.TH MCRYPT 3 "10 March 2002"
.SH NAME
libmcrypt \- encryption/decryption library
.SH SYNOPSIS
[see also
.I mcrypt.h
for more information]
.SH DESCRIPTION
The
.I libmcrypt
is a data encryption library.
The library is thread safe and provides encryption and decryption functions.
This version of the library supports many encryption algorithms and encryption
modes. Some algorithms which are supported:
SERPENT, RIJNDAEL, 3DES, GOST, SAFER+, CAST-256, RC2, XTEA, 3WAY,
TWOFISH, BLOWFISH, ARCFOUR, WAKE and more. 
.LP
OFB, CBC, ECB, nOFB, nCFB and CFB are the modes that all algorithms may function.
ECB, CBC, encrypt in blocks but CTR, nCFB, nOFB, CFB and OFB in bytes (streams). 
Note that CFB and OFB in the rest of the document represent the "8bit CFB or OFB" mode.
nOFB and nCFB modes represents a n-bit OFB/CFB mode, n is used to represent the
algorithm's block size.
The library supports an extra STREAM mode to include some stream algorithms
like WAKE or ARCFOUR.

In this version of the library all modes and algorithms are modular,
which means that the algorithm and the mode is loaded at run-time.
This way you can add algorithms and modes faster, and much easier.

.I LibMcrypt
includes the following symmetric (block) algorithms:

.B DES: 
The traditional DES algorithm designed by IBM and US NSA. 
Uses 56 bit key and 64 bit block. It is now considered a weak
algorithm, due to its small key size (it was never intended for use
with classified data).

.B 3DES or Triple DES:
DES but with multiple (triple) encryption. It encrypts the plaintext
once, then decrypts it with the second key, and encrypts it again with
the third key (outer cbc mode used for cbc). 
Much better than traditional DES since the key is now 168 bits (actually
the effective key length is 112 bits due to the meet-in-the-middle attack).

.B CAST-128:
CAST was designed in Canada by Carlisle Adams and Stafford Tavares.
The original algorithm used a 64bit key and block. The algorithm
here is CAST-128 (also called CAST5) which has a 128bit key and 64bit block size.

.B CAST-256:
CAST-256 was designed by Carlisle Adams. It is a symmetric cipher designed in
accordance with the CAST design procedure. It is an extention of
the CAST-128, having a 128 bit block size, and up to 256 bit key size.

.B xTEA:
TEA stands for the Tiny Encryption Algorithm. It is a feistel cipher 
designed by David Wheeler & Roger M. Needham.
The original TEA was intended for use in applications where
code size is at a premium, or where it is necessary for someone to remember 
the algorithm and code it on an arbitrary machine at a later time.
The algorithm used here is extended TEA and has a 128bit key size and
64bit block size.

.B 3-WAY:
The 3way algorithm designed by Joan Daemen. It uses key and block size
of 96 bits. 

.B SKIPJACK:
SKIPJACK was designed by the US NSA. It was part of the ill-fated
"Clipper" Escrowed Encryption Standard (EES) (FIPS 185) proposal. It
operates on 64bit blocks and uses a key of 80 bits. SKIPJACK is provided only as an extra module to libmcrypt.

.B BLOWFISH: 
The Blowfish algorithm designed by Bruce Schneier. It is better and faster 
than DES. It can use a key up to 448 bits.

.B TWOFISH:
Twofish was designed by Bruce Schneier, Doug Whiting, John Kelsey, Chris
Hall, David Wagner for Counterpane systems. Intended to be highly secure 
and highly flexible. It uses a 128bit block size and 128,192,256 bit key size.
(Twofish is the default algorithm)

.B LOKI97:
LOKI97 was designed by Lawrie Brown and Josef Pieprzyk. It has a 128-bit
block length and a 256bit key schedule, which can be initialized
using 128, 192 or 256 bit keys. It has evolved from the
earlier LOKI89 and LOKI91 64-bit block ciphers, with a strengthened key
schedule and a larger keyspace.

.B RC2:
RC2 (RC stands for Rivest Cipher) was designed by Ron Rivest. It uses block size of 64 bit and a key size
from 8 to 1024 bits. It is optimized for 16bit microprocessors (reflecting
its age). It is described in the RFC2268.

