Mercurial > hg > ucis.core

comparison NaCl/crypto_sign/edwards25519sha512batch.cs @ 20:c873e3dd73fe

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Added NaCl cryptography code
author Ivo Smits <Ivo@UCIS.nl>
date Mon, 15 Apr 2013 00:43:48 +0200
parents
children 7e9d1cfcc562
comparison
equal deleted inserted replaced
19:b9ef273964fd 20:c873e3dd73fe
1 using System;
2 using UCIS.NaCl;
3 using System.Runtime.InteropServices;
4
5 namespace UCIS.NaCl.crypto_sign {
6 public static class edwards25519sha512batch {
7 public const int SECRETKEYBYTES = 64;
8 public const int PUBLICKEYBYTES = 32;
9 public const int CRYPTO_BYTES = 64;
10
11 /*Arithmetic modulo the group order n = 2^252 + 27742317777372353535851937790883648493 = 7237005577332262213973186563042994240857116359379907606001950938285454250989 */
12
13 unsafe struct sc25519 {
14 public fixed UInt32 v[32]; //crypto_uint32 v[32];
15
16 static UInt32[] m = new UInt32[32] {0xED, 0xD3, 0xF5, 0x5C, 0x1A, 0x63, 0x12, 0x58, 0xD6, 0x9C, 0xF7, 0xA2, 0xDE, 0xF9, 0xDE, 0x14,
17 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10};
18
19 static UInt32[] mu = new UInt32[33] {0x1B, 0x13, 0x2C, 0x0A, 0xA3, 0xE5, 0x9C, 0xED, 0xA7, 0x29, 0x63, 0x08, 0x5D, 0x21, 0x06, 0x21,
20 0xEB, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x0F};
21
22 /* Reduce coefficients of r before calling reduce_add_sub */
23 static unsafe void reduce_add_sub(sc25519* r) {
24 int i, b = 0, pb = 0, nb;
25 Byte* t = stackalloc Byte[32];
26
27 for (i = 0; i < 32; i++) {
28 b = (r->v[i] < pb + m[i]) ? 1 : 0;
29 t[i] = (Byte)(r->v[i] - pb - m[i] + b * 256);
30 pb = b;
31 }
32 nb = 1 - b;
33 for (i = 0; i < 32; i++) r->v[i] = (uint)(r->v[i] * b + t[i] * nb);
34 }
35
36 /* Reduce coefficients of x before calling barrett_reduce */
37 static unsafe void barrett_reduce(sc25519* r, UInt32* x) { // const crypto_uint32 x[64]
38 /* See HAC, Alg. 14.42 */
39 UInt32* q2 = stackalloc UInt32[66]; // { 0 };
40 for (int z = 0; z < 66; z++) q2[z] = 0;
41 UInt32* q3 = q2 + 33;
42 UInt32* r1 = stackalloc UInt32[33];
43 UInt32* r2 = stackalloc UInt32[33]; // { 0 };
44 for (int z = 0; z < 33; z++) r2[z] = 0;
45 UInt32 carry;
46 int b, pb = 0;
47
48 for (int i = 0; i < 33; i++)
49 for (int j = 0; j < 33; j++)
50 if (i + j >= 31) q2[i + j] += mu[i] * x[j + 31];
51 carry = q2[31] >> 8;
52 q2[32] += carry;
53 carry = q2[32] >> 8;
54 q2[33] += carry;
55
56 for (int i = 0; i < 33; i++) r1[i] = x[i];
57 for (int i = 0; i < 32; i++)
58 for (int j = 0; j < 33; j++)
59 if (i + j < 33) r2[i + j] += m[i] * q3[j];
60
61 for (int i = 0; i < 32; i++) {
62 carry = r2[i] >> 8;
63 r2[i + 1] += carry;
64 r2[i] &= 0xff;
65 }
66
67 for (int i = 0; i < 32; i++) {
68 b = (r1[i] < pb + r2[i]) ? 1 : 0;
69 r->v[i] = (uint)(r1[i] - pb - r2[i] + b * 256);
70 pb = b;
71 }
72
73 /* XXX: Can it really happen that r<0?, See HAC, Alg 14.42, Step 3
74 * If so: Handle it here!
