Digital PDFs

AA-L674B-TC

April 1983

245 pages

Original

14MB

Document:	RSX-11M/M-PLUS Error Logging Manual
Order Number:	AA-L674B-TC
Revision:	0
Pages:	245
Original Filename:	http://bitsavers.org/pdf/dec/pdp11/rsx11m_s/RSX11M_V4.1_Apr83/6_SysMgmt/AA-L674B-TC_errLog_198304.pdf

OCR Text

RSX-11 M/M-PLUS
Error Logging Manual
Order No. AA-L674B-TC

RSX-11 M Version 4.1
RSX-11 M-PLUS Version 2.1

digital equipment corporation · maynard, massachusetts

First Printing, January 1982
Revised, April 1983
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 be used or copied only in accordance with the terms of such
license.
No responsibility is assumed for the use or reliability of software on
equipment that is not supplied by Digital Equipment Corporation or its
affiliated companies.

@ 1982, 1983 by Digital Equipment Corporation
All Rights Reserved.
Printed in U.S.A.

The postpaid READER'S COMMENTS form on the last page of this document
requests the user's critical evaluation to assist in preparing future
documentation.
The following are trademarks of Digital Equipment Corporation:
DEC
DEC/CMS
DEC/MMS
DECnet
DECsystem-10
DECSYSTEM-20
DEC US
DECwriter

DIBOL
Edu System
IAS
MAS SB US
PDP
PDT
RSTS

RSX
UNIBUS
VAX
VMS
VT

mamanma
ZK2344

HOW TO ORDER ADDITIONAL DOCUMENTATION
In Continental USA and Puerto Rico call

800-258-1710

In New Hampshire, Alaska, and Hawaii call 603-884-6660
In Canada call 613-234-7726 (Ottawa-Hull)
800-267-6146 (all other Canadian)

DIRECT MAIL ORDERS (USA & PUERTO RICO)*
Digital Equipment Corporation
P.O. Box CS2008
Nashua, New Hampshire 03061

DIRECT MAIL ORDERS (CANADA)
Digital Equipment of Canada Ltd.
940 Belfast Road
Ottawa, Ontario K1G 4C2
Attn: A&SG Business Manager

DIRECT MAIL ORDERS (INTERNATIONAL)
Digital Equipment Corporation
A&SG Business Manager
c/o Digital's local subsidiary or
approved distributor

*Any prepaid order from Puerto Rico must be placed
with the local Digital subsidiary (809-754-7575)
Internal orders should be placed through the Software Distribution Center (SOC), Digital Equipment
Corporation, Northboro, Massachusetts 01532

CONTENTS

Page
PREFACE

SUMMARY OF TECHNICAL CHANGES

CHAPTER 1
1.1
1.2
1. 2.1
1. 2. 2
1. 2. 3
1. 2. 4
1.3
1. 3 .1
1. 3 .2
1. 3. 3
1. 3. 4
CHAPTER 2

INTRODUCTION
THE PURPOSE OF ERROR LOGGING • • • • • •
• • 1-1
ERROR LOGGING OPERATION
• • • • • 1-1
Executive Routines
• • • •
• • • •
1-3
ERRLOG and ELI •
• • • •
1-4
RPT
• • • •
• • • • • • • 1-4
CFL
• • • • • • • • • • • •
• • 1-5
ERROR LOGGING OPTIONS
• • • •
• 1-6
Unexpected Traps or Interrupts
• • • • 1-6
Device Errors
• • • • • • • • • • • • 1-6
Interrupt Timeouts
• • • • • • •
• • • • 1-6
Memory Errors
• • • •
• • • • 1-6
ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)

INSTALLING ERRLOG AND ELI
2.1
2.2
USING ERRLOG AND ELI • • •
ELI SWITCHES • • • • • • • • . . . .
2.3
•
Logging Switches • • • •
2.3.1
ERROR Limiting Switches • • • • .
2.3.2
The Limit Switch • • • •
2.3.2.1
The Hard Limit Switch • • • • •
2.3.2.2
The Reset Switch • • • •
2.3.2.3
The Soft Limit Switch
2.3.2.4
2.3.3
File Naming Switches • • • • •
The Log Switch • • • • • • •
2.3.3.1
2.3.3.2
The Append Switch • • • • •
The Switch Switch
2.3.3.3
. .
2.3.3.4
The Backup Switch
Display Switch • • • •
. . . .
2.3.4
ERRLOG AND ELI MESSAGES
2.4
•
ELI Messages • • •
2.4.1
ERRLOG Messages • • • •
2.4.2
CHAPTER 3

• • 2-1
• 2-2
.

. 2-2

. 2-6

• • •
.

• 2-4
.

• • • • • • • 2-6
.

.
.

•
• •
• •
•
.
. .
. . . . .
. . . . .

• • • •

.
•
•
•
•
•
.
.

2-7
2-7
2-7
2-8
2-8
2-8
2-8
2-9
2-10

2-11
2-11

2-12

REPORT GENERATOR TASK (RPT)

3.1
INSTALLING AND RUNNING RPT • • • • • • •
• • • 3-1
USING RPT TO CREATE ERROR LOG REPORTS
• 3-2
3.2
3.2.1
The RPT Command Line • • • • • • • •
• • 3-2
Using Multiple Qualifiers in RPT Command Lines • 3-3
3.2.2
Using the Default RPT Command Line •
3.2.3
• • • 3-4
. . . . . . 3-5
RPT REPORT SWITCHES • • • •
3.3
Packet Selection Switches
. . . . . . . . 3-7
3.3.1
The Date Switch • • • • . . . .
. . 3-7
3.3.1.1
iii

CONTENTS
Page

The Device Switch • • • • • • • • • • • • •
3.3.1.2
The Packet Switch • • • • • • • • • • •
3.3.1.3
The Drive and Pack Serial Number Switch
3.3.1.4
The Type Switch
•
3.3.1.5
• • • • • •
The Volume Label Switch
3.3.1.6
• • • •
Report Format Switch
3.3.2
• • • • •
Brief Reports
•••••
3.3.2.l
Full Reports • • • • • •
3.3.2.2
Register Reports • • • • •
• • • •
3.3.2.3
3.3.2.4
No Report
• • • • • • • • • • •
Summary Switch (RSX-llM-PLUS only)
3.3.3
The All Qualifier
3.3.3.1
The Error Qualifier
• • • •
• •••••
3.3.3.2
3.3.3.3
The Geometry Qualifier •
• •••
The History Qualifier • • • • • • • • • • •
3.3.3.4
The None Qualifier • • • • • • • •
3.3.3.5
The Report Switch
• • • •
• ••••
3.3.4
3.3.4.1
Predefined Switch Strings
••••
User Defined Switch Strings
3.3.4.2
The Width Switch • •
3.3.5
ERLCNF REPORT MESSAGES •
3.4
ERLRPT REPORT MESSAGES •
3.5
CHAPTER 4

•
•
•
•

3-8
3-8
3-9
3-9
3-10
3-11
3-11
3-15
3-18
3-20
3-20
3-20
3-20
3-23
3-25
3-27
3-27
3-27
3-28
3-28
3-29
3-35

ERROR LOG CONTROL FILE ARCHITECTURE

TERMS AND CONCEPTS • • • • • • • • • • • • • • • • 4-i
CONTROL FILE MODULE ARCHITECTURE • • • • • • • • • 4-2
RSX-llM and RSX-llM-PLUS Control File Modules
• 4-3
4-10
Program Control Flow • • • • •
4-11
Compilation Paths
••••••
Modification and Recompilation • • • • • • • • 4-13
INTERNAL INTERFACES
• • • • • •
4-13
Interaction Between Dispatcher and Device-Level
Modules
• • • • • • • • • •
4-13
Interaction between DSP2Ml and ERM23 •
4-14
4.3.1.1
4.3.1.2
Interaction Between DSP2Pl and ERM23
4-15
4.4
DISPATCHING
• • • • • • • • • • • • • • •
4-16
4.4.1
Event-Level Dispatching
4-16
Device-Level Dispatching
4.4.2
4-18
CPU-level Dispatching
• • • • 4-19
4.4.3
SUPPORT OF NON-DIGITAL DEVICES • • • • • •
4-19
4.5
Error-logging of Unknown Devices •
4.5.1
4-19
Providing Driver Support for a Non-DIGITAL
4.5.2
Device • • • • • • • • • • • • • •
4-19
$BMSET on RSX-llM
4-20
4.5.2.1
$BMSET on RSX-llM-PLUS
4.5.2.2
4-20
$DVTMO and $DTOER on RSX-llM
4-20
4.5.2.3
$DVTMO and $DTOER on RSX-llM-PLUS
4.5.2.4
4-21
$DVERR and $DVCER on RSX-llM
4-21
4.5.2.5
4.5.2.6
$DVERR and $DVCER on RSX-llM-PLUS
4-22
$NSIER
4-22
4.5.2.7
$FNERL
4.5.2.8
4-22
$LOGER •
4-23
4.5.2.9
4.5.2.10
LOGTST
4-23
$CRPKT •
4-23
4.5.2.11
4.5.2.12
CALDEV on RSX-llM-PLUS
4-24
4-24
$QUPKT • • • • • • • • •
4.5.2.13
4.5.2.14
$QERMV • • • • • • • • •
4-25
Error-Logging Support for a Non-DIGITAL Device 4-25
4.5.3
4.5.3.1
How to Write a Device-Level Module • •
4-25
4.5.3.1.1
MODULE Statement • • • •
• •••
4-26
4.5.3.1.2
PROCEDURE Statement
4-26
4.5.3.1.3
SUBPACKET Declaration
4-26
4.5.3.1.4
Register Definitions •
4-27
4.1
4.2
4.2.l
4.2.2
4.2.3
4.2.4
4.3
4.3.1

CONTENTS
Page

4.5.3.1.5

Declaration of Local Work Variables and
Tables • • • • • • • • • • • • • • • • • •
4.5.3.1.6
Loading of the Intermodule Variables • • •
4.5.3.1.7
Determination of the Error Type • • • • •
4.5.3.1.8
Coroutine Back to Caller • • • • • • • • •
4.5.3.1.9
Perform the Bit-To-Text Translation and
Register Printing • • • • • • • • • •
4.5.3.1.10
Indicate Any Notes that are Required •
4.5.3.1.11
Exit the module • • • • • • • • •
4.5.3.2
How to Write a Notes Module • • • • • • • •
4.5.3.2.l
MODULE Statement • •
• • • •
4.5.3.2.2
PROCEDURE Statement
• • • •
4.5.3.2.3
Notes Heading
• • • • • •
4.5.3.2.4
Selecting a Note for Printing
4.5.3.2.5
Handling an Unknown Note Number
4.5.3.2.6
Getting the Next Note • • • • • •
4.5.3.2.7
Exit the Module • • • • • • • • •
4.5.3.3
MASSBUS and Non-MASSBUS Considerations •
4.5.3.4
Making the New Device-Level Module Known • •
4.6
CODE EXAMPLES • • • • • • • • • • • • • • •
4.6.1
RM02/03 Device-Level Module ERM23 • • • • • •
4.6.2
DSP2Ml Dispatcher Module for RSX-llM • • •
4.6.3
DSP2Pl Dispatcher Module for RSX-llM-PLUS
4.6.4
RM02/03 Notes Module NRM23 • • • • • •
4.6.5
Subpacket Definitions • • • • • • • • • •
4.6.5.1
Subpackets Declared by DISPATCH
4.6.5.2
Subpackets Declared by DSPlMl/DSPlPl •
4.6.5.3
Subpackets Declared by DSP2Ml/DSP2Pl • •
4.6.5.4
Subpackets Declared by DSP3Ml/DSP3Pl =
4.6.5.5
Subpackets Declared by DSP4Ml/DSP4Pl
4.6.5.6
Subpackets Declared by DSP5Ml/DSP5Pl •
4.6.5.7
Subpackets Declared by DSP6Ml/DSP6Pl • •
4.6.5.8
Subpackets Declared by DSP7Ml/DSP7Pl • •
CHAPTER 5

4-29
4-29
4-30
4-30
4-30
4-31
4-31
4-31
4-32
4-32
4-32
4-33
4-33
4-33
4-33
4-33
4-34
4-37
4-37
4-50
4-57
4-67
4-69
4-69
4-72
4-73

4-73
4-74
4-74
4-74
4-75

CONTROL FILE LANGUAGE GUIDE

CONTROL FILE OVERVIEW
5.1
. . . . . . . . . . . 5-1
5 .1.1
Report Generator General Processing • • • • • • 5-1
The General Format of an Error Log Packet
• 5-2
5 .1.2
5 .1. 3
Control File Language • • • • • • • •
• • 5-2
General Format of Control File Modules
5 .1.4
• • 5-2
5 .1. 5
Files • • • • • • • •
• •••
5-3
TYPES AND EXPRESSIONS
5.2
5-4
5.2.1
Data Types • •
• •••••••
• • 5-4
5.2.1.1
LOGICAL Type • • • • • • • • • • • • • •
5-4
5.2.1.2
STRING Type • • • •
• 5-4
ASCII Type • • •
5.2.1.3
5-5
Numeric Types
5.2.1.4
5-5
Field Types
• • • 5-7
5.2.1.5
5.2.1.6
POINTER Type •
5-7
RSX TIME Tvoe
5-7
5.2.1.7
VMS-TIME Type
5.2.1.8
• • 5-7
Variables
5-8
5.2.2
Literals
••••
5.2.3
5-9
Expressions
5.2.4
5-9
5.2.4.1
String Operators • • • • • •
5-9
Logical Operators • • • •
5.2.4.2
5-10
Relational Operators
5.2.4.3
5-11
Numeric Operators • • • • • • • • • • •
5.2.4.4
5-13
5.2.5
Operator Precedence
• • • •
5-15
FUNCTIONS • • • •
• • • • • • • • •
5-16
5.3
5.3.1
%CND Functions - Conditional Functions
5-17
%CNV Functions - Conversion Functions
5.3.2
5-17
5.3.2.1
%CNV Functions - Numeric Conversion Functions 5-17

CONTENTS
Page

%CNV Functions - Miscellaneous Conversion
Functions
• • • • • • • • • • • •
5.3.3
%COD Functions - Encoding Functions
5.3.4
%COM Functions -Computational Functions
5.3.5
%CTL Functions - RPT Control • • • •
5.3.6
%LOK Functions - Lookahead Functions
5.3.7
%PKT Functions - Packet Information • • • • •
5.3.8
%RPT Functions - Report Control
5.3.9
%STR Functions - String Handling
5.3.10
%TIM Functions - Time Handling • • • • • • • •
5.3.11
%USR Function - User I/O Function
5.4
DECLARATIONS • • • • • • •
5.4.1
Scope of Declarations
5.4.2
DECLARE Statement • • • • • • • •
5.4.3
PACKET Statement • • • • • • • •
5.4.4
SUBPACKET Statement • • • • • • •
• •••
5.4.5
Conditional Declarations •
• •••
5.5
ACTION STATEMENTS • • • • •
• ••••
5.5.1
SET Statement • • • • •
5.5.2
INCREMENT and DECREMENT Statements •
5.5.3
WRITE Statement
5. 5. 4
WRITE GROUP Statement • • • •
5.5.5
DECODE Statement •
5.6
CONTROL STATEMENTS ••
5.6.1
MODULE Statement ••
5.6.2
LITERAL Statement
5.6.3
CALL Statement ••
5.6.4
RETURN Statement
••••
5.6.5
PROCEDURE Statement
5.6.6
IF-THEN-ELSE Statement ••
5.6.7
CASE Statement • • • • • • • • • • • • • • • •
5.6.8
SELECT Statement • • • •
5.6.9
WHILE/UNTIL/DO Statements
5.6.10
LEAVE Statement
5.6.11
BEGIN-END Statement
5.6.12
Lexical Conditionals ••
5.7
TABLES • • • • • • • • • • • • • •
5.7.1
Table Structure
5.7.2
TABLE Statement
5.7.3
DYNAMIC TABLE Statement • • • •
5.7.4
FILE Statement • • • •
5.7.5
POINTER Statement
5.7.6
FIND Statement •
5.7.7
PUT Statement • • • • • • • • •
5.8
LISTS • • • • • •
• •••••
5.8.1
LIST Statement • • • • • • • • • • • • • •
5. 8. 2
SEARCH Statement • • • • •
5.9
SIGNALLING • • • • • • • • •
5.9.l
Signalling • • • •
5.9.2
ENABLE Statement •
5.9.3
SIGNAL Statement
5.9.4
SIGNAL STOP Statement • • • • • • • • • • • •
5.9.5
MESSAGE Statement
••••
5.9.6
CRASH Statement
•••••
5.10
PRINT FORMATTING • • • • • • • • • • • • •
5.10.1
FORMAT keyword string • • • • •
5.10.1.1
Control Directives •
5.10.1.2
Formatting Directives
5.10.1.3
Data-formatting Directives •
5.11
USER INTERFACE HANDLING
••••
5.11.l
Overview of User Interface Handling • • • • •
5.11.2
Command Mode • • • •
5.11.3
Option Mode • • • • •
5.12
ERLCFL REPORT MESSAGES • • •
5.3.2.2

5-19
5-19
5-20
5-21
5-21
5-22
5-22
5-23
5-25
5-25
5-26
5-26
5-26
5-28
5-29
5-29
5-30
5-31
5-31
5-31
5-31
5-32
5-32
5-32
5-33
5-33
5-33
5-34
5-34
5-34
5-35
5-35
5-36
5-36
5-36
5-36
5-37
5-37
5-37
5-38
5-39
5-39
5-39
5-39
5-40
5-40
5-40
5-40
5-40
5-41
5-41
5-41
5-41
5-41
5-42
5-42
5-42
5-43
5-43
5-43
5-43
5-44

CONTENTS

Page

APPENDIX A

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY

APPENDIX B

DRIVE SERIAL NUMBERS

APPENDIX C

ERROR LOG PACKET FORMAT

EXAMPLES
EXAMPLE 2-1
3-1
3-2
3-3
3-4
3-5
3-6
A-1
C-1

Error Logging Status • • •
Error Log Brief Report ••
Error Log Full Report
•
Error Log Register Report • • • •
Error Summary Report • • • • • • •
• •
Geometry Summary Report
History Summary Report • • •
• • •
Sample Execution of TUNE.CMD •
• • • •
Error Log Packet Format
. •

• • •
• • •
•

2-10
3-13
3-16
3-19
3-22
3-24
3-26

• • • • A-2
• • • • C-1

FIGURES
FIGURE

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

Error Logging System • • • • • • • • • • • • • • • 1-2
Structure of Error-Logging Packet • • • • • • • • 4-3
Compilation Path for RSX-llM Control File Modules 4-11
Compilation Path for RSX-llM-PLUS Control File
Modules
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
o
e
4-12

TABLES
TABLE

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

A-1
A-2
B-1

ELI Switches and Subswitches • • • • • • • • • • • 2-3
Error Logging Devices
• • • • • • • • • • • 2-5
RPT File Specification Defaults
• • • •
• 3-3
RPT Report Switches and Subswitches
• • • 3-5
Error Logging Code/Subcode Combinations • • • • 4-17
Event Types, Codes, and Their Dispatcher Modules 4-18
The DEVICE INFO Table • • • • • • • • • • • • • 4-34
Modules in-ERRLOG.ULB for RSX-llM
• A-4
Modules in ERRLOG.ULB for RSX-llM-PLUS • •
• A-5
Significant Digits in Drive Serial Numbers
•• B-1

vii

PREFACE

This manual contains information about operating the RSX-llM/M-PLUS
Error Logging System.
It explains how the Error Logger collects
information on system events and errors and how the Report Generator
and Control File produce various kinds of reports on those events and
errors.
It also includes information on the control file architecture
and on how to add user-written modules. The error logging system
allows you to monitor the reliability of the hardware on your system
and to set error limits and display messages on the console terminal
if the number of errors on a device exceeds those limits.
This manual assumes you are familiar with the following documents:
The RSX-llM/M-PLUS MCR Operations Manual
The RSX-llM/M-PLUS Utilities Manual
The RSX-llM-PLUS or RSX-llM System
Guide

Generation

and

Installation

The RSX-llM and RSX-llM-PLUS Information Directory and Master Index
define the intended readership for each manual in the documentation
set and provide a synopsis of each manual's contents.
When this
manual refers to other documents, consult the appropriate information
directory for information about the document.

INTENDED AUDIENCE
This manual is intended for Field Service personnel, system managers,
and others responsible for maintaining the integrity of hardware
devices connected to an RSX-llM or RSX-llM-PLUS system.
In addition to understanding the RSX-llM or RSX-llM-PLUS operating
system and the Error Logging System, you need a thorough knowledge of
the hardware devices that the Error
Logging System is monitoring.
This manual does not attempt to describe or explain the hardware
information that appears in the Error Log Reports.
For information
about a specific device, consult the hardware documentation for that
device.

STRUCTURE OF THIS DOCUMENT
Chapter 1 provides an overview of the purpose and function of the
Error Logging System.
It describes some features and limitations of
the system and explains the operating system resources that error
logging requires.
Chapter 2 describes the procedures for operating the Error Logger and
explains the Error Log Interface commands to control logging and
limiting.

Chapter 3 describes the procedures for operating the Report
and describes the report formatting available.

Generator

Chapter 4 explains the control file modules in detail, including flow
of
program
control,
interfaces
between
modules, and module
dispatching.
A knowledgeable system
programmer
can
use
the
information presented to add user-written modules to the Error Logging
System.
The chapter includes extensively annotated examples of
DIGITAL-supplied modules.
Chapter 5 describes the Control File Language, which is used to
control-file modules.

write

Appendix A describes the indirect command file, TUNE.CMD, that you can
use to remove devices from the Error Logging ULB and make it smaller.
Appendix B describes the formats used
DIGITAL devices.

for

drive

serial

numbers

Appendix C describes the formats for standard error log subpackets.

CONVENTIONS USED IN THIS DOCUMENT
Examples of Error Log Reports illustrate the operation of the Report
Generator.
They
do
not
attempt
to
explain
the specific
hardware-related events that the reports describe.
Black ink in command line descriptions designates
displays at the terminal.

what

the

computer

Red ink designates what the user enters at the terminal.
Square brackets [] enclose the optional parameters for an ELI, RPT, or
CFL command.
Uppercase characters in command lines or syntax descriptions
required syntax for the command.

indicate

Lowercase
selects .

the

user

Pink shading in text and examples indicates features that appear
on RSX-llM systems.

only

characters

indicate

variable

parameters

that

•qr~t. i~n~(ttrg. ··~iA.· t~~~ ;a,?ld;i· ~~a~;~',
O,U:' '~S~:-:-:1,l~t;P~QS:;· ~~:S~z~~··;·; .•

SUMMARY OF TECHNICAL CHANGES

The Error Logging System will now allow the hard and soft error limits
to be reached independently. Previously, reaching one of the limits
would disable logging of either kind of error on that device.
Now,
reaching the soft limit will not affect the logging of hard errors,
nor will reaching the hard limit affect the logging of soft errors.
Device timeouts are now logged as hard errors if unrecoverable, and as
soft errors if recoverablee
When generating a report, RPT looks first for LX: [l,6]ERRLOG.ULB.
it fails to find that file, it looks for LB: [l,6]ERRLOG.ULB.

The Executive ERROR module now resides in a directive
RSX-llM-PLUS systems and may reside in a directive common
systems.
Therefore, drivers that create data areas
information to be passed to the Executive ERROR module must
the data area in memory mapped by APR5.

common on
on RSX-llM
containing
not create

There are no user interface changes except for a number of
messages.

new

error

Two new chapters in this manual document the architecture of error log
control files and the Control file Language (CFL). These chapters
replace Appendix A in the previous version of this manual.
There have been a number of minor changes in CFL. Here is a list of
differences between CFL in RSX-llM V4.l and RSX-llM-PLUS V2.l and the
previous releases:
e

DYNAMIC TABLE statement:
DYNAMIC TABLE is a synonym for FILE. You should use this
statement in place of any FILE statement in new code.

•

new

FILE statement:
The FILE statement will be removed from a future release.
Please convert your code to use DYNAMIC TABLE instead of FILE.

•

%CNV$xxx functions:
The field width parameter is now optional and interacts with
the optional fill character parameter to determine whether the
resulting string- is printed as
is
or
is
leftor
right-justified.
In the earlier version, the digits in the
string were always right-justified and blank-filled if no
fill character was specified.

SUMMARY OF TECHNICAL CHANGES
e

%LOK$LENGTH function:
This function always returns the length of the data in a
packet or subpacket.
The length word for
the packet or
subpacket is not considered part of the data and is not
counted in determining the length value returned.

%LOK$BYTE, %LOK$WORD, %LOK$LONGWORD functions:
The offset parameter is the offset within the data of the
packet/subpacket at which the byte, word, or longword begins.
The offset unit is always in bytes, with the first byte of
data in the packet/subpacket being offset O.

%STR$UPCASE function:
STR$UPCASE accepts an ASCII string as a parameter, and returns
the ASCII string with all lowercase ASCII characters converted
to uppercase.

WRITE and WRITE GROUP statements:
Because of overlay restrictions, the following operators and
functions
cannot
be
used
in expressions in WRITE or
WRITE GROUP statements:
single and double operand numeric operators
the MA~Ctt operator
%COD$xxx functions
%CTL$xxx functions
%PKT$xxx functions
%RPT$xxx functions
%STR$xxx functions
%TIM$xxx functions
%USR$xxx functions

xii

CHAPTER l
INTRODUCTION

1.1

THE PURPOSE OF ERROR LOGGING

The RSX-llM/M-PLUS Error Logging System records information about
errors and events that occur on your system hardware, either for
immediate action or for later analysis and reporting.
Error logging
handles mass storage device (disk and tape) errors, as well as memory
errors. Since error logging is a part of the RSX-llM/M-PLUS system,
it is most effective for hardware errors that allow the system to
continue functioning.
Error logging is not used to detect information about operating system
failures or about device problems that cause the system to fail.
However,
it does provide information about what
I/O activities
occurred on a device at the time of an I/O failure.
If your system
includes the Crash Dump Analyzer (CDA) , CDA can provide reports on
operating system failures.
You can use Error Log Reports to determine that a device is having
problems before the device actually fails and causes you to lose data.
For example, a report showing a pattern of recurring errors from
different blocks on a single disk head may indicate that the head
needs to be replaced.

1.2

ERROR LOGGING OPERATION

The complete Error Logging System is composed of four tasks.
•

The Error Logger (ERRLOG)

•

The Error Log Interface (ELI)

•

The Report Generator (RPT)

•

The Control File Language Compiler (CFL)

When the executive or a device driver detects an error,
Executive
routines create an Error Log Packet in pool to describe the event.
(See Appendix C for a description of the Error Log packet.)
ERRLOG
then writes the packet from pool into the Error Log File on disk,
usually within a few seconds of when the packet is created.
Figure 1-1 shows the interaction of the
with routines in the Executive.

1-1

Error

Logging

System

tasks

ERROR LOG FILE

EXECUTIVE

ERROR LOG
REPORT

ERR LOG
PACKET-1
REPORT
GENERATOR

PACKET-2
COMMAND PKT

POOL

APT

ELI

COMMAND PKT

DRIVER

I-'

ERROR
module

1-i
~

I
IV

CFL
COMPILER

CONTROL
FILE

1-i
~

SOURCE.CNF

CONTROL
FILE
MODULE
ZK-495-81

Figure 1-1

Error Logging System

INTRODUCTION
ERRLOG receives user commands from the Error Log Interface
(ELI)
to
control ERRLOG operation.
These commands send error log packets
called command packets to the ERRLOG task.
The Report Generator
the Error Log File.

(RPT) generates reports from the

information

RPT uses a library of modules written in the Control File Language
(CFL)
to interpret data from the Error Log File and from user
commands. The CFL compiler is also part of the Error Logging System.
You can use CFL to recompile DIGITAL-supplied Control File Modules to
include patches to the modules supplied in the future.
You can also
use CFL to create and compile Control File Modules for devices other
than those DIGITAL supplies. Chapter 4 explains the control file
module architecture and
includes annotated DIGITAL control file
modules. Chapter 5 documents the Control File Language (CFL).

1.2.1

Executive Routines

Whenever the RSX-llM or RSX-llM-PLUS system is running and error
logging is active, routines in the Executive collect information from
device drivers and other tasks and write the information into error
log packets in system pool.
The Executive gathers information on the state of the registers when a
device error occurs, and includes information on system events, such
as device Mounts and Dismounts. You can also insert a text message
into the error log file using the MCR System Service Message command
(SSM).
(See the RSX-llM/M-PLUS MCR Operations Manual).
If Error Logging is not active on the system, the device drivers still
detect each hardware error, but the Executive does not create Error
Log packets.
The Error Logging System makes a distinction between hard errors and
soft errors.
Hard errors are those that cause an I/O operation to be
aborted because the device driver cannot recover from the error.
The
task that issued the I/O request receives an error code indicating
that the operation failed.
Soft errors are those from which the
device driver can recover.
The task that issued the I/O request does
not receive an error notification because the request eventually
succeeds.
The Error Logging System logs both hard and soft errors.
Thus, you
can have a system functioning properly, with no errors reported to any
tasks in the system, with errors still being encountered and logged.
Thus error logging terminology sometimes refers to errors as events:
they do not always mean an actual failure.
When Error Logging is active, the Executive writes the data from a
single event into one Error Log Packet and assigns a sequence number,
unique to that event, to the packet.
The Resource Monitoring Display
(RMD)
shows the highest assigned sequence number as ERRSEQ, the total
number of errors since error logging operations began.
When ERRLOG writes the packet in a file, the packet gets a number that
describes its location in the file relative to other packets. RPT
uses this number to refer to the event in later operations.
The
number does not change unless the organization of the file changes.
For example, if an earlier error log file is appended to the current
error log file, the packet numbers in the appended file will change.

1-3

INTRODUCTION
Thus, you can generate a brief format RPT report to determine the
packet numbers of the most significant errors on your system, and then
generate a full format report, by packet number, of only those errors.
The Executive includes a directive for error logging
(SMSG$)
that
sends Error Log Packets directly to the Error Logger.
(See the
RSX-llM/M-PLUS Executive Reference Manual for an explanation of how to
use this directive.) User tasks can use SMSG$ to communicate with the
Error Logger.

1.2.2

ERRLOG and ELI

ERRLOG writes the Error Log Packets from pool to the Error Log File in
binary format.
Only RPT can interpret and format data from the Error
Log File.
To issue a command to ERRLOG, type an ELI command to perform one of
the ERRLOG functions (logging, limiting, or file naming).
ELI sends
an error log command packet to ERRLOG with instructions on the
function to be performed, and ERRLOG returns the results, if any, to
ELI.
The ERRLOG task allows you to specify two files to contain the error
log packets written to disk.
ERRLOG uses the first file, the error
log file, unless an error is detected while ERRLOG writes to the file.
If an error
is detected,
ERRLOG switches to cne secona r11e, cne
backup file.
ELI commands allow you to establish or change the names
of the error log file and backup file.
The error logging system automatically limits the number of events it
logs on a given device.
This error limit can be changed dynamically
by ELI commands while error logging is running.
The system does
limiting in case the device starts to accumulate a large number of
errors. Without limiting in these cases, the error log file wo~ld
quickly become large and difficult to analyze.
The limiting does not
throw away useful information, because usually when a large number of
events occurs on a device, most of them are the same and you can
generalize from a report on a small number of the events.
After a device reaches a particular error limit, logging of that type
of error on the device stops until you reset the error count to zero
or raise the error limit.
ERRLOG sends a message to the console terminal or to any terminal that
has allocated the device, explaining that the device reached the error
limit.
Limiting does not affect operation of the device itself;
it
only starts or stops error logging on the device.

1.2.3

RPT

RPT creates reports on the data in the Error Log File, based on
information in the Error Log Control File and commands supplied by the
user. Modules in the Error Log Control File tell RPT how to interpret
and print entries from the Error Log File for a specific operating
system.
When you are ready to generate an Error Log Report, you can run RPT to
select the information you want to include in the report.
RPT can
generate reports in brief and full format on any collection of Error
Log Packets you select.
For example, you can select reports on a

1-4

INTRODUCTION
specific device by device name, device type, volume label, pack
identification, or drive serial number.
You can also select reports
of a specific error type or you can select a full report of all the
Error Log Packets in the Error Log File.
Error log reports can contain
device-supplied information.

both

context

information

and

Context information, which appears in full
format reports, contains
operating system version information and some information about the
CPU model. Context information on the I/O operation that encountered
the failure is recorded for device errors.
This information is useful
to correlate events recorded in the error log file with other events
in the system.
For example, hard I/O errors often cause the task
issuing the I/O request to exit with an error, since many tasks cannot
recover from I/O errors.
Information on the I/O operation is also
useful to determine the operation the device driver attempted at the
time of the failure.

This support is optional on RSX-llM.
In a full report, RPT also includes all device-supplied information,
including registers and any other information the device provides.
Each device supplies one or more machine words of information when an
error occurs.
RPT decodes each item of device information according
to the terminology used
in the device maintenance manual.
If
additional
information is useful to understand the significance of a
decoded item, that information is listed in parentheses.
Decoded items that are abnormal are flagged with a "*"
in reports.
These items may or may not represent error conditions, depending on
the state of the device.
Interpret items flagged with a "*" as "look
at me first".
RPT reports flag more than one item on most devices.
RPT reports also flag more than one item if a device encounters an
error or cannot perform an operation because of another error
condition. This condition occurs when an abnormal device status
condition causes an I/O function to fail.
The RPT report flags both
the I/O function failure and the abnormal device status.
An error type definition in the RPT report then boils all the
device-supplied information down to a single item reflecting the most
probable error reported by the device.

1. 2. 4

CFL

The Error Logging System includes a Control File Language compiler
(CFL)
used to recompile patched DIGITAL-written Control File Modules
or user-written modifications or additions to modules. Chapters 4 and
5 describe the operation of the CFL compiler and the DIGITAL-supplied
control file module for the RM02/RM03.

1-5

INTRODUCTION
1.3

ERROR LOGGING OPTIONS

Routines in the Executive respond to four types of errors:
•

Unexpected traps or interrupts

•

Device errors

•

Device timeouts

•

Memory errors

All systems that include Error Logging support

system

generation

~;~~~f~.:.:1f~~i~~f~l~~tr~~~~~sJi~;ij1;~~.:$~f~~~~4·1 errors·
However, support for logging
generation option on RSX-llM.

1.3.l

memory

errors

a separate system

Unexpected Traps or Interrupts

When your system includes Error Logging support, all unused system
vectors are filled with pointers to routines
in the Executive.
Therefore, routines in the Executive are called if a trap or interrupt
occurs to one of these unused vectors.
For example, a noisy
electrical environment or a static discharge may cause an unexpected
trap or interrupt to one of the unused vectors, or a valid interrupt
may be vectored to the wrong address.
In these cases,
the Executive
records this information.

1.3.2

Device Errors

Device errors are problems that a device encounters while carrying out
a software-requested operation. When a device error occurs and Error
Logging is active, the device driver calls Executive routines to
record the contents of the device registers or other hardware-supplied
information. The registers indicate the state of the device and
its
controller.
The routines also record information found in the actual
I/O request to the driver, such as the type of operation attempted.
This information aids you in the interpretation of the device error.

1.3.3

Interrupt Timeouts

Interrupt timeouts occur when the device that initiated an operation
fails to complete the operation within the length of time the driver
specified. Software timers that start when the transfer starts,
detect interrupt timeouts.
The system records the same information
for timeouts that it records for device errors.

1.3.4

Memory Errors

Memory errors occur when the parity bit stored with the data during a
write operation does not match the parity calculated when the data is
read. Some types of main memory use parity to ensure integrity of the
,.f:n,fp5mfi.V?n,!....!. ;~1;.}~~~~;q1::1:f:\1~~·tt~ sy~·~eiri~i. e;:x.fe.pt., t~e .pregenera ted sy~t~iµs
. 1.t),c].u~ S\1.pp;o,~::t.:·£,9.~:· l~C>,199:'t:ll9: memqry>p,ar.1·ty errors. The support is a
system generation option on RSX-llM.

1-6

CHAPTER 2
ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)

This chapter describes how to use the Error Log Task (ERRLOG) and the
Error Log Interface (ELI). Chapter 1 provided a general overview of
how ERRLOG and ELI work, along with the Report Generator
(RPT), to
form the complete Error Logging System.
ERRLOG gets event and status information from device drivers and the
executive in the form of Error Log Packets and writes the packets in
an Error Log File on disk. The executive also performs error limiting
to allow a maximum number of errors to be logged on each device before
logging stops.
ELI, the user interface to ERRLOG, includes switches to:
•

Start or stop logging or limiting

•

Change device error limits or error counts

•

Establish or change log file or backup file names

•

Display information about the error
device or of the entire system

logging

status

any

ERRLOG is the only part of the Error Logging System that must be
installed for error logging to occur.
You can install ELI when you
issue commands to ERRLOG and install RPT when you create reports.

2.1

INSTALLING ERRLOG AND ELI

To install the ERRLOG task, enter the following MCR command from a
privileged terminal or as an entry in the system startup command file:
INS $ERL IB§I)

To install ELI, enter the following MCR command from a privileged
terminal or as an entry in the system startup command file:
INS $EL I IDJ

If ELI is not installed, you can invoke it from a privileged
using the following MCR command:
RUN

$ELI (8ITJ

2-1

terminal

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
2.2

USING ERRLOG AND ELI

To invoke ELI after it is installed, issue the following
from any terminal:

MCR

command

ELI~

ELI>
You can use the ELI /SH switch to display error logging information
from any terminal.
However, you must use a privileged terminal to
execute any other ELI commands.
Enter each command on a separate line unless the
specifies otherwise.

command

description

The general format of an ELI command is:
[filespec]/switchl[ ••• switchn]
f ilespec
A device mnemonic or the name of an error log file, backup
or file to append to the current error log file.

file,

switches
Switches to set, change, or display ERRLOG operation.

(You

must

logging,

enter

specify at least one switch on each ELI c:om111and line.)

If you want to use only the ERRLOG defaults and start
the following ELI command:
/LOG (B.ITl

This command starts ERRLOG, using LB:[l,6]LOG.ERR as the default log
file and LB: [l,6]BACKUP.ERR as the default backup file. You must
specify the /LOG switch to use ERRLOG defaults.
The /LOG switch also starts error limiting to limit the number of hard
and soft errors ERRLOG records on each device before it stops logging
on that device. The default error limit, used when you begin limiting
with the /LOG switch, is five hard errors and eight soft errors for
each device. You can change these limits with the /HL or /SL switches
described in Section 2.3.2. However, you cannot use the switches to
change limits on the same command line as the /LOG switch.

2.3

ELI SWITCHES

This section describes the ELI switches and subswitches, divided
four types:
•

Logging switches

•

Limiting switches

•

File naming switches

•

Display switch

into

Remember that these switches only control operation of the Error
Logger.
Chapter 3 describes the RPT commands that generate actual
Error Log Reports. Chapter 5 describes the commands that control the
Control File Language Compiler.

2-2

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
Table 2-1 summarizes the ELI switches in alphabetical order.
ELI
syntax requires that you specify at least two characters of a switch
name and as many additional characters as it takes to make the switch
unique.
However,
the Logging and Limiting switches are called /LOG
and /LIM to make their names easier to remember.
Table 2-1
ELI Switches and Subswitches
Switch

Function

Subswitch

filespec/AP
(Append)
/DE
(Delete)

Appends the specified file
the current Error Log File.

Deletes the
specified
after appending
it
to
current Error Log File.

file
the

Sets the name for a backup file
to the next highest version of

f ilespec/BA
(Backup)

the file named.
device(s)/HL:n
(Hard Error Limit)

Set
limits
for hard
(unrecoverable) errors on a device.
You can use /SL, the Soft Error
Limits switch,
on the
same
command line.

/LIM
(Limiting)

Starts
the
use
of
error
limiting, using either default
limits or
those set with ELI
switches.
The
/LOG
switch
begins
error
limiting
by
default.

/-LIM
/NOL IM
(No Limiting)

Stops the use of
ing.

[filespec]/LOG
(Logging)

Begins error logger operation,
turns on error limiting
by
default,
and, if you specify a
file
name,
overrides
the
default name of the error log
file (LB: [l,6]LOG.ERR).
I f the
error log file already exists,
the /LOG
switch
uses
the
existing file.

error

limit-

/-LIM

Turns off error limiting while
the error logger is running.

/NV
(New Version)

Creates a new version of the
given file instead of using the
current version.
Stops error
logger operation
and turns off error limiting.

/-LOG
/NO LOG
(No Logging)

(continued on next page)

2-3

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
Table 2-1
(Cont.)
ELI Switches and Subswitches
Switch

Subswitch

Function

device(s)/RE
(Reset)

Resets the QIO and error counts
on the specified devices to
zero.

device(s)/SH
(Show)

Displays
error
logging
information for the specified
devices.
(If
you
do not
specify
device
names,
/SH
displays information for all
error logging devices on the
system.)

device(s)/SL:n
(Soft Error Limit)

Sets limits for
able) errors on
can use /HL,
Limit switch,
command line.)

f ilespec/SW
(Switch)

Copies current error log file
to the specified
file
and
transfers logging to that file.

2.3.1

soft
(recovera device.
(You
the Hard Error
on
the
same

/DE

Deletes the old file after the
/SW switch performs the copy
operation.

/NV

Creates a new version of the
specified
file
instead
of
appending data to the current
version.

Logging Switches
[filespec]/LOG
/-LOG
/NO LOG

ELI Logging Switches start or stop logging
devices in the system.
(See Table 2-2)

all

error

logging

Table 2-2 lists the device modules included in
the
original
LB: [l,6]ERRLOG.ULB as distributed with the Error Logging System.
However, if you have deleted any device modules from this ULB, using
the indirect command file described in Appendix A, your system will
not include support for those devices.
If you want error logging
support for the devices listed in Table 2-2, the Control File Module
listed with the device must be included in the ULB.
See Appendix A
for information on how to include and delete modules from the ULB.

2-4

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
Table 2-2
Error Logging Devices
Device

Control File Module

MLll
RK03/RK05
RK06/RK07
RL01/RL02
RMOS
RM02/RM03
RA80/RA81

EMLll
ERK OS
ERK67
ERL12
ERMOS
ERM23
MSCP80
MSC PAT
MSCPCE
MSC PEN
MSC PTO
DEVUDA
MSCP60
MSCPAT
MSCPCE
MSC PEN
MSC PTO

RA60

DE VU DA

MSCPSD
MSC PAT
MSCPCE
MSC PEN
MSC PTO
DE VU DA
ERM80
ERP07
ERP23
ERP456
ERSll
ERS34
ERXOl
ERX02
ET All
ETCll
ETSll
ETU58
ETU77
ET1645
ETU60
ET0310
ETSVOS

RC25/RD51/RX50

RM80
RP07
RP02/RP03
RP04/RP05/RP06
RSll
RS03/RS04
RXOl
RX02
TAU
TCll
TS11/TU80
TU58
TU77
TU16/TE16/TU45
TU60
TS03/TE10/TU10
TSVOS

The /LOG switch begins error logging operation and optionally allows
you to specify a file in which the error logger writes the data it
collects.
(See the file naming section below.)
If you specify an
existing file,
the /LOG switch appends new data to that file unless
you also specify the New Version switch (/NV) in the command line.
The /LOG switch also turns on error limiting, by default,
specify the No Limiting (/-LIM) switch to override it.
The NOLOG (/-LOG)
error limiting.

switch stops error logging and,

2-5

unless

default,

you
stops

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
/LOG Subswitches:
You can use the following subswitches on a command line with the
switch:

/LOG

/-LIM[IT]
The /-LIM subswitch turns off error
limiting.
This
subswitch overrides the default ERRLOG operation in which
/LOG automatically turns on error limiting.
/NV
The /NV subswitch causes the error logger to create a new
version of the error log file (either the file you specify
in the command line or the default error
log
file
LB: [l,6]LOG.ERR).
This subswitch overrides the default
operation in which the /LOG switch appends data to the
current version of the error log file.

2.3.2

ERROR Limiting Switches

The following switches control the error limiting operation of ERRLOG.
You can use them to start or stop error limiting or to change error
limits on specific devices. When a device reaches the user-specified
~LLvL

limit or Lhe Jefault error limit, ERRLOG displays the following

warning message on the console terminal or on any
allocated or attached the device:

that

has

In the message, xxxx is the type of limit (hard or soft) and n is
number to which the limit is set.

the

ERR LOG

**WARNING:

Device dd:

terminal

Exceeded (xxxx) Limit (n)

When the device reaches an error limit, error logging for that type of
error stops on the device until you reset the error and QIO counts to
0 or raise the error limit.
You can reset the error and QIO counts to zero with the ELI /RE
switch.
Mounting or dismounting the device or rebooting the system
also resets the error and QIO counts to zero.
However, using the
/-LOG switch to stop logging does not reset the error and QIO counts.
Logging on a device stops only when the device reaches both of the
limits set for hard and soft errors.
If, for example, the device
reaches its limits for hard errors but not for soft errors,
it will
continue to log soft errors until the soft error limit is also
reached.

2.3.2.1

The Limit Switch

/LIM
/-LIM
/NOL IM
The /LIM switch starts or stops use of error limits. These limits are
set by default for all devices on the system when you enable error
logging or they are set for individual devices with the hard and soft
limit switches described below.
The /LIM switch does not activate
error logging if it is not currently active on the system.
2-6

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
When you specify the /LOG switch to begin error logging,
it
automatically starts error limiting on all error logging devices
unless you inhibit limiting with the /-LIM switch.

2.3.2.2

The Hard Limit Switch

devl: [, ••• devn] :/HL:n
The /HL switch sets limits for the number of hard errors that error
logging records on the device specified.
Hard errors occur on a
device when an I/O operation fails and cannot be recovered by the
device driver.
You can set hard error limits for more than one device
in the same command line, as long as the limits are the same.
The
default hard error limit on each device is five.
The value n can be 0 to 255.
If you set the limit to 255,
logging
continues without stopping (the limit is infinite).
If the limit is
set to 0, no errors will be logged.
Subswitch:
You can u~e the following switch as a subswitch on a command line with
the /HL switch:
/SL:n
In this way, you can set both hard and soft error limits for
devices on the same command line.

2.3.2.3

The Reset Switch

devl:[, ••• devn:]/RE[SET]
The /RE switch resets the QIO count and error count for the specified
devices to zero.
You can specify up to 14 devices in one command
line. You cannot reset QIO and error counts on all devices in the
system at once by specifying the /RE switch without specifying
devices.
When ERRLOG resets the counts to
message on the Console Terminal:

zero,

displays

the

following

ERRLOG -- Error and QIO counts reset for ddnn:

2.3.2.4

The Soft Limit Switch

devl: [, ••• devn:]/SL:n
The /SL switch sets limits for soft errors. Soft errors occur on a
device when an I/O operation fails, but succeeds in a subsequent retry
attempt. You can set soft error limits for more than one device in
the same command line, as long as the limit is the same. The default
soft error limit for each device is eight.
The value n can be 0 to 255.
If you set the limit to 255,
logging
continues without stopping (the limit is infinite).
If the limit is
set to 0, no errors will be logged.

2-7

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
Subswitch:
You can use the following subswitch on a command
switch:

line

with

the

/SL

/HL
In this way, you can set both hard and soft error limits for
devices on the same command string.

2.3.3

File Naming Switches

The following sections describe switches that establish and change the
names of Error Log Files and Backup Files.

2.3.3.1

The Log Switch

[filespec]/LOG
The /LOG switch, which also initializes the error logger, sets the
name of the error log file that the error logger uses.
If you specify
an existing error log file, the default operation is to append data to
the current version of that file. To override the default, specify
the /NV switch. Tne error logger ~nen creates and writes data in a
new version of the file.
This switch does not work when error logging
is already active on your system.
The default error log file
specification is LB: [l,6]LOG.ERR.
The /LOG switch also specifies
LB: [l,6]BACKUP.ERR as the backup file. See Section 2.3.3.4 for more
information.

2.3.3.2

The Append Switch

filespec/AP[PEND]
The /AP switch appends the specified file to the end of the current
log file.
Error logging must be active for this switch to work.
The default operation is to append the specified file to
error log file and to keep the appended file.

the

current

Subswitch:
You can use the following subswitch on the command line with
switch:

the

/AP

/DE[LETE]
The /DE subswitch causes the error logger to delete the
specified file after it copies the file to the end of the
current error log file.

2.3.3.3

The Switch Switch

[filespec]/SW[ITCH]
The /SW switch copies the current error
specify and begins logging in that

2-8

log file to the file you
file.
The default operation

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
appends data to an existing version of the file and preserves the
version of the error log file.

old

Subswitches:
You can use the following subswitches on the command line with the /SW
switch:
/NV
The /NV subswitch creates a new version of the file you
specify.
This subswitch overrides the default operation in
which the /SW switch appends data to the latest version of
the file.
/DE[LETE]
The /DE subswitch causes the error logger to delete the
current error log file after it copies the file to the new
file you specify.

2.3.3.4

The Backup Switch

filespec/BA[CKUP]
The /BA switch specifies the file to be used as a backup file
if the
Error Logger cannot write to the current log file. By default, the
backup file is LB: [l,6]BACKUP.ERR.
The backup file specification is kept, but no file
is created until
needed.
You may wish to have your backup file on a different device
from the current log file. By default, both files are on pseudo
device LB:.
When the Error Logger cannot write to the current log file, it creates
and opens the backup file and writes to it.
At that point, you no
longer have a backup file, and the Error Logger displays the following
message on the Console Terminal:
ERRLOG

Log file error - logging continuing on backup file

After error logging switches to the backup file, there is no longer
backup file available.

The error logger uses the specified backup file as the current error
log file.
It does not rename the file to LOG.ERR, even though the
file is now the error log file.
At this point, you should specify a new backup file,
using the /BA
switch. Otherwise, if error logging cannot write to the new log file,
it will not be able to continue by writing in a backup file.
If the error logger tries to switch logging to
file, it displays the following message:

nonexistent

backup

ERRLOG -- Backup file error - logging discontinued
When that happens, logging stops and must be restarted.
If you create the backup file on a disk other than the disk containing
the error log file, this ensures that logging will continue even if
the disk with the error log file develops problems.

2-9

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
2.3.4

Display Switch

The /SH switch allows you to display
error logging on the system.

information

the

status

[devl, ••• devn]/SH[OW]
/SH[OW]
The /SH switch allows you to display information on the status of
error logging on the system. The /SH switch displays Error Logging
information on the devices specified (up to 14). If the command does
not specify devices, the Error Logger displays information on all
error logging devices in the system.
Example 2-1
illustrates the
output from the operation of the /SH switch:
Example 2-1

Error Logging Status

Error Logging Status
Logging: On

12-JAN-82 00:51:54

Limiting: On

Log File: LB:[l,6]LOG.ERR

File ID: DR3: 32,252

Backup File: LB:[l,6]BACKUP.ERR
Device
Name

Hard Error
Count/Limit

Soft Error
Count/Limit

QIO

MMO:
MMl:
MM2:
MM3:

0./5.
0./5.
0./5.
0./5.

0./8.
0./8.
0 ./8.
0 ./8.

23.
9776.

DBO:
DBl:
DB2:

0./5.
0./5.
0./5.

0 ./8.
0 ./8.
* 8./8.

14144.

DRO:
DRl:
DR2:
DR3:

0./5.
0./5.
0./5.
0./5.

0 ./8.
0 ./8.
0 ./8.
0./8.

o.
o.
164234.
625364.

DSO:
DSl:

0./5.
0./0.

0 ./8.
0./0.

DKO:
DKl:

0./5.
0./5.

0 ./8.
0./8.

OMO:
DMl:

0./5.
0./5.

0 ./8.
0./8.

DLO:
DLl:

0./5.
0./5.

0 ./8.
0 ./8.

DTO:
DTl:
DT2:
DT3:

0./5.
0./5.
0./5.
0./5.

0 ./8.
0 ./8.
0 ./8.
0 ./8.

DYO:
DYl:

0./5.
0./5.

0 ./8.
0./8.

1.
1.

DDO:
DDl:

0./5.
0./5.

0 ./8.
0./8.

o.
o.

2-10

Count

o.
o.
o.

46528.

130.

o.
o.
o.
o.
o.
o.
0.

o.
o.

(Offline)

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
If you specify device names in the /SH switch, the output is the same
as Example 2-1, except that the display only includes information on
the devices you specified.
The asterisk next to the soft error limit for
DB2:
indicates that
DB2:
reached the soft error limit and logging of soft errors stopped.
Note that the logging of hard errors will continue on DB2: until
the
hard error limit is reached.
The display continues to record additional QIOs on the device,
even
after logging stops because the Executive maintains the QIO count.
Therefore, the ratio of errors to QIOs on the device
necessarily give you a statistical error percentage=

2.4

does

not

ERRLOG AND ELI MESSAGES

ERRLOG displays messages on the console terminal when errors occur
during an operation.
In some cases, ERRLOG displays messages on any
terminal that has allocated or attached the device on which the error
occurs.
ELI displays messages on the terminal that invoked it.
This
section describes the messages,
their causes, and possible user
response.

2.4.1

ELI Messages

ELI -- ERRLOG not installed
Explanation:

ERRLOG is not installed on the system.

User Action:
Install ERRLOG from a privileged terminal and issue
the ELI command again.
ELI -- Failed to communicate with ERRLOG
Explanation:
ELI could not communicate
Executive directive (SMSG$).
User Action:

Fatal error.

with

ERRLOG

using

the

No user action is possible.

ELI -- File name must be specified
Explanation:
You used a Backup, Append, or Switch switch without
specifying a file name.
User Action:
Reenter the ELI command with
specification.

appropriate

file

ELI -- Get Command Line error
Explanation:

The Get Command Line procedure failed.

User Action:
operation.

This may

temporary

2-11

condition.

Retry

the

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
ELI -- Illegal switch combination
Explanation: You used an ELI switch in
combination
with
subswitches other than those allowed on a command string with
that switch.
(See Table 3-1.)
User Action: Reenter the command string, specifying a legal
combination of switches on each string. Use a separate command
string for additional switches, if necessary.
ELI -- Maximum number of devices exceeded
Explanation: You attempted to reset QIO and error counts on more
than 14 devices in one command string.
User Action:
less.

Specify the /Reset Switch again, with 14 devices or

ELI -- Switch requires device name (ddnn:) only
Explanation: You specified both a device name and UFO
file name an ELI switch that only accepts a device name.
User Action:

Reenter the command;

and/or

omit the UFD and file name.

ELI -- Syntax error
Explanation: You used an illegal switch or file specification or
made some other syntactical error.
User Action: Reenter
string syntax.

2.4.2

the

command,

using

the

proper

command

ERRLOG Messages

ERRLOG -- Backup file error - logging discontinued
Explanation: ERRLOG encountered an error when it wrote in the
log file.
It then tried to write in the backup file, but could
not. This error occurs if you fail to establish a new backup
file after ERRLOG switches logging to the backup file.
User Action:
Issue an ELI /BA command to establish a new
file and restart logging.

backup

ERRLOG -- Device not in system
Explanation: ERRLOG tried to use a device that
system configuration.

not

the

User Action: Check to be sure you specified the correct device
and reenter the command.
If the device is correct, no user
action is possible.

2-12

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
ERRLOG -- Error and QIO counts reset for ddnn:
Explanation:
reset.

