Update README.md

fix build test

.gitea/workflows/build_test.yaml

@@ -0,0 +1,17 @@
+name: build_test
+on: [push]
+
+jobs:
+  build:
+    runs-on: arch
+    steps:
+      - name: fix package cache
+        run: rm -fr /var/cache/pacman/pkg/* && pacman -Syy archlinux-keyring --needed --noconfirm && pacman-key --init && pacman-key --populate archlinux
+      - name: get dependencies
+        run: pacman -Sy nodejs make nasm binutils --needed --noconfirm
+
+      - name: Checkout code
+        uses: actions/checkout@v4
+      
+      - name: Build
+        run: make -j$(nproc)

Makefile

@@ -8,6 +8,11 @@ DEBUG_LD_FLAGS := -g
 DEBUG_NASM_FLAGS := -g -F dwarf
 
 # check for avx2 support
+ifeq ($(shell grep -o 'avx2[^ ]*' /proc/cpuinfo | head -n 1),avx512)
+    NASM_FLAGS += -DAVX2
+endif
+
+# check for avx512 support
 ifeq ($(shell grep -o 'avx512[^ ]*' /proc/cpuinfo | head -n 1),avx512)
     NASM_FLAGS += -DAVX512
 endif

README.md

@@ -0,0 +1,39 @@
+<img src="https://git.pupes.org/repo-avatars/e4ede9d30f070c9e191eace5a88dcaa40434b9cadf60204122fab5a83aec9a9f" alt="pulsar in asm game of life" width="196"></img>
+[![build test](https://git.pupes.org/PoliEcho/asm-game-of-life/actions/workflows/build_test.yaml/badge.svg)](https://git.pupes.org/PoliEcho/asm-game-of-life/actions?workflow=build_test.yaml)
+# AMD64 Assembly Game of life
+
+> [!WARNING]  
+> this program will break if your terminal is too small
+
+## Dependencies  
+> AMD64 Linux Kernel  
+### Build only  
+> nasm  
+> ld  
+> make  
+
+## Download
+[releases](https://git.pupes.org/PoliEcho/asm-game-of-life/releases)
+
+## Build  
+```shell
+make
+```
+
+## Controls
+| key    | action                  |
+|--------|-------------------------|
+| arrows | move cursor             |
+| ENTER  | invert cell             |
+| j/k    | change simulation speed |
+| p      | start/stop simulation   |
+| q      | quit                    |
+
+
+## Warning  
+delays in TTY may be diferent depending on cpu clockspeed use j/k to adjust
+
+## Notes  
+if screen does not clear properly after loading just move the cursor around a bit  
+looks best in TTY  
+it sometimes SEGFAULTS in TTY on start but only on non debug builds and it is not consistent

demo.txt

@@ -0,0 +1 @@
+ssh demo@demo.pupes.org # password: "demo"

src/drawing.asm

@@ -14,27 +14,45 @@ section .bss
 	simulation_running: RESB 1
 
 	next_frame_ptr: RESQ 1
+section .data 
+	extern simulation_speed
 
 section .rodata 
 	clear: 		db ESC_CHAR, "[2J", 0
+	global reset
 	reset:		db ESC_CHAR, "[0m", 0
 	resetLen:	equ $-reset-1
 	global resetLen
 
 	home_cursor:	db ESC_CHAR, "[H", 0
 
-	statusbar:	db ESC_CHAR, "[30;100m", "Use arrow keys to move cursor, enter to invert cell j/k to change simulation speed, p to       simulation", 0
-	START_STOP_pos: equ $-statusbar-17
+	statusbar:	db ESC_CHAR, "[32;100m", "Use arrow keys to move cursor, enter to invert cell j/k to change simulation speed, p to       simulation. Frame Delay multiplier:", 0
+	statusbarLen:	equ $-statusbar
+	START_STOP_pos: equ statusbarLen-42
 	
