一、实验目的 通过实际编程了解MD5算法的加密和解密过程,加深对Hash算法的认识。
二、实验原理 Hash函数是将任意长的数字串转换成一个较短的定长输出数字串的函数,输出的结果称为Hash值。Hash函数具有如下特点:
(1)快速性:对于任意一个输入值x,由Hash函数H(x),计算Hash值y,即y=H(x),是非常容易的。
(2)单向性:对于任意一个输出值y,希望反向推出输入值x,使得y=H(x),是非常困难的。
(3)无碰撞性:对任意给定的数据块x,希望找到一个y,满足H(x)=H(y),且x≠y,具有计算的不可行性。
Hash函数可用于数字签名、消息的完整性检测、消息的起源认证检测等。现在常用的Hash算法有MD5、SHA-1等。下面从MD5入手来介绍Hash算法的实现机制。
MD系列单向散列函数是由Ron Rivest设计的,MD5算法对任意长度的输入值处理后产生128位的输出值。MD5算法的实现步骤如下:
在MD5算法中,首先需要对信息进行填充,使其字节长度与448模512同余,即信息的字节长度扩展至nx512+448,11为一个正整数。填充的方法如下:在信息的后面填充第一位为1,其余各位均为0,直到满足上面的条件时才停止用0对信息填充。然后,再在这个结果后面附加一个以64位二进制表示的填充前信息长度。经过这两步的处理,现在的信息字节长度为n*512+448+64=(n+1)*512,即长度恰好是512的整数倍,这样做的目的是为满足后面处理中对信息长度的要求。n个分组中第q个分组表示为Yq。MD5中有A、B、C、D,4个32位被称作链接变量的整数参数,它们的初始值分别为:
A=01234567,B=89abcdef,C=fedcba98,D=76543210
当设置好这4个链接变量后,就开始进入算法的4轮循环运算。循环的次数是信息中512位信息分组数目。首先将上面4个链接变量复制到另外4个变量中:A到AA,B到BB,C到CC,D到DD,以备后面进行处理。
然后进入主循环,主循环有4轮,每轮循环都很相似。第1轮进行16次操作,每次操作对A、B、C和D中的其中3个作一次非线性函数运算,然后将所得结果加上第4个变量,文本的一个子分组和一个常数。再将所得结果向左循环移S位,并加上A、B、C或D其中之一。最后用该结果取代A、B、C或D其中之一。
以下是每次操作中用到的4个非线性函数(每轮一个)。
M[j]表示在第q个512位数据块中的第j个32位子分组,0≤j≤15。
常数T[i]可以有如下选择,在第i步中,T[i]是4294967296*abs(sin(i))的整数部分(注:4294967296=232。),i的单位是弧度。其中,T[i]是32位的随机数源,它消除了输入数据中任何规律性的特征。
表4-1说明了四轮主循环中每轮16步操作的具体步骤。
所有这些完成之后,将A、B、C、D分别加上AA、BB、CC、DD。然后用下一分组数据继续运行算法,最后的输出是A、B、C和D的级联。
算法分析
根据附录中有关MD5算法的头文件md5.h和实现文件md5.c,根据所提供的文件分析MD5算法的实现过程。
下面简单介绍所用到的结构体变量和函数(见表1-4)。程序中用到的结构体变量如下:
typedef struct md5_state{ ulong64 length; ulong32 state[4],curlen; unsigned char buf[64]; }md5_state;
length记录已经处理过的位数,curlen记录已经处理过的字节数,数组state存储上面所说的4个链接变量,buf作为处理过程中的缓存。
程序中用到的函数如下:
(1)void md5_init(md5_state * md)
函数名称:初始化函数。
参数说明:
md指向一个上面所提到的结构体变量。初始化时把curlen和length置为0,并把4个链接变量储存到state中。
(2)int md5_process(md5_state * md,const unsigned char * buf ,unsigned long len)
函数名称:处理函数。
参数说明:
md指向经过初始化函数处理过的一个结构体变量。
buf指向待处理的信息。
len是buf中信息的长度,以字节为单位。
这个函数对待处理的信息以512位为单位进行压缩,不足的部分存储在结构体的buf中,并且用len来指示信息的末尾,这样下次调用时会接着上一次的结果进行。
(3)int md5_done(md5_state * md,unsigned char * Hash)
函数名称:完成函数。
参数说明:md指向上面所处理过的结构体。
Hash指向存储结果的缓冲区。 这个函数对未完成的信息先进行padding操作,然后处理,并把最终结果存在Hash指向的缓冲区中。
(4)int md5_test(void)
函数名称:测试函数。 这个函数对上面的3个函数进行测试。函数内部定义了一组信息和Hash结果一一对应的数组。通过调用上面的3个函数,并把结果和正确结果相比较,可以判断程序正确与否。
实验代码:
//MD5文件头 #ifndef CRYPT_H_ #define CRYPT_H_ #include <assert.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <time.h> #include <ctype.h> #include <limits.h> /* Use small code where possible */ #define SMALL_CODE /* Enable self-test test vector checking */ #define LTC_TEST #define MD5 #ifdef __cplusplus extern "C" { #endif /* version */ #define CRYPT 0x0096 #define SCRYPT "0.96" /* error codes [will be expanded in future releases] */ enum { CRYPT_OK=0, /* Result OK */ CRYPT_FAIL_TESTVECTOR, /* Algorithm failed test vectors */ CRYPT_INVALID_ARG, /* Generic invalid argument */ }; /* type of argument checking, 0=default, 1=fatal and 2=none */ #define ARGTYPE 0 #ifdef _MSC_VER #define CONST64(n) n ## ui64 typedef unsigned __int64 ulong64; #else #define CONST64(n) n ## ULL typedef unsigned long long ulong64; #endif typedef unsigned long ulong32; /* Controls endianess and size of registers. Leave uncommented to get platform neutral [slower] code */ /* detect x86-32 machines somewhat */ #if defined(INTEL_CC) || (defined(_MSC_VER) && defined(WIN32)) || (defined(__GNUC__) && (defined(__DJGPP__) || defined(__CYGWIN__) || defined(__MINGW32__) || defined(__i386__))) #define ENDIAN_LITTLE #define ENDIAN_32BITWORD #endif /* detects MIPS R5900 processors (PS2) */ #if (defined(__R5900) || defined(R5900) || defined(__R5900__)) && (defined(_mips) || defined(__mips__) || defined(mips)) #define ENDIAN_LITTLE #define ENDIAN_64BITWORD #endif /* #define ENDIAN_LITTLE */ /* #define ENDIAN_BIG */ /* #define ENDIAN_32BITWORD */ /* #define ENDIAN_64BITWORD */ #if (defined(ENDIAN_BIG) || defined(ENDIAN_LITTLE)) && !(defined(ENDIAN_32BITWORD) || defined(ENDIAN_64BITWORD)) #error You must specify a word size as well as endianess in mycrypt_cfg.h #endif #if !(defined(ENDIAN_BIG) || defined(ENDIAN_LITTLE)) #define ENDIAN_NEUTRAL #endif #ifdef ENDIAN_NEUTRAL #define STORE32L(x, y) \ { (y)[3] = (unsigned char)(((x)>>24)&255); (y)[2] = (unsigned char)(((x)>>16)&255); \ (y)[1] = (unsigned char)(((x)>>8)&255); (y)[0] = (unsigned char)((x)&255); } #define LOAD32L(x, y) \ { x = ((unsigned long)((y)[3] & 255)<<24) | \ ((unsigned long)((y)[2] & 255)<<16) | \ ((unsigned long)((y)[1] & 255)<<8) | \ ((unsigned long)((y)[0] & 255)); } #define STORE64L(x, y) \ { (y)[7] = (unsigned char)(((x)>>56)&255); (y)[6] = (unsigned char)(((x)>>48)&255); \ (y)[5] = (unsigned char)(((x)>>40)&255); (y)[4] = (unsigned char)(((x)>>32)&255); \ (y)[3] = (unsigned char)(((x)>>24)&255); (y)[2] = (unsigned char)(((x)>>16)&255); \ (y)[1] = (unsigned char)(((x)>>8)&255); (y)[0] = (unsigned char)((x)&255); } #define LOAD64L(x, y) \ { x = (((ulong64)((y)[7] & 255))<<56)|(((ulong64)((y)[6] & 255))<<48)| \ (((ulong64)((y)[5] & 255))<<40)|(((ulong64)((y)[4] & 255))<<32)| \ (((ulong64)((y)[3] & 255))<<24)|(((ulong64)((y)[2] & 255))<<16)| \ (((ulong64)((y)[1] & 255))<<8)|(((ulong64)((y)[0] & 255))); } #define STORE32H(x, y) \ { (y)[0] = (unsigned char)(((x)>>24)&255); (y)[1] = (unsigned char)(((x)>>16)&255); \ (y)[2] = (unsigned char)(((x)>>8)&255); (y)[3] = (unsigned char)((x)&255); } #define LOAD32H(x, y) \ { x = ((unsigned long)((y)[0] & 255)<<24) | \ ((unsigned long)((y)[1] & 255)<<16) | \ ((unsigned long)((y)[2] & 255)<<8) | \ ((unsigned long)((y)[3] & 255)); } #define STORE64H(x, y) \ { (y)[0] = (unsigned char)(((x)>>56)&255); (y)[1] = (unsigned char)(((x)>>48)&255); \ (y)[2] = (unsigned char)(((x)>>40)&255); (y)[3] = (unsigned char)(((x)>>32)&255); \ (y)[4] = (unsigned char)(((x)>>24)&255); (y)[5] = (unsigned