[CRYPTO] Remove unused iv field from context structure
[linux-2.6.git] / crypto / serpent.c
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: Ruben Jesus Garcia Hernandez <ruben@ugr.es>, 18.10.2004
10  *               Based on code by hvr
11  *
12  * This program is free software; you can redistribute it and/or modify
13  * it under the terms of the GNU General Public License as published by
14  * the Free Software Foundation; either version 2 of the License, or
15  * (at your option) any later version.
16  */
17
18 #include <linux/init.h>
19 #include <linux/module.h>
20 #include <linux/errno.h>
21 #include <asm/byteorder.h>
22 #include <linux/crypto.h>
23
24 /* Key is padded to the maximum of 256 bits before round key generation.
25  * Any key length <= 256 bits (32 bytes) is allowed by the algorithm.
26  */
27
28 #define SERPENT_MIN_KEY_SIZE              0
29 #define SERPENT_MAX_KEY_SIZE             32
30 #define SERPENT_EXPKEY_WORDS            132
31 #define SERPENT_BLOCK_SIZE               16
32
33 #define PHI 0x9e3779b9UL
34
35 #define keyiter(a,b,c,d,i,j) \
36         b ^= d; b ^= c; b ^= a; b ^= PHI ^ i; b = rol32(b,11); k[j] = b;
37
38 #define loadkeys(x0,x1,x2,x3,i) \
39         x0=k[i]; x1=k[i+1]; x2=k[i+2]; x3=k[i+3];
40
41 #define storekeys(x0,x1,x2,x3,i) \
42         k[i]=x0; k[i+1]=x1; k[i+2]=x2; k[i+3]=x3;
43
44 #define K(x0,x1,x2,x3,i)                                \
45         x3 ^= k[4*(i)+3];        x2 ^= k[4*(i)+2];      \
46         x1 ^= k[4*(i)+1];        x0 ^= k[4*(i)+0];
47
48 #define LK(x0,x1,x2,x3,x4,i)                            \
49                                         x0=rol32(x0,13);\
50         x2=rol32(x2,3); x1 ^= x0;       x4  = x0 << 3;  \
51         x3 ^= x2;       x1 ^= x2;                       \
52         x1=rol32(x1,1); x3 ^= x4;                       \
53         x3=rol32(x3,7); x4  = x1;                       \
54         x0 ^= x1;       x4 <<= 7;       x2 ^= x3;       \
55         x0 ^= x3;       x2 ^= x4;       x3 ^= k[4*i+3]; \
56         x1 ^= k[4*i+1]; x0=rol32(x0,5); x2=rol32(x2,22);\
57         x0 ^= k[4*i+0]; x2 ^= k[4*i+2];
58
59 #define KL(x0,x1,x2,x3,x4,i)                            \
60         x0 ^= k[4*i+0]; x1 ^= k[4*i+1]; x2 ^= k[4*i+2]; \
61         x3 ^= k[4*i+3]; x0=ror32(x0,5); x2=ror32(x2,22);\
62         x4 =  x1;       x2 ^= x3;       x0 ^= x3;       \
63         x4 <<= 7;       x0 ^= x1;       x1=ror32(x1,1); \
64         x2 ^= x4;       x3=ror32(x3,7); x4 = x0 << 3;   \
65         x1 ^= x0;       x3 ^= x4;       x0=ror32(x0,13);\
66         x1 ^= x2;       x3 ^= x2;       x2=ror32(x2,3);
67
68 #define S0(x0,x1,x2,x3,x4)                              \
69                                         x4  = x3;       \
70         x3 |= x0;       x0 ^= x4;       x4 ^= x2;       \
71         x4 =~ x4;       x3 ^= x1;       x1 &= x0;       \
72         x1 ^= x4;       x2 ^= x0;       x0 ^= x3;       \
73         x4 |= x0;       x0 ^= x2;       x2 &= x1;       \
74         x3 ^= x2;       x1 =~ x1;       x2 ^= x4;       \
75         x1 ^= x2;
76
77 #define S1(x0,x1,x2,x3,x4)                              \
78                                         x4  = x1;       \
79         x1 ^= x0;       x0 ^= x3;       x3 =~ x3;       \
80         x4 &= x1;       x0 |= x1;       x3 ^= x2;       \
81         x0 ^= x3;       x1 ^= x3;       x3 ^= x4;       \
82         x1 |= x4;       x4 ^= x2;       x2 &= x0;       \
83         x2 ^= x1;       x1 |= x0;       x0 =~ x0;       \
84         x0 ^= x2;       x4 ^= x1;
85
86 #define S2(x0,x1,x2,x3,x4)                              \
