Digital PDFs

AA-H270A-TC

June 1979

86 pages

Original

2.9MB

Document:	RSX-11M/M-PLUS Error Logging Reference Manual Version 3.2/1.0
Order Number:	AA-H270A-TC
Revision:	0
Pages:	86
Original Filename:	http://bitsavers.org/pdf/dec/pdp11/rsx11m_s/RSX11M_V3.2_Jun79/6_Maintenance/AA-H270A-TC_RSX-11M_Error_Logging_Ref_Jun79.pdf

OCR Text

RSX-11 M/M-PLUS
Error Logging
Reference Manual
Order No. AA-H270A-TC

RSX-11M Version 3.2
RSX-11 M-P LUS Version 1.0

To order additional copies of this document, contact the Software Distribution
Center, Digital Equipment Corporation, Maynard, Massachusetts 01754

digiital equipmE!nt corporation · maynard, massachusetts

First Printing, June 1979
The information in this document is subject to change without notice
and should not be construed as a commitment by Digital Equipment
Corporation. Digital Equipment Corporation assumes no responsibility
for any errors that may appear in this document.
The software described in this document is furnished under a license
and may only be used or copied in accordance with the terms of such
li.cense.
No responsibility is assumed for the use or reliability of software on
equipment that is not supplied by DIGITAL or its affiliated companies.

~'I

(£') 1979 by Digital Equipment Corporation

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

DIGITAL
DEC
PDP
DEC US
UNIBUS
COMPUTER LABS
COMTEX
DDT
DECCOMM
ASSIST-11
VAX
DECnet
DATATRIEVE

DECsystem-lo
DECtape
DIBOL
EDUSYSTEM
FLIP CHIP
FOCAL
INDAC
LAB-8
DECSYSTEM-20
RTS-8
VMS
IAS
TRAX

MASS BUS
OMNIBUS
OS/8
PHA
RSTS
RSX
TYPESET-8
TYPESET-11
TMS-11
ITPS-10
SBI
PDT

1/80-14

!lllm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .llmlll................................ . .

CONTENTS
Page

PREFACE
CH.APTER

CHAPTER

INTRO DUCT ION

1-1

1.1
1.2
1. 2 .1
1. 2. 2
1. 2. 3
1. 2. 4

THE PURPOSES OF ERROR LOGGING
THE FUNCTIONS OF ERROR LOGGING
Error Logging Options
Information Gathered
The Error Log File
Formatting and Report Generation

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

HOW ERROR LOGGING WORKS

2-1

2.1

2.2.3

EXECUTIVE FEATURES
Memory Allocation Routines
Error Message Formatting Routines
Error Message Queuing Routine
TASK INTERACTION
ERRLOG Files (ERR.TMP and ERROR.TMP)
PSE File (ERROR.SYS)
SYE Report Files

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

OPERATING PROCEDURES

3-1

3.1
3.2
3.3
3.3.l
3.3.2
3.4

RUNNING ERRLOG {THE ERROR LOGGER)
RUNNING PSE (THE PRE-FORMATTER)
USING SYE (THE REPORT GENERATOR)
Invoking SYE
The SYE Command Line
RUNNING ERF {THE SHUTDOWN TASK)

3-1
3-2
3-3
3-3
3-4
3-8

THE ERROR LOG REPORTS

4-1

4.1
4.2
4.2.1
4.2.2
4.2.3
4 •.2.4
4.3
4.4

GENERATING ERROR LOG REPORTS
INDIVIDUAL ERROR REPORTS
Reports For Unexpected Traps or Interrupts
Device Error Reports
Device Interrupt Timeout Reports
Memory Parity Error Reports
OPTIONS FOR SYE RP SWITCH
SUMMARY REPORTS

4-1
4-1
4-1
4-2
4-5
4-5
4-6
4-8

ERROR LOGGING TASK MESSAGES

5-1

5.1
5. l. l
5 .1. 2
5.2
5.3
5.3.l
5.3.2
5.3.3
5.4

ERRLOG MESSAGES
Informational Messages
Error Messages
PSE MESSAGES
SYE MESSAGES
Command Line Errors
File Service Errors
Additional Messages
ERF MESSAGES

5-1
5-1
5-1
5-4
5-6
5-6
5-7
5-8
5-8

2 .1.1
2 .1. 2
2 .1. 3
2.2

2.2.1
2.2.2
CHAPTER

CH.~PTER

cm\PTER

iii

CONTENTS
Page

APPENDIX A

MODIFYING ERL AND SYE AT TASK BUILD

A-1

ERL
SYE

A-1
A-1

APPENDIX B

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS

B-1

APPENDIX C

ERROR.SYS FILE RECORD FORMATS

C-1

APPENDIX D

DEVICE DRIVER MODIFICATION

D-1

DEVICE DRIVER MODIFICATIONS
EXECUTIVE ROUTINES
$BMSET
$DVCER
$DTOER
$IODON
INT SE$
Additional Executive Routines

D-1
D-2
D-2
D-2

A.l
A.2

D.l
D.2
D.2.1
D.2.2
D.2.3
D.2.4

D.2.5
D.2.6

INDEX

D-3
D-4
D-4
D-4

Index-1
FIGURES

FIGURE

2-1
2-2

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

4-5
4-6
4-7
4-8

Executive Error Routines
Task Interaction
Report of an Unexpected Trap or Interrupt
Report of a Device Hardware Error
Report of a Memory Parity Error
Device Mounted Report
Configuration Change Report
Time Change Report
System Service Report
Summary Error Report

2-2
2-4
4-2

4-3
4-5

4-7
4-7
4-8
4-8

4-9

TABLES
TABLE

5-1

Source of Error Codes in ERRLOG Messages

5-2

PREFACE

MANUAL OBJECTIVES AND INTENDED AUDIENCE
This manual contains information needed to operate the error logging
subsystem.
It tells the user how to produce error logging reports.
These reports allow the user to monitor the hardware reliability of an
RSX-llM/M-PLUS system.
They al~o provide DIGITAL field service
eng:tneers with a convenient history of system performance.
This manual assumes a familiarity with the following documents:
The RSX-llM/M-PLUS MCR Operations Manual
The RSX-llM Utilities Manual
The RSX--llM-PLUS or
Guide

RSX-llM

System

Generation

and

Management

STRUCTURE OF THIS DOCUMENT
Chapter 1 describes the purposes and functions of error logging.
Chapter 2 describes how the Executive logging features and the
logging tasks interface to produce the final, readable reports.
Chapter 3 describes the procedures for
tasks.

operating

the

error

error
logging

Chapter 4 describes the error log reports generated by the subsystem.
Chapter 5 describes the messages generated by error logging tasks.
Appendix A explains how to modify the analyzer task SYE and the
logging task ERL at task build time.

error

Appendix B describes the record formats of temporary files produced by
the error logging task ERRLOG.
Appendix C describes record formats contained in the file ERROR.SYS.
Appendix D discusses modifications made to mass storage device drivers
to •~nable error logging. It also discusses several Executive routines
supplied for the error logging subsystem.

ASSOCIATED DOCUMENTS
The RSX-llM and RSX-llM-PLUS documentation directories define the
intended readership of each manual in the documentation set and
provide a brief synopsis of each manual's contents.
References are
made in the error logging manual to several other documents; readers
should consult the appropriate directory for information about these
documents.
The RSX-llM and RSX-llM-PLUS Mini-References contain a
section summarizing Error Logging Task operations.

CHAPTER 1
INTRODUCTION

1.1

THE PURPOSES OF ERROR LOGGING

The RSX-llM/M-PLUS error logging subsystem monitor~ the hardware
reliability of an RSX-llM/M-PLUS operating system.
It continually
receives informatie>n from the executive about every disk, DECtape,
magnetic tape,
and memory parity error, regardless of whether or not
the error is recoverable.
Error logging records these events and at
user-determined intervals, a
report-generating task can be run to
produce individual error and/or summary reports on some or all of the
errors.
Even without error logging, the device driver automatically
retries recoverable~ errors;
but after a successful recovery, the user
might be unaware that the error ever occurred.
Error logging may be used on any RSX-llM/M-PLUS system that has 24K
words or more of memory.
It is a system generation option for
RSX-llM.
Error logging reports are useful for efficient maintenance of the
hardware on which the RSX-llM/M-PLUS system runs.
Problems such as
line noise, static discharges, or inherently error-prone media,
for
instance,
can cause reco~erable errors on systems that are otherwise
functioning normally.
By studying error logging reports, the user can
learn to distinguish these kinds of errors from those that might be
symptoms of an impending device failure.
On the other hand,
some
recoverable errors that are insignificant in themselves might be
related to other, more serious errors, the effects of which are not
immediately apparent.
Information contained in the reports about each
error and about the! status of the system when the error occurred may
alert the system manager to an unforeseen hardware problem.
If the error reports seem to indicate an impending failure, the system
manager should contact a DIGITAL field service engineer who will use
the reports to help in diagnosing the problem.
Sometimes a device
fails so quickly that error log reports cannot be used to predict the
failure in time to prevent it.
In this case,
the field service
engineer can determine the cause more quickly if a report is available
which describes the errors that may have occurred immediately prior to
the failure.

1.2

THE FUNCTIONS OF ERROR LOGGING

If error logging is not activated on the system,
the RSX-llM/M-PLUS
Executive detects each hardware error, then either ignores it if it
causes no immediate problem or, when appropriate, retries the function

1-1

INTRODUCTION

that caused the error. The user normally has no means of knowing that
such an error occurred.
However,
when error logging has been
activated,
the
error logging subsystem performs the following
operations:
1.

Receives information from the Executive
about errors which were detected.

device

Stores the information in a file.

Formats the information and produces an error report.

drivers

Control of error logging is shared between routines in the Executive
module ERROR and four error logging tasks.
These routines and the
tasks ERRLOG, PSE, SYE and ERF, work together to carry out the
operations described above.

Error Logging Options

1.2.l

Error logging routines in the Executive can respond to four
types of
errors, depending on what options were chosen at System Generation:
l.

Unexpected traps or interrupts,

Device errors,

Interrupt timeouts, and

Memory parity errors.

If the error logging option was generated into the Executive and the
ERRLOG task was activated, all unused vectors are filled with pointers
to error logging routines.
Therefore,
when an unusually noisy
electrical environment or a static discharge causes an unexpected trap
or inte~rupt to occur to an unused vector, or a valid interrupt to be
vectored to the wrong address, Executive routines can determine and
record the incorrectly used vector.
Device errors are conditions that cause a device controller to
interrupt with its error bit set. When a device error occurs and
error logging is active the device driver calls the Executive to
record the contents of the device registers, which indicate the state
of the device and its controller.
In addition, these routines record
information
found
in
the actual
I/O request to aid
in the
interpretation.of the device error.
An interrupt timeout occurs when a device on which a transfer was
initiated fails
to
interrupt within a specified amount of time.
Interrupt timeouts are detected by software timers started when the
transfer
is initiated.
The system records the same information for
interrupt timeouts as
it does for device errors,
except no I/O
activity is reported.
Some main memory uses byte parity to ensure the integrity of
information.
The system generates parity for both data and addresses
on transfers to memory;
this parity is then checked on all transfers
from memory.
A parity error occurs when this check fails.

1-2

INTRODUCTION
1.2.2

Information Gathered

The error information gathered by the Executive routines (the state of
the registers when a device error occurs, for example) provides a
"snapshot" of the relevant parts of the system at the time of the
error.
In addition to all the system information, error routines
identify the type of error and the associated device
(for device
errors and interrupt timeouts), record the time the error occurred and
assigrn a sequence number to the error.
The system also generates a
System
Service Message if the MCR SSM command was used.
On
RSX-llM-PLUS systems the
error
log
report
includes
a
CPU
identification
and
information on whether the system has been
reconfigured.

1.2.3

The Error Log File

The f:irst two operations, error detection and the collection of error
information,
are continual processes done within the Executive or
device drivers.
When a certain amount of error information has been
gathE~red,
the Executive calls the error logging task ERRLOG to copy
the information to a permanent error log file.
The data may be copied
to a file on the system disk or to a file on a removable disk or
DECtape.

1.2.4

Formatting and Report Generation

When it is convenient or necessary to generate an error log
report,
the user first runs the pre-formatting task PSE to make a preliminary
consolidation of the raw data contained in one or more error log
files.
The PSE output file becomes input to SYE, the task that
actually generates the finished error logging report.
SYE is capable of generating individual error reports and/or summary
reports.
Among many other options, the user may specify a report that
covers a certain time period, a certain device or group of devices, or
a certain type of error.
See Chapter 4 for a detailed description of
the error reports that SYE can generate.
Because the raw error log data may be written to a
removable volume,
the user can generate the reports either on site or at any other
RSX-llM installation that has the PSE and SYE tasks.

1-3

ttq

-I'

CHAPTER 2
HOW ERROR LOGGING WORKS

2.1

EXECUTIVE FEATURES

The RSX-llM/M-PLUS Executive is responsible for
o

Detecting memory parity
errors,

and

unexpected

trap

interrupt

Gathering volatile data that reflects, as
closely
as
possible, the state of the system at the time of each error,
o

Controlling the amount of system dynamic memory used for
error logging, and activating the ERRLOG task which logs the
errors

A set of common error routines performs these functions for all four
types of detected errors (unexpected traps or interrupts, device
errors, interrupt timeouts, and memory parity errors).
The routines use system information made available by the Executive.
When an error occurs, an Executive routine notes the contents of the
I/O ac:tive bitmap for subsequent analysis. The bitmap could show that
the 1~rror was related to activity on some other device. If error
logging of undefined traps or interrupts has been generated into the
system, all unused vectors are filled with pointers to the error
routines. Therefore, if a device improperly interrupts through an
unused vector, an error routine is able to record the address of the
vector.
The common error routines are called every time an error occurs.
If
the error logging option for the respective error has been generated
into the system, the routines perform the following three error
logging functions:
1.

One routine allocates memory for error logging activity.

Another gathers relevant error data and formats the raw error
messages.

The last routine queues the formatted error
memory in preparation for logging by ERRLOG.

Figur1~ 2-1 is a flow chart diagram of the Executive
described in this section.