char)(((x)>>16)&255); \ (y)[6] = (unsigned char)(((x)>>8)&255); (y)[7] = (unsigned char)((x)&255); } #define LOAD64H(x, y) \ { x = (((ulong64)((y)[0] & 255))<<56)|(((ulong64)((y)[1] & 255))<<48) | \ (((ulong64)((y)[2] & 255))<<40)|(((ulong64)((y)[3] & 255))<<32) | \ (((ulong64)((y)[4] & 255))<<24)|(((ulong64)((y)[5] & 255))<<16) | \ (((ulong64)((y)[6] & 255))<<8)|(((ulong64)((y)[7] & 255))); } #endif /* ENDIAN_NEUTRAL */ #ifdef ENDIAN_LITTLE #define STORE32H(x, y) \ { (y)[0] = (unsigned char)(((x)>>24)&255); (y)[1] = (unsigned char)(((x)>>16)&255); \ (y)[2] = (unsigned char)(((x)>>8)&255); (y)[3] = (unsigned char)((x)&255); } #define LOAD32H(x, y) \ { x = ((unsigned long)((y)[0] & 255)<<24) | \ ((unsigned long)((y)[1] & 255)<<16) | \ ((unsigned long)((y)[2] & 255)<<8) | \ ((unsigned long)((y)[3] & 255)); } #define STORE64H(x, y) \ { (y)[0] = (unsigned char)(((x)>>56)&255); (y)[1] = (unsigned char)(((x)>>48)&255); \ (y)[2] = (unsigned char)(((x)>>40)&255); (y)[3] = (unsigned char)(((x)>>32)&255); \ (y)[4] = (unsigned char)(((x)>>24)&255); (y)[5] = (unsigned char)(((x)>>16)&255); \ (y)[6] = (unsigned char)(((x)>>8)&255); (y)[7] = (unsigned char)((x)&255); } #define LOAD64H(x, y) \ { x = (((ulong64)((y)[0] & 255))<<56)|(((ulong64)((y)[1] & 255))<<48) | \ (((ulong64)((y)[2] & 255))<<40)|(((ulong64)((y)[3] & 255))<<32) | \ (((ulong64)((y)[4] & 255))<<24)|(((ulong64)((y)[5] & 255))<<16) | \ (((ulong64)((y)[6] & 255))<<8)|(((ulong64)((y)[7] & 255))); } #ifdef ENDIAN_32BITWORD #define STORE32L(x, y) \ { unsigned long __t = (x); memcpy(y, &__t, 4); } #define LOAD32L(x, y) \ memcpy(&(x), y, 4); #define STORE64L(x, y) \ { (y)[7] = (unsigned char)(((x)>>56)&255); (y)[6] = (unsigned char)(((x)>>48)&255); \ (y)[5] = (unsigned char)(((x)>>40)&255); (y)[4] = (unsigned char)(((x)>>32)&255); \ (y)[3] = (unsigned char)(((x)>>24)&255); (y)[2] = (unsigned char)(((x)>>16)&255); \ (y)[1] = (unsigned char)(((x)>>8)&255); (y)[0] = (unsigned char)((x)&255); } #define LOAD64L(x, y) \ { x = (((ulong64)((y)[7] & 255))<<56)|(((ulong64)((y)[6] & 255))<<48)| \ (((ulong64)((y)[5] & 255))<<40)|(((ulong64)((y)[4] & 255))<<32)| \ (((ulong64)((y)[3] & 255))<<24)|(((ulong64)((y)[2] & 255))<<16)| \ (((ulong64)((y)[1] & 255))<<8)|(((ulong64)((y)[0] & 255))); } #else /* 64-bit words then */ #define STORE32L(x, y) \ { unsigned long __t = (x); memcpy(y, &__t, 4); } #define LOAD32L(x, y) \ { memcpy(&(x), y, 4); x &= 0xFFFFFFFF; } #define STORE64L(x, y) \ { ulong64 __t = (x); memcpy(y, &__t, 8); } #define LOAD64L(x, y) \ { memcpy(&(x), y, 8); } #endif /* ENDIAN_64BITWORD */ #endif /* ENDIAN_LITTLE */ #ifdef ENDIAN_BIG #define STORE32L(x, y) \ { (y)[0] = (unsigned char)(((x)>>24)&255); (y)[1] = (unsigned char)(((x)>>16)&255); \ (y)[2] = (unsigned char)(((x)>>8)&255); (y)[3] = (unsigned char)((x)&255); } #define LOAD32L(x, y) \ { x = ((unsigned long)((y)[0] & 255)<<24) | \ ((unsigned long)((y)[1] & 255)<<16) | \ ((unsigned long)((y)[2] & 255)<<8) | \ ((unsigned long)((y)[3] & 255)); } #define STORE64L(x, y) \ { (y)[0] = (unsigned char)(((x)>>56)&255); (y)[1] = (unsigned char)(((x)>>48)&255); \ (y)[2] = (unsigned char)(((x)>>40)&255); (y)[3] = (unsigned char)(((x)>>32)&255); \ (y)[4] = (unsigned char)(((x)>>24)&255); (y)[5] = (unsigned char)(((x)>>16)&255); \ (y)[6] = (unsigned char)(((x)>>8)&255); (y)[7] = (unsigned char)((x)&255); } #define LOAD64L(x, y) \ { x = (((ulong64)((y)[0] & 255))<<56)|(((ulong64)((y)[1] & 255))<<48) | \ (((ulong64)((y)[2] & 255))<<40)|(((ulong64)((y)[3] & 255))<<32) | \ (((ulong64)((y)[4] & 255))<<24)|(((ulong64)((y)[5] & 255))<<16) | \ (((ulong64)((y)[6] & 255))<<8)|(((ulong64)((y)[7] & 255))); } #ifdef ENDIAN_32BITWORD #define STORE32H(x, y) \ { unsigned long __t = (x); memcpy(y, &__t, 4); } #define LOAD32H(x, y) \ memcpy(&(x), y, 4); #define STORE64H(x, y) \ { (y)[7] = (unsigned char)(((x)>>56)&255); (y)[6] = (unsigned char)(((x)>>48)&255); \ (y)[5] = (unsigned char)(((x)>>40)&255); (y)[4] = (unsigned char)(((x)>>32)&255); \ (y)[3] = (unsigned char)(((x)>>24)&255); (y)[2] = (unsigned char)(((x)>>16)&255); \ (y)[1] = (unsigned char)(((x)>>8)&255); (y)[0] = (unsigned char)((x)&255); } #define LOAD64H(x, y) \ { x = (((ulong64)((y)[7] & 255))<<56)|(((ulong64)((y)[6] & 255))<<48)| \ (((ulong64)((y)[5] & 255))<<40)|(((ulong64)((y)[4] & 255))<<32)| \ (((ulong64)((y)[3] & 255))<<24)|(((ulong64)((y)[2] & 255))<<16)| \ (((ulong64)((y)[1] & 255))<<8)|(((ulong64)((y)[0] & 255))); } #else /* 64-bit words then */ #define STORE32H(x, y) \ { unsigned long __t = (x); memcpy(y, &__t, 4); } #define LOAD32H(x, y) \ { memcpy(&(x), y, 4); x &= 0xFFFFFFFF; } #define STORE64H(x, y) \ { ulong64 __t = (x); memcpy(y, &__t, 8); } #define LOAD64H(x, y) \ { memcpy(&(x), y, 8); } #endif /* ENDIAN_64BITWORD */ #endif /* ENDIAN_BIG */ #define BSWAP(x) ( ((x>>24)&0x000000FFUL) | ((x<<24)&0xFF000000UL) | \ ((x>>8)&0x0000FF00UL) | ((x<<8)&0x00FF0000UL) ) #ifdef _MSC_VER /* instrinsic rotate */ #include <stdlib.h> #pragma intrinsic(_lrotr,_lrotl) #define ROR(x,n) _lrotr(x,n) #define ROL(x,n) _lrotl(x,n) #elif defined(__GNUC__) && defined(__i386__) && !defined(INTEL_CC) static inline unsigned long ROL(unsigned long word, int i) { __asm__("roll %%cl,%0" :"=r" (word) :"0" (word),"c" (i)); return word; } static inline unsigned long ROR(unsigned long word, int i) { __asm__("rorl %%cl,%0" :"=r" (word) :"0" (word),"c" (i)); return word; } #else /* rotates the hard way */ #define ROL(x, y) ( (((unsigned long)(x)<<(unsigned long)((y)&31)) | (((unsigned long)(x)&0xFFFFFFFFUL)>>(unsigned long)(32-((y)&31)))) & 0xFFFFFFFFUL) #define ROR(x, y) ( ((((unsigned long)(x)&0xFFFFFFFFUL)>>(unsigned long)((y)&31)) | ((unsigned