alMml

digihl equipmentcorpomtion
 Copyright 1969 by
Digital Equipment Corporation
PDP is a registered trademark
of Digital Equipment, Corporation

The material in this handbook is for information pur-

poses only and is subject to change without notice.
 TABLE OF CONTENTS
CHAPIER 1 lNTRODUCTlOR
PDP-11 SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .:.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . 1
UNIBUS . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . , . . . . . . . . .. . . . . . . . . . . . . . . . 1
KAl 1 PROCESSOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .?. 1
Priority Interrupts ,...............,..... . . . . . . . . . . . . . . . . .. . . . . . . .. . . . . . . . . . . . . . . . 1
Reentrant Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Registers . . . .. . . . . .. . . . . . . . . .:... . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 2
Instruction Set . . . . . . . . . .._........................................................
Addressing . ..~...............................................~........,...,......... z
Asynchronous Operation ...... ........... ........ ............ ...... ......... L 2
PACKAGJNG . . . . . . . . . . . . . . ... . .. . . . . . . . . . . . . . . . . . . ..I . . . . . . . . . . . . . . 2
SOmARE ........................ .: ........................................................ 3
CHAPTER 2 SYSTEM INTRODUCTION
SYSTEM DEFINITION .................................................................. 5
SYSTEM COMPONENTS ............................................................... 5
UNIBUS ................................................... . .................................. 5
Single Bus ................................................................. . .......... 5
Bidirectional Lines ............ .......................................... ..* .... 5
Master-Slave Relation ........................... .............................. 5
Interlocked .Communication ................................................
Dynamic Master-Slave Relation ........................................... :
KAll CENTRAL PROCESSOR ...................................................... 6
General Registers ................................................................ l 6
Central Processor Status Register ........................................ 6
CORE MEMORY ............................................................................. 6
PERIPHERAL DEVICES ................................................................ 7
SYSTEM INTERACTION ................................................................
TRANSFER OF BUS MASTER ...................................................... 5
PRIORITY STRUCTURE ............ .................................................... 7
NPR Requests ....................................... .I ...........................
Interrupt Requests .............................................................. :
CHAPTER 3 ADDRESSING MODES
INTRODUCTION ............................. . ..................... .; ..................... 11
ADDRESS FIELDS ........................................................ . ............... 11
General Register Addressing ....................... .:. ...................... 11
*Indexed Addressing ........................ . ..................................... 12
Autoincrement Mode Addressing ........................................ 12
Autodecrement Addressing .................... . ............................. 12
STACK PROCESSING .................................................................. 13
USE OF THE PC AS A GENERAL REGISTER .......................... _... 13
Immediate Addressing ......................................................... 13
Absolute Addressing ......................................................... .:. 13
Relative Addressing .............................................................. 14
Deferred Relative Addressing .......................................... ..:. 14
USE OF THE SP AS A GENERAL REGISTER ................................ 14
. DOUBLE OPERAND ADDRESSING .............................................. 14
&iAPTER 4 INSTRUCTlOil Sk
INSTRUCTION TIMING ................................................................ 17
NOTATION .................................................................................... 17

Ill
. .
DOUBLE OPERAND-INSTRUCTIONS .......................... i.. ............... 17
Arithmetic Operations .................. 1........................................ 18
Boolean Instructions ............................................................. 20
BRANCHES .......................................................................... .:. ..... 21
Unconditional Branch ..........................................................
I Simple Conditional Branches .............................................. ;:
Signed Conditional Branches .............................................. 23
Unsigned Conditional Branches ..........................................
JUMP .................................................................................... 9:
SUBROUTINES ............................................................................. 27
Examples ............................................................................ 28
SINGLE OPERAND INSTRUCTIONS ............................................
Multiple Precision Operations .............................................. zi
Rotates ................................................................................ 33
Shifts .................................................................................. 34
Examples ............................................................................ 36
BYTE OPERATIONS ...................................................................... 36
Double Operand Byte Instructions ...................................... 36
Example .......... . ................................................................... 37
Single Operand Instructions ................................................ 38
CONDITION CODE OPERATORS ................................................
MISCELLANEOUS CONTROL INSTRUCTIONS .............................. , i!
PROCESSOR TRAPS .................................................................... 41
Trap Instructions ............................ . ................................... 41
Stack Overflow Trap ............................................................
Bus Error Traps .................................................................. ii
Trace Traps .......................................................................... 43
CHAPTER 5 I ADDRESS ALLOCATION
ADDRESS MAP ............................................................................ 45
Interrupt and Trap Vector .................................................... 46
Processor Stack and General Storage ..................................
Peripheral Registers ............................................................ z
CORE MEMORY ............................................................................ 46
Read-Write Core Memory .................................. , ................. 46
Read-Only Core Memory ...................................................... g
Wordlet Memory ..................................................................
CHAPTER 6 PROGRAMMING OF PERIPHERALS
DEViCE REGISTERS ...................................................... ..... . ......... ;; -
CONTROL & STATUS REGISTERS ...............................................
Device Function Bits ....................................... :.. ................ $
Memory Extension ....................................................... . ......
Done Enable and Interrupt Enable ......................................
Condition Bits ............................................................. ..i .... ii
Unit Bits .............................................................................. g
Error Bits .............. I ..:. .........................................................
DATA BUFFER REGISTERS .................................................. .: ...... 48
PROGRAMMING EXAMPLES--NON INTERRUPT ........................ 48
INTERRUPT STRUCTURE ............................................................ 50
PROGRAMMING EXAMPLE .......................................................... 51
CHAPTER 7 PERIPHERAL BULLETINS .
TELETYPE (MODEL LT33-DC/DD) .............................................. ;;
Size ......................................................................................
Power Requirement ............................................................ 53

IV
 . \
TELETYPE CONTROL (MODEL KLll) ......................................... 53
Teletype Control .......................................................... r ....... 53
Keyboard/Reader Operation ................................................ 53
Registers (TKS, TKB) .................................................. 54
Teleprinter/Punch ................................................................ 54
Registers (TPS, TPB) .................................................... .55
Programming Example ........................................................ 55
Peripheral Address Assignments .......................................... 55
Mounting ...................................... ....................................... 55
HIGH-SPEED PERFORATED TAPE READER (MODEL PCll) ...... 55
Tape Reader ........................................................................ 55
Registers (PRS, PRB) ................................................... 56
Programming Example ................................................ 56
Peripheral Address Assignments ................................ 56
Tape Punch .......................................................................... 56
Registers (PPS, PPB) .................. ., ................................ 57
Programming Example ................................................
Peripheral Address Assignments .................................. E
Mounting ............................................................................... 57
Environmental ....................................................................... 58
Line Frequency Clock (Model KWll-L) ............ I,....................
Register ..................................................................... .:. .... . .... 5588
Peripheral Address Assignments .......................................... 58
Mounting ............................................................................... 58
Vector Address .................................................................... 58
Priority Level ........................................................................ 58
CHAPTER 8 DESCRIPTION OF THE UNIBUS (
GENERAL CONCEPTS OF THE UNIBUS ...................................... 59
Single Bus ............................................................................ 59
Bidirectional Bus ................................................................ 59
Master-Slave Relation .......................................................... 59
. Interlocked Communication .................................................. 60
Dynamic Master-Slave Relation ............................................. 60
UNIBUS SIGNALS ........................................................................ 60
NON-INTERRUPT SIGNALS ..........................................................
Data Lines .................................. ..r.~. ................................... :?I
Address Lines ......................................................................
Control Lines .:. ..................................................................... E
Master Sync &‘Slave Sync ..................................................
Parity Available & Parity Bit ................................................ ::
Initialization ........................................................................ 61
Spare 1 & Spare 2 .............................................................. ,61
INTERRUPT SIGNALS ....................................... *. ....................... 61
Bus Request Lines ........................................................... 9..
I Bus Grant Lines ................. r.. .............................................. z:
Non-Processor Request ........................................................ 61
Non-Processor Grant ............................................................ 61
Selection Acknowledge ........................................................
Interrupt (and) Bus Busy .................................................... 6":
UNIBUS DATA TRANSFER OPERATIONS .................................... 61
DATO and DATOB ................................................................
DATI and DATIP .................................................................. :z
Examples of Data Transfers ......... . .................. ........ . ........... 62
Signal Description of Data Transfers ......... . ........................ 63

V
 UNIBUS CONTROL ............. ’.........................................................
Priority Arbitration .................................................................
Selection of Next m m ..............................................
Interrupt Sequence ................................... . ..........................
Example of Interrupt, etc. ..................................................
Example of NPR Operation ..................................................
CHAPTER 9 WI’TERF~MC
REGISTERS .................................................................................. 69
BUS DRIVERS AND RECEIVERS .................................................. 69
ADDRESS SELECTOR ..................................................................
INTERRUPT CONTRbL ................................... .............................
DEVICE CONTROL LOGIC ............................................................ 76
CHAPTER 10 CONFlGUMTICM AND MSTALLATIOW PlANNlm;
MODULAR CONSTRUCTION ........................................................
MOUNTING BOXES AND CABINETS ............................................
. PDP-11 Tabletop Box for 11/20, Etc. ..................................
PDP-11 Basic Mounting Box ................................................
PDP-11 Tabletop Extension Mounting Box ..........................
PDP-1 l-Freestanding Base Cabinet ............................. . ......
Freestanding Programmer’s Table .........................................
SYSTEM UNITS AND CABLES ....................................................
Peripheral Mounting Unit ....................................................
Blank System Unit ........................................................... .: ...
Unibus Module ....................................................................
Unibus Cable ...... :. ................................................................
CABLE REQUIREMENTS ..............................................................
PDP-11/20 POWER REQUIREMENTS ..........................................
TELETYPE REQUIREMENTS ........................................................
ENVIRONMENTAL REQUIREMENTS ............................................
INSTALLATION PROCEDURE ..................... ................................
CHAPTER 11 PAPER TAPE SOFlWARE SYSTEM
PTS FEATURES ............................................................................
PAL-11A Assembler ..............................................................
.
ED11 Editor ..........................................................................
ODT On-Line Debugging ......................................................
IOX Input/Output, etc. ........................................................
Math Package .......................................................................
Loaders ................................................................................
Core Dump Routines ..................................................... . ......
CHAPTER 12 THE OPERATOR’S CONSOLE
\ CQNSOLE ELEMENTS ..................................................................
Indicator Lights ..................................................................... :;
Register Displays ..................................................................
Switch Register ....................................................................
Control Switches .......................................................... i’. .....
x;
a7
.. CONTROL SWITCH OPERATION .................................................. 89
APPENDIX A-PDP-11 INSTRUCTION REPERTOIRE .................. 91
APPENDIX B-ADDRESSING SUMMARY .................... ..- ............. 95
ADDRESSING MODES ......................................... . ......................
General Register Addressing ................................................ E
PC Register Addressing ............. \. ........... . ........................... 95

--
VI
INSTRUCTION FORMATS ._.___...........___........................................ 95

APPENDIX C-ADDRESS MAP . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . 97

APPENDIX D-UNIBUS OPERATIONS .......... ........ ............ ........ .. 99
DATA-TRANSFERS .__............._._..................................................... 99
DATI and DATIP . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DATO and DATOB . . . .._.......................................................... / lzo”
F’TR-PRIORITY TRANSFER . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . ..__...................... 101
INTR-INTerRupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 102
GENERAL NOTES ON THE BUS OPERATIONS _.__._,...._...__.,.,,...... 102

VII
The PDP-11”is available in two versions-PDP-ll/lO and PDP-
11/20. The basic PDP-ll/lO contains 1,024 words of read only
memory in conjunction with 128 words of read/write memory and
the basic PDP-ll/PO includes 4,096,words of read/write memory.

VIII
 CHAPTER 1
INTRODUCTION
This publication is a handbook for Digital Equipment Corporation’s PDP-11.
It provides a comprehensive overview of the system structure, the instruction
repertoire, input/output programming, peripherals, general interfacing, soft-
ware, and console operation.

PDP-11 is Digital’s answer to the demand for a modular system for real-time
data acquisition, analysis and control. PDP-11 systems can handle a wide
variety of real-time control applications-each system being individually
tailored from a comprehensive array of modular building blocks. Digital is
unique among manufacturers of small-scale computers-in its ability to pro-
vide not only fast and efficient processing units, but also a large family of its
own compatible I/O devices including A/D and D/A converters, magnetic
tape, disk storage, paper tape, and displays, as well as a wide range of
general-purpose modules. This capability offers the user a hew, more efficient
approach to real-time systems.
The following paragraphs introduce the new PDP-11 by way of highlighting
several of the important design features that set it apart from other machines
in its class. Subsequent chapters of this manual place these features in their
proper context and provide detailed descriptions of each.

PDP-11 SYSTEMS
The PDP-11 is available in two versions designated as PDP-ll/ 10 and PDP-
11/20. The PDP-ll/ 10 contains a KAll processor, 1,024 words of 16-bit
read-only memory, and 128 16-bit words of read-write memory. The basic
PDP-ll/PO contains a KAll processor and 4,096 words of 16.bit read-write
core memory, a programmer’s console, and an ASR-33 Teletype. Both ver-
sions can be similarly expanded with either read-write or read-only memory
and peripheral devices.

UNIBUS _
Unibus is the name given to the single bus structure of the PDP-11. The
processor, memory and all peripheral devices share the same high-speed
bus. The Unibus enables the processor to view peripheral devices-as active
memory locations which perform special functions. Peripherals can thus be
addressed as memory. In other words, memory reference insfructions can
operate directly on’control, status, or data registers in peripheral devices.
Data transfers from input to output devices can bypass the processor com-
pletely.

KAll PROCESSOR
The KAll processor incorporates a unique combination of powerful features
not previously available in ldbit computers.
Priority Interrupts-A four-level automatic priority interrupt system permits
the processor to respond automatically to conditions outside the system, or
in the processor itself. Any number of separate devices can be attached to
each level.
Each perkpheral device in a PDP-11 system has a hardware pointer to its own
unique pair of memory locations which, in turn, point to the device’s service
routine. This unique identification eliminates the need for polling of devices

1
 to identify an interrupt, since the interrupt servicing hardware selects and
begins executing the appropriate service routine.
The device’s interrupt priority and service routine priority are independent.
This allows dynamic adjustment of system behavior in response to real-time
conditions.

The interrupt system allows the processor continually to compare its own
priority levels with the levels of any interrupting devices and to acknowledge
the device with the highest level above the processor’s priority level. Servic-
ing an interrupt for a device can be interrupted for servicing a higher priority
device. Service to the lower priority device can be resumed aUtOmaticallY
upon completion of the higher level servicing. Such a process, called nested
interrupt servicing, can be carried out to any level.

Rentrant Code-Both the interrupt handling hardware and the subroutine

call hardware are designed to facilitate writing reentrant code for the
PDP-11. This type of code allows use of a single copy of a given subroutine
or program to be shared by more than one process or task. This reduces,the
amount of core needed for multi-task applications such as the concurrent
servicing of many peripheral devices.
General Register-The PDP-11 is equipped with eight general registers. All
are program-accessible and can be used as accumulators, as pointers to
memory locations, or as full-word index registers. Six registers are used for
general-purpose access while the seventh and eighth registers are used as
a stack pointer and program counter respectively.
Instruction Set-An important feature of the PDP-11 instruction set is the
availability of double operand instructions. These instructions allow memory-
to-memory processing and eliminate the need to use registers for Storage of
intermediate results. By using double operand instructions, every memory
location can be treated as an accumulator. This significantly reduces the
length of programs by eliminating load and store operations associated with
single operand machines.
Addressing-Much of the power of the PDP-11 is derived from its wide range
of addressing capabilities. PDP-11 addressing modes include list sequential
addressing, full address indexing, full 16qbit word addressing, 8-bit byte
addressing, stack addressing, and direct addressing to 32K words.
Variable length instruction formatting allows a minimum number of bits to
be used for each addressing mode. This results in efficient use of program
storage space.
Asynchronous Operation-The PDP-11’s memory and processor operations
are asynchronous. As a result, I/O devices transferring directly to or from
memory may steal memory cycles during instruction operation.

PACKAGING ’
The PDP-11 has adopted a modular approach to allow custom configuring Of
systems, easy expansion, and easy servicing. Systems are composed of basic
building blocks, called System Units, which are completely independent sub-
systems connected only by pluggable Unibus and power connections.
There is no fixed wiring between them. An example of this type of subsystem
is a 4,096-word memory module.
System Units can be mounted in many combinations within the PDP-11
hardware, since there are no fixed positions for memory or l/O device con-
trollers. Additional units can be mounted easily and connected to the system

.
in the field. In case maintenance is required, defective System Units can be
replaced with spares and operation resumed within a few minutes.
-E ’
A corn-e package of user-oriented software includes:
Absolute assembler providing object and source listings
Stritig-oriented editor
Debugging routines capable of operating in a priority interrupt environ-
mint
Input/output handlers for standard peripherals
Relocatable integer and floating point math library

3
All PDP-11 processors, memories and peripherals are electrically
and mechanically modular subsystems .supported in System Units.
which are simply plugged together to form a computer tailored to
user needs.

‘4
 I
CHAPTER i
SYSTEM INTRODUCTION
SYSTEM DEFINITION
Digital Equipment Corporation’s PDP-11 is a 16.bit, general-purpose, parallel-
logic computer using two’s complement arithmetic. The PDP-11 is a variable ’
word length processor which .directly addresses 32,768 16.bit words or
65,536 8-bit bytes. All communication between system components is done
on a single- high-speed bus called a Unibus. Standard features of the system
include eight general-purpose registers which can be used as accumulators.
index registers, or address pointers, and a multi’level automatic priority in-
terrupt system.

SYSTEM COMPONENTS
UNIBW-There are five concepts that are very important for understanding
both the hardware and software implications of the Unibus.

Single Bus--The Unibus is a single, common path that connects the central
processor memory, and all peripherals. Addresses, data, and control informa-
tion are sent along the 56 lines of the bus.

The form of communication is the same for every device on the Unibus. The
processor uses the same set of signals to communicate with memory as with
peripheral devices. Peripheral devices also use this set of signals when com-
municating with the processor, memory, or other peripheral devices.

Peripheral device registers may be manipulated as flexibly as core memory

by the central processor. All the instructions that can be applied to data in
core memory can be applied equally well to data in peripheral device regis-,
.ters. This is an especially powerful feature, considering the special capability
of PDP-11 instructions to process data in any memory location as though it
were an accumulator.
Bidirectiona) Lines--Unibus lines are bidirectional, so that the same signals
which are received as input can be driven as output. This means that a
peripheral device register can be either read or set by the central processor
or other peripheral devices; thus, the same register can be used. for both
input and output functions.

Master-Slave Relation-Communication between two devices on the bus is

in the form of a master-slave relationship. At any point in time, there is one
device that has control of the bus. This controlling device is termed the
“bus master.” The master device controls the bus when communicating with
another device on the bus, termed the “slave.” A typical example of this
relationship is the processor, as master, fetching an instruction from mem-
ory (which is always a slave). Another example is the disk, as master, trans-
ferring data to memory, as slave.
interlocked Communication--Communication on tliq Unibus is interlocked
so that for each control signal issued by the master device, there must be-a
response from the slave in order to complete the transfer. Therefore, com-
munication is independent of the physical bus length and the response time
of the master and slave devices. The maximum transfer rate on the Unibus
is one 16-bit word every 750 nanoseconds, or 1.3 million 16-bit words- per
second.
Dvnamlc
Master-tive
#?e!&eMastar-slave relationships are dynamic. The
,processor, for example, may p&s bus control to a disk. The disk, as master,
could then communicate with a slave memory bank.
Since the Unibus is used by the processor and all I/O devices, there is a
priority structure to determine which device gets control of the bus. There
fore, every device on the Unibus which is capable of becoming bus master
has a ‘Priority assigned to it. When two devices which are kapable of becorn-
ing a bus master request use of the bus simultaneously, the device with the
higher priority will receive control first. Details of what conditions must be
satisfied before a device will get control of the bus are given in the section
on System Interaction.
,+(A11 CENTRAL PROCESSOR-There are four major features which are of
particular interest to the programmer: l), the General Registers: 2), the
Processor Status Word; (3), the Addressing Modes; and 4), the Instruction
Set. The addressing modes and the instruction set of the PDP-11 processor
will be discussed in detail in Chapters 3 and 4.
‘Ganeral Registers-The KAll processor contains eight 16.bit general regis-
Ms. These eight general registers (referred to as RO, Rl, . . . . . R7) may
be used as accumulators, as index registers, or as stack pointers. One of
these registers, R7, is reserved as. a program counter (PC). Generally, the
PC holds the address of the next instruction, but it may point to data or
to an address of data. The register R6 has the special function of processor
stack pointer.
Central Processor Status Register-The Central Processor Status Register
(PS) contains information on the current priority of the processor, the result
of previous operations, and an indicator for detecting the execution of an
instruction to be trapped during program debugging. The priority of the
central processor can be set under program control to any one of eight.
levels. This information is held in bits 5, 6, and 7 of the PS.
Four bits of the PS are assigned to monitoring different results of previous
instructions. These bits are set as follows:
Z-if the result was zero
N-if the result was negative .
C-if the operation resulted in a carry from the most significant bit
V-if the operation resulted in an arithmetic overflow
The T bit is used in program debugging and can be set or cleared under pro-
gram control. If this bit is set, when an instruction is fetched from memory
a processor trap will’ be caused by the completion of the instruction’s
execution.

Central Processor Status Register (PS)

CORE MEI;(ORy-The PDP-11 allows both 16.bit word and 8-bit byte ad-
dressing. The address space may be filled by core memory and peripheral
device registers. The top 4,096 words generally are reserved for peripheral
device registers. The remainder of address space can be used for read-write
core memory or read-only core memory.
Read-write core memory is currently available in 4,096 1Qbit word segments.
This memory has a cycle time of 1.2. microseconds and an access time of
500 nanoseconds. It is a standard part of a PDP-ll/PO system.

6
Read-only core memory (ROM) is available in 1,024 16 bit-word segments.
The access time of the ROM is 500 nanoseconds. Memory is also available in
128 16-bit word segments with a 2.0 microsecond cycle time. Both 1,024
words of read-only memory and 128 words of read-write memory mount in
a single System Unit and are a standard part of the PDP-ll/lO system.
PERIPHERAL DEVICES-The ASR-33 Teletype with low-speed paper tape
reader and punch is provided in the basic PDP-11/20 system. Options for the
.PDP-11 include a paper tape reader capable of reading 300 characters per
second, a paper tape punch with an output capacity of 50 characters per
second, and additional Teletype units. Provision is made for the addition
of numerous peripheral devices. These include standard DEC peripherals as
well as other devices which will be unique to the PDP-11.

