MD5算法 C语言 (从OpenSSL库中分离算法:一)
OpenSSL简介：
OpenSSL 是用于传输层安全性 (TLS) 和安全套接字层 (SSL) 协议的一个强大、商业级和功能齐全的工具包，它也是一个通用的密码学库。包含有RSA、SM4、DES、AES等诸多加密算法。
OpenSSL GitHub地址如下：
GitHub - openssl/openssl: TLS/SSL and crypto library
在日常的开发工作中，有时只想用OpenSSL库中的一种算法，此时调用整个OpenSSL库，往往是没必要的；再或者在嵌入式平台使用某一算法，那我们只移植这个算法，没有必要移植整个OpenSSL库。
MD5简介
MD5全称是第5代信息摘要算法，由英文拼写(Message-Digest Algorithm)缩略而成，即MD5，既然是第5代，那么之前还有MD4、MD3、MD2，前几代算法有的经过一段时间的应用，有的只是昙花一现，而MD5可以说是一棵常青树，虽然它也存在被碰撞的安全隐患。
MD5是一种被广泛使用的密码散列函数，可以产生出一个128位（16字节）的散列值（hash value），用于确保信息传输完整一致。
移植过程
一、下载代码
从OpenSSL的Github仓库下载代码，可以下载master分支，也可以下载最新的release版本。将下载的文件解压，得到代码。
重点关注红框目录：
- crypto目录内是各种加密算法
- include目录内是各加密算法对外接口的头文件
二、准备环境
新建Visual Studio C++ 控制台项目"md5test"，编译选项是：x86\Debug
三、准备文件
- 复制OpenSSL源码中：/crypto/md5文件夹到VS工程代码目录下
- 复制OpenSSL源码中：/include/openssl/md5.h文件到VS工程代码目录中md5文件夹内
- 复制OpenSSL源码中：/include/crypto/md32_common.h文件到VS工程代码目录中md5文件夹内
- 复制完成后，VS工程代码目录md5文件夹内文件如下：
删除目录内的如下本项目中无用文件：
- asm
- build.info
- md5_sha1.c
四、修改代码
修改调用文件
在VS工程中含有main函数的主文件:md5test.cpp中，增加包含md5.h头文件
#include "md5/md5.h"
在main函数中添加如下代码
const unsigned char Data[5] = "abcd"; //待计算md5的数据
unsigned char md[MD5_DIGEST_LENGTH] = { 0 }; //md5存储的变量
unsigned char* P = MD5(Data, strlen((const char*)Data), md); //计算md5
for (int i = 0; i < MD5_DIGEST_LENGTH; i++) { //输出md5
printf("%x-", *(md + i));
}
修改后的md5test.cpp代码如下：
#include <iostream>
#include "md5/md5.h"
int main()
{
const unsigned char Data[5] = "abcd"; //待计算md5的数据，若要计算文件的md5，则Data为从文件中读取的数据。
unsigned char md[MD5_DIGEST_LENGTH] = { 0 }; //md5存储的变量
unsigned char* P = MD5(Data, strlen((const char*)Data), md); //计算md5
for (int i = 0; i < MD5_DIGEST_LENGTH; i++) { //输出md5
printf("%02x-", *(md + i));
}
}
修改md5.h
删除条件编译 OPENSSL_MD5_H
删除条件编译 OPENSSL_NO_DEPRECATED_3_0
删除条件编译 OPENSSL_NO_DEPRECATED_3_0
删除条件编译 OPENSSL_NO_MD5
删除头文件# include <openssl/macros.h>
删除头文件# include <openssl/opensslconf.h>
删除头文件# include <openssl/e_os2.h>
删除函数类型宏OSSL_DEPRECATEDIN_3_0
修改或删除后的文件如下：
# pragma once
# include <stddef.h>
# ifdef __cplusplus
extern "C" {
# endif
# define MD5_DIGEST_LENGTH 16
/*
* !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
* ! MD5_LONG has to be at least 32 bits wide. !
