Central Processing Unit

• Introduction

• General Register Organization

• Stack Organization

• Instruction Formats

• Addressing Modes

• Data Transfer and Manipulation

• Program Control

• Reduced Instruction Set Computer

Ch8. Central Processing Unit 1 /48

 Introduction

Major Components of CPU

• Storage Components
Registers
Flags

• Execution (Processing) Components

Arithmetic Logic Unit(ALU)
Arithmetic calculations, Logical computations, Shifts/Rotates

• Transfer Components
Bus

• Control Components
Control Unit Register
File ALU

Control Unit

Ch8. Central Processing Unit 2 /48

 Registers
 In Basic Computer, there is only one general purpose register, the
Accumulator (AC)
 In modern CPUs, there are many general purpose registers
 It is advantageous to have many registers
– Transfer between registers within the processor are relatively fast
– Going “off the processor” to access memory is much slower

 How many registers will be the best ?

Ch8. Central Processing Unit 3 /48

 General Register Organization

General Register Organization

Input
Clock

R1
R2
R3
R4
R5
R6
R7
Load
(7 lines)
SELA { MUX MUX } SELB
3x8
A bus B bus
decoder

SELD
OPR ALU

Output

Ch8. Central Processing Unit 4 /48

 Control

Operation of Control Unit

The control unit
Directs the information flow through ALU by
- Selecting various Components in the system
- Selecting the Function of ALU
Example: R1  R2 + R3
[1] MUX A selector (SELA): BUS A  R2
[2] MUX B selector (SELB): BUS B  R3
[3] ALU operation selector (OPR): ALU to ADD
[4] Decoder destination selector (SELD): R1  Out Bus
3 3 3 5
Control Word SELA SELB SELD OPR

Encoding of register selection fields

Binary
Code SELA SELB SELD
000 Input Input None
001 R1 R1 R1
010 R2 R2 R2
011 R3 R3 R3
100 R4 R4 R4
101 R5 R5 R5
110 R6 R6 R6
111 R7 R7 R7

Ch8. Central Processing Unit 5 /48

 Control

ALU Control
Encoding of ALU operations OPR
Select Operation Symbol
00000 Transfer A TSFA
00001 Increment A INCA
00010 ADD A + B ADD
00101 Subtract A - B SUB
00110 Decrement A DECA
01000 AND A and B AND
01010 OR A and B OR
01100 XOR A and B XOR
01110 Complement A COMA
10000 Shift right A SHRA
Examples of ALU Microoperations 11000 Shift left A SHLA
Symbolic Designation
Microoperation SELA SELB SELD OPR Control Word
R1  R2  R3 R2 R3 R1 SUB 010 011 001 00101
R4  R4  R5 R4 R5 R4 OR 100 101 100 01010
R6  R6 + 1 R6 - R6 INCA 110 000 110 00001
R7  R1 R1 - R7 TSFA 001 000 111 00000
Output  R2 R2 - None TSFA 010 000 000 00000
Output  Input Input - None TSFA 000 000 000 00000
R4  shl R4 R4 - R4 SHLA 100 000 100 11000
R5  0 R5 R5 R5 XOR 101 101 101 01100
6 /48
Ch8. Central Processing Unit
 Stack Organization

Register Stack Organization

Stack
- Very useful feature for nested subroutines, nested interrupt services
- Also efficient for arithmetic expression evaluation
- Storage which can be accessed in LIFO
- Pointer: SP
- Only PUSH and POP operations are applicable stack Address
Flags 63
Register Stack FUL EMPTY
L

Stack pointer
4
SP C 3
Push, Pop operations 6 bits B 2
A 1
/* Initially, SP = 0, EMPTY = 1, FULL = 0 */ 0
PUSH POP DR
SP  SP + 1 DR  M[SP]
M[SP]  DR SP  SP  1
If (SP = 0) then (FULL  1) If (SP = 0) then (EMPTY  1)
EMPTY  0 FULL  0
Ch8. Central Processing Unit 7 /48
 Stack Organization