2-1

messages

routine

activity

HOW ERROR LOGGING WORKS

EXECUTIVE ERROR ROUTINES

ERROR
OCCURS

ADD 1 TO
ERROR
COUNT
MEMORY
ALLOCATION
ROUTINE

EXIT TO
CALLER
(ERROR NOT
LOGGED)

OTHER SYSTEM
AND ERROR
DATA
DEVICE
REGISTERS

1/0 ACTIVE
BITMAP
FORMATTING
ROUTINE
RAW ERROR
MESSAGE
QUEUING
ROUTINE i . . - . - - - - - - - - - - '

QUEUE FOR
LOGGING BY
TIME INTERVAL
(5 SECONDS)

YES

LOG
IMMEDIATELY

TASK
ERR LOG

ERR.TMP

Figure 2-1

Executive Error Routines

2-2

HOW ERROR LOGGING WORKS
2.1.l

'--".

Memory Allocation Routines

The memory allocation routines use the dynamic storage region
(DSR)
for
logging error information.
The routine places a limit on the
amount of DSR space available.
Normally each error results in a
record being placed in this memory space. On device errors a counter
(the error sequence number) is incremented by one. Nondevice errors,
such as parity errors, do not increment the counter. The task ERRLOG
is activated every five seconds to write to disk all the error records
in the DSR. This frees the space for future error records.
However,
when a burst of errors occurs that causes the available space to be
exceeded, the error records that caused the overflow are not placed in
the memory space.
In this case, ERRLOG is activated immediately.
Subsequent errors are not logged until ERRLOG frees DSR space by
writing the records at which time normal error recording resumes.
The
error sequence number continues to be incremented by one for each
device error even if the error is not logged. When ERRLOG writes the
error records to disk, the memory space is freed and normal recording
of error records resumes.

2.1.2

Error Message Formatting Routines

When an error occurs, the formatting routine for that type
creates a message containing the following information:

2.1.3

error

•

The error's sequence number

•

The date and time the error occurred

•

A code that classifies the error

•

Whether the device is mounted or dismounted

•

CPU identificationl

•

Configuration informationl

•

Any additional information needed to describe the error.

Appendix B contains further details about the
and formatting routines.

'-";

error

message

formats

Error Message Queuing Routine

The queuing routine provides the link between Executive error logging
activity and the task ERRLOG. After a raw message has been formatted,
the routine enters it into a memory queue, which ERRLOG periodically
writes to a file
(ERR.TMP).
ERRLOG scans the queue and logs the
contents to a file <>n disk. This logging occurs in a maximum of 5
seconds after the occurrence of an error, and sooner if a burst of
errors causes the queue to approach its limit. The queue containing
the error data is deallocated each time ERRLOG logs its contents.

l For RSX-llM-PLUS only

2-3

HOW ERROR LOGGING WORKS
2.2

TASK INTERACTION

The primary function of the error logging tasks is to maintain the
error information gathered by the Executive in order to produce a
report. ERRLOG preserves the Executive's information and passes it on
to PSE when that task is activated. PSE prepares the data for input
to SYE, which selectively generates the final, readable report.
The
:following sections describe the file activity and error logging task
interaction. Figure 2-2 demonstrates that file activity.

ERR LOG
1. WRITES ERROR MESSAGES TO RAW
DATA FILE
2. UPON REQUEST FROM PSE:
A. CREATES "SNAPSHOT" COPY
OF RAW DATA
B. DELETES CURRENT RAW
DATA
C. INITIALIZES NEW COPY OF
ERR.TMP

ERROR.TMP
2A.

COPY

28.

DELETE

2C.

INITIALIZE
NEW
ERR.TMP

PSE
REQUESTS ERRLOG TO:

1. CREATE ERROR.TMP
2. PREANALYZE RAW DATA,
GENERATING OR EXTENDING
ERROR.SYS

NOTE
DO NOT RUN PSE UNLESS
ERRLOG IS RUNNING
SVE
GENERATES SPECIFIED REPORTS

LISTING
FILES

ERF
II

ERR LOG
REPORTS

SHUTS OFF I I ERR LOG

Figure 2-2

Task Interaction

The error logging subsystem on RSX-llM/M-PLUS uses various defaults
for its file specifications.
The default device and directory are
SYO: and [1,6]. The tasks use several standard file names (including
ERR.TMP, ERROR.TMP and ERROR.SYS).
Users should not attempt to
override these defaults (although it is possible to do, so in some
cases).
One exception, however, is the device to which ERRLOG writes
the raw data file ERR.TMP;
it may be desirable to specify a
dedicated, nonsystem device for this file.
(This might be useful, for
instance, if the ERR.TMP files have to be formatted at a different
installation.)

2-4

••

BOW ERROR LOGGING WORKS
2.2.l

ERRLOG Files (ERR.TMP and ERROR.TMP)

Each time the RUN ERRLOG command is issued,
(usually at System
Startup), ERRLOG creates a file called ERR.TMP.
On RSX-llM and
M-PLUS, ERRLOG writes system configuration records to this file for
each Error Logging device.
PSE uses the configuration records to
determine the hardware configuration of the system for which PSE
preformats the data.
ERRLOG then periodically writes to the file the raw error data it
obtalns from the Executive.l By default, the data is transferred to
the file SYO:[l,6]ERR.TMP.
(1,6] and ERR.TMP are the only valid
values for the directory and filename, but the device field may be
changed at task build time.
When the user runs PSE, ERRLOG renames the ERR.TMP data file to
ERROH.TMP, makes it available to PSE and then creates a new ERR.TMP
file, which includes the required system configuration information.
Only one ERR.TMP file should exist at any given time.
NOTE
There may be more than one ERR.TMP file
if the system has crashed or has been
shut down improperly. In this case, PIP
should be used to change the name to
ERROR.TMP, but care should be taken to
preserve
the correct version number
order.

Appendix B describes the format of ERROR.TMP and ERR.TMP.

2.2.2

PSE File (ERROR.SYS)

PSE (the Pre-formatter Task) requests the file ERROR.TMP from ERRLOG,
preliminarily formats the raw error data, creates a file called
ERROH.SYS and then deletes ERROR.TMP. If one or more ERROR.SYS files
alrec1dy exist, PSE appends the newly pre-formatted data to the highest
version of the file. The data in ERROR.SYS is in a format that SYE
can use to generate its reports.
When supplying the PSE command line, the nonprivileged user must
override the output file's default device, directory and filename
(SYO:[l,6]ERROR.SYS). Attempting to write to [1,6) causes a PRIVILEGE
VIOLATION error when a nonprivileged user runs PSE.
(Note that the
input file specified in the SYE command line must be the same as the
PSE output file;
refer to Sections 3.2 and 3.3.)
Appendix D describes the format of the ERROR.SYS files in detail.

1 On an RSX-llM-PLUS system, if you reconfigure the system, the
log f:ile will include information on the new configuration.

2-5

error

HOW ERROR LOGGING WORKS
2.2.3

SYE Report Files

The output file from PSE, ERROR.SYS, becomes the input file
to SYE.
The name of the output file that SYE produces is automatically
generated. Refer to Chapter 4 for a detailed description of the
contents of the output file that SYE generates. This file is the end
product of the RSX-llM error logging software.

2-6

CHAPTER 3
OPERATING PROCEDURES

This chapter describes the operating procedures for
the three error
logging tasks
(EHRLOG, PSE and SYE) and for the task that terminates
error logging (ERF).
Error logging must be properly generated
into
the system before the user can invoke these tasks (see the RSX-llM or
RSX-llM-PLUS System Generation Manual).
Error logging requires very
little operator intervention.
After error logging has been activated either manually from a terminal
or automatically by a command in the system startup command file, its
operation is transparent to the user until error logging reports are
gene.rated.
Each error logging task issues a message whenever
it encounters a
condition
(not
related to errors
it
is logging)
that halts or
interferes with the
task's operation.
also
issue
The
tasks
informative messages
(when a
new file
is opened in response to a
request from PSE, for example).
Both types of messages are
listed,
for each of the error logging tasks.

3.1

RUNNING ERRLOG

(THE ERROR LOGGER)

ERRLOG is the only task that needs to
error logging to function.

installed

permanently

for

To install ERRLOG,
enter one of the
following commands,
either
manually from a terminal or as an entry in the system start-up command
file::
MCR> INS $ERL
The file ERL.TSK contains the ERRLOG task image.
To ensure that error
logging runs efficiently, ERL.TSK must be installed in a partition
that meets the following criteria:
•

Other tasks in the partition must be checkpointable.
If
ERL.TSK
is
installed
in a
user-controlled partition, all
other tasks must be checkpointable.

•

ERRLOG should not run in the same partition as PSE and
unless the partition is large and system-controlled.

ERF

•

ERRLOG must not run in the same partition as FllACP since
uses its services.

3-1

OPERATING PROCEDURES
To activate error logging, enter one of the following commands, either
manually from a
privileged terminal or as an entry in the system
start-up command file:
MCR>RUN ERRLOG
The System Startup Command File (LB: [l,2]STARTUP.CMD) may include the
above command.
If so, error logging is activated automatically at
system startup.
Note that ERRLOG requires that the directory [1,6] exist on the error
log device.
The directory [1,6] must be built with full access (RWED)
for PSE to run as a nonprivileged task.
The system disk is the
default error log device.
If it does not exist, ERRLOG returns the
following message:
ERL -- GET DIRECTORY ID FAILED, CODE -26
Then issue the following command to create the necessary directory:
MCR> UFO SY: [ 1 , 6]

3.2

RUNNING PSE (THE PRE-FORMATTER)

To install and run PSE, type the following:
>INS $PSE
>PSE
As distributed, PSE builds to run in an 8K partition (normally the GEN
partition),
providing enough space for 102.
devices on a mapped
system or 103. devices on an unmapped system.
To increase the amount
of space available to PSE for additional devices, append the /INC=xxx
keyword to the Install command.
(Refer to the RSX-llM/M-PLUS MCR
Operations Manual for
further information on the /INC switch to the
MCR INStall command.) PSE needs additional space at the rate of
lO(decimal)
words per unit,
where a unit is each device that could
cause a loggable error.
When running
in a user-controlled
(task)
partition,
PSE itself determines the amount of space available and
then uses the amount of space it needs.
PSE is a nonprivileged task.
However, you can change the ERL build
file to restrict PSE from being run on a nonprivileged terminal.
The directory [1,6] must be built nonprivileged for PSE to run from a
nonprivileged terminal.
Then ERL must be built so that the files it
creates in [l,6] (ERR.TMP and ERROR.TMP) can be read and deleted by
all users
(the world).
When this is done, PSE, which reads and
deletes ERR.TMP and ERROR.TMP can be run from any terminal.
If ERL is
built to generate ERR.TMP and ERROR.TMP files that cannot be read and
deleted by all users, PSE must be run from a privileged terminal.
To change the way ERL creates the files, edit the ERLBLD.CMD file
change the global definition FILPRO (the file protection word).

and

If FILPR0=60000, all users have all privileges and the world can read
and delete ERR.TMP and ERROR.TMP.
In this case, PSE can run from any
terminal.
(See Appendix A.)
(See Chapter 4 of the IAS/RSX-11 I/O Operations Reference Manual for a
description of default file protection routines and the protection
mask.)

3-2

e•e••._•

lmll. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11111111. . .

1
• • •. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

OPERATING .PROCEDURES
The pre-formatter task responds with the prompt PSE> and waits for the
user to type a command line. The format of the PSE command line is:
outdev: [ufd]file.typ=indev:
The output file in this command is described by
a
standard
RSX-llM/M-PLUS file specifier.
The input file specifier (indev:)
consists only of th~~ input device since the filename (assigned by the
task ERRLOG) is always ERROR.TMP and the UFO is always [1,6].
If you are running PSE from a nonprivileged terminal, you cannot use
the default UIC [1,6] as the output UIC. This results in a privilege
violation.
Both the call to PSE and the file specifiers may be
same line; that is~

included

the

>PSE outdev: [ufd]file.typ=indev:
The defaults for omitted fields are:
Default
outdev:
uf d
file
.typ
indev:

SYO:
[1,6]
ERROR
.SYS
SYO:

To use all the defaults for both the input and output file specifiers,
press the carriage return key in response to the PSE> prompt. When
PSE prompts again, type CTRL/Z to return to MCR.
>INS $PSE
>PSE<CR>
PSE><CR>
PSE> "z
>REM PSE

;INSTALL PSE FROM A PRIVILEGED TERMINAL
;RUN PSE FROM ANY TERMINAL
;GENERATE PRE-FORMATTED FILE
;EXIT FROM PSE
;REMOVE PSE

Note that PSE cannot be run with the RUN$ format.
It must
installed from a privileged terminal and invoked as PSE <CR>.

3.3

USING SYE (THE REPORT GENERATOR)

The following sections describe the methods of invoking SYE and
of the command line to generate the desired reports.

Invoking SYE

3.3.l

There are two methods for invoking SYE:
1.

Type the following commands at a privileged terminal:
>INS :?SYE
>SYE

Type the following command at any terminal:
>RUN $SYE

3-3

forms

OPERATING PROCEDURES
The report generator task, SYE, responds with a message and the prompt
SYE>, and waits for the user to type a command line.

For example:
>RUN $SYE
SYE>

3.3.2

The SYE Command Line

The format of the SYE command line is
outdev: [ufd]file.typ=indev: [ufd]file.typ/switchl ••• /switchn
If SYE has been installed,
both the call to SYE
specifiers may be included on the same lin~;
that is:

and

the

file

>SYE outdev: [ufd]file.typ=indev: [ufd]file.typ/switchl ••• /switchn
The default values for the output file specifier are:
Default
outdev:

SYO:

uf d

user uic

file

ERRLOG

typ

.LST

The input file is described by a standard RSX-llM file specifier;
it
must be a file previously created as an output file by PSE (see
Section 3.2).
The default values for the input file specifier are:
Field

Default

indev:

SYO:

uf d

[ l, 6]

file

ERROR

.typ

.SYS

switches

/-RP/SP

To use all the defaults for . both the input and the output file
specifiers,
press the carriage return key in response to the SYE
prompt.
When SYE prompts again, type CTRL/Z to return to the system
monitor.
For example:
SYE> <CR>
SYE> "'z

3-4

Lt.·

#••

OPERATING PROCEDURES
Every invocation of SYE produces summary error reports. These summary
error reports are described in detail in Section 4.4 of Chapter 4.
Through the use of either of two command line switches, /RP and /DV,
the user can cause additional error reports to be generated along with
the summary reports.
An option under the SYS class of the RP switch
allows you to further define the kinds of information SYE reports.
As
shown below, the /RP switch allows the creation of general classes of
error reports while the /DV switch generates reports specific to a
particular device type or device unit.
Switch
/RP [:class]

Description
The /RP switch requests individual error reports;
the class parameter is used to specify a general
classification of errors to be reported.
If no
class parameter is specified with the /RP switch,
SYE reports on all classes.
Class has one of the following values:
null

All classes.

