xref: /openbmc/linux/crypto/serpent_generic.c (revision ae213c44)
1 /*
2  * Cryptographic API.
3  *
4  * Serpent Cipher Algorithm.
5  *
6  * Copyright (C) 2002 Dag Arne Osvik <osvik@ii.uib.no>
7  *               2003 Herbert Valerio Riedel <hvr@gnu.org>
8  *
9  * Added tnepres support:
10  *		Ruben Jesus Garcia Hernandez <ruben@ugr.es>, 18.10.2004
11  *              Based on code by hvr
12  *
13  * This program is free software; you can redistribute it and/or modify
14  * it under the terms of the GNU General Public License as published by
15  * the Free Software Foundation; either version 2 of the License, or
16  * (at your option) any later version.
17  */
18 
19 #include <linux/init.h>
20 #include <linux/module.h>
21 #include <linux/errno.h>
22 #include <asm/byteorder.h>
23 #include <linux/crypto.h>
24 #include <linux/types.h>
25 #include <crypto/serpent.h>
26 
27 /* Key is padded to the maximum of 256 bits before round key generation.
28  * Any key length <= 256 bits (32 bytes) is allowed by the algorithm.
29  */
30 
31 #define PHI 0x9e3779b9UL
32 
33 #define keyiter(a, b, c, d, i, j) \
34 	({ b ^= d; b ^= c; b ^= a; b ^= PHI ^ i; b = rol32(b, 11); k[j] = b; })
35 
36 #define loadkeys(x0, x1, x2, x3, i) \
37 	({ x0 = k[i]; x1 = k[i+1]; x2 = k[i+2]; x3 = k[i+3]; })
38 
39 #define storekeys(x0, x1, x2, x3, i) \
40 	({ k[i] = x0; k[i+1] = x1; k[i+2] = x2; k[i+3] = x3; })
41 
42 #define store_and_load_keys(x0, x1, x2, x3, s, l) \
43 	({ storekeys(x0, x1, x2, x3, s); loadkeys(x0, x1, x2, x3, l); })
44 
45 #define K(x0, x1, x2, x3, i) ({				\
46 	x3 ^= k[4*(i)+3];        x2 ^= k[4*(i)+2];	\
47 	x1 ^= k[4*(i)+1];        x0 ^= k[4*(i)+0];	\
48 	})
49 
50 #define LK(x0, x1, x2, x3, x4, i) ({					   \
51 							x0 = rol32(x0, 13);\
52 	x2 = rol32(x2, 3);	x1 ^= x0;		x4  = x0 << 3;	   \
53 	x3 ^= x2;		x1 ^= x2;				   \
54 	x1 = rol32(x1, 1);	x3 ^= x4;				   \
55 	x3 = rol32(x3, 7);	x4  = x1;				   \
56 	x0 ^= x1;		x4 <<= 7;		x2 ^= x3;	   \
57 	x0 ^= x3;		x2 ^= x4;		x3 ^= k[4*i+3];	   \
58 	x1 ^= k[4*i+1];		x0 = rol32(x0, 5);	x2 = rol32(x2, 22);\
59 	x0 ^= k[4*i+0];		x2 ^= k[4*i+2];				   \
60 	})
61 
62 #define KL(x0, x1, x2, x3, x4, i) ({					   \
63 	x0 ^= k[4*i+0];		x1 ^= k[4*i+1];		x2 ^= k[4*i+2];	   \
64 	x3 ^= k[4*i+3];		x0 = ror32(x0, 5);	x2 = ror32(x2, 22);\
65 	x4 =  x1;		x2 ^= x3;		x0 ^= x3;	   \
66 	x4 <<= 7;		x0 ^= x1;		x1 = ror32(x1, 1); \
67 	x2 ^= x4;		x3 = ror32(x3, 7);	x4 = x0 << 3;	   \
68 	x1 ^= x0;		x3 ^= x4;		x0 = ror32(x0, 13);\
69 	x1 ^= x2;		x3 ^= x2;		x2 = ror32(x2, 3); \
70 	})
71 
72 #define S0(x0, x1, x2, x3, x4) ({			\
73 					x4  = x3;	\
74 	x3 |= x0;	x0 ^= x4;	x4 ^= x2;	\
75 	x4 = ~x4;	x3 ^= x1;	x1 &= x0;	\
76 	x1 ^= x4;	x2 ^= x0;	x0 ^= x3;	\
77 	x4 |= x0;	x0 ^= x2;	x2 &= x1;	\
78 	x3 ^= x2;	x1 = ~x1;	x2 ^= x4;	\
79 	x1 ^= x2;					\
80 	})
81 
82 #define S1(x0, x1, x2, x3, x4) ({			\
83 					x4  = x1;	\
84 	x1 ^= x0;	x0 ^= x3;	x3 = ~x3;	\
85 	x4 &= x1;	x0 |= x1;	x3 ^= x2;	\
86 	x0 ^= x3;	x1 ^= x3;	x3 ^= x4;	\
87 	x1 |= x4;	x4 ^= x2;	x2 &= x0;	\
88 	x2 ^= x1;	x1 |= x0;	x0 = ~x0;	\