87                                         x3 =~ x3;       \
88         x1 ^= x0;       x4  = x0;       x0 &= x2;       \
89         x0 ^= x3;       x3 |= x4;       x2 ^= x1;       \
90         x3 ^= x1;       x1 &= x0;       x0 ^= x2;       \
91         x2 &= x3;       x3 |= x1;       x0 =~ x0;       \
92         x3 ^= x0;       x4 ^= x0;       x0 ^= x2;       \
93         x1 |= x2;
94
95 #define S3(x0,x1,x2,x3,x4)                              \
96                                         x4  = x1;       \
97         x1 ^= x3;       x3 |= x0;       x4 &= x0;       \
98         x0 ^= x2;       x2 ^= x1;       x1 &= x3;       \
99         x2 ^= x3;       x0 |= x4;       x4 ^= x3;       \
100         x1 ^= x0;       x0 &= x3;       x3 &= x4;       \
101         x3 ^= x2;       x4 |= x1;       x2 &= x1;       \
102         x4 ^= x3;       x0 ^= x3;       x3 ^= x2;
103
104 #define S4(x0,x1,x2,x3,x4)                              \
105                                         x4  = x3;       \
106         x3 &= x0;       x0 ^= x4;                       \
107         x3 ^= x2;       x2 |= x4;       x0 ^= x1;       \
108         x4 ^= x3;       x2 |= x0;                       \
109         x2 ^= x1;       x1 &= x0;                       \
110         x1 ^= x4;       x4 &= x2;       x2 ^= x3;       \
111         x4 ^= x0;       x3 |= x1;       x1 =~ x1;       \
112         x3 ^= x0;
113
114 #define S5(x0,x1,x2,x3,x4)                              \
115         x4  = x1;       x1 |= x0;                       \
116         x2 ^= x1;       x3 =~ x3;       x4 ^= x0;       \
117         x0 ^= x2;       x1 &= x4;       x4 |= x3;       \
118         x4 ^= x0;       x0 &= x3;       x1 ^= x3;       \
119         x3 ^= x2;       x0 ^= x1;       x2 &= x4;       \
120         x1 ^= x2;       x2 &= x0;                       \
121         x3 ^= x2;
122
123 #define S6(x0,x1,x2,x3,x4)                              \
124                                         x4  = x1;       \
125         x3 ^= x0;       x1 ^= x2;       x2 ^= x0;       \
126         x0 &= x3;       x1 |= x3;       x4 =~ x4;       \
127         x0 ^= x1;       x1 ^= x2;                       \
128         x3 ^= x4;       x4 ^= x0;       x2 &= x0;       \
129         x4 ^= x1;       x2 ^= x3;       x3 &= x1;       \
130         x3 ^= x0;       x1 ^= x2;
131
132 #define S7(x0,x1,x2,x3,x4)                              \
133                                         x1 =~ x1;       \
134         x4  = x1;       x0 =~ x0;       x1 &= x2;       \
135         x1 ^= x3;       x3 |= x4;       x4 ^= x2;       \
136         x2 ^= x3;       x3 ^= x0;       x0 |= x1;       \
137         x2 &= x0;       x0 ^= x4;       x4 ^= x3;       \
138         x3 &= x0;       x4 ^= x1;                       \
139         x2 ^= x4;       x3 ^= x1;       x4 |= x0;       \
140         x4 ^= x1;
141
142 #define SI0(x0,x1,x2,x3,x4)                             \
143                         x4  = x3;       x1 ^= x0;       \
144         x3 |= x1;       x4 ^= x1;       x0 =~ x0;       \
145         x2 ^= x3;       x3 ^= x0;       x0 &= x1;       \
146         x0 ^= x2;       x2 &= x3;       x3 ^= x4;       \
147         x2 ^= x3;       x1 ^= x3;       x3 &= x0;       \
148         x1 ^= x0;       x0 ^= x2;       x4 ^= x3;
149
150 #define SI1(x0,x1,x2,x3,x4)                             \
151         x1 ^= x3;       x4  = x0;                       \
152         x0 ^= x2;       x2 =~ x2;       x4 |= x1;       \
153         x4 ^= x3;       x3 &= x1;       x1 ^= x2;       \
154         x2 &= x4;       x4 ^= x1;       x1 |= x3;       \
155         x3 ^= x0;       x2 ^= x0;       x0 |= x4;       \
