#include "dt_inc.s"
.data
.org 0x0
arr:
	.fill 2
	.fill 3
	.fill 5
	.fill 7
	.fill 11
	.fill 13
	.fill 17
	.fill 19
	.fill 23
	.fill 29
	.fill 31
	.fill 37
	.fill 41
	.fill 43
	.fill 47
	.fill 53

str:
	.ascii "hier das ergebnis: "
.text
.org 0x0
start:
	br+ main
	br+ main
	ret
main:
	call+ u_init
	call+ u_recv_byte
	call u_send_newline

	; benchprolog
	call t_init
	call t_stop
	ldis r1, 0
	call t_valset
	call t_start
	; /benchprolog

	; max start
	ldis r1, 16
	ldil r2, arr@lo
	ldih r2, arr@hi

	call+ max
	; max end

	; benchepilog
	push r0
	call+ t_init
	call+ t_stop
	call+ t_valget
	subi r0, r0, 0xc ; offset abziehen
	pop r3
	push r0
	push r3
	; /benchepilog

	call u_init ; weil r10 ueberschrieben wird
	ldis r1, str@lo
	ldih r1, str@hi
	ldis r2, 19
	call u_send_string

	pop r1
	call+ u_send_uint
	call u_send_newline
	pop r1
	call+ u_send_uint
	call u_send_newline

	br+ main

max:
	; r1 = len
	; r2 = arr
	; r4 = arr[x]
	; r0 = x = arr[0]
	ldw r0, 0(r2)

	subi r5, r1, 1
	andx r5, 0x3

	cmpi r5, 0
	breq+ loop0

	cmpi r5, 1
	addideq r1, r1, 3-1
	subideq r2, r2, 12
	breq+ loop1

	cmpi r5, 2
	addideq r1, r1, 2-1
	subideq r2, r2, 8
	breq+ loop2

	cmpi r5, 3
	; addideq r1, r1, 1-1
	subideq r2, r2, 4
	breq+ loop3

loop0:
	addi r1, r1, 4-1
maxloop:
	subi r1, r1, 4
	retzs-

	ldw r4, 4(r2)
	cmp r0, r4
	movlt r0, r4

loop1:
	ldw r4, 8(r2)
	cmp r0, r4
	movlt r0, r4

loop2:
	ldw r4, 12(r2)
	cmp r0, r4
	movlt r0, r4

loop3:
	ldw r4, 16(r2)
	cmp r0, r4
	movlt r0, r4

	addi r2, r2, 4*4
	br+ maxloop