Include ssh.h for the crc32() and crc32_update() prototypes.
[u/mdw/putty] / sshdes.c
CommitLineData
374330e2 1#include <assert.h>
2#include "ssh.h"
3
033b4cef 4
d1e726bc 5/* des.c - implementation of DES
6 */
7
374330e2 8/*
d1e726bc 9 * Description of DES
1608bfe3 10 * ------------------
d1e726bc 11 *
12 * Unlike the description in FIPS 46, I'm going to use _sensible_ indices:
13 * bits in an n-bit word are numbered from 0 at the LSB to n-1 at the MSB.
14 * And S-boxes are indexed by six consecutive bits, not by the outer two
15 * followed by the middle four.
16 *
17 * The DES encryption routine requires a 64-bit input, and a key schedule K
18 * containing 16 48-bit elements.
19 *
20 * First the input is permuted by the initial permutation IP.
21 * Then the input is split into 32-bit words L and R. (L is the MSW.)
22 * Next, 16 rounds. In each round:
23 * (L, R) <- (R, L xor f(R, K[i]))
24 * Then the pre-output words L and R are swapped.
25 * Then L and R are glued back together into a 64-bit word. (L is the MSW,
26 * again, but since we just swapped them, the MSW is the R that came out
27 * of the last round.)
28 * The 64-bit output block is permuted by the inverse of IP and returned.
29 *
30 * Decryption is identical except that the elements of K are used in the
31 * opposite order. (This wouldn't work if that word swap didn't happen.)
32 *
33 * The function f, used in each round, accepts a 32-bit word R and a
34 * 48-bit key block K. It produces a 32-bit output.
35 *
36 * First R is expanded to 48 bits using the bit-selection function E.
37 * The resulting 48-bit block is XORed with the key block K to produce
38 * a 48-bit block X.
39 * This block X is split into eight groups of 6 bits. Each group of 6
40 * bits is then looked up in one of the eight S-boxes to convert
41 * it to 4 bits. These eight groups of 4 bits are glued back
42 * together to produce a 32-bit preoutput block.
43 * The preoutput block is permuted using the permutation P and returned.
44 *
45 * Key setup maps a 64-bit key word into a 16x48-bit key schedule. Although
46 * the approved input format for the key is a 64-bit word, eight of the
47 * bits are discarded, so the actual quantity of key used is 56 bits.
48 *
49 * First the input key is converted to two 28-bit words C and D using
50 * the bit-selection function PC1.
51 * Then 16 rounds of key setup occur. In each round, C and D are each
52 * rotated left by either 1 or 2 bits (depending on which round), and
53 * then converted into a key schedule element using the bit-selection
54 * function PC2.
55 *
56 * That's the actual algorithm. Now for the tedious details: all those
57 * painful permutations and lookup tables.
58 *
59 * IP is a 64-to-64 bit permutation. Its output contains the following
60 * bits of its input (listed in order MSB to LSB of output).
61 *
62 * 6 14 22 30 38 46 54 62 4 12 20 28 36 44 52 60
63 * 2 10 18 26 34 42 50 58 0 8 16 24 32 40 48 56
64 * 7 15 23 31 39 47 55 63 5 13 21 29 37 45 53 61
65 * 3 11 19 27 35 43 51 59 1 9 17 25 33 41 49 57
66 *
67 * E is a 32-to-48 bit selection function. Its output contains the following
68 * bits of its input (listed in order MSB to LSB of output).
69 *
70 * 0 31 30 29 28 27 28 27 26 25 24 23 24 23 22 21 20 19 20 19 18 17 16 15
71 * 16 15 14 13 12 11 12 11 10 9 8 7 8 7 6 5 4 3 4 3 2 1 0 31
72 *
73 * The S-boxes are arbitrary table-lookups each mapping a 6-bit input to a
74 * 4-bit output. In other words, each S-box is an array[64] of 4-bit numbers.
75 * The S-boxes are listed below. The first S-box listed is applied to the
76 * most significant six bits of the block X; the last one is applied to the
77 * least significant.
78 *
79 * 14 0 4 15 13 7 1 4 2 14 15 2 11 13 8 1
80 * 3 10 10 6 6 12 12 11 5 9 9 5 0 3 7 8
81 * 4 15 1 12 14 8 8 2 13 4 6 9 2 1 11 7
82 * 15 5 12 11 9 3 7 14 3 10 10 0 5 6 0 13
83 *
84 * 15 3 1 13 8 4 14 7 6 15 11 2 3 8 4 14
85 * 9 12 7 0 2 1 13 10 12 6 0 9 5 11 10 5
86 * 0 13 14 8 7 10 11 1 10 3 4 15 13 4 1 2
87 * 5 11 8 6 12 7 6 12 9 0 3 5 2 14 15 9
88 *
89 * 10 13 0 7 9 0 14 9 6 3 3 4 15 6 5 10
90 * 1 2 13 8 12 5 7 14 11 12 4 11 2 15 8 1
91 * 13 1 6 10 4 13 9 0 8 6 15 9 3 8 0 7
92 * 11 4 1 15 2 14 12 3 5 11 10 5 14 2 7 12
93 *
94 * 7 13 13 8 14 11 3 5 0 6 6 15 9 0 10 3
95 * 1 4 2 7 8 2 5 12 11 1 12 10 4 14 15 9
96 * 10 3 6 15 9 0 0 6 12 10 11 1 7 13 13 8
97 * 15 9 1 4 3 5 14 11 5 12 2 7 8 2 4 14
98 *
99 * 2 14 12 11 4 2 1 12 7 4 10 7 11 13 6 1
100 * 8 5 5 0 3 15 15 10 13 3 0 9 14 8 9 6
101 * 4 11 2 8 1 12 11 7 10 1 13 14 7 2 8 13
102 * 15 6 9 15 12 0 5 9 6 10 3 4 0 5 14 3
103 *
104 * 12 10 1 15 10 4 15 2 9 7 2 12 6 9 8 5
105 * 0 6 13 1 3 13 4 14 14 0 7 11 5 3 11 8
106 * 9 4 14 3 15 2 5 12 2 9 8 5 12 15 3 10
107 * 7 11 0 14 4 1 10 7 1 6 13 0 11 8 6 13
108 *
109 * 4 13 11 0 2 11 14 7 15 4 0 9 8 1 13 10
110 * 3 14 12 3 9 5 7 12 5 2 10 15 6 8 1 6
111 * 1 6 4 11 11 13 13 8 12 1 3 4 7 10 14 7
112 * 10 9 15 5 6 0 8 15 0 14 5 2 9 3 2 12
113 *
114 * 13 1 2 15 8 13 4 8 6 10 15 3 11 7 1 4
115 * 10 12 9 5 3 6 14 11 5 0 0 14 12 9 7 2
116 * 7 2 11 1 4 14 1 7 9 4 12 10 14 8 2 13
117 * 0 15 6 12 10 9 13 0 15 3 3 5 5 6 8 11
118 *
119 * P is a 32-to-32 bit permutation. Its output contains the following
120 * bits of its input (listed in order MSB to LSB of output).
