Note: Not yet caught up with trunk. git-svn-id: svn://svn.code.sf.net/p/irrlicht/code/branches/ogl-es@6114 dfc29bdd-3216-0410-991c-e03cc46cb475
		
			
				
	
	
		
			630 lines
		
	
	
		
			23 KiB
		
	
	
	
		
			C++
		
	
	
	
	
	
			
		
		
	
	
			630 lines
		
	
	
		
			23 KiB
		
	
	
	
		
			C++
		
	
	
	
	
	
| /*
 | |
|  ---------------------------------------------------------------------------
 | |
|  Copyright (c) 2002, Dr Brian Gladman <                 >, Worcester, UK.
 | |
|  All rights reserved.
 | |
| 
 | |
|  LICENSE TERMS
 | |
| 
 | |
|  The free distribution and use of this software in both source and binary
 | |
|  form is allowed (with or without changes) provided that:
 | |
| 
 | |
|    1. distributions of this source code include the above copyright
 | |
|       notice, this list of conditions and the following disclaimer;
 | |
| 
 | |
|    2. distributions in binary form include the above copyright
 | |
|       notice, this list of conditions and the following disclaimer
 | |
|       in the documentation and/or other associated materials;
 | |
| 
 | |
|    3. the copyright holder's name is not used to endorse products
 | |
|       built using this software without specific written permission.
 | |
| 
 | |
|  ALTERNATIVELY, provided that this notice is retained in full, this product
 | |
|  may be distributed under the terms of the GNU General Public License (GPL),
 | |
|  in which case the provisions of the GPL apply INSTEAD OF those given above.
 | |
| 
 | |
|  DISCLAIMER
 | |
| 
 | |
|  This software is provided 'as is' with no explicit or implied warranties
 | |
|  in respect of its properties, including, but not limited to, correctness
 | |
|  and/or fitness for purpose.
 | |
|  ---------------------------------------------------------------------------
 | |
|  Issue Date: 26/08/2003
 | |
| 
 | |
|  This is a byte oriented version of SHA2 that operates on arrays of bytes
 | |
|  stored in memory. This code implements sha256, sha384 and sha512 but the
 | |
|  latter two functions rely on efficient 64-bit integer operations that
 | |
|  may not be very efficient on 32-bit machines
 | |
| 
 | |
|  The sha256 functions use a type 'sha256_ctx' to hold details of the
 | |
|  current hash state and uses the following three calls:
 | |
| 
 | |
|        void sha256_begin(sha256_ctx ctx[1])
 | |
|        void sha256_hash(const unsigned char data[],
 | |
|                             unsigned long len, sha256_ctx ctx[1])
 | |
|        void sha256_end(unsigned char hval[], sha256_ctx ctx[1])
 | |
| 
 | |
|  The first subroutine initialises a hash computation by setting up the
 | |
|  context in the sha256_ctx context. The second subroutine hashes 8-bit
 | |
|  bytes from array data[] into the hash state withinh sha256_ctx context,
 | |
|  the number of bytes to be hashed being given by the the unsigned long
 | |
|  integer len.  The third subroutine completes the hash calculation and
 | |
|  places the resulting digest value in the array of 8-bit bytes hval[].
 | |
| 
 | |
|  The sha384 and sha512 functions are similar and use the interfaces:
 | |
| 
 | |
|        void sha384_begin(sha384_ctx ctx[1]);
 | |
|        void sha384_hash(const unsigned char data[],
 | |
|                             unsigned long len, sha384_ctx ctx[1]);
 | |
|        void sha384_end(unsigned char hval[], sha384_ctx ctx[1]);
 | |
| 
 | |
|        void sha512_begin(sha512_ctx ctx[1]);
 | |
|        void sha512_hash(const unsigned char data[],