long)(x)<<(unsigned long)(32-((y)&31)))) & 0xFFFFFFFFUL) #endif #define ROL64(x, y) \ ( (((x)<<((ulong64)(y)&63)) | \ (((x)&CONST64(0xFFFFFFFFFFFFFFFF))>>((ulong64)64-((y)&63)))) & CONST64(0xFFFFFFFFFFFFFFFF)) #define ROR64(x, y) \ ( ((((x)&CONST64(0xFFFFFFFFFFFFFFFF))>>((ulong64)(y)&CONST64(63))) | \ ((x)<<((ulong64)(64-((y)&CONST64(63)))))) & CONST64(0xFFFFFFFFFFFFFFFF)) #undef MAX #undef MIN #define MAX(x, y) ( ((x)>(y))?(x):(y) ) #define MIN(x, y) ( ((x)<(y))?(x):(y) ) typedef struct md5_state { ulong64 length; ulong32 state[4], curlen; unsigned char buf[64]; }md5_state; extern void md5_init(md5_state * md); extern int md5_process(md5_state * md, const unsigned char *buf, unsigned long len); extern int md5_done(md5_state * md, unsigned char *hash); extern int md5_test(void); #if ARGTYPE == 0 #include <signal.h> /* this is the default LibTomCrypt macro */ extern void crypt_argchk(char *v, char *s, int d); #define _ARGCHK(x) if (!(x)) { crypt_argchk(#x, __FILE__, __LINE__); } #elif ARGTYPE == 1 /* fatal type of error */ #define _ARGCHK(x) assert((x)) #elif ARGTYPE == 2 #define _ARGCHK(x) #endif #ifdef __cplusplus } #endif #endif /* CRYPT_H_ */
主函数:
#include "md5.h" #ifdef MD5 #define F(x,y,z) (z ^ (x & (y ^ z))) #define G(x,y,z) (y ^ (z & (y ^ x))) #define H(x,y,z) (x^y^z) #define I(x,y,z) (y^(x|(~z))) #define FF(a,b,c,d,M,s,t) \ a = (a + F(b,c,d) + M + t); a = ROL(a, s) + b; #define GG(a,b,c,d,M,s,t) \ a = (a + G(b,c,d) + M + t); a = ROL(a, s) + b; #define HH(a,b,c,d,M,s,t) \ a = (a + H(b,c,d) + M + t); a = ROL(a, s) + b; #define II(a,b,c,d,M,s,t) \ a = (a + I(b,c,d) + M + t); a = ROL(a, s) + b; #define max 10240 #ifdef SMALL_CODE static const unsigned char Worder[64] = { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, 1,6,11,0,5,10,15,4,9,14,3,8,13,2,7,12, 5,8,11,14,1,4,7,10,13,0,3,6,9,12,15,2, 0,7,14,5,12,3,10,1,8,15,6,13,4,11,2,9 }; static const unsigned char Rorder[64] = { 7,12,17,22,7,12,17,22,7,12,17,22,7,12,17,22, 5,9,14,20,5,9,14,20,5,9,14,20,5,9,14,20, 4,11,16,23,4,11,16,23,4,11,16,23,4,11,16,23, 6,10,15,21,6,10,15,21,6,10,15,21,6,10,15,21 }; static const ulong32 Korder[64] = { 0xd76aa478UL, 0xe8c7b756UL, 0x242070dbUL, 0xc1bdceeeUL, 0xf57c0fafUL, 0x4787c62aUL, 0xa8304613UL, 0xfd469501UL, 0x698098d8UL, 0x8b44f7afUL, 0xffff5bb1UL, 0x895cd7beUL, 0x6b901122UL, 0xfd987193UL, 0xa679438eUL, 0x49b40821UL, 0xf61e2562UL, 0xc040b340UL, 0x265e5a51UL, 0xe9b6c7aaUL, 0xd62f105dUL, 0x02441453UL, 0xd8a1e681UL, 0xe7d3fbc8UL, 0x21e1cde6UL, 0xc33707d6UL, 0xf4d50d87UL, 0x455a14edUL, 0xa9e3e905UL, 0xfcefa3f8UL, 0x676f02d9UL, 0x8d2a4c8aUL, 0xfffa3942UL, 0x8771f681UL, 0x6d9d6122UL, 0xfde5380cUL, 0xa4beea44UL, 0x4bdecfa9UL, 0xf6bb4b60UL, 0xbebfbc70UL, 0x289b7ec6UL, 0xeaa127faUL, 0xd4ef3085UL, 0x04881d05UL, 0xd9d4d039UL, 0xe6db99e5UL, 0x1fa27cf8UL, 0xc4ac5665UL, 0xf4292244UL, 0x432aff97UL, 