]> Git Repo - secp256k1.git/blame - src/ecdsa_impl.h
Fix possible integer overflow in DER parsing
[secp256k1.git] / src / ecdsa_impl.h
CommitLineData
71712b27 1/**********************************************************************
fea19e7b 2 * Copyright (c) 2013-2015 Pieter Wuille *
71712b27
GM
3 * Distributed under the MIT software license, see the accompanying *
4 * file COPYING or http://www.opensource.org/licenses/mit-license.php.*
5 **********************************************************************/
6
0a433ea2 7
abe2d3e8
DR
8#ifndef SECP256K1_ECDSA_IMPL_H
9#define SECP256K1_ECDSA_IMPL_H
7a4b7691 10
f24041d6 11#include "scalar.h"
11ab5622
PW
12#include "field.h"
13#include "group.h"
14#include "ecmult.h"
949c1ebb 15#include "ecmult_gen.h"
11ab5622 16#include "ecdsa.h"
607884fc 17
6efd6e77
GM
18/** Group order for secp256k1 defined as 'n' in "Standards for Efficient Cryptography" (SEC2) 2.7.1
19 * sage: for t in xrange(1023, -1, -1):
20 * .. p = 2**256 - 2**32 - t
21 * .. if p.is_prime():
22 * .. print '%x'%p
23 * .. break
24 * 'fffffffffffffffffffffffffffffffffffffffffffffffffffffffefffffc2f'
25 * sage: a = 0
26 * sage: b = 7
27 * sage: F = FiniteField (p)
28 * sage: '%x' % (EllipticCurve ([F (a), F (b)]).order())
29 * 'fffffffffffffffffffffffffffffffebaaedce6af48a03bbfd25e8cd0364141'
30 */
dd891e0e 31static const secp256k1_fe secp256k1_ecdsa_const_order_as_fe = SECP256K1_FE_CONST(
4732d260
PW
32 0xFFFFFFFFUL, 0xFFFFFFFFUL, 0xFFFFFFFFUL, 0xFFFFFFFEUL,
33 0xBAAEDCE6UL, 0xAF48A03BUL, 0xBFD25E8CUL, 0xD0364141UL
34);
f24041d6 35
6efd6e77
GM
36/** Difference between field and order, values 'p' and 'n' values defined in
37 * "Standards for Efficient Cryptography" (SEC2) 2.7.1.
38 * sage: p = 0xFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFEFFFFFC2F
39 * sage: a = 0
40 * sage: b = 7
41 * sage: F = FiniteField (p)
42 * sage: '%x' % (p - EllipticCurve ([F (a), F (b)]).order())
43 * '14551231950b75fc4402da1722fc9baee'
44 */
dd891e0e 45static const secp256k1_fe secp256k1_ecdsa_const_p_minus_order = SECP256K1_FE_CONST(
4732d260
PW
46 0, 0, 0, 1, 0x45512319UL, 0x50B75FC4UL, 0x402DA172UL, 0x2FC9BAEEUL
47);
f24041d6 48
3bb9c447
PW
49static int secp256k1_der_read_len(const unsigned char **sigp, const unsigned char *sigend) {
50 int lenleft, b1;
51 size_t ret = 0;
52 if (*sigp >= sigend) {
53 return -1;
26320197 54 }
3bb9c447
PW
55 b1 = *((*sigp)++);
56 if (b1 == 0xFF) {
57 /* X.690-0207 8.1.3.5.c the value 0xFF shall not be used. */
58 return -1;
26320197 59 }
3bb9c447
PW
60 if ((b1 & 0x80) == 0) {
61 /* X.690-0207 8.1.3.4 short form length octets */
62 return b1;
26320197 63 }
3bb9c447
PW
64 if (b1 == 0x80) {
65 /* Indefinite length is not allowed in DER. */
66 return -1;
67 }
68 /* X.690-207 8.1.3.5 long form length octets */
3cb057f8 69 lenleft = b1 & 0x7F; /* lenleft is at least 1 */
3bb9c447
PW
70 if (lenleft > sigend - *sigp) {
71 return -1;
72 }
73 if (**sigp == 0) {
74 /* Not the shortest possible length encoding. */
75 return -1;
76 }
77 if ((size_t)lenleft > sizeof(size_t)) {
269d4227
GM
78 /* The resulting length would exceed the range of a size_t, so
79 * certainly longer than the passed array size.
