1
2 // Copyright in this code is held by Dr B.R. Gladman but free direct or
3 // derivative use is permitted subject to acknowledgement of its origin
4 // and subject to any constraints placed on the use of the algorithm by
5 // its designers (if such constraints may exist, this will be indicated
6 // below).
7 //
8 // Dr. B. R. Gladman . 25th January 2000.
9 //
10 // This is an implementation of Serpent, an encryption algorithm designed
11 // by Anderson, Biham and Knudsen and submitted as a candidate for the
12 // Advanced Encryption Standard programme of the US National Institute of
13 // Standards and Technology.
14 //
15 // The designers of Serpent have not placed any constraints on the use of
16 // this algorithm.
17
18 #include <sys/types.h>
19 #include <sys/endian.h>
20
21 #include <crypto/serpent/serpent.h>
22
23 // Partially optimised Serpent S Box boolean functions derived
24 // using a recursive descent analyser but without a full search
25 // of all subtrees. This set of S boxes is the result of work
26 // by Sam Simpson and Brian Gladman using the spare time on a
27 // cluster of high capacity servers to search for S boxes with
28 // this customised search engine.
29 //
30 // Copyright: Dr B. R Gladman (gladman@seven77.demon.co.uk)
31 // and Sam Simpson (s.simpson@mia.co.uk)
32 // 17th December 1998
33 //
34 // We hereby give permission for information in this file to be
35 // used freely subject only to acknowledgement of its origin
36
37 // 15 terms
38
39 #define rotr(x,n) (((x) >> ((int)((n) & 0x1f))) | ((x) << ((int)((32 - ((n) & 0x1f))))))
40 #define rotl(x,n) (((x) << ((int)((n) & 0x1f))) | ((x) >> ((int)((32 - ((n) & 0x1f))))))
41
42 #define sb0(a,b,c,d,e,f,g,h) \
43 { t1 = a ^ d; \
44 t2 = a & d; \
45 t3 = c ^ t1; \
46 t6 = b & t1; \
47 t4 = b ^ t3; \
48 t10 = ~t3; \
49 h = t2 ^ t4; \
50 t7 = a ^ t6; \
51 t14 = ~t7; \
52 t8 = c | t7; \
53 t11 = t3 ^ t7; \
54 g = t4 ^ t8; \
55 t12 = h & t11; \
56 f = t10 ^ t12; \
57 e = t12 ^ t14; \
58 }
59 // 15 terms
60
61 #define ib0(a,b,c,d,e,f,g,h) \
62 { t1 = ~a; \
63 t2 = a ^ b; \
64 t3 = t1 | t2; \
65 t4 = d ^ t3; \
66 t7 = d & t2; \
67 t5 = c ^ t4; \
68 t8 = t1 ^ t7; \
69 g = t2 ^ t5; \
70 t11 = a & t4; \
71 t9 = g & t8; \
72 t14 = t5 ^ t8; \
73 f = t4 ^ t9; \
74 t12 = t5 | f; \
75 h = t11 ^ t12; \
76 e = h ^ t14; \
77 }
78
79 // 14 terms!