.B ARCFOUR:
RC4 was designed by Ron Rivest. For several years this algorithm was
considered a trade secret and details were not available. In September 1994 someone
posted the source code in the cypherpunks mailing list. Although the source
code is now available RC4 is trademarked by RSADSI so a compatible cipher named
ARCFOUR is included in the mcrypt distribution. It is a stream cipher
and has a maximum key of 2048 bits.

.B RC6:
RC6 was designed by Ron Rivest for RSA labs. In mcrypt it uses block size of 128 bit and
a key size of 128/192/256 bits.
Refer to RSA Labs and Ron Rivest for any copyright, patent or license issues
for the RC6 algorithm. RC6 is provided only as an extra module to libmcrypt.

.B RIJNDAEL:
Rijndael is a block cipher, designed by Joan Daemen and Vincent Rijmen,
and was approved for the USA's NIST Advanced Encryption Standard, FIPS-197.
The cipher has a variable block length and key length. Rijndael can be implemented very efficiently on a wide range
of processors and in hardware. The design of Rijndael was strongly influenced
by the design of the block cipher Square. 
There exist three versions of this algorithm, namely:
.B RIJNDAEL-128 (the AES winner)
,
.B RIJNDAEL-192
,
.B RIJNDAEL-256
The numerals 128, 192 and 256 stand for the length of the block size.

.B MARS:
MARS is a 128-bit block cipher designed by IBM as a candidate for the 
Advanced Encryption Standard. Refer to IBM for any copyright, patent or license issues
for the MARS algorithm. MARS is provided only as an extra module to libmcrypt.

.B PANAMA:
PANAMA is a cryptographic module that can be used both as a
cryptographic hash function and as a stream cipher. It designed by Joan 
Daemen and Craig Clapp. PANAMA (the stream cipher) is included in
libmcrypt.

.B WAKE:
WAKE stands for Word Auto Key Encryption, and is an encryption system for 
medium speed encryption of blocks and of high security. 
WAKE was designed by David J. Wheeler. It is intended to 
be fast on most computers and relies on repeated table use and having a 
large state space. 

.B SERPENT:
Serpent is a 128-bit block cipher designed by Ross Anderson, Eli Biham and
Lars Knudsen as a candidate for the Advanced Encryption Standard.
Serpent's design was limited to well understood mechanisms, so that
could rely on the wide experience of block cipher cryptanalysis, and
achieve the highest practical level of assurance that no shortcut attack will
be found. Serpent has twice as many rounds as are necessary, to block all
currently known shortcut attacks. Despite these exacting design constraints,
Serpent is faster than DES. 

.B IDEA:
IDEA stands for International Data Encryption Algorithm and was designed
by Xuejia Lai and James Massey. It operates on 64bit blocks and uses a key of 128 bits.
Refer to Ascom-Tech AG for any copyright, patent or license issues
for the IDEA algorithm. IDEA is provided only as an extra module to libmcrypt.

.B ENIGMA (UNIX crypt):
A one-rotor machine designed along the lines of Enigma but considerable
trivialized. Very easy to break for a skilled cryptanalyst.
I suggest against using it. Added just for completeness.

.B GOST:
A former soviet union's algorithm. An acronym for "Gosudarstvennyi Standard" 
or Government Standard. It uses a 256 bit key and a 64 bit block.
 The S-boxes used here are described in the Applied Cryptography book
by Bruce Schneier. They were used in an application for the Central Bank
of the Russian Federation. 
 Some quotes from gost.c:
The standard is written by A. Zabotin (project leader), G.P. Glazkov,
and V.B. Isaeva.  It was accepted and introduced into use by the
action of the State Standards Committee of the USSR on 2 June 1989 as
No. 1409.  It was to be reviewed in 1993, but whether anyone wishes
to take on this obligation from the USSR is questionable.
 This code is based on the 25 November 1993 draft translation
by Aleksandr Malchik, with Whitfield Diffie, of the Government
Standard of the U.S.S.R. GOST 28149-89, "Cryptographic Transformation
Algorithm", effective 1 July 1990.  (Whitfield.Diffie@eng.sun.com)
Some details have been cleared up by the paper "Soviet Encryption
Algorithm" by Josef Pieprzyk and Leonid Tombak of the University
of Wollongong, New South Wales.  (josef/leo@cs.adfa.oz.au)