75 */
76
77 reduce_add_sub(r);
78 reduce_add_sub(r);
79 }
80
81 /*
82 static int iszero(const sc25519 *x)
83 {
84 // Implement
85 return 0;
86 }
87 */
88
89 public static unsafe void sc25519_from32bytes(sc25519* r, Byte* x) { //const unsigned char x[32]
90 UInt32* t = stackalloc UInt32[64]; // { 0 };
91 for (int i = 0; i < 32; i++) t[i] = x[i];
92 for (int i = 32; i < 64; i++) t[i] = 0;
93 barrett_reduce(r, t);
94 }
95
96 public static unsafe void sc25519_from64bytes(sc25519* r, Byte* x) { //const unsigned char x[64]
97 UInt32* t = stackalloc UInt32[64]; // { 0 };
98 for (int i = 0; i < 64; i++) t[i] = x[i];
99 barrett_reduce(r, t);
100 }
101
102 /* XXX: What we actually want for crypto_group is probably just something like
103 * void sc25519_frombytes(sc25519 *r, const unsigned char *x, size_t xlen)
104 */
105
106 public static unsafe void sc25519_to32bytes(Byte* r, sc25519* x) { //unsigned char r[32]
107 for (int i = 0; i < 32; i++) r[i] = (Byte)x->v[i];
108 }
109
110 public static unsafe void sc25519_add(sc25519* r, sc25519* x, sc25519* y) {
111 for (int i = 0; i < 32; i++) r->v[i] = x->v[i] + y->v[i];
112 for (int i = 0; i < 31; i++) {
113 uint carry = r->v[i] >> 8;
114 r->v[i + 1] += carry;
115 r->v[i] &= 0xff;
116 }
117 reduce_add_sub(r);
118 }
119
120 public static unsafe void sc25519_mul(sc25519* r, sc25519* x, sc25519* y) {
121 UInt32* t = stackalloc UInt32[64];
122 for (int i = 0; i < 64; i++) t[i] = 0;
123
124 for (int i = 0; i < 32; i++)
125 for (int j = 0; j < 32; j++)
126 t[i + j] += x->v[i] * y->v[j];
127
128 /* Reduce coefficients */
129 for (int i = 0; i < 63; i++) {
130 uint carry = t[i] >> 8;
131 t[i + 1] += carry;
132 t[i] &= 0xff;
133 }
134
135 barrett_reduce(r, t);
136 }
137
138 public static unsafe void sc25519_square(sc25519* r, sc25519* x) {
139 sc25519_mul(r, x, x);
140 }
141 }
142 struct ge25519 {
143 public fe25519 x;
144 public fe25519 y;
145 public fe25519 z;
146 public fe25519 t;
147
148 struct ge25519_p1p1 {
149 public fe25519 x;
150 public fe25519 z;
151 public fe25519 y;
152 public fe25519 t;
153 }
154 struct ge25519_p2 {
155 public fe25519 x;
156 public fe25519 y;
157 public fe25519 z;
158 }
159
160 /* Windowsize for fixed-window scalar multiplication */
161 const int WINDOWSIZE = 2; //#define WINDOWSIZE 2 /* Should be 1,2, or 4 */
162 const int WINDOWMASK = ((1 << WINDOWSIZE) - 1); //#define WINDOWMASK ((1<<WINDOWSIZE)-1)
163
164 /* packed parameter d in the Edwards curve equation */
165 static Byte[] ecd = new Byte[32] {0xA3, 0x78, 0x59, 0x13, 0xCA, 0x4D, 0xEB, 0x75, 0xAB, 0xD8, 0x41, 0x41, 0x4D, 0x0A, 0x70, 0x00,
166 0x98, 0xE8, 0x79, 0x77, 0x79, 0x40, 0xC7, 0x8C, 0x73, 0xFE, 0x6F, 0x2B, 0xEE, 0x6C, 0x03, 0x52};
167
168 /* Packed coordinates of the base point */
169 static Byte[] ge25519_base_x = new Byte[32] {0x1A, 0xD5, 0x25, 0x8F, 0x60, 0x2D, 0x56, 0xC9, 0xB2, 0xA7, 0x25, 0x95, 0x60, 0xC7, 0x2C, 0x69,
170 0x5C, 0xDC, 0xD6, 0xFD, 0x31, 0xE2, 0xA4, 0xC0, 0xFE, 0x53, 0x6E, 0xCD, 0xD3, 0x36, 0x69, 0x21};
171 static Byte[] ge25519_base_y = new Byte[32] {0x58, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66,
172 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66, 0x66};
173 static Byte[] ge25519_base_z = new Byte[32] { 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
174 static Byte[] ge25519_base_t = new Byte[32] {0xA3, 0xDD, 0xB7, 0xA5, 0xB3, 0x8A, 0xDE, 0x6D, 0xF5, 0x52, 0x51, 0x77, 0x80, 0x9F, 0xF0, 0x20,
