Network Working Group J. Linn
Request for Comments: 1964 OpenVision Technologies
Category: Standards Track June 1996
The Kerberos Version 5 GSS-API Mechanism
Status of this Memo
This document specifies an Internet standards track protocol for the
Internet community, and requests discussion and suggestions for
improvements. Please refer to the current edition of the "Internet
Official Protocol Standards" (STD 1) for the standardization state
and status of this protocol. Distribution of this memo is unlimited.
ABSTRACT
This specification defines protocols, procedures, and conventions to
be employed by peers implementing the Generic Security Service
Application Program Interface (as specified in RFCs 1508 and 1509)
when using Kerberos Version 5 technology (as specified in RFC 1510).
ACKNOWLEDGMENTS
Much of the material in this memo is based on working documents
drafted by John Wray of Digital Equipment Corporation and on
discussions, implementation activities, and interoperability testing
involving Marc Horowitz, Ted Ts'o, and John Wray. Particular thanks
are due to each of these individuals for their contributions towards
development and availability of GSS-API support within the Kerberos
Version 5 code base.
1. Token Formats
This section discusses protocol-visible characteristics of the GSS-
API mechanism to be implemented atop Kerberos V5 security technology
per RFC-1508 and RFC-1510; it defines elements of protocol for
interoperability and is independent of language bindings per RFC-
1509.
Tokens transferred between GSS-API peers (for security context
management and per-message protection purposes) are defined. The
data elements exchanged between a GSS-API endpoint implementation and
the Kerberos KDC are not specific to GSS-API usage and are therefore
defined within RFC-1510 rather than within this specification.
Linn Standards Track [Page 1]
RFC 1964 Kerberos Version 5 GSS-API June 1996
To support ongoing experimentation, testing, and evolution of the
specification, the Kerberos V5 GSS-API mechanism as defined in this
and any successor memos will be identified with the following Object
Identifier, as defined in RFC-1510, until the specification is
advanced to the level of Proposed Standard RFC:
{iso(1), org(3), dod(5), internet(1), security(5), kerberosv5(2)}
Upon advancement to the level of Proposed Standard RFC, the Kerberos
V5 GSS-API mechanism will be identified by an Object Identifier
having the value:
{iso(1) member-body(2) United States(840) mit(113554) infosys(1)
gssapi(2) krb5(2)}
1.1. Context Establishment Tokens
Per RFC-1508, Appendix B, the initial context establishment token
will be enclosed within framing as follows:
InitialContextToken ::=
[APPLICATION 0] IMPLICIT SEQUENCE {
thisMech MechType
-- MechType is OBJECT IDENTIFIER
-- representing "Kerberos V5"
innerContextToken ANY DEFINED BY thisMech
-- contents mechanism-specific;
-- ASN.1 usage within innerContextToken
-- is not required
}
The innerContextToken of the initial context token will consist of a
Kerberos V5 KRB_AP_REQ message, preceded by a two-byte token-id
(TOK_ID) field, which shall contain the value 01 00.
The above GSS-API framing shall be applied to all tokens emitted by
the Kerberos V5 GSS-API mechanism, including KRB_AP_REP, KRB_ERROR,
context-deletion, and per-message tokens, not just to the initial
token in a context establishment sequence. While not required by
RFC-1508, this enables implementations to perform enhanced error-
checking. The innerContextToken field of context establishment tokens
for the Kerberos V5 GSS-API mechanism will contain a Kerberos message
(KRB_AP_REQ, KRB_AP_REP or KRB_ERROR), preceded by a 2-byte TOK_ID
field containing 01 00 for KRB_AP_REQ messages, 02 00 for KRB_AP_REP
messages and 03 00 for KRB_ERROR messages.
Linn Standards Track [Page 2]
RFC 1964 Kerberos Version 5 GSS-API June 1996
1.1.1. Initial Token
Relevant KRB_AP_REQ syntax (from RFC-1510) is as follows:
AP-REQ ::= [APPLICATION 14] SEQUENCE {
pvno [0] INTEGER, -- indicates Version 5
msg-type [1] INTEGER, -- indicates KRB_AP_REQ
ap-options[2] APOptions,
ticket[3] Ticket,
authenticator[4] EncryptedData
}
APOptions ::= BIT STRING {
reserved (0),
use-session-key (1),
mutual-required (2)
}
Ticket ::= [APPLICATION 1] SEQUENCE {
tkt-vno [0] INTEGER, -- indicates Version 5
realm [1] Realm,
sname [2] PrincipalName,
enc-part [3] EncryptedData
}
-- Encrypted part of ticket
EncTicketPart ::= [APPLICATION 3] SEQUENCE {
flags[0] TicketFlags,
key[1] EncryptionKey,
crealm[2] Realm,
cname[3] PrincipalName,
transited[4] TransitedEncoding,
authtime[5] KerberosTime,
starttime[6] KerberosTime OPTIONAL,
endtime[7] KerberosTime,
renew-till[8] KerberosTime OPTIONAL,
caddr[9] HostAddresses OPTIONAL,
authorization-data[10] AuthorizationData OPTIONAL
}
-- Unencrypted authenticator
Authenticator ::= [APPLICATION 2] SEQUENCE {
authenticator-vno[0] INTEGER,
crealm[1] Realm,
cname[2] PrincipalName,
cksum[3] Checksum OPTIONAL,
cusec[4] INTEGER,
ctime[5] KerberosTime,
Linn Standards Track [Page 3]
RFC 1964 Kerberos Version 5 GSS-API June 1996
subkey[6] EncryptionKey OPTIONAL,
seq-number[7] INTEGER OPTIONAL,
authorization-data[8] AuthorizationData OPTIONAL
}
For purposes of this specification, the authenticator shall include
the optional sequence number, and the checksum field shall be used to
convey channel binding, service flags, and optional delegation
information. The checksum will have a type of 0x8003 (a value being
registered within the Kerberos protocol specification), and a value
field of at least 24 bytes in length. The length of the value field
is extended beyond 24 bytes if and only if an optional facility to
carry a Kerberos-defined KRB_CRED message for delegation purposes is
supported by an implementation and active on a context. When
delegation is active, a TGT with its FORWARDABLE flag set will be
transferred within the KRB_CRED message.