The error and QIO counts for a

User Action: No user
informational message.

action

given

necessary;

device

were

This

ERRLOG -- Error Log packet too long
Explanation: ERRLOG encountered an Error Log Packet that was too
large. The error log packet was corrupt.
User Action: If the Error Logging System includes user-generated
error log packets, check the code to make sure none of the
packets are too long. Otherwise, submit an SPR.
ERRLOG -- Failed to assign LUN
Explanation: ERRLOG tried to assign a Logical Unit Number to a
terminal to send a notification message and the assignment
failed. This occurs when a device exceeds the error limit set
for it and ERRLOG tries to notify the terminal or task that has
the device allocated or attached.
User Action: No user action is necessary.
The
limiting
operation succeeded. This informational message tells you ERRLOG
was unable to notify the allocating terminal.
ERRLOG -- File I/O error
Explanation: ERRLOG tried to execute a Switch or Append command
and could not open the new file or copy the old file to the new
one. When this error occurs, logging continues in the original
log file.
User Action: No action is required to continue
the Switch or Append command.

logging.

Retry

ERRLOG -- Log file error - logging continued on backup file
Explanation: An error occurred when ERRLOG tried to write in the
Error Log File.
The logging operation transferred to write in
the backup file. The backup file becomes the log file, but
retains the given name.
User Action: Issue an ELI command to establish a new backup
file.
Otherwise, if ERRLOG gets an error when it writes in the
new file (the previous backup file), it will not find a backup
file to use.
ERRLOG -- Logging already active
Explanation: ERRLOG received an ELI
when logging was running.
User Action:

command

begin

logging

No user action is necessary to continue logging.

2-13

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
ERRLOG -- Logging initialized
Explanation: When ELI starts ERRLOG operation, using the /LOG
switch, ERRLOG displays this message on the Console Terminal.
User Action: No user
informational message.

action

necessary.

This

The ERRLOG task is not currently

running

your

ERRLOG -- Logging not active
Explanation:
system.

User Action:
Issue an ELI /LOG
terminal and retry the operation.

command

from

privileged

ERRLOG -- Logging stopped
Explanation: When ELI stops ERRLOG operation, using the /-LOG
switch, ERRLOG displays this message on the Console Terminal.
User Action: No user
informational message.
ERRLOG -- No data
Explanation:

action

necessary.

This

subp~cket

ERRLOG tried to use a corrupted data subpacket.

User Action:
If the Error Logging System includes a user-written
control file module to generate error log packets, check the
code. Otherwise, submit an SPR.
ERRLOG -- No device subpacket
Explanation:

ERRLOG tried to use a corrupted device subpacket.

user Action:
If the Error Logging System includes a user-written
control file module to generate error log packets, check the
code.
Otherwise, submit an SPR.
ERRLOG -- Privilege violation
Explanation: You tried to issue a privileged ELI command (to set
or change ERRLOG operations)
from a nonprivileged terminal.
Nonprivileged users can only issue ELI Show commands.
User Action:
commands.

Log

privileged

terminal

and

issue

ERRLOG -- Task subpacket corrupted
Explanation:

ERRLOG tried to use a corrupted task subpacket.

User Action:

Submit an SPR.

2-14

the

ERROR LOG TASK (ERRLOG) AND ERROR LOG INTERFACE (ELI)
ERRLOG -- Unable to open file
Explanation:
ERRLOG could not open the
logging.
ERRLOG then transfers logging
immediately.
User Action:
file.
ERRLOG

Issue an ELI command

log file to begin
to the backup file

establish

unknown

new

backup

command

packet

-- Unknown command packet subtype

Explanation:
subtype.

ERRLOG

encountered

User Action:
If the Error Logging System includes a user-written
control file module to generate Error Log Packets, check the
code. Otherwise, submit an SPR.
ERRLOG -- **WARNING:

Device ddnn:

exceeded xx Limit (x)

Explanation: Device ddnn exceeded the error limit set with an
ELI Hard or Soft Limit switch or the default error limit of five
hard errors and eight soft errors.
User Action:
Check to see if the number of errors indicates a
serious hardware malfunction. To continue logging on the device,
reset the QIO and error counts to zero with the /Reset switch or
change the limits using the /HL or /SL switch.

2-15

CHAPTER 3
REPORT GENERATOR TASK (RPT)

This chapter describes how to use the Report Generator Task
create Error Log Reports.

(RPT)

Chapter 1 provided an overview of the interaction of elements in the
Error Logging System (the Error Log Control File and the Control File
Language Compiler). The RPT switches described in this chapter use
modules from the Error Log Control File to determine how to interpret
and format information from the error log file.
(See Chapter 2 for a
description or how the Error Logger creates the error log file.) RPT
and modules in the error log control file work together to interpret
the information in the error log file and define an event that occurs
on a device. They do not analyze the event itself or attempt to
diagnose hardware failures.
All RPT reports use the same entry number to refer the same Error Log
Packet, so you can use RPT brief reports to isolate a device or
specific events occurring on that device, and then specify entry
numbers to generate a full report on only the specific events you want
to look at in more detail. Note, however, that some ELI commands may
change the packet number associated with an event. For.example,
appending a file to the error log file will change the packet numbers
in the appended file.

3.1

INSTALLING AND RUNNING RPT

Since RPT is a nonprivileged task, any user can use it to create Error
Log Reports when it is installed on the system.
To install RPT, enter
the following MCR command from a privileged terminal or as an entry in
the system startup command file:
>INS

$RPT~

If RPT is not installed, you can invoke it from
the following MCR command:
>RUN
RPT>

any

terminal,

using

$RPT~

To invoke RPT when it is installed, issue the
from any terminal:
>RPT8filJ
RPT>

3-1

following

MCR

command

REPORT GENERATOR TASK (RPT)
3.2

USING RPT TO CREATE ERROR LOG REPORTS

The Error Log Control file needs at least two types
from RPT switches to generate Error Log Reports:

•

How to select which Error Log Packets to analyze

•

How to format the Error Log Packets

Switches on the RPT command line provide this
information,
independent of the file specification they accompany.

3.2.1

information

which

The RPT Command Line

The only element you must specify in an RPT command line is the equals
sign
(=).
All other file and switch specifications in the command
line are optional.
The general format of an RPT command line is
[reporttile[/switches]]=[inputfile[/switches]]
reportf ile
The name of the listing file that contains the Error Log Report.
Instead of a report file, you can specify TI:
to send the report
to your terminal.
On RSX-llM-PLUS systems with transparent
spooling you can specify LP:
to send the report to the line
printer.
switches
Optional switches to control how RPT selects,
formats,
and
(on
RSX-llM-PLUS)
summarizes
information from the error log file.
You can use the same switches with either the report file
specification or the input file specification on the command
line. RPT uses the switches
in the order you specify, but
ignores which file specification they accompany.
input file
The only input file you can specify in the command line is
Error Log File, the disk file that the Error Logger creates.

the

RPT also uses a universal library of compiled control file modules as
input.
RPT looks first for the file LX: [l,6]ERRLOG.ULB.
If it does
not find it, RPT looks for the file LB: [l,6]ERRLOG.ULB.
Use pseudo
device LX:
if you wish to save space on LB:.
RPT includes this file
by default and you cannot specify or change it from the command line,
so it is not part of the format described above.
RPT can, however, prompt for the name of a universal library.
If you
want RPT to prompt you for the universal library name, you must edit
the RPTBLD.BLD file and make the value of USERCM non-zero, then relink

3-2

REPORT GENERATOR TASK (RPT)
RPT.
If you do make this alteration, note that it has the additional
effect of preventing you from issuing an RPT command line from the MCR
level. That is, the following is the way to invoke RPT.
>RPT
CTL> (universal library filespec)
RPT> command line
The RPT input and output files described above assume the defaults
listed in Table 3-1, unless you specify otherwise in the command line.
Table 3-1
RPT File Specification Defaults
Report File

Universal
Library Filel

Input File

Device:

SYO:

LB:

LX:, LB:

UIC:

Current UIC

[1,6]

File Name:

ERRREPORT

LOG

ERRLOG

File Type:

.LST

.ERR

.ULB

Version:

new

latest

1. Not specified by user.

3.2.2

Using Multiple Qualifiers in RPT Command Lines

You can only specify each RPT switch once in a command line. However,
some switches provide an alternative syntax that allows you to specify
more than one argument for the switch.
To specify more than one argument for an RPT switch, use the following
command syntax:
/switch: {qualifierl,qualifier2 ••• qualifiern)
The parentheses, which are a required part of the command syntax,
allow RPT to use more than one qualifier for the switch.
If you do
not specify the parenthesis, RPT displays the following message on
your terminal and exits:
ERLCNF-F-SYNTAXERR command line syntax error
For example, to specify a report on more
following RPT switch:

than

one

devicer

use

the

/DE: (DB,DM2:,DR)
RPT generates a report on all the DB and DR devices in your system, as
well as device DM2:.

3-3

REPORT GENERATOR TASK (RPT)
The switches that permit you to specify multiple
way are:

3.2.3

qualifiers

The DEVICE switch

The PACKET switch

•

The SERIAL switch (one drive and one pack serial number)

The TYPE switch

this

Using the Default RPT Command Line

To use the RPT default command line, enter the following command:
RPT>= ifil]
This command causes RPT to use the file specification defaults (listed
in Table 3.1)
and switch defaults (listed below).
In general, this
command creates a brief format report, without any summaries, using
all of the Error Log Packets in the error log file.
The RPT default command line invokes the following switches:
/F[ORMAT] :B[RIEF]
Creates a brief format report containing one line for each
log packet described in the report.
(See Section 3.3.2.)

error

/T[YPE]:A[LL]
Creates a report on packets describing all
types of events:
peripheral, processor, memory, control, and system information
packets.
(See Section 3.3.1.5.)
/DA [TE] :R [ANGE]:*:*
Creates a report on packets of all dates.

(See Section 3.3.1.1.)

/DE[VICE] :ALL
Creates a report on all error logging devices in the system
/PA[CKET] :*:*
Creates a report for all packet numbers.

(See Section 3.3.1.3.)

/Sl1[f"JMARY] :R[ONE]

:'noes not ,create summary reports of statis.tfoa1 info,rma,tioti' on. all
packets included ·in the report. This switch is' available only. on·
.RSX'.7'.'.llM-P;LQS syst(:!llls•
(See ·section 3. 3.•.'J•J
/W[IDTH] :W[IDE]

Creates a wide (132 column)

report.

3-4

REPORT GENERATOR TASK (RPT)

3.3

RPT REPORT SWITCHES

This section describes the RPT switches, according to the RPT
These switches tell RPT how to perform
requirement that they fulfill.
the following tasks:
•

Select packets

•

Format packets

•

Summarize information from packets (on RSX-llM-PLUS only)

RPT syntax only requires that you specify enough characters in a
command or qualifier to make it unique. For example, you can specify
/T for the /TYPE switch, but you must specify /SU for the /SUMMARY
switch to distinguish it from the /SERIAL switch.
The command line examples used throughout this chapter highlight the
command they describe.
Any switches not explained in the command
descriptions assume the default values described in Section 3.2.3.
Table 3-2 summarizes the RPT report switches in alphabetical order.
Table 3-2
RPT Report Switches and Subswitches
Switch

Qualifiers

/DA:quali f ier
(Date)

P [REVIOUS] : ndays
R[ANGE] :start:end
T[ODAY]
Y[ESTERDAY]

Select packet based on date.

/DE:qualifier
(Device)

device name ( ,s)
ALL
-

Select packets based on device.

Function

Describes how RPT formats the error
log packets.

/F:qualif ier
(Format)

/P;qualifie:r
(Packet number)

B [RIEF]

Display
packets in brief
(one line for each packet).

F[ULL]

Display all of the information
the specified packet.

N[ONE]

Does not display information on a
packet-by-packet basis.

R[EGISTER]

Displays the
same information as
the FULL qualifier, but shows only
the device registers on packets for
peripheral errors.

nnnn.nnn

Select
number.

nnnn.nnn[:mmmm.mmm]

packets

based

format
in

packet

(continued on next page)

3-5

REPORT GENERATOR TASK (RPT)
Table 3-2
(Cont.)
RPT Report Switches and Subswitches
Switch

Qualifiers

Function

/R:qualifier
(Report)

DAY
MONTH
SYSTEM
WEEK
user_string

Invokes
a predefined string
of
switches
for
RPT
to use. The
qualifier
can be
one of
four
DIGITAL-defined strings or a userdefined switch string.

/SE:qualifier
(Serial number)

D [RIVE] :number
P[ACK]:number

Selects
packets
based on drive
and/or pack serial number. The pack
serial number is supplied only on
MSCP and last-track devices •
.selects.· , · t1Je · typt;{:of s wnmary .rE'.prirf·
. RPT: ·generates· ·
{bn
RS:>C:-::llM•.;.Pt.os' ·
sy·ste;mp prtlyJ ~···

alt summ~:ry r,epoirt~
to'ry• {.)E!r r:<;>fr:,.. a f1<::l; ·~~t.ry'.L!:

selects

Creates a summary
on devi:ce errors.

r~port .basedopl~.

qrei3t~~. a' spmmary .r ~~~pr.t; based,pn:
dis'k: g~o~t,ry ($1ect,o·r; tor •. t;rack,,.· ·for.
e:x,p~;p:~~). ~··~:---~,,?

/T~qu~l i fiP.r

Selects
type.

(Type)

packets

v· .,

based

packet

A[LL]

Selects all packets in the Error Log
File.

C[ONTROL]

Selects command packets from
Error Log Interface (ELI).

E[RRORS]

Selects packets from the
memory, and peripherals.

M[EMORY]

Selects packets from events that
occur in memory
(such as memory
parity errors).

PE[RIPHERAL]

Selects packets from all peripheral
devices that support Error Logging.
This qualifier does not
display
system information
(such as mounts
or dismounts).

the

processor,

(continued on next page)

3-6

REPORT GENERATOR TASK (RPT)
Table 3-2
{Cont.)
RPT Report Switches and Subswitches
Qualifiers

Switch
/T:qualifier
(Type) (Cont.)

Function

PR[OCESSOR]

Selects packets from events that
occur in the CPU, such asunknown
interrupts.

S [YSTEM_INFO]

Selects packets from events that
occur on the system but are not
specifically tied to a single piece
of hardware (such as time changes,
system service messages, mounts and
dismounts) •

/V:volume label
(Volume label)

Selects
label.
N[ARROW]
W[IDE]

/W:qualifier
(Width)

packets

based

volume

Selects the width of the report RPT
creates (80 or 132 columns).
The
narrow width qualifier is ignored on
summary reports.

Packet Selection Switches

3.3.1

The following switches tell RPT how to select which Error Log Packets
to report on.
This selection is based on an attribute of the device
or the packet or on the date and time that the packet was created.

3.3.1.1

The Date Switch

/DA[TE]:qualifier
QUALIFIERS:
P[REVIOUS] :n days
R[ANGE] :start date:end date
T[ODAY]
Y[ESTERDAY]
DEFAULT:
/DA:R:*:*
The /DATE switch allows you to select packets based on
an event occurred.
This switch includes qualifiers to
of dates or to specify a particular day.
DIGITAL also
strings to use with the /REPORT switch that use the
,.....,..o=ii .... o

V.01""\.l"'"\V-I-~

""'.&...~~'-'Iii;

1--.;;;:;l:''-'.L.\...~

-i=".,..
.L.VJ...

4-h,.....,

'-1.lC

r"\.¥,..,."l::T; ,.....,~
tJ.L'lli;;;;"V..LVU.:::>

.. _.,. _ _ ~

,..._

,,,_,....._.,_\...

WCC1'.

V.L

UIV1Jl-1Je

the date that
specify a range
supplies switch
/DATE switch to

The RANGE qualifier accepts starting and ending dates in the
RSX format:
DD-MMM-YY
(DD-MMM-YY HH:MM:SS)

3-7

standard

REPORT GENERATOR TASK (RPT)
However, if you specify the second format, with time as well as
the parentheses are a required part of the syntax.

date,

When you use the starting date and ending date format,
the starting
date rounds off to a time of 00:00:00 and the ending date rounds off
to 23:59:59.
The asterisk (*) used at the beginning of a range specification
indicates any date through the specified ending date. For example,
*:12-JAN-82 specifies all of the packets from the beginning of the
error log file through January 12, 1982.
The asterisk (*) used at the end of a range specification indicates
any date since the specified beginning date. For example, 4-FEB-82:*
specifies all of the packets from 00:00:00 on February 4, 1982 through
the end of the error log file.

The Device Switch

3.3.1.2

/DE[VICE] :qualifier
QUALIFIERS:
device name(,s)
ALL
DEFAULT:
/DE:ALL
The /DEVICE switch allows you to select packets for a particular
device, for more than one device, or for all the devices on the
system. You can specify more than one device w1tn the /DEVICE switch
by using the special syntax described in Section 3.2.2.
RPT uses the following conventions for device names with
switch:

the

/DEVICE

Meaning

Mnemonic
dd

Selects all devices with the mnemonic dd.

ddnn:

Selects the device with the mnemonic dd and the
number nn.

For example, /DE:DM selects all DM devices, and /DE: (DM,DB2:)
all DM devices and device DB2:.

3.3.1.3

unit

selects

The Packet Switch

P[ACKET] :nnnn.nn[:mmm.mm]
DEFAULT:
/P:*:*
The /PACKET switch allows you to select a packet or range of packets
by specifying the packet identification numbers. You can determine
the packet numbers you want to see by examining a brief report of all
packets.

3-8

REPORT GENERATOR TASK (RPT)
To select just one packet you specify one packet number. For example,
/PA:l23.4 selects only packet number 123.4.
To select a range of
packets, you specify the first and last packet numbers of that range:
/PA:l23.4:432.l selects all the packets from packet 123.4 through
packet 432. 1.
You can also specify more than one packet or packet range by using the
special syntax described in Section 3.2.2.
The asterisk (*) indicates an open-ended number.
You can select all
the packets before a particular number (*:235.3), or all the packets
after a particular number (235.3:*).

3.3.1.4

The Drive and Pack Serial Number Switch

/SE[RIAL]:qualifier
QUALIFIERS:
D[RIVE]:serial number
P[ACK] :serial number

{D[RIVE] :serial number,P[ACK]:serialnumber)
(P[ACK] :serial_number,D[RIVE] :serial number)
DEFAULT:
None
The /SERIAL switch allows you to select packets based on their drive
or pack serial number or both.
This switch only applies to peripheral
errors. You can select packets from any device that has a serial
number by drive serial number, but you can only select packets from
MSCP and last track devices by pack serial number.
Appendix B
explains where RPT gets drive serial numbers and lists the significant
digits in serial numbers for each error logging device.
You can specify one drive and one pack serial number or both in the
same command line by using the special syntax described in Section
3.2.2.

3.3.1.5

The Type Switch

/T [YPE]: [qualifier]
QUALIFIERS:
A[LL]
C[ONTROL]
E[RRORS]
M[EMORY]
PE[RIPHERAL]
PR[OCESSOR]
S[YSTEM_INFORMATION]
DEFAULT:
/T:A

3-9

REPORT GENERATOR TASK (RPT)
The /TYPE switch selects Error Log Packets based on their packet type.
You can select the following types of packets (or combination of
types) with the appropriate /TYPE switch qualifier:
Packet Type

Qualifier
ALL

All Error Log Packets in
File.

CONTROL

Error Log Command Packets
Error Log Interface (ELI).

sent

ERRORS

All Error Log Packets from
processor, and memory.

peripherals,

MEMORY

Error Log Packets from events that occur
in memory (such as memory parity errors).

PERIPHERAL

Error Log Packets from all peripheral
devices that support Error Logging. This
qualifier
does
not
display
system
information
(such
as
mounts
and
dismounts)
for
the
devices.
That
information
is
displayed
by
the
SYSTEM INFO qualifier.

PROCESSOR

Error Log Packets from events that
in

_,..,

SYSTEM INFORMATION

f-ho

-·,.-

f"Ofl
-•

C::11t""h

_ _ _ ,.,.

>=>C

the

11nlrnl"\<.Tn
-&•~:11.,1..&-T'l'l..a

Log

the

occur

; nt-o.,...,..nnt-C"
_..._,,_.._,...I:'- .......

......

Error Log Packets from events that occur
on the system, but are not specifically
tied to a single piece of hardware
(such
as time changes, system service messages
mounts, and dismounts).

You can specify more than one type of packet by using
syntax for the /TYPE switch, described in Section 3.2.2.

3.3.1.6

Error

the

special

The Volume Label Switch

/V[OLUME]:volumelabel
DEFAULT:
None
The /VOLUME switch selects packets for peripheral errors based on
volume label.

the

For example:
=/T:PE/V:ERRLOGSYS
This command line specifies that RPT find the device or devices
containing a volume with the label ERRLOGSYS and generate a report of
peripheral errors on those devices.
Since
the
/TYPE
switch
specification did not include system information, the report will not
include mounts or dismounts for the devices.

3-10

REPORT GENERATOR TASK (RPT)
Report Format Switch

3.3.2

/F[ORMAT]:qualifier
QUALIFIERS:
B [RIEF]
F [ULL]
R[EGISTER]
N[ONE]
DEFAULT:
/F:B
The /FORMAT switch tells RPT how to format a report from packets in
the error log file.
You can select reports in brief format (one line
for each error), in full format (all the information from the error
log packets specified)
or in register format
(dumping only the
registers for device errors).
The following sections describes
qualifiers to the /FORMAT switch.

3.3.2.1

Brief Reports

/F[ORMAT] :B[RIEF]
Brief reports are short, one-line
packets.

per

packet,

reports

selected

The brief report shown in Example 3-1 displays one line of information
about each of the error log packets specified in the RPT command line.
The following list describes the sections in the brief report.
The
numbers in the list reflect the numbers of the sections in Example
3-1. Note that all these examples show wide width reports.

The Error Log Packet Entry Number which describes the
relative position in the Error Log File. This number does
not change unless the file
is changed,
by an ELI/APPEND
command,
for example.
It is not changed by normal logging
into the file.

The date and time the packet was logged.

The type of entry in the error log file;
for
or soft device errors or system information.

The device on which the error occurred.

The error type as defined by the hardware information.
RPT
does not do any interpretation of these errors;
it merely
reports the hardware information.

(i)

Any other information error logging gathers on the error,
such as the I/O function that occurred at the time of the
error.

''.9:
.

''.'-'

~;<

ln;p·ut an:d reP:or't 'fTle spe.Cl'.fications:.

3-11

example,

hard

REPORT GENERATOR TASK (RPT)

CD · The format selection.
(I!)

The packet
created).

time

range

select ion

(time

the

The volume . label selection•
The drive and pack serial number. selections.
Types of summary reports selected·.
The' packet type selections.
The packet number select'ions ..
The device selections·.·
The number of packe~s. printed and proces:s·e<I,·.
The time RPT. report. generation be9~n.and>erided'!I

3-12

packets

were

Example 3-1
RSX-llM/M-Plus Error Logging System

Version V0-1

Time Stamp

Entry Type

4.4
5.1

04-JAN-1983 09:51:00
04-JAN-1983 09:51:06

Device Hard Error
Device Hard Error

DLOOO:
DLOOO:

Cover Open
Data CRC Error

Function
Function

Read Data
Read Data

5.2
10.3
42.2

04-JAN-1983 09:51:07
14-JAN-1983 14:20:18
15-FEB-1983 14:02:23

Device Hard Error
Device Soft Error
Device Soft Error

DLOOO:
DMOOl:
MSOOO:

Data CRC Error
Data Check
Uncorrectable Data

Function
Function
Function

Read Data
Read Data
Write

Select16n Information•
w
1

Report f fl e: DB2 :, [303 ( 12] .,..n.,,.,..,,.,,.t.:,1'•¥'
Input file:
DB2:(~03~1~]

Cf)

Report fortna~ selection:
BRIEF

CD>

Packet time selection$;
From * throu1lh *

Volume label seliect1on;
(not use(})

e serial
number selections:
Drive: (not used)
Pack:

Page

e
Device

•

Entry

1--'

ll-MAR-1983 08:27:50

Error Log Brief Report

(not used)

Summary selections:
No History
No Error
No Geometry

Error Type

Additional Information

Example 3-1
Seleetion ~nformation~

Cl)

Packet type selecHons:

Processor,

Memory

System Info
Peripheral
Control

41'

Packet number i:tehctions:
4.4

5.l
5.2
6. 4 .

10.3
42.2
.~· Device 'selections:
(ALL)

Number :of paekets printed / processed:

6. I 6.

Proces~ing began cit 11~MAR~1983 18:12:02

Processing ended at ll-MAR-1983 18:12:20

(Cont.)

Error Log Brief Report

REPORT GENERATOR TASK (RPT)
3.3.2.2

Full Reports

/F[ORMAT] :F[ULL]
Full reports provide a detailed listing of device events.
They list
and interpret all of the information collected in the Error Log
Packets they describe.
The full report, shown in Example 3-2, displays the complete contents
of error log packet number 4.4, a Cover Open Error described in the
brief report in Example 3-1.
The following
list describes the
sections of the full report.
The numbers on the list reflect the
numbered sections in Example 3-2 •
.,

The same identification information listed in
the brief report description.

System identification information including operating
and base level, CPU type and address mapping type.

Device identification information including the device name,
device type, volume label, controller, unit number, pack and
drive serial numbers, total I/O count on the device, and the
number of hard and soft errors logged previous to this one.

I/O operation identification includes the terminal and UIC
that initiated the operation, the task name, the beginning
physical memory address of the I/O buffer, the length of the
I/O request
(in bytes),
the maximum number of retries the
device driver allows for an I/O operation,
the number of
retries remaining, and the actual I/O operation taking place.

CS)

I/O operation information includes the device
and type of error as defined by the hardware.

The device error position information locates the error
cylinder, group, head, sector, and logical block number.

«i)

The device-supplied information includes a dump of the device
registers according to name, contents, and interpretation of
the bits in the registers.
The *
beside
some
bit
interpretations means that the condition is likely to have
contributed to the error.
It is a sign that you may want to
examine the condition.

The following RPT command line generated the full
3-2:
RPT>EXEMPF.RPT=RAISIN.LOG/PA:4.4/F:F

3-15

items

report

I/O

1-5

system

function
by

Example

Example 3 ... 2
RSX-llM/M-Plus Error Logging System

Entry

-4.4

Sequence

Version V0-1
1.

DLOOO:

Error Log Full Report

21-JAN ..·1982 06:54:13

Device Hard Error

Page

(Cover Open)

04-MAY-1981 09:51:00

Subunit

Pack SN

Drive SN

N/A

System Identification:

System

!dent

Processor

Mapping

CPU

Format

RSX-UM-PLUS

PDP-11/70

22-Bit

CPA

1 ..

Device Identification Information:
w

.......

°'

Device

Type

Volume Label

Controller

DLOOO:

RLOl

DL A

Unit
0

I/O Count
292.

I/O Operation Identification:

•

TI:

UIC

Task Name

TTOOO:

[003,054]

Device

CoritroHer

DROOO: DR L
MMOOO..:. MM T

Address

Length

••• BAD

340000

10240.

Unit

TI:

UIC

Task

Address

Length

COOOO:
TT003:

[031, (17~]
[031,076]

DROCFl

1212360
N/A

,N/A.

.. N/A

Maximum Retries

BRU'l')

Retries Remaining

Operation
IO.RLB ! IQ.X

512.

opera.ti on
lt)~RLB
IO~SPF

Hard Errors

Soft Errors

Example 3-2 (Cont.)

Error Log Full Report

I/O Operation Information:
~

Device Function

Type of Error

Read Data

Cover Open

Device Error Position Information:

Cylinder

Group

173.

N/A

H4~ad

Sector

Block

13.

6926.

Device Supplied Information:

w
1
......
......i

Ci)

Name

Value

Interpretation

RLCS

104335

*[15
] Composite Error
[ 9: 8] Drive Selected = 0
[ 6
] Interrupt Enabled
[ 3: l] Function
= Read Data

RLBA

043000

[15: 0] Bus Address Register

RLDA

126716

[15: 7] cylinder Address = 173.
[ 5: O] Sector Address
14.

RLMPl

133333

[12: O] Word count = 9685. words remaining

RLMP2

046074

*[14
]
*[10
]
[ 6
]
[ 4
]
[ 2: O]

current Head Error
Write Gate Error
Head Address = Upper head
Heads Out (over the disk)
Drive State = Seek

*[11
] Data CRC Error
[ 7
] Controller Ready
[ 5: 4] B.T>.17,BA16
01 (B)
[ 0
] Drive Ready

[ 6

* [11

[ 7
*[ 5
[ 3

] Head Selected

Spin Speed Error
Drive Type
= RLOl
Cover Open
Brushes Home

Lower head

REPORT GENERATOR TASK (RPT)
3.3.2.3

/F[ORMAT]:R[EGISTER]
Register reports contain the same information as full reports for
all
events except those that occur on peripherals.
Register reports list
the contents of all device registers for peripherals, but contain no
other information.
The register report in Example 3-3 includes only the register
of the full report for packet 4.4 (the Cover Open Error).

section

The following RPT command line generated the Register Report shown
Example 3-3:
RPT>EXEMPN.RPT=RAISIN.LOG/F:R/PA:4.4 '.BIT;

3-18

Example 3-3
Version V0-1

RSX-llM/M-Plus Error Logging System
Entry

4.4

Sequence

DLOOO:

Error Log Register Report
Page

22-JAN-1982 08:31:15

Device Hard Error

(Cover Open)

04-MAY~1981

09:51:00

Device Supplied Information:

w
1
......

Name

Value

Interpretation

RLCS

104335

]
*[15
[ 9: 8]
[ 6
]
[ 3: 1]

RLBA

043000

RLDA

126716

Composite Error
Drive Selected = 0
Interrupt Enabled
Function
= Read Data

133333

RLMP2

046074

(B)

[15: O] Bus Address Register
[15: 7] Cylinder Address
5: O] Sector Address

[

RLMPl

*[11
] Data CRC Error
[ 7
] Controller Ready
[ 5: 4] BAl 7 ,BA16
01
[ 0
] Drive Ready

[12: O] Word Count
]
*[14
]
*[10
[ 6
]
[ 4
]
[ 2: O]

= 173.

[ 6

] Head Selected

14.

= 9685. words remaining

Current Head Error
Write Gate Error
Head Address = Upper head
Heads Out (over the disk)
Drive State = Seek

*[11
[ 7
*[ 5
[ 3

Spin Speed Error
Drive Type
= RLOl
Cover Open
Brushes Home

Lower head

REPORT GENERATOR TASK (RPT)
3.3.2.4

No Report

/F[ORMAT]:N[ONE]
RPT does not generate a formatted output report on event information.
This switch satisfies the requirement to tell RPT how to format the
packets by telling it not to format the packets or produce a
packet-by-packet report.
It . i s useful ori 'RSX-HM-PLUS systems: w,tien,
you only want
generate a summary r'eport.

3 ~ 3 .}

swrimary 'Swi tCli '(RSX-llM-P.LUS only)
/SU [MMARY'] :qualifier
QOALIFI ERS':

A [I..I~J
E [RRORJ
G[EOMETRY]
H [ IST9RY] .
N[ONE]
DEFAULT:
/SU:N
The /SUMMARY switch, which. is only available on RSX-llM-PLOS systems,
tells RPT how to summarize the information from packets in the error
log file. Since the summaries are compilations of the data gathered
from the individual packets, the /SUMMARY switch tells RPT what
particular piece of information from the packets to use as the basis
for a summary report.
RPT cannot create summary reports in narrow width.
If you specify
narrow width, with the /W:N command, RPT formats the packet-by-packet
display in narrow width, but formats the summary in wide width.
The following sections describe the summary reports you
with /SUMMARY qualifiers.

3.3.3.l

can

generate

The All Qualifier

/SU[MMARY]:A[LL]
RPT generates summary reports sorted by history, er,ror., and ·geometry.
These summary reports are described . in Sections 3. 3. 3. 2 through

3.3.3.4.

l. 3.:3.2

The Error Qualifier

/SU[MMARY]:E[RROR]
RPT generates a summary report sorted· by error type.
The, error
summary, sorted by device, shows the .number of times each error
occurred on the device. The Count column of the summary te'11's the
number, of times· the error occurr,ed. .Ex~mple 3-4 shows the summary
section of an error summa'.Fy report.··

3-20

REPORT GENERATOR TASK (RPT)

~iJ9,: £·oj,:'iowin,g: .. R~T,:fc9mm~M~ gener:atea: :t}le .~~port· 'in. ·Exa~i>.~e :;J,~4/~::
:',>RP'r .·~:RROR,RPT·:.LOG=/SU:.E/F: N

:®.

·Wlien':you s.p.ec'Hy /FORMAT:NONE, RPT does·· hot .displa:v . . ·packe·ts
pa;cke·h·by-,.pacJ<e:.t .·basis: "as .show,Q, in: the '.Pr'ev.iqus exa111ples~·:.

3-21

·. ~

Example 3-4

Error Summary .Report· ·
:xi

tzl
'"C

Error summary (sorted by device):

'I'ype·

·Drive SN

Volume Label

Pack SN

Error. Type

'count J?irst/Last Occurrence.

Entry

:xi
t-3

G'l
tzl

FFF

N/A

Data Late

Ol.,-DEC_.1981 .15:39:26
01..;.PEC.•19.Sl. 15:.40:23

N
N

N/A
RS04
MMOOl:

TU45

N/A
148

WORKVOL
WORK VOL
<null label>.

N/A
N/A

N/A

Nonexistent Drive

Ol-DEC-1981 15:50:43
Ol-DEC.;,.1981 16:21:41
Ol-DEC""'l98l 16:2h45
Ol•DEC:"."1981 16:21:45

No error bit found
CRC Error (NRZI) ,
LRC Error {NRZl)

Oi.;,.DEC..;.1981 l6J0 5:·30
01.:.nEc~19a1 '16: os: 30.
Ol.:.DEC-198i 16:05;44
Ol-DEC.:..1981 16 :08:15

2.·2

3.4
4.3
7 .. 3

e.1

8.l

tzl

:xi

t-3
0

:xi
t-3

(/l

:;ii;:

5.2
5~2

5.3
7.1

:xi
'"C
t-3

REPORT GENERATOR TASK (RPT)

;

:RP~ g~fi·e.r~:tes a_ ·•sµ~ary. 'report: based on aeviqe geom~t:ry···;·cih9id:al· ·hioc~
or secl:o~, fo.r example) • The Error Count column. of; the srimmary tehs
!

how .many times .an error occurred in that device location. · ·
''<'.

·'.,

::The :fo~l:ow.in.g. RPT .. command generated the

. >RPT :E~ROR:RPT'.:uoc:;.~ /su: G/F :N '(Bfil) . :
i

;

••

·:l

·.>,

·.

. .

•

3-23

·'·'

report'. iri EX.ampLe ·;:. 3...:5:

:ti

Example 3-5

Cll

"'0:ti
~

C>
Cll

Cll

w
1

0
:ti

.i::..

)ii

CJ)

:;ii;:

N/A
N/A

WORKVOt
WORKVOL

N/A
N/A

:ti

REPORT GENERATOR TASK (RPT)
3.; 3·. 3. 4 ·... The ·fli sto.ry ·ouai:if{¢r
/SU{MMARY-} :HUSTORYJ

RPT gen~rates a sumniary report' sotted by device er'ror history~.
It.
displays ·the hard and soft error ·c.ount ~nd QlO count for every volume
used on each device.
'

·The fo.llowing· RPT

•

co~·and line 'gener.at:ed the report· in .Example. 3-6::

>~PT ~~RRORRPi:i:'. ~ LQG=/Su·: H/E':: N. ·~J.

3-25

.,,::a
tsJ

::a

to-i

Gl
tsJ

2!
tsJ

O"I

to-i

::a
to-i

)It

'°

.,,::ato-i

REPORT GENERATOR TASK (RPT)

::3.,{~::.;',~'.·,~,,'::<':'~.e.:::~~n~,~::oual'~,tX~l(<t '/.';;'','<'~,,,

' · :, ' ', : tsu:[i.fM1'.Rt:.l ~ N'toNeJ" ::>
)

~ .?'" -~

> ,'

~·i,'. :~

~ :·

.:·

,·"

.·

.~- :·,·i"•> ..

:, ,

: ,; . , >

.'.

,, :;~?,://?':">,,:,' <f,?:',,

".',»:,::·/' ' /·',
, ,
/

~·;"

, / , ,..

.. . . • i
;

'.:''

<' • ,.
l

;,''

~P'P'.··9,oes·.n~t.·'·9,~H1:et.a·#e:,:. ·:a;•··'· .~UinmarY.·:·<:r.~pt(.ft.:~.:',/'.':•,;,a~~e~e,~I.',, /'tili'$',,,-,~u~~}1'i''~;f'.'.

satisf ie~ '..·the · ·RJ?T · , ,requirem,e,nt, o:rtcRSX... :VJ::M~PEUS that ,th¢' 'c9mmand · line,,
,spe?i,f,'y '.h<?. Vf':,·t9 s.~,~ar.tze» ~he , :~_nf,'?.~inabi:,on ·~rom, Eri:"~',!'.,.' ~~.9 ~~:tc~,r~e;/ ·

3.3.4

The Report Switch
/R[EPORT]:defined report string
DEFAULT:
None

The /REPORT switch invokes a predefined string of switches for RPT to
use.
This switch string usually defines a particular type of report,
such as a report for a particular time period.
The string contains
any legal combination of RPT switches. The string cannot include the
/REPORT switch.
The /REPORT switch allows you to access a file that contains the
switch combinations you use frequently and lets you invoke the
switches, using the string name, instead of reentering the switches
explicitly.
RPT uses the normal default values described in Section 3.2.1 for all
switches not defined in the switch string if the switches have
defaults.
The DIGITAL and user-defined switch strings are found in the Control
File Module, PARSEM, or in a disk file, LB: [l,6]ERRDEFINE.CFS,
respectively. The /REPORT switch first searches PARSEM, where it
finds DIGITAL-defined strings.
If the string is not defined there,
RPT searches ERRDEFINE.CFS.
Since RPT looks in the Control File Module first, you cannot redefine
the DIGITAL-supplied strings unless you alter the control file module.
DIGITAL does not recommend that you alter control file modules.
You
can change the definitions for DIGITAL-supplied strings by slightly
altering their names and inserting the switch under the new name in
ERRDEFINE.CFS.

3.3.4.1

Predefined Switch Strings

DIGITAL supplies four
/REPORT switch.

predefined

switch

strings

o,n,~.:~sx-:;1~-iP,tJ.i~~~s;~~.t:,~,:4.f:he.o J>;wi:t~ ,s:br'.ing,S, .•d,.eftne.:.;, ,, ·~ ..

DAY - /FO:d?U;LL/SU:·rA:I;L/DA:'fODA,Y

•

WEEK - /SU: (HISTORY, ERROR)/DA:'PREVIOUS: 7

, • , MONTH.·~ /SU:.(HISTORY,ERROR)/DA:;P:REVIOUS:Jl
e

SYStrEM - /SU: (HISTORY, ERROR)

3-27

use

with

the

REPORT GENERATOR TASK (RPT)
On RSX-llM systems, the switch strings define:

DAY - /FORMAT:FULL/DA:TODAY

WEEK - /DA:PREVIOUS:7

MONTH - /DA:PREVIOUS:31

•

SYSTEM - uses all default switches

Note that the names of the predefined switch strings must
in full.
They cannot be abbreviated.

3.3.4.2

entered

User Defined Switch Strings

You can name and define your own switch strings to use with the
/REPORT switch by creating and editing LB: [l,6]ERRDEFINE.CFS and
inserting the switch strings you want to define.
Entries in this file must be in the form:
'switchname' ,'switchstring'
Note that single quotation marks are a required part of the syntax.
s~r:i tchna~e

The name of the switch string you are defining.
This name
becomes the qualifier to the /REPORT switch when you want to
invoke the string.
(The name must be nine characters or less.)
switchstring
The RPT switches you select to generate the report.
string must be 80 characters or less.)

(The

switch

For t:0v::1mnlt:0_ i f vnn w::1nt- t-n nt:0nt:0r::1t-t:0 a brief rPnort- of nPrinhpr;:il
erro;~----~~-· aii ~ the .. DB-d~~i~~~-~~-your system, ~ait-ERRDEFINE:CFS-~~d
insert the following line:
'DB','/FO:B/TY:PE/DE:DB'
You can then create this report with the following RPT command:
RPT>outfile=infile/R:DB
When you invoke a user-defined string, you must enter the full
string name.

3.3.5

The Width Switch
/W[IDTH] :qualifier
QUALIFIERS:
N[ARROW]
W[ IDE]

DEFAULT:
/W:W

3-28

switch

REPORT GENERATOR TASK (RPT)
The /WIDTH switch allows you to set the line width of the report RPT
generates to narrow (80 columns} or wide (132 columns) • The basic
report format does not change when RPT creates a narrow report.
Instead, each long line of the report wraps onto the next line at an
appropriate place.

·.N~.£~·~··..~~~·t. ·.s#ar·~.··•;:r;~P~.;.t~;.;~~i~::~~f;v·~~.~~~~.·:.::~~:.~k[tit#)~ij?;~~~.~~~?i1i;··;:';~;.:c··t.:• ·~~~~'.i···
.porti<>rl . . 9:e .the~e.·.:r.~p()r,~s ,;\l,s.: a~~ilY:·.~· ·;W~Q~:J~~.~~·' · /;.:•· :.\·:.:'.··"
:.·;::..:J.
•..-.·>:,,·.:.
<

•

3.4

''.;

,~,:

••••

•

••

"~''''

,~'.,,f

~.',,;-,· __ ,,_ *:: .. ·;.

'" =,:,

;.,~~··:.

,:, •.,,..:::.• ,"<-:,'<""

·:,,A·... ,x.='""'' ~.,,~~;,,<::_,-;;::,,;,~·:-.,·;;,

<.;.,

:,.:··~'.·:..,w",,".l"

..b·1:. ·o:.···.'.\..:•. ::.:.:.·:

:........ ~ ··'"--'~

.,.

-,~:

::w. ~

.•

ERLCNF REPORT MESSAGES

The Error Log Control File displays messages on your terminal if
errors occur during report generation.
The messages include an
abbreviation, a severity level code for
the
error
(warning,
informational, or fatal), and text describing the error.
In some cases, RPT also writes the message in the Error Log Report, if
it explains an error that appears in the report. For example, when
RPT fails to find a control file module for a device you specify,
it
displays a message on your terminal and in the report that includes
the error message.
This section lists the ERLCNF messages, along with possible causes and
methods for recovery.
The following are Fatal ERLCNF errors:
ERLCNF-F-ARGNOTUNQ, Argument specification <argument> is not unique
Explanation:
argument to
argument.

You did not specify enough characters in a switch
make it unique.
It can be confused with another

User Action:

Check the argument syntax and reenter the command.

ERLCNF-W-BADSUBPKT, Possible corruption in the <packetname>
in item <item label>

subpacket

Explanation: RPT found something in the subpacket that appeared
to be abnormal.
The file may be corrupted or it may be an
internal error within RPT.
User Action: You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other information you have.
You can create a
dump of the packet using the starting virtual block number of the
packet: the nnn portion of the packet number nnn.m.
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-W-DUILLFORM, MSCP format code <code> is undefined
Explanation: This may be an internal error within RPT.
indicates a format code in the RASO packet that is corrupted.

If you do, send
User Action: You should never see this message.
in an SPR along with a dump of the packet that generated the
(See BADSUBPKT
message and any other information you have
description).
3-29

REPORT GENERATOR TASK (RPT)
ERLCNF-F-ILLARGCOM, Illegal argument combination
Explanation:
You specified an illegal combination
with a switch.
User Action:

arguments

Check the syntax and reenter the command.

ERLCNF-F-ILLFILSPC, Illegal file specification - <filename>
Explanation: You used an illegal file specification with an
report generating command.
User Action:

RPT

Check the syntax and try the operation again.

ERLCNF-W-ILLPACCOD, Illegal code in packet <packetid>, Code
Explanation: The major code for the indicated packet
the range that RPT can handle.

= <xx>
is

beyond

User Action:
You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-F-ILLPACRAN, Illegal packet range - LOW = <xx>, HIGH
Explanation:

to b~ packeL

The RPT Packet Selection switch requires
-

- -- - - ! .J:: .! :::itJ~l,,;.LL.LI,,;

= <xx>
arguments

.c - _.. __ - .._

LUL.llldLe

User Action:
Determine the correct number
for
the packet
want to display, check the syntax and reenter the command.
ERLCNF-W-ILLPACSBC, Illegal subcode in packet <packetid>, Code
Subcode = <xx>
Explanation: The subcode for the indicated packet is beyond
range that RPT can handle.

you

<xx>,
the

You should never see this message. If you do, send
User Action:
in an SPR along with a dump of the packet that generated the
message and any other information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-F-ILLSWTARG, Illegal switch argument - <argument>
Explanation:
RPT recognized the switch argument, but
that the argument is incorrect in the context given.
User Action:

Check the syntax and reenter the command.

3-30

determined

REPORT GENERATOR TASK (RPT)
ERLCNF-F-INTERROOl, Internal error detected at position number <n>
Explanation: This is an internal RPT error.
PARSECLST and PARSECTION error messages.

It occurs with

the

User Action: You should never see this message.
If you do, send
in an SPR and the command line that generated the message and any
other information you have.
ERLCNF-F-MODNOTFND, Module not found - <module>
Explanation:
find it.

RPT searched ERRLOG.ULB for the module and did

not

User Action: You should never see this message. If you do, send
in an SPR and the command line that generated the message. Be
sure to include the name of the module that was missing.
ERLCNF-F-MULARGSPC, Argument <argument> specified multiple times
Explanation:
once ..

You specified an

RPT

switch

argument

User Action:

Check the syntax and reenter the command.

than

ERLCNF-F-MULSWTSPC, Switch <switch> specified multiple times
Explanation: You entered the specified switch more than once on
the same RPT command line. RPT only allows you to specify each
switch once.
User Action: Check the syntax and reenter the command. use the
special syntax for multiple switch specifications described in
Chapter 3 if the switch allows it.
ERLCNF-W-NODACSPRT, No IO_ACTIVITY support, packet

= (packet)

Explanation: This is usually caused by enabling I/O activity
support on RSX-llM systems without enabling the corresponding
support in the error log control file.
If you do, send
User Action: You should never see this message.
in an SPR along with a dump of the packet that generated the
message and any other information you have
(See BADSUBPKT
description) •
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-F-NODIDPACK, No Device_ID subpacket
Explanation:

This is an internal error within RPT.

User Action: You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other information you have.

3-31

REPORT GENERATOR TASK (RPT)
ERLCNF-W-NODRIVSZ, No drive of size <size> for mnemonic <ddnn>;
EUNKWN
Explanation:

using

This may be an internal error within RPT.

User Action:
You should never see this message if you have all
DIGITAL hardware.
If you have non-DIGITAL hardware, and you
receive this message, it is caused by a disagreement between
RPT's table of device sizes and the actual size of the device.
See Section 4.5.3.4 for information on changing the table of
device sizes.
ERLCNF-W-NODRIVTYP, No drive type <type>
EUNKWN

for

mnemonic

<dd>;

using

Explanation: This may be an internal error within RPT. From the
mnemonic, the drive appears to be a MASSBUS device. However, RPT
does not recognize the device type as a MASSBUS device.
User Action:
You should never see this message if you have only
DIGITAL hardware.
If you have non-DIGITAL hardware, the error is
caused by disagreement between RPT's table of device sizes and
the
size of the actual device.
See Section 4.5.3.4 for
information on changing the table of device sizes.
ERLCNF-F-NOINPFILE.

No input file specified

Explanation:
RPT did not find an input file on the command line.
This mess~ge occurs when you failed to specify an equals (=) sign
in the command.
User Action:

Check the syntax and reenter the command.

ERLCNF-W-NONOTES, No notes available for device <devicename>
Explanation: RPT includes a facility for displayinq notes at the
bottom of Full or Register reports.- This internal-error message
indicates that a device which did not have an associated NOTES
module.
If you do, send
User Action: You should never see this message.
in an SPR along with a dump of the packet that generated the
message and any other information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-F-NOREMATCH, No predefined switch string for <string>
Explanation:
RPT did not find the defined report string you used
in a /R[EPORT]
command, either in ERRDEFINE.CFS or among the
DIGITAL-defined report strings. Remember to use the entire name
of the DIGITAL or user-defined string.
User Action:

Check the syntax and reenter the command.

3-32

REPORT GENERATOR TASK (RPT)
ERLCNF-F-OPNINPFIL, Failed to open the input file
Explanation: RPT could not open the input file specified.
This
message is accompanied by the FILERRCOD information message, that
displays the FCS error code from the file.
Oser Action: Check the FCS error code and retry after correcting
the indicated condition.
ERLCNF-F-OPNREPFIL, Failed to open the report file
Explanation: RPT could not open the report
(output}
file
specified.
This
message is accompanied by the FILERRCOD
information message, that displays the FCS error code from the
report file.
User Action: Check the FCS error code and retry after correcting
the indicated condition.
ERLCNF-F-OPNUSRFIL, Failed to open the user file

Explanation:

RPT could not open the user file

specified.

This

message is accompanied by the FILERRCOD information message, that
displays the FCS error code from the file.
User Action: Check the FCS error code and retry after correcting
the indicated condition.
ERLCNF-F-SWTNOTUNQ, Switch specification <switch> is not unique
Explanation:
You did not specify enough characters of a switch
to make it unique.
It can be confused with another switch.
User Action:

Check the switch syntax and reenter the command.

ERLCNF-F-SYNTAXERR, Command line syntax error
Explanation:
Some element of the command line does not have
correct syntax.
User Action:

the

Check the syntax and reenter the command.

ERLCNF-F-TOOFEWARG, Too few arguments in switch <switch name>
Explanation: You specified a switch that requires
arguments, without specifying enough arguments.
User Action:

one

Check the syntax and reenter the command.

ERLCNF-F-UNKNWARG, Unknown argument - <argument>
Explanation:

You specified an argument that is unknown to RPT.

Oser Action:

Check the syntax and reenter the command.

3-33

REPORT GENERATOR TASK (RPT)
ERLCNF-W-UNKNWNDEV, Device mnemonic <dd> is unknown;

Explanation:

using EUNKWN

This may be an internal error within RPT.

User Action:

You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other
information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-W-UNKNWNNOT, No note number <number> for device <devicename>

Explanation: RPT includes a facility for displaying notes at the
bottom of reports.
This internal error message indicates that a
device tried to print a note that was not available.
User Action:

You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other
information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered,
or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-F-UNKNWSWT, Unknown switch - <switchname>

Explanation:

You specified an unknown RPT switch.

User Action:

,,"'"'""',.....i,..
VL.l\,,,;"""J~

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

,.. .. ,. ............ _ ••

_ ..... ...:.1

_ _ _ _ .._ _ _

\.LJ.C

..::ii:J.L.l\...Uh

UL1U.

.LCCll\..C.L

.&-\....ll...lJC

------~

\....VllllllQLJU •

The following are ERLCNF Warning messages:
ERLCNF-W-USEEUNKWN, Module <modulenarne> not found;

using EUNKWN

Explanation:

RPT was not able to find the module specified in
the Error Logging Universal Library and went to the EUNKWN module
instead.
This causes a formatted dump of the device register to
appear
in the report.
This message usually occurs if you tune
your ULB and eliminate the module for a device you want to use~

User Action:

Retune the ULB to include the missing module.

The following are ERLCNF Informational messages.
They accompany other
ERLCNF messages to give you additional
information. They do not
affect RPT operation.
ERLCNF-I-FILERRCOD, File error code

= <errorcode>

Explanation:
file.

This message displays the FCS error code
It accompanies messages on file access failures.

User Action:

None

necessary.

message.

3-34

This

for

informational

REPORT GENERATOR TASK (RPT)
ERLCNF-I-PARSECLST, PARSE.SECTION_LIST = <buf>
Explanation: This is an internal error within RPT.
accompanies the INTERROOl message described above.

This message

User Action:

You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
(See BADSUBPKT
message and any other information you have
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.
ERLCNF-I-PARSECTION, PARSE.SECTION

= <buf>

Explanation: This is an internal error
within
accompanies the INTERROOl message described above.

RPT.

User Action:

You should never see this message.
If you do, send
in an SPR along with a dump of the packet that generated the
message and any other information you have
(See BADSUBPKT
description).
If the message refers to a packet that you have altered, or a
module that you wrote, correct the module, recompile and add it
to the library.

3.5

ERLRPT REPORT MESSAGES

Most of the following error messages are either associated with errors
in the control file module that RPT is interpreting, or internal RPT
errors.
If the message refers to a control file module that you have altered,
or a module that you wrote and added to the error logging system,
correct the error, recompile the module, and add it to the library.
The module in which the error occurred is specified in the first (or
top)
line of the execution stack dump produced by RPT.
This
information appears on the report file and on the terminal from which
RPT is being run.
If the message refers to a DIGITAL-supplied module or is an internal
RPT error, please submit an SPR and include a listing of the error log
report file produced by RPT.
ERLRPT-F-ACCUDFVAR, Attempt to access undefined variable.
Explanation: A control file module
variable which had not been defined.

User Action:

attempted

Correct the user-written module or
for DIGITAL-supplied modules.

3-35

access

submit

a
SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-BADDIGIT, Invalid numeric digit in conversion.
Explanation: A numeric literal or the ASCII string argument for
the %COD$0CTAL, %COD$DECIMAL, %COD$HEX, %COD$BCD, %COD$BINARY, or
%COD$MACHINE function contained an illegal character for the
specified radix or was null or blank.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-BITFLDSIZ, Bit or field too large in extraction operation.
operation
Explanation: The bit or field in an extraction
exceeded the size of the value on which the extraction was
performed.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-BITTOOHIG, Bit number too large for specified storage unit.
Explanation: The bit number specified by the character string
portion of a #BI, #WI, #LI, #QI, or #VI numeric literal was too
large for the specified value size.
User Action: Correcc cne user-wr1ccen module or
for DIGITAL-supplied modules.

~ubmlt

SPR

ERLRPT-F-CASENOMAT, CASE selection expression has no matching value.
Explanation: No match was found for the value of the selector
expression in a CASE statement, and no ELSE clause was specified
in the CASE statement.
User Action:

Correct the user-written module or

C--

Y""\.T,...Tm"A.T

.LVL

U.Ll..l.L.l.M.1..o-i:>U~~.L.LCU

.- •• _ _ ,.;_~

..,,.,_~

••

1""",...

submit

SPR

1UVUU.LC.::>e

ERLRPT-F-CONTROLFI, Could not open control file.
Explanation:

The control file module could not be opened.

User Action: If using the default control file library, check to
see that it is in either LX: [1,6] or LB: [l,6] and is not locked,
and that you have read acces to it. If using a user specified
control file, check to see that it is not locked and that you
have read access.
ERLRPT-F-COROUMIS, COROUTINE
stack frame.

statement

executed

with

COROUTINE

Explanation: A COROUTINE
statement
was
executed
without
specifying a coroutine in the corresponding CALL statement.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-36

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-CRASH, Control file requested abort.
Explanation:
module.

The CRASH statement was executed by a control

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

file
SPR

ERLRPT-F-DATNOTEXI, Data declaration is longer than data.
Explanation: The amount of data specified in a PACKET or
SUBPACKET declaration was larger than the amount of data in the
PACKET or SUBPACKET. This condition may be due to an error in
the control file module or an error in the error log packet being
analyzed.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-DECAGAIN,
Group
Redeclaration illegal.

declaration

submit

already

SPR

declared.

Explanation: A
DECLARE,
PACKET,
SUBPACKET,
TABLE,
or
DYNAMIC TABLE statement was executed with a group name that was
already-defined.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-DECGRPCTX, Group in DECODE statement has no context.
Explanation: The group in the DECODE statement was a TABLE,
DYNAMIC TABLE, or PACKET or SUBPACKET with the REPEATED attribute
for which the current record context was not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-DECNOBIT, No BIT declaration
item.

corresponding

submit

SPR

DECODE

list

Explanation: The bit number specified for a data item in the
DECODE statement had no corresponding BIT declaration for the
data item in the specified group.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-DECNOTEXT, No bit to text transiation for DECODE iist item.
Explanation: The BIT declaration corresponding to the bit number
specified for a data item in the DECODE statement, had no print
expression.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-37

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-DEFCASELS, No match
conditional definition.

for

expression

control

CASE

Explanation: No match was found for the value of the selector
expression in a CASE conditional definition and no ELSE clause
was specified.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLRPT-F-DEFNOCONT, Attempt to access data in variable in
null context.

