Tutorial Emu8086
Tutorial Emu8086
Tutorial Emu8086
TUTORIAL
Emulador Emu8086 do
Microprocessador 8086
2
This tutorial is intended for those who are not familiar with assembler at all, or have a very distant idea about it. Of
course if you have knowledge of some other programming language (Basic, C/C++, Pascal...) that may help you a
lot. But even if you are familiar with assembler, it is still a good idea to look through this document in order to
study Emu8086 syntax.
It is assumed that you have some knowledge about number representation (HEX/BIN), if not it is highly
recommended to study Numbering Systems Tutorial before you proceed.
Assembly language is a low level programming language. You need to get some knowledge about computer
structure in order to understand anything. The simple computer model as I see it:
The system bus (shown in yellow) connects the various components of a computer.
The CPU is the heart of the computer, most of computations occur inside the CPU.
RAM is a place to where the programs are loaded in order to be executed.
8086 CPU has 8 general purpose registers, each register has its own name:
Despite the name of a register, it's the programmer who determines the usage for each general purpose register. The
main purpose of a register is to keep a number (variable). The size of the above registers is 16 bit, it's something
like: 0011000000111001b (in binary form), or 12345 in decimal (human) form.
4 general purpose registers (AX, BX, CX, DX) are made of two separate 8 bit registers, for example if AX=
0011000000111001b, then AH=00110000b and AL=00111001b. Therefore, when you modify any of the 8 bit
registers 16 bit register is also updated, and vice-versa. The same is for other 3 registers, "H" is for high and "L" is
for low part.
Because registers are located inside the CPU, they are much faster than memory. Accessing a memory location
requires the use of a system bus, so it takes much longer. Accessing data in a register usually takes no time.
Therefore, you should try to keep variables in the registers. Register sets are very small and most registers have
special purposes which limit their use as variables, but they are still an excellent place to store temporary data of
calculations.
SEGMENT REGISTERS
Although it is possible to store any data in the segment registers, this is never a good idea. The segment registers
have a very special purpose - pointing at accessible blocks of memory.
Segment registers work together with general purpose register to access any memory value. For example if we
would like to access memory at the physical address 12345h (hexadecimal), we should set the DS = 1230h and SI =
0045h. This is good, since this way we can access much more memory than with a single register that is limited to
16 bit values.
CPU makes a calculation of physical address by multiplying the segment register by 10h and adding general
purpose register to it (1230h * 10h + 45h = 12345h):
IP register always works together with CS segment register and it points to currently executing instruction.
Flags Register is modified automatically by CPU after mathematical operations, this allows to determine the type
of the result, and to determine conditions to transfer control to other parts of the program.
Generally you cannot access these registers directly.
Memory Access
To access memory we can use these four registers: BX, SI, DI, BP.
Combining these registers inside [ ] symbols, we can get different memory locations. These combinations are
supported (addressing modes):
Displacement can be a immediate value or offset of a variable, or even both. It's up to compiler to calculate a single
immediate value.
Displacement can be inside or outside of [ ] symbols, compiler generates the same machine code for both ways.
Generally the compiler takes care about difference between d8 and d16, and generates the required machine code.
By default DS segment register is used for all modes except those with BP register, for these SS segment register is
used.
There is an easy way to remember all those possible combinations using this chart:
You can form all valid combinations by taking only one item from each column or skipping the column by not
taking anything from it. As you see BX and BP never go together. SI and DI also don't go together. Here is an
example of a valid addressing mode: [BX+5].
5
The value in segment register (CS, DS, SS, ES) is called a "segment",
and the value in purpose register (BX, SI, DI, BP) is called an "offset".
When DS contains value 1234h and SI contains the value 7890h it can be also recorded as 1234:7890. The physical
address will be 1234h * 10h + 7890h = 19BD0h.
For example:
BYTE PTR [BX] ; byte access.
or
WORD PTR [BX] ; word access.
Emu8086 supports shorter prefixes as well:
sometimes compiler can calculate the data type automatically, but you may not and should not rely on that when
one of the operands is an immediate value.
MOV instruction
• The source operand can be an immediate value, general-purpose register or memory location.
• Both operands must be the same size, which can be a byte or a word.
REG: AX, BX, CX, DX, AH, AL, BL, BH, CH, CL, DH, DL, DI, SI, BP, SP.
The MOV instruction cannot be used to set the value of the CS and IP registers.
You can copy & paste the above program to Emu8086 code editor, and press [Compile and Emulate] button (or
press F5 key on your keyboard).
The Emulator window should open with this program loaded, click [Single Step] button and watch the register
values.
1. Select the above text using mouse, click before the text and drag it down until everything is selected.
As you may guess, ";" is used for comments, anything after ";" symbol is ignored by compiler.
Actually the above program writes directly to video memory, so you may see that MOV is a very powerful
instruction.
Variables
Variable is a memory location. For a programmer it is much easier to have some value be kept in a variable named
"var1" then at the address 5A73:235B, especially when you have 10 or more variables.
name DB value
name DW value
name - can be any letter or digit combination, though it should start with a letter. It's possible to declare
unnamed variables by not specifying the name (this variable will have an address but no name).
value - can be any numeric value in any supported numbering system (hexadecimal, binary, or decimal), or "?"
symbol for variables that are not initialized.
As you probably know from part 2 of this tutorial, MOV instruction is used to copy values from source to
destination.
Let's see another example with MOV instruction:
#MAKE_COM#
ORG 100h
VAR1 DB 7
var2 DW 1234h
Copy the above code to Emu8086 source editor, and press F5 key to compile and load it in the emulator. You
should get something like:
8
As you see this looks a lot like our example, except that variables are replaced with actual memory locations. When
compiler makes machine code, it automatically replaces all variable names with their offsets. By default segment is
loaded in DS register (when COM files is loaded the value of DS register is set to the same value as CS register -
code segment).
In memory list first row is an offset, second row is a hexadecimal value, third row is decimal value, and last row
is an ASCII character value.
Compiler is not case sensitive, so "VAR1" and "var1" refer to the same variable.
The offset of var2 is 0109h, and full address is 0B56:0109, this variable is a WORD so it occupies 2 BYTES. It is
assumed that low byte is stored at lower address, so 34h is located before 12h.
You can see that there are some other instructions after the RET instruction, this happens because disassembler has
no idea about where the data starts, it just processes the values in memory and it understands them as valid 8086
instructions (we will learn them later). You can even write the same program using DB directive only:
#MAKE_COM#
ORG 100h
DB 0A0h
DB 08h
DB 01h
DB 8Bh
DB 1Eh
DB 09h
DB 01h
DB 0C3h
DB 7
DB 34h
DB 12h
9
Copy the above code to Emu8086 source editor, and press F5 key to compile and load it in the emulator. You
should get the same disassembled code, and the same functionality!
