Digital PDFs

DEC-11-OMDRA-B-D

November 1975

308 pages

Original

11MB

Document:	RSX-11M I/O Drivers Reference Manual Version 2
Order Number:	DEC-11-OMDRA-B-D
Revision:	0
Pages:	308
Original Filename:	http://bitsavers.org/pdf/dec/pdp11/rsx11m_s/RSX11M_V2/DEC-11-OMDRA-B-D_Drvref_1975.pdf

OCR Text

RSX-11M
1/0 Drivers Reference Manual
Order No. DEC-11-0MDRA-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 form 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
COM TEX
DDT
DECCOMM

DECsystem-lo
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.

1/76-15

CONTENTS

Page

PREFACE

CHAPTER

0.1

MANUAL OBJECTIVES AND READER ASSUMPTIONS

xvii

0.2

STRUCTURE OF THE DOCUMENT

xvii

0.3

CONVENTIONS USED IN THIS MANUAL

xix

RSX-llM INPUT/OUTPUT

1-1

1.1

OVERVIEW OF RSX-llM I/0

1-1

1. 2

PHYSICAL, LOGICAL, AND VIRTUAL I/O

1-2

1. 3

RSX-llM DEVICES

1-2

1. 4
1. 4 .1
1. 4. 2
1. 4. 3

LOGICAL UNITS
Logical Unit Number
Logical Unit Table
Changing LUN Assignments

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

1. 5
1. 5.1
1. 5. 2
1. 5. 3

ISSUING AN I/O REQUEST
QIO Macro Format
Significant Events
System Traps

1-6
1-10
1-10

1.6

DIRECTIVE PARAMETER BLOCKS

1-11

1. 7
1. 7 .1
1. 7. 2
1. 7. 3
1. 7. 4
1. 7. 5
L 7e6
1. 7. 7

1-12
I/0-RELATED MACROS
1-13
The QIO$ Macro: Issuing an I/O Request
1-14
The DIR$ Macro: Executing a Directive
The .MCALL Directive: Retrievinq System Macros 1-14
The ALUN$ Macro: Assigning a LUN 1-15
The GLUN$ Macro: Retrieving LUN Information
1-17
The ASTX$S Macro: Terminating AST Service
1-19
The WTSE$ Macro: Waiting for an Event Flag
1-19

1. 8
1. 8 .1
1. 8. 2
1. 8. 3
1. 8. 4
1. 8. 5
1. 8. 6
i. 0. 1

STANDARD I/O FUNCTIONS
IO.ATT: Attaching to an I/O Device
IO.DET: Detaching from an I/O Device
IO.KIL: Canceling I/O Requests
IO.RLB: Reading a Logical Block
IO.RVB: Reading a Virtual Block
IO.WLB: Writing a Logical Block
IO.WVB: Writing a Virtual Block

1-20
1-21
1-22
1-22
1-22
1-23
1-23
1-23

1. 9

I/O COMPLETION

1-24

1.10
1.10 .1
1.10.2

RETURN CODES
Directive Conditions
I/O Status Conditions

1-25
1-26

TERMINAL DRIVER

2-1

2.1
2 .1.1

INTRODUCTION
ASR-33/35 Teletypes

2-1
2-2

iii

1-7

1-27

CONTENTS (Cont.)
Page
2 .1. 2
2 .1. 3

2-2
2-2
2-2

2.1. 6
2.1. 7

KSR-33/35 Teletypes
LA30 DECwriters
LA36 DECwr iter
RT02 Alphanumeric Display Terminal and RT02-C
Badge Reader/Alphanumeric Display Terminal
VT05B Alphanumeric Display Terminal
VT50 Alphanumeric Display Terminal

2.2

GET LON INFORMATION MACRO

2-3

2.3

QIO MACRO

2-3

2.4

STATUS RETURNS

2-4

2.5
2.5.1
2.5.2

CONTROL CHARACTERS AND SPECIAL KEYS
Control Characters
Special Keys

2-7
2-7
2-8

2.6

VERTICAL FORMAT CONTROL

2-9

2.7
2.7.1
2.7.3

TERMINAL INTERFACES
DHll Asynchronous Serial Line Multiplexer
DJll Asynchronous Serial Line Multiplexer
DLll Asynchronous Serial Line Interface

2-10
2-10
2-10
2-10

2.8
2.8.1
2.8.2
2.8.3

PROGRAMMING HINTS
Terminal Line Truncation
ESCape Code Conversion
RT02-C Control Function

2-11
2-11
2-11
2-11

DISK DRIVERS

3-1

3.1
3 .1.1
3.1. 2
3 .1. 3
3 .1. 4
3.1. 5
3.1. 6
3 .1. 7

INTRODUCTION
RFll/RSll Fixed-Head Disk
RP04 Pack Disk
RS03 Fixed-Head Disk
RS04 Fixed-Head Disk
RK11/RK05 Cartridge Disk
RP11/RP03 or RP02 Pack Disk
RXll/RXOl Flexible Disk

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

3.2

GET LON INFORMATION MACRO

3-3

3.3
3.3.1
3.3.2

QIO MACRO
Standard QIO Functions
Device-Specific QIO Functions

3-3
3-3
3-4

3.4

STATUS RETURNS

3-5

DECTAPE DRIVER

4-1

4.1

INTRODUCTION

4-1

4.2

GET LON INFORMATION MACRO

4-1

4.3
4.3.1
4.3.2

QIO MACRO
Standard QIO Functions
Device-Specific QIO Functions

4-2
4-2
4-2

4.4
4.4.1

STATUS RETURNS
DECtape Recovery Procedures

4-3
4-5

2 .1. 4
2 .1. 5

2.7.2

CHAPTER

2-2

2-3
2-3

CONTENTS (Cont. )
Page

CHAPTER

4.4.2

Select Recovery

4-6

4.5
4.5.i
4.5.2
4.5.3
4.5.4

PROGRAMMING HINTS
DECtape Transfers
Reverse Reading and Writing
Speed Considerations When Reversing Direction
Aborting a Task

4-6
4-6
4-6
4-6
4-7

MAGNETIC TAPE DRIVERS

5-1

5.1
5 .1.1
5 .1. 2

INTRODUCTION
TU10/TS03 Magnetic Tape
TU16 Magnetic Tape

5-1
5-1
5-2

5.2

GET LON INFORMATION MACRO

5-2

5.3
5.3.l
5.3.2
5.3.2.1
5.3.2.2
5.3.2.3

QIO MACRO
Standard QIO Functions
Device-Specific QIO Functions
IO.RWD
IO.RWU
IO.SEC

5-2
5-3
5-3
5-4
5-4
5-4

5.4
5.4.1
5.4.2

STATUS RETURNS
Select Recovery
Retry Procedures for Reads and Writes

5-8
5-10
5-11

5.5
5.5.1
5.5.2
5.5.3
5.5.4

PROGRAMMING HINTS
Block Size
Importance of Resetting Tape Characteristics
Aborting a Task
Writing an Even-Parity Zero

5-11
5-11
5-11
5-11
5-11

CASSETTE DRIVER

6-1

6.1

INTRODUCTION

6-1

6.2

GET LON INFORMATION MACRO

6-1

6.3
6.3.l
6.3.2

QIO MACRO
Standard QIO Functions
Device-Specific QIO Functions

6-2
6-2

6.4
6.4.1

STATUS RETURNS
Cassette Recovery Procedures

6-3
6-5

6.5

STRUCTURE OF CASSETTE TAPE

6-5

6.6
6.6.l
6.6.2
6.6.3
6.6.4
6.6.5
6.6.6

PROGRAMMING HINTS
Importance of Rewinding
End-of-File and IO.SPF
The Space Functions, IO.SPB and IO.SPF
Verification of Write Operations
Block Length
Logical End-of-tape

6-6
6-7
6-7
6-7
6-7
6-7
6-7

LINE PRINTER DRIVER

7-1

7.1
7 .1.1
7 .1. 2

INTRODUCTION
LPll Line Printer
LSll Line Printer

7-1
7-1
7-2

6-2

CONTENTS (Cont.)
Page

CHAPTER

7 .1. 3

LVll Line Printer

7-2

7.2

GET LUN INFORMATION MACRO

7-2

7.3

QIO MACRO

7-2

7.4
7.4.1

STATUS RETURNS
Ready Recovery

7-3
7-4

7.5

VERTICAL FORMAT CONTROL

7-5

7.6
7.6.l
7.6.2
7.6.3

PROGRAMMING HINTS
RUBOUT Character
Print Line Truncation
Aborting a Task

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

CARD READER DRIVER

8-1

8.1

INTRODUCTION

8-1

8.2

GET LUN INFORMATION MACRO

8-1

8.3
8.3.1
8.3.2

QIO MACRO
Standard QIO Functions
Device-Specific QIO Function

8-2
8-2
8-2

8.4
8.4.1
8.4.2
8.4.3

STATUS RETURNS
Card Input Errors and Recovery
Ready and Card Reader Check Recovery
I/O Status Conditions

8-3
8-3
8-6
8-7

8.5
8.5.1

FUNCTIONAL CAPABILITIES
Control Characters

8-8
8-8

8.6
8.6.1
8.6.2

CARD READER DATA FORMATS
Alphanumeric Format (026 and 029)
Binary Format

8-9
8-9
8-10

8.7
8.7.1
8.7.2

PROGRAMMING HINTS
Input Card Limitation
Aborting a Task

8-10
8-10
8-11

MESSAGE-ORIENTED COMMUNICATION DRIVERS

9-1

9.1
9 .1.1
9 .1. 2
9 .1. 3
9 .1. 4
9 .1. 5

INTRODUCTION
DAll-B Parallel Interprocessor Link
DLll-E Asynchronous Line Interface
DPll Synchronous Line Interface
DQll Synchronous Line Interface
DUll Synchronous Line Interface

9-1
9-2
9-2
9-2
9-3
9-3

9.2

GET LUN INFORMATION MACRO

9-3

9.3
9.3.1
9.3.2
9.3.2.1
9.3.2.2
9.3.2.3
9.3.2.4
9.3.2.5

QIO MACRO
Standard QIO Functions
Device-Specific QIO Functions
IO.FDX
IO.HDX
IO.INL and IO.TRM
IO.RNS
IO.RWD

9-4
9-4
9-5
9-5
9-5
9-5
9-6
9-6

CONTENTS (Cont.)
Page

CHAPTER

9 3.2.6
9.3.2.7

IO. SYN
IO.WNS

9-6
9-6

9.4

STATUS RETURNS

9-7

9.5
9.5.1
9.5.2
9.5.3
9.5.4
9.5.5
9.5.6

PROGRAMMING HINTS
Transmission Validation
Redundancy Checking
Half-Duplex and Full-Duplex Considerations
Low-Traffic Sync Character Considerations
Vertical Parity Support
Importance of IO.INL

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

9.6

PROGRAMMING EXAMPLE

9-10

ANALOG-TO-DIGITAL CONVERTER DRIVERS

10-1

10.1
10.1.1
10.1.2

INTRODUCTION
AFCll Analog-to-Digital Converter
ADOl-D Analog-to-Digital Converter

10-1
10-1
10-1

10.2

GET LUN INFORMATION MACRO

10-2

10.3
10.3.1
10.3.2

QIO MACRO
Standard QIO Function
Device-Specific QIO Function

10-2
10-2

10.4
10.4.1
10.4.2
10.4.3
10.4.4

10.4.6
10.4.7

FORTRAN INTERFACE
10-3
Synchronous and Asynchronous Process Control I/O 10-3
10-3
The isb Status Array
10-4
FORTRAN Subroutine Summary
AIRD/AIRDW: Performing Input of Analog Data
in Random Sequence
AISQ/AISQW: Reading Sequential Analog Input
Channels
10-5
ASADLN: Assigning a LUN to the ADOl-D
10-6
ASAFLN: Assigning a LUN to the AFCll
10-6

10.5
10.5.l

STATUS RETURNS
FORTRAN Interface Values

10.6
10.6.l

FUNCTIONAL CAPABILITIES
Control and Data Buffers

10-9

10.7
10.7.1
10.7.2
10.7.3
10.7.4

PROGRAMMING HINTS
Use of A/D Gain Ranges
Identical Channel Numbers on the AFCll
AFCll Sampling Rate
Restricting the Number of ADOl-D Conversions

10-9
10-9
10-9
10-9
10-10

UNIVERSAL DIGITAL CONTROLLER DRIVER

11-1

11.1
11.1.1
11.1. 2
11.1.2.1
11.1. 2. 2

INTRODUCTION
Creating the UDCll Driver
Accessing UDCll Modules
Driver Services
Direct Access

11-1
11-1
11-2
11-2
11-3

11. 2

GET LUN INFORMATION MACRO

11-3

11. 3

QIO MACRO

11-3

10.4.5

CHAPTER

vii

9-9

10-2

10-7
10-8

10-9

CONTENT s (Cont . )
Page

11.3.1
11. 3. 2
11.3.2.1
11.3.2.2
11.3.2.3
11.3.2.4
11.3.2.5
11. 4
11.4.1
11.4.1.1
11.4.1.2
11. 4. 2
11. 4. 3
11.4.3.l
11.4.3.2
11.4.3.3

Standard QIO Function
Device-Specific QIO Functions
Contact Interrupt Digital Input
(W733 Modules)
Timer (W734 I/O Counter Modules)
Latching Digital Output (M685, M803,
and M805 Modules)
Analog-to-Digital Converter (ADUOl Module)
ICSll Analog-to-Digital Converter
(!AD-IA Module)

11-3
11-3
11-5

DIRECT ACCESS
Defining the UDCll Configuration
Assembly Procedure for UDCOM.MAC
Symbols Defined by UDCOM.MAC
Including UDCll Symbolic Definitions in
the System Object Module Library
Referencing the UDCll through a Common Block
Creating a Global Common Block
Making the Common Block Resident
Linking a Task to the UDCll Common Block

11-8

11-7
11-7
11-7

11-8

11-9

11-9
11-10

11-11
11-11

11-12
11-13
11-13

11.5.20
11.5.21

11-14
FORTRAN INTERFACE
Synchronous and Asynchronous Process Control I/O 11-14
11-14
The isb Status Array
11-15
FORTRAN Subroutine Summary
AIRD/AIRDW: Performing Input of Analog
11-17
Data in Random Sequence
11-17
AISQ/AISQW: Reading Sequential Analog
Input Channels
11-18
AO/AOW: Performing Analog Output
11-18
ASUDLN: Assigning a LUN to UDO:
11-19
CTDI: Connecting to Contact Interrupts
11-19
CTTI: Connecting to Timer Interrupts
11-20
DFDI: Disconnecting from Contact Interrupts
11-21
DFTI: Disconnecting from Timer Interrupts
DI/DIW: Reading Several Contact Sense
11-21
Fields
DOL/DOLW: Latching or Unlatching Several
11-21
Fields
11-22
DOM/DOMW: Pulsing Several Fields
11-22
RCIPT: Reading a Contact Interrupt Point
RDCS: Reading Contact Interrupt
11-23
Change-of-State Data from a Circular Buffer
RODI: Reading Contact Interrupt Data
11-24
From a Circular Buffer
RDTI: Reading Timer Interrupt Data From
11-25
a Circular Buffer
RDWD: Reading a Full Word of Contact
11-26
Interrupt Data from the Circular Buffer
11-26
RSTI: Reading a Timer Module
11-26
SCTI: Initializing a Timer Module

11. 6
11. 6. 1

STATUS RETURNS
FORTRAN Interface Values

11-27
11-29

11. 7
11. 7 .1
11. 7. 2
11. 7. 3

PROGRAMMING HINTS
Checkpointable Tasks
Numbering Conventions
Processing Circular Buffer Entries

11-29
11-29
11-30

11. 5
11. 5 .1
11. 5. 2
11. 5. 3
11. 5. 4
11. 5. 5
11.5.6
11. 5. 7
11. 5. 8
11. 5. 9
11.5.10
11. 5 .11
11.5.12
11.5.13
11.5.14
11.5.15
11.5.16
11.5.17
11.5.18
11.5.19

viii

11-30

CONTENTS

(Cont. )

Page
CHAPTER

LABORATORY PERIPHERAL SYSTEMS DRIVERS

12-1

12.1
12.l.l
12.1.2

INTRODUCTION
ARll Laboratory Peripheral System
LPSll Laboratory Peripheral System

12-1
12-2
12-2

12.2

GET LUN INFORMATION MACRO

12-2

12.3
12.3.l
12.3.2
12.3.2.l
12.3.2.2
12.3.2.3
12.3.2.4
12.3.3

QIO MACRO
Standard QIO Function
Device-Specific QIO Functions (Immediate)
IO.LED
IO.REL
IO.SDI
IO. SDO
Device-Specific QIO Functions (Synchronous)

12.3.3.2
12.3.3.3
12.3.3.4
12.3.3.5
12.3.4
12.3.4.l

IO. HIS
IO.MDA
IO.MDI
IO.MOO
Device-Specific QIO Function (IO.STP)
IO.STP

12-2
12-2
12-3
12-3
12-4
12-4
12-4
12-4
12-5
12-6
12-7
12-7
12-7
12-8
12-8

12.4
12.4.l
12.4.2
12.4.3
12.4.4
12.4.5
12.4.6
12.4.7
12.4.8

12.4.16
12.4.17
12.4.18
12.4.19
12.4.20

FORTRAN INTERFACE
12-8
The isb Status Array
12-8
Synchronous Subroutines
12-9
FORTRAN Subroutine Summary
12-lC
ADC: Reading a Single A/D Channel
12-lJ
ADJLPS: Adjusting Buffer Pointers
12-1~
ASLSLN: Assigning a LUN to LSO:
12-1~
ASARLN: Assigning a LUN to ARO:
12-13
CVSWG: Converting a Switch Gain A/D Value to
Floating-Point
12-13
DRS: Initiating Synchronous Digital Input
12-14
Sampling
12-lE
HIST: Initiating Histogram Sampling
IDIR: Reading Digital Input
12-18
IDOR: Writing Digital Output
12-18
IRDB: Reading Data from an Input Buffer
12-19
L~D:
Displaying in LED Lights
12-19
LPSTP: Stopping an In-Progress Synchronous
Fune tion
12-20
PUTD: Putting a Data Item into an Output Buffer 12-20
RELAY: Latching an Output Relay
12-20
RTS:
Initiating Synchronous A/D Sampling
12-21
SDAC: Initiating Synchronous D/A Output
12-23
SDO: Initiating Synchronous Digital Output
12-24

12.5
12.5.l
12.5.2
12.5.3
12.5.4

STATUS RETURNS
IE.RSU: Resource in Use
Second I/O Status Word
IO.ADS and ADC Errors
FORTRAN Interface Values

12-26
12-28
12-29
12-30
12-30

12.6
12.6.1
12.6.2
12.6.3
12.6.4

PROGRAMMING HINTS
The LPSll Clock and Sampling Rates
Importance of the I/O Status Block
Buffer Management
Use of ADJLPS for Input and Output

12-31
12-31
12-32
12-32
12-33

IO.ADS

12.4.9
12.4.10
12.4.11
12.4.12
12.4.13

12.4.14
12.4.15

CONTENTS (Cont.)
Page
CHAPTER

CHAPTER

PAPER TAPE READER/PUNCH DRIVERS

13-1

13 .1

INTRODUCTION

13-1

13.2

GET LON INFORMATION MACRO

13-1

13. 3

QIO MACRO

13-2

13. 4
13.4.1
13.4.2

STATUS RETURNS
Error Conditions
Ready Recovery

13-2
13-4
13-4

13. 5
13.5.1
13.5.2

PROGRAMMING HINTS
Special Action Resulting from Attach and Detach
Reading Past End-of-Tape

13-4
13-4
13-4

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

14-1

14.1
14.1.1
14.1.1.1
14.1.1.2
14 .1. 2
14.1.3
14. 1. 4

INTRODUCTION
Hardware Configuration
Address Assignments
Supported I/O Modules
Alternate ICSll Support
Software Support
UDCll Software Compatibility

14-1
14-1
14-1
14-2
14-3
14-4
14-6

14.2

LON INFORMATION

14-6

14.3
14.3.1
14.3.2
14.3.3
14.3.4
14.3.5
14.3.6
14.3.6.1
14.3.6.2
14.3.6.3
14.3.6.4
14.3.6.5
14.3.6.6
14.3.6.7
14.3.7
14.3.7.1
14.3.7.2
14.3.7.3
14.3.7.4
14.3.7.5
14.3.7.6
14.3.8
14.3.9
14.3.9.l
14.3.9.2
14.3.10
14.3.10.1
14.3.10.2

ASSEMBLY LANGUAGE INTERFACE
General Error Status Returns
A/D Input - Read Multiple A/D Channels
Analog Output
Single-Shot Digital Output - Multi-Point
Bistable Digital Output - Multi-Point
Unsolicited Interrupt Processing
Connect to Digital Interrupts
Disconnect from Digital Interrupts
Connect to Counter Module Interrupts
Set Counter Initial Value
Disconnect from Counter Interrupts
Connect to Terminal Interrupts
Disconnect from Terminal Input
Activating a Task by Unsolicited Interrupts
Link a Task to Digital Interrupts
Link a Task to Counter Interrupts
Link a Task to Terminal Interrupts
Link a Task to Error Interrupts
Read Activating Data
Unlink a Task from Interrupts
Terminal Output
Maintenance Functions
Disable Hardware Error Reporting
Enable Hardware Error Reporting
Special Functions
I/O Rundown
Kill I/O

14-6
14-10
14-10
14-12
14-13
14-13
14-14
14-16
14-17
14-17
14-18
14-19
14-19
14-20
14-20
14-21
14-22
14-22
14-23
14-24
14-25
14-27
14-28
14-28
14-29
14-29
14-29
14-29

14.4
14.4.1
14.4.2
14. 4. 3

FORTRAN INTERFACE
14-29
Synchronous and Asynchronous Process Control I/O 14-31
Return Status Reporting
14-31
Optional Arguments
14-33

CONTENTS (Cont.)
Page
14.4.4

Assigning Default Logical and Physical
Units for ICS/ICR Input and Output
14-34
14.4.5
Analog Input
14-35
14.4.5.l
Random Channel Sequence
14-36
14.4.5.2
Sequential Channel Sequence
14-38
14.4.6
Analog Output - Multi-channel
14-40
14.4.7
Digital Output - Bistable Multiple Fields
14-42
14.4.8
Digital Input
14-43
14.4.8.l
Digital Sense Multiple Fields
14-43
14.4.8.2
Digital Interrupt Single-Point
14-44
14.4.9
Digital Output Momentary - Multiple Fields
14-45
14.4.10
Remote Terminal Output
14-46
14.4.11
Unsolicited Interrupt Data - Continual Monitoring14-47
14-48
14.4.11.1 Connect a Buffer for Receiving Digital Data
14.4.11.2 Reading Digital Interrupt Data
14-49
14.4.11.3 Disconnect a Buffer from Digital Interrupts
14-52
14.4.11.4 Connect a Buffer for Receiving Counter Data
14-52
14.4.11.5 Read Counter Data from the Circular Buffer
14-53
14.4.11.6 Miscellaneous Counter Routines
14-54
14.4.11.7 Disconnect a Buffer from Counter Interrupts
14-55
14.4.11.8 Connect a Circular Buffer to Terminal Interrupts 14-56
14.4.11.9 Read a Character from the Terminal Buffer
14-57
14.4.11.10 Disconnect a Circular Buffer from Terminal Input 14-57
14.4.11.11 Programming Example
14-58
14.4.12
Unsolicited Interrupt Processing - Task
Activation
14-60
14.4.12.l Link a Task to Interrupts
14-60
14.4.12.2 Read Activation Data
14-61
14.4.12.3 Remove Interrupt Linkage to a Task
14-63
14.4.13
Maintenance Functions
14-64
14.4.13.1 Place Selected Unit in Offline Status
14-65
14.4.13.2 Return a Device to Online Status
14-65

APPENDIX

14.5
14.5.1
14.5.2
14.5.3
14.5.4
14.5.5

ERROR DETECTION AND RECOVERY
Serial Line Errors
Power-fail at a Remote Site
Power Recovery at the Processor
Unit in Offline Status
Error Data - ICSR and !CAR Registers

14-65
14-66
14-66
14-67
14-67
14-67

14.6
14.6.l
14.6.2
14.6.2.1
14.6.2.2
14.6.2.3

DIRECT ACCESS
Linking a Task to the ICS/ICR Common Block
Accessing the I/O Page
Mapping Table Format
I/O Page Global Definitions
Sample Subroutine

14-69
14-70
14-71
14-71
14-73
14-73

14.7
14.7.1
14.7.2
14.7.2.1
14.7.2.2
14.7.2.3
14.7.2.4
14.7.2.5
14.7.2.6
14.7.2.7

CONVERSION OF EXISTING SOFTWARE
Features
Module Support
!AD-IA A/D Converter and IMX-IA Multiplexer
16-Bit Binary Counter
Bistable Digital Output
Momentary Digital Output
Noninterrupting Digital Input
Analog Output
Interrupting Digital Input

14-74
14-75
14-75
14-75
14-75
14-75
14-76
14-76
14-76
14-76

SUMMARY OF IO FUNCTIONS

A-1

A.l

ANALOG-TO-DIGITAL CONVERTER DRIVERS

A-1

CONTENTS (Cont.)
Page
A.2

CARD READER DRIVER

A-1

A.3

CASSETTE DRIVER

A-1

A.4

COMMUNICATION DRIVERS {MESSAGE-ORIENTED)

A-2

A.5

DECTAPE DRIVER

A-2

A.6

DISK DRIVERS

A-2

A.7

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

A-3

A.8

LABORATORY PERIPHERAL SYSTEMS DRIVERS

A-4

A.9

LINE PRINTER DRIVER

A-4

A.10

MAGNETIC TAPE DRIVERS

A-5

A.11

PAPER TAPE READER/PUNCH DRIVERS

A-5

A.12

TERMINAL DRIVER

A-5

A.13

UNIVERSAL DIGITAL CONTROLLER DRIVER

A-6

I/O FUNCTION AND STATUS CODES

B-1

B.l
B.1.1
B. l. 2

I/O STATUS CODES
I/O Status Error Codes
I/O Status Success Codes

B-1
B-1
B-3

B.2
B.2.1
B.2.2

DIRECTIVE CODES
Directive Error Codes
Directive Success Codes

B-3
B-3
B-3

B.3
B.3.1
B.3.2
B.3.3
B.3.4
B.3.5

B-3
B-4
B-4
B-4
B-4

B.3.6
B.3.7
B.3.8
B.3.9
B.3.10
B.3.11
B.3.12

I/0 FUNCTION CODES
Standard I/O Function Codes
Specific A/D Converter I/O Function Codes
Specific Card Reader I/O Function Codes
Specific Cassette I/O Function Codes
Specific Communications (Message-Oriented) I/O
Function Codes
Specific DECtape I/O Function Codes
Specific Disk I/O Function Codes {RXOl)
Specific ICS/ICR I/O Function Codes
Specific LPS I/O Function Codes
Specific Magtape I/O Function Codes
Specific Terminal I/O Function Codes
Specific UDC I/O Function Codes

B-5
B-5
B-5
B-5
B-7
B-7
B-7
B-8

APPENDIX

RSX-llM PROGRAMMING EXAMPLE

C-1

APPENDIX

GLOSSARY OF RSX-llM TERMS

D-1

APPENDIX

xii

CONTENTS (Cont.)
FIGURES

Number
1-1

Logical Unit Table

1-5

1-2

QIO Directive Parameter Block

1-12

5-1

Determination of Tape Characteristics for the
TUlO
5-6

5-2

Determination of Tape Characteristics for the
TU16
5-6

6-1

Structure of Cassette Tape

6-6

14-lA

Mapping Table Format

14-72

14-lB

Mapping Table Entry Format

14-72

TABLES

Number

Page

1-1

Directive Returns

1-27

1-2

I/O Status Returns

1-28

2-1

Standard Terminal Devices

2-1

2-2

Standard Communication Line Interfaces

2-2

2-3

Standard and Device-Specific QIO
Functions for Terminals

2-4

Terminal Status Returns

2-5

Terminal Control Characters

2-7

2-6

Special Terminal Keys

2-8

2-7

Vertical Format Control Characters

2-9

3-1

Standard Disk Devices

3-1

3-2

Standard QIO Functions for Disks

3-4

3-3

Device-Specific QIO Functions for
the RXOl Disk Driver

3-5

3-4

Disk Status Returns

3-5

4-1

Standard QIO Functions for DECtape

4-2

Device-Specific Functions for DECtape

4-3

DECtape Status Returns

4-3

xiii

CONTENTS (Cont.)
Page
5-1

Standard Magtape Devices

5-2

Standard QIO Functions for Magtape

5-3

Device-Specific QIO Functions for Magtape

5-4

Magtape Status Returns

5-8

6-1

Standard QIO Functions for Cassette

6-2

Device-Specific QIO Functions for Cassette

6-3

Cassette Status Returns

6-3

7-1

Standard Line Printer Devices

7-1

7-2

Standard QIO Functions for Line Printers

7-3

Line Printer Status Returns

7-3

7-4

Vertical Format Control Characters

7-5

8-1

Standard QIO Functions for the Card Reader

8-2

Device-Specific QIO Function for the Card Reader 8-2

8-3

Card Reader Switches and Indicators

8-4

Card Reader Status Returns

8-7

8-5

Card Reader Control Characters

8-9

8-6

Translation from DEC026 or DEC029 to ASCII

8-9

9-1

Message-Oriented Communication Interfaces

9-1

9-2

Standard QIO Functions for Communication
Interfaces

9-4

Device-Specific QIO Functions for Communication
Interfaces

9-5

9-4

Communication Status Returns

9-7

10-1

Standard Analog-to-Digital Converters

10-1

10-2

Standard QIO Function for the A/D Converters

10-2

10-3

Device-Specific QIO Function for the A/D
Converters

10-2

10-4

A/D Conversion Control Word

10-3

10-5

Contents of First Word of isb

10-4

10-6

FORTRAN Interface Subroutines for the AFCll and
ADOl-D

10-4

10-7

A/D Converter Status Returns

10~1

10-8

FORTRAN Interface Values

10=8

9-3

xiv

CONTENTS (Cont.)
Page

il-i

Standard QIO Function for the UDCll

11-3

11-2

Device-Specific QIO Functions for the UDCll

11-4

11-3

A/D Conversion Control Word

11-5

11-4

Contents of First Word of isb

11-15

11-5

FORTRAN Interface Subroutines for the UDCll

11-15

11-6

UDCll Status Returns

11-27

11-7

FORTRAN Interface Values

11-29

12-1

Laboratory Peripheral Systems

12-1

Standard QIO Function for the Laboratory
Peripheral Systems

12-2

Device-Specific QIO Functions for the
Laborarory Peripheral Systems (Immediate)

12-3

Device-Specific QIO Functions for the
Laboratory Peripheral Systems (Synchronous)

12-4

Device-Specific QIO Function for the
Laboratory Peripheral Systems (IO.STP}

12=8

12-6

Contents of First Word of isb

12-9

12-7

FORTRAN Interface Subroutines for the
Laboratory Peripheral Systems

12-10

12-8

Laboratory Peripheral Systems Status Returns

12-26

12-9

Returns to Second Word of I/O Status Block

12-29

12-10

FORTRAN Interface Values

12-31

13-1

Standard QIO Functions for the Paper
Tape Reader/Punch

13-2

Paper Tape Reader/Punch Status
Returns

13-2

14-1

ICS/ICR Address Assignments

14-1

14-2

Summary of ICS/ICR-11 QIO Functions

14-6

14-3

Sample ICS/ICR Configuration

14-10

14-4

FORTRAN Interface

14-30

14-5

Return Status Summary

14-32

14-6

A/D Conversion Control Word

14-36

14-7

ICSR Contents

14-67

14-8

ICAR Contents

14-68

12-3
12-4
12-5

13-2

PREFACE

0.1

MANUAL OBJECTIVES AND READER ASSUMPTIONS

This manual is designed to provide all information necessary to
interface directly with the I/O device drivers supplied as part of the
RSX-llM system.
It is intended for use by experienced RSX-llM
programmers who want to take advantage of the time and/or space
savings which result from direct use of the I/O drivers.
The orientation of this manual is tutorial, but it does not attempt to
introduce the reader to all areas of RSX-llM input/output operations.
Readers are expected to be familiar with the RSX-llM Executive
Reference Manual (DEC-11-0MERA-A-D) and to have some experience with
the Task Buiider and either FORTRAN IV or MACR0-11 assembly language.
Readers should also be familiar with the PDP-11 terminology presented
in the PDP-11 Processor Handbook and the PDP-11 Peripherals Handbook.
Users of RSX-llM who do not require such detailed knowledge of the I/O
drivers can use the device independent services provided by File
Control Services (FCS) as documented in the RSX-11 I/O Operations
Reference Manual {DEC~ll~OMFSA-A-D) .
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.

0.2

STRUCTURE OF THE DOCUMENT

This manual has three basic components:
1.

Chapter 1 provides an overview of RSX-llM input/output
operations.
It introduces the reader to the use of logical
unit numbers, directive parameter blocks, and macro calls.
It describes all of the I/O functions common to a variety of
devices, and summarizes standard error and status conditions
relating to completion of I/O requests.

Chapters 2 through 14 describe the use of all device
supported by RSX-llM. These include the following:

xvii

drivers

Device

Chapter
2

Terminals and terminal
communications line interfaces

Disks

DECtape

Magnetic tape

Cassette

Line printer

Card reader

Message-oriented communications
line interfaces

Analog-to-digital converters

Universal digital controller

Laboratory peripheral systems

Paper tape reader/punch

Industrial control local and remote
subsystems

Each of these chapters is structured in similar fashion and focuses on
the following basic elements:
. Description of the device, including information on physical
characteristics such as speed, capacity, access, and usage
. Summary of standard functions supported by
descriptions of device-specific functions

the

devices

and

. Discussion of special characters, carriage control codes, and
functional characteristics, if relevant
. Summary of error and status conditions returned on acceptance
or rejection of I/O requests
. Description of programming hints
under RSX-llM
3.

for

users

the

device

Appendixes A through D provide quick reference material on
I/O functions and status codes, a glossary of RSX-llM terms,
and an example of RSX-llM I/O operations. These include the
following:
Appendix

Contents

Summary of I/O
by dev i'ce

I/O function
codes

Programming example

functions
and

status

Glossary of RSX-llM terms

xviii

0.3

CONVENTIONS USED IN THIS MANUAL

There are a number of conventions and assumptions used in this manual
to present syntax and program coding examples. These are described in
the following list.
1.

Br ac ke ts ( [] }
parameters.

syntactic

models

enclose

optional

The following example illustrates this format:
ASTX$S
2.

[err]

Braces ( { } } in syntactic models indicate that
items must be selected, as in the following:

one

the

~DOM ~

CALL)
«inm,icont,idata, [idx], [isb], [lun]>
{DOMWJ
3.

An ellipsis ( .•. } in a syntactic model or coding example
indicates that parameters have been omitted. As used in this
manual, an ellipsis in a QIO macro call indicates omission of
standard QIO parameters described in section 1.4. This is
illustrated below:
QIO$C

optional

arguments

may

IO.KIL,9.

Certain parameters are required but ignored by RSX-llM; this
is necessary to maintain compatibility with RSX-llD. For
example, in the following, the priority specification (fourth
parameter) is ignored:
QIO$C

IO.ATT,6,,,,ASTOl

Commas indicating null trailing
omitted, as in the following:
QIO$C

null

Consecutive commas in a coding example
indicate
arguments. The following illustrates this usage:
QIO$C

IO.RLV, ..• ,<stadd,size>

IO.WLB,8.,EV,,IOSB,ASTX,<IOBUF,NBUF>

With the exception of MACR0-11 coding examples, all numbers
in the text of this manual are assumed to be decimal; octal
radix is explicitly declared as in the following:
An illegal logical block number has been specified
DECtape. The number exceeds 577 (1101 octal}.

for

In MACR0-11 coding examples, all numbers are assumed to be
octal;
decimal radix is explicitly designated by following
the number with a decimal point, as in the following example:
QIO$C

IO.RDB,14.,,,IOSB,,<IOBUF,80.>

xix

In FORTRAN subroutine models, parameters which begin with the
letters i through n indicate integer variables, as in the
following example:
CALL DRS (ibuf ,ilen,imode,irate,iefn,imask,isb,
[nbuf], is tart], [is top])
In general, where both i and n prefixes are used in a call,
the i form indicates the name of an array and the n form
specifies the size of the array.
All integer arrays and variables are assumed to occupy one
storage word per variable (i.e., INTEGER*2) and all real
arrays and variables are assumed to occupy two storage words
per variable {i.e., REAL*4).

CHAPTER 1
RSX-llM INPUT/OUTPUT

1.1

OVERVIEW OF RSX-llM I/O

The RSX-llM real-time Executive supports a wide variety of PDP-11
input and output devices, including disks, DECtapes, magnetic tapes,
tape cassettes, line printers, card readers, and such laboratory and
industrial devices as analog-to-digital converters, universal digital
controllers, and laboratory peripheral systems.
Drivers for these
devices are supplied by Digital Equipment Corporation as part of the
RSX-llM system software. This manual describes all of the device
drivers supported by RSX-llM and the characteristics, functions, error
conditions, and programming hints associated with each.
PDP-11
devices not described in this manual can be added to basic RSX-llM
configurations, but users must develop and maintain their own drivers
for these devices.
Input/output operations under RSX-llM are extremely flexible and are
as device- and function-independent as possible. Programs issue I/O
requests to logical units which have been previously associated with
particular physical device units.
Each program or task is able to
establish its own correspondence between physical device units and
logical unit numbers (LUNs). I/O requests are queued as issued; they
are subsequently processed according to the relative priority of the
tasks which issued them. I/O requests can be issued from MACR0-11 or
FORTRAN tasks by means of the File Control Services (for appropriate
devices), or can be interfaced directly to an I/O driver by means of
the QIO system directive.
All of the I/O services described in this manual are requested by the
user in the form of QIO system directives. A function code included
in the QIO directive indicates the particular input or output
operation to be performed. I/O functions can be used to request such
operations as:
. attaching or detaching a physical device unit for a
exclusive use

task's

. reading or writing a logical or virtual block of data
. canceling a task's I/O requests
A wide variety of device-specific input/output operations (e.g.,
reading DECtape in reverse, rewinding cassette tape) can also be
specified via QIO directives.

1-1

RSX-llM INPUT/OUTPUT
1.2

PHYSICAL, LOGICAL, AND VIRTUAL I/O

There are three possible modes in which an I/O
place. These are physical, logical, and virtual.

transfer

can

take

Physical I/O concerns reading and writing data in the actual physical
units accepted by the hardware (e.g., sectors on a disk). For most
devices, physical I/O is identical to logical I/O. For example, the
RK05 disk has sectors of 256 words, the same number as RSX-llM logical
blocks for all disks. Thus, in this case, a logical block maps
directly into a physical block. For other devices, the mapping is not
one to one. The RFll disk, for example, is word-addressable; however
no physical I/O may be done with the RFll. Data is always written in
256-word logical blocks. Another example is the RXOl flexible disk.
Data is recorded in physical sectors of 64 words each. Logical blocks
are made up of four physical sectors.
Logical I/O concerns reading and writing data in blocks that are
convenient for the operating system. In most cases, logical blocks
map directly into physical blocks.
For block-structured devices
(e.g., disks), logical blocks are numbered beginning at zero (0). For
nonblock-structured devices (e.g., terminals), logical blocks are not
addressable.
Virtual I/O concerns reading and writing data to open files. In this
case, the executive maps virtual blocks into logical blocks. For
file-structured devices (disks or DECtapes) virtual blocks are logical
blocks, but are numbered starting from one (1) and are relative to the
file rather than to the device. For nonfile-structured devices, the
mapping from virtual block to logical block is direct.

1.3

RSX-llM DEVICES

The devices listed below are supported by RSX-llM.
Drivers are
supplied for each of these devices, and I/O operations for them are
described in detail in subsequent chapters of this manual.
1.

A variety of terminals, including the following:
• ASR-33 and ASR-35 Teletypes (1)
• KSR-33 and KSR-35 Teletypes (1)
• LA30 DECwriters (serial and parallel)
. LA36 DECwriter
. VT05B Alphanumeric Display Terminal
. VT50 Alphanumeric Display Terminal
. RT02 Data Entry Terminal
. RT02-C Badge Reader and Data Entry Terminal

(1) Teletype is a registered trademark of the Teletype Corporation.

1-2

RSX-llM INPUT/OUTPUT
These terminals are supported on the
line interfaces:

following

asynchronous

. DJll Asynchronous Communications Line Interface Multiplexer

. DHll and DHll-DMll-BB
Interface Multiplexer

Asynchronous

. DLll-A,
DLll-B,
DLll-C,
and
Communications Line Interfaces
2.

Communications
DLll-D

Asynchronous

A variety of disks, including the following:
. RFll/RSll Fixed-Head Disk
. RP11/RP02 Pack Disk
. RP11/RP03 Pack Disk
. RXll/RXOl Flexible Disk
. RP04 Pack Disk
. RS03 Fixed-Head Disk
. RS04 Fixed-Head Disk
. RK11/RK05 Cartridge Disk

TC11/TU56 DECtape

Two types of magnetic tape:
. TU16 Magnetic Tape
. TMll/TUlO Magnetic Tape

TAll Tape Cassette

Three line printers:
. LPll Line Printer
. LSll Line Printer
. LVll Line Printer

CRll Card Reader

Synchronous and asynchronous line interfaces:
. DLll-E Asynchronous Communication Line Interface
. DPll Synchronous Communication Line Interface
. DUll Synchronous Communication Line Interface

Two analog-to-digital converters:
. AFCll Analog-to-Digital Converter
. ADOl-D Analog-to-Digital Converter

1-3

Line

RSX-llM INPUT/OUTPUT
10.

UDCll Universal Digital Controller

11.

Laboratory Peripheral Systems
. ARll Laboratory Peripheral System
. LPSll Laboratory Peripheral System

12.

Paper Tape Devices
• PCll Paper Tape Reader/Punch
. PRll Paper Tape Reader

13.

1.4

ICS/ICR Industrial Control Local and Remote Subsystems

LOGICAL UNITS

This section describes the construction of the logical unit table
the use of logical unit numbers.

1.4.1

and

Logical Unit Number

A logical unit number or LON is a number which is associated with a
physical device unit during RSX-llM I/O operations. For example, LON
1 might be associated with one of the terminals in the system, LUNs 2,
3, 4, and 5 with DECtape drives, and LUNs 6, 7, and 8 with disk units.
The association is a dynamic one; each task running in the system can
establish its own correspondence between LUNs and physical device
units, and can change any LON-physical device unit association at any
time.
The flexibility of this association contributes heavily to
RSX-llM device independence.
A logical unit number is simply a short name used to represent a
logical unit-physical device unit association. Once the association
has been made, the LON provides a direct and efficient mapping to the
physical device unit, and eliminates the necessity to search the
device tables whenever the system encounters a reference to a physical
device unit.
The user should remember that, although a LON-physical device unit
association can be changed at any time, reassignment of a LON at run
time causes pending I/O requests for the previous LON assignment to be
cancelled.
It is the user's responsibility to verify that all
outstanding I/O requests for a LON have been serviced before that LON
is associated with another physical device unit.

1.4.2

Logical Unit Table

There is one logical unit table (LOT} for each task running in an
RSX-llM system.
This table is a variable-length block contained in
the task header. Each LOT contains sufficient 2-word entries for the
number of logical units specified by the user at task build time.

1-4

RSX-llM INPUT/OUTPUT
Each entry or slot contains a pointer to the physical device unit
currently associated with that LUN. Whenever a user issues an I/O
request, RSX-llM matches the appropriate physical device unit to the
LUN specified in the call by indexing into the logical unit table by
the number supplied as the LUN. Thus if the call specifies 6 as the
LUN, RSX-llM accesses the sixth 2-word entry in the LUT and associates
the I/O request with the physical device unit to which the entry
points.
The number of LUN assignments valid for a task ranges from
zero to 255, but cannot be greater than the number of LUNs specified
at task build time.
Figure 1-1 illustrates a typical logical unit table.
Number of LUNs

l
LUN 1

1-----1------- -

LUN 2

LUN 3

UCB

)

(

UCB

)

---

0
1------

- -

LUN 4

( ____
uc_s)
Figure 1-1
Logical Unit Table

Word 1 of each active (assigned) 2-word entry in the logical unit
table points to the unit control block (UCB) of the physical device
unit with which the LUN is associated.
This linkage may
be
indirect - that is, the user may force redirection of references from
one unit to another unit via the MCR command, REDIRECT.
Word 2 of
each entry is reserved for mountable devices.

1.4.3

Changing LUN Assignments

Logical unit numbers have no significance until they are associated
with a physical device unit by means of one of the methods described
below:
1.

At task build time, the user can specify an ASG keyword
option, which associates a physical device unit with a
logical unit number referenced in the task being built.

The user or system operator can issue a REASSIGN command to
MCR;
this command reassigns a LUN to another physical
device unit and thus changes the LON-physical device unit
correspondence.
Note that this reassignment has no effect
on the in-core image of a task.

1-5

RSX-llM INPUT/OUTPUT
3.

1.5

At run time, a task can dynamically change a LUN assignment
by issuing the ASSIGN LUN system directive, which changes
the association of a LUN with a physical device unit during
task execution.

ISSUING AN I/O REQUEST

User tasks perform I/O in the RSX-llM system by submitting requests
for I/0 service in the form of QIO system directives. See Chapter 2
of the RSX-llM Executive Reference Manual for a complete description
of RSX-llM system directives.
0

In RSX-llM, as in most multiprogramming systems, tasks do not normally
access
physical
device units directly.
Instead, they utilize
input/output services provided by the Executive, since it
can
effectively multiplex the use of physical device units over many
users. The RSX-llM Executive routes I/O requests to the appropriate
device driver and queues them according to the priority of the
requesting task.
I/O operations proceed concurrently with other
activities in an RSX-llM system.
After an I/O request has been queued, the system does not wait for the
operation to complete. If at any time the user task which issued the
QIO request cannot proceed until the I/O operation has completed, it
should specify an event flag (see sections 1.5.l and 1.5.2) in the QIO
request and should issue a WAITFOR system directive which specifies
the same event flag at the point where synchronization must occur.
The task then waits for completion of I/O by waiting for the specified
event flag to be set.
Each QIO directive must supply sufficient information to identify and
queue the I/O request. The user may also want to include parameters
to receive error or status codes and to specify the address of an
asynchronous system trap service routine.
Certain types of I/O
operations require the specification of device-dependent information
as well. Typical QIO parameters are the following:
. I/O function to be performed
• Logical unit number associated with the physical device
to be accessed
. Optional event flag number for synchronizing
processing

I/O

unit

completion

. Optional address of the I/O status block to which information
indicating successful or unsuccessful completion is returned
. Optional address of an asynchronous system trap service
routine to be entered on completion of the I/O request
. Optional device- and function-dependent parameters specifying
such items as the starting address of a data buffer, the size
of the buffer, and a block number
A set of system macros which facilitate the issuing of QIO directives
is supplied with the RSX-llM system. These macros, which reside in
the System Macro Library (SY: [l,l]RSXMAC.SML), must be made available
to the source program by means of the MACR0-11 Assembler directive

1-6

RSX-llM INPUT/OUTPUT
.MCALL. The function of .MCALL is described in section 1.7.3.
Several of the first six parameters in the QIO directive are optional,
but space for these parameters must be reserved.
During expansion of a QIO macro, a value of zero is defaulted for all
null
(omitted)
parameters.
Inclusion
of
the
deviceand
function-dependent parameters depends on the physical device unit and
function specified.
If the user wanted to specify only an I/O
function code, a LON, and an address for an asynchronous system trap
service routine, the following might be issued:
QIO$C

IO.ATT,6,,,,ASTOX

where IO.ATT is the I/O function code for attach, 6 is the LON, ASTOX
is the AST address, and commas indicate null arguments for the event
flag number, the request priority, and the address of the I/O status
biock.
No additional device- or function-dependent parameters are
required for an attach function. The C form of the QIO$ macro is used
here and in most of the examples included in Chapter 1. Section 1.7
describes the three legal forms of the macro.
For convenience, any comma may be omitted if no parameters appear to
the right of it.
The command above could therefore be issued as
follows, if the asynchronous system trap was not desired.
QIO$C

IO. ATT, 6

All extra commas have been dropped. If, however, a parameter appears
to the right of any place-holding comma, that comma must be retained.

1.5.l

QIO Macro Format

The arguments for a specific QIO macro call may be different for each
I/O device accessed and for each I/O function requested. The general
format of the call is, however, common to all devices and is as
follows:
QIO$C

fnc,lun, [efn], [pri], [isb], [ast], [<pl,p2, ... ,p6>]

where brackets ([]} enclose optional or function-dependent parameters.
If function-dependent parameters <pl, .•. ,p6> are required, these
parameters must be enclosed within angle brackets (<>}. The following
paragraphs summarize the use of each QIO parameter. Section 1.7
discusses different forms of the QIO$ macro itself.
The fnc parameter is a symbolic name representing the I/O function
be performed. This name is of the form:

IO.xxx
where xxx identifies the particular I/O operation. For example, a QIO
request to attach the physical device unit associated with a LON
specifies the function code:
IO.ATT
A QIO request to cancel (or kill) all I/O requests for a specified LUN
begins in the following way:
QIO$C

IO.KIL, ...

1-7

RSX-llM INPUT/OUTPUT
The fnc parameter specified in the QIO request is stored internally as
a function code in the high-order byte and modifier bits in the
low-order byte of a single word. The function code is in the range
zero through 31 and is a binary value supplied by the system to match
the symbolic name specified in the QIO request.
The correspondence
between global symbolic names and function codes is defined in the
system object module library. Local symbolic definitions may also be
obtained via the FILIO$ and SPCIO$ macros which reside in the System
Macro Library and are summarized in Appendix A.
Several similar
functions may have identical function codes, and may be distinguished
only by their modifier bits. For example, the DECtape read logical
forward and read logical reverse functions have the same function
code. Only the modifier bits for these two operations are stored
differently.
The lun parameter represents the logical unit number (LUN) of the
associated physical device unit to be accessed by the I/O request.
The association between the physical device unit and the LUN is
specific to the task which issues the I/O request, and the LUN
reference is usually device-independent. An attach request to the
physical device unit associated with LUN 14 begins in the following
way:
QIO$C

IO.ATT,14., .•.

Because each task has its own logical unit table (LUT)
in which the
physical device unit-LUN correspondences are established, the legality
of a lun parameter is specific to the task which includes this
parameter in a QIO request.
In general, the LUN must be in the
following range:
0 ~ LUN ~length of task's LUT (if nonzero)

The number of LUNs specified in the logical unit table of a particular
task cannot exceed 255.
The efn parameter is a number representing the event flag to be
associated with the I/O operation. It may optionally be included in a
QIO request. The specified event flag is cleared when the I/0 request
is queued and is set when the I/O operation has completed. This
allows the task to use the WAITFOR system directive to synchronize I/O
programming by suspending execution to wait for an I/O operation to
complete and efn to be set.
However, if the task continues to
execute, it may test the event flag whenever it chooses by using the
READ ALL EVENT FLAGS system directive. If the user specifies an event
flag number, this number must be in the range 1 through 64. If an
event flag specification is not desired, efn can be omitted or can be
supplied with a value of zero. Event flags 1 through 32 are local
(specific to the issuing task); event flags 33 through 64 are global
(shared by all tasks in the system). Flags 25 through 32 and 57
through 64 are reserved for use by system software.
Within these
bounds, the user can specify event flags as desired to synchronize I/O
completion and task execution. Section 1.5.2 provides a more detailed
explanation of event flags and significant events.
The optional pri parameter is supplied only to make RSX-llM QIO
requests compatible with RSX-llD.
A specific priority cannot be
associated solely with the I/O request specified in the QIO macro
call.
An RSX-llM I/O request automatically assumes the priority of
the requesting task. For consistency with RSX-llD, it is recommended
that pri be valid, but the user should be aware that RSX-llM does not

1-8

RSX-llM INPUT/OUTPUT
use this specification in any way.
RSX-llD priorities must be in
range 1 through 250, and zero can be supplied to indicate the priority
of the requesting task. A value of zero or a null specification is
recommended for all RSX-llM use.
The optional isb parameter identifies the address of the I/O status
block (I/O status double-word) associated with the I/O request. This
block is a 2-word array in which a code representing the final status
of the I/O request is returned on completion of the operation. This
code is a binary value that corresponds to a symbolic name of the form
IS.xxx (for successful returns) or IE.xxx (for error returns). The
binary error code is returned to the low-order byte of the first word
of the status block.
It can be tested symbolically, by name. For
example, the symbolic status IE.BAD is returned if a bad parameter is
encountered.
The following illustrates the examination of the I/O
status block, IOST, to determine if a bad parameter has been detected.
QIO$C
WTSE$C

IO.ATT,14.,2,,IOST, ..•
2

CMPB
BEQ

#IE.BAD,IOST
ERROR

The correspondence between global symbolic names and I/O completion
codes is defined in the system object module library. Local symbolic
definitions, which are summarized in Appendix B, may also be obtained
via the IOERR$ macro which resides in the System Macro Library.
Certain device-dependent information is returned to the high-order
byte of the first word of isb on completion of the I/O operation. If
a read or write operation is successful, the second word is also
significant.
For example, in the case of a read function on a
terminal, the number of bytes typed before a carriage return is
returned in the second word of isb. If a Magtape unit is the device
and a write function is specified, this number represents the number
of bytes actually transferred. The status block can be omitted from a
QIO request if the user does not intend to test for successful
completion of the request.
The optional ast parameter specifies the address of a service routine
to be entered when an asynchronous system trap occurs. Section 1.5.3
discusses the use of asynchronous system traps, and section 2.2.5 of
the RSX-llM Executive Reference Manual describes traps in detail. If
the user wants to interrupt his task to execute special code on
completion of an I/O request, an asynchronous system trap routine can
be specified in the QIO request. When the specified I/O operation
completes, control branches to this routine at the software priority
of the requesting task. The asynchronous code beginning at address
ast is then executed, much as an interrupt service routine would be.
If the user does not want to perform asynchronous processing, the ast
parameter can be omitted or a value of zero specified in the QIO macro
call.
The additional QIO parameters, <pl,p2, •.. ,p6>, are dependent on the
particular function and device specified in the I/O request. Between
zero and six parameters can be included, depending on the particular
I/O function.
Rules for including these parameters and legal values
are described in subsequent chapters of this manual.

1-9

RSX-llM INPUT/OUTPUT
1.5.2

Significant Events

"Significant event" is a term used in real-time systems to indicate a
change in system status. In RSX-llM, a significant event is declared
when an I/O operation completes.
This signals the system that a
change in status has occurred and indicates that the Executive should
review the eligibility of all tasks in the system to determine which
task should run next. The use of significant events helps cooperating
tasks in a real-time system to communicate with each other and thus
allows these tasks to control their own sequence of execution
dynamically.
Significant events are normally set by system directives, either
directly or indirectly, by completion of a specified function. Event
flags associated with tasks may be used to indicate which significant
event has occurred. Of the 64 event flags available in RSX-llM, the
flags numbered 1 through 32 are local to an individual task and are
set or reset only as a result of that task's operation. The event
flags numbered 33 through 64 are common to all tasks.
Flags 25
through 32 and 57 through 64 are reserved for RSX-llM system software
use.
An example of the use of significant events follows. A task issues a
QIO directive with an efn parameter specified. A WAITFOR directive
follows the QIO and specifies as an argument the same event flag
number.
The event flag is cleared when the I/O request is queued by
the Executive, and the task is suspended when it executes the WAITFOR
directive until the event flag is set and a significant event is
declared at the completion of the I/O request. The task resumes when
the appropriate event flag is set, and execution resumes at the
instruction following the WAITFOR directive. During the time that the
task is suspended, tasks with priorities lower than that of the
suspended task have a chance to run, thus increasing throughput in the
system.

1.5.3

System Traps

System traps are used to interrupt task execution and to cause a
transfer of control to another memory location for special processing.
Traps are handled by the RSX-llM Executive and are relevant only to
the task in which they occur.
To use a system trap, a task must
contain a trap service routine which is automatically entered when the
trap occurs.
There are two types of system traps - synchronous and asynchronous.
Both are used to handle error or event conditions, but the two traps
differ in their relation to the task which is running when they are
detected.
Synchronous traps signal error conditions within the
executing task. If the same instruction sequence were repeated, the
same synchronous trap would occur.
Asynchronous traps signal the
completion of an external event such as an I/O operation.
An
asynchronous system trap (AST) usually occurs as the result of the
initiation or completion of an external event rather than a program
condition.
The Executive queues ASTs in a first-in-first-out queue for each task
and monitors all asynchronous service routine operations. Because
asynchronous traps are the end result of I/0-related activity, they
cannot be controlled directly by the task which receives them.

1-10

RSX-llM INPUT/OUTPUT
However, the task may, under certain circumstances, block honoring an
AST to prevent simultaneous access to a critical data region. When
access to the critical data region has been completed, the queued ASTs
may again be honored. The DSAR$S (DISABLE AST RECOGNITION) and ENAR$S
(ENABLE AST RECOGNITION) system directives provide the mechanism for
accomplishing this.
An example of an asynchronous trap condition is
the completion of an I/O request. The timing of such an operation
clearly cannot be predicted by the requesting task. If an AST service
routine is not specified in an I/O request, a trap does not occur and
normal task execution continues.
Asynchronous system traps associated with I/O requests enable the
requesting task to be truly event-driven. The AST service routine
contained in the initiating task is executed as soon as possible,
consistent with the system's priority structure. The use of the AST
routine to service I/O related events provides a response time which
is considerably better than a polling mechanism, and provides for
better overlap processing than the simple QIO and WAITFOR sequence.
Asynchronous system traps also provide an ideal mechanism for use in
multiple buffering of I/O operations.
All ASTs are inserted in a first-in-first-out queue on a per task
basis as they occur (i.e., the event which they are to signal has
expired). They are effected one at a time whenever the task does not
have ASTs disabled and is not already in the process of executing an
AST service routine. The process of effecting an AST involves storing
certain information on the task's stack, including the task's four
WAITFOR mask words, the Directive Status Word (DSW), the PS, the PC
and
any trap dependent parameters.
The task's general-purpose
registers RO-RS are not saved and thus it is the responsibility of the
AST service routine to save and restore the registers it uses. After
an AST is processed, the trap-dependent parameters (if any) must be
removed from the task's stack and au AST SERVICE EXIT directive
executed. The ASTX$S macro described in section 1.7.6 of this manual
is used to issue the AST SERVICE EXIT directive. On AST service exit,
control is returned to another queued AST, the executing task, or
another task which has been waiting to run. Section 2.2.5 of the
RSX-llM Executive Reference Manual describes in detail the purpose of
AST service routines and all system directives used to handle them.

1.6

DIRECTIVE PARAMETER BLOCKS

A directive parameter block (DPB) is a fixed-length area of contiguous
memory which contains the arguments specified in a system directive
macro call. The DPB for a QIO directive has a length of 12 words. It
is generated as the result of the expansion of a QIO macro call. The
first byte of the DPB contains the directive identification code (DIC)
- always 1 for QIO.
The second byte contains the size of the
directive parameter block in words - always 12 for QIO.
During
assembly of a user task containing QIO requests, the MACR0-11
Assembler generates a directive parameter block for each I/O request
specified in a QIO macro call. At run time, the Executive uses the
arguments stored in each DPB to create, for each request, an I/O
packet in system dynamic storage. The packet is entered by priority
into a queue of I/O requests for the specified physical device unit.
This queue is created and maintained by the RSX-llM Executive and is
ordered by the priority of the tasks which issued the requests.
The
I/O drivers examine their respective queues for the I/O request with
the highest priority capable of being executed.
This request is

1-11

RSX-llM INPUT/OUTPUT
de-queued (removed from the queue) and the I/O operation is performed.
The process is then repeated until the queue is emptied of all
requests.
After the I/O request has been completed, the Executive declares a
significant event and may set an event flag, cause a branch to an
asynchronous system trap service routine, and/or return the I/O
status, depending on the arguments specified in the original QIO macro
call. Figure 1-2 illustrates the layout of a sample DPB.