89 	x0 ^= x2;	x4 ^= x1;			\
90 	})
91 
92 #define S2(x0, x1, x2, x3, x4) ({			\
93 					x3 = ~x3;	\
94 	x1 ^= x0;	x4  = x0;	x0 &= x2;	\
95 	x0 ^= x3;	x3 |= x4;	x2 ^= x1;	\
96 	x3 ^= x1;	x1 &= x0;	x0 ^= x2;	\
97 	x2 &= x3;	x3 |= x1;	x0 = ~x0;	\
98 	x3 ^= x0;	x4 ^= x0;	x0 ^= x2;	\
99 	x1 |= x2;					\
100 	})
101 
102 #define S3(x0, x1, x2, x3, x4) ({			\
103 					x4  = x1;	\
104 	x1 ^= x3;	x3 |= x0;	x4 &= x0;	\
105 	x0 ^= x2;	x2 ^= x1;	x1 &= x3;	\
106 	x2 ^= x3;	x0 |= x4;	x4 ^= x3;	\
107 	x1 ^= x0;	x0 &= x3;	x3 &= x4;	\
108 	x3 ^= x2;	x4 |= x1;	x2 &= x1;	\
109 	x4 ^= x3;	x0 ^= x3;	x3 ^= x2;	\
110 	})
111 
112 #define S4(x0, x1, x2, x3, x4) ({			\
113 					x4  = x3;	\
114 	x3 &= x0;	x0 ^= x4;			\
115 	x3 ^= x2;	x2 |= x4;	x0 ^= x1;	\
116 	x4 ^= x3;	x2 |= x0;			\
117 	x2 ^= x1;	x1 &= x0;			\
118 	x1 ^= x4;	x4 &= x2;	x2 ^= x3;	\
119 	x4 ^= x0;	x3 |= x1;	x1 = ~x1;	\
120 	x3 ^= x0;					\
121 	})
122 
123 #define S5(x0, x1, x2, x3, x4) ({			\
124 	x4  = x1;	x1 |= x0;			\
125 	x2 ^= x1;	x3 = ~x3;	x4 ^= x0;	\
126 	x0 ^= x2;	x1 &= x4;	x4 |= x3;	\
127 	x4 ^= x0;	x0 &= x3;	x1 ^= x3;	\
128 	x3 ^= x2;	x0 ^= x1;	x2 &= x4;	\
129 	x1 ^= x2;	x2 &= x0;			\
130 	x3 ^= x2;					\
131 	})
132 
133 #define S6(x0, x1, x2, x3, x4) ({			\
134 					x4  = x1;	\
135 	x3 ^= x0;	x1 ^= x2;	x2 ^= x0;	\
136 	x0 &= x3;	x1 |= x3;	x4 = ~x4;	\
137 	x0 ^= x1;	x1 ^= x2;			\
138 	x3 ^= x4;	x4 ^= x0;	x2 &= x0;	\
139 	x4 ^= x1;	x2 ^= x3;	x3 &= x1;	\
140 	x3 ^= x0;	x1 ^= x2;			\
141 	})
142 
143 #define S7(x0, x1, x2, x3, x4) ({			\
144 					x1 = ~x1;	\
145 	x4  = x1;	x0 = ~x0;	x1 &= x2;	\
146 	x1 ^= x3;	x3 |= x4;	x4 ^= x2;	\
147 	x2 ^= x3;	x3 ^= x0;	x0 |= x1;	\
148 	x2 &= x0;	x0 ^= x4;	x4 ^= x3;	\
149 	x3 &= x0;	x4 ^= x1;			\
150 	x2 ^= x4;	x3 ^= x1;	x4 |= x0;	\
151 	x4 ^= x1;					\
152 	})
153 
154 #define SI0(x0, x1, x2, x3, x4) ({			\
155 			x4  = x3;	x1 ^= x0;	\
156 	x3 |= x1;	x4 ^= x1;	x0 = ~x0;	\
157 	x2 ^= x3;	x3 ^= x0;	x0 &= x1;	\
158 	x0 ^= x2;	x2 &= x3;	x3 ^= x4;	\
159 	x2 ^= x3;	x1 ^= x3;	x3 &= x0;	\
160 	x1 ^= x0;	x0 ^= x2;	x4 ^= x3;	\
161 	})
162 
163 #define SI1(x0, x1, x2, x3, x4) ({			\
164 	x1 ^= x3;	x4  = x0;			\
165 	x0 ^= x2;	x2 = ~x2;	x4 |= x1;	\
166 	x4 ^= x3;	x3 &= x1;	x1 ^= x2;	\
167 	x2 &= x4;	x4 ^= x1;	x1 |= x3;	\
168 	x3 ^= x0;	x2 ^= x0;	x0 |= x4;	\
169 	x2 ^= x4;	x1 ^= x0;			\
170 	x4 ^= x1;					\
171 	})
172 
173 #define SI2(x0, x1, x2, x3, x4) ({			\
174 	x2 ^= x1;	x4  = x3;	x3 = ~x3;	\
175 	x3 |= x2;	x2 ^= x4;	x4 ^= x0;	\
176 	x3 ^= x1;	x1 |= x2;	x2 ^= x0;	\
177 	x1 ^= x4;	x4 |= x3;	x2 ^= x3;	\
178 	x4 ^= x2;	x2 &= x1;			\
179 	x2 ^= x3;	x3 ^= x4;	x4 ^= x0;	\
180 	})
181 
182 #define SI3(x0, x1, x2, x3, x4) ({			\
183 					x2 ^= x1;	\
184 	x4  = x1;	x1 &= x2;			\
185 	x1 ^= x0;	x0 |= x4;	x4 ^= x3;	\
186 	x0 ^= x3;	x3 |= x1;	x1 ^= x2;	\
187 	x1 ^= x3;	x0 ^= x2;	x2 ^= x3;	\
188 	x3 &= x1;	x1 ^= x0;	x0 &= x2;	\
189 	x4 ^= x3;	x3 ^= x0;	x0 ^= x1;	\
190 	})
191 
192 #define SI4(x0, x1, x2, x3, x4) ({			\
193 	x2 ^= x3;	x4  = x0;	x0 &= x1;	\
194 	x0 ^= x2;	x2 |= x3;	x4 = ~x4;	\
195 	x1 ^= x0;	x0 ^= x2;	x2 &= x4;	\
196 	x2 ^= x0;	x0 |= x4;			\
