libgcrypt-1.4.6/doc/gcrypt.texi

\input texinfo                  @c -*- Texinfo -*-
@c %**start of header
@setfilename gcrypt.info
@include version.texi
@settitle The Libgcrypt Reference Manual
@c Unify some of the indices.
@syncodeindex tp fn
@syncodeindex pg fn
@c %**end of header
@copying
This manual is for Libgcrypt
(version @value{VERSION}, @value{UPDATED}),
which is GNU's library of cryptographic building blocks.

Copyright @copyright{} 2000, 2002, 2003, 2004, 2006, 2007, 2008, 2009 Free Software Foundation, Inc.

@quotation
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2 of the License, or (at your
option) any later version. The text of the license can be found in the
section entitled ``GNU General Public License''.
@end quotation
@end copying

@dircategory GNU Libraries
@direntry
* libgcrypt: (gcrypt).  Cryptographic function library.
@end direntry



@c
@c Titlepage
@c
@setchapternewpage odd
@titlepage
@title The Libgcrypt Reference Manual
@subtitle Version @value{VERSION}
@subtitle @value{UPDATED}
@author Werner Koch (@email{wk@@gnupg.org})
@author Moritz Schulte (@email{mo@@g10code.com})

@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@ifnothtml
@summarycontents
@contents
@page
@end ifnothtml


@ifnottex
@node Top
@top The Libgcrypt Library
@insertcopying
@end ifnottex


@menu
* Introduction::                 What is Libgcrypt.
* Preparation::                  What you should do before using the library.
* Generalities::                 General library functions and data types.
* Handler Functions::            Working with handler functions.
* Symmetric cryptography::       How to use symmetric cryptography.
* Public Key cryptography::      How to use public key cryptography.
* Hashing::                      How to use hash and MAC algorithms.
* Random Numbers::               How to work with random numbers.
* S-expressions::                How to manage S-expressions.
* MPI library::                  How to work with multi-precision-integers.
* Prime numbers::                How to use the Prime number related functions.
* Utilities::                    Utility functions.
* Architecture::                 How Libgcrypt works internally.

Appendices

* Self-Tests::                  Description of the self-tests.
* FIPS Mode::                   Description of the FIPS mode.
* Library Copying::             The GNU Lesser General Public License
                                says how you can copy and share Libgcrypt.
* Copying::                     The GNU General Public License says how you
                                can copy and share some parts of Libgcrypt.

Indices

* Figures and Tables::          Index of figures and tables.
* Concept Index::               Index of concepts and programs.
* Function and Data Index::     Index of functions, variables and data types.

@end menu

@ifhtml
@page
@summarycontents
@contents
@end ifhtml


@c **********************************************************
@c *******************  Introduction  ***********************
@c **********************************************************
@node Introduction
@chapter Introduction

Libgcrypt is a library providing cryptographic building blocks.

@menu
* Getting Started::             How to use this manual.
* Features::                    A glance at Libgcrypt's features.
* Overview::                    Overview about the library.
@end menu

@node Getting Started
@section Getting Started

This manual documents the Libgcrypt library application programming
interface (API).  All functions and data types provided by the library
are explained.

@noindent
The reader is assumed to possess basic knowledge about applied
cryptography.

This manual can be used in several ways.  If read from the beginning
to the end, it gives a good introduction into the library and how it
can be used in an application.  Forward references are included where
necessary.  Later on, the manual can be used as a reference manual to
get just the information needed about any particular interface of the
library.  Experienced programmers might want to start looking at the
examples at the end of the manual, and then only read up those parts
of the interface which are unclear.


@node Features
@section Features

Libgcrypt might have a couple of advantages over other libraries doing
a similar job.

@table @asis
@item It's Free Software
Anybody can use, modify, and redistribute it under the terms of the GNU
Lesser General Public License (@pxref{Library Copying}).  Note, that
some parts (which are in general not needed by applications) are subject
to the terms of the GNU General Public License (@pxref{Copying}); please
see the README file of the distribution for of list of these parts.

@item It encapsulates the low level cryptography
Libgcrypt provides a high level interface to cryptographic
building blocks using an extensible and flexible API.

@end table

@node Overview
@section Overview

@noindent
The Libgcrypt library is fully thread-safe, where it makes
sense to be thread-safe.  Not thread-safe are some cryptographic
functions that modify a certain context stored in handles.  If the
user really intents to use such functions from different threads on
the same handle, he has to take care of the serialization of such
functions himself.  If not described otherwise, every function is
thread-safe.

Libgcrypt depends on the library `libgpg-error', which
contains common error handling related code for GnuPG components.

@c **********************************************************
@c *******************  Preparation  ************************
@c **********************************************************
@node Preparation
@chapter Preparation

To use Libgcrypt, you have to perform some changes to your
sources and the build system.  The necessary changes are small and
explained in the following sections.  At the end of this chapter, it
is described how the library is initialized, and how the requirements
of the library are verified.

@menu
* Header::                      What header file you need to include.
* Building sources::            How to build sources using the library.
* Building sources using Automake::  How to build sources with the help of Automake.
* Initializing the library::    How to initialize the library.
* Multi-Threading::             How Libgcrypt can be used in a MT environment.
* Enabling FIPS mode::          How to enable the FIPS mode.
@end menu


@node Header
@section Header

All interfaces (data types and functions) of the library are defined
in the header file @file{gcrypt.h}.  You must include this in all source
files using the library, either directly or through some other header
file, like this:

@example
#include <gcrypt.h>
@end example

The name space of Libgcrypt is @code{gcry_*} for function
and type names and @code{GCRY*} for other symbols.  In addition the
same name prefixes with one prepended underscore are reserved for
internal use and should never be used by an application.  Note that
Libgcrypt uses libgpg-error, which uses @code{gpg_*} as
name space for function and type names and @code{GPG_*} for other
symbols, including all the error codes.

@noindent
Certain parts of gcrypt.h may be excluded by defining these macros:

@table @code
@item GCRYPT_NO_MPI_MACROS
Do not define the shorthand macros @code{mpi_*} for @code{gcry_mpi_*}.

@item GCRYPT_NO_DEPRECATED
Do not include defintions for deprecated features.  This is useful to
make sure that no deprecated features are used.
@end table

@node Building sources
@section Building sources

If you want to compile a source file including the `gcrypt.h' header
file, you must make sure that the compiler can find it in the
directory hierarchy.  This is accomplished by adding the path to the
directory in which the header file is located to the compilers include
file search path (via the @option{-I} option).

However, the path to the include file is determined at the time the
source is configured.  To solve this problem, Libgcrypt ships with a small
helper program @command{libgcrypt-config} that knows the path to the
include file and other configuration options.  The options that need
to be added to the compiler invocation at compile time are output by
the @option{--cflags} option to @command{libgcrypt-config}.  The following
example shows how it can be used at the command line:

@example
gcc -c foo.c `libgcrypt-config --cflags`
@end example

Adding the output of @samp{libgcrypt-config --cflags} to the compilers
command line will ensure that the compiler can find the Libgcrypt header
file.

A similar problem occurs when linking the program with the library.
Again, the compiler has to find the library files.  For this to work,
the path to the library files has to be added to the library search path
(via the @option{-L} option).  For this, the option @option{--libs} to
@command{libgcrypt-config} can be used.  For convenience, this option
also outputs all other options that are required to link the program
with the Libgcrypt libraries (in particular, the @samp{-lgcrypt}
option).  The example shows how to link @file{foo.o} with the Libgcrypt
library to a program @command{foo}.

@example
gcc -o foo foo.o `libgcrypt-config --libs`
@end example

Of course you can also combine both examples to a single command by
specifying both options to @command{libgcrypt-config}:

@example
gcc -o foo foo.c `libgcrypt-config --cflags --libs`
@end example

@node Building sources using Automake
@section Building sources using Automake

It is much easier if you use GNU Automake instead of writing your own
Makefiles.  If you do that, you do not have to worry about finding and
invoking the @command{libgcrypt-config} script at all.
Libgcrypt provides an extension to Automake that does all
the work for you.

@c A simple macro for optional variables.
@macro ovar{varname}
@r{[}@var{\varname\}@r{]}
@end macro
@defmac AM_PATH_LIBGCRYPT (@ovar{minimum-version}, @ovar{action-if-found}, @ovar{action-if-not-found})
Check whether Libgcrypt (at least version
@var{minimum-version}, if given) exists on the host system.  If it is
found, execute @var{action-if-found}, otherwise do
@var{action-if-not-found}, if given.

Additionally, the function defines @code{LIBGCRYPT_CFLAGS} to the
flags needed for compilation of the program to find the
@file{gcrypt.h} header file, and @code{LIBGCRYPT_LIBS} to the linker
flags needed to link the program to the Libgcrypt library.
@end defmac

You can use the defined Autoconf variables like this in your
@file{Makefile.am}:

@example
AM_CPPFLAGS = $(LIBGCRYPT_CFLAGS)
LDADD = $(LIBGCRYPT_LIBS)
@end example

@node Initializing the library
@section Initializing the library

Before the library can be used, it must initialize itself.  This is
achieved by invoking the function @code{gcry_check_version} described
below.

Also, it is often desirable to check that the version of
Libgcrypt used is indeed one which fits all requirements.
Even with binary compatibility, new features may have been introduced,
but due to problem with the dynamic linker an old version may actually
be used.  So you may want to check that the version is okay right
after program startup.

@deftypefun {const char *} gcry_check_version (const char *@var{req_version})

The function @code{gcry_check_version} initializes some subsystems used
by Libgcrypt and must be invoked before any other function in the
library, with the exception of the @code{GCRYCTL_SET_THREAD_CBS} command
(called via the @code{gcry_control} function).
@xref{Multi-Threading}.

Furthermore, this function returns the version number of the library.
It can also verify that the version number is higher than a certain
required version number @var{req_version}, if this value is not a null
pointer.
@end deftypefun

Libgcrypt uses a concept known as secure memory, which is a region of
memory set aside for storing sensitive data.  Because such memory is a
scarce resource, it needs to be setup in advanced to a fixed size.
Further, most operating systems have special requirements on how that
secure memory can be used.  For example, it might be required to install
an application as ``setuid(root)'' to allow allocating such memory.
Libgcrypt requires a sequence of initialization steps to make sure that
this works correctly.  The following examples show the necessary steps.

If you don't have a need for secure memory, for example if your
application does not use secret keys or other confidential data or it
runs in a controlled environment where key material floating around in
memory is not a problem, you should initialize Libgcrypt this way:

@example
  /* Version check should be the very first call because it
     makes sure that important subsystems are intialized. */
  if (!gcry_check_version (GCRYPT_VERSION))
    @{
      fputs ("libgcrypt version mismatch\n", stderr);
      exit (2);
    @}
        
  /* Disable secure memory.  */
  gcry_control (GCRYCTL_DISABLE_SECMEM, 0);

  /* ... If required, other initialization goes here.  */

  /* Tell Libgcrypt that initialization has completed. */
  gcry_control (GCRYCTL_INITIALIZATION_FINISHED, 0);
@end example


If you have to protect your keys or other information in memory against
being swapped out to disk and to enable an automatic overwrite of used
and freed memory, you need to initialize Libgcrypt this way:

@example
  /* Version check should be the very first call because it
     makes sure that important subsystems are intialized. */
  if (!gcry_check_version (GCRYPT_VERSION))
    @{
      fputs ("libgcrypt version mismatch\n", stderr);
      exit (2);
    @}

@anchor{sample-use-suspend-secmem}
  /* We don't want to see any warnings, e.g. because we have not yet
     parsed program options which might be used to suppress such
     warnings. */
  gcry_control (GCRYCTL_SUSPEND_SECMEM_WARN);

  /* ... If required, other initialization goes here.  Note that the
     process might still be running with increased privileges and that 
     the secure memory has not been intialized.  */

  /* Allocate a pool of 16k secure memory.  This make the secure memory
     available and also drops privileges where needed.  */
  gcry_control (GCRYCTL_INIT_SECMEM, 16384, 0);

@anchor{sample-use-resume-secmem}
  /* It is now okay to let Libgcrypt complain when there was/is
     a problem with the secure memory. */
  gcry_control (GCRYCTL_RESUME_SECMEM_WARN);

  /* ... If required, other initialization goes here.  */

  /* Tell Libgcrypt that initialization has completed. */
  gcry_control (GCRYCTL_INITIALIZATION_FINISHED, 0);
@end example

It is important that these initialization steps are not done by a
library but by the actual application.  A library using Libgcrypt might
want to check for finished initialization using:

@example
  if (!gcry_control (GCRYCTL_INITIALIZATION_FINISHED_P))
    @{
      fputs ("libgcrypt has not been initialized\n", stderr);
      abort ();
    @}       
@end example

Instead of terminating the process, the library may instead print a
warning and try to initialize Libgcrypt itself.  See also the section on
multi-threading below for more pitfalls.



@node Multi-Threading
@section Multi-Threading

As mentioned earlier, the Libgcrypt library is 
thread-safe if you adhere to the following requirements:

@itemize @bullet
@item
If your application is multi-threaded, you must set the thread support
callbacks with the @code{GCRYCTL_SET_THREAD_CBS} command
@strong{before} any other function in the library.

This is easy enough if you are indeed writing an application using
Libgcrypt.  It is rather problematic if you are writing a library
instead.  Here are some tips what to do if you are writing a library:

If your library requires a certain thread package, just initialize
Libgcrypt to use this thread package.  If your library supports multiple
thread packages, but needs to be configured, you will have to
implement a way to determine which thread package the application
wants to use with your library anyway.  Then configure Libgcrypt to use
this thread package.

If your library is fully reentrant without any special support by a
thread package, then you are lucky indeed.  Unfortunately, this does
not relieve you from doing either of the two above, or use a third
option.  The third option is to let the application initialize Libgcrypt
for you.  Then you are not using Libgcrypt transparently, though.

As if this was not difficult enough, a conflict may arise if two
libraries try to initialize Libgcrypt independently of each others, and
both such libraries are then linked into the same application.  To
make it a bit simpler for you, this will probably work, but only if
both libraries have the same requirement for the thread package.  This
is currently only supported for the non-threaded case, GNU Pth and
pthread.  Support for more thread packages is easy to add, so contact
us if you require it.

@item
The function @code{gcry_check_version} must be called before any other
function in the library, except the @code{GCRYCTL_SET_THREAD_CBS}
command (called via the @code{gcry_control} function), because it
initializes the thread support subsystem in Libgcrypt.  To
achieve this in multi-threaded programs, you must synchronize the
memory with respect to other threads that also want to use
Libgcrypt.  For this, it is sufficient to call
@code{gcry_check_version} before creating the other threads using
Libgcrypt@footnote{At least this is true for POSIX threads,
as @code{pthread_create} is a function that synchronizes memory with
respects to other threads.  There are many functions which have this
property, a complete list can be found in POSIX, IEEE Std 1003.1-2003,
Base Definitions, Issue 6, in the definition of the term ``Memory
Synchronization''.  For other thread packages, more relaxed or more
strict rules may apply.}.

@item
Just like the function @code{gpg_strerror}, the function
@code{gcry_strerror} is not thread safe.  You have to use
@code{gpg_strerror_r} instead.

@end itemize


Libgcrypt contains convenient macros, which define the
necessary thread callbacks for PThread and for GNU Pth:

@table @code
@item GCRY_THREAD_OPTION_PTH_IMPL

This macro defines the following (static) symbols:
@code{gcry_pth_init}, @code{gcry_pth_mutex_init},
@code{gcry_pth_mutex_destroy}, @code{gcry_pth_mutex_lock},
@code{gcry_pth_mutex_unlock}, @code{gcry_pth_read},
@code{gcry_pth_write}, @code{gcry_pth_select},
@code{gcry_pth_waitpid}, @code{gcry_pth_accept},
@code{gcry_pth_connect}, @code{gcry_threads_pth}.

After including this macro, @code{gcry_control()} shall be used with a
command of @code{GCRYCTL_SET_THREAD_CBS} in order to register the
thread callback structure named ``gcry_threads_pth''.

@item GCRY_THREAD_OPTION_PTHREAD_IMPL

This macro defines the following (static) symbols:
@code{gcry_pthread_mutex_init}, @code{gcry_pthread_mutex_destroy},
@code{gcry_pthread_mutex_lock}, @code{gcry_pthread_mutex_unlock},
@code{gcry_threads_pthread}.

After including this macro, @code{gcry_control()} shall be used with a
command of @code{GCRYCTL_SET_THREAD_CBS} in order to register the
thread callback structure named ``gcry_threads_pthread''.
@end table

Note that these macros need to be terminated with a semicolon.  Keep
in mind that these are convenient macros for C programmers; C++
programmers might have to wrap these macros in an ``extern C'' body.


@node Enabling FIPS mode
@section How to enable the FIPS mode
@cindex FIPS mode
@cindex FIPS 140

Libgcrypt may be used in a FIPS 140-2 mode.  Note, that this does not
necessary mean that Libcgrypt is an appoved FIPS 140-2 module.  Check the
NIST database at @url{http://csrc.nist.gov/groups/STM/cmvp/} to see what
versions of Libgcrypt are approved.

Because FIPS 140 has certain restrictions on the use of cryptography
which are not always wanted, Libgcrypt needs to be put into FIPS mode
explicitly.  Three alternative mechanisms are provided to switch
Libgcrypt into this mode:

@itemize
@item 
If the file @file{/proc/sys/crypto/fips_enabled} exists and contains a
numeric value other than @code{0}, Libgcrypt is put into FIPS mode at
initialization time.  Obviously this works only on systems with a
@code{proc} file system (i.e. GNU/Linux).

@item 
If the file @file{/etc/gcrypt/fips_enabled} exists, Libgcrypt is put
into FIPS mode at initialization time.  Note that this filename is
hardwired and does not depend on any configuration options.

@item 
If the application requests FIPS mode using the control command
@code{GCRYCTL_FORCE_FIPS_MODE}.  This must be done prior to any
initialization (i.e. before @code{gcry_check_version}).

@end itemize

@cindex Enforced FIPS mode

In addition to the standard FIPS mode, Libgcrypt may also be put into
an Enforced FIPS mode by writing a non-zero value into the file
@file{/etc/gcrypt/fips_enabled}.  The Enforced FIPS mode helps to
detect applications which don't fulfill all requirements for using
Libgcrypt in FIPS mode (@pxref{FIPS Mode}).

Once Libgcrypt has been put into FIPS mode, it is not possible to
switch back to standard mode without terminating the process first.
If the logging verbosity level of Libgcrypt has been set to at least
2, the state transitions and the self-tests are logged.



@c **********************************************************
@c *******************  General  ****************************
@c **********************************************************
@node Generalities
@chapter Generalities

@menu
* Controlling the library::     Controlling Libgcrypt's behavior.
* Modules::                     Description of extension modules.
* Error Handling::              Error codes and such.
@end menu

@node Controlling the library
@section Controlling the library

@deftypefun gcry_error_t gcry_control (enum gcry_ctl_cmds @var{cmd}, ...)

This function can be used to influence the general behavior of
Libgcrypt in several ways.  Depending on @var{cmd}, more
arguments can or have to be provided.

@table @code
@item GCRYCTL_ENABLE_M_GUARD; Arguments: none
This command enables the built-in memory guard.  It must not be used to
activate the memory guard after the memory management has already been
used; therefore it can ONLY be used at initialization time.  Note that
the memory guard is NOT used when the user of the library has set his
own memory management callbacks.

@item GCRYCTL_ENABLE_QUICK_RANDOM; Arguments: none
This command inhibits the use the very secure random quality level
(@code{GCRY_VERY_STRONG_RANDOM}) and degrades all request down to
@code{GCRY_STRONG_RANDOM}.  In general this is not recommened.  However,
for some applications the extra quality random Libgcrypt tries to create
is not justified and this option may help to get better performace.
Please check with a crypto expert whether this option can be used for
your application.

This option can only be used at initialization time.


@item GCRYCTL_DUMP_RANDOM_STATS; Arguments: none
This command dumps randum number generator related statistics to the
library's logging stream.

@item GCRYCTL_DUMP_MEMORY_STATS; Arguments: none
This command dumps memory managment related statistics to the library's
logging stream.

@item GCRYCTL_DUMP_SECMEM_STATS; Arguments: none
This command dumps secure memory manamgent related statistics to the
library's logging stream.

@item GCRYCTL_DROP_PRIVS; Arguments: none
This command disables the use of secure memory and drops the priviliges
of the current process.  This command has not much use; the suggested way
to disable secure memory is to use @code{GCRYCTL_DISABLE_SECMEM} right
after initialization.

@item GCRYCTL_DISABLE_SECMEM; Arguments: none
This command disables the use of secure memory.  If this command is
used in FIPS mode, FIPS mode will be disabled and the function
@code{gcry_fips_mode_active} returns false.  However, in Enforced FIPS
mode this command has no effect at all.

Many applications do not require secure memory, so they should disable
it right away.  This command should be executed right after
@code{gcry_check_version}. 

@item GCRYCTL_INIT_SECMEM; Arguments: int nbytes
This command is used to allocate a pool of secure memory and thus
enabling the use of secure memory.  It also drops all extra privileges
the process has (i.e. if it is run as setuid (root)).  If the argument
@var{nbytes} is 0, secure memory will be disabled.  The minimum amount
of secure memory allocated is currently 16384 bytes; you may thus use a
value of 1 to request that default size.

@item GCRYCTL_TERM_SECMEM; Arguments: none
This command zeroises the secure memory and destroys the handler.  The
secure memory pool may not be used anymore after running this command.
If the secure memory pool as already been destroyed, this command has
no effect.  Applications might want to run this command from their
exit handler to make sure that the secure memory gets properly
destroyed.  This command is not necessarily thread-safe but that
should not be needed in cleanup code.  It may be called from a signal
handler.

@item GCRYCTL_DISABLE_SECMEM_WARN; Arguments: none
Disable warning messages about problems with the secure memory
subsystem. This command should be run right after
@code{gcry_check_version}.

@item GCRYCTL_SUSPEND_SECMEM_WARN; Arguments: none
Postpone warning messages from the secure memory subsystem. 
@xref{sample-use-suspend-secmem,,the initialization example}, on how to
use it. 

@item GCRYCTL_RESUME_SECMEM_WARN; Arguments: none
Resume warning messages from the secure memory subsystem.
@xref{sample-use-resume-secmem,,the initialization example}, on how to
use it.

@item GCRYCTL_USE_SECURE_RNDPOOL; Arguments: none
This command tells the PRNG to store random numbers in secure memory.
This command should be run right after @code{gcry_check_version} and not
later than the command GCRYCTL_INIT_SECMEM.  Note that in FIPS mode the
secure memory is always used.

@item GCRYCTL_SET_RANDOM_SEED_FILE; Arguments: const char *filename
This command specifies the file, which is to be used as seed file for
the PRNG.  If the seed file is registered prior to initialization of the
PRNG, the seed file's content (if it exists and seems to be valid) is
fed into the PRNG pool.  After the seed file has been registered, the
PRNG can be signalled to write out the PRNG pool's content into the seed
file with the following command.


@item GCRYCTL_UPDATE_RANDOM_SEED_FILE; Arguments: none
Write out the PRNG pool's content into the registered seed file.

Multiple instances of the applications sharing the same random seed file
can be started in parallel, in which case they will read out the same
pool and then race for updating it (the last update overwrites earlier
updates).  They will differentiate only by the weak entropy that is
added in read_seed_file based on the PID and clock, and up to 16 bytes
of weak random non-blockingly.  The consequence is that the output of
these different instances is correlated to some extent.  In a perfect
attack scenario, the attacker can control (or at least guess) the PID
and clock of the application, and drain the system's entropy pool to
reduce the "up to 16 bytes" above to 0.  Then the dependencies of the
inital states of the pools are completely known.  Note that this is not
an issue if random of @code{GCRY_VERY_STRONG_RANDOM} quality is
requested as in this case enough extra entropy gets mixed.  It is also
not an issue when using Linux (rndlinux driver), because this one
guarantees to read full 16 bytes from /dev/urandom and thus there is no
way for an attacker without kernel access to control these 16 bytes.

@item GCRYCTL_SET_VERBOSITY; Arguments: int level
This command sets the verbosity of the logging.  A level of 0 disables
all extra logging whereas positive numbers enable more verbose logging.
The level may be changed at any time but be aware that no memory
synchronization is done so the effect of this command might not
immediately show up in other threads.  This command may even be used
prior to @code{gcry_check_version}.

@item GCRYCTL_SET_DEBUG_FLAGS; Arguments: unsigned int flags
Set the debug flag bits as given by the argument.  Be aware that that no
memory synchronization is done so the effect of this command might not
immediately show up in other threads.  The debug flags are not
considered part of the API and thus may change without notice.  As of
now bit 0 enables debugging of cipher functions and bit 1 debugging of
multi-precision-integers.  This command may even be used prior to
@code{gcry_check_version}.

@item GCRYCTL_CLEAR_DEBUG_FLAGS; Arguments: unsigned int flags
Set the debug flag bits as given by the argument.  Be aware that that no
memory synchronization is done so the effect of this command might not
immediately show up in other threads.  This command may even be used
prior to @code{gcry_check_version}.

@item GCRYCTL_DISABLE_INTERNAL_LOCKING; Arguments: none
This command does nothing.  It exists only for backward compatibility.

@item GCRYCTL_ANY_INITIALIZATION_P; Arguments: none
This command returns true if the library has been basically initialized.
Such a basic initialization happens implicitly with many commands to get
certain internal subsystems running.  The common and suggested way to
do this basic intialization is by calling gcry_check_version.

