Digital PDFs

DEC-11-OMGIA-B-D

November 1975

255 pages

Original

10.0MB

Document:	RSX-11M System Generation Manual Version 2
Order Number:	DEC-11-OMGIA-B-D
Revision:	0
Pages:	255
Original Filename:	http://bitsavers.org/pdf/dec/pdp11/rsx11m_s/RSX11M_V2/DEC-11-OMIGIA-B-D_SYSGEN_1975.pdf

OCR Text

RSX-11M
System Generation Manual
Order No. DEC-11-0MGIA-B-D

RSX-llM Version 2

digital equipment corporation · maynard. massachusetts

First Printing, November 1974
Revised, September 1975
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 described in this document is furnished under a license
and may only be used or copied in accordance to the terms of such
license.
Digital Equipment Corporation assumes no responsibility for the use
or reliability of its software on equipment that is not supplied by
Digital.

@ 1974, 1975 by Digital Equipment Corporation

The postage prepaid READER'S COMMENTS torm on the last page of this
document requests the user's critical evaluation to assist us in
preparing future documentation.
The following are trademarks of Digital Equipment Corporation:

DIGITAL
DEC
PDP
DECUS
UNIBUS
COMPUTER LABS
COMTEX
DDT
DECCOMM

DECsystem-IO
DECtape
DIBOL
EDU SYSTEM
FLIP CHIP
FOCAL
INDAC
LAB-8

MASS BUS
OMNIBUS
OS/8
PHA
RSTS
RSX
TYPESET-8
TYPESET-11

LIMITED RIGHTS LEGEND
Contract No.
Contractor or Subcontractor:

Digital Equipment Corporation

All the material contained herein is considered limited rights data
under such contract.

CONTENTS

Page
PREFACE

vii

0.1

MANUAL OBJECTIVES AND READER ASSUMPTIONS
STRUCTURE OF THE DOCUMENT
ASSOCIATED DOCUMENTS
THE BASELINE SYSTEM
USING THIS MANUAL TO ACQUIRE SYSTEM
GENERATION SKILLS

vii
vii
viii

INTRODUCTION TO SYSTEM GENERATION

1-1

1.1
1. 2
1. 2 .1

SYSTEM GENERATION OVERVIEW
FUNCTIONAL FLOW OF SYSTEM GENERATION
Configuration and Assembly of the
Executive - Phase 1
Setting Up Partitions, Building and
Installing Tasks - Phase 2
DIFFERENCE BETWEEN MAPPED AND UNMAPPED
SYSTEM

1-1
1-1

1-4

GETTING STARTED

2-1

2.1
2 .1.1
2.1.1.1
2.1.1.2
2 .1. 2
2.1.2.1
2.1.2.2
2.L2.3
2.1.2.4
2.1.2.5

CONTENTS OF THE DISTRIBUTION KITS
RK Release Kits
RK Disk Release Kit Contents - Summary
RK Magtape Release Kit Contents - Summary
RK Release Contents - Details
Executive Source Disk Contents
Unmapped Object Disk Contents
Mapped Object Disk Contents
MCR, FCP, and DMP Source Disk Contents
Device Combinations Supported by RK
Distribution Kits
RP Magtape Release Kit
RP Magtape Release Contents
PRESERVING DISTRIBUTION DISKS
Preserving the RK Distribution Disks
Preserving from an RK05 to DECtape
Preserving from an RK05 to Cassette
Preserving from an RK05 to an RKOS
Preserving from an RK05 to Floppy Disk
Preserving from an RK05 to Magtape
Preserving the RP Distribution Kit
Restoring Damaged Volumes
Cassette Restoration
DECtape Restoration
Disk-to-Disk Restoration
Floppy Disk-to-RK Disk Restoration
Magtape-to-RK Disk Restoration
Exiting from PRESRV
Formatting Disks
BOOTSTRAPPING THE BASELINE SYSTEM

2-1
2-1
2-1
2-2
2-3
2-3
2-5

0.2
0.3
0.4
0.5
CHAPTER

1. 2. 2
1.3
CHAPTER

2 .1. 3

2.1.3.1

2.2
2.2.1
2.2.1.1
2.2.1.2
2.2.1.3
2.2.1.4
2.2.1.5
2.2.2
2.2.3
2.2.3.l
2.2.3.2
2.2.3.3
2.2.3.4
2.2.3.5
2.2.4
2.2.5
2.3

iii

Vl.11.

1-2
1-3

2-7
2-7
2-7
2-8
2-9
2-9
2-11
2-11
2-11
2-11
2-11
2-12
2-12
2-12
2-14
2-14
2-14
2-14

2-14
2-14
2-15

CONTENTS (Cont.)
Page
2-15
2-15
2-16
2-16

2.4

Start-up Using BM792-YB Bootstrap ROM
Start-up Using MRll-DB Bootstrap ROM
Start-up Using the BM873-YA Bootstrap ROM
Start-up Using the BM873-YB Bootstrap ROM
Start-up Using the M9301-YA or M9301-YB
Bootstrap ROM
CONSOLE DISPLAY

SYSGEN DIALOGUE DETAILS

3-1

3.1
3.2
3.3
3.4
3.5

SYSGEN DIALOGUE
SYSGEN.CMD FILE DETAILS
SGN QUERY LIST
SYSGEN2.CMD FILE DETAILS
SGN ERROR MESSAGES

3-1
3-2
3-14
3-57
3-82

GENERATING LARGER SYSTEMS

4-1

4.1
4 .1.1
4 .1. 2
4 .1. 3
4 .1. 4
4.2

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

4.8
4.9
4.9.l
4.9.2
4.9.2.1
4.9.2.2
4.9.2.3
4.9.3
4.9.4
4.9.5

GENERATING NON-MINIMUM SYSTEMS
Unmapped System - Executive Less Than BK
Unmapped System - Executive Greater Than BK
Mapped Systems
System-Controlled Partitions
BUILDING SYSTEMS FOR RESIDENCE ON OTHER
DEVICES
SUPPORTING MORE THAN ONE SYSTEM ON THE
SAME VOLUME
INCREASING FILE SYSTEM THROUGHPUT
Building a Faster File System
File Control Primitives Dynamic Memory
Requirements
INSTALLING AND USING RESIDENT LIBRARIES
AND COMMON BLOCKS
Setting Up a Common Partition
Installing a Library or Common Block
Eliminating a Library or Common Block
BUILDING A FORTRAN OTS MEMORY-RESIDENT
LIBRARY
BUILDING AN FCS POSITION-INDEPENDENT
MEMORY-RESIDENT LIBRARY
PRINT SPOOLER (PRT)
BUILDING MULTI-USER SYSTEMS
Setting Up a Multi-User System
Installing and Naming Tasks
Fixed Partition Systems
System-Controlled Partitions
Task Invocation
Memory Compaction
Round-Robin Scheduler
Checkpointing During Terminal Input Wait

VIRTUAL MONITOR CONSOLE ROUTINE (VMR)

5-1

5.0
5.1
5.2
5.3
5.3.1

INTRODUCTION
INITIATING VMR
VMR FILE SPECIFIERS
COMMAND SUMMARY
Command Syntax

5-1
5-1
5-3
5-4
5-5

2.3.l
2.3.2
2.3.3
2.3.4
2.3.5

CHAPTER

4.3
4.4
4.4.l
4.4.2
4.5
4.5.1
4.5.2
4.5.3
4.6
4.7

CHAPTER

2-17
2-18

4-8
4-9
4-10
4-10

4-11
4-12
4-12
4-12
4-12
4-12
4-14
4-15
4-16
4-16
4-16
4-17
4-17
4-17
4-18
4-18
4-18

CONTENTS

(Cont.)
Page

APPENDIX

5.3.2
5.3.3
5.4
5.4.1
5.4.2
5.4.3
5.4.4
5.4.5
5.5

Keywords
Comments
VMR COMMANDS
Command Description Format
Inibialization Commands
Informational Commands
Task Control Commands
System Maintenance Command
VMR ERROR MESSAGE SUMMARY

5-6
5-6
5-6
5-7
5-8
5-25
5-29
5-41
5-43

PRES RV

A-1
A-1
A-2
A-3
A-4
A-4

A.5.2
A.5.2.1
A.5.2.2

INTRODUCTION TO PRESRV
PRESRV SERVICES
PRESRV File Command Strings
PRESRV Copy Switches
PRESRV Command Switches
PRESRV FILE COMMAND STRING SWITCHES DETAILED DESCRIPTION
Device Specific Switches
TUlO 7-Channel Tape Switches
Phase-Encoded Tape Switch (/PE)
Format Disk (/FO)
Administrative Switches
Abort on Hard I/O Error (/ER)
Rewind at Completion (/RW)
Verify Switch (/VE)
Formatting Switches
Blocking Switch (/BL:nnn)
Files-11 Format Switch (/FI)
Logical Tape Format Switch (/TP)
Image Mode Switch (/IM}
PRESRV COMMAND SWITCHES - DETAILED
DESCRIPTION
Hardware-Specific Switches
Set Control Status Register Address
Set Physical Unit Number
Set Vector Address
Set TM02 Unit Number
Stand-Alone Switches
Help Switch
Listing Switch
Time Switch
Boot Switch
PRESRV OPERATOR INTERFACE
Bootstrapping PRESRV
Computing Required Blocks for Buffering
Booting Another System or Stand-Alone
Program
Operator Messages
Action Messages
Error Messages

SYSTEM CONVENTIONS

B-1

B.l

RSX-llM UIC CONVENTIONS

B-1

.h. • .J.

A. 2
A.2.1

A.2.2
A. 2. 3
A. 3

A.3.1
A.3.1.1
A.3.1.2
7\

£"1 •

..J •

.L •

..J

A. 3. 2
A.3.2.1
A.3.2.2
A.3.2.3
A. 3. 3
A.3.3.l
A.3.3.2
A.3.3.3
A.3.3.4
A. 4

A.4.1
A.4.1.1
A. L l. 2
A.4.1.3
A.4.1.4
A. 4. 2
A.4.2.1
A.4.2.2
A.4.2.3
A.4.2.4
A. 5
A.5.1
A.5.1.1
A.5.1.2

APPENDIX

A-5
A-5
A-5
A-6
A-6
A-7
A-7
A-7
A-7
A-8
A-9
A-10
A-10
A-11
A-11
A-11

A-12
A-12

A-12
A-13
A-13
A-13

A-13
A-13

A-14
A-14

A-14
A-15
A-17
A-17
A-17
A-17

CONTENTS (Cont.)
Page

APPENDIX

APPENDIX
APPENDIX

B.1.1
B. l. 2
B.2

Group Number
Member Number
NAMING CONVENTIONS

B-1
B-2
B-3

RSX-llM OPERATION

C-1

C.l
C.2
C.3
C.3.1
C.3.2
C.4
C.5

BOOTING A WRITE-LOCKED DISK
RUNNING OUT OF DYNAMIC MEMORY
DYNAMIC MEMORY REQUIREMENTS
Sources Requiring Dynamic Memory
Space Saving Hints
PARITY MEMORY SUPPORT
SAVING THE SYSTEM WITH A KWll-Y WATCHDOG
TIMER

C-1
C-1
C-2
C-2
C-3
C-4

ZAP UTILITY PROGRAM

D-1

D.l
D.2
D.3
D.3.1
D.3.2
D.3.3
D.4
D.4.1
D.4.2
D.4.3
D.5

INTRODUCTION TO ZAP
INITIATING ZAP
FILE SPECIFIERS
Format of File Specifier
Defaults in File Specifiers
Command Switches
COMMAND SYNTAX
Overview
Specification of Addresses
Commands
ZAP ERROR MESSAGES

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

BAD BLOCK LOCATOR PROGRAM

E-1

E.l
E.2

FUNCTIONAL DESCRIPTION
BAD BLOCK ERROR MESSAGES

E-1
E-2

SYSTEM DATA STRUCTURES AND SYMBOLIC
DEFINITIONS

F-1

EXAMPLE SYSGEN

G-1

C-4

INDEX-1

INDEX

FIGURES
FIGURE

2-1
2-2

4-1
4-2

Baseline RK Unmapped System Partition
Layout
Baseline RK Mapped System Partition
Layout
Typical Unmapped 24K System
Typical Unmapped 28K System

2-4

2-6
4-3
4-4

TABLES
TABLE

A-1
B-1
D-2

Default Blocking Factors
System Identification Codes

ZAP Commands

A-9
B-4
D-5

PREFACE

0.1

MANUAL OBJECTIVES AND READER ASSUMPTIONS

System generation is a fairly complicated process.
To understand
fully what is taking place requires that the user be knowledgeable in
the operation of RSX-llM and its component tasks.
Refer to section
0.3 below for further information regarding related manuals and
recommended reading.
CAUTION

No user should undertake
a
system
generation without adequate preplanning.
This includes a thorough study of the
SGN program queries in Chapter 3 and the
answers the user will provide, so that
the Executive options, processor options
and peripheral complement are
those
describing his intended target system.
In addition, a preliminary layout of
memory partitions is essentiai if the
later phases of the system generation
process are to proceed smoothly.

0.2

STRUCTURE OF THE DOCUMENT

This manual provides information on how to generate an RSX-llM system
tailored to a particular installation's hardware configuration. As
delivered, RSX-llM includes an operational Executive, called the
Baseline System.
The Baseline System is described in section 0.4
below and is initiated by following the procedures discussed in
section 2.3.
Chapter 1 contains an overview of the system generation process.
It
is intended to familiarize the reader with the various phases involved
in producing a tailored system.
Chapter 2 describes the contents of the distribution kit and
information on getting started.

provides

The actual procedures involved in system generation are provided in
Chapter 3.
The step-by-step process followed in generating a system
is described.
vii

Chapter
4
covers
procedures
multi-function systems.

needed

for

generating

larger

Chapter 5 describes the commands and operation of Virtual MCR (VMR}, a
program which is used in the final stage of system generation to
initialize the user's target system.
The remaining sections comprise appendices describing the Volume
Preservation Program (PRESRV), UIC conventions, RSX-llM operational
hints, the File Patch Tility (ZAP), RSX-llM control blocks, and a
sample system generation.

0.3

ASSOCIATED DOCUMENTS

Other manuals closely allied to the purposes of this document are
described briefly in the RSX-llM/RSX-llS Documentation Directory,
Order No. DEC-11-0MUGA-B-D. The Documentation Directory defines the
intended readership of each manual in the RSX-llM/RSX-llS set and
provides a brief synopsis of each manual's contents.

0.4

THE BASELINE SYSTEM

RSX-llM efficiently services a wide range of
PDP-11
hardware
configurations.
Tailoring a system to a local set of hardware and
performance requirements is accomplished through the system generation
process.
The tailoring process is run on an RSX-llM system called the Baseline
System.
The Baseline System is a fully operational RSX-llM system
built explicitly for generating target systems. It is not designed
for general use due to the small size of the dynamic memory pool. The
Baseline System contains the following Executive options and device
support:
Executive options in the Baseline System
Files-11 Ancillary Control Processor (FllACP)
Checkpointing
Nonresident task support
Address checking (mapped system only)
I/O rundown
External MCR functions
Task termination and device-not-ready messages
Get Partition Parameters directive
Get Task Parameters directive
Install, Request, and Remove on Exit directives support
Logical device assignment support
256 words of dynamic memory

viii

Processor options in the Baseline System
For unmapped RK disk systems:
16K of memory
For unmapped RP02, RP03, and RP04 systems:
24K of memory
For all mapped systems:
24K of memory
Memory management unit (KT-11)
Peripheral device support in the RK disk Baseline System
2 RKOS disk drives
2 TAll cassette drives
1 TU56 DECtape drive
1 RXOl floppy disk drive
1 TUlO or TS03 mag tape drive

1 TU16 mag tape drive
1 LPll or LSll line printer
1 terminal

Peripheral device support for the RP disk Baseline System
1 RP02, RP03 or RP04 disk drive
1 TUlO or TS03 magtape drive
1 TU16 magtape drive
1 LPll or LSll line printer
1 terminal

0.5

USING THIS MANUAL TO ACQUIRE SYSTEM GENERATION SKILLS

RSX-llM system generation, being based on conditional assemblies,
affords the user a vast number of configuration options. Most users,
however, will not have a need to exploit all the flexibility inherent
in the RSX-llM system generation process. To assist the user in his
option selection, the entire system generation process is under the
direction of two indirect Monitor Console Routine (MCR) command files.
Learning how to generate a system is best achieved by active
experience with the indirect command files. It is suggested that the
user read through the manual (excluding appendices) before attempting
a system generation for his target system.

CHAPTER 1
INTRODUCTION TO SYSTEM GENERATION

1.1

SYSTEM GENERATION OVERVIEW

The object of the system generation procedure (SYSGEN) is to create an
RSX-llM system tailored to the user's local hardware configuration and
performance requirements. This chapter provides an outline of the
stages
involved
in
performing system generation.
The system
generation procedures are contained in Chapter 3.
The system generation process implies variability
categories of choices are available:
1.

Executive service options,

Processor options,

Peripheral devices,

Partition configurations, and

Member tasks.

A system is tailored to local needs by selecting
five categories.

1.2

and

choice.

options

from

Five

these

FUNCTIONAL FLOW OF SYSTEM GENERATION

The system generation process (SYSGEN) can be viewed as having two
milestones.
The particular set of user actions and system output in
achieving a milestone is considered a phase of SYSGEN. The milestones
are:
1.

Configuration and assembly of the Executive, and

Building the Executive, privileged tasks, and non-privileged
tasks;
setting
up
partitions, installing tasks, and
bootstrapping and saving the system.

The entire SYSGEN process, of which these phases are components, is
directed by two indirect MCR command files. The file SYSGEN.CMD
directs the first phase; SYSGEN2.CMD directs the second phase.

1-1

INTRODUCTION TO SYSTEM GENERATION
1.2.1

Configuration and Assembly of the Executive - Phase 1

Using the Baseline Executive, which exists on the distribution disks,
the user conducts this phase by answering interactive queries posed by
the System Generation Configuration Query Program (SGN).
Answering
these questions determines the:
Executive service options,
Processor options, and
Peripheral complement
that will be incorporated into the Executive being created to
the user's configuration.

service

NOTE
The SGN task is the means for specifying
all Executive options, processor options
and peripherals to be generated into the
target system. Thus, it is necessary to
thoroughly read Chapter 3
and
the
RSX-llM Executive Reference Manual to
know what and how to specify the options
required in the target Executive. It is
during SGN that the
Processor
and
Executive options are determined; also
determined are the number and type of
disks, tapes, printers, etc., that are
to be generated into the target system.
User interaction with the SGN program results in the creation of
files on the disk:
1.

RSXMC.MAC

SYSTB.MAC

RSXASM.CMD

RSXBLD.CMD

ICTAB.MAC

RSXMC.MAC contains the conditional assembly
produce the user-tailored Executive.

parameters

necessary

five

SYSTB.MAC is a file of device table descriptions needed to service the
peripheral configuration.
RSXASM.CMD is a MACR0-11 indirect command
assembly of the user-tailored Executive.

that

directs

the

RSXBLD.CMD is a Task Builder indirect command file which
building of the user-tailored Executive.

directs

the

1-2

file

INTRODUCTION TO SYSTEM GENERATION
ICTAB.MAC is a MACR0-11 source file which defines the module layout of
an ICS/ICR subsystem. It is generated only when the user selects the
ICS/ICR subsystem for inclusion in the target system.
After the queries posed by SGN have been answered and SGN has created
its five output files, the user-tailored Executive is assembled via
the indirect command file RSXASM.CMD.
Both the configuration and
assembly are directed by the indirect MCR command file SYSGEN.CMD.
Thus, Phase 1 is initiated by:
1.

Mounting the correct disk,

Bootstrapping the Baseline System,

Entering the time and date, and

Initiating the indirect MCR command file, SYSGEN.CMD.

1.2.2

Setting Up Partitions, Building and Installing Tasks - Phase 2

During this phase, the user-tailored Executive is
built,
the
privileged tasks are built, and, under certain conditions, some
non-privileged system tasks may also have to be built.
At the
completion of the build of the Executive, the user has an operational
Executive, and the value of the symbol $SYTOP is known.
The symbol $SYTOP represents the address of the last word in the
user-tailored Executive, enabling the user to determine the size of
the Executive. A complete operational system consists of:
1.

An Executive,

Privileged and non-privileged system tasks,

User tasks, and

A set of memory partitions into which tasks can be
and run.

installed

This phase is directed by the indirect MCR command file SYSGEN2.CMD,
which, once initiated, completely guides the user through the building
and installation of requisite tasks.
Having the value of $SYTOP, the user can lay out memory into the
partitions and subpartitions needed to meet the application and
performance requirements of his system. Also, the requirements of the
privileged and non-privileged system tasks can be met.
The building of privileged and non-privileged system tasks occurs
after the user has established his partition boundaries, since these
tasks ultimately execute in partitions defined by the user.
During this phase, an operational Executive was created and placed
into a file on disk.
Note that the system generation process was
being carried out under the Baseline System Executive.

1-3

INTRODUCTION TO SYSTEM GENERATION
Using Virtual MCR {VMR} and the disk image of the target system, the
user can now establish the partitions and install the tasks needed to
complete the operational system.
Having created the operational system, complete with partitions and
tasks, the user now must bootstrap the system, using the MCR Boot
command, and then save the system, using the MCR Save command with the
/WB switch appended.
MCR saves the system and then automatically
re-boots it.
The running of VMR to set partitions and to install tasks, and the
bootstrapping and saving of the system, are carried out under the
direction of the indirect command file named SYSGEN2.CMD.

1.3

DIFFERENCE BETWEEN MAPPED AND UNMAPPED SYSTEM

RSX-llM is designed to run on almost all models of the PDP-11
processor (11/04/05/10/15/20/35/40/45/50/70}.
The basic addressing
scheme of the processor instructions allows direct addressing of up to
32K words {64K bytes) of memory. In order to use memories larger than
32K words, the memory management unit {KT-11)
is available on the
PDP-11/35/40/45/50/70. This unit maps a virtual address (0-32K words)
into a physical address (0-124K words on all processors embodying the
KT-11, except the 11/70 which is expandable to 2048K words). An
unmapped system does not use the KT-11, whereas, the mapped system
does.
This difference has an impact on the building of tasks for
RSX-llM. On an unmapped system, tasks must be linked to the physical
base address of the partition in which they run. However, in a mapped
system, all tasks are mapped with the KT-11, and the virtual base
address of the task is always zero. If it is desired to run a task
built for an unmapped system in a partition with a different base
address, the task must be rebuilt to link it to the physical base
address of the new partition. This is not necessary on a mapped
system,
since
the physical base address of the partition is
transparent to the user because the KT-11 converts virtual addresses
to
physical addresses.
All tasks in a mapped system (except
privileged tasks) are built with a virtual base address of zero.
Thus, any task on a mapped system can be run in any partition large
enough for it, whether or not it is the partition for which the task
was built. This affects the SYSGEN procedure, since the alteration of
the partitions in an unmapped system may force the user to rebuild all
the utility tasks, whereas, in a mapped system, this will not happen.

1-4

CHAPTER 2
GETTING STARTED

2.1

CONTENTS OF THE DISTRIBUTION KITS

RSX-llM is distributed in three different kits:
1.

Four RK disks,

Two 600-foot, 800-BPI tapes of RK disk images, and

Four 600-foot, 800-BPI tapes containing an image of an
RP03, and RP04 disk, and a source tape.

2.1.1

RP02,

RK Release Kits

2.1.1.1

RK Disk Release Kit Contents - Summary

The RK disk distribution
identified as:

kit

contains

The Executive Source Disk,

The Unmapped Object Disk,

m\..~

.:J.

.LUC

Mapped Object T'I;

C'.' J,

.L..l .L.11;;1 n.

four

RK05

disk

cartridges

and

The MCR, FCP, and DMP Source Disk.

For a given generation type, mapped system, or unmapped system, only
two of these disks are used:
the Executive source disk and the
applicable object disk. Users building a mapped system must have a
hardware configuration with a minimum of 24K of memory and a Memory
Management Unit. The MCR, FCP, and DMP source disk is used if the
user wishes to modify MCR or FCP. DMP is supplied as an example of
how a typical MCR utility program is written.
The Executive source disk is used during the first phase of SYSGEN.
Among its tiles are:
1.

The Baseline System Executive,

The text for
Program (SGN) ,

the

System

2-1

Generation

Configuration

Query

GETT ING STARTED
3.

A subset of the system tasks,

Complete source code for the Executive,

The SGN task, and

The system
SYSGEN.CMD.

generation

indirect

MCR

command

Output files created during the first phase of system
generated on the source disk.

file

named

generation

are

The object disk (mapped or unmapped) is used for the final
SYSGEN and contains among its files:
1.

phases

The Baseline System Executive and system tasks;
The unmapped object disk contains a 16K unmapped Baseline
System and the object libraries for an unmapped MCR and FCP.
The mapped object disk contains a 24K mapped Baseline
and the object libraries for a mapped MCR and FCP.

System

Task build command files to create all system tasks;

Object library files for all system tasks;

The system generation indirect MCR command file, SYSGEN2.CMD.

and

In addition, the concatenated Executive object file, RSXllM.OBJ, the
Executive task build command file, RSXBLD.CMD, and the ICS/ICR table
module, ICTAB.MAC, created on the source disk during phase 1, are
copied to the object disk.
During the latter part of SYSGEN, the
user-tailored Executive is created on the object
disk,
which
ultimately becomes the user's system disk.
To obtain a detailed list of
commands:

the

contents

the

disks,

use

the

RUN $PIP
PIP>DKO: [*,*]/LI
This command sequence produces, via the console terminal,
directory listing of the volume mounted on DKO:.

2.1.1.2

complete

RK Magtape Release Kit Contents - Summary

The RK magtape release kit consists of two 600-foot, 9-track, 800-BPI
magtapes.
The magtapes contain four RK disk images. These four
images correspond to the four disks described in the previous section
on the RK disk release kit. The two magtapes are:
1.

Executive source and
files:

unmapped

object

tape

RSX11MBL12EX - Executive source disk image
RSX11MBL12UN - Unmapped object disk image

2-2

containing

the

GETTING STARTED
2.

Mapped object and MCR, FCP, and DMP
the files:

source

tape

containing

RSX11MBL12MP - Mapped object disk image
RSX11MBL12MF - MCR, FCP, and DMP source disk image
Before beginning a SYSGEN, the appropriate disks must be copied
RK cartridges using the Volume Preservation Program (PRESRV).
procedure is described in section 2.2.3.5.

2.1.2

onto
This

RK Release Contents - Details

2.1.2.1

Executive Source Disk Contents

The Executive source disk contains the following:
UIC
[ 1, 1]

[1,2]
[1,50]
[11,10]
[11;20]

[11,24]
[11,40]
[200,200]

CONTENTS
System and Executive MACRO libraries
Error message text file, SYSGEN text file, startup
command file
Baseline unmapped Executive and limited task complement
Executive source modules
Directory for unmapped Executive object modules
Directory for mapped Executive object modules
Directory for Executive correction files
SYSGEN part 1 command file

The RSX-llM system distributed on this disk has a
tasks provided.