 	
 	start_str:	db "START", 0
 	stop_str:	db "STOP ", 0
 
+	alive_switch_statement:
+		dq step_simulation.die; 0
+		dq step_simulation.die; 1
+		dq step_simulation.live; 2
+		dq step_simulation.live; 3
+		dq step_simulation.die; 4
+		dq step_simulation.die; 5
+		dq step_simulation.die; 6
+		dq step_simulation.die; 7
+		dq step_simulation.die; 8
+
+
 section .text 
 extern print_str
 extern string_copy
 extern memory_set
+extern memory_copy
 extern alloc
+extern unsigned_int_to_ascii
 
 global init_gameboard
 init_gameboard:
@@ -81,6 +99,7 @@ print_game_ui:
 	push rdi
 	add rdi, [gameboard_size]
 	sub di, [term_cols]
+	push rdi
 	add rdi, START_STOP_pos
 	
 	mov cl, [simulation_running]
@@ -93,75 +112,145 @@ print_game_ui:
 	.end_simulation_running_check:
 	call string_copy
 
+	pop rdi
+	add rdi, statusbarLen 
+
+	movss xmm0, [simulation_speed]
+
+	cvttss2si rsi, xmm0
+	push rsi
+	push rdi
+	call unsigned_int_to_ascii
+	pop rdi 
+	add rdi, rax
+
+	mov byte [rdi], '.'
+	inc rdi
+
+	pop rax
+	cvtsi2ss xmm1, rax
+	
+	subss xmm0, xmm1; get only value after decimal point 
+	mov rax, 10 
+	cvtsi2ss xmm1, rax
+
+	mulss xmm0, xmm1
+	cvttss2si rsi, xmm0; get first decimal point as int 
+
+	call unsigned_int_to_ascii; rdi already set
+	
+
 	pop rdi
 	call print_str
 	
 	
 	ret
 
-%macro check_if_hashtag 1
-	cmp r8, %1
-	jl +7
-	cmp r9, %1 
-	ja +5
+%macro check_if_hashtag 2
+	cmp %1, r8
+	jl .no_count_%2
+	cmp %1, r9
+	ja .no_count_%2
 	mov r11b, [%1]
 	cmp r11b, '#' 
-	jne +2 
+	jne .no_count_%2
 	inc dl
+	.no_count_%2:
 %endmacro
 
 global step_simulation:
-step_simultion:
+step_simulation:
 	mov rdi, [next_frame_ptr]; destination
 	mov rsi, [gameboard_ptr]; source 
 	mov rcx, [gameboard_size]; number of iterations
-
+	
 	mov r8, rsi; store lowest address posible so we are not checking out of bounds
 	mov r9, rsi
 	add r9, rcx; store higest address posible so we are not checking out of bounds
 
-	mov r10, [term_cols]
+	xor r10, r10
+	mov r10w, [term_cols]
 	;mov r11, [term_rows] this register has been confiscated since i cannot use ah because of error: cannot use high byte register in rex instruction
 
+	sub rcx, r10; remove status bar
+
 	xor rax, rax; this shouldn't be needed but just to be sure
 	xor r11, r11
 	xor rdx, rdx; we will use dl as # counter 
 	.for_every_column_on_gameboard:
-	mov al, [rdi]; NOTE to self if i need extra register i can shift this to ah and free up r11
+	xor dl, dl
+	mov al, [rsi]; NOTE to self if i need extra register i can shift this to ah and free up r11
 	
 	
-	inc rdi
-	check_if_hashtag rdi
-	dec rdi
+	inc rsi
+	check_if_hashtag rsi, 1; check column to the to the right
+	dec rsi
 
-	check_if_hashtag rdi-1
+	dec rsi
+	check_if_hashtag rsi, 2; check the one to the left
+	inc rsi
 
