/* rmd160.c - RIPE-MD160 * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. * * This file is part of GnuPG. * * GnuPG is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * GnuPG is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA */ #include #include #include #include #include #include "util.h" #include "memory.h" #include "rmd.h" #include "cipher.h" /* for rmd160_hash_buffer */ #include "algorithms.h" #include "bithelp.h" /********************************* * RIPEMD-160 is not patented, see (as of 25.10.97) * http://www.esat.kuleuven.ac.be/~bosselae/ripemd160.html * Note that the code uses Little Endian byteorder, which is good for * 386 etc, but we must add some conversion when used on a big endian box. * * * Pseudo-code for RIPEMD-160 * * RIPEMD-160 is an iterative hash function that operates on 32-bit words. * The round function takes as input a 5-word chaining variable and a 16-word * message block and maps this to a new chaining variable. All operations are * defined on 32-bit words. Padding is identical to that of MD4. * * * RIPEMD-160: definitions * * * nonlinear functions at bit level: exor, mux, -, mux, - * * f(j, x, y, z) = x XOR y XOR z (0 <= j <= 15) * f(j, x, y, z) = (x AND y) OR (NOT(x) AND z) (16 <= j <= 31) * f(j, x, y, z) = (x OR NOT(y)) XOR z (32 <= j <= 47) * f(j, x, y, z) = (x AND z) OR (y AND NOT(z)) (48 <= j <= 63) * f(j, x, y, z) = x XOR (y OR NOT(z)) (64 <= j <= 79) * * * added constants (hexadecimal) * * K(j) = 0x00000000 (0 <= j <= 15) * K(j) = 0x5A827999 (16 <= j <= 31) int(2**30 x sqrt(2)) * K(j) = 0x6ED9EBA1 (32 <= j <= 47) int(2**30 x sqrt(3)) * K(j) = 0x8F1BBCDC (48 <= j <= 63) int(2**30 x sqrt(5)) * K(j) = 0xA953FD4E (64 <= j <= 79) int(2**30 x sqrt(7)) * K'(j) = 0x50A28BE6 (0 <= j <= 15) int(2**30 x cbrt(2)) * K'(j) = 0x5C4DD124 (16 <= j <= 31) int(2**30 x cbrt(3)) * K'(j) = 0x6D703EF3 (32 <= j <= 47) int(2**30 x cbrt(5)) * K'(j) = 0x7A6D76E9 (48 <= j <= 63) int(2**30 x cbrt(7)) * K'(j) = 0x00000000 (64 <= j <= 79) * * * selection of message word * * r(j) = j (0 <= j <= 15) * r(16..31) = 7, 4, 13, 1, 10, 6, 15, 3, 12, 0, 9, 5, 2, 14, 11, 8 * r(32..47) = 3, 10, 14, 4, 9, 15, 8, 1, 2, 7, 0, 6, 13, 11, 5, 12 * r(48..63) = 1, 9, 11, 10, 0, 8, 12, 4, 13, 3, 7, 15, 14, 5, 6, 2 * r(64..79) = 4, 0, 5, 9, 7, 12, 2, 10, 14, 1, 3, 8, 11, 6, 15, 13 * r0(0..15) = 5, 14, 7, 0, 9, 2, 11, 4, 13, 6, 15, 8, 1, 10, 3, 12 * r0(16..31)= 6, 11, 3, 7, 0, 13, 5, 10, 14, 15, 8, 12, 4, 9, 1, 2 * r0(32..47)= 15, 5, 1, 3, 7, 14, 6, 9, 11, 8, 12, 2, 10, 0, 4, 13 * r0(48..63)= 8, 6, 4, 1, 3, 11, 15, 0, 5, 12, 2, 13, 9, 7, 10, 14 * r0(64..79)= 12, 15, 10, 4, 1, 5, 8, 7, 6, 2, 13, 14, 0, 3, 9, 11 * * * amount for rotate left (rol) * * s(0..15) = 11, 14, 15, 12, 5, 8, 7, 9, 