175 0x7D, 0xE3, 0xAB, 0x64, 0x8E, 0x4E, 0xEA, 0x66, 0x65, 0x76, 0x8B, 0xD7, 0x0F, 0x5F, 0x87, 0x67};
176
177 /* Packed coordinates of the neutral element */
178 static Byte[] ge25519_neutral_x = new Byte[32]; // { 0 };
179 static Byte[] ge25519_neutral_y = new Byte[32] { 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
180 static Byte[] ge25519_neutral_z = new Byte[32] { 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
181 static Byte[] ge25519_neutral_t = new Byte[32]; // { 0 };
182
183 static unsafe void p1p1_to_p2(ge25519_p2* r, ge25519_p1p1* p) {
184 fe25519.fe25519_mul(&r->x, &p->x, &p->t);
185 fe25519.fe25519_mul(&r->y, &p->y, &p->z);
186 fe25519.fe25519_mul(&r->z, &p->z, &p->t);
187 }
188
189 static unsafe void p1p1_to_p3(ge25519* r, ge25519_p1p1* p) {
190 p1p1_to_p2((ge25519_p2*)r, p);
191 fe25519.fe25519_mul(&r->t, &p->x, &p->y);
192 }
193
194 /* Constant-time version of: if(b) r = p */
195 static unsafe void cmov_p3(ge25519* r, ge25519* p, Byte b) {
196 fe25519.fe25519_cmov(&r->x, &p->x, b);
197 fe25519.fe25519_cmov(&r->y, &p->y, b);
198 fe25519.fe25519_cmov(&r->z, &p->z, b);
199 fe25519.fe25519_cmov(&r->t, &p->t, b);
200 }
201
202 /* See http://www.hyperelliptic.org/EFD/g1p/auto-twisted-extended-1.html#doubling-dbl-2008-hwcd */
203 static unsafe void dbl_p1p1(ge25519_p1p1* r, ge25519_p2* p) {
204 fe25519 a, b, c, d;
205 fe25519.fe25519_square(&a, &p->x);
206 fe25519.fe25519_square(&b, &p->y);
207 fe25519.fe25519_square(&c, &p->z);
208 fe25519.fe25519_add(&c, &c, &c);
209 fe25519.fe25519_neg(&d, &a);
210
211 fe25519.fe25519_add(&r->x, &p->x, &p->y);
212 fe25519.fe25519_square(&r->x, &r->x);
213 fe25519.fe25519_sub(&r->x, &r->x, &a);
214 fe25519.fe25519_sub(&r->x, &r->x, &b);
215 fe25519.fe25519_add(&r->z, &d, &b);
216 fe25519.fe25519_sub(&r->t, &r->z, &c);
217 fe25519.fe25519_sub(&r->y, &d, &b);
218 }
219
220 static unsafe void add_p1p1(ge25519_p1p1* r, ge25519* p, ge25519* q) {
221 fe25519 a, b, c, d, t, fd;
222 fixed (Byte* ecdp = ecd) fe25519.fe25519_unpack(&fd, ecdp);
223
224 fe25519.fe25519_sub(&a, &p->y, &p->x); // A = (Y1-X1)*(Y2-X2)
225 fe25519.fe25519_sub(&t, &q->y, &q->x);
226 fe25519.fe25519_mul(&a, &a, &t);
227 fe25519.fe25519_add(&b, &p->x, &p->y); // B = (Y1+X1)*(Y2+X2)
228 fe25519.fe25519_add(&t, &q->x, &q->y);
229 fe25519.fe25519_mul(&b, &b, &t);
230 fe25519.fe25519_mul(&c, &p->t, &q->t); //C = T1*k*T2
231 fe25519.fe25519_mul(&c, &c, &fd);
232 fe25519.fe25519_add(&c, &c, &c); //XXX: Can save this addition by precomputing 2*ecd
233 fe25519.fe25519_mul(&d, &p->z, &q->z); //D = Z1*2*Z2
234 fe25519.fe25519_add(&d, &d, &d);
235 fe25519.fe25519_sub(&r->x, &b, &a); // E = B-A
236 fe25519.fe25519_sub(&r->t, &d, &c); // F = D-C
237 fe25519.fe25519_add(&r->z, &d, &c); // G = D+C
238 fe25519.fe25519_add(&r->y, &b, &a); // H = B+A
239 }
240
241 /* ********************************************************************
242 * EXPORTED FUNCTIONS
243 ******************************************************************** */
244
245 /* return 0 on success, -1 otherwise */
246 public unsafe static Boolean ge25519_unpack_vartime(ge25519* r, Byte* p) { //const unsigned char p[32]
247 Boolean ret;
248 fe25519 t, fd;
249 fe25519.fe25519_setone(&r->z);
250 fixed (Byte* ecdp = ecd) fe25519.fe25519_unpack(&fd, ecdp);
251 Byte par = (Byte)(p[31] >> 7);