-
-	add rdi, r10
+	add rsi, r10
 	
-	check_if_hashtag rdi
+	check_if_hashtag rsi, 3; check the one to the down
 
+	inc rsi
+	check_if_hashtag rsi, 4; check the one to the down-right
+	dec rsi
+
+	dec rsi
+	check_if_hashtag rsi, 5; check the one to the down-left
+	inc rsi
+
+	sub rsi, r10
+
+
+	sub rsi, r10	
+	check_if_hashtag rsi, 6; check the one to the up
+
+	inc rsi
+	check_if_hashtag rsi, 7; check the one to the up-right
+	dec rsi
+
+	dec rsi
+	check_if_hashtag rsi, 8; check the one to the up-left
+	inc rsi
+
+	add rsi, r10
+
+	cmp al, '#'
+	jne .dead_cell
+
+	jmp [alive_switch_statement+(rdx*8)]
+
+	.die:
+	mov byte [rdi], 0x20; SPACE
+	jmp .end_check
+
+	.live:
+	mov byte [rdi], '#'
+	jmp .end_check
+
+	.dead_cell:
+	cmp dl, 3 
+	jne .fill_space_there
+	mov byte [rdi], '#'
+	jmp .end_check
+	.fill_space_there:
+	mov byte [rdi], 0x20; SPACE
+
+	.end_check:
+	dec rcx
 	inc rdi
-	check_if_hashtag rdi
-	dec rdi
+	inc rsi
+	test rcx, rcx
+	jnz .for_every_column_on_gameboard
 
-	check_if_hashtag rdi-1
+	mov rsi, [next_frame_ptr]; source
+	mov rdi, [gameboard_ptr]; destination 
+	mov rdx, [gameboard_size]; number of iterations
+	sub rdx, r10; remove statusbar
 
-	sub rdi, r10
-
-
-	sub rdi, r10	
-	check_if_hashtag rdi
-
-	inc rdi
-	check_if_hashtag rdi
-	dec rdi
-
-	check_if_hashtag rdi-1
-
-	add rdi, r10
-
-
-	; TODO create jump table
+	call memory_copy
 
 	ret

src/input.asm

@@ -1,8 +1,8 @@
 %include "symbols.asm"
 
 section .bss
-	cursor_rows: RESW 1; TODO DONT FORGET TO INICIALIZE
-	cursor_cols: RESW 1
+	alignb 16
+	termios: RESZ 1; 60 bytes is needed i use 64 for alligment and it is easier to work with
 
 	extern multipurpuse_buf
 
@@ -13,13 +13,25 @@ section .bss
 
 	extern simulation_running
 
+	extern running_in_tty
+section .data 
+	cursor_rows: dw 1
+	cursor_cols: dw 1
+
+
+	global simulation_speed
+	simulation_speed dd 1.0
 section .rodata
+	extern reset
 
 	cursor_up: db ESC_CHAR, "[1A", 0
 	cursor_down: db ESC_CHAR, "[1B", 0
 	cursor_right: db ESC_CHAR, "[1C", 0
 	cursor_left: db ESC_CHAR, "[1D", 0
-	
+
+	cursor_color: db ESC_CHAR, "[45m", 0
+
+	speed_multiplier dd 50.0
 
 	arrow_switch_statement:
 		dq handle_user_input.arrow_up
@@ -30,84 +42,166 @@ section .rodata
 section .text
 
 extern print_str
+extern step_simulation
+extern unsigned_int_to_ascii
+extern print_game_ui
+extern string_copy
 
 global handle_user_input
 handle_user_input:; main loop of the program
+	push r12
+	push r13
+	push r14
+	
+
+	lea r12, [multipurpuse_buf]
 
 	.main_loop:
+	xor r13, r13
+
+	; put the cursor where it should be 
+	call print_cursor
+	
+	
 
 	xor rax, rax
-	mov qword [multipurpuse_buf], rax; zeroout the buffer
+	mov qword [r12], rax; zeroout the buffer
 
 	mov rax, SYS_POLL
-	mov rdi, STDIN
+	mov dword [r12], STDIN; create pollfd struct
+	mov word [r12+4], POLLIN
+	mov rdi, r12
 	mov rsi, 1; only one file descriptor is provided
-	mov rdx, 0; no timeout. maybe use this for final sleep but run if user inputs something TODO
+	movss xmm0, [speed_multiplier]
+	movss xmm1, [simulation_speed]
+	mulss xmm0, xmm1; callculate sleep lenght
+	cvttss2si rdx, xmm0; truncate and copy to rdx
+	mov r14, rdx
 	syscall
 
