Programming Format and Global Structure

General Format

- though the assembler language is freeform like C, we adhere to a standard format for clarity
- general form of the lines of a program are denoted by fields

LABEL       FIELD2      OPERAND(S)  ; comment
|           |           |           |
           Convenient tab stops

• indentation separates fields
• align fields for readability
• blank lines are ignored
• ";" denotes the remaining text on the line is a comment
• attempt to make useful comments on every line of code
• assembler is case insensitive, VARIABLE is equivalent to variable
• number literals default to base-10
• literals prefixed with a 0 and postfixed with a h are interpreted as hexadecimal

LABEL Field

Form:

<label>     <pseudo-instruction/assembler directive>

-or-

<label>:    <instruction>

• <LABEL> corresponds to a variable name, procedure name, or a branching label -->
• The first character may not be a digit, the allowable characters are letters, numbers, and the special characters @, $, _, and ?.
• Max label length is 31 characters
• Labels prefixed with the special characters are used for the system, by convention
• There are some language reserved labels, like dataseg

FIELD2 Field

Contains:

1. pseudo-operations (assembler directives)
   non-executable used by assembler
   not translated into machine code
   UPPERCASE convention
2. executable operations (opcodes)
   converted to machine code
   1 opcode per line with 0+ operands
   lowercase convention

Global Structure

segments in an assembly program: STACK, DATA, CODE

Form:

<name>      SEGMENT     <option1>       <option2>  ...  <optionN>
.
.
.
<name> ENDS

The Stack Segment

• procedure calls
• interrupts
• runtime stack

example:

StackSeg    SEGMENT     STACK       para        'STACK'
            DB          32          DUP("STACK**")
StackSeg    ENDS

Option 1 - STACK
Must contain STACK so it can be combined with other stack segments by the assembler.

Option 2 - para
Sets the kind of address at which hte physical address begins.
para is default

Option 2 - 'STACK'
Assigns the stack a name used by a debugger
Optional, not always used.

Shortcut definition

.STACK N

Where N is the number of bytes to reserve. All defaults are used.

The Data Segment

example:

DatSeg      SEGMENT     para    'DATA'
.
.
.
DatSeg      ENDS

Shortcut definition

.DATA

The Code Segment

example:

CodeSeg     SEGMENT     para    'CODE'
            ASSUME  ss: StackSeg    ...
.
.
.
CodeSeg     ENDS

• every program must have a code segment
• other sements may be nested in the code segment, but this is bad form
• combinations may not exceed 64k
• may be subdivided into 0+ procedures

There are 14 segment registers on the 8086

1. cs - pointer to active code segment
2. ss - pointer to active stack segment
3. ds - pointer to active data segment
4. es - pointer to active extended segment

The last line ends with:

        END     <startlabel>

example:

program:
.
.
.
        END     program

Defining Data and Variables

The CODE Segment

Interrupts/DOS Calls

Return to CSC2252

Copyright © 1999, Bruce Lin