Word

size of DPB +

Byte

DIC for QIO
directive

I/O function

lun

logical unit number

efn

event flag number

fnc

~?_,_,,

modifiers
,

~reserv~

LLL.L.L.L...L :il

priority

pri

isb

address of I/O
status block

ast

address of
asynchronous trap
service routine

devicedependent
parameters

Figure 1-2
QIO Directive Parameter Block
1.7

I/0-RELATED MACROS

There are several system macros supplied with the RSX-llM system which
are used to issue and return information about I/O requests. These
macros reside in the System Macro Library and must be made available
during assembly via the MACR0-11 assembler directive .MCALL.
There are three distinct forms of most of the system directive macros
discussed in this section. The following list summarizes the forms of
QIO$, but the characteristics of each form also apply to ALUN$, GLUN$,
and other system directive macros described below.
1.

QIO$ generates a directive parameter block for the I/O
request
at
assembly
time, but does not provide the
instructions necessary to execute the request. This form of
the request is actually executed using the DIR$ macro.

QIO$S generates a directive parameter block for the I/O
request on the stack, and also generates code to execute the
request. This is a useful form for reentrant, sharable code
since the DPB is generated dynamically at execution time.

1-12

RSX-llM INPUT/OUTPUT
3.

QIO$C generates a directive parameter block for the I/O
request at assembly time, and also generates code to execute
the request. The DPB is generated in a separate program
section called $DPB$$.
This approach incurs little system
overhead and is useful when an I/O request is executed from
only one place in the program.

Parameters for both the QIO$ and QIO$C forms of the macro must be
valid expressions to be used in assembler data-generating directives
such as .WORD and .BYTE. Parameters for the QIO$S form must be valid
source
operand
address
expressions
to be used in assembler
instructions such as MOV and MOVB. The following example references
the same parameters in the three distinct forms of the macro call.
QIO$

IO.RLB,6,2,,,ASTOl, ••.

QIO$C

IO.RLB,6,2,,,ASTOl, .•.

QIO$S

#IO.RLB,#6,#2,,,#ASTOl, •..

Only the QIO$S form of the macro produces the DPB dynamically.
The
other
two
forms
generate
the
DPB
at assembly time.
The
characteristics and use of these different forms are described in
greater detail in the RSX-llM Executive Reference Manual.
The following Executive directives and assembler macros are
in this section:

1.7.1

QIO$, which is used to request an I/O
parameters for that request.

DIR$, which specifies the address of a directive parameter
block as its argument, and generates code to execute the
directive.

.MCALL, which is used to make available from the System Macro
Library all macros referenced during task assembly.

ALON$, which is used to associate a logical unit number
a physical device unit at run time.

GLUN$, which requests that the information about a physical
device unit associated with a specified LON be returned to a
user-specified buffer.

ASTX$S, which is used to terminate
execution
asynchronous system trap (AST) service routine.

WTSE$, which instructs the system to suspend execution of the
issuing task until a specified event flag is set.

The QIO$ Macro:

operation

described
and

supply

with

Issuing an I/O Request

As described in section 1.7, there are three distinct forms of the
QIO$ macro.
QIO$S generates a DPB for the I/O request on the stack,
and also generates code to execute the request. QIO$C generates a DPB
and code, but the DPB is generated in a separate program section.
QIO$ generates only the DPB for the I/O request.
This form of the
macro call is used in conjunction with DIR$ (see section 1.7.2) to

1-13

RSX-llM INPUT/OUTPUT
execute an I/O request. In the following example, the DIR$ macro
actually generates the code to execute the QIO$ directive. It
provides no QIO parameters of its own, but references the QIO
directive parameter block at address QIOREF by supplying this label as
an argument.
QIOREF:

QIO$

IO.RLB,6,2,,,ASTOl, •••

CREATE QIO DPB

READl:

DIR$

#QIOREF

ISSUE I/0 REQUEST

READ2:

DIR$

#QIOREF

ISSUE I/0 REQUEST

1.7.2

The DIR$ Macro:

Executing a Directive

The DIR$ (execute directive) macro has been implemented to allow a
task to reference a previously defined directive parameter block
without requiring that it specify all of the parameters of that macro
again. It is issued in the form:
DIR$
where:

1.7.3

[addr] [,err]

addr

is the address of a directive parameter block to be
used in the directive. If addr is not included, the
DPB itself or the address of the DPB is assumed to
already be on the stack.

err

is an optional argument which specifies the address of
an error routine to which control branches if the
directive is rejected. The branch occurs via a JSR PC,
err.

The .MCALL Directive:

Retrieving System Macros

.MCALL is a MACR0-11 assembler directive which is used to retrieve
macros from the System Macro Library (SY: [l,l]RSXMAC.SML) for use
during assembly. It must be included in every user task which invokes
system macros.
.MCALL is usually placed at the beginning of a user
task source module and specifies, as arguments in the call, all system
macros which must be made available from the library.
The following example illustrates the use of this directive:
.MCALL

QIOREF: QIO$

QIO$,QIO$S,DIR$,WTSE$S

MAKE MACROS AVAILABLE

IO.RLB,6,2,,,ASTOl, • . .

CREATE ONLY QIO DPB

1-14

RSX-llM INPUT/OUTPUT
READl:

DIR$

READ2:

QIO$S

#QIOREF

ISSUE I/O REQUEST

#IO.ATT,#14.,#8.,,,#AST02

CREATE DPB ON STACK
AND ISSUE REQUEST

As many macro references as can fit on a line can be included in a
single .MCALL directive.
There is no limit to the number of .MCALL
directives that can be specified.

1.7.4

The ALUN$ Macro:

Assigning a LON

The ASSIGN LUN macro is used to associate a logical unit number with a
physical device unit at run time. All three forms of the macro call
may be used. ASSIGN LUN does not request I/O for the physical device
unit, nor does it attach the unit for exclusive use by the issuing
task. It simply establishes a LON-physical device unit relationship,
so that when the task requests I/O for that particular LUN, the
associated physical device unit is referenced. The macro is issued
from a MACR0-11 program in the following way:
ALUN$
where:

lun,dev,unt

lun

is the logical unit number to be
specified physical device unit.

associated

dev

is the device name of the physical device or a logical
device name assigned to a physical device (see MCR ASN
command) •

unt

is the unit number of that device specified above.

For example, to associate LUN 10 with terminal unit 2,
macro call could be issued by the task:
ALUN$C

the

with

the

following

10.,TT,2

A unit number of 0 represents unit O for multi-unit devices such as
disk, DECtape, or terminals; it indicates the single available unit
for devices without multiple units, such as card readers and line
printers.
The following
list
contains
physical
alphabetically, that may be included as
standard devices supported by RSX-llM.

device
names,
dev parameters

Device
AD

ADOl-D Analog-to-Digital Converter

AFCll Analog-to-Digital Converter

ARll Laboratory Peripheral System

1-15

listed
for all

RSX-llM INPUT/OUTPUT
CR

CRll Card Reader

TAll/TU60 Tape Cassette

RP04 Pack Disk

RFll/RSll Fixed-Head Disk

RK11/RK05 Cartridge Disk

RP02/RP03 Pack Disk

RS03 and RS04 Fixed-Head Disks

TC11/TU56 DECtape

RXll/RXOl Flexible Disk

ICS/ICR Industrial Control Local and Remote Subsystems

LPll, LSll, and LVll Line Printers

LPSll Laboratory Peripheral System

TU16 Magnetic Tape

TMll/TUlO Magnetic Tape

PCll Paper Tape Punch

PCll or PRll Paper Tape Reader

Terminals

UDCll Universal Digital Controller

DAll-B Parallel Unibus Link

DLll-E Asynchronous Communication Line Interface

DPll Synchronous Communication Line Interface

DQll Synchronous Communication Line Interface

DUll Synchronous Communication Line Interface

A pseudo-device is a logical device which can normally be redirected
by the operator to another physical device unit at any time, without
requiring changes in programs which reference the pseudo-device.
Dynamic redirection of a physical device unit affects all tasks in the
system; reassignment by means of the MCR REASSIGN command affects
only one task. The following pseudo-devices are supported by RSX-llM:
Device
CL

Console listing, normally the line printer

Console output, normally the main operator's console

Pseudo-input terminal, normally the terminal from which
a task was requested

1-16

RSX-llM INPUT/OUTPUT
SY

System default device, normally the disk from which the
system was bootstrapped

The pseudo-device TI cannot be redirected, since such redirection
would have to be handled on a per task rather than a system wide basis
(i.e., change the TI device for one task without affecting the TI
assignments for other tasks).
Logical devices are SYSGEN options of RSX-llM that allow the user to
assign logical names to physical devices by means of the MCR command
ASN.
See the RSX-llM Operator's Procedures Manual for a full
description.
The example included below illustrates the use of the three
the ALUN$ macro.

forms

DATA DEFINITIONS
ASSIGN: ALUN$

10.,TT,2

GENERATE DPB

EXECUTABLE SECTION

1.7.5

DIR$

#ASSIGN

EXECUTE DIRECTIVE

ALUN$C

10.,TT,2

GENERATE DPB IN SEPARATE PROGRAM
SECTION, THEN GENERATE CODE TO
EXECUTE THE DIRECTIVE

ALUN$S

#10.,#"TT,#2

GENERATE DPB ON STACK, THEN
EXECUTE DIRECTIVE

The GLUN$ Macro:

Retrieving LUN Information

The GET LUN INFORMATION macro requests that information about a
LUN-physical device unit association be returned in a 6-word buffer
specified by the issuing task. All three forms of the macro call may
be used. It is issued from a MACR0-11 program in the following way:
GLUN$
where:

lun,buf

lun

is the logical unit number associated with the physical
device unit for which information is requested.

buf

is the 6-word buffer to which information is -returned.

1-17

RSX-llM INPUT/OUTPUT
For example, to request information on the disk unit
LON 8, the following call is issued:
GLUN$C

associated

with

8., IOBUF

The 6-word buffer contains the following indicators on
the directive:

completion

Contents
Name of device associated with lun

0
1

Unit number of associated device

Driver flag value, indicating that the driver
is resident (always returned- as 128 (200
octal) in RSX-llM)
0

Unit record-oriented
device
reader, line printer) ( 1 = yes)

Carriage-control device (e.g.,
terminal) (1 = yes)

Terminal device (1

yes)

Directory
( 1 = yes)

(e.g.,

Single directory device (l= yes)

Sequential device (1 = yes)

6-12

device

card

(e.g.,
line

printer,

disk)

DECtape,

Reserved

Device mountable as a communications channel
for Digital network support (e.g., DPll, DUll)
( 1 = yes)

Device mountable as a FILES-11
d i s k ) ( 1 = yes )

device

(e.g.,

Device mountable (logical OR of
14) ( 1 = yes)

bits

and

Undefined (included for RSX-llD compatibility)

Default buffer size for device
of line for terminal)

(e.g.,

The example included below illustrates the use of the three
the GLUN$ macro.

1-18

length

forms

RSX-llM INPUT/OUTPUT
DATA DEFINITIONS
GETLUN: GLUN$

6,DSKBUF

GENERATE DPB

EXECUTABLE SECTION

1.7.6

DIR$

#GETLUN

EXECUTE DIRECTIVE

GLUN$C

6,DSKBUF

GENERATE DPB IN SEPARATE PROGRAM
SECTION, THEN GENERATE CODE TO
EXECUTE THE DIRECTIVE

GLUN$S

#6,#DSKBUF

GENERATE DPB ON STACK, THEN
EXECUTE DIRECTIVE

The ASTX$S Macro:

Terminating AST Service

The AST SERVICE EXIT macro is used to terminate execution of an
asynchronous system trap (AST) service routine. Only the ASTX$S form
of this macro is provided; ASTX$ and ASTX$C are unsupported forms of
the macro call. The macro is issued in the following way:
ASTX$S
where:

err

[err]

is an optional argument which specifies the address of
an error routine to which control branches if the
directive is rejected.

On completion of the operation specified in this macro call, if
another AST is queued and asynchronous system traps have not been
disabled, then the next AST is immediately entered.
Otherwise, the
task's state before the AST was entered is restored (it is the AST
service routine's responsibility to save and restore the registers it
uses) •

1.7.7

The WTSE$ Macro:

Waiting for an Event Flag

The WAIT FOR SINGLE EVENT FLAG macro instructs the system to suspend
execution of the issuing task until the event flag specified in the
macro call is set. This macro is extremely useful in synchronizing
activity on completion of an I/O operation. All three forms of the
macro call may be used. It is issued as follows:
WTSE$
where:

efn

efn
is the event flag number

1-19

RSX-llM INPUT/OUTPUT
WTSE$ causes the task to be blocked from execution until the specified
event flag is set.
Frequently, an efn parameter is also included in a
QIO$ macro call, and the event flag is set on completion of the
I/O
operation specified
in that call.
The following example illustrates
task blocking until the setting of the specified event flag occurs.
This example also illustrates the use of the three forms of the macro
call.

DATA DEFINITIONS
WAIT:
IOSB:

WTSE$
.BLKW

GENERATE DPB
I/O STATUS BLOCK

EXECUTABLE SECTION

1.8

ASSIGN LUN 14 TO MAGTAPE UNIT ZERO
ATTACH DEVICE
EXECUTE WAITFOR DIRECTIVE

ALUN$S
QIO$C
DIR$

#14.,#"MM
IO.ATT,14.,5
#WAIT

QIO$S

#IO.RLB,#14.,#2,,#IOSB,#ASTX,<#BUF,#80.>
; READ RECORD, USE EFN2

WTSE$S

QIO$C

IO.WLB,14.,3,,IOSB,ASTOl,<BUF,80.>
; WRITE RECORD, USE EFN3

WTSE$C
QIO$C

3
IO.DET,14.

WAIT FOR READ TO COMPLETE

WAIT FOR WRITE TO COMPLETE
DETACH DEVICE

STANDARD I/O FUNCTIONS

There are a large number of input/output operations that can be
specified by means of the QIO directive. A particular operation can
be requested by including the appropriate function code as the first
parameter of a QIO macro call. Certain functions are standard.
These
functions are almost totally device-independent and can thus be
requested for nearly every device described in this manual. Others
are device-dependent and are specific to the operation of only one or
two I/O devices.
This section summarizes the function codes and
characteristics of the following device-independent I/O operations:

1-20

RSX-llM INPUT/OUTPUT
. attach to an I/O device
detach from an I/O device
. cancel I/O requests
. read a logical block
. read a virtual block
. write a logical block
. write a virtual block
For certain physical device units discussed in this manual, a standard
I/O function may be described as being a NOP. This means that no
operation is performed as a result of specifying the function, and an
I/O status code of rs.sue is returned in the I/O status block
specified in the QIO macro call.
In the following descriptions and in formats shown in subsequent
chapters, the five parameters represented by the ellipsis ( ... ) are as
explained in section 1.5.1.

1.8.1

IO.ATT:

Attaching to an I/O Device

The function code IO.ATT is specified by a user task when that task
requires exclusive use of an I/O device.
This function code is
included as the first parameter of a QIO macro call in the following
way:
QIO$C

IO.ATT, .••

Successful completion of an IO.ATT request causes the specified
physical device unit to be dedicated for exclusive use by the issuing
task. This enabies the task to process input or output in an unbroken
stream and is especially useful on sequential, nonfile-oriented
devices such as terminals, card readers, and line printers.
An
attached physical device unit remains under control of the issuing
task until it is explicitly detached by that task. The same LUN must
be specified for both the attach and detach functions.
While a physical device unit is attached, the I/O driver for that unit
dequeues only I/O requests issued by the task that issued the attach.
Thus, a request to attach a device unit already attached by another
task will not be processed until the attachment is broken and no
higher priority request exists for the unit. A LUN that is associated
with an attached physical device unit may not be reassigned by means
of an ASSIGN LUN directive.
If the task which issued an attach function exits or is aborted before
it issues a corresponding detach, the RSX-llM Executive automatically
detaches the physical device unit.

1-21

RSX-llM INPUT/OUTPUT
1.8.2

IO.DET:

Detaching from an I/O Device

The function code IO.DET is used to detach a physical device unit
which has been previously attached by means of an IO.ATT request for
exclusive use of the issuing task. This function code is included as
the first parameter of a QIO macro call in the following way:
QIO$C

IO.DET, •.•

The LON specifications of both IO.ATT and IO.DET must be the same, as
in the following example, which also illustrates the use of "S" forms
of several macro calls.

LOOP:

1.8.3

.MCALL
ALUN$S

ALUN$S,QIO$S
#14., # 11 CR

ASSOCIATE CARD READER WITH LON 14

QIO$S

UO.ATT I #14.

ATTACH CARD READER

QIO$S

#I 0 • RLB I # 14 • I • • •

QIO$S

#IO. DET I U4.

IO.KIL:

Canceling I/O Requests

READ CARD

DETACH CARD READER

The function code IO.KIL is issued by a task to cancel all of that
task's I/O requests for a particular physical device unit, including
all pending and active requests. This results in the status code
IE.ABO being returned in the I/O status block and the event flag being
set (if specified) for the respective requests, but does not initiate
any asynchronous system trap (AST) service routine that may have been
specified. Whether the current request is actually cancelled depends
on the device.
Because file-structured devices (disk and DECtape}
operate quickly, IO.KIL is a NOP for these devices and simply causes
the return of IS.SOC in the I/O status block.
This function code is included as the first parameter of a
in the following way:
QIO$C

QIO

macro

IO. KIL I • • •

IO.KIL is useful in such special cases as canceling an I/O request on
a physical device unit from which a response is overdue (i.e., a read
on a terminal).

1.8.4

IO.RLB:

Reading a Logical Block

The function code IO.RLB is specified by a task to read a block of
data from the physical device unit specified in the macro call. This
function code is included as the first parameter of a QIO macro in the
following way:
QTO$C

IO.RLB; ... ,<stadd,size,pn>

1-22

RSX-llM INPUT/OUTPUT
where:

stadd

is the starting address of the data buffer.

size

is the data buffer size in bytes.

1.8.5

IO.RVB:

represents one to four optional parameters, used to
specify such additional information as block numbers
for certain devices.

Reading a Virtual Block

The function code IO.RVB is used to read a virtual block of data from
the physical device unit specified in the macro call. A "virtual"
block indicates a relative block position within a file and is
identical to a "logical" block for such sequential devices as
terminals and card readers. It is recommended that all tasks use
virtual rather than logical reads.
However, if a virtual read is
issued for a file-structured device (disk or DECtape) , the user must
ensure that a file is open on the specified physical device unit.
This function code is included as the first parameter of a QIO macro
call in the following way:
QIO$C
where:

stadd

is the starting address of the data buffer.

size

is the data buffer size in bytes.

1.8.6

IO.RVB, •.• ,<stadd,size,pn>

IO.WLB:

represents one to four optional parameters, used to
specify such additional information as block numbers
for certain devices.

Writing a Logical Block

The function code IO.WLB is specified by a task to write a block of
data to the physical device unit specified in the macro call. This
function code is included as the first parameter of a QIO macro call
in the following way:
QIO$C
where:

stadd

is the starting address of the data buffer.

size

is the data buffer in bytes.

1.8.7

IO.WLB, •.• ,<stadd,size,pn>

IO.WVB:

represents one to four optional parameters, used to
specify such additional information as block numbers or
format control characters for certain devices.

Writing a Virtual Block

The function code IO.WVB is used to write a virtual block of data to a
physical device unit.
A "virtual" block indicates a relative block
position within a file and is identical to a "logical" block for such
sequential devices as terminals and line printers. It is recommended
that all tasks use virtual rather than logical writes. However, if a

1-23

RSX-llM INPUT/OUTPUT
virtual write is issued for a file-structured device (disk or
DECtape), the user must ensure that a file is open on the specified
physical device unit.
This function code is included as the first
parameter of a QIO macro call in the following way:
QIO$C
where:

stadd

is the starting address of the data buffer.

size

is the data buffer size in bytes.

1.9

IO.WVB, ... ,<stadd,size,pn>

represents one to four optional parameters, used to
specify such additional information as block numbers or
format control characters for certain devices.

I/O COMPLETION

When an I/O request has been completed, either successfully or
unsuccessfully, one or more actions may be taken by the Executive.
Selection of return conditions depends on the parameters included in
the QIO macro call. There are three major returns:
1.

A significant event is declared on completion of an I/O
operation.
If an efn parameter was included in the I/O
request, the corresponding event flag is set.

If an isb parameter was specified in the QIO macro call, a
code identifying the type of. success or failure is returned
in the low-order byte of the first word of the I/O status
block at the location represented by isb.
This status return code is of the form IS.xxx (success) or
IE.xxx (error).
For example, if the device accessed by the
I/O request is not ready, a status code of IE.DNR is returned
in isb. The section below (Return Codes) summarizes general
codes returned by most of the drivers described in this
manual.
If the isb parameter was omitted, the requesting task cannot
determine whether the I/0 request was successfully completed.
A carry clear return from the directive itself simply means
that the directive was accepted and the I/O request was
queued, not that the actual input/output operation was
successfully performed.

If an ast parameter was specified in the QIO macro call, a
branch to the asynchronous system trap (AST) service routine
which begins at the location identified by ast occurs on
completion of the I/O operation. See section 1.5.3 for a
detailed description of AST service routines.

1-24

RSX-llM INPUT/OUTPUT
1.10

RETURN CODES

There are two kinds of status conditions
RSX-llM when they occur in I/O requests:

recognized

. Directive conditions, which indicate
rejection of the QIO directive itself

the

and

handled

acceptance

. I/O status conditions, which indicate the success or
of the I/O operation
Directive conditions relevant to I/O operations may
the following:

indicate

failure
any

. directive acceptance
. invalid buffer specification

• invalid efn parameter
. invalid lun parameter
invalid DIC number or DPB size
. unassigned LUN
insufficient memory
A code indicating the acceptance or rejection of a directive is
returned to the directive status word at symbolic location $DSW. This
location can be tested to determine the type of directive condition.
I/O conditions indicate the success or failure of the I/O operation
specified in the QIO directive.
I/O driver errors include such
conditions as device not ready, privilege violation, file already
open, or write-locked device. If an isb parameter is included in the
QIO directive, identifying the address of a 2-word I/O status block,
an I/0 status code is returned in the low-order byte of the first word
of this block on completion of the I/O operation.
This code is a
binary value which corresponds to a symbolic name of the form IS.xxx
or IE.xxx. The low-order byte of the word can be tested symbolically,
by name, to determine the type of status return. The correspondence
between global symbolic names and directive and I/O completion status
codes is defined in the system object module library. Local symbolic
definitions may also be obtained via the DRERR$ and IOERR$ macros
which reside in the System Macro Library and are summarized in
Appendix B.
Binary values of status codes always have the following meaning:
Meaning

Code
Positive (greater than zero}

Successful completion

Zero

Operation still pending

Negative

Unsuccessful completion

A pending operation means that the I/O request is still in the queue
of requests for the respective driver, or the driver has not yet
completely serviced the request.

1-25

RSX-llM INPUT/OUTPUT
1.10.l

Directive Conditions

Table 1-1 summarizes the directive conditions which may be encountered
in QIO directives.
The acceptance condition is first, followed by
error codes indicating various reasons for rejection, in alphabetical
order.
Table 1-1
Directive Conditions
Code

rs.sue

Reason
Directive accepted
The first six parameters of the QIO directive were
valid, and sufficient dynamic memory was available to
allocate an I/O packet. The directive is accepted.

IE.ADP

Invalid address
The I/O status block or the QIO DPB was outside of the
issuing task's address space or was not aligned on a
word boundary.

IE.IEF

Invalid event flag number
The efn specification in a QIO directive was less
zero or greater than 64.

IE.ILU

than

Invalid logical unit number
The lun specification in a QIO directive was invalid
for the issuing task.
For example, there were only
five logical unit numbers associated with the task, and
the value specified for lun was greater than five.

IE.SOP

Invalid DIC number or DPB size
The directive identification code {DIC) or the size of
the directive parameter block (DPB) was incorrect; the
legal range for a DIC is from 1 through 127, and all
DIC values must be odd.
Each individual directive
requires a DPB of a certain size. If the size is not
correct for the particular directive, this code is
returned. The size of the QIO DPB is always 12 words.

IE.ULN

I Unassigned LUN
The logical unit number in the QIO directive was not
associated with a physical device unit. The user may
recover from this error by issuing a valid ASSIGN LUN
directive and then reissuing the rejected directive.

1-26

RSX-llM INPUT/OUTPUT
Table 1-1 (Cont.)
Directive Conditions
Code

Reason

IE. UPN

Insufficient dynamic memory
There was not enough dynamic memory to allocate an I/O
packet for the I/O request. The user can try again
later by blocking the task with a WAITFOR SIGNIFICANT
EVENT directive.
Note that WAITFOR SIGNIFICANT EVENT
is the only effective way for the issuing task to block
its execution, since other directives that could be
used for this purpose themselves require dynamic memory
for their execution (e.g., MARK TIME).

1.10.2

I/0 Status Conditions

The following list summarizes status codes which may be returned in
the I/O status block specified in the QIO directive on completion of
the I/O request. The I/O status block is a 2-word block with the
following format:
The low-order byte of the first word receives a status code
of the form IS.xxx or IE.xxx on completion of the I/O
operation.
The high-order byte of
the
first
word
is
usually
device-dependent;
in cases where the user might find
information in this byte helpful, this manual identifies that
information •
. The second word contains the number of bytes transferred or
processed if the operation is successful and involves reading
or writing.
If the isb parameter of the QIO directive is omitted, this information
is not returned.
The following illustrates a sample 2-word
completion of a terminal read operation:

I/O

status

block

Word 0
1

Number of bytes read

where -10 is the status code for IE.EOF (end of file). If this code
is returned, it indicates that input was terminated by typing CTRL/Z,
which is the end-of-file termination sequence on a terminal.
To test for a particular error condition, the user generally compares
the low-order byte of the first word of the I/O status block with a
symbolic value as in the following:
CMPB

#IE.DNR,IOSB

1-27

RSX-llM INPUT/OUTPUT
However, to test for certain types of successful completion of the I/O
operation, the entire word value must be compared. For example, if a
carriage return terminated a line of input from the terminal, a
successful completion code of IS.CR is returned in the I/O status
block. If an ESCape or ALTMODE character was the terminator, a code
of IS.ESC is returned. To check for either of these codes, the user
should first test the low-order byte of the first word of the block
for rs.sue and then test the fulf word for IS.CR or IS.ESC.
Note that all three of the following comparisons will test equal since
the low-order byte in all cases is +l.
CMP

#IS.CR,IOSB

CMP

#IS.ESC,IOSB

CMPB

#IS.SUC,IOSB

In the case of a successful completion where the carriage return is
the terminal indicator {IS.CR), the following illustrates the status
block:
1

word O
1

Byte

Number of bytes read

where 15 is the octal code for carriage return and +l
code for successful completion.

the

status

The codes described in Table 1-2 are general status codes which apply
to the majority of devices presented in subsequent chapters. Error
codes specific to only one or two drivers are described only in
relation to the devices for which they are returned. The list below
describes successful and pending codes first, then error codes in
alphabetical order.
Table 1-2
I/O Status Conditions
Reason

Code

rs.sue

I
IIS.PND

Successful completion
The I/O operation specified in the QIO directive was
completed successfully.
The second word of the I/O
status block can be examined to determine the number of
bytes processed, if the operation involved reading or
writing.
I/0 request pending
The I/O operation specified in the QIO directive has
not yet been executed. The I/O status block is filled
with zeros.

1-28

RSX-llM INPUT/OUTPUT
Table 1-2 (Cont.)
I/O Status Conditions
Reason

IE.ABO

Operation aborted
The specified I/O operation was cancelled via IO.KIL
while in progress or while still in the I/O queue.

IE.ALN

File already open
The task attempted to open a file on the physical
device unit associated with the specified LUN, but a
file has already been opened by the issuing task on
that LUN.

IE.BAD

Bad parameter
An illegal specification was supplied for one or more
of the device-dependent QIO parameters (words 6-11).
For example, a bad channel number or gain code was
specified
in
an
analog-to-digital converter I/O
operation.

IE.ELK

Illegal block number
An illegal block
number
was
specified
for
a
file-structured physical device unit.
This code is
returned, for example, if block 4800 is specified for
an RKOS disk, on which legal block numbers extend from
zero through 4799.

IE.BYT

Byte-aligned buffer specified.

IE.DAA

Byte alignment was specified for a buffer, but only
word alignment is legal for the physical device unit.
For example, a disk function requiring word alignment
was requested, but the buffer was aligned on a byte
boundary. Alternately, the length of a buffer was not
an appropriate multiple of bytes. For example, all
RP03 disk transfers must be an even multiple of four
bytes.
Device already attached
The physical device unit specified in an
IO.ATT
function was already attached to the issuing task.
This code indicates that the issuing task has already
attached the desired physical device unit, not that the
unit was attached by another task.

IE.DNA

Device not attached
IO.DET
The physical device unit specified in an
function was not attached to the issuing task. This
code has no bearing on the attachment status with
respect to other tasks.

1-29

RSX-llM INPUT/OUTPUT
Table 1-2 (Cont.)
I/O Status Conditions
Code
IE.DNR

Reason
Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.
This code is of ten returned as the result of an
interrupt timeout, that is, a "reasonable" amount of
time has passed, and the physical device unit has not
responded.

IE.EOF

End-of-file encountered
An end-of-file mark, record, or control
recognized on the input device.

IE.IFC

character

was

Illegal function
A function code was specified in an I/O request that
was illegal for the specified physical device unit.
This code is returned if the task attempts to execute
an illegal function or if, for example, a read function
is requested on an output-only device, such as the line
1
I printer.

IE.NLN

File not open
The task attempted to close a file on the physical
device unit associated with the specified LUN, but no I
file was currently open on that LUN.

IE.NOD

Insufficient buffer space
Dynamic storage space has been depleted, and there was
insufficient buffer space available to allocate a
secondary control block.
For example, if a task
attempts to open a file, buffer space for the window
and file control block must be supplied by
the
Executive.
This code is returned when there is not
enough space for such an operation.

IE.OFL

Device off-line
The physical device unit associated with the LUN
specified in the QIO directive was not on-line. When
the system was booted, a device check indicated that
this physical device unit was not in the configuration.

IE.OVR

Illegal read overlay request
A read overlay was requested and the physical device
unit specified in the QIO directive was not the
device unit from which the task was installed.
I physical
The read overlay function can only be executed on the
physical device unit from which the
task
image
containing the overlays was installed.

1-30

RSX-llM INPUT/OUTPUT
Table 1-2 {Cont.)
I/O Status Conditions
Code
IE.FRI

Reason
Privilege violation
The task which issued a request was not privileged to
execute that request. For example, for the UDCll and
LPSll, a checkpointable task attempted to connect to
interrupts
or
to execute a synchronous sampling
function.

IE.SPC

Illegal address space
The buffer requested for a read or write request was
partially or totally outside the address space of the
issuing task. Alternately a byte count of zero was
specified.

IE.VER

Unrecoverable error
After the system's standard number of retries have been
attempted upon encountering an error, the operation
I still could not be completed. This code is returned in
the case of parity, CRC, or similar errors.

IE.WLK

Write-locked device
The task attempted to write on a write-locked
unit.

j device

1-31

physical

CHAPTER 2
TERMINAL DRIVER

2.1

INTRODUCTION

The terminal driver provides support for a variety of terminal devices
under RSX-llM.
Table 2-1 summarizes the terminals supported, and
subsequent sections describe these devices in greater detail.
Table 2-1
Standard Terminal Devices
Model

Column Width

Character·set

Baud Range

ASR-33/35

llO

KSR-33/35

llO

LA30-P

300

LA30-S

ll0-300

LA36

80-132

64-96*

ll0-300

RT02

ll0-1200

RT02-C

110-1200

VT05B

64*

ll0-2400

VT50

ll0-9600

Where appropriate, terminals must be set to transmit only upper-case
alphabetic characters.
Input lines can be at most 80 bytes; longer
input lines are truncated.
The terminal driver
supports
the
communication
line
interfaces
summarized in Table 2-2.
These
interfaces are described
in
greater
detail
in
section 2.7.
Programming is identical for all.

* Both upper and lower case input are supported.

2-1

TERMINAL DRIVER
Table 2-2
Standard Communication Line Interfaces
Model

Type

DHll

16-line multiplexer or

DHll-DMll-BB

16-line multiplexer with modem control

DJll

16-line multiplexer

DLll-A/B/C/D

Single-line interfaces

2.1.1

ASR-33/35 Teletypes

The ASR-33 and ASR-35 Teletypes are asynchronous hard-copy
No paper tape reader or punch capability is supported.

2.1.2

terminals.

KSR-33/35 Teletypes

The KSR-33 and KSR-35 Teletypes are asynchronous, hard-copy terminals.

2.1.3

LA30 DECwriters

The LA30 DECwriter is an asynchronous, hard-copy terminal that is
capable of producing an original and one copy. It is particularly
appropriate for systems requiring large numbers of printer-terminals.
The LA30-P is a parallel model and the LA30-S is a serial model.

2.1.4

LA36 DECwriter

The LA36 DECwriter is an asynchronous terminal which produces hard
copy and operates in serial mode. It has an impact printer capable of
generating multipart and special preprinted forms.
Both upper-case
and lower-case characters can be received and transmitted.

2.1.5

RT02 Alphanumeric Display Terminal
Alphanumeric Display Terminal

and

RT02-C

Badge

Reader/

The RT02 is a compact, alphanumeric display terminal designed for
applications in which source data is primarily numeric. A shift key
permits the entry of 30
discrete characters, including upper-case
alphabetic characters. The RT02 can, however, receive and display 64
characters.
The RT02-C model also contains a badge reader. This option provides a
reliable method of identifying and controlling access to the PDP-11 or
to a secure facility. Furthermore, data in a format corresponding to
that of a badge (22-column fixed data) can be entered very quickly.

2-2

TERMINAL DRIVER
2.1.6

VT05B Alphanumeric Display Terminal

The VT05B is an alphanumeric display terminal that consists of a CRT
a1splay and a self-contained keyboard. From a programming point of
view, it is equivalent to other terminals, except that the VT05B
offers direct cursor addressing.

2.1.7

VT50 Alphanumeric Display Terminal

The VT50 is an alphanumeric display terminal that consists of a CRT
display and a keyboard. It is similar to the VT05B in operation, but
does not offer direct cursor addressing.

2.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) contains the following
information for terminals.
A setting of 1
indicates that the
described characteristic is true for terminals.
Bit

Setting

Meaning

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 are undefined; word 5 indicates the default buffer
size for the device, for terminals the width of the terminal carriage
or display screen.

2. 3

QIO MACRO

Table 2-3 lists the standard and device-specific functions of the QIO
macro that are valid for terminals.

2-3

TERMINAL DRIVER
Table 2-3
Standard and Device-Specific QIO Functions for Terminals
STANDARD FUNCTIONS:
Format

Function

QIO$C IO.ATT,...

Attach device

QIO$C IO.DET,...

Detach device

QIO$C IO.KIL,...

Cancel I/O requests

QIO$C IO.RLB, •.. ,<stadd,size>

Read logical block
(Read typed input into buffer)

QIO$C IO.RVB, .•• ,<stadd,size>

Read virtual block
(Read typed input into buffer)

QIO$C IO.WLB, •.. ,<stadd,size,vfc>

Write logical block
(Print buffer contents)

QIO$C IO.WVB, ••• ,<stadd,size,vfc>

Write virtual block
(Print buffer contents)

DEVICE-SPECIFIC FUNCTIONS:
QIO$C IO.RAL, ... ,<stadd,size>

Read logical block. Pass all bits.
Do not intercept control characters
or mask out parity bit.

QIO$C IO.WAL, ... ,<stadd,size>

Write logical block.
Pass
all
bits.
Do not intercept control
characters or add parity bit.

where:

stadd

is the starting address of the data buffer (may be on a
byte boundary) •

size

is the data buffer size in bytes (must be greater
zero) •

vfc

is a vertical format control character from Table 2-7.

than

The effect of IO.KIL on an in-progress request depends upon whether
the request is for input or output. If it is for input (i.e., IO.RLB,
IO.RVB, or IO.RAL), the request is forced to terminate, IE.ABO is
returned, and the second word of the I/O status block contains the
number of bytes already typed. If the request is for output (i.e.,
IO.WLB, IO.WVB, or IO.WAL), the transfer is terminated, and IS.SUC is
returned.

2.4

STATUS RETURNS

Table 2-4 lists error and status conditions that are returned
terminal driver described in this chapter.

2-4

the

TERMINAL DRIVER
Table 2-4
Terminal Status Returns
Code

Reason

IE.EOF

Successful completion on a read with End-of-file
The line of input read from the terminal was terminated
with the end-of-file character CTRL/Z.

rs.sue

Successful completion
The operation specified in the QIO directiv~ was
completed successfully. . The second word of the I/O
status block can be examined to determine the number of
I bytes
processed, if the operation involved reading or
writing.

IS.CR

Successful completion on a read
The line of input read from the terminal was terminated
by a carriage return.

IS.ESC

Successful completion on a read
The line of input read from the terminal was terminated
by an ESCape or ALTMODE character.

IS.PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed. The I/O status block if filled with 1
zeros.

IE.ABO

Operation aborted
The specified I/O operation was cancelled via
while in progress or while in the I/O queue.

IO.KIL

IE.DAA

Device already attached.
The physical device unit specified in an
IO.ATT
function was already attached by the issuing task.
This code indicates that the issuing task has already
attached the desired physical device unit, not that the
unit was attached by another task.

IE.DNA

Device not attached
IO.DET
The physical device unit specified in an
function was not attached by the issuing task. This
code has no bearing on the attachment status of other
tasks.

2-5

TERMINAL DRIVER
Table 2-4 (Cont.)
Terminal Status Returns
Code

Reason

IE.DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.
This code is returned to indicate one of the following
conditions:
. A timeout occurred on the physical
unit (i.e., an interrupt was lost) •

device

. An attempt was made to perform a transfer on
a remote DHll line without carrier present.
IE.IFC

Illegal function
A function code was specified in an
was illegal for terminals.

IE.NOD

I/O

request

that

Buffer allocation failure
Dynamic storage has been depleted, and there was I
insufficient space available to allocate a buffer for
an input request (i.e., all input is buffered in the
terminal driver).

IE.OFL

Device off-line
The physical device unit associated with the LON
specified in the QIO directive was not on-line. When
the system was booted, a device check indicated that
this physical device unit was not in the configuration.

IE.SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task. Alternately, a byte count of zero was
specified.

The following illustrates the contents of
return of an IS.ESC code:

word O
1

1+1

the

I/O

status

block

Byte

Number of bytes read

where 33 is the octal representation of the ESCape or ALTMODE
character, and +l is the status code for successful completion
(IS.SOC). The contents of this block on return of IS.CR are the same,
except that the high-order byte of word 0 contains 15, the octal code
for carriage return. Unlike other RSX-llM return codes, IS.CR and
IS.ESC are word values, rather than byte values. The low-order byte

2-6

TERMINAL DRIVER
simply indicates successful completion, and the high-order byte is
required to show the specific type. To test for an IS.ESC or IS.CR
code, the user can first test the low-order byte of the first word of
the I/O status block J..VJ. rs. sue, and then test the full word for
IS.ESC or IS.CR.
.c~-

2.5

CONTROL CHARACTERS AND SPECIAL KEYS

This section describes the particular meaning of special terminal
control characters and keys for RSX-llM. Note that control characters
and special keys are not recognized by the driver during a read/pass
all request (IO.RAL).

2.5.1

Control Characters

A control character is input from a terminal by holding the control
key (CTRL) down while typing one other key. Two of the control
characters described in Table 2-5, CTRL/U and CTRL/Z, are echoed on
the terminal printer as AU and AZ respectively.
Other control
characters are recognized by the terminal driver but are not printing
characters and are therefore not echoed.
Table 2-5
Terminal Control Characters

I Character

Meaning
Typing CTRL/C on the terminal causes unsolicited input
on that terminal to be directed to the Monitor Console
Routine (MCR). When the unsolicited input completes,
it is passed to the MCR dispatcher. "MCR>" is echoed
when the terminal is ready to accept the unsolicited
input.

CTRL/C

CTRL/I

Typing CTRL/I initiates a horizontal tab, and the
terminal spaces to the next tab stop. Tabs are set at
every eighth character position.

CTRL/J

Typing CTRL/J is equivalent to typing the LINE FEED key
on the terminal.

CTRL/K

Typing CTRL/K initiates a vertical
terminal performs four line feeds.

CTRL/L

Typing CTRL/L initiates a form feed, and the terminal
performs eight line feeds. Paging is not performed.

CTRL/M

Typing CTRL/M is equivalent to typing the carriage
RETURN key on the terminal (See section 2.5.2).

CTRL/O

Typing CTRL/O suppresses output being sent to
a
terminal by the current I/O request. For attached
terminals, CTRL/O remains in effect,
and
output
continues to be suppressed until any of the following
occur:

2-7

tab,

and

the

TERMINAL DRIVER
Table 2-5 (Cont.)
Terminal Control Characters
Character

Meaning
The terminal is detached
Solicited input is entered
Unsolicited input is entered
Another CTRL/O character is typed
For unattached terminals, CTRL/O suppresses output
only the current output buffer.

previously

CTRL/Q

Typing CTRL/Q resumes terminal
suspended by means of CTRL/S.

CTRL/S

Typing CTRL/S causes terminal output to
Output is resumed by typing CTRL/Q.

CTRL/U

Typing CTRL/U before typing a line terminator causes
previously typed characters to be deleted back to the
beginning of the line.
The system
echoes
this
character as AU, followed by a carriage return and a
line feed. This allows the line to be retyped.

I CTRL/Z

Typing CTRL/Z indicates an end-of-file for the current
terminal input.
It signals MAC, PIP, TKB, and other
system tasks that terminal input is complete and the
task should exit. The system echoes this character as
AZ followed by a carriage return and a line feed.

2.5.2

output

for

suspended.

Special Keys

The ESCape, carriage RETURN, and RUBOUT keys have special significance
for terminal input, as described in Table 2-6.
A line can be
terminated by an ESCape (or ALTMODE) character, by a carriage RETURN,
by CTRL/Z, or by completely filling the input buffer (i.e., exhausting
the byte count before a line terminator is typed).
The standard
buffer size for a terminal can be determined by issuing a GET LON
INFORMATION system directive and examining word 5 of the information
buffer.
Table 2-6
Special Terminal Keys
Key

Meaning

ESC

Typing ESCape or ALTMODE signals the terminal driver
that there is no further input on the current line.
This line terminator allows further input on the same
line since the carriage or cursor is not returned to
the first column position.

RETURN

Typing RETURN terminates the current line and causes
the carriage or cursor to return to the first column on
the line.
- - ·-.
·-··

2-8

' ..

. . . ,.,

TERMINAL DRIVER
Table 2-6 (Cont.)
Special Terminal Keys
Key

Meaning

RUBOUT

Typing RUBOUT deletes the last character typed on an
input line. Only characters typed since the last line
terminator may be deleted. Several characters can be
deleted in sequence by typing successive RUBOUTs. The
first RUBOUT echoes as a backslash (\) , followed by the
character which has been deleted. Subsequent RUBOUTs
cause only the deleted character to be echoed.
The
next character typed which is not a RUBOUT causes
another \ followed by the new character to be echoed.
The following example illustrates rubbing out ABC and
then typing CBA:
ABC\CBA\CBA
The second backslash is not displayed if a line
terminator is typed after rubbing out the characters on
a line, as in the following:
ABC\CBA

2.6

VERTICAL FORMAT CONTROL

Table 2-7 below summarizes the meanings of all characters used for
vertical format control on the terminal. Any one of these characters
can be specified as the value of the vfc parameter in an IO.WLB or
IO.WVB function.
Table 2-7
Vertical Format Control Characters
Octal
Value

Character

040

blank

SINGLE SPACE - Output a line feed, print
the
contents of the buffer, and output a carriage
return. Normally, printing immediately follows
the previously printed line.

060

zero

DOUBLE SPACE - Output two line feeds, print the
contents of the buffer, and output a carriage
return. Normally, the buffer contents are printed
two lines below the previously printed line.

061

one

PAGE EJECT - Output eight line feeds, print the
contents of the buffer, and output a carriage
return.

053

plus

OVERPRINT - Print the contents of the buffer and
output a carriage return, normally overprinting
the previous line.

Meaning

2-9

TERMINAL DRIVER
Table 2-7 {Cont.)
Vertical Format Control Characters
Octal
Value

Character

Meaning

044

dollar
sign

PROMPTING OUTPUT - Output a line feed and print
the contents of the buffer. This mode of output
is intended for use with a terminal where a
prompting message is output and input is then read
on the same line.

000

null

INTERNAL VERTICAL FORMAT - The buffer contents are
printed
without
addition of vertical format
control characters. In this mode, more than one
line
of guaranteed contiguous output can be
printed per I/O request.

All other vertical format control characters are interpreted as blanks
{octal 040).

2.7

TERMINAL INTERFACES

This section summarizes the characteristics of the three types
standard communication line interfaces supported by RSX-llM.

2.7.1

DHll Asynchronous Serial Line Multiplexer

The DHll multiplexer interfaces up to 16
asynchronous
serial
communications lines for terminal use. The DHll supports programmable
baud rates with no parity. The DMll-BB option may be included to
provide modem control for dial-up lines.
These lines must be
interfaced via Bell 103 or equivalent modems.

2.7.2

DJll Asynchronous Serial Line Multiplexer

The DJll multiplexer interfaces as many as 16 asynchronous serial
lines to the PDP-11 for local terminal communications. The DJll does
not provide a dial-up capability but supports jumper selectable baud
rates.

2.7.3

DLll Asynchronous Serial Line Interface

The DLll supports a single asynchronous serial line and handles
full-duplex communication between the PDP-11 and a terminal. There
are 13 standard baud rates available to DLll users (40-9600 baud).
Four versions of the DLll interface are supported by RSX-llM for
terminal use: DLll-A, DLll-B, DLll-C, and DLll-D.
The DLll-E is
supported only for message-oriented communications and is described in
Chapter 9.

2-10

TERMINAL DRIVER
2.8

PROGRAMMING HINTS

This section contains information on important considerations relevant
to users of the terminal driver described in this chapter.

2.8.1

Terminal Line Truncation

If the number of characters to be printed exceeds the line length of
the physical device unit, the terminal driver discards the excess
characters until it receives one that instructs it to return to
horizontal position 1. The user can determine that this will happen
by examining word 5 of the information buffer returned by the GET LON
INFORMATION system directive.

2.8.2

ESCape Code Conversion

An ESCape or ALTMODE character code of 33, 175, or 176 is converted
internally to 33 before it is returned to the user on input.

2.8.3

RT02-C Control Function

When sending a control character (e.g., vertical tab}
to the RT02-C
Badge Reader and Data Entry Terminal, the high-order bit (bit 7) of
the byte must be set to one. This causes the terminal driver not to
recognize the character.
In the case of a vertical tab, 213 octal
must be output rather than 13 octal.

2-11

CHAPTER 3
DISK DRIVERS

3.1

INTRODUCTION

the RSX-llM disk drivers support the disks summarized in Table
Subsequent sections describe these devices in greater detail.

3-1.

Table 3-1
Standard Disk Devices
MODEL

RPM

SURFACES

RFll/RSll

1800

128

2048.

262,144.

RP04

3600

411

5632.

43,980,288.

RS03

3600

4096.

262,144.

RS04

3600

8192.

524,288.

RK11/RK05

1500

200

3072.

1,228,800.

RP11E/RP02

2400

200

2560.

10,240,000.

RPllC/RP03

RXll/RXOl

"'Ii A f'\ /"\

CYLINDERS WORDS/
TRACK

Ann

'1c::t:n

WORDS/
DRIVE

'HI

A Q ()

() () ()

':tu u

..::;v

"'i:UU

.&.. .JUU o

'Vf""%VVfVVVe

360

1664.

128,128.

All of the disks described in this chapter are accessed in essentially
the same manner. Up to eight disks of each type (except RXOl) may be
connected to their respective
controllers.
Disks
and
other
file-structured media under RSX-llM are divided logically into a
series of 256-word blocks.

3.1.1

RFll/RSll Fixed-Head Disk

The RFll controller/RSll fixed-head disk provides random-access bulk
storage. It features fast track-switching time and a redundant set of
timing tracks. The RFll/RSll is unique because the hardware can
automatically perform a spiral read across disk platters.

3-1

DISK DRIVERS
3.1.2

RP04 Pack Disk

The RP04 (RH11-RH70 controller/RP04 pack disk) pack disk consists of
19 data surfaces and a moving read/write head. It is similar to the
RP11-C/RP03, but has twice the capacity.
The RP04 offers large
capacity storage with rapid access time.

3.1.3

RS03 Fixed-Head Disk

The RS03 (RH11-RH70 controller/RS03 fixed-head disk) is a fixed head
disk which offers speed and efficiency. With 64 tracks per platter,
and recording on one surface, the RS03 has a capacity of 262,144
words.

3.1.4

RS04 Fixed-Head Disk

The RS04 (RH11-RH70 controller/RS04 fixed-head disk) is similar to the
RS03 disk, and interfaces to the same controller; but provides twice
the number of words per track by recording on both surfaces of the
platter, and thus twice the capacity.

3.1.5

RKll/RKOS Cartridge Disk

The RKll controller/RKOS DECpack cartridge disk is an economical
storage
system
for
medium-volume, random-access storage.
The
removable disk cartridge offers the flexibility of large off-line
capacity with rapid transfers of files between on- and off-line units
without necessitating copying operations.

3.1.6

RP11/RP03 or RP02 Pack Disk

The RPll controller/RP02 or RP03 pack disk consists of 20 data
surfaces and a moving read/write head. The RP03 has twice as many
cylinders, and thus, double the capacity of the RP02.
Only an even
number of words can be transferred in a read/write operation.

3.1.7

RXll/RXOl Flexible Disk

The RXll controller/RXOl flexible disk is an economical storage system
for low-volume, random-access storage. Data is stored in 26 64-word
sectors per track; there are 77 tracks per disk.
Data may be
accessed by physical sector or logical block. If logical or virtual
block I/O is selected, the driver reads four physical sectors.
These
sectors are interleaved to optimize data transfer. The next logical
sector that falls on a new track is skewed by six sectors to allow for
track
to
track switch time.
Physical block I/O provides no
interleaving or skewing and provides access to all 2002 sectors on the
disk.
Logical or virtual I/O starts on track one and provides access
to 494 logical blocks.

3-2

DISK DRIVERS
3.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) con ta ins the fol lowing
information for disks.
A bit setting of l indicates that the
described characteristic is true for disks.
Bit

Setting

Record-oriented device

Carriage-control device

Terminal device

,
..L.

n.: ---'----··
LJ.LL
~1..,; 1..VLY

Single-directory device

Sequential device

6-12

Pe served

Device mountable as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

Meaning

...:1---!--

u~v .LI..;~

Words 3 and 4 of the buffer are undefined;
word
default buffer size, which is 512 for all disks.

indicates

the

QIO MACRO

This section summarizes the
functions for disk drivers.

3.3.l

standard,

and

device-specific

QIO

Standard QIO Functions

Table 3-2 lists the standard functions of the QIO macro that are valid
for disks.

3-3

DISK DRIVERS
Table 3-2
Standard QIO Functions for Disks
Format

Function

QIO$C IO.ATT, .••

Attach device*

QI0$C IO.DET, •••

Detach device

QIO$C IO.KIL, .••

Not applicable (NOP}

QIO$C IO.RLB, ... ,<stadd,size,,blkh,blkl>

Read logical block

QIO$C IO.RVB, .•. ,<stadd,size,,blkh.blkl>

Read virtual block

QIO$C IO.WLB, ..• ,<stadd,size,,blkh,blkl>

Write logical block

QIO$C IO.WVB, ... ,<stadd,size,,blkh,blkl>

Write virtual block

where:

stadd

is the starting address of the data buffer (must be
a word boundary} .

size

is the data buffer size in bytes (must be even, greater
than zero, and, for the RP02 and RP03, also a multiple
of four bytes} •

blkh/blkl

are block high and block low, combining to form a
double-precision
number
that
indicates
the
logical/virtual block address on the disk where the
transfer starts;
blkh represents the high eight bits
of the address, and blkl the low 16 bits.

IO.RVB and IO.WVB are associated with file operations (see the RSX-11
I/O Operations Reference Manual}. For these functions to be executed,
a file must be open on the specified LUN if the volume associated with
the LUN is mounted. Otherwise, the virtual I/O request is converted
to a logical I/O request using the specified block numbers.

3.3.2

Device-Specific QIO Functions

The device-specific functions of the QIO macro are valid for the
only; they are shown in Table 3-3.

RXll

*Only unmounted volumes may be attached.
An attempt to attach a
mounted volume will result in an IE.PR! status being returned in the
I/O status doubleword.

3-4

DISK DRIVERS
Table 3-3
Device-Specific Functions for the RXOl Disk Driver
QIO$C IO.RPB, ... ,<stadd,size,,,pbn> Read physical block
QIO$C IO.WDD, .•• ,<stadd,size,,,pbn> Write physical block (with deleted
data mark)
QIO$C IO.WPB, ... ,<stadd,size,,,pbn> Write physical block
where:

3.4

stadd is the starting address of the data buffer (must be
a word boundary) .

size

is the data buffer size
greater than zero) .

and

pbn

is the physical block number where the transfer
(must be in the range 0 to 2001.).

bytes

must

even

starts

STATUS RETURNS

The error and status conditions listed in Table 3-4
the disk drivers described in this chapter.

are

returned

Table 3-4
Disk Status Returns
Code

Reason

rs.sue

Successful completion
The operation specified in the QIO directive was
completed successfully.
The second word of the I/O
status block can be examined to determine the number of
I

bvtPs
-- ..t. -= · -

nro~PssPn.
- - •

if rhP onPrArion
or
c - - - - - - --involvPn
-- - - - - - -rPAnina
- - - - -- I

i:- -

- - -

writing.
IS.PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed. The I/O status block is filled with
zeros.

IS.RDD

Deleted data mark read
A deleted record was encountered while executing an
IO.RPB function.
The second word of the I/O status
block can be examined to determine the number of bytes
processed.

IE .ALN

File already open
The task attempted to open a file on the physical
device unit associated with specified LUN, but a file
has already been opened by the issuing task on that
LUN.

3-5

DISK DRIVERS
Table 3-4 (Cont.)
Disk Status Returns
Code
IE.BLK

Reason
Illegal block number
An illegal logical block number was specified.
This
code would be returned, for example, if block 4800 were
specified for an RKOS disk, on which legal block
numbers extend from zero through 4799.

IE. BYT

Byte-aligned buffer specified
Byte alignment was specified for a buffer, but only
word alignment is legal for disk. Alternatively, the
length of a buffer is not an appropriate number of
bytes.
For example, all RP03 and RP02 disk transfers
must be multiples of four bytes.

IE.DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.

IE. IFC

Illegal function
A function code was specified in an I/O request that is
illegal for disks.

I IE. NLN

File not open
The task attempted to close a file on the physical
device unit associated with the specified LUN, but no
file was currently open on that LUN.

IE. NOD

IE. OFL

IE.OVR

Illegal read overlay request
A read overlay was requested, and the physical device
unit specified in the QIO directive was not the
physical device unit from which the task was installed.
The read overlay function can only be executed on the
I physical device unit from which the task image
containing the overlays was installed.

3-6

DISK DRIVERS
Table 3-4 (Cont.)
Disk Status Returns
Code

Reason

IE.PRI

Privilege violation
The task which issued the request was not privileged to
execute that request. For disk, this code is returned
if a nonprivileged task attempts to read or write a
mounted volume directly (i.e., using IO.RLB or IO.WLB).
Also, this code is returned if any task attempts to
attach a mounted volume.

IE.SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task. Alternately, a byte count of zero was
specified.

IE.VER

Unrecoverable error
j After

the system's standard number of retries has been
attempted upon encountering an error, the operation
1
still could not be completed. For disk, unrecoverable
I errors are
usually parity errors.
1

IE.WLK

Write-locked device

I The task attempted to
I physically write-locked.
1

write

disk

that

was

When a disk I/O error condition is detected, an error is usually not
returned immediately. Instead, RSX-llM attempts to recover from most
errors by retrying the function as many as eight times. Unrecoverable
errors are generally parity, timing, or other errors caused by a
hardware malfunction.

3-7

CHAPTER 4
DECTAPE DRIVER

4.1

INTRODUCTION

The RSX-llM DECtape driver supports the TCll-G dual DECtape controller
with up to three additional dual DECtape transports. The TCll-G is a
dual-unit, bidirectional, magnetic-tape transport system for auxiliary
data storage.
DECtape is formatted to store data at fixed positions
on the tape, rather than at unknown or variable positions as on
coventional magnetic tape. The system uses redundant recording of the
mark, timing, and data tracks to increase reliability.
Each reel
contains 578 logical blocks. As with disk, each of these blocks can
be accessed separately, and each contains 256 words.

4.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive
(the first characterics word} contains the following
information for DECtapes. A bit setting of 1 indicates that the
described characteristic is true for DECtapes.
Bit

Setting

Meaning

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 of the buffer are undefined;
default buffer size, for DECtape 512 bytes.

4-1

word

indicates

the

DECTAPE DRIVER
4. 3

QIO MACRO

This section summarizes standard and device-specific QIO functions for
the DECtape driver.

4.3.l

Standard QIO Functions

Table 4-1 lists the standard functions of the QIO macro that are valid
for DECtape.
Table 4-1
Standard QIO Functions for DECtape
Function

Format
QIO$C IO.ATT, •••

Attach device*

QIO$C IO.DET, •••

Detach device

QIO$C IO.KIL, •••

Not applicable (NOP)

QIO$C IO.RLB, ••• ,<stadd,size,,,lbn>

Read logical block (forward)

QIO$C IO.RVB, •.• ,<stadd,size,,,lbn>

Read virtual block (forward)

QIO$C IO.WLB, ••• ,<stadd,size,,,lbn>

Write logical block (forward)

QIO$C IO.WVB, ••. ,<stadd,size,,,lbn>

Write virtual block (forward)

where:

stadd

is the starting address of the data buffer (must be
a word boundary) •

size

is the data buffer size in
greater than zero) •

even

and

lbn

is the logical block number on the DECtape where
transfer starts (must be in the range 0-577).

the

bytes

(must

IO.RVB and IO.WVB are associated with file operations (see the RSX-llM
I/O Operations Reference Manual). For these functions to be executed,
a file must be open on the specified LUN if the volume associated with
the LUN is mounted. Otherwise, the virtual I/O request is converted
to a logical I/O request using the specified block numbers.

4.3.2

Device-Specific QIO Functions

The device-specific functions of the QIO
DECtape are shown in Table 4-2.

macro

that

are

valid

for

*Only unmounted volumes mav be attached.
An attempt to attach a
mounted volume will result in an IE.PR! status being returned in the
I/C status doubleword.

