
                           
                                
                                              
                                     
                                   

                              flat assembler 1.49
                              Programmer's Manual


Table of contents


Chapter 1  Introduction

        1.1  Compiler overview
        1.1.1  System requirements
        1.1.2  Executing compiler from command line
        1.1.3  Compiler messages
        1.1.4  Output formats

        1.2  Assembly syntax
        1.2.1  Instruction syntax
        1.2.2  Data definitions
        1.2.3  Constants and labels
        1.2.4  Numerical expressions
        1.2.5  Jumps and calls
        1.2.6  Size settings

Chapter 2  Instruction set

        2.1  Intel Architecture instructions
        2.1.1  Data movement instructions
        2.1.2  Type conversion instructions
        2.1.3  Binary arithmetic instructions
        2.1.4  Decimal arithmetic instructions
        2.1.5  Logical instructions
        2.1.6  Control transfer instructions
        2.1.7  I/O instructions
        2.1.8  Strings operations
        2.1.9  Flag control instructions
        2.1.10  Conditional operations
        2.1.11  Miscellaneous instructions
        2.1.12  System instructions
        2.1.13  FPU instructions
        2.1.14  MMX instructions
        2.1.15  SSE instructions
        2.1.16  SSE2 instructions
        2.1.17  AMD 3DNow! instructions

        2.2  Control directives
        2.2.1  Repeating blocks of instructions
        2.2.2  Conditional assembly
        2.2.3  Other directives

        2.3  Preprocessor directives
        2.3.1  Including source files
        2.3.2  Symbolic constants
        2.3.3  Macroinstructions
        2.3.4  Structures

        2.4  Formatter directives
        2.4.1  MZ executable
        2.4.2  Portable Executable
        2.4.3  Common Object File Format
        2.4.4  Executable and Linkable Format


Chapter 1  Introduction


This chapter contains all the most important information you need to begin
using the flat assembler. If you are experienced assembly language programmer,
you should read at least this chapter before using this compiler.


1.1  Compiler overview

Flat assembler is a fast assembly language compiler for the Intel Architecture
processors, which does multiple passes to optimize the size of generated
machine code. It is self-compilable and versions for different operating
systems are provided. All the versions are designed to be used from the system
command line and they should not differ in behavior.


1.1.1  System requirements

All versions require the Intel Architecture 32-bit processor (at least 80386),
although they can produce programs for Intel Architecture 16-bit processors,
too. DOS version requires an OS compatible with MS DOS 2.0, Windows version
requires a Win32 console compatible with 3.1 version.


1.1.2  Executing compiler from command line

To execute flat assembler from the command line you need to provide two
parameters - first should be name of source file, second should be name of
destination file. After displaying short information about the program name
and version, compiler will read the data from source file and compile it.
When the compilation is successful, compiler will write the generated code
to the destination file and display the summary of compilation process;
otherwise it will display the information about error that occurred.
  The source file should be a text file, and can be created in any text
editor. Line breaks are accepted in both DOS and Unix standards, tabulators
are treated as spaces.
  There are no additional command line options, flat assembler requires only
the source code to include the information it really needs. For example, to
specify output format you specify it by using the "format" directive at the
beginning of source.


1.1.3  Compiler messages

As it is stated above, after the successful compilation compiler displays the
compilation summary. It includes the information of how many passes was done,
how much time it took, and how many bytes were written into destination file.
Here is an example of the compilation summary:

flat assembler  version 1.49
38 passes, 5.3 seconds, 77824 bytes.

In case of error during the compilation process, program will display an error
message. For example, when compiler can't find the input file, it will display
the following message:

flat assembler  version 1.49
error: source file not found.

If the error is connected with a specific part of source code, the source line
that caused the error will be also displayed. Also placement of this line in
the source is given to help you finding this error, for example:

flat assembler  version 1.49
example.asm [3]:
        mob     ax,1
error: illegal instruction.

It means that in the third line of the "example.asm" file compiler has
encountered an unrecognized instruction. When the line that caused error
contains a macroinstruction, also the line in macrinstruction definition
that generated the erroneous instruction is displayed:

flat assembler  version 1.49
example.asm [6]:
        stoschar 7
example.asm [3] stoschar [1]:
        mob     al,char
error: illegal instruction.

It means that the macroinstruction in the sixth line of the "example.asm" file
generated an unrecognized instruction with the first line of its definition.


1.1.4  Output formats

By default, when there is no "format" directive in source file, flat
assembler simply puts generated instruction codes into output, creating this
way flat binary file. By default it generates 16-bit code, but you can always
turn it into the 16-bit or 32-bit mode by using "use16" or "use32" directive.
Some of the output formats switch into 32-bit mode, when selected - more
information about formats which you can choose can be found in 2.4.
  All output code is always in the order in which it was entered into the
source file.


1.2  Assembly syntax

The information provided below is intended mainly for the assembler
programmers that have been using some other assembly compilers before.
If you are beginner, you should look for the assembly programming tutorials.
  Flat assembler by default uses the Intel syntax for the assembly
instructions, although you can customize it using the preprocessor
capabilities (macroinstructions and symbolic constants). It also has its own
set of the directives - the instructions for compiler.
  All symbols defined inside the sources are case-sensitive.


1.2.1  Instruction syntax

Instructions in assembly language are separated by line breaks, and one
instruction is expected to fill the one line of text. If line contains
a semicolon (except for the semicolons in quoted strings), the rest of this
line is the comment and compiler ignores it. If line contains "\"
characters, the next line is attached at this point. Line should not contain
anything but comments (started with semicolon) after the "\" character.
  Every instruction consists of the mnemonic and the various number of
operands, separated with commas. The operand can be register, immediate value
or a data addressed in memory, it can also be preceded by size operator to
define or override its size (table 1.1). Names of available registers you can
find in table 1.2, their sizes cannot be overridden. Immediate value can be
specified by any numerical expression.
  When operand is a data in memory, the address of that data (also any
numerical expression, but it may contain registers) should be enclosed in
square brackets or preceded by "ptr" operator. For example instruction
"mov eax,3" will put the immediate value 3 into the EAX register, instruction
"mov eax,[7]" will put the 32-bit value from the address 7 into EAX and the
instruction "mov byte [7],3" will put the immediate value 3 into the byte at
address 7, it can also be written as "mov byte ptr 7,3". To specify which
segment register should be used for addressing, segment register name followed
by a colon should be put just before the address value (inside the square
brackets or after the "ptr" operator).

   Table 1.1  Size Operators
  Ŀ
   Operator  Bits  Bytes 
  ͵
   byte      8     1     
   word      16    2     
   dword     32    4     
   fword     48    6     
   pword     48    6     
   qword     64    8     
   tword     80    10    
   dqword    128   16    
  

   Table 1.2  Registers
  Ŀ
   Type     Bits                                                 
  ͵
            8     al    cl    dl    bl    ah    ch    dh    bh   
   General  16    ax    cx    dx    bx    sp    bp    si    di   
            32    eax   ecx   edx   ebx   esp   ebp   esi   edi  
  Ĵ
   Segment  16    es    cs    ss    ds    fs    gs               
  Ĵ
   Control  32    cr0         cr2   cr3   cr4                    
  Ĵ
   Debug    32    dr0   dr1   dr2   dr3               dr6   dr7  
  Ĵ
   FPU      80    st0   st1   st2   st3   st4   st5   st6   st7  
  Ĵ
   MMX      64    mm0   mm1   mm2   mm3   mm4   mm5   mm6   mm7  
  Ĵ
   SSE      128   xmm0  xmm1  xmm2  xmm3  xmm4  xmm5  xmm6  xmm7 
  


1.2.2  Data definitions

To define data or reserve a space for it, use one of the directives listed in
table 1.3. The data definition directive should be followed by one or more of
numerical expressions, separated with commas. These expression define the
values for data cells of size depending on which directive is used. For
example "db 1,2,3" will define the three bytes of values 1, 2 and 3
respectively.
  The "db" and "du" directives also accept the quoted string values of any
length, which will be converted into chain of bytes when "db" is used and into
chain of words with zeroed high byte when "du" is used. For example "db 'abc'"
will define the three bytes of values 61, 62 and 63.
  The "dp" directive and its synonym "df" accept the values consisting of two
numerical expressions separated with colon, the first value will become the
high word and the second value will become the low double word of the far
pointer value. Also "dd" accepts such pointers consisting of two word values
separated with colon. The "dt" directive accepts only floating point values
and creates data in FPU double extended precision format.
  The "file" is a special directive and its syntax is different. This
directive includes a chain of bytes from file and it should be followed by the
quoted file name, then optionally numerical expression specifying offset in
file preceded by the colon, then - also optionally - comma and numerical
expression specifying count of bytes to include (if no count is specified, all
data up to the end of file is included).
  The data reservation directive should be followed by only one numerical
expression, and this value defines how many cells of the specified size should
be reserved. All data definition directives also accept the "?" value, which
means that this cell should not be initialized to any value and the effect is
the same as by using the data reservation directive. The uninitialized data
may not be included in the output file, so its values should be always
considered unknown.

   Table 1.3  Data directives
  Ŀ
   Size     Define  Reserve 
   (bytes)  data    data    
  ͵
   1        db      rb      
            file            
  Ĵ
   2        dw      rw      
            du              
  Ĵ
   4        dd      rd      
  Ĵ
   6        dp      rp      
            df      rf      
  Ĵ
   8        dq      rq      
  Ĵ
   10       dt      rt      
  


1.2.3  Constants and labels

In the numerical expressions you can also use constants or labels instead of
numbers. To define the constant or label you should use the specific
directives. Each label can be defined only once and it is accessible from the
any place of source (even before it was defined). Constant can be redefined
many times, but in this case it is accessible only after it was defined, and
is always equal to the value from last definition before the place where it's
used. When constant is defined only once in source, it's - like the label -
accessible from anywhere.
  The definition of constant consists of name of the constant followed by the
"=" character and numerical expression, which after calculation will become
the value of constant. This value is always calculated at the time the
constant is defined. For example you can define "count" constant by using the
directive "count = 17", and then use it in the assembly instructions, like
"mov cx,count" - which will become "mov cx,17" during the compilation process.
  There are different ways to define labels. The simplest is to follow the
name of label by the colon, this directive can even be followed by the other
instruction in the same line. It defines the label whose value is equal to
offset of the point where it's defined. This method is usually used to label
the places in code. The other way is to follow the name of label (without a
colon) by some data directive. It defines the label with value equal to
offset of the beginning of defined data, and remembered as a label for data
with cell size as specified for that data directive in table 1.3.
  The label can be treated as constant of value equal to offset of labelled
code or data. For example when you define data using the labelled directive
"char db 224", to put the offset of this data into BX register you should use
"mov bx,char" instruction, and to put the value of byte addressed by "char"
label to DL register, you should use "mov dl,[char]" (or "mov dl,ptr char").
But when you try to assemble "mov ax,[char]", it will cause an error, because
fasm compares the sizes of operands, which should be equal. You can force
assembling that instruction by using size override: "mov ax,word [char]", but
remember that this instruction will read the two bytes beginning at "char"
address, while it was defined as a one byte.
  The last and the most flexible way to define labels is to use "label"
directive. This directive should be followed by the name of label, then
optionally size operator (it can be preceded by a colon) and then - also
optionally "at" operator and the numerical expression defining the address at
which this label should be defined. For example "label wchar word at char"
will define a new label for the 16-bit data at the address of "char". Now the
instruction "mov ax,[wchar]" will be after compilation the same as
"mov ax,word [char]". If no address is specified, "label" directive defines
the label at current offset. Thus "mov [wchar],57568" will copy two bytes
while "mov [char],224" will copy one byte to the same address.
  The label whose name begins with dot is treated as local label, and its name
is attached to the name of last global label (with name beginning with
anything but dot) to make the full name of this label. So you can use the
short name (beginning with dot) of this label anywhere before the next global
label is defined, and in the other places you have to use the full name. Label
beginning with two dots are the exception - they are like global, but they
don't become the new prefix for local labels.
  The "@@" name means anonymous label, you can have defined many of them in
the source. Symbol "@b" (or equivalent "@r") references the nearest preceding
anonymous label, symbol "@f" references the nearest following anonymous label.
These special symbol are case-insensitive.
  The "load" directive allows to define constant with binary value loaded from
a file during the assembly process. This directive should be followed by the
name of constant, then optionally size operator, then "from" operator and
quoted file name, which can be also followed by a colon and numerical
expression specifying offset in file. The size operator has unusual meaning in
this case - it states how many bytes (up to 8) have to be loaded to form the
binary value of constant. If no size operator is specified, one byte is loaded
(thus value is in range from 0 to 255). With giving only the offset value
instead of quoted file name you can also load a value from the already
assembled code. The given offset should be a valid address in currently
generated code space, the loaded data cannot exceed current offset.


1.2.4  Numerical expressions

In the above examples all the numerical expressions were the simple numbers,
constants or labels. But they can be more complex, by using the arithmetical
or logical operators for calculations at compile time. All these operators
with their priority values are listed in table 1.4.
The operations with higher priority value will be calculated first, you can
of course change this behaviour by putting some parts of expression into
parenthesis. The "+", "-", "*" and "/" are standard arithmetical operations,
"mod" calculates the remainder from division. The "and", "or", "xor", "shl",
"shr" and "not" perform the same logical operations as assembly instructions
of those names. The "rva" is specific to PE output format and performs the
conversion of an address into the RVA.
  The numbers in the expression are by default treated as a decimal, binary
numbers should have the "b" letter attached at the end, octal number should
end with "o" letter, hexadecimal numbers should begin with "0x" characters
(like in C language) or with the "$" character (like in Pascal language) or
they should end with "h" letter. Also quoted string, when encountered in
expression, will be converted into number - the first character will become
the least significant byte of number.
  The numerical expression used as an address value can also contain any of
general registers used for addressing, they can be added and multiplied by
appropriate values, as it is allowed for Intel Architecture instructions.
  There are also two special symbols that can be used inside the numerical
expression. First is "$", which is always equal to the value of current
offset. Second is "%", which is the number of current repeat in parts of code
that are repeated using some special directives (see 2.2).
  Any numerical expression can also consist of single floating point value
(flat assembler does not allow any floating point operations at compilation
time) in the scientific notation, they can end with the "f" letter to be
recognized, otherwise they should contain at least one of the "." or "E"
characters. So "1.0", "1E0" and "1f" define the same floating point value,
while simple "1" defines an integer value.