252 fe25519.fe25519_unpack(&r->y, p);
253 fe25519.fe25519_square(&r->x, &r->y);
254 fe25519.fe25519_mul(&t, &r->x, &fd);
255 fe25519.fe25519_sub(&r->x, &r->x, &r->z);
256 fe25519.fe25519_add(&t, &r->z, &t);
257 fe25519.fe25519_invert(&t, &t);
258 fe25519.fe25519_mul(&r->x, &r->x, &t);
259 ret = fe25519.fe25519_sqrt_vartime(&r->x, &r->x, par);
260 fe25519.fe25519_mul(&r->t, &r->x, &r->y);
261 return ret;
262 }
263
264 public static unsafe void ge25519_pack(Byte* r, ge25519* p) { //unsigned char r[32]
265 fe25519 tx, ty, zi;
266 fe25519.fe25519_invert(&zi, &p->z);
267 fe25519.fe25519_mul(&tx, &p->x, &zi);
268 fe25519.fe25519_mul(&ty, &p->y, &zi);
269 fe25519.fe25519_pack(r, &ty);
270 r[31] ^= (Byte)(fe25519.fe25519_getparity(&tx) << 7);
271 }
272
273 public static unsafe void ge25519_add(ge25519* r, ge25519* p, ge25519* q) {
274 ge25519_p1p1 grp1p1;
275 add_p1p1(&grp1p1, p, q);
276 p1p1_to_p3(r, &grp1p1);
277 }
278
279 public static unsafe void ge25519_double(ge25519* r, ge25519* p) {
280 ge25519_p1p1 grp1p1;
281 dbl_p1p1(&grp1p1, (ge25519_p2*)p);
282 p1p1_to_p3(r, &grp1p1);
283 }
284
285 public static unsafe void ge25519_scalarmult(ge25519* r, ge25519* p, sc25519* s) {
286 int i, j, k;
287 ge25519 g;
288 fixed (Byte* ge25519_neutral_xp = ge25519_neutral_x) fe25519.fe25519_unpack(&g.x, ge25519_neutral_xp);
289 fixed (Byte* ge25519_neutral_yp = ge25519_neutral_y) fe25519.fe25519_unpack(&g.y, ge25519_neutral_yp);
290 fixed (Byte* ge25519_neutral_zp = ge25519_neutral_z) fe25519.fe25519_unpack(&g.z, ge25519_neutral_zp);
291 fixed (Byte* ge25519_neutral_tp = ge25519_neutral_t) fe25519.fe25519_unpack(&g.t, ge25519_neutral_tp);
292
293 ge25519[] pre = new ge25519[(1 << WINDOWSIZE)];
294 ge25519 t;
295 ge25519_p1p1 tp1p1;
296 Byte w;
297 Byte* sb = stackalloc Byte[32];
298 sc25519.sc25519_to32bytes(sb, s);
299
300 // Precomputation
301 pre[0] = g;
302 pre[1] = *p;
303 for (i = 2; i < (1 << WINDOWSIZE); i += 2) {
304 fixed (ge25519* prep = pre) {
305 dbl_p1p1(&tp1p1, (ge25519_p2*)(prep + i / 2));
306 p1p1_to_p3(prep + i, &tp1p1);
307 add_p1p1(&tp1p1, prep + i, prep + 1);
308 p1p1_to_p3(prep + i + 1, &tp1p1);
309 }
310 }
311
312 // Fixed-window scalar multiplication
313 for (i = 32; i > 0; i--) {
314 for (j = 8 - WINDOWSIZE; j >= 0; j -= WINDOWSIZE) {
315 for (k = 0; k < WINDOWSIZE - 1; k++) {
316 dbl_p1p1(&tp1p1, (ge25519_p2*)&g);
317 p1p1_to_p2((ge25519_p2*)&g, &tp1p1);
318 }
319 dbl_p1p1(&tp1p1, (ge25519_p2*)&g);
320 p1p1_to_p3(&g, &tp1p1);
321 // Cache-timing resistant loading of precomputed value:
322 w = (Byte)((sb[i - 1] >> j) & WINDOWMASK);
323 t = pre[0];
324 for (k = 1; k < (1 << WINDOWSIZE); k++)
325 fixed (ge25519* prekp = &pre[k]) cmov_p3(&t, prekp, (k == w) ? (Byte)1 : (Byte)0);
326
327 add_p1p1(&tp1p1, &g, &t);
328 if (j != 0) p1p1_to_p2((ge25519_p2*)&g, &tp1p1);
329 else p1p1_to_p3(&g, &tp1p1); /* convert to p3 representation at the end */
330 }
331 }
332 r->x = g.x;
333 r->y = g.y;
334 r->z = g.z;
335 r->t = g.t;
336 }
337
338 public unsafe static void ge25519_scalarmult_base(ge25519* r, sc25519* s) {
339 /* XXX: Better algorithm for known-base-point scalar multiplication */
340 ge25519 t;
341 fixed (Byte* ge25519_base_xp = ge25519_base_x) fe25519.fe25519_unpack(&t.x, ge25519_base_xp);
342 fixed (Byte* ge25519_base_yp = ge25519_base_y) fe25519.fe25519_unpack(&t.y, ge25519_base_yp);