156         x2 ^= x4;       x1 ^= x0;                       \
157         x4 ^= x1;
158
159 #define SI2(x0,x1,x2,x3,x4)                             \
160         x2 ^= x1;       x4  = x3;       x3 =~ x3;       \
161         x3 |= x2;       x2 ^= x4;       x4 ^= x0;       \
162         x3 ^= x1;       x1 |= x2;       x2 ^= x0;       \
163         x1 ^= x4;       x4 |= x3;       x2 ^= x3;       \
164         x4 ^= x2;       x2 &= x1;                       \
165         x2 ^= x3;       x3 ^= x4;       x4 ^= x0;
166
167 #define SI3(x0,x1,x2,x3,x4)                             \
168                                         x2 ^= x1;       \
169         x4  = x1;       x1 &= x2;                       \
170         x1 ^= x0;       x0 |= x4;       x4 ^= x3;       \
171         x0 ^= x3;       x3 |= x1;       x1 ^= x2;       \
172         x1 ^= x3;       x0 ^= x2;       x2 ^= x3;       \
173         x3 &= x1;       x1 ^= x0;       x0 &= x2;       \
174         x4 ^= x3;       x3 ^= x0;       x0 ^= x1;
175
176 #define SI4(x0,x1,x2,x3,x4)                             \
177         x2 ^= x3;       x4  = x0;       x0 &= x1;       \
178         x0 ^= x2;       x2 |= x3;       x4 =~ x4;       \
179         x1 ^= x0;       x0 ^= x2;       x2 &= x4;       \
180         x2 ^= x0;       x0 |= x4;                       \
181         x0 ^= x3;       x3 &= x2;                       \
182         x4 ^= x3;       x3 ^= x1;       x1 &= x0;       \
183         x4 ^= x1;       x0 ^= x3;
184
185 #define SI5(x0,x1,x2,x3,x4)                             \
186                         x4  = x1;       x1 |= x2;       \
187         x2 ^= x4;       x1 ^= x3;       x3 &= x4;       \
188         x2 ^= x3;       x3 |= x0;       x0 =~ x0;       \
189         x3 ^= x2;       x2 |= x0;       x4 ^= x1;       \
190         x2 ^= x4;       x4 &= x0;       x0 ^= x1;       \
191         x1 ^= x3;       x0 &= x2;       x2 ^= x3;       \
192         x0 ^= x2;       x2 ^= x4;       x4 ^= x3;
193
194 #define SI6(x0,x1,x2,x3,x4)                             \
195                         x0 ^= x2;                       \
196         x4  = x0;       x0 &= x3;       x2 ^= x3;       \
197         x0 ^= x2;       x3 ^= x1;       x2 |= x4;       \
198         x2 ^= x3;       x3 &= x0;       x0 =~ x0;       \
199         x3 ^= x1;       x1 &= x2;       x4 ^= x0;       \
200         x3 ^= x4;       x4 ^= x2;       x0 ^= x1;       \
201         x2 ^= x0;
202
203 #define SI7(x0,x1,x2,x3,x4)                             \
204         x4  = x3;       x3 &= x0;       x0 ^= x2;       \
205         x2 |= x4;       x4 ^= x1;       x0 =~ x0;       \
206         x1 |= x3;       x4 ^= x0;       x0 &= x2;       \
207         x0 ^= x1;       x1 &= x2;       x3 ^= x2;       \
208         x4 ^= x3;       x2 &= x3;       x3 |= x0;       \
209         x1 ^= x4;       x3 ^= x4;       x4 &= x0;       \
210         x4 ^= x2;
211
212 struct serpent_ctx {
213         u32 expkey[SERPENT_EXPKEY_WORDS];
214 };
215
216
217 static int serpent_setkey(void *ctx, const u8 *key, unsigned int keylen, u32 *flags)
218 {
219         u32 *k = ((struct serpent_ctx *)ctx)->expkey;
220         u8  *k8 = (u8 *)k;
221         u32 r0,r1,r2,r3,r4;
222         int i;
223
224         if ((keylen < SERPENT_MIN_KEY_SIZE)
225                         || (keylen > SERPENT_MAX_KEY_SIZE))
226         {
227                 *flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
228                 return -EINVAL;
229         }
230
231         /* Copy key, add padding */
232
233         for (i = 0; i < keylen; ++i)
234                 k8[i] = key[i];