SPR

group

with

Explanation: The control file module attempted to access a
variable in a TABLE, DYNAMIC TABLE, or PACKET or SUBPACKET with
the REPEATED attribute for which the current record context was
not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

declarations

has

ERLRPT-F-DEFNOSTAK, Declaration stack overflow.
Explanation:
over flowed •

The stack

used

for

processing

User Action: Edit RPTBLD.CMD to increase the psect extension for
psect DCSTKO, and rebuild RPT.
ERLRPT-F-DEFSTKUND, Internal error - Declaration stack underflow •
--

m\....:-

CtA,t-'..LCUICl'-.&.VUi

.LU.&.;:,

.1.;:,

User Action:

Please submit an SPR with any information you have.

.;_,_ ____ ,

-----

.&.U'-CJ.1.ICl..L

CJ. I.VJ.

•• :1..i...:-

.., •• _ , __ ~,.:-- ...

W.&. 1..U.LIJ

nnm

n.r .L.

ERLRPT-F-DIVZERO, Attempt to divide by zero.
Explanation:

A control file module attempted to divide by zero.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-EXEINVCOD, Internal error - Execution stack entry has invalid
code.
Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-EXEINVPOS, Internal error - INPUT file has
value.

invalid

position

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

3-38

REPORT GENERATOR TASK (RPT)
ERLRPT-F-EXPGRPNOC, Attempt to reference
context.

POINTER

for

group

without

Explanation: A control file module attempted to reference the
POINTER special variable for a TABLE, DYNAMIC TABLE, or PACKET or
SUBPACKET with the REPEATED attribute for -which the current
record context was not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-EXPINVCOD, Internal error - Invalid expression item
expression.

code

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-EXPINVTYP, Internal error
expression.

invalid

symbol

data

type

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-EXPNORSYM,
expression.

Symbol

without

read

access

referenced

Explanation: A control file module attempted to read a variable
defined in a DECLARE statement, which had not been initialized.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-EXPUDFGRP, Undefined group referenced in expression.
Explanation: A control file module
group which had not been defined.

attempted

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-EXPUDFSYM,
evaluation.

Undefined

Explanation: A control
undefined symbol.

symbol
file

referenced

module

attempted

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

reference

submit

a
SPR

expression

access

submit

an
SPR

ERLRPT-F-EXPVALOVR, Value stack overflow during expression evaluation.
Explanation: The stack
used
expressions has overflowed.

for

processing

values

and

User Action: Edit RPTBLD.CMD to increase the psect extension for
psect VLSTKO, and rebuild RPT.

3-39

REPORT GENERATOR TASK (RPT)
ERLRPT-F-EXPVARNOC, Attempt to
expression.

access

variable

without

context

Explanation: A control file module attempted to reference a
variable in a TABLE, DYNAMIC TABLE, or PACKET or SUBPACKET with
the REPEATED attribute for which the current record context was
not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FILERCLOS, File close error.
Explanation:
file.

An error occurred when RPT

attempted

User Action:
Check for file access conflicts, device errors,
low pool condition.

ERLRPT-F-FILERREAD, File read error.
Explanation:
file.

An error occurred when

RPT

attempted

read

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLRPT-F-FILERSPAN, Records in file are not allowed to span blocks.
Explanation:
The span block attribute of the error log file
being analy7.eo was set,
F.LT creates the error loa file with this
attribute set, and neither ELI, ERRLOG, nor RPT will modify it,
but some other task may have.
User Action:
Use ELI to
(re)start error logging with a new
version of the error log file,
then use PIP to append the
PIP may produce the
previous version to the new version.
following warning message:
PIP -- Input files have conflicting attributes
This message can be ignored.
ERLRPT-F-FILERWRIT, File write error.
Explanation:
file.

An error occurred when RPT attempted to write to

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLRPT-F-FILINTOPN, Internal error -

file already open.

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

3-40

REPORT GENERATOR TASK (RPT)
ERLRPT-F-FILINVCOD, Internal error - invalid file code
operation.

for

specified

Explanation:

This is an internal error within RPT.

user Action:

Please submit an SPR with any information you have.

ERLRPT-F-FILINVMOD, Control file library has invalid module name table
format.
Explanation: The control file library has an invalid module name
table format. The control file must be a universal library.
User Action: Make sure that the
universal library and rerun RPT.

control

file

valid

ERLRPT-F-FILNOTCTX, Operation requires that dynamic file have context.
Explanation: A control file module executed a POINTER DELETE or
POINTER MOVE statement on a DYNAMIC TABLE for which the current
record context was not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-FILNOTEXI, Internal error - declared dynamic
exist.

submit

SPR

file

does

not

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-FILNOTVIR, Could not create virtual address space for
table.

module

Explanation: RPT could not dynamically extend its address
to create room for the module table.

space

User Action: If the maximum task size for the partition is less
than 32K, use the MCR command SET /MAXEXT or DCL command SET
SYSTEM/EXTENSION LIMIT to increase the maximum task size, or run
RPT in a different partition.
ERLRPT-F-FILTOOBIG, File too large to read.
Explanation: RPT cannot analyze error log files which are larger
than 65535 blocks.
User Action:

Use ELI to create new error log files more often.

ERLRPT-F-FINDFIELD, FIELD in FIND statement does not have
type.

valid

data

Explanation: A control file module executed a
where the specified FIELD was not NUMERIC,
RSXTIME, VMSTIME, or LOGICAL.

FIND statement
STRING, ASCII,

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

3-41

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-FINDNOCON, FIND statement
context.

not

valid

group

with

Explanation: A control file module executed a FIND statement for
a TABLE or DYNAMIC TABLE attribute for which the current record
context was not valid.
User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

WRITE

ERLRPT-F-FORCLSNUL, FORMAT clause null.

Explanation: A control file module executed
WRITE GROUP statement with a null FORMAT clause.

User Action:

submit

Correct the user-written module or
for DIGITAL-supplied modules.

ERLRPT-F-FORFIELDS, FORMAT error - Field too narrow
print.

for

SPR

variable

Explanation:

A control
file module executed a
WRITE GROUP
statement where the width specified by a !DP directive was too
short for the corresponding variable.

User ~cticn:

Correct the user-written module or

o:>UUUl.L I..

.._

SPR

field

!OF

... L . - !

for DIGITAL-supplied modules.
ERLRPT-F-FORFIELDW, FORMAT error - Name too
directive.

long

for

Explanation~

A control
file module executed a
WRITE GROUP
statement where the width specified in a !DF directive was less
than the length of the name of the corresponding variable

User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FORINVCHA, FORMAT error -

Invalid character in FORMAT clause.

Explanation: A control
WRITE GROUP
statement
non-printing character.

or
module executed
a
WRITE
a FORMAT clause containing a

file
with

User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

ERLRPT-F-FORINVDIR, FORMAT error -

submit

SPR

Invalid format directive code.

Explanation: A control file module executed
a
WRITE
or
WRITE GROUP statement with a FORMAT clause containing an invalid
format directive.
User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

3-42

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-FORINVVTY, FORMAT error - Attempt to output invalid
type.

variable

Explanation: A control file module executed
a
WRITE
or
WRITE GROUP statement with a FORMAT clause containing a !DP
directive for which the corresponding variable was the wrong
type.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FORLINEOV, FORMAT error - Line overflow in FORMAT clause.
Explanation: A control module executed a WRITE or WRITE GROUP
statement
during which the output buffer overflowed -while
processing the FORMAT clause.
The output buffer
is
132
characters wide.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-FORNOARG, FORMAT error - Format
argument.

directive

submit

missing

SPR

required

Explanation: A control file module executed
a
WRITE
or
WRITE GROUP statement with a FORMAT clause containing an !FC or
!FS directive with no numeric argument.
User Action:
Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FORNONAME, FORMAT error - request to print a field name for a
value.
Explanation: A control file module executed
a
WRITE
or
WRITE GROUP statement with a FORMAT clause containing a !DF
directive matched with a value rather than a variable.
user Action:
Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FORNOREAD, FORMAT error - Attempt to print a variable without
read access.
Explanation: A control file module executed
a
WRITE
or
WRITE GROUP statement which attempted to print a variable without
read access.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

WRITE

ERLRPT-F-FORNOTASC, FORMAT clause not ASCII.
Explanation: A control file module executed
a
WRITE GROUP statement with a non-ASCII FORMAT clause.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-43

submit

SPR

REPORT GENERATOR TASK {RPT)
ERLRPT-F-FUNDATNOT, Specified (sub)packet
offset.

not

large

enough

for

Explanation: A control file module executed
a
look-ahead
function where the value of the offset argument was larger than
the specified PACKET or SUBPACKET.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-FUNFIELDS,
Invalid
conversion function.

conversion

code

submit

argument

SPR

time

Explanation: A control file module executed a time conversion
function with an illegal value for the conversion code argument.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FUNINVPOI, Invalid string pointer value in string function.
Explanation: A control file module executed a
%STR$PARSE or
%STR$QUOTE function where the value of the pointer argument was
larger than the length of the string argument.
User ~ction; Correct th8 u:::H:::=t.-wr..u:~en module or
for DIGITAL-supplied modules.
ERLRPT-F-FUNNOTCHA, Argument to STR$CHAR is not
character.
Expianation: The value ot the argument
function must be in the range 0 to 127(10).
User Action:
-Fn'I'"
_....,_

... .&.a..L.I

for

Correct the user-written module or

nT~T'l''aT -C!nnnl; OM
LI"'-""~

.._.,Wl;'I:"~.&.'""''\.A

submit

valid

SPR

range

for

the
submit

%STR$CHAR.
an

SPR

n'\nMnl
OC'
Ll.&'-''-AU.•'-"'"' e

ERLRPT-F-FUNNOTIMP, Function not implemented.
Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-FUNQUOODD, Quote string in STR$QUOTE function must have
length.

even

Explanation: A control file module executed
a
%STR$QUOTE
function, where the quote string argument was not an even length.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-FUNSTRSIZ, Output string from string function too large.
Explanation: A control file module executed a string
which resulted in a string longer than 255 characters.

function

Oser Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-44

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-FUNWRONGA, Incorrect number of arguments in function call.
Explanation: A control file module executed a function call with
the wrong number of arguments.
User Action: Correct the user-written moduie or
for DIGITAL-supplied modules.

submit

SPR

reference

ERLRPT-F-GROUPDEF, Attempt to reference undefined group.
Explanation:
A control file module
undefined group.

attempted

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-GROUPNOC, POINTER
context.

statement

executed

submit

group

without

SPR

Explanation: A control file module executed a POINTER statement
on TABLE or DYNAMIC TABLE for which the current record context

was not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-HEAPOVERF, Heap too small to hold value.

submit

SPR

Overflow.

Explanation:
The heap used for processing values and expressions
has overflowed.
User Action:
Edit RPTBLD.CMD to increase the psect extension for
psect VHEAPO, and rebuild RPT.
ERLRPT-F-INCFORWRI,
statement.

Too

few

FORMAT

expressions

WRITE GROUP

Explanation: A control file module executed a
WRITE GROUP
statement which did not have two FORMAT expressions in the FORMAT
clause.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLRPT-F-INCRDECRL, Numeric variable in INCREMENT or DECREMENT
than value.

SPR

larger

Explanation: A control file module executed an INCREMENT or
DECREMENT statement on a variable which was larger than a word.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-45

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-INCRDECRN, Variable in INCREMENT or DECREMENT
numeric.

statement

not

Explanation: A control file module executed
DECREMENT statement on a non-numeric variable.

INCREMENT

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

not

valid

ERLRPT-F-INCRDECRV, Variable in INCREMENT or DECREMENT
read-only.

Explanation: A control file module executed an INCREMENT or
DECREMENT statement on a variable which was not both readable and
writeable.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-INTINVDEC, Internal error - Invalid declaration entry type in
WRITEGROUP.
Explanation:

This is an internal error within RPT.

Oser Action:

Please submit an SPR with any information you have.

ERLRPT-F-INTVALSTK, Internal error value stack.

statement

left

information

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-INVRADCNV,
conversion.

Internal

error

Invalid

radix

code

for

Explanation:

This is an internal error within RPT.

Oser Action:

Please submit an SPR with any information you have.

ERLRPT-F-LEAVENOC, LEAVE statement executed outside of
block.
Explanation: A control file module executed a
which was not inside a loop statement block.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-LISTNOEXP, No expression in
variable.

LIST

for

conditional

LEAVE

statement,

submit

corresponding

SPR

Explanation: A control file module executed a SEARCH statement
in which a match was found, but there were not enough expressions
in the list element for the number of variables specified in the
GET clause of the SEARCH statement.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-46

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-LISTNOMAT, Too
referenced LIST.

many

expressions

statement

for

Explanation: A control file module executed a SEARCH statement
in which there were too many search expressions for the specified
LIST.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-LISTNOTDF,
defined.

Group

referenced

submit

SPR

statement

not

Explanation: A control file module executed a SEARCH statement
in which the name specified for the LIST was not defined.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-LISTNOTLS, Group referenced in
LIST.

submit

statement

SPR

not

Explanation: A control file module executed a SEARCH statement
in which the name specified for the LIST was not defined as a
list.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-MATDIFTYP, Values of differing type cannot be matched.
Explanation: A control file module executed a
which tried to match values of differing types.

MATCH

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

statement
an

SPR

ERLRPT-F-MATVALSIZ, Values of different size cannot be matched.
Explanation: A control file module executed a
which tried to match values of differing size.

MATCH

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLRPT-F-MEMALLFAI, Memory allocation failure memory.

statement

insufficient

Explanation: RPT could not dynamically extend its address
to create room for DYNAMIC TABLES or control file modules.

SPR

virtual
space

User Action: If the maximum task size for the partition is less
than 32K, use the MCR command SET /MAXEXT or DCL command SET
SYSTEM/EXTENSION LIMIT to increase the maximum task size, or run
RPT in a different partition. If this occurs while generating
summaries f·or large numbers of packets, try reducing the amount
of data needed by using RPT switches to reduce the number of
packets analyzed for each summary.

3-47

REPORT GENERATOR TASK (RPT)
ERLRPT-F-MEMINIFAI, Memory allocation initialization failure.
Explanation:
RPT could not dynamically extend its address
to create room for its data structures.

space

User Action:
If the maximum task size for the partition is less
than 32K,
use the MCR command SET /MAXEXT or DCL command SET
SYSTEM/EXTENSION LIMIT to increase the maximum task size, or
run
RPT in a different partition.
ERLRPT-F-MODLOAGRP, Undefined group referenced by module to be loaded.
Explanation:
The control file module being loaded, attempted
reference an undefined group.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

Explanation:
The control file module being loaded, attempted
reference an undefined symbol.

ERLRPT-F-MODLOASYM, Undefined symbol in module to be loaded.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-MODNAMENL, Module name cannot be null.
Explanation: A control file module attempted to access
control file module which had a null or blank name.

another

User Action:
Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-MODNOMEMf Insufficient free memory to load module=
Explanation:
RPT could not dynamically extend its address
to create room for control file modules.

space

User Action:
If the maximum task size for the partition is less
than 32K,
use the MCR command SET /MAXEXT or DCL command SET
SYSTEM/EXTENSION LIMIT to increase the maximum task size, or run
RPT in a different partition.
If this occurs while generating
summaries for large numbers of packets, try reducing the amount
of data needed by using other switches to reduce the number of
packets analyzed for each summary.
ERLRPT-F-MODSTART, Starting module for execution not found.
Explanation:
The control file
named DISPATCH.

library

must

contain

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-48

submit

module
an

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-MODZERO, Attempt to modulus by zero.
Explanation:
zero.

A control file module attempted to perform a MOD by

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-NOMORESTK, Execution stack overflow.
Explanation:

RPT's execution stack has overflowed.

User Action: Edit RPTBLD=CMD to increase the psect extension for
psect XCSTKO, and rebuild RPT.
ERLRPT-F-NOSTACKE, Internal error empty stack.

Pop

from

execution

stack

with

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-NOTDYNFIL, Dynamic file operation performed on invalid group.
Explanation: A control file module specified a group which was
not
defined as DYNAMIC TABLE in a statement or operation
requiring ·a DYNAMIC_TABLE.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-NOTPOINT, POINTER LOAD or MOVE executed
variable.

submit

with

Explanation: A control file module executed a
MOVE with a variable which was not a pointer.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

SPR

non-pointer

POINTER

LOAD

submit

SPR

ERLRPT-F-NOTPOIVAR, POINTER LOAD with no pointer variable specified.
Explanation: A control file module executed a
MOVE with no variable specified.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

POINTER

LOAD

submit

SPR

ERLRPT-F-NUMINVOPR, Invalid numeric double-operand operation code.
Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

3-49

REPORT GENERATOR TASK (RPT)
ERLRPT-F-OPRINVLOG, Attempt to perform logical operation on an invalid
type.
Explanation: A control
logical operation with
LOGICAL.

file module attempted to perform a
operands that were neither NUMERIC nor

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-OPRNOTIMP, Operation not implemented.
Explanation: A control file module attempted to perform a
multiplication where both operands were larger than a word value.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-PACKETSIZ, Il+egal packet size.
Explanation: The size of an error log packet was zero
cause the packet to cross a block boundary.
User Action:

would

Please submit an SPR with any information you have.

ERLRPT-F-POISETGRP, POINTER variable is
POINTER ••• LOAD or MOVE.

not

from

correct

group

Explanation: A control file module executed a POINTER LOAD or
MOVE statement in which the optional pointer variable was not a
pointer to the specified DYNAMIC TABLE.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLRPT-F-POISETMOD, POINTER variable is from wrong module
LOAD or MOVE.

SPR

POINTER

Explanation: A control file module executed a POINTER LOAD or
MOVE statement in which the DYNAMIC TABLE pointed to by the
optional pointer variable was not in the same module as the
DYNAMIC_TABLE specified in the POINTER statement.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLRPT-F-POISETSIZ, GROUP too small for POINTER in POINTER
or MOVE.

SPR

LOAD

Explanation: A control file module executed a POINTER LOAD or
MOVE statement in which the optional pointer variable was
pointing past the end of the specified DYNAMIC TABLE.
User Action: Correct the user-written module or
for DIGITAL-supplied modules,

3-50

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-PROCNAMEN, Null procedure name.
Explanation: A control file module specified
procedure name in a CALL or ENABLE statement.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

null

submit

blank
an

SPR

ERLRPT-F-RAD50BYTE, Cannot convert a byte using RAD50 conversion.
Explanation: A control file module attempted to convert an ASCII
string or numeric literal to a BYTE using RAD50 conversion.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-RELINVCOD, Invalid relational operator.
Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-RETURNNOC, A RETURN was executed with no corresponding CALL.
Explanation: A control file module executed a
outside of a procedure or coroutine.

RETURN

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-SELECTNOM, SELECT statement index has no
block.

submit

matching

statement
an

SPR

statement

Explanation: A control file module executed a SELECT statement
with no statement block to match the value of the numeric control
expression.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-SIGNOTASC, Parameter or message in SIGNAL-class statement not
ASCII.
Explanation: A control
file
module
executed
a
SIGNAL,
SIGNAL_STOP, or MESSAGE statement with a non-ASCII argument.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-SIGTOOBIG, Message and parameters in
too long.
Explanation:
SIGNAL STOP,
concatenated
characters.

submit

SIGNAL-class

SPR

statement

A control
file
module
executed
a
SIGNAL,
or MESSAGE statement in which the length of the
message and parameters was
longer
than
255

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-51

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-SIGTOOMAN, Cannot issue a SIGNAL during SIGNAL processing.
Explanation: A control file module executed a
SIGNAL
or
previous SIGNAL or
SIGNAL STOP statement while processing a
SIGNAL-STOP.
User Action:
Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-STANOTIMP, Statement not implemented.
Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-STANOTVAL, Internal error -

invalid statement code.

Explanation:

This is an internal error within RPT.

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-SUBEXTBIG, Substring extraction end element exceeds string.

Explanation:
___ .__ ..... _:

i:>UUi:>l..1. .LU'::j

file module
-··._ A___ control
.... .; -•u which the

attempted

'l:hl..1.Qvl...LV11

perform

exceeded the end cf

the string.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

Explanation:
packet.

The current subpacket, exceeded the bounds

the

User Action:

Please submit an SPR with any information you have.

ERLRPT-F-SUBPKTSIZ, Illegal subpacket size.

ERLRPT-F-UNDEFPROC, Specified procedure not found.
Explanation: A control file module has
executed
statement, and the specified procedure was not found.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

CALL

submit

SPR

ERLRPT-F-UNDMODULE, Specified module not found.
Explanation: A control
file module has
executed
statement, and the specified module was not found.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-52

submit

CALL

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-VALSTKOVR, Value stack overflow.
Explanation: The stack
used
expressions has overflowed.

for

processing

values

and

User Action: Edit RPTBLD.CMD to increase the psect extension for
psect VLSTKO, and rebuild RPT.
ERLRPT-F-VALUESIZE, Value in expression is too large.
Explanation: A control file module evaluated an expression
which an intermediate value or the final value was too large.

user Action: Correct the user-written module or
for DIGITAL-supplied modules.

SPR

submit

ERLRPT-F-VALUETYPE, Value in expression is wrong type.
Explanation: A control file module evaluated an expression in
which an intermediate value or the final value was the wrong
type.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLRPT-F-VARNOCONT,
context.

Attempt

access

variable

submit

SPR

group

without

Explanation: A control file module attempted to reference a
variable for a TABLE, DYNAMIC TABLE, or PACKET or SUBPACKET with
the REPEATED attribute for which the current record context was
not valid.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-VARNODATA, Attempt to access variable in group with no data.
Explanation: A control file module attempted to reference a
variable for a TABLE, DYNAMIC TABLE, or PACKET or SUBPACKET with
no data.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLRPT-F-VARNOTDAT, Attempt to load data into a BIT or FIELD variable.
Explanation: A control file module attempted to load a value
into a BIT or FIELD in a group, rather than into the data item
for which the BIT or FIELD was defined.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

3-53

submit

SPR

REPORT GENERATOR TASK (RPT)
ERLRPT-F-WRITEACCV, Attempt
variable.

load

value

into

non-writable

Explanation: A control file module attempted to
into a data item in a PACKET, SUBPACKET or TABLE.

load

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

3-54

value
an

SPR

CHAPTER 4

ERROR LOG CONTROL FILE ARCHITECTURE

This chapter describes the architecture of Error Log Control Files. A
knowledgeable system programmer can use the information presented here
to add user-written modules to the Error Logging System.
This chapter includes the following major sections:

4.1

•

Terms and Concepts
Defines the most
concepts presented in this chapter.

important

•

Control File Module Architecture
Describes RSX-llM and
RSX-llM-PLUS control file modules, the flow of program control
through the modules, module compilation paths, and recompiling
modules after modifications.

•

Internal Interfaces -- Describes interaction
file modules, with examples.

between

•

Module Dispatching
Explains
CPU-level dispatching.

and

•

Support of Non-DIGITAL Devices -- Provides the information you
need
to
include driver and error logging support for
non-DIGITAL devices.

•

Error Logging System Source Code Examples
Includes
annotated listings of source code for four modules:
ERM23,
DSP2Ml, DSP2Pl, and NRM23.
The source code is keyed to
discussions in the Internal
Interfaces section
(4.3) and
Support of Non-DIGITAL Devices section (4.5) of this chapter.

event-level

terms

and

control

device-

TERMS AND CONCEPTS

Here are definitions of
presented in this chapter:

the

most

important

terms

and

concepts

•

Control File - A collection of modules that together perform a
function, such as processing error log files.

•

Module - A component of the Error Logging System.
There are
three kinds of modules:
source modules, which have the file
type .CNF, object modules, which have the file type .ICF, and
listing modules, which have the file type .LST. Module names
that end in Ml are generally common to both RSX-llM and
RSX-llM-PLUS systems
(for example, DEVSMl), except where the
module name has an alternate Pl ending
(for example, DSP2Ml
and DSP2Pl).
In this case, module names that end in Ml are
for RSX-llM systems only and those that end in Pl are for
RSX-llM-PLUS systems only.

4-1

ERROR LOG CONTROL FILE ARCHITECTURE
File Language - The language in which control files
• Control
are written.
The Control File Language (CFL) is described in
Chapter 5.
•

Error Log File - The file that contains the raw error logging
data.
One record in the file corresponds to one event.
The
default specification for this file is LB:[l,6]LOG.ERR.

•

Event - Something that is logged in the error log file.
An
event may be the recording of an actual device error or it
could be some informational data, such as a device mount or a
change in system time.

•

Packets/Subpackets - Each record (or event) is also a packet.
A packet begins with a length word and is followed by data,
which can consist of zero or more subpackets.
A subpacket
also consists of a length word followed by data.
Every packet
in the Error Logging System contains at least one subpacket.
NOTE
The packet length word begins the packet, but it is
not part of the packet;
the packet length word is
kept by FCS.
Therefore, the packet length word is not
included in the length of the packet.
However, the
subpacket length words are part of the packet and are
included in its length. Thi~
the Error Logging System.
See Figure 4-1 for the general structure of
packet.

4.2

error-logging

CONTROL FILE MODULE ARCHITECTURE

The Error Logging System is modular:
that is,
information and
dependencies specific to different devices are isolated in modules
written for each device.
This section describes the architecture of
the RSX-llM and RSX-llM-PLUS control file modules:
the modules
themselves, the flow of program control through the modules,
the
compilation paths, and modifying and recompiling the modules.

4-2

ERROR LOG CONTROL FILE ARCHITECTURE

PACKET LENGTH
SUBPACKET LENGTH

HEADER SUBPACKET - - - - - - - 1

SUBPACKET LENGTH

TASK SUBPACKET

SUBPACKET LENGTH

DEVICE !DENT!F!CAT!ON SUBPACKET

1-_ _s_u_s_P_A_c_K_E_T_L_E_N_G_T_H_---1

DEVICE OPERATION SUBPACKET

~
I

I
I

DEVICE ACTIVITY SUBPACKET _ _ _...,.·
(OPTIONAL ON RSX-11 M)

I
I
S_U_B_P_A_C_K_E_T_L_E_N_G_T_H_ _

! - i_ _

II
SUBPACKET LENGTH

DATA SUBPACKET - - - - - - - -

ZK-i 111-82

Figure 4-1 - Structure of Error-Logging Packet

4.2.1

RSX-llM and RSX-llM-PLOS Control File Modules

Here are short descriptions of the function of each of the RSX-llM and
RSX-llM-PLUS control file modules.
Remember that modules with names
that end with "Ml" are either common to both RSX-llM and RSX-llM-PLUS
or are for RSX-llM only, and modules with names ending with "Pl" are
for RSX-llM-PLUS only.
DISPATCH
DISPATCH is the root module for the Error Logging System.
It
declares all commonly used variables, calls the INITMl module to
initialize the system, and then calls the PARSEM module to obtain
and parse the command line.
DISPATCH then requests the records
4-3

ERROR LOG CONTROL FILE ARCHITECTURE
from the input log file, declares the common subpackets
(HEADER,
TASK, DEVICE ID, DEVICE OPERATION, and DEVICE ACTIVITY (optional
on RSX-llM, standard on RSX-llM-PLUS)) for each record, computes
the correct dispatcher module name, and calls that module. When
all the records are processed, it calls the summary modules if
requested
(RSX-llM-PLUS
only),
and
finally
calls
the
FINLMl/FINLPl module to clean up.
See Section 4.6.5 for the
definitions of the standard DIGITAL subpackets. Dispatching is
described in more detail in Section 4.4.
PARS EM
PARSEM declares variables local to the processing of the command
line and calls the PARSlM module to obtain the command line.
It
then calls the PARS2M module to process any switches and the
PARS3M module to open the various files.
PARSEM also provides
commonly used parsing routines to the other parsing modules.
PARSlM
PARSlM initializes parsing variables and gets the command line
from RPT.
It then breaks all of the file specifications out of
the command line, leaving all of the switches.
PARSlM then
searches for the /REPORT switch.
If it finds the switch, PARSlM
replaces it with the specified string of predefined switches.
PARS2M
PARS2M gets a switch from the string of switches produced by
PARSlM.
It then checks the switch for ambiguity and calls PRS2AM
to process the switch.
If PRS2AM does not recognize the switch,
it is passed to the PRS2BM module.
PARS2M repeats this process
until all switches have been processed.
PRS2AM

PRS2AM processes
PACKET.

the

following

switches:

DATE,

DEVICE,

and

PRS2BM
PRS2BM processes the following switches:
FORMAT, SERIAL, SUMMARY
(RSX-llM-PLUS only), TYPE, VOLUME, and WIDTH.
PARS 3M
PARS3M applies the default values to any switches that
specified and opens the specified files.

were

not

SELTMl
SELTMl is called by DISPATCH to determine if the current packet
meets the selection criteria of the command line switches.
DSPlMl/DSPlPl
The DSPlMl/DSPlPl modules process Error Log Control events
(See
Section 4.4.1).
These modules declare the DATA subpacket for
each type of event and process the event to completion, calling
the formatter modules to print the common data if the FULL report
format is specified.

4-4

ERROR LOG CONTROL FILE ARCHITECTURE
DSP2Ml/DSP2Pl
The DSP2Ml/DSP2Pl modules process Device Error events
(See
Section 4.4.1). These modules call DEVSMl to determine the name
of the device-level module required to process the event and then
calls that module as a co-routine and passes control to it. The
device-level module declares the DATA subpacket and then extracts
information from the registers of the logged device so it can
provide additional selection information. When the device-level
module returns control to DSP2Ml/DSP2Pl, it performs the last of
the selection tests and makes the decision whether to continue
with this event or not.
If DSP2Ml/DSP2Pl decides to continue,
and if the FULL report format has been specified, DSP2Ml/DSP2Pl
calls the formatter modules to print the common information.
Once printing is completed, control returns to the device-level
module, which prints the device registers.
If the BRIEF report has been specified, DSP2Ml/DSP2Pl still must
decide whether to continue, but there is no need for the
formatter modules and DSP2Ml/DSP2Pl does its own printing.

:arr ,Jiax...:,11t1~,i>~us /on1,~, . . :·<>n.<;e:·:, ~{1;t. •¥~e.· · . Pt·.i:nt)/n:~· ··: i .s: '. ·: ·c.Orlip1,e~.ed:/( ·:;t.f{~.
1
·:::.:i!'?~J,i~:.-~N~i*t~::~i~~~:.!'t~!~a~~~-r~::~~~~ %$:
·peiqy~~;te~.' ~:~lntri:;::t't¥ .. iq.. the: ';fi. U~~ri .f~·~es! ::·.·

DSP3Ml/DSP3Pl
The DSP3Ml/DSP3Pl modules process Device Information events
(See
Section
4.4.1).
They
perform the same function as the
DSP2Ml/DSP2Pl modules, but for device errors not related to I/O.
These modules are required only if you have a TU78 or MSCP (Mass
Storage Control Protocol) device.
DSP4Ml/DSP4Pl
The DSP4Ml/DSP4Pl modules process Device Control
Information
events
(See Section 4.4.1).
DSP4Ml/DSP4Pl calls DEVSMl to get
the type of device associated with the device mnemonic.
Mount, dismount and reset operations have no DATA subpacket. The
formatter modules print the information if the FULL report mode
is specified;
otherwise, the module does all the printing
itself.
Like the DSP2Pl and DSP3Pl modules, DSP4Pl records
summary information if requested.
The Block Replacement event does have a DATA subpacket which is
processed entirely by this module.
This type of event does not
contribute to summaries.
DSP5Ml/DSP5Pl
The DSP5Ml/DSP5Pl modules process events detected by the CPU (See
Section 4.4.3). DSP5Ml/DSP5Pl gets the CPU type from the HEADER
subpacket declared by DISPATCH and calls
the
appropriate
CPU-level module as a co-routine if the event was a memory parity
error. The processing then proceeds much like that for device
errors.
If the event was an unknown interrupt, the
processes the DATA subpacket itself.

4-5

module

declares

and

ERROR LOG CONTROL FILE ARCHITECTURE
DSP6Ml/DSP6Pl
The DSP6Ml/DSP6Pl modules process System Control
Information
events
(See Section 4.4).
There is no DATA subpacket associated
with the power recovery event.
The formatter modules print the
common information if in FULL-report mode;
otherwise, the module
does all the printing itself.
DSP7Ml/DSP7Pl
The DSP7Ml/DSP7Pl modules process Control Information events (See
Section 4.4).
These modules declare the DATA subpacket for each
type of event and process the event to completion, calling the
formatter modules to print the common data if the FULL report
format is specified.
F INLM l/F INL P 1
FINLMl or FINLPl is called by DISPATCH to clean up after all
error log events are processed.

the

dn; RSx'..o>1:l.M--:P~US' :<itlly, · FI,NLP{:~\so~ out~ut~' the..tinal 'page

o'f th~
Tbi Si . rpag~·: 1contai.ns 'SUqb: lin:for!matii:on as the
1
dommafki\ p. Unederitiered 'by'
\}ser,,; , the . £i les :used) t~. . swi t.ch
•~t;a:it'~s1,i ·;,~tihe;:! ·'.dumb~r of· ~verit'~; :proc~ssed, ·and· ·~qw· ·i:Qng 'iit took,···. tg

'errirk>tr:: ! :;1btj : :t:ieplo:r'.t:..

·~~~ei'l=.ait~. '. qbe .• :~epcirt:.
T""I••, ........ ,

·;th1e

· · · ·'

··

/'r"I••, r.'T1l•,

.c l"J.1.L'll'l.1./ .c r1.1.n1·1.1.

FMlNMl/FMlWMl are formatter modules.
They print
information at
the top of each page of a FULL report.
The information comes
mostly from the HEADER subpacket.
FMlNMl prints reports
in
NARROW format and FMlWMl prints reports in WIDE format.
f'M2CM1

FM2CM1 is one of the formatter modules.
It prints the Requesting
Task section of a FULL report.
The information comes from the
TASK subpacket.
FM2CM1 prints reports in both NARROW and WIDE
formats.
FM3CM1
FM3CM1
is a
formatter
module.
It
prints
the
Device
The
Identification
Information section of a
FULL report.
information comes from the DEVICE ID subpacket.
FM3CM1 prints
reports in both NARROW and WIDE formats ..
FM4NM1/FM4WM1
FM4NM1/FM4WM1 are formatter modules.
They print
the
I/O
Operation
Identification
section
of a FULL report.
The
information comes from the DEVICE OPERATION subpacket.
FM4NM1
prints reports
in NARROW format, and FM4WM1 prints reports in
WIDE format.
Optionally, FM4NM1/FM4WM1 also prints the Concurrent I/O Activity
section of a FULL report.
The information comes from the DEVICE
ACTIVITY subpacket.
See Section 4.1 for more information.

4-6

ERROR LOG CONTROL FILE ARCHITECTURE
FMTNPl/FMTWPl
FMTNPl/FMTWPl are formatter modules.
They print the first page
of a FULL report,
that is,
all of the information from the
HEADER, TASK, DEVICE ID, DEVICE OPERATION, and DEVICE ACTIVITY
subpackets.
FMTNPl prints reports in NARROW format, and FMTWPl
prints reports in WIDE format.
INITMl
INITMl initializes variables to be used later in the control
file.
It sets up the page-top banners,
formatter module
selectors, and WRITE GROUP format statements, based on whether
the report is NARROW-or WIDE.
DEVSMl
DEVSMl
is
called
by
DSP2Ml/DSP2Pl,
DSP3Ml/DSP3Pl,
and
DSP4Ml/DSP4Pl to provide certain device-related information.
DSP2Ml/DSP2Pl and DSP3Ml/DSP3Pl call
it to find,
among other
things,
the name of the device-level module that should help
process the event. DSP4Ml/DSP4Pl calls DEVSMl to find out the
name of the device associated with a device mnemonic.
If the device mnemonic is DU, DEVSMl then calls DEVUDA to do most
of the processing.
DEVUDA
DEVUDA is called only by DEVSMl.
It
processing of events on MSCP devices.

assists

DEVSMl

the

ERRORM
ERRORM is the error
Whenever a SIGNAL
error.

processor for
the Error Logging System.
or SIGNAL STOP occurs, ERRORM processes the

SMRYEP
SMRYEP prints Error summaries on RSX-llM-PLUS only.
DISPATCH
calls SMRYEP after all packets have been processed if an Error
summary was requested.
SMRYGP
SMRYGP prints Geometry summaries on RSX-llM-PLUS only.
DISPATCH
calls SMRYGP after all packets have been processed if a Geometry
summary was requested.
SMRYHP
SMRYHP prints History summaries on RSX-llM-PLUS only.
DISPATCH
calls SMRYHP after all packets have been processed if a History
summary was requested.
CPU-level modules
There are five CPU-level modules, all with names derived from
their associated processors.
They are called as co-routines by
DSP5Ml/DSP5Pl to process memory parity errors~
•

Ell'34

(RSX-llM only) - 'Process,es errors from th'e PDP-11/34.

4-7

ERROR LOG CONTROL FILE ARCHITECTURE
•

Ell44 - Processes errors from the PDP-11/44.

E1160 (RSX-llM only} - Processes errors, from the PDP,;..11/60,.

•

Ell7X - Processes errors from the PDP-11/70 and PDP-11/74.

•

EllXX - Processes errors from all other PDP-11 processors.

EUNKWN
EUNKWN is a universal device-level or CPU-level module.
EUNKWN
is called if a particular device-level module is unavailable, or
if the device mnemonic is unknown to the Error Logging System.
EUNKWN is also called if the CPU type is unknown.
EUNKWN produces a formatted dump of the data, showing the
relative offset within the data, and the data itself, in word,
high-byte, and low-byte formats (all octal), and in binary word
format.
See Section 4.5 for more information on error logging
from unknown devices.
DMPALL
DMPALL is similar to EUNKWN.
DMPALL is called if the packet
cannot be processed due to an error in format or structure.
DISPATCH calls DMPALL if the packet fails any sanity check.
DMPALL prodQces a
formatt~a
dump of
the data: showing the
relative offset within the data, and the data itself, in word,
high-byte, and low-byte formats (all octal), and in binary word
format.

Device-Level Modules

Device-level modules contain
details
of
the
bit-to-text
translation
for
all
supported
error-logging
devices.
DSP2Ml/DSP2Pl and DSP3Ml/DSP3Pl call them as co-routines.
Their
names are derived from the names of the associated devices.
See Section 4.5 for
constructed, and how
unsupported devices.

information on
you can write

how these modules are
device-level modules for

Here are the standard error-logging device-level modules:

• EMLll - Processes MLll errors
• ERK05 - Processes RK05 errors

• ERK67 - Processes RK06 and RK07 errors
• ERL12 - Processes RLOl and RL02 errors
• ERM05 - Processes RM05 errors
• ERM23 - Processes RM02 and RM03 errors
• ERM80 - Processes RM80 errors

• ERP07 - Processes RP07 errors
• ERP23 - Processes RP02 and RP03 errors
• ERP456 - Processes RP04, RP05 and RP06 errors
4-8

ERROR LOG CONTROL FILE ARCHITECTURE

• ERSll - Processes RSll errors
• ERS34 - Processes RS03 and RS04 errors
• ERXOl - Processes RXOl errors

• ERX02 - Processes RX02 errors
• ET0310 - Processes TS03, TElO, and TUlO errors
• ET1645 - Processes TE16, TU16, and TU45 errors
• ETAll - Processes TAll errors
• ETCll - Processes TCll errors
Processes TSll, TU80 errors

ETSll

•

ETSVOS - Processes TSVOS errors

•

ETU58 - Processes TU58 errors

•

ETU77 - Processes TU77 errors

•

MSCPAT - Processes MSCP Attention errors

•

MSCPE - Processes MSCP controller errors

•

MSCP60 - Processes MSCP RA60 errors

•

MSCP80 - Processes MSCP RA80/RA81 errors

•

MSCPEN - Processes MSCP End Packet errors

MSCPSD - Processes MSCP Small Disk (RC25/RD51/RD50) errors

•

MSCPTO - Processes MSCP Timeout errors

Notes Modules
Notes modules contain notes for error conditions that need
additional explanation.
Notes modules are device-specific and
have names derived from the names of the associated device-level
module.
See Section 4.5.3.2 for more information on how these
modules are constructed.
Here are the standard error-logging notes modules:

• NMLll - Processes MLll notes
• NRK67 - Processes RK06 and RK07 notes
• NRM05 - Processes RMOS notes
• NRM23 - Processes RM02 and RM03 notes
• NT0310 - Processes TS03, TElO, TUlO notes
• NTSll - Processes TSll, TU80 notes

4-9

ERROR LOG CONTROL FILE ARCHITECTURE
Program Control Flow

4.2.2

Here is a description of the general flow of program
the control file modules:

control

through

RPT opens the control file.
In most cases,
this is the
default control file LX:
or LB: [l,6]ERRLOG.ULB.
If you wish
to use some other filespec, RPT must be rebuilt to prompt for
the new name of the file.

RPT creates the module table in its dynamic work space.
This
table contains an entry for each module in the control file
universal library.

RPT loads the DISPATCH module and transfers
ENTRY procedure.

The ENTRY procedure is very similar to the root module in
most MACR0-11 programs.
This procedure declares most of the
commonly used data structures.
It then enables the ERROR 1
procedure in the ERRORM module as an error handler.
Next, Tt
gets and parses the command line by calling the SETUP
procedure in the PARSEM module.
That done, it performs some
general initialization with the INIT 1 procedure from INITMl.
ENTRY
then
sets
up a
loop which steps through the
PACKET RANGE file, extracting pairs of packet ranges which
are fed back to RPT.
The next step, performed for each
fJd(.;kt!i:. r'.C:t(1l::J~, is Lo loop through the
current packet .Lau'='c,
requesting each packet in turn and calling the DISPATCH
procedure i.n the DI~PA~CH . mod.ule. . \ften! :
1?ac~tjs
a~
Praicket. ra,ng~s .hp~e .b7e;n, ;r~~ur:~t~l1, ,Ltj ~~~~r;~t,e;=;1 .sp¥rm~r;ii~st 1(o.q.
RSX-HM-PLUS only), if reqU,e:s.ted ,. by . ~a;l;l.1;ng the. : pr;~~d.ure~

control

alJr '.

~~...

'.i

'c4-ii , i 1 • ••

it~·

, ·

i:....!.:i" :....lr~!ul · , '-rlf~~nhnf ,

~uri..ERROR,,: :>2v,M GE°11~RJ:,,: .aHvt ·~~ ~J-~~~lxt
1

•

·.

~4n;?:-,Sf'!.l\~f\~ .. ;r.~~~-~Y.~l,~:" . . ~~!p;~lk~.:t¥~" ,,.!;2

" ,. . !du;~w ;. ·~· , .. 1

i~ ~~~·x.~c1 ,~:1·~~~~P,,

~,!~.";f~:~w•~!.t.~4~"'~}:!~.

E'~i~_;l+0~~~g;;i·lll·t~~;t~Ji1li~i.~M~"~+:1~1•~¥#~·:%;"li!1t",J"f·;r:cJ .• t• .•1j111,

the

!;;:;.·.p.,.,;s."';1;:;:.,9

The DISPATCH procedure in the DISPATCH module declares all of
the common subpackets.
These are, in order of appearance,
HEADER,
TASK,
DEVICE ID,
DEVICE OP,
and
IO ACTIVITY
(conditionally
supported
on
RSX~llM}.
Each of these
subpackets has a mask bit
in the HEADER subpacket which
indicates the presence of the subpacket.
If the bit is set,
the subpacket is present and therefore declared.
If the bit
is not set, the subpacket is not present and consequently not
declared.
Note that the HEADER subpacket must always be
present.
As each subpacket is declared, various tests are
performed that must be passed or the entire packet
is
rejected.
These tests are for the various selection criteria
that the user can specify using command line qualifiers.
If
the tests are passed, the procedure then computes the name of
the appropriate dispatcher module.
The dispatcher module
name is derived by concatenating the following elements:
•

The string "DSP"

The event
decimal

The string "Ml"

code

(HEADER.CODE_TYPE)
(for RSX-llM)

or "Pl"

converted

ASCII

(for RSX-llM-PLUS)

For example, an event with a code of 5 would be dispatched to
the module DSP5Ml or DSP5Pl, depending on the operating
system.

4-10

ERROR LOG CONTROL FILE ARCHITECTORE
6.

4.2.3

The dispatcher modules (or modules they may call) handle the
declaration of the DATA subpacket if one is present. The
dispatcher modules also perform further selection tests, as
appropriate.
Eventually, the dispatcher module decides
whether or not information about the event should be printed.
BRIEF format reports are printed entirely by the dispatcher
module. FULL and REGISTER format reports are printed by a
combination of:
•

One or more of the RSX-llM formatter modules
(FMlNMl,
FMlWMl, FM2CM1, FM3CM1, FM4NM1, FM4WM1), or one of the
RSX-llM-PLUS formatter modules (FMTNPl, or FMTWPl)

•

The appropriate dispatcher module

•

A device-level module (if it is a device error), or
CPU-level module (if it is a processor or memory error)

Compilation Paths

For both the RSX-llM and RSX-llM-PLUS operating systems, the DISPATCH
module must be compiled first.
The next modules to be compiled are at
the next level, namely (for RSX-llM) ERRORM, DSP2Ml, DSPlMl, DSPSMl,
and
PARSEM using as input the symbol file produced from the
compilation of DISPATCH. Modules in the same group, such as ERRORM,
INITMl, SELTMl, and FINLMl all use the same input symbol file (in this
case, DISPATCH) and can be compiled in any order.
Figures 4-2 and 4-3 indicate the compiiation paths for the RSX-llM and
RSX-llM-PLUS modules, respectively.
DISPATCH

ERRORM
INITM1
SELTM1
FINLM1

DSP2M1

DSP1M1
DSP3M1
DSP4M1
DSP6M1
DSP7M1

DSP5M1

DEVSM1

FM1NM1
FM1WM1
FM2CM1
FM3NM1
FM3WM1
FM4NM1
FM4WM1

DEVICE
LEVEL
MODULES

E11XX
E1134
E1144
E1160
E117X

DEVUDA

PARSEM

PARS1M

PARS2M

PARS3M

I
PRS2AM
PRS2BM

ZK-1112-82

Figure 4-2 - Compilation Path for RSX-llM Control File Modules

4-11

ERROR LOG CONTROL FILE ARCHITECTURE

ER~ORM

. OSP2El'

.OPSlP1

·oSP3P1 ·

, , . JNIJP1.

SELJMl /,

DSf>4P·t·.··
, ·. QSP.6P1 ... ·

F!Nt..P1 . '

. ·'.OSP7P1

' 'FMTN,PJ
FM7WP1.

Where modules in the figure are connected by vertical lines, the upper
module is compiled first.
The lower module or modules are then
compiled using the symbol file produced by the module at the next
higher level.
Therefore, again using an RSX-llM example, the DSP2Ml
module is compiled using the symbol file
from DISPATCH,
the DEVSMl
mu<lule i::s compiled u::sing the ::symbol file fr:om DSP2Ml, dud ::so

Oll.

Many of the modules in the RSX-llM error log control file have a
common source with RSX-llM-PLUS modules. Compile-time conditionals in
some of these modules generate variants specific to each operating
system.
Modules used
in the RSX-llM Error Logging System must be
compiled using the following compile-time literal declarations:
Option>LITERAL SUPPORT.RSX llM =TRUE
Option>LITERAL SUPPORT.RSX-llM PLUS= FALSE
Option> LITERAL SUPPORT.IO ACTIVITY= FALSE
Ah+-;,..,..., I
"I:'""' .... ",.., I

The declaration,
LITERAL SUPPORT.IO ACTIVITY =FALSE
can be changed to TRUE to enable processing of I/O
activity
subpackets.
If you choose to do this, you must recompile all control
file modules and generate a new system after defining the symbol
E$$ACT in RSXMC.

Modules ;qsed . in the .RSX-llM~PLUS Errdr Logging ~ystem .~u~t
.~s;o9 .t~~ f'opoo/*r;I9, :99~Pi.fe+;t:ime +i't~ral :?7clar~F~f?~~i:~j £. "·
:}ept;fon> iL.:I T!!R).:l:. :. S~PP()~T;~RS* •. ·l:]:M .;=•··. F:AL~E: ., .. ,

\Opt~o1:i:>it'l~R$ .:S:Df?.~qR;!{',.~sx--ll;M'. ;P;Lu~·' ~· ~jRU)!i;: ''
• option>'.Lr'r:E~AL;soPfO~T•IO Xcr1v;·i~.~r::=: 'I'RpE·i,: : :.
·:option>:;, · · ·. · ; : ~: '·
.. ,,_ ·:. : »; ;; · ·. •. · : . · ·,.
·~
See Chapter 5 for a description of the Control File Language
these declarations.

4-12

used

ERROR LOG CONTROL FILE ARCHITECTURE
4.2.4

Modification and Recompilation

You can modify any control file module.
After doing so, you
recompile the module and replace it in the control file library.

must

There is one very important
control file module:

any

rule

remember

when

modifying

IF
your modification to a module creates new groups, tables, or
dynamic tables, OR creates a new variable within any of
these structures, OR reorders a variable within any of these
structures
THEN
you must also recompile all modules on the same branch
the tree at levels lower than the modified module.

END IF
Note that recompilation of lower-level modules is not necessary if you
modify the run-time logic.
For example, modifying the statement
IF %STR$LENGTH(PARSE.SWITCH_LIST) EQ 0
to
IF %STR$LENGTH(PARSE.SWITCH_LIST) EQ 1
in the PARSEM module would not require recompiling any of the lower
level modules, namely PARSlM, PARS2M, PARS3M, PRS2AM, or PRS2BM.
However, changing the line to read
IF %STR$LENGTH(PARSE.SWT_LIST) EQ 1
and creating the new variable SWT LIST in the PARSE group requires
recompilation of the lower-level modules.
This is because the
information in the symbol file consists of
group
names
(in
alphabetical order) and the variables defined within the group (in the
order declared).
The compiler uses the information from the input
symbol file to compute relative group and variable numbers for use
when a module references a group and/or variable declared in a higher
level module.
These group and variable numbers, rather than the
names, are used to resolve references to groups and variables when a
module is loaded. Defining new groups, or variables within a group,
changes the relative order of these symbols.

4.3

INTERNAL INTERFACES

This section discusses the specifics of various internal interfaces of
the Error Log Control File modules.
All of the modules used as
examples in this chapter appear at the end of this chapter.

4.3.1

Interaction Between Dispatcher and Device-Level Modules

The following two sections describe,
in detail, the interaction
between a dispatcher module and a device-level module, using the
processing of an RM03 error as an example. Section 4.3.1.1 describes
the interaction using an RSX-llM module (DSP2Ml). Section 4.3.L2
4-13

ERROR LOG CONTROL FILE ARCHITECTURE

the, "di:ffei:~nce·s· ~tween
~SX:""l.lM-:-PLUS. n\J><iul~ (.DSt>,2:P:tr.;;

describes

~he: '.:~~ispa-tphe~.;:Jno~µ,:l;~s ":.;y~i119:, ·an;
. '.·. ';>;• ... <;; ::·: .. ,,•
.•

For this detailed examination, the following discussion refers to the
ERM23 device-level module code in Section 4.6.1 and the DSP2Ml
dispatcher module code in Section 4.6.2. Both the discussion here and
the code in those two sections are keyed to each other by the module
names (either ERM23 or DSP2Ml) and numbers that look like this:
Ct.
You may wish to remove the pages for Sections 4.6.1 and 4.6.2 from
your book for easier reference in following the interaction between
these ·two modules.

4.3.1.1 Interaction between DSP2Ml and ERM23 - Processing in DSP2Ml
begins with DISPATCH having declared all subpackets, except for the
DATA subpacket.
All subpackets, except for the TASK subpacket, are
needed for a device error.
DSP2Ml begins by determining that
peripheral errors are requested and that the subcode is valid.
Having
completed these checks, DSP2Ml calls DEVSMl (DSP2Ml
DEVSMl
returns
device
information
in
three
variables.
INTERMOD DEVERR.DISP NAME contains the name of the device-level module
needed ~o process the DATA subpacket,
in this
case
"ERM23".
INTERMOD DEVERR.DRIVE TYPE contains the string "RM03" for the drive
type.
The last variable, INTERMOD DEVERR.ALT NAME contains the string
"RM02/03" for the alternate drive-type.
(Thi alternate name variable

t»).

.!_

--.L

----!I

.:11 ____ !

J.o::>

UV\..

Ui:>t::U

UULJ.lJ':::f

------

_ _ _ _ _ _ _ .:

Ut::VJ.1,.;t;:

,.:1 _ _ _ ,!

t::LLVL

.1:-'LVl,.;l:::i:>i:>.Lll'::j•/

After returning from DEVSMl the NOTE NUMBERS file is cleared
(DSP2Ml
f) ) .
This deletes any records that may remain there from previous
events.
The next step establishes the coroutine relationship with
the
device-level moau1e
(DSP2Ml ii;.
Tne DEVICE ERROR procedure in the
DSP2Ml module is one partner while the DEVICE ENTRY procedure in the
ERM23 module is the other.
Control passes to the DEVICE ERROR
procedure.
The first thing the DEVICE ERROR procedure does is pass control to its
partner
(DSP2Ml
G) . Module ERM23 receives control at the beginning
of the DEVICE_ENTRY procedure (ERM23 f) ) .
DEVICE ENTRY proceeds to declare the DATA subpacket (ERM23 C) ) .
Once
this Ts completed the INTERMOD DEVERR variables are filled in (ERM23
0 and f) ) . All of the variables must be filled in.
If the
information
for
a
particular variable is unavailable or not
applicable,
use
the
string
"N/A".
For
the
variable
INTERMOD DEVERR.ERROR CYLINDER the string "???" also has a special
meaning;
it indicates to DSP2Ml that the section titled Device Error
Position Information is to be suppressed.
Once the INTERMOD DEVERR variables are all
filled
in,
DEVICE ENTRY
will coroutine back DEVICE ERROR (ERM23 ~ ) • DEVICE_ERROR regains
control where it left off (DSP2Ml 0 ) .
DSP2Ml then performs the serial number tests,
if required, after
having first
initialized the variable INTERMOD DEVERR.REJECT FLAG to
FALSE.
If the serial number test
is
faTled,
the
variable
INTERMOD DEVERR.REJECT FLAG is set to TRUE.
The path through the two modules now depends on the report format,
either BRIEF,
FULL, REGISTER or NONE.
Following are explanations of
each of these paths.

4-14

ERROR LOG CONTROL FILE ARCHITECTURE
BRIEF
The REJECT FLAG variable is tested (DSP2Ml (i) ) •
If TRUE,
nothing is output. If FALSE, one line is output which contains
the information required for a BRIEF report. In either case, the
variable INTERMOD DEVERR.PRINT FLAG is set to FALSE.
FULL
The REJECT FLAG variable is tested (DSP2Ml f) ) .
If TRUE,
nothing is-output and the PRINT FLAG variable is set to FALSE.
If REJECT FLAG is FALSE, the four formatter modules are called to
print the information from the common subpackets. The DSP2Ml
module then prints (still on the first page)
the information
passed back in the INTERMOD DEVERR variables filled in by ERM23.
DSP2Ml generates a page break, then prints a header on the second
page. When done DSP2Ml sets the PRINT FLAG variable to TRUE.
REGISTER
The REGISTER path is almost identical to the FULL path. The only
difference is that the page containing all of the common
information is not printed. The header on the page containing
the register translation supplies a summary of the information
instead.
NONE
The NONE path sets the PRINT FLAG variable to FALSE (DSP2Ml Ci) ) ~

All of these paths converge again at DSP2Ml
At this point
control once again passes to the DEVICE_ENTRY procedure in ERM23.
The first thing the DEVICE ENTRY procedure does upon regaining control
is test the PRINT FLAG -variable (ERM23
If it is FALSE, the
module exits (ERM23 -4D ) .