* !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
*/
# define MD5_LONG unsigned int
# define MD5_CBLOCK 64
# define MD5_LBLOCK (MD5_CBLOCK/4)
typedef struct MD5state_st {
MD5_LONG A, B, C, D;
MD5_LONG Nl, Nh;
MD5_LONG data[MD5_LBLOCK];
unsigned int num;
} MD5_CTX;
int MD5_Init(MD5_CTX *c);
int MD5_Update(MD5_CTX *c, const void *data, size_t len);
int MD5_Final(unsigned char *md, MD5_CTX *c);
unsigned char *MD5(const unsigned char *d, size_t n,
unsigned char *md);
void MD5_Transform(MD5_CTX *c, const unsigned char *b);
# ifdef __cplusplus
}
# endif
修改md5_dgst.c
删除头文件# include "internal/deprecated.h"
删除头文件# include <openssl/opensslv.h>
修改或删除后的文件如下：
#include <stdio.h>
#include "md5_local.h"
/*
* Implemented from RFC1321 The MD5 Message-Digest Algorithm
*/
#define INIT_DATA_A (unsigned long)0x67452301L
#define INIT_DATA_B (unsigned long)0xefcdab89L
#define INIT_DATA_C (unsigned long)0x98badcfeL
#define INIT_DATA_D (unsigned long)0x10325476L
int MD5_Init(MD5_CTX *c)
{
memset(c, 0, sizeof(*c));
c->A = INIT_DATA_A;
c->B = INIT_DATA_B;
c->C = INIT_DATA_C;
c->D = INIT_DATA_D;
return 1;
}
#ifndef md5_block_data_order
# ifdef X
# undef X
# endif
void md5_block_data_order(MD5_CTX *c, const void *data_, size_t num)
{
const unsigned char *data = data_;
register unsigned MD32_REG_T A, B, C, D, l;
# ifndef MD32_XARRAY
/* See comment in crypto/sha/sha_local.h for details. */
unsigned MD32_REG_T XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7,
XX8, XX9, XX10, XX11, XX12, XX13, XX14, XX15;
# define X(i) XX##i
# else
MD5_LONG XX[MD5_LBLOCK];
# define X(i) XX[i]
# endif
A = c->A;
B = c->B;
C = c->C;
D = c->D;
for (; num--;) {
(void)HOST_c2l(data, l);
X(0) = l;
(void)HOST_c2l(data, l);
X(1) = l;
/* Round 0 */
R0(A, B, C, D, X(0), 7, 0xd76aa478L);
(void)HOST_c2l(data, l);
X(2) = l;
R0(D, A, B, C, X(1), 12, 0xe8c7b756L);
(void)HOST_c2l(data, l);
X(3) = l;
R0(C, D, A, B, X(2), 17, 0x242070dbL);
(void)HOST_c2l(data, l);
X(4) = l;
R0(B, C, D, A, X(3), 22, 0xc1bdceeeL);
(void)HOST_c2l(data, l);
X(5) = l;
R0(A, B, C, D, X(4), 7, 0xf57c0fafL);
(void)HOST_c2l(data, l);
X(6) = l;
R0(D, A, B, C, X(5), 12, 0x4787c62aL);
(void)HOST_c2l(data, l);
X(7) = l;
R0(C, D, A, B, X(6), 17, 0xa8304613L);
(void)HOST_c2l(data, l);
X(8) = l;
R0(B, C, D, A, X(7), 22, 0xfd469501L);
(void)HOST_c2l(data, l);
X(9) = l;
R0(A, B, C, D, X(8), 7, 0x698098d8L);
(void)HOST_c2l(data, l);
X(10) = l;
R0(D, A, B, C, X(9), 12, 0x8b44f7afL);
(void)HOST_c2l(data, l);
X(11) = l;
R0(C, D, A, B, X(10), 17, 0xffff5bb1L);
(void)HOST_c2l(data, l);
X(12) = l;