235         if (i < SERPENT_MAX_KEY_SIZE)
236                 k8[i++] = 1;
237         while (i < SERPENT_MAX_KEY_SIZE)
238                 k8[i++] = 0;
239
240         /* Expand key using polynomial */
241
242         r0 = le32_to_cpu(k[3]);
243         r1 = le32_to_cpu(k[4]);
244         r2 = le32_to_cpu(k[5]);
245         r3 = le32_to_cpu(k[6]);
246         r4 = le32_to_cpu(k[7]);
247
248         keyiter(le32_to_cpu(k[0]),r0,r4,r2,0,0);
249         keyiter(le32_to_cpu(k[1]),r1,r0,r3,1,1);
250         keyiter(le32_to_cpu(k[2]),r2,r1,r4,2,2);
251         keyiter(le32_to_cpu(k[3]),r3,r2,r0,3,3);
252         keyiter(le32_to_cpu(k[4]),r4,r3,r1,4,4);
253         keyiter(le32_to_cpu(k[5]),r0,r4,r2,5,5);
254         keyiter(le32_to_cpu(k[6]),r1,r0,r3,6,6);
255         keyiter(le32_to_cpu(k[7]),r2,r1,r4,7,7);
256
257         keyiter(k[  0],r3,r2,r0,  8,  8); keyiter(k[  1],r4,r3,r1,  9,  9);
258         keyiter(k[  2],r0,r4,r2, 10, 10); keyiter(k[  3],r1,r0,r3, 11, 11);
259         keyiter(k[  4],r2,r1,r4, 12, 12); keyiter(k[  5],r3,r2,r0, 13, 13);
260         keyiter(k[  6],r4,r3,r1, 14, 14); keyiter(k[  7],r0,r4,r2, 15, 15);
261         keyiter(k[  8],r1,r0,r3, 16, 16); keyiter(k[  9],r2,r1,r4, 17, 17);
262         keyiter(k[ 10],r3,r2,r0, 18, 18); keyiter(k[ 11],r4,r3,r1, 19, 19);
263         keyiter(k[ 12],r0,r4,r2, 20, 20); keyiter(k[ 13],r1,r0,r3, 21, 21);
264         keyiter(k[ 14],r2,r1,r4, 22, 22); keyiter(k[ 15],r3,r2,r0, 23, 23);
265         keyiter(k[ 16],r4,r3,r1, 24, 24); keyiter(k[ 17],r0,r4,r2, 25, 25);
266         keyiter(k[ 18],r1,r0,r3, 26, 26); keyiter(k[ 19],r2,r1,r4, 27, 27);
267         keyiter(k[ 20],r3,r2,r0, 28, 28); keyiter(k[ 21],r4,r3,r1, 29, 29);
268         keyiter(k[ 22],r0,r4,r2, 30, 30); keyiter(k[ 23],r1,r0,r3, 31, 31);
269
270         k += 50;
271
272         keyiter(k[-26],r2,r1,r4, 32,-18); keyiter(k[-25],r3,r2,r0, 33,-17);
273         keyiter(k[-24],r4,r3,r1, 34,-16); keyiter(k[-23],r0,r4,r2, 35,-15);
274         keyiter(k[-22],r1,r0,r3, 36,-14); keyiter(k[-21],r2,r1,r4, 37,-13);
275         keyiter(k[-20],r3,r2,r0, 38,-12); keyiter(k[-19],r4,r3,r1, 39,-11);
276         keyiter(k[-18],r0,r4,r2, 40,-10); keyiter(k[-17],r1,r0,r3, 41, -9);
277         keyiter(k[-16],r2,r1,r4, 42, -8); keyiter(k[-15],r3,r2,r0, 43, -7);
278         keyiter(k[-14],r4,r3,r1, 44, -6); keyiter(k[-13],r0,r4,r2, 45, -5);
279         keyiter(k[-12],r1,r0,r3, 46, -4); keyiter(k[-11],r2,r1,r4, 47, -3);
280         keyiter(k[-10],r3,r2,r0, 48, -2); keyiter(k[ -9],r4,r3,r1, 49, -1);
281         keyiter(k[ -8],r0,r4,r2, 50,  0); keyiter(k[ -7],r1,r0,r3, 51,  1);
282         keyiter(k[ -6],r2,r1,r4, 52,  2); keyiter(k[ -5],r3,r2,r0, 53,  3);
283         keyiter(k[ -4],r4,r3,r1, 54,  4); keyiter(k[ -3],r0,r4,r2, 55,  5);
284         keyiter(k[ -2],r1,r0,r3, 56,  6); keyiter(k[ -1],r2,r1,r4, 57,  7);
285         keyiter(k[  0],r3,r2,r0, 58,  8); keyiter(k[  1],r4,r3,r1, 59,  9);
286         keyiter(k[  2],r0,r4,r2, 60, 10); keyiter(k[  3],r1,r0,r3, 61, 11);
287         keyiter(k[  4],r2,r1,r4, 62, 12); keyiter(k[  5],r3,r2,r0, 63, 13);
288         keyiter(k[  6],r4,r3,r1, 64, 14); keyiter(k[  7],r0,r4,r2, 65, 15);
289         keyiter(k[  8],r1,r0,r3, 66, 16); keyiter(k[  9],r2,r1,r4, 67, 17);
290         keyiter(k[ 10],r3,r2,r0, 68, 18); keyiter(k[ 11],r4,r3,r1, 69, 19);
291         keyiter(k[ 12],r0,r4,r2, 70, 20); keyiter(k[ 13],r1,r0,r3, 71, 21);
292         keyiter(k[ 14],r2,r1,r4, 72, 22); keyiter(k[ 15],r3,r2,r0, 73, 23);
293         keyiter(k[ 16],r4,r3,r1, 74, 24); keyiter(k[ 17],r0,r4,r2, 75, 25);
294         keyiter(k[ 18],r1,r0,r3, 76, 26); keyiter(k[ 19],r2,r1,r4, 77, 27);