121 *
122 * 16 25 12 11 3 20 4 15 31 17 9 6 27 14 1 22
123 * 30 24 8 18 0 5 29 23 13 19 2 26 10 21 28 7
124 *
125 * PC1 is a 64-to-56 bit selection function. Its output is in two words,
126 * C and D. The word C contains the following bits of its input (listed
127 * in order MSB to LSB of output).
128 *
129 * 7 15 23 31 39 47 55 63 6 14 22 30 38 46
130 * 54 62 5 13 21 29 37 45 53 61 4 12 20 28
131 *
132 * And the word D contains these bits.
133 *
134 * 1 9 17 25 33 41 49 57 2 10 18 26 34 42
135 * 50 58 3 11 19 27 35 43 51 59 36 44 52 60
136 *
137 * PC2 is a 56-to-48 bit selection function. Its input is in two words,
138 * C and D. These are treated as one 56-bit word (with C more significant,
139 * so that bits 55 to 28 of the word are bits 27 to 0 of C, and bits 27 to
140 * 0 of the word are bits 27 to 0 of D). The output contains the following
141 * bits of this 56-bit input word (listed in order MSB to LSB of output).
142 *
143 * 42 39 45 32 55 51 53 28 41 50 35 46 33 37 44 52 30 48 40 49 29 36 43 54
144 * 15 4 25 19 9 1 26 16 5 11 23 8 12 7 17 0 22 3 10 14 6 20 27 24
145 */
146
1608bfe3 147/*
148 * Implementation details
149 * ----------------------
150 *
151 * If you look at the code in this module, you'll find it looks
152 * nothing _like_ the above algorithm. Here I explain the
153 * differences...
154 *
155 * Key setup has not been heavily optimised here. We are not
156 * concerned with key agility: we aren't codebreakers. We don't
157 * mind a little delay (and it really is a little one; it may be a
158 * factor of five or so slower than it could be but it's still not
159 * an appreciable length of time) while setting up. The only tweaks
160 * in the key setup are ones which change the format of the key
161 * schedule to speed up the actual encryption. I'll describe those
162 * below.
163 *
164 * The first and most obvious optimisation is the S-boxes. Since
165 * each S-box always targets the same four bits in the final 32-bit
166 * word, so the output from (for example) S-box 0 must always be
167 * shifted left 28 bits, we can store the already-shifted outputs
168 * in the lookup tables. This reduces lookup-and-shift to lookup,
169 * so the S-box step is now just a question of ORing together eight
170 * table lookups.
171 *
172 * The permutation P is just a bit order change; it's invariant
173 * with respect to OR, in that P(x)|P(y) = P(x|y). Therefore, we
174 * can apply P to every entry of the S-box tables and then we don't
175 * have to do it in the code of f(). This yields a set of tables
176 * which might be called SP-boxes.
177 *
178 * The bit-selection function E is our next target. Note that E is
179 * immediately followed by the operation of splitting into 6-bit
180 * chunks. Examining the 6-bit chunks coming out of E we notice
181 * they're all contiguous within the word (speaking cyclically -
182 * the end two wrap round); so we can extract those bit strings
183 * individually rather than explicitly running E. This would yield
184 * code such as
185 *
186 * y |= SPboxes[0][ (rotl(R, 5) ^ top6bitsofK) & 0x3F ];
187 * t |= SPboxes[1][ (rotl(R,11) ^ next6bitsofK) & 0x3F ];
188 *
189 * and so on; and the key schedule preparation would have to
190 * provide each 6-bit chunk separately.
191 *
192 * Really we'd like to XOR in the key schedule element before
193 * looking up bit strings in R. This we can't do, naively, because
194 * the 6-bit strings we want overlap. But look at the strings:
195 *
196 * 3322222222221111111111
197 * bit 10987654321098765432109876543210
198 *
199 * box0 XXXXX X
200 * box1 XXXXXX
201 * box2 XXXXXX
202 * box3 XXXXXX
203 * box4 XXXXXX
204 * box5 XXXXXX
205 * box6 XXXXXX
206 * box7 X XXXXX
207 *
208 * The bit strings we need to XOR in for boxes 0, 2, 4 and 6 don't
209 * overlap with each other. Neither do the ones for boxes 1, 3, 5
210 * and 7. So we could provide the key schedule in the form of two
211 * words that we can separately XOR into R, and then every S-box
212 * index is available as a (cyclically) contiguous 6-bit substring
213 * of one or the other of the results.
214 *
215 * The comments in Eric Young's libdes implementation point out
216 * that two of these bit strings require a rotation (rather than a
217 * simple shift) to extract. It's unavoidable that at least _one_
218 * must do; but we can actually run the whole inner algorithm (all
219 * 16 rounds) rotated one bit to the left, so that what the `real'
220 * DES description sees as L=0x80000001 we see as L=0x00000003.
221 * This requires rotating all our SP-box entries one bit to the
222 * left, and rotating each word of the key schedule elements one to
223 * the left, and rotating L and R one bit left just after IP and
224 * one bit right again just before FP. And in each round we convert
225 * a rotate into a shift, so we've saved a few per cent.
226 *
227 * That's about it for the inner loop; the SP-box tables as listed
228 * below are what I've described here (the original S value,
229 * shifted to its final place in the input to P, run through P, and
230 * then rotated one bit left). All that remains is to optimise the
231 * initial permutation IP.
