xref: /plan9-contrib/sys/src/libmp/amd64/mpvecadd.s (revision 272efad760864ee41cfe633b56aea9b4f5cf3ae7)
1/*
2 *	mpvecadd(mpdigit *a, int alen, mpdigit *b, int blen, mpdigit *sum)
3 *
4 *		sum[0:alen] = a[0:alen-1] + b[0:blen-1]
5 *
6 *	prereq: alen >= blen, sum has room for alen+1 digits
7 */
8TEXT	mpvecadd(SB),$0
9
10	MOVL	alen+8(FP),DX
11	MOVL	blen+24(FP),CX
12/*	MOVL	a+0(FP),SI */
13	MOVQ	RARG, SI
14	MOVQ	b+16(FP),BX
15	SUBL	CX,DX
16	MOVQ	sum+32(FP),DI
17	XORL	BP,BP			/* this also sets carry to 0 */
18
19	/* skip addition if b is zero */
20	TESTL	CX,CX
21	JZ	_add1
22
23	/* sum[0:blen-1],carry = a[0:blen-1] + b[0:blen-1] */
24_addloop1:
25	MOVL	(SI)(BP*4), AX
26	ADCL	(BX)(BP*4), AX
27	MOVL	AX,(DI)(BP*4)
28	INCL	BP
29	LOOP	_addloop1
30
31_add1:
32	/* jump if alen > blen */
33	INCL	DX
34	MOVL	DX,CX
35	LOOP	_addloop2
36
37	/* sum[alen] = carry */
38_addend:
39	JC	_addcarry
40	MOVL	$0,(DI)(BP*4)
41	RET
42_addcarry:
43	MOVL	$1,(DI)(BP*4)
44	RET
45
46	/* sum[blen:alen-1],carry = a[blen:alen-1] + 0 */
47_addloop2:
48	MOVL	(SI)(BP*4),AX
49	ADCL	$0,AX
50	MOVL	AX,(DI)(BP*4)
51	INCL	BP
52	LOOP	_addloop2
53	JMP	_addend
54
55