[RFCs/IDs] [Plain Text] [Tracker] [WG] [Email] [Diff1] [Diff2] [Nits]
Versions: 00 01 02 03 04 05 06 07 RFC 3961
INTERNET DRAFT K. Raeburn
Kerberos Working Group MIT
Document: draft-ietf-krb-wg-crypto-01.txt May 3, 2002
expires November 3, 2002
Encryption and Checksum Specifications
for Kerberos 5
Abstract
This document describes a framework for defining encryption and
checksum mechanisms for use with the Kerberos protocol [Kerb],
defining an abstraction layer between the Kerberos protocol and
related protocols, and the actual mechanisms themselves. Several
mechanisms are also defined in this document. Some are taken from
RFC 1510, modified in form to fit this new framework, and
occasionally modified in content when the old specification was
incorrect. New mechanisms are presented here as well. This document
does NOT indicate which mechanisms may be considered "required to
implement" or deprecated.
Comments should be sent to the editor, or to the IETF Kerberos
working group (ietf-krb-wg@anl.gov).
Status
This document is an Internet-Draft and is in full conformance with
all provisions of Section 10 of RFC2026 [RFC2026]. Internet-Drafts
are working documents of the Internet Engineering Task Force (IETF),
its areas, and its working groups. Note that other groups may also
distribute working documents as Internet-Drafts. Internet-Drafts are
draft documents valid for a maximum of six months and may be updated,
replaced, or obsoleted by other documents at any time. It is
inappropriate to use Internet-Drafts as reference material or to cite
them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.html.
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
Raeburn [Page 1]
INTERNET DRAFT May 2002
Table of Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . 2
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1. Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Encryption algorithm profile . . . . . . . . . . . . . . . . . . 4
3. Checksum mechanism attributes . . . . . . . . . . . . . . . . . . 9
4. Simplified profile for CBC ciphers with key derivation . . . . . 10
4.1. A key derivation function . . . . . . . . . . . . . . . . . . . 10
4.2. Simplified profile parameters . . . . . . . . . . . . . . . . . 12
4.3. Cryptosystem profile based on simplified profile . . . . . . . 13
4.4. Checksum profiles based on simplified profile . . . . . . . . . 14
5. Profiles for Kerberos encryption systems . . . . . . . . . . . . 15
5.1. null . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.2. DES-based encryption systems . . . . . . . . . . . . . . . . . 16
5.3. Triple-DES Encryption with Key Derivation . . . . . . . . . . . 22
6. Profiles for Kerberos checksums . . . . . . . . . . . . . . . . . 24
6.1. RSA MD4- and MD5-Based Checksums . . . . . . . . . . . . . . . 24
6.2. CRC-32 Checksum . . . . . . . . . . . . . . . . . . . . . . . . 27
6.3. DES MAC checksums . . . . . . . . . . . . . . . . . . . . . . . 28
6.4. HMAC-SHA1-DES3-KD Checksum . . . . . . . . . . . . . . . . . . 29
7. Use of Kerberos encryption outside this specification . . . . . . 29
8. Assigned Numbers . . . . . . . . . . . . . . . . . . . . . . . . 30
9. Implementation Notes . . . . . . . . . . . . . . . . . . . . . . 32
10. Security Considerations . . . . . . . . . . . . . . . . . . . . 32
11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 33
12. Editor's address . . . . . . . . . . . . . . . . . . . . . . . . 34
13. Full Copyright Statement . . . . . . . . . . . . . . . . . . . . 34
A. Test vectors . . . . . . . . . . . . . . . . . . . . . . . . . . 35
A.1. n-fold . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
A.2. mit_des_string_to_key . . . . . . . . . . . . . . . . . . . . . 35
A.3. DES3 DR and DK . . . . . . . . . . . . . . . . . . . . . . . . 37
A.4. DES3string_to_key . . . . . . . . . . . . . . . . . . . . . . . 38
A.5. Modified CRC-32 . . . . . . . . . . . . . . . . . . . . . . . . 39
B. Significant Changes from RFC 1510 . . . . . . . . . . . . . . . . 39
C. Document History (delete before publication) . . . . . . . . . . 40
Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Normative References . . . . . . . . . . . . . . . . . . . . . . . . 43
Informative References . . . . . . . . . . . . . . . . . . . . . . . 44
Raeburn [Page 2]
INTERNET DRAFT May 2002
Introduction
The Kerberos protocols are designed to encrypt messages of arbitrary
sizes, using block encryption ciphers, or less commonly, stream
encryption ciphers. Encryption is used to prove the identities of
the network entities participating in message exchanges. However,
nothing in the Kerberos protocol requires any specific encryption
algorithm be used, as long as certain operations are available in the
algorithm that is used.
The following sections specify the encryption and checksum mechanisms
currently defined for Kerberos, as well as a framework for defining
future mechanisms. The encoding, chaining, padding and other
requirements for each are described. Test vectors for several
functions are given in appendix A.
1. Concepts
Both encryption and checksum mechanisms are defined in terms of
profiles, detailed in later sections. Each specifies a collection of
operations and attributes that must be defined for a mechanism. A
Kerberos encryption or checksum mechanism specification is not
complete if it does not define all of these operations and
attributes.
An encryption mechanism must provide for confidentiality and
integrity of the original plaintext. (Integrity checking may be
achieved by incorporating a checksum, if the encryption mode does not
provide an integrity check itself.) It must also provide non-
malleability [Bellare98, Dolev91]. Use of a random confounder
prepended to the plaintext is recommended. It should not be possible
to determine if two ciphertexts correspond to the same plaintext,
without knowledge of the key.
A checksum mechanism [1] must provide proof of the integrity of the
associated message, and must preserve the confidentiality of the
message in case it is not sent in the clear. It should be infeasible
to find two plaintexts which have the same checksum. It is NOT
required that an eavesdropper be unable to determine if two checksums
are for the same message; it is assumed that the messages themselves
will be visible to any such eavesdropper.
Due to advances in cryptography, it is considered unwise by some
cryptographers to use the same key for multiple purposes. Since keys
are used in performing a number of different functions in Kerberos,
it is desirable to use different keys for each of these purposes,
even though we start with a single long-term or session key.
Raeburn [Page 3]
INTERNET DRAFT May 2002
We do this by enumerating the different uses of keys within Kerberos,
and making the "usage number" an input to the encryption or checksum
mechanisms; this enumeration is outside the scope of this document.
Later sections of this document define simplified profile templates
for encryption and checksum mechanisms that use a key derivation
function applied to a CBC mode (or similar) cipher and a checksum or
hash algorithm.
We distinguish the "base key" specified by other documents from the
"specific key" to be used for a particular instance of encryption or
checksum operations. It is expected but not required that the
specific key will be one or more separate keys derived from the
original protocol key and the key usage number. The specific key
should not be explicitly referenced outside of this document. The
typical language used in other documents should be something like,
"encrypt this octet string using this key and this usage number";
generation of the specific key and cipher state (described in the
next section) are implicit. The creation of a new cipher-state
object, or the re-use of one from a previous encryption operation,
may also be explicit.
New protocols defined in terms of the Kerberos encryption and
checksum types should use their own key usage values. Key usages are
unsigned 32 bit integers; zero is not permitted.
All data is assumed to be in the form of strings of octets or 8-bit
bytes. Environments with other byte sizes will have to emulate this
behavior in order to get correct results.
2. Encryption algorithm profile
An encryption mechanism profile must define the following attributes
and operations. The operations must be defined as functions in the
mathematical sense: no additional or implicit inputs (such as
Kerberos principal names or message sequence numbers) are permitted.
protocol key format
This describes what octet string values represent valid keys. For
encryption mechanisms that don't have perfectly dense key spaces,
this will describe the representation used for encoding keys. It
need not describe specific values that are not valid or desirable
for use; such values should be avoid by all key generation
routines.
specific key structure
This is not a protocol format at all, but a description of the
keying material derived from the chosen key and used to encrypt or
decrypt data or compute or verify a checksum. It may, for
Raeburn [Page 4]
INTERNET DRAFT May 2002
example, be a single key, a set of keys, or a combination of the
original key with additional data. The authors recommend using
one or more keys derived from the original key via one-way
functions.
required checksum mechanism
This indicates a checksum mechanism that must be available when
this encryption mechanism is used. Since Kerberos has no built in
mechanism for negotiating checksum mechanisms, once an encryption
mechanism has been decided upon, the corresponding checksum
mechanism can simply be used.
key-generation seed length, K
This is the length of the random bitstring needed to generate a
key with the encryption scheme's random-to-key function (described
below). This must be a fixed value so that various techniques for
producing a random bitstring of a given length may be used with
key generation functions.
key generation functions
Keys must be generated in a number of cases, from different types
of inputs. All function specifications must indicate how to
generate keys in the proper wire format, and must avoid generation
of keys that significantly compromise the confidentiality of
encrypted data, if the cryptosystem has such. Entropy from each
source should be preserved as much as possible. Many of the
inputs, while unknown, may be at least partly predictable (e.g., a
password string is likely to be entirely in the ASCII subset and
of fairly short length in many environments; a semi-random string
may include timestamps); the benefit of such predictability to an
attacker must be minimized.
string-to-key (UTF-8 string, UTF-8 string, opaque)->(protocol-key)
This function generates a key from two UTF-8 strings and an
opaque octet string. One of the strings is normally the
principal's pass phrase, but is in general merely a secret
string. The other string is a "salt" string intended to
produce different keys from the same password for different
users or realms. While the strings provided will use UTF-8
encoding, no specific version of Unicode should be assumed; all
valid UTF-8 strings should be allowed.