197 	x0 ^= x3;	x3 &= x2;			\
198 	x4 ^= x3;	x3 ^= x1;	x1 &= x0;	\
199 	x4 ^= x1;	x0 ^= x3;			\
200 	})
201 
202 #define SI5(x0, x1, x2, x3, x4) ({			\
203 			x4  = x1;	x1 |= x2;	\
204 	x2 ^= x4;	x1 ^= x3;	x3 &= x4;	\
205 	x2 ^= x3;	x3 |= x0;	x0 = ~x0;	\
206 	x3 ^= x2;	x2 |= x0;	x4 ^= x1;	\
207 	x2 ^= x4;	x4 &= x0;	x0 ^= x1;	\
208 	x1 ^= x3;	x0 &= x2;	x2 ^= x3;	\
209 	x0 ^= x2;	x2 ^= x4;	x4 ^= x3;	\
210 	})
211 
212 #define SI6(x0, x1, x2, x3, x4) ({			\
213 			x0 ^= x2;			\
214 	x4  = x0;	x0 &= x3;	x2 ^= x3;	\
215 	x0 ^= x2;	x3 ^= x1;	x2 |= x4;	\
216 	x2 ^= x3;	x3 &= x0;	x0 = ~x0;	\
217 	x3 ^= x1;	x1 &= x2;	x4 ^= x0;	\
218 	x3 ^= x4;	x4 ^= x2;	x0 ^= x1;	\
219 	x2 ^= x0;					\
220 	})
221 
222 #define SI7(x0, x1, x2, x3, x4) ({			\
223 	x4  = x3;	x3 &= x0;	x0 ^= x2;	\
224 	x2 |= x4;	x4 ^= x1;	x0 = ~x0;	\
225 	x1 |= x3;	x4 ^= x0;	x0 &= x2;	\
226 	x0 ^= x1;	x1 &= x2;	x3 ^= x2;	\
227 	x4 ^= x3;	x2 &= x3;	x3 |= x0;	\
228 	x1 ^= x4;	x3 ^= x4;	x4 &= x0;	\
229 	x4 ^= x2;					\
230 	})
231 
232 static void __serpent_setkey_sbox(u32 r0, u32 r1, u32 r2, u32 r3, u32 r4, u32 *k)
233 {
234 	k += 100;
235 	S3(r3, r4, r0, r1, r2); store_and_load_keys(r1, r2, r4, r3, 28, 24);
236 	S4(r1, r2, r4, r3, r0); store_and_load_keys(r2, r4, r3, r0, 24, 20);
237 	S5(r2, r4, r3, r0, r1); store_and_load_keys(r1, r2, r4, r0, 20, 16);
238 	S6(r1, r2, r4, r0, r3); store_and_load_keys(r4, r3, r2, r0, 16, 12);
239 	S7(r4, r3, r2, r0, r1); store_and_load_keys(r1, r2, r0, r4, 12, 8);
240 	S0(r1, r2, r0, r4, r3); store_and_load_keys(r0, r2, r4, r1, 8, 4);
241 	S1(r0, r2, r4, r1, r3); store_and_load_keys(r3, r4, r1, r0, 4, 0);
242 	S2(r3, r4, r1, r0, r2); store_and_load_keys(r2, r4, r3, r0, 0, -4);
243 	S3(r2, r4, r3, r0, r1); store_and_load_keys(r0, r1, r4, r2, -4, -8);
244 	S4(r0, r1, r4, r2, r3); store_and_load_keys(r1, r4, r2, r3, -8, -12);
245 	S5(r1, r4, r2, r3, r0); store_and_load_keys(r0, r1, r4, r3, -12, -16);
246 	S6(r0, r1, r4, r3, r2); store_and_load_keys(r4, r2, r1, r3, -16, -20);
247 	S7(r4, r2, r1, r3, r0); store_and_load_keys(r0, r1, r3, r4, -20, -24);
248 	S0(r0, r1, r3, r4, r2); store_and_load_keys(r3, r1, r4, r0, -24, -28);
249 	k -= 50;
250 	S1(r3, r1, r4, r0, r2); store_and_load_keys(r2, r4, r0, r3, 22, 18);
251 	S2(r2, r4, r0, r3, r1); store_and_load_keys(r1, r4, r2, r3, 18, 14);
252 	S3(r1, r4, r2, r3, r0); store_and_load_keys(r3, r0, r4, r1, 14, 10);
253 	S4(r3, r0, r4, r1, r2); store_and_load_keys(r0, r4, r1, r2, 10, 6);
254 	S5(r0, r4, r1, r2, r3); store_and_load_keys(r3, r0, r4, r2, 6, 2);
255 	S6(r3, r0, r4, r2, r1); store_and_load_keys(r4, r1, r0, r2, 2, -2);
256 	S7(r4, r1, r0, r2, r3); store_and_load_keys(r3, r0, r2, r4, -2, -6);
257 	S0(r3, r0, r2, r4, r1); store_and_load_keys(r2, r0, r4, r3, -6, -10);
258 	S1(r2, r0, r4, r3, r1); store_and_load_keys(r1, r4, r3, r2, -10, -14);
259 	S2(r1, r4, r3, r2, r0); store_and_load_keys(r0, r4, r1, r2, -14, -18);
260 	S3(r0, r4, r1, r2, r3); store_and_load_keys(r2, r3, r4, r0, -18, -22);
261 	k -= 50;
262 	S4(r2, r3, r4, r0, r1); store_and_load_keys(r3, r4, r0, r1, 28, 24);