The checksum value field's format is as follows:
Byte Name Description
0..3 Lgth Number of bytes in Bnd field;
Currently contains hex 10 00 00 00
(16, represented in little-endian form)
4..19 Bnd MD5 hash of channel bindings, taken over all non-null
components of bindings, in order of declaration.
Integer fields within channel bindings are represented
in little-endian order for the purposes of the MD5
calculation.
20..23 Flags Bit vector of context-establishment flags,
with values consistent with RFC-1509, p. 41:
GSS_C_DELEG_FLAG: 1
GSS_C_MUTUAL_FLAG: 2
GSS_C_REPLAY_FLAG: 4
GSS_C_SEQUENCE_FLAG: 8
GSS_C_CONF_FLAG: 16
GSS_C_INTEG_FLAG: 32
The resulting bit vector is encoded into bytes 20..23
in little-endian form.
24..25 DlgOpt The Delegation Option identifier (=1) [optional]
26..27 Dlgth The length of the Deleg field. [optional]
28..n Deleg A KRB_CRED message (n = Dlgth + 29) [optional]
In computing the contents of the "Bnd" field, the following detailed
points apply:
(1) Each integer field shall be formatted into four bytes, using
little-endian byte ordering, for purposes of MD5 hash
computation.
Linn Standards Track [Page 4]
RFC 1964 Kerberos Version 5 GSS-API June 1996
(2) All input length fields within gss_buffer_desc elements of a
gss_channel_bindings_struct, even those which are zero-valued,
shall be included in the hash calculation; the value elements of
gss_buffer_desc elements shall be dereferenced, and the
resulting data shall be included within the hash computation,
only for the case of gss_buffer_desc elements having non-zero
length specifiers.
(3) If the caller passes the value GSS_C_NO_BINDINGS instead of
a valid channel bindings structure, the Bnd field shall be set
to 16 zero-valued bytes.
In the initial Kerberos V5 GSS-API mechanism token (KRB_AP_REQ token)
from initiator to target, the GSS_C_DELEG_FLAG, GSS_C_MUTUAL_FLAG,
GSS_C_REPLAY_FLAG, and GSS_C_SEQUENCE_FLAG values shall each be set
as the logical AND of the initiator's corresponding request flag to
GSS_Init_sec_context() and a Boolean indicator of whether that
optional service is available to GSS_Init_sec_context()'s caller.
GSS_C_CONF_FLAG and GSS_C_INTEG_FLAG, for which no corresponding
context-level input indicator flags to GSS_Init_sec_context() exist,
shall each be set to indicate whether their respective per-message
protection services are available for use on the context being
established.
When input source address channel binding values are provided by a
caller (i.e., unless the input argument is GSS_C_NO_BINDINGS or the
source address specifier value within the input structure is
GSS_C_NULL_ADDRTYPE), and the corresponding token received from the
context's peer bears address restrictions, it is recommended that an
implementation of the Kerberos V5 GSS-API mechanism should check that
the source address as provided by the caller matches that in the
received token, and should return the GSS_S_BAD_BINDINGS major_status
value if a mismatch is detected. Note: discussion is ongoing about
the strength of recommendation to be made in this area, and on the
circumstances under which such a recommendation should be applicable;
implementors are therefore advised that changes on this matter may be
included in subsequent versions of this specification.
1.1.2. Response Tokens
A context establishment sequence based on the Kerberos V5 mechanism
will perform one-way authentication (without confirmation or any
return token from target to initiator in response to the initiator's
KRB_AP_REQ) if the mutual_req bit is not set in the application's
call to GSS_Init_sec_context(). Applications requiring confirmation
that their authentication was successful should request mutual
authentication, resulting in a "mutual-required" indication within
KRB_AP_REQ APoptions and the setting of the mutual_req bit in the
Linn Standards Track [Page 5]
RFC 1964 Kerberos Version 5 GSS-API June 1996
flags field of the authenticator checksum. In response to such a
request, the context target will reply to the initiator with a token
containing either a KRB_AP_REP or KRB_ERROR, completing the mutual
context establishment exchange.
