1 /*
2  * Copyright 2013 The Android Open Source Project
3  *
4  * Redistribution and use in source and binary forms, with or without
5  * modification, are permitted provided that the following conditions are met:
6  *     * Redistributions of source code must retain the above copyright
7  *       notice, this list of conditions and the following disclaimer.
8  *     * Redistributions in binary form must reproduce the above copyright
9  *       notice, this list of conditions and the following disclaimer in the
10  *       documentation and/or other materials provided with the distribution.
11  *     * Neither the name of Google Inc. nor the names of its contributors may
12  *       be used to endorse or promote products derived from this software
13  *       without specific prior written permission.
14  *
15  * THIS SOFTWARE IS PROVIDED BY Google Inc. ``AS IS'' AND ANY EXPRESS OR
16  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
17  * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
18  * EVENT SHALL Google Inc. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
19  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
20  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
21  * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
22  * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
23  * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
24  * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
25  */
26 
27 // This is an implementation of the P256 elliptic curve group. It's written to
28 // be portable 32-bit, although it's still constant-time.
29 //
30 // WARNING: Implementing these functions in a constant-time manner is far from
31 //          obvious. Be careful when touching this code.
32 //
33 // See http://www.imperialviolet.org/2010/12/04/ecc.html ([1]) for background.
34 
35 #include <assert.h>
36 #include <stdint.h>
37 #include <string.h>
38 #include <stdio.h>
39 
40 #include "mincrypt/p256.h"
41 
42 const p256_int SECP256r1_n =  // curve order
43   {{0xfc632551, 0xf3b9cac2, 0xa7179e84, 0xbce6faad, -1, -1, 0, -1}};
44 
45 const p256_int SECP256r1_p =  // curve field size
46   {{-1, -1, -1, 0, 0, 0, 1, -1 }};
47 
48 const p256_int SECP256r1_b =  // curve b
49   {{0x27d2604b, 0x3bce3c3e, 0xcc53b0f6, 0x651d06b0,
50     0x769886bc, 0xb3ebbd55, 0xaa3a93e7, 0x5ac635d8}};
51 
p256_init(p256_int * a)52 void p256_init(p256_int* a) {
53   memset(a, 0, sizeof(*a));
54 }
55 
p256_clear(p256_int * a)56 void p256_clear(p256_int* a) { p256_init(a); }
57 
p256_get_bit(const p256_int * scalar,int bit)58 int p256_get_bit(const p256_int* scalar, int bit) {
59   return (P256_DIGIT(scalar, bit / P256_BITSPERDIGIT)
60               >> (bit & (P256_BITSPERDIGIT - 1))) & 1;
61 }
62 
p256_is_zero(const p256_int * a)63 int p256_is_zero(const p256_int* a) {
64   int i, result = 0;
65   for (i = 0; i < P256_NDIGITS; ++i) result |= P256_DIGIT(a, i);
66   return !result;
67 }
68 
69 // top, c[] += a[] * b
70 // Returns new top
mulAdd(const p256_int * a,p256_digit b,p256_digit top,p256_digit * c)71 static p256_digit mulAdd(const p256_int* a,
72                          p256_digit b,
73                          p256_digit top,
74                          p256_digit* c) {
75   int i;
76   p256_ddigit carry = 0;
77 
78   for (i = 0; i < P256_NDIGITS; ++i) {
79     carry += *c;
80     carry += (p256_ddigit)P256_DIGIT(a, i) * b;
81     *c++ = (p256_digit)carry;
82     carry >>= P256_BITSPERDIGIT;
83   }
84   return top + (p256_digit)carry;
85 }
86 
87 // top, c[] -= top_a, a[]
subTop(p256_digit top_a,const p256_digit * a,p256_digit top_c,p256_digit * c)88 static p256_digit subTop(p256_digit top_a,
89                          const p256_digit* a,
90                          p256_digit top_c,
91                          p256_digit* c) {
92   int i;
93   p256_sddigit borrow = 0;
94 
95   for (i = 0; i < P256_NDIGITS; ++i) {
96     borrow += *c;
97     borrow -= *a++;
98     *c++ = (p256_digit)borrow;
99     borrow >>= P256_BITSPERDIGIT;
100   }
101   borrow += top_c;
102   borrow -= top_a;
103   top_c = (p256_digit)borrow;
104   assert((borrow >> P256_BITSPERDIGIT) == 0);
105   return top_c;
106 }
107 
108 // top, c[] -= MOD[] & mask (0 or -1)
109 // returns new top.