263 	S5(r3, r4, r0, r1, r2); store_and_load_keys(r2, r3, r4, r1, 24, 20);
264 	S6(r2, r3, r4, r1, r0); store_and_load_keys(r4, r0, r3, r1, 20, 16);
265 	S7(r4, r0, r3, r1, r2); store_and_load_keys(r2, r3, r1, r4, 16, 12);
266 	S0(r2, r3, r1, r4, r0); store_and_load_keys(r1, r3, r4, r2, 12, 8);
267 	S1(r1, r3, r4, r2, r0); store_and_load_keys(r0, r4, r2, r1, 8, 4);
268 	S2(r0, r4, r2, r1, r3); store_and_load_keys(r3, r4, r0, r1, 4, 0);
269 	S3(r3, r4, r0, r1, r2); storekeys(r1, r2, r4, r3, 0);
270 }
271 
272 int __serpent_setkey(struct serpent_ctx *ctx, const u8 *key,
273 		     unsigned int keylen)
274 {
275 	u32 *k = ctx->expkey;
276 	u8  *k8 = (u8 *)k;
277 	u32 r0, r1, r2, r3, r4;
278 	int i;
279 
280 	/* Copy key, add padding */
281 
282 	for (i = 0; i < keylen; ++i)
283 		k8[i] = key[i];
284 	if (i < SERPENT_MAX_KEY_SIZE)
285 		k8[i++] = 1;
286 	while (i < SERPENT_MAX_KEY_SIZE)
287 		k8[i++] = 0;
288 
289 	/* Expand key using polynomial */
290 
291 	r0 = le32_to_cpu(k[3]);
292 	r1 = le32_to_cpu(k[4]);
293 	r2 = le32_to_cpu(k[5]);
294 	r3 = le32_to_cpu(k[6]);
295 	r4 = le32_to_cpu(k[7]);
296 
297 	keyiter(le32_to_cpu(k[0]), r0, r4, r2, 0, 0);
298 	keyiter(le32_to_cpu(k[1]), r1, r0, r3, 1, 1);
299 	keyiter(le32_to_cpu(k[2]), r2, r1, r4, 2, 2);
300 	keyiter(le32_to_cpu(k[3]), r3, r2, r0, 3, 3);
301 	keyiter(le32_to_cpu(k[4]), r4, r3, r1, 4, 4);
302 	keyiter(le32_to_cpu(k[5]), r0, r4, r2, 5, 5);
303 	keyiter(le32_to_cpu(k[6]), r1, r0, r3, 6, 6);
304 	keyiter(le32_to_cpu(k[7]), r2, r1, r4, 7, 7);
305 
306 	keyiter(k[0], r3, r2, r0, 8, 8);
307 	keyiter(k[1], r4, r3, r1, 9, 9);
308 	keyiter(k[2], r0, r4, r2, 10, 10);
309 	keyiter(k[3], r1, r0, r3, 11, 11);
310 	keyiter(k[4], r2, r1, r4, 12, 12);
311 	keyiter(k[5], r3, r2, r0, 13, 13);
312 	keyiter(k[6], r4, r3, r1, 14, 14);
313 	keyiter(k[7], r0, r4, r2, 15, 15);
314 	keyiter(k[8], r1, r0, r3, 16, 16);
315 	keyiter(k[9], r2, r1, r4, 17, 17);
316 	keyiter(k[10], r3, r2, r0, 18, 18);
317 	keyiter(k[11], r4, r3, r1, 19, 19);
318 	keyiter(k[12], r0, r4, r2, 20, 20);
319 	keyiter(k[13], r1, r0, r3, 21, 21);
320 	keyiter(k[14], r2, r1, r4, 22, 22);
321 	keyiter(k[15], r3, r2, r0, 23, 23);
322 	keyiter(k[16], r4, r3, r1, 24, 24);
323 	keyiter(k[17], r0, r4, r2, 25, 25);
324 	keyiter(k[18], r1, r0, r3, 26, 26);
325 	keyiter(k[19], r2, r1, r4, 27, 27);
326 	keyiter(k[20], r3, r2, r0, 28, 28);
327 	keyiter(k[21], r4, r3, r1, 29, 29);
328 	keyiter(k[22], r0, r4, r2, 30, 30);
329 	keyiter(k[23], r1, r0, r3, 31, 31);
330 
331 	k += 50;
332 
333 	keyiter(k[-26], r2, r1, r4, 32, -18);
334 	keyiter(k[-25], r3, r2, r0, 33, -17);
335 	keyiter(k[-24], r4, r3, r1, 34, -16);
336 	keyiter(k[-23], r0, r4, r2, 35, -15);
337 	keyiter(k[-22], r1, r0, r3, 36, -14);
338 	keyiter(k[-21], r2, r1, r4, 37, -13);
339 	keyiter(k[-20], r3, r2, r0, 38, -12);
340 	keyiter(k[-19], r4, r3, r1, 39, -11);
341 	keyiter(k[-18], r0, r4, r2, 40, -10);
342 	keyiter(k[-17], r1, r0, r3, 41, -9);
343 	keyiter(k[-16], r2, r1, r4, 42, -8);
344 	keyiter(k[-15], r3, r2, r0, 43, -7);
345 	keyiter(k[-14], r4, r3, r1, 44, -6);