295         keyiter(k[ 20],r3,r2,r0, 78, 28); keyiter(k[ 21],r4,r3,r1, 79, 29);
296         keyiter(k[ 22],r0,r4,r2, 80, 30); keyiter(k[ 23],r1,r0,r3, 81, 31);
297
298         k += 50;
299
300         keyiter(k[-26],r2,r1,r4, 82,-18); keyiter(k[-25],r3,r2,r0, 83,-17);
301         keyiter(k[-24],r4,r3,r1, 84,-16); keyiter(k[-23],r0,r4,r2, 85,-15);
302         keyiter(k[-22],r1,r0,r3, 86,-14); keyiter(k[-21],r2,r1,r4, 87,-13);
303         keyiter(k[-20],r3,r2,r0, 88,-12); keyiter(k[-19],r4,r3,r1, 89,-11);
304         keyiter(k[-18],r0,r4,r2, 90,-10); keyiter(k[-17],r1,r0,r3, 91, -9);
305         keyiter(k[-16],r2,r1,r4, 92, -8); keyiter(k[-15],r3,r2,r0, 93, -7);
306         keyiter(k[-14],r4,r3,r1, 94, -6); keyiter(k[-13],r0,r4,r2, 95, -5);
307         keyiter(k[-12],r1,r0,r3, 96, -4); keyiter(k[-11],r2,r1,r4, 97, -3);
308         keyiter(k[-10],r3,r2,r0, 98, -2); keyiter(k[ -9],r4,r3,r1, 99, -1);
309         keyiter(k[ -8],r0,r4,r2,100,  0); keyiter(k[ -7],r1,r0,r3,101,  1);
310         keyiter(k[ -6],r2,r1,r4,102,  2); keyiter(k[ -5],r3,r2,r0,103,  3);
311         keyiter(k[ -4],r4,r3,r1,104,  4); keyiter(k[ -3],r0,r4,r2,105,  5);
312         keyiter(k[ -2],r1,r0,r3,106,  6); keyiter(k[ -1],r2,r1,r4,107,  7);
313         keyiter(k[  0],r3,r2,r0,108,  8); keyiter(k[  1],r4,r3,r1,109,  9);
314         keyiter(k[  2],r0,r4,r2,110, 10); keyiter(k[  3],r1,r0,r3,111, 11);
315         keyiter(k[  4],r2,r1,r4,112, 12); keyiter(k[  5],r3,r2,r0,113, 13);
316         keyiter(k[  6],r4,r3,r1,114, 14); keyiter(k[  7],r0,r4,r2,115, 15);
317         keyiter(k[  8],r1,r0,r3,116, 16); keyiter(k[  9],r2,r1,r4,117, 17);
318         keyiter(k[ 10],r3,r2,r0,118, 18); keyiter(k[ 11],r4,r3,r1,119, 19);
319         keyiter(k[ 12],r0,r4,r2,120, 20); keyiter(k[ 13],r1,r0,r3,121, 21);
320         keyiter(k[ 14],r2,r1,r4,122, 22); keyiter(k[ 15],r3,r2,r0,123, 23);
321         keyiter(k[ 16],r4,r3,r1,124, 24); keyiter(k[ 17],r0,r4,r2,125, 25);
322         keyiter(k[ 18],r1,r0,r3,126, 26); keyiter(k[ 19],r2,r1,r4,127, 27);
323         keyiter(k[ 20],r3,r2,r0,128, 28); keyiter(k[ 21],r4,r3,r1,129, 29);
324         keyiter(k[ 22],r0,r4,r2,130, 30); keyiter(k[ 23],r1,r0,r3,131, 31);
325
326         /* Apply S-boxes */
327
328         S3(r3,r4,r0,r1,r2); storekeys(r1,r2,r4,r3, 28); loadkeys(r1,r2,r4,r3, 24);
329         S4(r1,r2,r4,r3,r0); storekeys(r2,r4,r3,r0, 24); loadkeys(r2,r4,r3,r0, 20);
330         S5(r2,r4,r3,r0,r1); storekeys(r1,r2,r4,r0, 20); loadkeys(r1,r2,r4,r0, 16);
331         S6(r1,r2,r4,r0,r3); storekeys(r4,r3,r2,r0, 16); loadkeys(r4,r3,r2,r0, 12);
332         S7(r4,r3,r2,r0,r1); storekeys(r1,r2,r0,r4, 12); loadkeys(r1,r2,r0,r4,  8);
333         S0(r1,r2,r0,r4,r3); storekeys(r0,r2,r4,r1,  8); loadkeys(r0,r2,r4,r1,  4);
334         S1(r0,r2,r4,r1,r3); storekeys(r3,r4,r1,r0,  4); loadkeys(r3,r4,r1,r0,  0);
335         S2(r3,r4,r1,r0,r2); storekeys(r2,r4,r3,r0,  0); loadkeys(r2,r4,r3,r0, -4);
336         S3(r2,r4,r3,r0,r1); storekeys(r0,r1,r4,r2, -4); loadkeys(r0,r1,r4,r2, -8);
337         S4(r0,r1,r4,r2,r3); storekeys(r1,r4,r2,r3, -8); loadkeys(r1,r4,r2,r3,-12);
338         S5(r1,r4,r2,r3,r0); storekeys(r0,r1,r4,r3,-12); loadkeys(r0,r1,r4,r3,-16);
339         S6(r0,r1,r4,r3,r2); storekeys(r4,r2,r1,r3,-16); loadkeys(r4,r2,r1,r3,-20);
340         S7(r4,r2,r1,r3,r0); storekeys(r0,r1,r3,r4,-20); loadkeys(r0,r1,r3,r4,-24);
341         S0(r0,r1,r3,r4,r2); storekeys(r3,r1,r4,r0,-24); loadkeys(r3,r1,r4,r0,-28);
342         k -= 50;
343         S1(r3,r1,r4,r0,r2); storekeys(r2,r4,r0,r3, 22); loadkeys(r2,r4,r0,r3, 18);
