Internet Draft                                 Editor:  Blake Ramsdell,
draft-ietf-smime-cert-03.txt                   Worldtalk
March 24, 1998
Expires in six months


                 S/MIME Version 3 Certificate Handling

Status of this memo

This document is an Internet-Draft. 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."

To view the entire list of current Internet-Drafts, please check
the "1id-abstracts.txt" listing contained in the Internet-Drafts
Shadow Directories on ftp.is.co.za (Africa), ftp.nordu.net
(Northern Europe), ftp.nis.garr.it (Southern Europe), munnari.oz.au
(Pacific Rim), ftp.ietf.org (US East Coast), or ftp.isi.edu
(US West Coast).


1. Overview

S/MIME (Secure/Multipurpose Internet Mail Extensions), described in
[SMIME-MSG], provides a method to send and receive secure MIME
messages. Before using a public key to provide security services, the
S/MIME agent MUST certify that the public key is valid. S/MIME agents
MUST use X.509 certificates to validate public keys as described in
the ITU-T Recommendation X.509 (1997) [X.509] and the Internet Public
Key Infrastructure (PKIX) X.509 Certificate and CRL Profile [KEYM].
S/MIME agents MUST meet the S/MIME-specific requirements documented in
this I-D in addition to those stated in [X.509] and [KEYM].

This specification is compatible with the Cryptographic Message Syntax
[CMS] in that it uses the data types defined by CMS. It also inherits
all the varieties of architectures for certificate-based key
management supported by CMS. Note that the method S/MIME messages make
certificate requests is defined in [SMIME-MSG].



1.1 Definitions

For the purposes of this draft, the following definitions apply.

ASN.1: Abstract Syntax Notation One, as defined in ITU-T X.680-689.

Attribute Certificate (AC): An X.509 AC is a separate structure from a
subject's public key X.509 Certificate.  A subject may have multiple
X.509 ACs associated with each of its public key X.509 Certificates.
Each X.509 AC binds a SEQUENCE OF Attributes with one of the subject's
public key X.509 Certificates.  The X.509 AC syntax is defined in
[X.509]

BER: Basic Encoding Rules for ASN.1, as defined in ITU-T X.690.

Certificate: A type that binds an entity's distinguished name to a
public key with a digital signature. This type is defined in ITU-T
X.509 [X.509]. This type also contains the distinguished name of the
certificate issuer (the signer), an issuer-specific serial number, the
issuer's signature algorithm identifier, a validity period, and
extensions as defined in [KEYM].

Certificate Revocation List (CRL): A type that contains information
about certificates whose validity an issuer has prematurely revoked.
The information consists of an issuer name, the time of issue, the
next scheduled time of issue, a list of certificate serial numbers and
their associated revocation times, and extensions as defined in
[KEYM]. The CRL is signed by the issuer. The type intended by this
specification is the one defined in [KEYM].

DER: Distinguished Encoding Rules for ASN.1, as defined in ITU-T
X.690.


1.2 Compatibility with Prior Practice of S/MIME

S/MIME version 3 agents should attempt to have the greatest
interoperability possible with S/MIME version 2 agents. S/MIME version
2 is described in RFC 2311 through RFC 2315, inclusive. RFC 2311 also
has historical information about the development of S/MIME.


1.3 Terminology

Throughout this draft, the terms MUST, MUST NOT, SHOULD, and SHOULD
NOT are used in capital letters. This conforms to the definitions in
[MUSTSHOULD]. [MUSTSHOULD] defines the use of these key words to help
make the intent of standards track documents as clear as possible. The
same key words are used in this document to help implementors achieve
interoperability.


1.4 Discussion of This Draft

This draft is being discussed on the "ietf-smime" mailing list.
To subscribe, send a message to:
     ietf-smime-request@imc.org
with the single word
     subscribe
in the body of the message. There is a Web site for the mailing list
at <http://www.imc.org/ietf-smime/>.


2. CMS Options

The CMS message format allows for a wide variety of options in content
and algorithm support. This section puts forth a number of support
requirements and recommendations in order to achieve a base level of
interoperability among all S/MIME implementations. Most of the CMS
format for S/MIME messages is defined in [SMIME-MSG].


2.1 CertificateRevocationLists

Receiving agents MUST support for the Certificate Revocation List
(CRL) format defined in [KEYM]. If sending agents include CRLs in
outgoing messages, the CRL format defined in [KEYM] MUST be used.

All agents MUST validate CRLs and check certificates against CRLs, if
available, in accordance with [KEYM].

Receiving agents MUST recognize CRLs in received S/MIME messages.