11, 13, 14, 15, 6, 7, 9, 8 * s(16..31) = 7, 6, 8, 13, 11, 9, 7, 15, 7, 12, 15, 9, 11, 7, 13, 12 * s(32..47) = 11, 13, 6, 7, 14, 9, 13, 15, 14, 8, 13, 6, 5, 12, 7, 5 * s(48..63) = 11, 12, 14, 15, 14, 15, 9, 8, 9, 14, 5, 6, 8, 6, 5, 12 * s(64..79) = 9, 15, 5, 11, 6, 8, 13, 12, 5, 12, 13, 14, 11, 8, 5, 6 * s'(0..15) = 8, 9, 9, 11, 13, 15, 15, 5, 7, 7, 8, 11, 14, 14, 12, 6 * s'(16..31)= 9, 13, 15, 7, 12, 8, 9, 11, 7, 7, 12, 7, 6, 15, 13, 11 * s'(32..47)= 9, 7, 15, 11, 8, 6, 6, 14, 12, 13, 5, 14, 13, 13, 7, 5 * s'(48..63)= 15, 5, 8, 11, 14, 14, 6, 14, 6, 9, 12, 9, 12, 5, 15, 8 * s'(64..79)= 8, 5, 12, 9, 12, 5, 14, 6, 8, 13, 6, 5, 15, 13, 11, 11 * * * initial value (hexadecimal) * * h0 = 0x67452301; h1 = 0xEFCDAB89; h2 = 0x98BADCFE; h3 = 0x10325476; * h4 = 0xC3D2E1F0; * * * RIPEMD-160: pseudo-code * * It is assumed that the message after padding consists of t 16-word blocks * that will be denoted with X[i][j], with 0 <= i <= t-1 and 0 <= j <= 15. * The symbol [+] denotes addition modulo 2**32 and rol_s denotes cyclic left * shift (rotate) over s positions. * * * for i := 0 to t-1 { * A := h0; B := h1; C := h2; D = h3; E = h4; * A' := h0; B' := h1; C' := h2; D' = h3; E' = h4; * for j := 0 to 79 { * T := rol_s(j)(A [+] f(j, B, C, D) [+] X[i][r(j)] [+] K(j)) [+] E; * A := E; E := D; D := rol_10(C); C := B; B := T; * T := rol_s'(j)(A' [+] f(79-j, B', C', D') [+] X[i][r'(j)] [+] K'(j)) [+] E'; * A' := E'; E' := D'; D' := rol_10(C'); C' := B'; B' := T; * } * T := h1 [+] C [+] D'; h1 := h2 [+] D [+] E'; h2 := h3 [+] E [+] A'; * h3 := h4 [+] A [+] B'; h4 := h0 [+] B [+] C'; h0 := T; * } */ /* Some examples: * "" 9c1185a5c5e9fc54612808977ee8f548b2258d31 * "a" 0bdc9d2d256b3ee9daae347be6f4dc835a467ffe * "abc" 8eb208f7e05d987a9b044a8e98c6b087f15a0bfc * "message digest" 5d0689ef49d2fae572b881b123a85ffa21595f36 * "a...z" f71c27109c692c1b56bbdceb5b9d2865b3708dbc * "abcdbcde...nopq" 12a053384a9c0c88e405a06c27dcf49ada62eb2b * "A...Za...z0...9" b0e20b6e3116640286ed3a87a5713079b21f5189 * 8 times "1234567890" 9b752e45573d4b39f4dbd3323cab82bf63326bfb * 1 million times "a" 52783243c1697bdbe16d37f97f68f08325dc1528 */ static void burn_stack (int bytes) { char buf[150]; wipememory(buf,sizeof buf); bytes -= sizeof buf; if (bytes > 0) burn_stack (bytes); } void rmd160_init( RMD160_CONTEXT *hd ) { hd->h0 = 0x67452301; hd->h1 = 0xEFCDAB89; hd->h2 = 0x98BADCFE; hd->h3 = 0x10325476; hd->h4 = 0xC3D2E1F0; hd->nblocks = 0; hd->count = 0; } /**************** * Transform the message X which consists of 16 32-bit-words */ static void transform( RMD160_CONTEXT *hd, byte *data ) { u32 a,b,c,d,e,aa,bb,cc,dd,ee,t; #ifdef BIG_ENDIAN_HOST u32 x[16]; { int i; byte *p2, *p1; for(i=0, p1=data, p2=(byte*)x; i < 16; i++, p2 += 4 ) { p2[3] = *p1++; p2[2] = *p1++; p2[1] = *p1++; p2[0] = *p1++; } } #else #if 0 u32 *x =(u32*)data; #else /* this version is better because it is always aligned; * The performance penalty on a 586-100 is about 6% which * is acceptable - because the data is more local it might * also be possible that this is faster on some machines. * This function (when compiled with -02 on gcc 2.7.2) * executes on a 586-100 (39.73 bogomips) at about 1900kb/sec; * [measured with a 4MB data and "gpgm --print-md rmd160"] */ u32 x[16]; memcpy( x, data, 64 ); #endif #endif #define K0 0x00000000 #define K1 0x5A827999 #define K2 0x6ED9EBA1 #define K3 0x8F1BBCDC #define K4 0xA953FD4E #define KK0 0x50A28BE6 #define KK1 0x5C4DD124 #define KK2 0x6D703EF3 #define KK3 0x7A6D76E9 #define KK4 0x00000000 #define F0(x,y,z) ( (x) ^ (y) ^ (z) ) #define F1(x,y,z) ( ((x) & (y)) | (~(x) & (z)) ) #define F2(x,y,z) ( ((x) | ~(y)) ^ (z) ) #define F3(x,y,z) ( ((x) & (z)) | ((y) & ~(z)) ) #define F4(x,y,z) ( (x) ^ ((y) | ~(z)) ) #define R(a,b,c,d,e,f,k,r,s) do { t = a + f(b,c,d) + k + x[r]; \ a = rol(t,s) + e; \ c = rol(c,10); \ } while(0) /* left lane */ a = hd->h0; b = hd->h1; c = hd->h2; d = hd->h3; e = hd->h4; R( a, b, c, d, e, F0, K0, 0, 11 ); R( e, a, b, c, d, F0, K0, 1, 14 ); R( d, e, a, b, c, F0, K0, 2, 15 ); R( c, d, e, a, b, F0, K0, 3, 12 ); R( b, c, d, e, a, F0, K0, 4, 5 ); R( a, b, c, d, e, F0, K0, 5, 8 ); R( e, a, b, c, d, F0, K0, 6, 7 ); R( d, e, a, b, c, F0, K0, 7, 9 ); R( c, d, e, a, b, F0, K0, 8, 11 ); R( b, c, d, e, a, F0, K0, 9, 13 ); R( a, b, c, d, e, F0, K0, 10, 14 ); R( e, a, b, c, d, F0, K0, 11, 15 ); R( d, e, a, b, c, F0, K0, 12, 6 ); R( c, d, e, a, b, F0, K0, 13, 7 ); R( b, c, d, e, a, F0, K0, 14, 9 ); R( a, b, c, d, e, F0, K0, 15, 8 ); R( e, a, b, c, d, F1, K1, 7, 7 ); R( d, e, a, b, c, F1, K1, 4, 6 ); R( c, d, e, a, b, F1, K1, 13, 8 ); R( b, c, d, e, a, F1, K1, 1, 13 ); R( a, b, c, d, e, F1, K1, 10, 11 ); R( e, a, b, c, d, F1, K1, 6, 9 ); R( d, e, a, b, c, F1, K1, 15, 7 ); R( c, d, e, a, b, F1, K1, 3, 15 ); R( b, c, d, e, a, F1, K1, 12, 7 ); R( a, b, c, d, e, F1, K1, 0, 12 ); R( e, a, b, c, d, F1, K1, 9, 15 ); R( d, e, a, b, c, F1, K1, 5, 9 ); R( c, d, e, a, b, F1, K1, 2, 11 ); R( b, c, d, e, a, F1, K1, 14, 7 ); R( a, b, c, d, e, F1, K1, 11, 13 ); R( e, a, b, c, d, F1, K1, 8, 12 ); R( d, e, a, b, c, F2, K2, 3, 11 ); R( c, d, e, a, b, F2, K2, 10, 13 ); R( b, c, d, e, a, F2, K2, 14, 6 ); R( a, b, c, d, e, F2, K2, 4, 7 ); R( e, a, b, c, d, F2, K2, 9, 14 ); R( d, e, a, b, c, F2, K2, 15, 9 ); R( c, d, e, a, b, F2, K2, 8, 13 ); R( b, c, d, e, a, F2, K2, 1, 15 ); R( a, b, c, d, e, F2, K2, 2, 14 ); R( e, a, b, c, d, F2, K2, 7, 8 ); R( d, e, a, b, c, F2, K2, 0, 13 ); R( c, d, e, a, b, F2, K2, 6, 6 ); R( b, c, d, e, a, F2, K2, 