+	mov sil, [running_in_tty]
+	test sil, sil
+	jz .skip_tty_mul
+	push rax
+	mov rax, 2500; magic number
+	xor rdx,rdx
+	mul r14
+	mov r14, rax
+	pop rax
+
+	.skip_tty_mul:
 	test rax, rax; SYS_POLL returns 0 when no change happens within timeout
-	jz .no_input 
+	jz .no_input
+
+	.repeat_read:
+
+	xor rax, rax
+	mov qword [r12], rax; zeroout the buffer
 	
 	mov rax, SYS_READ
 	mov rdi, STDIN
-	lea rsi, [multipurpuse_buf]
+	lea rsi, [r12]
 	mov rdx, 8; size of multipurpuse buffer
 	syscall; read user input
 
-	mov rax, [multipurpuse_buf]
-	shr rax, 5; we need only 3 bytes for this inpus sceame
+	cmp rax, EAGAIN
+	jne .handle_input
+	mov al, [running_in_tty]
+	test al, al
+	jz .no_input
+	; this runs only if running in tty
 
-	cmp eax, 0x001B5B44; check if input is more than left arrow
-	ja .handle_single_byte_chars
+	test r14, r14
+	jz .no_input; timeout 
+	dec r14
+	jmp .repeat_read
+
+	.handle_input:
+	mov rax, [r12]
+	
+	cmp eax, 0x00415B1B; check if input is more than left arrow
+	jl .handle_single_byte_chars
+
+	bswap eax
+
+	sub eax, 0x1B5B4100
+	shr eax, 8
+	cmp al, 3
+	ja .no_input
+
+	mov r9w, [term_rows]
+	dec r9w
+	mov r10w, [term_cols]
 
-	sub eax, 0x1B5B41
 	jmp [arrow_switch_statement+(rax*8)]; lets hope this works
 
 	.arrow_up:
 	dec word [cursor_rows]
+	jnz .move_cursor_up
+	inc word [cursor_rows]
+	jmp .end_input_handling
+	.move_cursor_up:
 	lea rdi, [cursor_up]
 	call print_str
-	jmp .no_input
+	jmp .end_input_handling
+
 	.arrow_down:
-	inc word [cursor_rows]
+	mov r8w, [cursor_rows]
+	inc r8w
+	cmp word r8w, r9w
+	ja .end_input_handling
+	mov word [cursor_rows], r8w
 	lea rdi, [cursor_down]
 	call print_str
-	jmp .no_input
+	jmp .end_input_handling
+
 	.arrow_right:
- 	inc word [cursor_cols]
+ 	mov r8w, [cursor_cols]
+	inc r8w
+	cmp word r8w, r10w
+	ja .end_input_handling
+	mov word [cursor_cols], r8w
 	lea rdi, [cursor_right]
 	call print_str
-	jmp .no_input
+	jmp .end_input_handling
+
 	.arrow_left:
 	dec word [cursor_cols]
+	jnz .move_cursor_left 
+	inc word [cursor_cols]
+	jmp .end_input_handling
+	.move_cursor_left:
 	lea rdi, [cursor_left]
 	call print_str
-	jmp .no_input
+	jmp .end_input_handling
 
 	.handle_single_byte_chars:
 
-	shr eax, 2; get the char to al 
 
 	cmp al, 0xa; NEWLINE (enter key)
 	jne .check_p
 
 	xor rax, rax; zeroout rax
 	mov ax, [cursor_rows]
-	mul dword [term_cols]
-	add rax, [cursor_cols]
+	dec ax
+	mul word [term_cols]
+	mov cx, [cursor_cols]
+	dec cx
+	add ax, cx
 	
-	lea rdi, [gameboard_ptr+rax]
+	mov rdi, [gameboard_ptr]
+	add rdi, rax
 	mov cl, [rdi]
 	cmp cl, '#'
 	je .hashtag_present
 
 	mov byte [rdi], '#'
-	jmp .no_input
+	jmp .end_input_handling
 
 	.hashtag_present:
 
 	mov byte [rdi], ' '
-	jmp .no_input
+	jmp .end_input_handling
 
 	.check_p:
 	cmp al, 'p'
@@ -115,27 +209,202 @@ handle_user_input:; main loop of the program
 
 	xor byte [simulation_running], 0x01; switch simulation on or off
 
-	jmp .no_input
+	jmp .end_input_handling	
 
 	.check_j:
 	cmp al, 'j'
 	jne .check_k
 
-	; TODO implement simulation speed
+	movss xmm0, [simulation_speed]
+	mov eax, 0x3DCCCCCD; 0.1f
+	vmovd xmm1, eax
 