@item GCRYCTL_INITIALIZATION_FINISHED; Arguments: none
This command tells the libray that the application has finished the
intialization.

@item GCRYCTL_INITIALIZATION_FINISHED_P; Arguments: none
This command returns true if the command@*
GCRYCTL_INITIALIZATION_FINISHED has already been run.

@item GCRYCTL_SET_THREAD_CBS; Arguments: struct ath_ops *ath_ops
This command registers a thread-callback structure.
@xref{Multi-Threading}.

@item GCRYCTL_FAST_POLL; Arguments: none
Run a fast random poll.

@item GCRYCTL_SET_RNDEGD_SOCKET; Arguments: const char *filename
This command may be used to override the default name of the EGD socket
to connect to.  It may be used only during initialization as it is not
thread safe.  Changing the socket name again is not supported.  The
function may return an error if the given filename is too long for a
local socket name.

EGD is an alternative random gatherer, used only on systems lacking a
proper random device.

@item GCRYCTL_PRINT_CONFIG; Arguments: FILE *stream
This command dumps information pertaining to the configuration of the
library to the given stream.  If NULL is given for @var{stream}, the log
system is used.  This command may be used before the intialization has
been finished but not before a gcry_version_check.

@item GCRYCTL_OPERATIONAL_P; Arguments: none
This command returns true if the library is in an operational state.
This information makes only sense in FIPS mode.  In contrast to other
functions, this is a pure test function and won't put the library into
FIPS mode or change the internal state.  This command may be used before
the intialization has been finished but not before a gcry_version_check.

@item GCRYCTL_FIPS_MODE_P; Arguments: none
This command returns true if the library is in FIPS mode.  Note, that
this is no indication about the current state of the library.  This
command may be used before the intialization has been finished but not
before a gcry_version_check.  An application may use this command or
the convenience macro below to check whether FIPS mode is actually
active.

@deftypefun int gcry_fips_mode_active (void)

Returns true if the FIPS mode is active.  Note that this is
implemented as a macro.
@end deftypefun



@item GCRYCTL_FORCE_FIPS_MODE; Arguments: none
Running this command puts the library into FIPS mode.  If the library is
already in FIPS mode, a self-test is triggered and thus the library will
be put into operational state.  This command may be used before a call
to gcry_check_version and that is actually the recommended way to let an
application switch the library into FIPS mode.  Note that Libgcrypt will
reject an attempt to switch to fips mode during or after the intialization.

@item GCRYCTL_SELFTEST; Arguments: none
This may be used at anytime to have the library run all implemented
self-tests.  It works in standard and in FIPS mode.  Returns 0 on
success or an error code on failure.


@end table

@end deftypefun

@node Modules
@section Modules

Libgcrypt supports the use of `extension modules', which
implement algorithms in addition to those already built into the library
directly.

@deftp {Data type} gcry_module_t
This data type represents a `module'.
@end deftp

Functions registering modules provided by the user take a `module
specification structure' as input and return a value of
@code{gcry_module_t} and an ID that is unique in the modules'
category.  This ID can be used to reference the newly registered
module.  After registering a module successfully, the new functionality
should be able to be used through the normal functions provided by
Libgcrypt until it is unregistered again.

@c **********************************************************
@c *******************  Errors  ****************************
@c **********************************************************
@node Error Handling
@section Error Handling

Many functions in Libgcrypt can return an error if they
fail.  For this reason, the application should always catch the error
condition and take appropriate measures, for example by releasing the
resources and passing the error up to the caller, or by displaying a
descriptive message to the user and cancelling the operation.

Some error values do not indicate a system error or an error in the
operation, but the result of an operation that failed properly.  For
example, if you try to decrypt a tempered message, the decryption will
fail.  Another error value actually means that the end of a data
buffer or list has been reached.  The following descriptions explain
for many error codes what they mean usually.  Some error values have
specific meanings if returned by a certain functions.  Such cases are
described in the documentation of those functions.

Libgcrypt uses the @code{libgpg-error} library.  This allows to share
the error codes with other components of the GnuPG system, and to pass
error values transparently from the crypto engine, or some helper
application of the crypto engine, to the user.  This way no
information is lost.  As a consequence, Libgcrypt does not use its own
identifiers for error codes, but uses those provided by
@code{libgpg-error}.  They usually start with @code{GPG_ERR_}.

However, Libgcrypt does provide aliases for the functions
defined in libgpg-error, which might be preferred for name space
consistency.


Most functions in Libgcrypt return an error code in the case
of failure.  For this reason, the application should always catch the
error condition and take appropriate measures, for example by
releasing the resources and passing the error up to the caller, or by
displaying a descriptive message to the user and canceling the
operation.

Some error values do not indicate a system error or an error in the
operation, but the result of an operation that failed properly.

GnuPG components, including Libgcrypt, use an extra library named
libgpg-error to provide a common error handling scheme.  For more
information on libgpg-error, see the according manual.

@menu
* Error Values::                The error value and what it means.
* Error Sources::               A list of important error sources.
* Error Codes::                 A list of important error codes.
* Error Strings::               How to get a descriptive string from a value.
@end menu


@node Error Values
@subsection Error Values
@cindex error values
@cindex error codes
@cindex error sources

@deftp {Data type} {gcry_err_code_t}
The @code{gcry_err_code_t} type is an alias for the
@code{libgpg-error} type @code{gpg_err_code_t}.  The error code
indicates the type of an error, or the reason why an operation failed.

A list of important error codes can be found in the next section.
@end deftp

@deftp {Data type} {gcry_err_source_t}
The @code{gcry_err_source_t} type is an alias for the
@code{libgpg-error} type @code{gpg_err_source_t}.  The error source
has not a precisely defined meaning.  Sometimes it is the place where
the error happened, sometimes it is the place where an error was
encoded into an error value.  Usually the error source will give an
indication to where to look for the problem.  This is not always true,
but it is attempted to achieve this goal.

A list of important error sources can be found in the next section.
@end deftp

@deftp {Data type} {gcry_error_t}
The @code{gcry_error_t} type is an alias for the @code{libgpg-error}
type @code{gpg_error_t}.  An error value like this has always two
components, an error code and an error source.  Both together form the
error value.

Thus, the error value can not be directly compared against an error
code, but the accessor functions described below must be used.
However, it is guaranteed that only 0 is used to indicate success
(@code{GPG_ERR_NO_ERROR}), and that in this case all other parts of
the error value are set to 0, too.

Note that in Libgcrypt, the error source is used purely for
diagnostic purposes.  Only the error code should be checked to test
for a certain outcome of a function.  The manual only documents the
error code part of an error value.  The error source is left
unspecified and might be anything.
@end deftp

@deftypefun {gcry_err_code_t} gcry_err_code (@w{gcry_error_t @var{err}})
The static inline function @code{gcry_err_code} returns the
@code{gcry_err_code_t} component of the error value @var{err}.  This
function must be used to extract the error code from an error value in
order to compare it with the @code{GPG_ERR_*} error code macros.
@end deftypefun

@deftypefun {gcry_err_source_t} gcry_err_source (@w{gcry_error_t @var{err}})
The static inline function @code{gcry_err_source} returns the
@code{gcry_err_source_t} component of the error value @var{err}.  This
function must be used to extract the error source from an error value in
order to compare it with the @code{GPG_ERR_SOURCE_*} error source macros.
@end deftypefun

@deftypefun {gcry_error_t} gcry_err_make (@w{gcry_err_source_t @var{source}}, @w{gcry_err_code_t @var{code}})
The static inline function @code{gcry_err_make} returns the error
value consisting of the error source @var{source} and the error code
@var{code}.

This function can be used in callback functions to construct an error
value to return it to the library.
@end deftypefun

@deftypefun {gcry_error_t} gcry_error (@w{gcry_err_code_t @var{code}})
The static inline function @code{gcry_error} returns the error value
consisting of the default error source and the error code @var{code}.

For @acronym{GCRY} applications, the default error source is
@code{GPG_ERR_SOURCE_USER_1}.  You can define
@code{GCRY_ERR_SOURCE_DEFAULT} before including @file{gcrypt.h} to
change this default.

This function can be used in callback functions to construct an error
value to return it to the library.
@end deftypefun

The @code{libgpg-error} library provides error codes for all system
error numbers it knows about.  If @var{err} is an unknown error
number, the error code @code{GPG_ERR_UNKNOWN_ERRNO} is used.  The
following functions can be used to construct error values from system
errno numbers.

@deftypefun {gcry_error_t} gcry_err_make_from_errno (@w{gcry_err_source_t @var{source}}, @w{int @var{err}})
The function @code{gcry_err_make_from_errno} is like
@code{gcry_err_make}, but it takes a system error like @code{errno}
instead of a @code{gcry_err_code_t} error code.
@end deftypefun

@deftypefun {gcry_error_t} gcry_error_from_errno (@w{int @var{err}})
The function @code{gcry_error_from_errno} is like @code{gcry_error},
but it takes a system error like @code{errno} instead of a
@code{gcry_err_code_t} error code.
@end deftypefun

Sometimes you might want to map system error numbers to error codes
directly, or map an error code representing a system error back to the
system error number.  The following functions can be used to do that.

@deftypefun {gcry_err_code_t} gcry_err_code_from_errno (@w{int @var{err}})
The function @code{gcry_err_code_from_errno} returns the error code
for the system error @var{err}.  If @var{err} is not a known system
error, the function returns @code{GPG_ERR_UNKNOWN_ERRNO}.
@end deftypefun

@deftypefun {int} gcry_err_code_to_errno (@w{gcry_err_code_t @var{err}})
The function @code{gcry_err_code_to_errno} returns the system error
for the error code @var{err}.  If @var{err} is not an error code
representing a system error, or if this system error is not defined on
this system, the function returns @code{0}.
@end deftypefun


@node Error Sources
@subsection Error Sources
@cindex error codes, list of

The library @code{libgpg-error} defines an error source for every
component of the GnuPG system.  The error source part of an error
value is not well defined.  As such it is mainly useful to improve the
diagnostic error message for the user.

If the error code part of an error value is @code{0}, the whole error
value will be @code{0}.  In this case the error source part is of
course @code{GPG_ERR_SOURCE_UNKNOWN}.

The list of error sources that might occur in applications using
@acronym{Libgcrypt} is:

@table @code
@item GPG_ERR_SOURCE_UNKNOWN
The error source is not known.  The value of this error source is
@code{0}.

@item GPG_ERR_SOURCE_GPGME
The error source is @acronym{GPGME} itself.

@item GPG_ERR_SOURCE_GPG
The error source is GnuPG, which is the crypto engine used for the
OpenPGP protocol.

@item GPG_ERR_SOURCE_GPGSM
The error source is GPGSM, which is the crypto engine used for the
OpenPGP protocol.

@item GPG_ERR_SOURCE_GCRYPT
The error source is @code{libgcrypt}, which is used by crypto engines
to perform cryptographic operations.

@item GPG_ERR_SOURCE_GPGAGENT
The error source is @command{gpg-agent}, which is used by crypto
engines to perform operations with the secret key.

@item GPG_ERR_SOURCE_PINENTRY
The error source is @command{pinentry}, which is used by
@command{gpg-agent} to query the passphrase to unlock a secret key.

@item GPG_ERR_SOURCE_SCD
The error source is the SmartCard Daemon, which is used by
@command{gpg-agent} to delegate operations with the secret key to a
SmartCard.

@item GPG_ERR_SOURCE_KEYBOX
The error source is @code{libkbx}, a library used by the crypto
engines to manage local keyrings.

@item GPG_ERR_SOURCE_USER_1
@item GPG_ERR_SOURCE_USER_2
@item GPG_ERR_SOURCE_USER_3
@item GPG_ERR_SOURCE_USER_4
These error sources are not used by any GnuPG component and can be
used by other software.  For example, applications using
Libgcrypt can use them to mark error values coming from callback
handlers.  Thus @code{GPG_ERR_SOURCE_USER_1} is the default for errors
created with @code{gcry_error} and @code{gcry_error_from_errno},
unless you define @code{GCRY_ERR_SOURCE_DEFAULT} before including
@file{gcrypt.h}.
@end table


@node Error Codes
@subsection Error Codes
@cindex error codes, list of

The library @code{libgpg-error} defines many error values.  The
following list includes the most important error codes.

@table @code
@item GPG_ERR_EOF
This value indicates the end of a list, buffer or file.

@item GPG_ERR_NO_ERROR
This value indicates success.  The value of this error code is
@code{0}.  Also, it is guaranteed that an error value made from the
error code @code{0} will be @code{0} itself (as a whole).  This means
that the error source information is lost for this error code,
however, as this error code indicates that no error occurred, this is
generally not a problem.

@item GPG_ERR_GENERAL
This value means that something went wrong, but either there is not
enough information about the problem to return a more useful error
value, or there is no separate error value for this type of problem.

@item GPG_ERR_ENOMEM
This value means that an out-of-memory condition occurred.

@item GPG_ERR_E...
System errors are mapped to GPG_ERR_EFOO where FOO is the symbol for
the system error.

@item GPG_ERR_INV_VALUE
This value means that some user provided data was out of range.

@item GPG_ERR_UNUSABLE_PUBKEY
This value means that some recipients for a message were invalid.

@item GPG_ERR_UNUSABLE_SECKEY
This value means that some signers were invalid.

@item GPG_ERR_NO_DATA
This value means that data was expected where no data was found.

@item GPG_ERR_CONFLICT
This value means that a conflict of some sort occurred.

@item GPG_ERR_NOT_IMPLEMENTED
This value indicates that the specific function (or operation) is not
implemented.  This error should never happen.  It can only occur if
you use certain values or configuration options which do not work,
but for which we think that they should work at some later time.

@item GPG_ERR_DECRYPT_FAILED
This value indicates that a decryption operation was unsuccessful.

@item GPG_ERR_WRONG_KEY_USAGE
This value indicates that a key is not used appropriately.

@item GPG_ERR_NO_SECKEY
This value indicates that no secret key for the user ID is available.

@item GPG_ERR_UNSUPPORTED_ALGORITHM
This value means a verification failed because the cryptographic
algorithm is not supported by the crypto backend.

@item GPG_ERR_BAD_SIGNATURE
This value means a verification failed because the signature is bad.

@item GPG_ERR_NO_PUBKEY
This value means a verification failed because the public key is not
available.

@item GPG_ERR_NOT_OPERATIONAL
This value means that the library is not yet in state which allows to
use this function.  This error code is in particular returned if
Libgcrypt is operated in FIPS mode and the internal state of the
library does not yet or not anymore allow the use of a service.

This error code is only available with newer libgpg-error versions, thus
you might see ``invalid error code'' when passing this to
@code{gpg_strerror}.  The numeric value of this error code is 176.

@item GPG_ERR_USER_1
@item GPG_ERR_USER_2
@item ...
@item GPG_ERR_USER_16
These error codes are not used by any GnuPG component and can be
freely used by other software.  Applications using Libgcrypt
might use them to mark specific errors returned by callback handlers
if no suitable error codes (including the system errors) for these
errors exist already.
@end table


@node Error Strings
@subsection Error Strings
@cindex error values, printing of
@cindex error codes, printing of
@cindex error sources, printing of
@cindex error strings

@deftypefun {const char *} gcry_strerror (@w{gcry_error_t @var{err}})
The function @code{gcry_strerror} returns a pointer to a statically
allocated string containing a description of the error code contained
in the error value @var{err}.  This string can be used to output a
diagnostic message to the user.
@end deftypefun


@deftypefun {const char *} gcry_strsource (@w{gcry_error_t @var{err}})
The function @code{gcry_strerror} returns a pointer to a statically
allocated string containing a description of the error source
contained in the error value @var{err}.  This string can be used to
output a diagnostic message to the user.
@end deftypefun

The following example illustrates the use of the functions described
above:

@example
@{
  gcry_cipher_hd_t handle;
  gcry_error_t err = 0;

  err = gcry_cipher_open (&handle, GCRY_CIPHER_AES, 
                          GCRY_CIPHER_MODE_CBC, 0);
  if (err)
    @{
      fprintf (stderr, "Failure: %s/%s\n",
               gcry_strsource (err),
               gcry_strerror (err));
    @}
@}
@end example

@c **********************************************************
@c *******************  General  ****************************
@c **********************************************************
@node Handler Functions
@chapter Handler Functions

Libgcrypt makes it possible to install so called `handler functions',
which get called by Libgcrypt in case of certain events. 

@menu
* Progress handler::            Using a progress handler function.
* Allocation handler::          Using special memory allocation functions.
* Error handler::               Using error handler functions.
* Logging handler::             Using a special logging function.
@end menu

@node Progress handler
@section Progress handler

It is often useful to retrieve some feedback while long running
operations are performed.

@deftp {Data type} gcry_handler_progress_t
Progress handler functions have to be of the type
@code{gcry_handler_progress_t}, which is defined as:

@code{void (*gcry_handler_progress_t) (void *, const char *, int, int, int)}
@end deftp

The following function may be used to register a handler function for
this purpose.

@deftypefun void gcry_set_progress_handler (gcry_handler_progress_t @var{cb}, void *@var{cb_data})

This function installs @var{cb} as the `Progress handler' function.
It may be used only during initialization.  @var{cb} must be defined
as follows:

@example
void
my_progress_handler (void *@var{cb_data}, const char *@var{what},
                     int @var{printchar}, int @var{current}, int @var{total})
@{
  /* Do something.  */
@}
@end example

A description of the arguments of the progress handler function follows.

@table @var
@item cb_data
The argument provided in the call to @code{gcry_set_progress_handler}.
@item what
A string identifying the type of the progress output.  The following
values for @var{what} are defined:

@table @code
@item need_entropy
Not enough entropy is available.  @var{total} holds the number of
required bytes.

@item primegen
Values for @var{printchar}:
@table @code
@item \n
Prime generated.
@item !
Need to refresh the pool of prime numbers.
@item <, >
Number of bits adjusted.
@item ^
Searching for a generator.
@item .
Fermat test on 10 candidates failed.
@item :
Restart with a new random value.
@item +
Rabin Miller test passed.
@end table

@end table

@end table
@end deftypefun

@node Allocation handler
@section Allocation handler

It is possible to make Libgcrypt use special memory
allocation functions instead of the built-in ones.

Memory allocation functions are of the following types:
@deftp {Data type} gcry_handler_alloc_t
This type is defined as: @code{void *(*gcry_handler_alloc_t) (size_t n)}.
@end deftp
@deftp {Data type} gcry_handler_secure_check_t
This type is defined as: @code{int *(*gcry_handler_secure_check_t) (const void *)}.
@end deftp
@deftp {Data type} gcry_handler_realloc_t
This type is defined as: @code{void *(*gcry_handler_realloc_t) (void *p, size_t n)}.
@end deftp
@deftp {Data type} gcry_handler_free_t
This type is defined as: @code{void *(*gcry_handler_free_t) (void *)}.
@end deftp

Special memory allocation functions can be installed with the
following function:

@deftypefun void gcry_set_allocation_handler (gcry_handler_alloc_t @var{func_alloc}, gcry_handler_alloc_t @var{func_alloc_secure}, gcry_handler_secure_check_t @var{func_secure_check}, gcry_handler_realloc_t @var{func_realloc}, gcry_handler_free_t @var{func_free})
Install the provided functions and use them instead of the built-in
functions for doing memory allocation.  Using this function is in
general not recommended because the standard Libgcrypt allocation
functions are guaranteed to zeroize memory if needed.

This function may be used only during initialization and may not be
used in fips mode.


@end deftypefun

@node Error handler
@section Error handler

The following functions may be used to register handler functions that
are called by Libgcrypt in case certain error conditions occur.  They
may and should be registered prior to calling @code{gcry_check_version}.

@deftp {Data type} gcry_handler_no_mem_t
This type is defined as: @code{int (*gcry_handler_no_mem_t) (void *, size_t, unsigned int)}
@end deftp
@deftypefun void gcry_set_outofcore_handler (gcry_handler_no_mem_t @var{func_no_mem}, void *@var{cb_data})
This function registers @var{func_no_mem} as `out-of-core handler',
which means that it will be called in the case of not having enough
memory available.  The handler is called with 3 arguments: The first
one is the pointer @var{cb_data} as set with this function, the second
is the requested memory size and the last being a flag.  If bit 0 of
the flag is set, secure memory has been requested.  The handler should
either return true to indicate that Libgcrypt should try again
allocating memory or return false to let Libgcrypt use its default
fatal error handler.
@end deftypefun

@deftp {Data type} gcry_handler_error_t
This type is defined as: @code{void (*gcry_handler_error_t) (void *, int, const char *)}
@end deftp

@deftypefun void gcry_set_fatalerror_handler (gcry_handler_error_t @var{func_error}, void *@var{cb_data})
This function registers @var{func_error} as `error handler',
which means that it will be called in error conditions.
@end deftypefun

@node Logging handler
@section Logging handler

@deftp {Data type} gcry_handler_log_t
This type is defined as: @code{void (*gcry_handler_log_t) (void *, int, const char *, va_list)}
@end deftp

@deftypefun void gcry_set_log_handler (gcry_handler_log_t @var{func_log}, void *@var{cb_data})
This function registers @var{func_log} as `logging handler', which means
that it will be called in case Libgcrypt wants to log a message.  This
function may and should be used prior to calling
@code{gcry_check_version}.
@end deftypefun

@c **********************************************************
@c *******************  Ciphers  ****************************
@c **********************************************************
@c @include cipher-ref.texi
@node Symmetric cryptography
@chapter Symmetric cryptography

The cipher functions are used for symmetrical cryptography,
i.e. cryptography using a shared key.  The programming model follows
an open/process/close paradigm and is in that similar to other
building blocks provided by Libgcrypt.

@menu
* Available ciphers::           List of ciphers supported by the library.
* Cipher modules::              How to work with cipher modules.
* Available cipher modes::      List of cipher modes supported by the library.
* Working with cipher handles::  How to perform operations related to cipher handles.
* General cipher functions::    General cipher functions independent of cipher handles.
@end menu

@node Available ciphers
@section Available ciphers

@table @code
@item GCRY_CIPHER_NONE
This is not a real algorithm but used by some functions as error return.
The value always evaluates to false.

@item GCRY_CIPHER_IDEA
@cindex IDEA
This is the IDEA algorithm.  The constant is provided but there is
currently no implementation for it because the algorithm is patented.

@item GCRY_CIPHER_3DES
@cindex 3DES
@cindex Triple-DES
@cindex DES-EDE
@cindex Digital Encryption Standard
Triple-DES with 3 Keys as EDE.  The key size of this algorithm is 168 but
you have to pass 192 bits because the most significant bits of each byte
are ignored.

@item GCRY_CIPHER_CAST5
@cindex CAST5
CAST128-5 block cipher algorithm.  The key size is 128 bits.
	
@item GCRY_CIPHER_BLOWFISH
@cindex Blowfish
The blowfish algorithm. The current implementation allows only for a key
size of 128 bits.

@item GCRY_CIPHER_SAFER_SK128
Reserved and not currently implemented.

@item GCRY_CIPHER_DES_SK	  
Reserved and not currently implemented.
 
@item  GCRY_CIPHER_AES        
@itemx GCRY_CIPHER_AES128
@itemx GCRY_CIPHER_RIJNDAEL
@itemx GCRY_CIPHER_RIJNDAEL128
@cindex Rijndael
@cindex AES
@cindex Advanced Encryption Standard
AES (Rijndael) with a 128 bit key.

@item  GCRY_CIPHER_AES192     
@itemx GCRY_CIPHER_RIJNDAEL192
AES (Rijndael) with a 192 bit key.

@item  GCRY_CIPHER_AES256 
@itemx GCRY_CIPHER_RIJNDAEL256
AES (Rijndael) with a 256 bit key.
    
@item  GCRY_CIPHER_TWOFISH
@cindex Twofish
The Twofish algorithm with a 256 bit key.
    
@item  GCRY_CIPHER_TWOFISH128
The Twofish algorithm with a 128 bit key.
    
@item  GCRY_CIPHER_ARCFOUR   
@cindex Arcfour
@cindex RC4
An algorithm which is 100% compatible with RSA Inc.'s RC4 algorithm.
Note that this is a stream cipher and must be used very carefully to
avoid a couple of weaknesses. 

@item  GCRY_CIPHER_DES
@cindex DES
Standard DES with a 56 bit key. You need to pass 64 bit but the high
bits of each byte are ignored.  Note, that this is a weak algorithm
which can be broken in reasonable time using a brute force approach.

@item  GCRY_CIPHER_SERPENT128
@itemx GCRY_CIPHER_SERPENT192
@itemx GCRY_CIPHER_SERPENT256
@cindex Serpent
The Serpent cipher from the AES contest.

@item  GCRY_CIPHER_RFC2268_40
@itemx GCRY_CIPHER_RFC2268_128
@cindex rfc-2268
@cindex RC2
Ron's Cipher 2 in the 40 and 128 bit variants.  Note, that we currently
only support the 40 bit variant.  The identifier for 128 is reserved for
future use.

@item GCRY_CIPHER_SEED
@cindex Seed (cipher)
A 128 bit cipher as described by RFC4269.

@item  GCRY_CIPHER_CAMELLIA128
@itemx GCRY_CIPHER_CAMELLIA192
@itemx GCRY_CIPHER_CAMELLIA256
@cindex Camellia
The Camellia cipher by NTT.  See
@uref{http://info.isl.ntt.co.jp/@/crypt/@/eng/@/camellia/@/specifications.html}.

