Unit III Control Unit Design
Unit III Control Unit Design
Unit III Control Unit Design
DESIGN
•INTRODUCTION
•CONTROL TRANSFER
•FETCH CYCLE
•INSTRUCTION INTERPRETATION AND
EXECUTION
•HARDWIRED CONTROL
•MICROPROGRAMMED CONTROL
INTRODUCTION
• Important component of CPU is the controller
Microprogram
TERMINOLOGY
- Program stored in memory that generates all the control signals required
to execute the instruction set correctly
- Consists of microinstructions
Microinstruction
- Contains a control word and a sequencing word
Control Word - All the control information required for one clock cycle
Sequencing Word - Information needed to decide
the next microinstruction address
- Vocabulary to write a microprogram
Dynamic Microprogramming
- Computer system whose control unit is implemented with
a microprogram in WCS
- Microprogram can be changed by a systems programmer or a user
TERMINOLOGY
Sequencer (Micro program Sequencer)
A Micro program Control Unit that determines
the Microinstruction Address to be executed
in the next clock cycle
- In-line Sequencing
- Branch
- Conditional Branch
- Subroutine
- Loop
- Instruction OP-code mapping
Controller functions
1.Fetch and instruction sequencing (fetch cycle)-Generates
control signal to fetch instruction from memory and the
sequence of operations involved in processing an instruction
2.Instruction interpretation and execution (execution cycle)-Tasks
involved are
» Interpreting the operand addressing mode implied in the
operation code and fetching the operands
» Sequencing the successive micro operations on the data
path to execute the operation code specified in the
instruction
3.Interrupt processing (interrupt cycle)-Process interrupt. Tasks
are
» Suspend execution of current program
» Save context
» Set PC to start address of interrupt handler routine
» Process interrupt
» Restore context and continue interrupted program
Interaction Between Data and
Control Units
Digital Systems can be partitioned in a Datapath and Control Unit
Control Unit
CONTROL TRANSFER
• CPU fetches and
executes each instruction
CPU Cycles
of the program
• Successively goes F E Comments
through fetch, execute
and interrupt cycles 0 0 Not used
• Two flip-flops marked F
and E identify each of 1 0 Fetch Cycle
these 3 cycles
0 1 Execute
• Two general operations Cycle
• Instruction control transfer
• Program control transfer
1 1 Interrupt
Cycle
Instruction Cycle (with Interrupts) -
State Diagram
9
Instruction Control Transfer
• PC holds the address of the instruction to be executed
• During fetch cycle PC is incremented to hold the address of
the next instruction
• Assuming each instruction has a length of one word, then
PC incremented by one, that is,PC <= PC + 1;
• Length of a variable length instruction needs to be specified
in some bits of the operation code field of instruction
• Transfer of control to non sequential instruction occurs
during execution cycle of conditions branch with specified
condition satisfied or by an unconditional branch instruction
Example
• Jump X (Unconditional branch to X) or
• Jump ZX (Branch to X if the result of last arithmetic
operation is zero)
• X stored in the operand field of instruction register
Instruction Control Transfer
• For most instructions: fetch instruction, fetch
operands, execute, store.
• An abstract view of the implementation of the
MIPS subset showing the major functional units
and the major connections between them
• Missing Multiplexers, and some Control lines for read and write.
PROGRAM CONTROL TRANSFER
• While program P1 is running ,CPU is under control of
P1
• Transfer of program control may occur in 2 situations
1. There is an instruction in P1 which calls a subroutine
denoted as program P2
JMPSUBX, where X denotes starting address of P2
2. While an instruction in P1 is getting fetched and
executed ,an interrupt flag gets set, where X denotes
starting address of ISR(P2). After execution of ISR
restore the execution which is saved in a temporary
register
• Prior to transferring control to P2 ,the CPU status of
P1 is stored in stack (push and pop operation)
FETCH CYCLE
• The instruction whose address is determined by
the PC is obtained from the memory
• Loaded into the IR.