80
81 #define sb1(a,b,c,d,e,f,g,h) \
82 { t1 = ~a; \
83 t2 = b ^ t1; \
84 t3 = a | t2; \
85 t4 = d | t2; \
86 t5 = c ^ t3; \
87 g = d ^ t5; \
88 t7 = b ^ t4; \
89 t8 = t2 ^ g; \
90 t9 = t5 & t7; \
91 h = t8 ^ t9; \
92 t11 = t5 ^ t7; \
93 f = h ^ t11; \
94 t13 = t8 & t11; \
95 e = t5 ^ t13; \
96 }
97
98 // 17 terms
99
100 #define ib1(a,b,c,d,e,f,g,h) \
101 { t1 = a ^ d; \
102 t2 = a & b; \
103 t3 = b ^ c; \
104 t4 = a ^ t3; \
105 t5 = b | d; \
106 t7 = c | t1; \
107 h = t4 ^ t5; \
108 t8 = b ^ t7; \
109 t11 = ~t2; \
110 t9 = t4 & t8; \
111 f = t1 ^ t9; \
112 t13 = t9 ^ t11; \
113 t12 = h & f; \
114 g = t12 ^ t13; \
115 t15 = a & d; \
116 t16 = c ^ t13; \
117 e = t15 ^ t16; \
118 }
119
120 // 16 terms
121
122 #define sb2(a,b,c,d,e,f,g,h) \
123 { t1 = ~a; \
124 t2 = b ^ d; \
125 t3 = c & t1; \
126 t13 = d | t1; \
127 e = t2 ^ t3; \
128 t5 = c ^ t1; \
129 t6 = c ^ e; \
130 t7 = b & t6; \
131 t10 = e | t5; \
132 h = t5 ^ t7; \
133 t9 = d | t7; \
134 t11 = t9 & t10; \
135 t14 = t2 ^ h; \
136 g = a ^ t11; \
137 t15 = g ^ t13; \
138 f = t14 ^ t15; \
139 }
140
141 // 16 terms
142
143 #define ib2(a,b,c,d,e,f,g,h) \
144 { t1 = b ^ d; \
145 t2 = ~t1; \
146 t3 = a ^ c; \
147 t4 = c ^ t1; \
148 t7 = a | t2; \
149 t5 = b & t4; \
150 t8 = d ^ t7; \
151 t11 = ~t4; \
152 e = t3 ^ t5; \
153 t9 = t3 | t8; \
154 t14 = d & t11; \
155 h = t1 ^ t9; \
156 t12 = e | h; \
157 f = t11 ^ t12; \
158 t15 = t3 ^ t12; \
159 g = t14 ^ t15; \
160 }
161
162 // 17 terms
163
164 #define sb3(a,b,c,d,e,f,g,h) \
165 { t1 = a ^ c; \
166 t2 = d ^ t1; \
167 t3 = a & t2; \
168 t4 = d ^ t3; \
169 t5 = b & t4; \
170 g = t2 ^ t5; \
171 t7 = a | g; \
172 t8 = b | d; \
173 t11 = a | d; \
174 t9 = t4 & t7; \
175 f = t8 ^ t9; \
176 t12 = b ^ t11; \
177 t13 = g ^ t9; \
178 t15 = t3 ^ t8; \
179 h = t12 ^ t13; \
180 t16 = c & t15; \
181 e = t12 ^ t16; \
182 }
183
184 // 16 term solution that performs less well than 17 term one
185 // in my environment (PPro/PII)
186
187 #if 0
188
189 #define sb3(a,b,c,d,e,f,g,h) \
190 { t1 = a ^ b; \
191 t2 = a & c; \
192 t3 = a | d; \
193 t4 = c ^ d; \
194 t5 = t1 & t3; \
195 t6 = t2 | t5; \
196 g = t4 ^ t6; \
197 t8 = b ^ t3; \
198 t9 = t6 ^ t8; \
199 t10 = t4 & t9; \