.B SAFER:
SAFER (Secure And Fast Encryption Routine) is a block cipher developed
by Prof. J.L. Massey at the Swiss Federal Institute of Technology.
There exist four versions of this algorithm, namely:
.B SAFER K-64
,
.B SAFER K-128
,
.B SAFER SK-64
and
.B SAFER SK-128.
The numerals 64 and 128 stand for the length of the user-selected
key, 'K' stands for the original key schedule and 'SK' stands for the
strengthened key schedule (in which some of the "weaknesses" of the
original key schedule have been removed). In mcrypt only SAFER SK-64 and
SAFER SK-128 are used.

.B SAFER+:
SAFER+ was designed by Prof. J.L. Massey, Prof. Gurgen H. Khachatrian and
Dr. Melsik K. Kuregian for Cylink. SAFER+ is based on the existing SAFER
family of ciphers and provides for a block size of 128bits and 128, 192
and 256 bits key length.


.LP
A short description of the modes supported by libmcrypt:

.B STREAM: 
The mode used with stream ciphers. In this mode the keystream from
the cipher is XORed with the plaintext. Thus you should NOT ever use the
same key.

.B ECB: 
The Electronic CodeBook mode. It is the simplest mode to use with a 
block cipher. Encrypts each block independently. It is a block mode
so plaintext length should be a multiple of blocksize (n*blocksize).

.B CBC:
The Cipher Block Chaining mode. It is better than ECB since the plaintext
is XOR'ed with the previous ciphertext. A random block should be placed as the
first block (IV) so the same block or messages always encrypt to something
different. It is a block mode so plaintext length should be a multiple 
of blocksize (n*blocksize).

.B CFB:
The Cipher-Feedback Mode (in 8bit). This is a self-synchronizing
stream cipher implemented from a block cipher. This is the best mode
to use for encrypting strings or streams. This mode requires an IV.

.B OFB:
The Output-Feedback Mode (in 8bit). This is a synchronous
stream cipher implemented from a block cipher. It is intended for use
in noisy lines, because corrupted ciphertext blocks do not corrupt the
plaintext blocks that follow. Insecure (because used in 8bit mode) so it is
recommended not to use it. Added just for completeness.

.B nOFB:
The Output-Feedback Mode (in nbit). n Is the size of the block of the
algorithm. This is a synchronous stream cipher implemented from a block
cipher. It is intended for use in noisy lines, because corrupted ciphertext
blocks do not corrupt the plaintext blocks that follow. This mode
operates in streams.

.B nCFB:
The Cipher-Feedback Mode (in nbit). n Is the size of the block of the
algorithm. This is a self synchronizing stream cipher implemented from a block
cipher. This mode operates in streams.

.B CTR:
The Counter Mode. This is a stream cipher implemented from a block
cipher. This mode uses the cipher to encrypt a set of input blocks, called
counters, to produce blocks that will be XORed with the plaintext.
In libmcrypt the counter is the given IV which is incremented at each step.
This mode operates in streams.

.B Error Recovery in these modes:
If bytes are removed or lost from the file or stream in ECB, CTR, CBC and OFB modes,
are impossible to recover, although CFB and nCFB modes will recover. If some
bytes are altered then a full block of plaintext is affected in ECB, nOFB and CTR modes,
two blocks in CBC, nCFB and CFB modes, but only the corresponding byte in OFB mode.

.LP
Encryption can be done as follows:

A call to function:
.B    MCRYPT mcrypt_module_open( char *algorithm, char* algorithm_directory, 
.B				char* mode, char* mode_directory);

This function associates the algorithm and the mode specified. 
The name of the algorithm is specified in algorithm, eg "twofish", and the algorithm_directory
is the directory where the algorithm is (it may be null if it is
the default). The same applies for the mode.
The library is closed by calling mcrypt_module_close(), but you should
not call that function if mcrypt_generic_end() is called before.
Normally it returns an encryption descriptor,
or MCRYPT_FAILED on error.

