OpenSSL使用EVP与算法API进行对称加密

Hi i have installed openssl on my linux machine and going through the header files and documentation (which is highly insufficint :( ).

嗨，我已经在我的linux机器上安装了openssl并通过头文件和文档（这是非常不充分的:(）。

i am trying to build a project(in 'c') which uses symmetric crypto algos (i am focusing on aes256cbc). The problem is i am confused as in how to use the library functions in my code.

我正在尝试构建一个使用对称加密算法的项目（在'c'中）（我专注于aes256cbc）。问题是我很困惑如何在我的代码中使用库函数。

For my implementation of aes256cbc i can directly use the functions defined in the 'aes.h' header file(which appeared to me at the first place).

对于我的aes256cbc实现，我可以直接使用'aes.h'头文件中定义的函数（首先在我看来）。

But on googling i came accross some tutorial for this which are using 'evp.h' functions to do this http://saju.net.in/code/misc/openssl_aes.c.txt

但在google搜索中，我遇到了一些教程，使用'evp.h'函数来做这个http://saju.net.in/code/misc/openssl_aes.c.txt

Is there a specific reason for this or directly accessing the aes.h functions is better.

是否有特定原因或直接访问aes.h函数更好。

And also if someone can point me to a good documentation/tutorial of any kind on using the crypto library of openssl will be much appreciated.

而且如果有人能指出我使用openssl的加密库的任何类型的良好文档/教程将非常感激。

many thanks

非常感谢

P.S forgive me if i am being naive

如果我天真，请原谅我

2 个解决方案

#1

Using the EVP API has the advantage that you can use the same API for all the symmetric ciphers that OpenSSL supports, in a generic way. This makes it way easier to replace the algorithm used, or make the algorithm user-configurable at a later stage. Most of the code you write is not specific to the encryption algorithm you selected.

使用EVP API的优势在于，您可以以通用方式为OpenSSL支持的所有对称密码使用相同的API。这使得更换所使用的算法变得更容易，或者在稍后阶段使算法可由用户配置。您编写的大多数代码并非特定于您选择的加密算法。

Here's a simple example for encryption with AES-256 in CBC mode:

以下是在CBC模式下使用AES-256进行加密的简单示例：

#include <stdio.h>
#include <openssl/evp.h>

int main()
{
    EVP_CIPHER_CTX ctx;
    unsigned char key[32] = {0};
    unsigned char iv[16] = {0};
    unsigned char in[16] = {0};
    unsigned char out[32]; /* at least one block longer than in[] */
    int outlen1, outlen2;

    EVP_EncryptInit(&ctx, EVP_aes_256_cbc(), key, iv);
    EVP_EncryptUpdate(&ctx, out, &outlen1, in, sizeof(in));
    EVP_EncryptFinal(&ctx, out + outlen1, &outlen2);

    printf("ciphertext length: %d\n", outlen1 + outlen2);

    return 0;
}

For simplicity, I omitted error handling.

为简单起见，我省略了错误处理。

IMO one of the most important pieces of documentation on OpenSSL is Network Security with OpenSSL by Viega/Messier/Chandra. It is from 2002 (0.9.7), so does not cover changes to OpenSSL during the last 10 years, but it is IMO still a less painful way to learn OpenSSL than by using only the manual pages.

IMO关于OpenSSL最重要的文档之一是Viega / Messier / Chandra的OpenSSL网络安全。它是从2002年（0.9.7）开始，因此不包括过去10年中对OpenSSL的更改，但IMO仍然是一种学习OpenSSL的痛苦方式，而不仅仅是使用手册页。

#2

Currently OpenSSL wiki has good documentation on how to use the EVP family of functions: http://wiki.openssl.org/index.php/EVP

目前，OpenSSL wiki有关于如何使用EVP系列函数的良好文档：http：//wiki.openssl.org/index.php/EVP

The other upside of using the EVP over algorithm API is that EVP will automatically use hardware acceleration (like AES-NI instruction set) if available. With algorithm API you need to enable it manually.

使用EVP over算法API的另一个好处是EVP将自动使用硬件加速（如AES-NI指令集）（如果可用）。使用算法API，您需要手动启用它。

#1