200 e = t1 ^ t10; \
201 t12 = g & e; \
202 f = t9 ^ t12; \
203 t14 = b | d; \
204 t15 = t4 ^ t12; \
205 h = t14 ^ t15; \
206 }
207
208 #endif
209
210 // 17 terms
211
212 #define ib3(a,b,c,d,e,f,g,h) \
213 { t1 = b ^ c; \
214 t2 = b | c; \
215 t3 = a ^ c; \
216 t7 = a ^ d; \
217 t4 = t2 ^ t3; \
218 t5 = d | t4; \
219 t9 = t2 ^ t7; \
220 e = t1 ^ t5; \
221 t8 = t1 | t5; \
222 t11 = a & t4; \
223 g = t8 ^ t9; \
224 t12 = e | t9; \
225 f = t11 ^ t12; \
226 t14 = a & g; \
227 t15 = t2 ^ t14; \
228 t16 = e & t15; \
229 h = t4 ^ t16; \
230 }
231
232 // 15 terms
233
234 #define sb4(a,b,c,d,e,f,g,h) \
235 { t1 = a ^ d; \
236 t2 = d & t1; \
237 t3 = c ^ t2; \
238 t4 = b | t3; \
239 h = t1 ^ t4; \
240 t6 = ~b; \
241 t7 = t1 | t6; \
242 e = t3 ^ t7; \
243 t9 = a & e; \
244 t10 = t1 ^ t6; \
245 t11 = t4 & t10; \
246 g = t9 ^ t11; \
247 t13 = a ^ t3; \
248 t14 = t10 & g; \
249 f = t13 ^ t14; \
250 }
251
252 // 17 terms
253
254 #define ib4(a,b,c,d,e,f,g,h) \
255 { t1 = c ^ d; \
256 t2 = c | d; \
257 t3 = b ^ t2; \
258 t4 = a & t3; \
259 f = t1 ^ t4; \
260 t6 = a ^ d; \
261 t7 = b | d; \
262 t8 = t6 & t7; \
263 h = t3 ^ t8; \
264 t10 = ~a; \
265 t11 = c ^ h; \
266 t12 = t10 | t11;\
267 e = t3 ^ t12; \
268 t14 = c | t4; \
269 t15 = t7 ^ t14; \
270 t16 = h | t10; \
271 g = t15 ^ t16; \
272 }
273
274 // 16 terms
275
276 #define sb5(a,b,c,d,e,f,g,h) \
277 { t1 = ~a; \
278 t2 = a ^ b; \
279 t3 = a ^ d; \
280 t4 = c ^ t1; \
281 t5 = t2 | t3; \
282 e = t4 ^ t5; \
283 t7 = d & e; \
284 t8 = t2 ^ e; \
285 t10 = t1 | e; \
286 f = t7 ^ t8; \
287 t11 = t2 | t7; \
288 t12 = t3 ^ t10; \
289 t14 = b ^ t7; \
290 g = t11 ^ t12; \
291 t15 = f & t12; \
292 h = t14 ^ t15; \
293 }
294
295 // 16 terms
296
297 #define ib5(a,b,c,d,e,f,g,h) \
298 { t1 = ~c; \
299 t2 = b & t1; \
300 t3 = d ^ t2; \
301 t4 = a & t3; \
302 t5 = b ^ t1; \
303 h = t4 ^ t5; \
304 t7 = b | h; \
305 t8 = a & t7; \
306 f = t3 ^ t8; \
307 t10 = a | d; \
308 t11 = t1 ^ t7; \
309 e = t10 ^ t11; \
310 t13 = a ^ c; \
311 t14 = b & t10; \
312 t15 = t4 | t13; \
313 g = t14 ^ t15; \
314 }
315
316 // 15 terms
317
318 #define sb6(a,b,c,d,e,f,g,h) \
319 { t1 = ~a; \
320 t2 = a ^ d; \
321 t3 = b ^ t2; \
322 t4 = t1 | t2; \
323 t5 = c ^ t4; \