346 	keyiter(k[-13], r0, r4, r2, 45, -5);
347 	keyiter(k[-12], r1, r0, r3, 46, -4);
348 	keyiter(k[-11], r2, r1, r4, 47, -3);
349 	keyiter(k[-10], r3, r2, r0, 48, -2);
350 	keyiter(k[-9], r4, r3, r1, 49, -1);
351 	keyiter(k[-8], r0, r4, r2, 50, 0);
352 	keyiter(k[-7], r1, r0, r3, 51, 1);
353 	keyiter(k[-6], r2, r1, r4, 52, 2);
354 	keyiter(k[-5], r3, r2, r0, 53, 3);
355 	keyiter(k[-4], r4, r3, r1, 54, 4);
356 	keyiter(k[-3], r0, r4, r2, 55, 5);
357 	keyiter(k[-2], r1, r0, r3, 56, 6);
358 	keyiter(k[-1], r2, r1, r4, 57, 7);
359 	keyiter(k[0], r3, r2, r0, 58, 8);
360 	keyiter(k[1], r4, r3, r1, 59, 9);
361 	keyiter(k[2], r0, r4, r2, 60, 10);
362 	keyiter(k[3], r1, r0, r3, 61, 11);
363 	keyiter(k[4], r2, r1, r4, 62, 12);
364 	keyiter(k[5], r3, r2, r0, 63, 13);
365 	keyiter(k[6], r4, r3, r1, 64, 14);
366 	keyiter(k[7], r0, r4, r2, 65, 15);
367 	keyiter(k[8], r1, r0, r3, 66, 16);
368 	keyiter(k[9], r2, r1, r4, 67, 17);
369 	keyiter(k[10], r3, r2, r0, 68, 18);
370 	keyiter(k[11], r4, r3, r1, 69, 19);
371 	keyiter(k[12], r0, r4, r2, 70, 20);
372 	keyiter(k[13], r1, r0, r3, 71, 21);
373 	keyiter(k[14], r2, r1, r4, 72, 22);
374 	keyiter(k[15], r3, r2, r0, 73, 23);
375 	keyiter(k[16], r4, r3, r1, 74, 24);
376 	keyiter(k[17], r0, r4, r2, 75, 25);
377 	keyiter(k[18], r1, r0, r3, 76, 26);
378 	keyiter(k[19], r2, r1, r4, 77, 27);
379 	keyiter(k[20], r3, r2, r0, 78, 28);
380 	keyiter(k[21], r4, r3, r1, 79, 29);
381 	keyiter(k[22], r0, r4, r2, 80, 30);
382 	keyiter(k[23], r1, r0, r3, 81, 31);
383 
384 	k += 50;
385 
386 	keyiter(k[-26], r2, r1, r4, 82, -18);
387 	keyiter(k[-25], r3, r2, r0, 83, -17);
388 	keyiter(k[-24], r4, r3, r1, 84, -16);
389 	keyiter(k[-23], r0, r4, r2, 85, -15);
390 	keyiter(k[-22], r1, r0, r3, 86, -14);
391 	keyiter(k[-21], r2, r1, r4, 87, -13);
392 	keyiter(k[-20], r3, r2, r0, 88, -12);
393 	keyiter(k[-19], r4, r3, r1, 89, -11);
394 	keyiter(k[-18], r0, r4, r2, 90, -10);
395 	keyiter(k[-17], r1, r0, r3, 91, -9);
396 	keyiter(k[-16], r2, r1, r4, 92, -8);
397 	keyiter(k[-15], r3, r2, r0, 93, -7);
398 	keyiter(k[-14], r4, r3, r1, 94, -6);
399 	keyiter(k[-13], r0, r4, r2, 95, -5);
400 	keyiter(k[-12], r1, r0, r3, 96, -4);
401 	keyiter(k[-11], r2, r1, r4, 97, -3);
402 	keyiter(k[-10], r3, r2, r0, 98, -2);
403 	keyiter(k[-9], r4, r3, r1, 99, -1);
404 	keyiter(k[-8], r0, r4, r2, 100, 0);
405 	keyiter(k[-7], r1, r0, r3, 101, 1);
406 	keyiter(k[-6], r2, r1, r4, 102, 2);
407 	keyiter(k[-5], r3, r2, r0, 103, 3);
408 	keyiter(k[-4], r4, r3, r1, 104, 4);
409 	keyiter(k[-3], r0, r4, r2, 105, 5);
410 	keyiter(k[-2], r1, r0, r3, 106, 6);
411 	keyiter(k[-1], r2, r1, r4, 107, 7);
412 	keyiter(k[0], r3, r2, r0, 108, 8);
413 	keyiter(k[1], r4, r3, r1, 109, 9);
414 	keyiter(k[2], r0, r4, r2, 110, 10);
415 	keyiter(k[3], r1, r0, r3, 111, 11);
416 	keyiter(k[4], r2, r1, r4, 112, 12);
417 	keyiter(k[5], r3, r2, r0, 113, 13);
418 	keyiter(k[6], r4, r3, r1, 114, 14);
419 	keyiter(k[7], r0, r4, r2, 115, 15);
420 	keyiter(k[8], r1, r0, r3, 116, 16);
421 	keyiter(k[9], r2, r1, r4, 117, 17);
422 	keyiter(k[10], r3, r2, r0, 118, 18);