CD ) .

If the PRINT FLAG variable is TRUE, ERM23 performs the bit-to-text
translation -of the registers. Following that, any required notes are
indicated by PUTs to the NOTE NUMBERS file specifying the note index
{ERM23 ~ ) • The module then exits {ERM23 4D ) .
When ERM23 exits DSP2Ml regains control and the coroutine partnership
is broken (DSP2Ml ~ ) • The DEVICE ERROR procedure then checks for
entries in the NOTE NUMBERS file.
If there are any, DSP2Ml computes
the name of the notes file. The name of a notes module is the same as
its corresponding device-level module except the first character of
the module name is the letter "N", in this case NRM23. The notes
module is then called to print the requested notes.

4.3.1.2 Interaction Between DSP2Pl and ERM23 - The relationship and
flow of control between the DSP2Ml/DSP2Pl module and device-level
modules is identical. They both pass the same information and control
back and forth at the same points. However, there are differences in
the modules themselves. This section explains those differences.
As in the previous discussion, you may wish to remove the pages for
Sections 4.6.1 and 4.6.3 from your book for easier reference in
following the interaction between these two modules.

4-15

ERROR LOG CONTROL FILE ARCHITECTURE
Following are the differences between the DSP2Ml and DSP2Pl modules.

t» ) .

DSP2Pl declares the logical variable INDICATE.TAPE FLAG (DSP2Pl
This variable is set by DEVSMl to indicate whether or not the device
is a magtape. The variable is used later in processing summary
information.
If the packet is not rejected and if the report format is not NONE,
the variable REPORT.PRINT COUNT is incremented (DSP2Pl ~ ) • This
variable keeps a count of how many events were printed (as opposed to
looked at, which is a separate tally).
After the device-level module has printed (if instructed to)
and has
exited back to DSP2Pl, the UPDATE RECORD procedure in DSP2Pl is called
(DSP2Pl C) ) .

The UPDATE RECORD procedure tests to see if an ERROR summary was
requested
(DSP2Pl Gt ) •
If not, processing goes on to the GEOMETRY
section.
If an ERROR summary was requested, DSP2Pl searches the ERROR INFO E
file to see if an error having the same error type has been
encountered.
If so, the record in the file describing that type of
error is updated to show that one more error occurred, and when it
occurred.
If no such error is found in the file, a new record that
describes the error is added to the file.
Processing then goes on to
the GEOMETRY section.
The UPDATE RECORD procedure then tests to see if a GEOMETRY
was requested (DSP2Pl
If not, the procedure exits.

€t).

summary

Updating the ERROR INFO G file is much the same as updating the
ERROR INFO E file~
The only difference is that the information
recorded is somewhat different.
In particular, the GEOMETRY summary
records information regarding where on the device the error occurred.
It is for this reason that we need to know whether or not the device
is a magtape; magtapes have no valid geometry information.

4.4

DISPATCHING

This section discusses module dispatching. Dispatching happens at two
levels: event-level dispatching and device- or CPU-level dispatching.

4.4.1

Event-Level Dispatching

All events that occur in the Error Logging System are assigned a
unique
combination
of
code
and subcode.
These code/subcode
combinations can be found in the file EPKDF.MAC
(EPKDF$ macro in
EXEMC.MLB)
along with the definition of the structure of error log
packets.
See Appendix C for a listing of EPKDF$.
Table 4-1
summarizes the error logging code/subcode combinations.

4-16

ERROR LOG CONTROL FILE ARCHITECTURE
Table 4-1
Error Logging Code/Subcode Combinations
Subcode

Code
1. Error Log Control

1.
2.
3.
4.
5.
6.

Error Log Status Change
Switch Logging Files
Append Fi le
Declare Backup File
Show {not logged)
Change Limits

2. Device Errors

1. Device Hard Error
2. Device Soft Error
3. Device Interrupt Timeout
(hard)

~~(~,j~:J~\f

,. '' . . ·.~·,~~;~~t~i~:~~

's·.···· Device Interrupt Timeout
(soft)

3. Device Information

1. Device Information Message

4. Device Control Information

1.
2.
3.
4.

5. CPU-Detected Errors

1. Memory Error
2. Unexpected Interrupt

6. System Control Information

1. Power Recovery

7. Control Information

1.
2.
3.
4.
5.

Device Mount
Device Dismount
Device Counts Reset
Block Replacement

Time Change
System Crash
Device Driver Load
Device Driver Unload
Message

Each code group is processed by one of the dispatcher modules.
These
modules are named DSPlMl/DSPlPl, DSP2Ml/DSP2Pl, ••• , DSP7Ml/DSP7Pl.
The name of the dispatcher module is derived on the fly in the
DISPATCH module's DISPATCH procedure by concatenating the following
elements:
•

The string "DSP"

•

The event code (HEADER.CODE_TYPE) converted to ASCII decimal

•

The string "Ml" (for RSX-llM) or "Pl" (for RSX-llM-PLUS)

The single-digit ASCII conversion of the code value (obtained from the
HEADER subpacket)
is required because the RSX-llM/M-PLUS Librarian
utility LBR allows a maximum of six Radix-50 characters for a module
name.
The code value 9 is currently unused; values 0 and 8 are
reserved.
Once the dispatcher module has been called it checks to see if this
type of event was requested.
If the event type was not requested, the
module returns, effectively ignoring the entry.
Event types are
requested by using the /TYPE command line qualifier.
The event types,
codes, and the dispatcher modules that process them, are listed in
Table 4-2.
4-17

ERROR LOG CONTROL FILE ARCHITECTURE
Table 4-2
Event Types, Codes, and Their Dispatcher Modules
Type

Codes

ALL

1-7

DSPlMl/DSPlPl, ••• , DSP7Ml/DSP7Pl

CONTROL

DSPlMl/DSPlPl

ERRORS

2,3,5

DSP2Ml/DSP2Pl, DSP3Ml/DSP3Pl, DSP5Ml/DSP5Pl

MEMORY

DSP5Ml/DSP5Pl

PERIPHERAL

2,3

DSP2Ml/DSP2Pl, DSP3Ml,DSP3Pl

PROCESSOR

DSP5Ml/DSP5Pl

SYSTEM INFO

4,6,7

DSP4Ml/DSP4Pl, DSP6Ml/DSP6Pl, DSP7Ml/DSP7Pl

Dispatcher Modules

Once the dispatcher module determines that this type of event was
If it is not,
requested, it checks to see if the subcode is in range.
the event is rejected with an error message.
At this point dispatcher modules may declare and print the DATA
subpacket themselves or may call lower level modules to do so. The
error-logging dispatcher modules handle all of the printing for the
BRIEF report mode.
If the FULL report mode is specified, the
dispatcher modules call one or more of the following modules to print
the common portions of the event:
System, Width

Formatter Module(s)

RSX-llM, NARROW

FMlNMl, FM2CM1, FM3CM1 and FM4NM1

RSX-TlM, WIDE

FMlWMl, FM2CM1, FM3CM1 and FM4WM1

The dispatcher module may print the rest of the event itself
with a lower-level module.

4.4.2

work

Device-Level Dispatching

Device-level dispatching is performed with the assistance of the
DEVSMl module.
This module is called by DSP2Ml/DSP2Pl, DSP3Ml/DSP3Pl
and DSP4Ml/DSP4Pl and determines, among other things, the correct
device-level module for the event.
Here is a description of how DEVSMl works (see the source code for
exception cases;
this discussion addresses only usual cases). The
first thing DEVSMl checks is whether there is a DEVICE ID subpacket.
If no DEVICE ID subpacket is found, an error results.
Once past that
check, DEVMSl-uses the device mnemonic to search the DEVICE INFO
table.
If the device is not found, DEVSMl specifies the EUNKWN module
in the variable INTERMOD DEVERR.DISP NAME.
Assuming that the mnemonic is recognized, DEVSMl tests to see if
(a)
the mnemonic is that of a MASSBUS device, and (b) there is a DATA
subpacket.
Assuming both are true, DEVSMl looks ahead into the DATA
subpacket to obtain the MASSBUS Drive Type from the logged registers.
4-18

ERROR LOG CONTROL FILE ARCHITECTURE
This value is unique for each MASSBUS device.
Once this value is
obtained, the DEVICE INFO table is searched again, this time using the
drive-type value as the key.
Assuming this search turns up a match,
the variable INTERMOD DEVERR.DISP NAME is filled in with the module
name specified by the resulting record in the table.
If there is no DATA subpacket, or if the device is not a MASSBUS
device, the search of the table ends up pointing to the first record
that matched on the specified mnemonic.
DEVSMl performs a further
search of the table based on the mnemonic as well as the device size
(which is provided in the variable DEVICE ID.DEV TYPE). The variable
INTERMOD DEVERR.DISP NAME is then filled
in -with the module name
specified in the record that is the result of this search.

4.4.3

CPU-level Dispatching

CPU-level dispatching is performed by DSP5Ml/DSP5Pl.
The HEADER
subpacket contains a variable called PROC TYPE that indicates the type
of processor the error was logged on.
DSP5Ml/DSP5Pl uses that
variable to search a table that contains module names associated with
the CPU-type value.

4.5

SUPPORT OF NON-DIGITAL DEVICES

This section explains what you have to
support for non-DIGITAL devices.

provide

error-logging

Adding error-logging support for a non-DIGITAL device consists of
either one or three steps, depending on the desired level of support.
The first step is to include error-logging support in the driver.
Without this support no information can be logged for the device. For
full error-logging support, you must perform two more steps: write a
device-level module for the new device, and add it to the control file
library and make the Error Logging System aware of the new module.
The following sections show you what
steps.

4.5.1

you

need

accomplish

these

Error-logging of Unknown Devices

The Error Logging System can handle entries from devices unknown to
the system. Entries from an unknown device are handled by the EUNKWN
module, which functions as a universal device-level module.
For a
BRIEF report, EUNKWN will pass back "N/A" in the INTERMOD DEVERR
variables to indicate that the information is not available.
For a
FULL report, EUNKWN prints the device registers in a dump-style format
where the bit-to-text translation would normally take place.
The rest
of the report is unchanged.

4.5.2

Providing Driver Support for a Non-DIGITAL Device

The Executive module ERROR contains the routines to be used by a
driver to log device errors. A device error in this sense can be a
real error, a timeout, or perhaps an informational message.
The
following sections discuss the routines in general. See the code in
[11,lO]ERROR.MAC for more detail.

4-19

ERROR LOG CONTROL FILE ARCHITECTURE
For the most part, driver support is the same for RSX-llM and
Where there are differences, the full discussion is
RSX-llM-PLUS.
repeated.

4.5.2.1 $BMSET on RSX-llM - On RSX-llM, the $BMSET coroutine raises
the processor priority to 7 (to lock out interrupts) and then calls
the caller to start the I/O function.
When the re-called caller
returns, $BMSET lowers the processor priority to O, thus allowing
interrupts once again.
INPUTS
None
OUTPUTS
None

the

.4 .s. 2'. 2 · $BMSET .· on : R~;...llM-PLu·s :- on· · ·.R·sx-1'.1~".'"PLU:Sr. ·. · ·
$BM~J~'!T::.·
coro·u.tine· .raises' the· processorpr.ior:~t.y· 'to» 7 . . {t'O :l()Ck..·ou·t· i)\te,rrtipts).,,.
:set~· the ''intertu:pt active" b~t S2.~CT inS .. S'I'Z.of 'th~· SCB, .·'and
then:
caU s . the cal,l·er ·to start the I/O f:unction. . Wh:en· the: re.-·ca:lJ~ :caJl:e;r·
re~urns, · $BMSET I owe.rs th.e .;pr,ocessor. pr ior;\~.Y " .t~: . . · O., :thu.s. . all.o:w,:,in:9.
interrupts once again.
· ··
INPUTS
R4' = SCB Address
OUTPUTS
The "interrupt active" bit S2.ACT is set

4.5.2.3 $DVTMO and $DTOER on RSX-llM - On RSX-llM, the routine $DVTMO
logs device timeouts at PRO, and the routine $DTOER logs device
timeouts at device priority. The routines behave identically, except
that $DTOER disables the device interrupt and lowers the processor
priority to 0.
If the symbol D$$IAG is defined, the routines test to see if the
timeout is a diagnostic function.
Diagnostic functions are never
logged.
The routines load the error code and subcode in
falls into the routine $DVCER.

and

Rl = I/O Packet Address (if D$$IAG defined
function)

and

the

routine

INPUTS
R2
R4

CSR Address

= SCB Address

OUTPUTS
a

diagnostic

C = O, if D$$IAG not defined or not a diagnostic function.
Create an error log packet and fill it in.
Put a pointer to
the packet (S.BMSV) in the SCB and set the "error in progress"
bit SP.EIP in S.PRI.

4-20

ERROR LOG CONTROL FILE ARCHITECTURE
C = 1, If D$$IAG defined and a diagnostic function.
Set the
"error in progress" bit SP.EIP in S.PRI. Do not create an
error log packet.

·«4.5·.~2.,4 ' ...$DVTMO 'and·' $0.!'QER' on ... , RS~.;.llM-PLUS·.~ On RSX-UM.:..:PLUSj
the
.,rQutine, .••. $PVTMQ. lqgs. dev.i9e ...timeout·s·.~t 'PRo·, Qnd ·th~ r.outine $DTOER
·logs. ·.dE?v'ice ·.timeout~ .at'·deyice. · ptior~·t:t.y.·· The. routines' .··.behave:
i<I(imH.ca:lly.···<:e,xcept· ... :th~t . . :$DTOER· .. dJsabtes ·. :t:,l)e··· <le,vice': interrupt. and
·il~:W·E:·i::s·.· ~·):l:e·· P~:?ce$s9x·. ::p~i<;>'.r<i,ty .. :t.o··:o·~

.:Th~,'..rp·utiries·.::cJ.eai·. ~hf;\.~~.·~pt~~·r.up.t.··a<;tiV:en':b:it: .$2.:~/~cT.t~:::tti~··· ·Scs·:···•. ·:wo·ra'..

<~e, ro~:;i·o~~- {oad , tne ~r ~or , ~J~~::'ai>d :~~bd~e. iflio / ~~n~

·:faH:s ''into tne· · r·ou.t"lne :$ovc·1~1t{. ·.'. · ·.. · · ·. : ·.

:::

o,),::~'':~-~<>~·:~.:
.·~·

«<,,·RS·,~. UCB.· Address
'~o

,'',

;,";<."<

''/,;

<,·

';;'',.</,

,':'.~ ':,, ',,:<·;, «,

tti.> ( rd4title

<" '. ··: ·· ··· <. · ·:

......

"' '
;<'<;;,~,../

>,.'<.<'·,
',

',~'.~~-~,·:,~,v",' ', ,·>t.~ ,·', .<~

.- ··:tt'.~. ONR.'.·.·a·na·.· 37.7. ·<roe'vlce'
·~... :r.10: ·Packet. Address . ·

',~ .<~,,;;:.,"'·v,

Y,··'y'~:;

' ..

:·,''

"bot R~~4y):

, ·· ·

¢··:; (l,i:,:~+· ... ~ot,•..a.;dia~no;~ti9, ·~,;~u:nc.ti,.on~.· ·.... "·C:r~ate ...an.: ·eri;o:r.:>.·109·

· pp,cket

.and ".fill it in.
'.'i()··~.th~.>$CB~:ana··set.·t:l:le
S •.. P.Rl~ : .

'Put. a pointe.r to :th.e packet.{S .. BMSV).
·'iri' ·Pro<jre.~s".. bit' ·s;.?~E .:J:P'.. in
. ,

. . .

4.5.2.5 $DVERR and $DVCER on RSX-llM - $DVERR and $DVCER are the
same;
$DVCER is the routine name, and $DVERR is a synonym. This
routine logs device errors. If an error is already in progress on the
device, it will be ignored. If not, $DVCER allocates an error log
packet and fills it in with the context of the current transfer. Note
that this routine requires that there be an I/O packet associated with
this error. See the routine $LOGER (Section 4.5.2.9) to log an error
where there is no I/O active on the device.
INPUTS
R4
RS

SCB Address
UCB Address

OUTPUTS
If no error is already in progress on this device, allocate an
error log packet, fill it in, point the SCB to the packet, and
set the "error in progress" bit.
If an error is in progress on this device, this routine
no-op.

4-21

ERROR LOG CONTROL FILE ARCHITECTURE
. 4h. 5~.2. 6

$DVERR. an·d· $ovbEk. on asx..:f18..:etus .. - ·;$:0VfERR. ·anCi $DV'C~R · ai·e:··the'.·

same r
.$DVCER 'is the· ro·uti ne · ~name} '.and $,I>VE~R r:s. ·,a S•Y'r\QO:~ ·• . :Thi.s
routin:e· 1:09$ device. err.or:s. If. a() ,e~·r:o1 r: .i9 a:1:r.e'.a(ly: :in ·pro<.tr:es:s<on\'. the.
dev.ic.e, lt . w'ill ·be.· ignored. ·i,f. .n9t',1·. $PV;CER.'.a:1<1,,<~cat.~$: .a·n, ·;~r:r'or; .1:o'g:
pa,cket anq .fi11s, it' ·in. with. the ·co·n·text of .the: qu:r:rent .tra.nde:t";: : .Notte·
.·tha,t:: ·this· ·.rou't:fne· r.equ iX:e:s. ·<that :t.h~:re. ~~: ·an. ;.!/{)'. ··~a:c;k·et, .:a'.s:s.p:ci/a;~ )~i;th>
''t'.hli:s,. ~r;r,,o;r .~; ·~,,see..th;e, ,ro;u,ti~~ne; 1$,IfG~R; J;~~t',H:>:~ r4'.. 5:·;2,.;~~>.: :t~ ?~:. ;a::n: ;~rtr;9:r~,:
1
1
1
1
.{J
Nblt:e~. ·that: . on RSX-1'.lM-P!LiUS ~.· '.ittfior~t:iio:ni 'ah,p:·uit'. bdnburlr"eAtl ·fri/ld i ·rldtli~Vli1t;y/
1

rn !'°i;t1t ~ f'T'.l, 1. 11 ji; Li ilJi r;JHHJ;;
···tt\m~;tC~s• J al', ·i~e&:.~• ·tj j.l·•j(r•·J /j • ·~.:·.••: ffl' JJJril1i tti··i.· ! ·1·1 [
·~r;rrrer:c l'~\ 'If ~~ "]lb,

s' .

tt~i~ 1 :l;;J:r~~~~ :~~e;r; ft'i-ffi ·
',,;;::.J:"•t'.! ::i7i•I!f'. ;no :e;r!r~r; is :1a;
1

l>•:ic.>~1rlrdr'' "Im;

'Jlilll~~l,~~t"
I

,,,

yt 1in! .:

:;11 1

rr·imt~ ~ 11

:o~i:lejs"s:<;~ '!th!ilsi ~rv::x~i,;, ·Jall~~t
\

~i'n!!:t •ltbe \

•! it;)\

r;·~

w ~1~Yf fa~ H,~ ~id·ft

$NSIER - This routine logs nonsense interrupts.
identifies the interrupting vector and logs the error.

4.5.2.7

The

routine

INPUTS
@(SP)

Contains bits 06:04 of the unused vector number.

OUTPUTS
If a nonsense interrupt is in the
increment the interrupt count.

process

being

logged,

If this is the beginning of the processing of a nonsense
interrupt, identify the vector and create and queue an error
log packet.

$FNERL - This routine is called at I/O completion, or when it
is necessary to queue an error log packet after a successful recovery
of a mid-transfer error.
In essence, this routine completes the
processing of an error.
4.5.2.8

The routine first inserts the error retry information. It then tests
to
see
if this was a hard (unrecoverable) error or a soft
(recoverable) error, and updates the packet accordingly.
All errors
are assumed to be "hard" up to this point. Depending on the result of
that test, $FNERL tests against the appropriate limit to see if the
limit has already been met. If the limit had been previously met the
packet is discarded. If not, $FNERL updates the appropriate error
count, logs the packet, and sets the SCB to show that the processing
of this error has been completed.

4-22

ERROR LOG CONTROL FILE ARCHITECTURE
INPUTS

RO = First I/O Status word
R2 = Starting and Final error
update limits)
R3
Error Log Packet Address
R4
SCB Address or 0
RS
UCB Address

retry

counts

(if R4

= 0)

(if

not

OUTPUTS
Either queue or discard the error log packet (depending on the
limits)
and set the SCB to indicate that no error is being
processed.

4.S.2.9 $LOGER - Drivers use $LOGER to create an error log packet
when no I/O is present, such as when a driver receives an unsolicited
interrupt from a device that contains
information that should be
logged.
$LOGER creates the packet normally, but the driver is
responsible for filling in the DATA subpacket information.
Otherwise,
processing is similar to $DVERR.
INPUTS
Rl
R4
RS

Length of data to be logged (in bytes)
SCB Address (If O, then no I/O packet is present)
UCB Address

OUTPUTS

1, Error cannot be logged for some reason

C
Rl
R3

O, Error can be logged
Address of DATA area in the packet

= Address of Error Log packet

4.S.2.10 LOGTST - This routine is not for
use by drivers.
Other
routines in the ERROR module call LOGTST to see if an error can or
should be logged.

4.S.2.11 $CRPKT - This routine creates an error log packet.
It is
called as part of the $SMSG directive processing, and by other
Executive routines as part of the processing of a memory error,
nonsense interrupt, time change, power fail recovery, or device error.
In general, the routine determines the required format and size of the
packet, allocates the required amount of pool, and then fills in the
packet.
It obtains information from SYSCOM,
the appropriate DCBs,
UCBs, SCBs, TCBs, VCBs, and the I/O packet, as required.
Note that a HEADER subpacket is
crash will result if $CRPKT
subpacket".

A forced system
always required.
detects the condition of "no HEADER

Note also that on RSX-llM,. information ·about concurrent I/O activity
on other devices can· also be optionally logged. Do this by defining
the symbol E$'$ACT and doing a new system.
(You must recompile the
error-logging control files, as well.)

4-23

ERROR LOG CONTROL FILE ARCHITECTURE

.On: 1 .RSX-llM~·P·uhs:, .· i.n·fo~afi·6tr. abox1t·coricU,r'r:ent l/b:· :a~tfvity' on- ··oth~er- .
I

devib~s ',i&' 'alwa-ys ~~CJ~<?' 1.Jf: ada:i;tfol'); <t·~ ·the a·¢tiv1ty <m.- the 4~-vfoe

q-ues;~t<J~:..:

·· .•:.

...

in.

INPUTS
RO
Rl
R2
R3
R4
RS

Packet Code and Subcode (See EPKDF for details)
Length of DATA subpacket
Control Mask word (See EPKDF for details)
Beginning Address of data for DATA subpacket
TCB Address (for TASK subpacket)
UCB Address (for DEVICE IDENTIFICATION subpacket)

OUTPUTS
RO
Rl
R2
R3
R4

Unchanged
Beginning Address of data in the DATA subpacket
Unchanged
Beginning Address of Error Log packet
Unchanged

C = O, A packet was created
C = 1, A packet was not created
RS = Unchanged

, ··~Ai.A.,c:..,.~,~;~h"

-:1·,.:,

••.J:-,.c..•·..a...6""

,..·~i.Y.·-no:...,,. ~·..- -~~," ""1 ·,,~,< ~,~,Y'TC': ,
'""'Clr..l.l5J',W"W

Vt.&

.&.'t..1..n.-.i...a...1.·10-~..UVU'

-;-

~ ;:ni£":v \.,,,,.'..u(· .~;r'~..~·,1
v1·.1i,

.\..\.U.A.~.J;..~l.&---:'J:':U,V>Jj

1"''.ri-O"t.~

:1"'11:
"-""1.Ll;VCIV

=.-;-·.11<--. ~.-. 1°-~;•~;W
\wo\Cl..L"""~..Lq.,\...~ ..P

'tbe :,.,11J09ic·a l :unit n·umbeir; folr tlrle 9iven UC•B.'
1

···:'INPUT$
,~:',j
I

l· <>

'< <;

RO! - Pointer ilnto J:;pe · Erro;r ~pg packet

Rl: - ocs· ·Address

· ~q ;~ii:rt:.: t:Q ·ee~t;;J:l~te.•..i~.-•:s~~"q~:: . ~Qg,:·,p~cke'.t
4.S.2.13 $QUPKT - This routine queues an Error Log packet.
If there
is no other packet in the queue, $QUPKT requests the error logger task
(ERRLOG) with a delay of 2 seconds.
If there is another entry in the
queue,
$QUPKT requests ERRLOG to run immediately. Command packets
(from ELI) always cause ERRLOG to run immediately.
INPUTS
R3 = Pointer to packet for insertion in queue
OUTPUTS
None

4-24

ERROR LOG CONTROL FILE ARCHITECTURE
4.5.2.14 $QERMV - This routine removes an entry from the error log
queue and transfers it to a user buffer.
It is called only by ERRLOG.
INPUTS
R4
R5

Length of user buffer
Address of user buffer

OUTPUTS
Rl
R4
R5

Length of packet
Unchanged
Unchanged

O, Packet was successfully removed

C
1, Either no packet to remove or packet was too long.
If
Rl <> O, the packet was too long and Rl contains the packet
length.
If Rl = O, then there was no packet to remove.

4.5.3

Error-Logging Support for a Non-DIGITAL Device

Full error-logging support requires two steps beyond driver support.
The first step is to write the device-level module for the new device.
This module contains the detailed instructions on how to interpret the
logged information, that is, the bit-to-text translation information
for the device registers.
The information common to all events is
interpreted by the DIGITAL-supplied modules.
The second step is to add the new module to the control file library
and make the Error Logging System aware of the new module. The
following sections explain these steps in detail.

4.5.3.1 How to Write a Device-Level Module - This section explains
the general structure of device-level modules, using the RM02/03
module ERM23 as an example. Section 4.6.1 is an annotated listing of
ERM23;
Section 4.6.4 is an annotated listing of the notes module for
the RM02/03. Both the discussion here and the code in those two
sections are keyed to each other by the module names (either ERM23 or
DSP2Ml) and numbers that look like this: . , •
You may wish to remove the pages for Sections 4.6.1 and 4.6.4 from
your book for easier reference in following the interaction between
these two modules.
In general, the flow of a device-level module proceeds as follows:
.,

MODULE statement followed by module header

PROCEDURE statement

SUBPACKET declaration

Declaration of local work variables and table declarations

(i)

Intermodule variable loading

Error-type determination

4-25

ERROR LOG CONTROL FILE ARCHITECTURE

Ci)

Coroutine back to caller

Ci)

Bit-to-text translation and register printing

Note requirements indicated

Exit the module.

Each of these procedures are described in the following sections.

4.5.3.1.1

MODULE Statement -

0
The MODULE statement for a device-level module must.be of the form:
MODULE modulename 'ident'
The module name must match the name specified for this device in the
DISP NAME field of the DEVICE INFO table in the DEVSMl module (See
SectTon 4.5.3.4). Generally, the-module name begins with the letter
"E", followed by five or fewer letters indicating the device or
devices served by the module. For example, the ERM23 module handles
the RM02 and RM03 disks, while the ERP456 module serves the RP04, RP05
and RP06 disks.
The ident field is exactly what it implies, an identification value
that is stored in the module. Generally, the ident begins with a
letter that identifies the operating system the module is intended to
be used with, such as "M", followed by a version and update number in
the standard DIGITAL style.
The module header follows.
This includes the copyright
author, date written, and audit trails of modifications.

4.5.3.1.2

PROCEDURE Statement -

The PROCEDURE statement for a device-level module must be of the form:
PROCEDURE DEVICE ENTRY
The procedure name must be DEVICE ENTRY.
This name is hard coded into
the DSP2Ml/DSP2Pl and DSP3Ml/DSP3Pl dispatcher modules.

4.5.3.1.3

SUBPACKET Declaration -

The device-level module is responsible for the declaration of the
device data
(usually in the form of registers) •
The SUBPACKET
declaration defines the number of registers, how they are printed, and
the bit-to-text translations for the various bits and fields of the
registers. The general format of the statements is as follows:

4-26

ERROR LOG CONTROL FILE ARCHITECTURE

SUBPACKET subpacket_name
reg_ name:

aux label:

= DISP.NEXT_PACKET NAMED

WORD MACHINE
BIT [15]:
BIT (14]:
FIELD (12:2]:

BIT (11]:

reg_ name:

'true text' ;
'true-text',
'false text' ;
'Bits I2 and 13 = '
%CNV_$BINARY(Subpacket_name.aux_label, 2 f
(B) 1

I 0 I)

;

true text'

;

WORD MACH !NE

END_PACKET ;

The subpacket name is usually REGISTER, although this name is not
required.
DlSP.NEXT PACKET is a variable that contains the subpacket
number of the 'data' subpacket and has been set up by the preceding
modules.
The NAMED qualifier indicates to RPT that the register
labels are to be saved for later printing.
What follows next are the definitions of the registers and their
and fields.
The end of the subpacket declaration is
'END PACKET ;'.

4.5.3.1.4

indicated

the

bits

statement

The label assigned to a register provides both a reference to the
register
(a variable name) and a name for the register when printing.
The register name is printed later on (if you specified a FULL format
report).
In most cases, the Error Logging System uses the same
register names used by DIGITAL field service hardware documents.
For the RM02/03, the first register declared looks like:
RMCSl:

WORD MACHINE
BIT [15]:
RMCSl TRE: BIT [14]:
BIT [13]:
BIT [12]:
BIT [11]:
BIT [10]:

RMCSl BA:

FIELD [8:2]:

'*Special Condition set' ;
'*Transfer Error' ;
'*MASSBUS Control Bus Parity Err'
'*Unused bit set' ;
' Drive Available',
'*Drive not Available (other port using it)'
' UNIBUS B Selected for Data Transfer',
' UNIBUS A Selected for Data Transfer' ;
I
BA17:BA16 = I
%CNV $BINARY(REGISTER.RMCS1 BA, 2, '0')
I

BIT

7]:

BIT

6]:

(B)'

;

Controller Ready',
' Controller not Ready' ;
' Interrupt Enabled',
' Interrupt not Enabled' ;

4-27

ERROR LOG CONTROL FILE ARCHITECTURE
The first line indicates that the name of the register
is RMCSl and
that it is a WORD in length (16 bits). The MACHINE qualifier states
that, when printed, the register is to be formatted
in the native
radix for the machine that the report is being generated on. The
native radix for the PDP-11 is octal, and for the VAX-11, hexadecimal.
Other print qualifiers are available to change the radix, such as HEX,
OCTAL, DECIMAL, BCD, BINARY, and RAD50.
The second line defines bit 15 of the register RMCSl,
including when
it is to be printed and what is to be printed. Only one text string
is provided.
This indicates that the bit is to be printed only when
true (set). Otherwise, nothing is printed for that bit.
Bit 14 has a label of RMCSl TRE.
Labels assigned to bits and fields
are never printed. They are allowed so you can reference the bit or
field as a variable.
As with bit 15, the text for this bit is printed
only if the bit is set.
Bit 11 has two text arguments.
The first argument is printed if the
bit is set and the second argument is printed if the bit is reset.
In
other words, this bit will always be printed.
Bits 8 and 9 are defined to be a FIELD with the
RMCSl BA.
The arguments for a field are as follows:

variable

name

FIELD [starting_bit_number:number_of_bits]: 'other string',
'O strTng',
! l - st r: i (J y ! ,
'2-string',

'N_string'
Tne 0 string is printed it the value of the field
is zero.
The
1 string is printed if the value of the field is 1, and so on. The
other string is printed
if the field has a value that has no
corresponding text string. Note that for the field RMCSl BA there is
l'nlt7
..,a..a..a...z

:>n
\.A.I..&

l'f-'ho,..

.,, .... ,..;..,,..

""""'.1.&'-.&.._a.J'-.a..~l..&';je

'T''hovo~n,..o

•1..1'-.L..'-.&..'-'.L..\,,.O.f

+-'h;.,,

\.-J.J..L...;i

~;01,::i

.L..&.~..L.U

;""

.Lft.:1

~1·.•~••t:!'

\A..LYY\.A_J~

..... .,..;;:-+-,.,.,::i

l:".L...l..L.&\..CUe

A technique that is used in the DIGITAL device-level modules is to
declare a field over any contiguous unused bits. The other string is
defined to be 'Unused bits set', and the 0 string is defined to be
NULL
(the null, or zero length string).- If the field has the value
zero, nothing is printed.
If, however, any of the bits are set, the
field appears in the report.
Note that all of the text strings associated with bits and fields have
as their first character either a space or an asterisk. RPT, when
printing the text for a bit or field, removes the first character of
the string and places it in front of the bit or field position
indicator. An asterisk signals some kind of special condition.
For
example, bit 11 of RMCSl can print one of two ways, either as:

[11] Drive Available
or as:

*[11] Drive not Available (other port using it)
Remember that the asterisk does not necessarily indicate an error,
just something interesting.
A blank in front of the position
indicator means a normal or status condition.

4-28

ERROR LOG CONTROL FILE ARCHITECTURE
You can use IF ••. THEN •.. ELSE, CASE, and SELECT
statements
to
conditionalize the declaration of the subpacket.
The statement blocks
in these structures must be enclosed by BEGIN and END.
You can use
variables previously declared in the subpacket even though the
declaration of the subpacket is not complete. Also note the use of
the %LOK (lookahead) functions in various device-level modules.
They
look into a subpacket before it is declared, usually to produce
variables to control the declaration.
Note the variable REGISTER.LENGTH towards the end of the subpacket
declaration in ERM23.
This variable was created when the SUBPACKET
statement was executed.
The variable name
is
of
the
form
subpacket name.LENGTH
and contains the number of bytes in the
subpacket:

4.5.3.1.5

Declaration of Local Work variables and Tables -

0
The device-level module often needs some local variables and tables.
These
are
generally
defined after the end of the subpacket
declaration, although this is not required. Remember, however, that
variables must be declared in a module before they can be used.

4.5.3.1.6

Loading of the Intermodule Variables -

The DISPATCH module declares a collection of variables having the
group name INTERMOD DEVERR.
Some of these ASCII string variables pass
information from the device-level modules back to their caller.
The
variables that must be filled in are:
e

INTERMOD DEVERR.DRIVE SN

INTERMOD DEVERR.DEV FUNCTION

• INTERMOD DEVERR.PHYS ONIT
• INTERMOD DEVERR.ERROR CYLINDER
• INTERMOD DEVERR.ERROR SECTOR
• INTERMOD DEVERR.ERROR HEAD
• INTERMOD DEVERR.ERROR GROUP
•

INTERMOD DEVERR.BLOCK NUMBER

• INTERMOD DEVERR.ERROR TYPE
•

INTERMOD DEVERR.DRIVE TYPE
details on this variable.)

(See

Section

4.5.3.3

for

This section of the module is where these variables are filled
in.
Use the string
'N/A' if the information is either not applicable or
not available. Note that for certain devices, most notably magnetic
tapes, the ERROR CYLINDER variable is filled in with the string '???'.
This flag tells the dispatcher module to suppress the printing of the
section entitled Device Error Position Information.
Note that one of
the variables to be filled in contains the error type. See the next
section for more details on how the error type is determined.
4-29

ERROR LOG CONTROL FILE ARCHITECTURE
4.5.3.1.7

Determination of the Error Type -

The error-type definition is essentially a determination of the most
likely problem as indicated by the error bits for a given event.
It
is not a determination of 'what broke', but rather an indication of
'what happened'.
The error type is determined solely on the basis of
the bits in the current event.
No inter-event analysis is performed.
The error type is determined by a precedence parse of the various
error bits found
in the device registers.
The DECODE statement, in
conjunction with IF ••• THEN ••. ELSE-type constructs, is used to search
the bits in a specific order.
The first condition found to be true
stops the search.

4.5.3.1.8

Coroutine Back to Caller -

Once all of the intermodule variables have been filled in, a coroutine
statement returns control to the device module's caller.
The caller
examines the returned information and determines whether to continue
processing the event.
Nothing has been printed up to this point in
the processing of this event.
If the decision is not to proceed (to reject the event), the caller
(a)
sets the variable INTERMOD DEVERR.PRINT FLAG to FALSE and, (b)
coroutines back to the device-level module.
If the decision is to proceed, the caller performs some or all of the
printinq, dependinq on whether the print format is FULL or BRIEF.
If
the FULL format is specified, the caller (a) prints everything except
the device registers, (b) sets the variable INTERMOD DEVERR.PRINT FLAG
to TRUE, and (c) coroutines back to the device-level-module.
If- the
format
is BRIEF,
the caller (a) performs all required printing, (b)
sets the variabie INTERMOD DEVERR.PRINT FLAG to FALSE
and
(c)
coroutines back to the device=level module~
When the device-level module regains control
it examines the print
flag.
If TRUE, the module prints the device registers and generates
any required note indicators.
If the print flag is FALSE, the module
exits.

4.5.3.1.9

Perform the Bit-To-Text Translation and Register Printing -

If the variable INTERMOD DEVERR.PRINT FLAG is TRUE the device-level
module
prints
the device registers and performs the required
bit-to-text translation.
This is done by executing a WRITE statement
(to produce column headers) followed by a WRITE GROUP statement.
The
WRITE GROUP statement references the subpacket name specified in the
SUBPACKET statement.
It also uses two variables, REPORT.W G F 1 and
REPORT.W G F 2, as format strings. These variables are initTaiized by
the
INITMl -module and contain the format strings for printing the
register data in either WIDE or NARROW format.
If you need to print
data that does not conform to the formats defined by these variables,
you can define your own format.
You can test the 109ical variable
REPORT.WIDE
to determine whether a WIDE or NARROW report was
requested.

4-30

ERROR LOG CONTROL FILE ARCHITECTURE
If the variable INTERMOD DEVERR.PRINT FLAG is FALSE, the
module exits.

4.5.3.1.10

device-level

Indicate Any Notes that are Required

The Error Logging System can print notes for certain conditions that
need additional explanation.
If you need such notes, you can create a
notes module (See Section 4.5.3.2 for details) and include it in the
library.
You can then request a note by referencing it from the
device-level module.
You request a note by performing a PUT into the NOTE NUMBERS file
specifying the note number in the NOTE NUMBERS.INDEX variable. For
example, the RM02/03 device-level module can optionally generate a
note if certain unused bits in the RMDA register are set. This is
done with with the code:
If the unused bits

5 to 7 are set in the RMDA register.

IF (REGISTER.RMDA [5:3] NE #BB'O')
THEN
Print the note saying that it may cause an invalid
sector address to be recognized resulting in a
possible invalid address error.
PUT NOTE NUMBERS INDEX
END IF ;

When the device-level module exits, the caller tests to see whether
any notes were requested.
If notes were requested, the dispatcher
strips the first character from the device-level module's name and
replaces it with the letter 'N'. For example, the notes module for
ERM23 (the RM02/03 device-level module)
is NRM23.
The dispatcher
calls the notes module, which determines which notes were requested
and prints them.
Multiple notes can be
requested.

4.5.3.1.11

requested.

They

are

printed

the

order

Exit the module -

When everything is done, the device-level module exits.
Exiting a
module implies a RETURN to the module's caller. Exiting from a
device-level module also breaks the coroutine relationship.

4.5.3.2 How to Write a Notes Module - This section explains the
structure of a notes module using the RM02/03 notes module as an
example. Section 4.6.4 contains an annotated listing of this module.

4-31

ERROR LOG CONTROL FILE ARCHITECTURE
Here, in general, is the flow of a notes module:

MODULE statement followed by module header

PROCEDURE statement

Notes heading

Selection of a note for printing

(it

Handling of an unknown note number

Getting the next note

Exit the module

These sections are now explained in detail.

4.5.3.2.1

MODULE Statement -

0
The MODULE statement for a notes module must be of the form:
MODULE module_name 'ident' ;
The module name of a notes module is related to its corresponding
device-level module name by replacing the first letter of the
device-level module's name with the letter 'N' to get the notes module
name. This conv.ention must be followed, because the notes module name
is derived from the name of the device-level module and is never
looked up in a table.
See Section 4.5.3.1 for an explanation of the
MODULE statement.

4.5.3.2.2

'ident'

field

the

PROCEDURE Statement -

The PROCEDURE statement for a notes module must be
form:

the

following

PROCEDURE NOTES
The procedure name must be NOTES.
This is
wired
DSP2Ml/DSP2Pl and DSP3Ml/DSP3Pl dispatcher modules.

4.5.3.2.3

into

the

Notes Heading -

The notes heading declares what is about to be printed.
Notice that
notes appear directly following the register interpretation in FULL
and REGISTER reports only.

4-32

ERROR LOG CONTROL FILE ARCHITECTURE
4.5.3.2.4

Selecting a Note for Printing -

Notes are selected for printing by testing the NOTE NUMBERS file for
context after performing a POINTER NOTE NUMBERS FlRST operation.
If
records remain (that is, if there is context) a SELECT is performed on
the variable NOTE NUMBERS.INDEX.
This variable indicates which note
to print.

4.5.3.2.5

Handling an Unknown Note Number -

The ELSE clause of the SELECT statement traps unknown note numbers.
A
SIGNAL is performed using the 'UNKNWNNOT' error indication. The note
number and the drive type are passed to the error handler as string
arguments.

4.5.3.2.6

Getting the Next Note -

The next note is obtained by POINTER NOTE NUMBERS NEXT.
This causes
RPT to point to the next record in the NOTE NUMBERS file.
If another
record exists, the NOTE NUMBERS file has context at the top of the
WHILE ••• DO loop; otherwise there will be no context, which means that
there will be no more notes.

4.5.3.2.7

Exit the Module -

When everything is done, the notes module
implies a RETURN to the module's caller.

exits.

Exiting

module

4.5.3.3 MASSBUS and Non-MASSBOS Considerations - All device-level
modules work essentially the same way.
The only exception is that
MASSBUS modules are not required
to
fill
in
the
variable
INTERMOD_DEVERR.DRIVE_TYPE, whereas non-MASSBUS modules are.
This exception has to do with mixed MASSBUS configurations.
With
mixed configurations, the Executive's database may not match the
actual configurations.
A mismatch can happen if unit plugs have been
inadvertently swapped.
The Error Logging System deals with this possibility as follows:
1.

When a device's mnemonic is found in the DEVICE INFO table in
module DEVSMl, the MASSBUS FLAG is checked.
If-it is TRUE, a
lookahead into the device registers returns the device's
DRIVE TYPE.

The DEVICE INFO table is then searched again to find a record
having that drive type.

4-33

ERROR LOG CONTROL FILE ARCHITECTURE

The Error Logging System then dispatches to the module
corresponding to the actual registers logged, not to the
module indicated by the mnemonic provided by the Executive.

For MASSBUS devices, the Error Logging System uses the device name
provided by the DEVICE INFO table.
This name will always be correct,
as each MASSBUS device has a unique drive-type value.
If there is a
mismatch between the mnemonic supplied and the device type as
determined by examining the registers, the device-type field
in the
printed report is preceded by an asterisk.
For
non-MASSBUS
devices,
it
is
the
device-level
module's
responsibility to supply correct drive-type information.
The DEVSMl
module fills in the value based on the device's mnemonic and size, but
sometimes this information is not accurate. The RK03 and RK05 are
examples of where this is necessary. Both RK03 and RK05 device errors
are processed by the ERK05 module.
The ERK05 module figures out,
based on the device registers, which kind of drive it is and fills in
the DRIVE TYPE variable accordingly.
Another example is DU devices.
In this case, the Error Logging System is only concerned that the
device mnemonic is DU.
It is up to the modules that handle these
devices to provide the drive-type information.

4.5.3.4 Making the New Device-Level Module Known - The Error Logging
System is made aware of a new device-level module by adding a record
to the DEVICE INFO table in the DEVSMl module. A section of the table
is reproduced-in Table 4-3.
Table 4-3
The DEVICE INFO Table
TABLE DEVICE INFO
MNEMONIC
PRINT NAME
ALT PRINT NAME
DISP NAME
SIZEMASSBUS FLAG
DRIVE TYPE
BEGIN TABLE

: ASCII [2] ;
: ASCII [6] ;
: ASCII [12) ;
: ASCII [6j ;
:LONGWORD ;
: LOGICAL
:BYTE

'CT', 'TU60',

'TU60',

'DB', 'RP04',
I DB I '
'RPOS',
'RP06',
I DB''

'RP04/05',
'RP04/05 1 ,
'RP06',

DD'' 'TU58',

'DF', 'RFll',

Device mnemonic
Name for printing
Alternate name for printing
Name of aevice moauie
Size of device
True i f a MASSBUS device
MASSBUS device type number
'ETAll',

#LD'O',

FALSE, #BO'O'

ERP456 I '
ERP456 I '
I ERP456 I '

#LD'l71798',
#LD'l71798',
#LD'340670',

TRUE,
TRUE,
TRUE,

'TU58',

'ETU58',

#LD'512',

FALSE, #BO'O'

'RFll',

'ERSll',

#LD'-1',

FALSE, #BO'O'

'RK03/05',

'ERKOS',

#LD' 4800 I '

FALSE, :jtBO'O'

DL'' 'RLOl',
DL'' 'RL02',

'RLOl',
'RL02',

'ERL12',
'ERL12',

#LD'l0240',
tLD I 20480 I '

FALSE, #BO'O'
FALSE, tBO'O'

'DM', 'RK06',
DM'' 'RK07',

'RK06',
'RK07',

'ERK67',
'ERK67',

#LD'27126',
#LD'53790',

FALSE, #BO'O'
FALSE, #BO'O'

I
I

DK''

RKOS I '

;

:fl:BO' 20'
#B0 1 21 1
#B0'22'

(continued on next page)

4-34

ERROR LOG CONTROL FILE ARCHITECTURE

Table 4-3 (Cont.)
The DEVICE INFO Table
'DP' I

'RP03' I

RP03 I I

'ERP23',

#LD'80000',

FALSE, #BO'O'

DR I I
DR' I
I DR' I
'DR' I
I DR I I

RM02 I I
'RM03',
'RMOS' I
'RM80',
I RP07 I I

'RM02/03',
'RM02/03',
'RMOS',
'RM80',
'RP07',

'ERM23' I
'ERM23',
'ERMOS',
'ERM80',
'ERP07' I

#LD'l31680' I
#LD'l31680' I
#LD'500384',
#LD'242606' I
#LD'l008000',

TRUE,
TRUE,
TRUE,
TRUE,
TRUE,

#B0'25'
#B0'24'
#B0'27'
#B0'26'
#B0'42'

IDS I I
IDS',
'DS' I
IDS I,

I RS03 1 ,
'RS03',
'RS04',
I RS04 I ,

'RS03/04' I
'RS03/04',
'RS03/04',
'RS03/04',

'ERS34',
I ERS34 I ,
I ERS34 I ,
'ERS34 1 ,

#LD'l024',
#LD'l024',
#LD'2048',
#LD 1 2048',

TRUE,
TRUE,
TRUE,
TRUE,

#BO'O'
#BO' l'
#BO' 2'
#B0'3'

'TU56',

'ETCll',

#LD'576',

FALSE, #BO'O'

I
I

'DT', 'TU56',

The columns of the table, taken from left to right, correspond to the
declared items MNEMONIC, PRINT N&~E, ALT PRINT N&M.E, DISP NAME, SIZE,
MASSBUS FLAG, and DRIVE TYPE. Following are explanations of each of
these declared items. MNEMONIC
The mnemonic is a two-character ASCII field that is the device
mnemonic, as found
in the Device Control Block (DCBj • Records
should be kept in alphabetical order by mnemonic.
PRINT NAME
This six-character ASCII field identifies the particular device.
This field
is used in the printing of the Device Identification
Information section of FULL or REGISTER reports whenever the
device registers are available.
In general, this field is used
unless devices are being mounted or dismounted.
In those cases,
the device registers are not available and, depending on the
device, there may be insufficient information to completely
identify a device. When this occurs, the ALT PRINT NAME field is
used instead.
ALT PRINT NAME
This twelve-character ASCII field identifies the device when the
device registers are not available,
usually for mounts and
dismounts.
In these cases, depending on the device, there may be
insufficient information to identify a device completely.
For
example, when an RP04 is mounted, the only information available
that can identify the device is the mnemonic DB and the device
size.
This information is the same for an RP04 and an RP05.
In
this case, the ALT PRINT NAME field is used, which identifies the
device as an RP04/05.
DISP NAME
This six-character ASCII
field
identifies the name of the
device-level module used to process error-logging entries for the
particular device.

4-35

ERROR LOG CONTROL FILE ARCHITECTURE
SIZE
This longword specifies the number of blocks on the device.
There are two special values associated with this field:
a value
of zero (0) indicates that the device is a magtape, and a value
of -1 indicates there is no fixed size for the device.
DEVSMl
will not
correctly
handle
combinations
of
fixedand
variable-size devices having the same mnemonic.
MASSBUS FLAG
This logical value indicates whether or not the device is a
MASSBUS device.
Set it TRUE for MASSBUS devices, and FALSE for
any other devices.
DRIVE TYPE
This byte specifies the MASSBUS drive-type value.
Each MASSBUS
device has a unique value which is available in the low byte of
the drive-type register.
If the record
is not for a MASSBUS
device, this field should be zero (0).
Once the record has been added to the source module
(use SLP so
multiple corrections can be easily merged) the DEVSMl module must be
recompiled.
The first step in this process is to extract the symbol
file for
the DSP2Ml module
(or DSP2Pl for RSX-llM-PLUS) from the
ERRLOGETC.ULB library.
The command should be:
>LBR DSP2Ml.SYM=ERRLOGETC.ULB/EX:DSP2Ml (for RSX-llM)

'"'"YO

> LBR OSP2P;t.SYM=ER~LOGETC.OLB/Ei*:DSPal?l ;(;for 1~S~;::11,M-;~LUS,)
Once this is done, DEVSMl can
sequence is:

recompiled.

The

RSX-llM

command

new

variables

> CFL
CFL>DEVSMl,DEVSMl,DEVSMl=DEVSMl,DSP2Ml
Option>LITERAL SUPPORT.RSX llM =TRUE
Ont-inn"!. r

-i..----··

T m'C'n 11 T'

.u.1.1...wr.nu

c-nnnl'\r>11'

nc<v-, i u

....n.;.i:-.i:-v•''-•L'....,.A

.i...i..c-..

nr nc<

.i:-.uv ....

Option>LIT~RAL SJPPCRT.IO iCTI~ITY

Option>/

t:'l 11 r

c<t:'

•.n.u...,.ci

= FALSE

CFL> "z

The :RSX-·llM-PLUS. COqU.nar;ld sequeO:ce. is:.·

There is no need to recompile the DEVUDA module as
are created in this process.

The updated DEVSMl module can be replaced in the control file
with the command:

library

>LBR ERRLOG.ULB/RP=DEVSMl.ICF
Once this is done, the Error Logging System will be able to associate
the mnemonic of the device with a module used to process entries for
that device.

4-36

ERROR LOG CONTROL FILE ARCHITECTURE
At this point you should include the device-level module
(and notes
module,
if required) in the error log library. This is done by using
the command:
>LBR ERRLOG.ULB/IN=device_level module.ICF[,notes module.ICF]
The EUNKWN module is used (with a warning message) if an attempt is
made to process an error log entry for a device that is listed in the
DEVICE INFO table and whose corresponding device-level module is
unavailable.

4=6

CODE EXAMPLES

The following sections consist of examples of source code from the
Error Logging System. These examples are annotated for use with the
preceding narrative text.
They are written in the Control File
Language, which is documented in the next chapter. The examples in
this chapter are:
•

ERM23 device-level module for RM02s and RM03s

•

DSP2Ml dispatcher module for RSX-llM

:~£;~';:;~-~~:~i~,~:/~;~,~~~,~~b,~~:::;::~~~#:~'.~i},~<l#J~~~,~:Cf~,~i:PJ:i~s'' _
•

4.6.1

NRM23 notes module for RM02s and RM03s

RM02/03 Device-Level Module ERM23

Following is an annotated listing of ERM23.MAC,
module for the RM02 and RM03 disk drives.

the

device-level

0
MODULE ERM23 'MOl.01'

;

ERROR LOG CONTROL FILE MODULE: RM02, RM03
COPYRIGHT (c) 1981 BY
DIGITAL EQUIPMENT CORPORATION, MAYNARD
MASSACHUSETTS. ALL RIGHTS RESERVED.
THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED
AND COPIED ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE
AND WITH THE INCLUSION OF THE ABOVE COPYRIGHT NOTICE. THIS
SOFTWARE OR ANY OTHER COPIES THEREOF, MAY NOT BE PROVIDED OR
OTHERWISE MADE AVAILABLE TO ANY OTHER PERSON.
NO TITLE TO AND
OWNERSHIP OF THE SOFTWARE IS HEREBY TRANSFERED.
THE INFORMATION IN THIS DOCUMENT IS SUBJECT TO CHANGE WITHOUT
NOTICE AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY
EQUIPMENT CORPORATION.
DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF
ITS SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY DIGITAL.
VERSION 01.01
ROBERT E. LI

08-JAN-81

4-37

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

This is one of the many device modules, which is called by the device
error dispatcher (DSP2cl) or device information dispatcher (DSP3cl)
to process all the device dependent information.
Modified by:
CBP

Correct BAE/CS3 register logic

PROCEDURE DEVICE ENTRY
This procedure, which is called via COROUTINE statement from a dispatch
module, declares and translates all ·device registers or data fields of
the data subpacket. The intermodule variables required by the dispatch
modules are stuffed with the appropriate values, followed by a COROUTINE
back to the dispatch module. The dispatch module then COROUTINES back to
this routine a second time, at a point where a write group is used to
print the details of a FULL or REGISTER report.
BEGIN
! Declare a variable to hold the length of the subpacket.

DECLARE PACKET LENGTH ;
TEMP
-:BYTE
END_DECLARE ;
!
!

Now get the lengt:n of t:ne UATH. subpacket. Remembe.L L11dt.
is in bytes and includes two bytes for the length word.

SET PACKET_LENGTH.TEMP TO %LOK_$LENGTH(DISP.NEXT_PACKET)

U-1t:

rt:t.urn~a

valu~

;

Define the data subpacket offsets and all the print information.

SUBPACKET REGISTER

DISP.NEXT PACKET NAMED

RMCSl:

WORD MACHINE
BIT [15]:
RMCSl TRE: BIT (14]:
BIT [13]:
BIT [12]:
BIT [11]:
BIT [10]:
RMCSl BA:

'*Special Condition set' ;
'*Transfer Error' ;
'*MASSBUS Control Bus Parity
'*Unused bit set' ;
' Drive Available',
'*Drive not Available (other
' Unibus B Selected for Data
' Unibus A Selected for Data

port using it)'
Transfer',
Transfer' ;

FIELD [ 8: 2] :

' BA17,BA16 = '
%CNV $BINARY(REGISTER.RMCS1 BA, 2, '0')

BIT

7]:

BIT

6]:

Controller Ready',
' Controller not Ready'
' Interrupt Enabled',
' Interrupt not Enabled' ;
' Function = '
I INTERMOD_DEVERR.DEV FUNCTION
' *Go bi t on ' ;

RMCSl FN:

Err'

FIELD [1:5]:
BIT [ 0]:

(B)'

4-38

;

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module
RMWC:

RM.BA:
RMDA:

WORD MACHINE ;
FIELD [O: 16]:
WORD MACHINE ;
FIELD [0:16]:

RMDA HD:

WORD MACHINE ;
FIELD [13:3]:
FIELD [8:5]:

RMDA SEC:

F IE LD [ 5 : 3 ] :
FIELD [0:5]:

RMCS2:
RMCS2 WC:

WORD MACHINE
BIT [ 15] :
BIT [ 14]:
BIT [13]:
BIT [ 12] :
BIT [11]:
BIT [ 101 :
BIT [ 9]:
BIT [ 81 :
'DTm
OJ...I.