   Table 1.4  Arithmetical and logical operators by priority
  Ŀ
   Priority  Operators    
  ͵
   0         +  -         
  Ĵ
   1         *  /         
  Ĵ
   2         mod          
  Ĵ
   3         and  or  xor 
  Ĵ
   4         shl  shr     
  Ĵ
   5         not          
  Ĵ
   6         rva          
  


1.2.5  Jumps and calls

The operand of any jump or call instruction can be preceded not only by the
size operator, but also by one of the operators specifying type of the jump:
"near" of "far". For example, when assembler is in 16-bit mode, instruction
"jmp dword [0]" will become the far jump and when assembler is in 32-bit mode,
it will become the near jump. To force this instruction to be treated
differently, use the "jmp near dword [0]" or "jmp far dword [0]" form.
  When operand of near jump is the immediate value, assembler will generate
the shortest variant of this jump instruction if possible (but won't create
32-bit instruction in 16-bit mode nor 16-bit instruction in 32-bit mode,
unless there is a size operator stating it). By specifying the size operator
you can force it to always generate long variant (for example "jmp word 0" in
16-bit mode and "jmp dword 0" in 32-bit mode) or to always generate short
variant and terminate with an error when it's impossible (for example
"jmp byte 0").


1.2.6  Size settings

When instruction uses some memory addressing, by default the shorter 8-bit
form is generated if only address value fits in range, but it can be
overridden using the "word" or "dword" operator before the address inside the
square brackets (or after the "ptr" operator).
  Instructions "adc", "add", "and", "cmp", "or", "sbb", "sub" and "xor" with
first operand being 16-bit or 32-bit are by default generated in shortened
8-bit form when the second operand is immediate value fitting in the range
for signed 8-bit values. It also can be overridden by putting the "word" or
"dword" operator before the immediate value.
  Immediate value as an operand for "push" instruction without a size operator
is by default treated as a word value if assembler is in 16-bit mode and as a
double word value if assembler is in 32-bit mode, shorter 8-bit form of this
instruction is used if possible, "word" or "dword" size operator forces the
"push" instruction to be generated in longer form for specified size. "pushw"
and "pushd" mnemonics force assembler to generate 16-bit or 32-bit code
without forcing it to use the longer form of instruction.


Chapter 2  Instruction set


This chapter provides the detailed information about the instructions and
directives supported by flat assembler. Directives for defining constants and
labels were already discussed in 1.2.3, all other directives will be described
later in this chapter.


2.1  Intel Architecture instructions

In this section you can find both the information about the syntax and
purpose the assembly language instructions. If you need more technical
information, look for the Intel Architecture Software Developer's Manual.
  Assembly instructions consist of the mnemonic (instruction's name) and from
zero to three operands. If there are two or more operands, usually first is
the destination operand and second is the source operand. Each operand can be
register, memory or immediate value (see 1.2 for details about syntax of
operands). After description of each instruction there are provided examples
of different combinations of operands (if the instruction has any).
  Some instructions act as prefixes and can be followed by other instruction
in the same line, and there can be more than one prefix in a line. Each name
of the segment register is also a mnemonic of instruction prefix, altough it
is recommended to use segment overrides inside the square brackets instead of
these prefixes.


2.1.1  Data movement instructions

"mov" transfers a byte, word or double word from the source operand to the
destination operand. It can transfer data between general registers, from
the general register to memory, or from memory to general register, but it
cannot move from memory to memory. It can also transfer an immediate value to
general register or memory, segment register to general register or memory,
general register or memory to segment register, control or debug register to
general register and general register to control or debug register. The "mov"
can be assembled only if the size of source operand and size of destination
operand are the same. Below are the examples for each of the allowed
combinations:

    mov bx,ax       ; general register to general register
    mov [char],al   ; general register to memory
    mov bl,[char]   ; memory to general register
    mov dl,32       ; immediate value to general register
    mov [char],32   ; immediate value to memory
    mov ax,ds     ; segment register to general register
    mov [bx],ds     ; segment register to memory
    mov ds,ax       ; general register to segment register
    mov ds,[bx]     ; memory to segment register
    mov eax,cr0     ; control register to general register
    mov cr3,ebx     ; general register to control register

  "xchg" swaps the contents of two operands. It can swap two byte operands,
two word operands or two double word operands. Order of operands is not
important. The operands may be two general registers, or general register
with memory. For example:

    xchg ax,bx      ; swap two general registers
    xchg al,[char]  ; swap register with memory

  "push" decrements the stack frame pointer (ESP register), then transfers
the operand to the top of stack indicated by ESP. The operand can be memory,
general register, segment register or immediate value of word or double word
size. If operand is an immediate value and no size is specified, it is by
default treated as a word value if assembler is in 16-bit mode and as a double
word value if assembler is in 32-bit mode. "pushw" and "pushd" mnemonics are
variants of this instruction that store the values of word or double word size
respectively. If more operands follow in the same line (separated only with
spaces, not commas), compiler will assemble chain of the "push" instructions
with these operands. The examples are with single operands:

    push ax         ; store general register
    push es         ; store segment register
    push [bx]       ; store memory
    push 1000h      ; store immediate value

  "pusha" saves the contents of the eight general register on the stack.
This instruction has no operands. There are two version of this instruction,
one 16-bit and one 32-bit, assembler automatically generates the appropriate
version for current mode, but it can be overridden by using "pushaw" or
"pushad" mnemonic to always get the 16-bit or 32-bit version. The 16-bit
version of this instruction pushes general registers on the stack in the
following order: AX, CX, DX, BX, the initial value of SP before AX was pushed,
BP, SI and DI. The 32-bit version pushes equivalent 32-bit general registers
in the same order.
  "pop" transfers the word or double word at the current top of stack to the
destination operand, and then increments ESP to point to the new top of stack.
The operand can be memory, general register or segment register. "popw" and
"popd" mnemonics are variants of this instruction for restoring the values of
word or double word size respectively. If more operands separated with spaces
follow in the same line, compiler will assemble chain of the "pop"
instructions with these operands.

    pop bx          ; restore general register
    pop ds          ; restore segment register
    pop [si]        ; restore memory

  "popa" restores the registers saved on the stack by "pusha" instruction,
except for the saved value of SP (or ESP), which is ignored. This instruction
has no operands. To force assembling 16-bit or 32-bit version of this
instruction use "popaw" or "popad" mnemonic.


2.1.2  Type conversion instructions

The type conversion instructions convert bytes into words, words into double
words, and double words into quad words. These conversion can be done using
the sign extension or zero extension. The sign extension fills the extra bits
of the larger item with the value of the sign bit of the smaller item, the
zero extension simply fills them with zeros.
  "cwd" and "cdq" double the size of value AX or EAX register respectively
and store the extra bits into the DX or EDX register. The conversion is done
using the sign extension. These instructions have no operands.
  "cbw" extends the sign of the byte in AL throughout AX, and "cwde" extends
the sign of the word in AX throughout EAX. These instruction also have no
operands.
  "movsx" converts a byte to word or double word and a word to double word
using the sign extension. "movzx" does the same, but it uses the zero
extension. The source operand can be general register or memory, while the
destination operand must be a general register. For example:

    movsx ax,al         ; byte register to word register
    movsx edx,dl        ; byte register to double word register
    movsx eax,ax        ; word register to double word register
    movsx ax,byte [bx]  ; byte memory to word register
    movsx edx,byte [bx] ; byte memory to double word register
    movsx eax,word [bx] ; word memory to double word register


2.1.3  Binary arithmetic instructions

"add" replaces the destination operand with the sum of the source and
destination operands and sets CF if overflow has occurred. The operands may
be bytes, words or double words. The destination operand can be general
register or memory, the source operand can be general register or immediate
value, it can also be memory if the destination operand is register.

    add ax,bx       ; add register to register
    add ax,[si]     ; add memory to register
    add [di],al     ; add register to memory
    add al,48       ; add immediate value to register
    add [char],48   ; add immediate value to memory

  "adc" sums the operands, adds one if CF is set, and replaces the destination
operand with the result. Rules for the operands are the same as for the "add"
instruction. An "add" followed by multiple "adc" instructions can be used to
add numbers longer than 32 bits.
  "inc" adds one to the operand. It does not affect CF. The operand can be
general register or memory, size of operand can be byte, word or double word.

    inc ax          ; increment register by one
    inc byte [bx]   ; increment memory by one

  "sub" subtracts the source operand from the destination operand and replaces
the destination operand with the result. If a borrow is required, the CF is
set. Rules for the operands are the same as for the "add" instruction.
  "sbb" subtracts the source operand from the destination operand, subtracts
one if CF is set, and stores the result to the destination operand. Rules for
the operands are the same as for the "add" instruction. A "sub" followed by
multiple "sbb" instructions may be used to subtract numbers longer than 32
bits.
  "dec" subtracts one from the operand. It does not affect CF. Rules for the
operand are the same as for the "inc" instruction.
  "cmp" subtracts the source operand from the destination operand. It updates
the flags as the "sub" instruction, but does not alter the source and
destination operands. Rules for the operands are the same as for the "sub"
instruction.
  "neg" subtracts a signed integer operand from zero. The effect of this
instructon is to reverse the sign of the operand from positive to negative or
from negative to positive. Rules for the operand are the same as for the "inc"
instruction.
  "xadd" exchanges the destination operand with the source operand, then loads
the sum of the two values into the destination operand. Rules for the operands
are the same as for the "add" instruction.
  All the above binary arithmetic instruction update SF, ZF, PF and OF flags.
SF is always set to the same value as the sign bit of the result, ZF is set
when all bits of result are zero, PF is set when low order eight bits of
result contain an even number of set bits, OF is set if result is too large a
positive number or too small a negative number (excluding sign bit) to fit in
destination operand.
  "mul" performs an unsigned multiplication of the operand and the
accumulator. If the operand is a byte, the processor multiplies it by the
contents of AL and returns the 16-bit result to AH and AL. If the operand is a
word, the processor multiplies it by the contents of AX and returns the 32-bit
result to DX and AX. If the operand is a double word, the processor multiplies
it by the contents of EAX and returns the 64-bit result in EDX and EAX. "mul"
sets CF and OF when the upper half of the result is nonzero, otherwise they
are cleared. Rules for the operand are the same as for the "inc" instruction.
  "imul" performs a signed multiplication operation. This instruction has
three variations. First has one operand and behaves in the same way as the
"mul" instruction. Second has two operands, in this case destination operand
is multiplied by the source operand and the result replaces the destination
operand. Destination operand must be a general register, it can be word or
double word, source operand can be general register, memory or immediate
value. The immediate value can be a byte, in this case processor automatically
does the sign extension to it before performing the multiplication. Third form
has three operands, the destination operand must be a general register, word
or double word in size, source operand can be general register or memory, and
third operand must be an immediate value. The source operand is multiplied by
the immediate value and the result is stored in the destination register.
All the three forms calculate the product to twice the size of operands and
set CF and OF when the upper half of the result is nonzero, but second and
third form truncate the product to the size of operands. So second and third
forms can be also used for unsigned operands because, whether the operands
are signed or unsigned, the lower half of the product is the same.
Below are the examples for all three forms:

    imul bl         ; accumulator by register
    imul word [si]  ; accumulator by memory
    imul bx,cx      ; register by register
    imul bx,[si]    ; register by memory
    imul bx,10      ; register by immediate value
    imul ax,bx,10   ; register by immediate value to register
    imul ax,[si],10 ; memory by immediate value to register

  "div" performs an unsigned division of the accumulator by the operand.
The dividend (the accumulator) is twice the size of the divisor (the operand),
the quotient and remainder have the same size as the divisor. If divisor is
byte, the dividend is taken from AX register, the quotient is stored in AL and
the remainder is stored in AH. If divisor is word, the upper half of dividend
is taken from DX, the lower half of dividend is taken from AX, the quotient is
stored in AX and the remainder is stored in DX. If divisor is double word,
the upper half of dividend is taken from EDX, the lower half of dividend is
taken from EAX, the quotient is stored in EAX and the remainder is stored in
EDX. Rules for the operand are the same as for the "mul" instruction.
  "idiv" performs a signed division of the accumulator by the operand.
It uses the same registers as the "div" instruction, and the rules for
the operand are the same.


2.1.4  Decimal arithmetic instructions

Decimal arithmetic is performed by combining the binary arithmetic
instructions (already described in the prior section) with the decimal
arithmetic instructions. The decimal arithmetic instructions are used to
adjust the results of a previous binary arithmetic operation to produce a
valid packed or unpacked decimal result, or to adjust the inputs to a
subsequent binary arithmetic operation so the operation will produce a valid
packed or unpacked decimal result.
  "daa" adjusts the result of adding two valid packed decimal operands in
AL. "daa" must always follow the addition of two pairs of packed decimal
numbers (one digit in each half-byte) to obtain a pair of valid packed
decimal digits as results. The carry flag is set if carry was needed.
This instruction has no operands.
  "das" adjusts the result of subtracting two valid packed decimal operands
in AL. "das" must always follow the subtraction of one pair of packed decimal
numbers (one digit in each half-byte) from another to obtain a pair of valid
packed decimal digits as results. The carry flag is set if a borrow was
needed. This instruction has no operands.
  "aaa" changes the contents of register AL to a valid unpacked decimal
number, and zeroes the top four bits. "aaa" must always follow the addition
of two unpacked decimal operands in AL. The carry flag is set and AH is
incremented if a carry is necessary. This instruction has no operands.
  "aas" changes the contents of register AL to a valid unpacked decimal
number, and zeroes the top four bits. "aas" must always follow the
subtraction of one unpacked decimal operand from another in AL. The carry flag
is set and AH decremented if a borrow is necessary. This instruction has no
operands.
  "aam" corrects the result of a multiplication of two valid unpacked decimal
numbers. "aam" must always follow the multiplication of two decimal numbers
to produce a valid decimal result. The high order digit is left in AH, the
low order digit in AL. The generalized version of this instruction allows
adjustment of the contents of the AX to create two unpacked digits of any
number base. The standard version of this instruction has no operands, the
generalized version has one operand - an immediate value specifying the
number base for the created digits.
  "aad" modifies the numerator in AH and AL to prepare for the division of two
valid unpacked decimal operands so that the quotient produced by the division
will be a valid unpacked decimal number. AH should contain the high order
digit and AL the low order digit. This instruction adjusts the value and
places the result in AL, while AH will contain zero. The generalized version
of this instruction allows adjustment of two unpacked digits of any number
base. Rules for the operand are the same as for the "aam" instruction.