As you may guess, the compiler just converts the program source to the set of bytes, this set is called machine
code, processor understands the machine code and executes it.
ORG 100h is a compiler directive (it tells compiler how to handle the source code). This directive is very important
when you work with variables. It tells compiler that the executable file will be loaded at the offset of 100h (256
bytes), so compiler should calculate the correct address for all variables when it replaces the variable names with
their offsets. Directives are never converted to any real machine code.
Why executable file is loaded at offset of 100h? Operating system keeps some data about the program in the first
256 bytes of the CS (code segment), such as command line parameters and etc.
Though this is true for COM files only, EXE files are loaded at offset of 0000, and generally use special segment
for variables. Maybe we'll talk more about EXE files later.
Arrays
Arrays can be seen as chains of variables. A text string is an example of a byte array, each character is presented as
an ASCII code value (0..255).
b is an exact copy of the a array, when compiler sees a string inside quotes it automatically converts it to set of
bytes. This chart shows a part of the memory where these arrays are declared:
You can access the value of any element in array using square brackets, for example:
MOV AL, a[3]
You can also use any of the memory index registers BX, SI, DI, BP, for example:
MOV SI, 3
MOV AL, a[SI]
If you need to declare a large array you can use DUP operator.
The syntax for DUP:
for example:
c DB 5 DUP(9)
is an alternative way of declaring:
c DB 9, 9, 9, 9, 9
Of course, you can use DW instead of DB if it's required to keep values larger then 255, or smaller then -128. DW
cannot be used to declare strings!
The expansion of DUP operand should not be over 1020 characters! (the expansion of last example is 13 chars), if
you need to declare huge array divide declaration it in two lines (you will get a single huge array in the memory).
There is LEA (Load Effective Address) instruction and alternative OFFSET operator. Both OFFSET and LEA
can be used to get the offset address of the variable.
LEA is more powerful because it also allows you to get the address of an indexed variables. Getting the address of
the variable can be very useful in some situations, for example when you need to pass parameters to a procedure.
Reminder:
In order to tell the compiler about data type,
these prefixes should be used:
For example:
BYTE PTR [BX] ; byte access.
or
WORD PTR [BX] ; word access.
Emu8086 supports shorter prefixes as well:
sometimes compiler can calculate the data type automatically, but you may not and should not rely on that when
one of the operands is an immediate value.
ORG 100h
ORG 100h
11
RET
VAR1 DB 22h
END
These lines:
LEA BX, VAR1
MOV BX, OFFSET VAR1
are even compiled into the same machine code: MOV BX, num
num is a 16 bit value of the variable offset.
Please note that only these registers can be used inside square brackets (as memory pointers): BX, SI, DI, BP!
(see previous part of the tutorial).
Constants
Constants are just like variables, but they exist only until your program is compiled (assembled). After definition of
a constant its value cannot be changed. To define constants EQU directive is used:
For example:
k EQU 5
MOV AX, k
MOV AX, 5
You can view variables while your program executes by selecting "Variables" from the "View" menu of emulator.
To view arrays you should click on a variable and set Elements property to array size. In assembly language there
are not strict data types, so any variable can be presented as an array.
12
You can edit a variable's value when your program is running, simply double click it, or select it and click Edit
button.
It is possible to enter numbers in any system, hexadecimal numbers should have "h" suffix, binary "b" suffix, octal
"o" suffix, decimal numbers require no suffix. String can be entered this way:
'hello world', 0
(this string is zero terminated).
Interrupts
Interrupts can be seen as a number of functions. These functions make the programming much easier, instead of
writing a code to print a character you can simply call the interrupt and it will do everything for you. There are also
interrupt functions that work with disk drive and other hardware. We call such functions software interrupts.
Interrupts are also triggered by different hardware, these are called hardware interrupts. Currently we are
interested in software interrupts only.
To make a software interrupt there is an INT instruction, it has very simple syntax:
INT value
The following example uses INT 10h sub-function 0Eh to type a "Hello!" message. This functions displays a
character on the screen, advancing the cursor and scrolling the screen as necessary.
13
Copy & paste the above program to Emu8086 source code editor, and press [Compile and Emulate] button. Run
it!
To make programming easier there are some common functions that can be included in your program. To make
your program use functions defined in other file you should use the INCLUDE directive followed by a file name.
Compiler automatically searches for the file in the same folder where the source file is located, and if it cannot find
the file there - it searches in Inc folder.
Currently you may not be able to fully understand the contents of the emu8086.inc (located in Inc folder), but it's
OK, since you only need to understand what it can do.
To use any of the functions in emu8086.inc you should have the following line in the beginning of your source file:
include 'emu8086.inc'
• PUTC char - macro with 1 parameter, prints out an ASCII char at current cursor position.
• PRINTN string - macro with 1 parameter, prints out a string. The same as PRINT but automatically adds
"carriage return" at the end of the string.
To use any of the above macros simply type its name somewhere in your code, and if required parameters, for
example:
include emu8086.inc
ORG 100h
GOTOXY 10, 5
When compiler process your source code it searches the emu8086.inc file for declarations of the macros and
replaces the macro names with real code. Generally macros are relatively small parts of code, frequent use of a
macro may make your executable too big (procedures are better for size optimization).
• PRINT_STRING - procedure to print a null terminated string at current cursor position, receives address
of string in DS:SI register. To use it declare: DEFINE_PRINT_STRING before END directive.
• PTHIS - procedure to print a null terminated string at current cursor position (just as PRINT_STRING),
but receives address of string from Stack. The ZERO TERMINATED string should be defined just after
the CALL instruction. For example:
CALL PTHIS
db 'Hello World!', 0
• GET_STRING - procedure to get a null terminated string from a user, the received string is written to
buffer at DS:DI, buffer size should be in DX. Procedure stops the input when 'Enter' is pressed. To use it
declare: DEFINE_GET_STRING before END directive.
• CLEAR_SCREEN - procedure to clear the screen, (done by scrolling entire screen window), and set
cursor position to top of it. To use it declare: DEFINE_CLEAR_SCREEN before END directive.
• SCAN_NUM - procedure that gets the multi-digit SIGNED number from the keyboard, and stores the
result in CX register. To use it declare: DEFINE_SCAN_NUM before END directive.
15
• PRINT_NUM - procedure that prints a signed number in AX register. To use it declare:
DEFINE_PRINT_NUM and DEFINE_PRINT_NUM_UNS before END directive.
• PRINT_NUM_UNS - procedure that prints out an unsigned number in AX register. To use it declare:
DEFINE_PRINT_NUM_UNS before END directive.