The third argument, the octet string, may be used to pass
mechanism-specific parameters in to this function. Since doing
so implies knowledge of the specific encryption system, it is
intended that generating non-default parameter values be an
uncommon operation, and that normal Kerberos applications be
able to treat this parameter block as an opaque object supplied
Raeburn [Page 5]
INTERNET DRAFT May 2002
by the KDC or defaulted to some mechanism-specific constant
value.
This should be a one-way function, so that compromising a
user's key in one realm does not compromise the user's key in
another realm, even if the same password (but a different salt)
is used.
random-to-key (bitstring[K])->(protocol-key)
This function generates a key from a random bit string of a
specific size. It may be assumed that all the bits of the
input string are equally random, even though the entropy
present in the random source may be limited.
key-derivation (protocol-key, integer)->(specific-key)
In this function, the integer input is the key usage value as
described above; the usage values must be assumed to be known
to an attacker. The specific-key output value was described in
section 1.
string-to-key parameter format
This describes the format of the block of data that can be passed
to the string-to-key function above to configure additional
parameters for that function. Along with the mechanism of
encoding parameter values, bounds on the allowed parameters should
also be described to avoid allowing a spoofed KDC to compromise
the user's password. It may be desirable to construct the
encoding such that values weakening the resulting key unacceptably
cannot be encoded, if practical.
Tighter bounds might be permitted by local security policy, or to
avoid excess resource consumption; if so, recommended defaults for
those bounds should be given in the specification. The
description should also outline possible weaknesses that may be
caused by not applying bounds checks or other validation to a
parameter string received from the network.
As emntioned above, this should be considered opaque to most
normal applications.
default string-to-key parameters (octet string)
This default value for the "params" argument to the string-to-key
function is to be used when the application protocol (Kerberos or
otherwise) does not explicitly set the parameter value. As
indicated above, this parameter block should be treated as an
opaque object in most cases.
Raeburn [Page 6]
INTERNET DRAFT May 2002
cipher state
This describes any information that can be carried over from one
encryption or decryption operation to the next, for use in
conjunction with a given specific key. For example, a block
cipher used in CBC mode may put an initial vector of one block in
the cipher state.
This state must be non-empty, and must influence encryption so as
to require that messages be decrypted in the same order they were
encrypted, if the cipher state is carried over from one encryption
to the next. Distinguishing out-of-order or missing messages from
corrupted messages is not required; if desired, this can be done
at a higher level by including sequence numbers and not "chaining"
the cipher state between encryption operations.
The cipher state may not be reused in multiple encryption or
decryption operations; these operations all generate a new cipher
state that may be used for following operations using the same key
and operation.
The contents of the cipher state must be treated as opaque outside
of encryption system specifications.
initial cipher state (specific-key, direction)->(state)
This describes the generation of the initial value for the cipher
state if it is not being carried over from a previous encryption
or decryption operation.
This describes any initial state setup needed before encrypting
arbitrary amounts of data with a given specific key; the specific
key and the direction of operations to be performed (encrypt
versus decrypt) must be the only input needed for this
initialization.
This state should be treated as opaque in any uses outside of an
encryption algorithm definition.
IMPLEMENTATION NOTE: [Kerb1510] was vague on whether and to what
degree an application protocol could exercise control over the
initial vector used in DES CBC operations. Some existing
implementations permit the setting of the initial vector. This
new specification does not permit application control of the
cipher state (beyond "initialize" and "carry over from previous
encryption"), since the form and content of the initial cipher
state can vary between encryption systems, and may not always be a
single block of random data.
New Kerberos application protocols should not assume that they can
Raeburn [Page 7]
INTERNET DRAFT May 2002
control the initial vector, or that one even exists. However, a
general-purpose implementation may wish to provide the capability,
in case applications explicitly setting it are encountered.
encrypt (specific-key, state, octet string)->(state, octet string)
This function takes the specific key, cipher state, and plaintext
as input, and generates ciphertext and a new cipher state as
outputs. If the basic encryption algorithm itself does not
provide for integrity protection (as DES in CBC mode does not do),
then some form of MAC or checksum must be included that can be
verified by the receiver. Some random factor such as a confounder
should be included so that an observer cannot know if two messages
contain the same plaintext, even if the cipher state and specific
keys are the same. The exact length of the plaintext need not be
encoded, but if it is not and if padding is required, the padding
must be added at the end of the string so that the decrypted
version may be parsed from the beginning.
The specification of the encryption function must not only
indicate the precise contents of the output octet string, but also
the output cipher state, if that state is not empty. The
application protocol may carry forward the output cipher state
from one encryption with a given specific key to another; the
effect of this "chaining" must be defined.
Assuming correctly-produced values for the specific key and cipher
state, no input octet string may result in an error indication.
decrypt (specific-key, state, octet string)->(state, octet string)
This function takes the specific key, cipher state, and ciphertext
as inputs, and verifies the integrity of the supplied ciphertext.
If the ciphertext's integrity is intact, this function produces
the plaintext and a new cipher state as outputs; otherwise, an
error indication must be returned, and the data discarded.
The result of the decryption may be longer than the original
plaintext, for example if the encryption mode adds padding to
reach a multiple of a block size. If this is the case, any extra
octets must be after the decoded plaintext. An application
protocol which needs to know the exact length of the message must
encode a length or recognizable "end of message" marker within the
plaintext. [2]
As with the encryption function, a correct specification for this
function must indicate not only the contents of the output octet
string, but also the resulting cipher state.
These operations and attributes are all that should be required to
Raeburn [Page 8]
INTERNET DRAFT May 2002
support Kerberos and various proposed preauthentication schemes.
A document defining a new encryption type should also describe known
weaknesses or attacks, so that its security may be fairly assessed,
and should include test vectors or other validation procedures for
the operations defined. Specific references to information readily
available elsewhere are sufficient.
3. Checksum mechanism attributes
A checksum mechanism profile must define the following attributes and
operations:
associated encryption algorithm(s)
This essentially indicates the type of encryption key this
checksum mechanism can be used with. A single checksum mechanism
may have more than one associated encryption algorithm if they
share the same wire key format, string-to-key function, and key
derivation function. (This combination means that, for example, a
checksum type and password are adequate to get the specific key
used to compute a checksum.)
get_mic function
This function generates a MIC token for a given specific key (see
section 2), and message (represented as an octet string), that may
be used to verify the integrity of the associated message. This
function is not required to return the same deterministic result
on every use; it need only generate a token that the verify_mic
routine can check.
The output of this function will also dictate the size of the
checksum.
verify_mic function
Given a specific key, message, and MIC token, this function
ascertains whether the message integrity has been compromised.
For a deterministic get_mic routine, the corresponding verify_mic
may simply generate another checksum and compare them.
The get_mic and verify_mic operations must be able to handle inputs
of arbitrary length; if any padding is needed, the padding scheme
must be specified as part of these functions.
These operations and attributes are all that should be required to
support Kerberos and various proposed preauthentication schemes.
As with encryption mechanism definition documents, documents defining
new checksum mechanisms should indicate validation processes and
Raeburn [Page 9]
INTERNET DRAFT May 2002
known weaknesses.
4. Simplified profile for CBC ciphers with key derivation
The profile outlines in sections 2 and 3 describes a large number of
operations that must be defined for encryption and checksum
algorithms to be used with Kerberos. We describe here a simpler
profile from which both encryption and checksum mechanism definitions
can be generated, filling in uses of key derivation in appropriate
places, providing integrity protection, and defining multiple
operations for the cryptosystem profile based on a smaller set of
operations given in the simplified profile. Not all of the existing
cryptosystems for Kerberos fit into this simplified profile, but we
recommend that future cryptosystems use it or something based on it.
[3]
Not all of the operations in the complete profiles are defined
through this mechanism; several must still be defined for each new
algorithm pair.
4.1. A key derivation function
Rather than define some scheme by which a "protocol key" is composed
of a large number of encryption keys, we use keys derived from a base
key to perform cryptographic operations. The base key must be used
only for generating the derived keys, and this derivation must be
non-invertible and entropy-preserving. Given these restrictions,
compromise of one derived key does not compromise the other subkeys.
Attack of the base key is limited, since it is only used for
derivation, and is not exposed to any user data.
Since the derived key has as much entropy as the base keys (if the
cryptosystem is good), password-derived keys have the full benefit of
all the entropy in the password.