@end table

@node Cipher modules
@section Cipher modules

Libgcrypt makes it possible to load additional `cipher modules'; these
ciphers can be used just like the cipher algorithms that are built
into the library directly.  For an introduction into extension
modules, see @xref{Modules}.

@deftp {Data type} gcry_cipher_spec_t
This is the `module specification structure' needed for registering
cipher modules, which has to be filled in by the user before it can be
used to register a module.  It contains the following members:

@table @code
@item const char *name
The primary name of the algorithm.
@item const char **aliases
A list of strings that are `aliases' for the algorithm.  The list must
be terminated with a NULL element.
@item gcry_cipher_oid_spec_t *oids
A list of OIDs that are to be associated with the algorithm.  The
list's last element must have it's `oid' member set to NULL.  See
below for an explanation of this type.
@item size_t blocksize
The block size of the algorithm, in bytes.
@item size_t keylen
The length of the key, in bits.
@item size_t contextsize
The size of the algorithm-specific `context', that should be allocated
for each handle.
@item gcry_cipher_setkey_t setkey
The function responsible for initializing a handle with a provided
key.  See below for a description of this type.
@item gcry_cipher_encrypt_t encrypt
The function responsible for encrypting a single block.  See below for
a description of this type.
@item gcry_cipher_decrypt_t decrypt
The function responsible for decrypting a single block.  See below for
a description of this type.
@item gcry_cipher_stencrypt_t stencrypt
Like `encrypt', for stream ciphers.  See below for a description of
this type.
@item gcry_cipher_stdecrypt_t stdecrypt
Like `decrypt', for stream ciphers.  See below for a description of
this type.
@end table
@end deftp

@deftp {Data type} gcry_cipher_oid_spec_t
This type is used for associating a user-provided algorithm
implementation with certain OIDs.  It contains the following members:
@table @code
@item const char *oid
Textual representation of the OID.
@item int mode
Cipher mode for which this OID is valid.
@end table
@end deftp

@deftp {Data type} gcry_cipher_setkey_t
Type for the `setkey' function, defined as: gcry_err_code_t
(*gcry_cipher_setkey_t) (void *c, const unsigned char *key, unsigned
keylen)
@end deftp

@deftp {Data type} gcry_cipher_encrypt_t
Type for the `encrypt' function, defined as: gcry_err_code_t
(*gcry_cipher_encrypt_t) (void *c, const unsigned char *outbuf, const
unsigned char *inbuf)
@end deftp

@deftp {Data type} gcry_cipher_decrypt_t
Type for the `decrypt' function, defined as: gcry_err_code_t
(*gcry_cipher_decrypt_t) (void *c, const unsigned char *outbuf, const
unsigned char *inbuf)
@end deftp

@deftp {Data type} gcry_cipher_stencrypt_t
Type for the `stencrypt' function, defined as: gcry_err_code_t
(*gcry_@/cipher_@/stencrypt_@/t) (void *c, const unsigned char *outbuf, const
unsigned char *, unsigned int n)
@end deftp

@deftp {Data type} gcry_cipher_stdecrypt_t
Type for the `stdecrypt' function, defined as: gcry_err_code_t
(*gcry_@/cipher_@/stdecrypt_@/t) (void *c, const unsigned char *outbuf, const
unsigned char *, unsigned int n)
@end deftp

@deftypefun gcry_error_t gcry_cipher_register (gcry_cipher_spec_t *@var{cipher}, unsigned int *algorithm_id, gcry_module_t *@var{module})

Register a new cipher module whose specification can be found in
@var{cipher}.  On success, a new algorithm ID is stored in
@var{algorithm_id} and a pointer representing this module is stored
in @var{module}.
@end deftypefun

@deftypefun void gcry_cipher_unregister (gcry_module_t @var{module})
Unregister the cipher identified by @var{module}, which must have been
registered with gcry_cipher_register.
@end deftypefun

@deftypefun gcry_error_t gcry_cipher_list (int *@var{list}, int *@var{list_length})
Get a list consisting of the IDs of the loaded cipher modules.  If
@var{list} is zero, write the number of loaded cipher modules to
@var{list_length} and return.  If @var{list} is non-zero, the first
*@var{list_length} algorithm IDs are stored in @var{list}, which must
be of according size.  In case there are less cipher modules than
*@var{list_length}, *@var{list_length} is updated to the correct
number.
@end deftypefun

@node Available cipher modes
@section Available cipher modes

@table @code
@item GCRY_CIPHER_MODE_NONE
No mode specified.  This should not be used.  The only exception is that
if Libgcrypt is not used in FIPS mode and if any debug flag has been
set, this mode may be used to bypass the actual encryption.

@item GCRY_CIPHER_MODE_ECB
@cindex ECB, Electronic Codebook mode
Electronic Codebook mode.  

@item GCRY_CIPHER_MODE_CFB
@cindex CFB, Cipher Feedback mode
Cipher Feedback mode.  The shift size equals the block size of the
cipher (e.g. for AES it is CFB-128).

@item  GCRY_CIPHER_MODE_CBC
@cindex CBC, Cipher Block Chaining mode
Cipher Block Chaining mode.

@item GCRY_CIPHER_MODE_STREAM
Stream mode, only to be used with stream cipher algorithms.

@item GCRY_CIPHER_MODE_OFB
@cindex OFB, Output Feedback mode
Output Feedback mode.

@item  GCRY_CIPHER_MODE_CTR
@cindex CTR, Counter mode
Counter mode.

@end table

@node Working with cipher handles
@section Working with cipher handles

To use a cipher algorithm, you must first allocate an according
handle.  This is to be done using the open function:

@deftypefun gcry_error_t gcry_cipher_open (gcry_cipher_hd_t *@var{hd}, int @var{algo}, int @var{mode}, unsigned int @var{flags})

This function creates the context handle required for most of the
other cipher functions and returns a handle to it in `hd'.  In case of
an error, an according error code is returned.

The ID of algorithm to use must be specified via @var{algo}.  See
@xref{Available ciphers}, for a list of supported ciphers and the
according constants.

Besides using the constants directly, the function
@code{gcry_cipher_map_name} may be used to convert the textual name of
an algorithm into the according numeric ID.

The cipher mode to use must be specified via @var{mode}.  See
@xref{Available cipher modes}, for a list of supported cipher modes
and the according constants.  Note that some modes are incompatible
with some algorithms - in particular, stream mode
(@code{GCRY_CIPHER_MODE_STREAM}) only works with stream ciphers. Any
block cipher mode (@code{GCRY_CIPHER_MODE_ECB},
@code{GCRY_CIPHER_MODE_CBC}, @code{GCRY_CIPHER_MODE_CFB},
@code{GCRY_CIPHER_MODE_OFB} or @code{GCRY_CIPHER_MODE_CTR}) will work
with any block cipher algorithm.

The third argument @var{flags} can either be passed as @code{0} or as
the bit-wise OR of the following constants.

@table @code
@item GCRY_CIPHER_SECURE
Make sure that all operations are allocated in secure memory.  This is
useful when the key material is highly confidential.
@item GCRY_CIPHER_ENABLE_SYNC
@cindex sync mode (OpenPGP)
This flag enables the CFB sync mode, which is a special feature of
Libgcrypt's CFB mode implementation to allow for OpenPGP's CFB variant. 
See @code{gcry_cipher_sync}.
@item GCRY_CIPHER_CBC_CTS
@cindex cipher text stealing
Enable cipher text stealing (CTS) for the CBC mode.  Cannot be used
simultaneous as GCRY_CIPHER_CBC_MAC.  CTS mode makes it possible to
transform data of almost arbitrary size (only limitation is that it
must be greater than the algorithm's block size).
@item GCRY_CIPHER_CBC_MAC
@cindex CBC-MAC
Compute CBC-MAC keyed checksums.  This is the same as CBC mode, but
only output the last block.  Cannot be used simultaneous as
GCRY_CIPHER_CBC_CTS.
@end table
@end deftypefun 

Use the following function to release an existing handle:

@deftypefun void gcry_cipher_close (gcry_cipher_hd_t @var{h})

This function releases the context created by @code{gcry_cipher_open}.
It also zeroises all sensitive information associated with this cipher
handle.
@end deftypefun

In order to use a handle for performing cryptographic operations, a
`key' has to be set first:

@deftypefun gcry_error_t gcry_cipher_setkey (gcry_cipher_hd_t @var{h}, const void *@var{k}, size_t @var{l})

Set the key @var{k} used for encryption or decryption in the context
denoted by the handle @var{h}.  The length @var{l} (in bytes) of the
key @var{k} must match the required length of the algorithm set for
this context or be in the allowed range for algorithms with variable
key size.  The function checks this and returns an error if there is a
problem.  A caller should always check for an error.

@end deftypefun

Most crypto modes requires an initialization vector (IV), which
usually is a non-secret random string acting as a kind of salt value.
The CTR mode requires a counter, which is also similar to a salt
value.  To set the IV or CTR, use these functions:

@deftypefun gcry_error_t gcry_cipher_setiv (gcry_cipher_hd_t @var{h}, const void *@var{k}, size_t @var{l})

Set the initialization vector used for encryption or decryption. The
vector is passed as the buffer @var{K} of length @var{l} bytes and
copied to internal data structures.  The function checks that the IV
matches the requirement of the selected algorithm and mode.
@end deftypefun

@deftypefun gcry_error_t gcry_cipher_setctr (gcry_cipher_hd_t @var{h}, const void *@var{c}, size_t @var{l})

Set the counter vector used for encryption or decryption. The counter
is passed as the buffer @var{c} of length @var{l} bytes and copied to
internal data structures.  The function checks that the counter
matches the requirement of the selected algorithm (i.e., it must be
the same size as the block size).
@end deftypefun

@deftypefun gcry_error_t gcry_cipher_reset (gcry_cipher_hd_t @var{h})

Set the given handle's context back to the state it had after the last
call to gcry_cipher_setkey and clear the initialization vector.

Note that gcry_cipher_reset is implemented as a macro.
@end deftypefun

The actual encryption and decryption is done by using one of the
following functions.  They may be used as often as required to process
all the data.

@deftypefun gcry_error_t gcry_cipher_encrypt (gcry_cipher_hd_t @var{h}, unsigned char *{out}, size_t @var{outsize}, const unsigned char *@var{in}, size_t @var{inlen})

@code{gcry_cipher_encrypt} is used to encrypt the data.  This function
can either work in place or with two buffers.  It uses the cipher
context already setup and described by the handle @var{h}.  There are 2
ways to use the function: If @var{in} is passed as @code{NULL} and
@var{inlen} is @code{0}, in-place encryption of the data in @var{out} or
length @var{outsize} takes place.  With @var{in} being not @code{NULL},
@var{inlen} bytes are encrypted to the buffer @var{out} which must have
at least a size of @var{inlen}.  @var{outsize} must be set to the
allocated size of @var{out}, so that the function can check that there
is sufficient space. Note that overlapping buffers are not allowed.

Depending on the selected algorithms and encryption mode, the length of
the buffers must be a multiple of the block size.

The function returns @code{0} on success or an error code.
@end deftypefun


@deftypefun gcry_error_t gcry_cipher_decrypt (gcry_cipher_hd_t @var{h}, unsigned char *{out}, size_t @var{outsize}, const unsigned char *@var{in}, size_t @var{inlen})

@code{gcry_cipher_decrypt} is used to decrypt the data.  This function
can either work in place or with two buffers.  It uses the cipher
context already setup and described by the handle @var{h}.  There are 2
ways to use the function: If @var{in} is passed as @code{NULL} and
@var{inlen} is @code{0}, in-place decryption of the data in @var{out} or
length @var{outsize} takes place.  With @var{in} being not @code{NULL},
@var{inlen} bytes are decrypted to the buffer @var{out} which must have
at least a size of @var{inlen}.  @var{outsize} must be set to the
allocated size of @var{out}, so that the function can check that there
is sufficient space.  Note that overlapping buffers are not allowed.

Depending on the selected algorithms and encryption mode, the length of
the buffers must be a multiple of the block size.

The function returns @code{0} on success or an error code.
@end deftypefun


OpenPGP (as defined in RFC-2440) requires a special sync operation in
some places.  The following function is used for this:

@deftypefun gcry_error_t gcry_cipher_sync (gcry_cipher_hd_t @var{h})

Perform the OpenPGP sync operation on context @var{h}.  Note that this
is a no-op unless the context was created with the flag
@code{GCRY_CIPHER_ENABLE_SYNC}
@end deftypefun

Some of the described functions are implemented as macros utilizing a
catch-all control function.  This control function is rarely used
directly but there is nothing which would inhibit it:

@deftypefun gcry_error_t gcry_cipher_ctl (gcry_cipher_hd_t @var{h}, int @var{cmd}, void *@var{buffer}, size_t @var{buflen})

@code{gcry_cipher_ctl} controls various aspects of the cipher module and
specific cipher contexts.  Usually some more specialized functions or
macros are used for this purpose.  The semantics of the function and its
parameters depends on the the command @var{cmd} and the passed context
handle @var{h}.  Please see the comments in the source code
(@code{src/global.c}) for details.
@end deftypefun

@deftypefun gcry_error_t gcry_cipher_info (gcry_cipher_hd_t @var{h}, int @var{what}, void *@var{buffer}, size_t *@var{nbytes})

@code{gcry_cipher_info} is used to retrieve various
information about a cipher context or the cipher module in general.

Currently no information is available.
@end deftypefun

@node General cipher functions
@section General cipher functions

To work with the algorithms, several functions are available to map
algorithm names to the internal identifiers, as well as ways to
retrieve information about an algorithm or the current cipher context.

@deftypefun gcry_error_t gcry_cipher_algo_info (int @var{algo}, int @var{what}, void *@var{buffer}, size_t *@var{nbytes})

This function is used to retrieve information on a specific algorithm.
You pass the cipher algorithm ID as @var{algo} and the type of
information requested as @var{what}. The result is either returned as
the return code of the function or copied to the provided @var{buffer}
whose allocated length must be available in an integer variable with the
address passed in @var{nbytes}.  This variable will also receive the
actual used length of the buffer. 

Here is a list of supported codes for @var{what}:

@c begin constants for gcry_cipher_algo_info
@table @code
@item GCRYCTL_GET_KEYLEN:
Return the length of the key. If the algorithm supports multiple key
lengths, the maximum supported value is returned.  The length is
returned as number of octets (bytes) and not as number of bits in
@var{nbytes}; @var{buffer} must be zero.

@item GCRYCTL_GET_BLKLEN:
Return the block length of the algorithm.  The length is returned as a
number of octets in @var{nbytes}; @var{buffer} must be zero.

@item GCRYCTL_TEST_ALGO:
Returns @code{0} when the specified algorithm is available for use.
@var{buffer} and @var{nbytes} must be zero.
 
@end table  
@c end constants for gcry_cipher_algo_info

@end deftypefun
@c end gcry_cipher_algo_info

@deftypefun {const char *} gcry_cipher_algo_name (int @var{algo})

@code{gcry_cipher_algo_name} returns a string with the name of the
cipher algorithm @var{algo}.  If the algorithm is not known or another
error occurred, the string @code{"?"} is returned.  This function should
not be used to test for the availability of an algorithm.
@end deftypefun

@deftypefun int gcry_cipher_map_name (const char *@var{name})

@code{gcry_cipher_map_name} returns the algorithm identifier for the
cipher algorithm described by the string @var{name}.  If this algorithm
is not available @code{0} is returned.
@end deftypefun

@deftypefun int gcry_cipher_mode_from_oid (const char *@var{string})

Return the cipher mode associated with an @acronym{ASN.1} object
identifier.  The object identifier is expected to be in the
@acronym{IETF}-style dotted decimal notation.  The function returns
@code{0} for an unknown object identifier or when no mode is associated
with it.
@end deftypefun


@c **********************************************************
@c *******************  Public Key  *************************
@c **********************************************************
@node Public Key cryptography
@chapter Public Key cryptography

Public key cryptography, also known as asymmetric cryptography, is an
easy way for key management and to provide digital signatures.
Libgcrypt provides two completely different interfaces to
public key cryptography, this chapter explains the one based on
S-expressions.

@menu
* Available algorithms::        Algorithms supported by the library.
* Used S-expressions::          Introduction into the used S-expression.
* Public key modules::          How to work with public key modules.
* Cryptographic Functions::     Functions for performing the cryptographic actions.
* General public-key related Functions::  General functions, not implementing any cryptography.

* AC Interface::                Alternative interface to public key functions.
@end menu

@node Available algorithms
@section Available algorithms

Libgcrypt supports the RSA (Rivest-Shamir-Adleman) algorithms as well
as DSA (Digital Signature Algorithm) and Elgamal.  The versatile
interface allows to add more algorithms in the future.

@node Used S-expressions
@section Used S-expressions

Libgcrypt's API for asymmetric cryptography is based on data structures
called S-expressions (see
@uref{http://people.csail.mit.edu/@/rivest/@/sexp.html}) and does not work
with contexts as most of the other building blocks of Libgcrypt do.

@noindent
The following information are stored in S-expressions:

@itemize @asis
@item keys

@item plain text data

@item encrypted data

@item signatures

@end itemize

@noindent
To describe how Libgcrypt expect keys, we use examples. Note that
words in
@ifnottex
uppercase
@end ifnottex
@iftex
italics
@end iftex
indicate parameters whereas lowercase words are literals.

Note that all MPI (multi-precision-integers) values are expected to be in
@code{GCRYMPI_FMT_USG} format.  An easy way to create S-expressions is
by using @code{gcry_sexp_build} which allows to pass a string with
printf-like escapes to insert MPI values.

@menu
* RSA key parameters::  Parameters used with an RSA key.
* DSA key parameters::  Parameters used with a DSA key.
* ECC key parameters::  Parameters used with ECC keys.
@end menu

@node RSA key parameters
@subsection RSA key parameters

@noindent
An RSA private key is described by this S-expression:

@example
(private-key
  (rsa
    (n @var{n-mpi})
    (e @var{e-mpi})
    (d @var{d-mpi})
    (p @var{p-mpi})
    (q @var{q-mpi})
    (u @var{u-mpi})))
@end example

@noindent
An RSA public key is described by this S-expression:

@example
(public-key
  (rsa
    (n @var{n-mpi})
    (e @var{e-mpi})))
@end example


@table @var
@item n-mpi
RSA public modulus @math{n}.
@item e-mpi
RSA public exponent @math{e}.
@item d-mpi
RSA secret exponent @math{d = e^{-1} \bmod (p-1)(q-1)}.
@item p-mpi
RSA secret prime @math{p}.
@item q-mpi
RSA secret prime @math{q} with @math{p < q}.
@item u-mpi
Multiplicative inverse @math{u = p^{-1} \bmod q}.
@end table

For signing and decryption the parameters @math{(p, q, u)} are optional
but greatly improve the performance.  Either all of these optional
parameters must be given or none of them.  They are mandatory for
gcry_pk_testkey.

Note that OpenSSL uses slighly different parameters: @math{q < p} and 
 @math{u = q^{-1} \bmod p}.  To use these parameters you will need to
swap the values and recompute @math{u}.  Here is example code to do this:

@example
  if (gcry_mpi_cmp (p, q) > 0)
    @{
      gcry_mpi_swap (p, q);
      gcry_mpi_invm (u, p, q);
    @}
@end example




@node DSA key parameters
@subsection DSA key parameters

@noindent
A DSA private key is described by this S-expression:

@example
(private-key
  (dsa
    (p @var{p-mpi})
    (q @var{q-mpi})
    (g @var{g-mpi})
    (y @var{y-mpi})
    (x @var{x-mpi})))
@end example

@table @var
@item p-mpi
DSA prime @math{p}.
@item q-mpi
DSA group order @math{q} (which is a prime divisor of @math{p-1}).
@item g-mpi
DSA group generator @math{g}.
@item y-mpi
DSA public key value @math{y = g^x \bmod p}.
@item x-mpi
DSA secret exponent x.
@end table

The public key is similar with "private-key" replaced by "public-key"
and no @var{x-mpi}.


@node ECC key parameters
@subsection ECC key parameters

@noindent
An ECC private key is described by this S-expression:

@example
(private-key
  (ecc
    (p @var{p-mpi})
    (a @var{a-mpi})
    (b @var{b-mpi})
    (g @var{g-point})
    (n @var{n-mpi})
    (q @var{q-point})
    (d @var{d-mpi})))
@end example

@table @var
@item p-mpi
Prime specifying the field @math{GF(p)}.
@item a-mpi
@itemx b-mpi
The two coefficients of the Weierstrass equation @math{y^2 = x^3 + ax + b}
@item g-point
Base point @math{g}.
@item n-mpi
Order of @math{g}
@item q-point
The point representing the public key @math{Q = dP}.
@item d-mpi
The private key @math{d}
@end table

All point values are encoded in standard format; Libgcrypt does
currently only support uncompressed points, thus the first byte needs to
be @code{0x04}.

The public key is similar with "private-key" replaced by "public-key"
and no @var{d-mpi}.

If the domain parameters are well-known, the name of this curve may be
used.  For example

@example
(private-key
  (ecc
    (curve "NIST P-192")
    (q @var{q-point})
    (d @var{d-mpi})))
@end example

The @code{curve} parameter may be given in any case and is used to replace
missing parameters.

@noindent
Currently implemented curves are:
@table @code
@item NIST P-192
@itemx 1.2.840.10045.3.1.1
@itemx prime192v1
@itemx secp192r1
The NIST 192 bit curve, its OID, X9.62 and SECP aliases.

@item NIST P-224
@itemx secp224r1
The NIST 224 bit curve and its SECP alias.

@item NIST P-256
@itemx 1.2.840.10045.3.1.7
@itemx prime256v1
@itemx secp256r1
The NIST 256 bit curve, its OID, X9.62 and SECP aliases.

@item NIST P-384
@itemx secp384r1
The NIST 384 bit curve and its SECP alias.

@item NIST P-521
@itemx secp521r1
The NIST 521 bit curve and its SECP alias.

@end table
As usual the OIDs may optionally be prefixed with the string @code{OID.}
or @code{oid.}.



@node Public key modules
@section Public key modules

Libgcrypt makes it possible to load additional `public key
modules'; these public key algorithms can be used just like the
algorithms that are built into the library directly.  For an
introduction into extension modules, see @xref{Modules}.

@deftp {Data type} gcry_pk_spec_t
This is the `module specification structure' needed for registering
public key modules, which has to be filled in by the user before it
can be used to register a module.  It contains the following members:

@table @code
@item const char *name
The primary name of this algorithm.
@item char **aliases
A list of strings that are `aliases' for the algorithm.  The list
must be terminated with a NULL element.
@item const char *elements_pkey
String containing the one-letter names of the MPI values contained in
a public key.
@item const char *element_skey
String containing the one-letter names of the MPI values contained in
a secret key.
@item const char *elements_enc
String containing the one-letter names of the MPI values that are the
result of an encryption operation using this algorithm.
@item const char *elements_sig
String containing the one-letter names of the MPI values that are the
result of a sign operation using this algorithm.
@item const char *elements_grip
String containing the one-letter names of the MPI values that are to
be included in the `key grip'.
@item int use
The bitwise-OR of the following flags, depending on the abilities of
the algorithm:
@table @code
@item GCRY_PK_USAGE_SIGN
The algorithm supports signing and verifying of data.
@item GCRY_PK_USAGE_ENCR
The algorithm supports the encryption and decryption of data.
@end table
@item gcry_pk_generate_t generate
The function responsible for generating a new key pair.  See below for
a description of this type.
@item gcry_pk_check_secret_key_t check_secret_key
The function responsible for checking the sanity of a provided secret
key.  See below for a description of this type.
@item gcry_pk_encrypt_t encrypt
The function responsible for encrypting data.  See below for a
description of this type.
@item gcry_pk_decrypt_t decrypt
The function responsible for decrypting data.  See below for a
description of this type.
@item gcry_pk_sign_t sign
The function responsible for signing data.  See below for a description
of this type.
@item gcry_pk_verify_t verify
The function responsible for verifying that the provided signature
matches the provided data.  See below for a description of this type.
@item gcry_pk_get_nbits_t get_nbits
The function responsible for returning the number of bits of a provided
key.  See below for a description of this type.
@end table
@end deftp

@deftp {Data type} gcry_pk_generate_t
Type for the `generate' function, defined as: gcry_err_code_t
(*gcry_pk_generate_t) (int algo, unsigned int nbits, unsigned long
use_e, gcry_mpi_t *skey, gcry_mpi_t **retfactors)
@end deftp

@deftp {Data type} gcry_pk_check_secret_key_t
Type for the `check_secret_key' function, defined as: gcry_err_code_t
(*gcry_pk_check_secret_key_t) (int algo, gcry_mpi_t *skey)
@end deftp

@deftp {Data type} gcry_pk_encrypt_t
Type for the `encrypt' function, defined as: gcry_err_code_t
(*gcry_pk_encrypt_t) (int algo, gcry_mpi_t *resarr, gcry_mpi_t data,
gcry_mpi_t *pkey, int flags)
@end deftp

@deftp {Data type} gcry_pk_decrypt_t
Type for the `decrypt' function, defined as: gcry_err_code_t
(*gcry_pk_decrypt_t) (int algo, gcry_mpi_t *result, gcry_mpi_t *data,
gcry_mpi_t *skey, int flags)
@end deftp

@deftp {Data type} gcry_pk_sign_t
Type for the `sign' function, defined as: gcry_err_code_t
(*gcry_pk_sign_t) (int algo, gcry_mpi_t *resarr, gcry_mpi_t data,
gcry_mpi_t *skey)
@end deftp

@deftp {Data type} gcry_pk_verify_t
Type for the `verify' function, defined as: gcry_err_code_t
(*gcry_pk_verify_t) (int algo, gcry_mpi_t hash, gcry_mpi_t *data,
gcry_mpi_t *pkey, int (*cmp) (void *, gcry_mpi_t), void *opaquev)
@end deftp

@deftp {Data type} gcry_pk_get_nbits_t
Type for the `get_nbits' function, defined as: unsigned
(*gcry_pk_get_nbits_t) (int algo, gcry_mpi_t *pkey)
@end deftp

@deftypefun gcry_error_t gcry_pk_register (gcry_pk_spec_t *@var{pubkey}, unsigned int *algorithm_id, gcry_module_t *@var{module})

Register a new public key module whose specification can be found in
@var{pubkey}.  On success, a new algorithm ID is stored in
@var{algorithm_id} and a pointer representing this module is stored
in @var{module}.
@end deftypefun

@deftypefun void gcry_pk_unregister (gcry_module_t @var{module})
Unregister the public key module identified by @var{module}, which
must have been registered with gcry_pk_register.
@end deftypefun

@deftypefun gcry_error_t gcry_pk_list (int *@var{list}, int *@var{list_length})
Get a list consisting of the IDs of the loaded pubkey modules.  If
@var{list} is zero, write the number of loaded pubkey modules to
@var{list_length} and return.  If @var{list} is non-zero, the first
*@var{list_length} algorithm IDs are stored in @var{list}, which must
be of according size.  In case there are less pubkey modules than
*@var{list_length}, *@var{list_length} is updated to the correct
number.
@end deftypefun

@node Cryptographic Functions
@section Cryptographic Functions

@noindent
Note that we will in future allow to use keys without p,q and u
specified and may also support other parameters for performance
reasons. 