"71.
I

J •

BIT

6]:

BIT

51 :

BIT

4] :

RMCS2 UN:

BIT [ 3]:
FIELD [0:3]:

RMDS:

WORD MACHINE

RMDS ERR:

BIT
BIT
BIT
BIT

[14]:

BIT
BIT
BIT
BIT

[ 10] :
[ 9]:
[ 8]:
[ 7]:

BIT [15]:
[13]:

[12]:
[11]:

BIT
6]:
FIELD [ 1 : 5] :
BIT [ 0]:
RMERl:
WORD MACHINE
RMERl DCK: BIT [15]:
BIT [14]:
BIT [13]:
BIT [12]:
BIT [11]:

BIT
BIT
BIT
BIT

[10]:
[ 9]:
[ 8]:
[ 7] :

%CNV $DECIMAL P(%COM $NEGATE(REGISTER.RMWC), 6)
I I words remaining'-;
' Bus Address Register'
'*Unused bits set', NULL ;
' Track Address = '
I %CNV_$DECIMAL_P(REGISTER.RMDA_HD, 2)
'*Unused bits set (see note)', NULL
' Sector Address = '
%CNV_$DECIMAL_P(REGISTEReRMDA_SEC; 2)
'*Data Late' ;
'*Write Check Error'
'*Parity Error';
'*Nonexistent Drive'
'*Nonexistent Memory'
'*Program Error' ;
'*Missed Transfer' ;
'*MASSBUS Data Bus Parity Error' ;
' Output Ready (silo contains data)',
' Output not Ready (silo empty)'
' Input Ready (silo not full)',
' Input not Ready (silo full)' ;
' Controller Clear '
J ' ( c 1 ears a 11 d r i v es as we 11 ) ' ;
*Parity Test set (even parity)',
' Parity Test reset (odd parity)'
'*Bus Address Increment Inhibit'
' Drive Selected = '
INTERMOD DEVERR.PHYS UNIT
' Attention Active' ;
'*Error (RMERl,2 have bits set)' ;
' Position in Progress' ;
' Medium Online', '*Medium not Online'
' Drive is Write Locked',
' Drive is Write Enabled' ;
' Last Sector Transfered (last of the pack)'
' Programmable {ports program selectable)'
' Drive Present' , '*Drive not Present' ;
' Drive Ready',
' Drive not Ready' ;
' Volume Valid', '*Volume not Valid'
'*Unused bits set', NULL ;
' Drive in Offset Mode',
' Drive not in Offset Mode'
'*Data Check' ;
'*Drive Unsafe' ;
'*Operation Incomplete'
=*Drive Timing Error; ;
'*Write Lock Error' ;
'*Invalid Address Error'
'*Address Overflow Error'
'*Header CRC Error' ;
'*Header Compare Error' ;

4-39

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module
RMERl ECH: BIT
BIT
BIT
BIT
BIT
BIT
BIT
RMAS:

RMLA ANG:
RMDB:
RMMRl:

4] :

3]:
2]:
1] :
0] :

WORD MACHINE
FIELD [8: 8]:
BIT
BIT
BIT
BIT
BIT
BIT
BIT
BIT

RMLA:

6] :

5]:

7]:
6]:
5]:
4]:
3]:
2]:
l]:
0]:

WORD MACHINE
FIELD [ 11: 5] :
FIELD [ 6: 5] :
FIELD [0:6]:
WORD MACHINE
FIELD [0:16]:
WORD MACHINE
BIT [15]:

BIT [14]:

BIT [13):

'*ECC Hard Error' ;
'*Write Clock Fail'
'*Format Error' ;
'*Parity Error' ;
'*Register Modification Refused'
'*Illegal Register' ;
'*Illegal Function' ;
%CND $IF(REGISTER.RMDT [11],
NULL,'*Unused bits set'),
NULL ;
' Unit #7 Attention'
' Unit #6 Attention'
' Unit #5 Attention'
' Unit #4 Attention'
' Unit #3 Attention'
' Unit #2 Attention'
' Unit #1 Attention'
' Unit #0 Attention'
'*Unused bits set', NULL ;
' Sector Count = '
j %CNV_$DECIMAL_P(REGISTER.RMLA_ANG, 2)
*Unused bits set', NULL;
' Data Buffer contents'
%CND $IF(REGISTER.RMMR1 MM,
T Debug Clock set',-NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Debug Clock resetT, NULL)
%CND $IF(REGISTER.RMMR1 MM,
I
Debug Clock Enabled', NULL),
%CND $IF(REGIS~F.R_RMMR1 MM!
I
Debug Clock Disabled', NULL)
%CND $IF(REGISTER.RMMR1 MM,
1 Diagnostic End of-Block set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T

BIT [12]:

BIT [11]:

BIT [10]:

BIT [ 9] :

BIT [ 8]:

BIT [ 7]:

Diagnostic End of-Block reset' ;NULL)

%CND $IF(REGISTER.RMMR1 MM,
T Search Time Out dTsabled', NULL),
%CND $IF(REGISTER.RMMR1 MM,
1
Search Time Out enabled', NULL)
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Clock-set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Clock-reset', NULL)
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Read Data set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
' MaTntenance Read Data-reset' ,NULL)
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Unit Ready', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Unit Not Ready', NULL)
%CND $IF(REGISTER.RMMR1 MM,
1 Maintenance On Cylinder', NULL),
%CND $IF(REGISTER.RMMR1 MM,
1 Maintenance not On Cylinder' ,NULL)
%CND $IF(REGISTER.RMMR1 MM,
1
*Maintenance Seek Error', NULL) ;

4-40

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module
BIT

6]:

BIT

5]:

BIT
BIT

4]:
3]:

BIT [ 2]:

BIT [ l]:

RMMRl MM:

BIT [ 0]:

RMDT:

WORD MACHINE
BIT [15]:
BIT [14]:
BIT [13]:
BIT [12]:
BIT [11]:

RMDT TYP:
RMSN:

RMOF:

FIELD [9: 2]:
FIELD [0:8]:
WORD MACHINE ;
FIELD [0:16]:
WORD MACHINE ;
FIELD [13:3]:
BIT [12]:
BIT [11]:
BIT [10]:
FIELD [8:2]:
BIT [ 7]:
FIELD [0:7]:

RMDC:
RMDC DC:
RMHR:
RMMR2:

WORD MACHINE
FIELD [10:6]:
FIELD (0:10]:
WORD MACHINE ;
FIELD [0:16]:
WORD MACHINE
BIT [ 15] :
BIT [ 14] :
BIT [13]:
BIT [ 12]:
BIT [11]:
BIT [10]:

%CND $IF(REGISTER.RMMR1 MM,
T*Maintenance Drive-Fault 1 ,NULL)
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Sector Pulse set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Sector Pulse reset', NULL)
'*Unused bit set' ;
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Write-Protect', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Write-Enabled', NULl)
;
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Index-Pulse set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
,-Maintenance Index Pulse reset', NULL)
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Sector Compare set', NULL),
%CND $IF(REGISTER.RMMR1 MM,
T Maintenance Sector Compare reset', NULL);
' Diagnostic Mode on',
' Diagnostic Mode off' ;
'*Drive not Sector Addressable' ;
'*Unit is a Tape Drive' ;
NULL, '*Unit is not a Moving Head Device'
'*Unused bit set' ;
' DRQ on (dual port unit)',
' DRQ off (single port unit)'
'*Unused bits set', NULL ;
' Drive Type = '
INTERMOD DEVERR.DRIVE TYPE
' Drive Serial Number = '
%CNV_$BCD(REGISTER.RMSN,4)

(BCD) I

'*Unused bits set', NULL
' 16 Bit Data Format',
'*18 Bit Data Format' ;
' ECC Inhibit', 1 ECC enabled'
' Header Compare Inhibit',
' Header Compare Enabled' ;
'*Unused bits set', NULL ;
' Offset Direction = Forward',
' Offset Direction = Reverse'
'*Unused bits set', NULL;
'*Unused bits set', NULL ;
' Desired Cylinder = '
%CNV_$DECIMAL_P(REGISTER.RMDC_DC, 4)
' Holding Register contents'

;

' Port A Reques~ for Service'
' Port B Request for Service'
' Control Select Tag on' ;
%CND $IF(REGISTER.RMMR1 MM,
' Test Sequencer BranchTng on', NULL)
' Control or Cylinder Tag on'
' Control or Head Tag on' ;

4-41

ERROR LOG CONTROL FILE ARCHITECTURE

RMMR2 MBL: FIELD [0:10]:

%CND $IF(REGISTER.RMMR1 MM,
' Maintenance Bus Lines-= '
%CNV $BINARY(REGISTER.RMMR2 MBL, 10, '0')
'

(Bf'

, NULL)
RMER2:

RMECl:
RMECl PS:
RMEC2:

;

WORD MACHINE
BIT [15]:
BIT [14]:
BIT [13]:
BIT [12]:
BIT [11]:
BIT [10]:
FIELD [8:2]:
BIT [ 7]:
FIELD [4: 3]:
BIT [ 3]:
FIELD [0:3]:

'*Bad Sector Detected (hdr bit)' ;
'*Seek Incomplete' ;
'*Operator Plug Error (removed)' ;
'*Invalid Command (VV bit reset)' ;
'*Loss of System Clock' ;
'*Loss of Bit Clock' ;
'*Unused bits set', NULL
'*Device Check' ;
'*Unused bits set', NULL
'*Data Parity Error' ;
'*Unused bits set', NULL

WORD MACHINE
FIELD [13:3]:
FIELD [0:13]:

'*Unused bits set', NULL ;
' ECC Position = ' I VAR.ECCPS

WORD MACHINE ;
FIELD [11:5]:
FIELD [ 0: 11] :

'*Unused bits set', NULL ;
' ECC Pattern
' I VAR.ECCPAT

IF DEVICE OP.FLG BAE AND {PACKET_LENGTH.TEMP EQ tBD'46')
If the RH70 flag
is true and the packet length is 22 registers,
declare the BAE and CS3 registers. Note that the packet length check
is necessary because umapped RSX systems will not log BAE and CS3
even if the controller is an RH70.
THEN
BEGIN
RHBAE:

WORD MACHINE ;
FIELD [6:10]:
RMBAE EXT: FIELD [0:6]:
RMCS3:

RMCS3 !PC
END ;
END IF
END PACKET

'*Unused bits set', NULL ;
' BA21 through BA16 = '
I %CNV_$BINARY(REGISTER.RMBAE_EXT, 6, '0')

WORD MACHINE
BIT [15]:
BIT [14]:
BIT [13]:
BIT [12]:
BIT [11]:
BIT [10]:
FIELD [7: 3]:
BIT [ 6]:

'*Address Parity Error' ;
'*Data Parity Error, Odd Word'
'*Data Parity Error, Even Word' ;
'*Write Check Error, Odd Word' ;
'*Write Check Error, Even Word'
' Double Word Transfered' ;
'*Unused bits set', NULL
' Interrupt Enabled',
' Interrupt not Enabled'
F !ELD [ 4: 2] :
'*Unused bits set', NULL
FIELD [0:4]:
' Inverse Parity Check Bits = '
%CNV_$BINARY(REGISTER.RMCS3_IPC, 4, '0')

{B) I

;

4-42

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

Declare all variables needed for the subpacket print information.
DECLARE VAR
ECCPS:
ECCPAT:
END DECLARE

ASCII [22]
ASCII [ 22]

ECC position.
ECC pattern.

! Create the device function code conversion table.

TABLE FUNCTION
FUN CODE:
FUN-TEXT:
BEGIN TABLE
4f:BO'OO',
4f:B0'02' I
4f:B0'03' I
4f:B0'04',
4f:B0'05',
4f:B0'06',
4f:B0'07',
4f:BO'l0',
#BO'll',
4f:B0' 14 I I
4f:B0'24' I
4f:B0'25',
4f:B0'30',
4!=B0'31' I
4f:B0'34',
4f:B0'35',
END_TABLE ;

BYTE MACHINE
ASCII

[ 27]

;

'No Operation'
'Seek Command'
'Recalibrate' ;
'Drive Clear' ;
'Release (dual port)'
'Offset Command' ;
'Return to Centerline'
'Read in Preset' ;
'Pack-Acknowledge'
'Search Command' ;
'Write Check Data' ;
'Write Check Header and Data'
'Write Data' ;
'Write Header and Data'
'Read Data' ;
'Read Header and Data' ;

Calculate the ECC Position.
Determine if the ECC position is normal (not used), has an illegal
value, points to the starting bit within the sector or is irrelevent.
IF REGISTER.RMECl PS LE 4f:WD'4128'
THEN
At this point, the ECC position is within range (0. to 4128.).
Next, find out if the ECC position counter (register) was used.
If the ECC position register value equals an octal 4066, it
indicates the register was initialized but not used.
SET VAR.ECCPS TO %CND $IF(REGISTER.RMEC1 PS EQ 4f:W0'4066',
'Normal', %CNV_$DECIMAL_P(REGISTER.RMECl_PS, 6)) ;
ELSE
SET VAR.ECCPS TO 'Outside of legal range'
END IF

4-43

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

If the error was a non-correctable hard error or Error Correction
was inhibited, then the ECC position and ECC pattern are irrelevant.
IF (REGISTER.RMERl ECH EQ TRUE)
THEN
BEGIN
SET VAR.ECCPS TO 'Irrelevant (ECH set)'
SET VAR.ECCPAT TO 'Irrelevant (ECH set)'
END
END IF i
IF (REGISTER.RMOF [11] EQ TRUE)
THEN
BEGIN
SET VAR.ECCPS TO 'Irrelevant (EC! set)'
SET VAR.ECCPAT TO 'Irrelevant (EC! set)'
END
ELSE
SET VAR.ECCPAT TO %CNV_$0CTAL(REGISTER.RMEC2 [0:11]' 4, '0')

(O) I

END IF

The following will use the register information to determine the
value of the intermodule variables, which are needed by the
' dispatcher and stuff these accordingly
The variables are:
INTERMOD DEVERR.DRIVE SN
INTERMOD-DEVERR.DEV FUNCTION
INTERMOD-DEVERR.PHYS UNIT
INTERMOD-DEVERR.ERROR CYLINDER
INTERMOD-DEVERR.ERROR-SECTOR
!NTERMOD-DEVERR.ERROR-HEAD
INTERMOD-DEVERR.ERROR-GROUP
INTERMOD-DEVERR.BLOCK-NUMBER
INTERMOD-DEVERR.ERROR-TYPE

(not applicable to this device)

Return the drive serial number.
SET INTERMOD DEVERR.DRIVE SN TO %CNV_$BCD(REGISTER.RMSN, 12, '
! Lookup the function code in the function table.

4-44

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

FIND FUNCTION FUN_CODE = REGISTER.RMCSl_FN ;
!
! Check if a match is found between the register and the table,
IF FUNCTION.CONTEXT
THEN
Yes, return the associated function text in the variable.
SET INTERMOD DEVERR.DEV FUNCTION TO FUNCTION.FUN TEXT
ELSE
Otherwise, return text indicating an invalid function.
SET INTERMOD DEVERR.DEV FUNCTION TO 'Invalid function'

;

END IF ;
!

Return the physical unit number.

SET INTERMOD DEVERR.PHYS UNIT TO %CNV_$DECIMAL(REGISTER.RMCS2_UN, 1)
DISK GEOMETRY INFORMATION.
Calculate the intermodule variables for LBN, GROUP, CYLINDER, TRACK,
and SECTOR address, initially assumrning the error packet was NOT caused
by a data error.
Calculate LBN using the formula •••
LBN

= ( CYLINDER ADRS * number of SECTORS/CYL +
HEAD ADRS * number of SECTORS/TRACK +
SECTOR_ADRS }

SET INTERMOD DEVERR.BLOCK NUMBER TO
%CNV-$DECIMAL P(
(REGISTER.RMDC DC* #LD'l60' +
REGISTER.RMDA-HD * #WD'32'
+
REGISTER.RMDA-SEC } ,
9)
Initialize GROUP.

(not applicable to this device)

SET INTERMOD DEVERR.ERROR GROUP TO 'N/A'
!

Initialize CYLINDER.

SET INTERMOD DEVERR.ERROR CYLINDER TO
%CNV_$DECIMAL_P(REGISTER.RMDC_DC, 3)
!

Initialize TRACK (head).

SET INTERMOD DEVERR.ERROR HEAD TO
%CNV_$DECIMAL_P(REGISTER.RMDA_HD, 2)
Initialize SECTOR.
SET INTERMOD DEVERR.ERROR SECTOR TO
%CNV_$DECIMAL_P(REGISTER.RMDA_SEC, 2}

4-45

;

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

Correct the geometry information if necessary.
Upon a data error, the hardware will update the GROUP, CYLINDER, TRACK and
SECTOR to point to the sector following the sector in error.
In order to
make the intermodule variables for GROUP, CYLINDER, TRACK, SECTOR and LBN
point to the media address causing a data error, they are corrected (backed
off by 1) using the following algorithm.
Was it a data error ?
Yes, it was a data error.
Decrement LBN.
Was SECTOR = 0 ?
Yes, SECTOR = 0.
SECTOR = SECTORMAX.
Was TRACK = O?
Yes, TRACK = o.
TRACK = TRACKMAX.
Decrement CYLINDER.
No, TRACK NOT = 0.
Decrement TRACK.
No, SECTOR NOT = 0.
Decrement SECTOR.
No, it was not a data error.

(check error bits)
(correction (backoff) is needed)
(recalculate pointing to previous BLK)
(sector underflow boundry?}
(underflow sector and borrow from TRK)
(underflow the sector}
(track underflow boundry?}
(underflow TRK and borrow from CYL)
(underflow the track}
(borrow from CYL for TRK}
(no undeflow of TRK}
(Simply, with no borrow from CYL}
(no underflow at all)
(point to the previous block)
(no correction (backoff) needed)

Was it a data error?
IF REGISTER.RMERl DCK OR REGISTER.RMERl ECH OR REGISTER.RMCS2 WC
THEN
! Yes,

it was a data error.

(LBN and geometry information needs correction)

BEGIN
Correct the LBN by recalculating (backed off by one block} •
!

SET !NTERMOD DEVERR.BLOCK NUMBER TO
%CNV-$DECIMAL P(
(REGISTER.RMDC DC * #LD'l60' +
REGISTER.RMDA-HD * #WD'32'
+
REGISTER.RMDA-SEC ) -1,
n'
:7 I

•

Was the sector address zero? (Sector underflow?)
IF REGISTER.RMDA SEC EQ #BD'OO'
THEN
! Yes, it was zero.

(so undeflow the sector and borrow from track)

BEGIN
! Underflow the sector.

SET INTERMOD DEVERR.ERROR SECTOR TO '31.'
! Was track

;

(head) address zero? (track underflow?)

IF REGISTER.RMDA HD EQ #BD'OO'
THEN
! Yes, the track was 0, so underflow the track
! and borrow from the cylinder.

4-46

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

BEGIN
Underflow the track (head) •
SET INTERMOD DEVERR.ERROR HEAD TO '4.'
1 Borrow from the cylinder.
1

SET INTERMOD DEVERR.ERROR CYLINDER TO
%CNV_$DECIMAL_P(REGISTER.RMDC_DC - 1, 3)
END
ELSE
1 No, the track was not zero. Simply decrement it.

(no track underflow)

SET INTERMOD DEVERR.ERROR HEAD TO
%CNV_$DECIMAL=P(REGISTER.RMDA_HD - 1, 2)
END IF
END ;
ELSE

;

No, the sector address was not zero. Simply decrement it.
(no sector underflow)
SET INTERMOD DEVERR.ERROR SECTOR TO
%CNV_$DECIMAL_P(REGISTER.RMDA_SEC - 1, 2)
END IF
END
END IF

Find the reason causing this error packet and set the variable
accordingly.
!

IF REGISTER.RMCSl TRE
THEN
BEGIN
IF NOT REGISTER.RMDS ERR
THEN
DECODE
INTERMOD DEVERR.ERROR TYPE = REGISTER ;
RMCS2 [15] ;
Data Late
RMCS2 [ 14] ;
Write Check Error
RMCS2 [13]
U.B. Parity Error
RMCS2 [12]
Nonexistent Drive
RMCS2 [11]
Nonexistent Memory
RMCS2 [10]
Program Error
RMCS2 [ 9]
Missed Transfer
RMCS2 [ 8]
MASSBUS Data Bus Parity Error
END DECODE

4-47

ERROR LOG CONTROL FILE ARCHITECTURE
Di:-uir::e:>-Le:>ui?1_ "'10d1Jl~

ELSE
DECODE
INTERMOD DEVERR.ERROR TYPE = REGISTER ;
RMER2 [15) ;
Bad Sector Detected (Hdr bit)
RMER2 (14)
Seek Incomplete
RMER2 [13)
Operator Plug Error (removed)
RMER2 (12]
Invalid Command (VV bit reset)
RMER2 [11]
Lost of System Clock
RMER2 [ 10]
Lost of Bit Clock
RMER2 [ 7]
' Device Check
RMER2 [ 3]
Data Parity Error
RMERl [ 6]
ECC Hard Error
RMERl (15]
Data Check
RMERl [14]
Drive Unsafe
RMERl [13]
Operation Incomplete
RMERl (12]
Drive Timing Error
RMERl [11]
Write Lock Error
RMERl (10]
Invalid Address Error
RMERl [ 9]
Address Overflow Error
RMERl [ 8]
Header CRC Error
RMERl [ 7)
Header Compare Error
RMERl [ 5]
Write Clock Fail
RMERl [ 4]
Format Error
RMERl [ 3)
Parity Error
RMERl ( 2]
Register Modification Refused
RMERl [ l]
Illegal Register
RMERl [ O]
Illegal Function
END DECODE
END IF
END ;
ELSE
DECODE
INTERMOD DEVERR.ERROR TYPE = REGISTER ;
NOT RMDS
[12]
Medium not Online
NOT RMDS
[ 8]
Drive not Present
NOT RMDS
[ 6]
Volume not Valid
RMCSl [13j
MASSBUS Control Bus Parity
Drive not Available
NOT RMCSl [ 11]
Controller not Ready
NOT RMCSl [ 7]
END DECODE ;
END IF ;

~rror

IF (INTERMOD DEVERR.ERROR TYPE EQ NULL)
THEN
SET INTERMOD DEVERR.ERROR TYPE TO 'No error bit found'
END IF

All the intermodule variables have been stuffed,
coroutine caller (calling dispatch module).

so return to the

COROUTINE ;
The dispatcher
returns control to this module here, with the flag
INTERMOD DEVERR.PRINT FLAG set to either TRUE or FALSE.
If the
flag is TRUE, a FULL or REGISTER report is in progress, the banner
has been printed, and this module prints device registers (or data
fields for packet oriented devices)
Otherwise, this module does
not print anything, and simply exits back to the dispatcher.
The
width of the report
(80/132)
is controlled by dispatcher defined
format variables REPORT.W G F 1 and REPORT.W G F 2 based on the
user specified /WIDTH switch~

4-48

ERROR LOG CONTROL FILE ARCHITECTURE
Device-Level Module

0
IF INTERMOD DEVERR.PRINT FLAG
THEN
BEGIN
! Print the header for the Name, Value and Interpretation fields.
WRITE
FORMAT
'!5FCName!l3FCValue!25FCinterpretation!2FL'
Print the registers according to the format variable (80/132)
provided by the dispatcher.
WRITE GROUP REGISTER
FORMAT
Print format for the register name
and it's associated value.
REPORT.W_G_F_l,
! Print format for the exploded bits and fields.
REPORT.W G F 2
If there are any NOTES to be printed, this is where the
PUT of note indicies is done on the note file.
When the
return from this module is done, the dispatching module
examines the note file to determine if the note module
NRM23
should be called to print the notes specified by
index number.
If the unused bits

5 to 7 are set in the RMDA register.

IF (REGISTER.RMDA [5:3] NE iBB'O')
THEN
Print the note saying that it may cause an invalid
! sector address to be recognized resulting in a
! possible invalid address error.
PUT NOTE NUMBERS INDEX
END IF

END
END IF
END-;

END MODULE

4-49

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM Dispatcher Module
4.6.2

DSP2Ml Dispatcher Module for RSX-llM

Following is an annotated listing of the DSP2Ml dispatcher module
RSX-llM.
MODULE DSP2Ml 'MOl.00'

for

;

ERROR LOG CONTROL FILE MODULE: DSP2Ml
COPYRIGHT (c) 1981 BY
DIGITAL EQUIPMENT CORPORATION, MAYNARD
MASSACHUSETTS.
ALL RIGHTS RESERVED.
THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED
AND COPIED ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE
AND WITH THE INCLUSION· OF THE ABOVE COPYRIGHT NOTICE. THIS
SOFTWARE OR ANY OTHER COPIES THEREOF, MAY NOT BE PROVIDED OR
OTHERWISE MADE AVAILABLE TO ANY OTHER PERSON.
NO TITLE TO AND
OWNERSHIP OF THE SOFTWARE IS HEREBY TRANSFERED.
THE INFORMATION IN THIS DOCUMENT IS SUBJECT TO CHANGE WITHOUT
NOTICE AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL
EQUIPMENT CORPORATION.
DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF
ITS SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY DIGITAL.
VERSION 01. 00
C. PUNTAM

22-SEP-80

This module is called to process Device Error packets.
Module Name:

DSP 2 M 1

Module Prefix: ----------!
Error Code:

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

Operating System: ------------!
Packet Format:

-----------------!

The following Error Subcodes are defined:
Subcode

Mnemonic

E $SDVH
E-$SDVS
E=$STMO

2
3

Meaning
Device Hard Error
Device Soft Error
Device Interrupt Timeout

Define any literals used in this module.
LITERAL DSP2 SUB ANY.FORMAT 1 =
'I/O Operation InformatTon:!lFL'
•--- --------- ------------!2FL'
'!5FCDevice Function!38FCType of Error!2FL'
'!5FC!30DP!38FC!30DP!3FL' ;

4-50

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM Dispatcher Module
LITERAL DSP2 SUB ANY.FORMAT 2

:~~~~~~ ~~~~~ ~~~~~~~~ ~~=~~~~~~~~~i~~~:

'!5FCCylinder!l5FCGroup!22FCHead!28FCSector!36FCBlock!2FL'
I !5FC!8DP!l5FC!5DP!22FC!4DP!28FC!6DP!36FC!lODP!2FL'
PROCEDURE START MOD
BEGIN
! Create the Subcode Conversion table.
TABLE SUBCODE
NUMBER
TEXT

:WORD ;
: ASCII [ 18 j

BEGIN TABLE

i-;

2,
3,
END TABLE

'Device Hard Error'
'Device Soft Error'
'Device Timeout' ;

First check to see if PERIPHERAL errors are selected. If they are not,
simply return. Also determine the packet subtype. If it is a known
subtype code, then proceed. Otherwise it is an error.
IF NOT REPORT.PERIPHERAL
THEN
This type of packet has not been selected for printing.
RETURN
END IF ;
FIND SUBCODE NUMBER = HEADER.CODE SUBTYPE
IF NOT SUBCODE.CONTEXT
THEN
BEGIN
SIGNAL 'ILLPACSBC' PARAMETERS
REPORT.PACKET !DENT,
%CNV $DECIMAL(HEADER.CODE TYPE, 3) I
%CNV=$DECIMAL(HEADER.CODE=SUBTYPE, 3)
RETURN
END ;
END IF

0
Find the device name by calling the DEVICE NAME procedure.
CALL MODULE 'DEVSMl' PROCEDURE 'DEVICE NAME'

Prepare the NOTE_NUMBERS file for any notes that may be requested.
POINTER NOTE NUMBERS CLEAR

4-51

ERROR LOG CONTROL FILE ARCHITECTURE
""'! _ ....... __.__,__ .._

U..t..>:>tJO. '-'-u<;;:;:.1..

Now set up the procedure DEVICE ERROR and the appropriate device level
module as a coroutine pair.
CALL MODULE INTERMOD DEVERR.DISP NAME PROCEDURE 'DEVICE ENTRY'
COROUTINE 'DEVICE ERROR'
END ;
PROCEDURE tiEVICE E~ROR
BE.GIN
!

The following is used to format the output· for the
'Device Hard Error', 'Device Soft Error' and
'Device Interrupt Timeout' Device Error packets.
The DEVICE ID subpacket contains information about the .
device on which the error occured.
The DEVICE OP subpacket contains information about the I/O
Operation In progress on the device at the time of the error.

Obtain information from the coroutine partner.
COROUTINE ;

Assume the serial number test will succeed or be irrelevent.
SET INTERMOD DEVERR.REJECT FLAG TO FALSE ;
! Now test to see if this device passes the drive serial number test.

IF REPORT.DRIVE SN VALID AND
(INTERMOD_DEVERR.DRIVE_SN NE %CNV_$BCD(REPORT.DRIVE_SN, 12))
THEN
Indicate that the test failed.
SET INTERMOD DEVERR.REJECT FLAG TO TRUE
END_IF ;
Determine the type of report and format the output
accordingly.
CASE REPORT.MODE OF
['BRIEF']:
! The BRIEF report is one line long.

BE.GIN
Now ouput the information based on the result of the test.

4-52

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM Dispatcher Module

IF NOT INTERMOD DEVERR.REJECT FLAG
THEN
Now output the brief report.
WRITE
REPORT.PACKET !DENT,
%CNV $RSX TIME(HEADER.TIME STAMP, 0),
SUBCODE.TEXT,
DISP;DEVICE STRING,
INTERMOD DEVERR.ERROR TYPE,
. 'Function = ' I IN.TERMOD_DEVERR.DEV_J?UNCTION
FORMAT
REPORT.BRIEF FORMAT
END_IF ;
! Now go back to the partner. It. wi 11 simply return
! without printing.
!
SET INTERMOD DEVERR.PRINT FLAG TO FALSE
END ;

['FULL', 'REGISTERS']:
The FULL report contains detailed information
! about the error.

BEGIN
Now output the information based on the result of the test.

IF NOT INTERMOD DEVERR.REJECT FLAG
THEN
Output the first page if the report type is 'FULL'.
BEGIN
IF REPORT.MODE EQ 'FULL'
THEN
Now output the information for the standard subpackets.
BEGIN
CALL MODULE REPORT.FULL MOD l PROCEDURE
CALL MODULE REPORT.FULL MOD 2 PROCEDURE 'OUTPUT PACKETS'
CALL MODULE REPORT.FULL MOD 3 PROCEDURE 'OUTPUT PACKETS'
CALL MODULE REPORT.FULL MOD 4 PROCEDURE 'OUTPUT PACKETS'

4-53

ERROR LOG CONTROL FILE ARCHITECTURE
K~X-11M oispatcner Module

Now output the Data subpacket information.
WRITE
INTERMOD DEVERR.DEV FUNCTION,
INTERMOD-DEVERR.ERROR TYPE
FORMAT
DSP2 SUB ANY.FORMAT 1
Now output the Device Error Position information
if it is applicable.
IF INTERMOD DEVERR.ERROR CYLINDER NE '???'
THEN
WRITE
INTERMOD DEVERR.ERROR CYLINDER,
INTERMOD-DEVERR.ERROR-GROUP,
INTERMOD-DEVERR.ERROR-HEAD,
INTERMOD-DEVERR.ERROR-SECTOR,
INTERMOD-DEVERR.BLOCK-NUMBER
FORMAT
DSP2 SUB ANY.FORMAT 2
END IF :
A full report is wanted, so print the record I.D.
and header.
WRITE
REPORT.PACKET !DENT
FORMAT

END

'!lFP!SFCEntry !DP!22FC(continued) !3FL'
'Device Supplied Information:!FL' I
'------ -------- ------------!2FL' ;

ELSE
Only a register dump is requested, therefore print
the banner from the full report.
WRITE
REPORT.PACKET !DENT,
%CNV $DECIMAL-P(HEADER.ERROR SEQ, 8),
DISP:DEVICE STRING,
SUBCODE.TEXT,
%CND $IF((INTERMOD DEVERR.ERROR TYPE NE NULL) 1
T (' I INTERMOD_DEVERR.ERROR_TYPE I ')',
NULL) ,
%CNV_$RSX_TIME(HEADER.TIME_STAMP, 0)

4-54

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM Dispatcher Module
FORMAT
Select the format statement
based on the desired width.
%CND_$IF(REPORT.WIDE,
!

! WIDE is selected.

'!lFP!SFCEntry !7DP!20FCSequence !9DP' I
'!40FC!6DP!48FC!l8DP!DP!2FS!20DP!3FL' I
'Device Supplied Information:!FL' I
'------ -------- ------------!2FL',
! NARROW is selected •
.!

'!lFP!SFGEntry !7DP!20FCSequence !9DP!40FC!6DP!FL'
'!5FC!l8DP!DP!2FS!20DP!3FL' I
'Device Supplied Information:!FL' I
'------ -----~-- --~---------!2FL') ;
END~IF

;

! Now indicate we want to have the device module print.

SET INTERMOD DEVERR.PRINT 'FLAG TO TRUE ;
END
ELSE
The packet was rejected. Don't print anything.
SET INTERMOD DEVERR.PRINT FLAG TO FALSE ;
END IF ;
! Now go back to the partner. It will output the device registers
! if the print flag is true.

END ;

[I NONE'] :

If the report type is NONE, output nothing.

SET INTERMOD DEVERR.PRINT FLAG TO FALSE ;
END CASE

Now COROUTINE back to the partner. It will print if instructed to do so.
COROUTINE ;

4-55

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM Dispatcher Module

Now see i f any notes were requested and print them if there were.
IF NOTE NUMBERS.CONTEXT THEN
BEGIN
SET INTERMOD_DEVERR.DISP_NAME .TO 'N' L
%STR_$REMAINING ( INTERMOD _DEVERR. DISP_NAME, 2)
I.F %PKT_$MODULE ( INTERMOD_DEVERR.DISP_NAME)
THEN
CALL MODULE INTERMOD DEVERR.DISP NAME PROCEDURE 'NOTES'
ELSE
SIGNAL 'NONOTES' PARAMETERS INTERMOD DEVERR.DRIVE TYPE ;

END IF ;
END ;
END MODULE

DSP2Ml.CNF

4-56

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM- PLUS Dispatcher Module
4 .6;3 · · oSP2Pl. 1>1~patchei ·~oa.\il;e for RSX-l:iM-PbOS ·

·Foll·owf~~ ··:i.~ an annotatedu listing ·0£ the nsnPi' dispatcher .module· for
RSX_;llM--PLUS,.
,'MODULE' DSP2P~ •'POL oo·· . i

. ! ERROR L.OG. CQNT.ROL. FfLE MODUL~! DSP2Pl

f . ..

: .. · '·.· . .

·!

\!.
!·
''.(

DIGlTAL EQU:i.PMENT CORP,PRATipN, MAXNARD·
. ~ASSAGHUS}Sf'TS •

·:A.LL R:I;GH;TS· RESERV~D.;,

·.·· ·.·.·.'

.... ·.

THiS· ·. sOFTWARE 1s· ···FqR~.ISH.l!1o'.: 9~PER•·f1.~Ip~~$~. AND.MAY"sE·usEa ·

·''.t~ '~

{" ·; >, ,,,. '' :;;: • ' . ::<' • .: "., •: - : :.':>. :"·'~'·; .:"':<:"';< ~;'::' :·' .:, ·.'.'.' ,:'·'.'. :~,·· .:~

·::oN:I..i. ~N ·. AccoRDANCE wr.Pa ·'THE· TERMS· dF sucH >LICENSE.·
! 'ANO WITH . THE ,INCLU.SIQN :.' QF.: 'THE .isov.e COPYRfG.HT NOTICE;. .THIS·.
·t .· s9F'l'WJUtE ·.· :·02· ·A~Y.<.oirli:ER··.· ·eot>1~$• ·.'THER'EOF,:··MA:Y··. NOT· ·st.·· .~R<W;tQ.ED o~·
: 1:. OTHf!~W:iS:i!!'·~Ap·E.• · .AVAiL.Aij~E.· TO·.·Ab!Y. :.QTHER'·PERS.ON.·, >·NP··· Tl.TLE TO .·AND. ·
::.r.mfNERSHIP·. OF.,"THE.:SOFTWARE>'IS,\M·.&:REB.Y.··TRAN:SFERED:.;:.>' ':. \'
:.:.·i ··AND , ·cOJ?!Ei> ·

',>

• < :: ' '

;uu '."

~'."::',: :'~~

.< .' '.~''. '<,

<: ''

>': ~ ':'·.,' ;,~" '·: •"\'

< • • ·,, ' •• ,' :·: X :,'.', '"; ·';

•

~',, '. • ' ·~: ••

:. ; '•: • '. '. ·' ··'.. ';,

.:!.: .T:H'E :·rN:FOR~~Tlq~ .·· ·+N·.:~11.rn:.;ooco~i~N.'t· is<~uB;J.EC.T''. to .:CH.ANGE. ·WITHOUT:.·

:.·!'.\NQTICE . AND.:SHOULl)'':NOT::'BE·.. ·coN'STROED.'AS A .COMMITMENT'·· •ifY' .lHGITAL
·:· .t EQUIPMEf\JT.·CQRP,ORA'J;'iO~.;, .>:.
.
.
::1\.' . .

·.. · ...· .... ·..

,. •\

"r.·orGn~L asSui'iF!S NO 'RESPONSrs·i'Li~Y· FOR. Ttrn"·cj~E· oR .RELiA:s1't1TY OF
!··.~1.S SOFTWAlU~·QN .EQUIPMENT THAT.JS NO.T SOPPLIEP.BY OIG·ITAt~.

:l 11·~RsiON<oY ~-·oo >' \
.1:
. '!.

.c.•... Po..Nr.'1M
·Thi~ mod·t;tl~ . iscal,led to ;process Devi~e· Erfor packets.

·!
:·J:. ,··

.Modui'e'·l?reflx:::.:.:..:.: __:.:~::..--1

·'f

i ·!

:r·.·.·. ~~:Jt;~;~;::1:-:;:}i:tr=tJi.i.:.·· •.
,.!

•'.

. '!

'!.

.. ·· . ' ~ .. '.. .

.•

·1.
. : pac.k~t ·J"orm·~:t:;::·.\~:".........:.~.,."'."·,~~-~--.:.~-. t.
t
, <-.h'" ··,.-.> '-"<·~·-<.·., * '· ..''.-~:.,
<
'~.·
f Th~ fol lbwing Err Or Subcode~ : are de:ffn.ed.·;·
'

•

;.. 'i . . , < :· •• •

.
--·----'

!
! .

·1

3·

•••• : •

' •• •

• :

•

' ,•

': • • :i '

,..

·.

• ~

' ': .

.< •

·> ., '

< ••

•

..· .. ··

.. .

.:..--~-·~--

. .

. I

•

-~.--;-.~-.-

E.$SPVH

E:-::-$.SDVS· .
E-$STMO

E=$SUNS.
!
4
!
!
· ! Define any literals used in this· module.

.Dev ,i c.e :11a rd .Error,
Device· Soft'··Error·
· 'oevice Inter.rupt Timeout·
Onsolicited Interrupt

LITERAL DSP2 SUB ANY.FORMAT 1

'r/o operation InformatTon: t lFL'
'--- --------- --~---------!2FL'
'!SFCDevice Function!38FCType of
I

!5FC!30DP!38FC!30DP!3FL 1

~rror!2FL'

;

4-57

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM-PLUS Dispatcher Module
tITERAL' DSP2 ~UB ANi.• FO~MAT 2 '
. '.·· .· '.. '
•Device <Error ,po:sitiol'l }nf6rmatt69!TlFL·• · . . ,. ..
I,... _ _. ...... _
- - - -..;.;..:.. -,...-,_._ __ ...,. ', _ _.,'."" _ _ _ _ _...:;_"_ ! ,2FL I
' • '

. · · · ..· ,
..

. .·'
'

. '.

'.·! s:Fccyl ind.er;! 15FCGro~1i!: 22FOHead !'~a:pc~ector ! 36J?CB1oc.k'12FtJ

., !5FC!8DPtl$FCJSDP!22F(!!4DP·!'28FC?6l>P:t36FC!100P.!W~ 1·:. ·,

:.·P1~'(')CEDURE START>MOD' L

... ·:.

.·

•

.1 •

', . :

• •

· .•.

.-·

..J~EGIN

,J.

·! ·Create tl"fe ;suhc.ode conveis~~h: :tabl:er.1
'!, · , ..' ..... : ' ': · .... : .
'VABL'l~·

SUBCOD·E ·;.
..· ·N·UMBER. . .
''."·.TEXT· ,. /
·&'.EGIN ·TABJ,,E.•

. .· C .
":" '1

... .....

·3:::

.·'

' 2' .·

. 4·1

· .. ·· .... ·•··. ·

·.:WORD.';

•. )';

. :·I;.
:'AS.Cir .+1sr: ';'. <·: '\ . '

,' .·

'., ,·, L < " '·.

;··,, • .

>·~:~~vi«~e Ha. ···.r.a.'.:LE.··.'..r1..m
. ·.:.·. r.:.·.·.·. ·
.

' ' oaV''i.ce 'soft: 'Erfrbr I' ' ..; ' '

·'.:··~.ti~v:i,~.e: ·~itne,~.u;tj·~ }':,: .. '" ,.
··;~ur>i'Qu:s

Tntiekrfrj\i.E>t\

E;ND TABLE· { , '
~

~,··,

.·

. ,

:!,' ·~:·

,::·'.>

! .Creat;:e ·~· flag that .indicates wh~the:r~ ·the
DECLARE INDICA'.rE

TAPE FLAG

:LOOICAL

E'ND_ D,E:CL~RE ;

First check to see if PERlBHE.RAL erJr!orrs are
simply return. Also determine the f',ac;ket subtype.
subtype code., then pxoce'ed •' Otherwi'isbi i.t. i:s: an e.i::r:(>J:tl.•H
~

;

t ~

IF NOT REPORT.PERIPHERAL

This type of packet has not been[ ;Selected
RETURN

END_IF ;
.FIND SUBCOOE NUMBER

= HEADER.CODE sua~lPE ;
-

•I

IF NOT SUBCODE.CONTEXT

THEN
BEGnz
.
.
.
.
S IG-~·~.L '. lL_:LP ~CSBC •.· P~~Al't~TE;RS
REPORT.PACKET IDEN:T:,: . .

~ .~ ;·
.. l ;J;

%CNV $DEC!MAL~H·EADERI.CODE Tl'P!E,!.·.i3).··:;

%CNV-:$0'.ECIMAL{JiEADER;.COJ)E._$U~f~t
·.3)
-·.. . . . . .· , . ' .,,._ . ', ;:'.,: ;·
,. .Jf/';'.ii:..i:l'·:t;
-~<>

RETURN
END i
.ENO_IF

4-58

··.

>I<

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM-PLUS Dispatcher Module
,·

'·

;

name by calling the DEVJCE--NAME procedure~

! ,Now find the device

'CAL'(. MODULE 1DEVSM1' PROCEDURE 'DEVICE_NAME'' ,;

1 Prepare the NOTE NUMBERS file for any notes ,that may ,be requested.
'f

'POINTER NOTE NUMBERS ,CLEAR

r, Now set. up the procedure DEVICE ERROR and the appropriate device level
i·:nt.o<Jule as a. coroutine pair.

':1 '

CALG. MOPOI,E. JNTERMOD DEVERR. DISP .NAME PROCEDURE/ DEVICE_ENTRY'
. . . COROUTlNE' '<DBYTCE ERROR' ; .-:.
EMO :

. .

PR~E;.DOgJ::

DEVI'CE_.:_ERROR

BEGTN ·

·1 ... Tlje ·fo116wiJ:)<Lls. used. to' .format,. the. oµtput. fo.r the .. ·..· · . .·...
·1· ·'.Device .Hard. Error''·· •oev:ice . Soft .. E:rror'·•, .·.·oevice, ·tnter,+upt fim~9ut•

i::. j:lnd, 'SPu.r iow:r):nterrupt' DeV>ice· E:rror PC:lcke,ts •.
l ' .·.·.•· ...··

.·.

··.. •·

'' ' ' '

'·.

•'

.·· ·.· ' ' '

' ' .·· ...... ··•

. i.:·The,Df?Vl<=E IO sµbpacket ·.co9tains· in>tormation ·(;lbout the
!·. d.evice on which the errO'r occ·ured.
' '
'' '
! The<l/O. ACTIVITY: subpacket contains· information.· .a.bout .all .
.l · other "concurrent I/O· in· the. systetth .·

·1 The ·DEVICE OP subpacket contains information about the I/O
Operation In progress· on the device at the time of the error. ·
!.
2 ·,Obtain information from the coroutine . partner ..·.
! '
CO ROOT IN~. ;

t<. Assume ,the 'serial number .test will .succee;d or be

irre·levent~

'S.tT . INTERMOD_ DEVERR. RE,]'ECT....E'LAG TO .·FALSE ;.
' !, ·. ' '. ' '' ' ' .· '
' ; ' .·.· ' ' ''
''' ' .. · '

'[·'.N9w ,t;:e$t: to, see if.' tbis ·devite P.asses .th~ .dr.ive .!$:er.ial~ number· test,.
l'" .,

IF: f{EPORT.oR.'1vE SN< VALlD" AND
'
' '
.
·\iNT.ERMOD· DEVERR"~DRIVE SN NE %CNv:·$BCD(REPORT.DRIVE SN, 12))
,

', ,.,,.

~·

. .

,=, ':"""'"'

-~--

""';""'

THEN,
. Indicate· that the. test failed.
SET INTERMOD_DEVERR.REJECT_FLAG TO.TRUE

.EtiD IF ; .
Determine the type of xeport and format the. output

f accordingly.
CASE REPORT.MODE OF
['BRIEF I 1 :
! .The BRIEF report is one line long.
BEGIN

Now ouput the information based on the result of the test.

4-59

ERROR LOG CONTROL FILE ARCHITECTURE

IF. NOT INTERM.OD.- iDEVERR
~REJECT.
·FLAG
.
.
'
'
THEN
'

I '

"! ·Now. output 'the brief report~
!
BEGIN.
WRITE
REPO.RT··~PACKET ID'ENT,

. %CNV '$RSX.TlME(HEADE~.•.'.TIME STAM:Pi O).,
·. SUBCODE. TEXT~ . . ·
· .· ·. ' .. ·~ . ·.
•
DISP·. OEV!CE • STRING,
.
It-iTERMOp DEVERR. ERROR. TY:PE, . . . . . •,

. .. •
...., '. Functioil ::=:. ~ . J 'IN'!'E·RMC?D:_;DE{IERR:•J)EV...:.E"Y'NC'l'IO~·

, ,<

FORMAT
.

'..

. . . .·

..• . .

. ..

.REPORT .'BRIEF: . FORMAT :; .

N~w increment• t}1e printed pa.cket:

INCREMENT

REPO~T. PRINT ·COUNT

; '.

ENO ;·

· E~m·. IF ;
! Now go

~ithout

back .to th.~ par:t.n~r. J,t w.tll si,mply ~~.tux.Q
printing.
·
·

SET INTERMOD DEVERR.PRINT FLAG TO FALSE
END ;

['FULL', 'REGISTERS']:
The
!

about the error.

BEGIN
! Now output the information based on the result of the test.._
!
IF NOT INTERMOO_OEVERR.REJEC!l'_FLAG
THEN

Output the first page if the report type .. is 'FULL•.

BEGIN
IF REPORT.MODE EQ 'FULJ:;'

'°·:.

·\,

''•

I•

;. :'
••

;_..,

··Now output· . the. information• for··the .standard· s~bpac:ke~~~·
· l·

BEGIN
CALL MODULE REPORT.FULL_MOO PROcEDUR~
! . Now ·Ou:tput the. D~ta'..;s~~packet. informat'ioihu··..
..!

WRITE
·. iNTERMOD

bEVER:R. DEV: .. FUNCTION, .

lNTERMOD-DEVER1~. ERROR .TYPE .

FORMAT .

. . ... . . .

. .. ·• '-

DSPi SUB ANY~FORMAT' 1
. .

4-60

~· '.

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM-PLUS Dispatcher Module

. ~aw . . ou.tp'u:~· th~,.·,·~;v'fc~:,:·:E!r·tor.·:·Posi.tloli:. i'nf()rmat!on

.:·;·:.i~·lt:.::·~;s ··?PP~ic~~fi:>l~·· ... ·

. .

. 'i¢ lN~ER:Aoo~oE.VERi.. ,~R~OR~
THEN ..

l,=·.,'

4-61

ERROR LOG CONTROL FILE ARCHITECTURE
ocv

11~A'

nrrTC

•\.-.1..:0.

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

............ i.....

n.;_""""_ ...... ,,,....i.....~.,_, .....

._it:''-- ........... .__

,,._..:1 .... 1,....,
... ._

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

FORMAT
Select the format ~tatement
on the desired width.

ba~ed

%CND_$1P(REP0RT.WlDE,
!

! WIDE is selected.
!

•11FP15FCEntry !7DP120FCSequence !9DP' I
'!40FC!6DP!48FC!18DP!DP!2FS!20DP!3FL' I
'Device Supplied Information: !FL' 1· .
'--~--- -------- ~----~----~-!2FL't

! NARROW is s~lected •.

!lFP!SFCEntry l7DP ! 2oFCSequence ! 9DP !40FC! 61,)P! FL'
I. ·... ·
.
'Device Supplied Inforination:'!FL'. I
l

!5FC!l8DP!DP!2BS!20DP!3~L'

' __ ....;___ -------- ------.....,..-...:.. ... -! 2FL' ·)
7

END _,IF

;

Now increment the printed packet count and tell the deviCe
! module we wan't it to pr int.

INCREMENT REPORT. PRINT_couwr

SET Il~TERMOD_D~V'ER~. Pi:<INT~:tLAl; TO TRUE
END

ELSE

We don't want to print because the packet was rejected ..
SET

TN'l'F.RMOD_DEVERR.PR!NT_FLAG TO E'ALSE

;

END IF ;
Now go back to the partner. It will output the device registers
.:.c
J.J..

.a...1-.-

\..llt::

--.:-.&...

.1;-'.LJ.U\..

,c:, __

J...l.C2'j

!J.i:>

..t...-·-\...L\..11:;:.

END ;

[I NONE']:

! If the report type is NONE, output nothing.
SET INTERMOD DEVERR.PRINT FLAG :ro FALSE ;

END_CASE ;
! Now COROUTINE back to the partner •
.!

COROUTINE ;

4-62

will print if instructed to ~o so.

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM-PLUS Dispatcher Module

(
! Tesi to·

se~ ··If: tli~ . pac}tet .w~s

IF ·NOT INTERMOD DEVERR.RE:JECT FLAG
.!·

Update tne· files •

. ,! .
CALL '.PROCEDURE' •upoATE RECORD'

! .

·; r:i :«N.OTE~NUMBERS .. 'C.oN~~XT "l'fi.E'.t•t
. :~E~lN: .. : i :' : .. ; .. ; . . · . . · :. ·. : .... : . . . . · . • . ·.· .... , . ,
. '.·&~!f;. :I·N'l'~RMODl_D:E;'/~R~·~DISP~~l\f1E'.'rQ .:'·N', J;

>.: /:·:

. ~.S~R·. •$.REMAINJNG.(IN.TER.MOD •DEV.ERR •. DJ:SP. NAME,,
,·.-

•

>·

••

,.·

:,._

.··:

IF %:PK.T_~MODtJLE,:'(I:.NTERMOD~DEVERR'•Pl:Rt>~t4iiME)
(

« , .,

·THEN·

·.CALL·. MOOULE INTERMOb_DEVERR ,.pISP_NAME PROCEDURE

NOTES''

<".,,

ELSE ..
·SIGNAL

·NONOTES 1 PARAMETERS INTERMOD OEVERR~'DRlVE

TYPE

END IF ,

. E:ND;-;·;.
END. IF ;.

ENO:-;

··

:PIWCEDURE UPDATE RECORD .
B:EGI:N , ·
.. .
.

( Thi~: ·~pto6ed~r·e. :is us~ ··i:o upd'ate ao error :~.YPe·: ·.record ·: i:f. a. ·r;ecor(l fo~', the
·!.~type:· ''oC:e'r;tor· :exists·~· Iif. rt :d.oe'.s O()t .e?Cis,t:,;i a, record: •lS' ·:created:.·'.:.
J

! .· Fir~t

see if' a:. .,record :exists Jn the E,RROR ::ttiFO·E 'fHe •that.•mat:cb~·s;
on th~·: f.oil~V?ing ke:ys!··: · ·
......
'. · · ·
.,

·!

Device name.
Device type
1'
·,'•

.Pack SN

or:ive· SN
Volume label

E.r r.o:r type

4-63

ERROR LOG CONTROL FILE ARCHITECTURE
"'"'"''
,.,._~
""'r"'T,....
.L.'-tr....J.i~"-'...&."'-,_ ..... ,...,,.i.:.&..1"-'t...J

"""'!----..A--\...-... ..,...

._.._...;,,J::"i:.....'-"'--i. .i.C.._

•11-,..:J ... ,'"':'
~

... V".A\.A.....,,_

TU.EN

:. !
, ••. ~::d ·Thls:·type 'of S·um~ary is desi.red. ·
~ .·~i- ~.,'
! ~ ', : .
·~
',; ;, ,

•

••

•

::\\' ..,.'.~EG.IN· ·
,,:: : C:\·)~<HNTER: ~R:R'OR tNFO,.:_E .. FlRS'J:' ·i

:~}.,!:;:No!\' try_to :f,ind.·a re~ord th'at matches on all of the keys.

··.:·:\\":·~!\". :. \ .... ·.· ·. :. ' .•.. ·...·
,\\;''·J!'!·INO·. ERROR.··IN'FO .E ··.

·~ . ·.'':'· :'N'ArIE'~ 'DISP.DEVICE STRING,

;\<~::·•·:
· ·"

•;:•\

.
·.:oEvr~e .TYPE = %CND)IF(I·NTERMOD ·DEVERR.~H~MATC~ FLAG,
... -:-:: ·: . . ' . . :- . lNTER:MOD DEVERR.DRIVE' TYPE,-.
PAC'K. SN -= DEVICE . ;m. PACK SN, - . .
'.DRIVE SN.~ 'INTERMOD DEVERR.DRIVE SN·,
VOLUME LABEL = DEVICE IO.VOLUME LABEL,
'ERROR .TYPE. :::: . .f NTERf•tOp=DEVERR. ERROR~TYPE

See if ·there·~as.. a matCh~ ·
IF. ·.ERROR_ JNFO_ E. CONTEXT
'.THEN

There was a match. Update the record to
show that this error occurred.
BEGIN
INCREMENT ERROR_INE'O_E.ERROR_COUNT ;

IF DISP.PACKET DATE LT ERROR INFO E.FIRST DATE
THEN
BEGIN
SET ERROR_INFO_E.FIRST_DATE TO DISP.PACKET_DATE ;
SET ERROR INFO E.FIRST PACKET TO REPORT.PACKET !DENT
END

END IF ;
IF DISP •. PACKET DATE GT ERROR_}NFO_E.LAST_DATE
THEN

BEGIN
. SET· 'ERROR lNFO'""-E.LAST_DATiE TO DISP.PACKET_DATE

; ·

SET "ERROR~INFO~E.;t.AST:__PA<~Ji<·ET;TO REPORT.PACKET IDENT ;·

4-64

ERROR LOG CONTROL FILE ARCHITECTURE
RSX-llM-PLUS Dispatcher Module

POT ,ERROR tf1JFO E
NAME :; DISP.DEVICE, ,STRING,
. . , . . •.. ,
:.
. .·
·J:)EVICE ·TYPE 7'. %Ct.JD-$)FlJNTERMQJ:)• OEVERR~(~HS,~A;TCH :~r.!(1,,
. ·.

· ·. ·

=
=

· . .

