Some improvements to musl libc build script.

[slunkcrypt.git] / README.md

---
title: "![SlunkCrypt](etc/img/SlunkCrypt-Logo.png)"
---


Introduction
============

SlunkCrypt is an experimental cross-platform cryptography library and command-line tool. A fully-featured GUI is provided for the Windows platform.

Please refer to the section [*encryption algorithm*](#encryption-algorithm) for more details!


Legal Warning
=============

Use of SlunkCrypt may be illegal in countries where encryption is outlawed. We believe it is legal to use SlunkCrypt in many countries all around the world, but we are not lawyers, and so if in doubt you should seek legal advice before downloading it. You may find useful information at [cryptolaw.org](http://www.cryptolaw.org/), which collects information on cryptography laws in many countries.


System Requirements
===================

The SlunkCrypt library and the command-line application currently run on the following platforms:

* **Microsoft Windows** (Windows XP SP-3, or later) — 32-Bit (i686) and 64-Bit (x86-64)
* **Linux** (kernel version 3.17, or later) — 32-Bit (i686) and 64-Bit (x86-64)
* **Various BSD flavors** (tested on NetBSD 9.2, FreeBSD 13.0 and OpenBSD 7.0) — 32-Bit (i686) and 64-Bit (x86-64)
* **Solaris** (tested on Solaris 11.4 and OmniOS/illumos) — 32-Bit (i686) and 64-Bit (x86-64)
* **GNU/Hurd** (tested on Debian GNU/Hurd 0.9) — 32-Bit (i686)
* **Haiku** (tested on Haiku R1/b3) — 32-Bit (i686) and 64-Bit (x86-64)
* **Mac OS X** (tested on “Big Sur”) — Intel-based (x86-64) and Apple Silicon (AArch64)

The SlunkCrypt GUI application currently runs on the following platforms:

* **Microsoft Windows** with .NET Framework 4.5 — can be installed on Windows Vista, or later


GUI Usage
=========

This is how the graphical user interface (GUI) for SlunkCrypt looks on [Windows 11](https://i.pinimg.com/736x/0f/57/4c/0f574c445b497cc4748a9e1ab7491582--taps-funny-stuff.jpg):

![](etc/img/SlunkCrypt-GUI.jpg)

Prerequisites
-------------

Please be sure to install the **.NET Framework 4.5**, or any later *.NET Framework 4.x* version, before running the SlunkCrypt GUI application:  
<https://dotnet.microsoft.com/download/dotnet-framework>

***Note:*** If you are running Windows 8 or later, then almost certainly a suitable version of the .NET Framework is already installed &#128526;

Settings
--------

The following settings can be adjusted in the `slunkcrypt-gui.exe.config` configuration file:

- **`DisableBusyIndicator`:**  
  If set to `true`, the “busy indicator” animation will be *disabled* on application startup &mdash; default value: `false`.

- **`ThreadCount`:**  
  Specifies the number of worker threads to use &mdash; default value: `0` (i.e. detect the number of available processors and create one thread for each processor).

- **`KeepIncompleteFiles`:**  
  If set to `true`, incomplete or corrupted output files will *not* be deleted &mdash; default value: `false`.

Command-line Usage
==================

This section describes the SlunkCypt command-line application.

Synopsis
--------

The SlunkCypt command-line program is invoked as follows:

    slunkcrypt --encrypt [pass:<pass>|file:<file>] <input.txt> <output.enc>
    slunkcrypt --decrypt [pass:<pass>|file:<file>] <input.enc> <output.txt>
    slunkcrypt --make-pw [<length>]

Commands
--------

One of the following commands **must** be chosen:

- **`--encrypt` (`-e`):**  
  Run application in ***encrypt*** mode. Reads the given *plaintext* and generates *ciphertext*.
- **`--decrypt` (`-d`):**  
  Run application in ***decrypt*** mode. Reads the given *ciphertext* and restores *plaintext*.
- **`--make-pw` (`-p`):**  
  Generate a "strong" random passphrase, suitable for use with SlunkCrypt.
- **`--self-test` (`-t`):**  
  Run the application in ***self-test*** mode. Program will exit after all test are completed.

Options
-------

The following command-line options are available:

- **`pass:<pass>`**:  
  * Specifies the "secret" passphrase directly on the command-line. This is considered *insecure*.
- **`file:<file>`**:  
  * Specifies a file to read the passphrase from. Only the *first* line of the file will be read!
  * *Note:* It is also possible to specify **`-`** in order to read the passphrase from the *stdin*.
- **`<input>`**:  
  * In ***encrypt*** mode &ndash; specifies the *plaintext* file (unencrypted information) that is to be encrypted.
  * In ***decrypt*** mode &ndash; specifies the *ciphertext* file (result of encryption) that is to be decrypted.
- **`<output>`**:  
  * In ***encrypt*** mode &ndash; specifies the file where the *ciphertext* (result of encryption) will be stored.
  * In ***decrypt*** mode &ndash; specifies the file where the *plaintext* (unencrypted information) will be stored.
- **`<length>`**:  
  * Specifies the length of the passphrase to be generated. If *not* specified, defaults to 24.

### Remarks {.unlisted}

- The same passphrase must be used for both, ***encrypt*** and ***decrypt*** mode. The decryption of the ciphertext will only be possible, if the "correct" passphrase is known. It is recommended to choose a "random" password that is at least 12 characters in length and consists of a mix of upper-case characters, lower-case characters, digits as well as special characters.

- Passing the passphrase directly on the command-line is insecure, because the full command-line may be visible to other users!

