RFC 5272
Certificate Management over CMS (CMC)
4. PKI Responses
Two types of PKI Responses exist. This section gives the details on both types.4.1. Simple PKI Response
Clients MUST be able to process the Simple PKI Response. The Simple PKI Response consists of a SignedData with no EncapsulatedContentInfo and no SignerInfo. The certificates requested in the PKI Response are returned in the certificate field of the SignedData. Clients MUST NOT assume the certificates are in any order. Servers SHOULD include all intermediate certificates needed to form complete certification paths to one or more trust anchors, not just the newly issued certificate(s). The server MAY additionally return CRLs in the CRL bag. Servers MAY include the self-signed certificates. Clients MUST NOT implicitly trust included self-signed certificate(s) merely due to its presence in the certificate bag. In the event clients receive a new self-signed certificate from the server, clients SHOULD provide a mechanism to enable the user to use the certificate as a trust anchor. (The Publish Trust Anchors control (Section 6.15) should be used in the event that the server intends the client to accept one or more certificates as trust anchors. This requires the use of the Full PKI Response message.)
4.2. Full PKI Response
Clients MUST be able to process a Full PKI Response. The Full PKI Response consists of a SignedData or AuthenticatedData encapsulating a PKIResponse content type. The certificates issued in a PKI Response are returned in the certificates field of the immediately encapsulating SignedData. Clients MUST NOT assume the certificates are in any order. Servers SHOULD include all intermediate certificates needed to form complete chains to one or more trust anchors, not just the newly issued certificate(s). The server MAY additionally return CRLs in the CRL bag. Servers MAY include self-signed certificates. Clients MUST NOT implicitly trust included self-signed certificate(s) merely due to its presence in the certificate bag. In the event clients receive a new self-signed certificate from the server, clients MAY provide a mechanism to enable the user to explicitly use the certificate as a trust anchor. (The Publish Trust Anchors control (Section 6.15) exists for the purpose of allowing for distribution of trust anchor certificates. If a trusted anchor publishes a new trusted anchor, this is one case where automated trust of the new trust anchor could be allowed.)4.2.1. PKIResponse Content Type
The PKIResponse content type is used for the Full PKI Response. The PKIResponse content type is identified by: id-cct-PKIResponse ::= {id-pkix id-cct(12) 3 } The ASN.1 structure corresponding to the PKIResponse content type is: PKIResponse ::= SEQUENCE { controlSequence SEQUENCE SIZE(0..MAX) OF TaggedAttribute, cmsSequence SEQUENCE SIZE(0..MAX) OF TaggedContentInfo, otherMsgSequence SEQUENCE SIZE(0..MAX) OF OtherMsg } ReponseBody ::= PKIResponse Note: In [RFC2797], this ASN.1 type was named ResponseBody. It has been renamed to PKIResponse for clarity and the old name kept as a synonym.
The fields in PKIResponse have the following meaning:
controlSequence is a sequence of controls. The controls defined in
this document are found in Section 6. Controls can be defined by
other parties. Details on the TaggedAttribute structure are found
in Section 3.2.1.1.
cmsSequence is a sequence of [CMS] message objects. See
Section 3.2.1.3 for more details.
otherMsgSequence is a sequence of arbitrary data objects. Data
objects placed here are referred to by one or more controls. This
allows for controls to use large amounts of data without the data
being embedded in the control. See Section 3.2.1.4 for more
details.
Processing of PKIResponse by a recipient is as follows:
1. All controls should be examined and processed in an appropriate
manner. The appropriate processing is to complete processing at
this time, to ignore the control, or to place the control on a
to-do list for later processing.
2. Additional processing of non-element items includes the saving of
certificates and CRLs present in wrapping layers. This type of
processing is based on the consumer of the element and should not
be relied on by generators.
No processing is required for cmsSequence or otherMsgSequence members
of the PKIResponse, if items are present and are not referenced by a
control. In this case, the cmsSequence and otherMsgSequence members
are to be ignored.
5. Application of Encryption to a PKI Request/Response
There are occasions when a PKI Request or Response must be encrypted
in order to prevent disclosure of information in the PKI Request/
Response from being accessible to unauthorized entities. This
section describes the means to encrypt Full PKI Requests and
Responses (Simple PKI Requests cannot be encrypted). Data portions
of PKI Requests and Responses that are placed in the cmsSequence
field can be encrypted separately.