To use any of the above procedures you should first declare the function in the bottom of your file (but before
END!!), and then use CALL instruction followed by a procedure name. For example:
include 'emu8086.inc'
ORG 100h
DEFINE_SCAN_NUM
DEFINE_PRINT_STRING
DEFINE_PRINT_NUM
DEFINE_PRINT_NUM_UNS ; required for print_num.
DEFINE_PTHIS
First compiler processes the declarations (these are just regular the macros that are expanded to procedures). When
compiler gets to CALL instruction it replaces the procedure name with the address of the code where the procedure
is declared. When CALL instruction is executed control is transferred to procedure. This is quite useful, since even
if you call the same procedure 100 times in your code you will still have relatively small executable size. Seems
complicated, isn't it? That's ok, with the time you will learn more, currently it's required that you understand the
basic principle.
Most Arithmetic and Logic Instructions affect the processor status register (or Flags)
16
As you may see there are 16 bits in this register, each bit is called a flag and can take a value of 1 or 0.
• Carry Flag (CF) - this flag is set to 1 when there is an unsigned overflow. For example when you add
bytes 255 + 1 (result is not in range 0...255). When there is no overflow this flag is set to 0.
• Zero Flag (ZF) - set to 1 when result is zero. For none zero result this flag is set to 0.
• Sign Flag (SF) - set to 1 when result is negative. When result is positive it is set to 0. Actually this flag
take the value of the most significant bit.
• Overflow Flag (OF) - set to 1 when there is a signed overflow. For example, when you add bytes 100 +
50 (result is not in range -128...127).
• Parity Flag (PF) - this flag is set to 1 when there is even number of one bits in result, and to 0 when there
is odd number of one bits. Even if result is a word only 8 low bits are analyzed!
• Auxiliary Flag (AF) - set to 1 when there is an unsigned overflow for low nibble (4 bits).
• Interrupt enable Flag (IF) - when this flag is set to 1 CPU reacts to interrupts from external devices.
• Direction Flag (DF) - this flag is used by some instructions to process data chains, when this flag is set to
0 - the processing is done forward, when this flag is set to 1 the processing is done backward.
REG, memory
memory, REG
REG, REG
memory, immediate
REG, immediate
REG: AX, BX, CX, DX, AH, AL, BL, BH, CH, CL, DH, DL, DI, SI, BP, SP.
After operation between operands, result is always stored in first operand. CMP and TEST instructions affect flags
only and do not store a result (these instruction are used to make decisions during program execution).
• AND - Logical AND between all bits of two operands. These rules apply:
1 AND 1 = 1
1 AND 0 = 0
0 AND 1 = 0
0 AND 0 = 0
1 OR 1 = 1
1 OR 0 = 1
0 OR 1 = 1
0 OR 0 = 0
As you see we get 1 every time when at least one of the bits is 1.
• XOR - Logical XOR (exclusive OR) between all bits of two operands. These rules apply:
1 XOR 1 = 0
1 XOR 0 = 1
0 XOR 1 = 1
0 XOR 0 = 0
As you see we get 1 every time when bits are different from each other.
REG
memory
REG: AX, BX, CX, DX, AH, AL, BL, BH, CH, CL, DH, DL, DI, SI, BP, SP.
REG
memory
REG: AX, BX, CX, DX, AH, AL, BL, BH, CH, CL, DH, DL, DI, SI, BP, SP.
Controlling the program flow is a very important thing, this is where your program can make decisions according to
certain conditions.
• Unconditional Jumps
The basic instruction that transfers control to another point in the program is JMP.
JMP label
To declare a label in your program, just type its name and add ":" to the end, label can be any character
combination but it cannot start with a number, for example here are 3 legal label definitions:
label1:
label2:
a:
Label can be declared on a separate line or before any other instruction, for example:
x1:
MOV AX, 1
ORG 100h
calc:
ADD AX, BX ; add BX to AX.
JMP back ; go 'back'.
stop:
Of course there is an easier way to calculate the some of two numbers, but it's still a good example of JMP
20
instruction.
As you can see from this example JMP is able to transfer control both forward and backward. It can jump
anywhere in current code segment (65,535 bytes).
Unlike JMP instruction that does an unconditional jump, there are instructions that do a conditional jumps
(jump only when some conditions are in act). These instructions are divided in three groups, first group
just test single flag, second compares numbers as signed, and third compares numbers as unsigned.
JC , JB, JNAE Jump if Carry (Below, Not Above Equal). CF = 1 JNC, JNB, JAE
JNC , JNB, Jump if Not Carry (Not Below, Above CF = 0 JC, JB, JNAE
JAE Equal).
•
As you can see there are some instructions that do that same thing, that's correct, they even are assembled
into the same machine code, so it's good to remember that when you compile JE instruction - you will get
it disassembled as: JZ.
Different names are used to make programs easier to understand and code.
ZF = 1
Jump if Less or Equal (<=).
JLE , JNG or JNLE, JG
Jump if Not Greater (not >).
SF <> OF
CF = 0
Jump if Above (>).
JA , JNBE and JNA, JBE
Jump if Not Below or Equal (not <=).
ZF = 0
CF = 1
Jump if Below or Equal (<=).
JBE , JNA or JNBE, JA
Jump if Not Above (not >).
ZF = 1
Generally, when it is required to compare numeric values CMP instruction is used (it does the same as SUB
(subtract) instruction, but does not keep the result, just affects the flags).
Another example:
it's required to compare 7 and 7,
7-7=0
the result is zero! (Zero Flag is set to 1 and JZ or JE will do the jump).
22
Here is an example of CMP instruction and conditional jump:
include emu8086.inc
ORG 100h
stop:
Try the above example with different numbers for AL and BL, open flags by clicking on [FLAGS] button,
use [Single Step] and see what happens, don't forget to recompile and reload after every change (use F5
shortcut).
•
All conditional jumps have one big limitation, unlike JMP instruction they can only jump 127 bytes
forward and 128 bytes backward (note that most instructions are assembled into 3 or more bytes).
o Get a opposite conditional jump instruction from the table above, make it jump to label_x.
Here is an example:
23
not_equal:
stop:
END
Another, yet rarely used method is providing an immediate value instead of a label. When immediate value starts
with a '$' character relative jump is performed, otherwise compiler calculates instruction that jumps directly to
given offset. For example:
ORG 100h
RET
END
Procedures
Procedure is a part of code that can be called from your program in order to make some specific task. Procedures
make program more structural and easier to understand. Generally procedure returns to the same point from where
it was called.
name PROC
24
; here goes the code
; of the procedure ...
RET
name ENDP
name - is the procedure name, the same name should be in the top and the bottom, this is used to check correct
closing of procedures.
Probably, you already know that RET instruction is used to return to operating system. The same instruction is used
to return from procedure (actually operating system sees your program as a special procedure).