@noindent

Some functions operating on S-expressions support `flags', that
influence the operation.  These flags have to be listed in a
sub-S-expression named `flags'; the following flags are known:

@table @code
@item pkcs1
Use PKCS#1 block type 2 padding.
@item no-blinding
Do not use a technique called `blinding', which is used by default in
order to prevent leaking of secret information.  Blinding is only
implemented by RSA, but it might be implemented by other algorithms in
the future as well, when necessary.
@end table

@noindent
Now that we know the key basics, we can carry on and explain how to
encrypt and decrypt data.  In almost all cases the data is a random
session key which is in turn used for the actual encryption of the real
data.  There are 2 functions to do this:

@deftypefun gcry_error_t gcry_pk_encrypt (@w{gcry_sexp_t *@var{r_ciph},} @w{gcry_sexp_t @var{data},} @w{gcry_sexp_t @var{pkey}})

Obviously a public key must be provided for encryption.  It is
expected as an appropriate S-expression (see above) in @var{pkey}.
The data to be encrypted can either be in the simple old format, which
is a very simple S-expression consisting only of one MPI, or it may be
a more complex S-expression which also allows to specify flags for
operation, like e.g. padding rules.

@noindent
If you don't want to let Libgcrypt handle the padding, you must pass an
appropriate MPI using this expression for @var{data}:

@example 
(data
  (flags raw)
  (value @var{mpi}))
@end example

@noindent
This has the same semantics as the old style MPI only way.  @var{MPI} is
the actual data, already padded appropriate for your protocol.  Most
systems however use PKCS#1 padding and so you can use this S-expression
for @var{data}:

@example 
(data
  (flags pkcs1)
  (value @var{block}))
@end example

@noindent
Here, the "flags" list has the "pkcs1" flag which let the function know
that it should provide PKCS#1 block type 2 padding.  The actual data to
be encrypted is passed as a string of octets in @var{block}.  The
function checks that this data actually can be used with the given key,
does the padding and encrypts it.

If the function could successfully perform the encryption, the return
value will be 0 and a new S-expression with the encrypted result is
allocated and assigned to the variable at the address of @var{r_ciph}.
The caller is responsible to release this value using
@code{gcry_sexp_release}.  In case of an error, an error code is
returned and @var{r_ciph} will be set to @code{NULL}.

@noindent
The returned S-expression has this format when used with RSA:

@example
(enc-val
  (rsa
    (a @var{a-mpi})))
@end example

@noindent
Where @var{a-mpi} is an MPI with the result of the RSA operation.  When
using the Elgamal algorithm, the return value will have this format:

@example
(enc-val
  (elg
    (a @var{a-mpi})
    (b @var{b-mpi})))
@end example

@noindent
Where @var{a-mpi} and @var{b-mpi} are MPIs with the result of the
Elgamal encryption operation.
@end deftypefun
@c end gcry_pk_encrypt

@deftypefun gcry_error_t gcry_pk_decrypt (@w{gcry_sexp_t *@var{r_plain},} @w{gcry_sexp_t @var{data},} @w{gcry_sexp_t @var{skey}})

Obviously a private key must be provided for decryption.  It is expected
as an appropriate S-expression (see above) in @var{skey}.  The data to
be decrypted must match the format of the result as returned by
@code{gcry_pk_encrypt}, but should be enlarged with a @code{flags}
element:

@example
(enc-val
  (flags)
  (elg
    (a @var{a-mpi})
    (b @var{b-mpi})))
@end example

@noindent
Note that this function currently does not know of any padding
methods and the caller must do any un-padding on his own.

@noindent
The function returns 0 on success or an error code.  The variable at the
address of @var{r_plain} will be set to NULL on error or receive the
decrypted value on success.  The format of @var{r_plain} is a
simple S-expression part (i.e. not a valid one) with just one MPI if
there was no @code{flags} element in @var{data}; if at least an empty
@code{flags} is passed in @var{data}, the format is:

@example
(value @var{plaintext})
@end example
@end deftypefun
@c end gcry_pk_decrypt


Another operation commonly performed using public key cryptography is
signing data.  In some sense this is even more important than
encryption because digital signatures are an important instrument for
key management.  Libgcrypt supports digital signatures using
2 functions, similar to the encryption functions:

@deftypefun gcry_error_t gcry_pk_sign (@w{gcry_sexp_t *@var{r_sig},} @w{gcry_sexp_t @var{data},} @w{gcry_sexp_t @var{skey}})

This function creates a digital signature for @var{data} using the
private key @var{skey} and place it into the variable at the address of
@var{r_sig}.  @var{data} may either be the simple old style S-expression
with just one MPI or a modern and more versatile S-expression which
allows to let Libgcrypt handle padding:

@example 
 (data
  (flags pkcs1)
  (hash @var{hash-algo} @var{block}))
@end example

@noindent
This example requests to sign the data in @var{block} after applying
PKCS#1 block type 1 style padding.  @var{hash-algo} is a string with the
hash algorithm to be encoded into the signature, this may be any hash
algorithm name as supported by Libgcrypt.  Most likely, this will be
"sha256" or "sha1".  It is obvious that the length of @var{block} must
match the size of that message digests; the function checks that this
and other constraints are valid.

@noindent
If PKCS#1 padding is not required (because the caller does already
provide a padded value), either the old format or better the following
format should be used:

@example
(data
  (flags raw)
  (value @var{mpi}))
@end example

@noindent
Here, the data to be signed is directly given as an @var{MPI}.

@noindent
The signature is returned as a newly allocated S-expression in
@var{r_sig} using this format for RSA:

@example
(sig-val
  (rsa
    (s @var{s-mpi})))
@end example

Where @var{s-mpi} is the result of the RSA sign operation.  For DSA the
S-expression returned is:

@example
(sig-val
  (dsa
    (r @var{r-mpi})
    (s @var{s-mpi})))
@end example

Where @var{r-mpi} and @var{s-mpi} are the result of the DSA sign
operation.  For Elgamal signing (which is slow, yields large numbers
and probably is not as secure as the other algorithms), the same format is
used with "elg" replacing "dsa".
@end deftypefun
@c end gcry_pk_sign

@noindent
The operation most commonly used is definitely the verification of a
signature.  Libgcrypt provides this function:

@deftypefun gcry_error_t gcry_pk_verify (@w{gcry_sexp_t @var{sig}}, @w{gcry_sexp_t @var{data}}, @w{gcry_sexp_t @var{pkey}})

This is used to check whether the signature @var{sig} matches the
@var{data}.  The public key @var{pkey} must be provided to perform this
verification.  This function is similar in its parameters to
@code{gcry_pk_sign} with the exceptions that the public key is used
instead of the private key and that no signature is created but a
signature, in a format as created by @code{gcry_pk_sign}, is passed to
the function in @var{sig}.

@noindent
The result is 0 for success (i.e. the data matches the signature), or an
error code where the most relevant code is @code{GCRYERR_BAD_SIGNATURE}
to indicate that the signature does not match the provided data.

@end deftypefun
@c end gcry_pk_verify

@node General public-key related Functions
@section General public-key related Functions

@noindent
A couple of utility functions are available to retrieve the length of
the key, map algorithm identifiers and perform sanity checks:

@deftypefun {const char *} gcry_pk_algo_name (int @var{algo})

Map the public key algorithm id @var{algo} to a string representation of
the algorithm name.  For unknown algorithms this functions returns the
string @code{"?"}.  This function should not be used to test for the
availability of an algorithm.
@end deftypefun

@deftypefun int gcry_pk_map_name (const char *@var{name})

Map the algorithm @var{name} to a public key algorithm Id.  Returns 0 if
the algorithm name is not known.
@end deftypefun

@deftypefun int gcry_pk_test_algo (int @var{algo})

Return 0 if the public key algorithm @var{algo} is available for use.
Note that this is implemented as a macro.
@end deftypefun


@deftypefun {unsigned int} gcry_pk_get_nbits (gcry_sexp_t @var{key})

Return what is commonly referred as the key length for the given
public or private in @var{key}.
@end deftypefun

@deftypefun {unsigned char *} gcry_pk_get_keygrip (@w{gcry_sexp_t @var{key}}, @w{unsigned char *@var{array}})

Return the so called "keygrip" which is the SHA-1 hash of the public key
parameters expressed in a way depended on the algorithm.  @var{array}
must either provide space for 20 bytes or be @code{NULL}. In the latter
case a newly allocated array of that size is returned.  On success a
pointer to the newly allocated space or to @var{array} is returned.
@code{NULL} is returned to indicate an error which is most likely an
unknown algorithm or one where a "keygrip" has not yet been defined.
The function accepts public or secret keys in @var{key}.
@end deftypefun

@deftypefun gcry_error_t gcry_pk_testkey (gcry_sexp_t @var{key})

Return zero if the private key @var{key} is `sane', an error code otherwise.
Note that it is not possible to check the `saneness' of a public key.

@end deftypefun


@deftypefun gcry_error_t gcry_pk_algo_info (@w{int @var{algo}}, @w{int @var{what}}, @w{void *@var{buffer}}, @w{size_t *@var{nbytes}})

Depending on the value of @var{what} return various information about
the public key algorithm with the id @var{algo}.  Note that the
function returns @code{-1} on error and the actual error code must be
retrieved using the function @code{gcry_errno}.  The currently defined
values for @var{what} are:

@table @code
@item GCRYCTL_TEST_ALGO:
Return 0 if the specified algorithm is available for use.
@var{buffer} must be @code{NULL}, @var{nbytes} may be passed as
@code{NULL} or point to a variable with the required usage of the
algorithm. This may be 0 for "don't care" or the bit-wise OR of these
flags:

@table @code
@item GCRY_PK_USAGE_SIGN 
Algorithm is usable for signing.
@item GCRY_PK_USAGE_ENCR 
Algorithm is usable for encryption.
@end table

Unless you need to test for the allowed usage, it is in general better
to use the macro gcry_pk_test_algo instead.

@item GCRYCTL_GET_ALGO_USAGE:
Return the usage flags for the given algorithm.  An invalid algorithm
return 0.  Disabled algorithms are ignored here because we
want to know whether the algorithm is at all capable of a certain usage.

@item GCRYCTL_GET_ALGO_NPKEY
Return the number of elements the public key for algorithm @var{algo}
consist of.  Return 0 for an unknown algorithm.

@item GCRYCTL_GET_ALGO_NSKEY
Return the number of elements the private key for algorithm @var{algo}
consist of.  Note that this value is always larger than that of the
public key.  Return 0 for an unknown algorithm.

@item GCRYCTL_GET_ALGO_NSIGN
Return the number of elements a signature created with the algorithm
@var{algo} consists of.  Return 0 for an unknown algorithm or for an
algorithm not capable of creating signatures.

@item GCRYCTL_GET_ALGO_NENC
Return the number of elements a encrypted message created with the algorithm
@var{algo} consists of.  Return 0 for an unknown algorithm or for an
algorithm not capable of encryption.
@end table

@noindent
Please note that parameters not required should be passed as @code{NULL}.
@end deftypefun
@c end gcry_pk_algo_info


@deftypefun gcry_error_t gcry_pk_ctl (@w{int @var{cmd}}, @w{void *@var{buffer}}, @w{size_t @var{buflen}})

This is a general purpose function to perform certain control
operations.  @var{cmd} controls what is to be done. The return value is
0 for success or an error code.  Currently supported values for
@var{cmd} are:

@table @code
@item GCRYCTL_DISABLE_ALGO
Disable the algorithm given as an algorithm id in @var{buffer}.
@var{buffer} must point to an @code{int} variable with the algorithm id
and @var{buflen} must have the value @code{sizeof (int)}.

@end table
@end deftypefun
@c end gcry_pk_ctl

@noindent
Libgcrypt also provides a function to generate public key
pairs:

@deftypefun gcry_error_t gcry_pk_genkey (@w{gcry_sexp_t *@var{r_key}}, @w{gcry_sexp_t @var{parms}})

This function create a new public key pair using information given in
the S-expression @var{parms} and stores the private and the public key
in one new S-expression at the address given by @var{r_key}.  In case of
an error, @var{r_key} is set to @code{NULL}.  The return code is 0 for
success or an error code otherwise.

@noindent
Here is an example for @var{parms} to create an 2048 bit RSA key:

@example
(genkey
  (rsa
    (nbits 4:2048)))
@end example

@noindent
To create an Elgamal key, substitute "elg" for "rsa" and to create a DSA
key use "dsa".  Valid ranges for the key length depend on the
algorithms; all commonly used key lengths are supported.  Currently
supported parameters are:

@table @code
@item nbits
This is always required to specify the length of the key.  The argument
is a string with a number in C-notation.  The value should be a multiple
of 8.

@item curve @var{name}
For ECC a named curve may be used instead of giving the number of
requested bits.  This allows to request a specific curve to override a
default selection Libgcrypt would have taken if @code{nbits} has been
given.  The available names are listed with the description of the ECC
public key parameters.

@item rsa-use-e
This is only used with RSA to give a hint for the public exponent. The
value will be used as a base to test for a usable exponent. Some values
are special:

@table @samp
@item 0
Use a secure and fast value.  This is currently the number 41.
@item 1
Use a value as required by some crypto policies.  This is currently
the number 65537.
@item 2
Reserved
@item > 2
Use the given value.
@end table

@noindent
If this parameter is not used, Libgcrypt uses for historic reasons
65537.

@item qbits
This is only meanigful for DSA keys.  If it is given the DSA key is
generated with a Q parameyer of this size.  If it is not given or zero 
Q is deduced from NBITS in this way:
@table @samp
@item 512 <= N <= 1024
Q = 160
@item N = 2048
Q = 224
@item N = 3072
Q = 256
@item N = 7680
Q = 384
@item N = 15360
Q = 512
@end table
Note that in this case only the values for N, as given in the table,
are allowed.  When specifying Q all values of N in the range 512 to
15680 are valid as long as they are multiples of 8.

@item transient-key
This is only meaningful for RSA and DSA keys.  This is a flag with no
value.  If given the RSA or DSA key is created using a faster and a
somewhat less secure random number generator.  This flag may be used
for keys which are only used for a short time and do not require full
cryptographic strength.

@item domain
This is only meaningful for DLP algorithms.  If specified keys are
generated with domain parameters taken from this list.  The exact
format of this parameter depends on the actual algorithm.  It is
currently only implemented for DSA using this format:

@example
(genkey
  (dsa
    (domain
      (p @var{p-mpi})
      (q @var{q-mpi})
      (g @var{q-mpi}))))
@end example

@code{nbits} and @code{qbits} may not be specified because they are
derived from the domain parameters.

@item derive-parms
This is currently only implemented for RSA and DSA keys.  It is not
allowed to use this together with a @code{domain} specification.  If
given, it is used to derive the keys using the given parameters.

If given for an RSA key the X9.31 key generation algorithm is used
even if libgcrypt is not in FIPS mode.  If given for a DSA key, the
FIPS 186 algorithm is used even if libgcrypt is not in FIPS mode.

@example
(genkey
  (rsa
    (nbits 4:1024)
    (rsa-use-e 1:3)
    (derive-parms
      (Xp1 #1A1916DDB29B4EB7EB6732E128#)
      (Xp2 #192E8AAC41C576C822D93EA433#)
      (Xp  #D8CD81F035EC57EFE822955149D3BFF70C53520D
            769D6D76646C7A792E16EBD89FE6FC5B605A6493
            39DFC925A86A4C6D150B71B9EEA02D68885F5009
            B98BD984#)
      (Xq1 #1A5CF72EE770DE50CB09ACCEA9#)
      (Xq2 #134E4CAA16D2350A21D775C404#)
      (Xq  #CC1092495D867E64065DEE3E7955F2EBC7D47A2D
            7C9953388F97DDDC3E1CA19C35CA659EDC2FC325
            6D29C2627479C086A699A49C4C9CEE7EF7BD1B34
            321DE34A#))))
@end example

@example
(genkey
  (dsa
    (nbits 4:1024)
    (derive-parms
      (seed @var{seed-mpi}))))
@end example


@item use-x931
@cindex X9.31
Force the use of the ANSI X9.31 key generation algorithm instead of
the default algorithm. This flag is only meaningful for RSA and
usually not required.  Note that this algorithm is implicitly used if
either @code{derive-parms} is given or Libgcrypt is in FIPS mode.

@item use-fips186
@cindex FIPS 186
Force the use of the FIPS 186 key generation algorithm instead of the
default algorithm.  This flag is only meaningful for DSA and usually
not required.  Note that this algorithm is implicitly used if either
@code{derive-parms} is given or Libgcrypt is in FIPS mode.  As of now
FIPS 186-2 is implemented; after the approval of FIPS 186-3 the code
will be changed to implement 186-3.


@item use-fips186-2
Force the use of the FIPS 186-2 key generation algorithm instead of
the default algorithm.  This algorithm is slighlty different from
FIPS 186-3 and allows only 1024 bit keys.  This flag is only meaningful
for DSA and only required for FIPS testing backward compatibility.


@end table
@c end table of parameters

@noindent
The key pair is returned in a format depending on the algorithm.  Both
private and public keys are returned in one container and may be
accompanied by some miscellaneous information.

@noindent
As an example, here is what the Elgamal key generation returns:

@example
(key-data
  (public-key
    (elg
      (p @var{p-mpi})
      (g @var{g-mpi})
      (y @var{y-mpi})))
  (private-key
    (elg
      (p @var{p-mpi})
      (g @var{g-mpi})
      (y @var{y-mpi})
      (x @var{x-mpi})))
  (misc-key-info
    (pm1-factors @var{n1 n2 ... nn}))
@end example

@noindent
As you can see, some of the information is duplicated, but this
provides an easy way to extract either the public or the private key.
Note that the order of the elements is not defined, e.g. the private
key may be stored before the public key. @var{n1 n2 ... nn} is a list
of prime numbers used to composite @var{p-mpi}; this is in general not
a very useful information and only available if the key generation
algorithm provides them.  
@end deftypefun
@c end gcry_pk_genkey

@node AC Interface
@section Alternative Public Key Interface

This section documents the alternative interface to asymmetric
cryptography (ac) that is not based on S-expressions, but on native C
data structures.  As opposed to the pk interface described in the
former chapter, this one follows an open/use/close paradigm like other
building blocks of the library.

@strong{This interface has a few known problems; most noteworthy an
inherent tendency to leak memory.  It might not be available in
forthcoming versions of Libgcrypt.}


@menu
* Available asymmetric algorithms::  List of algorithms supported by the library.
* Working with sets of data::   How to work with sets of data.
* Working with IO objects::     How to work with IO objects.
* Working with handles::        How to use handles.
* Working with keys::           How to work with keys.
* Using cryptographic functions::  How to perform cryptographic operations.
* Handle-independent functions::  General functions independent of handles.
@end menu

@node Available asymmetric algorithms
@subsection Available asymmetric algorithms

Libgcrypt supports the RSA (Rivest-Shamir-Adleman)
algorithms as well as DSA (Digital Signature Algorithm) and Elgamal.
The versatile interface allows to add more algorithms in the future.

@deftp {Data type} gcry_ac_id_t

The following constants are defined for this type:

@table @code
@item GCRY_AC_RSA
Rivest-Shamir-Adleman
@item GCRY_AC_DSA
Digital Signature Algorithm
@item GCRY_AC_ELG
Elgamal
@item GCRY_AC_ELG_E
Elgamal, encryption only.
@end table
@end deftp

@node Working with sets of data
@subsection Working with sets of data

In the context of this interface the term `data set' refers to a list
of `named MPI values' that is used by functions performing
cryptographic operations; a named MPI value is a an MPI value,
associated with a label.

Such data sets are used for representing keys, since keys simply
consist of a variable amount of numbers.  Furthermore some functions
return data sets to the caller that are to be provided to other
functions.

This section documents the data types, symbols and functions that are
relevant for working with data sets.

@deftp {Data type} gcry_ac_data_t
A single data set.
@end deftp

The following flags are supported:

@table @code
@item GCRY_AC_FLAG_DEALLOC
Used for storing data in a data set.  If given, the data will be
released by the library.  Note that whenever one of the ac functions
is about to release objects because of this flag, the objects are
expected to be stored in memory allocated through the Libgcrypt memory
management.  In other words: gcry_free() is used instead of free().

@item GCRY_AC_FLAG_COPY
Used for storing/retrieving data in/from a data set.  If given, the
library will create copies of the provided/contained data, which will
then be given to the user/associated with the data set.
@end table

@deftypefun gcry_error_t gcry_ac_data_new (gcry_ac_data_t *@var{data})
Creates a new, empty data set and stores it in @var{data}.
@end deftypefun

@deftypefun void gcry_ac_data_destroy (gcry_ac_data_t @var{data})
Destroys the data set @var{data}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_set (gcry_ac_data_t @var{data}, unsigned int @var{flags}, char *@var{name}, gcry_mpi_t @var{mpi})
Add the value @var{mpi} to @var{data} with the label @var{name}.  If
@var{flags} contains GCRY_AC_FLAG_COPY, the data set will contain
copies of @var{name} and @var{mpi}.  If @var{flags} contains
GCRY_AC_FLAG_DEALLOC or GCRY_AC_FLAG_COPY, the values
contained in the data set will be deallocated when they are to be
removed from the data set.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_copy (gcry_ac_data_t *@var{data_cp}, gcry_ac_data_t @var{data})
Create a copy of the data set @var{data} and store it in
@var{data_cp}.  FIXME: exact semantics undefined.
@end deftypefun

@deftypefun {unsigned int} gcry_ac_data_length (gcry_ac_data_t @var{data})
Returns the number of named MPI values inside of the data set
@var{data}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_get_name (gcry_ac_data_t @var{data}, unsigned int @var{flags}, char *@var{name}, gcry_mpi_t *@var{mpi})
Store the value labelled with @var{name} found in @var{data} in
@var{mpi}.  If @var{flags} contains GCRY_AC_FLAG_COPY, store a copy of
the @var{mpi} value contained in the data set.  @var{mpi} may be NULL
(this might be useful for checking the existence of an MPI with
extracting it).
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_get_index (gcry_ac_data_t @var{data}, unsigned int flags, unsigned int @var{index}, const char **@var{name}, gcry_mpi_t *@var{mpi})
Stores in @var{name} and @var{mpi} the named @var{mpi} value contained
in the data set @var{data} with the index @var{idx}.  If @var{flags}
contains GCRY_AC_FLAG_COPY, store copies of the values contained in
the data set. @var{name} or @var{mpi} may be NULL.
@end deftypefun

@deftypefun void gcry_ac_data_clear (gcry_ac_data_t @var{data})
Destroys any values contained in the data set @var{data}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_to_sexp (gcry_ac_data_t @var{data}, gcry_sexp_t *@var{sexp}, const char **@var{identifiers})
This function converts the data set @var{data} into a newly created
S-Expression, which is to be stored in @var{sexp}; @var{identifiers}
is a NULL terminated list of C strings, which specifies the structure
of the S-Expression.

Example:

If @var{identifiers} is a list of pointers to the strings ``foo'' and
``bar'' and if @var{data} is a data set containing the values ``val1 =
0x01'' and ``val2 = 0x02'', then the resulting S-Expression will look
like this: (foo (bar ((val1 0x01) (val2 0x02))).
@end deftypefun

@deftypefun gcry_error gcry_ac_data_from_sexp (gcry_ac_data_t *@var{data}, gcry_sexp_t @var{sexp}, const char **@var{identifiers})
This function converts the S-Expression @var{sexp} into a newly
created data set, which is to be stored in @var{data};
@var{identifiers} is a NULL terminated list of C strings, which
specifies the structure of the S-Expression.  If the list of
identifiers does not match the structure of the S-Expression, the
function fails.
@end deftypefun

@node Working with IO objects
@subsection Working with IO objects

Note: IO objects are currently only used in the context of message
encoding/decoding and encryption/signature schemes.