SYSTEM INTERACTION
At any point in time only one device can be in control of the bus, or be bus
master. The master communicates with another device on the bus which is
called the slave. Usually, the established master will communicate with the
slave in the form of data transfers.
Full 16-bit words or 8-bit bytes of information can be transferred on the bus
between the master and the slave. The information can be instructions, ad-
dresses, or data. This type of ‘operation occurs when the processor, as
master, is fetching instructions, operands, and data from memory, and re-
storing the results into memory after execution of instructions. Pure data
transfers occur between a disk control and memory.
TRANSFER OF BUS MASTER-When a device (other than the central pro-
cessor) is capable of becoming bus master and requests use of the bus, it is
generally for one of two purposes: 1) to make a non-processor transfer of
data directly to or from memory, or 2) to interrupt program execution and
force the processor to branch to a specific address where an interrupt
service routine is located.
PRIORITY STRUCTURE-When a device capable of becoming. bus master
requests use of the bus, the handling of that request depends on the loca-
tion of that device in the priority structure. These factors must be considered
to determine the priority of the request;
1. The processor’s priority can be set under program control to one of
eight levels using bits 7, 6, and 5 in the processor status register.
These three bits set a priority level that inhibits granting of bus re-
quests on lower levels.
2. Bus requests from external devices can be made on one of five re-
quest lines. A non-processor request (NPR) has the highest priority,
and its request is honored by the processor between bus cycles of
an instruction execution. Bus request 7 (BR7) is the next highest
--priority, and BR4 is the lowest. The four lower level priority requests
are honored by the processor between instructions. When the pro-
cessor’s priority is set to a level, for example 6, all bus requests on
BR6 and below are ignored.
3. When more than one device is connected to the same bus request
(BR) line, a device nearer the central processor has a higher priority
than a device farther away. Any number of devices can be connected
to a given BR or NPR line.
Once’s device other than the processor has control of the bus, it is for one
of two types of requests: 1) NPR Request, 2) ‘Interrupt Request.

7
NPR Requeata-NPR data transfers can be made between any two peripheral
devices without the supervision of the processor. Normally, NPR transfers
are between a mass storage device, such as a disk, and core memory.
The structure of the bus also permits device-to-device trat’ISfer% allowing.
customer-designed peripheral controllers to access other devices such as
disks directly. -
An NPR device has very fast access to the bus and can transfer at high data
rates once it has control. The processor state is not affected by the transfer:
therefore the processor can relinquish control while an instruction is in
Progress. This can occur at the end of any bus cycle except in between a
read-modify-write-sequence. (See Chapter 8 for details). In the PDP-11, an
NPR device can gain bus control in 3.5 microseconds or less. An NPR device
in COritrOl of the bus may transfer ldbit words from memory at memory
speed or every 1.2 microseconds in the PDP-ll/EO or every 1.0 microseconds
in the PDP-ll/lO.
IIIterrUpt Requests-Devices that request interrupts on the bus request lines
(BR7, BR6, BR5, BR4) can take advantage of the power and flexibility of
the processor. The entire instruction set is available for manipulating data
and status registers. When a device servicing program must be run, the task
currently under way in the central processor is interrupted and the device
service routine is initiated. Once the device request has been satisfied, the
processor returns to the interrupted task.
In the PDP-11, the return address for the interrupted routine and the proces-
sor status word are held in a “stack.” A stack is a dvnamic seauential
list of data with special provision for access from one end. A stack-is also
called a “push down” or “LIFO” (Last-In First-Out) list. Storaee and re-
trieval from stacks is called “pushing” and “popping” respecti&ly. These
operations are illustrated in Figure 2-1.
In the PDP-11, a stack is automatically maintained by the hardware for inter-
rupt processing. Thus, higher level requests can interrupt the processing of
lower level interrupt service, and automatically return control to the lower
level interrupt service routines when the higher level servicing is completed.
Here is an example of this procedure. A peripheral requires service and
requests use of the bus at one of the.BR levels (BR7, BR6, BR5, BR4). The
operations undertaken to “service” the device are as follows:

,E0

I l.AN EMPTY
STACK
3.PlJStlING
~tiW&CNTO
ANOTHER
THE

E2 E3 E3

El El

E0

4. ANOTHER
PUSH

Fig 2-1
El
5 POP
El

E0

Illustration
I3
6.
E4

EO

PUSH

of Push and Pop Operations

8
Priorities permitting, the processor relinquishes the bus to that
device.
When the device has control of the bus, it sends the processor an
interrupt command with the address of the words in memory con-
taining the address and status of the appropriate device service
routine.
The processor then “pushes”- first, the current central processor
status (PS) and then, the current program counter (PC) onto the
processor stack.
The new. PC and PS (the “interrupt vector”) are taken from the loca-
tion specified by the.device and the next location. and the device

Figure 2-2 Nested device Servicing

9
 service routine is begun. Note that those operations all occur auto-
matically and that no device-polling is required to determine which
service routine to execute.
5. 7.2 microseconds is the time interval between the central Processor’s
receiving the interrupt command and the fetching of the first inStruC-
tion. This assumes there were no NPR transfers during this time.
6. The device service routine can resume the interrupted process by
executing the RTI (Return from interrupt) instrudion which ‘ipops”
the processor stack back into the PC and PS. This requires 4.5
microseconds if there are no intervening NPR’s.
7, A device service routine can be interrupted in turn by a sufficiently
high priority bus request any time after completion of its first in-
struction.
8. lf such an interrupt occurs, the PC and PS of the device service .
routine are aUtOITX3tiCally pushed into the stack and the new device
routine initiated as above. This “nesting” of priority interrupts can
go on to any level, limited only by the core available for the stack.
More commonly, this process will nest only four levels deep since
there are four levels of BR signals. An example of nested device
servicing is shown in Figure 2-2. A rough core map is given for each
step of the process. The SP points to the top word of the stack as
shown.

10
 CHAPTER 3
ADDRESSING jdODES‘
Most data in a program is structured in some way-in a table, in a stack, in
a table of addresses, op perhap$ in a small set of frequently-used variables‘
local to a limited region of a program. The PDP-11 handles these common
data structures -with addressing modes specifically designed for each kind
of access. In addition, addressing for unstructured data is general enotfgh *’
to permit direct random ac%ess to all of core. Memory is not brokeri up into
pages and fields (often awkward and wasteful of core storage).
Addressing in the PDP-11 is dohe through the general registers. PrOWems
requiring several stacks can use the general registers for stack pointers.
Those requiring many local variables can use general registers as accumu-
lators. The general registers can be used interchangeably as index ,registen
or as sequential list pointers to access tabularrdata. Address arithmetic may
be done directly in the general registers. ,
ADDRESS FIELDS
PDP-11 instruction words contain a 6-bit address field divided into two sub-
fields selecting the general register and the mode generating the effective
address.

The register subfield specifies which of the eight general registers is to be

used in the address calculation. The mode subfield indicates how this register
is to be used in determining the operand. These modes will be described
in the following paragraphs.
GENERAL REGISTER ADDRESSING-The general registers will be used .as
simple accumulators for operating on frequently-accessed variables. In this
mode, the operand is held directly in the general register. The general reg-
isters are in iast memory, (280-nanosecond cycle time) resulting in a speed
improvement for operations on these variables.
PAL-11, the PDP-11 assembler, interprets instructions of the form
OPR R
as general register operations. R has been defineb as a register name and
OPR is used to represent a general instruction mnemonic. ihe address field
for general register operations is

Operands that are pointed to -by addresses (indirect or deferred) are de-
noted to PAL-11 by the @ symbol. Thus, instructions of the form
*.
OPR @R
specify deferred register addressing and have the following address field.

11
Deferred register addressing may also be selected in PAL-11 by the form
OPR (R).
INDEXED AD~RE&G-T~IZ general fegisters may be used as index reg
jr&en to per&t random access of items in tables or stacks of data. InStrUC-
tions of the form
OPR X(R)
specify indexed mode addressing. The effective address is the sum of X
and the contents of the specified general- register R.
The index word containing X follows the instructi& word.

Index mode addressing can be deferred to permit access of data elements

through tables or stacks of their addresses. The address field for index de-
ferred mode is

It is specified by instructions of the form

OPR @X(R) _

AUTOINCREMENT ADDRESSlNCiAutoincrement addressing provides for

automatic stepping of a pointer through sequential elements of a table
of operands. In this mode, the address of the operand is taken from the
general register and then the contents of the register are stepped (incre-
mented by one or two) to address/the next word or byte depending upon
whether the instruction operates on byte or word data. Instructions of the
form *

01% (RI+
specify autoincrement addressing. The address field for autoincrement ad-
dressing is

This mode may also be deferred. Instructions of the form

OPR @(Wk
specify deferred autoincrement addressing and assemble with the followihg
address field.

AUTODECREMENT ADDRESSlN*Autode&ement addressing steps the spe-

cified general register to the next lower byte (decrement by one) or word

12
 (decrement by two) address and-uses the new contents of the general reg-
ister as the operand address. Instructions of the form
-0PR -(RI
specify autodecrement addressing. The address field for autodecrement ad- ’
dressing is

This mode also may be deferred and specified by instructions of the form
OPR @ -(R). When deferred the address field is

STACK PROCESSING
The combination of autoincrement addressing in which the general register is
stepped forward after the operand address is determined and autodecrement
addressing in which the general register is stepped backward before the
operand address is determined is the basic requirement for convenient low
overhead stack operations.
The PDP-11 has extensive stack processing capabilities. The stack pointer
(SP), R6, maintains a stack for the nested handling of interrupts. All of the
general registers can maintain stacks under program control. Elements in
the middle of stacks may be accessed through indexed addressing. This
provides for convenient access of dynamically assigned temporary storage,
especially useful in nested procedures.

USE OF THE PC AS A GENERAL REGISTER

There are special implications in the use of the addressing modes already
described when applied to the PC (R7). The use of the PC with the address-
ing modes described above generates immediate, absolute, relative, and
deferred relative addressing.
IMMEDIATE ADDRESSING-Immediate addressing provides time and space
improvement for access of constant operands by including the constant in .
the instruction. The instruction word referencing an immediate operand
specifies autoincrement addressing through the program counter. The ad-
dress field would be

The program counter points to the word after the instruction word. The con-
tents of this word are therefore used as the operand and the program counter
. is stepped to the next word. PAL-11 recognizes address expressions of the
form “#n” as immediate operands and codes them with the address field
shown above followed by a word of data (n).
A full word is assembled for immediate operands even in byte instructions
so that instruction words are always fetched from even locations.
ABSOLUTE ADDRESSING-The contents of the location following the instruc

13
tion word may be taken as the address of an operand by specifying deferral
in immediate mode addressing. That is, instructions of the form

refer to the operand at, address A. PAL-11 assembles address cxprwssions of

this form into an address field

followed by a word containing the o erand address.

P
MUTmE ADDRESSIF&-Relative addressing specifies the operand address
relative to the instruction location. This is accomplished by using the pc as
an index regkter. The PC is considered as a base address. The of&&, the
distance betwe’en the. location of the operand and the PC, is held in the
index word of the instruction. PAL-11 assembles instructions of the form

OPR A
(where A has not been assigned as a name of a general register) as an
instruction word with the address field

followed by an index word of the form

k-f OFTM*IIIID.2

DEFERRED RELATIVE ADDRESSING-Deferral of relative addressing permits

access to data through memory locations holding operand addresses. The
“@I” character specifies deferred addressing: i.e., OPR @A. The address field
for deferred relative addressing is

USE OF THE SP AS A GENERAL REGISTER

The processor stack pointer will in most cases be the general register used
in PDP-11 stack operations. Note that the content of SP, (SP), refers to the
top element of the stack, that -(SP) will push data onto the stack, that
(SP)+ will pop data off the stack, and that X(SP) will permit random access
of items on the stack. Since the SP is used by the processor for interrupt
handling, it has a special attribute: autoincrements and autodecrements are
always done in steps of two. Byte operations using the SP in this way will
simply leave odd addresses unmodified.

DOUBLE OPFRAND ADDRESSING

Operations which imply two operands such as add, subtract .and compare
are presented in the PDP11 by instructions which specify two addresses. The *
instruction word for 6uch operations is of the form

Instruction Word-Double Operand Instructions

14
and is followed by index words and immediate operands for the source and
destination address fields as appropriate. Source address calculations are
performed before destination address calculations. Since each operand may
be anywhere in core storage or in the general registers, each memory location
is thus effectively provided with the arithmetic capabilities of an accumulator.
Further, since peripheral device registers and memory location are addressed
in the same way, the contents of peripheral data buffers can be stored or
loaded directly to and from memory without use of any general register. This
means that interrupt routines can be executed without saving and restoring
any of the general registers.

19
 CHAPTER 4
INSTRUCTION SET
This chapter Presents the order code for the PDP-11. Each PDP-11 instruc-
tion is described in terms of five parameters: operation, effect on condition
codes, base timing, assembler mnemonics, and octal representation. Special
comments are included where appropriate.

NOTATION
The following notations will be used in this section:
(XXX) : The contents of XXX
src : The Source Address
dst : The Destination Address
A : Boolean “AND” Function
V : Boolean “OR” Function
tf : Boolean “Exclusive OR” Function
: Boolean ‘NOT” Function (Complement)
: “becomes”
: “is popped from the stack”
i : “is pushed onto the stack”

INSTRUCTION TIMING
The PDP-11 is an asynchronous processor in which, in many cases, memory
and processor operations are overlapped. The execution time for an instruc-
tion is the sum of a basic instruction time and the time to determine and
fetch the source and/or destination operands. The following table shows the
addressing times required for the various modes of addressing source and
destination operands. The instruction time for each operation is given
(throughout this chapter) for the 11/20 configuration. All times stated are
subject to +20% variation.
ADDRESSING FORM TIMING
(src or dst) src bs)t dst Wt
R 0 0
(RI or @R 1.5 1.4*
1.5 1.4’
L”c’Rf 1.4*
@(W + ::7” 2.6*
G?-(R) 2.6*
BASE(R) 2:; 2.6*
@BASE(R) or @(R) 3.9 3.8’
l dst time is .4 ws. less than listed time if instruction was a
CoMPare. CoMPare Byte
Bit Test, Bit Test Byte
TeST, or TeST Byte
none of which ever modify the destination word.
t referencing bytes at odd addresses adds 0.6~s to sn and dst times.

DOUBLE OPERAND INSTRUCTIONS-Double Operand Instructions are repre-

sented in assembly language as:
OPR src, dst
,’ where src and dst are the addresses of the source and destination operands
respectively. The execution time for these operations is comprised of the
source time, the destination time; and the instruction time. The source and
destination times depend on addressing modes and are described in the pre.
ceding table.

17
Arithmetic Operations-

Operation: (src) + (dst)

Condition Codes:
Z: set if (src) = 0; cleared otherwise
N: set if (src) < 0; cleared otherwise
C: not affected
V: cleared
Description: Moves the source operand to the destination location. The pre-
vious contents of the destination are lost. The contents of the source are
not affected.
The MOV instruction is a generalization of ‘load,” “store,” “setup,” ‘push,”
“pop,’ and interregister transfer operations.
General registers may be loaded with the contents of memory addresses with
instructions of the form:
MOV src, R
Registers may be loaded with a counter, and pointer values with MOV in-
structions:
MOV #n. R
iwhich loads the number n into register R)
Operands may be pushed onto a stack by:
MOV src, -(R)
and may be popped off a stack by:
MOV (R)+, dst

Interregister transfers are simply:

MOV RA, RB
(RA and RB are general registers)
Memory-to-memory transfers may be done with the MOV instruction in the
general form:
MOV src, dst
ADD ADD WC. dsi 2.3~

0 , 6 WC dst ,
I I I I I 1 I I
(5 12 11 6 5 0

Operation: (src) + (dst) + (dst)

Condition Codes: Z: set if result = 0; cleared otherwise
N: set if result < 0; cleared otherwise
C: set if there was a carry from the most significant bit
of the result; cleared otherwise
V: set if there was arithmetic overflow as a result of the
operation, that is, if both dperands were of the same
sign and the result was of the opposite sign; cleared
otherwise

18
 Description: Adds the source operand to the destination operand and stores
the re&lt at the destination address. The original contents-of the destination
are lost. Ttr, ktmcI of the sowe are not a#ectod. Two’s cwpt addi-
tion is perfoti.
The ADD instruction inckdes as special cases the “add-to-register,” “add-t@
memory,” and ‘Md-reister-to-rwister” functions:
Add-to&gister ADD WC, R ,
A&l-to*mory ADD R, +t
Add Register-to-Register ADD RA. RB

Arithmetic may also be done directly in memory by the general form ADD
instruction
ADD src, dst
.
Use of this form saves considerable loadindand storing of accumulators.

Two special cases of the ADD instruction are particularly useful in coppilers,
interpreters, and other stack arithmeti processes:
ADD (R +, (RI
(where R is th II stack pointer)
which replaces the top two elements ‘pf the stack with their sum; and ADD
src. (R), which increases the top eletient of the stack by the contents of
the source address.

The “Add Immediate” operation is y&t another special case of this general-
ized .ADD iristruction:
ADD #n, dst *

Immediate operations are useful in dealing with constant operinds. Note

that:
ADD #n. R
steps the register R (which may be an index register) through n addresses
eliminating the need for a special “add-to-index register” instruction.

All these special cases of the ADD instruction apply equally well to the other
double operand instruetions that follow.
suBtract SUB WC, drt 2.3 YI

1 , 6 WC drt
I t I I I t II t I I t I
15 12 11 6 5 0

Operation: (dst) - (src) + (dst) [in detail, (dst) + - (src) + 1 + (dst)]

Condition Codes: 2: set if result = 0; cleared otherwise
N: set if result < 0; cleared otherwise
C: cleared if there was a carry from the most significant
bit of the result: set othen@se
V: set if there was arithmetic overflow as a result of the
operation, that is, if-the operands were of opposite
4 signs and the sign of source was the Same as the
sign of the result; cleared otherwise.

Description: Subtract; the source operand from the destination operand and
leaves the result at the destination address. The original contents of the
destination are lost. The contents of the source are not affected.

19
 COMParo CMP rrc.dst 2.3ur’

01 2 WC dst
I I I I 1 1 1 I I I t
6 5 0
(5 12 11

.Operation: (src) - (dst) [in detail, (src) + - (dst) + 11

Condition Codes: Z: set if result = 0; cleared otherwise
N: set if ‘result < 0; cleared ptherwise
C: cleared if there was a carry from the most significant
bit of the result; set otherwise
V: set if there was arithmetic overflow; that is, operands
were of opposite signs and the sign of the destination
was the same as the sign of the result; cleared
otherwise.
.
Description: Arithmetically compares the source and destination operands.
Affects neither operand. The only action is to set the condition codes
appropriately.
Boolean Instructions-These instructions have the same format- as the
double operand arithmetic group. They permit operations on data at the
bit level.
Bll set BIS src.dst 2.31~~

0 , 5
It~IJ*I~~I’~l
15 12 11 6 5 0

Operation: (src) V (dst) + (dst)

Condition Codes: ,Z: set if resu’lt = 0; cleared otherwise
N: set if high-order bit of result set; cleared otherwise
C: not affected
V: cleared
Description: Performs “Inclusive OR” transfer between the source and des-
tination operands and leaves the result at the destination address; that is,
corresponding bits set in the source are set in the destination. The original
contents of the destination are lost. The source is not affected.
Bit Clear BIC src, dst 2 9us

0 4 src dst
1 I L I t I t t I I I
15 12 11 6 5 0

Operation: - (src) A (dst) + (dst)

Conditions Codes: Z: set if result = 0; cleared otherwise

N: set if high-order bit of result set; cleared otherwise
C: not affected
V: cleared .

Description: The BIC instructi,on clears each bit in the destination that car-
responds to a set bit in the source. The original contents of the destination
are lost. The contents of the sources are unaffected.
*There is no read/modify/write cycle in the CMP, BIT. and TST operations. This.iaves
0.4 ws in all destination address modes except register mode.

20 ’
 Bll Test BIT rrc,dst .2.9***

* 0, 3 WC dd ,
I I I I I I I I 1 I
. 15 12 11 6 5 0

Operation: (src) A (dst)

Condition Codes: Z: set if result = 0; cleared otherwise

N: set if high-order bit of result set; cleared otherwise
C: not affected
V: cleared
Description: Performs logical “and” comparison of the source and destination
operands and modifies condition codes accordingly. Neither the source nor
destination operands are affected.
The BIT instruction may be used to test whether any of the corresponding
bits that are set in the destination are also set in the source or whether
all corresponding bits set in the destination are clear in the source.
Note.that the operations of BIS, BIC, and BIT are parallel in that the same
mask may be used to set, clear and test the state of particular bits in a word.
BRANCHES-Branches have the instruction format
Opwotion exx IOC Instruction Time
operation code offs*1
I II I I I I II 1 I,1 1 II’1 I
ts - 6 7 0

The offset is treated as a signed two’s complement displacement to be mul-

tiplied by 2 and applied to the program counter. The program counter points
to the next word in sequence. The effect is to cause the next instruction to
be taken from an address, “lot”, located up to 127. words back (-254
bytes) or 128 wordsahead (+ 256 bytes) of the branch instruction. PAL-11
gives an error indication in the instruction if “lot” is outside this range.
.
The PDP-11 assembler handles address arithmetic for the user and com-
putes and assembles the proper offsets field for branch instructions in the
form
Bxx lot
where lot is the address to which the branch is to be made. The branch
instructions have no effect on condition codes.
Unconditional Branch-
BRbnch Wncondilionol) BR IOC 2.6~s

01 lOI I I Ol I4 I I I I I I
15 6 7 0

Operation: lot + (PC)

Description: Provides a way of transferring program control within a limited
range with a one word instruction. The execution time is equal to the in-
struction time (2.6~s) for the operation.