-	jmp .no_input
+	addss xmm0, xmm1
+	; wont check for overflows since the user would need be crazy to reach that number
+	movss [simulation_speed], xmm0
+
+	jmp .end_input_handling	
 
 	.check_k:
 	cmp al, 'k'
-	jne .no_input
+	jne .check_q
+	
+	movss xmm0, [simulation_speed]
+	mov eax, 0x3DCCCCCD; 0.1f
+	vmovd xmm1, eax
 
-	; TODO implement simulation speed
+	subss xmm0, xmm1
 
+	ucomiss xmm0, xmm1; check if number is smaller than 0.1
+	jb .end_input_handling
+	movss [simulation_speed], xmm0
+
+	jmp .end_input_handling
+
+	.check_q:
+	cmp al, 'q'
+	jne .end_input_handling	
+	jmp .function_exit
+	ret; exit if q pressed
+
+	.end_input_handling:
+	mov r13b, 1
 	.no_input:
 
-		
+	mov al, [simulation_running]
+	test al, al
+	jz .dont_step
+	call step_simulation
+	mov r13b, 1
+	.dont_step:
+	
+	test r13b, r13b
+	jz .main_loop
+	call print_game_ui
 	jmp .main_loop
 
+.function_exit:
+	pop r14
+	pop r13
+	pop r12
+	ret
+
+	
+global disable_canonical_mode_and_echo
+disable_canonical_mode_and_echo:
+	
+	mov rax, SYS_IOCTL
+	mov rdi, STDIN
+	mov rsi, TCGETS
+	lea rdx, [termios]
+	syscall 
+
+	; save original termios struct
+	%ifdef AVX2
+		%ifdef AVX512
+			vmovdqa64 zmm0, [termios]
+		%else
+			vmovdqa ymm0, [termios]
+			vmovdqa ymm1, [termios+32]						
+		%endif 
+	%else
+		vmovdqa xmm0, [termios]
+		vmovdqa xmm1, [termios+16]
+		vmovdqa xmm2, [termios+32]
+		vmovdqa xmm3, [termios+64]
+	%endif
+
+	
+	mov eax, [termios+12]; get c_lflag
+	and eax, NOT_ECHO; disable ECHO
+	and eax, NOT_ICANON; disable ICANON
+	mov [termios+12], eax
+
+	mov rax, SYS_IOCTL
+	mov rdi, STDIN
+	mov rsi, TCSETS
+	lea rdx, [termios]
+	syscall 
+
+
+	; load original termios struct
+	%ifdef AVX2
+		%ifdef AVX512
+			vmovdqa64 [termios], zmm0
+		%else
+			vmovdqa [termios], ymm0
+			vmovdqa [termios+32], ymm1						
+		%endif 
+	%else
+		vmovdqa [termios], xmm0
+		vmovdqa [termios+16], xmm1
+		vmovdqa [termios+32], xmm2
+		vmovdqa [termios+64], xmm3
+	%endif
+
+
+	ret
+
+global reset_terminal
+reset_terminal:
+	mov rax, SYS_IOCTL
+	mov rdi, STDIN
+	mov rsi, TCSETS
+	lea rdx, [termios]
+	syscall 
 	ret
 	