SYS [:type]

All system errors
(error logging
startup,
power
fails,
and PSE
entries) which are significant to
DIGITAL
service groups such as
field service and software support.
It also reports on configuration
changes on RSX-llM-PLUS.
Type has
values:

one

the

following

NULL - All types of system entries
PSE

- Entries from PSE operation

STA

- Error Logging start-up

MOU

- device mounts and dismounts

*CFG - configuration information
TIM

- system time changes

MSG - system service message
* M-PLUS only
HOW

All hardware errors. When class is
HOW, an additional parameter can be
specified further
defining
the
errors
to
be
reported.
This
parameter takes the form:
/RP:HDW[:type]

3-5

OPERATING PROCEDURES
where
Type is one of the following:

TMO

null
(no entry)

All types of hardware
errors

OSK

All disk errors

MAG

All magnetic tape
DECtape errors

MEM

All cache
and
main
memory parity errors

Interrupt time out errors
disks and tapes, in the form

and

from

/RP : TMO [ : type]
Type is one of the following:
OSK

disk information only

MAG

tape information only

/SU

Create summary report

/QU

Create short summary report

/-RP
or
/NORP

Do not include individual error
produces a summary report only.
default.

reports.
/-RP is

SYE
the

/DV:dev[n]

Include in the report only those errors
occurred on a specified device type or
specified device unit.

that
on a

you

can

/DV:DK requests that error statistics on all
or RK05 units be provided

RK03

Besides the
specify:

standard

device

types,

CMM

Communications devices

UDI

Undefined interrupts

PWR

Powerfail

For example:

/DV:DKl requests that only the error
for RK03 or RK05 unit l be provided.

3-6

statistics

OPERATING PROCEDURES
The rema1n1ng switches that can be specified on the
line are:
/BEG:time-and-day

'--''

SYE

command

Include in the report only those errors
logged after the specified time and date.
The format of the time-and-date parameter is:
dd-mon-yr:hh:mm:ss
where all the numbers are decimal and leading
zeros are not required. Null equals zero.
The field terminator (:) must be entered for
a null field.
The hours
(hh)
field
is
expressed on the basis of a 24-hour clock.
Example:
/BEG:22-NOV-78:20:30:0 includes those errors
that occurred after 8:30 PM on November 22,
1978.
The default is to include all errors of the
specified type, no matter when they occurred.

/END:time-and-date

Include in the report only those errors
logged prior to the specified time and date.
The format of the time-and-date parameter is:
dd-mon-yr:hh:mm:ss
where all the numbers are decimal and leading
zeros are not required. Null equals zero.
The field terminator (:) must be entered for
a null field. ' The hours
(hh)
field
is
expressed on the basis of a 24-hour clock.
Example:
/END:l8-0CT-78:18:00:00 includes those errors
that occurred before 6:00 PM on October 18,
1978.
The default is to include all errors of the
specified type, no matter when they occurred.

/HEL[P]

The /HELP switch causes SYE
operating instructions.

display

SYE

To obtain
this
information,
the
user
specifies the /HELP switch on its own in
response to the SYE prompt as follows:
SYE>/HELP <CR>
/SP

Spool output file.
This is the default.
No
file
is printed if spooling has not been
generated into the system.

3-7

OPERATING PROCEDURES

SYE command line examples:

•

>SYE<CR>
SYE>
Produces file ERRLOG.LST containing a summary report of all
errors in file SYO:[l,6] ERROR.SYS. Spools ERRLOG.LST to the
line printer.

)SYE /RP<CR>
SYE>
Produces file ERRLOG.LST containing individual reports and a
summary of all errors in file SYO:[l,6]ERROR.SYS. Spools
ERRLOG.LST to the line printer.

)SYE /DV:DBO<CR>
SYE>
Produces file ERRLOG.LST containing individual reports and a
summary of only DBO
(RP04, RPOS, RP06)
errors in file
SYO:[l,6] ERROR.SYS. Spool ERRLOG.LST to the line printer.

3 .• 4

RUNNING ERF

(THE SHUTDOWN TASK)

To terminate error logging, run the task ERF by typing
pair of commands from a privileged terminal:
>INS $ERF
>ERF
Note that the command RUN $ERF is not acceptable.

3-8

the

following

CHAPTER 4
THE ERROR LOG REPORTS

4.1

'-"'''

GENERATING ERROR LOG REPORTS

To generate error log reports, you must first run PSE from any
terminal, unless you changed the file protection mask for ERR.TMP in
the ERLBLD file to restrict PSE from running from a nonprivileged
terminal.
In that case, the terminal must be privileged. PSE
performs a preliminary formatting of the data in files created by the
error logging task ERRLOG. The default PSE output file (ERROR.SYS)
contains the error data which becomes the input file to SYE.
SYE, the Report Generator, can be run from any terminal. When the PSE
output file is available, the user must decide what kind of reports
are needed and then select the appropriate switches for the SYE
command line. For example, the user can select a time frame that the
report is to encompass and can specify that the report include only
those errors associated with a certain device type or unit. Refer to
Section 3.3 for a description of SYE operating procedures and all the
available command line switch options.
SYE produces an error report in the form of a printed listing or a
listing file or both.
By default, SYE creates a file called
ERRLOG.LST on the system disk; this file contains summary information
on all of the errors recorded in the ERROR.SYS file. Both PSE and SYE
are generated if you select these options in Phase 3 of System
Generation.

4.2

INDIVIDUAL ERROR REPORTS

Error logging detects and records four kinds of errors:
unexpected
traps or interrupts, device errors, interrupt timeouts, and memory
parity errors (refer to Section 1.2.1).
This section illustrates
sample reports on each kind of error.

4.2.1

Reports For Unexpected Traps or Interrupts

Figure 4-1 is a report of an interrupt through an unused vector. Such
interrupts are also called traps in SYE reports. The description of
the report refers to each line of text by number, where the first line
described is line l.

4-1

THE ERROR LOG REPORTS
COMMAND
1
2
3

************************** ENTRY 12. *****************************
UNEXPECTED TRAP OR INTERRUPT
LOGGED 28-JAN-79 16:24:00
ON -- CPA-- Configuration Entry #121
*******************************************************************

4
5
6
7
8

SOFTWARE STATUS
VECTOR WHERE TRAP OR INTERRUPT OCCURRED
PC
PRIOR THIS ERROR
PSW
PRIOR THIS ERROR
ERRORS MISSED WHILE THIS ERROR LOGGED

9
10

DEVICE I/O ACTIVITY
204

200
177000
000004
0

1 RSX-llM-PLUS only
Figure 4-1

Report of an Unexpected Trap or Interrupt

•

Line 1 gives the record's entry number. The first error to
occur after error logging has been initialized is assigned an
entry number of 1. The entry number is then incremented by 1
every time a loggable error occurs.

•

Line 2 describes the type of error.

•

Line 3 gives the date and time the error occurred, and,
on
RSX-llM-PLUS systems,
identification information on the CPU
and system configuration changes.

The next block of text (lines 4 through 8) contains information
the software status of the trap or interrupt:
which

•

Line 5 provides the address of the vector through
interrupt occurred.

•

Line 6 is the value of the program counter immediately
to the error.

prior

•

Line 7 is the processor status word immediately prior to
unexpected trap or interrupt.

the

•

Line 8 notes
system was
detection of
UNIBUS noise

the

the number of errors that occurred while the
logging this error.
(This count permits the
bursts of illegal interrupts possibly caused by
corruption.)

The last block of data, headed "DEVICE I/O ACTIVITY",
lists all
vectors with I/O activity on the UNIBUS at the time of the error.
list includes the interrupt address of the device that caused
error.

4.2.2

about

the
The
the

Device Error Reports

Figure 4-2 is a typical error log report on a device error.
(The
figure also serves as a sample interrupt timeout report, since the
report format is identical for both types of error.)
The following
description of the device error report refers to each line of text by
number, where the first line described is line 1.

4-2

.,..__.,._______________________________..

_mll~-------------

lml'

THE ERROR LOG REPORTS
1
2
3

*************************** ENTRY 36. *****************************

DEVICE HARDWARE ERROR
LOGGED
17-FEB-79 13:39:59

ON CPA -- Configuration Entry #21

*******************************************************************
4
5
6
7
8

UNIT IDENTIFICATION
UNIT LOGICAL NAME
UNIT PHYSICAL NAME
DEVICE TYPE
VOLUME LABEL

9
10
11
12
15

SOFTWARE STATUS
TASK NAME
TASK UIC
TASK START ADDRESS
IO FUNCTION ISSUED

16
17
18
19
20
21
22

DEVICE REGISTERS
RKDS
004723
RKER
000002
RKCS
100744
RKWC
000000
RKBA
132211
RKDA
000023

23 MEDIA ADDRESS
24
CYLINDER
25
TRACK
26
SECTOR
27
LOGICAL BLOCK

DKO
DKO
(CONTROLLER-0
UNIT-0)
RK05/RK05F/RK03 (DISK)

••• VFY
1,1
373400
READ

CHECKSUM ERROR
READ

AT IO STAR'r
0
1

AFTER ERROR
0
1

28
29

ERROR DIAGNOSIS
NOT RECOVERED AFTER 9. RETRIES

30
31

DEVICE I/O ACTIVITY
220

1 RSX-llM-PLUS only

Figure 4-2

Report of a Device Hardware Error

The first two lines of text, offset by two
information about the error:

star

lines,

give

header

1 gives the
record's entry number and identification
• Line
information on the CPU and configuration for RSX-llM-PLUS

systems.
•

Line 2 describes the type of error

•

Line 3 gives the date and time the error occurred.

The next block of data
device:
•

(UNIT

(DEVICE HARDWARE ERROR).

IDENTIFICATION)

describes

the

erring

Line !> (UNIT LOGICAL NAME) shows the device mnemonic (DK)
and
unit logical number (0).
The unit logical number ranges from
0 to n, where n is an octal number
representing the total
number,
minus
one,
of such devices supported by all
controllers at a particular installation.

4-3

THE ERROR LOG REPORTS
•

Line 6 (UNIT PHYSICAL NAME) shows the device mnemonic,
the
physical unit number, and the controller number.
The physical
unit number can range from 0 to 7 for each controller.

•

Line 7 (DEVICE TYPE) shows the DIGITAL
(RK05/RK05F/RK03).

•

Line 8 provides the volume label
identification for RSX-llM
systems.
(If there is no volume label, the field remains
blank.)

name

for

the

device

The next block of data (SOFTWARE STATUS) describes the user task
initiated the I/O request that failed.
•

Line 10 provides the 6-character name of
initiated the erring I/O request ( ••• VFY).

the

task

that
that

11 is the User Identification Code (UIC) under which
• Line
task was running.

the

12 shows the task's physical start address, that is,
• Line
base address of the task in memory.

the

• Line 15 describes the I/O function requested by the task.
The report then describes, in lines 16 through 22, the state of the
device registers at the time of the error. The first column lists the
register names;
the second shows the contents of the registers.
The
device register names are those used in the PDP-11 Peripherals
Handbook. The third column, following some registers
(lines 18 and
19)
explains the meaning of the error bits in register RKER and RKCS,
respectively.
NOTE
The function that failed,
as described
by the register contents in lines 16-22
may differ from the function requested
by the task.
For example, if a task requests a WRITE
to a DECtape, the driver must perform a
READ BLOCK NUMBER function, which could
then cause an error.
The next block of text (MEDIA ADDRESS) shows the physical disk address
at the start of the I/O operation and at the time of the error (this
section does not appear for DECtape, magnetic tape,
and cassette
errors).
Line 27 (LOGICAL BLOCK) specifies the logical block number
requested at I/O start time and the logical block number computed from
the device registers at error time.
The next block, (ERROR DIAGNOSIS), (lines 28 and
29)
show that the
error was not successfully recovered and indicates how many times the
device driver retried the operation. This line alternatively shows
the number of times the driver retried the error if recovery was
successful.
The last block of text (lines 30 and 31) lists the vectors with active
data transfers on the UNIBUS at the time of the error. The list
includes the interrupt vector address of the device that caused the
error.
The number on the second line (220) is the vector address

4-4

THE ERROR LOG REPORTS
through which a transfer complete or data
fielded for the device.
~

error

interrupt

would

The block headed "VECTORS WITH ACTIVE IO" does not appear if there was
no I/O activity when the error occurred. Also, error reports on
interrupt timeouts do not include this information because the long
time element involved makes it irrelevant.

Device Interrupt Timeout Reports

4.2.3

Interrupt timeout reports have the same format as device error
reports.
Refer to Section 4.2.2 for a description of the various
entries.

Memory Parity Error Reports

4.2.4

Figure 4-3 illustrates a report of a memory parity error.
The
following description of the report refers to each line of text by
number, where the first line described is line 1.
1
2
3

****************************** ENTRY 524. ***********************
MEMORY PARITY ERROR
LOGGED
5-JUN-79 13:03:36

******************************************************************
4
5

6
7
8
9
10
11

SOFTWARE STATUS
TASK NAME
TT61
TASK START ADDRESS
721200
TASK SIZE (WORDS)
7000
PC PRIOR TO THIS ERROR 121212
PSW PRIOR TO THIS ERROR 170000
PARTITION NAME
GEN
PARTITION START ADDRESS 721200

12
13
14
15
16
17
18
19
20
21

MEMORY REGISTERS
(11/70 TYPE)
LOW ERROR ADDRESS
HIGH ERROR ADDRESS
MEMORY SYSTEM ERROR

22
23

ERROR DIAGNOSIS
MEMORY PARITY ERROR IN LOCATION 722410

WITH MEMORY MANAGEMENT REGISTER 5 ACTIVE IN USER MODE

MEMORY CONTROL
MEMORY MAINTENANCE
MEMORY HIT/MISS

Figure 4-3

122410
000003
104404

000001
000000
000043

MAIN MEMORY EVEN WORD
ERROR IN MAINTENANCE
CPU ERROR
CPU ABORT
DISABLE TRAPS

Report of a Memory Parity Error

The first three lines of text, offset by two star lines,
information about the error:
•

Line 1 gives the record's entry number.

