1
Lecture 6Lecture 6
Conditional ProcessingConditional Processing
Assembly Language for
Intel-Based Computers,
4th edition
Kip R. Irvine
2
OutlineOutline
Boolean and Comparison Instructions
Conditional Jumps Conditional Loops High-Level Logic Structures
3
Logic InstructionsLogic Instructions Syntax for AND, OR, XOR, and TEST:
op-code destination, source AND, OR and XOR perform the Boolean bitwise
operation and store the result into destination. TEST is just an AND but the result is not stored Both operands must be of the same type
either byte, word or doubleword Both operands cannot be memory at same time
again: memory to memory operations are forbidden They modify OF, CF, SF, ZF, PF according to the
result of the operation (as usual)
4
Logic Instructions (cont.)Logic Instructions (cont.)
The source is often an immediate operand called a bit mask: used to fix certain bits to 0 or 1
To clear a bit we use an AND since: 0 AND b = 0 (b is cleared) 1 AND b = b (b is conserved)
Ex: to clear the sign bit of AL without affecting the others, we do: AND al,7Fh ;msb of AL is cleared
since 7Fh = 0111 1111b
5
Logic Instructions (cont.)Logic Instructions (cont.)
To set (ie: set to 1) certain bits, we use OR: 1 OR b = 1 (b is set) 0 OR b = b (b is conserved)
To set the sign bit of AH, we do: OR ah,80h
To test if ECX=0 we can do: OR ecx,ecx since this does not change the number in ECX
and set ZF=1 if and only if ecx=0
6
Logic Instructions (cont.)Logic Instructions (cont.)
XOR can be used to complement, conserve, or clear certain bits because: b XOR 1 = NOT(b) (b is complemented) b XOR 0 = b (b is conserved) b XOR b = 0 (b is cleared)
Ex: to initialize a register to 0 we can use a two-byte instruction: XOR ax,ax instead of the three-byte instruction: MOV ax,0
This was faster on 8088's
7
Logic Instructions (cont.)Logic Instructions (cont.)
Using OR and XOR in strange ways: XOR ax, ax MOV ax, 0
OR ax, ax CMP ax, 0bytes 2 386/88 3 cycles 4 cycles286 2 cycles 2 cycles386 2 cycles 2 cycles486 1 cycle 1 cycle
Machine code 33 C0 B8 0000
0B C0 83 F8 00
8
Logic Instructions (cont.)Logic Instructions (cont.)
To convert from upper case to lower case we can use the usual method: ADD dl,20h
But "A" = 0100 0001b "a" = 0110 0001b OR dl,20h ;converts from upper to lower case AND dl,0DFh ;converts from lower to upper case
since DFh = 1101 1111b OR dl,20h and AND dh, 20h
leaves lower case characters in dl unchanged and upper case characters in dh unchanged
Bit 5
9
Logic Instructions (cont.)Logic Instructions (cont.)
To invert all the bits (ones complement): NOT destination does not affect any flag and destination cannot
be an immediate operand
Reflection What are the easiest ways to clearing or
setting individual CPU flags?
10
The CMP instructionThe CMP instruction CMP destination,source
Performs: destination - source but does not store the result of this subtraction into destination
But the flags are affected like SUB Same restrictions on operands as SUB Very often used just before performing a
conditional jump CMP is similar to TEST except TEST does
AND operation
11
Conditional JumpsConditional Jumps Jumps Based On . . .
Specific flags Equality Unsigned comparisons Signed Comparisons
Applications Encrypting a String Bit Test (BT) Instruction
12
JJcondcond Instruction Instruction
A conditional jump instruction branches to a label when specific register or flag conditions are met
Examples: JB, JC jump to a label if the Carry flag is set JE, JZ jump to a label if the Zero flag is set JS jumps to a label if the Sign flag is set JNE, JNZ jump to a label if the Zero flag is clear JECXZ jumps to a label if ECX equals 0
17
Applications Applications (1 of 5) (1 of 5)
cmp eax,ebxja Larger
• Task: Jump to a label if unsigned EAX is greater than EBX
• Solution: Use CMP, followed by JA
cmp eax,ebxjg Greater
• Task: Jump to a label if signed EAX is greater than EBX
• Solution: Use CMP, followed by JG
18
Applications Applications (2 of 5) (2 of 5)
cmp eax,Val1jbe L1 ; below or equal
• Jump to label L1 if unsigned EAX is less than or equal to Val1
cmp eax,Val1jle L1
• Jump to label L1 if signed EAX is less than or equal to Val1
19
Applications Applications (3 of 5) (3 of 5)
mov Large,bxcmp ax,bxjna Nextmov Large,ax
Next:
• Compare unsigned AX to BX, and copy the larger of the two into a variable named Large
mov Small,axcmp bx,axjnl Nextmov Small,bx
Next:
• Compare signed AX to BX, and copy the smaller of the two into a variable named Small
20
Applications Applications (4 of 5) (4 of 5)
cmp WORD PTR [esi],0je L1
• Jump to label L1 if the memory word pointed to by ESI equals Zero
test DWORD PTR [edi],1jz L2
• Jump to label L2 if the doubleword in memory pointed to by EDI is even
21
Applications Applications (5 of 5) (5 of 5)
and al,00001011b ; clear unwanted bitscmp al,00001011b ; check remaining bitsje L1 ; all set? jump to L1
• Task: Jump to label L1 if bits 0, 1, and 3 in AL are all set.