0xab9423a7UL, 0xfc93a039UL, 0x655b59c3UL, 0x8f0ccc92UL, 0xffeff47dUL, 0x85845dd1UL, 0x6fa87e4fUL, 0xfe2ce6e0UL, 0xa3014314UL, 0x4e0811a1UL, 0xf7537e82UL, 0xbd3af235UL, 0x2ad7d2bbUL, 0xeb86d391UL }; #endif #include <signal.h> #if (ARGTYPE == 0) void crypt_argchk(char *v, char *s, int d) { fprintf(stderr, "_ARGCHK '%s' failure on line %d of file %s\n", v, d, s); (void)raise(SIGABRT); } #endif void zeromem(void *dst, size_t len) { unsigned char *mem = (unsigned char *)dst; _ARGCHK(dst != NULL); while (len-- > 0) *mem++ = 0; } void burn_stack(unsigned long len) { unsigned char buf[32]; zeromem(buf, sizeof(buf)); if (len > (unsigned long)sizeof(buf)) burn_stack(len - sizeof(buf)); } /*this function compress 512bits once,and it is called by md5_process and md5_done*/ #ifdef CLEAN_STACK static void _md5_compress(md5_state *md, unsigned char *buf) #else static void md5_compress(md5_state *md, unsigned char *buf) #endif { ulong32 i, W[16], a, b, c, d; #ifdef SMALL_CODE ulong32 t; #endif /* copy the state into 512-bits into W[0..15] */ for (i = 0; i < 16; i++) { LOAD32L(W[i], buf + (4*i)); } /* copy state */ a = md->state[0]; b = md->state[1]; c = md->state[2]; d = md->state[3]; #ifdef SMALL_CODE for (i = 0; i < 16; ++i) { /*第一轮变换*/ FF(a,b,c,d,W[Worder[i]],Rorder[i],Korder[i]); t = d; d = c; c = b; b = a; a = t; } for (; i < 32; ++i) { /*第二轮变换*/ GG(a,b,c,d,W[Worder[i]],Rorder[i],Korder[i]); t = d; d = c; c = b; b = a; a = t; } for (; i < 48; ++i) { /*第三轮变换*/ HH(a,b,c,d,W[Worder[i]],Rorder[i],Korder[i]); t = d; d = c; c = b; b = a; a = t; } for (; i < 64; ++i) { /*第四轮变换*/ II(a,b,c,d,W[Worder[i]],Rorder[i],Korder[i]); t = d; d = c; c = b; b = a; a = t; } #else FF(a,b,c,d,W[0],7,0xd76aa478UL) /*第一轮变换开始*/ FF(d,a,b,c,W[1],12,0xe8c7b756UL) FF(c,d,a,b,W[2],17,0x242070dbUL) FF(b,c,d,a,W[3],22,0xc1bdceeeUL) FF(a,b,c,d,W[4],7,0xf57c0fafUL) FF(d,a,b,c,W[5],12,0x4787c62aUL) FF(c,d,a,b,W[6],17,0xa8304613UL) FF(b,c,d,a,W[7],22,0xfd469501UL) FF(a,b,c,d,W[8],7,0x698098d8UL) FF(d,a,b,c,W[9],12,0x8b44f7afUL) FF(c,d,a,b,W[10],17,0xffff5bb1UL) FF(b,c,d,a,W[11],22,0x895cd7beUL) FF(a,b,c,d,W[12],7,0x6b901122UL) FF(d,a,b,c,W[13],12,0xfd987193UL) FF(c,d,a,b,W[14],17,0xa679438eUL) FF(b,c,d,a,W[15],22,0x49b40821UL) /*第一轮变换结束*/ GG(a,b,c,d,W[1],5,0xf61e2562UL) /*第二轮变换开始*/ GG(d,a,b,c,W[6],9,0xc040b340UL) GG(c,d,a,b,W[11],14,0x265e5a51UL) GG(b,c,d,a,W[0],20,0xe9b6c7aaUL) GG(a,b,c,d,W[5],5,0xd62f105dUL) GG(d,a,b,c,W[10],9,0x02441453UL) GG(c,d,a,b,W[15],14,0xd8a1e681UL) GG(b,c,d,a,W[4],20,0xe7d3fbc8UL) GG(a,b,c,d,W[9],5,0x21e1cde6UL) GG(d,a,b,c,W[14],9,0xc33707d6UL) GG(c,d,a,b,W[3],14,0xf4d50d87UL) GG(b,c,d,a,W[8],20,0x455a14edUL) GG(a,b,c,d,W[13],5,0xa9e3e905UL) GG(d,a,b,c,W[2],9,0xfcefa3f8UL) GG(c,d,a,b,W[7],14,0x676f02d9UL) GG(b,c,d,a,W[12],20,0x8d2a4c8aUL) /*第二轮变换结束*/ HH(a,b,c,d,W[5],4,0xfffa3942UL) /*第三轮变换开始*/ HH(d,a,b,c,W[8],11,0x8771f681UL) HH(c,d,a,b,W[11],16,0x6d9d6122UL) HH(b,c,d,a,W[14],23,0xfde5380cUL) HH(a,b,c,d,W[1],4,0xa4beea44UL) HH(d,a,b,c,W[4],11,0x4bdecfa9UL) HH(c,d,a,b,W[7],16,0xf6bb4b60UL) HH(b,c,d,a,W[10],23,0xbebfbc70UL) HH(a,b,c,d,W[13],4,0x289b7ec6UL) HH(d,a,b,c,W[0],11,0xeaa127faUL) HH(c,d,a,b,W[3],16,0xd4ef3085UL) HH(b,c,d,a,W[6],23,0x04881d05UL) HH(a,b,c,d,W[9],4,0xd9d4d039UL) HH(d,a,b,c,W[12],11,0xe6db99e5UL) HH(c,d,a,b,W[15],16,0x1fa27cf8UL) HH(b,c,d,a,W[2],23,0xc4ac5665UL) /*第三轮变换结束*/ II(a,b,c,d,W[0],6,0xf4292244UL) /*第四轮变换开始*/ II(d,a,b,c,W[7],10,0x432aff97UL) II(c,d,a,b,W[14],15,0xab9423a7UL) II(b,c,d,a,W[5],21,0xfc93a039UL) II(a,b,c,d,W[12],6,0x655b59c3UL) II(d,a,b,c,W[3],10,0x8f0ccc92UL) II(c,d,a,b,W[10],15,0xffeff47dUL) II(b,c,d,a,W[1],21,0x85845dd1UL) II(a,b,c,d,W[8],6,0x6fa87e4fUL) II(d,a,b,c,W[15],10,0xfe2ce6e0UL) II(c,d,a,b,W[6],15,0xa3014314UL) II(b,c,d,a,W[13],21,0x4e0811a1UL) II(a,b,c,d,W[4],6,0xf7537e82UL) II(d,a,b,c,W[11],10,0xbd3af235UL) II(c,d,a,b,W[2],15,0x2ad7d2bbUL) II(b,c,d,a,W[9],21,0xeb86d391UL) /*第四轮变换结束*/ #endif md->state[0] = md->state[0] + a; md->state[1] = md->state[1] + b; md->state[2] = md->state[2] + c; md->state[3] = md->state[3] + d; } #ifdef CLEAN_STACK static void md5_compress(md5_state *md, unsigned char *buf) { _md5_compress(md, buf); burn_stack(sizeof(ulong32) * 21); } #endif void md5_init(md5_state * md) { _ARGCHK(md != NULL); /* md->state[0] = 0x12114567UL; //四个链接变量 md->state[1] = 0x89abcdefUL; md->state[2] = 0xfedcba98UL; md->state[3] = 0x76543210UL; */ md->state[0] = 0x67452301UL; md->state[1] = 0xefcdab89UL; md->state[2] = 0x98badcfeUL; md->state[3] = 0x10325476UL; md->curlen = 0; md->length = 0; } /*假设信息的长度为n×512+l(0<=l<512)。这个函数处理前面的n×512比特,留下的由md5_done函数完成。 */ int md5_process (md5_state * md, const unsigned char *buf, unsigned long len) { unsigned long n; _ARGCHK(md != NULL); _ARGCHK(buf != NULL); if (md->curlen > sizeof(md->buf)) { return CRYPT_INVALID_ARG; } while (len > 0) { if (md->curlen == 0 && len >= 64) { md5_compress (md, (unsigned char *)buf); md->length += 64 * 8; buf += 64; len -= 64; } else { n = MIN(len, (len - md->curlen)); memcpy(md->buf + md->curlen, buf, (size_t)n); md->curlen += n; buf+= n; len-= n; if (md->curlen == len) { md5_compress (md, md->buf); md->length += 8*len; md->curlen = 0; } } } return CRYPT_OK; } int md5_done(md5_state * md, unsigned char *hash) { int i; _ARGCHK(md != NULL); _ARGCHK(hash != NULL); if (md->curlen >= sizeof(md->buf)) { return CRYPT_INVALID_ARG; } md->length += md->curlen * 8; /* append the '1' bit */ md->buf[md->curlen++] = (unsigned char)0x80; /* if the length is currently above 56 bytes we append zeros * then compress. Then we can fall back to padding zeros and length * encoding like normal. */ if (md->curlen > 56) { while (md->curlen < 64) { md->buf[md->curlen++] = (unsigned