2.1.5  Logical instructions

"not" inverts the bits in the specified operand to form a one's
complement of the operand. It has no effect on the flags. Rules for the
operand are the same as for the "inc" instruction.
  "and", "or" and "xor" instructions perform the standard
logical operations. They update the SF, ZF and PF flags. Rules for the
operands are the same as for the "add" instruction.
  "bt", "bts", "btr" and "btc" instructions operate on a single bit which can
be in memory or in a general register. The location of the bit is specified
as an offset from the low order end of the operand. The value of the offset
is the taken from the second operand, it either may be an immediate byte or
a general register. These instructions first assign the value of the selected
bit to CF. "bt" instruction does nothing more, "bts" sets the selected bit to
1, "btr" resets the selected bit to 0, "btc" changes the bit to its
complement. The first operand can be word or double word.

    bt  ax,15        ; test bit in register
    bts word [bx],15 ; test and set bit in memory
    btr ax,cx        ; test and reset bit in register
    btc word [bx],cx ; test and complement bit in memory

  "bsf" and "bsr" instructions scan a word or double word for first set bit
and store the index of this bit into destination operand, which must be
general register. The bit string being scanned is specified by source operand,
it may be either general register or memory. The ZF flag is set if the entire
string is zero (no set bits are found); otherwise it is cleared. If no set bit
is found, the value of the destination register is undefined. "bsf" scans from
low order to high order (starting from bit index zero). "bsr" scans from high
order to low order (starting from bit index 15 of a word or index 31 of a
double word).

    bsf ax,bx        ; scan register forward
    bsr ax,[si]      ; scan memory reverse

  "shl" shifts the destination operand left by the number of bits specified
in the second operand. The destination operand can be byte, word, or double
word general register or memory. The second operand can be an immediate value
or the CL register. The processor shifts zeros in from the right (low order)
side of the operand as bits exit from the left side. The last bit that exited
is stored in CF. "sal" is a synonym for "shl".

    shl al,1         ; shift register left by one bit
    shl byte [bx],1  ; shift memory left by one bit
    shl ax,cl        ; shift register left by count from cl
    shl word [bx],cl ; shift memory left by count from cl

  "shr" and "sar" shift the destination operand right by the number of bits
specified in the second operand. Rules for operands are the same as for the
"shl" instruction. "shr" shifts zeros in from the left side of the operand as
bits exit from the right side. The last bit that exited is stored in CF.
"sar" preserves the sign of the operand by shifting in zeros on the left side
if the value is positive or by shifting in ones if the value is negative.
  "shld" shifts bits of the destination operand to the left by the number
of bits specified in third operand, while shifting high order bits from the
source operand into the destination operand on the right. The source operand
remains unmodified. The destination operand can be a word or double word
general register or memory, the source operand must be a general register,
third operand can be an immediate value or the CL register.

    shld ax,bx,1     ; shift register left by one bit
    shld [di],bx,1   ; shift memory left by one bit
    shld ax,bx,cl    ; shift register left by count from cl
    shld [di],bx,cl  ; shift memory left by count from cl

  "shrd" shifts bits of the destination operand to the right, while shifting
low order bits from the source operand into the destination operand on the
left. The source operand remains unmodified. Rules for operands are the same
as for the "shld" instruction.
  "rol" and "rcl" rotate the byte, word or double word destination operand
left by the number of bits specified in the second operand. For each rotation
specified, the high order bit that exits from the left of the operand returns
at the right to become the new low order bit. "rcl" additionally puts in CF
each high order bit that exits from the left side of the operand before it
returns to the operand as the low order bit on the next rotation cycle. Rules
for operands are the same as for the "shl" instruction.
  "ror" and "rcr" rotate the byte, word or double word destination operand
right by the number of bits specified in the second operand. For each rotation
specified, the low order bit that exits from the right of the operand returns
at the left to become the new high order bit. "rcr" additionally puts in CF
each low order bit that exits from the right side of the operand before it
returns to the operand as the high order bit on the next rotation cycle.
Rules for operands are the same as for the "shl" instruction.
  "test" performs the same action as the "and" instruction, but it does not
alter the destination operand, only updates flags. Rules for the operands are
the same as for the "and" instruction.
  "bswap" reverses the byte order of a 32-bit general register: bits 0 through
7 are swapped with bits 24 through 31, and bits 8 through 15 are swapped with
bits 16 through 23. This instruction is provided for converting little-endian
values to big-endian format and vice versa.

    bswap edx        ; swap bytes in register


2.1.6  Control transfer instructions

"jmp" unconditionally transfers control to the target location. The
destination address can be specified directly within the instruction or
indirectly through a register or memory, the acceptable size of this address
depends on whether the jump is near or far (it can be specified by preceding
the operand with "near" or "far" operator) and whether the instruction is
16-bit or 32-bit. Operand for near jump should be "word" size for 16-bit
instruction or the "dword" size for 32-bit instruction. Operand for far jump
should be "dword" size for 16-bit instruction or "pword" size for 32-bit
instruction. A direct "jmp" instruction includes the destination address as
part of the instruction, the operand specifying address should be the
numerical expression for near jump, or two numerical expressions separated
with colon for far jump, the first specifies selector of segment, the second
is the offset within segment. An indirect "jmp" instruction obtains the
destination address indirectly through a register or a pointer variable, the
operand should be general register or memory. See also 1.2.5 for more details.

    jmp 100h         ; direct near jump
    jmp 0FFFFh:0     ; direct far jump
    jmp ax           ; indirect near jump
    jmp pword [ebx]  ; indirect far jump

  "call" transfers control to the procedure, saving on the stack the address
of the instruction following the "call" for later use by a "ret" (return)
instruction. Rules for the operands are the same as for the "jmp" instruction,
but the "call" has no short variant of direct instruction and thus it not
optimized.
  "ret", "retn" and "retf" instructions terminate the execution of a procedure
and transfers control back to the program that originally invoked the
procedure using the address that was stored on the stack by the "call"
instruction. "ret" is the equivalent for "retn", which returns from the
procedure that was executed using the near call, while "retf" returns from
the procedure that was executed using the far call. These instructions default
to the size of address appropriate for the current code setting, but the size
of address can be forced to 16-bit by using the "retw", "retnw" and "retfw"
mnemonics, and to 32-bit by using the "retd", "retnd" and "retfd" mnemonics.
All these instructions may optionally specify an immediate operand, by adding
this constant to the stack pointer, they effectively remove any arguments that
the calling program pushed on the stack before the execution of the "call"
instruction.
  "iret" returns control to an interrupted procedure. It differs from "ret" in
that it also pops the flags from the stack into the flags register. The flags
are stored on the stack by the interrupt mechanism. It defaults to the size of
return address appropriate for the current code setting, but it can be forced
to use 16-bit or 32-bit address by using the "iretw" or "iretd" mnemonic.
  The conditional transfer instructions are jumps that may or may not transfer
control, depending on the state of the CPU flags when the instruction
executes. The mnemonics for conditional jumps may be obtained by attaching
the condition mnemonic (see table 2.1) to the "j" mnemonic,
for example "jc" instruction will transfer the control when the CF flag is
set. The conditional jumps can be near and direct only, and can be optimized
(see 1.2.5), the operand should be an immediate value specifying target
address.

   Table 2.1  Conditions
  Ŀ
   Mnemonic  Condition tested       Description            
  ͵
   o         OF = 1                 overflow               
  Ĵ
   no        OF = 0                 not overflow           
  Ĵ
   c                                carry                  
   b         CF = 1                 below                  
   nae                              not above nor equal    
  Ĵ
   nc                               not carry              
   ae        CF = 0                 above or equal         
   nb                               not below              
  Ĵ
   e         ZF = 1                 equal                  
   z                                zero                   
  Ĵ
   ne        ZF = 0                 not equal              
   nz                               not zero               
  Ĵ
   be        CF or ZF = 1           jump if below or equal 
   na                               not above              
  Ĵ
   a         CF or ZF = 0           above                  
   nbe                              not below nor equal    
  Ĵ
   s         SF = 1                 sign                   
  Ĵ
   ns        SF = 0                 not sign               
  Ĵ
   p         PF = 1                 parity                 
   pe                               parity even            
  Ĵ
   np        PF = 0                 not parity             
   po                               parity odd             
  Ĵ
   l         SF xor OF = 1          less                   
   nge                              not greater nor equal  
  Ĵ
   ge        SF xor OF = 0          greater or equal       
   nl                               not less               
  Ĵ
   le        (SF xor OF) or ZF = 1  less or equal          
   ng                               not greater            
  Ĵ
   g         (SF xor OF) or ZF = 0  greater                
   nle                              not less nor equal     
  

  The "loop" instructions are conditional jumps that use a value placed in
CX (or ECX) to specify the number of repetitions of a software loop. All
"loop" instructions automatically decrement CX (or ECX) and terminate the
loop (don't transfer the control) when CX (or ECX) is zero. It uses CX or ECX
whether the current code setting is 16-bit or 32-bit, but it can be forced to
us CX with the "loopw" mnemonic or to use ECX with the "loopd" mnemonic.
"loope" and "loopz" are the synonyms for the same instruction, which acts as
the standard "loop", but also terminates the loop when ZF flag is set.
"loopew" and "loopzw" mnemonics force them to use CX register while "looped"
and "loopzd" force them to use ECX register. "loopne" and "loopnz" are the
synonyms for the same instructions, which acts as the standard "loop", but
also terminate the loop when ZF flag is not set. "loopnew" and "loopnzw"
mnemonics force them to use CX register while "loopned" and "loopnzd" force
them to use ECX register. Every "loop" instruction needs an operand being an
immediate value specifying target address, it can be only short jump (in the
range of 128 bytes back and 127 bytes forward from the address of instruction
following the "loop" instruction).
  "jcxz" branches to the label specified in the instruction if it finds a
value of zero in CX, "jecxz" does the same, but checks the value of ECX
instead of CX. Rules for the operands are the same as for the "loop"
instruction.
  "int" activates the interrupt service routine that corresponds to the
number specified as an operand to the instruction, the number should be in
range from 0 to 255. The interrupt service routine terminates with an "iret"
instruction that returns control to the instruction that follows "int".
"int3" mnemonic codes the short (one byte) trap that invokes the interrupt 3.
"into" instruction invokes the interrupt 4 if the OF flag is set.
  "bound" verifies that the signed value contained in the specified register
lies within specified limits. An interrupt 5 occurs if the value contained in
the register is less than the lower bound or greater than the upper bound. It
needs two operands, the first operand specifies the register being tested,
the second operand should be memory address for the two signed limit values.
The operands can be "word" or "dword" in size.

    bound ax,[bx]    ; check word for bounds
    bound eax,[esi]  ; check double word for bounds


2.1.7  I/O instructions

  "in" transfers a byte, word, or double word from an input port to AL, AX,
or EAX. I/O ports can be addressed either directly, with the immediate byte
value coded in instruction, or indirectly via the DX register. The destination
operand should be AL, AX, or EAX register. The source operand should be an
immediate value in range from 0 to 255, or DX register.

    in al,20h        ; input byte from port 20h
    in ax,dx         ; input word from port addressed by dx

  "out" transfers a byte, word, or double word to an output port from AL, AX,
or EAX. The program can specify the number of the port using the same methods
as the "in" instruction. The destination operand should be an immediate value
in range from 0 to 255, or DX register. The source operand should be AL, AX,
or EAX register.

    out 20h,ax       ; output word to port 20h
    out dx,al        ; output byte to port addressed by dx


2.1.8  Strings operations

The string operations operate on one element of a string. A string element
may be a byte, a word, or a double word. The string elements are addressed by
SI and DI (or ESI and EDI) registers. After every string operation SI and/or
DI (or ESI and/or EDI) are automatically updated to point to the next element
of the string. If DF (direction flag) is zero, the index registers are
incremented, if DF is one, they are decremented. The amount of the increment
or decrement is 1, 2, or 4 depending on the size of the string element. Every
string operation instruction has short forms which have no operands and use
SI and/or DI when the code type is 16-bit, and ESI and/or EDI when the code
type is 32-bit. SI and ESI by default address data in the segment selected
by DS, DI and EDI always address data in the segment selected by ES. Short
form is obtained by attaching to the mnemonic of string operation letter
specifying the size of string element, it should be "b" for byte element,
"w" for word element, and "d" for double word element. Full form of string
operation needs operands providing the size operator and the memory addresses,
which can be SI or ESI with any segment prefix, DI or EDI always with ES
segment prefix.
  "movs" transfers the string element pointed to by SI (or ESI) to the
location pointed to by DI (or EDI). Size of operands can be byte, word, or
double word. The destination operand should be memory addressed by DI or EDI,
the source operand should be memory addressed by SI or ESI with any segment
prefix.

    movs byte [di],[si]        ; transfer byte
    movs word [es:di],[ss:si]  ; transfer word
    movsd                      ; transfer double word

  "cmps" subtracts the destination string element from the source string
element and updates the flags AF, SF, PF, CF and OF, but it does not change
any of the compared elements. If the string elements are equal, ZF is set,
otherwise it is cleared. The first operand for this instruction should be the
source string element addressed by SI or ESI with any segment prefix, the
second operand should be the destination string element addressed by DI or
EDI.