+print_cursor:
+	push r12
+	push r13
+	lea r12, [multipurpuse_buf]
+	mov rdi, r12
+	mov word [rdi], 0x5B1B; will store ESC_CHAR, '[' they have to be in reverse order here due to little endian
+	add rdi, 2
+	push rdi
+	xor rsi, rsi
+	mov si, [cursor_rows]
+	call unsigned_int_to_ascii
+	pop rdi
+	add rdi, rax; add lenght of string to pointer 
+	mov byte [rdi], ';'
+	inc rdi
+	push rdi
+	mov si, [cursor_cols]
+	call unsigned_int_to_ascii
+	pop rdi
+	add rdi, rax
+	mov byte [rdi], 'H'
+	inc rdi
+	mov byte [rdi], 0; null terminate
+
+	mov rdi, r12
+	call print_str
+
+	; write there real cursor
+	xor r13, r13
+	mov rdi, r12
+	lea rsi, [cursor_color]
+	call string_copy
+	mov rdi, r12
+	add r13, rax
+	add rdi, r13
+
+	xor rax, rax
+
+	mov ax, [cursor_rows]
+	dec ax
+	mul word [term_cols]; get index of character
+	add ax, [cursor_cols]
+	dec ax
+
+	xor rsi, rsi
+
+	add rax, [gameboard_ptr]
+	mov sil, [rax]; now we got the character in si
+	mov byte [rdi], sil
+	inc rdi
+	inc r13
+
+	lea rsi, [reset]
+	call string_copy
+	mov rdi, r12
+	add r13, rax
+	add rdi, r13
+
+	lea rsi, [cursor_left]
+	call string_copy
+	mov rdi, r12
+	add r13, rax
+	add rdi, r13
+	
+	inc rdi
+	mov byte [rdi], 0; null terminate
+
+	mov rdi, r12
+	call print_str
+	
+	pop r13
+	pop r12
+	ret
 	

src/main.asm

@@ -4,7 +4,7 @@
 
 section .bss
 	global multipurpuse_buf
-	multipurpuse_buf: RESB 8
+	multipurpuse_buf: RESB 16
 	
 	global term_rows
 	term_rows: RESW 1
@@ -19,11 +19,18 @@ section .bss
 
 	extern cursor_rows
 	extern cursor_cols
-
+	
+	global running_in_tty
+	running_in_tty: RESB 1
 	
 section .rodata
 	extern resetLen
 
+	hide_cursor: db ESC_CHAR, "[?25l", 0
+	show_cursor: db ESC_CHAR, "[?25h", 0
+
+	help_text: db "asm-game-of-life [args]",0xA,"-h	display this help menu",0xA,"Controls:",0xA,"use arrow keys to move around",0xA,"ENTER to invert cell",0xA,"p     to START/STOP simulation",0xA,"k     to increase simulation speed",0xA,"j     to decrese simulation speed",0xA, 0
+
 section .text
 extern print_str
 extern unsigned_int_to_ascii
@@ -33,6 +40,11 @@ extern alloc
 extern init_gameboard
 extern print_game_ui
 
+extern handle_user_input
+
+extern disable_canonical_mode_and_echo
+extern reset_terminal
+
 global _start
 _start:
 	; get terminal dimensions
@@ -50,11 +62,50 @@ _start:
 
 	; handle args
 	pop rcx; get argc (number of arguments)
+	pop rax; get rid of program name arugument
 	cmp rcx, 1 
 	jle .no_arguments_provided
-	; TODO hanndle arguments
+	dec rcx
+	.handle_arg:
+	pop rax
+	mov word di, [rax]
+
+	cmp di, 0x682D; check if -h was passed
+	jne .next_arg
+	lea rdi, [help_text]
+	call print_str
+	jmp .exit_program
+
+	.next_arg:
+	dec rcx
+	test rcx, rcx
+	jnz .handle_arg
+	
 	.no_arguments_provided:
 
+	
+	pop rax; get rid of null termination of argv 
+	; handle enviroment vars
+
+	.handle_env:
+	pop rax
+	test rax, rax; test if we reached end of envs
+	jz .no_envs
+
+	mov dword edi, [rax]
+	cmp edi, 0x4D524554; check for "TERM" inverted becose endiannes
+	jne .handle_env
+	mov qword rdi, [rax+5]; remove the TERM= part this should never segfault since there sould allwas be other data behind enviroment vars and i dont mind garbage 
+	mov rsi, 0xffffffffff
+	and rdi, rsi
+	mov rsi, 0x78756e696c
+	cmp rdi, rsi; check for "linux"
+	jne .no_envs
+	mov byte [running_in_tty], 1
+
+	.no_envs:
+
+
 	call init_alloc
 