limited

number

The tasks are:
BOO
DMO
IND
INI
INS

MCR
MOU
SAV
TKN
UFO
FCP
SGN
PIP
FLX
BI GMAC
MAC
EDI
SLP
VFY

- MCR Boot command
- MCR Dismount command
- MCR indirect command file processor
- MCR Disk Initialization command
- ~~~ Install command
- Monitor console routine
- MCR Mount command
- MCR Save command
- Task termination notification routine
- MCR User File Directory command
- 2K file sys tern
- System Generation Configuration Query Program
- Peripheral Interchange Program
- File Transfer Program
- 14K macro assembler
- 8K macro assembler
- Line Text Editor rogram
- Source Language Input Program
- File Structure Verf ication Program

The partition layout for the system on the source
Figure 2-1.

2-3

disk

shown

GETTING STARTED
22K

1300008
UNASSIGNED
(6K)

16K

1000008

PAR14K
Partition
(14K)
GEN
Subpartition
(8K)

SK
SYSPAR Partition
301008
DRVPAR Partition
253008

Executive

Figure 2-1
Baseline RK Unmapped System Partition Layout
This system will run on a machine with 16K of memory or more.
The
subpartition GEN is used for
the assembler on 16K systems, and
partition PAR14K is used for the assembler on systems with more than
16K of memory.
On 16K systems, it is not possible to produce both
object modules and listings during a single assembly of the Executive.
Two assemblies are required; one to produce the object modules, and
one to produce the listings. On systems with more than 16K of memory,
this restriction does not exist.
The PAR14K partition must not be used to run tasks on a
since it maps into 6K of non-existent memory.
Devices generated into this system are:
2 RKOS drives
1 TU56 DECtape drive
2 TAll cassette drives
1 RXOl floppy disk drive
1 TUlO or TS03 magtape drive
1 TU16 magtape drive
1 LPll or LSll line printer
1 terminal

2-4

16K

machine,

GETTING STARTED
2.1.2.2

Unmapped Object Disk Contents

The unmapped object disk contains the following:
CONTENTS

DIC
[ 1, 1]

System object library, Executive symbol definition
library, ODT and TRACE object modules, and concatenated
object module files for the FORTRAN object time system
(OTS)

[ l, 2]

Error message text file, startup command file

[1,3]

Directory for lost files found by VFY

[1,4]

Directory for Snap-Shot and Post-Mortem Dumps

[ l: 20]

Object module libraries, task build command files: and
overlay description files for all DEC-supported tasks,
including the unmapped versions of the privileged tasks

[1,30]

Directory for task maps

[1,50]

RSX-llM Baseline System Executive and all DEC-supplied
tasks built to run with the Baseline System Executive

[l,60]

Directory for RSX-llS Executive and tasks

[200,200] SYSGEN Part 2 command file and source modules for
building resident FORTRAN OTS and file control services
(FCS) libraries
The partition layout and device configuration
of
the
system
distributed on the unmapped object disk is identical to that of the
source disk. This disk is used to build a new, unmapped RSX-llM
system to user specifications.

2.1.2.3

Mapped Object Disk Contents

The mapped object disk
UIC

following:

CONTENTS

[ 1, 1]

System object library, Executive symbol definition
library, ODT and TRACE object modules, and concatented
object module files for the FORTRAN object time system
(OTS)

[ 1, 2]

Error message text file, startup command file

[1,3]

Directory for lost files found by VFY

[1,4]

Directory for Snap Shot and Post-Mortem Dumps

[1,20]

Object module libraries and overlay description
for all DEC-supported non-privileged tasks

2-5

files

GETTING STARTED
[1,24]

Object module libraries and overlay
for all DEC-supported privileged
command files for all tasks

description files
tasks, task build

[1,34]

Directory for task maps

[l,54]

RSXllM mapped 24K Baseline System Executive and
DEC-supplied tasks built to run with the Executive

[1,64]

Directory for RSX-llS Executive and tasks

all

[200,200] SYSGEN part 2 command file and source modules for
building resident FORTRAN OTS and file control services
(FCS) libraries.
The device configuration for this system is identical to that of the
system on the source disk. The partition layout appears in Figure
2-2.
24K

SYSPAR Partition
(2K)

22K

1400008
1300008

UNASSIGNED
{6K)
16K

1000008

PAR14K
Partition
{14K)
GEN
Subpartition
(SK)

BK
DRVPAR Partition

Executive

Figure 2-2
Baseline RK Mapped System Partition Layout
The minimum configuration supported by the mapped RSX-llM system is
24K words.
The partitioning scheme is such that all tasks except
BIGMAC and BIGTKB run in partition GEN. BIGMAC and BIGTKB are built
to run in partition PAR14K.

2-6

GETTING STARTED
2.1.2.4

MCR, FCP, and DMP Source Disk Contents

The MCR, FCP, and DMP source disk contains the following:
CONTENTS

UIC
[ 1, l]
[ 1, 2]

[1,50]
[12,10]
[12,20]
[12,24]
[12,30]
[12,34]
(13,10]
[13,20]
[13,24]
[13;30]

[13,34]
[35,10]
[35,20]
[35,30]

System and Executive macro libraries
Error message text file, startup command file
Baseline unmapped Executive and limited task complement
MCR source modules
Unmapped MCR assembly command file
Mapped MCR assembly command file
Directory for unmapped MCR listing files
Directory for mapped MCR listing files
FCP source modules
Unmapped FCP assembly command file
Mapped FCP assembly command file
Directory for unmapped FCP listing files
Directory for mapped FCP listing files
DMP source file
DMP assembly command file
Directory for DMP listing files

The RSX-llM system distributed on this disk is identical to the one on
the Executive source disk.

2.1.2.5

Device Combinations Supported by RK Distribution Kits

The following list contains the valid combinations of devices for use
in generating a target system.
The user must have one of the
following combinations to successfully complete a system generation
with the RK kits:
1.

RK05 disk and TAll cassette (working storage)

RK05 disk and TU56 DECtape (working storage)

3. RK05 disk and RXOl floppy disk (working storage)

2.1.3

RKOS disk and TUlO

RKOS disk and TU16 mag tape (working storage)

RK05 disk and TS03 mag tape (working storage)

Two RK05
mounted)

disks

'ft'\~N .... ~'I""\"UIQ':j'--Q.t"C

(source

'"-~" .. n.
lr.;._U'::::f
"'"",.
\WV._

and

storage)

object

disk

simultaneously

RP Magtape Release Kit

The RP magtape release kit consists of four magtapes.
One tape
contains the MCR, FCP, and DMP source files, and the other three tapes
contain (on separate tapes) RP02, RP03, and RP04 disk images in PRESRV
format.
The source tape is in FLX format and is needed only if the
user wishes to modify MCR or FCP or to list the DMP task for
examination.
Each RP disk image contains a running unmapped and
mapped Executive and all tasks.
The
unmapped
Executive
is

2-7

GETTING STARTED
bootstrapped in from a hardware bootstrap. In addition to the two
systems, the disks contain all of the files necessary to build a
target system.
The user may generate a mapped system while running
either the unmapped or mapped Executives or may generate an unmapped
system while running either the unmapped or mapped Executive. The
only difference between the three system disk images is that they have
the appropriate disk driver for the disk on which they are to run.

2.1.3.1

RP Magtape Release Contents

The disk images in the RP magtape release kit contain the following:
UIC

CONTENTS

[ 1, 1]

Executive and system macro libraries, system object
library, Executive symbol definition library, FORTRAN
OTS object modules, ODT and TRACE debugging aids

[1,2]

SYSGEN text file, system error
command file

[1,3]

Directory for lost files found by VFY

[ 1, 4]

Directory for Snap Shot and Post-Mortem Dumps

[l,20]

Unmapped
system
privileged
task
object
module
libraries, non-privileged task object module libraries,
and unmapped system task build command and overlay
descriptor files

[ 1, 24]

Mapped system privileged task object module libraries,
and mapped system task build command and overlay
descriptor files.

[l,30]

Directory for unmapped system task build maps

[l,34]

Directory for mapped system task build maps

[l,50]

RSX-llM unmapped Baseline System Executive and all
system-supplied tasks built to run with this Executive.

[ 1, 54]

RSX-llM mapped Baseline System Executive and
all
system-supplied tasks built to run with this Executive.

[l,60]

Directory for unmapped RSX-llS Executive and tasks

(1,64]

Directory for mapped RSX-llS Executive and tasks

[11,10]

Executive source files

[11,20]

Directory for unmapped Executive object files

[11,24]

Directory for mapped Executive object files

[11,30]

Directory for unmapped Executive listing files

[11,34]

Directory for mapped Executive listing files

[11,40]

Directory for Executive source correction files

2-8

message

file,

startup

GETTING STARTED
[12,10]

Directory for MCR source files

[12,20]

Unmapped MCR assembly command file and
build command file

object

library

[12,24]

Mapped MCR assembly

object

module

command

file

and

library build command file
[12,30]

Directory for unmapped MCR listing files

[12,34]

Directory for mapped MCR listing files

[13,10]

Dirctory for FCP source files

[13,20]

Unmapped FCP assembly command file

[13,24]

Mapped FCP assembly command file

[13,30]

Directory for unmapped FCP listing files

[13,34]

Directory for mapped FCP listing files

[35,10]

DMP source files

[35,20]

DMP assembly command file

[35,30]

Directory for DMP listing files

i200,200j SYSGEN

part 1 and part 2 command files, FORTRAN
resident library file and FCS resident library file.

The device configuration for this kit is:
1 RP02, RP03, or RP04 disk drive
l TUlO or TS03 magtape drive
1 TU16 magtape drive
1 LPll line printer
1 terminal
The partition layouts for the unmapped and mapped systems are
identical to those for the RK distribution disks. Both the mapped and
unmapped Executives are supplied with running copies of all the system
tasks.

2.2

2.2.1

PRESERVING DISTRIBUTION DISKS

Preserving the RK Distribution Disks

If the release kit is received on RK disk cartridges, the user is
strongly urged to back up the distribution disks prior to their actual
use in system generation. Since these disks are actively manipulated
during the generation of a user-tailored system, as well as during
system operation with the Baseline System Executive, inadvertent
corruption is possible during the user's initial learning experiences
with RSX-llM. To facilitate preservation of various storage media,
Digital provides a stand-alone utility called the Volume Preservation
Program (PRESRV). PRESRV is described in Appendix A. The user must

2-9

GETTING STARTED
always preserve the object disk on which he plans to generate his
system, since the target system will replace the Baseline System on
that disk. The system generation procedure is designed to run only on
the Baseline System. The source disks are not corrupted by the system
generation procedure;
however, it is good practice to preserve them
too.
Before preserving the RK release disks, it is advisable to verify that
they can be read on the user's disk drives by invoking VFY. All disks
contain the task VFY. One of the functions of VFY is to read all
allocated blocks on a disk to locate blocks which cannot be read; VFY
also determines the file to which each bad block is allocated.
After
bootstrapping a disk, run VFY and specify the /RC (read check) switch.
If any blocks are unreadable, the task will report them.
VFY is
described in the RSX-llM Utilities Procedures Manual. If the disk or
disks are unreadable, the user should immediately call the local
Digital software support specialist for assistance.
A minimum RK system must
complements:

contain

one

the

following

peripheral

An RKOS disk and a dual TU56 DECtape unit,

An RKOS disk and a dual TAll cassette tape unit,

Dual RKOS units,

An RK05 and an RXOl floppy disk, or

An RKOS disk and a TUlO, TU16, or TS03 magtape drive.

To preserve the disks in the distribution kit, use the bootstrapping
procedure described in section 2.3 to boot the unmapped object disk.
Then, using the MCR Boot command, bootstrap the Volume Preservation
Program (PRESRV). The full sequence is described below.
Bootstrap the system;

the following display results:

RSXllM V02 BL12 16K
>RED DKO:=SYO:
>MOU SYO:
>@STARTUP
>@<EOF>
Bootstrap PRESRV using the following MCR commands:
>INS [l,SO]BOO*
>BOO [l,SO]PRESRV*
PRESRV is bootstrapped and the following display occurs on the console
terminal:
RSXllS VO! BL12
VOLUME PRESERVATION PROGRAM - V02.l
FOR HELP TYPE /HE
nn. - BLOCKS AVAILABLE FOR BUFFERING
PRE>

*Contained on the RK mapped object disk under UIC [1,54].

2-10

GETTING STARTED
indicating that the PRESRV program is ready for input.
The message
"nn. - BLOCKS AVAILABLE FOR BUFFERING" indicates how many 256 word
blocks are available for buffering. Depending on the specific backup
medium being used (cassette, DECtape, RK05 disk, or magtape), select
one of the procedures described in sections 2.2.1.1 through 2.2.1.5.

2.2.1.1

Preserving from an RK05 to DECtape
(16K machines)
(all others)

PRE>DTO:label/BL:l/VE=DKO:/BL:l
PRE>DTO:label/BL:4/VE=DKO:/BL:4

where label is the name to be given to the file on DECtape.

2.2.1.2

Preserving from an RK05 to Cassette
(16K machines)
(all others)

PRE>CTO:label/BL:l/VE=DKO:/BL:l
PRE>DTO:label/BL:3/VE=DKO:/BL:3

where label is the name to be given to the file on the cassette.

2.2.1.3

Preserving from an RK05 to an RK05

PRE>DKl:/BL:l/VE/FO=DKO:/BL:l
PRE>DK1:/BL:l2/FO/VE=DKO:/BL:l2.

(16K machines)
(all others)

•

2.2.1.4

Preserving from an RKOS to Floppy Disk
(16K machines)
(all others)

PRE>DXO:label/BL:l/VE=DKO:/BL:l
pre>dxO:label/BL:6./VE=DKO/BL:l2.

2.2.1.5

Preserving from an RKOS to Magtape

The minimum RK magtape system is 24K.
PRE>MTO:label/BL:l2./VE=DKO:/BL:l2.

(TUlO and TS03)

PRE>MMO:/VEC=224/CSR=l72440
PRE>MMO:label/BL:l2./VE=DKO:/BL:l2.

(TU16)

NOTE
For TU16's at non-standard vectors or
CSR'S, the first line should be modified
and
CSR
to
reflect
the
vector
assignment.
See
Appendix A for a
complete discussion.

2-ll

GETTING STARTED
The first DECtape, cassette, or magtape used in preserving a volume
contains a bootstrappable image of PRESRV;
thus, if the system disk
is inadvertently destroyed, it can be restored by bootstrapping from
the first tape created. Such bootstrapping automatically initiates
PRESRV, which can then be used to restore the original disk contents.
PRESRV copies to and from any supported disk or tape and has a number
of options. The complete documentation for PRESRV appears in Appendix
A.

2.2.2

Preserving the RP Distribution Kit

The RP02, RP03, and RP04 distribution kit is supplied on 600-foot,
9-track, 800-BPI magtapes. To create an RP disk, the proper file must
be copied from tape to the disk. This is done using PRESRV.
The
first step is to select the proper tape and mount it on drive zero.
Mount the RP disk on drive zero and allow it to reach operating speed.
Using
the bootstrap ROM on the host computer, bootstrap PRESRV from
the distributed magtape.
The command strings presented below will copy the disk image from tape
to disk.
For RP02 and RP03 disks:
PRE>DP:/FO/VE=MTO:RSX11MBL12
PRE>DP:/FO/VE=MMO:RSX11MBL12

(TUlO and TS03)
(TU16)

For RP04 disks:
PRE>DB:/BL:6/FO/VE=MTO:RSX11MBL12
PRE>DB:/BL:6/FO/VE=MMO:RSX11MBL12

2.2.3

(TUlO and TS03)
(TU16)

Restoring Damaged Volumes

If a system disk is destroyed on a single RK05 system,
it must be
re-created via a backup tape that has been created by PRESRV. The
bootstrapping of a backup tape, with the exception of cassettes (which
are discussed below),
is identical to that of the Baseline System.
Refer to section 2.3. Bootstrapping a backup tape created via PRESRV
produces the following console display:
RSX-llS VOl BL12
VOLUME PRESERVATION PROGRAM - V02.l
FOR HELP TYPE /HE
nn. - BLOCKS AVAILABLE FOR BUFFERING
PRE>
PRESRV is now ready to accept commands. To restore the contents of a
set of backup tapes, assuming the disk and tapes are mounted on unit 0
of their respective controllers, enter
the commands given in the
appropriate section (2.2.3.l through 2.2.3.5).

2.2.3.1

Cassette Restoration

PRE>DKO:/VE/BL:l/FO=CTO:label
PRE>DKO:/VE/FO=CTO:label

( 16K machines)
(all others)
2-12

GETTING STARTED
where label is the name that was specified when the disk was preserved
to the cassette.
Cassette bootstrapping without a cassette ROM
bootstrap requires an extensive key-in via
the console
switch
register.
On a 11/05/10/20/35/40/45/50/70, proceed as follows:
1.

Press HALT.

Enter the address 001000 in the console switch
press LOAD ADRS.

Enter the first value from the right-hand column of the table
below in the console switch register; lift the DEP switch.

Enter the next value from the right-hand column of the
in the console switch register;
lift the DEP switch.

Repeat Step 4 for the remaining 26 values in the table.

Enter the address 001000 in the console switch
press LOAD ADRS.

Lift the HALT switch.

Press START.

and

table

and

The cassette will be bootstrapped, and PRESRV will display its initial
message and prompt.
ADDRESS VALUE TABLE
001000 012700
001002 177500
001004 005010
001006 010701
001010 062701
001012 000052
001014 012702
001016 000375
001020 112103
001022 112110
001024 100413
001026 130310
001030 001776
001032 105202
001034 100772
001036 116012
001040 000002
001042 120337
001044 000000
001046 001767
001050 000000
001052 000755
001054 005710
001056 100774
001060 005007
001062 017640
001064 002415
001066 112024

2-13

GETTING STARTED
2.2.3.2

DECtape Restoration

PRE>DKO:/VE/BL:l/FO=DTO:label/TP
PRE>DKO:/VE/FO=DTO:label/TP

(16K machines)
(all others)

where label is the name that was specified when the disk was preserved
to DECtape.

2.2.3 •. 3

Disk-to-Disk Restoration

PRE>DKl:/BL:l/VE/FO=DKO:/BL:l
PRE>DK1:/BL:l2./VE/FO=DKO:/BL:l2.

2.2.3.4

(16K machines)
(all others)

Floppy Disk-to-RK Disk Restoration

PRE>DKO:/VE/FO/BL:l.=DXO:label/TP
PRE>DKO:/VE/FO/BL:l2.=DXO:label/TP

(16K machines)
(all others)

where label is the name that was specified when the disk was preserved
to floppy disk.

2.2.3.5

Magtape-to-RK Disk Restoration

PRE>DKO:/VE/FO=MTO:label
PRE>DKO:/VE/FO=MTO:label

(TUlO and TS03)
(TU16)

where label is the name that was specified when the disk was preserved
to mag tape.
It is advisable to create a backup disk.
In the event that
inadvertent destruction of a system disk occurs, a backup will already
be available, and a second backup can be created at will.
Systems with more than one disk unit, such that disk
backup medium, may restore damaged system disks
existing backup disk.

2.2.4

is
by

the normal
copying the

Exiting from PRESRV

Since PRESRV is a stand-alone program, the only way to exit from it is
to bootstrap in a new system or stand-alone program.

2.2.5

Formatting Disks

Formatted disks are required to restore or create a disk copy of the
preserved distribution kit disks.
Disks that have never been used
require formatting before any data can be written on them. Formatting
a disk causes timing and addressing information to be recorded on the
disk and is accomplished using the stand-alone PRESRV program.
The
command
lines given in section 2.2.3 include the switch /FO,
indicating that a format pass should be made before copying the backup
to the disk. This will ensure that all data is readable on the disk.

2-14

GETTING STARTED
2.3

BOOTSTRAPPING THE BASELINE SYSTEM

The procedure for bootstrapping depends on which hardware bootstrap is
present on the host processor. On the 11/05/10/15/20/35/40/45/50/70
processors, the operator must first halt the processor by pressing the
HALT key, set the console switches to the hardware bootstrap ROM
address, press the LOAD ADRS key, and then press the START key.
The
RSX-llM system message is then displayed at the console terminal. The
step-by-step procedure is carried out as described in section 2.3.l,
2.3.2, or 2.3.3.
On an 11/04 processor having the M9301-YA/YB ROM,
the operator must press the BOOT button and enter (via the console
terminal)
the device name and unit number of the peripheral device
from which the bootstrap is to occur. The step-by-step procedure is
described in section 2.3.4.

Start-uE Usin9 the BM792-.YB Boo ts tr a12 ROM

")
'l
"" • ..;J • .J..

Depress HALT.

Set the console switches to 173100.

Depress LOAD ADRS.

Enter the address of the load device in the console switches:
RKll disk

177406

RFll disk

177462

RPll disk

176716

TCll DECtape

177344

Lift up the HALT key.

Press START.

The start-up sequence shown in section 2.4 is then
console terminal.

2.3.2

displayed

the

Start-u:e Using the MRll-DB Boo tstra12 ROM
l.

Depress HALT.

Enter the address of the load device in the console switches:
RFll disk

173100

RKll disk

l 73ll0

TCll DECtape

173120

TMll mag tape

173136

RPll disk

173154

2-15

GETTING STARTED
3.

Depress LOAD ADRS.

Lift up the HALT key.

Press START.

The start-up sequence shown in section 2.4 is then
console terminal.

2.3.3

displayed

the

Start-up Using the BM873-YA Bootstrap ROM
1.

Depress HALT.

Enter the address of the load device in the console switches:
RFll disk

173000

RKll disk

173010

TCll DECtape

173030

TMll mag tape

173050

RPll disk

173100

RCll disk

173144

KLll/DLll

173210

TAll cassette

173230

PCll papertape 173312
3.

Depress LOAD ADRS.

Lift up the HALT key.

Press START.

The start-up sequence shown in section 2.4 is then
console terminal.

2.3.4

displayed

the

Start-up Using the BM873-YB Bootstrap ROM
1.

Depress HALT.

Enter the address of the load device in the console switches:
RH11/RH70/RS03/RS04 disk (unit 0)

173000

RH11/RH70/RS03/RS04 disk (unit in
switch register)

173002

RKll disk (unit 0)

173030

2-16

GETTING STARTED
RKll disk (unit in switch register)

173032

TCll DECtape (unit 0)

173070

TMll/TMAll magtape

173110

RFll disk

173136

RHll/RH70/TM02/TU16 magtape (unit 0)

173150

RCll disk

173212

RHll/RH70 device combination (unit zero)

173230

RH11/RH70 device combination {unit in
switch register)

173232

RHll/RH70/RP04 disk (unit 0)

173320

RHll/RH70/RP04 disk (unit in
switch register)

173322

RP11/RP02/RP03 disk (unit 0)

173350

RPll/RP02/RP03 disk (unit in
switch register}

173352

KLll/DLll console terminal reader

173510

TAll cassette (unit zero}

173524

TAll cassette (unit in switch register)

173526

PCll papertape reader

173620

Depress LOAD ADRS.
(For units other than
number in switch register.}

Lift up the HALT key.

Press START.

The start-up sequence shown in section 2.4 is then
console terminal.

2.3.5

zero,

enter

displayed

unit

the

Start-up Using the M9301-YA or M9301-YB Bootstrap ROM
1.

Press the BOOT switch.

In response to the "$" prompt on the console terminal,
one of the following:
DB - RJP04 disk

M9301-YB only

DK - RKll disk
DP - RPll (RP02/03) disk
OS - RS03/04 disk

M9301-YB only

2-17

enter

GETTING STARTED
OT - TCll DECtape
DX - RXll floppy disk
CT - TAll cassette tape
MC - Mixed massbus device

M9301-YB only

MM - TM02/TU16 magtape

M9301-YB only

MT - TMll/TUlO magtape
PR - PCll/PRll paper tape reader
TT - KLll/DLll console terminal reader
followed by an optional, single octal digit for bootstrapping
from units other than zero, and terminated by a carriage
return.
The start-up sequence shown in section 2.4 is then
console terminal.

2.4

displayed

the

CONSOLE DISPLAY

If the system being bootstrapped was built on a system with a
different device configuration than the current host configuration, a
series of display lines appears on the console in the following form:
DEVICE ddnn:

NOT IN CONFIGURATION

where:
dd = device name
nn
unit number
Messages of this form indicate
present on the host system.

which

devices

are

not

physically

Then the system displays:
RSX-llM <version-number> <base level> nnnnK [MAPPED]
>RED ddnn:=SYO:
>MOU ddnn:
>@[l,2]STARTUP
where:
<version number> - The RSX-llM release number.
<base level>

- The RSX-llM base level (Release 2 is BL12) •

nnnnK

- Size of memory on hardware configuration
from which the system was bootstrapped.

MAPPED

- MAPPED is displayed for systems built to
use the memory management unit.

- Load device name.

2-18

GETTING STARTED
nn

- Load device unit.

[l,2]STARTUP

- Start-up MCR command file which contains
user determined commands to initialize the
system. File may be null.

The system is now ready to run. The RED and MOU commands displayed
are internally generated Redirect and Mount commands. The Redirect
command establishes the load device as the system device (SY:), and
the Mount command mounts the load device as a Files-11 volume. The
@[l,2]STARTUP command is also internally generated and is used to
execute a set of MCR commands which initialize the user's system
(e.g., to request that the operator set the time and date and then
mount a set of other disks to run other tasks}.

2-19

CHAPTER 3
SYSGEN DIALOGUE DETAILS

3.1

SYSGEN DIALOGUE

This chapter provides details of the SYSGEN dialogue.
the process is discussed:

Each

phase

Section 3.2 covers phase 1,
Section 3.3 covers the SGN query program, and
Section 3.4 covers phase 2.
All text, questions, and commands displayed during the running of the
command file are described.
However, during any single run of the
command file, many of these questions and commands will not be
executed, since they depend on the machine configuration that the
system generation is being run on and the target machine the system is
being built for. Appendix G contains the console output for an actual
system generation run.
At various places during the execution of the SYSGEN.CMD
and
SYSGEN2.CMD files, the user will be asked questions to which he must
reply Y<CR> (YES) , N<CR> (NO) , or simply <CR> (NO) , where <CR>
represents the carriage return character. Questions are distinguished
by a leading asterisk (*)
and trailing"?
[Y/N]:".
Command file
processing is suspended until an answer is keyed in.
In many places in the description of tne command files, two very
similar commands will be shown together, the only difference being the
UIC specified in the command (e.g., [1,20] and [l,24]). Both commands
are shown for completeness.
However, the user should keep in mind
during the running of the command files that only one of the two
commands will actually be issued. The specific command issued is a
function of whether the host system is mapped or unmapped and whether
the target system is mapped or unmapped.
The SYSGEN command files have been designed
to
inhibit
the
installation and removal of tasks already installed in the host
system. This makes it possible to perform a SYSGEN on a multi-user
system without affecting the tasks in that system. By using the
Assign (ASN) command to define a logical device (SY:) which is
assigned to a physical device containing a release disk, a SYSGEN may
be accomplished on-line with a multi-user system.

3-1

SYSGEN DIALOGUE DETAILS
3.2

SYSGEN.CMD FILE DETAILS

This section describes the entire SYSGEN.CMD indirect command file by
adding descriptive text to the actual command file to clarify the
SYSGEN process. The command file is presented in upper-case
text;
the descriptive text is presented in upper- and lower-case text.

SYSGEN PART 1

COPYRIGHT {C) 1975, DIGITAL EQUIPMENT CORP., MAYNARD MASS. 01754.
DETERMINE SYSTEM FEATURES AND ASSEMBLE THE EXECUTIVE
IF YOU ARE PLANNING TO INCLUDE ICS/ICR OR UDCll SUPPORT YOU
SHOULD CONSULT THE LATEST RELEASE NOTES BEFORE PROCEEDING FURTHER.
i

* DO YOU WISH TO CONTINUE? [Y/N]:
LONG DIALOGUE PRINTS A DESCRIPTION OF EVERY STEP IN THIS
SYSGEN COMMAND FILE. ON THE OTHER HAND, SHORT DIALOGUE
PROVIDES VIRTUALLY NO EXPLANATORY TEXT.
;

* DO YOU WANT LONG DIALOGUE? [Y/N]:
For
the experienced programmer who has performed many
SYSGENs, much of the descriptive text in the file may be
suppressed by answering N to this question.
It
is
recommended that the less experienced user answer Y, since
it will lead him through the system generation process
correctly.
*ARE YOU BUILDING AN RSXllS SYSTEM? [Y/N]:
The selection of features and assembly of the Executive are
controlled by the SYSGEN.CMD command file. This question is
used later to select the correct SYSGEN phase 2 command file
for RSX-llM or RSX-llS.
* ARE YOU BUILDING A MAPPED SYSTEM? [Y/N]:
This query determines the UIC settings that will
be
generated during the SYSGEN process. See Appendix B for UIC
conventions used on the distribution kit disks.
*ARE YOU RUNNING ON A MACHINE WITH AT LEAST 24K WORDS? [Y/N]:
The Baseline System Executive is set up with a 14K partition
which can be used if the host machine has at least 24K of
memory.
If this is the case, SYSGEN will use the larger and
faster version of MACR0-11 during the generation process.
*WILL YOUR SYSTEM INCLUDE ICS/ICR-11 SUPPORT? [Y/N]:
When generating ICS/ICR-11 support into the Executive,
device common blocks which map to the ICS/ICR-11 device
registers must be copied from the source disk to the disk on

3-2

SYSGEN DIALOGUE DETAILS
which SYSGEN phase 2 will be run. This question will
the proper files to be copied at the right time.

cause

*WILL YOUR SYSTEM INCLUDE UDCll SUPPORT? [Y/N]:
When generating UDCll support into the Executive, device
common blocks which map to the UDCll device registers must
be copied from the source disk to the disk on which SYSGEN
phase 2 will be run. This question will cause the proper
files to be copied at the right time.
This question will
not be asked if ICS/ICR-11 support was selected, since
ICS/ICR-11 support precludes UDCll support.

INSTALL THE NECESSARY TASKS
SET /UIC=[l,50]
SET /UIC=[l,54]
The SET command appears frequently in the SYSGEN.CMD file.
Generally, its use in this file is to establish a default
UIC.
This
makes
it
unnecessary
to
include
UIC
specifications in subsequent commands.
The UIC [1,50] is the one under which unmapped system tasks
are stored.
The UIC [1,54] is the one under which mapped
system tasks are stored. Only one of the two above commands
will be issued, depending on the type of system the SYSGEN
is being run on. Similarly, at various places in the r11e
where both of the commands appear, only one will be issued
during an actual SYSGEN.
INS PIP
The Baseline System has a limited pool of dynamic memory,
and since installed tasks consume dynamic memory even when
they are inactive, the strategy is to install tasks only
when they are needed and remove them as soon as practical.
Tasks are installed and removed only if they are not
installed in the host system. This makes it possible to
perform SYSGENs on multi-user systems.
DELETE FILES FROM PAST SYSGENS
NOTE: ON VIRGIN DISKS, NONE OF THE FILES WE TRY TO
DELETE WILL EXIST. PIP WILL GIVE ERROR MESSAGES
THAT SHOULD INDICATE THAT THE FILE DOES NOT EXIST.
In generating a system, several files are created.
These
files are of no use after successfully completing a SYSGEN.
If another SYSGEN is to be performed from a disk on which a
previous SYSGEN was performed, these files should be deleted
to free valuable disk space.
The following
commands
accomplish the desired cleanup.
SET /UIC= [ 11, 20]
By convention, the following files in an unmapped system are
stored under UIC [11,20]:

3-3

SYSGEN DIALOGUE DETAILS
1.

The assembly command
(RSXASM.CMD),

file

assemble

the

Executive

The build command
(RSXBLD.CMD),

file

build

the

Executive

Executive object modules created during the assembly
the Executive,

A file of concatenated Executive object modules built by
PIP and used by the Librarian to build the Executive
object module library, and

Files for the ICS/ICR-11 or UDCll used to build device
common block9 for tasks which reference the device.

PIP RSXASM.CMD;*/DE
PIP RSXBLD.CMD;*/DE
PIP *.OBJ;*/DE
PIP *.UDC;*/DE
PIP *.ICR;*/DE
The preceding PIP commands delete:
1.

All versions of the assembly command file,

All versions of the task build command file,

All object module files,

All versions of UDCll modules, and

All versions of ICS/ICR-11 modules.
On a release kit disk that has never had a SYSGEN
performed on it, no residual files from a previous
SYSGEN exist.
Thus, the attempted deletion is
flagged by PIP. For example, the command line:
>PIP RSXASM.CMD;*/DE
produces:
PIP -- NO SUCH FILE(S)
SY: [ll,20]RSXASM.CMD;*
These are warning diagnostics and do not affect the
SYSGEN currently in progress.

SET /UIC= [ 11, 24]
PIP RSXASM.CMD;*/DE
PIP RSXBLD.CMD;*/DE
PIP *.OBJ;*/DE
PIP *.UDC;*/DE
PIP *.ICR;*/DE
It cannot be known whether the previous SYSGEN done on the
source disk was for a mapped or an unmapped system. The
preceding commands delete the files generated for building a
mapped system.

3-4

SYSGEN DIALOGUE DETAILS
SET /UIC=[ll,30]
PIP UDCOM.LST;*/DE
PIP ICTAB.LST;*/DE,ICOM.LST;*
SET /UIC=[ll,34]
~IP UDCOM.LST;*/DE
PIP ICTAB.LST;*/DE,ICOM.LST;*
The previous commands delete listing
ICR/ICS-11 or UDCll common blocks
unmapped areas on the disk.

files generated for
for both mapped and

SET /U!C=[ll,10]
By convention, UIC (11,10] is used for the device table file
(SYSTB.MAC), prefix file
(RSXMC.MAC), and the ICR/ICS-11
table file produced during the running of the SGN query
program. The output of SGN is used to configure the system.
PIP SYSTB.MAC;*/DE
PIP RSXMC.MAC;*/DE
PIP ICTAB.MAC;*/DE
Delete all existing versions of the
prefix file, and ICS/ICR-11 table.

device

table

file,

NOW WE RUN THE SYSGEN PROGRAM TO SELECT THE EXECUTIVE
FEATURES AND DEVICES WE WANT IN THE NEW SYSTEM
NOTE: IF YOU ARE RUNNING ON A 16K WORD MACHINE, YOU SHOULD
SPECIFY THAT NO ASSEMBLY LISTINGS BE GENERATED.
IF YOU ARE RUNNING ON A MACHINE WITH AT LEAST 24K WORDS,
YOU MAY GET ASSEMBLY LISTINGS. HOWEVER, THE FOLLOWING
CONDITIONS APPLY:
1) RK SYSTEM DISK - DIRECT LISTINGS TO LP: OR DKl:
2) RP02/03, RP04 SYSTEM DISK - DIRECT LISTINGS TO LP: OR SY:
RUN $SGN
The SGN query program is run to produce the five files
mentioned previously. This process consists of a number of
interactive queries. The operator-supplied answers to these
queries provide SGN with the information necessary to
construct its output files. These files are then used by
MACR0-11 and the Task Builder to create the user-tailored
system.
The production of the Executive map MUST be specified as "Y"
(yes), since it is required in phase 2 of SYSGEN to
determine partition layout.
The Baseline System
therefore, spooling
specified.

does not contain a print spooler;
of the listings or map should not be

SGN presents queries which, when answered, define:
Executive options,
Processor options, and the
Peripheral complement

3-5

SYSGEN DIALOGUE DETAILS
for the user-tailored system. For a detailed explanation of
SGN dialogue details, see section 3.3. For brevity, these
queries, which are essentially self-explanatory, are omitted
here.
If answers to the SGN queries are unsatisfactory,
the
command file provides a means of orderly exit. Abort SGN by
typing CTRL/C and <CR>; when MCR prompts, type ABO.
This
command will abort the SGN program. Allow the indirect file
processor to process the next commands in the file and wait
until the next question is asked. Answer N to this question
and a clear exit can be made (see next question).
AT THIS POINT WE RENAME THE ASSEMBLY COMMAND FILE
AND THE SYSTEM BUILD COMMAND FILE TO
[11,20]
[11,24]
PIP [11,20]/RE=RSXASM.CMD;*,RSXBLD.CMD;*
PIP [11,24]/RE=RSXASM.CMD;*,RSXBLD.CMD;*
SGN runs under UIC [11,10] and produces five files which are
recorded in the [11,10] file directory. Command files are
conventionally stored in [11,20] or [11,24]. The preceding
PIP commands, only one of which is actually issued, puts the
command files into the proper directory.
*DID YOU ANSWER THE SYSGEN QUESTIONS TO YOUR SATISFACTION? [Y/N]:
If, for any reason, the SGN query program included incorrect
responses that would result in a user-tailored Executive
which did not reflect the user's intentions, an "N"
(no)
response to this query will effect a controlled exit and
permit the user to restart the entire system generation
process.
If N <CR> is answered, the following lines are
displayed:
WE WILL EXIT NOW SO THAT YOU MAY RESTART THIS COMMAND FILE
FROM THE BEGINNING. THIS WILL ALLOW A CLEAN UP OF THE
DISK AND A NEW RUN OF SGN TO SELECT THE PROPER OPTIONS
FOR YOUR SYSTEM.

'
REM
PIP
If no errors occurred during SGN, the
displayed.

following

lines

are

*ARE YOU RUNNING AN RP02/03 OR RP04 AS THE SYSTEM DISK? [Y/N]:
This question is asked to determine whether the output from
the assembly of the Executive will have to be transferred to
another medium and to determine where the optional source
listings are going.
If this is answered "Y", the next line
displayed begins with the assembly of the Executive.
If
not, the next question is displayed.
*ARE YOU GENERATING ASSEMBLY LISTINGS ON DKl? [Y/N]:
For RK disk SYSGENs, it is possible to specify DKl: as the
output device
for
listing files.
If
this question is

3-6

SYSGEN DIALOGUE DETAILS
answered "Y<CR>", the disk will be mounted, and the
directory will be added for the listing files
following commands.
PLACE A DISK IN DRIVE i

proper
by the

(WRITE-ENABLED) •

;

* IS THE DISK READY IN DRIVE l? [Y/N]
This question causes the indirect file processor to suspend
itself until the disk reaches operating speed. When the
disk is ready, answer "Y" to this question.
*IS THE DISK IN DRIVE 1 INITIALIZED AS A FILES-11 VOLUME? [Y/N]:
If the disk is not initialized, an "N" answer will cause the
disk to be initialized by the following commands, which
install task INI, initialize the disk, and then remove !NI.
If the disk is already initialized, processing continues at
the Mount (MOU) command.
INS [l,50]INI
INS [l,54]INI
!NI DKl:
REM !NI
Whether or not the disk needs to
processing continues at this point.

initialized,

DKl:

MOU DKi:
INS [l,50]UFD
INS [l,54]UFD
UFO DKl: [11, 30]
UFO DKl: [ 11, 3 4]
REM ur:u
flt:ll'"\

Processing of all disk types continues at this point.
NOW WE ASSEMBLE THE EXECUTIVE
SET /UIC=[l,50]
SET /UIC=[l,54]
INS MAC
INS BIGMAC
The proper version of MACR0-11 is installed, depending on
whether or not the host system is mapped, and whether or not
it has at least 24K of memory. BIGMAC is the assembler of
choice and is always chosen when at least 24K of memory is
available.
SET /UIC=[ll,20]
SET /UIC=[ll,24]
TIM
MAC @RSXASM
TIM
The Set command establishes the default UIC.
MACR0-11 is
requested to run with RSXASM as an input indirect command
file. Note that MACR0-11 is running
under [11,20] or

3-7

SYSGEN DIALOGUE DETAILS
[11,24] and, since RSXASM is part of an
indirect file
request,
.CMD is the default file type.
The assembly will
require from 15 minutes to 2 hours to complete, depending
upon the hardware available and the configuration of the
target system.
If the UDCll or ICS/ICR-11 was selected for
the
target
system, the following text and commands are displayed:
NOW WE ASSEMBLE THE MODULES NECESSARY FOR TASKS TO ACCESS
UDCll OR ICS/ICR-11 I/0 MODULE REGISTERS WITHIN THE I/O
PAGE.
I/O MODULE TYPES ACCESSED DIRECTLY ARE:
UDCll:
ANALOGUE OUTPUTS
SINGLE SHOT DIGITAL OUTPUTS
DIGITAL INPUTS
COUNTER INPUTS
ICS/ICR-11
DIGITAL INPUT
COUNTER INPUT
If the ICS/ICR-11 was selected, the
asked:

following

question

*IS DIRECT ACCESS TO THE ICS/ICR-11 I/0 MODULES REQUIRED? [Y/N]:
If the UCDll was selected, the following question is asked:
*IS DIRECT ACCESS TO THE UDC-11 I/O MODULES REQUIRED? [Y/N]:
If either of the above questions
following text is displayed:

answered

IN AN UNMAPPED SYSTEM ACCESS TO I/O MODULES IS
THROUGH UNRESTRICTED REFERENCE TO THE I/O PAGE
OR VIA A GLOBAL COMMON BLOCK THAT OCCUPIES THE SET
OF PHYSICAL LOCATIONS RESERVED FOR THE DEVICE REGISTERS. IF YOU WISH TO. ACCESS THE DEVICE THROUGH THE
COMMON BLOCK, YOU MUST EDIT THE FILE RSXMC.MAC UNDER
UIC [11,10], AS SHOWN IN THE FOLLOWING EXAMPLE:
For UDCll
*PL U$$Dll
U$$Dll=O
*I U$$DCM=O
*ED

3-8

"Y",

the

SYSGEN DIALOGUE DETAILS
For ICS/ICR-11
*PL I$$Cll
I$$Cll=O
*I I$$COM=O

*ED
In either case, the following question is asked:
* DO YOU HAVE TO EDIT FILE RSXMC.MAC? [Y/N]:
If this question is answered "Y", the following commands are
issued:
INS [l,SO]EDI
INS [l,54]EDI
EDI [11,lO]RSXMC.MAC
The user should perform the edit specified. The Line Text
Editor (EDI) will be removed and the correct modules will be
assembled.
REM ••• EDI
MAC
MAC
MAC
MAC
MAC
MAC

[ll,24]ICTAB.ICR,[ll,34]ICTAB/-SP=[ll,10]ICTAB
[ll,24]ICOM.ICR,[ll,34]ICOM/-SP=[ll,10]RSXMC,ICOM
[ll,20]ICTAB.ICR,[ll,30]ICTAB/-SP=[ll,10]ICTAB
[ll,20]ICOM.ICR,[ll,30]ICOM/-SP=[ll,10]RSXMC,ICOM
[ll,24jUDCOM.UDC,[ll,34]UDCOM/-SP=[ll,10]RSXMC,UDCOM
[ll,20]UDCOM.UDC,[ll,30]UDCOM/-SP=[ll,10]RSXMC,UDCOM
Whether or not UDCll or ICS/ICR-11 support is selected,
processing continues here.

REM

all

••• MAC

DMO DK!:
NOW WE BUILD THE CONCATENATED OBJECT MODULE FILE
FOR THE EXECUTIVE.
The assembly of the Executive has resulted in the creation
of a number of individual object module files.
These
individual files can be concatenated into a single file with
a substantial savings in disk space. Later, during the
build phase, the Task Builder processes a library produced
from this concatenated file to create the task image of the
user-tailored Executive.
PIP RSXllM.OBS=*.OBJ
PIP .OBJ/RE=RSXllM.OBS
All existing object files cataloged under
the default UIC
are concatenated as a result of this command. The file
RSXllM.OBS is then renamed RSXllM.OBJ
For purposes of SYSGEN, a host system is defined as
the following:
1.

2 RK disks;

1 RK disk, 1 TAll dual cassette;

3-9

one

SYSGEN DIALOGUE DETAILS
3.

1 RK disk, 1 TU56 DECtape;

1 RK disk, 1 TUlO, TU16 or TS03 magtape;

1 RK disk, 1 RXOl floppy disk;

1 RP02, RP03, or RP04 disk.

Queries are posed until one of these configuration types is
established.
For disk-tape configurations,
two files,
RSXBLD.CMD and RSXllM.OBJ, must be copied to backup tape
for later restoration to the object disk.
In addition, if
the UDCll was included, the common block module UDCOM.UDC
must be copied.
If the ICS/ICR-11 was included, the files
ICOM.ICR and ICTAB.ICR must be copied.
NOW WE COPY THE BUILD COMMAND FILE, THE ASSEMBLY PREFIX
FILE, THE UDC OR ICS/ICR FILES AND THE CONCATENETED
OBJECT MODULES TO SECONDARY STORAGE.
;

*IS YOUR SECONDARY STORAGE DEVICE AN RK CARTRIDGE DISK? [Y/N]:
The following instructions and commands are executed if
host system is a 2-RK disk system.

the

PUT THE UNMAPPED OBJECT DISK IN DRIVE 1
PUT THE MAPPED OBJECT DISK IN DRIVE 1
AND MAKE SURE THAT THE DRIVE IS WRITE ENABLED.

* IS THE DISK IN DRIVE 1 READY? [Y/N] :
MOU DKl:
SET /UIC=[l,20]
SET /UIC=[l,24]
PIP DK1:/NV=[ll,20]RSXBLD.CMD,RSX11M.OBJ
PIP DK1:/NV=[ll,24]RSXBLD.CMD,RSX11M.OBJ
PIP DK1:/NV=[ll,24]UDCOM.UDC
PIP DK1:/NV=[ll,24]ICOM.ICR,ICTAB.ICR
PIP DK1:/NV=[ll,20]UDCOM.UDC
PIP DK1:/NV=[ll,20]ICOM.ICR,ICTAB.ICR
PIP DKl: [200,200]/NV=[ll,lO]RSXMC.MAC
DMO DKl:
End of 2-RK system processing.
after the RXOl questions.

The

display

begins

INS [l,SO]FLX
INS [l,54]FLX
All other configurations, except the RXOl, require the File
Transfer Program (FLX)
to copy the files to the medium.
Thus, the proper version of FLX is installed.
*IS YOUR SECONDARY STORAGE DEVICE A CASSETTE TAPE? [Y/N]:
End of RK-CT processing. The next display begins after
the
RXOl questions. The following instructions and commands are
executed if the host is an RK-CT system.

3-10

SYSGEN DIALOGUE DETAILS
LOA CT:
;

; PUT A BLANK CASSETTE IN DRIVE 0
;

*IS THE BLANK CASSETTE IN DRIVE 0 READY? [Y/N]:
FLX CTO:/ZE
FLX CTO:/VE/BS:512.=RSXBLD.CMD/RS,RSX11M.OBJ
FLX CTO:/VE/BS:512.=[ll,10]RSXMC.MAC/RS
FLX CTO:/VE/BS:512./FB=UDCOM.UDC/RS
FLX CTO:/VE/BS:512./FB=ICTAB.ICR,ICOM.ICR/RS
End of RK-CT system processing.
after the RXOl questions.

The

display

begins

*IS YOUR SECONDARY STORAGE DEVICE A DECTAPE? [Y/N]:
The following instuctions and commands are executed
host is an RK-DT system.

the

LOA OT:
; PUT A BLANK DECTAPE IN DRIVE

;

*IS THE BLANK DECTAPE IN DRIVE 0 READY? [Y/N]:
FLX DTO:/ZE
FLX DTO:/DO=RSXBLD.CMD/RS,RSXllM.OBJ
FLX DTO:/DO/FB=UDCOM.UDC/RS
FLX DTO:/DO/FB=ICTAB.ICR,ICOM.ICR/RS
FLX DTO:/DO=[ll,lO]RSXMC.MAC/RS
The following instructions and command are executed
host is an RP02/RP03 or RP04 system.
COPY THE EXECUTIVE OBJECT MODULE FILE, TASK BUILD
COMMAND FILE AND UDC OR ICS/ICR FILES TO THE PROPER
DIRECTORY ..
;

SET /UIC=[l,20]
SET /UIC=[l,24]
PIP /NV=[ll,20]RSX11M.OBJ,RSXBLD.CMD
PIP /NV=[ll,24]RSX11M.OBJ,RSXBLD.CMD
PIP /NV=[ll,20]ICTAB.ICR,ICOM.ICR
PIP /NV=[ll,20]UDCOM.UDC
PIP /NV=[ll,24]ICTAB.ICR,ICOM.ICR
PIP /NV=[ll,24]UDCOM.UDC
PIP (200,200]/NV=[ll,lO]RSXMC.MAC
REM •.. PIP
REM •.• FLX
*DO YOU WANT TO GO DIRECTLY TO THE NEXT PHASE OF THIS SYSGEN? [Y/N]:
Since all files needed for a SYSGEN reside on an RP disk, it