13, 5 ); R( a, b, c, d, e, F2, K2, 11, 12 ); R( e, a, b, c, d, F2, K2, 5, 7 ); R( d, e, a, b, c, F2, K2, 12, 5 ); R( c, d, e, a, b, F3, K3, 1, 11 ); R( b, c, d, e, a, F3, K3, 9, 12 ); R( a, b, c, d, e, F3, K3, 11, 14 ); R( e, a, b, c, d, F3, K3, 10, 15 ); R( d, e, a, b, c, F3, K3, 0, 14 ); R( c, d, e, a, b, F3, K3, 8, 15 ); R( b, c, d, e, a, F3, K3, 12, 9 ); R( a, b, c, d, e, F3, K3, 4, 8 ); R( e, a, b, c, d, F3, K3, 13, 9 ); R( d, e, a, b, c, F3, K3, 3, 14 ); R( c, d, e, a, b, F3, K3, 7, 5 ); R( b, c, d, e, a, F3, K3, 15, 6 ); R( a, b, c, d, e, F3, K3, 14, 8 ); R( e, a, b, c, d, F3, K3, 5, 6 ); R( d, e, a, b, c, F3, K3, 6, 5 ); R( c, d, e, a, b, F3, K3, 2, 12 ); R( b, c, d, e, a, F4, K4, 4, 9 ); R( a, b, c, d, e, F4, K4, 0, 15 ); R( e, a, b, c, d, F4, K4, 5, 5 ); R( d, e, a, b, c, F4, K4, 9, 11 ); R( c, d, e, a, b, F4, K4, 7, 6 ); R( b, c, d, e, a, F4, K4, 12, 8 ); R( a, b, c, d, e, F4, K4, 2, 13 ); R( e, a, b, c, d, F4, K4, 10, 12 ); R( d, e, a, b, c, F4, K4, 14, 5 ); R( c, d, e, a, b, F4, K4, 1, 12 ); R( b, c, d, e, a, F4, K4, 3, 13 ); R( a, b, c, d, e, F4, K4, 8, 14 ); R( e, a, b, c, d, F4, K4, 11, 11 ); R( d, e, a, b, c, F4, K4, 6, 8 ); R( c, d, e, a, b, F4, K4, 15, 5 ); R( b, c, d, e, a, F4, K4, 13, 6 ); aa = a; bb = b; cc = c; dd = d; ee = e; /* right lane */ a = hd->h0; b = hd->h1; c = hd->h2; d = hd->h3; e = hd->h4; R( a, b, c, d, e, F4, KK0, 5, 8); R( e, a, b, c, d, F4, KK0, 14, 9); R( d, e, a, b, c, F4, KK0, 7, 9); R( c, d, e, a, b, F4, KK0, 0, 11); R( b, c, d, e, a, F4, KK0, 9, 13); R( a, b, c, d, e, F4, KK0, 2, 15); R( e, a, b, c, d, F4, KK0, 11, 15); R( d, e, a, b, c, F4, KK0, 4, 5); R( c, d, e, a, b, F4, KK0, 13, 7); R( b, c, d, e, a, F4, KK0, 6, 7); R( a, b, c, d, e, F4, KK0, 15, 8); R( e, a, b, c, d, F4, KK0, 8, 11); R( d, e, a, b, c, F4, KK0, 1, 14); R( c, d, e, a, b, F4, KK0, 10, 14); R( b, c, d, e, a, F4, KK0, 3, 12); R( a, b, c, d, e, F4, KK0, 12, 6); R( e, a, b, c, d, F3, KK1, 6, 9); R( d, e, a, b, c, F3, KK1, 11, 13); R( c, d, e, a, b, F3, KK1, 3, 15); R( b, c, d, e, a, F3, KK1, 7, 7); R( a, b, c, d, e, F3, KK1, 0, 12); R( e, a, b, c, d, F3, KK1, 13, 8); R( d, e, a, b, c, F3, KK1, 5, 9); R( c, d, e, a, b, F3, KK1, 10, 11); R( b, c, d, e, a, F3, KK1, 14, 7); R( a, b, c, d, e, F3, KK1, 15, 7); R( e, a, b, c, d, F3, KK1, 8, 12); R( d, e, a, b, c, F3, KK1, 12, 7); R( c, d, e, a, b, F3, KK1, 4, 6); R( b, c, d, e, a, F3, KK1, 9, 15); R( a, b, c, d, e, F3, KK1, 1, 13); R( e, a, b, c, d, F3, KK1, 2, 11); R( d, e, a, b, c, F2, KK2, 15, 9); R( c, d, e, a, b, F2, KK2, 5, 7); R( b, c, d, e, a, F2, KK2, 1, 15); R( a, b, c, d, e, F2, KK2, 3, 11); R( e, a, b, c, d, F2, KK2, 7, 8); R( d, e, a, b, c, F2, KK2, 14, 6); R( c, d, e, a, b, F2, KK2, 6, 6); R( b, c, d, e, a, F2, KK2, 