80 */
3bb9c447 81 return -1;
26320197 82 }
3bb9c447 83 while (lenleft > 0) {
3bb9c447 84 ret = (ret << 8) | **sigp;
3bb9c447
PW
85 (*sigp)++;
86 lenleft--;
87 }
3cb057f8
TR
88 if (ret > (size_t)(sigend - *sigp)) {
89 /* Result exceeds the length of the passed array. */
90 return -1;
91 }
3bb9c447
PW
92 if (ret < 128) {
93 /* Not the shortest possible length encoding. */
94 return -1;
95 }
96 return ret;
97}
98
99static int secp256k1_der_parse_integer(secp256k1_scalar *r, const unsigned char **sig, const unsigned char *sigend) {
100 int overflow = 0;
101 unsigned char ra[32] = {0};
102 int rlen;
103
104 if (*sig == sigend || **sig != 0x02) {
105 /* Not a primitive integer (X.690-0207 8.3.1). */
26320197
GM
106 return 0;
107 }
3bb9c447
PW
108 (*sig)++;
109 rlen = secp256k1_der_read_len(sig, sigend);
110 if (rlen <= 0 || (*sig) + rlen > sigend) {
111 /* Exceeds bounds or not at least length 1 (X.690-0207 8.3.1). */
26320197
GM
112 return 0;
113 }
3bb9c447
PW
114 if (**sig == 0x00 && rlen > 1 && (((*sig)[1]) & 0x80) == 0x00) {
115 /* Excessive 0x00 padding. */
26320197
GM
116 return 0;
117 }
3bb9c447
PW
118 if (**sig == 0xFF && rlen > 1 && (((*sig)[1]) & 0x80) == 0x80) {
119 /* Excessive 0xFF padding. */
26320197
GM
120 return 0;
121 }
3bb9c447
PW
122 if ((**sig & 0x80) == 0x80) {
123 /* Negative. */
124 overflow = 1;
125 }
126 while (rlen > 0 && **sig == 0) {
127 /* Skip leading zero bytes */
128 rlen--;
129 (*sig)++;
130 }
131 if (rlen > 32) {
132 overflow = 1;
f24041d6 133 }
3bb9c447
PW
134 if (!overflow) {
135 memcpy(ra + 32 - rlen, *sig, rlen);
136 secp256k1_scalar_set_b32(r, ra, &overflow);
137 }
138 if (overflow) {
139 secp256k1_scalar_set_int(r, 0);
140 }
141 (*sig) += rlen;
142 return 1;
143}
144
145static int secp256k1_ecdsa_sig_parse(secp256k1_scalar *rr, secp256k1_scalar *rs, const unsigned char *sig, size_t size) {
146 const unsigned char *sigend = sig + size;
147 int rlen;
148 if (sig == sigend || *(sig++) != 0x30) {
149 /* The encoding doesn't start with a constructed sequence (X.690-0207 8.9.1). */
26320197
GM
150 return 0;
151 }
3bb9c447
PW
152 rlen = secp256k1_der_read_len(&sig, sigend);
153 if (rlen < 0 || sig + rlen > sigend) {
154 /* Tuple exceeds bounds */
155 return 0;
f24041d6 156 }
3bb9c447
PW
157 if (sig + rlen != sigend) {
158 /* Garbage after tuple. */
26320197
GM
159 return 0;
160 }
3bb9c447
PW
161
162 if (!secp256k1_der_parse_integer(rr, &sig, sigend)) {
26320197
GM
163 return 0;
164 }
3bb9c447 165 if (!secp256k1_der_parse_integer(rs, &sig, sigend)) {
26320197
GM
166 return 0;
167 }
3bb9c447
PW
168
169 if (sig != sigend) {
170 /* Trailing garbage inside tuple. */
171 return 0;
172 }
173
d41e93a5 174 return 1;
607884fc
PW
175}
176
dd891e0e 177static int secp256k1_ecdsa_sig_serialize(unsigned char *sig, size_t *size, const secp256k1_scalar* ar, const secp256k1_scalar* as) {
f24041d6 178 unsigned char r[33] = {0}, s[33] = {0};
f24041d6 179 unsigned char *rp = r, *sp = s;
788038d3 180 size_t lenR = 33, lenS = 33;