4-5

give

header

THE ERROR LOG REPORTS
•

Line 2 describes the type of error (MEMORY PARITY ERROR).

•

Line 3 gives the date and
supplies identification on
for RSX-llM-PLUS systems.

time the error occurred, and
the CPU and system configuration

The next block of data (SOFTWARE STATUS) describes the user task
was active when the error occurred:
•

Line 5 shows the name of the task.

•

Line 6 gives the task's physical start address;
base address of the task in memory.

that

that is,

the

•

Line 7 gives the size of the task in words.

•

Lines 8 and 9 give the task's Program Counter (PC)
Processor Status word (PSW) before the error occurred.

and

•

Line 10 and 11 provide the name of the partition containing
the task, and the physical start address of the partition.

The next block (MEMORY REGISTERS) describes the state of the memory
registers at the time of the error. The register names listed in the
first column are those used in the PDP-11 Processor Handbooks.
The
second column shows the contents of the registers; and the third
column describes individual bits in the MEMORY SYSTEM ERROR register
and the MEMORY CONTROL register.
PDP-11/40 and /45 Processors Only - When a report describes a memory
parity error that occurred on a PDP-11/40 or /45 processor, the report
displays the address as well as the contents of each register.
Lines 22 through 24 provide an error diagnosis:
•

Line 23 gives the physical location in memory in which the
parity
error
occurred.
PDP-11/40
and /45 Processors
Only - When the parity error occurred on a PDP-11/40 or /45
processor, this line gives the range of addresses in which the
error occurred, rather than a precise location. For example,
the line might read:
PARITY ERROR IN LOCATIONS 210000 TO 213777

Line 24 appears for mapped systems only:
•

4.3

This line gives the number of the memory management register
group active at the time of the error and the operating mode
of the task that was active when the error occurred.

OPTIONS FOR SYE RP SWITCH

Figures 4-4 through 4-8 are examples of the reports SYE generates with
various options for the RP switch.

4-6

THE ERROR LOG REPORTS
1 **************************** ENTRY 197. *******************************
2 DEVICE MOUNTED MESSAGE
3 LOGGED
22-DEC-78 01:35:52 ON CPA -- Configuration Entry #21

"-"

.•

***;r*****************************************************************
4 DEV JC CE MOUNTlm-- DKO: DBLWAK

UIC= [ 1, 1]

UCB ADDR.=022252

1 RSX-llM-PLUS only
Figure 4-4

•

Device Mounted Report

Line 2 shows the option with the RP:SYS switch
generate the message.
In this case,
the
included:

which caused SYE to
SYE command could have

/RP:SYS:MOU
/RP:SYS
or
/RP
Line 3 gives the day and time the action occurred
RSX-llM-PLUS, a message on the system configuration.

and,

Line 4 gives specific information on which device was mounted.
1 *************************** ENTRY 204. *******************************
2 SYSTEM CONFIGURATION MODIFIED
3 LOGGED
22-DEC-78 03:20:11 -- CONFIGURATION ENTRY #2.1

*********************************************************************
4 TT56:

YHC15:

ONLINE,ACCPATH,MTD-LGD,DRIVER

1 RSX-llM-PLUS only
Figure 4-5

Configuration Change Report

Line 2 shows the option with the RP:SYS switch that caused SYE to
generate this message. The following SYE commands can generate this
report:
/RP:SYS:CFG
/RP:SYS
/RP
Note that this switch (CFG) only works on RSX-llM-PLUS systems,
include re-configuration.

which

Line 3 shows the day and time the action occurred and the entry number
for the last configuration change.
Line 4 shows the actual change that was made in the system.

4-7

THE ERROR LOG REPORTS
1 **************************** ENTRY 196.
2 SYSTEM TIME CHANGE
3 LOGGED 22-DEC-78 01:00:33

*******************************

**********************************************************************
Figure 4-6

Time Change Report

Line 2 shows the option with the RP:SYS switch that caused SYE to
generate this report.
The following SYE commands can generate this
report:
/RP:SYS:TIM
/RP:SYS
/RP
1 ************************** ENTRY 3.
2 SYSTEM SERVICE MESSAGE
3 LOGGED
ll-JUN-79 10:16:17

*******************************

*******************************************************************
4 TASK INITIATING MESSAGE:
5 THE MESSAGE IS BELOW:

••• MCR

6 THIS TEXT WAS INSERTED IN THE ERROR LOG REPORT BY AN MCR SSM COMMAND

Figure 4-7

System Service Report

Line 2 shows the option with the
message.

switch

that

caused

SYE

Line 3 shows the day and time the action occurred.
Line 4 shows the task that initiated the message;

in this case MCR

Line 6 is the message text.
The following SYE command can generate this report.
/RP:SYS:MSG
/RP:SYS
/RP

4.4

SUMMARY REPORTS

Figure 4-8 is an example of the summary section of a report.
This
part of the report contains summary information about all the errors
described in the report as a whole (but not necessarily about all the
errors in the input file to SYE).

4-8

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
1
.

THE ERROR LOG REPORTS

*********************************************************************
SYSTEM ERROR REPORT SUMMARY TOTALS

*********************************************************************
2
3
4
5
6
7
8
9

DEVICE ERRORS
MEMORY PARITY ERRORS
CACHE PARITY ERRORS
INTERRUPT TIMEOUT ERRORS
UNEXPECTED TRAPS OR INTERRUPTS
POWER FAILS
ENTRIES MISSING (SATURATION)
TOTAL ERROR CPO COUNT 1

COMMAND LINE USED
INPUT l?ILE
OUTPUT FILE
DATE OF FIRST ENTRY
DATE OF LAST ENTRY
ENTRIES PROCESSED
UNRECOGNIZABLE ENTRIES

12
13
14
15
16

80.
O.
0.
O~

O.
O.
O.
80.
ERKOS.LST/-SP=[304,324]DKDBDS.SYS
SYO: [304,324]DKDBDS.SYS;l
SY0:[304,324]ERK05.LST;2
17-FEB-77 05:58:16
17-FEB-77 15:39:08
94.

**********************************************************************
17

DETAILED DEVICE ERROR REPORT SUMMARIES

**********************************************************************

18
19
20
21
22
23

PROCESSOR!
UNIT LOGICAL NAMEl
UNIT PHYSICAL NAMEl
DEVICE TYPEl
VOLUME LABELl

25
26
27
28
29
30
31

UNIT LOGICAL NAME
UNIT PHYSICAL NAME
DEVICE TYPE

CPA
DS2
DS2
CONTROLLER-0
RS04
JEPBACKUP

UNIT-2

DATA LATE

[HARD]
O.

DKO
DKO
CONTROLLER-0 UNIT-0
RK05/RK05F/RK03
[HARD]
MOTION/SEEK
o.
CHECKSUM OR CRC
9.
DATA LATE
9.

1 RSX-llM-PLUS only

Figure 4-8

Summary Error Report

4-9

[SOFT]
35.

[SOFT]
1.

o.
0.

THE ERROR LOG REPORTS
Lines 2 through
statistics:

the
number

summary

report

provide

various

described

error

•

Line 2 is the
report.

device

errors

•

Line 3 records the total
described in the report.

number

•

Line 4 shows the total number of cache parity errors.

•

Line 5 is the number of interrupt timeouts described
report.

•

Line 6 is the number of unexpected traps or interrupts.

•

Line 7 gives the number of power fails.

•

Line 8 is the number of errors that were not logged because
another error was already being recorded.
Lines 10 through 16
of the report summarize details of the execution of SYE.

•

Line 9 provides the total errors which
processor for an RSX-llM-PLUS system.

•

Line 10 is the command line issued to SYE.

•

Line 11 describes the complete file specifier of the input
file and line 12 describes the complete file specifier of the
output file.

•

Lines 13 and 14 give the dates and times of the first and last
entries in the input file.

•

The number of entries formatted in the report appears in
15.

•

Line 16 is
SYE.
An
version of
one of the

memory

parity

occurred

this
errors

this

each

line

the number of unrecognizable errors encountered by
unrecognizable error is any entry that the current
SYE cannot format.
This situation can occur for
following reasons~

An old version of SYE has been run.

The entry refers to a device not supported by SYE.
Such
an entry may be encountered if the site has implemented
error logging on a device SYE does not recognize.

SYE has encountered a data structure (field format) error.
Such encounters indicate that the wrong version of the
pre-formatter PSE was used.

Line 17 marks the beginning of a series of detailed device error
report
summaries.
Each such summary provides error statistics
relati~g
to an individual device.
Figure 4-8 shows two
such
summaries,
the first on lines 18 through 24, and the second on lines
25 through 31.
•

Line 18 shows the processor
(on
which the report was generated.

RSX-llM-PLUS

•

Lines 19 and 25 show the unit logical name.

systems)

for

4-10

...............................................................................................u•u......-..tt•••u";~._..,

THE ERROR LOG REPORTS
•

Lines 20 and 26 give the physical unit numbers and
numbers of the respective devices.

•

Lines 21 and 27 show the DIGITAL name for the device.

•

Line 22, on RSX-llM-PLUS systems, provides
for the device.

the

controller

Volume

Label

The remaining sections in each detailed device error summary (lines 21
and· 24 in the first instance and lines 28 through 31 in the second)
show the number of hard (non-recovered) and soft
(recovered)
errors
described in this report.
The types of errors which can appear in
these sections, examples of which appear on lines 24 and
29 through
31, are described below.
Headi!!,2

Description

l CHECKSUM OR CRC

The controller detected either a data
checksum
error or a hard error-check
correction error.

2 CONTROLLER PARITY

The hardware controller detected a
error within itself.

3 DATA LAT.E

The controller detected a data
error
caused by one of two possible situations:

parity

•

Data was ready for memory when memory
had not emptied its hardware buffer
(silo), or

•

The controller was ready for data,
but
memory's
silo
had no data to be
written.

4 DATA PARITY

The controller
error.

5 DEV-TO-MEM PARITY

The controller detected a parity error
while the controller was performing a
Write or Write-Check command.

6 DRIVE

A drive specific error,
for
drive timing error, occurred.

7 DRIVE SELECT

A drive selection error,
such as the
selection of multiple heads, has occurred.

8 HEADER

The controller detected a unit error that
prevents
a
successful data transfer.
Specific errors
include search errors,
header compare errors, and format errors.

9 INTERRUPT TIMEOUT

An initiated operation did not cause an
interrupt
within
a
specified
time
interval.

10 MOTION/SEEK

The
controller
detected
mechanical
problems on a device unit, which cause the
unit to produce errors such as
seek
incomplete,
seek
error,
or operation
incomplete.

4-11

detected

data

parity

example

THE ERROR LOG REPORTS

Heading

Description

11 NONEXISTENT MEMORY

The controller
error.

12 POWER ERROR

The controller detected an
to power difficulties.

13 SOFTWARE ERROR

Corrupted service
controller error.

14 UNIT UNSAFE

The controller detected an error that
caused a selected device to be unusable.

15 WRITE CHECK

A write check error occurred.

16 WRITE LOCKED

A write operation
write-locked unit.

detected

bus

timeout

error

routines

was

caused

attempted

17 UNCATEGORIZED ERRORS Any error that does not fit in one of
above categories.

4-12

a
the

CHAPTER 5
ERROR LOGGING TASK MESSAGES

5.1

5.1.l

ERRLOG MESSAGES

Informational Messages

The following two messages are informational
ERRLOG's normal operation.

messages

issued

during

ERL -- ERROR LOG INITIALIZED
The file ERR.TMP has been opened and initialized successfully.
The message occurs either when the RUN ERRLOG command is issued.
or after ERRLOG has renamed the current ERR.TMP file to ERROR.TMP
and opened a new ERR.TMP.
ERL -- LOGGING ENDED AFTER ddd ERRORS
ERRLOG issues this message when error logging is shut down, where
"ddd" is the number of errors that were logged to the current
file. This count will have been zeroed if PSE had received a
file from ERRLOG while logging was active; therefore it does not
necessarily indicate the total number of errors logged since
ERRLOG was initialized. Note that if ERRLOG is shut down and an
error occurs, the error count is updated but the error is not
logged.

5.1.2

Error Messages

The error logging subsystem terminates in response to an error
condition unless the description of the associated error message
states otherwise. All error codes are decimal numbers.
See the
IAS/RSX-11
I/O Operations Reference Manual, Appendix I, for a
description of the error codes.
Table 5-1 illustrates sources for error logging messages.

5-1

ERROR LOGGING TASK MESSAGES
Table 5-1
Source of Error Codes in ERRLOG Messages
Kind of number

Source
File Control Service (FCS)

Negative decimal number

Driver/File Control Primitives (FCP)

Positive decimal number

File Descriptor Block (FOB)

Positive decimal number

ERL -- ASSIGN LUN FAILURE
Explanation: An error occurred when ERRLOG tried to assign LUN 4
while creating file ERR.TMP. Logging is not initialized.
User Action:

Retry the ERRLOG operation.

ERL -- DEVICE FULL. MOUNT NEW VOLUME OR
"REA ERRLOG 4 DDU:",THEN "RUN ERRLOG"
Explanation: While trying to write an error to the file, ERRLOG
found the device full.
The buffer containing the most recent
error logged is requeued in memory, and error logging is shut
down.
User Action:
In order to reinitialize logging, either
a
different device with available space must be used, or the full
volume must be dismounted, a new one mounted and the task ERRLOG
restarted;
the system will then reinitialize error logging.
(There may be errors queued in the dynamic pool ready to be
logged.)
ERL -- DEVICE RECORD PUT ERROR, CODE ddd
Explanation: During ERR.TMP initialization, a file system
was detected in writing a device configuration record.

error

ERL -- ERROR ON REOPEN, CODE ddd
Explanation: ERRLOG was unable to open an existing, initialized
ERR.TMP file in order to write an error message to it.
ERL -- ERR.TMP NOT RENAMED (ERR.TMP:

ddd, ERROR.TMP:

ddd)

Explanation: ERRLOG failed to rename the ERR.TMP file either
when passing the file contents to PSE or when error logging was
being shutdown. The codes are from the respective files' File
Descriptor Blocks (FDBs)
(the code may be 0 for ERROR.TMP,
indicating the problem was only with ERR.TMP).
In this case,
error logging is not shutdown since a new file will be created.
User Action:
Once the reason for the rename failure
is
determined, the file may be renamed to ERROR.TMP using PIP, and
subsequently formatted (see the RSX-llM Utilities Manual).