Environment
-----------

The following environment variables may be used:

- **`SLUNK_PASSPHRASE`**:  
  Specifies the "secret" passphrase. This environment variables is only evaluated, if the passphrase was **not** specified on the command-line.  
  Passing the passphrase via environment variable is considered more secure, because environment variables are not normally visible to other (unprivileged) users.

- **`SLUNK_KEEP_INCOMPLETE`**:  
  If set to a *non-zero* value, incomplete or corrupted output files will **not** be deleted automatically. By default, the files will be deleted.

- **`SLUNK_THREADS`**:  
  Specifies the number of worker threads to use. By default, SlunkCrypt detects the number of available processors and creates one thread for each processor.

Examples
--------

Here are some examples on how to use the SlunkCrypt command-line application:

### Example #1 {.unlisted}

1. Let's generate a new random (secure) password first:

       slunkcrypt --make-pw

   *Example output:*

       cdG2=fh<C=3[SSCzf[)iDjIV

2. Now, encrypt the plaintext message, using the generated password:

       slunkcrypt --encrypt pass:"cdG2=fh<C=3[SSCzf[)iDjIV" plaintext.txt ciphertext.enc

   Optionally, let's have a look at the ciphertext:

       hexdump -C ciphertext.enc

3. Finally, decrypt the ciphertext, using the same password as before:

       slunkcrypt --decrypt pass:"cdG2=fh<C=3[SSCzf[)iDjIV" ciphertext.enc plaintext.out

   Optionally, verify that the decrypted file is identical to the original:

       sha256sum -b plaintext.txt plaintext.out

### Example #2 {.unlisted}

1. Generate a new password and store it to a text file:

       slunkcrypt --make-pw > passwd.txt

   Optionally, output the generated password to the terminal:

       cat passwd.txt

2. Encrypt file by reading the password from the text file:

       slunkcrypt --encrypt file:passwd.txt plaintext.txt ciphertext.enc

### Example #3 {.unlisted}

1. Generate a new password directly to an environment variable:

       MY_PASSWD="$(slunkcrypt --make-pw)"

   Optionally, output the generated password to the terminal:

       echo "${MY_PASSWD}"

2. Encrypt file by reading the password from the *stdin*:

       slunkcrypt --encrypt - plaintext.txt ciphertext.enc <<< "${MY_PASSWD}"


Encryption algorithm
====================

The SlunkCrypt algorithm is based on core concepts of the well-known [**Enigma**](https://en.wikipedia.org/wiki/Enigma_machine) machine but with numerous improvements, largely inspired by R. Anderson's [***“A Modern Rotor Machine”***](https://rdcu.be/cBo8y):

- The original Enigma machine had only *three* (or, in some models, *four*) rotors, plus a static "reflector" wheel. In SlunkCrypt, we uses **256** simulated rotors for an improved security. Furthermore, the original Enigma machine supported only 26 distinct symbols, i.e. the letters `A` to `Z`. In SlunkCrypt, we use **256** distinct symbols, i.e. the byte values `0x00` to `0xFF`, which allows the encryption (and decryption) of arbitrary streams of bytes, rather than just plain text. Of course, SlunkCrypt can encrypt (and decrypt) text files as well.

- In the original Enigma machine, the signal passes through the rotors *twice*, once in forward direction and then again in backwards direction &ndash; thus the "reflector" wheel. This way, the Enigma's encryption was made *involutory*, i.e. encryption and decryption were the same operation. While this was highly convenient, it also severely weakened the cryptographic strength of the Enigma machine, because the number of possible permutations was reduced drastically! This is one of the main reasons why the Enigma machine eventually was defeated. In SlunkCrypt, the signal passes through the simulated rotors just *once*, in order to maximize the number of possible permutations. This eliminates the most important known weakness of the Enigma machine. Obviously, in SlunkCrypt, separate "modes" for encryption and decryption need to be provided, since encryption and decryption *no* longer are the same operation.

- In the original Enigma machine, the rightmost rotor was moved, by one step, after every symbol. Meanwhile, all other rotors were moved, by one step, *only* when their right-hand neighbor had completed a full turn &ndash; much like the odometer in a car. The fact that most of the rotors remained in the same "static" position most of the time was an important weakness of the Enigma machine. Also, the sequence of the Enigma's rotor positions started to repeat after only 16,900 characters. SlunkCrypt employs an improved stepping algorithm, based on a ***linear-feedback shift register* (LSFR)**, ensuring that *all* rotors move frequently and in a "randomized" unpredictable pattern. The rotor positions of SlunkCrypt practically *never* repeat.

- The internal wiring of each of the original Enigma machine's rotors was *fixed*. Each rotor "type" came with a different internal wiring (i.e. permutation). Some models had up to eight rotor "types" to choose from, but only three or four rotors were used at a time. Nonetheless, the internal wiring (i.e. permutation) of each of the supplied rotors was **not** modifiable. This severely restricted the key space of the Enigma machine, as far as the rotors are concerned, because *only* the order of the rotors and the initial position of each rotor could be varied. In SlunkCrypt, a fully *randomized* wiring (i.e. permutation) is generated from the password for each of the 256 simulated rotors. The initial rotor positions are *randomized* as well.

- SlunkCrypt does **not** currently implement the *plugboard* (“Steckerbrett”) of the original Enigma machine. That is because, even though the plugboard has a large key space, it is just a *fixed* substitution cipher that does *not* contribute much to the cryptographic strength of the Enigma machine. In fact, the plugboard could be "erased" by Welchman's [diagonal board](https://en.wikipedia.org/wiki/Bombe#Stecker_values).


Programming Interface (API)
===========================

This section describes the SlunkCypt library interface for software developers.