18c329c5
PW
181 secp256k1_scalar_get_b32(&r[1], ar);
182 secp256k1_scalar_get_b32(&s[1], as);
f24041d6
PW
183 while (lenR > 1 && rp[0] == 0 && rp[1] < 0x80) { lenR--; rp++; }
184 while (lenS > 1 && sp[0] == 0 && sp[1] < 0x80) { lenS--; sp++; }
26320197 185 if (*size < 6+lenS+lenR) {
74a2acdb 186 *size = 6 + lenS + lenR;
d41e93a5 187 return 0;
26320197 188 }
0a07e62f
PW
189 *size = 6 + lenS + lenR;
190 sig[0] = 0x30;
191 sig[1] = 4 + lenS + lenR;
192 sig[2] = 0x02;
193 sig[3] = lenR;
f24041d6 194 memcpy(sig+4, rp, lenR);
0a07e62f
PW
195 sig[4+lenR] = 0x02;
196 sig[5+lenR] = lenS;
f24041d6 197 memcpy(sig+lenR+6, sp, lenS);
d41e93a5 198 return 1;
0a07e62f
PW
199}
200
dd891e0e 201static int secp256k1_ecdsa_sig_verify(const secp256k1_ecmult_context *ctx, const secp256k1_scalar *sigr, const secp256k1_scalar *sigs, const secp256k1_ge *pubkey, const secp256k1_scalar *message) {
792bcdb0 202 unsigned char c[32];
dd891e0e 203 secp256k1_scalar sn, u1, u2;
b4ceedf1 204#if !defined(EXHAUSTIVE_TEST_ORDER)
dd891e0e 205 secp256k1_fe xr;
b4ceedf1 206#endif
dd891e0e
PW
207 secp256k1_gej pubkeyj;
208 secp256k1_gej pr;
792bcdb0 209
18c329c5 210 if (secp256k1_scalar_is_zero(sigr) || secp256k1_scalar_is_zero(sigs)) {
d41e93a5 211 return 0;
26320197 212 }
607884fc 213
18c329c5 214 secp256k1_scalar_inverse_var(&sn, sigs);
f24041d6 215 secp256k1_scalar_mul(&u1, &sn, message);
18c329c5 216 secp256k1_scalar_mul(&u2, &sn, sigr);
792bcdb0 217 secp256k1_gej_set_ge(&pubkeyj, pubkey);
a9b6595e 218 secp256k1_ecmult(ctx, &pr, &pubkeyj, &u2, &u1);
ce7eb6fb
PW
219 if (secp256k1_gej_is_infinity(&pr)) {
220 return 0;
221 }
b4ceedf1
AP
222
223#if defined(EXHAUSTIVE_TEST_ORDER)
224{
225 secp256k1_scalar computed_r;
b4ceedf1
AP
226 secp256k1_ge pr_ge;
227 secp256k1_ge_set_gej(&pr_ge, &pr);
228 secp256k1_fe_normalize(&pr_ge.x);
229
230 secp256k1_fe_get_b32(c, &pr_ge.x);
678b0e54 231 secp256k1_scalar_set_b32(&computed_r, c, NULL);
b4ceedf1
AP
232 return secp256k1_scalar_eq(sigr, &computed_r);
233}
234#else
18c329c5 235 secp256k1_scalar_get_b32(c, sigr);
ce7eb6fb 236 secp256k1_fe_set_b32(&xr, c);
13278f64 237
3627437d
GM
238 /** We now have the recomputed R point in pr, and its claimed x coordinate (modulo n)
239 * in xr. Naively, we would extract the x coordinate from pr (requiring a inversion modulo p),
240 * compute the remainder modulo n, and compare it to xr. However:
241 *
242 * xr == X(pr) mod n
243 * <=> exists h. (xr + h * n < p && xr + h * n == X(pr))
244 * [Since 2 * n > p, h can only be 0 or 1]
245 * <=> (xr == X(pr)) || (xr + n < p && xr + n == X(pr))
246 * [In Jacobian coordinates, X(pr) is pr.x / pr.z^2 mod p]
247 * <=> (xr == pr.x / pr.z^2 mod p) || (xr + n < p && xr + n == pr.x / pr.z^2 mod p)
248 * [Multiplying both sides of the equations by pr.z^2 mod p]
249 * <=> (xr * pr.z^2 mod p == pr.x) || (xr + n < p && (xr + n) * pr.z^2 mod p == pr.x)
250 *
251 * Thus, we can avoid the inversion, but we have to check both cases separately.