Memory Stack Organization

1000
Memory with Program, Data, Program
PC (instructions)
and Stack Segments
Data
AR (operands)
SP 3000
stack
3997
3998
3999
4000
4001
- A portion of memory is used as a stack with a
processor register as a stack pointer Stack grows
In this direction
- PUSH: SP  SP - 1
M[SP]  DR
- POP: DR  M[SP]
SP  SP + 1
- Most computers do not provide hardware to check stack overflow (full
stack) or underflow (empty stack)  must be done in software

Ch8. Central Processing Unit 8 /48

 Stack Organization

Reverse Polish Notation

• Arithmetic Expressions: A + B
A+B Infix notation
+AB Prefix or Polish notation
AB+ Postfix or reverse Polish notation
- The reverse Polish notation is very suitable for stack
manipulation
• Evaluation of Arithmetic Expressions
Any arithmetic expression can be expressed in parenthesis-free
Polish notation, including reverse Polish notation

(3 * 4) + (5 * 6)  34*56*+

6
4 5 5 30
3 3 12 12 12 12 42
3 4 * 5 6 * +

Ch8. Central Processing Unit 9 /48

 Processor Organization
 In general, most processors are organized in one of 3 ways

– Single register (Accumulator) organization

» Basic Computer is a good example
» Accumulator is the only general purpose register

– General register organization

» Used by most modern computer processors
» Any of the registers can be used as the source or destination
for computer operations

– Stack organization
» All operations are done using the hardware stack
» For example, an OR instruction will pop the two top elements
from the stack, do a logical OR on them, and push the result on
the stack

Ch8. Central Processing Unit 10 /48

 Instruction Format

Instruction Format
• Instruction Fields
OP-code field - specifies the operation to be performed
Address field - designates memory address(es) or a processor register(s)
Mode field - determines how the address field is to be interpreted (to
get effective address or the operand)
• The number of address fields in the instruction format
depends on the internal organization of CPU

• The three most common CPU organizations:

Single accumulator organization:
ADD X /* AC  AC + M[X] */
General register organization:
ADD R1, R2, R3 /* R1  R2 + R3 */
ADD R1, R2 /* R1  R1 + R2 */
MOV R1, R2 /* R1  R2 */
ADD R1, X /* R1  R1 + M[X] */
Stack organization:
PUSH X /* TOS  M[X] */
ADD
Ch8. Central Processing Unit 11 /48
 Instruction Format

Three, and Two-address Instructions

• Three-Address Instructions
Program to evaluate X = (A + B) * (C + D) :
ADD R1, A, B /* R1  M[A] + M[B] */
ADD R2, C, D /* R2  M[C] + M[D] */
MUL X, R1, R2 /* M[X]  R1 * R2 */

- Results in short programs

- Instruction becomes long (many bits)

• Two-Address Instructions
Program to evaluate X = (A + B) * (C + D) :

MOV R1, A /* R1  M[A] */

ADD R1, B /* R1  R1 + M[A] */
MOV R2, C /* R2  M[C] */
ADD R2, D /* R2  R2 + M[D] */
MUL R1, R2 /* R1  R1 * R2 */
MOV X, R1 /* M[X]  R1 */

Ch8. Central Processing Unit 12 /48

 Instruction Format

One, and Zero-address Instructions

• One-Address Instructions
- Use an implied AC register for all data manipulation
- Program to evaluate X = (A + B) * (C + D) :
LOAD A /* AC  M[A] */
ADD B /* AC  AC + M[B] */
STORE T /* M[T]  AC */
LOAD C /* AC  M[C] */
ADD D /* AC  AC + M[D] */
MUL T /* AC  AC * M[T] */
STORE X /* M[X]  AC */
• Zero-Address Instructions
- Can be found in a stack-organized computer
- Program to evaluate X = (A + B) * (C + D) :
PUSH A /* TOS  A */
PUSH B /* TOS  B */
ADD /* TOS  (A + B) */
PUSH C /* TOS  C */
PUSH D /* TOS  D */
ADD /* TOS  (C + D) */
MUL /* TOS  (C + D) * (A + B) */
POP X /* M[X]  TOS */
Ch8. Central Processing Unit 13 /48
 Addressing Modes

