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
232int __serpent_setkey(struct serpent_ctx *ctx, const u8 *key,
233		     unsigned int keylen)
234{
235	u32 *k = ctx->expkey;
236	u8  *k8 = (u8 *)k;
237	u32 r0, r1, r2, r3, r4;
238	int i;
239
240	/* Copy key, add padding */
241
242	for (i = 0; i < keylen; ++i)
243		k8[i] = key[i];
244	if (i < SERPENT_MAX_KEY_SIZE)
245		k8[i++] = 1;
246	while (i < SERPENT_MAX_KEY_SIZE)
247		k8[i++] = 0;
248
249	/* Expand key using polynomial */
250
251	r0 = le32_to_cpu(k[3]);
252	r1 = le32_to_cpu(k[4]);
253	r2 = le32_to_cpu(k[5]);
254	r3 = le32_to_cpu(k[6]);
255	r4 = le32_to_cpu(k[7]);
256
257	keyiter(le32_to_cpu(k[0]), r0, r4, r2, 0, 0);
258	keyiter(le32_to_cpu(k[1]), r1, r0, r3, 1, 1);
259	keyiter(le32_to_cpu(k[2]), r2, r1, r4, 2, 2);
260	keyiter(le32_to_cpu(k[3]), r3, r2, r0, 3, 3);
261	keyiter(le32_to_cpu(k[4]), r4, r3, r1, 4, 4);
262	keyiter(le32_to_cpu(k[5]), r0, r4, r2, 5, 5);
263	keyiter(le32_to_cpu(k[6]), r1, r0, r3, 6, 6);
264	keyiter(le32_to_cpu(k[7]), r2, r1, r4, 7, 7);
265
266	keyiter(k[0], r3, r2, r0, 8, 8);
267	keyiter(k[1], r4, r3, r1, 9, 9);
268	keyiter(k[2], r0, r4, r2, 10, 10);
269	keyiter(k[3], r1, r0, r3, 11, 11);
270	keyiter(k[4], r2, r1, r4, 12, 12);
271	keyiter(k[5], r3, r2, r0, 13, 13);
272	keyiter(k[6], r4, r3, r1, 14, 14);
273	keyiter(k[7], r0, r4, r2, 15, 15);
274	keyiter(k[8], r1, r0, r3, 16, 16);
275	keyiter(k[9], r2, r1, r4, 17, 17);
276	keyiter(k[10], r3, r2, r0, 18, 18);
277	keyiter(k[11], r4, r3, r1, 19, 19);
278	keyiter(k[12], r0, r4, r2, 20, 20);
279	keyiter(k[13], r1, r0, r3, 21, 21);
280	keyiter(k[14], r2, r1, r4, 22, 22);
281	keyiter(k[15], r3, r2, r0, 23, 23);
282	keyiter(k[16], r4, r3, r1, 24, 24);
283	keyiter(k[17], r0, r4, r2, 25, 25);
284	keyiter(k[18], r1, r0, r3, 26, 26);
285	keyiter(k[19], r2, r1, r4, 27, 27);
286	keyiter(k[20], r3, r2, r0, 28, 28);
287	keyiter(k[21], r4, r3, r1, 29, 29);
288	keyiter(k[22], r0, r4, r2, 30, 30);
289	keyiter(k[23], r1, r0, r3, 31, 31);
290
291	k += 50;
292
293	keyiter(k[-26], r2, r1, r4, 32, -18);
294	keyiter(k[-25], r3, r2, r0, 33, -17);
295	keyiter(k[-24], r4, r3, r1, 34, -16);
296	keyiter(k[-23], r0, r4, r2, 35, -15);
297	keyiter(k[-22], r1, r0, r3, 36, -14);
298	keyiter(k[-21], r2, r1, r4, 37, -13);
299	keyiter(k[-20], r3, r2, r0, 38, -12);
300	keyiter(k[-19], r4, r3, r1, 39, -11);
301	keyiter(k[-18], r0, r4, r2, 40, -10);
302	keyiter(k[-17], r1, r0, r3, 41, -9);
