-
Notifications
You must be signed in to change notification settings - Fork 42
/
mul_bmi2_amd64.h
112 lines (111 loc) · 2.01 KB
/
mul_bmi2_amd64.h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
#define mulBMI2(a0,a1,a2,a3, rb) \
MOVQ a0, DX \
MOVQ $0, R13 \
MULXQ 0+rb, R8, R9 \
MULXQ 8+rb, AX, R10 \
ADDQ AX, R9 \
MULXQ 16+rb, AX, R11 \
ADCQ AX, R10 \
MULXQ 24+rb, AX, R12 \
ADCQ AX, R11 \
ADCQ $0, R12 \
ADCQ $0, R13 \
\
MOVQ a1, DX \
MOVQ $0, R14 \
MULXQ 0+rb, AX, BX \
ADDQ AX, R9 \
ADCQ BX, R10 \
MULXQ 16+rb, AX, BX \
ADCQ AX, R11 \
ADCQ BX, R12 \
ADCQ $0, R13 \
MULXQ 8+rb, AX, BX \
ADDQ AX, R10 \
ADCQ BX, R11 \
MULXQ 24+rb, AX, BX \
ADCQ AX, R12 \
ADCQ BX, R13 \
ADCQ $0, R14 \
\
MOVQ a2, DX \
MOVQ $0, CX \
MULXQ 0+rb, AX, BX \
ADDQ AX, R10 \
ADCQ BX, R11 \
MULXQ 16+rb, AX, BX \
ADCQ AX, R12 \
ADCQ BX, R13 \
ADCQ $0, R14 \
MULXQ 8+rb, AX, BX \
ADDQ AX, R11 \
ADCQ BX, R12 \
MULXQ 24+rb, AX, BX \
ADCQ AX, R13 \
ADCQ BX, R14 \
ADCQ $0, CX \
\
MOVQ a3, DX \
MULXQ 0+rb, AX, BX \
ADDQ AX, R11 \
ADCQ BX, R12 \
MULXQ 16+rb, AX, BX \
ADCQ AX, R13 \
ADCQ BX, R14 \
ADCQ $0, CX \
MULXQ 8+rb, AX, BX \
ADDQ AX, R12 \
ADCQ BX, R13 \
MULXQ 24+rb, AX, BX \
ADCQ AX, R14 \
ADCQ BX, CX
#define gfpReduceBMI2() \
\ // m = (T * N') mod R, store m in R8:R9:R10:R11
MOVQ ·np+0(SB), DX \
MULXQ 0(SP), R8, R9 \
MULXQ 8(SP), AX, R10 \
ADDQ AX, R9 \
MULXQ 16(SP), AX, R11 \
ADCQ AX, R10 \
MULXQ 24(SP), AX, BX \
ADCQ AX, R11 \
\
MOVQ ·np+8(SB), DX \
MULXQ 0(SP), AX, BX \
ADDQ AX, R9 \
ADCQ BX, R10 \
MULXQ 16(SP), AX, BX \
ADCQ AX, R11 \
MULXQ 8(SP), AX, BX \
ADDQ AX, R10 \
ADCQ BX, R11 \
\
MOVQ ·np+16(SB), DX \
MULXQ 0(SP), AX, BX \
ADDQ AX, R10 \
ADCQ BX, R11 \
MULXQ 8(SP), AX, BX \
ADDQ AX, R11 \
\
MOVQ ·np+24(SB), DX \
MULXQ 0(SP), AX, BX \
ADDQ AX, R11 \
\
storeBlock(R8,R9,R10,R11, 64(SP)) \
\
\ // m * N
mulBMI2(·p2+0(SB),·p2+8(SB),·p2+16(SB),·p2+24(SB), 64(SP)) \
\
\ // Add the 512-bit intermediate to m*N
MOVQ $0, AX \
ADDQ 0(SP), R8 \
ADCQ 8(SP), R9 \
ADCQ 16(SP), R10 \
ADCQ 24(SP), R11 \
ADCQ 32(SP), R12 \
ADCQ 40(SP), R13 \
ADCQ 48(SP), R14 \
ADCQ 56(SP), CX \
ADCQ $0, AX \
\
gfpCarry(R12,R13,R14,CX,AX, R8,R9,R10,R11,BX)