Confidentiality is provided by wrapping the PKI Request/Response (a
SignedData) in an EnvelopedData. The nested content type in the
EnvelopedData is id-SignedData. Note that this is different from
S/MIME where there is a MIME layer placed between the encrypted and
signed data. It is recommended that if an EnvelopedData layer is
applied to a PKI Request/Response, a second signature layer be placed outside of the EnvelopedData layer. The following figure shows how this nesting would be done: Normal Option 1 Option 2 ------ -------- -------- SignedData EnvelopedData SignedData PKIData SignedData EnvelopedData PKIData SignedData PKIData Note: PKIResponse can be substituted for PKIData in the above figure. Options 1 and 2 prevent leakage of sensitive data by encrypting the Full PKI Request/Response. An RA that receives a PKI Request that it cannot decrypt MAY reject the PKI Request unless it can process the PKI Request without knowledge of the contents (i.e., all it does is amalgamate multiple PKI Requests and forward them to a server). After the RA removes the envelope and completes processing, it may then apply a new EnvelopedData layer to protect PKI Requests for transmission to the next processing agent. Section 7 contains more information about RA processing. Full PKI Requests/Responses can be encrypted or transmitted in the clear. Servers MUST provide support for all three options. Alternatively, an authenticated, secure channel could exist between the parties that require confidentiality. Clients and servers MAY use such channels instead of the technique described above to provide secure, private communication of Simple and Full PKI Requests/ Responses.6. Controls
Controls are carried as part of both Full PKI Requests and Responses. Each control is encoded as a unique OID followed by the data for the control (see syntax in Section 3.2.1.1. The encoding of the data is based on the control. Processing systems would first detect the OID (TaggedAttribute attrType) and process the corresponding control value (TaggedAttribute attrValues) prior to processing the message body.
The OIDs are all defined under the following arc:
id-pkix OBJECT IDENTIFIER ::= { iso(1) identified-organization(3)
dod(6) internet(1) security(5) mechanisms(5) pkix(7) }
id-cmc OBJECT IDENTIFIER ::= { id-pkix 7 }
The following table lists the names, OID, and syntactic structure for
each of the controls described in this document.
Identifier Description OID ASN.1 Structure Section
--------------------------------------------------------------------
id-cmc-statusInfo id-cmc 1 CMCStatusInfo 6.1.2
id-cmc-identification id-cmc 2 UTF8String 6.2.3
id-cmc-identityProof id-cmc 3 OCTET STRING 6.2.2
id-cmc-dataReturn id-cmc 4 OCTET STRING 6.4
id-cmc-transactionId id-cmc 5 INTEGER 6.6
id-cmc-senderNonce id-cmc 6 OCTET STRING 6.6
id-cmc-recipientNonce id-cmc 7 OCTET STRING 6.6
id-cmc-addExtensions id-cmc 8 AddExtensions 6.5.2
id-cmc-encryptedPOP id-cmc 9 EncryptedPOP 6.7
id-cmc-decryptedPOP id-cmc 10 DecryptedPOP 6.7
id-cmc-lraPOPWitness id-cmc 11 LraPOPWitness 6.8
id-cmc-getCert id-cmc 15 GetCert 6.9
id-cmc-getCRL id-cmc 16 GetCRL 6.10
id-cmc-revokeRequest id-cmc 17 RevokeRequest 6.11
id-cmc-regInfo id-cmc 18 OCTET STRING 6.12
id-cmc-responseInfo id-cmc 19 OCTET STRING 6.12
id-cmc-queryPending id-cmc 21 OCTET STRING 6.13
id-cmc-popLinkRandom id-cmc 22 OCTET STRING 6.3.1
id-cmc-popLinkWitness id-cmc 23 OCTET STRING 6.3.1
id-cmc-popLinkWitnessV2 id-cmc 33 OCTET STRING 6.3.1.1
id-cmc-confirmCertAcceptance id-cmc 24 CMCCertId 6.14
id-cmc-statusInfoV2 id-cmc 25 CMCStatusInfoV2 6.1.1
id-cmc-trustedAnchors id-cmc 26 PublishTrustAnchors 6.15
id-cmc-authData id-cmc 27 AuthPublish 6.16
id-cmc-batchRequests id-cmc 28 BodyPartList 6.17
id-cmc-batchResponses id-cmc 29 BodyPartList 6.17
id-cmc-publishCert id-cmc 30 CMCPublicationInfo 6.18
id-cmc-modCertTemplate id-cmc 31 ModCertTemplate 6.5.1
id-cmc-controlProcessed id-cmc 32 ControlsProcessed 6.19
id-cmc-identityProofV2 id-cmc 34 IdentityProofV2 6.2.1
Table 1: CMC Control Attributes
6.1. CMC Status Info Controls
The CMC Status Info controls return information about the status of a client/server request/response. Two controls are described in this section. The Extended CMC Status Info control is the preferred control; the CMC Status Info control is included for backwards compatibility with RFC 2797. Servers MAY emit multiple CMC status info controls referring to a single body part. Clients MUST be able to deal with multiple CMC status info controls in a PKI Response. Servers MUST use the Extended CMC Status Info control, but MAY additionally use the CMC Status Info control. Clients MUST be able to process the Extended CMC Status Info control.6.1.1. Extended CMC Status Info Control
The Extended CMC Status Info control is identified by the OID: id-cmc-statusInfoV2 ::= { id-cmc 25 } The Extended CMC Status Info control has the ASN.1 definition: CMCStatusInfoV2 ::= SEQUENCE { cMCStatus CMCStatus, bodyList SEQUENCE SIZE (1..MAX) OF BodyPartReference, statusString UTF8String OPTIONAL, otherInfo OtherStatusInfo OPTIONAL } OtherStatusInfo ::= CHOICE { failInfo CMCFailInfo, pendInfo PendInfo, extendedFailInfo ExtendedFailInfo } PendInfo ::= SEQUENCE { pendToken OCTET STRING, pendTime GeneralizedTime } ExtendedFailInfo ::= SEQUENCE { failInfoOID OBJECT IDENTIFIER, failInfoValue ANY DEFINED BY failInfoOID }
BodyPartReference ::= CHOICE {
bodyPartID BodyPartID,
bodyPartPath BodyPartPath
}
The fields in CMCStatusInfoV2 have the following meaning:
cMCStatus contains the returned status value. Details are in
Section 6.1.3.
bodyList identifies the controls or other elements to which the
status value applies. If an error is returned for a Simple PKI
Request, this field is the bodyPartID choice of BodyPartReference
with the single integer of value 1.
statusString contains additional description information. This
string is human readable.
otherInfo contains additional information that expands on the CMC
status code returned in the cMCStatus field.