5-2

) . $I

mil

ERROR LOGGING TASK MESSAGES
ERL -- ERR.TMP OPEN FAILURE, CODE ddd
Explanation:
The file system reported
a
failure
during
initialization when ERRLOG tried to create the file ERR.TMP.
ERL -- FILE CLOSE ERROR, CODE ddd
Explanation: After ERRLOG wrote
initialization information or
error
records,
FCS returned an error when it tried to close
ERR.TMP. The file will be locked and truncated;
error
records
may therefore be missing from the end of the file.

User Action:
Use PIP to unlock the file for
formatting or for
further
logging
if the error is not persistent (see the RSX-11
Utilities Manual) •
ERL -- GET DIRECTORY ID FAILED, CODE ddd
Explanation:
Failure during creation of a new ERR.TMP file.
If
the code is -26, the directory [1,6) does not exist on the system
disk.
(Refer to Section 3.1.)
User Action:

Create UFO and retry operation.

ERL -- INITIALIZATION PUT ERROR, CODE ddd
Explanation:
During file initialization, an error
the first attempt to write to the ERR.TMP file.

occurred

ERL -- MARKED FOR REMOVE, LOGGING NOT STARTED
Explanation:
This message will appear if the command RUN $ERL
has been issued to start up error logging.
ERRLOG must be
permanently installed;
therefore it may not be executed by the
default "install/request/remove" method of MCR.
User Action:
Restart error logging using commands in Section 3.1
of this manual.
ERL -- NO ERROR FILE
Explanation:
This error is unlikely to occur.
It indicates that
some task has zeroed an internal file ID pointer.
ERRLOG exits
when it encounters this situation.

'-"''

User Action:

Rerun ERRLOG to activate error logging.

ERL -- OUTPUT ERROR, CODE ddd
Explanation: This error was caused by an attempt to write data
to the file ERR.TMP after the occurrence of a loggable error.
ERL -- TASK NAME NOT "ERRLOG"
Explanation:
This message appears when the user tries
to
initialize ERRLOG after the task has been installed incorrectly.
User Action:
"ERRLOG 11 •

Install the

error

5-3

logging

task

under

the

name

ERROR LOGGING TASK MESSAGES
ERL -- TERMINAL NOT PRIVILEGED

5.2

Explanation:
The terminal from which ERRLOG is initialized
be privileged.

must

User Action:
operation.

Log

privileged

terminal

and

retry

the

semantic

error

when

input.

Retype

the

input

file

PSE MESSAGES

PSE -- COMMAND STRING PARSE ERROR
Explanation: PSE encountered a syntax
examining an input command line.
User Action:
PSE prompts
corrected command line.

for

further

PSE -- DELETE ERROR ON INPUT FILE
Explanation:
When PSE finished
processing
ERROR.TMP, the task was unable to delete it.
User Action:

the

Use PIP to delete ERROR.TMP.

PSE -- ERROR LOGGER DID NOT PROVIDE A FILE
Explanation:
PSE was able to communicate with ERRLOG, but ERRLOG
did not make a file available within 30 seconds. This message is
also displayed when the RUN$PSE command format
is used to
indicate that PSE cannot run.
User Action:

Run PSE again in the form:

PSE -- lST RECORD WAS NOT AN
PRE-ANALYZED

INITIALIZATION

INS $PSE <CR>
RECORD

FILE

PSE<CR>
CANNOT

Explanation:
The first
record of
a
file
must
be
an
initialization record.
Both the input and the output files are
closed. The input file is not deleted.
User Action:
Rerun PSE, making sure you are giving PSE the
correct input file.
In other words, the file must be output from
ERRLOG and not output from PSE or SYE.
If the error occurs
again, delete the oldest version of ERROR.TMP.
PSE -- INPUT FILE ERROR
Explanation: An error was encountered when PSE tried to open or
obtain data from the input file ERROR.TMP. The input and output
files are closed. The input file ERROR.TMP is not deleted.
User Action:
ERROR.TMP.

Rerun PSE.

the

error

occurs

again,

delete

PSE -- OUTPUT DEVICE IS FULL
Explanation:
The output device became full while PSE was writing
to the output file.
Both the input and output files were closed.
ERROR.TMP is not deleted.
User Action:

Select another volume for the output file.

5-4

ERROR LOGGING TASK MESSAGES
PSE -- OUTPUT FILE ERROR
gxplanation:
An error was encountered while PSE was accessing
the output file.
Both the input and output files are closed.
ERROR.TMP is not deleted.
User Action:
UFD.

Run PSE with output file sent

unprivileged

PSE -·- RECORD SIZE WAS INCORRECT
l~xplanation:

When PSE reads a record from the
input file,
it
verifies the size of the record.
If the size is wrong, it
ignores the record and continues to process the rest of the input
file.

PSE -- TOO MANY DEVICE DESCRIPTOR ENTRIES.
A LARGER PARTITION
DEVIC.E = device
IDENTIFIER = id

RECOMMEND THAT PSE RUN

Explanation:
PSE has run out of storage space in which to hold
descriptor entries for
the device named in the message.
PSE
continues to process the input file, but it is unable to process
any more errors for the named device.
User Action:
In the future, run PSE for this configuration in a
larger partition,
or specify the /INC=xxx switch when PSE is
installed in a system controlled partition.
PSE requires 10
(decimal)
words.of space for each device for which errors are to
be logged.
PSE -- UNABLE TO CLOSE THE INPUT FILE
Explanation:
PSE is unable to close the
The file is not deleted.

input

file

ERROR.TMP.

PSE -- UNABLE TO CLOSE THE OUTPUT FILE
Explanation:
PSE is unable to close the output file.
The
message generally indicates that the output device is full and
that the last block of the file will be missing.
User Action:

Make more space available on your output device.

PSE -- NO ERROR.TMP FILE PROCESSED
Explanation:

PSE cannot locate any files named ERROR.TMP.

User Action: Verify that the correct input device was
or that ERRLOG has been initialized.

specified

PSE -- UNKNOWN DEVICE OR CONTROLLER IDENTIFIER
IDENTIFIER = id
Explanation:
PSE has received a device descriptor error entry
that does not match any device described in the device tables.
PSE ignores that entry record and continues to process the rest
of the file.

5-5

ERROR LOGGING TASK MESSAGES
PSE -- UNABLE TO RECOGNIZE ERROR ENTRY CODE
ENTRY CODE = code
Explanation: PSE must recognize the entry code contained within
each record in order to know how to process the record.
If PSE
does not recognize the entry code,
it ignores the record and
continues to process the rest of the Input file.
PSE -- UNABLE TO REQUEST A FILE FROM ERRLOG
Explanation:
PSE tried to request ERRLOG to make a file
available for formatting.
This error indicates that ERRLOG has
not been initialized.
User Action:

5.3

5.3.1

Run ERRLOG and then run PSE again.

SYE MESSAGES

Command Line Errors

1~he

messages described in this section occur when
error in the command line.

SYE

encounters

than

SYE -- COMMAND STRING ERROR
ERROR NUMBER x
Explanation:
SYE has encountered an error
file) when trying to obtain a command line.

(other

end

User Action:
Refer to the IAS/RSX-11 I/O Operations Reference
Manual, Appendix I, for an explanation of the error code "x", a
decimal number.
SYE
text

COMMAND STRING ERROR
Explanation:
SYE encountered an error when it tried to parse the
command line.
"text" is the command line that contains the
error.
User Action:

SYE
text

Check syntax and re-enter command.

COMMAND STRING SYNTAX ERROR
Explanation:
SYE encountered an error in the syntax of the
command line;
"text" is the erring portion of the command line.
User Action:

Reenter the command, using the correct syntax.

SYE -- ILLEGAL REPORT SWITCH
/REPORT: text
Explanation:
Illegal values ("text") have been supplied with the
report (/RP) switch.
User Action:

See Section 3.3.

5-6

ERROR LOGGING TASK MESSAGES
SYE -- ILLEGAL SWITCH COMBINATION
NO OUTPUT CAN BE GENERATED
Explanation:
combination.

The SYE command line specified

User Action:

Reenter the command line.

illegal

switch

SYE -- NOT CATEGORIZED
Explanation:
device, but
categories.

There was a hard or soft error on the specified
the error did not fit in any of the 16 defined

SYE -- NO ARGUMENT FOR DEVICE SWITCH
Explanation: The SYE command line did not specify a group of
devices or a specific device in the device switch (that is, /DV).
User Action:

5.3.2

Reenter the command line.

File Service Errors

The following messages are caused by file failures.
Each message
contains the line "FATAL ERROR - x" where x is a decimal number that
represents an error. code returned by File Control Services (FCS). See
the IAS/RSX-11 I/O Operations Reference Manual, Appendix I.
SYE -- DEVICE ERROR INPUT FILE
FATAL ERROR - x.
filename
Explanation: An error (other than end of file) occurred when SYE
attempted to read data from the input file. SYE closes all files
and prompts for the next command line.
SYE -- DEVICE ERROR INPUT FILE
NO SUCH FILE
filename
Explanation: The input file specified in the SYE command line
does not exist.
SYE closes all files and prompts for the next
command line.
SYE -- DEVICE ERROR OUTPUT FILE
FATAL ERROR - x.
filename
Explanation: SYE encountered an error when attempting to write
data to the output file. SYE closes all files and prompts for
the next command line.
SYE -- OPEN FAILURE ON INPUT DEVICE
FATAL ERROR - x.
filename
Explanation: SYE encountered an error when it tried to open the
input file.
SYE closes all files and prompts for the next
command line.

5-7

ERROR LOGGING TASK MESSAGES
SYE -- OPEN FAILURE ON OUTPUT DEVICE
FATAL ERROR - x.
filename
Explanation:
SYE encountered an error when it tried to open the
output file.
SYE closes all files and prompts for the next
command line.
SYE -- SUMMARY TEMP FILE I/O ERROR REPORT CONTINUES WITHOUT SUMMARIES
Explanation:
SYE encountered an error while reading or writing
to file SYESUM.TMP. This file is a temporary file created when
SYE runs out of available work space in memory.

5.3.3

Additional Messages

SYE -- CONFIGURATION BUFFER FILLED
Explanation: Due to a large number of configuration records in
the input file,
a buffer in SYE is filled to capacity and the
configuration information reported may be incomplete.
SYE
TEMPORARY FILE ERROR
REPORT CONTINUES WITHOUT SUMMARIES
Explanation:
SYE was unable to produce a complete summary report
because it encountered an error in a temporary file containing
data for the device summary reports.
SYE produces a
report
without the device summaries.
User Action:
SYE -- x.

None required

PAGES IN REPORT

Explanation:
SYE displays this message when it has completed its
formatting of the complete report.
"x." is the decimal number of
pages in the report.
User Action:

5.4

None required.

ERF MESSAGES

The shutdown task ERF always displays a message in response to a
request to run;
the message is either an error message or a
confirmation of shutdown.
ERF -- "ERRLOG" NOT INSTALLED
Explanation:

The task ERRLOG has not been installed.

User Action:

Install ERL and run again.

ERF -- ERROR LOGGING NOT STARTED
Explanation:
Error logging was not
does not need to be shutdown.

initialized,

and

therefore

5-8

................................................................................ ..........................._.1•m
~

ERROR LOGGING TASK MESSAGES
ERF -- REQUESTED "ERRLOG" TO STOP LOGGING
~·

Explanation: The normal confirmation message.
ERF has exited
and ERRLOG should display its own termination message shortly.
User Action:

None.

This is the ERF success message.

ERF -- TERMINAL NOT PRIVILEGED
Explanation:
User Action:

ERF must be requested from a privileged terminal.

Run ERF from privileged terminal.

'-''

5-9

APPENDIX A
MODIFYING ERL AND SYE AT TASK BUILD

A.l

ERL

The command file ERLBLD.CMD contains instructions to the task builder
to create the ERRLOG task.
There is a global definition which may be
changed which affects the way the PSE runs.

This definition (FILPRO) is the file protection mask for ERL-generated
files [l,6]ERR.TMP •~nd [l,6]ERROR.TMP. These are files that are later
read and deleted by PSE.
Thus if the file protection mask extends read and delete privileges to
the world, PSE can be run from a nonprivileged terminal.
FILPR0=60000 extends all privileges to system,
read and delete privileg~s to the world.

·~

owner

and

group

and

Change the value of FILPRO to 170000 to require PSE to run from a
privileged terminal.
See the IAS/RSX-11 I/O Operations Reference
Manua,!, Chapter 4, the section describing default file protection
routines for a complete description of the file protection mask.

A. 2

:SYE

The command file SYEBLD.CMD contains instructions to the task builder
to create the ••• SYE task.
There are several option statements in the
command file that may be changed before task build to modify the
standard analyzer task SYE:
1.

The page width and buffer size for output from the terminal
message processor to the terminals.
These values are defined
by the following TKB option statements:
GBLDEl~=PT$WTH: nnn

{page width)

and
(buffer size)

EXTSCT=TBUF:nnn

where nnn is octal and must be the same in
The default value is 110 (octal).

both

directives.

The value assigned to nnn does not affect task size.

A-1

MODIFYING ERL AND SYE AT TASK BUILD
2.

The page width and buffer size for the output from the error
processors.
These values are defined by the following TKB
option statements:
GBLDEF=PL$WTH:nnn

(page width)

and
EXTSCT=OBUF:nnn

(buffer size)

where nnn is octal and must be the same in
The default value is 120 (octal).

both

directives.

Note that the value of nnn affects the size of the task.
Figure A-1 lists the contents of the command file SYEBLD.CMD.