is possible to invoke SYSGEN phase 2 immediately.
If this
question is answered "Y", the following commands and
text
are displayed.
@[200,200]SYSGEN2
@[200,200]11SGEN

3-11

SYSGEN DIALOGUE DETAILS
For RSX-llM SYSGEN,
the command file selected will be
SYSGEN2.CMD.
For RSX-llS SYSGEN, the command file selected
will be llSGEN.CMD. After phase 2, the command file will
exit.
WHEN THIS COMMAND FILE IS FINISHED, YOU MAY CONTINUE WITH
THE NEXT PHASE OF THIS SYSGEN WITH THE FOLLOWING COMMAND:
If SYSGEN phase 2 on an RP disk is not selected, it may be
done manually at a later time by following the directions
given at the end of the command file.
At this point, RP
processing terminates and the command file will exit.
* IS YOUR SECONDARY STORAGE DEVICE A MAGTAPE? [Y/N]:
The following instructions and commands are executed if
host is an RK magtape (TUlO, TS03, TU16) system.

the

MOUNT A MAGTAPE IN DRIVE 0
;

*IS THE MAGTAPE IN DRIVE ZERO READY? [Y/N]:
* IS THE MAGTAPE DRIVE A TU16? [Y/N] :
The device mnemonic for TU16 magtape is MM.
The device
mnemonic for TUlO and TS03 is MT.
This question is used to
select the proper sequence of commands.
LOA MT:
FLX MTO:/ZE
FLX MTO: [1,20]/DO=SY:RSXBLD.CMD/RS,RSXllM.OBJ
FLX MTO: (1,20]/DO=SY: (11,lO]RSXMC.MAC/RS
FLX MTO: (1,24]/DO=SY:RSXBLD.CMD/RS,RSXllM.OBJ
FLX MTO: (1,24]/DO=SY: (11,lO]RSXMC.MAC/RS
FLX MTO: (1,20]/DO/FB=SY:UDCOM.UDC/RS
FLX MTO: (1,20]/DO/FB=SY:ICTAB.ICR,ICOM.ICR/RS
FLX MTO: (1,24]/DO/FB=SY:UDCOM.UDC/RS
FLX MTO: (1,24]/DO/FB=SY:ICTAB.ICR,ICOM.ICR/RS
The above commands are issued for TUlO and TS03
The next display begins after the RXOl questions.

magtape.

LOA MM:
FLX MMO:/ZE
FLX MMO: (1,20]/DO=SY:RSXBLD.CMD/RS,RSXllM.OBJ
FLX MMO: (1,20]/DO=SY: (11,lO]RSXMC.MAC/RS
FLX MMO: (1,24]/DO=SY:RSXBLD.CMD/RS,RSXllM.OBJ
FLX MMO: (1,24]/DO=SY: (11,lO]RSXMC.MAC
FLX MMO: (1,20]/DO/FB=SY:UDCOM.UDC/RS
FLX MMO: (1,20]/DO/FB=SY:ICTAB.ICR,ICOM.ICR/RS
FLX MMO: [1,24]/DO/FB=SY:UDCOM.UDC/RS
FLX MMO: (1,24]/DO/FB=SY:ICTAB.ICR,ICOM.ICR/RS
The above commands are issued for TU16 magtape.
display begins after the RXOl questions.

The

* IS YOUR SECONDARY STORAGE DEVICE AN RXOl FLEXIBLE DISK? [Y/N]:
The following instructions and commands are executed if
host is an RK-RXOl system.

3-12

the

SYSGEN DIALOGUE DETAILS
LOA DX:
;

; PLACE A BLANK FLEXIBLE DISK IN DRIVE 0
;

*IS THE FLEXIBLE DISK IN DRIVE 0 READY? [Y/N]:
SET /UIC=[l,50]
SET /UIC=[l,54]
INS IN!
INI DXO:
REM ••. INI
MOU DXO:
INS UFO
SET /UIC=[l,20]
UFO DXO: [l,20]
SET /UIC=[l,24]
UFO DXO: [1,24]
REM ••. UFO
PIP DX0;=[11:20]RSXllM.OBJ;RSXBLD.CMD
PIP DXO:=[ll,24]RSX11M.OBJ,RSXBLD.CMD
PIP DXO:=[ll,20]UDCOM.UDC
PIP DXO:=[ll,20]ICTAB.ICR,ICOM.ICR
PIP DXO:=[ll,24]UDCOM.UDC
PIP DXO:=[ll,24]ICTAB.ICR,ICOM.ICR
PIP DXO:=[ll,lO]RSXMC.MAC
OMO DXO:
End of RK-RXOl
... FLX" below~

processing .

Processing

begins

with

"REM

If none of the supported configurations are selected,
the
following text is displayed and the questions are repeated.
***
E R R 0 R
***
THE CONFIGURATION YOU ARE RUNNING ON IS NOT
A STANDARD SYSGEN CONFIGURATION.
YOU MUST HAVE ONE OF THE
COMBINATIONS OF DEVICES SPECIFIED ABOVE. WE WILL
REPEAT THE CONFIGURATION QUESTIONS.
Normal processing continues
processing is completed.

here

after

specific

device

REM ..• FLX
REM ••. PIP
The FLX and PIP tasks are removed;
their function
is
complete, and
they are removed to bring the system back to
its initial state.

DISMOUNT THE SYSTEM DISK AND MOUNT THE
UNMAPPED OBJECT DISK
MAPPED OBJECT DISK
IN DRIVE 0 AND BOOT IN THE SYSTEM ON THAT DISK
WHEN THE SYSTEM IS UP, SET THE TIME AND DATE
AND INITIATE THE INDIRECT COMMAND FILE

3-13

SYSGEN DIALOGUE DETAILS
[200,200JSYSGEN2.CMD. FOR EXAMPLE:
[200,200JllSGEN.CMD. FOR EXAMPLE:
>TIM 10:20 9/15/75
The above text is displayed only for RK-disk host systems.
>@SYSGEN2
>@llSGEN
SET /UIC=[200,200J

3.3

SGN QUERY LIST

The SGN task consists of a series of questions which the user answers.
At the completion of SGN, the user has, in effect, specified his:
1.

Executive options,

Processor options,

Peripheral complement.

This section contains every question (listed in order of appearance}
in the dialogue conducted by SGN.
The dialogue itself has two
options, long and short. Short dialogue consists only of the queries.
Long dialogue precedes each query with explanatory text.
In the text which follows, upper case is produced by the SGN task.
Any additional explanation required appears in lower case, prefaced
with the heading:
Additional explanation:
***BEGIN SGN***
Additional explanation:
The System Generation Configuration Query Program (SGN) has
started execution.
A slight pause occurs while necessary
files are opened.
EVERY QUERY INCLUDES REQUIRED ANSWER OPTIONS.

OPTIONS ARE:

[Y/NJ

FOR YES/NO

[DDUJ

IF A DEVICE-UNIT IS REQUIRED

(OJ

IF REQUIRED NUMBER IS ASSUMED OCTAL

(DJ

IF REQUIRED NUMBER IS ASSUMED DECIMAL

THE OPTIONS WHICH APPLY TO A GIVEN QUERY ARE IMPLIED IN THE QUERY.
A
CARRIAGE RETURN WILL DEFAULT AN ANSWER TO NO FOR [Y/NJ AND ZERO FOR A
NUMERIC. NO DEFAULT APPLIES FOR THE DEVICE-UNIT.

3-14

SYSGEN DIALOGUE DETAILS
THE NUMERIC DEFAULT, [D]
OR
[O], CAN BE OVERRIDDEN BY PREFIXING
NUMBERS TO BE ACCEPTED AS OCTAL WITH A NUMBER SIGN (#} OR SUFFIXING
THOSE TO BE ACCEPTED AS DECIMAL WITH A DECIMAL POINT.
SOME QUERIES SUBSEQUENT TO KEY QUERIES HAVE PRESTORED REPLIES.
THESE
REPLIES CAN BE OVERRIDDEN BY TERMINATING A REPLY TO A KEY QUERY WITH
ALTMODE.
ANSWERS WHICH ARE IN ERROR ARE IMMEDIATELY FLAGGED BY AN ERROR MESSAGE
DISPLAY;
SGN WILL WAIT TO ACCEPT THE CORRECTED ANSWER.
SGN PROVIDES
LONG OR SHORT DIALOGUE OPTIONS.
SHORT DIALOGUE LISTS THE QUERY ONLY.
LONG DIALOGUE PROVIDES EXPLANATORY TEXT WITH EACH QUERY.
THE DIALOGUE
PROPER NOW COMMENCES.
LONG DIALOGUE RESULTS IN THE DISPLAY OF EXPLANATORY TEXT
PRECEDING EACH QUERY.
SHORT DIALOGUE OMITS THE TEXT.
LONG
DIALOGUE IS AVAILABLE ON A PER QUERY BASIS BY ENTERING ESCAPE
OR ALTMODE FOLLOWING THE DISPLAY OF THE QUERY.
LONG DIALOGUE DESIRED? [Y/N]:
Additional explanation:
Once long dialogue is chosen, the explanatory text is
displayed
prior
to
each query and cannot be
suppressed
(except on a line-by-line basis with
CTRL/O).
Selective short dialogue, which is used in
the descriptions which follow, produces explanatory
text following the query and is the recommended mode
of operation.
ASSEMBLY LISTING FILES DESIRED? [Y/N]:
SGN PRODUCES INDIRECT FILES WHICH ARE SUBSEQUENTLY USED TO
ASSEMBLE THE USER-TAILORED EXECUTIVE.
LISTING PRODUCTION IS
TIME CONSUMING AND IS NOT RECOMMENDED.
IF AN 8K ASSEMBLER IS
BEING USED (WHICH IS REQUIRED ON A 16K SYSTEM) YOU CANNOT HAVE
BOTH A LISTING FILE AND AN OBJECT FILE DURING THE SAME RUN.
Additional explanation:
Two passes are required to produce object files dnu
listings on a 16K system or any system using the 8K
assembler.
The first pass can be used to produce the
object modules.
Then
the file RSXASM should be
edited
to produce listing files.
The next two
queries do not appear unless assembly listings were
requested.
WHAT IS TO BE THE LISTING DEVICE? [DDU] :
NOTE, THE DDU REPLY IS NOT TERMINATED WITH A
INDICATE LINE PRINTER ZERO ENTER:

COLON.

LPO
IF SPOOLED OUTPUT IS DESIRED ANSWER THIS QUERY WITH:

SYO

3-15

THUS

SYSGEN DIALOGUE DETAILS
Additional explanation:
The listing files need not be directed to a line
printer.
They can be directed to a terminal or a
disk, e.g., TTO: or DKl:.
SPOOL LISTING FILE? [Y/N]:
SPOOLING WILL ONLY ACTUALLY OCCUR FOR DIRECTORY DEVICES.
Additional explanation:
The Baseline System does not have a print spooler.
Therefore, the user should select no spooling of
either listing files or map files.
TASK BUILDER MAP DESIRED? [Y/N]:
A TASK BUILDER MAP IS STRONGLY RECOMMENDED.
ESTABLISHING
PARTITIONS AND DIAGNOSING SYSTEM FAILURES REQUIRE THIS MAP.
Additional explanation:
The only way to lay out partitions for a target
system is to have the value of the last address used
by the Executive. This value can only be obtained
from the Executive memory allocation map. The next
two queries do not appear unless a map was requested.
WHAT IS TO BE THE LISTING DEVICE? [DDU]:
NOTE, THE DDU REPLY IS NOT TERMINATED WITH A
INDICATE LINE PRINTER ZERO ENTER:

COLON.

THUS

LPO
IF SPOOLED OUTPUT IS DESIRED ANSWER THIS QUERY WITH:
SYO
Additional explanation:
The listing files need not be directed to a line
printer.
They can be directed to a terminal or a
disk, e.g., TTO: or DKl:.
SPOOL MAP FILE? [Y/N] :
SPOOLING WILL ONLY ACTUALLY OCCUR FOR DIRECTORY DEVICES.
THE PRECEDING QUERIES DO NOT
AFFECT
THE
USER-TAILORED
EXECUTIVE.
THE EXECUTIVE OPTION QUERIES THAT FOLLOW WHICH ARE
PRECEDED BY A SINGLE ASTERISK ARE INCLUDED IN THE BASELINE
SYSTEM.
THE FEATURE, HOWEVER, IS NOT ESSENTIAL TO THE
OPERATION OF A MINIMUM SYSTEM. DOUBLE ASTERISKED QUERIES MUST
BE SELECTED FOR FULL SUPPORT OF THE MINIMUM SYSTEM. QUERIES
WITH NO ASTERISKS ARE NOT IN THE MINIMUM SYSTEM AND ARE
OPTIONAL.

3-16

SYSGEN DIALOGUE DETAILS
BEGIN SELECTION OF EXECUTIVE OPTIONS.
ARE YOU BUILDING AN RSX-llS SYSTEM? [Y/N]:
AS DISTRIBUTED, THE RSX~llM SOURCE DISK PROVIDES FOR THE
GENERATION OF AN RSX-llS EXECUTIVE.
IN ORDER TO GENERATE SUCH
A SYSTEM YOU MUST HAVE AN RSX-llS LICENSE AND CANNOT INCLUDE
FILES-11, CHECKPOINTING, NONRESIDENT TASK, OR DYNAMIC MEMORY
ALLOCATION SUPPORT.
Additional explanation:
If an RSX-llS system is being generated,
next four queries do not appear.

then

the

**DO YOU WANT FILES-11 ACP SUPPORT? [Y/N]:
FAILURE TO INCLUDE THIS OPTION ELIMINATES ALL THE SERVICES
DESCRIBED IN THE RSX-11 I/O OPERATIONS REFERENCE MANUAL. SINCE
DEC-SUPPLIED SOFTWARE MAKES USE OF THESE SERVICES, THEIR
ELIMINATION IS NOT RECOMMENDED.
**DO YOU WANT CHECKPOINTING? [Y/N]:
IF TASKS ARE TO BE DISPLACED FROM A PARTITION TO YIELD THE
PROCESSOR RESOURCES TO A TASK OF HIGHER PRIORITY ABLE TO USE
THE PARTITION, THEN CHECKPOINTING IS REQUIRED.
Additional explanation:
For large systems, checkpointing may not be required
if MCR, TKN, and FCP each operates in its own
dedicated partition:
thus, checkpointing may be
omitted. Note that, in a minimum system, these tasks
must run in the same partition and,
therefore,
checkpointing is required.
If checkpointing is selected, then the following
query is posed, and the nonresident task query does
not appear.
DO YOU WANT TASKS TO BE CHECKPOINTABLE DURING TERMINAL INPUT? [Y/N]:
SELECTION OF THIS OPTION ALLOWS CHECKPOINTABLE TASKS TO BE
DISPLACED FROM MEMORY DURING TERMINAL INPUT BY HIGHER PRIORITY
TASKS.
**DO YOU WANT NON-RESIDENT TASK SUPPORT?
IF SELECTED, THE SYSTEM WILL CONTAIN A TASK LOADER.
NOT
INCLUDING THIS LOADER MEANS TASKS MUST BE PERMANENTLY RESIDENT
IN THE GENERATED SYSTEM.
DO YOU WANT MEMORY MANAGEMENT UNIT SUPPORT? [Y/N]:
REQUIRED IF THE
MANAGEMENT UNIT.

USER-TAILORED

3-17

TARGET

SYSTEM

HAS

MEMORY

SYSGEN DIALOGUE DETAILS
Additional explanation:
If memory management
is selected and an RSX-llM
system is being generated, then the following query
is posed.
DO YOU WANT DYNAMIC MEMORY ALLOCATION SUPPORT? [Y/N]:
THIS OPTION ALLOWS AS MANY TASKS AS WILL FIT IN A PARTITION TO
BE SIMULTANEOUSLY RESIDENT.
THE ALLOCATION OF PARTITION MEMORY
IS AUTOMATIC AND IS CARRIED OUT DYNAMICALLY BY THE RSX-llM
EXECUTIVE.
Additional explanation:
If dynamic memory allocation is selected,
the next
query is also posed.
This feature may not be
included in an RSX-llS system.
DO YOU WANT AUTOMATIC MEMORY COMPACTION? [Y/N]:
AUTOMATIC MEMORY
COMPACTION
MINIMIZES
THE
EFFECTS
OF
FRAGMENTATION IN DYNAMICALLY ALLOCATED PARTITIONS.
IT IS
INVOKED
WHEN
FREE
SPACE
EXISTS
BETWEEN
RESIDENT
NONCHECKPOINTABLE TASKS AND TASKS ARE WAITING TO BE LOADED INTO
THE PARTITION.
Additional explanation:
Memory compaction is
implemented as a nonresident
task called the Shuffler. Whenever it is necessay to
compact memory in a partition,
the Shuffler
is
automatically loaded and proceeds to move
tasks
toward the front of the partition in an attempt
to
provide better memory utilization.
DO YOU WANT ADDRESS CHECKING? [Y/N]:
IF ADDRESS CHECKING IS INCLUDED,
USER GENERATED ADDRESSES
PASSED IN EXECUTIVE DIRECTIVES WILL BE SUBJECT TO BOUNDARY,
ALIGNMENT, AND MODULUS CHECKS.
Additional explanation:
This feature is automatically included if the
system contains a memory management unit.

target

**DO YOU WANT I/O RUNDOWN? [Y/N]:
I/O RUNDOWN INSURES THE INTEGRITY OF AN RSX-llM SYSTEM BY
MAKING SURE ALL OUTSTANDING TASK RELATED I/O OPERATIONS ARE
CLEANED UP BEFORE ALLOWING THE TASK TO EXIT.
OMISSION OF THIS
FEATURE SHOULD BE CAREFULLY CONSIDERED.
Additional explanation:
This feature is automatically included if the
target
system
contains a memory management unit.
Its
omission is recommended only in RSX-llS systems.

3-18

SYSGEN DIALOGUE DETAILS
DO YOU WANT NETWORK ACP SUPPORT?

[Y/N]:

AS DISTRIBUTED, THE RSX-llM SOURCE DISK PROVIDES FOR THE
GENERATION OF AN RSX-llM OR RSX-llS EXECUTIVE THAT CONTAINS
SUPPORT FOR NETWORK COMMUNICATIONS.
IN ORDER TO GENERATE SUCH
A SYSTEM YOU MUST HAVE THE APPROPRIATE NETWORK SOFTWARE
LICENSE.
Additional explanation:
Network software
provides
the
capability
for
simultaneously operating tasks on different systems
to establish logical communication links and
to
-exchange data.
If this option is not selected, the
next seven queries are omitted.
DO YOU WANT SUPPORT FOR REMOTE TASK LOADING? [Y/N] :
SELECTION OF THIS OPTION PROVIDES SUPPORT FOR THE LOADING
TASKS FROM A MASS STORAGE DEVICE ON ANOTHER SYSTEM.

Additional explanation:
This query is posed only for RSX-llS systems.
selected, the next query is also posed.

DO YOU WANT REMOTE TASK CHECKPOINTING SUPPORT? [Y/N]:
THIS OPTION ALLOWS TASKS THAT ARE REMOTELY LOADED TO BE
CHECKPOINTED BY HIGHER PRIORITY TASKS.
CHECKPOINT STORAGE
SPACE IS PROVIDED ON THE HOST MASS STORAGE DEVICE.
Additional explanation:
Remote task checkpointing on slow communication links
can be a lengthy process. Selection of this feature
should be given careful thought.
SHOULD THE NETWORK ACP BE OVERLAID? [Y/N]:
IN AN RSX-llM SYSTEM THE NETWORK ACP MAY BE
OVERLAID.
OVERLAYING CONSERVES APPROXIMATELY lK OF MEMORY AT THE COST OF
A SLIGHT DEGRADATION IN PERFORMANCE.
Additional explanation:
This query is posed only for RSX-llM systems.
The
network ACP cannot be overlaid in an RSX-llS system.
DOES THE TARGET SYSTEM HAVE A KGll CRC OPTION? [Y/N]:
THIS HARDWARE OPTION PROVIDES EFFICIENT CALCULATION OF
BLOCK CHECK CHARACTERS.

MESSAGE

Additional explanation:
This option cannot be selected unless the hardware is
actually present.

3-19

SYSGEN DIALOGUE DETAILS
WHAT IS THE TIMEOUT PERIOD FOR NETWORK MESSAGES? [D] :
NETWORK MESSAGE TIMEOUT IS PROVIDED TO DETECT COMMUNICATION
LINK AND REMOTE SYSTEM FAILURES. TYPICAL TIMES (IN SECONDS)
ARE ON THE ORDER OF TWICE THE LONGEST MESSAGE ROUND TRIP TIME.
RESPONSE LIMITS: 3.-60.
WHAT IS THE MAXIMUM NETWORK DATA MESSAGE SIZE IN BYTES? [D] :
TYPICAL NETWORK MESSAGE LENGTHS ARE 80.

TO 120.

BYTES.

RESPONSE LIMITS: 64.-528.
HOW MANY LOGICAL CHANNELS CAN BE ACTIVE SIMULTANEOUSLY? [D] :
THE COMMUNICATION BETWEEN TASKS IN A NETWORK IS VIA LOGICAL
CHANNELS.
ONE LOGICAL CHANNEL IS REQUIRED FOR EACH PAIR OF
COMMUNICATING TASKS.
RESPONSE LIMITS: 1.-254.
DO YOU WANT AST SUPPORT? [Y/N]:
IF YOU INTEND TO USE EITHER THE PRINT SPOOLER OR THE FORTRAN
ISA PROCESS CONTROL SUBROUTINE CALLS, THEN ASYNCHRONOUS SYSTEM
TRAP PROCESSING IS REQUIRED.
SEE THE RSX-llM
EXECUTIVE
REFERENCE MANUAL FOR A DETAILED DESCRIPTION OF THE AST FACILITY
PROVIDED IN RSX-llM/S.
Additional explanation:
AST support will be selected
automatically
Floating
Point Processor or network support
selected.

if
is

**DO YOU WANT SUPPORT FOR EXTERNAL MCR FUNCTIONS? [Y/N]:
THIS OPTION IS ALWAY REQUIRED IN AN RSX-llM SYSTEM AND NEVER IN
AN RSX-llS SYSTEM.
Additional explanation:
External MCR functions are tasks that
have
a
6-character name that starts with" ... ".
If this
option is selected {RSX-llM system only) , these tasks
may be initiated by simply typing their name.
**DO YOU WANT TASK TERMINATION AND DEVICE NOT READY MESSAGES? [Y/N]:
ABNORMAL TASK TERMINATION AND DEVICE NOT
AUTOMATICALLY
PRODUCED BY THE SYSTEM
SELECTED.

READY MESSAGES ARE
IF THIS OPTION IS

Additional explanation:
If this option is included in an RSX-llS system, then
the Task Termination Notification Task (TKN) must
also be included in the system for
the messages to
actually be printed.

3-20

SYSGEN DIALOGUE DETAILS
DO YOU WANT POWERFAIL RECOVERY? [Y/N] :
IF THE POWER FAILS AND THIS OPTION HAS BEEN OMITTED, THE SYSTEM
WILL
HALT
OPERATIONS.
OTHERWISE
SYSTEM
OPERATION IS
AUTOMATICALLY COMMENCED WHEN POWER IS RESTORED.
**DO YOU WANT GET PARTITION PARAMETERS DIRECTIVE? [Y/N]:
REQUIRED IF DEC-SUPPLIED PROGRAM
TKB, MAC, ETC.) IS TO BE RUN.

DEVELOPMENT

SOFTWARE

(I.E.,

DO YOU WANT GET SENSE SWITCHES DIRECTIVE? [Y/N]:
THIS DIRECTIVE PROVIDES THE CAPABILITY TO READ THE PROCESSOR
SWITCH REGISTER AND RETURN ITS CONTENTS TO THE REQUESTING TASK.
**DO YOU WANT GET TASK PARAMETERS DIRECTIVE? [Y/N]:
REQUIRED IF DEC-SUPPLIED PROGRAM
TKB, MAC, ETC.) IS TO BE RUN.

DEVELOPMENT

SOFTWARE

(I.E.,

DO YOU WANT SEND/RECEIVE DIRECTIVES? [Y/N]:
THESE DIRECTIVES PROVIDE THE CAPABILITY FOR CO-OPERATING TASKS
TO COMMUNICATE VIA SYSTEM MAINTAINED MESSAGE QUEUES.
IF THE
TARGET SYSTEM WILL CONTAIN THE PRINT SPOOLER, THIS OPTION MUST
BE INCLUDED.
*DO YOU WANT INSTALL, REQUEST, AND REMOVE ON EXIT SUPPORT? [Y/N]:
THIS FACILITY OF MCR CAN SUBSTANTIALLY REDUCE THE NUMBER OF
INSTALLED TASKS NEEDED FOR CONVENIENT SYSTEM OPERATION. SEE
DESCRIPTION OF THE RUN COMMAND IN THE RSX-llM OPERATOR'S
PROCEDURES MANUAL FOR DETAILS.
*DO YOU WANT LOGICAL DEVICE ASSIGNMENT SUPPORT? [Y/N]:
LOGICAL DEVICE ASSIGNMENT PROVIDES THE CAPABILITY TO ASSOCIATE
A LOGICAL NAME TO A PHYSICAL DEVICE UNIT. ASSIGNMENTS MAY BE
MADE ON A PER TERMINAL OR SYSTEM WIDE BASIS.
DO YOU WANT TERMINAL~HOST SYNCHRONIZATION SUPPORT? [Y/NJ:
TERMINAL-HOST SYNCHRONIZATION ALLOWS OUTPUT TO BE TURNED OFF
(CTRL S)
AND ON (CTRL Q) FROM A TERMINAL. THIS MAY BE USEFUL
FOR HIGH BAUD RATE DISPLAY TERMINALS AND IS REQUIRED IF THE
TARGET SYSTEM CONTAINS VTSO TERMINALS WITH THE COPIER OPTION.
DO YOU WANT SETTABLE CASE CONVERSION FOR TERMINAL INPUT? [Y/N]:
NORMALLY THE TERMINAL DRIVER ACCEPTS AND ECHOS CHARACTERS AS
RECEIVED.
SELECTION OF THIS OPTION PROVIDES THE CAPABILITY TO
SELECTIVELY ENABLE A TERMINAL SO THAT THE TERMINAL DRIVER WILL
AUTOMATICALLY CONVERT ALL LOWER CASE INPUT TO UPPER CASE.
Additional explanation:
Enabling a
terminal
for
case
conversion
accomplished via the MCR Set/Lower command.

3-21

SYSGEN DIALOGUE DETAILS
DO YOU WANT TRANSPARENT TERMINAL READ/WRITE SUPPORT? [Y/N]:
TERMINAL INPUT IS USUALLY FILTERED ON BOTH INPUT AND OUTPUT.
THIS IS NECESSARY TO RECOGNIZE CONTROL CHARACTERS THAT HAVE
SPECIAL MEANING
(I.E., ALTMODE OR ESC)
OR REQUIRE
FILL
CHARACTERS (I.E., CARRIAGE RETURN).
THIS OPTION PROVIDES
SUPPORT FOR I/O FUNCTIONS
(IO.RAL AND IO.WAL) THAT BYPASS
CHARACTER FILTERING.
DO YOU WANT EXECUTIVE LEVEL ROUND ROBIN SCHEDULING? [Y/N]:
ROUND ROBIN SCHEDULING PROVIDES THE CAPABILITY FOR A GROUP OF
TASKS WITHIN A RANGE OF PRIORITIES TO SHARE THE AVAILABLE
PROCESSING TIME.
EACH PRIORITY CLASS WITHIN THE RANGE IS ROUND
ROBIN SCHEDULED AT EACH SCHEDULE INTERVAL (TIME SLICE) .

Additional explanation:
Round-robin scheduling refers to the algorithm where
each task of a group of tasks is successively given
preference in allocating available processor time.
The actual algorithm used in RSX-llM is to locate the
first task within each priority class that is not
blocked from executing, and to move that task to the
end of its respective priority class.
Compute-bound
tasks tend to actually round robin, whereas I/0-bound
tasks tend to remain at the front of their respective
priority class.
If this option is not
queries are omitted.

selected,

the

three

HOW MANY CLOCK TICKS ARE TO EXPIRE BETWEEN SCHEDULE INTERVALS? [DJ :
ROUND ROBIN SCHEDULING IS EXECUTED ON A PERIODIC BASIS.
TYPICAL
SCHEDULING
INTERVALS
ARE ON THE ORDER OF 100
MILLISECONDS.
(6 CLOCK TICKS FOR A 60 CYCLE LINE FREQUENCY
CLOCK.)
RESPONSE LIMITS: 1.-1000.
WHAT IS THE HIGHEST PRIORITY CLASS TO BE CONSIDERED? [D]:
ROUND ROBIN SCHEDULING IS EXECUTED FOR A
CLASSES. TYPICAL RANGES ARE 1. TO 100.

RANGE

PRIORITY

RESPONSE LIMITS: 1.-249.
WHAT IS THE LOWEST PRIORITY CLASS TO BE CONSIDERED? [D]:
ARE YOU PLANNING TO INCLUDE A USER WRITTEN DRIVER? [Y/N]:
ADDITION OF A USER WRITTEN DRIVER REQUIRES THAT ADEQUATE SPACE
BE ALLOCATED FOR INTERRUPT VECTORS DURING THE ASSEMBLY OF THE
GENERATED EXECUTIVE. ACTUAL DRIVER INCLUSION OCCURS WHEN THE
TARGET EXECUTIVE IS BUILT.

Additional explanation:
If a user-written driver is not
the next query is omitted.

3-22

being

incorporated,

SYSGEN DIALOGUE DETAILS
WHAT IS THE ADDRESS OF THE HIGHEST DEVICE INTERRUPT VECTOR?

[O]:

DO YOU WANT TO INCLUDE THE EXECUTIVE DEBUGGING TOOL? [Y/N]:
THE EXECUTIVE DEBUGGING TOOL PROVIDES A SUBSET ODT-11 THAT RUNS
THIS TUUL ALLOWS
A ~AKT u~ Ttl~ RSX-llM EXECUTIVE.
I/O DRIVERS, AND
INTERACTIVE DEBUGGING OF EXECUTIVE MODULES,
INTERRUPT SERVICE ROUTINES.

Additional explanation:
See the RSX-llM Guide To Writing An
subset description.

I/O

Driver

for

DO YOU WANT REPORT AT SYSTEM CRASHES? [Y/N]:
IF THIS OPTION IS NOT SELECTED AND THE SYSTEM CRASHES, THEN THE
SYSTEM SIMPLY HALTS.
Additional explanation:
If the crash display option is not selected, then the
next query does not appear.
WHAT IS THE OUTPUT DEVICE CSR ADDRESS? [0]:
THE DEVICE ON WHICH THE CRASH NOTIFICATION APPEARS MUST BE A
PAPER TAPE, TELETYPE, OR LINE PRINTER COMPATIBLE DEVICE; THAT
IS, THE DEVICE MUST TRANSMIT A SINGLE BYTE AT A TIME.
THE
NORMAL CRASH OUTPUT DEVICE IS THE CONSOLE TERMINAL WHICH HAS A
CSR ADDRESS 177564.
DO YOU WANT A PANIC DUMP ROUTINE? [Y/N]:
IF THIS OPTION IS NOT SELECTED AND A SYSTEM FAILURE OCCURS,
IT
MAY NOT BE POSSIBLE FOR THE USER TO CAPTURE VITAL DATA ON THE
STATE OF THE SYSTEM AT THE TIME THE FAILURE OCCURRED.
Additional explanation:
The panic dump routine is normally used in debugging
new drivers or orivileqed tasks.
It provides a means
of generating a-formatted dump of memory on a line
printer or a terminal.
If the dump routine query is
not selected, then the next query will not appear.
WHAT IS THE OUTPUT CSR ADDRESS? [O]:
THE DEVICE ON WHICH THE FORMATTED DUMP APPEARS MUST BE A PAPER
TAPE, TELETYPE, OR LINE PRINTER COMPATIBLE DEVICE; THAT IS,
THE DEVICE MUST TRANSMIT A SINGLE BYTE AT A TIME.
THE NORMAL
CRASH DUMP DEVICE IS A LINE PRINTER WHICH HAS A CSR ADDRESS OF
177514.
**WHAT IS THE SIZE OF THE DYNAMIC STORAGE REGION IN WORDS? [D]:
A NON ZERO RESPONSE TO THIS QUERY IS REQUIRED.
A MINIMUM OF
256 WORDS WILL BE ACCEPTED TO ENSURE MINIMUM OPERATIONS.
1024
WORDS IS A REASONABLE ESTIMATE FOR ADEQUATE PERFORMANCE OF A
SET OF USER TASKS MAKING FULL USE OF THE BASELINE SYSTEM.

3-23

SYSGEN DIALOGUE DETAILS
RESPONSE LIMITS: 256.-16384.

MINUS SIZE OF EXECUTIVE

Additional explanation:
The amount of dynamic memory needed depends on the
configuration of the machine and the number of tasks
installed.
See
Appendix
C
for
a
complete
description.
DO YOU WANT ROTATING PATTERN IN DATA LIGHTS? [Y/N]:
IF SELECTED, THE SYSTEM WILL, WHEN IDLE, CAUSE THE DATA LIGHTS
TO FLICKER IN A PRONOUNCED CIRCULAR ROTATION. THE OPTION
FUNCTIONS PROPERLY ONLY ON PDP-11/45/50/70 SYSTEMS.
END EXECUTIVE SERVICE OPTIONS.
BEGIN PROCESSOR OPTIONS.
WHAT IS THE SIZE OF PHYSICAL MEMORY IN 1024 WORD BLOCKS? [D]:
THE MINIMUM MEMORY IS 16K FOR AN RSX-llM SYSTEM AND SK FOR AN
RSX-llS SYSTEM.
THE MAXIMUM MEMORY SIZE IS 1024K FOR AN
RSX-llM SYSTEM AND 124K FOR AN RSX-llS SYSTEM.
IN ORDER TO
UTILIZE MEMORY ABOVE 28K THE TARGET SYSTEM MUST CONTAIN A
MEMORY MANAGEMENT UNIT.
RESPONSE LIMITS: 8.-1024.
IS THE TARGET PROCESSOR AN LSI-11? [Y/N]:
THE
RSX-llS IS CAPABLE OF RUNNING ON ALL PDP-11 PROCESSORS.
LSI-11
IS THE SMALLEST MEMBER OF THE PDP-11 FAMILY OF
PROCESSORS AND IS CONSTRUCTED USING LARGE SCALE INTEGRATED
CIRCUIT TECHNOLOGY.
Additional explanation:
This query is posed only if an RSX-llS system is
being built and memory management was not selected.
If answered affirmatively, the next question posed is
that concerning FIS support.
DO YOU WANT EAE SUPPORT? [Y/N]:
THE EXTENDED ARITHMETIC ELEMENT (EAE) (KEll-A/B) IS AN OPTION
WHICH PERFORMS MULTIPLICATION, DIVISION, MULTIPLE POSITION
SHIFTS AND NORMALIZATION SIGNIFICANTLY FASTER THAN SOFTWARE
ROUTINES.
IT CONNECTS DIRECTLY TO THE UNIBUS AND IS PROGRAMMED
AS A PERIPHERAL, ALLOWING OVERLAP BETWEEN CPU
AND
EAE
OPERATIONS.
Additional explanation:
If the generated
system
will
contain
memory
management unit support, this query is not posed.
If
answered affirmatively, the next three queries are
omitted.

3-24

SYSGEN DIALOGUE DETAILS
DO YOU WANT FLOATING POINT PROCESSOR SUPPORT? [Y/N]:
THE PDP-11 FLOATING POINT PROCESSOR IS AN OPTIONAL ARITHMETIC
PROCESSOR.
IT
PERFORMS
ALL
FLOATING POINT ARITHMETIC
OPERATIONS AND CONVERTS DATA BETWEEN INTEGER AND FLOATING POINT
FORMATS.
Additional explanation:
the next
If this query is answered affirmatively,
query is posed, and
the FIS query is omitted.
the FIS
Otherwise, the next query is skipped, and
query is posed.
IS THE TARGET PROCESSOR A PDP-11/45/50/70? [Y/N]:
DO YOU WANT FIS SUPPORT? [Y/N]:
INCLUDES
THE FIS OPTION (WHICH REQUIRES THE EIS OPTION)
HARDWARE FLOATING POINT ADD, SUBTRACT, MULTIPLY, AND DIVIDE
INSTRUCTIONS.
DO YOU HAVE A PROGRAMMABLE CLOCK? [Y/N]:
RSX-llM/S REQUIRES A REAL-TIME CLOCK FOR ITS OPERATION. EITHER
THE
KWll-P
PROGRAMMABLE FREQUENCY CLOCK OR KWll-L LINE
FREQUENCY CLOCK MAY BE USED.
Additional explanation:
This query and the next query are not posed if the
target processor
is an LSI-11. Otherwise, if the
answer is affirmative, then the next query is posed.
WHAT IS THE NUMBER OF INTERRUPTS PER SECOND? [D]:
THE NUMBER OF INTERRUPTS SPECIFIED
INTO 10 0 0 0 ( 10} .

MUST

EVENLY

DIVISIBLE

RESPONSE LIMITS:l.-1000.

If the programmable clock option was not selected,
the previous query is replaced by the following
query.
IS YOUR LINE FREQUENCY 50 CYCLE? [Y/N] :
LINE FREQUENCY IS EITHER 50 OR 60 CYCLE.
Additional explanation:
This query is posed if the
target processor is an
LSI-11 or does not contain a programmable clock.
The
line frequency must be either 50 or 60 cycles per
second.

3-25

SYSGEN DIALOGUE DETAILS
DO YOU WANT KWll-Y WATCHDOG TIMER SUPPORT? [Y/N]:
THE KWll-Y IS A WATCHDOG TIMER THAT IS CONSTANTLY ATTEMPTING TO
TIME OUT SYSTEM ACTIVITY. THE RSX-llM/S EXECUTIVE REFRESHES
THE TIMER AT EVERY CLOCK INTERRUPT. IF THE SYSTEM FAULTS OR A
HARDWARE FAILURE OCCURS AND THE TIMER ELAPSES AN EXTERNAL
SIGNAL IS GENERATED.
Additional explanation:
If this option is selected, the generated system will
only run on configurations that contain a KWll-Y.
DO YOU WANT PARITY MEMORY SUPPORT? [Y/N]:
PARITY MEMORY PROVIDES CIRCUITRY TO CHECK THE INTEGRITY OF
STORED DATA.
PARITY CHECKING BITS ARE GENERATED WHEN DATA
WORDS ARE WRITTEN INTO MEMORY AND CHECKED WHEN THE DATA WORDS
ARE SUBSEQUENTLY READ OUT OF MEMORY.
Additional explanation:
The parity memory option provides the system with
detection of errors in the memory system. If this
query is answered affirmatively, the next query is
posed.
IS THE TARGET PROCESSOR A PDP-11/70? [Y/N]:
THE PDP-11/70 PROVIDES ADDITIONAL PARITY CHECKING CIRCUITRY.
THE BUFFER MEMORY (CACHE) USED TO SPEED UP MAIN MEMORY
OPERATIONS IS ALSO CAPABLE OF GENERATING AND CHECKING PARITY.
END PROCESSOR OPTIONS
BEGIN PERIPHERAL OPTIONS
HOW MANY ADOl-D A/D CONVERTERS DO YOU HAVE? [D] :
THE ADOl-D IS A MULTICHANNEL ANALOG DATA ACQUISITION
WHICH INTERFACES DIRECTLY TO PDP-11 COMPUTERS.

OPTION

THE STANDARD ADOl-D CONSISTS OF AN EXPANDABLE SOLID-STATE INPUT
MULTIPLEXER, PROGRAMMABLE INPUT RANGE SELECTOR, A/D CONVERTER,
CONTROL, AND BUS INTERFACE. THE MULTIPLEXER CAN BE EXPANDED BY
ADDING 4-CHANNEL MODULES UP TO 32 CHANNELS. AN EXPANSION
MULTIPLEXER MAY BE ADDED TO PROVIDE A MAXIMUM CONFIGURATION OF
64 CHANNELS.
Additional explanation:
If the answer to this query is non-zero, then the
next three queries appear. Otherwise they are not
displayed. If the reply is terminated with ALTMODE,
the subsidiary queries for this question are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT ADOl-D? [0]:130
130 IS THE STANDARD VECTOR ADDRESS FOR THE ADOl-D

3-26

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the reply to the query for the number of ADOl-Ds
was greater than 1, then this query is repeated by
SGN~
On the second and subsequent displays of this
query, however, the user is required to supply an
interrupt vector address.
WHAT IS ITS CSR ADDRESS? [0]:176770
176770 IS THE STANDARD CSR ADDRESS FOR THE ADOl-D
Additional explanation:
This query is repeated for each of the ADOl-Ds after
the first; the user supplies the CSR address.
HOW MANY A/D CHANNELS

IT HAVE? [D] :

RESPONSE LIMITS: 1.-64.
Additional explanation:
This query appears as many times as the number of
ADOl-Ds specified.
Note that the queries for the
interrupt vector, CSR, and the number of channels
occur one after the other for each ADOl-D. The
repeat sequence repeats all three of the rrnor;
....... oc::
_...,.
~--

HOW MANY AFCll A/D CONVERTERS DO YOU HAVE? [D]:
THE AFCll IS A DIFFERENTIAL ANALOG INPUT SUB-SYSTEM FOR
IDACS-11
INDUSTRIAL
CONTROL
SYSTEMS.
THE AFCll SYSTEM
MULTIPLEXES UP TO 1024 DIFFERENTIAL INPUT ANALOG SIGNALS,
SELECTS
GAIN,
AND
PERFORMS
A 13-BIT ANALOG-TO-DIGITAL
CONVERSION AT A 200 CHANNEL PER SECOND RATE UNDER PROGRAM
CONTROL.

Additional explanation:
If the answer to this query is non-zero, then the
next three queries appear. ucnerwise cney are not
displayed. If the reply is terminated with ALTMODE,
the
subsidiary queries for this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT AFCll? [0]:134
134 IS THE STANDARD VECTOR ADDRESS FOR THE AFCll.
Additional explanation:
If the reply to the query for the number of AFClls
was greater than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however, the user is required to supply an
interrupt vector address.

3-27

SYSGEN DIALOGUE DETAILS
WHAT IS ITS CSR ADDRESS? [O] :172570
172570 IS THE STANDARD CSR ADDRESS FOR THE AFCll.
Additional explanation:
This query is repeated for each of the AFClls
the first;
the user supplies the CSR address.

after

HOW MANY A/D CHANNELS DOES IT HAVE? [D] :
RESPONSE LIMITS: 1.-1024.
Additional explanation:
This query appears as many times as the number of
AFClls specified.
Note that the queries for the
interrupt vector, CSR, and
the number of channels
occur one after the other for each AFCll. The repeat
sequence repeats all three of the queries.
HOW MANY ARll LABORATORY PERIPHERAL SYSTEMS DO YOU HAVE? [D] :
THE ARll LAB PERIPHERAL SYSTEM IS A LOW-COST REALTIME SUBSYSTEM
WHICH INTERFACES TO THE PDP-11 FAMILY OF COMPUTERS. THE SYSTEM
INCLUDES A 10-BIT A/D CONVERTOR, PROGRAMMABLE REALTIME CLOCK,
AND DISPLAY CONTROL.
Aditional explanation:
If the answer to this query is non-zero,
then the
next seven queries appear. Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
then subsidiary queries for
this query are not
pre-answered.
WHAT IS THE CLOCK VECTOR ADDRESS OF THE NEXT ARll? [O]:
THE CLOCK VECTOR ADDRESS FOR THE ARll IS ASSIGNED FROM THE
FLOATING VECTOR SPACE STARTING AT 300. CONSULT FIELD SERVICE
TO DETERMINE LOCATION IF UNKNOWN.
Additional explanation:
If the number of ARlls specified was greater than
this query is repeated by SGN for each ARll.

WHAT IS ITS CSR ADDRESS? [0]:170400
170400 IS THE STANDARD CSR ADDRESS FOR THE ARll.
Additional explanation:
If the number of ARlls specified was greater than 1,
this query is repeated by SGN for each ARll. On the
second and subsequent displays of this query however,
the user is required to supply the CSR address.

3-28

SYSGEN DIALOGUE DETAILS
WHAT IS THE POLARITY MASK WORD? [O]:
THE POLARITY MASK WORD CORRESPONDS ONE-TO-ONE WITH THE 16
POSSIBLE CHANNELS.
A 1-BIT DESIGNATES A UNI-POLAR CHANNEL. A
0-BIT DESIGNATES A BI-POLAR CHANNEL.
IS THE D/A OPTION PRESENT? [Y/N] :
WHAT IS THE MAXIMUM NUMBER OF SAMPLES PER SECOND? [D] :
THE QUERY RESPONSE VALUE MUST BE EVENLY DIVISIBLE INTO 10000.
RESPONSE LIMITS: 1.-2500.
IS THE DRll-K OPTION PRESENT? [Y/N]:
THE ARll DOES NOT INCLUDE THE 16-BIT DIGITAL I/0 OPTION THAT IS
AVAILABLE WITH THE LPSll.
THE SAME FUNCTIONALITY, HOWEVER, M1\V
BE ACHIEVED BY USING A DRll-K THAT IS JUMPER SELECTED TO USE
THE 16 INDIVIDUAL INPUT LINES WITHOUT INTERRUPTS.
,-.a.C"l.~

Additional explanation:
This query appears as many times as
the number of
ARlls specified.
Note that the queries for interrupt
vector, CSR, number of A/D channels, polarity mask
word, D/A option, maximum sample rate, and DRll-K
option occur one after the other for each ARll.
The
repeat sequence repeats all seven queries.
If this
query is answered affirmatively, then the next query
is also posed.
WHAT IS ITS CSR ADDRESS? [O] :170420
170420 IS THE STANDARD CSR ADDRESS FOR THE DRll-K.
HOW MANY CM/CRll CARD READERS DO YOU HAVE? [D]:
THE CRll CARD READER READS EIA STANDARD 80-COLUMN PUNCHED DATA
CARDS AT 300 CARDS PER MINUTE; THE CMll-F READS 80-COLUMN
MARK-SENSE CARDS, WHICH CAN HAVE PUNCHED HOLES, AT 285 CARDS
PER l'lJ.l" u J.. .c:...
UT"t.."fTTmr.1

Additional explanation:
If the answer to this query is non-zero,
then
the
next two queries appear.
Otherwise
they are not
displayed.
If the reply is terminated with ALTMODE,
the subsidiary queries for
this question are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT CM/CRll? [O] :230
230 IS THE STANDARD VECTOR ADDRESS FOR THE CM/CRll.
Additional explanation:
If the reply to the query for the number of CM/CRlls
was greater
than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however,
the user
is required to supply an
interrupt vector address.

3-29

SYSGEN DIALOGUE DETAILS
WHAT IS ITS CSR ADDRESS? [OJ :177160
177160 IS THE STANDARD CSR ADDRESS FOR THE CM/CRll.
Additional explanation:
This query appears as many times as the number of
CM/CRlls specified.
Note that the queries for the
interrupt vector and CSR occur one after the other
for each CM/CRll. The repeat sequence repeats both
queries.
HOW MANY TAll DUAL CASSETTES DO YOU HAVE? [DJ:
THE TAll MAGNETIC TAPE CASSETTE SYSTEM IS A DUAL-DRIVE,
REEL-TO-REEL UNIT DESIGNED TO REPLACE PAPER TAPE.
ITS TWO
DRIVES RUN NON-SIMULTANEOUSLY
USING
PROPRIETARY
DIGITAL
EQUIPMENT
CORPORATION
PHILIPS-TYPE
CASSETTES.
THE TAll
INCLUDES A CONTROL UNIT AND A DUAL TAPE TRANSPORT.
Additional explanation:
If the answer to this query is non-zero,
then the
next two queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT TAll? [OJ:260
260 IS THE STANDARD VECTOR ADDRESS FOR THE TAll.
Additional explanation:
If the reply to the query for the number of TAlls was
greater
than 1, then this query is repeated by SGN.
On the second and subsequent display of this query,
however,
t_he user is required to supply an interrupt
vector address.
WHAT IS ITS CSR ADDRESS? [OJ:l77500
177500 IS THE STANDARD CSR ADDRESS FOR THE TAll.
Additional explanation:
This query appears as many times as the number of
TAllS specified.
Note that the queries for the
interrupt vector and CSR occur one after the other
for each TAll.
The repeat sequence repeats both
queries.
HOW MANY RJ/RWP04 DISK PACK CONTROLLERS DO YOU HAVE? [DJ:
THE RJ/RWP04 IS A 19 DATA SURFACE, REMOVABLE, RANDOM ACCESS
BULK STORAGE SYSTEM WITH A STORAGE CAPACITY OF 43,980,288 WORDS
PER PACK. THE RJ/RWP04 INCLUDES A CONTROL UNIT AND THE FIRST
DISK PACK DRIVE. THE SYSTEM IS EXPANDABLE UP TO EIGHT DRIVES.

3-30

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the answer to this query is non-zero, then the
next six queries appear.
Otherwise they are not
displayed. If the reolv is terminated with ALTMODEj
the- subsidiary queries for this query are not
pre-answered.
DO YOU WANT OFFSET RECOVERY SUPPORT? [Y/N]:
OFFSET RECOVERY IS A TECHNIQUE FOR READING DATA IN INSTANCES
WHERE IT CANNOT OTHERWISE BE RECOVERED. IT INVOLVES MOVING THE
DISK READ HEAD IN SMALL INCREMENTS ABOUT CYLINDER CENTERLINE IN
AN ATTEMPT TO PICK UP A STRONGER SIGNAL AND THUS ALLOW DATA TO
BE RECOVERED.
Additional explanation:
Offset
recovery
is
especially
important
in
environments where disk packs are frequently moved
from drive to drive. It is also recommended where a
higher degree of data reliability is required.
WHAT IS THE VECTOR ADDRESS OF THE NEXT RJ/RWP04? [0]:254
254 IS THE STANDARD VECTOR ADDRESS FOR THE RJ/RWP04.
Additional

explanation~

If the reply to the query for the number of RJ/RWP04s
was greater than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however, the user is required to supply an
interrupt vector address.
WHAT IS THE CSR ADDRESS? [0]:176700
176700 IS THE STANDARD CSR ADDRESS FOR THE RJ/RWP04.
Additional explanation:
This query is repeated Ior each of tne RJ/RWP04s
after the first;
the user supplies the CSR address.
HOW MANY DRIVES DOES IT HAVE? [D]:
Additional explanation:
This query appears as many times
RJ/RWP04 controllers specified.

the

number

WHAT IS THE PHYSICAL UNIT NUMBER OF THE FIRST DRIVE? [D]:
THE TYPICAL RESPONSE IS ZERO.
A NON-ZERO REPLY SHOULD BE
ARRIVED
AT
AFTER
CONSULTATION
WITH
A
DEC
SOFTWARE
REPRESENTATIVE.
Additional explanation:
This query appears as many times as the number of
RJ/RWP04
controllers
specified.
Note that the
queries for the interrupt vector, CSR, the number of

3-31

SYSGEN DIALOGUE DETAILS
drives, and physical unit number occur one after
the
other for each RJ/RWP04 controller.
The repeat
sequence repeats all three of the queries.
IS THE NEXT DRIVE AN RP04? [Y/N]:
Additional explanation:
This query is repeated as many times as the number of
drives previously specified.
HOW MANY RFll FIXED HEAD DISK CONTROLLERS DO YOU HAVE? [D]:
THE RFll-A IS A FAST, LOW-COST, RANDOM-ACCESS BULK-STORAGE
SYSTEM THAT PROVIDES 262,144 17-BIT WORDS (16 DATA BITS AND 1
PARITY BIT)
OF STORAGE.
UP TO EIGHT RSll DISKS CAN BE
CONTROLLED BY ONE RFll CONTROL FOR A TOTAL OF 2,047,102 WORDS
OF STORAGE. AN RFll-A INCLUDES A CONTROL UNIT AND THE FIRST
DISK DRIVE.
Additional explanation:
If the answer to this query is non-zero,
then the
next two queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT RFll? [0]:204
204 IS THE STANDARD VECTOR ADDRESS FOR THE RFll.
Additional explanation:
If the reply to the query for the number of RFlls was
greater
than 1, then this query is repeated by SGN.
On the second and subsequent displays of this query,
however,
the user is required to supply an interrupt
vector address.
WHAT IS ITS CSR ADDRESS? [O] :177460
177460 IS THE STANDARD CSR ADDRESS FOR THE RFll.
Additional explanation:
This query is repeated for each of the RFlls after
the first;
the user supplies the CSR address. Note
that the queries for interrupt vector and CSR address
The repeat
occur one after the other for each RFll.
sequence repeats both queries.
HOW MANY RKll CARTRIDGE DISK CONTROLLERS DO YOU HAVE? [D]:
THE RKll-D CARTRIDGE DISK DRIVE AND CONTROL IS A COMPLETE MASS
STORAGE SYSTEM FOR MEDIUM VOLUME, RANDOM ACCESS DATA STORAGE.
THE SYSTEM INCLUDES A MODULAR MASS STORAGE DEVICE UTILIZING
REMOVABLE
DISK CARTRIDGES AND A COMPLETE EASY-TO-PROGRAM
CONTROL. A DISK CARTRIDGE HOLDS OVER 1.2 MILLION WORDS.
THE
SYSTEM IS EXPANDABLE UP TO 9.6 MILLION WORDS PER CONTROL (8
DISKS) • AN RKll-D INCLUDES A CONTROL UNIT AND THE FIRST DISK
DRIVE.

3-32

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the answer to this query is non-zero, then the
next four queries appear.
Otherwise they are not
displayed~
If the reply is terminated with ALTMODE;
the
subsidiary queries for this query are not
pre-answered.
DO YOU WANT WRITE CHECK SUPPORT? [Y/N]:
WRITE CHECK INCREASES THE DATA RELIABILITY OF RK05 DISKS AT THE
COST OF A DEGRADATION IN DISK THROUGHPUT. WRITE CHECK, IF
SELECTED, MAY BE ENABLED/DISABLED DYNAMICALLY ON A PER DRIVE
BASIS VIA THE MCR SET COMMAND.
Additional explanation:
Write check 4 0 the process whereby the
driver
performs a read after each write and checks that the
data actually matches. If the data cannot be read or
does
not match the data written, the data is
rewritten and another read attempt is performed.
WHAT IS THE VECTOR ADDRESS OF THE NEXT RKll? [0]:220
220 IS THE STANDARD VECTOR ADDRESS FOR THE RKll.
Additional explanation:
If the reply to the query for the number of RKll-Ds
was greater than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however, the user is required to supply an
interrupt vector address.
WHAT IS ITS CSR ADDRESS? [0]:177404
177404 IS THE STANDARD CSR ADDRESS FOR THE RKll.
Additional explanation:
This query is repeated for each of cne KKiiS
the first; the user supplies the CSR address.

after

HOW MANY DRIVES DOES IT HAVE? [D] :
Additional explanation:
This query appears as many times as the number of
RKll-Ds specified.
Note that the queries for the
interrupt vector, CSR, and the number of drives occur
one after the other for each RKll-D controller. The
repeat sequence repeats all three of the queries.
HOw MANY RPllC/E DISK CONTROLLERS DO YOU HAVE? [D]:
THE RPll-C/E IS A COMPLETE MASS STORAGE SYSTEM USING A MAGNETIC
DISK PACK WITH 20 DATA SURFACES AND A MOVING READ/WRITE HEAD.
THE RPll-C/E INCLUDES A CONTROL UNIT AND THE FIRST DISK PACK
DRIVE.
THE SYSTEM IS EXPANDABLE UP TO 8 DRIVES, EACH HAVING A
CAPACITY OF 20,480,000 (RP03)
OR 10,240,000 (RPR02) 16-BIT

3-33

SYSGEN DIALOGUE DETAILS
WORDS. ACCESS TIMES ARE 29 MSEC. AVERAGE LATERAL (CYLINDER TO
CYLINDER) AND 12.5 MSEC. HALF ROTATIONAL.
Additional explanation:
If the answer to this query is non-zero, then the
next four queries appear.
Otherwise they are not
displayed. If the reply is terminated with ALTMODE,
the
subsidiary queries for this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT RPllC/E? [O] :254
254 IS THE STANDARD VECTOR ADDRESS FOR THE RPllC/E.
Additional explanation:
If the reply to the query for the number of RPCllC/Es
was greater than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however, the user is required to supply an
interrupt vector address.
WHAT IS ITS CSR ADDRESS? [0]:176714
176714 IS THE STANDARD CSR ADDRESS FOR THE RPllC/E.
Additional explanation:
This query is repeated for each of the RPllC/Es after
the first; the user supplies the CSR address.
HOW MANY DRIVES DOES IT HAVE? [D] :
Additional explanation:
This query appears as many times as the number of
RPllC/E controllers specified. Note that the queries
for the interrupt vector, CSR, and the number of
drives occur one after the other for each RPllC/E
controller. The repeat sequence repeats all three of
the queries.
IS THE NEXT DRIVE AN RPR02? [Y/N]:
Additional explanation:
This query is repeated as many times as the number of
drives previously specified.
HOW MANY RJ/RWS03-04 FIXED HEAD DISK CONTROLLERS DO YOU HAVE?[D]:
THE RJ/RWS03-04 USES THE RHll CONTROLLER. THE RJ/RWS03 HAS A
CAPACITY OF 256K WORDS; THE RJ/RWS04 HAS A CAPACITY OF 512K
WORDS. TRANSFER RATE rs 4 MICROSECONDS PER WORD WITH AN
AVERAGE LATENCY OF 8 MILLISECONDS.

3-34

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the answer to this query is non-zero, then the
next five queries appear.
Otherwise they are not
a1splayed.
If the reply is terminated with ALTMODE:
the
subsidiary queries for this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT RJ/RWS03-04? [OJ :204
204 IS THE STANDARD VECTOR ADDRESS FOR THE RJ/RWS03-04.
Additional explanation:
If the reply to the query for the number
of
RJ/RWS03-04s was greater than 1, then this query is
repeated by SGN.
On the second and subsequent
displays of this query, however, the user is required
to supply an interrupt vector address.
WHAT IS ITS CSR ADDRESS? [OJ:l72040
172040 IS THE STANDARD CSR ADDRESS FOR THE RJ/RWS03-04.
Additional explanation:
This query is repeated for each of
after the first;

the

RJ/RWS03-04s

the user supplies the CSR address.

HOW MANY DRIVES DOES IT HAVE? [DJ:
Additional explanation:
This query appears as many times as
RJ/RWS03-04 controllers specified.

the

number

SHOULD

WHAT IS THE PHYSICAL UNIT NUMBER OF THE FIRST DRIVE? [D] :
THE TYPICAL RESPONSE IS

ARRIVED
AT
AFTER
REPRESENTATIVE.

ZERO;

CONSULTATION

NON-ZERO

WITH

DEC

SOFTWARE

Additional explanation:
This query is repeated as many times as the number of
RJ/RWS03-04 controllers specified.
Note that the
queries for the interrupt vector, CSR, number of
drives, and the unit number of first drive occur one
after the other for each RJ/RWS03-04 controller. The
repeat sequence repeats all four of the queries.
IS THE NEXT DRIVE AN RS03? [Y/N]:
Additional explanation:
RS03s and RS04s can be mixed on the same controller.
This query is repeated as many times as the number of
drives previously specified.

3-35

SYSGEN DIALOGUE DETAILS
HOW MANY TCll DECTAPE CONTROLLERS DO YOU HAVE? [D] :
THE TCll-G IS A DUAL-UNIT BIDIRECTIONAL MAGNETIC-TAPE TRANSPORT