4-2

DECTAPE DRIVER
Table 4-2
Device-Specific Functions for DECtape
Format

Function

QIO$C IO.RLV, ••. ,<stadd,size,,,lbn>

Read logical block (reverse)

QIO$C IO.WLV, ••• ,<stadd,size,,,lbn>

Write logical block (reverse)

Where:

4.4

stadd

is the starting address of the data buffer (must be
a word boundary) .

size

is the data buffer size in
greater than zero).

even

and

lbn

is the logical block number on the DECtape where
transfer starts (must be in the range 0-577).

the

bytes

(must

STATUS RETURNS

The error and status conditions listed in Table 4-3
the DECtape driver described in this chapter.

are

returned

Table 4-3
DECtape Status Returns
Code

Reason

rs.sue

Successful completion
The operation specified in the QIO directive was
completed successfully.
The second word of the !/O 1
status block can be examined to determine the number of I
bytes processed, if the operation involved reading or

I writing.

IS. PND

I/O request pending

The operation specified in the QIO directive has not
yet been executed. The I/O status block is filled with
zeros.
IE .ALN

File already open
The task attempted to open a file on the physical
device unit associated with the specified LUN, but a
file has already been opened by the issuing task on
that LUN.

IE. BLK

Illegal block number.
An illegal logical block number was specified
DECtape. The number exceeds 577 (1101 octal).

4-3

for

DECTAPE DRIVER
Table 4-3 {Cont.)
DECtape Status Returns
Code
IE. BYT

Reason
Byte-aligned buffer specified.
Byte alignment was specified for a buffer, but only
word alignment is legal for DECtape. Alternately, the
length of the buffer is not an even number of bytes.

IE.DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.

IE. IFC

Illegal function
A function code was specified in an I/O request that is
illegal for DECtape.

IE. NLN

File not open
The task attempted to close a file on the physical
device unit associated with the specified LUN, but no
I file was currently open on that LUN.

IE. NOD

I IE.OFL

IE.OVR

Illegal read overlay request
A read overlay was requested and the physical device
unit specified in the QIO directive was not the
physical device unit from which the task was installed.
The read overlay function can only be executed on the
physical device unit from which the
task
image
containing the overlays was installed.

IE. PRI

Privilege violation
The task which issued the request was not privileged to
execute that request.
For DECtape, this code is
returned when a nonprivileged task attempts to read or
write a mounted volume directly {i.e., IO.RLB, IO.RLV,
IO.WLB, or IO.WLV). Also, this code is returned if any
I task at. tempts to attach a mounted volume.

4-4

DECTAPE DRIVER
Table 4-3 (Cont.)
DECtape Status Returns
Code

Reason

IE. SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task. Alternately, a byte count of zero was
spec i f ied.

IE. VER

Unrecoverable error
After the system's standard number of retries has been
attempted upon encountering an error, the operation
still could not be completed. For DECtape, this code
is
returned
to
indicate
any of the following
I conditions.

I · A parity error was encountered.
j •

The task attempted a forward
past block 577 {iiOl octal).

multi-block

transfer

I.I past
The task attempted a backward multi-block transfer
block zero.
IE. WLK

I Write-locked device
The task attempted to write on a DECtape unit that
physically write-locked.

4.4.1

was

DECtape Recovery Procedures

When a DECtape I/O error condition is detected, RSX-llM attempts to
recover from the condition by retrying the function as many as five
times. Unrecoverable errors are generally parity, mark track, or
other errors caused by a faulty recording medium or a hardware
malfunction. An unrecoverable error condition also occurs when a read
or write operation is performed past the last block of the DECtape on
a forward operation, or the first block of the DECtape on a reverse
operation.
In addition to the standard error conditions, an unrecoverable error
is reported when the "rock count" exceeds eight. The rock count is
the number of times the DECtape driver reverses the direction of the
tape while looking for a block number. Assume that the block numbers
on a portion of DECtape are 99, 96, and 101, where one bit was dropped
from block number 100, making it 96. If an I/O request is received
for block 100 and the tape is positioned at block 99, the driver
starts searching forward for block 100.
The first block to be
encountered is 96 and because the driver is searching for block 100 in
a forward direction and 96 is less than 100, the search continues
forward. Block 101 is the next block, and because number 101 is
greater than 100, the driver reverses the direction of the tape and

4-5

DECTAPE DRIVER
starts to search backward. The next block number in this direction is
96 and direction is reversed again, because 100 is greater than 96.
To prevent the DECtape from being hung in this position, continually
rocking between block numbers 96 and 100, a maximum rock count of
eight has been established.

4.4.2

Select Recovery

If the DECtape unit is in an off-line condition when the I/O function
is performed, the message shown below is output on the operator's
console.

*** DTn:

-- SELECT ERROR

where n is the unit number of the drive that is currently off-line.
The user should respond by placing the unit to REMOTE. The driver
retries the function, from the beginning, once every second.
It
displays the message once every 15 seconds until the appropriate
DECtape unit is selected. A select error may also occur when there
are two drives with the same unit number or when no drive has the
appropriate unit number.

4.5

PROGRAMMING HINTS

This
section
considerations
this chapter.

4.5.1

contains
information
on
important
programming
relevant to users of the DECtape driver described in

DECtape Transfers

If the transfer length on a write is less than 256 words, a partial
block is transferred with zero fill for the rest of the physical
block. If the transfer length on a read is less than 256 words, only
the number of words specified is transferred. If the transfer length
is greater than 256 words, more than one physical
block
is
transferred.

4.5.2

Reverse Reading and Writing

The DECtape driver supports reverse reading and writing, because these
functions speed up data transfers in some cases. A block should
normally be read in the same direction in which it was written. If a
block is read from a DECtape into memory in the opposite direction
from that in which it was written, it is reversed in memory (e.g.,
word 255 becomes word O, and 254 becomes word 1). If this occurs, the
user must then reverse the data within memory.

4.5.3

Speed Considerations When Reversing Direction

It is possible to reverse direction at any time while reading or
writing DECtape.
However, the user should understand that reversing

4-6

DECTAPE DRIVER
direction substantially slows down the movement of the tape.
Because
DECtape must be moving at a certain minimum speed before reading or
writing can be performed, a tape block cannot be accessed immediately
after reversing direction. Two blocks must be bypassed before a read
or write function can be executed, to give the tape unit time to build
up to normal access speed. Furthermore, when a request is issued to
read or write in a certain direction, the tape first begins to move in
that direction, then starts detecting block numbers. The following
examples illustrate these principles.
If a DECtape is positioned at block number 12 and the driver receives
a request to read block 10 forward, the tape starts to move forward,
in the direction requested. When block number 14 is encountered, the
driver reverses the direction of the tape, since 14 is greater than
10. The search continues backward, and block numbers 11 and 10 are
encountered.
Because the direction must be reversed and the driver
requires two blocks to build up sufficient speed Ior reading, block
number 9 and 8 are also bypassed in the backward direction. Then the
direction is reversed and the driver encounters blocks 8 and 9 forward
before reaching block number 10 and executing the read request.

4.5.4

Aborting a Task

If a task is aborted while waiting for a unit to be
DECtape driver recognizes this fact within one second.

4-7

selected,

the

CHAPTER 5
MAGNETIC TAPE DRIVERS

5.1

INTRODUCTION

RSX-llM supports three magnetic tape devices:
the TUlO, the TS03, and
the TU16.
Table 5-1 summarizes these devices and subsequent sections
describe them in greater detail.
Table 5-1
Standard Magtape Devices
TUlO

TU16

TS03

Number of channels

7 or 9

Recording density, in frames
per inch

For 7-channel:
200, 556, or 800;
for 9-channel:
800

800 or 1600

800

Tape speed, in inches per
second

Maximum data transfer rate,
in bytes per second

36,000

For 800 bpi:
12,000
36,000;
for 1600 bpi:
72,000

Recording Method

NRZI

NRZI or
Phase
Encoding

NRZI

Programming for Magtape
is quite similar
to programming for
the
magnetic
tape cassette
(see Chapter 6).
Unlike cassette, however,
Magtape can handle variable-length records and allows
the user
to
select a parity mode.
RSX-llM does not support a file structure for Magtape.

5.1.1

TU10/TS03 Magnetic Tape

The TU10/TS03 consists of a
TMll
transport.
It
is a low-cost,

controller with a TUlO or TS03
high performance system for serial

5-1

MAGNETIC TAPE DRIVERS
storage
of
large
volumes
of
data
and
industry-compatible format.
All recording is
inverted (NRZI).

5.1.2

programs
in
an
non-return-to-zero,

TU16 Magnetic Tape

The TU16 consists of an RH11/RH70 controller, a TM02 formatter, and a
TU16 transport.
It is quite similar to the TUlO but is a Massbus
device, with a common controller, a specialized formatter, and a
drive.
Recording is either 800 bpi NRZI or 1600 bpi phase-encoded
(PE) •
5.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) contains the following
information for Magtapes. A bit setting of 1 indicates that the
described characteristic is true for Magtapes.
Setting

Meaning

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 of the buffer are undefined;
default buffer size, for Magtapes 512 bytes.

5.3

word

indicates

the

QIO MACRO

This section summarizes standard and device-specific QIO functions for
the Magtape drivers.

5-2

MAGNETIC TAPE DRIVERS
5.3.1

Standard QIO Functions

Table 5-2 lists the standard functions of the QIO macro that are valid
for Mag tape.

Table 5-2
Standard QIO Functions for Magtape
Format

Function

QIO$C IO.ATT, ...

Attach device

QIO$C IO. DET, •..

Detach device

QIO$C IO.KIL, •••

Cancei I/O requests

QIO$C IO.RLB, .•• ,<stadd,size>

Read logical block
(read tape into buffer)

QIO$C IO.RVB, •.. ,<stadd,size>

Read virtual block
(read tape into buffer)

QIO$C IO.WLB, ... ,<stadd,size>

Write logical block
(write buffer contents to tape)

QIO$C IO.WVB, ••. ,<stadd,size>

Write virtual block
(write buffer contents to tape)

where:

stadd

is the starting address of the data buffer {must be
a word boundary} •

size

is the data buffer size in bytes (must be even, greater
than zero, and, for a write, must be at least 14
bytes) •

IO.KIL does not cancel an in progress request unless
has occurred.

5.3.2

select

error

Device-Specific QIO Functions

Table 5-3 lists the device-specific functions of the QIO macro that
are valid for Magtape. Additional details on certain functions appear
below.

5-3

MAGNETIC TAPE DRIVERS
Table 5-3
Device-Specific QIO Functions for Magtape
Format

Function

QIO$C IO. EOF, .••

Write end-of-file mark (tape mark)

QIO$C IO. RWD I • • •

Rewind unit

QIO$C IO. RWU I • • •

Rewind and turn unit off-line

QIO$C IO. SEC I • . . .

Sense tape characteristics

QIO$C IO.SMO, •.. ,<cb>

Mount tape and set tape characteristics

QIO$C IO.SPB, ••• ,<nbs>

Space blocks

QIO$C IO.SPF, ••. ,<nes>

Space files

QIO$C IO.STC, .•• ,<cb>

Set tape characteristics

where:

represents the characteristic bits to set.

nbs

is the number of blocks to space
forward, negative i f reverse).

past

(positive

. nes

is the number of EOF marks to space past
forward, negative if reverse).

(positive

5.3.2.1 IO.RWD - Completion of IO.RWD means that the rewind has been
initiated. However, a request for the same unit will be queued by the
driver until load point (BOT) is reached.

5.3.2.2 IO.RWU - IO.RWU is normally used when operator
intervention
is required
(e.g.,
to load a new tape).
The operator must turn the
unit back on-line manually before subsequent operations can proceed.

5.3.2.3 IO.SEC - This function returns the
tape characteristics in
the second I/O status word.
The tape characteristic bits are defined
as follows:
Can Be Set by
IO.SMO and IO.STC

Meaning When Set

For TUlO, 556 bpi density
(seven-channel).
For
TU16, reserved.

For TUlO, 200 bpi density
(seven-channel).
For
TU16, reserved.

5-4

MAGNETIC TAPE DRIVERS
Can Be Set by
IO.SMO and IO.STC

Meaning When Set
2

For TUlO, core-dump mode
(seven-channel, see below).
For TU16, reserved.

Even parity (default is odd).

Tape is past EOT.

Last tape command encountered
EOF (unless last command was
backspace) .

Writing is prohibited.

Writing with extended interrecord gap is prohibited
(i.e., no recovery is attempted
after write error).

Select error on unit (reserved
for driver; always 0 when read
by user) .

Unit is rewinding (reserved for
driver; always 0 when read by
user).

Tape is physically write-locked.

For TU10/TS03, reserved.
1600 bpi, density.

For TUlO, drive is seven-channel.
For TU16, reserved.

Tape is at load point (BOT).

Tape is at end-of-volume (EOV).

Tape is past EOV (reserved for driver; always 0 when read by
user) .

For TU16,

In core-dump mode (TUlO only, 800 bpi density, and seven-channel),
each eight-bit byte is written on two tape frames, four bits per
frame. In other modes on seven-channel tape, only six low-order bits
per byte are written.
The effect of these settings is illustrated in Figure 5-1 for the TUlO
and in Figure 5-2 for TU16.

5-5

MAGNETIC TAPE DRIVERS

SET EVEN
PARITY

yes

SET 800 BPI,
NINE-CHANNEL

yes

SET 556 BPI,
SEVEN-CHANNEL

yes

SET 800 BPI,
SEVEN-CHANNEL,
CORE-DUMP MODE

yes

SET 200 BPI,
SEVEN-CHANNEL

yes

SET 800 BPI,
SEVEN-CHANNEL

Figure 5-1
Determination of Tape Characteristics
for the TUlO

5-6

MAGNETIC TAPE DRIVERS

SET 1600 BPI ___y_e_s_ _ _~

SET 800 BPI

yes

SET
EVEN PARITY

Figure 5-2
Determination of Tape Characteristics
for the TU16

5-7

MAGNETIC TAPE DRIVERS
5.4

STATUS RETURNS

The error and status conditions listed in Tabel 5-4
the Magtape drivers described in this chapter.

are

returned

Table 5-4
Magtape Status Returns
Code

rs.sue

Reason
Successful completion
The operation specified in the QIO directive was
completed successfully.
The second word of the I/O
status block can be examined to determine the number of
bytes processed, if operation involved reading or
writing. This code is also returned if nbs equals zero
in an IO.SPB function or if nes equals zero in an
IO.SPF function.

IS.PND

I/O request pending

I The operation specified in the QIO
yet been completed.
with zeros.
IE.ABO

The

directive has not
I/O status block is filled

Operation aborted
The specified I/O operation was cancelled via IO.KIL
while in progress or while still in the I/O queue.

IE.BBE

Bad block
A bad block was encountered while reading or writing
and the error persists after nine retries. The number
of bytes transferred is returned in the second word of
the I/O status block.
For TMll, IE.BBE may also
indicate that a bad tape error (BTE)
has
been
encountered while reading or spacing.

IE.BYT

Byte-aligned buffer specified
Byte alignment was specified for a buffer, but only
word alignment is legal for Magtape. Alternatively,
the length of a buffer is not an even number of bytes.

IE.DAA

Device already attached
The physical device unit specified in an
IO.ATT
function was already attached by the issuing task.
This code indicates that the issuing task has already
attached the desired physical device unit, not that the
unit was attached by another task.

IE.DAO

Data overrun
On a read, a record exceeded the

stated

buffer

size.

I The final portion of the buffer is checked for parity,
I but is not read into memory.

5-8

MAGNETIC TAPE DRIVERS
Table 5-4 (Cont.)
Magtape Status Returns
Code

Reason

IE.DNA

Device not attached
The physical device unit specified in an
IO.DET
function was not attached by the issuing task. This
code has no bearing on the attachment status of other
tasks.

IE.DNR

I . While trying to read or space,
I

IE.EOF

the

driver

detected

. The "LOAD" switch on the physical drive was
to the off position.

switched

blank tape .

End-of-file encountered
An end-of-file (tapemark) was encountered.

IE.EQT

End-of-tape encountered
The
end-of-tape
(physical
end-of-volume)
was
encountered while the tape was moving in the forward
1 directione
A ten-foot length of tape is provided past
EOT to be used for writing data and markers, such as
volume trailer labels. The IE.EQT code will continue
to be returned in the I/O status block until the EOT
marker is passed in the reverse direction.

IE.EOV

End-of-volume encountered
On
a
forward
spacing
function,
the
logical
end-of-volume was encountered. An end-of-volume is two
consecutive
end-of-file
marks
(EOF) ,
or
a
beginning-of-tape mark (BOT) followed by an EOF. The
tape is normally left positioned between the two marks.

IE.FHE

Fatal hardware error
Nonrecoverable hardware malfunction.

IE.IFC

Illegal function
A function code was specified in an I/O request that is
illegal for Magtape.

5-9

MAGNETIC TAPE DRIVERS
Table 5-4 {Cont.}
Magtape Status Returns
Code
IE. OFL

Reason
Dev ice off-line
The physical device unit associated with the LUN
specified in the QIO directive was not on-line. When
the system was booted, a device check indicated that
this physical device unit was not in the configuration.

IE. SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task. For Magtape, this code is also returned
if a byte count of zero was specified or if the user
attempted to write a block that was less than 14 bytes
long.

IE. VER

Unrecoverable error
After the system's standard number of retries has been I
attempted upon encountering an error, the operation
still could not be completed. For Magtape, this code
is returned in the case of CRC or checksum errors or
when a tape block could not be read.

IE .WLK

Write-locked device
The task attempted to write on a Magtape unit that was
physically
write-locked.
Alternately,
tape
characteristic bit 6 was set by the software to
write-lock the unit logically.

After read and write functions, the second I/O status word contains
the number of bytes actually processed by the function. After spacing
functions, it contains the number of blocks of files spaced over. The
EOF mark counts as one block. If an EOF mark is encountered by a read
operation, the second I/O status word will contain an octal two.

5.4.1

Select Recovery

If a request fails because the desired unit is off-line, no drive has
the desired unit number, or has its power off, the following message
is output on the operator's console:

*** MTn:

SELECT ERROR

Where n is the unit number of the specified drive. The driver checks
the unit for readiness and repeats the message every 15 seconds until
the requesting task is aborted or the unit is made available. In the
latter case, the driver then proceeds with the request.

5-10

MAGNETIC TAPE DRIVERS
5.4.2

Retry Procedures for Reads and Writes

If an error occurs during a read (e.g., vertical parity error}, the
recovery procedure depends on the type of magtape in use. A bad tape
error on a TU10/TS03 results in an immediate return of the error code
IE.BEE.
All other read errors for both the TU10/TS03 and TU16 are
retried by backspacing one record and then rereading the record in
question.
If the error persists after nine retries, IE.EBE is
returned.
Write recovery is the same for both the TU10/TS03 and TU16.
When a
write operation fails, the driver attempts to avoid the bad spot on
the tape by means of an extended interrecord gap (IRG).
This means
that it backspaces, makes the IRG just before the record three inches
longer, and then retries the write. If the error persists after nine
retries, IE.BEE is returned. The requesting task can use IO.STC to
prohibit writing with an extended interrecord gap.
In this case,
IE.EBE is returned as soon as a write fails.

5.5

PROGRAMMING HINTS

This
section
considerations
chapter.

5.5.1

contains
information
on
important
programming
relevant to users to Magtape drivers described in this

Block Size

Each block must contain an even number of bytes, at least 14 for a
write and at most 65,534. It is more reasonable, however, to work
with a block size of approximately 2,048 bytes.

5.5.2

Importance of Resetting Tape Characteristics

A task that uses Magtape should always set the tape characteristics to
the proper value before beginning I/O operations. The task cannot be
certain in what state a previous task left these characteristics.
It
is also possible that an operator might have changed the Magtape unit
selection. If the selection switch is changed, the new physical
device unit may not correspond to the characteristics of the unit
described by the respective unit control block.

5.5.3

Aborting a Task

If a task is aborted while waiting for a Magtape unit to be
the Magtape driver recognizes this fact within one second.

5.5.4

selected,

Writing an Even-Parity Zero

If an even-parity zero were written normally, it would appear to the
drive as blank tape.
It is therefore converted to 20 (octal). If
this conversion is undesirable, the user must ensure that
no
even-parity zeros are output on the tape.

5-11

CHAPTER 6
CASSETTE DRIVER

6.1

INTRODUCTION

RSX-llM supports the TAll magnetic tape cassette (a TAll controller
with a TU60 dual transport).
Programming for cassette is quite
similar to programming for Magtape (see Chapter 5). The TAll system
is a dual-drive, reel-to-reel unit designed to replace paper tape.
Its two drives run nonsimultaneously, using Digital Proprietary
Philips-type cassettes.
The maximum capacity of a cassette, in bytes, is 92,000 (minus 300 per
file gap and 46 per interrecord gap). It can transfer data at speeds
of up to 562 uyL~b per second. Recording density ranges from 350 to
700 bits ber inch, depending on tape postion.

6.2

GET LON INFORMATION MACRO

Word 2 of the buffer filled by the GET LON INFORMATION system
directive (the first characteristics word) contains the following
information for cassettes. A bit setting of 1 indicates that the
described characteristic is true for cassettes.
Setting

Meaning

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

6-1

CASSETTE DRIVER
Words 3 and 4 of the buffer are undefined;
default buffer size, for cassettes 128 bytes.

6. 3

word

indicates

the

QIO MACRO

This section summarizes standard and device-specific QIO functions for
the cassette driver.

6.3.1

Standard QIO Functions

Table 6-1 lists the standard functions of the QIO macro that are valid
for cassette.
Table 6-1
Standard QIO Functions for Cassette
Format

Function

QIO$C IO. ATT, •.•

Attach device

QIO$C IO. DET, •••

Detach device

QIO$C IO. KIL I • • •

Cancel I/O requests

QIO$C IO.RLB, ••• ,<stadd,size>

Read logical block
(read tape into buffer)

QIO$C IO.RVB, ••• ,<stadd,size>

Read virtual block
(read tape into buffer)

QIO$C IO.WLB, ••. ,<stadd,size>

Write logical block
(write buffer contents to
tape)

QIO$C IO.WVB, •.. ,<stadd,size>

Write virtual block
(write buffer contents to
tape)

where:

stadd

is the starting address of the data buffer (may be on a
byte boundary) •

size

is the data buffer size in bytes (must be greater
zero) •

than

IO.KIL does not affect in progress-requests.

6.3.2

Device-Specific QIO Functions

Table 6-2 lists the device-specific functions of the QIO macro that
are valid for cassette.
The section on programming hints below
provides more detailed information about certain functions.

6-2

CASSETTE DRIVER
Table 6-2
Device-Specific QIO Functions for Cassette
Format

Function

QIO$C IO. EOF, ••.

Write end-of-file gap

QIO$C IO.RWD, ••.

Rewind unit

QIO$C IO.SPB, .•• ,<nbs>

Space blocks

QIO$C IO.SPF, ••• ,<nes>

Space files

where:

6.4

nbs

is the number of blocks to space
forward, negative if reverse j •

past

(positive

nes

is the number of EOF gaps to space
forward, negative if reverse).

past

(positive

are

returned

STATUS RETURNS

The error and status conditions listed in Table 6-3
the cassette driver described in this chapter.
Table 6-3
Cassette Status Returns
Code

Reason

rs.sue

Successful completion
The operation specified in the QIO directive was
completed successfully.
The second word of the I/O
status block can be examined to determine the number of I
bytes processed, if the operation involved reading or
or the number of blocks or files spaced, if
I writing,
the operation involved spacing blocks or files.
1

IS.PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed. The I/O status block is filled with
zeros.

IE .ABO

Operation aborted
The specified I/O operation was
while still in the I/O queue.

IE. DAA

cancelled

via

IO.KIL

6-3

CASSETTE DRIVER
Table 6-3 (Cont.)
Cassette Status Returns
Code
IE. DAO

Reason
Data overrun
The driver was not able to sustain
required by the TAll controller.

IE. DNA

the

data

rate

Device not attached
The physical device unit specified by an
IO. DET
function was not attached by the issuing task. This
code has no bearinq on the attachment status of other
tasks.

IE. DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.
This code is returned to indicate one of the following
conditions:

I . The cassette has not been physically inserted .
. The unit is off-line .
• A timeout occurred on the physical device unit (i.e.,
an interrupt was lost).
IE. EOF

End-of-file encountered
An end-of-file gap was recognized on the cassette tape.
This code is returned if an EOF gap is encountered
during a read or if the cassette is physically removed
during an I/O operation.

IE. EOT

End-of-tape encountered
While reading or writing, clear trailer at end-of-tape
(EOT) was encountered. Unlike Magtape, writing beyond
EOT is not permitted on cassettes. This condition is
always sensed on a write before it would be sensed on a
read of the same section of tape.
If IE.EQT is
returned
during
a write, the cassette head has
encountered EOT before finishing the writing of the
last block.
It is recommended that the user rewrite
the block on another cassette in its entirety.

IE. IFC

Illegal function
A function code was specified in an I/O request that is
illegal for cassette.

IE. OFL

6-4

CASSETTE DRIVER
Table 6-3 (Cont.)
Cassette Status Returns
Code

Reason

IE.SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task. Alternately, a byte count of zero was
specified on a transfer.

IE.VER

Nonrecoverable error
This code is returned when a block check error occurs
(see section 6.6.5).
The cyclic redundancy check
(CRC), a two-byte value located at the end of each
block, is a checksum that is tested during all read
operations to ensure that data is read correctly. This
is returned if a read request did not specify exactly
the number of bytes of data in the record on tape.
If
a nonrecoverable error is returned, the user may
attempt recovery by spacing backward one block and
retrying the read operation.

IE.WLK
The task attempted to write on a cassette unit that was
physically write-locked.
This code may be returned
after an IO.WLB, IO.WVB, or IO.EOF function.

6.4.1

Cassette Recovery Procedures

If an error occurs during a read or write operation, the operation
should be retried several times. The recommended maximum number of
retries is nine for a read and three for a write because each retry
involves backspacing, which does not always position the tape in the
same place. More than three retries of a write operation may destroy
previously written data. For example, to retry a write, it is best to
space two blocks in reverse, then space one block forward.
This
insures the tape is in the proper position to rewrite the block that
encountered the error.
After read and write functions, the second I/O status word contains
the number of bytes actually processed by the function. After spacing
functions, it contains the number of blocks or files actually spaced.

6.5

STRUCTURE OF CASSETTE TAPE

Figure 6-1 illustrates a general structure for cassette
different structure can be employed if the user wishes.

tape.

Here the tape consists of blocks of data interspersed with sections of
clear tape that serve as leader, trailer, interrecord gaps (IRGs), and
end-of-file gaps.

6-5

CASSETTE DRIVER
The logical end-of-tape in this case consists of a sentinel label
record, rather than the conventional oroup of end-of-file gaps. Each
file must contain at least one block. The size of each block depends
upon the number of bytes the user specifies when writing the block.

IRGs
EOT

BOT

\-~~~~-""""-~~~~-)
FILE 1

\..__~~---'V'-~~~~J''---'
FILE 2
LEOT

150 FEET

Figure 6-1
Structure of Cassette Tape

Meaning

Abbreviation

6.6

Clear leader

BOT

Physical beginning-of-tape

LPG

Load point gap (blan~ tape written
before the first retrievable record)

File label record

REC

Fixed-length record (data)

EOF

End-of-file gap

IRG

Interrecord gap

SLR

Sentinel label record

LEOT

Logical end-of-tape

EQT

Physical end-of-tape

Clear trailer

driver

PROGRAMMING HINTS

This
section
considerations
this chapter.

information
on
important
programming
contains
relevant to users of the cassette driver described in

6-6

CASSETTE DRIVER
6.6.l

Importance of Rewinding

The first cassette operation performed on a tape must always be a
rewind to ensure that the tape is positioned to a known place. When
it is positioned in clear tape there is no way to determine whether it
is in leader at the beginning-of-tape (BOT) or in trailer at the
end-of-tape (EOT).

6.6.2

End-of-File and IO.SPF

The hardware senses end-of-file (EOF) as a timeout.
When IO.SPF is
issued in the forward direction (nes is positive), the tape is
positioned two-thirds of the way from the beginning of the final file
gap.
.LIJ
effect,
this
is all
the way through the file gap. When
IO.SPF is issued in the reverse direction (nes is negative), the tape
is positioned one-third of the way from the beginning of the final
file gap (i.e., two thirds of the way from the beginning of the last
file spaced). Therefore to correctly position the tape for a read or
write after issuing IO.SPF in reverse, the user should issue IO.SPB
forward for one block, followed by IO.SPB in reverse for one block.
6.6.3

The Space Functions, IO.SPB and IO.SPF

IO.SPB always stops in an IRGj IO.SPF in an EOF gapse
Neither space
function actually takes effect until data is encountered.
For
example, suppose the tape is positioned in clear leader at BOT and the
user requests that one block be spaced forward. The drive passes over
the remaining leader until it reaches data, passes one block, and
stops in the IRG. Similarly, if the same command is issued when the
tape is at BOT on a blank tape or a tape containing only EOF gaps, the
function does not terminate until EOT.

6.6.4

Verification of Write Operations

Certain errors, such as cyclic redundancy check, are detected on
but not write operations.
Therefore, to ensure reliability
recording, it is recommended that the user perform a read
verification of every write operation.

6.6.5

of
as

Block Length

The user must specify the exact number of bytes per block when
requesting read or write operations. An attempt to read a block with
an incorrect byte count causes an unrecoverable error
(see section
6.4) to occur.

6.6.6

Logical End-of-Tape

The conventional method of signaling logical end-of-tape by multiple
EOF gaps is inadequate for cassettes. This is because multiple EOF
gaps are not distinguishable from each other.
For example, two
sequential EOF gaps would be read as three instead of two. Also
spacing functions, since they are triggered by encountering data, can
not recognize multiple EOF gaps. Consequently, the use of a sentinel
or key record to signal logical end-of-tape is recommended.
6-7

CHAPTER 7
LINE PRINTER DRIVER

7.1 INTRODUCTION
The RSX-llM line printer driver supports the line printers summarized
in Table 7-1.
Subsequent sections of this chapter describe these
printers in greater detail.
Table 7-1
Standard Line Printer Devices
Model
--

Column Width

Character Set

Lines per Minute

LPll-F

170-1110

LPll-H

170-1110

LPll-J

132

170-1110

LPll-K

132

170-1110

LPll-R

132

1110

LPll-S

132

1110

LPll-V

132

300

LPll-W

132

300

LSll

132

60-200

LVll

132

500

7.1.l LPll LINE PRINTER DRIVER
The LPll is a high-speed line printer available in a variety of
models. The entire LPll model line consists of impact printers, using
one hammer per column and a revolving drum with upper-case and
optional lower-case characters.
The LPll-R and LPll-S are fully
buffered models which operate at a standard speed of 1110 lines per
minute. The other LPll models have 20-character print buffers. These
printers are therefore able to print at full speed if the printed line
is no longer than 20 characters. Lines which exceed this maximum are
printed at a slower rate. Forms with up to six parts may be used for
multiple copies.

7-1

LINE PRINTER DRIVER
7.1.2

LSll Line Printer

The LSll is a medium-speed line printer. It has a 20-character print
buffer, and lines of 20 characters or less are printed at a rate of
200 lines per minute. Longer lines are printed at a slower rate.
RSX-llM does not support the LSll expanded character set feature.

7.1.3 LVll Line Printer
The LVll is a fully-buffered, electrostatic printer-plotter which
operates at a standard rate of 500 lines per minute. RSX-llM supports
only the LVll print capability, not the plotter mode.

7.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) contains the following
information for line printers. A bit setting of 1 indicates that the
described characteristic is true for line printers.
Bit

Settin9

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications channel

Device mountable as a FILES-11 volume

Device mountable

Meaning

Words 3 and 4 of the buffer are undefined;
word 5 indicates the
default size for the device, for line printers the width of the
printer carriage (i.e., 80 or 132).

7.3 QIO MACRO
Table 7-2 lists the standard functions of the QIO macro that are valid
for line printers.

7-2

LINE PRINTER DRIVER
Table 7-2
Standard QIO Functions for Line Printers
Format

Function

QIO$C IO.ATT, .••

Attach device

QIO$C IO. DET I • • •

Detach device

QIO$C IO.KIL, ..•

Cancel I/O requests

QIO$C IO.WLB, .•• ,<stadd,size,vfc>

Write logical block
(print buffer contents)

QIO$C IO.WVB, •.• ,<stadd,size,vfc>

Write virtual block
(print buffer contents)

where: stadd

is the starting address of the data buffer (may be on a
byte boundary) •

size

is the data buffer size in bytes (must be greater
zero) •

than

vfc

is a vertical format control character from Table 7-4.

IO.KIL does not cancel an in progress request unless the line printer
is in an offline condition because of a power failure or a paper jam
or because it is out of paper.
The line printer driver supports no device-specific functions.

7.4 STATUS RETURNS
Table 7-3 lists the error and status conditions that are
the line printer driver described in this chapter.

returned

Table 7-3
Line Printer Status Returns
Code
IS.SOC

IS. PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed.
The I/O status block is filled
with zeros.

7-3

LINE PRINTER DRIVER
Table 7-3 (Cont.)
Line Printer Status Returns
Code

Reason

IE .ABO

Operation aborted
The specified I/O operation was cancelled
progress or while in the I/O queue.

IE. DAA

while

Device already attached
The physical device unit specified in an IO.ATT
function was already attached by the issuing task.
This code indicates that the issuing task has already
attached the desired physical device unit, not that
the unit was attached by another task.

IE. DNA

Device not attached
The physical device unit specified an IO.DET function
was not attached by the issuing task. This code has
no bearing on the attachment status of other tasks.

I IE. IFC

Illegal function
A function code was specified in an I/O
is illegal for line printers.

IE. OFL

request

that I

Device off-line
The physical device unit associated with the LUN
specified in the QIO directive was not on-line. When
the system was booted, a device check indicated that
this
physical
device
unit
was
not
in
the
configuration.

IE. SPC

Illegal address space
The buffer specified for a write request was partially
or totally outside the address space of the issuing
task.
Alternately, a byte count
of
zero
was
specified.

7.4.1 Ready Recovery
If any of the following conditions occur:
Paper jam
Printer out of paper
Printer turned off-line
Power failure
the driver determines that the line printer is off-line,
following message is output on the operator's console:
***T.Pn:

-- NOT RRADY

7-4

and

the

LINE PRINTER DRIVER
where n is the unit number of the line printer that is not ready. The
driver retries the function which encountered the error condition from
the beginning, once every second. It displays the message every 15
seconds until the line printer is readied. If a power failure occurs
while printing a line, the entire line is reprinted from the beginning
when power is restored.

7.5 VERTICAL FORMAT CONTROL
Table 7-4 summarizes the meaning of all characters used for vertical
format control on the line printer. Any one of these characters can
be specified as the vfc parameter in an IO.WLB or IO.WVB function.
Table 7-4
Vertical Format Control Characters
Octal
Value

Character

040

blank

I 060

I zero

DOUBLE SPACE: output two line feeds,
print tne
contents of the buffer, and output a carriage
return.
Normally, the buffer
contents
are
printed two lines below the previously printed
line.

061

one

; rdoutput a
feed, -n,...
the I
.l::"lcontents of the buffer, and output a carriage
return. Normally, the contents of the buffer are
printed on the first line of the next page.

053

I plus

OVERPRINT: print the contents of the buffer and
perform a carriage return, normally overprinting
the previous line.

Meaning
SINGLE SPACE: output a line feed, print the
contents of the buffer, and output a carriage
return. Normally, printing immediately follows
the previously printed line.

PAGE

EJECT:

4=~-~

LVl.111

.LL.I.'-'

044

dollar
sign

PROMPTING OUTPUT: output a line feed and then
print the contents of the buffer.

000

null

INTERNAL VERTICAL FORMAT:
the buffer contents
are printed without addition of vertical format
control characters. In this mode, more than one
line of guaranteed contiguous output can be
printed per I/O request.

All other vertical format control characters are interpreted as blanks
{octal 040).

7.6 PROGRAMMING HINTS
This
section
contains
information
on
important
programming
considerations relevant to users of the line printer driver described
in this chapter.

7-5

LINE PRINTER DRIVER
7.6.1 RUBOUT Character

The line printer driver discards the ASCII character code 177 during
output, because a RUBOUT on the LSll printer causes a RUBOUT of the
hardware print buffer.

7.6.2 Print Line Truncation

If the number of characters to be printed exceeds the width of the
print carriage, the driver discards excess characters until it
receives one that instructs it to empty the buffer and return to
horizontal position 1.
The user can determine that truncation will
occur by issuing a GET LUN INFORMATION system directive and exam1n1ng
word 5 of the information buffer. This word contains the width of the
print carriage in bytes.

7.6.3 Aborting a Task

If a task is aborted while waiting for the line printer to be readied,
the line printer driver recognizes this fact within one second.

7-6

CHAPTER 8
CARD READER DRIVER

8. 1 INTRODUCTION
The RSX-llM card reader driver supports the CRll card reader.
This
reader is a virtually jam-proof device which reads EIA standard
80-column punched cards at the rate of 300 per minute. The hopper can
hold 600 cards.
This device uses a vacuum picker which provides
extreme tolerance to damaged cards and makes card wear insignificant.
Cards are riffled in the hopper to prevent sticking. The reader uses
a strong vacuum to deliver the bottom card. It has a very short card
track, so only one card is in motion at a time.

8.2 GET LUN'INFORMATION MACRO
Word 2 of the buffer filled by the GET LON INFORMATION system
directive (the first characteristics word) contains the following
information for card readers. A bit setting of l indicates that the
described characteristic is true for card readers.
Meaning

Bit

Setting

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mount.able as a communications
channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 of the buffer are undefined;
word 5 indicates
default buffer size, which is 80 bytes for the card reader.

8-1

the

CARD READER DRIVER
8. 3 QIO MACRO
This section summarizes standard and device-specific QIO functions for
the card reader driver.

8.3.l Standard QIO Functions
Table 8-1 lists the standard functions of the QIO macro that are valid
for the card reader.
Table 8-1
Standard QIO Functions for the Card Reader
Format

Fune tion

QIO$C IO. ATT, ••.

Attach device

QIO$C IO. DET, ...

Detach device

QIO$C IO.KIL, •..

Cancel I/O requests

QIO$C IO.RLB, ..• ,<stadd,size>

Read logical block
(alphanumeric)

QIO$C IO.RVB, •.• ,<stadd,size>

Read virtual block
(alphanumeric)

where:

stadd

is the starting address of the data buffer (may be on
a byte boundary) •

size

is the data buffer size in bytes (must be greater than
zero).

IO.KIL does not cancel an in progress request unless the card reader
is in an offline condition because of a pick, read, stack or hopper
check, because of power failure, or because the RESET button has not
been depressed.

8.3.2 Device-Specific QIO Function
The device-specific function of the QIO macro that is
card reader is shown in Table 8-2.

valid

for

the

stadd

is the starting address of the data buffer (may be
a byte boundary) •

size

is the data buffer size in bytes (must be greater than
zero).

Table 8-2
Device-Specific QIO Function for the Card Reader
Format

Function

QIO$C IO.RDB, ... ,<stadd,size>
where:

Read logical block (binary)

8-2

CARD READER DRIVER
8.4 STATUS RETURNS
There are a wide var1ecy of error conditions and recovery procedures
related to the use of the card reader. This section describes the
three major ways in which the system reports error conditions.
1.

Lights and indicators on the card reader panel are turned on
or off to indicate particular operational problems such as
read, pick, stack or hopper checks. Switches are available
to turn the reader power on and off and to allow the user to
reset after correcting an error condition.

A message is output on the operator's console if
checks or power problems occur.

An !/O completion code is returned in the low-order byte of
the first word of the I/O status block specified in the QIO
macro to indicate success or failure on completion of an I/O
function.

operational

The following subsections describe each of these returns in detail.

8.4.l Card Input Errors and Recovery
The table included below describes all external lights and switches
used to indicate to the operator that a hardware problem has occurred
and must be corrected. There are two classes of hardware errors:
Those requiring the operator to ready the reader and
operation again.

try

the

Those requiring the operator to remove the last card from the
output stacker, to replace it in the input hopper, and to try
the operation again.
In the first case, the card reader was unable to read the current
card.
In the second, the card was read incorrectly and must be
The
reader driver
automatically restarts a read operation within one second after the
cards have been replaced in the input hopper.
Table 8-3 summarizes the functions of lights and indicators on the
front panel of the card reader.
It discusses common operational
errors which might be encountered while reading cards and recovery
procedures associated with these error conditions.

8-3

CARD READER DRIVER
Table 8-3
Card Reader Switches and Indicators
Indicator

Description

POWER
switch

pushbutton
indicator
switch
(alternate
action:
pressed for
both ON and
OFF)

READ
CHECK
indicator

white light

Action

Recovery

Card may have been
read incorrectly;
restore power if
possible by depressing the POWER
When indicator is
off, depressing switch switch; insert the
applies power to
card again as the
reader and causes
first card in the
input hopper, and
associated indicator to light.
press the RESET
switch; in some
When indicator is
cases, it may be
lit, depressing
necessary to
restart the program.
switch removes all
power from reader and
causes indicator to
go out.

Controls application
of all power to the
card reader.

When lit, this light
indicates that the
card just read may be
torn on the leading or
trailing edges, or
that the card may
have punches in
column positions 0
or 81.
Because READ CHECK
indicates an error
condition, whenever
this indicator is
lit, it causes the
card reader to stop
operation and extinguishes the RESET
indicator.

8-4

Card was read incorrectly; duplicate if
necessary, insert
the card again as
the first card in the
input hopper and
press the RESET
switch.

CARD READER DRIVER
Table 8-3 (Cont.}
Card Reader Switches and Indicators
Ind ica. tor

Description

PICK
CHECK
indicator

white light

Action
When lit, this light
indicates that the
card reader failed to
move a card into the
read station after
it received a READ
COMMAND from the
controller.

Recovery
Card could not be
read; press the
RESET switch to try
again or remove the
cards from the input
hopper, smooth the
leading edges, replace, and then
press the RESET
switch.

Stops card reader
operation and extinguishes RESET
indicator.
STACK
CHECK
indicator

white light

When lit, this light
indicates that the
previous card was not
properly seated in
the output stacker
and therefore may be
badly mutilated.
Stops card reader
operation and extinguishes RESET
indicator.

Card may have been
read incorrectly and
is not positioned
properly in the output stacker; duplicate the card if it
is damaged; insert
the card again as
the first card in
the input hopper and
press the RESET
switch.

HOPPER
CHECK
indicator

white light

When lit, this light
Card may have been
indicates that either read incorrectly;
the input hopper is
empty the stacker or
empty or that the out- fill the hopper; input stacker is full.
sert the card again
as the fist card in
the input hopper and
press the RESET
switch.

STOP
switch

momentary
pushbutton/
indicator
switch
(red light)

When depressed,
immediately lights
and drops the READY
line, thereby extinguishing the RESET
indicator. Card
reader operation then
stops as soon as the
card currently in the
read station has been
read.
This switch has no
effect on the system
power; it only stops
the current operation.

8-5

CARD READER DRIVER
Table 8-3 (Cont.)
Card Reader Switches and Indicators
Indicator

Description

RESET
switch

momentary
pushbutton/
indicator
switch
(green
1 ight)

Act ion

Recovery

When depressed and
released, clears all
error flip-flops and
initializes card
reader logic. Associated RESET indicator
lights to indicate
that the READY signal
is applied to the controller.
The RESET indicator
goes out whenever the
STOP switch is depressed or whenever
an error indicator
lights (READ CHECK,
PICK CHECK, STACK
CHECK, or HOPPER
CHECK) .

8.4.2 Ready and Card Reader Check Recovery
If any of the following conditions occurs:
POWER failure
reset switch not pressed (reader offline)
the driver determines that the card reader is not ready, and the
following message is output on the operator's console:

*** CRn: -- NOT READY
If any of the following conditions occurs:
Pick error (PICK CHECK)
Read error (READ CHECK)
Output stacker error (STACK CHECK)
Input hopper out of cards (HOPPER CHECK)
Output stacker full (HOPPER CHECK)
the driver determines that a card reader check has occurred,
following message is output on the operator's console:

and

the

*** CRn: -- READ FAILURE. CHECK HARDWARE STATUS
where n is the unit number of the card reader that is not ready.
The
operator should correct the error and press RESET:
the driver

8-6

CARD READER DRIVER
attempts the function from the beginning, once every second.
It
displays the message once every 15 seconds until the card reader is
readied. In all cases except pick error, the last card read should be
reinserted in the input hopper, as
8.4.l.

8.4.3 I/O Status Conditions

The error and status conditions listed in Table 8-4
the card reader driver described in this chapter.

are

returned

Table 8-4
Card Reader Status Returns
Code

Reason

rs.sue

IS.PND

I/0 request pending

I The opera ti on specified in the QIO directive
yet been executed.
with zeros.
IE .ABO

has not
I/O status block is filled

Operation aborted
i

IE. DAA

The specified I/O operation was cancelled
progress or while still in the I/O queue.

while

Device already attached

IE. DNA

The

The physical device unit specified in an IO.ATT
function was already attached by the issuing task.
This code indicates that the_is~u~ng_task ~as alre~dy
attached the desired physical device unit, not that
the unit was attached by another task.
Device not attached
The physical device unit specified in an IO. DET
function was not attached by the issuing task. This
code has no bearing on the attachment status of other
tasks.

IE. EOF

End-of-file encountered
An end-of-file control card was recognized.

IE. IFC

Illegal function
A function code was specified in an I/O
is illegal for card readers.

8-7

request

that

CARD READER DRIVER
Table 8-4 (Cont.)
Card Reader Status Returns
Code

Reason

IE.NOD

Buffer allocation failure
Dynamic storage space has been depleted, and there was
insufficient buffer space available to allocate a card
buffer (i.e., cards are read into a driver buffer,
translated, and then moved to the user buffer).

IE.OFL

Device off-line
The physical device unit associated with the LUN
specified in the QIO directive was not on-line. When
the system was booted, a device check indicated that
this
physical
device
unit
was
not
in
the
configuration.

IE.SPC

Illegal address space
The buffer specified for a read request was partially
or totally outside the address space of the issuing
task.
Alternately, a byte count
of
zero
was
specified.

8.5 FUNCTIONAL CAPABILITIES
The card reader driver can perform the following functions:
1.

Read cards in DEC026 format and translate to ASCII.

Read cards in DEC029 format and translate to ASCII.

Read cards in binary format.

If the QIO macro specifies the IO.RLB or IO.RVB function, the driver
interpets all data as alphanumeric (026 or 029 format). As explained
below, control characters indicate whether 026 or 029 is desired.
If
the QIO macro specifies IO.ROB, the driver interprets all data,
including 026 and 029 control characters, as binary.

8.5.1 Control Characters
Table 8-5 lists the multipunched cards that the card reader driver
recognizes as control characters. They are never transferred to the
user's buffer or included in the count of transferred bytes in
alphanumeric mode. In binary mode the only control card recognized is
binary EOF.

8-8

CARD READER DRIVER
Table 8-5
Card Reader Control Characters
Punches

Columns

Meaning

End-of-file (alphanumeric)

12-11-0-1-6-7-8-9
12-11-0-1-6-7-8-9

(all 8 punches in
the first 8 columns)

End-of-file (binary)

12-2-4-8

026-coded cards follow

12-0-2-4-6-8

029-coded cards follow

DEC026 is the default translation mode
when
the
system
is
bootstrapped. This mode remains in errect unt11 exp11c1tiy changed by
a control card indicating that DEC029 cards will follow.
After
encountering a DEC029 control card, the driver translates all cards in
DEC029 format unless another DEC026 control card is encountered. This
card
overrides
the 029 mode specification and indicates that
subsequent cards are to be translated in 026 format.
Control
characters are addressed to the card reader itself, and remain in
effect even when the reader is attached and subsequently detached.

8.6 CARD READER DATA FORMATS
The card reader reads data in either alphanumeric or binary format.

8.6.l Alphanumeric Format (026 and 029)
Table 8-6 summarizes the translation from DEC026 or DEC029 card
to ASCII.

codes

Table 8-6
Translation from DEC026 or DEC029 to ASCII

Character

{
}
SPACE

$
%

AND
(
)

NonParity
ASCII
173
175
040
041
042
043
044
045
046
047
050
051
052
053

DEC029

DEC026

12 0

11 0

none
12 8 7
8 7
8 3
11 8 3
0 8 4
12
8 5
12 8 5
11 8 5

none
12 8 7
0 8 5
0 8 6
11 8 3
0 8 7
11 8 7
8 6
0 8 4
12 8 4
11 8 4
12

11 8 4
12 8 6

8-9

Character

0
1
2

3
4
5
6
7
8

NonParity
ASCII
054
055
056
057
060
061
062
063
064
065
066

DEC029

DEC026

0 8 3

12 8 3
0 1
0
1
2
3
4
5
6

067

070
071

8
9

CARD READER DRIVER
Table 8-6 (Cont.)
Translation from DEC026 or DEC029 to ASCII

Character

<
>
?
@

A
B

D
E

F
G
H
I
J
K
L

NonParity
ASCII

DEC029

DEC026

Character

072
073
074
075
076
077
100
101
102
103
104
105
106
107
110
111
112
113

8 2
11 8 6
12 8 4
8 6
0 8 6
0 8 7
8 4
12 1
12 2
12 3
12 4
12 5
12 6
12 7
12 8
12 9
11 1
11 2

M
N
0
p

114

11 3

11 8 2
0 8 2
12 8 6
8 3
11 8 6
12 8 2
8 4
12 1
12 2
12 3
12 4
12 5
12 6
12 7
12 8
12 9
11 1
11 2
11 3

T
u

w
x
y

z
[

]

or t
_ or+-

NonParity
ASCII
115
116
117
120
121
122
123
124
125
126
127
130
131
132
133
134
135
136
137

DEC029

DEC026

11 4
11 5

11 4
11 5
11 6
11 7
11 8
11 9
0 2
0 3
0 4
0 5
0 6
0 7
0 8
0 9
11 8 5
8 7
12 8 5
8 5
8 2

11 6
11 7
11 8
11 9
0 2
0 3
0 4
0 5
0 6
0 7
0 8
0 9
12 8 2
0 8 2
11 8 2
11 8 7
0 8 5

8.6.2 Binary Format
In RSX-llM binary format, the data are not packed, but are transferred
exactly as read, one card column per word. Because each word has 16
bits and each card column represents only 12, the data from the column
are stored in the rightmost 12 bits of the word. The word's remaining
four bits contain zeros.

8.7 PROGRAMMING HINTS
This
section
contains
information
on
important
programming
considerations relevant to users of the card reader driver described
in this chapter. Section 8.4 contains information on operational
error-recovery procedures which might be important from a programming
point of view.

8.7.1 Input Card Limitation
Only one card can be read with a single QIO macro call. A request to
read more than 80 bytes or columns, the length of a single card, does
not result in a multiple card transfer.
Only 80 columns are
processed. It is possible to read fewer than 80 columns of card input
with a QIO read function. The user can specify that only the first 10
columns, for example, of each card are to be read.

8-10

CARD READER DRIVER
8.7.2 Aborting a Task
If a task which is waiting for
the
aborted, the card reader driver
second.

8-11

card reader to
recognizes this

be readied is
fact within one

CHAPTER 9
MESSAGE-ORIENTED COMMUNICATION DRIVERS

9.1

INTRODUCTION

RSX-llM supports a variety of communication line interfaces
synchronous
and
asynchronous,
single-line
and
multiplexers,
character-oriented and message-oriented. These are used for terminal
communications,
remote job entry, multicomputer interfaces, and
laboratory and industrial control cbmmunications. Communications line
interfaces can be roughly divided into two categories:
Terminal (character-oriented) communications devices
Multicomputer (message-oriented) communications devices
Chapter 2 describes the character-oriented asynchronous communications
line interfaces used primarily for terminal communications. The
PDP-11 PERIPHERALS HANDBOOK contains more detail on these devices.
This chapter describes in some detail the RSX-llM message-oriented
synchronous and asynchronous communication line interfaces. These are
used most frequently in multicomputer communications.
Character-oriented communications devices include the DHll, DJll,
DLll-A, DLll-B, DLll-C and DLll-D interfaces. These are asynchronous
multiplexers and single-line interfaces which are
used
almost
exclusively for termimal communications. Transfers on all of these
interfaces are performed one character at a time.
None of the
interfaces in this category have drivers of their own (i.e., they are
supported via the terminal driver), and none can be accessed directly
as RSX-llM devices.
Message-oriented communications line interfaces are used primarily to
link two separate but complementary computer systems. One system must
serve as the transmitting device and the other as the receiving
device.
Devices in this category include the synchronous and
asynchronous single-line interfaces summarized in Table 9-1.
Table 9-1
Message-Oriented Communication
Model

Interfac~s

Fune t ion

DAll-B

Asynchronous, parallel Single-line interface

DLll-E

Asynchronous

Single-line interface

9-1

MESSAGE-ORIENTED COMMUNICATION DRIVERS
Table 9-1 (Cont.)
Message-Oriented Communication Interfaces
Model

Function

DPll

Synchronous

Single-line interface

DQll

Synchronous

Single-line interface

DUll

Synchronous

Single-line interface

The message-oriented communication line interfaces are used
to transfer large blocks of data.

primarily

Whereas the character-oriented interfaces can only be
accessed
indirectly through the terminal driver, the DAll-B, DLll-E, DPll, DQll
and DUll allow I/O requests to be queued directly for them.
These
devices have drivers of their own and can be accessed by means of the
logical device names listed in Table 1-1. These names can be used in
assigning LUNs via the ASSIGN LUN system directive, at task build or
via the REASSIGN MCR command.
The following subsections briefly
discuss the message-oriented interfaces supported for RSX-llM.

9.1.1

DAll-B Synchronous Line Interface

The DAll-B provides a bit parallel, direct memory access interface
between two PDP-11 computer systems. Data transfers are performed a
word at a time and are made directly between the memories of the two
systems. The maximum transfer rate is 500,000 baud, and is adjustable
by the user to match the system configuration requirements.
Being a
parallel device, the DAll-B does not utilize sync characters. The
interface is half-duplex and transfers data in blocks of up to 32K
words.
The DAll-B requires two cooperating computers to effect a data
transfer.
In order to control the physical link between the
computers, the device driver contains its own simple line protocol.
This protocol requires one system to issue a receive QIO and the other
to issue a transmit QIO before any data is actually transferred.

9.1.2

DLll-E Asynchronous Line Interface

The DLll-E is an asynchronous, serial-bit, single-line interface.
It
is a block-transfer device used for remote terminal and multicomputer
communications. Baud rates are selectable between 50 and 9600, and
full data set control is supported.

9.1.3

DPll Synchronous Line Interface

The DPll provides a program interrupt interface between a PDP-11 and a
serial synchronous line.
This interface facilitates the use of the
PDP-11 in remote batch processing, remote data collection, and remote
concentration applications. The modem control feature allows the DPll
to be used in switched or dedicated configurations.

9-2

MESSAGE-ORIENTED COMMUNICATION DRIVERS
On the DPll, baud rates are selectable between 2000 and 19,200.
The
programmer can select a specific sync character which is used to
synchronize the transmitting and receiving systems.

9.1.4

DQll Synchronous Line Interface

The DQll provides a direct memory access interface between a PDP-11
and
a
serial
synchronous
line.
The
direct memory access
characteristic of the DQll allows the device to operate at speeds
higher than those of program interrupt devices, and with a lower
interrupt overhead. Modem control of the DQll allows the device to be
used in switched or dedicated configurations.
The DQll handles data rates from 2400 baud to 1!000!000 baud.
The
limiting rate is determined by the modem and data set interface level
converters.
The DQll sync character is programmable in the same manner as the DPll
and the DUll.
The maximum data block length transmitted is 65,536
characters.

9.1.5

DUll Synchronous Line Interface

The DUll synchronous line interface is a single-line communications
device which provides a program-controlled interface between the
PDP-11 and a serial synchronous line. The PDP-11 can be interfaced
with a high-speed line to perform remote batch processing, remote data
collection, and remote concentration applications. Modem control is a
standard feature of the DUll and allows the device to be used in
switched or dedicated configurations. The DUll transmits data at a
maximum rate of 9600 baud; this rate is limited by modern and data set
interface level converters.
The DUll can be programmed to accept any user-defined sync character.
The use of the sync character is the same for the DUll and the DPll.

9.2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) contains the following
information for message-oriented communication interfaces.
A bit
setting of 1 indicates that the described characteristic is true for
the interfaces described in this chapter.
Bit

Setting

Meaning

Record-oriented device

Carriage-control device

Terminal device

Directory device

9-3

MESSAGE-ORIENTED COMMUNICATION DRIVERS
Bit

Setting

Meaning

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 a.re undefined, and word 5 has a special meaning for
the
DLll-E, DQll, DPll, and the DUll interfaces.
Byte 0 of word 5
contains the number of sync characters to be transmitted before a
synching
message (e.g., after line turn around in half-duplex
operation), and byte 1 is used as a sync counter.
9.3

QIO MACRO

This section summarizes the standard and device-specific functions of
the QIO macro that are valid for the communication interfaces
described in this chapter.

9.3.1

Standard QIO Functions

Table 9-2 lists the standard functions of the QIO macro that are valid
for the communication devices.
Table 9-2
Standard QIO Functions for Communication Interfaces
Function

Format
QIO$C IO. ATT, .•.

Attach device*

QIO$C IO. DET, .•.

:Cetach device

CIO$C IO.KIL, ...

Cancel I/O requests

QI0$C IO.RLB, ... ,<stadd,size>

Read logical block (stripping
sync)

QIO$C IO.WLB, ... ,<stadd,size>

Write logical block (preceded by
syncs)

where:

stadd is the starting address of the data buffer (may be on a
byte boundary) .
size

is the data buffer size in bytes (must be greater than
zero) •

*Only unmounted channels may be attached.
An attempt to attach a
mounted channel will result in an IE.PR! status being returned in the
I/O status duubleword.

9-4

MESSAGE-ORIENTED COMMUNICATION DRIVERS
9.3.2

Device-Specific QIO Functions

The specific functions of the QIO macro that are
communication line interfaces are shown in Table 9-3.

valid

for

the

Table 9-3
Device-Specific QIO Functions for Communication Interfaces
Format

Function

QIO$C IO. FDX

Set device to full-duplex mode

QIO$C IO. HDX, .•.

Set device to half-duplex mode

QIO$C IO. INL, ...

Initialize device
characteristics

QIO$C IO.RNS, •.. ,<stadd,size>

Read logical block, without stripping
sync
characters (transparent mode).
Not applicable to DAll-B or DQll

QIO$C IO.SYN, ••. ,<syn>

Specify sync character

QIO$C IO. TRM, .••

Terminate communication,
from physical channel

QIO$C IO.WNS, ... ,<stadd,size>

Write logical block without preceding
sync
characters (transparent mode).
Not applicable to DAll-B.

where:

and

set

device

disconnecting

stadd is the starting address of the data buffer (may be on a
byte boundary) •
size

is the data buffer size in bytes (must be greater
zero) .

syn

is the sync character, expressed as an octal value.

The device-specific functions listed in Table
greater detail below.

9-3

are

described

than

9.3.2.1 IO.FOX - The IO.FOX QIO function is used to set the mode on a
DLll-E, DPll, DQll, or DUll unit to full-duplex. The IO.FOX function
code can be combined (ORed) with the IO.SYN function code, if desired,
to set the operational characteristics of the physical device unit.

9.3.2.2 IO.HDX - The IO.HDX QIO function is used to set the mode on a
DLll-E, DPll, DQll, or DUll unit to half-duplex. The IO.HDX function
code can be combined (ORed together) with the IO.SYN function code, if
desired, to set the operational characteristics of the physical device
unit.