To generate a derived key from a base key, we generate a pseudorandom
octet string, using an algorithm DR described below, and generate a
key from that octet string using a function dependent on the
encryption algorithm; the input length needed for that function,
which is also dependent on the encryption algorithm, dictates the
length of the string to be generated by the DR algorithm (the value
"k" below).
Derived Key = DK(Base Key, Well-Known Constant)
DK(Key, Constant) = random-to-key(DR(Key, Constant))
DR(Key, Constant) = k-truncate(E(Key, Constant))
Raeburn [Page 10]
INTERNET DRAFT May 2002
Here DR is the random-octet generation function described below, and
DK is the key-derivation function produced from it. In this
construction, E(Key, Plaintext) is a block cipher, Constant is a
well-known constant determined by the specific usage of this
function, and k-truncate truncates its argument by taking the first k
bits. Here, k is the key generation seed length needed for the
encryption system.
The output of the DR function is a string of bits; the actual key is
produced by applying the cryptosystem's random-to-key operation on
this bitstring.
If the output of E is shorter than k bits, then some entropy in the
key will be lost. If the Constant is smaller than the block size of
E, then it must be padded so it may be encrypted.
In either of these situations, a variation of the above construction
is used, where the folded Constant is encrypted, and the resulting
output is fed back into the encryption as necessary (the | indicates
concatentation):
K1 = E(Key, n-fold(Constant))
K2 = E(Key, K1)
K3 = E(Key, K2)
K4 = ...
DR(Key, Constant) = k-truncate(K1 | K2 | K3 | K4 ...)
n-fold is an algorithm which takes m input bits and ``stretches''
them to form n output bits with equal contribution from each input
bit to the output, as described in [Blumenthal96]:
We first define a primitive called n-folding, which takes a
variable-length input block and produces a fixed-length output
sequence. The intent is to give each input bit approximately
equal weight in determining the value of each output bit. Note
that whenever we need to treat a string of octets as a number, the
assumed representation is Big-Endian -- Most Significant Byte
first.
To n-fold a number X, replicate the input value to a length that
is the least common multiple of n and the length of X. Before
each repetition, the input is rotated to the right by 13 bit
positions. The successive n-bit chunks are added together using
1's-complement addition (that is, with end-around carry) to yield
a n-bit result....
Raeburn [Page 11]
INTERNET DRAFT May 2002
Test vectors for n-fold are supplied in Appendix A. [4]
In this document, n-fold is always used to produce n bits of output,
where n is the block size of E.
The size of the Constant must not be larger than the block size of E,
because reducing the length of the Constant by n-folding can cause
collisions.
If the size of the Constant is smaller than the block size of E, then
the Constant must be n-folded to the block size of E. This string is
used as input to E. If the block size of E is less than the key
size, then the output from E is taken as input to a second invocation
of E. This process is repeated until the number of bits accumulated
is greater than or equal to the key size of E. When enough bits have
been computed, the first k are taken as the derived key.
Since the derived key is the result of one or more encryptions in the
base key, deriving the base key from the derived key is equivalent to
determining the key from a very small number of plaintext/ciphertext
pairs. Thus, this construction is as strong as the cryptosystem
itself.
4.2. Simplified profile parameters
These are the operations and attributes that must be defined:
protocol key format
string-to-key function
default string-to-key parameters
key-generation seed length, k
random-to-key function
As above for the normal encryption mechanism profile.
unkeyed hash algorithm, H
This should be a collision-resistant hash algorithm with fixed-
size output, suitable for use in an HMAC [HMAC]. It must support
inputs of arbitrary length.
encryption block size, n
encryption/decryption functions, E and D
These are basic encryption and decryption functions for messages
of sizes that are multiples of the block size. No integrity
checking or confounder should be included here. They take as
input the IV or similar data, a protocol-format key, and a octet
string, returning a new IV and octet string.
The encryption function is not required to use CBC mode, but is
Raeburn [Page 12]
INTERNET DRAFT May 2002
assumed to be using something with similar properties. In
particular, prepending a one-block confounder to the plaintext
should alter the entire ciphertext (comparable to choosing and
including a random initial vector for CBC mode).
While there are still a number of properties to specify, they are
fewer and simpler than in the full profile.
4.3. Cryptosystem profile based on simplified profile
The above key derivation function is used to produce three
intermediate keys. One is used for computing checksums of
unencrypted data. The other two are used for encrypting and
checksumming plaintext to be sent encrypted.
The ciphertext output is the concatenation of the output of the basic
encryption function E and an HMAC using the specified hash function
H, both applied to the plaintext with a one-block random confounder
prefix and sufficient padding to bring it to a multiple of the
encryption algorithm's block size. Decryption is performed by
stripping off the HMAC, decrypting the remainder, and verifying the
HMAC. The cipher state is an initial vector, initialized to zero.
cryptosystem from simplified profile
---------------------------------------------------------------------------------
protocol key format As given.
specific key structure Three protocol-format keys: { Kc, Ke, Ki }.
key-generation seed As given.
length
required checksum The checksum mechanism defined by the
mechanism simplified checksum profile given later.
cipher state CBC initial vector (one block), initialized
to all zero.
encryption function
C1 = E(Ke, conf | plaintext | pad, oldstate.ivec)
H1 = HMAC(Ki, conf | plaintext | pad)
ciphertext = C1 | H1
newstate.ivec = last block of C1
Raeburn [Page 13]
INTERNET DRAFT May 2002
cryptosystem from simplified profile
---------------------------------------------------------------------------------
decryption function P1 = D(Ke, C1, oldstate.ivec)
if (H1 != HMAC(Ki, P1)) report error
newstate.ivec = last block of C1
default string-to-key As given.
params
key generation functions:
string-to-key function As given.
random-to-key function As given.
key-derivation function The "well-known constant" used for the DK function is
the key usage number, expressed as four octets in big-
endian order, followed by one octet indicated below.
Kc = DK(base-key, usage | 0x99);
Ke = DK(base-key, usage | 0xAA);
Ki = DK(base-key, usage | 0x55);
4.4. Checksum profiles based on simplified profile
When an encryption system is defined using the simplified profile
given in section 4.2, a checksum algorithm may be defined for it as
follows:
checksum mechanism from simplified profile
----------------------------------------------
associated cryptosystem as defined above
get_mic HMAC(Kc, message)
verify_mic get_mic and compare
The HMAC function and key Kc are as described in section 4.3.
Raeburn [Page 14]
INTERNET DRAFT May 2002
5. Profiles for Kerberos encryption systems
These are the currently defined encryption systems for Kerberos. The
astute reader will notice that some of them do not fulfill all of the
requirements outlined above. These weaker encryption systems are
defined for backwards compatibility; newer implementations should
attempt to make use of the stronger encryption systems when possible.
The full list of current encryption type number assignments is given
in section 8.
5.1. null
If no encryption is in use, the encryption system is said to be the
NULL encryption system. In the NULL encryption system there is no
checksum, confounder or padding. The ciphertext is simply the
plaintext. The null Key is used by the null encryption system and is
zero octets in length.
This encryption system should not be used for protection of data. It
exists primarily to associate with the rsa-md5 checksum type, but may
also be useful for testing protocol implementations.
null
------------------------------------------------
protocol key format zero-length bit string
specific key structure empty
required checksum rsa-md5
mechanism
key-generation seed 0
length
cipher state none
initial cipher state none
encryption function identity
decryption function identity, no integrity
check
default string-to-key none
params
Raeburn [Page 15]
INTERNET DRAFT May 2002
null
------------------------------------------------
key generation functions:
string-to-key empty string
random-to-key empty string
key-derivation empty string
The null encryption algorithm is assigned the etype value zero (0).
5.2. DES-based encryption systems
These encryption systems encrypt information under the Data
Encryption Standard [DES77] using the cipher block chaining mode
[DESM80]. A checksum is computed as described below and placed in
the cksum field. DES blocks are 8 bytes. As a result, the data to
be encrypted (the concatenation of confounder, checksum, and message)
must be padded to an 8 byte boundary before encryption. The values
of the padding bytes are unspecified.
Plaintext and DES ciphertext are encoded as blocks of 8 octets which
are concatenated to make the 64-bit inputs for the DES algorithms.
The first octet supplies the 8 most significant bits (with the
octet's MSB used as the DES input block's MSB, etc.), the second
octet the next 8 bits, ..., and the eighth octet supplies the 8 least
significant bits.
Encryption under DES using cipher block chaining requires an
additional input in the form of an initialization vector; this vector
is specified for each encryption system, below.
The DES specifications identify some 'weak' and 'semi-weak' keys;
those keys shall not be used for encrypting messages for use in
Kerberos. Additionally, because of the way that keys are derived for
the encryption of checksums, keys shall not be used that yield 'weak'
or 'semi-weak' keys when eXclusive-ORed with the hexadecimal constant
0xF0F0F0F0F0F0F0F0.