Addressing Modes

• Addressing Modes

* Specifies a rule for interpreting or modifying the

address field of the instruction (before the operand
is actually referenced)

* Variety of addressing modes

- to give programming flexibility to the user

- to use the bits in the address field of the
instruction efficiently

Ch8. Central Processing Unit 14 /48

 Addressing Modes

Types of Addressing Modes

• Implied Mode
Address of the operands are specified implicitly
in the definition of the instruction
- No need to specify address in the instruction
- EA = AC, or EA = Stack[SP]
- Examples from Basic Computer
CLA, CME, INP

• Immediate Mode
Instead of specifying the address of the operand,
operand itself is specified
- No need to specify address in the instruction
- However, operand itself needs to be specified
- Sometimes, require more bits than the address
- Fast to acquire an operand

Ch8. Central Processing Unit 15 /48

 Addressing Modes

Types of Addressing Modes

Register Mode
Address specified in the instruction is the register address
- Designated operand need to be in a register
- Shorter address than the memory address
- Saving address field in the instruction
- Faster to acquire an operand than the memory addressing
- EA = IR(R) (IR(R): Register field of IR)

• Register Indirect Mode

Instruction specifies a register which contains
the memory address of the operand
- Saving instruction bits since register address
is shorter than the memory address
- Slower to acquire an operand than both the
register addressing or memory addressing
- EA = [IR(R)] ([x]: Content of x)

• Autoincrement or Autodecrement Mode

- When the address in the register is used to access memory, the
value in the register is incremented or decremented by automatically
Ch8. Central Processing Unit 16 /48
 Addressing Modes

Types of Addressing Modes

• Direct Address Mode
Instruction specifies the memory address which
can be used directly to access the memory
- Faster than the other memory addressing modes
- Too many bits are needed to specify the address
for a large physical memory space
- EA = IR(addr) (IR(addr): address field of IR)

• Indirect Addressing Mode

The address field of an instruction specifies the address of a memory
location that contains the address of the operand
- When the abbreviated address is used large physical memory can be
addressed with a relatively small number of bits
- Slow to acquire an operand because of an additional memory access
- EA = M[IR(address)]

Ch8. Central Processing Unit 17 /48

 Addressing Modes

Types of Addressing Modes

• Relative Addressing Modes
The Address fields of an instruction specifies the part of the address
(abbreviated address) which can be used along with a designated
register to calculate the address of the operand
- Address field of the instruction is short
- Large physical memory can be accessed with a small number of
address bits
- EA = f(IR(address), R), R is sometimes implied

3 different Relative Addressing Modes depending on R;

* PC Relative Addressing Mode (R = PC)
- EA = PC + IR(address)
* Indexed Addressing Mode (R = IX, where IX: Index Register)
- EA = IX + IR(address)
* Base Register Addressing Mode
(R = BAR, where BAR: Base Address Register)
- EA = BAR + IR(address)

Ch8. Central Processing Unit 18 /48

 Addressing Modes

Addressing Modes - examples

Address Memory
200 Load to AC Mode
PC = 200
201 Address = 500
R1 = 400 202 Next instruction

XR = 100
399 450
AC 400 700

Addressing Effective Content

Mode Address of AC 500 800
Direct address 500 /* AC  (500) */ 800
Immediate operand - /* AC  500 */ 500
Indirect address 800 /* AC  ((500)) */ 300 600 900
Relative address 702 /* AC  (PC+500) */ 325
Indexed address 600 /* AC  (RX+500) */ 900
Register - /* AC  R1 */ 400
Register indirect 400 /* AC  (R1) */ 700 702 325
Autoincrement 400 /* AC  (R1)+ */ 700
Autodecrement 399 /* AC  -(R) */ 450
800 300
Ch8. Central Processing Unit 19 /48
 Data Transfer and Manipulation

Data Transfer Instructions

• Typical Data Transfer Instructions
Name Mnemonic
Load LD
Store ST
Move MOV
Exchange XCH
Input IN
Output OUT
Push PUSH
Pop POP

