global	main
extern	printf
extern	inicjuj_czas
extern	drukuj_czas

size	equ	 16*1048576

section	.data
zero:	dq	0.0
align	16
one:	dq	1.0, 1.0
two:	dq	2.0, 2.0
align	32
four:	dq	4.0, 4.0, 4.0, 4.0
init:	dq	1.0, 2.0, 3.0, 4.0

str:		db	'%lf %lf %lf',10,0

section	.bss
align 32	
va:	resq	size
vb:	resq	size
vc:	resq	size

section	.text

main:
sub	rsp,8

call	fill3

call	inicjuj_czas

call	vv3

call	drukuj_czas

call	print_result3

add	rsp,8
ret

vv1:
mov r12,va
mov r13,vb
mov r14,vc
mov r15d,size
_petla:
cmp r15d,0
je _koniec
movq xmm0,[r12]
mulsd xmm0,xmm0
movq xmm1,[r13]
mulsd xmm1,xmm1
addsd xmm1,xmm0
sqrtsd xmm0,xmm1
movq [r14],xmm0
dec r15d
add r12,8
add r13,8
add r14,8
jmp _petla
_koniec:
ret

;do uzupelnienia

ret

vv2:
mov r12,va
mov r15d,size
_petla2:
cmp r15d,0
je _koniec2
movq xmm0,[r12]
mulsd xmm0,xmm0
movq xmm1,[r12+8]
mulsd xmm1,xmm1
addsd xmm1,xmm0
sqrtsd xmm0,xmm1
movq [r12+16],xmm0
dec r15d
add r12,24
jmp _petla2
_koniec2:
ret

;do uzupelnienia

ret

vv3:
mov r12,va
mov r15d,size/2
_petla3:
cmp r15d,0
je _koniec3
movapd xmm0,[r12]
mulpd xmm0,xmm0
movapd xmm1,[r12+16]
mulpd xmm1,xmm1
addpd xmm1,xmm0
sqrtpd xmm0,xmm1
movapd [r12+32],xmm0
dec r15d
add r12,48
jmp _petla3
_koniec3:
ret

;do uzupelnienia

ret


vv4:
mov	r15d,size/4
mov	rsi,va
vv_loop4:
vmovapd	ymm0,[rsi]
vmovapd	ymm1,[rsi+32]
vmulpd	ymm0,ymm0
vmulpd	ymm1,ymm1
vaddpd	ymm1,ymm0
vsqrtpd	ymm0,ymm1
add	rsi,96
vmovapd	[rsi-32],ymm0
dec	r15d
jne	vv_loop4
ret


fill1:
xor	r15d,r15d
movq	xmm0,[zero]
movq	xmm1,[one]
fill_loop1:
addsd	xmm0,xmm1
movq	[va+8*r15d],xmm0
movq	[vb+8*r15d],xmm0
inc	r15d
cmp	r15d,size
jne	fill_loop1
ret

fill2:
mov	r15d,size
mov	r13,va
movq	xmm0,[zero]
movq	xmm1,[one]
fill_loop2:
addsd	xmm0,xmm1
movq	[r13],xmm0
movq	[r13+8],xmm0
add	r13,24
dec	r15d
jne	fill_loop2
ret

fill3:
mov	r15d,size/2
mov	r13,va
movapd	xmm0,[init]
movapd	xmm1,[two]
fill_loop3:
movapd	[r13],xmm0
movapd	[r13+16],xmm0
addpd	xmm0,xmm1
add	r13,48
dec	r15d
jne	fill_loop3
ret

fill4:
mov	r15d,size/4
mov	r13,va
vmovapd	ymm0,[init]
vmovapd	ymm1,[four]
fill_loop4:
vmovapd	[r13],ymm0
vmovapd	[r13+32],ymm0
vaddpd	ymm0,ymm1
add	r13,96
dec	r15d
jne	fill_loop4
ret

print_result1:
xor	r15d,r15d
sub	rsp,8
print_loop1:
mov	rdi,str
movq	xmm0,[va+8*r15d]
movq	xmm1,[vb+8*r15d]
movq	xmm2,[vc+8*r15d]
mov	eax,3
call	printf
inc	r15d
cmp	r15d,size
jne	print_loop1
add	rsp,8
ret

print_result2:
sub	rsp,8
mov	r15d,size
mov	r13,va
print_loop2:
mov	rdi,str
movq	xmm0,[r13]
movq	xmm1,[r13+8]
movq	xmm2,[r13+16]
mov	eax,3
call	printf
add	r13,24
dec	r15d
jne	print_loop2
add	rsp,8
ret

print_result3:
sub	rsp,8
mov	r15d,size/2
mov	r13,va
print_loop3:
xor	r14d,r14d
print_loop3a:
mov	rdi,str
movq	xmm0,[r13+r14]
movq	xmm1,[r13+r14+16]
movq	xmm2,[r13+r14+32]
mov	eax,3
call	printf
add	r14d,8
cmp	r14d,16
jne	print_loop3a
add	r13,48
dec	r15d
jne	print_loop3
add	rsp,8
ret

print_result4:
sub	rsp,8
mov	r15d,size/4
mov	r13,va
print_loop4:
xor	r14d,r14d
print_loop4a:
mov	rdi,str
movq	xmm0,[r13+r14]
movq	xmm1,[r13+r14+32]
movq	xmm2,[r13+r14+64]
mov	eax,3
call	printf
add	r14d,8
cmp	r14d,32
jne	print_loop4a
add	r13,96
dec	r15d
jne	print_loop4
add	rsp,8
ret