344         S2(r2,r4,r0,r3,r1); storekeys(r1,r4,r2,r3, 18); loadkeys(r1,r4,r2,r3, 14);
345         S3(r1,r4,r2,r3,r0); storekeys(r3,r0,r4,r1, 14); loadkeys(r3,r0,r4,r1, 10);
346         S4(r3,r0,r4,r1,r2); storekeys(r0,r4,r1,r2, 10); loadkeys(r0,r4,r1,r2,  6);
347         S5(r0,r4,r1,r2,r3); storekeys(r3,r0,r4,r2,  6); loadkeys(r3,r0,r4,r2,  2);
348         S6(r3,r0,r4,r2,r1); storekeys(r4,r1,r0,r2,  2); loadkeys(r4,r1,r0,r2, -2);
349         S7(r4,r1,r0,r2,r3); storekeys(r3,r0,r2,r4, -2); loadkeys(r3,r0,r2,r4, -6);
350         S0(r3,r0,r2,r4,r1); storekeys(r2,r0,r4,r3, -6); loadkeys(r2,r0,r4,r3,-10);
351         S1(r2,r0,r4,r3,r1); storekeys(r1,r4,r3,r2,-10); loadkeys(r1,r4,r3,r2,-14);
352         S2(r1,r4,r3,r2,r0); storekeys(r0,r4,r1,r2,-14); loadkeys(r0,r4,r1,r2,-18);
353         S3(r0,r4,r1,r2,r3); storekeys(r2,r3,r4,r0,-18); loadkeys(r2,r3,r4,r0,-22);
354         k -= 50;
355         S4(r2,r3,r4,r0,r1); storekeys(r3,r4,r0,r1, 28); loadkeys(r3,r4,r0,r1, 24);
356         S5(r3,r4,r0,r1,r2); storekeys(r2,r3,r4,r1, 24); loadkeys(r2,r3,r4,r1, 20);
357         S6(r2,r3,r4,r1,r0); storekeys(r4,r0,r3,r1, 20); loadkeys(r4,r0,r3,r1, 16);
358         S7(r4,r0,r3,r1,r2); storekeys(r2,r3,r1,r4, 16); loadkeys(r2,r3,r1,r4, 12);
359         S0(r2,r3,r1,r4,r0); storekeys(r1,r3,r4,r2, 12); loadkeys(r1,r3,r4,r2,  8);
360         S1(r1,r3,r4,r2,r0); storekeys(r0,r4,r2,r1,  8); loadkeys(r0,r4,r2,r1,  4);
361         S2(r0,r4,r2,r1,r3); storekeys(r3,r4,r0,r1,  4); loadkeys(r3,r4,r0,r1,  0);
362         S3(r3,r4,r0,r1,r2); storekeys(r1,r2,r4,r3,  0);
363
364         return 0;
365 }
366
367 static void serpent_encrypt(void *ctx, u8 *dst, const u8 *src)
368 {
369         const u32
370                 *k = ((struct serpent_ctx *)ctx)->expkey,
371                 *s = (const u32 *)src;
372         u32     *d = (u32 *)dst,
373                 r0, r1, r2, r3, r4;
374
375 /*
376  * Note: The conversions between u8* and u32* might cause trouble
377  * on architectures with stricter alignment rules than x86
378  */
379
380         r0 = le32_to_cpu(s[0]);
381         r1 = le32_to_cpu(s[1]);
382         r2 = le32_to_cpu(s[2]);
383         r3 = le32_to_cpu(s[3]);
384
385                                  K(r0,r1,r2,r3,0);
386         S0(r0,r1,r2,r3,r4);     LK(r2,r1,r3,r0,r4,1);
387         S1(r2,r1,r3,r0,r4);     LK(r4,r3,r0,r2,r1,2);
388         S2(r4,r3,r0,r2,r1);     LK(r1,r3,r4,r2,r0,3);
389         S3(r1,r3,r4,r2,r0);     LK(r2,r0,r3,r1,r4,4);
390         S4(r2,r0,r3,r1,r4);     LK(r0,r3,r1,r4,r2,5);
391         S5(r0,r3,r1,r4,r2);     LK(r2,r0,r3,r4,r1,6);
392         S6(r2,r0,r3,r4,r1);     LK(r3,r1,r0,r4,r2,7);
393         S7(r3,r1,r0,r4,r2);     LK(r2,r0,r4,r3,r1,8);
394         S0(r2,r0,r4,r3,r1);     LK(r4,r0,r3,r2,r1,9);
395         S1(r4,r0,r3,r2,r1);     LK(r1,r3,r2,r4,r0,10);
396         S2(r1,r3,r2,r4,r0);     LK(r0,r3,r1,r4,r2,11);
397         S3(r0,r3,r1,r4,r2);     LK(r4,r2,r3,r0,r1,12);
398         S4(r4,r2,r3,r0,r1);     LK(r2,r3,r0,r1,r4,13);
399         S5(r2,r3,r0,r1,r4);     LK(r4,r2,r3,r1,r0,14);
400         S6(r4,r2,r3,r1,r0);     LK(r3,r0,r2,r1,r4,15);
401         S7(r3,r0,r2,r1,r4);     LK(r4,r2,r1,r3,r0,16);
402         S0(r4,r2,r1,r3,r0);     LK(r1,r2,r3,r4,r0,17);
403         S1(r1,r2,r3,r4,r0);     LK(r0,r3,r4,r1,r2,18);
404         S2(r0,r3,r4,r1,r2);     LK(r2,r3,r0,r1,r4,19);
405         S3(r2,r3,r0,r1,r4);     LK(r1,r4,r3,r2,r0,20);
406         S4(r1,r4,r3,r2,r0);     LK(r4,r3,r2,r0,r1,21);
407         S5(r4,r3,r2,r0,r1);     LK(r1,r4,r3,r0,r2,22);
408         S6(r1,r4,r3,r0,r2);     LK(r3,r2,r4,r0,r1,23);
409         S7(r3,r2,r4,r0,r1);     LK(r1,r4,r0,r3,r2,24);