The fields in OtherStatusInfo have the following meaning:
failInfo is described in Section 6.1.4. It provides an error code
that details what failure occurred. This choice is present only
if cMCStatus contains the value failed.
pendInfo contains information about when and how the client should
request the result of this request. It is present when the
cMCStatus is either pending or partial. pendInfo uses the
structure PendInfo, which has the fields:
pendToken is the token used in the Query Pending control
(Section 6.13).
pendTime contains the suggested time the server wants to be
queried about the status of the certification request.
extendedFailInfo includes application-dependent detailed error
information. This choice is present only if cMCStatus contains
the value failed. Caution should be used when defining new values
as they may not be correctly recognized by all clients and
servers. The CMCFailInfo value of internalCAError may be assumed
if the extended error is not recognized. This field uses the type
ExtendedFailInfo. ExtendedFailInfo has the fields:
failInfoOID contains an OID that is associated with a set of
extended error values.
failInfoValue contains an extended error code from the defined
set of extended error codes.
If the cMCStatus field is success, the Extended CMC Status Info
control MAY be omitted unless it is the only item in the response.
6.1.2. CMC Status Info Control
The CMC Status Info control is identified by the OID:
id-cmc-statusInfo ::= { id-cmc 1 }
The CMC Status Info control has the ASN.1 definition:
CMCStatusInfo ::= SEQUENCE {
cMCStatus CMCStatus,
bodyList BodyPartList,
statusString UTF8String OPTIONAL,
otherInfo CHOICE {
failInfo CMCFailInfo,
pendInfo PendInfo } OPTIONAL
}
The fields in CMCStatusInfo have the following meaning:
cMCStatus contains the returned status value. Details are in
Section 6.1.3.
bodyList contains the list of controls or other elements to which
the status value applies. If an error is being returned for a
Simple PKI Request, this field contains a single integer of value
1.
statusString contains additional description information. This
string is human readable.
otherInfo provides additional information that expands on the CMC
status code returned in the cMCStatus field.
failInfo is described in Section 6.1.4. It provides an error
code that details what failure occurred. This choice is
present only if cMCStatus is failed.
pendInfo uses the PendInfo ASN.1 structure in Section 6.1.1. It
contains information about when and how the client should
request results of this request. The pendInfo field MUST be
populated for a cMCStatus value of pending or partial. Further
details can be found in Section 6.1.1 (Extended CMC Status Info
Control) and Section 6.13 (Query Pending Control ).
If the cMCStatus field is success, the CMC Status Info control MAY be
omitted unless it is the only item in the response. If no status
exists for a Simple or Full PKI Request, then the value of success is
assumed.
6.1.3. CMCStatus Values
CMCStatus is a field in the Extended CMC Status Info and CMC Status
Info controls. This field contains a code representing the success
or failure of a specific operation. CMCStatus has the ASN.1
structure:
CMCStatus ::= INTEGER {
success (0),
-- reserved (1),
failed (2),
pending (3),
noSupport (4),
confirmRequired (5),
popRequired (6),
partial (7)
}
The values of CMCStatus have the following meaning:
success indicates the request was granted or the action was
completed.
failed indicates the request was not granted or the action was not
completed. More information is included elsewhere in the
response.
pending indicates the PKI Request has yet to be processed. The
requester is responsible to poll back on this Full PKI request.
pending may only be returned for certification request operations.
noSupport indicates the requested operation is not supported.
confirmRequired indicates a Confirm Certificate Acceptance control
(Section 6.14) must be returned before the certificate can be
used.
popRequired indicates a direct POP operation is required
(Section 6.3.1.3).
partial indicates a partial PKI Response is returned. The requester
is responsible to poll back for the unfulfilled portions of the
Full PKI Request.
6.1.4. CMCFailInfo
CMCFailInfo is a field in the Extended CMC Status Info and CMC Status
Info controls. CMCFailInfo conveys more detailed information
relevant to the interpretation of a failure condition. The
CMCFailInfo has the following ASN.1 structure:
CMCFailInfo ::= INTEGER {
badAlg (0),
badMessageCheck (1),
badRequest (2),
badTime (3),
badCertId (4),
unsupportedExt (5),
mustArchiveKeys (6),
badIdentity (7),
popRequired (8),
popFailed (9),
noKeyReuse (10),
internalCAError (11),
tryLater (12),
authDataFail (13)
}
The values of CMCFailInfo have the following meanings:
badAlg indicates unrecognized or unsupported algorithm.
badMessageCheck indicates integrity check failed.
badRequest indicates transaction was not permitted or supported.
badTime indicates message time field was not sufficiently close to
the system time.
badCertId indicates no certificate could be identified matching the
provided criteria.
unsupportedExt indicates a requested X.509 extension is not
supported by the recipient CA.
mustArchiveKeys indicates private key material must be supplied.
badIdentity indicates identification control failed to verify.
popRequired indicates server requires a POP proof before issuing
certificate.
popFailed indicates POP processing failed.
noKeyReuse indicates server policy does not allow key reuse.
internalCAError indicates that the CA had an unknown internal
failure.
tryLater indicates that the server is not accepting requests at this
time and the client should try at a later time.
authDataFail indicates failure occurred during processing of
authenticated data.