324 f = b ^ t5; \
325 t13 = ~t5; \
326 t7 = t2 | f; \
327 t8 = d ^ t7; \
328 t9 = t5 & t8; \
329 g = t3 ^ t9; \
330 t11 = t5 ^ t8; \
331 e = g ^ t11; \
332 t14 = t3 & t11; \
333 h = t13 ^ t14; \
334 }
335
336 // 15 terms
337
338 #define ib6(a,b,c,d,e,f,g,h) \
339 { t1 = ~a; \
340 t2 = a ^ b; \
341 t3 = c ^ t2; \
342 t4 = c | t1; \
343 t5 = d ^ t4; \
344 t13 = d & t1; \
345 f = t3 ^ t5; \
346 t7 = t3 & t5; \
347 t8 = t2 ^ t7; \
348 t9 = b | t8; \
349 h = t5 ^ t9; \
350 t11 = b | h; \
351 e = t8 ^ t11; \
352 t14 = t3 ^ t11; \
353 g = t13 ^ t14; \
354 }
355
356 // 17 terms
357
358 #define sb7(a,b,c,d,e,f,g,h) \
359 { t1 = ~c; \
360 t2 = b ^ c; \
361 t3 = b | t1; \
362 t4 = d ^ t3; \
363 t5 = a & t4; \
364 t7 = a ^ d; \
365 h = t2 ^ t5; \
366 t8 = b ^ t5; \
367 t9 = t2 | t8; \
368 t11 = d & t3; \
369 f = t7 ^ t9; \
370 t12 = t5 ^ f; \
371 t15 = t1 | t4; \
372 t13 = h & t12; \
373 g = t11 ^ t13; \
374 t16 = t12 ^ g; \
375 e = t15 ^ t16; \
376 }
377
378 // 17 terms
379
380 #define ib7(a,b,c,d,e,f,g,h) \
381 { t1 = a & b; \
382 t2 = a | b; \
383 t3 = c | t1; \
384 t4 = d & t2; \
385 h = t3 ^ t4; \
386 t6 = ~d; \
387 t7 = b ^ t4; \
388 t8 = h ^ t6; \
389 t11 = c ^ t7; \
390 t9 = t7 | t8; \
391 f = a ^ t9; \
392 t12 = d | f; \
393 e = t11 ^ t12; \
394 t14 = a & h; \
395 t15 = t3 ^ f; \
396 t16 = e ^ t14; \
397 g = t15 ^ t16; \
398 }
399
400 #define k_xor(r,a,b,c,d) \
401 { a ^= ctx->l_key[4 * r + 8]; \
402 b ^= ctx->l_key[4 * r + 9]; \
403 c ^= ctx->l_key[4 * r + 10]; \
404 d ^= ctx->l_key[4 * r + 11]; \
405 }
406
407 #define k_set(r,a,b,c,d) \
408 { a = ctx->l_key[4 * r + 8]; \
409 b = ctx->l_key[4 * r + 9]; \
410 c = ctx->l_key[4 * r + 10]; \
411 d = ctx->l_key[4 * r + 11]; \
412 }
413
414 #define k_get(r,a,b,c,d) \
415 { ctx->l_key[4 * r + 8] = a; \
416 ctx->l_key[4 * r + 9] = b; \
417 ctx->l_key[4 * r + 10] = c; \
418 ctx->l_key[4 * r + 11] = d; \
419 }
420
421 // the linear transformation and its inverse
422
423 #define rot(a,b,c,d) \
424 { a = rotl(a, 13); \
425 c = rotl(c, 3); \
426 d ^= c ^ (a << 3); \
427 b ^= a ^ c; \
428 d = rotl(d, 7); \
429 b = rotl(b, 1); \
430 a ^= b ^ d; \
431 c ^= d ^ (b << 7); \
432 a = rotl(a, 5); \
433 c = rotl(c, 22); \