Relevant KRB_AP_REP syntax is as follows:
AP-REP ::= [APPLICATION 15] SEQUENCE {
pvno [0] INTEGER, -- represents Kerberos V5
msg-type [1] INTEGER, -- represents KRB_AP_REP
enc-part [2] EncryptedData
}
EncAPRepPart ::= [APPLICATION 27] SEQUENCE {
ctime [0] KerberosTime,
cusec [1] INTEGER,
subkey [2] EncryptionKey OPTIONAL,
seq-number [3] INTEGER OPTIONAL
}
The optional seq-number element within the AP-REP's EncAPRepPart
shall be included.
The syntax of KRB_ERROR is as follows:
KRB-ERROR ::= [APPLICATION 30] SEQUENCE {
pvno[0] INTEGER,
msg-type[1] INTEGER,
ctime[2] KerberosTime OPTIONAL,
cusec[3] INTEGER OPTIONAL,
stime[4] KerberosTime,
susec[5] INTEGER,
error-code[6] INTEGER,
crealm[7] Realm OPTIONAL,
cname[8] PrincipalName OPTIONAL,
realm[9] Realm, -- Correct realm
sname[10] PrincipalName, -- Correct name
e-text[11] GeneralString OPTIONAL,
e-data[12] OCTET STRING OPTIONAL
}
Values to be transferred in the error-code field of a KRB-ERROR
message are defined in RFC-1510">[RFC-1510], not in this specification.
Linn Standards Track [Page 6]
RFC 1964 Kerberos Version 5 GSS-API June 1996
1.2. Per-Message and Context Deletion Tokens
Three classes of tokens are defined in this section: "MIC" tokens,
emitted by calls to GSS_GetMIC() (formerly GSS_Sign()) and consumed
by calls to GSS_VerifyMIC() (formerly GSS_Verify()), "Wrap" tokens,
emitted by calls to GSS_Wrap() (formerly GSS_Seal()) and consumed by
calls to GSS_Unwrap() (formerly GSS_Unseal()), and context deletion
tokens, emitted by calls to GSS_Delete_sec_context() and consumed by
calls to GSS_Process_context_token(). Note: References to GSS-API
per-message routines in the remainder of this specification will be
based on those routines' newer recommended names rather than those
names' predecessors.
Several variants of cryptographic keys are used in generation and
processing of per-message tokens:
(1) context key: uses Kerberos session key (or subkey, if
present in authenticator emitted by context initiator) directly
(2) confidentiality key: forms variant of context key by
exclusive-OR with the hexadecimal constant f0f0f0f0f0f0f0f0.
(3) MD2.5 seed key: forms variant of context key by reversing
the bytes of the context key (i.e. if the original key is the
8-byte sequence {aa, bb, cc, dd, ee, ff, gg, hh}, the seed key
will be {hh, gg, ff, ee, dd, cc, bb, aa}).
1.2.1. Per-message Tokens - MIC
Use of the GSS_GetMIC() call yields a token, separate from the user
data being protected, which can be used to verify the integrity of
that data as received. The token has the following format:
Byte no Name Description
0..1 TOK_ID Identification field.
Tokens emitted by GSS_GetMIC() contain
the hex value 01 01 in this field.
2..3 SGN_ALG Integrity algorithm indicator.
00 00 - DES MAC MD5
01 00 - MD2.5
02 00 - DES MAC
4..7 Filler Contains ff ff ff ff
8..15 SND_SEQ Sequence number field.
16..23 SGN_CKSUM Checksum of "to-be-signed data",
calculated according to algorithm
specified in SGN_ALG field.
Linn Standards Track [Page 7]
RFC 1964 Kerberos Version 5 GSS-API June 1996
GSS-API tokens must be encapsulated within the higher-level protocol
by the application; no embedded length field is necessary.
1.2.1.1. Checksum
Checksum calculation procedure (common to all algorithms): Checksums
are calculated over the data field, logically prepended by the first
8 bytes of the plaintext packet header. The resulting value binds
the data to the packet type and signature algorithm identifier
fields.
DES MAC MD5 algorithm: The checksum is formed by computing an MD5
RFC-1321">[RFC-1321] hash over the plaintext data, and then computing a DES-CBC
MAC on the 16-byte MD5 result. A standard 64-bit DES-CBC MAC is
computed per [FIPS-PUB-113], employing the context key and a zero IV.
The 8-byte result is stored in the SGN_CKSUM field.