A DES key is 8 octets of data. This consists of 56 bits of actual
key data, and 8 parity bits, one per octet. The key is encoded as a
series of 8 octets written in MSB-first order. The bits within the
key are also encoded in MSB order. For example, if the encryption
key is (B1,B2,...,B7,P1,B8,...,B14,P2,B15,...,B49,P7,B50,...,B56,P8)
where B1,B2,...,B56 are the key bits in MSB order, and P1,P2,...,P8
are the parity bits, the first octet of the key would be
Raeburn [Page 16]
INTERNET DRAFT May 2002
B1,B2,...,B7,P1 (with B1 as the most significant bit). See the
[DESM80] introduction for reference.
Encryption data format
The format for the data to be encrypted includes a one-block
confounder, a checksum, the encoded plaintext, and any necessary
padding, as described in the following diagram. The msg-seq field
contains the part of the protocol message which is to be encrypted.
+-----------+----------+---------+-----+
|confounder | checksum | msg-seq | pad |
+-----------+----------+---------+-----+
One generates a random confounder of one block, placing it in
confounder; zeroes out the checksum field (of length appropriate to
exactly hold the checksum to be computed); calculates the appropriate
checksum over the whole sequence, placing the result in check; adds
the necessary padding; then encrypts using the specified encryption
type and the appropriate key.
String to key transformation
To generate a DES key from a UTF-8 text string (password), a "salt"
is concatenated to the text string, and then padded with zero-valued
octets to a multiple of 8 octets.
This string is then fan-folded and eXclusive-ORed with itself to form
an 8 byte DES key. Before eXclusive-ORing a block, every byte is
shifted one bit to the left to leave the lowest bit zero. The key is
the "corrected" by correcting the parity on the key, and if the key
matches a 'weak' or 'semi-weak' key as described in the DES
specification, it is eXclusive-ORed with the constant
0x00000000000000F0. This key is then used to generate a DES CBC
checksum on the initial string with the salt appended. The result of
the CBC checksum is the "corrected" as described above to form the
result which is returned as the key.
Pseudocode follows:
key_correction(key) {
fixparity(key);
if (is_weak_key_key(key))
key = key XOR 0xF0;
return(key);
}
Raeburn [Page 17]
INTERNET DRAFT May 2002
mit_des_string_to_key(string,salt) {
odd = 1;
s = string | salt;
tempkey = NULL;
pad(s); /* with nulls to 8 byte boundary */
for (8byteblock in s) {
if (odd == 0) {
odd = 1;
reverse(8byteblock)
}
else odd = 0;
left shift every byte in 8byteblock one bit;
tempkey = tempkey XOR 8byteblock;
}
tempkey = key_correction(tempkey);
key = key_correction(DES-CBC-check(s,tempkey));
return(key);
}
des_string_to_key(string,salt,params) {
if (length(params) == 0)
type = 0;
else if (length(params) == 1)
type = params[0];
else
error("invalid params");
if (type == 0)
mit_des_string_to_key(string,salt);
else if (type == 1)
afs_des_string_to_key(string,salt);
else
error("invalid params");
}
The AFS string-to-key algorithm is not defined here, but a parameter
block containing a byte value of one (1) is reserved for its use.
5.2.1. DES with MD5
The des-cbc-md5 encryption mode encrypts information under DES in CBC
mode with an all-zero initial vector, with an MD5 checksum (described
in [MD5-92]) computed and placed in the checksum field.
Raeburn [Page 18]
INTERNET DRAFT May 2002
The encryption system parameters for des-cbc-md5 are:
des-cbc-md5
--------------------------------------------------------------------
protocol key format 8 bytes, parity in low bit of each
specific key structure copy of original key
required checksum rsa-md5-des
mechanism
key-generation seed 8 bytes
length
cipher state 8 bytes (CBC initial vector)
initial cipher state all-zero
encryption function des-cbc(confounder | checksum | msg | pad,
ivec=oldstate)
where
checksum = md5(confounder | 0000...
| msg | pad)
newstate = last block of des-cbc output
decryption function decrypt encrypted text and verify checksum
newstate = last block of ciphertext
default string-to-key empty string
params
key generation functions:
string-to-key des_string_to_key
random-to-key copy input, then fix parity bits (discards
low bit of each input byte)
key-derivation identity
The des-cbc-md5 encryption type is assigned the etype value three
(3).
Raeburn [Page 19]
INTERNET DRAFT May 2002
5.2.2. DES with MD4
The des-cbc-md4 encryption mode also encrypts information under DES
in CBC mode, with an all-zero initial vector. An MD4 checksum
(described in [MD4-92]) is computed and placed in the checksum field.
des-cbc-md4
--------------------------------------------------------------------
protocol key format 8 bytes, parity in low bit of each
specific key structure copy of original key
required checksum rsa-md4-des
mechanism
key-generation seed 8 bytes
length
cipher state 8 bytes (CBC initial vector)
initial cipher state all-zero
encryption function des-cbc(confounder | checksum | msg | pad,
ivec=oldstate)
where
checksum = md4(confounder | 0000...
| msg | pad)
newstate = last block of des-cbc output
decryption function decrypt encrypted text and verify checksum
newstate = last block of ciphertext
default string-to-key empty string
params
key generation functions:
string-to-key des_string_to_key
random-to-key copy input, then fix parity bits
key-derivation identity
The des-cbc-md4 encryption algorithm is assigned the etype value two
(2).
Raeburn [Page 20]
INTERNET DRAFT May 2002
5.2.3. DES with CRC
The des-cbc-crc encryption type uses DES in CBC mode, with a 4-octet
CRC-based checksum computed as described in section 6.2. Note that
this is not a standard CRC-32 checksum, but a slightly modified one.
Unless otherwise specified, the key should be used as the
initialization vector, unlike for the other Kerberos DES encryption
schemes. The other details of the encryption of this data are
identical to those for the des-cbc-md5 encryption mode.
des-cbc-crc
--------------------------------------------------------------------
protocol key format 8 bytes, parity in low bit of each
specific key structure copy of original key
required checksum rsa-md5-des
mechanism
key-generation seed 8 bytes
length
cipher state 8 bytes (CBC initial vector)
initial cipher state copy of original key
encryption function des-cbc(confounder | checksum | msg | pad,
ivec=oldstate)
where
checksum = crc(confounder | 00000000
| msg | pad)
newstate = last block of des-cbc output
decryption function decrypt encrypted text and verify checksum
newstate = last block of ciphertext
default string-to-key empty string
params
key generation functions:
string-to-key des_string_to_key
random-to-key copy input, then fix parity bits
Raeburn [Page 21]
INTERNET DRAFT May 2002
des-cbc-crc
--------------------------------------------------------------------
key-derivation identity
The des-cbc-crc encryption algorithm is assigned the etype value one
(1).
5.3. Triple-DES Encryption with Key Derivation
This encryption type is based on the Triple DES cryptosystem in
Outer-CBC mode, and the HMAC-SHA1 message authentication algorithm.
A Triple DES key is the concatenation of three DES keys as described
above for des-cbc-md5. A Triple DES key is generated from random
data by creating three DES keys from separate sequences of random
data.
Encrypted data using this type must be generated as described in
section 4.3. If the length of the input data is not a multiple of
the block size, zero octets must be used to pad the plaintext to the
next eight-octet boundary. The confounder must be eight random
octets (one block).
The simplified profile for Triple DES, with key derivation as defined
in section 4, is as follows:
des3-cbc-hmac-sha1-kd
------------------------------------------------
protocol key format 24 bytes, parity in low
bit of each
key-generation seed 21 bytes
length
hash function SHA-1
block size 8 bytes
default string-to-key none
params
encryption and triple-DES encrypt and
decryption functions decrypt, in outer-CBC
mode
Raeburn [Page 22]
INTERNET DRAFT May 2002
des3-cbc-hmac-sha1-kd
------------------------------------------------
key generation functions:
random-to-key see below
string-to-key DES3string-to-key (see
below)
The des3-cbc-hmac-sha1-kd encryption type is assigned the value
sixteen (16).
5.3.1. Triple DES Key Production (random-to-key, string-to-key)
The 168 bits of random key data are converted to a protocol key value
as follows. First, the 168 bits are divided into three groups of 56
bits, which are expanded individually into 64 bits as follows:
1 2 3 4 5 6 7 p
9 10 11 12 13 14 15 p
17 18 19 20 21 22 23 p
25 26 27 28 29 30 31 p
33 34 35 36 37 38 39 p
41 42 43 44 45 46 47 p
49 50 51 52 53 54 55 p
56 48 40 32 24 16 8 p
The "p" bits are parity bits computed over the data bits. The output
of the three expansions are concatenated to form the protocol key
value.
When the HMAC-SHA1 of a string is computed, the key is used in the
protocol key form.
The string-to-key function is used to transform UTF-8 passwords into
DES3 keys. The DES3 string-to-key function relies on the "N-fold"
algorithm and DK function, described in section 4.