252 * secp256k1_gej_eq_x implements the (xr * pr.z^2 mod p == pr.x) test.
253 */
ce7eb6fb 254 if (secp256k1_gej_eq_x_var(&xr, &pr)) {
6c476a8a 255 /* xr * pr.z^2 mod p == pr.x, so the signature is valid. */
ce7eb6fb
PW
256 return 1;
257 }
4732d260 258 if (secp256k1_fe_cmp_var(&xr, &secp256k1_ecdsa_const_p_minus_order) >= 0) {
6c476a8a 259 /* xr + n >= p, so we can skip testing the second case. */
ce7eb6fb 260 return 0;
4adf6b2a 261 }
4732d260 262 secp256k1_fe_add(&xr, &secp256k1_ecdsa_const_order_as_fe);
ce7eb6fb 263 if (secp256k1_gej_eq_x_var(&xr, &pr)) {
3627437d 264 /* (xr + n) * pr.z^2 mod p == pr.x, so the signature is valid. */
ce7eb6fb
PW
265 return 1;
266 }
267 return 0;
b4ceedf1 268#endif
607884fc
PW
269}
270
dd891e0e 271static int secp256k1_ecdsa_sig_sign(const secp256k1_ecmult_gen_context *ctx, secp256k1_scalar *sigr, secp256k1_scalar *sigs, const secp256k1_scalar *seckey, const secp256k1_scalar *message, const secp256k1_scalar *nonce, int *recid) {
792bcdb0 272 unsigned char b[32];
dd891e0e
PW
273 secp256k1_gej rp;
274 secp256k1_ge r;
275 secp256k1_scalar n;
792bcdb0
GM
276 int overflow = 0;
277
a9b6595e 278 secp256k1_ecmult_gen(ctx, &rp, nonce);
50eb498e 279 secp256k1_ge_set_gej(&r, &rp);
50eb498e
PW
280 secp256k1_fe_normalize(&r.x);
281 secp256k1_fe_normalize(&r.y);
282 secp256k1_fe_get_b32(b, &r.x);
18c329c5 283 secp256k1_scalar_set_b32(sigr, b, &overflow);
25e3cfbf
AP
284 /* These two conditions should be checked before calling */
285 VERIFY_CHECK(!secp256k1_scalar_is_zero(sigr));
286 VERIFY_CHECK(overflow == 0);
287
26320197 288 if (recid) {
269d4227
GM
289 /* The overflow condition is cryptographically unreachable as hitting it requires finding the discrete log
290 * of some P where P.x >= order, and only 1 in about 2^127 points meet this criteria.
291 */
a9f5c8b8 292 *recid = (overflow ? 2 : 0) | (secp256k1_fe_is_odd(&r.y) ? 1 : 0);
26320197 293 }
18c329c5 294 secp256k1_scalar_mul(&n, sigr, seckey);
a9f5c8b8 295 secp256k1_scalar_add(&n, &n, message);
18c329c5
PW
296 secp256k1_scalar_inverse(sigs, nonce);
297 secp256k1_scalar_mul(sigs, sigs, &n);
a9f5c8b8 298 secp256k1_scalar_clear(&n);
2f6c8019
GM
299 secp256k1_gej_clear(&rp);
300 secp256k1_ge_clear(&r);
18c329c5 301 if (secp256k1_scalar_is_zero(sigs)) {
d41e93a5 302 return 0;
26320197 303 }
18c329c5
PW
304 if (secp256k1_scalar_is_high(sigs)) {
305 secp256k1_scalar_negate(sigs, sigs);
26320197 306 if (recid) {
50eb498e 307 *recid ^= 1;
26320197 308 }
50eb498e 309 }
eb0be8ee 310 return 1;
0a07e62f
PW
311}
312
abe2d3e8 313#endif /* SECP256K1_ECDSA_IMPL_H */
This page took 0.090984 seconds and 4 git commands to generate.