Clients MUST use revocation information included as a CRL in an S/MIME
message when verifying the signature and certificate path validity in
that message.  Clients SHOULD store CRLs received in messages for use
in processing later messages.

Clients MUST handle multiple valid Certificate Authority (CA)
certificates containing the same subject name and the same public keys
but with overlapping validity intervals.


2.2 CertificateChoices

Receiving agents MUST support X.509 v1 and X.509 v3 certificates. See
[KEYM] for details about the profile for certificate formats. End
entity certificates MAY include an Internet mail address, as described
in section 3.1.

Receiving agents SHOULD support X.509 attribute certificates.


2.2.1 Historical Note About CMS Certificates

The CMS message format supports a choice of certificate formats for
public key content types: X.509, PKCS #6 Extended Certificates and
X.509 Attribute Certificates. The PKCS #6 format is not in widespread
use. In addition, X.509 v3 certificate extensions address much of the
same functionality and flexibility as was intended in the PKCS #6.
Thus, sending and receiving agents MUST NOT use PKCS #6 extended
certificates.


2.3 CertificateSet

Receiving agents MUST be able to handle an arbitrary number of
certificates of arbitrary relationship to the message sender and to
each other in arbitrary order. In many cases, the certificates
included in a signed message may represent a chain of certification
from the sender to a particular root. There may be, however,
situations where the certificates in a signed message may be unrelated
and included for convenience.

Sending agents SHOULD include any certificates for the user's public
key(s) and associated issuer certificates. This increases the
likelihood that the intended recipient can establish trust in the
originator's public key(s). This is especially important when sending
a message to recipients that may not have access to the sender's
public key through any other means or when sending a signed message to
a new recipient. The inclusion of certificates in outgoing messages
can be omitted if S/MIME objects are sent within a group of
correspondents that has established access to each other's
certificates by some other means such as a shared directory or manual
certificate distribution. Receiving S/MIME agents SHOULD be able to
handle messages without certificates using a database or directory
lookup scheme.

A sending agent SHOULD include at least one chain of certificates up
to, but not including, a Certificate Authority (CA) that it believes
that the recipient may trust as authoritative. A receiving agent
SHOULD be able to handle an arbitrarily large number of certificates
and chains.

Clients MAY send CA certificates, that is, certificates that are self-
signed and can be considered the "root" of other chains. Note that
receiving agents SHOULD NOT simply trust any self-signed certificates
as valid CAs, but SHOULD use some other mechanism to determine if this
is a CA that should be trusted.

Receiving agents MUST support chaining based on the distinguished name
fields. Other methods of building certificate chains may be supported
but are not currently recommended.

Receiving agents SHOULD support X.509 attribute certificates.  At a
minimum, receiving agents SHOULD at least support the decoding of
X.509 attribute certificates.  Please note that there is no
requirement that the same CA create both the public key X.509
Certificate and X.509 attribute certificate(s) for a user.  Each
organization's local policy will define how X.509 attribute
certificates are validated and used.  The implications of performing
multiple certification path validations should be considered when
defining local policy.  Exchanges between a subject and the CA dealing
with the generation of X.509 attribute certificates are outside the
scope of this specification.


3. Distinguished Names in Certificates


3.1 Using Distinguished Names for Internet Mail

The format of an X.509 certificate includes fields for the subject
name and issuer name. The subject name identifies the owner of a
particular public key/private key pair while the issuer name is meant
to identify the entity that "certified" the subject (that is, who
signed the subject's certificate). The subject name and issuer name
are defined by X.509 as Distinguished Names.

Distinguished Names are defined by a ITU-T standard X.501 [X.501]. A
Distinguished Name is broken into one or more Relative Distinguished
Names. Each Relative Distinguished Name is comprised of one or more
Attribute-Value Assertions. Each Attribute-Value Assertion consists of
a Attribute Identifier and its corresponding value information, such
as CountryName=US. Distinguished Names were intended to identify
entities in the X.500 directory tree [X.500]. Each Relative
Distinguished Name can be thought of as a node in the tree which is
described by some collection of Attribute-Value Assertions. The entire
Distinguished Name is some collection of nodes in the tree that
traverse a path from the root of the tree to some end node which
represents a particular entity.

The goal of the directory was to provide an infrastructure to uniquely
name every communications entity everywhere. However, adoption of a
global X.500 directory infrastructure has been slower than expected.
Consequently, there is no requirement for X.500 directory service
provision in the S/MIME environment, although such provision would
almost undoubtedly be of great value in facilitating key management
for S/MIME.