MD2.5 algorithm: The checksum is formed by first DES-CBC encrypting a
16-byte zero-block, using a zero IV and a key formed by reversing the
bytes of the context key (i.e. if the original key is the 8-byte
sequence {aa, bb, cc, dd, ee, ff, gg, hh}, the checksum key will be
{hh, gg, ff, ee, dd, cc, bb, aa}). The resulting 16-byte value is
logically prepended to the to-be-signed data. A standard MD5
checksum is calculated over the combined data, and the first 8 bytes
of the result are stored in the SGN_CKSUM field. Note 1: we refer to
this algorithm informally as "MD2.5" to connote the fact that it uses
half of the 128 bits generated by MD5; use of only a subset of the
MD5 bits is intended to protect against the prospect that data could
be postfixed to an existing message with corresponding modifications
being made to the checksum. Note 2: This algorithm is fairly novel
and has received more limited evaluation than that to which other
integrity algorithms have been subjected. An initial, limited
evaluation indicates that it may be significantly weaker than DES MAC
MD5.
DES-MAC algorithm: A standard 64-bit DES-CBC MAC is computed on the
plaintext data per [FIPS-PUB-113], employing the context key and a
zero IV. Padding procedures to accomodate plaintext data lengths
which may not be integral multiples of 8 bytes are defined in [FIPS-
PUB-113]. The result is an 8-byte value, which is stored in the
SGN_CKSUM field. Support for this algorithm may not be present in
all implementations.
1.2.1.2. Sequence Number
Sequence number field: The 8 byte plaintext sequence number field is
formed from the sender's four-byte sequence number as follows. If
the four bytes of the sender's sequence number are named s0, s1, s2
Linn Standards Track [Page 8]
RFC 1964 Kerberos Version 5 GSS-API June 1996
and s3 (from least to most significant), the plaintext sequence
number field is the 8 byte sequence: (s0, s1, s2, s3, di, di, di,
di), where 'di' is the direction-indicator (Hex 0 - sender is the
context initiator, Hex FF - sender is the context acceptor). The
field is then DES-CBC encrypted using the context key and an IV
formed from the first 8 bytes of the previously calculated SGN_CKSUM
field. After sending a GSS_GetMIC() or GSS_Wrap() token, the sender's
sequence number is incremented by one.
The receiver of the token will first verify the SGN_CKSUM field. If
valid, the sequence number field may be decrypted and compared to the
expected sequence number. The repetition of the (effectively 1-bit)
direction indicator within the sequence number field provides
redundancy so that the receiver may verify that the decryption
succeeded.
Since the checksum computation is used as an IV to the sequence
number decryption, attempts to splice a checksum and sequence number
from different messages will be detected. The direction indicator
will detect packets that have been maliciously reflected.
The sequence number provides a basis for detection of replayed
tokens. Replay detection can be performed using state information
retained on received sequence numbers, interpreted in conjunction
with the security context on which they arrive.
Provision of per-message replay and out-of-sequence detection
services is optional for implementations of the Kerberos V5 GSS-API
mechanism. Further, it is recommended that implementations of the
Kerberos V5 GSS-API mechanism which offer these services should honor
a caller's request that the services be disabled on a context.
Specifically, if replay_det_req_flag is input FALSE, replay_det_state
should be returned FALSE and the GSS_DUPLICATE_TOKEN and
GSS_OLD_TOKEN stati should not be indicated as a result of duplicate
detection when tokens are processed; if sequence_req_flag is input
FALSE, sequence_state should be returned FALSE and
GSS_DUPLICATE_TOKEN, GSS_OLD_TOKEN, and GSS_UNSEQ_TOKEN stati should
not be indicated as a result of out-of-sequence detection when tokens
are processed.
1.2.2. Per-message Tokens - Wrap
Use of the GSS_Wrap() call yields a token which encapsulates the
input user data (optionally encrypted) along with associated
integrity check quantities. The token emitted by GSS_Wrap() consists
of an integrity header whose format is identical to that emitted by
GSS_GetMIC() (except that the TOK_ID field contains the value 02 01),
followed by a body portion that contains either the plaintext data
Linn Standards Track [Page 9]
RFC 1964 Kerberos Version 5 GSS-API June 1996
(if SEAL_ALG = ff ff) or encrypted data for any other supported value
of SEAL_ALG. Currently, only SEAL_ALG = 00 00 is supported, and
means that DES-CBC encryption is being used to protect the data.
The GSS_Wrap() token has the following format:
Byte no Name Description
0..1 TOK_ID Identification field.
Tokens emitted by GSS_Wrap() contain
the hex value 02 01 in this field.
2..3 SGN_ALG Checksum algorithm indicator.
00 00 - DES MAC MD5
01 00 - MD2.5
02 00 - DES MAC
4..5 SEAL_ALG ff ff - none
00 00 - DES
6..7 Filler Contains ff ff
8..15 SND_SEQ Encrypted sequence number field.
16..23 SGN_CKSUM Checksum of plaintext padded data,
calculated according to algorithm
specified in SGN_ALG field.
24..last Data encrypted or plaintext padded data
GSS-API tokens must be encapsulated within the higher-level protocol
by the application; no embedded length field is necessary.
1.2.2.1. Checksum
Checksum calculation procedure (common to all algorithms): Checksums
are calculated over the plaintext padded data field, logically
prepended by the first 8 bytes of the plaintext packet header. The
resulting signature binds the data to the packet type, protocol
version, and signature algorithm identifier fields.