The n-fold algorithm is applied to the password string concatenated
with a salt value. For 3-key triple DES, the operation will involve
a 168-fold of the input password string, to generate an intermediate
key, from which the user's long-term key will be derived with the DK
function. The DES3 string-to-key function is shown here in
pseudocode:
Raeburn [Page 23]
INTERNET DRAFT May 2002
DES3string-to-key(passwordString, salt, params)
if (params != emptyString)
error("invalid params");
s = passwordString + salt
tmpKey = random-to-key(168-fold(s))
key = DK (tmpKey, KerberosConstant)
No weak-key checking is performed. The KerberosConstant value is the
byte string {0x6b 0x65 0x72 0x62 0x65 0x72 0x6f 0x73}. These values
correspond to the ASCII encoding for the string "kerberos".
6. Profiles for Kerberos checksums
These are the checksum types currently defined for Kerberos. The
full list of current checksum type number assignments is given in
section 8.
6.1. RSA MD4- and MD5-Based Checksums
6.1.1. RSA MD4 Cryptographic Checksum Using DES
The RSA-MD4-DES checksum calculates a keyed collision-proof checksum
by prepending an 8 octet confounder before the text, applying the RSA
MD4 checksum algorithm [MD4-92], and encrypting the confounder and
the checksum using DES in cipher-block-chaining (CBC) mode using a
variant of the key, where the variant is computed by eXclusive-ORing
the key with the constant 0xF0F0F0F0F0F0F0F0. [5] The initialization
vector should be zero. The resulting checksum is 24 octets long.
This checksum is tamper-proof and believed to be collision-proof.
The DES specifications identify some weak keys' and 'semi-weak keys';
those keys shall not be used for generating RSA-MD4 checksums for use
in Kerberos.
rsa-md4-des
----------------------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-md4, des-cbc-crc
get_mic des-cbc(key XOR 0xF0F0F0F0F0F0F0F0,
conf | rsa-md4(conf | msg),
ivec=0)
verify_mic decrypt and verify rsa-md4 checksum
The rsa-md4-des checksum algorithm is assigned a checksum type number
of three (3).
Raeburn [Page 24]
INTERNET DRAFT May 2002
6.1.2. The RSA MD5 Checksum
The RSA-MD5 checksum calculates a checksum using the RSA MD5
algorithm [MD5-92]. The algorithm takes as input an input message of
arbitrary length and produces as output a 128-bit (16 octet)
checksum. RSA-MD5 is believed to be collision-proof. However, since
it is unkeyed, it must be used with caution. Currently it is used by
some implementations in places where the checksum itself is part of a
larger message that will be encrypted. Its use is not recommended.
rsa-md5
----------------------------------------------
associated cryptosystem null
get_mic rsa-md5(msg)
verify_mic get_mic and compare
The rsa-md5 checksum algorithm is assigned a checksum type number of
seven (7).
6.1.3. RSA MD5 Cryptographic Checksum Using DES
The RSA-MD5-DES checksum calculates a keyed collision-proof checksum
by prepending an 8 octet confounder before the text, applying the RSA
MD5 checksum algorithm, and encrypting the confounder and the
checksum using DES in cipher-block-chaining (CBC) mode using a
variant of the key, where the variant is computed by eXclusive-ORing
the key with the hexadecimal constant 0xF0F0F0F0F0F0F0F0. The
initialization vector should be zero. The resulting checksum is 24
octets long. This checksum is tamper-proof and believed to be
collision-proof.
The DES specifications identify some 'weak keys' and 'semi-weak
keys'; those keys shall not be used for encrypting RSA-MD5 checksums
for use in Kerberos.
rsa-md5-des
----------------------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-md4, des-cbc-crc
get_mic des-cbc(key XOR 0xF0F0F0F0F0F0F0F0,
conf | rsa-md5(conf | msg))
verify_mic decrypt and verify rsa-md5 checksum
Raeburn [Page 25]
INTERNET DRAFT May 2002
rsa-md5-des
----------------------------------------------------------------
The rsa-md5-des checksum algorithm is assigned a checksum type number
of eight (8).
6.1.4. The RSA MD4 Checksum
The RSA-MD4 checksum calculates a checksum using the RSA MD4
algorithm [MD4-92]. The algorithm takes as input an input message of
arbitrary length and produces as output a 128-bit (16 octet)
checksum. RSA-MD4 is believed to be collision-proof.
rsa-md4
------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-
md4, des-cbc-crc
get_mic md4(msg)
verify_mic compute checksum and
compare
The rsa-md4 checksum algorithm is assigned a checksum type number of
two (2).
6.1.5. RSA MD4 Cryptographic Checksum Using DES alternative
The RSA-MD4-DES-K checksum calculates a keyed collision-proof
checksum by applying the RSA MD4 checksum algorithm and encrypting
the results using DES in cipher block chaining (CBC) mode using a DES
key as both key and initialization vector. The resulting checksum is
16 octets long. This checksum is tamper-proof and believed to be
collision-proof. Note that this checksum type is the old method for
encoding the RSA-MD4-DES checksum and it is no longer recommended.
Raeburn [Page 26]
INTERNET DRAFT May 2002
rsa-md4-des-k
----------------------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-md4, des-cbc-crc
get_mic des-cbc(key, md4(msg), ivec=key)
verify_mic decrypt, compute checksum and compare
The rsa-md4-des-k checksum algorithm is assigned a checksum type
number of six (6).
6.2. CRC-32 Checksum
This CRC-32 checksum calculates a checksum based on a cyclic
redundancy check as described in ISO 3309 [CRC], modified as
described below. The resulting checksum is four (4) octets in
length. The CRC-32 is neither keyed nor collision-proof; thus, the
use of this checksum is not recommended. An attacker using a
probabilistic chosen-plaintext attack as described in [SG92] might be
able to generate an alternative message that satisfies the checksum.
The use of collision-proof checksums is recommended for environments
where such attacks represent a significant threat.
The CRC-32 checksum used in the des-cbc-crc encryption mode is
identical to the 32-bit FCS described in ISO 3309 with two
exceptions: the sum with the all-ones polynomial times x**k is
omitted, and the final remainder is not ones-complemented. ISO 3309
describes the FCS in terms of bits, while this document describes the
Kerberos protocol in terms of octets. To disambiguate the ISO 3309
definition for the purpose of computing the CRC-32 in the des-cbc-crc
encryption mode, the ordering of bits in each octet shall be assumed
to be LSB-first. Given this assumed ordering of bits within an
octet, the mapping of bits to polynomial coefficients shall be
identical to that specified in ISO 3309.
crc32
------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-
md4, des-cbc-crc
get_mic crc32(msg)
verify_mic compute checksum and
compare
Raeburn [Page 27]
INTERNET DRAFT May 2002
The crc32 checksum algorithm is assigned a checksum type number of
one (1).
6.3. DES MAC checksums
6.3.1. DES CBC checksum
The DES-MAC checksum is computed by prepending an 8 octet confounder
to the plaintext, performing a DES CBC-mode encryption on the result
using the key and an initialization vector of zero, taking the last
block of the ciphertext, prepending the same confounder and
encrypting the pair using DES in cipher-block-chaining (CBC) mode
using a a variant of the key, where the variant is computed by
eXclusive-ORing the key with the constant 0xF0F0F0F0F0F0F0F0. The
initialization vector should be zero. The resulting checksum is 128
bits (16 octets) long, 64 bits of which are redundant. This checksum
is tamper-proof and collision-proof.
The DES specifications identify some "weak" and "semiweak" keys;
those keys shall not be used for generating DES-MAC checksums for use
in Kerberos, nor shall a key be used whose variant is "weak" or
"semi-weak".
des-mac
----------------------------------------------------------------
associated des-cbc-md5, des-cbc-md4, des-cbc-crc
cryptosystem
get_mic des-cbc(key XOR 0xF0F0F0F0F0F0F0F0,
conf | des-mac(key, conf | msg, ivec=0),
ivec=0)
verify_mic decrypt, compute DES MAC using confounder,
compare
The des-mac checksum algorithm is assigned a checksum type number of
four (4).
6.3.2. DES CBC checksum alternative
The DES-MAC-K checksum is computed by performing a DES CBC-mode
encryption of the plaintext, and using the last block of the
ciphertext as the checksum value. It is keyed with an encryption key
and an initialization vector; any uses which do not specify an
additional initialization vector will use the key as both key and
Raeburn [Page 28]
INTERNET DRAFT May 2002
initialization vector. The resulting checksum is 64 bits (8 octets)
long. This checksum is tamper-proof and collision-proof. Note that
this checksum type is the old method for encoding the DESMAC checksum
and it is no longer recommended.
The DES specifications identify some "weak keys"; those keys shall
not be used for generating DES-MAC checksums for use in Kerberos.
des-mac-k
----------------------------------------------------------------
associated cryptosystem des-cbc-md5, des-cbc-md4, des-cbc-crc
get_mic des-mac(key, msg, ivec=key or given)
verify_mic compute MAC and compare
The des-mac-k checksum algorithm is assigned a checksum type number
of five (5).
6.4. HMAC-SHA1-DES3-KD Checksum
This checksum type is defined as outlined in section 4.2 above, using
the des3-hmac-sha1-kd encryption algorithm parameters from section
5.3. The checksum is thus a SHA-1 HMAC using the computed key Kc
over the message to be protected.