21
gbnpk corldttbnrit Bran&es-Conditioned branches combine in one instruc-
tion a conditional sMp, unconditional branch sequence.
.
Timing for the conditional branches is shown as execution time if the con-.
dition is not met, followed by the execution time if the condition is met (end
a program branch occurs).
Branch on Eauol(z.ro) BEQIOC ~.SILS,~,~.~ILS

offrrt
0 , ,O I I I I I 4 I I I I, I, 1
15 6 7 0

Operation: lot + (PC) if Z = 1

Description: Tests the state of the Z-bit and causes a. branch if Z is set. It
is used to test equality following a CMP operation, to test that no bits set
in the destination were also set in the source fdllowing a BIT operation, and
generally, to test that the result of the previouq operation was zero.
Thus the sequence
CMP A,B ; compare A and B
BEQ C ; branch if they are equal
will branch to C if A = B (A - B A 0)
and the sequence
ADD A.B ; addAtoB
BEQ C ; branch if the result = 0
will branch to C if A + B = 0.
Branch on Not EqualGk~) BNE IOC 1.5113.2.6 ILS

offset
0
0 1 0
I1 1
I I I I I I III I
15 B 7 0

Operation: lot + (PC) if Z = 0

Description: Tests the state of the Z-bit and causes a branch if the Z-bit is
clear. BNE is the complementary -operation to BEQ. It is. used to test in-
equality following a CMP, to test that some bits set in the destination were
also set in the source, following a BIT end, generally, to test that the result
of the previous operation was not zero.
Branch on Minus BMI IQC 1.5u.s, 2.6~

1 I 0 I I 0 III 4 I Ill offsri II I I

15 B 7 0

Operation: lot --, (PC) if N = 1

Description: Tests the state of the’N-bit and causes a branch if N is set. It
is used to test the sign (most significant bit) of the result of the previous
operation.
Branch on PLUS 6PL IOC t.5 AU, 2.61~

0ffS.t
1 I 0 I, 0 I II 0 I I1 I I1 1
15 -. B 7 0
Operation: lot + (PC) if N = 0.
Description: Tests the state of the N-bit and causes a branch if N is,clear.
BPL is the complementary operation to BMI.

Branch on Carry Set BCS lot 1.5~s ,Z.~ALS

1 , 0, ,
31 I4 I I I I I I I
t5 8 7 0

Operation: lot + (PC) if C = 1

Description: Tests the state of the C-bit and causes a branch if C is set. It
is used to test for a carry in the result of a previous operation.

Bmnch on Carry Clear BCC IOC t.5 U.S. 2.61~5

, 0 3 0
1 I I I I I I I IOffset III I I
15 8 7 0

Operation: lot + (PC) if C’= 0

Description: Tests the state of the C-bit and causes a branch if C is clear.
BCC is the complementary operation to BCS.

Branch on overflow set BVS IOC 1.5us,2.6us

offset
1 , 0 I I 2 I I 4 I I I I t I 1 I
15 8 7 0

Operation: lot + (PC) if V = 1

Description: Tests the state of the V-bit (overflow) and causes a branch. if
the V-bit is set. BVS is used to detect arithmetic overflow in the previous
operation.

eranch on Overflow clear WC IOC 1.5us.2.61~~

oftset
1 I 0
I 121 PI I I I I I I 1 I
15 8 7 0

Operation: lot + (PC) if V = 0

Description: Tests the state of the V-bit and causes a branch if the V-bit is
clear. BVC is the complementary operation to BVS.

Signed Condiiional Branches--Particular combinations of the condition code

bits are tested with the signed conditioned branches. These instructions are
used to test the results of instructions in which the operands were consid-
ered as signed (two’s complement) values.

Note that the sense of signed comparisons differs from that of unsigned
comparisons in that in signed 16-bit, two’s complement arithmetic the
sequence of values is as follows:

23
 largest .. ......... .... . . . . . . . . . . . . 077777
077776
positive .
.
.
000001
000000
177777
177776
negative

lQo00l
smallest 100000
whereas in unsigned 16-bit arithmetic the sequence is considered to be .

highest . . . . ~. . . . . . . . . . . . . . . . . . . . . . . . . 177777
.

000002
00000 1
lowest _. .._. . . . _.. ._ ,. .___, 000000

Branch on Less ThadZero) BLT lot l.Sir. 2.6~

offset
01 0 2 4
I I I II I I I I I I I
15 8 7 0

Operation: lot + (PC) if N V V = 1

Description: Causes a branch if the “Exclusive OR” of the N- and V-bits are’
1. Thus BLT will always branch following an operation that added two neg
ative numbers, even if overflow occurred.
In particular, BLT will always cause a branch if it follows a CMP instruction
operating on a negative source and a positive destination (even if overfloti
occurred). Further, BLT will never cause a branch when it follows a CMP
instruction operating on a positive source and negative destination. BLT
will not cause a branch if the result .of the previous operation was zero
(without overflow).

Branch on Greater than or Equal (Zero) BGE lot 1.5 rrs.2.6~~

offset
0 1 0
I I 2, IO I I I I I I
15 8 7 0

Operation: lot + (PC) if N V V = 0

Description: Causes a branch if N and V are either both clear or both set.
BGE is .the complementary operation to BLT. Thus BGE will always cause
a branch when it follows an operation that caused addition to two positive
numbers. BGE will also cause a branch on a zero result.

24
 Branch on Less than or Equal(Z*ro) BLE lot 1.5~~. 2.6~

offset
01 to, I ‘31 14 I t I I I
i5 07 0

Operation: lot + (PC) if Z v (N V V) = 1

Description: Operation of BLE is similar to that of BLT but in addition will

cause a branch if the result of the previous operation was zero.

Branch on Greater Than (Zero) BGT lot 1.51u,2.6u~

0, 0 I I 3 I II 0 11 1
offset ’ 1 ”
15 8 7 0

Operation: lot -+ (PC) if Z v (N tt V) = 0

Description: Operation of BGT is’ similar to BGE, except that BGT will not
cause a branch on a zero result.
Unsigned Conditional Branches-The Unsigned Conditional Branches pro-
vide a means of testing the result of comparison operations in which the
operands are considered as unsigned values.
Branch on Hlgher BHI l.Sur,2.6ur

offset
1 , , 0
I, 1 I I ‘0
11.1 I I I I I
‘t 5 8 7 0

Gperation: lot + (PC) if both C and Z = 0

Description: Causes a branch if the previous operation caused neither a carry
nor a zero result. This will happen in comparison (CMP) operations as long
as the source has a higher unsigned value than the destination.

Branchon Lower of same BLOS lot 1.5~. 2.61~s

offset
I,0 f ,
1 I' I I I I I 1 1
(5 8 7 0

Operation: lot + (PC) if C v Z = 1

Description: Causes a branch if the previous operation caused either a carry
or a zero result. BLOS is the complementary operation to BHI. The branch ’
will occur in comparison operations as long as the source is equal to, or has
a lower unsigned value than, the destination.
Comparison of unsigned values with the CMP instruction can be tested for
“higher or same” and “higher” by a simple test of the C-bit. For convenience,
the mnemonics BHIS (Branch on Higher. or Same) and BLOS (Branch on
Lower Or Same) have been defined such that BHIS = BCC and BLO = BCS.

Bmnch on Higher Q Same BHIS IQC - 1.5~. 2.6~

t 1 offset
0 3 0
I, Ii I I I, I I , I
15 -8 7 0

Operation: lot + (PC) if C = 0

Description: BHIS is the same instruction as BCC

25
 BLO IOC t.5 YS, 2.6~s
Branch on LOwn
Offset
t 1 0, 1 , 3 , 1 4 I I I I I 1.1
95 e 7 0

Operation: lot + (PC) if C = 1

Description: BLO is the same instruction as BCS
q The following example illustrates the use of some of the instructions and
addressing modes described thus far. Two new instructions are used: INC
(INCrement) and ASL (Arithmetic Shift Left) which respectively, add 1 (INC)
and multiply an operand by 2 (ASL). Their operation is fully described later
in this chapter.
This example demonstrates the generation of a table (histogram) that shows
the frequency of occurrence of each value in another table (within a range
of values l-100). Histogram generation (including initialization) requires
22 words. Values outside the range l-1OB are ignored.
HIST: MOV #OTABLE, RO ;set up to clear output table
MOV #-loo., Rl ;lOO entries in output table
CLOOP: ‘yg p+ ;clear next entry
;check if done
BNE CLOOP ;if not, continue clearing
MOV #ITABLE, RO ;set up input pointer
MOV # -lOOO., Rl ;length of table
MOV #lOO., R2 ;max input value
HLOOP: MOV (RO)+, R4 ;get next input value
BLE NOCOUNT ;ignore if less than or equal zero
CMP R4, R2 ;check against max value
BGT NOCOUNT ;ignore if greater
ASL R4 ;2 bytes per table entry
INC OTABLE (R4) ;increment proper element
NOCOUNT: INC- Rl ;input done?
BNE HLOOP ;if not, continue scanning
HALT ;histogram complete

The JUMP Instruction-JMP (JUMP) provides more flexible program branch-

ing then is provided with the branch instructions. Control may be transferred
to any location in memory (no range limitation) and can be accomplished
with the full flexibility of the PDP-11 addressing modes.

JUMP JMP dst 1.2US

dst
1
01 ‘0
I I 0
I I I I I I I I t 1
15 6 5 0

Operation: dst + (PC)

Conditioned Codes: not affected
Description: Register mode is illegal in JMP instructions and will cause an
“illegal instruction” condition. (Program control cannot be transferred to a
register.) Register deferred mode is legal and will cause program control to
be transferred to the address held in the specified register. Note that instruc-
tions are word data and must therefore be fetched from an even-numbered

26
 address. A “boundary error” condition will result when the processor at-
tempts to fetch an instruction from an odd address.
Deferred index mode JMP instructions permit transfer of control to the
address contained in a selectable element of a. table of dispatch vectors.
SUBROUTINES-The subroutine call in the PDP-II provides for automatic
nesting of subroutines, regntrancy, and multiple entry points. Subroutines
may call other subroutines (or indeed themselves) to any level of nesting
without making special provision for storage of return addresses at each
level of subroutine call. The subroutine calling mechanism modifies no fixed
location in memory and thus also provides for reentrancy. This allows one
copy of a subroutine to be shared among several interrupting processes.

-Q dsi
01 0 I I 4 I I, I I, I I,,
15 9 6 6 3 0

Operation: dst + (tmp) (tmp.is an internal processor register)

Oes) & (push reg contents onto processor stack)
(W -+ (rea) (PC holds location following JSR; this address
_ (tmp) + Up now put in reg)
Condition Codes: not affected
Description: Execution time for JSR is the sum of instruction and destination
times. In execution of the JSR, the old’contents of the specified-register,
(the “linkage pointer”), are automatically pushed onto the ‘processor stack
and new linkage information placed in the register. Thus subroutines nested
within subroutines to any depth may all be called with the same linkage
re&ter. There is no need either to plan the maximum depth at which any
particular subroutine will be called or to include instructions in each routine
to save and restore the linkage pointer. Further, since all linkages are saved
in a reentrant manner-on the processor stack-execution of a subroutine
may be interrupted, the same subroutine reentered and executed by an
interrupt service routine. Execution of the initial subroutine can then be
resumed when other requests are satisfied. This process (called nesting) can
proceed to any level.
A subroutine called with a JSR reg, dst instruction can access the arugments
following the call with either autoincrement addressing, (reg) +, (if argu-
ments are accessed sequentially) or by indexed addressing, X(reg), (if ac-
cessed in random order). These addressing modes may also be deferred,
@(reg)+ and @X(reg) if the parameters are operand addresses rather than
the operands themselves.
JSR PC, dst is a special case of the PDP-11 subroutine call suitable for
subroutine calls that transmit parameters through the general registers. No
register except the program counter is modified by this call.
Another special case of the JSR instruction is JSR PC, (SP)+ which ex-
changes the top element of the processor stack and the contents of the
program counter. Use of this instruction allows two routines to swap pro-
- gram control and resume operation when recalled where they left off. Such
-routines are called “co-routines.”
Return from a subroutine is done by the RTS instruction: RTS reg loads the
&tents of the reg into the PC and pops the top element of the processor
stack into the specified register.

27
 ReTurn from Sutwoutine 3.5&S

I
15 3 2 0

Operation: (reg) + (PC)

T (ret3
Condition Codes: not affected
Description: Loads content of reg into PC and pops the top element of the.
Processor stack into the specified register. Execution time for RTS is equal
to the basic instruction time.
Return from a subroutine is typically made through the same register that
was used in its call. Thus, a subroutine called with a JSR PC, dst exists
with a RTS PC and a subroutine called with a JSR R5, dst, picks up param-
eters with addressing modes (R5)+, X(R5);or @X(R5) and finally exists
with a RTS R5.

Programming Examples of the Use of Subroutines-

1. Passing arguments in subroutine calls-The subroutine TOLER
checks each element in an array of unsigned integers to determine
whether any elements are outside specified limits. If all are within
tolerance, the value 0 is returned in the register RO. If TOLER find
an element out of tolerance, it returns the address of the bad
element + 2 in RO. The calling sequence for TOLER is:
JSR R5, TOLER .
. WORD ARRAY ;address of array ts be
‘I
:checked (*WORD expres-
ision-defines a word equal
;to the value of the expres-
;sion)
. WORD -LENGTH ;minus # of items in array
/
. WORD HILIM ;upper limit of tolerance
. WORD LOLIM ;lower limit of tolerance
;subroutine returns here
;Tolerance Check-Array Elements Within Limits?
TOLER: , MOV (R5)+, RO ;get array address
MOV (R5)+, Rl ;get minus the length
MOV (R5)+, R2 :get high tolerance limit
MOV (R5)+, R3 ;get low tolerance limit
TLOOP: MOV (RO)+; R4 ;get next element of array
CMP R4, R2 ;check it against high limit
BHI TEXIT ;leave routine if higher
CMP R4, R3 ;check it against low limit
BLO TEXIT ;leave routine if lower
INC Rl ;increment count, check
;whether at end of array
; ;contihue if not at end yet
BNE TLOOP
CLR RO ;exit with RO = 0 if all ok
TEXIT: RTS R5 ;return, RO holds poirter
:or 0

28
The instruction INC Rl increases the contents of Rl by 1 and the instruction
CLR RO zeroes the register RO

2. Saving and restoring registers on the stack-This subroutine pushes

RO-R5 onto the stack. It is called by:

JSR R5, SAVE

SAVE: MOV R4, -(SP) ;R5 was pushed by the JSR
MOV R3, -(SP) ;R5 will be at the bottom
;of the stack
\ MOV R2, -(SP) ;R4, R3, R2, Rl, and RO
;in order
MOV Rl, -(SP) ;will be above it
MOV RO, -(SP) ;RO is at the top of the
;stack
JMP R5 ;R5 holds ‘the return ad-
;dress

The TST operation is equivalent to comparing the operand with 0, i.e.,

TST opr = CMP opr, #0
The only effect is to set the appropriate condition codes.

The following example illustrates a subroutine to restore RO-R5 from the

stack.

REST: TST (SP) + ;this increments the SP by 2

MOV (SP)+, RO ;the registers are restored
MOV (SP)+, Rl :in reverse order to that in
MOV (SP)+, R2 which
MOV (SP)+, R3 ;they were put on the stack
MOV (SP)+, R4 ;R5 is loaded into the PC
RTS R5 and the old R5 restored

The operation TST (SP)+ removes the top element on the stack. At the time
it is used, the top element holds the contents of R5 that were saved by the
call to REST. Since R5 is to be loaded with the value saved on the stack
by SAVE, this information is not needed.

3. Stacks, recursion, and nesting -The following subroutine converts

an unsigned binary integer to a string of typed ASCII characters. In
the routine, the remainders of successive divisions by 10 are saved
and then typed in reverse order.

The operation of the subroutine is to call a part of itself (begin-

ning with DECREM) repeatedly until a zero quotient is calculated
by an integer divide subroutine, IDIVR. At each iteration, the dividend
is divided by 10, the resulting quotient replaces the dividend, and
the remainder is pushed onto the processor stack. The processor
stack thus holds interleaved data (remainders) and control informa-
tion (return addresses from calls to DECPNT and DECREM) when
the quotient finally comes up as 0 and the branch is made to
DECTIY. The portion of the routine beginning at DEClTY then pops
a remainder from the stack, converts it to an ASCII character, types
it and then returns control to DECTTY (with RTS PC) until the stack
is reduced finally to its state immediately after the call to DECPNT.

29
 At this point execution of RTS PC returns control to the main
program.

A character is typed in DECTY by loading the teleprinter buffer

(TPB) and waiting for the teleprinter READY flag, the most significant
bit of the low-order byte of the teleprinter status word (TPS),
to be set.

The symbols CR and LF are assumed equal to the ASCII repre-

sentations for carriage return and line feed respectively.

This’subroutine types the unsigned integer in RO. It illustrates recursion and

the use of stacks.

, DECPNT: MOV #lo., R2 ;set up divisor of 10

DECREM: JSR PC, IDIVR ;subroutine divides (RO) by
;W’)
MOV Rl, -(SP) ;quotient is in RO, remain-
;der is in Rl
TST RO ;after pushing remainder
;onto stack test quotient
BEQ DECTTY ;if the quotient is 0, we‘re
;done getting remainders
JSR PC, DECREM ;if not try again
DECTTY: MOV (SP)+, RO ;get next remainder
ADD #60, RO ;make an ASCII character
l-rYOUT: MOV RO, TPB ;type the ASCII character in
;RO
l-rYLUp: TST TPS ’ ;wait for the teleprinter to
;be done
BPL ll-YLlJP ;TPS is negative when the
;TP is done
CMP. #CR, RO ;was the character of a car-
;riage return
BEQ TI-YLF ;if not: return, if so; get a
;line feed
RTS PC ;returns either to DEClTY
;or main program
T-IYLF: MOV ‘#LF, TPB ;type a line feed
BR TTYLUP ;and wait for it to be corn-
;pleted
4. Multiple entry points-In the example that follows, the subroutines
described above are used to type out all the entries in a table of
unsigned integers. that are not within specified tolerance.

The subroutine TOLER is entered at TOLER for initialization and at

TLOOP to pick up each bad entry of the array after the first one.

The subroutine DECPNT is entered at DECPNT to print the value of

’ the unsigned binary number held in RO and at TlYOUT to print the
ASCII character held in RO. TTYOUT prints the carriage return, line
feed sequence when it sees the carriage return character.

This routine types all out-of-tolerance elements of an integer array.

The program starts at TYPOUT.

30
 TYPFIN: HALT ;suspend processor opera-
;tion, wait for key continue
TYPOUT: JSR R5, TOLER ;get address of bad item:
- ;initialization entry
. WORD ARRAY ;address of array
. WORD -LENGTH ;-length of array
. WORD HILIM ;high limit
. WORD LOLIM ;low limit
TYPCHK: BEQ TYPFIN ;Z-bit is set if no more out
;of limits
JSR R5, SAVE ;an element is out of limits,
;save registers
MOV -(RO), RO ;RO holds address + 2, get
;operand into RO
JSR PC, DECPNT Tprint out number
MOV #CR, RO ;type CR, LF
JSR PC, TTYOUT mote use of second entry
;point
JSR R5, REST ;restore registers
JSR R5, TLOOP ;continue searching array,
;alternate entry
BR TYPCHK ;another bad element?
SINGLE OPERAND INSTRUCTIONS-Single Operand Instructions are repre-
sented as:
OPeRotion OPR dst Instruction Time
operation code dst
I I’ I I I I t ,I I I I 1 I I I I I
15 6 5’ 0

The execution time for single operand instructions is the sum of the basic
instruction time and destination a$dress time for the operation.
General Operations-
CLMR CLR dd 2.3~

0 1 0 I 51 I 101 dd
I I I I 1
15 6 5 0

Operation: 0 + (dst)
Condition Codes: Z: set
N: cleared
C: cleared
V: cleared
Description: Zeroes the specified destination.
INCrement = _ Hcdst 2.3~

drt --
0 1 0
I I
5
II I
2.
I I l t I 11
15 6 5 0

Operation: (dst) + 1 + (dst)

Condition Codes: Z: set if the result is 0; cleared otherwise
N: set if t&e result is < 0; cleared otherwise
C: not affected
V: set if (dst) held 077777; cleared otherwise
Description: Adds 1 to the contents of the destination.

31
 DECrment DEC drt 2.3~

01 10, I 15‘ I 13, I I I I I

15 6 5 ‘0

Operation: (dst) - 1 + (dst)

Condition Codes 2: set if the result is 0; cleared otherwise
N:-set if the result is < 0;cleared otherwise
V: not affected
C: set if (dst) was 100000; cleared otherwise
Description: Subtracts 1 from the contents of the destination.
NEGate NEG dst 2.3~

drt
4
0 1 0
I I 5
t I I I I I I I I I
15 6 5 0

Operation: - (dst) + (dst)

Condition Codes: as in SUB dst, #0
z: set if the result is 0; cleared otherwise
N: set if the result is < 0; cleared otherwise
c: cleared if the result is 0; set otherwise
v: set if the result is 100000; cleared otherwise
Description: Replaces the contents of the destination address by their two’s
complement. (However, 100000. is replaced by itself-in two’s complement
notation the most negative number has no positive counterpart.)
To-ST TST dst 2.3~s *

dst
0 1 0 5 7
I t 1. I , I I I’1 I I I
15 6 5 0

Operation: 0 - (dst)
Condition Codes: as in CMP #0, dst
Z: set if the result is 0; cleared otherwise
N: set if the result is < 0; cleared otherwise
C: cleared
V: cleared
Description: Sets the condition codes Z and N according to the contents of
the destination address.
COMplement COM dst 2.31~ *

dst
01 IO, I 15, I I’, t I I I I
15 6 5 0

Operation: - (dst) + (dst)

Condition Codes: Z: set if result is 0; cleared otherwise
N: set if most significant bit. of result set; cleared other-
wise
C: set
V: cleared
Description: Replaces:. the co.nten)s. of tl$’ destination address by their
logical complement (each: bit .equal- to. 0 is .set.and each bit e&al to1 is
cleared).
l See the note for the CMP instruction.

32
Multiple Precision Operations-It is sometimes convenient
to d6 arithmetic
on operands considered as multiple words.‘The PDP-11 makes special pro-
vision for such operations with the instructions ADC (ADd Carry) and SBC
(SuBtract Carry).
ADd carry ADC dst 2.3YS

dst
5 5 I I
01 IO, 1 I I I I 1 1 I
15 6 5 0

Operation: (dst) + (F) + (dst)

Condition Cobes: Z: set if result = 0; cleared otherwise
N: set if result < 0; cleared otherwise
c: set if (dst) was 177777 and (C) was 1; cleared other-
wise
V: set if (dst) was 077777 and (C) was 1; cleared other-
wise.
Description: Adds the contents of the C-bit into the destination. This permits
the carry from the addition of the two low-order words to be carried into the
high-order result.
Double precision addition may be done with the following instruction se-
quence:
ADD AO, BO ; add low-order parts
ADC Bl, ; add carry into high-order
ADD Al,Bl ; add high-order parts
suatract Carry SEC dst 2.3~

dst
0 1 0 ,.. 5 6
I I I I I I I I J
t5 6 5 0

Operation: (dst) - (C) -, (dst)

Condition Codes: Z: set if the result 0;‘cleared otherwise
N: set if the result < 0; cleared otherwise
C: cleared if the result is 0 and C = 1; set otherwise
V: set if the result is 100000; cleared otherwise
Description: Subtracts the contents of the C-bit from the-destination. This
permits the carry from the subtraction of two low-order wdtds to be sub-
tracted from the high-order part of the result.
Double precision subtraction is done by:-
SUB AO, BO
SBC Bl
SUB Al, Bl
Double precision negation is accomplisheg bith:
NEG BO ;negate low-order part; sets C unless BO = 0
SBC Bl ;makes “NEG Bl” = “COMB Bl” unless BO = 0
l
NEG Bl ;negate high-order part

Rotates-Testing of sequential bits of’s word and detailed bit manipulation

are aided with rotate operations. The instructions ROR (RDtate Right) and
ROL (ROtate Left) cause the C-bit of the status register to be effectively
appended to the destination operand in circular bit shifting.