• Solution: Clear all bits except bits 0, 1,and 3. Then compare the result with 00001011 binary.
22
Your turn . . .Your turn . . . Write code that jumps to label L1 if either
bit 4, 5, or 6 is set in the BL register. Write code that jumps to label L1 if bits 4,
5, and 6 are all set in the BL register. Write code that jumps to label L3 if EAX is
negative.
23
Encrypting a StringEncrypting a String
KEY = 239 ; can be any byte valueBUFMAX = 128.databuffer BYTE BUFMAX+1 DUP(0)bufSize DWORD BUFMAX
.codemov ecx,bufSize ; loop countermov esi,0 ; index 0 in buffer
L1:xor buffer[esi],KEY ; translate a byteinc esi ; point to next byteloop L1
The following loop uses the XOR instruction to transform every character in a string into a new value.
24
String Encryption ProgramString Encryption Program
Tasks: Input a message (string) from the user Encrypt the message Display the encrypted message Decrypt the message Display the decrypted message
View the Encrypt.asm program's source code. Sample output:
Enter the plain text: Attack at dawn.
Cipher text: «¢¢Äîä-Ä¢-ïÄÿü-Gs
Decrypted: Attack at dawn.
26
LOOPZ and LOOPELOOPZ and LOOPE Syntax:
LOOPE destinationLOOPZ destination
Logic: ECX ECX – 1 if ECX > 0 and ZF=1, jump to destination
Useful when scanning an array for the first element that does not match a given value.
27
LOOPNZ and LOOPNELOOPNZ and LOOPNE
LOOPNZ (LOOPNE) is a conditional loop instruction
Syntax: LOOPNZ destinationLOOPNE destination
Logic: ECX ECX – 1; if ECX > 0 and ZF=0, jump to destination
Useful when scanning an array for the first element that matches a given value.
28
LOOPNZ ExampleLOOPNZ Example
.dataarray SWORD -3,-6,-1,-10,10,30,40,4sentinel SWORD 0.code
mov esi,OFFSET arraymov ecx,LENGTHOF array
next:test WORD PTR [esi],8000h ; test sign bitpushfd ; push flags on stackadd esi,TYPE arraypopfd ; pop flags from stackloopnz next ; continue loopjnz quit ; none foundsub esi,TYPE array ; ESI points to value
quit:
The following code finds the first positive value in an array:
29
Your turn . . .Your turn . . .
.dataarray SWORD 50 DUP(?)sentinel SWORD 0FFFFh.code
mov esi,OFFSET arraymov ecx,LENGTHOF array
L1: cmp WORD PTR [esi],0 ; check for zero
(fill in your code here)
quit:
Locate the first nonzero value in the array. If none is found, let ESI point to the sentinel value:
30
. . . (solution). . . (solution).dataarray SWORD 50 DUP(?)sentinel SWORD 0FFFFh.code
mov esi,OFFSET arraymov ecx,LENGTHOF array
L1: cmp WORD PTR [esi],0 ; check for zeropushfd ; push flags on stackadd esi,TYPE arraypopfd ; pop flags from stackloope L1 ; continue loopjz quit ; none foundsub esi,TYPE array ; ESI points to value
quit:
31
Conditional StructuresConditional Structures
• Block-Structured IF Statements
• Compound Expressions with AND
• Compound Expressions with OR
• WHILE Loops
• Table-Driven Selection
32
Block-Structured IF StatementsBlock-Structured IF Statements
Assembly language programmers can easily translate
logical statements written in C++/Java into assembly
language. For example:
mov eax,op1cmp eax,op2jne L1mov X,1jmp L2
L1: mov X,2L2:
if( op1 == op2 ) X = 1;else X = 2;
33
Your turn . . .Your turn . . .