423 	keyiter(k[11], r4, r3, r1, 119, 19);
424 	keyiter(k[12], r0, r4, r2, 120, 20);
425 	keyiter(k[13], r1, r0, r3, 121, 21);
426 	keyiter(k[14], r2, r1, r4, 122, 22);
427 	keyiter(k[15], r3, r2, r0, 123, 23);
428 	keyiter(k[16], r4, r3, r1, 124, 24);
429 	keyiter(k[17], r0, r4, r2, 125, 25);
430 	keyiter(k[18], r1, r0, r3, 126, 26);
431 	keyiter(k[19], r2, r1, r4, 127, 27);
432 	keyiter(k[20], r3, r2, r0, 128, 28);
433 	keyiter(k[21], r4, r3, r1, 129, 29);
434 	keyiter(k[22], r0, r4, r2, 130, 30);
435 	keyiter(k[23], r1, r0, r3, 131, 31);
436 
437 	/* Apply S-boxes */
438 	__serpent_setkey_sbox(r0, r1, r2, r3, r4, ctx->expkey);
439 
440 	return 0;
441 }
442 EXPORT_SYMBOL_GPL(__serpent_setkey);
443 
444 int serpent_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
445 {
446 	return __serpent_setkey(crypto_tfm_ctx(tfm), key, keylen);
447 }
448 EXPORT_SYMBOL_GPL(serpent_setkey);
449 
450 void __serpent_encrypt(struct serpent_ctx *ctx, u8 *dst, const u8 *src)
451 {
452 	const u32 *k = ctx->expkey;
453 	const __le32 *s = (const __le32 *)src;
454 	__le32	*d = (__le32 *)dst;
455 	u32	r0, r1, r2, r3, r4;
456 
457 /*
458  * Note: The conversions between u8* and u32* might cause trouble
459  * on architectures with stricter alignment rules than x86
460  */
461 
462 	r0 = le32_to_cpu(s[0]);
463 	r1 = le32_to_cpu(s[1]);
464 	r2 = le32_to_cpu(s[2]);
465 	r3 = le32_to_cpu(s[3]);
466 
467 					K(r0, r1, r2, r3, 0);
468 	S0(r0, r1, r2, r3, r4);		LK(r2, r1, r3, r0, r4, 1);
469 	S1(r2, r1, r3, r0, r4);		LK(r4, r3, r0, r2, r1, 2);
470 	S2(r4, r3, r0, r2, r1);		LK(r1, r3, r4, r2, r0, 3);
471 	S3(r1, r3, r4, r2, r0);		LK(r2, r0, r3, r1, r4, 4);
472 	S4(r2, r0, r3, r1, r4);		LK(r0, r3, r1, r4, r2, 5);
473 	S5(r0, r3, r1, r4, r2);		LK(r2, r0, r3, r4, r1, 6);
474 	S6(r2, r0, r3, r4, r1);		LK(r3, r1, r0, r4, r2, 7);
475 	S7(r3, r1, r0, r4, r2);		LK(r2, r0, r4, r3, r1, 8);
476 	S0(r2, r0, r4, r3, r1);		LK(r4, r0, r3, r2, r1, 9);
477 	S1(r4, r0, r3, r2, r1);		LK(r1, r3, r2, r4, r0, 10);
478 	S2(r1, r3, r2, r4, r0);		LK(r0, r3, r1, r4, r2, 11);
479 	S3(r0, r3, r1, r4, r2);		LK(r4, r2, r3, r0, r1, 12);
480 	S4(r4, r2, r3, r0, r1);		LK(r2, r3, r0, r1, r4, 13);
481 	S5(r2, r3, r0, r1, r4);		LK(r4, r2, r3, r1, r0, 14);
482 	S6(r4, r2, r3, r1, r0);		LK(r3, r0, r2, r1, r4, 15);
483 	S7(r3, r0, r2, r1, r4);		LK(r4, r2, r1, r3, r0, 16);
484 	S0(r4, r2, r1, r3, r0);		LK(r1, r2, r3, r4, r0, 17);
485 	S1(r1, r2, r3, r4, r0);		LK(r0, r3, r4, r1, r2, 18);
486 	S2(r0, r3, r4, r1, r2);		LK(r2, r3, r0, r1, r4, 19);
487 	S3(r2, r3, r0, r1, r4);		LK(r1, r4, r3, r2, r0, 20);
488 	S4(r1, r4, r3, r2, r0);		LK(r4, r3, r2, r0, r1, 21);
489 	S5(r4, r3, r2, r0, r1);		LK(r1, r4, r3, r0, r2, 22);
490 	S6(r1, r4, r3, r0, r2);		LK(r3, r2, r4, r0, r1, 23);
491 	S7(r3, r2, r4, r0, r1);		LK(r1, r4, r0, r3, r2, 24);
492 	S0(r1, r4, r0, r3, r2);		LK(r0, r4, r3, r1, r2, 25);
493 	S1(r0, r4, r3, r1, r2);		LK(r2, r3, r1, r0, r4, 26);
494 	S2(r2, r3, r1, r0, r4);		LK(r4, r3, r2, r0, r1, 27);
495 	S3(r4, r3, r2, r0, r1);		LK(r0, r1, r3, r4, r2, 28);
