ARM

The first encounter

Authors: Nemanja Perovic, [email protected] Prof. Dr. Veljko Milutinovic, [email protected]

1

What Is ARM?
Advanced RISC Machine

First RISC microprocessor for commercial use

Market-leader for low-power and cost-sensitive embedded applications

ARM Powered Products

Features
Architectural simplicity

which allows
Very small implementations which result in Very low power consumption

The History of ARM

Developed at Acorn Computers Limited, of Cambridge, England, between 1983 and 1985

Problems with CISC:

Slower then memory parts Clock cycles per instruction

The History of ARM (2)

Solution the Berkeley RISC I:
Competitive
Easy to develop (less than a year) Cheap Pointing the way to the future

ARM Architecture
Typical RISC architecture:
Large uniform register file
Load/store architecture Simple addressing modes Uniform and fixed-length instruction fields

ARM Architecture (2)

Enhancements:
Each instruction controls the ALU and shifter
Auto-increment and auto-decrement addressing modes Multiple Load/Store Conditional execution

ARM Architecture (3)

Results:
High performance Low code size

Low power consumption

Low silicon area

Pipeline Organization
Increases speed most instructions executed in single cycle Versions:

3-stage (ARM7TDMI and earlier)

5-stage (ARMS, ARM9TDMI) 6-stage (ARM10TDMI)

10

Pipeline Organization (2)

3-stage pipeline: Fetch Decode - Execute Three-cycle latency, one instruction per cycle throughput
i n s t r u c t i o n

Fetch
i+1

Decode Fetch
i+2

Execute Decode Fetch Execute Decode Execute

cycle

t+1

t+2

t+3

t+4

11

Pipeline Organization (3)

5-stage pipeline:

Stages:
Fetch Decode Execute Buffer/data

Reduces work per cycle => allows higher clock frequency

Separates data and instruction memory => reduction of CPI (average number of clock Cycles Per Instruction)

Write-back

12

Pipeline Organization (4)

Pipeline flushed and refilled on branch, causing execution to slow down Special features in instruction set eliminate small jumps in code to obtain the best flow through pipeline

13

Operating Modes
Seven operating modes:

User
Privileged:
System (version 4 and above) FIQ IRQ Abort

exception modes

Undefined
Supervisor
14

Operating Modes (2)

User mode:

Exception modes:

Normal program execution mode

System resources unavailable Mode changed by exception only

Entered upon exception

Full access to system resources Mode changed freely

15

Exceptions
Exception Reset Undefined instruction Software interrupt Mode Supervisor Undefined Supervisor Priority 1 6 6 IV Address 0x00000000 0x00000004 0x00000008

Prefetch Abort
Data Abort Interrupt Fast interrupt

Abort
Abort IRQ FIQ

5
2 4 3

0x0000000C
0x00000010 0x00000018 0x0000001C

Table 1 - Exception types, sorted by Interrupt Vector addresses

16

ARM Registers
31 general-purpose 32-bit registers
16 visible, R0 R15

Others speed up the exception process

17

ARM Registers (2)

Special roles:

Hardware
R14 Link Register (LR): optionally holds return address for branch instructions R15 Program Counter (PC)

Software
R13 - Stack Pointer (SP)

18

ARM Registers (3)

Current Program Status Register (CPSR)

Saved Program Status Register (SPSR)

On exception, entering mod mode:

(PC + 4) LR
CPSR SPSR_mod PC IV address

R13, R14 replaced by R13_mod, R14_mod

In case of FIQ mode R7 R12 also replaced
19

ARM Registers (4)

System & User
R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 (PC) CPSR

FIQ
R0 R1 R2 R3 R4 R5 R6 R7_fiq R8_fiq R9_fiq R10_fiq R11_fiq R12_fiq R13_fiq R14_fiq R15 (PC) CPSR SPSR_fiq

Supervisor
R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_svc R14_svc R15 (PC) CPSR SPSR_svc

Abort
R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_abt R14_abt R15 (PC) CPSR SPSR_abt

IRQ
R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_irq R14_irq R15 (PC) CPSR SPSR_irq

Undefined
R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13_und R14_und R15 (PC) CPSR SPSR_und
20

Instruction Set
Two instruction sets:

ARM
Standard 32-bit instruction set

THUMB
16-bit compressed form Code density better than most CISC

Dynamic decompression in pipeline

21

ARM Instruction Set

Features:

Load/Store architecture
3-address data processing instructions

Conditional execution
Load/Store multiple registers Shift & ALU operation in single clock cycle

22

ARM Instruction Set (2)

Conditional execution:

Each data processing instruction prefixed by condition code

Result smooth flow of instructions through pipeline

16 condition codes:
MI PL negative positive or zero
overflow

EQ NE

equal not equal unsigned higher or same

unsigned lower

HI LS

unsigned higher

GT LE

signed greater than

unsigned lower or same

signed greater than or equal signed less than

signed less than or equal

always special purpose

CS

VS

GE

AL

CC

VC

no overflow

LT

NV

23

ARM Instruction Set (3)