• Data Transfer Instructions with Different Addressing Modes

Assembly
Mode Convention Register Transfer
Direct address LD ADR AC M[ADR]
Indirect address LD @ADR AC  M[M[ADR]]
Relative address LD $ADR AC  M[PC + ADR]
Immediate operand LD #NBR AC  NBR
Index addressing LD ADR(X) AC  M[ADR + XR]
Register LD R1 AC  R1
Register indirect LD (R1) AC  M[R1]
Autoincrement LD (R1)+ AC  M[R1], R1  R1 + 1
Autodecrement LD -(R1) R1  R1 - 1, AC  M[R1]
Ch8. Central Processing Unit 20 /48
 Data Transfer and Manipulation

Data Manipulation Instructions

• Three Basic Types: Arithmetic instructions
Logical and bit manipulation instructions
Shift instructions
• Arithmetic Instructions
Name Mnemonic
Increment INC
Decrement DEC
Add ADD
Subtract SUB
Multiply MUL
Divide DIV
Add with Carry ADDC
Subtract with Borrow SUBB
Negate(2’s Complement) NEG

• Logical and Bit Manipulation Instructions • Shift Instructions

Name Mnemonic Name Mnemonic
Clear CLR Logical shift right SHR
Complement COM Logical shift left SHL
AND AND Arithmetic shift right SHRA
OR OR Arithmetic shift left SHLA
Exclusive-OR XOR Rotate right ROR
Clear carry CLRC Rotate left ROL
Set carry SETC Rotate right thru carry RORC
Complement carry COMC Rotate left thru carry ROLC
Enable interrupt EI
Disable interrupt DI
Ch8. Central Processing Unit 21 /48
 Flag Processor Status Word
 In Basic Computer, the processor had several (status) flags – 1 bit
value that indicated various information about the processor’s state – E,
FGI, FGO, I, IEN, R
 In some processors, flags like these are often combined into a register
– the processor status register (PSR); sometimes called a processor
status word (PSW)
 Common flags in PSW are
– C (Carry): Set to 1 if the carry out of the ALU is 1
– S (Sign): The MSB bit of the ALU’s output
Status Flag Circuit
– Z (Zero): Set to 1 if the ALU’s output is all 0’s
A B
– V (Overflow): Set to 1 if there is an overflow 8 8
c7
c8 8-bit ALU
F7 - F0
V Z S C
F7
Check for 8
zero output
F 22 /48
Ch8. Central Processing Unit
 Program Control

Program Control Instructions

+1
In-Line Sequencing (Next instruction is fetched
from the next adjacent location in the memory)
PC
Address from other source; Current Instruction,
Stack, etc; Branch, Conditional Branch,
Subroutine, etc

• Program Control Instructions

Name Mnemonic
Branch BR
Jump JMP
Skip SKP
Call CALL
Return RTN
Compare(by  ) CMP
Test(by AND) TST
* CMP and TST instructions do not retain their
results of operations (  and AND, respectively).
They only set or clear certain Flags.
Ch8. Central Processing Unit 23 /48
 Program Control

Conditional Branch Instructions

Mnemonic Branch condition Tested condition
BZ Branch if zero Z=1
BNZ Branch if not zero Z=0
BC Branch if carry C=1
BNC Branch if no carry C=0
BP Branch if plus S=0
BM Branch if minus S=1
BV Branch if overflow V=1
BNV Branch if no overflow V=0
Unsigned compare conditions (A - B)
BHI Branch if higher A>B
BHE Branch if higher or equal A  B
BLO Branch if lower A<B
BLOE Branch if lower or equal A  B
BE Branch if equal A=B
BNE Branch if not equal AB
Signed compare conditions (A - B)
BGT Branch if greater than A>B
BGE Branch if greater or equalA  B
BLT Branch if less than A<B
BLE Branch if less or equal A  B
BE Branch if equal A=B
BNE Branch if not equal AB
Ch8. Central Processing Unit 24 /48
 Program Control