303	keyiter(k[-16], r2, r1, r4, 42, -8);
304	keyiter(k[-15], r3, r2, r0, 43, -7);
305	keyiter(k[-14], r4, r3, r1, 44, -6);
306	keyiter(k[-13], r0, r4, r2, 45, -5);
307	keyiter(k[-12], r1, r0, r3, 46, -4);
308	keyiter(k[-11], r2, r1, r4, 47, -3);
309	keyiter(k[-10], r3, r2, r0, 48, -2);
310	keyiter(k[-9], r4, r3, r1, 49, -1);
311	keyiter(k[-8], r0, r4, r2, 50, 0);
312	keyiter(k[-7], r1, r0, r3, 51, 1);
313	keyiter(k[-6], r2, r1, r4, 52, 2);
314	keyiter(k[-5], r3, r2, r0, 53, 3);
315	keyiter(k[-4], r4, r3, r1, 54, 4);
316	keyiter(k[-3], r0, r4, r2, 55, 5);
317	keyiter(k[-2], r1, r0, r3, 56, 6);
318	keyiter(k[-1], r2, r1, r4, 57, 7);
319	keyiter(k[0], r3, r2, r0, 58, 8);
320	keyiter(k[1], r4, r3, r1, 59, 9);
321	keyiter(k[2], r0, r4, r2, 60, 10);
322	keyiter(k[3], r1, r0, r3, 61, 11);
323	keyiter(k[4], r2, r1, r4, 62, 12);
324	keyiter(k[5], r3, r2, r0, 63, 13);
325	keyiter(k[6], r4, r3, r1, 64, 14);
326	keyiter(k[7], r0, r4, r2, 65, 15);
327	keyiter(k[8], r1, r0, r3, 66, 16);
328	keyiter(k[9], r2, r1, r4, 67, 17);
329	keyiter(k[10], r3, r2, r0, 68, 18);
330	keyiter(k[11], r4, r3, r1, 69, 19);
331	keyiter(k[12], r0, r4, r2, 70, 20);
332	keyiter(k[13], r1, r0, r3, 71, 21);
333	keyiter(k[14], r2, r1, r4, 72, 22);
334	keyiter(k[15], r3, r2, r0, 73, 23);
335	keyiter(k[16], r4, r3, r1, 74, 24);
336	keyiter(k[17], r0, r4, r2, 75, 25);
337	keyiter(k[18], r1, r0, r3, 76, 26);
338	keyiter(k[19], r2, r1, r4, 77, 27);
339	keyiter(k[20], r3, r2, r0, 78, 28);
340	keyiter(k[21], r4, r3, r1, 79, 29);
341	keyiter(k[22], r0, r4, r2, 80, 30);
342	keyiter(k[23], r1, r0, r3, 81, 31);
343
344	k += 50;
345
346	keyiter(k[-26], r2, r1, r4, 82, -18);
347	keyiter(k[-25], r3, r2, r0, 83, -17);
348	keyiter(k[-24], r4, r3, r1, 84, -16);
349	keyiter(k[-23], r0, r4, r2, 85, -15);
350	keyiter(k[-22], r1, r0, r3, 86, -14);
351	keyiter(k[-21], r2, r1, r4, 87, -13);
352	keyiter(k[-20], r3, r2, r0, 88, -12);
353	keyiter(k[-19], r4, r3, r1, 89, -11);
354	keyiter(k[-18], r0, r4, r2, 90, -10);
355	keyiter(k[-17], r1, r0, r3, 91, -9);
356	keyiter(k[-16], r2, r1, r4, 92, -8);
357	keyiter(k[-15], r3, r2, r0, 93, -7);
358	keyiter(k[-14], r4, r3, r1, 94, -6);
359	keyiter(k[-13], r0, r4, r2, 95, -5);
360	keyiter(k[-12], r1, r0, r3, 96, -4);
361	keyiter(k[-11], r2, r1, r4, 97, -3);
362	keyiter(k[-10], r3, r2, r0, 98, -2);
363	keyiter(k[-9], r4, r3, r1, 99, -1);
364	keyiter(k[-8], r0, r4, r2, 100, 0);