9.3.2.3

function

IO.INL and IO.TRM - These two QIO functions have the same
code but different modifier bits.
IO.INL is used to

9-5

MESSAGE-ORIENTED COMMUNICATION DRIVERS
initialize a physical device unit for use as a communications link.
It turns the device on-line, sets device characteristics, and ensures
that the appropriate data terminal is ready.
IO.TRM disconnects the
device.
If the device has a dial-up interface, it also hangs up the
line.

9.3.2.4 IO.RNS - The IO.RNS QIO function is used to read a logical
block of data, without stripping the sync characters which may precede
the data. A similar function is IO.RLB, which is non-transparent, in
that it causes sync characters preceding the data message to be
stripped.
IO.RLB is used at the start of a segmented data request, in
which the block might have the following layout:

s s

where:

cs I cs I
7
8

DATA

I cs I

S is a sync character
H is a header character
CS is a validity check character

The programmer must strip sync characters from the beginning of a data
block in this way. Stripping only at the beginning of a read allows a
later character which happens to have the same binary value as a sync
character to be read without stripping. IO.RLB is used to read a
logical block with leading sync characters stripped~ IO.RNS is used
to read the block without stripping leading sync characters. Since
the DAll-B is a parallel device and there are no sync characters, it
treats the latter as if it were IO.RLB. Generally, IO.RLB should be
used.

9.3.2.5 IO.RWD - Completion of the IO.RWD means that rewind has been
initiated and the magtape controller is free. Additional operations
may then be initiated. Requests for the same drive are queued by the
driver until beginning-of-tape (BOT) is reached.

9.3.2.6 IO.SYN - This QIO function allows the programmer to specify
the sync character to be recognized when an IO.RLB or IO.WLB function
is performed. IO.SYN can be combined (ORed together) with IO.HDX to
set the characteristics of the physical device unit.

9.3.2.7 IO.WNS - This QIO function causes a logical block to be
written with no preceding sync characters. To ensure that the two
systems involved in a communication are synchronized, two or more sync
characters are transmitted by one system and received by the other

9-6

MESSAGE-ORIENTED COMMUNICATION DRIVERS
before any other message can be sent. IO.WLB is used to write a block
of data, preceded by sync characters; IO.WNS is used to perform a
block transfer without sending sync characters first.
Since the
DAll-B is a parallel device and there are no sync characters, it
treats the latter as if it were IO.WLB. Generally, IO.WLB should be
used.

9.4

STATUS RETURNS

The error and status conditions listed in Table 9-4 are
the communication drivers described in this chapter.

returned

Table 9-4
Communication Status Returns
Code

Reason

rs.sue

Successful completion
The operation specified in the QIO directive was completed
successfully.
The second word of the I/O status block can
be examined to determine the number of bytes processed, if
the operation involved reading or writing.

IS.PND

I/O request pending

IE.BCC

The operation specified in the QIO directive has not yet
been executed. The I/O status block is filled with zeros.
Block check error
When the Cyclic Redundancy Check (CRC) option is present on
the DQll, a check character is appended to each message
transmitted. The receiver of the messages recalculates the
check character and compares it with the one transmitted.
This error code is returned when the two check characters do
not match, and represents a transmission error.

IE.DAO

Data overrun
Due to UNIBUS traffic or a modem problem, the
DQll
controller was unable to maintain the data rate required to
prevent data loss (i.e., the receipt of another byte before
processing of a previous byte was completed).

IE.DNR

Device not ready
The physical device unit specified in the QIO directive was
not ready to perform the desired I/O operation. This code
is returned to indicate one of the following conditions:
. The physical device unit could not be
the circuit could not be completed).

9-7

initialized

(i.e.,

MESSAGE-ORIENTED COMMUNICATION DRIVERS
Table 9-4 (Cont.)
Communication Status Returns
Code

Reason

IE. DNR

(Cont.)

IE. IFC

The transmission of a character was not followed by an
interrupt within the period of time selected as the device
timeout
period.
This
timeout occurs only when a
transmission is in progress and the interrupt marking
completion of a message does not occur. The appropriate
response to this condition is to attempt to resynchronize
the device by initializing and accepting the next request.
A timeout does not occur on a read.
If the receiving
device is not ready, the transfer will not be initiated by
the transmitting device. Once the transfer is initiated,
however,
it will complete either by satisfying the
reouested byte count or by timing out.
Illegal function
A function code was specified in an I/O request that
illegal for message-oriented communication devices.

IE.OFL

Device off-line
The physical device unit associated with the LUN specified
in the QIO directive was not on-line. When the system was
booted, a device check indicated that this physical device
unit was not in the configuration.

IE. SPC

Illegal address space
The buffer specified for a read or write request was
partially or totally outside the address space of the
issuing task.
Alternately, a byte count of zero was
spec i f ied.

IE. VER

Nonrecoverable error (DAllB only)
The data transfer terminated before all of the data has been
transmitted.
The error code is returned on transmit when
both systems attempt to transmit at the same time.
This
condition is detected by the device protocol. The error
code is returned on receive when the transmit data count of
the
transmitting side does not equal the data count
specified by the receive QIO.

9.5

PROGRAMMING HINTS

This
section
contains
information
on
important
programming
considerations relevant to users of lhe message-oriented communication
interfaces described in this chapter.

9-8

MESSAGE-ORIENTED COMMUNICATION DRIVERS
9.5.1

Transmission Validation

Because there is no way for the transmitting device to verify that the
data block has successfully arrived at the receiving device unless the
receiver responds, the transmitter assumes that any message which is
clocked out on the line (without line or device outage) has been
successfully transmitted. As soon as the receiver is able to satisfy
a read request, it returns a successful status code (IS.SUC) in the
I/O status block. Of course, only the task which receives the message
can determine whether or not the message has actually been transmitted
accurately.
The receiving device should be ready to receive
request) at the time the transmission is sent.

9.5.2

data

(with

read

Redundancy Checking

By the nature of message-oriented communications, only the task which
receives a communication can determine whether or not the message was
received successfully. The transmitter simply transfers data, without
validation of any kind.
It is therefore the responsibility of the
communicating tasks which use the device to check the accuracy of the
transmission. A simple validity check is a checksum-type longitudinal
redundancy check. A better approach to validating data is the use of
a cyclic redundancy check (CRC). A CRC can be computed in software or
with a hardware device, such as the KG-11 communications arithmetic
option.

9.5.3

Half-Duplex and Full-Duplex Considerations

Because there is a single I/O request queue, only one QIO request can
be performed at a time. It is therefore not possible, through QIOs,
for a device to send and receive data at the same time.
Also, since
timeouts are not set LuL receive functions, a receive QIO is
terminated only by receiving a message from the remote system, or by
issuing an IO.KIL QIO for the device. Therefore, if no message is
transmitted by the remote system, a receive will not terminate, and no
further I/O can be performed on that device until the receive is
killed by issuing an IO.KIL QIO.
Both half-duplex and full-duplex lines can be used with the DLll-E,
DUll, DPll, and DQll.
The mode is settable by using IO.FOX for
full-duplex and IO.HDX for half-duplex.
In half-duplex mode, the
modem signal RTS
(Request To Send) is cleared after each "transmit
message". In full-duplex, this signal is always left on.
Using
full-duplex mode eliminates modem delays in transmission, but requires
full-duplex hardware and communication links.
Only half-duplex mode is available with
nature of the hardware.

9.5.4

the

DAll-B

because

the

Low-Traffic Sync Character Considerations

If message traffic on a line
communications device should

is
be

9-9

low, each message sent from a
preceded by a sync train. This

MESSAGE-ORIENTED COMMUNICATION DRIVERS
enables the controller to resynchronize if a message is "broken"
(i.e., part or all of it is lost in transmission}. Correspondingly,
every message received by a communications device under low-traffic
conditions, when messages are not contiguous (back-to-back), should be
read via an IO.RLB (read, strip sync) function.
This requires that
the first character in the data message itself not have the binary
value of the sync character.

9.5.5

Vertical Parity Support

Vertical parity is not supported by the DAll-B, DLll-E, DPll, DQll, or
DUll. Codes are assumed to be eight-bit only.

9.5.6

Importance of IO.INL

After the type of communication line has been determined, and after
IO.SYN has specified the sync character, it is extremely important
that IO.INL be issued before any transfers occur. This ensures that
appropriate parameters are initialized and that the interface is
properly conditioned. Note that IO.INL provides the only means of
setting device characteristics, such as sync character. For this
reason, IO.INL should always be used immediately prior to the first
transfer over a newly-activated link.
9.6

PROGRAMMING EXAMPLE

The following example illustrates the initialization,
device
parameters, and transmission of a block of
message-oriented communication device.

TXAST:

setting of
data on a

.MCALL

ALUN$S,QIO$S

ALUN$S
QIO$S
QIO$S
QIO$S

#1,#"XP,#O
USE LUNl FOR DPll
#IO.HDX!IO.SYN,#l,,,,,<#226>
SET CEVICE PARAMETERS
#IO.INL,#1
PUT DEVICE ON LINE
#IO.WLB,#1,,,#TXSTS,#TXAST,<#TXBUF,#lOO>; SEND A BLOCK

CMPB

#IS.SUC&377,@(SP)+

BEQ

10$

9-10

WAS DA'l'A CLOCKED OUT
SUCCESSFULLY?
IF SO, SET UP FOR NEXT
BLOCK

CHAPTER 10
ANALOG-TO-DIGITAL CONVERTER DRIVERS

10.l

INTRODUCTION

The AFCll and ADOl-D analog-to-digital (A/D) converters are used for
the acquisition of industrial and laboratory analog data. Although
each has its own driver, programming for both is quite similar and
both are multichannel, programmable gain devices. The ADOl-D should
not be confused with the ADUOl, a UDC module, which is described in
Chapter 11. Table 10-1 compares the AFCll and the ADOl-D briefly, and
subsequent sections describe these devices in greater detail.
Table 10-1
Standard Analog-to-Digital Converters
ADOl-D

AFCll
Maximum sampling rate (points
per second)

200 (20 per single)
channel

Approximately
10,000

Number of bits

13 or 14

10 or 11

Maximum number of analog channels
that can be multiplexed

10.1.l

AFCll Analog-to-Digital Converter

The AFCll is a differential analog input subsystem for industrial
data-acquisition and control systems. It multiplexes signals, selects
gain, and performs a 13- or 14-bit analog-to-digital conversion under
program control.
With the use of appropriate signal-conditioning
modules, the system can intermix and accept low-level, high-level, and
current inputs, with a high degree of noise immunity.
10.1.2

ADOl-D Analog-to-Digital Converter

The ADOl-D is an extremely fast analog data-acquisition system.
It
multiplexes signals, selects gain, and performs a 10- or 11-bit
analog-to-digital conversion under program control.
The ADOl-D is
normally unipolar, but an optional sign-bit facilitates bipolar
operation.

10-1

ANALOG-TO-DIGITAL CONVERTER DRIVERS
10.2

GET LUN INFORMATION MACRO

If a GET LUN INFORMATION system directive is issued for a LUN
associated with an analog-to-digital converter, word 2 (the first
characteristics word) contains all zeros, words 3 and 4 are undefined,
and word 5 is not significant, since there is no concept of a default
buffer size for analog-to-digital converters.

10.3

QIO MACRO

This section summarizes standard and device-specific QIO functions for
analog-to-digital converter drivers.

10.3.l

Standard QIO Function

The standard function that is valid for
is shown in Table 10-2.

analog-to-digital

converters

Table 10-2
Standard QIO Function for the A/D Converters
Format

Function

QI0$C IO.KIL, .•.

Cancel I/O requests

Since all requests are processed within a small amount of time, no
in-~rogress
request is ever cancelled. This function simply cancels
all queued requests.

10.3.2

Device-Specific QIO Function

The device-specific function of the QIO macro that
analog-to-digital converters is shown in Table 10-3.

valid

for

Table 10-3
Device-Specific QIO Function for the A/D Converters
Format

Function

QIO$C IO.RBC, ... ,<stadd,size,stcnta>
where:

Initiate multiple A/D
conversions

stadd

is the starting address of the data buffer (must be
a word boundary) .

size

is the control buffer size in bytes (must be even and
greater than zero);
the data buffer is the same size.

stcnta

is the starting address of the control buffer (must be
on a word boundary); each control buffer word must be
constructed as shown in Table 10-4.

10-2

ANALOG-TO-DIGITAL CONVERTER DRIVERS
Table 10-4
A/D Conversion Control Word
Bits

Meaning

0-11

Channel number

Range:

12-15

Gain value for this
sample, expressed as
a bit pattern as
follows

Gain:

15
0
0

0
0
0
0

0
0

10.4

ADOl-D

AFCll

0-1023

Range:
Gain:

0
0
0
1
1
1
1

0
1
1

1
0
1
0
1
0
1

2
illegal
illegal
10
20
illegal
illegal
50
100
illegal
illegal
200
1000
illegal
illegal

2
4

0
1
1

0-63

1
1

l
l

0
0

1
1

1
u
"

1
1
1
1

0
0
1
1

0
1
0
1

illegal
illegal
illegal
illegal
illegal
illegal
illegal
illegal
illegal
illegal
illegal
illegal

FORTRAN INTERFACE

A coiiection of FORTRAN-callable subroutines provide FORTRAN

programs
access to the AFCll and the ADOl-D. These are described in this
section. All are reentrant and may be placed in a resident library.

10.4.1

Synchronous and Asynchronous Process Control I/O

The ISA standard provides for synchronous and asynchronous I/O.
Synchronous I/O is indicated by appending a "W" to the name of the
subroutine (e.g., AISQ/AISQW). The synchronous call suspends task
execution until the I/O operation is complete. If the asynchronous
form is used, execution continues and the calling program must
periodically test the status word for completion.

10.4.2

The isb Status Array

The isb (I/0 status block) parameter is a two-word integer array that
contains the status of the FORTRAN call, in accordance with ISA
convention. This array serves two purposes:
1.

It is the 2-word I/O status block to which the driver returns
a status code on completion of an I/O operation.

10-3

ANALOG-TO-DIGITAL CONVERTER DRIVERS
2.

The first word of isb receives a status code from the FORTRAN
interface in ISA-compatible format, with the exception of the
I/O pending condition, which is indicated by a status of
zero. The ISA standard code for this condition is +2.

The meaning of the contents of isb varies, depending on the FORTRAN
call that has been executed, but Table 10-5 lists certain general
principles that apply.
The section describing each
subroutine
provides further details.
Table 10-5
Contents of First Word of isb
Contents

Meaning

isb(l) = 0

Operation pending;

isb(l) = 1

Successful completion

isb(l) = 3

< isb(l) -< 300

isb(l) > 300

I/0 in progress

Interface subroutine unable to generate
directive or number of samples is
zero

oro

error I

QIO directive rejected and actual
code = -(isb(l) - 3)

Driver rejected request and actual error
code = -(isb(l) - 300)

Unless otherwise specified, the value of isb(2) is the value
by the driver to the second word of the I/O status block.

returned

FORTRAN interface subroutines depend on asynchronous system traps to
set their status.
Thus, if the trap mechanism is disabled, proper
status cannot be set.

10.4.3

FORTRAN Subroutine Summary

Table 10-6 lists the FORTRAN interface subroutines supported
AFCll and ADOl-D under RSX-llM.

for

the

Table 10-6
FORTRAN Interface Subroutines for the AFCll and ADOl-D
Subroutine

Function

AIRD/AIRDW

Perform input of analog data
sequence

AISQ/AISQ\Ai

Read a series of sequential analog input
channels

ASADLN

Assign a LUN to the ADOl-D

ASAFLN

Assign a LUN to the AFCll

10-4

random

ANALOG-TO-DIGITAL CONVERTER DRIVERS
The following subsections briefly describe the function and format of
each FORTRAN subroutine call. Note the use of ASADLN and ASAFLN to
assign a default logical unit number.

10.4.4

AIRD/AIRDW:
Sequence

Performing

Input

Analog

Data

The ISA standard AIRD/AIRDW FORTRAN subroutines input analog
random sequence. These calls are issued as follows:
CALL

where:

{AIRD }
AIRDW

Random
data

( inm, icont, idata, [ isb], [lun])

inm

specifies the number of analog input channels.

icont

is an integer array containing terminal connection
data - channel number
(right-justified in bits 0-11)
and gain (bits 12-15), as shown in Table 10-4.

idata

is an integer array to receive the converted values.

isb

is a two-word integer array
status is returned

lun

is the logical unit number.

which

the

subroutine

The isb array has the standard meaning defined in section 10.4.2.
If
inm = 0, then isb(l) = 3. The contents of idata are undefined if an
error occurs.

10.4.5

AISQ/AISQW:

Reading Sequential Analog Input Channels

The ISA standard AISQ/AISQW FORTRAN subroutines read a series of
sequential analog input channels. These calls are issued as follows:
CALL

where:

{ AISQ }
AISQW

(inm,icont,idata, [isb], [lun])

inm

specifies the number of analog input channels.

icont

is an integer array containing terminal connection
data - channel number
(right-justified in bits 0-11)
and gain (bits 12-15), as shown in Table 10-4.

idata

is an integer array to receive the converted values.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

10-5

array

which

the

subroutine

ANALOG-TO-DIGITAL CONVERTER DRIVERS
For sequential analog input, channel number is computed in steps of
one, beginning with the value specified in the first element of icont.
The channel number field is ignored in all other elements of the
array.
The gain used for each conversion is taken from the respective element
in icont. Thus, even though the channel number is ignored in all but
the first element of icont, the gain must be specified for each
conversion to be performed.
The isb array has the standard meaning defined in section 10.4.2.
If
inm = 0, then isb(l) = 3. The contents of idata are undefined if an
error occurs.

10.4.6

ASADLN:

Assigning a LUN to the ADOl-D

The ASADLN FORTRAN subroutine assigns the specified LUN to the ADOl-D
and defines it as the default logical unit number to be used whenever
a LUN specification is omitted from an AIRD(W)/AISQ(W) subroutine
call. It is issued as follows:
CALL ASADLN (lun, [ isw], [ iun])
where:

lun

is the logical unit number to be assigned to the ADOl-D
and defined as the default unit.

isw

is an integer variable to which the result
ASSIGN LUN system directive is returned.

iun

is the unit number to be assigned.
value of 0 is assumed.

unspecified,

the
a

Only the LUN specified in the last call to ASADLN or ASAFLN is defined
as the default unit.

10.4.7

ASAFLN:

Assigning a LUN to the AFCll

The ASAFLN FORTRAN subroutine assigns the specified LUN to the AFCll
and defines it as the default logical unit number to be used whenever
a LUN specification is omitted from an AIRD(W)/AISQ(W)
subroutine
call. It is issued as follows:
CALL ASAFLN ( lun, [ isw] , [ iun])
where:

lun

is the logical unit number to be assigned to AFCll
defined as the default unit.

and

isw

is an integer variable to which the status
ASSIGN LUN system directive is returned.

the

iun

is the unit number to be assigned.
value of 0 is assumed.

from

unspecified,

Only the LUN specified in the last call to ASAFLN or ASADLN is defined
as the default unit.

10-6

ANALOG-TO-DIGITAL CONVERTER DRIVERS
10.5

STATUS RETURNS

The error and status conditions listed in Table 10-7 are returned
-'----L-·the analog~to~digital converter drivers described in LllLO::. \,;lldJJLt::!L
L..1-!-

Table 10-7
A/D Converter Status Returns
Code

Reason

IS. sue

IS. PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed. The I/O status block is filled with
zeros.

IE .ABO

Operation aborted
The specified I/O operation was
while still in the I/O queue.

IE. BAD

cancelled

via

IO.KIL

Bad parameter
An illegal specification was supplied for one or more
of the device-dependent QIO parameters (words 6-11).
For the analog-to-digital
converters,
this
code
indicates that a bad channel number or gain code was
specified in the control buffer.

IE. BYT

~yte-a~igned buffer specified
Hyte a11gnment was specified for a data or control
buffer,
but
only
word
alignment is legal for
analog-to-digital converters. Alternately, the length
of the data and control buffer is not an even number of
bytes.

IE. DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.
For the AFCll, this code is returned if an interrupt
timeout occurred or the power failed. In the case of
the ADOl-D, which is not operated in interrupt mode,
this code indicates a software timeout occurred (i.e.
a conversion did not complete within 30 microseconds).

IE. IFC

Illegal function
A function code was specified in an I/O request that is
illegal for analog-to-digital converters.

10-7

ANALOG-TO-DIGITAL CONVERTER DRIVERS
Table 10-7 (Cont.)
A/D Converter Status Returns
Code

Reason

IE. OFL

IE. SPC

Illegal address space
The data or control buffer specified for a conversion
request was partially or totally outside the address
space of the issuing task. Alternately, a byte count
of zero was specified.

FORTRAN interface
section 10.5.1.

10.5.1

values

for

these

subroutines

are

presented

FORTRAN Interface Values

The values listed in Table 10-8 are
calls.

returned

Table 10-8
FORTRAN Interface Values
Status Return

rs.sue
IS. PND
IE .ABO
IE. ADP
IE. BAD
IE. BYT
IE. DAO
IE. DNR
IE. IEF
IE. IFC
IE. ILU
IE. NOD
IE. ONP
IE. PRI
IE. RSU
IE. SDP
IE. SPC
IE. ULN
IE. UPN

FORTRAN Value
+01
+00
+315
+101
+301
+319
+313
+303
+100
+302
+99
+323
+305
+316
+317
+102
+306
+08
+04

10-8

FORTRAN

subroutine

ANALOG-TO-DIGITAL CONVERTER DRIVERS
10.6

FUNCTIONAL CAPABILITIES

The AFCll and ADOl-D operate only in multi-sample mode, because the
user can simulate single-sample mode by simply specifying one sample.
Multi-sample mode permits many channels to be sampled at approximately
the same time without requiring the user to queue multiple I/O
requests.
The maximum number of channels in the configuration is specified at
system-generation time. This value is stored in the respective AFCll
and ADOl-D unit control blocks.

10.6.l

Control and Data Buffers

The user must define two buffers of equal size, the control buffer and
the data buffer.
The former contains the control words needed to
perform one A/D conversion per channel specified. Each control word
indicates the channel to be sampled and the gain to be applied (see
Table 10-4).
The data buffer receives the results of the conversions. Each result
is placed in the data buffer location that corresponds to the control
word that specified it.

10.7

PROGRAMMING HINTS

This
section
contains
information
on
important
programming
considerations relevant to users of the analog-to-digital converter
drivers described in this chapter.

10.7.l

Use of A/D Gain Ranges

Note that the A/D gain ranges overlap. The key to successful use of
the A/D converters is to change to a higher gain whenever a full-scale
reading is imminent and to change to a lower gain whenever the last
A/D value recorded was less than half of full scale. This method
maintains maximum resolution while avoiding saturation.

10.7.2

Identical Channel Numbers on the AFCll

When requesting sampling of more than one channel, the user should not
specify multiple sampling of a single channel without 10 or more
intervening samples on other channels. This ensures 50 milliseconds
between samples on a single channel. If sampling occurs more often
than this on a single channel, partial results are returned (see
10.7.3 below).

10.7.3

AFCll Sampling Rate

Although the AFCll can sample a maximum of 200 points per second, a
single channel can only be sampled at 20 points per second. Because

10-9

ANALOG-TO-DIGITAL CONVERTER DRIVERS
the channel capacitor needs 50 milliseconds to recharge after each
conversion, more frequent sampling may result in partial readings. If
this occurs, the user will receive no indication that information is
being lost.
To ensure that information is not lost on any one
channel, the user should sample approximately ten other channels
before returning to the first one.

10.7.4

Restricting the Number of ADOl-D Conversions

The ADOl-D is an extremely fast device, providing a 25-microsecond
conversion rate, and is driven programmably to minimize system
overhead. However, an excessive number of conversions in a single
request essentially locks out the rest of the system because the
driver does not return control to the system until it has finished all
the specified conversions. No other task can run, although interrupts
can still occur and are processed.

10-10

CHAPTER 11
UNIVERSAL DIGITAL CONTROLLER DRIVER

11.1

INTRODUCTION

The UDCll is a digital input/output system for industrial and process
control applications.
It interrogates and/or drives up to 252
directly addressable digital sense and/or control modules. The UDCll
operates under program control as a high-level digital multiplexer,
interrogating digital inputs and driving digital outputs.
The UDC driver will support either the UDCll or ICSll subsystem.
ICSll (Industrial Control Subsystem) operates as an input/output
device that is functionally similar to the UDCll. A maximum of 16 I/O
modules can be placed in one ICSll subsystem. up to 12 ICSlls can be
interfaced to one computer system.
The ICSll subsystem is also
supported by the ICS/ICR-11 driver described in chapter 14. The
reader should consult that chapter for a comparison of driver
features.
While performing analog-to-digital conversions, the UDCll driver can
handle other functions, such as contact or timer interrupts or
latching output. These functions are performed immediately, without
requiring any in progress analog-to-digital conversions to first be
completed.
Unlike other RSX-llM I/O device drivers, the UDCll driver is neither a
multicontroller nor a multiunit driver.

11.1.1 Creating the UDCll Driver
Each installation must assemble the driver source module with a prefix
file that defines the particular hardware configuration. The prefix
file is created during system generation according to the user's
response to questions relating to the UDCll. This file is named
RSXMC.MAC
and
includes
symbolic
definitions
of
the
UDCll
configuration.
These definitions encode the relative module number
and the number of modules for each generic type specified in the
system generation dialog. The encoding has the following format:
8

number of modules

!starting module numberl

11-1

UNIVERSAL DIGITAL CONTROLLER DRIVER
One or more of the following symbols is generated:
Symbol

Module Type

U$$ADM
U$$AOM
U$$CIM
U$$CSM
U$$LTM
U$$SSM
U$$TIM

Analog input
Analog output
Contact interrupt
Contact sense input
Latching digital output
Single-shot digital output
Timer (I/O counter)

Note that all modules of a given type must be
sequential slots.

11.1.2

installed

together

Accessing UDCll Modules

RSX-llM provides two methods of accessing the UDCll:
1.

A QIO macro call issued to the driver

Restricted direct access by any task to
dedicated to the UDCll

I/O

page

registers

The first method, access through the driver, is required to service
interrupting modules and to set and record the state of latching
digital output modules.
The second method, direct access, is a high-speed, low-overhead way to
service noninterrupting modules.
The following functions may be
performed in this manner:
. Analog output
. Contact sense input
. Single-shot digital output
. Read a contact interrupt module
. Read a timer module

11.1.2.1
modules:

Driver Services - The driver services the following types of

Contact interrupt

Timer (I/O counter)

Analog input

Latching digital output

Contact and timer interrupts need not be serviced by a single task.
One task may be connected to contact interrupts, and another to timer
interrupts. A nonprivileged task can connect to either or both of

11-2

UNIVERSAL DIGITAL CONTROLLER DRIVER
these classes by providing a circular buffer to receive interrupt
information and an event flag to allow triggering of the task whenever
a buffer entry is made.

11.1.2.2 Direct Access - A global common block within
the I/O page
provides restricted direct access to the UDCll device registers. In a
mapped system, the length of the block is set to prevent access to
other device registers. In an unmapped system, the use of the common
block is optional. Section 11.4 explains direct access more fully.

11.2

GET LUN INFORMATION MACRO

If a GET LUN INFORMATION system a1rect1ve is issued for a LUN
associated with the UDCll, word 2 (the first characteristics word)
contains all zeros, words 3 and 4 are undefined, and word 5 is not
significant, since there is no concept of a default buffer size for
universal digital controllers.

11.3

QIO MACRO

This section summarizes standard and device-specific QIO functions for
the UDCll driver.
In issuing them, note the numbering conventions
described in 11.7.2.

11.3.l Standard QIO Function

The standard function that is valid for the UDCll is

shown

Table

11-1.

Table 11-1
Standard QIO Function for the UDCll
Format

Function

QIO$C IO.KIL, •••

cancel I/O requests

IO.KIL cancels all queued requests and disconnects all interrupt
connections, but does not stop any I/O that is currently in progress.

11.3.2 Device-Specific QIO Functions

Table 11-2 summarizes device-specific QIO functions that are supported
for the UDCll.

11-3

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-2
Device-Specific QIO Functions for the UDCll
Format

Function

QIO$C IO.CCI, ... ,<stadd,sizb,tevf>

Connect a buffer to contact
interrupts

QIO$C IO.CTI, •.. ,<stadd,sizb,tevf,arv>

Connect a buffer to timer
interrupts

QIO$C IO.DCI,...

Disconnect a buffer from
contact interrupts

QIO$C IO.DTI,...

Disconnect a buffer from timer
interrupts

QIO$C IO.ITI, ... ,<mn,ic>

Initialize a timer

QIO$C IO.MLO, ... ,<opn,pp,dp>

Open or close latching digital
output points

QIO$C IO.RBC, ... ,<stadd,size,stcnta>

Initiate multiple A/D
conversions

where:

stadd

is the starting address of the data buffer (must be
a word boundary) •

sizb

is the data buffer size in bytes (must be even and
large enough to include a 2-word buffer header plus one
data entry; the buffer may cross a 4K boundary).

tevf

is the trigger event flag number (in

range

through

64) •

arv

is the starting address of the table
values (must be on a word boundary).

initial/reset

is the module number.

is the initial count.

opn

is the first latching digital output point number,
which must be on a module boundary (evenly divisible by
16) .

is the 16-bit mask.

is the data pattern.

size

is the control buffer size in bytes (must be even and
greater than zero); the data buffer is the same size.

stcnta is the starting address of the control buffer (must be
on a word boundary); each control buffer word must be
constructed as shown in Table 11-3.
The following sections describe the functions listed in Table 11-2.

11-4

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-3
A/D Conversion Control Word
Bits

Meaning

ADUOl

0-11

Channel number

Range:

12-15

Gain value for this
sample, expressed as
a bit pattern as
follows

Gain:

0
0
0
0
0
0
0
0
1
1
1

0
0
0
0
1
1
1
1
0
0
0

0
0
1
1
0
0
1
1
0
0
1

l
l

0
l

l
0

0
1
0
1
0
1
0
1
0
1
0
l

1
1
1

0
1
1

0-4095

1
2
illegal
illegal
10
20
illegal
illegal
50
100
illegal
illegal
200
1000
illegal
illegal

1
0
1

11.3.2.1 Contact Interrupt Digital Input (W733 Modules) - Digital
input and change of state information from contact interrupt modules
is reported in a requester-provided circular buffer.
The buffer
consists of a 2-word header, followed by a data area in the following
format:
1 ' driver index
2 user index
3 en t:ry
4 en t:ry

.
.
.

Whenever a change of state occurs in one or more contact points an
interrupt is generated.
The UDCll driver gains control, determines
whether the change of state is of interest (i.e., a contact closure
and point closing
(PCL)
is set on the module), and then optionally
makes an entry in the data area of the buffer, updates the index words
and sets the trigger event flag of the connected task.
Each entry consists of five words in the following format:

11-5

UNIVERSAL DIGITAL CONTROLLER DRIVER
Word

Contents

Entry existence indicator

Change-of-state (COS) indicator

Module data (current point values)

Module number (interrupting module)

Generic code (interrupting module)

The driver enters data in the location currently indicated by the
driver index. This pointer can be considered as a FORTRAN index into
the buffer, i.e., the first location of the buffer is associated with
the index 1.
The beginning of the data area is the location of the
first entry (index 3).
Entries are made in a circular fashion,
starting at the beginning of the data area, filling in order of
increasing memory address to the end of the data area, and then
wrapping around from the end to the beginning of the data area.
It is expected that the connected task will maintain its own pointer
(the user index)
to the location in the buffer where it is next to
retrieve contact interrupt data. When a task is triggered by the
driver, it should process data in the buffer starting at the location
indicated by its pointer and continuing in a circular fashion until
the two pointers are equal or a zero entry existence indicator is
encountered. Equality of pointers means that the connected task has
retrieved all the contact interrupt information that the driver has
entered into the buffer.
The entry existence indicator is set nonzero when a buffer entry is
made.
When a requester has removed or processed an entry, he must
clear the existence indicator in order to free the buffer entry
position.
If data input occurs in a burst sufficient to overrun the buffer, data
is discarded and a count of data overruns is incremented. The nonzero
entry existence indicator also serves as an overrun indicator.
A
positive value {+l) indicates no overruns between entries; a negative
value is the two's complement of the number of times data have been
discarded between entries.
The module number indicates a module on which a change of state in the
direction of interest has been recognized for one or more discrete
points. The direction of the change may be from 0 to 1 or 1 to 0,
depending on the PCL (point closing) and POP (point opening) module
jumpers. The change of state {COS) indicator specifies which point or
points of the module have changed state.
The bit position of an on-bit in the COS indicator provides the
low-order bits (3-0) of a point number and the module number provides
the high order bits (15-4). The module data indicates the logical
value (polarity) of each point in the module at the time of the
interrupt.
Contact interrupt data can be reported to only one task.
The
functions IO.CCI and IO.DCI in Table 11-2 are provided to enable a
task to connect and disconnect from contact interrupts.
If the
connection is successful, the second word of the I/O status block
contains the number of words passed per interrupt in the low-order
byte and the initi~l FOR~RAN index to the beginning of the data area
in the high-order byte.

11-6

UNIVERSAL DIGITAL CONTROLLER DRIVER
NOTE
The size of the data area
multiple of the entry size.

must

11.3.2.2 Timer (W734 I/O Counter Modules) - A timer
(I/O counter)
module is a clock that is initialized (loaded), counts up or down, and
then causes an interrupt. The UDCll driver treats such modules in a
way similar to that in which it handles contact interrupts. The
requester provides a circular buffer similar to that for contact
interrupts.
Each entry consists of four words in the following
format:
Word

Contents

Entry existence indicator

Module data (current value)

Module number (interrupting module)

Generic code (interrupting module)

The IO.CTI function in Table 11-2 enables a task to connect to timer
interrupts.
The table of initial/reset values is used to initially
load the timers and to reload them on interrupt (overflow). The table
contains one word for each timer module. The contents of the first
word are used to load the first module, and so forth.
If a timer has
a nonzero value when it interrupts, it is not reloaded, so that
self-clocking modules and modules that interrupt on half count can
continue counting from the initial value.
The IO.DTI function in Table 11-2 disconnects a task from timer
interrupts, and the IO.ITI function provides the capability of
initializing a single timer. Requests to initialize a counter are
valid only if the issuing task has connected a buffer for receiving
counter interrupts.
NOTE
The size of the data area
multiple of the entry size.

must

11.3.2.3 Latching
Digital
Output
(M685,
M803,
and
M805
Modules) - Each module has 16 latching digital output points. The
IO.MLO function in Table 11-2 opens or closes a set of up to 16
points.
Bit n of the mask and data pattern corresponds to the point
opn + n.
If a bit in the mask is set, the corresponding point is
opened or closed, depending on whether the corresponding bit in the
data pattern is clear or set.
If a bit in the mask is clear,
the
corresponding point remains unaltered.

11.3.2.4 Analog-to-Digital Converter
(ADUOl Module) - Each
ADUOl
module has eight analog input channels. The IO.RSC function in Table
11-2 initiates A/D conversions on multiple ADUOl input channels.

11-7

UNIVERSAL DIGITAL CONTROLLER DRIVER
Restrictions on maximum sampling rates are the same as defined for the
AFCll in Chapter 10.

11.3.2.5 ICSll Analog-to-Digital Converter
(IAD-IA module) - Each
IAD-IA module has eight analog input channels. The channel capacity
may be expanded to 120 by the addition of IMX-IA multiplexers.
Each
multiplexer adds 16 input channels to the converter. Restrictions on
maximum sampling rates are the same as defined for the AFCll in
Chapter 10. The IAD-IA module preempts eight module slots regardless
of the number of IMX-IA multiplexers installed.

For addressing purposes, each converter occupies a block of 120
channels.
Thus, A/D converter 0 is addressed by referencing channels
0 through 119; A/D converter 1 is addressed by referencing channels
120
through 239 etc.
When fewer than seven multiplexers are
installed, not all addresses within the block are valid.

11.4

DIRECT ACCESS

Section 11.1.2 describes UDCll functions that may be performed by
referencing a module through its physical address in the I/O page.
Under RSX-llM such access is accomplished by one of the following
methods:

A privileged task or any task running in an unmapped system
has unrestricted access to the I/O pa1e and may therefore
access each module by absolute address.

Using the Task Builder, a task may link to a global common
area whose physical address limits span a set of locations in
the I/O page. This method applies to either a mapped or
unmapped system.

The latter method allows a task to be transported to any other system
simply
by relinking.
Further, in a mapped system the memory
management hardware will abort all references to device registers
outside the physical address limits of the common block.
The operations required to implement each method may be summarized
follows:
1.

Unrestricted access to the I/O page
a.

An object module is created which defines the UDCll
configuration through a list of absolute global addresses
and addressing limits for each module type.

The object module is included in the system library file.

A task is created containing the appropriate global
references.
Such references are resolved when the task
builder automatically searches the system library file.

Steps a and b are executed once, during system generation (see RSX-llM
System Generation Manual).
Step c is performed each time a task is
created that references the UDCll.

11-8

UNIVERSAL DIGITAL CONTROLLER DRIVER
2.

Access to the I/O page through a Global Common Block:
a.

An object module is created which defines the UDCll
configuration
through a list of relocatable global
addresses and addressing limits for each module type.

The object module is linked, using the Task Builder,
create an image of the Global Common block on disk.

The SET command is used to define
resides on the I/O page.

The INSTALL MCR command is used to make the Global Common
Block resident in memory.

A task is created containing the appropriate global
references.
Such references are resolved by directing
the Task Builder to link the Task to the common block.

common

block

to
that

Steps a through d are executed once, during system generation. Step e
is performed each time a task is created that references the UDCll
common block. The following paragraphs describe each step in detail.

11.4.1

Defining the UDCll Configuration

The source module UDCOM.MAC*, when assembled with the proper prefix
file, provides global definitions for the following parameters:
The starting address of each module type.
The highest point number within a given module type.
The highest module number within a given module type.
The last two parameters are absolute quantities that may be used to
prevent a task from referencing a module that is nonexistent or out of
limits.

By means of conditional assemo1y tne list of addresses may be created
as absolute symbols defining locations in the I/O page or as symbols
within a relocatable program section to be used when building and
linking to the UDCll Global Common area.

11.4.1.1 _A_s_s_e_m_b_l_y_P_r_o_c_e_d_u_r_e_f_o_r__U_D_C_O_M_._M_iA_C_ - UDCOM. MAC is assembled
with the RSX-llM configuration parameters contained in the file
RSXMC.MAC.

To create relocatable module addresses either the parameter U$$DCM or
M$$MGE must be defined.
M$$MGE will be included in RSXMC.MAC if
memory management was specified when the system was qenerated.
If
not,
the user should edit the file to include the following
definition:
U$$DCM=O

* This module

resides on the RK05 cartridge of the RSX-llM RK
For RP distribution
distribution kit labeled EXECUTIVE SOURCE.
kits, it resides on the RP image. The file is located under UIC
[11,10].
11-9

UNIVERSAL DIGITAL CONTROLLER DRIVER
The file may then be assembled using the MCR command:
>MAC UDCOM,UDLST=[ll,lO]RSXMC,UDCOM
This command invokes the MACR0-11 assembler which processes the input
files RSXMC.MAC and UDCOM.MAC to create UDCOM.OBJ and UDLST.LST.
To create absolute module addresses, both of the above parameters must
be undefined. Edit RSXMC.MAC, if necessary, to remove definitions and
then invoke the MACR0-11 assembler with the following MCR command:
>MAC UDCDF,UDLST=[ll,lO]RSXMC,UDCOM
In this sequence the files UDCDF.OBJ and UDLST.LST are created from
the specified source modules. UDCDF.OBJ contains the module addresses
in absolute form.

11.4.1.2 Symbols Defined by UDCOM.MAC - This
symbolic definitions created by UDCOM.MAC.
The following symbols define the absolute or
the first module of a given type:

section

relocatable

Symbol

Module Type

$.ADM
$.AOM
$.CIM
$.CSM
$.LTM
$.SSM
$.TIM

Analog input
Analog output
Contact interrupt
Contact sense input
Latching digital output
Single-shot digital output
Timer (I/O counter)

The addresses in relocatable form are defined
named UDCOM having the attributes:
REL OVR D
GBL -

lists

the

address

program

section

relocatable
overlaid
data
global scope

Note that these attributes correspond to those
common block within a Fortran program.
In either the absolute or relocatable case,
referenced by the corresponding symbolic
module index.

attached

named

individual modules are
address plus a relative

The following symbols define the highest digital point within a module
type:
Symbol

Module Type

P$.CIM
P$.CSM
P$.LTM
P$.SSM

Contact interrupt
Contact sense input
Latching digital output
Single-shot digital output

11-10

UNIVERSAL DIGITAL CONTROLLER DRIVER
The highest point number is defined relative to the first point on the
first module of a specific type.
For example if two contact interrupt modules are installed: the symbol
P$.CIM will have an octal value of 37.
The following symbols define the highest module number within a
module type.
Symbol

Module Type

M$ .ADM
M$.AOM
M$.CIM
M$.CSM
M$.LTM
M$.SSM
M$.TIM

Analog input
Analog output
Contact interrupt
Contact sense input
Latching digital output
Single-shot digital
Timer (I/O counter)

given

The highest module number is defined relative to the first module of a
given type.
Thus, based on the previous example, M$.CIM will have a
value of l.

11.4.2

Including UDCll Symbolic
Module Library

Definitions

the

System

Object

As described in 11.4, a task having unrestricted access to the I/O
page may reference a UDCll module by absolute address. The object
module UDCDF contains symbolic definitions of
absolute
module
addresses and may be included in the System Object ~odule Library:
SY: [l,l]SYSLIB.OLB
The Task Builder automatically searches this file to resolve any
undefined globals remaining after all input files have been processed.
The following example illustrates the procedure for including the file
UDCDF.OBJ in the library.
>SET /UIC=[l,l]
>LBR SYSLIB/IN=[200,200]UDCCF
The SET MCR command is issued to establish the current UIC as [1,1).
Next, the RSXllM Librarian is invoked and instructed, through the use
of the /IN switch to include the object module UDCDF.OBJ in the file
SYS LIB. OLB.

11.4.3

Referencing the UDCll through a Global Common Block

The following sections define the procedure for creating a Global
Common block in the I/O PAGE, making the block resident in memory, and
creating a task which references UDCll modules within the block.
Examples are given for both mapped and unmapped systems.

11-11

UNIVERSAL DIGITAL CONTROLLER DRIVER
11.4.3.1 Creating a Global Common Block - The following sequence
illustrates the use of the object file UDCOM.OBJ to create a disk
image of the global common area in a mapped system.
>SET /UIC=[l,l]
>TKB
TKB>UDCOM/MM,LP:,SY:UDCOM/PI/-HD=[200,200]UDCOM
TKB>/
ENTER OPTIONS:
TKB>PAR=UDCOM:O:lOOO
TKB>STACK=O
TKB>/
In the above example, a current UIC of [1,1] is established and the
Task Builder is initiated. The initial input line to the Task Builder
specifies the following files:
A core image output file to be named UDCOM.TSK
A memory map output to the line printer
A symbol table file to be named UDCOM.STB
All files reside on SY: under UIC [1,1].
UDCOM.OBJ containing the UDCll address
values, constitutes the input.
The switches specified for the output
information to the Task Builder:

The single input file,
definitions as relocatable

files

convey

the

following

/MM

indicates that the core image of the common block
reside on a system with Memory Management.

will

/PI

indicates that the core image is position independent;
that is the virtual address of the common block may
appear on any 4K boundary within a task's address
space.

/-HD

indicates that the core image will not contain a
header.
A header is only required for a core image
file that is to be installed and executed as a task.

A single line of option input must be entered to eliminate the default
memory allocation for the stack area.
The following sequence illustrates the corresponding procedure for
unmapped system:

>SET /UIC=[l,l]
>TKB
TKB>UDCOM/-MM,LP:,SY:UDCOM/PI/-HD=[200,200]UDCOM
TKB>/
ENTER OPTIONS:
TKB>STACK=O
TKB>PAR=UDCOM:l71000:1000
TKB>/
Again the task builder is requested to produce a core image and symbol
table file under the UIC [1,1] and a map file on the line printer from
the input file UDCOM.OBJ.
The output file switches convey the
following information:

11-12

UNIVERSAL DIGITAL CONTROLLER DRIVER
/-MM

indicates that the core image of the common block
reside on an unmapped system.

will

/PI

Indicates that the core image is position independent.
In an unmapped system the core image is fixed in the
same address space for all tasks; however, the global
symbols defined in the symbol table file retain the
relocatable attribute.

/-HD

indicates that a core image without a header is
created.

The PAR option specifies the base and length of the common area to
coincide with the standard UDCll addresses in the I/O page. All
references to the common block by tasks will be resolved within this
region.

11.4.3.2 Making the Common Block Resident - The following SET command
creates a UDCll common block residing in the I/O page for a mapped
system:
>SET /MAIN=UDCOM:7710:10:DEV
The corresponding command in an unmapped system is:
>SET /MAIN=UDCOM:l710:10:DEV
The preceding sequence specifies the allocation of a common block in
the I/O page whose physical address limits correspond to the UDCll
standard locations. Note that the address bounds and length are
defined in units of 32 words.
The command
>INS [l,l]UDCOM
declares the common block resident in the system.

11.4.3.3 Linking a Task to the UDCll Common Block - A task may access
UDCll modules by linking to the common block as follows:
TKB>TASK,LP:=TASK.OBJ
TKB>/
ENTER OPTIONS:
TKB> COMMON=UDCOM:RW
TKB>/
The above sequence is valid for either a mapped or unmapped system.
In both cases the Task Builder will link the task to the common block
by relocating the Global symbol definitions contained in UDCOM.STB.
If
memory
management is present, the Executive will map the
appropriate physical locations into the task's virtual addressing
space when the task is made active.

11-13

UNIVERSAL DIGITAL CONTROLLER DRIVER
11.5

FORTRAN INTERFACE

A collection of FORTRAN-callable subroutines provide FORTRAN programs
access to the UDCll. These are described in this section. All are
reentrant and may be placed in a resident library.
Instead of using the FORTRAN-callable subroutines described in this
section, a FORTRAN program may use the global common feature described
in section 11.4 to reference UDCll modules directly in the I/O page,
as shown in the following example:

c
c
c
c
c

UDCll GLOBAL COMMON
COMMON /UDCOM/ ICSM(lO) ,IAO(lO)
READ CONTACT SENSE MODULE l DIRECTLY
ICS=ICSM(l)

Note that the position of each module type must correspond to
sequence in which storage is allocated in the common statements.

the

11.5.l Synchronous and Asynchronous Process Control I/O
The ISA standard provides for synchronous and asynchronous process
I/O.
Synchronous I/O is indicated by appending a "W" to the name of
the subroutine (e.g., AO/AOW). But due to the fact that nearly all
UDCll I/O operations are performed immediately, in most cases the "W"
form of the call is retained only for compatibility and has no meaning
under RSX-llM.
In the case of A/D input, however, the "W" from is
significant: the synchronous call suspends task execution until input
is complete.
If the asynchronous form is used, execution continues
and the calling program must periodically test the status word for
completion.

11.5.2

The isb Status Array

The isb (I/C status block) parameter is a 2-word integer array that
contains the status of the FORTRAN call, in accordance with ISA
(Instrument Society of America) convention.
This array serves two
purposes:
l.

It is the 2-word I/O status block to which the driver
returns an I/O status code on completion of an I/O
operation.

The first word of isb receives a status code from the
FORTRAN interface in ISA-compatible format, with the
exception of the I/0 pending condition,
which
is
indicated by a status of zero. The ISA standard code for
this condition is +2.

The meaning of the contents of isb varies, depending on the FORTRAN
call that has been executed, but Table 11-4 lists certain general
principles that apply. The section describing each subroutine gives
more details.

11-14

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-4
Contents of First Word of isb
Contents

Meaning

isb(l)

Operation pending;

isb(l)

Successful completion

isb(l)

Interface
subroutine
unable
to
generate QIO directive or number of
points requested is zero

I/O in progress

3 < isb(l) < 300

QIO directive rejected and
error code = -{isb(l) - 3)

actual I

isb(l) > 300

Driver rejected request and
error code = -(isb(l) - 300)

actual

In some cases, the values or states of points being read,
latched are returned to isb word 2.

pulsed,

FORTRAN interface subroutines for analog input depend on asynchronous
system traps to set their status. Thus, if the trap mechanism is
disabled, proper status cannot be set.
For direct access calls (indicated in Table 11-5 below), errors are
detected and returned by the FORTRAN interface subroutine itself,
rather than by the driver. Although the use of a two-word status
block is therefore unnecessary, these errors are returned in standard
format to retain compatibility with functions called through QIO
directives and handled by other drivers. Errors of this type that may
be returned are:
isb ( 1) = 3

Number of points requested is
zero

isb{l)

Invalid UDCll module

+321

11.5.3 FORTRAN Subroutine Summary
Table 11-5 lists the FORTRAN interface subroutines supported for
the
UDCll under RSX-llM.
(D)
indicates a direct access call and the
optional logical unit number for such a call may be specified to
retain compatibility with RSX-llD, but this specification is ignored.
Table 11-5
FORTRAN Interface Subroutines for the UDCll
Subroutine

Function

AIRD/AIRDW

Perform input
sequence

AISQ/AISQW

Read a series of
channels

11-15

analog

random

analog

input

data

sequential

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-5 (Cont.)
FORTRAN Interface Subroutines for the UDCll
Subroutine
AO/AOW

Function
Perform analog output on

several

channels

( D)

ASUDLN

Assign a LUN to the UDCll

CTDI

Connect a circular buffer
contact interrupt data

CT'l I

Connect a circular buffer to receive
interrupt data

DFDI

Disconnect a buffer from contact interrupts

receive
timer

I Disconnect a buffer from timer interrupts

DFTI
DI/DIW

Read several 16-point contact sense

fields

( D)

I Latch or unlatch several 16-point fields

DOL/DOLW

I DOM/DO MW

Pulse several 16-point fields (D)

RC I PT

Read the state of
interrupt point (D)

RDCS

Read the contents of a contact interrupt
circular buffer, returning data on only
those points that have changed state.

RDDI

Read the contents of a contact interrupt
circular buffer, one point for each call

I RDTI

Read the contents of a timer interrupt
circular buffer, one entry for each call

RDWD

Read the contents of a contact interrupt
circular
buffer, returning 16 bits of
module
data
and
change-of-state
information.

RSTI

Read a single timer module (D)

SCTI

Set a timer module to an initial value

single

contact

The following subsections briefly describe the function and format of
each FORTRAN subroutine call.
Note the use of ASUDLN to specify a
default logical unit number. Also consider the numbering conventions
described in 11.7.2.
The following FORTRAN functions do not perform
facilitate conversions between BCD and binary.
Convert four BCD digits to a binary number:
IBIN

= KBCD2B(IBCD}

11-16

I/O

directly,

but

UNIVERSAL DIGITAL CONTROLLER DRIVER
Convert a binary number to four BCD digits:
IBCD

11.5.4

= KB2BCD(IBIN)

AIRD/AIRDW:
Sequence

Performing

Input

Analog

Data

The ISA standard AIRD/AIRDW FORTRAN subroutines input analog
random sequence. These calls are issued as follows:
}AIRD }
)AIRDW

CALL
where:

Random
data

(inm,icont,idata, [isb] ,lun)

inm

specifies the number of analog input channels.

icont

is an integer array containing terminal connection
data - channel number
(right-justified in bits 0-11)
and gain (bits 12-15), as shown in Table 11-3.

idata

is an integer array to receive the converted values.

isb

is a two-word integer array
status is returned.

lun

is the logical unit number.

which

the

subroutine

NOTE
lun is a required parameter.

The isb array has the standard meaning defined in section 11.5.2.
If
inm
0, then isb(l) = 3. The contents of idata are undefined if an
error occurs.

11.5.5

AISQ/AISQW:

Reading Sequential Analog Input Channels

The ISA standard AISQ/AISQW FORTRAN subroutines read a series of
sequential analog input channels. These calls are issued as follows:
AISQ }
{ AISQW

CALL
where:

(inm,icont,idata, [isb] ,lun)

inm

specifies the number of analog input channels.

icont

is an integer array containing terminal connection
data - channel number
(right-justified in bits 0-11)
and gain (bits 12-15), as shown in Table 11-3.

idata

is an inteqer array to receive the converted values.

isb

is a two-word integer array
status is returned.

lun

is the logical unit number.

11-17

which

the

subroutine

UNIVERSAL DIGITAL CONTROLLER DRIVER
NOTE
lun is a required parameter.
For sequential analog input, channel number is computed in steps of
one, beginning with the value specified in the first element of icont.
The channel number field is ignored in all other elements of the
array.
The gain used for each conversion is taken from the respective element
in icont. Thus, even though the channel number is ignored in all but
the first element of icont, the qain must be specified for each
conversion to be performed.
The isb array has the standard meaning defined in section 11.5.2.
If
inm
0, then isb(l) = 3. The contents of idata are undefined if an
error occurs.

11.5.6

AO/AOW:

Performing Analog Output

The ISA standard AO/AOW FORTRAN subroutines initiate analoq output
several channels. These calls are issued as follows:
CALL

where:

(inm,icont,idata, [isb], [lun])

inm

specifies the number of analog output channels.

icont

is an inteqer array containinq the channel numbers.

idata

is an inteqer array containing
settinqs, in the ranqe 0-1023.

isb

is a two-word inteqer array
status is returned.

lun

is the loqical unit number (ignored if present).

the
which

output
the

voltaqe

subroutine

The isb array has the standard meaninq defined in section 11.5.2.

11.5.7

ASUDLN:

Assiqninq a LUN to the UDCll

The ASUDLN FORTRAN subroutine assiqns the specified LUN to the
specified unit and defines it as the default logical unit number to be
used whenever a LUN specification is omitted from a UDCll subroutine
call. It is issued as follows:
CALL ASUDLN (lun, [ isw], [ iun])
where:

lun

is the logical unit number to be assigned
specified unit, and defined as the default.

11-18

the

UNIVERSAL DIGITAL CONTROLLER DRIVER

11.5.8

isw

is an integer variable to which the result
ASSIGN LUN system directive is returned.

iun

is an integer defining the UDCll unit
number is specified, 0 is assumed.

CTDI:

number.

of
If

the
no

Connecting to Contact Interrupts

The CTDI FORTRAN subroutine connects a task to contact interrupts and
specifies a circular buffer to receive contact interrupt data. The
length of this buffer can be computed by considering the following:
. Rate at which contact module interrupts occur
. Number of modules that can interrupt simultaneously
. Rate at which the circular buffer is emptied
The UDCll driver generates a five-word entry for each contact
interrupt and the interface subroutine itself requires 10 words of
additional storage. Thus the isz parameter, described below, can be
computed as follows:
isz = (10 + 5 * n)
where n is the number of entries in the buffer and isz is expressed in
words.
The call is issued as follows:
CALL C'I'DI

where:

(ibuf ,isz,iev, [isb], [lun])

ibuf

is an integer
interrupt data.

isz

is the length of the array in
size of 15.

iev

is the trigger event flag number. The specified event
flag is set whenever the driver inserts an entry in the
data buffer.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

array

that

array

receive

words,

with

which

the

contact

minimum

subroutine

The isb array has the standard meaning defined in section 11.5.2.

11.5.9

CTTI:

Connecting to Timer Interrupts

The CTTI FORTRAN subroutine connects a task to timer interrupts and
specifies a circular buffer to receive timer interrupt data. The
length of this buffer can be computed by considering the following:
Rate at which timer module interrupts occur

11-19

UNIVERSAL DIGITAL CONTROLLER DRIVER
Number of modules that can interrupt simultaneously
Rate at which the circular buffer is emptied
The UDCll driver generates a four-word entry for each timer interrupt
and the interface subroutine itself requires 8 words of additional
storage. Thus the isz parameter, described below, can be computed as
follows:
isz

(8 + 4 * n)

where n is the number of entries in the buffer and isz is expressed in
words.
When a timer module interrupt occurs, the driver resets the count to
an initial value, normally that specified in iv. The initial value
for a specific module can be modified by calling the SCTI subroutine
{see section 11. 5 .19) •
The call is issued as follows:
CALL CTTI { ibuf, isz, iev, iv, [ isb], [lun])
where:

ibuf

is an integer array that is to receive timer
data.

isz

is the length of the array in
size of 12.

iev

is a trigger event flag number.
The specified event
flag is set whenever the driver inserts an entry in the
data buffer.

is an integer array which contains the initial timer
module values, with one entry for each timer module,
where entry n corresponds to timer module number n-1.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

array

words,

which

with

the

interrupt
a

minimum

subroutine

The isb array has the standard meaning defined in section 11.5.2.

11.5.10

DFDI:

Disconnecting from Contact Interrupts

The DFDI FORTRAN subroutine disconnects
interrupts. It is issued as follows:

task

from

which

the

contact

CALL DFDI {[isb],[lun])
where:

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

11-20

array

subroutine

UNIVERSAL DIGITAL CONTROLLER DRIVER
The isb array has the standard meaning defined in section 11.5.2.

11.5.11

DFTI:

Disconnecting from Timer Interrupts

The DFTI FORTRAN subroutine disconnects a task from timer
It is issued as follows:

interrupts.

CALL DFTI ( [ isb] '[lun])
where:

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

array

which

the

subroutine

The isb array has the standard meaning defined in section 11.5.2

11.5.12

DI/DIW:

Reading Several Contact Sense Fields

The ISA standard DI/DIW FORTRAN subroutines read several
contact sense fields. These calls are issued as follows:
(DI }

CALL
where:

16-point

(inm,icont,idata,isb,[lun])

f DIW'

inm

specifies the number of fields to be read.

icont

is an integer array containing the initial point number
"-i....of each field 1...V
UC::
L c::au.

idata

is an integer array that is to receive the input
16 bits of contact data for each field read.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number (ignored if present).

---..:J

array

which

the

data,

subroutine

The isb array has the standard meaning defined in section 11.5.2.
11.5.13

DOL/DOLW:

Latching or Unlatching Several Fields

The ISA standard DOL/DOLW FORTRAN subroutines latch or unlatch one
more 16-point fields. These calls are issued as follows:
CALL

~ DOL

(inm,icont,idata,imsk, [isb], [lun])

lDoLw
where:

inm

specifies the
unlatched.

icont

is an integer array containing the initial point number
of each 16-point field.

number

11-21

fields

latched

UNIVERSAL DIGITAL CONTROLLER DRIVER
idata

is an integer array which specifies the points to be
latched or unlatched;
bit n of idata corresponds to
point number icont + n; if the corresponding bit in
imsk is set, the bit is changed; a bit value of 1
indicates latching, and 0 unlatching;
each entry in
the array specifies a string of 16 points.

imsk

is an integer array in which bits are set to indicate
points
whose
states
are to be changed in the
corresponding idata bits;
each entry in the array
specifies a 16-bit mask word.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

array

which

the

subroutine

The isb array has the standard meaning defined in section 11.5.2.

11.5.14

DOM/DOMW:

Pulsing Several Fields

The ISA standard DOM/DOMW FORTRAN subroutines pulse several 16-bit
fields (one-shot digital output points). These calls are issued as
follows:
DOM }
{ DOMW

CALL
where:

(inm,icont,idata, [idx], [isb], [lun])

inm

specifies the number of fields to be pulsed.

icont

is an integer array containing the initial point number
of each 16-point field.

idata

is an integer array which specifies the points to be
pulsed;
bit n of idata corresponds to point number
icont + n.

idx

is a dummy argument retained for compatibility with
existing Instrument Society of America standard FORTRAN
process control calls.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number (ignored if present).

array

which

the

subroutine

The isb array has the standard meaning defined in section 11.5.2.