INTERMOD D.EVERR.DRIVE. TYPE., -

PACK. SN
DEVICE· ·ID;.,pACK SN:,. '.· 'DIHVE SN
Il\lTERMori. D.~VERR.ORIVE SN,:'
: .N0LUME :LABEL ·= DEVICE 1'D ~VOLUME LABEL,;'
E,RROR TiPE. = .ua'i'E.RMQO"'.°".DEVER,R,. J!'.RR()R' !TYPE:,·
. ERROR---COPNT :~ l•r . . , ... -. · ,, .. , · · · . ··· . .,... '' ·
FIRST=DATE
:=. otsP ••PACKET; DA~E .
.'
:.
· .. "':' :: , ... : ·
' •'
LAS;T•DATE :7'.· .• nrs,P,P.AC:K~T· [)ATE,:''
E'lRS!ii_i:i?ACf{E·T' ."'.': -~E:"~~'l'.•.PA_C•KET~~DJ~N:T ;,;

... -··

........ ...

:' LAS:T::, PAC-~ET. :.= RE{:10RT.·.;:PAGMEW:.:_IOEJi.T·.,

,' .}'

:':fa~p: :r:t

::"·' :,''

_,.

; . ··

.p-7;"·

;·! /Ft·r--Eit ,.~·,···s~e ,i ·f--; ~ t·e~b·rd
.}· on.: the... following

·oevj,'ce. name
· · .: Devic~ tY.P·e ..
Eack. SN.:•:
·. Drive .SN.

V::olume ianei

...BlopJ{ number

-·'!'

ih~~~./'/:
. l' ;'J!hi,s type· of:
. •!
" ' .: .': ;, '
BJ!JG!N: ; ·

· ·J?o INTER ERROR I NFo G .FIR ST ; .
. · .f,'Now.'):'r:y:t:Q Jfn(:l a;·~ req·ord

tiiat.'

FIND.. ERROR INFO G .
'.NAME =· Dl'SB.DEV'ICE 'S,TRING,

DEVICE 'l'YPK= %CND7 $TF{INTERMOD DEVERR.MISMATCH_iFLA~,

.
- INTERMOD DEVERR.DRIVE TYPE,
PACK SN = DEVICE ID. PACK SN, DRIVE SN = INTERMOD DEVERR.DRIVE SN,
VOLUME LABEL . DEVICE ID .VOLUME LABEL,
BLOCK_NUMBER =· '1NTERMOD_pEVERR.BLOCK~NUMBE.R

See if there was a match.

IF ERROR_INFO_G.CONTEXT

4-65

'·~~:i
.

ERROR LOG CONTROL FILE ARCHITECTURE

THEN

. ! There · was:· ·a , ma.tch. Update the re co rd to

show. that, thi? ·error o·ccµrred.

!
INCREME,NT ERROR INFO' .G •.'.ERROR_COUNT ; .

·ELSE

·1 'This. is th~ f'ir:st .error. of' t'his. Jcint;i. c:r~ate ..a r~cord in the
! : ERROR_:.IN'F.0;.;..G: :file .that describes this error.

~·~T; ERR9R_:INFOiG

. ·.. . : ..· ..·. ..
. .....
•NAME ::: · 'DISP.DEVICE STRING,.·
; . ;. . . . . . .
.. .
..
: '.;·DEVICE·' TY'.P·E:
%CNb-~$1F'( lN:TERtiOD DEVERR~Mts:MA'l'eH F~AGi
.. '' '
' -- .
: ' .· . -. ;INTE·RMOD DEVERR •.DRIVE ·TYPE,PACK SN.=· :DEVICE ID.PACK' SN,- . . .
.
.:~.
. oifrvlr SN. :lNTERttoo· ·DEV.ERR~DRIVE· SN,·.
VOLUME ·LABEL =· .DEV ICE 10 .. VOLlJME LABEL,
.
ERROR JiE.AD. = INT;ERMOD--0.EVERR ~;ERROR HEAD;·
ERROR....,.GRQUP. = INTERMOD DEV·ERR. ER.ROR GROUP,
ERROR-CYLINDER ·= · !NTERMOD. DEVERR'. ERROR CYLINDER,
ERROR-SECT.OR ::: INTERMOD OE\/'ERR.ERROR· SECTOR, .
BLOCK-NUMBER
INTERMOD-DEVJ!:RR~BLOCK-NOMaER.~

·=

ERROR=COUNT

=. l

;

- ·

END IF.;
END-;
END IF ;
END-;

END MODULE

DSP2Pl.CNF

4-66

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
4.6.4

RM02/03 Notes Module NRM23

Following is an annotated listing of the notes module for the RM02 and
RM03 disk drives.

0
MODULE NRM23 'MOl.00'

;

ERROR LOG CONTROL FILE MODULE:

RM02, RM03 Notes

COPYRIGHT (c) 1981 BY
DIGITAL EQUIPMENT CORPORATION, MAYNARD
MASSACHUSETTS.
ALL RIGHTS RESERVED.
' THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED
AND COPIED ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE
AND WITH THE INCLUSION OF THE ABOVE COPYRIGHT NOTICE.
THIS
SOFTWARE OR ANY OTHER COPIES THEREOF, MAY NOT BE PROVIDED OR
OTHERWISE MADE AVAILABLE TO ANY OTHER PERSON.
NO TITLE TO AND
OWNERSHIP OF THE SOFTWARE IS HEREBY TRANSFERED.
THE INFORMATION IN THIS DOCUMENT IS SUBJECT TO CHANGE WITHOUT
NOTICE AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL
EQUIPMENT CORPORATION.
' DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF
ITS SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY DIGITAL.
VERSION 01.00
R. Ryan

30-Jun-81

This is one of the many device modules,which is called by the device
error dispatcher (DSP2cl) or device information dispatcher (DSP3cl)
to process notes for all the device dependent information.

PROCEDURE NOTES
!

! This procedure, which is called from the DSP2cl module, processes any
! requests for notes.
!

BEGIN
Print the NOTE header and define the format for the NOTE section.

WRITE FORMAT
'!3FLNotes on RM02, RM03 errors:!2FL'
POINTER NOTE NUMBERS FIRST
WHILE NOTE NUMBERS.CONTEXT DO

4-67

ERROR LOG CONTROL FILE ARCHITECTURE
Nnh=i.s

Mna111-P

BEGIN
SELECT NOTE NUMBERS.INDEX OF
! Note number 1.
BEGIN
WRITE FORMAT
I*
RMDA bits 5,6,7 are unused, however if they are'
WRITE FORMAT
set, they will be interpreted as the high order'
WRITE FORMAT
bits of the sector address.
This may result in'
WRITE FORMAT
an Invalid Address Error.! 3FL' ;
END
ELSE