Implement the following pseudocode in
assembly language. All values are unsigned:
cmp ebx,ecxja nextmov eax,5mov edx,6
next:
if( ebx <= ecx ){ eax = 5; edx = 6;}
(There are multiple correct solutions to this problem.)
34
Your turn . . .Your turn . . .
Implement the following pseudocode in assembly
language. All values are 32-bit signed integers:mov eax,var1cmp eax,var2jle L1mov var3,6mov var4,7jmp L2
L1: mov var3,10L2:
if( var1 <= var2 ) var3 = 10;else{ var3 = 6; var4 = 7;}
(There are multiple correct solutions to this problem.)
35
Compound Expression with ANDCompound Expression with AND (1 of 3) (1 of 3)
When implementing the logical AND operator, consider
that HLLs use short-circuit evaluation
In the following example, if the first expression is false, the
second expression is skipped:
if (al > bl) AND (bl > cl) X = 1;
36
Compound Expression with ANDCompound Expression with AND (2 of 3) (2 of 3)
cmp al,bl ; first expression...ja L1jmp next
L1:cmp bl,cl ; second expression...ja L2jmp next
L2: ; both are truemov X,1 ; set X to 1
next:
if (al > bl) AND (bl > cl) X = 1;
This is one possible implementation . . .
37
Compound Expression with ANDCompound Expression with AND (3 of 3) (3 of 3)
cmp al,bl ; first expression...jbe next ; quit if falsecmp bl,cl ; second expression...jbe next ; quit if falsemov X,1 ; both are true
next:
if (al > bl) AND (bl > cl) X = 1;
But the following implementation uses 29% less code by reversing the first relational operator. We allow the program to "fall through" to the second expression:
38
Your turn . . .Your turn . . .
Implement the following pseudocode in
assembly language. All values are unsigned:
cmp ebx,ecxja nextcmp ecx,edxjbe nextmov eax,5mov edx,6
next:
if( ebx <= ecx && ecx > edx )
{ eax = 5; edx = 6;}
(There are multiple correct solutions to this problem.)
39
Compound Expression with ORCompound Expression with OR (1 of 2) (1 of 2) When implementing the logical OR operator, consider that
HLLs use short-circuit evaluation
In the following example, if the first expression is true, the
second expression is skipped:
if (al > bl) OR (bl > cl) X = 1;
40
Compound Expression with ORCompound Expression with OR (1 of 2) (1 of 2)
cmp al,bl ; is AL > BL?ja L1 ; yescmp bl,cl ; no: is BL > CL?jbe next ; no: skip next statement
L1: mov X,1 ; set X to 1next:
if (al > bl) OR (bl > cl) X = 1;
We can use "fall-through" logic to keep the code as short as possible:
41
WHILE LoopsWHILE Loops
while( eax < ebx)eax = eax + 1;
A WHILE loop is really an IF statement followed by the body of the loop, followed by an unconditional jump to the top of the loop. Consider the following example:
top:cmp eax,ebx ; check loop conditionjae next ; false? exit loopinc eax ; body of loopjmp top ; repeat the loop
next:
This is a possible implementation:
42
Your turn . . .Your turn . . .
top:cmp ebx,val1 ; check loop conditionja next ; false? exit loopadd ebx,5 ; body of loopdec val1jmp top ; repeat the loop
next:
while( ebx <= val1){
ebx = ebx + 5;val1 = val1 - 1
}
Implement the following loop, using unsigned 32-bit integers:
43
Using the .IF DirectiveUsing the .IF Directive
Runtime Expressions Relational and Logical
Operators MASM-Generated Code .REPEAT Directive .WHILE Directive
44
Runtime ExpressionsRuntime Expressions
.IF eax > ebxmov edx,1
.ELSEmov edx,2
.ENDIF
• .IF, .ELSE, .ELSEIF, and .ENDIF can be used to evaluate runtime expressions and create block-structured IF statements.
• Examples:
• MASM generates "hidden" code for you, consisting of code labels, CMP and conditional jump instructions.
.IF eax > ebx && eax > ecxmov edx,1
.ELSEmov edx,2
.ENDIF
46
MASM-Generated CodeMASM-Generated Code
mov eax,6cmp eax,val1jbe @C0001 mov result,1
@C0001:
.data
val1 DWORD 5
result DWORD ?
.code
mov eax,6
.IF eax > val1
mov result,1
.ENDIF
Generated code:
MASM automatically generates an unsigned jump (JBE) because val1 is unsigned.
47
MASM-Generated CodeMASM-Generated Code
mov eax,6cmp eax,val1jle @C0001 mov result,1
@C0001:
.data
val1 SDWORD 5
result SDWORD ?