232 *
233 * IP is not an arbitrary permutation. It has the nice property
234 * that if you take any bit number, write it in binary (6 bits),
235 * permute those 6 bits and invert some of them, you get the final
236 * position of that bit. Specifically, the bit whose initial
237 * position is given (in binary) as fedcba ends up in position
238 * AcbFED (where a capital letter denotes the inverse of a bit).
239 *
240 * We have the 64-bit data in two 32-bit words L and R, where bits
241 * in L are those with f=1 and bits in R are those with f=0. We
242 * note that we can do a simple transformation: suppose we exchange
243 * the bits with f=1,c=0 and the bits with f=0,c=1. This will cause
244 * the bit fedcba to be in position cedfba - we've `swapped' bits c
245 * and f in the position of each bit!
246 *
247 * Better still, this transformation is easy. In the example above,
248 * bits in L with c=0 are bits 0x0F0F0F0F, and those in R with c=1
249 * are 0xF0F0F0F0. So we can do
250 *
251 * difference = ((R >> 4) ^ L) & 0x0F0F0F0F
252 * R ^= (difference << 4)
253 * L ^= difference
254 *
255 * to perform the swap. Let's denote this by bitswap(4,0x0F0F0F0F).
256 * Also, we can invert the bit at the top just by exchanging L and
257 * R. So in a few swaps and a few of these bit operations we can
258 * do:
259 *
260 * Initially the position of bit fedcba is fedcba
261 * Swap L with R to make it Fedcba
262 * Perform bitswap( 4,0x0F0F0F0F) to make it cedFba
263 * Perform bitswap(16,0x0000FFFF) to make it ecdFba
264 * Swap L with R to make it EcdFba
265 * Perform bitswap( 2,0x33333333) to make it bcdFEa
266 * Perform bitswap( 8,0x00FF00FF) to make it dcbFEa
267 * Swap L with R to make it DcbFEa
268 * Perform bitswap( 1,0x55555555) to make it acbFED
269 * Swap L with R to make it AcbFED
270 *
271 * (In the actual code the four swaps are implicit: R and L are
272 * simply used the other way round in the first, second and last
273 * bitswap operations.)
274 *
275 * The final permutation is just the inverse of IP, so it can be
276 * performed by a similar set of operations.
277 */
278
d1e726bc 279typedef struct {
280 word32 k0246[16], k1357[16];
371e569c 281 word32 iv0, iv1;
d1e726bc 282} DESContext;
374330e2 283
d1e726bc 284#define rotl(x, c) ( (x << c) | (x >> (32-c)) )
285#define rotl28(x, c) ( ( (x << c) | (x >> (28-c)) ) & 0x0FFFFFFF)
374330e2 286
32874aea 287static word32 bitsel(word32 * input, const int *bitnums, int size)
288{
d1e726bc 289 word32 ret = 0;
290 while (size--) {
32874aea 291 int bitpos = *bitnums++;
292 ret <<= 1;
293 if (bitpos >= 0)
294 ret |= 1 & (input[bitpos / 32] >> (bitpos % 32));
d1e726bc 295 }
296 return ret;
297}
374330e2 298
32874aea 299void des_key_setup(word32 key_msw, word32 key_lsw, DESContext * sched)
300{
d1e726bc 301
302 static const int PC1_Cbits[] = {
32874aea 303 7, 15, 23, 31, 39, 47, 55, 63, 6, 14, 22, 30, 38, 46,
304 54, 62, 5, 13, 21, 29, 37, 45, 53, 61, 4, 12, 20, 28
d1e726bc 305 };
306 static const int PC1_Dbits[] = {
32874aea 307 1, 9, 17, 25, 33, 41, 49, 57, 2, 10, 18, 26, 34, 42,
308 50, 58, 3, 11, 19, 27, 35, 43, 51, 59, 36, 44, 52, 60
d1e726bc 309 };
1608bfe3 310 /*
311 * The bit numbers in the two lists below don't correspond to
312 * the ones in the above description of PC2, because in the
313 * above description C and D are concatenated so `bit 28' means
314 * bit 0 of C. In this implementation we're using the standard
315 * `bitsel' function above and C is in the second word, so bit
316 * 0 of C is addressed by writing `32' here.
317 */
d1e726bc 318 static const int PC2_0246[] = {
32874aea 319 49, 36, 59, 55, -1, -1, 37, 41, 48, 56, 34, 52, -1, -1, 15, 4,
320 25, 19, 9, 1, -1, -1, 12, 7, 17, 0, 22, 3, -1, -1, 46, 43
d1e726bc 321 };
322 static const int PC2_1357[] = {
32874aea 323 -1, -1, 57, 32, 45, 54, 39, 50, -1, -1, 44, 53, 33, 40, 47, 58,
324 -1, -1, 26, 16, 5, 11, 23, 8, -1, -1, 10, 14, 6, 20, 27, 24
d1e726bc 325 };
32874aea 326 static const int leftshifts[] =
327 { 1, 1, 2, 2, 2, 2, 2, 2, 1, 2, 2, 2, 2, 2, 2, 1 };
d1e726bc 328
329 word32 C, D;
330 word32 buf[2];
331 int i;
332
333 buf[0] = key_lsw;
334 buf[1] = key_msw;
335
336 C = bitsel(buf, PC1_Cbits, 28);
337 D = bitsel(buf, PC1_Dbits, 28);
338
339 for (i = 0; i < 16; i++) {
32874aea 340 C = rotl28(C, leftshifts[i]);
341 D = rotl28(D, leftshifts[i]);
342 buf[0] = D;
343 buf[1] = C;
344 sched->k0246[i] = bitsel(buf, PC2_0246, 32);
345 sched->k1357[i] = bitsel(buf, PC2_1357, 32);
d1e726bc 346 }
374330e2 347