Getting started
---------------

In order to use the SlunkCypt library in your C++ code, include **`<slunkcrypt.hpp>`** header and instantiate the appropriate SlunkCypt classes:

### Example #1 {.unlisted}

Here is a simple example on how to use the SlunkCrypt [**`Encryptor`**](#encryptor) class:

    #include <slunkcrypt.hpp>
    #include <fstream>
    #include <iostream>

    int main()
    {
        /* Open input and output files here */
        uint8_t buffer[BUFF_SIZE];
        slunkcrypt::Encryptor slunk_encrypt(passphrase);
        while (input.good())
        {
            input.read(reinterpret_cast<char*>(buffer), BUFF_SIZE);
            if ((!input.bad()) && (input.gcount() > 0))
            {
                if (!slunk_encrypt.inplace(buffer, (size_t)input.gcount()))
                {
                    /* Implement error handling here */
                }
                output.write(reinterpret_cast<char*>(buffer), count);
            }
        }
        std::cout << std::hex << slunk_encrypt.get_nonce() << std::endl;
    }

### Example #2 {.unlisted}

Here is a simple example on how to use the SlunkCrypt [**`Decryptor`**](#decryptor) class:

    #include <slunkcrypt.hpp>
    #include <fstream>
    #include <iostream>

    int main()
    {
        /* Open input and output files here */
        uint8_t buffer[BUFF_SIZE];
        slunkcrypt::Decryptor slunk_decrypt(passphrase, nonce);
        while (input.good())
        {
            input.read(reinterpret_cast<char*>(buffer), BUFF_SIZE);
            if ((!input.bad()) && (input.gcount() > 0))
            {
                if (!slunk_decrypt.inplace(buffer, (size_t)input.gcount()))
                {
                    /* Implement error handling here */
                }
                output.write(reinterpret_cast<char*>(buffer), count);
            }
        }
    }

C++11 API
---------

This section describes the "high-level" C++11 API of the SlunkCrypt library. All SlunkCrypt classes live in the **`slunkcrypt`** namespace.

### Encryptor

Class for *encrypting* data using the SlunkCrypt library.

#### Constructor

Create and initialize a new **``Encryptor``** instance. Also generated a new, random nonce.

    Encryptor::Encryptor(
      const std::string &passwd,
      const size_t thread_count = 0U
    );

***Parameters:***

  * `passwd`  
    The password to "protect" the message. The password is given as an `std::string`, e.g. UTF-8 encoded characters. The same password **may** be used to encrypt *multiple* messages. Also, the same password **must** be used for both, encryption *and* decryption; it will *only* be possible decrypt the ciphertext, if the "correct" password is known. The password must be kept confidential under all circumstances!
     
    *Note:* In order to thwart *brute force* attacks, it is recommended to choose a "random" password that is at least 12 characters in length and that consists of upper-case characters, lower-case characters, digits as well as other "special" characters.

  * `thread_count`  
    Specifies the number of worker threads to use (optional). By default, SlunkCrypt detects the number of available processors and creates one thread for each processor.

***Exceptions:***

  * Throws `std::runtime_error`, if the nonce could not be generated, or if the SlunkCrypt context could not be allocated.

#### Encryptor::process() [1]

Encrypt the next message chunk, using separate input/output buffers.

    bool process(
      const uint8_t *const input,
      uint8_t *const output,
      size_t length
    );

***Parameters:***

 * `input`  
    A pointer to the *input* buffer containing the next chunk of the plaintext to be encrypted. The plaintext is given as a byte array (`uint8_t`). This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially.
    
    The *input* buffer must contain *at least* `length` bytes of data. If the buffer is longer than `length` bytes, then only the first `length` bytes will be processed and the remainder is ignored!

 * `output`  
    A pointer to the *output* buffer where the ciphertext chunk that corresponds to the given plaintext chunk will be stored. The ciphertext is stored as a byte array (`uint8_t`); it has the same length as the plaintext data.
    
    The *output* buffer must provide sufficient space for storing *at least* `length` bytes of encrypted data. If the buffer is longer than `length` bytes, then only the first `length` bytes of the buffer will be filled with encrypted data!

 * `length`  
   The length of the plaintext chunk contained in the *input* buffer given by the `input` parameter, in bytes. At the same time, this determines the minimum required size of the *output* buffer given by the `output` parameters, in bytes.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

#### Encryptor::process() [2]

Encrypt the next message chunk, using separate input/output containers (`std::vector`).

    bool process(
      const std::vector<uint8_t> &input,
      std::vector<uint8_t> &output
    );

***Parameters:***

 * `input`  
    A reference to the `std::vector<uint8_t>` instance containing the next chunk of the plaintext to be encrypted. This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially.

 * `output`  
    A reference to the `std::vector<uint8_t>` instance where the ciphertext that corresponds to the given plaintext will be stored.
    
    The `output.size()` must be *greater than or equal* to `input.size()`. If the `output.size()` is larger than the `input.size()`, then only the first `input.size()` elements of `output` will be filled with encrypted data!

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned. The function fails, if the *output* `std::vector` is too small.

#### Encryptor::inplace() [1]

Encrypt the next message chunk, using a single buffer.

    bool inplace(
      uint8_t *const buffer,
      size_t length
    );

***Parameters:***

 * `buffer`  
    A pointer to the buffer initially containing the next chunk of the plaintext to be encrypted. The plaintext is given as a byte array (`uint8_t`). This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially. The ciphertext chunk that corresponds to the given plaintext chunk will be stored to the *same* buffer, thus replacing the plaintext data.
    
    The buffer must initially contain *at least* `length` bytes of input data; the first `length` bytes of the buffer will be overwritten with the encrypted data. If the buffer is longer than `length` bytes, then only the first `length` bytes will be processed and overwritten.

