| Network Working Group | R. Barnes |
| Internet-Draft | O. Friel |
| Intended status: Informational | Cisco |
| Expires: January 17, 2019 | July 16, 2018 |
Usage of PAKE with TLS 1.3
draft-barnes-tls-pake-04
The pre-shared key mechanism available in TLS 1.3 is not suitable for usage with low-entropy keys, such as passwords entered by users. This document describes an extension that enables the use of password-authenticated key exchange protocols with TLS 1.3.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on January 17, 2019.
Copyright (c) 2018 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
DISCLAIMER: This is a work-in-progress draft and has not yet seen significant security analysis. It should not be used as a basis for building production systems.
In some applications, it is desireable to enable a client and server to authenticate to one another using a low-entropy pre-shared value, such as a user-entered password.
In prior versions of TLS, this functionality has been provided by the integration of the Secure Remote Password PAKE protocol (SRP) [RFC5054]. The specific SRP integration described in RFC 5054 does not immediately extend to TLS 1.3 because it relies on the Client Key Exchange and Server Key Exchange messages, which no longer exist in 1.3.
TLS 1.3 itself provides a mechanism for authentication with pre-shared keys (PSKs). However, PSKs used with this protocol need to be “full-entropy”, because the binder values used for authentication can be used to mount a dictionary attack on the PSK. So while the TLS 1.3 PSK mechanism is suitable for the session resumption cases for which it is specified, it cannot be used when the client and server share only a low-entropy secret.
Enabling TLS to address this use case effectively requires the TLS handshake to execute a password-authenticated key establishment (PAKE) protocol. This document describes a TLS extension pake that can carry data necessary to execute a PAKE.
This extension is generic, in that it can be used to carry key exchange information for multiple different PAKEs. We assume that the client and server have pre-negotiated a choice of PAKE (and any required parameters) in addition to the password itself. As a first case, this document defines a concrete protocol for executing the SPAKE2+ PAKE protocol [I-D.irtf-cfrg-spake2].
The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in [RFC2119].
The mechanisms described in this document also apply to DTLS 1.3 [I-D.ietf-tls-dtls13], but for brevity, we will refer only to TLS throughout.
In order to use this protocol, a TLS client and server need to have pre-provisioned the values required to execute the protocol:
Servers will of course have multiple instances of this configuration information for different clients. Clients may also have multiple identities, even within a given server. We assume that in either case, a single opaque “identity” value is sufficient to identify the required parameters.
A client offers to authenticate with PAKE by including a pake extension in its ClientHello. The content of this exension is a PAKEClientHello value, providing a list of identities under which the client can authenticate, and for each identity, the client’s first message from the underlying PAKE protocol.
If a client sends the pake extension, then it MAY also send the key_share and pre_shared_key extensions, to allow the server to choose an authentication mode. Unlike PSK-based authentication, however, authentication with PAKE cannot be combined with the normal TLS ECDH mechanism. Forward secrecy is provided by the PAKE itself.
struct {
opaque identity<0..2^16-1>;
opaque pake_message<1..2^16-1>;
} PAKEShare;
struct {
PAKEShare client_shares<0..2^16-1>;
} PAKEClientHello;
A server that receives a pake extension examines the list of client shares to see if there is one with an identity the server recognizes. If so, the server may indicate its choice of PAKE authentication by including a pake extension in its ServerHello. The content of this exension is a PAKEServerHello value, specifying the identity value for the password the server has selected, and the server’s first message in the PAKE protocol.
Use of PAKE authenication is compatible with standard certificate-based authentication of both clients and servers. If a server includes an pake extension in its ServerHello, it may still send the Certificate and CertificateVerify messages, and/or send a CertificateRequest message to the client.
If a server uses PAKE authentication, then it MUST NOT send an extension of type key_share, pre_shared_key, or early_data.
struct {
PAKEShare server_share;
} PAKEServerHello;
Based on the messages exchanged in the ClientHello and ServerHello, the client and server execute the specified PAKE protocol to derive a shared key. This key is used as the ECHD(E) input to the TLS 1.3 key schedule.
As with client authentication via certificates, the server has not authenticated the client until after it has received the client’s Finished message. When a server negotiates the use of this mechanism for authentication, it MUST NOT send application data before it has received the client’s Finished message.
In order to be usable with the pake extension, a PAKE protocol must specify some syntax for its messages, and the protocol itself must be compatible with the message flow described above. A specification describing the use of a particular PAKE protocol with TLS must provide the following details:
The underlying cryptographic protocol must be compatible with the message flow described above:
In addition, to be compatible with the security requirements of TLS 1.3, PAKE protocols defined for use with TLS 1.3 MUST provide forward secrecy.
Several current PAKE protocols satisfy these requirements, for example:
In order to use SPAKE2+, a TLS client and server need to have pre-provisioned the values required to execute the SPAKE2+ protocol (see Section 3.1 of [I-D.irtf-cfrg-spake2]):
Note that the hash function H might be different from the hash function associated with the ciphersuite negotiated by the two parties. The hash function H MUST be a hash function suitable for hashing passwords, e.g., Argon2 or scrypt [I-D.irtf-cfrg-argon2] [RFC7914].