R0(B, C, D, A, X(11), 22, 0x895cd7beL);
(void)HOST_c2l(data, l);
X(13) = l;
R0(A, B, C, D, X(12), 7, 0x6b901122L);
(void)HOST_c2l(data, l);
X(14) = l;
R0(D, A, B, C, X(13), 12, 0xfd987193L);
(void)HOST_c2l(data, l);
X(15) = l;
R0(C, D, A, B, X(14), 17, 0xa679438eL);
R0(B, C, D, A, X(15), 22, 0x49b40821L);
/* Round 1 */
R1(A, B, C, D, X(1), 5, 0xf61e2562L);
R1(D, A, B, C, X(6), 9, 0xc040b340L);
R1(C, D, A, B, X(11), 14, 0x265e5a51L);
R1(B, C, D, A, X(0), 20, 0xe9b6c7aaL);
R1(A, B, C, D, X(5), 5, 0xd62f105dL);
R1(D, A, B, C, X(10), 9, 0x02441453L);
R1(C, D, A, B, X(15), 14, 0xd8a1e681L);
R1(B, C, D, A, X(4), 20, 0xe7d3fbc8L);
R1(A, B, C, D, X(9), 5, 0x21e1cde6L);
R1(D, A, B, C, X(14), 9, 0xc33707d6L);
R1(C, D, A, B, X(3), 14, 0xf4d50d87L);
R1(B, C, D, A, X(8), 20, 0x455a14edL);
R1(A, B, C, D, X(13), 5, 0xa9e3e905L);
R1(D, A, B, C, X(2), 9, 0xfcefa3f8L);
R1(C, D, A, B, X(7), 14, 0x676f02d9L);
R1(B, C, D, A, X(12), 20, 0x8d2a4c8aL);
/* Round 2 */
R2(A, B, C, D, X(5), 4, 0xfffa3942L);
R2(D, A, B, C, X(8), 11, 0x8771f681L);
R2(C, D, A, B, X(11), 16, 0x6d9d6122L);
R2(B, C, D, A, X(14), 23, 0xfde5380cL);
R2(A, B, C, D, X(1), 4, 0xa4beea44L);
R2(D, A, B, C, X(4), 11, 0x4bdecfa9L);
R2(C, D, A, B, X(7), 16, 0xf6bb4b60L);
R2(B, C, D, A, X(10), 23, 0xbebfbc70L);
R2(A, B, C, D, X(13), 4, 0x289b7ec6L);
R2(D, A, B, C, X(0), 11, 0xeaa127faL);
R2(C, D, A, B, X(3), 16, 0xd4ef3085L);
R2(B, C, D, A, X(6), 23, 0x04881d05L);
R2(A, B, C, D, X(9), 4, 0xd9d4d039L);
R2(D, A, B, C, X(12), 11, 0xe6db99e5L);
R2(C, D, A, B, X(15), 16, 0x1fa27cf8L);
R2(B, C, D, A, X(2), 23, 0xc4ac5665L);
/* Round 3 */
R3(A, B, C, D, X(0), 6, 0xf4292244L);
R3(D, A, B, C, X(7), 10, 0x432aff97L);
R3(C, D, A, B, X(14), 15, 0xab9423a7L);
R3(B, C, D, A, X(5), 21, 0xfc93a039L);
R3(A, B, C, D, X(12), 6, 0x655b59c3L);
R3(D, A, B, C, X(3), 10, 0x8f0ccc92L);
R3(C, D, A, B, X(10), 15, 0xffeff47dL);
R3(B, C, D, A, X(1), 21, 0x85845dd1L);
R3(A, B, C, D, X(8), 6, 0x6fa87e4fL);
R3(D, A, B, C, X(15), 10, 0xfe2ce6e0L);
R3(C, D, A, B, X(6), 15, 0xa3014314L);
R3(B, C, D, A, X(13), 21, 0x4e0811a1L);
R3(A, B, C, D, X(4), 6, 0xf7537e82L);
R3(D, A, B, C, X(11), 10, 0xbd3af235L);
R3(C, D, A, B, X(2), 15, 0x2ad7d2bbL);
R3(B, C, D, A, X(9), 21, 0xeb86d391L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
#endif
修改md5_local.h
删除头文件# include <openssl/e_os2.h>
修改头文件# include <openssl/md5.h>改为#include "md5.h"
修改头文件# include "crypto/md32_common.h"改为#include "md32_common.h"
修改或删除后的文件如下：
#include <stdlib.h>
#include <string.h>
#include "md5.h"
#ifdef MD5_ASM