SYSTEM FOR AUXILIARY DATA STORAGE.
THE SYSTEM STORES INFORMATION AT FIXED POSITIONS ON MAGNETIC
TAPE AS IN MAGNETIC DISK OR DRUM STORAGE DEVICES, RATHER THAN
AT UNKNOWN OR VARIABLE POSITIONS AS IN CONVENTIONAL MAGNETIC
TAPE SYSTEMS.
A DECTAPE SYSTEM CONSISTS OF UP TO 4 DUAL TRANSPORTS, AND A
CONTROL UNIT (WHICH WILL BUFFER AND CONTROL INFORMATION FOR UP
TO FOUR DUAL TRANSPORTS) . A TCll-G INCLUDES A CONTROL UNIT AND
THE FIRST DUAL TAPE TRANSPORT.
Additional explanation:
If the answer to this query is non-zero,
then the
next three queries appear. Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the subsidiary queries for
this query will not be
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT TCll? [0] :214
214 IS THE STANDARD VECTOR ADDRESS FOR THE TCll.
Additional explanation:
If the reply to the query for the number of TCll-Gs
was greater
than 1, then this query is repeated by
SGN. On the second and subsequent displays of this
query, however,
the user
is required to supply an
interrupt vector address.
WHAT IS ITS COMMAND REGISTER ADDRESS? [O] :177342
177342 IS THE STANDARD COMMAND REGISTER ADDRESS FOR THE TCll.
Additional explanation:
This query is repeated for each of the TCll-Gs after
the first;
the user supplies the command register
address.
HOW MANY DRIVES DOES IT HAVE? [D] :
Additional explanation:
This query appears as many times as the number of
TCll-G controllers specified. Note that the queries
for the interrupt vector, command register, and
the
number of drives occur one after the other for each
TCll controller. The repeat sequence repeats all
three of the queries.
HOW MANY RXll FLEXIBLE DISK CONTROLLERS DO YOU HAVE? [D]:
THE RXll IS AN INDUSTRY-COMPATIBLE DISK STORAGE SYSTEM FOR
FLEXIBLE (FLOPPY) DISKS. AN RXll IS CAPABLE OF CONTROLLING TWO
FLEXIBLE DISK DRIVES EACH HAVING A CAPACITY OF 124,128 16-BIT
WORDS.

3-36

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the answer to this query is non-zero,
then the
next three queries appear. Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the subsidiary queries for
this query will not be
pre-answered.
WHAT IS THE VECTOR OF THEN NEXT RXll? [O] :264
264 IS THE STANDARD VECTOR ADDRESS FOR THE RXll.
Additional explanation:
If the reply to the query for the number of RXlls was
greater than 1, then this query is repeated by SGN.
On the second and subsequent displays of this query,
however, the user is required to supply an interrupt

vector address.
WHAT IS ITS CSR ADDRESS? [0]:177170
177170 IS THE STANDARD CSR ADDRESS FOR THE RXll.
Additional explanation:
This query is repeated for each of the RXlls
the first;
the user supplies the CSR address.

after

HOW MANY DRIVES DOES IT HAVE? [D]:
Additional explanation:
This query appears as many times as the number of
RXll controllers specified.
Note that the queries
for the interrupt vector, CSR, and the number of
drives occur one after the other for each RXll
controller. The repeat sequence repeats all three of
the queries.
HOW MANY LP/LS/LVll LINE PRINTERS DO YOU HAVE? [D]:
THE LPll HIGH SPEED LINE PRINTER IS AVAILABLE IN SIX MODELS.
LPll-F
LPll-H
LPll-J
LPll-K
LPll-R
LPll-S
LPll-V
LPll-W

80 COLUMN, 64 CHARACTER
80 COLUMN, 96 CHARACTER
132 COLUMN, 64 CHARACTER
132 COLUMN, 96 CHARACTER
132 COLUMN, 64 CHARACTER (HEAVY DUTY, HIGH SPEED)
132 COLUMN, 96 CHARACTER (HEAVY DUTY, HIGH SPEED)
132 COLUMN, 64 CHARACTER
132 COLUMN, 96 CHARACTER

THE LSll LINE PRINTER IS A MEDIUM-SPEED IMPACT DOT-MATRIX LINE
PRINTER FOR THE PDP-11 FAMILY. THE LINE PRINTER PRINTS AT A
RATE OF 165 CHARACTERS PER SECOND AT 10 CHARACTERS PER INCH
WITH UP TO 132 CHARACTERS PER LINE.
THE LVll IS A HIGH-SPEED ELECTROSTATIC PRINTER-PLOTTER.
THE PRINT OPTION IS SUPPORTED UNDER RSX-llM/S.

3-37

ONLY

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the answer to this query is non-zero,
then the
next five queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
DO YOU WANT FAST PRINTER SUPPORT? [Y/N] :
TO ENSURE MAXIMUM PRINT THROUGHPUT ON
SELECT THIS OPTION.

THE

LPll-R

LPll-S,

WHAT IS THE VECTOR ADDRESS OF THE NEXT LP/LS/LVll? [0]:200
200 IS THE STANDARD VECTOR ADDRESS FOR THE LP/LS/LPll.
Additional explanation:
If the reply to the query for
the number
of
LP/LS/LVlls was greater
than 1, then this query is
repeated by SGN.
On the second and subsequent
displays of this query, however, the user is required
to supply an interrupt vector address.
WHAT IS ITS CSR ADDRESS? [0]:177514
177514 IS THE STANDARD CSR ADDRESS FOR THE LP/LS/LVll.
Additional explanation:
This query is repeated for each of the LP/LS/LVlls
after the first;
the user supplies the CSR address.

rs IT AN LSll OR LPll-V/W? [Y/N]:
FAILURE TO PROPERLY
FAILURES.

SPECIFY

PRINTER

TYPE

MAY

CAUSE

OUTPUT

Additional explanation:
This query is repeated for each LP/LS/LVll specified.
IS IT A 132 COLUMN PRINTER? [Y/N]:
FAILURE TO PROPERLY SPECIFY LINE WIDTH WILL
TRUNCATION. A NO ANSWER DEFAULTS TO 80 COLUMNS.

CAUSE

BUFFER

Additional explanation:
This query appears as many times as the number of
printers specified.
Note that the queries for the
interrupt vector, CSR, LSll, and line width, occur
one after the other for each printer specified. The
repeat sequence repeats all four of the queries.
HOW MANY LPSll LABORATORY PERIPHERAL SYSTEMS DO YOU HAVE? [D] :
THE LPSll LABORATORY PERIPHERAL SYSTEM IS A REAL-TIME SUBSYSTEM
WHICH INTERFACES TO THE PDP-11 FAMILY OF COMPUTERS. THE SYSTEM

3-38

SYSGEN DIALOGUE DETAILS
HOUSES A 12-BIT A/D CONVERTER, PROGRAMMABLE REAL-TIME
DISPLAY CONTROLLER AND A 16-BIT DIGITAL I/0 OPTION.

CLOCK,

Additional explanation:
If the answer to this query is non-zero, then the
next eight queries appear. Otherwise they are not
displayed. If the reply is terminated with ALTMODE,
the
subsidiary queries for this query are not
pre-answered.
WHAT IS THE CLOCK VECTOR ADDRESS OF THE NEXT LPSll? [O]:
THE CLOCK VECTOR ADDRESS FOR THE LPSll IS ASSIGNED FROM THE
FLOATING VECTOR SPACE STARTING AT 300. CONSULT FIELD SERVICE
TO DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the number of LPSlls specified was greater than 1,
this query is repeated by SGN for each LPSll.
WHAT IS ITS CSR ADDRESS?

[0]:170400

170400 IS THE STANDARD CSR ADDRESS FOR THE LPSll.
Additional explanation:
This query is repeated for each of the LPSlls
the first; the user supplies the CSR address.

after

HOW MANY A/D CHANNELS DOES IT HAVE? [D]:
RESPONSE LIMITS: 1.-64.
Additional explanation:
If the number of A/D channels is 16 or less, the
following query is posed; otherwise it is omitted.
IS THE GAIN RANGING OPTION PRESENT? [Y/N]:
Additional explanation:
If the gain ranging option is present, then the
following query is posed; otherwise it is omitted.
WHAT IS THE POLARITY MASK WORD? [O]:
THE POLARITY MASK WORD CORRESPONDS ONE-TO-ONE WITH THE 16
POSSIBLE CHANNELS. A 1-BIT DESIGNATES A UNI-POLAR CHANNEL. A
0-BIT DESIGNATES A BI-POLAR CHANNEL.
IS THE D/A OPTION PRESENT? [Y/N]:
Additional explanation:
If the D/A option is present, then the
query is posed; otherwise it is omitted.

3-39

following

SYSGEN DIALOGUE DETAILS
HOW MANY D/A CHANNELS DOES IT HAVE? [D] :
RESPONSE LIMITS: 1.-10.
Additional explanation:
The first two D/A channels correspond to the X and Y
registers of the display controller. Up to eight
additional D/A channels may be specified if the
external D/A option is present.
WHAT IS THE MAXIMUM NUMBER OF SAMPLES PER SECOND? [D] :
THE QUERY RESPONSE VALUE MUST BE EVENLY DIVISIBLE INTO 10000.
RESPONSE LIMITS: 1.-2500.
Additional explanation:
This query appears as many times as the number of
LPSlls
specified.
Note
that the queries for
interrupt vector, CSR, number of A/D channels, gain
ranging, D/A option, and maximum sample rate occur
one after the other for each LPSll.
The repeat
sequence repeats all six of the queries.
HOW MANY TJ/TWU16 MAGTAPE CONTROLLERS DO YOU HAVE? [D] :
THE TJ/TWU16 IS A MAGNETIC TAPE SYSTEM IDEALLY SUITED FOR
WRITING, READING, AND STORING LARGE VOLUMES OF DATA AND
PROGRAMS IN A SERIAL MANNER.
THE 10 1/2-INCH TAPE REELS
CONTAIN UP TO 2400 FEET OF TAPE. THE TAPE IS 9-TRACK, RECORDED
EITHER NRZI OR PHASE ENCODED WITH RECORDING DENSITIES OF 800 OR
1600 BPI.
Additional explanation:
If the response to this query is non-zero,
then the
next four queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT TJ/TWU16? [0]:224
224 IS THE STANDARD VECTOR ADDRESS FOR THE TJ/TWU16.
Additional explanation:
If the reply to the number of TJ/TWU16s was greater
than 1,
this query is repeated by SGN for each
TJ/TWU16. On the second and subsequent displays of
this query, however, the user is required to supply
an interrupt vector address.
WHAT IS ITS CSR ADDRESS? [O] :172440
172440 IS THE STANDARD CSR ADDRESS FOR THE TJ/TWU16.

3-40

SYSGEN DIALOGUE DETAILS
Additional explanation:
This query is repeated for each of the TJ/TWU16s
after the first;
the user supplies the CSR address.
HOW MANY DRIVES DOES IT HAVE? [D]:
Additional explanation:
This query appears as many times
TJ/TWU16 controllers specified.

the

number

WHAT IS THE PHYSICAL UNIT NUMBER OF THE TM02 FORMATTER? [D]:
A TYPICAL RESPONSE IS ZERO;
A NON-ZERO RESPONSE SHOULD BE
ARRIVED
AT
AFTER
CONSULTATION
WITH
A
DEC
SOFTWARE
REPRESENTATIVE.
Additional explanation:
This query appears as many times as the number of
TJ/TWU16s specified.
Note that the queries for the
interrupt vector, CSR, the number of drives, and
the
physical unit number occur one after the other for
each TJ/TWU16. The repeat sequence repeats all four
of the queries.
HOW MANY TM/TMAll MAGTAPE CONTROLLERS DO YOU HAVE? [D]:
THE TM/TMAll IS A MAGNETIC TAPE SYSTEM IDEALLY SUITED FOR
WRITING, READING, AND STORING LARGE VOLUMES OF DATA AND
PROGRAMS IN A SERIAL MANNER.
THE SYSTEM READS AND WRITES IN
INDUSTRY-COMPATIBLE FORMAT.
TUlO DRIVES SUPPORT 10 1/2-INCH
REELS CONTAINING UP TO 2400 FEET OF TAPE UPON WHICH OVER 180
MILLION BITS OF DATA CAN BE STORED ON HIGH DENSITY 9-TRACK TAPE
OR OVER 140 MILLION BITS CAN BE STORED ON HIGH DENSITY 7-TRACK
TAPE.
TS03 DRIVES SUPPORT 7-INCH REELS CONTAINING UP TO 600
FEET OF TAPE IN HIGH DENSITY 9-TRACK FORMAT.
Additional explanation:

If the response to this query is non-zero,
c.ueu
the
next three queries appear; otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT TM/TMAll?

[0] :224

224 IS THE STANDARD VECTOR ADDRESS FOR THE TM/TMAll.
Additional explanation:
If the reply to the query for the number of TM/TMAlls
was greater than 1, this query is repeated by SGN for
each TM/TMAll. On the second and subsequent displays
of this query, however,
the user
is required to
supply an interrupt vector address.
WHAT IS ITS COMMAND REGISTER ADDRESS?

3-41

[0]:172522

SYSGEN DIALOGUE DETAILS
172522 IS THE STANDARD COMMAND REGISTER ADDRESS FOR THE TM/TMAll.
Additional explanation:
This query is repeated for each of the TM/TMAlls
after the first;
the user supplies the command
register address.
HOW MANY DRIVES DOES IT HAVE?[D]:
Additional explanation:
This query appears as many times as the number of
TM/TMAlls specified.
Note that the queries for the
interrupt vector, command register, and the number of
drives occur one after the other for each TM/TMAll.
The repeat sequence repeats all three of the queries.
HOW MANY PCll PAPER TAPE READER/PUNCHES DO YOU HAVE? [D]:
THE PCll IS A HIGH SPEED PAPER TAPE READER/PUNCH THAT IS
CAPABLE OF READING EIGHT-HOLE UNOILED PERFORATED PAPER TAPE AT
300 CHARACTERS PER SECOND, AND PUNCHING TAPE AT 50 CHARACTERS
PER SECOND.
Additional explanation:
If the answer to this query is non-zero, then the
next two queries appear.
Otherwise they are not
displayed. If the reply is teminated with ALTMODE,
the subsidiary queries for this query will not be
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT PCll? [0]70
70 IS THE STANDARD VECTOR ADDRESS FOR THE PCll.
Additional explanation:
If the reply to the query for the number of PClls was
greater than 1, then this query is repeated by SGN.
On the second and subsequent displays of this query,
however, the user is required to supply an interrupt
vector address.
WHAT IS ITS CSR ADDRESS? [0]:177550
177550 IS THE STANDARD CSR ADDRESS FOR THE PCll.
Additional explanation:
This query is repeated for each of the PClls after
the first; the user supplies the CSR address. Note
that queries for interrupt vector and CSR occur one
after the other for each PCll. The repeat sequence
repeats both queries.
HOW MANY PRll PAPER TAPE READERS DO YOU HAVE {EXCLUDING PCllS)? [DJ:

3-42

SYSGEN DIALOGUE DETAILS
THE PRll IS A HIGH SPEED PAPER TAPE READER THAT IS CAPABLE OF
READING
EIGHT=HOLE UNOILED PERFORATED PAPER TAPE AT 300
CHARACTERS PER SECOND.

70 IS THE STANDARD VECTOR ADDRESS FOR THE PRll.
Additional explanation:
If the reply to the query for the number of PRlls was
greater than 1, then this query is repeated by SGN.
On the second and subsquent displays of this query,
however,
the user is required to supply an interrupt
vector address.
WHAT IS ITS CSR ADDRESS? [0]:177550

177550 IS THE STANDARD CSR ADDRESS FOR THE PRll.
Additional explanation:
This query is repeated for each of the PRlls after
the first;
the user supplies the CSR address. Note
that queries for interrupt vector and CSR occur one
after the for each PRll. The repeat sequence repeats
both queries.
HOW MANY DLll LINE INTERFACES DO YOU HAVE (INCLUDING CONSOLE)? [D]:
NOTE

THE OPERATOR CONSOLE
VIA A DLll.

MUST

ATTACHED

THE DLll SERIES OF ASYNCHRONOUS SINGLE LINE INTERFACES HANDLE
FULL OR HALF DUPLEX COMMUNICATION BETWEEN A WIDE VARIETY OF
SERIAL COMMUNICATION CHANNELS AND A PDP-11 COMPUTER.
WITH A DLll INTERFACE, A PDP-11 COMPUTER CAN COMMUNICATE WITH A
LOCAL TERMINAL SUCH AS A CONSOLE DECWRITER.
Additional explanation:
If the answer to this query is non-zero,
then the
next two queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the subsidiary queries for
this query will not be
pre-answered.

3-43

SYSGEN DIALOGUE DETAILS
WHAT IS THE VECTOR ADDRESS OF THE NEXT DLll? [0]:60
60 IS THE STANDARD VECTOR ADDRESS FOR THE FIRST
DLll.
ADDITIONAL VECTORS ARE ASSIGNED FROM THE FLOATING VECTOR SPACE
STARTING AT 300. CONSULT FIELD SERVICE TO DETERMINE LOCATIONS
IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DLlls was
greater
than 1, then this query is repeated by SGN.
On the second and subsequent displays of this query,
however,
the user is required to supply an interrupt
vector address.
WHAT IS ITS CSR ADDRESS? [OJ :177560
177560 IS
ADDITIONAL
TO 176770.
UNKNOWN.

THE STANDARD CSR ADDRESS FOR THE FIRST DLll.
DLll CSR ADDRESSES ARE ASSIGNED IN THE RANGE 176000
CONSULT FIELD SERVICE TO DETERMINE LOCATIONS IF

Additional explanation:
This query is repeated for each of the DLlls after
the first;
the user supplies the CSR address. Note
that queries for interrupt vector and CSR occur one
after
the other for each DLll. The repeat sequence
repeats both of the queries.
HOW MANY DHll ASYNCHRONOUS LINE MULTIPLEXERS DO YOU HAVE? [D] :
THE DHll MULTIPLEXER CONNECTS THE PDP-11 WITH 16
SERIAL
COMMUNICATIONS
LINES
OPERATING WITH
PROGRAMMABLE PARAMETERS.
THESE PARAMETERS ARE:

ASYNCHRONOUS
INDIVIDUALLY

NUMBER OF STOP BITS
TRANSMITTER SPEED (BAUD)
RECEIVER SPEED (BAUD)
Additional explanation:
If the answer to the this query is non-zero, then the
next six queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE TOTAL NUMBER OF REMOTE DIAL-UP LINES? [D]:
REMOTE LINES WILL OCCUPY CONSECUTIVE LINE LOCATIONS, AND IF
MORE THAN 16 REMOTE LINES ARE SPECIFIED THEY MUST ALSO OCCUPY
CONSECUTIVE DHll'S. ASSIGNMENT OF LINES IS BY GROUPS OF 4.
WHAT IS THE TOTAL NUMBER OF LOCAL LINES? [D] :
LOCAL LINES WILL ALSO OCCUPY CONSECUTIVE LINE AND DHll'S.
THEIR ASSIGNMENT FOLLOWS THE ASSIGNMENT OF ANY REMOTE LINES
SPECIFIED.

3-44

SYSGEN DIALOGUE DETAILS
LOCAL LINES CAN BE EITHER EIA OR 20 MILLIAMP.
EACH TYPE MUST
HAVE AN APPROPRIATE INTERFACE, AND THE INTERFACE SERVICES A
GROUP OF FOUR LINES OF THE SAME TYPE.
THE RESPONSE TO THE
LOCAL LINE QUERY MAY BE LARGER THAN THE ACTUAL NUMBER OF LOCAL
LINES, SINCE ALLOCATION IS BY GROUPS OF FOUR.
HENCE IF SIX EIA
LINES ARE INSTALLED, TWO LINE GROUP SLOTS, WHICH CONTAIN EIGHT
LINES, ARE NEEDED.
THE REQUIRED RESPONSE TO THE LOCAL LINE QUERY CAN BE CALCULATED
BY SUMMING THE NUMBER OF EIA AND 20 MILLIAMP LINE GROUPS (EACH
LINE GROUP CONTAINS SLOTS FOR FOUR LINES)
AND MULTIPLYING BY
FOUR.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DHll? [O]:
THE VECTOR ADDRESS FOR ~HE DHll IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DHlls was
greater
than 1,
then this query is repeated by SGN
for each DHll.
WHAT IS ITS CSR ADDRESS? [0]:
THE CSR ADDRESS FOR THE DHll IS ASSIGNED FROM THE FLOATING
ADDRESS SPACE STARTING AT 160010. CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DHlls was
greater
than 1,
then this query is repeated by SGN
for each DHll.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DMll-BB?

[O]:

THE VECTOR ADDRESS FOR THE DMll-BB IS ASSIGNED FROM THE
FLOATING VECTOR SPACE STARTING AT 300. CONSULT FIELD SERVICE
TO DETERMINE THE LOCATION IF UNKNOWN.
IF A USER HAS SPECIFIED
THE EXISTENCE OF REMOTE DIAL-UP LINES THE DMll-BB INTERFACE IS
REQUIRED.
ONE DMll-BB CAN SERVICE UP TO 16 REMOTE DIAL-UP
LINES.
Additional explanation:
If the reply to the query for the number of remote
lines was non-zero, this query is repeated by SGN for
each DMll-BB that is required.
WHAT IS ITS CSR ADDRESS?
170500

[0]:170500

rs THE STANDARD CSR ADDRESS FOR THE DMll-BB.

3-45

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the reply to the query for the number of remote
lines was non-zero, this query is repeated by SGN for
each DMll-BB that is required. Note that the queries
for remote dial-up lines, local lines, DHll interrupt
vector, DHll CSR, DMll-BB interrupt vector, and
DMll-BB CSR occur one after the other for each DHll.
The repeat sequence repeats all six of the queries.
HOW MANY DJll ASYNCHRONOUS LINE MULTIPLEXERS DO YOU HAVE? [D] :
THE DJll IS A MULTIPLEXER INTERFACE BETWEEN 16 ASYNCHRONOUS
SERIAL DATA-COMMUNICATION CHANNELS AND THE PDP-11 UNIBUS. THE
DJll IS A UNIT WHOSE CHARACTER FORMATS AND OPERATING SPEEDS ARE
JUMPER OR STRAP SELECTABLE IN GROUPS OF FOUR LINES.
Additional explanation:
If the answer to this query is non-zero, then the
next three queries appear; otherwise they are not
displayed.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DJll? [O]:
THE VECTOR ADDRESS FOR THE DJll IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DJlls was
non-zero, this query is repeated by SGN for each
DJll.
WHAT IS ITS CSR ADDRESS? [0]:
THE CSR ADDRESS FOR THE DJll IS ASSIGNED FROM THE FLOATING
ADDRESS SPACE STARTING AT 160010. CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DJlls was
non-zero, this query is repeated by SGN for each
DJll.
HOW MANY LINES DOES IT HAVE? [D]:
Additional explanation:
This query appears as many times as the number of
DJlls specified.
Note that the queries for the
interrupt vector, CSR, and the number of lines occur
one after the other for each DJll.
The repeat
sequence repeats all three of the queries.

3-46

SYSGEN DIALOGUE DETAILS
HOW MANY ICS/ICR-11 INDUSTRIAL SUBSYSTEMS DO YOU HAVE? [D] :
THE ICS-11 AND ICR-11 ARE PROCESS I/O SUBSYSTEMS WHICH PROVIDE
ANALOG AND DIGITAL INTERFACE CAPABILITIES BETWEEN THE PDP-11
COMPUTER SYSTEMS AND TYPICAL INDUSTRIAL INSTRUMENTATION AND
TRANSDUCERS.
THE ICS-11 IS A LOCAL SUBSYSTEM FOR DIRECT
CONNECTION TO THE PDP-11 UNIBUS. THE ICR-11 MAY BE REMOTELY
LOCATED FROM THE PDP-11 UP TO A MAXIMUM OF 6000 FEET (1830
METERS) AND CONNECTED VIA A HIGH SPEED SERIAL LINE INTERFACE.
EACH ICS/ICR-11 CAN ACCOMMODATE UP TO 16 FUNCTIONAL I/O
MODULES.
RSX-llM/S SUPPORTS ANY MIXTURE OF ICS-11 AND ICR-11
SUBSYSTEMS TO A MAXIMUM OF 12.
Additional explanation:
If the answer to this query is non-zero,
then the
next 20 queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
DO YOU WANT TASK ACTIVATION FROM UNSOLICITED INTERRUPTS? [Y/N] :
THE RSX-llM/S ICS/ICR-11 DRIVER PROVIDES THE CAPABILITY TO
INITIATE A TASK AS THE RESULT OF AN UNSOLICITED (EXTERNAL)
INTERRUPT.
IF THIS OPTION IS SELECTED, SUPPORT IS PROVIDED TO
ALLOW TASKS TO BE LINKED TO INTERRUPTS AND TO OBTAIN ACTIVATION
DATA WHEN INITIATED~
WHAT IS THE VECTOR ADDRESS OF THE NEXT ICS/ICR-11? 10]:234
234 IS THE STANDARD VECTOR ADDRESS FOR THE FIRST ICS/ICR-11.
ADDITIONAL VECTORS ARE ASSIGNED FROM THE FLOATING VECTOR SPACE
STARTING AT 300. CONSULT FIELD SERVICE TO DETERMINE LOCATIONS
IF UNKNOWN.
Additional explanation:
If the reply to the query for
the number
of
ICS/ICR-lls was greater than 1, then this query is

repeated

~u~.

the

second

ana

subsequent

displays of the query, however, the user is required
to supply an interrupt vector address.
WHAT IS ITS CSR ADDRESS? [0]:171776
171776 IS THE STANDARD CSR ADDRESS FOR THE FIRST ICS/ICR-11.
ADDITIONAL ICS/ICR-11 CSR ADDRESSES ARE ASSIGNED STARTING AT
171776 AND PROCEED DOWNWARD AT EIGHT BYTE INTERVALS
(E.G.
171766, 171756, 171746, ETC.).
Additional explanation:
This query is repeated for each of the ICS/ICR-lls
after the first;
the user supplies the CSR address.
IS IT AN ICR-11? [Y/N]:

3-47

SYSGEN DIALOGUE DETAILS
Additional explanation:
If the reply to the query for the number
of
ICS/ICR-lls was non-zero, this query is repeated by
SGN for each ICS/ICR-11.
HOW MANY A/D CONVERTOR MODULES DOES THIS SUBSYSTEM HAVE? [D]:
THE IAD-IA IS A MERCURY
DIGITAL CONVERTOR WITH
INPUT.

WETTED FLYING CAPACITOR ANALOG TO
8 CHANNELS OF WIDE RANGE DIFFERENTIAL

NOTE
MODULE NUMBERS REPRESENT THE RELATIVE
POSITION WITHIN AN ICS/ICR-11 AND START
AT ZERO AND PROCEED TO 15.
Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non-zero value, then the next two queries are
posed.
WHAT IS THE MODULE NUMBER OF THE NEXT A/D CONVERTOR? [D] :
Additional explanation:
This query is repeated as many times as the number of
A/D convertors specified.
HOW MANY MULTIPLEXERS ARE ASSOCIATED WITH IT? [D]:
THE IMX-IA IS A 16 CHANNEL EXPANDER MULTIPLEXER FOR THE IAD-IA
A/D CONVERTOR.
UP TO 7 IMX-IA MODULES MAY BE ADDED TO EACH
IAD-IA A/D CONVERTOR. ALL IMX-IA MODULES MUST BE PHYSICALLY
ADJACENT TO THEIR ASSOCIATED IAD-IA A/D CONVERTOR.
Additional explanation:
This query is repeated as many times as the number of
A/D converters specified. Note that the queries for
module number and number of multiplexers occur one
after the other for each A/D convertor. The repeat
sequence repeats both queries.
HOW MANY D/A CONVERTOR MODULES DOES THIS SUBSYSTEM HAVE? [D] :
THE IDA-OA IS A 4 CHANNEL DIGITAL TO ANALOG CONVERTOR
OF DRIVING 0 TO 10 VOLTS OR 0 TO 50 MILLIAMPS.

CAPABLE

Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non-zero value, then the next query is posed.

3-48

SYSGEN DIALOGUE DETAILS
WHAT IS THE MODULE NUMBER OF THE NEXT D/A CONVERTOR? [DJ :
Additional explanation:
This query is repeated as many times as the number of
D/A convertors specified.
HOW MANY SINGLE-SHOT OUTPUT MODULES DOES THIS SUBSYSTEM HAVE? (DJ :
THE SINGLE-SHOT OUTPUTS AVAILABLE FOR THE ICS/ICR-11 INCLUDE
THE !DC-OB DC VOLTAGE OUTPUT AND THE !AC-OB AC VOLTAGE OUTPUT
MODULES.
Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non=zero value, then the next query is posed.
WHAT IS THE MODULE NUMBER OF THE NEXT SINGLE-SHOT OUTPUT MODULE? [DJ:
Additional explanation:
This query is repeated as_ many times as the number of
single-shot output modules specified.
HOW MANY BI-STABLE OUTPUT MODULES DOES THIS SUBSYSTEM HAVE? [D]:
THE BI-STABLE DIGITAL OUTPUT MODULES FOR THE ICS/ICR-11 INCLUDE
THE IRL-OA AND IRL-OB RELAY OUTPUTS, AND THE IAC-OA AC VOLTAGE
OUTPUTS.
Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non-zero value, then the next query is posed.
WHAT IS THE MODULE NUMBER OF THE NEXT BI-STABLE OUTPUT MODULE? [D] :
Additional explanation:
This query is repeated as many times as the number of
bi-stable output modules specified.
HOW MANY DIGITAL SENSE MODULES DOES THIS SUBSYSTEM HAVE? [DJ:
THE DIGITAL SENSE MODULES FOR THE ICS/ICR-11 INCLUDE THE !DC-IA
ISOLATED DC INPUT, THE IDC-ID NON-ISOLATED INPUT, AND THE
IAC-IA AC INPUT.
Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non-zero value, then the next query is posed.
WHAT IS THE MODULE NUMBER OF THE NEXT DIGITAL SENSE MODULE? [DJ:

3-49

SYSGEN DIALOGUE DETAILS
Additional explanation:
This query is repeated as many times as the number of
digital sense modules specified.
HO~

MANY DIGITAL INTERRUPT MODULES DOES THIS SUBSYSTEM HAVE? [DJ:
THE DIGITAL INTERRUPT MODULES FOR THE ICS/ICR-11 INCLUDE THE
IDC-iB ISOLATED DC INTERRUPT, THE IDC-IE NON-ISOLATED DC
INTERRUPT, AND THE IAC-IB AC INTERRUPT.
Additional

explanation~

This query appears as many times as the number of
ICS/ICR-lls specified.
If this query is answered
with a non-zero value, then the next query is posed.
WHAT IS THE MODULE NUMBER OF THE NEXT DIGITAL INTERRUPT MODULE? [DJ:
Additional explanation:
This query is repeated as many times as the number of
digital interrupt modules specified.
HOW MANY I/O COUNTER MODULES DOES THIS SUBSYTEM HAVE? [DJ:
THE IDC-IC IS A 16 BIT BINARY UP-COUNTER WITH PRESCALAR
AND BOTH EXTERNAL TRIGGER AND OVERFLOW OUTPUTS.

INPUT

Additional explanation:
This query appears as many times as the number of
ICS/ICR-lls specified.
Note that the queries for
interrupt vector, CSR, and ICS/ICR-11 and the key
queries
for
"A/D
converters,
D/A
converters,
single-shot outputs, bi-stable
outputs,
digital
sense, digital interrupt, and I/O counters occur one
after the other for each ICS/ICR-11.
The repeat
sequence repeats all ten queries. If the answer to
this query is non-zero, then the next query is posed.
WHAT IS THE MODULE NUMBER OF THE NEXT I/O COUNTER MODULE? [DJ:
Additional explanation:
This query is repeated as many times as the number of
I/O counter modules specified.
DO YOU WANT MODULE STATUS RESTORED ON POWER AND ERROR RECOVERY? [Y/N]:
THE RSX-llM/S ICS/ICR-11 DRIVER WILL OPTIONALLY RESTORE THE
STATE OF ALL BI-STABLE DIGITAL OUTPUTS, ANALOG OUTPUTS, AND I/O
PRESCALAR VALUES ON POWER RECOVERY AND RECOVERY FROM SUSTAINED
ICR-11 SERIAL TRANSMISSION ERRORS.
Additional explanation:
This query is posed only if the target system
contains any bistable output, D/A converter, or I/O
counter modules.

3-50

SYSGEN DIALOGUE DETAILS
HOW MANY SERIAL LINE ERRORS ARE ACCEPTABLE IN ONE SECOND? [D] :
THE RSX-llM/S ICS/ICR-11 DRIVER WILL PLACE AN ICR-11 SUBSYSTEM
OFFLINE IF THE SERIAL LINE ERRORS EXCEED A PREDEFINED LIMIT
DURING ONE SECOND OF ELAPSED TIME.
THE SUBSYSTEM
WILL
AUTOMATICALLY BE PLACED ONLINE WHEN SERIAL LINE ERRORS HAVE
STOPPED.
RESPONSE LIMITS: 1.-5000.
Additional explanation:
This query is posed only if
contains an ICR-11 subsystem.

the

target

system

HOW MANY UDCll UNIVERSAL DIGITAL CONTROLLERS DO YOU HAVE? [D]:
THE UDCll IS AN INPUT/OUTPUT OPTION FOR INDUSTRIAL AND PROCESS
CONTROL APPLICATIONS THAT USE THE PDP-11 COMPUTER.
THE UDCll
INTERROGATES OR DRIVES UP TO 252 DIRECTLY ADDRESSABLE DIGITAL
SENSE AND CONTROL FUNCTIONAL I/0 MODULES OR UP TO 4032
INDIVIDUAL DIGITAL POINTS.
RSX-llM/S SUPPORTS ONLY ONE UDCll CONTROLLER.
ANSWERS TO THIS QUERY ARE 0 AND 1.

THE

ONLY

VALID

Additional explanation:
If the answer to this query is non-zero,
then
the
next 17 queries appear.
Otherwise
they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS FOR THE UDCll? [O] :234
234 IS THE STANDARD VECTOR ADDRESS FOR THE UDCll.
WHAT IS ITS CSR ADDRESS?

[OJ :171776

171776 IS THE STANDARD CSR ADDRESS FOR THE UDCll.
Additional explanation:
The UDCll can support up to 252 modules, which can be
classified into seven types:
1.
2.
3.
4.
5.

6.
7.

Analog input;
Analog output;
Digital interrupt;
Digital sense;
Bistable output;
Single-shot output, and
I/O counters.

RSX-llM/RSX-llS requires that modules of
the same
type occupy consecutive locations within the UDCll.
Thus, if there are seven analog input modules and the
first module starts at slot 0, then the last must be
at slot 6.
Obviously, module types cannot overlap,
though unassigned slots are permitted.

3-51

SYSGEN DIALOGUE DETAILS
The queries
boundaries.

which

establish

the

module

HOW MANY ADUOl ANALOG INPUT MODULES DO YOU HAVE?
WHAT IS THE MODULE NUMBER OF THE FIRST ANALOG INPUT? [D] :
HOW MANY IAD-IA ANALOG INPUT MODULES DO YOU HAVE? [D]:
WHAT IS THE MODULE NUMBER OF THE FIRST ANALOG INPUT? [D] :
HOW MANY ANALOG OUTPUT MODULES DO YOU HAVE? [D] :
WHAT IS THE MODULE NUMBER OF THE FIRST ANALOG OUTPUT? [D]:
HOW MANY DIGITAL INTERRUPT MODULES DO YOU HAVE? [D]:
WHAT IS THE MODULE NUMBER OF THE FIRST DIGITAL INTERRUPT? [D]:
HOW MAY DIGITAL SENSE MODULES DO YOU HAVE? [D]:
WHAT IS THE MODULE NUMBER OF THE FIRST DIGITAL SENSE INPUT? [D]:
HOW MANY BISTABLE OUTPUT MODULES DO YOU HAVE? [D]:
WHAT IS THE MODULE NUMBER OF THE FIRST BISTABLE OUTPUT? [D] :
HOW MANY SINGLE-SHOT OUTPUT MODULES DO YOU HAVE? [D] :
WHAT IS THE MODULE NUMBER OF THE FIRST SINGLE-SHOT OUTPUT? [D] :
HOW MANY I/0 COUNTER MODULES DO YOU HAVE? [D] :
WHAT IS THE MODULE NUMBER OF THE FIRST I/O COUNTER? [D] :
HOW MANY DAll-B PARALLEL LINE INTERFACES DO YOU HAVE? [D] :
THE DAll-B UNIBUS LINK IS A HIGH-SPEED, HALF-DUPLEX CHANNEL FOR
CONNECTING TWO PDP-11 COMPUTERS OVER A DISTANCE OF UP TO 100
FEET (30 METERS).
USING THE DIRECT-MEMORY-ACCESS FACILITIES OF
EACH COMPUTER, IT IS CAPABLE OF TRANSFERRING BLOCKS OF DATA AT
HIGH DATA RATES WITH MINIMAL PROCESSOR OVERHEAD.
Additional explanation:
If the answer to this query is non-zero,
then the
next two queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the subsidiary queries for
this query will not be
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DAll-B? [0]:124
124 IS THE STANDARD VECTOR ADDRESS FOR THE DAll-B.
Additional explanation:
If the reply to the query for the number of DAll-Bs
was greater
than 1, then this query is repeated by
SGN. On the second and subsquent displays of this
query, however,
the user is required to supply an
interrupt vector address.

3-52

SYSGEN DIALOGUE DETAILS
WHAT IS ITS CSR ADDRESS? [OJ :172414
172414 IS THE STANDARD CSR ADDRESS FOR THE DA-llB.
Additional explanation:
This query is repeated for each of the DAll-Bs after
the first;
the user supplies the CSR address. Note
that queries for interrupt vector and CSR occur one
after the other for each DAll-B. The repeat sequence
repeats both queries.
HOW MANY DLll-E ASYNCHRONOUS LINE INTERFACES DO YOU HAVE? [DJ :
THE DLll SERIES OF ASYNCHRONOUS SINGLE LINE INTERFACES HANDLE
FULL OR HALF DUPLEX COMMUNICATION BETWEEN A WIDE VARIETY OF
SERIAL COMMUNICATION CHANNELS AND A PDP-11 COMPUTER.
Additional explanation:
If the answer to this query is non-zero, then the
next four queries appear.
Otherwise they are not
displayed.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DLll-E? [OJ:
THE VECTOR ADDRESS FOR THE DLll-E IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DLll-Es
was greater than 1, this query is repeated by SGN for
each DLll-E.
WHAT IS ITS CSR ADDRESS? [OJ:
THE CSR ADDRESS FOR THE DLll-E IS ASSIGNED FROM THE ADDRESS
SPACE STARTING AT 175610. CONSULT FIELD SERVICE TO DETERMINE
THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DLll-Es
was greater than 1, this query is repeated for each
DLll-E.
IS IT A HALF DUPLEX LINE? [Y/N]:
Additional explanation:
If the reply to the query for the number of DLll-Es
was greater than 1, this query is repeated for each
DLll-E.

3-53

SYSGEN DIALOGUE DETAILS
HOW MANY SYNC CHARACTERS ARE REQUIRED IN A SYNC LEADER? [D] :
RESPONSE LIMITS: 3.-14.
Additional explanation:
If the reply to the query for the number of DLll-Es
was greater
than 1, this query is repeated for each
DLll-E.
Note that the queries for interrupt vector,
CSR, half duplex, and sync leader occur one after the
other for each DLll-E. The repeat sequence repeats
all four queries.
HOW MANY DPll SYNCHRONOUS LINE INTERFACES DO YOU HAVE? [DJ:
THE DPll PROVIDES A DOUBLE-BUFFERED PROGRAM INTERRUPT INTERFACE
BETWEEN A PDP-11 AND A SERIAL SYNCHRONOUS LINE.
THE DPll INTERFACE HANDLES A WIDE VARIETY OF TERMINAL AND LINE
DISCIPLINES
(I.E.,
LINE CONTROL PROCEDURES AND ERROR CONTROL
TECHNIQUES). A PROGRAMMER CAN VARY SYNC CHARACTER, CHARACTER
SIZE, AND CAN ACCESS MODEM CONTROL LEADS. AUTOMATIC SYNC
CHARACTER STRIPPING AND AUTOMATIC IDLING ARE ALSO PROGRAM
SELECTABLE.
WHILE IDLING, THE DPll TRANSMITS THE CONTENTS OF
THE SYNC BUFFER.
Additional explanation:
If the answer to this query is non-zero,
then the
next four queries appear.
Otherwise they are not
displayed.
If the reply is terminated with ALTMODE,
the
subsidiary queries for
this query are not
pre-answered.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DPll? [OJ:
THE VECTOR ADDRESS FOR THE DPll IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DPlls was
greater
than 1,
this query is repeated by SGN for
each DPll.
WHAT IS ITS CSR ADDRESS? [0]:174770
174770 IS THE STANDARD CSR ADDRESS FOR THE DPll.
Additional explanation:
This query is repeated for each of the DPlls
the first;
the user supplies the CSR address.

after

IS IT A HALF DUPLEX LINE? [Y/N] :
Additional explanation:
If the reply to the query for the number of DPlls was
greater
th0
1,
this query is repeated by SGN for
each DPlJ..

3-54

SYSGEN DIALOGUE DETAILS
HOW MANY SYNC CHARACTERS ARE REQUIRED IN A SYNC LEADER?

[DJ :

RESPONSE LIMITS: 3.-14.
Additional explanation:
If the reply to the guery for the number of DPlls was
greater than 1, this query is repeated for each DPll.
Note that the queries for interrupt vector, CSR, half
duplex, and sync leader occur one after the other for
each DPll.
The repeat sequence repeats all four
queries.
HOW MANY DQll SYNCHRONOUS LINE INTERFACES DO YOU HAVE?

[DJ :

THE DQll IS A HIGH-SPEED, DOUBLE BUFFERED COMMUNICATIONS DEVICE
DESIGNED
TO INTERFACE THE PDP-11 PROCESSOR TO A SERIAL
SYNCHRONOUS COMMUNICATION LINE.
USING THE DIRECT-MEMORY-ACCESS
FACILITIES OF THE PDP-11, IT IS CAPABLE OF TRANSFERRING BLOCKS
OF DATA AT HIGH DATA RATES WITH MINIMAL PROCESSOR OVERHEAD.
Additional explanation:
If the answer to this query
next six queries appear.
displayed.

is non-zero,
then
the
Otherwise
they are not

WHAT IS THE VECTOR ADDRESS OF THE NEXT DQll? [OJ:
THE VECTOR ADDRESS FOR THE DQll IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DQlls was
greater
than 1,
this query is repeated by SGN for
each DQll.
WHAT IS ITS CSR ADDRESS? [OJ:
THE CSR ADDRESS FOR THE DQll rs ASSIGNED FROM THE FLOATING
ADDRESS SPACE STARTING AT 160010. CONSULT FIELD SERVICE TO
DETERMINE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DQlls was
greater
than 1,
this query is repeated by SGN for
each DQll.
IS IT A HALF DUPLEX LINE? [Y/NJ:
Additional explanation:
If the reply to the query for the number of DQlls was
greater
than 1,
this query is repeated by SGN for
each DQll.

3-55

SYSGEN DIALOGUE DETAILS
DOES IT HAVE THE ERROR DETECTION EXPANDER UNIT? [Y/N]:
DOES IT HAVE THE CHARACTER DETECTION EXPANDER UNIT? [Y/N]:
HOW MANY SYNC CHARACTERS ARE REQUIRED IN A SYNC LEADER? [Y/N] :
RESPONSE LIMITS: 3.-14.
Additional explanation:
If the reply to the query for the number of DQlls was
greater than 1, this query is repeated for each DQll.
Note that the queries for interrupt vector, CSR, half
duplex, error detection, character detection, and
sync leader occur one after the other for each DQll.
The repeat sequence repeats all six queries.
HOW MANY DUll ASYNCHRONOUS LINE INTERFACES DO YOU HAVE? [D]:
THE DUll PROVIDES A DOUBLE-BUFFERED PROGRAM INTERFACE BETWEEN A
PDP-11 AND A SERIAL SYNCHRONOUS LINE.
THE DUll PROVIDES A FLEXIBLE INTERFACE IN THAT THE PROGRAMMER
CAN SPECIFY THE SYNC AND CHARACTER SIZE.
MODEM CONTROL
CAPABILITY ALLOWS THE USE OF THE DUll IN SWITCHED OR DEDICATED
APPLICATIONS.
THE INTERFACE PROVIDES STRIP SYNC AND IDLE SYNC
FEATURES TO SIMPLIFY PROGRAMMING AND REMOVE PROCESSOR OVERHEAD.
Additional explanation:
If the answer to this query is non-zero,
then the
next four queries appear;
otherwise they are not
displayed.
WHAT IS THE VECTOR ADDRESS OF THE NEXT DUll?

[ 0] :

THE VECTOR ADDRESS FOR THE DUll IS ASSIGNED FROM THE FLOATING
VECTOR SPACE STARTING AT 300.
CONSULT FIELD SERVICE TO
DETERMINE THE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DUlls was
greater
than 1,
this query is repeated by SGN for
each DUll.
WHAT IS ITS CSR ADDRESS?

[O]:

THE CSR ADDRESS FOR THE DUll IS ASSIGNED FROM THE FLOATING
ADDRESS SPACE STARTING AT 160010. CONSULT FIELD SERVICE TO
DETERMINE LOCATION IF UNKNOWN.
Additional explanation:
If the reply to the query for the number of DUlls was
greater than 1, this guery is repeated for each DUll.

3-56

SYSGEN DIALOGUE DETAILS
IS IT A HALF DUPLEX LINE? [Y/N]:
Additional explanation:
If the reply to the query for the number of DUlls was
greater
than 1,
this query is repeated by SGN for
each DUll.
HOW MANY SYNC CHARACTERS ARE REQUIRED IN A SYNC LEADER? [D]:
RESPONSE LIMITS: 3.-14.
Additional explanation:
If the reply to the query for the number of DUlls was
greater than 1, this query is repeated for each DUll.
Note that the queries for interrupt vector, CSR, half
duplexf and sync leader occur one after the other for
each DUll. The repeat sequence repeats all four
queries.

3.4

SYSGEN2.CMD FILE DETAILS

The user-tailored Executive is built under
the direction of the
indirect file SYSGEN2.CMD.
At the completion of the assembly and
configuration phases, directions were given to bootstrap from the
proper object disk, set the time, and initiate the indirect file
SYSGEN2.CMD. The narrative which follows is an examination of the
SYSGEN2.CMD file.
The conventions used in describing the SYSGEN.CMD
file apply.
SYSGEN PART 2

COPYRIGHT (C) 1975, DIGITAL EQUIPMENT CORP., MAYNARD MASS. 01754
BUILD THE EXEC AND ALL REQUIRED SYSTEM TASKS
THIS COMMAND FILE WILL PRODUCE THE TARGET SYSTEM ON THE DISK
PACK CURRENTLY MOUNTED AS SY:.
IT WILL REPLACE THE BASELINE
SYSTEM THAT YOU ARE CURRENTLY RUNNING. YOU MUST PRESERVE
THIS DISK BEFORE PROCEEDING ANY FURTHER IF YOU HAVEN'T
DONE SO ALREADY.
;

*DO YOU WANT TO CONTINUE WITH SYSGEN PART 2? [Y/N]:
This question allows the user
to exit so that he can
preserve the disk before modifying it in building his target
system.

3-57

SYSGEN DIALOGUE DETAILS
LONG DIALOGUE PRINTS A DESCRIPTION OF EVERY STOP IN THIS
SYSGEN 2 COMMAND FILE. ON THE OTHER HAND, SHORT DIALOGUE
PROVIDES VIRTUALLY NO EXPLANTORY TEXT.
;

* DO YOU WANT LONG DIALOGUE?

(Y/N] :

As
in SYSGEN Phase 1, long
dialogue
descriptive
text which
the experienced
dispense with.

displays
much
user may wish to

LONG DIALOGUE PRINTS A DESCRIPTION OF EVERY STEP IN THIS
SYSGEN 2 COMMAND FILE.
ON THE OTHER HAND, SHORT DIALOGUE
PROVIDES VIRTUALLY NO EXPLANATORY TEXT.

* ARE YOU BUILDING A MAPPED SYSTEM?

(Y/N] :

The answer supplied to this query determines
the defaults
which will be used
for accessing and storing files.
See
Appendix B for UIC conventions.

* ARE YOU RUNNING ON A MACHINE WITH A LINE PRINTER? (Y/N] :
ASN TTO:=LI:
ASN LP:=LI:
Later, the user is asked if Task Builder maps are desired.
These maps are displayed on the line printer if this query
is answered affirmatively.
Otherwise they will be displayed
on the operator's console.
The correct Assign (ASN) command
is issued to assign the logical device LI:
to the selected
listing device.

* ARE YOU RUNNING ON A MACHINE WITH AT LEAST 24K WORDS?

[Y/N]:

This question is used to determine which Task Builder
to
use.
For
16K systems,
the SK Task Builder is used.
For
larger systems, the 14K Task Builder is used, which results
in a significant speed improvement in the building of all
tasks.
In addition,
it is not possible
to build
the
non-privileged
tasks with
the
BK Task Builder.
Thus, if
this question is
answered
negatively,
no
questions
concerning the non-privileged tasks will be asked.

* ARE YOU RUNNING AN RP02, RP03, OR RP04 AS THE SYSTEM DISK? (Y/N]:
For RK system disks, space is at a premium, and many files
are deleted
immediately after use.
RP disks, on the other
hand, have from 8 to 36 times as much space available.
This
question controls the deletion of task images and object
module libraries during the SYSGEN phase 2 process.

* DO YOU WANT SUPPORT FOR DIRECT ACCESS TO ICS/ICR-11 MODULES? [Y/N]:
* DO YOU WANT SUPPORT FOR DIRECT ACCESS TO UDCll MODULES? [Y/N]:
These two questions are used to determine
if
the modules
UDCOM.UDC,
ICTAB.ICR,
and
ICOM.ICR must be copied off the
backup media.
If either of
these questions
is answered
affirmatively, the next question is displayed.

3-58

SYSGEN DIALOGUE DETAILS

* DO YOU WANT TO BUILD THE UDC OR ICS/ICR DEVICE COMMON BLOCK?

[Y/N]:

This question is used to determine whether the common
should be built.

block

SET /UIC=[l,50]
SET /UIC=[l,54]
; INSTALL TASKS NEEDED TO DO THE WORK
INS
INS
INS
INS
INS
INS

PIP
PIP/PAR=PAR14K
LBR
LBR/PAR=PAR14K
EDI
FLX

INSTALL
INSTALL
INSTALL
INSTALL
INSTALL
INSTALL

PIP
PIP
LIBRARIAN
LIBRARIAN
EDITOR
FI LEX

System tasks are stored under uiL
Li,~UJ
for
an unmapped
system and
[1,54] for a mapped system.
The correct UIC is
established as the default UIC, and the specified tasks are
installed because
they will be needed by SYSGEN2.CMD as it
proceeds.
Each of these tasks will be
installed here and
later removed only if they are not already installed in the
host system.

* HAVE YOU ALREADY BUILT THE EXEC?

[Y/N]:

This question asks whether or not
the Executive has been
linked
together by the Task builder.
SYSGEN phase 1 does
not link the Executive, it only assembles
it.
Therefore,
the first
time you encounter this question, it should be
answered negatively.
Typically, before a partition can be laid out,
the
user
needs a Task Builder map of the newly built user-tailored
Executive.
This occurs later in SYSGEN2.CMD and may require
a lengthy off-line analysis of partition layout for the
target system.
After analyzing
the partition
layout,
SYSGEN2.CMD can be re-initiated.
If,
at this point, an
affirmative answer is typed, the process of building
the
user-tailored Executive is skipped, and SYSGEN2.CMD advances
to
the dialogue for
the partitioning phase of system
generation.
AS DISTRIBUTED ON THIS OBJECT DISK, THE SYSTEM OBJECT
LIBRARY {SYSLIB) DOES NOT CONTAIN THE FORTRAN RUN TIME
SYSTEM MODULES, LPS-11 MODULES OR THE PURDUE ISA MODULES
FOR THE UDCll OR ICR/ICS-11 SUBSYSTEMS.
THEY MUST BE
ADDED IF YOU PLAN TO USE THESE ROUTINES IN YOUR TARGET SYSTEM.
SET /UIC=[l,l]
* DO YOU WANT TO ADD THE FORTRAN RUNTIME MODULES TO SYSLIB? [Y/N]:
LBR SYSLIB/RP=FOROTS
As delivered, the system object library
{SYSLIB)
does not
contain
the FORTRAN run-time modules.
They exist on the
disk as a set of concatenated object modules ready to add to
SYSLIB.

3-59

SYSGEN DIALOGUE DETAILS
* DOES YOUR TARGET SYSTEM HAVE THE FLOATING POINT UNIT? [Y/N] :
LBR SYSLIB/RP=FORFPU
* DOES YOUR TARGET SYSTEM HAVE THE FLOATING INSTRUCTION SET? [Y/N]:
LBR SYSLIB/RP=FORFIS
* DOES YOUR TARGET SYSTEM HAVE THE EXTENDED INSTRUCTION SET? [Y/N] :
LBR SYSLIB/RP=FOREIS
* DOES YOUR TARGET SYSTEM HAVE THE EXTENDED ARITHMETIC ELEMENT? [Y/N]:
LBR SYSLIB/RP=FOREAE
LBR SYSLIB/RP=FORNHD
The FORTRAN run-time system supports five different types of
arithmetic routines.
The one used depends on the hardware
present in the target machine.
The floating point unit
provides the fastest subroutines, the FIS is next, then EIS,
EAE, and slowest is the non-hardware assisted subroutines.
The use of one of the subroutine options precludes the use
of the other options.
>PIP FOROTS.OBJ;*/DE,FORFPU;*,FORFIS;*,FOREIS;*,FOREAE;*,FORNHD;*
Since space on the RK object disk is limited, once the
FORTRAN library is added to SYSLIB, all of the concatenated
object modules are deleted. This occurs even if no FORTRAN
modules are added to SYSLIB. They are not deleted on RP
host systems.
* DO YOU WANT TO ADD THE PURDUE ISA ROUTINES TO SYSLIB? [Y/N]:
LBR SYSLIB/RP=PURISA
* DOES YOUR TARGET SYSTEM HAVE THE UDCll SUBSYSTEM? [Y/N]:
LBR SYSLIB/RP=PURUDC
* DOES YOUR TARGET SYSTEM HAVE THE ICR/ICS-11 SUBSYSTEM? [Y/N]:
LBR SYSLIB/RP=PURICR
PIP PURISA.OBJ;*/DE,PURUDC;*,PURICR;*
*DOES YOUR TARGET SYSTEM HAVE THE LPS-11 SUBSYSTEM? [Y/N]:
LBR SYSLIB/RP=LPS
PIP LPS.OBJ;*/DE
By answering the above questions for the target system, the
appropriate modules are added to SYSLIB. The files are then
deleted to make space for the new Executive and privileged
tasks. They are not deleted on RP host systems.
SET /UIC=[l,20]
SET /UIC=[l,24]
NOW WE COPY THE EXEC BUILD FILE, THE ASSEMBLY PREFIX FILE, THE
UDC OR ICS/ICR FILES AND THE CONCATENATED EXECUTIVE OBJECT
MODULES FROM THE SECONDARY STORAGE MEDIUM TO THE SYSTEM DISK.
IF DURING SYSGEN PART 1 YOU ANSWERED THAT YOUR SECONDARY
STORAGE MEDIUM WAS AN RK DISK CARTRIDGE, THESE FILES
WERE COPIED TO THE SYSTEM DISK AT THAT TIME.
;

* ARE THE FILES ALREADY ON THIS DISK? [Y/N]:
Three files, RSXBLD.CMD, RSXllM.OBJ, and RSXMC.MAC must be