343 fixed (Byte* ge25519_base_zp = ge25519_base_z) fe25519.fe25519_unpack(&t.z, ge25519_base_zp);
344 fixed (Byte* ge25519_base_tp = ge25519_base_t) fe25519.fe25519_unpack(&t.t, ge25519_base_tp);
345 ge25519_scalarmult(r, &t, s);
346 }
347 }
348 unsafe struct fe25519 {
349 public fixed UInt32 v[32]; // crypto_uint32 v[32];
350
351 const int WINDOWSIZE = 4; //#define WINDOWSIZE 4 /* Should be 1,2, or 4 */
352 const int WINDOWMASK = ((1 << WINDOWSIZE) - 1); //#define WINDOWMASK ((1<<WINDOWSIZE)-1)
353
354 static unsafe void reduce_add_sub(fe25519* r) {
355 for (int rep = 0; rep < 4; rep++) {
356 UInt32 t = r->v[31] >> 7;
357 r->v[31] &= 127;
358 t *= 19;
359 r->v[0] += t;
360 for (int i = 0; i < 31; i++) {
361 t = r->v[i] >> 8;
362 r->v[i + 1] += t;
363 r->v[i] &= 255;
364 }
365 }
366 }
367
368 unsafe static void reduce_mul(fe25519* r) {
369 for (int rep = 0; rep < 2; rep++) {
370 UInt32 t = r->v[31] >> 7;
371 r->v[31] &= 127;
372 t *= 19;
373 r->v[0] += t;
374 for (int i = 0; i < 31; i++) {
375 t = r->v[i] >> 8;
376 r->v[i + 1] += t;
377 r->v[i] &= 255;
378 }
379 }
380 }
381
382 /* reduction modulo 2^255-19 */
383 unsafe static void freeze(fe25519* r) {
384 UInt32 m = (r->v[31] == 127) ? 1u : 0;
385 for (int i = 30; i > 1; i--)
386 m *= (r->v[i] == 255) ? 1u : 0;
387 m *= (r->v[0] >= 237) ? 1u : 0;
388
389 r->v[31] -= m * 127;
390 for (int i = 30; i > 0; i--)
391 r->v[i] -= m * 255;
392 r->v[0] -= m * 237;
393 }
394
395 /*freeze input before calling isone*/
396 unsafe static Boolean isone(fe25519* x) {
397 bool r = x->v[0] == 1;
398 for (int i = 1; i < 32; i++) r &= (x->v[i] == 0);
399 return r;
400 }
401
402 /*freeze input before calling iszero*/
403 unsafe static Boolean iszero(fe25519* x) {
404 bool r = (x->v[0] == 0);
405 for (int i = 1; i < 32; i++) r &= (x->v[i] == 0);
406 return r;
407 }
408
409
410 unsafe static Boolean issquare(fe25519* x) {
411 Byte[] e = new Byte[32] { 0xf6, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x3f }; /* (p-1)/2 */
412 fe25519 t;
413
414 fixed (Byte* ep = e) fe25519_pow(&t, x, ep);
415 freeze(&t);
416 return isone(&t) || iszero(&t);
417 }
418
419 public static unsafe void fe25519_unpack(fe25519* r, Byte* x) { //const unsigned char x[32]
420 for (int i = 0; i < 32; i++) r->v[i] = x[i];
421 r->v[31] &= 127;
422 }
423
424 /* Assumes input x being reduced mod 2^255 */
425 public static unsafe void fe25519_pack(Byte* r, fe25519* x) { //unsigned char r[32]
426 for (int i = 0; i < 32; i++)
427 r[i] = (byte)x->v[i];
428
429 /* freeze byte array */
430 UInt32 m = (r[31] == 127) ? 1u : 0; /* XXX: some compilers might use branches; fix */
431 for (int i = 30; i > 1; i--)
432 m *= (r[i] == 255) ? 1u : 0;
433 m *= (r[0] >= 237) ? 1u : 0;
434 r[31] -= (byte)(m * 127);
435 for (int i = 30; i > 0; i--)
436 r[i] -= (byte)(m * 255);
437 r[0] -= (byte)(m * 237);
438 }
439
440 public static unsafe void fe25519_cmov(fe25519* r, fe25519* x, Byte b) {
441 Byte nb = (Byte)(1 - b);
442 for (int i = 0; i < 32; i++) r->v[i] = nb * r->v[i] + b * x->v[i];
443 }
444
445 public static unsafe Byte fe25519_getparity(fe25519* x) {
446 fe25519 t = new fe25519();
447 for (int i = 0; i < 32; i++) t.v[i] = x->v[i];
448 freeze(&t);
449 return (Byte)(t.v[0] & 1);
450 }
451
452 public static unsafe void fe25519_setone(fe25519* r) {
453 r->v[0] = 1;
454 for (int i = 1; i < 32; i++) r->v[i] = 0;
455 }
456
457 static unsafe void fe25519_setzero(fe25519* r) {