The hmac-sha1-des3-kd checksum algorithm is assigned a checksum type
number of twelve (12).
7. Use of Kerberos encryption outside this specification
Several Kerberos-based application protocols and preauthentication
systems have been designed and deployed that perform encryption and
message integrity checks in various ways. While in some cases there
may be good reason for specifying these protocols in terms of
specific encryption or checksum algorithms, we anticipate that in
many cases this will not be true, and more generic approaches
independent of particular algorithms will be desirable. Rather than
having each protocol designer reinvent schemes for protecting data,
using multiple keys, etc, we have attempted to present in this
section a general framework that should be sufficient not only for
the Kerberos protocol itself but also for many preauthentication
systems and application protocols, while trying to avoid some of the
assumptions that can work their way into such protocol designs.
Raeburn [Page 29]
INTERNET DRAFT May 2002
Some problematic assumptions we've seen (and sometimes made) include:
that a random bitstring is always valid as a key (not true for DES
keys with parity); that the basic block encryption chaining mode
provides no integrity checking, or can easily be separated from such
checking (not true for many modes in development that do both
simultaneously); that a checksum for a message always results in the
same value (not true if a confounder is incorporated); that an
initial vector is used (may not be true if a block cipher in CBC mode
is not in use).
Such assumptions, while they may hold for any given set of encryption
and checksum algorithms, may not be true of the next algorithms to be
defined, leaving the application protocol unable to make use of those
algorithms without updates to its specification.
The Kerberos protocol uses only the attributes and operations
described in sections 2 and 3. Preauthentication systems and
application protocols making use of Kerberos are encouraged to use
them as well. The specific key and string-to-key parameters should
generally be treated as opaque. While the string-to-key parameters
are manipulated as an octet string, the representation for the
specific key structure is implementation-defined; it may not even be
a single object.
While we don't recommend it, some application protocols will
undoubtedly continue to use the key data directly, even if only in
some of the currently existing protocol specifications. An
implementation intended to support general Kerberos applications may
therefore need to make the key data available, as well as the
attributes and operations described in sections 2 and 3. [6]
8. Assigned Numbers
The following encryption type numbers are already assigned or
reserved for use in Kerberos and related protocols.
Encryption type etype block minimum confounder section
value size pad size size
----------------------------------------------------------------------
NULL 0 1 0 0 5.1
des-cbc-crc 1 8 4 8 5.2.3
des-cbc-md4 2 8 0 8 5.2.2
des-cbc-md5 3 8 0 8 5.2.1
des3-cbc-sha1-kd 16 8 0 8 5.3
Other numbers have been reserved for use in encryption systems not
Raeburn [Page 30]
INTERNET DRAFT May 2002
defined here. Encryption type numbers are unfortunately overloaded
on occasion in Kerberos-related protocols, so some of the reserved
numbers do not and will not correspond to encryption systems fitting
the profile presented here.
Encryption type etype value comment
----------------------------------------------------------------------
[reserved] 4
des3-cbc-md5 5
[reserved] 6
des3-cbc-sha1 7
dsaWithSHA1-CmsOID 9 (pkinit)
md5WithRSAEncryption-CmsOID 10 (pkinit)
sha1WithRSAEncryption-CmsOID 11 (pkinit)
rc2CBC-EnvOID 12 (pkinit)
rsaEncryption-EnvOID 13 (pkinit from PKCS#1 v1.5)
rsaES-OAEP-ENV-OID 14 (pkinit from PKCS#1 v2.0)
des-ede3-cbc-Env-OID 15 (pkinit)
rc4-hmac 23 (swift)
rc4-hmac-exp 24 (swift)
subkey-keynaterial 65 (opaque mhur)
The following checksum type numbers are assigned or reserved. As
with encryption type numbers, some overloading of checksum numbers
has occurred.
Checksum type sumtype checksum section
value size
----------------------------------------------------------------------
CRC32 1 4 6.2
rsa-md4 2 16 6.1.4
rsa-md4-des 3 24 6.1.1
des-mac 4 16 6.3.1
des-mac-k 5 8 6.3.2
rsa-md4-des-k 6 16 6.1.5
rsa-md5 7 16 6.1.2
rsa-md5-des 8 24 6.1.3
rsa-md5-des3 9 24
hmac-sha1-des3-kd 12 20 6.4
hmac-sha1-des3 13 20
sha1 (unkeyed) 14 20
[reserved] 0x8003 ? [GSS-KRB5]
Management of these assignments is not being delegated to IANA at
Raeburn [Page 31]
INTERNET DRAFT May 2002
this time.
9. Implementation Notes
The "interface" described here is the minimal information that must
be defined to make a cryptosystem useful within Kerberos in an
interoperable fashion. It is not an attempt to define a complete API
for cryptographic functionality within Kerberos. Actual
implementations providing clean APIs will probably find it useful to
make additional information available, which should be possible to
derive from a specification written to the framework given here. For
example, an application designer may wish to determine the largest
number of bytes that can be encrypted without overflowing a certain
size output buffer, or conversely, the maximum number of bytes that
might be obtained by decrypting a given ciphertext message. (In
fact, an implementation of the GSS-API Kerberos mechanism [GSS-KRB5]
will require some of these.)
The presence of a mechanism in this document should not be taken as
an indication that it must be implemented for compliance with any
specification; required mechanisms will be specified elsewhere.
Indeed, some of the mechanisms described here for backwards
compatibility are now considered rather weak for protecting critical
data.
10. Security Considerations
Recent years have brought advancements in the ability to perform
large-scale attacks against DES, to such a degree that it is not
considered a strong encryption mechanism any longer; triple-DES is
generally preferred in its place, despite the poorer performance.
See [ESP-DES] for a summary of some of the potential attacks, and
[EFF-DES] for a detailed discussion of the implementation of
particular attack. However, most Kerberos implementations still have
DES as their primary interoperable encryption type.
DES has a small number of "weak" and "semiweak" keys, but the use of
triple-DES makes no effort to avoid them. The nature of the weak
keys is such that it is extremely unlikely that they will weaken the
triple-DES encryption -- only slightly more likely than having the
middle of the three sub-keys match one of the other two, which
effectively converts the encryption to single-DES. (Which is another
case we make no effort to avoid.)
The true CRC-32 checksum is not collision-proof; an attacker could
use a probabilistic chosen-plaintext attack to generate a valid
message even if a confounder is used [SG92]. The use of collision-
proof checksums is of course recommended for environments where such
Raeburn [Page 32]
INTERNET DRAFT May 2002
attacks represent a significant threat. The "simplifications" (read:
bugs) introduced when CRC-32 was implemented for Kerberos cause
leading zeros to effectively be ignored, so messages differing only
in leading zero bits will have the same checksum.
[HMAC] and [IPSEC-HMAC] discuss weaknesses of the HMAC algorithm.
Unlike [IPSEC-HMAC], this specification does not use the suggested
truncation of the HMAC output. As pointed out in [IPSEC-HMAC], SHA-1
was not developed to be used as a keyed hash function, which is a
criterion of HMAC. [HMAC-TEST] contains test vectors for HMAC-SHA-1.
The mit_des_string_to_key function was originally constructed with
the assumption that all input would be ASCII; it ignores the top bit
of each input byte. Folding with XOR is also not an especially good
mixing mechanism in terms of preserving randomness.
The n-fold function used in the string-to-key operation for des3-cbc-
hmac-sha1-kd was designed to cause each bit of input to contribute
equally to the output; it was not designed to maximize or equally
distribute randomness in the input, and there are conceivable cases
of partially structured input where randomness may be lost. This
should only be an issue for highly structured passwords, however.
[RFC1851] discusses the relative strength of triple-DES encryption.
The relative slow speed of triple-DES encryption may also be an issue
for some applications.
This document, like the Kerberos protocol, completely ignores the
notion of limiting the amount of data a key may be used with to a
quantity based on the robustness of the algorithm or size of the key.
It is assumed that any defined algorithms and key sizes will be
strong enough to support very large amounts of data, or they will be
deprecated once significant attacks are known.
This document also places no bounds on the amount of data that can be
handled in various operations. In order to avoid denial of service
attacks, implementations will probably want to restrict message sizes
at some higher level.
11. Acknowledgments
This document is an extension of the encryption specification
included in [Kerb1510] by B. Clifford Neuman and John Kohl, and much
of the text of the background, concepts, and DES specifications are
drawn directly from that document.
The abstract framework presented in this document was put together by
Jeff Altman, Sam Hartman, Jeff Hutzelman, Cliff Neuman, Ken Raeburn,
Raeburn [Page 33]
INTERNET DRAFT May 2002
and Tom Yu, and the details were refined several times based on
comments from John Brezak and others.
Marc Horowitz wrote the original specification of triple-DES and key
derivation in a pair of Internet Drafts (under the names draft-
horowitz-key-derivation and draft-horowitz-kerb-key-derivation) which
were later folded into a draft revision of [Kerb1510], from which
this document was later split off.