365	keyiter(k[-7], r1, r0, r3, 101, 1);
366	keyiter(k[-6], r2, r1, r4, 102, 2);
367	keyiter(k[-5], r3, r2, r0, 103, 3);
368	keyiter(k[-4], r4, r3, r1, 104, 4);
369	keyiter(k[-3], r0, r4, r2, 105, 5);
370	keyiter(k[-2], r1, r0, r3, 106, 6);
371	keyiter(k[-1], r2, r1, r4, 107, 7);
372	keyiter(k[0], r3, r2, r0, 108, 8);
373	keyiter(k[1], r4, r3, r1, 109, 9);
374	keyiter(k[2], r0, r4, r2, 110, 10);
375	keyiter(k[3], r1, r0, r3, 111, 11);
376	keyiter(k[4], r2, r1, r4, 112, 12);
377	keyiter(k[5], r3, r2, r0, 113, 13);
378	keyiter(k[6], r4, r3, r1, 114, 14);
379	keyiter(k[7], r0, r4, r2, 115, 15);
380	keyiter(k[8], r1, r0, r3, 116, 16);
381	keyiter(k[9], r2, r1, r4, 117, 17);
382	keyiter(k[10], r3, r2, r0, 118, 18);
383	keyiter(k[11], r4, r3, r1, 119, 19);
384	keyiter(k[12], r0, r4, r2, 120, 20);
385	keyiter(k[13], r1, r0, r3, 121, 21);
386	keyiter(k[14], r2, r1, r4, 122, 22);
387	keyiter(k[15], r3, r2, r0, 123, 23);
388	keyiter(k[16], r4, r3, r1, 124, 24);
389	keyiter(k[17], r0, r4, r2, 125, 25);
390	keyiter(k[18], r1, r0, r3, 126, 26);
391	keyiter(k[19], r2, r1, r4, 127, 27);
392	keyiter(k[20], r3, r2, r0, 128, 28);
393	keyiter(k[21], r4, r3, r1, 129, 29);
394	keyiter(k[22], r0, r4, r2, 130, 30);
395	keyiter(k[23], r1, r0, r3, 131, 31);
396
397	/* Apply S-boxes */
398
399	S3(r3, r4, r0, r1, r2); store_and_load_keys(r1, r2, r4, r3, 28, 24);
400	S4(r1, r2, r4, r3, r0); store_and_load_keys(r2, r4, r3, r0, 24, 20);
401	S5(r2, r4, r3, r0, r1); store_and_load_keys(r1, r2, r4, r0, 20, 16);
402	S6(r1, r2, r4, r0, r3); store_and_load_keys(r4, r3, r2, r0, 16, 12);
403	S7(r4, r3, r2, r0, r1); store_and_load_keys(r1, r2, r0, r4, 12, 8);
404	S0(r1, r2, r0, r4, r3); store_and_load_keys(r0, r2, r4, r1, 8, 4);
405	S1(r0, r2, r4, r1, r3); store_and_load_keys(r3, r4, r1, r0, 4, 0);
406	S2(r3, r4, r1, r0, r2); store_and_load_keys(r2, r4, r3, r0, 0, -4);
407	S3(r2, r4, r3, r0, r1); store_and_load_keys(r0, r1, r4, r2, -4, -8);
408	S4(r0, r1, r4, r2, r3); store_and_load_keys(r1, r4, r2, r3, -8, -12);
409	S5(r1, r4, r2, r3, r0); store_and_load_keys(r0, r1, r4, r3, -12, -16);
410	S6(r0, r1, r4, r3, r2); store_and_load_keys(r4, r2, r1, r3, -16, -20);
411	S7(r4, r2, r1, r3, r0); store_and_load_keys(r0, r1, r3, r4, -20, -24);
412	S0(r0, r1, r3, r4, r2); store_and_load_keys(r3, r1, r4, r0, -24, -28);
413	k -= 50;
414	S1(r3, r1, r4, r0, r2); store_and_load_keys(r2, r4, r0, r3, 22, 18);