on the system disk to complete SYSGEN phase 2. In addition,
the files for the UDC or ICR/ICS-11 common blocks must also
be on the system disk if user tasks are going to access the
device registers. For RK-RK and RP host systems, these
files are already on the disk and need not be copied. For
other configurations, they must be transferred to the system

3-60

SYSGEN DIALOGUE DETAILS
disk. RSXBLD.CMD is the command file that directs the Task
Builder in constructing the user-tailored Executive. It was
built by SGN. RSXllM.OBJ is the concatenated object module
file produced by PIP from the user-tailored Executive object
modules during the assembly run following the completion of
the dialogue program. RSXMC.MAC is the assembly prefix file
produced by SGN.
It is used to control the assembly of
options selected during the running of SGN. They were saved
on the backup medium and now must be copied onto the object
disk where they will be used to build the user-tailored
Executive. For RP system disks, this question will not be
issued;
instead, control will pass immediately to the point
where the three files are purged and the Executive library
is created.
* IS YOUR SECONDARY STORAGE DEVICE A CASSETTE TAPE? [Y/N] :
If the host system is an RK-cassette system,
the following
commands will copy the required files from cassette to the
system disk.

MOUNT THE CASSETTE FROM SYSGEN PART 1 IN DRIVE 0
;

*IS THE CASSETTE IN DRIVE 0 READY? [Y/N]:
LOA CT:
FLX SY:=CTO:RSXBLD.CMD,RSXllM.OBJ
FLX SY:/FB=CTO:ICTAB.ICR,ICOM.ICR
FLX SY:/FB=CTO:UDCOM.UDC
FLX SY: [200,200]=CTO:RSXMC.MAC
End of RK-cassette commands.
where the files are purged.

Control passes

the

point

* IS YOUR SECONDARY STORAGE DEVICE A DECTAPE? [Y/N] :
If the host system is an RK-DECtape system, the following
commands will copy the specified files from DECtape to the
system disk.
MOUNT

mC.Jr.'4

.1.Il.C.

nr.'l,..,ml\nn

U.C.\..J.h.t'.C:.

r."I T"\r'\••

r .rtv1•1

SYSGEN PART l IN DRIVE 0

* IS THE DECTAPE IN DRIVE 0 READY? [Y/N] :
LOA DT:
FLX SY:=DTO:RSXBLD.CMD,RSXllM.OBJ
FLX SY:/FB=DTO:ICTAB.ICR,ICOM.ICR
FLX SY:/FB=DTO:UDCOM.UDC
FLX SY: [200,200]=DTO:RSXMC.MAC
End of RK-DECtape commands.
where the files are purged.

Control

passes

the

point

*IS YOUR SECONDARY STORAGE DEVICE A MAGTAPE? [Y/N]:
If the host system is an RK-magtape system, the following
commands will copy the specified files from magtape to the
system disk.

3-61

SYSGEN DIALOGUE DETAILS
MOUNT THE MAGTAPE FROM SYSGEN PART 1 IN DRIVE 0
;

* IS THE MAGTAPE IN DRIVE 0 READY? [Y/N] :
* IS THE MAGTAPE DRIVE A TU16? [Y/N]:
If the magtape is a TU16, the following commands are issued:
LOA MM:
FLX SY:=MMO:RSXBLD.CMD,RSXllM.OBJ
FLX SY:/FB=MMO:ICTAB.ICR,ICOM.ICR
FLX SY:/FB=MMO:UDCOM.UDC
FLX SY: [200,200]=MMO:RSXMC.MAC
If the magtape is a TUlO or ·Ts03, the following commands are
issued:
LOA MT:
FLX SY:=MTO:RSXBLD.CMD,RSXllM.OBJ
FLX SY:/FB=MTO:ICTAB.ICR,ICOM.ICR
FLX SY:/FB=MTO:UDCOM.UDC
FLX SY: [200,200]=MTO:RSXMC.MAC
End of RK magtape commands.
where the files are purged.

Control

passes

the

point

* IS YOUR SECONDARY STORAGE DEVICE AN RXOl FLEXIBLE DISK? [Y/N]:
If the host system is an RK floppy disk system,
the
following commands will copy the specified files from floppy
disk to the system disk.
LOA DX:
;

; MOUNT THE FLEXIBLE DISK FROM SYSGEN PART l IN DRIVE 0
;

* rs THE FLEXIBLE DISK IN DRIVE 0 READY?

[Y/N]:

MOU DXO:
PIP SY:/NV=DXO:RSXllM.OBJ,RSXBLD.CMD
PIP SY:/NV=DXO:ICTAB.ICR,ICOM.ICR
PIP SY:/NV=DXO:UDCOM.UDC
PIP SY: [200,200]/NV=DXO:RSXMC.MAC
DMO DXO:
End of RK floppy disk commands.
where the files are purged.

Control passes to the point

If none of the supported configurations is specified,
the
following error message is issued and the questions are
repeated to determine the secondary storage medium used
during SYSGEN phase 1.

***

ERROR

***

YOU HAVE NOT CHOSEN ONE OF THE STANDARD SYSGEN SECONDARY
STORAGE DEVICES. WE WILL REPEAT THE QUESTIONS.

3-62

SYSGEN DIALOGUE DETAILS
Control passes to this location after the files RSXBLD.CMD,
RSXllM.OBJ, RSXMC.MAC and, optionally, ICTAB.ICR, ICOM.ICR,
and UDCOM.UDC, are copied to the system disk.
Or, in the
case of RP systems or 2-RK disk systems, control passes to
this location immediately on determining the system type.
REM •.. FLX
REM LOA
The following commands are issued
which are no longer needed.
PIP
PIP
PIP
PIP

delete

the

drivers

[l,54]CTDRV.*;*/DE,DTDRV.*;*,DXDRV.*;*
[l,54]MMDRV.*;*/DE,MTDRV.*;*,LOA.*;*
[l,SO]CTDRV.*;*/DE,DTDRV.*;*,DXDRV.*;*
[l,SO]MMDRV.*;*/DE,MTDRV.*;*,LOA.*;*

DELETE ALL BUT LATEST VERSION OF EXEC BUILD COMMAND FILE,
ASSEMBLY PREFIX FILE, AND ALL EXEC OBJECT LIBRARIES IF ANY.
PIP RSXBLD.CMD/PU
PIP RSXllM.OLB;*/DE
PIP [200,200]RSXMC.MAC/PU
If a previous SYSGEN was not done on this disk, the file
RSXllM.OLB will not exist. PIP will issue a warning message
to this effect.
NOW WE BUILD THE EXEC LIBRARY
Following the assembly of the user-tailored Executive, the
individual object modules were concatenated into a single
file named RSXllM.OBJ. Concatenation was performed to save
space
on
the
secondary medium.
This file must be
transformed into a library file for processing by the Task
Builder (TKB).
LBR RSXllM/CR:80.=RSXllM.OBJ ·
PIP RSXllM.OBJ;*/DE
!DELETE OBJECT MODULE FILE
The preceding two commands create a library file named
RSXllM.OLB from the concatenated object file and delete the
no-longer-needed concatenated object file, thus freeing disk
space.
* DO YOU WISH TO ADD USER WRITTEN DRIVER(S) TO THE EXEC? [Y/N]:
As part of the SYSGEN process, the user may include his own
specialized drivers in the Executive. If the user answers
affirmatively to this question, the comwand file will lead
him along in the steps required to assemble the driver and
device table modules, insert them into the Executive object
library, and modify the Executive build command file to
include the driver when the Executive is built.
See the
RSX-llM Guide to Writing an I/O Driver for additional
information on writing drivers and including them in the
system.

3-63

SYSGEN DIALOGUE DETAILS
SET /UIC=[200,200]
WE WILL PAUSE WHILE YOU ASSEMBLE YOUR DRIVER(S} AND USRTB
MODULE. THE EXEC ASSEMBLY PREFIX FILE RSXMC.MAC IS ALREADY
LOCATED UNDER UIC (200,200]. ASSUMING YOUR DRIVER(S) NAME IS
XXDRV, WHERE XX IS THE DEVICE NAME (E.G., DK}, THE FOLLOWING
COMMANDS WILL ASSEMBLE THE DRIVER(S} AND THE USRTB MODULE.
>RUN $MAC
>RUN $BIGMAC
MAC>XXDRV=[l,l]EXEMC/ML, [200,200]RSXMC,XXDRV
MAC>USRTB=[l,l]EXEMC/ML,[200,200]RSXMC,USRTB
MAC>AZ
At this point, the indirect file will pause while
assembles his drivers and user device tables.

the

user

NOW YOU MUST ADD YOUR DRIVER(S} AND USRTB MODULE
TO THE EXEC OBJECT MODULE LIBRARY. THE FOLLOWING IS AN EXAMPLE:
LBR>RSXllM/RP=[200,200]XXDRV,USRTB
LBR>AZ
SET /UIC=[l,20]
SET /UIC=[l,24]
LBR
When the Librarian prompts for input, the user should
in a command to include his drivers and device table.

type

YOU MUST NOW ADD COMMANDS TO INCLUDE YOUR DRIVER(S} AND USRTB
MODULE INTO THE EXEC BY EDITING THE EXEC TASK BUILD COMMAND FILE.
TO ADD DRIVER(S) INSERT COMMANDS OF THE FORM:
RSXllM/LB:XXDRV
INTO THE COMMAND FILE IN THE PLACE WHERE THE
OTHER DRIVERS ARE REFERENCED. XXDRV REPRESENTS THE NAME OF
YOUR DRIVER(S}. IN ADDITION, LOCATE THE LINE IN WHICH THE
MODULE SYSTB IS REFERENCED AND ADD A REFERENCE TO YOUR
USRTB MODULE IMMEDIATELY AFTER IT. E.G.,
[l,20]RSX11M/LB:LOADR:NULTK:SYSTB:USRTB:SYTAB:INITL
[l,24]RSX11M/LB:LOADR:NULTK:SYSTB:USRTB:SYTAB:INITL
THEN LOCATE THE LINE:
GBLDEF=$USRTB:O
AND DELETE IT.
EDI RSXBLD.CMD
When the Line Text Editor (EDI}
perform the required edits.

3-64

prompts,

the

user

should

SYSGEN DIALOGUE DETAILS
YOUR DRIVERS WILL AUTOMATICALLY BE LINKED WITH THE EXECUTIVE
YOU ARE BUILDING.
Whether or not the user is adding a driver to the Executive,
processing continues at this point.
SET /UIC=[l,50]
SET /UIC=[l,54]
INS TKB
INS BIGTKB
SET /UIC=[l,20]
SET /UIC=[l,24]
;

; NOW WE BUILD THE EXECUTIVE
;

TKB @RSXBLD
The correct version of the Task Builder is installed and
used
to guide the Executive.
The Task Builder (TKB) is
controlled by the command file RSXBLD.CMD, which is created
during the running of SGN (and possibly modified to include
other drivers).
The following text is only displayed when the UDCll or
ICS/ICR-11 is being included in the user's target system and
the device common block is to be built.
NOW WE BUILD THE UDC OR ICS/ICR DEVICE
COMMON BLOCK. THE DEVICE COMMON BLOCK RESIDES
ON THE I/O PAGE AND IS REFERENCED WHENEVER A TASK
DIRECTLY ACCESSES ANY I/O MODULES.
;

SET /UIC=[l,24]
SET /UIC=[l,20]
TKB @UDCBLD
TKB @ICBLD
SET /UIC=[l,l]
TOO

uun

CVCTT0/nn-r1
~n1Tnnu
Tnn
~i~uio1nr-tir'VJi~v~.i~n

LBR SYSLIB.RP=[l,20]IDCOM.UDC
LBR SYSLIB/RP=[l,24]ICTAB.ICR
LBR SYSLIB/RP=[l,20]ICTAB.ICR
PIP [1,24]*.ICR;*/DE,ICBLD.CMD;*
PIP [1,20]*.ICR;*/DE,ICBLD.CMD;*
PIP [1,24]*.UDC;*/DE,UDCBLD.CMD;*
PIP [1,20]*.UDC;*/DE,UDCBLD.CMD;*
Whether or not the UDC or ICS/ICR is being included
target system, processing continues at this point.
REM ... LBR
YOU NOW NEED THE MAP OF THE EXECUTIVF TO LAYOUT
THE PARTITIONS IN YOUR NEW SYSTEM.
;

* DO YOU HAVE THE MAP ALREADY? [Y/N]:
PIP LI:=[l,30]RSX11M.MAP
PIP LI:=[l,34]RSX11M.MAP

3-65

the

SYSGEN DIALOGUE DETAILS
At this point, the user must find the symbol $SYTOP in the
module INITL in the map. The symbol $SYTOP provides the
value of the highest memory location occupied by the
user-tailored Executive and is used for partition layouts.
In an unmapped system, tasks are linked to the physical base
address of the partition in which they run. To create an
unmapped user-tailored system, the task build command files
for FCP, MCR and TKTN must be edited if $SYTOP is above
30100(8). If $SYTOP is above 40000(8), all other task build
command files must be edited, and the associated tasks rr.ust
be rebuilt. In a mapped system, tasks are linked to a
virtual base address rather than a physical base address
and, thus, do not depend on the physical base address of the
partition.
LOOK IN THE MAP OF THE EXEC FOR THE MODULE INITL WHICH IS
LOCATED AT THE END OF THE MAP. THE SYMBOL $SYTOP
INDICATES THE LAST ADDRESS ALLOCATED TO THE EXEC. USER
PARTITIONS MAY START AT ANY LOCATION BEYOND THIS ADDRESS
AS LONG AS THEY FALL ON A 64 BYTE BOUNDARY. SEE THE
SYSGEN MANUAL FOR A DESCRIPTION OF HOW TO SET UP PARTITIONS.
YOU MAY WISH TO STOP AT THIS TIME TO LAY OUT YOUR PARTITIONS.
;

* DO YOU WANT TO CONTINUE AND BUILD THE TASKS NOW? [Y/N]:
If the user wishes to ponder his partition layout, it is
possible at this point to interrupt the SYSGEN phase 2
process. See Chapter 4 for information on laying out
partitions.
At
a later time, the user may restart
SYSGEN2.CMD by answering the question:
* HAVE YOU ALREADY BUILT THE EXECUTIVE? [Y/N]:
If the answer is affirmative, control will pass directly to
the point where the system tasks are built. If the user
specifies an exit from SYSGEN phase 2 at this time, the
following text is displayed.
WE WILL EXIT NOW. WHEN YOU ARE READY TO RESUME
THE BUILDING OF THE SYSTEM, EXECUTE THIS INDIRECT COMMAND FILE
AND ANSWER Y TO THE FOLLOWING QUESTION:
>*HAVE YOU ALREADY BUILT THE EXEC? [Y/N]:
The SYSGEN process will shut down, allowing the user to lay
out the partitions.
After laying out partitions, he can
resume the SYSGEN process by rebootstrapping the system and
starting SYSGEN2.CMD again.
If the user elects to continue with
processing continues at this point.

3-66

the

SYSGEN

process,

SYSGEN DIALOGUE DETAILS
WE WILL NOW EDIT THE TASK BUILD COMMAND FILES
AND BUILD THE REQUIRED TASKS.
THE TASKS MCR, FllACP, AND TKTN NORMALLY RUN IN PARTITION
SYSPAR WHICH IS 10000(8) BYTES LONG.
IF YOU
WISH TO CHANGE THE PARTITION NAME OR PUT ANY OF THESE TASKS
IN OTHER PARTITIONS, YOU MUST EDIT THE CORRECT COMMAND FILE.
IN ADDITION, FOR UNMAPPED SYSTEMS, PARTITION
SYSPAR IS ASSUMED TO START AT 30100.
IF IN YOUR
TARGET SYSTEM, SYSPAR'S BASE ADDRESS HAS CHANGED,
YOU MUST EDIT THESE TASK BUILD FILES.
The privileged
tasks must always be rebuilt when
the
Executive has been rebuilt. The privileged tasks are linked
to the Executive symbol
table file and map into
the
Executive address space so that they may access various
system lists and data structures.
If the Executive is
changed, these addresses change;
thus, the tasks have to be
rebuilt.
The Line Text Editor (EDI) may be used to examine the PAR
command line in each file that may require modification.
Every file in question must be checked, since a Partitions
command which does not reflect the requirements of the
user-tailored system can produce tasks which will not run.
A typical sequence as it would appear on the console is:
>EDI BOOBLD.CMD
[PAGE
l]
*L PAR=
PAR=GEN:40000:40000
*KILL
EDI>~Z

[EXIT]
For an unmapped system, tasks built to run
the Task Builder command:

SYSPAR

have

PAR=SYSPAR:30100:7700
Those built to run in the GEN subpartition specify:
PAR=GEN:40000:40000
For mapped systems, the Partitions commands are:
PAR=SYSPAR:O:lOOOO
PAR=GEN:0:40000
If alteration is necessary, the change may oe accomplished
as shown below.
Suppose, for example, the GEN subpartition
is to be 50000(8) words in length. The above sequence would
be changed to:
>EDI BOOBLD.CMD
[PAGE l]
*L PAR
PAR=GEN:40000:40000

3-67

SYSGEN DIALOGUE DETAILS
*C/0:4/0:5
PAR=GEN:40000:50000
*EX
[EXIT]
SET /UIC=[l,20]
SET /UIC=[l,24]
*DO YOU WISH TO BUILD THE BIG (4.SK) AND FASTER FILE SYSTEM? [Y/N]:
It is possible to build different file systems: one 2K in
size, and relatively heavily overlaid, and one a minimum of
4K in size, which is less overlaid and provides faster
processing of file access requests.
As a result of the
answer to this question, one of the following two questions
is asked.

* DO YOU HAVE TO EDIT THE 2K FCP TASK BUILD COMMAND FILE?

[Y/N] :

The default partition for
the 2K FCP is SYSPAR.
For
unmapped systems, it is assumed to begin at 30100(8) and is
7700(8) words long. For mapped systems, SYSPAR is 10000(8)
words long.
BIGFCP'S BUILD FILE SELECTS PARTITION FCPPAR
WHICH STARTS AT 100000(8) AND IS
22000(8) LONG.
IN MOST CASES, THESE LIMITS WILL
HAVE BE TO ALTERED FOR YOUR SYSTEM.

* DO YOU HAVE TO EDIT THE 4.SK FCP TASK BUILD COMMAND FILE?

[Y/N]:

The default partition for the big FCP is FCPPAR.
Its size
is specified as 22000(8) words;
for unmapped systems, it is
based at 100000(8). The size of BIGFCP may be altered, as
described in section 4.4.1.
Depending on the answer to the
appropriate preceding question, one of the following
two
commands may be issued to allow editing of the correct task
build command file.
EDI FCPBLD.CMD
EDI BIGFCPBLD.CMD
* DO YOU HAVE TO EDIT THE TASK BUILD FILES FOR MCR AND TKTN? [Y/N]:
If the preceding question is answered
following two commands are issued.

affirmatively,

EDI MCRBLD.CMD
EDI TKNBLD.CMD
THE FOLLOWING TASKS PROCESS ADDITIONAL MCR COMMANDS.
BOO, OMO, AT., INI, INS, MOU AND SAV.
THEY ARE BUILT TO RUN IN PARTITION GEN WHICH IS 8K
WORDS (40000 BASE 8 BYTES) LONG.
IF YOU WISH TO CHANGE THE
PARTITION IN WHICH ANY OF THEM RUN, YOU MUST EDIT THE
APPROPRIATE BUILD COMMAND FILE.
IN ADDITION, FOR UNMAPPED SYSTEMS PARTITION GEN
IS ASSUMED TO START AT 40000.
IF IN YOUR TARGET

3-68

the

SYSGEN DIALOGUE DETAILS
SYSTEM GEN'S BASE ADDRESS HAS CHANGED, YOU MUST
EDIT THE BUILD COMMAND FILES FOR ALL OF THESE TASKS.
;

* DO YOU HAVE TO EDIT ANY FILES FOR MCR TASKS? [Y/N]:
If the previous question is answered affirmatively,
the
following questions are issued;
depending on the user's
response, the EDI command may or may not be issued.
* DO YOU WANT TO EDIT TASK BOO'S BUILD COMMAND FILE?
EDI BOOBLD.CMD
* DO YOU WANT TO EDIT TASK DMO'S BUILD COMMAND FILE?
EDI DMOBLD.CMD
*DO YOU WANT TO EDIT TASK AT.'S BUILD COMMAND FILE?
EDI INDBLD.CMD
* DO YOU WANT TO EDIT TASK !NI'S BUILD COMMAND FILE?
EDI INIBLD.CMD
* DO YOU WANT TO EDIT TASK INS'S BUILD COMMAND FILE?
EDI INSBLD.CMD
*DO YOU WANT TO EDIT TASK MOU'S BUILD COMMAND FILE?
EDI MOUBLD.CMD
* DO YOU WANT TO EDIT TASK SAV'S BUILD COMMAND FILE?
EDI SAVBLD.CMD
PIP *.CMD/PU

[Y/N]:
[Y/N]:
[Y/N]:
[Y/N]:
[Y/N]:
[Y/N]:
[Y/N]:

The following text is always displayed.
TASK PMD IS THE POST-MORTEM AND SNAP DUMP PROCESSOR.
IT IS A PRIVILEGED TASK AND FOR CORRECT OPERATION IT SHOULD
RUN IN A PARTITION WITH ONLY CHECKPOINTABLE TASKS.
;

* DO YOU WANT TO BUILD TASK PMD? [Y/N] :
If the previous question was answered affirmatively, the
following question will be displayed. The EDI command will
depend on the answer to the edit question.
NOTE
This sequence of two questions and an
EDI
command is repeated many times
The
further on in the command file.
the
same;
operation
is
always
therefore, a description will not be
given for each successive one.
* DO YOU HAVE TO EDIT PMD'S BUILD COMMAND FILE? [Y/N]:
EDI PMDBLD.CMD
The task SHF ... , referred to below, may only run in a mapped
system. Therefore, the text associated with it will only be
displayed during the building of mapped systems.
TASK SHF IS THE SYSTEM CONTROLLED PARTITION TASK
COMPACTION ROUTINE.
IT WILL AGGLOMERATE FREE SPACE
IN A SYSTEM CONTROLLED PARTITION ON REQUEST FROM
THE EXECUTIVE. YOU MUST HAVE SELECTED THIS FEATURE
DURING SYSGEN PART 1 FOR THIS TASK TO BE CALLED

3-69

SYSGEN GIALOGUE DETAILS

BY THE EXECUTIVE.

IT NORMALLY RUNS IN PARTITION SYSPAR.

;

* DO YOU WANT TO BUILD TASK SHF? [Y/N]:
* DO YOU HAVE TO EDIT TASK SHF'S BUILD COMMAND FILE?

[Y/N]:

EDI SHFBLD.CMD
The next portion of text is displayed only if the user
specifies that the host system has at least 24K of memory.
There are two reasons for this:
1. Many of the tasks require BIGTKB to be built.
2. For unmapped systems with only 16K of memory, it is
not possible to run the tasks at any base address
other than 40000 (BK).
THE UTILITY TASKS (MAC, PIP, TKB ETC.) ARE ALREADY BUILT TO RUN IN
PARTITION GEN (EXCEPT BIGMAC AND BIGTKB WHICH ARE BUILT TO RUN IN
PAR14K) . IF YOU WISH TO CHANGE THE PARTITION IN WHICH ANY OF THEM
RUN YOU MUST EDIT THE APPROPRIATE TASK'S BUILD COMMAND FILE.
The following note appears only when building mapped systems.
NOTE: FOR MAPPED SYSTEMS IT ISN'T NECESSARY
TO REBUILD ANY OF THESE TASKS SINCE YOU
MAY EXPLICITLY SPECIFY THE PARTITION IN
WHICH THEY RUN AT INSTALL TIME.
NONE OF THE TASK BUILD COMMAND FILES BUILDS CHECKPOINTABLE TASKS
EXCEPT FOR THE PRINT SPOOLER TASK PRT. IF YOU WISH TO BUILD ANY
TASK CHECKPOINTABLE YOU MUST EDIT THE TASK BUILD COMMAND FILE
TO CHANGE THE /-CP SWITCH ON THE TASK FILE SPECIFICATION TO /CP.
NOTE: CHECKPOINTABLE TASKS REQUIRE MORE DISK SPACE
THAN NON-CHECKPOINTABLE TASKS.
The following paragraph of text appears only
unmapped systems.

when

building

FOR UNMAPPED SYSTEMS PARTITION GEN
IS ASSUMMED TO START AT 40000.
IF IN YOUR TARGET SYSTEM GEN'S
BASE ADDRESS HAS CHANGED, YOU
MUST EDIT THE TASK BUILD COMMAND
FILES AND REBUILD EVERY UTILITY
TASK YOU PLAN TO RUN ON YOUR
TARGET SYSTEM.
THE PRINT SPOOLER IS NOT DISTRIBUTED AS A TASK IMAGE. IF YOU
WISH TO USE A PRINT SPOOLER IN YOUR TARGET SYSTEM, YOU MUST
RXPLICITLY BUILD IT.
If you wish to build the print spooler,
affirmatively to the next question.

3-70

you

must

answer

SYSGEN DIALOGUE DETAILS

* DO YOU WANT TO REBUILD ANY OF THE UTILITY TASKS? [Y/N]:
* DO YOU WANT TO BUILD TASK PRT (PRINT SPOOLER)? [Y/N]:
* DO YOU WANT TO EDIT TASK PRT'S BUILD COMMAND FILE? [Y/N]:
EDI PRTBLD.CMD
*DO YOU WANT TO REBUILD TASK UFD? [Y/N]:
* DO YOU HAVE TO EDIT TASK UFD'S BUILD COMMAND FILE? [Y/N]:
EDI UFDBLD.CMD
* DO YOU WANT TO REBUILD TASK MAC? [Y/N]:
* DO YOU WANT TO EDIT 'rASK MAC'S BUILD COMl>iAND FILE? [Y/N]:
EDI MACBLD.CMD
* DO YOU WANT TO REBUILD TASK BIGMAC? [Y/N]:
* DO YOU WANT TO EDIT TASK BIGMAC'S BUILD COMMAND FILE? [Y/N]:
EDI BIGMACBLD.CMD
* DO YOU WANT TO REBUILD TASK TKB? [Y/N]:
* DO YOU WANT TO EDIT TASK TKB'S BUILD COMMAND FILE? [Y/N]:
EDI TKBBLD.CMD
*DO YOU WANT TO REBUILD TASK BIGTKB? [Y/N]:
* DO YOU WANT TO EDIT TASK BIGTKB'S BUILD COMMAND FILE? (Y/N] =
EDI BIGTKBBLD.CMD
* DO YOU WANT TO REBUILD TASK PIP? [Y/N] :
* DO YOU WANT TO EDIT TASK PIP'S BUILD COMMAND FILE? [Y/N]:
EDI PIPBLD.CMD
* DO YOU WANT TO REBUILD TASK FLX? [Y/N]:
*DO YOU WANT TO EDIT TASK FLX'S BUILD COMMAND FILE? [Y/N]:
EDI FLXBLD.CMD
* DO YOU WANT TO REBUILD TASK VFY? [Y/N] :
* DO YOU WANT TO EDIT TASK VFY'S BUILD COMMAND FILE? [Y/N]:
EDI VFYBLD.CMD
* DO YOU WANT TO REBUILD TASK DMP? [Y/N] :
* DO YOU WANT TO EDIT TASK DMP'S BUILD COMMAND FILE? [Y/N]:
EDI DMPBLD.CMD
* DO YOU WANT TO REBUILD TASK EDI? [Y/N]:
* DO YOU WANT TO EDIT TASK EDI'S BUILD COMMAND FILE? [Y/N]:
EDI EDIBLD.CMD
* DO YOU WANT TO REBUILD TASK SLP? [Y/N] :
* DO YOU WANT TO EDIT TASK SLP'S BUILD COMMAND FILE? [Y/N]:
EDI SLPBLD.CMD
* DO YOU WANT TO REBUILD TASK LBR? [Y/N] :
*DO YOU WANT TO EDIT TASK LBR'S BUILD COMMAND FILE? [Y/N]:
EDI LBRBLD.CMD
*DO YOU WANT TO REBUILD TASK FOR? [Y/N]:
r
1 * DO YOU WANT TO
l .I. I l'I! J ;
EDI FORBLD.CMD
* DO YOU WANT TO REBUILD TASK CRF? [Y/N]:
* DO YOU WANT TO EDIT TASK CRF'S BUILD COMMAND FILE? [Y/N]:
EDI CRFBLD.CMD
* DO YOU WANT TO REBUILD TASK ZAP? [Y/N] :
*DO YOU WANT TO EDIT TASK ZAP'S BUILD COMMAND FILE? [Y/N]:
EDI ZAPBLD.CMD
* DO YOU WANT TO REBUILD TAS~ BAD? [Y/N] :
* DO YOU WANT TO EDIT TASK BAD'S BUILD COMMAND FILE? [Y/N]:
EDI BADBLD.CMD
* DO YOU WANT TO REBUILD TASK VMR? [Y/N] :
* DO YOU WANT TO EDIT TASK VMR'S BUILD COMMAND FILE? [Y/N]:
EDI VMRBLD.CMD
PIP *.CMD/PU
REM ... EDI
'1'.T

/'t.i

Whether or not the utility tasks are to be built, processing
will resume here.

3-71

SYSGEN DIALOGUE DETAILS
SET /UIC=[l,50]
SET /UIC=[l,54]
;

; INSTALL THE TASK BUILDER
INS TKB
INS BIGTKB
The SYSGEN procedure selects the proper version of the Task
Builder, if the Task Builder is not already installed. This
could be the case if SYSGEN2.CMD was terminated after
building the Executive.
* DO YOU WANT TO PRINT THE MAPS FOR THE PRIVILEGED TASKS? [Y/N] :
* DO YOU WANT TO PRINT THE MAPS FOR THE UTILITY TASKS? [Y/N] :
The task build command files always generate task build
maps.
In most cases, these maps are of little use to the
user and are not needed. Therefore, a negative response is
usually given to both questions.
The second question,
referring to the utility tasks, is issued only if the
utility tasks are being built.
THE TASK MAPS WILL BE PRINTED ON
THE CONSOLE DEVICE
THE LINE PRINTER
SET /UIC=[l,20]
PIP PRT.OLB;*/DE,SUBMAC;*,FULMAC;*,TKB;*,PIP;*,FLX;*,VFY;*,DMP;*
PIP EDI.OLB;*/DE,SLP;*,LBR;*,FOR;*,CRF;*,ZAP;*,VMR;*,PIPUTL;*,BAD.OBJ;*
When not building the utility tasks on an RK host system,
the object library files are immediately deleted to provide
raore space on the disk.
The next ten commands control processing for building the
print spooler.
All other tasks to be built are controlled
similarly. Therefore, an explanation of the processing will
be given only for the print spooler.
NOW WE BUILD THE REQUESTED UTILITY TASKS
SET /UIC=[l,20]
Utility tasks, object library files, and overlay description
files reside under UIC [1,20].
All tasks but the print spooler are built to run with the
Baseline System.
Therefore, if the task is going to be
rebuilt and the host is an RK system where disk space is at
a premium, the task file is deleted before the new task is
built. A command of the form:
PIP [l,5x]PRT.TSK;*/DE
will be issued to delete the task. The symbol x represents
0 or 4 for unmapped or mapped systems, respectively.

3-72

SYSGEN DIALOGUE DETAILS
TKB @[l,20]PRTBLD
TKB @[l,24]PRTBLD
If the task is to be built, one of the two previous commands
will be issued.
The command files for building tasks for
unmapped systems reside under UIC [1,20]. Those for mapped
systems reside under UIC [1,24].
SET /UIC=[l,30]
SET /UIC=[l,34]
Task Builder maps are generated under UIC
[1,30]
for
unmapped systems and [1,34] for mapped systems. Thus, the
correct default UIC is selected by one of the two commands
above.
PIP LI:=PRT.MAP
This command will be issued to
specified to be printed.

the

map

was

The task build maps are always deleted, whether or not
have been printed.

they

PIP PRT.MAP;*/DE

PIP [l,20]PRT.OLB;*
For RK host systems, whether or not the task was built,
object module library will be deleted.
SE'I· /UIC= [ 1, 20]
PIP [l,SO]MAC.TSK;*/DE
PIP [l,54]MAC.TSK;*/DE
TKB @[l,20]MACBLD
TKB @[l,24]MACBLD
SET /UIC=.[1,30]
SET /UIC=(l,34]
PIP LI: =MAC. MP~P
PIP MAC.MAP;*/DE
PIP [l,20]SUBMAC.OLB;*/DE
SET /UIC=[l,20]
PIP (l,SO]BIGMAC.TSK;*/DE
PIP [l,54]BIGMAC.TSK;*/DE
TKB @[l,20]BIGMACBLD
TKB @[l,24]BIGMACBLD
SET /UIC=(l,30]
SET /UIC=[l,34]
PIP LI:=BIGMAC.MAP
PIP BIGMAC.MAP;*/DE
PIP [l,20]FULMAC.OLB;*/DE
SET /UIC=(l,20]
PIP [l,SO]TKB.TSK/DE
PIP [l,54]TKB.TSK/DE
TKB @[l,20]TKBBLD
TKB @[l,24]TKBBLD
SET /UIC=(l,30]
SET /UIC=(l,34]
PIP LI:=TKB.MAP
PIP TKB.MAP;*/DE
SET /UIC=[l,20]

3-73

the

SYSGEN DIALOGUE DETAILS
Note:

Since the big Task Builder is used to rebuild
the utility
tasks,
it cannot be deleted at this point on RK host
systems.
It is deleted after all tasks are built, only if a
new version was built for the target system.
AlRo, the Task
Builder library is used by tht CRF task and will not be
deleted until after CRF is processed.

TKB @[l,20]BIGTKBBLD
TKB @[l,24]BIGTKBBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=BIGTKB.MAP
PIP BIGTKB.MAP;*/DE
SET /UIC=[l,50]
SET /UIC=[l,54]
PIP PIP.TSK;l/RE=PIP.TSK
s E'l' I u I c = [1 , 2 0 ]
Note:

Since PIP is used during the building of tasks, it cannot be
deleted at this time on RK host systems. When the target
system is bootstrapped and saved, all tasks residing under
UIC
[l,50]
or
(1,54]
should be purged
to destroy old
versions and to free disk space.

TKB @[l,20]PIPBLD
TKB @[l,24]PIPBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=PIP.MAP
PIP PIP.MAP;*/DE
PIP [l,20]PIP.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]FLX.TSK;*/DE
PIP [l,54]FLX.TSK;*/DE
TKB @[l,20]FLXBLD
TKB @[l,24]FLXBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=FLX.MAP
PIP FLX.MAP;*/DE
PIP [l,20]FLX.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]VFY.TSK;*/DE
PIP [l,54]VFY.TSK;*/DE
TKB @[1,20]VFYBLD
TKB @[l,24]VFYBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=VFY.MAP
PIP VFY. MJ.'.P; */DE
PIP [l,20]VFY.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]DMP.TSK;*/DE
PIP [l,54]DMP.TSK;*/DE
TKB @[l,20]DMPBLD
TKB @[l,24]DMPBLD
SET /UIC=[l,30]
SE'!' /UIC= [ 1, 34]
PIP LI:=DMP.MAP
PIP DMP.MAP;*/DE

3-74

SYSGEN DIALOGUE DETAILS
PIP [l,20]DMP.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]EDI.TSK;*/DE
PIP [l,54]EDI.TSK;*/DE
?'11 Rnnu.n
TKB ra'- Lr -1 r_ .! - - - - - -

TKB @[ 1 , 2 4 ] ED IB LD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI: =EDI. MAP
PIP EDI.MAP;*/DE
PIP [l,20]EDI.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]SLP.TSK;*/DE
PIP [l,54]SLP.TSK;*/DE
TKB @[l,20]SLPBLD
TKB @[l,24]SLPBLD
SET /UIC=[l,30]
SET /UIC;::; [l, 34]

PIP LI:=SLP.MAP
PIP SLP.MAP;*/DE
PIP [l,20]SLP.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]LBR.TSK;*/DE
PIP [l,54]LBR.TSK;*/DE
TKB @[l,20]LBRBLD
TKB @[l,24]LBRBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=LBR.MAP
PIP LBR.MAP;*/DE
PIP [l,20]LBR.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]FOR.TSK;*/DE
PIP [l,54]FOR.TSK;*/DE
TKB @[l,20]FORBLD
TKB @[l,24]FORBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
pTn
.1.r LI:=FOR.MAP
PIP FOR.MAP;*/DE
PIP [l,20]FOR.OLB;*/DE
SET /UIC=[l,20]

PIP [l,50]CRF.TSK;*/DE
PIP [l,54]CRF.TSK;*/DE
TKB @[l,20]CRFBLD
TKB @[l,24]CRFBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=CRF.MAP
PIP CRF.MAP;*/DE
PIP [l,20]CRF.OLB;*/DE,TKB;*
SET /UIC=[l,20]
PIP [l,50]ZAP.TSK;*/DE
PIP [l,54]ZAP.TSK;*/DE
TKB @[l,20]ZAPBLD
TKB @[l,24]ZAPBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=ZAP.MAP
PIP ZAP.MAP;*/DE

3-75

SYSGEN DIALOGUE DETAILS
PIP [l,20]ZAP.OLB;*/DE
SET /UIC=[l,20]
PIP [l,50]BAD.TSK;*/DE
PIP [l,54]BAD.TSK;*/DE
TKB @[l,20]BADBLD
TKB @[l,24]BADBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=BAD.MAP
PIP BAD.MAP;*/DE
PIP [l,20]BAD.OBJ;*/DE
SET /UIC=[l,20]
Note:

Since the VMR task will be needed to install tasks in the
target system before bootstrapping the system, the current
task image is not deleted at this point on RK host systems.
It should be handled similarly to PIP.

TKB @[l,20]VMRBLD
TKB @[l,24]VMRBLD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=VMR.MAP
PIP VMR.MAP;*/DE
PIP [l,20]VMR.OLB;*/DE
PIP [l,20]PIPUTL.OLB;*/DE
This completes the processing of the utility tasks.
The
file PIPUTL.OLB contains subroutines used by many of the
utility tasks. Thus, this file is deleted after all tasks
are built.
Whether or not the utility tasks are built, processing
continues at this point to build all the privileged tasks.
NOW WE BUILD THE PRIVILEGED AND MCR TASKS
The processing of privileged tasks is similar
to the
foregoing.
Therefore, only one task will be described
fully.
SET /UIC=[l,20]
SET /UIC=[l,24]
One of the two commands above will be issued to select the
proper default UIC for
the object library,
task build
command, and overlay description files.

3-76

SYSGEN DIALOGUE DETAILS
TKB @BOOBLD.CMD
PIP BOO.OLB;*/DE
The above two commands build the tasks and delete the object
library file. The object library file is deleted only on RK
disk host systems.
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=BOO.MAP
The correct map UIC is set, and, if
printed.

selected,

the

map

PIP 800.MAP;*/DE
The task map is always deleted.
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @DMOBLD.CMD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=DMO.MAP
PIP DMO.MAP;*/DE
SET /UIC=[l,50]
SET /UIC=[l,54]
PIP FCP.TSK;l/RE=FCP.TSK
The file control primitives (FCP) task is always built with
a specific version number;
therefore, to prevent the new
task file from deleting the current one,
the old FCP.TSK
file is renamed with a version number of one (1).
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @FCPBLD
TKB @BIGFCPBLD
The selected version of the file control primitives is
built.
PIP FCP.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=FCP.MAP
PIP LI:=BIGFCP.MAP
PIP FCP.MAP;*/DE
PIP BIGFCP.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @INDBLD.CMD
SET /UIC=[l,30]
SET /UIC=(l,34]
PIP LI:=IND.MAP
PIP IND.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @INIBLD. CMD
PIP INI.OLB;*/DE
SET /UIC=[l,30]

3-77

SYSGEN DIALOGUE DETAILS
SET /UIC=[l,34]
PIP LI:=INI.MAP
PIP IN!. MAP;* /DE·
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @INSBLD.CMD
PIP INS.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=INS.MAP
PIP INS.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @MCRBLD.CMD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=MCR.MAP
PIP MCR.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @MOUBLD.CMD
PIP MOU.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=MOU.MAP
PIP MOU.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @SAVBLD.CMD
PIP SAV.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=SAV.MAP
PIP SAV.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @TKNBLD.CMD
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=TKN.MAP
PIP TKN.MAP;*/DE
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @UFDBLD.CMD
PIP UFD.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=UFD.MAP
PIP UFD.MAP;*/DE

The following commands are issued only if the Post-mortem
and Snap Shot Dump tasks were selected to be built.
SET /UIC=[l,20]
SET /UIC=[l,24]
TKB @PMDBLD
PIP PMD.OLB;*/DE
SET /UIC=[l,30]
SET /UIC=[l,34]
PIP LI:=PMD.MAP
PIP PMD.MAP;*/DE

3-78

SYSGEN DIALOGUE DETAILS
The following commands are
issued only when building a
mapped system and the task compaction routine was selected
to be built.
SET /UIC=[l,24J
TKB @SHFBLD
SET /UIC=[l,34]
PIP LI:=SHF.MAP
PIP SHF.MAP;*/DE
At this point, all tasks have been built.
If
the host
system is running on an RK disk, the Task Builders are
purged, and all object module libraries, task build command
files, and overlay description files are deleted, except for
the Executive object library and task build command file.
PIP [l,SO]BIGTKB.TSK/PU,TKB.TSK/PU
PIP [l,54]BIGTKB.TSK/PU,TKB.TSK/PU
SET /UIC=[l,20]
PIP [200,200]/RE=RSXBLD.CMD
PIP [200,200]/RE=RSXllM.OLB
PIP *.*;*/DE
PIP /RE=[200,200]RSXBLD.CMD
PIP /RE=[200,200]RSX11M.OLB
SET /UIC=[l,24]
PIP [200,200]/RE=RSXBLD.CMD
PIP [200,200]/RE=RSXllM.OLB
PIP *.*;*/DE
PIP /RE=[200,200]RSXBLD.CMD
PIP /RE=[200,200]RSX11M.OLB
SET /UIC=[l,50]
SET /UIC=[l,54]
The following text and questions appear only
mapped target systems.

when

building

YOU NOW MUST SPECIFY WHETHER THE MEMORY SIZE WHICH YOU
SPECIFIED FOR YOUR TARGET SYSTEM IN SGN DURING SYSGEN
PHASE ONE IS GREATER THAN 32K WORDS
;

*IS YOUR TARGET SYSTEM LARGER THAN 32K WORDS? [Y/Nj:
If the answer to this question is affirmative, the following
text is displayed:
IF YOU EVER WISH TO HAVE TASKS FIXED ABOVE 32K WORDS IN AN
IMAGE OF YOUR TARGET SYSTEM, OR IF THE MCR TASK ..• SAV
(WHICH NORMALLY RUNS IN GEN) IS TO EVER RUN ABOVE
32K WORDS, YOU MUST CREATE A LARGER COPY OF YOUR SYSTEM
WITH PIP AT THIS TIME.
;

* DO YOU WISH TO CREATE A LARGER COPY OF YOUR SYSTEM? [Y/N]:
If the answer to this question is affirmative, the following
text is diplayed:

3-79

SYSGEN DIALOGUE DETAILS
AT THIS TIME WE WILL INVOKE PIP TO ALLOW YOU TO CREATE A
LARGER COPY OF YOUR SYSTEM. TO DO SO, ENTER THE FOLLOWING
COMMAND:
PIP>RSXllM.SYS/NV/CO/BL:N.=RSXllM.TSK
PIP>"' Z
TO CREATE A SYSTEM IMAGE IN WHICH TASK IMAGES MAY BE FIXED UP
TOM K WORDS, OR IN WHICH THE TOP OF ••• SAV MAY BE AS HIGH
AS M K WORDS, YOU MAY CALCULATE A VALUE FOR N IN THE ABOVE
COMMAND AS FOLLOWS
N=(4*M)+2
FOR EXAMPLE, TO CREATE A COPY OF THE SYSTEM WHICH PROVIDES
THE AFOREMENTIONED SUPPORT UP TO 40K WORDS, THE COMMAND WOULD
BE:
PIP>RSX11M.SYS/NV/CO/BL:l62.=RSX11M.TSK
PIP>"z
ONCE PIP IS INVOKED YOU MUST CREATE A FILE RSXllM.SYS
FOR SUBSEQUENT USE BY VMR.
NOTE: SAVED IMAGES ON 11/70 ARE CURRENTLY
SUPPORTED ONLY TO A MAXIMUM OF 124K WORDS
(I.E. , N=4 98.) •
;

PIP
The next lines are displayed only if the size of the system
is less than 32K or if the user did not want to extend the
size of the system image file.
WE WILL NOW MAKE A COPY OF YOUR TARGET SYSTEM

'
PIP
RSXllM.SYS/NV=RSXllM.TSK
No matter how large the target system is, the processing
text and commands continues at this point.

SET /UIC=[l,50]
SET /UIC=[l,54]
RSXllM.TSK IS A BACKUP COPY OF THE EXECUTIVE YOU JUST BUILT.
IF YOU DO NOT WISH TO RETAIN IT, WE WILL DELETE IT NOW.
;

* DO YOU WISH TO RETAIN RSXllM.TSK AS A SYSTEM BACKUP? [Y/N]:
The normal reason for retaining.the RSXllM.TSK file is to be
able to set up two or more systems:
one for program
development, and one for production; or
two systems with
different partition layouts.
If the answer to this question
is negative, the following command is issued:
PIP RSXllM.TSK;*/DE
In all SYSGENs, the following text is displayed to guide the
user in setting up partitions and installing tasks into the
target system Executive.

3-80

SYSGEN DIALOGUE DETAILS
YOUR TARGET SYSTEM IS NOW READY TO SET UP PARTITIONS AND INSTALL
THE REQUIRED TASKS. THE PROCEDURE TO FOLLOW ONCE VIRTUAL MCR
IS RUNNING IS:
1)
EXTENDED POOL SPACE TO BASE OF FIRST PARTITION
2) SET UP YOUR PARTITIONS
2A) SET UP DEVICE COMMON BLOCKS FOR UDC OR ICS/ICR.
3)
INSTALL TASKS (AT LEAST FCP, INS, MOU,
AND MCR MUST BE INSTALLED)
4)
EXIT FROM VIRTUAL MCR AND BOOT IN YOUR TARGET SYSTEM
SET UP THE TARGET SYSTEM BASED ON THE FOLLOWING EXAMPLE.
VMR EXAMPLE:
>RUN $VMR
RUN VIRTUAL MCR
ENTER FILENAME:RSXllM.SYS
VMR PROMPTS FOR FILE NAME
EXTEND POOL TO FIRST PARTITION BAS:
VMR>SET /POOL=400
VMR>SET /MAIN=SYSPAR:l300:100:TASK ! SET UP SYSTEM PARTITION
VMR>SET /MAIN=PAR14K:400:700:TASK ! SET UP 14K PARTITION
VMR>SET /SUB=PAR14K:GEN:400:400 ! SET UP SK PARTITION
VMR>SET /MAIN=ICOM:7710:10:DEV
VMR>SET /MAIN=UDCOM:7710:10:DEV
VMR>SET /MAIN=ICOM:l7010:10:DEV
VMR>SET /MAIN=UDCOM:l710:10:DEV
VMR>INS [l,l]ICOM
SET COMMON RESIDENT
VMR>INS [l,l]UDCOM
SET COMMON RESIDENT
VMR>INS BOO
VMR>INS DMO
VMR>INS FCP
VMR>INS BIGFCP
VMR>INS IND
VMR>INS INI
VMR>INS INS
VMR>INS MCR
VMR>INS MOU
VMR>INS SAV
VMR>INS TKN
VMR>INS UFD
VMR>f "Z

INSTALL BOOT
INSTALL DISMOUNT
INSTALL FILE SYSTEM
INSTALL FILE SYSTEM
INSTALL INDIRECT FILE PROCESSOR
INSTALL INITVOL
INSTALL INSTALL
INSTALL MCR
INSTALL MOUNT
INSTALL SAVE
INSTALL TASK TERMINATION TASK
INSTALL USER FILE DIRECTORY BUILD El
EXIT FROM VIRTUAL MCR

END OF VMR EXAMPLE
NOW WE WILL RUN VIRTUAL MCR SO THAT YOU MAY INITIALIZE
YOUR SYSTEM. BE CAREFUL TO INSTALL EITHER FCP (2K FILE SYSTEM)
OR BIGFCP (4K FILE SYSTEM) BEFORE BOOTING AND SAVING OR YOU MAY
HAVE TO START SYSGEN2 ALL OVER.
;

RUN $VMR
WHEN THE INDIRECT FILE EXITS, BOOT IN YOUR TARGET SYSTEM,
SET THE DATE AND TIME, AND SAVE THE SYSTEM WITH A BOOTSTRAP.
WHEN YOUR SYSTEM REBOOTS ITSELF, PURGE THE OLD TASK FILES.
E.G.,
>BOOT RSXllM
RSXllM V02 BL12
>TIM 10/15/75 12:00:00
>SAV /WB

3-81

SYSGEN DIALOGUE DETAILS
RSXllM V02 BL12
>RED DKO:=SY:
>MOU
>@[l,2]STARTUP
>@ <EOF>
>SET /UIC=[l,50]
>RUN $PIP
PI~ *.*/PU
PIP> AZ

28K

>
INS
INS
REM
REM

[l,SO]BOO; -1
[l,54]800; -1
•.• PIP
.•. TKB
The user should bootstrap the user-tailored Executive, using
the previous commentary as a model.

3.5

SGN ERROR MESSAGES

The System Generation Configuration Query Program (SGN)
issues a
message whenever it detects an error in the response to a query.
These messages are listed below in alphabetical order. Most messages
apply to a number of queries. Those which are specific to a single
query are so noted in the explanatory text following the message. The
"Action" paragraph is included to suggest how to correct the error and
proceed with the system generation process.
SGN -- EXPLICIT DECIMAL AND OCTAL IS INVALID
An explicit specification of both decimal and octal is an
invalid response to a numeric query. For example, the
user entered #n.;
the numeric operators"#" and " " are
mutually contradictory.
Action: Re-enter correct numeric value.
SGN -- FATAL I/0 ERROR ASSEMBLER INDIRECT FILE
SGN was unable to create the RSXASM.CMD file.
and other fatal I/0 errors, can result from:
1.

A full disk - no room remains,

No UIC under which to store the file,

A full index file,

The disk being write-protected, or

A write error.

3-82

This error,

SYSGEN DIALOGUE DETAILS
Action: Each of the above conditions has a
of occurrence. Check for:

low

probability

UIC
Disk write protect
No room (index file or disk)
Write error
Corrective action is to establish the UIC, write
enable the disk, reorganize the disk, or call the
Digital field service representative.
SGN -- FATAL I/0 ERROR CONFIGURATION PARAMETER FILE
SGN was unable to create the RSXMC.MAC file.
lm 19;.i -9 Action: See FATAL
INDIRECT FILE message above.

I/0

ERROR

ASSEMBLER

SGN -- FATAL I/O ERROR DEVICE TABLE FILE
SGN was unable to create the SYSTB.MAC file.
Action:

See FATAL I/O ERROR ASSEMBLER INDIRECT FILE message
above.

SGN -- FATAL I/O ERROR MAPPING TABLE FILE
SGN was unable to create file ICTAB.MAC.
Action:

See FATAL I/O ERROR ASSEMBLER INDIRECT FILE message
above.

SGN -- FATAL I/0 ERROR SYSGEN TEXT FILE.
SGN was unable to read the SYSGEN.TXT file.
Action:

See FATAL I/0 ERROR ASSEMBLER INDIRECT FILE message
above.

SGN -- FATAL I/0 ERROR TASK BUILDER INDIRECT FILE
SGN was unable to create the RSXBLD.CMD file.
Action:

See FATAL I/0 ERROR ASSEMBLER INDIRECT FILE message
above.

SGN -- ILLEGAL ALPHABETIC ARGUMENT
A supplied argument should have
illegal alphabetic combination.
Action:

been

numeric

ddnn

was

Re-enter correct argument value.

SGN -- ILLEGAL DEVICE SPECIFICATION
A device specification in
entered.
Action:

form

other

Re-enter correct argument value.

3-83

than

SYSGEN DIALOGUE DETAILS
SGN -- ILLEGAL NUMBER OF LOCAL LINES
The number of DHlls specified cannot support the number of
local lines specified.
Action:

Check number of DHlls and the combination of remote
and local line calculations.
Determine correct
number of local lines and re-enter correct value.

SGN -- ILLEGAL NUMBER OF REMOTE LINES
The number of DHlls cannot support the
lines specified.
Action:

number

Check number of DHlls with remote
and re-enter correct value.

remote

lines

specified

duplication

SGN -- ILLEGAL VECTOR ADDRESS
Vectors are checked for duplication;
occurs, this message is issued.
Action:

Determine correct vector address and
value.

re-enter

the

SGN -- INSUFFICIENT ARGUMENTS
This message is for
future
additions
to
Currently, no query accepts multiple responses.
Action:

syntax.

Notify Digital field service representative.

SGN -- INSUFFICIENT UNIBUS MAPPING REGISTERS
UNIBUS mapping registers are statically allocated by SGN
for PDP-11/70 systems with extended memory. All UNIBUS
NPR devices (excluding the DHll) require allocated mapping
registers and there are insufficient registers for the
specified device configuration.
Action:

Abort SGN and rerun, reducing the number of UNIBUS
NPR devices generated into the target system.

SGN -- INVALID NUMBER OF MODULES
The number of modules specified for an ICS/ICR-11 is not
possible.
An ICS/ICR-11 can have a maximum of 16
functional modules.
Action:

Check total number of
re-enter correct value.

modules

specified

and

SGN -- LOW PRIORITY CLASS MUST BE LESS THAN HIGH CLASS
The lowest priority class specified for
round-robin
scheduling is a higher priority than the previously
specified highest priority.
Action:

Re-enter value specifying a priority less
equal to the highest priority class.

3-84

than

SYSGEN DIALOGUE DETAILS
SGN

NUMERIC RESPONSE OUT OF RANGE
The required numeric
permissible range.
Action:

value

does

Check value, particularly for
and re-enter correct value.

not

lie

within

number

base

the

error;

SGN -- ODD NUMBER IS INVALID RESPONSE
Query requires an even number reply.
Action:

Determine correct value and re-enter.

SGN -- OVERLAPPING MODULE RANGE SPECIFIED
The ranges specified for the UDCll modules have
one overlapping specification.
Action:

Check module range
correct values.

specifications

and

least

re-enter

SGN -- SYNTAX ERROR
Form or format of reply is invalid.
Action:

Re-enter reply correctly and proceed.

SGN -- TOO
A maximum of three RHll controllers is permitted.
Action:

Review configuration requirements to meet
maximum allowable number of RHll controllers.

the

SGN -- YES-NO QUESTIONS MUST BE ANSWERED WITH Y,N, OR <CR>
None of the three permissible replies has been entered for
a Y/N query.
Action:
SGN -- 10000.

Enter proper reply.

MUST BE AN INTEGRAL MULTIPLE OF INTERRUPTS PER SECOND

The basic clock count rate is 10 kilohertz. The Executive
cannot deliver interrupts at a rate not evenly divisible
into 10000.
Action:
SGN -- 10000.

Enter a clock rate which is evenly
10000.

MUST BE AN

INTEG~AL

divisible

into

MULTIPLE OF SAMPLES PER SECOND

The LPSll basic clock rate is 10 kilohertz. The Executive
cannot deliver interrupts at a rate not evenly divisible
into 10000.
Action:

Enter a sample rate which is evenly divisible
10000.

3-85

into

CHAPTER 4
GENERATING LARGER SYSTEMS

4.1

GENERATING NON-MINIMUM SYSTEMS

In a 16K system, the user must observe the partition boundaries, as
defined in the Baseline System.
Specifically, SYSPAR starts at
30100(8), and GEN starts at 40000(8). There is no need for the DRVPAR
partition, since it exists as an artifice which allows generating a
large number of configurations from the same kit. Thus, there is no
room for expansion or movement of partitions. But if the system has
more than 16K, the user has the option of considering a different
partition layout. The consequences of this flexibility depend on the
size of the user-tailored Executive and whether the system is mapped
or unmapped.
Given the fact that the user has more than 16K of memory
(and
this
usually means he has memory in 8K multiples}, he should be aware of
certain Digital-supplied software options before laying out his
partitions. These options are described below:
A.

There are two versions of Task Builder and MACR0-11.
The
standard versions operate with ~K of memory. The larger
versions (named BIGTKB and BIGMAC) need a 14K partition, but
run considerably faster
than the standard versions.
In
addition, BIGMAC supports the floating point instruction set
and storage allocation psuedo op-codes. BIGTKB is required
for rebuilding the Task Builder itself. BIGMAC is built by
the command file BIGMACBLD.CMD, and BIGTKB is built by the
command file BIGTKBBLD.CMD.

FCP may be built in two versions: one that operates in 2K,
and another
that operates in at least 4K.
The 4K version
operates two to three times faster than the 2K version.
The
Baseline System comes with a 2K FCP.
BIGFCPBLD.CMD is the
Task Builder command file used to build the big FCP.

Not~ also that MCR, FCP, and
TKN share a partition in a
m1n1mum 16K system.
If more memory is available, the user
can consider dedicating a partition to FCP.
Depending on the