11.5.15

RCIPT:

Reading a Contact Interrupt Point

The RCIPT FORTRAN subroutine reads the state
interrupt point. It is issued as follows:
CALL RCIPT (ipt,isb,[lun])

11-22

single

contact

UNIVERSAL DIGITAL CONTROLLER DRIVER
where:

ipt

is the number of the point to be read;
points are
numbered sequentially from 0, the first point on the
first contact interrupt module.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number (ignored if present).

array

which

the

subroutine

The isb array has the same basic meaning defined in section
However,
isb word 2 is set to one of the following
representing the state of the point:
Setting

11.5.16

11.5.2.
values,

Meaning

.FALSE.

(0)

Point is open

.TRUE.

(-1)

Point is closed

RDCS: Reading Contact Interrupt Change-of-State Data from a
Circular Buffer

The RDCS FORTRAN subroutine reads contact interrupt data from a
circular buffer that was specified in a CTDI call (see 11.5.8 above).
It does no actual input or output,
but
rather
performs
a
point-by-point scan of an interrupt entry in the buffer, returning the
state of each point that has changed state, as a logical value.
The
trigger event flag which was specified in the CTDI call is cleared
when the "buffer empty" condition is detected.
On the initial call to RDCS, the module number, module data, and
change-of-state word of the next interrupt entry are read from the
circular buffer and stored for subsequent reference.
The subroutine
then searches the entry change-of-state word until a nonzero point is
encountered. The point number is ccmputed and returned to the caller
along with the state of the point. Scanning for po:nts that have
changed state resumes on the next call;
all other points are
bypassed.
The next entry is automatically read when the caller has
received all change-of-state information from the current entry. If a
valid entry is not found, ipt is set negative and ict (if specified)
is either assigned a value of zero or an overrun count maintained by
the UDCll driver.
If ict is zero, no further entries remain. A
nonzero value indicates that the driver received more data than could
be stored in the buffer, and ict represents the number of entries that
were discarded.
The RDCS call is issued as follows:
CALL RDCS (ipt,ival,[ict])
where:

ipt

is a variable to which the digital input
is returned; it may be set as follows:

point

number

ipt < 0 if no valid entry is found
(i.e., no
interrupt data currently in buffer, or overrun
detected).
One of the following
values
is
returned to indicate the condition detected:

11-23

UNIVERSAL DIGITAL CONTROLLER DRIVER
-1

Buffer empty
Overrun detected

-2

ipt => 0 if the value indicated is a point number
that has changed state; the state is returned to
ival.
ival is a variable to which the state of
returned; it may be set as follows:

11.5.17

.FALSE.

(0)

.TRUE.

(-1)

the

point

i f the point is open
i f the point is closed

ict

is an integer variable for receiving the overrun count.
A nonzero positive count indicates that the driver was
unable to store the number of interrupts indicated.

RODI:

Reading Contact Interrupt Data From a Circular Buffer

The RODI FORTRAN subroutine reads contact interrupt data from a
circular buffer that was specified in a CTDI call (see 11.5.8 above}.
It does no actual input or output,
but
rather
performs
a
point-by-point scan of an interrupt entry in the buffer, returning the
state of each point as a logical value. The trigger event flag which
was specified in the CTDI call is also cleared.
On the initial call to RODI the module number and data of the next
interrupt entry are read from the circular buffer and stored for
subsequent reference. The subroutine then sets the current data bit
number n to zero, examines the state of data bit n, and converts bit n
to a point number via the following formula:
ipt = module number

* 16 + n

On each subsequent call, n is incremented by one and then data bit n
is examined in the stored module data. When n reaches 16, it is reset
to zero and an attempt is made to read the next interrupt entry from
the circular buffer.
If a valid entry is not found, ipt is set
negative and ict (if specified} is either assigned a value of zero or
an overrun count maintained by the UDCll driver. If ict is zero, no
further entries remain. A nonzero value indicates that the driver
received more data than could be stored in the buffer, and ict
represents the number of entries that were discarded.
The RODI call is issued as follows:
CALL RODI (ipt,ival,[ict]}
where:

ipt

is a variable to which the digital input
is returned; it may be set as follows:

point

number

(i.e. , no
ipt < 0 if no valid entry is found
interrupt data currently in buffer, or buffer
empty}. One of the following values is returned
to indicate the condition detected:
-l=Buffer empty
-2=0verrun detected

11-24

UNIVERSAL DIGITAL CONTROLLER DRIVER
ipt => 0 if the value indicated is a point number;
the state is returned to ival
iv al

is a variable to which the state of
returned; it may be set as follows:
.FALSE.

( 0)

the

point

if the point is open

.TRUE. (-1) if the point is closed

11.5.18

ict

is a variable to which the overrun count may be
returned;
a nonzero positive count indicates that the
driver was unable to store the number of entries
indicated.

RDTI:

Reading Timer Interrupt Data From a Circular Buffer

The RDTI FORTRAN subroutine reads timer interrupt data from a circular
buffer that was specified in a CTTI call (see 11.5.9 above). It does
no actual input or output, but rather performs a scan of each entry in
the buffer, returning the timer value for each call. The trigger
event flag which was specified in the CTTI call is also cleared.
When a timer module interrupt occurs, the UDCll driver resets the
count to an initial value, usually that specified in the iv array on
the CTTI call. The initial value can be modified for a specific
module by calling the subroutine SCTI (see section 11.5.19).
The RDTI call is issued as follows:
CALL RDTI (imod,itm,[ivrn])
where:

imod

is a variable to which the module number
it may be set as follows:

returned;

• imod < 0 if no valid entry is found
(i.e. , no
interrupt data currently in buffer, or buffer
empty). One of the following values is returned
to indicate the condition detected:
-l=Buf fer empty
-2=0verrun detected
imod > 0 if the
module number;
returned in itm

entry is valid, indicating a
the value of the timer module is

itm

is a variable to which the timer value is returned.

ivrn

is a variable to which the overrun count may be
returned;
a nonzero positive count indicates that the
driver was unable to store the number of values
indicated.

11-25

UNIVERSAL DIGITAL CONTROLLER DRIVER
11.5.19 RDWD: Reading a Full Word of Contact Interrupt Data from the
Circular Buffer
The RDWD FORTRAN subroutine reads a full word of contact interrupt and
change-of-state data from the circular buffer that was specified in a
CTDI call (see section 11.5.8). It does no actual input or output,
but rather performs a scan of each entry, returning the state of a
module and optionally, the change-of-state data for each call.
The
trigger event flag specified in the call to CTDI is cleared.
The call to RDWD is issued as follows:
CALL RDWD (imod,ist,[ivrn] ,[icos])
where:

imod is a variable to which the module number
it may be set as follows:

returned;

imod < 0 if no valid entry is found (i.e., no
interrupt data currently in buffer or overrun
detected).
One of the following
values
is
returned to indicate the condition detected:
-l=Buf fer empty
-2=0verrun detected

11.5.20

ist

is a variable to which the module data is returned.

ivrn

is a variable to which the overrun count may be
returned; a nonzero, positive count indicates that the
driver was unable to store the number of entries
indicated.

icos

is a variable to which the change-of-state data is
returned.
One bit is set for each point that has
changed state in the direction indicated by the "point
open"
(POP) or "point closed"
(PCL) jumpers on the
module.

RSTI:

Reading a Timer Module

The RSTI FORTRAN subroutine reads a single timer module.
as follows:
where:

It is issued

CALL RSTI (imod,isb,[lun])
imod
is the module number of the timer to be read.
isb

is a 2-word integer
status is returned.

array

which

the

subroutine

lun

is the logical unit number (ignored if present).

The isb array has the standard meaning defined in section 11.5.2.

11.5.21

SCTI:

Initializing a Timer Module

The SCTI FORTRAN subroutine sets a timer module to an

11-26

initial

value.

UNIVERSAL DIGITAL CONTROLLER DRIVER
It is issued as follows:
CALL SCTI ( imod, ival, [ isb], [lun])
where:

imod

is the module number of the timer to be set.

iv al

is the initial timer value.

isb

is a 2-word integer
status is returned.

lun

is the logical unit number.

array

which

the

subroutine

The isb array has the standard meaning defined in section 11.5.2.
Calls to initialize a counter are valid only if the issuing task has
connected a buffer for receiving counter interrupts via a call to
CTTI.

11.6

STATUS RETURNS

Table 11-6 lists the error and status conditions that are returned
the UDCll driver described in this chapter:

Table 11-6
UDCll Status Returns
Code
IS. sue

Reason
Sucessful completion
The operation specified in the QIO directive was
completed successfully.
The second word of the I/O
status block can be examined to determine the number of
samples completed or converted.

IS. PND

I/O request pending
The operation specified in the QIO directive has not
yet been executed. The I/O status block is filled with
zeros.

IE .ABO

Operation aborted
The specified I/O operation was
while still in the I/O queue.

IE. BAD

cancelled

via

IO.KIL

Bad parameter
An illegal specification was supplied for one or more
of the device-dependent QIO parameters (words 6-11).
For the UDCll, this code indicates an illegal channel
number or gain code for the ADUOl.

11-27

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-6 (Cont.)
UDCll Status Returns
Code
IE. BYT

Reason
Byte-aligned buffer specified
Byte alignment was specified for a buffer but only word
alignment is legal for the UDCll. Alternately, the
length of a buffer was not an even number of bytes.

IE. CON

Connect error
The task attempted to connect to contact or timer
interrupts, but the interrupt was already connected to
another task. Only one task can be connected to a
timer or contact interrupt. Alternately a task which
was not connected attempted to disconnect from contact
or timer interrupts.

IE. DNR

Device not ready
The physical device unit specified in the QIO directive
was not ready to perform the desired I/O operation.
For the ADUOl, this code is returned if an interrupt
timeout occurred or the power failed.

IE. IEF

Invalid event flag number
The trigger event flag number specified
function was not in the range 1 to 64.

I IE. IFC

connect

Illegal function
A function code was included in an I/O request that is
illegal for the UDCll, or a request to initialize a
counter (IO.ITI) was issued by a task that was not
connected to receive counter interrupts. The function
may also refer to a UDCll feature which was not
specified at system generation.

IE. MOD

Invalid UDCll module
On latching output, the user specified a starting point
number
which
was
not legal (defined at system
generation) or was not evenly divisible by 16.

IE. OFL

IE. FRI

Privilege violation
The task which issued the request was not privileged to
execute
that request.
For the UDCll, this code
indicates that a checkpointable task attempted to
connect to timer or contact interrupts.

11~28

UNIVERSAL DIGITAL CONTROLLER DRIVER
Table 11-6 (Cont.)
UDCll Status Returns
Code
IE. SPC

Reason
Illegal address space
The specified control, data, or interrupt buffer was
partially or totally outside the address space of the
issuing task. Alternately, the interrupt buffer was
too small for a single data entry (6 words for timer
interrupts and 7 words for contact interrupts) or a
byte count of zero was specified.

FORTRAN interface values for these status
section 11.6.1.
11.6.1

returns

presented

FORTRAN Interface Values

The values listed in Table 11-7 are returned in
calls.
Table 11-7
FORTRAN Interface Values

FORTRAN

Status Return

FORTRAN Value

IS. sue
IS. PND
IE .ABO
IE .ADP

+01
+00
+315
+101

IE. BAD

..L.

IE. BYT
IE. DAO

+319
+313
+303
+100
+302
+99
+321
+305
+316
+317
+102
+306
+08
+04

IE. DNR

IE. IEF
IE. IFC
IE. ILU

IE.MOD
IE .ONP
IE. PRI
IE. RSU
IE. SDP
IE. SPC
IE. ULN
IE. UPN

11.7

are

subroutine

-::in l

..JV..L.

PROGRAMMING HINTS

This
section
contains
information
on
important
programming
considerations relevant to users of the UDCll driver described in this
chapter.

11.7.1

Checkpointable Tasks

Since checkpointable tasks are not allowed to have more than one
outstanding I/O request, a task that issues a request to connect to
timer or contact interrupts must not be checkpointable.

11-29

UNIVERSAL DIGITAL CONTROLLER DRIVER
11.7.2

Numbering Conventions

Numbering is relative. Module numbers start at 0, beginning with
first module of a given type.

the

Channel numbers also start at 0, with channel 0 as the first channel
on the first module of a given type. For the ADUOl, channel 8 is the
first channel on the second analog output module.
Each IAD-IA module installed in an ICSll subsystem occupies 120
channels (regardless of the number of multiplexers installed). In
this case, channel 120 is the first channel on the second IAD-IA A/D
converter.
Point numbers start at 0, with point 0 as the first point on the first
module of a given type. For instance, point 20 (octal) is the first
point of the second contact sense module (i.e., relative module number
1) •

11.7.3

Processing Circular Buffer Entries

Circular buffer entries should be processed in the following sequence.
1.

Execute a WAITFOR system directive using the trigger
event flag specified in the subroutine called to connect
the circular buffer (CTTI or CTDI).

Repeatedly call the appropriate subroutine to read the
circular buffer until all entries have been obtained, and
ipt indicates that the buffer is empty (-1).

Perform any other processing and return to step 1.

11-30

CHAPTER 12
LABORATORY PERIPHERAL SYSTEMS DRIVERS

12.1

INTRODUCTION

The LPSll and ARll Laboratory Peripheral Systems are
modular,
real-time subsystems used for
the acquisition and/or output of
laboratory analog data. Table 12-1 compares the LPSll with the ARll.
Table 12-1
Laboratory Peripheral Systems
LPSll

Analog-to-Digital Conversion
(with sample and hold
circuitry)

12 bits of precision
16-channel multiplexer
with gain ranging

10 bits of precision
16-channel multiplexer
without gain ranging

Maximum of 64 channels
without gain ranging
Prograrmnable Real-Time Clock

Yes

Digital-to-Analog OUtput

12 bits of precision
10 channels (including display)

10 bits of precision
2 channels (including
display)

Display Control

4096 by 4096 dot matrix

1024 by 1024 dot matrix

Digital I/O Option

16 digital points
and programmable
relays

16 digital points
(available with
DRll-K option)

At system generation, the user can specify the following:
Number of A/D channels
Presence or absence of the gain ranging option
(LPSAM-SG)
(LPSll only)
and
the polarity of each channel (uni- or
bi-polar) •
Presence or absence of the external D/A option
(LPSVC
LPSDA), and if present, the number of D/A channels.
Clock preset value

12-1

and

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.1.1

ARll Laboratory Peripheral System

The ARll is a one-module, real-time analog subsystem that interfaces
to the PDP-11 family of computers via a "hex" small peripheral
controller slot. The system is a subset of the LPSll, and as such,
enjoys
the
same degree of flexibility.
The ARll includes a
16-channel, 10-bit A/D converter with sample-and-hold, a programmable
real-time clock with one external input, and a display control with
two 10-bit D/A converters.

12.1.2

LPSll Laboratory Peripheral System

The LPSll is a high-performance, modular, real-time subsystem with the
flexibility of serving a variety of applications, including biomedical
research, analytical instrumentation, data collection and reduction,
monitoring,
data
logging, industrial testing, engineering, and
technical education. The basic subsystem, built in a compact size and
designed for easy interface with external instrumentation, includes a
12-bit A/D converter, a programmable real-time clock, with two Schmitt
triggers, a display controller with two 12-bit D/A converters, and a
16-bit digital I/O option. Up to nine different option types may be
added to the basic package.

12.2

GET LUN INFORMATION MACRO

If a GET LUN INFORMATION system directive is issued for a LUN
associated with a Laboratory Peripheral System, word 2 (the first
characteristics word) contains all zeros, words 3 and 4 are undefined,
and word 5 contains a 16-bit buffer preset value that controls the
rate of the real-time clock interrupts, as explained in section
12.6.1.

12.3

QIO MACRO

This section summarizes standard and device-specific QIO functions for
the Laboratory Peripheral System drivers.

12.3.l

Standard QIO Function

Table 12-2 lists the standard function of the QIO macro that is
for the Laboratory Peripheral Systems.
Table 12-2
Standard QIO Function for
Laboratory Peripheral Systems
Format

Function

QIO$C IO.KIL, ...

Cancel I/0 requests

IO.KIL cancels all queued and in-progress I/O requests.

12-2

valid

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.3.2

Device-Specific QIO Functions (Immediate)

Except for IO.STP (see section 12.3.4), all device-specific functions
of the QIO macro that are valid for the Laboratory Peripheral Systems
are either immediate or synchronous. Each immediate function performs
a
complete operation, whereas each synchronous function simply
initiates an operation synchronized to the real-time clock.
Table
12-3 lists the immediate functions.
Table 12-3
Device-Specific Functions for the
Laboratory Peripheral Systems (Immediate)
Format

Function

QIO$C IO.LED, ••• ,<int,num>

Display number in LED lights
( LPSll only)

QIO$C IO.REL, ••• ,<rel,pol>

Latch output relay (LPSll only)

QIO$C IO.SDI, ••• ,<mask>

Read digital input register

QIO$C IO.SDO, ••• ,<mask,data>

Write digital output register

where:

int

is the 16-bit signed binary integer to display.

num

is the LED digit number where the decimal point is
to be placed.

rel

is the relay number (zero or one) •

pol

is the polarity
closed) •

mask

is the mask word.

data

is the data word.

(zero

for

open,

nonzero

for

The following subsections describe the functions listed above.

12.3.2.1 IO.LED - This LPSll-only function displays a 16-bit signed
binary integer in the light-emitting diode (LED) lights. The number
is displayed with a leading blank (positive number) or minus sign
(negative number) followed by five nonzero-suppressed decimal digits
that represent the magnitude of the number. LED digits are numbered
from right to left, starting at 1.
The number may be displayed with or without a decimal point.
If the
parameter num is a number from 1 to 5, then the corresponding LED
digit is displayed with a decimal point to the right of the digit. If
the LED digit number is not a number from 1 to 5, then no decimal
point is displayed.

12-3

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.3.2.2 IO.REL - This LPSll-only function opens or closes
the
programmable relays in the digital I/O status register. Approximately
300 milliseconds are required to open or close a relay.
The driver
imposes no delays when executing this function.
Thus it is the
responsibility of the user to insure that adequate time has elapsed
between the opening and closing of a relay.

12.3.2.3 IO.SDI - This function reads data qualified by a mask word
from the digital input register. The mask word contains a 1 in each
bit position from which data is to be read.
All other bits are
zero-filled and the resulting value is returned in the second I/O
status word.
The operation performed is:
RETURN VALUE=MASK.AND.INPUT REGISTER

12.3.2.4 IO.SDO - This function writes data qualified by a mask word
into the digital output register. The mask word contains a 1 in each
bit position that is to be written. The data word specifies the data
to be written in corresponding bit positions.
The operation performed is:
NEW REGISTER=<MASK.AND.DATA>.OR.<<.NOT.MASK>.AND.OLD REGISTER>

12.3.3

Device-Specific QIO Functions (Synchronous)

Table 12-4 lists the synchronous, device-specific functions of the QIO
macro that are valid for the Laboratory Peripheral Systems.

Table 12-4
Device-Specific QIO Functions for the
Laboratory Peripheral Systems (Synchronous)
Function

Format
QIO$C IO.ADS, .•• ,<stadd,size,pnt,
ticks,bufs,chna>

Initiate A/D sampling

QIO$C IO.HIS, ••• ,<stadd,size,pnt,
ticks,bufs>

Initiate histogram sampling
(LPSll only)

QI0$C IO.MDA, ••. ,<stadd,size,pnt,
ticks,bufs,chnd>

Initiate D/A output

QI0$C IO.MDI, •.. ,<stadd,size,pnt,
ticks,bufs,mask>

Initiate digital input
sampling

QIO$C IO.MDO, .•• ,<stadd,size,pnt,
ticks,bufs,mask>

Initiate digital output

12-4

LABORATORY PERIPHERAL SYSTEMS DRIVERS
where:

stadd

is the starting address of the
(must be on a word boundary) .

data

buffer

size

is the data buffer size in
greater than zero and a
bytes) •

pnt

is
the
digital
point
numbers
(byte
0 - starting input/output point number; byte
1 - input point number to stop the function).

ticks

is the number of real-time clock
ticks
between
samples
or
data transfers, as
appropriate.

buf s

is the number of data buffers to transfer.

chna

is
the
analog-to-digital
conversion
specification.
Byte 0 contains the starting
channel number. For LPSll this must be in
the range of 0-63;
for ARll the range is
0-15. If the LPSll gain ranging option is
present, the channel number must be in the
range of 0-15, and bits 4 and 5 specify the
gain code.

bytes (must be
multiple of four

Byte 1 contains the number of consecutive
analog-to-digital channels to sample. For
LPSll this must be in the range of 1-64; for
ARll the range is 1-16.
chnd

is the digital-to-analog
output
channel
specification.
Byte 0 contains the starting
channel number. For LPSll this must be in
the range of 0-9; for the ARll the range is
0-1.

Byte 1 contains the number of consecutive
channels to be output. For LPSll this must
be in the range of 1-10; for ARll the range
lS

mask

1-2.

is the mask word.

The following subsections describe the functions listed above.

12.3.3.1 IO.ADS - This function reads one or more A/D channels at
precisely timed intervals, with or without auto gain-ranging. If two
or more channels are specified, all are sampled at approximately the
same time, once per interval.
Sampling may be started when the request is dequeued or when a
specified digital input point is set. A digital output point may
optionally be set when sampling is started.
Sampling may
be
terminated by a program request (IO.STP or IO.KIL), by the clearing of
a digital input point, or by the collection of a specified number of
buffers of data.

12-5

LABORATORY PERIPHERAL SYSTEMS DRIVERS
All input is double-buffered with respect to the user task. Each time
a half buffer of data has been collected, the user task is notified
(via the setting of an event flag)
that data is available to be
processed while the driver fills the other half of the buffer. If the
user task does not respond quickly enough, a data overrun may result.
This occurs if the driver attempts to put another data item in the
user buffer when no space is available.
The subfunction modifier bits are identical to those described in
section 12.3.3.2. In addition, setting bit 3 to a 1 means LPSll auto
gain-ranging is requested. Bit 3 is ignored for the ARll. If bits 7
and 6 are both set to 1, the digital input point and digital output
point number are assumed to be the same.
If LPSll auto gain-ranging is used, the LPSAM-SG hardware option must
be present and specified at system generation. The auto gain-ranging
algorithm causes a channel to be sampled at the highest gain at which
saturation does not occur. If the gain-ranging option is present and
auto gain-ranging is not specified in bit 3 of the subfunction code,
then bits 4 and 5 of the starting channel number specify the gain at
which samples are to be converted. Gain codes are as follows:
Code

Gain

00
01
10

16
64

Data words written into the user buffer contain the converted value in
bits 0-11 and the gain code, as shown below, in bits 12-15:
Code
0000
0001
0010
OOll

16
64

If the LPSAM-SG option is present, then the band pass filter jumpers
must not be clipped.
Also, each channel must have been defined as
uni- or bi-polar at system generation.
The ARll always returns data which is equivalent to an LPSll gain of
1.
Channel polarity must always be specified for the ARll at system
generation since this operation is software selectable.

12.3.3.2 IO.HIS - This LPSll-only function measures the elapsed time
between a series of events by means of Schmitt trigger one. Each time
a sample is to be taken, a counter is incremented and Schmitt trigger
one is tested. If it has fired, then the counter is written into the
user buffer and reset to zero. Thus the data item returned to the
user is the number of sample intervals between Schmitt trigger
firings.
If the counter overflows before Schmitt trigger one fires, then a zero
value is written into the user buffer. Sampling may be started and
stopped as
described
in
section
12.3.3.1.
All
input
is
double-buffered with respect to the user task.

12-6

LABORATORY PERIPHERAL SYSTEMS DRIVERS
The subfunction modifier bits appear below.
the action listed in the right-hand column.
Bit

A setting of 1

indicates

Meaning

0-3

Unused

Stop on number of buffers

Stop on digital input point clear

Set digital output point at start of operation

Start on digital input point
set
specification means start immediately)

zero

12.3.3.3 IO.MDA - This function writes data into one or more external
D/A converters at precisely timed intervals. If two or more channels
are specified, all are written at approximately the same time, once
per interval.
Output may be started or stopped as described in
section 12.3.3.1. All output is double-buffered with respect to the
user task.
D/A converters 0 and l correspond to the X and Y registers of the
display control. Note that there are no specific driver functions to
set the display status register. This is reserved for the user.
D/A
converters 2 through 9 correspond to the LPSll, LPSDA external D/A
option.
The subfunction modifier bits are
section 12.3.3.2.

identical

those

described

12.3.3.4 IO.MDI - This function provides the capability to read data
qualified by a mask word from the digital input register at precisely
timed intervals. Sampling may be started and stopped as described in
section 12.3.3.1.
All input is double-buffered with respect to the
user task.

The mask word contains a 1 in each bit position from which data is
be read. All other bits are zero.

The subfunction modifier bits are
section 12.3.3.2.

identical

those

described

12.3.3.5 IO.MOO - This function writes data qualified by a mask word
into the digital output register at precisely timed intervals. Output
may be started and stopped as described in section 12.3.3.1.
All
output is double-buffered with respect to the user task.
The subfunction modifier bits are
section 12.3.3.2.

identical

12-7

those

described

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.3.4

Device-Specific QIO Function (IO.STP}

Table 12-5 lists the device-specific function of the QIO macro,
is valid for the Laboratory Peripheral Systems.

which

Table 12-5
Device-Specific QIO Function for the
Laboratory Peripheral Systems (IO.STP}
Format

Function

QIO$C IO.STP, ... ,<stadd>

Stop in-progress request

where:

stadd

is the buffer address of the function to stop
(must be the same as the address specified in
the initiating request}.

12.3.4.1 IO.STP - IO.STP stops a single in-progress
synchronous
request.
It is unlike IO.KIL in that it only cancels the specified
request, whereas IO.KIL cancels all requests.

12.4

FORTRAN INTERFACE

A collection of FORTRAN-callable subroutines provide FORTRAN programs
access to the Laboratory Peripheral Systems.
These routines are
described in this section.
Some of these routines may be called from FORTRAN as
either
subroutines or functions.
All are reentrant and may be placed in a
resident library.

12.4.1

The isb Status Array

The isb (I/0 status block} parameter is a two-word integer array that
contains the status of the FCRTRAN call, in accordance with ISA
convention. This array serves two purposes:
1.

It is the 2-word I/0 status block to which the driver returns
an I/O status code on completion of an I/O operation.

The first word of the isb receives a status code from the
FORTRAN
interface
in
ISA-compatible format, with the
exception of the I/O pending condition, which is indicated by
a status of zero. The ISA standard code for this condition
is +2.

The meaning of its contents varies, depending on the FORTRAN call that
has been executed, but Table 12-6 lists certain general principles
that apply. The sections describing individual subroutines provide
more details.

12-8

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-6
Contents of First Word of isb
Meaning

Contents
isb(l} = 0

Operation pending;

isb(l} = 1

Successful completion

isb(l} = 3

Interface subroutine unable to generate QIO
directive, or illegal time or buffer value

<= isb(l) -< 300

isb(l) > 300

I/O in progress

directive rejected
code = - ( i sb ( 1} - 3)

and

actual

error

Driver rejected request
code = - ( isb ( 1) - 300)

and

actual

error

QIO

FORTRAN interface routines depend on asynchronous system traps to set
their status. Thus, if the trap mechanism is disabled, proper status
cannot be set.

12.4.2

Synchronous Subroutines

RTS, DRS, HIST (LPSll only), SDO, and SDAC are FORTRAN subroutines
that
initiate synchronous functions.
When they are used, the
appropriate laboratory peripheral system driver and the FORTRAN
program communicate by means of a caller-specified data buffer of the
following format:
Buffer Header

Current Buffer Pointer
Address of Second I/O Status Word
Address of End of Euf fer + 1
Address of Start of Data

Start of Data
Half Buffer
End of Buffer
The buffer header is initialized when the synchronous function
initiation routine is called. The length of the buffer must be even
and greater than or equal to six. An even length is required so that
the buffer is exactly divisible into half buffers.
The drivers perform double buffering within the half buffers.
Each
time a driver fills or empties a half buffer, it sets a user-specified
event flag to notify the user task that more data is available or
needed.
The user task responds by putting more data into the buffer
or by removing the data now available.
If the user task does not respond quickly enough, a data overrun may
result.
This occurs if the driver attempts to put another data item

12-9

LABORATORY PERIPHERAL SYSTEMS DRIVERS
in the user buffer when no space is available (i.e., the buffer is
full of data) or if the driver attempts to obtain the next data item
from the user buffer when none is available (i.e., the buffer is
empty) •
All synchronous functions may be initiated immediately or when· a
specified digital input point is set (i.e., a start button is pushed).
They may be terminated by any combination of a program request, the
processing of the reguired number of full buffers of data, or the
clearing of a specified digital input point (i.e., a stop button is
pushed).
A digital output point may also optionally be set at the
start of a synchronous function. This could be used, for example, as
a signal to start a test instrument.

12.4.3

FORTRAN Subroutine Summary

Table 12-7 lists the FORTRAN interface subroutines supported for the
Laboratory Peripheral Systems under RSX-llM. S and F indicate whether
they can be called as subroutines or functions.
Table 12-7
FORTRAN Interface Subroutines for Laboratory Peripheral Systems
Subroutine

Function

ADC

Read a single A/D channel (F,S)

ADJLPS

Adjust buffer pointers (S)

ASARLN

Assign a LUN to ARO:

(S)

ASLSLN

Assign a LUN to LSO:

(S)

CVSWG

Convert a switch gain A/D value to floating-point (F}

DRS

Initiate synchronous digital input sampling (S}

HIST

Initiate histogram sampling (S}

!DIR

Read digital input

!DOR

Write digital output (F,S}

IRDB

Read data from
(F,S)

LED

Display number in LED lights (S}

LPSTP

Stop an in-progress synchronous function (S)

PUTD

Put data into a synchronous function output buffer (S}

RELAY

Latch an output relay (S)

RTS

J--·---------------------·----··---------·---·····-------------

......__-~~·-·-----

(LPSll only}

(F,S)

synchronous

function

buffer

(LPSll only)

Initiate synchronous A/D sampling {S)

12-10

input

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-7 (Cont.)
FORTRAN Interface Subroutines for Laboratory Peripheral Systems
Subroutine

Function

SDAC

Initiate synchronous D/A output

SDO

Initiate synchronous digital output

( s)
( s)

The following subsections briefly describe the function and format
each FORTRAN subroutine call.

12.4.4

ADC:

Reading a Single A/D Channel

The ADC FORTRAN subroutine or function reads a single converted value
from an A/D channel.
If the gain-ranging option is present in the
LPSll hardware, the channel may be converted at a specific gain or the
driver can select the best gain (the gain providing the most
significance). The converted value is returned as a normalized
floating-point number. The call is issued as follows:
CALL ADC { ichan, [var] , [ iga in] , [ isb])

where:

ichan

specifies the A/D channel to be converted.

var

is a floating-point variable that receives
the converted value in floating-point format.

igain

specifies the gain at which the specified A/D
channel is to be converted. The default is
1.
If specified, igain
may
have
the
following values:
Gain
Auto gain-ranging
(driver
provides
gain
that
significance)

A gain of
driver.
isb

the

ARll

is a 2-word integer array to
subroutine status is returned.

which

the

12-11

always

used

selects
most

LABORATORY PERIPHERAL SYSTEMS DRIVERS
The isb array has the standard meaning described in section 12.4.l.
When the function form of the call is used, the value of the function
is the same as that returned in var. If this value is negative, an
error has occurred during the A/D conversion {see section 12.5.3).
Otherwise, this value is a floating-point number calculated from the
following formula:
var = (64

12.4.5

ADJLPS:

* converted value) I

conversion gain

Adjusting Buffer Pointers

The ADJLPS FORTRAN subroutine adjusts buffer pointers for a buffer
that a laboratory peripheral system driver is either synchronously
filling or emptying. It is usually called when indexing is being used
for direct access to the data in a buffer.
When data in a buffer is to be processed only once, the IRDB and PUTD
routines may be used. In some cases, however, it is useful to leave
data in the buffer until processing is complete. The user program may
process the data directly, then call ADJLPS to free half the buffer.
Use of the routine for synchronous output functions is quite similar.
When a half buffer of data is ready for output, ADJLPS is called to
make the half buffer available.
When ADJLPS is used for either input or output, care must be taken to
insure that the program stays in sync with the driver. If the program
loses its position with respect to the driver, the function must be
stopped and restarted. An attempt to over-adjust will cause a 3 to be
returned in isb(l) and no adjustment to take place.
The call is issued as follows:
CALL ADJLPS (ibuf,iadj,[isb]}
where:

ibuf

is an integer array which was previously
specified in a synchronous input or output
function.

iadj

specifies the adjustment to be applied to the
buffer pointers. For an input function this
specifies the number of data values that have
been removed from the data buffer. For an
output function this specifies the number of
data values that have been put into the data
buffer.

isb

is a 2-word integer array to
subroutine status is returned.

which

The isb array has the standard meaning described in section 12.4.l.

12-12

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.4.6

ASLSLN:

Assigning a LUN to LSO:

The ASLSLN FORTRAN subroutine assigns a logical unit number
(LUN)
to
the LPSll. It must be called prior to executing any other Laboratory
Peripheral Systems FORTRAN function or subroutine.
Subsequent calls
to other interface routines then implicitly reference the LPSll via
the LUN assigned.
The call is issued as follows:
CALL ASLSLN (lun,[isb])
where:

lun

is the number of the LUN to
LSO:

isb

is a 2-word integer array to
subroutine status is returned.

assigned

which

the

The isb array has the standard meaning described in section 12.4.1.

12.4.7

ASARLN:

Assigning a LUN to ARO:

The ASARLN FORTRAN subroutine assigns a logical unit number
(LUN)
to
the ARll.
It must be called prior to executing any other laboratory
peripheral system FORTRAN function or subroutine. Subsequent calls to
other interface routines then implicitly reference the ARll via the
LUN assigned.
The call is issued as follows:
CALL ASARLN (lun,[isb])
where:

lun

is the number of the LUN to
ARO:

isb

is a 2-word integer array to
subroutine status is returned.

assigned

which

the

The isb array has the standard meaning described in section 12.4.1.

12.4.8

CVSWG:

Converting a Switch Gain A/D Value to Floating-Point

The CVSWG FORTRAN subroutine converts an A/D value from a synchronous
A/D sampling function to a floating-point number. Each data item
returned by a laboratory peripheral system driver consists of a gain
code and converted value packed in a single word
(see section
12.3.3.l). This form is not readily usable by FORTRAN, but is much
more efficient than converting each value to floating-point in the
driver. This routine unpacks the gain code and value, then converts
the result to a floating-point number. It may be conveniently used in
conjunction with the IRDB routine (see section 12.4.13).
The call is issued as follows:
CVSWG (ival)

12-13

LABORATORY PERIPHERAL SYSTEMS DRIVERS
where:

iv al

is the value to be converted to floating
point. Its format must be that returned by a
synchronous A/D sampling
function.
The
conversion is performed according to the
following formula:
var = (64

* converted value)/conversion

gain

For the various gain codes,
var

= x * converted value

as shown below:
x

12.4.9

DRS:

Initiating Synchronous Digital Input Sampling

The DRS FORTRAN subroutine reads data qualified by a mask word from
the digital input register at precisely timed intervals. Sampling may
be started or stopped as for RTS (see section 12.4.18) and all input
is double-buffered with respect to the user task.
Data may be
sequentially retrieved from the data buffer via the IRDB routine (see
section 12.4.12), or the ADJLPS routine (see section 12.4.5) may be
used in conjunction with direct access to the input data. The call is
issued as follows:
CALL DRS (ibuf,ilen,imode,irate,iefn,imask,isb,[nbuf],
[is tart], [is top])
where:

ibuf

is an integer array that is
input data values.

ilen

specifies the length uf ibuf (must
and greater than or equal to six).

imode

specifies the start, stop, and sampling mode.
Its value is encoded by adding together the
appropriate function selection values shown
below.

12-14

receive
be

the
even

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Function
Selection
Value

Meaning

128

Start on digital
set

Set digital
start

Stop on
clear

Stop on number of buffers

input

output

digital

point

input

point

Thus a value of 192 for imode specifies:
The sampling is to be started when a specified digital
point is set.

input

A digital output point is to be set when sampling is started.
Sampling will be stopped via a program request.
irate

is a 2-word integer array that specifies the
time interval between digital input samples.
The first word specifies the interval units
as follows:
irate ( 1)

Unit

Real-time clock ticks

Milliseconds

Seconds

Minutes

The second word specifies
the
interval
magnitude as a 16-bit unsigned integer.
iefn

specifies the number of the event flag that
is to be set each time a half buffer of data
has been collected.

imask

specifies the
read.

isb

is a 2-word integer array to
subroutine status is returned.

nbuf

specifies the number of buffers of data to be
collected.
It is needed only if a function
selection value of 16 has been added into
imode.

12-15

digital

input

points

which

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
is tart

specifies the digital input pointer number to
be
used to trigger sampling and/or the
digital output point number to be set when
sampling is started. It is needed only if a
function selection value of 128 or 64 has
been added into imode.

is top

specifies the digital input point number to
be used to stop sampling. It is needed only
if a function selection value of 32 has been
added into imode.

When sampling is in progress, the first word of the isb array is zero
and the second word contains the number of data values currently in
the buffer.

12.4.10

HIST:

Initiating Histogram Sampling (LPSll only)

The HIST FORTRAN subroutine measures the elapsed time between a series
of events via Schmitt trigger one.
Each time a sample is to be taken, a counter is incremented and
Schmitt trigger one is tested. If it has fired, then the counter is
written into the user buffer and the counter is reset to zero.
Thus
the data returned to the user is the number of sample intervals
between Schmitt trigger firings.
If the counter overflows before
Schmitt trigger one fires, a zero value is written into the user
buffer. Sampling may be started and stopped as for RTS (see section
12.4.18) and all input is double-buffered with respect to the user
task. The call is issued as follows:
CALL HIST (ibuf,ilen,imode,irate,iefn,isb,[nbuf],
[is tart], [is top])
where:

ibuf

is an integer array that is
input data values.

ilen

specifies the length of ibuf (must
and greater than or equal to six).

imode

specifies the start, stop and sampling mode.
Its
value
is
encoded
by
adding the
appropriate function selection values shown
below:
Function
Selection
Value

the

even

Meaning

128

Start of digital
set

Set digital
start

12-16

receive

input

output

point

LABORATORY PERIPHERAL SYSTEMS DRIVERS

irate

Stop on
clear

digital

input

Stop on number of buffers

point

is a 2-word integer array that specifies the
time interval between samples.
The first
word specifies the interval units as follows:
irate(l)

Unit

Real-time clock ticks

Milliseconds

Seconds

Minutes

The second word specifies
the
interval
magnitude as a 16-bit signed integer.
iefn

specifies the number of the event flag that
is to be set each time a half buffer of data
has been collected.

isb

is a 2-word integer array to
subroutine status is returned.

nbuf

specifies the number or butters of data to be
collected.
It is needed only if a function
selection value of 16 has been added into
imode.

is tart

specifies the digital input point number to
be
used to trigger sampling and/or the
digital output point number to be set when
sampling is started. It is needed only if a
function selection value of 128 or 64 has
been added into imode.

is top

specifies the digital input point number to
be used to stop sampling. It is needed only
if a function selection value of 32 has been
added into imode.

which

the

The isb array has the standard meaning described in section 12.4.1.
When sampling is in progress, the first word of the isb array is zero
and the second word contains the number of data values currently in
the buffer.

12-17

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.4.11

IDIR:

Reading Digital Input

The IDIR FORTRAN subroutine or function reads the digital input
register as an unsigned binary integer or as four binary-coded decimal
(BCD) digits. In the latter case, the BCD digits are converted to a
binary integer before the value is returned to the caller. The call
is issued as follows:
CALL IDIR ( imode, [ ival], [ isb])
where:

imode

specifies the mode in which the digital input
register is to be read.
If imode equals
zero, then the digital input register is read
as four BCD digits and converted to a binary
integer. Otherwise it is read as a 16-bit
unsigned binary integer.

iv al

is a variable that receives the value read.

isb

is a 2-word integer array to
subroutine status is returned.

which

the

The isb array has the standard meaning described in section 12.4.1.
When the function form of the call is used, the value of the
is the same as that returned in ival.

12.4.12

IDOR:

function

Writing Digital Output

The IDOR FORTRAN subroutine or function clears or sets bits in the
digital output register. The caller provides a mask word and output
mode. Bits in the digital output registers corresponding to the bits
specified in the mask word are either set or cleared according to the
specified mode. The call is issued as follows:
CALL IDOR (imode,imask,[newval] ,[isb])
where:

imode

specifies whether the bits specified by imask
are to be cleared or set in the digital
output register. If imode equals zero, then
the bits are to be cleared. Otherwise they
are to be set.

imask

specifies the bits to be cleared or set in
the digital output register.
It may be
conveniently specified as an octal constant.

newval

is a variable that receives
(actual)
value written into
output register.

isb

is a 2-word integer array to
subroutine status is returned.

12-18

the
the

updated
digital

which

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
The isb array has the standard meaning described in section 12.4.1.
When the function form of the call is used, the value of the
is the same as that returned in newval.

12.4.13

IRDB:

function

Reading Data from an Input Buffer

The IRDB FORTRAN subroutine or function retrieves data sequentially
from
a
buffer that a laboratory peripheral system driver is
synchronously filling. If no data is available when the call is
executed, the contents of the next location in the data buffer are
returned without updating the buffer pointers. The call is issued as
follows:
CALL IRDB (ibuf,[ival])
where:

ibuf

is an integer array which was previously
specified in a synchronous input sampling
request (i.e., DRS, HIST, or RTS).

iv al

is a variable that receives the next value in
the data buffer.

When the function form of the call is used, the value of the
is the same as that returned in ival.

12.4.14

LED:

function

Displaying in LED Lights (LPSll only)

The LED FORTRAN subroutine displays a 16-bit signed binary integer in
the LED lights.
The number is displayed with a leading blank
(positive number) or minus (negative number), followed by five
non-zero suppressed decimal digits that represent the magnitude of the
number. LED digits are numbered right to left starting at 1 and
continuing to 5.
The number may be displayed with or without a
decimal point. The call is issued as follows:
CALL LED (ival,[idec],[isb])
where:

ival

is the variable
displayed.

idec

specifies the position of the decimal point.
A value of 1 to 5 specifies that a decimal
point is to be displayed. All other values
specify that no decimal point is to be
displayed.

isb

is a 2-word integer array to
subroutine status is returned.

whose

value

which

The isb array has the standard meaning described in section 12.4.1.

12-19

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
For example, the following call
CALL LED (-55,2)
would cause -0005.5 to be displayed in the LED lights.

12.4.15

LPSTP:

Stopping an In-Progress Synchronous Function

The LPSTP FORTRAN subroutine selectively stops
request. The call is issued as follows:

single

synchronous

CALL LPSTP (ibuf)
where:

12.4.16

ibuf

PUTD:

is an integer array that specifies a
that
was
previously
specified
synchronous initiation request.

buffer
in
a

Putting a Data Item into an Output Buffer

The PUTD FORTRAN subroutine puts data sequentially into a buffer that
a laboratory peripheral system driver is synchronously emptying. If
no free space is available, no operation is performed.
The call is
issued as follows:
CALL PUTD (ibuf ,ival)
where:

12.4.17

ibuf

is an integer array which was previously
specified in a synchronous output request
( SDO or SDAC) .

iv al

is a variable whose value is to be placed in
the next free location in the data buffer.

RELAY:

Latching an Output Relay (LPSll only)

The RELAY FORTRAN subroutine opens or closes the
call is issued as follows:

LPSll

relays.

The

CALL RELAY {irel,istate,[isb])
where:

irel

specifies which relay is to be opened or
closed {one for relay one, two for relay
two) .

istate

specifies whether the relay is to be opened
or closed. If istate equals zero, the relay
is to be opened.
Otherwise it is to be
closed.

isb

is a 2-word integer array to
subroutine status is returned.

which

The isb array has the standard meaning described in section 12.4.1.

12-20

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.4.18

RTS:

Initiating Synchronous A/D Sampling

The RTS FORTRAN subroutine reads one or more A/D channels at precisely
timed
intervals, with or without auto gain-ranging.
The auto
gain-ranging algorithm (LPSll only) causes the channels to be sampled
at the highest gain at which saturation does not occur.
Sampling may be started when the interface subroutine is called or
when a specified digital input point is set. A digital output point
may optionally be set when sampling is started.
Sampling may be
terminated by a program request (stop-in-progress request or kill
I/O), the clearing of a digital input point, or the collection of a
specified number of buffers of data.
All input is double-buffered with respect to the user task. Each time
a half buffer of data has been collected; the user task is notified
(via the setting of an event flag)
that data is available to be
processed while the driver fills the other half of the buffer. Data
may be sequentially retrieved from the data buffer via the IRDB
routine (see section 12.4.12, or the ADJLPS routine (see section
12.4.5) may be used in conjunction with direct access to the input
data. The call is issued as follows:
CALL RTS (ibuf,ilen,imode,irate,iefn,ichan,nchan,isb,
[nbuf], [is tart], [is top])
where:

ibuf

is an integer array that is
converted data values.

ilen

specifies the length of ibuf (must
and greater than or equal to six).

imode

specifies the start, stop, and sampling mode.
Its value is encoded by adding together the
appropriate function selection values
as
shown below:
Fune tion
Selection
Value

receive

the

even

Meaning

128

Start on digital
set

Set digital
start

Stop on
clear

Stop on number of buffers

Auto gain-ranging (LPSll only)

12-21

input

output

digital

point

input

point

LABORATORY PERIPHERAL SYSTEMS DRIVERS
irate

is a 2-word integer array that specifies the
time interval between A/D samples. The first
word specifies the interval unit as follows:
irate (1)

Unit

Real-time clock ticks

Milliseconds

Seconds

Minutes

The second word specifies
the
interval
magnitude as a 16-bit unsigned integer.
iefn

specifies the number of the event flag that
is to be set each time a half buffer of data
has been collected.

ichan

specifies the starting A/D channel of the
block of channels to be sampled synchronously
(must be between 0 and 63 for LPSll and
between 0 and 15 for ARll).

nchan

specifies the number of A/D channels to be
sampled (must be between 1 and 64 for LPSll
and between 1 and 16 for ARll).

isb

is a 2-word integer array to
subroutine status is returned.

nbuf

specifies the number of buffers of data that
are to be collected. It is needed only if a
function selection value of 16 has been added
in to imode.

is tart

is top

specifies the digital input point number to
be used to stop sampling. It is needed only
if a function selection value of 32 has been
added into imode.

which

the

The values listed for ichan and nchan above, are the maximum allowable
for each of the devices. In practice, they are constrained by the
number of channels available as specified during SYSGEN.
The isb parameter
12.4.1.

has

the

standard

meaning

described

section

When sampling is in progress, the first word of the isb array is zero
and the second word contains the number of data values currently in
the buffer.

12-22

LABORATORY PERIPHERAL SYSTEMS DRIVERS
12.4.19

SDAC:

Initiating Synchronous D/A Output

The SDAC FORTRAN subroutine writes data into one or more external D/A
converters at precisely timed intervals. Output may be started and
stopped as for RTS (see section 12.4.18) and
all
input
is
double-buffered with respect to the user task. One full buffer of
data must be available when synchronous output is initiated.
After SDAC has initiated output and the specified event flag has been
set to notify the task that free buffer space is available, the PUTD
routine (see section 12.4.16) may be used to put data values
sequentially into the output data buffer. The ADJLPS routine (see
section 12.4.5) may be used in conjunction with direct access to the
output data buffer. The SDAC call is issued as follows:
CALL SDAC (ibuf ,ilen,imode,irate,iefn,ichan,nchan,isb,
[nbuf], [is tart], [is top])
where:

ibuf

is an integer array that contains the
data values.

ilen

specifies the length of ibuf (must
and greater than or equal to six) .

imode

specifies the start, stop and sampling mode.
Its value is encoded by adding together the
appropriate function selection values
as
shown below:
Function
Selection
Values

irate

output
be

even

Meaning

128

Start on digital
set

Set digital
start

Stop on
clear

Stop on number of buffers

input

output

digital

point

input

point

is a 2-word integer array that specifies the
time interval between D/A outputs. The first
word specifies the interval units as follows:
irate ( 1)

Unit

Real-time clock ticks

Milliseconds

Seconds

Minutes

The second word specifies
the
interval
magnitude as a 16-bit unsigned integer.

12-23

LABORATORY PERIPHERAL SYSTEMS DRIVERS
iefn

specifies the number of the event flag that
is to be set each time a half buffer of data
has been output.

ichan

specifies the starting D/A channel of the
block
of
channels
to be written into
synchronously (must be between 0 and 9 for
LPSll and be 0 or 1 for ARll).

nchan

specifies the number of D/A channels to be
written into (must be between 1 and 10 for
LPSll and be 1 or 2 for ARll).

isb

is a 2-word integer array to
subroutine status is returned.

nbuf

specifies the number of buffers of data to be
output.
It is needed only if a function
selection value of 16 has been added into
imode.

is tart

is top

specifies the digital input point number to
be used to stop sampling. It is needed only
if a function selection value of 32 has been
added into imode.

which

the

The isb array has the standard meaning described in section 12.4.1.
When sampling is in progress, the first word of the isb array is zero
and the second word contains the number of free positions in the
buffer.

12.4.20

SDO:

Initiating Synchronous Digital Output

The SDO FORTRAN subroutine writes data qualified by a mask word into
the digital output register at precisely timed intervals. Sampling
may be started and stopped as for RTS (see section 12.4.18) and all
input is double-buffered with respect to the user task. One full
buffer of data must be available when output is initiated.
After SDO has initiated output and the specified event flag has been
set to notify the task that free buffer space is available, the PUTD
routine (see section 12.4.16) may be used to put data values
sequentially into the output data buffer. The ADJLPS routine (see
section 12.4.5) may be used in conjunction with direct access to the
output data buffer. The SDO call is issued as follows:
CALL SDO (ibuf,ilen,imode,irate,iefn,imask,isb,[nbuf],
[ ist.ar t], [is top])

12-24

LABORATORY PERIPHERAL SYSTEMS DRIVERS
where:

ibuf

is an integer array that contains the digital
output values.

ilen

specifies the length or iouf (must
and greater than or equal to six) .

imode

specifies the start, stop, and sampling mode.
Its value is encoded by adding together the
appropriate function selection values
as
shown below:
Fune ti on
Selection
Value

irate

even

Meaning

128

Start on digital
set

input

Set digital
start

Stop on
clear

Stop on number of buffers

output

digital

point

input

point

is a 2-wora integer array that specifies the
time interval between digital outputs. The
first word specifies the interval units as
follows:
irate (1)

Unit

Real-time clock ticks

Milliseconds

Seconds

Minutes

The second word specifies
the
interval
magnitude as a 16-bit unsigned integer.
iefn

specifies the number of the event flag that
is to be set each time a half buffer of data
has been output.

imask

specifies the digital output points that are
to
be written.
It may be conveniently
specified as an octal constant.

isb

is a 2-word integer array to
subroutine status is returned.

nbuf

specifies the number of buffers of data to be
output.
It is needed only if a function
selection value of 16 has been added into
imode.

12-25

which

the

LABORATORY PERIPHERAL SYSTEMS DRIVERS
is tart

is top

specifies the digital input point number to
be used to stop sampling. It is needed if a
function selection value of 32 has been added
into imode.

The isb parameter
12.4.1.

has

the

standard

meaning

described

section

When sampling is in progress, the first word of the isb array is zero
and the second word contains the number of free positions in the
buffer.

12.5

STATUS RETURNS

The error and status conditions listed in Table 12-8 are returned by
the Laboratory Peripheral System drivers described in this chapter.
Table 12-8
Laboratory Peripheral Systems Status Returns
Code

rs.sue

IS.PND

I/0 request pending
The operation specified in the QIO
not yet been completed.

IE.ABO

IE.BAD

directive

has

The specified I/O operation was canceled
IO.KIL or IO.STP) while in progress.

(via

Operation aborted

Bad parameter
An illegal specification was supplied for one or
more of the device-dependent QIO parameters (words
6-11). The second I/O status word is filled with
zeros.

12-26

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-8 (Cont.)
Laboratory Peripheral Systems Status Returns
Code

Reason

IE.BYT

Byte-aligned buffer specified
Byte alignment was specified for a data buffer but
only
word alignment is legal for Laboratory
Peripheral Systems. Alternately, the length of a
buffer is not an even number of bytes.

IE.DAO

Data overrun
For Laboratory Peripheral Systems, the driver
attempted to get a value from the user buffer when
none was available or attempted to put a value in
the user buffer when no space was available.

IE.DNR

Device not ready
The physical device unit specified in the QIO
directive was not ready to perform the desired I/0
operation.
For Laboratory Peripheral Systems,
this code is returned if a device time-out occurs
I while a function is in progress. The second I/O
status word contains the number of free positions
in the buffer, as appropriate.

IE.IEF

Invalid event flag number
An invalid event flag number wab specified in a
synchronous function (i.e., an event flag number
that was not in the range l to 64).

IE.IFC

Illegal function

IE.NOD

A function code was included in an I/O
that is illegal for the LPSll or ARll.

request

Insufficient buffer space
Dynamic storage space has been depleted, and there
is insufficient buffer space available to allocate
a secondary control block for a
synchronous
function.

IE.OFL

Device off-line
The physical device unit associated with the LUN
specified in the QIO directive was not on-line.
When the system was booted, a device
check
indicated that this physical device unit was not
in the configuration.

12-27

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-8 (Cont.)
Laboratory Peripheral Systems Status Returns
Code
IE .ONP

Reason
Option not present
An option dependent function or subfunction was
requested,
and the required feature was not
specified at system generation. For example the
gain-ranging option or D/A option is not present.
The second I/0 status word contains zeros.

IE. PIH

Privilege violation
The task which issued the request
was
not
privileged
to
execute
that
request.
For
Laboratory Peripheral Systems, a checkpointable
task attempted to execute a synchronous sampling
function.

IE. RSU

Resource in use
A resource needed by the function requested in the I
QIO directive was being used (see section 12.5.1).

IE. SPC

Illegal address space
The buffer specified for a read or write request
was partially or totally outside the address space
of the issuing task. Alternately a byte count of
zero was specified.
The second I/O status word
contains zeros.

FORTRAN interface values for these status
section 12.5.4.

12. 5. 1

returns

are

presented

IE. RSU

IE.RSU is returned if a function requests a resource that is currently
being used.
The requesting task may repeat the request at a later
time or take any alternative action required.
Because certain functions do not need such resources, they never cause
this code to be returned. Other functions return this code under the
following conditions:
Function

When IE. RSU Is Returned

IO. SDO

One or more spec i f ied digital output bi ts are in use

IO.ADS

Digital output point (if specified) is in use

IO.HIS

Digital output point (if specified} is in use

IO.MDA

Digital output point (if specified) is in use

12-28

LABORATORY PERIPHERAL SYSTEMS DRIVERS
IO.MDI

Digital output point (if specified)
points to be sampled are in use

,.."

u1g1cai output point (if specified) or output
be written are in use

••T"'\i""'\

J.U • l'lLJU

digital

input

bits

The following components of the Laboratory Peripheral Systems are each
considered a single resource:
Resource

When Sharable

The A/D Converter
and clock

Always sharable.

Each bit in the
digital output
register

Never sharable.

Each bit in the
digital input
register

Always sharable when used by IO.SDI
or for start/stop conditions (specified
in subfunction modifier bits), even when
in
use
by another function;
when
specified by a synchronous digital input
function, not sharable with another such
function.

Each resource is allocated on a first-come-first-served basis (i.e.,
when a conflict arises, the most recent request is rejected with a
status of IE.RSU).

12.5.2

Second I/O Status Word

On successful completion of a function specified in a QIO macro call,
the rs.sue code is returned to the first word of the I/O status block
Table 12-9 lists the contents of the second word of the status
on successful completion for each function.

block,

Table 12-9
Returns to Second Word of I/O Status Block
Successful
Function

Contents of Second Word

IO.KIL

Number of data values before I/O was canceled

IO. LED

Zero

IO.REL

Zero

IO.SDI

Masked value read from digital input register

IO.SDO

Updated value written into digital output register

12-29

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-9 (Cont.)
Returns to Second Word of I/O Status Block
Successful
Function

Contents of Second Word

IO.ADS

Number of data values remaining in buffer

IO. HIS

Number of data values remaining in buffer

IO.MDA

Number of free positions in buffer

IO.MDI

Number of data values remaining in buff er

IO.MOO

Number of free positions in buffer

IO. STP

Zero

When IE.BAD is returned, the second I/O status word contains zero.
Laboratory Peripheral Systems drivers return the IE.BAD code under the
following conditions:
Function

When IE.BAD is Returned

IO.REL

Relay number not 0 or 1

IO.ADS
IO.MDA

No I/O status block, illegal digital I/O point
number, or illegal channel number

IO.HIS
IO.MDI
IO.MOO

No I/O status block or illegal
digital I/O point number

12.5.3

IO.ADS and ADC Errors

While IO.ADS or the ADC FORTRAN subroutine is converting a sample, two
error conditions may arise. Both of these conditions are reported to
the user by placing illegal values in the data buffer. A -1 (177777
octal)
is placed in the buffer if an A/D conversion does not complete
within 30 microseconds. A -2 (177776 octal) is placed in the buffer
if an error occurs during an A/D conversion (LPSll only).

12.5.4

FORTRAN Interface Values

The values listed in Table 12-10 are returned
calls.

12-30

FORTRAN

subroutine

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Table 12-10
FORTRAN Interface Values
Status Return

FORTRAN Value

Is.sue

+01
+00
+315
+101
+334
+301
+319
+313
+303
+100
+302
+99
+323
+365
+305
+316
+317
+102
+306
+08
+04

IS. PND
IE .ABO
IE .ADP
IE .ALN
IE .BAD
IE. BYT
IE. DAO
IE .DNR
IE. IEF

IE.IFC
IE. ILU
IE .NOD
IE .OFL
IE .ONP
IE. PRI
IE .RSU
IE. SDP
IE. SPC
IE. ULN
IE. UPN

12.6

PROGRAMMING HINTS

This
section
contains
information
on
important
programming
considerations relevant to users of the Laboratory Peripheral Systems
drivers described in this chapter.

12.6.l

The LPSll/ARll Clock and Sampling Rates

The basic real-time clock frequency (count rate} for all synchronous
functions is always lOKHz. Device characteristics word 4 contains a
16-bit buffer preset value, set dynamically or at system generation,
that controls the rate of "ticks" (i.e., the rate at which the clock
interrupts}. The quotient that results when this value is divided
into lOKHz is the rate of "ticks". For example, if this value is 2,
the "tick" rate is 5KHz. The user may use a GET LUN INFORMATION
system directive to examine the value and a SET /BUF MCR function to
modify it while the system is running.
The ticks parameter in a synchronous function specifies the number of
"ticks" between samples or data transfers. The value of ticks is a
16-bit number. Thus 65,536 discrete sampling frequencies are possible
for each of 65,536 different "tick" rates. This provides a maximum
single-channel sample rate of 1 sample every 100
microseconds
(possible in hardware but impractical in software} and a minimum of 1
sample every 429,495 seconds. A single-channel rate greater than 2KHz
is possible, but not recommended.

12-31

LABORATORY PERIPHERAL SYSTEMS DRIVERS
The figures below represent initial timing tests run under RSX-llM.
It should be noted that no computation was performed on the data other
than continuously removing it from or inserting it into the data
buffer.
The following data is for the LPSll
management and no gain-ranging option.