 * `length`  
   The length of the plaintext chunk initially contained in the input/output buffer given by the `buffer` parameter, in bytes. At the same time, this determines the portion of the input/output buffer that will be overwritten with encrypted data, in bytes.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

#### Encryptor::inplace() [2]

Encrypt the next message chunk, using a single container (`std::vector`).

    bool inplace(
      std::vector<uint8_t> &buffer
    );

***Parameters:***

 * `buffer`  
    A reference to the `std::vector<uint8_t>` initially containing the next chunk of the plaintext to be encrypted. This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially. The ciphertext chunk that corresponds to the given plaintext chunk will be stored to the *same* `std::vector<uint8_t>`, thus replacing all the plaintext data.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

#### Encryptor::get_nonce()

Retrieve the random nonce that is used to encrypt the message.

    uint64_t get_nonce();

***Return value:***

  * Returns the nonce that is used to encrypt the message. The purpose of the nonce is to ensure that each message will be encrypted differently, even when the same password is used to encrypt multiple (possibly identical) messages. Therefore, a new random nonce must be chosen for each message! It is not necessary to keep the nonce confidential, but the same nonce must be used for both, encryption and decryption. Typically, the nonce is stored/transmitted alongside the ciphertext.
    
    *Note:* The `Encryptor` class automatically generates a new, random nonce for each message to be encrypted. Use *this* function to retrieve that nonce, so that it can be passed to `Decryptor` for decryption later.

### Decryptor

Class for *decrypting* data using the SlunkCrypt library.

#### Constructor

Create and initialize a new **``Decryptor``** instance.

    Decryptor::Decryptor(
      const std::string &passwd,
      const uint64_t nonce,
      const size_t thread_count = 0U
    );

***Parameters:***

  * `passwd`  
    The password to "protect" the message. The password is given as an `std::string`, e.g. UTF-8 encoded characters. The same password **may** be used to encrypt *multiple* messages. Also, the same password **must** be used for both, encryption *and* decryption; it will *only* be possible decrypt the ciphertext, if the "correct" password is known. The password must be kept confidential under all circumstances!
    
    *Note:* In order to thwart *brute force* attacks, it is recommended to choose a "random" password that is at least 12 characters in length and that consists of upper-case characters, lower-case characters, digits as well as other "special" characters.

  * `nonce`  
    The *nonce* (number used once) to be used for the decryption process. The purpose of the nonce is to ensure that each message will be encrypted differently, even when the same password is used to encrypt *multiple* (possibly identical) messages. Therefore, a new *random* nonce **must** be chosen for each message! It is *not* necessary to keep the nonce confidential, but the same nonce **must** be used for both, encryption *and* decryption. Typically, the nonce is stored/transmitted alongside the ciphertext.
    
    *Note:* The `Encryptor` class automatically generates a new, random nonce for each message to be encrypted. Use `Encryptor::get_nonce()` to retrieve that nonce, so that it can be passed to `Decryptor` for decryption later.

  * `thread_count`  
    Specifies the number of worker threads to use (optional). By default, SlunkCrypt detects the number of available processors and creates one thread for each processor.

***Exceptions:***

  * Throws `std::runtime_error`, if the SlunkCrypt context could not be allocated.

#### Decryptor::process() [1]

Decrypt the next message chunk, using separate input/output buffers.

    bool process(
      const uint8_t *const input,
      uint8_t *const output,
      size_t length
    );

***Parameters:***

 * `input`  
    A pointer to the *input* buffer containing the next chunk of the ciphertext to be decrypted. The ciphertext is given as a byte array (`uint8_t`).
    
    The *input* buffer must contain *at least* `length` bytes of data. If the buffer is longer than `length` bytes, then only the first `length` bytes will be processed and the remainder is ignored!

 * `output`  
    A pointer to the *output* buffer where the plaintext chunk that corresponds to the given ciphertext chunk will be stored. The plaintext is stored as a byte array (`uint8_t`); it has the same length as the ciphertext data.
    
    The *output* buffer must provide sufficient space for storing *at least* `length` bytes of decrypted data. If the buffer is longer than `length` bytes, then only the first `length` bytes of the buffer will be filled with decrypted data!

 * `length`  
   The length of the ciphertext chunk contained in the *input* buffer given by the `input` parameter, in bytes. At the same time, this determines the minimum required size of the *output* buffer given by the `output` parameters, in bytes.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

#### Decryptor::process() [2]

Decrypt the next message chunk, using separate input/output containers (`std::vector`).

    bool process(
      const std::vector<uint8_t> &input,
      std::vector<uint8_t> &output
    );

***Parameters:***

 * `input`  
    A reference to the `std::vector<uint8_t>` instance containing the next chunk of the ciphertext to be decrypted.

 * `output`  
    A reference to the `std::vector<uint8_t>` instance where the plaintext that corresponds to the given ciphertext will be stored.
    
    The `output.size()` must be *greater than or equal* to `input.size()`. If the `output.size()` is greater than the `input.size()`, then only the first `input.size()` elements of `output` will be filled with decrypted data!

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned. The function fails, if the *output* `std::vector` is too small.

#### Decryptor::inplace() [1]

Decrypt the next message chunk, using a single buffer.

    bool inplace(
      uint8_t *const buffer,
      size_t length
    );

***Parameters:***

 * `buffer`  
    A pointer to the buffer initially containing the next chunk of the ciphertext to be decrypted. The ciphertext is given as a byte array (`uint8_t`). The plaintext that corresponds to the given ciphertext will be stored to the *same* buffer, replacing the plaintext data.
    