415	S2(r2, r4, r0, r3, r1); store_and_load_keys(r1, r4, r2, r3, 18, 14);
416	S3(r1, r4, r2, r3, r0); store_and_load_keys(r3, r0, r4, r1, 14, 10);
417	S4(r3, r0, r4, r1, r2); store_and_load_keys(r0, r4, r1, r2, 10, 6);
418	S5(r0, r4, r1, r2, r3); store_and_load_keys(r3, r0, r4, r2, 6, 2);
419	S6(r3, r0, r4, r2, r1); store_and_load_keys(r4, r1, r0, r2, 2, -2);
420	S7(r4, r1, r0, r2, r3); store_and_load_keys(r3, r0, r2, r4, -2, -6);
421	S0(r3, r0, r2, r4, r1); store_and_load_keys(r2, r0, r4, r3, -6, -10);
422	S1(r2, r0, r4, r3, r1); store_and_load_keys(r1, r4, r3, r2, -10, -14);
423	S2(r1, r4, r3, r2, r0); store_and_load_keys(r0, r4, r1, r2, -14, -18);
424	S3(r0, r4, r1, r2, r3); store_and_load_keys(r2, r3, r4, r0, -18, -22);
425	k -= 50;
426	S4(r2, r3, r4, r0, r1); store_and_load_keys(r3, r4, r0, r1, 28, 24);
427	S5(r3, r4, r0, r1, r2); store_and_load_keys(r2, r3, r4, r1, 24, 20);
428	S6(r2, r3, r4, r1, r0); store_and_load_keys(r4, r0, r3, r1, 20, 16);
429	S7(r4, r0, r3, r1, r2); store_and_load_keys(r2, r3, r1, r4, 16, 12);
430	S0(r2, r3, r1, r4, r0); store_and_load_keys(r1, r3, r4, r2, 12, 8);
431	S1(r1, r3, r4, r2, r0); store_and_load_keys(r0, r4, r2, r1, 8, 4);
432	S2(r0, r4, r2, r1, r3); store_and_load_keys(r3, r4, r0, r1, 4, 0);
433	S3(r3, r4, r0, r1, r2); storekeys(r1, r2, r4, r3, 0);
434
435	return 0;
436}
437EXPORT_SYMBOL_GPL(__serpent_setkey);
438
439int serpent_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
440{
441	return __serpent_setkey(crypto_tfm_ctx(tfm), key, keylen);
442}
443EXPORT_SYMBOL_GPL(serpent_setkey);
444
445void __serpent_encrypt(struct serpent_ctx *ctx, u8 *dst, const u8 *src)
446{
447	const u32 *k = ctx->expkey;
448	const __le32 *s = (const __le32 *)src;
449	__le32	*d = (__le32 *)dst;
450	u32	r0, r1, r2, r3, r4;
451
452/*
453 * Note: The conversions between u8* and u32* might cause trouble
454 * on architectures with stricter alignment rules than x86
455 */
456
457	r0 = le32_to_cpu(s[0]);
458	r1 = le32_to_cpu(s[1]);
459	r2 = le32_to_cpu(s[2]);
460	r3 = le32_to_cpu(s[3]);
461
462					K(r0, r1, r2, r3, 0);
463	S0(r0, r1, r2, r3, r4);		LK(r2, r1, r3, r0, r4, 1);
464	S1(r2, r1, r3, r0, r4);		LK(r4, r3, r0, r2, r1, 2);
465	S2(r4, r3, r0, r2, r1);		LK(r1, r3, r4, r2, r0, 3);
466	S3(r1, r3, r4, r2, r0);		LK(r2, r0, r3, r1, r4, 4);
467	S4(r2, r0, r3, r1, r4);		LK(r0, r3, r1, r4, r2, 5);
468	S5(r0, r3, r1, r4, r2);		LK(r2, r0, r3, r4, r1, 6);
