What is assembler?
I always read things related to assembler and get confused. At first I thought this was a mess and that assembler was the same thing as assembly. But it seems that's not it.
What is assembler? And what is your relationship with assembly?
4 answers
Assemble
Assembler, as its name says, is a assembler, and Not a compiler, although it works very similarly. It takes a text that is a programming code and turns it into binary code (machine code). What differs from a compiler is precisely that the instructions in the language have a one-to-one relationship in binary code. Parsing and transforming code into an assembler is much simpler than a compiler (the lexer is more or less equal, the parser is simpler, and semantic analysis in general does not exist).
Assembly
The low-level programming language is the assembly (assembly language), which is assembled by a assembler (the suffix er in English is equal to our dor and indicates an agent of an action). We use capital letters because it's a first name.
There are several dialects for each architecture physical (x86 processor, ARM, MIPS, etc.), or virtual (JVM, CIL, and many dynamic languages have their own Assembly). The semantics depend on the architecture, but the syntax does not, and the specific assembler can adopt whatever they prefer.
Some people confuse machine and Assembly code. The first is binary, the Assembly is at a level that humans (normal, although some will say that these are not so normal like this :P) understand. Are mnemonics that define the instructions that the processor should run.
Assembly is all imperative and each mnemonic is a very simple statement manipulating a value in registers, moving the data between registers and memory and controlling the execution flow in the simplest possible way, no abstractions, No syntax sugar, no ready design patterns.
When the person says they will program in assembler they make the same mistake as the person who says they will Program in Visual Study.
Assembly
But assembly in lowercase is that file with binary CLR code (aka .NET).
Examples
X86-64 Assembly code example Intel syntax:
; Assembler (x86) version of 99 Bottles of beer
;
; This version is for NASM compiler but doesn't use any
; macros, just all basic instructions for x86 assembler.
; Also only putchar() function is used to print character
; onto screen, and the whole rest is in code.
;
; nasm -fwin32 99.asm
; gcc -o 99.exe 99.obj
global _main
extern _putchar
segment .data
_line_1_1 db ' bottles of beer on the wall, ', 0
_line_1_2 db ' bottles of beer.', 13, 10, 0
_line_2_1 db 'Take one down and pass it around, ', 0
_line_2_2 db ' bottles of beer on the wall.', 13, 10, 13, 10, 0
_line_2_2_one db ' bottle of beer on the wall.', 13, 10, 13, 10, 0
_ending_lines db '1 bottle of beer on the wall, 1 bottle of beer.', 13, 10
db 'Take one down and pass it around, no more bottles of beer on the wall.', 13,
10, 13, 10
db 'No more bottles of beer on the wall, no more bottles of beer. ', 13, 10
db 'Go to the store and buy some more, 99 bottles of beer on the wall.', 13, 10, 0
segment .text
; this function converts integer in range 0-99 to string
_integer_to_string:
mov eax, dword [esp + 08h] ; get the vavlue
mov ecx, 10 ;
sub edx, edx
div ecx ; divide it by 10
mov ecx, dword [esp + 04h] ; get the output offset
test eax, eax ; is greater than 9
jz .skip_first_digit ; skip saving 0 char if no
add al, 030h ; convert number to ascii char
mov byte [ecx], al ; save
inc ecx ; increase pointer
jmp .dont_test_second_digit ;
.skip_first_digit: ; only if less then 10
test edx, edx
jz .skip_second_digit
.dont_test_second_digit: ; if it was greater than 10
add dl, 030h ; than second digit must by
mov byte [ecx], dl ; written at no condition
inc ecx
.skip_second_digit: ; only skip if value was 0
mov byte [ecx], ah ; save the null ending char
retn 4 ; ret and restore stack
; function prints null-terminated line to stdout
_show_line:
push edi ; function save registers
push esi
mov edi, dword [esp + 0Ch] ; get the pointer to string
sub eax, eax ; look for zeros
sub ecx, ecx
dec ecx ; set ecx to -1
repnz scasb ; search for 0 in string
neg ecx
sub ecx, 2 ; get the string length w/o zero
mov esi, dword [esp + 0Ch] ; get pointer once again
.putchar_loop:
push ecx ; keep the counter
lodsb ; get the char
push eax
call _putchar ; print char to stdout
add esp, 4 ; correct stack
pop ecx ; get back the counter
dec ecx
jnz .putchar_loop ; if not last char then get next
pop esi ; restore registers
pop edi
retn 4
; prints string for only one number
_bottles:
push ebp ; keep the offset to call params
mov ebp, esp
sub esp, 4 ; reserve one local variable
mov eax, dword [ebp + 08h] ; get number of bottles
dec eax ; is it 1?
jnz .more_than_one ; nope, it's not
push _ending_lines ; print the last lines
call _show_line
jmp .end ; exit function
.more_than_one:
inc eax ; get the original value
push eax ; convert it to string
lea eax, [ebp - 04h]
push eax ; string will be stored here
call _integer_to_string
lea eax, [ebp - 04h]
push eax
call _show_line ; 'xx'
push _line_1_1
call _show_line ; ' bottles of beer on the wall, '
lea eax, [ebp - 04h]
push eax
call _show_line ; 'xx'
push _line_1_2
call _show_line ; ' bottles of beer.'
mov eax, dword [ebp + 08h]
dec eax ; in second line the value is one less
push eax
lea eax, [ebp - 04h]
push eax
call _integer_to_string ; convert it to string
push _line_2_1
call _show_line ; 'Take one down and pass it around, '
lea eax, [ebp - 04h]
push eax
call _show_line ; 'xx'
cmp dword [ebp + 08h], 2
jnz .second_line_for_more_than_one
push _line_2_2_one ; ' bottle of beer on the wall.'
jmp .show_line
.second_line_for_more_than_one:
push _line_2_2 ; ' bottles of beer on the wall.'