@deftp {Data type} {gcry_ac_io_t}
@code{gcry_ac_io_t} is the type to be used for IO objects.
@end deftp

IO objects provide an uniform IO layer on top of different underlying
IO mechanisms; either they can be used for providing data to the
library (mode is GCRY_AC_IO_READABLE) or they can be used for
retrieving data from the library (mode is GCRY_AC_IO_WRITABLE).

IO object need to be initialized by calling on of the following
functions:

@deftypefun void gcry_ac_io_init (gcry_ac_io_t *@var{ac_io}, gcry_ac_io_mode_t @var{mode}, gcry_ac_io_type_t @var{type}, ...);
Initialize @var{ac_io} according to @var{mode}, @var{type} and the
variable list of arguments.  The list of variable arguments to specify
depends on the given @var{type}.
@end deftypefun

@deftypefun void gcry_ac_io_init_va (gcry_ac_io_t *@var{ac_io}, gcry_ac_io_mode_t @var{mode}, gcry_ac_io_type_t @var{type}, va_list @var{ap});
Initialize @var{ac_io} according to @var{mode}, @var{type} and the
variable list of arguments @var{ap}.  The list of variable arguments
to specify depends on the given @var{type}.
@end deftypefun

The following types of IO objects exist:

@table @code
@item GCRY_AC_IO_STRING
In case of GCRY_AC_IO_READABLE the IO object will provide data from a
memory string.  Arguments to specify at initialization time:
@table @code
@item unsigned char *
Pointer to the beginning of the memory string
@item size_t
Size of the memory string
@end table
In case of GCRY_AC_IO_WRITABLE the object will store retrieved data in
a newly allocated memory string.  Arguments to specify at
initialization time:
@table @code
@item unsigned char **
Pointer to address, at which the pointer to the newly created memory
string is to be stored
@item size_t *
Pointer to address, at which the size of the newly created memory
string is to be stored
@end table

@item GCRY_AC_IO_CALLBACK
In case of GCRY_AC_IO_READABLE the object will forward read requests
to a provided callback function.  Arguments to specify at
initialization time:
@table @code
@item gcry_ac_data_read_cb_t
Callback function to use
@item void *
Opaque argument to provide to the callback function
@end table
In case of GCRY_AC_IO_WRITABLE the object will forward write requests
to a provided callback function.  Arguments to specify at
initialization time:
@table @code
@item gcry_ac_data_write_cb_t
Callback function to use
@item void *
Opaque argument to provide to the callback function
@end table
@end table

@node Working with handles
@subsection Working with handles

In order to use an algorithm, an according handle must be created.
This is done using the following function:

@deftypefun gcry_error_t gcry_ac_open (gcry_ac_handle_t *@var{handle}, int @var{algorithm}, int @var{flags})

Creates a new handle for the algorithm @var{algorithm} and stores it
in @var{handle}.  @var{flags} is not used currently.

@var{algorithm} must be a valid algorithm ID, see @xref{Available
asymmetric algorithms}, for a list of supported algorithms and the
according constants.  Besides using the listed constants directly, the
functions @code{gcry_pk_name_to_id} may be used to convert the textual
name of an algorithm into the according numeric ID.
@end deftypefun

@deftypefun void gcry_ac_close (gcry_ac_handle_t @var{handle})
Destroys the handle @var{handle}.
@end deftypefun

@node Working with keys
@subsection Working with keys

@deftp {Data type} gcry_ac_key_type_t
Defined constants:

@table @code
@item GCRY_AC_KEY_SECRET
Specifies a secret key.
@item GCRY_AC_KEY_PUBLIC
Specifies a public key.
@end table
@end deftp

@deftp {Data type} gcry_ac_key_t
This type represents a single `key', either a secret one or a public
one.
@end deftp

@deftp {Data type} gcry_ac_key_pair_t
This type represents a `key pair' containing a secret and a public key.
@end deftp

Key data structures can be created in two different ways; a new key
pair can be generated, resulting in ready-to-use key.  Alternatively a
key can be initialized from a given data set.

@deftypefun gcry_error_t gcry_ac_key_init (gcry_ac_key_t *@var{key}, gcry_ac_handle_t @var{handle}, gcry_ac_key_type_t @var{type}, gcry_ac_data_t @var{data})
Creates a new key of type @var{type}, consisting of the MPI values
contained in the data set @var{data} and stores it in @var{key}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_key_pair_generate (gcry_ac_handle_t @var{handle}, unsigned int @var{nbits}, void *@var{key_spec}, gcry_ac_key_pair_t *@var{key_pair}, gcry_mpi_t **@var{misc_data})

Generates a new key pair via the handle @var{handle} of @var{NBITS}
bits and stores it in @var{key_pair}.

In case non-standard settings are wanted, a pointer to a structure of
type @code{gcry_ac_key_spec_<algorithm>_t}, matching the selected
algorithm, can be given as @var{key_spec}.  @var{misc_data} is not
used yet.  Such a structure does only exist for RSA.  A description
of the members of the supported structures follows.

@table @code
@item gcry_ac_key_spec_rsa_t
@table @code
@item gcry_mpi_t e
Generate the key pair using a special @code{e}.  The value of @code{e}
has the following meanings:
@table @code
@item = 0
Let Libgcrypt decide what exponent should be used.
@item = 1
Request the use of a ``secure'' exponent; this is required by some
specification to be 65537.
@item > 2
Try starting at this value until a working exponent is found.  Note
that the current implementation leaks some information about the
private key because the incrementation used is not randomized.  Thus,
this function will be changed in the future to return a random
exponent of the given size.
@end table
@end table
@end table

Example code:
@example
@{
  gcry_ac_key_pair_t key_pair;
  gcry_ac_key_spec_rsa_t rsa_spec;

  rsa_spec.e = gcry_mpi_new (0);
  gcry_mpi_set_ui (rsa_spec.e, 1);

  err = gcry_ac_open  (&handle, GCRY_AC_RSA, 0);
  assert (! err);

  err = gcry_ac_key_pair_generate (handle, 1024, &rsa_spec,
                                   &key_pair, NULL);
  assert (! err);
@}
@end example
@end deftypefun


@deftypefun gcry_ac_key_t gcry_ac_key_pair_extract (gcry_ac_key_pair_t @var{key_pair}, gcry_ac_key_type_t @var{which})
Returns the key of type @var{which} out of the key pair
@var{key_pair}.
@end deftypefun

@deftypefun void gcry_ac_key_destroy (gcry_ac_key_t @var{key})
Destroys the key @var{key}.
@end deftypefun

@deftypefun void gcry_ac_key_pair_destroy (gcry_ac_key_pair_t @var{key_pair})
Destroys the key pair @var{key_pair}.
@end deftypefun

@deftypefun gcry_ac_data_t gcry_ac_key_data_get (gcry_ac_key_t @var{key})
Returns the data set contained in the key @var{key}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_key_test (gcry_ac_handle_t @var{handle}, gcry_ac_key_t @var{key})
Verifies that the private key @var{key} is sane via @var{handle}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_key_get_nbits (gcry_ac_handle_t @var{handle}, gcry_ac_key_t @var{key}, unsigned int *@var{nbits})
Stores the number of bits of the key @var{key} in @var{nbits} via @var{handle}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_key_get_grip (gcry_ac_handle_t @var{handle}, gcry_ac_key_t @var{key}, unsigned char *@var{key_grip})
Writes the 20 byte long key grip of the key @var{key} to
@var{key_grip} via @var{handle}.
@end deftypefun

@node Using cryptographic functions
@subsection Using cryptographic functions

The following flags might be relevant:

@table @code
@item GCRY_AC_FLAG_NO_BLINDING
Disable any blinding, which might be supported by the chosen
algorithm; blinding is the default.
@end table

There exist two kinds of cryptographic functions available through the
ac interface: primitives, and high-level functions.

Primitives deal with MPIs (data sets) directly; what they provide is
direct access to the cryptographic operations provided by an algorithm
implementation.

High-level functions deal with octet strings, according to a specified
``scheme''.  Schemes make use of ``encoding methods'', which are
responsible for converting the provided octet strings into MPIs, which
are then forwared to the cryptographic primitives.  Since schemes are
to be used for a special purpose in order to achieve a particular
security goal, there exist ``encryption schemes'' and ``signature
schemes''.  Encoding methods can be used seperately or implicitly
through schemes.

What follows is a description of the cryptographic primitives.

@deftypefun gcry_error_t gcry_ac_data_encrypt (gcry_ac_handle_t @var{handle}, unsigned int @var{flags}, gcry_ac_key_t @var{key}, gcry_mpi_t @var{data_plain}, gcry_ac_data_t *@var{data_encrypted})
Encrypts the plain text MPI value @var{data_plain} with the key public
@var{key} under the control of the flags @var{flags} and stores the
resulting data set into @var{data_encrypted}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_decrypt (gcry_ac_handle_t @var{handle}, unsigned int @var{flags}, gcry_ac_key_t @var{key}, gcry_mpi_t *@var{data_plain}, gcry_ac_data_t @var{data_encrypted})
Decrypts the encrypted data contained in the data set
@var{data_encrypted} with the secret key KEY under the control of the
flags @var{flags} and stores the resulting plain text MPI value in
@var{DATA_PLAIN}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_sign (gcry_ac_handle_t @var{handle}, gcry_ac_key_t @var{key}, gcry_mpi_t @var{data}, gcry_ac_data_t *@var{data_signature})
Signs the data contained in @var{data} with the secret key @var{key}
and stores the resulting signature in the data set
@var{data_signature}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_verify (gcry_ac_handle_t @var{handle}, gcry_ac_key_t @var{key}, gcry_mpi_t @var{data}, gcry_ac_data_t @var{data_signature})
Verifies that the signature contained in the data set
@var{data_signature} is indeed the result of signing the data
contained in @var{data} with the secret key belonging to the public
key @var{key}.
@end deftypefun

What follows is a description of the high-level functions.

The type ``gcry_ac_em_t'' is used for specifying encoding methods; the
following methods are supported:

@table @code
@item GCRY_AC_EME_PKCS_V1_5
PKCS-V1_5 Encoding Method for Encryption.  Options must be provided
through a pointer to a correctly initialized object of type
gcry_ac_eme_pkcs_v1_5_t.

@item GCRY_AC_EMSA_PKCS_V1_5
PKCS-V1_5 Encoding Method for Signatures with Appendix.  Options must
be provided through a pointer to a correctly initialized object of
type gcry_ac_emsa_pkcs_v1_5_t.
@end table

Option structure types:

@table @code
@item gcry_ac_eme_pkcs_v1_5_t
@table @code
@item gcry_ac_key_t key
@item gcry_ac_handle_t handle
@end table
@item gcry_ac_emsa_pkcs_v1_5_t
@table @code
@item gcry_md_algo_t md
@item size_t em_n
@end table
@end table

Encoding methods can be used directly through the following functions:

@deftypefun gcry_error_t gcry_ac_data_encode (gcry_ac_em_t @var{method}, unsigned int @var{flags}, void *@var{options}, unsigned char *@var{m}, size_t @var{m_n}, unsigned char **@var{em}, size_t *@var{em_n})
Encodes the message contained in @var{m} of size @var{m_n} according
to @var{method}, @var{flags} and @var{options}.  The newly created
encoded message is stored in @var{em} and @var{em_n}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_decode (gcry_ac_em_t @var{method}, unsigned int @var{flags}, void *@var{options}, unsigned char *@var{em}, size_t @var{em_n}, unsigned char **@var{m}, size_t *@var{m_n})
Decodes the message contained in @var{em} of size @var{em_n} according
to @var{method}, @var{flags} and @var{options}.  The newly created
decoded message is stored in @var{m} and @var{m_n}.
@end deftypefun

The type ``gcry_ac_scheme_t'' is used for specifying schemes; the
following schemes are supported:

@table @code
@item GCRY_AC_ES_PKCS_V1_5
PKCS-V1_5 Encryption Scheme.  No options can be provided.
@item GCRY_AC_SSA_PKCS_V1_5
PKCS-V1_5 Signature Scheme (with Appendix).  Options can be provided
through a pointer to a correctly initialized object of type
gcry_ac_ssa_pkcs_v1_5_t.
@end table

Option structure types:

@table @code
@item gcry_ac_ssa_pkcs_v1_5_t
@table @code
@item gcry_md_algo_t md
@end table
@end table

The functions implementing schemes:

@deftypefun gcry_error_t gcry_ac_data_encrypt_scheme (gcry_ac_handle_t @var{handle}, gcry_ac_scheme_t @var{scheme}, unsigned int @var{flags}, void *@var{opts}, gcry_ac_key_t @var{key}, gcry_ac_io_t *@var{io_message}, gcry_ac_io_t *@var{io_cipher})
Encrypts the plain text readable from @var{io_message} through
@var{handle} with the public key @var{key} according to @var{scheme},
@var{flags} and @var{opts}.  If @var{opts} is not NULL, it has to be a
pointer to a structure specific to the chosen scheme (gcry_ac_es_*_t).
The encrypted message is written to @var{io_cipher}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_decrypt_scheme (gcry_ac_handle_t @var{handle}, gcry_ac_scheme_t @var{scheme}, unsigned int @var{flags}, void *@var{opts}, gcry_ac_key_t @var{key}, gcry_ac_io_t *@var{io_cipher}, gcry_ac_io_t *@var{io_message})
Decrypts the cipher text readable from @var{io_cipher} through
@var{handle} with the secret key @var{key} according to @var{scheme},
@var{flags} and @var{opts}.  If @var{opts} is not NULL, it has to be a
pointer to a structure specific to the chosen scheme (gcry_ac_es_*_t).
The decrypted message is written to @var{io_message}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_sign_scheme (gcry_ac_handle_t @var{handle}, gcry_ac_scheme_t @var{scheme}, unsigned int @var{flags}, void *@var{opts}, gcry_ac_key_t @var{key}, gcry_ac_io_t *@var{io_message}, gcry_ac_io_t *@var{io_signature})
Signs the message readable from @var{io_message} through @var{handle}
with the secret key @var{key} according to @var{scheme}, @var{flags}
and @var{opts}.  If @var{opts} is not NULL, it has to be a pointer to
a structure specific to the chosen scheme (gcry_ac_ssa_*_t).  The
signature is written to @var{io_signature}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_data_verify_scheme (gcry_ac_handle_t @var{handle}, gcry_ac_scheme_t @var{scheme}, unsigned int @var{flags}, void *@var{opts}, gcry_ac_key_t @var{key}, gcry_ac_io_t *@var{io_message}, gcry_ac_io_t *@var{io_signature})
Verifies through @var{handle} that the signature readable from
@var{io_signature} is indeed the result of signing the message
readable from @var{io_message} with the secret key belonging to the
public key @var{key} according to @var{scheme} and @var{opts}.  If
@var{opts} is not NULL, it has to be an anonymous structure
(gcry_ac_ssa_*_t) specific to the chosen scheme.
@end deftypefun

@node Handle-independent functions
@subsection Handle-independent functions

These two functions are deprecated; do not use them for new code.

@deftypefun gcry_error_t gcry_ac_id_to_name (gcry_ac_id_t @var{algorithm}, const char **@var{name})
Stores the textual representation of the algorithm whose id is given
in @var{algorithm} in @var{name}.  Deprecated; use @code{gcry_pk_algo_name}.
@end deftypefun

@deftypefun gcry_error_t gcry_ac_name_to_id (const char *@var{name}, gcry_ac_id_t *@var{algorithm})
Stores the numeric ID of the algorithm whose textual representation is
contained in @var{name} in @var{algorithm}. Deprecated; use
@code{gcry_pk_map_name}.
@end deftypefun

@c **********************************************************
@c *******************  Hash Functions  *********************
@c **********************************************************
@node Hashing
@chapter Hashing

Libgcrypt provides an easy and consistent to use interface for hashing.
Hashing is buffered and several hash algorithms can be updated at once.
It is possible to compute a MAC using the same routines.  The
programming model follows an open/process/close paradigm and is in that
similar to other building blocks provided by Libgcrypt.

For convenience reasons, a few cyclic redundancy check value operations
are also supported.

@menu
* Available hash algorithms::   List of hash algorithms supported by the library.
* Hash algorithm modules::      How to work with hash algorithm modules.
* Working with hash algorithms::  List of functions related to hashing.
@end menu

@node Available hash algorithms
@section Available hash algorithms

@c begin table of hash algorithms
@cindex SHA-1
@cindex SHA-224, SHA-256, SHA-384, SHA-512
@cindex RIPE-MD-160
@cindex MD2, MD4, MD5
@cindex TIGER
@cindex HAVAL
@cindex Whirlpool
@cindex CRC32
@table @code
@item GCRY_MD_NONE
This is not a real algorithm but used by some functions as an error
return value.  This constant is guaranteed to have the value @code{0}.

@item GCRY_MD_SHA1
This is the SHA-1 algorithm which yields a message digest of 20 bytes.
Note that SHA-1 begins to show some weaknesses and it is suggested to
fade out its use if strong cryptographic properties are required.

@item GCRY_MD_RMD160
This is the 160 bit version of the RIPE message digest (RIPE-MD-160).
Like SHA-1 it also yields a digest of 20 bytes.  This algorithm share a
lot of design properties with SHA-1 and thus it is advisable not to use
it for new protocols.

@item GCRY_MD_MD5
This is the well known MD5 algorithm, which yields a message digest of
16 bytes.  Note that the MD5 algorithm has severe weaknesses, for
example it is easy to compute two messages yielding the same hash
(collision attack).  The use of this algorithm is only justified for
non-cryptographic application.


@item GCRY_MD_MD4
This is the MD4 algorithm, which yields a message digest of 16 bytes.
This algorithms ha severe weaknesses and should not be used.

@item GCRY_MD_MD2
This is an reserved identifier for MD-2; there is no implementation yet.
This algorithm has severe weaknesses and should not be used.

@item GCRY_MD_TIGER
This is the TIGER/192 algorithm which yields a message digest of 24 bytes.

@item GCRY_MD_HAVAL
This is an reserved value for the HAVAL algorithm with 5 passes and 160
bit. It yields a message digest of 20 bytes.  Note that there is no
implementation yet available.

@item GCRY_MD_SHA224
This is the SHA-224 algorithm which yields a message digest of 28 bytes.
See Change Notice 1 for FIPS 180-2 for the specification.

@item GCRY_MD_SHA256
This is the SHA-256 algorithm which yields a message digest of 32 bytes.
See FIPS 180-2 for the specification.

@item GCRY_MD_SHA384
This is the SHA-384 algorithm which yields a message digest of 48 bytes.
See FIPS 180-2 for the specification.

@item GCRY_MD_SHA512
This is the SHA-384 algorithm which yields a message digest of 64 bytes.
See FIPS 180-2 for the specification.

@item GCRY_MD_CRC32
This is the ISO 3309 and ITU-T V.42 cyclic redundancy check.  It yields
an output of 4 bytes.  Note that this is not a hash algorithm in the
cryptographic sense.

@item GCRY_MD_CRC32_RFC1510
This is the above cyclic redundancy check function, as modified by RFC
1510.  It yields an output of 4 bytes.  Note that this is not a hash
algorithm in the cryptographic sense.

@item GCRY_MD_CRC24_RFC2440
This is the OpenPGP cyclic redundancy check function.  It yields an
output of 3 bytes.  Note that this is not a hash algorithm in the
cryptographic sense.

@item GCRY_MD_WHIRLPOOL
This is the Whirlpool algorithm which yields a message digest of 64
bytes.

@end table
@c end table of hash algorithms

@node Hash algorithm modules
@section Hash algorithm modules

Libgcrypt makes it possible to load additional `message
digest modules'; these digests can be used just like the message digest
algorithms that are built into the library directly.  For an
introduction into extension modules, see @xref{Modules}.

@deftp {Data type} gcry_md_spec_t
This is the `module specification structure' needed for registering
message digest modules, which has to be filled in by the user before
it can be used to register a module.  It contains the following
members:

@table @code
@item const char *name
The primary name of this algorithm.
@item unsigned char *asnoid
Array of bytes that form the ASN OID.
@item int asnlen
Length of bytes in `asnoid'.
@item gcry_md_oid_spec_t *oids
A list of OIDs that are to be associated with the algorithm.  The
list's last element must have it's `oid' member set to NULL.  See
below for an explanation of this type.  See below for an explanation
of this type.
@item int mdlen
Length of the message digest algorithm.  See below for an explanation
of this type.
@item gcry_md_init_t init
The function responsible for initializing a handle.  See below for an
explanation of this type.
@item gcry_md_write_t write
The function responsible for writing data into a message digest
context.  See below for an explanation of this type.
@item gcry_md_final_t final
The function responsible for `finalizing' a message digest context.
See below for an explanation of this type.
@item gcry_md_read_t read
The function responsible for reading out a message digest result.  See
below for an explanation of this type.
@item size_t contextsize
The size of the algorithm-specific `context', that should be
allocated for each handle.
@end table
@end deftp

@deftp {Data type} gcry_md_oid_spec_t
This type is used for associating a user-provided algorithm
implementation with certain OIDs.  It contains the following members:

@table @code
@item const char *oidstring
Textual representation of the OID.
@end table
@end deftp

@deftp {Data type} gcry_md_init_t
Type for the `init' function, defined as: void (*gcry_md_init_t) (void
*c)
@end deftp

@deftp {Data type} gcry_md_write_t
Type for the `write' function, defined as: void (*gcry_md_write_t)
(void *c, unsigned char *buf, size_t nbytes)
@end deftp

@deftp {Data type} gcry_md_final_t
Type for the `final' function, defined as: void (*gcry_md_final_t)
(void *c)
@end deftp

@deftp {Data type} gcry_md_read_t
Type for the `read' function, defined as: unsigned char
*(*gcry_md_read_t) (void *c)
@end deftp

@deftypefun gcry_error_t gcry_md_register (gcry_md_spec_t *@var{digest}, unsigned int *algorithm_id, gcry_module_t *@var{module})

Register a new digest module whose specification can be found in
@var{digest}.  On success, a new algorithm ID is stored in
@var{algorithm_id} and a pointer representing this module is stored
in @var{module}.
@end deftypefun

@deftypefun void gcry_md_unregister (gcry_module_t @var{module})
Unregister the digest identified by @var{module}, which must have been
registered with gcry_md_register.
@end deftypefun

@deftypefun gcry_error_t gcry_md_list (int *@var{list}, int *@var{list_length})
Get a list consisting of the IDs of the loaded message digest modules.
If @var{list} is zero, write the number of loaded message digest
modules to @var{list_length} and return.  If @var{list} is non-zero,
the first *@var{list_length} algorithm IDs are stored in @var{list},
which must be of according size.  In case there are less message
digests modules than *@var{list_length}, *@var{list_length} is updated
to the correct number.
@end deftypefun

@node Working with hash algorithms
@section Working with hash algorithms

To use most of these function it is necessary to create a context;
this is done using:

@deftypefun gcry_error_t gcry_md_open (gcry_md_hd_t *@var{hd}, int @var{algo}, unsigned int @var{flags})

Create a message digest object for algorithm @var{algo}.  @var{flags}
may be given as an bitwise OR of constants described below.  @var{algo}
may be given as @code{0} if the algorithms to use are later set using
@code{gcry_md_enable}. @var{hd} is guaranteed to either receive a valid
handle or NULL.

For a list of supported algorithms, see @xref{Available hash
algorithms}.

The flags allowed for @var{mode} are:

@c begin table of hash flags
@table @code
@item GCRY_MD_FLAG_SECURE
Allocate all buffers and the resulting digest in "secure memory".  Use
this is the hashed data is highly confidential.

@item GCRY_MD_FLAG_HMAC
@cindex HMAC
Turn the algorithm into a HMAC message authentication algorithm.  This
only works if just one algorithm is enabled for the handle.  Note that
the function @code{gcry_md_setkey} must be used to set the MAC key.
The size of the MAC is equal to the message digest of the underlying
hash algorithm.  If you want CBC message authentication codes based on
a cipher, see @xref{Working with cipher handles}.

@end table
@c begin table of hash flags

You may use the function @code{gcry_md_is_enabled} to later check
whether an algorithm has been enabled.

@end deftypefun
@c end function gcry_md_open

If you want to calculate several hash algorithms at the same time, you
have to use the following function right after the @code{gcry_md_open}:

@deftypefun gcry_error_t gcry_md_enable (gcry_md_hd_t @var{h}, int @var{algo})

Add the message digest algorithm @var{algo} to the digest object
described by handle @var{h}.  Duplicated enabling of algorithms is
detected and ignored.
@end deftypefun

If the flag @code{GCRY_MD_FLAG_HMAC} was used, the key for the MAC must
be set using the function:

@deftypefun gcry_error_t gcry_md_setkey (gcry_md_hd_t @var{h}, const void *@var{key}, size_t @var{keylen})

For use with the HMAC feature, set the MAC key to the value of
@var{key} of length @var{keylen} bytes.  There is no restriction on
the length of the key.
@end deftypefun


After you are done with the hash calculation, you should release the
resources by using:

@deftypefun void gcry_md_close (gcry_md_hd_t @var{h})

Release all resources of hash context @var{h}.  @var{h} should not be
used after a call to this function.  A @code{NULL} passed as @var{h} is
ignored.  The function also zeroises all sensitive information
associated with this handle.


