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
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
|
#include "fp_amd64.h"
#define _fqAdd(c,a,b) \
_fpAdd( 0+c, 0+a, 0+b) \
_fpAdd(16+c,16+a,16+b)
#define _fqSub(c,a,b) \
_fpSub( 0+c, 0+a, 0+b) \
_fpSub(16+c,16+a,16+b)
#define _fqMulBmi2(c, a, b) \
MOVL $0, R15 \
\ // T0 = a0 * b0, R11:R10:R9:R8 <- 0+ra:8+ra * 0+rb:8+rb
MOVQ 0+b, DX \
MULXQ 0+a, R8, R9 \
MULXQ 8+a, R10, AX \
ADDQ R10, R9 \
MOVQ 8+b, DX \
MULXQ 8+a, R10, R11 \
ADCQ AX, R10 \
MULXQ 0+a, DX, AX \
ADCQ $0, R11 \
ADDQ DX, R9 \
\
\ // T1 = a1 * b1, R15:R14:R13:R12 <- 16+ra:24+ra * 16+rb:24+rb
MOVQ 16+b, DX \
MULXQ 16+a, R12, R13 \
ADCQ AX, R10 \
MULXQ 24+a, R14, AX \
ADCQ $0, R11 \
MOVQ 24+b, DX \
ADDQ R14, R13 \
MULXQ 24+a, R14, R15 \
ADCQ AX, R14 \
ADCQ $0, R15 \
MULXQ 16+a, DX, AX \
ADDQ DX, R13 \
ADCQ AX, R14 \
ADCQ $0, R15 \
\
\ // c0 = T0 - T1 = a0*b0 - a1*b1
SUBQ R12, R8 \
SBBQ R13, R9 \
SBBQ R14, R10 \
SBBQ R15, R11 \
\
SHLQ $1, R10, R11 \
SHLQ $1, R9, R10 \
MOVQ 16+b, DX \
BTRQ $63, R9 \
\
\ // T0 = a0 * b1, R15:R14:R13:R12 <- 0+ra:8+ra * 16+rb:24+rb
MULXQ 0+a, R12, R13 \
BTRQ $63, R11 \
SBBQ $0, R10 \
SBBQ $0, R11 \
MULXQ 8+a, R14, AX \
ADDQ R14, R13 \
MOVQ 24+b, DX \
MULXQ 8+a, R14, R15 \
ADCQ AX, R14 \
ADCQ $0, R15 \
MULXQ 0+a, DX, AX \
ADDQ DX, R13 \
ADCQ AX, R14 \
ADCQ $0, R15 \
\
\ // Reducing and storing c0
ADDQ R8, R10 \
ADCQ R9, R11 \
BTRQ $63, R11 \
ADCQ $0, R10 \
ADCQ $0, R11 \
\
\ // T1 = a1 * b0, R12:R11:R10:R9 <- 16+ra:24+ra * 0+rb:8+rb
MOVQ 0+b, DX \
MULXQ 16+a, R8, R9 \
MOVQ R10, 0+c \
MULXQ 24+a, R10, AX \
ADDQ R10, R9 \
MOVQ 8+b, DX \
MOVQ R11, 8+c \
MULXQ 24+a, R10, R11 \
ADCQ AX, R10 \
ADCQ $0, R11 \
MULXQ 16+a, DX, AX \
ADDQ DX, R9 \
ADCQ AX, R10 \
ADCQ $0, R11 \
\
\ // c1 = T0 + T1 = a0*b1 + a1*b0
ADDQ R12, R8 \
ADCQ R13, R9 \
ADCQ R14, R10 \
ADCQ R15, R11 \
\
\ // Reducing and storing c1
SHLQ $1, R10, R11 \
SHLQ $1, R9, R10 \
BTRQ $63, R9 \
BTRQ $63, R11 \
ADCQ R10, R8 \