33
 ROR dst 2.3~
ROtaA Right
dst
01 lOI I I 6 I .I 0 , I I I I 1
15 6 5 0

Condition Codes: Z: set if all bits of result = 0; cleared otherwise.

N: set if the high-order bit of the result is set; cleared
otherwise
C: loaded with the low-order bit of the destination
V: loaded with the Exclusive OR of the N-bit and C-bit
(as set by the completion of the rotate operation).
Description: Rotates all bits of the destination right one place. Bit 0 is loaded
into the C-bit of the status word and the previous contents of the’C-bit are
loaded into bit 15 of the destination. ’
ROtato Left ROL dst 2.3~

I 1I
dst
0 1 0
I 16, I I 1 I I 1 I 1
15 6 5 0

Condition Codes: Z: set if all bits of the result word = 0; cleared other-
wise
N: set if the highlorder bit of the result word is set;
cleared otherwise
C: loaded with the high-order bit of the destination
V: loaded with the Exclusive OR of the N-bit and C-bit
(as set by the completion of the rotate operation)
Description: Rotates all bits of the destination left one place. Bit 15 is loaded
into the C-bit of the status word and the previous contents of the C-bit are
loaded into bit 0 of the destination.
SWAO Bytes SWAB dst 2.3~

dst
01 10, I 10, .I 131 I I I I
15 6 5 0

Condition Codes: Z: set if low-order byte of result = 0; cleared otherwise

N: set if high-order bit of low-order byte (bit 7) of result
is set: cleared otherwise
C: cleared
V: cleared
Description: Exchanges high-order byte and low-order byte of the destination
word (dst must be a word address).
Shifts-Scaling data by factors of 2 is accomplished by the shift instructions:
ASR-Arithmetic Shift Right
ASL-Arithmetic Shift Left
The sign bit (bit 15) of the operand is replicated in shifts to the right. The
low-order bit is filled with 0 in shifts to the left. Bits shifted out of the C-bit
are lost.
/
34
 - Arithmetic Shift Right ASR dst 2.3 us

dst
01 , 0, 1 , 6, 1 , 2 , I t I e I
15 6 5 0

Condition Codes: Z: set if the result = 0; cleared otherwise

N: set if the high-order bit of the result is set; cleared
otherwise
C: loaded from the low-order bit of the destination
V: loaded from the Exclusive OR of the N-bit and C-bit
.(as set by the completion,of the shift operation)
. Description: Shifts all bits of the destination right one ‘place. Bit 15 is rePli-
cated. The C-bit is loaded from bit 0 of the destination. ASR performs signed
division of fhe destination by 2.
Arithmetic Shift Left ASL dst 2.31~

dst
01 , 0 , 1 , 6 , 1 , 3, 1 I I I I
(5 6 5 0

Condition Codes: Z: set if the result = 0; cleared otherwise

N: set if the high-order bit of the result is set; cleared
otherwise
C: loaded with the high-order bit of the destination
V: loaded with the Exclusive OR of the N-bit and C-bit
(as set by the completion of the shift operation)

Description: Shifts all bits of the destination left one place. Bit 0 is loaded
with a 0. The C-bit of the status word is loaded from the most significant bit
of the destination. ASL performs a signed multiplication of the destination
by 2.

Multiple precision shifting is done with a sequence of shifts and rotates.

Double Precision Right Shift:

ASR Al; low-order bit of Al to C-bit
ROR AO; C-bit to high-order bit of.AO

Double Precision Left Shift:

ASL AO;. high-order bit of A0 to C-bit
ROL Al; C-bit to low-order bit of Al

Normalization of operands (scaling of the operand until the operand taken

as a 15-bit fraction with sign is in the range - l/e < operand < l/e) pro-
ceeds as follows:

NORM: ASL A ; shift O’s into low-order bit

BEQ NFIN ; if the result is 0, the operation is
; complete
BVC NORM ’ ; if the sign did not change, continue
ROR A ;restore the sign
BR NDONE ; normalization complete
NFtNy.3 i ’ ROR ‘A ” : &store the sign: 000000 or 100000
: 1 “‘ 1 ? ,ASR A, f atiiYmplicate R?OOOOOO or 140000
\
NDONE: . . .
Double precision norma’lization proceeds similarly: ,

DNORM: ASL A0 ; double precision left shift

ROL Al
BEQ DZERO ; high order result O?. if so, check.10~
BVC DNORM ; if the sign did not change, continue
ROR Al ; restore the sign
BR \ DNDONE ; normalization complete
DZERO: TST A0 ; low order zero, too?
BNE DN0RM ; if not, continue normalization
ROR Al ; restore the sign; 000000 or 100000
ASR Al ; and replicate it; 000000 or 140000
DNDONE: :. .
The following example illustrates the use of shifts and rotates in a 16-bit un-
signed integer multiply subroutine. Access of operands through address
parameters following the subroutine is also shown. The multiplication takes
115170 &s in in-line code. The entire subroutine as shown below takes
approximately 200 ps and requires 16 words. The calling sequence is JSR
R5, MULT. I

. WORD MCAND ; address of multiplicand ,

. WORD MPLIER ; address of multiplier
. WORD PROD ;.address of product
MULT: CLR RO
MOV @ (R5) +, Rl ; get multiplier into Rl
MOV @ (R5) +, R2 ; get multiplicahd into R2
MOV #-169 R3 ; set counter
M LOOP: ASL RO ; double prec shift
ROL Rl ; shift and add multiply
BCC NOADD ; most significant bit governs add
ADD R2, RO ; if set add in multiplicand
ADC Rl ; keep 32-bit product
NOADD: INC R3 ; done?
BNE MLOOP ; if not continue
MOV (R5) +, R2 ; get address to store prod.
MOV RO, (R2) + ; put low-order away, move to high
MOV Rl, (R2) ; put high-order away
RTS R5 ; return to calling program
Bni OPERATIONS--The PDP-11 processor includes a full complement of
instructions that manipulate byte operands. Addressing is byte-oriented so
that instructions for byte manipulation are straightforward. In addition, byte
instructions with autoincrement or autodecrement direct addressing cause
the specified register to be stepped by one to point to the next byte of data.
Byte operations in register mode access the low-order byte of the specified
register. These provisions enable the PDP-11 to perform as either a word or
byte processor.
Timing of byte instructions is the same as for word instructions except that
an additional 0.6 ps is required for access of bytes at odd addresses.
Double Operand Byte Instructions-

Move !sytr MOVB WC, dst 2.3~

WC dst
1 , t
I I I I I I I’, I I I I I
. 15 t2 11 6 5 0

-36
Operation: (src) + (dst)
Condition Codes: Set on the byte result as in MOV

Description: Same as MOV instruction. The MOVB instruction in register mode

(unique among byte operations) extends the most significant bit of the byte
register (sign extension). Otherwise MOVB operates on bytes exactly as MOV
operates on words.

Coware Byte CMPS src,dst 2.3us*

src dst
‘I I 12 -I 1 I I 1 1 I I , I
t5 12 H 6 5 0

Operation: (src) - (dst) ; in detail (src) + - (dst) + 1

Condition Codes: Set on the byte result as in CMP

Description: Same as CMP instruction.

811Set Byte BISB src,dst 2.3~

SW dst
t1 ,s,- I I I I
I I I I I
I5 12 11 6 5 0

Operation: (src) V (dst) + (dst)

Condition Codes: Set on the byte result as in BIS

Description: Same as ,BIS.
,
Bll CIm3rByte BICB crc .dat 2.31~
SK dst
114, , I I I I I I I I
t5 12 H 6 5 0

Operation: - (src) (dst) i, (dst)

Condition Codes: set on the byte result as in BIC
Description: Same as BIC.

Bit Tart Byte BIT6 m,drt 2.31~~

41 , 3, I I UCI 1 I drt
I I I I
t5 12 11 6 5 0

Operation: (src) (dst)

Condition Codes: Set on the byte result as in BIT

Description: Same as BIT.

The following subroutine scans a packed character string of variable length
lines, removes blanks and unpacks the string to left-justified length lines.
INSTRING is the address of the INput STRING, OUTSTRING is the address
of,the OUTput String. EOLCHAR. SPCHAR.‘and EORCHAR are the End Of
Line CHARacter, Space CHARacter, and End of Record CHARacter respec-
tively.
* See the note for the CMP instruction.

37 \
LNLINE is the Length of uNpacked LINES. The routine requires 24 words.

EDIT: MOV # INSTRING, RO ; set up input byte pointer

MOV #OUTSTRING, Rl ; set up output byte pointer
MOV #EOLCHAR, R2 ; put high use constant in reg.
MOV #SPCHAR, R3 ; to save time in loop
NOLINE: MOV #LNLINE, R4 ; R4 holds # char left in line
NXTCHR: MOVB (RO) + ,R5 ; get next byte
CMP R5, R2 ; end of line?
BEQ FILINE ; if yes, fill line
CMP R5, R3 ; blank?
BEQ NXTCHR ; if yes, skip character
DEC R4 ; decrement # of characters left in line
MOVB R5, (Rl) + ; move byte to output string
BR NXTCHR ; continue
FILINE: CLRB (Rl) + ; put a blank byte in output
DEC R4 ; decrement # char left
BNE FILINE ; continue if not end
CHKEND: CMPB (RO), #EORCHAR ; end of record?
BNE NULINE ; if not EOR, start next line

Single Operand Byte Instructions-

CLeaR Byte CLRB dst _ 2.3~

dst
‘I lOI I 15, I rot I I I I I
t5 6 5 0

Operation: 0 + (dst)
Condition Codes: Set on the byte result as in CLR
Description: Same as CLR

INCrement Byte INCB dst 2.3~

dst
‘I 101 I 15, I 121 I I I # I
15 6 5 0

Operation: (dst) + 1 + (dst)

Condition Codes: Set on the byte result as in INC
Description: Same as INC. The carry from a byte does not affect any other
byte.
OECrcmcnt Byte OECBdst 2.3~

dst
1 1 oi 1 5 3
I I I I 1 I I
15 6 5 0

bperation: (dst) - 1 + (dst)

Condition Codes: Set on the byte result as in DEC

Description: Same as DEC.

38
 ~.
NEGote Byte NEGB -dst 23ns

dst
‘I 101 I Ii, I 141 I I I t
I5 6 5 0
-. l

Operation: -(dst) + (dst) ; in detail, - (dst) f. 1 + (dst)

Condition Codes: Set on the byte result as NEG

Description: Same as NEG.

T&T Byte TSTB dst 2.3~(1*

dst
‘I lOI 5, I 17, I I I I
15 6 5 0

Operation: 0 - (dst)
Condition Codes: Set on the byte result as TST

Description: Same as TST.

cohtp*m(Hlt Byte COMB dst 2.3~

dst
‘I 101 I 151% I I’, I I I I I

15 6 5 0

Operation: - (dst) + (dst)

Condition Codes: Set on the byte result as COM
Description: Same as COM.
AOd Carry Byte ADCB dst 2.3~

dst
‘I lOI I 15, I I 5 ,
I I I I
15 6 5 0
‘\ Operation: (dst) + (C) + (dst)

Condition Codes: Set on the byte result as ADC

Description: Same as ADC.
SuBtract Carry Byte iBC0 dst 2.3~~

I , 0 dst
I 15, I I 6 I I I , I , ,
15 6 5 0

Operation: (dst) - (C) + (dst)

Condition Codes: Set on the byte result as SBC
Description: Same as SBC.
ROtate R’ght Byte RORB dst 2.3u+
I
dst
‘I 10, I 16, I ,“I I I I 1 n
15 - 6 5 0

Operation: as in ROR.on byte operands

Condition Codes: Set on the byte result as ROR
Description: Same as ROR
l See the note for the CMP instruction.

39
 . ,
mtate Left Byte ROLB drt ’ ~ 2.3~1 Q

1
15
, 0 I I t 1 I I .I 1 1 1 ’ ’ 1
6 3
6 5
dst ,

Operation: as in ROL on byte operands

Condition Codes: set on the byte results as ROL
Description: Same as ROL
Arithmetic Shift Riaht Bvte ASRB dst 2.3~7

1 1 0
I 16’1 I I
2
I
I

15 6 5 0

Operation: as in ASR on byte operands

ConditionCodes: set on the byte result as ASR
Description: Same as ASR
Arithmetic Shift Left Byte ASLB dst 2.3u**
dst
‘I 101 I I ‘61 I 131 I I I I I
t5 6 5 0

Operation: as in ASL on byte operands

Condition Codes: set on the byte results as ASL
Description: Same as ASL

CONDITION CODE OPERATORS--Condition code operators set and clear con-

dition code bits. Selectable combinations of these bits may be cleared or set .
together in one instruction.

Condition Code Qpemtors

15 5 4 3 2 1 0

Condition code bits corresponding to bits in the condition code operator

(bits 3-O; N, Z, V, C) are modified according to the sense of bit 4, the set/
clear bit of the operator. The following mnemonics are permanent symbols
in the assembler:
Mnemonic Operation Op Code Mnemonic Qperetion Op Code
CLC Clear C 000241 SEC Set C 000261
CLV Clear V 000242 SEV Set V 000262
Clear Z oo(2244 SEZ Set Z 000264
K Clear N 000250 SEN Set N 000270
Timing for all condition code operators is the basic instruction time (1.5~s)
for the operators. (The codes 000240 and 000260 are the shortest “no+opera-
tion” instructions.)
T Shift and rotate operations r,equire an additional 0.6 we to etceee bytes at odd
addresses.

40
 Combinations of the above set or clear operations may be ORed together to
form new instruction mnemonics. For example: CLCV = CLC ! CLV. The new
instruction clears C and V bits. (‘I!” signifies “inclusive or” in PAL-11.)

MISCELLANEOUS CONTROL iNSTRUCTIONS

RESet ExTernol bus RESET 20 ms

01 IO, I 101 I 101 I lOI I I 5 ,

15 0

Condition Codes: not affected

Description: Sends an INIT pulse along the Unibus by the processor. AlI
devices on the bus are reset to their state at power-up.
\
WAit for IntempT WAIT 1.8 IL*

15 0

Condition Codes: not affected

Description: Provides a way for the processor to relinquish use of the bus
while it waits for an external interrupt. Having been given a WAIT command,
the processor will not compete for bus use by fetching instructions or
operands from memory. This permits higher transfer rates between a device
and memory, since no processor-induced latencies will be encountered by
bus requests from the device. In WAIT, as in all instructions, the PC points
to the next instruction following the WAIT operation.
Thus when an interrupt causes the PC and PS to be pushed onto the proces-
sor stack, the address of the next instruction following the WAIT is saved.
The exit from the interrupt routine (i.e. execution of an RTI instruction) will
cause resumption of the interrupted process at the instruction following the
WAIT.

HALT HALT l.SlLS

01 101 I IO, I ,Ol I 101 I 101

45 0

Condition Codes: not affected

Description: Causes the processor operation to cease. The console is given
control of the bus. The console data lights display the contents of RO; the
console address lights display the address of the halt instruction. Transfers
on the Unibus are terminated immediately. The PC points to the next in-
struction to be executed. Pressing the continue key on the console causes
processor operation to resume. No INIT signal is given.
Processor Traps -Processor Traps are internally generated interrupts.
Error conditions, completion of an instrustion in trace mode (i.e. T-bit of
status word set), and certain instructions cause traps. As in interrupts, the
current PC and PS are saved on the processor stack and a new PC and PS
are loaded from the appropriate trap (interrupt) vector. See Appendix C for
a summary of Trap Vector Addresses.
,Trapf Instructions-Trap Instructions provide for calls to emulators, i/O
monitors, debugging packages, and user-defined interpreters.

41
 EMulotw Traps EMT xycx 9.9 us

XXI
‘I 101 ,4, 0, , 1 , , 1 , ,

15 9 7 0

Operation: (PS) J, SP
WI J SP
(30) + PC
(3?) + PS

Condition Codes: loaded from trap vector.

.
Description: Performs a trap sequence with a trap vector address of 30.
All operation codes from 104000 to 104377 are EMT calls. The low-order
byte, bits O-7 of the EMT instructions, may be used to transmit information
to the emulating routine (e.g., function to be performed). The trap vector
for EMT is at address 30. The new PC is taken from the Word at address 30;
the new central processor status (PS) is taken from the word at address 32.

TRAP TRAP xxx 9.9JLs

.‘I 10, 4 4 xxx

I I I I I
15 9 7 0

Operation: as in E&IT except the trap vector is located at 34.

Condition Codes: loaded from trap vector.

Description: Performs a trap sequence with a trap vector address of 34.

’ Operation codes from 104400 to 104777 are TRAP instructions. TRAPS and
EMTs are identical in operation, except that the trap vector for TRAP is at
address 34.

l/O Trap IOT 9.9LS

01 IO, I 101 I 101 I 101 I 141

15 0

Operation: as EMT except the trap vector is located at address 20 and no

information is transmitted in the low byte. ,
Condition Codes: loaded from trap vector.

Description: Used to call the I/O executive routine IOX.

No defined mnemonic 000003 9.9us

“I_-,04 I 101 I 101 I IO, I I 3 I

15 0

Operation: Same as IOT except that trap vector is located at address 14.

Condition Codes: loaded from trap vector.

Description: Used to call debugging aids. The user is cautioned against

employing code 000003 in programs run under these debugging aids.

42
 R.Turn from Interrupt RTI 4.Bus

/ 01 ,o, I IO, I 101 I 101 I a21

IS 0

Operation: SP T (PC), SP t (PS).

Condition Codes: loaded from processor stack.
Description: Used to exit from an interrupt or TRAP service routine. The PC
and PS are restored (popped) from the processor stack.
Instruction traps are also caused by attempts to execute instruction codes
reserved for future processor expansion (reserved instructions) ‘Or instruc-
tions with illegal addressing modes (illegal instructions). Order codes not
corresponding to any of the instructions described above are considered to
be reserved instructions. Illegal instructions are JMP and JSR with register
mode destinations. Reserved and illegal instruction traps occur as described
under EMT, but trap through vectors at addresses 10 and 04 respectively.
Stack Overflow Trap-Stack.Overflow Trap is a processor trap through the ,
vector at address 4. It is caused by referencing addresses below 400, through
the processor stack pointer R6 (SP) in autodecrement or autodecrement de-
ferred addressing. The instruction causing the overflow is completed before
the trap is made.
Bus Error Traps-Bus Error Traps are:
1. Boundary Errors-attempts to reference word operands at odd ad-
dresses.
2. Time-Out Errors-attempts to reference addresses on the bus that
made no response within 10 ps. In general, these are caused by at-
tempts to reference nonexistent memory, and attempts to referf?nce
nonexistent peripheral devices.
Bus error traps cause processor traps through the trap vector address 4.
Trace Trap-Trace Trap enables bit 4 of the PS word and causes processor -
traps at the end of instruction executions. The instruction that is executed
after the instruction that set the T-bit will proceed to completion and then
cause a processor trap through the trap vector at address 14.
The following are special cases and are detailed in subsequent paragraphs.
1. The traced instruction cleared the T-bit.
2. The traced instruction set the T-bit.
3. The traced instruction caused an instruction trap.
4. The traced instruction caused a bus error trap.
5. The traced instruction caused a stack overflow trap.
6. The process was interrupted between the time the T-bit was set and
the fetching of the instruction that was to be traced.
I
7. The traced instruction was a WAIT.
8. The traced instruction was a HALT.
An instruction that cleared the T-bit-Upon fetching the traced instruction
an internal flag, the trace flag, was set. The trap will still occur at the end
of execution of this instruction. The stacked status word, however, will have
a clear T-bit.

43
An instruction that set the T-bit-Since the T-bit was already set, Settiflg it
again has no effect.
An instruction that caused an Instruction Trap--The instruction trap is
sprung and the entire routine for the service trap is executed. If the service
routine exists with an RTI or in any other way restores the stacked status
word, the T-bit is set again, the instruction following the traced instruction
is executed and, unless it ,is one of the special cases noted above, a trace
trap occurs.

An instruction that caused a Bus Error-This is treated as in an Instruction

Trap. The only difference is that the error service is not as likely to exit
with an RTI, so that the trace trap may not occur.
An instruction that caused a stack overflow-The instruction COmpleteS
execution as usual-the Stack Overflow does not cause a trap. The Trace
Trap Vector is loaded into the PC and PS, and the old PC and, PS are pushed
onto the stack. Stack Overflow occurs again, and this time the trap is made.,
An interrupt between setting of the T-bit and fetch of the traced instruction
The entire interrupt service routine is executed and then the T-bit is set again
by the exiting RTI. The traced instruction is executed (if there have been no
other interrupts) and, unless it is a special case noted above, causes a trace
trap.
Note’that no interrupts are acknowledged between the time of fetching any
trapped instruction (including one that is trapped by reason of the T-bit being
set) and completing execution of the first instruction of the trap service.
A WAIT-The trap occurred immediately. The address of the next instruction
is saved on the stack. .
A HALT-The processor halts. When the continue. key on the console is
pressed, the instruction following,the HALT is fetched and executed. Unless
it is one of the exceptions noted above, the trap occurs immediately follow-
ing execution.
Trap priorities-In case multiple processor trap conditions occur simultane-
ously the following order of priorities is observed (from high to low):
1. Bus Errors
2. instruction Traps
3. Trace Trap
4. Stack Overflow Trap
The details on the trace trap process have been described in the trace trap
operational description which includes cases in which an instruction being
traced causes a bus error, instruction trap, or a stack overflow trap,
If a bus error is caused by the.trap process handling instruction traps, trace
traps, stack overflow traps, or a previous bus error, the processor is halted.
If a stack overflow is caused by the trap process in handling bus errors, in-
struction traps, or trace traps, the process is completed and then the stack
overflow trap is sprung.

44
 CHAPTER 5
ADDRESS ALLOCATION
The PDP-11 provides for a very flexible addressing structure. Both 16-bit
words and 8-bit bytes can be directly addressed. Addresses are 1Bbits long
allowing for direct addressing of 32,768 words or 65,536 bytes.