 | |
|                             unsigned long len, sha512_ctx ctx[1]);
 | |
|        void sha512_end(unsigned char hval[], sha512_ctx ctx[1]);
 | |
| 
 | |
|  In addition there is a function sha2 that can be used to call all these
 | |
|  functions using a call with a hash length parameter as follows:
 | |
| 
 | |
|        int sha2_begin(unsigned long len, sha2_ctx ctx[1]);
 | |
|        void sha2_hash(const unsigned char data[],
 | |
|                             unsigned long len, sha2_ctx ctx[1]);
 | |
|        void sha2_end(unsigned char hval[], sha2_ctx ctx[1]);
 | |
| 
 | |
|  My thanks to Erik Andersen <andersen@codepoet.org> for testing this code
 | |
|  on big-endian systems and for his assistance with corrections
 | |
| */
 | |
| 
 | |
| /* define the hash functions that you need          */
 | |
| 
 | |
| #define SHA_2           /* for dynamic hash length  */
 | |
| #define SHA_256
 | |
| #define SHA_384
 | |
| #define SHA_512
 | |
| 
 | |
| #include <string.h>     /* for memcpy() etc.        */
 | |
| #include <stdlib.h>     /* for _lrotr with VC++     */
 | |
| 
 | |
| #include "sha2.h"
 | |
| #include "../os.h"
 | |
| 
 | |
| /*  BYTE ORDER IN 32-BIT WORDS
 | |
| 
 | |
|     To obtain the highest speed on processors with 32-bit words, this code
 | |
|     needs to determine the byte order of the target machine. The following
 | |
|     block of code is an attempt to capture the most obvious ways in which
 | |
|     various environemnts define byte order. It may well fail, in which case
 | |
|     the definitions will need to be set by editing at the points marked
 | |
|     **** EDIT HERE IF NECESSARY **** below.  My thanks to Peter Gutmann for
 | |
|     some of these defines (from cryptlib).
 | |
| */
 | |
| 
 | |
| #define BRG_LITTLE_ENDIAN   1234 /* byte 0 is least significant (i386) */
 | |
| #define BRG_BIG_ENDIAN      4321 /* byte 0 is most significant (mc68k) */
 | |
| 
 | |
| #ifdef __BIG_ENDIAN__
 | |
| #define PLATFORM_BYTE_ORDER BRG_BIG_ENDIAN
 | |
| #else
 | |
| #define PLATFORM_BYTE_ORDER BRG_LITTLE_ENDIAN
 | |
| #endif
 | |
| 
 | |
| #ifdef _MSC_VER
 | |
| #pragma intrinsic(memcpy)
 | |
| #endif
 | |
| 
 | |
| #define rotr32(x,n)   (((x) >> n) | ((x) << (32 - n)))
 | |
| 
 | |
| #if !defined(bswap_32)
 | |
| #define bswap_32(x) irr::os::Byteswap::byteswap(x)
 | |
| #endif
 | |
| 
 | |
| #if (PLATFORM_BYTE_ORDER == BRG_LITTLE_ENDIAN)
 | |
| #define SWAP_BYTES
 | |
| #else
 | |
| #undef  SWAP_BYTES
 | |
| #endif
 | |
| 
 | |
| #if defined(SHA_2) || defined(SHA_256)
 | |
| 
 | |
| #define SHA256_MASK (SHA256_BLOCK_SIZE - 1)
 | |
| 
 | |
| #if defined(SWAP_BYTES)
 | |
| #define bsw_32(p,n) { int _i = (n); while(_i--) p[_i] = bswap_32(p[_i]); }
 | |
| #else
 | |
| #define bsw_32(p,n)
 | |
| #endif
 | |
| 
 | |
| /* SHA256 mixing function definitions   */
 | |
| 
 | |
| #if 0
 | |
| 
 | |
| #define ch(x,y,z)       (((x) & (y)) ^ (~(x) & (z)))
 | |
| #define maj(x,y,z)      (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z)))
 | |
| 
 | |
| #else   /* Thanks to Rich Schroeppel and Colin Plumb for the following      */
 | |
| 
 | |
| #define ch(x,y,z)       ((z) ^ ((x) & ((y) ^ (z))))
 | |
| #define maj(x,y,z)      (((x) & (y)) | ((z) & ((x) ^ (y))))
 | |
| 
 | |
| #endif
 | |
| 
 | |
| #define s256_0(x) (rotr32((x),  2) ^ rotr32((x), 13) ^ rotr32((x), 22))
 | |
| #define s256_1(x) (rotr32((x),  6) ^ rotr32((x), 11) ^ rotr32((x), 25))
 | |
| #define g256_0(x) (rotr32((x),  7) ^ rotr32((x), 18) ^ ((x) >>  3))