    The buffer must initially contain *at least* `length` bytes of input data; the first `length` bytes of the buffer will be overwritten with the encrypted data. If the buffer is longer than `length` bytes, then only the first `length` bytes will be processed and overwritten.

 * `length`  
   The length of the ciphertext chunk initially contained in the input/output buffer given by the `buffer` parameter, in bytes. At the same time, this determines the portion of the input/output buffer that will be overwritten with decrypted data, in bytes.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

#### Decryptor::inplace() [2]

Decrypt the next message chunk, using a single container (`std::vector`).

    bool inplace(
      std::vector<uint8_t> &buffer
    );

***Parameters:***

 * `buffer`  
    A reference to the `std::vector<uint8_t>` initially containing the next chunk of the ciphertext to be decrypted. The plaintext that corresponds to the given ciphertext will be stored to the *same* `std::vector<uint8_t>`, replacing all the ciphertext data.

***Return value:***

  * If successful, `true` is returned; otherwise `false` is returned.

C99 API
-------

This section describes the "low-level" C99 API of the SlunkCypt library.

### Functions

The SlunkCypt library defines the following functions:

#### slunkcrypt_alloc()

Allocate and initialize a new SlunkCrypt encryption/decryption context.

    slunkcrypt_t slunkcrypt_alloc(
        const uint64_t nonce,
        const uint8_t *const passwd,
        const size_t passwd_len,
        const int mode
    );

***Parameters:***

  * `nonce`  
    The *nonce* (number used once) to be used for the encryption/decryption process. The purpose of the nonce is to ensure that each message will be encrypted differently, even when the same password is used to encrypt *multiple* (possibly identical) messages. Therefore, a new *random* nonce **must** be chosen for each message to be encrypted! It is *not* necessary to keep the nonce confidential, but the same nonce **must** be used for both, encryption *and* decryption. Typically, the nonce is stored/transmitted alongside the ciphertext.
    
    *Note:* It is recommended to generate a random nonce via the [`slunkcrypt_generate_nonce()`](#slunkcrypt_generate_nonce) function for each message!
  
  * `passwd`  
    The password to "protect" the message. The password is given as a byte array (`uint8_t`), e.g. UTF-8 encoded characters; a terminating NULL character is *not* required, as the length of the password is specified explicitly. The same password **may** be used to encrypt *multiple* messages. Also, the same password **must** be used for both, encryption *and* decryption; it will *only* be possible decrypt the ciphertext, if the "correct" password is known. The password must be kept confidential under all circumstances!
    
    *Note:* In order to thwart *brute force* attacks, it is recommended to choose a "random" password that is at least 12 characters in length and that consists of upper-case characters, lower-case characters, digits as well as other "special" characters.
  
  * `passwd_len`  
    The length of password given by the `passwd` parameter, in bytes, **not** counting a terminating NULL character. The minimum/maximum length of the password are given by the `SLUNKCRYPT_PWDLEN_MIN` and `SLUNKCRYPT_PWDLEN_MAX` constants, respectively.

  * `mode`  
    The mode of operation. Use `SLUNKCRYPT_ENCRYPT` in order to set up this context for *encryption*, or use `SLUNKCRYPT_DECRYPT` in order to set up this context for *decryption*.

***Return value:***

  * If successful, a handle to the new SlunkCrypt context is return; otherwise `SLUNKCRYPT_NULL` is returned.
    
    *Note:* Applications **should** treat `slunkcrypt_t` as an *opaque* handle type. Also, as soon as the SlunkCrypt context is *not* needed anymore, the application **shall** call [`slunkcrypt_free()`](#slunkcrypt_free) in order to "erase" and de-allocate that context. If a SlunkCrypt context is *not* de-allocated properly, it will result in a memory leak!

#### slunkcrypt_alloc_ext()

Allocate and initialize a new SlunkCrypt encryption/decryption context with additional parameters.

    slunkcrypt_t slunkcrypt_alloc_ext(
        const uint64_t nonce,
        const uint8_t *const passwd,
        const size_t passwd_len,
        const int mode,
        const slunkparam_t *const param
    );

***Parameters:***

  * `nonce`  
    The *nonce* (number used once) to be used for the encryption/decryption process. The purpose of the nonce is to ensure that each message will be encrypted differently, even when the same password is used to encrypt *multiple* (possibly identical) messages. Therefore, a new *random* nonce **must** be chosen for each message to be encrypted! It is *not* necessary to keep the nonce confidential, but the same nonce **must** be used for both, encryption *and* decryption. Typically, the nonce is stored/transmitted alongside the ciphertext.
    
    *Note:* It is recommended to generate a random nonce via the [`slunkcrypt_generate_nonce()`](#slunkcrypt_generate_nonce) function for each message!
  
  * `passwd`  
    The password to "protect" the message. The password is given as a byte array (`uint8_t`), e.g. UTF-8 encoded characters; a terminating NULL character is *not* required, as the length of the password is specified explicitly. The same password **may** be used to encrypt *multiple* messages. Also, the same password **must** be used for both, encryption *and* decryption; it will *only* be possible decrypt the ciphertext, if the "correct" password is known. The password must be kept confidential under all circumstances!
    
    *Note:* In order to thwart *brute force* attacks, it is recommended to choose a "random" password that is at least 12 characters in length and that consists of upper-case characters, lower-case characters, digits as well as other "special" characters.
  
  * `passwd_len`  
    The length of password given by the `passwd` parameter, in bytes, **not** counting a terminating NULL character. The minimum/maximum length of the password are given by the `SLUNKCRYPT_PWDLEN_MIN` and `SLUNKCRYPT_PWDLEN_MAX` constants, respectively.