A call to function:
.B    int mcrypt_generic_init( MCRYPT td, void *key,
.B                               int lenofkey,
.B                               void *IV);

This function initializes all buffers for the specified thread
The maximum value of lenofkey should be the one obtained by 
calling mcrypt_get_key_size() and every value smaller than this is legal.
Note that Lenofkey should be specified in bytes not bits.
The IV should normally have the size of the algorithms block size, but
you must obtain the size by calling mcrypt_get_iv_size().
IV is ignored in ECB. IV MUST exist in CFB, CBC, STREAM, nOFB and
OFB modes. It needs to be random and unique (but not secret). The same IV must be used
for encryption/decryption. 
After calling this function you can use the descriptor for encryption
or decryption (not both). 
Returns a negative value on error.

To encrypt now call:

.B    int mcrypt_generic( MCRYPT td, void *plaintext, int len);

This is the main encryption function. td is the encryption descriptor
returned by mcrypt_generic_init(). Plaintext is the plaintext you
wish to encrypt and len should be the length (in bytes) of the
plaintext and it should be k*algorithms_block_size if used in a mode
which operated in blocks (cbc, ecb, nofb), or whatever when used in
cfb or ofb which operate in streams. The plaintext is replaced by the
ciphertext. Returns 0 on success.

To decrypt you can call:

.B    int mdecrypt_generic( MCRYPT td, void *ciphertext, int len);

The decryption function. It is almost the same with mcrypt_generic.
Returns 0 on success.

When you're finished you should call:

.B    int mcrypt_generic_end( MCRYPT td);

This function terminates encryption specified by the encryption descriptor (td).
Actually it clears all buffers, and closes all the modules used.
Returns a negative value on error.
.B This function is deprecated.
Use mcrypt_generic_deinit() and mcrypt_module_close() instead.

.B    int mcrypt_generic_deinit( MCRYPT td);

This function terminates encryption specified by the encryption descriptor (td).
Actually it clears all buffers. The difference with mcrypt_generic_end() is that
this function does not close the modules used. Thus you should use mcrypt_module_close().
Using this function you gain in speed if you use the same modules for several 
encryptions.
Returns a negative value on error.

.B    int mcrypt_module_close( MCRYPT td);

This function closes the modules used by the descriptor td.

.P
These are some extra functions that operate on modules that have been opened:
These functions have the prefix mcrypt_enc_*. 

.B    int mcrypt_enc_set_state(MCRYPT td, void *state, int size);
This function sets the state of the algorithm. Can
be used only with block algorithms and certain modes like CBC, CFB etc. 
It is usefully if you want to restart or start a different encryption quickly. 
Returns zero on success. The state is the output of mcrypt_enc_get_state().

.B    int mcrypt_enc_get_state(MCRYPT td, void *state, int *size);
This function returns the state of the algorithm. Can
be used only certain modes and algorithms. The size will hold the
size of the state and the state must have enough bytes to hold it.
Returns zero on success.

.B    int mcrypt_enc_self_test( MCRYPT td);

This function runs the self test on the algorithm specified by the descriptor
td. If the self test succeeds it returns zero. 

.B    int mcrypt_enc_is_block_algorithm_mode( MCRYPT td);

Returns 1 if the mode is for use with block algorithms, otherwise
it returns 0. (eg. 0 for stream, and 1 for cbc, cfb, ofb)

.B    int mcrypt_enc_is_block_algorithm( MCRYPT td);

Returns 1 if the algorithm is a block algorithm or 0 if it is a stream
algorithm.

.B    int mcrypt_enc_is_block_mode( MCRYPT td);

Returns 1 if the mode outputs blocks of bytes or 0 if it outputs bytes.
(eg. 1 for cbc and ecb, and 0 for cfb and stream)

.B    int mcrypt_enc_get_block_size( MCRYPT td);

Returns the block size of the algorithm specified by the encryption descriptor
in bytes. The algorithm MUST be opened using mcrypt_module_open().

.B    int mcrypt_enc_get_key_size( MCRYPT td); 

Returns the maximum supported key size of the algorithm specified by the encryption descriptor
in bytes. The algorithm MUST be opened using mcrypt_module_open().

.B    int* mcrypt_enc_get_supported_key_sizes( MCRYPT td, int* sizes)

Returns the key sizes supported by the algorithm specified by the encryption descriptor. 
If sizes is zero and returns NULL then all key sizes between 1 and mcrypt_get_key_size() are
supported by the algorithm. If it is 1 then only the mcrypt_get_key_size()
size is supported and sizes[0] is equal to it. If it is greater than 1 then
that number specifies the number of elements in sizes which are the key
sizes that the algorithm supports. The returned value is allocated with
malloc, so you should not forget to free it.