@end deftypefun

Often you have to do several hash operations using the same algorithm.
To avoid the overhead of creating and releasing context, a reset function
is provided:

@deftypefun void gcry_md_reset (gcry_md_hd_t @var{h})

Reset the current context to its initial state.  This is effectively
identical to a close followed by an open and enabling all currently
active algorithms.
@end deftypefun


Often it is necessary to start hashing some data and then continue to
hash different data.  To avoid hashing the same data several times (which
might not even be possible if the data is received from a pipe), a
snapshot of the current hash context can be taken and turned into a new
context:

@deftypefun gcry_error_t gcry_md_copy (gcry_md_hd_t *@var{handle_dst}, gcry_md_hd_t @var{handle_src})

Create a new digest object as an exact copy of the object described by
handle @var{handle_src} and store it in @var{handle_dst}.  The context
is not reset and you can continue to hash data using this context and
independently using the original context.
@end deftypefun


Now that we have prepared everything to calculate hashes, it is time to
see how it is actually done.  There are two ways for this, one to
update the hash with a block of memory and one macro to update the hash
by just one character.  Both methods can be used on the same hash context.

@deftypefun void gcry_md_write (gcry_md_hd_t @var{h}, const void *@var{buffer}, size_t @var{length})

Pass @var{length} bytes of the data in @var{buffer} to the digest object
with handle @var{h} to update the digest values. This
function should be used for large blocks of data.
@end deftypefun

@deftypefun void gcry_md_putc (gcry_md_hd_t @var{h}, int @var{c})

Pass the byte in @var{c} to the digest object with handle @var{h} to
update the digest value.  This is an efficient function, implemented as
a macro to buffer the data before an actual update. 
@end deftypefun

The semantics of the hash functions do not provide for reading out intermediate
message digests because the calculation must be finalized first.  This
finalization may for example include the number of bytes hashed in the
message digest or some padding.

@deftypefun void gcry_md_final (gcry_md_hd_t @var{h})

Finalize the message digest calculation.  This is not really needed
because @code{gcry_md_read} does this implicitly.  After this has been
done no further updates (by means of @code{gcry_md_write} or
@code{gcry_md_putc} are allowed.  Only the first call to this function
has an effect. It is implemented as a macro.
@end deftypefun

The way to read out the calculated message digest is by using the
function:

@deftypefun {unsigned char *} gcry_md_read (gcry_md_hd_t @var{h}, int @var{algo})

@code{gcry_md_read} returns the message digest after finalizing the
calculation.  This function may be used as often as required but it will
always return the same value for one handle.  The returned message digest
is allocated within the message context and therefore valid until the
handle is released or reseted (using @code{gcry_md_close} or
@code{gcry_md_reset}.  @var{algo} may be given as 0 to return the only
enabled message digest or it may specify one of the enabled algorithms.
The function does return @code{NULL} if the requested algorithm has not
been enabled.
@end deftypefun

Because it is often necessary to get the message digest of one block of
memory, a fast convenience function is available for this task: 

@deftypefun void gcry_md_hash_buffer (int @var{algo}, void *@var{digest}, const void *@var{buffer}, size_t @var{length});

@code{gcry_md_hash_buffer} is a shortcut function to calculate a message
digest of a buffer.  This function does not require a context and
immediately returns the message digest of the @var{length} bytes at
@var{buffer}.  @var{digest} must be allocated by the caller, large
enough to hold the message digest yielded by the the specified algorithm
@var{algo}.  This required size may be obtained by using the function
@code{gcry_md_get_algo_dlen}.

Note that this function will abort the process if an unavailable
algorithm is used.
@end deftypefun

@c ***********************************
@c ***** MD info functions ***********
@c ***********************************

Hash algorithms are identified by internal algorithm numbers (see
@code{gcry_md_open} for a list).  However, in most applications they are
used by names, so two functions are available to map between string
representations and hash algorithm identifiers.

@deftypefun {const char *} gcry_md_algo_name (int @var{algo})

Map the digest algorithm id @var{algo} to a string representation of the
algorithm name.  For unknown algorithms this function returns the
string @code{"?"}.  This function should not be used to test for the
availability of an algorithm.
@end deftypefun

@deftypefun int gcry_md_map_name (const char *@var{name})

Map the algorithm with @var{name} to a digest algorithm identifier.
Returns 0 if the algorithm name is not known.  Names representing
@acronym{ASN.1} object identifiers are recognized if the @acronym{IETF}
dotted format is used and the OID is prefixed with either "@code{oid.}"
or "@code{OID.}".  For a list of supported OIDs, see the source code at
@file{cipher/md.c}. This function should not be used to test for the
availability of an algorithm.
@end deftypefun

@deftypefun gcry_error_t gcry_md_get_asnoid (int @var{algo}, void *@var{buffer}, size_t *@var{length})

Return an DER encoded ASN.1 OID for the algorithm @var{algo} in the
user allocated @var{buffer}. @var{length} must point to variable with
the available size of @var{buffer} and receives after return the
actual size of the returned OID.  The returned error code may be
@code{GPG_ERR_TOO_SHORT} if the provided buffer is to short to receive
the OID; it is possible to call the function with @code{NULL} for
@var{buffer} to have it only return the required size.  The function
returns 0 on success.

@end deftypefun


To test whether an algorithm is actually available for use, the
following macro should be used:

@deftypefun gcry_error_t gcry_md_test_algo (int @var{algo}) 

The macro returns 0 if the algorithm @var{algo} is available for use.
@end deftypefun

If the length of a message digest is not known, it can be retrieved
using the following function:

@deftypefun {unsigned int} gcry_md_get_algo_dlen (int @var{algo})

Retrieve the length in bytes of the digest yielded by algorithm
@var{algo}.  This is often used prior to @code{gcry_md_read} to allocate
sufficient memory for the digest.
@end deftypefun


In some situations it might be hard to remember the algorithm used for
the ongoing hashing. The following function might be used to get that
information:

@deftypefun int gcry_md_get_algo (gcry_md_hd_t @var{h})

Retrieve the algorithm used with the handle @var{h}.  Note that this
does not work reliable if more than one algorithm is enabled in @var{h}.
@end deftypefun

The following macro might also be useful:

@deftypefun int gcry_md_is_secure (gcry_md_hd_t @var{h})

This function returns true when the digest object @var{h} is allocated
in "secure memory"; i.e. @var{h} was created with the
@code{GCRY_MD_FLAG_SECURE}.
@end deftypefun

@deftypefun int gcry_md_is_enabled (gcry_md_hd_t @var{h}, int @var{algo})

This function returns true when the algorithm @var{algo} has been
enabled for the digest object @var{h}.
@end deftypefun



Tracking bugs related to hashing is often a cumbersome task which
requires to add a lot of printf statements into the code.
Libgcrypt provides an easy way to avoid this.  The actual data
hashed can be written to files on request.

@deftypefun void gcry_md_debug (gcry_md_hd_t @var{h}, const char *@var{suffix})

Enable debugging for the digest object with handle @var{h}.  This
creates create files named @file{dbgmd-<n>.<string>} while doing the
actual hashing.  @var{suffix} is the string part in the filename.  The
number is a counter incremented for each new hashing.  The data in the
file is the raw data as passed to @code{gcry_md_write} or
@code{gcry_md_putc}.  If @code{NULL} is used for @var{suffix}, the
debugging is stopped and the file closed.  This is only rarely required
because @code{gcry_md_close} implicitly stops debugging.
@end deftypefun


The following two deprecated macros are used for debugging by old code.
They shopuld be replaced by @code{gcry_md_debug}.

@deftypefun void gcry_md_start_debug (gcry_md_hd_t @var{h}, const char *@var{suffix})

Enable debugging for the digest object with handle @var{h}.  This
creates create files named @file{dbgmd-<n>.<string>} while doing the
actual hashing.  @var{suffix} is the string part in the filename.  The
number is a counter incremented for each new hashing.  The data in the
file is the raw data as passed to @code{gcry_md_write} or
@code{gcry_md_putc}.
@end deftypefun


@deftypefun void gcry_md_stop_debug (gcry_md_hd_t @var{h}, int @var{reserved})

Stop debugging on handle @var{h}.  @var{reserved} should be specified as
0.  This function is usually not required because @code{gcry_md_close}
does implicitly stop debugging.
@end deftypefun


@c **********************************************************
@c *******************  Random  *****************************
@c **********************************************************
@node Random Numbers
@chapter Random Numbers

@menu
* Quality of random numbers::   Libgcrypt uses different quality levels.
* Retrieving random numbers::   How to retrieve random numbers.
@end menu

@node Quality of random numbers
@section Quality of random numbers

@acronym{Libgcypt} offers random numbers of different quality levels:

@deftp {Data type} gcry_random_level_t
The constants for the random quality levels are of this enum type.
@end deftp

@table @code
@item GCRY_WEAK_RANDOM
For all functions, except for @code{gcry_mpi_randomize}, this level maps
to GCRY_STRONG_RANDOM.  If you do not want this, consider using
@code{gcry_create_nonce}.
@item GCRY_STRONG_RANDOM
Use this level for session keys and similar purposes.
@item GCRY_VERY_STRONG_RANDOM
Use this level for long term key material.
@end table

@node Retrieving random numbers
@section Retrieving random numbers

@deftypefun void gcry_randomize (unsigned char *@var{buffer}, size_t @var{length}, enum gcry_random_level @var{level})

Fill @var{buffer} with @var{length} random bytes using a random quality
as defined by @var{level}.
@end deftypefun

@deftypefun {void *} gcry_random_bytes (size_t @var{nbytes}, enum gcry_random_level @var{level})

Convenience function to allocate a memory block consisting of
@var{nbytes} fresh random bytes using a random quality as defined by
@var{level}.
@end deftypefun

@deftypefun {void *} gcry_random_bytes_secure (size_t @var{nbytes}, enum gcry_random_level @var{level})

Convenience function to allocate a memory block consisting of
@var{nbytes} fresh random bytes using a random quality as defined by
@var{level}.  This function differs from @code{gcry_random_bytes} in
that the returned buffer is allocated in a ``secure'' area of the
memory.
@end deftypefun

@deftypefun void gcry_create_nonce (unsigned char *@var{buffer}, size_t @var{length})

Fill @var{buffer} with @var{length} unpredictable bytes.  This is
commonly called a nonce and may also be used for initialization
vectors and padding.  This is an extra function nearly independent of
the other random function for 3 reasons: It better protects the
regular random generator's internal state, provides better performance
and does not drain the precious entropy pool.

@end deftypefun



@c **********************************************************
@c *******************  S-Expressions ***********************
@c **********************************************************
@node S-expressions
@chapter S-expressions

S-expressions are used by the public key functions to pass complex data
structures around.  These LISP like objects are used by some
cryptographic protocols (cf. RFC-2692) and Libgcrypt provides functions
to parse and construct them.  For detailed information, see
@cite{Ron Rivest, code and description of S-expressions,
@uref{http://theory.lcs.mit.edu/~rivest/sexp.html}}.

@menu
* Data types for S-expressions::  Data types related with S-expressions.
* Working with S-expressions::  How to work with S-expressions.
@end menu

@node Data types for S-expressions
@section Data types for S-expressions

@deftp {Data type} gcry_sexp_t
The @code{gcry_sexp_t} type describes an object with the Libgcrypt internal
representation of an S-expression.
@end deftp

@node Working with S-expressions
@section Working with S-expressions

@noindent
There are several functions to create an Libgcrypt S-expression object
from its external representation or from a string template.  There is
also a function to convert the internal representation back into one of
the external formats:


@deftypefun gcry_error_t gcry_sexp_new (@w{gcry_sexp_t *@var{r_sexp}}, @w{const void *@var{buffer}}, @w{size_t @var{length}}, @w{int @var{autodetect}})

This is the generic function to create an new S-expression object from
its external representation in @var{buffer} of @var{length} bytes.  On
success the result is stored at the address given by @var{r_sexp}. 
With @var{autodetect} set to 0, the data in @var{buffer} is expected to
be in canonized format, with @var{autodetect} set to 1 the parses any of
the defined external formats.  If @var{buffer} does not hold a valid
S-expression an error code is returned and @var{r_sexp} set to
@code{NULL}.
Note that the caller is responsible for releasing the newly allocated
S-expression using @code{gcry_sexp_release}.
@end deftypefun

@deftypefun gcry_error_t gcry_sexp_create (@w{gcry_sexp_t *@var{r_sexp}}, @w{void *@var{buffer}}, @w{size_t @var{length}}, @w{int @var{autodetect}}, @w{void (*@var{freefnc})(void*)})

This function is identical to @code{gcry_sexp_new} but has an extra
argument @var{freefnc}, which, when not set to @code{NULL}, is expected
to be a function to release the @var{buffer}; most likely the standard
@code{free} function is used for this argument.  This has the effect of
transferring the ownership of @var{buffer} to the created object in
@var{r_sexp}.  The advantage of using this function is that Libgcrypt
might decide to directly use the provided buffer and thus avoid extra
copying.
@end deftypefun

@deftypefun gcry_error_t gcry_sexp_sscan (@w{gcry_sexp_t *@var{r_sexp}}, @w{size_t *@var{erroff}}, @w{const char *@var{buffer}}, @w{size_t @var{length}})

This is another variant of the above functions.  It behaves nearly
identical but provides an @var{erroff} argument which will receive the
offset into the buffer where the parsing stopped on error.
@end deftypefun

@deftypefun gcry_error_t gcry_sexp_build (@w{gcry_sexp_t *@var{r_sexp}}, @w{size_t *@var{erroff}}, @w{const char *@var{format}, ...})

This function creates an internal S-expression from the string template
@var{format} and stores it at the address of @var{r_sexp}. If there is a
parsing error, the function returns an appropriate error code and stores
the offset into @var{format} where the parsing stopped in @var{erroff}.
The function supports a couple of printf-like formatting characters and
expects arguments for some of these escape sequences right after
@var{format}.  The following format characters are defined:

@table @samp
@item %m
The next argument is expected to be of type @code{gcry_mpi_t} and a copy of
its value is inserted into the resulting S-expression.
@item %s
The next argument is expected to be of type @code{char *} and that
string is inserted into the resulting S-expression.
@item %d
The next argument is expected to be of type @code{int} and its value is
inserted into the resulting S-expression.
@item %b
The next argument is expected to be of type @code{int} directly
followed by an argument of type @code{char *}.  This represents a
buffer of given length to be inserted into the resulting S-expression.
@item %S
The next argument is expected to be of type @code{gcry_sexp_t} and a
copy of that S-expression is embedded in the resulting S-expression.
The argument needs to be a regular S-expression, starting with a
parenthesis.

@end table

@noindent
No other format characters are defined and would return an error.  Note
that the format character @samp{%%} does not exists, because a percent
sign is not a valid character in an S-expression.
@end deftypefun

@deftypefun void gcry_sexp_release (@w{gcry_sexp_t @var{sexp}})

Release the S-expression object @var{sexp}.  If the S-expression is
stored in secure memory it explicitly zeroises that memory; note that
this is done in addition to the zeroisation always done when freeing
secure memory.
@end deftypefun


@noindent
The next 2 functions are used to convert the internal representation
back into a regular external S-expression format and to show the
structure for debugging.

@deftypefun size_t gcry_sexp_sprint (@w{gcry_sexp_t @var{sexp}}, @w{int @var{mode}}, @w{char *@var{buffer}}, @w{size_t @var{maxlength}})

Copies the S-expression object @var{sexp} into @var{buffer} using the
format specified in @var{mode}.  @var{maxlength} must be set to the
allocated length of @var{buffer}.  The function returns the actual
length of valid bytes put into @var{buffer} or 0 if the provided buffer
is too short.  Passing @code{NULL} for @var{buffer} returns the required
length for @var{buffer}.  For convenience reasons an extra byte with
value 0 is appended to the buffer.

@noindent
The following formats are supported:

@table @code
@item GCRYSEXP_FMT_DEFAULT
Returns a convenient external S-expression representation.

@item GCRYSEXP_FMT_CANON
Return the S-expression in canonical format.

@item GCRYSEXP_FMT_BASE64
Not currently supported.

@item GCRYSEXP_FMT_ADVANCED
Returns the S-expression in advanced format.
@end table
@end deftypefun

@deftypefun void gcry_sexp_dump (@w{gcry_sexp_t @var{sexp}})

Dumps @var{sexp} in a format suitable for debugging to Libgcrypt's
logging stream.
@end deftypefun

@noindent
Often canonical encoding is used in the external representation.  The
following function can be used to check for valid encoding and to learn
the length of the S-expression"

@deftypefun size_t gcry_sexp_canon_len (@w{const unsigned char *@var{buffer}}, @w{size_t @var{length}}, @w{size_t *@var{erroff}}, @w{int *@var{errcode}})

Scan the canonical encoded @var{buffer} with implicit length values and
return the actual length this S-expression uses.  For a valid S-expression
it should never return 0.  If @var{length} is not 0, the maximum
length to scan is given; this can be used for syntax checks of
data passed from outside.  @var{errcode} and @var{erroff} may both be
passed as @code{NULL}.

@end deftypefun


@noindent
There are functions to parse S-expressions and retrieve elements:

@deftypefun gcry_sexp_t gcry_sexp_find_token (@w{const gcry_sexp_t @var{list}}, @w{const char *@var{token}}, @w{size_t @var{toklen}})

Scan the S-expression for a sublist with a type (the car of the list)
matching the string @var{token}.  If @var{toklen} is not 0, the token is
assumed to be raw memory of this length.  The function returns a newly
allocated S-expression consisting of the found sublist or @code{NULL}
when not found.
@end deftypefun


@deftypefun int gcry_sexp_length (@w{const gcry_sexp_t @var{list}})

Return the length of the @var{list}.  For a valid S-expression this
should be at least 1.
@end deftypefun


@deftypefun gcry_sexp_t gcry_sexp_nth (@w{const gcry_sexp_t @var{list}}, @w{int @var{number}})

Create and return a new S-expression from the element with index @var{number} in
@var{list}.  Note that the first element has the index 0.  If there is
no such element, @code{NULL} is returned.
@end deftypefun

@deftypefun gcry_sexp_t gcry_sexp_car (@w{const gcry_sexp_t @var{list}})

Create and return a new S-expression from the first element in
@var{list}; this called the "type" and should always exist and be a
string. @code{NULL} is returned in case of a problem.
@end deftypefun

@deftypefun gcry_sexp_t gcry_sexp_cdr (@w{const gcry_sexp_t @var{list}})

Create and return a new list form all elements except for the first one.
Note that this function may return an invalid S-expression because it
is not guaranteed, that the type exists and is a string.  However, for
parsing a complex S-expression it might be useful for intermediate
lists.  Returns @code{NULL} on error.
@end deftypefun


@deftypefun {const char *} gcry_sexp_nth_data (@w{const gcry_sexp_t @var{list}}, @w{int @var{number}}, @w{size_t *@var{datalen}})

This function is used to get data from a @var{list}.  A pointer to the
actual data with index @var{number} is returned and the length of this
data will be stored to @var{datalen}.  If there is no data at the given
index or the index represents another list, @code{NULL} is returned.
@strong{Caution:} The returned pointer is valid as long as @var{list} is
not modified or released.

@noindent
Here is an example on how to extract and print the surname (Meier) from
the S-expression @samp{(Name Otto Meier (address Burgplatz 3))}:

@example
size_t len;
const char *name;

name = gcry_sexp_nth_data (list, 2, &len);
printf ("my name is %.*s\n", (int)len, name);
@end example
@end deftypefun

@deftypefun {char *} gcry_sexp_nth_string (@w{gcry_sexp_t @var{list}}, @w{int @var{number}})

This function is used to get and convert data from a @var{list}. The
data is assumed to be a Nul terminated string.  The caller must
release this returned value using @code{gcry_free}.  If there is
no data at the given index, the index represents a list or the value
can't be converted to a string, @code{NULL} is returned.
@end deftypefun

@deftypefun gcry_mpi_t gcry_sexp_nth_mpi (@w{gcry_sexp_t @var{list}}, @w{int @var{number}}, @w{int @var{mpifmt}})

This function is used to get and convert data from a @var{list}. This
data is assumed to be an MPI stored in the format described by
@var{mpifmt} and returned as a standard Libgcrypt MPI.  The caller must
release this returned value using @code{gcry_mpi_release}.  If there is
no data at the given index, the index represents a list or the value
can't be converted to an MPI, @code{NULL} is returned.
@end deftypefun


@c **********************************************************
@c *******************  MPIs ******** ***********************
@c **********************************************************
@node MPI library
@chapter MPI library

@menu
* Data types::                  MPI related data types.
* Basic functions::             First steps with MPI numbers.
* MPI formats::                 External representation of MPIs.
* Calculations::                Performing MPI calculations.
* Comparisons::                 How to compare MPI values.
* Bit manipulations::           How to access single bits of MPI values.
* Miscellaneous::               Miscellaneous MPI functions.
@end menu

Public key cryptography is based on mathematics with large numbers.  To
implement the public key functions, a library for handling these large
numbers is required.  Because of the general usefulness of such a
library, its interface is exposed by Libgcrypt. 
In the context of Libgcrypt and in most other applications, these large
numbers are called MPIs (multi-precision-integers).

@node Data types
@section Data types

@deftp {Data type} {gcry_mpi_t}
This type represents an object to hold an MPI.
@end deftp

@node Basic functions
@section Basic functions

@noindent
To work with MPIs, storage must be allocated and released for the
numbers.  This can be done with one of these functions:

@deftypefun gcry_mpi_t gcry_mpi_new (@w{unsigned int @var{nbits}})

Allocate a new MPI object, initialize it to 0 and initially allocate
enough memory for a number of at least @var{nbits}.  This pre-allocation is
only a small performance issue and not actually necessary because
Libgcrypt automatically re-allocates the required memory.
@end deftypefun

@deftypefun gcry_mpi_t gcry_mpi_snew (@w{unsigned int @var{nbits}})

This is identical to @code{gcry_mpi_new} but allocates the MPI in the so
called "secure memory" which in turn will take care that all derived
values will also be stored in this "secure memory".  Use this for highly
confidential data like private key parameters.
@end deftypefun

@deftypefun gcry_mpi_t gcry_mpi_copy (@w{const gcry_mpi_t @var{a}})

Create a new MPI as the exact copy of @var{a}.
@end deftypefun


@deftypefun void gcry_mpi_release (@w{gcry_mpi_t @var{a}})

Release the MPI @var{a} and free all associated resources.  Passing
@code{NULL} is allowed and ignored.  When a MPI stored in the "secure
memory" is released, that memory gets wiped out immediately.
@end deftypefun

@noindent
The simplest operations are used to assign a new value to an MPI:

@deftypefun gcry_mpi_t gcry_mpi_set (@w{gcry_mpi_t @var{w}}, @w{const gcry_mpi_t @var{u}})

Assign the value of @var{u} to @var{w} and return @var{w}.  If
@code{NULL} is passed for @var{w}, a new MPI is allocated, set to the
value of @var{u} and returned.
@end deftypefun

@deftypefun gcry_mpi_t gcry_mpi_set_ui (@w{gcry_mpi_t @var{w}}, @w{unsigned long @var{u}})

Assign the value of @var{u} to @var{w} and return @var{w}.  If
@code{NULL} is passed for @var{w}, a new MPI is allocated, set to the
value of @var{u} and returned.  This function takes an @code{unsigned
int} as type for @var{u} and thus it is only possible to set @var{w} to
small values (usually up to the word size of the CPU).
@end deftypefun

@deftypefun void gcry_mpi_swap (@w{gcry_mpi_t @var{a}}, @w{gcry_mpi_t @var{b}})

Swap the values of @var{a} and @var{b}.
@end deftypefun

@node MPI formats
@section MPI formats

@noindent
The following functions are used to convert between an external
representation of an MPI and the internal one of Libgcrypt.

@deftypefun gcry_error_t gcry_mpi_scan (@w{gcry_mpi_t *@var{r_mpi}}, @w{enum gcry_mpi_format @var{format}}, @w{const unsigned char *@var{buffer}}, @w{size_t @var{buflen}}, @w{size_t *@var{nscanned}})

Convert the external representation of an integer stored in @var{buffer}
with a length of @var{buflen} into a newly created MPI returned which
will be stored at the address of @var{r_mpi}.  For certain formats the
length argument is not required and should be passed as @code{0}.  After a
successful operation the variable @var{nscanned} receives the number of
bytes actually scanned unless @var{nscanned} was given as
@code{NULL}. @var{format} describes the format of the MPI as stored in
@var{buffer}:

@table @code
@item GCRYMPI_FMT_STD
2-complement stored without a length header.

@item GCRYMPI_FMT_PGP
As used by OpenPGP (only defined as unsigned). This is basically
@code{GCRYMPI_FMT_STD} with a 2 byte big endian length header.

@item GCRYMPI_FMT_SSH
As used in the Secure Shell protocol.  This is @code{GCRYMPI_FMT_STD}
with a 4 byte big endian header.

@item GCRYMPI_FMT_HEX
Stored as a C style string with each byte of the MPI encoded as 2 hex
digits.  When using this format, @var{buflen} must be zero.