characteristics of his applications, substantial throughput
improvements can be realized.

The tasks AT., CRF, LBR, MAC, PIP, TKB, and VFY run such that
they determine the size of the partition in which they are
installed and make use of any additional space beyond their
dimension of 8K for
symbol
table space. Thus, they will
benefit from being installed in a partition larger
than 8K.
Similarly, BIGTKB
and
BIGMAC
will
benefit if they are

4-1

GENERATING LARGER SYSTEMS
installed in a partition larger than 14K. If, during a task
build,
assembly, or library operation, a symbol table
overflow occurs, the symbol table space may be expanded by
using a larger partition.
E.

Multi-user systems are easy to generate if there is enough
memory to support more than one system task at a time. A
multi-user system is one which has more than one terminal and
whose terminals may be used simultaneously for program
development or running other tasks. Given enough memory and
a suitable processor, RSX-llM may satisfactorily support from
2 to 16 or more users who are editing, compiling, task
building, debugging, and running production tasks.
System tuning, however, is application sensitive.
Each
installation must establish optimum tradeoffs of hardware and
software. The partitioning layouts offered in the following
sections are mainly designed to minimize the duration of the
SYSGEN process itself, while maintaining a balance with known
performance and size requirements of system software.

4.1.1

Unmapped System - Executive Less Than 8K

All system tasks for an unmapped system are built to run at specific
physical memory addresses and must be rebuilt if they are to run
elsewhere. The tasks distributed on the release disks fall into two
classes:
1.

Those built to run at 30100(8)
partition) , and

Those built to
partition).

run

(6K)

40000(8)

(which

(SK)

(which

the
is

SYSPAR
the

GEN

If, after the creation of the user-tailored Executive, $SYTOP exceeds
30100(8), but is less than 40000(8), the simplest strategy is to move
the SYSPAR partition. This means that only the task build command
files for MCR, FCP, and TKN need to be edited. Figure 4-1 depicts a
suggested layout for a 24K system where $SYTOP lies between 30100(8)
and 40000(8).

4-2

GENERATING LARGER SYSTEMS
24K

140000(8)
SYSPAR PARTITION
(2K)

22K

130000(8)
USER
ALLOCATABLE
(6K)

16K

100000(8)

PAR14K
PARTITION
(14K)
GEN
SUBPARTITION
(8K)

40000(8)
E~cmI~

(8K)

Figure 4-1
Typical Unmapped 24K System
Figure 4-1 shows that SYSPAR was moved, requiring the editing of the
task build command files for MCR, FCP, and TKN. The partition PAR14K
is now a real partition (rather than a pseudo partition, as was the
case in a 16K system) and is properly placed to run BIGTKB without
rebuilding. Furthermore, GEN has not been moved;
therefore, no
system task other than the privileged tasks running in GEN need be
rebuilt. Finally, a user-allocatable subpartition of 6K is available
in PAR14K. Of course 1 the ~ser is free to define this 6K memory block
into any configuration of subpartitions that meets his requirements.
4-2 illustrates a suggested 28K unmapped system.
It is
identical to Figure 4-1 in its allocation of the 0-24K block of
memory. The additional 4K is allocated as a main partition which can
be used by the print spooler (which actually requires only 2K) or
which can be partitioned to suit user requirements.
Figur~

If the print spooler is operating in a dedicated partition, the
printer to which output is being spooled will print at a maximum
throughtput rate. This is the reason for allocating the partitions as
shown in Figure 4-2.

4-3

GENERATING LARGER SYSTEMS
28K

160000(8)
PAR4K PARTITION
(4K)

24K

140000(8)

SYSPAR PARTITION
(2K)

22K

130000(S)

USER
ALLOCATABLE
(6K)
16K

100000(8)

PAR14K
PARTITION
(14K)
GEN
SUBPARTITION
(SK)

40000(8)
EXECUTIVE

{SK)

0 ....._____________, 0

Figure 4-2
Typical Unmapped 28K System
Both of these layouts are attempts to balance SYSGEN requirements with
known system software requirements.
Local user requirements may
dictate an alternative partitioning scheme.
The following system
requirements for unmapped systems should, however, be considered:
1.

Most Digital non-privileged
software
requires
an
BK
partition.
This software assumes that its name is GEN and
that it starts at 40000(8),

MCR, FCP, and TKN require a 2K partition. These tasks assume
that its name is SYSPAR and that it starts at 30100(8).

BIGTKB and BIGMAC require 14K. These tasks assume that the
partition is named PAR14K and that it starts at 40000(8).

Changing a starting address or default partition name requires that
the task build command files be edited and the tasks involved be
rebuilt.
Changing the partition's name does not require a rebuild, since the
Install command allows the partition name to be specified. However,
the base addresses of the new partition and that for which the task
was built must be the same, since unmapped system tasks are linked to
physical addresses.

4-4

GENERATING LARGER SYSTEMS
4.1.2

Unmapped System - Executive Greater Than SK

If the value of $SYTOP exceeds 40000(8) on an unmapped system, all
system tasks must be rebuilt. Since TKB cannot be rebuilt without
BIGTKB, a 24K system is,
in general,
the m1n1mum system for a
user-tailored Executive which exceeds SK.* The layout of partitions
becomes a user responsibility, limited only by the stated requirements
for system software and application performance goals.

4.1.3

Mapped Systems

The minimum memory size for a mapped system is 24K. The user with a
mapped system will not have to rebuild non-privileged tasks if his
partition boundaries move. This is true because non-privileged tasks
on a mapped system are built to run at a virtual base address of zero,
rather than at a physical base address. Thus, the value of $SYTOP is
much less critical in a mapped system.
Privileged tasks, as with an unmapped system, must always be rebuilt
because
they
are linked to the Executive symbol table file.
Non-privileged tasks need be rebuilt only if the user changes any of
the task's attributes. The Task Builder establishes task attributes
in building a task. The following are defined task attribu~es and
their associated Task Builder switches.
DEFINED ATTRIBUTE

DEFAULT

Task is an ancillary control
processor

-AC

Input file has concatenated
object modules

Task is checkpointable

-CP

Task contains a debugging aid

-DA

Task uses extended arithmetic element (EAE)

-EA

Task uses 11/45 floating point processor (FPP)

-FP

Task image is to contain a header

Input file is a library file

-LB

* New systems exist in multiples of SK and 16K memory stacks.
Existing systems of 20K could support a user-tailored Executive which
exceeded SK.

4-5

GENERATING LARGER SYSTEMS
MM

System has memory management unit (KTll}

MM**

Input file contains an overlay description

-MP

Task is position independent

-PI

Post mortem dump requested on SST abort

-PM

Task has privileged access rights

-PR

Short memory allocation file is requested

-SH

Task p-sections are allocated sequentially

-SQ

Task is traceable

-TR

XT;n

Task Builder exits after n diagnostics

-XT

4.1.4

System-controlled Partitions

RSX-llM supports two kinds of partitions: user-controlled (or
partitions, and system-controlled partitions.

fixed)

A user-controlled partition can be subdivided into a maximum of seven
fixed subpartitions.
A single task may execute in the entire
partition or up to seven tasks may be simultaneously executing in the
subpartitions.
Execution of tasks requiring the entire partition and
those requiring subpartitions is mutually exclusive.
The main attribute of a user-controlled main partition or subpartition
is that only one task can be executing in the partition at once,
regardless of memory requirments. This can lead to substantial waste
in memory resources, but gives the user absolute control over system
activity.
System-controlled partitions are contiguous areas of memory that are
managed and dynamically allocated to tasks by the Executive. As many
tasks as will fit are allowed to
run
simultaneously
in
a
system-controlled partition. Tasks are allocated a contiguous region
in the partition and are relocated with the memory management unit.
Disregarding checkpointing for the moment,
the
allocation
of
user-controlled
partitions and the subpartitions is simple and
straightforward. Tasks are simply granted residency on the basis of
their priority. When a task is made active, it is merged into a queue
of tasks waiting to gain residency in the partition.
When the
paritition is freed, the highest priority task is granted control of
the partition.
Allocation of a system-controlled partition (again
disregarding
checkpointing)
involves the examination of the current state of the
partition to determine whether there is enough contiguous space to
load a task.
This is accomplished by examining a list of allocated
areas in the partition and calculating the size of the gaps that exist
between allocated areas. The list of allocated areas is sorted by

** The default switch setting is /MM if the system on which the Task
Builder resides has memory management unit; otherwise, the default is
/-MM.

4-6

GENERATING LARGER SYSTEMS
physical memory address, so computation of gap size is simple.
The
first gap that is big enough is allocated to the task, and its
allocation descriptor is merged into the list of allocated areas.
If a biq enouoh oao cannot be found and memory compaction has not been
generated
into- - the
system,
then
the- task is placed in a
priority-ordered list of tasks waiting to be loaded into
the
partition.
Each time a task exits and frees memory in the partition,
the partition wait queue is examined and an attempt is made to
allocate memory to the highest priority waiting task. If memory
compaction has been generated into the system and the shuffler task is
installed, the Executive will activate the shuffler to compact memory,
and an attempt will be made to load tasks again.
RSX-llM
supports
checkpointing
in
both
user-controlled
and
system-controlled partitions. The objective is to avoid preempting a
lower priority task, unless a higher priority task can actually be
brought in and make use of the freed memory. This optimizes the use
of computer system resources while maintaining a priority scheduling
di SC ipline.
Checkpointing in a user-controlled partition occurs under one
conditions:

two

A task requires the user-controlled main partition and is of
a higher priority than any task currently occupying it or any
subpartition of the partition.
Furthermore,
all
the
occupying tasks must be checkpointable, and no task can have
checkpointing disabled. If all these conditions hold, then
all the occupying tasks are rolled out to disk, and the
higher priority task is granted control of
the
main
partition.

A task requires a subpartition of the user-controlled main
partition, and a lower priority task occupies the partition
or the subpartition into which the task is to be loaded.
Furthermore, the occupying task must be checkpointable and
have checkpointing enabled. If all these conditions hold,
then the occupying task is rolled out to disk, and the higher
priority task is granted control of the partition.

Checkpointing in a system-controlled partition occurs as the result of
an allocation failure, as described above. That is, an attempt is
made to allocate a contiguous region of a system-controlled partition
to a task, and an unoccupied memory area of sufficient size cannot be
found. In this case, the list of allocated areas in the partition is
re-examined to determine whether a free space of sufficient size can
be formed by checkpointing one or more tasks. As with user-controlled
partitions, each task considered for checkpointihg must be of a lower
priority, it must be checkpointable, and
it
must
not
have
checkpointing disabled.
The list is scanned from the beginning, looking for a series of
non-contiguous
tasks, where each task satisfies the checkpoint
criteria. If the aggregate sum of the memory occupied by such a
series of tasks is adequate to satisfy the memory requirement for the
higher priority task, then the non-contiguous tasks are rolled out,
and the higher priority task is allocated the freed space and brought
into memory. If such a series of tasks cannot be found, and the
shuffler task cannot compact memory sufficiently to reclaim enough
memory, then no task is rolled out, and the higher pr ior-i ty task is

4-7

GENERATING LARGER SYSTEMS
placed in the partition wait queue.
Note, in this case, that
checkpointing of any task does not improve the situation, because
there is not sufficient contiguous memory that can be reclaimed to
satisfy the requirements of the high priority task.
RSX-llM's philosphy of checkpointing in system-controlled partitions
is to avoid preempting memory unless it can actually be used. No
lower priority task is checkpointed unless it is known beforehand that
enough contiguous space will be made available to load a higher
priority task.
Since the Executive controls the system-controlled partition, the user
should be aware of the following feature. If the partition is the
last real memory partition in the system (not including device
partitions), the partition's size will be dynamically altered at
bootstrap time to use all the memory from the beginning of the
partition to the end of physical memory.
Thus, if a user bootstraps his system on a machine with more or less
memory than he usually runs, the partition will be adjusted to the
proper size. In addition, if the user purchases more memory, a system
generation is not required to make use of this additional memory,
since the additional memory is included automatically.
System-controlled partitions are especially useful in multi-user
systems.
By installing all the user tasks in the partition, space is
used as needed. One user can be running an SK task, such as PIP,
while another user is running a larger task, such as BIGTKB; neither
user needs a dedicated partition. Later, a different task combination
is possible.

4.~

BUILDING SYSTEMS FOR RESIDENCE ON OTHER DEVICES

RSX-llM may use any file-structured device for system residence.
Thus, in addition to the RK05, RP02, RP03, and RP04 disks, the system
may reside on RFll, RS03/RS04, RXOl, or DECtape devices.
A system
including
all
privileged
and
non-privileged
tasks
requires
approximately 2200 blocks of storage. This storage requirement must
be considered when using a single platter RFll, RS03/RS04, RXOl or
DECtape as a system residence device, since these devices have less
than the required space.
It is fairly easy to put a system on another device.

The steps are:

Format the disk, if necessary.

Initialize the disk as a Files-11 volume.

Mount the disk and establish the following UICs on it.
[1,1]

[1,2], [1,4] and [l,50] or [1,54]

PIP all files in UICs [1,1], [1,2], and [1,50] or [1,54]
the system disk to the new device, as shown below:
>SET /UIC=[l,l]
>PIP DSO:=DKO:*.*
>PIP DSO: [l,2]=DKO: [1,2] *·*
>PIP DSO: [l,50]=DKO: [l,50]*.*

4-8

from

GENERATING LARGER SYSTEMS
5.

You will need a virgin image of the target system Executive
to bootstrap on the new device. Thus, in SYSGEN2 you should
specify that a copy be made of the Executive.
Using PIP,
transfer file RSXllM.TSK to the new disk and rename it
RSXllM.SYS. Be sure that it is a contiguous file. If it is
not, make it contiguous by issuing the following command:
PIP [l,50]RSX11M.SYS/CO=[l,50]RSX11M.SYS
NOTE
The target system Executive must have
the device on which it is to reside
generated into the system;
otherwise
the device driver will not be available
to load tasks.

Using the Assign (ASN) command, create the logical
name SY: and assign it to the new disk, as follows:

device

>ASN DSO:=SY:
7.

Run VMR to set up the partitions
tasks.

and

install

the

required

Bootstrap in the new system and
block, as shown below.

save

bootstrap

with

>SAV /WB

4.3

SUPPORTING MORE THAN ONE SYSTEM ON THE SAME VOLUME

Supporting more
straightforward.
[2,54].

than one system on the same device is fairly
Create a UIC for the new system, such as [2,50] or

Copy all files from [1,50] or [1,54]
to the new UIC (or
rename them into the new DIC). Note that tasks may not be
installed in two systems at the same time;
therefore, all
tasks to be used in both systems must be copied.

Create the file RSXllM.SYS from the RSXllM.TSK file.

Run VMR, set up partitions, and install tasks.

Bootstrap the system and save it (with a bootstrap block,
desired) •

There are now two
systems
on
the
same
volume.
If
the
install-run-and-remove option of the Run command is to work properly,
the default UIC for the task image (specified by $ in RUN $xxx) must
be set up for the proper UIC. This may be done using ZAP. Get a map
of MCR and locate the object module RPSOV, set up a relocation
register to the block and starting address of RPSOV, and open the
location at offset 36(8) from the beginning of the module.
This is
the beginning of a text string (1,50] or (1,54], depending on whether
it is an unmapped or mapped system. Change the string to the correct

4-9

GENERATING LARGER SYSTEMS
UIC.

Note that the length of the string must not be changed.
EXAMPLE:
>UFD SY: [2,54]
>SET /UIC=[2,54]
>PIP =[l,54]
>PIP RSXllM.SYS:*/DE
>PIP RSXllM.SYS=RSXllM.TSK
>VMR
ENTER FILENAME: RSXllM
VMR> SET /MAIN=SYSPAR:400:100:TASK
VMR> SET /MAIN=GET:500:1100:SYS
VMR> SET /POOL=400
VMR> INS FCP
VMR> INS INS
VMR> INS MCR
VMR> INS MOU
VMR> INS SAV
VMR> "'z
>ZAP
ZAP>MCR
55:102054:1R
1,36\
055: 102112\133
055:102113\061
-062

-x

>BOO
RSXllM V02 BL12
>SAV
RSXllM V02 BL12 64K MAPPED
>RED DKO:=SYO:
>MOU DKO:
>@[l,2]STARTUP
>@<EOF>

4.4

INCREASING FILE SYSTEM THOUGHPUT

4.4.1

Building a Faster File System

As distributed, the SYSGEN command files build a 2K or 4.5K file
system. The 4.5K file system is two to three times faster than the 2K
system for certain operations. However, increased speed in directory
operations is possible by increasing the number of directory blocks
read in at one time. This becomes important on large disks, such as
the RP02, RP03, and RP04.
To increase or decrease the number of directory blocks read in at one
time,
the user must edit the file BIGFCPBLD.CMD. The size of BIGFCP
can be reduced to 4K by eliminating the buffer allocated to directory
blocks, or it may be increased by .25K for each additional block to be
used. The program section name is $$BUF3.
To reduce the size of
BIGFCP to 4K, delete the line:
EXTSCT=$$BJF3:2000

4-10

GENERATING LARGER SYSTEMS
To increase the number of blocks read, edit the line to increase the
PSECT by the number of additional bytes to be read. Keep in mind that
directory blocks are read in multiples of 1000(8) bytes;
thus, the
PSECT should be altered only by increments of 1000(8). For example,
the s ta temen t:
EXTSCT=$$BUF3:6000
will read in six directory blocks at a time.
NOTE
The file system task should not share a
partition with any task that performs
file-structured
.I/O.
Any
other
configuration will cause a malfunction
of the RSX-llM system. The operation of
RSX-llM is heavily dependent on the file
system.
Nearly
all
tasks
request
processing from it. Thus, for a system
to operate properly, the file system
must be in a partition where it can
react promptly. Most tasks request an
I/O operation through FCP, and while
this I/O operation is in progress, they
cannot be checkpointed. Thus, if FCP is
running in the same partition; it will
not be able to get in to process the
request.

4.4.2

File Control Primitives Dynamic Memory Requirements

A special feature of the file control primitives (FCP)
task allows
directory operations to be speeded up.
However, as many as 54
(decimal) words of dynamic memory may be required for each mounted
volume.
This feature allows directories to be pre-accessed, as follows:
the
first c1me a directory operation is performed on a given directory
(master file directory or user file directory), the directory is
accessed.
An 18-word file control block is stored in the directory
least-recently-used (LRU) list. The LRU list is actually part of the
file control block list pointed to by the volume control block.
Pre-accessing directories in this way generally saves three disk
accesses for every directory operation. The search of the master file
directory for the identification of the user file directory is
eliminated completely, thus eliminating the read of the MFD file
header and MFD data block.
Accessing the user file directory
precludes the necessity for reading the UFD's file header.
The dynamic memory required to effect this time saving may be
intolerable for users with limited dynamic memory space, because up to
three 18-word LRU entries may be required for each mounted volume.
The maximum number of directory LRU-list entries that will be used by
FCP is set at volume mount time by use of the /LRU parameter. The
default is currently set to 3; however, it may be overridden.

4-11

GENERATING LARGER SYSTEMS
4.5

INSTALLING AND USING RESIDENT LIBRARIES AND COMMON BLOCKS

RSX-llM supports resident libraries and common blocks which may be
shared by separate tasks. Resident libraries and common blocks must
be installed in common partitions.

4.5.1

Settiug up a Common Partition

A common partition is established through an MCR
specifies that the partition type is "COM".
following command:
>SET

/MAIN=FORRES:l000:300:COM
'

will set up a main common partition
100000, that is 30000 bytes long.

named

To eliminate a common partition, use the SET
command.

4.5.2

Set Command which
For example, the

"FORRES",

starting

/NOMAIN

SET

/NOSUB

Installing a Library or Common Block

Both the task and symbol table files (.TSK and .STB) must be stored
under UIC [1,1].
Tasks using the common block or resident library
must specify such use during task build. Before a task referencing a
library or common block can be installed, the common block or library
must be installed. This is done with the MCR Install command, as
shown below:
INS

(l,l]FORRES/PAk=iORRES

This command causes the library or common block to be loaded into the
partition.
Tasks that reference the library or common block can then
be installed.

4.5.3

Eliminating a Library or Common Block

A common block or library can be eliminated only through the
command. For example, the command:

MCR

Set

SET /NOMAIN=FORRES
will eliminate the library FORRES (installed as shown in section 4.5.2
above) and the partition FORRES.
Care should be exercised when
eliminating a library or common block to remove any tasks which
reference it befo·re removing the library.

4.6

BUILDING A FORTRAN OTS MEMORY-RESIDENT LIBRARY

In order to save memory space, a user may wish to build a shareable
(resident) library containing frequently-used FORTRAN routines.

4-12

GENERATING LARGER SYSTEMS
A MACR0-11 source file designed to aid the user in building a resident
This file, named
FORTRAN library is distributed with RSX-llMo
As distributed, this
FORRES.MAC, is located under UIC [200,200].
module generates an 8K library for a system with no extended
arithmetic support.
While it is not necessary to build the resident library during system
generation, memory space for it should be considered at that time.
Otherwise, it may be difficult to install the library.
Building an optimal resident library is an interactive process.
The
file FORRES.MAC must be edited with consideration for size and program
requirements.
For most efficient use of virtual memory, the library should be as
close as possible to a 4K boundary, without exceeding that boundary.
The following steps are a guide to tailoring the FORTRAN library to
individual system requirements.
This process assumes that a system
generation has already been done, that all tasks are installed, and
that the system has been saved.
1.

For easy reference to the
FORRES.MAC may be useful:

MACR0-11

file,

listing

>SET /UIC=[200,200J
>PIP LP:=FORRES.MAC
2.

The MACR0-11 file may now be edited to insert or remove entry
points for various routines.

Assemble the MACR0-11 file:
>MAC FORRES=FORRES

Task build the file, specifying the /-HD and /-PI switches on
the symbol table file. The only Task Builder option required
is STACK=O. The appropriate command sequence is shown below:
TKB>,LP:,SY:FORRES/-HD/-PI=FORRES
TKB>/
ENTER OPTIONS:
TKB>STACK=O
TKB>//

Check the Task Builder map to obtain the size of the library
(in octal bytes), as shown in the line "R/W MEM LIMITS:". If
the library will not fit into the partition planned for it,
examine the rest of the map to obtain the sizes of individual
modules. Remember that the library should be as close as
possible to a 4K boundary without exceeding that boundary.

Edit (or re-edit)
the MACR0-11 file, adding or removing
semicolons in accordance with size and usage constraints.

Re-assemble the MACR0-11 file and task build it again in
order to verify the size of the library. The short TKB map
may be specified, as follows:
>MAC FORRES=FORRES
>TKB
TKB>,TI:/SH,SY:FORRES/-HD/-PI=FORRES

4-13

GENERATING LARGER SYSTEMS
TKB>/
ENTER OPTIONS:
TKB> STACK=O
TKB>//
8.

When the library is the desired size, task build it in UIC
[1,1), specifying all output files.
In addition to the
STACK= option, the PAR= option must be used, as shown below:
SET /UIC=[l,l]
>TKB
TKB>FORRES,TI:/SH,SY:FORRES/-HD/-PI=[200,200]FORRES
TKB>/
ENTER OPTIONS:
TKB> STACK=O
TKB> PAR=FORRES:base:length
TKB>//
(If there is a system resident
LIBR=SYSRES:RO may also be used.)

TKB

option

11. Now perform a SAVE, or edit the [1,2] STARTUP.CMD
file to install FORRES at bootstrap time.

command

library,

the

Allocate space for the library, as shown below:
>SET /MAIN=FORRES:base:length:COM

10. Install the library using the following command:
>INS FORRES

4.7

BUILDING AN FCS POSITION-INDEPENDENT MEMORY-RESIDENT LIBRARY

In order to save memory, a
user
may
wish
to
produce
a
position-independent library containing all the FCS modules required
by a set of programs that are intended to run simultaneously.
The MACR0-11 source file, [200,200] FCSRES.MAC, is provided to aid the
user in building an FCS resident library. As distributed, this file
produces a 4K library and contains every module in FCS.
A user
wishing to produce a smaller version of FCSRES may do so by editing
the FCSRES.MAC file, eliminating the .GLOBL references for those
modules to be excluded.
It should be understood that the number of FCS modules that goes into
a resident library is dependent on which programs are going to use FCS
and what modules those programs require. An essential point to bear
in mind is that a program which links to FCSRES will not be able to
link in missing FCS modules from the SYSLIB.OLB object library.
Such
a program cannot link with FCSRES. It must instead link all of FCS
from SYSLIB.

4-14

GENERATING LARGER SYSTEMS
To build FCSRES:
1.

Edit the FCSRES.MAC filei if desired.

Assemble it, as shown below:
MAC>FCSRES=FCSRES

Task build it with the following command sequence:
>SET /UIC=[l,l]
)TKB
TKB>FCSRES/-HD/PI,LP:,SY:FCSRES=[200,200]FCSRES
TKB>/
ENTER OPTIONS:
TKB>STACK=O
TKB>PAR=FCSRES:base:size
TKB>//
The base address for FCSRES will always be 0 for a mapped
system.
For an unmapped system, it will be the actual
common block base address. The size of the partition may
be assumed to be 4K.
After one task build, a more
accurate size can be given.

Set up a common block, as shown below:
>SET /MAIN=FCSRES:l400:200:COM
The above example sets up a 4K common block at 140000.

Install the library using the following command:
>INS [l,l]FCSRES

Save the system or edit the [1,2] STARTUP.CMD command file to
install FCSRES at bootstrap time.

If a FORTRAN resident library is also desired in conjunction with an
FCS resident library, build FCSRES first, then include the Task
Builder option statement:
LIBR=FCSRES:RO
in the task build commands for FORRES. This, of course, results in a
considerably smaller FORRES.
By properly tailoring the FORTRAN
resident library, a very flexible 4K FORRES using a 4K FCSRES can be
built.
This would allow the use of FCSRES by itself when FORRES was
also not required.

4.8

PRINT SPOOLER (PRT}

A print spooler (PRT) task is provided which can be built to print
listings simultaneously with the operation of other tasks. The
MACR0-11 assembler, Task Builder~ and FORTRAN compilers will try to
spool all listing or map files generated by them onto disk. If the
print spooler is not installed, the files will remain on disk. If the
spooler is installed, it will be started, thus printing any files in
its receive list (input queue). When a file has been printed, it will

4-15

GENERATING LARGER SYSTEMS
be deleted. Using the print spooler speeds up the operation of
MACR0-11, the Task Builder, and compilers, since they do not have to
wait for I/O to complete on the relatively slow (compared to disk)
line printer.
For optimum operation, the print spooler should be
installed in its own partition.
The print spooler requires a 2K
partition in which to run.
It may be built checkpointable or not
checkpointable. However, if it is built non-checkpointable, it will
run considerably faster because it will double-buffer printing. The
Executive limits checkpointable tasks to one outstanding I/O request.
Non-checkpointable tasks do not have this restriction.
It is possible for the user to build a non-deleting print spooler.
This is accomplished by editing the task build command file PRTBLD.CMD
during SYSGEN phase 2. When the Line Text Editor (EDI) prompts, enter
the following commands:
F GBLPAT
GBLPAT=PRT:$DELET:l
*C/l/01
GBLPAT=PRT:$DELET:O
*EX

4.9

BUILDING MULTI-USER SYSTEMS

As mentioned before, RSX-llM is capable of supporting more than one
terminal which is running tasks. Such systems are called multi-user
systems. Depending on the processor, disks, and memory size, RSX-llM
can support from 2 to 16 or more users at the same time.

4.9.l

Setting Up a Multi-User System

At system generation, more than one terminal should be generated into
the system. These terminals may consist of any combination of DLlls,
DJlls, or DHlls. DHlls may have either local or dial-up lines.
When the partitions are laid out, the system manager has the choice of
setting
up
fixed
partitions
or system-controlled partitions.
System-controlled partitions must be selected at system generation
time.
In a fixed partition system, each user terminal is allocated a
dedicated partition in which to run its tasks. In a system-controlled
partition, the Executive parcels out memory to user tasks, as
required. This scheme is much more flexible than a fixed partition
scheme, since only the space needed for the current task mix is used.
In a fixed partition system, much space could be wasted if partitions
are made big enough to run tasks such as BIGTKB and BIGMAC; most work
is done using smaller tasks such as EDI or PIP.

4.9.2

Installing and Naming Tasks

A single task image may be installed any number of times
as long as it is not checkpointable.

Rsx~llM,

Checkpointable tasks must be copied to allow multiple users to run
them.
For example, the indirect file processor task (AT.} is
checkpointable, and copies of it must be made to allow more than one

4-16

GENERATING LARGER SYSTEMS
user to run it.
Thus, there
should
be
INDO.TSK,
INDl.TSK
- •.• INDn.TSK, where the number n after the file name "IND" indicates
the terminal to which the task image belongs.
INDO should be
installed as AT.O, INDl.TSK as AT.I, and INDn.TSK should be installed
as AT.n. The parameter n may be any octal number from 0-377(8}.
Similarly, other checkpointable tasks should be copied and named
accordingly {a checkpointable MACR0-11 in file MACn.TSK would be named
MACn). The applicable convention is that the task name should be
built from the MCR command name used to invoke it (e.g., MAC) and the
terminal number from which it is to run.

4.9.2.1

Fixed Partition Systems

For fixed partitions, all tasks to be started by MCR, rather than a
Run command, should be installed in the above manner. In additionf
for the install-run-and-remove option of the Run command to work
properly on fixed partition systems, the task INS should be installed
in each user partition dedicated to a terminal. Tasks started by the
install-run-and-remove command will automatically be named TTn, where
n is the unit number of the terminal from which it was started.
The
task will run in the partition that the task INSn was running in at
the time it installed the task. The partition and naming conventions
may be overridden by explicitly naming the task and partition in the
Run command.

4.9.2.2

System-Controlled Partitions

For multi-user systems with a system-controlled partition,
the
Executive performs the naming and creating of tasks. Except for
checkpointable tasks, only one copy of each task needs to be installed
in the system-controlled partition. The default name should be that
used in single-user systems {e.g., MACR0-11 is ••• MAC).

4.9.2.3

Task Invocation

When the user types the command to invoke a task (e.g., >MAC A,A=B),
the system executes the following algorithim, where XXX is the command
name:
1.

Search for task .•• XXX. If found and not active, start
and assign TI: to the terminal which issued the command.
active, issue "task active" message.

it
If

Otherwise:
2.

Search for task XXXn. If found and not active, start it
assign TI: to the terminal which issued the command.
Otherwise:

and

If active, issue "task active" message.

If task XXXn is not found and task ••• XXX is found, and task
••. XXX is installed in a system-controlled partition, then
duplicate the task control block entry for task ••• XXX, name

4-17

GENERATING LARGER SYSTEMS
the task xxxn, start it, and assign TI:
to the invoking
terminal.
On task exit, deallocate the temporary task
control block entry.

4.9.3

If tasks •.. XXX and XXXn are not found, issue
SYSTEM" message.

"TASK

NOT

For the install-run-and-remove option of the Run command, the
task is always named TTn, where n is the unit number of the
terminal from which it was requested. It is installed in the
partition that task INS is running in, unless overridden by
the /PAR= option.

Memory Compaction

The shuffler task SHF ••• works in conjunction with the Executive to
assure that space in a system-controlled partition is used to its
fullest. Thus, if small blocks of memory exist where none are large
enough for a task to run in, but the total free space is large enough,
the Executive will call task SHF ••. to move all tasks down in memory
to reclaim the free space into one contiguous piece big enough to
accommodate the task. This option must be selected at SYSGEN time and
the task SHF...
must be built. SHF ••• should be installed in a
user-controlled partition with tasks that execute for short periods of
time, thus allowing it to respond quickly to Executive requests. It
should not run in a system-controlled partition or in the same
partition as FllACP or any other ACP task.

4.9.4

Round-Robin Scheduler

There is an option at system generation time to include a round-robin
scheduler into the Executive.
This scheduler time-slices the CPU
among tasks in a priority range selected at system generation time.
It ensures that all tasks of equal priority share in the CPU
resources. But, since RSX-llM is a priority-structured system, higher
priority tasks will be allocated more of the CPU time than lower
priority tasks. Tasks not in the priority range are unaffected.

4.9.5

Checkpointing During Terminal Input Wait

A SYSGEN option exists which allows tasks that are checkpointable to
be displaced to the disks while they are waiting for input from a
terminal. This feature allows more copies of terminal I/0-bound tasks
(e.g., EDI} to be run than normally could be run in a given amount of
memory. However, the tasks must be built as checkpointable, which
requires more disk space than non-checkpointable tasks.
When a checkpointable task is waiting for terminal input, its
effective
priority is zero (its real priority never changes).
Therefore, lower real priority tasks ready to run will cause a higher
real
priority task that is waiting for terminal input to be
checkpointed. When the terminal input to the checkpointed task
completes, the higher priority task will be brought back into memory,
causing other lower priority checkpointable tasks to be checkpointed,

4-18

GENERATING LARGER SYSTEMS

if necessary, to make room for it. A task which has been checkpointed
while awaiting terminal input will not be brought back into memory
until its terminal request has been satisfieda This is true even if
there is enough memory available for it.

4-19

CHAPTER 5
VIRTUAL MONITOR CONSOLE ROUTINE

5.0

(VMR)

INTRODUCTION

The Virtual Monitor Console Routine (VMR) is an RSX-llM system progra.
that allows the complete on-line generation of RSX-llM and RSX-llS
systems. As its name suggests, VMR provides the capability to execute
MCR commands that are directed to the disk image of a system.
VMR supports a compatible subset of the MCR commands described in the
RSX-llM Operator's Procedures Manual.
All supported commands have
identical syntax and semantics as their MCR counterpart with one
exception, which is the Save command.

5.1

INITIATING VMR

There are six methods for initiating VMR. These methods are described
below.
The first four methods apply when VMR is installed and ready
to be executed.
The remaining two cause VMR to be installed;
executed, and then removed on exit.
Method 1:
>VMR @indirect
VMR is loaded, executes the specified indirect command file,
then exits.
Method 2:
>VMR
VMR is loaded and responds with the following prompt:
ENTER FILENAME:
Method 3:
>RUN ••• VMR
VMR is loaded and responds with the following prompt:
ENTER FILENAME:

5-1

and

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Method 4:
>RUN ... VMR/UIC=[g,m]
The UIC under which VMR executes is explicitly specified for this
run only. Normally, VMR executes with the default UIC associated
with the initiating terminal. VMR is loaded and responds with
the following prompt:
ENTER FILENAME:
Method 5:
>RUN $VMR
VMR is installed from the system UIC ([1,50] or [1,54]),
and responds with the following prompt:

loaded,

ENTER FILENAME:
Method 6:
>RUN $VMR/UIC=[g,m]
The UIC under which VMR executes is explicitly specified for this
run only. Normally, VMR executes with the default UIC associated
with the initiating terminal. VMR is installed from the system
UIC ( [ 1, 50] or [ 1, 54]) , loaded, and responds with the following
prompt:
ENTER FILENAME:
In all methods of initiation except method 1, VMR prompts with:
ENTER FILENAME:
At this point, the user enters the file specifier for the file that
contains the image of the RSX-llM or RSX-llS system he wishes to
modify. A symbol table file corresponding to the system image file
with the same filename, but of type .STB, must be under the same UIC.
(See exception described for the /IM switch below.) For example, if
the file specifier RSXllS.SYS was entered in response to the
ENTER FILENAME:
prompt, there must be a corresponding file named RSXllS.STB
contains the symbol definitions for the RSXllS.SYS system image.
The system image and symbol table files are
produced by the RSX-llM Task Builder.

the

standard

that
format

Upon receiving a valid file specifier, VMR proceeds to open and read
the symbol table file;
it also opens the system image file and
verifies that the system image file is indeed a system image.
VMR
then prompts with
VMR>
and awaits the entry of a command. The user may enter any
command or terminate VMR by typing CTRL/Z*.

valid

*CTRL/Z is typed by simultaneously depressing the CTRL and Z keys.

5-2

VMR

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

If method 1 is used to initiate VMR, the specified file is
to contain a series of commands to be executed by VMR.
The first record in the file must contain a
system image file that is to be modified.

file

considered

specifier

for

the

In addition to accepting an indirect file during initiation, the
may also enter

user

@f ilespec if ier
anytime that VMR is prompting for input. This directs VMR to obtain
input from the specified file until an end-of-file is reached and then
to return to the terminal for further input. This feature of VMR is
very useful in cases where a series of Install commands, for example,
are to be repeated for several generated systems.

5.2

VMR FILE SPECIFIERS

File specifiers accepted by VMR are in the standard RSX-llM form:
dev:[g,m]filename.type;ver/sw
where:
dev:

is the device-unit on which the
volume
containing the desired file is mounted, for
example, DKO: or DTl:. The name consists of
two ASCII characters followed, by an optional
1- or 2-digit octal unit number.
Default:

[g ,m]

SYO:

is the user identification
code
(UIC),
consisting of a group number and a member
number, associated with
the
user
file
directory (UFD) containing the desired file.
Default:

filename

the UIC under which VMR is running.

is the name of the file.
In RSX-llM, a
filename can be up to nine alphanumeric
characters in length. Filename and type are
always separated by a period (.)-.
Default:

typ

none;

must be specified.

is the the file type, consisting
three alphanumeric characters.

Default: SYS for system image files, and CMD
for indirect files.
ver

is the version number in the range 0

through

77777(8).

Default:
/sw

latest version.

is a 2-character
switch option.

5-3

ASCII

name

identifying

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Only one switch option (/IM) is provided with VMR, and it is used in
conjunction with the Save command to write non-RSX-llM/RSX-llS system
images to secondary media in bootstrappable format. The switch may be
specified only on the system image file and overrides the requirement
for the symbol table file.
When the /IM switch is specified, the only command VMR will accept is
the Save command. The /IM switch is useful in cases where RSX-llM is
used as a host system to develop and create small, core-only systems
that must be loaded from a serial media such as paper tape.

5.3

COMMAND SUMMARY

VMR commands are classified into four categories:
initialization,
informational, task control, and system maintenance. The following
command summary lists each command by category, the page on which it
may be found, and a brief description of the command.
Initialization Commands
ASSIGN

[5-8 ] Assign a logical name to a physical device.

INSTALL

[5-10] Install a task into the system image.

SET

[5-15] Set system parameters.

TIME

[5-24] Establish or display the time as stored in
system image.

the

Informational Commands
of
peripheral
system on the

DEVICES

list
[5-25] Display the
by the
recognized
terminal.

LUNS

[5-26] Display the list of LUN assignments for
indicated task on the entering terminal.

PARTITIONS

[5-27] Display the list of partition
the entering terminal.

TASKLIST

[5-28] Display the system
entering terminal.

task

devices
entering

definitions

directory

the

Task Control Commands
ALTER

[5-29] Alter the priority of a task.

CANCEL

[5-30] Cancel time-based
task.

FIX

[5-31] Fix a task in
system image).

5-4

initiation

memory

(task

requests
is

copied

for

into

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

REASSIGN

[5-32] Change LON assignment(s).

REDIRECT

[5-33] Redirect all I/O requests
device to another.

REMOVE

[5-35] Remove a task from the system image.

RUN

[5-36] Schedule a task's activation. The task may run
immediately when the system is bootstrapped,
after a delay, or in synchronization with the
Periodic
rescheduling
is
system
clock.
optional.

UNFIX

[5-40] Unfix a task in the system image.

from

one

physical

System Maintenance Command
SAVE

5.3.1

[5-41] Write the system image of memory onto
secondary media in bootstrappable format.

Command Syntax

It is not necessary to type in the entire command name when entering a
command.
VMR requires only the first three letters of a command,
followed by the command parameters. If parameters are entered, they
must be preceded by at least one blank or tab.
The following example shows how the Time command may be specified.
Note that brackets are used in the example to indicate that the "E" is
optional; the square brackets are not actually entered as part of the
command name.
VMR>TIM[E]<CR>*
14:00:04 22-AUG-73
Thus, either
't'71.An,mTnAt:t / r t n '
v !'!.['\./ .L .1.l'l.C:.. '~.['\./

or
VMR>TIM<CR>
is acceptable.

