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
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
|
// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
//
// Permission to use, copy, modify, and/or distribute this software for any
// purpose with or without fee is hereby granted, provided that the above
// copyright notice and this permission notice appear in all copies.
//
// THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
// WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
// MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
// ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
// WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
// ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
// OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
// ----------------------------------------------------------------------------
// Multiply z := x * y
// Inputs x[8], y[8]; output z[16]
//
// extern void bignum_mul_8_16_alt
// (uint64_t z[static 16], uint64_t x[static 8], uint64_t y[static 8]);
//
// Standard x86-64 ABI: RDI = z, RSI = x, RDX = y
// Microsoft x64 ABI: RCX = z, RDX = x, R8 = y
// ----------------------------------------------------------------------------
#include "s2n_bignum_internal.h"
.intel_syntax noprefix
S2N_BN_SYM_VISIBILITY_DIRECTIVE(bignum_mul_8_16_alt)
S2N_BN_SYM_PRIVACY_DIRECTIVE(bignum_mul_8_16_alt)
.text
// These are actually right
#define z rdi
#define x rsi
// This is moved from rdx to free it for muls
#define y rcx
// Other variables used as a rotating 3-word window to add terms to
#define t0 r8
#define t1 r9
#define t2 r10
// Macro for the key "multiply and add to (c,h,l)" step
#define combadd(c,h,l,numa,numb) \
mov rax, numa; \
mul QWORD PTR numb; \
add l, rax; \
adc h, rdx; \
adc c, 0
// A minutely shorter form for when c = 0 initially
#define combadz(c,h,l,numa,numb) \
mov rax, numa; \
mul QWORD PTR numb; \
add l, rax; \
adc h, rdx; \
adc c, c
// A short form where we don't expect a top carry
#define combads(h,l,numa,numb) \
mov rax, numa; \
mul QWORD PTR numb; \
add l, rax; \
adc h, rdx
S2N_BN_SYMBOL(bignum_mul_8_16_alt):
_CET_ENDBR
#if WINDOWS_ABI
push rdi
push rsi
mov rdi, rcx
mov rsi, rdx
mov rdx, r8
#endif
// Copy y into a safe register to start with
mov y, rdx
// Result term 0
mov rax, [x]
mul QWORD PTR [y]
mov [z], rax
mov t0, rdx
xor t1, t1
// Result term 1
xor t2, t2
combads(t1,t0,[x],[y+8])
combadz(t2,t1,t0,[x+8],[y])
mov [z+8], t0
// Result term 2
xor t0, t0
combadz(t0,t2,t1,[x],[y+16])
combadd(t0,t2,t1,[x+8],[y+8])
combadd(t0,t2,t1,[x+16],[y])
mov [z+16], t1
// Result term 3
xor t1, t1
combadz(t1,t0,t2,[x],[y+24])
combadd(t1,t0,t2,[x+8],[y+16])
combadd(t1,t0,t2,[x+16],[y+8])
combadd(t1,t0,t2,[x+24],[y])
mov [z+24], t2
// Result term 4
xor t2, t2
combadz(t2,t1,t0,[x],[y+32])
combadd(t2,t1,t0,[x+8],[y+24])
combadd(t2,t1,t0,[x+16],[y+16])
combadd(t2,t1,t0,[x+24],[y+8])
combadd(t2,t1,t0,[x+32],[y])
mov [z+32], t0
// Result term 5
xor t0, t0
combadz(t0,t2,t1,[x],[y+40])
combadd(t0,t2,t1,[x+8],[y+32])
combadd(t0,t2,t1,[x+16],[y+24])
combadd(t0,t2,t1,[x+24],[y+16])
combadd(t0,t2,t1,[x+32],[y+8])
combadd(t0,t2,t1,[x+40],[y])
mov [z+40], t1
// Result term 6
xor t1, t1
combadz(t1,t0,t2,[x],[y+48])
combadd(t1,t0,t2,[x+8],[y+40])
combadd(t1,t0,t2,[x+16],[y+32])
combadd(t1,t0,t2,[x+24],[y+24])
combadd(t1,t0,t2,[x+32],[y+16])
combadd(t1,t0,t2,[x+40],[y+8])
combadd(t1,t0,t2,[x+48],[y])
mov [z+48], t2
// Result term 7
xor t2, t2
combadz(t2,t1,t0,[x],[y+56])
combadd(t2,t1,t0,[x+8],[y+48])
combadd(t2,t1,t0,[x+16],[y+40])
combadd(t2,t1,t0,[x+24],[y+32])
combadd(t2,t1,t0,[x+32],[y+24])
combadd(t2,t1,t0,[x+40],[y+16])
combadd(t2,t1,t0,[x+48],[y+8])
combadd(t2,t1,t0,[x+56],[y])
mov [z+56], t0
// Result term 8
xor t0, t0
combadz(t0,t2,t1,[x+8],[y+56])
combadd(t0,t2,t1,[x+16],[y+48])
combadd(t0,t2,t1,[x+24],[y+40])
combadd(t0,t2,t1,[x+32],[y+32])
combadd(t0,t2,t1,[x+40],[y+24])
combadd(t0,t2,t1,[x+48],[y+16])
combadd(t0,t2,t1,[x+56],[y+8])
mov [z+64], t1
// Result term 9
xor t1, t1
combadz(t1,t0,t2,[x+16],[y+56])
combadd(t1,t0,t2,[x+24],[y+48])
combadd(t1,t0,t2,[x+32],[y+40])
combadd(t1,t0,t2,[x+40],[y+32])
combadd(t1,t0,t2,[x+48],[y+24])
combadd(t1,t0,t2,[x+56],[y+16])
mov [z+72], t2
// Result term 10
xor t2, t2
combadz(t2,t1,t0,[x+24],[y+56])
combadd(t2,t1,t0,[x+32],[y+48])
combadd(t2,t1,t0,[x+40],[y+40])
combadd(t2,t1,t0,[x+48],[y+32])
combadd(t2,t1,t0,[x+56],[y+24])
mov [z+80], t0
// Result term 11
xor t0, t0
combadz(t0,t2,t1,[x+32],[y+56])
combadd(t0,t2,t1,[x+40],[y+48])
combadd(t0,t2,t1,[x+48],[y+40])
combadd(t0,t2,t1,[x+56],[y+32])
mov [z+88], t1
// Result term 12
xor t1, t1
combadz(t1,t0,t2,[x+40],[y+56])
combadd(t1,t0,t2,[x+48],[y+48])
combadd(t1,t0,t2,[x+56],[y+40])
mov [z+96], t2
// Result term 13
xor t2, t2
combadz(t2,t1,t0,[x+48],[y+56])
combadd(t2,t1,t0,[x+56],[y+48])
mov [z+104], t0
// Result term 14
combads(t2,t1,[x+56],[y+56])
mov [z+112], t1
// Result term 11
mov [z+120], t2
// Return
#if WINDOWS_ABI
pop rsi
pop rdi
#endif
ret
#if defined(__linux__) && defined(__ELF__)
.section .note.GNU-stack,"",%progbits
#endif
|
