Assembly Language Programming

CS208

Assembly Language

Assembly language allows us to use convenient abbreviations (called mnemonics) for machine language operations and memory locations. Each assembly language is specific to a particular hardware architecture, and can only be used on a machine of that architecture. An assembly language program must be translated into machine code before it can be executed. The program that tells the computer how to perform the translation is called an assembler.

Assembly Language

When a processor chip is designed, it is designed to understand and execute a set of machine code instructions (OpCodes) unique to that chip. One step up from machine code is assembly code. Each machine code instruction is given a mnemonic (name), so that it is easier for human beings to write code. There is generally a one-to-one correspondence between the assembly languages mnemonic instructions and the machine language numeric instructions.

Model Assembly Instructions

Our assembly language instructions have two parts:

The operation code specifies the operation the computer is to carry out (add, compare, etc) An address that allows the instruction to refer to a location in main memory

The CPU runs each instruction in the program, starting with instruction 0, using the fetch-decode-execute cycle.

Review of the Fetch-Decode-Execute Cycle

The CPU fetches the next instruction from the address contained in the Program Counter and places the instruction in the Instruction Register.
When a program starts, the program counter contains 0, so the instruction at address 0 is fetched.

Immediately after the instruction fetch, the CPU adds 1 word to the contents of the Program Counter, so that it will contain the address of the next sequential instruction.

Review of the Fetch-Decode-Execute Cycle

The CPU decodes the instruction in the Instruction Register and determines what operations need to be done and what the address is of any operand that will be used. The specified operation is executed (add, compare, etc).
After execution of the instruction has been completed the cycle starts all over again (unless the instruction terminates the program).

CPU

CPU Registers

The Instruction Register (IR) contains the actual instruction which is currently being executed by the CPU.

The Status Register records the result of comparing the contents of register A with the contents of register B.
The Program Counter (PC) contains the address of the next instruction to be executed by the program.

CPU Registers

Registers A & B hold the operands for each arithmetic operation (ie. the values on which the operation will be performed). After the operation has been carried out, the result is always stored in Register B. Therefore, after an arithmetic operation has been performed, the second operand is no longer stored in Register B, because it has been overwritten by the result of the operation.

CPU Registers

After a comparison has been done, the Status Register will hold a code that stores the results of the comparison.
The results are coded as follows:
-1 0 1 if (A < B) if (A = B) if (A > B)

Model Assembly Language Instructions

Operation STP LDA What it means to the CPU Stop the program Load register A with value from a specified memory location

LDB
STR

INP
PNT

Load register B with value from a specified memory location Store register B value to a specified memory location Store data input by user to a specified memory location Print the value stored in a specified memory location to the screen

Model Assembly Language Instructions

Operation
JLT JGT JEQ JMP CMP

What it means to the CPU

Jump if less than (Status register = -1) to a specified memory location Jump if greater than (Status register = 1) to a specified memory location Jump if equal (Status register = 0) to a specified memory location Unconditional jump to a specified memory location Compare register A to register B and set Status Register value

Model Assembly Language Instructions

Operation
ADD SUB MUL DIV MOD

What it means to the CPU

Add (register A + register B) and store sum in register B Subtract (register A - register B) and store difference in register B Multiply (register A * register B) and store product in register B Divide for quotient (register A/register B) and store quotient in register B Divide for remainder (register A/register B) and store remainder in register B

Steps to write Assembly Programs

Create C++ Program (only the statements between the { and } brackets are needed) Translate each C++ statement to the equivalent assembly statement(s)

Number the assembly language program starting from 0

Replace memory names by memory address numbers of empty memory cell Resolve jumps (replace with number of memory cell jumping to)

C++ to Assembly Language

Statement #include Assembly equivalent none

void main()
const int, double, char

none
value in memory cell address of memory cell

cin
cout assignment (=) val3 = val1 + val2

INP
PNT LDA Val1 LDB Val2 ADD STR Val3 STP

Sample Program #1
Program #1. Write an assembly language program that will get a number as input from the user, and output its square to the user.

Sample Program #1
Step 1: Write an algorithm to describe the steps needed to solve our problem.

Algorithm:
1. Input a number and store it in memory. 2. Compute the square by multiplying the number times itself. 3. Output the results.

Sample Program #1
Step 2: Write the C++ code
{ int Number , Square; cout << "Enter a number: "; cin >> Number; Square = Number * Number ; cout << Square; }

Sample Program #1
Step 3: Translate C++ code to assembly { cout << "Enter a number: "; cin >> Number;

INP number

square = number * number;

LDA number LDB number MUL STR square

PNT square STP

cout << Square; }

Sample Program #1
Step 4: Number assembly code lines starting from 0

0 1 2 3 4 5 6

INP LDA LDB MUL STR PNT STP

number number number square square

Sample Program #1
Step 5: Replace memory names with empty memory locations after STP
0 1 2 3 4 5 6 INP LDA LDB MUL STR PNT STP number 7 number 7 number 7 square 8 square 8

Sample Program #1
Step 6: Final Assembly code

INP LDA LDB MUL STR PNT STP

7 7 7 8 8

Running the Code on the Model Assembler

Type the code on the previous slide into a file (use Notepad).
Save the file as sample1.txt in the same directory as the assembler.exe file Double click assembler.exe Press ENTER Type the filename sample1.txt Press r to run

Before running the code, the screen will look like this:

Your Assembly Code

After running the code, the screen will look like this:

Results of Program Run

C++ Decisions to Assembly Language

C++ Statement if ( Num < 10 ) cout << Num; Assembly equivalent LDA Num LDB Ten CMP [test condition] JLT Then block address JMP address of statement after Then block PNT Num [Then block]

C++ Decisions to Assembly Language

Pascal Statement if ( Num < 10 ) cout << Num; else cout << 0;

Assembly equivalent LDA Num LDB Ten CMP [Test condition] JLT Then block address PNT Zero [Else block] JMP Address of statement after Then block PNT Num [Then block]

Sample Program #2
Program #2. Write an assembly program that will get a number from the user, and determine if the number is evenly divisible by 5.
Output zero (false) if the number is NOT evenly divisible by 5 or one (true) if the number IS evenly divisible.

Sample Program #2
Step 1: Write the algorithm to describe the steps needed to solve our problem.
1. Read in a number and store it in memory. 2. Determine if input number is evenly divisible by 5. 2.1 Divide input number by 5 to get the remainder. 2.2 Compare remainder to 0. If remainder equals 0, the number is evenly divisible. If the remainder does not equal 0, the number NOT evenly divisible. 3. Output the results 3.1 If evenly divisible, output 1. 3.2 If NOT evenly divisible, output 0.

Sample Program #2
Step 2: Write the C++ code

{
const int Zero = 0; const int One = 1; const int Five = 5; int number, rem; cout << "Enter number: "; cin >> number; rem = number % Five; if (rem = Zero) cout << One; else cout << Zero; }

Sample Program #2
Step 3: Translate C++ code to Assembly
cout << "Enter number: "; cin >> number;

INP number

rem = number % Five ;

LDA number LDB Five MOD STR rem

Sample Program #2
Step 3: continued if (rem = Zero) cout << One; else cout << Zero; LDA Zero LDB rem CMP JEQ then block address PNT Zero else block JMP address after then block PNT One then block STP

Sample Program #2
Step 4: Number assembly code lines starting from 0 0 1 2 3 4 5 6 7 8 9 10 11 12 INP LDA LDB MOD STR LDA LDB CMP JEQ PNT JMP PNT STP number number Five

rem Zero rem

condition

then block address Zero else block address after then block One then block

Sample Program #2
Step 5: Replace names by cell numbers after STP

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

INP LDA LDB MOD STR LDA LDB CMP JEQ PNT JMP PNT STP 5 0 1

number 16 number 16 Five 13 rem 17 Zero 14 rem 17

then block address Zero 14 address after then block One 15 Five Zero One number rem

Sample Program #2
Step 5: Replace jumps by instruction numbers

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

INP LDA LDB MOD STR LDA LDB CMP JEQ PNT JMP PNT STP 5 0 1

16 16 13 17 14 17

address of then block 11 14 else block address after then block 12 15 then block

Sample Program #2
Step 6: Final Assembly code

INP LDA LDB MOD STR LDA LDB CMP JEQ PNT JMP PNT STP 5 0 1

16 16 13 17 14 17 11 14 12 15

C++ Decisions to Assembly Language

C++ Statement while ( Num < 10 ) cout << Num; Assembly equivalent LDA Num LDB Ten CMP test condition JLT to While Block JMP to stmt after While Block PNT Num While Block JMP to test condition stmt after While Block

Sample Program #3
Program #3. Write a program to display a count by fives to 100.

Sample Program #3
Step 1: Write an algorithm to describe the steps needed to solve our problem 1. Set Count to start at 0

2. While Count is less than 100

2.1 Add 5 to the Count and store the sum back into the Count

2.2 Display the Count

Sample Program #3
Step 2: Write C++ code { const int Five = 5; const int Hundred = 100; int Count = 0; while (Count < Hundred) { Count = Count + 5; cout >> Count; } }

Sample Program #3
Step 3: Translate C++ code to assembly
while (Count < Hundred) { Count = Count + 5; cout >> Count; } LDA Count LDB Hundred CMP JLT to while block JMP to stmt after while block LDB Five ADD STR Count PNT Count JMP to test condition test condition

while block

stmt after while block

Sample Program #3
Step 4: Number assembly code lines from 0
0 LDA Count test condition 1 LDB Hundred 2 CMP 3 JLT to while block 4 JMP to stmt after while block 5 LDB Five while block 6 ADD 7 STR Count 8 PNT Count 9 JMP to test condition 10 STP stmt after while block

Sample Program #3
Step 5: Replace memory names with empty memory locations after STP
0 LDA Count 13 1 LDB Hundred 12 2 CMP 3 JLT to while block 4 JMP to stmt after while block 5 LDB Five 11 6 ADD 7 STR Count 13 8 PNT Count 13 9 JMP to test condition 10 STP 11 5 Five 12 100 Hundred 13 0 Count

Sample Program #3
Step 5: Replace memory names with empty memory locations after STP
0 LDA 13 test condition 1 LDB 12 2 CMP 3 JLT to while block 5 4 JMP stmt after while block 10 5 LDB 11 while block 6 ADD 7 STR 13 8 PNT 13 9 JMP to test condition 0 10 STP stmt after while block 11 5 12 100 13 0

Sample Program #3
Final Assembly Code: LDA 13 LDB 12 CMP JLT 5 JMP 10 LDB 11 ADD STR 13 PNT 13 JMP 0 STP 5 100 0