 | |
| #define g256_1(x) (rotr32((x), 17) ^ rotr32((x), 19) ^ ((x) >> 10))
 | |
| 
 | |
| /* rotated SHA256 round definition. Rather than swapping variables as in    */
 | |
| /* FIPS-180, different variables are 'rotated' on each round, returning     */
 | |
| /* to their starting positions every eight rounds                           */
 | |
| 
 | |
| #define h2(i) p[i & 15] += \
 | |
|     g256_1(p[(i + 14) & 15]) + p[(i + 9) & 15] + g256_0(p[(i + 1) & 15])
 | |
| 
 | |
| #define h2_cycle(i,j)  \
 | |
|     v[(7 - i) & 7] += (j ? h2(i) : p[i & 15]) + k256[i + j] \
 | |
|         + s256_1(v[(4 - i) & 7]) + ch(v[(4 - i) & 7], v[(5 - i) & 7], v[(6 - i) & 7]); \
 | |
|     v[(3 - i) & 7] += v[(7 - i) & 7]; \
 | |
|     v[(7 - i) & 7] += s256_0(v[(0 - i) & 7]) + maj(v[(0 - i) & 7], v[(1 - i) & 7], v[(2 - i) & 7])
 | |
| 
 | |
| /* SHA256 mixing data   */
 | |
| 
 | |
| const sha2_32t k256[64] =
 | |
| {   n_u32(428a2f98), n_u32(71374491), n_u32(b5c0fbcf), n_u32(e9b5dba5),
 | |
|     n_u32(3956c25b), n_u32(59f111f1), n_u32(923f82a4), n_u32(ab1c5ed5),
 | |
|     n_u32(d807aa98), n_u32(12835b01), n_u32(243185be), n_u32(550c7dc3),
 | |
|     n_u32(72be5d74), n_u32(80deb1fe), n_u32(9bdc06a7), n_u32(c19bf174),
 | |
|     n_u32(e49b69c1), n_u32(efbe4786), n_u32(0fc19dc6), n_u32(240ca1cc),
 | |
|     n_u32(2de92c6f), n_u32(4a7484aa), n_u32(5cb0a9dc), n_u32(76f988da),
 | |
|     n_u32(983e5152), n_u32(a831c66d), n_u32(b00327c8), n_u32(bf597fc7),
 | |
|     n_u32(c6e00bf3), n_u32(d5a79147), n_u32(06ca6351), n_u32(14292967),
 | |
|     n_u32(27b70a85), n_u32(2e1b2138), n_u32(4d2c6dfc), n_u32(53380d13),
 | |
|     n_u32(650a7354), n_u32(766a0abb), n_u32(81c2c92e), n_u32(92722c85),
 | |
|     n_u32(a2bfe8a1), n_u32(a81a664b), n_u32(c24b8b70), n_u32(c76c51a3),
 | |
|     n_u32(d192e819), n_u32(d6990624), n_u32(f40e3585), n_u32(106aa070),
 | |
|     n_u32(19a4c116), n_u32(1e376c08), n_u32(2748774c), n_u32(34b0bcb5),
 | |
|     n_u32(391c0cb3), n_u32(4ed8aa4a), n_u32(5b9cca4f), n_u32(682e6ff3),
 | |
|     n_u32(748f82ee), n_u32(78a5636f), n_u32(84c87814), n_u32(8cc70208),
 | |
|     n_u32(90befffa), n_u32(a4506ceb), n_u32(bef9a3f7), n_u32(c67178f2),
 | |
| };
 | |
| 
 | |
| /* SHA256 initialisation data */
 | |
| 
 | |
| const sha2_32t i256[8] =
 | |
| {
 | |
|     n_u32(6a09e667), n_u32(bb67ae85), n_u32(3c6ef372), n_u32(a54ff53a),
 | |
|     n_u32(510e527f), n_u32(9b05688c), n_u32(1f83d9ab), n_u32(5be0cd19)
 | |
| };
 | |
| 
 | |
| sha2_void sha256_begin(sha256_ctx ctx[1])
 | |
| {
 | |
|     ctx->count[0] = ctx->count[1] = 0;
 | |
|     memcpy(ctx->hash, i256, 8 * sizeof(sha2_32t));
 | |
| }
 | |
| 
 | |
| /* Compile 64 bytes of hash data into SHA256 digest value   */
 | |
| /* NOTE: this routine assumes that the byte order in the    */
 | |
| /* ctx->wbuf[] at this point is in such an order that low   */
 | |
| /* address bytes in the ORIGINAL byte stream placed in this */
 | |
| /* buffer will now go to the high end of words on BOTH big  */
 | |
| /* and little endian systems                                */
 | |
| 
 | |
| sha2_void sha256_compile(sha256_ctx ctx[1])
 | |
| {   sha2_32t    v[8], j, *p = ctx->wbuf;
 | |
| 
 | |
|     memcpy(v, ctx->hash, 8 * sizeof(sha2_32t));
 | |
| 
 | |
|     for(j = 0; j < 64; j += 16)
 | |
|     {
 | |
|         h2_cycle( 0, j); h2_cycle( 1, j); h2_cycle( 2, j); h2_cycle( 3, j);
 | |
|         h2_cycle( 4, j); h2_cycle( 5, j); h2_cycle( 6, j); h2_cycle( 7, j);
 | |