PROC and ENDP are compiler directives, so they are not assembled into any real machine code. Compiler just
remembers the address of procedure.
Here is an example:
ORG 100h
CALL m1
MOV AX, 2
RET ; return to operating system.
m1 PROC
MOV BX, 5
RET ; return to caller.
m1 ENDP
END
The above example calls procedure m1, does MOV BX, 5, and returns to the next instruction after CALL: MOV
AX, 2.
There are several ways to pass parameters to procedure, the easiest way to pass parameters is by using registers,
here is another example of a procedure that receives two parameters in AL and BL registers, multiplies these
parameters and returns the result in AX register:
ORG 100h
MOV AL, 1
MOV BL, 2
CALL m2
CALL m2
CALL m2
25
ORG 100h
MOV AL, 1
MOV BL, 2
CALL m2
CALL m2
CALL m2
CALL m2
m2 PROC
MUL BL ; AX = AL * BL.
RET ; return to caller.
m2 ENDP
END
In the above example value of AL register is update every time the procedure is called, BL register stays
unchanged, so this algorithm calculates 2 in power of 4,
so final result in AX register is 16 (or 10h).
ORG 100h
END
"b." - prefix before [SI] means that we need to compare bytes, not words. When you need to compare words add
"w." prefix instead. When one of the compared operands is a register it's not required because compiler knows the
size of each register.
The Stack
Stack is an area of memory for keeping temporary data. Stack is used by CALL instruction to keep return address
for procedure, RET instruction gets this value from the stack and returns to that offset. Quite the same thing
happens when INT instruction calls an interrupt, it stores in stack flag register, code segment and offset. IRET
instruction is used to return from interrupt call.
PUSH REG
PUSH SREG
PUSH memory
PUSH immediate
POP REG
POP SREG
POP memory
Notes:
The stack uses LIFO (Last In First Out) algorithm, this means that if we push these values one by one into the
stack: 1, 2, 3, 4, 5 the first value that we will get on pop will be 5, then 4, 3, 2, and only then 1.
It is very important to do equal number of PUSHs and POPs, otherwise the stack maybe corrupted and it will be
impossible to return to operating system. As you already know we use RET instruction to return to operating
system, so when program starts there is a return address in stack (generally it's 0000h).
PUSH and POP instruction are especially useful because we don't have too much registers to operate with, so here
is a trick:
• Restore the original value of the register from stack (using POP).
Here is an example:
ORG 100h
ORG 100h
28
END
The exchange happens because stack uses LIFO (Last In First Out) algorithm, so when we push 1212h and then
3434h, on pop we will first get 3434h and only after it 1212h.
The stack memory area is set by SS (Stack Segment) register, and SP (Stack Pointer) register. Generally operating
system sets values of these registers on program start.
• Add 2 to SP register.
The current address pointed by SS:SP is called the top of the stack.
For COM files stack segment is generally the code segment, and stack pointer is set to value of 0FFFEh. At the
address SS:0FFFEh stored a return address for RET instruction that is executed in the end of the program.
You can visually see the stack operation by clicking on [Stack] button on emulator window. The top of the stack is
marked with "<" sign.
Macros
Macros are just like procedures, but not really. Macros look like procedures, but they exist only until your code is
compiled, after compilation all macros are replaced with real instructions. If you declared a macro and never used it
in your code, compiler will simply ignore it. emu8086.inc is a good example of how macros can be used, this file
contains several macros to make coding easier for you.
Macro definition:
<instructions>
ENDM
29
MOV AX, p1
MOV BX, p2
MOV CX, p3
ENDM
ORG 100h
MyMacro 1, 2, 3
MyMacro 4, 5, DX
RET
• To mark the end of the procedure, you should type the name of the procedure before the ENDP
directive.
Macros are expanded directly in code, therefore if there are labels inside the macro definition you may get
"Duplicate declaration" error when macro is used for twice or more. To avoid such problem, use LOCAL directive
followed by names of variables, labels or procedure names. For example:
MyMacro2 MACRO
LOCAL label1, label2
CMP AX, 2
JE label1
CMP AX, 3
JE label2
label1:
INC AX
label2:
ADD AX, 2
ENDM
ORG 100h
MyMacro2
MyMacro2
RET
If you plan to use your macros in several programs, it may be a good idea to place all macros in a separate file.
Place that file in Inc folder and use INCLUDE file-name directive to use macros. See Library of common
functions - emu8086.inc for an example of such file.
Usually, when a computer starts it will try to load the first 512-byte sector (that's Cylinder 0, Head 0, Sector 1) from
any diskette in your A: drive to memory location 0000h:7C00h and give it control. If this fails, the BIOS tries to
use the MBR of the first hard drive instead.
This tutorial covers booting up from a floppy drive, the same principles are used to boot from a hard drive. But
using a floppy drive has several advantages:
• You can keep your existing operating system intact (Windows, DOS...).
#MAKE_BOOT#
Copy the above example to Emu8086 source editor and press [Compile and Emulate] button. The Emulator
automatically loads ".boot" file to 0000h:7C00h.
You can run it just like a regular program, or you can use the Virtual Drive menu to Write 512 bytes at 7C00h to
the Boot Sector of a virtual floppy drive (FLOPPY_0 file in Emulator's folder).
After writing your program to the Virtual Floppy Drive, you can select Boot from Floppy from Virtual Drive
menu.
If you are curious, you may write the virtual floppy (FLOPPY_0) or ".boot" file to a real floppy disk and boot your
computer from it, I recommend using "RawWrite for Windows" from:
http://uranus.it.swin.edu.au/~jn/linux/rawwrite.htm
(recent builds now work under all versions of Windows!)
Note: however, that this .boot file is not an MS-DOS compatible boot sector (it will not allow you to read or write
data on this diskette until you format it again), so don't bother writing only this sector to a diskette with data on it.
As a matter of fact, if you use any 'raw-write' programs, such at the one listed above, they will erase all of the data
anyway. So make sure the diskette you use doesn't contain any important data.
32
".boot" files are limited to 512 bytes (sector size). If your new Operating System is going to grow over this size,
you will need to use a boot program to load data from other sectors. A good example of a tiny Operating System
can be found in "Samples" folder as:
micro-os_loader.asm
micro-os_kernel.asm
To create extensions for your Operating System (over 512 bytes), you can use ".bin" files (select "BIN Template"
from "File" -> "New" menu).
To write ".bin" file to virtual floppy, select "Write .bin file to floppy..." from "Virtual Drive" menu of emulator:
Sector at:
Cylinder: 0
Head:0
Sector: 1
• Floppy disk has 2 sides, and there are 2 heads; one for each side (0..1), the drive heads move above the
surface of the disk on each side.