9, 14); R( a, b, c, d, e, F2, KK2, 11, 12); R( e, a, b, c, d, F2, KK2, 8, 13); R( d, e, a, b, c, F2, KK2, 12, 5); R( c, d, e, a, b, F2, KK2, 2, 14); R( b, c, d, e, a, F2, KK2, 10, 13); R( a, b, c, d, e, F2, KK2, 0, 13); R( e, a, b, c, d, F2, KK2, 4, 7); R( d, e, a, b, c, F2, KK2, 13, 5); R( c, d, e, a, b, F1, KK3, 8, 15); R( b, c, d, e, a, F1, KK3, 6, 5); R( a, b, c, d, e, F1, KK3, 4, 8); R( e, a, b, c, d, F1, KK3, 1, 11); R( d, e, a, b, c, F1, KK3, 3, 14); R( c, d, e, a, b, F1, KK3, 11, 14); R( b, c, d, e, a, F1, KK3, 15, 6); R( a, b, c, d, e, F1, KK3, 0, 14); R( e, a, b, c, d, F1, KK3, 5, 6); R( d, e, a, b, c, F1, KK3, 12, 9); R( c, d, e, a, b, F1, KK3, 2, 12); R( b, c, d, e, a, F1, KK3, 13, 9); R( a, b, c, d, e, F1, KK3, 9, 12); R( e, a, b, c, d, F1, KK3, 7, 5); R( d, e, a, b, c, F1, KK3, 10, 15); R( c, d, e, a, b, F1, KK3, 14, 8); R( b, c, d, e, a, F0, KK4, 12, 8); R( a, b, c, d, e, F0, KK4, 15, 5); R( e, a, b, c, d, F0, KK4, 10, 12); R( d, e, a, b, c, F0, KK4, 4, 9); R( c, d, e, a, b, F0, KK4, 1, 12); R( b, c, d, e, a, F0, KK4, 5, 5); R( a, b, c, d, e, F0, KK4, 8, 14); R( e, a, b, c, d, F0, KK4, 7, 6); R( d, e, a, b, c, F0, KK4, 6, 8); R( c, d, e, a, b, F0, KK4, 2, 13); R( b, c, d, e, a, F0, KK4, 13, 6); R( a, b, c, d, e, F0, KK4, 14, 5); R( e, a, b, c, d, F0, KK4, 0, 15); R( d, e, a, b, c, F0, KK4, 3, 13); R( c, d, e, a, b, F0, KK4, 9, 11); R( b, c, d, e, a, F0, KK4, 11, 11); t = hd->h1 + d + cc; hd->h1 = hd->h2 + e + dd; hd->h2 = hd->h3 + a + ee; hd->h3 = hd->h4 + b + aa; hd->h4 = hd->h0 + c + bb; hd->h0 = t; } /* Update the message digest with the contents * of INBUF with length INLEN. */ static void rmd160_write( RMD160_CONTEXT *hd, byte *inbuf, size_t inlen) { if( hd->count == 64 ) { /* flush the buffer */ transform( hd, hd->buf ); burn_stack (108+5*sizeof(void*)); hd->count = 0; hd->nblocks++; } if( !inbuf ) return; if( hd->count ) { for( ; inlen && hd->count < 64; inlen-- ) hd->buf[hd->count++] = *inbuf++; rmd160_write( hd, NULL, 0 ); if( !inlen ) return; } while( inlen >= 64 ) { transform( hd, inbuf ); hd->count = 0; hd->nblocks++; inlen -= 64; inbuf += 64; } burn_stack (108+5*sizeof(void*)); for( ; inlen && hd->count < 64; inlen-- ) hd->buf[hd->count++] = *inbuf++; } /**************** * Apply the rmd160 transform function on the buffer which must have * a length 64 bytes. Do not use this function together with the * other functions, use rmd160_init to initialize internal variables. * Returns: 16 bytes in buffer with the mixed contentes of buffer. */ void rmd160_mixblock( RMD160_CONTEXT *hd, char *buffer ) { char *p = buffer; transform( hd, buffer ); #define X(a) do { *(u32*)p = hd->h##a ; p += 4; } while(0) X(0); X(1); X(2); X(3); X(4); #undef X } /* The routine terminates