434 }
435
436 #define irot(a,b,c,d) \
437 { c = rotr(c, 22); \
438 a = rotr(a, 5); \
439 c ^= d ^ (b << 7); \
440 a ^= b ^ d; \
441 d = rotr(d, 7); \
442 b = rotr(b, 1); \
443 d ^= c ^ (a << 3); \
444 b ^= a ^ c; \
445 c = rotr(c, 3); \
446 a = rotr(a, 13); \
447 }
448
449 // initialise the key schedule from the user supplied key
450
serpent_set_key(serpent_ctx * ctx,const u_int8_t in_key[],int key_len)451 void serpent_set_key(serpent_ctx *ctx, const u_int8_t in_key[], int key_len)
452 {
453 u_int32_t i,lk,a,b,c,d,e,f,g,h;
454 u_int32_t t1,t2,t3,t4,t5,t6,t7,t8,t9,t10,t11,t12,t13,t14,t15,t16;
455
456 if(key_len > 256)
457 return;
458
459 i = 0; lk = (key_len + 31) / 32;
460
461 while(i < lk)
462 {
463 ctx->l_key[i] = htole32(*((const u_int32_t *)(in_key + 4 * i))); i++;
464 }
465
466 if(key_len < 256)
467 {
468 while(i < 8)
469
470 ctx->l_key[i++] = 0;
471
472 i = key_len / 32; lk = 1 << key_len % 32;
473
474 ctx->l_key[i] = (ctx->l_key[i] & (lk - 1)) | lk;
475 }
476
477 t1 = ctx->l_key[2] ^ ctx->l_key[4] ^ ctx->l_key[6] ^ 0x9e3779b9;
478 t2 = ctx->l_key[3] ^ ctx->l_key[5] ^ ctx->l_key[7] ^ 0x9e3779b9;
479
480 for(i = 0; i < 132; i += 2)
481 {
482 ctx->l_key[i + 8] = rotr(i ^ ctx->l_key[i] ^ t2, 21);
483
484 t1 ^= ctx->l_key[i + 2] ^ ctx->l_key[i + 8];
485
486 ctx->l_key[i + 9] = rotr((i + 1) ^ ctx->l_key[i + 1] ^ t1, 21);
487
488 t2 ^= ctx->l_key[i + 3] ^ ctx->l_key[i + 9];
489 }
490
491 k_set( 0,a,b,c,d);sb3(a,b,c,d,e,f,g,h);k_get( 0,e,f,g,h);
492 k_set( 1,a,b,c,d);sb2(a,b,c,d,e,f,g,h);k_get( 1,e,f,g,h);
493 k_set( 2,a,b,c,d);sb1(a,b,c,d,e,f,g,h);k_get( 2,e,f,g,h);
494 k_set( 3,a,b,c,d);sb0(a,b,c,d,e,f,g,h);k_get( 3,e,f,g,h);
495 k_set( 4,a,b,c,d);sb7(a,b,c,d,e,f,g,h);k_get( 4,e,f,g,h);
496 k_set( 5,a,b,c,d);sb6(a,b,c,d,e,f,g,h);k_get( 5,e,f,g,h);
497 k_set( 6,a,b,c,d);sb5(a,b,c,d,e,f,g,h);k_get( 6,e,f,g,h);
498 k_set( 7,a,b,c,d);sb4(a,b,c,d,e,f,g,h);k_get( 7,e,f,g,h);
499 k_set( 8,a,b,c,d);sb3(a,b,c,d,e,f,g,h);k_get( 8,e,f,g,h);
500 k_set( 9,a,b,c,d);sb2(a,b,c,d,e,f,g,h);k_get( 9,e,f,g,h);
501 k_set(10,a,b,c,d);sb1(a,b,c,d,e,f,g,h);k_get(10,e,f,g,h);
502 k_set(11,a,b,c,d);sb0(a,b,c,d,e,f,g,h);k_get(11,e,f,g,h);
503 k_set(12,a,b,c,d);sb7(a,b,c,d,e,f,g,h);k_get(12,e,f,g,h);