Subroutine Call and Return

• Subroutine Call Call subroutine
Jump to subroutine
Branch to subroutine
Branch and save return address
• Two Most Important Operations are Implied;
* Branch to the beginning of the Subroutine
- Same as the Branch or Conditional Branch

* Save the Return Address to get the address

of the location in the Calling Program upon
exit from the Subroutine

• Locations for storing Return Address CALL

SP  SP - 1
• Fixed Location in the subroutine (Memory) M[SP]  PC
• Fixed Location in memory PC  EA
• In a processor Register
• In memory stack RTN
- most efficient way PC  M[SP]
SP  SP + 1

Ch8. Central Processing Unit 25 /48

 Program Control

Program Interrupt -types of interrupts

External interrupts
External Interrupts initiated from the outside of CPU and Memory
- I/O Device → Data transfer request or Data transfer complete
- Timing Device → Timeout
- Power Failure
- Operator
Internal interrupts (traps)
Internal Interrupts are caused by the currently running program
- Register, Stack Overflow
- Divide by zero
- OP-code Violation
- Protection Violation
Software Interrupts
Both External and Internal Interrupts are initiated by the computer HW.
Software Interrupts are initiated by the executing an instruction.
- Supervisor Call → Switching from a user mode to the supervisor mode
→ Allows to execute a certain class of operations
which are not allowed in the user mode
Ch8. Central Processing Unit 26 /48
 Program Control

Interrupt Procedure
Interrupt Procedure and Subroutine Call
- The interrupt is usually initiated by an internal or
an external signal rather than from the execution of
an instruction (except for the software interrupt)

- The address of the interrupt service program is

determined by the hardware rather than from the
address field of an instruction

- An interrupt procedure usually stores all the

information necessary to define the state of CPU
rather than storing only the PC.

The state of the CPU is determined from;

Content of the PC
Content of all processor registers
Content of status bits
Many ways of saving the CPU state
depending on the CPU architectures
Ch8. Central Processing Unit 27 /48
 RIS
C
RISC : Historical Background
IBM System/360, 1964
 The real beginning of modern computer architecture

 Distinction between Architecture and Implementation

 Architecture: The abstract structure of a computer

seen by an assembly-language programmer

Hardware
-program
Compiler
High-Level Instruction
Language Hardware
Set
Architecture
Hardware

Implementation

 Continuing growth in semiconductor memory and microprogramming

 A much richer and complicated instruction sets
 CISC(Complex Instruction Set Computer)

Ch8. Central Processing Unit 28 /48

 Arguments Advanced at That Time
 Richer instruction sets would simplify compilers
 Richer instruction sets would alleviate the software crisis
– move as much functions to the hardware as possible

 Richer instruction sets would improve architecture quality

Ch8. Central Processing Unit 29 /48

 RISC

Architecture Design Principles - In 70’s

• Large microprograms would add little or nothing

to the cost of the machine
 Rapid growth of memory technology
 Large General Purpose Instruction Set

• Microprogram is much faster than the machine instructions

 Microprogram memory is much faster than main memory
 Moving the software functions into
microprogram for the high performance machines

• Execution speed is proportional to the program size

 Architectural techniques that led to small program
 High performance instruction set

• Number of registers in CPU has limitations

 Very costly
 Difficult to utilize them efficiently

Ch8. Central Processing Unit 30 /48

 RISC

Four Modern Architectures in 70’s

DEC Xerox Intel

IBM 370/168 VAX-11/780 Dorado iAPX-432
Year 1973 1978 1978 1982
# of instrs. 208 303 270 222
Control mem. size 420 Kb 480 Kb 136 Kb 420 Kb
Instr. size (bits) 16-48 16-456 8-24 6-321
Technology ECL MSI TTL MSI ECL MSI NMOS VLSI
Execution model reg-mem reg-mem stack stack
mem-mem mem-mem mem-mem
reg-reg reg-reg
Cache size 64 Kb 64 Kb 64 Kb 64 Kb