|         h2_cycle( 8, j); h2_cycle( 9, j); h2_cycle(10, j); h2_cycle(11, j);
 | |
|         h2_cycle(12, j); h2_cycle(13, j); h2_cycle(14, j); h2_cycle(15, j);
 | |
|     }
 | |
| 
 | |
|     ctx->hash[0] += v[0]; ctx->hash[1] += v[1]; ctx->hash[2] += v[2]; ctx->hash[3] += v[3];
 | |
|     ctx->hash[4] += v[4]; ctx->hash[5] += v[5]; ctx->hash[6] += v[6]; ctx->hash[7] += v[7];
 | |
| }
 | |
| 
 | |
| /* SHA256 hash data in an array of bytes into hash buffer   */
 | |
| /* and call the hash_compile function as required.          */
 | |
| 
 | |
| sha2_void sha256_hash(const unsigned char data[], unsigned long len, sha256_ctx ctx[1])
 | |
| {   sha2_32t pos = (sha2_32t)(ctx->count[0] & SHA256_MASK),
 | |
|              space = SHA256_BLOCK_SIZE - pos;
 | |
|     const unsigned char *sp = data;
 | |
| 
 | |
|     if((ctx->count[0] += len) < len)
 | |
|         ++(ctx->count[1]);
 | |
| 
 | |
|     while(len >= space)     /* tranfer whole blocks while possible  */
 | |
|     {
 | |
|         memcpy(((unsigned char*)ctx->wbuf) + pos, sp, space);
 | |
|         sp += space; len -= space; space = SHA256_BLOCK_SIZE; pos = 0;
 | |
|         bsw_32(ctx->wbuf, SHA256_BLOCK_SIZE >> 2)
 | |
|         sha256_compile(ctx);
 | |
|     }
 | |
| 
 | |
|     memcpy(((unsigned char*)ctx->wbuf) + pos, sp, len);
 | |
| }
 | |
| 
 | |
| /* SHA256 Final padding and digest calculation  */
 | |
| 
 | |
| static sha2_32t  m1[4] =
 | |
| {
 | |
|     n_u32(00000000), n_u32(ff000000), n_u32(ffff0000), n_u32(ffffff00)
 | |
| };
 | |
| 
 | |
| static sha2_32t  b1[4] =
 | |
| {
 | |
|     n_u32(80000000), n_u32(00800000), n_u32(00008000), n_u32(00000080)
 | |
| };
 | |
| 
 | |
| sha2_void sha256_end(unsigned char hval[], sha256_ctx ctx[1])
 | |
| {   sha2_32t    i = (sha2_32t)(ctx->count[0] & SHA256_MASK);
 | |
| 
 | |
|     bsw_32(ctx->wbuf, (i + 3) >> 2)
 | |
|     /* bytes in the buffer are now in an order in which references  */
 | |
|     /* to 32-bit words will put bytes with lower addresses into the */
 | |
|     /* top of 32 bit words on BOTH big and little endian machines   */
 | |
| 
 | |
|     /* we now need to mask valid bytes and add the padding which is */
 | |
|     /* a single 1 bit and as many zero bits as necessary.           */
 | |
|     ctx->wbuf[i >> 2] = (ctx->wbuf[i >> 2] & m1[i & 3]) | b1[i & 3];
 | |
| 
 | |
|     /* we need 9 or more empty positions, one for the padding byte  */
 | |
|     /* (above) and eight for the length count.  If there is not     */
 | |
|     /* enough space pad and empty the buffer                        */
 | |
|     if(i > SHA256_BLOCK_SIZE - 9)
 | |
|     {
 | |
|         if(i < 60) ctx->wbuf[15] = 0;
 | |
|         sha256_compile(ctx);
 | |
|         i = 0;
 | |
|     }
 | |
|     else    /* compute a word index for the empty buffer positions  */
 | |
|         i = (i >> 2) + 1;
 | |
| 
 | |
|     while(i < 14) /* and zero pad all but last two positions      */
 | |
|         ctx->wbuf[i++] = 0;
 | |
| 
 | |
|     /* the following 32-bit length fields are assembled in the      */
 | |
|     /* wrong byte order on little endian machines but this is       */
 | |
|     /* corrected later since they are only ever used as 32-bit      */
 | |
|     /* word values.                                                 */
 | |
| 
 | |
|     ctx->wbuf[14] = (ctx->count[1] << 3) | (ctx->count[0] >> 29);
 | |
|     ctx->wbuf[15] = ctx->count[0] << 3;
 | |
| 
 | |
|     sha256_compile(ctx);
 | |
| 
 | |
