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
|
/*
* Copyright 2017-2021 The OpenSSL Project Authors. All Rights Reserved.
* Copyright 2014 Cryptography Research, Inc.
*
* Licensed under the OpenSSL license (the "License"). You may not use
* this file except in compliance with the License. You can obtain a copy
* in the file LICENSE in the source distribution or at
* https://www.openssl.org/source/license.html
*
* Originally written by Mike Hamburg
*/
#include "openssl/macros.h"
#include "internal/numbers.h"
#ifndef UINT128_MAX
/* No support for 128 bit ints, so do nothing here */
NON_EMPTY_TRANSLATION_UNIT
#else
# include "../field.h"
void gf_mul(gf_s * RESTRICT cs, const gf as, const gf bs)
{
const uint64_t *a = as->limb, *b = bs->limb;
uint64_t *c = cs->limb;
uint128_t accum0 = 0, accum1 = 0, accum2;
uint64_t mask = (1ULL << 56) - 1;
uint64_t aa[4], bb[4], bbb[4];
unsigned int i, j;
for (i = 0; i < 4; i++) {
aa[i] = a[i] + a[i + 4];
bb[i] = b[i] + b[i + 4];
bbb[i] = bb[i] + b[i + 4];
}
for (i = 0; i < 4; i++) {
accum2 = 0;
for (j = 0; j <= i; j++) {
accum2 += widemul(a[j], b[i - j]);
accum1 += widemul(aa[j], bb[i - j]);
accum0 += widemul(a[j + 4], b[i - j + 4]);
}
for (; j < 4; j++) {
accum2 += widemul(a[j], b[i - j + 8]);
accum1 += widemul(aa[j], bbb[i - j + 4]);
accum0 += widemul(a[j + 4], bb[i - j + 4]);
}
accum1 -= accum2;
accum0 += accum2;
c[i] = ((uint64_t)(accum0)) & mask;
c[i + 4] = ((uint64_t)(accum1)) & mask;
accum0 >>= 56;
accum1 >>= 56;
}
accum0 += accum1;
accum0 += c[4];
accum1 += c[0];
c[4] = ((uint64_t)(accum0)) & mask;
c[0] = ((uint64_t)(accum1)) & mask;
accum0 >>= 56;
accum1 >>= 56;
c[5] += ((uint64_t)(accum0));
c[1] += ((uint64_t)(accum1));
}
void gf_mulw_unsigned(gf_s * RESTRICT cs, const gf as, uint32_t b)
{
const uint64_t *a = as->limb;
uint64_t *c = cs->limb;
uint128_t accum0 = 0, accum4 = 0;
uint64_t mask = (1ULL << 56) - 1;
int i;
for (i = 0; i < 4; i++) {
accum0 += widemul(b, a[i]);
accum4 += widemul(b, a[i + 4]);
c[i] = accum0 & mask;
accum0 >>= 56;
c[i + 4] = accum4 & mask;
accum4 >>= 56;
}
accum0 += accum4 + c[4];
c[4] = accum0 & mask;
c[5] += accum0 >> 56;
accum4 += c[0];
c[0] = accum4 & mask;
c[1] += accum4 >> 56;
}
void gf_sqr(gf_s * RESTRICT cs, const gf as)
{
const uint64_t *a = as->limb;
uint64_t *c = cs->limb;
uint128_t accum0 = 0, accum1 = 0, accum2;
uint64_t mask = (1ULL << 56) - 1;
uint64_t aa[4];
unsigned int i;
/* For some reason clang doesn't vectorize this without prompting? */
for (i = 0; i < 4; i++)
aa[i] = a[i] + a[i + 4];
accum2 = widemul(a[0], a[3]);
accum0 = widemul(aa[0], aa[3]);
accum1 = widemul(a[4], a[7]);
accum2 += widemul(a[1], a[2]);
accum0 += widemul(aa[1], aa[2]);
accum1 += widemul(a[5], a[6]);
accum0 -= accum2;
accum1 += accum2;
c[3] = ((uint64_t)(accum1)) << 1 & mask;
c[7] = ((uint64_t)(accum0)) << 1 & mask;
accum0 >>= 55;
accum1 >>= 55;
accum0 += widemul(2 * aa[1], aa[3]);
accum1 += widemul(2 * a[5], a[7]);
accum0 += widemul(aa[2], aa[2]);
accum1 += accum0;
accum0 -= widemul(2 * a[1], a[3]);
accum1 += widemul(a[6], a[6]);
accum2 = widemul(a[0], a[0]);
accum1 -= accum2;
accum0 += accum2;
accum0 -= widemul(a[2], a[2]);
accum1 += widemul(aa[0], aa[0]);
accum0 += widemul(a[4], a[4]);
c[0] = ((uint64_t)(accum0)) & mask;
c[4] = ((uint64_t)(accum1)) & mask;
accum0 >>= 56;
accum1 >>= 56;
accum2 = widemul(2 * aa[2], aa[3]);
accum0 -= widemul(2 * a[2], a[3]);
accum1 += widemul(2 * a[6], a[7]);
accum1 += accum2;
accum0 += accum2;
accum2 = widemul(2 * a[0], a[1]);
accum1 += widemul(2 * aa[0], aa[1]);
accum0 += widemul(2 * a[4], a[5]);
accum1 -= accum2;
accum0 += accum2;
c[1] = ((uint64_t)(accum0)) & mask;
c[5] = ((uint64_t)(accum1)) & mask;
accum0 >>= 56;
accum1 >>= 56;
accum2 = widemul(aa[3], aa[3]);
accum0 -= widemul(a[3], a[3]);
accum1 += widemul(a[7], a[7]);
accum1 += accum2;
accum0 += accum2;
accum2 = widemul(2 * a[0], a[2]);
accum1 += widemul(2 * aa[0], aa[2]);
accum0 += widemul(2 * a[4], a[6]);
accum2 += widemul(a[1], a[1]);
accum1 += widemul(aa[1], aa[1]);
accum0 += widemul(a[5], a[5]);
accum1 -= accum2;
accum0 += accum2;
c[2] = ((uint64_t)(accum0)) & mask;
c[6] =
|