*The angle brackets are used to denote that the enclosed character(s)
are not taken literally as part of the command syntax. For example,
<CR> indicates that a carriage return terminates the input line.

5-5

VIRTUAL MONITOR CONSOLE ROUTINE
5.3.2

(VMR}

Keywords

Some commands use keywords which generally apply to a command
argument.
A keyword consists of a slash (/), followed by an ASCII
identification, and optionally followed by an equal sign (=) and the
value of the keyword, as follows:
/Keyword=value
Keywords can be submitted in any order.
As an example of keyword
usage, the Install command requires a filename argument specifying the
task to be installed. Keywords can be appended to the filename.
One
such keyword, /TASK, can specify the name under which the task is to
be installed. Thus:
VMR>INS JIM/TASK=SUPER
will cause the task contained in the file named JIM.TSK to be
installed with the name SUPER. Keywords are command specific and,
thus, are defined with each command.

5.3.3

Comments

VMR will treat a line of text as a comment if the first character in
the line is a semicolon (;). In addition, the exclamation mark (!)
may be used to delimit comments on a command line.
The first
exclamation mark starts the comment, and the next exclamation mark, or
the end of the line, terminates the comment. Any text between the two
exclamation marks is ignored, as shown below:
>;

THIS LINE IS A COMMENT

>TAS !THIS IS A COMMENT STRING!
This feature is especially useful in providing
files.

5.4

comments

indirect

VMR COMMANDS

Command names are at least three characters in length.
VMR accepts
the first three characters of a command and then searches for a blank,
tab, carriage return, or ESCape character. Therefore, embedded blanks
are not allowed in a command name.
If a command is entered
incorrectly, an error message will be displayed at the entering
terminal.
Syntactical descriptions of commands and messages described
chapter observe the following notational conventions:

this

Lower case indicates a variable whose actual value is
determined when the command is entered or when the message is
issued. For example, the value of "taskname" depends on the
name of the task associated with the command or message.

Brackets [] enclose optional items.
A syntactical element
enclosed in brackets (.e.g., [dt]) may or may not be included
in the command, at the user's option. In the UIC keyword,

5-6

VIRTUAL MONITOR CONSOLE ROUTINE
square brackets are part of the
exception is noted as it occurs.
3.

(VMR)

required

syntax.

This

Unless explicitly qualified, all numeric values required in a
command may be entered as decimal or octal. Decimal values
are indicated by a trailing period;
octal values are
indicated by the absence of a trailing period. Thus,
255.
and
377
have the same value.
value 377 is octal.)

5.4.1

(The value 255.

is decimal, while

the

Command Description Format

Each command is described using
that do not apply are omitted.

four

subheadings.

Any

subheadings

COMMAND NAME
This section is headed by the command name in capital letters.
right margin; the minimal VMR command mnemonic is shown.

On the

Following the heading, the action of the command is described.
Format:
The command format is given, and all parameters are described.
Examples:
Example are shown.
Command-Related Error Messages:
Error messages specific to the command appear in this section.
The command may also produce messages listed in section 5.5
(common error messages), but these are not listed here.
Notes:
A list of special considerations that may prove helpful
assisting the programmer in the proper use- of the command.

5-7

VIRTUAL MONITOR CONSOLE ROUTINE
5.4.2

(VMR)

Initialization Commands

ASSIGN

ASN

The Assign command is used to define, delete, or display logical
device assignments.
Logical device assignments are a convenient way
to associate logical names with physical devices. There are two types
of logical device assignments:
local and global.
Only global
assignments are supported by VMR (see the
RSX-llM
Operator's
Procedures Manual for a complete description of local assignments).
Logical device names may be the same as physical device names, or they
may be purely random, as long as they follow the syntax for device
names. This syntax requires a 2-character ASCII device name followed
by an optional 1- or 2-digit octal unit number. When referring to a
logical device, the name must be terminated by a colon, just as any
physical device.
Format:
ASN ppnn:=llnn:/GBL
Assign logical name llnn
assignment is global.

physical

device

ddnn.

The

ASN /GBL
Display all global assignments on the entering terminal.
ASN =llnn:/GBL
Delete the global logical device assignment of logical
llnn.
ASN =/GBL
Delete all global logical assignments.
where:
pp - physical device name.
nn - unit number.
11 - logical device name.
Example:
VMR>ASN DK:=GBO:/GBL !DEFINE GLOBAL DEVICE GBO:
VMR>ASN /GBL
!LIST ALL ASSIGNMENTS
GBO: DKO:
!GLOBAL
VMR>ASN =/GBL
!ELIMINATE ALL GLOBAL ASSIGNMENTS

5-8

name

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Command-Related Error Messages:
VMR -- LOGICAL DEVICE NOT IN SYSTEM
The specified logical device name has not
been
previously defined and, therefore, cannot be found in
the logical device assignment table.
VMR -- PSEUDO DEVICE ASSIGNMENT ERROR
A logical device name may not be assigned to a
device.

pseudo

Notes:
1.

VMR supports only global assignments, and
syntax.

5-9

/GBL

required

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

INSTALL

INS

The Install command is used to make tasks known to the system and to
load resident libraries into common partitions. Filename is the only
required argument. The keywords are optional arguments f6r parameters
that were either not specified at task-build time, or are to be
overridden during the performance of the Install command. The process
of task installation results in a task control block (TCB) being
constructed and recorded in the system image. Once installed, a task
may be acted upon by specifying its name in an appropriate command.
Format:
INS[TALL] filespec[/keyword] [/keyword] ...
Keywords:
/CKP
option
/INC = extension size
/PAR
pname
/PMD
option
/PRI
number
/TASK
taskname
/UIC = [group,member]
where:
filespec - dev: [group,member]filename.type;version
The standard defaults for file specifiers
TSK is the default type.
CKP

- Checkpoint option.
If CKP=YES (default
for
checkpointable
tasks),
checkpointing
of
a
checkpointable task will be allowed. If CKP=NO,
checkpointing will be disabled for the task.
Default:

INC

apply.

yes, for checkpointable tasks.

- The number of words to be allocated to a task
which is run in a system-controlled partition.
The extension size is usually expressed in octal,
but may also be expressed in decimal by appending
a period. This keyword will override the EXTTSK
parameter used at task-build time and is used for
tasks which determine the size of the partition
in which they are running in order to decide how
much dynamic space they may allocate to symbol
tables or buffers. Tasks such as MACR0-11, Task
Builder, PIP, LBR, and VFY, use the space from
the end of the task image to the end of the
partition as dynamic memory. In effect, the INC
keyword defines the number of words in this
dynamic region.
Default:
zero.

size specified at

5-10

task-build

time

VIRTUAL MONITOR CONSOLE ROUTINE
PAR

- A partition name which overrides the partition
specified at task-build time.
In an unmapped
system, the partition specified must have the
same base address as the partition the task was
built for.
Default:

PMD

the Task Builder uses the name GEN.

- Post-Mortem
Dump
option.
If
PMD=YES,
a
Post-Mortem dump will be requested for the task
if it aborts due to an SST error condition.
If
PMD=NO (the default at task-build time), no dump
will be requested at SST abort time.
This
keyword will override the /PM switch specified at
task-build time.
Default:

specified at task-build time.

Task priority. The value range is
250(10). Number conventions apply.

PRI

Default:
TASK

(VMR)

50.

through

(decimal).

- Taskname. The name by which the task
referenced.

Default: name specified at task-build time or
the first six characters of the tllename.
(See
the RSX-llM System Generation Manual for an
explanation of the multi-user system default.)
UIC

- [group,member]. The square brackets are required
syntax.
This UIC is the default for the task.
It can be overridden in the Run command.
The
task UIC determines the file protection class
(system, owner, group, world) of the task, thus
directly influencing file access.
Default:

default from task build.

Examples:
VMR>INSTALL SCAN
Install task SCAN.

Priority and UIC are defaulted.

VMR>INSTALL DKl: [l,54]SCAN.TSK;4/PRI=l03
Install task SCAN, with file type TSK, version 4 on device
DKl:, group number 1, member number 54, and a priority of
103(8).
VMR>INS MAC/INC=4096./PAR=SYSCTL
Install task MAC in the system-controlled partition
and increase the task size by 4096. words.

5-11

SYSCTL

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Command Error Messages:
VMR -- BASE ADDRESS MUST BE ON 4K BOUNDARY
The base address of the task is not on a 4K
Applicable only to mapped systems.

boundary.

VMR -- BASE MISMATCH COMMON BLOCK <common-name>
The base address of the partition does not match that
of the common block.
Applicable only to unmapped
systems.
VMR -- CHECKPOINT AREA TOO SMALL
The area allocated for checkpointing the task is
smaller than the partition into which it is being
installed.
VMR -- COMMON BLOCK IS TASK PARTITION <common-name>
A task's request for access to a common block is
rejected because the partition requested is a task
partition.
VMR -- COMMON BLOCK NOT LOADED <common-name>
The specified common block is linked to the task, but
has not been loaded into the system. Install the
common block, then install the task.
VMR -- COMMON BLOCK OCCUPIED
An attempt has been made to load a common block
is already occupied.

which

VMR -- COMMON BLOCK PARAMETER MISMATCH <common-name>
Parameters of a common block do not match those in the
task's label block.
VMR -- DEVICE ddnn:

NOT IN SYSTEM

Device specified during task build by the
keyword option does not exist in the system.

Assign

VMR -- FILE NOT CONTIGUOUS
An attempt is being made to install a task from a
non-contiguous file.
Task images may exist only in
contiguous files.
VMR -- FILE NOT TASK IMAGE
Data in the label block is not correct, indicating the
file is not a task image.
VMR -- ILLEGAL DEVICE/VOLUME
The device specified is not
device.

5-12

valid

task

residence

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR}

VMR -- ILLEGAL FIRST APR
The Task Builder allows the user to specify the
virtual base address of a task image as a multiple of
4K words.
Privileged tasks
normally
start
at
100000(8) in order to allow them to be mapped into the
first 16K of the Executive, but they may also start at
120000(8} or 140000(8}.
Non-privileged tasks must
start at O.
VMR -- ILLEGAL PRIORITY
The value of the priority in the command
range, i.e., not 1 to 250(10), inclusive.

out

VMR -- LENGTH MISMATCH COMMON BLOCK <blockname>
The length parameter for the common
block,
as
described in the label block for the task image, does
not match the corresponding length parameter defined
in the system. A task's label block data must match
system data for that task before it can be installed.
VMR -- NOT ENOUGH APRS FOR TASK IMAGE
A privileged task built to run using APR 4, 5, or 6 as
its base is too large. Using APR4, the task is larger
than 12K; using APR5, the task is larger than 8K;
using APR6, the task is larger than 4K. Applies only
to mapped systems.
VMR -- NO ROOM AVAILABLE IN STD FOR NEW TASK
This message means that there is no dynamic memory
available to build the system task directory (STD)
entry (task control block}; therefore, no task can be
installed.
VMR -- PARTITION NOT COMMON
A partition specified for a common
defined as a common partition.

library

not

VMR -- PRIVILEGED TASK LARGER THAN 12K
Privileged tasks have a maximum size of 12K.
VMR -- SPECIFIED PARTITION FOR COMMON BLOCK
A task is being installed into a common block.
VMR -- SPECIFIED PARTITION TOO SMALL
The task being installed is larger than the
into which it is being installed.

partition

VMR -- TASK AND PARTITION BASES MISMATCH
The base of the partition does not match that of the
task
being installed.
Applies only to unmapped
systems.

5-13

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- TASK IMAGE I/0 ERROR
The Install command cannot read the task image file.
VMR -- TASK IMAGE CURRENTLY INSTALLED
The requested task image, which is checkpointable, is
already installed.
Checkpointable tasks may only be
singly installed;
other tasks may
be
multiply
installed.
VMR -- TASK NAME ALREADY IN USE
An attempt has been made to install a task
same name as one already in the system.

with

the

VMR -- TOO MANY COMMON BLOCK REQUESTS
A task is limited to three common blocks.
VMR -- TOO MANY LUNS
A task has requested more
assigned.

than

255(10)

LUNs

VMR -- UNDEFINED COMMON BLOCK <blockname>
A task references a common block that is not defined
in the system. Usually, this indicates that the task
was built for another system.
Notes:
1.

In order to install a task, the specified residence device
(i.e., the device from which the task is being installed)
must exist in both the host system and the system image.
This is necessary in order to read the task image and to
create the task control block. In most cases, the device is
defaulted to SY:, and since all systems have an SY:, this
default is always valid. However, a problem sometimes arises
in building RSX-llS systems when the residence device exists
only in the host system. The intent is to install and then
to fix the task in the system image, but the install cannot
be completed because the residence device does not exist in
both systems.
This problem cpn be solved with the Assign
(ASN) command. Suppose that a task exists on DK2: and that
it is desired to install and fix the task in an RSX-llS
system that does not contain a DK2:.
A logical assignment
can be made in the system image that associates the device
name DK2: with any valid device (e.g., TTO:); then the task
can be installed and fixed.
The following VMR command
sequence:
VMR>ASN TT:=DK2:/GBL
VMR>INS DK2:TASK
VMR>FIX TASK
will solve the problem.

5-14

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

SET

The Set command makes it possible to alter a number of system-wide
parameters. In particular, the Set command enables the user to:
1.

Establish the default buffer size for a device;

Establish terminal device characteristics;

Establish a terminal as privileged;

Characterize a terminal as a slave terminal (i.e., a terminal
from which the system will not accept unsolicited input);

Establish a default UIC for a terminal;

Create partitions or subpartitions;

Add space to the dynamic storage region, and

Cause all writes to an RKOS disk to be automatically followed
by a write check to ensure that the data was written
correctly.

Format:
SET /Keyword=values
Only one keyword per command is permitted.
All the keywords
except BUF, POOL, and UIC may be prefixed by NO, which negates
the effect of the keyword. Specifying NO in conjunction with the
keywords BUF, POOL, and UIC produ~es a syntax error.
The keywords for setting device characteristics follow:
/BUF

= dev : [size]

/LA30S [= dev:]
/LOWER [= dev:]
/PRIV

[= dev:]

/SLAVE [= dev:]
/SPEED = dev: [recv:xmit]
/UIC

[= [group,member]]

/UIC

[= [[group ,member] : ] dev:]

/VTOSB [= dev:]
The effects of these keywords are as follows:
BUF

Set or display the default buffer size of the
specified device.
If the size is not entered,
this keyword results in the current size being
displayed.
BUF
is
particularly useful for
defining line printer width (80 or 132 columns).

5-15

VIRTUAL MONITOR CONSOLE ROUTINE
LA30S

(VMR)

- The LA30S keyword establishes
the
specified
terminal device as an LA30S.
If the device
parameter (= dev:) is omitted, all LA30S terminals
are displayed.

NOLA30S - Resets the terminal to other than an LA30S.
If
the device parameter (= dev:) is omitted, all
non-LA30S terminals will be displayed.
LOWER

- The LOWER keyword specifies
that
lower-case
characters are not to be converted to upper-case
when received from the specified terminal. If the
device
parameter
(= dev:)
is
omitted, all
terminals for which lower-case characters are not
converted are displayed.

NOLOWER - Resets the terminal so that lower-case characters
received on input will be automatically converted
to upper-case and echoed as upper-case.
The
absence of the device parameter (= dev:) results
in a display of all terminals for which conversion
to upper-case is specified.
PRIV

- The PRIV keyword will set the identified device to
a privile~ed terminal.
If the device parameter
(= dev:)
is omitted, VMR
will
display
all
terminals that are privileged.

NOPRIV

- The NOPRIV keyword sets the identified terminal to
non-privileged status. The absence of the device
parameter (= dev:) results in a display of all
non-privileged terminals.

SLAVE

- The SLAVE keyword establishes the
identified
device as a terminal that can only enter data if
it is solicited from a task;
thus, unsolicited
input is always rejected.
The absence of the
device parameter (= dev:) will result in the
display of all terminals currently classified as
slaves.

NOSLAVE - Resets slave status. The absence of the device
parameter (= dev:) results in the display of all
non-slave terminals.
SPEED

- The SPEED keyword establishes the receive and
transmit baud rate for terminals attached to the
system through a DHll multiplexer. The parameter
"recv" is the receive baud rate for the DHll,
while "xmit" is the transmit baud rate for the
DHll.
Both must be specified in setting the
speed.
If neither is specified, the current
settings will be displayed. Valid baud rates are:
0, 110, 150, 200, 300, 600, 1200, 1800, 2400,
4800, 9600, A (user-specified baud rate A), and B
(user-specified baud rate B).

VT05B

- The VT05B keyword sets the terminal as a VTOSB (a
2400 baud VT05). If the device parameter (= dev:)
is omitted, all terminals classified as VT05Bs
will be displayed.

5-16

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

NOVTOSB - Resets the terminal to other than a VTOSB. If the
device
parameter
(= dev:)
is
omitted, all
terminals classified as non-VTOSB terminals will
be displayed.
UIC

- The specified UIC is established as the default
UIC for the specified terminal. All tasks that
are to run from the terminal, after the system is
bootstrapped,
will run with the UIC of the
terminal, unless explicitly overridden with the
UIC switch in the Run command.
Furthermore,
external* MCR function tasks are requested with
this UIC. If a UIC is not specified, the current
UIC will be displayed. If the UIC is specified
with the parameter dev:, the default UIC for the
specified terminal will be set. If the UIC is not
specified, but the parameter dev: is, the default
UIC for the specified terminal will be displayed.

The keywords for altering memory allocation are:
/MAIN

pname[:base:size:type]

/SUB

= mname:pname[:base:size]

/POOL

[=top]

These keywords are used as follows:

MAIN

- This keyword is used to establish a main partition.
pname

The 1- to 6-character
partition name.

alphanumeric

:base

Physical base
address
of
the
partition, specified as a number of
64-byte blocks.

;size

Size of the partition, specified as
a number of 64-byte blocks.

:type

TASK
for a user-controlled task
partition,
SYS
for
a
system-controlled task partition,
COM for a common partition, or DEV
for a common partition mapping into
the
device
registers.
User-controlled
and
system-controlled task partitions
are used for user tasks.
COM
partitions are used for resident
libraries and common blocks.
DEV
partitions
are
used
by
non-privileged tasks
to
access
device registers in the external
page, such as those for the UDCll
and ICS/ICR-11.

* These are the MCR Mount, Install, Dismount, and UFD commands
run as tasks), ann all system-supplied software.
5-17

(which

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Base and size may be entered in any one of the
following formats; the value nnnn is a number to
which a multiplication factor is applied
to
determine the value used.
Format

Calculated Value

nnnn
nnnn.
nnnK
nnn.K

Octal (nnnn*lOO)
Decimal (nnnn.*64.)
Octal K (nnn*4000)
Decimal K (nnn.*2048.)

Valid ranges of calculated values are:
O<=base<l 77777
O.<=base<65535
OK<=base<3 777K
O.K<=base<2047.K

Octal
Decimal
Octal K
Decimal K

O<size<2000
O.<size<l024.
OK<size<40K
O.K<size<32.K

For system-controlled partitions, partition
size may range from O<size<l024.K, or any
valid representation of these limits.
Example:
Any of the following inputs for
allocate a 2048. byte partition.

size

will

40
32.
lK
l.K

If base, size, and type are not specified,
VMR displays the values of these parameters
for the named partition.
If the
named
partition has subpartitions, the parameters
of all its subpartitions will
also
be
displayed.
NOMAIN

- pname
The NOMAIN keyword eliminates the named
from the system.

SUB

partition

- This keyword is used to establish a subpartition.
Subpartitions
may
not
be
established
in
system-controlled main partitions.
mname

The
1to
6-character
main
partition name.
The subpartition
being
defined
will
become
a
subpartition
in the named main
partition.

:pname

The 1- to 6-character subpartition
name.
The
subpartition
being
defined is a subpartition of the
main partition specified in mname.

5-18

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

:base

Physical base
address
of
the
subpartition, specified as a number
of 64-byte blocks.

:size

Size of the suboartition. soecified
as a number of G4-byte.bio~k~. -- -

Base and size may be entered in any one of the
following formats:
the value nnnn is a number to
which a multiplication factor is applied
to
determine the value used.
Format

Calculated Value

nnnn
nnnn.
nnnK
nnn.K

·Octal (nnnn*lOO}
Decimal (nnnn.*64.)
Octal K (nnn*4000)
Decimal K 1.-..-. ... *'lf\AO

\!lU!lo··"-V"%UoJ

Valid ranges of values are:
Octal
Decimal
Octal K
Decimal K

O<=base<l 77777
O.<=base<65535
OK<=base<3777K
O.K<=base<2047.K

O<size<2000
O.<size<l024.
OK<size<40K
O.K<size<32.K

Example:
Any of the following inputs for
allocate a 2048. byte partition.

size

will

40
32.
lK
l.K

If base and size are omitted, the current values
will be displayed for the named subpartition.
NOSUB

- =mname:pname
Eliminates the named subpartition from the system.

POOL

- This keyword is used to increase the size
dynamic storage region.
top

the

- First location in memory to be used for
user
partitions.
This is given in
number of 64-byte blocks.
It can be
given in octal or decimal. If top is
not given, the system will print out the
current top of memory, the size of the
longest block of pool space in words,
and the total number of words in the
pool.
The format is:
POOL = top:max:total
where top is given in units of
blocks.
Max and total are
number of words.

5-19

64-byte
decimal

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

The keyword for specifying write check on writes
is:

RKOS

disks

/WCHK[=dev:J
WCHK

Specifies the RKOS disk which is to have all
writes followed by a write check. This feature
provides high reliability for data transfers to
the RKOS disk. If the device parameter (= dev:)
is omitted, all RKOS disk drives with write check
enabled will be displayed.

NOWCHK -

Specifies that write check is to be disabled for
the specified device.
If the device parameter
(= dev:) is not specified, the RKOS disks which
have write check disabled will be displayed.

Command Error Messages:
VMR -- ALIGNMENT ERROR
An attempt has been made to create a partition, but
the base address or size conflicts with existing
partitions or physical memory size.
VMR -- DEVICE NOT DHll
An attempt has been made to set the baud rate for a
terminal which is not attached to a DHll multiplexer.
VMR -- DEVICE NOT TERMINAL
An
attempt
has
been
made
to
set
characteristics on a non-terminal device.

terminal

VMR -- NON EXISTENT MEMORY
An attempt has been made
non-existent memory.

define

partition

An attempt has been made to define a partition with
name that is already in use.

VMR -- PARTITION ALREADY EXISTS

VMR -- SPACE USED
An attempt has been made to create a partition
subpartition in a storage area already occupied.

VMR -- TASK INSTALLED IN PARTITION
An attempt has been made to
containing installed tasks.

eliminate

partition

VMR -- TOO MANY SUBPARTITIONS
A main partition is limited
subpartitions.

5-20

maximum

seven

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- UNKNOWN MAIN PARTITION
An attempt has been made to define a subpartition of a
non-existent main partition.
VMR -- WRITE CHECK NOT SUPPORTED
An attempt has been made to enable write check on a
device which is not an RK05. Write check is currently
supported only for RKOSs.
Notes:
1.

If a task is currently installed in a partition and an
attempt is made to eliminate the partition (/NOMAIN, /NOSUB),
the command will be rejected.

If a main partition is eliminated, all its subpartitions
also eliminated.

If ·a subpartition is being eliminated, only the specified
subpartition will be eliminated. A main partition that has
subpartitions with tasks installed will not be allowed to be
eliminated until those tasks are removed.

When defining a partition (main or sub), the name must not
already be defined as a partition.
In addition, main
partitions may not overlap any other main partition, nor may
subpartitions
of
a
main partition overlap any other
subpartitions of the main partition.

The numeric value convention applies to all the
numeric parameters.

All partitions except LDR are considered user partitions by
the Pool command.
Thus, SYSPAR is included as a user
partition.

Once space has been allocated to the dynamic storage
it can never be recovered for use in partitions.

System-controlled partitions may not have
subpartitions
explicitly defined by the Set command. The Executive defines
subpartitions of the system-controlled partition dynamically,
as needed, for tasks installed in the system-controlled
partition.

Examples:
Example 1:
VMR>SET /BUF=LPO:
BUF=LP0:132.
Display the default buffer size of LPO:.

5-21

Set

are

command

region,

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Example 2:
VMR>SET /BUF=TT1:40.
Set TTl:

buffer size to 40(10).

Example 3:
VMR>SET /LA30S=TT2:
VMR>SET /LA30S
LA30S=TTO:
LA30S=TT2:
Set TT2: as an LA30S device.
Display all LA30Ss.
Example 4:
VMR>SET /NOLA30S=TTO:
Set TTO: to other than an LA30S.
Example 5:
VMR>SET /PRIV=TTO:
Set TTO: to a privileged terminal.
Example 6:
VMR>SET /PRIV
PRIV=TTO:
PRIV=TTl:
Display all privileged terminals.
Example 7:
VMR>SET /SLAVE=TT3:
Set TT3:

to a slave terminal-

Example 8:
VMR>SET /VT05B
VT05B=TT4:
VT05B=TT5:
VT05B=TT6:
Display all terminals which are VT05Bs.
Example 9:
VMR>SET /MAIN=SYSPAR:420:140:TASK
Define a main partition, called SYSPAR, as a user-controlled
partition whose base address is 42000(8), and whose length
is 14000(8) bytes.

5-22

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

Example 10:
VMR>SET /SUB=SYSPAR:MCRPAR:445:56
Define a subpartition, called MCRPAR, of main partition
SYSPAR whose base address is 44500{8), and whose length is
5600(8) bytes.
Example 11:
VMR>SET /MAIN=SYSPAR
MAIN=SYSPAR:0420:0140:TASK
SUB=SYSPAR:MCRPAR:0445:0056
Display the parameters of the main partition SYSPAR and
of its subpartitions.

all

Example 12:
VMR>SET /NOSUB=SYSPAR:MCRPAR
Eliminate subpartition MCRPAR.
Example 13:
VMR>SET /NOMAIN=SYSPAR
Eliminate main partition SYSPAR.
Example 14:
VMR>SET /MAIN=GEN:l000:6000:SYS
Define a main partition, called GEN, as a system-controlled
partition whose base address is 100000(8), and whose length
is 600000(8) bytes.
Example 15:
VMR>SET /POOL=420

Set the beginning of user partitions to 42000.
Example 16:
VMR>SET /POOL
POOL=0420:150.:420.
The first location for
user partitions is 42000;
the
longest free block is 150. words; and the total of all pool
space is 420. words.
Example 17:
VMR>SET /WCHK=DKl:
Enable write check on DKl:.

5-23

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

TIME

TIM

The Time command makes it possible to:
Set the current time of day in the system image;
cur~ent

Set the

date in the system image, or

Display the time and date stored in the system image.
Format:
TIM[E]

[hrs:mins[:secs]]

[mnth/day/year]

where:
hrs
mins

=Hours (range 0-23);
Minutes (range 0-59);

secs = Seconds (range 0-59)

(optional);

mnth =Month (range 1-12);
day
year

= Day (range 1-31), and
Year (range 0-99).
Year is relative to 1900;
1974 is entered as 74.

thus,

Examples:
Example 1:
VMR>TIM
10:23:31 03-May-75
Display the time and date stored in the system image.
Example 2:
VMR>TIM 1:30 4/7/75
In the system image, set the time to 01:30:00 and the date to
07-May-75.
Notes:
1.

If neither time nor date is specified, the time and date will
be displayed on the entering terminal.

If the time and date are specified, the clock and calendar
will be set;
if only the time is specified, the clock will
be set;
if only the date is specified, the calendar will be
set.

The order of time and date on the command is not significant.
Either order will be accepted.

All numeric quantities are decimal.
necessary (or permitt~d).

5-24

No terminating period is

VIRTUAL MONITOR CONSOLE ROUTINE

5.4.3

(VMR)

Informational Commands

DEVICES

DEV

The Devices command displays, on the entering terminal, the
names of all device-units specified in the system image.

symbolic

Format:
DEV [ICES]

Example:
VMR>DEV
DKO:

DKl:
DFO:
DTl:
DT2:

DT 3:

OFF LINE

LPO:
TTO:
TTl:

TIO:
COO:

TTO:

CLO:
SYO:

LPO:
DKO:

Notes:
1.

A device name in the second column is the device to which the
corresponding device in the first column has been redirected.

OFFLINE is an optional message which indicates that, although
the system tables contain entries for this device, the host
configuration on which the system was last saved did not
contain the related device.

5-25

VIRTUAL MONITOR CONSOLE ROUTINE
LOGICAL UNIT NUMBERS

(VMR)
LUN

This command is used to determine which physical devices a task has
statically
assigned.
A
list
of
physical device units and
corresponding logical unit numbers is displayed on the entering
terminal. This list indicates which devices were assigned to the task
specified in the LUNs command.
Format:
LUN[S] taskname
Example:
VMR>LUN XKE
TIO:

SYO:
CLO:

TT3:

2
3
4

TT3:

DKO:
TT3:

6
7

The

display

shows
and
assigned to task XKE.
1 , 2, 3 , 4, 5, 6 ,

static
7;
no

assignments
for
LUNs
other LUNs are statically

Command Error Messages:
VMR -- NO LUNS
The task which is the argument of the LUNs command
does not have any logical units. This is not an error
message, but rather, an indication that there are no
assignments to display.

5-26

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

PARTITIONS

PAR

The Partition Definitions command displays, on the entering
a description of each memory partition in the system image.

terminal,

The following is displayed:
1.

Partition name;

Partition base address (octal);

Partition size (octal);

Partition kind, main partition or subpartition, and

Partition type, TASK (user-controlled), COM (common),
(device registers), or SYS (system-controlled).

DEV

Format:
PAR [TIT IONS]
Example:
VMR>PAR
LOR
023544
SYS PAR 030100
TKNPAR 030100
GEN
040000
COM PAR 140000
DE VP AR 160000

001000
007700
000400
100000
020000
020000

MAIN
MAIN
SUB
MAIN
MAIN
MAIN

TASK
TASK
TASK
SYS
COM
DEV

Notes:
1.

The display contains the name, base, size, kind, and type
the partition in columns 1 through 5, respectively.

Partitions of type TASK are user-controlled task partitions.
COM partitions are used for reentrant libraries and common
data areas. DEV partitions are used to allow tasks to
communicate with specific device registers, such as the UDCll
and ICS/ICR-11 industrial control subsystems. SYS partitions
are system-controlled task partitions.

5-27

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

TASK LIST

TAS

The Task List
description of
contains:

terminal,
a
command displays, on the entering
each task installed in the system image. The display

Task name;

Task version identification;

Partition name;

Task priority;

Size of task in bytes (octal);

Load device identification;

Disk address logical block number (octal), and

Task memory state.

Format:
TAS [KLIST]
ExamEle:
VMR>TAS
LOADER
. TKTN.
... MCR
... MOU

OOOlM PARl
OOOlM GEN
OOOlM MCR
OOOlM PAR3K

248.
249 .
221.
221.

010600
002200
002200
001400

DKO:
DKO:
DKO:
DKO:

-00000010572
-00000011112
-00000010511
-00000011255

FIXED

Notes:
1.

The display contains, in columnar form, from left to right:
task name, task version identification, partition name, task
priority, task size, load device identification, logical
block number on the load device, and task memory state.

FIXED indicates that the task is fixed in the system image.

5-28

VIRTUAL MONITOR CONSOLE ROUTINE
5.4.4

(VMR)

Task Control Commands

ALTER

ALT

This command provides the capability to
installed task.

change

the

priority

Format:
ALT[ER]

taskname /keyword=value

Keyword:
/PRI

priority

where:
taskname - name of task whose priority is to be altered.
priority - new priority to
task.

assigned

the

specified

Example:
VMR>ALT TEST /PRI=250
Alter the priority of task TEST to 250(8).

Notes:
1.

Priority must be less than or equal to 250(10)
than 0. The numeric conventions apply.

5-29

and

greater

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

CANCEL

CAN

The Cancel command provides the capability to cancel the time-based
initiation requests for a task.
These requests result from a Run
directive, or any of the time-synchronized variations of the VMR Run
command.

Format:
CAN[CEL] taskname

Example:
VMR>CAN XKE
Cancel all periodic rescheduling and
requests for task XKE.

5-30

time-based

initiation

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

FIX-IN-MEMORY

FIX

The Fix-in-Memory command provides the capability to fix a task in its
partition.
The task image is copied into the system image, and its
partition is marked busy.
Format:
FIX taskname
Example:
VMR>FIX XKE
FIX task XKE in its partition.
Command Error Messages:
VMR -- TASK ALREADY FIXED
The task used as the argument of the
already fixed in the system image.

Fix

command

VMR -- TASK CHECKPOINTABLE
A checkpointable task cannot be fixed.
Notes:
1.

Fixed tasks remain physically in the system image even after
they exit; thus, they need not be reloaded when requested to
run. Their ability to respond to realtime dem&nds is,
therefore, increased.
Only an Unfix command can place the
task in a state which will permit subsequent freeing of the
occupied space in the system image.

Tasks may not be fixed in system-controlled partitions.

Checkpointable tasks are not fixable.

A task must be installed before it can be fixed.

5-31

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

REASSIGN

REA

The Reassign command provides the capability to reassign a task's
logical unit numbers (LUNs) from one physical device-unit to another.
The reassignments are made in the task image.
Format:
REA[SSIGN] taskname lun nud:
where:
lun
nud

logical unit number
new device-unit

Examples:
Example 1:
VMR>REA JOE 3 TTO:
Reassign LUN 3 of task JOE to device TTO:.
Example 2:
VMR>REA BILL 2 TTl:
Reassign LUN 2 of task BILL to device TTl:.
Command Error Messages:
VMR -- LUN OUT OF RANGE
An attempt has been made to reassign a logical unit
number which is greater than the maximum number of
logical unit numbers allocated during task build.
Notes:
1.

The Reassign command performs re-assignments in the indicated
task's image.

The Reassign command does not affect a system image resident
task;
hence, Reassign has no effect on the assignments of a
task that is fixed in the system image. The Redirect command
is used for dynamic device reconfiguration.

5-32

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

REDIRECT

RED

The Redirect command provides the capability to redirect all I/O
requests directed to a physical device-unit from that unit to another.
Format:
RED[IRECT] nud:=old:
where:
nud:

new device-unit;

this is the redirected-to device.

old:

old device-unit;

this is the redirected-from device.

Example 1:
VMR>RED TT3:=TT6:
Redirect all I/O requests for device TT6:

to device TT3:.

Example 2:
VMR>RED TT:=LP:
Redirect all I/O requests for device LPO:

to device TTO:.

Command Error Messages:
VMR -- CIRCULAR REDIRECT ERROR
An attempt has been made to
that a circular list of
result.

redirect a device such
redirected devices would

VMR -- DEVICE NOT REDIRECTABLE
The specified old device may not be redirected.
VMR -- DEVICE ddnn:

NOT IN SYSTEM

New device in Redirect command is not in
(does not exist in the device tables).

the

system

VMR -- OLD DEVICE ATTACHED
An attempt has
device.

been

made

redirect

attached

pseudo

device

VMR -- PSEUDO DEVICE REDIRECT ERROR
An attempt has been made to redirect a
to another pseudo device.

5-33

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- TI REDIRECT ERROR

The pseudo device TI:

may not be redirected.

Notes:
1.

The device TI:

may not be redirected.

A pseudo device may
device.

An attached device may not be redirected.

not

5-34

redirected

another

pseudo

VIRTUAL MONITOR CONSOLE ROUTINE
REMOVE

(VMR)
REM

The Remove command provides the capability to delete a task name from
the system task directory (STD), actually removing the task from the
system image.
Format:
REM[OVE] taskname
Example:
VMR>REM SCAN
Remove task SCAN from the system image.
Notes:
1.

The process of removing a task from the system makes a task
unknown to the system. ·rt is the complement of the Install
command.

If a task that is the object of a Remove command is fixed, it
is automatically unfixed and then removed.

5-35

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

RUN

The Run command initiates the execution of a task.
may be:

With Run,

task

Started immediately at bootstrap time;
Started a time increment from the system image time;
Started a time increment from clock unit synchronization, or
Started at an absolute time of day relative to the
time.

system

image

All the Run commands modify the system. image in such a way that the
desired action will occur when the system image is subsequently
bootstrapped.
All these option are available with or without optional rescheduling.
The Run command has four format variations, as described below.
Formats:
1.

Run immediately at bootstrap time.
RUN taskname [/Keyword] [/Keyword]
Keywords:
/RSI

nnnnu

/UIC

[group,member]

where:
taskname

1- to 6-character taskname.

RSI

The reschedule interval. The format is the
same as for dtime detailed below under Format
2. The reschedule interval specifies how
often the task is to be rerun. Thus, each
time the specified time interval elapses, an
initiation request is made for the specified
task.
Default:

UIC

no rescheduling.

Keyword for the UIC parameter.
Legal group
and member numbers range from 1 through
377(8). The square brackets are required
syntax. This is the UIC under which the task
will be requested. The UIC determines which
files the task may access.
Default:
time.

the value established at task-build

5-36

VIRTUAL MONITOR CONSOLE ROUTINE
2.

(VMR)

Run a time increment from system image time.
RUN taskname dtime [/Keyword] [/Keyword]
Keywords:
/RSI

nnnnu

/UIC

[group,member]

where:
taskname
dtime

1- to 6-character taskname.
= The time at which the task will be

initiated
in time units from system image time. It is
of the form nnnnu:

A time parameter consists of:
a.

A magnitude field, and

A units field.

where:
nnnn

the magnitude
units, with:

T for
S for
M for
H for

ticks,
seconds,
minutes, and
hours.

The legal value of the magnitude is related
to the value of the units field, which is
encoded as:
T = Ticks. A tick is a clock interrupt,
and the rate at which interrupts
occur depends on the type of clock
installed in the system.
For a line-frequency clock, the tick
rate is either 50 or 60 per second,
corresponding to the line frequency.
For a programmable clock, a maximum
of
1000
ticks
per
second is
available.
(Frequency is selectable
at system generation.)
S

Seconds.

Minutes.

Hours.

5-37

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

The magnitude is the number of units to be
clocked,
but the magnitude value cannot
exceed 24 hours in the specified units.

Units = T. Any positive value
(maximum of 15 bits).

valid

Units = S. Any positive value
(maximum of 15 bits).

valid

Units

1440.

Units

24.

RSI

See Format 1 above.

UIC

See Format 1 above.

is maximum magnitude.
is maximum magnitude.

Run a time increment from clock unit synchronization.
The task start time is determined by first waiting for the
system image's next time unit occurrence (hour, minute,
second, tick), and then waiting for the specified time
increment to elapse.
Run taskname sync [dtime] [/Keyword] [/Keyword]
Keywords:
/RSI = nnnnu
/UIC

[group,member]

where:
taskname = 1- to 6-character taskname.

sync

H =synchronize on
M = synchronize on
S = synchronize on
T
synchronize on

dtime

As defined in Format 2 above;
dtime, if
present, is added to the synchronization unit
to produce the actual run time for running
the task.

RSI

= See Format 1 above.

UIC

= See Format 1 above.

the next hour,
the next minute,
the next second,
the next tick.

Run at an absolute time of day relative to the
time.
RUN taskname atime [/Keyword] [/Keyword]
Keywords:
/RSI = nnnnu
/UIC = [group,rnernber]

5-38

system

image

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

where:
taskname

1- to 6-character taskname.

a time

Absolute time of day when this task will be
initiated relative to the system image time.
Input format is hh:mm:ss where:
hh
mm
SS

hours (decimal assumed) ,
minutes (decimal assumed) ,
seconds (decimal assumed).

RSI

See Format 1 above.

UIC

See Format 1 above.

Examples:
VMR>RUN XKE 15M
RUN task XKE 15 minutes from system image time.
VMR>RUN XKE 15M/RSI=90S/UIC=[3,1J
RUN task XKE 15 minutes from system image time, rescheduling
it every 90 seconds with a user identification code of
r 1 _, ,
L -

T ..... J

•

Command Error Messages:
VMR -- INVALID TIME PARAMETER
A time field is incorrect.
Notes:
1.

The Cancel command may be used to cancel all initiation
requests for a task, including those to run immediately when
the system is bootstrapped.

When a Run command is issued for an absolute time of day, the
absolute time is converted to a time increment from the
current time stored in the system image.
Thus,
if the
current time in the system image is 8:00 and a Run command is
issued with an absolute time of 9:00 specified, the task will
run one hour from the time the system is bootstrapped,
regardless of whether the time is reset.

All time parameters are converted
to
double-precision
magnitude, in ticks. The number of ticks per second is caKen
from the system image. This value will be the one in use the
last time the system was saved, or the value specified during
SYSGEN phase 2 for the virgin system image. Note that it the
system is bootstrapped on a machine with a clock different
from the one indicated in the system image
(i.e., with a
programmable clock in lieu of a line clock, or vice versa),
the ticks will be measured on the machine's clock; thus. the
effective time will be different from that specified to VMR.

5-39

VIRTUAL MONITOR CONSOLE ROUTINE

{VMR}

UNFIX

UNF

The Unfix command provides the capability to free a fixed task
the system image. Unfix is the complement of the Fix command.

from

Format:
UNF[IA] taskname
Example:
VMR>UNFIX XKE
Unfix task XKE, freeing the partition in which it resides.
Command Error Messages:
VMR -- TASK NOT FIXED
An attempt has been made to unfix a task which is
fixed.

not

Notes:
1.

If a task exits or aborts, but is fixed, it will still remain
in control of the system image memory in which it resides.

5-40

VIRTUAL MONITOR CONSOLE ROUTINE
5.4.5

(VMR)

System Maintenance Command

L SAV

SAVE

The VMR Save command is used to write the image of a system onto a
sequential medium in bootstrappable format. This command is intended
to provide the user with a facility to save an RSX-llS or foreign