The use of Distinguished Names in accordance with the X.500 directory
is not very widespread. By contrast, Internet mail addresses, as
described in RFC 822 [RFC-822], are used almost exclusively in the
Internet environment to identify originators and recipients of
messages. However, Internet mail addresses bear no resemblance to
X.500 Distinguished Names (except, perhaps, that they are both
hierarchical in nature). Some method is needed to map Internet mail
addresses to entities that hold public keys. Some people have
suggested that the X.509 certificate format should be abandoned in
favor of other binding mechanisms. Instead, S/MIME keeps the X.509
certificate and Distinguished Name mechanisms while tailoring the
content of the naming information to suit the Internet mail
environment.

End-entity certificates MAY contain an Internet mail address as
described in [RFC-822]. The address must be an "addr-spec" as defined
in Section 6.1 of that specification.  The email address SHOULD be in
the subjectAltName extension, and SHOULD NOT be in the subject
distinguished name.

Receiving agents MUST recognize email addresses in the subjectAltName
field. Receiving agents MUST recognize email addresses in the
Distinguished Name field.

Sending agents SHOULD make the address in the From or Sender header in
a mail message match an Internet mail address in the signer's
certificate. Receiving agents MUST check that the address in the From
or Sender header of a mail message matches an Internet mail address in
the signer's certificate, if mail addresses are present in the
certificate. A receiving agent SHOULD provide some explicit alternate
processing of the message if this comparison fails, which may be to
display a message that shows the recipient the addresses in the
certificate or other certificate details.

All subject and issuer names MUST be populated (i.e. not an empty
SEQUENCE) in S/MIME-compliant v3 X.509 Certificates, except that the
subject DN in a user's (i.e. end-entity) certificate MAY be an empty
SEQUENCE in which case the subjectAltName extension will include the
subject's identifier and MUST be marked as critical.


3.2 Required Name Attributes

Receiving agents MUST support parsing of zero, one, or more instances
of each of the following set of name attributes within the
Distinguished Names in certificates.

Guidelines for the inclusion, omission, and ordering of name
attributes during the creation of a distinguished name will most
likely be dictated by the policies associated with the certification
service which will certify the corresponding name and public key.

countryName
stateOrProvinceName
localityName
commonName
title
organizationName
organizationalUnitName
streetAddress
postalCode
telephoneNumber
emailAddress

All attributes other than emailAddress are described in X.520 [X.520].
emailAddress is an IA5String that can have multiple attribute values.


4. Certificate Processing

A receiving agent needs to provide some certificate retrieval
mechanism in order to gain access to certificates for recipients of
digital envelopes. There are many ways to implement certificate
retrieval mechanisms. X.500 directory service is an excellent example
of a certificate retrieval-only mechanism that is compatible with
classic X.500 Distinguished Names. The PKIX Working Group is
investigating other mechanisms. Another method under consideration by
the IETF is to provide certificate retrieval services as part of the
existing Domain Name System (DNS). Until such mechanisms are widely
used, their utility may be limited by the small number of
correspondent's certificates that can be retrieved. At a minimum, for
initial S/MIME deployment, a user agent could automatically generate a
message to an intended recipient requesting that recipient's
certificate in a signed return message.

Receiving and sending agents SHOULD also provide a mechanism to allow
a user to "store and protect" certificates for correspondents in such
a way so as to guarantee their later retrieval. In many environments,
it may be desirable to link the certificate retrieval/storage
mechanisms together in some sort of certificate database. In its
simplest form, a certificate database would be local to a particular
user and would function in a similar way as a "address book" that
stores a user's frequent correspondents. In this way, the certificate
retrieval mechanism would be limited to the certificates that a user
has stored (presumably from incoming messages).  A comprehensive
certificate retrieval/storage solution may combine two or more
mechanisms to allow the greatest flexibility and utility to the user.
For instance, a secure Internet mail agent may resort
to checking a centralized certificate retrieval mechanism for a
certificate if it can not be found in a user's local certificate
storage/retrieval database.

Receiving and sending agents SHOULD provide a mechanism for the import
and export of certificates, using a CMS certs-only message. This
allows for import and export of full certificate chains as opposed to
just a single certificate. This is described in [SMIME-MSG].


4.1 Certificate Revocation Lists

A receiving agent SHOULD have access to some certificate-revocation
list (CRL) retrieval mechanism in order to gain access to certificate-
revocation information when validating certificate chains. A receiving
or sending agent SHOULD also provide a mechanism to allow a user to
store incoming certificate-revocation information for correspondents
in such a way so as to guarantee its later retrieval. However, it is
always better to get the latest information from the CA than to get
information stored away from incoming messages.