DES MAC MD5 algorithm: The checksum is formed by computing an MD5
hash over the plaintext padded data, and then computing a DES-CBC MAC
on the 16-byte MD5 result. A standard 64-bit DES-CBC MAC is computed
per [FIPS-PUB-113], employing the context key and a zero IV. The 8-
byte result is stored in the SGN_CKSUM field.
MD2.5 algorithm: The checksum is formed by first DES-CBC encrypting a
16-byte zero-block, using a zero IV and a key formed by reversing the
bytes of the context key (i.e., if the original key is the 8-byte
sequence {aa, bb, cc, dd, ee, ff, gg, hh}, the checksum key will be
{hh, gg, ff, ee, dd, cc, bb, aa}). The resulting 16-byte value is
logically pre-pended to the "to-be-signed data". A standard MD5
checksum is calculated over the combined data, and the first 8 bytes
of the result are stored in the SGN_CKSUM field.
Linn Standards Track [Page 10]
RFC 1964 Kerberos Version 5 GSS-API June 1996
DES-MAC algorithm: A standard 64-bit DES-CBC MAC is computed on the
plaintext padded data per [FIPS-PUB-113], employing the context key
and a zero IV. The plaintext padded data is already assured to be an
integral multiple of 8 bytes; no additional padding is required or
applied in order to accomplish MAC calculation. The result is an 8-
byte value, which is stored in the SGN_CKSUM field. Support for this
lgorithm may not be present in all implementations.
1.2.2.2. Sequence Number
Sequence number field: The 8 byte plaintext sequence number field is
formed from the sender's four-byte sequence number as follows. If
the four bytes of the sender's sequence number are named s0, s1, s2
and s3 (from least to most significant), the plaintext sequence
number field is the 8 byte sequence: (s0, s1, s2, s3, di, di, di,
di), where 'di' is the direction-indicator (Hex 0 - sender is the
context initiator, Hex FF - sender is the context acceptor).
The field is then DES-CBC encrypted using the context key and an IV
formed from the first 8 bytes of the SEAL_CKSUM field.
After sending a GSS_GetMIC() or GSS_Wrap() token, the sender's
sequence numbers are incremented by one.
1.2.2.3. Padding
Data padding: Before encryption and/or signature calculation,
plaintext data is padded to the next highest multiple of 8 bytes, by
appending between 1 and 8 bytes, the value of each such byte being
the total number of pad bytes. For example, given data of length 20
bytes, four pad bytes will be appended, and each byte will contain
the hex value 04. An 8-byte random confounder is prepended to the
data, and signatures are calculated over the resulting padded
plaintext.
After padding, the data is encrypted according to the algorithm
specified in the SEAL_ALG field. For SEAL_ALG=DES (the only non-null
algorithm currently supported), the data is encrypted using DES-CBC,
with an IV of zero. The key used is derived from the established
context key by XOR-ing the context key with the hexadecimal constant
f0f0f0f0f0f0f0f0.
1.2.3. Context deletion token
The token emitted by GSS_Delete_sec_context() is based on the packet
format for tokens emitted by GSS_GetMIC(). The context-deletion
token has the following format:
Linn Standards Track [Page 11]
RFC 1964 Kerberos Version 5 GSS-API June 1996
Byte no Name Description
0..1 TOK_ID Identification field.
Tokens emitted by
GSS_Delete_sec_context() contain
the hex value 01 02 in this field.
2..3 SGN_ALG Integrity algorithm indicator.
00 00 - DES MAC MD5
01 00 - MD2.5
02 00 - DES MAC
4..7 Filler Contains ff ff ff ff
8..15 SND_SEQ Sequence number field.
16..23 SGN_CKSUM Checksum of "to-be-signed data",
calculated according to algorithm
specified in SGN_ALG field.
SGN_ALG and SND_SEQ will be calculated as for tokens emitted by
GSS_GetMIC(). The SGN_CKSUM will be calculated as for tokens emitted
by GSS_GetMIC(), except that the user-data component of the "to-be-
signed" data will be a zero-length string.
2. Name Types and Object Identifiers
This section discusses the name types which may be passed as input to
the Kerberos V5 GSS-API mechanism's GSS_Import_name() call, and their
associated identifier values. It defines interface elements in
support of portability, and assumes use of C language bindings per
RFC-1509. In addition to specifying OID values for name type
identifiers, symbolic names are included and recommended to GSS-API
implementors in the interests of convenience to callers. It is
understood that not all implementations of the Kerberos V5 GSS-API
mechanism need support all name types in this list, and that
additional name forms will likely be added to this list over time.
Further, the definitions of some or all name types may later migrate
to other, mechanism-independent, specifications. The occurrence of a
name type in this specification is specifically not intended to
suggest that the type may be supported only by an implementation of
the Kerberos V5 mechanism. In particular, the occurrence of the
string "_KRB5_" in the symbolic name strings constitutes a means to
unambiguously register the name strings, avoiding collision with
other documents; it is not meant to limit the name types' usage or
applicability.