@item GCRYMPI_FMT_USG
Simple unsigned integer.
@end table

@noindent
Note that all of the above formats store the integer in big-endian
format (MSB first).
@end deftypefun


@deftypefun gcry_error_t gcry_mpi_print (@w{enum gcry_mpi_format @var{format}}, @w{unsigned char *@var{buffer}}, @w{size_t @var{buflen}}, @w{size_t *@var{nwritten}}, @w{const gcry_mpi_t @var{a}})

Convert the MPI @var{a} into an external representation described by
@var{format} (see above) and store it in the provided @var{buffer}
which has a usable length of at least the @var{buflen} bytes. If
@var{nwritten} is not NULL, it will receive the number of bytes
actually stored in @var{buffer} after a successful operation.
@end deftypefun

@deftypefun gcry_error_t gcry_mpi_aprint (@w{enum gcry_mpi_format @var{format}}, @w{unsigned char **@var{buffer}}, @w{size_t *@var{nbytes}}, @w{const gcry_mpi_t @var{a}})

Convert the MPI @var{a} into an external representation described by
@var{format} (see above) and store it in a newly allocated buffer which
address will be stored in the variable @var{buffer} points to.  The
number of bytes stored in this buffer will be stored in the variable
@var{nbytes} points to, unless @var{nbytes} is @code{NULL}.
@end deftypefun

@deftypefun void gcry_mpi_dump (@w{const gcry_mpi_t @var{a}})

Dump the value of @var{a} in a format suitable for debugging to
Libgcrypt's logging stream.  Note that one leading space but no trailing
space or linefeed will be printed.  It is okay to pass @code{NULL} for
@var{a}.
@end deftypefun


@node Calculations
@section Calculations

@noindent
Basic arithmetic operations:

@deftypefun void gcry_mpi_add (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}})

@math{@var{w} = @var{u} + @var{v}}.
@end deftypefun


@deftypefun void gcry_mpi_add_ui (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{unsigned long @var{v}})

@math{@var{w} = @var{u} + @var{v}}.  Note that @var{v} is an unsigned integer.
@end deftypefun


@deftypefun void gcry_mpi_addm (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}}, @w{gcry_mpi_t @var{m}})

@math{@var{w} = @var{u} + @var{v} \bmod @var{m}}.
@end deftypefun

@deftypefun void gcry_mpi_sub (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}})

@math{@var{w} = @var{u} - @var{v}}.
@end deftypefun

@deftypefun void gcry_mpi_sub_ui (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{unsigned long @var{v}})

@math{@var{w} = @var{u} - @var{v}}.  @var{v} is an unsigned integer.
@end deftypefun

@deftypefun void gcry_mpi_subm (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}}, @w{gcry_mpi_t @var{m}})

@math{@var{w} = @var{u} - @var{v} \bmod @var{m}}.
@end deftypefun

@deftypefun void gcry_mpi_mul (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}})

@math{@var{w} = @var{u} * @var{v}}.
@end deftypefun

@deftypefun void gcry_mpi_mul_ui (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{unsigned long @var{v}})

@math{@var{w} = @var{u} * @var{v}}.  @var{v} is an unsigned integer.
@end deftypefun

@deftypefun void gcry_mpi_mulm (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{gcry_mpi_t @var{v}}, @w{gcry_mpi_t @var{m}})

@math{@var{w} = @var{u} * @var{v} \bmod @var{m}}.
@end deftypefun

@deftypefun void gcry_mpi_mul_2exp (@w{gcry_mpi_t @var{w}}, @w{gcry_mpi_t @var{u}}, @w{unsigned long @var{e}})

@c FIXME: I am in need for a real TeX{info} guru:
@c I don't know why TeX can grok @var{e} here.
@math{@var{w} = @var{u} * 2^e}.
@end deftypefun

@deftypefun void gcry_mpi_div (@w{gcry_mpi_t @var{q}}, @w{gcry_mpi_t @var{r}}, @w{gcry_mpi_t @var{dividend}}, @w{gcry_mpi_t @var{divisor}}, @w{int @var{round}})

@math{@var{q} = @var{dividend} / @var{divisor}}, @math{@var{r} =
@var{dividend} \bmod @var{divisor}}.  @var{q} and @var{r} may be passed
as @code{NULL}.  @var{round} should be negative or 0.
@end deftypefun

@deftypefun void gcry_mpi_mod (@w{gcry_mpi_t @var{r}}, @w{gcry_mpi_t @var{dividend}}, @w{gcry_mpi_t @var{divisor}})

@math{@var{r} = @var{dividend} \bmod @var{divisor}}.
@end deftypefun

@deftypefun void gcry_mpi_powm (@w{gcry_mpi_t @var{w}}, @w{const gcry_mpi_t @var{b}}, @w{const gcry_mpi_t @var{e}}, @w{const gcry_mpi_t @var{m}})

@c I don't know why TeX can grok @var{e} here.
@math{@var{w} = @var{b}^e \bmod @var{m}}.
@end deftypefun

@deftypefun int gcry_mpi_gcd (@w{gcry_mpi_t @var{g}}, @w{gcry_mpi_t @var{a}}, @w{gcry_mpi_t @var{b}})

Set @var{g} to the greatest common divisor of @var{a} and @var{b}.  
Return true if the @var{g} is 1.
@end deftypefun

@deftypefun int gcry_mpi_invm (@w{gcry_mpi_t @var{x}}, @w{gcry_mpi_t @var{a}}, @w{gcry_mpi_t @var{m}})

Set @var{x} to the multiplicative inverse of @math{@var{a} \bmod @var{m}}.
Return true if the inverse exists.
@end deftypefun


@node Comparisons
@section Comparisons

@noindent
The next 2 functions are used to compare MPIs:


@deftypefun int gcry_mpi_cmp (@w{const gcry_mpi_t @var{u}}, @w{const gcry_mpi_t @var{v}})

Compare the multi-precision-integers number @var{u} and @var{v}
returning 0 for equality, a positive value for @var{u} > @var{v} and a
negative for @var{u} < @var{v}.
@end deftypefun

@deftypefun int gcry_mpi_cmp_ui (@w{const gcry_mpi_t @var{u}}, @w{unsigned long @var{v}})

Compare the multi-precision-integers number @var{u} with the unsigned
integer @var{v} returning 0 for equality, a positive value for @var{u} >
@var{v} and a negative for @var{u} < @var{v}.
@end deftypefun


@node Bit manipulations
@section Bit manipulations

@noindent
There are a couple of functions to get information on arbitrary bits
in an MPI and to set or clear them:

@deftypefun {unsigned int} gcry_mpi_get_nbits (@w{gcry_mpi_t @var{a}})

Return the number of bits required to represent @var{a}.
@end deftypefun

@deftypefun int gcry_mpi_test_bit (@w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Return true if bit number @var{n} (counting from 0) is set in @var{a}.
@end deftypefun

@deftypefun void gcry_mpi_set_bit (@w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Set bit number @var{n} in @var{a}.
@end deftypefun

@deftypefun void gcry_mpi_clear_bit (@w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Clear bit number @var{n} in @var{a}.
@end deftypefun

@deftypefun void gcry_mpi_set_highbit (@w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Set bit number @var{n} in @var{a} and clear all bits greater than @var{n}.
@end deftypefun

@deftypefun void gcry_mpi_clear_highbit (@w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Clear bit number @var{n} in @var{a} and all bits greater than @var{n}.
@end deftypefun

@deftypefun void gcry_mpi_rshift (@w{gcry_mpi_t @var{x}}, @w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Shift the value of @var{a} by @var{n} bits to the right and store the
result in @var{x}.
@end deftypefun

@deftypefun void gcry_mpi_lshift (@w{gcry_mpi_t @var{x}}, @w{gcry_mpi_t @var{a}}, @w{unsigned int @var{n}})

Shift the value of @var{a} by @var{n} bits to the left and store the
result in @var{x}.
@end deftypefun

@node Miscellaneous
@section Miscellaneous

@deftypefun gcry_mpi_t gcry_mpi_set_opaque (@w{gcry_mpi_t @var{a}}, @w{void *@var{p}}, @w{unsigned int @var{nbits}})

Store @var{nbits} of the value @var{p} points to in @var{a} and mark
@var{a} as an opaque value (i.e. an value that can't be used for any
math calculation and is only used to store an arbitrary bit pattern in
@var{a}).

WARNING: Never use an opaque MPI for actual math operations.  The only
valid functions are gcry_mpi_get_opaque and gcry_mpi_release.  Use
gcry_mpi_scan to convert a string of arbitrary bytes into an MPI.

@end deftypefun

@deftypefun {void *} gcry_mpi_get_opaque (@w{gcry_mpi_t @var{a}}, @w{unsigned int *@var{nbits}})

Return a pointer to an opaque value stored in @var{a} and return its
size in @var{nbits}.  Note that the returned pointer is still owned by
@var{a} and that the function should never be used for an non-opaque
MPI.
@end deftypefun

@deftypefun void gcry_mpi_set_flag (@w{gcry_mpi_t @var{a}}, @w{enum gcry_mpi_flag @var{flag}})

Set the @var{flag} for the MPI @var{a}.  Currently only the flag
@code{GCRYMPI_FLAG_SECURE} is allowed to convert @var{a} into an MPI
stored in "secure memory".
@end deftypefun

@deftypefun void gcry_mpi_clear_flag (@w{gcry_mpi_t @var{a}}, @w{enum gcry_mpi_flag @var{flag}})

Clear @var{flag} for the multi-precision-integers @var{a}.  Note that
this function is currently useless as no flags are allowed.
@end deftypefun

@deftypefun int gcry_mpi_get_flag (@w{gcry_mpi_t @var{a}}, @w{enum gcry_mpi_flag @var{flag}})

Return true when the @var{flag} is set for @var{a}.
@end deftypefun

@deftypefun void gcry_mpi_randomize (@w{gcry_mpi_t @var{w}}, @w{unsigned int @var{nbits}}, @w{enum gcry_random_level @var{level}})

Set the multi-precision-integers @var{w} to a random value of
@var{nbits}, using random data quality of level @var{level}.  In case
@var{nbits} is not a multiple of a byte, @var{nbits} is rounded up to
the next byte boundary.  When using a @var{level} of
@code{GCRY_WEAK_RANDOM} this function makes use of
@code{gcry_create_nonce}.
@end deftypefun

@c **********************************************************
@c ******************** Prime numbers ***********************
@c **********************************************************
@node Prime numbers
@chapter Prime numbers

@menu
* Generation::                  Generation of new prime numbers.
* Checking::                    Checking if a given number is prime.
@end menu

@node Generation
@section Generation

@deftypefun gcry_error_t gcry_prime_generate (gcry_mpi_t *@var{prime},unsigned int @var{prime_bits}, unsigned int @var{factor_bits}, gcry_mpi_t **@var{factors}, gcry_prime_check_func_t @var{cb_func}, void *@var{cb_arg}, gcry_random_level_t @var{random_level}, unsigned int @var{flags})

Generate a new prime number of @var{prime_bits} bits and store it in
@var{prime}.  If @var{factor_bits} is non-zero, one of the prime factors
of (@var{prime} - 1) / 2 must be @var{factor_bits} bits long.  If
@var{factors} is non-zero, allocate a new, @code{NULL}-terminated array
holding the prime factors and store it in @var{factors}.  @var{flags}
might be used to influence the prime number generation process.
@end deftypefun

@deftypefun gcry_error_t gcry_prime_group_generator (gcry_mpi_t *@var{r_g}, gcry_mpi_t @var{prime}, gcry_mpi_t *@var{factors}, gcry_mpi_t @var{start_g})

Find a generator for @var{prime} where the factorization of
(@var{prime}-1) is in the @code{NULL} terminated array @var{factors}.
Return the generator as a newly allocated MPI in @var{r_g}.  If
@var{start_g} is not NULL, use this as the start for the search.
@end deftypefun

@deftypefun void gcry_prime_release_factors (gcry_mpi_t *@var{factors})

Convenience function to release the @var{factors} array.
@end deftypefun

@node Checking
@section Checking

@deftypefun gcry_error_t gcry_prime_check (gcry_mpi_t @var{p}, unsigned int @var{flags})

Check wether the number @var{p} is prime.  Returns zero in case @var{p}
is indeed a prime, returns @code{GPG_ERR_NO_PRIME} in case @var{p} is
not a prime and a different error code in case something went horribly
wrong.
@end deftypefun

@c **********************************************************
@c ******************** Utilities ***************************
@c **********************************************************
@node Utilities
@chapter Utilities

@menu
* Memory allocation:: Functions related with memory allocation.
@end menu

@node Memory allocation
@section Memory allocation

@deftypefun {void *} gcry_malloc (size_t @var{n})

This function tries to allocate @var{n} bytes of memory.  On success
it returns a pointer to the memory area, in an out-of-core condition,
it returns NULL.
@end deftypefun

@deftypefun {void *} gcry_malloc_secure (size_t @var{n})
Like @code{gcry_malloc}, but uses secure memory.
@end deftypefun

@deftypefun {void *} gcry_calloc (size_t @var{n}, size_t @var{m})

This function allocates a cleared block of memory (i.e. initialized with
zero bytes) long enough to contain a vector of @var{n} elements, each of
size @var{m} bytes.  On success it returns a pointer to the memory
block; in an out-of-core condition, it returns NULL.
@end deftypefun

@deftypefun {void *} gcry_calloc_secure (size_t @var{n}, size_t @var{m})
Like @code{gcry_calloc}, but uses secure memory.
@end deftypefun

@deftypefun {void *} gcry_realloc (void *@var{p}, size_t @var{n})

This function tries to resize the memory area pointed to by @var{p} to
@var{n} bytes.  On success it returns a pointer to the new memory
area, in an out-of-core condition, it returns NULL.  Depending on
whether the memory pointed to by @var{p} is secure memory or not,
gcry_realloc tries to use secure memory as well.
@end deftypefun

@deftypefun void gcry_free (void *@var{p})
Release the memory area pointed to by @var{p}.
@end deftypefun

@c **********************************************************
@c *****************  Architecure Overview  *****************
@c **********************************************************
@node Architecture
@chapter Architecture

This chapter describes the internal architecture of Libgcrypt.

Libgcrypt is a function library written in ISO C-90.  Any compliant
compiler should be able to build Libgcrypt as long as the target is
either a POSIX platform or compatible to the API used by Windows NT.
Provisions have been take so that the library can be directly used from
C++ applications; however building with a C++ compiler is not supported.

Building Libgcrypt is done by using the common @code{./configure && make}
approach.  The configure command is included in the source distribution
and as a portable shell script it works on any Unix-alike system.  The
result of running the configure script are a C header file
(@file{config.h}), customized Makefiles, the setup of symbolic links and
a few other things.  After that the make tool builds and optionally
installs the library and the documentation.  See the files
@file{INSTALL} and @file{README} in the source distribution on how to do
this.

Libgcrypt is developed using a Subversion@footnote{A version control
system available for many platforms} repository.  Although all released
versions are tagged in this repository, they should not be used to build
production versions of Libgcrypt.  Instead released tarballs should be
used.  These tarballs are available from several places with the master
copy at @indicateurl{ftp://ftp.gnupg.org/gcrypt/libgcrypt/}.
Announcements of new releases are posted to the
@indicateurl{gnupg-announce@@gnupg.org} mailing list@footnote{See
@url{http://www.gnupg.org/documentation/mailing-lists.en.html} for
details.}.


@float Figure,fig:subsystems
@caption{Libgcrypt subsystems}
@center @image{libgcrypt-modules, 150mm,,Libgcrypt subsystems}
@end float

Libgcrypt consists of several subsystems (@pxref{fig:subsystems}) and
all these subsystems provide a public API; this includes the helper
subsystems like the one for S-expressions.  The API style depends on the
subsystem; in general an open-use-close approach is implemented.  The
open returns a handle to a context used for all further operations on
this handle, several functions may then be used on this handle and a
final close function releases all resources associated with the handle.

@menu
* Public-Key Subsystem Architecture::              About public keys.
* Symmetric Encryption Subsystem Architecture::    About standard ciphers.
* Hashing and MACing Subsystem Architecture::      About hashing.
* Multi-Precision-Integer Subsystem Architecture:: About big integers.
* Prime-Number-Generator Subsystem Architecture::  About prime numbers.
* Random-Number Subsystem Architecture::           About random stuff.
@c * Helper Subsystems Architecture::                 About other stuff.
@end menu



@node Public-Key Subsystem Architecture
@section Public-Key Architecture

Libgcrypt implements two interfaces for public key cryptography: The
standard interface is PK interface using functions in the
@code{gcry_pk_} name space.  The AC interface in an alternative one
which is now deprecated and will not be further described.  The AC
interface is also disabled in FIPS mode.

Because public key cryptography is almost always used to process small
amounts of data (hash values or session keys), the interface is not
implemented using the open-use-close paradigm, but with single
self-contained functions.  Due to the wide variety of parameters
required by different algorithms S-expressions, as flexible way to
convey these parameters, are used.  There is a set of helper functions
to work with these S-expressions.
@c see @xref{S-expression Subsystem Architecture}.

Aside of functions to register new algorithms, map algorithms names to
algorithms identifiers and to lookup properties of a key, the
following main functions are available:

@table @code

@item gcry_pk_encrypt 
Encrypt data using a public key.

@item gcry_pk_decrypt 
Decrypt data using a private key.

@item gcry_pk_sign 
Sign data using a private key.

@item gcry_pk_verify
Verify that a signature matches the data.

@item gcry_pk_testkey
Perform a consistency over a public or private key.

@item gcry_pk_genkey
Create a new public/private key pair.

@end table

With the help of the module registration system all these functions
lookup the module implementing the algorithm and pass the actual work
to that module.  The parsing of the S-expression input and the
construction of S-expression for the return values is done by the high
level code (@file{cipher/pubkey.c}).  Thus the internal interface
between the algorithm modules and the high level functions passes data
in a custom format.  The interface to the modules is published
(@file{gcrypt-modules.h}) so that it can used to register external
implementations of algorithms with Libgcrypt.  However, for some
algorithms this module interface is to limited and thus for the
internal modules an extra interface is sometimes used to convey more
information.

By default Libgcrypt uses a blinding technique for RSA decryption to
mitigate real world timing attacks over a network: Instead of using
the RSA decryption directly, a blinded value @math{y = x r^{e} \bmod n}
is decrypted and the unblinded value @math{x' = y' r^{-1} \bmod n}
returned.  The blinding value @math{r} is a random value with the size
of the modulus @math{n} and generated with @code{GCRY_WEAK_RANDOM}
random level.

@cindex X9.31
@cindex FIPS 186
The algorithm used for RSA and DSA key generation depends on whether
Libgcrypt is operated in standard or in FIPS mode.  In standard mode
an algorithm based on the Lim-Lee prime number generator is used.  In
FIPS mode RSA keys are generated as specified in ANSI X9.31 (1998) and
DSA keys as specified in FIPS 186-2.



@node Symmetric Encryption Subsystem Architecture
@section Symmetric Encryption Subsystem Architecture

The interface to work with symmetric encryption algorithms is made up
of functions from the @code{gcry_cipher_} name space.  The
implementation follows the open-use-close paradigm and uses registered
algorithm modules for the actual work.  Unless a module implements
optimized cipher mode implementations, the high level code
(@file{cipher/cipher.c}) implements the modes and calls the core
algorithm functions to process each block.

The most important functions are:

@table @code

@item gcry_cipher_open
Create a new instance to encrypt or decrypt using a specified
algorithm and mode.

@item gcry_cipher_close
Release an instance.

@item gcry_cipher_setkey
Set a key to be used for encryption or decryption. 

@item gcry_cipher_setiv
Set an initialization vector to be used for encryption or decryption.

@item gcry_cipher_encrypt
@itemx gcry_cipher_decrypt 
Encrypt or decrypt data.  These functions may be called with arbitrary
amounts of data and as often as needed to encrypt or decrypt all data.

@end table

There are also functions to query properties of algorithms or context,
like block length, key length, map names or to enable features like
padding methods.



@node Hashing and MACing Subsystem Architecture
@section Hashing and MACing Subsystem Architecture

The interface to work with message digests and CRC algorithms is made
up of functions from the @code{gcry_md_} name space.  The
implementation follows the open-use-close paradigm and uses registered
algorithm modules for the actual work.  Although CRC algorithms are
not considered cryptographic hash algorithms, they share enough
properties so that it makes sense to handle them in the same way.
It is possible to use several algorithms at once with one context and
thus compute them all on the same data.

The most important functions are:

@table @code
@item gcry_md_open
Create a new message digest instance and optionally enable one
algorithm.  A flag may be used to turn the message digest algorithm
into a HMAC algorithm.

@item gcry_md_enable
Enable an additional algorithm for the instance.

@item gcry_md_setkey
Set the key for the MAC.

@item gcry_md_write
Pass more data for computing the message digest to an instance.

@item gcry_md_putc
Buffered version of @code{gcry_md_write} implemented as a macro.

@item gcry_md_read
Finalize the computation of the message digest or HMAC and return the
result.

@item gcry_md_close
Release an instance

@item gcry_md_hash_buffer
Convenience function to directly compute a message digest over a
memory buffer without the need to create an instance first.

@end table

There are also functions to query properties of algorithms or the
instance, like enabled algorithms, digest length, map algorithm names.
it is also possible to reset an instance or to copy the current state
of an instance at any time.  Debug functions to write the hashed data
to files are available as well.



@node Multi-Precision-Integer Subsystem Architecture
@section Multi-Precision-Integer Subsystem Architecture

The implementation of Libgcrypt's big integer computation code is
based on an old release of GNU Multi-Precision Library (GMP).  The
decision not to use the GMP library directly was due to stalled
development at that time and due to security requirements which could
not be provided by the code in GMP.  As GMP does, Libgcrypt provides
high performance assembler implementations of low level code for
several CPUS to gain much better performance than with a generic C
implementation.

@noindent
Major features of Libgcrypt's multi-precision-integer code compared to
GMP are:

@itemize
@item 
Avoidance of stack based allocations to allow protection against
swapping out of sensitive data and for easy zeroing of sensitive
intermediate results.

@item
Optional use of secure memory and tracking of its use so that results
are also put into secure memory.

@item
MPIs are identified by a handle (implemented as a pointer) to give
better control over allocations and to augment them with extra
properties like opaque data.

@item
Removal of unnecessary code to reduce complexity.

@item
Functions specialized for public key cryptography.

@end itemize



@node Prime-Number-Generator Subsystem Architecture
@section Prime-Number-Generator Subsystem Architecture

Libgcrypt provides an interface to its prime number generator.  These
functions make use of the internal prime number generator which is
required for the generation for public key key pairs.  The plain prime
checking function is exported as well.

The generation of random prime numbers is based on the Lim and Lee
algorithm to create practically save primes.@footnote{Chae Hoon Lim
and Pil Joong Lee. A key recovery attack on discrete log-based shemes
using a prime order subgroup. In Burton S. Kaliski Jr., editor,
Advances in Cryptology: Crypto '97, pages 249­-263, Berlin /
Heidelberg / New York, 1997. Springer-Verlag.  Described on page 260.}
This algorithm creates a pool of smaller primes, select a few of them
to create candidate primes of the form @math{2 * p_0 * p_1 * ... * p_n
+ 1}, tests the candidate for primality and permutates the pool until
a prime has been found.  It is possible to clamp one of the small
primes to a certain size to help DSA style algorithms.  Because most
of the small primes in the pool are not used for the resulting prime
number, they are saved for later use (see @code{save_pool_prime} and
@code{get_pool_prime} in @file{cipher/primegen.c}).  The prime
generator optionally supports the finding of an appropriate generator.

@noindent
The primality test works in three steps:

@enumerate
@item
The standard sieve algorithm using the primes up to 4999 is used as a
quick first check.

@item
A Fermat test filters out almost all non-primes.

@item
A 5 round Rabin-Miller test is finally used.  The first round uses a
witness of 2, whereas the next rounds use a random witness.

@end enumerate

To support the generation of RSA and DSA keys in FIPS mode according
to X9.31 and FIPS 186-2, Libgcrypt implements two additional prime
generation functions: @code{_gcry_derive_x931_prime} and
@code{_gcry_generate_fips186_2_prime}.  These functions are internal
and not available through the public API.



@node Random-Number Subsystem Architecture
@section Random-Number Subsystem Architecture

Libgcrypt provides 3 levels or random quality: The level
@code{GCRY_VERY_STRONG_RANDOM} usually used for key generation, the
level @code{GCRY_STRONG_RANDOM} for all other strong random
requirements and the function @code{gcry_create_nonce} which is used
for weaker usages like nonces.  There is also a level
@code{GCRY_WEAK_RANDOM} which in general maps to
@code{GCRY_STRONG_RANDOM} except when used with the function
@code{gcry_mpi_randomize}, where it randomizes an
multi-precision-integer using the @code{gcry_create_nonce} function.

@noindent
There are two distinct random generators available: 

@itemize
@item 
The Continuously Seeded Pseudo Random Number Generator (CSPRNG), which
is based on the classic GnuPG derived big pool implementation.
Implemented in @code{random/random-csprng.c} and used by default.
@item
A FIPS approved ANSI X9.31 PRNG using AES with a 128 bit key. Implemented in
@code{random/random-fips.c} and used if Libgcrypt is in FIPS mode.
@end itemize

@noindent
Both generators make use of so-called entropy gathering modules:

@table @asis
@item rndlinux
Uses the operating system provided
@file{/dev/random} and @file{/dev/urandom} devices.

@item rndunix
Runs several operating system commands to collect entropy from sources
like virtual machine and process statistics.  It is a kind of
poor-man's @code{/dev/random} implementation. It is not available in
FIPS mode.

@item rndegd
Uses the operating system provided Entropy Gathering Daemon (EGD).
The EGD basically uses the same algorithms as rndunix does.  However
as a system daemon it keeps on running and thus can serve several
processes requiring entropy input and does not waste collected entropy
if the application does not need all the collected entropy. It is not
available in FIPS mode.