# if defined(__i386) || defined(__i386__) || defined(_M_IX86) || \
defined(__x86_64) || defined(__x86_64__) || defined(_M_AMD64) || defined(_M_X64)
# define md5_block_data_order ossl_md5_block_asm_data_order
# elif defined(__ia64) || defined(__ia64__) || defined(_M_IA64)
# define md5_block_data_order ossl_md5_block_asm_data_order
# elif defined(__sparc) || defined(__sparc__)
# define md5_block_data_order ossl_md5_block_asm_data_order
# endif
#endif
void md5_block_data_order(MD5_CTX *c, const void *p, size_t num);
#define DATA_ORDER_IS_LITTLE_ENDIAN
#define HASH_LONG MD5_LONG
#define HASH_CTX MD5_CTX
#define HASH_CBLOCK MD5_CBLOCK
#define HASH_UPDATE MD5_Update
#define HASH_TRANSFORM MD5_Transform
#define HASH_FINAL MD5_Final
#define HASH_MAKE_STRING(c,s) do { \
unsigned long ll; \
ll=(c)->A; (void)HOST_l2c(ll,(s)); \
ll=(c)->B; (void)HOST_l2c(ll,(s)); \
ll=(c)->C; (void)HOST_l2c(ll,(s)); \
ll=(c)->D; (void)HOST_l2c(ll,(s)); \
} while (0)
#define HASH_BLOCK_DATA_ORDER md5_block_data_order
#include "md32_common.h"
/*-
#define F(x,y,z) (((x) & (y)) | ((~(x)) & (z)))
#define G(x,y,z) (((x) & (z)) | ((y) & (~(z))))
*/
/*
* As pointed out by Wei Dai, the above can be simplified to the code
* below. Wei attributes these optimizations to Peter Gutmann's
* SHS code, and he attributes it to Rich Schroeppel.
*/
#define F(b,c,d) ((((c) ^ (d)) & (b)) ^ (d))
#define G(b,c,d) ((((b) ^ (c)) & (d)) ^ (c))
#define H(b,c,d) ((b) ^ (c) ^ (d))
#define I(b,c,d) (((~(d)) | (b)) ^ (c))
#define R0(a,b,c,d,k,s,t) { \
a+=((k)+(t)+F((b),(c),(d))); \
a=ROTATE(a,s); \
a+=b; };
#define R1(a,b,c,d,k,s,t) { \
a+=((k)+(t)+G((b),(c),(d))); \
a=ROTATE(a,s); \
a+=b; };
#define R2(a,b,c,d,k,s,t) { \
a+=((k)+(t)+H((b),(c),(d))); \
a=ROTATE(a,s); \
a+=b; };
#define R3(a,b,c,d,k,s,t) { \
a+=((k)+(t)+I((b),(c),(d))); \
a=ROTATE(a,s); \
a+=b; };
修改md5_one.c
删除头文件# include "internal/deprecated.h"
删除头文件# include <openssl/crypto.h>
修改头文件# include <openssl/md5.h>改为#include "md5.h"
删除条件编译 CHARSET_EBCDIC
删除头文件# include <openssl/ebcdic.h>
删除代码中全部的OPENSSL_cleanse();
*解释：经查阅openssl代码中的OPENSSL_cleanse函数声明，此函数作用等同与标准C中的memset(p,0,size)功能，是将参数一内存清0，本文件中调用该函数的地方，其参数1均为栈空间，函数结束后将自动回收，因此没必要手动清0，因此该函数可以不调用，若不放心，可以用memset函数替代。
修改或删除后的文件如下：
#include <stdio.h>
#include <string.h>
#include "md5.h"
unsigned char *MD5(const unsigned char *d, size_t n, unsigned char *md)
{
MD5_CTX c;
static unsigned char m[MD5_DIGEST_LENGTH];
if (md == NULL)
md = m;
if (!MD5_Init(&c))
return NULL;
#ifndef CHARSET_EBCDIC
MD5_Update(&c, d, n);
#else
{
char temp[1024];
unsigned long chunk;
while (n > 0) {
chunk = (n > sizeof(temp)) ? sizeof(temp) : n;