.show_line:
call _show_line
.end:
leave
retn 4
; main function, the command line arguments are not important
_main:
pushad
mov ecx, 99 ; printf from 99
.main_loop:
push ecx
push ecx
call _bottles ; print lines for this value
pop ecx
loop .main_loop ; if still greater than zero
popad
sub eax, eax ; That's all folks!
retn
No ARM:
;99 Bottles of Beer generator
;For ARM processors running RISCOS
;Using built in BASIC assembler
;
MOV R7, #99 ;bottle count kept in R7
MOV R12, R14 ;store caller return address
.beginverse ;(_prints verses then returns to caller_)
BL bottlesofbeer
ADR R0, onthewall
SWI "OS_Write0" ;prints string at address in R0
BL bottlesofbeer
SWI "OS_NewLine"
ADR R0, take
SWI "OS_Write0"
SUBS R7,R7,#1
BLNE bottlesofbeer ;beer left
BLEQ nobeer ;no beer left
ADR R0, onthewall
SWI "OS_Write0"
SWI "OS_NewLine"
SWI "OS_NewLine"
BNE beginverse ;go again if there's beer left
BL buymorebeer ;print last verse
MOV PC, R12 ;exit to caller
.bottlesofbeer ;(_prints "x bottle(s) of beer"_)
MOV R0, R7 ;arg1- number of bottles
ADR R1, bottlenum ;arg2- buffer address
MOV R2, #3 ;arg3- buffer length
SWI "OS_ConvertInteger3" ;convert number of beers to string
SWI "OS_Write0" ;and print it
CMP R7, #1
ADR R0, bottles ;
ADREQ R0, bottle ;bottles is replaced with bottle if 1 bottle left
SWI "OS_Write0"
ADR R0, ofbeer
SWI "OS_Write0"
CMP R1, #0 ;unset zero flag so "nobeer" doesnt execute after return
MOV PC, R14 ;return
.buymorebeer ;(_prints final verse_)
MOV R11, R14 ;save return address
BL nobeer
ADR R0, onthewall
SWI "OS_Write0"
ADR R0, comma
SWI "OS_Write0"
BL nobeer
SWI "OS_NewLine"
ADR R0, gotostore
SWI "OS_Write0"
MOV PC, R11 ;return to saved address
.nobeer ;(_prints "no more bottles of beer"_)
ADR R0, nomore
SWI "OS_Write0"
ADR R0, bottles
SWI "OS_Write0"
ADR R0, ofbeer
SWI "OS_Write0"
MOV PC, R14
;string components
.ofbeer
EQUS "of beer" ;string contents
EQUB 0 ;zero terminator
.onthewall
EQUS " on the wall "
EQUB 0
.bottle
EQUS " bottle "
EQUB 0
.bottles
EQUS " bottles "
EQUB 0
.take
EQUS "Take one down and pass it around, "
EQUB 0
.nomore
EQUS "no more"
EQUB 0
.bottlenum
EQUS " "
EQUB 0
.comma
EQUS ","
EQUB 0
.gotostore
EQUS "Go to the store and buy some more...99 bottles of beer."
EQUB 0
Source .
Did you notice why they say you should comment on your code? That's where you should be.
Related questions:
What is assembler? And what is your relationship with assembly?
Assemble it's a compiler. It converts code written in the language Assembly for native code.
Assembly is a programming language. But it is not a typical language. It is characterized by being a low-level programming language composed of rigid and simple format instructions that do not allow substructures and by some labels (labels that are targets for bypass instructions). It usually (almost always, but there are some cases that it doesn't) is mapped one-to-one into instructions to be executed by the processor (a processor instruction = an instruction in assembly).
Each statement is defined by a mnemonic. For example, in a MOV eax, 1 or JN algum_label Statement, MOV and JN are the mnemonics, which are the ones that also define the name of the statement in question.
Assembler is the program that converts the assembly code to the instructions themselves (encoded as a sequence of bytes). That is, the assembler is the compiler. Despite this, confusion between the terms assembly and assembler is common, and many people talk about " to program in assembler " when in fact it should be " to program in assembly".
Since there are several types of processors, each with its own instruction set, this means that for each processor we have at least one assembly dialect. Different assembler developers may use different notations or different mnemonics for instructions, and therefore, even in the same architecture, there may be several distinct dialects of assembly.
I found this definition very relevant:
Machine Code:
it is the output compiled by an assembler compiler.
Assembly
it is the machine-readable form of the code.
Assembly Language
refers to a specific machine code language with x86 assembly.
Assemble
it is the tool used to compile source code into source code. machine.
Assembler Language
it is the language used by any assembler assembler.
I translated this answer SO-en , I thought it was good, but my English is not "the best"...