|     /* extract the hash value as bytes in case the hash buffer is   */
 | |
|     /* mislaigned for 32-bit words                                  */
 | |
|     for(i = 0; i < SHA256_DIGEST_SIZE; ++i)
 | |
|         hval[i] = (unsigned char)(ctx->hash[i >> 2] >> (8 * (~i & 3)));
 | |
| }
 | |
| 
 | |
| sha2_void sha256(unsigned char hval[], const unsigned char data[], unsigned long len)
 | |
| {   sha256_ctx  cx[1];
 | |
| 
 | |
|     sha256_begin(cx); sha256_hash(data, len, cx); sha256_end(hval, cx);
 | |
| }
 | |
| 
 | |
| #endif
 | |
| 
 | |
| #if defined(SHA_2) || defined(SHA_384) || defined(SHA_512)
 | |
| 
 | |
| #define SHA512_MASK (SHA512_BLOCK_SIZE - 1)
 | |
| 
 | |
| #define rotr64(x,n)   (((x) >> n) | ((x) << (64 - n)))
 | |
| 
 | |
| #if !defined(bswap_64)
 | |
| #define bswap_64(x) ((((sha2_64t)(bswap_32((sha2_32t)(x)))) << 32) | (bswap_32((sha2_32t)((x) >> 32))))
 | |
| #endif
 | |
| 
 | |
| #if defined(SWAP_BYTES)
 | |
| #define bsw_64(p,n) { int _i = (n); while(_i--) p[_i] = bswap_64(p[_i]); }
 | |
| #else
 | |
| #define bsw_64(p,n)
 | |
| #endif
 | |
| 
 | |
| /* SHA512 mixing function definitions   */
 | |
| 
 | |
| #define s512_0(x) (rotr64((x), 28) ^ rotr64((x), 34) ^ rotr64((x), 39))
 | |
| #define s512_1(x) (rotr64((x), 14) ^ rotr64((x), 18) ^ rotr64((x), 41))
 | |
| #define g512_0(x) (rotr64((x),  1) ^ rotr64((x),  8) ^ ((x) >>  7))
 | |
| #define g512_1(x) (rotr64((x), 19) ^ rotr64((x), 61) ^ ((x) >>  6))
 | |
| 
 | |
| /* rotated SHA512 round definition. Rather than swapping variables as in    */
 | |
| /* FIPS-180, different variables are 'rotated' on each round, returning     */
 | |
| /* to their starting positions every eight rounds                           */
 | |
| 
 | |
| #define h5(i) ctx->wbuf[i & 15] += \
 | |
|     g512_1(ctx->wbuf[(i + 14) & 15]) + ctx->wbuf[(i + 9) & 15] + g512_0(ctx->wbuf[(i + 1) & 15])
 | |
| 
 | |
| #define h5_cycle(i,j)  \
 | |
|     v[(7 - i) & 7] += (j ? h5(i) : ctx->wbuf[i & 15]) + k512[i + j] \
 | |
|         + s512_1(v[(4 - i) & 7]) + ch(v[(4 - i) & 7], v[(5 - i) & 7], v[(6 - i) & 7]); \
 | |
|     v[(3 - i) & 7] += v[(7 - i) & 7]; \
 | |
|     v[(7 - i) & 7] += s512_0(v[(0 - i) & 7]) + maj(v[(0 - i) & 7], v[(1 - i) & 7], v[(2 - i) & 7])
 | |
| 
 | |
| /* SHA384/SHA512 mixing data    */
 | |
| 
 | |
| const sha2_64t  k512[80] =
 | |
| {
 | |
|     n_u64(428a2f98d728ae22), n_u64(7137449123ef65cd),
 | |
|     n_u64(b5c0fbcfec4d3b2f), n_u64(e9b5dba58189dbbc),
 | |
|     n_u64(3956c25bf348b538), n_u64(59f111f1b605d019),
 | |
|     n_u64(923f82a4af194f9b), n_u64(ab1c5ed5da6d8118),
 | |
|     n_u64(d807aa98a3030242), n_u64(12835b0145706fbe),
 | |
|     n_u64(243185be4ee4b28c), n_u64(550c7dc3d5ffb4e2),
 | |
|     n_u64(72be5d74f27b896f), n_u64(80deb1fe3b1696b1),
 | |
|     n_u64(9bdc06a725c71235), n_u64(c19bf174cf692694),
 | |
|     n_u64(e49b69c19ef14ad2), n_u64(efbe4786384f25e3),
 | |
|     n_u64(0fc19dc68b8cd5b5), n_u64(240ca1cc77ac9c65),
 | |
|     n_u64(2de92c6f592b0275), n_u64(4a7484aa6ea6e483),
 | |
|     n_u64(5cb0a9dcbd41fbd4), n_u64(76f988da831153b5),
 | |
|     n_u64(983e5152ee66dfab), n_u64(a831c66d2db43210),
 | |
|     n_u64(b00327c898fb213f), n_u64(bf597fc7beef0ee4),
 | |
|     n_u64(c6e00bf33da88fc2), n_u64(d5a79147930aa725),
 | |
|     n_u64(06ca6351e003826f), n_u64(142929670a0e6e70),
 | |
|     n_u64(27b70a8546d22ffc), n_u64(2e1b21385c26c926),
 | |
|     n_u64(4d2c6dfc5ac42aed), n_u64(53380d139d95b3df),
 | |
|     n_u64(650a73548baf63de), n_u64(766a0abb3c77b2a8),