For purposes of clarification to GSS-API implementors, this section's
discussion of some name forms describes means through which those
forms can be supported with existing Kerberos technology. These
discussions are not intended to preclude alternative implementation
strategies for support of the name forms within Kerberos mechanisms
or mechanisms based on other technologies. To enhance application
Linn Standards Track [Page 12]
RFC 1964 Kerberos Version 5 GSS-API June 1996
portability, implementors of mechanisms are encouraged to support
name forms as defined in this section, even if their mechanisms are
independent of Kerberos V5.
2.1. Mandatory Name Forms
This section discusses name forms which are to be supported by all
conformant implementations of the Kerberos V5 GSS-API mechanism.
2.1.1. Kerberos Principal Name Form
This name form shall be represented by the Object Identifier {iso(1)
member-body(2) United States(840) mit(113554) infosys(1) gssapi(2)
krb5(2) krb5_name(1)}. The recommended symbolic name for this type
is "GSS_KRB5_NT_PRINCIPAL_NAME".
This name type corresponds to the single-string representation of a
Kerberos name. (Within the MIT Kerberos V5 implementation, such
names are parseable with the krb5_parse_name() function.) The
elements included within this name representation are as follows,
proceeding from the beginning of the string:
(1) One or more principal name components; if more than one
principal name component is included, the components are
separated by `/`. Arbitrary octets may be included within
principal name components, with the following constraints and
special considerations:
(1a) Any occurrence of the characters `@` or `/` within a
name component must be immediately preceded by the `\`
quoting character, to prevent interpretation as a component
or realm separator.
(1b) The ASCII newline, tab, backspace, and null characters
may occur directly within the component or may be
represented, respectively, by `\n`, `\t`, `\b`, or `\0`.
(1c) If the `\` quoting character occurs outside the contexts
described in (1a) and (1b) above, the following character is
interpreted literally. As a special case, this allows the
doubled representation `\\` to represent a single occurrence
of the quoting character.
(1d) An occurrence of the `\` quoting character as the last
character of a component is illegal.
Linn Standards Track [Page 13]
RFC 1964 Kerberos Version 5 GSS-API June 1996
(2) Optionally, a `@` character, signifying that a realm name
immediately follows. If no realm name element is included, the
local realm name is assumed. The `/` , `:`, and null characters
may not occur within a realm name; the `@`, newline, tab, and
backspace characters may be included using the quoting
conventions described in (1a), (1b), and (1c) above.
2.1.2. Host-Based Service Name Form
This name form has been incorporated at the mechanism-independent
GSS-API level as of GSS-API, Version 2. This subsection retains the
Object Identifier and symbolic name assignments previously made at
the Kerberos V5 GSS-API mechanism level, and adopts the definition as
promoted to the mechanism-independent level.
This name form shall be represented by the Object Identifier {iso(1)
member-body(2) United States(840) mit(113554) infosys(1) gssapi(2)
generic(1) service_name(4)}. The previously recommended symbolic
name for this type is "GSS_KRB5_NT_HOSTBASED_SERVICE_NAME". The
currently preferred symbolic name for this type is
"GSS_C_NT_HOSTBASED_SERVICE".
This name type is used to represent services associated with host
computers. This name form is constructed using two elements,
"service" and "hostname", as follows:
service@hostname
When a reference to a name of this type is resolved, the "hostname"
is canonicalized by attempting a DNS lookup and using the fully-
qualified domain name which is returned, or by using the "hostname"
as provided if the DNS lookup fails. The canonicalization operation
also maps the host's name into lower-case characters.
The "hostname" element may be omitted. If no "@" separator is
included, the entire name is interpreted as the service specifier,
with the "hostname" defaulted to the canonicalized name of the local
host.
Values for the "service" element will be registered with the IANA.
2.1.3. Exported Name Object Form for Kerberos V5 Mechanism
Support for this name form is not required for GSS-V1
implementations, but will be required for use in conjunction with the
GSS_Export_name() call planned for GSS-API Version 2. Use of this
name form will be signified by a "GSS-API Exported Name Object" OID
value which will be defined at the mechanism-independent level for
Linn Standards Track [Page 14]
RFC 1964 Kerberos Version 5 GSS-API June 1996
GSS-API Version 2.
This name type represents a self-describing object, whose framing
structure will be defined at the mechanism-independent level for
GSS-API Version 2. When generated by the Kerberos V5 mechanism, the
Mechanism OID within the exportable name shall be that of the
Kerberos V5 mechanism. The name component within the exportable name
shall be a contiguous string with structure as defined for the
Kerberos Principal Name Form.
In order to achieve a distinguished encoding for comparison purposes,
the following additional constraints are imposed on the export
operation:
(1) all occurrences of the characters `@`, `/`, and `\` within
principal components or realm names shall be quoted with an
immediately-preceding `\`.
(2) all occurrences of the null, backspace, tab, or newline
characters within principal components or realm names will be
represented, respectively, with `\0`, `\b`, `\t`, or `\n`.
(3) the `\` quoting character shall not be emitted within an
exported name except to accomodate cases (1) and (2).
2.2. Optional Name Forms
This section discusses additional name forms which may optionally be
supported by implementations of the Kerberos V5 GSS-API mechanism.