496 	S4(r0, r1, r3, r4, r2);		LK(r1, r3, r4, r2, r0, 29);
497 	S5(r1, r3, r4, r2, r0);		LK(r0, r1, r3, r2, r4, 30);
498 	S6(r0, r1, r3, r2, r4);		LK(r3, r4, r1, r2, r0, 31);
499 	S7(r3, r4, r1, r2, r0);		K(r0, r1, r2, r3, 32);
500 
501 	d[0] = cpu_to_le32(r0);
502 	d[1] = cpu_to_le32(r1);
503 	d[2] = cpu_to_le32(r2);
504 	d[3] = cpu_to_le32(r3);
505 }
506 EXPORT_SYMBOL_GPL(__serpent_encrypt);
507 
508 static void serpent_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
509 {
510 	struct serpent_ctx *ctx = crypto_tfm_ctx(tfm);
511 
512 	__serpent_encrypt(ctx, dst, src);
513 }
514 
515 void __serpent_decrypt(struct serpent_ctx *ctx, u8 *dst, const u8 *src)
516 {
517 	const u32 *k = ctx->expkey;
518 	const __le32 *s = (const __le32 *)src;
519 	__le32	*d = (__le32 *)dst;
520 	u32	r0, r1, r2, r3, r4;
521 
522 	r0 = le32_to_cpu(s[0]);
523 	r1 = le32_to_cpu(s[1]);
524 	r2 = le32_to_cpu(s[2]);
525 	r3 = le32_to_cpu(s[3]);
526 
527 					K(r0, r1, r2, r3, 32);
528 	SI7(r0, r1, r2, r3, r4);	KL(r1, r3, r0, r4, r2, 31);
529 	SI6(r1, r3, r0, r4, r2);	KL(r0, r2, r4, r1, r3, 30);
530 	SI5(r0, r2, r4, r1, r3);	KL(r2, r3, r0, r4, r1, 29);
531 	SI4(r2, r3, r0, r4, r1);	KL(r2, r0, r1, r4, r3, 28);
532 	SI3(r2, r0, r1, r4, r3);	KL(r1, r2, r3, r4, r0, 27);
533 	SI2(r1, r2, r3, r4, r0);	KL(r2, r0, r4, r3, r1, 26);
534 	SI1(r2, r0, r4, r3, r1);	KL(r1, r0, r4, r3, r2, 25);
535 	SI0(r1, r0, r4, r3, r2);	KL(r4, r2, r0, r1, r3, 24);
536 	SI7(r4, r2, r0, r1, r3);	KL(r2, r1, r4, r3, r0, 23);
537 	SI6(r2, r1, r4, r3, r0);	KL(r4, r0, r3, r2, r1, 22);
538 	SI5(r4, r0, r3, r2, r1);	KL(r0, r1, r4, r3, r2, 21);
539 	SI4(r0, r1, r4, r3, r2);	KL(r0, r4, r2, r3, r1, 20);
540 	SI3(r0, r4, r2, r3, r1);	KL(r2, r0, r1, r3, r4, 19);
541 	SI2(r2, r0, r1, r3, r4);	KL(r0, r4, r3, r1, r2, 18);
542 	SI1(r0, r4, r3, r1, r2);	KL(r2, r4, r3, r1, r0, 17);
543 	SI0(r2, r4, r3, r1, r0);	KL(r3, r0, r4, r2, r1, 16);
544 	SI7(r3, r0, r4, r2, r1);	KL(r0, r2, r3, r1, r4, 15);
545 	SI6(r0, r2, r3, r1, r4);	KL(r3, r4, r1, r0, r2, 14);
546 	SI5(r3, r4, r1, r0, r2);	KL(r4, r2, r3, r1, r0, 13);
547 	SI4(r4, r2, r3, r1, r0);	KL(r4, r3, r0, r1, r2, 12);
548 	SI3(r4, r3, r0, r1, r2);	KL(r0, r4, r2, r1, r3, 11);
549 	SI2(r0, r4, r2, r1, r3);	KL(r4, r3, r1, r2, r0, 10);
550 	SI1(r4, r3, r1, r2, r0);	KL(r0, r3, r1, r2, r4, 9);
551 	SI0(r0, r3, r1, r2, r4);	KL(r1, r4, r3, r0, r2, 8);
552 	SI7(r1, r4, r3, r0, r2);	KL(r4, r0, r1, r2, r3, 7);
553 	SI6(r4, r0, r1, r2, r3);	KL(r1, r3, r2, r4, r0, 6);
554 	SI5(r1, r3, r2, r4, r0);	KL(r3, r0, r1, r2, r4, 5);
555 	SI4(r3, r0, r1, r2, r4);	KL(r3, r1, r4, r2, r0, 4);
556 	SI3(r3, r1, r4, r2, r0);	KL(r4, r3, r0, r2, r1, 3);
557 	SI2(r4, r3, r0, r2, r1);	KL(r3, r1, r2, r0, r4, 2);
558 	SI1(r3, r1, r2, r0, r4);	KL(r4, r1, r2, r0, r3, 1);
559 	SI0(r4, r1, r2, r0, r3);	K(r2, r3, r1, r4, 0);
560 
561 	d[0] = cpu_to_le32(r2);
562 	d[1] = cpu_to_le32(r3);
563 	d[2] = cpu_to_le32(r1);
564 	d[3] = cpu_to_le32(r4);
565 }
566 EXPORT_SYMBOL_GPL(__serpent_decrypt);
567 
568 static void serpent_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
569 {