ADCQ R11, R9 \
BTRQ $63, R9 \
ADCQ $0, R8 \
ADCQ $0, R9 \
MOVQ R8, 16+c \
MOVQ R9, 24+c
#define _fqMulLeg(c, a, b) \
_fpMulLeg(R10, R9, R8, 0+a, 0+b) \
_fpMulLeg(R13,R12,R11,16+a,16+b) \
MOVQ $0,R14 \
SUBQ R11, R8 \
SBBQ R12, R9 \
SBBQ R13,R10 \
SBBQ $0,R14 \
SHLQ $1,R10 \
BTRQ $63, R9 \
ADCQ R10, R8 \
ADCQ R14, R9 \
MOVQ R8, R14 \
MOVQ R9, R15 \
_fpMulLeg(R10, R9, R8, 0+a,16+b) \
_fpMulLeg(R13,R12,R11,16+a, 0+b) \
ADDQ R11, R8 \
ADCQ R12, R9 \
ADCQ R13,R10 \
SHLQ $1,R10 \
BTRQ $63, R9 \
ADCQ R10, R8 \
ADCQ $0, R9 \
MOVQ R14, 0+c \
MOVQ R15, 8+c \
MOVQ R8,16+c \
MOVQ R9,24+c
#define _fqSqrBmi2(c,a) \
\ // t0 = R9:R8 = a0 + a1, R14:CX = a1
MOVQ 0+a, R10 \
MOVQ 16+a, R14 \
SUBQ R14, R10 \
MOVQ 8+a, R11 \
MOVQ 24+a, CX \
SBBQ CX, R11 \
\
BTRQ $63, R11 \
SBBQ $0, R10 \
\
\ // t1 = R11:R10 = a0 - a1
MOVQ R10, DX \
MOVQ 0+a, R8 \
ADDQ R14, R8 \
MOVQ 8+a, R9 \
ADCQ CX, R9 \
\
\ // c0 = t0 * t1 = (a0 + a1)*(a0 - a1), CX:R14:R13:R12 <- R9:R8 * R11:R10
MULXQ R8, R12, R13 \
SBBQ $0, R11 \
MULXQ R9, R14, AX \
MOVQ R11, DX \
ADDQ R14, R13 \
MULXQ R9, R14, CX \
MOVQ 8+a, R9 \
ADCQ AX, R14 \
ADCQ $0, CX \
MULXQ R8, DX, AX \
MOVQ 0+a, R8 \
ADDQ DX, R13 \
ADCQ AX, R14 \
ADCQ $0, CX \
\
\ // t2 = R9:R8 = 2*a0
ADDQ R8, R8 \
ADCQ R9, R9 \
\
\ // Reducing and storing c0
SHLQ $1, R14, CX \
SHLQ $1, R13, R14 \
BTRQ $63, R13 \
BTRQ $63, CX \
ADCQ R14, R12 \
ADCQ CX, R13 \
BTRQ $63, R13 \
ADCQ $0, R12 \
ADCQ $0, R13 \
MOVQ R12, 0+c \
MOVQ R13, 8+c \
\
\ // c1 = 2a0 * a1, CX:R14:R11:R10 <- R9:R8 * 16+ra:24+ra
MOVQ 16+a, DX \
MULXQ R8, R10, R11 \
MULXQ R9, R14, AX \
ADDQ R14, R11 \
MOVQ 24+a, DX \
MULXQ R9, R14, CX \
ADCQ AX, R14 \
ADCQ $0, CX \
MULXQ R8, DX, AX \
ADDQ DX, R11 \
ADCQ AX, R14 \
ADCQ $0, CX \
\
\ // Reduce and store c1
SHLQ $1, R14, CX \
SHLQ $1, R11, R14 \
BTRQ $63, R11 \
BTRQ $63, CX \
ADCQ R14, R10 \
ADCQ CX, R11 \
BTRQ $63, R11 \
ADCQ $0, R10 \
ADCQ $0, R11 \
MOVQ R10, 16+c \
MOVQ R11, 24+c
#define _fqSqrLeg(c,a) _fqMulLeg(c,a,a)
|