410         S0(r1,r4,r0,r3,r2);     LK(r0,r4,r3,r1,r2,25);
411         S1(r0,r4,r3,r1,r2);     LK(r2,r3,r1,r0,r4,26);
412         S2(r2,r3,r1,r0,r4);     LK(r4,r3,r2,r0,r1,27);
413         S3(r4,r3,r2,r0,r1);     LK(r0,r1,r3,r4,r2,28);
414         S4(r0,r1,r3,r4,r2);     LK(r1,r3,r4,r2,r0,29);
415         S5(r1,r3,r4,r2,r0);     LK(r0,r1,r3,r2,r4,30);
416         S6(r0,r1,r3,r2,r4);     LK(r3,r4,r1,r2,r0,31);
417         S7(r3,r4,r1,r2,r0);      K(r0,r1,r2,r3,32);
418
419         d[0] = cpu_to_le32(r0);
420         d[1] = cpu_to_le32(r1);
421         d[2] = cpu_to_le32(r2);
422         d[3] = cpu_to_le32(r3);
423 }
424
425 static void serpent_decrypt(void *ctx, u8 *dst, const u8 *src)
426 {
427         const u32
428                 *k = ((struct serpent_ctx *)ctx)->expkey,
429                 *s = (const u32 *)src;
430         u32     *d = (u32 *)dst,
431                 r0, r1, r2, r3, r4;
432
433         r0 = le32_to_cpu(s[0]);
434         r1 = le32_to_cpu(s[1]);
435         r2 = le32_to_cpu(s[2]);
436         r3 = le32_to_cpu(s[3]);
437
438                                 K(r0,r1,r2,r3,32);
439         SI7(r0,r1,r2,r3,r4);    KL(r1,r3,r0,r4,r2,31);
440         SI6(r1,r3,r0,r4,r2);    KL(r0,r2,r4,r1,r3,30);
441         SI5(r0,r2,r4,r1,r3);    KL(r2,r3,r0,r4,r1,29);
442         SI4(r2,r3,r0,r4,r1);    KL(r2,r0,r1,r4,r3,28);
443         SI3(r2,r0,r1,r4,r3);    KL(r1,r2,r3,r4,r0,27);
444         SI2(r1,r2,r3,r4,r0);    KL(r2,r0,r4,r3,r1,26);
445         SI1(r2,r0,r4,r3,r1);    KL(r1,r0,r4,r3,r2,25);
446         SI0(r1,r0,r4,r3,r2);    KL(r4,r2,r0,r1,r3,24);
447         SI7(r4,r2,r0,r1,r3);    KL(r2,r1,r4,r3,r0,23);
448         SI6(r2,r1,r4,r3,r0);    KL(r4,r0,r3,r2,r1,22);
449         SI5(r4,r0,r3,r2,r1);    KL(r0,r1,r4,r3,r2,21);
450         SI4(r0,r1,r4,r3,r2);    KL(r0,r4,r2,r3,r1,20);
451         SI3(r0,r4,r2,r3,r1);    KL(r2,r0,r1,r3,r4,19);
452         SI2(r2,r0,r1,r3,r4);    KL(r0,r4,r3,r1,r2,18);
453         SI1(r0,r4,r3,r1,r2);    KL(r2,r4,r3,r1,r0,17);
454         SI0(r2,r4,r3,r1,r0);    KL(r3,r0,r4,r2,r1,16);
455         SI7(r3,r0,r4,r2,r1);    KL(r0,r2,r3,r1,r4,15);
456         SI6(r0,r2,r3,r1,r4);    KL(r3,r4,r1,r0,r2,14);
457         SI5(r3,r4,r1,r0,r2);    KL(r4,r2,r3,r1,r0,13);
458         SI4(r4,r2,r3,r1,r0);    KL(r4,r3,r0,r1,r2,12);
459         SI3(r4,r3,r0,r1,r2);    KL(r0,r4,r2,r1,r3,11);
460         SI2(r0,r4,r2,r1,r3);    KL(r4,r3,r1,r2,r0,10);
461         SI1(r4,r3,r1,r2,r0);    KL(r0,r3,r1,r2,r4,9);
462         SI0(r0,r3,r1,r2,r4);    KL(r1,r4,r3,r0,r2,8);
463         SI7(r1,r4,r3,r0,r2);    KL(r4,r0,r1,r2,r3,7);
464         SI6(r4,r0,r1,r2,r3);    KL(r1,r3,r2,r4,r0,6);
465         SI5(r1,r3,r2,r4,r0);    KL(r3,r0,r1,r2,r4,5);
466         SI4(r3,r0,r1,r2,r4);    KL(r3,r1,r4,r2,r0,4);
467         SI3(r3,r1,r4,r2,r0);    KL(r4,r3,r0,r2,r1,3);
468         SI2(r4,r3,r0,r2,r1);    KL(r3,r1,r2,r0,r4,2);
469         SI1(r3,r1,r2,r0,r4);    KL(r4,r1,r2,r0,r3,1);
470         SI0(r4,r1,r2,r0,r3);    K(r2,r3,r1,r4,0);
471
472         d[0] = cpu_to_le32(r2);
473         d[1] = cpu_to_le32(r3);
474         d[2] = cpu_to_le32(r1);
475         d[3] = cpu_to_le32(r4);
476 }
477
478 static struct crypto_alg serpent_alg = {
479         .cra_name               =       "serpent",
480         .cra_flags              =       CRYPTO_ALG_TYPE_CIPHER,
481         .cra_blocksize          =       SERPENT_BLOCK_SIZE,
482         .cra_ctxsize            =       sizeof(struct serpent_ctx),
483         .cra_module             =       THIS_MODULE,
484         .cra_list               =       LIST_HEAD_INIT(serpent_alg.cra_list),
485         .cra_u                  =       { .cipher = {