Tom Yu provided the text describing the modifications to the standard
CRC algorithm as Kerberos implementations actually use it.
Miroslav Jurisic provided information for one of the UTF-8 test cases
for the string-to-key functions.
Marcus Watts and Louis LeVay caught some errors in an earlier draft.
12. Editor's address
Kenneth Raeburn
Massachusetts Institute of Technology
77 Massachusetts Avenue
Cambridge, MA 02139
raeburn@mit.edu
13. Full Copyright Statement
Copyright (C) The Internet Society (2002). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
Raeburn [Page 34]
INTERNET DRAFT May 2002
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE."
A. Test vectors
This section provides test vectors for various functions defined or
described in this document. For convenience, most inputs are ASCII
strings, though some UTF-8 samples are be provided for string-to-key
functions. Keys and other binary data are specified as hexadecimal
strings.
A.1. n-fold
The n-fold function is defined in section 4.1. As noted there, the
sample vector in the original paper defining the algorithm appears to
be incorrect. Here are values provided by Marc Horowitz:
64-fold("012345") =
64-fold(303132333435) = be072631276b1955
56-fold("password") =
56-fold(70617373776f7264) = 78a07b6caf85fa
64-fold("Rough Consensus, and Running Code") =
64-fold(526f75676820436f6e73656e7375732c20616e642052756e
6e696e6720436f6465) = bb6ed30870b7f0e0
168-fold("password") =
168-fold(70617373776f7264) =
59e4a8ca7c0385c3c37b3f6d2000247cb6e6bd5b3e
192-fold("MASSACHVSETTS INSTITVTE OF TECHNOLOGY"
192-fold(4d41535341434856534554545320494e5354495456544520
4f4620544543484e4f4c4f4759) =
db3b0d8f0b061e603282b308a50841229ad798fab9540c1b
A.2. mit_des_string_to_key
The function mit_des_string_to_key is defined in section 5.2. We
present here several test values, with some of the intermediate
results. The fourth test demonstrates the use of UTF-8 with three
characters. The last two tests are specifically constructed so as to
trigger the weak-key fixups for the intermediate key produced by fan-
folding; we have no test cases that cause such fixups for the final
key.
Raeburn [Page 35]
INTERNET DRAFT May 2002
UTF-8 encodings used in test vector:
eszett C3 9F s-caron C5 A1 c-acute C4 87
g-clef F0 9D 84 9E
Test vector:
salt: "ATHENA.MIT.EDUraeburn"
415448454e412e4d49542e4544557261656275726e
password: "password" 70617373776f7264
fan-fold result: c01e38688ac86c2e
intermediate key: c11f38688ac86d2f
DES key: cbc22fae235298e3
salt: "WHITEHOUSE.GOVdanny" 5748495445484f5553452e474f5664616e6e79
password: "potatoe" 706f7461746f65
fan-fold result: a028944ee63c0416
intermediate key: a129944fe63d0416
DES key: df3d32a74fd92a01
salt: "EXAMPLE.COMpianist" 4558414D504C452E434F4D7069616E697374
password: g-clef f09d849e
fan-fold result: 3c4a262c18fab090
intermediate key: 3d4a262c19fbb091
DES key: 4ffb26bab0cd9413
salt: "ATHENA.MIT.EDUJuri" + s-caron + "i" + c-acute
415448454e412e4d49542e4544554a757269c5a169c487
password: eszett c39f
fan-fold result: b8f6c40e305afc9e
intermediate key: b9f7c40e315bfd9e
DES key: 62c81a5232b5e69d
salt: "AAAAAAAA" 4141414141414141
password: "11119999" 3131313139393939
fan-fold result: e0e0e0e0f0f0f0f0
intermediate key: e0e0e0e0f1f1f101
DES key: 984054d0f1a73e31
Raeburn [Page 36]
INTERNET DRAFT May 2002
salt: "FFFFAAAA" 4646464641414141
password: "NNNN6666" 4e4e4e4e36363636
fan-fold result: 1e1e1e1e0e0e0e0e
intermediate key: 1f1f1f1f0e0e0efe
DES key: c4bf6b25adf7a4f8
A.3. DES3 DR and DK
These tests show the derived-random and derived-key values for the
des3-hmac-sha1-kd encryption scheme, using the DR and DK functions
defined in section 5.3.1. The input keys were randomly generated;
the usage values are from this specification.
key: dce06b1f64c857a11c3db57c51899b2cc1791008ce973b92
usage: 0000000155
DR: 935079d14490a75c3093c4a6e8c3b049c71e6ee705
DK: 925179d04591a79b5d3192c4a7e9c289b049c71f6ee604cd
key: 5e13d31c70ef765746578531cb51c15bf11ca82c97cee9f2
usage: 00000001aa
DR: 9f58e5a047d894101c469845d67ae3c5249ed812f2
DK: 9e58e5a146d9942a101c469845d67a20e3c4259ed913f207
key: 98e6fd8a04a4b6859b75a176540b9752bad3ecd610a252bc
usage: 0000000155
DR: 12fff90c773f956d13fc2ca0d0840349dbd39908eb
DK: 13fef80d763e94ec6d13fd2ca1d085070249dad39808eabf
key: 622aec25a2fe2cad7094680b7c64940280084c1a7cec92b5
usage: 00000001aa
DR: f8debf05b097e7dc0603686aca35d91fd9a5516a70
DK: f8dfbf04b097e6d9dc0702686bcb3489d91fd9a4516b703e
key: d3f8298ccb166438dcb9b93ee5a7629286a491f838f802fb
usage: 6b65726265726f73
DR: 2270db565d2a3d64cfbfdc5305d4f778a6de42d9da
DK: 2370da575d2a3da864cebfdc5204d56df779a7df43d9da43
Raeburn [Page 37]
INTERNET DRAFT May 2002
key: b55e983467e551b3e5d0e5b6c80d45769423a873dc62b30e
usage: 636f6d62696e65
DR: 0127398bacc81a2a62bc45f8d4c151bbcdd5cb788a
DK: 0126388aadc81a1f2a62bc45f8d5c19151bacdd5cb798a3e
key: c1081649ada74362e6a1459d01dfd30d67c2234c940704da
usage: 0000000155
DR: 348056ec98fcc517171d2b4d7a9493af482d999175
DK: 348057ec98fdc48016161c2a4c7a943e92ae492c989175f7
key: 5d154af238f46713155719d55e2f1f790dd661f279a7917c
usage: 00000001aa
DR: a8818bc367dadacbe9a6c84627fb60c294b01215e5
DK: a8808ac267dada3dcbe9a7c84626fbc761c294b01315e5c1
key: 798562e049852f57dc8c343ba17f2ca1d97394efc8adc443
usage: 0000000155
DR: c813f88b3be2b2f75424ce9175fbc8483b88c8713a
DK: c813f88a3be3b334f75425ce9175fbe3c8493b89c8703b49
key: 26dce334b545292f2feab9a8701a89a4b99eb9942cecd016
usage: 00000001aa
DR: f58efc6f83f93e55e695fd252cf8fe59f7d5ba37ec
DK: f48ffd6e83f83e7354e694fd252cf83bfe58f7d5ba37ec5d
A.4. DES3string_to_key
These are the keys generated for some of the above input strings for
triple-DES with key derivation as defined in section 5.3.1.
salt: "ATHENA.MIT.EDUraeburn"
passwd: "password"
key: 850bb51358548cd05e86768c313e3bfef7511937dcf72c3e
salt: "WHITEHOUSE.GOVdanny"
passwd: "potatoe"
key: dfcd233dd0a43204ea6dc437fb15e061b02979c1f74f377a
Raeburn [Page 38]
INTERNET DRAFT May 2002
salt: "EXAMPLE.COMbuckaroo"
passwd: "penny"
key: 6d2fcdf2d6fbbc3ddcadb5da5710a23489b0d3b69d5d9d4a
salt: "ATHENA.MIT.EDUJuri" + s-caron + "i" + c-acute
passwd: eszett
key: 16d5a40e1ce3bacb61b9dce00470324c831973a7b952feb0
salt: "EXAMPLE.COMpianist"
passwd: g-clef
key: 85763726585dbc1cce6ec43e1f751f07f1c4cbb098f40b19
A.5. Modified CRC-32
Below are modified-CRC32 values for various ASCII strings. Only the
printable ASCII characters are checksummed, no C-style trailing zero-
valued octet. The 32-bit modified CRC and the sequence of output
bytes as used in Kerberos are shown. (The octet values are separated
here to emphasize that they are octet values and not 32-bit numbers,
which will be the most convenient form for manipulation in some
implementations. The bit and byte order used internally for such a
number is irrelevant; the octet sequence generated is what is
important.)
mod-crc-32("foo") = 33 bc 32 73
mod-crc-32("test0123456789") = d6 88 3e b8
mod-crc-32("MASSACHVSETTS INSTITVTE OF TECHNOLOGY") = f7 80 41 e3
mod-crc-32(8000) = 4b 98 83 3b
mod-crc-32(0008) = 32 88 db 0e
mod-crc-32(0080) = 20 83 b8 ed
mod-crc-32(80) = 20 83 b8 ed
mod-crc-32(80000000) = 3b b6 59 ed
mod-crc-32(00000001) = 96 30 07 77
B. Significant Changes from RFC 1510
The encryption and checksum mechanism profiles are new. The old
specification defined a few operations for various mechanisms, but
Raeburn [Page 39]
INTERNET DRAFT May 2002
didn't outline what should be required of new mechanisms in terms of
abstract properties, nor how to ensure that a mechanism specification
is complete enough for interoperability between implementations. The
new profiles do differ from the old specification in a few ways:
Some message definitions in [Kerb1510] could be read as permitting
the initial vector to be specified by the application; the text
was too vague. It is specifically not permitted in this
specification. Some encryption algorithms may not use
initialization vectors, so relying on chosen, secret
initialization vectors for security is unwise. Also, the
prepended confounder in the existing algorithms is roughly
equivalent to a per-message initialization vector that is revealed
in encrypted form. However, carrying state across from one
encryption to another is explicitly permitted through the opaque
"cipher state" object.