If additional failure reasons are needed, they SHOULD use the
ExtendedFailureInfo item in the Extended CMC Status Info control.
However, for closed environments they can be defined using this type.
Such codes MUST be in the range from 1000 to 1999.
6.2. Identification and Identity Proof Controls
Some CAs and RAs require that a proof-of-identity be included in a
certification request. Many different ways of doing this exist with
different degrees of security and reliability. Most are familiar
with a bank's request to provide your mother's maiden name as a form
of identity proof. The reasoning behind requiring a proof-of-
identity can be found in Appendix C of [CRMF].
CMC provides a method to prove the client's identity based on a
client/server shared-secret. If clients support the Full PKI
Request, clients MUST implement this method of identity proof
(Section 6.2.2). Servers MUST provide this method, but MAY
additionally support bilateral methods of similar strength.
This document also provides an Identification control
(Section 6.2.3). This control is a simple method to allow a client
to state who they are to the server. Generally, a shared-secret AND
an identifier of that shared-secret are passed from the server to the
client. The identifier is placed in the Identification control, and
the shared-secret is used to compute the Identity Proof control.
6.2.1. Identity Proof Version 2 Control
The Identity Proof Version 2 control is identified by the OID:
id-cmc-identityProofV2 ::= { id-cmc 34 }
The Identity Proof Version 2 control has the ASN.1 definition:
IdentifyProofV2 ::= SEQUENCE {
hashAlgID AlgorithmIdentifier,
macAlgID AlgorithmIdentifier,
witness OCTET STRING
}
The fields of IdentityProofV2 have the following meaning:
hashAlgID is the identifier and parameters for the hash algorithm
used to convert the shared-secret into a key for the MAC
algorithm.
macAlgID is the identifier and the parameters for the message
authentication code algorithm used to compute the value of the
witness field.
witness is the identity proof.
The required method starts with an out-of-band transfer of a token
(the shared-secret). The shared-secret should be generated in a
random manner. The distribution of this token is beyond the scope of
this document. The client then uses this token for an identity proof
as follows:
1. The PKIData reqSequence field (encoded exactly as it appears in
the Full PKI Request including the sequence type and length) is
the value to be validated.
2. A hash of the shared-secret as a UTF8 string is computed using
hashAlgID.
3. A MAC is then computed using the value produced in Step 1 as the
message and the value from Step 2 as the key.
4. The result from Step 3 is then encoded as the witness value in
the Identity Proof Version 2 control.
When the server verifies the Identity Proof Version 2 control, it
computes the MAC value in the same way and compares it to the witness
value contained in the PKI Request.
If a server fails the verification of an Identity Proof Version 2
control, the CMCFailInfo value MUST be present in the Full PKI
Response and MUST have a value of badIdentity.
Reuse of the shared-secret on certification request retries allows the client and server to maintain the same view of acceptable identity proof values. However, reuse of the shared-secret can potentially open the door for some types of attacks. Implementations MUST be able to support tokens at least 16 characters long. Guidance on the amount of entropy actually obtained from a given length token based on character sets can be found in Appendix A of [PASSWORD].6.2.2. Identity Proof Control
The Identity Proof control is identified by the OID: id-cmc-identityProof ::= { id-cmc 3 } The Identity Proof control has the ASN.1 definition: IdentifyProof ::= OCTET STRING This control is processed in the same way as the Identity Proof Version 2 control. In this case, the hash algorithm is fixed to SHA-1 and the MAC algorithm is fixed to HMAC-SHA1.6.2.3. Identification Control
Optionally, servers MAY require the inclusion of the unprotected Identification control with an Identification Proof control. The Identification control is intended to contain a text string that assists the server in locating the shared-secret needed to validate the contents of the Identity Proof control. If the Identification control is included in the Full PKI Request, the derivation of the key in Step 2 (from Section 6.2.1) is altered so that the hash of the concatenation of the shared-secret and the UTF8 identity value (without the type and length bytes) are hashed rather than just the shared-secret. The Identification control is identified by the OID: id-cmc-identification ::= { id-cmc 2 } The Identification control has the ASN.1 definition: Identification ::= UTF8String
6.2.4. Hardware Shared-Secret Token Generation
The shared-secret between the EE and the server is sometimes computed using a hardware device that generates a series of tokens. The EE can therefore prove its identity by transferring this token in plain text along with a name string. The above protocol can be used with a hardware shared-secret token generation device by the following modifications: 1. The Identification control MUST be included and MUST contain the hardware-generated token. 2. The shared-secret value used above is the same hardware-generated token. 3. All certification requests MUST have a subject name, and the subject name MUST contain the fields required to identify the holder of the hardware token device. 4. The entire certification request MUST be shrouded in some fashion to prevent eavesdropping. Although the token is time critical, an active eavesdropper cannot be permitted to extract the token and submit a different certification request with the same token value.6.3. Linking Identity and POP Information