 | |
|     n_u64(81c2c92e47edaee6), n_u64(92722c851482353b),
 | |
|     n_u64(a2bfe8a14cf10364), n_u64(a81a664bbc423001),
 | |
|     n_u64(c24b8b70d0f89791), n_u64(c76c51a30654be30),
 | |
|     n_u64(d192e819d6ef5218), n_u64(d69906245565a910),
 | |
|     n_u64(f40e35855771202a), n_u64(106aa07032bbd1b8),
 | |
|     n_u64(19a4c116b8d2d0c8), n_u64(1e376c085141ab53),
 | |
|     n_u64(2748774cdf8eeb99), n_u64(34b0bcb5e19b48a8),
 | |
|     n_u64(391c0cb3c5c95a63), n_u64(4ed8aa4ae3418acb),
 | |
|     n_u64(5b9cca4f7763e373), n_u64(682e6ff3d6b2b8a3),
 | |
|     n_u64(748f82ee5defb2fc), n_u64(78a5636f43172f60),
 | |
|     n_u64(84c87814a1f0ab72), n_u64(8cc702081a6439ec),
 | |
|     n_u64(90befffa23631e28), n_u64(a4506cebde82bde9),
 | |
|     n_u64(bef9a3f7b2c67915), n_u64(c67178f2e372532b),
 | |
|     n_u64(ca273eceea26619c), n_u64(d186b8c721c0c207),
 | |
|     n_u64(eada7dd6cde0eb1e), n_u64(f57d4f7fee6ed178),
 | |
|     n_u64(06f067aa72176fba), n_u64(0a637dc5a2c898a6),
 | |
|     n_u64(113f9804bef90dae), n_u64(1b710b35131c471b),
 | |
|     n_u64(28db77f523047d84), n_u64(32caab7b40c72493),
 | |
|     n_u64(3c9ebe0a15c9bebc), n_u64(431d67c49c100d4c),
 | |
|     n_u64(4cc5d4becb3e42b6), n_u64(597f299cfc657e2a),
 | |
|     n_u64(5fcb6fab3ad6faec), n_u64(6c44198c4a475817)
 | |
| };
 | |
| 
 | |
| /* Compile 64 bytes of hash data into SHA384/SHA512 digest value  */
 | |
| 
 | |
| sha2_void sha512_compile(sha512_ctx ctx[1])
 | |
| {   sha2_64t    v[8];
 | |
|     sha2_32t    j;
 | |
| 
 | |
|     memcpy(v, ctx->hash, 8 * sizeof(sha2_64t));
 | |
| 
 | |
|     for(j = 0; j < 80; j += 16)
 | |
|     {
 | |
|         h5_cycle( 0, j); h5_cycle( 1, j); h5_cycle( 2, j); h5_cycle( 3, j);
 | |
|         h5_cycle( 4, j); h5_cycle( 5, j); h5_cycle( 6, j); h5_cycle( 7, j);
 | |
|         h5_cycle( 8, j); h5_cycle( 9, j); h5_cycle(10, j); h5_cycle(11, j);
 | |
|         h5_cycle(12, j); h5_cycle(13, j); h5_cycle(14, j); h5_cycle(15, j);
 | |
|     }
 | |
| 
 | |
|     ctx->hash[0] += v[0]; ctx->hash[1] += v[1]; ctx->hash[2] += v[2]; ctx->hash[3] += v[3];
 | |
|     ctx->hash[4] += v[4]; ctx->hash[5] += v[5]; ctx->hash[6] += v[6]; ctx->hash[7] += v[7];
 | |
| }
 | |
| 
 | |
| /* Compile 128 bytes of hash data into SHA256 digest value  */
 | |
| /* NOTE: this routine assumes that the byte order in the    */
 | |
| /* ctx->wbuf[] at this point is in such an order that low   */
 | |
| /* address bytes in the ORIGINAL byte stream placed in this */
 | |
| /* buffer will now go to the high end of words on BOTH big  */
 | |
| /* and little endian systems                                */
 | |
| 
 | |
| sha2_void sha512_hash(const unsigned char data[], unsigned long len, sha512_ctx ctx[1])
 | |
| {   sha2_32t pos = (sha2_32t)(ctx->count[0] & SHA512_MASK),
 | |
|              space = SHA512_BLOCK_SIZE - pos;
 | |
|     const unsigned char *sp = data;
 | |
| 
 | |
|     if((ctx->count[0] += len) < len)
 | |
|         ++(ctx->count[1]);
 | |
| 
 | |
|     while(len >= space)     /* tranfer whole blocks while possible  */
 | |
|     {
 | |
|         memcpy(((unsigned char*)ctx->wbuf) + pos, sp, space);
 | |
|         sp += space; len -= space; space = SHA512_BLOCK_SIZE; pos = 0;
 | |
|         bsw_64(ctx->wbuf, SHA512_BLOCK_SIZE >> 3);
 | |
|         sha512_compile(ctx);
 | |
|     }
 | |
| 
 | |
|     memcpy(((unsigned char*)ctx->wbuf) + pos, sp, len);
 | |
| }
 | |
| 
 | |
| /* SHA384/512 Final padding and digest calculation  */
 | |
| 