486         .cia_min_keysize        =       SERPENT_MIN_KEY_SIZE,
487         .cia_max_keysize        =       SERPENT_MAX_KEY_SIZE,
488         .cia_setkey             =       serpent_setkey,
489         .cia_encrypt            =       serpent_encrypt,
490         .cia_decrypt            =       serpent_decrypt } }
491 };
492
493 static int tnepres_setkey(void *ctx, const u8 *key, unsigned int keylen, u32 *flags)
494 {
495         u8 rev_key[SERPENT_MAX_KEY_SIZE];
496         int i;
497
498         if ((keylen < SERPENT_MIN_KEY_SIZE)
499             || (keylen > SERPENT_MAX_KEY_SIZE)) {
500                 *flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
501                 return -EINVAL;
502         } 
503
504         for (i = 0; i < keylen; ++i)
505                 rev_key[keylen - i - 1] = key[i];
506  
507         return serpent_setkey(ctx, rev_key, keylen, flags);
508 }
509
510 static void tnepres_encrypt(void *ctx, u8 *dst, const u8 *src)
511 {
512         const u32 * const s = (const u32 * const)src;
513         u32 * const d = (u32 * const)dst;
514
515         u32 rs[4], rd[4];
516
517         rs[0] = swab32(s[3]);
518         rs[1] = swab32(s[2]);
519         rs[2] = swab32(s[1]);
520         rs[3] = swab32(s[0]);
521
522         serpent_encrypt(ctx, (u8 *)rd, (u8 *)rs);
523
524         d[0] = swab32(rd[3]);
525         d[1] = swab32(rd[2]);
526         d[2] = swab32(rd[1]);
527         d[3] = swab32(rd[0]);
528 }
529
530 static void tnepres_decrypt(void *ctx, u8 *dst, const u8 *src)
531 {
532         const u32 * const s = (const u32 * const)src;
533         u32 * const d = (u32 * const)dst;
534
535         u32 rs[4], rd[4];
536
537         rs[0] = swab32(s[3]);
538         rs[1] = swab32(s[2]);
539         rs[2] = swab32(s[1]);
540         rs[3] = swab32(s[0]);
541
542         serpent_decrypt(ctx, (u8 *)rd, (u8 *)rs);
543
544         d[0] = swab32(rd[3]);
545         d[1] = swab32(rd[2]);
546         d[2] = swab32(rd[1]);
547         d[3] = swab32(rd[0]);
548 }
549
550 static struct crypto_alg tnepres_alg = {
551         .cra_name               =       "tnepres",
552         .cra_flags              =       CRYPTO_ALG_TYPE_CIPHER,
553         .cra_blocksize          =       SERPENT_BLOCK_SIZE,
554         .cra_ctxsize            =       sizeof(struct serpent_ctx),
555         .cra_module             =       THIS_MODULE,
556         .cra_list               =       LIST_HEAD_INIT(serpent_alg.cra_list),
557         .cra_u                  =       { .cipher = {
558         .cia_min_keysize        =       SERPENT_MIN_KEY_SIZE,
559         .cia_max_keysize        =       SERPENT_MAX_KEY_SIZE,
560         .cia_setkey             =       tnepres_setkey,
561         .cia_encrypt            =       tnepres_encrypt,
562         .cia_decrypt            =       tnepres_decrypt } }
563 };
564
565 static int __init init(void)
566 {
567         int ret = crypto_register_alg(&serpent_alg);
568
569         if (ret)
570                 return ret;
571
572         ret = crypto_register_alg(&tnepres_alg);
573
574         if (ret)
575                 crypto_unregister_alg(&serpent_alg);
576
577         return ret;
578 }
579
580 static void __exit fini(void)
581 {
582         crypto_unregister_alg(&tnepres_alg);
583         crypto_unregister_alg(&serpent_alg);
584 }
585
586 module_init(init);
587 module_exit(fini);
588
589 MODULE_LICENSE("GPL");
590 MODULE_DESCRIPTION("Serpent and tnepres (kerneli compatible serpent reversed) Cipher Algorithm");
591 MODULE_AUTHOR("Dag Arne Osvik <osvik@ii.uib.no>");
592 MODULE_ALIAS("tnepres");