    cmpsb                      ; compare bytes
    cmps word [ds:si],[es:di]  ; compare words
    cmps dword [fs:esi],[edi]  ; compare double words

  "scas" subtracts the destination string element from AL, AX, or EAX
(depending on the size of string element) and updates the flags AF, SF, ZF,
PF, CF and OF. If the values are equal, ZF is set, otherwise it is cleared.
The operand should be the destination string element addressed by DI or EDI.

    scas byte [es:di]          ; scan byte
    scasw                      ; scan word
    scas dword [es:edi]        ; scan double word

  "stos" places the value of AL, AX, or EAX into the destination string
element. Rules for the operand are the same as for the "scas" instruction.
  "lods" places the source string element into AL, AX, or EAX. The operand
should be the source string element addressed by SI or ESI with any segment
prefix.

    lods byte [ds:si]          ; load byte
    lods word [cs:si]          ; load word
    lodsd                      ; load double word

  "ins" transfers a byte, word, or double word from an input port addressed
by DX register to the destination string element. The destination operand
should be memory addressed by DI or EDI, the source operand should be the DX
register.

    insb                       ; input byte
    ins word [es:di],dx        ; input word
    ins dword [edi],dx         ; input double word

  "outs" transfers the source string element to an output port addressed by
DX register. The destination operand should be the DX register and the source
operand should be memory addressed by SI or ESI with any segment prefix.

    outs dx,byte [si]           ; output byte
    outsw                       ; output word
    outs dx,dword [gs:esi]     ; output double word

  The repeat prefixes "rep", "repe"/"repz", and "repne"/"repnz" specify
repeated string operation. When a string operation instruction has a repeat
prefix, the operation is executed repeatedly, each time using a different
element of the string. The repetition terminates when one of the conditions
specified by the prefix is satisfied. All three prefixes automatically
decrease CX or ECX register (depending whether string operation instruction
uses the 16-bit or 32-bit addressing) after each operation and repeat the
associated operation until CX or ECX is zero. "repe"/"repz" and
"repne"/"repnz" are used exclusively with the "scas" and "cmps" instructions
(described below). When these prefixes are used, repetition of the next
instruction depends on the zero flag (ZF) also, "repe" and "repz" terminate
the execution when the ZF is zero, "repne" and "repnz" terminate the execution
when the ZF is set.

    rep  movsd       ; transfer multiple double words
    repe cmpsb       ; compare bytes until not equal


2.1.9  Flag control instructions

The flag control instructions provide a method for directly changing the
state of bits in the flag register. All instructions described in this
section have no operands.
  "stc" sets the CF (carry flag) to 1, "clc" zeroes the CF, "cmc" changes the
CF to its complement. "std" sets the DF (direction flag) to 1, "cld" zeroes
the DF, "sti" sets the IF (interrupt flag) to 1 and therefore enables the
interrupts, "cli" zeroes the IF and therefore disables the interrupts.
  "lahf" copies SF, ZF, AF, PF, and CF to bits 7, 6, 4, 2, and 0 of the
AH register. The contents of the remaining bits are undefined. The flags
remain unaffected.
  "sahf" transfers bits 7, 6, 4, 2, and 0 from the AH register into SF, ZF,
AF, PF, and CF.
  "pushf" decrements "esp" by two or four and stores the low word or
double word of flags register at the top of stack, size of stored data
depends on the current code setting. "pushfw" variant forces storing the
word and "pushfd" forces storing the double word.
  "popf" transfers specific bits from the word or double word at the top
of stack, then increments "esp" by two or four, this value depends on
the current code setting. "popfw" variant forces restoring from the word
and "popfd" forces restoring from the double word.


2.1.10  Conditional operations

  The instructions obtained by attaching the condition mnemonic (see table
2.1) to the "set" mnemonic set a byte to one if the condition is true and set
the byte to zero otherwise. The operand should be an 8-bit be general register
or the byte in memory.

    setne al         ; set al if zero flag cleared
    seto byte [bx]   ; set byte if overflow

  "salc" instruction sets the all bits of AL register when the carry flag is
set and zeroes the AL register otherwise. This instruction has no arguments.
  The instructions obtained by attaching the condition mnemonic to the "cmov"
mnemonic transfer the word or double word from the general register or memory
to the general register only when the condition is true. The destination
operand should be general register, the source operand can be general register
or memory.

    cmove ax,bx      ; move when zero flag set
    cmovnc eax,[ebx] ; move when carry flag cleared

  "cmpxchg" compares the value in the AL, AX, or EAX register with the
destination operand. If the two values are equal, the source operand is
loaded into the destination operand. Otherwise, the destination operand is
loaded into the AL, AX, or EAX register. The destination operand may be a
general register or memory, the source operand must be a general register.

    cmpxchg dl,bl    ; compare and exchange with register
    cmpxchg [bx],dx  ; compare and exchange with memory

  "cmpxchg8b" compares the 64-bit value in EDX and EAX registers with the
destination operand. If the values are equal, the 64-bit value in ECX and EBX
registers is stored in the destination operand. Otherwise, the value in the
destination operand is loaded into EDX and EAX registers. The destination
operand should be a quad word in memory.

    cmpxchg8b [bx]   ; compare and exchange 8 bytes


2.1.11  Miscellaneous instructions

"nop" instruction occupies one byte but affects nothing but the instruction
pointer. This instruction has no operands and doesn't perform any operation.
  "ud2" instruction generates an invalid opcode exception. This instruction
is provided for software testing to explicitly generate an invalid opcode.
This is instruction has no operands.
  "xlat" replaces a byte in the AL register with a byte indexed by its value
in a translation table addressed by BX or EBX. The operand should be a byte
memory addressed by BX or EBX with any segment prefix. This instruction has
also a short form "xlatb" which has no operands and uses the BX or EBX address
in the segment selected by DS depending on the current code setting.
  "lds" transfers a pointer variable from the source operand to DS and the
destination register. The source operand must be a memory operand, and the
destination operand must be a general register. The DS register receives the
segment selector of the pointer while the destination register receives the
offset part of the pointer. "les", "lfs", "lgs" and "lss" operate identically
to "lds" except that rather than DS register the ES, FS, GS and SS is used
respectively.

    lds bx,[si]      ; load pointer to ds:bx

  "lea" transfers the offset of the source operand (rather than its value)
to the destination operand. The source operand must be a memory operand, and
the destination operand must be a general register.

    lea dx,[bx+si+1] ; load effective address to dx

  "cpuid" returns processor identification and feature information in the
EAX, EBX, ECX, and EDX registers. The information returned is selected by
entering a value in the EAX register before the instruction is executed.
This instruction has no operands.
  "pause" instruction delays the execution of the next instruction an
implementation specific amount of time. It can be used to improve the
performance of spin wait loops. This instruction has no operands.
  "enter" creates a stack frame that may be used to implement the scope rules
of block-structured high-level languages. A "leave" instruction at the end of
a procedure complements an "enter" at the beginning of the procedure to
simplify stack management and to control access to variables for nested
procedures. The "enter" instruction includes two parameters. The first
parameter specifies the number of bytes of dynamic storage to be allocated on
the stack for the routine being entered. The second parameter corresponds to
the lexical nesting level of the routine, it can be in range from 0 to 31.
The specified lexical level determines how many sets of stack frame pointers
the CPU copies into the new stack frame from the preceding frame. This list
of stack frame pointers is sometimes called the display. The first word (or
double word when code is 32-bit) of the display is a pointer to the last stack
frame. This pointer enables a "leave" instruction to reverse the action of the
previous "enter" instruction by effectively discarding the last stack frame.
After "enter" creates the new display for a procedure, it allocates the
dynamic storage space for that procedure by decrementing ESP by the number of
bytes specified in the first parameter. To enable a procedure to address its
display, "enter" leaves BP (or EBP) pointing to the beginning of the new stack
frame. If the lexical level is zero, "enter" pushes BP (or EBP), copies SP to
BP (or ESP to EBP) and then subtracts the first operand from ESP. For nesting
levels greater than zero, the processor pushes additional frame pointers on
the stack before adjusting the stack pointer.

    enter 2048,0     ; enter and allocate 2048 bytes on stack


2.1.12  System instructions

"lmsw" loads the operand into the machine status word (bits 0 through 15 of
CR0 register), while "smsw" stores the machine status word into the
destination operand. The operand can be a 16-bit or 32-bit general register
or the word in memory.

    lmsw ax          ; load machine status from register
    smsw [bx]        ; store machine status to memory

  "lgdt" and "lidt" instructions load the values in operand into the global
descriptor table register or the interrupt descriptor table register
respectively. "sgdt" and "sidt" store the contents of the global descriptor
table register or the interrupt descriptor table register in the destination
operand. The operand should be a 6 bytes in memory.

    lgdt [ebx]       ; load global descriptor table

  "lldt" loads the operand into the segment selector field of the local
descriptor table register and "sldt" stores the segment selector from the
local descriptor table register in the operand. "ltr" loads the operand into
the segment selector field of the task register and "str" stores the segment
selector from the task register in the operand. Rules for operand are the same
as for the "lmsw" instruction.
  "lar" loads the access rights from the segment descriptor specified by
the selector in source operand into the destination operand and sets the ZF
flag. The operands can be both words or double words. The source operand may
be a general register or memory. The destination operand should be a general
register.

    lar ax,[bx]      ; load access rights into word
    lar eax,edx      ; load access rights into double word

  "lsl" loads the segment limit from the segment descriptor specified by the
selector in source operand into the destination operand and sets the ZF flag.
Rules for operand are the same as for the "lar" instruction.
  "verr" and "verw" verify whether the code or data segment specified with
the operand is readable or writable from the current privilege level. The
operand should be a word, it can be general register or memory. If the segment
is accessible and readable (for "verr") or writable (for "verw") the ZF flag
is set, otherwise it's cleared. Rules for operand are the same as for the
"lldt" instruction.
  "arpl" compares the RPL (requestor's privilege level) fields of two segment
selectors. The first operand contains one segment selector and the second
operand contains the other. If the RPL field of the destination operand is
less than the RPL field of the source operand, the ZF flag is set and the RPL
field of the destination operand is increased to match that of the source
operand. Otherwise, the ZF flag is cleared and no change is made to the
destination operand. The destination operand can be a word general register
or memory, the source operand must be a general register.

    arpl bx,ax       ; adjust RPL of selector in register
    arpl [bx],ax     ; adjust RPL of selector in memory

  "clts" clears the TS (task switched) flag in the CR0 register. This
instruction has no operands.
  "lock" prefix causes the processor's bus-lock signal to be asserted during
execution of the accompanying instruction. In a multiprocessor environment,
the bus-lock signal insures that the processor has exclusive use of any shared
memory while the signal is asserted. The "lock" prefix can be prepended only
to the following instructions and only to those forms of the instructions
where the destination operand is a memory operand: "add", "adc", "and", "btc",
"btr", "bts", "cmpxchg", "cmpxchg8b", "dec", "inc", "neg", "not", "or", "sbb",
"sub", "xor", "xadd" and "xchg". If the "lock" prefix is used with one of
these instructions and the source operand is a memory operand, an undefined
opcode exception may be generated. An undefined opcode exception will also be
generated if the "lock" prefix is used with any instruction not in the above
list. The "xchg" instruction always asserts the bus-lock signal regardless of
the presence or absence of the "lock" prefix.
  "hlt" stops instruction execution and places the processor in a halted
state. An enabled interrupt, a debug exception, the BINIT, INIT or the RESET
signal will resume execution. This instruction has no operands.
  "invlpg" invalidates (flushes) the TLB (translation lookaside buffer) entry
specified with the operand, which should be a memory. The processor determines
the page that contains that address and flushes the TLB entry for that page.
  "rdmsr" loads the contents of a 64-bit MSR (model specific register) of the
address specified in the ECX register into registers EDX and EAX. "wrmsr"
writes the contents of registers EDX and EAX into the 64-bit MSR of the
address specified in the ECX register. "rdtsc" loads the current value of the
processor's time stamp counter from the 64-bit MSR into the EDX and EAX
registers. The processor increments the time stamp counter MSR every clock
cycle and resets it to 0 whenever the processor is reset. "rdpmc" loads the
contents of the 40-bit performance monitoring counter specified in the ECX
register into registers EDX and EAX. These instructions have no operands.
  "wbinvd" writes back all modified cache lines in the processor's internal
cache to main memory and invalidates (flushes) the internal caches. The
instruction then issues a special function bus cycle that directs external
caches to also write back modified data and another bus cycle to indicate that
the external caches should be invalidated. This instruction has no operands.
  "rsm" return program control from the system management mode to the program
that was interrupted when the processor received an SMM interrupt. This
instruction has no operands.
  "sysenter" executes a fast call to a level 0 system procedure, "sysexit"
executes a fast return to level 3 user code. The addresses used by these
instructions are stored in MSRs. These instructions have no operands.