[1,24]
S Y E B L D • C MD
;USE: RSXllM SYE ERROR LOG DISPLAY PROGRAM
;THIS COMMAND FILE WILL BUILD THE ••• SYE TASK FROM THE
;SYEBLD.ODL FILE FOR A MAPPED SYSTEM
;

[l,54]SYE/MM/CP/-FP,MP: [l,34]SYE/CR/MA/-SP=IN:SYEBLD/MP
;

STACK=96
ASG=TI:l:5
TASK= ••• SYE
PAR=GEN:0:40000
;

;THIS DEFINES THE PAGE WIDTH FOR OUTPUT TO THE TERMINALS
;FROM THE TERMINAL MESSAGE PROCESSOR - THIS HAS NO EFFECT ON THE
;OUTPUT OF THE ERROR DISPLAYS
;ALL VALUES ARE IN OCTAL
;

GBLDEF=PT$WTH:ll0

Figure A-1

SYEBLD.CMD

A-2

MODIFYING ERL AND SYE AT TASK BUILD
;THIS IS THE SIZE OF THE TERMINAL BUFFER
;THIS VALUE MUST BE EQUAL TO THE VALUE USED FOR PT$WTH
;

EXTSCT=TBUF:llO
;

;THIS DEFINES THE PAGE WIDTH FOR THE OUTPUT FROM THE ERROR
;PROCESSORS
;

;ALL VALUES ARE OCTAL
GBLDE:F=PL$WTH:l20
;

;THIS IS THE EXPANSION OF THE OUTPUT BUFFER
;THIS VALUE MUST BE THE SAME AS PL$WTH
;

EXTSCT=OBUF:l20
;

;THIS IS THE DEFINITION OF THE PAGE LENGTH FOR THE OUTPUT FROM
;THE ERROR PROCESSORS
;

;ALL VALUES TYPED ARE OCTAL
;

GBLDE:F=PL$ LGH: 7 4
;
~·

/
Figure A-1 (Cont.)

SYEBLD.CMD

Figure A-2 lists the contents of ERLBLD.CMD.

BOB LYONS APR l, 76
BLOOl

ADDED "/" AT END OF OPTIONS

RA027

ADD GBLDEF FILE PROTECTION MASK
FOR ERR.TMP&ERROR.TMP

;·

[l,54]ERL/PR/MM,MP:[l,34]ERL/CR/MA/-SP=IN:ERLBLD/MP
UNITS=4
ASG=U: 2
ASG=SY:4
GBLDEF=FILPR0:6000
STACK=64
TASK==ERRRLOG
PAR=GEN:0:40000
UIC=l[l,6]
PRI=l45
I
Figure A-2 ERLBLD.CMD

A-3

APPENDIX B
ERR.TMP AND ERROR.TMP FILE RECORD FORMATS

This Appendix illustrates record formats that appear in a temporary
file (either ERR.TMP or ERROR.TMP) created by the error logging task
ERRLOG. Printouts generated by the RSX-llM/M-PLUS DMP utility program
of ~~ither temporary file contain the record formats described in the
following tables. Specifically, the tables describe:
•

Initial configuration records (Table B-1),

•

Device initialization records (Table B-2),

•

Device hardware and interrupt timeout errors (Table B-3),

•

Unexpected trap or interrupt errors (Table B-20),

•

Device driver load and unload errors (Table B-21),

•

Memory parity errors (Table B-22), and

•

Power fail records (Table B-23).

Tables B-4 through B-18 explain the device-dependent parameters
in T4:ible B-3.
Users should refer to these tables when implementing
features into the error logging system.

new

In printouts generated by the DMP program, the fecord
precedes each variable length record description.

size

shown

devices
in

bytes

Reserved Words - The term "reserved" means that the corresponding word
contains a value for a specific purpose; its contents must not be
altered.

Blank Words - The term "blank" means that the corresponding
allocated for future use; its contents have no meaning.

word

was

Time Indications - Time indications are recorded as bytes or words.
In general, byte entries are made by Executive routines; word entries
are made by the get time system directive.

B-1

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-1
ERR.TMP and ERROR.TMP File Record Format
Initial Configuration Record
Offset
Value

Word
0
1

Size of record in bytes
(Reserved)
Error code (EC.IN!)*
Year
Month
Day
Hour
Minute
Second
RSX-llM version number
Error logger Identification
Error logger Identification
Initial start (0·) 'PSE' request ( 2)
Configuration code
(Blank)
(Blank)
(Blank)
(Blank)

2
3
4
5
6
7
8
9
10
11
.12

13
14
15
16

Definition

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24
+26
+30
+32
+34
+36
+40

E.SIZE
E.CODE
E.TIME

E.WHY
E.NAME

*EC.INI=40
Configuration Codes:
0 - Unmapped
2 -

18 bit addressing

4 -

22 bit addressing
Table B-2
ERR.TMP and ERROR.TMP File Record Format
Device Initialization Record
Off set
Value

Size of record in bytes
Init. SUB-CODE* Error code (EC.IN!)
UCB address
Device name in ASCII
Physical unit number System unit number
Vector/4
System controller number
Device CSR Address
IOABM mask
Device control words (CWl)
Device control words (CW2)
Device control words (CW3)
Device control words (CW4)
(Blank)
(Blank)
(Blank)

0
1
2

3
4

5
6
7

8
.9
10
11
12
13
14

1
2

+O
+2
+4
+6
+10
+12
+14
+16
+20

Definition
E.SIZE
E.CODE

+22
+24
+26
+30

+32
+34

Device is loaded for error logging (E.DDW).
Device is unloaded for error logging (E.RDDW).

B-2

I I $¥..llh•lli

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-3
ERR.TMP and ERROR.TMP File Record Format
Device Hardware and Interrupt Timeout Errors
~

Off set
Value

Word
Size of :record in bytes
(Reserved)
Entry Code (FC.DVC*)
Second
Minute
Hour
Day
Month
Year
Error sequence number
Active bit map
Maximum retr ieis
Retries left
Task priority
I/O in Queue
Task name
Task name
Base address of task
Programmer group
Programmer code
UCB address
Function code
I/O Packet Parameters
(device-dependent)*
(device-dependent)
(device-dependent)
(device-dependent)
(device-dependent)
(device-dependent)
Count of registers
Device registers

0
l

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

'-'i
n

Device registers

+2
+4

Definition
E.SIZE
E.CODE
E.TIME

+6
+10

+12
+14
+16

+20
+22

E.SEQ
E.ABM
E.RTRY
E.IOC(21)
E.TASK

+24
+26

+30
+32

+34

E.PAR
E.UIC
E.UCB
E.FNC

+36

+40
+42

+44
+46

+50
+52

+54
+56

E.RCNT
E .REGS

*FC.DVC=l

See I/O Packet Parameters in Tables B-4 through B-16.
Table B-4
I/O Packet Parameters
RK03, RK05, RK05F
Off set
Value
15
16
17
18
19
20
21

Transfer memory address bits
Transfer memory address bits
Transfer size in bytes
(Reserved)
11
RKDA" for I/O GO
LBN or transfer start
(Reserved)

(High)
(Low)

+36
+40

+42
+44

+46
+50

+52

B-3

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-5
I/O Packet Parameters
RP02, RP03
Offset
Value

Word
.15

16
17
18
19

20
21

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RPCA" for I/O GO
"RPDA" for I/O GO
(Reserved)

+36
+40
+42
+44
+46
+SO
+52

Table B-6
I/O Packet Parameters
RFll
Offset
Value

Word
15
16
17
18

19
20
21

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RFDAE" for I/O GO
"RFDAR" for I/O GO
(Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-7
I/O Packet Parameters
RS03, RS04

Off set
Value

Word
15
16
17
18
19
20
21

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
(Reserved)
"RSDA" for I/O GO
(Reserved)

B-4

+36
+40
+42
+44
+46
+50
+52

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-8
I/O Packet Parameters
RP04, RP05, RP06
Offset
Value

Word
15
16
17
18
19
20
21

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RPDC" for I/O GO
"RPDA" for I/O GO
(Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-9
I/O Packet Parameters
RK06
Offset
Value

Word
15
16
17
18
19
20
21

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RKDA" for I/O GO
"RKDC" for I/O GO
(Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-10
I/O Packet Parameters
TAll
Off set
Value

Word
15
16
17
18
19
20
21

Space count if spacing function or
"KTAPR" contents to transfer data through if
data function
Virtual buffer address of transfer
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

B-5

+36
+40
+42
+44
+46
+SO
+52

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-11
I/O Packet Parameters
TS03, TElO, TUlO
Off set
Value

Word
15
16
17
18

19
20
21

Space count if space function or transfer memory
address (High) if data function
Transfer memory address (Low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-12
I/O Packet Parameters
TS04
Offset
Value
15
16
17
18

19
20
21

Transfer memory address bits (high)
Transfer memory address bits (low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table B-13
I/O Packet Parameters
TE16, TU16, TU45
Off set
Value

Word
15
16
17
18

19
20
21

Space count if space function or transfer memory
Address (High) if data function
Transfer memory address (Low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

+36
+40
+42
+44
+46
+50
+52

B-6

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-14
I/O Packet Parameters
TU58
Off set
Value

Word
15
16
17
18

19
20
21

Relocation Bias of Data Buffer
Buffer address of I/O Transfer
Number of bytes to transfer
Not used
Low byte must be O. High byte not used
Logical or physical block number
Not used

+36
+40
+42
+44
+46
+50
+52

Table B-15
I/O Packet Parameters
TCll
Off set
Value

Word·
15
16
17
18

19
20
21

Transfer memory address (High)
Transfer memory address (Low)
Transfer size in bytes
(Reserved)
(Reserved)
LBN for transfer start
(Reserved)

+36

+40
+42
+44
+46

+50
+52

Table B-16
I/O Packet Parameters
RLOl, RL02
Off set
Value

Word
7
8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Total transfer size in byies
Seek difference count
"RLDA" for I/O GO
This transfer size in bytes
(Reserved)

•

B-7

+36

+40
+42
+44
+46
+50
+52

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-17
I/O Packet Parameters
RM03, RP07
Offset
Value

Word
7
8
9

10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RMDC" f~r I/O GO
"RMDA" for I/O GO
{Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-18
I/O Packet Parameters
RXOl
Offset
Value

Word
7
8
9

10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
{Reserved)
Current transfer size in bytes
"RXDA" for I/O GO
(Reserved)

+36
+40
+42
+44
+46
+50
+52

Table B-19
I/O Packet Parameters
RX02
Off set
Value

Word
7
8
9

10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
Status register after interrupt
Logical or physical block number
Current transfer size in bytes
Track number and sector number

B-8

+36
+40
+42
+44
+46
+50
+52

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-20
ERR.TMP and ERROR.TMP File Record Format
Unexpected Trap or Interrupt Error
Off set
Value

Word
0
1
2
3
4
5
6
7

Size of record in bytes
(Reserved)
Error Code (EC.NSI)
Minute
Second
Day
Hour
Year
Month
Error Sequence Number
Active bit map
Missed count
Vector/4 of Trap

+O
+2
+4
+6
+10
+12
+14
+16

8
9

Processor Status before trap
Program Counter before trap

+20
+22

Definition
E.SIZE
E.CODE
E.TIME
E.SEQ
E.ABM
E.VCTR(l6)
E.LOST(l7)
E.OPC

Table B-21
ERR.TMP and ERROR.TMP File Record Format
Device Driver Load or Unload
Off set
Value

Word
0
1
2
3
4
5
6

7
8
9
10
11
12
13
14
15
16

Size of record in bytes
(200)
Error code (EC.IN!)
Year
Month
Day
Hour
Minute
Second
RSX-llM version number
Error logger Identification
Error logger Identification
Load(4)/Unload (10) code
Configuration code
Driver name
(Blank)
(Blank)
(Blank)

B-9

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24
+26
+30
+32
+34
+36
+40

Definition
E.SIZE
E.CODE
E.TIME

E.WHY
E.NAME

ERR.TMP AND ERROR.TMP FILE RECORD FORMATS
Table B-22
ERR.TMP and ERROR.TMP File Record Format
Memory Parity Errors
Off set
Value
0
1
2

3
4

6
7

8
9
10
11
12
13
14
15
16
17
18

Size
Processor Type
Entry Code (EC.MPG*)
MINUTE
SECOND
DAY
HOUR
YEAR
MONTH
ERROR SEQUENCE NUMBER
TRAP PC
TRAP PS
FIRST WORD TASK NAME
SECOND WORD TASK NAME
FIRST WORD PARTITION NAME
SECOND WORD PARTITION NAME
PARTITION BASE ADDRESS
PARTITION SIZE
MEMORY PARITY CSR's

19
20
21
22
23
24
25
26
27
28

29
30
31
32
33
34
35

Cache parity CSR's

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24
+26
+30
+32
+34
+36
+40
+42
+44
+46
+50
+52
+54
+56
+60
+62
+64
+66
+70
+72
+74

+100

+102
+104

II
II

+106

+110
*EC.MPG=2

B-10

•'-

ERR •• TMP AND ERROR.TMP FILE RECORD FORMATS

Table B-23
ERRoTMP and ERROR.TMP File Record Format
Power Fail Record
Off set
Value

Word
0
l
2

3
4

5
6

Size of record in bytes
Reserved Error Code (EC.PWR)
Year
Month
Day
Hour
Minute
Second
EC.PWR=42

B-11

+O
+2
+4
+6
+10
+12
+14
+16

Definition
E.SIZE
E.CODE
E.TIME

APPENDIX C

ERROR.SYS FILE RECORD FORMATS

This Appendix defines record formats contained in the file ERROR.SYS.
If the report generating task SYE is not available, the user can
consult these record formats to analyze the contents of ERROR.SYS.
These formats should also be studied when adding features to the
existing error logg~ng system.

"'-"'

A user can run the system utility DMP to obtain a
listing of
ERROR.SYS.
(See the RSX-11 Utilities Manual.) In such a printout, the
record size in bytes precedes each variable length record.
Each record consists of one or more fields.
The first field in every
record is a header field that contains basic error information and
optionally points to additional fields.
The optional fields can
follow in any order.
Tables D-1 through D-6 define the six header
field formats available for different types of errors;
these errors
are:
•

Hardware or interrupt timeout errors (Table C-1),

•

Memory parity errors (Table C-2),

•

Error logging startup errors (Table C-3),

•

Error logging termination errors

•

Power fail records (Table C-5)

•

Errors caused by PSE's removal of
C-6), and

•

Unexpected trap or interrupt errors (Table C-7).

(Table C-4),

the

ERR.TMP

file

(Table

The first 8 words of any header field always contain the same type
information.

Tables C-8, C-25, C-26, and C-27 define four fields that may or may
not be part of the record, depending on the type of error described.
(Tables C-9 through C-20
describe
device-dependent
parameters
referenced in Table C-8.) The four additional fields are:
•

A program field, format O,

•

A program field, format 2 (Table C-25),

•

A bus activity field

•

A device register field

(Table C-8),

(Table C-26), and
(Table C-27).