 | |
| static sha2_64t  m2[8] =
 | |
| {
 | |
|     n_u64(0000000000000000), n_u64(ff00000000000000),
 | |
|     n_u64(ffff000000000000), n_u64(ffffff0000000000),
 | |
|     n_u64(ffffffff00000000), n_u64(ffffffffff000000),
 | |
|     n_u64(ffffffffffff0000), n_u64(ffffffffffffff00)
 | |
| };
 | |
| 
 | |
| static sha2_64t  b2[8] =
 | |
| {
 | |
|     n_u64(8000000000000000), n_u64(0080000000000000),
 | |
|     n_u64(0000800000000000), n_u64(0000008000000000),
 | |
|     n_u64(0000000080000000), n_u64(0000000000800000),
 | |
|     n_u64(0000000000008000), n_u64(0000000000000080)
 | |
| };
 | |
| 
 | |
| static void sha_end(unsigned char hval[], sha512_ctx ctx[1], const unsigned int hlen)
 | |
| {   sha2_32t    i = (sha2_32t)(ctx->count[0] & SHA512_MASK);
 | |
| 
 | |
|     bsw_64(ctx->wbuf, (i + 7) >> 3);
 | |
| 
 | |
|     /* bytes in the buffer are now in an order in which references  */
 | |
|     /* to 64-bit words will put bytes with lower addresses into the */
 | |
|     /* top of 64 bit words on BOTH big and little endian machines   */
 | |
| 
 | |
|     /* we now need to mask valid bytes and add the padding which is */
 | |
|     /* a single 1 bit and as many zero bits as necessary.           */
 | |
|     ctx->wbuf[i >> 3] = (ctx->wbuf[i >> 3] & m2[i & 7]) | b2[i & 7];
 | |
| 
 | |
|     /* we need 17 or more empty byte positions, one for the padding */
 | |
|     /* byte (above) and sixteen for the length count.  If there is  */
 | |
|     /* not enough space pad and empty the buffer                    */
 | |
|     if(i > SHA512_BLOCK_SIZE - 17)
 | |
|     {
 | |
|         if(i < 120) ctx->wbuf[15] = 0;
 | |
|         sha512_compile(ctx);
 | |
|         i = 0;
 | |
|     }
 | |
|     else
 | |
|         i = (i >> 3) + 1;
 | |
| 
 | |
|     while(i < 14)
 | |
|         ctx->wbuf[i++] = 0;
 | |
| 
 | |
|     /* the following 64-bit length fields are assembled in the      */
 | |
|     /* wrong byte order on little endian machines but this is       */
 | |
|     /* corrected later since they are only ever used as 64-bit      */
 | |
|     /* word values.                                                 */
 | |
| 
 | |
|     ctx->wbuf[14] = (ctx->count[1] << 3) | (ctx->count[0] >> 61);
 | |
|     ctx->wbuf[15] = ctx->count[0] << 3;
 | |
| 
 | |
|     sha512_compile(ctx);
 | |
| 
 | |
|     /* extract the hash value as bytes in case the hash buffer is   */
 | |
|     /* misaligned for 32-bit words                                  */
 | |
|     for(i = 0; i < hlen; ++i)
 | |
|         hval[i] = (unsigned char)(ctx->hash[i >> 3] >> (8 * (~i & 7)));
 | |
| }
 | |
| 
 | |
| #endif
 | |
| 
 | |
| #if defined(SHA_2) || defined(SHA_384)
 | |
| 
 | |
| /* SHA384 initialisation data   */
 | |
| 
 | |
| const sha2_64t  i384[80] =
 | |
| {
 | |
|     n_u64(cbbb9d5dc1059ed8), n_u64(629a292a367cd507),
 | |
|     n_u64(9159015a3070dd17), n_u64(152fecd8f70e5939),
 | |
|     n_u64(67332667ffc00b31), n_u64(8eb44a8768581511),
 | |
|     n_u64(db0c2e0d64f98fa7), n_u64(47b5481dbefa4fa4)
 | |
| };
 | |
| 
 | |
| sha2_void sha384_begin(sha384_ctx ctx[1])
 | |
| {
 | |
|     ctx->count[0] = ctx->count[1] = 0;
 | |
|     memcpy(ctx->hash, i384, 8 * sizeof(sha2_64t));
 | |
| }
 | |
| 
 | |
| sha2_void sha384_end(unsigned char hval[], sha384_ctx ctx[1])
 | |
| {
 | |
|     sha_end(hval, ctx, SHA384_DIGEST_SIZE);
 | |
| }
 | |
| 
 | |
| sha2_void sha384(unsigned char hval[], const unsigned char data[], unsigned long len)
 | |
| {   sha384_ctx  cx[1];