2.1.13  FPU instructions

The FPU (Floating-Point Unit) instructions operate on the floating-point
values in three formats: single precision (32-bit), double precision (64-bit)
and double extended precision (80-bit). The FPU registers form the stack and
each of them holds the double extended precision floating-point value. When
some values are pushed onto the stack or are removed from the top, the FPU
registers are shifted, so ST0 is always the value on the top of FPU stack, ST1
is the first value below the top, etc. The ST0 name has also the synonym ST.
  "fld" pushes the floating-point value onto the FPU register stack. The
operand can be 32-bit, 64-bit or 80-bit memory location or the FPU register,
it's value is then loaded onto the top of FPU register stack (the ST0
register) and is automatically converted into the double extended precision
format.

    fld dword [bx]   ; load single prevision value from memory
    fld st2          ; push value of st2 onto register stack

  "fld1", "fldz", "fldl2t", "fldl2e", "fldpi", "fldlg2" and "fldln2" load the
commonly used contants onto the FPU register stack. The loaded constants are
+1.0, +0.0, lb 10, lb e, pi, lg 2 and ln 2 respectively. These instructions
have no operands.
  "fild" convert the singed integer source operand into double extended
precision floating-point format and pushes the result onto the FPU register
stack. The source operand can be a 16-bit, 32-bit or 64-bit memory location.

    fild qword [bx]  ; load 64-bit integer from memory

  "fst" copies the value of ST0 register to the destination operand, which
can be 32-bit or 64-bit memory location or another FPU register. "fstp"
performs the same operation as "fst" and then pops the register stack,
getting rid of ST0. "fstp" accepts the same operands as the "fst" instruction
and can also store value in the 80-bit memory.

    fst st3          ; copy value of st0 into st3 register
    fstp tword [bx]  ; store value in memory and pop stack

  "fist" converts the value in ST0 to a signed integer and stores the result
in the destination operand. The operand can be 16-bit or 32-bit memory
location. "fistp" performs the same operation and then pops the register
stack, it accepts the same operands as the "fist" instruction and can also
store integer value in the 64-bit memory, so it has the same rules for
operands as "fild" instruction.
  "fbld" converts the packed BCD integer into double extended precision
floating-point format and pushes this value onto the FPU stack. "fbstp"
converts the value in ST0 to an 18-digit packed BCD integer, stores the result
in the destination operand, and pops the register stack. The operand should be
an 80-bit memory location.
  "fadd" adds the destination and source operand and stores the sum in the
destination location. The destination operand is always an FPU register, if
the source is a memory location, the destination is ST0 register and only
source operand should be specified. If both operands are FPU registers, at
least one of them should be ST0 register. An operand in memory can be a
32-bit or 64-bit value.

    fadd qword [bx]  ; add double precision value to st0
    fadd st2,st0     ; add st0 to st2

  "faddp" adds the destination and source operand, stores the sum in the
destination location and then pops the register stack. The destination operand
must be an FPU register and the source operand must be the ST0. When no
operands are specified, ST1 is used as a destination operand.

    faddp            ; add st0 to st1 and pop the stack
    faddp st2,st0    ; add st0 to st2 and pop the stack

"fiadd" instruction converts an integer source operand into double extended
precision floating-point value and adds it to the destination operand. The
operand should be a 16-bit or 32-bit memory location.

    fiadd word [bx]  ; add word integer to st0

  "fsub", "fsubr", "fmul", "fdiv", "fdivr" instruction are similar to "fadd",
have the same rules for operands and differ only in the perfomed computation.
"fsub" substracts the source operand from the destination operand, "fsubr"
substract the destination operand from the source operand, "fmul" multiplies
the destination and source operands, "fdiv" divides the destination operand by
the source operand and "fdivr" divides the source operand by the destination
operand. "fsubp", "fsubrp", "fmulp", "fdivp", "fdivrp" perform the same
operations and pop the register stack, the rules for operand are the same as
for the "faddp" instruction. "fisub", "fisubr", "fimul", "fidiv", "fidivr"
perform these operations after converting the integer source operand into
floating-point value, they have the same rules for operands as "fiadd"
instruction.
  "fsqrt" computes the square root of the value in ST0 register, "fsin"
computes the sine of that value, "fcos" computes the cosine of that value,
"fchs" complements its sign bit, "fabs" clears its sign to create the absolute
value, "frndint" rounds it to the nearest integral value, depending on the
current rounding mode. "f2xm1" computes the exponential value of 2 to the
power of ST0 and substracts the 1.0 from it, the value of ST0 must lie in the
range -1.0 to +1.0. All these instruction store the result in ST0 and have no
operands.
  "fsincos" computes both the sine and the cosine of the value in ST0
register, stores the sine in ST0 and pushes the cosine on the top of FPU
register stack. "fptan" computes the tangent of the value in ST0, stores the
result in ST0 and pushes a 1.0 onto the FPU register stack. "fpatan" computes
the arctangent of the value in ST1 divided by the value in ST0, stores the
result in ST1 and pops the FPU register stack. "fyl2x" computes the binary
logarithm of ST0, multiplies it by ST1, stores the result in ST1 and pop the
FPU register stack; "fyl2xp1" performs the same operation but it adds 1.0 to
ST0 before computing the logarithm. "fprem" computes the remainder obtained
from dividing the value in ST0 by the value in ST1, and stores the result
in ST0. "fprem1" performs the same operation as "fprem", but it computes the
remainder in the way specified by IEEE Standard 754. "fscale" truncates the
value in ST1 and increases the exponent of ST0 by this value. "fxtract"
separates the value in ST0 into its exponent and significand, stores the
exponent in ST0 and pushes the significand onto the register stack. "fnop"
performs no operation. These instruction have no operands.
  "fxch" exchanges the contents of ST0 an another FPU register. The operand
should be an FPU register, if no operand is specified, the contents of ST0 and
ST1 are exchanged.
  "fcom" and "fcomp" compare the contents of ST0 and the source operand and
set flags in the FPU status word according to the results. "fcomp"
additionally pops the register stack after performing the comparision. The
operand can be a single or double precision value in memory or the FPU
register. When no operand is specified, ST1 is used as a source operand.

    fcom             ; compare st0 with st1
    fcomp st2        ; compare st0 with st2 and pop stack

  "fcompp" compares the contents of ST0 and ST1, sets flags in the FPU status
word according to the results and pops the register stack twice. This
instruction has no operands.
  "fucom", "fucomp" and "fucompp" performs an unordered comparision of two FPU
registers. Rules for operands are the same as for the "fcom", "fcomp" and
"fcompp", but the source operand must be an FPU register.
  "ficom" and "ficomp" compare the value in ST0 with an integer source operand
and set the flags in the FPU status word according to the results. "ficomp"
additionally pops the register stack after performing the comparision. The
integer value is converted to double extended precision floating-point format
before the comparision is made. The operand should be a 16-bit or 32-bit
memory location.

    ficom word [bx]  ; compare st0 with 16-bit integer

  "fcomi", "fcomip", "fucomi", "fucomip" perform the comparision of ST0 with
another FPU register and set the ZF, PF and CF flags according to the results.
"fcomip" and "fucomip" additionaly pop the register stack after performing the
comparision. The instructions obtained by attaching the FPU condition mnemonic
(see table 2.2) to the "fcmov" mnemonic transfer the specified FPU register
into ST0 register if the fiven test condition is true. These instruction
allow two different syntaxes, one with single operand specifying the source
FPU register, and one with two operands, in that case destination operand
should be ST0 register and the second operand specifies the source FPU
register.

    fcomi st2        ; compare st0 with st2 and set flags
    fcmovb st0,st2   ; transfer st2 to st0 if below

   Table 2.2  FPU conditions
  Ŀ
   Mnemonic  Condition tested  Description            
  ͵
   b         CF = 1            below                  
   e         ZF = 1            equal                  
   be        CF or ZF = 1      below or equal         
   u         PF = 1            unordered              
   nb        CF = 0            not below              
   ne        ZF = 0            not equal              
   nbe       CF and ZF = 0     not below nor equal    
   nu        PF = 0            not unordered          
  

  "ftst" compares the value in ST0 with 0.0 and sets the flags in the FPU
status word according to the results. "fxam" examines the contents of the ST0
and sets the flags in FPU status word to indicate the class of value in the
register. These instructions have no operands.
  "fstsw" and "fnstsw" store the current value of the FPU status word in the
destination location. The destination operand can be either a 16-bit memory or
the AX register. "fstsw" checks for pending umasked FPU exceptions before
storing the status word, "fnstsw" does not.
  "fstcw" and "fnstcw" store the current value of the FPU control word at the
specified destination in memory. "fstcw" checks for pending umasked FPU
exceptions before storing the control word, "fnstcw" does not. "fldcw" loads
the operand into the FPU control word. The operand should be a 16-bit memory
location.
  "fstenv" and "fnstenv" store the current FPU operating environment at the
memory location specified with the destination operand, and then mask all FPU
exceptions. "fstenv" checks for pending umasked FPU exceptions before
proceeding, "fnstenv" does not. "fldenv" loads the complete operating
environment from memory into the FPU. "fsave" and "fnsave" store the current
FPU state (operating environment and register stack) at the specified
destination in memory and reinitializes the FPU. "fsave" check for pending
unmasked FPU exceptions before proceeding, "fnsave" does not. "frstor"
loads the FPU state from the specified memory location. All these instructions
need an operand being a memory location.
  "finit" and "fninit" set the FPU operating environment into its default
state. "finit" checks for pending unmasked FPU exception before proceeding,
"fninit" does not. "fclex" and "fnclex" clear the FPU exception flags in the
FPU status word. "fclex" checks for pending unmasked FPU exception before
proceeding, "fnclex" does not. "wait" and "fwait" are synonyms for the same
instruction, which causes the processor to check for pending unmasked FPU
exceptions and handle them before proceeding. These instruction have no
operands.
  "ffree" sets the tag associated with specified FPU register to empty. The
operand should be an FPU register.
  "fincstp" and "fdecstp" rotate the FPU stack by one by adding or
substracting one to the pointer of the top of stack. These instruction have no
operands.


2.1.14  MMX instructions

The MMX instructions operate on the packed integer types and use the MMX
registers, which are the low 64-bit parts of the 80-bit FPU registers. Because
of this MMX instructions cannot be used at the same time as FPU instructions.
They can operate on packed bytes (eight 8-bit integers), packed words (four
16-bit integers) or packed double words (two 32-bit integers), use of packed
formats allows to perform operations on multiple data at one time.
  "movq" copies a quad word from the source operand to the destination
operand. At least one of the operands must be a MMX register, the second one
can be also a MMX register or 64-bit memory location.

    movq mm0,mm1     ; move quad word from register to register
    movq mm2,[ebx]   ; move quad word from memory to register

  "movd" copies a double word from the source operand to the destination
operand. One of the operands must be a MMX register, the second one can be a
general register or 32-bit memory location. Only low double word of MMX
register is used.
  All general MMX operations have two operands, the destination operand should
be a MMX register, the source operand can be a MMX register or 64-bit memory
location. Operation is performed on the corresponding data elements of the
source and destination operand and stored in the data elements of the
destination operand. "paddb", "paddw" and "paddd" perform the addition of
packed bytes, packed words, or packed double words.  "psubb", "psubw" and
"psubd" perform the substraction of appropriate types. "paddsb", "paddsw",
"psubsb" and "psubsw" perform the addition or substraction of packed bytes
or packed words with the signed saturation. "paddusb", "paddusw", "psubusb",
"psubusw" are analoguous, but with unsigned saturation. "pmulhw" and "pmullw"
performs a signed multiply of the packed words and store the high or low words
of the results in the destination operand. "pmaddwd" performs a multiply of
the packed words and adds the four intermediate double word products in pairs
to produce result as a packed double words. "pand", "por" and "pxor" perform
the logical operations on the quad words, "pandn" peforms also a logical
negation of the destination operand before performing the "and" operation.
"pcmpeqb", "pcmpeqw" and "pcmpeqd" compare for equality of packed bytes,
packed words or packed double words. If a pair of data elements is equal, the
corresponding data element in the destination operand is filled with bits of
value 1, otherwise it's set to 0. "pcmpgtb", "pcmpgtw" and "pcmpgtd" perform
the similar operation, but they check whether the data elements in the
destination operand are greater than the correspoding data elements in the
source operand. "packsswb" converts packed signed words into packed signed
bytes, "packssdw" converts packed signed double words into packed signed
words, using saturation to handle overflow conditions. "packuswb" converts
packed signed words into packed unsigned bytes. Converted data elements from
the source operand are stored in the low part of the destination operand,
while converted data elements from the destination operand are stored in the
high part. "punpckhbw", "punpckhwd" and "punpckhdq" interleaves the data
elements from the high parts of the source and destination operands and
stores the result into the destination operand. "punpcklbw", "punpcklwd" and
"punpckldq" perform the same operation, but the low parts of the source and
destination operand are used.

    paddsb mm0,[esi] ; add packed bytes with signed saturation
    pcmpeqw mm3,mm7  ; compare packed words for equality

  "psllw", "pslld" and "psllq" perform logical shift left of the packed words,
packed double words or a single quad word in the destination operand by the
amount specified in the source operand. "psrlw", "psrld" and "psrlq" perform
logical shift right of the packed words, packed double words or a single quad
word. "psraw" and "psrad" perform arithmetic shift of the packed words or
double words. The destination operand should be a MMX register, while source
operand can be a MMX register, 64-bit memory location, or 8-bit immediate
value.

    psllw mm2,mm4    ; shift words left logically
    psrad mm4,[ebx]  ; shift double words right arithmetically

  "emms" makes the FPU registers usable for the FPU instructions, it must be
used before using the FPU instructions if any MMX instructions were used.