ebcdic2ascii(temp, d, chunk);
MD5_Update(&c, temp, chunk);
n -= chunk;
d += chunk;
}
}
#endif
MD5_Final(md, &c);
return md;
}
修改md32_common.h
删除头文件# include <openssl/crypto.h>
删除代码中全部的OPENSSL_cleanse();
*解释：经查阅openssl代码中的OPENSSL_cleanse函数声明，此函数作用等同与标准C中的memset(p,0,size)功能，是将参数一内存清0，本文件中调用该函数的地方，其参数1均为栈空间，函数结束后将自动回收，因此没必要手动清0，因此该函数可以不调用，若不放心，可以用memset函数替代。
修改或删除后的文件如下：
#if !defined(DATA_ORDER_IS_BIG_ENDIAN) && !defined(DATA_ORDER_IS_LITTLE_ENDIAN)
# error "DATA_ORDER must be defined!"
#endif
#ifndef HASH_CBLOCK
# error "HASH_CBLOCK must be defined!"
#endif
#ifndef HASH_LONG
# error "HASH_LONG must be defined!"
#endif
#ifndef HASH_CTX
# error "HASH_CTX must be defined!"
#endif
#ifndef HASH_UPDATE
# error "HASH_UPDATE must be defined!"
#endif
#ifndef HASH_TRANSFORM
# error "HASH_TRANSFORM must be defined!"
#endif
#ifndef HASH_FINAL
# error "HASH_FINAL must be defined!"
#endif
#ifndef HASH_BLOCK_DATA_ORDER
# error "HASH_BLOCK_DATA_ORDER must be defined!"
#endif
#define ROTATE(a,n) (((a)<<(n))|(((a)&0xffffffff)>>(32-(n))))
#if defined(DATA_ORDER_IS_BIG_ENDIAN)
# define HOST_c2l(c,l) (l =(((unsigned long)(*((c)++)))<<24), \
l|=(((unsigned long)(*((c)++)))<<16), \
l|=(((unsigned long)(*((c)++)))<< 8), \
l|=(((unsigned long)(*((c)++))) ) )
# define HOST_l2c(l,c) (*((c)++)=(unsigned char)(((l)>>24)&0xff), \
*((c)++)=(unsigned char)(((l)>>16)&0xff), \
*((c)++)=(unsigned char)(((l)>> 8)&0xff), \
*((c)++)=(unsigned char)(((l) )&0xff), \
l)
#elif defined(DATA_ORDER_IS_LITTLE_ENDIAN)
# define HOST_c2l(c,l) (l =(((unsigned long)(*((c)++))) ), \
l|=(((unsigned long)(*((c)++)))<< 8), \
l|=(((unsigned long)(*((c)++)))<<16), \
l|=(((unsigned long)(*((c)++)))<<24) )
# define HOST_l2c(l,c) (*((c)++)=(unsigned char)(((l) )&0xff), \
*((c)++)=(unsigned char)(((l)>> 8)&0xff), \
*((c)++)=(unsigned char)(((l)>>16)&0xff), \
*((c)++)=(unsigned char)(((l)>>24)&0xff), \
l)
#endif
/*
* Time for some action :-)
*/
int HASH_UPDATE(HASH_CTX *c, const void *data_, size_t len)
{
const unsigned char *data = data_;
unsigned char *p;
HASH_LONG l;
size_t n;
if (len == 0)
return 1;
l = (c->Nl + (((HASH_LONG) len) << 3)) & 0xffffffffUL;
if (l < c->Nl) /* overflow */
c->Nh++;
c->Nh += (HASH_LONG) (len >> 29); /* might cause compiler warning on
* 16-bit */
c->Nl = l;
n = c->num;
if (n != 0) {
p = (unsigned char *)c->data;
if (len >= HASH_CBLOCK || len + n >= HASH_CBLOCK) {
memcpy(p + n, data, HASH_CBLOCK - n);
HASH_BLOCK_DATA_ORDER(c, p, 1);
n = HASH_CBLOCK - n;
data += n;
len -= n;
c->num = 0;
/*
* We use memset rather than OPENSSL_cleanse() here deliberately.