In a Full PKI Request, identity information about the client is carried in the signature of the SignedData containing all of the certification requests. Proof-of-possession information for key pairs, however, is carried separately for each PKCS #10 or CRMF certification request. (For keys capable of generating a digital signature, the POP is provided by the signature on the PKCS #10 or CRMF request. For encryption-only keys, the controls described in Section 6.7 are used.) In order to prevent substitution-style attacks, the protocol must guarantee that the same entity generated both the POP and proof-of-identity information. This section describes two mechanisms for linking identity and POP information: witness values cryptographically derived from the shared-secret (Section 6.3.1.3) and shared-secret/subject distinguished name (DN) matching (Section 6.3.2). Clients and servers MUST support the witness value technique. Clients and servers MAY support shared-secret/subject DN matching or other bilateral techniques of similar strength. The idea behind both mechanisms is to force the client to sign some data into each certification request that can be directly associated with the
shared-secret; this will defeat attempts to include certification requests from different entities in a single Full PKI Request.6.3.1. Cryptographic Linkage
The first technique that links identity and POP information forces the client to include a piece of information cryptographically derived from the shared-secret as a signed extension within each certification request (PKCS #10 or CRMF).6.3.1.1. POP Link Witness Version 2 Controls
The POP Link Witness Version 2 control is identified by the OID: id-cmc-popLinkWitnessV2 ::= { id-cmc 33 } The POP Link Witness Version 2 control has the ASN.1 definition: PopLinkWitnessV2 ::= SEQUENCE { keyGenAlgorithm AlgorithmIdentifier, macAlgorithm AlgorithmIdentifier, witness OCTET STRING } The fields of PopLinkWitnessV2 have the following meanings: keyGenAlgorithm contains the algorithm used to generate the key for the MAC algorithm. This will generally be a hash algorithm, but could be a more complex algorithm. macAlgorithm contains the algorithm used to create the witness value. witness contains the computed witness value. This technique is useful if null subject DNs are used (because, for example, the server can generate the subject DN for the certificate based only on the shared-secret). Processing begins when the client receives the shared-secret out-of-band from the server. The client then computes the following values: 1. The client generates a random byte-string, R, which SHOULD be at least 512 bits in length. 2. The key is computed from the shared-secret using the algorithm in keyGenAlgorithm.
3. A MAC is then computed over the random value produced in Step 1,
using the key computed in Step 2.
4. The random value produced in Step 1 is encoded as the value of a
POP Link Random control. This control MUST be included in the
Full PKI Request.
5. The MAC value produced in Step 3 is placed in either the POP Link
Witness control or the witness field of the POP Link Witness V2
control.
* For CRMF, the POP Link Witness/POP Link Witness V2 control is
included in the controls field of the CertRequest structure.
* For PKCS #10, the POP Link Witness/POP Link Witness V2 control
is included in the attributes field of the
CertificationRequestInfo structure.
Upon receipt, servers MUST verify that each certification request
contains a copy of the POP Link Witness/POP Link Witness V2 control
and that its value was derived using the above method from the
shared-secret and the random string included in the POP Link Random
control.
The Identification control (see Section 6.2.3) or the subject DN of a
certification request can be used to help identify which shared-
secret was used.
6.3.1.2. POP Link Witness Control
The POP Link Witness control is identified by the OID:
id-cmc-popLinkWitness ::= { id-cmc 23 }
The POP Link Witness control has the ASN.1 definition:
PopLinkWitness ::= OCTET STRING
For this control, SHA-1 is used as the key generation algorithm.
HMAC-SHA1 is used as the mac algorithm.
6.3.1.3. POP Link Random Control
The POP Link Random control is identified by the OID:
id-cmc-popLinkRandom ::= { id-cmc 22 }
The POP Link Random control has the ASN.1 definition:
PopLinkRandom ::= OCTET STRING
6.3.2. Shared-Secret/Subject DN Linking
The second technique to link identity and POP information is to link
a particular subject distinguished name (subject DN) to the shared-
secrets that are distributed out-of-band and to require that clients
using the shared-secret to prove identity include that exact subject
DN in every certification request. It is expected that many client-
server connections that use shared-secret-based proof-of-identity
will use this mechanism. (It is common not to omit the subject DN
information from the certification request.)
When the shared-secret is generated and transferred out-of-band to
initiate the registration process (Section 6.2), a particular subject
DN is also associated with the shared-secret and communicated to the
client. (The subject DN generated MUST be unique per entity in
accordance with the CA policy; a null subject DN cannot be used. A
common practice could be to place the identification value as part of
the subject DN.) When the client generates the Full PKI Request, it
MUST use these two pieces of information as follows:
1. The client MUST include the specific subject DN that it received
along with the shared-secret as the subject name in every
certification request (PKCS #10 and/or CRMF) in the Full PKI
Request. The subject names in the certification requests MUST
NOT be null.
2. The client MUST include an Identity Proof control (Section 6.2.2)
or Identity Proof Version 2 control (Section 6.2.1), derived from
the shared-secret, in the Full PKI Request.