ADDRESS MAP
As a result of the organization of the PDP-11, bus addresses serve several
functions. A map of possible PDP-11 bus address allocation is shown

BUS ADDRESS CONTENT

0
Processor.
Program Counter Trap Vectors
and Device
Processor Status Word Interrupt
Vectors

4608 Stack Pointer Overflow Limit

Stacks, Program and Data Storage

160bOOs
Typical
Registers for
. Programmed
Status Register and Transfer
Data Buffer Register Device

.
Device Address Register Typical
Word Count Register Registers
Memory Address Register for a
Control and Status Registers Block
Transfer
* Device

1777778
Figure 5-1
Simplified Address Allocation Map

45
in Figure 5-1. Three areas of addresses of particular interest to the Pro-
grammers are: 1) Interrupt and Trap VeMors; 2) Processor Stack and General
Storage; and 3) Peripheral Device Registers.
INTERRUPT AND TRAP VECTORS-Addresses between lOCatiOn zero and
location 4001 are generally reserved for interrupt and trap vectors.
PROCESSOR STACK AND GENERAL STORAGE-Addresses between 4001
and the limit of implemented core are available for the processor stack or
other programs and data. The highest address in ttiis region is 157777*.
PERIPHERAL DEVICE REGISTERS-Addresses above 160000, generally are
reserved for peripheral device status, control, and data registers. The general
registers and the processor status can be addressed from the program
console using addresses in this area.
A more detailed address allocation map can be found in Appendix D.
CORE MEMORY
The three types of core memory that can be used in a PDP-11 system are:
1) Read-Write Core Memory: 2) Read-Only Core Memory; and 3) Wordlet
Memory. These memories can be located anywhere in address space provided
they do not overlap. They do not have to be in continuous address locations.’
MMll-E READ WRITE CORE MEMORY-The MMll-E has the following
specifications: ,
Capacity: 4,096 l&bit words or 8,192 8-bit bytes -
Cycle Time: 1.2 microseconds
Access Time: 500 nanoseconds
Configuration: Planer 3-wire, 3-D using 22 mil cores
Packaging: One standard PDP-11 System Unit
interface: Designed to work with PDP-11 bus, l-FL-compatible
MRll-A READ-ONLY CORE MEMORY (ROM)-The ROM has the following
specifications:
Capacity: 1,024 l&bit words or 2,048 8-bit bytes
Access Time: 500 nanoseconds
Configuration: P-piece core with wire braid, 256 wires, 64 cores
Packaging: 3/4 of one standard PDP-11 System Unit
Interface: Designed to work with PDP-11 bus, TTL-compatible
MWllA WORDLET MEMORY-The wordlet memory is used with ROM sys-
tems and provides read-write memory capacity for temporary data and in- ’
struction storage.
Capacity: 128 16.bit words or 256 8-bit bytes
Cycle Time: 2.0 microseconds
Access Time: 1.0 miorosecond
Configuration: 5-Wire, 3D
Packaging: l/4 standard PDP-11 single System Unit-
Interface: The wordlet memory will work with the ROM and interfaces
through the ROM System Unit to the PDP-11 bus.

46
 CHAPiER 6
PROGRAMMING OF PERIPHERALS
Programming of peripherals is extremely simple in the PDP-11-a special
class of instructions to deal with input/output operations is unnecessary.
The Unibus permits a unified addressing structure in which control, status,
and data registers for peripheral devices are directly addressed as memory
locations. Therefore all operations on these registers, such as transferring
information into or out of them or manpulating data within them, are per-
formed by normal memory reference instruction.
The ability to use all memory reference instructions on peripheral device.
registers greatly increases the flexibility of input/output programming. ln-
formation in a device register can be compared directly with a value and a
branch made on the result.
CMP #lOl, PRB
BEQ SERVICE
In this case the program looks for 101, from the paper tape reader data
buffer, and branches if it finds it. There is no need to transfer the informa-
tion into an intermediate register for comparison.

When the character is of interest, a memory reference instruction can trans-

fer the character into a user buffer in core or in another peripheral device.

MOV PRB, LOC

This instruction transfers a character from the paper tape reader buffer into
a user-defined location.
All arithmetic operations can be performed on a peripheral device register.
ADD #lo, -DSX
This instruction will add lo8 to a display’s x-deflection register.
All peripheral device registers can be treated as accumulators. There is no
need to funnel all data transfers, arithmetic operations, and comparisons
through a single or small number of accumulator registers.

DEVICE REGISTERS
All peripheral devices are specified by a set of registers which are addressed
‘as core memory and manipulated as flexibly as an accumulator. There are
two types of registers associated with each device: 1) Control and Status.Reg-
isters (CSR); and 2) Data Registers.
CONTROL AND STATUS REGISTERS (CSR)-Each peripheral has one or more
control and status registers which contain all the information necessary to
communicate with that device. The general form of a control and status
register is shown below.

General Control and Status Register

, This general form does not necessarily apply to any device, but is presented
as a format which could be used as a guideline for designing peripheral

47
devices. Many devices will require less than sixteen status bits. Other devices
will require more than sixteen bits and therefore will require additional status
and control registers.

Device Function Bits-These three bits specify operations that a device iS

to perform. An example of one operation for a paper tape read&r is read
one character. For a disk one operation would be read a block of words from
memory and store them on the disk.

Memory Extension Bits-These two bits are resewed for future expansion.
They will allow devices to ‘use a full 18 bits to specify addresses on the bus.

Done Enable and Error Enable Bits-These two bits are independently Pro-
grammable. If bit 6 is set, an interrupt will occur as a reSUlt of a function
done condition. If bit 5 is set, an interrupt will occur as the result of an
error condition. This occurs when one or more of the error bits is sat to a
one. To initiate an interrupt routine to read from the paper tape reader,
the instruction
MOV #lOl, PRS
could be used. This sets bit 0 and bit 6 of the paper tape reader control and
status register. Setting bit 0 starts the read operation and setting bit 6
enables an interrupt to occur when the read operation’is complete.

Condition Bits--The CSR contains a DONE bit, a READY bit, or a DONE-

BUSY pair of bits, depending on the device. These bits are set and cleared
by the hardware, but may be queried by the program to determine the
availability of the device. For example, the teleprinter status register (TPS)
has a READY bit (7) that is cleared on request for output and then set when
output is complete. The keyboard status register (TKS) has a DONE-BUSY
pair (Bits 7 and 11) that distinguishes between no input (DONE = BUSY
= 0), input under way (DONE = 0, BUSY = 1). and input complete (DONE
= 1, BUSY = 0).
.
The DONE bit could be used to control an input loop for reading from- the
paper tape reader as follows:
LOOP: TSTB PRS ; test low byte of paper tape status register
BPL LOOP ; branch back if DONE bit (bit 7) is not set

Unit Bits-Some peripheral systems have more than one device per control.
For example, a disk system can have multiple surfaces per control and an
analog-to-digital converter can have multiple channels. The unit bits select
the proper surface or channel.

Error Bits--Generally there. is an individual bit associated with a specific

error. When more bits are required for errors, they can be obtained by ex-
panding the error section in the word or by using another status word.

Example of Control and Status Register -The high-speed paper tape reader
control and status register (PRS) is as follows:

These bits may be read or set by instructions which use the appropriate
effective address. Bit 0 of the PRS is the function bit for reading one char-

48
 acter. Incrementing the PRS will set bit 0 and cause one character to be
read. The instruction
INC PRS
performs that function. MOV #l, PRS does the same thing but takes one
more word.

DATA BUFFER REGISTERS-Each device has at least one buffer register

for temporarily storing data to be transfer into or out of the computer. The
number and type of data registers is a function of the device. The paper
tape reader and punch use single 8-bit data buffer registers. A disk would
use I&bit data registers and some devices may use two ldbit registers for
data buffers.

PROGRAMMING EtiMPLES
PROGRAM CONTROLLED DATA TRANSFER WITH THE INTERRUPT DISABLED
-Single character I/O devices (teletype, paper tape reader/punch) have an
addressable register buffer through which data is transferred. For input, the
data buffer register is the source operand of the instruction used to get the
. data; for output, it% the destination operand. For example assuming the
paper tape reader interrupt is not enabled, character input could proceed
as follows:

MOV R, -(SP) ;
save R on the stack
MOV #8UFFER, R ;
pointer to input buffer into register R
START: INC PRS ;
start up reader
LOOP: BIT PRS, # 100200 ;
test DONE and ERROR bits
BEQ LOOP ;
branch back if none on yet
BMI ERROR ;
branch to error routine if minus
MOVB PRB, (R)+ ;
move byte from device buffer reg-
ister to user’s buffer and increment
;
;
pointer
CMP #LIMIT R, ;
check for end of buffer
BGE START ;
get next character.
MOV (SP)+, R restore R
;
-
Character output to the paper tape punch might be executed as follows:
MOV RO, L(SP) ; save RO
MOV Rl, -(SP) ; save Rl
MOV NCHAR, RO ; *number of characters into RO
MOV BUFFER, Rl ; user buffer address into Rl
LOOP: BIT PPS, #100200 ; test device ready and error bits
BEQ LOOP ; fall through if on
BMI ERROR ; branch on error
MOVB (Rl)+, PPB ; output character, increment pointer
DEC RO ; decrement character counter (and
; set condition codes)
BGT LOOP ; repeat if greater than zero
MOV (SP)+, RO ; restore RO
MOV (SP)+, Rl ; restore Rl

BLOCK TRANSFER WITH THE INTERRUPT DISABLED-High-speed block

transfer devices use the Unibus to make data transfers between the device
and core memory. These devices are provided with addressible registers
that control the flow of data.

49
A tvoical
_. set might be:
1. Control-and status register
2. Memory address register
3. Word count register
4. Device address register

Loading the device address register would in general initiate the transfer,
which then proceeds without processor intervention. The device issues non-
processor requests for the Unibus that, when granted, allow direct data
transfer between the device and memory. These requests are interleaved
with processor. requests for the bus. If very fast transfer is required, the
processor may execute a WAIT instruction after starting the block transfer.

The DONE or appropriate -error bits are set in the CSR with Completion of
the transfer or when an error occurs. These may be enabled to cause an
interrupt or may be tested to determine when the device needs assistance.
A block transfer could be executed as follows:
’ MOV #401, DKS, ; read block of data (function 1)
; from unit 1
MOV #BUFADR, DKMA ; buffer address to memory ad-
; dress register
MOV #BUFCNT, DKWC ; word count to word count register
MOV #BLKNO, DKDA ; block number to device address
; register, which starts the trans-
; fer

.
: when data is needed.
LOOP: BIT #DKMSK, DKS ; test done bit and error bits
BEQ LOOP ; branch back if none on \
BIT #DKEMSK, DKS ; test for any error bits
BNE ERROR ; branch if any on
; data is now in buffer at BUFADR

INTERRUPT STRUCTURE
If the appropriate interrupt enable bit is on, in,the control and status register
of a device, transition from 0 to 1 of the DONE or READY bit causes an
interrupt request to be issued to the processor. Also if DONE or READY’ is a _
1 when the interrupt enable is turned on, an interrupt request is made. If
the device makes the request at a priority greater than that at which the
processor is running and no other conflicts exist, the request is granted and
the interrupt sequence takes place:
a. the current program counter and processor status are pushed onto
the processor stack;
.b. the new PC and PS are loaded from a pair of locations (the interrupt
vector) in low core unique to the interrupting device.
Since each device has a unique interrupt vector which dispatches control to
the appropriate interrupt handling routine immediately, no device polling is
required.. Furthermore, since the PS contains the processor priority, the
priority at which an interrupt request is serviced can be set under program
control and is independent of the priority of the interrupt request. The

50
ReTurn fmm Interrupt instruction is used to reverse the action of the
interrupt sequence. The top two words on the stack are popped into the PC
and PS, returning control to the interrupted sequence.
PROGRAMMING EXAMPLE
A paper tape reader interrupt service could appear as follows:
First the user must initialize the service routine by specifying an address
pointer and a word count
INIF MOV #BUFADR, #0 ; set up address pointer
POINTR = . - 2 ; in third word of MOV instruction.
MOV #CNTR, #0 ; set up character count in
CRCNT=.-2 ; third word of MOV instruction.
MOV #lOl, PRS ; read a character with interrupt
_ ; enabled.
When the interrupt request occurs and is acknowledged, the processor stores
the current PC and PS on the stack. Next it picks up the interrupt vector or
new PC and PS beginning at location 70~. The next instruction executed is
the first instruction of the device service routine at PRSER.
PRSER: TST PRS test for error
BMI ERROR I branch to error routine if
bit 15 of PRS is set.
MOVB PRB, @POINTR ; move character (byte)
from reader to buffer
INC POINTR increment pointer \
DEC CRCNT decrement character count
BEQ DONE branch when input done
. INC PRS start reader for next character
DONE: RTI ; return from interrupt

51
The DIGITAL M225 module contains 8 high speed general-purpose
registers. The M225 general registers provide program flexibility
when used as accumulators, index registers, and pointers to data
words.

52
 CHAPTER 7
TELETYPE (MODEL LT33-DC/DD)
The standard Teletype Model 33 ASR (Automatic Send-Receive) can be used
to type in or print out information at a rate of up to ten characters per sec-
ond, or to read in or punch out perforated paper tape at a ten characters
per second rate. Signals transferred between the 33 ASR and the control
logic are standard serial, 11-unit code Teletype signals. The signals consist
of “marks” and “spaces” which correspond to idle and bias current in the
Teletype serial line, and to O’s and l’s in the control and computer. The
start mark and subsequent eight bits are each one unit of time duration and
are followed by the stop mark which is two units.
The 8-bit-code used by the Model 33 ASR Teletype unit is the America1
Standard Code for Information Interchange (ASCII) modified. To convert the
ASCII code to Teletype code, add 200 octal (ASCII + 200s = Teletype).
The Model 33 ASR can generate all assigned codes except 340 through
374 and 376. The Model 33 ASR can detect all characters, but does not
interpret all codes that it can generate as commands. The standard number
of characters printed per line is 72. The sequence for proceeding to the next
line is a carriage return followed by a line feed. Punched tape format is as
follows:
Tape Channel 87 654 S 321

Binary Code 10 110 100

(Punch = 1)
Octal Code 2 6 4
(S = Sprocket)
SIZE- Floor space approximately 221/” wide, 181/2” deep
Cable length 8 feet
MODEL POWER REQUIREMENTS
LT33-DC 115 v 510% 60 kO.45 Hz
LT33-DD 230 V -t-iO% 50 kO.75 Hz
TELETYPE CONTROL (MODEL KLll)
TELETYPE CONTROL-Serial information read or written by a Teletype unit
is assembled or disassembled by the control for parallel transfer on the
Unibus. The control also provides the flags which cause a priority interrupt
and indicate the availability of the teletype.
KEYBOARD/READER-The Teletype control contains an 8-bit buffer (TKB)
which assembles and holds the code ‘for the last character struck on the
keyboard or read from the tape. Teletype characters from the keyboard/
reader are received serially by the 8-bit -shift register TKB. The code of -a
Teletype character is loaded into the TKB so that “spaces” correspond to
binary O’s and holes, “marks,” correspond to binary 1’s. Upon program
command, the contents of the TKB may be transferred in parallel to a
memory location or a general register.
A character is read from the low-speed reader by setting the Teletype reader
enable bit, (RDR ENB), to a 1. This sets the busy bit (BUSY) to a 1. When a
Teletype character starts to enter, the control de-energizes a relay in the

53
 Teletype unit to release the tape feed latch. When releasea the latch
mechanism stops tape motion only when a complete character has been
sensed, and before sensing of the next character is statthd. When the charac-
ter is available in buffer (TKB), the busy bit (BUSY) i$ cleared and the done
flag (DONE) is set. If the interrupt is enabled, a request is made for the bus
at level 4 (BR4). The interrupt vector is at location 60,. The DONE bit is
cleared by any instruction which reads the contents of the buffer (TKB) into
the processor. If the DONE flag is cleared before the interrupt is granted, no
interrupt will occur. The keyboard must be read within 18 milliseconds of
DONE to ensure no loss of information.
Registers1
Teletype Keyboard Status (TKS)

IS 1t 7 6 0

I 0*
l-BUSY ’ $&TENB L- ROR
ENB

Bit
0 RDR ENB Requests that one character be read from the
reader; set from the bus: (Note: Setting RDR
ENB causes tape to advance by one character
which is shifted into TK8 if DONE is cleared.)
Receipt of START bit on the serial input line sets
BUSY, clears RDR ENB and clears TKB.
6 INT ENB O-No interrupt; l-Attach the keyboard and
reader to the priority interrupt system at bus
request level 4.
7 DONE Character available; cleared by reading the buf-
fer (TKB).
11 BUSY Character is being read; set by RDR EN8 going
’ to a 1. Cleared by DONE going to a 1.
I The following notation will be used throyghout this chapter for describing registers.
0 -A power clear sets this bit to 0.
1 -A power clear sets this bit to 1.
l -This bit can only be read from the bus.
$-This bit can only be set from the bus. If it is read, it will always appear
as zero.
Teletype Keyboard Buffer (TKB).

*
&BIT CHARACTER
I I
15 B 7 0

TELEPRINTER/PUNCH-On program command, a character is sent in parel-

lel from a memory location (or a general register) to the TPB shift register
for transmission to the teleprinter/punch unit. The control generates the
. .
start “space,” then shifts the eight bits serially into the Teletype u&,-and
then generates the stop “marks.” This transfer of information from the TPB
into the teleprinter/punch unit is accomplished at the normal Teletype rate
and requires 100 milliseconds for completion. The READY flag in the tele-
printer/punch indicates that the TP8 is ready to receive a new character. A
maintenance mode is.provided which connects the TPB eutput to the TKB
input so that the parallel serial and serial parallel shifting may be verified.

54
Registers
Teleprinter Status Word (TPS)

7 6 2
*

L I-
I 1 0 0 1

IN-f EN6 L MAINTENANCE

READY CONTROL

Bit
2 MAINT Maintenance function which connects TPB serial
output to TKB serial input.
6 INT ENB O-No interrupt; 1 -attaches the Teleprinter to
the priority interrupt system at BR4.
7 READY Set by punch/printer DONE; cleared by loading
the teleprinterbuffer (TPB).
Teleprinter Buffer (TPB)

15 8
I 7
a-BIT CHARACTER OATA

0
l

PRDGRAMMING EXAMPLE-To read i character from tape and echo it on

the printer:
ECHO: INC TKS ; set RDR ENB
c
TSTB TKS ; test for DONE set
BPL .-2 ; test again if not set
TSTB TPS ; test for printer READY set
BPL .-2 ; test again if not set
MOVB TKB, TPB ; put input character into’output
buffer to be printed
BR ECHO ; return for another character
PERIPHERAL ADDRESS ASSIGNMENTS
TKS 177560
TKB 177562
TPS 177564
TPB 177566
VECTOR ADDRESS Keyboard/ Reader 60
Teleprinter/ Punch 64
PRIORITY LEVEL set to BR&Teletype printer is lower than the Teletype
key board
MOUNTING-Requires one small peripheral controller mounting s,pace (l/4
of a DDll or one of two such spaces in KAll)

HIGH-SPEED PERFORATED TAPE READER PUNCH ‘AND

CONTROL (TYPE PCll)
TAPE READER-This device senses 8-hole perforated paper or Mylar tape
photo-electrically at 300 characters per second. The reader control requests
reader movement, transfers data from the reader into the reader buffer
(PRB), and signals the computer when incoming data is present. It does this

55
by setting a DONE bit. If the interrupt is enabled and the interrupt is granted,
the processor traps to location 70, and may immediately begin executing the
service routine for the paper tape reader.

Reghters
Paper Tape Reader Status Word (PRS)

15 11 7 6 0
* it *
0 0 0 0 0

L
l- ERROR L BUSY
L DONE
LINT EN6 ROR
RDR
ENB

Bll ~
0 RDR ENB Requests read of next character; can be set from
bus only if ERROR = 0. Clears PRB, sets BUSY.
6 INT ENB O-No interrupt; 1-attached to priority interrupt
system at BR4. (Note: Interrupt occurs when INT
ENB is a 1 and either the error flag, ERROR, or
the done flag, DONE, becomes a 1.)
7 DONE Set by character available; cleared by reading the
paper, tape reader buffer (PRB).
11 BUSY Set by RDR ENB going to a 1; cleared by DONE
going to a 1. J

15 ERROR Error Flag - Set or cleared by out-of-tape sensor *

or off line switch.

Paper Tape Reader Buffer (PRB)

it
E-BIT CHARACTER
I I
15 8 7 0

PROGRAMMING EXAMPLE/Gape reading subroutine (not using interrupt):

READ: INCB PRS ; enable reader
TEST: BIT $100 200 PRS ; test for error or done
BEQ TEST ; branch back if not done
BML ERROR ; branch if error = 1
MOVB PRB. RO ;~ get character from buffer
RTS R ; return to caller
ERROR: (message type out routine)
HALT ; wait for operator intervention
JMP READ ; try again when continue switch is hit.

TAPE PUNCH-This option of a Royal McBee paper tape punch that per-
forates B-hole tape at a rate of 50 characters per second. Information to be
punched on a line of tape is loaded in an B-bit punch buffer (PPB) from a
memory location or one of the general registers. The punch flag, READY,
becomes a 1 at the completion of punching action, signaling new information
may be transferred into the punch buffer and punching initiated.

56
Registers
Paper Tape Punch Status Word (PPS)

Bit
6 INT ENB O-No Interrupt; l-Attached fo priority iflterrUFt
system. (Note: An interrupt occurs when 1NT EN8
is a 1 and either the ERROR flag or the READY flag
becomes a ‘1.) i
7 READY Set by punch done; cleared by loading the paper
tape punch buffer (PPB).
15 ERROR Error Flag-Set by out-of-tape sensor: or unit power
off switch.

Paper Tape Punch Buffer (PPB)

I II &BIT CHARACTER DATA l

I
1
(5 s 7 0

Loading the buffer initiates punching.

PROCRAMMING EXAMPLE
PUNCH: BIT # 100200, PPS ; test for ready or error
BEQ PUNCH
BMI ERROR
MOV RO. PPB
RTS R ;
ERROR: (message type out)
HALT; wait for operator to fix punch
JMP PUNCH; try again when Continue is hit.

PERIPHERAL ADDRESS ASSIGNMENTS

PRS 177550
PRB 177552
PPS 177554
PPB 177556

VECTOR ADDRESSEGReader 70
Punch 74

PRIORITY LEVEL-Set to BR4. Punch is lower than reader.

MOUNTING-Electromechanical assembly-EIA Standard 19” rack, lOI/,”

vertical mounting space, by 171/2” deep.
PCll-M Controller--One small peripheral controller mounting space (l/4
of DDll or one of two such places in KAll).