The TLS client and server roles map to the A and B roles in the SPAKE2+ specification, respectively. The identity of the server is the domain name sent in the server_name extension of the ClientHello message. The identity of the client is an opaque octet string, specified in the spake2 ClientHello extension, defined below.
From the shared password, each party computes two shared integers w0 and w1 by running the following algorithm twice (changing the context value each time):
struct {
uint16 context;
opaque client\_identity<0..255>;
opaque server\_name<0..255>;
opaque password<0..255>;
} PasswordInput;
Servers MUST store only the value w0 and the product L = w1*G, where G is the fixed generator of the group. Clients will need to have access to the values w0 and w1 directly, but SHOULD generate these values dynamically, rather than caching them.
The content of a pake_message in a ClientHello is the client’s key share T. The value T is computed as specified in [I-D.irtf-cfrg-spake2], as T = w*M + X, where M is a fixed value for the DH group and X is the public key of a fresh DH key pair. The format of the key share T is the same as for a KeyShareEntry.key_exchange value from the same group.
The content of a pake_message in a ServerHello is the server’s key share S. The value S is computed as specified in [I-D.irtf-cfrg-spake2], as S = w*N + Y, where N is a fixed value for the DH group and Y is the public key of a fresh DH key pair. The format of the key share S is the same as for a KeyShareEntry.key_exchange value from the same group.
Based on these messages, both the client and server can compute the two shared values as specified in [I-D.irtf-cfrg-spake2].
| Name | Value | Client | Server |
|---|---|---|---|
| Z | x*y*G | x*(S - w0*N) | x*(T - w0*M) |
| V | w1*y*G | w1*(S - w0*N) | y*L |
The following value is used as the (EC)DHE input to the TLS 1.3 key schedule:
K = H(Z || V)
Here H is the hash function corresponding to the TLS cipher suite in use and || represents concatenation of octet strings.
Many of the security properties of this protocol will derive from the PAKE protocol being used. Security considerations for PAKE protocols are noted in Section 5.
The mechanism defined in this document does not provide protection for the client’s identity, in contrast to TLS client authentication with certificates.
TLS servers that offer this mechanism can be used by third party attackers as an oracle for two questions:
The former is signaled by whether the server returns a pake extension.
[[TODO: Similar to https://tools.ietf.org/html/rfc5054#section-2.5.1.3, the server could run through a complete handshake calculation and fail at the end so that the attacker only knows that the identity/password pair is incorrect, but does not know if the identity is recognized or not. This requires that the server can interpret the pake_message and ascertain the associated PAKE algorithm, group parameters, etc., which requires a reworking of some text in this draft as the identity is currently defined as providing a map to said group parameters. This is related to the discussion in the Open Items section.]]
The latter is signaled by whether the connection succeeds. These oracles are all-or-nothing: If the attacker does not have the correct identity or password, he does not learn anything about the correct value.
For the most part, the security properties of the password-based authentication described in this document are the same as those described in the Security Considerations of [I-D.irtf-cfrg-spake2]. The TLS Finished MAC provides the key confirmation required for the security of the protocol. Note that all of the elements covered by the example confirmation hash listed in that document are also covered by the Finished MAC:
The x and y values used in the SPAKE2 protocol MUST have the same ephemerality properties as the key shares sent in the key_shares extension. In particular, x and y MUST NOT be equal to zero. This ensures that TLS sessions using SPAKE2 have the same forward secrecy properties as sessions using the normal TLS (EC)DH mechanism.
It is possible that a client may know the password to use, but may not know in advance which PAKE protocols(s) a particular server supports. A potential solution to this is similar to TLS1.3 ClientHello key_share operation: the client may send an empty client_shares vector in its PAKEClientHello extension. The server can then send an HelloRetryRequest indicating which PAKE protocol, and associated group parameters, the client should use. The client then sends another ClientHello that includes pake_message in the PAKEClientHello extension calculated using the correct algorithm. This requires definition of a suitable field for transporting PAKE algorithm and group parameters.
As an optimaisation, similar to TLS1.3 key_share operation, the client could guess the PAKE protocol and include a pake_message derived from its guess in the initial ClientHello. If the server does not support the selected PAKE protcol (or protocol group parameter, etc.), the server can send an HelloRetryRequest indicating the supported PAKE protocol and group parameters. Note: it is TBD if sending two different pake_messages derived from two different protocol and/or group parameters in two different ClientHello messages constitutes a significant attack vector. This needs cryptographic review.
This document requests that IANA add a value to the TLS ExtensionType Registry with the following contents:
| Value | Extension Name | TLS 1.3 | Reference |
|---|---|---|---|
| TBD | pake | CH, SH | RFC XXXX |
[[ RFC EDITOR: Please replace “TBD” in the above table with the value assigned by IANA, and replace “XXXX” with the RFC number assigned to this document. ]]
| [I-D.ietf-tls-dtls13] | Rescorla, E., Tschofenig, H. and N. Modadugu, "The Datagram Transport Layer Security (DTLS) Protocol Version 1.3", Internet-Draft draft-ietf-tls-dtls13-28, July 2018. |
| [I-D.irtf-cfrg-spake2] | Ladd, W. and B. Kaduk, "SPAKE2, a PAKE", Internet-Draft draft-irtf-cfrg-spake2-05, February 2018. |
| [RFC2119] | Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997. |