It is recognized that some of the name forms cited here are derived
from UNIX(tm) operating system platforms; some listed forms may be
irrelevant to non-UNIX platforms, and definition of additional forms
corresponding to such platforms may also be appropriate. It is also
recognized that OS-specific functions outside GSS-API are likely to
exist in order to perform translations among these forms, and that
GSS-API implementations supporting these forms may themselves be
layered atop such OS-specific functions. Inclusion of this support
within GSS-API implementations is intended as a convenience to
applications.
2.2.1. User Name Form
This name form shall be represented by the Object Identifier {iso(1)
member-body(2) United States(840) mit(113554) infosys(1) gssapi(2)
generic(1) user_name(1)}. The recommended symbolic name for this
type is "GSS_KRB5_NT_USER_NAME".
This name type is used to indicate a named user on a local system.
Linn Standards Track [Page 15]
RFC 1964 Kerberos Version 5 GSS-API June 1996
Its interpretation is OS-specific. This name form is constructed as:
username
Assuming that users' principal names are the same as their local
operating system names, an implementation of GSS_Import_name() based
on Kerberos V5 technology can process names of this form by
postfixing an "@" sign and the name of the local realm.
2.2.2. Machine UID Form
This name form shall be represented by the Object Identifier {iso(1)
member-body(2) United States(840) mit(113554) infosys(1) gssapi(2)
generic(1) machine_uid_name(2)}. The recommended symbolic name for
this type is "GSS_KRB5_NT_MACHINE_UID_NAME".
This name type is used to indicate a numeric user identifier
corresponding to a user on a local system. Its interpretation is
OS-specific. The gss_buffer_desc representing a name of this type
should contain a locally-significant uid_t, represented in host byte
order. The GSS_Import_name() operation resolves this uid into a
username, which is then treated as the User Name Form.
2.2.3. String UID Form
This name form shall be represented by the Object Identifier {iso(1)
member-body(2) United States(840) mit(113554) infosys(1) gssapi(2)
generic(1) string_uid_name(3)}. The recommended symbolic name for
this type is "GSS_KRB5_NT_STRING_UID_NAME".
This name type is used to indicate a string of digits representing
the numeric user identifier of a user on a local system. Its
interpretation is OS-specific. This name type is similar to the
Machine UID Form, except that the buffer contains a string
representing the uid_t.
3. Credentials Management
The Kerberos V5 protocol uses different credentials (in the GSSAPI
sense) for initiating and accepting security contexts. Normal
clients receive a ticket-granting ticket (TGT) and an associated
session key at "login" time; the pair of a TGT and its corresponding
session key forms a credential which is suitable for initiating
security contexts. A ticket-granting ticket, its session key, and
any other (ticket, key) pairs obtained through use of the ticket-
granting-ticket, are typically stored in a Kerberos V5 credentials
cache, sometimes known as a ticket file.
Linn Standards Track [Page 16]
RFC 1964 Kerberos Version 5 GSS-API June 1996
The encryption key used by the Kerberos server to seal tickets for a
particular application service forms the credentials suitable for
accepting security contexts. These service keys are typically stored
in a Kerberos V5 key table, or srvtab file. In addition to their use
as accepting credentials, these service keys may also be used to
obtain initiating credentials for their service principal.
The Kerberos V5 mechanism's credential handle may contain references
to either or both types of credentials. It is a local matter how the
Kerberos V5 mechanism implementation finds the appropriate Kerberos
V5 credentials cache or key table.
However, when the Kerberos V5 mechanism attempts to obtain initiating
credentials for a service principal which are not available in a
credentials cache, and the key for that service principal is
available in a Kerberos V5 key table, the mechanism should use the
service key to obtain initiating credentials for that service. This
should be accomplished by requesting a ticket-granting-ticket from
the Kerberos Key Distribution Center (KDC), and decrypting the KDC's
reply using the service key.
4. Parameter Definitions
This section defines parameter values used by the Kerberos V5 GSS-API
mechanism. It defines interface elements in support of portability,
and assumes use of C language bindings per RFC-1509.
4.1. Minor Status Codes
This section recommends common symbolic names for minor_status values
to be returned by the Kerberos V5 GSS-API mechanism. Use of these
definitions will enable independent implementors to enhance
application portability across different implementations of the
mechanism defined in this specification. (In all cases,
implementations of GSS_Display_status() will enable callers to
convert minor_status indicators to text representations.) Each
implementation should make available, through include files or other
means, a facility to translate these symbolic names into the concrete
values which a particular GSS-API implementation uses to represent
the minor_status values specified in this section.
It is recognized that this list may grow over time, and that the need
for additional minor_status codes specific to particular
implementations may arise. It is recommended, however, that
implementations should return a minor_status value as defined on a
mechanism-wide basis within this section when that code is accurately
representative of reportable status rather than using a separate,
implementation-defined code.
