/usr/include/x86_64-linux-gnu/qcc/CryptoECCfp.h is in liballjoyn-common-dev-1604 16.04a-3.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 | #ifndef FIELD_256_H
#define FIELD_256_H
/**
* @file aj_crypto_fp.h Header file for field arithmetic for ECC.
*/
/******************************************************************************
* Copyright AllSeen Alliance. All rights reserved.
*
* Permission to use, copy, modify, and/or distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
******************************************************************************/
#include <qcc/platform.h>
#include <qcc/String.h>
#include <stdio.h>
#include <string.h>
using namespace std;
namespace qcc {
typedef enum { B_FALSE, B_TRUE } boolean_t;
/* Digit types for multiprecision integers.*/
typedef uint64_t digit_t;
typedef const uint64_t digit_tc;
typedef int64_t sdigit_t;
/* Number of bits in the large integer radix, i.e., digits are from the set {0, ..., 2^(RADIX_BITS) - 1}. */
#define RADIX_BITS (64)
/* The zero digit_t. */
#define DIGIT_ZERO ((digit_t) 0)
/* Convert a bitlength to the number of digit_t's required to represent it. */
#define NBITS_TO_NDIGITS(x) (((x) + RADIX_BITS - 1) / (RADIX_BITS))
/* Number of temps required by field arithmetic functions */
#define P256_TEMPS (2 * P256_DIGITS)
/* Number of digits required to represent a field element. */
#define P256_DIGITS ((256 + RADIX_BITS - 1) / RADIX_BITS)
/* Swap two values of the same type. */
#ifdef __cplusplus
#define SWAP(a, b) std::swap(a, b)
#else
#define SWAP(a, b) { \
(a) = (a) ^ (b); \
(b) = (a) ^ (b); \
(a) = (a) ^ (b); \
}
#endif
/* Multiprecision type to represent 256-bit field elements */
typedef digit_t digit256_t[P256_DIGITS];
typedef const digit256_t digit256_tc;
/**
* Add two field elements (modular addition).
*
* @param[in] addend1 The first addend.
* @param[in] paddend1 The second addend.
* @param[out] sum The sum addend1 + paddend1 (mod p256).
*
*/
void fpadd_p256(digit256_tc addend1, digit256_tc paddend1, digit256_t sum);
/**
* Set a field element to the value zero.
*
* @param[in,out] a The value to be zeroed.
*
* @remarks
* This function uses a platform-specific secure zero function,
* to ensure it will not be optimized away.
*/
void fpzero_p256(digit256_t a);
/**
* Test whether a field element is zero.
*
* @param[in] a The field element to test.
*
* @return TRUE if a is zero, and FALSE if a is nonzero.
*/
bool fpiszero_p256(digit256_t a);
/**
* Get the value P256, the prime that defines the field.
*
* @param[out] a The field element that will be set to P256.
*
*/
void fpgetprime_p256(digit256_t a);
/**
* Test whether a 256-bit value is a valid element of the field defiend
* by the prime P256.
*
* @param[in] a The field element to test.
*
* @return TRUE if a is in [0, P256-1], and FALSE otherwise.
*/
boolean_t fpvalidate_p256(digit256_tc a);
/**
* Test whether a 256-bit value is in [0, modulus-1].
*
* @param[in] a The 256-bit value to test.
*
* @return TRUE if a is in [0, modulus-1], and FALSE otherwise.
*/
boolean_t validate_256(digit256_tc a, digit256_tc modulus);
/**
* Test whether a digit is zero, in constant time.
*
* @param[in] x the digit to test.
*
* @return TRUE if x == 0, FALSE otherwise.
*/
boolean_t is_digit_zero_ct(digit_t x);
/**
* Test whether a digit is nonzero, in constant time.
*
* @param[in] x the digit to test.
*
* @return TRUE if x != 0, FALSE otherwise.
*/
boolean_t is_digit_nonzero_ct(digit_t x);
/**
* Field subtraction (modular subtraction).
*
* @param[in] minuend The field element to subtract from.
* @param[in] subtrahend The field element to subtract.
* @param[out] difference The output difference, minuend - subtrahend (mod p256).
*/
void fpsub_p256(digit256_tc minuend, digit256_tc subtrahend, digit256_t difference);
/**
* Negate a field element.
*
* @param[in,out] a The field element to be negated.