504 k_set(13,a,b,c,d);sb6(a,b,c,d,e,f,g,h);k_get(13,e,f,g,h);
505 k_set(14,a,b,c,d);sb5(a,b,c,d,e,f,g,h);k_get(14,e,f,g,h);
506 k_set(15,a,b,c,d);sb4(a,b,c,d,e,f,g,h);k_get(15,e,f,g,h);
507 k_set(16,a,b,c,d);sb3(a,b,c,d,e,f,g,h);k_get(16,e,f,g,h);
508 k_set(17,a,b,c,d);sb2(a,b,c,d,e,f,g,h);k_get(17,e,f,g,h);
509 k_set(18,a,b,c,d);sb1(a,b,c,d,e,f,g,h);k_get(18,e,f,g,h);
510 k_set(19,a,b,c,d);sb0(a,b,c,d,e,f,g,h);k_get(19,e,f,g,h);
511 k_set(20,a,b,c,d);sb7(a,b,c,d,e,f,g,h);k_get(20,e,f,g,h);
512 k_set(21,a,b,c,d);sb6(a,b,c,d,e,f,g,h);k_get(21,e,f,g,h);
513 k_set(22,a,b,c,d);sb5(a,b,c,d,e,f,g,h);k_get(22,e,f,g,h);
514 k_set(23,a,b,c,d);sb4(a,b,c,d,e,f,g,h);k_get(23,e,f,g,h);
515 k_set(24,a,b,c,d);sb3(a,b,c,d,e,f,g,h);k_get(24,e,f,g,h);
516 k_set(25,a,b,c,d);sb2(a,b,c,d,e,f,g,h);k_get(25,e,f,g,h);
517 k_set(26,a,b,c,d);sb1(a,b,c,d,e,f,g,h);k_get(26,e,f,g,h);
518 k_set(27,a,b,c,d);sb0(a,b,c,d,e,f,g,h);k_get(27,e,f,g,h);
519 k_set(28,a,b,c,d);sb7(a,b,c,d,e,f,g,h);k_get(28,e,f,g,h);
520 k_set(29,a,b,c,d);sb6(a,b,c,d,e,f,g,h);k_get(29,e,f,g,h);
521 k_set(30,a,b,c,d);sb5(a,b,c,d,e,f,g,h);k_get(30,e,f,g,h);
522 k_set(31,a,b,c,d);sb4(a,b,c,d,e,f,g,h);k_get(31,e,f,g,h);
523 k_set(32,a,b,c,d);sb3(a,b,c,d,e,f,g,h);k_get(32,e,f,g,h);
524
525 return;
526 }
527
528 // encrypt a block of text
529
serpent_encrypt(serpent_ctx * ctx,const u_int8_t in_blk[],u_int8_t out_blk[])530 void serpent_encrypt(serpent_ctx *ctx, const u_int8_t in_blk[],
531 u_int8_t out_blk[])
532 {
533 u_int32_t a,b,c,d,e,f,g,h;
534 u_int32_t t1,t2,t3,t4,t5,t6,t7,t8,t9,t10,t11,t12,t13,t14,t15,t16;
535
536 a = htole32(*((const u_int32_t *)(in_blk )));
537 b = htole32(*((const u_int32_t *)(in_blk + 4 )));
538 c = htole32(*((const u_int32_t *)(in_blk + 8 )));
539 d = htole32(*((const u_int32_t *)(in_blk + 12)));
540
541 k_xor( 0,a,b,c,d); sb0(a,b,c,d,e,f,g,h); rot(e,f,g,h);
542 k_xor( 1,e,f,g,h); sb1(e,f,g,h,a,b,c,d); rot(a,b,c,d);
543 k_xor( 2,a,b,c,d); sb2(a,b,c,d,e,f,g,h); rot(e,f,g,h);
544 k_xor( 3,e,f,g,h); sb3(e,f,g,h,a,b,c,d); rot(a,b,c,d);
545 k_xor( 4,a,b,c,d); sb4(a,b,c,d,e,f,g,h); rot(e,f,g,h);
546 k_xor( 5,e,f,g,h); sb5(e,f,g,h,a,b,c,d); rot(a,b,c,d);