469	S6(r2, r0, r3, r4, r1);		LK(r3, r1, r0, r4, r2, 7);
470	S7(r3, r1, r0, r4, r2);		LK(r2, r0, r4, r3, r1, 8);
471	S0(r2, r0, r4, r3, r1);		LK(r4, r0, r3, r2, r1, 9);
472	S1(r4, r0, r3, r2, r1);		LK(r1, r3, r2, r4, r0, 10);
473	S2(r1, r3, r2, r4, r0);		LK(r0, r3, r1, r4, r2, 11);
474	S3(r0, r3, r1, r4, r2);		LK(r4, r2, r3, r0, r1, 12);
475	S4(r4, r2, r3, r0, r1);		LK(r2, r3, r0, r1, r4, 13);
476	S5(r2, r3, r0, r1, r4);		LK(r4, r2, r3, r1, r0, 14);
477	S6(r4, r2, r3, r1, r0);		LK(r3, r0, r2, r1, r4, 15);
478	S7(r3, r0, r2, r1, r4);		LK(r4, r2, r1, r3, r0, 16);
479	S0(r4, r2, r1, r3, r0);		LK(r1, r2, r3, r4, r0, 17);
480	S1(r1, r2, r3, r4, r0);		LK(r0, r3, r4, r1, r2, 18);
481	S2(r0, r3, r4, r1, r2);		LK(r2, r3, r0, r1, r4, 19);
482	S3(r2, r3, r0, r1, r4);		LK(r1, r4, r3, r2, r0, 20);
483	S4(r1, r4, r3, r2, r0);		LK(r4, r3, r2, r0, r1, 21);
484	S5(r4, r3, r2, r0, r1);		LK(r1, r4, r3, r0, r2, 22);
485	S6(r1, r4, r3, r0, r2);		LK(r3, r2, r4, r0, r1, 23);
486	S7(r3, r2, r4, r0, r1);		LK(r1, r4, r0, r3, r2, 24);
487	S0(r1, r4, r0, r3, r2);		LK(r0, r4, r3, r1, r2, 25);
488	S1(r0, r4, r3, r1, r2);		LK(r2, r3, r1, r0, r4, 26);
489	S2(r2, r3, r1, r0, r4);		LK(r4, r3, r2, r0, r1, 27);
490	S3(r4, r3, r2, r0, r1);		LK(r0, r1, r3, r4, r2, 28);
491	S4(r0, r1, r3, r4, r2);		LK(r1, r3, r4, r2, r0, 29);
492	S5(r1, r3, r4, r2, r0);		LK(r0, r1, r3, r2, r4, 30);
493	S6(r0, r1, r3, r2, r4);		LK(r3, r4, r1, r2, r0, 31);
494	S7(r3, r4, r1, r2, r0);		K(r0, r1, r2, r3, 32);
495
496	d[0] = cpu_to_le32(r0);
497	d[1] = cpu_to_le32(r1);
498	d[2] = cpu_to_le32(r2);
499	d[3] = cpu_to_le32(r3);
500}
501EXPORT_SYMBOL_GPL(__serpent_encrypt);
502
503static void serpent_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
504{
505	struct serpent_ctx *ctx = crypto_tfm_ctx(tfm);
506
507	__serpent_encrypt(ctx, dst, src);
508}
509
510void __serpent_decrypt(struct serpent_ctx *ctx, u8 *dst, const u8 *src)
511{
512	const u32 *k = ctx->expkey;
513	const __le32 *s = (const __le32 *)src;
514	__le32	*d = (__le32 *)dst;
515	u32	r0, r1, r2, r3, r4;
516
517	r0 = le32_to_cpu(s[0]);
518	r1 = le32_to_cpu(s[1]);
519	r2 = le32_to_cpu(s[2]);
520	r3 = le32_to_cpu(s[3]);
521
522					K(r0, r1, r2, r3, 32);
523	SI7(r0, r1, r2, r3, r4);	KL(r1, r3, r0, r4, r2, 31);
524	SI6(r1, r3, r0, r4, r2);	KL(r0, r2, r4, r1, r3, 30);
525	SI5(r0, r2, r4, r1, r3);	KL(r2, r3, r0, r4, r1, 29);
526	SI4(r2, r3, r0, r4, r1);	KL(r2, r0, r1, r4, r3, 28);