PDP-11/40

with

memory

Analog rates:
1 request for 1 channel at 2.5KHz
1 request for 2 channels at 2.0KHz (aggregate 4KHz}
2 requests for 1 channel at 2.0KHz (aggregate 4KHz)
Digital rates:
1 request for 2 channels at 2.5KHz (aggregate 5KHz}
The following data is for the ARll on a PDP-11/40 with no
management, no digital I/O option, and no uni-polar sampling.

memory

Analog rates:
1 request for 1 channel at 3.3KHz
1 request for 2 channels at 2.5KHz (aggregate 5.0KHz}

2 requests for 1 channel at 2.5KHz (aggregate 5.0KHz)
Digital rate:
2 requests for 2 channels at 3.3KHz (aggregate 6.6KHz)

12.6.2

Importance of the I/O Status Block

An I/O status block must be specified with every synchronous function.
If the first I/O status word is nonzero, the request has been
terminated and the value indicates the reason for termination.
Otherwise, the request is in progress, and the second I/O status word
contains the number of data values remaining in the buffer (or the
number of free positions in the buffer for IO.MDA and IO.MOO}.

12.6.3

Buffer Management

The buffer unload protocol for synchronous input functions
is
described below. The user constructs a five-word block that contains
the following:
IOSB:
CURPT:
LSTPT:
FSTPT:

.BLKW
.WORD
.WORD
.WORD

I/0 STATUS DOUBLE-WORD
ADDRESS OF BUFFER
ADDRESS OF END OF BUFFER
ADDRESS OF BUFFER

2
BUFFER
BUFFER+n
BUFFER

12-32

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Two of these words are required by the driver (I/O status block) and
the remaining three by the user to unload data values from the buffer.
The user then issues the I/O request with the appropriate parameters
and the address of the above block as the I/O status block. The QIO
directive zeros both I/O status words to initialize them.
If the driver accepts the request, it sets up a write pointer to the
first word in the user buffer.
Thus the user has a buffer read
pointer and the driver has a buffer write pointer. The user and the
driver share the second I/O status word, which is the number of data
words in the buffer that contain data.
Each time the driver attempts to put a sample value into the buffer,
it increments the second I/O status word and compares the result with
the size of the buffer. If the result is greater, buffer overrun has
occurred and the request is terminated. Otherwise the vaiue is stored
in the buffer at the address specified by the driver's write pointer
and the write pointer is updated.
If the value stored in the user buffer fills half of the buffer, the
event flag specified in the I/O request is set in order to notify the
user that a half buffer of data is available to be processed.
Each
time the user task is awakened, it should execute the following code:
AO Ohl
;CLEAR EFN
TST
IOSB+2
;ANY DATA IN BUFFER?
BEQ
;IF EQ NO
30$
MOV
@CURPT,RO
;GET NEXT VALUE FROM BUFFER
DEC
IOSB+2
;REDUCE NUMBER OF ENTRIES
ADD
;UPDATE BUFFER READ POINTER
#2,CURPT
CURPT,LSTPT
CMP
;END OF BUFFER?
BLOS
;IF LOS NO
20$
MOV
FSTPT,CURPT
;RESET BUFFER READ POINTER
Process data value
BR
;TRY AGAIN
10$
mcmn
;REQUEST TERMINATED?
IOSB
BNE
;IF NE YES
40$
WTSE$S
;WAIT FOR EFN
#EFN
BR
5$
i
Determine reason for termination
~TOQCC

~u~~y~

20$:
30$:

40$:

tr~~"

i~io

For IO.MDA and IO.MDO, this protocol differs slightly. The user task
maintains a write pointer and the driver a read pointer. The entire
buffer must be full when the request is executed.

12.6.4

Use of ADJLPS for Input and Output

The following FORTRAN example illustrates the
using ADJLPS for synchronous input and output.

12-33

proper

protocol

for

LABORATORY PERIPHERAL SYSTEMS DRIVERS
Synchronous input:
DIMENSION IBF(l004) ,IERR(2) ,INTVL(2)

C INITIATE SYNCHRONOUS A/D SAMPLING,

INTVL(l)=2
INTVL(2)=5
CALL RTS(IBF,1004,160,INTVL,IEFN,6,6,IERR,50,l6,l5)

C INITIALIZE HALF BUFFER INDEX

INDX=4

c
c WAIT FOR HALF BUFFER OF DATA
c
c

CALL WAITFR(IEFN)

C CLEAR EVENT FLAG

CALL CLREF(IEFN)

c
c PROCESS HALF BUFFER OF DATA
c

SUM=O
DO 20 I=l,500
SUM=SUM+CVSWG(IBF(I+INDX))
CONTINUE
AVERG=SUM/500

C FREE HALF BUFFER FOR MORE DATA

CALL ADJLPS(IBF,500)

C ADJUST BUFFER INDEX

c
c

INDX=INDX+500
IF(INDX.GE.1004)

INDX=4

C CHECK IF ANOTHER HALF BUFFER OF DATA IS AVAILABLE

IF(IERR(2) .GE.500 GO TO 15
IF(IERR(l) .NE.0) GO TO end of sampling
GO TO 10
Synchronous output:

DIMENSION IBF(l004) ,IERR(2) ,INTVL(2)

C FIRST BUFFER OF DATA MUST BE AVAILABLE AT START

C THIS EXAMPLE ASSUMES FIRST BUFFER IS FULL AT START

C START SYNCHRONOUS DIGITAL OUTPUT FUNCTION

c
c

INTVL (1) =2
INTVL(2)=5
CALL SDO(IBF,1004,160,INTVL,IEFN,MASK,IERR,50,16,15)

C INITIALIZE HALF BUFFER INDEX

12-34

LABORATORY PERIPHERAL SYSTEMS DRIVERS

c
INDX=4

C WAITFOR ROOM IN BUFFER

CALL WAITFR(IEFN)

c
C CLEAR EVENT FLAG

c
15

CALL CLREF(IEFN)

c
C CALCULATE VALUES TO PUT IN BUFFER

X=(Y+2)*Z
DO 20 I=l,500
IBF(I+INDX)=X**5/A
CONTINUE

C SIGNIFY ANOTHER HALF BUFFER IS FULL

CALL ADJLPS(IBF,500)

c
C ADJUST BUFFER INDEX

INDX=INDX+5 0 0
IF(INDX.GE.1004} INDX=4

C CHECK IF ANOTHER HALF BUFFER IS EMPTY

IF(IERR(2) .GE.500) GO TO 15
IF(IERR(l) .NE.0) GO TO end of sampling
GO TO 10
NOTE
In both the examples above, care is
taken to ensure that the program stay
"in sync" with the driver.
If the
program
"loses"
its
position with
respect to the driver the function must
be stopped and restarted since this is
the only way to recover. Caution should
be exercised to ensure that the program
sequence above be used
to avoid
a
possible loss of data.

12-35

CHAPTER 13
PAPER TAPE READER/PUNCH DRIVERS

13.l

INTRCDUCTION

The RSX-llM paper tape reader/punch drivers support the PCll paper
tape reader/punch and the PRll paper tape reader. The PCll is a
high-speed reader/punch capable of reading eight-hole,
unoiled,
perforated paper tape at 300 characters per second, and punching tape
at 50 characters per second. The PRll has the same characteristics as
the paper tape reader portion of the PCll. All transfers are image
mode only, with no interpretation of data.

13~2

GET LUN INFORMATION MACRO

Word 2 of the buffer filled by the GET LUN INFORMATION system
directive (the first characteristics word) contains the following
information for paper tape devices. A bit setting of 1 indicates that
the described characteristic is true for these devices.
Bit

Setting

Meaning:

Record-oriented device

Carriage-control device

Terminal device

Directory device

Single-directory device

Sequential device

6-12

Reserved

Device mountable as a communications channel

Device mountable as a FILES-11 volume

Device mountable

Words 3 and 4 of the buffer are undefined;
word 5 indicates
default buffer size, which is 64 bytes for paper tape devices.

13-1

the

PAPER TAPE READER/PUNCH DRIVERS
13.3

QIO MACRO

Table 13-1 lists the standard functions of
valid for the paper tape reader/punch.

the

QIO

macro

that

are

Table 13-1
Standard QIO Functions for the Paper Tape Reader/Punch
Format

Function

QI0$C
QI0$C
QI0$C
QIO$C
QI0$C
QIO$C
QIO$C

IO.ATT, •••
IO.DET, ••.
IO.KIL, •..
IO.RLB, .•. ,<stadd,size>
IO.RVB, .•• ,<stadd,size>
IO.WLB, ..• ,<stadd,size>
IO.WVB, .•. ,<stadd,size>

where:

stadd

is the starting address of the data buffer {may be
on a byte boundary)

size

is the data buffer size in bytes {must be
than zero)

Attach device
Detach device
Cancel I/O requests
Read logical block (reader only)
Read virtual block (reader only)
Write logical block (punch only)
Write virtual block (punch only)

greater

IO.KIL never cancels an in-progress read request.
In-progress write
requests are cancelled only when the punch driver is waiting for the
punch to become ready at the start of a transfer.
The paper tape drivers support no device-specific functions.

13.4

STATUS RETURNS

Table 13-2 lists error and status conditions that are returned by
paper tape reader/punch drivers.

the

Table 13-2
Paper Tape Reader/Punch Status Returns
Code

Reason

rs.sue

Successful completion.
The operation specified in the QIO directive was completed I
The second word of the I/O status block can be
examined to determine the number of bytes processed,
if the
operation involved reading or writing.

I successfully.
IS.PND

I/O request pending.
The operation specified in the QIO directive has not yet been I
executed. The I/O status block is filled with zeros.

IE.ABO

Operation aborted.
The I/O request was cancelled
still in the I/O queue.

13-2

while

progress

while

PAPER TAPE READER/PUNCH DRIVERS
Table 13-2 (Cont.)
Paper Tape Reader/Punch Status Returns
Code
IE.DAA

i Reason

Device already attached.
The physical device unit specified in an IO.ATT function was
already attached by the issuing task. This code indicates
that the issuing task has already attached the desired
physical device unit, not that the unit was attached by
another task.

IE. DNA

Device not attached.
The physical device unit specified in an IO.DET function was
not attached by the issuing task. This code has no bearing
on the attachment status of other tasks.

IE. DNR

Device not ready.
The reader and punch drivers return this code when a time-out
occurs.
The reader driver also returns this code when an
error condition (see Section 13.4.1)
is encountered before
the initiation of the first transfer after an ATTACH command
has been issued.

IE. EOF

I End-of-file encountered.
The reader driver encountered an error condition (see Section II
13.4.1) at a time other than the initiation of the first read
after a valid ATTACH command. The second word of the I/O I
status buffer contains a count of bytes successfully read !
before the error condition was encountered.

IE. IFC

Illegal function.
An illegal function code was specified in an I/O request that
is not legal for the respective paper tape drivers.

IE. OFL

Device

The physical device unit associated with the LUN specified in
the QIO directive was not on-line.
When the system was
booted, a device check indicated that this physical device
unit was not in the configuration.
IE. SPC

Illegal address space.
The buffer specified for a read or write request was
partially or totally outside the address space of the issuing
task. Alternatively, a byte count of zero was specified.

IE. VER

Unrecoverable hardware error {punch only).
The punch driver encountered an error condition (see Section
13.4.1) at a time other than the initiation of a transfer.
Section 13.4.2 describes the action of the punch driver when
an error condition is encountered upon the initiation of a
transfer.

13-3

PAPER TAPE READER/PUNCH DRIVERS
13.4.1

Error Conditions

There are four error conditions that are
paper tape drivers. These conditions are:

the

When the punch driver encounters an error condition upon
initiation of a transfer, the following message is displayed:

the

No tape

Read~r

Power low

Hardware malfunction

13.4.2

indistinguishable

off-line

Ready Recovery

*** PPn:

-- NOT READY

where n is the unit number of the paper tape punch that is not ready.
This message is repeated every 15 seconds until the error condition is
corrected, or until the I/O request is cancelled.
When the error
condition has been corrected, the transfer will begin within one
second.

13.5

PROGRAMMING HINTS

This
section
contains
information
on
important
programming
considerations relevant to users of the paper tape drivers described
in this chapter.

13.5.1

Special Action Resulting from Attach and Detach

When an Attach or Detach is issued to the punch, the punch driver
initiates a transfer of 170 (decimal) nulls. Upon the first read
after an attach to the reader, all nulls preceding the first non-null
character on the tape are read and discarded by the reader driver.

13.5.2

Reading Past End-of-Tape

When the reader driver reads past the physical end-of-tape, it
normally generates at least two incorrect data bytes. These bytes are
included in the byte count returned by the driver. User software that
does not prevent reads past the physical end-of-tape should discard at
least the last six characters in the buffer when IE.EOF is returned by
the driver.

13-4

CHAPTER 14
INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

14.1

INTRODUCTION

ICSll and ICRll are local, and remote process I/O
subsystems
respectively.
They operate under program control as devices capable
of interrogating digital and analog input, and driving digital and
analog output.

"
14.1.1

Hardware Configuration

A single l.Li::> or l.LK control.l.er can handle up to 16 I/0 modules in duy
configuration;
a module contains 16 bits of input or output data,
providing a total of 256 digital points. Up to 12 ICR or res units
are supported.
The ICS/ICR driver is tailored to the user's needs,
interactively, through the SYSGEN
(System
Generation
program)
dialogue. The driver is capable of handling any combination of !CR or
ICS controllers installed on a single system.

14.1.1.1 Address Assignments - Each ICRllA Unibus interface or ICSll
file oox must be configured for individually addressable interrupt
vectors, Control and Status Registers (!CSR), and module Address
Registers (!CAR) as shown below.
Table 14-1
I CS/ I CR Address Assignments
MODULE ADDRESSES

I CSR/ I CAR ADDR. INTERRUPT VECT.

171000-171036

171770-171776

234-236

171040-171076

171760-171766

XXX-XXX+2

171100-171136

171750-171756

XXX+4-XXX+6

171140-171176

171740-171746

XXX+l0-XXX+l2

171200-171236

171730-171736

XXX+l4-XXX+l6

ICS/ICR UNIT NO.

14-1

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-1 {Cont.)
ICS/ICR Address Assignments
ICS/ICR UNIT NO.

MODULE ADDRESSES

I CSR/ I CAR ADDR. INTERRUPT VECT.

171240-171276

171720-171726

XXX+20-XXX+22

171300-171336

171710-171716

XXX+24-XXX+26

171340-171376

171700-171706

XXX+30-XXX+32

171400-171436

171670-171676

XXX+34-XXX+36

171440-171476

171660-171666

XXX+40-XXX+42

171500-171536

171650-171656

XXX+44-XXX+46

171540-171576

171640-171646

XXX+50-XXX+52

NOTES
nnnnn6
nnnnn4

= Control and Status Register
= Address Register

Additional controllers
are
vector addresses above 300.

assigned

14.1.1.2 Supported I/O Modules - The following modules, all optional,
are supported by the ICS/ICR driver.
D/A Converters
IDA-OA -

4-channel digital-to-analog converter.

A/D Converters
IAD-IA IMX-IA -

analog-to-digital
8-channel wide-range differential
converter.
16-channel flying capacitor relay multiplexer.

Counters
IDC-IC -

16-bit binary counter.

Bistable Digital Outputs
IDC-OA IAC-OA IRL-OA IRL-OB -

D/C flip/flop driver.
A/C flip/flop driver.
Latching relay output.
Flip/flop relay output.

Momentary Digital Output
IDC-OB IAC-OB -

D/C single-shot driver.
A/C single-shot driver.

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Digital Inputs (Noninterrupting)*
IDC-IA IDC-ID
IAC-IA -

D/C voltage sense input.
D/C voltage input module.
A/C voltage input module.

Digital Inputs (Interrupting)
IDC-IB IAC-IB -

D/C voltage interrupt input.
A/C voltage interrupt input.

Terminal Input/Output
110 CPS Remote Terminal Interface to ICRll.

14.1.2

Alternate ICSll Support

The ICSll Industrial Control Subsystem is supported either by the
UDCll or ICS/ICR-11 device driver. If the system does not have an
ICRll controller, and if a driver of minimum size is required, then
UDCll support should be considered. The hardware requirements for
such support are as follows:
1.

Each file box must be assigned to the same
address (normally 234).

interrupt

vector

The control and status register within each file box must
appear at the same address within the I/O page (normally
171776).

If support of the IAD-IA A/D

converter

required,

the

following

module addressing and installation conventions are imposed:
1.

Each IAD-IA converter and its associated IMX-IA
relay
multiplexer is assigned a fixed block of 120 logical channel
numbers. No more than 32 IAD-IA converters may be insta11ed
in a single system. Based on this convention, A/D converter
0 occupies channels 0-119, A/D converter 1 occupies 120-239
etc.

Regardless of the actual number of IMX-IA multiplexers
preempts a block of eight
installed,
each
converter
contiguous module slots.

The slots reserved for all A/D converters and
must occupy a block of contiguous module slots.

multiplexers

If necessary, the vector and address changes can be made by Field
Service personnel.
Assuming the hardware configuration is correct,
the user can implement the desired UDCll software support by answering
all SYSGEN questions relating to the UDCll in the affirmative.
If the additional ICS/ICR-11 driver features are required (at a
commensurate increase in the memory requirements), then each ICSll
*Note that noninterrupting input modules are accessed directly by a
task.
Hence, while FORTRAN interface routines are available, no
support for such modules is included in the driver.

14-3

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
file box must be configured for individually addressable interrupt
vectors and control status registers. This change can be performed by
Field Service personnel.
The
necessary
software
support
is
incorporated by answering all SYSGEN questions relating to the
ICS/ICR-11 in the affirmative.
The additional ICSll capabilities provided by
may be summarized as follows:
1.

Multi-controller, parallel operation.

Increased A/D conversion throughput.

Activation of
counters.

No requirement to
contiguous slots.

directly

driver

interrupts

same

type

Section 14.7 summarizes the software differences between the
ICS/ICR drivers in detail.

UDC

and

install

from

ICS/ICR-11

digital

14.1.3

tasks

the

modules

the

Software Support

ICS/ICR operations are divided into two categories:
1.

Functions performed directly by any task.

Functions requiring driver services.

Direct functions are accomplished through memory references to the
ICS/ICR registers on the I/O page. In a protected system any task may
gain restricted access to the device registers by linking to a global
common block that resides within the appropriate physical memory
limits. Direct functions consist of:
1.

Reading counter modules.

Reading any digital input module.

Driver requests are divided into the following categories:
1.

Noninterrupting output functions
a.

Bistable (flip/flop) digital output

Analog output

Momentary (single-shot) digital output

Requests for interrupting functions
a.

Analog input

Remote terminal output

14-4

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
3.

Requests for unsolicited interrupts
a.

Digital interrupts

Counter interrupts

Remote terminal input

Remote unit or serial line errors

With the exception of A/D input and remote terminal
functions are complete upon return to the user's task.

output,

all

Under RSX-llM noninterrupting output functions
are
immediately
submitted to the controller through a circular buffer that is filled
at driver level and emptied at interrupt level. A QIO is considered
successfully completed when the request is inserted in the circular
buffer.
The following operations are in this category:
1.

Bistable digital outputs

Analog outputs

Momentary digital outputs

Interrupting functions are those operations that generate an interrupt
within some fixed time after initiation. The driver allows a list of
multiple transactions to be specified in a single QIO.
Each
transaction is initiated in sequence without waiting for the preceding
interrupt, until either the list is exhausted or all modules of the
specified type are active.
The following operations are in this
category:
1.

A/D inputs

Remote terminal output

Unsolicited interrupts may require no initiation by the processor and
occur at indeterminate intervals. The following functions are in this
category:
1.

Interrupting digital inputs

Counter modules

Remote terminal input

Error interrupts

All unsolicited interrupt data, except for errors, may be placed in a
task-provided circular buffer. On interrupt, an event flag specified
by the task is set. s~ch data for each module type is supplied to
only one task per controller. In addition, the driver will activate
selected tasks on the occurrence of digital or terminal input
interrupts.
Error interrupts are described further on in this chapter.

14-5

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Terminal support is restricted to passing terminal data between the
device and a task.
The only special character is Control-C (003),
which may cause a user-specified task to be made ac~ive. There is no
other special processing for terminal I/0 except that the parity bit
is removed. This is similar to the terminal driver function of
IO.RAL.
1.

MCR is not invoked.

Characters are not echoed.

Carriage control is not performed.

TABs, RUBOUTs, etc.

Line terminators are not recognized.

Fill characters are not generated.

14.1.4

are not recognized.

UDCll Software Compatibility

Many of the MACRO and FORTRAN interfaces described in the following
paragraphs are fully compatible with existing UDCll applications
software; however, the user should consult section 14.7 for a summary
of differences that do exist between UDC and ICS/ICR software.

14.2

LUN INFORMATION

A request for logical unit information returns
the
device-dependent data in words 2 through 5 of the buffer:

14.3

WD 02

WD 03

undefined

WD 04

undefined

WD 05

following

ASSEMBLY LANGUAGE INTERFACE

Table 14-2 summarizes standard
supported by the ICS/ICR driver.

and

device-specific

QIO

functions

Table 14-2
Summary of ICS/ICR-11 QIO Functions
QIO$C IO.CCI, ... <stadd,sizb,tevf>

Connect
buffer
a
digital interrupts

QIO$C IO.CTI, •.. <stadd,sizb,tevf,arv>

Connect
buffer
a
counter interrupts

QIO$C IO.CTY, ... <stadd,sizb,tevf>

buffer
Connect
a
terminal interrupts

14-6

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-2 (Cont.)
Summary of ICS/ICR-11 QIO Functions

I QI0$C IO.DC!, ...

Disconnect a buffer
digital interrupts

from

QI0$C IO. DTI, ...

Disconnect a buffer
counter interrupts

from

QIO$C IO. DTY, ...

Disconnect a buffer
terminal interrupts

from

QI0$C IO.FLN, .•.

Set

QI0$C IO.ITI, ... <mn,ic>

Initialize

QI0$C IO. LDI, ... < tname, [ tevf] ,pn ,csm>

Link task
interrupts

QI0$C IO.LKE, ... <tname,,[tevf]>

Link
task
interrupts

error

QIO$C IO.LT!, ..• <tname,, [tevf] ,en, [arv]>

Link task
interrupts

counter

QIO$C IO.LTY, ... <tname,, [tevf]>

Link
task
to
remote
terminal interrupts

QIO$C IO.MLO, ... <opn,pp,dp>

Open or close bistable
digital output points

QIO$C IO.MSO, ••• <opn,dp>

Pulse single-shot digital
output points

QIO$C IO.NLK, .•. <tname>

Unlink a task
interrupts

from

QIO$C IO. NLN, ...

Place ICS/ICR
online

controller I

QI0$C IO.RAD, •.• <stadd>

Read activating data

QI0$C IO.RBC, ... <stadd,size,stcnta>

multiple
Initiate
conversions

~I0$C

Perform analog output

IO.SAO, ..• <chn,vout>

offline

controller

counter

digital

all

A/D

QI0$C IO.UDI, •.• <tname>

Unlink
a
task
digital interrupts

from

QI0$C IO.UER, ... <tname>

Unlink a task from
interrupts

error

QIO$C IO.UTI, ... <tname>

Unlink
a
task
counter interrupts

from

QIO$C IO.UTY, ... <tname>

Unlink
a
task
from
terminal interrupts.

QIO$C IO.WLB, ... <stadd,sizb>

Transmit data to the
Remote terminal

14-7

!CR

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Where:
arv

is the starting address of a buffer containing initial
or reset counter values. The buffer must be aligned on
a word boundary.

chn

is the D/A channel number

is the counter number

csm

is the change-of-state mask

is the binary data pattern

is the initial count

is the module number

opn

is the first latching digital output point
This value must be on a module boundary
divisible by 16}

sizb

is the data buffer size in bytes.
For a circular
buffer connected to unsolicited interrupts, this value
must be even and large enough to include one entry plus
the 2-word header

size

is the data and control buffer size in bytes.
This
value must be an even number that is greater than zero.

stadd

is the starting address of the data.buffer (must be
a word boundary}

staddb

is the starting address of the terminal output
May be aligned on a byte boundary

stcnta

is the starting address of the control buffer (must be
on a word boundary}; each control buffer word must be
constructed as described in paragraph 14.3.2

tevf

is an event flag number in the range 1 to 64

tname

is a 2-word task name composed of 1 to
characters in RADIX-50 format

vout

is a binary number between 0 and 1023.
converted to an analog output

that is

number.
(evenly

buffer.

alphanumeric
to

The following sections contain a detailed description of
each
function.
In the discussion of QIO request parameters, the following
conventions apply.
All numbering is relative.
Module numbers start at 0 beginning with the first module of a given
type. Increasing module numbers correspond to increasing physical bus
addresses.
Channel numbers start at 0, with channel 0 as the first channel on the
first module of a given type.

14-8

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Point numbers start at 0 with point 0 as the first point on the first
module of a given type. Points within a module are numbered "from
right to left" in increasing order.
It should be remembered that there is no requirement for modules of a
given type to occupy contiguous slots; thus, for example, digital
points 15(10} and 16(10} need not reside on physically adjacent
modules.
It is assumed that the number of points or channels per module is a
constant for each generic type. Specifically the following weights
apply:
1.

Each Digital I/O Module contains 16 points.

Each Counter Module contains 1 channel.

Each D/A Module contains 4 channels.

Each A/D Converter contains 120 channels.

As stated above, an A/D converter is assigned a block of 120 channels.
The number of channels in use within the block depends on the number
of multiplexers installed. The driver will reject an attempt to
address a nonexistent channel.
The table below illustrates the relationship between physical slot
numbers,
bus
addresses
and
relative
addresses for a given
configuration. It is referred to in the following discussion.
As noted, a block of 120 relative addresses is reserved for each A/D
converter.
The converter and multiplexer in slots 10 and 11 contain
channels 0 through 24. The converter in slot 17 contains channels 120
through 127.
An attempt to access a nonexistent channel (e.g.,
channel 30 or channel 129} will be rejected by the driver.
The user should observe that the flip-flop drivers in slots 13 and 15
contain relative point numbers 0 through 15, and 16 through 29
In general, the
although the modules are not physically adjacent.
relationship between slot number, module type, bus address, and
relative address is as follows:
1.

A set of contiguous relative addresses is defined for each
module of a given type that is installed. Each relative
address, when qualified by type, uniquely identifies a
digital point or channel.

A set of slot numbers and bus addresses, possibly not
contiguous, is occupied by all modules of a given type. Such
addresses may be assigned solely on the basis of hardware and
installation considerations.
Increasing relative addresses
correspond to increasing bus addresses.

14-9

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-3
Sample ICS/ICR Configuration
Unit:

Slot Number
9.
10.
11.
12.
13.
14.
15.
16.
17.

14.3.1

Type

Bus Address

Relative Addresses

171000
171002

0
0-119.

171006
171010
171012
171014
171016
171020

1.
0-15.
0-3
16.-31.
4-7.
120.-239.

Counter
A/D Converter
A/D Multiplexer
Counter
Flip Flop driver
D/A Converter
Flip Flop driver
D/A Converter
A/D Converter

------

General Error Status Returns

The following error status returns apply uniformly to all requests:
IE.ABO - Operation aborted. The specified operation was cancelled via
IO.KIL or the request timed out while the unit was offline.
IE.OFL - Controller offline. The physical device unit associated with
the LUN specified in the QIO directive was not online. An
ICS/ICR controller may be offline because a device check
during bootstrap load has indicated that the controller is
not in the configuration.
IE.DNR - Controller not ready.
been detected.

A nonrecoverable controller error

has

IE.IFC - Illegal function. A function code was included in an I/O
request that is illegal for the ICS/ICR. The function may
also refer to an ICS/ICR module type or function that was not
specified during system generation.

14.3.2

A/D Input - Read Multiple A/D Channels

This function provides the capability of reading several A/D channels
at any permissible gain. The driver is capable of initiating parallel
transfers when more than one A/D converter is installed in a file box;
however, only one interrupt module request (remote terminal or A/D)
may be in progress at a given time.
QIO DPB format:
QIO$C IO.RBC, ... <stadd,size,stcnta>
where:
stadd

the starting address of the data buffer {must be
word boundary) .

14-10

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
size

the data buffer size in bytes (must be even and greater
than zero);
the control buffer is the same size.

stcnta

the starting aaaress of the control buffer (must be on
a word boundary);
each control buffer word must be
constructed as shown in Table 14-6.

Return Status:

rs.sue - Function successfully completed.
IE.BAD - Illegal channel or gain code specified.
IE.BYT - Data buffer is byte aligned. Alternatively, the
of the buffer is not an even number of bytes.

length

IE.DNR - Device not
occurred.

interrupt

timeout

ready.

A/D

converter

Note that the second I/O Status word contains
number of conversions successfully completed.

count

the

One control word is paired with each data word.
That is, the data
appearing in a data array element is obtained using the gain and
channel number specified in the corresponding element of the control
array.
Control words specify the gain and channel in the following
format:
Bits

Meaning

0-11

Channel Number range: 0-1919

12-15

Gain value for
this sample. The
binary value is
as follows:

GAIN

IAD-IA

0
0

0
0
0
0

0
0

1
2

0
0
0

0
0
0
1
1
1
1
1
1
1
1

1
1
1
1
0
0
0
0

1
1
1
1

1
1
0
0
1
1
0
0

1
1
0
0
1
1

1
0
1
0
1
0
1
0
1
0
1

ILLEGAL
ILLEGAL
10

20
ILLEGAL
ILLEGAL
50
100
ILLEGAL
ILLEGAL
200

1000

ILLEGAL
ILLEGAL

Upon receipt and validation of the parameters within the I/O
the driver will initiate the following sampling procedure:

14-11

packet,

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
1.

The control word is fetched and tested for validity (i.e.,
for legal gain and channel). If an error is encountered or
no further control words remain, processing is terminated as
described in Step 4.

Assuming the A/D converter board is idle, the driver starts
the conversion, sets this resource busy and returns to Step
1. If the converter is busy the driver returns control to
the system after saving the data required to initiate the
conversion when the channel becomes idle.

On the occurrence of an A/D interrupt, the interrupt service
routine initiates the appropriate processing at the noninterrupt level that will either set the channel idle or
initiate a previous request stored during Step 2. The
occurrence of the latter results in processing of additional
control words as described in Step 1.

A/D requests
conditions:

are

terminated

under

any

the

All control words have been processed.

A hardware error has occurred.

An error in a control word has been detected.

following

Regardless of the cause, the driver cannot complete request processing
until all pending A/D transfers have gone to completion.
Because of overlapped processing, multiple errors can occur (e.g., a
hardware error and an erroneous control word). The driver will return
the status associated with the earliest transaction that caused an
error condition. Thus, at the user interface, the driver will appear
to execute all conversions sequentially.

14.3.3

Analog Output

This function provides the capability
output channel to a specified voltage.

setting

single

analog

QIO DPB format:
QIO$C IO.SAO, ... <chn,vout>
where:
chn

the output channel number

vout

the output voltage representation

Output voltage varies linearly with the binary input to the channel,
where 0 to plus ten volts (+lOv.) is represented by integers from 0 to
1023.

14-12

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Return Status:
rs.sue - Function submitted for output to controller.
IE.MOD - Nonexistent D/A channel was specified.
The second I/O status word is zero.

14.3.4

Single-Shot Digital Output - Multi-Point

This function provides the capability of pulsing a field of up to 16
momentary digital output points.
Fields must be aligned on module
boundaries.
QIO DPB format:
QIO$e IO.MSO, ... <opn,dp>
where:
opn

the starting digital output point number. Point number
must be aligned on a module boundary (i.e., must be a
multiple of 16).

the 16-bit mask. One point is pulsed corresponding
each bit set in the mask word.

Return Status:
rs.sue - Function submitted for output to the controller.
IE.MOD - Invalid starting point number specified.
Point
nonexistent or not aligned on a module boundary.

14.3.5

Bistable Digital Output - Multi-Point

This function provides the capability of setting or resetting a field
of up to 16 bistable digital output points. Fields must be aligned on
a module boundary.
QIO DPB format:
QIO$e IO.MLO, •.. <opn,pp,dp>
where:
opn

the starting digital output point number.

the 16-bit mask.

the data pattern.

A bit is set in the mask word for each point that may change state.
The state of points corresponding to reset mask bits is unaltered.
When the mask bit is set, the output will be "closed" if the data
is set and "open" if the data bit is clear.

14-13

bit

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Return Status:
IS.sue - Function submitted for output to the controller.
IE.MOD - Invalid starting point number specified. Point does not
exist or is not aligned on a module boundary.

14.3.6

Unsolicited Interrupt Processing

Unsolicited interrupts consist of the following:
1.

Digital interrupts

Counter interrupts

Remote terminal input

Hardware Errors

Based on the type of interrupt, the driver may
interrupt data in one or more of the following ways:

dispose

the

The data may be furnished to a task that has issued a request
to continually monitor such information.

A task may be activated by a specific input.
That is, a
dormant task can be requested to run, or an event flag may be
set if the task is currently active.

The driver will allow continual monitoring for digital, counter, and
terminal inputs with the provision that, for each controller, only one
task per module type may receive such inputs.
Task activation is permitted for digital, terminal, and
error
interrupts. The processing related to hardware errors is discussed in
section 14.5. Activation of tasks by digital counter and terminal
inputs is covered in section 14.3.7.
The driver functions described in the following paragraphs allow a
task to continually receive interrupt data. To monitor such data a
task must provide:
1.

A buffer that is filled by the driver and emptied by the task
in circular fashion.

An event flag that will be set upon the
interrupt.

The driver will connect a single task
interrupts from a specific module type.

14-14

per

occurrence

controller

each

receive

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
The buffer to be connected has the format shown below:
FORTRAN

Index

Contents

driver index

user index

entry

.
The buffer consists of a two-word header containing the driver and
user index, as shown, followed by a data area that is subdivided into
fixed length entries. Each entry consists of an entry existence
indicator followed by one or more words of device-dependent data.
Such information usually consists of module data, relative module
number, and a code identifying a module type. On the occurrence of an
interrupt, the driver enters data in the location currently indicated
by the driver index. This index can be considered as a FORTRAN index
into the buffer.
That
the first location in the buffer
is
"'°' ,
associated with the index 1.
The beginning of the data area is
associated with the first entry, index 3.
Entries are made in a
circular fashion starting at the beginning of the data area, filling
in order of increasing memory address, and wrapping around to the
beginning of the data area when there is insufficient space for an
entry at the end. Note that the size of the data area must be an
integer multiple of the entry size.
1

J..

It is expected that the connected task will maintain the user index,
ensuring that it indicate where, in the buffer, the task is to process
interrupt data next.
When the task is activated by the driver, it should process data in
the buffer starting at the location indicated by its pointer, and
continuing in circular fashion until an existence indicator is
encountered that is zero.
The existence indicator is set to +l when a buffer entry is made.
Except to record a hardware error, the contents of an entry are not
altered by the driver if the indicator is nonzero.
Hence, when a
requester has removed or processed the entry, he must clear the
existence indicator in order to free the buffer entry position.
If
the driver detects a nonzero indicator, i.e., data input has occurred
in a burst sufficient to overrun the buffer, the data is discarded and
a count of data overruns is incremented. The count is maintained in
the entry existence indicator which, as noted above, is set to +l to
indicate no overruns between entries, +2 to indicate a hardware error
entry, or a negative value recording the two's complement of the
number of times data has been discarded between entries. The overrun
count will never be allowed to wrap around to a positive value.
In the event of a nonrecoverable controller error {remote unit
power-fail or hard data error) all connected tasks are activated with
the following entry in the circular buffer:

14-15

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

WD 00

Hardware error indicator (+2)

WD nn
WD nn+l
WD nn+2

Contents of !CSR register
Physical unit number
Generic code indicator
set to 177770(8)

= offset to module data word.

This entry is always placed
status.

the

buffer

regardless

overflow

The error flags are obtained from the controller !CSR word at the time
the error was detected (see Table 14-7).

14.3.6.1 Connect to Digital Interrupts - This
single task to receive digital interrupt data.

function

allows

connected

(must

QIO DPB format:
QIO IO.CCI, .•. <stadd,sizb,tevf>
where:
stadd

starting address of buffer to
word aligned)

sizb

length of buffer in bytes
buffer length is 14 bytes

tevf

trigger event flag number

(must

even).

Minimum

Status:

rs.sue - Function successfully completed.

Second I/O status word
contains the number of words passed per interrupt in the
low byte, and the initial FORTRAN index in the high
byte.

IE.BYT - Buffer address is byte
number of bytes.

aligned

length

odd

IE.CON - Interrupt already connected to another task.
IE.IEF - Illegal event flag was specified i.e., not in the
1-64.

range

IE.PR! - Task checkpointable and not fixed in memory.
IE.SPC - Interrupt circular buffer was not wholly within the
address space of the task. Alternatively, the buffer
was too small for a single data entry (7 words minimum).

14-16

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Entry Format:
WD
WD
WD
WD
WD

00 01 02 03 04 -

Existence Indicator
Change of state indicator
Module data
Relative module number
Generic Code 1, 2 or 3, indicating a digital interrupt

The contents
previously.

the

existence

indicator

have

been

described

The change-of-state indicator records those bits for which a change of
state in the direction of interest has been detected. The direction
of the change may be from 0 to 1 (point closed (PCL)) or 1 to 0 (point
open (POP)) depending upon the PCL or POP jumper connections on the
digital interrupt module. The driver will assume that at least one of
these signals is always asserted.
The relative module number indicates the module on which the change of
state was recognized.
The module data word records data received at the time
was serviced.

the

interrupt

The generic code identifies the type of module that caused the
interrupt. A digital interrupting module may have the value 1, 2 or 3
as selected by user-installed jumpers on the module.

14.3.6.2 Disconnect from Digital Interrupts - This function allows
task to terminate the processing of digital interrupt data.

QIO DPB format:
QIO$C IO.DCI, ...
Return Status:

IS.SOC - Function successfully completed.
zero.
IE.CON - Task was not connected.

14.3.6.3
a single

Second I/O status

word

Second I/O status word is zero.

Connect to Counter Module Interrupts - This function
receive counter interrupt dat~.

allows

t~sk-to

QIO DPB format:
QIO$C IO.CTI, ... <stadd,sizb,tevf,arv>
where:
stadd

starting address
aligned)

sizb

length of buffer in bytes
buffer length is 12 bytes.

tevf

trigger event flag number

arv

starting address of table
(wust be word aligned)

14-17

circular

buffer

(must

initial

(must

word

even).

Minimum

counter

values

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Word 03 defines an array of initial counter values.
One entry is
required for each counter installed in a physical unit. Entries are
paired with modules in logically ascending sequence. The counter is
set to the initial value upon receipt of the connect function and
whenever an overflow interrupt occurs (i.e., when the count reaches
zero).
Return Status:

rs.sue - Function successfully completed.

The second I/O status
word contains the number of words passed per interrupt
in the low byte, and the initial FORTRAN index in the
high byte.

IE.BYT - Buffer address is byte
number of bytes.

aligned

length

odd

IE.CON - Interrupt already connected to another task.
IE.IEF - Illegal event flag was specified, i.e., not in the range
1-64.
IE.PR! - Task checkpointable and not fixed in memory.
IE.SPC - Interrupt circular buffer or table of initial values was
not wholely within the address space of the task.
Alternately, the buffer was too small for a single data
entry (6 words minimum).
Entry Format:
WD 00 - Existence indicator
WD 01 - Module data
WD 02 - Relative module number
WD 03 - Generic code (4, 5 or 6)

14.3.6.4 Set Counter Initial Value - This function allows a counter.
initial value to be established.
A task need not be connected to
counter interrupts to perform this function.
QIO DPB format:
QIO$C IO.ITI, •.. <mn,ic>
where:
mn

relative module number

new initial count

Return Status:

rs.sue - New value submitted for output to the

controller.
second word of I/O status is set to zero.

IE.MOD - Nonexistent module number specified.

14-18

The

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Upon receipt of the request, the new initial value is immediately
queued
for
output
to
the controller.
The counter will be
reinitialized with this value on overflow if a task is connected to
counter interrupts.

14.3.6.5 Disconnect from Counter Interrupts - This function allows
task to terminate counter interrupt processing.

QIO DPB format:
QIO$C IO. DTI, ...
Return Status:

rs.sue - Function successfully completed.
status is set to zero.
IE.CON - Task was not connected to timer interrupts.
After disconnect is complete, counters will
initial value at the time of the interrupt.

not

reset

the

14.3.6.6 Connect to Terminal Interrupts - This function allows a task
to receive terminal inputs from the selected ICRll controller.
QIO DPB format:
QIO$C IO.CTY, .•• <stadd,sizb,tevf>
where:
stadd

address of circular buffer (must be word aligned)

sizb

length of buffer (must be even).
is 12 bytes.
'--.:---\,..L J.'j'jt:L

-----'t:Vt:Ul.

.C1--

J...Lal:::j

Minimum buffer length

-1-.--

- · ..
11UlllUl:L

Return Status:

rs.sue - Function successfully completed.

The second I/O status
word contains the number of words passed per interrupt
in the low byte, and the initial FORTRAN index in the
high byte.

IE.BYT - Buffer is byte aligned or length is
bytes

odd

number

IE.CON - Interrupt already connected to another task.
IE.IEF - Illegal event flag was specified, i.e., not in the range
1 to 64.
IE.MOD - Nonexistent device.

Controller is ICSll.

IE.SPC - Interrupt circular buffer was not wholly within the
address space of the task. Alternatively, the buffer
was too small for a single entry (6 words minimum).

14-19

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Entry Format:
WD 00 - Existence indicator
WD 01 - High byte = 0, low byte = terminal input character
WD 02 - Relative module number (normally 0)
WD 03 - Generic code indicator (normally 0)
Note that words 2 and 3 are nonzero only when the entry
the result of a nonrecoverable controller error.

was

made

All remote terminal data is conveyed to the requesting task as
but with the parity bit removed.

input,

NOTE
Remote terminal input is not
the driver.

echoed

Disconnect from Terminal Input - This function allows a task
to discontinue the processing of terminal input.

14.3.6.7

QIO DPB format:
QIO$C IO.DTY, ••.
Return Status:
IS.SOC - Function successfully completed.
status is set to zero.

The second word of I/O

IE.CON - Task was not connected to remote terminal interrupts.

14.3.7

Activating a Task by Unsolicited Interrupts

The functions described
capability of:

the

following

paragraphs

provide

Activating a task in response to unsolicited interrupts.

Interrogating
activation.

Removing a task from the activation list.

the

driver

determine

the

reason

the

for

The QIO DPB parameters specify the task name, an optional trigger
event flag to be set if the task is active, and device-dependent
parameters that identify the interrupt source. A task is linked to
interrupts (i.e., made eligible for activation) provided that:
1.

the resource exists,

the task is installed, and

no other task is linked to the resource.

14-20

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
If another task is linked to the resource, the driver will reject the
request with a status of resource-in-use (IE.RSU). A resource is
defined as a single interrupt point, remote terminal (Control-C input
only), or counter module.
On the occurrence of the appropriate interrupt, the task is made
active if dormant; otherwise, a trigger event flag, if specified, is
set. The task may interrogate the driver to determine the conditions
that caused activation, and to signify interrupt recognition. The
function of the event flag is to allow such a task to recognize an
event that has occurred while the task was active. Recognition is
ensured prior to the completion of task execution by issuing the Exit
If system directive followed by the Clear Event Flag directive.
The linkage between a task and a specific interrupt is removed
issuing the appropriate unlink request via the QIO directive.

Only one task may be associated with each interrupt source (i.e., one
task per digital interrupt point, terminal input, or counter module.
NOTE
The MCR command REMOVE automatically
unlinks a task from all interrupts.

14.3.7.1 Link a Task to Digital Interrupts - The following function
allows a task to be activated on the occurrence of digital interrupts.
QIO DPB format:
QIO$C IO.LOI, ••• <tname,, [tevf] ,pn,csm>
where:

tname

a 1- to 6-character alphanumeric task name
Radix-SO format

tevf

trigger event flag (O=none)

point number (must be aligned on a module boundary)

csm

change-of-state mask

2-word,

The change-of-state mask indicates those bits for which a change of
state in the direction specified by the PCL and POP jumpers causes the
task to be activated.
Only one task may be linked to a given
interrupt point. A zero change of state mask is not permitted.
Return Status

rs. sue

Function successfully completed.
I/O status is set to zero.

IE. BAD

Change-of-state mask set to zero.

IE. IEF

Trigger event flag not in the range 0-64.

14-21

The second

word

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
IE .MOD

Nonexistent module or point.
boundary.

Not aligned on a

IE .NOD

Insufficient dynamic
control block.

IE .NST

Task "tname" is not installed.

IE. RSU

One or more of the specified points is in use by other
tasks.

14.3.7.2
task to
module.

memory

allocate

module

secondary

Link a Task to Counter Interrupts - This function allows a
be activated by means of an interrupt from a single counter

QIO DPB format:
QIO$C IO. LTI, .•• <tname,, [ tevf] ,en, [ ic] >
where:
tname

= a 1- to 6-character alphanumeric task

name

2-word

Rad ix-50 format.

tevf

trigger event flag (O=none)

relative module number

counter value (optional)

The counter value if nonzero, is used to reinitialize the module in a
manner similar to that described for the Set Counter function in
section 14.3.6.4. Initialization may be bypassed by setting this
parameter to zero.
Return Status:

rs.sue

Function successfully completed.
I/O status is set to zero.

IE. IEF

Trigger event flag parameter not in the range 0-64.

IE .MOD

Nonexistent module specified.

IE .NOD

Insufficient dynamic memory to
control block.

IE. RSU

Counter is linked to another task.

The second

allocate

word

secondary

14.3.7.3 Link a Task to Terminal Interrupts - This function allows a
task to be activated by means of an interrupt from a terminal module.
The task will be activated only in response to the Control-C character
(octal 003).
QIO DPB format:
QI0$C IO.LTY, ... <tname,, [tevf]>

14-22

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
where:
tname

a 1- to 6-character alphanumeric task name

2-word,

Radix-50 formate

tevf

trigger event flag (O=none).

Return Status:
IS. sue

Function successfully completed.
I/O status is zero.

IE. IEF

Trigger event flag parameter not in the range 0-64.

IE .MOD

Nonexistent module (unit is ICSll controller).

IE .NOD

Insufficient dynamic
control block.

IE. NST

Task "tname" is not installed.

IE. RSU

Remote terminal is linked to another task.

storage

The second

allocate

word

secondary

14.3.7.4 Link a Task to Error Interrupts - This function allows a
single task ~v be activated whenever a remote unit power-fail or
nonrecoverable serial line error is detected on any or all remote
units in a system.
Only one task within a system may be linked to
error interrupts. Once linked, the selected task may receive error
reports from any ICR controller.
QIO DPB format:
QIO$C IO. LKE, •.• <tname,, [ tevf] >
where:
tname

a 1- to 6-character alphanumeric task
Radix-50 format.

tevf

trigger event flag (0

name

2-word

none)

Return Status:
IS. sue

Function successfully completed.
I/O status is zero.

IE. IEF

Trigger event flag parameter not in the range 0-64.

IE. IFC

No ICRll subsystems are installed.

IE .NOD

Insufficient dynamic
control block.

IE. NST

Task "tname" is not installed.

IE. RSU

Another task is linked to error interrupts.

14-23

storage

The second

allocate

word

secondary

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.3.7.5 Read Activating Data - This function allows
determine the conditions that caused it to be activated.

task

QIO DPB format:
QIO$C IO.RAD, ••. <stadd>
where:
stadd

address of 6-word buffer
(must be word aligned).

receive

activation

data

The buffer receives data in the following format:
WD 00

- Activation indicator

WD 01

- Physical unit number

WD 02

- Generic Code

WD 03

- Relative module number

WD 04

- Hardware dependent data

WD 05

- Hardware dependent data

The activation indicator is similar in function to the existence
indicator used when reading circular buffer entries. The indicator is
set to +l on the occurrence of an interrupt to which the reguesting
task is linked, and the appropriate data is stored. The indicator is
cleared when the data is solicited by the task.
If an interrupt
linked to the task occurs and the parameter is nonzero then the
previously stored data is not modified and the driver sets this
element with the two's complement of the number of linked interrupts
not recorded.
The physical unit number specifies the controller
interrupt.
The generic code is identical to that specified
entries, namely:

that

received

the

for

circular

buffer

generic

code

will

0 - Terminal (Control-C)
1,2,3 - Digital interrupt
4,5,6 - Counter interrupt
177770 - Fatal controller error
Hardware-dependent data is
consist of the following:

associated

with

Terminal:
WD 04

Terminal buffer contents (low byte)

WD 05

undefined

14-24

and

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Digital Interrupts:
WD 04

Module data

WD 05

Change-of-state indicator

Counter:
WD 04

Module data

WD 05

undefined

Fatal Controller Error:
WD04

Contents of ICSR register (see Table 14-7)

WD05

Contents of ICAR register (see Table 14-8)

Return Status:

rs.sue -

Function successfully completed.
I/O status is zero.

IE.BYT -

Buffer address is aligned on an odd byte boundary.

IE.NLK -

Task "tname" was not linked to interrupts.

IE.SPC -

Buffer not totally within the task's address space.

The

second

14.3.7.6 Unlink a Task from Interrupts - The functions
the following paragraphs provide the capability of:

word

described

Unlinking a task from all interrupts on a controller,

Selectively unlinking a task from interrupts by module type.

Unlink a Task from All Interrupts

This function unlinks a task from all interrupts on a given controller
and from error interrupts.
QIO DPB format:
QIO$C IO.NLK, ••• <tname>
where:
tname

1- to 6-character
Radix-50 format.

alphanumeric

task

name

rs.sue -

Function successfully completed.
I/O status is zero.

The

second

IE.NLK -

Task "tname" was not linked to interrupts.

2-word

Return Status:

14-25

word

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
b.

Unlink a Task from all Digital Interrupts

This function provides the capability of unlinking
digital interrupt points on a controller.

task

from

all

a 1- to 6-character alphanumeric task
Radix-50 format.

name

rs.sue -

Function successfully completed.
I/O status is zero.

second

word

IE.NLK -

Task "tname" was not linked to the specified
interrupt.

class

IE.NST -

Task not installed.

IE.MOD -

Nonexistent module type specified.

task

from

all

a 1- to 6-character alphanumeric task
Radix-50 format.

name

rs.sue -

Function successfully completed.
I/O status is zero.

The

second

word

IE.NLK -

Task "tname"
interrupts.

the

specified

IE.NST -

Task not installed.

IE.MOD -

Nonexistent module type specified.

unlinking

QIO DPB format:
QIO$C IO.UDI, .•• <tname>
where:
tname

2-word

Return Status:

The

Unlink a Task from Counter Interrupts

This function provides the capability of unlinking
counter module interrupts.

QIO DPB format:
QIO$C IO.UTI, ... <tname>
where:
tname

2-word

Return Status:

was

not

linked

Unlink a Task from Terminal Interrupts

This function provides
terminal interrupts.

the

capability

14-26

task

from

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
QIO DPB format:
QIO$C IOQUTY, ... <tname>
where:
tname

a 1- to 6-character alphanumeric task
Radix-SO format.

name

2-word

rs.sue -

Function successfully completed.
I/O status is zero.

The

second

word

IE.NLK -

Task "tname"
interrupts.

the

specified

IE.NST -

Task not installed.

IE.MOD -

Nonexistent module specified (i.e., device is an
controller) .

Re tu r n S ta tu s :

was

not

linked

ICSll

Unlink a Task from Error Interrupts

This function.provides the capability of unlinking
error interrupts.

task

from

all

QIO DPB format:
QIO$C IO.UER, ..• <tname>
where:
tname

1- to 6-character
Radix-SO format.

alphanumeric

task

name

2-word

rs.sue -

Function successfully completed.
I/O status is zero. -

The

second

IE.IFC -

No ICRll controllers exist in the system.

IE.NLK -

Task "tname" was not linked to error interrupts.

IE.NST -

Task not installed.

Return Status:

14.3.8

word

Terminal Output

This function allows a task to perform output to the terminal device.
Characters are output exactly as they appear in the buffer. The
carriage control parameter is not recognized. It should be noted that
only one interrupt module request per controller (terminal or A/D) may
be in progress at a given time. Thus, the driver will not initiate an
A/D operation on a given controller, until any terminal output in
progress for that controller has been completed.

14-27

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
QIO DPB format:
QIO$C IO.WLB, ..• <staddb,sizb>
where:
staddb

Buffer address (may be odd)

sizb

Byte count (may be odd)

Return Status:

Is.sue

Function successfully completed. Second word
status contains the number of bytes output.

IE.MOD

Nonexistent hardware function.
an ICSll controller.

14.3.9

I/O

Request was issued for

Maintenance Functions

The functions described below allow a privileged task to enable and
disable error reporting while troubleshooting or maintenance on a
remote unit is in progress.

14.3.9.1 Disable Hardware Error Reporting - This function allows a
privileged task to disable error reporting and error interrupts, and
restrict access to the controller while remote unit troubleshooting or
module calibration is in progress (see section 14.5.1). Upon receipt
and validation of the request, error interrupts are disabled and
subsequent controller timeouts are ignored. The occurrence of device
timeout while A/D conversion or remote terminal input is in progress
results in termination of the request with the error code IE.ABO.
When error reporting is disabled in this manner, access to the
controller for input or output to I/O modules is restricted to
privileged tasks. All other requests not requiring the transmission
of data to or from the device, are permitted for all tasks. Such
requests are as follows:
from

digital,

counter,

remote

terminal

Disconnect
interrupts

Unlink from interrupts

Read activating data

Link to digital, remote terminal, or error interrupts

Connect a buffer to digital or remote terminal interrupts

All other requests not issued by a privileged task are
the error code IE.DNR.
QIO DPB format:
QIO$C IO.FLN, •..

14-28

rejected

with

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Return Status:

rs.sue

Function successfully completed.

IE.FLN

Unit already offline.

IE.PR!

Task not privileged.

14.3.9.2 Enable Hardware Error Reporting - This function allows a
privileged task to enable error reporting and device error interrupts.
Upon receipt and validation of the function, all device error
interrupts are enabled and the unit is marked online. These actions
are performed regardless of the current state of the unit.

QIO DPB format:
QIO$C IO.NLN, •..
Return Status:

rs.sue

Function successfully completed.

IE.PR!

Task not privileged.

14.3.10

Special Functions

14.3.10.l I/0 Rundown - An I/O rundown request from
the Executive
will automatically cause the task to be disconnected from all
interrupts. The rundown operation is not finished until any A/D input
in progress for the task, has been completed.

14.3.10.2 Kill I/O - The kill I/O function allows a task to initiate
I/O rundown processing for itself on any device. Request processing
is identical to that described for I/O rundown.

QIO DPB format:
QIO$C IO.KIL, ...
Return Status:

rs.sue 14.4

Function successfully completed.

FORTRAN INTERFACE

The following table lists the FORTRAN interface subroutines supported
for the ICS/ICR subsystem.
(D) indicates a direct access call.
Unless specifically noted, all subroutines are reentrant (but not
necessarily position-independent) and may be placed in an absolute
resident library.

14-29

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-4
FORTRAN Interface
Subroutine

Fune ti on

AIRD/AIRDW

Input analog data
sequence.

AISQ/AISQW

Read a series of sequential analog
random gain.

AO/AOW

Perform analog output on several channels.

ASICLN/
ASUDLN

Assign a LUN to an ICS/ICR controller.

CTDI

Connect a circular buffer to receive digital
data.

interrupt

CTTI

Connect a circular buffer to receive counter
data.

interrupt

CTTY

Connect a circular
term in al data.

DFDI

Disconnect a buffer from digital interrupts.

DFTI

Disconnect a buffer from counter interrupts.

DFTY

Disconnect a buffer from remote terminal interrupts.

DI/DIW

Read several 16-point digital sense fields (D).

DOL/DOLW

Latch or
fields.

DOM/DO MW

Pulse multiple
fields.

LNK

Link a task to unsolicited interrupts.

OF LIN

Suppress error reporting.
status.

ON LIN

Enable error reporting.

RC I PT

Read a single digital interrupt point (D).

RDA CT

Read interrupt activation data.

RDDI

Read the digital interrupt circular buffer.

RDTI

Read the counter interrupt circular buffer.

RDCS

Read digital interrupt circular buffer. Return data on
only those points for which a change of state has been
recognized.

from

unlatch

multiple

buffer

several

16-point

14-30

channels

input

channels

receive

ICRll

random
at

remote

16-point

bistable

output

momentary

digital

output

ready

Place

unit

not

Return unit to ready status.

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-4 (Cont.)
FORTRAN Interface
Subroutine

Function

RDWD

Read digital interrupt circular buffer.
data word.

Return a

RSTI

Read a single counter module (D).

RTO/RTOW

Perform output to a remote ICRll terminal.

UNLNK

Unlink a task from unsolicited interrupts.

14.4.1

Synchronous and Asynchronous Process Control I/O

full

The Instrument Society of America (ISA) standard provides
for
synchronous and asynchronous I/O.
Synchronous I/O is indicated by
appending a W to the name of the subroutine (e.g., AO/AOW).
Except
for analog input and terminal output, all QIOs issued by the process
control subroutines are serviced immediately by the driver and are
complete upon return to the issuing task. In such cases there is no
functional difference between the synchronous and asynchronous forms;
however, both forms of the name are recognized. In the case of A/D
input and terminal output, the subroutines are functionally distinct.
If the asynchronous form is used, execution continues and the calling
program must periodically test the status word for completion.

14.4.2

Return Status Reporting

The I/O status parameter is a 2-word integer array. The first element
of the array receives the status of the FORTRAN call in accordance
with ISA convention.
This array serves two purposes:
1.

It is the 2-word I/O status block to which the driver returns
an I/O status code on completion of an I/O request.

The first word of the status block receives a status code
from the FORTRAN interface subroutine in ISA-compatible
format, with the exception of the I/O pending condition,
which is indicated by a status of zero. The ISA standard
code for this condition is +2.

For asynchronous analog input and terminal output, status is set by
means of an asynchronous trap, therefore the trap mechanism must be
enabled while these functions are in progress.
For compatibility, the two-word status block is also required for
status returned by the direct access calls. Errors of this type that
may be returned are:
Word 1

Number of points requested is zero.

Word 1

+321

Invalid ICS/ICR module.

14-31

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
The status code must be interpreted in the context of the function
requested; however, the following general conditions will apply:
Meaning

Contents of Status Word 1
0

Operation pending, I/O in progress

Successful completion

Error in a calling argument has been
detected by the interface subroutine

3<Word l< 300.

QIO directive rejected.
code = -(WORD 1 - 3)

Word 1

> 300

Actual

Request rejected by driver.
error code = -(WORD 1 - 300)

error
Actual

Table 14-5 lists all possible status values:
the FORTRAN value,
assembly language mnemonic, actual value and related definition.
Table 14-5
Return Status Summary
ASSEMBLY
LANGUAGE
VALUE

ASSEMBLY
LANGUAGE
MNEMONIC

DEFINITION

IS. PND

Operation pending.

IS. sue

Successful completion.

none

Error detected in FORTRAN calling
sequence.

-1

IE. UPN

Insufficient dynamic
allocate I/O packet.

-5

IE. ULN

Unassigned LUN.

-6

IE. LNL

LUN usage interlocked.

+99

-96

IE. ILU

Inv al id LUN.

+100

-97

IE. IEF

Invalid event flag number.

+101

-98

IE.ADP

Part of DPB
out
addressing space.

+102

-99

IE. SDP

Invalid DIC or DPB size.

+301

-1

IE. BAD

Bad parameters.

+302

-2

IE. IFC

Invalid I/O function code.

+303

-3

IE. DNR

Device not ready.