This is an unknown note number.
SIGNAL 'UNKNWNNOT' PARAMETERS

~~~~~~~~c:~~~~:(~~:~DN~~~~RS.INDEX, 3),
i~i~~nvu

u~v~~~.u~iv~

iir~

END SELECT

POINTER NOTE NUMBERS NEXT
END

END

END MODULE

NRM23.CNF

4-68

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
4.6.5

Subpacket Definitions

The following sections list the DIGITAL-standard subpackets.
listed under the modules that declare them.

They are

HEADER
subpacket
4.6.5.1 Subpackets Declared by DISPATCH - The
It describes the
contains
information
largely
from
SYSCM.
characteristics of the system and packet. This subpacket is always
required.
SUBPACKET HEADER
SUBPKT FLAGS
SFLG HDR
SFLG-TSK
SFLG-DID
SFLG-DOP
SFLG-DAC
SF LG-DAT
SFLG-MBC
SFLG-CMD
SFLG-ZER
OP SYS
FORMAT ID
OP SYS-ID
CONTEXT CODE
cc NOR
CC-STA
CC-CD A
FLAGS
FLG ADR
FLG-COU
FLG-QBS
ENTRY SEQ
ERROR-SEQ
CODE TYPE
CODE-SUBTYPE
TIME-STAMP
PROC-TYPE
RESERVED
PROC ID
URM CPU
END_PACKET ;

DISP.NEXT PACKET
:WORD ;
:BIT [O]
:BIT [l]
:BIT [2]
:BIT [3]
:BIT [4]
:BIT [5]
:BIT [13]
:BIT [14]
:BIT [15]

E$HSBF
Header Subpacket
Task Subpacket
Device Id. Subpacket
Device Op. Subpacket
Device Ac. Subpacket
Data Subpacket
22-bit Massabus Controller
Command Subpacket
' I/O counts zeroed
E$HSYS
E$HIDN
E$HSID
E$HCTX
Normal Entry
Start Entry
CDA Entry
E$HFLG
Addressing mode
Error Counts supplied
Q-BUS system
E$HENS
E$HERS
E$HTYC
E$HTYS
E$HTIM
E$HPTY
Reserved byte
E$HURM
Processor Identifier

:BYTE ;
: ASCII [4]
:BYTE ;
:BIT [O]
:BIT [l]
:BIT [2]
:BYTE ;
: FI ELD [ 0: 2]
:BIT [2]
:BIT [3]
:WORD
:WORD
:BYTE
:BYTE
:RSX TIME
:BYTE
:BYTE ;
:WORD ;
:FIELD [O: 4]

The TASK subpacket contains information either about the task that
logged the packet, or the task that caused the packet to be logged.
SUBPACKET TASK
DISP.NEXT PACKET ;
TASK NAME
:LONGWORD ;
UIC
:WORD ;
UIC MEMBER
: FIELD [ 0: 8]
UIC-GROUP
:FIELD [8:8]
TI DEV
: ASCII [2)
TI-UNIT
:BYTE ;
FLAGS
:BYTE ;
:BIT [0]
FLG PRV
FLG PRI
:BIT [l]

E$TTSK
E$TUIC
Member number in UIC
Group number in UIC
E$TTID
E$TTIU
E$TFLG
Privileged Task
Privileged Terminal

The DEVICE_ID subpacket contains information about the device on which
the error occurred.

4-69

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
DISP.NEXT PACKET
:ASCII (2]
:BYTE
:BYTE
:BYTE
:BYTE

SUBPACKET DEVICE ID
MNEMONIC
LOGICAL UNIT
CONTROLLER NUM
PHYS UNIT
PHYS-SUB UNIT

E$ILDV
E$ILUN
E$IPCO
E$IPUN
E$IPSU

$IF SUPPORT.RSX_llM PLUS
$THEN
IF OP SYS.SUFFIX EQ 'P'
THEN
BEGIN
PHYS DEV MN EM ON
END

: ASCII [2]

E$IPDV

:BYTE ;
:BIT [ 0]

E$IFLG
Subcontroller device

END IF
$END IF

DEV FLAGS
DFLG SUB

$IF SUPPORT.RSX llM PLUS

$THEN
DFLG NUX

:BIT [l]

No UCB extension

:BYTE ;
: ASCII [ 12]
:LONGWORD
:WORD ;
:LONGWORD
:LONGWORD
:BYTE
:BYTE

Reserved byte
E$IVOL
E$IPAK
E:;iIDCL
E$IDTY
E$IOPR
E$IERS
E$IERH

$END_ IF
RESERVED
VOLUME LABEL
PACK SN
DEV TYPE CLASS
DEV-TYPEIO COUNT
SOFT ERCNT
HARO-ERCNT

$IF SUPPORT.RSX llM PLUS

$THEN
IF OP SYS.SUFFIX EQ 'P'
THEN
BEGIN
WRD XFR COUNT
CYL-CRS-COUNT
END

:LONGWORD
:LONGWORD

E$IBLK
E$ICYL

END IF
$END_IF
END PACKET

4-70

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
The DEVICE_OP subpacket contains information about the
operation.
SUBPACKET DEVICE OP
TASK NlLME
UIC
UIC MEMBER
UIC-GROUP
TI DEV
T.I-UNIT
RESERVED
IO FUNCTION
SF IQX
SF-IQQ
SF-IQUMD
FLAGS FLG TRA
FLG-DMA
FLG-BAE
RESERVED
XFER ADDRESS 1
XFERl HIGH 2
XFERl-HIGH-6
XFER ADDRESS 2
XFER2 TAUB
XFER BYTE-COUNT
RETRIES LEFT
MAX RETRIES
END PACKET ;

DISP.NEXT PACKET
:LONGWORD ;
:WORD ;
: FIELD [ 0 : 8]
: FIELD [ 8 : 8 ]
: ASCII [2]
:BYTE ;
:BYTE ;
:WORD ;
:BIT [O]
:BIT (l]
:BIT [2]
:BYTE ;
:BIT [O]
:BIT [l]
:BIT [2]
:BYTE ;
:WORD ;
:FIELD [4: 2]
: FIELD [ 0: 6]
:WORD ;
: F I ELD

[ 0 : 13 ]

:WORD
:BYTE
:BYTE

requested

E$0TSK
E$0UIC
Member number in UIC
' Group number in UIC
E$0TID
E$0TIU
Reserved Byte
E$0FNC
IQ.X subfunction bit
IQ.Q subfunction bit
IQ.UMD subfunction bit
E$0FLG
Transfer operation
DMA device
22 bit addressing device
Reserved Byte
E$0ADD + 0
High Order 2 bits of address
High Order 6 bits of address
E$0ADD + 2
T. A. ..1.u uni ts of bytes
E$0SIZ
E$0RTY
E$0RTY+l

The IO ACTIVITY subpacket contains information about other
on in the system at the time the error was detected.
SUBPACKET IO ACTIVITY
MNEMONICLOGICAL UNIT
CONTROLLER NUM
PHYS UNIT
PHYS-SUB UNIT

I/O

going

DISP.NEXT PACKET REPEATED ;
: ASCII [2J
E$ALDV
:BYTE
E$ALUN
:BYTE
E$APCO
:BYTE
E$APUN
:BYTE
E$APSU

$IF SUPPORT.RSX llM PLUS
$THEN
IF OP SYS.SUFFIX EQ 'P'
THEN
BEGIN
PHYS DEV MNEMON
END

: ASCII

[2]

E$IPDV

END IF
$END_IF
DEV FLAGS
r...r.tT

LJ.Cl.JU

:BYTE ;
t""rTn

OUD

E$IFLG
Subcontroller device

;BIT [0]

4-71

ERROR LOG CONTROL FILE ARCHITECTURE
N0t-1?~

Mnn n l P

$IF SUPPORT.RSX_llM_PLUS
$THEN
DFLG NUX

:BIT [l]

No UCB extension

:BYTE ;
:LONGWORD
:WORD ;
: FIELD [ 0 : 8]
:FIELD [8:8]
: ASCII (2]
:WORD ;
:BIT (0)
:BIT [l]
:BIT (2]
:BYTE ;
:BIT [0]
:BIT [l]
:BIT [2]
:BYTE ;
:WORD ;
:FIELD (4: 2]
: F IE LO [ 0 : 6 ]
:WORD ;
:FIELD [0:13]
:WORD ;

E$ATIU
E$ATSK
E$AUIC
Member number in UIC
Group number in UIC
E$ATID
E$AFNC
IQ.X subfunction bit
IQ.Q subfunction bit
IQ.OMO subfunction bit
E$AFLG
Transfer operation
OMA device
22 bit addressing device
Reserved Byte
E$AADD + 0
High Order 2 bits of address
High Order 6 bits of address
E$AADD + 2
T. A. in units of bytes
E$ASIZ

$END_IF
TI UNIT
TASK NAME
UIC
UIC MEMBER
UIC-GROUP
TI DEV
IO-FUNCTION
SF IQX
SF-IQQ
SF-IQUMD
FLAGS FLG TRA
FLG-DMA
FLG BAE
RESERVED
XFER ADDRESS 1
XFERl HIGH 2
XFERl-HIGH-6
XFER ADDRESS 2
XFER2 TAUB
XFER BYTE-COUNT
END PACKET

4.6.5.2 Subpackets Declared by DSPlMl/DSPlPl - There are several
different
DATA subpackets declared by DSPlMl/DSPlPl.
Here are
descriptions of each of them.
The following DATA subpacket
(Code
1, Subcode
information about a "Status change" operation:

contains

SUBPACKET DATA
D!SP.NEXT PACKET
LIMIT CODE
:BYTE
LOG CODE
:BYTE ;
:BYTE ;
FLAGS
FLG CRE
:BIT [O]
FILE SPEC LEN
:BYTE ;
FILE-SPEC: ASCII [ 80]
END PACKET ;
The following DATA subpacket
(Code
1, Subcode
information about a "Switch Logging Files" operation:
SUBPACKET DATA
DISP.NEXT PACKET
:WORD ;
RESERVED
FLAGS
:BYTE ;
:BIT [0]
FLG CRE
FLG-DEL
:BIT [l]
FILE SPEC LEN
:BYTE ;
FILE-SPEC: ASCII [80]
END PACKET

4-72

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
The following DATA subpacket
(Code
1,
Subcode
information about an "Append to File" operation:

contains

DISP.NEXT PACKET
SUBPACKET DATA
RESERVED
:WORD ;
:BYTE ;
FLAGS
:BIT [0]
FLG CRE
FLG-DEL
:BIT [l]
:BYTE ;
FILE SPEC LEN
FILE-SPEC: ASCII [ 80]
END _PACKET ;
The following DATA subpacket
(Code
1, Subcode
information about a "Set Backup File" operation:
DISP.NEXT PACKET
SUBPACKET DATA
:WORD
RESERVED
:BYTE ;
FLAGS
:BYTE ;
FILE SPEC LEN
FILE-SPEC: ASCII [80]
END_PACKET ;
The following DATA subpacket
(Code
1, Subcode
information about a "Change Limits" operation:
SUBPACKET DATA = DISP.NEXT PACKET REPEATED
HARD LIM FLAG
:BYTE ;
NEW LIMH
:BIT [O]
HARD LIMIT
:BYTE ;
SOFT-LIM FLAG
:BYTE ;
NEW LIMS
:BIT [0]
SOFT LIMIT
:BYTE ;
MNEMONIC
: ASCII [ 2]
LOGICAL UNIT
:BYTE ;
~BYTE ;
RESERVED
END PACKET

4.6.5.3 Subpackets
for device errors
that is specific
device-level module

Declared by DSP2Ml/DSP2Pl - The DATA subpackets
(Code = 2, Subcodes = 1, 2, 3) contain information
to each device.
Please see the
appropriate
for the format of the DATA subpacket.

4.6a5.4 Subpackets Declared by DSP3Ml/DSP3Pl - The DATA subpackets
for device information messages
(Code
3,
Subcode = 1) contain
information that
is specific to each device.
Please see
the
appropriate device-level module for the format of the DATA subpacket.

4-73

ERROR LOG CONTROL FILE ARCHITECTURE
Nnh:~~

Mnn11l P

4.6.5.5 Subpackets Declared by DSP4Ml/DSP4Pl - There is no DATA
subpacket for "Mount", "Dismount", and "Device Info Reset" events
(Code = 4, Subcodes = 1, 2, 3).
The following DATA subpacket
(Code
4, Subcode
information about a "Block Replacement" operation:

contains

SUBPACKET DATA = DISP.NEXT PACKET ;
FLAGS
:WORD ;
PRIMARY RBN
:BIT [0]
SUCCESS:BIT [l]
LBN
:LONGWORD
NEW RBN
:LONGWORD
OLD-RBN
:LONGWORD
END PACKET

4.6.5.6 Subpackets Declared by DSPSM1/DSP5Pl - This DATA subpacket
(Code
5, Subcode = 1) contains information about a "Memory Parity
Error" event:
SUB PACKET REGISTER
RESERVED
RESERVED
P CSROO
P CSROl
P-CSR02
P-CSR03
P-CSR04
P-CSR05
P-CSR06
P-CSR07
P-CSRCS
P-CSR09
P-CSRlO
P-CSRll
P-CSR12
T"\-,,,.,n '1 ..,

I:'

~.::>1'.1.J

P-CSR14
P-CSR15
LOW ERR
HIGHERR
CACHERR
CSHCTRL
CSHMAIN
CACHHIT
END PACKET ;

DISP.NEXT PACKET NAMED
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD
:WORD

Mask for Cache Registers
Mask for Parity CSR's
Memory Parity CSR 00
Memory i:>ari ty CSR Ol
Memory Parity CSR 02
Memory Parity CSR 03
Memory Parity CSR 04
Memory Parity CSR 05
Memory Parity CSR 06
Memory Parity CSR 07
MPmory Parity CSR 08
Memory Parity CSR 09
Memory Parity CSR 10
Memory Parity CSR 11
Memory Parity CSR 12
D~ r i t-v CSR 13
Memory -----.J.
Memory Parity CSR 14
Memory Parity CSR 15
Low Error Address Register
High Error Address Register
Cache Error Register
Cache Control Register
Cache Maintenance Register
Cache Hit/Miss Register

The following DATA subpacket (Code
5, Subcode
information about an "Unknown Interrupt" event:

contains

SUBPACKET DATA = DISP.NEXT PACKET
VECTOR OVER FOUR
:BYTE
LOST INT
:BYTE ;
END PACKET

4.6.5.7 Subpackets Declared by DSP6Ml/DSP6Pl - The "Power
event (Code = 6, Subcode = 1) has no DATA subpacket.

4-74

Recovery"

ERROR LOG CONTROL FILE ARCHITECTURE
Notes Module
4.6.5.8 Subpackets Declared by DSP7Ml/DSP7Pl - The following DATA
subpacket
(Code
7, Subcode = 1) contains information about a "Time
Change" event:
SUBPACKET DATA
NEW TIME
END_PACKET ;

DISP.NEXT PACKET
:RSX TIME

The following DATA subpacket
(Code
7,
information about a "System Crash" event:

Subcode

contains

The following DATA subpacket (Code = 7,
Subcodes
3,
4)
information about a "Driver Load" or "Driver Unload" event:

contains

SUBPACKET DATA
CRASH TIME
OP SYS
FORMAT ID
OP SYS-ID
TASK NAME
TI DEV
TI-UNIT
FLAGS
KERNEL APR5
URM
URM CPU

DISP.NEXT PACKET
:RSX TIME
:BYTE ;
:BYTE ;
: ASCII [ 4]
:LONGWORD ;
: ASCII [ 2] ;
:BYT£ ;
:BYTE ;
:LONGWORD
:WORD ;
: FIELD

[ 0: 4]

END_PACKET ""i

SUBPACKET DATA = DISP.NEXT PACKET
DRIVER NAME
:ASCII [2]
END_PACKET-;

;

The following DATA subpacket
(Code
7,
information about a "System Message" event:
SUBPACKET DATA = DISP.NEXT PACKET ;
MESSAGE LEN
:WORD ;
MESSAGE-TEXT
:ASCII [80]
END PACKET

4-75

Subcode

contains

CHAPTER 5
CONTROL FILE LANGUAGE GUIDE

This chapter describes the Control File Language used by the Report
Generator
of the Error Logging System.
The preceding chapter
describes how to include support for non-DIGITAL devices in the Error
Logging System.

5.1

CONTROL FILE OVERVIEW

The control file for the Report Generator (RPT) describes the format
of the error log file and the format of reports based on the file.
The actions specified are executed for each event to produce a report.
The control file specifies the format of the data in an error log
packet, and the output format of the report.
In addition, the control
file specifies information on the accumulation of summary information,
how to derive additional information, and the handling of selection
criteria for reports.
Control file modules are ASCII text files containing a series of
statements written in the Control File Language (CFL). The CFL
compiler produces intermediate form modules which are placed in a
universal library. This library is the control file.

5.1.1

Report Generator General Processing

RPT is an interpreter for the intermediate form modules contained in
the control file.
RPT processes control file modules which can, in
turn, process error log files.
The control file module, written in
CFL, specifies the processing to be performed on the current packet.
The processing usually involves calling subroutines from
other
modules.
CFL includes primitives that stop packet processing on an error
because the packet does not meet specified selection criteria.

5-1

CONTROL FILE LANGUAGE GUIDE
5.1.2

The General Format of an Error Log Packet

Here is the general format of an error log packet:

+-------------------------------+
I Packet Length
I
+-------------------------------+
Length of Subpacket l
I
l -------------------------------+

Data for Subpacket 1

+-------------------------------+
I Length of Subpacket 2
I
+-------------------------------+
Data for Subpacket 2

+-------------------------------+

+-------------------------------+
! Length of Subpacket n
!
T-------------------------------•
I

Data for Subpacket n

+-------------------------------+
Each subpacket contains a different type of data.
The information in
the subpackets, taken together, describes an event loggea oy cne error
logging system.
The control file contains primitives for describing a
subpacket so that its contents can be symbolically manipulated. Other
primitives describe the entire packet as a unit.

5.1.3

Control File Language

CFL is a speci~lized language designed for this application.
It is a
statement-oriented, block-structured language similar in concept to
Pascal and Algol. Unlike BLISS, CFL is not expression-oriented.
CFL
does differentiate between statements and expressions.
However, CFL has some of the capabilities of BLISS in that expressions
can contain conditionals. This feature handles the more complex data
formats of error log files. CFL does not include the full set of
primitives required of an ordinary general-purpose language. A number
of specialized primitives speed up and simplify the handling of common
error log data formats.

5.1.4

General Format of Control File Modules

Each control file module contains lines of ASCII text.
Each line is a
sequence of elements - keywords, variable names, numbers, operators,
and the like.
Spaces and tabs separate atomic items, such as keywords
or names.
Excess spaces and tabs are ignored and can be used freely
for formatting.

5-2

CONTROL FILE LANGUAGE GUIDE
Insert comments in the module text by prefixing
exclamation point
(!).
The compiler ignores any
following an exclamation point.
The three basic
declarations.

elements

CFL

are

statements,

them with an
text on a line
expressions

A statement describes an action.
Statements begin
keyword and are terminated by a semicolon (;).

An expression describes a computation.
Expressions are
terminated by any nonexpression keyword, or by a comma (,).
A nonexpression keyword is a keyword that is not valid in an
expression=
Expressions can also be terminated like a
statement, with a semicolon (;).
Expressions can also be
included
in
an expression-list, using conditionals to
determine whether or not a given statement or expression is
to be executed. Expression-lists consist of a fixed number
of expressions, separated by commas (,).

A declaration defines the content of packets and subpackets
and defines groups of tables for evaluating packets and
subpackets. The syntax of a declaration differs for each
use.
Group declarations start w1~n DECLARE and end with
END DECLARE. Table declarations begine with TABLE and end
with
END_TABLE.
Dynamic table declarations begin with
DYNAMIC TABLE and end with END TABLE.
Packet and subpacket
declarations begin with PACKET and SUBPACKET, respectively,
and end with END PACKET.

5.1.5

with

and
a

Files

CFL can obtain input from and direct output to any one
files.
The files have the following internal names:
e

set

INPUT
The data input file.
Data packets are read from the file.
The control file can open this file,
close it, and read
packets from it.

REPORT
The report output file.
Lines of ASCII text are written to
this file.
The control file can open this file, close it, and
write ASCII text to it. Automatic paging is available for
this file.

•

USER
The user prompting file.
The user can be prompted for
input,
and input read from this file.
The control file can open this
file, close it, and read ASCII data from this file,
with
optional prompting.

COMMAND
The command input file.
The user can be prompted for
input,
and
input read from this file.
The control file can read
ASCII data from this file, with optional prompting.

5-3

CONTROL FILE LANGUAGE GUIDE

ERROR
The error output file.
to this file.

5.2

The control file can write ASCII

data

TYPES AND EXPRESSIONS

RPT types data to allow easy manipulation of error log information.
Expressions describe data values used by RPT.
This section describes
the attributes of supported data types and the format of expressions.

5.2.1

Data Types

CFL supports seven data types:
logical, string, numeric,
field,
pointer, RSX TIME and VMS TIME.
The evaluation of an expression
results in a value. This value is one of the supported data types.
The data type of an expression is determined from its context.
The only automatic conversions are from string values to numeric
values, and conversions of numeric values between different numeric
types and field types.
The following sections describe the types in detail.

5.2.1.1 LOGICAL Type - The LOGICAL type expresses the Boolean values,
TRUE and FALSE. A LOGICAL type is equivalent to a BIT type. No other
automatic conversions are performed to or from LOGICAL types.
Express
the literal values for this type with the keywords TRUE and FALSE.

5.2.1.2 STRING Type - The STRING type represents strings of binary
bytes.
Literal values for the string type cannot be represented.
String operations must be used to construct string literals.
For purposes of conversion, numeric values are considered exactly
equivalent to strings.
The length of the string is the number of
bytes used to represent a value of the numeric type. For example, a
WORD
is
equivalent to a string of length 2.
The following
equivalences are used:
Type

Equivalent string

BYTE
WORD
LONGWORD
QUADWORD

String
String
String
String

of
of
of
of

length
length
length
length

2
4
8

Strings of length four or less are converted to numeric types by
appending leading zero bytes to form a longword. Strings of length
five to eight are converted to numeric types by appending leading zero
bytes to form a quadword. Strings of length greater than eight are
not converted to numeric types.
String declaration requires specification of the maximum size
string. The syntax for string declaration is:
STRING[ size ]

5-4

the

CONTROL FILE LANGUAGE GUIDE
If the string is a variable, it can contain any number of elements up
to and including the specified maximum.
If the string is part of a
data declaration,
it contains exactly the number of characters
specified.

5.2.1.3 ASCII Type - The ASCII type represents character strings.
ASCII string literals are represented by character strings enclosed in
a pair of apostrophes ('string'). Two successive apostrophes in a
quoted string represent a single quoted apostrophe. Therefore the
string 'ABC''DE' represents the string literal ABC'DE.
The keyword
NULL represents the null string. There is no automatic conversion to
or from ASCII strings. You cannot use the quotation mark
en; to
enclose such strings, nor does the quotation mark require flagging.
ASCII string declaration requires specification of the maximum
of the string. Specify the maximum string length as follows:

length

ASCII[ size ]
If the ASCII string is a variable,
it can contain any number of
characters up to the specified maximum.
If the ASCII string is a data
item, it must contain exactly the number of characters specified.

5.2.1.4 Numeric Types - Numeric data types represent numbers for
computation.
The numeric types are distinguished only by the length
of the bit field used to contain the number. A BYTE is a one-byte
field, a WORD a two-byte field, a LONGWORD a four-byte field, and a
QUADWORD an eight-byte field.
For purposes of conversion, the numeric
types are considered equivalent to strings, with the length determined
by the type.
The special numeric type VALUE indicates a natural machine value.
VALUE is a WORD on the PDP-11 and a LONGWORD on the VAX-11.
Numeric types have a default output radix of decimal.
expressing a numeric type is:

The syntax

A
for

type option,option, •••
The valid options are divided into radix options and attribute
They are
options.
The radix options determine the print radix.
DECIMAL, OCTAL, HEX, BCD, BINARY, or RAD50.
The attribute options are WIDTH and FILL.
The WIDTH option has the format
WIDTH=n
and specifies the width of print field in characters.
The FILL option has the format
FILL='character'
and specifies the fill character.

5-5

CONTROL FILE LANGUAGE GUIDE
Here are the default print field width and
choice of radix:
Radix

Fill

fill

character

for

Print Field Width
BYTE

WORD

LONGWORD

QUADWORD

DECIMAL space

OCTAL
HEX
BCD

IO I

3
2
2

6
4
4

11
8
8

22
16
16

BINARY

IO I

RAD50

space

n/a

IQ I
IO I

each

The default radix is DECIMAL.
For example, here is the specification for a LONGWORD to be printed in
BCD using leading spaces and a field six characters wide:
LONGWORD BCD,WIDTH=6,FILL='

The special radix MACHINE is the normal radix used to express values
for the host machine. The MACHINE radix is OCTAL for the PDP-11 and
HEX for the VAX-11.
You can express numeric literals in a number of ways. A sequence of
digits is, by default, interpreted as a VALUE numeric literal. The
number is assumed to be decimal.
Express a numeric literal of a
specified type and radix as follows:

# <type indicator> '<character string>'
The character string is interpreted according to the specification
given by the type indicator.
The type indicator is a one- or
two-character string specifying the type of the number and the radix

which

interpret the character string.

The first character of

the trpe indicator is the type of the number, as shown:
B
W
L
Q
V

Byte
Word
Longword
Quadword
Value

The second character of the type indicator is the radix in which to
interpret the character string.
If the radix is not specified, the
character string is assumed to be decimal. The valid radix indicators
are:
A
B
D
I
0
R
X

ASCII
Binary
Decimal
Bit value
Octal
Radix 50
Hexadecimal

A minus sign (-) preceding any character string interpreted as a
number indicates the two's complement of that number in the indicated
radix, that is, binary, octal, decimal, and hexadecimal.

5-6

CONTROL FILE LANGUAGE GUIDE
For example, the character string to represent a
the octal value "17" would be:

byte

that

contains

#BO'l7'
while the character string to represent a word
"-16." would be:

containing

the

value

#W'-16'
The bit value radix indicator specifies that the quoted number is a
decimal number representing a bit position. The value of the literal
is 2 raised to the power of the bit position.

5.2.1.5 Field Types - The field types represent fields of numeric
types. The BIT type represents a single bit of a numeric type, and is
equivalent to a LOGICAL type.
The FIELD
type
represents
a
one-or-more-bit field of a numeric type, and is equivalent to a
numeric type. A field type is always a field of a numeric variable.
There are no literal values for field types.
The syntax for expressing a BIT type is:
BIT[ bit number ]
The syntax for expressing a FIELD type is:
FIELD[ low bit number :

field length in bits ]

In either case a field type is declared directly following the numeric
type of which it is a field.

5.2.1.6 POINTER Type - The POINTER type is a table pointer.
Use it
to declare variables that temporarily store pointers for later use.
The POINTER type cannot be converted to or from any other type. There
is no literal representation of the POINTER type.
The value of the POINTER type is specific to a given table. A POINTER
variable containing a value specifying a table entry for a given table
cannot be used to reference an entry in another table.
The variable
can, however, be loaded with another value referencing an entry in
another table.

5.2.1.7 RSX TIME Type - The RSX TIME type represents a time in RSX
format.
RSX time is represented as six sequential bytes, containing
the year since 1900, the month, day, hour, minute, and second in that
order.
This is a compression of the format returned by the Executive
GTIM$ directive.
The RSX TIME type can only be printed or compared to
types, or converted using one of the %TIM functions.

other

RSX TIME

5.2.1.8 VMS TIME Type - The VMS TIME type represents a time in VMS
format.
VMS time is represented as a quadword containing the time in
hundreds of nanoseconds since 17 November 1858.

5-7

CONTROL FILE LANGUAGE GUIDE
The VMS TIME type can only be printed or compared to
types, or converted using one of the %TIM functions.

5.2.2

other

VMS TIME

variables

The named variable is the fundamental unit for data manipulation.
Named variables are defined in a given module, and available to that
module and any modules called by the module.
Named variables are
declared in named groups. The full name of a variable, that is, the
name by which it is referenced, is the name of the group, a period
(.), and the nafue of the variable in the group:
<group name>.<variable name>
The group and variable names cannot be more than 15 characters. Names
can include the characters A through Z, the numbers 0 through 9, the
dollar sign ($), and the underscore ( ) • The leading character of a
name must be alphabetic.
Use the-same syntax to reference data in
either packets or subpackets.
The CFL compiler assigns each variable a type through declarations.
(See Section 5.4 for a description of the declaration process.)
Variables that are not fixed-length, such as ASCII and STRING type
variables, are assigned a maximum length as well. The variable can
contain any amount of data that fits in its maximum length specified.
A field in the current record of a table can be referenced in the same
manner as a variable, as follows:
<table name>.<field name>
The field value referenced is the specified field in the current
record of the table.
If there is no current record for the table, an
error results.
Several special variables provide information about a group, packet,
subpacket, or table. Reference the special variables as follows:
<group,

[sub]packet or table name>.<special variable name>

The special variables are described below:
e

LENGTH
LENGTH is the length of the data in the group in addressable
units of the host machine (bytes for the PDP-11 and VAX-11).
LENGTH includes all repetitions for repeated data or records
for tables.

POINTER
POINTER returns the current pointer for the specified group.
POINTER is not valid for any data structure that would not
have a current record context. This includes variables and
non-repeated data.
There is a current record context for
tables and repeated data.

CONTEXT
CONTEXT returns a logical value. If the specified group has a
current record context, the value is TRUE.
If the specified
group has no current record context, the value is FALSE.

5-8

CONTROL FILE LANGUAGE GUIDE
e

COUNT
COUNT returns a numeric value representing the number of
records in a group. For groups of variables, it is always 1.
For packets or subpackets, it is the number of repetitions of
the data.
For tables it is the number of records in the
table.

5.2.3

Literals

Literal values can be assigned symbolic names (See Section 5.6.2 for
information on the LITERAL statement). These symbolic names have the
same syntax restrictions as variable names.
Literal names are
considered equivalent to their values in expressions.

5.2.4

Expressions

Expressions describe a computation through a sequence of operands and
operators.
Operands are variables or literals. Operators direct the
computation.
Expression evaluation is from left to right, and
operator precedence is observed.
Use parentheses, (() and ()), to
override precedence.
Operators are either unary, which means that they take one operand, or
binary, which means that they take two operands.
Unary operators can
either precede the operand, in which case they are called prefix
operators, or succeed the operand, in which case they are called
suffix operators.
Binary operators are always between the two
operands.
Operators are type-specific, that is, they operate between two
elements of a specific type to produce a result of a specific type.
The elements of an expression can be any of the following:
•

Literals
Literals express fixed values of a given type.

•

Variables
Variables reference previously computed values.

•

Subexpressions in parentheses
Any valid expression enclosed in parentheses can be used as an
element to an operator.

Functions
A function is a predefined computation.

(See Section 5.3.)

The following sections describe the operators in detail.

5.2.4.1 String Operators - String operators produce either binary or
ASCII string results.
The result is a string with the same type as
the operand string or strings. The string operators are:

5-9

CONTROL FILE LANGUAGE GUIDE
•

String concatenation -

I binary operator

The string concatenation operator concatentates the first
operand with the second.
Both strings must be of the same
type.
For example, the following expression:
'ABC'

'DEF'

produces the string:
'ABCDEF'
•

Substring extraction - <n:m> unary suffix operator
The substring extraction operator produces the string formed
by character n and the next m elements. String element
numbers start with 1. For exampae, the expression:
'ABCDEFGH'<4:3>
produces the string:
'DEF'
Both n and m must be word values.

•

Element Extraction - <n> unary suffix operator
The element extraction operator is a special case of the
substring extraction operator which extracts the nth element
as a single-character substring. For example, the expression:
'ABCDEFGH'<4>
produces the string:

'D'

5.2.4.2 Logical Operators - Logical operators perform operations on
logical variables, or compare string or numeric variables to yield
logical results.
Here are the logical operators:
•

Logical AND binary operator
The logical AND operator does a
logical AND of
expressions. For example, the following expression:

the

two

the

two

TRUE AND FALSE
produces the logical value FALSE.
•

Logical OR binary operator
The logical
expressions.

OR operator does a
logical OR of
For example, the following expression:

TRUE OR FALSE
produces the logical value TRUE.

5-10

CONTROL FILE LANGUAGE GUIDE
•

Logical Exclusive-OR (XOR) binary operator
The logical exclusive-OR operator does a logical exclusive-OR
of
the
two
expressions.
For example, the following
expression:
TRUE XOR TRUE
produces the logical value FALSE.

•

Logical NOT unary prefix operator
The logical NOT operator produces the logical complement of
single variable. For example, the following expression:

NOT TRUE
produces the logical value FALSE.
There is also a logical
numeric expressions:
•

Bit extraction -

operator

produce

logical

results

from

[n] unary suffix operator

The bit extraction operator is TRUE if and only if bit n of
the longword expression operand is set. For example, the
following expression:
#WO' 305'

[4]

produces the logical value FALSE. The binary value
305 is 011000110, with the fourth bit clear.
numbered from O.

of octal
Bits are

5.2.4.3 Relational Operators - Relational operators compare string,
time, or numeric operands.
The comparisons are string comparisons if
both operands are string or ASCII string operands.
The comparisons
are numeric comparisons if one operand is numeric and the other is
either numeric or string. You cannot compare ASCII string operands
and numeric operands.
The comparisons are time comparisons if both
operands are times of the same type.
You cannot compare different
types of time.
In numeric comparisons, the larger numeric value is greater.
In string comparisons, CFL stops at the first two characters that
don't match and performs an ASCII sort.
That is, CFL compares the
ASCII values of the characters.
NOTE
Although Z is greater than A in ASCII,
an ASCII sort is not the same as an
alphabetical sort.
Any lowercase letter
has greater value than any uppercase
letter,
for
instance,
but
any
alphabetical character has greater value
than any numerical character, and so
forth. See any standard reference.

5-11

CONTROL FILE LANGUAGE GUIDE
If one string is longer than the other and the shorter string has the
same leading elements as the longer, the longer string is greater.
In time comparisons, later times are greater.
Here are the relational operators:
•

Equality (EQ) binary operator
The equality operator is TRUE if and only if the operands
equal. For example, the following expression:

are

#WD'l23' EQ #WD'355'
produces the logical value FALSE.
•

Inequality (NE) binary operator
The inequality operator is TRUE if and only if the operands
are not equal. For example, the following expression:
'ABCDEF' NE 'ABC'
produces the logical value TRUE.

•

Greater-than (GT) binary operator
The greater-than operator is TRUE if and u11.J.y .1.L Lilt= f lrst
operand is greater than the second operand. For example, the
following expression:

123 GT 355
produces the logical value FALSE.
•

Less-than (LT) binary operator
The less-than operator is TRUE if and only if the first
operand is less than the second operand. Fer example, the
following expression:
'ABCDEF' LT 'ABCZZZ'
produces the logical value TRUE.

•

Greater-than-or-equal (GE) binary operator
The greater-than-or-equal operator is TRUE if and only if the
first operand is greater than or equal to the second operand.
For example, the following expression:
45 GE 45
produces the logical value TRUE.

•

Less-than-or-equal (LE) binary operator
The less-than-or-equal operator is TRUE if and only if the
first operand is less than or equal to the second operand.
For example, the following expression:
'Z I

LE 'A'

produces the logical value FALSE.

5-12

CONTROL FILE LANGUAGE GUIDE
•

String-matching (MATCH) binary operator
The MATCH operator compares strings. The strings are examined
to determine which is shorter. The shorter string is compared
character-by-character to the
longer
string.
If
all
characters in the shorter string match with characters in the
longer string, then the strings are equal and the value is
TRUE.
This means a null string always matches any other
string. For example:
'ABCDEF' MATCH 'AB'
produces the logical value TRUE and the expression:
'ABCDEF' MATCH 'ABCX'
produces the logical value FALSE.

5.2.4.4 Numeric Operators - Numeric operators operate on
variables as unsigned longwords. The numeric operators are:
•

numeric

Field extraction ([n:rn]) suffix unary operator
The field extraction operator produces the longword formed by
taking the m-bit field beginning at bit n in the longword.
Bit positions are numbered from least significant to most
significant, beginning with O. For example, the expression:
#W0'357'

[3:6]

produces the octal 35.
Octal 357 has the binary value
011101111.
Bit three and the next six bits have the binary
value 011101, or octal 35.
•

Logical SHIFT binary operator
The SHIFT operator produces the first operand shifted by the
number of bit positions specified by the second operand. Each
left shift of one bit is the equivalent of multiplying by 2
and each right shift of one bit is the equivalent of dividing
by 2.
Indicate a left shift by making the second operand
positive, and a right shift by making the second operand
negative.
If the second operand is zero, nothing shifts. The
shifting is logical shifting; there is no sign extension on
right shifts.
For example the following expression:
#WD'205' SHIFT 2
produces decimal 820, which is decimal 205 multiplied by 4.

•

Multiplication (*) binary operator
The multiplication operator produces the product of the two
operands.
The result of the multiplication is truncated to
the 32 low-order bits.
For example, the expression:
5

* 3

produces decimal 15.

5-13

CONTROL FILE LANGUAGE GUIDE
•

Division (/) binary operator
The division operator produces the integer quotient of the two
operands.
For example, the expression:
15 I #B'2'
produces decimal 7.

•

Modulus (MOD) binary operator
The MOD operator produces the
division of the two operands.

remainder

the

integer

zero

the

For example, the expression:
15 MOD 2
produces decimal 1.
NOTE
A division or modulus operation
divisor causes an error.
•

with

Addition (+) binary operator
The addition operator produces the sum of the
The sum is truncated to the 32 low-order bits.

two

operands.

For
5 + 12

produces decimal 17.
•

Subtraction (-) binary operator
The subtraction operator, or minus, produces the difference of
the two operands.
The difference is truncated to the 32
low-order bits.
For example, the expression:
12 - 3

produces the decimal value 9.
•

Negation (-) unary prefix operator
The negation operator, or minus, produces the two's complement
of the operand. For example, the expression:
- f:B'8'

produces the decimal value -8.
NOTE
The minus is both a unary and binary operator.

5-14

CONTROL FILE LANGUAGE GUIDE
Other numeric operators perform bitwise logical operations between two
numeric operands. That is, rather than comparing the numeric operands
as numbers, these operators compare the numeric operators bit by bite
Here are the bitwise logical operators:
•

Bitwise AND binary operator
The bitwise AND operator produces the bitwise logical AND
the two operands. For example, the following expression:

#B0'41' AND #B0'3'
produces the octal value 1. The binary value of octal 41 is
00100001 and the binary value of octal 3 is 00000011. The
bitwise AND operation determines that the least significant
bit is set in both operands and returns the binary value
0000001, or octal 1.
•

Bitwise OR binary operator
The bitwise OR operator produces the bitwise logical OR of the
two operands. For example, the following expression:
#B0'41' OR #B0'3'
produces the octal value 43. That is, the bitwise OR of the
binary values returns the binary value 00100011, or octal 43.

•

Bitwise Exclusive-OR (XOR) binary operator
The bitwise
Exclusive-OR
expression:

XOR operator produces the
bitwise
logical
of the two operands. For example, the following

#B0'41' XOR #B0'3'
That
is,
the
bitwise
produces the octal value 42.
exclusive-OR of the binary values returns the binary value
00100010, or octal 42.
•

Bitwise complement (NOT) unary operator
The bitwise NOT operator produces the bitwise complement
(logical negation) of the operand. For example, the following
expression:
NOT #B0'41'
produces the octal value 336. That is, the binary value of
octal 41 is 00100001 and its complement is 11011110, or octal
3360.

5.2.5

Operator Precedence

Operations occur in the order defined by operator precedence unless
overridden using parentheses. Operator precedence in CFL is the same
as in most other languages, such as FORTRAN.
Operators with higher
precedence are evaluated before operators with lowec precedence. For
example, the expression:
A + B

* C

5-15

CONTROL FILE LANGUAGE GUIDE
is evaluated as A + (B * C) rather than (A + B)
* C, because the
multiplication operator, *, has higher precedence than the addition
operator, +.
In general expressions are evaluated from left to right,
taking
into
account operator precedence,
unless overridden by
parentheses.
Operators are classified into the categories shown in the following
table,
listed in order of decreasing precedence.
The order in which
operators are listed within a category is not significant.
Highest precedence class - prefix/suffix unary operators:
NOT
[n :m]
[ n]

<n:m>
<n>

Numeric negation
Logical or numeric bitwise negation
Numeric field extraction
Logical value extraction
Substring extraction
Element Extraction

Multiplication precedence class - numeric binary operators:

SHIFT
MOD

Numeric multiplication
Numeric division
Numeric logical shifting
Numeric modulus

Addition precedence class - numeric binary operators:

Numeric addition
Numeric subtraction

Logical operation class
AND
OR
XOR

Logical and bitwise logical AND
Logical and bitwise logical OR
Logical and bitwise logical XOR

Relational class EQ
NE
GT
GE
LT
LE

5. 3

- logical/bitwise logical operators:

logical comparison operators:

Equality
Inequality
Greater than
Greater than or equal to
Less than
Less than or equal to

FUNCTIONS

Functions provide special computations or special values not otherwise
available to a program written in CFL.
A reference to a function in an expression has the format:
% <function

(argument [ = value ] ,

•••

Function names have the format % <class name>$<function name>.
Some
functions require arguments. Functions return a value of a type that
is fixed for a given function.
For example, the following function:
%PKT$IDENT
returns the identification code of the current data packet as an ASCII
string.

5-16

CONTROL FILE LANGUAGE GUIDE
The following sections list by class name the CFL
values they returno

5.3.1

functions

and

the

%CND Functions - Conditional Functions

The %CND functions select one of a set of expressions for evaluation.
You can state criteria to select one of the arguments to be evaluated
in a given context.
NOTE
All expressions are evaluated before
determining a
result.
This means all
expressions must be valid
for
any
possible
value
of
the
~ogical
expression.
That
is,
%CND$IF 1s not
entirely equivalent to an IF-THEN-ELSE
statement, and
%CND$SELECT
is
not
entirely
equivalent
to
a
SELECT
statement.
The functions are:
•

%CND$IF(<logical>,<true exp>,<false exp>)
Evaluates the specified logical expression~
If the expression
is true,
the true expression is returned as the value of the
function.
If the expression is false, the false expression is
returned as the value of the function.

•

%CND$SELECT(<selector>,<exp else>,<exp O>,<exp l>, ••• )
Evaluates the specified selector expression.
If the value of
the expression is zero, the exp 0 expression is returned as
the value of the function.
If the value of the expression is
one,
the exp 1 expression is returned as the value of the
function.
In general, if the value of the selector expression
is n,
the value of exp n is returned as the value of the
expression.
If no expression is provided corresponding to the
value of the selector expression, the value of exp else is
returned as the value of the expression.

5.3.2

%CNV Functions - Conversion Functions

The %CNV functions convert expressions to ASCII strings. This is done
primarily for printing.
The conversions allow specification of the

output radix, leading fill character (if any), and

number

digits

converted.

5.3.2.1 %CNV Functions - Numeric Conversion Functions - The numeric
conversion functions convert numeric values to ASCII strings in the
radix of the specific function.
The syntax of these functions
is as
follows:
%CNV$xxx(<numeric_value>[,<field_width>[,<fill_character>]])

5-17

CONTROL FILE LANGUAGE GUIDE
where xxx is the radix.
If no field width is specified, the default
is O.
If no fill character
is specified, the default is the null
character.
The field-width and the fill character control the length
of the returned string and justification-of the digits in the string.
The numeric conversion functions behave as follows.
CFL converts the
numeric value to an ASCII string using the appropriate radix, and
calculates the number of resulting digits.
The following algorithm
formats the returned string.
if field width

= 0 then

return a string of length number_of_digits containing only the
converted digits
else
if number_of_digits > field width then
return a string of length field width filled with asterisks
else
if fill character

= null character then

return a string of length field width with the digits left
justified and pad the string with trailing blanks
else
return a string of length field width with the digits right
justified preceded by the specified fill_character
•

%CNV$0CTAL(<numeric_value>[,<field_width>[,<fill_character>]])
Converts the number from binary to ASCII octal representation.

= %CNV$DECIMAL(<numeric_value>[,<field_width>[,<fill_character>jjj
the
Converts
representation.
•

number

from

binary

ASCII

decimal

%CNV$DECIMAL_P(<numeric value>[,<field width>[,<fillcharacter>]])
This function is identical to %CNV$DECIMAL function,
except
that it appends a decimal point to the end of the output ASCII
string.
The decimal point is not counted in the field width.

•

%CNV$HEX(<numeric_value>[,<field_width>[,<fill_character>]])
Converts the number
representation.

•

from

binary

ASCII

%CNV$BCD(<numeric_value>[,<field_width>[,<fill_character>]])
This function is identical to %CNV$HEX. Converts
from binary to ASCII hexadecimal representation.

•

hexadecimal

the

number

%CNV$BINARY(<numeric_value>[,<field_width>[,<fill_character>]])
Converts
the
representation.

number

from

5-18

binary

ASCII

binary

CONTROL FILE LANGUAGE GUIDE
•

%CNV$MACHINE(<numeric_value>[,<field_width>[,<fill_character>]])
Converts the number from binary to ASCII representation in the
natural machine radix, which is octal for the PDP-11 and
hexadecimal for the VAX-11.

•

%CNV$RAD50(<numeric value>[,<field_width>[,<fill_character>]])
This function converts a numeric type to an ASCII string using
Radix-50 conversion.
The numeric value must be a word,
longword, or quadword.

5.3.2.2 %CNV Functions
Miscellaneous Conversion
number of other conversion functions are available:
•

Functions - A

%CNV$STRING(<string>)
Performs a hexadecimal conversion of the specified
ASCII •

•

string

%CNV$RSX_TIME(<RSX time value>[,<fields>])
Converts the RSX time value to a string of the format:
yy-mrnm-dd hh:mm:ss.
The optional fields numeric parameter
specifies the number of fields of the date to be converted.
To convert only the date, specify 3. To convert the date and
time, exclusive of the seconds, specify 5.
The default is
full date and time expressed in six fields.

•

%CNV$VMS_TIME(<VMS time value>[,<fields>])
Converts the VMS time value to a string of the format:
yy-mmm-dd hh:mm:ss.
The optional fields numeric parameter
specifies the number of fields of the date to be converted.
To convert only the date, specify 3.
To convert the date and
time, exclusive of the seconds, specify 5.
The default is
full date and time expressed in six fields.

5.3.3

%COD Functions - Encoding Functions

The encoding functions convert ASCII strings into numeric
Various functions do the conversion using different radixes.
•

values.

%COD$0CTAL(<string>)
Converts the string to a VALUE using octal radix. The string
may contain only the digits 0 through 7 and optional leading
spaces or the minus (-).

%COD$DECIMAL(<string>)
Converts the string to a VALUE using decimal radix.
The
string may contain only the digits 0 through 9 and optional
leading spaces or the minus (-).

%COD$HEX(<string>}
Converts the string to a VALUE using hexadecimal radix.
The
string may contain only the digits 0 through 9 and A through
E, and optional leading spaces or the minus (-).

5-19

CONTROL FILE LANGUAGE GUIDE

%COD$BCD(<string>)
Same as %COD$HEX.
Converts the string to a VALUE using
hexadecimal radix.
The string may contain only the digits 0
through 9 and A through E, and optional leading spaces or the
minus (-).

•

%COD$BINARY(<string>)
Converts the string to a VALUE using binary radix. The string
may contain only the digits 0 and 1, and optional leading
spaces or the minus (-).

%COD$MACHINE(<string>)
Converts the string to a VALUE using the natural radix for the
MACHINE, which is OCTAL for the PDP-11 and HEX for the VAX-11 •

•

%COD$RSX_TIME(<string>)
Converts the string to a date in RSX format.
The string must
be of the form dd-mmm-yy [hh:mm[:ss]].
The date and time can
occur in either order;
the seconds (:ss) are optional.
The
default for the time fields is 00:00:00.

%COD$VMS_TIME(<string>)
Converts the string to a date in VM~ rormat.
The string must
be of the form dd-mmm-yy hh:mrn:ss.
The date and time can
occur in either order;
the seconds (:ss) are optional.

5.3.4

%COM Functions -Computational Functions
•

%COM$AND(<numeric expression> , <numeric expression> )

Both
expressions must be machine VALUEs or shorter. This function
is used primarily for overlay reasons on the PDP-11.
•

%COM$HARDWARE( <numeric expression> )
Returns the ASCII character corresponding to the numeric
expression
in
the
DEC
hardware
alphabet,
which is
ABCDEFHJKLMNPRSTUV,
numbered
from
0
through
17.
A
%COM$HARDWARE( 0 ) returns an "A".

•

%COM$LONGWORD( <value> , <bit> , <value> , <bit> •••
Returns a LONGWORD value.
Each value is shifted by the
specified number of bits and then all the values are ORed •

•

%COM$NEGATE( <value>
Returns the negative of the specified
two's complement of the value.

•

value.

This

the

%COM$NULL(<expression>)
Returns a TRUE if the result of the expression is a value of
length zero, a FALSE if the result of the expression is a
value with length other than zero.

5-20

CONTROL FILE LANGUAGE GUIDE
5.3.5

%CTL Functions - RPT Control
•

%CTL$0PEN(<file>,<file spec>,<default spec>)
Opens the file using the file specification and ~ne default
file specification.
The value of the function is the fully
qualified file specification for the file.

%CTL$STATUS(<file>)
Returns the value TRUE if the file is open and
file is not open.

•

FALSE

file

system

Returns the value TRUE if the specified file is at EOF,
if the file is not at EOF.

FALSE

%CTL$EOF(<file>)

%CTL$CLOSE(<file>)
Closes the file. The value of the function is the
records written to the file.

•

the

%CTL$FILE STATUS
Returns the numeric status value returned by the
after the last file open operation.

number

%CTL$INPUT( <low> , <high>)
Sets the lowest and highest packets to be processed by RPT.
Returns a TRUE if both packet specifications are syntactically
correct and a FALSE if either is not syntactically correct.
This implicitly sets the processing direction, as well,
because if the high packet is lower than the low packet, the
file is processed backwards.
A null packet specification
takes the default, the beginning of file for the low packet
and the end of file for the high packet.

5.3.6

%LOK Functions - Lookahead Functions

The %LOK functions obtain information in undeclared data packets or
subpackets.
There is a %LOK function for each of the data types
supported for lookahead. All offsets are byte offsets. Here are the
functions:
•

%LOK$BYTE(<subpacket_number>,<offset>)
Returns the specified byte from the current data packet.
subpacket number is the number of the subpacket from which
data is to be obtained.
If the subpacket number is zero,
data is obtained from the packet itself. The offset is
byte offset in the subpacket for the data item.

•

The
the
the
the

%LOK$WORD(<subpacket_number>,<offset>)
Returns the specified word from the current data packet.
subpacket number is the number of the subpacket from which
data is to be obtained.
If the subpacket number is zero,
data is obtained from the packet itself. The offset is
byte offset in the subpacket for the data item.

5-21

The
the
the
the

CONTROL FILE LANGUAGE GUIDE
•

%LOK$LONGWORD(<subpacket_number>,<offset>)
Returns the specified longword from the current data packet.
The subpacket number is the number of the subpacket from which
the data is to be obtained.
If the subpacket number is zero,
the data is obtained from the packet itself. The offset is
the byte offset in the subpacket for the data item.

•

%LOK$LENGTH(<subpacket_number>)
Returns the length of the data in the specified subpacket.
The subpacket number is the number of the subpacket whose
length is to be returned.
If the subpacket number is zero,
the length of the data packet is returned.

%PKT Functions - Packet Information

5.3.7

The %PKT functions obtain information about the current packet:
e

%PKT$MODULE(<module name>)
Returns the value TRUE if the specified module exists
control file, and FALSE if it does not exist.

the

%PKT$IDENT
Attempts to get the next packet from the input file in the
range specified by %CTL$INPUT and makes it the current packet.
If no more packets exist within that range, a null string is
returned.
Otherwise,
%PKT$IDENT returns the current packet
identification as a fixed-length ASCII string of
eight
characters.

%RPT Functions - Report Control

5.3.8

The %RPT functions control report generation:
•

%RPT$PAGE_SIZE(<lines>)
The default page size is 57 lines of text plus headers and a
form feed.
%RPT$PAGE SIZE changes the number of lines per
page to the specified value.
If the value is zero, the page
size is infinite. The function returns the previous number of
lines per page before the function was executed.

%RPT$PAGE_DEFAULT
Returns the default number of lines per page of RPT, which
decimal 57.

%RPT$PAGE_CURRENT
Returns the current number of lines per page.

%RPT$PAGE_REMAINING
Returns the number of lines remaining on the current page.

5-22

CONTROL FILE LANGUAGE GUIDE

%RPT$LINE_SKIP(<interval>,<lines>)
Causes RPT to skip the specified number of lines every
interval number of lines.
If the interval is zero, automatic
line skipping is suppressed.
The function value is the
previous interval.

%RPT$LINE_REMAINING
Returns the number of lines remaining in the current interval.

%RPT$COMMAND
Returns the command line as a string.

%RPT$ IDE NT
Returns RPT ident as a string.

%RPT$STATUS(<status>)
Sets the exit status of RPT to the specified status value if
it is more severe than the current exit status.
If it is not
more severe, no action is taken. The actual status value is
determined by the control files using this function based on
the value given by the <status> argument.
A status value is
considered a SUCCESS, or TRUE, status if the low bit is 1, and
a FAILURE, or FALSE, statue if the low bit is 0.
The following algorithm, where NEW STATUS is the value of the
<status> argument and EXIT STATUS-is the current exit status,
is used to update the exit status:
IF NEW STATUS
THEN
BEGIN
IF EXIT_STATUS AND ( NEW_STATUS GT EXIT_STATUS
THEN
SET EXIT STATUS TO NEW STATUS
END
ELSE
BEGIN
IF EXIT_STATUS OR ( NEW_STATUS GT EXIT_STATUS
THEN
SET EXIT STATUS TO NEW STATUS
END ;
The function returns to original value of
than the potentially updated EXIT STATUS.

5.3.9

EXIT STATUS

rather

%STR Functions - String Handling

The %STR functions manipulate ASCII and binary strings:
•

%STR$TRAIL(<string>,<element>)
Removes all trailing repetitions of the specified element from
the specified string.
The value of the function is the
original string without the trailing characters.

%STR$LENGTH(<string>)
Returns the length of the specified string as a numeric.

5-23

CONTROL FILE LANGUAGE GUIDE
•

%STR$PARSE(<string>,<pointer>,<control>)
%STR$PARSE performs a simple parse by returning a pointer to
the end of the substring beginning at the specified pointer
position in the string and terminated by any of the characters
in the control string or by the end of the string.

•

%STR$QUOTE(<string>,<pointer>,<control>,<quote>)
STR$QUOTE performs a simple parse with quote characters.
STR$QUOTE works the same as STR$PARSE except STR$QUOTE handles
quote characters. The quote argument is a character string of
two characters. For clarity, the two characters should match
in some way, but this is not required.
If the first character
of the pair is encountered, checking for control characters
stops until the second character of the pair appears.
For
example,
the quote string
'<>' causes anything between a
left-angle bracket (<) and a right-angle bracket
(>)
to be
considered as "quoted" and treated as a unit.

%STR$REMAINING(<string>,<pointer>)
Returns the substring of the specified string consisting of
all characters including and following the specified pointer
position.

%STR$MATCH(<string>,<string>)
Performs an element-by-element comparison of the two strings.
The comparison continues only as long as there are elements to
compare. That is, with strings of different lengths,
the
comparison stops with the last element in the shorter string.
%STR$MATCH returns TRUE if the elements match and FALSE if
they do not.

•

%STR$SEARCH(<string>,<pointer>,<string>)
Searches the first string, beginning at the specified pointer
position,
for the second string. The pointer returns to the
position in the first string at which the second string
begins.
If the second string isn't found, %STR$SEARCH returns
a zero.

•

%STR$PAD(<string>,<paddingstring>,<lead>,<trail>)
%STR$PAD creates a new string consisting of the specified
string padded with the single-character padding string. The
lead and trail numeric expressions specify how many padding
characters you wish to lead or trail the original string.

•

%STR$FILE(<string>,<pointer>)
%STR$FILE assumes the pointer is at the beginning of file
specification.
It
returns a pointer to the character
following the file specification.
If the string pointed to is
not a valid file specification, %STR$FILE returns a zero.

%STR$UPCASE( <value> )
Returns the specified
converted to uppercase.

string

with

all

lowercase

letters

%STR$CHAR( <value> )
Returns the character corresponding to the specified value.

5-24

CONTROL FILE LANGUAGE GUIDE
5.3.10

%TIM Functions - Time Handling

The %TIM functions manipulate times.
Remember that
include both date and time unless otherwise noted.
functions:

the time values
Here are the %TIM

RSX Time Functions:

%TIM$RSX_CURRENT
Returns the current date and time as a value in RSX format.

%TIM$RSX_DATE(<RSX time value>)
Returns the date only in RSX format.

•

%TIM$RSX_VMS(<VMS time>)
Returns an RSX time value corresponding to the
time.

specified

VMS

%TIM$RSX_NULL
Returns a null RSX time value=
spaces.

This value prints as all blank

VMS Time Functions:
e

%TIM$VMS_CURRENT
Returns the current date and time as a value in VMS format •

•

%TIM$VMS_DA'rE(<VMS time value>)
Returns the date only in VMS format.

%TIM$VMS_PLUS(<VMS time>,<days>)
Returns a VMS time value containing the
the specified number of days.

specified

time

plus

Returns a VMS time value containing the specified
the specified number of days.

time

minus

%TIM$VMS_MINUS(<VMS time>,<days>)

%TIM$VMS_RSX(<RSX time>)
Returns a VMS time value corresponding to
time.

specified

RSX

%TIM$VMS_NULL
Returns a null VMS time value.
spaces.

5.3.11

the

This value prints as all blank

%USR Function - User I/O Function

The %USR function performs input and output to and from
RPT:

5-25

the

user

CONTROL FILE LANGUAGE GUIDE
•

%USR$STRING(<file>,<prompt>,<maximum length>)
Writes out the prompt string if the specified file is a
terminal, and reads a string input whose maximum length is
specified by the length parameter.
If the specified output
file is not a terminal, there is no prompt and only the read
is performed.

5.4

DECLARATIONS

This section describes the declaration of variables and data items. A
declaration includes print-formatting information along with the
definition of data items; it is different from declaration in most
languages.

5.4.1

Scope of Declarations

Data items can be referenced during the scope of the declaration, that
is, from the point they are declared until the declaration is
discarded.
If a declaration is made in a given procedure, data items
can be referenced in the defining procedure or any procedure called by
it. What about coroutine interactions.

5.4.2

DECLARE Statement

The DECLARE statement begins the declaration of a block of
Here is the format:

variables.

DECLARE <group name> [ NAMED ] ;
<variable name>
<type> [: <print expressions>]
<variable name>
<type> [: <print expressions>]
<variable name>
<type> [: <print expressions>]
END_DECLARE ;
The group name is the prefix name by which the group variables are
referenced.
Each of the group variables is referenced using the name:
<group name>.<variable name>
The variable type is one of the RPT data types:
ASCII, NUMERIC, FIELD, RSX_TIME or VMS TIME.

LOGICAL,

STRING,

The optional NAMED qualifier specifies that the symbol names are to be
kept and used with the WRITE GROUP statement FORMAT clause qualifiers
that print a symbol name.
The optional print expressions specify expressions to be evaulated and
printed if the group is printed using the WRITE GROUP statement.
If
you specify more than one expression, separate them by commas.
The print information consists of one or more expressions separated by
commas.
If the WRITE GROUP statement is used to print the data group,
the print expressions-are evaluated and printed.
Declaration of a numeric type can be followed by declaration of one or
more field types.
The field types are considered fields of the
preceding numeric type declaration.

5-26

CONTROL FILE LANGUAGE GUIDE
Print information for field types is handled specially. The variable
names for field types do not appear when the WRITE GROUP statement is
used, and the print expressions following a field type declaration are
considered to apply to the preceding numeric type.
Of the print
expressions following a field type, one is selected based on the field
value.
For BIT fields, the first print expression is used if the BIT is TRUE,
and the second if the BIT is FALSE.
For FIELD fields, the first print expression is used if no other print
expres~ion
applies.
The second print expression is used if the FIELD
value is zero, the third if the field value is one, the fourth if the
field value is two, and so one
When a print expression is printed for a field type, it is printed
the following format:
p[h:l]

ttt ••• ttt

using the following symbols:
p -

The leading character of the print expression.
as a prefix to the print field.

This appears

h -

The high bit number of the FIELD, or the bit number
BIT.
The brackets are printed.

the

1 -

The low bit number of the FIELD.
The "l" field
leading colon are printed as blanks for BITs.

and

the

t -

The trailing characters of the print expression.
The
trailing characters are any characters following the first
character.
For example, take the following definition:
DECLARE EXAMPLE
VARIABLE 1
FIELD [6:2]

WORD :
'*The value of this field

3I 1
I
I

The value of this field is 0'
The value of this field is l'

When EXAMPLE is printed with a WRITE GROUP statement,
the
field will be printed as follows, depending on whether bits
6 through 7 contain the value 0, 1, 2, or 3:
7:

The value of this field is 0

The value of this field is 1

* [ 7:

The value of this field is 2 or 3
NOTE

The field is declared in the form [6:2],
meaning that it starts at bit 6 and is
two bits long.
However,
the
print
format is expressed
in the form [ 7:
6], meaning that it consists of bits 6
through 7.

5-27

CONTROL FILE LANGUAGE GUIDE
5.4.3

PACKET Statement

The PACKET statement declares an
format:
PACKET

<name>
<name>
<name>
<name>
END_PACKET ;

REPEATED
<type>
<type>
<type>

input

data

packet.

Here

the

] [ NAMED ] ;
<print information>
<print information>
<print information>

The REPEATED data attribute is optional.
A PACKET without this
attribute specifies a single packet. The REPEATED specifies that the
data in the packet is repeated. The number of repetitions is computed
by dividing the packet length by the length of the data items.
Note
that the items must be referenced as for a DYNAMIC TABLE; they cannot
be referenced directly.
The optional NAMED qualifier specifies that the symbol names are to be
kept and used with the WRITE GROUP statement FORMAT clause qualifiers
that pr.int a symbol name.
Declaration of data defines the special variable LENGTH, referenced as
the data items themselves would be referenced:
<data group name>.LENGTH
Note that the
data group.

Each of the data item names is declared, along with the type and print
information.
The name is the name by which the data element is
referenced.

is Lhe

t:: - .... -- -

L U Lill d l.

....

L U L

.:I -

U d l. d

-- -

I::: J. l:::llll::: lJ l.

...-

~ -

....- -

.L I::: .1.. I::: L I::: lJ I,,; I::: o

<data group name>.<data element name>
The data item type is declared as specified
Types.

Section

5.2.1!

Data

The special variable name RESERVED in place of an element declaration
specifies a sequence of undefined values.
The type declaration
specifies the length of the undefined area. This cannot be a field
type. The syntax is as follows:
PACKET

<name> <data organization> ;
<name> : <type> : <print information>
RESERVED : <type>
<name>
<type>
<print information>
END_PACKET ;

A RESERVED declaration in a PACKET or SUBPACKET indicates to the
compiler
(and RPT) that the area is currently unused, but to use its
length in determining the size of the PACKET or SUBPACKET and the
offsets of elements following the RESERVED declaration. Use RESERVED
either to reserve space for future use or to force word-boundary
alignments.
Note that an element name could be used in these situations, but that
RESERVED serves as a documentation aid, and saves having to define
unique element names if there are multiple unused areas in a packet or
subpacket.

5-28

CONTROL FILE LANGUAGE GUIDE
5.4.4

SUBPACKET Statement

The SUBPACKET statement declares an input data SUBPACKET.
format:
SUBPACKET <name> = <expression> <data attribute>
<name>
<type> [: <print information>]
<name>
<type> [: <print information>]
<name>
<type> [: <print information>]
END_PACKET ;

Here is the
NAMED ]

The attribute, if present,
is REPEATED.
A SUBPACKET without an
attribute
specifies
a single subpacket.
The leading REPEATED
attribute specifies that the data in the subpacket is repeated.
The
number of repetitions is computed by dividing the subpacket length by
the length of the data items.
Note that the items must be referenced
as for a DYNAMIC_TABLE; they cannot be referenced directly.
The optional NAMED qualifier specifies that the symbol names are kept
and used with the WRITE GROUP statement FORMAT clause qualifiers that
print a symbol name.
The handling of SUBPACKET is otherwise the same as for PACKET.

5.4.5

Conditional Declarations

RPT provides a mechanism for conditional declaration of data items.
Conditional declaration can only be used for data, that is, PACKET and
SUBPACKET declarations.
FIELD and BIT declarations cannot cross
conditionals.
All FIELD and BIT declarations must be in the same
conditional as their data item.
The conditionals allowed in declarations are as follows:

IF
IF has the following syntax:
<name>
<name>
<name>

<type> [: <print information>]
<type> [: <print information>]
<type> [: <print information>]

IF <expression>
THEN
BEGIN
<name>
<name>
<name>
END
ELSE
BEGIN
<name>
<name>
<name>
END
END IF
<name>
<name>
<name>

<type> [: <print information>]
<type> [: <print information>]
<type> [: <print information>]

If the expression is TRUE, the THEN clause is defined.
If the
expression is FALSE,
the ELSE clause is defined. The ELSE
clause is optional.
5-29

CONTROL FILE LANGUAGE GUIDE
e

CASE
CASE has the following syntax:
<name>
<name>
<name>

<type> [: <print information>]
<type> [: <print information>]
<type> [: <print information>]

CASE <expression> OF
<expression> , <expression> ,
BEGIN
<name>
<name>
<name>
END

<type> [: <print information>]
<type> [: <print information>]
<type> [: <print information>]

<expression>
BEGIN
<name>
<name>
<name>
END
[<expression>

••• ]

<expression> ,

••• ]

... ]

BEGIN
<name>
<name>
<name>
END

<type> [ : <print information>]
<type> [ : <print information>]
<type> [ : <print information>]

BEGIN
<name>
<name>
<name>
END

<type> [ : <print information>]
<type> [ : <print information>]
<type> [ : <print information>]

r.1r rt'C'
i.:.u..;;Ji.:.

END CASE
<name>
<name>
<name>

<type> [: <print information>]
<type> [: <print information>]
<type> [: <print information>]

The expression in the CASE statement is evaluated and the
expression lists searched to find a matching expression value.
If a match is found the declaration is made.
If no matching
expression value is found, the optional ELSE clause is
executed. Otherwise, an error occurs.

5.5

ACTION STATEMENTS

Action statements perform processing. CFL has a limited set of action
statements because it is a simple, special-purpose language.
The
statements provided have capabilities designed to make the handling of
error log data as simple as possible.

5-30

CONTROL FILE LANGUAGE GUIDE
5.5.1

SET Statement

SET sets the value of a variable to
Here is the format:

the

results

computation.

SET <variable name> TO <expression> ;
The expression is evaluated using the type of the specified
and the variable is set to the value of the expression.

5.5.2

variable,

INCREMENT and DECREMENT Statements

INCREMENT and DECREMENT adjust the value of numeric variables of
length VALUE or less by a value.
The value defaults to 1. The format
of the statements are:
INCREMENT <variable name>

BY <numeric expression>

DECREMENT <variable name>

BY <numeric expression>

The value of the variable is increased bi the value of the numeric
expression for the INCREMENT statement, and decreased by the value of
the numeric expression for the DECREMENT statement.

5.5.3

WRITE Statement

WRITE writes information to a specified ouput.
WRITE <expression> , •••

Here is the format:

TO <output> FORMAT <format> ;

The expressions are printed in the order specified.
The optional
output clause can contain either the specifications REPORT or ERROR.
ERROR directs output to the invoking terminal. REPORT is the default
and directs output to the output report.
The REPORT file must be
open. The ERROR file is always open.
The format is described in
section 5.10, Print Formatting.

5.5.4

WRITE GROUP Statement

WRITE GROUP writes a decoded block of data.
define the formatting.
Here is the format:

The

data

definitions

WRITE_GROUP <group name> TO <output> FORMAT <format> , <format> ;
The group name is the name of the group of variables or data items to
be written. The optional output specification is the same as for the
WRITE statement, and has the same defaults. The format is described
in Section 5.10, Print Formatting.
The group name can be followed by a symbol list:
<group name> <symbol name>,<symbol name>, •••
In this case only the specified symbols are listed.
The first FORMAT clause is for printing all data
items,
FORMAT clause is for printing all BIT and FIELD items.

5-31

the

second

CONTROL FILE LANGUAGE GUIDE
DECODE Statement

5.5.5

The DECODE
translation.

statement
performs
specialized
The statement has the form:

declaration-to-text

<variable name> = <group name>
<bit number>
<data item>
<bit number>
<data item>
<bit number>
<data item>
END_DECODE ;

DECODE

Each of the data items must be a data item in the specified group.
The bit numbers are numbers of bits in the data item.
The DECODE
statement processes the data items in the order specified, checking
the specified bits.
If a bit is found to be TRUE, the corresponding
bit-to-text translation for that bit is performed, and the result
returned
in the specified variable.
This completes the statement. A
data item can be preceded by a NOT, which indicates the specified bit
must be FALSE for the bit-to-text translation to be performed.
If no
bit-to-text translation is found, the null string is returned.

5.6

CONTROL STATEMENTS

Control statements direct RPT.
Control statements define and
invoke
procedures and control termination of the procedure. Other control

statements also conditionally control the execution cf statements in a
procedure.

These are called conditional statements.

5.6.1

MODULE Statement

•
• " " ' rT r n
l"!VLJULl.C:.

~ - - ,J.ClJ
- - .. C;:)
- UC\,.;

L\..UC
1..... -

-UQlllC
- - -

- .C
VJ.

~
\.... \..l!C

-1uv1
- ~
•• ,J..C
...nA

\...
- J..
~ -H'::j
IJC

- -

,, __ _

.: ,.I.CU.
- ...:2

"-Vlll~J..

llCJ..C

format:
MODULE <name> <ident> ( attribute , attribute
MODULE must be the first statement in any module.
end with an END_MODULE ; statement.
The module name specifies the name to be
inserted into the control file library.
The module ident is a quoted string
header.

used
be

...

Each

module

must

when

the

module

inserted

the

The optional module attributes specify information to
processing the module. Two attributes are recognized:

;

)

module
used

KEEP
KEEP specifies that if a module cache
is used by RPT,
the
module should be kept because it is likely to be used again.

FLUSH
FLUSH specifies that if a module cache is used by RPT,
the
module should be flushed because it is unlikely to be used
again.

5-32

CONTROL FILE LANGUAGE GUIDE
5.6.2

LITERAL Statement

LITERAL assigns a name to a literal value. All LITERAL statements in
a module must precede any PROCEDURE statement.
The format of a
LITERAL statement is:
LITERAL <group>.<name>

= <compiletime constant expression> ;

The name is equivalent to the value of the compiletime constant
expression, and can be used in any expression or compiletime constant
expression to represent the specified value.

5.6.3

CALL Statement

CALL invokes a subroutine.

CALL has the following format:

CALL [ MODULE <module name expression> ]
PROCEDURE <procedure name expression>
[ COROUTINE <procedure name expression> ] ;
The optional module name expression specifies the module to be called.
If the module name is not specified, the specified procedure is
assumed to be in the current module. The procedure name specifies the
procedure to call.
The optional COROUTINE argument specifies that the two
called
procedures are coroutines. The first called procedure is specified as
in the normal form of a CALL statement. The procedure specified using
the COROUTINE keyword is executed first.
That procedure can then
execute a statement that passes control to the other procedure
specified in the CALL. None of the declarations are lost.
The two procedures can trade control back and forth using the
COROUTINE statement. Each time a COROUTINE statement is executed, the
other procedure resumes execution from the point of the last COROUTINE
statement.
If one of the procedures returns, control passes to the
other as if a COROUTINE statement were executed.
When both have
completed, the coroutines exit to the caller.

5.6.4

RETURN Statement

RETURN forces a return from a
RETURN is optional at the
RETURN statement is:

procedure to the calling procedure.
end of a procedure. The format of the

RETURN ;
This terminates the current
calling procedure.

5.6.5

procedure

and

returns

control

the

PROCEDURE Statement

PROCEDURE declares the beginning of a procedure.
format:
PROCEDURE <name> <statement block>

5-33

It has the following

CONTROL FILE LANGUAGE GUIDE
The procedure name cannot be more than
include the characters A through Z,
dollar sign ($),and the underscore ( ).
name must be alphabetic.
-

15 characters.
Names can
the numbers 0 through 9, the
The leading character of a

The statement block is executed as the named procedure.

5.6.6

IF-THEN-ELSE Statement

IF-THEN-ELSE is the most basic
conditional
statement.
Other
conditional statements are provided to simplify the handling of common
situations that would be
cumbersome
with
IF-THEN-ELSE.
The
IF-THEN-ELSE has the following format:
IF <logical expression> THEN <block>
[ELSE <block>;] END_IF ;
If the logical expression is TRUE,
the block following the THEN
statement is executed.
If the logical expression is FALSE, the block
following the ELSE statement is executed.
The ELSE clause is
optional.
If it is not specified, no action is performed if the
expression is FALSE.
Each block consists of a single statement. Using BEGIN-END, a block
can contain a compound statement. See Section 5.6.11 on BEGIN-END
statements for a description of how to use BEGIN-END statements to
make multiple statements appear as a single logical entity to
conditional statements.

5.6.7

CASE Statement

CASE selects one of a set of possible outcomes based on an expression.
The format of a CASE statement is as follows:
CASE <expression> OF
[ <expression>
<expression>
<expression>
[ <expression>
[ <expression>
<expression>
[ELSE <block> ;]
END CASE

, ...
I

...
• • •

]
]
]

This executes the block corresponding to the first expression equal to
the selector numeric expression.
If ELSE is specified, it is executed
if no expression matches.

5.6.8

SELECT Statement

SELECT is a special case of CASE.
SELECT selects one of a
of blocks.
The general format is:
SELECT <numeric expression> OF
<block>
<block> ;
<block> ;
[ELSE <block> ;]
END_SELECT;

5-34

given

set

CONTROL FILE LANGUAGE GUIDE
SELECT selects the nth block, where n is the value of the numeric
expression and is greater than or equal to 1.
If the last block is
preceded by ELSE, the block is executed if and only if the value of
the numeric expression exceeds the number of blocks supplied.

5.6.9

WHILE/UNTIL/DO Statements

The WHILE/UNTIL/DO statements control conditional looping. To specify
a conditional loop, specify a block of statements to be conditionally
executed and an expression to control the execution.
The DO statement specifies the block of statements to be conditionally
executed.
The WHILE statement specifies an expression to be considered satisfied
if it is TRUE.
The UNTIL statement specifies an expression to be considered satisfied
if it is FALSE.
A DO statement must be specified with a WHILE or UNTIL statement. The
block of statements specified by the DO statement is executed until
the condition specified by the WHILE or UNTIL statement is no longer
satisfied.
If DO WHILE or DO UNTIL is specified, the DO is executed once before
testing the condition.
If WHILE DO or UNTIL DO is specified, the
condition is tested first before executing the DO statement.
This
leads to the following statement combinations:
DO <block> WHILE <expression> ;
Executes the block once, then evaluates the expression.
If the
expression is TRUE, the block is repeated.
1I ~ne expression is
FALSE, execution continues following the WHILE statement.
DO <block> UNTIL <expression> ;
Executes the block once, then evaluates the expression.
If the
expression is FALSE, the block is repeated.
If the expression is
TRUE, execution continues following the UNTIL statement.
WHILE <expression> DO <block>
Evaluates the expression. If the expression is TRUE, the block is
executed and the process repeated.
If the expression is FALSE,
execution continues following the DO statement.
UNTIL <expression> DO <block>
Evaluates the expression. If the expression is FALSE, the block is
executed and the process repeated.
If the expression is TRUE,
execution continues following the DO statement.

5.6.10

LEAVE Statement

LEAVE immediately terminates the current DO statement.