-\
57
 ENVIRONMENTAL
55”-100°F
20% -95% RH (without condensation) ~-

MODEL DESCRIPTION POWER REQUIREMENTS

PC11 Reader, Punch & Control 115+10% 60 Hz
PCllA Reader, Punch &Control 115+-10% 50 Hz
PRll Reader 81 Control 115-c10°h 50-60 Hz
LINE FREQUENCY CLOCK (TYPE KWll-L)
The KWll=L real timeclock provides a method of measuring time intervals
at line frequency. This clock consists of a frequency source and control logic.
When enabled the clock causes an interrupt every 16.6 or 20 milliseconds,
depending upon Iflie frequency.
Register
Line Time Clock Status Register (LKS)
7 6
- (I
0 0 I
LINTR ENB
t CLOCK

Bit
6 INTR ENB When set, an interrupt will occur every time CLOCK goes true.
Cleared by program or reset or start sequence.
7 CLOCK Set to 1 every 16.6 milliseconds (60 Hz) or 20 milliseconds (50
Hz). Cleared by reading LKS, RESET or pressing ttie START
switch.
PERIPHERAL ADDRESS ASSIGNMENTS
LKS 177546
VECTOR ADDRESS 100
PRIORITY LEVEL BR6
MOUNTING-This option plugs into the KAll processor.

.
58
 CHAPTER 8 -~
DESCRIPTION OF THE UNIBUS
Communication between all system units in a PDP-11 configuration is done
by a single common bus: the Unibus. All communication-both instructions
and logical operations-is defined by a set of 56 signals. This set of 56 sig
nals is used for program controlled data transfers, direct memory data trans-
fers, priority bus control, and program interrupt.

This chapter presents the concepts of the Unibus and how they affect pro-
gram software and interfacing hardware. The use of the 56 bus signals to
effect.data transfers and to control bus use is also described.

GENERAL CONCEPTS OF THE UNIBUS

There are five major aspects of the Unibus that affect both software and
hardware considerations in the PDP-11.
SINGLE BU!j-The set of 56 signals that comprise the Unibus is the one
and only bus connecting all peripheral devices, memories, and the central
processor. Thus, to every device there exists a single set of signals by which
it can be interrogated by the processor or other devices, or be used by the
device itself to transfer data to and from memory.
The processor uses this same set of signals to communicate with all mem-
ories and devices. The important point here is that the form- of the com-
munication used by processor and peripheral devices is identical. Conse-
quently; the same set of program instructions used to reference memory
is used to reference peripheral devices. (A look at the PDP-11 instruction
set will reveal that there are no explicit l/O instructions.)
Peripheral devices in a PDP-11 system are designed to respond to the Unibus
in the same manner as memory. Device status registers, device con-
trol registers, and device data registers are each assigned unique “memory”
addresses. For example, the instruction MOVB RO, PUNCH would load the
punch buffer register with an 8-bit character contained in RO. Other in-
structions would monitor the punch status and the program could deter-
mine when the punching operation was complete.
BIDIRECTIONAL BUS-Unibus bus signals are bidirectional-the signal re-
ceived as an-input can be driven as an output, as shown in Figure 8-1.

r--- ----- Aurn7

RECEIVE. BUS SIGNAL

DRtVE BUS StGNAL

I

I
DEVtCE LDGtC
l- ----------- J

Figure 8-l Bidirectional Nature of the Bus

MASTER-SLAVE RELATION-At any one point in time, there is one device,
called the master, that has control of the bus. The master device controls

59
 the bus to communicate with other devices, call,ed slaves, on the bus. An
example of this relationship is the processor (master) fetching an instruction
from memory (which is always a slave).
INTERLOCKED COMMUNICATION-For erich control signal issued by the
master device, there is a response from the slave; thus bus communication
is independent of the physical bus length and the response time of the mas-
ter and slave devices. Also, master-slave relationships can exist in nearly
any combination between fast-responding and slow-responding devices.
* DYNAMIC MASTER-SLAVE RELATION-Master-slave relationships are dy-
namic. The processor, for example, can pa&s bus control to a disk. The disk,
as master, could then communicate with a slave memory bank.
UNIBUS SIGNALS
The 56 Unibus signals can be divided into two major groups-the interrupt
group and the non-interrupt group. The interrupt group can then be sub-
divided into two classes-the request and control class and the grant class.
All bus signals except the grant class are bidirectional in nature and are
% connected to every device (though they may not be used by every device).
The grant signals, because of their special nature in priority bus control
(to be explained later), are bussed through each device and are unidiiectional
in nature.
NON-INTERRUPT SIGNALS
Data Lines (0 < 15:OO >)-(Note that the notation A <a:b> specifies
b - a + 1 signal lines which are named Aa through Ab.) The 16 data lines
are used to transfer information between master and slave. This is the bit
format:

I HIGH BYTE I LOW BYTE 1

(5 8 7 0

Address Lines (A < 17:00 >)-The 18 address lines are used by the master
device’to select the slave (a unique core memory or device register address)
with which it will be communicating. This is the bit format of the 18 signals:

A < 15:Ol > are used to specify a unique 16-bit word group. In byte opera-
tions, A00 is used to specify the byte being referenced. If a word is refer-
enced at X (X must be even, since words can be addressed on even bound-
aries only), the low byte can be referenced at X and the high byte at X + 1.
A < 15:00 > are supplied by the software as memory reference addresses.
Al7 and Al6 are used as extended memory bits for relocation and as pro-
tection schemes in future systems. In the PDP-11/20 and the PDP-ll/lO,
Al7 and Al6 are asserted or forced to 1 whenever an attempt is made to
reference a memory location where A15’= Al4 = Al3 = 1. Thus the hard-
ware converts the 18bit software address to a full Is-bit-bus address.
An address map is shown in Figure 8-2.

60
 oooo#O-017777
trt
scFrwAF& AOORESS

MEMORY
4K
BANK
HAROWARE

oooooo-own7
ADDRESS
.

02CWO-037777 02ooOO-037777

El I
I
2nd
MEMORY
4K
0Aw

I
I
t40000-t57777 t40000-157777
7th 4K
MEMORY BANK _
t6oooo-t77777 760000-777777
PERIPHERAL
BANK
El

Figure 8-2 Address Map .

The peripheral bank is composed of the processor’s fast memory, status

register, console switch register, and all device registers.

Control Lines (C < 1:0 > )-These two bus signals are coded by the master
device to indicate to the slave one of four possible data transfer operations.

Master Synchronization and Slave Synchronization (MSYN, SSYN)-MYSN

is a control signal used by the master to indicate to the slave that address
and control information is present. SSYN is the slave’s response- to MSYN

lnititiliiation (INIT)-This signal is a power clear signal asserted by the con-

sole and the processor which is used to reset peripheral devices.
PA, PB. SPl. SP2-These lines are not implemented on the PDP-ll/lO or
PDP-ll/PO.

INTERRUPT SIGNALS
Bus Request Lines (BR i: 7:4 > )-These four bus signals are used by
peripheral devices to request control of the bus.
Bus Grant Lines (BG < 7:4 >-)-These signals are the processor’s response
to a BR. They will be asserted only at the end of instruction execution.

Non-Processor Request (NPR)-This is a bus request from -a peripheral

device to the processor.
Non-Processor Grant (NPG)-This isthe processor’s response to an NPR. It
occurs at the end of bus cycles within the instruction execution.
Selection Acknowledge (SACK)-SACK is asserted by a bus-requesting device
that has received a bus grant. Bus control will pass to this device when the
current master of the bus completes its operations.

INTERRUPT (INTR)-This signal is asserted by the master to start program

interruption in the processor.
Bus Busy (BBSY)-This signal denotes bus in use by a master device.

UNIBUS DATA TRANSFER OPERATIONS

Direction of data transfers on the Unibus is defined in relation to the master

61
device. A data transfer from processor to memory (always a slave) is “data
out,” and a transfer from memory to processor is “data in.”

TYPES OF DATA TRANSFERS-The type of data transfer being made between

master and slave is determined by the C lines coded as follows: _

DATO AND DATOB-The DATO and DATOB operations are used to transfer
data out of the master to the slave. DATO is used to transfera word to the
address specified by A < .17:01 >. The slave ignores A00 and the data ap
pears onD < 15:00>. DATOB is used to transfer a byte of data to the ad-
dress specified by A < 17:OO >. A00 = 0 indicates, the low byte, and data
appears on D < 07:OO >; A00 = 1 indicates the high byte, and data appears
on D < 15:08 >.

DATI AND DATIP-The DATI and DATIP operations transfer data from a slave
whose address is specified on A < 17:Ol > into the master. Both transfers
are made in words on D < 15:00 >. In destructive read-out devices, DATI
commands a read-write operation, while a DATIP commands a read operation
only and sets a pause flag. When the device receives the subsequent DATO
or DATOB and its pause flag is set, the usual read cycle is skipped and an
immediate write cycle is initiated. Thus, DATlPs are immediately followed
by a DATO or DATOB to effect a read-modify-write data exchange. In non-
destructive read-out devices, DATI and DATIP are treated identically.

This diagram illustrates the data flow in the four data transfers:
DATI OR DATIP
DATA= D<l5:00> I
15 8.1 7 0
SLAM ~~EGISTER
I
HIGH BYTE , LOW BYTE
I
A A
DAToBI\AOO DATOBhm
DATA=D<t5:08 > DATA=D <OX00 >
DATA=CKl5:00>
Figure 8-3 Data Flow

Note that all transfers into the master are word operations; it is up to the
master to accept the appropriate byte. On a DATOB, the master must place
the byte on the appropriate data lines; the slave must accept the proper byte.

DATA TRANSFER EXAMPLES-The bus operations used by the processor for

a typical instruction sequence illustrates how the data transfer operations
are used. The “program” starts at location 1000:
1000: INCB @RO
ADD #3, @RO
where RO contains 500 and location 500 contains 10023. The result of this

62
I instruction sequence will leave 10027 in location 500. In binary form, this
coding appears as:
1000: 105210 ;INCB @RO
1002: 062710 ;ADD (PC)+, @RO
1004: 000003 ;3
. The following table lists the bus operations that result as a consequence
of these two instructions:
Processor Cycle Bus Operation Bus Address Dais Transferred _
1. Fetch DATI (PC),= 001000 105210
2. Destination DATIP (RO) = 000500 010923
3. Execute DATOB (RO) = 000500 000024
4. Fetch DATI (lqz) = 001002 062710
5. Source DATI (PC)= 001004 000003
6. Destination DATIP (RO) = 000500 010024
7. Execute DATO (RO) = 000500 010027
Note that instep 3, it is inconsequential what data appears on D < 15:OB >;
the slave accepts only the modified low byte.

A second example of bus ‘operation compares the contents of the Teletype

keyboard data buffer whose address is 177560 with the ASCII value for the
letter “A.”

200: CMPB @#177560, #301

This instruction is assembled in three words as follows:

200: 123727 ;CMPB @@7)+, WV+
202: 177560 ;Address of data buffer
204: 000301 ;301

The processor will execute this instruction with these cycles:

Processor Cycle Bus Operation Bus Address Data Transferred
1. Fetch DATI (PC) = 200 123727 ’
2. Source DATI (PC)=202 177560
3. Source DATI 777560 ASCII
4. Destination DATI (PC) = 204 000301
5. Execute none - condition codes set internally.

Note that in step 3, the soware specified address 177560 was converted to
the bus address 777560.

SIGNAL DESCRIPTION OF DATA TRANSFERS-Figure 8.4(a) shows the sig

nal flow between master and slave during a DATO operation. (The sequence
is similar for DATOB except that only a byte of information is transferred.)
The master sets Control for DATO, sets Address for the unique slave address,
and sets Data for the information to be transferred. The master then asserts
MSYN. This signal is received by the slave that recognizes its address: it
responds by accepting the data arid asserting SSYN. SSYN is received by the
master which then negates Control, Address, Data, and MSYN. The slave
sees MSYN negated and negates SSYN. The master device continues its
operation when it sees SSYN negated.

63
 \
MASTER SLAVE

OPERATION: DAl-0

A.C,D
MSYN ‘4

rg------ SSYN

A’C’D 3
SSYN
I

Figure 8-4(a)
The flow of signals for DATI is shown in Figure 8.4(b). (DATIP is similar
except that the internal operation of the slave device is modified.) The master
sets Control for DATI, sets Address for the slave to be se’lected, and asserts
MSYN. The selected slave responds by setting Data for the information re-
quested and asserts SSYN. The master sees SSYN, accepts the data, and
then negates Control, Address, and MSYN. The slave sees MSYN negated
and negates SSYN. The master continues when it sees SSYN negated.

A more detailed signal sequence for the DATI, DATIP, DATO, and DATOB bus
operations can be found in Appendix D.
MASTER SLAVE
/
OPERATION: DATI

W
MSYN

, SSYNaD
v

$% _
ssVN6

i ’

Figure 8-4(b)

UNIBUS CONTROL OPERATIONS

The following section will deal with how a device becomes master of the bus
.and how control of the bus is transferred from one device to another. TWO
additional bus operations will be presented-the PTR (Priority Transfer) and
INTR (Interrupt).
In normal operation, the processor is bus master, fetching instructions and
operands from memory. Other devices on the bus have the capability of
becoming bus master, and use the bus for one of two purposes: l), to gain
direct memory access or 2). to interrupt program execution and force the
processor to branch to a specific address:

PRIORITY ARBITRATION-Transfer of bus control from one device to another

is determined by a priority scheme in which three factors must be considered.
First, the processor’s priority is determined by bits 7, 6, and 5 in the pro-

64
 cessor status register. These three bits set a priority level that inhibits
granting of bus requests-on lower levels.

Second, bus requests from external devices can be made on oni of five
request lines. NPR has the highest priority, and its request is honored by the
processor between bus cycles of an instruction execution. BR7 is the next
highest; BR4 is the lowest. These four lower level requests are honored by
the processor between instructions, except when the instruction currently
being executed causes an internal trap (either an error or trap instruction).
In this case, BR requests will not be honored until completion of the first
instruction after the trap sequence. Thus if two requests are made to the
processor for bus control, the higher of the two requests will be honored first.
Third, in response to a bus request, the processor may honor the request by
asserting a bus grant (BG) corresponding to the line on which the bus re-
quest was made. This signal is passed serially through each device in the
system. If a device had made a request, it would. block the grant signal
and prevent it from reaching the following devices. Thus, in this “pass-the-
pulse” chain, the device that is closest to the processor has the highest
~ priority on that request level.
This table lists device priorities:
Highest: Devices on NPR
Processor when priority = 111
Devices on BR7
Processor when priority = 110
Devices on BR6
Processor when priority = 101
Devices on BR5
Processor when priority = 100
Devices on BR4
Processor when priority = 011
Internal options
Processor when priority = 010
Internal options I
Processor when priority = 001
Internal options
Lowest: Processor when priority = 000
When the processor’s priority is set at N, all requests for bus control at
level N and below are ignored.
SELECTION OF NEXT BUS MASTER-The signal sequence by which a device
becomes selected as next bus master is the PTR (Priority Transfer) bus ,
operation. Note that this operation does not actually transfer bus control:
it only selects a device as next bus master. It takes one additional condition
to complete the transfer: the current bus master must complete its bus
operations. The signal that indicates this is BBSY. Thus, when a device makes
an NP,R or BR request to the processor for bus control, it waits until it first
becomes selected as next bus master by the PTR operation and second, it
no longer senses BBSY, The negation of the BBSY signal indicates that
the current master has completed its bus operation. The selected device
now becomes bus master and asserts BBSY itself.
INTERRUPT SEQUENCE---Once the device has bus control and is asserting
BBSY itself, it‘is sole user of the bus until it releases its control. This release
of control can be made either actively or passively. Passive release is realized

65
by negating BBSY. Bus control will then pass to either a device that was
selected in the meantime by another PTR sequence or back to the processor,
which will continue where it was interrupted. Active release of bus control
is realized through the INTR bussequence.
The INTR (interrupt) operation is used by the bus master to transfer to the
processor a memory address (called the interrupt vector). Two consecutive
words, the starting address of an interrupt service routine and a new status
word, are stored at the interrupt vector address. After the INTR sequence is
Complete, the Processor automatically becomes bus master and begins a trap
sequence in which it stores the current value of the PC and PS on the stack
and fetches a new PC and PS from the location pointed to by the interrupt
vector. Thus, the next instruction executed is the Start of, the interrupt
service routine.
It is illegal to issue an INTR command after gaining control of the bus by
requesting on an NPR line. NPR requests are granted during instruction
execution and external bus masters must restrict their bus use* to nonpro-
cessor activities.

Interrupt Servicing Sequence Example-The following is an example of the .*

INTR sequence. , _

When a peripheral requires service and requests control of the bus with a
BR signal, the operations undertaken to “service” the device are as follows:
l Gain Control of the Bus-When the processor has no higher priority tasks
to complete, it relinquishes the bus to that device. Higher priority items are
(in order of priority):
1. Acknowledging an NPR request
2. Handling a processor error (illegal instructions, requirements for non-
existent memory, etc.)
3. Completing the current instruction
4. Acknowledging a trace trap
5. Continuing a higher priority process
6. Acknowledging a higher level BR signal
7. Acknowledging same level BR signals for devices closer to the processor
l Do INTR Sequence-when the device has cpntrol of the bus, it initiates
an INTR sequence, transferring to the processor the interrupt vector address
which specifies two words in memory containing the address and status of
the appropriate device service routine.

0 Push Old Interrupt Vector Onto Stack-The processor then “pushes”-

first, the current central processor status (PS) and then the current program
counter (PC) onto the processor stack.
l Fetch New Interrupt Vector-The new PC and PS (the “interrupt vector”)
are taken from the address specified by the device, and the device service
routine is begun. Note that those operations all occur automatically and that
no device polling is required to determine which service routine to execute.
Example of NPR Operation-Disk operation gives an example of a device
‘which uses the bus for direct memory access. Under program.control, the
processor would initialize registers in the disk control that specify word count
(WC, number of words in block of data to be transferred), memory address
(MA, the address at which the block of /data is found or is loaded), and Track
Address (TA, the point on the disk where the block of data starts). Also, the

66
program would set certain function bits in the disk’s command and status
register that specify a read or write function. For this example, assume the
disk was set to read. *
Once the disk’s control registers are initialized, the disk control logic’starts
a search for the requested data. (fhe processor in the meantime has con-
tinued in its program execution.) When the disk has found the data, it
assembles the first l&bit word from the disk surface into its data register.
The disk now requests bus control via the NPR request line. The processor,
when it has completed its current bus cycle of the current instruction and
no higher NPR requests exist, grants control of the bus to the disk. The disk,
as bus master, effects a DATO bus operation, transferring the contents to
its data buffer to the core address held in its MA. The MA is now incremented
and the WC is decremented. When the DATO operation is COftydete, the disk
passively releases control of the bus.
When the second word has been assembled, the disk again requests bus
control, does a data transfer, and then releases bus control. This cycle is
repeated until the WC reaches zero. At this point, the disk has completed
the transfer that was requested.
To notify the program that the transfer is finished, the disk initiates a request
for bus control at the BR level, gains control when higher priority requests
are satisfied, and does an immediate INTR to the processor and causes the
program to branch to a specific service program (as described in the previous
example). \
Details of the INTR and PTR bus operations can be found in Appendix D.

67
 ,

,I ;

The plug-in console board with modular construction is supplied

in the basic 11/20 configuration. In addition to aiding program-
ming, console contributes to ease of maintenance on the PDP-11. *

68
 CHAPTER 9
Interfacing
A typical device bus interface as shown in Figure 9-l is composed of five
major components: 1). Registers; 2), Bus Drivers and Receivers; 3). Address
Selector; 4). interrupt Control; and 5), Device Control Logic.

REGISTERS
Each device is assigned bus addresses at which the program can inter-
rogate and/or load the device status, control, and data registers. The stan-
dardized mapping for these registers and the bit assignments of the corn- .
mandlstatus register (CSR) were given in Chapters 5 and 6.

As shown in Figure 9-1, all information flow between the device logic and ’
the Unibus is done through the registers. In general, registers are designed
to be both loadable and readable from the bus. This allows the program to
use such instructions as ADD RO, REG, or INC REG. However, registers can
be “one-sided,” either “read-only” or “write-only.” Examples of read-only
bits are the DONE and BUSY flags in the device’s CSR. These bits are de-
rived from the internal state of the device log& and are not under direct
program control. Write-only registers are used when it is unnecessary to
read back information. Attempting to read such a register would result in an .
all-zero transfer. The instructions effective with this type of register are then
limited to those which load the register such as MOV RO, REG, or CLR REG
(as opposed to ADD REG, RO, or INC REG).

Figure 9.1 Typical Peripheral Device Interface

BUS DRIVERS AND RECEIVERS

To maintain the transmission-line characteristics of the Unibus, special cir-
cuits are required to pass signals to and from the bus. The majority of bus
signals (all except the five grant lines) are received, driven and terminated
as shown in Figure 9-2.

69
 I
I
I
I
I
I
I
I M930

L- --- -J i WI--
DRIVER

Rl , R2=190fi 5% 1/4W
R3. R4 = 390A 5% iI4W

Figure 9.2 Typical Unibus Line

Information is received from the bus using gates which have a high input
impedance and proper logic thresholds. High input levels must be greater
than 2.5 V with an input current less than 160 pa. Low level input must be
less than 1.4 V with an input current greater than 0 pa.
information transmitted on the bus must be driven with open collector drivers
capable of sinking 50 ma with a collector voltage of less than .8 V. Output
leakage current must be less than 25 ~a.
In PDP-11 systems, the bus signals are terminated at both ends by resistor
dividers provided on the M930 module. Physically, an M930 is located in
the processor; another is located at the last unit on the bus. A bus signal
sits at logical “0” (inactive, or negated state) at a voltage of 3.4 V. A bus
line is at logical “1” (active, or asserted) when it is pulled to ground.
Drivers and receivers meeting these specifications are available on the
M783, M784 and M785 modules as shown in Figures 9-3, 9-4 and 9-5.

70
Ml05 ADDREsS SELECTOR
Tho ‘Ml05 Address Selector as shown in Figure 9-6 is used to provide gating
signals for up to four device registers. The selector decodes the 18-bit bus
address on A < 17:00 > as follows:

Figure 9.3 M783 Unibus Drivers ix

Figure 9.4 M784 Unibus Receivers

 . Figure 9.5 M785 Unibus Drivers and Receivers

A00 is used for byte control. A01 and A02 are decoded to provide one of
four addresses. A < 1203 > are determined by jumpers on the card. When
the jumper is in, the selector will look for a 0 on that address line-
A < 17:13 > must all be l’s-(this defines the external bank). Other bus
inputs to the selector are C < 1:0 > and MSYN. The single bus output is
SSYN. The user signals are SELECT 0. 2, 4, and 6 (corresponding to the
decoding of A02 and AOl, one of which is asserted when A < 17:13 > are all-
l’s and A < 12:03 > compare with the state of the jumpers. Other user sig-
nals are OUT HIGH (gate data into high byte), OUT LOW (gate data into low
byte), and IN (gate data onto the bus). The equations for these last three
signals are as follows: .
OUT HIGH = DATO + DATOB;AOO
OUT LOW = DATO + DATOB*m
IN = DATI + DATIP

where “+” means a logical or and “*” means a logical and.