2.1.15  SSE instructions

The SSE extension adds more MMX instructions and also introduces the
operations on packed single precision floating point values. The 128-bit
packed single precision format consists of four single precision floating
point values. The 128-bit SSE registers are designed for the purpose of
operations on this data type.
  "movaps" and "movups" transfer a double quad word operand containing packed
single precision values from source operand to destination operand. At least
one of the operands have to be a SSE register, the second one can be also a
SSE register or 128-bit memory location. Memory operands for "movaps"
instruction must be aligned on boundary of 16 bytes, operands for "movups"
instruction don't have to be aligned.

    movups xmm0,[ebx]  ; move unaligned double quad word

  "movlps" moves packed two single precision values between the memory and the
low quad word of SSE register. "movhps" moved packed two single precision
values between the memory and the high quad word of SSE register. One of the
operands must be a SSE register, and the other operand must be a 64-bit memory
location.

    movlps xmm0,[ebx]  ; move memory to low quad word of xmm0
    movhps [esi],xmm7  ; move high quad word of xmm7 to memory

  "movlhps" moves packed two single precision values from the low quad word
of source register to the high quad word of destination register. "movhlps"
moves two packed single precision values from the high quad word of source
register to the low quad word of destination register. Both operands have to
be a SSE registers.
  "movmskps" transfers the most significant bit of each of the four single
precision values in the SSE register into low four bits of a general register.
The source operand must be a SSE register, the destination operand must be a
general register.
  "movss" transfers a single precision value between source and destination
operand (only the low double word is trasferred). At least one of the operands
have to be a SSE register, the second one can be also a SSE register or 32-bit
memory location.

    movss [edi],xmm3   ; move low double word of xmm3 to memory

  Each of the SSE arithmetic operations has two variants. When the mnemonic
ends with "ps", the source operand can be a 128-bit memory location or a SSE
register, the destination operand must be a SSE register and the operation is
performed on packed four single precision values, for each pair of the
corresponding data elements separately, the result is stored in the
destination register. When the mnemonic ends with "ss", the source operand
can be a 32-bit memory location or a SSE register, the destination operand
must be a SSE register and the operation is performed on single precision
values, only low double words of SSE registers are used in this case, the
result is stored in the low double word of destination register. "addps" and
"addss" add the values, "subps" and "subss" substract the source value from
destination value, "mulps" and "mulss" multiply the values, "divps" and
"divss" divide the destination value by the source value, "rcpps" and "rcpss"
compute the approximate reciprocal of the source value, "sqrtps" and "sqrtss"
compute the square root of the source value, "rsqrtps" and "rsqrtss" compute
the approximate reciprocal of square root of the source value, "maxps" and
"maxss" compare the source and destination values and return the greater one,
"minps" and "minss" compare the source and destination values and return the
lesser one.

    mulss xmm0,[ebx]   ; multiply single precision values
    addps xmm3,xmm7    ; add packed single precision values

  "andps", "andnps", "orps" and "xorps" perform the logical operations on
packed single precision values. The source operand can be a 128-bit memory
location or a SSE register, the destination operand must be a SSE register.
  "cmpps" compares packed single precision values and returns a mask result
into the destination operand, which must be a SSE register. The source operand
can be a 128-bit memory location or SSE register, the third operand must be an
immediate operand selecting code of one of the eight compare conditions
(table 2.3). "cmpss" performs the same operation on single precision values,
only low double word of destination register is affected, in this case source
operand can be a 32-bit memory location or SSE register. These two
instructions have also variants with only two operands and the condition
encoded within mnemonic. Their mnemonics are obtained by attaching the
mnemonic from table 2.3 to the "cmp" mnemonic and then attaching the "ps" or
"ss" at the end.

    cmpps xmm2,xmm4,0  ; compare packed single precision values
    cmpltss xmm0,[ebx] ; compare single precision values

   Table 2.3  SSE conditions
  Ŀ
   Code  Mnemonic  Description             
  ͵
   0     eq        equal                   
   1     lt        less than               
   2     le        less than or equal      
   3     unord     unordered               
   4     neq       not equal               
   5     nlt       not less than           
   6     nle       not less than nor equal 
   7     ord       ordered                 
  

  "comiss" and "ucomiss" compare the single precision values and set the ZF,
PF and CF flags to show the result. The destination operand must be a SSE
register, the source operand can be a 32-bit memory location or SSE register.
  "shufps" moves any two of the four single precision values from the
destination operand into the low quad word of the destination operand, and any
two of the four values from the source operand into the high quad word of the
destination operand. The destination operand must be a SSE register, the
source operand can be a 128-bit memory location or SSE register, the third
operand must be an 8-bit immediate value selecting which values will be moved
into the destination operand. Bits 0 and 1 select the value to be moved from
destination operand to the low double word of the result, bits 2 and 3 select
the value to be moved from the destination operand to the second double word,
bits 4 and 5 select the value to be moved from the source operand to the third
double word, and bits 6 and 7 select the value to be moved from the source
operand to the high double word of the result.

    shufps xmm0,xmm0,10010011b ; shuffle double words

  "unpckhps" performs an interleaved unpack of the values from the high parts
of the source and destination operands and stores the result in the
destination operand, which must be a SSE register. The source operand can be
a 128-bit memory location or a SSE register. "unpcklps" performs an
interleaved unpack of the values from the low parts of the source and
destination operand and stores the result in the destination operand,
the rules for operands are the same.
  "cvtpi2ps" converts packed two double word integers into the the packed two
single precision floating point values and stores the result in the low quad
word of the destination operand, which should be a SSE register. The source
operand can be a 64-bit memory location or MMX register.

    cvtpi2ps xmm0,mm0  ; convert integers to single precision values

  "cvtsi2ss" converts a double word integer into a single precision floating
point value and stores the result in the low double word of the destination
operand, which should be a SSE register. The source operand can be a 32-bit
memory location or 32-bit general register.

    cvtsi2ss xmm0,eax  ; convert integer to single precision value

  "cvtps2pi" converts packed two single precision floating point values into
packed two double word integers and stores the result in the destination
operand, which should be a MMX register. The source operand can be a 64-bit
memory location or SSE register, only low quad word of SSE register is used.
"cvttps2pi" performs the similar operation, except that truncation is used to
round a source values to integers, rules for the operands are the same.

    cvtps2pi mm0,xmm0  ; convert single precision values to integers

  "cvtss2si" convert a single precision floating point value into a double
word integer and stores the result in the destination operand, which should be
a 32-bit general register. The source operand can be a 32-bit memory location
or SSE register, only low double word of SSE register is used. "cvttss2si"
performs the similar operation, except that truncation is used to round a
source value to integer, rules for the operands are the same.

    cvtss2si eax,xmm0  ; convert single precision value to integer

  "pextrw" copies the word in the source operand specified by the third
operand to the destination operand. The source operand must be a MMX register,
the destination operand must be a 32-bit general register (but only the low
word of it is affected), the third operand must an 8-bit immediate value.

    pextrw eax,mm0,1   ; extract word into eax

  "pinsrw" inserts a word from the source operand in the destination operand
at the location specified with the third operand, which must be an 8-bit
immediate value. The destination operand must be a MMX register, the source
operand can be a 16-bit memory location or 32-bit general register (only low
word of the register is used).

    pinsrw mm1,ebx,2   ; insert word from ebx

  "pavgb" and "pavgw" compute average of packed bytes or words. "pmaxub"
return the maximum values of packed unsigned bytes, "pminub" returns the
minimum values of packed unsigned bytes, "pmaxsw" returns the maximum values
of packed signed words, "pminsw" returns the minimum values of packed signed
words. "pmulhuw" performs a unsigned multiply of the packed words and stores
the high words of the results in the destination operand. "psadbw" computes
the absolute differences of packed unsigned bytes, sums the differences, and
stores the sum in the low word of destination operand. All these instructions
follow the same rules for operands as the general MMX operations described in
previous section.
  "pmovmskb" creates a mask made of the most significant bit of each byte in
the source operand and stores the result in the low byte of destination
operand. The source operand must be a MMX register, the destination operand
must a 32-bit general register.
  "pshufw" inserts words from the source operand in the destination operand
from the locations specified with the third operand. The destination operand
must be a MMX register, the source operand can be a 64-bit memory location or
MMX register, third operand must an 8-bit immediate value selecting which
values will be moved into destination operand, in the similar way as the third
operand of the "shufps" instruction.
  "movntq" moves the quad word from the source operand to memory using a
non-temporal hint to minimize cache pollution. The source operand should be a
MMX register, the destination operand should be a 64-bit memory location.
"movntps" stores packed single precision values from the SSE register to
memory using a non-temporal hint. The source operand should be a SSE register,
the destination operand should be a 128-bit memory location. "maskmovq" stores
selected bytes from the first operand into a 64-bit memory location using a
non-temporal hint. Both operands should be a MMX registers, the second operand
selects wich bytes from the source operand are written to memory. The
memory location is pointed by DI (or EDI) register in the segment selected
by DS.
  "prefetcht0", "prefetcht1", "prefetcht2" and "prefetchnta" fetch the line
of data from memory that contains byte specified with the operand to a
specified location in hierarchy.  The operand should be an 8-bit memory
location.
  "sfence" performs a serializing operation on all instruction storing to
memory that were issued prior to it. This instruction has no operands.
  "ldmxcsr" loads the 32-bit memory operand into the MXCSR register. "stmxcsr"
stores the contents of MXCSR into a 32-bit memory operand.
  "fxsave" saves the current state of the FPU, MXCSR register, and all the FPU
and SSE registers to a 512-byte memory location specified in the destination
operand. "fxrstor" reloads data previously stored with "fxsave" instruction
from the specified 512-byte memory location. The memory operand for both those
instructions must be aligned on 16 byte boundary, it should declare operand
of no specified size.


2.1.16  SSE2 instructions

The SSE2 extension introduces the operations on packed double precision
floating point values, extends the syntax of MMX instructions, and adds also
some new instructions.
  "movapd" and "movupd" transfer a double quad word operand containing packed
double precision values from source operand to destination operand. These
instructions are analogous to "movaps" and "movups" and have the same rules
for operands.
  "movlpd" moves double precision value between the memory and the low quad
word of SSE register. "movhpd" moved double precision value between the memory
and the high quad word of SSE register. These instructions are analogous to
"movlps" and "movhps" and have the same rules for operands.
  "movmskpd" transfers the most significant bit of each of the two double
precision values in the SSE register into low two bits of a general register.
This instruction is analogous to "movmskps" and has the same rules for
operands.
  "movsd" transfers a double precision value between source and destination
operand (only the low quad word is trasferred). At least one of the operands
have to be a SSE register, the second one can be also a SSE register or 64-bit
memory location.
  Arithmetic operations on double precision values are: "addpd", "addsd",
"subpd", "subsd", "mulpd", "mulsd", "divpd", "divsd", "sqrtpd", "sqrtsd",
"maxpd", "maxsd", "minpd", "minsd", and they are analoguous to arithmetic
operations on single precision values described in previous section. When the
mnemonic ends with "pd" instead of "ps", the operation is performed on packed
two double precision values, but rules for operands are the same. When the
mnemonic ends with "sd" instead of "ss", the source operand can be a 64-bit
memory location or a SSE register, the destination operand must be a SSE
register and the operation is performed on double precision values, only low
quad words of SSE registers are used in this case.
  "andpd", "andnpd", "orpd" and "xorpd" perform the logical operations on
packed double precision values. They are analoguous to SSE logical operations
on single prevision values and have the same rules for operands.
  "cmppd" compares packed double precision values and returns and returns a
mask result into the destination operand. This instruction is analoguous to
"cmpps" and has the same rules for operands. "cmpsd" performs the same
operation on double precision values, only low quad word of destination
register is affected, in this case source operand can be a 64-bit memory or
SSE register. Variant with only two operands are obtained by attaching the
condition mnemonic from table 2.3 to the "cmp" mnemonic and then attaching
the "pd" or "sd" at the end.
  "comisd" and "ucomisd" compare the double precision values and set the ZF,
PF and CF flags to show the result. The destination operand must be a SSE
register, the source operand can be a 128-bit memory location or SSE register.
  "shufpd" moves any of the two double precision values from the destination
operand into the low quad word of the destination operand, and any of the two
values from the source operand into the high quad word of the destination
operand. This instruction is analoguous to "shufps" and has the same rules for
operand. Bit 0 of the third operand selects the value to be moved from the
destination operand, bit 1 selects the value to be moved from the source
operand, the rest of bits are reserved and must be zeroed.
  "unpckhpd" performs an unpack of the high quad words from the source and
destination operands, "unpcklpd" performs an unpack of the low quad words from
the source and destination operands. They are analoguous to "unpckhps" and
"unpcklps", and have the same rules for operands.
  "cvtps2pd" converts the packed two single precision floating point values to
two packed double precision floating point values, the destination operand
must be a SSE register, the source operand can be a 64-bit memory location or
SSE register. "cvtpd2ps" converts the packed two double precision floating
point values to packed two single precision floating point values, the
destination operand must be a SSE register, the source operand can be a
128-bit memory location or SSE register. "cvtss2sd" converts the single
precision floating point value to double precision floating point value, the
destination operand must be a SSE register, the source operand can be a 32-bit
memory location or SSE register. "cvtsd2ss" converts the double precision
floating point value to single precision floating point value, the destination
operand must be a SSE register, the source operand can be 64-bit memory
location or SSE register.
  "cvtpi2pd" converts packed two double word integers into the the packed
double precision floating point values, the destination operand must be a SSE
register, the source operand can be a 64-bit memory location or MMX register.
"cvtsi2sd" converts a double word integer into a double precision floating
point value, the destination operand must be a SSE register, the source
operand can be a 32-bit memory location or 32-bit general register. "cvtpd2pi"
converts packed double precision floating point values into packed two double
word integers, the destination operand should be a MMX register, the source
operand can be a 128-bit memory location or SSE register. "cvttpd2pi" performs
the similar operation, except that truncation is used to round a source values
to integers, rules for operands are the same. "cvtsd2si" converts a double
precision floating point value into a double word integer, the destination
operand should be a 32-bit general register, the source operand can be a
64-bit memory location or SSE register. "cvttsd2si" performs the similar
operation, except that truncation is used to round a source value to integer,
rules for operands are the same.
  "cvtps2dq" and "cvttps2dq" convert packed single precision floating point
values to packed four double word integers, storing them in the destination
operand. "cvtpd2dq" and "cvttpd2dq" convert packed double precision floating
point values to packed two double word integers, storing the result in the low
quad word of the destination operand. "cvtdq2ps" converts packed four
double word integers to packed single precision floating point values.
"cvtdq2pd" converts packed two double word integers from the low quad word
of the source operand to packed double precision floating point values.
For all these instruction destination operand must be a SSE register, the
source operand can be a 128-bit memory location or SSE register.
  "movdqa" and "movdqu" transfer a double quad word operand containing packed
integers from source operand to destination operand. At least one of the
operands have to be a SSE register, the second one can be also a SSE register
or 128-bit memory location. Memory operands for "movdqa" instruction must be
aligned on boundary of 16 bytes, operands for "movdqu" instruction don't have
to be aligned.
  "movq2dq" moves the contents of the MMX source register to the low quad word
of destination SSE register. "movdq2q" moves the low quad word from the source
SSE register to the destination MMX register.

    movq2dq xmm0,mm1   ; move from MMX register to SSE register
    movdq2q mm0,xmm1   ; move from SSE register to MMX register