FORTRAN
INTERFACE
VALUE

14-32

storage

user's

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-5 (Cont.)
Return Status Summary
----

FORTRAN
INTERFACE
VALUE

ASSEMBLY
LANGUAGE
VALUE

ASSEMBLY
LANGUAGE
MNEMONIC

DEFINITION

+306

-6

IE. SPC

Illegal buffer.

+315

-15

IE .ABO

Request aborted.

+316

-16

IE. PRI

Privilege violation.

+317

-17

IE. RSU

Resource in use.

+319

-19

IE. BYT

Buffer address or length is odd.

+321

-21

IE .MOD

Illegal module number.

+322

-22

IE.CON

Another task already connected to
interrupts.

+323

-23

IE.NOD

Insufficient dynamic memory to
allocate secondary control block.

+379

-79

IE. NLK

linked
Task not
interrupts.

+380

-80

IE. FLN

ICRll already of fl ine.

+381

-81

IE.NST

Task is not installed.

+397

-97

IE. IEF

Invalid event flag number.

14.4.3

ICS/ICR

Optional Arguments

The calling sequences discussed in subsequent sections frequently
contain optional arguments.
These arguments are enclosed in square
brackets within the calling sequence description.
A
statement
containing such arguments may be written with these parameters deleted
by truncating the argument list if the optional parameters are at the
end of the calling sequence, or replacing them with commas if they are
embedded elsewhere in the list. Consider the routine XYZ below having
two optional arguments.
CALL XYZ ( ibuf [, ilen]

[, ival])

If the argument ival is to be omitted then the calling sequence
be:

would

CALL XYZ(IBUF,ILEN)
When an optional argument in the middle of the list is to be omitted
it is replaced with a comma. Consider the routine XYZ, above. The
following statement is used to omit the parameter ilen
CALL XYZ(IBUF,,IVAL)

14-33

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.4

Assigning Default Logical and Physical Units for ICS/ICR Input

an.Q_~t.E._l:!~---=---ASICLN/ASUDLN

The following subroutines must be called to assign and record a
default LUN and physical unit if either parameter is to be unspecified
in subsequent FORTRAN calls for which these parameters are optional.
Calling Sequence:
CALL ASICLN([lun]
CALL ASUDLN([lun]

[,idsw]
[,idsw]

[ , i un t] )
[ , i un t] )

Before a task can issue the call to ASUDLN, the ASN command must be
issued through MCR to assign logical device UDnn to the appropriate
physical ICS/ICR unit.
Argument Description:
lun

An optional integer variable whose value is the number
of the LUN to be assigned to the physical unit
specified by iunt or unit 0. If unspecified, no LUN is
assigned.

idsw

An optional integer variable to receive the
the assign lun directive.

iunt

An optional integer variable that specifies the unit
number to be assigned. Assumed to be zero if omitted.

result

Return Status:
The following values are returned to idsw:
+l

Assignment or function successfully completed.

-5

LON usage is interlocked because LON is assigned
to a device that is attached to another device or
a file is currently open on the LON

-96

Invalid LON

The call to ASUDLN assigns a LON to logical device UD:
and is
provided for compatibility with existing UDCll software. The call to
ASICLN assigns a LON to device IC:.
Upon successful issuance of the Assign LON directive, the subroutine
executes a Get LUN Information directive to obtain the actual unit
numbers to be saved. It is therefore possible to alter the default
physical unit referenced in a direct access call, by means of the ASN
MCR function, provided that such logical assignments are done before
the task is made active.

14-34

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Examples:
1.

Assign LUN 5 to ICR unit 3.

10
2.

CALL ASICLN (5,IERR,3)
IF(IERR) 20,10,10
--------------

Assign LUN 1 to logical device UD:, unit 0
a.

The following MCR command is issued to create logical
device UDO:, and assign all references to physical device
ICl:.
>ASN ICl:

UD:

The FORTRAN call
CALL ASUDLN (1)
assigns logical device UDO: to LUN 1.
Because of the
previous ASN command the Executive will assign this LUN
to physical device ICl: and return a value of one (1)
for
the
unit number in response to the GET LUN
Information directive. This value will be stored and
later referenced whenever the physical unit number is
unspecified in any of the FORTRAN calls that reference
the I/O page directly.

14.4.5

Analog Input

The following routines provide the capability of performing A/D input:
AIRD/AIRDW -

ISA Standard call to read multiple channels in random
order. This call requires one or more control variables
containing A/D channel and gain in the format shown in
Table 14-6.

AISQ/AISQW -

ISA Standard call
sequential order.

14-35

read

multiple

channels

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-6
A/D Conversion Control Word
Bits

Meaning

0-11

Channel number

Range:

12-15

Gain value for this
sample, expressed as
a bit pattern as
follows

Gain:

!AD-IA

0
0
0
0
0
0
0
0
1
1
1
1

0
0
0
0
1
1
1
1
0
0
0
0

0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1

0
0
1
1
0
0
1
1
0
0
1
1
0

1
1
1

1
1

0-1919

2
illegal
illegal
10
20
illegal
illegal
50
100
illegal
illegal
200
1000
illegal
illegal

14.4.5.1 AIRD/AIRDW: Analog Input-Random Channel Sequence - The ISA
standard call provides the capability of reading multiple A/D channels
in random sequence.
CALL AIRD(inm,icont,idata[ ,isb] ,lun)
or
CALL AIRDW(inm,icont, ... etc.}
Argument Descriptions:
inm

Integer variable specifying the number of
be read.

icont

An integer array of size inm containing control data in
the format shown in table 14-6.

idat

An integer array of dimension inm to receive the
converted values. Each element in the array is paired
with a control element in icont that defines the
channel and gain.

isb

An optional 2-word integer array to receive the results
of the call as follows:

channels

Conversion successfully completed.
The second
word contains the number of channels converted.

Number of channels requested was zero.

14-36

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
+4

Insufficient
packet.

LUN was

+99

Invalid LUN.

+301

At least one invalid control word was specified.
The second I/O status word contains the number of
channels successfully converted.

+303

Device not ready.
Interrupt response was not
received from an A/D channel within one second
after initiation. The second word of I/0 status
contains
the number of channels successfully
converted.

+306

Control or data buffer
user's addressing space.

+319

Control or data buffer is byte aligned.

lun

dynamic

storage

not

allocate

wholly

within

I/O

the

An integer variable specifying the ICS/ICR logical unit
number. This parameter is required.

Example:
The following example illustrates how A/D throughput can be enhanced
when several IAD-IA A/D Converters are in a system.
This is
accomplished by means of interleaved samples that initiate parallel
conversions on each module.
Samples are to be obtained from 12
channels on three IAD-IA A/D converter modules at a gain of 1.

c
c PROGRAM TO SAMPLE 12 A/D CHANNELS
c IN RANDOM SEQUENCE FOR MAXIMUM
c THRUPUT.
c
c CHANNELS TO BE SAMPLED:
c
c
0

c
c
c
c
c
c
c
c

c
c

1
2
3
120
121
122
123
240
241
242
243

-A/D MODULE 0

-A/D MODULE 1

-A/D MODULE 2

c INTERLEAVED SEQUENCE FOR MAXIMUM

c THRUPUT.
c

c
c

120
240

14-37

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

c
c

c
c
c

c
c

121
241
2
122
242
3

123
243

C THE FORTRAN CONVENTION FOR ARRAY
C STORAGE CAN BE USED TO REPRESENT
C THE ABOVE SEQUENCE IN AN N X I INTEGER
C CONTROL ARRAY. WHERE:

c
c
N
NUMBER OF MODULES TO BE SAMPLED
c
I
NUMBER OF SAMPLES PER/MODULE
c
c ALLOCATE STORAGE FOR CONTROL ARRAY

c
c

DIMENSION ICONT (3,4)

C INITIALIZE CONTROL ARRAY FOR IAD-IA MODULE 0

c
c

DATA ICONT(l,l) ,ICONT{l,2) ,ICONT(l,3) ,ICONT{l,4)/0,1,2,3/

C INITIALIZE CONTROL ARRAY FOR IAD-IA MODULE 1

DATA ICONT(2,l) ,ICONT(2,2) ,ICONT(2,3) ,ICONT(2,4)/120,121,122,123/

c
C INITIALIZE CONTROL ARRAY FOR IAD-IA MODULE 2

DATA ICONT(3,l) ,ICONT(3,2) ,ICONT(3,3) ,ICONT(3,4)/240,241,242,243/

C ALLOCATE STORAGE FOR DATA ARRAY
C IN SIMILAR FASHION TO FACILITATE
C CHANNEL REFERENCES

c
C

DIMENSION IDATA (3,4)
BEGIN EXECUTABLE STATEMENTS

c
C INITIATE A/D SYNCHRONOUS CONVERSION ON LUN 3

CALL AIRDW(l2,ICONT,IDATA,,3)

14.4.5.2

AISQ/AISQW:

Analog Input-Sequential Channel Sequence

The ISA standard call described below provides the capability of
sampling multiple A/D channels in sequential order. Channels are
sampled in increments of one, beginning with the channel specified
in
icont(l).

14-38

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
CALL AISQ(inm,icont,idata [ ,isb] ,lun)
or
CALL AISQW(inm,icont ... etc.)
Argument Descriptions:
inm

Integer variable specifying the number of
be read.

icont

An integer array of size inm containing initial channel
in the first element only, and gain in the format shown
in Table 14-6 in the remaining elements.

idat

An integer array of size inm to receive the converted
values.
Each element is paired with the corresponding
control element in icont that defines
the
gain
parameter.

elements

Channels are sampled sequentially starting with
first channel specified in element 1 of icont.
isb

the

An optional 2-word integer array to receive the results
of the call as follows:
+l

Conversion successfully completed.
The second
word contains the number of channels converted.

Number of channels requested was zero

Insufficient
packet

LUN was not assigned

+99

Invalid LUN

+301

At least one invalid control word was specified.
The second I/O status word contains the number of
channels successfully converted.

+303

Device not ready.
Interrupt response was not
received' from an A/D channel with in one second
after initiation. The second word of I/O status
contains
the number of channels successfully
converted.

+306

Control or data buffer
user's addressing space

+319

Control or data buffer is byte aligned

lun

dynamic

storage

not

allocate

wholly

within

I/O

the

An integer vaiable containing the logical unit number.
This parameter is required.

Example:
The following example illustrates the procedure for
sequential
sampling.
Five channels are converted at gains of 1, 2, 20, 50, and
1000, starting at channel 3.

14-39

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

c
C ALLOCATE SPACE FOR STATUS ARRA

DIMENSION !SB (2)

C ALLOCATE SPACE FOR CONTROL ARRAY
C AND ESTABLISH INITIAL VALUES

DIMENSION ICONT(S)
DATA ICONT(l) ,ICONT{2) ,ICONT{3)/0000003,0010000,0050000/
DATA ICONT{4) ,ICONT{S)/0100000,0150000/

C ALLOCATE SPACE FOR DATA ARRAY

DIMENSION !DAT (5)

c
C INITIATE SEQUENTIAL, ASYNCHRONOUS CONVERSION
C VIA LUN 1

CALL AISQ{S,ICONT,IDAT,ISB,l)
IF{ISB{l) .NE.0) GO TO 20
{continue processing)

c
C TEST CONVERSION STATUS

c
20

GO TO 10
{test for errors or process converted data)

END

14.4.6

Analog Output - AO/AOW: Multi-Channel

This ISA standard routine is called to output voltage from multiple
D/A channels.
Calling Sequence:
CALL AO ( inm, icnt, idat [, isb] [ ,lun])
or
CALL AOW{inm,icnt •.. etc.)
Argument Descriptions:
inm

Integer variable containing the number of channels to
be output.

icnt

Integer array containing the channel numbers to receive
outpute

14-40

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
idat

Integer array containing the output voltage setting as
a value between 0 and 1023 where:
0 = 0 volts de and
1023

isb

= +9.99 volts (full scale).

Optional 2-word integer array to receive status.
of the following values is returned
in isb(l).
second element is always zero.

One
The

Function successfully completed.

Zero channels requested.

Insufficient dynamic storage to allocate an I/O
packet.

LON was not assigned.

+99

Invalid LON.

+303

Controller not ready.

+321

Nonexistent channel specified.

lun

Optional integer variable containing the logical unit
number.

Example:
Output the variable voltages contained
channels 2 and 3 respectively.

IV(l)

c
C ALLOCATE DATA ARRAY

DIMENSION IV(2)

C ALLOCATE CONTROL ARRAY

c
c

DIMENSION ICNT(2)

C ALLOCATE STATUS ARRAY

DIMENSION ISB(2)

C INITIALIZE CONTROL ARRAY

DATA ICNT(l) ,ICNT(2)/2,3/

c
C PERFORM A/D OUTPUT VIA LUN 3

CALL AOW(2,ICNT,IV,ISB,3)
IF (ISB(l) .GE.3) go to error processor

14-41

and

IV(2)

D/A

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.7

Digital Output - DOL/DOLW:

Bistable Multiple Fields

The following ISA standard call provides the capability of latching or
unlatching multiple 16-point bistable digital output fields.
Calling Sequence
CALL DOL(inm,icnt,idat,imsk[,isb] [,lun]}
or
CALL DOLW(inm,icnt •.. etc.}
Argument Descriptions:
inm

Integer variable specifying the number of fields to
latched or unlatched.

icnt

Integer array containing the initial point within
field.

each

id at

Integer array containing binary data that defines
points within the field to be latched or unlatched.
The state of each bit is interpreted as follows:

imsk

latch point.

unlatch point.

Integer array containing binary data that defines
points within the field for which a change of state is
permitted.
A bit set to 1 defines a point that may assume the
state defined by the corresponding bit in idat. A 0
bit specifies a point for which no change of state is
permitted.

isb

lun

Optional 2-word integer array to receive the results of
the call. Status is returned in isb(l} as shown below.
isb(2} is always zero.
+l

Function successfully completed.

Zero points specified.

Insufficient dynamic storage to
packet.

LUN not assigned.

+99

Inv al id LUN.

allocate

I/O

+303

Controller not ready.

+321

Nonexistent point number specified. One
points within the field do not exist.

Optional integer specifying the Logical Unit Number.

14-42

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Example:
Reset points 0,1,20 and 21

DIMENSION ICNT(2) ,IDAT(2) ,IMSK(2)

C INITIALIZE THE CONTROL ARRAY

DATA ICNT(l) ,ICNT(2)/0,20/

C INITIALIZE MASK ARRAY TO EFFECT A
C CHANGE-OF-STATE ONLY ON THE SPECIFIED
C POINTS.

DATA IMSK(l) ,IMSK(2)/000003,0000003/

c
C RESET THE SPECIFIED POINTS.
C TO LUN 3.

ICR IS ASSIGNED

CALL DOLW(2,ICNT,IDAT,IMSK,,3)

14.4.8

Digital Input

Both of the following subroutines perform their functions
through
direct access to the ICS/ICR hardware registers. Therefore, the
physical unit number replaces LUN in the calling sequences described
below. Note that any need for conversion of BCD encoded digital input
into binary, can be accomplished through the FORTRAN function
IBIN=KBCD2B (!BCD).
Binary data can be converted to BCD through the FORTRAN function
IBCD=KB2BCD (IBIN).
NOTE
When the physical
unit
number
is
explicitly
included
in the calling
sequence, it cannot be reassigned by the
MCR command ASN.

14.4.8.1 DI/DIW:
Digital
Input - Digital Sense Multiple
Fields
- This ISA standard subroutine provides the capability of reading
multiple 16-point contact sense fields.

14-43

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Calling Sequence:
CALL DI { inm, icnt, idat [, isb] [, iun])
or
CALL DIW{inm,icnt ... etc.)
Argument Descriptions:
inm

Integer variable specifying the number of fields to
read.

icnt

Integer array containing the initial
each field.

idat

Integer array to receive the input data

isb

Optional, 2-word integer array to receive the results
of the call.
The status is returned in isb (1) as
follows:

point

number

Function succesfully completed.

Zero points requested.

+321 -

Nonexistent point requested. One or more points within
the 16-bit field does not exist.

iun

Optional integer variable specifying the physical
number.

unit

Example:
Read two contact sense fields starting at points 3 and 27 on
unit IC 2: •

physical

DIMENSION ICNT{2), IDAT(2} ,ISB{2)
DATA
ICNT(l) ,ICNT{2)/3,27/

CALL DI (2,ICNT,IDAT,ISB,2)
IF (ISB(l) .GE.3) go to error procedure

14.4.8.2 RCIPT: Digital Input - Digital Interrupt Single-Point - The
following subroutine returns the state of a single digital interrupt
point as a logical value.
Calling Sequence:
CALL RCIPT {ipt,isb[,iun])
Argument Descriptions:
ipt

Integer variable defining the point to be read.

14-44

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
isb
+l

+321 iun

2-word integer array to receive status
follows. Status is returned to isb(l).

and

data

Function successfully completed. Data is
isb(2) as a logical value, where:

returned

. TRUE.

(-1)

Point closed .

. FALSE.

(0)

Point open .

Nonexistent point specified.
Optional integer variable defining
number.

the

physical

unit

Example:
Read the state of contact interrupt point 3 on unit 0.
DIMENSION !SB (2)

CALL RCIPT (3,ISB,0)
IF (ISB(2) .EQ •. FALSE.) go to point open routine.

14.4.9

Digital Output Momentary - DOM/DOMW:

Multiple Fields

This ISA standard call allows multiple 16-bit fields to be pulsed.
Calling Sequence:
CALL DOM (inm,icnt,idat[,idx] [,isb] [,lun])
or
CALL DOMW (inm,icnt .•. etc.)
Argument Descriptions:
inm

Integer variable specifying the number of fields to
pulsed.

icnt

Integer array containing
field.

idat

Integer array defining the points to be pulsed. A bit
1s set corresponding to each point that is to be
triggered.

idx

Optional
dummy
integer
variable
retained
compatibility with the standard form of the call.

isb

Optional 2-word integer array to receive the results of
the call as follows in isb(l), isb(2) is set to zero.

14-45

the

initial

point

be
each

for

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
+l

Function successfully completed.

Number of fields to be output is zero.

Insufficient dynamic storage to allocate on I/O packet.

LUN not assigned.

+99

Invalid LUN.

+303 -

Controller not ready

+321 -

Nonexistent point specified.
a field do not exist.

lun

Optional integer variable
number.

One or more points within

defining

the

logical

unit

Example:
Pulse momentary digital output fields defined by points 20, 37
on LUN 1.

and

DIMENSION ICONT(3) ,IDAT(3)
DATA ICONT(l) ,ICONT(2) ,ICONT(3)/20,37,0/
CALL DOM(3,ICONT,IDAT,,l)

14.4.10

Remote Terminal Output - RTO/RTOW

The following function provides the capability of transmitting a
character string to a remote ICRll terminal. Both synchronous and
asynchronous forms are supported.
CALL RTO ( ibc, idat [, isb] [ ,lun])
or
CALL RTOW (ibc,idat .•.• etc.)
Argument Descriptions:
ibc

Integer variable specifying
output.

the

number

bytes

idat

Byte array (LOGICAL
string to be output.

containing

the

character

isb

Optional, 2-word integer array to receive the results
of the call in isb(l) as follows.
isb(2) is set to the
number of bytes actually transferred to ~he device.

14-46

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
0

Operation pending.

Function successfully completed.

Zero bytes to be transmitted.

Insufficient dynamic storage to allocate I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+303 -

Device not ready.
Terminal failed to respond with
BUFFER EMPTY within 1 second after character was
transmitted.

+306 -

Part or all of buffer is
addressing space.

+321 -

Nonexistent module.

lun

out

the

issuing

task's

Device is ICSll.

Integer variable defining the logical unit number.

Example:
Output a character string to
assigned to LUN 3.

remote

terminal

via

the

ICR

unit

CALL RTOW(32,'APPLY +5 VOLTS TO A/D CHANNEL 10' ,,3)

14.4.11

Unsolicited Interrupt Data - Continual Monitoring

Subroutines are provided, that permit a FORTRAN program to continually
monitor unsolicited interrupt data supplied to a user circular buffer
as described in paragraph 14.3.6. Such routines allow the program to
connect a buffer for input, disconnect the buffer upon completion and
read and return the buffer contents in a format suitable for FORTRAN
processing.
The calls summarized below perform these functions for
interrupting digital input modules, counters, and remote terminal
inputs:
Interrupting Digital Inputs
CTDI

Connect a buffer to receive digital interrupts.

RDDI

Read the state of a single interrupting point.

RDCS

Read the state of a single interrupting point for which
a change of state has been detected.

RDWD

Read 16
buffer.

DFDI

Disconnect a buffer from digital interrupts.

bits

interrupt

14-47

data

from

the

circular

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Counter Modules
CTTI

Connect a buffer to receive counter interrupts.

RDTI

Read the counter circular buffer.

DFTI

Disconnect a buffer from counter interrupts

Remote Terminal Input
CTTY

Connect a buffer to receive remote terminal inputs.

RDTY

Read remote terminal data from the circular buffer.

DFTY

Disconnect a buffer from remote terminal interrupts.

14.4.11.1 CTDI: Connect a Buffer for Receiving Digital Interrupt
Data - The following routine allows a task to provide a circular
buffer that will receive digital interrupt data, and define an event
flag that will be set upon the occurrence of each interrupt.
Calling Sequence:
CALL CTDI ( ibuf, isz, iev [, isb] [ ,1 un])
Argument Descriptions:
ibuf

An integer array making up the circular buffer that
to receive interrupt data.

isz

Integer variable specifying the length of the
buffer in words.

iev

Integer variable specifying the event flag that is to
be set whenever the driver receives an interrupt from a
digital input module.

isb

Optional, 2-word integer array to receive the results
of the call.
The status values specified below are
returned to isb(l).

circular

- Function successfully completed. isb(2) receives the
number of words passed per interrupt in the low byte.

- Insufficient
packet.

- Unassigned LUN.

dynamic

I/O

+99

- Invalid LUN.

+306

- Part of buffer is out of the user's address space
buffer is too small to accommodate a single entry.

+316

- Privilege violation - task is checkpointable and
fixed in memory.

not

14-48

storage

allocate

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
+319

Buffer address or length is an odd number of bytes.

+322

- Another task is already connected to interrupts.

+397

- Invalid event flag specified

lun

Integer variable specifying the logical unit number.

The space allocated for the circular buffer must be large enough to
accommodate at least one 5-word entry plus an additional 10 words of
storage that are required by the subroutines that read circular buffer
contents.
Thus the buffer allocation specified by the integer
variable isz may be computed as
isz

(10 + 5 * n}

where n is the number of entries to be contained in the buffer and isz
is expressed in words.

14.4.11.2 Reading Digital Interrupt Data - Each of the following
routines reads data that has been stored in the circular buffer and
performs the following common processing:
1.

Detects, and optionally reports, the occurrence of an error
entry that has been placed in the buffer by the driver
because of a nonrecoverable device fault (e.g., fatal serial
line error or remote power-fail}.

Clears the trigger event flag when no further entries
to be processed.

Clears and optionally reports any overrun conditions.

Only one of the following three routines can be invoked
task.
a.

RDDI:

remain

single

Read Digital Interrupt Data from a Circular Buffer

The RDDI FORTRAN subroutine reads contact interrupt data from a
circular buffer that was specified in a CTDI call (see 14.4.11.1
above}. It does no actual input or output, but rather performs a
point-by-point scan of an interrupt entry in the buffer, returning the
state of each point as a logical value.
On the initial call to RDDI, the module number and data of the next
interrupt entry are read from the circular buffer and stored for
subsequent reference. The subroutine then sets the current data bit
number n to zero, examines the state of data bit n, and converts bit n
to a point number via the following formula:
ipt

= module number * 16 + n

On each subsequent call, n is incremented by one and then data-bit n
is examined in the stored module data. When n reaches 16, it is reset
to zero and an attempt is made to read the next interrupt entry from
the circular buffer.
If a valid entry is not found, ipt is set
negative and ict (if specified} is either assigned a value of zero or
an overrun count that is maintained py the ICS/ICR driver. If ict is

14-49

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
zero, no further entries remain. A nonzero value indicates that the
driver received more data than could be stored in the buffer, and ict
represents the number of entries that were discarded.
The variable ict, receives the control register contents that are set
by the driver as described in section 14.3.6 whenever a nonrecoverable
controller error occurs.
Calling Sequence:
CALL RDDI (ipt,ival[,ict])
Argument Descriptions:
is a variable to which the digital input
is returned. It may be set as follows:

ipt

point

ipt < 0 if no val id en try is found

The specific value of ipt reflects the error
detected as follows:
-1 -

-2 -3 2.

iv al

no data ( i . e . ,
no
currently in buffer)
overrun
hardware error

.FALSE.

(0)

.TRUE.

(-1)

that

interrupt

ipt => 0 if the value indicated is a point
the state is returned to ival.

is a variable to which the state of
returned; it may be set as follows:

the

was
data

number;
point

if the point is open
if the point is closed

Optional integer variable to receive the overrun count.
or the contents of the CSR register on the occurrence
of a fatal controller error. Otherwise set to zero.

ic t

number

RDCS:

Read Digital Interrupt Points That Have Changed State

The RDCS FORTRAN subroutine returns data in the format of subroutine
RDDI as described above except that only points that have changed
state are processed, resulting in significantly improved throughput
and reduced processing overhead for the calling task.
Processing specific to the routine is as follows:
On the initial call, the module number, module data and change of
state information are read from the circular buffer and stored for
later reference. The subroutine then sets the current data bit number
n to zero and begins scanning the change-of-state word until a nonzero
bit is found. The point number and current state are then reported as
previously described.
If no change of state is found or when no
further bits remain to be processed, the next entry is fetched as
described above.

14-50

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
The processing of error conditions is identical to subroutine RDDI.
Calling Sequence:
CALL RDCS (ipt,ival[,ict])
Argument Descriptions:
ipt

Integer variable to receive the digital
number. It may be set as follows:
1.

-1 -

iv al

no data
overrun
hardware error

ipt => 0 if the value indicated is a point
the state is returned to ival.

number,

Integer variable to receive the state of the point as a
logical value where:

ic t

point

ipt < 0 if no valid entry is found (i.e., overrun,
error or no data in buffer). The specific value of
ipt reflects the error that was detected
as
follows:
-2 -3 -

input

. FALSE.

( 0)

point open

.TRUE.

(-1)

point closed

Optional integer variable. A nonzero value indicates
that the variable has been set with an overrun count
returned by the driver, or with the contents of the CSR
register on the occurrence of a fatal controller error.
Otherwise set to zero.
RDWD:

Read a Full Word of Digital Interrupt Data

The following subroutine is called to return a full word of digital
interrupt data from the circular buffer, and optionally change of
state information. A new entry is read for each call;
hence,
throughput is high when processing is contingent upon several possible
conditions within a module.
Calling Sequence:
CALL RDWD (imod,ival[,ict][,icos])
Argument Descriptions:
imod

an integer variable to receive
status as follows:
1.

the

module

imod < 0 if no data is present
condition or error was detected

The specific value of ipt reflects the
was detected as follows:
-1 -2 -3 2.

number
an

overrun

error

that

no data
overrun
hardware error

imod => 0 Module number.
14-51

Interrupt data is in ival

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
iv al

Integer variable to receive the digital interrupt data.

ic t

icos

Optional integer variable to receive change-of-state
information.
Bits set to a 1 correspond to points for
which a change of state has been recorded.

14.4.11.3 DFDI: Disconnect a Buffer from Digital Interrupts - The
following routine is called to disconnect a task's circular buffer
from digital interrupts.
Calling Sequence:
CALL DFDI ( [ isb] [ ,lun]}
Argument Descriptions:
isb

lun

Optional 2-word integer array to receive the results of
the call as follows.
isb(l) is always zero.
+l

Function successfully completed.

Insufficient dynamic storage to allocate I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+322 -

Task not connected to interrupts.
Optional
number.

integer

variable

containing

logical

unit

14.4.11.4 CTTI: Connect a Buffer for Receiving Counter Data - The
following subroutine may be called to connect a circular buffer that
is to receive counter data and to define an event flag that is to be
set upon occurrence of each interrupt.
Calling Sequence:
CALL CTTI (ibuf,isz,iev,iv[,isb] [,lun]}
Argument Descriptions:
ibuf

An integer array making up the circular buffer that
to receive interrupt data.

isz

Integer variable specifying the length of the
buffer in words.

iev

Integer variable defining an event flag that is to be
set whenever the driver receives an interrupt from a
counter module.

14-52

circular

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
iv

Integer array of initial counter values.
One element
is required for each counter in the physical unit. The
value is used to initialize and reset the counter when
a value of ~~~v ~u reached. This parameter may be
reset for a specific module through a call to SCTI.

isb

Optional 2-word integer array to receive the results of
the call.
The status values specified below are
returned to isb(l}.

lun

Function successfully completed. isb(2}
receives the
number of words passed per interrupt in the low byte.

Insufficient dynamic storage to allocate an I/0 packet.

Unassigned LUN.

+99

Invalid LUN.

+303 -

Controller not ready.

+306 -

Part of buffer is out of the user's address space
buffer is too small to accommodate a single entry.

+316 -

Privilege violation - task is
fixed in memory.

not

+319 -

Buffer address or length is an odd number of bytes.

+322 -

Another task is already connected to interrupts.

+397 -

Invalid event flag specified.

checkpointable

and

Integer variable specifying the logical unit number.

The space allocated for the circular buffer must be large enough to
accommodate at least one 4-word entry plus an additional 8 words of
storage required by the subroutine that reads buffer contents (RDTI).
The buffer allocation specified by the variable isz may be computed as
isz = {8 + 4 * n)
where n is the number of entries to be contained in the buffer.

14.4.11.5 RDTI: Read Counter Data from the Circular
following call returns counter interrupt data from
buffer. A new entry is read on each call.
Calling Sequence:
CALL RDTI ( imod, ival [, ic t]}

14-53

Buffer - The
the circular

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Argument Descriptions:
Integer variable to receive module number and status as
follows:

imod

imod < 0 No data in buffer, data overrun or error
condition
detected The specific value of ipt
reflects the error that was detected as follows:
-1 -2 -3 -

no data
overrun
hardware error

imod => 0 Module number of counter.
is in ival.
to

receive

Integer variable
interrupt.

ict

Optional integer variable to receive the overrun count,
or the !CSR contents returned by the driver on the
occurrence of a fatal hardware error.
Otherwise, set
to zero.

Miscellaneous Counter Routines

Read a Counter Module

RSTI:

counter

The following routine directly accesses a counter register
its current value.

data

iv al

14.4.11.6

the

Interrupt data

return

Calling Sequence:
CALL RSTI (imod,isb,[,iun])
Argument Descriptions:
imod

An integer variable
counter to be read.

the

isb

A two-word integer array to receive status and data
follows. Status is returned to isb(l).

Function successfully completed.
isb(2).

+321 -

Nonexistent module specified.

iun

containing

number

Data is

ICS/ICR

The following routine may be called by any task to revise the
value that is used to activate a counter.

initial

SCTI:

specifying

returned

the

Optional integer variable
physical unit number.

the

Reset a Counter Initial Value

14-54

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Calling Sequence:
CALL SCTI ( imod, iv al [, isb] [ ,lun])
Argument Descriptions:
imod

Integer variable specifying the relative module
of the counter to be reset.

iv al

Integer value specifying the new initial value.

isb

Optional 2-word integer array to
follows.
isb(2) is always zero.

receive

number

status

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+303 -

Controller not ready.

+321 -

Nonexistent module specified.
Optional integer specifying the logical unit number.

lun

14.4.11.7 DFTI: Disconnect a Buffer
following subroutine is called to
buffer from interrupts.

from Counter Interrupts - The
disconnect the task's circular

Calling Sequence:
CALL DFTI ( [ isb] [ ,lun])
Argument Descriptions:
Optional 2-word integer array to
follows.
isb(2) is always zero.

isb

lun

receive

status

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+322 -

Task was not connected to interrupts.
Opti~nal

integer variable specifying the

Number.

14-55

Logical

Unit

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.11.8 CTTY: Connect
a
Circular
Buffer
to
Terminal
Interrupts - The following routine allows a task to provide a circular
buffer to receive remote terminal input data, and to define an event
flag that is set on the occurrence of each interrupt.
Calling Sequence:
CALL CTTY ( ibuf, isz, iev [, isb] [ ,lun])
Argument Descriptions:
The following arguments are identical in form and function
described for subroutine CTDI (section 14.4.11.1).

those

buffer,

that

ibuf

An integer array making up the
receives interrupt data.

isz

Length of the circular buffer in words.

iev

Event flag to be set on each terminal interrupt.

circular

Buffer size is computed as
isz

(8 + 4 * n)

where n is the number of entries that can be stored in the buffer.
isb

lun

Optional 2-word integer array to receive the results of
the call.
The status values specified below are
re turned to. isb ( 1) .
+l

Function successfully completed. isb(2) receives the
number of words passed per interrupt in the low byte.

Insufficient dynamic storage to allocate an I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+306 -

Part of buffer is out of the user's address space
buffer is too small to accommodate a single entry.

+316 -

Privilege violation - task is
fixed in memory.

and

not

+319 -

Buffer address not on a word boundary or length
odd number of bytes.

+321 -

Nonexistent module specified.

+322 -

Another task is already connected to interrupt.

+397 -

Invalid event flag specified.
Logical unit number.

14-56

checkpointable

Unit is ICSll.

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.11.9 RDTY: Read a Character from the Terminal Buffer - This
subroutine retrieves a single character from the terminal circular
buffer on each call.
Calling Sequence:
CALL RDTY (ind,ichr[,ivr])
Argument Descriptions:
ind

An integer variable to receive status as follows:
1.

=O character retrieved from buffer is in ichr

<0 no data in buffer, overrun, or hardware error
The specific value of ind reflects the
was detected as follows:
-1 -2 -3 -

error

that

no data
overrun
hardware error

ichr

Logical * 1 or integer variable to receive the terminal
data.
If an integer is specified only the low byte
will be set.

ivr

Optional integer variable to receive the overrun count,
or the ICSR contents on the occurrence of a fatal
hardware error. Otherwise set to zero.

14.4.11.10 DFTY: Disconnect a
Circular
Input - The following routine disconnects
from terminal inputs.

Buffer
from
Terminal
a task's circular buffer

Calling Sequence:
CALL DFTY ( [ isb] [ ,lun])
Argument Descriptions:
isb

lun

Optional, 2-word integer array to receive status
isb(l) as follows.
isb(2) is always set to zero.

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+322 -

Task was not connected to interrupts.
An optional integer array specifying the
number.

14-57

logical

unit

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.11.11 Programming Example - The following are excerpts from a
FORTRAN program that is to monitor a remote terminal for input and
echo the received characters when a carriage return is detected.

c
C

SPECIFY BYTE FORMAT FOR TERMINAL DATA
LOGICAL*l TCHR

c
C
C

ALLOCATE STORAGE FOR THE TERMINAL
BUFFER

c
C
C
C

DIMENSION IBUF(32)
ALLOCATE STORAGE FOR THE PACKED
INPUT DATA SO THAT IT IS ALIGNED
ON A WORD BOUNDARY

•C

DIMENSION ICHR(40)
DIMENSION TCHR(80)
EQUIVALENCE (TCHR,ICHR)

c
C
C

ALLOCATE STORAGE FOR A
2-WORD STATUS BLOCK
DI MENS ION I SB ( 2)
INITIALIZE ICRll LOGICAL UNIT(?) AND
TRIGGER EVENT FLAG NUMBER(2)
DATA IEV, LUN/2, 7/

c
C
C

c
c

CONNECT THE TASK TO TERMINAL
INPUTS. IF CONNECT FAILS--STOP 1
CALL CTTY (IBUF,32,IEV,ISB,LUN)
IF (ISB(l) .GE.3) STOP 1

C
C
C
C
C

10--POLL THE CIRCULAR BUFFER
FOR DATA. ECHO THE LINE WHEN
80 CHARACTERS ARE RECEIVED
OR A CARRIAGE RETURN IS
DETECTED.

DO 70 I

20--WAIT FOR TRIGGER EVENT FLAG

CALL WAITFR (IEV)

C
C

30--PACK THE CIRCULAR BUFFER DATA
INTO THE BYTE ARRAY

CALL RDTY (ISB,TCHR(I), IVR)

DISPATCH ON ERROR CONDITION

c
c

c
c
c
c
c
c

= 1,80

GO TO (20,50,40)-ISB
GO TO 60
40--REPORT HARDWARE FAULT
14-58

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

c
40

CALL ALARM (IVR)

GO TO 30

50--REPORT OVERRUN CONDITION

CALL LOST (IVR)

GO TO 30

C
C
C

60--CHECK FOR CARRIAGE RETURN,
EXIT TO ECHO ROUTINE IF
PRESENT

IF (TCHR(I) .EQ."15) GO TO 80

CONTINUE

80--FALL THROUGH TO ECHO A LINE

c
c
c

c
C

CALL RTOW (I,TCHR,,LUN)
DISCONNECT TERMINAL BUFFER, EXIT
CALL DFTY (,LUN)
CALL EXIT
END

The procedure for reading the buffer
summarized as follows:

the

example

the

call

Wait for the trigger event flag
connect the buffer.

Upon regaining control, call the appropriate routine to read
the buffer until one of the following terminal conditions is
detected:
All data has been read,

An overrun count is detected,

A fatal error is encountered.

may

specified

above

On the occurrence of 2a or 2b, perform any appropriate
processing; then return to scan for additional data.

If a hardware error is detected, use the ICSR register
contents
for
further
fault
analysis and warning as
appropriate. In the event of such an error, the event flag
will not be set by the driver again unless normal service is
resumed.

The calling task should not execute the Wait-For
until the buffer-empty condition is detected.

14-59

directive

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.12

Unsolicited Interrupt Processing - Task Activation

The following routines provide the capability of linking a task to an
interrupt, soliciting information from the driver concerning how the
task was activated, and unlinking a task from all interrupts.

14.4.12.1 LNK: Link a Task to Interrupts - This subroutine allows
any
installed task to be activated on the occurrence of any
unsolicited interrupt.
Calling Sequence:
CALL LNK (tnam,iprm[,isb][,lun])
Argument Descriptions:
tnam

Real variable containing task name in RADIX-SO format.

iprm

5-word integer array containing the following data:

iprm(l)

Interrupt class.

May be one of the following:

0 - Digital interrupts
1 - Counters

2 - Remote terminal (Control-C only)
3 - Error interrupts.
iprm ( 2)

Reserved

iprm ( 3)

Optional event flag set if task to be activated is
dormant when the interrupt occurs.

iprm(4)
iprm ( 5)

Hardware-dependent parameters as follows:

Interrupt Class

Parameter Contents

Digital

iprm ( 4)
iprm ( 5)

Point number
Change-of-state mask

Counter

iprm(4)
iprm ( 5)

Module number
Counter initial value

Remote Terminal

iprm ( 4)

not used

Error

iprm(S)

not used

isb

not

Optional 2-word integer array to receive status
isb{l) as follows.
isb(2) is always set to zero.
+l

Function successfully completed.

Unrecognized interrupt class specified.

Insufficient dynamic storage to allocate I/O packet.

14-60

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
+8

Unassigned LUN.

+99

Invalid LUN.

+301 -

Task tnam not installed.

+303 -

Controller not ready.

+317 -

Resource in
interrupt.

+323 -

Insufficient dynamic
control block.

+397 -

Invalid event flag number specified.

iun

use.

Other

task

memory

already

linked

allocate

secondary

Optional integer specifying the logical unit number.

Example:
Link task ALARM to report fatal hardware
malfunction on any ICRll physical unit.

errors

arising

from

DIMENSION IPRM(5)

C
C
C
C

INITIALIZE PARAMETER ARRAY WITH:
l. OPERATION CODE
2. RESERVED ELEMENT CLEARED
3. GLOBAL EVENT FLAG

DATA IPRM{l), IPRM(2), IPRM(3)/3,0,64/
DATA ALARM/6RALARM /

CALL LNK (ALARM,IPRM,,7)

14.4.12.2 RDACT: Read Activation Data - The following call allows a
task to determine the interrupt conditions that caused it to become
active.
Calling Sequence:
CALL RDACT ( iprm [, isb] [ ,lun])
Argument Descriptions:
iprm

6-word integer array to receive activatin data in
following format.

iprm(l)

Activation indicator (see section 14.3.7.5).

iprm ( 2)

Physical unit number of ICR.

14-61

the

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Generic code.

iprm ( 3)

Set to one of the following values.

0 - Remote terminal
1,2,3 - Digital interrupt
4,5,6 - Counter interrupt
177770 - Fatal hardware error
iprm ( 4)

Relative module number.

iprm ( 5)
iprm(6)

Hardware-dependent data.

The following data is
module.
Module Type

returned

based

Generic Code

upon

the

type

interrupt

Parameter Contents

Remote Terminal

iprm ( 5)
iprm(6)

terminal input character
undefined

Digital Interrupt

1,2,3

iprm(5)
iprm(6)

module data
change-of-state data

Counter

4,5,6

iprm(5)
iprm ( 6)

value of the counter at interrupt
undefined

Error

177770

iprm ( 5)
iprm(6)

contents of ICSR
contents of !CAR.

isb

Optional 2-word integer array to receive
isb(l) as follows.
isb(2) is set to zero.

status

Function successfully completed.

Insufficient dynamic storage to allocate I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+306 -

iprm array
space.

+319

Address of iprm is odd.

+379 -

Task not linked to ICS/ICR interrupts.

lun

not

fully

within

the

task's

Optional integer variable specifying the
number.

addressing

logical

unit

Example:
The following is an excerpt from a program that reads activating data
into array IACT and conditionally exits if the event flag (IEFN)
specified in a previous link request is not set.

14-62

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

c
C

ALLOCATE SPACE FOR DATA ARRAY
DIMENSION IACT(6)

CALL RDACT (IACT,,7)

c
C

c
c

C
C

c
C
C

CLOSE ALL FILES
CALL CLOSE(l)
CALL CLOSE ( 2)
EXIT IF TRIGGER EVENT FLAG IS NOT SET
ELSE CLEAR EVENT FLAG AND RESTART.
CALL EXITIF (IEFN)
FLAG WAS SET.
CON'I' INUE .

CLEAR IT AND

CALL CLREF (IEFN)
GO TO 10
STOP
END
The foregoing example illustrates the following considerations when
task is made active by ICS/ICR interrupts:

To avoid race conditions, the Exit-If directive should be
used
to
test
the state of the event flag and conditionally
exit.
Issuing a Test Event Flag directive followed by an
Exit would cause a
flag set condition occurring after the
test to go unrecognized.

Use of the Exit-If directive bypasses the closure of all
files
that is normally done automatically by the FORTRAN
object time system when the program executes a STOP or CALL
EXIT statement.
Thus,
to exit cleanly, the program must
explicitly close all files before invoking the directive.

14.4.12.3 UNLNK:
Remove Interrupt Linkage to a Task - The following
call removes all linkage between a task and ICS/ICR interrupts.
Calling Sequence:
CALL UNLNK (tnam,iprm, [isb], [lun])

14-63

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Argument Descriptions:
tnam

Real variable containing task name in Radix-SO format.

iprm

Integer variable containing the interrupt
be one of the following:

class.

May

0 - Digital interrupts
1 - Counters
2 - Remote terminal
3 - Error interrupts
4 - All interrupts.
Optional, 2-word integer array to receive the results
of the call in isb(l) as follows.
isb(2) is set to
zero.

isb

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

Unassigned LUN.

+99

Invalid LUN.

+379 -

Task not linked to ICS/ICR interrupts.

+380 -

Task not installed.

lun

Integer variable specifying the logical unit number.

Example:
Remove the linkage between task ALARM and all ICS/ICR interrupts.
DATA ALARM/6RALARM /
CALL UNLNK (ALARM,,,7)

14.4.13

Maintenance Functions

The following functions cause the ICS/ICR driver to suppress or enable
hardware error reporting while online maintenance and troubleshooting
is in progress as described in paragraph 14.3.9.
OF LIN

- Place selected unit offline.

ON LIN

- Return selected unit to online status.

These calls may be issued only by a privileged task.

14-64

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.4.13.l OFLIN: Place Selected Unit
in
Offline
following call is executed to set a controller offline:
CALL OFLIN ([isb]

Status - The

[,lun])

Argument Descriptions:
Optional 2-word integer array to receive the results of
the call in isb(l) as follows.
isb(2) is always zero.

isb
+l

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

LUN not assigned.

+99

Invalid LUN.

+316 -

Issuing task not privileged.

+380 -

Device already offline.

lun

Optional integer variable
logical unit number.

the

ICS/ICR

14.4.13.2 ONLIN: Return a Device to Online Status - The
call will return the selected unit to online status.

following

CALL ONLIN ( [ isb]

specifying

[ ,lun])

Argument Descriptions:
isb

14.5

Optional 2-word integer array to receive the results of
the call in isb(l) as follows.
isb(2) is always zero.
+l

Function successfully completed.

Insufficient dynamic storage to allocate an I/O packet.

LUN not assigned.

+99

Invalid LUN.

+316 -

Issuing task not privileged.

ERROR DETECTION AND RECOVERY

Error Detection
contingencies.

and

recovery

procedures

encompass

Nonrecoverable serial line errors

Power-fail at the remote station

Power recovery at the processor

No response from an interrupting module

14-65

the

following

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
The first two conditions are dealt with in a manner similar to other
types of unsolicited interrupts. Specifically, such occurrences may
cause a task to be activated, and are reported to all tasks that are
connected to digital, counter or terminal input.
The following
paragraphs discuss specific driver activity relating to each error
condition.
14.5.1

Serial Line Errors

The
driver
detects
nonrecoverable
serial
line
errors.
A
nonrecoverable error condition is defined as the occurrence of a
predetermined number of error interrupts in an interval of 1 second or
no response from the controller upon initiation of an output data
transfer via the serial line. The occurrence of such a condition
causes the driver to perform as follows:
1.

Place the controller in a "not ready" status

Disable further error interrupts

Report the condition to the task that is linked to errors,
and to any tasks connected to receive unsolicited interrupt
data from the faulty unit.
Subsequent QIO requests that
transfer data to or from the unit are rejected with a status
of IE.DNR.

Requests for interrupting modules that are pending (A/D converters and
terminal output) are allowed to time out with the error code IE.DNR.
The serial line error rate required to consider the link inoperative
may be specified by the user at the time of system generation.
After reporting the error as described above, the driver will
automatically remove the "not ready" status when the error condition
is not detected at the end of any I-second interval.
If requested
during system generation, the state of the following remote modules
will be restored as described.
1.

Bistable outputs - set to last recorded state

Counters - reinitialized to last initial value

Analog outputs - restored to last output value.

14.5.2

Power-fail at a Remote Site

The detection of AC low from the remote site will immediately trigger
the processing described in section 14.5.1. The absence of AC-low
will automatically return the unit to the "ready" status.
If specified, the state of the following remote modules will
restored as described:
1.

Bistable outputs - set to last recorded state

Counters - reinitialized to last initial value

Analog outputs - restored to last output value.

14-66

then

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.5.3

Power Recovery at the Processor

Power recovery by the processor will initiate the activity described
in section 14.5.2 for both local and remote file boxes. However,
power recovery processing at the processor will not be reported to a
task that is linked to error interrupts or connected to receive
unsolicited interrupt data.
14.5.4

Unit in Offline Status

A unit that is offline (see section 14.3.9.1)
is considered to be
under manual control for purposes of diagnosis and maintenance. Under
these conditions, error reporting as described in section 14.5.1 is
unnecessary and frequently undesirable since fault indications are
generally a by-product of these activities (i.e., a remote unit is
shut down to install an I/O module)
not the result of a genuine
controller fault.
Furthermore, to permit the operation of diagnostic software, it is
advisable to attempt to service all QIO requests regardless of the
controller status. Consequently, under these circumstances, error
reporting and detection are modified as follows when the controller is
of fl ine:
1.

Access to the controller with the intention of transmitting
data to or from the device is restricted to privileged tasks.

The task linked to error interrupts and any tasks receiving
interrupt data are not notified of remote power-fail or fatal
serial line errors.

All device error interrupts become disabled.

An attempt is made to service all QIO requests if issued by a
privileged task.
If such requests time out (i.e., A/D
converter or remote terminal output), they are terminated
with the error code IE.ABO rather than with IE.DNR. No
hardware errors are reported for I/O requests that are
normally
completed
immediately (e.g., bistable digital

14.5.5

Error Data - ICSR and ICAR Registers

Whenever a reportable error occurs, the driver returns the contents of
the appropriate control and status register (ICSR) and, in.some cases,
the contents of the address register
(ICAR)
to assist in fault
diagnosis.
Tables 14-7 and 14-8 describe the contents of these
registers.
Table 14-7
ICSR Contents
BIT

NAME

READ/WRITE

DESCRIPTION

OUTPUT BUSY

Indicates output
new data.

MAINT

R/W

Maintenance.

NOT USED

Always set to 1.

ERROR

Indicates occurrence of communication
serial line error. Reset when ICAR is
read.

14-67

buffer

cannot

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
Table 14-7 (Cont.)
ICSR Contents
BIT

NAME

READ/WRITE

DESCRIPTION

MA INT

R/W

Maintenance.

PWR FAIL

Remote Power Supply AC LO indicator.

TBMT INT EN

R/W

Enables bit 15 of ICAR to interrupt.

MA INT

R/W

Maintenance.

MOD INT

Indicates I/O Module requires
servicing.

RESET

Resets all I/O modules.

interrupt

Always read

TTY ENABLE

PWR FAIL INT ENABLE R/W

Enables bit 10 to interrupt.

BMT INT ENABLE

R/W

Enables
complement
interrupt.

MOD INT ENABLE

R/W

Enables Bit 7 to interrupt.

ERROR INT ENABLE

R/W

Enables Bit 12 to interrupt.

RIF

R/W

Resets the interrupting module's flag
when set and the module is addressed.
This clearing action also resets the RIF
bit.

R/W

Activates TTY mode, disables I/O mode.

15 to

bit

Table 14-8
ICAR Contents
BIT

NAME

DESCRIPTION

TBMT

Indicates TTY output buffer
new data.

PCL

Pulse closed. This bit is set by a
jumper on a digital interrupt module.
This jumper is removed
if
contact
closures are not of interest to the
user.

POP

Pulse Opened. This bit is set by a
jumper on a digital interrupt module.
This jumper is removed if
contacts
opening are not of interest to the user.

Indicates terminal character has been
received.
Cleared by reading terminal
character.

11-08

Generic Code

A 4-Bit binary code that identifies
the
type
of module requesting
interrupt.

07~00

Module Address

8-Bit address of the
the interrupt.
14-68

module

can

the
the

requesting

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.6

DIRECT ACCESS

Section 14.1.3 notes those ICS/ICR-11 functions that may be performed
by referencing a module through its physical address in the I/O page.
Under RSX-llM such access is accomplished by one of the following
methods:
1.

A privileged task or any task running in an unmapped system
has unrestricted access to the I/O page and may therefore
access each module by absolute address.

Using the Task Builder, a task may link to a global common
area whose physical address limits span a set of locations in
the I/O page. This method applies to either a mapped or
unmapped system.

_. . . -t...--

The latter method allows a task to be transported to any other i:JYi:Jl...l::lll
simply by relinking.
Moreover, in a mapped system the memory
management hardware aborts all references to device registers outside
the physical address limits of the common block.
Because the software allows arbitrary module placement,
direct
reference, in either case, must be accomplished by translating a
relative module number to a physical or virtual register address
within the I/O page.
This translation or mapping is performed by
means of a table (ICTAB.MAP} that is created during system generation,
and inserted in the system object module library.
The operations required to implement each method may be summarized
follows:
1.

Unrestricted access to the I/O page
a.

Based upon the user's response to the ICS/ICR SYSGEN
queries, the MACRO source file ICTAB.MAC is automatically
created nder UIC [11,10] on the source disk. This file
contains tables that describe the physical location of
each counter, digital interrupt, and digital sense module
in the target system.

ICTAB.MAC is assembled rur eventual
system object module library.

The MACRO source file ICOM.MAC, under UIC [11,10] on the
source disk, is assembled to generate global definitions
for the first ICS/ICR address on the I/O page and the
number of ICS/ICR controllers in the target system. The
resulting object file is incorporated in the system
library file.

A task is built containing the appropriate
global
references.
Such references are resolved when the Task
Builder automatically searches the system library.

inclusion

Steps a, b, and c are executed once. Step d is performed each time
task that references the ICS/ICR-11 is created.
2.

the

Access to the I/O page through a Global Common Block:

Steps la and lb are performed.

File !COM.MAC under UIC [11,10] is

14-69

assembled

define

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
the first ICS/ICR module address as a relocatable value,
the number of I/O page locations required, and the number
of controllers present on the target system.
c.

File !COM.OBJ, created in step b, is linked using the
Task Builder to create an image of the Device Common
Block on Disk.

The SET and INSTALL MCR or VMR commands are used to
allocate space for the common block and declare the block
resident in the target system.

A task is created containing the appropriate global
references to the common block and mapping table. Common
block references are resolved by directing the Task
Builder to link the Task to the device common block
(ICOM). The mapping table reference is resolved from the
system library module ICTAB.

The detailed procedure for creating the necessary object files and
device common block is performed automatically as part of the system
generation process, and is described fully in the RSX-llM System
Generation
Manual
(DEC-11-0MGIA-C-D)
(SYSGEN).
Therefore, the
discussion in the following paragraphs is limited to procedures for
linking to the device common block, and using the file ICTAB.MAC to
determine module addresses within the I/O page.
NOTE
ICS/ICR inputs are not valid until 3ms
after power recovery at the processor.
Tasks that
are
referencing
inputs
directly may establish a power recovery
AST entry point that suspends
task
execution
for
the
necessary
time
interval.
14.6.l

Linking a Task to the ICS/ICR Common Block

Once the device common block has been created, a task may access
ICS/ICR modules by linking to the common block. This can be done by
using the Task Builder commands shown in the following example.
TKB>TASK,LP:=TASK.OBJ
TKB>/
ENTER OPTIONS:
TKB> COMMON=ICOM:RO
TKB>/
The illustration is valid for either a mapped or unmapped system.
In
both cases the Task Builder links the task to the common block by
relocating the global symbol definitions contained in the common block
symbol table file ICOM.STB located under UIC [1,1].
If memory
management is present, the Executive will map the appropriate physical
locations into the task's virtual addressing space when the task is
made active.

14-70

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.6.2

Accessing the I/O Page

After the task has been linked to the I/O

through

reference

the

device

page,

common

either

directly

block, access to specific

ICS/ICR counter, or digital input modules during task execution
three-step process:

The task generates a request for module data by specifying
module type, relative module number and physical unit number.

The data contained in module ICTAB is accessed
to
translate
the arguments of step 1 to a physical offset from the ICS/ICR
base address on the I/O page.

The ICS/ICR base address, defined
in
the common block or
system library module that was created from file ICOM.MAC, is
added to the offset to compute a physical or virtual address
and the module data is read.

The next few paragraphs describe the
format of the system library
module ICTAB, and common block module ICOM in detail.
A sample MACRO
subroutine that references these modules is then presented.

14.6.2.1 Mapping Table Format - The mapping table created by SYSGEN
(file ICTAB.MAC)
is used
to translate module type, relative moauie
number, and physical unit number for counter, digital
interrupt,
and
digital
sense modules,
to
the physical or virtual address of the
module on the I/O page.
This module must be assembled and inserted in
the system object module library before the standard FORTRAN callable
routines can be used to read digital input and counter modules.
The
table contains one set of entries for each physical unit.
The entry
sets are arranged in order of ascending unit number
(Figure 14-lA).
Entries within each unit are arranged in sequence by module type as
shown in this figure.
The structure of each entry is depicted in Figure 14-lB.
Entries are
18 bytes long.
Byte 0 contains the highest number of modules of a
given type that can be referenced for the controller.
Bytes 2 through
17, when indexed by relative module numbers, yield a value between 0
and 255 representing the physical location of the module within
the
set of external page addresses allocated to the ICS/ICR-11.
The following global symbols are defined by this module:
.ICTAB

Location of mapping tables

I.CTBL

Length in bytes of one set of entries

14-71

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

DIGITAL SENSE
DIGITAL INTERRUPT

UNIT ~
MAPPING TABLE

COUNTER MODULES

...

INCREASING
MEMORY
ADDRESSES

DIGITAL SENSE
DIGITAL INTERRUPT

UNIT n
MAPPING TABLE

COUNTER MODULES

Figure 14-lA
Mapping Table Format

BYTE
MAX. MODULE NO.

RESERVED

PHYSICAL MODULE NO. PHYSICAL MODULE NO.
INCREASING
MEMORY
ADDRESSES

Figure 14-lB
Mapping Table Entry Format

14-72

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.6.2.2 I/O Page Global Definitions - As previously
mentioned,
module ICOM contains symbolic definitions for I/O page references that
are resolved either through unrestricted access or by means of a
device common block that is resident on the I/O page. The procedures
for implementing either method are carried out
during
system
generation. Upon completion, the following global symbols are defined
and later referenced by the FORTRAN callable subroutines:
.ICMD =
I$$Cll

First ICS/ICR virtual or physical
the I/O page.

address

within

= Number of ICS/ICR controllers

If the global common block was built, the definitions above are
contained in the symbol table file that was created by the Task
Builder; otherwise, they are included in the system object module
library.*

14.6.2.3 Sample Subroutine - The following subroutine, residing in
the system library, utilizes the modules previously described, to read
ICS/ICR module data.
READ ICS/ICR-11 DIRECT ACCESS INPUTS
LOCAL DATA
ADDRESS OF ICS/ICR-11 MAPPING TABLES
.ENABL

LSB

• REPT

12 .

.WORD

.ICTAB+<I.CTBL+N>

N=O
ICMAP:

N=N+l
.ENDR
;+
**-.RDIC-READ ICS/ICR-11 DIRECT ACCESS INPUTS
THIS SUBROUTINE IS CALLED TO TRANSLATE RELATIVE MODULE NUMBER
TO PHYSICAL EXTERNAL PAGE ADDRESS AND READ THE MODULE DATA.
INPUTS:
RO
Rl

RELATIVE MODULE NUMBER
MODULE CODE
WHERE:
0

CONTACT SENSE

* The definitions are included in module ICOM in the system library or
in the STB file ICOM.STB under UIC[l,l] on the system disk. The STB
file is automatically referenced by the Task Builder in response to
the use of the LIBR keyword.

14-73

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
1 = CONTACT INTERRUPTS
2 = COUNTERS
STACK SETUP IS AS
(SP)+OO
(SP)+02
(SP)+04
OUTPUTS:

FOLLOWS:
RETURN TO CALLER
I/O STATUS BLOCK ADDRESS (NOT REFERENCED).
PHYSICAL UNIT NUMBER

C/CLEAR
RO

MODULE DATA

C/SET:
NONEXISTENT PHYSICAL UNIT NUMBER OR MODULE SPECIFIED
;-

• RDIC: :
MOV
CMP
BLO
ASL

4(SP),R2
#I$$Cll-l ,R2
10$
R2

MOV

ICMAP(R2) ,R2

ASL
ADD

Rl
Rl,R2

ASL
ASL
ASL
ADD
TSTB
SEC
BEQ
INCB
CMPB
BLO
INC
ADD
CLR
BISB
ASL
MOV

Rl
Rl
Rl
Rl,R2
(R2)
10$
RO
(R2)+,RO
10$
R2
RO,R2
RO
{R2) ,RO
RO
.ICMD(RO) ,RO

GET PHYSICAL UNIT NUMBER
LEGAL UNIT NUMBER?
IF LO NO
CONVERT PHYSICAL UNIT NUMBER TO WORD
OFFSET
GET ADDRESS OF MAPPING TABLE
ENTRIES FOR THIS UNIT
CONVERT CODE TO WORD OFFSET
MULTIPLY OFFSET BY 9 AND ADD
TO TABLE ADDRESS

COMPUTE OFFSET TO TABLE
MODULE EXIST?
ASSUME NO
IF EQ NO
CONVERT TO NUMBER OF MODULES
LEGAL MOCULE NUMBER?
IF LO NO
POINT TO TABLE ENTRIES
OFFSET TO MODULE NUMBER
SET FOR MOVB WITHOUT SIGN EXTEND
GET INDEX TO MODULE
CONVERT TO WORD OFFSET
GET MODULE DATA

10$:
RETURN
.END

14.7

CONVERSION OF EXISTING SOFTWARE

The following paragraphs are
intended as guidance in converting
existing UDC or res software to run under the ICS/ICR-11 driver and
associated FORTRAN support routines. The differences described here
are restricted
to module support and features
that would affect
existing software.
New features, unsupported in previous systems, are
not discussed.

