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
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
|
#include <stdint.h>
#include <stdlib.h>
#include <limits.h>
#include "bstea.h"
/* a key schedule constant - 32/golden-ratio */
static const uint32_t delta = 0x9e3779b9;
/* v points to the wordsize-way vectorized plaintext,
* k to the vectorized key */
/* input quantities are disposed in the following way:
v0 <- v[0..31] k0 <- k[0..31] k2 <- k[64..95]
v1 <- v[32..63] k1 <- k[32..63] k3 <- k[96..127]
*/
void encrypt(parallel_blocks_t v, const parallel_keys_t k, unsigned int r)
{
/* Stride 32 between consecutive words in input quantities */
# define offset_v0 0
# define offset_v1 32
# define offset_k0 0
# define offset_k1 32
# define offset_k2 64
# define offset_k3 96
vector_width_t carry;
vector_width_t axorb;
vector_width_t aandb;
vector_width_t ai;
vector_width_t bi;
vector_width_t borrow;
vector_width_t notaandb;
vector_width_t v1_lshift_4[32];
vector_width_t v1_plus_sum[32]; /* term two */
vector_width_t v1_rshift_5[32];
vector_width_t v1_lshift_4_plus_k0[32]; /* term one */
vector_width_t v1_rshift_5_plus_k1[32]; /* term three */
vector_width_t v0_lshift_4[32];
vector_width_t v0_plus_sum[32]; /* term two */
vector_width_t v0_rshift_5[32];
vector_width_t v0_lshift_4_plus_k2[32]; /* term one */
vector_width_t v0_rshift_5_plus_k3[32]; /* term three */
vector_width_t shift;
int i;
/* setup */
uint32_t sum = 0;
for (i = 0; i < 32; ++i)
v1_lshift_4[i] = v1_plus_sum[i] = v1_rshift_5[i] = \
v1_lshift_4_plus_k0[i] = v1_rshift_5_plus_k1[i] = \
v0_lshift_4[i] = v0_plus_sum[i] = v0_rshift_5[i] = \
v0_lshift_4_plus_k2[i] = v0_rshift_5_plus_k3[i] = 0;
while (r > 0) {
sum += delta;
/* lshift v1 by 4 */
shift = 4;
for (i = 31; i >= 0; i--)
v1_lshift_4[i] = (i >= shift) ? v[offset_v1 + i - shift] : 0;
/* add k0 to v1_lshift_4 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v1_lshift_4[i];
bi = k[offset_k0 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_lshift_4_plus_k0[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* add delta sum to v1 */
carry = 0;
for (i = 0;i < 32;++i) {
/* VECTOR_AT_ONE where the ith bit of the sum is set */
/*
* Each iteration follows the first 32 elements
* in the expansion of multiples of 32/golden-ratio,
* or 32/(1+sqrt(5)/2
*/
ai = TERNARY((sum & (1<<i)), VECTOR_AT_ONE, VECTOR_AT_ZERO);
bi = v[offset_v1 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_plus_sum[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* rshift v1 by 5 */
shift = 5;
for (i = 0; i < 32; ++i)
v1_rshift_5[i] = (i < (32 - shift)) ? v[offset_v1 + i + shift] : 0;
/* add k1 to v1_rshift_5 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v1_rshift_5[i];
bi = k[offset_k1 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_rshift_5_plus_k1[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* xor the three terms and increment v0 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v1_lshift_4_plus_k0[i] ^ v1_plus_sum[i] ^ v1_rshift_5_plus_k1[i];
bi = v[offset_v0 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v[offset_v0 + i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* lshift v0 by 4 */
shift = 4;
for (i = 31; i >= 0; i--)
v0_lshift_4[i] = (i >= shift) ? v[offset_v0 + i - shift] : 0;
/* add k2 and v0_lshift_4 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v0_lshift_4[i];
bi = k[offset_k2 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_lshift_4_plus_k2[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* add delta sum to v0 */
carry = 0;
for (i = 0;i < 32;++i) {
/* VECTOR_AT_ONE where the ith bit of the sum is set */
ai = TERNARY((sum & (1<<i)), VECTOR_AT_ONE, VECTOR_AT_ZERO);
bi = v[offset_v0 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_plus_sum[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* rshift v0 by 5 */
shift = 5;
for (i = 0; i < 32; ++i)
v0_rshift_5[i] = (i < (32 - shift)) ? v[offset_v0 + i + shift] : 0;
/* add k3 to v0_rshift_5 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v0_rshift_5[i];
bi = k[offset_k3 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_rshift_5_plus_k3[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* xor the three terms and increment v1 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v0_lshift_4_plus_k2[i] ^ v0_plus_sum[i] ^ v0_rshift_5_plus_k3[i];