.B    int mcrypt_enc_get_iv_size( MCRYPT td); 

Returns size of the IV of the algorithm specified by the encryption descriptor
in bytes. The algorithm MUST be opened using mcrypt_module_open().
If it is '0' then the IV is ignored in that algorithm. IV is used in CBC,
CFB, OFB modes, and in some algorithms in STREAM mode.

.B    int mcrypt_enc_mode_has_iv( MCRYPT td); 

Returns 1 if the mode needs an IV, 0 otherwise. Some 'stream' algorithms
may need an IV even if the mode itself does not need an IV.

.B    char* mcrypt_enc_get_algorithms_name( MCRYPT td);

Returns a character array containing the name of the algorithm.
The returned value is allocated with malloc, so you should not forget to 
free it.

.B    char* mcrypt_enc_get_modes_name( MCRYPT td);

Returns a character array containing the name of the mode.
The returned value is allocated with malloc, so you should not forget to 
free it.

.P
These are some extra functions that operate on modules:
These functions have the prefix mcrypt_module_*.

.B    int mcrypt_module_self_test (char* algorithm, char* directory);

This function runs the self test on the specified algorithm. If the self
test succeeds it returns zero. 

.B    int mcrypt_module_is_block_algorithm_mode( char* algorithm, char* directory);

Returns 1 if the mode is for use with block algorithms, otherwise
it returns 0. (eg. 0 for stream, and 1 for cbc, cfb, ofb)

.B    int mcrypt_module_is_block_algorithm( char* mode, char* directory);

Returns 1 if the algorithm is a block algorithm or 0 if it is a stream
algorithm.

.B    int mcrypt_module_is_block_mode( char* mode, char* directory);

Returns 1 if the mode outputs blocks of bytes or 0 if it outputs bytes.
(eg. 1 for cbc and ecb, and 0 for cfb and stream)

.B    int mcrypt_module_get_algo_block_size( char* algorithm, char* directory);

Returns the block size of the algorithm.

.B    int mcrypt_module_get_algo_key_size( char* algorithm, char* directory);

Returns the maximum supported key size of the algorithm.

.B    int* mcrypt_module_get_algo_supported_key_sizes( char* algorithm, char* directory, int* sizes);

Returns the key sizes supported by the algorithm. If sizes is zero and 
returns NULL then all key sizes between 1 and mcrypt_get_key_size() are
supported by the algorithm. If it is 1 then only the mcrypt_get_key_size()
size is supported and sizes[0] is equal to it. If it is greater than 1 then
that number specifies the number of elements in sizes which are the key
sizes that the algorithm supports. This function differs to
mcrypt_enc_get_supported_key_sizes(), because the return value here is
allocated (not static), thus it should be freed.

.LP

.B char** mcrypt_list_algorithms ( char* libdir, int* size);

Returns a pointer to a character array containing all
the mcrypt algorithms located in the libdir, or if it is NULL, in
the default directory. The size is the number of the character arrays.
The arrays are allocated internally and should be freed by using mcrypt_free_p().

.B char** mcrypt_list_modes ( char* libdir, int *size);

Returns a pointer to a character array containing all
the mcrypt modes located in the libdir, or if it is NULL, in
the default directory. The size is the number of the character arrays.
The arrays should be freed by using mcrypt_free_p().

.B void mcrypt_free_p (char **p, int size);

Frees the pointer to array returned by previous functions.

.B void mcrypt_free (void *ptr);

Frees the memory used by the pointer.

.B    void mcrypt_perror(int err);

This function prints a human readable description of the error 'err' in the stderr.
The err should be a value returned by mcrypt_generic_init().

.B    const char* mcrypt_strerror(int err);

This function returns a human readable description of the error 'err'.
The err should be a value returned by mcrypt_generic_init().

.B    int mcrypt_mutex_register ( void (*mutex_lock)(void) , void (*mutex_unlock)(void) );

This function is only used in multithreaded application and only if compiled
with dynamic module loading support. This is actually used
internally in libltdl. Except for the dynamic module loading libmcrypt is 
thread safe.

