RFC 8422

Elliptic Curve Cryptography (ECC) Cipher Suites for Transport Layer Security (TLS) Versions 1.2 and Earlier

📅 published August 2018, from ietf

RFC 8422 💖 Part 1 of 6

  • Title: Elliptic Curve Cryptography (ECC) Cipher Suites for Transport Layer Security (TLS) Versions 1.2 and Earlier 🔐
  • Authors: Y. Nir (Check Point), S. Josefsson (SJD AB), M. Pegourie-Gonnard (ARM) 👑
  • Series: Internet Engineering Task Force (IETF), Request for Comments: 8422
  • Category: Standards Track 🌈
  • ISSN: 2070-1721
  • Date: August 2018 🦋
  • Obsoletes: RFC 4492
  • Updates: none listed

Abstract ✨

okay bestie, this doc is all about ECC-based key exchange for TLS 💅 It covers ECDHE key agreement during a TLS handshake, plus ECDSA and EdDSA as the authentication mechanisms.

Heads up: this document obsoletes RFC 4492 🫶

Status of This Memo 🎀

It's an Internet Standards Track doc, so no cap, this one is official.

It's a product of the Internet Engineering Task Force (IETF) 🌸 It represents the consensus of the IETF community, got public review, and was approved for publication by the Internet Engineering Steering Group (IESG). For more on Internet Standards, check Section 2 of RFC 7841.

For the current status, any errata, and how to give feedback, go to https://www.rfc-editor.org/info/rfc8422.

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. Please read these carefully, since they explain your rights and restrictions. Code Components pulled out of this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions, and they come without warranty, as the Simplified BSD License describes.

Table of Contents 📖

1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   4
  1.1.  Conventions Used in This Document . . . . . . . . . . . .   4
2.  Key Exchange Algorithm  . . . . . . . . . . . . . . . . . . .   4
  2.1.  ECDHE_ECDSA . . . . . . . . . . . . . . . . . . . . . . .   6
  2.2.  ECDHE_RSA . . . . . . . . . . . . . . . . . . . . . . . .   7
  2.3.  ECDH_anon . . . . . . . . . . . . . . . . . . . . . . . .   7
  2.4.  Algorithms in Certificate Chains  . . . . . . . . . . . .   7
3.  Client Authentication . . . . . . . . . . . . . . . . . . . .   8
  3.1.  ECDSA_sign  . . . . . . . . . . . . . . . . . . . . . . .   8
4.  TLS Extensions for ECC  . . . . . . . . . . . . . . . . . . .   9
5.  Data Structures and Computations  . . . . . . . . . . . . . .  10
  5.1.  Client Hello Extensions . . . . . . . . . . . . . . . . .  10
    5.1.1.  Supported Elliptic Curves Extension . . . . . . . . .  11
    5.1.2.  Supported Point Formats Extension . . . . . . . . . .  13
    5.1.3.  The signature_algorithms Extension and EdDSA  . . . .  13
  5.2.  Server Hello Extension  . . . . . . . . . . . . . . . . .  14
  5.3.  Server Certificate  . . . . . . . . . . . . . . . . . . .  15
  5.4.  Server Key Exchange . . . . . . . . . . . . . . . . . . .  16
    5.4.1.  Uncompressed Point Format for NIST Curves . . . . . .  19
  5.5.  Certificate Request . . . . . . . . . . . . . . . . . . .  20
  5.6.  Client Certificate  . . . . . . . . . . . . . . . . . . .  21
  5.7.  Client Key Exchange . . . . . . . . . . . . . . . . . . .  22
  5.8.  Certificate Verify  . . . . . . . . . . . . . . . . . . .  23
  5.9.  Elliptic Curve Certificates . . . . . . . . . . . . . . .  24
  5.10. ECDH, ECDSA, and RSA Computations . . . . . . . . . . . .  24
  5.11. Public Key Validation . . . . . . . . . . . . . . . . . .  26
6.  Cipher Suites . . . . . . . . . . . . . . . . . . . . . . . .  26
7.  Implementation Status . . . . . . . . . . . . . . . . . . . .  27
8.  Security Considerations . . . . . . . . . . . . . . . . . . .  27
9.  IANA Considerations . . . . . . . . . . . . . . . . . . . . .  28
10. References  . . . . . . . . . . . . . . . . . . . . . . . . .  29
  10.1.  Normative References . . . . . . . . . . . . . . . . . .  29
  10.2.  Informative References . . . . . . . . . . . . . . . . .  31
Appendix A.  Equivalent Curves (Informative)  . . . . . . . . . .  32
Appendix B.  Differences from RFC 4492  . . . . . . . . . . . . .  33
Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . .  34
Authors' Addresses  . . . . . . . . . . . . . . . . . . . . . . .  34

1. Introduction 🎀

okay so this doc adds ECC support to TLS. It applies to TLS versions 1.0 [RFC2246], 1.1 [RFC4346], and 1.2 [RFC5246]. ECC in TLS 1.3 is defined in [TLS1.3], and it's explicitly out of scope here 🙈 Specifically, this document defines:

  • the use of the ECDHE key agreement scheme with ephemeral keys to establish the TLS premaster secret, and
  • the use of ECDSA and EdDSA signatures for authentication of TLS peers.

Here's the tour 🦋 Section 2 gives an overview of ECC-based key exchange algorithms for TLS. Section 3 covers using ECC certificates for client authentication. The TLS extensions that let a client negotiate specific curves and point formats are in Section 4. Section 5 specifies the data structures an ECC-based handshake needs, how they're encoded in TLS messages, and how those messages get processed. Section 6 defines ECC-based cipher suites and picks out a small subset as recommended for all implementations of this specification. Section 8 talks about security considerations. Section 9 covers IANA considerations for the name spaces created by this document's predecessor. Appendix B lists the differences from [RFC4492], the doc this one replaces.