There are 3 devices attached to the emulator: Traffic Lights, Stepper-Motor and Robot. You can view devices using
"Virtual Devices" menu of the emulator.
In general, it is possible to use any x86 family CPU to control all kind of devices, the difference maybe in base I/O
port number, this can be altered using some tricky electronic equipment. Usually the ".bin" file is written into the
Read Only Memory (ROM) chip, the system reads program from that chip, loads it in RAM module and runs the
program. This principle is used for many modern devices such as micro-wave ovens and etc...
Traffic Lights
34
Usually to control the traffic lights an array (table) of values is used. In certain periods of time the value is read
from the array and sent to a port. For example:
table DW 100001100001b
DW 110011110011b
DW 001100001100b
DW 011110011110b
start:
MOV SI, 0
next_value:
LOOP next_value
; ==========================
PAUSE PROC
; store registers:
PUSH CX
PUSH DX
PUSH AX
; restore registers:
POP AX
POP DX
POP CX
RET
PAUSE ENDP
; ==========================
Stepper-Motor
The motor can be half stepped by turning on pair of magnets, followed by a single and so on.
The motor can be full stepped by turning on pair of magnets, followed by another pair of magnets and in the end
followed by a single magnet and so on. The best way to make full step is to make two half steps.
Robot
Complete list of robot instruction set is given in I/O ports section of Emu8086 reference.
To control the robot a complex algorithm should be used to achieve maximum efficiency. The simplest, yet very
inefficient, is random moving algorithm, see robot.asm in Samples folder.
It is also possible to use a data table (just like for Traffic Lights), this can be good if robot always works in the same
surroundings.
Operand types:
REG: AX, BX, CX, DX, AH, AL, BL, BH, CH, CL, DH, DL, DI, SI, BP, SP.
Notes:
• When two operands are required for an instruction they are separated by comma. For example:
REG, memory
• When there are two operands, both operands must have the same size (except shift and rotate instructions).
For example:
AL, DL
DX, AX
m1 DB ?
AL, m1
m2 DW ?
AX, m2
memory, immediate
REG, immediate
memory, REG
REG, SREG
• Some examples contain macros, so it is advisable to use Shift + F8 hot key to Step Over (to make macro
code execute at maximum speed set step delay to zero), otherwise emulator will step through each
instruction of a macro. Here is an example that uses PRINTN macro:
•
• #make_COM#
• include 'emu8086.inc'
• ORG 100h
• MOV AL, 1
• MOV BL, 2
• PRINTN 'Hello World!' ; macro.
• MOV CL, 3
• PRINTN 'Welcome!' ; macro.
RET
38
These marks are used to show the state of the flags:
Some instructions generate exactly the same machine code, so disassembler may have a problem decoding to
your original code. This is especially important for Conditional Jump instructions (see "Program Flow
Control" in Tutorials for more information).
Example:
MOV AX, 15 ; AH = 00, AL = 0Fh
AAA ; AH = 01, AL = 05
RET
C Z S O P A
r ? ? ? ? r
ASCII Adjust before Division. Prepares two BCD values for division.
Algorithm:
• AL = (AH * 10) + AL
• AH = 0
Algorithm:
AAM No operands • AH = AL / 10
• AL = remainder
Example:
MOV AL, 15 ; AL = 0Fh
39
AAM ; AH = 01, AL = 05
RET
C Z S O P A
? r r ? r ?
Algorithm:
Example:
MOV AX, 02FFh ; AH = 02, AL = 0FFh
AAS ; AH = 01, AL = 09
RET
C Z S O P A
r ? ? ? ? r
Add.
REG, Algorithm:
memory
memory, operand1 = operand1 + operand2
REG
ADD REG, REG Example:
memory, MOV AL, 5 ; AL = 5
immediate ADD AL, -3 ; AL = 2
REG, RET
immediate C Z S O P A
r r r r r r
REG, Logical AND between all bits of two operands. Result is stored in operand1.
memory
AND memory, These rules apply:
REG
40
memory, 1 AND 0 = 0
immediate 0 AND 1 = 0
REG, 0 AND 0 = 0
immediate
Example:
MOV AL, 'a' ; AL = 01100001b
AND AL, 11011111b ; AL = 01000001b ('A')
RET
C Z S O P
0 r r 0 r
Transfers control to procedure, return address is (IP) is pushed to stack. 4-byte address may be
entered in this form: 1234h:5678h, first value is a segment second value is an offset (this is a
far call, so CS is also pushed to stack).
Example:
#make_COM#
ORG 100h ; for COM file.
procedure
name CALL p1
label ADD AX, 1
CALL RET ; return to OS.
4-byte
address
p1 PROC ; procedure declaration.
MOV AX, 1234h
RET ; return to caller.
p1 ENDP
C Z S O P A
unchanged
CLC No operands CF = 0
C
0
Clear Direction flag. SI and DI will be incremented by chain instructions: CMPSB, CMPSW,
LODSB, LODSW, MOVSB, MOVSW, STOSB, STOSW.
CLD No operands Algorithm:
DF = 0
D
41
if CF = 1 then CF = 0
CMC No operands if CF = 0 then CF = 1
C
r
Compare.
Algorithm:
operand1 - operand2
REG,
memory result is not stored anywhere, flags are set (OF, SF, ZF, AF, PF, CF) according to result.
memory,
REG Example:
CMP REG, REG MOV AL, 5
memory, MOV BL, 5
immediate CMP AL, BL ; AL = 5, ZF = 1 (so equal!)
REG, RET
immediate
C Z S O P A
r r r r r r
Example:
see cmpsw.asm in Samples.
C Z S O P A
r r r r r r
Example:
MOV AL, 0Fh ; AL = 0Fh (15)
DAA ; AL = 15h
RET
C Z S O P A
r r r r r r
Example:
MOV AL, 0FFh ; AL = 0FFh (-1)
DAS ; AL = 99h, CF = 1
RET
C Z S O P A
r r r r r r
REG Decrement.
DEC memory Algorithm:
43
Example:
MOV AL, 255 ; AL = 0FFh (255 or -1)
DEC AL ; AL = 0FEh (254 or -2)
RET
Z S O P A
r r r r r
CF - unchanged!
Unsigned divide.
Algorithm:
when operand is a byte:
AL = AX / operand
AH = remainder (modulus)
when operand is a word:
AX = (DX AX) / operand
REG DX = remainder (modulus)
DIV memory Example:
MOV AX, 203 ; AX = 00CBh
MOV BL, 4
DIV BL ; AL = 50 (32h), AH = 3
RET
C Z S O P A
? ? ? ? ? ?
Example:
MOV AX, 5
HLT No operands HLT
C Z S O P A
unchanged
Signed divide.