Ch8. Central Processing Unit 31 /48

 Complex Instruction Set Computer

 These computers with many instructions and addressing modes

came to be known as Complex Instruction Set Computers (CISC)

 One goal for CISC machines was to have a machine language

instruction to match each high-level language statement type

Ch8. Central Processing Unit 32 /48

 Variable Length Instructions
 The large number of instructions and addressing modes led CISC
machines to have variable length instruction formats
 The large number of instructions means a greater number of bits to
specify them
 In order to manage this large number of opcodes efficiently, they were
encoded with different lengths:
– More frequently used instructions were encoded using short opcodes.
– Less frequently used ones were assigned longer opcodes .
 Also, multiple operand instructions could specify different addressing
modes for each operand
– For example,
» Operand 1 could be a directly addressed register,
» Operand 2 could be an indirectly addressed memory location,
» Operand 3 (the destination) could be an indirectly addressed
register.
 All of this led to the need to have different length instructions in different
situations, depending on the opcode and operands used
Ch8. Central Processing Unit 33 /48
 Complex Instruction Set Computer

 The problems with CISC computers are

– The complexity of the design may slow down the processor,

– The complexity of the design may result in costly errors in the
processor design and implementation,
– Many of the instructions and addressing modes are used rarely, if
ever

Ch8. Central Processing Unit 34 /48

 RISC

Summary: Criticisms on CISC

High Performance General Purpose Instructions
- Complex Instruction
→ Format, Length, Addressing Modes
→ Complicated instruction cycle control due to the complex
decoding HW and decoding process
- Multiple memory cycle instructions
→ Operations on memory data
→ Multiple memory accesses/instruction
- Microprogrammed control is necessity
→ Microprogram control storage takes
substantial portion of CPU chip area
→ Semantic Gap is large between machine
instruction and microinstruction

- General purpose instruction set includes all the features

required by individually different applications
→ When any one application is running, all the features
required by the other applications are extra burden to the application
Ch8. Central Processing Unit 35 /48
 Reduced Instruction Set Computers
 In the late ‘70s and early ‘80s there was a reaction to the
shortcomings of the CISC style of processors
 Reduced Instruction Set Computers (RISC) were proposed as an
alternative
 The underlying idea behind RISC processors is to simplify the
instruction set and reduce instruction execution time

 RISC processors often feature:

– Few instructions
– Few addressing modes
– Only load and store instructions access memory
– All other operations are done using on-processor registers
– Fixed length instructions
– Single cycle execution of instructions
– The control unit is hardwired, not microprogrammed
Ch8. Central Processing Unit 36 /48
 RISC

Characteristics of Initial RISC Machines

IBM 801 RISC I MIPS

Year 1980 1982 1983
Number of
instructions 120 39 55
Control memory
size 0 0 0
Instruction
size (bits) 32 32 32
Technology ECL MSI NMOS VLSI NMOS VLSI
Execution model reg-reg reg-reg reg-
reg

Ch8. Central Processing Unit 37 /48

 Registers

 By simplifying the instructions and addressing modes, there is space

available on the chip or board of a RISC CPU for more circuits than
with a CISC processor
 This extra capacity is used to
– Pipeline instruction execution to speed up instruction execution
– Add a large number of registers to the CPU

Ch8. Central Processing Unit 38 /48

 Pipelining

 A very important feature of many RISC processors is the ability to

execute an instruction each clock cycle

 This may seem nonsensical, since it takes at least once clock cycle
each to fetch, decode and execute an instruction.

 It is however possible, because of a technique known as pipelining

– We’ll study this in detail later

 Pipelining is the use of the processor to work on different phases of

multiple instructions in parallel

Ch8. Central Processing Unit 39 /48

 Pipelining
 For instance, at one time, a pipelined processor may be
– Executing instruction it
– Decoding instruction it+1
– Fetching instruction it+2 from memory

 So, if we’re running three instructions at once, and it takes an

average instruction three cycles to run, the CPU is executing an
average of an instruction a clock cycle