371e569c 348 sched->iv0 = sched->iv1 = 0;
d1e726bc 349}
374330e2 350
d1e726bc 351static const word32 SPboxes[8][64] = {
352 {0x01010400, 0x00000000, 0x00010000, 0x01010404,
32874aea 353 0x01010004, 0x00010404, 0x00000004, 0x00010000,
354 0x00000400, 0x01010400, 0x01010404, 0x00000400,
355 0x01000404, 0x01010004, 0x01000000, 0x00000004,
356 0x00000404, 0x01000400, 0x01000400, 0x00010400,
357 0x00010400, 0x01010000, 0x01010000, 0x01000404,
358 0x00010004, 0x01000004, 0x01000004, 0x00010004,
359 0x00000000, 0x00000404, 0x00010404, 0x01000000,
360 0x00010000, 0x01010404, 0x00000004, 0x01010000,
361 0x01010400, 0x01000000, 0x01000000, 0x00000400,
362 0x01010004, 0x00010000, 0x00010400, 0x01000004,
363 0x00000400, 0x00000004, 0x01000404, 0x00010404,
364 0x01010404, 0x00010004, 0x01010000, 0x01000404,
365 0x01000004, 0x00000404, 0x00010404, 0x01010400,
366 0x00000404, 0x01000400, 0x01000400, 0x00000000,
367 0x00010004, 0x00010400, 0x00000000, 0x01010004L},
d1e726bc 368
369 {0x80108020, 0x80008000, 0x00008000, 0x00108020,
32874aea 370 0x00100000, 0x00000020, 0x80100020, 0x80008020,
371 0x80000020, 0x80108020, 0x80108000, 0x80000000,
372 0x80008000, 0x00100000, 0x00000020, 0x80100020,
373 0x00108000, 0x00100020, 0x80008020, 0x00000000,
374 0x80000000, 0x00008000, 0x00108020, 0x80100000,
375 0x00100020, 0x80000020, 0x00000000, 0x00108000,
376 0x00008020, 0x80108000, 0x80100000, 0x00008020,
377 0x00000000, 0x00108020, 0x80100020, 0x00100000,
378 0x80008020, 0x80100000, 0x80108000, 0x00008000,
379 0x80100000, 0x80008000, 0x00000020, 0x80108020,
380 0x00108020, 0x00000020, 0x00008000, 0x80000000,
381 0x00008020, 0x80108000, 0x00100000, 0x80000020,
382 0x00100020, 0x80008020, 0x80000020, 0x00100020,
383 0x00108000, 0x00000000, 0x80008000, 0x00008020,
384 0x80000000, 0x80100020, 0x80108020, 0x00108000L},
d1e726bc 385
386 {0x00000208, 0x08020200, 0x00000000, 0x08020008,
32874aea 387 0x08000200, 0x00000000, 0x00020208, 0x08000200,
388 0x00020008, 0x08000008, 0x08000008, 0x00020000,
389 0x08020208, 0x00020008, 0x08020000, 0x00000208,
390 0x08000000, 0x00000008, 0x08020200, 0x00000200,
391 0x00020200, 0x08020000, 0x08020008, 0x00020208,
392 0x08000208, 0x00020200, 0x00020000, 0x08000208,
393 0x00000008, 0x08020208, 0x00000200, 0x08000000,
394 0x08020200, 0x08000000, 0x00020008, 0x00000208,
395 0x00020000, 0x08020200, 0x08000200, 0x00000000,
396 0x00000200, 0x00020008, 0x08020208, 0x08000200,
397 0x08000008, 0x00000200, 0x00000000, 0x08020008,
398 0x08000208, 0x00020000, 0x08000000, 0x08020208,
399 0x00000008, 0x00020208, 0x00020200, 0x08000008,
400 0x08020000, 0x08000208, 0x00000208, 0x08020000,
401 0x00020208, 0x00000008, 0x08020008, 0x00020200L},
d1e726bc 402
403 {0x00802001, 0x00002081, 0x00002081, 0x00000080,
32874aea 404 0x00802080, 0x00800081, 0x00800001, 0x00002001,
405 0x00000000, 0x00802000, 0x00802000, 0x00802081,
406 0x00000081, 0x00000000, 0x00800080, 0x00800001,
407 0x00000001, 0x00002000, 0x00800000, 0x00802001,
408 0x00000080, 0x00800000, 0x00002001, 0x00002080,
409 0x00800081, 0x00000001, 0x00002080, 0x00800080,
410 0x00002000, 0x00802080, 0x00802081, 0x00000081,
411 0x00800080, 0x00800001, 0x00802000, 0x00802081,
412 0x00000081, 0x00000000, 0x00000000, 0x00802000,
413 0x00002080, 0x00800080, 0x00800081, 0x00000001,
414 0x00802001, 0x00002081, 0x00002081, 0x00000080,
415 0x00802081, 0x00000081, 0x00000001, 0x00002000,
416 0x00800001, 0x00002001, 0x00802080, 0x00800081,
417 0x00002001, 0x00002080, 0x00800000, 0x00802001,
418 0x00000080, 0x00800000, 0x00002000, 0x00802080L},
d1e726bc 419
420 {0x00000100, 0x02080100, 0x02080000, 0x42000100,
32874aea 421 0x00080000, 0x00000100, 0x40000000, 0x02080000,
422 0x40080100, 0x00080000, 0x02000100, 0x40080100,
423 0x42000100, 0x42080000, 0x00080100, 0x40000000,
424 0x02000000, 0x40080000, 0x40080000, 0x00000000,
425 0x40000100, 0x42080100, 0x42080100, 0x02000100,
426 0x42080000, 0x40000100, 0x00000000, 0x42000000,
427 0x02080100, 0x02000000, 0x42000000, 0x00080100,
428 0x00080000, 0x42000100, 0x00000100, 0x02000000,
429 0x40000000, 0x02080000, 0x42000100, 0x40080100,
430 0x02000100, 0x40000000, 0x42080000, 0x02080100,
431 0x40080100, 0x00000100, 0x02000000, 0x42080000,
432 0x42080100, 0x00080100, 0x42000000, 0x42080100,
433 0x02080000, 0x00000000, 0x40080000, 0x42000000,
434 0x00080100, 0x02000100, 0x40000100, 0x00080000,
435 0x00000000, 0x40080000, 0x02080100, 0x40000100L},
d1e726bc 436
437 {0x20000010, 0x20400000, 0x00004000, 0x20404010,
32874aea 438 0x20400000, 0x00000010, 0x20404010, 0x00400000,