Algorithm:
when operand is a byte:
AL = AX / operand
AH = remainder (modulus)
when operand is a word:
AX = (DX AX) / operand
REG DX = remainder (modulus)
IDIV memory Example:
MOV AX, -203 ; AX = 0FF35h
MOV BL, 4
IDIV BL ; AL = -50 (0CEh), AH = -3 (0FDh)
RET
C Z S O P A
? ? ? ? ? ?
Signed multiply.
Algorithm:
when operand is a byte:
AX = AL * operand.
REG when operand is a word:
IMUL memory (DX AX) = AX * operand.
Example:
MOV AL, -2
MOV BL, -4
IMUL BL ; AX = 8
RET
44
C Z S O P A
r ? ? r ? ?
CF=OF=0 when result fits into operand of IMUL.
Increment.
Algorithm:
operand = operand + 1
REG Example:
INC memory MOV AL, 4
INC AL ; AL = 5
RET
Z S O P A
r r r r r
CF - unchanged!
Algorithm:
if OF = 1 then INT 4
Example:
INTO No operands
; -5 - 127 = -132 (not in -128..127)
; the result of SUB is wrong (124),
; so OF = 1 is set:
MOV AL, -5
SUB AL, 127 ; AL = 7Ch (124)
INTO ; process error.
RET
IRET No operands
45
Algorithm:
Pop from stack:
o IP
o CS
o flags register
C Z S O P A
popped
Short Jump if first operand is Above second operand (as set by CMP instruction). Unsigned.
Algorithm:
if (CF = 0) and (ZF = 0) then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 250
CMP AL, 5
JA label JA label1
PRINT 'AL is not above 5'
JMP exit
label1:
PRINT 'AL is above 5'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Above or Equal to second operand (as set by CMP instruction).
Unsigned.
Algorithm:
if CF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 5
CMP AL, 5
JAE label JAE label1
PRINT 'AL is not above or equal to 5'
JMP exit
label1:
PRINT 'AL is above or equal to 5'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Below second operand (as set by CMP instruction). Unsigned.
Algorithm:
if CF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
JB label ORG 100h
MOV AL, 1
CMP AL, 5
JB label1
PRINT 'AL is not below 5'
JMP exit
label1:
PRINT 'AL is below 5'
46
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Below or Equal to second operand (as set by CMP instruction).
Unsigned.
Algorithm:
if CF = 1 or ZF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 5
CMP AL, 5
JBE label JBE label1
PRINT 'AL is not below or equal to 5'
JMP exit
label1:
PRINT 'AL is below or equal to 5'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Equal to second operand (as set by CMP instruction).
Signed/Unsigned.
Algorithm:
if ZF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 5
CMP AL, 5
JE label JE label1
PRINT 'AL is not equal to 5.'
JMP exit
label1:
PRINT 'AL is equal to 5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Greater then second operand (as set by CMP instruction).
Signed.
Algorithm:
if (ZF = 0) and (SF = OF) then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 5
CMP AL, -5
JG label JG label1
PRINT 'AL is not greater -5.'
JMP exit
label1:
PRINT 'AL is greater -5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Greater or Equal to second operand (as set by CMP instruction).
Signed.
Algorithm:
if SF = OF then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, -5
JGE label JGE label1
PRINT 'AL < -5'
JMP exit
label1:
PRINT 'AL >= -5'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Less then second operand (as set by CMP instruction). Signed.
JL label
Algorithm:
48
Short Jump if first operand is Less or Equal to second operand (as set by CMP instruction).
Signed.
Algorithm:
if SF <> OF or ZF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, -2
CMP AL, 5
JLE label JLE label1
PRINT 'AL > 5.'
JMP exit
label1:
PRINT 'AL <= 5.'
exit:
RET
C Z S O P A
unchanged
Unconditional Jump. Transfers control to another part of the program. 4-byte address may be
entered in this form: 1234h:5678h, first value is a segment second value is an offset.
Algorithm:
always jump
Example:
include 'emu8086.inc'
#make_COM#
label ORG 100h
4-byte MOV AL, 5
JMP JMP label1 ; jump over 2 lines!
address
PRINT 'Not Jumped!'
MOV AL, 0
label1:
PRINT 'Got Here!'
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Above second operand (as set by CMP instruction).
Unsigned.
Algorithm:
JNA label if CF = 1 or ZF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
49
ORG 100h
MOV AL, 2
CMP AL, 5
JNA label1
PRINT 'AL is above 5.'
JMP exit
label1:
PRINT 'AL is not above 5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Above and Not Equal to second operand (as set by CMP
instruction). Unsigned.
Algorithm:
if CF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, 5
JNAE label JNAE label1
PRINT 'AL >= 5.'
JMP exit
label1:
PRINT 'AL < 5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Below second operand (as set by CMP instruction).
Unsigned.
Algorithm:
if CF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 7
CMP AL, 5
JNB label JNB label1
PRINT 'AL < 5.'
JMP exit
label1:
PRINT 'AL >= 5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Below and Not Equal to second operand (as set by CMP
instruction). Unsigned.
Algorithm:
if (CF = 0) and (ZF = 0) then jump
JNBE label Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 7
50
CMP AL, 5
JNBE label1
PRINT 'AL <= 5.'
JMP exit
label1:
PRINT 'AL > 5.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Equal to second operand (as set by CMP instruction).
Signed/Unsigned.
Algorithm:
if ZF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, 3
JNE label JNE label1
PRINT 'AL = 3.'
JMP exit
label1:
PRINT 'Al <> 3.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Greater then second operand (as set by CMP instruction).
Signed.
Algorithm:
if (ZF = 1) and (SF <> OF) then jump
Example:
include 'emu8086.inc'
JNG label
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, 3
JNG label1
PRINT 'AL > 3.'
51
JMP exit
label1:
PRINT 'Al <= 3.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Greater and Not Equal to second operand (as set by CMP
instruction). Signed.
Algorithm:
if SF <> OF then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, 3
JNGE label JNGE label1
PRINT 'AL >= 3.'
JMP exit
label1:
PRINT 'Al < 3.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Less then second operand (as set by CMP instruction).
Signed.
Algorithm:
if SF = OF then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, -3
JNL label JNL label1
PRINT 'AL < -3.'
JMP exit
label1:
PRINT 'Al >= -3.'
exit:
RET
C Z S O P A
unchanged
Short Jump if first operand is Not Less and Not Equal to second operand (as set by CMP
instruction). Signed.
Algorithm:
if (SF = OF) and (ZF = 0) then jump
Example:
include 'emu8086.inc'
JNLE label
#make_COM#
ORG 100h
MOV AL, 2
CMP AL, -3
JNLE label1
PRINT 'AL <= -3.'