Linn Standards Track [Page 17]
RFC 1964 Kerberos Version 5 GSS-API June 1996
4.1.1. Non-Kerberos-specific codes
GSS_KRB5_S_G_BAD_SERVICE_NAME
/* "No @ in SERVICE-NAME name string" */
GSS_KRB5_S_G_BAD_STRING_UID
/* "STRING-UID-NAME contains nondigits" */
GSS_KRB5_S_G_NOUSER
/* "UID does not resolve to username" */
GSS_KRB5_S_G_VALIDATE_FAILED
/* "Validation error" */
GSS_KRB5_S_G_BUFFER_ALLOC
/* "Couldn't allocate gss_buffer_t data" */
GSS_KRB5_S_G_BAD_MSG_CTX
/* "Message context invalid" */
GSS_KRB5_S_G_WRONG_SIZE
/* "Buffer is the wrong size" */
GSS_KRB5_S_G_BAD_USAGE
/* "Credential usage type is unknown" */
GSS_KRB5_S_G_UNKNOWN_QOP
/* "Unknown quality of protection specified" */
4.1.2. Kerberos-specific-codes
GSS_KRB5_S_KG_CCACHE_NOMATCH
/* "Principal in credential cache does not match desired name" */
GSS_KRB5_S_KG_KEYTAB_NOMATCH
/* "No principal in keytab matches desired name" */
GSS_KRB5_S_KG_TGT_MISSING
/* "Credential cache has no TGT" */
GSS_KRB5_S_KG_NO_SUBKEY
/* "Authenticator has no subkey" */
GSS_KRB5_S_KG_CONTEXT_ESTABLISHED
/* "Context is already fully established" */
GSS_KRB5_S_KG_BAD_SIGN_TYPE
/* "Unknown signature type in token" */
GSS_KRB5_S_KG_BAD_LENGTH
/* "Invalid field length in token" */
GSS_KRB5_S_KG_CTX_INCOMPLETE
/* "Attempt to use incomplete security context" */
4.2. Quality of Protection Values
This section defines Quality of Protection (QOP) values to be used
with the Kerberos V5 GSS-API mechanism as input to GSS_Wrap() and
GSS_GetMIC() routines in order to select among alternate integrity
and confidentiality algorithms. Additional QOP values may be added in
future versions of this specification. Non-overlapping bit positions
are and will be employed in order that both integrity and
Linn Standards Track [Page 18]
RFC 1964 Kerberos Version 5 GSS-API June 1996
confidentiality QOP may be selected within a single parameter, via
inclusive-OR of the specified integrity and confidentiality values.
4.2.1. Integrity Algorithms
The following Quality of Protection (QOP) values are currently
defined for the Kerberos V5 GSS-API mechanism, and are used to select
among alternate integrity checking algorithms.
GSS_KRB5_INTEG_C_QOP_MD5 (numeric value: 1)
/* Integrity using partial MD5 ("MD2.5") of plaintext */
GSS_KRB5_INTEG_C_QOP_DES_MD5 (numeric value: 2)
/* Integrity using DES MAC of MD5 of plaintext */
GSS_KRB5_INTEG_C_QOP_DES_MAC (numeric value: 3)
/* Integrity using DES MAC of plaintext */
4.2.2. Confidentiality Algorithms
Only one confidentiality QOP value is currently defined for the
Kerberos V5 GSS-API mechanism:
GSS_KRB5_CONF_C_QOP_DES (numeric value: 0)
/* Confidentiality with DES */
Note: confidentiality QOP should be indicated only by GSS-API calls
capable of providing confidentiality services. If non-zero
confidentiality QOP values are defined in future to represent
different algorithms, therefore, the bit positions containing those
values should be cleared before being returned by implementations of
GSS_GetMIC() and GSS_VerifyMIC().
4.3. Buffer Sizes
All implementations of this specification shall be capable of
accepting buffers of at least 16 Kbytes as input to GSS_GetMIC(),
GSS_VerifyMIC(), and GSS_Wrap(), and shall be capable of accepting
the output_token generated by GSS_Wrap() for a 16 Kbyte input buffer
as input to GSS_Unwrap(). Support for larger buffer sizes is optional
but recommended.
Linn Standards Track [Page 19]
RFC 1964 Kerberos Version 5 GSS-API June 1996
5. Security Considerations
Security issues are discussed throughout this memo.
6. References
RFC-1321">[RFC-1321]: Rivest, R., "The MD5 Message-Digest Algorithm", RFC
1321, April 1992.
RFC-1508">[RFC-1508]: Linn, J., "Generic Security Service Application Program
Interface", RFC 1508, September 1993.
RFC-1509">[RFC-1509]: Wray, J., "Generic Security Service Application Program
Interface: C-bindings", RFC 1509, September 1993.
RFC-1510">[RFC-1510]: Kohl, J., and C. Neuman, "The Kerberos Network
Authentication Service (V5)", RFC 1510, September 1993.
[FIPS-PUB-113]: National Bureau of Standards, Federal Information
Processing Standard 113, "Computer Data Authentication", May 1985.
AUTHOR'S ADDRESS
John Linn
OpenVision Technologies
One Main St.
Cambridge, MA 02142 USA
Phone: +1 617.374.2245
EMail: John.Linn@ov.com
Linn Standards Track [Page 20]