458 for (int i = 0; i < 32; i++) r->v[i] = 0;
459 }
460
461 public static unsafe void fe25519_neg(fe25519* r, fe25519* x) {
462 fe25519 t = new fe25519();
463 for (int i = 0; i < 32; i++) t.v[i] = x->v[i];
464 fe25519_setzero(r);
465 fe25519_sub(r, r, &t);
466 }
467
468 public static unsafe void fe25519_add(fe25519* r, fe25519* x, fe25519* y) {
469 for (int i = 0; i < 32; i++) r->v[i] = x->v[i] + y->v[i];
470 reduce_add_sub(r);
471 }
472
473 public static unsafe void fe25519_sub(fe25519* r, fe25519* x, fe25519* y) {
474 UInt32* t = stackalloc UInt32[32];
475 t[0] = x->v[0] + 0x1da;
476 t[31] = x->v[31] + 0xfe;
477 for (int i = 1; i < 31; i++) t[i] = x->v[i] + 0x1fe;
478 for (int i = 0; i < 32; i++) r->v[i] = t[i] - y->v[i];
479 reduce_add_sub(r);
480 }
481
482 public static unsafe void fe25519_mul(fe25519* r, fe25519* x, fe25519* y) {
483 UInt32* t = stackalloc UInt32[63];
484 for (int i = 0; i < 63; i++) t[i] = 0;
485 for (int i = 0; i < 32; i++)
486 for (int j = 0; j < 32; j++)
487 t[i + j] += x->v[i] * y->v[j];
488
489 for (int i = 32; i < 63; i++)
490 r->v[i - 32] = t[i - 32] + 38 * t[i];
491 r->v[31] = t[31]; /* result now in r[0]...r[31] */
492
493 reduce_mul(r);
494 }
495
496 public static unsafe void fe25519_square(fe25519* r, fe25519* x) {
497 fe25519_mul(r, x, x);
498 }
499
500 /*XXX: Make constant time! */
501 public static unsafe void fe25519_pow(fe25519* r, fe25519* x, Byte* e) {
502 /*
503 fe25519 g;
504 fe25519_setone(&g);
505 int i;
506 unsigned char j;
507 for(i=32;i>0;i--)
508 {
509 for(j=128;j>0;j>>=1)
510 {
511 fe25519_square(&g,&g);
512 if(e[i-1] & j)
513 fe25519_mul(&g,&g,x);
514 }
515 }
516 for(i=0;i<32;i++) r->v[i] = g.v[i];
517 */
518 fe25519 g;
519 fe25519_setone(&g);
520 fe25519[] pre = new fe25519[(1 << WINDOWSIZE)];
521 fe25519 t;
522 Byte w;
523
524 // Precomputation
525 fixed (fe25519* prep = pre) fe25519_setone(prep);
526 pre[1] = *x;
527 for (int i = 2; i < (1 << WINDOWSIZE); i += 2) {
528 fixed (fe25519* prep = pre) {
529 fe25519_square(prep + i, prep + i / 2);
530 fe25519_mul(prep + i + 1, prep + i, prep + 1);
531 }
532 }
533
534 // Fixed-window scalar multiplication
535 for (int i = 32; i > 0; i--) {
536 for (int j = 8 - WINDOWSIZE; j >= 0; j -= WINDOWSIZE) {
537 for (int k = 0; k < WINDOWSIZE; k++)
538 fe25519_square(&g, &g);
539 // Cache-timing resistant loading of precomputed value:
540 w = (Byte)((e[i - 1] >> j) & WINDOWMASK);
541 t = pre[0];
542 for (int k = 1; k < (1 << WINDOWSIZE); k++) fixed (fe25519* prekp = &pre[k]) fe25519_cmov(&t, prekp, (k == w) ? (Byte)1 : (Byte)0);
543 fe25519_mul(&g, &g, &t);
544 }
545 }
546 *r = g;
547 }
548
549 /* Return 0 on success, 1 otherwise */
550 public static unsafe Boolean fe25519_sqrt_vartime(fe25519* r, fe25519* x, Byte parity) {
551 /* See HAC, Alg. 3.37 */
552 if (!issquare(x)) return true;
553 Byte[] e = new Byte[32] { 0xfb, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x1f }; /* (p-1)/4 */
554 Byte[] e2 = new Byte[32] { 0xfe, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x0f }; /* (p+3)/8 */
555 Byte[] e3 = new Byte[32] { 0xfd, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x0f }; /* (p-5)/8 */
556 fe25519 p = new fe25519(); // { { 0 } };
557 fe25519 d;
558 fixed (Byte* ep = e) fe25519.fe25519_pow(&d, x, ep);
559 freeze(&d);
560 if (isone(&d))
561 fixed (Byte* e2p = e2) fe25519.fe25519_pow(r, x, e2p);