JMP exit
52
label1:
PRINT 'Al > -3.'
exit:
RET
C Z S O P A
unchanged
Algorithm:
if OF = 0 then jump
Example:
; -5 - 2 = -7 (inside -128..127)
; the result of SUB is correct,
; so OF = 0:
include 'emu8086.inc'
#make_COM#
ORG 100h
JNO label MOV AL, -5
SUB AL, 2 ; AL = 0F9h (-7)
JNO label1
PRINT 'overflow!'
JMP exit
label1:
PRINT 'no overflow.'
exit:
RET
C Z S O P A
unchanged
Short Jump if No Parity (odd). Only 8 low bits of result are checked. Set by CMP, SUB, ADD,
TEST, AND, OR, XOR instructions.
Algorithm:
if PF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000111b ; AL = 7
JNP label OR AL, 0 ; just set flags.
JNP label1
PRINT 'parity even.'
JMP exit
label1:
PRINT 'parity odd.'
exit:
RET
C Z S O P A
unchanged
Short Jump if Not Signed (if positive). Set by CMP, SUB, ADD, TEST, AND, OR, XOR
instructions.
Algorithm:
if SF = 0 then jump
Example:
JNS label
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000111b ; AL = 7
OR AL, 0 ; just set flags.
53
JNS label1
PRINT 'signed.'
JMP exit
label1:
PRINT 'not signed.'
exit:
RET
C Z S O P A
unchanged
Short Jump if Not Zero (not equal). Set by CMP, SUB, ADD, TEST, AND, OR, XOR
instructions.
Algorithm:
if ZF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000111b ; AL = 7
OR AL, 0 ; just set flags.
JNZ label JNZ label1
PRINT 'zero.'
JMP exit
label1:
PRINT 'not zero.'
exit:
RET
C Z S O P A
unchanged
include 'emu8086.inc'
#make_COM#
org 100h
JO label MOV AL, -5
SUB AL, 127 ; AL = 7Ch (124)
JO label1
PRINT 'no overflow.'
JMP exit
label1:
PRINT 'overflow!'
exit:
RET
C Z S O P A
unchanged
Short Jump if Parity (even). Only 8 low bits of result are checked. Set by CMP, SUB, ADD,
TEST, AND, OR, XOR instructions.
Algorithm:
if PF = 1 then jump
JP label Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000101b ; AL = 5
54
Short Jump if Parity Even. Only 8 low bits of result are checked. Set by CMP, SUB, ADD,
TEST, AND, OR, XOR instructions.
Algorithm:
if PF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000101b ; AL = 5
OR AL, 0 ; just set flags.
JPE label JPE label1
PRINT 'parity odd.'
JMP exit
label1:
PRINT 'parity even.'
exit:
RET
C Z S O P A
unchanged
Short Jump if Parity Odd. Only 8 low bits of result are checked. Set by CMP, SUB, ADD,
TEST, AND, OR, XOR instructions.
Algorithm:
if PF = 0 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 00000111b ; AL = 7
OR AL, 0 ; just set flags.
JPO label JPO label1
PRINT 'parity even.'
JMP exit
label1:
PRINT 'parity odd.'
exit:
RET
C Z S O P A
unchanged
Short Jump if Signed (if negative). Set by CMP, SUB, ADD, TEST, AND, OR, XOR
instructions.
Algorithm:
if SF = 1 then jump
Example:
JS label include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 10000000b ; AL = -128
OR AL, 0 ; just set flags.
JS label1
55
Short Jump if Zero (equal). Set by CMP, SUB, ADD, TEST, AND, OR, XOR instructions.
Algorithm:
if ZF = 1 then jump
Example:
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AL, 5
CMP AL, 5
JZ label JZ label1
PRINT 'AL is not equal to 5.'
JMP exit
label1:
PRINT 'AL is equal to 5.'
exit:
RET
C Z S O P A
unchanged
AH bit: 7 6 5 4 3 2 1 0
LAHF No operands [SF] [ZF] [0] [AF] [0] [PF] [1] [CF]
bits 1, 3, 5 are reserved.
C Z S O P A
unchanged
Example:
#make_COM#
ORG 100h
m DW 1234h
DW 5678h
END
Example:
#make_COM#
ORG 100h
m DW 1234h
END
Example:
#make_COM#
ORG 100h
m DW 1234h
DW 5678h
END
LEA SI, a1
MOV CX, 5
MOV AH, 0Eh
m: LODSB
INT 10h
57
LOOP m
RET
a1 DB 'H', 'e', 'l', 'l', 'o'
C Z S O P A
unchanged
RET
Decrease CX, jump to label if CX not zero and Equal (ZF = 1).
Algorithm:
• CX = CX - 1
• if (CX <> 0) and (ZF = 1) then
o jump
LOOPE label else
o no jump, continue
Example:
; Loop until result fits into AL alone,
; or 5 times. The result will be over 255
; on third loop (100+100+100),
; so loop will exit.
58
include 'emu8086.inc'
#make_COM#
ORG 100h
MOV AX, 0
MOV CX, 5
label1:
PUTC '*'
ADD AX, 100
CMP AH, 0
LOOPE label1
RET
C Z S O P A
unchanged
Decrease CX, jump to label if CX not zero and Not Equal (ZF = 0).
Algorithm:
• CX = CX - 1
• if (CX <> 0) and (ZF = 0) then
o jump
else
o no jump, continue
Example:
; Loop until '7' is found,
; or 5 times.
include 'emu8086.inc'
#make_COM#
LOOPNE label ORG 100h
MOV SI, 0
MOV CX, 5
label1:
PUTC '*'
MOV AL, v1[SI]
INC SI ; next byte (SI=SI+1).
CMP AL, 7
LOOPNE label1
RET
v1 db 9, 8, 7, 6, 5
C Z S O P A
unchanged
v1 db 9, 8, 7, 6, 5
C Z S O P A
unchanged
o SI = SI - 1
o DI = DI - 1
Example:
#make_COM#
ORG 100h
LEA SI, a1
LEA DI, a2
MOV CX, 5
REP MOVSB
RET
a1 DB 1,2,3,4,5
a2 DB 5 DUP(0)
C Z S O P A
unchanged
RET
a1 DW 1,2,3,4,5
a2 DW 5 DUP(0)
C Z S O P A
unchanged
Unsigned multiply.
Algorithm:
when operand is a byte:
AX = AL * operand.
when operand is a word:
(DX AX) = AX * operand.
REG Example:
MUL memory MOV AL, 200 ; AL = 0C8h
MOV BL, 4
MUL BL ; AX = 0320h (800)
RET
C Z S O P A
r ? ? r ? ?
CF=OF=0 when high section of the result is zero.
Example:
MOV AL, 5 ; AL = 05h
NEG AL ; AL = 0FBh (-5)
NEG AL ; AL = 05h (5)
RET
C Z S O P A
r r r r r r
No Operation.
