-2008 FFFF

Reserved local register /O space

2009 0000-200F FFFF

MSI Buffer RAM

2010 0000-2011 FFFF

NI Buffer RAM

2012 0000-2013 FFFF

SSC base address register

2014 0000

SSC configuration register

2014 0010

CDAL bus timeout control register

Diagnostic LED register

Reserved local register VO space

2014 0020
2014 0030
2014 0034-2014 0068

Diagnostic registers

2014 006C-2014 OOFF

Timer O control register

2014 0100

Timer 0 interval register

2014 0104

Timer 0 next interval register

2014 0108

Timer O 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 1/0 space

2014 0120-2014 O12F

MSIDB address decode match register

2014 0130

MSIDB address decode mask register

2014 0134

Reserved local register I/O space

2014 0138-2014 O13F

LIOD address decode match register

2014 0140

LIOD address 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 /O space

2014 0800-201F FFFF

Reserved local 1/0 space

2020 0000-2FFF FFFF

Local Q22-bus memory space

3000 0000-303F FFFF

Reserved local register /O space

3040 0000-3FFF FFFF

Address Assignments

A-5

A.3 Internal Processor Registers
Table A-5 lists the internal processor registers implemented in the CVAX
CPU chip and the SSC.

Table A-5:

KAG640 IPRs

Dec

Hex

Mnemonic

Type

Location

Kernel Stack Pointer

Executive Stack Pointer

KSP

r/'w

CVAX

ESP

r/w

CVAX

Supervisor Stack Pointer

SSpP

r/w

CVAX

3
4

User Stack Pointer

USP

riw

CVAX

Interrupt Stack Pointer

ISP

r/w

CVAX

Reserved

CVAX

Reserved

CvAX

Reserved

PO Base Register

POB

r/w

CVAX

CVAX
CVAX

PO Length Register

POLR

r/w

P1 Base Register

P1BR

r/w

CVAX

P1 Length Register

P1LR

r/w

CVAX

System Base Register

SBR

r/w

CVAX

System Length Register

SLR

r/w

CVAX

Reserved

CVAX

Reserved

CVAX

Process Control Block Base

PCBB

r/w

System Control Block Base

SCBB

r/w

CVAX
CVAX

Interrupt Priority Level

IPL

r/w

CVAX

AST Level

ASTLVL

r/w

Software Interrupt Request

SIRR

CVAX

Software Interrupt Summary

SISR

r/w

CVAX

Reserved

A-6

KA640 CPU System Maintenance

CVAX

Table A-5 (Cont.):

KA640 IPRs

Dec

Name

Hex

Mnemonic

Type

Location

Reserved

Interval Clock Control Status

ICCS

riw

CVAX

25
26

Next Interval Count

NICR

CVAX

Interval Count

ICR

CVAX

Time-of-year Register

TOY

r/w

SSC

28
29
30

1C
1D
1E

Console Storage Receiver Status
CSRS!
Console Storage Receiver Data ~ CSRD!
Console Storage Transmitter
CSTS!

T/w
r
r/w

SSC
SSC
SSC

31
32
33

1F
20
21

Console Storage Transmitter Data CSDB!
Console Receiver Control Status
RXCS
Console Receiver Data Buffer
RXDB

w
r/w
r

SsC
SSC
SSC

Console Transmitter Control

r/w

SSC
SSC

CVAX

Status

TXCS

Status

Console Transmitter Data Buffer

TXDB

Translation Buffer Disable

TBDR

r/w

CVAX

Cache Disable

CADR

r/w

CVAX

Machine Check Error Summary

MCESR

r/w

CVAX

Memory System Error

MSER

r/w

Reserved

CVAX

Reserved

Console Saved PC

SAVPC

CVAX

Console Saved PSL

SAVPSL

Reserved

CVAX

Reserved

CVAX

Reserved

CVAX

Reserved

/O System Reset Register

CVAX
CVAX

CVAX

IORESET

CVAX

Address Assignments

A-7

A.4 Global Q22-Bus Address Space Map
Table A—6 lists the addresses and memory contents of the Q22-bus memory
space.

Table A-6:

Q22-Bus Memory Space

Contents

Address

Q22-bus memory space (octal)

0000 0000-1777 7777

Table A-7 lists the contents and addresses of the Q22-bus I/O space with
BBS7 asserted.

Table A-7:

Q22-Bus I/0 Space with BBS7 Asserted

Contents
Q22-bus /O space (Octal)

Address
|

1776 0000-1777 7777

Reserved Q22-bus I/O space

1776 0000-1776 0007

Q22-bus floating address space

1776 0010-1776 3777

User reserved Q22-bus address space

1776 4000-1776 7777

Reserved and Q22-bus fixed address space

1777 0000-1777 7477

Interprocessor communication register

1777 7500

Reserved Q22-bus 1/O space

1777 7502-1777 1777

A-8

KA640 CPU System Maintenance

Appendix

Related Documentation
The following documents contain information relating to MicroVAX or
MicroPDP-11 systems.
Document Title

Order Number

Modules

CXA16 Technical Manual

EK-CAB16-TM

CXYO08 Technical Manual

EK-CXY08-TM

DEQNA Ethernet User’s Guide

EK-DEQNA-UG

DHV11 Technical Manual

EK-DHV11-TM

DLV11-J User’s Guide

EK-DLV1J-UG

DMV 11 Synchronous Controller Technical Manual

EK~-DMV11-TM

DMV 11 Synchronous Controller User’s Guide

EK-DMV11-UG

DPV11 Synchronous Controller Technical Manual

EK-DPV11-TM

DPV11 Synchronous Controller User’s Guide

EK-DPV11-UG

DRV11—-J Interface User’s Manual

EK-DRV1J-UG

DRV11-WA General Purpose DMA User’s Guide

EK-DRVWA-UG

DZQ11 Asynchronous Multiplexer Technical Manual

EK-DZQ11-TM

DZQ11 Asynchronous Multiplexer User’s Guide

EK-DZQ11-UG

DZV11 Asynchronous Multiplexer Technical Manual

EK-DZV11-TM

DZV11 Asynchronous Multiplexer User’s Guide

"EK-DZV11i-UG

IEU11-A/IEQ11-A User’s Guide

EK-IEUQ1-UG

KA630-AA CPU Module User’s Guide

EK-KA630-UG

KA640-AA CPU Module User’s Guide

EK-KA640-UG

KA650-AA CPU Module User’s Guide

EK-KA650-UG

KDA50-Q CPU Module User's Guide

EK-KDA5Q-UG

KDJ11-B CPU Module User’s Guide

EK-KDJ1B-UG

KDJ11-D/S CPU Module User’s Guide

EK-KDJ1D-UG

KDF11-BA CPU Module User’s Guide

EK-KDFEB-UG

KMV11 Programmable Communications Controller User’s Guide

EK-KMV11-UG

KMV11 Programmable Communications Controller Technical

EK-KMV11-TM

Manual