ARM instruction set

Data processing instructions

Data transfer instructions Block transfer instructions Branching instructions Multiply instructions Software interrupt instructions

24

Data Processing Instructions

Arithmetic and logical operations

3-address format:

Two 32-bit operands (op1 is register, op2 is register or immediate) 32-bit result placed in a register

Barrel shifter for op2 allows full 32-bit shift within instruction cycle
25

Data Processing Instructions (2)

Arithmetic operations:

ADD, ADDC, SUB, SUBC, RSB, RSC

Bit-wise logical operations:

AND, EOR, ORR, BIC

Register movement operations:

MOV, MVN

Comparison operations:

TST, TEQ, CMP, CMN

26

Data Processing Instructions (3)

Conditional codes

+
Data processing instructions

+
Barrel shifter

=
Powerful tools for efficient coded programs
27

Data Processing Instructions (4)

e.g.:

if (z==1) R1=R2+(R3*4)

compiles to
EQADDS R1,R2,R3, LSL #2

( SINGLE INSTRUCTION ! )

28

Data Transfer Instructions

Load/store instructions

Used to move signed and unsigned Word, Half Word and Byte to and from registers
Can be used to load PC (if target address is beyond branch instruction range)
LDR
LDRH LDRSH LDRB LDRSB

Load Word
Load Half Word Load Signed Half Word Load Byte Load Signed Byte

STR
STRH

Store Word
Store Half Word

STRSH Store Signed Half Word STRB STRSB Store Byte Store Signed Byte
29

Block Transfer Instructions

Load/Store Multiple instructions (LDM/STM) Whole register bank or a subset copied to memory or restored with single instruction
Mi LDM R0 R1 R2 Mi+14 Mi+15 R14 R15
30

Mi+1 Mi+2

STM

Swap Instruction
Exchanges a word between registers Two cycles but single atomic action Support for RT semaphores
R0
R1 R2

R7 R8

R15

31

Modifying the Status Registers

Only indirectly
R0 R1 MRS R7 CPSR SPSR MSR R8

MSR moves contents from CPSR/SPSR to selected GPR

MRS moves contents from selected GPR to CPSR/SPSR Only in privileged modes

R14 R15

32

Multiply Instructions
Integer multiplication (32-bit result)

Long integer multiplication (64-bit result)

Built in Multiply Accumulate Unit (MAC)

Multiply and accumulate instructions add product to running total

33

Multiply Instructions
Instructions:
MUL MULA
UMULL UMLAL SMULL SMLAL

Multiply Multiply accumulate

Unsigned multiply Unsigned multiply accumulate Signed multiply Signed multiply accumulate

32-bit result 32-bit result

64-bit result 64-bit result 64-bit result 64-bit result

34

Software Interrupt
SWI instruction

Forces CPU into supervisor mode Usage: SWI #n

31 28 27 24 23 0

Cond

Opcode

Ordinal

Maximum 224 calls Suitable for running privileged code and making OS calls
35

Branching Instructions
Branch (B): jumps forwards/backwards up to 32 MB Branch link (BL): same + saves (PC+4) in LR Suitable for function call/return Condition codes for conditional branches
36

Branching Instructions (2)

Branch exchange (BX) and Branch link exchange (BLX): same as B/BL + exchange instruction set (ARM THUMB) Only way to swap sets

37

Thumb Instruction Set

Compressed form of ARM

Instructions stored as 16-bit, Decompressed into ARM instructions and Executed

Lower performance (ARM 40% faster)

Higher density (THUMB saves 30% space) Optimal interworking (combining two sets) compiler supported
38

THUMB Instruction Set (2)

More traditional:

No condition codes
Two-address data processing instructions

Access to R0 R8 restricted to

MOV, ADD, CMP

PUSH/POP for stack manipulation

Descending stack (SP hardwired to R13)

39

THUMB Instruction Set (3)

No MSR and MRS, must change to ARM to modify CPSR (change using BX or BLX) ARM entered automatically after RESET or entering exception mode Maximum 255 SWI calls

40

The Next Step

New ARM Cortex family of processors

New NEON media and signal processing extensions Thumb-2 blended 16/32-bit instruction set for performance and low power Improved Interrupt handling

41

Summary
Adoption of ARM technology has increased efficiency and lowered costs ARM is the worlds leading architecture today

3 billion ARM Powered chips and counting

42

References
www.arm.com

ARM Limited ARM Architecture Reference Manual, Addison Wesley, June 2000
Trevor Martin The Insiders Guide To The Philips ARM7Based Microcontrollers, Hitex (UK) Ltd., February 2005 Steve Furber ARM System-On-Chip Architecture (2nd edition), Addison Wesley, March 2000

43

The End

Authors: Nemanja Perovic, [email protected] Prof. Dr. Veljko Milutinovic, [email protected]

44