Algorithm:
• Do nothing
Example:
; do nothing, 3 times:
NOP
NOP No operands
NOP
NOP
RET
C Z S O P A
unchanged
Logical OR between all bits of two operands. Result is stored in first operand.
These rules apply:
REG, 1 OR 1 = 1
memory 1 OR 0 = 1
memory, 0 OR 1 = 1
REG 0 OR 0 = 0
OR REG, REG
memory, Example:
immediate MOV AL, 'A' ; AL = 01000001b
REG, OR AL, 00100000b ; AL = 01100001b ('a')
immediate RET
C Z S O P A
0 r r 0 r ?
Example:
im.byte, AL MOV AX, 0FFFh ; Turn on all
im.byte, AX OUT 4, AX ; traffic lights.
OUT
DX, AL
DX, AX MOV AL, 100b ; Turn on the third
OUT 7, AL ; magnet of the stepper-motor.
C Z S O P A
unchanged
62
Pop all general purpose registers DI, SI, BP, SP, BX, DX, CX, AX from the stack.
SP value is ignored, it is Popped but not set to SP register).
Push all general purpose registers AX, CX, DX, BX, SP, BP, SI, DI in the stack.
Original value of SP register (before PUSHA) is used.
PUSHA No operands Note: this instruction works only on 80186 CPU and later!
Algorithm:
• PUSH AX
• PUSH CX
63
• PUSH DX
• PUSH BX
• PUSH SP
• PUSH BP
• PUSH SI
• PUSH DI
C Z S O P A
unchanged
Algorithm:
• SP = SP - 2
PUSHF No operands
• SS:[SP] (top of the stack) = flags
C Z S O P A
unchanged
Rotate operand1 left through Carry Flag. The number of rotates is set by operand2.
When immediate is greater then 1, assembler generates several RCL xx, 1 instructions
because 8086 has machine code only for this instruction (the same principle works for all other
shift/rotate instructions).
Algorithm:
memory, shift all bits left, the bit that goes off is set to CF and previous value of CF is inserted
immediate to the right-most position.
REG,
RCL immediate Example:
STC ; set carry (CF=1).
memory, CL MOV AL, 1Ch ; AL = 00011100b
REG, CL RCL AL, 1 ; AL = 00111001b, CF=0.
RET
C O
r r
OF=0 if first operand keeps original sign.
Rotate operand1 right through Carry Flag. The number of rotates is set by operand2.
Algorithm:
shift all bits right, the bit that goes off is set to CF and previous value of CF is
inserted to the left-most position.
memory,
immediate Example:
REG, STC ; set carry (CF=1).
RCR immediate MOV AL, 1Ch ; AL = 00011100b
RCR AL, 1 ; AL = 10001110b, CF=0.
memory, CL RET
REG, CL
C O
r r
OF=0 if first operand keeps original sign.
Z
r
Algorithm:
check_cx:
if CX <> 0 then
• do following chain instruction
• CX = CX - 1
chain • if ZF = 1 then:
REPE o go back to check_cx
instruction
else
o exit from REPE cycle
else
• exit from REPE cycle
Example:
see cmpsb.asm in Samples.
Z
r
Repeat following CMPSB, CMPSW, SCASB, SCASW instructions while ZF = 0 (result is Not
Equal), maximum CX times.
Algorithm:
check_cx:
if CX <> 0 then
• do following chain instruction
• CX = CX - 1
chain
REPNE
instruction • if ZF = 0 then:
o go back to check_cx
else
o exit from REPNE cycle
else
• exit from REPNE cycle
Z
r
Repeat following CMPSB, CMPSW, SCASB, SCASW instructions while ZF = 0 (result is Not
Zero), maximum CX times.
Algorithm:
check_cx:
if CX <> 0 then
• do following chain instruction
• CX = CX - 1
REPNZ
chain • if ZF = 0 then:
instruction o go back to check_cx
else
o exit from REPNZ cycle
else
• exit from REPNZ cycle
Z
r
if CX <> 0 then
• do following chain instruction
• CX = CX - 1
• if ZF = 1 then:
o go back to check_cx
else
o exit from REPZ cycle
else
• exit from REPZ cycle
Z
r
Example:
#make_COM#
ORG 100h ; for COM file.
No operands
RET or even CALL p1
immediate
ADD AX, 1
ROR
66
immediate Algorithm:
REG, shift all bits right, the bit that goes off is set to CF and the same bit is inserted to the
immediate left-most position.
Example:
memory, CL MOV AL, 1Ch ; AL = 00011100b
REG, CL ROR AL, 1 ; AL = 00001110b, CF=0.
RET
C O
r r
OF=0 if first operand keeps original sign.
AH bit: 7 6 5 4 3 2 1 0
SAHF No operands [SF] [ZF] [0] [AF] [0] [PF] [1] [CF]
bits 1, 3, 5 are reserved.
C Z S O P A
r r r r r r
C Z S O P A
r r r r r r
memory, CL RET
REG, CL
C O
r r
OF=0 if first operand keeps original sign.
memory, CL RET
REG, CL
C O
r r
OF=0 if first operand keeps original sign.
C
1
Set Direction flag. SI and DI will be decremented by chain instructions: CMPSB, CMPSW,
LODSB, LODSW, MOVSB, MOVSW, STOSB, STOSW.
Algorithm:
STD No operands DF = 1
D
1
REP STOSB
RET
a1 DB 5 dup(0)
C Z S O P A
unchanged
LEA DI, a1
MOV AX, 1234h
MOV CX, 5
REP STOSW
RET
a1 DW 5 dup(0)
69
C Z S O P A
unchanged
Subtract.
Algorithm:
REG, operand1 = operand1 - operand2
memory
memory, Example:
REG MOV AL, 5
SUB REG, REG SUB AL, 1 ; AL = 4
memory,
immediate RET
REG,
immediate C Z S O P A
r r r r r r
Logical AND between all bits of two operands for flags only. These flags are effected: ZF, SF,
PF. Result is not stored anywhere.
REG, Example:
memory MOV AL, 5
XCHG memory, MOV AH, 2
REG XCHG AL, AH ; AL = 2, AH = 5
REG, REG XCHG AL, AH ; AL = 5, AH = 2
RET
C Z S O P A
unchanged
RET
70
Logical XOR (Exclusive OR) between all bits of two operands. Result is stored in first
operand.
These rules apply:
REG, 1 XOR 1 = 0
memory 1 XOR 0 = 1
memory, 0 XOR 1 = 1
REG 0 XOR 0 = 0
XOR REG, REG
memory, Example:
immediate MOV AL, 00000111b
REG, XOR AL, 00000010b ; AL = 00000101b
immediate RET
C Z S O P A
0 r r 0 r ?