char)0; } md5_compress(md, md->buf); md->curlen = 0; } /* pad upto 56 bytes of zeroes */ while (md->curlen < 56) { md->buf[md->curlen++] = (unsigned char)0; } /* store length */ STORE64L(md->length, md->buf+56); md5_compress(md, md->buf); /* copy output */ for (i = 0; i < 4; i++) { STORE32L(md->state[i], hash+(4*i)); } #ifdef CLEAN_STACK zeromem(md, sizeof(md5_state)); #endif return CRYPT_OK; } int md5_test(void) { #ifndef LTC_TEST return CRYPT_NOP; #else static const struct { char *msg; unsigned char hash[16]; } tests[] = { { "", { 0xd4, 0x1d, 0x8c, 0xd9, 0x8f, 0x00, 0xb2, 0x04, 0xe9, 0x80, 0x09, 0x98, 0xec, 0xf8, 0x42, 0x7e } }, { "a", {0x0c, 0xc1, 0x75, 0xb9, 0xc0, 0xf1, 0xb6, 0xa8, 0x31, 0xc3, 0x99, 0xe2, 0x69, 0x77, 0x26, 0x61 } }, { "abc", { 0x90, 0x01, 0x50, 0x98, 0x3c, 0xd2, 0x4f, 0xb0, 0xd6, 0x96, 0x3f, 0x7d, 0x28, 0xe1, 0x7f, 0x72 } }, { "message digest", { 0xf9, 0x6b, 0x69, 0x7d, 0x7c, 0xb7, 0x93, 0x8d, 0x52, 0x5a, 0x2f, 0x31, 0xaa, 0xf1, 0x61, 0xd0 } }, { "abcdefghijklmnopqrstuvwxyz", { 0xc3, 0xfc, 0xd3, 0xd7, 0x61, 0x92, 0xe4, 0x00, 0x7d, 0xfb, 0x49, 0x6c, 0xca, 0x67, 0xe1, 0x3b } }, { "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789", { 0xd1, 0x74, 0xab, 0x98, 0xd2, 0x77, 0xd9, 0xf5, 0xa5, 0x61, 0x1c, 0x2c, 0x9f, 0x41, 0x9d, 0x9f } }, { "12345678901234567890123456789012345678901234567890123456789012345678901234567890", { 0x57, 0xed, 0xf4, 0xa2, 0x2b, 0xe3, 0xc9, 0x55, 0xac, 0x49, 0xda, 0x2e, 0x21, 0x07, 0xb6, 0x7a } }, { NULL, { 0 } } }; int i,j; unsigned char tmp[16]; md5_state md; for (i = 0; tests[i].msg != NULL; i++) { md5_init(&md); md5_process(&md, (unsigned char *)tests[i].msg, (unsigned long)strlen(tests[i].msg)); md5_done(&md, tmp); if (memcmp(tmp, tests[i].hash, 16) != 0) { return CRYPT_FAIL_TESTVECTOR; } for(j=0;j<16;j++) printf("%x",tmp[j]); printf("\n"); } return CRYPT_OK; #endif } #endif void md5(char *mssage) { md5_state md; int i; unsigned char hash[16]; md5_init(&md); md5_process(&md,(const unsigned char *)mssage,(unsigned long)strlen(mssage)); md5_done(&md,hash); printf("哈希值为:"); for(i=0;i<16;i++) printf("%x",hash[i]); } void main() { md5_test(); int flag=1,option; FILE *fp; char message[max]; while(flag) { printf("请选择待处理数据类型[0]字符串,[1]文件:"); scanf("%d",&option); if (option==0) { printf("请输入待处理的字符串: "); scanf("%s",message); md5(message); printf("\n"); } else if (option==1) { char filename[100]; printf("请输入文件名(路径): "); scanf("%s",filename); fp=fopen(filename,"r"); if(!fp) {printf("文件找不到!\n");continue;} while(!feof(fp)) { fgets(message,sizeof(message),fp); } md5(message); printf("\n"); } else { flag=0; } } }
运行结果:
小结:
MD5以512位分组来处理输入的信息,且每一分组又被划分为16个32位子分组,经过了一系列的处理后,算法的输出由四个32位分组组成,将这四个32位分组级联后将生成一个128位散列值;
SHA将输入流按照每块512位(64个字节)进行分块,并产生20个字节(160位)的被称为信息认证代码或信息摘要的输出。
MD5加密算法采用的是 小端存储模式,即高位数据放在高地址,低位数据放在低地址。
SHA加密算法采用的是 大端存储模式,即高位数据放在低地址,低位数据放在高地址。