  * `mode`  
    The mode of operation. Use `SLUNKCRYPT_ENCRYPT` in order to set up this context for *encryption*, or use `SLUNKCRYPT_DECRYPT` in order to set up this context for *decryption*.

  * `param`  
    Additional parameters used to initialize the SlunkCrypt context, given as a pointer to a [`slunkparam_t`](#slunkcrypt-parameters) struct. The memory for the struct must be allocated by the caller and SlunkCrypt does **not** take owner ship of this memory; it will *copy* the relevant fields. The caller is responsible to free the struct; it can be allocated with automatic storage duration.

***Return value:***

  * If successful, a handle to the new SlunkCrypt context is return; otherwise `SLUNKCRYPT_NULL` is returned.
    
    *Note:* Applications **should** treat `slunkcrypt_t` as an *opaque* handle type. Also, as soon as the SlunkCrypt context is *not* needed anymore, the application **shall** call [`slunkcrypt_free()`](#slunkcrypt_free) in order to "erase" and de-allocate that context. If a SlunkCrypt context is *not* de-allocated properly, it will result in a memory leak!

#### slunkcrypt_reset()
  
Re-initialize an existing SlunkCrypt encryption/decryption context.

    int slunkcrypt_reset(
        const slunkcrypt_t context,
        const uint64_t nonce,
        const uint8_t *const passwd,
        const size_t passwd_len,
        const int mode
    );

***Parameters:***

  * `context`  
    The existing SlunkCrypt context to be re-initialized. This must be a valid handle that was returned by a previous invocation of the [`slunkcrypt_alloc()`](#slunkcrypt_alloc) function.

  * *other parameters*:  
    Please refer to the [`slunkcrypt_alloc()`](#slunkcrypt_alloc) function for details!

***Return value:***

  * If successful, `SLUNKCRYPT_SUCCESS` is returned; otherwise `SLUNKCRYPT_FAILURE` or `SLUNKCRYPT_ABORTED` is returned.

#### slunkcrypt_free()

De-allocate an existing SlunkCrypt encryption/decryption context. This will "clear" and release any memory occupied by the context.

    void slunkcrypt_free(
        const slunkcrypt_t context
    );

***Parameters:***

 * `context`  
    The existing SlunkCrypt context to be de-allocated. This must be a valid handle that was returned by a previous invocation of the [`slunkcrypt_alloc()`](#slunkcrypt_alloc) function.
    
    *Note:* Once a handle has been passed to this function, that handle is *invalidated* and **must not** be used again!

#### slunkcrypt_generate_nonce()

Generate a new random *nonce* (number used once), using the system's "cryptographically secure" entropy source.

    int slunkcrypt_generate_nonce(
      int64_t *const nonce
    );

***Parameters:***

 * `nonce`  
   A pointer to a variable of type `int64_t` that receives the new random nonce.

***Return value:***

  * If successful, `SLUNKCRYPT_SUCCESS` is returned; otherwise `SLUNKCRYPT_FAILURE` or `SLUNKCRYPT_ABORTED` is returned.

#### slunkcrypt_process()

Encrypt or decrypt the next message chunk, using separate input/output buffers.

    int slunkcrypt_process(
        const slunkcrypt_t context,
        const uint8_t *const input,
        uint8_t *const output,
        size_t length
    );

***Parameters:***

 * `context`  
    The existing SlunkCrypt context to be used for processing the message chunk. This context will be updated.

    *Note:* This function operates either in "encryption" mode or in "decryption" mode, depending on how the given SlunkCrypt context has been [initialized](#slunkcrypt_alloc) or [re-initialized](#slunkcrypt_reset).

 * `input`  
    A pointer to the *input* buffer containing the next chunk of the plaintext to be encrypted (*encryption* mode), or the next chunk of the ciphertext to be decrypted (*decryption* mode). The input data is given as a byte array (`uint8_t`). This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially by this function.
    
    The *input* buffer must contain *at least* `length` bytes of data. If the buffer is longer than `length` bytes, then only the first `length` bytes will be processed; the remainder is ignored!

 * `output`  
    A pointer to the *output* buffer where the ciphertext that corresponds to the given plaintext chunk (*encryption* mode), or the plaintext that corresponds to the given ciphertext chunk (*decryption* mode) will be stored. The output data is stored as a byte array (`uint8_t`) and it always has the same length as the input data.
    
    The *output* buffer **must** provide sufficient space for storing *at least* `length` bytes. If the buffer is longer than `length` bytes, then only the first `length` bytes will be filled!

 * `length`  
   The length of the given plaintext chunk (*encryption* mode), or the length of the given ciphertext chunk (*decryption* mode) in the `input` buffer, in bytes. At the same time, this parameter determines the minimum required size of the `output` buffer, in bytes. If this parameter is set to *zero*, the function does nothing; this is *not* considered an error.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `SLUNKCRYPT_SUCCESS` is returned; otherwise `SLUNKCRYPT_FAILURE` or `SLUNKCRYPT_ABORTED` is returned.

#### slunkcrypt_inplace()

Encrypt or decrypt the next message chunk, using a *single* input/output buffer.

    int slunkcrypt_inplace(
        const slunkcrypt_t context,
        uint8_t *const buffer,
        size_t length
    );

***Parameters:***

 * `context`  
    The existing SlunkCrypt context to be used for processing the message chunk. This context will be updated.

    *Note:* This function operates either in "encryption" mode or in "decryption" mode, depending on how the given SlunkCrypt context has been [initialized](#slunkcrypt_alloc) or [re-initialized](#slunkcrypt_reset).