 As we’ll see when we cover it in depth, there are complications

– For example, what happens to the pipeline when the processor
branches

 However, pipelined execution is an integral part of all modern

processors, and plays an important role

Ch8. Central Processing Unit 40 /48

 Registers
 By having a large number of general purpose registers, a processor
can minimize the number of times it needs to access memory to load
or store a value
 This results in a significant speed up, since memory accesses are
much slower than register accesses
 Register accesses are fast, since they just use the bus on the CPU
itself, and any transfer can be done in one clock cycle
 To go off-processor to memory requires using the much slower
memory (or system) bus
 It may take many clock cycles to read or write to memory across the
memory bus
– The memory bus hardware is usually slower than the processor
– There may even be competition for access to the memory bus by other
devices in the computer (e.g. disk drives)
 So, for this reason alone, a RISC processor may have an advantage
over a comparable CISC processor, since it only needs to access
memory
– for its instructions, and
– occasionally to load or store a memory value
Ch8. Central Processing Unit 41 /48
 RISC

Utilizing RISC Registers –Register Window

<Weighted Relative Dynamic Frequency of HLL Operations>

Machine- Memory
Dynamic Instruction Reference
Occurrence Weighted Weighted
Pascal C Pascal C Pascal C
ASSIGN 45 38 13 13 14 15
LOOP 5 3 42 32 33 26
CALL 15 12 31 33 44 45
IF 29 43 11 21 7 13
GOTO 3
Other 6 1 3 1 2 1

The procedure (function) call/return is the most time-consuming

operations in typical HLL programs

Ch8. Central Processing Unit 42 /48

 RISC

Call-Return Behavior

Call-return behavior as a function of nesting depth and time

Ch8. Central Processing Unit 43 /48

 RISC

Register Window Approach

• Observations
- Weighted Dynamic Frequency of HLL Operations
 Procedure call/return is the most time consuming operations

- Locality of Procedure Nesting

 The depth of procedure activation fluctuates
within a relatively narrow range

- A typical procedure employs only a few passed

parameters and local variables
• Solution
- Use multiple small sets of registers (windows),
each assigned to a different procedure

- A procedure call automatically switches the CPU to use a different

window of registers, rather than saving registers in memory

- Windows for adjacent procedures are overlapped

to allow parameter passing

Ch8. Central Processing Unit 44 /48

 RISC

Circular Overlapped Register Windows

Ch8. Central Processing Unit 45 /48

 RISC

Overlapped Register Windows

R73 R15 Common
R10 to D and A
R6
4
R63 R73
Local to D
R58 R64
R57 R63
R58 Common to C and D
R48
R47 Proc D R57
Local to C
R42 R48
R41 R47
R42 Common to B and C
R32
R31 Proc C R41
Local to B
R26 R32
R25 R31
R16 Common to A and B
R2
R15 Proc B6 R25
Local to A
R10 R16
R9 R9 R15
Common to all Common to A and D
procedures R10
R0 R0 Proc A
Global
Ch8. Central Processing Unit registers 46 /48
 Overlapped Register Windows
 There are three classes of registers:
– Global Registers
» Available to all functions
– Window local registers
» Variables local to the function
– Window shared registers
» Permit data to be shared without actually needing to copy it
 Only one register window is active at a time
– The active register window is indicated by a pointer
 When a function is called, a new register window is activated
– This is done by incrementing the pointer
 When a function calls a new function, the high numbered registers of
the calling function window are shared with the called function as the
low numbered registers in its register window
 This way the caller’s high and the called function’s low registers
overlap and can be used to pass parameters and results
Ch8. Central Processing Unit 47 /48
 Overlapped Register Windows

 In addition to the overlapped register windows, the processor has

some number of registers, G, that are global registers
– This is, all functions can access the global registers.

 The advantage of overlapped register windows is that the processor

does not have to push registers on a stack to save values and to
pass parameters when there is a function call
– Conversely, pop the stack on a function return
 This saves
– Accesses to memory to access the stack.
– The cost of copying the register contents at all

 And, since function calls and returns are so common, this results in a
significant savings relative to a stack-based approach

Ch8. Central Processing Unit 48 /48