Receiving and sending agents SHOULD retrieve and utilize CRL
information every time a certificate is verified as part of a
certificate chain validation even if the certificate was already
verified in the past. However, in many instances (such as off-line
verification) access to the latest CRL information may be difficult or
impossible. The use of CRL information, therefore, may be dictated by
the value of the information that is protected. The value of the CRL
information in a particular context is beyond the scope of this draft
but may be governed by the policies associated with particular
certificate hierarchies.


4.2 Certificate Chain Validation

In creating a user agent for secure messaging, certificate, CRL, and
certificate chain validation SHOULD be highly automated while still
acting in the best interests of the user. Certificate, CRL, and chain
validation MUST be performed as per [KEYM] when validating a
correspondent's public key. This is necessary before using a public
key to provide security services such as: verifying a signature;
encrypting a content-encryption key (ex: RSA); or forming a pairwise
symmetric key (ex: Diffie-Hellman) to be used to encrypt or decrypt a
content-encryption key.

Certificates and CRLs are made available to the chain validation
procedure in two ways: a) incoming messages, and b) certificate and
CRL retrieval mechanisms. Certificates and CRLs in incoming messages
are not required to be in any particular order nor are they required
to be in any way related to the sender or recipient of the message
(although in most cases they will be related to the sender). Incoming
certificates and CRLs SHOULD be cached for use in chain validation and
optionally stored for later use. This temporary certificate and CRL
cache SHOULD be used to augment any other certificate and CRL
retrieval mechanisms for chain validation on incoming signed messages.


4.3 Certificate and CRL Signing Algorithms

Certificates and Certificate-Revocation Lists (CRLs) are signed by the
certificate issuer. A receiving agent MUST be capable of verifying the
signatures on certificates and CRLs made with id-dsa-with-sha1.

A receiving agent SHOULD be capable of verifying the signatures on
certificates and CRLs made with md2WithRSAEncryption,
md5WithRSAEncryption and sha-1WithRSAEncryption signature algorithms
with key sizes from 512 bits to 2048 bits described in [SMIME-MSG].


4.4 X.509 Version 3 Certificate Extensions

The X.509 v3 standard describes an extensible framework in which the
basic certificate information can be extended and how such extensions
can be used to control the process of issuing and validating
certificates. The PKIX Working Group has ongoing efforts to identify
and create extensions which have value in particular certification
environments. As such, there is
still a fair amount of profiling work to be done before there is
widespread agreement on which v3 extensions will be used. Further,
there are active efforts underway to issue X.509 v3 certificates for
business purposes. This draft identifies the minumum required set of
certificate extensions which have the greatest value in the S/MIME
environment. The basicConstraints, and keyUsage extensions are defined
in [X.509].

Sending and receiving agents MUST correctly handle the v3 Basic
Constraints Certificate Extension, the Key Usage Certificate
Extension, authorityKeyID, subjectKeyID, and the subjectAltNames when
they appear in end-user certificates. Some mechanism SHOULD exist to
handle the defined v3 certificate extensions when they appear in
intermediate or CA certificates.

Certificates issued for the S/MIME environment SHOULD NOT contain any
critical extensions other than those listed here. These extensions
SHOULD be marked as non-critical unless the proper handling of the
extension is deemed critical to the correct interpretation of the
associated certificate. Other extensions may be included, but those
extensions SHOULD NOT be marked as critical.


4.4.1 Basic Constraints Certificate Extension

The basic constraints extension serves to delimit the role and
position of an issuing authority or end-user certificate plays in a
chain of certificates.

For example, certificates issued to CAs and subordinate CAs contain a
basic constraint extension that identifies them as issuing authority
certificates. End-user subscriber certificates contain an extension
that constrains the certificate from being an issuing authority
certificate.

Certificates SHOULD contain a basicConstraints extension.


4.4.2 Key Usage Certificate Extension

The key usage extension serves to limit the technical purposes for
which a public key listed in a valid certificate may be used. Issuing
authority certificates may contain a key usage extension that
restricts the key to signing certificates, certificate revocation
lists and other data.

For example, a certification authority may create subordinate issuer
certificates which contain a keyUsage extension which specifies that
the corresponding public key can be used to sign end user certs and
sign CRLs.

If a key usage extension is included in a v3 X.509 Certificate, then
it MUST be marked as critical.


4.4.3 Subject Alternative Name Extension

The subject alternative name extension is used in S/MIME as the
preferred means to convey the RFC-822 email address(es) that
correspond to the entity for this certificate. Any RFC-822 email
addresses present MUST be encoded using the rfc822Name CHOICE of the
GeneralName type. Since the SubjectAltName type is a SEQUENCE OF
GeneralName, multiple RFC-822 email addresses MAY be present.