 * `buffer`  
    A pointer to the buffer containing the next chunk of the plaintext to be encrypted (*encryption* mode), or the next chunk of the ciphertext to be decrypted (*decryption* mode). The input data is given as a byte array (`uint8_t`). This can be arbitrary binary data, e.g. UTF-8 encoded text. NULL bytes are **not** treated specially by this function. The ciphertext that corresponds to the given plaintext chunk (*encryption* mode), or the plaintext that corresponds to the given ciphertext chunk (*decryption* mode) will be stored to the ***same*** buffer.
    
    The given buffer **must** initially contain *at least* `length` bytes of input data. The first `length` bytes in the buffer will be processed and will be *overwritten* with the corresponding output data. If the buffer is longer than `length` bytes, then only the first `length` bytes in the buffer will be processed; the remainder is ignored!

 * `length`  
   The length of the plaintext chunk (*encryption* mode), or the length of the ciphertext chunk (*decryption* mode) initially contained in the input/output buffer, in bytes.

   *Note:* It is recommended to process chunks of at least &sim;64 KB each, in order to take full advantage of multi-threading.

***Return value:***

  * If successful, `SLUNKCRYPT_SUCCESS` is returned; otherwise `SLUNKCRYPT_FAILURE` or `SLUNKCRYPT_ABORTED` is returned.

#### slunkcrypt_random_bytes()

Generate a sequence of random bytes, using the system's "cryptographically secure" entropy source.

    size_t slunkcrypt_random_bytes(
      uint8_t *const buffer,
      const size_t length
    );

***Parameters:***

 * `buffer`  
   A pointer to the *output* buffer where the random bytes will be stored.

   The *output* buffer **must** provide sufficient space for storing *at least* `length` bytes of random data. *At most* the first `length` bytes of the buffer will be filled with random data!

 * `length`  
   The number of random bytes to be generated. At the same time, this parameter determines the minimum required size of the `output` buffer, in bytes.

***Return value:***

  * The number of random bytes that have been generated and that have been stored to the `output` buffer buffer is returned.
    
    The number of generated random bytes can be *at most* `length`. Less than `length` random bytes will be generated, if and only if the the system's "cryptographically secure" entropy source could *not* provide the requested number of bytes at this time &ndash; in that case, you can try again later. The number of generated bytes can be as low as **0**.


#### slunkcrypt_bzero()

Erase the contents of a byte array, by overwriting it with *zero* bytes. Compiler optimizations will **not** remove the erase operation.

    void slunkcrypt_bzero(
      void *const buffer,
      const size_t length
    );

***Parameters:***

 * `buffer`  
   A pointer to the buffer whose content is to be erased.

   The buffer **must** be *at least* `length` bytes in size. If the buffer is longer than `length` bytes, then *only* the first `length` bytes of the buffer will be erased!

 * `length`  
   The size of the buffer to be erased, in bytes.

### Types

#### SlunkCrypt parameters

The `slunkparam_t` struct is used to pass additional parameters that will be used for initializing the SlunkCrypt context. It contains the following fields:

* `version` &ndash; The version of the parameter struct; **must** be set to *`SLUNKCRYPT_PARAM_VERSION`*.
* `thread_count` &ndash; The number of worker threads to use. If this parameter is set to **0**, which is the *default* value, then SlunkCrypt automatically detects the number of available (logical) processors and creates one thread for each processor. Also, the number of threads is capped to a maximum of `MAX_THREADS` (currently defined as **16**).

### Global variables

The SlunkCypt library defines the following global variables:

#### Version information

These variables can be used to determine the version of the SlunkCrypt library at runtime, using the [semantic versioning](https://semver.org/spec/v2.0.0.html) scheme:

* `const uint16_t SLUNKCRYPT_VERSION_MAJOR` &ndash; The current *major* version.
* `const uint16_t SLUNKCRYPT_VERSION_MINOR` &ndash; The current *minor* version.
* `const uint16_t SLUNKCRYPT_VERSION_PATCH` &ndash; The current *patch* version.
* `const char *SLUNKCRYPT_BUILD` &ndash; The build date and time, as a C string, in the *`"mmm dd yyyy, hh:mm:ss"`* format.

#### Abort request

If this flag is set to a non-zero value by the application, any ongoing SlunkCypt library invocation will be aborted as quickly as possible:

* `volatile int g_slunkcrypt_abort_flag` &ndash; The global abort flag.

  *Note:* Applications may set this flag in their signal handler, e.g. when a `SIGINT` is received, in order to "gracefully" shut down the SlunkCypt library. All long-running library functions will return `SLUNKCRYPT_ABORTED`, if the invocation was interrupted. The application still is responsible for [free](#slunkcrypt_free)'ing any SlunkCypt contexts that it allocated successfully!

### Constants

The SlunkCypt library defines the following constants:

#### Mode of operation

The SlunkCypt library supports the following modes of operation:

* `SLUNKCRYPT_ENCRYPT` &ndash; Run library in *encryption* mode, i.e. consume plaintext and produce ciphertext.
* `SLUNKCRYPT_DECRYPT` &ndash; Run library in *decryption* mode, i.e. consume ciphertext and produce plaintext.

#### Limits

The following limits are defined for the SlunkCypt library:

* `SLUNKCRYPT_PWDLEN_MIN` &ndash; The *minimum* required length of a password, currently **8** bytes.
* `SLUNKCRYPT_PWDLEN_MAX` &ndash; The *maximum* allowed length of a password, currently **256** bytes.

#### Error codes

SlunkCypt library functions that return an error code may return one of the following constants:

* `SLUNKCRYPT_SUCCESS` &ndash; The operation completed successfully.
* `SLUNKCRYPT_FAILURE` &ndash; The operation has failed.
* `SLUNKCRYPT_ABORTED` &ndash; The operation was aborted before completion, as requested by the application.