.LP
Some example programs follow here. Compile as "cc prog.c -lmcrypt", or
"cc prog.c -lmcrypt -lltdl" depending on your installation.
Libltdl is used for opening dynamic libraries (modules).

.nf
/* First example: Encrypts stdin to stdout using TWOFISH with 128 bit key and CFB */

#include <mcrypt.h>
#include <stdio.h>
#include <stdlib.h>
/* #include <mhash.h> */

main() {

  MCRYPT td;
  int i;
  char *key;
  char password[20];
  char block_buffer;
  char *IV;
  int keysize=16; /* 128 bits */

  key=calloc(1, keysize);
  strcpy(password, "A_large_key");

/* Generate the key using the password */
/*  mhash_keygen( KEYGEN_MCRYPT, MHASH_MD5, key, keysize, NULL, 0, password, strlen(password));
 */
  memmove( key, password, strlen(password));

  td = mcrypt_module_open("twofish", NULL, "cfb", NULL);
  if (td==MCRYPT_FAILED) {
     return 1;
  }
  IV = malloc(mcrypt_enc_get_iv_size(td));

/* Put random data in IV. Note these are not real random data, 
 * consider using /dev/random or /dev/urandom.
 */

  /*  srand(time(0)); */
  for (i=0; i< mcrypt_enc_get_iv_size( td); i++) {
    IV[i]=rand();
  }

  i=mcrypt_generic_init( td, key, keysize, IV);
  if (i<0) {
     mcrypt_perror(i);
     return 1;
  }

  /* Encryption in CFB is performed in bytes */
  while ( fread (&block_buffer, 1, 1, stdin) == 1 ) {
      mcrypt_generic (td, &block_buffer, 1);

/* Comment above and uncomment this to decrypt */
/*    mdecrypt_generic (td, &block_buffer, 1);  */

      fwrite ( &block_buffer, 1, 1, stdout);
  }

/* Deinit the encryption thread, and unload the module */
  mcrypt_generic_end(td);

  return 0;

}


/* Second Example: encrypts using CBC and SAFER+ with 192 bits key */

#include <mcrypt.h>
#include <stdio.h>
#include <stdlib.h>

main() {

  MCRYPT td;
  int i;
  char *key; /* created using mcrypt_gen_key */
  char *block_buffer;
  char *IV;
  int blocksize;
  int keysize = 24; /* 192 bits == 24 bytes */


  key = calloc(1, keysize);
  strcpy(key, "A_large_and_random_key"); 

  td = mcrypt_module_open("saferplus", NULL, "cbc", NULL);

  blocksize = mcrypt_enc_get_block_size(td);
  block_buffer = malloc(blocksize);
/* but unfortunately this does not fill all the key so the rest bytes are
 * padded with zeros. Try to use large keys or convert them with mcrypt_gen_key().
 */

  IV=malloc(mcrypt_enc_get_iv_size(td));

/* Put random data in IV. Note these are not real random data, 
 * consider using /dev/random or /dev/urandom.
 */

/* srand(time(0)); */
  for (i=0; i < mcrypt_enc_get_iv_size(td); i++) {
    IV[i]=rand();
  }

  mcrypt_generic_init( td, key, keysize, IV);

  /* Encryption in CBC is performed in blocks */
  while ( fread (block_buffer, 1, blocksize, stdin) == blocksize ) {
      mcrypt_generic (td, block_buffer, blocksize);
/*      mdecrypt_generic (td, block_buffer, blocksize); */
      fwrite ( block_buffer, 1, blocksize, stdout);
  }

/* deinitialize the encryption thread */
  mcrypt_generic_deinit (td);

/* Unload the loaded module */
  mcrypt_module_close(td);
  return 0;

}
.fi

.LP
The library does not install any signal handler. 
.LP
Questions about libmcrypt should be sent to:
.IP
mcrypt-dev@lists.hellug.gr
or, if this fails, to the author addresses given below.
The mcrypt home page is:
.IP
http://mcrypt.hellug.gr
.LP
.SH AUTHORS
Version 2.4
Copyright (C) 1998-1999 Nikos Mavroyanopoulos (nmav@hellug.gr).
.LP
Thanks to all the people who reported problems and suggested various
improvements for mcrypt; who are too numerous to cite here.
.LP
.\" end of man page