The server receiving this message MUST (a) validate the Identity
Proof control and then, (b) check that the subject DN included in
each certification request matches that associated with the shared-
secret. If either of these checks fails, the certification request
MUST be rejected.
6.3.3. Renewal and Rekey Messages
When doing a renewal or rekey certification request, linking identity
and POP information is simple. The client copies the subject DN for
a current signing certificate into the subject name field of each
certification request that is made. The POP for each certification
request will now cover that information. The outermost signature
layer is created using the current signing certificate, which allows
the original identity to be associated with the certification request. Since the name in the current signing certificate and the names in the certification requests match, the necessary linking has been achieved.6.4. Data Return Control
The Data Return control allows clients to send arbitrary data (usually some type of internal state information) to the server and to have the data returned as part of the Full PKI Response. Data placed in a Data Return control is considered to be opaque to the server. The same control is used for both Full PKI Requests and Responses. If the Data Return control appears in a Full PKI Request, the server MUST return it as part of the PKI Response. In the event that the information in the Data Return control needs to be confidential, it is expected that the client would apply some type of encryption to the contained data, but the details of this are outside the scope of this specification. The Data Return control is identified by the OID: id-cmc-dataReturn ::= { id-cmc 4 } The Data Return control has the ASN.1 definition: DataReturn ::= OCTET STRING A client could use this control to place an identifier marking the exact source of the private key material. This might be the identifier of a hardware device containing the private key.6.5. RA Certificate Modification Controls
These controls exist for RAs to be able to modify the contents of a certification request. Modifications might be necessary for various reasons. These include addition of certificate extensions or modification of subject and/or subject alternative names. Two controls exist for this purpose. The first control, Modify Certification Request (Section 6.5.1), allows the RA to replace or remove any field in the certificate. The second control, Add Extensions (Section 6.5.2), only allows for the addition of extensions.
6.5.1. Modify Certification Request Control
The Modify Certification Request control is used by RAs to change fields in a requested certificate. The Modify Certification Request control is identified by the OID: id-cmc-modCertTemplate ::= { id-cmc 31 } The Modify Certification Request has the ASN.1 definition: ModCertTemplate ::= SEQUENCE { pkiDataReference BodyPartPath, certReferences BodyPartList, replace BOOLEAN DEFAULT TRUE, certTemplate CertTemplate } The fields in ModCertTemplate have the following meaning: pkiDataReference is the path to the PKI Request containing certification request(s) to be modified. certReferences refers to one or more certification requests in the PKI Request referenced by pkiDataReference to be modified. Each BodyPartID of the certReferences sequence MUST be equal to either the bodyPartID of a TaggedCertificationRequest (PKCS #10) or the certReqId of the CertRequest within a CertReqMsg (CRMF). By definition, the certificate extensions included in the certTemplate field are applied to every certification request referenced in the certReferences sequence. If a request corresponding to bodyPartID cannot be found, the CMCFailInfo with a value of badRequest is returned that references this control. replace specifies if the target certification request is to be modified by replacing or deleting fields. If the value is TRUE, the data in this control replaces the data in the target certification request. If the value is FALSE, the data in the target certification request is deleted. The action is slightly different for the extensions field of certTemplate; each extension is treated individually rather than as a single unit. certTemplate is a certificate template object [CRMF]. If a field is present and replace is TRUE, it replaces that field in the certification request. If the field is present and replace is FALSE, the field in the certification request is removed. If the field is absent, no action is performed. Each extension is treated as a single field.
Servers MUST be able to process all extensions defined, but not prohibited, in [PKIXCERT]. Servers are not required to be able to process every X.509v3 extension transmitted using this protocol, nor are they required to be able to process other, private extensions. Servers are not required to put all RA-requested extensions into a certificate. Servers are permitted to modify RA-requested extensions. Servers MUST NOT alter an extension so as to reverse the meaning of a client-requested extension. If a certification request is denied due to the inability to handle a requested extension and a Full PKI Response is returned, the server MUST return a CMCFailInfo value with the value of unsupportedExt. If a certification request is the target of multiple Modify Certification Request controls, the behavior is: o If control A exists in a layer that contains the layer of control B, control A MUST override control B. In other words, controls should be applied from the innermost layer to the outermost layer. o If control A and control B are in the same PKIData (i.e., the same wrapping layer), the order of application is non-determinate. The same order of application is used if a certification request is the target of both a Modify Certification Request control and an Add Extensions control.6.5.2. Add Extensions Control
The Add Extensions control has been deprecated in favor of the Modify Certification Request control. It was replaced so that fields in the certification request other than extensions could be modified. The Add Extensions control is used by RAs to specify additional extensions that are to be included in certificates. The Add Extensions control is identified by the OID: id-cmc-addExtensions ::= { id-cmc 8 } The Add Extensions control has the ASN.1 definition: AddExtensions ::= SEQUENCE { pkiDataReference BodyPartID, certReferences SEQUENCE OF BodyPartID, extensions SEQUENCE OF Extension }
The fields in AddExtensions have the following meaning:
pkiDataReference contains the body part identity of the embedded
certification request.
certReferences is a list of references to one or more of the
certification requests contained within a PKIData. Each body part
identifier of the certReferences sequence MUST be equal to either
the bodyPartID of a TaggedCertificationRequest (PKCS #10) or the
certReqId of the CertRequest within a CertReqMsg (CRMF). By
definition, the listed extensions are to be applied to every
certification request referenced in the certReferences sequence.