5. Security Considerations

All of the security issues faced by any cryptographic application must
be faced by a S/MIME agent. Among these issues are protecting the
user's private key, preventing various attacks, and helping the user
avoid mistakes such as inadvertently encrypting a message for the
wrong recipient. The entire list of security considerations is beyond
the scope of this document, but some significant concerns are listed
here.

When processing certificates, there are many situations where the
processing might fail. Because the processing may be done by a user
agent, a security gateway, or other program, there is no single way to
handle such failures. Just because the methods to handle the failures
has not been listed, however, the reader should not assume that they
are not important. The opposite is true: if a certificate is not
provably valid and associated with the message, the processing
software should take immediate and noticable steps to inform the end
user about it.

Some of the many places where signature and certificate checking might
fail include:
- no Internet mail addresses in a certificate match the sender of a
message
- no certificate chain leads to a trusted CA
- no ability to check the CRL for a certificate
- an invalid CRL was received
- the CRL being checked is expired
- the certificate is expired
- the certificate has been revoked
There are certainly other instances where a certificate may be
invalid, and it is the responsibility of the processing software to
check them all thoroughly, and to decide what to do if the check
fails.


A. References

[CERTV2] "S/MIME Version 2 Certificate Handling", RFC 2312

[CMS] "Cryptographic Message Syntax", Internet Draft draft-housley-
smime-cms

[CRMF] "Certificate Request Message Format", Internet Draft draft-ietf-
pkix-crmf

[KEYM] "Internet Public Key Infrastructure X.509 Certificate and CRL
Profile", Internet-Draft draft-ietf-pkix-ipki-part1

[MUSTSHOULD] "Key words for use in RFCs to Indicate Requirement
Levels", RFC 2119

[RFC-822], "Standard For The Format Of ARPA Internet Text Messages",
RFC 822.

[SMIME-MSG] "S/MIME Version 3 Message Specification ", Internet Draft
draft-ietf-smime-msg

[X.500] ITU-T Recommendation X.500 (1997) | ISO/IEC 9594-1:1997,
Information technology - Open Systems Interconnection - The Directory:
Overview of concepts, models and services

[X.501] ITU-T Recommendation X.501 (1997) | ISO/IEC 9594-2:1997,
Information technology - Open Systems Interconnection - The Directory:
Models

[X.509] ITU-T Recommendation X.509 (1997) | ISO/IEC 9594-8:1997,
Information technology - Open Systems Interconnection - The Directory:
Authentication framework

[X.520] ITU-T Recommendation X.520 (1997) | ISO/IEC 9594-6:1997,
Information technology - Open Systems Interconnection - The Directory:
Selected attribute types.


B. Acknowledgements

This document is largely based on [CERTV2] written by Steve Dusse,
Paul Hoffman, Blake Ramsdell, and Jeff Weinstein.

Significant comments and additions were made by John Pawling and Jim
Schaad.


C. Needed changes

Names for chaining -- still no clear consensus
Key usage for signing / encrypting certificate explanation
4.4.3 -- do we need to further qualify the syntax for rfc822Name (no
wildcards or comments)?
We need a reference for attribute certificates


D. Changes from last draft

Revised section 1 (John Pawling)
Removed [PKCS-1] due to section 1 revisions (Blake Ramsdell)
Changed Certificate and Certificate Revocation List definitions to
identify extensions and refer to [KEYM] (John Pawling)
Another MUST to MAY for email address in certificates, section 2.2
(John Pawling)
Added Attribute Certificates to 2.2.1 (Jim Schaad)
Changed language regarding "proposed" v3 extensions in 2.2.1 (Jim
Schaad / John Pawling)
Reworded NULL-DN sentence in 3.1 to avoid confusion with ASN.1 NULL
(John Pawling)
PKCS #7 to CMS in section 4 (John Pawling)
Section 4.2, revised language to support Diffie-Hellman (John Pawling)
Removed section A which was basically restating things in other drafts
(John Pawling)
Section 2.1, removed language regarding nextUpdate field in CRL
(Elliott Ginsburg)
Wrote 4.4.3. Flames welcome. (Blake Ramsdell)
Fixed [CERTV2] to point to RFC 2312 (Blake Ramsdell)
Removed "old S/MIME" appendix (Jim Schaad, John Pawling, Paul Hoffman)


E. Editor's address

Blake Ramsdell
Worldtalk
13122 NE 20th St., Suite C
Bellevue, WA 98005
(425) 882-8861
blaker@deming.com