@item rndw32
Targeted for the Microsoft Windows OS.  It uses certain properties of
that system and is the only gathering module available for that OS.

@item rndhw
Extra module to collect additional entropy by utilizing a hardware
random number generator.  As of now the only supported hardware RNG is
the Padlock engine of VIA (Centaur) CPUs.  It is not available in FIPS
mode.

@end table


@menu
* CSPRNG Description::      Description of the CSPRNG.
* FIPS PRNG Description::   Description of the FIPS X9.31 PRNG.
@end menu


@node CSPRNG Description
@subsection Description of the CSPRNG

This random number generator is loosely modelled after the one
described in Peter Gutmann's paper: "Software Generation of
Practically Strong Random Numbers".@footnote{Also described in chapter
6 of his book "Cryptographic Security Architecture", New York, 2004,
ISBN 0-387-95387-6.}

A pool of 600 bytes is used and mixed using the core RIPE-MD160 hash
transform function.  Several extra features are used to make the
robust against a wide variety of attacks and to protect against
failures of subsystems.  The state of the generator may be saved to a
file and initially seed form a file.

Depending on how Libgcrypt was build the generator is able to select
the best working entropy gathering module.  It makes use of the slow
and fast collection methods and requires the pool to initially seeded
form the slow gatherer or a seed file.  An entropy estimation is used
to mix in enough data from the gather modules before returning the
actual random output.  Process fork detection and protection is
implemented.

@c FIXME:  The design and implementaion needs a more verbose description.

The implementation of the nonce generator (for
@code{gcry_create_nonce}) is a straightforward repeated hash design: A
28 byte buffer is initially seeded with the PID and the time in
seconds in the first 20 bytes and with 8 bytes of random taken from
the @code{GCRY_STRONG_RANDOM} generator.  Random numbers are then
created by hashing all the 28 bytes with SHA-1 and saving that again
in the first 20 bytes.  The hash is also returned as result.


@node FIPS PRNG Description
@subsection Description of the FIPS X9.31 PRNG

The core of this deterministic random number generator is implemented
according to the document ``NIST-Recommended Random Number Generator
Based on ANSI X9.31 Appendix A.2.4 Using the 3-Key Triple DES and AES
Algorithms'', dated 2005-01-31.  This implementation uses the AES
variant.

The generator is based on contexts to utilize the same core functions
for all random levels as required by the high-level interface.  All
random generators return their data in 128 bit blocks.  If the caller
requests less bits, the extra bits are not used.  The key for each
generator is only set once at the first time a generator context is
used.  The seed value is set along with the key and again after 1000
output blocks.

On Unix like systems the @code{GCRY_VERY_STRONG_RANDOM} and
@code{GCRY_STRONG_RANDOM} generators are keyed and seeded using the
rndlinux module with the @file{/dev/radnom} device. Thus these
generators may block until the OS kernel has collected enough entropy.
When used with Microsoft Windows the rndw32 module is used instead.

The generator used for @code{gcry_create_nonce} is keyed and seeded
from the @code{GCRY_STRONG_RANDOM} generator.  Thus is may also block
if the @code{GCRY_STRONG_RANDOM} generator has not yet been used
before and thus gets initialized on the first use by
@code{gcry_create_nonce}.  This special treatment is justified by the
weaker requirements for a nonce generator and to save precious kernel
entropy for use by the ``real'' random generators.

A self-test facility uses a separate context to check the
functionality of the core X9.31 functions using a known answers test.
During runtime each output block is compared to the previous one to
detect a stucked generator.

The DT value for the generator is made up of the current time down to
microseconds (if available) and a free running 64 bit counter.  When
used with the test context the DT value is taken from the context and
incremented on each use.

@c @node Helper Subsystems Architecture
@c @section Helper Subsystems Architecture
@c 
@c There are a few smaller subsystems which are mainly used internally by
@c Libgcrypt but also available to applications.
@c 
@c @menu
@c * S-expression Subsystem Architecture::   Details about the S-expression architecture.
@c * Memory Subsystem Architecture::         Details about the memory allocation architecture.
@c * Miscellaneous Subsystems Architecture:: Details about other subsystems.
@c @end menu
@c 
@c @node S-expression Subsystem Architecture
@c @subsection S-expression Subsystem Architecture
@c 
@c Libgcrypt provides an interface to S-expression to create and parse
@c them.  To use an S-expression with Libgcrypt it needs first be
@c converted into the internal representation used by Libgcrypt (the type
@c @code{gcry_sexp_t}).  The conversion functions support a large subset
@c of the S-expression specification and further fature a printf like
@c function to convert a list of big integers or other binary data into
@c an S-expression.
@c 
@c Libgcrypt currently implements S-expressions using a tagged linked
@c list.  However this is not exposed to an application and may be
@c changed in future releases to reduce overhead when already working
@c with canonically encoded S-expressions.  Secure memory is supported by
@c this S-expressions implementation.
@c 
@c @node Memory Subsystem Architecture 
@c @subsection Memory Subsystem Architecture 
@c 
@c TBD.
@c 
@c 
@c @node Miscellaneous Subsystems Architecture
@c @subsection Miscellaneous Subsystems Architecture
@c 
@c TBD.
@c 
@c 



@c **********************************************************
@c *******************  Appendices  *************************
@c **********************************************************

@c ********************************************
@node Self-Tests
@appendix Description of the Self-Tests

In addition to the build time regression test suite, Libgcrypt
implements self-tests to be performed at runtime.  Which self-tests
are actually used depends on the mode Libgcrypt is used in.  In
standard mode a limited set of self-tests is run at the time an
algorithm is first used.  Note that not all algorithms feature a
self-test in standard mode.  The @code{GCRYCTL_SELFTEST} control
command may be used to run all implemented self-tests at any time;
this will even run more tests than those run in FIPS mode.

If any of the self-tests fails, the library immediately returns an
error code to the caller.  If Libgcrypt is in FIPS mode the self-tests
will be performed within the ``Self-Test'' state and any failure puts
the library into the ``Error'' state.

@c --------------------------------
@section Power-Up Tests

Power-up tests are only performed if Libgcrypt is in FIPS mode.  

@subsection Symmetric Cipher Algorithm Power-Up Tests

The following symmetric encryption algorithm tests are run during
power-up:

@table @asis
@item 3DES
To test the 3DES 3-key EDE encryption in ECB mode these tests are
run:
@enumerate
@item
A known answer test is run on a 64 bit test vector processed by 64
rounds of Single-DES block encryption and decryption using a key
changed with each round.
@item
A known answer test is run on a 64 bit test vector processed by 16
rounds of 2-key and 3-key Triple-DES block encryption and decryptions
using a key changed with each round.
@item
10 known answer tests using 3-key Triple-DES EDE encryption, comparing
the ciphertext to the known value, then running a decryption and
comparing it to the initial plaintext.
@end enumerate
(@code{cipher/des.c:selftest})

@item AES-128
A known answer tests is run using one test vector and one test
key with AES in ECB mode. (@code{cipher/rijndael.c:selftest_basic_128})

@item AES-192
A known answer tests is run using one test vector and one test
key with AES in ECB mode. (@code{cipher/rijndael.c:selftest_basic_192})

@item AES-256
A known answer tests is run using one test vector and one test key
with AES in ECB mode. (@code{cipher/rijndael.c:selftest_basic_256})
@end table

@subsection Hash Algorithm Power-Up Tests

The following hash algorithm tests are run during power-up:

@table @asis
@item SHA-1
A known answer test using the string @code{"abc"} is run.
(@code{cipher/@/sha1.c:@/selftests_sha1})
@item SHA-224
A known answer test using the string @code{"abc"} is run.
(@code{cipher/@/sha256.c:@/selftests_sha224})
@item SHA-256
A known answer test using the string @code{"abc"} is run.
(@code{cipher/@/sha256.c:@/selftests_sha256})
@item SHA-384
A known answer test using the string @code{"abc"} is run.
(@code{cipher/@/sha512.c:@/selftests_sha384})
@item SHA-512
A known answer test using the string @code{"abc"} is run.
(@code{cipher/@/sha512.c:@/selftests_sha512})
@end table

@subsection MAC Algorithm Power-Up Tests

The following MAC algorithm tests are run during power-up:

@table @asis
@item HMAC SHA-1
A known answer test using 9 byte of data and a 64 byte key is run.
(@code{cipher/hmac-tests.c:selftests_sha1})
@item HMAC SHA-224
A known answer test using 28 byte of data and a 4 byte key is run.
(@code{cipher/hmac-tests.c:selftests_sha224})
@item HMAC SHA-256
A known answer test using 28 byte of data and a 4 byte key is run.
(@code{cipher/hmac-tests.c:selftests_sha256})
@item HMAC SHA-384
A known answer test using 28 byte of data and a 4 byte key is run.
(@code{cipher/hmac-tests.c:selftests_sha384})
@item HMAC SHA-512
A known answer test using 28 byte of data and a 4 byte key is run.
(@code{cipher/hmac-tests.c:selftests_sha512})
@end table

@subsection Random Number Power-Up Test

The DRNG is tested during power-up this way:

@enumerate
@item 
Requesting one block of random using the public interface to check
general working and the duplicated block detection.
@item
3 know answer tests using pre-defined keys, seed and initial DT
values.  For each test 3 blocks of 16 bytes are requested and compared
to the expected result.  The DT value is incremented for each block.
@end enumerate

@subsection Public Key Algorithm Power-Up Tests

The public key algorithms are tested during power-up:

@table @asis
@item RSA
A pre-defined 1024 bit RSA key is used and these tests are run
in turn:
@enumerate
@item 
Conversion of S-expression to internal format. 
(@code{cipher/@/rsa.c:@/selftests_rsa})
@item 
Private key consistency check.
(@code{cipher/@/rsa.c:@/selftests_rsa})
@item
A pre-defined 20 byte value is signed with PKCS#1 padding for SHA-1.
The result is verified using the public key against the original data
and against modified data.  (@code{cipher/@/rsa.c:@/selftest_sign_1024})
@item
A 1000 bit random value is encrypted and checked that it does not
match the orginal random value.  The encrtypted result is then
decrypted and checked that it macthes the original random value.
(@code{cipher/@/rsa.c:@/selftest_encr_1024})
@end enumerate

@item DSA
A pre-defined 1024 bit DSA key is used and these tests are run in turn:
@enumerate
@item
Conversion of S-expression to internal format.
(@code{cipher/@/dsa.c:@/selftests_dsa})
@item
Private key consistency check.
(@code{cipher/@/dsa.c:@/selftests_dsa})
@item
A pre-defined 20 byte value is signed with PKCS#1 padding for
SHA-1.  The result is verified using the public key against the
original data and against modified data.
(@code{cipher/@/dsa.c:@/selftest_sign_1024})
@end enumerate
@end table

@subsection Integrity Power-Up Tests

The integrity of the Libgcrypt is tested during power-up but only if
checking has been enabled at build time.  The check works by computing
a HMAC SHA-256 checksum over the file used to load Libgcrypt into
memory.  That checksum is compared against a checksum stored in a file
of the same name but with a single dot as a prefix and a suffix of
@file{.hmac}.


@subsection Critical Functions Power-Up Tests

The 3DES weak key detection is tested during power-up by calling the
detection function with keys taken from a table listening all weak
keys.  The table itself is protected using a SHA-1 hash.
(@code{cipher/@/des.c:@/selftest})



@c --------------------------------
@section Conditional Tests

The conditional tests are performed if a certain contidion is met.
This may occur at any time; the library does not necessary enter the
``Self-Test'' state to run these tests but will transit to the
``Error'' state if a test failed.

@subsection Key-Pair Generation Tests

After an asymmetric key-pair has been generated, Libgcrypt runs a
pair-wise consistency tests on the generated key.  On failure the
generated key is not used, an error code is returned and, if in FIPS
mode, the library is put into the ``Error'' state.

@table @asis
@item RSA
The test uses a random number 64 bits less the size of the modulus as
plaintext and runs an encryption and decryption operation in turn.  The
encrypted value is checked to not match the plaintext and the result
of the decryption is checked to match the plaintext.

A new random number of the same size is generated, signed and verified
to test the correctness of the signing operation.  As a second signing
test, the signature is modified by incrementing its value and then
verified with the expected result that the verification fails.
(@code{cipher/@/rsa.c:@/test_keys})
@item DSA
The test uses a random number of the size of the Q parameter to create
a signature and then checks that the signature verifies.  As a second
signing test, the data is modified by incrementing its value and then
verified against the signature with the expected result that the
verification fails.  (@code{cipher/@/dsa.c:@/test_keys})
@end table


@subsection Software Load Tests

Loading of extra modules into libgcrypt is disabled in FIPS mode and
thus no tests are
implemented. (@code{cipher/@/cipher.c:@/_gcry_cipher_register},
@code{cipher/@/md.c:@/_gcry_md_register},
@code{cipher/@/pubkey.c:@/_gcry_pk_register})


@subsection Manual Key Entry Tests

A manual key entry feature is not implemented in Libgcrypt.


@subsection Continuous RNG Tests

The continuous random number test is only used in FIPS mode.  The RNG
generates blocks of 128 bit size; the first block generated per
context is saved in the context and another block is generated to be
returned to the caller.  Each block is compared against the saved
block and then stored in the context.  If a duplicated block is
detected an error is signaled and the libray is put into the
``Fatal-Error'' state.
(@code{random/@/random-fips.c:@/x931_aes_driver})



@c --------------------------------
@section Application Requested Tests

The application may requests tests at any time by means of the
@code{GCRYCTL_SELFTEST} control command.  Note that using these tests
is not FIPS conform: Although Libgcrypt rejects all application
requests for services while running self-tests, it does not ensure
that no other operations of Libgcrypt are still being executed.  Thus,
in FIPS mode an application requesting self-tests needs to power-cycle
Libgcrypt instead.

When self-tests are requested, Libgcrypt runs all the tests it does
during power-up as well as a few extra checks as described below.

@subsection Symmetric Cipher Algorithm Tests

The following symmetric encryption algorithm tests are run in addition
to the power-up tests:

@table @asis
@item AES-128
A known answer tests with test vectors taken from NIST SP800-38a and
using the high level functions is run for block modes CFB and OFB.

@end table

@subsection Hash Algorithm Tests

The following hash algorithm tests are run in addition to the 
power-up tests:

@table @asis
@item SHA-1
@itemx SHA-224
@itemx SHA-256
@enumerate
@item
A known answer test using a 56 byte string is run.
@item 
A known answer test using a string of one million letters "a" is run.
@end enumerate
(@code{cipher/@/sha1.c:@/selftests_sha1},
@code{cipher/@/sha256.c:@/selftests_sha224},
@code{cipher/@/sha256.c:@/selftests_sha256})
@item SHA-384
@item SHA-512
@enumerate
@item
A known answer test using a 112 byte string is run.
@item 
A known answer test using a string of one million letters "a" is run.
@end enumerate
(@code{cipher/@/sha512.c:@/selftests_sha384},
@code{cipher/@/sha512.c:@/selftests_sha512})
@end table

@subsection MAC Algorithm Tests

The following MAC algorithm tests are run in addition to the power-up
tests:

@table @asis
@item HMAC SHA-1
@enumerate
@item
A known answer test using 9 byte of data and a 20 byte key is run.
@item
A known answer test using 9 byte of data and a 100 byte key is run.
@item
A known answer test using 9 byte of data and a 49 byte key is run.
@end enumerate
(@code{cipher/hmac-tests.c:selftests_sha1})
@item HMAC SHA-224
@itemx HMAC SHA-256
@itemx HMAC SHA-384
@itemx HMAC SHA-512
@enumerate
@item
A known answer test using 9 byte of data and a 20 byte key is run.
@item
A known answer test using 50 byte of data and a 20 byte key is run.
@item
A known answer test using 50 byte of data and a 26 byte key is run.
@item
A known answer test using 54 byte of data and a 131 byte key is run.
@item
A known answer test using 152 byte of data and a 131 byte key is run.
@end enumerate
(@code{cipher/@/hmac-tests.c:@/selftests_sha224},
@code{cipher/@/hmac-tests.c:@/selftests_sha256},
@code{cipher/@/hmac-tests.c:@/selftests_sha384},
@code{cipher/@/hmac-tests.c:@/selftests_sha512})
@end table


@c ********************************************
@node FIPS Mode
@appendix Description of the FIPS Mode

This appendix gives detailed information pertaining to the FIPS mode.
In particular, the changes to the standard mode and the finite state
machine are described.  The self-tests required in this mode are
described in the appendix on self-tests.

@c -------------------------------
@section Restrictions in FIPS Mode

@noindent
If Libgcrypt is used in FIPS mode these restrictions are effective:

@itemize
@item
The cryptographic algorithms are restricted to this list:

@table @asis
@item GCRY_CIPHER_3DES
3 key EDE Triple-DES symmetric encryption.
@item GCRY_CIPHER_AES128
AES 128 bit symmetric encryption.
@item GCRY_CIPHER_AES192
AES 192 bit symmetric encryption.
@item GCRY_CIPHER_AES256
AES 256 bit symmetric encryption.
@item GCRY_MD_SHA1
SHA-1 message digest.
@item GCRY_MD_SHA224
SHA-224 message digest.
@item GCRY_MD_SHA256
SHA-256 message digest.
@item GCRY_MD_SHA384
SHA-384 message digest.
@item GCRY_MD_SHA512
SHA-512 message digest.
@item GCRY_MD_SHA1,GCRY_MD_FLAG_HMAC
HMAC using a SHA-1 message digest.
@item GCRY_MD_SHA224,GCRY_MD_FLAG_HMAC
HMAC using a SHA-224 message digest.
@item GCRY_MD_SHA256,GCRY_MD_FLAG_HMAC
HMAC using a SHA-256 message digest.
@item GCRY_MD_SHA384,GCRY_MD_FLAG_HMAC
HMAC using a SHA-384 message digest.
@item GCRY_MD_SHA512,GCRY_MD_FLAG_HMAC
HMAC using a SHA-512 message digest.
@item GCRY_PK_RSA
RSA encryption and signing.         
@item GCRY_PK_DSA
DSA signing.
@end table

Note that the CRC algorithms are not considered cryptographic algorithms
and thus are in addition available.

@item
RSA key generation refuses to create a key with a keysize of
less than 1024 bits.  

@item
DSA key generation refuses to create a key with a keysize other
than 1024 bits.

@item
The @code{transient-key} flag for RSA and DSA key generation is ignored.

@item
Support for the VIA Padlock engine is disabled.

@item 
FIPS mode may only be used on systems with a /dev/random device.
Switching into FIPS mode on other systems will fail at runtime.

@item
Saving and loading a random seed file is ignored.

@item
An X9.31 style random number generator is used in place of the
large-pool-CSPRNG generator.

@item
The command @code{GCRYCTL_ENABLE_QUICK_RANDOM} is ignored.

@item
The Alternative Public Key Interface (@code{gcry_ac_xxx}) is not
supported and all API calls return an error.

@item
Registration of external modules is not supported.

@item 
Message digest debugging is disabled.

@item
All debug output related to cryptographic data is suppressed.

@item 
On-the-fly self-tests are not performed, instead self-tests are run
before entering operational state.

@item
The function @code{gcry_set_allocation_handler} may not be used.  If
it is used Libgcrypt disables FIPS mode unless Enforced FIPS mode is
enabled, in which case Libgcrypt will enter the error state.

@item
The digest algorithm MD5 may not be used.  If it is used Libgcrypt
disables FIPS mode unless Enforced FIPS mode is enabled, in which case
Libgcrypt will enter the error state.

@item 
In Enforced FIPS mode the command @code{GCRYCTL_DISABLE_SECMEM} is
ignored.  In standard FIPS mode it disables FIPS mode.

@item
A handler set by @code{gcry_set_outofcore_handler} is ignored.
@item
A handler set by @code{gcry_set_fatalerror_handler} is ignored.

@end itemize

Note that when we speak about disabling FIPS mode, it merely means
that the function @code{gcry_fips_mode_active} returns false; it does
not mean that any non FIPS algorithms are allowed.

@c ********************************************
@section FIPS Finite State Machine

The FIPS mode of libgcrypt implements a finite state machine (FSM) using
8 states (@pxref{tbl:fips-states}) and checks at runtime that only valid
transitions (@pxref{tbl:fips-state-transitions}) may happen.

@float Figure,fig:fips-fsm
@caption{FIPS mode state diagram}
@center @image{fips-fsm,150mm,,FIPS FSM Diagram}
@end float

@float Table,tbl:fips-states
@caption{FIPS mode states}
@noindent
States used by the FIPS FSM:
@table @asis

@item Power-Off 
Libgcrypt is not runtime linked to another application.  This usually
means that the library is not loaded into main memory.  This state is
documentation only.

@item Power-On
Libgcrypt is loaded into memory and API calls may be made.  Compiler
introducted constructor functions may be run.  Note that Libgcrypt does
not implement any arbitrary constructor functions to be called by the
operating system

@item Init
The Libgcrypt initialization functions are performed and the library has
not yet run any self-test.

@item Self-Test
Libgcrypt is performing self-tests.               

@item Operational
Libgcrypt is in the operational state and all interfaces may be used.

@item Error
Libgrypt is in the error state.  When calling any FIPS relevant
interfaces they either return an error (@code{GPG_ERR_NOT_OPERATIONAL})
or put Libgcrypt into the Fatal-Error state and won't return.  

@item Fatal-Error
Libgcrypt is in a non-recoverable error state and 
will automatically transit into the  Shutdown state.        

@item Shutdown
Libgcrypt is about to be terminated and removed from the memory. The
application may at this point still runing cleanup handlers.

@end table
@end float


@float Table,tbl:fips-state-transitions
@caption{FIPS mode state transitions}
@noindent
The valid state transitions (@pxref{fig:fips-fsm}) are:
@table @code
@item 1 
Power-Off to Power-On is implicitly done by the OS loading Libgcrypt as
a shared library and having it linked to an application.

@item 2
Power-On to Init is triggered by the application calling the
Libgcrypt intialization function @code{gcry_check_version}.

@item 3
Init to Self-Test is either triggred by a dedicated API call or implicit
by invoking a libgrypt service conrolled by the FSM.

@item 4 
Self-Test to Operational is triggered after all self-tests passed
successfully.  

@item 5
Operational to Shutdown is an artifical state without any direct action
in Libgcrypt.  When reaching the Shutdown state the library is
deinitialized and can't return to any other state again.

@item 6
Shutdown to Power-off is the process of removing Libgcrypt from the
computer's memory.  For obvious reasons the Power-Off state can't be
represented within Libgcrypt and thus this transition is for
documentation only.

@item 7
Operational to Error is triggered if Libgcrypt detected an application
error which can't be returned to the caller but still allows Libgcrypt
to properly run.  In the Error state all FIPS relevant interfaces return
an error code.

@item 8
Error to Shutdown is similar to the Operational to Shutdown transition
(5).

@item 9
Error to Fatal-Error is triggred if Libgrypt detects an fatal error
while already being in Error state.

@item 10
Fatal-Error to Shutdown is automatically entered by Libgcrypt 
after having reported the error.

@item 11
Power-On to Shutdown is an artifical state to document that Libgcrypt
has not ye been initializaed but the process is about to terminate.

@item 12
Power-On to Fatal-Error will be triggerd if certain Libgcrypt functions
are used without having reached the Init state.

@item 13
Self-Test to Fatal-Error is triggred by severe errors in Libgcrypt while
running self-tests.

@item 14
Self-Test to Error is triggred by a failed self-test.

@item 15
Operational to Fatal-Error is triggered if Libcrypt encountered a
non-recoverable error.

@item 16
Operational to Self-Test is triggred if the application requested to run
the self-tests again.

@item 17
Error to Self-Test is triggered if the application has requested to run
self-tests to get to get back into operational state after an error.

@item 18
Init to Error is triggered by errors in the initialization code.

@item 19
Init to Fatal-Error is triggered by non-recoverable errors in the
initialization code.

@item 20
Error to Error is triggered by errors while already in the Error
state.


@end table
@end float

@c ********************************************
@section FIPS Miscellaneous Information

Libgcrypt does not do any key management on itself; the application
needs to care about it.  Keys which are passed to Libgcrypt should be
allocated in secure memory as available with the functions
@code{gcry_malloc_secure} and @code{gcry_calloc_secure}.  By calling
@code{gcry_free} on this memory, the memory and thus the keys are
overwritten with zero bytes before releasing the memory.

For use with the random number generator, Libgcrypt generates 3
internal keys which are stored in the encryption contexts used by the
RNG.  These keys are stored in secure memory for the lifetime of the
process.  Application are required to use @code{GCRYCTL_TERM_SECMEM}
before process termination.  This will zero out the entire secure
memory and thus also the encryption contexts with these keys.



@c **********************************************************
@c *************  Appendices (license etc.)  ****************
@c **********************************************************
@include lgpl.texi

@include gpl.texi

@node Figures and Tables
@unnumbered List of Figures and Tables

@listoffloats Figure

@listoffloats Table

@node Concept Index
@unnumbered Concept Index

@printindex cp

@node Function and Data Index
@unnumbered Function and Data Index

@printindex fn



@bye

GCRYCTL_SET_RANDOM_DAEMON_SOCKET
GCRYCTL_USE_RANDOM_DAEMON
The random damon is still a bit experimental, thus we do not document
them.  Note that they should be used during initialization and that
these functions are not really thread safe.




@c  LocalWords:  int HD