If a certification request corresponding to bodyPartID cannot be
found, the CMCFailInfo with a value of badRequest is returned
referencing this control.
extensions is a sequence of extensions to be applied to the
referenced certification requests.
Servers MUST be able to process all extensions defined, but not
prohibited, in [PKIXCERT]. Servers are not required to be able to
process every X.509v3 extension transmitted using this protocol, nor
are they required to be able to process other, private extensions.
Servers are not required to put all RA-requested extensions into a
certificate. Servers are permitted to modify RA-requested
extensions. Servers MUST NOT alter an extension so as to reverse the
meaning of a client-requested extension. If a certification request
is denied due to the inability to handle a requested extension and a
response is returned, the server MUST return a CMCFailInfo with the
value of unsupportedExt.
If multiple Add Extensions controls exist in a Full PKI Request, the
exact behavior is left up to the CA policy. However, it is
recommended that the following policy be used. These rules would be
applied to individual extensions within an Add Extensions control (as
opposed to an "all or nothing" approach).
1. If the conflict is within a single PKIData, the certification
request would be rejected with a CMCFailInfo value of badRequest.
2. If the conflict is between different PKIData, the outermost
version of the extension would be used (allowing an RA to
override the requested extension).
6.6. Transaction Identifier Control and Sender and Recipient Nonce Controls
Transactions are identified and tracked with a transaction identifier. If used, clients generate transaction identifiers and retain their value until the server responds with a Full PKI Response that completes the transaction. Servers correspondingly include received transaction identifiers in the Full PKI Response. The Transaction Identifier control is identified by the OID: id-cmc-transactionId ::= { id-cmc 5 } The Transaction Identifier control has the ASN.1 definition: TransactionId ::= INTEGER The Transaction Identifier control identifies a given transaction. It is used by client and server to manage the state of an operation. Clients MAY include a Transaction Identifier control in a request. If the original request contains a Transaction Identifier control, all subsequent requests and responses MUST include the same Transaction Identifier control. Replay protection is supported through the use of the Sender and Recipient Nonce controls. If nonces are used, in the first message of a transaction, a Recipient Nonce control is not transmitted; a Sender Nonce control is included by the transaction originator and retained for later reference. The recipient of a Sender Nonce control reflects this value back to the originator as a Recipient Nonce control and includes its own Sender Nonce control. Upon receipt by the transaction originator of this response, the transaction originator compares the value of Recipient Nonce control to its retained value. If the values match, the message can be accepted for further security processing. The received value for a Sender Nonce control is also retained for inclusion in the next message associated with the same transaction. The Sender Nonce and Recipient Nonce controls are identified by the OIDs: id-cmc-senderNonce ::= { id-cmc 6 } id-cmc-recipientNonce ::= { id-cmc 7 } The Sender Nonce control has the ASN.1 definition: SenderNonce ::= OCTET STRING
The Recipient Nonce control has the ASN.1 definition:
RecipientNonce ::= OCTET STRING
Clients MAY include a Sender Nonce control in the initial PKI
Request. If a message includes a Sender Nonce control, the response
MUST include the transmitted value of the previously received Sender
Nonce control as a Recipient Nonce control and include a new value as
its Sender Nonce control.
6.7. Encrypted and Decrypted POP Controls
Servers MAY require that this POP method be used only if another POP
method is unavailable. Servers SHOULD reject all certification
requests contained within a PKIData if any required POP is missing
for any element within the PKIData.
Many servers require proof that the entity that generated the
certification request actually possesses the corresponding private
component of the key pair. For keys that can be used as signature
keys, signing the certification request with the private key serves
as a POP on that key pair. With keys that can only be used for
encryption operations, POP MUST be performed by forcing the client to
decrypt a value. See Section 5 of [CRMF] for a detailed discussion
of POP.
By necessity, POP for encryption-only keys cannot be done in one
round-trip, since there are four distinct steps:
1. Client tells the server about the public component of a new
encryption key pair.
2. Server sends the client a POP challenge, encrypted with the
presented public encryption key.
3. Client decrypts the POP challenge using the private key that
corresponds to the presented public key and sends the plaintext
back to the server.
4. Server validates the decrypted POP challenge and continues
processing the certification request.
CMC defines two different controls. The first deals with the
encrypted challenge sent from the server to the user in Step 2. The
second deals with the decrypted challenge sent from the client to the
server in Step 3.
The Encrypted POP control is used to send the encrypted challenge
from the server to the client as part of the PKIResponse. (Note that
it is assumed that the message sent in Step 1 above is a Full PKI
Request and that the response in Step 2 is a Full PKI Response
including a CMCFailInfo specifying that a POP is explicitly required,
and providing the POP challenge in the encryptedPOP control.)