C-1

ERROR.SYS FILE RECORD FORMATS
Reserved Words - The term "reserved" means that the correspon~ing word
contains a value for a specific purpose;
its contents must not be
altered.
Blank Words - The term "blank" means that the related
allocated for future use;
its contents have no meaning.

word

was

Table C-1
ERROR.SYS File Record Format
Header Field Hardware (001) or Interrupt Timeout (140) Errors
Format 0
Off set
Value

Word
0
1

2
3

4
5
6
7
8

9
10
11
12
13
14
15
16
17
18

Size of header field in bytes
Entry type (001/140) format (0)
Year
Month
Day
Hour
Minute
Second
Error sequence number
UDE (device type and class)
UDE (controller and unit number)
Device mnemonic (ASCII)
Unit logical number
Offset to register field
Offset to program field
(Reserved)
(Reserved)
Offset to bus activity field
(Reserved)

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24
+26
+30
+32
+34
+36
+40
+42
+44

Definition
A$SIZE
A$FMT(2),A$ENTY(3)
A$TIME

A$SEQN
A$DCDT(22),A$DTYP(23)
A$CNTN(24),A$UNTN(25)
A$DVNM
A$SYUN
A$DRF
A$PGF
A$BAF

Table C-2
ERROR.SYS File Record Format
Header Field Memory Parity Error (002) Format 1
Offset
Value

Word
0
1
2
3

4
5
6
7

8
9

10
11
12

Size of header Field in Bytes
Entry type (002)
Format (1)
Year
Month
Day
Hour
Minute
Second
Error sequence number
UDE (device class and type)
UDE (controller and unit number)
Offset to program field
Offset to register field

C-2

+o
+2
+4
+6

Definition
A$SIZE
A$FMT(2),A$ENTY(3)
A$TIME

+10
+12

+14
+16

+20
+22
+24
+26
+30

A$SEQN
A$DCDT(22),A$DTYP(23)
A$CNTN(24),A$UNTN(25)
A$MPF1
A$MRF1

ERROR.SYS FILE RECORD FORMATS
Table C-3
ERROR.SYS File Record Format
Header Field System Start Up (040) Format O
Off set
Value

Word
0
1
2

3
4
5

6
7
8
9
10
11

Size of header field in bytes
Entry type (040)
Format (0)
Year
Month
Day
Hour
Minute
Second
Error sequence number
(Reserved)
(Reserved)
(Reserved)

+O
+2
+4
+6

Definition
A$SIZE
A$FMT(2),A$ENTY(3)
A$TIME

+10

+12
+14
+16
+20
+22
+24
+26

A$SEQN

Table C-4
ERROR.SYS File Record Format
Header Field Error Logging Terminated (041) Format O
Off set
Value

Word
0
1
2
3
4
5
6
7

Size of header field in bytes
Entry type (041)
Format (0)
Year
Month
Day
Hour
Minute
Second

+O
+2
+4
+6
+10
+12
+14
+16

Definition
A$SIZE
A$FMT(2),A$ENTY(3)
A$TIME

Table C-5
ERROR.SYS File Record Format
Power Fail (042) Format 0
Offset
Value
0

1
2
3
4
5
6
7
8

Size of header field in bytes
Entry type (42) Format (0)
Year
Month
Day
Hour
Minute
Second
Error sequence number

C-3

+O
+2
+4
+6

Definition
A$SIZE
A$FMT(2) ,A$ENTY(3)
A$TIME

+10

+12
+14
+16
+18

A$SEQN

ERROR.SYS FILE RECORD FORMATS
Table C-6
ERROR.SYS File Record Format
Header Field Error File (ERR.TMP) Re-initialized (043) Format O
Off set
Value

Word
0
l
2
3
4

5
6
7

Size of header field in bytes
Entry type (043)
Format (0)
Year
Month
Day
Hour
Minute
Second
Error sequence number

+O
+2
+4
+6
+10
+12
+14
+16
+20

Definition
A$SIZE
A$FMT(2),A$ENTY(3)
A$TIME

A$SEQN

Table C-7
ERROR.SYS File Record Format
Header Field Unexpected Trap or Interrupt (141) Format 0
Off set
Value

Word
0
l
2
3
4
5

6
7
8
9
10
11
12
13

Size of header field in bytes
Entry type (141)
Format (0)
Year
Month
Day
Hour
Minute
Second
Error sequence number
Vector of undefined interrupt
Unexpected interrupts not logged
because of this log effort
PSW of interrupted process
PC of interrupted process
Offset to bus activity field

Definition

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24

A$SEQN
A$VCTR
A$UICT

+26
+30
+32

A$0PS
A$0PC
A$UIPT

A$SIZE
A$FMT ( 2)
A$TIME

,A$ENTY(3)

..
~

C-4

ERROR.SYS FILE RECORD FORMATS
Table c-8
ERROR.SYS File Record Format
Program Field J?ormat O {Hardware or Interrupt Timeout Errors)
The Program Field contains information
about the program that caused the error
or that was active when the error occurred.
Off set
Value

Word
Size of program field in bytes
Address mode*
Format (0)
Task name in RADSO
Task name in RADSO
Programmer group
Programmer code
Base program load address**
Function code
I/O packet information***
(Device-dependent)
(Device-dependent)
(Device-dependent)
(Device-dependent)
{Device-dependent)
(Device-dependent)
Maximum retries
Retries left****
I/O sti 11 in Queue

0
1
2
3
4
5
6
7

8
9
10

""'

12
13

14
15

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24
+26
+30
+32
+34
+36

Definition
P$FMT ( 2) , P$ADMD ( 3)
P$TN
P$UIC
P$TAO
P$FC
P$PARM

P$RTO
P$RQ

O = Unmapped; 2 = 18-bit Addressing; 4 = 22-bit Addressing.

Real if unmapped; 1/64 Real if 18-bit or 22-bit Addressing.

***

See I/O Packet Parameters, Tables C-9 through C-21.

***~:

If word is negative, error was not recovered.

Table C-9
I/O Packet Parameters
RK03, RKOS, RK05F
Off set
Value
7
8
9
10

11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
"RKDA" for I/O GO
LBN or transfer start
(Reserved)

C-5

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-10
I/O Packet Parameters
RP02, RP03
Offset
Value

Word
7
8

9
10
11
12
13

Transfer memory address bits {High)
Transfer memory address bits {Low)
Transfer size in bytes
{Reserved)
"RPCA" for I/O GO
"RPDA" for I/O GO
{Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-11
I/O Packet Parameters
RFll
Offset
Value

Word

7
8
9

10
11
12
13

Transfer memory address bits {High)
Transfer memory address bits (Low)
Transfer size in bytes
{Reserved)
"RFDAE" for I/O GO
"RFDAR" for I/O GO
{Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-12
I/O Packet Parameters
RS03, RS04
Offset
Value

Word

7
8
9

10
11

12
13

Transfer memory address bits {High)
Transfer memory address bits {Low)
Transfer size in bytes
(Reserved)
{Reserved)
"RSDA" for I/O GO
{Reserved)

C-6

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-13
I/O Packet Parameters
RP04, RPOS, RP06
Off set
Value

Word
7
8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved}
"RPDC" for I/O GO
"RPDA" for I/O GO
(Reserved}

+16
+20
+22
+24
+26
+30
+32

Table C-14
I/O Packet Parameters
RK06

~
Off set
Value

Word
7
B
9

10
11
~

12
13

Transfer memory address bits (High}
Transfer memory address bits (Low}
Transfer size in bytes
(Reserved)
"RKDA" for I/O GO
"RKDC" for I/O GO
(Reserved}

+16
+20
+22
+24
+26
+30
+32

Table C-15
I/O Packet Parameters
TAll
Off set
Value

Word

7
B

9
10
11
1:2
13

Space count if spacing function or
"KTAPR" contents to transfer data through if
data function
Virtual buffer address of transfer
Transfer size in bytes
(Reserved}
(Reserved}
(Reserved}
(Reserved}

C-7

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-16
I/O Packet Parameters
TS03, TElO, TUlO
Off set
Value

Word
7
8
9
10
11
12
13

Space count if space function or transfer memory
address (High) if data function
Transfer memory address (Low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-17
I/O Packet Parameters
TS04
Offset
Value

Word
15
16
17
18
19
20
21

Transfer memory address bits (high)
Transfer memory address bits (low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-18
I/O Packet Parameters
TE16, TU16, TU45
Off set
Value

7
8
9
10
11
12
13

Space count if space function or transfer memory
address (High) if data function
Transfer memory address (Low)
Transfer size in bytes
(Reserved)
(Reserved)
(Reserved)
(Reserved)

C-8

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-19
I/O Packet Parameters
TU58
Off set
Value

Word
15
16
17
18
19
20
21

Relocation·bias of data buffer
Buffer address of I/O transfer
Number of bytes to transfer
Not used
Low byte must be zero. High byte not used
Logical or physical block number used
Not used

+36
+40
+42
+44
+46
+50
+52

Table C-20
I/O Packet Parameters
TCll
Off set
Value

Word
7
8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
(Reserved)
(Reserved)
LBN for transfer start
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-21
I/O Packet Parameters
RLOl, RL02
Off set
Value

Word
7

8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Total transfer size in bytes
Seek difference count
"RLDA" for I/O GO
This transfer size in bytes
(Reserved)

C-9

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-22
I/O Packet Parameters
RM03, RP07
Off set
Value

Word
7
8
9
10
11
12
13

Transfer memory address bits {High)
Transfer memory address bits (Low)
Transfer size in bytes
{Reserved)
"RMDC" for I/O GO
"RMDA" for I/O GO
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-23
I/O Packet Parameters
RXOl
Off set
Value
7
8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits {Low)
Transfer size in bytes
(Reserved)
Current transfer size in bytes
"RXDA" for I/O GO
(Reserved)

+16
+20
+22
+24
+26
+30
+32

Table C-24
I/O Packet Parameters
RX02
Offset
Value

Word
7
8
9
10
11
12
13

Transfer memory address bits (High)
Transfer memory address bits (Low)
Transfer size in bytes
Status register often interrupt
Logical or physical block number
Current transfer size in bytes
Track number and sector number

C-10

+16
+20
+22
+24
+26
+30
+32

ERROR.SYS FILE RECORD FORMATS
Table C-25
ERROR.SYS File Record Format
Program Field Format 2 (Memory Parity Errors)
The Program Field contains information
about the program that caused the error or
that was active when the error occurred.
Off set
Value
0
1
2

3
4
5
6
7
8
9

+O
+2
+4
+6
+10
+12
+14
+16
+20
+22
+24

Size of program field in bytes
Address mode*
Format (2)
Task name in RADSO
Task name in RADSO
Base address of load **
Task maximum size **
Partition name in RADSO
Partition name in RADSO
Base address of partition
Processor Status word
Program Counter

Definition
P$FMT(2), P$ADMD(3)
P$TN
P$TA2
P$TS
P$PN
P$PS
P$PSW
P$PC

* 0 = Unmapped; 2 = 18-bit addressing; 4 = 22-bit addressing.
** Real if unmapped; l/64th real if 18-bit or 22-bit addressing.
Table C-26
ERROR.SYS File Record Format
Bus Activity Field Format O
The bus activity field contains information
on bus activity at the time of the error.
Offset
Value

Word
0
1
2

Size of bus activity field in bytes
(Res,erved)
Format (0)
Active vector/4
Active vector/4

Active vector/4

"
Active vector/4

+O
+2
+4

C-11

Definition

B$fmt
B$VEC

ERROR.SYS FILE RECORD FORMATS
Table C-27
ERROR.SYS File Record Format
Device Register Field Format O
The device register field contains the contents
of the control and status registers of the
device on which the error occurred.
Offset
Value

Word
0
1
2

Size of register field in bytes
(Reserved)
Format (0)
Device registers
II

Device registers

C-12

+O
+2

_ Jj. -

Definition

R$FMT
R$REG

APPENDIX D
DEVICE DRIVER MODIFICATION

This Appendix describes the modifications made to the mass storage
device drivers to enable error logging, as well as the common
Executive routines that are supplied for the error logging subsystem.
All error logging code is incorporated in the Executive or device
driver at system generation. In addition to building the necessary
device data structures, the system generation procedures enable all
error logging code by defining the following conditional assembly
symbols:

D.l

Symbo];

Errors

E$$DVC

Hardware
timeouts)

E$$NSI

Undefined interrupts

E$$PER

Memory parity errors

errors

(including

interrupt

DEVICE DRIVER MODIFICATIONS

A device driver that is modified to enable error logging performs
following:

the

•

Immediately before starting a device data type function,
it
calls the Executive routine "$BMSET" to set the appropriate
bit, thereby indicating that the device has a data function
active on the UNIBUS.

411

It defines the entry point for
system macro INTSE$.

It cal ls the Executive routine $DVCER if it discovers a
loggable
error
while executing the interrupt handling
routine. The routine $DVCER then logs the error.

interrupt

handling

via

the

It calls the Executive routine $DTOER if it discovers a
device timeout error.
The routine $DTOER then logs the
error.
•

For each error, it performs the required number of retries,
and records in R2 the maximum number of possible retries
(high byte) and the number of possible retries not taken (low
byte)e It then calls the Executive routine $!ODON.

Before attempting to modify a device driver, a programmer should study
a driver that has already been adapted for error logging, such as the
RK05 (DKDRV) device driver.