439 0x20004000, 0x00404010, 0x00400000, 0x20000010,
440 0x00400010, 0x20004000, 0x20000000, 0x00004010,
441 0x00000000, 0x00400010, 0x20004010, 0x00004000,
442 0x00404000, 0x20004010, 0x00000010, 0x20400010,
443 0x20400010, 0x00000000, 0x00404010, 0x20404000,
444 0x00004010, 0x00404000, 0x20404000, 0x20000000,
445 0x20004000, 0x00000010, 0x20400010, 0x00404000,
446 0x20404010, 0x00400000, 0x00004010, 0x20000010,
447 0x00400000, 0x20004000, 0x20000000, 0x00004010,
448 0x20000010, 0x20404010, 0x00404000, 0x20400000,
449 0x00404010, 0x20404000, 0x00000000, 0x20400010,
450 0x00000010, 0x00004000, 0x20400000, 0x00404010,
451 0x00004000, 0x00400010, 0x20004010, 0x00000000,
452 0x20404000, 0x20000000, 0x00400010, 0x20004010L},
d1e726bc 453
454 {0x00200000, 0x04200002, 0x04000802, 0x00000000,
32874aea 455 0x00000800, 0x04000802, 0x00200802, 0x04200800,
456 0x04200802, 0x00200000, 0x00000000, 0x04000002,
457 0x00000002, 0x04000000, 0x04200002, 0x00000802,
458 0x04000800, 0x00200802, 0x00200002, 0x04000800,
459 0x04000002, 0x04200000, 0x04200800, 0x00200002,
460 0x04200000, 0x00000800, 0x00000802, 0x04200802,
461 0x00200800, 0x00000002, 0x04000000, 0x00200800,
462 0x04000000, 0x00200800, 0x00200000, 0x04000802,
463 0x04000802, 0x04200002, 0x04200002, 0x00000002,
464 0x00200002, 0x04000000, 0x04000800, 0x00200000,
465 0x04200800, 0x00000802, 0x00200802, 0x04200800,
466 0x00000802, 0x04000002, 0x04200802, 0x04200000,
467 0x00200800, 0x00000000, 0x00000002, 0x04200802,
468 0x00000000, 0x00200802, 0x04200000, 0x00000800,
469 0x04000002, 0x04000800, 0x00000800, 0x00200002L},
d1e726bc 470
471 {0x10001040, 0x00001000, 0x00040000, 0x10041040,
32874aea 472 0x10000000, 0x10001040, 0x00000040, 0x10000000,
473 0x00040040, 0x10040000, 0x10041040, 0x00041000,
474 0x10041000, 0x00041040, 0x00001000, 0x00000040,
475 0x10040000, 0x10000040, 0x10001000, 0x00001040,
476 0x00041000, 0x00040040, 0x10040040, 0x10041000,
477 0x00001040, 0x00000000, 0x00000000, 0x10040040,
478 0x10000040, 0x10001000, 0x00041040, 0x00040000,
479 0x00041040, 0x00040000, 0x10041000, 0x00001000,
480 0x00000040, 0x10040040, 0x00001000, 0x00041040,
481 0x10001000, 0x00000040, 0x10000040, 0x10040000,
482 0x10040040, 0x10000000, 0x00040000, 0x10001040,
483 0x00000000, 0x10041040, 0x00040040, 0x10000040,
484 0x10040000, 0x10001000, 0x10001040, 0x00000000,
485 0x10041040, 0x00041000, 0x00041000, 0x00001040,
486 0x00001040, 0x00040040, 0x10000000, 0x10041000L}
d1e726bc 487};
374330e2 488
d1e726bc 489#define f(R, K0246, K1357) (\
490 s0246 = R ^ K0246, \
491 s1357 = R ^ K1357, \
492 s0246 = rotl(s0246, 28), \
493 SPboxes[0] [(s0246 >> 24) & 0x3F] | \
494 SPboxes[1] [(s1357 >> 24) & 0x3F] | \
495 SPboxes[2] [(s0246 >> 16) & 0x3F] | \
496 SPboxes[3] [(s1357 >> 16) & 0x3F] | \
497 SPboxes[4] [(s0246 >> 8) & 0x3F] | \
498 SPboxes[5] [(s1357 >> 8) & 0x3F] | \
499 SPboxes[6] [(s0246 ) & 0x3F] | \
500 SPboxes[7] [(s1357 ) & 0x3F])
501
502#define bitswap(L, R, n, mask) (\
503 swap = mask & ( (R >> n) ^ L ), \
504 R ^= swap << n, \
505 L ^= swap)
506
507/* Initial permutation */
508#define IP(L, R) (\
509 bitswap(R, L, 4, 0x0F0F0F0F), \
510 bitswap(R, L, 16, 0x0000FFFF), \
511 bitswap(L, R, 2, 0x33333333), \
512 bitswap(L, R, 8, 0x00FF00FF), \
513 bitswap(R, L, 1, 0x55555555))
514
515/* Final permutation */
516#define FP(L, R) (\
517 bitswap(R, L, 1, 0x55555555), \
518 bitswap(L, R, 8, 0x00FF00FF), \
519 bitswap(L, R, 2, 0x33333333), \
520 bitswap(R, L, 16, 0x0000FFFF), \
521 bitswap(R, L, 4, 0x0F0F0F0F))
522
32874aea 523void des_encipher(word32 * output, word32 L, word32 R, DESContext * sched)
524{
d1e726bc 525 word32 swap, s0246, s1357;
526
527 IP(L, R);
528
529 L = rotl(L, 1);
530 R = rotl(R, 1);
531
32874aea 532 L ^= f(R, sched->k0246[0], sched->k1357[0]);
533 R ^= f(L, sched->k0246[1], sched->k1357[1]);
534 L ^= f(R, sched->k0246[2], sched->k1357[2]);
535 R ^= f(L, sched->k0246[3], sched->k1357[3]);
536 L ^= f(R, sched->k0246[4], sched->k1357[4]);
537 R ^= f(L, sched->k0246[5], sched->k1357[5]);
538 L ^= f(R, sched->k0246[6], sched->k1357[6]);
539 R ^= f(L, sched->k0246[7], sched->k1357[7]);
540 L ^= f(R, sched->k0246[8], sched->k1357[8]);
541 R ^= f(L, sched->k0246[9], sched->k1357[9]);
d1e726bc 542 L ^= f(R, sched->k0246[10], sched->k1357[10]);
543 R ^= f(L, sched->k0246[11], sched->k1357[11]);
544 L ^= f(R, sched->k0246[12], sched->k1357[12]);
545 R ^= f(L, sched->k0246[13], sched->k1357[13]);
546 L ^= f(R, sched->k0246[14], sched->k1357[14]);
547 R ^= f(L, sched->k0246[15], sched->k1357[15]);