• The PC is then incremented to point to the next
instruction and switch over to execution cycle
• CPU enters fetch cycle if F=1 and E=0
C01: MAR PC
C02C03:MDR M(MAR);PC PC+1
C04:IR MDR
C05:F 0; E 1
Sequence of micro operations of fetch cycle
INSTRUCTION INTERPRETATION
AND EXECUTION
• Instruction loaded into Instruction Register (IR)
• Processor interprets instruction and performs
required actions
– Different for each instruction
• e.g. ADD X, R1 - add the contents of location X
to Register 1 , result in R1
• C61: MAR <- (IRaddress)
• C62: MDR <- (memory)
• C63: R1 <- R1 + (MDR)
• Note no overlap of micro-operations
Input-Output behavior of a control
unit
• Set of signals Cin1 input to the controller from the datapath includes those
from the operation code of IR, various flag registers of ALU,interrupt , etc.
dictate the flow of control for the algorithm implemented on the datapath for
an operation code
• Other input signals Data Path
Cin2 are those control
which control the operation
of the controller itself, Cin1 Cout1
such as start, stop, clock, etc.
• Cout1 fed at
appropriate Cin2 Cout2
point on the datapath Control
to sequence and control Unit
the micro operations
associated with operation code
• Other control signal Cout2 cover signals such as busy, operation complete,
etc transmitted to other control units
Implementation of Control Unit
COMPARISON OF CONTROL UNIT
IMPLEMENTATIONS
Control Unit Implementation
Combinational Logic Circuits (Hard-wired)
Control Data
Memory IR Status F/Fs
Microprogram
M Control Data
e
m
o IR Status F/Fs
r
y
C Control C
Next Address Storage C
M M D P CPU
Generation A (-program D
Logic s
R memory) R }
HARDWIRED CONTROL
• The controller is designed as a sequential logic circuit which
generates the specific sequence of control signals as its primary
output
C0=C01: MAR PC
C1=C02C03:MDR M(MAR);PC PC+1
C2=C04:IR MDR
C3=C05:F 0; E 1
Sequence of micro operations of fetch cycle
• Its output is decoded to derive the desired control signals in the
given sequence.
Controller Block Diagram
Detailed control sequencing
• ADD (R3), R1 High-level execution sequence:
• Fetch instruction.
• Fetch memory operand.
• Perform addition.
• Store result in R1.
Detailed control sequencing:
1: PCout, MARin, Read, Clear Y, Set carry-in to ALU, Add,
Zin
2: Zout, PCin, WMFC
3: MDRout, IRin
4: R3out, MARin, Read
5: R1out, Yin, WMFC
6: MDRout, Add, ZinZout, R1in, END
Methods for systematic design of
hardwired control logic
• Sequence counter method: To design
controller of moderate complexity
• Delay element method: Depends on the
use of clocked delay elements for
generating the sequence of control signals
• State table method: Employs the
algorithmic approach to sequential circuit
design using classical state table method
Controller specification for the Fetch
Sequence
• Address of next instruction is in PC (s01)
• Content of PC loaded into MAR (c01) ,s02
• Address (MAR) is placed on address bus
• Control unit issues READ command
• Result (data from memory) appears on data bus
• Data from data bus copied into MDR (Co2),s03
• PC incremented by 1 (in parallel with data fetch from memory)
(c03),s03
• Data (instruction) moved from MDR to IR (c04), S04
• MDR is now free for further data fetches (F 0 ; E 1) (c05) , s05
Sequence counter method
• Step 1: Identify the distinct phases in the flowchart. Employ log p
number of flip flops to handle p number distinct phases
Start R 1
Modulo –K
End S Reset counter
0
Clock
Reset
…..
Step 2: Identify the maximum number of distinct
steps, k, in each of the phases .Employ a mod k I/k Decoder
counter to generate control signals for each of the
k steps ……
Step 3 : design a combinational logic circuit to c1c2 ck
generate the sequence of control signals to
control the micro operations of each phase
Delay element method
Control unit based on delay element method
for the fetch cycle
Begin Ci,j
C01: MAR PC;
t0 (d1=t1-t0) Delay block Di
D1 C02: MDR M(MAR),
t1
c03: PC PC+1 ; (d2=t2-t1) Ci+1,j’
D2
t2 C04: IR MDR; (d3=t3-t2)
D3
t3 C05: F 0; E 1; (d4=t4-t3)
D4
t4
Execute
State table method
• The state Si (i=1,2…) has been marked
above each block of the flowchart.