D-1

DEVICE DRIVER MODIFICATION
D.2

EXECUTIVE ROUTINES

D.2.1

$BMSET

This routine is found in the ERROR module.
Calling sequence:
CALL $BMSET
Description:
;+
**-$BMSET-SET A DRIVER'S BIT IN THE I/O ACTIVE BITMAP
THIS COROUTINE RAISES PROCESSOR PRIORITY TO SEVEN AND
SETS THE MASK IN THE SCB IN $IOABM,
LETS THE CALLER START THE FUNCTION, AND THEN,
ALLOWS INTERRUPTS.
INPUTS:
R4=ADDRESS OF THE SCB
OUTPUTS:
$IOABM IS MODIFIED
;-

D.2.2

$DVCER

THIS ROUTINE IS FOUND IN THE ERROR module.
Calling sequence:
CALL $DVCER
Description:
;+
**-$DVCER-DEVICE ERROR BIT SET
THIS IS THE EMB FORMATTING ROUTINE WHEN DEVICE ERROR BIT
ERRORS ARE RECOGNIZED BY THE DRIVER. ON THE
FIRST OCCURRENCE OF AN ERROR, $DVCER ATTEMPTS TO LOG IT.
IF ERRORS OCCUR ON RETRIES, THEY ARE NOT LOGGED.
THE ERROR CODE "EC.DVC" IS PUSHED ON THE STACK,
THE ERROR IN PROGRESS BIT IS SET IN THE SCB, THE LENGTH OF THE
REQUIRED EMB IS CALCULATED AND "$ALEMB" IS CALLED. IF "$ALEMB" FAILS
TO ALLOCATE A PACKET FOR ANY REASON, "$DVCER" EXITS
AND THE POINTER IN THE SCB TO THE EMB IS CLEARED.
ELSE, THE SAVED $IOABM IS COPIED FROM THE SCB TO THE EMB AND
A POINTER TO THE EMB IS SAVED IN THE SCB. THE ERROR INFORMATION
INCLUDING DEVICE REGISTERS IS PUT INTO THE EMB AND THE
RETURN IS MADE.
INPUTS:
·~

D-2

DEVICE DRIVER MODIFICATION
R4=ADDRESS OF THE SCB
OUTPUTS:
IF SUCCESSFUL, THE EMB IS FILLED, AND THE SCB
CONTAINS A POINTER TO IT. AN ERROR IN PROGRESS BIT
IS SET IN THE SCB. ELSE, THE OCCURRENCE OF
THE ERROR IS COUNTED ONLY.
NOTE: ALL REGISTERS ARE PRESERVED
;-

D.2.3

$DTOER

THIS ROUTINE IS IN THE ERROR module.
Calling sequence:
CALL $DTOER

Descri.ption:
;+
**-$DTOER-DEVICE TIMEOUTS
THIS IS THE EMB FORMATTING ROUTINE WHEN
TIMEOUT ERRORS ARE RECOGNIZED BY THE DRIVER. ON THE
FIRST OCCURRENCE OF AN ERROR, "$DTOER" ATTEMPTS TO LOG IT.
IF ERRORS OCCUR ON RETRIES, THEY ARE NOT LOGGED.
THE ERROR CODE "EC.OTO" IS PUSHED ON THE STACK,
THE ERROR IN PROGRESS BIT IS SET IN THE SCB, THE LENGTH OF THE
REQUIRED EMB IS CALCUALTED AND "$ALEMB" IS CALLED. IF "$ALEMB" FAILS
TO ALLOCATE A PACKET FOR ANY REASON, "$DTOER" EXITS
AND THE POINTER IN THE SCB TO THE EMB IS CLEARED.
ELSE, THE SAVED "$IOABM" IS COPIED FROM THE SCB TO THE EMB AND
A POINTER TO THE EMB IS SAVED IN THE SCB. THE ERROR INFORMATION
INCLUDING DEVICE REGISTERS IS PUT INTO THE EMB AND THE
RETURN IS MADE. THE CONTENTS OF THE CSR THAT IS SAVED IS UNCHANGED
FROM THE TIME OF TIMEOUT. AFTER THE CSR IS SAVED, DEVICE INTERRUPTS
ARE DISABLED AND CPU PRIORITY IS LOWERED TO PRO.
INPUTS:
R2=ADDRESS OF THE CSR
R4=ADDRESS OF THE SCB
OUTPUTS:
C=O IF FUNCTION WAS NOT A USER-MODE DIAGNOSTIC FUNCTION
THE EMB IS FILLED AND THE SCB CONTAINS A POINTER
TO IT AND ERROR IN PROGRESS FLAG IS SET IN THE SCB.
C=l IF FUNCTION WAS A USER-MODE DIAGNOSTIC FUNCTION.
IN THIS CASE ONLY INTERRUPT ENABLE IS CLEARED AND
THE PRIORITY IS LOWERED TO O.

•

NOTE: ALL REGISTERS ARE PRESERVED
;-

D-3

DEVICE DRIVER MODIFICATION
D.2.4

$IODON

Refer to the RSX-llM Guide to Writing an I/O Driver.

D.2.5

INTSE$

$INTSE is the interrupt save routine for device drivers that support
error logging. To generate the required interrupt service code in the
driver, the programmer issues a call to the system macro INTSE$.
INTSE$ also generates the appropriate global symbols for interrupt
entry points.
The format of the INTSE$ macro is:
INTSE$

xx,pri,#ctlrs [,pssave,ucbsave]

where:
xx

is the 2-character device mnemonic.

pri

is the priority of the device (the priority that
be used in a call to $INTSE).

#ctlrs

is the number of controllers the driver services.

pssave

is an optional argument specifying a variable in which
to save the PS word. If omitted, a variable namea TEMP
is used.

ucbsave

is an optional argument specifying a vector in which to
store
the interrupting device's UCB address.
If
omitted, a vector named CNTBL is used.

Outputs:

is the controller index.

is the UCB address.

would

Example:
INTSE$ PP, PR4,P$$Pll

D.2.6

Additional Executive Routines

These routines exist within the Executive and are called by the above
mentioned routines.
This information is supplied to assist in the
handling and understanding of the internal workings of the RSX-llM
error logging sub-system.

D.2.6.1

$ALEMB or $ALEB1 - This routine is found in the ERROR module.

Calling sequence:
CALL $ALEMB
CALL $ALEB1

allocate EMB and increment error sequence number
allocate EMB and do not increment error sequence
number

D-4

ua4

2$0 42 ...

DEVICE DRIVER MODIFICATION
Description:
;+
**-$ALEMB-ALLOCATE AN ERROR MESSAGE BLOCK
**-·$ALEB1-ALLOCATE AN ERROR MESSAGE BLOCK (ALTERNATE ENTRY)

THIS ROUTINE IS CALLED BY ERROR SERVICING
ROUTINES. IT COUNTS THE OCCURRENCE OF THE ERROR AND TRIES
TO ALLOCATE A CORE BLOCK FROM THE POOL.
IF IT IS SUCCESFUL, IT FILLS IN THE ERROR CODE, THE TIME AND THE
ERROR SEQUENCE NUMBER. ELSE, IT RETURNS WITH B-SET.
INPUTS:
2(SP)=ERROR CODE
O(SP)=RETURN
Rl =SIZE OF THE EMB TO ALLOCATE
OU'I'PUTS:
IF B-CLEAR:
RO=ADDRESS OF THE FIRST UNFILLED BYTE
Rl=ADDRESS OF THE EMB
IF B-SET, UNSUCCESSFUL
NOTE: R2 AND R3 ARE DESTROYED BY $ALEMB THRU $ALOCB
;-

D.2.6.2

$QEMB - This routine is found in the ERROR module.

Calling sequence:

'-""'

CALL $QEMB
Description:

;+
**-$QEMB-QUEUE AN EMB
THIS IS THE COMMON POINT FOR ALL EMBS. THE EMB IS QUEUED
FIFO IN THE ERROR QUEUE. IF THERE ARE
ENOUGH BYTES OF EMBS IN THE POOL, THE LOGGER TASK IS AWAKENED.
ELSE, IF THE QUEUE WAS EMPTY, A SCHEDULE REQUEST IS MADE
SO A QUEUED EMB WILL BE WRITTEN WITHIN A TIME LIMIT.
ELSE, A RETURN rs MADE.
INPUTS:
Rl=ADDRESS OF THE EMB
omrPUTS:
THE EMB IS QUEUED. CONDITIONALLY, THE LOGGER
OR A SCHEDULE REQUEST IS MADE FOR THE LOGGER
NOTE: REGISTERS RO-R3 ARE DESTROYED
;-

D-5

rs WAKED

um:

-X-

za_;a.,

INDEX

A
$ALEB1, D-4
$ALEMB, D-4

B
/BEG switch, 3-7
Bitmap,
I/O active, 2-1
$BMSET, D-1, D-2

D
Data formats,
error logging, B-1
Defaults,
PSE I 3-3
SYE I 3-4, 3-5
oete1ction ,
error, 1--1
Device driver modification, 4-1 D-1
Device error reports, 4-2
Device errors, 1-2
OSK type, 3-6
$DTOER, D-3
/DV switch, 3-6
$DVCER, D-2

E
E$$DVC, D-1
E$$NSI, D-1
E$$PER, D-1
/END switch, 3-7
ERF, 3-8
ERF error messages, 5-8
ERR.TMP, 2-5
record formats, C-1
ERRJLOG, 1-3, 2-1
error messages, 5-1
files, 2-4
operating procedures, 3-1
Error detection, 1-1
Error log file, 1-3
Error log reports, 4-1
Error logging,
functions of 1-1
purposes of, 1-1
starting up, 3-1
tasks, 1-2
terminating, 3-8
Error logging data formats, B-1
Error messages, 5~1
ERF, 5-8
ERRLOG, 5-1
PSE, 5-4
SYE', 5-6

Error reports,
device, 4-3
interrupt timeout, 4-5
memory parity, 4-5
undefined interrupt, 4-1, 4-2
Error routines,
Executive, 1-2, 1-3, 2-1, D-2
ERROR.SYS, 2-4
record formats, C-1
ERROR.TMP, 2-5
record formats, B-1
Errors,
Device, 1-2
hard, 4-11
loggable, 1-2
memory parity, 1-2
nonrecovered, 4-11
recovered, 4-11
soft, 4-11
Example,
PSE, 3-3
Examples,
SYE I 3-8
Executive error routines,
1-2, 1-3, 2-1, D-2

F
File,
error log, 1-3
Filenames,
SYE output, 3-4
Files,
ERRLOG, 2-4
SYE report, 2-6
Formats,
ERR.TMP record, B-2
ERROR.SYS record, C-1
ERROR.TMP record, B-2
Formatting routine,
message, 2-3

G
Generating error reports, 1-3, 3-3
4-1

H
Hard errors, 4-11
HOW class, 3-5
/HELP switch, 3-7

I/O active bitmap, 2-1
Individual reports, 4-1

Index-1

INDEX

Procedures,
ERF operating, 3-8
ERRLOG operating, 3-1
PSE operating, 3-2
SYE operating, 3-3
PSE, 1-3, 2-5
error messages, 5-4
example, 3-3
operating procedures, 3-2

Interaction,
task, 2-4
Interrupt reports,
undefined, 4-1, 4-2
Interrupt timeout reports, 4-5
Interrupt timeouts, 1-2
Interrupts,
undefined, 1-2
$INTSE, D-4
INTSE$ D-1, D-4
$IODON, E-1, E-4, D-1, D-4

Q
$QEMB, D-5
Queuing routine,
message, 2-3

L
Log file,
error, 1-3
Log reports,
error, 4-1
Loggable errors, 1-2
Logging,
starting up, 3-1
terminating error, 3-8
Logging tasks,
error, 1-2

M
MAG type, 3-6
MEM type, 3-6
Memory allocation routines, 2-3
Memory parity errors, 1-2
Memory parity error reports,
4-5, 4-6
Message formatting routine, 2-3
Message queuing routine, 2-3
Messages,
ERF error, 5-8
ERRLOG error, 5-1
PSE error, 5-4
SYE error, 5-6
Modification,
device driver, D-1
Modifying SYE, A-1

Record formats,
ERR.TMP, B-1
ERROR.SYS, C-1
ERROR.TMP, B-1
Recovered errors, 4-11
Report files,
SYE, 2-6
Report generation, 1-3
Report switch, 3-5
Reports,
device error, 4-2
error log, 4-1
generating error, 3-3, 4-1
individual, 4-1
interrupt timeout, 4-5
memory parity error, 4-5
summary, 3-8, 4-8, 4-9
undefined interrupt, 4-1, 4-2
Routine,
memory allocation, 2-3
message formatting, 2-3
message queuing, 2-3
Routines,
Executive error, 1-2, 1-3, 2-1,
D-2
/RP switch, 3-5, 3-6

Nonrecovered errors, 4-11

0
Operating procedures, 3-1
ERF, 3-8
ERRLOG, 3-1
PSE, 3-2
SYE, 3-3

p
Parity errors,
memory, 1-2

Soft errors, 4-11
STA type, 3-5
Starting up error logging, 3-1
Summary reports, 3-8, 4-8, 4-9
Switches,
SYE, 3-5
SYE I 1-3 I 2-6
defaults, 3-4, 3-5
error messages, 5-6
examples, 3-8
modifying, A-1
operating procedures, 3-3
output filenames, 3-4

Index-2

INDEX

SYE (Cont.)
report files, 2-6
switches, 3-5
SYEBLD.CMD, A-2
SYS class, 3-5

Terminating error logging, 3-8
Timeout reports,
interrupt, 4-5
Timeouts,
interrupt, 1-2
TMO class, 3-6

·r
Task interaction~ 2-4
Tasks,
error logging, 1-2

u
Undefined interrupt reports, 4-1,
4-2
Undefined interrupts, 1-2

Index-3

RSX-llM/M-PLUS
Error Logging
Reference Manual
AA-H270A-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.

~.
cu

·=
C)

s:
0
0

Did you find errors in this manual?
number.

If so, specify the error and the

pagE~

Please indicate the type of reader that you most nearly represent.

•

[]
[]
[]
[]
[]
[]

Assembly language programmer
Higher-level language programmer
Occasional programmer (experienced)
User with little programming experience
Student programmer
Other (pl.ease specify) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

Name-----------------------------~

Date ______________________

Org1ani za tion.____________________________________________
Street
CitY-----------------------

State

___________ Zip Code _________
or
Country

-1

Do Not Tear- Fold Here and Tape

~nmnomo

""'

No Postage
Necessary
if Mailed in the
United States

111111

l
I

BUSINESS REPLY MAIL
FIRST CLASS PERMIT N0.33 MAYNARD MASS.
POSTAGE WILL BE PAID BY ADDRESSEE

RT/C SOFTWARE PUBLICATIONS

l~~

TW/A14

DIGITAL EQUIPMENT CORPORATION
1925 ANDOVER STREET
TEWKSBURY, MASSACHUSETTS

01876

I
I

r~
-

Do Not Tear - Fold Here

I
I
I
I
Allii,
I ,
I
Q)

1.e
...;i

I~ ~

I~00
I ..s:= .J
<
1=
u
I
I

i1;_. . ._ _ll!llllllll_ _ _ _ _ _ _ _. , ._ _ _ _ _ _ _ _ _llllllll_ _ _ _...___. __ _ _ _ _ _ _ __.......

•
1
•