Use of the M105, drivers, receivers and a flip-flop register is shown in- Fig
ure 9-7.

72
 EXT. CAP
I

-J f , 1 1 SELECT 2 H

EH
SELECT 4 H

SELECT 6 Ii

A62L J I
A61 L
AmL I
OUTHl6HH
GIL OUT LW Ii
CaL a IN tl

Figure 9.6 Ml05 Address Selector

4
M782 INTERRUPT CONTROL
The M782 Interrupt Control module contains the necessary logic circuits to
allow a peripheral device to gain bus control and perform a program inter-
rupt. The three circuits on this card are block diagrammed in Figure 9-8.
Note that only signals relevant to the user’s interface are shown; bus
signals SSYN, BBSY and SACK have been omitted for clarity.
The Master Control circuit is used to gain bus control. When INT and INT
ENB are asserted, a bus request is made on the request line to which BR
is jumpered. When the processor issues the corresponding grant and other
bus conditions are met, the MASTER signal is asserted, indicating that this
device now has bus control. Note that this circuit also can be used to gain
bus control on an NPR line for a device which requests the bus for direct
memory access.

73
 . .

-i
I
L--- -- ------2

Figure 9.7 Typical Peripheral Device Regker

74
 In addition to two Master Control circuits, a third logic network provides the
necessary signals and gating to perform the INTR bus operation. When either
of the START INTR signals is asserted, the INTR bus signal is asserted
along with a vector address qn D < 07:02 >. Bits 07:03 are determined by
jumpers on the card. A jumper “in” forces a 0 in that bit. Bit 2 is controlled
by Vector Bit 2. When the processor responds to the INTR signal by asserting
SSYN, the INTR DONE signal is asserted. This line is used to clear the
condition which asserted INTR START.

Figure 9.8 M782 Interrupt Control

Figure 9-9 shows a possible interconnection of the M782 to provide inde-

pendent interrupts for two possible conditions in a device: ERROR and DONE.
The ERROR and DONE signals shown in Figure 9-9 are signals from bits 15
and 7 in a device’s CSR. Likewise ERROR INT ENB and DONE INT ENB are
derived from the CSR. Both interrupts in this example are tied to the BR4
level: the corresponding grant line BG4 enters the ERROR Master Control and
is passed on to the DONE Master Control. Thus, ERROR ha% a slightly higher
priority interrupt level than DONE.
Both MASTER signals are tied to the INTR control. Thus, whenever either
ERROR or DONE gains bus control, an INTR operation is initiated. Note
that Vector Bit 2 is a 1 or 0 as a function of which master control
_ is interrupting. Also, INTR DONE is tied to MASTER CLEAR to clear the
master condition.
 ,
DEVICE CONTROL LOGIC
The type of control logic for a peripheral depends on .the nature of )h”
device. Digital offers a wide line of general-purpose logic IllO~Ul~S for IITF
plementing control logic. These modules are described in detail In another
Digital publication: The Logic Handbook.

Figure 9.9 Typical‘lnterconnection of M782 Interrupt Control

76 .
 CHAPTER 10
CONFlGURATtON AND INSTALLAflON PLANNING
MODULAR CONSTRUCTlbN
Physically, the PDP.11 is composed of a number of System Units. Each
System Unit is composed of three 8-slot connector blocks mounted end-to-
end as shown in Figure 10-l. The Unibus connects to the System Unit at
the lower left and at the upper left. Power also connects to the unit in the
leftmost black. A System Unit is connected to other System Units only via
the Unibus.

UNIBUS CONNECTION

p:*~~

Figure 10.1 System Unit

The remainder of the System Unit contams logic for the processor, memory
or an I/O device interface. This logic is composed of single height, double
height, or quad height modules which are 8.5 ” deep.
The use of System Units allows the PDP-11 to be optimally packaged for
each individual application. Up to six System Units can be mounted into a
single mounting box. For a basic PDP-ll/PO system, the processor/console
would fill 21/2 System Unit spaces and 4096 words of core memory would
fill one System Unit space. This leaves 21/, spaces for user-designated op
tions. This would allow the user to add 8,192 words of additional core
memory, a Teletype control, and a High-Speed Paper Tape Control, or 4,096
words of core memory and six Teletype interfaces, Larger systems will
require a BAll-EC or BAll-ES Extension Mounting Box which contains space
for six additional System Units.
The use of System Units also facilitates expansion of systems in the field
and service. To add an additional option to a PDP-11 system, the proper
System Unit is mounted in the Basic or Extension Mounting 80x and the
Unibus is extended. Servicing of the PDP-11 can be done by swapping
modules or by swapping System Units.
MOUNTING BOXES AND CABINETS
The PDP-11 is available as either a tabletop or rack-mounted configura-
tion. The rack-mounted configuration may be installed in a DEC cabinet or
mounted in a customer cabinet. The PDP-11 mounts in an EIA standard 19-
inch cabinet. The rack-mounted PDP-11 has tilt-slides as standard mount-
ing hardware.
1 The following mounting units and cabinets are‘available for PDP-II systems.
PDP-11 TABLETOP BOX AND POWER SUPPLY FOR 11/20, ll/lO SYSTEMS-
fBAll-CC AND H720)-This cover and box may be specified with a basic
i1/20 and ll/lO system and includes:
'1. H720 Power Supply
2. 15’ of power cord with ground wire

77
 + For 115 V standard, 3prong; U-ground, 15-ampere connectors
+ For 230 V pigtail leads on one end
3. Cooling Fans
4. Filter
5. Programmers Console with 11/20 or Turn-Key Console with ll/lO
Approximate Size-11" high, 26” wide, 24” deep. Figure 10-2 shows the
layout of this unit.

I
*
Figure 10.2 Table Top PDP-11 Dimensions
Approximate Weight-100 Ibs. (including CP, console and 4K core)
. Power-12OV +- 10%,47-63 Hz 6 amps. single phase
(BAll-CC
and H720-A)
230V -c 10%,47-63 Hz 3 amps. single phase
(BAll-CC and H720.B)
PDP-11BkC MOUNTINGBOXAND PGWERSUPPLY(BAll-CSAND H720)
-This basic mounting box may be specified with a basic 11/20 or a ll/lO
system and includes:
1. Tilt and Lock Chasis Slides
2. H720 Power Supply
3. 15’ of power cord with ground wire
+ For 115 V standard, 3-prong, Uground, 15-ampere connector
+ For 230 V pigtail lead% on one end I
4. Cooling Fans
5. Filter
6. Programmers Console ‘with 11/20 or Turn-Key Console with lI/IO
Approximate Size-lbl/2” high, 19" wide, 23” deep. Figures 16-3, IO-4 and
lo-5 show the layout of this unit and give slide dimensipns.

.
78
. Approximate Weight-90 Itis. (including CP, cdnsole and AK core)
Power-12OV zflO%,47-63 Hz 6 amps. single phase
@All-C5 and H720-A) X -
230V +10%,47-63 Hz 3 amps. single phase
(BAll-C5 and H720-B)

~
Figure 10.3 Rack Mountable PDP-11 Dimensions

Figure 10.5 Side View of Mounting Hardware

PDP-11TABLETOP UcFENSfdh MOUNTING BOX (BAll-EC)-The tabletop
Extension Box is supplied, when ordered, for mounting of up to 6 additional
System Units which can not be contained in the Basic Mounting Box. This
unit is supplied with:
1. 15' of power cord with ground wire
+ For 115 V standard, 3-prong, U-ground, 15-ampere connector
+ For 230 V pigtail leads on one end
2. Cooling Fans
3. Filter
4. Front Panel
5. Unibus Cable from Basic Mounting Box, 8’6” long

Approximate Size-11" high, 20” wide, 24” deep

Power-120 V -C lo%, 47-63 Hz 6 amps. r single phase
(when H720-A is added)
230 V + 10%,47-63 Hz 3 amps. single phase ‘)
(when H720-B is added)
PDP-Ll EXTENSIONMOUNTING BOX(BAll-ES)-The Extension Box is sup
plied, when ordered, for mounting of up to 6 additional System Units which
can not be contained in the Basic Mounting Box. This unit contains:
1. Tilt and Lock chassis slides
2. 15’ of power cord with ground wire
+ For 115 V standard, 3-prong, U-ground, 15-ampere connector
+ For 230 V pigtail leads on one end
3. Cooling Fans
4.. Filter
5. Front Panel
6. Bus Cable from Basic Box, 8’ 6” long
Approximate size--101/z” high, 19” wide, 23” ,deep
Power-120 V + lo%, 47-63 Hz 6 amps. single phase
(when H720-A is added).
230V +10%,47-63 Hz 3 amps. single phase
(when H720-B is added)
PDP-11FREESTANDING BASE CABINET (H969CA)This optional cabinet
cabinet can be used to mount the BAll-CS Basic Mounting Box and a
BAll-ES Extension Mounting Box supplied with Tilt and Lock chassis slides
in addition to other PDP-11 equipment. . .

Panel capacity is six lOI/” high mounting spaces, each of which is covered
with black plastic panels if equipment is not mounted-(5 panels, maximum,
supplied).
items supplied with the cabinet include:
1. H950-A Frame
2. H952-E Coasters
3. H-952-F Levelers
4. H-952-C Fan Assembly (in top of cabinet)
5. H-950-S Filter
6. PDP-11 Logo
7. H-950-B Rear Door
8. lOl/," Plastic Bezels, maximum of 5 supplied
9. Two H952-A End Panels

80
 10. H955)-D Mounting Panel Doors
11.. H952-B Stabilizer Feet
12. #7406782 Kick Plate
13. #7005909 Power Distribution Panel (ac ?rrd dc, mounted on upper
left side)
Approximate Size-22” wide, 39” deep (including stabilizer feet), 711/” high
Approximate height-150 Ibs. (without computer)

e Voltage-115 V 60 Hz (for fans)

230 Y 50 Hz (for fans)
PDP-11 POWER SUPPLY SUBSYSTEM H728-This Power supply is used in
the Basic and Extension Mounting boxes and supplies power to all devices
mounted in one of these boxes. It is included in. basic PDP-11 systems,
‘but must be ordered separately with a BAllES or BAllEC Extension Mount-
ing Box.
Approximate Size-161/,” wide, 8” high, 6” deep
Approximate Weight-25 Ibs.

Voltages-(specify input voltage)

IN 108V +-lo%, 47-63 Hz 6 amps (H720A)
120V *lo%. 47-63 Hz 6 amps l (H720A)
216V *lo%, 47-63 Hz 3 amps
228V +-lo%, 47-63 Hz 3 amps gEE;
240V *lo%, 47-63 Hz 3 amps. (H7208)

OUT +5v ,*50/o . 12 amps

-15v ‘5% 10 amps

+8RMS (unregulated) 1.5 amps,

-22v (unregulated) 1.0 amps
FREESTANDING PROGRAMMER’S TABLE (H952-HA)-This freestanding table
fits directly below the programmer’s console in the Freestanding Base
Cabinet and extends into the cabinet approximately 1”. The surface plate is
supported by its own adjustable height legs.
Approximate Size-20” extension from cabinet, 19” wide, 27” above floor

SYSTEM UNITS AND CABLES

The following items are available for mounting standard and special periph-
eral device logic into a PDP-11 system.

PERIPHERAL MOUNTING UNIT (DDll-A)-The DDll is a prewired System

Unit which allows standard small peripheral interfaces to be mounted in a
PDP-11 system. It accepts standard small peripheral interfaces (up to 4)
such as the KLll Teletype Control or the controller portion (PCll-M) of the
High Speed Reader/Punch. For mounting, it requires one-sixth (l/6) of a
BAll Mounting Box.
BLANK SYSTEM. UNIT (BBll)-The BBll consists of three 288-pin con-
nector blocks connected end-to-end. This unit Is unwired except for Unibus
and power connections and allows customer-built interfaces to be integrated
easily into a PDP-11 system. For mounting it requires one-sixth (l/6) of a
BAll Mounting Box.

81
 UNIBUS MODULE (M920)-The M920 is a double module which connects
the Unibus from one System Unit to the next within a Mounting Box. The
printed circuit cards are separated ‘by 1” for this-purpose. A single M920
will carry all 56 Unibus signals .and 14 grounds.

UNIBUS CABLE (BCllA)-The BCllA is a 120~conductor flexprint cable used

to connect System Units in .different mounting boxes or a peripheral device
which is removed from the mounting boxes.

The 120 signals consist of the 56 Unibus lines plus 64 grounds., Signals and +
grou,nds alternate to minimize cross talk. .

BCl lA-2. 2’
BCllA-5 5’
BCllA-BA 8’6”
BCllA-10 10’
BCl l A-15 15‘
BCl lA-25 25’

CABLE REQUIREMENTS
When an Extension Mounting Box is used, an external cable, the @CIlA, is
the only signal.conn@ion between mounting boxes. This external bus cable
may also be. used to connect other peripherals to the PDP11. The maximum
combined, internal and external, bus cable.length is 50’. -

’ PDP-11/20 POWER REQUlkEMENTS /

Input Voltage and Current-105-125 Vat, 6 amperes, 210-260 Vat 3 am-
peres, (single phase)
Line Frequency47-63 Hz
Pdwer Dissipation400 watts d
A standard 15-foot, 3-prong,- U-ground, 15-ampere, line cord is provided on
the rear of the PDP-11 for connection to the power source on 120 Vat
models. On 230 Vat models, a 15-foot, S-conductor cable with pigtails is
provided.

TELETYPE REQUIREMENTS
The standard Teletype requires a floor space approximately 221/2 inches
wide by 181/s inches deep. The Teletype cable length restricts its location to
within 8 feet of the si#e of the computer.

Input Voltage-115 Vat *lo%, 60 Hz kO.45 Hz, 230 Vat *lo%, 50 Hz

-CO.75 Hz
Line Current Drainl2.0 amperes
Power Dissipation-150 watts
The Teletype plugs into the rear of the PDP-11 Basic Mounting Box and is ’
turned ON and OFF by the~eower s,witch on the front panel of the PDP-11.

ENVIRONMENTAL REQUIREMENTS
The PDP-11 is designed to operate from +lO to +5O”C and with a relative
humidity of from 20 to 95% (without condensation).

82
.
. INSTALLATION .PROCEDlJRE
The PDP-11 is crated for shipment to the customer site to prevent damage.
Installation is provided by DEC personnel at the customers site.
Computer customers may send persorinel to instruction courses on camput&
operation, programming, and maintenance conducted regularly in Maynard,
Massachusetts, Palo Alto, California, and Reading, England.

83
The PDP-11 has adopted a moddlar packaging approach to allow
custom configuring of systems, easy expansion and easy servicing.

.. 84
 CHAPTER 11
PAPER TAPE SiFTWARE SYSTEM‘
PAPER TAPE SOFTWARE SYSTEM (PTS) ’
PTS is a compatible group of software packages designed to aid development
of PDP-11 application programs. A brief description of each item with its
major features is offered below with detailed programming information avail-
able in corresponding software user’s manuals.
PTS FEATURES
l 4K Absolute Assembler
l Symbolic Program Editor for editing of paper tape which is string oriented
0 On-Line Debugging Aid allowing rapid and accurate modification of assem-
bled programs
0 I/O Driver Routine.allowing subroutine level communication with periph-
eral devices. and double buffered input/output operation concurrent with
running programs
0 Floating Point Math Package’using both reentrant and relocatable code
0 General Utilities in.cluding loaders and dump routines
PALIlA ASSEMBLER-This two- or three-pass assembler runs on a PDP-11
with 4K words bf core memory and an ASR-33. It will also accommodate a
high-speed reader/punch. Optional outputs include the absolute object code,
an assembly listing containing each sdurce statement, and an indication of
any errors detected in the statement. A symkol table may be alphabetically
listed.
ED11 EDITOR-The PDP-11 Editor (EDll) allows the user to type identified
portions of source program on the teleprinter and to make corrections or
additions. This is accomplished by typing simple commands that cause the
Editor to-read, print, punch out on paper tape, search, delete and/or add to
the text of the program.
Use of the ED11 presupposes no special knowledge or technical skill beyond
that of the operation of explicitly defined one-character commands. The
commands are grouped according to function: input, positioning of the
current-character location pointer, output, search (which is done by charac-
ter string), insert, delete, and exchange of text portions.
ED11 uses 2,000 words of-core and requires an ASR-33 unit which includes
a printer, keyboard, paper tape reader and paper tape punch. Alternatively,
a KSR-33 may be used in conjunction with the high-speed paper tape’reader
and punch.
ODT-11ON-LINE DEBUGGING TECHNIQUE-ODT-11 is a core resident pro-
gram which allows the user to debug his binary programs at the console by
running them in specific segments and checking for expected results at vari-
ous points. If modification of the program is needed, the user can alter the
contents of the appropriate location by “opening’! it and typing in new data.
Two versions of ODT are available, one being a subset of the other. The
larger system uses 750 words of core and utilizes an ASR-33, or a KSR-33
and a high-speed paper tape punch and reader. The smaller version uses the
same peripherals and 500 words of core. Up to eight breakpoints can be set
using the larger .version of ODT, while one breakpoint is allowed in the
smaller version.

85
Debugging operations alternate between commands to ODT and the running
of the program to be debugged. Breakpoints are set in the user’s program by
ODT commands, and a command to run starts execution of the program.
When a breakpoint is encountered, the program run is suspended, and the
progress of its execution can be monitored and altered. This is accomplished
by using commands to open memory locations of interest, as well as special
registers.
An Operator may examine and change the operating priority of both ODT
and the user’s program, the mask and address range for searches, results
of logical and arithmetic operations, the SP and PC, and the general registers.
Other commands will search for values of specified bits of a word, or for
references to.an address within an address range, calculate 16-bit and 8-bit
offsets to an address and restart the running of the user’s program at any
address.
IOX input/Output Utility ‘Peripheral Driver-lox is a set of service routines
allowing sing&or double buffered I/O processing on an ASR-33 and/or a high-
speed paper tape reader and punch. This routme atlows the user to make
simple assembly language calls specifying devices and data forms to accom-
plish interrupt-controlled data transfer concurrent with execution of the run-
ning program. Multiple devices can be run simultaneously.
IOX frees the user from the details of dealing directly with the device and
allows development of programs which may be run under the direction of a
monitor with minimum modification.
IOX also provides some degree of real-time control by allowing pser programs
to be executed at priority leaIs at the completion of some device action or
data transfer.
MATH PACKAGE-A number of commonly used subroutines are available to
simplify programming. These routines are reentrant and relocatable to pro-
vide maximum flexibility. Arguments are treated as floating point numbers
with a signed 31.bit fraction and a signed 15-bit exponent. Subroutines sup-
plied include:
ADD
MULtiply
SUBtract
DlVide
SIN
cos
ATAN
FIX-FLOAT
FLOAT-FIX
NORmalize
(Integer MULtiply and DlVide are also supplied)
LOADiER%Two loaders are used:
l A Bootstrap loader loads the ABSolute loader and jumps to it.
l ABSolute loader loads PAL-1lA output, checks for checksum errors and ,’
jumps to a user program or halts when done.
CORE DUMP ROUTINES-Routines are provided which dump specified ’
ranges of core locations on paper tape in absolute format or on the tele-
printer in octal.

86
 CHAPTER 12 1. ’
THE OPERATOR’S CONSOiE
‘The PDP-11 Operator’s Console has been configured to achieve convenient
control of the system. Through switches and keys on the console, programs
or information can be manually inserted or modified. Also.indioator lamps on
, the console face display the status of the machine, the contents of t’he Bus
Address Register and the data at the output of the data paths. _

The consoie is shown in Figure 12-1.

kllilaliltlalll lcidial ll

Figure 12-1

CONSOLE ELEMENTS
The console has the following indicators and switches:
1. A bank.of 8 indicators, indicating the following conditions or oper-
‘ations: Fetch, Execute, Bus, .Run, Source, Destination and Address
(2 bits).
2. An l&bit Address Register Display
3. .A 18bit Data Register Display
4. An l&bit Switch Register
5. Control Switches:

a. LOAD ADDR (Load Address)

b. EXAM (Examine)
c. CONT (Continue)
,d. .ENABLE/ HALT
e. SIINST-S/CYCLE (Single Instruction/Single Cycle)
f. START
g. DEP (Deposit)

INDICATOR LIGHTS-The indicators signify specific machine functions,

operations, or states. Each is defined below.
1. Fetch-indicates that thdcentral processor is in the state of fetching
an instruction.
2. Execute-indicates that the central processor is in the state of
executing an instruction.

87
 3. Bus-indicates that a peripheral is controlling the bus. It is lit when
BBSY (Bus Busy) is asserted, unless the processor (which includes
the Console) is asserting BBSY.
4. Run-indicates that the processor is running. It monitors the cdntrol
flip-flop for the internal clock.
5. Source&ndicates that the central processor is. obtaining source
data except from an internal register.
6. Destination-indicates that the central processor is obtaining des
tination data (except from an internal register).
7. Address-identifies the source or destination address cycle of the
’ central processor, using two lights that are decoded zero, one, two,
or three. When references are made via the Unibus to the.addresses,
the lights tell the machine’s source or destination cycle. For an in-
ternal register reference, there is a “zeroth” addressing operation.
REGISTER DISPLAYS-The Operator’s Console has an l&bit Address Regis-
ter Display and a 16-bit Data Register Display. The Address Register Display
is tied directly to the output of an l&bit flip-flop register called the Bus
Address Register. This register displays the address of data examined or
deposited.
*The l&bit data register is divided on the face of the console by a line into
two 8-bit bytes. This register is tied to t,he output of the processor data paths
and will reflect the output of the processor adder.
SWITCH REGISTER-The PDP-ll/lO’and PDP-ll/PO can reference 216 bytes
addresses. However, the Unibus ,has expansion capability for 218 byte ad-
dresses. In order that the console can access the entire l&bit address
scheme, the switch register is 18 bits wide. These bits are assigned as 0
. through 17. The highest two are used only as addresses. A switch in the
“up” position is considered to have a “1” value and in the “down” position
to .have a “0” value. The condition of the 18 switches can be loaded into the
bus address register or any memory location by using the appropriate control
switches which are described below.
CONTROL SWITCHES-The switches listed in item 5 of the “Console
Elements” have these specific control functions:,
1. LOAD ADDR-transfers the contents of the l&bit switch register
into the bus address register.
2. EXAM-displays the contents of the location specified by the bus
address register.
3. DEP-deposits the contents of the low 16 bits of the switch register
into the address then displayed in the address register. (This switch
is actuated by raising it.)
4. ENABLE/HALT-allows or prevents running of programs. For a pro-
gram to run, theswitch must be in the ENABLE position (up). Placing
the switch in the HALT position (down) will halt the system.
5. START-starts executing a program when the ENABLE/HALT switch
is in the ENABLE position. When the START switch is depressed, it
asserts a system initialization signal; the system actually starts when
the switch is released. The processor will start executing at the
address which was last loaded by the LOAD ADDR key.
6. CONT-allows ‘the machine to continue without initialization from
whateyer state it was in when halted.
7. S/ INST-S/CYCLE-determines whether a single instruction or a
single bus cycle is performed when the CONT switch is depressed
while the machine is in the halt mode.