system (see /IM switch) onto a load medium that can be transported to
another system and bootstrapped using a hardware bootstrap or the /BO
switch in PRESRV.
System images created with the Save command on all media, except paper
tape, are in a format compatible with the logical tape format in
PRESRV.
Thus, any of PRESRV's logical tape operations may be
performed on system images created with the Save command. System
images created by the Save command always start at the beginning of
the medium.
On paper tape, system images are created in Absolute Loader format and
must be bootstrapped via the Absolute Loader.
Format:
SAV[E] ddnn: [label]
where:
ddnn:

Specifies the device-unit onto which the
is to be written. Valid devices are:
CT:
DT:
DX:
MM:
MT:
PP:

label

system

image

Tape cassette,
DECtape,
Flexible diskette,
TU16 magtape,
TU10/TS03 magtape, and
Paper tape.

The file label name must be specified for all secondary
media except paper tapew It consists of from one to
twelve alphanumeric characters.

Example:
VMR>SAV MT:MYSYS
In the example above, the system image is written onto
in bootstrappable format.
System changes made by
Redirect or other VMR commands are also saved with
system image.

MT:
the
the

Command Error Messages:
VMR -- DEVICE ddnn: NOT IN SYSTEM
The specified
system.

device

5-41

does

not

exist

the

host

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- ILLEGAL DEVICE VOLUME
The specified device was other than one of
devices specified above.

the

valid

output

device

VMR -- I/O ERROR ON OUTPUT DEVICE
An I/O error was encountered on
while writing the system image.

the

Notes:
1.

When the /LI command is used in PRESRV to list a medium
(other than paper tape)
containing a system image created
with the Save command, each system image will cause a line to
be printed which contains the following:
VM (for VMR),
indicated as the source device, the label specified in the
SAV command line, a block size of 512, the number of 512-byte
blocks in the system image (excluding boot and label blocks),
and the date and time that the system was saved.

When RSX-115 systems are saved to paper tape (without the /IM
switch specified at VMR initiation time), the last 1000(8)
bytes of the system image are not written out.
This is to
allow space to accommodate the Absolute Loader. When the /IM
switch has been specified with the system's file name, the
entire system image is written out to paper tape. In this
case, it is the user's responsibility to insure that the
system image is not so large that it will overlay the
Absolute Loader.

When bootstrapping systems on media other than paper tape,
regardless of the size oi the system image, the last 1000(8)
bytes of physical memory will not be loaded to accommodate
the bootstrap.

Systems saved on a TU10/TS03 9-track, magtape
booted from a TU16 drive, and vice-versa.

When the /IM switch has been specified with the system file
specifier, the Save command will always use a bootstrap which
does not enable memory management.

The maximum memory size of a mapped system is 124K words.
For physical and practical limitations, the Save command
imposes maximum sizes on saved images in mapped systems for
certain devices.
If a mapped system is saved, whose memory
size is greater than the maximum size imposed by the Save
command for the specified device, only the amount of memory
equal to the maximum size {starting at physical location 0)
is saved: no diagnostic message is issued. The devices for
which the Save command imposes a maximum size on saved system
images are:
Paper tape punch (PP) - 28K words,
Flexible disk (DX) - 123K words, and
Cassette tape (CT) - 28K words.

5-42

drive

can

VIRTUAL MONITOR CONSOLE ROUTINE
7.

5.5

(VMR)

None of the bootstraps require that the target RSX-llS system
have a driver for the bootstrap device.

VMR ERROR MESSAGE SUMMARY

All error messages
alphabetical order.

produced

VMR

appear

this

section

VMR -- ALIGNMENT ERROR
An attempt has been made to create a partition, but the
base address or size conflicts with existing partitions or
physical memory size.
VMR -- BASE ADDRESS MUST BE ON 4K BOUNDARY
The base address of the task is not
Applicable only to mapped systems.

boundary.

VMR -- BASE MISMATCH COMMON BLOCK <common-name>
The base address of the partition does not match that
the common block. Applicable only to unmapped systems.

VMR -- CHECKPOINT AREA TOO SMALL
The area allocated for checkpointing the task is smaller
than the partition into which it is being installed.
VMR -- CIRCULAR REDIRECT ERROR
An attempt has been made to redirect a device such
circular list of redirected devices would result.

that

VMR -- COMMAND I/O ERROR
An I/O error was detected during
file.

read

indirect

VMR -- COMMAND SYNTAX ERROR
command line
The command line printed has incorrect syntax.
VMR -- COMMON BLOCK IS TASK PARTITION <common-name>
A task's request for access to a common block is rejected
because the partition requested is a task partition.
VMR -- COMMON BLOCK NOT LOADED <common-name>
The specified common block is linked to the task, but it
has not been loaded into the system image. Install the
specified common block, then install the task.

5-43

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- COMMON BLOCK OCCUPIED
An attempt has been made to load a
already occupied.

common

block

that

those

the

The device specified in the command was not generated
the system image.

into

VMR -- COMMON BLOCK PARAMETER MISMATCH <common-name>
Parameters of a common block do
task's label block.
VMR -- DEVICE ddnn:

not

match

NOT IN SYSTEM

VMR -- DEVICE NOT DHll
An attempt has been made to set the baud rate for
terminal which is not attached to a DHll multiplexer.

VMR -- DEVICE NOT MOUNTED
The device specified
mounted.

the

file

specification

not

VMR -- DEVICE NOT REDIRECTABLE
The specified device may not be redirected.
VMR -- DEVICE NOT TERMINAL
An attempt has been made to set terminal
for a device which is not a terminal.

characteristics

VMR -- EXECUTIVE SIZE GREATER THAN 16K
The system image indicates that the Executive
than 16K words. The system is unusable.

larger

VMR -- FILE filename HAS ILLEGAL FORMAT
The specified symbol table file has an invalid
an RSX-llM/RSX-llS symbol table file.

format

for

References to the the indicated system image file, via
symbol table file, have retrieved inconsistent data.

the

VMR -- FILE filename HAS INCONSISTENT DATA

VMR -- FILE NOT CONTIGUOUS
An attempt has been made to install a task from
non-contiguous file. Task images must be contiguous.

VMR -- FILE NOT FOUND
The requested file is not in the directory on the specified
volume.

5-44

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- FILE NOT TASK IMAGE
Data in the label block is not correct, indicating that the
file is not a valid task image.
VMR -- ILLEGAL DEVICE/VOLUME
The device specified is not a valid task residence device.
VMR -- ILLEGAL ERROR-SEVERITY CODE code
This message indicates an internal failure in VMR.
If this
error
message
persists,
contact a Digital software
representative.
VMR -- ILLEGAL FILE SPECIFICATION
f ilespec
The indicated file specification has an illegal
contains wild cards.

format

VMR -- ILLEGAL FIRST APR
The Task Builder allows the user
to specify the virtual
base address of a mapped task image as a multiple of 4K
words.
Privileged tasks normally start at 100000(8)
in
order
to allow them to be mapped into the first 16K of the
Executive, but they may also start at 120000(8)
or
140000(8). Non-privileged tasks must start at O.
VMR -- ILLEGAL FUNCTION
A command has been entered which VMR does not recognize.
VMR -- ILLEGAL GET COMMAND LINE ERROR CODE
This error indicates a system failure.
If this message
persists, contact a Digital software representative.
VMR -- ILLEGAL KEYWORD VALUE
A value entered for a keyword exceeds its limits.
VMR ---ILLEGAL PRIORITY
The value of the priority in the command is out
i.e., not 1 to 250(10), inclusive.

range,

VMR -- ILLEGAL SWITCH
/switch
The indicated switch, in the context of
entered, is illegal.

5-45

the

command

line

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- INDIRECT COMMAND SYNTAX ERROR
command line
The indicated command line was
indirect file specifier.

entered

with

improper

VMR -- INDIRECT FILE DEPTH EXCEEDED
command line
The indicated command line contains more than two levels of
indirect files.
VMR -- INDIRECT FILE OPEN FAILURE
command line
The indirect command file in
could not be opened.

the

indicated

command

line

VMR -- INVALID KEYWORD
A keyword has been encountered which
context of the specified command.

illegal

the

VMR -- INVALID TIME PARAMETER
A time field is incorrect.
VMR -- INVALID UIC
A value of zero for either the group number or
number was detected.

the

member

VMR -- I/O ERROR ON INPUT FILE
filename
An input I/O error occurred on the indicated file.
VMR -- I/O ERROR ON OUTPUT DEVICE
An output
system.

I/O

error

was

encountered

while

saving

the

VMR -- I/O ERROR ON OUTPUT FILE filename
An output I/O error occurred on the indicated file.
VMR -- LENGTH MISMATCH COMMON BLOCK <common-name>
The length parameter for the common block, as described in
the label block for the task image, is different from the
corresponding length parameter defined in the system image.
A task's label block data must match system data for that
task before it can be installed.

5-46

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR}

VMR -- LOGICAL DEVICE NOT IN SYSTEM
The specified logical device is not defined in the
device tables.

logical

VMR -- LUN OUT OF RANGE
An attempt has been made to re-assign a LUN which is
greater than the maximum number of LUNs allocated during
task build.
VMR ..:,._ NO LUNS
The task specified as the argument of the LUNs command does
not have any logical units assigned. This is not an error
message; rather, it is an indication that there are no
assignments to display.
VMR -- NO POOL SPACE
A request for dynamic memory in the
cannot be satisfied.

system

image

VMR

VMR -- NO ROOM AVAILABLE IN STD FOR NEW TASK
This message means that there is no dynamic storage
available to make an entry in the system task directory
{STD}; therefore, no task can be installed.
VMR -- NON EXISTENT MEMORY
An attempt has been
non-existent memory.

made

define

partition

Because of the virtual base address of the task, there
not enough APRs left to map the entire task image.

are

VMR -- NOT ENOUGH APRS FOR TASK IMAGE

VMR -- OLD DEVICE ATTACHED
An attempt
device-unit.

been

made

redirect

attached

VMR -- OPEN FAILURE ON FILE filename
The indicated file could not be opened.
VMR -- PARTITION ALREADY EXISTS
An attempt has been made to define a partition with a
that is already in use.

name

VMR -- PARTITION NOT COMMON
A partition specified for a common library is
as a common partition.

5-47

not. defined

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMP -- PARTITION NOT IN SYSTEM
The partition name specified as an argument
cannot be found in the system.

command

VMR -- PRIVILEGED TASK LARGER THAN 12K
Privileged tasks have a maximum size of 12K.
VMR -- PSEUDO DEVICE ASSIGNMENT ERROR
A logical device name may
device.

not

assigned

pseudo

VMR -- PSEUDO DEVICE REDIRECT ERROR
An attempt has been made to redirect
another pseudo device.

pseudo

device

An attempt has been made to create a partition or
subpartition in a main storage area already occupied.

VMR -- SPACE USED

VMR -- SPECIFIED PARTITION FOR COMMON BLOCK
An attempt is being made to install
block.

task

common

VMR -- SPECIFIED PARTITION TOO SMALL
The task being installed is larger than the partition
which it is being installed.

into

VMR -- SYNTAX ERROR
This message generally indicates
that
the
required
information has been entered incorrectly. The user can
usually correct this condition by retyping the arguments
according to the command specifications.
VMR -- TASK ACTIVE
The task used as the argument of the command is active.
VMR -- TASK ALREADY FIXED
The task used as the argument of the Fix command is already
in the system image.
VMR -- TASK AND ~ARTITION BASES MISMATCH
The base of the partition does not match that of the task
being installed. Applicable only to unmapped systems.
VMR -- TASK IS CHECKPOINTABLE
A checkpointable task cannot be fixed in the system image.

5-48

VIRTUAL MONITOR CONSOLE ROUTINE

(VMR)

VMR -- TASK IMAGE I/O ERROR
The Install command cannot read the task image file.
VMR

TASK IMAGE CURRENTLY INSTALLED
The requested task image is already
checkpointable.

installed

and

VMR -- TASKS INSTALLED IN PARTITION
An attempt has been made
containing installed tasks.

eliminate

partition

VMR -- TASK NAME ALREADY IN USE
An attempt has been made to install a task having the
name as one

same

VMR -- TASK NOT FIXED
An attempt has been made to
fixed.

unfix

task

which

not

VMR -- TASK NOT IN SYSTEM
The referenced task has not been installed.
VMR -- TI REDIRECT ERROR
The psuedo device TI:

may not be redirected.

VMR -- TOO MANY COMMON BLOCK REQUESTS
A task is limited to three common blocks.
VMR -- TOO MANY LUNS
A task has requested more than 255(10) LUNs to be assigned.
VMR -- UNDEFINED COMMON BLOCK <common-name>
A task references a common block that is not defined in the
system image.
Usually this indicates that the task was
built to run in another system.
VMR -- UNKNOWN MAIN PARTITION
An attempt has been made to define
non-existent main partition.

subpartition

VMR -- WRITE CHECK NOT SUPPORTED FOR DEVICE
Write check is only supported for RKOS disks.

5-49

APPENDIX A
PRES RV

A.l

INTRODUCTION TO PRESRV

The Volume Preservation Program (PRESRV) is a complete, stand-alone
system with an integrated Executive. Its principal function is to
create copies of disk and magnetic tape volumes. The copy capability
is the basis for the _backup and restoration procedures used for
preserving images of the RSX-llM distribution kit disks. Both backup
and restoration of kit disks are copy operations carried out by
PRESRV.
PRESRV makes copies in three formats:

Files-11 1

Logical tape, and

Image mode.

Files-11 format copies are those in which all Files-11 information is
copied to the backup medium. The Files-11 information indicates which
blocks are allocated to files and includes the files themselves.
Blo~ks not allocated to files are not copied.
Logical tape format copies are those in which the medium assumes the
characteristics of a sequential medium (e.g., magnetic tape}. Thus, a
disk can be copied into logical tape format, and the structure on the
disk assumes sequential access of the data. Also, a Files-11 volume
can be copied to another disk in logical tape format.
The resulting
copy can then be treated in subsequent PRESRV operations as a logical
tape.
A logical tape consists of a boot block, a label block, a stand-alone
RSX-llM system image (first volume only), followed by the preserved
data. Bootstrapping the first volume of a logical tape will activate
PRESRV, so that the contents of the tape may be restored to the
original device type. When a tape is produced by PRESRV, be it
physical or logical, the user may specify a label of up to twelve
characters in length for the tape to be produced.
Image mode copies non-Files-11 disk volumes to other disks or to tape.
In this mode, non Files-11 volumes may be saved and restored. Image
mode copies all physical blocks on a device to the backup volume, as
opposed to Files-11 mode, which copies only those blocks that are in
use. Thus, Files-11 mode assumes knowledge of the disk structure,
while image mode does not.

A-1

PRES RV
PRESRV supports both multi-file volumes and multi-volume files.
Multiple
disk
representations on magtape are separated by an
end-of-file mark (tape mark). For output to magtape only, it is
assumed that the volume is correctly positioned at the load point,
initially, or after a tape mark, subsequently.
After each file is
written, two tape marks are written, and the magtape volume is then
positioned between the two tape marks for possible further processing.
NOTE
PRESRV determines the position of the
tape to decide whether or not to write
the boot block when copying to a tape.
If the above tape is positioned at BOT,
a boot block is written. If the tape is
not at BOT (e.g., one file is copied to
tape already), a boot block is not
written.
If the tape drive is off-line
when this check is made, the tape looks
like it is past BOT and, therefore,
PRESRV decides not to write a boot
block.
When PRESRV actually write~ to
tape, the driver detemines if the tape
is off-line and issues a message. When
the tape is put on-line, no boot block
is written.
Therefore, to be sure to get a boot
block on the tape, the tape must be
on-line and positioned at BOT before the
copy operation is started.
On input from magtape, there are two possibilities:
1.

If a file label is specified, the input volume
and a search is made for the named file.

rewound,

If no file label is specified, the next file on the tape is
taken as input.
It is assumed that the tape is properly
positioned (at the load point for the first file).

A file may extend across a volume boundary, but no new files may begin
on the second or subsequent volume of a multi-volume file.

A.2

PRESRV SERVICES

PRESRV provides the following services:
1.

Make image copies of volumes.

Make copies of volumes in Files-11 format.
Only the used
blocks of the Files-11 volume are copied, either to another
Files-11 volume, or to a logical tape. The logical tape may
be multi-volume.

Make copies of volumes formatted as logical tapes.

List labels from volumes formatted as logical tapes.

Format RKOS, RP02, RP03, and RP04 disks.

A-2

PRES RV
PRESRV acts on user requests which are in two forros:
1.

File command strings, and

Commands.

A.2.1

PRESRV File Command Strings

Any component of RSX-llM which needs to refer to files does so via
standard file command string having the following general format:

outflespcl.,,,outflespcn = inflespcl, ... inflespcn
PRESRV restricts this general form to one input and one output file
specifier. Thus, the PRESRV file command string has the general form:
outflespc=inflespc
Outflespc is an output file specifier, and inflespc is an input file
specifier.
In no case can the total length of the command string
exceed the maximum line length (80 bytes).
Each file specifier
format:

(whether

input

output)

has

the

following

dev: [g,m]filename.type;version/switch .•• /switchn
where:
dev:

The physical device on which the volume containing the
desired file is mounted, for example, DKO: or DTl:.
The name consists of two ASCII characters, followed by
an optional 1-digit (octal) unit number and a colon.
If no unit number is specified, unit 0 is assumed.
PRESRV is generated with units 0 and 1 for each type
of device.

[g,m]

Not used by PRESRV (must be null) .

filename

A label for a file written to a logical tape.
..1.!l
RSX-llM,
this field is used to name a file. A
filename can be up to nine alphanumeric characters in
length;
however, PRESRV allows file names of up to
twelve characters in length.
T-

PRESRV, however, does not use the filename as an
identifier for searching a directory to effect a name
mapping. The filename is used by PRESRV as a label
for a file on a logical tape. The label is then used
to extract files on multi-file volumes. PRESRV never
operates
on Files-11 files;
PRESRV only copies
volumes. The filename field enables the user to label
the copied volume.
type

Not used by PRESRV (must be null).

version

Not used by PRESRV (must be null).

A-3

PRES RV
/switch

A 2-character ASCII name identifying the
option.
The switch identifier may be followed
number of values.
In PRESRV, only the block
accepts
a value.
No PRESRV switch uses
negation.

switch
by any
switch
switch

Since PRESRV does not use the [group,member] designator, or
the type
or version component of the filename, the PRESRV filename c0mmand
string has the general format:
dev:filename/switchl .•. /switchn
The filename is used only to identify files on a logical tape.

A.2.2

PRESRV Copy Switches

The file specifier and any applied switches specify actions PRESRV eis
to initiate in its copy operation. Switches are divided into three
categories.
1.

Device Specific:
a.

Set 200-BPI mode (/20);

Set 556-BPI mode (/55);

Set 800-BPI mode (/80);

Set 1600-BPI mode, phase-encoded magtape (/PE), and

·e.

Format disk (/FO).

Administrative:
a.

Abort on hard I/O error (/ER);

Rewind at completion (/RW), and

Verify (/VE).

Formatting:
a.

Blocking (/BL:nn);

Files-11 format (/FI);

Logical tape format (/TP), and

Image mode (/IM).

The details of these switches are covered in section A.3.

A.2.3

PRESRV Command Switches

PRESRV
command
switches
are
divided
into
two
categories;
hardware-specific switches, which may all appear on one line, and
stand-alone switches, which are entered one at a time. Both types are

A-4

PRES RV
acted upon by PRESRV immediately.

command

until

completes

PRESRV

will

not

the current command.

another

Hardware-specific

switches are:

Set the address of the control status register for

device

TU16

tape

(/CSR);
2.

Set the physical unit number for a device (/UNIT);

Set the interrupt vector for a device (/VEC), and

Set the unit number of the TM02 formatter for the
drive (/TM02) •

Stand-alone switches are:
1.

Obtain a summary of PRESRV's services (/HE);
u-1 ........

UC~.t:-'

also called the

_,...,._.......,_ ...... ..:i.

\_;VllllllQUU j

Obtain a listing of all file labels on a logical tape (/LI);

Set the time and date (/TIM), and

Bootstrap a device (/BO).

Refer to section A.4 for details.

A.3

PRESRV FILE COMMAND STRING SWITCHES - DETAILED DESCRIPTION

A.3.1

Device-Specific Switches

A.3.1.1

TUlO 7-Channel Magtape Switches

Three switches:
/20
/55
/80
apply ·only to 7-channel TUlO magtapes operating in image copy mode.
Furthermore, when 7-channel tape is being used, the copy operation
must be accomplished in image mode.
If no tape density is defined,
the default density is 800 BPI.
These switches
specifiers.

may

applied

Example:
PRE>MTO:/VE=MTl:/80

A-5

either

input

output

file

PRES RV
NOTE
Logical tapes written
on
7-channel
drives are always in core-dump mode (two
4-bit frames per byte).

A.3.1.2

Phase-Encoded Tape Switch (/PE)

This switch indicates that magtape volumes to be written or read on a
TU16. drive are to be treated in phase-encoded mode (1600 BPI). This
switch may appear on either input or output file specifiers.
NOTE

A.3.1.3

File labels, bootstraps, and system
images are always written at 800
BPI.
This implies that
logical
tapes written with the /PE switch
may not be copied in image mode.

Logical tapes written on a TU16 are
bootstrappable on a TU16, TUlO, or
TS03, and logical tapes written on a
TUlO or TS03 are bootstrappable on a
TUlO, TS03, or TU16. However, only
800-BPI TU16 tapes are readable on a
TUlO or TS03.

Format Disk (/FO)

PRESRV is capable of writing the necessary timing and
sector
identification information on RK05, RP02, RP03, and RP04 disks to
prepare them before being used to store data.
This operation is
called formatting the volume and is specified by the /FO switch.
Formatting may be specified as a separate operation or in conjunction
with a copy operation. If a copy operation has been specified, the
format pass is made before the copy is made. The /FO switch may be
specified only on the output side of the command string.
Examples:
PRE>DKO:/FO
PRE>DPO:/FO/VE=MTO:DISK
When formatting starts, the message:

** BEGIN FORMATTING **
is displayed.

When formatting has been completed, the message

** END FORMATTING **
is displayed.

A-6

PRES RV
Before RP02 or RP03 disks may be formatted, formatting must be enabled
with the format enable switch on the RPll controller. This switch is
located .on the controller in the third bay from the bottom. The front
cover must be removed to expose the switch. PRESRV will prompt the
user to enable and disable formatting at the proper times with the
following messages:
ENABLE FORMATTING WITH SWITCH ON CONTROLLER. TYPE CR WHEN READY>
DISABLE FORMATTING WITH SWITCH ON CONTROLLER. TYPE CR WHEN READY>
When the switch is set properly, the user should
return key to notify PRESRV to continuea

A.3.2

press

the

carriage

Administrative Switches

A.3.2.l

Abort on Hard I/O Error (/ER)

If the /ER switch is attached to a file specifier, an unrecoverable
I/O error will cause the termination of the current PRESRV operation.
PRESRV prompts with:
PRE>
after issuing an error message.
The switch /ER applies to either input or output file specifiers.
Examples:
PRE>DTl:A/ER=DK:

A.3.2.2

Rewind at Completion (/RW)

The /RW switch will cause PRESRV to rewind any physical tape that has
completed the specified transfer. The /RW switch applies to either
input or output file specifiers.
Example:
PRE>MTl:FILEABC/RW=DPO:

A.3.2.3

Verify Switch (/VE)

The /VE switch will cause a second pass over the input and output
media so that every record written is read back and compared with the
corresponding record on the input device. A verification failure is
declared if the comparison fails. The /VE switch may be on either the
input or output side of the command line.
Example:
PRE>DKO:/VE=DKl:
PRE>DKO:=DKl:/VE

A-7

PRES RV
A.3.3

Formatting Switches

PRESRV's principal function is to make media copies.
/BL
/FI
/TP
/IM

Four switches:

(Blocking) ,
(Files-11),
(Logical tape), and
(Image mode),

are used to explicitly establish formats.
all devices supported by PRESRV.

Defaults

are

assumed

for

Devices handled by PRESRV fall into one of two classes:
1.

Tape devices, to which
cassette belong, and

the

TUlO,

TS03,

TU16,

and

TAll

Block-structured devices, to
belong.

which

DECtape

and

all

disks

Tape devices are always defaulted to logical tape format,
block-structured devices are always defaulted to Files-11 format.
exceptions to this rule exist:
1.

and
Two

If the input is in Files-11 format and the output is a
block-structured device without a switch, and the input and
output devices are not identical, then the output will be
defaulted to logical tape. Thus, the string:
PRE>DTO:A=DKO:
will create a logical tape on DTO: rather than a Files-11
format.
This choice was made since the typical transfer
between non-identical media is done to create backup copies
of the system.
The resulting DECtape(s), being in logical
tape format, contain a bootstrappable image of PRESRV on the
first volume;

If both the input and output devices are identical and
neither has a switch, the transfer will be in image mode,
producing an output format identical to the input format.

Blocking can be imposed on transfers on either logical tapes or
block-structured
devices.
A
block is a unit of 256 words,
representing the smallest amount of data that will be transferred by a
single read or write request issued by PRESRV.
Each device has a default blocking factor which is used by PRESRV if
no blocking switch is specified, and if buffer space is available to
perform the default blocking.
The defaults, shown in Table A-1, are overridden as follows:
1.

For block-structured devices the /BL switch can
either input or output file specifiers to
blocking factor.

For logical tape output,
blocking factor.

A-8

/BL

switch

can

be used on
establish a
establish

the

PRES RV
3.

For logical tape input, the blocking factor is always
from the file label; the /BL switch is ignored.

taken

Finally, if a copy is image mode, PRESRV will, for tapes, use
a blocking factor which uses all available buffering space.

Blocking calculations used to match available buffering space to
blocking specifications for a file specifier appear in section
A.3.3.1.
Table A-1
Default Blocking Factors
DEVICE

FORMAT CLASS

DB
DF
DK
DP
DS
DT
DX
MM
MT

•
Logical
tape
Block structured
Block structured
Block structured
Block structured
Block structured
Block structured
Block structured
Logical tape
Logical tape

A.3.3.l

Blocking Switch (/BL:nn)

DEFAULT BLOCK SIZE
L

22.
8.

12.
10.
8.

2.
6.
1.
1.

This switch is of the form:
/BL:nn
where:
is the standard switch separator.
nn

is the number of 256(10)-word blocks in
May be specified in octal or decimal
point) .

each

record.

The /BL switch can apply to either input or output file specifiers.
Example:
PRE>MTl:ABQRS/BL:20.=DBO:/BL:ll.
NOTE
1.

For image copies,
the
blocking
factor specified with a disk or
DECtape must be a submultiple of the
number
of blocks on the volume
(e.g., 4800 for DK).

If buffer space permits, use a large
blocking
factor
for DECtape to
improve operational efficiency.

A-9

PRES RV

A.3.3.2

Files-11 Format Switch (/FI)

This switch applies only to Files-11 devices (disks and DECtape). The
/FI switch is used to override a default that would result in a PRESRV
operation not intended by the user.
For example, when units are
identical and no switches are used, the transfer is in image mode. A
Files-11 transfer can be effected by using the /FI switch.
The /FI
switch may apply to either input or output file specifiers.
NOTE
If an image transfer
(DKO:=DKl:) fails
and the input volume is a Files-11
device, use an /FI transfer to copy only
blocks in use.
The blocks which are
unreadable may be free blocks which do
not contain data.
This ·type of cdpy
(/FI) takes more time and should be used
only if I/O errors occur in an image
mode copy.
Example:
PRE>DKO:/FI=DKl:/FI

A.3.3.3

Logical Tape Format Switch (/TP)

As with the /FI switch, the /TP switch is used to override an
inappropriate PRESRV default. On output, the /TP switch can apply to
any device, and the the result of the copy operation will be a logical
tape.
On input, the volume to which /TP is attached must,
logical tape format.

fact,

Example:
PRE>DKO:=DTO:A/TP
It is possible to copy specific files from one tape to another in
logical tape mode. This is useful if one wishes to combine a number
of small tapes (e.g., 600-ft.) into a large tape (e.g., 2400-ft.), or
to extract a few files onto a small tape from a large tape. This is
done by specifying a filename on the output side and, optionally, on
the input side. If no filename is given on the input side, the next
file on the tape is copied.
Examples:
PRE>MTl:ABC/TP=MTO:ABC/TP
PRE>MTl:DEF/TP=MTO:/TP
PRE MTl:GHI/TP=MTO:JKL/TP

A-10

PRES RV
A.3.3.4

Image Mode Switch {/IM)

This switch is used to override the Files-11 default in copying
disk-structured volumes to tape. Using the /IM switch, an operator is
capable of copying DOS-11, RSTS, MUMPS, or any other format disk to
tape.
Example:
PRE>MTO:A=DPO:/IM

A.4

PRESRV COMMAND SWITCHES - DETAILED DESCRIPTION

A.4.1

Hardware-Soecific Switches

During the system generation of PRESRV, the following CSR
addresses are established:
DEVICE

CSR

VECTOR

CT
DB
DF
DK
DP
DS
DT

177500
176700
177460
177404
176714
172040
177342
171700
172440
172522

260
300
204
220
254
310
214
264
320
224

MM
MT

and

vector

Two units (physical unit numbers 0 and 1) are generated for each
device type.
Both units of each device type are defined as being on
the same controller. Thus, if the CSR or vector is changed for one
unit, it is automatically changed for the other units.
If the generated CSR or vector address does not correspond to the
actual hardware configuration in use, the /CSR and /VEC commands can
be used to correct PRESRV-generated values.
The /UNIT switch is used to set the desired physical unit number for a
device mnemonic. Thus, for a system which has more than two drives of
any device type, it is possible to use them in PRESRV by setting the
physical unit number. For example:
DKO:/UNIT=2
will set DKO: to physical unit 2, and every reference to the name
DKO:
will actually reference physical unit 2 on the RKll controller.
Another use for the /UNIT switch is for mixed massbus peripherals. It
is possible to attach RP04, RS03/RS04 disk drives and TM02 tape
formatters to the same RH11/RH70 controller. By using the /UNIT and
/TM02 switches, it is possible to set up the mapping of device
mnemonics to the actual physical devices. The /TM02 switch is used to
set the unit number of the TM02 tape formatter on the RH controller.

A-11

PRES RV
The /UNIT switch is used to select the unit number of RP04, RS03, and
RS04 disk drives on the RH controller or the drive number of the TU16
tape drives on the TM02.
The /CSR, /VEC, /UNIT, and /TM02 switches may all be specified in
same command line.

A.4.1.1

the

Set Control Status Register Address

The format is:
dev:/CSR=nnnnnn
where:
dev: = The device whose CSR is to be changed.
nnnnnn = The octal address of CSR used by the device.
Example:
DSO:/CSR=l72040

A.4.1.2

Set Physical Unit Number

The format is:
dev:/UNIT=nn
where:
dev:
nn

The device-unit whose unit number is to be changed.
The new unit number for the device-unit.

Example:
DB1:/UNIT=3

A.4.1.3

Set Vector Address

The format is:
dev:/VEC=nnn
where:
dev: = The device-unit
changed.
nnn

whose

vector

The new octal vector address.

Example:
MMO:/VEC=224

A-12

address

PRES RV
A.4.1.4

Set TM02 Unit Number

The format is:
dev:/TM02=n
where:
dev:

The TM02 formatter whose position
controller is to be changed.

the

Example:
MMO:/TM02=2

A.4.2

Stand-Alone Switches

Stand-alone switches must be specified on separate command lines.

A.4.2.1

Help Switch

The Help switch, whose format is:
/HE or /HELP
causes a summary of the PRESRV commands, switches, and defaults to
printed on the console terminal.

A.4.2.2

Listing Switch

This switch is of the form:

dev:/LI
and lists all tape file labels on the specified device.
The device
must be formatted as a logical tape. Regardless of whether the tape
was created with the /PE switch specified or not, the tape labels are
always written in 800-BPI mode on 9-track drives. Thus, the /PE
switch ~s not needed to list the files on a tape created with the /PE
switch specified.

A.4.2.3

Time Switch

This switch is of the form:
/TIM=mm/dd/yy

hh:mm[:ss]

where:
mm
dd
year

month (1-12);
day (1-31);
year (0-99);

A-13

PRES RV
hh
mm

hour (0-23);
minutes (0-59), and
seconds (0-59) (optional).

All values are decimal and do not require a trailing decimal point.

A.4.2.4

Boot Switch

This switch is of the form:
dev:/BO
and will read the bootstrap block on the specified device into memory,
starting at location 0, and pass control to it.
This is a means of bootstrapping a system
there is no ROM bootstrap loader.

A.5

PRESRV OPERATOR INTERFACE

A.5.1

Bootstrapping PRESRV

from

device

for

which

To initiate the PRESRV version distributed with the Release 2 kits,
first bootstrap the RSX-llM system, as discussed in section 2.3. When
the system prompts with:

>
enter
>INS [l,50]BOO (Install Boot command)
>BOO [l, 50] PRESRV (RK Exec source, unmapped object or RP disks)
or
>INS [l,54]BOO (Install Boot command)
>BOO [l,54]PRESRV (RK mapped object disk)
To bootstrap PRESRV from logical tapes (e.g., RP distribution kit
tapes), use the bootstrap loader to bootstrap from the logical tape.
The PRESRV stand-alone system will replace the current Executive and
display an identification message of the form:
RSX-llS VOl BL12
VOLUME PRESERVATION PROGRAM - V02.l
FOR HELP TYPE /HE
nn. - BLOCKS AVAILABLE FOR BUFFERING
Following this message, PRESRV prompts with:
PRE>
indicating its readiness to accept commands.

A-14

PRES RV
PRESRV is built as a task running under the RSX-llS
the display:

Executive.

Thus

"RSX-llS VOl BL12"
indicates the version of the RSX-llS Executive in use.
version number of PRESRV.

V02.l

the

NOTE
It is advisable for
the
user
to
bootstrap
PRESRV
from its original
medium, configure it if necessary (using
/CSR,
etc.),
and write a DECtape,
magtape, cassette, or floppy disk by
preserving some Files-11 volume with a
command such as:
PRE>DTO:A=DKO:
Subsequent bootstrapping of the logical
tape
will
activate a user-tailored
version of PRESRV.
The data on the
logical tape is of no consequence.
If, in the course of trying to perform some function, PRESRV haics at
location 522, the system has crashed. The probable cause is improper
values for the CSR or vector for one of the specified devices in the
command string.
Set the correct CSR and vector addresses for the
devices in question and issue the command again.

A.5.1.1

Computing Required Blocks for Buffering

In the display of the number of blocks available for buffering, the
value nn.
is tne number of 256(10)-word blocks available for data
transfer and verification operations.
The user must use nn.
to
determine if a given PRESRV media copy has sufficient buffering space.
The space required depends on blocking factors and the use of the
verification switch.
Default block sizes for devices supported by
PRESRV are given in Table A-1.
The example presented below illustrates the method for determining the
blocks required for PRESRV copy operations. The rule is to establish
the blocks required for:
Input,
Output, and
Verification.
Example 1:
PRE>MTO:ABC=DKO:

A-15

PRES RV
This exampl~ creates a logical
Blocking requirements are:

tape

labeled

ABC

Input

12.

(DK default block size).

Output

(MT default block size).

Verification

(No ver f ica ti on spec i f ied) .

Total

13.

unit

MTO:.

Example 2:
PRE>DKO:

= MTO:XYZ

Copy file XYZ on MTO:

to Files-11 unit DKO:.

Input

Record size specified in file XYZ's label.

Output

12.

Verification

Total

12., plus record size of file XYZ.

PRE>DKO:/VE

MTO:IJK

(DK default block size).

Example 3:

Copies file IJK on MTO: to unit DKO:. Each record written
will be read and compared to its memory-resident image.
Input

Record size of file IJK.

Output

12.

(DK default block size).

Verification

12.

(Re-read block from DKO:).

Total

24., plus record size of file IJK.

DKO:

Example 4:
PRE>MMO:

= MMl:

Image copies MMl:
to MMO:.
The image copy mode will use all
available memory buffer space in creating blocks on MMO:. If any
input record is too large, the transfer will be aborted.
Example 5:
PRE>MMO:/VE = MMl:
Image copies MMl: to MMO:. Since verification is requested, one-half
of available memory buffer space will be used to form blocks to be
written to MMO:. If any input record is too large, the transfer will
be aborted.

A-16

PRES RV
A.5.1.2

Bootstrapping Another System or Stand-alone Program

Use the /BO switch to bootstrap another system or stand-alone program
if the device has a bootstrap block pointing to the desired system or
program.

A.5.2

Operator Messages

A.5.2.1

Action Messages

During the course of PRESRV's operation, it may be necessary for the
operator to take certain actions. At such times, PRESRV will issue
prompting messages to the user's terminal.
These messages are as
.i=,....,
, """""' .. _.,_ •
J..V..L..LVYYO o

**BEGIN VERIFICATION**
MOUNT OUTPUT VOLUME nn AND TYPE CR>
MOUNT INPUT VOLUME nn AND TYPE CR>
OUTPUT VOLUME WRITE-LOCKED.

TYPE CR WHEN READY>

VOLUME FAULTY - "R" TO RETRY, "C" TO CONTINUE>
ENABLE FORMATTING WITH SWITCH ON CONTROLLER.

TYPE CR WHEN READY>

DISABLE FORMATTING WITH SWITCH ON CONTROLLER.

TYPE CR WHEN READY>

The VOLUME FAULTY message is issued whenever a verification error or a
write error occurs and the /ER switch has not been specified.
NOTE
If PRESRV is ready to accept input and
the user types "ALTMODE", the current
operation will be aborted.

A.5.2.2

Error Messages

PRESRV error messages are listed alphabetically below.
CONFLICTING SWITCHES
Two switches which conflict in the operation to be performed have
been specified, .e.g, /TP and /FI.
Action:

Specify the correct switches.

DEVICE NOT FORMATTABLE
The device specified to be formatted is not one of the following:
RK03/05, RP02/03, or RP04. Only these devices may be formatted
by PRESRV.
Action:

Specify correct device mnemonic.

A-17

PRES RV
DEVICE OFFLINE
The specified device is not in the hardware configuration, or the
CSR is set improperly.
Action:

Set the CSR and vector for the device to the proper
addresses. Then retype command.

DICTIONARY READ ERROR
An error has occurred in reading
equivalent) from a logical tape.

dictionary

(storage

bitmap

Action: Retry or abort.
DRIVE NOT READY
The specified drive is not up to speed and ready.
Action:

Wait until the drive is ready.

ERROR IN READING COMMAND
There is some problem with terminal input, or a

was typed.

Action: Retype command.
EXPECTED EOF NOT FOUND
During verification of a tape-image copy, an EOF mark
from the input volume, but not from the output volume.
Action:

was

read

Retry the operation.

FATAL ERROR ON INPUT DEVICE
An error has occurred while reading the
tape-image copy.
Action:

input

volume

during

Retry the operation.

FATAL ERROR ON OUTPUT DEVICE
An error has occurred while reading or writing the output
during a tape image copy.
Action:

volume

Retry the operation.

ILLEGAL BLOCK COUNT
The number of blocks specified with the /BL switch
for the available buffer space, or it is 0.
Action:

too

Retry the operation with a valid block count.

A-18

high

PRES RV
INPUT DEVICE ERROR

nnnnn

An error has occurred in attempting to read data.
The number
provided is the octal logical block number (24 bits) when the
error was detected.
Act ion:
INPUT VOLUME nn.

None.
OUT OF SEQUENCE

The number of the input volume does not match the number
required volume.
Action:

the

Mount proper volume and proceed.

INSUFFICIENT BUFFER SPACE
A record (or all records) on an input tape has been found to be
larger than the available buffer space.
Alternatively, the
transfer specification requires more blocks than are available
(see section A.5.1.1).
Action:

Abort (or retry in the alternate case}.

LABEL OR BOOTSTRAP WRITE ERROR
An error has occurred while writing the initial part of a logical
tape.
Action:

Retry or abort the operation.

LUN ASSIGNMENT ERROR
A device and unit combination specified in a :Dmmand string is
either not represented in the current PRESRV system or it is not
a valid device for PRESRV operations.
Action:

Determine device status and try a proper command.

NO SUCH FILE
A file named in the command string was not found on
tape, or a magtape was at EOV for input.
Action:
OUTPUT DEVICE ERROR

(logical)

Retry the operation with proper filename.
nnnnnn

nnnnnn

An error occurred when PRESRV attempted to read or write data on
the output volume.
The number indicated is the octal logical
block number of the location where the error occurred.
Action:

PRESRV continues
specified.

A-19

unless

the

/ER

switch

was

PRES RV
OUTPUT VOLUME OVERFLOW
Output volume is a Files-11 device, and
than the output volume can hold.
Action:

input

has

blocks

Try some other transfer mode.

SYNTAX ERROR
Input not recognizable.
Action:

Retype command.

TAPE LABEL READ ERROR
An error has occurred in reading a logical tape label record.
Action:

Retry or abort the operation.

TAPE WAS WRITTEN FOR dd
The Files-11 volume being restored from a logical tape is not of
the same type as the one from which the tape was originally made.
The new volume will be given the bitmap of the old volume.
This
is only a warning message, not an error message. The characters
dd identify the intended Files-11 device.
TRANSFER SPECIFICATION MODE ERROR
Some inconsistency appears in the transfer command,
magnetic tape has an /FI switch associated with it.
Action:

e.g.,

Retype an acceptable command.

VERIFICATION ERROR

nnnnn

The data read back from the output volume does not match what was
written.
The number indicated is the octal logical block number
of the location where the error occurred.
Action:

Retry, continue, or abort the operation.

Any of the following three messages may appear if a volume assumed
be in Files-11 format is found not to be:

BITMAP FILE HOR READ ERROR
STORAGE BITMAP FILE READ ERROR
HOME BLOCK READ ERROR
Action:

Retype the command to specify the proper
mode, or wait until the device is ready.

A-20

transfer

APPENDIX B
SYSTEM CONVENTIONS

B.l

RSX-llM UIC CONVENTIONS

The RSX-llM Release 2 distribution kits observe a set of UIC
conventions for files on the disks. The purpose of these conventions
is to provide a uniform and consistent means of allocating, locating,
and maintaining files on disk.
On a disk, there are many user file directories
(UFD's)
under which
files are catalogued.
A particular UFD is referenced by specifying
the correct UIC. The UIC consists of two numbers separated by a comma
and enclosed in brackets (i.e., UIC [11,20]). The first number {11)
is the group member; the second number is the member number (20) and

indicates

the

specific

RSX-llM makes use of the
directories.

B.1.1