• This the state of the controller which
generates the control signals to control the
micro operations in the data path
Ij C’i
Combinational logic
FFs
Steps to design the Control
structure realizing flowchart
• State Assignment: States are assigned as
s1,s2,s3…;each such assignment specifies a
particular state of the controller at the specific
time step. State table derived from state
assignment
• State Minimization: A set of states {Sa,Sb,…Sc}
can be merged to a single state S’ if Si & Sj is
pair wise compatible
• State Encoding: State variables are defined and
states are encoded in terms of state variables
Hard-wired Control Unit-
advantages
1. Minimizes the average number of clock
cycles needed per instruction
2. occupies a relatively small area
(typically 10%) of the CPU chip area
3. High efficiency in terms of operation
speed
4. is to minimize cost of the circuit
Problems With Hard Wired
Designs
• Complex sequencing & micro-operation
logic
• Difficult to design and test
• Inflexible design
• Large design turn around time for complex
design
• Difficult to add new instructions
Tasks Done By Micro
programmed Control Unit
• Microinstruction sequencing
• Microinstruction execution
• Must consider both together
Microprogrammed Control
The concept of micro programmed control, employ the following
steps:
1. Any instruction to be executed on a CPU can be broken down
into a set of sequential micro operations – each specifying a
RTL operation on the data path. The set of micro operations to
be executed on the RTL components at any time step is
referred as microinstructions.
2. The sequence of control signals necessary to execute the
sequential microinstructions stored in ROM called control ROM
3. To implement an instruction on the data path , the control
signals stored in the ROM can be accessed
4. The control signals read from the ROM are used to control the
micro operations associated with a microinstruction to be
executed at any time step
5. The address of the next micro instruction is generated
6. The steps 3,4 and 5 are repeated till the set of sequential
microinstructions associated with the instruction is executed
Microprogrammed Control
(continued)
Control
inputs Status signals from datapath
Next-address
generator
Sequencer
Control address
register
Control address
Address
Control
memory
(R OM)
Data
Microinstruction
MICROINSTRUCTION FORMAT
Information in a Microinstruction
- Control Information
- Sequencing Information
- Constant
Information which is useful when feeding into the system
Field Encoding
Field A Field B
Field A Field B
2 bits 6 bits
2 bits 3 bits
2x4 6 x 64
2x4 3x8 Decoder Decoder
Decoder Decoder
Decoder and
1 of 4 1 of 8 selection logic
Functional Encoding
• The micro operations executed in a data
path of a CPU may be categorized under
different function names such as Shift
function, Add function, Logical functions,
Input-Output, etc
• Multiple control bits associated to control a
specified function
Resource encoding
• The CPU data path consists of a set of
interconnected RTL components.
• Each of these components or a subset of
such components viewed as a resource.
• If control signals associated with such a
resource are mutually exclusive, then they
can be encoded in a single control field.
Microinstruction Encoding
Direct Encoding
Microinstruction Encoding
Indirect Encoding
Micro instruction execution
Two basic actions
1. Generating next microinstruction address and
fetching micro program instruction from control
memory by this address
2. Executing the micro operations controlled by
different signals encoded in various control
fields of the microinstruction
1. Decoding and applying control signals on the CPU data
path
2. Executing the intended micro operations controlled by the
signals
3. Storing the output in the destination register specified in
the micro operation on the CPU data path
Sequencing
MICROINSTRUCTION SEQUENCING
Instruction code
Mapping
logic
Subroutine
register
Control address register (SBR)
(CAR)
Incrementer
select a status
bit
Microoperations
Branch address
First level
-Vertical format Micro program
Second level
-Horizontal format Nano program
- Interprets the microinstruction fields, thus converts a vertical
microinstruction format into a horizontal
nano instruction format.
11 bits
Control memory
2048 x 8
Microinstruction (8 bits)
Nanomemory address
Nanomemory
256 x 200