527	SI3(r2, r0, r1, r4, r3);	KL(r1, r2, r3, r4, r0, 27);
528	SI2(r1, r2, r3, r4, r0);	KL(r2, r0, r4, r3, r1, 26);
529	SI1(r2, r0, r4, r3, r1);	KL(r1, r0, r4, r3, r2, 25);
530	SI0(r1, r0, r4, r3, r2);	KL(r4, r2, r0, r1, r3, 24);
531	SI7(r4, r2, r0, r1, r3);	KL(r2, r1, r4, r3, r0, 23);
532	SI6(r2, r1, r4, r3, r0);	KL(r4, r0, r3, r2, r1, 22);
533	SI5(r4, r0, r3, r2, r1);	KL(r0, r1, r4, r3, r2, 21);
534	SI4(r0, r1, r4, r3, r2);	KL(r0, r4, r2, r3, r1, 20);
535	SI3(r0, r4, r2, r3, r1);	KL(r2, r0, r1, r3, r4, 19);
536	SI2(r2, r0, r1, r3, r4);	KL(r0, r4, r3, r1, r2, 18);
537	SI1(r0, r4, r3, r1, r2);	KL(r2, r4, r3, r1, r0, 17);
538	SI0(r2, r4, r3, r1, r0);	KL(r3, r0, r4, r2, r1, 16);
539	SI7(r3, r0, r4, r2, r1);	KL(r0, r2, r3, r1, r4, 15);
540	SI6(r0, r2, r3, r1, r4);	KL(r3, r4, r1, r0, r2, 14);
541	SI5(r3, r4, r1, r0, r2);	KL(r4, r2, r3, r1, r0, 13);
542	SI4(r4, r2, r3, r1, r0);	KL(r4, r3, r0, r1, r2, 12);
543	SI3(r4, r3, r0, r1, r2);	KL(r0, r4, r2, r1, r3, 11);
544	SI2(r0, r4, r2, r1, r3);	KL(r4, r3, r1, r2, r0, 10);
545	SI1(r4, r3, r1, r2, r0);	KL(r0, r3, r1, r2, r4, 9);
546	SI0(r0, r3, r1, r2, r4);	KL(r1, r4, r3, r0, r2, 8);
547	SI7(r1, r4, r3, r0, r2);	KL(r4, r0, r1, r2, r3, 7);
548	SI6(r4, r0, r1, r2, r3);	KL(r1, r3, r2, r4, r0, 6);
549	SI5(r1, r3, r2, r4, r0);	KL(r3, r0, r1, r2, r4, 5);
550	SI4(r3, r0, r1, r2, r4);	KL(r3, r1, r4, r2, r0, 4);
551	SI3(r3, r1, r4, r2, r0);	KL(r4, r3, r0, r2, r1, 3);
552	SI2(r4, r3, r0, r2, r1);	KL(r3, r1, r2, r0, r4, 2);
553	SI1(r3, r1, r2, r0, r4);	KL(r4, r1, r2, r0, r3, 1);
554	SI0(r4, r1, r2, r0, r3);	K(r2, r3, r1, r4, 0);
555
556	d[0] = cpu_to_le32(r2);
557	d[1] = cpu_to_le32(r3);
558	d[2] = cpu_to_le32(r1);
559	d[3] = cpu_to_le32(r4);
560}
561EXPORT_SYMBOL_GPL(__serpent_decrypt);
562
563static void serpent_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
564{
565	struct serpent_ctx *ctx = crypto_tfm_ctx(tfm);
566
567	__serpent_decrypt(ctx, dst, src);
568}
569
570static int tnepres_setkey(struct crypto_tfm *tfm, const u8 *key,
571			  unsigned int keylen)
572{
573	u8 rev_key[SERPENT_MAX_KEY_SIZE];
574	int i;
575
576	for (i = 0; i < keylen; ++i)
577		rev_key[keylen - i - 1] = key[i];
578
579	return serpent_setkey(tfm, rev_key, keylen);
580}
581
582static void tnepres_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
583{
584	const u32 * const s = (const u32 * const)src;
585	u32 * const d = (u32 * const)dst;
586