bi = v[offset_v1 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v[offset_v1 + i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
--r;
}
}
/* v points to the wordsize-way vectorized ciphertext,
* k to the vectorized key */
/* input quantities are disposed in the following way:
v0 <- v[0..31] k0 <- k[0..31] k2 <- k[64..95]
v1 <- v[32..63] k1 <- k[32..63] k3 <- k[96..127]
*/
void decrypt(parallel_blocks_t v, const parallel_keys_t k, unsigned int r)
{
# define offset_v0 0
# define offset_v1 32
# define offset_k0 0
# define offset_k1 32
# define offset_k2 64
# define offset_k3 96
vector_width_t carry;
vector_width_t axorb;
vector_width_t aandb;
vector_width_t ai;
vector_width_t bi;
vector_width_t borrow;
vector_width_t notaandb;
vector_width_t v1_lshift_4[32];
vector_width_t v1_plus_sum[32]; /* term two */
vector_width_t v1_rshift_5[32];
vector_width_t v1_lshift_4_plus_k0[32]; /* term one */
vector_width_t v1_rshift_5_plus_k1[32]; /* term three */
vector_width_t v0_lshift_4[32];
vector_width_t v0_plus_sum[32]; /* term two */
vector_width_t v0_rshift_5[32];
vector_width_t v0_lshift_4_plus_k2[32]; /* term one */
vector_width_t v0_rshift_5_plus_k3[32]; /* term three */
vector_width_t shift;
int i;
/* setup */
uint32_t sum = delta * r;
for (i = 0; i < 32; ++i)
v1_lshift_4[i] = v1_plus_sum[i] = v1_rshift_5[i] = \
v1_lshift_4_plus_k0[i] = v1_rshift_5_plus_k1[i] = \
v0_lshift_4[i] = v0_plus_sum[i] = v0_rshift_5[i] = \
v0_lshift_4_plus_k2[i] = v0_rshift_5_plus_k3[i] = 0;
while (r > 0) {
/* lshift v0 by 4 */
shift = 4;
for (i = 31; i >= 0; i--)
v0_lshift_4[i] = (i >= shift) ? v[offset_v0 + i - shift] : 0;
/* add k2 and v0_lshift_4 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v0_lshift_4[i];
bi = k[offset_k2 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_lshift_4_plus_k2[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* add delta sum to v0 */
carry = 0;
for (i = 0;i < 32;++i) {
/* VECTOR_AT_ONE where the ith bit of the sum is set */
ai = TERNARY((sum & (1<<i)), VECTOR_AT_ONE, VECTOR_AT_ZERO);
bi = v[offset_v0 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_plus_sum[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* rshift v0 by 5 */
shift = 5;
for (i = 0; i < 32; ++i)
v0_rshift_5[i] = (i < (32 - shift)) ? v[offset_v0 + i + shift] : 0;
/* add k3 to v0_rshift_5 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v0_rshift_5[i];
bi = k[offset_k3 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v0_rshift_5_plus_k3[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* xor the three terms and decrement v1 */
borrow = 0;
for (i = 0;i < 32;++i) {
ai = v[offset_v1 + i];
bi = v0_lshift_4_plus_k2[i] ^ v0_plus_sum[i] ^ v0_rshift_5_plus_k3[i];
notaandb = (ai ^ VECTOR_AT_ONE) & bi;
axorb = ai ^ bi;
v[offset_v1 + i] = axorb ^ borrow;
borrow = notaandb | ((ai ^ VECTOR_AT_ONE) & borrow) | (bi & borrow);
}
/* lshift v1 by 4 */
shift = 4;
for (i = 31; i >= 0; i--)
v1_lshift_4[i] = (i >= shift) ? v[offset_v1 + i - shift] : 0;
/* add k0 to v1_lshift_4 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v1_lshift_4[i];
bi = k[offset_k0 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_lshift_4_plus_k0[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* add delta sum to v1 */
carry = 0;
for (i = 0;i < 32;++i) {
/* VECTOR_AT_ONE where the ith bit of the sum is set */
ai = TERNARY((sum & (1<<i)), VECTOR_AT_ONE, VECTOR_AT_ZERO);
bi = v[offset_v1 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_plus_sum[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* rshift v1 by 5 */
shift = 5;
for (i = 0; i < 32; ++i)
v1_rshift_5[i] = (i < (32 - shift)) ? v[offset_v1 + i + shift] : 0;
/* add k1 to v1_rshift_5 */
carry = 0;
for (i = 0;i < 32;++i) {
ai = v1_rshift_5[i];
bi = k[offset_k1 + i];
aandb = ai & bi;
axorb = ai ^ bi;
v1_rshift_5_plus_k1[i] = axorb ^ carry;
carry &= axorb;
carry |= aandb;
}
/* xor the three terms and decrement v0 */
borrow = 0;
for (i = 0;i < 32;++i) {
ai = v[offset_v0 + i];
bi = v1_lshift_4_plus_k0[i] ^ v1_plus_sum[i] ^ v1_rshift_5_plus_k1[i];
notaandb = (ai ^ VECTOR_AT_ONE) & bi;
axorb = ai ^ bi;
v[offset_v0 + i] = axorb ^ borrow;
borrow = notaandb | ((ai ^ VECTOR_AT_ONE) & borrow) | (bi & borrow);
}
sum -= delta;
--r;
}
}
|