subM(const p256_int * MOD,p256_digit top,p256_digit * c,p256_digit mask)110 static p256_digit subM(const p256_int* MOD,
111                        p256_digit top,
112                        p256_digit* c,
113                        p256_digit mask) {
114   int i;
115   p256_sddigit borrow = 0;
116   for (i = 0; i < P256_NDIGITS; ++i) {
117     borrow += *c;
118     borrow -= P256_DIGIT(MOD, i) & mask;
119     *c++ = (p256_digit)borrow;
120     borrow >>= P256_BITSPERDIGIT;
121   }
122   return top + (p256_digit)borrow;
123 }
124 
125 // top, c[] += MOD[] & mask (0 or -1)
126 // returns new top.
addM(const p256_int * MOD,p256_digit top,p256_digit * c,p256_digit mask)127 static p256_digit addM(const p256_int* MOD,
128                        p256_digit top,
129                        p256_digit* c,
130                        p256_digit mask) {
131   int i;
132   p256_ddigit carry = 0;
133   for (i = 0; i < P256_NDIGITS; ++i) {
134     carry += *c;
135     carry += P256_DIGIT(MOD, i) & mask;
136     *c++ = (p256_digit)carry;
137     carry >>= P256_BITSPERDIGIT;
138   }
139   return top + (p256_digit)carry;
140 }
141 
142 // c = a * b mod MOD. c can be a and/or b.
p256_modmul(const p256_int * MOD,const p256_int * a,const p256_digit top_b,const p256_int * b,p256_int * c)143 void p256_modmul(const p256_int* MOD,
144                  const p256_int* a,
145                  const p256_digit top_b,
146                  const p256_int* b,
147                  p256_int* c) {
148   p256_digit tmp[P256_NDIGITS * 2 + 1] = { 0 };
149   p256_digit top = 0;
150   int i;
151 
152   // Multiply/add into tmp.
153   for (i = 0; i < P256_NDIGITS; ++i) {
154     if (i) tmp[i + P256_NDIGITS - 1] = top;
155     top = mulAdd(a, P256_DIGIT(b, i), 0, tmp + i);
156   }
157 
158   // Multiply/add top digit
159   tmp[i + P256_NDIGITS - 1] = top;
160   top = mulAdd(a, top_b, 0, tmp + i);
161 
162   // Reduce tmp, digit by digit.
163   for (; i >= 0; --i) {
164     p256_digit reducer[P256_NDIGITS] = { 0 };
165     p256_digit top_reducer;
166 
167     // top can be any value at this point.
168     // Guestimate reducer as top * MOD, since msw of MOD is -1.
169     top_reducer = mulAdd(MOD, top, 0, reducer);
170 
171     // Subtract reducer from top | tmp.
172     top = subTop(top_reducer, reducer, top, tmp + i);
173 
174     // top is now either 0 or 1. Make it 0, fixed-timing.
175     assert(top <= 1);
176 
177     top = subM(MOD, top, tmp + i, ~(top - 1));
178 
179     assert(top == 0);
180 
181     // We have now reduced the top digit off tmp. Fetch new top digit.
182     top = tmp[i + P256_NDIGITS - 1];
183   }
184 
185   // tmp might still be larger than MOD, yet same bit length.
186   // Make sure it is less, fixed-timing.