 	xor rax, rax
@@ -80,12 +131,23 @@ _start:
 	mov rdx, O_NONBLOCK
 	syscall
 
+	lea rdi, [hide_cursor]
+	call print_str
+
 	call print_game_ui	
 
+	call disable_canonical_mode_and_echo
 
+	call print_game_ui
 
+	call handle_user_input
 
+	call reset_terminal
 
+	lea rdi, [show_cursor]
+	call print_str
+
+	.exit_program:
 	mov rax, SYS_EXIT
     	mov rdi, 0             ; return code
     	syscall

src/str.asm

@@ -24,7 +24,7 @@ print_str: ; takes pointer to string in rdi and retuns in rax
     ret
 
 global unsigned_int_to_ascii
-unsigned_int_to_ascii: ; takes pointer to array in rdi and value stored in rsi DOES NOT BOUNDS CHECK
+unsigned_int_to_ascii: ; takes pointer to array in rdi and value stored in rsi DOES NOT BOUNDS CHECK return len of string in rax
     xor r11, r11
     mov rcx, 10
     mov rax, rsi
@@ -61,7 +61,7 @@ unsigned_int_to_ascii: ; takes pointer to array in rdi and value stored in rsi D
     ret
 
 global string_copy
-string_copy:; takes pointer to destination in rdi and pointer to source in rsi
+string_copy:; takes pointer to destination in rdi and pointer to source in rsi and return lenght in rax
 	
 	xor rax, rax
 	xor rcx, rcx
@@ -154,52 +154,16 @@ memory_copy:; takes  destination in rdi, source in rsi and lenght in rdx
 
 	mov r9, rdi
 
-	cmp rdx, 16
-	jnl .write_16_or_more_bytes 
-	mov rcx, rdx 
-	jmp .write_less_than_16_bytes
-	.write_16_or_more_bytes:
-	mov rax, rdi; move destination to rax
-	and rax, 0xF; offset is stored in rax
+	mov rcx, rdx
+	shr rcx, 3; calculate how many 8 byte chunks we need
+	and rdx, 0x7; calculate the reminder
 
-	
-	test al, al; check if resault is 0
-	jz .addr_is_16_Byte_alligned
-	
+	rep movsq; move as many 8 byte chunks as posible
 
-	mov cl, 16
-	sub cl, al; now offset to first higher 16 byte alligned address is stored in r8
-	movzx rcx, cl; remove ani posible garbage
-	
+	mov rcx, rdx
 
-	.write_less_than_16_bytes:
-	sub rdx, rcx; we will write these bytes now
-	
-	rep movsb
+	rep movsb; move the rest
 
-	.addr_is_16_Byte_alligned:
-	mov r10, rdx
-	shr r10, 4; set it to how many 128bit(16Byte) chunk we need 
-	test r10, r10; check if we need to write aditional 16 bytes at all
-	jz .function_exit
-		
-	.move_16_bytes:
-	movdqa xmm8, [rsi]
-	movdqa [rdi], xmm8
-	add rdi, 16
-	add rsi, 16
-	sub rdx, 16
 
-	cmp rdx, 16; test if rdx is less than 16
-	jge .move_16_bytes
-
-	.function_exit:
-
-	test rdx, rdx; test if rdx is 0
-	jz .true_function_exit
-	movzx rcx, dl
-	jmp .write_less_than_16_bytes
-
-	.true_function_exit:
-	mov rax, r9; return pointer to memory area same as memset in libc
+	mov rax, r9; return pointer to memory area same as memcpy in libc
 	ret	

src/symbols.asm

@@ -5,14 +5,22 @@ SYS_WRITE 	equ 1
 SYS_BRK		equ 12
 SYS_FCNTL	equ 72
 SYS_POLL	equ 7
+SYS_NANOSLEEP	equ 35
 
 STDIN		equ 0
 STDOUT		equ 1
 
 TIOCGWINSZ	equ 0x5413
-POLLIN		equ 1 
+TCGETS		equ 0x5401
+TCSETS		equ 0x5402
 F_SETFL		equ 4 
 O_NONBLOCK	equ 2048
+POLLIN		equ 0x0100; compensate for litle endian
+
+NOT_ECHO	equ -9
+NOT_ICANON	equ -3
+
+EAGAIN		equ -11
 
 ASCII_ZERO 	equ 48
 ESC_CHAR 	equ 27