570 	struct serpent_ctx *ctx = crypto_tfm_ctx(tfm);
571 
572 	__serpent_decrypt(ctx, dst, src);
573 }
574 
575 static int tnepres_setkey(struct crypto_tfm *tfm, const u8 *key,
576 			  unsigned int keylen)
577 {
578 	u8 rev_key[SERPENT_MAX_KEY_SIZE];
579 	int i;
580 
581 	for (i = 0; i < keylen; ++i)
582 		rev_key[keylen - i - 1] = key[i];
583 
584 	return serpent_setkey(tfm, rev_key, keylen);
585 }
586 
587 static void tnepres_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
588 {
589 	const u32 * const s = (const u32 * const)src;
590 	u32 * const d = (u32 * const)dst;
591 
592 	u32 rs[4], rd[4];
593 
594 	rs[0] = swab32(s[3]);
595 	rs[1] = swab32(s[2]);
596 	rs[2] = swab32(s[1]);
597 	rs[3] = swab32(s[0]);
598 
599 	serpent_encrypt(tfm, (u8 *)rd, (u8 *)rs);
600 
601 	d[0] = swab32(rd[3]);
602 	d[1] = swab32(rd[2]);
603 	d[2] = swab32(rd[1]);
604 	d[3] = swab32(rd[0]);
605 }
606 
607 static void tnepres_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
608 {
609 	const u32 * const s = (const u32 * const)src;
610 	u32 * const d = (u32 * const)dst;
611 
612 	u32 rs[4], rd[4];
613 
614 	rs[0] = swab32(s[3]);
615 	rs[1] = swab32(s[2]);
616 	rs[2] = swab32(s[1]);
617 	rs[3] = swab32(s[0]);
618 
619 	serpent_decrypt(tfm, (u8 *)rd, (u8 *)rs);
620 
621 	d[0] = swab32(rd[3]);
622 	d[1] = swab32(rd[2]);
623 	d[2] = swab32(rd[1]);
624 	d[3] = swab32(rd[0]);
625 }
626 
627 static struct crypto_alg srp_algs[2] = { {
628 	.cra_name		=	"serpent",
629 	.cra_driver_name	=	"serpent-generic",
630 	.cra_priority		=	100,
631 	.cra_flags		=	CRYPTO_ALG_TYPE_CIPHER,
632 	.cra_blocksize		=	SERPENT_BLOCK_SIZE,
633 	.cra_ctxsize		=	sizeof(struct serpent_ctx),
634 	.cra_alignmask		=	3,
635 	.cra_module		=	THIS_MODULE,
636 	.cra_u			=	{ .cipher = {
637 	.cia_min_keysize	=	SERPENT_MIN_KEY_SIZE,
638 	.cia_max_keysize	=	SERPENT_MAX_KEY_SIZE,
639 	.cia_setkey		=	serpent_setkey,
640 	.cia_encrypt		=	serpent_encrypt,
641 	.cia_decrypt		=	serpent_decrypt } }
642 }, {
643 	.cra_name		=	"tnepres",
644 	.cra_flags		=	CRYPTO_ALG_TYPE_CIPHER,
645 	.cra_blocksize		=	SERPENT_BLOCK_SIZE,
646 	.cra_ctxsize		=	sizeof(struct serpent_ctx),
647 	.cra_alignmask		=	3,
648 	.cra_module		=	THIS_MODULE,
649 	.cra_u			=	{ .cipher = {
650 	.cia_min_keysize	=	SERPENT_MIN_KEY_SIZE,
651 	.cia_max_keysize	=	SERPENT_MAX_KEY_SIZE,
652 	.cia_setkey		=	tnepres_setkey,
653 	.cia_encrypt		=	tnepres_encrypt,
654 	.cia_decrypt		=	tnepres_decrypt } }
655 } };
656 
657 static int __init serpent_mod_init(void)
658 {
659 	return crypto_register_algs(srp_algs, ARRAY_SIZE(srp_algs));
660 }
661 
662 static void __exit serpent_mod_fini(void)
663 {
664 	crypto_unregister_algs(srp_algs, ARRAY_SIZE(srp_algs));
665 }
666 
667 subsys_initcall(serpent_mod_init);
668 module_exit(serpent_mod_fini);
669 
670 MODULE_LICENSE("GPL");
671 MODULE_DESCRIPTION("Serpent and tnepres (kerneli compatible serpent reversed) Cipher Algorithm");
672 MODULE_AUTHOR("Dag Arne Osvik <osvik@ii.uib.no>");
673 MODULE_ALIAS_CRYPTO("tnepres");
674 MODULE_ALIAS_CRYPTO("serpent");
675 MODULE_ALIAS_CRYPTO("serpent-generic");
676