The control
expression in the associated UNTIL or WHILE statement is considered
satisfied and is not reevaluated.

5-35

CONTROL FILE LANGUAGE GUIDE
5.6.11

BEGIN-END Statement

BEGIN-END statements force a compound statement to be treated as one
statement for purposes of conditionals. For example, to process two
statements in the THEN clause of an IF statement, use the the
following construct:
IF <logical expression>
THEN
BEGIN
<statement l>
<statement 2>
END ;
ELSE
<statement>
END IF

5.6.12

Lexical Conditionals

Lexical conditionals perform conditional handling at compilation.
A
lexical conditional is valid wherever a statement is valid.
Lexical
conditionals have the following format:
$IF <compiletime constant expression>
$THEN
<stateme~t block>
[$ELSE
<statement block>]
$END_IF
The $ELSE block is optional.
If the compiletime constant expression
is TRUE, everything in the $THEN block is compiled and the $ELSE block
is not compiled.
If the compiletime constant expression is FALSE, the
$THEN block is not compiled and the $ELSE block, if present, is
compiled.
Lexical conditionals can be nested to any level.

5.7

TABLES

The table is one of the fundamental units of data organization for
RPT.
RPT uses tables to structure large amounts of data to be
referenced during report generation.

5.7.1

Table Structure

Tables are sets of similar records containing fields by which the
records can be referenced.
Tables, and the data in them, can either
be declared statically as part of the definition of a given control
file module, or dynamically during the operation of RPT.
Static
tables hold reference data, while dynamic tables store information
computed during the operation of RPT.
Each record in a table is a sequence of named fields.
The definition
of the table defines the names of the fields and their sequence.
Reference tables by name. Table names follow the ordinary rules for
naming groups.
The name cannot be more than 15 characters.
Names can

5-36

CONTROL FILE LANGUAGE GUIDE
include the characters A through Z, the numbers 0 through 9,
the
dollar sign
($), and the underscore ( } • The leading character of a
name must be alphabetic.
Fields in a table are also named.

Field names follow the same rules.

Table entries are manipulated by setting the current entry pointer for
a table, and then using either the table manipulation statements or
simple variable references to read or modify the data in the table.
The following sections describe each of
table-manipulation statements in detail.

5.7.2

the

table-definition

and

TABLE Statement

TABLE defines a static table.

The format of the statement is:

TABLE <table name> ;
<name>
<type> [: <print expressions>]
<name>
<type> [: <print expressions>]
<name>
<type> [: <print expressions>]
BEGIN TABLE
<value>
<value>
<value> ,
<value> I • • •
<value>
<value>
<value>
<value>
<value> I • • •
END_TABLE ;

...

The declaration list following the TABLE statement specifies each of
the fields,
their types, and print information. The format is the
same as for DECLARE. The list values are individual compile-time
constant expressions separated by commas.
Each sequence of list
values separated by commas (,)
and terminated by a semicolon
(;)
represents one TABLE record.
Each table value must be of the same
type as the corresponding declaration from the declaration list.
TABLE records cannot be modified at run time.

5.7.3

DYNAMIC TABLE Statement

DYNAMIC TABLE declares a dynamic table.
statement is as follows:
DYNAMIC TABLE <table name>
<name>
<type> [:
<name>
<type> [:
<name>
<type> [:
END_TABLE ;

The format of a DYNAMIC TABLE

;
<print expressions>]
<print expressions>]
<print expressions>]

Records are placed in the DYNAMIC TABLE at run time through use of the
PUT statment, and can be modified-by some of the POINTER statements.

5.7.4

FILE Statement

FILE is identical to DYNAMIC TABLE.
It is included for compatibility
only and you should use -DYNAMIC TABLE. The FILE statement may be
removed in a future release.

5-37

CONTROL FILE LANGUAGE GUIDE
FILE declares a dynamic table.
follows:

The format of a FILE statement

FILE <table name> ;
<name>
(type> [: <print expressions>]
<name>
<type> [: <print expressions>]
<name>
<type> [: <print expressions>]
ENDFILE
Records are placed in the FILE dynamically at run time through use of
the PUT statement and can be modified by some of the POINTER
statements.

5.7.5

POINTER Statement

POINTER adjusts the current pointer for a table
syntax:

using

the

following

POINTER <table name> <action> <optional pointer variable>
Each of the actions is described below:
•

FIRST
Sets the current table pointer to the first record of the
table.
If there is no next record, then the current table
pointer is set to null (see RESET).

•

NEXT
Sets the current table pointer to the next record of the
table.
Tf
there is no next record; then the current table
pointer is set to null (see RESET).

•

PREVIOUS
Sets the current table pointer to the previous record of the
table.
If you back up past the beginning, then the table
pointer is set to null (see RESET).

•

RESET
Sets the current table pointer to null, that is, there
table pointer.

•

LOAD <pointer variable>
Sets the current table pointer to the
variable.

value

the

pointer

CLEAR
The specified table must be a DYNAMIC TABLE.
Deletes all
records from the DYNAMIC TABLE and sets the current table
pointer to null (see RESET):

•

DELETE
The specified table must be a DYNAMIC TABLE.
Deletes the
current record and advances the pointer to the next record.
If there is no next record, then the current table pointer is
set to null (see RESET).

5-38

CONTROL FILE LANGUAGE GUIDE
•

MOVE <pointer variable>
The specified table must be a FILE. The record pointed to by
the pointer variable is moved to the current record position
and the current record and all following records are moved up
one record.
This is used mainly for sorting records in a
FILE.

5.7.6

FIND Statement

FIND finds a record in a table using one
format of the FIND statement is:
FIND <table> <f ield>=<value>

• • •

key

values.

The

SELECT <expression> ;

The table is searched until an entry with all specified fields having
the specified value is encountered. Tables are searched sequentially
from the current pointer position.
If no record is found, the current
pointer for the table is set to null.
If you specify the optional SELECT clause, a record does not satisfy
the search criteria unless the select expression, evaluated with the
current record for the table set to the specified record, is TRUE.

5.7.7

POT Statement

PUT creates a new record in a table. The specified table
DYNAMIC TABLE.
PUT has the following format:

must

, ...

PUT <table> <field>=<expression>

Sets the specified fields of the record to the values of the specified
expressions.
Note that all fields must be specified; none of the
fields of the record are optional.

5.8

LISTS

This section describes the expression-list-handling facilities of CFL.

5.8.1

LIST Statement

LIST declares a list of expression groups.
follows:

The format of LIST

LIST <list name> ;
<expression>
<expression>
<expression>

• • •

, ...
I

• • •

END_LIST ;
The expression lists can then
described in this section.

5-39

referenced

other

statements

CONTROL FILE LANGUAGE GUIDE
5.8.2

SEARCH Statement

SEARCH locates a specific entry in a LIST.
format:

SEARCH has

the

following

SEARCH <list name> <expression> , ••• , <expression>
GET <variable> , ••• <variable>
FLAG <variable> ;
The specified list is searched sequentially until an entry is found
where each of the SEARCH expressions is equal in value to the
corresponding LIST expression in the same expression list.
The
variables in the GET clause are then set to the corresponding
remaining expressions of the expression list, and the FLAG variable is
set to TRUE.
If no match is found, the variables specified in the GET
clause are unchanged, and the FLAG variable is set to FALSE.

5.9

SIGNALLING

This section describes the signalling facilities of CFL.

5.9.1

Signalling

signailing breaks the control flow in the report to handle special
conditions. Control goes to a special routine established by the user
called a handler routine. When a condition is signalled by the user
using the signalling statements, the most recently declared handler
routine is called. The handler routine can then take the appropriate
action.
Any routine can establish a handler routine.
When a condition is
signalled by the user, the user can optionally suppress the change in
the flow of control, and cause the handler to return to the routine
executing the signal.
When a condition is signalled, a message describing the event is
The message inserted
appended to the file ERROR, if the file exists.
in the ERROR file consists of a sequence of comma-delimited quoted
the
arguments to the SIGNAL-class
strings,
corresponding
to
statements.

5.9.2

ENABLE Statement

ENABLE has the following format:
ENABLE [ MODULE <expression>

PROCEDURE <expression> ;

The procedure becomes the condition handler for this procedure and all
called procedures, unless a called procedure in turn has an ENABLE.

5.9.3

SIGNAL Statement

SIGNAL has the following format:
SIGNAL <message code> PARAMETERS <expression list>

5-40

CONTROL FILE LANGUAGE GUIDE
The message code and expressions in the expression list are ASCII
strings.
When the SIGNAL statement is executed these expressions are
evaluated and the resulting ASCII strings are appended to the ERROR
file as quoted strings separated by commas, as follows:
'<message code>' ,'<expressionl>' ,'<expression2>' ••• ,'<expressionN>'
The signal-handling routine is then called. After execution of the
signal-handling procedure, execution resumes following the statement.

5.9.4

SIGNAL STOP Statement

The SIGNAL STOP statement is the same as the SIGNAL statement, except
that after execution of the signal-handling procedure, execution
resumes following the call to the procedure that executed the ENABLE
statement.
SIGNAL STOP <message code> PARAMETERS <expression list>

5.9.5

MESSAGE Statement

MESSAGE has the same format as SIGNAL.
It causes the appended string
to be placed in the ERROR file, but does not cause any signal
processing.
MESSAGE <message code> PARAMETERS <expression list>

5.9.6

CRASH Statement

CRASH causes an immediate abort of RPT. Use it in cases
handling where the signalling mechanism is inadequate.
statment has the following format:

of error
The CRASH

CRASH
CRASH causes
structures.

5.10

detailed

dump

many

RPT's

internal

data

PRINT FORMATTING

This section describes the output-formatting facilities of CFL.

The FORMAT keyword on the WRITE and WRITE GROUP statements expresses
output-formatting information. The keyword has the following syntax:
FORMAT <format string>
The format string can be any ASCII
after substitution specified by
have the format

string.
The string is output,
output directives. The directives

!n(mdd)

5-41

CONTROL FILE LANGUAGE GUIDE
where dd is the two-character directive, m is the optional argument,
and n is the optional repeat count. The parentheses need not be
included if there is no repeat count. This syntax is the same as that
for the VMS %FAO facility. A double-exclamation mark, !!, prints as a
single exclamation mark.
Multiple format strings can be specified one after the other separated
by the concatenation operator, the vertical bar <I, ASCII 174). They
are treated as one concatenated string.
The allowable directives are specified in the following sections.

5.10.1.1 Control Directives - The control directives control the
processing of the format string.
Here are the control directives:
!nCE

Repeat the FORMAT clause.

!nCF

When used with a WRITE
statement,
this
directive
terminates output if the values of all expressions have
been output. When used with a WRITE GROUP statement, this
directive terminates output if all fields in the specified
group have been output. The effect in both cases is to
terminate evaluation and output of the format string if
there are no more values to be output.

5.10.1.2 Formatting Directives - The formatting directives output
carriage-control information. Here are the formatting directives:
!nFC

Print following output beginning at column n.

!nFS

Space the current output print column forward n columns.

!nFL

Output n-1 blank lines.
Printing resumes on the line
following the blank lines.
The default, n=l, causes
output to begin on the line following the current line.

!nFP

Output a page break.

5.10.1.3 Data-formatting Directives - The data-formatting directives
control the output of data. Here are the data-formatting directives:
!nDF

Print the field name of the current output field.
The
argument specifies the field width to be used for the
name. The name is printed left-justified.

!nDP

Print the current output field.
The argument specifies the field width to be used for the
field.
For numeric fields,
the field width n must be
greater than the field width specified when the field was
defined.
If
all fields have been printed, output
terminates.

5-42

CONTROL FILE LANGUAGE GUIDE
5.11

USER INTERFACE HANDLING

This section describes the user interface to the compiler.

5.11.1

Overview of User Interface Handling

The compiler implements two user interface modes, command mode and
OPTION mode.
In OPTION mode, the compiler requests a command line,
followed by requests for OPTIONS, which are terminated when a line
beginning with a slash (/) is entered, at which point the compilation
takes place.

5.11.2

Command Mode

In command mode, the compiler requires a command line of the following
syntax:
<output>,<listing>,<symbols>=<input>,<symbols>
All files on the left of the equals (=) character are output
and all files to the right of the = character are input files.
The output file is the compiled
default file type .!CF.

module

output

The listing file is the compilation listing file.
file type .LST.

file.

It has the

files,
has

the

default

The symbols output file is the compilation symbol table, which must be
used as input to any compilation of a module to be called from this
module at execution time.
It has the default file type .SYM.
The input file is the .CNF source file.
The symbols input file is a compilation symbol table from
which calls the module being compiled at execution time.

5.11.3

the

module

Option Mode

In option mode, the compiler accepts the following option:
e

LITERAL
The LITERAL option has the form:
LITERAL group.name = value
This is the same syntax as for the LITERAL statement in the
source.
The LITERAL is declared for the duration of the
compilation. The only valid values are single items:
quoted
strings, numeric values, and logical TRUE and FALSE values.
Numeric values must be positive decimal values that are
treated as a machine values.

5-43

CONTROL FILE LANGUAGE GUIDE
5.12

ERLCFL REPORT MESSAGES

ERLCFL-F-ASCIIBIG, ASCII literal quoted string too long for type.
Explanation: An ASCII radix numeric literal in a control file
source module contains too many characters for the specified
numeric type.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-BADDIGIT, Invalid numeric digit in conversion.
Explanation: A numeric literal or the ASCII string argument for
the %COD$0CTAL, %COD$DECIMAL, %COD$HEX, %COD$BCD, %COD$BINARY, or
%COD$MACHINE function contains an illegal character for the
specified radix or was null or blank.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-BITFLDSIZ, Bit or field too large in extraction operation.
Explanation: The bit or field in an extraction operation exceeds
the size of the value on which the extraction l~ performed.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-BITNOPREC, A BIT or FIELD must have a preceding data item.
Explanation: A BIT or FIELD declaration in a control file source
module must be preceded by a data item within the declaration.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-BITNOTVAR, A BIT or FIELD not allowed on variable-length data
item.
1

Explanation: A BIT or FIELD declaration in a control file source
module is not allowed on a variable length data item.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-BITNUMINV, BIT number outside the declared data item.
Explanation: The bit number in a BIT declaration for a data item
is too large for the data item.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

5-44

submit

SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-BITTOOHIG, Bit number too large for specified storage unit.
Explanation: The bit number specified by the character string
portion of a #BI, #WI, #LI, #QI, or #VI numeric literal is too
large for the specified value size.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

not

SPR

ERLCFL-F-BYTERADSO, A BYTE data item cannot print in RADSO.
Explanation:

The print radix for a BYTE declaration can

RADSO.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-CASENOTDAT,
declaration.

declaration

clause

must

submit

data

Explanation: A declaration CASE clause attempted to declare data
but was not contained within a declaration.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-CFLINPUT, Could not open input source file.
Explanation: CFL could not open the input source file
on the command line.

specified

User Action: Check the file specification, and
you have access to the specified file.

sure

that

output

file

sure

that

output

file

make

ERLCFL-F-CFLISTING, Could not create listing output file.
Explanation: CFL could not create
specified on the command line.

the

listing

User Action: Check the file specification, and
you have access to the specified file.

make

ERLCFL-F-CFLMODULE, Could not create module output file.
Explanation: CFL could not create
specified on the command line.

the

module

User Action: Check the file specification, and
you have access to the specified file.

make

sure

that

ERLCFL-F-CFLSYMBOL, Could not open symbol file for input.
Explanation: CFL could not open the input symbol file
on the command line.

specified

User Action: Check the file specification, and
you have access to the specified file.

sure

5-45

make

that

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-CFLSYMOUT, Could not create symbol output file.
Explanation: CFL could not create
specified on the command line.

the

output

User Action: Check the file specification, and
you have access to the specified file.

file

symbol

make

sure

that

option

line

ERLCFL-F-CMDOPTERR, Option line syntax error.
Explanation:
input.

CFL encountered a syntax error on the

User Action:
file please
CFL again.

If the error occurred in a DIGITAL-supplied command
submit an SPR, otherwise correct the error and run

ERLCFL-F-CMDSPCERR, Command line syntax error.
Explanation:
input.

CFL encountered a syntax error on the command

line

User Action:
file please
CFL again.

If the error occurred in a DIGITAL-supplied command
submit an SPR, otherwise correct the error and run

ERLCFL-F-DECTOOBIG, Declaration too large, too many symbols.
Explanation:
compiled.

A declaration in a control file is too large to

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

Explanation: A control file module contains a division
in a compile-time constant expression.

zero

SPR

ERLCFL-F-D!VZERO,

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-EXPTOOBIG, Operator stack overflow.

Expression too complex.

Explanation: An expression in a control file
too complex to be compiled.

source

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-EXPTOOLAR, Operator stack overflow.

module

submit

is
SPR

Expression too complex.

Explanation: An expression in a control file
too complex to be compiled.

source

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

5-46

submit

module

submit

is
SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-FIELDBIG, FIELD exceeds size of the declared data item.
Explanation: A FIELD declaration in a control file source module
exceeds the bounds of its corresponding data item.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-FIELDBITI, FIELD starting bit is outside
item.
Explanation:

submit

the

declared

SPR

data

A FIELD declaration in a control file source module

exceeds the bounds of its corresponding data item~

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-FIELDSMAL, FIELD width must be at least one bit.
Explanation: A FIELD declaration in a control file source module
did not have a width specified.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-FILERCLOS, File close error.
Explanation:
file.

An error occurred when CFL

attempted

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLCFL-F-FILERREAD, File read error.
Explanation:
file.

An error occurred when

CFL

attempted

read

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLCFL-F-FILERWRIT, File write error.
Explanation:
file.

An error occurred when CFL attempted to write to

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLCFL-F-FILINTOPN, Internal error - file already open.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

5-47

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-FILINVCOD, Internal error - invalid file code
operation.

for

specified

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-FLUSHINV, FLUSH attribute not allowed with KEEP attribute.
Explanation: A control file source module had both the FLUSH and
KEEP module attributes specified.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-FUNFIELDS, Invalid
conversion function.

conversion

code

submit

argument

SPR

time

Explanation: A control file module contains a time conversion
function with an illegal value for the conversion code argument.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-FUNINVPOI, Invalid string pointer value in string function.
Explanation: A control file module contains a %STR$PARSE or
%STR$QUOTE function where the value of the pointer argument is
larger than the length of the string argument.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-FUNNOTCHA, Argument to STR$CHAR is not

submit

valid

SPR

range

for

..,..\...---.-.1---

.... uc::LJ. O.'- 0...0:::::.1. •

Explanation: The value of the argument
function must be in the range 0 to 127(10).

for

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

the
submit

%STR$CHAR
an

SPR

ERLCFL-F-FUNNOTCOM, Function call not allowed in compile-time constant
expression.
Explanation: A control file module contains a function call that
could not be evaluated at compile-time, where a compile-time
constant expression was required.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-FUNQUOODD, Quote string in STR$QUOTE function must have
length.

even

Explanation: A control file module contains
a
%STR$QUOTE
function, where the quote string argument is not an even length.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
5-48

submit

SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-FUNSTRSIZ, Output string from string function too large.
Explanation: A control file module executed a string
which resulted in a string longer than 255 characters.

function

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-FUNWRONGA, Incorrect number of arguments in function call.
Explanation: A control file source module
call with the wrong number of arguments.

contains

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

function
an

SPR

ERLCFL-F-FUNWRONGC, Incorrect number of arguments in function call.
Explanation: A control file source module
call with the wrong number of arguments.

contains

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-HEAPOVERF, Heap too small to hold value.

submit

function
an

SPR

Overflow.

Explanation: The heap used for processing values and expressions
has overflowed.
User Action: Edit CFLBLD.CMD to increase the psect extension for
psect VHEAPO, and rebuild CFL.
ERLCFL-F-IFNOTDATA,
declaration.

declaration

clause

must

data

Explanation: An IF clause cannot be used to declare data outside
of a declaration in a control file source module.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

control

file

ERLCFL-F-ILLCHAR, Illegal character in input.
Explanation:
An illegal character was found in
source module.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-INTEOTMAN, Internal - More than one operator on stack at term
end.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

5-49

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-INTEOTNUL, Internal - End of term reached with null
stack.

operator

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTEXPNOO, Internal - Operator missing from operator stack.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTFUNEND,
termination.

Internal

Stack

entry

missing

function

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTFUNMIS, Internal - Function
stack.

code

missing

from

operator

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTFUNNOT, Internal
termination.

Function

code

missing

function

Explanation:

This is an internai error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTOPRNOT, Internal - Operator outside of an expression term.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTPROUND, Internal
underflow.

Compiler

internal

production

stack

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INTSYMLNK, Internal error - Invalid symbol linkage
module.

setup

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you havee

5-50

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-INTWRONGP,
production.

Internal

Wrong

production

popped

internal

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INVFUNCT, Invalid function name specified.
Explanation:
A control file source module specified
function name.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

invalid

SPR

ERLCFL-F-INVNUMSIZ, Internal - A numeric variable has an invalid size.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-INVPOIACT, Invalid POINTER-statement action name.
Explanation: A control file source module specified
action for a POINTER statement.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-INVRADCNV,
conversion=

Internal

error

Invalid

submit

radix

invalid
an

SPR

code

for

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-KEEPINV, KEEP attribute not allowed with FLUSH attribute.
Explanation: A control file source module had both the FLUSH and
KEEP module attributes specified.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-LITINVTYP, Internal error invalid type.

Literal

submit

literal

SPR

table

has

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-LITNOVALU, Internal error - no value
value.

load

into

literal

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

5-51

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-MODATTRIN, Invalid module attribute name specified.
Explanation: A control file source module specified
module attribute.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

invalid
an

SPR

Explanation: A control file source module contains a modulus
zero in a compile-time constant expression.

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

SPR

ERLCFL-F-MODZERO, Attempt to modulus by zero.

submit

file

source

module

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLCFL-F-NOQUOTE, String literal missing closing quote.
Explanation: A string literal in a control
was not terminated by a closing quote.

ERLCFL-F-NULLOPERA, Internal
operator.

Null

suffix

operand

SPR

non-suffix

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-NUMFILLCH, A print fill character
character.

string

must

contain

Explanation: A print fill character in a declaration
control file source module must contain one charatcer.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

one
in

a
SPR

ERLCFL-F-NUMINVOPR, Invalid numeric double-operand operation code.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-OPRFLSCOM, Internal error - attempt to flush a CTCE operand.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

5-52

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-OPRINVLOG, Attempt to perform logical operation on an invalid
type.

Explanation:

A control file source module attempted to perform a
logical operation with operands that were neither NUMERIC nor
LOGICAL.

User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

ERLCFL-F-OPRNOTCOM,
compile-time.

Operator

CTCE

cannot

submit

SPR

evaluated

Explanation:

A compile-time constant expression in a control
file source module contains an operator which could not be
evaluated at compile-time.

User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-OPRNOTIMP, Operation not implemented.

Explanation: A control file source module attempted to perform a
multiplication where both operands were larger than a word value.
User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

The end-of-file was reached on a control
source module, before the end of the module was reached.

file

User Action:

SPR

ERLCFL-F-PARSEREOF, Premature EOF encountered.

Explanation:

Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLCFL-F-PROSTKOVR, Compiler internal production stack overflow.

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-PRSSTKOVR, Parse stack overflow.

Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-RAD50BYTE, Cannot convert a byte using RAD50 conversion.

Explanation: A control file source module attempted
an ASCII string or numeric literal to a BYTE
conversion.
User Action:

Correct the user-written module or
for DIGITAL-supplied modules.

5-53

to convert
using RAD50

submit

SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-RADINVCOD, Invalid radix code string in radix literal.
Explanation: A control file source module
radix code in a numeric literal.

contains

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-RADINVRAD,
literal.

Invalid

radix

invalid

submit

SPR

character

specified

radix

Explanation: A control file source module
radix code in a numeric literal.

contains

invalid

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-RADLITINV, Invalid literal type character in radix literal.
Explanation: A control file source module contains an invalid
type character in the radix portion of a numeric literal.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-RELINVCOD, Invalid relational operator.
Explanation:

This is an internal error within CFL.

User Action:

Please submit an SPR with any information you have.

ERLCFL-F-RESBITILL, A BIT
RESERVED.

FIELD

data

item

cannot

Explanation~
A control file source module
FIELD declaration for a RESERVED data item.

contains

User Action: Correct the user-written module or
for DIGITAL-supplied modules.
ERLCFL-F-RESINVPRT, Print expression
data.

list

Explanation: A control file source
expression for a RESERVED data item.

not

be
a

BIT

submit

SPR

allowed

module

declared

contains

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

RESERVED
a

print
an

SPR

ERLCFL-F-SPCNOTVAR, Special variable cannot be used in a CTCE.
Explanation:
predeclared
expression.

A control file source module used one of the
special
variables
in a compile-time constant

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

5-54

submit

SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-SRCDATERR, Failed to get creation date of source file
Explanation: An error occurred when CFL tried
creation date of the control file source module.

get

the

User Action: Check for file access conflicts, device errors,
low pool condition.

ERLCFL-F-SUBEXTBIG, Substring extraction end element exceeds string.
Explanation:
A control
file module attempted to perform a
substring extraction in which the substring exceeded the end of
the string.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-SYMNOTCOM, A variable is not valid in a compile-time constant
expression.
Explanation: A control file source module contains a variable in
a compile-time constant expression.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

ERLCFL-F-SYMNOTLIT, Specified LITERAL symbol name not part of
group.

SPR

LITERAL

Explanation: A control file module contains a
reference to a
LITERAL symbol which has not been defined for the specified
LITERAL group.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

Explanation: CFL encountered a syntax error while compiling
control file source module.

the

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

SPR

ERLCFL-F-SYNTAXERR, Syntax error.

submit

ERLCFL-F-TABLEBIG, TABLE element has too many literal values.
Explanation:
A control file source module contains a
element with too many values for the corresponding TABLE.
User Action:
Correct the user-written module or
for DIGITAL-supplied modules.

submit

TABLE
an

SPR

ERLCFL-F-TABLESMAL, TABLE element has too few literal values.
Explanation: A control file source module contains a
element with too few values for the corresponding TABLE.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

5-55

submit

TABLE
an

SPR

CONTROL FILE LANGUAGE GUIDE
ERLCFL-F-VALSTKOVR, Value stack overflow.
Explanation: The stack
used
expressions has overflowed.

for

processing

values

and

User Action: Edit CFLBLD.CMD to increase the psect extension for
psect VLSTKO, and rebuild CFL.
ERLCFL-F-VALUESIZE, Value in expression is too large.
Explanation: A
control
file
source
module
contains
a
compile-time constant expression in which an intermediate value
or the final value is too large.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-VALUETYPE, Value in expression is wrong type.
Explanation: A
control
file
source
module
contains
a
compile-time constant expression in which an intermediate value
or the final value is the wrong type.
User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-VARLITGRP, A variable name cannot have the same group name as
a LITERAL.
Explanation: A control file module contains a
the same group name as a LITERAL declaration.

declaration

with

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

submit

SPR

ERLCFL-F-WRITEDES, WRITE-class statement has invalid destination.
Explanation:
destination
statement.

A control file
module
contains
an
invalid
for the TO clause of a WRITE or WRITE GROUP

User Action: Correct the user-written module or
for DIGITAL-supplied modules.

5-56

submit

SPR

APPENDIX A
TUNING THE ERROR LOGGING UNIVERSAL LIBRARY

When RPT starts operation, it scans ERRLOG.ULB
and
generates
descriptors in its work space for all the modules it finds in the
library. RPT searches these descriptors for the appropriate modules
when it generates reports.
Therefore, if you remove modules for
devices that your system does not have, RPT operation is faster and
does not require as much memory or disk space.
An indirect command file, TUNE.CMD, is included with your Error
Logging System.
The command file lS ln the universal library
remove
LB:[l,6]ERRLOGETC.ULB.
This command file allows you to
unneeded modules from ERRLOG.ULB.
To use TUNE.CMD you must first extract it from the library, using
following LBR command:
LBR TUNE.CMD=LB: [l,6]ERRLOGETC.ULB/EX:TUNE

the

lBD)

The file, TUNE.CMD, is copied to your current directory
execute the command.

and

you

can

The file prompts you for a processor type and for a list of the
devices you want to use with Error Logging. It then removes the
devices you do not select from ERRLOG.ULB and writes a new ERRLOG.ULB
to the current UFD.
TUNE.CMD requires the actual physical names of the devices you select
{RLOl, RL02, for example). Since some modules in ERRLOG.ULB handle
more than one physical device type, the tune command only includes the
module once. If you select a device name more than once, the command
displays the following message on your terminal:
Module <module name> replaced.
The message is only for your information.
The command
continues to include the other modules you specified.

file

then

Example A-1 is an example of use of the RSX-llM version of TUNE.CMD.
Files for other systems are similar. This file creates a library for
an 11/70 with:
Devices
TU45
RP06
RP05
RM03
RM05
RM80
RS04

RK05
RK06
RK07
TU56
RX02
TU58

A-1

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY
Example A-1

Sample Execution of TUNE.CMD

@LB: [l,6]TUNE (@])
This command procedure is used to tune the Error Log Control file for
your system configuration. The procedure prompts for the location of
the master control file (the shipped file), the CPU type, and the
error logging devices available on your system. It then creates a new
version of the control file that contains only the required support.
The original control file is left unchanged.
All files are created in your default directory on the default
device. When cleaning up, all files with the extension of .ICF as
well as TEMPTUNEO.TMP;* and TEMPTUNEl.TMP;* are deleted.
;

* Continue? [Y /N]: Y (@])
;

* Location of master file [D: LB: [l,6]ERRLOG.ULB]
Now enter
CPU types.

the

CPU type.

[S]: (@])

Hit.the escape key for a list of legal

;

*Enter CPU type [S]: @W
The acceptable CPU types are:

.. .. '"" .....

J.J./u~,

J.J./U'tr

, ,
In. e
J..J../ u~'

J..J../ J..V

11/20,
11/35,
11/50,
11/74

11/23,
11/40,
11/55,

11/24,
11/44,
11/60,

11/34
11/45
11/70

In. II

;

* Enter CPU type [ s] : 11/70 ~
Now enter the devices in your configuration separated by commas.
Terminate by entering a period. Hit the escape key for a list of
acceptable device names.
;

*Enter device name(s)

[S]:

Below is a list of acceptable device names. If you have more than
one type of device listed as "x or y or z" you need enter only one.
For example, if you have RP04s and RP06s you need only enter RP04
or RP06 - not both.
The acceptable device names are:
TU56
TU58
TU60
RP04
RP07
RSll
RK03
RLOl
RK06

(DECtape)
(DECtape II)
(Cassette)
or RP05 or RP06
or RK05
or RL02
or RK07
(continued on next page)

A-2

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY
Example A-1 (Cont.)

Sample Execution of TUNE.CMD

RP02 or RP03
RM02 or RM03
RM05

RM80
RS03 or RS04
RA60
RA80/RA81
RC25 or RD51 or RXSO
RXOl
RX02
MLll
TE16 or TU16 or TU45
TU77
TElO or TUlO or TS03
TU78
TSll or TU80
;
TSVOS
*Enter device name(s) [S]: TU45,RP06,RP05,RM03,RM05,RM80 ~
* Enter device name(s) [S]: RS04,RK05,RK06,RK07 ,TU56,RX02,TU58 ifilIJ
* Enter device name(s) [S]: • ~
Extract the files from the master library.
LBR @TEMPTUNEO.TMP
Build the new library. Note that you may see messages like "Module
"XYZZY" replaced" if you have selected more than one device having
the same mnemonic. For example, selecting RK06 and RK07 support will
produce this message. This type of message can be ignored.
;

LBR @TEMPTUNEl.TMP
Module "ERP456" replaced
Module nERK67 " replaced
Module "NRK67 " replaced
Now clean up.
PIP TEMPLIB.ULB/TR
PIP ERRLOG.ULB/RE/NV=TEMPLIB.ULB
PIP *.ICF;*,TEMPTUNEO.TMP;*,TEMPTUNEl.TMP;*/DE
Finished.

You can then copy the new library to [1,6] where it will become the
default
library
for
RPT.
You should maintain the original
DIGITAL-supplied ERRLOG.ULB, either on LB:[l,6] or
in
another
location. You can then use TUNE.CMD again later, on the original ULB,
to add support for devices that you have taken out.
To list the modules in the library, use the following LBR command:

>LBR ERRLOG.ULB/LI :BIT;
Table A-1 lists
RSX-llM.

the

modules

the

DIGITAL-supplied

A-3

library

for

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY

Table A-1
Modules in ERRLOG.ULB for RSX-llM
Module' Name

Module Description

DEVSMl

Defines tables for device modules

DEVUDA

MSCP related module

DIS PAT

Entry modul~ for error log control file

DMPALL

Processes errors from unknown events

DSPlMl
DSP2Ml

Processes Error ·Logger Command Packets
error packets
Processes device information packets
Processes device control-information
Processes CPU/memory detected errors
Processes system control information
Processes control information packets
P~ocesses.device

DSP3M~

DSP4Ml
DSPSMl
DSP6Ml
DSP7Ml
EMLll
ERK OS
ERK67
ERL12
ERM05

ERM23
ERM80
ERP07
ERP23
ERP456
ERSll

·1

l
I

Control File Device Modules

ERS34

ERXOl
ERX02
ETAll
ETCll
ETSll
ETU58

ETU77 .
ET0310
ET1645
ETSV05
ER RO RM

Processes control file error conditions

EUNKWN

Processes errors from unknown devices

EllXX }Ell34
Ell44
Ell60
:
Ell7X ..

Processes CPU/memory packets

FINLMl

Cleans up control file after processing

FMlNMl
FMlWMl
FM2CM1
FM3CM1
FM4NM1
FM4WM1

Prints narrow cir wide width· reports

(continued on next page)
A-4

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY

Table A-1 (Cont.)
Modules in ERRLOG.ULB for RSX-llM
Module Name

Module Description

INITMl

Initializes control file modules

MSCPCE
MSC
PAT ~
MSCP80

:~~:~~ j

MSCP Device Modules

MSC PTO
MSCPSD
NMLll
NRK67
NRMOS
NRM23
NT0310
NTSll

PAR.SEM}
PARSlM
PARS2M
PARS 3M
PRS2AM
PRS2BM
SELTMl

Generates notes from device modules

RPT command line parser

Selects packets to process

:.\~~1b}:~.\Ai:2·":i1s~:s ··'the .m.odulEfs · 1;ri· tzhe o·:tGITAL-4s.upplied\ .· ~>ibrary
·:·
· \ .:. · · ··

jj:t.s~~.tif'l~~4US.•..>

·fcfr

~i~~~-~;f(~~~M_::,~~?~~· . . • ·~#~!~~t~~~'~!~~-~~r~~~- ·.·
'

MSCP felat~.d \module
'Ehtr}" mo(J.ule. for. er

for log control . file·

DMPALL

Processes errors from unknown events

.. 'DSPlPl
·osP2P1
DSP3Pl
DSP4Pl
DSPSPl
DSP6Pl
DSP7Pl

Processes Error Logger Command Packets
Processes device e~~or packets
Processes device information packets
Processes device control information
Processes CPU/memory detected errors
Processes system control information
Processes control information packets
(continued on next page)

A-5

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY
·Table A...:.'2· fCont.}
Modules in ERRLOG.ULB for RSX-llM-PLUS'

Module· Name

Module .Description

EMLll
ERK05
ERK67
ERL12

.ERM OS
ERM23

ERMSO ..
ERP07
. ERP23

FUe·: Device

ERP4S6 .
ERSll
ER.S34

.ERXOl
ERX02 .

. E,TAll
ETCH

ETSll
ETU58
ETU77
·,ET0310
''"'"ET 16,4 5
ETSVU!:>j

ERR ORM

Processes control file error conditions

EUNKWN

Processes ·errors from unknown devices

EllXX "')

Ell44 ).
Ell 7X

Processes CPU/memory packets

FINLPl

Cleans up· control file after processing

FMTNPl
FMTWPl

Pr in ts nar;row width reports
Prints wide width reports

INITMl

Initia,liz;e'.s'. control file modules

MSCPCE "\
MSCPAT

MSCP Dev ip:e; Modules

'.:~g:.:~g
·...J.·
MSC
PEN

MSCPTO . ·

MSCPSD

NRK67

: NRM(fS .·
· ·.· · .N· ·.·.ML
...·l·i.·..
·

NRM23'

·····1·····
..·.:.:.

NT0310

NTSll' ·...

{continued on next page)

A-6

TUNING THE ERROR LOGGING UNIVERSAL LIBRARY
Tcibltf:A-2 , (Conf.J
:·ERRLOG. Uf:..B
..... : ~od.U,le. ·Name · ·

.PAR.S lM. ·. , ·. •· ·.
"PARS2M' .·
·.p•. . A··.··. R
. .·.·.S.·.·•.·•...
E M·.. . ···.·.·.··.··.········.···1··
..··. ·..··

· RPT collimctnd line·· parser·

PARS 3M. ·. ·
":PRS2AM>: . :. •~.

PRSiBM: . ':) .. .

·sf4i;tY~;P:. . ·
$'r.t·R;Y,6l?'. .

~tint~\:E.rro.r: :.summ:~q.( Repo•r~ ··:· .·..
·~:•Pr..ini. s\·.·G·" ·e.~~~t~r.Y;>,-s. umqi<.trY, Re. p.q

,, '·.• ':.: ;o : :' ..

rr· · -:"·'•'

: · . \•,;.:;·:.":; ·"·SMRYap·· .. ; . ,,:; ~. :·::S: ,:.~~1;.,~·P;~... ;H.~.~jiqp[. ·~Uinm.~ry :.'.R~;p~t"~. ''
:·,

.;:

~,~, '.»_ ~- ';, ;,·=\~, '., -~~: -~ ·:, ~ : ,' ?·

~·' ~ -~ ;: '( .

;,. ·,

•• ,

~ ~. ~

A-7

.. :-:

APPENDIX B
DRIVE SERIAL NUMBERS

RPT reports drive serial numbers for those devices that have serial
numbers.
Table B-1 lists the drives that provide serial numbers and
the significant digits RPT uses from those serial numbers. The number
of digits varies by drive type. They appear in binary coded decimal
(BCD) format.
RPT obtains the serial number from the device electronics. The number
is selected through a series of jumpers within the device that are set
at manufacturing time. These jumpers match the low order digits of
the actual device serial number.
If any of these jumpers is altered,
the drive will have a different serial number, unless the new jumpers
are set to reflect the actual device serial number. Therefore, the
number RPT prints may not be the same as the number on the device
identification plate.
Table B-1
Significant Digits in Drive Serial Numbers
Device/Controller

Significant Digits in Serial Number
3 BCD digits
3 BCD digits
4 BCD digits
4 BCD digits
4 BCD digits
4 BCD digits
4 BCD digits
4 BCD digits
4 BCD digits

RK06
RK07
RM80
TU77
TU78
RP04
RPOS
RP06
RP07
RXOl
RX02
RM02
RM03
RMOS
TAll
RSll
RP02
RP03
MLll

12 bits,
12 bits,
16 bits,
16 bits,
16 bits,
16 bits,
16 bits,
16 bits,
16 bits,
N/A
N/A
16 bits,
16 bits,
16 bits,
N/A
N/A
N/A
N/A
16 bits,

TE16/RH11/RH70
TU16/RH11/RH7 0
TU45
TS03
TElO/TMBll
TUlO/TMBll

16 bits, 4 BCD digits
16 bits, 4 BCD digits
16 bits, 4 BCD digits
N/A
N/A
N/A

4 BCD digits
4 BCD digits
4 BCD digits

4 BCD digits

(continued on next page)
B-1

DRIVE SERIAL NUMBERS
Table B-1 (Cont.)
Significant Digits in Drive Serial Numbers
Device/Controller
RS03
RS04
TSll
TCll
RLOl
RL02
TSV05
RA60/RA80/RA81
RC25/RD51/RX50
TU80

Significant Digits in Serial Number

N/A
N/A
N/A
N/A
N/A
N/A
N/A
6 digits
6 digits
N/A

B-2

APPENDIX C
ERROR LOG PACKET FORMAT

Example C-1 shows the format of an error
described in the system macro EPKDF.MAC.

log

packet

memory,

When a device error is logged, the error log packet contains device
supplied information to describe the error. This information usually
consists of the device registers and some additional information
supplied by the system.
Error logging also writes context information into the error log
packet.
This information includes the time and date of the error,
information about the system that logged the error, and information
about the I/O operation that generated the error.
The error logging system also creates packets for events in the system
that are not errors; but are important to the interpretation of
errors, such as the time error logging starts or stops.
Example C-1
.IIF NDF S$$YDF ,

Error Log Packet Format
.NLIST

COPYRIGHT (c) 1983 BY
DIGITAL EQUIPMENT CORPORATION, MAYNARD,
MASSACHUSETTS.
ALL RIGHTS RESERVED.
THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED AND COPIED
ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE AND WITH THE
INCLUSION OF THE ABOVE COPYRIGHT NOTICE. THIS SOFTWARE OR ANY OTHER
COPIES THEREOF MAY NOT BE PROVIDED OR OTHERWISE MADE AVAILABLE TO ANY
OTHER PERSON. NO TITLE TO AND OWNERSHIP OF THE SOFTWARE IS HEREBY
TRANSFERRED.
THE INFORMATION IN THIS SOFTWARE IS SUBJECT TO CHANGE WITHOUT NOTICE
AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL EQUIPMENT
CORPORATION.
DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY
SOFTWARE ON EQUIPMENT WHICH IS NOT SUPPLIED BY DIGITAL.

ITS

IDENT /l.03/
Modified By:
C. PUTNAM

- 03-FEB-82
1.01
- Add M-PLUS Definition of 'NUX'

(No UCB extension) bit

(continued on next page)

C-1

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)
C. PUTNAM

- 13-SEP-82
1.02
- Add Q-BUS flag to HEADER subpacket

C. PUTNAM

- 06-JAN-83
1.03
- Add E$STMS (recovered timeout) device error
- Now E$STMO means unrecovered timeout device error

.MACRO

Error Log Packet Format

EPKDF$,L,B

; Error Message Block Definitions
;.ASECT
Header Subpacket

+-----------------------------------------------+
I Subpacket Length in Bytes
I
+-----------------------------------------------+
I Subpacket Flags
I
+-----------------------+-----------------------+
Format Identification
Operating System Code
+-----------------------+-----------------------+
/ Operating System Identification
I
I

+-----------------------+-----------------------+
I Flags
I Context Code
I
+-----------------------+-----------------------+
I Entry Sequence
I
+-----------------------------------------------+
I Error Sequence
I
+-----------------------+-----------------------+
Entry Type Subcode
Entry Type Code
+-----------------------+-----------------------+
I Time Stamp
I
I

+-----------------------+-----------------------+
Reserved
Processor Type
+-----------------------+-----------------------+
I
I
+-----------------------------------------------+
I

-- -

l:'L.U~t!::i::iUL

,e .! - - .L ! - .LUt!lll..1.L.L~dl..1.Vll

... .:1 -

1 r"T -

• • \.

\Ut'l:l"l}

.=O
E$HLGH: 'LI
E$HSBF: 'L'
E$HSYS: 'LI
E$HIDN: 'LI
E$HSID: 'L'
E$HCTX: 'L'
E$HFLG: 'LI
E$HENS:'L'
E$HERS: 'LI
E$HENC: 'LI
E$HTYC: 'LI
E$HTYS: 'L'
E$HTIM: 'LI
E$HPTY:'L'

.BLKW
.BLKW
.BLKB
.BLKB
.BLKB
.BLKB
.BLKB
.BLKW
.BLKW

1
1
1
1

.BLKB
.BLKB
.BLKB
• BLKB
.BLKB

1
1
6
1
1

1
1
1
1

Subpacket length in bytes
Subpacket Flags
Operating System Code
Format Identification
Operating System Identification
Context Code
Flags
Entry Sequence Number
Error Sequence Number
Entry Code
Entry Type Code
Entry Type Subcode
Time Stamp
Processor Type
Reserved

(continued on next page)

C-2

ERROR LOG PACKET FORMAT

Example C-1
E$HURM: 'L'

.BLKW

(Cont.)

Error Log Packet Format

; Processor Identification (URM)

.EVEN
E$HLEN: 'LI

Length

Subpacket Flags for E$HSBF
SM.ERR
SM.HOR
SM.TSK
SM.DID
SM.DOP
SM.DAC
SM.DAT
SM.MBC
SM.CMD
SM.ZER

='B'
= 'B'
='BI
='B'
='B'
= 'B'
='BI
= 'B'
='B'
='B'

1
1
2
4
10
20
40
20000
40000
100000

Error Packet
Header Subpacket
Task Subpacket
Device Identification Subpacket
Device Operation Subpacket
Device Activity Subpacket
Data Subpacket
22-bit massbus controller present
Error Log Command Packet
Zero I/O Counts

Codes for field E$HIDN
EH$FOR

='B'

Current packet format

Flags for the error log flags byte ($ERFLA) in the exec.
ES. INI
ES.DAT
ES.LIM
ES.LOG

='BI
= 'B'
='B'
= 'B'

1
2
4
10

Error
Error
Error
Error

log initialized
log receiving data packets
limiting enabled
logging enabled

Type and Subtype Codes for fields E$HTYC and E$HTYS
Symbols with names E$Cxxx are type codes for field E$HTYC,
symbols with names E$Sxxx are subtype codes for field E$HTYS.
E$CCMD
E$SSTA
E$SSWI
E$SAPP
E$SBAC
E$SSHO
E$SCHL

='B'
='B'
='B'
= 'B'
= 'B'
= 'B'
='B'

E$CERR
E$SDVH
E$SDVS
E$STMO
E$SUNS
E$STMS

= 'B'
='BI
= 'B'
= 'B'
= 'B'
='B'

E$CDVI
E$SDVI

= 'B'
='BI

E$CDCI
E$SMOU

= 'B'
= 'B'
= 'B'
= 'B'
='B'
='B'
='B'

E$SDMO

E$SRES
E$SRCT
E$CCPU
E$SMEM

1
2
3

4
5
6
2

1
2
3

4
5
3

1
4
1
2
3
4
5
1

Error Log Control
Error Log Status Change
Switch Logging Files
Append File
Declare Backup File
Show
Change Limits
Device Errors
Device Hard Error
Device Soft Error
Device Interrupt Timeout (HARD)
Device Unsolicited Interrupt
Device Interrupt Timeout (SOFT)
Device Information
Device Information Message
Device Control Information
Device Mount
Device Dismount
Device Count Reset
Block Replacement
CPU Detected Errors
Memory Error

(continued on next page)

C-3

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)

Error Log Packet Format

E$SINT

='B'

Unexpected Interrupt

E$CSYS
E$SPWR

='B'
='B'

System Control Information
Power Recovery

E$CCTL
E$STIM
E$SCRS
E$SLOA
E$SUNL
E$SHRC
E$SMES

='B'
='B'
='B'
='B'
='B'
='B'
='B'

7
1
2
3
4
5
6

Control Information
Time Change
System Crash
Device Driver Load
Device Driver Unload
Reconfiguration Status Change
Message

E$CSDE
E$SABO

='B'
='B'

10
1

Software Detected Events
Task Abort

Codes for Context Code entry E$HCTX
EH$NOR
EH$STA
EH$CRS

='B'
='B'
='B'

1
2
3

Normal Entry
Start Entry
Crash Entry

Codes for Flags entry E$HFLG
EH$VIR
EH$EXT
EH$COU
EH$QBS

='B'
='B'
='B'
='B'

1
2
4

Addresses are virtual
Addresses are extended
Error counts supplied
Q-BUS CPU

Task Subpacket

+-----------------------------------------------+
I Task Subpacket Length
I
+-----------------------------------------------+
I Task Name in RADSO
j

;.=O

+-----------------------------------------------+
I Task UIC
I
+-----------------------------------------------+
I Task TI: Device Name
I
+-----------------------+-----------------------+
I Flags
I Task TI: Unit Number j
+-----------------------+-----------------------+

E$TLGH: 'LI
E$TTSK:'L'
E$TUIC: 'LI
E$TTID: 'L'
E$TTIU: IL I
E$TFLG: 'LI

.BLKW
.BLKW
.BLKW
.BLKB
.BLKB
.BLKB

Task Subpacket Length
Task Name in RAD50
Task UIC
Task TI: Device Name
Task TI: Unit
Flags

1
2
1
2
1
1

.EVEN
E$TLEN:'L'
Flags for entry E$TFLG
ET$PRV

='B'

Task is Privileged
(continued on next page)
C-4

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)
ET$PRI

='B'

Error Log Packet Format
Terminal is Privileged

Device Identification Subpacket

+-----------------------------------------------+
j Device Identification Subpacket Length
I
+-----------------------------------------------+
I Device Mnemonic Name
I
+-----------------------+-----------------------+
I Controller Number
I Device Unit Number
I
+===~===~=--------------+~-~~--~~--------------=+

I Physical Subunit #

I Physical Unit #

+-----------------------+-----------------------+
I Physical Device Mnemonic (RSX-llM-PLUS only) I
+-----------------------+-----------------------+
I Reserved
I Flags
I
+-----------------------+-----------------------+

Volume Name of Mounted Volume

+-----------------------------------------------+
I Pack Identification
I
+-----------------------------------------------+
I Device Type Class
I
+-----------------------------------------------+
I Device Type
I
+-----------------------------------------------+
I I/O Operation Count Longword
I
+-----------------------+-----------------------+
I Hard Error Count
I Soft Error Count
I
+-----------------------+-----------------------+
I Blocks Transferred Count (RSX-llM-PLUS only) I

;

+-----------------------------------------------+
I Cylinders Crossed Count (RSX-llM-PLUS only) I
+-----------------------------------------------+

.=O
E$ILGH:'L'
E$ILDV: IL I
E$ILUN: 'L'
E$IPCO: IL I
E$IPUN: 'L'
E$IPSU: 'L'

.BLKW
.BLKW
.BLKB
.BLKB
.BLKB
.BLKB

1
1
1
1
1
1

Device Identification Subpacket Length
Device Mnemonic Name
Device Unit Number
Controller Number
Physical Unit Number
Physical Subunit Number

.IF DF R$$MPL
E$IPDV: 'LI
E$IFLG: 'L'
E$IVOL: IL I

.BLKW

Physical Device Mnemonic

.ENDC
.BLKB
.BLKB
.BLKB

R$$MPL
1
1
12.

Flags
Reserved
Volume Name
(continued on next page)

C-5

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)
E$IPAK: 'LI
E$IDEV: 'L'
E$IDCL: IL I
E$IDTY: IL I
E$IOPR: IL I
E$IERS: 'LI
E$IERH: 'LI

.BLKB

.BLKW
.BLKW
.BLKW
.BLKB
.BLKB

1
2
2
1
1

Error Log Packet Format
Pack Identification
Device Type
Device Type Class
Device Type
I/O Operation Count Longword
Soft Error Count
Hard Error Count

.IF DF R$$MPL
E$IBLK: 'LI
E$ICYL: 'LI

.BLKW
.BLKW

2
2

.ENDC

R$$MPL

Blocks transferred count
Cylinders crossed count

.EVEN
E$ILEN: 'L'

Subpacket Length

Flags for field E$IFLG
EI$SUB
EI$NUX

='B'
1
.IF DF R$$MPL
='B'
2
.ENDC ; R$$MPL

Subcontroller device
No UCB extension, data invalid

Device Operation Subpacket

+-----------------------------------------------+
I Device Operation Subpacket Length
I
+-----------------------------------------------+
I Task Name in RADSO
I
+-----------------------------------------------+
I Task UIC
I
+-----------------------------------------------+
I Task TI: Logical Device Mnemonic
I
+-----------------------+-----------------------+
I Reserved
I Task TI: Device Unit I
+-----------------------+-----------------------+
I I/O Function Code
I
+-----------------------+-----------------------+
I Reserved
I Operation Flags
I
+-----------------------+-----------------------+
I Transfer Operation Address
I
+-----------------------------------------------+
I Transfer Operation Byte Count
I
+-----------------------------------------------+
I Current Operation Retry Count
I
+-----------------------------------------------+
.=O
E$0LGN: 'LI
E$0TSK: 'L'
E$0UIC: 'L'
E$0TID: 'L'
E$0TIU: 'LI

.BLKW
.. BLKW
.BLKW
.BLKB
.BLKB
.BLKB

1
2
1

2
1
1

Subpacket Length
Task Name in RADSO
Task UIC
Task TI: Logical Device Mnemonic
Task TI: Logical Device Unit
Reserved
(continued on next page)

C-6

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)
E$0FNC: 'LI
E$0FLG: 'LI
E$0ADD: IL I
E$0SIZ:'L'
E$0RTY: 'L'

.BLKW
.BLKB
.BLKB
.BLKW
.BLKW
.BLKW

1
1
1

Error Log Packet Format
I/O Function Code
Operation Flags
Reserved
Transfer Operation Address
Transfer Operation Byte Count
Current Operation Retry Count

2
1
1

.EVEN
E$0LEN:'L'

Device Operation Subpacket Length

Flags for field E$0FLG
EO$TRA
EO$DMA
E0$EXT
EO$PIP

='B'
='B'
='B'
='B'

1
2
4

Transfer Operation
OMA Device
Extended Addressing Device
Device is positioning

I/O Activity Subpacket

;
.=O

+-----------------------------------------------+
j I/O Activity Subpacket Length
j
+-----------------------------------------------+

E$ALGH: 'LI

.BLKW

Subpacket Length

I/O Activity Subpacket Entry

+-----------------------------------------------+
I Logical Device Name Mnemonic
I
+-----------------------+-----------------------+
I Controller Number
I Logical Device Unit I
+-----------------------+-----------------------+
I Physical Subunit #
I Physical Unit Number I
+-----------------------+-----------------------+
I Physical Device Mnemonic (RSX-llM-PLUS only) I
+-----------------------+-----------------------+
I Task TI: logical unit I Device flags
I
+-----------------------+-----------------------+
I Requesting Task Name in RADSO
I
+-----------------------------------------------+
j Requesting Task UIC
I
+-----------------------------------------------+
I Task TI: Logical Device Name
I
+-----------------------------------------------+
I I/O Function Code
I
+-----------------------+-----------------------+
I Reserved
I Flags
I
+-----------------------+-----------------------+
I Transfer Operation Address
I
;
.=O

+-----------------------------------------------+
I Transfer Operation Byte Count
I
+-----------------------------------------------+
(continued on next page)

C-7

ERROR LOG PACKET FORMAT
Example C-1 (Cont.)
E$ALDV: 'L'
E$ALUN: 'L'
E$APCO: 'L'
E$APUN: 'L'
E$APSU:'L'

.BLKW
.BLKB
.BLKB
.BLKB
.BLKB

Error Log Packet Format
Logical Device Name Mnemonic
Logical Device Unit
Controller Number
Physical Unit Number
Physical Subunit Number

1
1
1
1
1

.IF DF R$$MPL
E$APDV: 'L'

.BLKW

Physical Device Mnemonic

1
1
2
1
1
1
1
1
2
1

Device flags
Task TI: Logical Unit
Requesting Task Name in RADSO
Requesting Task UIC
Task TI: Logical Device Name
I/O Function Code
Flags
Reserved
Transfer Operation Address
Transfer Operation Byte Count

.ENDC
E$ADFG:'L'
E$ATIU:'L'
E$ATSK: 'LI
E$AUIC: 'L'
E$ATID:'L'
E$AFNC: 'LI
E$AFLG: 'LI
E$AADD: 'LI
E$ASIZ: 'L'

.BLKB
.BLKB
.BLKW
.BLKW
.BLKW
.BLKW
.BLKB
.BLKB
.BLKW
.BLKW

.EVEN
SubpackeL Entry

E$ALEN:'L'

L~nyth

Flags for field E$ADFG
EA$SUB
EA$NUX

='BI
1
.IF DF R$$MPL
='BI
2
.ENDC ; R$$MPL

Subcontroller device
No UCB extension, data invalid

Flags for field E$AFLG
'C'7\
cmo7\
.UC'1~ .L l.'r1

EA$DMA
EA$EXT
EA$PIP

='B'
= 'B'
='B'
= 'B'

l
2
4

mv~~~~ov
.1..&..YL.l.,;,;,.&..~..I..

.PSECT
.MACRO
.ENDM

EPKDF$

X,Y

.ENDM
.IIF NDF S$$YDF

n~~v~~;~~

V,t',,;;,;.&.Y\.o..&.....,1.1

DMA Device
Device has Extended Addressing
Device is positioning

.LIST

C-8

INDEX

ACTION statement, 5-30
/AP switch, 2-3, 2-8
Append switch, 2-3, 2-8
Delete subswitch, 2-3, 2-8
Architecture
control file module, 4-1
error log control file, 4-1
/BA switch, 2-3, 2-9
Backup switch, 2-3, 2-9
BEGIN-END statement, 5-36
$BMSET, 4-20

Brief format report, 3-11
CALDEV, 4-24
CALL statement, 5-33
CASE statement, 5-34
CDA, 1-1
CFL, 1-1, 1-3, 1-5, 5-1
declaration
conditional, 5-29 to 5-30
data i tern, 5-26
definition, 5-3
scope, 5-26
variable, 5-26
definition, 5-2
expression
conditionals in, 5-2
list handling, 5-39
intermediate form modules,
5-1
lexical conditionals, 5-36
signalling, 5-40
spaces and tabs in text, 5-2
statement
ACTION, 5-30
BEGIN-END, 5-36
CALL, 5-33
CASE, 5-34
CONTROL, 5-32
CRASH, 5-41
DECLARE, 5-27
DECODE, 5-32
DECREMENT, 5-31
DYNAMIC TABLE, 5-37
ENABLE, 5-40
FILE, 5-37 to 5-38
FIND, 5-39
IF-THEN-ELSE, 5-34
INCREMENT, 5-31
LEAVE, 5-35
LIST, 5-39
LITERAL, 5-33
MESSAGE, 5-41
MODULE, 5-32

CFL
statement {Cont.)
PACKET, 5-28
POINTER, 5-38 to 5-39
PRINT FORMATTING, 5-41
PROCEDURE, 5-33 to 5-34
PUT, 5-39
RETURN, 5-33
SEARCH, 5-40
SELECT, 5-34 to 5-35
SET, 5-31
SIGNAL, 5-40 to 5-41
SIGNAL STOP, 5-41
SUBPACKET, 5-29
TABLE, 5-37
WHILE/UNTIL/DO, 5-35
WRITE, 5-31
WRITE GROUP, 5-31
table structure, 5-36
user interface
command mode, 5-43
option mode, 5-43
CFL command line
DEVSMl, 4-36
CFL command mode, 5-43
CFL comments, 5-3
CFL data type
ASCII I 5-5
automatic conversion, 5-4
binary byte, 5-4
bit, 5-4, 5-7
expression, 5-4
logical, 5-4
numeric, 5-5
attribute option, 5-5
byte, 5-5
default, 5-5
longword, 5-5
quadword, 5-5
radix option, 5-5
value, 5-5
word, 5-5
numeric field, 5-7
numeric literal, 5-6
pointer, 5-7
RSXTIME, 5-7
string, 5-4
numeric value, 5-4
VMSTIME, 5-7
CFL expression, 5-9
definition, 5-3
logical operators, 5-10
numeric operators, 5-13 to
5-15

Index-!

INDEX
CFL expression (Cont.)
relational operators, 5-11 to
5-13
string operators, 5-9
CFL file
command, 5-3
error, 5-4
input, 5-3
report, 5-3
user, 5-3
CFL function
%CND, 5-17
%CNV, 5-17 to 5-18
%CNV$RSXTIME, 5-19
%CNV$STRING, 5-19
%CNV$VMSTIME, 5-19
%COD, 5-19
%COM I 5-20
%CTL, 5-21
format, 5-16
%LOK, 5-21
%PKT, 5-22
%RPT, 5-22
%STR, 5-23 to 5-24
%TIM, 5-25
-t;Ui:>K1 5-25
CFL literals, 5-9
CFL named variable, 5-8
CONTEXT, 5-8
COUNT, 5-9
LENGTH, 5-8
POINTER, 5-8
CFL operands
literals, 5-9
variables, 5-9
CFL operator precedence, 5-15
r~T
....... ... ....

nn~;nn
..., I:' '- ... .._, & &

mn~o
&L& .._, '-4. ' - ,

Control File Language
definition
See also CFL, 4-2
Control File Language compiler
See also CFL
Control file module, 1-3
architecture, 4-1 to 4-2
compilation path, 4-11
CPU level module, 4-7
data subpacket, 4-3
definition, 4-1
device activity subpacket,
4-3
device ID subpacket, 4-3
device operation subpacket,
4-3
DEVSMl, 4-7
DEVUDA, 4-7
Dispatch, 4-3
dispatching, 4-1
DMPALL, 4-8
DSPlMl, 4-4
DSPlPl, 4-4
DSP2Ml, 4-5
DSP2Pl, 4-5

~-A~
"'Z wl

CFL primitives, 5-1
CFL statement
definition, 5-3
%CND function, 5-17
.CNF module, 4-1
%CNV function, 5-17 to 5-18
%CNV$RSXTIME, 5-19
%CNV$STRING, 5-19
%CNV$VMSTIME, 5-19
%COD, 5-19
%COM, 5-20
Compilation path
RSX-llM, 4-11
RSX-llM-PLUS, 4-12
Compiler conditional
literal declaration, 4-12
Computational function, 5-20
Concurrent I/O activity, 4-22,
4-24
optional logging, 4-23
Concurrent I/O error logging,
1-5
Conditional declaration, 5-30
Conditional function, 5-17
Control file
universal library, 5-1

Index-2

DSP4fvil, 4-5

DSP4Pl, 4-5
DSP5Ml, 4-5
DSP5Pl, 4-5
DSP6Ml, 4-6
DSP6Pl, 4-6
DSP7Ml, 4-6
DSP7Pl, 4-6
El134, 4-8
El144, 4-8
Ell60, 4-8
Ell7X, 4=8
EllXX, 4-8
ERRORM, 4-7
EUNKWN, 4-8
FINLMl, 4-6
FINLPl, 4-6
FMlNMl, 4-6
FMlWMl, 4-6
FM2CM1, 4-6
FM3CM1, 4-6
FM4NM1 I 4-6
FM4WM1, 4-6
FMTNPl, 4-7
format, 5-1
header subpacket, 4-3
INITMl, 4-7
interaction of interface,
4-13
literal declaration, 4-12
naming conventions, 4-1
Non-DIGITAL device module,
4-1
options, 4-12
PARSlM, 4-4
PARS2M, 4-4
PARS3M, 4-4
PARSEM, 4-4

INDEX
Control file module (Cont.)
program control flow, 4-10
PRS2AM, 4-4
PRS2BM, 4-4
recompilation, 4-13
SELTMl, 4-4
SMRYEP, 4-7
SMRYGP, 4-7
SMRYHP, 4-7
summary, 4-3
task subpacket, 4-3
CONTROL statementi 5-32
CPU-level
dispatching module, 4-11
DSP5Ml, 4-19
DSP5Pl, 4-19
CRASH statement, 5-41
$CRPKT, 4-23
%CTL, 5-21
/DA switch, 3-5, 3-8
Date switch, 3-5, 3-8
/DE switch, 3-5, 3-8
DECLARE statement, 5-26 to 5-27
DECODE statement, 5-32
DECREMENT statement, 5-31
Default fill character, 5-6
Default print field width, 5-6
Default radix, 5-6
Defined report string
DIGITAL, 3-27
user, 3-28
Device drivers
without error logging, 1-3
Device error logging, 4-21 to
4-22
concurrent I/O activity, 4-22
Device errors
error logging, 1-6
hardware register contents,
1-6
Device I/O activity, 1-1
Device information table, 4-34
ALTPRINTNAME field, 4-35
DISPNAME field, 4-35
DRIVETYPE field, 4-36
MASSBUSFLAG field, 4-36
MNEMONIC field, 4-35
PRINTNAME field, 4-35
SIZE field, 4-36
Device module
EMLll, 4-8
ERK05, 4-8
ERK67, 4-8
ERL12, 4-8
ERM05, 4-8
ERM23, 4-8
ERM80, 4-8
ERP07, 4-8
ERP23, 4-8
ERP456, 4-8
ERSll, 4-9
ERS34, 4-9

Device module (Cont.)
ERXOl, 4-9
ERX02, 4-9
ET0310, 4-9
ET1645, 4-9
ETAll, 4-9
ETCll, 4-9
ETSll, 4-9
ETSV05, 4-9
ETU58, 4-9
ETU77, 4-9
MSCP60: 4-9
MSCP80, 4-9
MSCPAT, 4-9
MSCPE, 4-9
MSCPEN, 4-9
MSCPSD, 4-9
MSCPTO, 4-9
Device switch, 3-5, 3-8
Device timeout, 4-20
Device timeout logging, 4-21
Device-level module, 4-11
adding to system, 4-25
addition to device
information table, 4-34
bit-to-text translation, 4-30
coroutine, 4-30
device data declaration, 4-26
device name, 4-26
error type, 4-30
EUNKWN, 4-37
exit, 4-31
* flag, 4-28
intermodule variable, 4-29
local work variable, 4-29
MASSBUS, 4-33 to 4-34
module statement, 4-26
Non-MASSBUS, 4-33 to 4-34
notes, 4-31
procedure statement, 4-26
register definitions, 4-27
subpacket statement, 4-26
Devices
error logging, 2-4
control file module, 2-4 to
2-5
DEVSMl, 4-7, 4-18, 4-34, 4-36
SLP, 4-36
DEVUDA, 4-7
Dispatch module, 4-3, 4-11
subpacket definition, 4-69 to
4-71

Dispatch module path
brief report, 4-15
full report, 4-15
no report, 4-15
register report, 4-15
Dispatch procedure, 4-10
Dispatching
event-level, 4-16
DMPALL, 4-8
Drive serial numbers
significant numbers, B-1

Index-3

INDEX
DSPlMl, 4-4
subpacket definition,
DSPlPl, 4-4
subpacket definition,
DSP2Ml, 4-5, 4-36, 4-50
subpacket definition,
DSP2Pl, 4-5, 4-36, 4-57
subpacket definition,
summary report, 4-16
DSP3Ml
subpacket definition,
DSP3Pl
subpacket definition,
DSP4Ml, 4-5
subpacket definition,
DSP4Pl, 4-5
subpacket definition,
DSP5Ml, 4-5, 4-19
subpacket definition,
DSP5Pl, 4-5, 4-19
subpacket definition,
DSP6Ml, 4-6
subpacket definition,
DSP6Pl, 4-6
subpacket definition,
DSP7Ml, 4-6
subpacket definition,
DSP7Pl, 4-6
subpacket definition,
$DTOER, 4-21
$DVCER, 4-21 to 4-22
$DVERR, 4-21 to 4-22
~1"'\~Fm&A'r\

y I.I V .I. l"I U I

~ -

4-72
4-72
to 4-56
4-73
to 4-66
4-73
4-73
4-73
4-74
4-74
4-74
4-74
4-74
4-74

Ct-.7;

f- ,...h

- ,.., .. - - ... ...

4-75
4-75

'l,
£. .!.

$DVTMO and $DTOER, 4-20
DYNAMIC TABLE statement, 5-37
Ell34~

ELI
Log switch (Cont.)
error limiting and, 2-2
New version subswitch, 2-3,
2-6
No limit subswitch, 2-3,
2-5 to 2-6
No limit switch, 2-3, 2-6
No log switch, 2-4
/NOLIM switch, 2-3, 2-6
/NOLOG switch, 2-4
nonprivileged command, 2-2
privileged commands, 2-2
/R switch, 2-4
/RE switch, 2-7
Reset switch, 2-4, 2-7
/SH switch, 2-4, 2-10 to 2-11
Show switch, 2-2, 2-4, 2-10
to 2-11
/SL switch, 2-4, 2-7
Soft error limit switch, 2-4,
2-7
subswitch summary, 2-3
/SW switch, 2-4, 2-8
switch functions, 2-1

4-8
Ell44, 4-8
Ell60, 4-8
Ell7X, 4-8
EllXX, 4-8
ELI, 1-1
/AP switch, 2-3, 2-8
Append switch, 2-3, 2-8
Delete subswitch, 2-3, 2-8
/BA switch, 2-3, 2-9
Backup switch, 2-3, 2-9
error messages
console terminal, 2-11
user terminal, 2-11
file naming switch, 2-8
Hard error limit switch, 2-3,
2-7
/HL switch, 2-3, 2-7
installing
pr iv i1 eg ed , 2-1
startup command file, 2-1
invoking, 2-2
pr iv i l eg ed , 2 -1
/LIM switch, 2-3, 2-6
Limit switch, 2-3, 2-6
/LOG switch, 2-2 to 2-5, 2-8
Log switch, 2-3 to 2-5, 2-8

~11m1'1'111 ::t W'"'1.7

- ............. '""" ~ J '

"")_ '")

.,,,

Switch switch, 2-4, 2-8
Delete subswitch, 2-4, 2-9
New version subswitch, 2-4,
2-9
using ERRLOG defaults, 2-2
ELI command
reset limit, 1-4
ELI show switch
QIO count, 2-11
ELI switches
type
display, 2-2
f i1 e nam i ng , 2 - 2
limiting, 2-2
1 ogg i ng , 2 - 2
EMLll, 4-8
ENABLE statement, 5-40
Encoding function, 5-19
Entry procedure, 4-10
EPKDF.MAC, 4-16
ERK05, 4-8
ERK67, 4-8
ERLl 2 I 4-8
ERLCFL Report Messages, 5-44
ERLCNF Report Messages, 3-29
fatal t 3-29
information, 3-34
ERLRPT Report Messages, 3-35
control file, 3-35
ERM05, 4-8
ERM23, 4-8
ERM23 device-level module, 4-37
to 4-49
ERM80, 4-8
ERP07, 4-8
ERP23, 4-8
ERP456, 4-8

Index-4

INDEX
ERRDEFINE.CFS, 3-27
ERRLOG, 1-1
Backup file, 2-9
backup file, 1-4
defaults, 2-2
ELI commands, 1-3 to 1-4
installing
privileged, 2-1
log file, 1-4
mandatory installation, 2-1
ERRLOG Messages, 2-12
ERRLOG.ULB, 3-2
module addition, 4-36
tuning, A-1
ERRLOGETC.ULG, 4-36
Error limit, 1-4
device mount, 2-6
logging and, 2-6
notification, 1-4
reset, 1-4, 2-7
reset counts, 2-6
system reboot, 2-6
Error Log
Interface
See also ELI
Error log
command packet, 1-4
control file, 1-3
definition, 4-1
devices, 2-4
file, 1-1
definition, 4-2
format, 5-1
naming, 2-8
RPT I 1-3
packet, 1-1
command, 1-3
creation, 4-23
definition, 4-2
ERRSEQ, 1-3
format, 1-4, 5-1, C-1
packet number
definition, 1-3
processing, 5-1
queue, 4-24
remove from queue, 4-25
SMSG$, 1-4
structure, 4-3
unsolicited interrupt, 4-23
packet number, 3-1
report
context information, 1-5
subpacket
definition, 4-2
system
task interaction, 1-2
Error log c
ontrol file
architecture, 4-1
Error Logger
See also ERRLOG
Error logging
device errors, 1-6

Error logging (Cont.)
executive support, 1-6
device errors, 1-6
device timeouts, 1-6
memory errors, 1-6
unexpected traps or
interrupts, 1-6
interrupt timeouts, 1-6
memory errors, 1-6
support on Non-DIGITAL device,
4-19
device level modulei 4-19
driver, 4-19
universal library entry,
4-19
Unexpected traps or
interrupts, 1-6
unknown device
See EUNKWN
unused vectors, 1-6
Error logging devices
control fiie module, 2-4 to
2-5
Error Logging System
hard error, 1-3
soft error, 1-3
Error logging system
CDA, 1-1
Control file language
compiler, 1-1
device I/O activity, 1-1
Error log interface, 1-1
Error logger, 1-1
executive routines, 1-1
mass storage device errors,
1-1
memory errors, 1-1
operating system failure, 1-1
Report generator, 1-1
ERROR module, 4-19
Error sequence number
See ERRSEQ
Error type definition, 1-5
ERRORM, 4-7
ERRSEQ
Append switch, 1-3
change, 1-3
RMD use, 1-3
ERSll, 4-9
ERS34, 4-9
ERXOl, 4-9
ERX02, 4-9
ET0310, 4-9
ET1645, 4-9
ETAll, 4-9
ETCll, 4-9
ETSll, 4-9
ETSV05, 4-9
ETU58, 4-9
ETU77, 4-9
EUNKWN, 4-8, 4-19, 4-37
Event
definition, 1-3, 4-2

Index-5

INDEX
Event-level dispatching, 4-16
control information, 4-17
CPU-detected errors, 4-17
device control information,
4-17
device errors, 4-17
device information, 4-17
dispatcher module, 4-18
error log control code, 4-17
event code, 4-18
event type, 4-18
format, 4-18
system control information,
4-17
Example
Brief format report, 3-13 to
3-14
device information table,
4-34
DSP2Ml dispatch module, 4-14
DSP2Ml dispatcher module,
4-50 to 4-56
DSP2Pl, 4-15
DSP2Pl dispatcher module,
4-57 to 4-66
T:'1T""lit.•'"''""

i::. r\l'l.G J ,

~ -

FM4WM1, 4-6, 4-11
FMTNPl, 4-7, 4-11
FMTPl, 4-11
FNlNMl, 4-6
$FNERL, 4-22
Format module, 4-11
Format switch, 3-5
full report, 3-15
no report, 3-19
Register report, 3-18
Full format report, 3-15
context information, 1-5
device error, 1-5
device supplied information,
1-5
* flag, 1-5
I/O operation, 1-5
Ha rd error
definition, 1-3
Hard error limit switch, 2-3,
2-7
/HL switch, 2-3, 2-7

"'r
.G 0

ERM23 device-level module,
4-37 to 4-49
ERM23 module, 4-14
ERM23 notes file, 4-31
error log packet format, 5-2
error log subpacket format,
5-2

Full format report, 3-16 to
3-17
notes module, 4-67 to 4-68
NRM23 notes module, 4-67 to
4-68
Register report, 3-19
RM02/03
register printing, 4-27
RM02/03 module, 4-38 to 4-49
RM03 module, 4-13
Example ERM23 module, 4-15
Executive error logging support,
1-6
Executive routines
error collection
device registers, 1-3
Executive Send Message
directive
See SMSG$
/F switch, 3-5
FILE statement, 5-37 to 5-38
FIND statement, 5-39
FINLMl, 4-6
FINLPl, 4-6
FMlNMl, 4-11
FMlWMl, 4-6, 4-11
FM2CM1, 4-6, 4-11
FM3CM1, 4-6, 4-11
FM4NM1, 4-6

.ICF module, 4-1
IF-THEN-ELSE 8~a~ement, 5-34
INCREMENT statement, 5-31
INITMl I 4-7
Installation
RPT, 3-1
Intermodule variable, 4-29
block number, 4-29
cylinder error, 4-29
device function, 4-29
drive serial number, 4-29
drive type, 4-29
error type, 4-29 to 4-30
group error, 4-29
head error, 4-29
MASSBUS, 4-33
Non-MASSBUS, 4-33
physical unit number, 4-29
sector error, 4-29
Intermodule variable
declaration, 4-14
Interrupt timeouts
error logging, 1-6
LBR
module name requirement, 4-17
LBR command line
device-level module, 4-37
DEVSMl, 4-36
DSP2Ml, 4-36
DSP2Pl, 4-36
LEAVE statement, 5-35
Lexical conditionals, 5-36
/LIM switch, 2-3
/LIM switch, 2-6
Limit switch, 2-3, 2-6
LIST statement, 5-39
Literal declaration, 4-12
LITERAL statement, 5-33

Index-6

INDEX
$NSIER, 4-22
NT0310, 4-9
NTSll, 4-9
Numeric conversion function,
5-17 to 5-18

/LOG switch, 2-3, 2-5, 2-8
Log switch, 2-3, 2-5, 2-8
New version subswitch, 2-3,
2-6
No limit subswitch, 2-3, 2-5
to 2-6
$LOGER, 4-23
LOGTST
ERROR routines, 4-23
%LOK, 5-21
Lookahead function, 5-21
• LST module, 4-1
LUN
calculation, 4-24
LX:, 3-2

Operating system failure, 1-1

Mass storage device errors, 1-1
MCR SSM command
See SSM command, 1-3
Memory errors, 1-1
error logging, 1-6
pregenerated systems, 1-6
MESSAGE statement, 5-41
Messages
ERLCFL, 5-44
Mixed MASSBUS configuration,
4-33
MODULE statement; 5-32
MSCP60, 4-9
MSCP80, 4-9
MSCPAT, 4-9
MSC PE, 4-9
MSC PEN, 4-9
MSCPSD, 4-9
MSCPTO, 4-9
Named variable, 5-8
NMLll, 4-9
No limit switch, 2-3, 2-6
No log switch, 2-3
/NOLIM switch, 2-3, 2-6
/NOLOG switch, 2-3
Nonsense interrupt logging,
4-22
Notes module
exit, 4-33
heading, 4-32
MODULE statement, 4-32
NMLll, 4-9
note name, 4-32
NRK6 7, 4-9
NRM05, 4-9
NRM2 3, 4-9
NT0310, 4-9
NTSll, 4-9
print declaration, 4-32
print number, 4-33
PROCEDURE statement, 4-32
unknown note, 4-33
user written, 4-31
NRK6 7 I 4-9
NRM05, 4-9
NRM23, 4-9

/P switch, 3-5, 3-8
Packet information function,
5-22
Packet number, 3-1
Packet number switch, 3-5, 3-8
Packet selection switches, 3-8
PACKET statement, 5-28
PARSlM, 4-4
PARS2M, 4-4
PARS3M, 4-4
PARSEM, 4-4
%PKT, 5-22
POINTER statement, 5-38 to 5-39
PRINT FORMATTING statement,
5-41
keyword string, 5-41 to 5-42
PROCEDURE statement, 5-33 to
5-34
Processor priority, 4-20
PRS2AM 1 4-4
PRS2BM, 4-4
PUT statement, 4-31, 5-39
$QERMV, 4-25
QUPKT, 4-24
/R switch, 2-4, 2-7, 3-6, 3-27
Register report, 3-18
Report control function, 5-22
Report Generator
See also RPT
Report switch, 3-6, 3-27
Reset switch, 2-4, 2-7
RETURN statement, 4-31, 5-33
RM02/03 module, 4-37 to 4-49
RM02/03 notes module, 4-67 to
4-68
%RPT I 5-22
RPT, 1-1
command line, 3-2
input file, 3-2
multiple qualifiers, 3-3
report file, 3-2
switches, 3-2
universal library, 3-2
concurrent I/O error logging,
1-5
/DA switch, 3-5, 3-8
Date switch, 3-5, 3-8
/DE switch, 3-5, 3-8
Device switch, 3-5, 3-8
multiple qualifiers, 3-4
error log control file, 1-4
error log file, 1-4
Error log packet

Index-7

INDEX
RPT
Error log packet (Cont.)
number, 3-1
error log packets, 1-4
error type definition, 1-5
/F switch, 3-5, 3-11
Format switch, 3-5, 3-11
brief report, 3-11
full report, 3-15
No report, 3-19
Register report, 3-18
information required, 3-2
format selection, 3-2
packet selection, 3-2
summary selection, 3-2
interpreter, 5-1
/P switch, 3-5, 3-8
Packet number switch, 3-5,
3-8
multiple qualifiers, 3-4
/R switch, 3-6, 3-27
report information selection,
1-4 to 1-5
Report switch, 3-6, 3-27
/SE switch, 3-6, 3-9
~eriai number switch, 3-6,
3-9
multiple qualifiers, 3-4
/SU switch, 3-6, 3-20
subswitch summary, 3-5
Summary switch, 3-6, 3-20
error report, 3-20
geometry report, 3-23
multiple qualifiers, 3-4
no report, 3-27
summary
switch
1-: _..._ __ .. ______ .....
11 .L ~ I.. V 1 y

1 t' J:JV 1 I.. t

/SE switch, 3-6, 3-9
SEARCH statement, 5-40
SELECT statement, 5-34 to 5-35
SELTMl, 4-4
Serial number switch, 3-6, 3-9
SET statement, 5-31
/SH switch, 2-4, 2-10 to 2-11
/Show switch, 2-10 to 2-11
Show switch, 2-4
SIGNAL statement, 5-40 to 5-41
SIGNAL STOP statement, 5-41
Signal! i ng
deinition, 5-40
/SL switch, 2-4, 2-7
SMRYEP, 4-7
SMRYGP, 4-7
SMRYHP, 4-7
SMSG$, 1-4
Soft error
definition, 1-3
Soft error limit switch, 2-4,
2-7
SSM command, 1-3
%STR, 5-23 to 5-24
String conversion function,
5-19
String handling function, 5-23
to 5-24
/SU switch, 3-6, 3-20
Subpacket definition, 4-69
dispatch module, 4-69 to 4-71
DSPlMl, 4-72
DSPlPl, 4-72
DSP2Ml, 4-73
DSP2Pl, 4-73
DSP3Ml, 4-73
1""\..., '°'-,;"',

Ui:>.l:".:J.l:".J. r

.:J - t£. ~

switch summary, 3-5
/T switch, 3-7, 3-9
Type switch, 3-7, 3-9
multiple qualifiers, 3-4
/V switch, 3-7, 3-10
Volume label switch, 3-7,
3-10
/W switch, 3-7, 3-28
Width switch, 3-7, 3-28
RPT control function, 5-21
RPT default command line, 3-4
Date switch, 3-4
Device switch, 3-4
Format switch, 3-4
Packet switch, 3-4
Summary switch, 3-4
Type switch, 3-4
Width switch, 3-4
RPT default file specification,
3-3
RPT report switch
format packet, 3-5
select packet, 3-5
summarize packet, 3-5
RPTBLD.BLD file, 3-2
RSXTIME, 5-7

.-, ""\

':1:- / .:J

DSP4Ml, 4-74
DSP4Pl, 4-74
DSP5Ml, 4-74
DSP5Pl, 4-74
DSP6Ml, 4-74
DSP6Pl, 4-74
DSP7Ml, 4-75
DSP7Pl, 4-75
SUBPACKET statement, 5-29
Summary switch, 3-6, 3-20
all report, 3-20
error report, 3-20
geometry report, 3-23
history report, 3-25
no report, 3-27
/SW switch, 2-4, 2-8
Switch switch, 2-4, 2-8
Delete subswitch, 2-4, 2-9
New version subswitch, 2-4,
2-9
System Service Message command
See SSM command
/T switch, 3-7, 3-9
Table
definition, 5-36

Index-8

INDEX
Table (Cont.)
RPT use, 5-36
TABLE statement, 5-37
%TIM, 5-25
Time handling function, 5-25
Type switch, 3-7, 3-9

USERCM, 3-2
%USR, 5-25
/V switch, 3-7
VMSTIME, 5-7
Volume label switch, 3-7

Unexpected traps or interrupts,
1-6
User defined switch string,
3-28
User I/O function, 5-25

/W switch, 3-7, 3-28
WHILE/UNTIL/DO statement, 5-35
Width switch, 3-7, 3-28
WRITE GROUP statement, 5-31
WRITE statement, 4-30, 5-31

Index-9

RSX-llM/M-PLCS
Error Logging Manual
AA-L674B-TC
READER'S COMMENTS
NOTE: This form is for document comments only. DIGITAL will use comments submitted on this form at the
company's discretion. If you require a written reply and are eligible to receive one under Software
Performance Report (SPR) service, submit your comments on an SPR form.

Did you find this manual understandable, usable, and well organized? Please make suggestions for improvement.

Did you find errors in this manual? If so, specify the error and the page number.

Please indicate the type of user/reader that you most nearly represent.
[_] Assembly language programmer
lJ Higher-level language programmer
L_] Occasional programmer (experienced)
:J User with little programming experience
,__ J Student programmer
U Other (please specify)

Organization
Street
C i t y - - - - - - - - - - - - - - - - - - - - - - - - - State _ _ _ _ _ _ Zip Code _ _ _ _ __
or Country

Do Not Tear - Fold Here and Tape -

~n~nomn

No Postage
Necessary
if Mai led in the
United States

111111

BUSINESS REPLY MAIL
FIRST CLASS PERMIT N0.33 MAYNARD MASS.
POSTAGE WILL BE PAID BY ADDRESSEE

BSSG PUBLICATIONS ZK1-3/ J35
DIGITAL EQUIPMENT CORPORATION
110 SPIT BROOK ROAD
NASHUA, NEW HAMPSHIRE 03061

Do Not Tear - Fold Here