548
549 L = rotl(L, 31);
550 R = rotl(R, 31);
551
32874aea 552 swap = L;
553 L = R;
554 R = swap;
d1e726bc 555
556 FP(L, R);
557
558 output[0] = L;
559 output[1] = R;
560}
374330e2 561
32874aea 562void des_decipher(word32 * output, word32 L, word32 R, DESContext * sched)
563{
d1e726bc 564 word32 swap, s0246, s1357;
374330e2 565
d1e726bc 566 IP(L, R);
374330e2 567
d1e726bc 568 L = rotl(L, 1);
569 R = rotl(R, 1);
374330e2 570
d1e726bc 571 L ^= f(R, sched->k0246[15], sched->k1357[15]);
572 R ^= f(L, sched->k0246[14], sched->k1357[14]);
573 L ^= f(R, sched->k0246[13], sched->k1357[13]);
574 R ^= f(L, sched->k0246[12], sched->k1357[12]);
575 L ^= f(R, sched->k0246[11], sched->k1357[11]);
576 R ^= f(L, sched->k0246[10], sched->k1357[10]);
32874aea 577 L ^= f(R, sched->k0246[9], sched->k1357[9]);
578 R ^= f(L, sched->k0246[8], sched->k1357[8]);
579 L ^= f(R, sched->k0246[7], sched->k1357[7]);
580 R ^= f(L, sched->k0246[6], sched->k1357[6]);
581 L ^= f(R, sched->k0246[5], sched->k1357[5]);
582 R ^= f(L, sched->k0246[4], sched->k1357[4]);
583 L ^= f(R, sched->k0246[3], sched->k1357[3]);
584 R ^= f(L, sched->k0246[2], sched->k1357[2]);
585 L ^= f(R, sched->k0246[1], sched->k1357[1]);
586 R ^= f(L, sched->k0246[0], sched->k1357[0]);
d1e726bc 587
588 L = rotl(L, 31);
589 R = rotl(R, 31);
590
32874aea 591 swap = L;
592 L = R;
593 R = swap;
d1e726bc 594
595 FP(L, R);
596
597 output[0] = L;
598 output[1] = R;
374330e2 599}
600
d1e726bc 601#define GET_32BIT_MSB_FIRST(cp) \
602 (((unsigned long)(unsigned char)(cp)[3]) | \
603 ((unsigned long)(unsigned char)(cp)[2] << 8) | \
604 ((unsigned long)(unsigned char)(cp)[1] << 16) | \
605 ((unsigned long)(unsigned char)(cp)[0] << 24))
606
607#define PUT_32BIT_MSB_FIRST(cp, value) do { \
608 (cp)[3] = (value); \
609 (cp)[2] = (value) >> 8; \
610 (cp)[1] = (value) >> 16; \
611 (cp)[0] = (value) >> 24; } while (0)
612
613static void des_cbc_encrypt(unsigned char *dest, const unsigned char *src,
32874aea 614 unsigned int len, DESContext * sched)
615{
d1e726bc 616 word32 out[2], iv0, iv1;
617 unsigned int i;
618
619 assert((len & 7) == 0);
620
371e569c 621 iv0 = sched->iv0;
622 iv1 = sched->iv1;
d1e726bc 623 for (i = 0; i < len; i += 8) {
32874aea 624 iv0 ^= GET_32BIT_MSB_FIRST(src);
625 src += 4;
626 iv1 ^= GET_32BIT_MSB_FIRST(src);
627 src += 4;
628 des_encipher(out, iv0, iv1, sched);
629 iv0 = out[0];
630 iv1 = out[1];
631 PUT_32BIT_MSB_FIRST(dest, iv0);
632 dest += 4;
633 PUT_32BIT_MSB_FIRST(dest, iv1);
634 dest += 4;
374330e2 635 }
371e569c 636 sched->iv0 = iv0;
637 sched->iv1 = iv1;
374330e2 638}
639
d1e726bc 640static void des_cbc_decrypt(unsigned char *dest, const unsigned char *src,
32874aea 641 unsigned int len, DESContext * sched)
642{
d1e726bc 643 word32 out[2], iv0, iv1, xL, xR;
644 unsigned int i;
645
646 assert((len & 7) == 0);
647
371e569c 648 iv0 = sched->iv0;
649 iv1 = sched->iv1;
d1e726bc 650 for (i = 0; i < len; i += 8) {
32874aea 651 xL = GET_32BIT_MSB_FIRST(src);
652 src += 4;
653 xR = GET_32BIT_MSB_FIRST(src);
654 src += 4;
655 des_decipher(out, xL, xR, sched);
656 iv0 ^= out[0];
657 iv1 ^= out[1];
658 PUT_32BIT_MSB_FIRST(dest, iv0);
659 dest += 4;
660 PUT_32BIT_MSB_FIRST(dest, iv1);
661 dest += 4;
662 iv0 = xL;
663 iv1 = xR;
374330e2 664 }
371e569c 665 sched->iv0 = iv0;
666 sched->iv1 = iv1;
374330e2 667}
668
d1e726bc 669static void des_3cbc_encrypt(unsigned char *dest, const unsigned char *src,
32874aea 670 unsigned int len, DESContext * scheds)
671{
d1e726bc 672 des_cbc_encrypt(dest, src, len, &scheds[0]);
673 des_cbc_decrypt(dest, src, len, &scheds[1]);
674 des_cbc_encrypt(dest, src, len, &scheds[2]);
374330e2 675}
676
033b4cef 677static void des_cbc3_encrypt(unsigned char *dest, const unsigned char *src,
32874aea 678 unsigned int len, DESContext * scheds)
679{
033b4cef 680 word32 out[2], iv0, iv1;
681 unsigned int i;
682
683 assert((len & 7) == 0);
684
371e569c 685 iv0 = scheds->iv0;
686 iv1 = scheds->iv1;
033b4cef 687 for (i = 0; i < len; i += 8) {
32874aea 688 iv0 ^= GET_32BIT_MSB_FIRST(src);
689 src += 4;
690 iv1 ^= GET_32BIT_MSB_FIRST(src);
691 src += 4;
692 des_encipher(out, iv0, iv1, &scheds[0]);
693 des_decipher(out, out[0], out[1], &scheds[1]);
694 des_encipher(out, out[0], out[1], &scheds[2]);
695 iv0 = out[0];
696 iv1 = out[1];
697 PUT_32BIT_MSB_FIRST(dest, iv0);
698 dest += 4;
699 PUT_32BIT_MSB_FIRST(dest, iv1);
700 dest += 4;
033b4cef 701 }
371e569c 702 scheds->iv0 = iv0;
703 scheds->iv1 = iv1;
033b4cef 704}
705
d1e726bc 706static void des_3cbc_decrypt(unsigned char *dest, const unsigned char *src,