562 else {
563 for (int i = 0; i < 32; i++)
564 d.v[i] = 4 * x->v[i];
565 fixed (Byte* e3p = e3) fe25519.fe25519_pow(&d, &d, e3p);
566 for (int i = 0; i < 32; i++)
567 r->v[i] = 2 * x->v[i];
568 fe25519_mul(r, r, &d);
569 }
570 freeze(r);
571 if ((r->v[0] & 1) != (parity & 1)) {
572 fe25519_sub(r, &p, r);
573 }
574 return false;
575 }
576
577 public static unsafe void fe25519_invert(fe25519* r, fe25519* x) {
578 fe25519 z2;
579 fe25519 z9;
580 fe25519 z11;
581 fe25519 z2_5_0;
582 fe25519 z2_10_0;
583 fe25519 z2_20_0;
584 fe25519 z2_50_0;
585 fe25519 z2_100_0;
586 fe25519 t0;
587 fe25519 t1;
588
589 /* 2 */
590 fe25519_square(&z2, x);
591 /* 4 */
592 fe25519_square(&t1, &z2);
593 /* 8 */
594 fe25519_square(&t0, &t1);
595 /* 9 */
596 fe25519_mul(&z9, &t0, x);
597 /* 11 */
598 fe25519_mul(&z11, &z9, &z2);
599 /* 22 */
600 fe25519_square(&t0, &z11);
601 /* 2^5 - 2^0 = 31 */
602 fe25519_mul(&z2_5_0, &t0, &z9);
603
604 /* 2^6 - 2^1 */
605 fe25519_square(&t0, &z2_5_0);
606 /* 2^7 - 2^2 */
607 fe25519_square(&t1, &t0);
608 /* 2^8 - 2^3 */
609 fe25519_square(&t0, &t1);
610 /* 2^9 - 2^4 */
611 fe25519_square(&t1, &t0);
612 /* 2^10 - 2^5 */
613 fe25519_square(&t0, &t1);
614 /* 2^10 - 2^0 */
615 fe25519_mul(&z2_10_0, &t0, &z2_5_0);
616
617 /* 2^11 - 2^1 */
618 fe25519_square(&t0, &z2_10_0);
619 /* 2^12 - 2^2 */
620 fe25519_square(&t1, &t0);
621 /* 2^20 - 2^10 */
622 for (int i = 2; i < 10; i += 2) { fe25519_square(&t0, &t1); fe25519_square(&t1, &t0); }
623 /* 2^20 - 2^0 */
624 fe25519_mul(&z2_20_0, &t1, &z2_10_0);
625
626 /* 2^21 - 2^1 */
627 fe25519_square(&t0, &z2_20_0);
628 /* 2^22 - 2^2 */
629 fe25519_square(&t1, &t0);
630 /* 2^40 - 2^20 */
631 for (int i = 2; i < 20; i += 2) { fe25519_square(&t0, &t1); fe25519_square(&t1, &t0); }
632 /* 2^40 - 2^0 */
633 fe25519_mul(&t0, &t1, &z2_20_0);
634
635 /* 2^41 - 2^1 */
636 fe25519_square(&t1, &t0);
637 /* 2^42 - 2^2 */
638 fe25519_square(&t0, &t1);
639 /* 2^50 - 2^10 */
640 for (int i = 2; i < 10; i += 2) { fe25519_square(&t1, &t0); fe25519_square(&t0, &t1); }
641 /* 2^50 - 2^0 */
642 fe25519_mul(&z2_50_0, &t0, &z2_10_0);
643
644 /* 2^51 - 2^1 */
645 fe25519_square(&t0, &z2_50_0);
646 /* 2^52 - 2^2 */
647 fe25519_square(&t1, &t0);
648 /* 2^100 - 2^50 */
649 for (int i = 2; i < 50; i += 2) { fe25519_square(&t0, &t1); fe25519_square(&t1, &t0); }
650 /* 2^100 - 2^0 */
651 fe25519_mul(&z2_100_0, &t1, &z2_50_0);
652
653 /* 2^101 - 2^1 */
654 fe25519_square(&t1, &z2_100_0);
655 /* 2^102 - 2^2 */
656 fe25519_square(&t0, &t1);
657 /* 2^200 - 2^100 */
658 for (int i = 2; i < 100; i += 2) { fe25519_square(&t1, &t0); fe25519_square(&t0, &t1); }
659 /* 2^200 - 2^0 */
660 fe25519_mul(&t1, &t0, &z2_100_0);
661
662 /* 2^201 - 2^1 */
663 fe25519_square(&t0, &t1);
664 /* 2^202 - 2^2 */
665 fe25519_square(&t1, &t0);
666 /* 2^250 - 2^50 */
667 for (int i = 2; i < 50; i += 2) { fe25519_square(&t0, &t1); fe25519_square(&t1, &t0); }
668 /* 2^250 - 2^0 */
669 fe25519_mul(&t0, &t1, &z2_50_0);
670
671 /* 2^251 - 2^1 */
672 fe25519_square(&t1, &t0);
673 /* 2^252 - 2^2 */
674 fe25519_square(&t0, &t1);
675 /* 2^253 - 2^3 */
676 fe25519_square(&t1, &t0);
677 /* 2^254 - 2^4 */
678 fe25519_square(&t0, &t1);
679 /* 2^255 - 2^5 */
680 fe25519_square(&t1, &t0);
681 /* 2^255 - 21 */
682 fe25519_mul(r, &t1, &z11);
683 }
684 }
685
686 public static unsafe void crypto_sign_keypair(out Byte[] pk, out Byte[] sk) {