187   addM(MOD, 0, tmp, subM(MOD, 0, tmp, -1));
188 
189   memcpy(c, tmp, P256_NBYTES);
190 }
p256_is_odd(const p256_int * a)191 int p256_is_odd(const p256_int* a) { return P256_DIGIT(a, 0) & 1; }
p256_is_even(const p256_int * a)192 int p256_is_even(const p256_int* a) { return !(P256_DIGIT(a, 0) & 1); }
193 
p256_shl(const p256_int * a,int n,p256_int * b)194 p256_digit p256_shl(const p256_int* a, int n, p256_int* b) {
195   int i;
196   p256_digit top = P256_DIGIT(a, P256_NDIGITS - 1);
197 
198   n %= P256_BITSPERDIGIT;
199   for (i = P256_NDIGITS - 1; i > 0; --i) {
200     p256_digit accu = (P256_DIGIT(a, i) << n);
201     accu |= (P256_DIGIT(a, i - 1) >> (P256_BITSPERDIGIT - n));
202     P256_DIGIT(b, i) = accu;
203   }
204   P256_DIGIT(b, i) = (P256_DIGIT(a, i) << n);
205 
206   top = (p256_digit)((((p256_ddigit)top) << n) >> P256_BITSPERDIGIT);
207 
208   return top;
209 }
210 
p256_shr(const p256_int * a,int n,p256_int * b)211 void p256_shr(const p256_int* a, int n, p256_int* b) {
212   int i;
213 
214   n %= P256_BITSPERDIGIT;
215   for (i = 0; i < P256_NDIGITS - 1; ++i) {
216     p256_digit accu = (P256_DIGIT(a, i) >> n);
217     accu |= (P256_DIGIT(a, i + 1) << (P256_BITSPERDIGIT - n));
218     P256_DIGIT(b, i) = accu;
219   }
220   P256_DIGIT(b, i) = (P256_DIGIT(a, i) >> n);
221 }
222 
p256_shr1(const p256_int * a,int highbit,p256_int * b)223 static void p256_shr1(const p256_int* a, int highbit, p256_int* b) {
224   int i;
225 
226   for (i = 0; i < P256_NDIGITS - 1; ++i) {
227     p256_digit accu = (P256_DIGIT(a, i) >> 1);
228     accu |= (P256_DIGIT(a, i + 1) << (P256_BITSPERDIGIT - 1));
229     P256_DIGIT(b, i) = accu;
230   }
231   P256_DIGIT(b, i) = (P256_DIGIT(a, i) >> 1) |
232       (highbit << (P256_BITSPERDIGIT - 1));
233 }
234 
235 // Return -1, 0, 1 for a < b, a == b or a > b respectively.
p256_cmp(const p256_int * a,const p256_int * b)236 int p256_cmp(const p256_int* a, const p256_int* b) {
237   int i;
238   p256_sddigit borrow = 0;
239   p256_digit notzero = 0;
240 
241   for (i = 0; i < P256_NDIGITS; ++i) {
242     borrow += (p256_sddigit)P256_DIGIT(a, i) - P256_DIGIT(b, i);
243     // Track whether any result digit is ever not zero.
244     // Relies on !!(non-zero) evaluating to 1, e.g., !!(-1) evaluating to 1.
245     notzero |= !!((p256_digit)borrow);
246     borrow >>= P256_BITSPERDIGIT;
247   }
248   return (int)borrow | notzero;
249 }
250 
251 // c = a - b. Returns borrow: 0 or -1.
p256_sub(const p256_int * a,const p256_int * b,p256_int * c)252 int p256_sub(const p256_int* a, const p256_int* b, p256_int* c) {
253   int i;
254   p256_sddigit borrow = 0;
255 
256   for (i = 0; i < P256_NDIGITS; ++i) {
257     borrow += (p256_sddigit)P256_DIGIT(a, i) - P256_DIGIT(b, i);
258     if (c) P256_DIGIT(c, i) = (p256_digit)borrow;
259     borrow >>= P256_BITSPERDIGIT;
260   }
261   return (int)borrow;
262 }
263 
264 // c = a + b. Returns carry: 0 or 1.
p256_add(const p256_int * a,const p256_int * b,p256_int * c)265 int p256_add(const p256_int* a, const p256_int* b, p256_int* c) {
266   int i;
267   p256_ddigit carry = 0;
268 
269   for (i = 0; i < P256_NDIGITS; ++i) {
270     carry += (p256_ddigit)P256_DIGIT(a, i) + P256_DIGIT(b, i);
271     if (c) P256_DIGIT(c, i) = (p256_digit)carry;
272     carry >>= P256_BITSPERDIGIT;
273   }
274   return (int)carry;
275 }
276 
277 // b = a + d. Returns carry, 0 or 1.