The Encrypted POP control is identified by the OID:
id-cmc-encryptedPOP ::= { id-cmc 9 }
The Encrypted POP control has the ASN.1 definition:
EncryptedPOP ::= SEQUENCE {
request TaggedRequest,
cms ContentInfo,
thePOPAlgID AlgorithmIdentifier,
witnessAlgID AlgorithmIdentifier,
witness OCTET STRING
}
The Decrypted POP control is identified by the OID:
id-cmc-decryptedPOP ::= { id-cmc 10 }
The Decrypted POP control has the ASN.1 definition:
DecryptedPOP ::= SEQUENCE {
bodyPartID BodyPartID,
thePOPAlgID AlgorithmIdentifier,
thePOP OCTET STRING
}
The encrypted POP algorithm works as follows:
1. The server randomly generates the POP Proof Value and associates
it with the request.
2. The server returns the Encrypted POP control with the following
fields set:
request is the original certification request (it is included
here so the client need not keep a copy of the request).
cms is an EnvelopedData, the encapsulated content type being id-
data and the content being the POP Proof Value; this value
needs to be long enough that one cannot reverse the value from
the witness hash. If the certification request contains a
Subject Key Identifier (SKI) extension, then the recipient
identifier SHOULD be the SKI. If the issuerAndSerialNumber
form is used, the IssuerName MUST be encoded as NULL and the
SerialNumber as the bodyPartID of the certification request.
thePOPAlgID identifies the algorithm to be used in computing the
return POP value.
witnessAlgID identifies the hash algorithm used on the POP Proof
Value to create the field witness.
witness is the hashed value of the POP Proof Value.
3. The client decrypts the cms field to obtain the POP Proof Value.
The client computes H(POP Proof Value) using the witnessAlgID and
compares to the value of witness. If the values do not compare
or the decryption is not successful, the client MUST abort the
enrollment process. The client aborts the process by sending a
request containing a CMC Status Info control with CMCFailInfo
value of popFailed.
4. The client creates the Decrypted POP control as part of a new
PKIData. The fields in the DecryptedPOP are:
bodyPartID refers to the certification request in the new PKI
Request.
thePOPAlgID is copied from the encryptedPOP.
thePOP contains the possession proof. This value is computed by
thePOPAlgID using the POP Proof Value and the request.
5. The server then re-computes the value of thePOP from its cached
value and the request and compares to the value of thePOP. If
the values do not match, the server MUST NOT issue the
certificate. The server MAY re-issue a new challenge or MAY fail
the request altogether.
When defining the algorithms for thePOPAlgID and witnessAlgID, care
must be taken to ensure that the result of witnessAlgID is not a
useful value to shortcut the computation with thePOPAlgID. The POP
Proof Value is used as the secret value in the HMAC algorithm and the
request is used as the data. If the POP Proof Value is greater than
64 bytes, only the first 64 bytes of the POP Proof Value is used as
the secret.
One potential problem with the algorithm above is the amount of state
that a CA needs to keep in order to verify the returned POP value.
The following describes one of many possible ways of addressing the
problem by reducing the amount of state kept on the CA to a single
(or small set) of values.
1. Server generates random seed x, constant across all requests.
(The value of x would normally be altered on a regular basis and
kept for a short time afterwards.)
2. For certification request R, server computes y = F(x,R). F can
be, for example, HMAC-SHA1(x,R). All that's important for
statelessness is that y be consistently computable with only
known state constant x and function F, other inputs coming from
the certification request structure. y should not be predictable
based on knowledge of R, thus the use of a one-way function like
HMAC-SHA1.
6.8. RA POP Witness Control
In a certification request scenario that involves an RA, the CA may
allow (or require) that the RA perform the POP protocol with the
entity that generated the certification request. In this case, the
RA needs a way to inform the CA that it has done the POP. The RA POP
Witness control addresses this issue.
The RA POP Witness control is identified by the OID:
id-cmc-lraPOPWitness ::= { id-cmc 11 }
The RA POP Witness control has the ASN.1 definition:
LraPopWitness ::= SEQUENCE {
pkiDataBodyid BodyPartID,
bodyIds SEQUENCE of BodyPartID
}
The fields in LraPOPWitness have the following meaning:
pkiDataBodyid contains the body part identifier of the nested
TaggedContentInfo containing the client's Full PKI Request.
pkiDataBodyid is set to 0 if the request is in the current
PKIData.
bodyIds is a list of certification requests for which the RA has
performed an out-of-band authentication. The method of
authentication could be archival of private key material,
challenge-response, or other means.
If a certification server does not allow an RA to do the POP verification, it returns a CMCFailInfo with the value of popFailed. The CA MUST NOT start a challenge-response to re-verify the POP itself.6.9. Get Certificate Control
Everything described in this section is optional to implement. The Get Certificate control is used to retrieve a previously issued certificate from a certificate repository. A CA, an RA, or an independent service may provide this repository. The clients expected to use this facility are those where a fully deployed directory is either infeasible or undesirable. The Get Certificate control is identified by the OID: id-cmc-getCert ::= { id-cmc 15 } The Get Certificate control has the ASN.1 definition: GetCert ::= SEQUENCE { issuerName GeneralName, serialNumber INTEGER } The fields in GetCert have the following meaning: issuerName is the name of the certificate issuer. serialNumber identifies the certificate to be retrieved. The server that responds to this request places the requested certificate in the certificates field of a SignedData. If the Get Certificate control is the only control in a Full PKI Request, the response should be a Simple PKI Response.
(page 49 continued on part 3)