*
* @return If a is less than or equal to modulus returns "1" (TRUE), else returns "0" (FALSE).
*/
boolean_t fpneg_p256(digit256_t a);
/**
* Divide a field element by two.
*
* @param[in] numerator The numerator in the division.
* @param[out] quotient The quotient: numerator/2 (mod p256).
* @param[in,out] temps Temporary space for use by this function, must have digit length P256_TEMPS.
*
*/
void fpdiv2_p256(digit256_tc numerator, digit256_t quotient, digit_t* temps);
/**
* Modular multiplication.
*
* @param[in] multiplier The multiplier.
* @param[in] multiplicand The multiplicand.
* @param[out] product The product multiplier*multiplicand (mod p256).
* @param[in,out] temps Temporary space for use by this function, must have digit length P256_TEMPS.
*
*/
void fpmul_p256(digit256_tc multiplier, digit256_tc multiplicand, digit256_t product, digit_t* temps);
/**
* Modular squaring.
*
* @param[in] multiplier The value to be squared.
* @param[out] product The square mutiplier*multiplier (mod p256).
* @param[in,out] temps Temporary space for use by this function, must have digit length P256_TEMPS.
*
*/
void fpsqr_p256(digit256_tc multiplier, digit256_t product, digit_t* temps);
/**
* Copy one field element to another.
*
* @param[in] src The source field element.
* @param[out] dst The destination field element.
*/
void fpcopy_p256(digit256_tc src, digit256_t dst);
/**
* Check whether two field elements are equal.
*
* @param[in] a The first field element to compare.
* @param[in] b The second field element to compare.
*
* @return TRUE if the element two elements are equal, FALSE otherwise.
*
* @remarks
* Note that all inputs must be fully reduced mod p256, e.g., p+1 and 1 will not be considered equal.
* This should not be an issue since all outputs of this implementation are fully reduced.
*/
boolean_t fpequal_p256(digit256_tc a, digit256_tc b);
/**
* Compute the multiplicative inverse of a field element.
*
* @param[in] a The element to be inverted.
* @param[out] inv The output result 1/a (mod p256).
* @param[in,out] temps Temporary space for use by this function, must have digit length P256_TEMPS.
*
*/
void fpinv_p256(digit256_tc a, digit256_t inv, digit_t* temps);
/**
* Set a field element to a single digit value.
*
* @param[in] dig0 The value to assign.
* @param[out] a The field element to be assigned.
*
* @remarks For example, fpset_p256((digit_t)1, a) sets a to the value 1.
*/
void fpset_p256(digit_t dig0, digit256_t a);
/**
* Swaps the byte order of the digits in a field element. The order of digits
* is not changed.
* i.e., fpdigitswap_p256(a) does a[i] = ByteSwap(a[i])
*
* @param[in,out] a The field element to have its digits swapped
*/
void fpdigitswap_p256(digit256_t a);
/**
* Create a field element x from a byte string.
* Input buffer must have length sizeof(digit256_t)
* Inputs larger than P256 will be reduced mod P256.
*
* @param[in] bytes The byte array to import.
* @param[out] x The field element to create.
* @param[in,out] temps Temporary space for use by this function, must have digit length P256_TEMPS.
* @param[in] is_bigendian TRUE if bytes has big endian byte ordering, or FALSE if little endian ordering.
*
*/
void fpimport_p256(const uint8_t* bytes, digit256_t x, digit_t* temps, bool is_bigendian);
/**
* Test whether a field element is a square.
*
* @param[in] a The element to test.
* @param[in,out] temps Temporary space for use by this function.
*
* @return
* TRUE if the element is a square mod P256, FALSE otherwise.
*
* @remarks
* temps must have digit length P256_TEMPS.
*/
boolean_t fpissquare_p256(digit256_tc a, digit_t* temps);
/**
* Compute the square root of a field element (known to be a square).
*
* @param[in] a The element to compute the square root of.
* @param[out] sqrt The output result sqrt(a) (mod p256).
* @param[in,out] temps Temporary space for use by this function.
*
* @remarks
* temps must have digit length P256_TEMPS.
* If a is not a square, the returned value is incorrect. See
* fpissquare_p256 to test whether a is a square.
*/
void fpsqrt_p256(digit256_tc a, digit256_t sqrt, digit_t* temps);
} /*namespace qcc*/
#endif /* FIELD_P256_H */
|