  All MMX instructions operating on the 64-bit packed integers (those with
mnemonics starting with "p") are extended to operate on 128-bit packed
integers located in SSE registers. Additional syntax for these instructions
needs an SSE register where MMX register was needed, and the 128-bit memory
location or SSE register where 64-bit memory location of MMX register were
needed. The exception is "pshufw" instruction, which doesn't allow extended
syntax, but has two new variants: "pshufhw" and "pshuflw", which allow only
the extended syntax, and perform the same operation as "pshufw" on the high
or low quad words of operands respectively. Also the new instruction "pshufd"
is introduced, which performs the same operation as "pshufw", but on the
double words instead of words, it allows only the extended syntax.

    psubb xmm0,[esi]   ; substract 16 packed bytes
    pextrw eax,xmm0,7  ; extract highest word into eax

  "paddq" performs the addition of packed quad words, "psubq" performs the
substraction of packed quad words, "pmuludq" performs an unsigned multiply
of low double words from each corresponding quad words and returns the results
in packed quad words. These instructions follow the same rules for operands as
the general MMX operations described in 2.1.14.
  "pslldq" and "psrldq" perform logical shift left or right of the double
quad word in the destination operand by the amount of bits specified in the
source operand. The destination operand should be a SSE register, source
operand should be an 8-bit immediate value.
  "punpckhqdq" interleaves the high quad word of the source operand and the
high quad word of the destination operand and writes them to the destination
SSE register. "punpcklqdq" interleaves the low quad word of the source operand
and the low quad word of the destination operand and writes them to the
destination SSE register. The source operand can be a 128-bit memory location
or SSE register.
  "movntdq" stores packed integer data from the SSE register to memory using
non-temporal hint. The source operand should be a SSE register, the
destination operand should be a 128-bit memory location. "movntpd" stores
packed double precision values from the SSE register to memory using a
non-temporal hint. Rules for operand are the same. "movnti" stores integer
from a general register to memory using a non-temporal hint. The source
operand should be a 32-bit general register, the destination operand should
be a 32-bit memory location. "maskmovdqu" stores selected bytes from the first
operand into a 128-bit memory location using a non-temporal hint. Both
operands should be a SSE registers, the second operand selects wich bytes from
the source operand are written to memory. The memory location is pointed by DI
(or EDI) register in the segment selected by DS and does not need to be
aligned.
  "clflush" writes and invalidates the cache line associated with the address
of byte specified with the operand, which should be a 8-bit memory location.
  "lfence" performs a serializing operation on all instruction loading from
memory that were issued prior to it. "mfence" performs a serializing operation
on all instruction accesing memory that were issued prior to it, and so it
combines the functions of "sfence" (described in previous section) and
"lfence" instructions. These instructions have no operands.


2.1.17  AMD 3DNow! instructions

The 3DNow! extension adds a new MMX instructions to those described in 2.1.14,
and introduces operation on the 64-bit packed floating point values, each
consisting of two single precision floating point values.
  These instructions follow the same rules as the general MMX operations, the
destination operand should be a MMX register, the source operand can be a MMX
register or 64-bit memory location. "pavgusb" computes the rounded averages
of packed unsigned bytes. "pmulhrw" performs a signed multiply of the packed
words, round the high word of each double word results and stores them in the
destination operand. "pi2fd" converts packed double word integers into
packed floating point values. "pf2id" converts packed floating point values
into packed double word integers using truncation. "pi2fw" converts packed
word integers into packed floating point values, only low words of each
double word in source operand are used. "pf2iw" converts packed floating
point values to packed word integers, results are extended to double words
using the sign extension. "pfadd" adds packed floating point values. "pfsub"
and "pfsubr" substracts packed floating point values, the first one substracts
source values from destination values, the second one substracts destination
values from the source values. "pfmul" multiplies packed floating point
values. "pfacc" adds the low and high floating point values of the destination
operand, storing the result in the low double word of destination, and adds
the low and high floating point values of the source operand, storing the
result in the high double word of destination. "pfnacc" substracts the high
floating point value of the destination operand from the low, storing the
result in the low double word of destination, and substracts the high floating
point value of the source operand from the low, storing the result in the high
double word of destination. "pfpnacc" substracts the high floating point value
of the destination operand from the low, storing the result in the low double
word of destination, and adds the low and high floating point values of the
source operand, storing the result in the high double word of destination.
"pfmax" and "pfmin" compute the maximum and minimum of floating point values.
"pswapd" reverses the high and low double word of the source operand. "pfrcp"
returns an estimates of the reciprocals of floating point values from the
source operand, "pfrsqrt" returns an estimates of the reciprocal square
roots of floating point values from the source operand, "pfrcpit1" performs
the first step in the Newton-Raphson iteration to refine the reciprocal
approximation produced by "pfrcp" instruction, "pfrsqit1" performs the first
step in the Newton-Raphson iteration to refine the reciprocal square root
approximation produced by "pfrsqrt" instruction, "pfrcpit2" performs the
second final step in the Newton-Raphson iteration to refine the reciprocal
approximation or the reciprocal square root approximation. "pfcmpeq",
"pfcmpge" and "pfcmpgt" compare the packed floating point values and sets
all bits or zeroes all bits of the correspoding data element in the
destination operand according to the result of comparision, first checks
whether values are equal, second checks whether destination value is greater
or equal to source value, third checks whether destination value is greater
than source value.
  "prefetch" and "prefetchw" load the line of data from memory that contains
byte specified with the operand into the data cache, "prefetchw" instruction
should be used when the data in the cache line is expected to be modified,
otherwise the "prefetch" instruction should be used. The operand should be an
8-bit memory location.
  "femms" performs a fast clear of MMX state. This instruction has no
operands.


2.2  Control directives

This section describes the directives that control the assembly process, they
are processed during the assembly and may cause some blocks of instructions
to be assembled differently or not assembled at all.


2.2.1  Repeating blocks of instructions

"times" directive repeats one instruction specified number of times. It
should be followed by numerical expression specifying number of repeats and
the instruction to repeat (optionally colon can be used to separate number and
instruction). When special symbol "%" is used inside the instruction, it is
equal to the number of current repeat. For example "times 5 db %" will define
five bytes with values 1, 2, 3, 4, 5. Recursive use of "times" directive is
also allowed, so "times 3 times % db %" will define six bytes with values
1, 1, 2, 1, 2, 3.
  "repeat" directive repeats the whole block of instructions. It should be
followed by numerical expression specifying number of repeats. Instructions
to repeat are expected in next lines, ended with the "end repeat" directive,
for example:

    repeat 8
        mov byte [bx],%
        inc bx
    end repeat

The generated code will store byte values from one to eight in the memory
addressed by BX register.
  Number of repeats can be zero, in that case the instructions are not
assembled at all.


2.2.2  Conditional assembly

"if" directive causes come block of instructions to be assembled only under
certain condition. It should be followed by logical expression specifying the
condition, instructions in next lines will be assembled only when this
condition is met, otherwise they will be skipped. The optional "else if"
directive followed with logical expression specifying additional condition
begins the next block of instructions that will be assembled if previous
conditions were not met, and the additional condition is met. The optional
"else" directive begins the block of instructions that will be assembled if
all the conditions were not met. The "end if" directive ends the last block of
instructions.
  You should note that "if" directive is processed at assembly stage and
therefore it doesn't affect any preprocessor directives, so if you put some
symbolic constants or macroinstructions inside such block, they will get
defined even when the condition is not met.
  The logical expression consist of logical values and logical operators. The
logical operators are "~" for logical negation, "&" for logical and, "|" for
logical or. The negation has the highest priority. Logical value can be a
numerical expression, it will be false if it is equal to zero, otherwise it
will be true. Two numerical expression can be compared using one of the
following operators to make the logical value: "=" (equal), "<" (less),
">" (greater), "<=" (less or equal), ">=" (greater or equal),
"<>" (not equal). The "eq" compares any two symbols whether they are exactly
the same. The "in" operator checks whether given symbol is a member of the
list of symbols following this operator, the list should be enclosed between
"<" and ">" characters, its members should be separated with commas. The
"eqtype" operator compares any two symbols whether they are of the same type.
The "used" operator should be followed by a symbol name, it checks whether the
given symbol is used somewhere (it returns correct result even if symbol is
used only after this check). The "defined" operator can be followed by any
expression, usually just by a single symbol name; it checks whether the given
expression contains only symbols that were defined prior to it.
  The following simple example uses the "count" constant that should be
defined somewhere in source:

    if count>0
        mov cx,count
        rep movsb
    end if

These two assembly instructions will be assembled only if the "count" constant
is greater than 0.
  The next example is more complex and assumes that the symbolic constant
"reg" has been defined:

    if reg in <cs,ds,es,fs,gs,ss>
        mov dx,reg
        add ax,dx
        shl ax,1
    else if reg eq ax
        shl ax,2
    else
        add ax,reg
        shl ax,1
    end if

The first block of instructions will be assembled only if the value of
"reg" is segment register, otherwise the second or third block will
assembled whether the value of "reg" is AX register or not.
  The last example shows how to conditionally define a variable that was not
defined yet:

    if ~defined var & used var
        var db 0
    end if

The directive that defines "var" will be processed only if such symbol was
not defined yet, but is used somewhere in source.

2.2.3  Other directives

"virtual" defines virtual data at specified address. This data won't be
included in the output file, but labels defined there can be used in other
parts of source. This directive can be followed by "at" operator and the
numerical expression specifying the address for virtual data, otherwise is
uses current address, the same as "virtual at $". Instructions defining data
are expected in next lines, ended with "end virtual" directive.
  This directive can be used to create union of some variables, for example:

    GDTR dp ?
    virtual at GDTR
        GDT_limit dw ?
        GDT_address dd ?
    end virtual

It defines two labels for parts of the 48-bit variable at "GDTR" address.
  It can be also used to define labels for some structures addressed by a
register, for example:

    virtual at bx
        LDT_limit dw ?
        LDT_address dd ?
    end virtual

With such definition instruction "mov ax,[LDT_limit]" will be assembled
to "mov ax,[bx]".
  Declaring defined data values or instructions inside the virtual block would
also be useful, because the "load" directive can be used to load the values
from the virtually generated code into a constants. This directive should be
used after the code it loads but before the virtual block ends, because it
can only load the values from the same code space. For example:

    virtual at 0
        xor eax,eax
        and edx,eax
        load zeroq dword from 0
    end virtual

The above piece of code will define the "zeroq" constant containing four bytes
of the machine code of the instructions defined inside the virtual block.
  "display" directive displays the message at the assembly time. It should
be followed by the quoted strings or byte values, separated with commas. It
can be used to display values of some constants, for example:

    d1 = '0'+ $ shr 12 and 0Fh
    d2 = '0'+ $ shr 8 and 0Fh
    d3 = '0'+ $ shr 4 and 0Fh
    d4 = '0'+ $ and 0Fh
    if d1>'9'
        d1 = d1 + 'A'-'9'-1
    end if
    if d2>'9'
        d2 = d2 + 'A'-'9'-1
    end if
    if d3>'9'
        d3 = d3 + 'A'-'9'-1
    end if
    if d4>'9'
        d4 = d4 + 'A'-'9'-1
    end if
    display 'Current offset is 0x',d1,d2,d3,d4,13,10

Instructions before the "display" directive calculate four digits of 16-bit
value and convert them into characters for displaying.


2.3  Preprocessor directives

All preprocessor directives are processed before the main assembly process,
and therefore are not affected by the control directives. At this time also
all comments are stripped out.


2.3.1  Including source files

"include" directive includes the specified source file at the position where
it is used. It should be followed by the quoted name of file that should be
included, for example:

    include 'macros.inc'

The whole included file is preprocessed before preprocessing the lines next
to the line containing the "include" directive. There are no limits to the
number of included files as long as they fit in memory.
  The quoted path can contain environment variables enclosed within "%"
characters, they will be replaced with their values inside the path, both the
"\" and "/" characters are allowed as a path separators. It concerns also
paths given with the "file" and "load" directives or in the command line.


2.3.2  Symbolic constants

The symbolic constants are different from the numerical constants, before the
assembly process they are replaced with their values everywhere in source
lines after their definitions, and anything can become their values.
  The definition of symbolic constant consists of name of the constant
followed by the "equ" directive. Everything that follows this directive will
become the value of constant. If the value of symbolic constant contains
other symbolic constants, they are replaced with their values before assigning
this value to the new constant. For example:

    d equ dword
    NULL equ d 0
    d equ edx

After there three definitions the value of "NULL" constant is "dword 0" and
the value of "d" is "edx". So, for example, "push NULL" will be assembled as
"push dword 0" and "push d" will be assembled as "push edx".
  "restore" directive allows to get back previous value of redefined symbolic
constant. It should be followed by one more names of symbolic constants,
separated with commas. So "restore d" after the above definitions will give
"d" constant back the value "dword". If there was no constant defined of given
name, "restore" won't cause an error, it will be just ignored.
  Symbolic constant can be used to adjust the syntax of assembler to personal
preferences. For example the following set of definitions provides the handy
shortcuts for all the size operators:

    b equ byte
    w equ word
    d equ dword
    p equ pword
    f equ fword
    q equ qword
    t equ tword
    x equ dqword

  Because symbolic constant may also have an empty value, it can be used to
allow the syntax with "offset" word before any address value:

    offset equ

After this definition "mov ax,offset char" will be valid construction for
copying the offset of "char" variable into "ax" register, because "offset" is
replaced with an empty value, and therefore ignored.
  Symbolic constants can also be defined with the "fix" directive, which has
the same syntax as "equ", but defines constants of high priority - they are
replaced with their symbolic values even before processing the preprocessor
directives and macroinstructions, the only exception is "fix" directive
itself, which has the highest possible priority, so it allows redefinition of
constants defined this way. But when such high priority constants are found
inside the value following the "fix" directive, they are replaced with their
values before assigning this value to the new constant.