* Using OPENSSL_cleanse() here could be a performance issue. It
* will get properly cleansed on finalisation so this isn't a
* security problem.
*/
memset(p, 0, HASH_CBLOCK); /* keep it zeroed */
} else {
memcpy(p + n, data, len);
c->num += (unsigned int)len;
return 1;
}
}
n = len / HASH_CBLOCK;
if (n > 0) {
HASH_BLOCK_DATA_ORDER(c, data, n);
n *= HASH_CBLOCK;
data += n;
len -= n;
}
if (len != 0) {
p = (unsigned char *)c->data;
c->num = (unsigned int)len;
memcpy(p, data, len);
}
return 1;
}
void HASH_TRANSFORM(HASH_CTX *c, const unsigned char *data)
{
HASH_BLOCK_DATA_ORDER(c, data, 1);
}
int HASH_FINAL(unsigned char *md, HASH_CTX *c)
{
unsigned char *p = (unsigned char *)c->data;
size_t n = c->num;
p[n] = 0x80; /* there is always room for one */
n++;
if (n > (HASH_CBLOCK - 8)) {
memset(p + n, 0, HASH_CBLOCK - n);
n = 0;
HASH_BLOCK_DATA_ORDER(c, p, 1);
}
memset(p + n, 0, HASH_CBLOCK - 8 - n);
p += HASH_CBLOCK - 8;
#if defined(DATA_ORDER_IS_BIG_ENDIAN)
(void)HOST_l2c(c->Nh, p);
(void)HOST_l2c(c->Nl, p);
#elif defined(DATA_ORDER_IS_LITTLE_ENDIAN)
(void)HOST_l2c(c->Nl, p);
(void)HOST_l2c(c->Nh, p);
#endif
p -= HASH_CBLOCK;
HASH_BLOCK_DATA_ORDER(c, p, 1);
c->num = 0;
#ifndef HASH_MAKE_STRING
# error "HASH_MAKE_STRING must be defined!"
#else
HASH_MAKE_STRING(c, md);
#endif
return 1;
}
#ifndef MD32_REG_T
# if defined(__alpha) || defined(__sparcv9) || defined(__mips)
# define MD32_REG_T long
/*
* This comment was originally written for MD5, which is why it
* discusses A-D. But it basically applies to all 32-bit digests,
* which is why it was moved to common header file.
*
* In case you wonder why A-D are declared as long and not
* as MD5_LONG. Doing so results in slight performance
* boost on LP64 architectures. The catch is we don't
* really care if 32 MSBs of a 64-bit register get polluted
* with eventual overflows as we *save* only 32 LSBs in
* *either* case. Now declaring 'em long excuses the compiler
* from keeping 32 MSBs zeroed resulting in 13% performance
* improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
* Well, to be honest it should say that this *prevents*
* performance degradation.
*/
# else
/*
* Above is not absolute and there are LP64 compilers that
* generate better code if MD32_REG_T is defined int. The above
* pre-processor condition reflects the circumstances under which
* the conclusion was made and is subject to further extension.
*/
# define MD32_REG_T int
# endif
#endif
五、编译运行
将md5.h,md5_dgst.c,md5_local.h,md5_one.c,md32_common.h文件，配置添加到VS工程中。
编译运行，输出md5结果如下：
e2-fc-71-4c-47-27-ee-93-95-f3-24-cd-2e-7f-33-1f-
后续会分享: 基于该剥离后的算法，计算小文件md5，计算大文件md5。
欢迎大家关注、留言讨论、可分享源码
1280
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