547 k_xor( 6,a,b,c,d); sb6(a,b,c,d,e,f,g,h); rot(e,f,g,h);
548 k_xor( 7,e,f,g,h); sb7(e,f,g,h,a,b,c,d); rot(a,b,c,d);
549 k_xor( 8,a,b,c,d); sb0(a,b,c,d,e,f,g,h); rot(e,f,g,h);
550 k_xor( 9,e,f,g,h); sb1(e,f,g,h,a,b,c,d); rot(a,b,c,d);
551 k_xor(10,a,b,c,d); sb2(a,b,c,d,e,f,g,h); rot(e,f,g,h);
552 k_xor(11,e,f,g,h); sb3(e,f,g,h,a,b,c,d); rot(a,b,c,d);
553 k_xor(12,a,b,c,d); sb4(a,b,c,d,e,f,g,h); rot(e,f,g,h);
554 k_xor(13,e,f,g,h); sb5(e,f,g,h,a,b,c,d); rot(a,b,c,d);
555 k_xor(14,a,b,c,d); sb6(a,b,c,d,e,f,g,h); rot(e,f,g,h);
556 k_xor(15,e,f,g,h); sb7(e,f,g,h,a,b,c,d); rot(a,b,c,d);
557 k_xor(16,a,b,c,d); sb0(a,b,c,d,e,f,g,h); rot(e,f,g,h);
558 k_xor(17,e,f,g,h); sb1(e,f,g,h,a,b,c,d); rot(a,b,c,d);
559 k_xor(18,a,b,c,d); sb2(a,b,c,d,e,f,g,h); rot(e,f,g,h);
560 k_xor(19,e,f,g,h); sb3(e,f,g,h,a,b,c,d); rot(a,b,c,d);
561 k_xor(20,a,b,c,d); sb4(a,b,c,d,e,f,g,h); rot(e,f,g,h);
562 k_xor(21,e,f,g,h); sb5(e,f,g,h,a,b,c,d); rot(a,b,c,d);
563 k_xor(22,a,b,c,d); sb6(a,b,c,d,e,f,g,h); rot(e,f,g,h);
564 k_xor(23,e,f,g,h); sb7(e,f,g,h,a,b,c,d); rot(a,b,c,d);
565 k_xor(24,a,b,c,d); sb0(a,b,c,d,e,f,g,h); rot(e,f,g,h);
566 k_xor(25,e,f,g,h); sb1(e,f,g,h,a,b,c,d); rot(a,b,c,d);
567 k_xor(26,a,b,c,d); sb2(a,b,c,d,e,f,g,h); rot(e,f,g,h);
568 k_xor(27,e,f,g,h); sb3(e,f,g,h,a,b,c,d); rot(a,b,c,d);
569 k_xor(28,a,b,c,d); sb4(a,b,c,d,e,f,g,h); rot(e,f,g,h);
570 k_xor(29,e,f,g,h); sb5(e,f,g,h,a,b,c,d); rot(a,b,c,d);
571 k_xor(30,a,b,c,d); sb6(a,b,c,d,e,f,g,h); rot(e,f,g,h);
572 k_xor(31,e,f,g,h); sb7(e,f,g,h,a,b,c,d); k_xor(32,a,b,c,d);
573
574 *((u_int32_t *)(out_blk )) = le32toh(a);
575 *((u_int32_t *)(out_blk + 4)) = le32toh(b);
576 *((u_int32_t *)(out_blk + 8)) = le32toh(c);
577 *((u_int32_t *)(out_blk + 12)) = le32toh(d);
578 }
579
580 // decrypt a block of text
581
serpent_decrypt(serpent_ctx * ctx,const u_int8_t in_blk[],u_int8_t out_blk[])582 void serpent_decrypt(serpent_ctx *ctx, const u_int8_t in_blk[],
583 u_int8_t out_blk[])
584 {
585 u_int32_t a,b,c,d,e,f,g,h;
586 u_int32_t t1,t2,t3,t4,t5,t6,t7,t8,t9,t10,t11,t12,t13,t14,t15,t16;
587
588 a = htole32(*((const u_int32_t *)(in_blk )));
589 b = htole32(*((const u_int32_t *)(in_blk + 4 )));
590 c = htole32(*((const u_int32_t *)(in_blk + 8 )));