Thread safety
-------------

The following functions are fully "thread-safe" and thus may safely be called by *any* thread at *any* time ***without*** the need for synchronization:

* `slunkcrypt_alloc()`
* `slunkcrypt_generate_nonce()`
* `slunkcrypt_random_bytes()`
* `slunkcrypt_bzero()`
* `Encryptor::Encryptor()`
* `Decryptor::Decryptor()`

The following functions are "reentrant" and thus may safely be called by *any* thread at *any* time ***without*** the need for synchronization &ndash; provided that each instance of `slunkcrypt_t`, `Encryptor` or `Decryptor` is "owned" by a *single* thread **and** that each instance will *exclusively* be access by its respective "owner" thread:

* `slunkcrypt_reset()`
* `slunkcrypt_free()`
* `slunkcrypt_process()`
* `slunkcrypt_inplace()`
* `Encryptor::process()`
* `Encryptor::inplace()`
* `Encryptor::get_nonce()`
* `Decryptor::process()`
* `Decryptor::inplace()`

***Note:*** If the same `slunkcrypt_t`, `Encryptor` or `Decryptor` instance needs to be shared across *multiple* threads (i.e. the same instance is accessed by *concurrent* threads), then the application **must** *serialize* any invocation of the above functions on that shared instance, by using a suitable synchronization mechanism! This can be achieved by using a [*mutex*](https://linux.die.net/man/3/pthread_mutex_lock).


Source Code
===========

The latest SlunkCrypt source code is available from the official Git mirrors at:

* <https://osdn.net/projects/slunkcrypt/scm/git/SlunkCrypt/>
* <https://gitlab.com/lord_mulder/slunkcrypt/>
* <https://bitbucket.org/muldersoft/slunkcrypt/>
* <https://repo.or.cz/slunkcrypt.git>
* <https://punkindrublic.mooo.com:3000/Muldersoft/SlunkCrypt>


Build Instructions
==================

SlunkCrypt can be built from the sources on Microsoft Windows or any POSIX-compatible platform, using a C-compiler that supports the C99 standard.

* **Microsoft Windows:**  
  Project/solution files for [Visual Studio](https://visualstudio.microsoft.com/) are provided. These should work “out of the box” with Visual Studio 2017 or any later version.  
  Just open the solution, select the “Release” configuration, choose the “x86” or “x64” platform, and finally press `F5`.  
  Visual Studio also is the only way to build the SlunkCrypt GUI, which is based on Microsoft.NET and Windows Presentation Foundation (WPF).

  Alternatively, SlunkCrypt can built using [Mingw-w64](https://www.mingw-w64.org/) (available via [MSYS2](https://www.msys2.org/)) or even [Cygwin](https://www.cygwin.com/) &ndash; see Linux instructions for details!

* **Linux:**  
  Please make sure that the *C compiler* (GCC or Clang) as well as *Make* are installed. Then simply run **`make -B`** from the project's base directory!

  If not already installed, the required build tools can usually be installed via your distribution's package manager.  
  For example, on Debian-based distributions, the command **`sudo apt install build-essential`** installs all the required build tools at once.

  In order to create a *fully-static* binary of SlunkCrypt that runs on ***any*** Linux distribution from the last decade, you can use [musl libc](https://musl.libc.org/):  
  `make -B CC=musl-gcc STATIC=1`

* **BSD and Solaris:**  
  SlunkCrypt can be built on various BSD flavors and Solaris, but the command **`gmake -B`** needs to be used here, since the native `make` doesn't work!  
  GNU Make can be installed from the package manager. For example, use **`pkg install gmake`** on FreeBSD or **`pkg_add gmake`** on OpenBSD.

* **Mac OS X:**  
  Once you have managed to find a terminal (or even better, connect via SSH), Mac OS X almost works like a proper operating system.  
  The Xcode command-line tools can be installed with the command **`xcode-select --install`**, if not present yet. Then just type **`make -B`** to build!

  *Hint:* If you want to build with GCC, which produces faster code than Apple's Xcode compiler, you may install it on Mac OS X via [Homebrew](https://formulae.brew.sh/formula/gcc).


License
=======

This work has been released under the **CC0 1.0 Universal** license.

For details, please refer to:  
<https://creativecommons.org/publicdomain/zero/1.0/legalcode>

Acknowledgement
---------------

SlunkCrypt incorporates code from the following *third-party* software projects:

* The "checksum" algorithm used by the SlunkCrypt command-line application was adapted from the **BLAKE2** reference C implementation.  
  ```
   BLAKE2 reference source code package - reference C implementations
   Copyright 2012, Samuel Neves <sneves@dei.uc.pt>.  You may use this under the
   terms of the CC0, the OpenSSL Licence, or the Apache Public License 2.0, at
   your option.  The terms of these licenses can be found at:
   - CC0 1.0 Universal : http://creativecommons.org/publicdomain/zero/1.0
   - OpenSSL license   : https://www.openssl.org/source/license.html
   - Apache 2.0        : http://www.apache.org/licenses/LICENSE-2.0

   More information about the BLAKE2 hash function can be found at
   https://blake2.net.
   ```

* ***Windows only:*** Builds of SlunkCypt that have multi-threading enabled use the **POSIX Threads for Windows (pthreads4w)** library.  
  ```
  Pthreads4w - POSIX Threads for Windows
  Copyright 1998 John E. Bossom
  Copyright 1999-2018, Pthreads4w contributors

  Homepage: https://sourceforge.net/projects/pthreads4w/
 
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at

      http://www.apache.org/licenses/LICENSE-2.0

  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License.
  ```


&marker;