14-74

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.7.1

Features

Principal features affecting existing software are:
1.

Support for the ICS/ICR-11 as a multi-unit,
device.

Removal of software restrictions
functionally similar modules.

the

multi-controller
placement

Multi-unit support affects any software that addresses modules outside
the range of a single file box. In general, such software must be
modified at the source level.
Unrestricted module placement affects MACR0-11 programs that directly
access digital input and counter modules. Such programs may utilize
the library routine described in section 14.3 to read data from these
modules.

14.7.2

Module Support

14.7.2.1

IAD-IA A/D Converter and IMX-IA Multiplexer

MACRO Interface: Identical to UDCll driver
FORTRAN Interface: Same as UDCll
Functional Differences:
The ICS/ICR-11 driver can initiate parallel conversions on each IAD-IA
in a file box that is referenced by a single QIO request. The UDCll
driver performs all conversions serially.

The ICS/ICR-11 driver supports any permissible configuration of IAD-IA
A/D converters and IMX-IA multiplexers. The UDCll driver requires
that eight module slots be reserved for each IAD-IA in the system
regardless of the actual number of multiplexers installed.

14.7.2.2

16-Bit Binary Counter

MACRO Interface:

Identical to UDCll driver

FORTRAN Interface: Same as UDCll; however, if the counter is read
through a call to RDTI then the task must be relinked to incorporate
the revised FORTRAN Interface routine.
Functional differences:
The ICS/ICR-11 driver permits any task to reset an initial counter
value (via FORTRAN call RSTI or through the IO.RTI QIO function). The
UDCll driver restricts this operation to a task that has connected to
counter interrupts.

14.7.2.3

Bistable Digital Output

MACRO Interface: Identical to UDCll
FORTRAN Interface: Identical to UDCll
Functional Differences:

None
14-75

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS
14.7.2.4

Momentary Digital Output

MACRO Interface:
User interface is via the QIO IO.MSC issued to the ICS/ICR-11 driver.
The UDCll driver does not support this function since the module may
be accessed directly through the UDC device common block.
FORTRAN Interface:
Identical to UDCll; however, existing FORTRAN tasks must be
to include ICS/ICR-11 FORTRAN interface routines.

relinked

Functional Differences:
Momentary output operations are now processed by the
driver, rather than through direct access to the I/O page.
14.7.2.5

ICS/ICR-11

Noninterrupting Digital Input

MACRO Interface:
MACRO Interface is by means of the ICS/ICR-11 device common block
mapping table described in section 14.6.

and

FORTRAN Interface: Identical to UDCll; however, existing tasks must
be relinked to include revised ICS/ICR-11 FORTRAN interface routines.
Functional Differences:

14.7.2.6

None

Analog Output

MACRO Interface:
User interface is via the QIO IO.SAO issued to the ICS/ICR-11 driver.
The UDCll driver does not support this function since the module may
be accessed directly through the UDC device common block.
FORTRAN Interface:
Identical to UDCll; however, existing FORTRAN tasks must be
to include ICS/ICR-11 FORTRAN interface subroutines.

relinked

Functional Differences:
Analog output operations are now processed by the
rather than through direct access to the I/O page.
14.7.2.7

ICS/ICR-11

driver

Interrupting Digital Input

MACRO Interface:

Identical to UDCll driver

FORTRAN Interface:
Identical to UDCll driver;
however, if digital inputs are read
through the call to RCIPT then the task must be relinked to
incorporate the revised ICS/ICR-11 ~ORTRAN interface routines.
Functional Differences:

None.

14-76

APPENDIX A
SUMMARY OF I/O FUNCTIONS

This appendix summarizes legal I/O functions for all device drive
described
in this manual.
Both devices and functions are list
alphabetically. The meanings of the five parameters represented
the ellipsis
( ... ) are described in section 1.5.1. The meanings
the function-specific parameters shown below are discussed
in
t
appropriate driver chapters.
The user may reference these functio
symbolically by invoking
the system macros FILI0$
(standard I
functions)
and SPCIO$ (special I/O functions), or by allowing them
be defined at task build time from the system object library.

A.l

ANALOG-TO-DIGITAL CONVERTER DRIVERS

IO. KIL, .••

Cancel I/O requests

IO.RBC, ••. ,<stadd,size,stcnta> Initiate an A/D conversion

A.2

CARD READER DRIVER

IO.ATT, ...

Attach device

IO.DET, •••

Detach device

IO.KIL, ..•

Cancel I/O requests

IO.RDB, ... ,<stadd,size>

Read logical block (binary)

IO.RLB, ..• ,<stadd,size>

Read logical block (alphanumeric)

IO.RVB, ..• ,<stadd,size>

Read virtual block (alphanumeric)

A.3

CASSETTE DRIVER

IO. ATT, •..

Attach device

IO. DET, .••

Detach device

IO. EOF, •••

Write end-of-file gap

IO.KIL, ••.

Cancel I/O requests

IO.RLB, ... ,<stadd,size>

Read logical block

IO.RVB, ..• ,<stadd,size>

Read virtual block

A-1

SUMMARY OF I/0 FUNCTIONS
IO.RWD, .••

Rewind tape

IO.SPB, ... ,<nbs>

Space blocks

IO.SPF, •.. ,<nes>

Space files

IO.WLB, ..• ,<stadd,size>

Write logical block

IO.WVB, ••. ,<stadd,size>

Write virtual block

A.4

COMMUNICATION DRIVERS (MESSAGE-ORIENTED)

IO.ATT, ••.

Attach device

IO.DET, .••

Detach device

IO. FDX, •..

Set device to full duplex mode

IO.HDX, ••.

Set device to half-duplex mode

IO. !NL, •.•

Initialize device and set device
characteristics

IO.RLB, ... ,<stadd,size>

Read logical block, stripping
sync characters

IO.RNS, ..• ,<stadd,size>

Read logical block, transparent mode

IO.SYN, ••• ,<syn>

Specify sync character

IO. TRM, •..

Terminate communication, disconnecting
from physical channel

IO.WLB, .•• ,<stadd,size>

Write logical block with sync leader

IO.WNS, •.. ,<stadd,size>

Write logical block, no sync leader

A.5

DECTAPE DRIVER

IO.RLB, ... ,<stadd,size,,,lbn>

Read logical block (forward)

IO.RLV, ••. ,<stadd,size,,,lbn>

Read logical block (reverse)

IO.RVB, ..• ,<stadd,size,,,lbn>

Read virtual block (forward)

IO.WLB, ... ,<stadd,size,,,lbn>

Write logical block (forward)

IO.WLV, ••. ,<stadd,size,,,lbn>

Write logical block (reverse)

IO.WVB, •.. ,<stadd,size,,,lbn>

Write virtual block (forward)

A.6

DISK DRIVER

IO.RLB, ... ,<stadd,size,,blkh,blkl>

Read logical block

IO.RPB, •.. ,<stadd,size,,,pbn>

Read physical block

A-2

SUMMARY OF I/O FUNCTIONS
IO.RVB, •.. ,<stadd,size,,blkh,blkl>

Read virtual block

IO.WDD, •.. ,<stadd,size,,,pbn>
deleted

Write

physical

block

(with

data mark)
IO.WLB, ... ,<stadd,size,,blkh,blkl>

Write logical block

IO.WPB, ... ,<stadd,size,,,pbn>

Write physical block

IO.WVB, •.. ,<stadd,size,,blkh,blkl>

Write virtual block

A.7

INDUSTRIAL CONTROL LOCAL AND REMOTE SUBSYSTEMS

IO.CCI, •.• <stadd,sizb,tevf>

Connect a
interrupts

buffer

digital

IO.CTI, ... <stadd,sizb,tevf,arv>

Connect a
interrupts

buffer

counter

IO.CTY, ..• <stadd,sizb,tevf>

Connect a
interrupts

buffer

terminal

IO. DC I

Disconnect a buffer from digital
interrupts

•••

IO. DT I, .••

Disconnect a buffer from
interrupts

IO. DTY, ••.

Disconnect a buffer from terminal
interrupts

IO. FLN I • • •

Set controller offline

IO.ITI, .•. <mn,ic>

Initialize a counter

IO. LDI, .•• < tname,, [ tevf] , pn ,csm>

Link task to digital interrupts

IO.LKE, ... <tname,, [tevf]>

Link task to error interrupts

I 0 . LT I , • . . < t name , , [ t e v f ] , c n , [ a r v ] >

Link task to counter interrupts

IO.LTY, ..• <tname,,[tevf]>

Link task
interrupts

IO.MLO, ... <opn,pp,dp>

Open or close
output points

remote

terminal

bistable

digital

IO.MSO, ... <opn,dp>

Pulse single-shot digital
points

output

IO.NLK, .•. <tname>

Unlink a task from all interrupts

IO. NLN, .••

Place ICR controller online

IO.RAD, ... <stadd>

Read activating data

IO.RBC, •.. <stadd,size,stcnta>

Initiate multiple A/D conversions

A-3

counter

SUMMARY OF I/O FUNCTIONS
IO.SAO, ... <chn,vout>

Perform analog output

IO.UDI, ..• <tname>

Unlink a
interrupts

task

from

digital

IO.UER, ... <tname>

Unlink
a
interrupts

task

from

error

IO.UTI, ..• <tname>

Unlink a
interrupts

task

from

counter

IO.UTY, ... <tname>

Unlink a task
interrupts

from

terminal

IO.WLB, ... <staddb,sizb>

Transmit data to the
terminal

A.8

!CR

LABORATORY PERIPHERAL SYSTEMS DRIVERS

IO.ADS, ... ,<stadd,size,pnt,
ticks,bufs,chna>

Perform A/D sampling

IO.HIS, ... ,<stadd,size,pnt,
ticks,bufs>

Perform histogram sampling

IO.KIL,...

Cancel I/O requests

IO.LED, ... ,<int,num>

Display number in LED lights

IO.MDA, ... ,<stadd,size,pnt,
ticks,bufs,chnd>

Perform D/A output

IO.MDI, ... ,<stadd,size,pnt,
ticks,bufs,mask>

Perform digital input sampling

IO.MDO, ... ,<stadd,size,pnt,
ticks,bufs,mask>

Perform digital output

IO.REL, ••. ,<rel,pol>

Latch output relay

IO.SDI, ••• ,<mask>

Read digital input register

IO.SDO, .•. ,<mask,data>

Write digital output register

IO.STP, ..• ,<stadd>

Stop in-progress request

A.9

LINE PRINTER DRIVER

IO.ATT, ..•

Attach device

IO.DET, .•.

Detach device

IO. KIL I • • •

Cancel I/O requests

IO.WLB, ... ,<stadd,size,vfc>

Write logical block

IO.WVB, ... ,<stadd,size,vfc>

Write virtual block

A-4

remote

SUMMARY OF I/O FUNCTIONS
A.10

MAGNETIC TAPE DRIVER

IO.ATT, •••

Attach device

IO.DET, .••

Detach device

IO.EOF, •••

Write end-of-file (tape mark)

IO.KIL, •••

Cancel I/O requests

IO.RLB, •.. ,<stadd,size>

Read logical block

IO.RVB, •.• ,<stadd,size>

Read virtual block

IO.RWD, •••

Rewind tape

IO. RWU, •••

Rewind and turn unit off-line

IO.SEC, •..

Read tape characteristics

IO.SMO, ••• ,<cb>

Mount tape and set tape characteristics

IO.SPB, •.• ,<nbs>

Space blocks

IO.SPF, ••• ,<nes>

Space files

-1-'

J.V.i:>J.\...1 • • • ,,l,.;U/

Set tape characteristics

IO.WLB, ••• ,<stadd,size>

Write logical block

IO.WVB, ••. ,<stadd,size>

Write virtual block

Tr\

rtmrt

A.11

PAPER TAPE READER/PUNCH DRIVERS

IO.ATT, •••

Attach device

IO.DET, •••

Detach device

IO. KIL, ••.

Cancel I/O Requests

IO.RLB, ••. ,<stadd,size>

Read logical block (reader only)

IO.RVB, ••. ,<stadd,size>

Read virtual block (reader only)

IO.WLB, •.• ,<stadd,size>

Write logical block (punch only)

IO.WVB, •.. ,<stadd,size>

Write virtual block (punch only)

A.12

TERMINAL DRIVER

IO.ATT, •.•

Attach device

IO. DET, •••

Detach device

IO.KIL, •••

Cancel I/O requests

IO.RAL, ... ,<stadd ,size>

Read logical block and pass all bits

A-5

SUMMARY OF I/O FUNCTIONS
IO.RLB, •.. ,<stadd,size>

Read logical block

IO.RVB, ... ,<stadd,size>

Read virtual block

IO.WAL, ..• ,<stadd,size>

Write logical block and pass all bits

IO.WLB, .•• ,<stadd,size,vfc>

Write logical block

IO.WVB, ..• ,<stadd,size,vfc>

Write virtual block

A.13

UNIVERSAL DIGITAL CONTROLLER DRIVER

IO.CCI, •.. ,<stadd,sizb,tevf>

Connect a buffer to contact
interrupts

IO.CTI, ••• ,<stadd,sizb,tevf,arv>

Connect a buffer to timer
interrupts

IO. DC I, •••

Disconnect a buffer from contact
interrupt

IO.DTI, •••

Disconnect a buffer from timer
interrupts

IO. !TI, .•. , <mn, ic>

Initialize a timer

IO.KIL, .••

Cancel I/O requests

IO.MLO, ••• ,<opn,pp,dp>

Open or close latching digital
output points

IO.RBC, ••• ,<stadd,size,stcnta>

Initiate multiple A/D conversions

A-6

APPENDIX B
I/O FUNCTION AND STATUS CODES

This appendix lists the numeric codes for all I/O functions, directive
status
returns, and I/0 completion status returns.
Lists are
organized in the following sequence:
I/O completion status codes
Directive status codes
Device-independent I/O function codes
Device-dependent I/O function codes
Device-dependent function codes are listed by device.
and codes are organized in alphabetical order.

Both

devices

For each code, the symbolic name is listed in form IO.xxx, IE.xxx, or
IS.xxx.
A brief description of the error or function is also
included. Both decimal and octal values are provided for all codes.
B.l

I/0 STATUS CODES

This section lists error and success codes which can be returned in
the I/O status block on completion of an I/O function. Ine codes
below may be referenced symbolically by invoking the system macro
IOERR$.

B.1.1

I/O Status Error Codes
Decimal

Octal

Meaning

IE .ABO

-15

177761

Operation aborted

IE.ALN

-34

177736

File already open

IE. BAD

-01

177777

Bad parameter

IE. BBE

-56

177710

Bad block

IE. BLK

-20

177754

Illegal block number

IE. BYT

-19

177755

Byte-ligned buffer specified

IE. CON

-22

177752

UDC connect error

Name

B-1

I/0 FUNCTION AND STATUS CODES
Name

Decimal

Octal

Meaning

IE. DAA

-08

177770

Device already attached

IE. DAO

-13

177763

Data overrun

IE. DNA

-07

177771

Device not attached

IE. DNR

-03

177775

Device not ready

IE. EOF

-10

177766

End-of-file encountered

IE. EOT

-62

177702

End-of-tape encountered

IE. EOV

-11

177765

End-of-volume encountered

IE.FHE

-59

177705

Fatal hardware error

IE. FLN
-81
of fl ine

177657

I CS/ I CR

IE. IFC

-2

177776

Illegal function

IE.MOD

-21

177753

Invalid UDC or ICS/ICR module

IE .NLK

-79

177661

Task not linked to specified
ICS/ICR interrupts

IE .NLN

-37

177733

File not open

IE.NOD

-23

177751

No dynamic memory available
to allocate a secondary
control block

IE.NST

-80

177660

Task specified in ICS/ICR
Link or Unlink request is not
installed

IE.OFL

-65

177677

Device off-line

IE.GNP

-05

177773

Illegal subfunction

IE .OVR

-18

177756

Illegal read overlay request

IE.PR!

-16

177760

Privilege violation

IE.RSU

-17

177757

Nonsharable resource in use

IE. SPC

-06

177772

Illegal address space

IE. VER

-04

177774

Unrecoverable error

IE. WLK

-12

177764

Write-locked device

B-2

controller

already

I/O FUNCTION AND STATUS CODES
B. l. 2

B.2

I/O Status Success Codes
Octal

Meaning

1
15

006401

Successful completion with
carriage return

1
33

015401

Successful completion
with ESCape

+00

000000

I/O request pending

IS. RDD

+02

000002

Deleted data mark read

rs.sue

+01

000001

Successful completion

Name

Decimal

IS.CR

Byte 0:
Byte 1:

IS.ESC

Byte 0:
Byte 1:

IS. PND

DIRECTIVES CODES

This section lists error and success codes which can
the directive status word at symbolic location
directive is issued.

LJ.

be returned
in
$DSW when a QIO

, Directive Error Codes

• .J..

Name

Decimal

Octal

Meaning

IE .ADP

-98

177636

Invalid address

IE. IEF

-97

177637

Invalid event

IE. ILU

-96

177640

Invalid logical unit number

IE. SDP

-99

177635

Invalid DIC number or DPB size

IE. ULN

-05

177773

Unassigned LUN

IE. UPN

-01

177777

Insufficient dynamic storage

-F 1 .,, ,...

.L-LU';:!

,..... ,,Tn h,.... ....

11\,..lJUUC:.L

B.2.2 Directive Success Codes

B.3

Name

Decimal

Octal

Meaning

rs.sue

+01

000001

Directive accepted

I/0 FUNCTION CODES

This section lists codes for all
functions.

standard

B-3

and

device-dependent

I/O

I/O FUNCTION AND STATUS CODES
B.3.1 Standard I/O Fune tion Codes
Name

Octal Words

IO.ATT

001400

Attach device

IO.DET

002000

Detach device

IO. KIL

000012

Cancel I/O requests

IO.RLB

001000

Read logical block

IO.RVB

010400

Read virtual block

IO.WLB

000400

Write logical block

IO.WVB

011000

Write virtual block

Octal Bytes
Code Subcode

Meaning

B.3.2

Specific A/D Converter I/O Function Codes

Name

Octal Words

IO.RBC

003000

Octal Bytes
Code Subcode
6

Meaning
Initiate an A/D conversion

B.3.3 Specific Card Reader I/O Function Codes
Name

Octal Words

IO.RDS

001200

Octal Bytes
Code Subcode

200

Meaning
Read logical block (binary)

B.3.4 Specific Cassette I/O Function Codes
Name

Octal Words

Octal Bytes
Code Subcode

IO.EOF

003000

Write end-of-file gap

IO.RWD

002400

Rewind tape

IO. SPB

002420

Space blocks

IO. SPF

002440

Space files

B-4

Meaning

I/O FUNCTION AND STATUS CODES
B.3.5 Specific Communication (Message-Oriented)

Name

Octal Words

Octal Bytes
Subcode
"'"~-

I/0 Function Codes

Meaning

'-VU'C

IO. FOX

003020

Set device to full-duplex mode

IO. HDX

003010

Set device to half-duplex mode

IO. INL

002400

Initialize device and
set device characteristics

IO. RNS

001020

Read logical block,
transparent mode

IO. SYN

003040

Specify sync character

IO.TRM

002310

Terminate communication,
d isconnec t.ing from physical
channel

IO.WNS

000420

Write logical block with no
sync leader

/r"\

Function Codes

UeJeU

Specific DECtape

Name

Octal Words

IO.RLV
IO.WLV

J./U

Octal Bytes
Code Subcode

Meaning

001100

100

Read logical block (reverse)

000500

100

Write logical block (reverse)

B.3.7

Specific Disk I/O Function Codes (RXOl)

Name

Octal Words

IO. RPB
IO. WPB

Octal Bytes
Code Subcode

Meaning

001040

Read physical block

000440

Write physical block

B.3.8

Specific ICS/ICR I/O Function Codes

Name

Octal Words

IO. CCI

Octal Bytes
Code Subcode

Meaning

014000

Connect a buff er to digital
interrupt input

IO.CTI

015400

Connect a counter

IO.CTY

003400

Connect a remote terminal

B-5

I/O FUNCTION AND STATUS CODES
IO. DCI

014400

Disconnect a buffer from
digital interrupt input

IO.DTI

016000

Disconnect a buffer from
counter input

IO. DTY

006400

Disconnect a buffer from
terminal input

IO. FLN

012400

Place selected unit offline

IO. ITI

017000

Initialize a counter

IO.LOI

007000

Link a task to digital
interrupts

IO.LKE

012000

Link a task to error
interrupts

IO. LTI

007400

Link a task to counter
interrupts

IO.LTY

010000

Link a task to terminal
interrupts

IO.MLO

006000

Open or close bistable digital
output points

IO.MSG

005000

Pulse single-shot digital
output points

IO.NLK

011400

Unlink a task from all
unsolicited interrupts

IO. NLN

017400

Place selected unit online

IO. RAD

010400

Read task activation data

IO.RBC

003000

Initiate multiple A/D
conversions

IO.SAO

004000

Perform analog output to
specified channel

IO.UDI

011410

Unlink a task from digital
interrupts

IO.UER

011440

Unlink a task from error
interrupts

IO. UTI

011420

Unlink a task from
counter interrupts

IO. UTY

011430

Unlink a task from
terminal interrupts

IO. WLB

000400

Output to remote terminal

B-6

I/O FUNCTION AND STATUS CODES
B.3.9 Spec i f ic LPS I/0 Function Codes
Octal Bytes
Code Subcode

Meaning

014000

Initialize A/D sampling

IO. HIS

015000

Initialize histogram sampling

IO. LED

012000

Display number in LED lights

IO.MDA

016000

Initialize D/A output

IO.MDI

014400

Initialize digital input
sampling

IO.MOO

015400

Initialize digital output

IO.REL

013400

Latch output relay

IO. SDI

013000

Read dig ital input register

IO. SDO

012400

Write digital output register

IO. STP

016400

Stop in-progress request

Name

Octal Words

IO.ADS

B.3.10 s,eec i f ic Magta:ee I/O Function Codes
Name

Octal Words

IO. EOF

Octal Bytes
Code Subcode

Meaning

003000

Write end-of-file gap

IO.RWD

002400

Rewind tape

IO. RWU

002540

140

Rewind and unload

IO.SEC

002520

120

Sense characteristics

IO.SMO

002560

160

Mount and set characteristics

IO. SPB

002420

Space blocks

IO.SPF

002440

Space files

IO. STC

002500

100

Set characteristics

B.3.11

Specific Terminal I/O Function Codes

Name

Octal Words

Octal Bytes
Code Subcode

Meaning

IO. RAL

001010

Read pass all bits

IO.WAL

000410

Write pass all bits

B-7

I/0 FUNCTION AND STATUS CODES
B.3.12 Specific UDC I/O Function Codes
Name

Octal Words

Octal Bytes
Code Subcode

Meaning

IO. CCI

014000

Connect a buffer to contact
interrupt digital input

IO.CTI

015400

Connect a timer

IO. DCI

014400

Disconnect a buffer from
contact interrupt digital
input

IO.OT!

016000

Disconnect a timer

IO. !TI

017000

Initialize a timer

IO.MLO

006000

Open or close latching
digital output points

IO.RBC

003000

Initiate multiple
A/D conversions

B-8

APPENDIX C
RSX-llM PROGRAMMING EXAMPLE

MESSAGE TRANSFER

MACRO M0710

10•0CT•74 10119

PAGE

.TITLE MESSAGE TRANSFER
.IDENT 1011

1
2
3
4

1974,

DIGITAL

EQUIPMENT

CORP,,

MAYNARD,

MASS,

THIS SOFTWARE IS FURNISHED TO PURCHASER UNDER A LICENSE FOR USE
ON A SINGLE COMPUTER SYSTEM AND CAN BE COPIED Cill!Tt! INCLUSION
OF DEC'S COPYRIGHT NOTICE) ONLY FOR USE IN SUCH SYSTEM, EXCE 0 T
AS HAY OTHERwISE BE PROVIDED IN WRITING BY DEC.

7
6
9

10
11
12

THE INFORMATION JN THIS DOCUMENT IS SUBJECT TO CHANGE WtTHOUT
NOTICE AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL
E~UIPMENT COROORATION.

13
14
15

DEC ~SSUHES NO ~ES~ONSIBILitY FOR THE
OF ITS SOFT~ARE ON EQUI?MENT WHICH IS

1b
17
18
19

USE OR RELIABILITY
NOT SJPPLifD BY DEC,

'IERSION 01

20
21

EARL WALDIN

22
23

DEMONSTRATION OF USE OF RSX•11M I/O

MACRO LIBRARY CALLS

27
28

.MCALL

ALUNSS,QIOSS,WTSE$S,WSIGSS

29
30
31
32
33
34

35
3b
37

38
39

40
!Jl
Ll2
Ll3
LILI

r LOCALLY DEFINED MACROS
• "1ACRO
JSR
.ENDM

CALL SUBR
?C,SJBR

rDEFINITION FOR SUBROUTINE CALLS

.MACRO
RTS
.ENDM

RETLJRN
PC

JSUBROJTI~E

LOCAL DAU

45
!ib

C-1

RETURN ~ACRO

RSX-llM PROGRAMMING EXAMPLE
• REAo-im.ATED STORAGE

tH
qs
q9

50 000000
51 000002

52
53
54
55
Sb
57 000004

000000
00000121

.WORD
.itlORD

rREAD STATUS BLOCK

0
0

l'IR ITE•RELATED STO~AGE
000000

MESSAGE TRANSFER
58 000006

RDSTSI

wRSTSs

.WORD

MACRO Ml1l710
00012100

JWRITE STATUS BLOCK

l1l

10•0CT•74 10119

.WORD

PAGE 1•1

b0
BUFFER STORAGE

b2
bl

&4 000010
b5 000132

BUF11
BUF21

.BLKB
.BLKB

82,
82,

rBUFFER 1
sBUFFER 2

&8
b9

I
I

**•SXFER•DEMONSTRATE USE OF RSX•11M I/0 BY OUTPUTTING RECORDS
FROM TII CUSER'S TE~MINAL) TO LINE PRINTER. REQUESTS ARE DOUB~E
r BUFFERED TO DEMONST~ATE HOW OPERATIONS MAY BE OVE~LAPPED.

70
71
72

••

73 000254

SXFERll

I
I

74 000274
75

77
78 000314
79 0003bb
80 000370
81 000374

103003

A~UNSS

ALUNIS

#1,#•TI,#0
#2,#•L.P

tLUN 1 IS Tll DEVICE
sLUN 2 IS LP01

READ A LINE FROM INOUT DEVICE, LUN

•
1011

000747

QIOSS
BCC
CALL
BR

20S
STCHK
101

Jlf DISPATCHED OK, CONTINUE
SCHECK STATUS
Jlf RECOVE~ABLE ERROR, TRY AGAIN

WTSESS

C~PB

ROSTS,#IS.SUC

JWAIT UNTIL 1 COMPLETE
rREAD SUCCESSFUL?

BEQ
JMP

30$

•IO.R~B,#1,#1,,#RDSTS,,<#BUf1,#80.>

83 00037&
84 000410

es 00041&
Sb 000420

2051
12b727
177304
0000006
001402
0001&7

1001

JCONTINUE IF SUCCESSfUL
sTERMINATE IF NOT SUCCESSFUL

000440

88 000424

0lb701
177352

30Ss

sGET ACTUAL BYTE COUNT IN R1

MOV

J BEGIN TO FILL SECOND BUFFER

91
92 000430

93 000502
94 01210504

103003

95 000510

000747

40S I

QIOSS
BCC
CALL
BR

#IO.RL.8,#1,#2,,#RDSTS,,<#BUF2,#80.>
50S
1CONTINUE IF DISPATCH OK
STCHK
SCHECK STATUS
411lS
JTRY AGAIN

97
98

START BJFFER 1 OUT

99
100
1 u 000512

102 0005b4
133 0005bb
104 000572
135
111lb
137
138

50SI

103003

QIOSS
BCC
CAL~

000747

#IO.~L.B1#2,j1,,#~RSTS 1 ,<#BJf1,R1r#40>

b0S
STCHK
50S

tCONTINUE IF NO DISPATCH ERROR
SCHECK STATUS
JTRY AGAIN

IS A SYNCHRONIZATION POINT. BOTH FUNCTIONS MJST CO~PLETE
BEFORE ANYTHING ELSE BEGINS.

T~IS

1"9
110

C-2

RSX-llM PROGRAMMING EXAMPLE
MESSAGE TRANSFER
111 000574
112 000606
113 000&14
114 000616
115 000&22
11b 000634

00$1
126727
111106

00000121G
001123
016702
177100
126727

WTSESS
CMPB

#2
~OSTS,#IS.SUC

SWAIT FOR 2 TO FILL
sSUCCESSfUL?

BNE
MOV

100$
ROSTS+2,R2

J IF NOT, CUSH
sGET COUNT FOR BUFFER 2

WTSESS

C~PB

~RSTS,#IS.SUC

JWAIT FOR 1 TO EMPTY
JSUCCESSFUL?

B~E

100$

sIF NOT, CRASH

177144
000000G
117 0~0642
118
119
120
121 000644
122 000716
123 000720

001110

124 000724

000747

125
12b 000726
127 001000
128 001002
129 00100&

103003

QIOSS
BCC
CALL

000747

FILL BUFFER 1, EMPTY BUFFER 2

7051
103003

GlIOSS
BCC
CALL
B~

80SI

#IO.RLB,#1,#1,,*RDSTS,,<#BUF1,#60.>
80S
JIF OK, CONTINUE
STCHK
J CHECK STATUS
70$
rTRY AGAIN
#IO.wLB,#2,#2,,*~RSTS,,<#BUF2,R21#40>

90S
STCHK
80S

JCONTINUE If SUCCESSFUL
rCHECK ST~TUS IF NOT SUCCESSFUL
rRETURN

130

131

THIS !S ALSO A SY~CHRON!ZAT!O~ POINT

132

133
134

135 001010
136 001022

90Ss
126727
17b752

137 001030

000000G
001015

138 001032

01b701
176744

139 00103&
140 001050

126727

WTSESS
CMPB

BNE
MOV

100!
RDSTS+2,R1

JIF NOT,

WTSESS

C~PB

~RSTS,#IS.SUC

JW4IT FOR BUFFER 2 TO EMPTY
JSUCCESSFUL?

~DSTS,•Is.suc

SWAIT FOR 1 TO FILL
JSUCCESSFUL?
CRAS~l

JGET ACTUAL BYTE COJNT IN R1

176730
000000G

141 00105&

001002

142 001060

000167

BNE

100$

JMP

40$

JTERMINATE IF NOT
JBACK INTO LOOP

177344

143

144
145
14b
147
148 001064

DON'T ATTEMPT TO RECOVER ERRORS
000004

1005 a

149
150
151

152
153
154

155
150

JCRASH TASK

!OT

s **• STCHK • ATTEMPT TO RECOVE~ DIRECTIVE DISPATCH ERROR ONLY IF
J IT INVOLVES DYNAMIC MEMORY ALLOCATION • OTHERWISE TERMINATE.
INPUTS I
CSP):~ETURN

C-3

ADDRESS

RSX-llM PROGRAMMING EXAMPLE
MACRO M071 Ill

MESSAGE TRANSFER
157

10•0CT•74 U11q

?AGE 1•3

OUTPUTS I

158
159
1&0
1 &1
1&2
1&3 0111111106
164 0111111174
1&5 fdlll111176
1&6 01111104
167
1&8 1111111106

126 727

·-

STC~l<t

111111001110G
11100000G
0111101114

111111001114

10St

16q
171il

NONE
CM?B

SOSW, UE.UPN

JBUFFER ALLOCATION F'AILURE?

BNE
wSIGSS
RETURN

10$

JIF NOT TERMINATE
JAWAIT SIGNIFICANT EVENT
JTRY AGAIN

.END

0111111254'

r CRASH TASK

IOT
SXFER

MESSAGE TRAt..ISFER
TABLE

SYMBOL

01110010R

BJF1

ABS.

E~RORS

F~EE

~OqE1

WRSTS
111111011104R
SDSW : ****** GX

STCHK

USARG: 11101111111112

~OSTS

1111111110111111
0111111111

nETECTED:

****** GX
Is.sue= ****** GX

IO.~LB=

11100132R
BJF2
IE°.J~N: ****** GX
IO.~LB: ****** GX

111211101110R
0010bbR

00111
0131
0

3586. wORDS

.~SGILitTTM:MSG.01111

C-4

SXFER

2'0111251.!RG

APPENDIX D
GLOSSARY OF RSX-llM TERMS

ASYNCHRONOUS SYSTEM TRAP (AST)
A system condition which occurs as a result of an external
significant event such as completion of an I/O request. On
occurrence of the significant event, control passes to an AST
service routine, and the AST is added to an Executive
first-in first-out queue for the task in which the service
routine appears.
ATTACH
Dedicate a physical device unit for exclusive use by the task
that requested attachment. Once the physical device unit has
been attached by a task, using an IO.ATT I/O function, only
that task can free the unit again for use by other tasks in
the system. Attachment request attempted to a device unit
already attached by another task will not be terminated until
the attachment request can finally be honored;
in other
words, the attachment request is terminated only when the
previous attachment is terminated, and no higher priority
attachment requests are queued.
DETACH
Free an attached physical device unit for use by tasks other
than the one that attached it. A physical device unit can
only be detached, by means of an IO.DET I/O function, by the
task that attached it, or by the Executive if the task is
terminated with the device still attached.
DIRECTIVE
A type of system meta-instruction which is used to provide a
facility inherent in the hardware by means of executive
requests issued to the RSX-llM Executive.
Directives are
usually invoked by means of execution of expanded code from
macros in the System Macro Library (RSXMAC.SML).
EVENT FLAG NUMBER
A number which can be specified in a QIO or other macro call
to indicate to the issuing task which significant event has

D-1

GLOSSARY OF RSX-llM TERMS
occurred. There are 64 event flags available in RSX-llM.
Flags numbered 1 through 32 are local to a task; 33 through
64 are common to all tasks.
Flags 25 through 32 and 57
through 64 are normally reserved for RSX-llM system software
use. Each of the available flags can be referenced by number
and can be used for communication and synchronization between
user tasks, or between tasks and executive service requests,
including I/O requests.
I/O STATUS BLOCK
A 2-word array (double-word) in which a code representing the
final status of an I/O request is returned, if the address of
the block is specified in the QIO directive parameter block
(DPB) which generated the request. A code identifying the
type of success or error is returned in the low-order byte of
the first word, optional device-dependent information in the
high-order byte of the first word, and the number of bytes
transferred on a read or write in the second word of the
block. Although the I/O Status Block is optional, it is the
only way a user can guarantee that he will know the outcome
of an I/O request.
LOGICAL ADDRESS
A logical address is a software representation of a hardware
address.
The use of the phrase "logical address" implies
that some mapping occurs between "true", or hardware address
and "artificial" or logical address. The reason for using
logical addresses is that they simplify the way one deals
with a hardware device or family of devices.
The logical address in RSX-llM refers to the relative
position of a logical block on a volume. The volume is
divided into logical blocks, each of which is assigned an
address called a logical block number
(LBN).
All mass
storage media are accessed by LBN (e.g., 17) rather than
physical address (e.g., cylinder 5, track 3, sector 7).
LOGICAL BLOCK
A logical block in RSX-llM parlance refers to 512 bytes of
storage which may be considered to be a discrete entity for
logical purposes. In fact, it might be composed of odd-sized
fragments of non-contiguous storage.
Actually, a logical
block generally refers to one or more physical blocks of a
formatted or block-structured mass memory which compose the
logical atom for access to the medium.
Logical block may
also refer to the in-core image of a logical block which is
or will be on a mass storage device.
The concept of logical block is useful on file-structured
devices, in that all such devices appear to share all these
characteristics except total number of blocks.

D-2

GLOSSARY OF RSX-llM TERMS
LOGICAL BLOCK NUMBER (LBN)
Sequential position of a logical nLock w1tn respect
to a
collection of such blocks (which may compose a volume).
If
the collection of blocks had been written in logical order on
a sequential medium, such as magnetic tape, the logical block
number for any block would be the true position of the block
on that medium, e.g., logical block 2 would be encountered
just after LBN 1 and just before LBN 3.
LOGICAL UNIT NUMBER (LUN)
A number associated with a physical device unit during a
task's I/O operations.
Each task in the system can establish
its own correspondence between LUNs and physical device
units.
MACRO
A system capability which allows a user to generate Assembler
instructions, data, or symbols in a predetermined format by
providing actual arguments to the Assembler in a macro call
included
in
a
MACR0-11
program.
Macros provide a
standardized means of obtaining access to system services or
resources by invocations from programs.
PRIORITY

A number associated with an RSX-llM task which
relative position of that task among all tasks in the system.
The priority is associated with a task at task build time and
may be changed at install or run time.
Legal priorities are
in range 1 through 250, with greater magnitude indicating
higher priority.
If two
tasks are identical in every way
(i.e.,
resources used, etc.)
except priority
and
are
initiated at the same time, the task with the higher priority
will complete first.
I/O requests issued by a
task assume
the priority of that task and are honored according to the
task's priority.
SIGNIFICANT EVENT
An event or condition which indicates a change in system
status.
In RSX-llM, a significant event is declared when an
I/O operation completes and in some other cases as well.
A
declaration
of a significant event indicates that the
Executive should review the eligibility of all tasks
in
the
system to determine which task should run next, since the
significant event might unblock the execution of a higher
priority task.
SYNCHRONOUS SYSTEM TRAP (SST)
A system condition which occurs as a result of an error or
fault within
the executing
task.
If the same instruction
sequence were repeated,
the same synchronous trap would
occur.
On recognition of a synchronous trap, control passes
to an SST service routine.
SSTs are not handled directly by
the Executive as ASTs are.
D-3

GLOSSARY OF RSX-llM TERMS

VIRTUAL ADDRESS
A number which indicates relative
position
within
a
collection of logically-related granules of a storage medium.
The fact that the medium itself may be virtual (e.g., 1
million bytes of addressable memory, but only 64K in core
memory, the remainder on mass storage)
is
of
little
consequence~
in
fact,
the
ability to deal with a
hierarchical or multi-level memory as if it were one medium
is one of the principal advantages of systems supporting
virtual addressing. In RSX-llM, virtual address generally
refers to relative position within a task image, while
VIRTUAL BLOCK NUMBER (VBN) refers to relative position within
a file.
VIRTUAL BLOCK
One of a collection of blocks which make up a user file (or
the core image of that file). The block is virtual only in
that its address (VBN) refers to position within a file
regardless of the file's allocation or placement on a storage
medium. When a user accesses a file, he can think of the
file as a virtual storage medium belonging to him. Virtual
addressing within that file could be considered to be
absolute addressing on a virtual medium.

D-4

INDEX

Aborting a task, 4-7, 5-11,
7-6, 8-11
Activating a task by unsolicited
interrupts, 14-20, 14-60
ADOl-D analog to digital converter, 10-1
ADOl-D conversions, restricting
the number of, 10-10
A/D conversion control word,
10-3, 14-35
A/D converter I/O function

codes, specific, B-4
A/D converter status returns,
10-7, 14-11
A/D functional capabilities,
10-9
A/D gain ranges, use of, 10-9
A/D programming hints, 10-9
A/D value, switch gain, 12-13
Address assignments for
ICS/ICR: 14-1
AFCll analog-to-digital
converter, 10-1, 10-9
Alphanumeric format (026 and
029), 8-9
ALUN$ macro, 1-15
Analog data, input of, 10-5,
11-17, 14-10, 14-35, 14-75
Analog input channels, reading
sequential, 10-5, 11-17,
14-38
Analog output, performing, 11-18
14-12, 14-40, 14-76
Analog-to-digital converters,
10=1, 11=8, 14-75
ASR-33/35 Teletypes, 2-2
Assembly language interface,
14-6
Assembly procedure for
UDCOM.MAC, 11-9
Assigning a LUN, 1-15
to ADOl-D, 10-6
to AFCll, 10-6
to ARO, 12-13
to the ICS/ICR, 14-34
to LSO, 12-13
to the UDCll, 11-18
AST service, terminating, 1-19
ASTX$$ macro, 1-19
Asynchronous line interface,
DLll-E, 9-2
Asynchronous process control
I/O, synchronous and, 10-3,
11-14, 14-31
Attaching to an I/O device, 1-21

Binary format, 8-10
Bistable digital output, 14-13
Block length, 6-7
Block reading,
logical, 1-22
virtual, 1-23
Block size, 5-11
Block writing,
logical, 1-23
virtual, 1-23
Buffer, circular, 11-30
Buffer management, 12-32
Buffer pointers, adjusting, 12-12
Buffer, reading data from an
input, 12-19
Buffers, control and data, 10-9

Cancelling I/O requests, 1-22
Card input errors and recovery,
8-3
Card limitation, input, 8-10
Card reader check recovery,
ready and, 8-6
Card reader data formats, 8-9
Cassette I/O function codes,
specific, B-4
Cassette recovery procedures, 6-5
Cassette tape, structure of, 6-5
Channel numbers on the AFCll,
identical, 10-9
Channel, reading a single A/D,
12-11
Channels, reading sequential
analog input, 11-17, 14-38
Characters, control, 2-7, 8-8
Checkpointable tasks, 11-29
Circular buffer, 11-30
Clock and sampling rates, 12-31
Code conversion, ESCape, 2-11
Codes, directive, B-3
Codes, I/O function, B-3
Codes, return, 1-25
Codes, specific communication I/O
function, B-5
Common block, 11-3, 11-9, 11-11,
14-69
Communication I/O function codes,
specific, B-5
Communications drivers programming
example, 9-10
Communications drivers programming
hints, 9-8
Conditions, directive, 1-26
Conditions, I/O status, 1-27
Connecting to contact interrupts,
11-19

Index-1

INDEX (Cont.)
Connecting to counter module
interrupts, 14-17, 14-52
Connecting to digital interrupts,
14-16, 14-48
Connecting to terminal
interrupts, 14-19, 14-56
Connecting to timer interrupts,
11-19
Contact interrupt data, reading,
11-22, 11-23, 11-24, 11-26
Contact interrupt digital input,
11-5
Contact interrupts, connecting
tO I 11-19
Contact interrupts, disconnecting from, 11-20
Contact sense fields, reading
several, 11-21
Control and data buffers, 10-9
Control characters, 2-7, 8-8
Control word, A/D conversion,
10-3, 14-36
Converter, analog-to-digital,
11-7, 11-8
Counter routines, miscellaneous,
14-54
Counter, setting initial value,
14-18, 14-54

Data formats, card reader, 8-9
DECtape I/O function codes,
specific, B-5
DECtape recovery procedures, 4-5
DECtape transfers, 4-6
DECwriters, 2-2
Default logical and physical
units for ICS/ICR, assigning
(ASICLN/ASUDLN) , 14-34
Detaching from an I/O device,
1-22
Device, attaching to an I/O,
1-21
Device, detaching from an I/O,
1-22
Device-specific QIO function,
2-3, 3-4, 4-2, 5-3, 6-2,
8-2, 9-5, 10-2, 11-4, 12-2
Device specific QIO functions
(synchronous) , 12-4
Devices, RSX-llM, 1-2
DHll asynchronous serial line
multiplexer, 2-10
Digital input, 14-16, 14-43,
14-76
Digital output, bistable
fields, 14-13, 14-42, 14-75
momentary, 14-45, 14-75
single-shot, 14-13

Digital sense, 14-43
Direct access, 11-3, 11-8, 14-69
Directive codes, B-3
Directive conditions, 1-26
Directive, .MCALL, 1-14
Directive parameter blocks, 1-11
Disconnecting from contact
interrupts, 11-20
Disconnecting from counter
interrupts, 14-19, 14-55
Disconnecting from digital
interrupts, 14-17, 14-52
Disconnecting from terminal
input, 14-20, 14-57
Disconnecting from timer
interrupts, 11-21
Disk, RFll/RSll fixed-head, 3-1
Disk, RP04 pack, 3-2
Disk, RS03 fixed-head, 3-2
Disk, RS04, fixed-head, 3-2
Disk, RK11/RK05 cartridge, 3-2
Disk, RP11/RP03 pack, 3-2
DJll asynchronous serial line
multiplexer, 2-10
DLll-E asynchronous line interface, 9-2
DPll synchronous line interface,
9-2
DUll synchronous line interface,
9-3

End-of-file and IO.SPF, 6-7
End-of-tape, logical, 6-7
Error codes, directive, B-3
Error codes, I/O s'tatus, B-1
Error data - ICSR and !CAR
registers, 14-67
Error detection and recovery,
14-65
Error reporting,
disable hardware, 14-28
enable hardware, 14-29
Errors, IO.ADS and ADC, 12-30
Errors and recovery, card input,
8-3
Errors, serial line, 14-66
ESCape code conversion, 2-11
Even-parity zero, writing an,
5-11
Event flag, waiting for an, 1-19
Events, significant, 1-10

Floating-point, 12-13
Format, binary, 8-10
Format control, vertical, 7-5

Index-2

INDEX (Cont. )
Format, QIO macro, 1-7
Formats, card reader data, 8-9
FORTRAN interface, 10-3, 11-14,
12-8, 14-29
FORTRAN interface values, 10-8,
11-29, 12-30
FORTRAN subroutine summary,
10-4, 11-15, 12-10, 14-30
Full-duplex considerations, 9-9
Functions, summary of, I/O, A-1

Gain A/D value, switch, 12-13
Gain ranges, use of A/D, 10-9
Get LUN information macro, 1-17,
2-3, 3-3, 4-1, 5-2, 6-1,
7-2, 8-1, 9-3, 10-2, 11-3,
12-2, 13-1
Global common block, 11-9, 11-11,
11-13, 14-69
GLUN$ macro, 1-17

Half-duplex considerations, 9-9
Hardware configuration, 14-1,
14-10
Histogram sampling, 12-16

!CAR contents, 14-68
ICSll support, alternate, 14-3
ICS/ICR address assignments,
14-1
ICSR contents, 14-67
Initializing a timer module,
11-26
Input buffer, reading data from
an, 12-19
Input, contact interrupt digital,
11-5
Input of analog data, 10-5,
11-17, 14-10, 14-35, 14-75
Input, reading digital, 12-18
Interrupt processing, unsolicited, 14-14, 14-20, 14-47,
14-60
I/O completion, 1-24
I/O function codes, B-3
I/O function codes, specific,
A/D converter, B-4
cassette, B-4
communication, B-5
DECtape, B-5
disk (RXOl) , B-5
LPS, B-7
magtape, B-7

I/O function codes, specific, (Cont.)
terminal, B-7
UDC, B-8
I/O function and status codes,
B-1
I/O functions, standard, 1-20,
B-4
I/O functions, summary of, A-1
I/O page, accessing the, 14-71
I/O page global definitions,
14-73
I/O, physical, logical, and
virtual, 1-2
I/O related macros, 1-12
I/O request, issuing an, 1-6,
1-13
I/O requests, cancelling, 1-22
I/O rundown, 14-29
I/O status block, 12-32
I/O status codes, B-1
I/O status conditions, 1-27, 8-7
I/O status word, 12-29
I/O, synchronous and asynchronous
process control, 11-14
IO.ADS and ADC errors, 12-30
IO.ADS function, 12-5
IO.ATT, 1-21
IO.DET, 1-22
IO.HIS function, 12-6
IO.INL, 9-5, 9-10
IO.KIL, 1-22
IO.LED function, 12-3
IO.MDA function; 12-7
IO.MDI function, 12-7
IO.MOO function, 12-7
12-4
IO.RLB, 1-22
IO.RNS, 9-6
IO.RVB, 1-23
IO.RWU, 5-4
IO.SDI function, 12-4
IO.SDO function, 12-4
IO.SEC, 5-4
IO.SPB and IO.SPF, space
functions, 6-7
IO.SPF, end-of-file and, 6-7
IO.SPF, space functions, IO.SPB
and, 6-7
IO.STP function, 12-8
IO.SYN QIO function, 9-6
IO.TRM QIO functions, 9-5
IO.WLB, 1-23
IO.WNS QIO function, 9-6
IO.WVB, 1-23
Isb status array, 10-3, 11-15,
12-8
Issuing an I/O request, 1-6, 1-13

Index-3

INDEX (Cont.)
Keys, special, 2-8
Kill I/O, 14-29
KSR-33/35 Teletypes, 2-2

Latching an output relay, 12-20
Latching digital output, 11-7
Latching or unlatching several
fields, 11-21
LA30 DECwriters, 2-2
LA36 DECwriter, 2-2
LED lights, displaying in, 12-19
Library, system object module,

Macros, I/0-related, 1-12
Magtape I/O function codes,
specific, B-7
Maintenance functions, 14-28,
14-64
Mapping table format, 14-71
.MCALL directive, 1-14
Modules, supported I/O, 14-2
Multiplexer, DHll asynchronous
serial line, 2-10
Multiplexer, DJll asynchronous
serial line, 2-10

11-11

Linking a task to counter
interrupts, 14-22
Linking a task to digital
interrupts, 14-21
Linking a task to the ICS/ICR
conunon block, 14-70
Linking a task to interrupts,
14-60
Linking a task to terminal
interrupts, 14-22
Linking a task to the UDCll
conunon block, 11-13
Logical block, reading a, 1-22
Logical block, writing a, 1-23
Logical Unit Number, 1-4
Logical unit table, 1-4
Logical units, 1-4
LSll line printer, 7-2
LPS I/O function codes, specific,
B-7
LPSll, FORTRAN interface subroutines, 12-10
LPll line printer, 7-1
LPS status returns, 12-26
LUN, assigning a, 1-15
to ADOl-D, 10-6
to AFCll, 10-6
to an LPS, 12-13
to a UDC, 11-18
LUN assignments, changing, 1-5
LUN information, 1-17, 14-6
LUN information macro, get,
1-17, 2-3, 3-3, 4-1, 5-2,
6-1, 7-2, 8-1, 9-3, 10-2,
11-3, 12-2, 13-1
LVll line printer, 7-2

Macro
ALUN$, 1-15
ASTX$S, 1-19
GLUN$, 1-17
WTSE$, 1-21
QIO$, 1-13

Number, logical unit, 1-4
Number of ADOl-D conversions,
restricting the, 10-10
Numbering conventions, 11-30

Offline status, placing selected
unit in, 14-65
Online status, returning a device
tO I 14-65
Operator intervention, 5-4
Output buffer, 12-20
Output, latching digital, 11-7
Output relay, latching an, 12-20
Output, synchronous D/A, 12-23
Output, synchronous digital,
12-24
Output, writing digital, 12-18
Output, use of ADJLPS for input
and, 12-33

Parameter blocks, directive, 1-11
Parity support, vertical, 9-10
Power-fail at a remote site,
14-66
Power recovery at the processor,
14-67
Print line truncation, 7-6
Printer,
LPll line, 7-1
LSll line, 7-2
LVll line, 7-2
Process control I/O, synchronous
and asynchronous, 10-3, 11-14,
14-31
Programming examples, 9-10, 14-58,
C-1
Progranuning hints, 2-11, 4-6,
5-11, 6-6, 7-5, 8-10, 9-8,
10-9, 11-29, 12-31, 13-4
Pulsing several fields, 11-22

Index-4

INDEX (Cont.)
QIO functions, summary of
ICS/ICR-11, 14-7
QIO macro, 1-7, 1-13, 2-3, 3-3,
4-2, 5-2, 6-2, 7-2, 8-2,
9-4, 10-2, 11-3, 12-2,
13-2
QIO functions, device-specific,
2-3, 3-4, 4-2, 5-3, 6-2,
8-2, 10-2, 11-3, 12-3
QIO functions, standard, 1-20,
3-3, 4-2, 5-3, 6-2, 8-2,
9-4, 10-2, 11-3, 12-2

Rate, AFCll, sampling, 10-9
Rates, clock and sampling,
12-31
Reading activating data, 14-24,
14-61
Reading A/D channels, 14-10
Reading a contact interrupt
point 1 11-22
Reading counter data from the
circular buffer, 14-53
Reading digital interrupt data,
14-49
Reading a logical block, 1-22
Reading a single A/D channel,
12-11
Reading from the terminal
buffer, 14-57
Reading a timer module, 11-26
Reading a virtual block, 1-23
Reading contact interrupt
~-~-

u~~~,

,,_~A

~~-~~

Reading data from an input
buffer, 12-19
Reading digital input, 12-18,
14-49
Reading sequential analog
input channels, 10-5,
11-17, 14-38
Reading several contact
sense fields, 11-21
Reading timer interrupt data,
11-25
Ready recovery, 7-4, 8-6
Recovery, card input errors
and, 8-3
Recovery procedures, cassette,
6-5
Recovery procedures, DECtape,

4-5
Recovery, ready, 7-4, 8-6
Recovery, select, 4-6, 5-10
Redundancy checking, 9-9
Relay, latching an output,
12-20

Retrieving LUN information, 1-17
Retrieving system macros, 1-14
Retry procedures for magtape,
5-11
Return codes, 1-25
Returns, status, 1-27, 2-4, 3-5,
4-3, 5-8, 6-3, 7-3, 8-3,
9-7, 10-7, 11-27, 12-26,
13-2, 14-10, 14-31
Reverse reading and writing, 4-6
Rewinding, importance of, 6-7
RFll/RSll fixed=head disk, 3=1
RP04 pack disk, 3-2
RS03 fixed-head disk, 3-2
RS04 fixed-head disk, 3-2
RKll/RKOS cartridge disk, 3-2
RP11-C/RP03 pack disk, 3-2
RSX-llM devices, 1-2
RSX-llM I/O, overview of, 1-1
RT02 alphanumeric display
terminal, 2-2
RT02-C badge reader/alphanumeric
display terminal, 2-2
RT02-C control function, 2-11
RUBOUT character, 7-6
RXll/RXOl flexible disk, 3-2

Sampling, histogram, 12-16
Sampling rate, AFCll, 10-9
Sampling rates, clock and, 12-31
Sampling, synchronous A/D, 12-?1
Sampling, synchronous digital
input, 12-14
Schmitt trigger, 12-6
Select recovery, 4-6, 5-10
Significant event, 1-10
Single A/D channel, reading a,
12-11
Single-shot digital output, 14-13
Software support for ICS/ICR,
14-4
Space functions, IO.SPB and
IO.SPF, 6-7
Special keys, 2-8
Standard I/O functions, 1-23,
B-4
Status array, isb, 10-3, 11-15,
12-8
Status block, I/O, 12-32
Status codes, I/O function and,
B-1
Status condition, I/O, 1-27, 8-7
Status error codes, I/O, B-1
Status returns, 1-27, 2-4, 3-5,
4-3, 5-8, 6-3, 7-3, 8-3,
9-7, 10-7, 11-27, 12-26,
13-2, 14-10, 14-31
Index-5

INDEX (Cont.)
Status success codes, I/O, B-3
Status word, I/O, 12-29
Success codes, directive, B-3
Success codes, I/O status,
B-3
Switch gain A/D value, 12-13
Symbols defined by UDCOM.MAC,
11-10
Sync character considerations,
low-traffic, 9-9
Synchronous A/D sampling, 12-21
Synchronous and asynchronous
process control I/O, 10-3,
11-14, 14-31
Synchronous D/A output, 12-23
Synchronous, device-specific
QIO functions, 12-4
Synchronous digital input
sampling, 12-14
Synchronous digital output,
12-24
Synchronous function, inprogress, 12-20
Synchronous line interface,
DPll, 9-2
Synchronous line interface,
DUll, 9-3
Synchronous subroutines, 12-9
System macros, retrieving, 1-14
System object module library,

Timer module,
initializing a, 11-26
reading a, 11-26
Transmission validation, 9-9
Traps, system, 1-10
Truncation, print line, 7-6
TUlO magnetic tape, 5-1
TU16 magnetic tape, 5-2

UDCll configuration, defining
the, 11-9
UDCll driver, creating the, 11-1
UDCll driver services, 11-2,
14-3, 14-75
UDCll I/O function codes,
specific, B-8
UDCll modules, accessing, 11-2
UDCll programming hints, 11-29
UDCll software compatibility
with ICS/ICR, 14-6
UDCll status returns, 11-27
UDCll symbolic definitions, 11-11
Unlatching several fields,
latching or, 11-21
Unlink a task from interrupts,
14-25, 14-63

11-11

System traps, 1-10

Tape characteristics, 5-5
Tape characteristics, resetting,
5-11

Tape, structure of cassette, 6-5
Tape, TUlO magnetic, 5-1
Tape, TU16 magnetic, 5-2
Teletypes,
ASR-33/35, 2-2
KSR-33/35 I 2-2
Terminal,
RT02 alphanumeric display, 2-2
RT02-C badge reader/alphanumeric display, 2-2
VT05B alphanumeric display, 2-3
VT50 alph~numeric display, 2-3
Terminal line truncation, 2-11
Terminal output, 14-27, 14-46
Terminating AST service, 1-19
Timer, 11-7
Timer interrupt data, reading,

Verification of write operations,
6-7
Vertical format control, 7-5
Vertical parity support, 9-10
Virtual block,
reading a, 1-23
writing a, 1-23
VT05B alphanumeric display
terminal, 2-3
VT50 alphanumeric display
terminal, 2-3

Waiting for an event flag, 1-19
Write operations, verification
of, 6-7
Writes, retry procedures for
reads and, 5-11
Writing digital output, 12-18
Writing an even-parity zero, 5-11
Writing a logical block, 1-23
Writing a virtual block, 1-23

11-25

Timer interrupts,
connecting to, 11-19
disconnecting from, 11-21
Index-6

RSX-llM I/O Drivers
Reference Manual
DEC-11-0MDRA-B-D
READER'S COMMENTS
NOTE:

This form is for document corrments only. Problems
with software should be reported on a Software
Problem Repcrt (SPR) form

Did you find errors in this manual?

If so, specify by page.

Did you find this manual understandable, usable, and well-organized?
Please make suggestions for improvement.

Is there sufficient documentation on associated system programs
required for use of the software described in this manual? If not,
what material is missing and where should it be placed?

Please indicate the type of user/reader that you most nearly represent.

0
0

Assembly language programmer
Higher-level language programmer
Occasional programmer (experienced)
User with little programming experience
Student programmer
Non-programmer interested in computer concepts and capabilities

CitY~~~~~~~~~~~~~--State~~~~~~-Zip

Code __~~--~--~
or
Country

If you require a written reply,

please check here.

[]

---------------------------------------------------Fold IIere------------------------------------------------------------

--------------------------------------- Do Not Tear - Fold IIere and Staple ----------------------------------------------FIRST CLASS
PERMIT NO. 33
MAYNARD, MASS.
~ESS REPLY MAIL
)STAGE STAMP NECESSARY IF MAILED IN THE UNITED STATES

e will be paid by:

~nmnomn
Software Communications
P. 0. Box F
Maynard, Massachusetts 01754

digital equipment corporation

Printed in U.S.A.