687 sc25519 scsk;
688 ge25519 gepk;
689
690 sk = new Byte[SECRETKEYBYTES];
691 pk = new Byte[PUBLICKEYBYTES];
692 randombytes.generate(sk);
693 fixed (Byte* skp = sk) crypto_hash.sha512.crypto_hash(skp, skp, 32);
694 sk[0] &= 248;
695 sk[31] &= 127;
696 sk[31] |= 64;
697
698 fixed (Byte* skp = sk) sc25519.sc25519_from32bytes(&scsk, skp);
699
700 ge25519.ge25519_scalarmult_base(&gepk, &scsk);
701 fixed (Byte* pkp = pk) ge25519.ge25519_pack(pkp, &gepk);
702 }
703
704 public static unsafe Byte[] crypto_sign(Byte[] m, Byte[] sk) {
705 if (sk.Length != SECRETKEYBYTES) throw new ArgumentException("sk.Length != SECRETKEYBYTES");
706 Byte[] sm = new Byte[m.Length + 64];
707 UInt64 smlen;
708 fixed (Byte* smp = sm, mp = m, skp = sk) crypto_sign(smp, out smlen, mp, (ulong)m.Length, skp);
709 return sm;
710 }
711 public static unsafe void crypto_sign(Byte* sm, out UInt64 smlen, Byte* m, UInt64 mlen, Byte* sk) {
712 sc25519 sck, scs, scsk;
713 ge25519 ger;
714 Byte* r = stackalloc Byte[32];
715 Byte* s = stackalloc Byte[32];
716 Byte* hmg = stackalloc Byte[crypto_hash.sha512.BYTES];
717 Byte* hmr = stackalloc Byte[crypto_hash.sha512.BYTES];
718
719 smlen = mlen + 64;
720 for (UInt64 i = 0; i < mlen; i++) sm[32 + i] = m[i];
721 for (int i = 0; i < 32; i++) sm[i] = sk[32 + i];
722 crypto_hash.sha512.crypto_hash(hmg, sm, mlen + 32); /* Generate k as h(m,sk[32],...,sk[63]) */
723
724 sc25519.sc25519_from64bytes(&sck, hmg);
725 ge25519.ge25519_scalarmult_base(&ger, &sck);
726 ge25519.ge25519_pack(r, &ger);
727
728 for (int i = 0; i < 32; i++) sm[i] = r[i];
729
730 crypto_hash.sha512.crypto_hash(hmr, sm, mlen + 32); /* Compute h(m,r) */
731 sc25519.sc25519_from64bytes(&scs, hmr);
732 sc25519.sc25519_mul(&scs, &scs, &sck);
733
734 sc25519.sc25519_from32bytes(&scsk, sk);
735 sc25519.sc25519_add(&scs, &scs, &scsk);
736
737 sc25519.sc25519_to32bytes(s, &scs); /* cat s */
738 for (UInt64 i = 0; i < 32; i++)
739 sm[mlen + 32 + i] = s[i];
740 }
741
742 public static unsafe Byte[] crypto_sign_open(Byte[] sm, Byte[] pk) {
743 if (pk.Length != PUBLICKEYBYTES) throw new ArgumentException("pk.Length != PUBLICKEYBYTES");
744 Byte[] m = new Byte[sm.Length - 64];
745 UInt64 mlen;
746 fixed (Byte* smp = sm, mp = m, pkp = pk) {
747 if (crypto_sign_open(mp, out mlen, smp, (ulong)sm.Length, pkp) != 0) return null;
748 }
749 return m;
750 }
751 public static unsafe int crypto_sign_open(Byte* m, out UInt64 mlen, Byte* sm, UInt64 smlen, Byte* pk) {
752 mlen = 0;
753 if (smlen < 64) return -1;
754
755 Byte* t1 = stackalloc Byte[32], t2 = stackalloc Byte[32];
756 ge25519 get1, get2, gepk;
757 sc25519 schmr, scs;
758 Byte* hmr = stackalloc Byte[crypto_hash.sha512.BYTES];
759
760 if (ge25519.ge25519_unpack_vartime(&get1, sm)) return -1;
761 if (ge25519.ge25519_unpack_vartime(&gepk, pk)) return -1;
762
763 crypto_hash.sha512.crypto_hash(hmr, sm, smlen - 32);
764
765 sc25519.sc25519_from64bytes(&schmr, hmr);
766 ge25519.ge25519_scalarmult(&get1, &get1, &schmr);
767 ge25519.ge25519_add(&get1, &get1, &gepk);
768 ge25519.ge25519_pack(t1, &get1);
769
770 sc25519.sc25519_from32bytes(&scs, &sm[smlen - 32]);
771 ge25519.ge25519_scalarmult_base(&get2, &scs);
772 ge25519.ge25519_pack(t2, &get2);
773
774 if (m != null) for (UInt64 i = 0; i < smlen - 64; i++) m[i] = sm[i + 32];
775 mlen = smlen - 64;
776
777 return crypto_verify._32.crypto_verify(t1, t2);
778 }
779 }
780 }