587	u32 rs[4], rd[4];
588
589	rs[0] = swab32(s[3]);
590	rs[1] = swab32(s[2]);
591	rs[2] = swab32(s[1]);
592	rs[3] = swab32(s[0]);
593
594	serpent_encrypt(tfm, (u8 *)rd, (u8 *)rs);
595
596	d[0] = swab32(rd[3]);
597	d[1] = swab32(rd[2]);
598	d[2] = swab32(rd[1]);
599	d[3] = swab32(rd[0]);
600}
601
602static void tnepres_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
603{
604	const u32 * const s = (const u32 * const)src;
605	u32 * const d = (u32 * const)dst;
606
607	u32 rs[4], rd[4];
608
609	rs[0] = swab32(s[3]);
610	rs[1] = swab32(s[2]);
611	rs[2] = swab32(s[1]);
612	rs[3] = swab32(s[0]);
613
614	serpent_decrypt(tfm, (u8 *)rd, (u8 *)rs);
615
616	d[0] = swab32(rd[3]);
617	d[1] = swab32(rd[2]);
618	d[2] = swab32(rd[1]);
619	d[3] = swab32(rd[0]);
620}
621
622static struct crypto_alg srp_algs[2] = { {
623	.cra_name		=	"serpent",
624	.cra_driver_name	=	"serpent-generic",
625	.cra_priority		=	100,
626	.cra_flags		=	CRYPTO_ALG_TYPE_CIPHER,
627	.cra_blocksize		=	SERPENT_BLOCK_SIZE,
628	.cra_ctxsize		=	sizeof(struct serpent_ctx),
629	.cra_alignmask		=	3,
630	.cra_module		=	THIS_MODULE,
631	.cra_u			=	{ .cipher = {
632	.cia_min_keysize	=	SERPENT_MIN_KEY_SIZE,
633	.cia_max_keysize	=	SERPENT_MAX_KEY_SIZE,
634	.cia_setkey		=	serpent_setkey,
635	.cia_encrypt		=	serpent_encrypt,
636	.cia_decrypt		=	serpent_decrypt } }
637}, {
638	.cra_name		=	"tnepres",
639	.cra_flags		=	CRYPTO_ALG_TYPE_CIPHER,
640	.cra_blocksize		=	SERPENT_BLOCK_SIZE,
641	.cra_ctxsize		=	sizeof(struct serpent_ctx),
642	.cra_alignmask		=	3,
643	.cra_module		=	THIS_MODULE,
644	.cra_u			=	{ .cipher = {
645	.cia_min_keysize	=	SERPENT_MIN_KEY_SIZE,
646	.cia_max_keysize	=	SERPENT_MAX_KEY_SIZE,
647	.cia_setkey		=	tnepres_setkey,
648	.cia_encrypt		=	tnepres_encrypt,
649	.cia_decrypt		=	tnepres_decrypt } }
650} };
651
652static int __init serpent_mod_init(void)
653{
654	return crypto_register_algs(srp_algs, ARRAY_SIZE(srp_algs));
655}
656
657static void __exit serpent_mod_fini(void)
658{
659	crypto_unregister_algs(srp_algs, ARRAY_SIZE(srp_algs));
660}
661
662module_init(serpent_mod_init);
663module_exit(serpent_mod_fini);
664
665MODULE_LICENSE("GPL");
666MODULE_DESCRIPTION("Serpent and tnepres (kerneli compatible serpent reversed) Cipher Algorithm");
667MODULE_AUTHOR("Dag Arne Osvik <osvik@ii.uib.no>");
668MODULE_ALIAS_CRYPTO("tnepres");
669MODULE_ALIAS_CRYPTO("serpent");
670MODULE_ALIAS_CRYPTO("serpent-generic");
671