PROWAREtech

articles » current » assembly » x86 » tutorial » page-06

Intel IA-32 Assembly Tutorial - A Guide to the Basics of x86 Assembly - Page 06

Language Elements (Data-Related Operators and Directives, More on Instructions: JMP and LOOP).

Data-Related Operators and Directives

OFFSET operator returns the offset of a data label. The offset represents the distance in bytes of the label from the beginning of the data segment.

	.data
	byt BYTE ?
	wrd WORD ?
	dbl DWORD ?
	array WORD 1,2,3,4,5,6,7,8,9
	.code
	mov esi,OFFSET byt       ;assume byt located at 00808000h = ESI
	mov esi,OFFSET wrd       ;00808001h = ESI
	mov esi,OFFSET dbl       ;00808003h = ESI
	mov esi,OFFSET array + 6 ;0080800Dh = ESI

The ALIGN directive is used to align a variable on an even-numbered address in memory because the CPU can access these variables faster.


	.data
	byt BYTE ?       ;00808000h
	ALIGN 2
	wrd WORD ?       ;00808002h
	dbl DWORD ?      ;00808004h

Use the PTR directive to override the default size of an operand.

	.data
	dbl DWORD 12345678h
	wdbl WORD 5678h,1234h
	.code
	mov al,BYTE PTR dbl     ;al = 78h
	mov al,BYTE PTR [dbl+1] ;al = 56h
	mov al,BYTE PTR [dbl+2] ;al = 34h
	mov al,BYTE PTR [dbl+3] ;al = 12h
	mov eax,DWORD PTR wdbl  ;eax = 12345678h

The TYPE operator returns the size in bytes of a single element of a variable.

The LENGTHOF operator returns the number of elements in an array.

Using the SIZEOF operator is the same as multiplying LENGTHOF by TYPE.

The LABEL directive can be used to construct a larger integer from to smaller ones and vice-versa.

	.data
	wordword LABEL DWORD
	var1 WORD 5678h
	var2 WORD 1234h
	.code
	mov eax,wordword ;EAX = 12345678h

Indirect Addressing

By using a register as a pointer to access an array is called indirect addressing.

	.data
	val BYTE Fh
	.code
	mov esi,OFFSET val
	mov al,[esi]       ;al = Fh, [esi] is a pointer and this only works on registers

The assembler does not know whether ESI points to a BYTE, WORD or DWORD. The PTR operator tells the assembler what a registry points to. Note: EAX, EBX, ECX, EDX, etc. could have been used with the same results.

	inc [esi]		;DOES NOT WORK!
	inc BYTE PTR [esi]

Indirect operands are needed to deal with arrays.

	.data
	bytearray BYTE 1,2,3,4,5,6,7,8
	wordarray WORD 10,11,12,13,14,15,16
	.code
	mov esi,OFFSET bytearray
	mov al,[esi]    ;AL = 1
	add esi,1       ;increment by 1 (byte)
	mov al,[esi]    ;AL = 2
	add esi,1       ;increment by 1 (byte)
	mov al,[esi]    ;AL = 3

	mov esi,OFFSET wordarray
	mov al,[esi]    ;AL = 10
	add esi,2       ;increment by 2 (bytes)
	mov al,[esi]    ;AL = 11
	add esi,2       ;increment by 2 (bytes)
	mov al,[esi]    ;AL = 12

An indexed operand adds a constant to a register to generate an effective address.

	.data
	bytearray BYTE 1,2,3,4,5,6,7,8
	wordarray WORD 10,11,12,13,14,15,16
	.code
	mov esi,OFFSET bytearray
	mov al,[esi]     ;AL = 1
	mov al,[esi+1]   ;AL = 2
	mov al,[esi+2]   ;AL = 3

	mov esi,OFFSET wordarray
	mov al,[esi]     ;AL = 10
	mov al,[esi+2]   ;AL = 11
	mov al,[esi+4]   ;AL = 12

A variable that contains the address of another variable is a pointer.

	.data
	bytearray BYTE "Hello World!",0
	ptrbyte DWORD OFFSET bytearray

This little code snippet demonstrates loading the address of a variable that points to a DWORD into the ESI register. It then increments the variable pointed to by one and then moves that value into the EAX register.

	mov esi,[esp+8]
	inc DWORD PTR [esi]
	mov eax,DWORD PTR [esi]

This code snippet demonstrates an easy way of traversing an array of DWORD values. It uses EAX × 4 (bytes) as an index into the array.


	mov esi,[esp+8]
	xor eax, eax              ; eax = 0
loop_start:
	inc DWORD PTR [esi+eax*4] ; increment the value in the array by one
	inc eax                   ; increment the index into the array
	cmp eax, 100
	jne loop_start            ; stop looping when eax = 100

The TYPEDEF operator creates a user-defined type. TYPEDEF is useful for creating pointer variables. This example creates new data types that are pointers:

	PBYTE TYPEDEF PTR BYTE
	PWORD TYPEDEF PTR WORD
	PDWORD TYPEDEF PTR DWORD

The JMP instruction jumps unconditionally to a code label and the LOOP instruction repeats a block of statements a specified number of times. A LOOP instruction must jump between -128 to +127 bytes. Instructions have an average size of 3 bytes.

	.data
	.code
lblBegin:
	xor edx,edx     ;edx = 0
	mov ecx,10      ;ecx is used by LOOP as a counter
lblLoop:
	inc edx
	loop lblLoop    ;first subtract 1 from ecx then if ecx is zero stop looping otherwise jump to lblLoop
	jmp lblBegin    ;jmp unconditionally to the label lblBegin

Example copying a string from a source to a destination using the LOOP instruction:

	.data
	sour BYTE "copy me to a new place in memory.",0
	dest BYTE SIZEOF sour DUP(0),0
	.code
	xor esi,esi
	mov ecx,SIZEOF sour
lblLoop:
	mov al,sour[esi]
	mov dest[esi],al
	add esi,1
	loop lblLoop
<<<[Page 6 of 15]>>>

This site uses cookies. Cookies are simple text files stored on the user's computer. They are used for adding features and security to this site. Read the privacy policy.
CLOSE