.code
mov eax,6
.IF eax > val1
mov result,1
.ENDIF
Generated code:
MASM automatically generates a signed jump (JLE) because val1 is signed.
48
MASM-Generated CodeMASM-Generated Code
mov ebx,5mov eax,6cmp eax,ebxjbe @C0001 mov result,1
@C0001:
.data
result DWORD ?
.code
mov ebx,5
mov eax,6
.IF eax > ebx
mov result,1
.ENDIF
Generated code:
MASM automatically generates an unsigned jump (JBE) when both operands are registers . . .
49
MASM-Generated CodeMASM-Generated Code
mov ebx,5mov eax,6cmp eax,ebxjle @C0001 mov result,1
@C0001:
.data
result SDWORD ?
.code
mov ebx,5
mov eax,6
.IF SDWORD PTR eax > ebx
mov result,1
.ENDIF
Generated code:
. . . unless you prefix one of the register operands with the SDWORD PTR operator. Then a signed jump is generated.
50
.REPEAT Directive.REPEAT Directive
; Display integers 1 – 10:
mov eax,0.REPEAT
inc eaxcall WriteDeccall Crlf
.UNTIL eax == 10
Executes the loop body before testing the loop condition associated with the .UNTIL directive.
Example:
51
.WHILE Directive.WHILE Directive
; Display integers 1 – 10:
mov eax,0.WHILE eax < 10
inc eaxcall WriteDeccall Crlf
.ENDW
Tests the loop condition before executing the loop body The .ENDW directive marks the end of the loop.
Example:
52
MacrosMacros
Introducing Macros Defining Macros Invoking Macros Macro Examples Nested Macros Example Program: Wrappers
53
Introducing MacrosIntroducing Macros
A macro1 is a named block of assembly language statements.
Once defined, it can be invoked (called) one or more times.
During the assembler's preprocessing step, each macro call is expanded into a copy of the macro.
The expanded code is passed to the assembly step, where it is checked for correctness.
1Also called a macro procedure.
54
Defining MacrosDefining Macros
• A macro must be defined before it can be used.
• Parameters are optional.
• Each parameter follows the rules for identifiers. It is a string that is assigned a value when the macro is invoked.
• Syntax:
macroname MACRO [parameter-1, parameter-2,...]
statement-list
ENDM
55
mNewLine Macro ExamplemNewLine Macro Example
mNewLine MACRO ; define the macrocall Crlf
ENDM.data
.codemNewLine ; invoke the macro
This is how you define and invoke a simple macro.
The assembler will substitute "call crlf" for "mNewLine".
56
mPutChar MacromPutChar Macro
mPutchar MACRO charpush eaxmov al,charcall WriteCharpop eax
ENDM
Writes a single character to standard output.
Definition:
.codemPutchar 'A'Invocation:
1 push eax1 mov al,'A'1 call WriteChar1 pop eax
Expansion:viewed in the listing file
57
Invoking MacrosInvoking Macros (1 of 2) (1 of 2)
When you invoke a macro, each argument you pass matches a declared parameter.
Each parameter is replaced by its corresponding argument when the macro is expanded.
When a macro expands, it generates assembly language source code.
Arguments are treated as simple text by the preprocessor.
58
Invoking MacrosInvoking Macros (2 of 2) (2 of 2)Relationships between macros, arguments, and parameters:
59
mWriteStr MacromWriteStr Macro (1 of 2) (1 of 2)
mWriteStr MACRO bufferpush edxmov edx,OFFSET buffercall WriteStringpop edx
ENDM.datastr1 BYTE "Welcome!",0.codemWriteStr str1
Provides a convenient way to display a string, by passing the string name as an argument.
60
mWriteStr MacromWriteStr Macro (2 of 2) (2 of 2)
1 push edx1 mov edx,OFFSET str11 call WriteString1 pop edx
The expanded code shows how the str1 argument replaced the parameter named buffer:
mWriteStr MACRO bufferpush edxmov edx,OFFSET buffercall WriteStringpop edx
ENDM
61
Invalid ArgumentInvalid Argument
If you pass an invalid argument, the error is caught when the expanded code is assembled.
Example:
.codemPutchar 1234h
1 push eax1 mov al,1234h ; error!1 call WriteChar1 pop eax
mPutchar MACRO charpush eaxmov al,charcall WriteCharpop eax
ENDM
62
Blank ArgumentBlank Argument
If you pass a blank argument, the error is also caught when the expanded code is assembled.
Example:
.codemPutchar
1 push eax1 mov al,1 call WriteChar1 pop eax
mPutchar MACRO charpush eaxmov al,charcall WriteCharpop eax
ENDM