32874aea 707 unsigned int len, DESContext * scheds)
708{
d1e726bc 709 des_cbc_decrypt(dest, src, len, &scheds[2]);
710 des_cbc_encrypt(dest, src, len, &scheds[1]);
711 des_cbc_decrypt(dest, src, len, &scheds[0]);
374330e2 712}
713
033b4cef 714static void des_cbc3_decrypt(unsigned char *dest, const unsigned char *src,
32874aea 715 unsigned int len, DESContext * scheds)
716{
033b4cef 717 word32 out[2], iv0, iv1, xL, xR;
718 unsigned int i;
719
720 assert((len & 7) == 0);
721
371e569c 722 iv0 = scheds->iv0;
723 iv1 = scheds->iv1;
033b4cef 724 for (i = 0; i < len; i += 8) {
32874aea 725 xL = GET_32BIT_MSB_FIRST(src);
726 src += 4;
727 xR = GET_32BIT_MSB_FIRST(src);
728 src += 4;
729 des_decipher(out, xL, xR, &scheds[2]);
730 des_encipher(out, out[0], out[1], &scheds[1]);
731 des_decipher(out, out[0], out[1], &scheds[0]);
732 iv0 ^= out[0];
733 iv1 ^= out[1];
734 PUT_32BIT_MSB_FIRST(dest, iv0);
735 dest += 4;
736 PUT_32BIT_MSB_FIRST(dest, iv1);
737 dest += 4;
738 iv0 = xL;
739 iv1 = xR;
033b4cef 740 }
371e569c 741 scheds->iv0 = iv0;
742 scheds->iv1 = iv1;
033b4cef 743}
744
371e569c 745static void *des3_make_context(void)
746{
747 return smalloc(3*sizeof(DESContext));
748}
374330e2 749
371e569c 750static void *des3_ssh1_make_context(void)
32874aea 751{
371e569c 752 /* Need 3 keys for each direction, in SSH1 */
753 return smalloc(6*sizeof(DESContext));
d39f364a 754}
755
371e569c 756static void *des_make_context(void)
0d7c43a6 757{
371e569c 758 return smalloc(sizeof(DESContext));
0d7c43a6 759}
760
371e569c 761static void *des_ssh1_make_context(void)
32874aea 762{
371e569c 763 /* Need one key for each direction, in SSH1 */
764 return smalloc(2*sizeof(DESContext));
d39f364a 765}
766
371e569c 767static void des3_free_context(void *handle) /* used for both 3DES and DES */
32874aea 768{
371e569c 769 sfree(handle);
d39f364a 770}
771
371e569c 772static void des3_key(void *handle, unsigned char *key)
32874aea 773{
371e569c 774 DESContext *keys = (DESContext *) handle;
d39f364a 775 des_key_setup(GET_32BIT_MSB_FIRST(key),
371e569c 776 GET_32BIT_MSB_FIRST(key + 4), &keys[0]);
32874aea 777 des_key_setup(GET_32BIT_MSB_FIRST(key + 8),
371e569c 778 GET_32BIT_MSB_FIRST(key + 12), &keys[1]);
32874aea 779 des_key_setup(GET_32BIT_MSB_FIRST(key + 16),
371e569c 780 GET_32BIT_MSB_FIRST(key + 20), &keys[2]);
781}
782
783static void des3_iv(void *handle, unsigned char *key)
784{
785 DESContext *keys = (DESContext *) handle;
786 keys[0].iv0 = GET_32BIT_MSB_FIRST(key);
787 keys[0].iv1 = GET_32BIT_MSB_FIRST(key + 4);
d39f364a 788}
789
371e569c 790static void des_key(void *handle, unsigned char *key)
0d7c43a6 791{
371e569c 792 DESContext *keys = (DESContext *) handle;
0d7c43a6 793 des_key_setup(GET_32BIT_MSB_FIRST(key),
371e569c 794 GET_32BIT_MSB_FIRST(key + 4), &keys[0]);
0d7c43a6 795}
796
371e569c 797static void des3_sesskey(void *handle, unsigned char *key)
32874aea 798{
371e569c 799 DESContext *keys = (DESContext *) handle;
800 des3_key(keys, key);
801 des3_key(keys+3, key);
374330e2 802}
803
371e569c 804static void des3_encrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 805{
371e569c 806 DESContext *keys = (DESContext *) handle;
807 des_3cbc_encrypt(blk, blk, len, keys);
374330e2 808}
809
371e569c 810static void des3_decrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 811{
371e569c 812 DESContext *keys = (DESContext *) handle;
813 des_3cbc_decrypt(blk, blk, len, keys+3);
374330e2 814}
815
371e569c 816static void des3_ssh2_encrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 817{
371e569c 818 DESContext *keys = (DESContext *) handle;
819 des_cbc3_encrypt(blk, blk, len, keys);
033b4cef 820}
821
371e569c 822static void des3_ssh2_decrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 823{
371e569c 824 DESContext *keys = (DESContext *) handle;
825 des_cbc3_decrypt(blk, blk, len, keys);
033b4cef 826}
827
371e569c 828static void des_ssh2_encrypt_blk(void *handle, unsigned char *blk, int len)
0d7c43a6 829{
371e569c 830 DESContext *keys = (DESContext *) handle;
831 des_cbc_encrypt(blk, blk, len, keys);
0d7c43a6 832}
833
371e569c 834static void des_ssh2_decrypt_blk(void *handle, unsigned char *blk, int len)
0d7c43a6 835{
371e569c 836 DESContext *keys = (DESContext *) handle;
837 des_cbc_decrypt(blk, blk, len, keys);
0d7c43a6 838}
839
32874aea 840void des3_decrypt_pubkey(unsigned char *key, unsigned char *blk, int len)
841{
7cca0d81 842 DESContext ourkeys[3];
843 des_key_setup(GET_32BIT_MSB_FIRST(key),