The use of key derivation is new.
Several new methods are introduced, including generation of a key
in wire-protocol format from random input data.
The means for influencing the string-to-key algorithm are laid out
more clearly.
other new methods?
Triple-DES support is new.
The des-cbc-crc and CRC descriptions have been updated to align them
with existing implementations.
[Kerb1510] had no indication what character set or encoding might be
used for pass phrases and salts.
In [Kerb1510], key types, encryption algorithms and checksum
algorithms were only loosely associated, and the association was not
well described. In this specification, key types and encryption
algorithms have a one-to-one correspondence, and associations between
encryption and checksum algorithms are described so that checksums
can be computed given negotiated keys, without requiring further
negotiation for checksum types.
C. Document History (delete before publication)
Changes from kerberos-revisions-## to crypto-00 draft:
Collected DES encryption algorithms together under one subsection.
Raeburn [Page 40]
INTERNET DRAFT May 2002
Replaced salt type number with string-to-key parameter octet-string.
Incorporated Tom Yu's description of our "modified" (i.e., broken)
CRC algorithm.
Changes from -00 to -01:
Added "Significant Changes from RFC 1510" and "Document History"
appendixes.
Fixed up some incorrect section number references.
Collected RSA checksum algorithms together under one section, and
likewise for DES MAC checksum algorithms.
Changed DES encryption description to indicate that the trailing
padding is included in the checksum calculation.
Updated introduction.
Dropped combine-keys function. It'll be needed eventually, but this
isn't the best place for it.
Added request for weaknesses, attacks, test vectors in follow-on
documents defining cipher suites, and references on weaknesses for
suites defined here.
We are not assigning the number registrations to IANA at this time.
The get_mic function need not have a fixed size output. (If it is
required, someone remind me why.) The hash function used in the
simplified profile must, on the other hand.
Cipher state description tightened up a bit: Split description into
state and initial-state. Added direction (encrypt vs decrypt) as an
input parameter for initialization, operations performed must be
consistent with that. Cipher state can be used as input to
encrypt/decrypt only once; operations output the new state to be used
next time, if chaining at all. Chaining is required to influence the
output, to detect message stream modifications; previously this was
explicitly not required.
Moved some descriptive text out of tables.
Fixed some speling probelms.
Raeburn [Page 41]
INTERNET DRAFT May 2002
Notes
[1] While Message Authentication Code (MAC) or Message Integrity
Check (MIC) would be more appropriate terms for many of the
uses in this document, we continue to use the term "checksum"
for historical reasons.
[2] In the case of Kerberos, the encrypted objects will generally
be ASN.1 DER encodings, which contain indications of their
length in the first few octets.
[3] As of the time of this writing, some new modes of operation
have been proposed, some of which may permit encryption and
integrity protection simultaneously. After some of these
proposals have been subjected to adequate analysis, we may
wish to formulate a new simplified profile based on one of
them.
[4] It should be noted that the sample vector in Appendix B.2 of
the original paper appears to be incorrect. Two independent
implementations from the specification (one in C by Marc
Horowitz, and another in Scheme by Bill Sommerfeld) agree on
a value different from that in [Blumenthal96].
[5] A variant of the key is used to limit the use of a key to a
particular function, separating the functions of generating a
checksum from other encryption performed using the session
key. The constant 0xF0F0F0F0F0F0F0F0 was chosen because it
maintains key parity. The properties of DES precluded the
use of the complement. The same constant is used for similar
purpose in the Message Integrity Check in the Privacy
Enhanced Mail standard.
[6] Perhaps one of the more common reasons for directly
performing encryption is direct control over the negotiation
and to select a "sufficiently strong" encryption algorithm
(whatever that means in the context of a given application).
While Kerberos directly provides no facility for negotiating
encryption types between the application client and server,
there are other means for accomplishing similar goals. For
example, requesting only "strong" session key types from the
KDC, and assuming that the type actually returned by the KDC
will be understood and supported by the application server.
Raeburn [Page 42]
INTERNET DRAFT May 2002
Normative References
@@ NOTE: draft-rfc-editor-rfc2223bis-01 specifies separate "normative
references" and "informative references" sections. This draft has
not yet become an RFC as of the time of this writing.
This section copied from kerberos-revisions draft. Drop the ones we
don't need, add anything new that we do need. Move informational-
only references to the next section. Update old I-D references to
RFCs, or find other sources.
[Bellare98]
Bellare, M., Desai, A., Pointcheval, D., and P. Rogaway,
"Relations Among Notions of Security for Public-Key Encryption
Schemes". Extended abstract published in Advances in Cryptology-
Crypto 98 Proceedings, Lecture Notes in Computer Science Vol.
1462, H. Krawcyzk ed., Springer-Verlag, 1998.
[Blumenthal96]
Blumenthal, U., "A Better Key Schedule for DES-Like Ciphers",
Proceedings of PRAGOCRYPT '96, 1996.
[CRC]
International Organization for Standardization, "ISO Information
Processing Systems - Data Communication - High-Level Data Link
Control Procedure - Frame Structure," IS 3309, 3rd Edition,
October 1984.
[DES77]
National Bureau of Standards, U.S. Department of Commerce, "Data
Encryption Standard," Federal Information Processing Standards
Publication 46, Washington, DC, 1977.
[DESM80]
National Bureau of Standards, U.S. Department of Commerce, "DES
Modes of Operation," Federal Information Processing Standards
Publication 81, Springfield, VA, December 1980.
[Dolev91]
Dolev, D., Dwork, C., Naor, M., "Non-malleable cryptography",
Proceedings of the 23rd Annual Symposium on Theory of Computing,
ACM, 1991.
[HMAC]
Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed-Hashing
for Message Authentication", RFC 2104, February 1997.
[MD4-92]
Rivest, R., "The MD4 Message Digest Algorithm," RFC 1320, MIT
Laboratory for Computer Science, April 1992.
[MD5-92]
Rivest, R., "The MD5 Message Digest Algorithm," RFC 1321, MIT
Laboratory for Computer Science, April 1992.
Raeburn [Page 43]
INTERNET DRAFT May 2002
[RFC2026]
Bradner, S., "The Internet Standards Process -- Revisions 3," RFC
2026, October 1996.
[SG92]
Stubblebine, S., and V. D. Gligor, "On Message Integrity in
Cryptographic Protocols," in Proceedings of the IEEE Symposium on
Research in Security and Privacy, Oakland, California, May 1992.
Informative References
[EFF-DES]
Electronic Frontier Foundation, "Cracking DES: Secrets of
Encryption Research, Wiretap Politics, and Chip Design", O'Reilly
& Associates, Inc., May 1998.
[ESP-DES]
Madson, C., and N. Doraswamy, "The ESP DES-CBC Cipher Algorithm
With Explicit IV", RFC 2405, November 1998.
[GSS-KRB5]
Linn, J., "The Kerberos Version 5 GSS-API Mechanism," RFC 1964,
June 1996.
[HMAC-TEST]
Cheng, P., and R. Glenn, "Test Cases for HMAC-MD5 and HMAC-SHA-1",
RFC 2202, September 1997.
[IPSEC-HMAC]
Madson, C., and R. Glenn, "The Use of HMAC-SHA-1-96 within ESP and
AH", RFC 2404, November 1998.
[Kerb]
Neuman, C., Kohl, J., Ts'o, T., Yu, T., Hartman, S., and K.
Raeburn, "The Kerberos Network Authentication Service (V5)",
draft-ietf-krb-wg-kerberos-clarifications-00.txt, February 22,
2002. Work in progress.
[Kerb1510]
Kohl, J., and C. Neuman, "The Kerberos Network Authentication
Service (V5)", RFC 1510, September 1993.
Raeburn [Page 44]
Html markup produced by rfcmarkup 1.70, available from
http://tools.ietf.org/tools/rfcmarkup/