p256_add_d(const p256_int * a,p256_digit d,p256_int * b)278 int p256_add_d(const p256_int* a, p256_digit d, p256_int* b) {
279   int i;
280   p256_ddigit carry = d;
281 
282   for (i = 0; i < P256_NDIGITS; ++i) {
283     carry += (p256_ddigit)P256_DIGIT(a, i);
284     if (b) P256_DIGIT(b, i) = (p256_digit)carry;
285     carry >>= P256_BITSPERDIGIT;
286   }
287   return (int)carry;
288 }
289 
290 // b = 1/a mod MOD, binary euclid.
p256_modinv_vartime(const p256_int * MOD,const p256_int * a,p256_int * b)291 void p256_modinv_vartime(const p256_int* MOD,
292                          const p256_int* a,
293                          p256_int* b) {
294   p256_int R = P256_ZERO;
295   p256_int S = P256_ONE;
296   p256_int U = *MOD;
297   p256_int V = *a;
298 
299   for (;;) {
300     if (p256_is_even(&U)) {
301       p256_shr1(&U, 0, &U);
302       if (p256_is_even(&R)) {
303         p256_shr1(&R, 0, &R);
304       } else {
305         // R = (R+MOD)/2
306         p256_shr1(&R, p256_add(&R, MOD, &R), &R);
307       }
308     } else if (p256_is_even(&V)) {
309       p256_shr1(&V, 0, &V);
310       if (p256_is_even(&S)) {
311         p256_shr1(&S, 0, &S);
312       } else {
313         // S = (S+MOD)/2
314         p256_shr1(&S, p256_add(&S, MOD, &S) , &S);
315       }
316     } else {  // U,V both odd.
317       if (!p256_sub(&V, &U, NULL)) {
318         p256_sub(&V, &U, &V);
319         if (p256_sub(&S, &R, &S)) p256_add(&S, MOD, &S);
320         if (p256_is_zero(&V)) break;  // done.
321       } else {
322         p256_sub(&U, &V, &U);
323         if (p256_sub(&R, &S, &R)) p256_add(&R, MOD, &R);
324       }
325     }
326   }
327 
328   p256_mod(MOD, &R, b);
329 }
330 
p256_mod(const p256_int * MOD,const p256_int * in,p256_int * out)331 void p256_mod(const p256_int* MOD,
332               const p256_int* in,
333               p256_int* out) {
334   if (out != in) *out = *in;
335   addM(MOD, 0, P256_DIGITS(out), subM(MOD, 0, P256_DIGITS(out), -1));
336 }
337 
338 // Verify y^2 == x^3 - 3x + b mod p
339 // and 0 < x < p and 0 < y < p
p256_is_valid_point(const p256_int * x,const p256_int * y)340 int p256_is_valid_point(const p256_int* x, const p256_int* y) {
341   p256_int y2, x3;
342 
343   if (p256_cmp(&SECP256r1_p, x) <= 0 ||
344       p256_cmp(&SECP256r1_p, y) <= 0 ||
345       p256_is_zero(x) ||
346       p256_is_zero(y)) return 0;
347 
348   p256_modmul(&SECP256r1_p, y, 0, y, &y2);  // y^2
349 
350   p256_modmul(&SECP256r1_p, x, 0, x, &x3);  // x^2
351   p256_modmul(&SECP256r1_p, x, 0, &x3, &x3);  // x^3
352   if (p256_sub(&x3, x, &x3)) p256_add(&x3, &SECP256r1_p, &x3);  // x^3 - x
353   if (p256_sub(&x3, x, &x3)) p256_add(&x3, &SECP256r1_p, &x3);  // x^3 - 2x
354   if (p256_sub(&x3, x, &x3)) p256_add(&x3, &SECP256r1_p, &x3);  // x^3 - 3x
355   if (p256_add(&x3, &SECP256r1_b, &x3))  // x^3 - 3x + b
356     p256_sub(&x3, &SECP256r1_p, &x3);
357 
358   return p256_cmp(&y2, &x3) == 0;
359 }
360 
p256_from_bin(const uint8_t src[P256_NBYTES],p256_int * dst)361 void p256_from_bin(const uint8_t src[P256_NBYTES], p256_int* dst) {
362   int i;
363   const uint8_t* p = &src[0];
364 
365   for (i = P256_NDIGITS - 1; i >= 0; --i) {
366     P256_DIGIT(dst, i) =
367         (p[0] << 24) |
368         (p[1] << 16) |
369         (p[2] << 8) |
370         p[3];
371     p += 4;
372   }
373 }
374