the computation */ static void rmd160_final( RMD160_CONTEXT *hd ) { u32 t, msb, lsb; byte *p; rmd160_write(hd, NULL, 0); /* flush */; t = hd->nblocks; /* multiply by 64 to make a byte count */ lsb = t << 6; msb = t >> 26; /* add the count */ t = lsb; if( (lsb += hd->count) < t ) msb++; /* multiply by 8 to make a bit count */ t = lsb; lsb <<= 3; msb <<= 3; msb |= t >> 29; if( hd->count < 56 ) { /* enough room */ hd->buf[hd->count++] = 0x80; /* pad */ while( hd->count < 56 ) hd->buf[hd->count++] = 0; /* pad */ } else { /* need one extra block */ hd->buf[hd->count++] = 0x80; /* pad character */ while( hd->count < 64 ) hd->buf[hd->count++] = 0; rmd160_write(hd, NULL, 0); /* flush */; memset(hd->buf, 0, 56 ); /* fill next block with zeroes */ } /* append the 64 bit count */ hd->buf[56] = lsb ; hd->buf[57] = lsb >> 8; hd->buf[58] = lsb >> 16; hd->buf[59] = lsb >> 24; hd->buf[60] = msb ; hd->buf[61] = msb >> 8; hd->buf[62] = msb >> 16; hd->buf[63] = msb >> 24; transform( hd, hd->buf ); burn_stack (108+5*sizeof(void*)); p = hd->buf; #ifdef BIG_ENDIAN_HOST #define X(a) do { *p++ = hd->h##a ; *p++ = hd->h##a >> 8; \ *p++ = hd->h##a >> 16; *p++ = hd->h##a >> 24; } while(0) #else /* little endian */ #define X(a) do { *(u32*)p = hd->h##a ; p += 4; } while(0) #endif X(0); X(1); X(2); X(3); X(4); #undef X } static byte * rmd160_read( RMD160_CONTEXT *hd ) { return hd->buf; } /**************** * Shortcut functions which puts the hash value of the supplied buffer * into outbuf which must have a size of 20 bytes. */ void rmd160_hash_buffer( char *outbuf, const char *buffer, size_t length ) { RMD160_CONTEXT hd; rmd160_init( &hd ); rmd160_write( &hd, (byte*)buffer, length ); rmd160_final( &hd ); memcpy( outbuf, hd.buf, 20 ); } /**************** * Return some information about the algorithm. We need algo here to * distinguish different flavors of the algorithm. * Returns: A pointer to string describing the algorithm or NULL if * the ALGO is invalid. */ const char * rmd160_get_info( int algo, size_t *contextsize, byte **r_asnoid, int *r_asnlen, int *r_mdlen, void (**r_init)( void *c ), void (**r_write)( void *c, byte *buf, size_t nbytes ), void (**r_final)( void *c ), byte *(**r_read)( void *c ) ) { static byte asn[15] = /* Object ID is 1.3.36.3.2.1 */ { 0x30, 0x21, 0x30, 0x09, 0x06, 0x05, 0x2b, 0x24, 0x03, 0x02, 0x01, 0x05, 0x00, 0x04, 0x14 }; if( algo != 3 ) return NULL; *contextsize = sizeof(RMD160_CONTEXT); *r_asnoid = asn; *r_asnlen = DIM(asn); *r_mdlen = 20; *(void (**)(RMD160_CONTEXT *))r_init = rmd160_init; *(void (**)(RMD160_CONTEXT *, byte*, size_t))r_write = rmd160_write; *(void (**)(RMD160_CONTEXT *))r_final = rmd160_final; *(byte *(**)(RMD160_CONTEXT *))r_read = rmd160_read; return "RIPEMD160"; }