 | |
| 
 | |
|     sha384_begin(cx); sha384_hash(data, len, cx); sha384_end(hval, cx);
 | |
| }
 | |
| 
 | |
| #endif
 | |
| 
 | |
| #if defined(SHA_2) || defined(SHA_512)
 | |
| 
 | |
| /* SHA512 initialisation data   */
 | |
| 
 | |
| const sha2_64t  i512[80] =
 | |
| {
 | |
|     n_u64(6a09e667f3bcc908), n_u64(bb67ae8584caa73b),
 | |
|     n_u64(3c6ef372fe94f82b), n_u64(a54ff53a5f1d36f1),
 | |
|     n_u64(510e527fade682d1), n_u64(9b05688c2b3e6c1f),
 | |
|     n_u64(1f83d9abfb41bd6b), n_u64(5be0cd19137e2179)
 | |
| };
 | |
| 
 | |
| sha2_void sha512_begin(sha512_ctx ctx[1])
 | |
| {
 | |
|     ctx->count[0] = ctx->count[1] = 0;
 | |
|     memcpy(ctx->hash, i512, 8 * sizeof(sha2_64t));
 | |
| }
 | |
| 
 | |
| sha2_void sha512_end(unsigned char hval[], sha512_ctx ctx[1])
 | |
| {
 | |
|     sha_end(hval, ctx, SHA512_DIGEST_SIZE);
 | |
| }
 | |
| 
 | |
| sha2_void sha512(unsigned char hval[], const unsigned char data[], unsigned long len)
 | |
| {   sha512_ctx  cx[1];
 | |
| 
 | |
|     sha512_begin(cx); sha512_hash(data, len, cx); sha512_end(hval, cx);
 | |
| }
 | |
| 
 | |
| #endif
 | |
| 
 | |
| #if defined(SHA_2)
 | |
| 
 | |
| #define CTX_256(x)  ((x)->uu->ctx256)
 | |
| #define CTX_384(x)  ((x)->uu->ctx512)
 | |
| #define CTX_512(x)  ((x)->uu->ctx512)
 | |
| 
 | |
| /* SHA2 initialisation */
 | |
| 
 | |
| sha2_int sha2_begin(unsigned long len, sha2_ctx ctx[1])
 | |
| {   unsigned long   l = len;
 | |
|     switch(len)
 | |
|     {
 | |
|         case 256:   l = len >> 3;
 | |
|         /* Falls through. */
 | |
|         case  32:   CTX_256(ctx)->count[0] = CTX_256(ctx)->count[1] = 0;
 | |
|                     memcpy(CTX_256(ctx)->hash, i256, 32); break;
 | |
|         case 384:   l = len >> 3;
 | |
|         /* Falls through. */
 | |
|         case  48:   CTX_384(ctx)->count[0] = CTX_384(ctx)->count[1] = 0;
 | |
|                     memcpy(CTX_384(ctx)->hash, i384, 64); break;
 | |
|         case 512:   l = len >> 3;
 | |
|         /* Falls through. */
 | |
|         case  64:   CTX_512(ctx)->count[0] = CTX_512(ctx)->count[1] = 0;
 | |
|                     memcpy(CTX_512(ctx)->hash, i512, 64); break;
 | |
|         default:    return SHA2_BAD;
 | |
|     }
 | |
| 
 | |
|     ctx->sha2_len = l; return SHA2_GOOD;
 | |
| }
 | |
| 
 | |
| sha2_void sha2_hash(const unsigned char data[], unsigned long len, sha2_ctx ctx[1])
 | |
| {
 | |
|     switch(ctx->sha2_len)
 | |
|     {
 | |
|         case 32: sha256_hash(data, len, CTX_256(ctx)); return;
 | |
|         case 48: sha384_hash(data, len, CTX_384(ctx)); return;
 | |
|         case 64: sha512_hash(data, len, CTX_512(ctx)); return;
 | |
|     }
 | |
| }
 | |
| 
 | |
| sha2_void sha2_end(unsigned char hval[], sha2_ctx ctx[1])
 | |
| {
 | |
|     switch(ctx->sha2_len)
 | |
|     {
 | |
|         case 32: sha256_end(hval, CTX_256(ctx)); return;
 | |
|         case 48: sha_end(hval, CTX_384(ctx), SHA384_DIGEST_SIZE); return;
 | |
|         case 64: sha_end(hval, CTX_512(ctx), SHA512_DIGEST_SIZE); return;
 | |
|     }
 | |
| }
 | |
| 
 | |
| sha2_int sha2(unsigned char hval[], unsigned long size,
 | |
|                                 const unsigned char data[], unsigned long len)
 | |
| {   sha2_ctx    cx[1];
 | |
| 
 | |
|     if(sha2_begin(size, cx) == SHA2_GOOD)
 | |
|     {
 | |
|         sha2_hash(data, len, cx); sha2_end(hval, cx); return SHA2_GOOD;
 | |
|     }
 | |
|     else
 | |
|         return SHA2_BAD;
 | |
| }
 | |
| 
 | |
| #endif
 | |
| 
 |