591 d = htole32(*((const u_int32_t *)(in_blk + 12)));
592
593 k_xor(32,a,b,c,d); ib7(a,b,c,d,e,f,g,h); k_xor(31,e,f,g,h);
594 irot(e,f,g,h); ib6(e,f,g,h,a,b,c,d); k_xor(30,a,b,c,d);
595 irot(a,b,c,d); ib5(a,b,c,d,e,f,g,h); k_xor(29,e,f,g,h);
596 irot(e,f,g,h); ib4(e,f,g,h,a,b,c,d); k_xor(28,a,b,c,d);
597 irot(a,b,c,d); ib3(a,b,c,d,e,f,g,h); k_xor(27,e,f,g,h);
598 irot(e,f,g,h); ib2(e,f,g,h,a,b,c,d); k_xor(26,a,b,c,d);
599 irot(a,b,c,d); ib1(a,b,c,d,e,f,g,h); k_xor(25,e,f,g,h);
600 irot(e,f,g,h); ib0(e,f,g,h,a,b,c,d); k_xor(24,a,b,c,d);
601 irot(a,b,c,d); ib7(a,b,c,d,e,f,g,h); k_xor(23,e,f,g,h);
602 irot(e,f,g,h); ib6(e,f,g,h,a,b,c,d); k_xor(22,a,b,c,d);
603 irot(a,b,c,d); ib5(a,b,c,d,e,f,g,h); k_xor(21,e,f,g,h);
604 irot(e,f,g,h); ib4(e,f,g,h,a,b,c,d); k_xor(20,a,b,c,d);
605 irot(a,b,c,d); ib3(a,b,c,d,e,f,g,h); k_xor(19,e,f,g,h);
606 irot(e,f,g,h); ib2(e,f,g,h,a,b,c,d); k_xor(18,a,b,c,d);
607 irot(a,b,c,d); ib1(a,b,c,d,e,f,g,h); k_xor(17,e,f,g,h);
608 irot(e,f,g,h); ib0(e,f,g,h,a,b,c,d); k_xor(16,a,b,c,d);
609 irot(a,b,c,d); ib7(a,b,c,d,e,f,g,h); k_xor(15,e,f,g,h);
610 irot(e,f,g,h); ib6(e,f,g,h,a,b,c,d); k_xor(14,a,b,c,d);
611 irot(a,b,c,d); ib5(a,b,c,d,e,f,g,h); k_xor(13,e,f,g,h);
612 irot(e,f,g,h); ib4(e,f,g,h,a,b,c,d); k_xor(12,a,b,c,d);
613 irot(a,b,c,d); ib3(a,b,c,d,e,f,g,h); k_xor(11,e,f,g,h);
614 irot(e,f,g,h); ib2(e,f,g,h,a,b,c,d); k_xor(10,a,b,c,d);
615 irot(a,b,c,d); ib1(a,b,c,d,e,f,g,h); k_xor( 9,e,f,g,h);
616 irot(e,f,g,h); ib0(e,f,g,h,a,b,c,d); k_xor( 8,a,b,c,d);
617 irot(a,b,c,d); ib7(a,b,c,d,e,f,g,h); k_xor( 7,e,f,g,h);
618 irot(e,f,g,h); ib6(e,f,g,h,a,b,c,d); k_xor( 6,a,b,c,d);
619 irot(a,b,c,d); ib5(a,b,c,d,e,f,g,h); k_xor( 5,e,f,g,h);
620 irot(e,f,g,h); ib4(e,f,g,h,a,b,c,d); k_xor( 4,a,b,c,d);
621 irot(a,b,c,d); ib3(a,b,c,d,e,f,g,h); k_xor( 3,e,f,g,h);
622 irot(e,f,g,h); ib2(e,f,g,h,a,b,c,d); k_xor( 2,a,b,c,d);
623 irot(a,b,c,d); ib1(a,b,c,d,e,f,g,h); k_xor( 1,e,f,g,h);
624 irot(e,f,g,h); ib0(e,f,g,h,a,b,c,d); k_xor( 0,a,b,c,d);
625
626 *((u_int32_t *)(out_blk )) = le32toh(a);
627 *((u_int32_t *)(out_blk + 4)) = le32toh(b);
628 *((u_int32_t *)(out_blk + 8)) = le32toh(c);
629 *((u_int32_t *)(out_blk + 12)) = le32toh(d);
630 }
631
632