2.3.3  Macroinstructions

"macro" directive allows you to define your own complex instructions, called
macroinstructions, using which can greatly simplify the process of
programming. In its simplest form it's similar to symbolic constant
definition. For example the following definition defines a shortcut for the
"test al,0xFF" instruction:

    macro tst {test al,0xFF}

After the "macro" directive there is a name of macroinstruction and then its
contents enclosed between the "{" and "}" characters. You can use "tst"
instruction anywhere after this definition and it will be assembled as
"test al,0xFF". Defining symbolic constant "tst" of that value would give the
similar result, but the difference is that the name of macroinstruction is
recognized only as an instruction mnemonic. Also, macroinstructions are
replaced with corresponding code even before the symbolic constants are
replaced with their values. So if you define macroinstruction and symbolic
constant of the same name, and use this name as an instruction mnemonic, it
will be replaced with the contents of macroinstruction, but it will be
replaced with value if symbolic constant if used somewhere inside the
operands.
  The definition of macroinstruction can consist of many lines, because
"{" and "}" characters don't have to be in the same line as "macro" directive.
For example:

    macro stos0
     {
        xor al,al
        stosb
     }

The macroinstruction "stos0" will be replaced with these two assembly
instructions anywhere it's used.
  Like instructions which needs some number of operands, the macroinstruction
can be defined to need some number of arguments separated with commas. The
names of needed argument should follow the name of macroinstruction in the
line of "macro" directive and should be separated with commas if there is more
than one. Anywhere one of these names occurs in the contents of
macroinstruction, it will be replaced with corresponding value, provided when
the macroinstruction is used. Here is an example of a macroinstruction that
will do data alignment for binary output format:

    macro align value { rb (value-1)-($+value-1) mod value }

When the "align 4" instruction is found after this macroinstruction is
defined, it will be replaced with contents of this macroinstruction, and the
"value" will there become 4, so the result will be "rb (4-1)-($+4-1) mod 4".
  If a macroinstruction is defined that uses an instruction with the same name
inside its definition, the previous meaning of this name is used. Useful
redefinition of macroinstructions can be done in that way, for example:

    macro mov op1,op2
     {
      if op1 in <ds,es,fs,gs,ss> & op2 in <cs,ds,es,fs,gs,ss>
        push  op2
        pop   op1
      else
        mov   op1,op2
      end if
     }

This macroinstruction extends the syntax of "mov" instruction, allowing both
operands to be segment registers. For example "mov ds,es" will be assembled as
"push es" and "pop ds". In all other cases the standard "mov" instruction will
be used. The syntax of this "mov" can be extended further by defining next
macroinstruction of that name, which will use the previous macroinstruction:

    macro mov op1,op2,op3
     {
      if op3 eq
        mov   op1,op2
      else
        mov   op1,op2
        mov   op2,op3
      end if
     }

It allows "mov" instruction to have three operands, but it can still have two
operands only, because when macroinstruction is given less arguments than it
needs, the rest of arguments will have empty values. When three operands are
given, this macroinstruction will become two macroinstructions of the previous
definition, so "mov es,ds,dx" will be assembled as "push ds", "pop es" and
"mov ds,dx".
  When it's needed to provide macroinstruction with argument that contains
some commas, such argument should be enclosed between "<" and ">" characters.
If it contains more than one "<" character, the same number of ">" should be
used to tell that the value of argument ends.
  "purge" directive allows removing the last definition of specified
macroinstruction. It should be followed by one or more names of
macroinstructions, separated with commas. If such macroinstruction has not
been defined, you won't get any error. For example after having the syntax of
"mov" extended with the macroinstructions defined above, you can disable
syntax with three operands back by using "purge mov" directive. Next
"purge mov" will disable also syntax for two operands being segment registers,
and all the next such directives will do nothing.
  If after the "macro" directive you enclose some group of arguments' names in
square brackets, it will allow giving more values for this group of arguments
when using that macroinstruction. Any more argument given after the last
argument of such group will begin the new group and will become the first
argument of it. That's why after closing the square bracket no more argument
names can follow. The contents of macroinstruction will be processed for each
such group of arguments separately. The simplest example is to enclose one
argument name in square brackets:

    macro stoschar [char]
     {
        mov al,char
        stosb
     }

This macroinstruction accepts unlimited number of arguments, and each one
will be processed into these two instructions separately. For example
"stoschar 1,2,3" will be assembled as the following instructions:

    mov al,1
    stosb
    mov al,2
    stosb
    mov al,3
    stosb

  There are some special directives available only inside the definitions of
macroinstructions. "local" directive defines local names, which will be
replaced with unique values each time the macroinstruction is used. It should
be followed by names separated with commas. This directive is usually needed
for the constants or labels that macroinstruction defines and uses internally.
For example:

    macro movstr
     {
        local move
      move:
        lodsb
        stosb
        test al,al
        jnz move
     }

Each time this macroinstruction is used, "move" will become other unique name
in its instructions, so you won't get an error you normally get when some
label is defined more than once.
  "forward", "reverse" and "common" directives divide macroinstruction into
blocks, each one processed after the processing of previous is finished. They
differ in behavior only if macroinstruction allows multiple groups of
arguments. Block of instructions that follows "forward" directive is processed
for each group of arguments, from first to last - exactly like the default
block (not preceded by any of these directives). Block that follows "reverse"
directive is processed for each group of argument in reverse order - from last
to first. Block that follows "common" directive is processed only once,
commonly for all groups of arguments. Local name defined in one of the blocks
is available in all the following blocks when processing the same group of
arguments as when it was defined, and when it is defined in common block it is
available in all the following blocks not depending on which group of
arguments is processed.
  Here is an example of macroinstruction that will create the table of
addresses to strings followed by these strings:

    macro strtbl name,[string]
     {
      common
        label name dword
      forward
        local label
        dd label
      forward
        label db string,0
     }

First argument given to this macroinstruction will become the label for table
of addresses, next arguments should be the strings. First block is processed
only once and defines the label, second block for each string declares its
local name and defines the table entry holding the address to that string.
Third block defines the data of each string with the corresponding label.
  The directive starting the block in macroinstruction can be followed by the
first instruction of this block in the same line, like in the following
example:

    macro stdcall proc,[arg]
     {
      reverse push arg
      common call proc
     }

This macroinstruction can be used for calling the procedures using STDCALL
convention, arguments are pushed on stack in the reverse order. For example
"stdcall foo,1,2,3" will be assembled as:

    push 3
    push 2
    push 1
    call foo

  If some name inside macroinstruction has multiple values (it is either one
of the arguments enclosed in square brackets or local name defined in the
block following "forward" or "reverse" directive) and is used in block
following the "common" directive, it will be replaced with all of its values,
separated with commas. For example the following macroinstruction will pass
all of the additional arguments to the previously defined "stdcall"
macroinstruction:

    macro invoke proc,[arg]
     { common stdcall [proc],arg }

It can be used to call indirectly (by the pointer stored in memory) the
procedure using STDCALL convention.
  Inside macroinstruction also special operator "#" can be used. This
operator causes two names to be concatenated into one name. It can be useful,
because it's done after the arguments and local names are replaced with their
values. The following macroinstruction will generate the conditional jump
according to the "cond" argument:

    macro jif op1,cond,op2,label
     {
        cmp op1,op2
        j#cond label
     }

For example "jif ax,ae,10h,exit" will be assembled as "cmp ax,10h" and
"jae exit" instructions.
  To make macroinstruction behaving differently when some of the arguments are
of some special type, for example a quoted strings, you can use "eqtype"
comparision operator. Here's an example of utilizing it to distinguish a
quoted string from an other argument:

    macro message arg
     {
      if arg eqtype ""
        local str
        jmp   @f
        str   db arg,0Dh,0Ah,24h
        @@:
        mov   dx,str
      else
        mov   dx,arg
      end if
        mov   ah,9
        int   21h
     }

The above macro is designed for displaying messages in DOS programs. When the
argument of this macro is some number, label, or variable, the string from
that address is displayed, but when the argument is a quoted string, the
created code will display that string followed by the carriage return and
line feed.


2.3.4  Structures

"struc" directive is a special variant of "macro" directive that is used to
define data structures. Macroinstruction defined using the "struc" directive
must be preceded by a label (like the data definition directive) when it's
used. This label will be also attached at the beginning of every name starting
with dot in the contents of macroinstruction. The macroinstruction defined
using the "struc" directive can have the same name as some other
macroinstruction defined using the "macro" directive, structure
macroinstruction won't prevent the standard macroinstruction being processed
when there is no label before it and vice versa. All the rules concerning
standard macroinstructions apply to structure macroinstructions.
  Here is the sample of structure macroinstruction:

    struc point x,y
     {
        .x dw x
        .y dw y
     }

For example "my point 7,11" will define structure labelled "my", consisting of
two variables: "my.x" with value 7 and "my.y" with value 11.
  Next example shows how to extend the data definition directive "db" with
ability to calculate the size of defined data by using the structure
macroinstruction:

    struc db [data]
     {
      common
        label .data byte
        db data
        .size = $-.data
     }

With such definition for example "msg db 'Hello!',13,10" will define also
"msg.size" constant, equal to the size of defined data in bytes and also
additional label "msg.data", which will be recognized as a label for data of
byte size.
  Defining data structures addressed by registers or absolute values should be
done using the "virtual" directive with structure macroinstruction
(see 2.2.3).


2.4  Formatter directives

"format" directive followed by the format identifier allows to select the
output format. This directive should be put at the beginning of the source.
Default output format is a flat binary file, it can also be selected by using
"format binary" directive.
  "use16" and "use32" directives force the assembler to generate 16-bit or
32-bit code, omitting the default setting for selected output format.
  "org" directive sets address at which the following code is expected to
appear in memory. It should be followed by numerical expression specifying
the address. You can also use this directive in the "$=" form followed by
numerical expression.
  Below are described different output formats with the directives specific to
these formats.


2.4.1  MZ executable

To select the MZ output format, use "format MZ" directive. The default code
setting for this format is 16-bit.
  "segment" directive defines a new segment, it should be followed by label,
which value will be the number of defined segment, optionally "use16" or
"use32" word can follow to specify whether code in this segment should be
16-bit or 32-bit. The origin of segment is aligned to paragraph (16 bytes).
All the labels defined then will have values relative to the beginning of this
segment.
  "entry" directive sets the entry point for MZ executable, it should be
followed by the far address (name of segment, colon and the offset inside
segment) of desired entry point.
  "stack" directive sets up the stack for MZ executable. It can be followed by
numerical expression specifying the size of stack to be created automatically
or by the far address of initial stack frame when you want to set up the stack
manually. When no stack is defined, the stack of default size 4096 bytes will
be created.
  "heap" directive should be followed by a 16-bit value defining maximum size
of additional heap in paragraphs (this is heap in addition to stack and
undefined data). Use "heap 0" to always allocate only memory program really
needs. Default size of heap is 65535.


2.4.2  Portable Executable

To select the Portable Executable output format, use "format PE" directive, it
can be followed by additional format settings: use "console", "GUI" or
"native" operator selects the target subsystem (floating point value
specifying subsystem version can follow), "DLL" marks the output file as a
dynamic link library. Then can follow the "at" operator and the numerical
expression specifying the base of PE image and then optionally "on" operator
followed by the quoted string containing file name selects custom MZ stub for
PE program (when specified file is not a MZ executable, it is treated as a
flat binary executable file and converted into MZ format). The default code
setting for this format is 32-bit.
  "section" directive defines a new section, it should be followed by quoted
string defining the name of section, then one or more section flags can
follow. Available flags are: "code", "data", "readable", "writeable",
"executable", "shareable", "discardable", "notpageable". Among with flags also
one of the special PE data identifiers can be specified to mark the whole
section as a special data, possible identifiers are "export", "import",
"resource" and "fixups". If the section is marked to contain fixups, they are
generated automatically and no more data needs to be defined in this section.
The origin of section is aligned to page (4096 bytes).
  "entry" directive sets the entry point for Portable Executable, the value of
entry point should follow.
  "stack" directive sets up the size of stack for Portable Executable, value
of stack reserve size should follow, optionally value of stack commit
separated with comma can follow. When stack is not defined, it's set by
default to size of 4096 bytes.
  "heap" directive chooses the size of heap for Portable Executable, value of
heap reserve size should follow, optionally value of heap commit separated
with comma can follow. When no heap is defined, it is set by default to size
of 65536 bytes, when size of heap commit is unspecified, it is by default set
to zero.
  "data" directive begins the definition of special PE data, it should be
followed by one of the data identifiers ("export", "import", "resource" or
"fixups") or by the number of data entry in PE header. The data should be
defined in next lines, ended with "end data" directive. When fixups data
definition is chosen, they are generated automatically and no more data needs
to be defined there.


2.4.3  Common Object File Format

To select Common Object File Format, use "format COFF" or "format MS COFF"
directive whether you want to create classic or Microsoft's COFF file. The
default code setting for this format is 32-bit.
  "section" directive defines a new section, it should be followed by quoted
string defining the name of section, then one or more section flags can
follow. Available flags are: "code" and "data" for both COFF variants,
"readable", "writeable", "executable", "shareable", "discardable" and
"notpageable" only for Microsoft COFF variant.
  "extrn" directive defines the external symbol, it should be followed by the
name of symbol and optionally the size operator specifying the size of data
labelled by this symbol. The name of symbol can be also preceded by quoted
string containing name of the external symbol and the "as" operator.
  "public" directive declares the existing symbol as public, it should be
followed by the name of symbol, optionally it can be followed by the "as"
operator and the quoted string containing name under which symbol should be
available as public.


2.4.4  Executable and Linkable Format

To select ELF output format, use "format ELF" directive. The default code
setting for this format is 32-bit.
  "section" directive defines a new section, it should be followed by quoted
string defining the name of section, then can follow one or both of the
"executable" and "writeable" flags. The origin of section is aligned to
double word (4 bytes).
  "extrn" and "public" directives have the same meaning and syntax as when the
COFF output format is selected (described in previous section).
  To create executable file, use "format ELF executable" directive. It allows
to use "entry" directive followed by the value to set as entry point of
program. On the other hand it makes "extrn" and "public" directives
unavailable. "section" directive in this case can be followed only by one or
more section flags and its origin is aligned to page (4096 bytes). Available
flags for section are: "readable", "writeable" and "executable".


EOF