32874aea 844 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[0]);
845 des_key_setup(GET_32BIT_MSB_FIRST(key + 8),
846 GET_32BIT_MSB_FIRST(key + 12), &ourkeys[1]);
7cca0d81 847 des_key_setup(GET_32BIT_MSB_FIRST(key),
32874aea 848 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[2]);
7cca0d81 849 des_3cbc_decrypt(blk, blk, len, ourkeys);
ee5c1422 850 memset(ourkeys, 0, sizeof(ourkeys));
7cca0d81 851}
852
32874aea 853void des3_encrypt_pubkey(unsigned char *key, unsigned char *blk, int len)
854{
6e522441 855 DESContext ourkeys[3];
856 des_key_setup(GET_32BIT_MSB_FIRST(key),
32874aea 857 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[0]);
858 des_key_setup(GET_32BIT_MSB_FIRST(key + 8),
859 GET_32BIT_MSB_FIRST(key + 12), &ourkeys[1]);
6e522441 860 des_key_setup(GET_32BIT_MSB_FIRST(key),
32874aea 861 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[2]);
6e522441 862 des_3cbc_encrypt(blk, blk, len, ourkeys);
ee5c1422 863 memset(ourkeys, 0, sizeof(ourkeys));
6e522441 864}
865
9dda6459 866void des3_decrypt_pubkey_ossh(unsigned char *key, unsigned char *iv,
867 unsigned char *blk, int len)
868{
869 DESContext ourkeys[3];
870 des_key_setup(GET_32BIT_MSB_FIRST(key),
871 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[0]);
872 des_key_setup(GET_32BIT_MSB_FIRST(key + 8),
873 GET_32BIT_MSB_FIRST(key + 12), &ourkeys[1]);
874 des_key_setup(GET_32BIT_MSB_FIRST(key + 16),
875 GET_32BIT_MSB_FIRST(key + 20), &ourkeys[2]);
371e569c 876 ourkeys[0].iv0 = GET_32BIT_MSB_FIRST(iv);
877 ourkeys[0].iv1 = GET_32BIT_MSB_FIRST(iv+4);
9dda6459 878 des_cbc3_decrypt(blk, blk, len, ourkeys);
ee5c1422 879 memset(ourkeys, 0, sizeof(ourkeys));
9dda6459 880}
881
882void des3_encrypt_pubkey_ossh(unsigned char *key, unsigned char *iv,
883 unsigned char *blk, int len)
884{
885 DESContext ourkeys[3];
886 des_key_setup(GET_32BIT_MSB_FIRST(key),
887 GET_32BIT_MSB_FIRST(key + 4), &ourkeys[0]);
888 des_key_setup(GET_32BIT_MSB_FIRST(key + 8),
889 GET_32BIT_MSB_FIRST(key + 12), &ourkeys[1]);
890 des_key_setup(GET_32BIT_MSB_FIRST(key + 16),
891 GET_32BIT_MSB_FIRST(key + 20), &ourkeys[2]);
371e569c 892 ourkeys[0].iv0 = GET_32BIT_MSB_FIRST(iv);
893 ourkeys[0].iv1 = GET_32BIT_MSB_FIRST(iv+4);
9dda6459 894 des_cbc3_encrypt(blk, blk, len, ourkeys);
ee5c1422 895 memset(ourkeys, 0, sizeof(ourkeys));
9dda6459 896}
897
65a22376 898static const struct ssh2_cipher ssh_3des_ssh2 = {
371e569c 899 des3_make_context, des3_free_context, des3_iv, des3_key,
900 des3_ssh2_encrypt_blk, des3_ssh2_decrypt_blk,
033b4cef 901 "3des-cbc",
371e569c 902 8, 168, "triple-DES"
033b4cef 903};
904
0d7c43a6 905/*
906 * Single DES in ssh2. It isn't clear that "des-cbc" is an official
907 * cipher name, but ssh.com support it and apparently aren't the
908 * only people to do so, so we sigh and implement it anyway.
909 */
910static const struct ssh2_cipher ssh_des_ssh2 = {
68a49acb 911 des_make_context, des3_free_context, des3_iv, des_key,
371e569c 912 des_ssh2_encrypt_blk, des_ssh2_decrypt_blk,
0d7c43a6 913 "des-cbc",
371e569c 914 8, 56, "single-DES"
0d7c43a6 915};
916
65a22376 917static const struct ssh2_cipher *const des3_list[] = {
0a3f1d48 918 &ssh_3des_ssh2
919};
920
65a22376 921const struct ssh2_ciphers ssh2_3des = {
0a3f1d48 922 sizeof(des3_list) / sizeof(*des3_list),
923 des3_list
924};
925
0d7c43a6 926static const struct ssh2_cipher *const des_list[] = {
927 &ssh_des_ssh2
928};
929
930const struct ssh2_ciphers ssh2_des = {
931 sizeof(des3_list) / sizeof(*des_list),
932 des_list
933};
934
65a22376 935const struct ssh_cipher ssh_3des = {
371e569c 936 des3_ssh1_make_context, des3_free_context, des3_sesskey,
937 des3_encrypt_blk, des3_decrypt_blk,
938 8, "triple-DES"
374330e2 939};
940
371e569c 941static void des_sesskey(void *handle, unsigned char *key)
32874aea 942{
371e569c 943 DESContext *keys = (DESContext *) handle;
944 des_key(keys, key);
945 des_key(keys+1, key);
9697bfd2 946}
947
371e569c 948static void des_encrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 949{
371e569c 950 DESContext *keys = (DESContext *) handle;
951 des_cbc_encrypt(blk, blk, len, keys);
9697bfd2 952}
953
371e569c 954static void des_decrypt_blk(void *handle, unsigned char *blk, int len)
32874aea 955{
371e569c 956 DESContext *keys = (DESContext *) handle;
957 des_cbc_decrypt(blk, blk, len, keys+1);
9697bfd2 958}
959
65a22376 960const struct ssh_cipher ssh_des = {
371e569c 961 des_ssh1_make_context, des3_free_context, des_sesskey,
962 des_encrypt_blk, des_decrypt_blk,
963 8, "single-DES"
9697bfd2 964};