88
. .
When the system is running a program, the LOAD ADDR, EXAM, and DE-
POSIT functions are disabled to prevent disrupting the,program. When the
machine is to be halted, the ENABLE/HALT switch is thrown to the halt
position. The machine will halt either at the end of the current instruction,
or at the end of the current bus cycle, depending upon the position of the
S/ INST-S/CYCLE switch.

-OPERATING THE CONTROL SWlTCHES

When the PDP-11 has been halted, it is possible to examine and update bus
locations. To examine a specific location, the operator sets the switches of
the switch register to correspond to the location’s address. The operator then
presses LOAD ADDR, which will transfer the contents of the switch register
into the bus address register. The location of the address to be examined is
then displayed in the address register display. The operator then depresses
EXAM. The data in that location will appear in the data register display.
If the operator then depresses EXAM again, the, bus address register will be
incremented by 2 to the next word address and the new location will be
examined. In the PDP-11, the bus address register will always be pointing to
the data currently displayed in the data register,.The incrementation occurs
when the EXAM switch is depressed, and then the location is examined.
The examine function has been designed so that if LOAD ADDR and then *
EXAM are depressed, the address register will not be incremented. In this
case, the location reflected in the address register display is examined
directly. However, on the second (and successive) depressings of EXAM, the
bus address register is incremented. This will continue for successive de-
pressings as long as another control switch is not depressed.
If.the operator finds an incorrect entry in the data register, he can enter new
data there by putting it in the switch register and raising the DEP key. The
address register will not increment when this.data is deposited. Therefore.
when the operator presses the EXAM key, he can examine the data he just
deposited. However, when he presses EXAM again, the system will increment.
If the operator attempts to examine data from, or deposit data into, a non-
existent memory location, the “time out” feature will cause an error flag. The
data register will then reflect location 4, the trap location, for references to
nonexistent locations. To verify this condition, the operator should try to
deposit some number other than four in that location; if four is still indi-
cated, this would indicate that either nothing is assigned to that location, or
that whatever is assigned to that location is not working properly.
When doing consecutive examines or consecutive deposits, the address will
increment by 2, to successive word locations. However, if the programmer is
examining the fast registers (the “scratch pad” memory), the system only
increments by 1. The reason for this is that once the switch register is set
properly, the programmer can then use the four least significant bitsof the
switch register in examining fast memory registers from the front panel.
To start a PDP-11 program, the programmer loads the starting address of
the program in the switch register, depresses LOAD ADDR, and after ensur-
ing that the ENABLE/HALT switch is in the ENABLE position, depresses
ST;:. The program will ,start to run as soon as the START switch is re-

The Run indicator lamp is driven off the flip-flop that controls the clock.
Normally, when the system is running, not only will this light be on, .but the

89
other lights; (Fetch, Execute, Source, Destination; the.Address lights, and the
Address and Data registers) will be flickering. If the run light is on, and none
ofthe other indicators are flickering, the system could. be executing a “wait”
instruction which waits for an interrupt.
While in the halt mode;if the operator wishes to do a siqgle instruction, he
places the S/INST-S/CYCLE switch in the S/lNST position and depresses
CONT. When CONT is depressed, the console momentarily passes control to
the processor, allowing the machine to execute one instruction before regain- .
.ing control. Each time the CONT switch is depressed, the machine will
execute one instruction.
Similarly, if the operator wishes to have the machine perform a single bus
cycle, he places the S/INST-S/CYCLE switch in the S/CYCLE position and
presses CONT. The machine will then perform one complete bus cycle and
halt. The operator cannot do an,examine or deposit function at the end of a
single bus cycle unless the cycle ends coincidental with the end of an in-
struction. This prevents altering machine flow. Only when the machine is at
the end of an instruction and in the halt mode’can the examine or deposit
functions operate.
To start the machine running its program again, the operator places the
ENABLE/HALT switch in the ENABLE position, and depresses the CONT
switch. ‘.

90
 APPENDIX A-PDFIi INSTRUCTION REPERTOIRE

instruction Codes
Mnemonic Operation OP Code ZNCV Timing

DOUBLE OPERAND GROUP: OPR scr;dst .

MOV(B) MOVe (Byte) .lSSDD / d-0 2.3
(src) + (dst)
CMP(B) CoMPare (Byte) -2SSDD / r’ fI r/ 2.3;
/ Cm) - (dst)
BIT(B) Blt Test (Byte) .3SSDD 4 i-0 2.9*
(src) A (dst)
BIG(B) Blt Clear (Byte) 4SSDD ,,’ i-0 2.9
- (src) A (dst) + (dst)
BIS(B) Blt Set (Byte) .5SSDD / d-0 2.3
(src) v (dst) + (dst) ’
ADD ADD ’ 06SSDD r/ r’ / ,/ 2.3
suBtra(@ + (dst) -, (dst)
SUB 16SSDD / / / / 2.3
’ (dst)‘- (src) + (dst)

CONDITIONAL BRANCHES: Bxx lot

BR * BRanch (unconditionally) 0004xx 2.6-
lad+ (PC) -. -
BNE Bratch if Not Equal (Zero) 0010xx 2.6-
lot + (PC) if Z = 0
BEQ Branch if Equal (Zero). 0014xX 2.6-
loc+(PC)ifZ=l
BGE Branch‘if Greater or Equal (Zero) 002oxx 2.6 -
loc+(PC)ifNVV=O)
BLT Branch if Less Than (Zero) 0024Xx 2.6-
loc+(PC)ifNYV=l
BGT Branch if Greater Than (Zero) 003oxx 2.6 -
loc+(PC)ifZv(NVt~O)
BLE Branch if Less Than or Equal (Zero) 0034xX 2.6 -
loc+(PC)ifZv(NVV)=l
BPL Branch if PLUS ioooxx i.6- ’
loc+(PC)ifN=O b
BMI Branch if Minus 1004xX - 2.6 -
loc+(PC)ifN=l
BHI Branch if Higher 1010xx - 2.6 -,
lot-* (PC) ifCvZ=O
BLOS Branch if Lower-o? Same 1014xX - 2.6 -
loc+(PC)ifCvZ=l
BVC Branch if overflow Clear 102OXX - 2.6 -
- loc+(PC)ifV=O
BVS Branch if overflow Set 1024Xx \- 2.6 -
lot+ (PC) if V= 1
, BCC Branch if Carry Clear 103oxx - 2.6 -
(or BHIS) lot+ (PC) ifC=O
BCS Branch if Carry Set 1034xX - 2.6 -
(or BLO) lot + (PC) if C = 1

91
 SUBROUTINE CALL: JSR reg,dst
1JSR Jump to SubRoutine 004RDD - 4.2
(dstb (tmp), 0%) J
(PC> + (w3), Ww) + (PC)
SUBROUTINE RETURN: RTS reg
RTS ReTurn from Subroutine 00020R - 3.5
(w) + PC. t (w3)
SINGLE OPERAND GROUP: OPR dst
CLR(B) CLeaR (Byte) .050DD 1000 2.3
0 + (dst)
< COM(B) COMplement (Byte) -051DD 4 400 2.3
- (dst) + (dst)
INC(B) INCrement (Byte) -052DD
(dst) ,+ 1 + (dst)
DEC(B) DECrement (Byte) .053DD
(dst) 1 i 4 (dst)
NEG(B) NEGate (Bvte) .05,4DD
+‘(dstj + 1 + (dst)
ADW) ADd Carry (Byte) .055DD
(dst) + (Q --, (dst)
SBC(B) SuBtract Carry (Byte) .056DD
(dst) - (C) + (dst)
TST(B) TeST (Byte) .057DD
0 - (dst)
ROR(B) Rotate Right (Byte). .060DD
- rotate right 1 place with C
ROL(B) ROtate Left (Byte) .061DD
rotate left 1 place with C
A5R( B) Arithmetic Shift Right (Byte) .062DD r’ r’/ d 2.3”
shift right with sign extension
ASL( B) Arithmetic Shift Left (Byte) .063DD 4’4 r/ 4 2.3”
shift left with lo-order zero
JMP JUMP OOOlDD - 1.2
GM) + (PC)
SWAB SWAP Bytes 0003DD ,.‘t/OO 2.3 _
bytes of a word are exchanged
. CONDITION CODE OPERATORS: OPR 1.5
Condition Code Operators set or clear combinations of condition code bits.
Selected bits are set if S = 1 and cleared otherwise. Condition code bits cor-
responding to bits set as marked in the word below are set or cleared.
CONDITION CODE OPERATORS;

0 0 - 0 2 4SNZVC
1 I 1 I I I I i
15 54324 0

Thus SEC ‘= 000261 sets the C bit and has no effect on the other condition
code bits (CLC = 000241 clears the C Bit)
OPERATE GROUP: OPR-
HALT HALT 000000 l.8 I
processor stops; (RO) and the HALT address in lights
WAIT WAIT 00000 1 1.8
processor releases l&s, waits for interrupt

92
RTI ReTurn from Interrupt 000002 4 d / / 4.8
t Pa f (W
IOT Input/Output Trap 000004 d/v’/ 8.9
(PSI 4 s (PC) 4, (20) + (PC), (22) + (PS)
RESET RESET 000005 - 20 ms.
an INIT oulse;s issued bv the CP
EMT - EMulator Trap’ 104400--1104377
(PS) 4, (PC) 4, (30) + (PC), (32) + (PS)’
TRAP TRAP 104400-104777 4
PSI 4, (PC) 4 r (34) + (PC), (36) + (PSI

NOTATION:
1. for order codes -
word/byte bit, set for byte (+lOOOOO)
. SS-source field,
DD-destination field
XX-offset (8 bit)
2. for operations
A and,
V or,
not,
(’ contents of,
XOR
& “is pushed onto the processor stack”
-r “the contents
popped and becomes”
of the top of the processor stack is

+ “becomes”
3. for timing
* 0.4 ‘p.s less if not register mode
- 0.9 ps less if conditions for branch not met
0 1.2 r.~smore if addressing odd byte
/ (0.6 r~s additional in addressing odd bytes otherwise)
4. for condition codes
f ~;a;~;$=W

1 set

93
 /

94
 APPENDIX R-ADDRESSING SUhMlARY
ADDRESSING. ‘MODES.

S,C or’ dst

GENERAL REGISTER ADDRESSING

Timing (ws)
Mode Description SymboJii src dst
0 ’ register 00’ 00
1 registerdeferred $ R or (R) 1.5 1.4
2 auto increment CR) + 1.5 1.4
auto increment deferred @ (RI + 2.7 2.6
auto decrement - CR) 1.5 1.4
5 auto decrementdeferred @ - 09 2.7 2.6
6 indexed X CR) 2.7 2.6
7 indexed deferred @ X (RI or @ CR) 3.9 3.8

PC REGISTER ADDRESSING
Timing w)
Mode Description Symbolic dst
immediate E 1.4
f absolute F?YA 2.7 2.6
.6 relative 2.7 2.6
7 relative deferred @A 3.9 3.8

INSTRUCTION FORMATS

DOUBLE OPERAND GROUP: OPR src,dsl

CODE Y
I PI 4 11’15 I I I I I dsi, I I
15 12 11 6 5 0

CONDITIONAL BRANCHES: Sax lot (loc=bffset-2)+ +2)

SUBROUTlNE CALL: JSR r*g.dsl

SUBROUTINE RETURN; RTS rr9

0 , 0 I, 0 I I t 2 I I I 0 II 1 r*g II
15 3 2 0
.

SINGLE OPERAND GROUP: OPR drt

OP COOE
I
15 6 s, 0

CONDITION CODE OPERATORS:

0 0 0“ 2 4SNZVC
I I 1 I I 1 I I
15 5432t 0

.
 APPENDIX C-ADDRESS MAP .
0 USER DEVfCE INTERRUPT VECTOR
4 BUS ERROR, ILLEGAL INSTRUCTION, STACK OIKRFLOW TRAP
VECTOR
10 RESERVED INSTRUCTIONS TRAP VECTOR
14 CODE 000003 AND TRACE TRAP VECTOR
IOT INSTUCTION TRAP VECTOR
i: POWER FAIL INTERRUPT VECTOR
30 .EMT INSTRUCTION TRAP VECTOR
34 TRAP INSTUCTION TRAP VECTOR
40
44
SYSTEM SOFTWARE COMMUNQTION
50,
54 1
TELEPRINTER INTERRUPT YECTOR
z TELETYPE KEYBOARD AND LOW SPEED READER INTERRUPT
VECTOR
70 HIGH SPEED PAPER TAPE PUNCH INTERRUPT VECTOR
74 HIGH SPEED PAPER TAPE READER INTERRUPT VECTOR
.
.
. (additional interrupt vectors)
.
.
.
400
.
.
. PROCESSOR STACK
PROGRAM AND DATA
RESIDENT SYSTEM SOFTWARE
*
.
. (ABSOLUTE LOADER, BOOTSTRAP, I/O EXECUTIVE)
(end of implemented storage)
160000
.
. SMALL READ-ONLY STORAGE UNITS

.
. OTHER PERIPHERAL DEVICE REGISTERS

177550 HIGH SPEED READER AND PUNCH DEVICE STATUS AND BUFFER
REGISTERS

97
177560 TELElYP;- KEYBOARD AND PUNCH DEVCE STATUS AND EUFFER
REGISTER -’
.
.
.

*
177576
17-7600
.
RESERVEP FOR EXPANSION OF PROCESSOR REGISTERS
177677
177700
.
. GENERAL REGISTERS RO - R7

1777;6 CENTRAL PROCESSOR STATUS REGISTER (PS)

.
 .
APPENDIX l+UNlBUS OPERATIONS
There are ‘six bus operations: four to effect data transfers, on9 to transfer
bus control, and one to effect a program interrupt. This appendix describes
the signal interaction on the Unibus to perform these six operations.

DATA TRANSFERS
The four data transfers use the C lines coded as follows:
Cl co
0 0 DATI-DATa In
0 1 DATIP-DATa In, Pause
1 0 DATO-DATa Out
1' 1 DATOB-DATa Out, Byte

DATI AND DATIP-These two bus operations transfer data from a Slave
whose address is specified by A < 17:Ol > into the master. Both transfers
are made in words on D, < 15:OC >. In destructive read-out devices,
DATI commands a read-restore operation, while DATIP commands a read-
pause operation and the setting of a pause flag. DATlPs are to be followed
by a DATO or DATOB to effect a read-modify.write data exchange. In non _
destructive read-out devices, DATI and DATIP are treated identically. The
sequence of operations is as follows:
1. Master puts address on A, 0 or 1 on C, and waits 150 nanoseconds.
(75 nanoseconds for deskewing address + 75 nanoseconds for ad-
.
dress decoding).
2. Master asserts MSYN.
3. Slave decodes address, sees 0 er 1 on C, and MSYN and ldegins read
cycle (flip-flop register would simply gate flop outputs to bus).
4. Slave completes read cycle, outputs data to D lines, and asserts
SSYN. If the slave is a destructive read-out device, it now restores
data on a OATI: it sets a pause flag on a DATIP.
Figure D-l shows the signals for a DATI operation.
a
SIGNALS Al MASTER

ADtRESS-CONTROL AT ! I

jR
DATA
JT
MSYN
L .

SSYN IR

SIGNALS AT SLAVE

ADDRESS-CONTROL IR 1

DATA p T I
I
MSYN IR I
SSYN
I
MEMORY CYCLE

T= SIGNAL AS TFIANSMITTEO
,
R*SIGNAL AS RECEIVED

Figure D-l DATI Operation

+

99
 l

5. Master sees SSYN and waits 75 nanoseconds, minimum (data des-

kewing + internal gating deskewing).
6. Master.strobes data, drops MSYN, and waits 75 nanoseconds (des-
kew address).
7. Masfer drops A and C and waits for SSYN to fall.
8. Slave sees MSYN fall and drops SSYN and D lines.
9. Master sees SSYN fall, signaling-end of bus operation.
NOTES:
1. Step 1 of the next data transfer may begin at step 7 of the current DATI or
DATI P.
‘2. Step 2 of the next data transfer may begin at step 9 of the current DATI or
DATIP.

MT0 AND DATOB-These two bus operations transfer data out of the mas-
ter to the slave. DATO is used to transfer a word to the address specified
by A < 17:Ol >. The slave ignores A00 and the data appear. on D < 15:00 >.
DATOB is used to transfer a byte to the add<ess specified by A < 17:00 >.
A00 = 0 indicates the low byte and data appears on D < 07:OO >; A00 f 1
indicates high byte and data appears on D < 15:08 >. The sequence of op
eration is as follows:
1. Master puts address on A, data on D, 2 or 3 on C, and waits 150
nanoseconds (75 nanoseconds for deskewing address + 75 nano-
seconds for address decoding).
. 2. Master asserts MSYN.
3. Slave decodes address, sees 2 or 3 on C and MSYN and strobes in
word or byte. When slave has taken data, it asserts SSYN. If the slave
is a destructive read-out device and its pause flag is set (by DATIP),
slave begins write cycle; if .not, slave must first do a read cycle to
clear the memory cell and then a write.
4. Master sees SSYN and drops MSYN and waits 75 nanosecondi (des-
kew address). e
Master drops A, D, and C, and waits for SSYN to fall.
2: Slave sees MSYN fall and drops SSYN.
7. Master sees SSYN fall, signaling end of bus operation.
Figure D-2 shows the signals for a DATO operation.
DATO

SIGNALS AT MAST&
.

DATA IT I
I
MSYN IT

IR I
SSYN

SIGNALS AT SLAVE
I
ADDRESS-CONTROL IR

BATA IR 1

MSYN IR I

SSYN * ‘T

MEMORY CYCLE -1

Figure D-2 DATO Operation

100
 \
NOTES:
1. Step 1 of the next-data transfer-may begin at step 5 of the current DATO or
DATOB.
2. Step .2 of the next data transfer may begin at step 7 of the current DATO or
DATOB.
\
PTR-PRIORITY TRANSFER
This bus operation is used to pass control of the bus from one master to
another. The steps which fbllow are performed simultaneously with the data
transfers:
0. Current master device always has BBSY asserted.
1. Requesting device asserts its assigned BR line.
2. Processor sees BR asserted, determines which BR is highest, and
asserts the corresponding BG line if the processor’s Current Prior@
level’allow that level of bus request.
3. Each device that receives the BG passes it on to the next . device
unless it itself is requesting.
4. The BG is propagated a!ong the priority chain until it reaches the
first .requesting device. This device becomes selected as next bus
master and does not allow the BG to pass to succeeding devices.
5. The selected device asserts SACK and drops its BR, .and waits for
BBSY,,BG, and SSYN to drop. .
6. The processor sees SACK and drops BG.
The device which is current master completes its data transfers,
. 7.
drops BBSY, and ceases to be bus master. .
8. The selected device sees BG, BBSY, and SSYN drop, becomes bus
master, asserts BBSY, drops SACK, and begins data transfers.
9. New master relinquishes bus control, either to the processor or to a
requesting device, by dropping BBSY at the end of its last bus op
eration. This is termed a passive release of bus control.
NOTES:
1. NPR bus requests ore handled as above.
2. Processor defers action on BR <7:4> until last bus cycle of an instruction
execution or interrupt sequence, NPR is acted upon immediately.
. 3. Processor becomes bus master and asserts BBSY whenever it sees BBSY = 0
end no other’ device has been selected or is being selected as next bus master.
4. Processor will not execute step 2 if SACK is asserted. See note 2 under INTR.

Figure D-3 shows the signals for a PTR operation. I

PTR
SIGNALS AT DEVICE

BR JT . I

BG II3 I

SACK IT

SIGNALS AT PROCESSOR

BR JR I

‘0G .
SACK IR

T= SIGNAL AS TRANSMITTED
R * SIGNAL AS RECEIVED

Figure D-3 PTR Operation

101
INTR-lNTerRupt
This bus operation is initiated by a master immediately after receiving bus
control to effect a program interrupt in the processor. It proceeds as follows:
0. Device has become bus master via PTR and BBSY is as&ted.
Master puts interrupt vector address on D and asserts INTR.
:: Processor sees INTR and waits 75 nanoseconds (deskewdata).
Processor strobes data and asserts SSYN.
:: Master sees SSYN, drops INTR, -D, and BBSY. The master has now
relinquised bus control directly to the processor. The INTR sequence
is termed an active release of bus control.
- 5. Processor sees INTR drop and drops SSYN and. enters interrupt
sequence to update PC and PS.

1. Step 1 must be made simultaneously with step 8 of FTR; that-is, SACK cannot
be dropped until INTR i,s asserted.
2.’ When the processor’sees SACK drop. it waits 75 nanoseconds (deskew). If, at
that time, INTR = 1. the processor issues no SG’s until the interrupt sequence
is complete.

Figure D-4 shows the signals for the INTR operation.

~
SIGNALS Al MASTER
BBSY T
JR

DATA IT 1

INTR IT

SMN rR 1

-%NALS AT FIWCESSOR
BBSY -i UT

MTA -R I

INTR jR 1

SBYN IT

T. SIGNAL AS TRANSMITTED
R * SIGNAL AS RECEIVED

Figure D-4 INTR Operation

GENERAL NOTES ON THE BUS OPEliATlONS

1. A master device doing a read-modify-write operation must keep bus
control BBSY asserted for both bus transactions (both the DATIP
. and the DATO or DATOB). This is the one case where an NPR request
will not be honored between bus transactions.
2. A device becomes master by the PTR operation. If ‘the request for
bus control was made on the NPR line, bus control must be released
passively (by dropping BBSY). Bus control is then passed either back
to the processor to execute the next bus cycle of the instruction or
to another device requesting on the NPR tine. If a device becomes
master via a,BR request line, control may be passed actively back
to the processor by using the INTR operation or passively (by drop
pitig BBSY); If control is given up actively, only NPR~requests will‘be
honored during the interrupt sequence of.updating the PC and PS.
If control is given up passively, control may pass either to +he
processor to fetch the next instruction or to an NPR requesting
device.

io3
 ’ I

The PDP-11provides Direct Device Addressing. All memory-and

devices on the Unibus are directly addressable and may be op
erated upon by all computer instructions. Direct device to device
transfers are possible.

,
.

104