Implementing this spec requires familiarity with TLS, TLS extensions [RFC4366], and ECC 💜

1.1. Conventions Used in This Document

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here. 🌈

2. Key Exchange Algorithm 🦋

This doc defines three new ECC-based key exchange algorithms for TLS ✨ They all use Ephemeral ECDH (ECDHE) to compute the TLS premaster secret, and they only differ in the mechanism (if any) used to authenticate them. Deriving the TLS master secret from the premaster secret, and then generating the bulk encryption/MAC keys and initialization vectors, works the same no matter which key exchange algorithm you pick. ECC doesn't change any of that 💅

Table 1 is the quick summary 👑 All of these key exchange algorithms provide forward secrecy if and only if fresh ephemeral keys are generated and used, and also destroyed after use.

  +-------------+------------------------------------------------+
  | Algorithm   | Description                                    |
  +-------------+------------------------------------------------+
  | ECDHE_ECDSA | Ephemeral ECDH with ECDSA or EdDSA signatures. |
  | ECDHE_RSA   | Ephemeral ECDH with RSA signatures.            |
  | ECDH_anon   | Anonymous ephemeral ECDH, no signatures.       |
  +-------------+------------------------------------------------+

                Table 1: ECC Key Exchange Algorithms

These key exchanges are basically the same vibe as DHE_DSS, DHE_RSA, and DH_anon, respectively 🫶

With ECDHE_RSA, a server can reuse its existing RSA certificate and easily meet a constrained client's elliptic curve preferences (see Section 4). The catch is that the server's computational cost is higher for ECDHE_RSA than for the traditional RSA key exchange, which doesn't provide forward secrecy 🙈

The anonymous key exchange algorithm doesn't authenticate the server or the client. Like other anonymous TLS key exchanges, it's subject to man-in-the-middle attacks, so applications using TLS with this algorithm SHOULD provide authentication by other means. > This is important! Don't skip it 💀

      Client                                        Server
      ------                                        ------
      ClientHello          -------->
                                               ServerHello
                                              Certificate*
                                        ServerKeyExchange*
                                      CertificateRequest*+
                           <--------       ServerHelloDone
      Certificate*+
      ClientKeyExchange
      CertificateVerify*+
      [ChangeCipherSpec]
      Finished             -------->
                                        [ChangeCipherSpec]
                           <--------              Finished
      Application Data     <------->      Application Data

           * message is not sent under some conditions
           + message is not sent unless client authentication
             is desired

        Figure 1: Message Flow in a Full TLS 1.2 Handshake

Figure 1 shows every message involved in the TLS key establishment protocol (aka full handshake) 📊 ECC only directly affects the ClientHello, the ServerHello, the server's Certificate message, the ServerKeyExchange, the ClientKeyExchange, the CertificateRequest, the client's Certificate message, and the CertificateVerify. Next, we dig into the ECC key exchange algorithm in more detail, covering the content and processing of these messages. For ease of exposition, client authentication and its messages (marked with a '+' in Figure 1) wait until Section 3, and the optional ECC-specific extensions (which impact the Hello messages) wait until Section 4 🧚

2.1. ECDHE_ECDSA

In ECDHE_ECDSA, the server's certificate MUST contain an ECDSA- or EdDSA-capable public key. 🔐

The server sends its ephemeral ECDH public key, plus a specification of the matching curve, in the ServerKeyExchange message. These parameters MUST be signed with ECDSA or EdDSA using the private key that corresponds to the public key in the server's Certificate.

The client generates an ECDH key pair on the same curve as the server's ephemeral ECDH key, and sends its public key in the ClientKeyExchange message.

Both client and server perform an ECDH operation (see Section 5.10) and use the resulting shared secret as the premaster secret ✨

2.2. ECDHE_RSA

This key exchange is the same as ECDHE_ECDSA, except that the server's certificate MUST contain an RSA public key authorized for signing, and the signature in the ServerKeyExchange message must be computed with the corresponding RSA private key. 🌸

To be continued in Part 2 💖

2.3. ECDH_anon

NOTE: okay bestie, even though the name starts with "ECDH_" (no E), the key in ECDH_anon is ephemeral just like the keys in ECDHE_RSA and ECDHE_ECDSA. 💅 The naming copies DH_anon, where the key is also ephemeral but the name doesn't show it. 🙈

In ECDH_anon, the server's Certificate, the CertificateRequest, the client's Certificate, and the CertificateVerify messages MUST NOT be sent. 🚫 No certs, no requests, no verify, nothing. Just vibes. ✨

The server MUST send an ephemeral ECDH public key and a specification of the corresponding curve in the ServerKeyExchange message. 🔑 These parameters MUST NOT be signed. 🙅‍♀️

The client generates an ECDH key pair on the same curve as the server's ephemeral ECDH key and sends its public key in the ClientKeyExchange message. 🎀

Both client and server perform an ECDH operation and use the resultant shared secret as the premaster secret. 🦋 All ECDH calculations are performed as specified in Section 5.10. 💖

2.4. Algorithms in Certificate Chains

This specification does not impose restrictions on signature schemes used anywhere in the certificate chain. 🌈 The previous version of this document required the signatures to match, but that restriction came from earlier TLS versions, and it's lifted here, same as it was in RFC 5246. 🫶

3. Client Authentication

This document defines a client authentication mechanism named after the type of client certificate involved: ECDSA_sign. 🧚 The ECDSA_sign mechanism works with any of the non-anonymous ECC key exchange algorithms described in Section 2, and also with other non-anonymous (non-ECC) key exchange algorithms defined in TLS. ✨

Note that client certificates with EdDSA public keys also use this mechanism. 💜

The server can request ECC-based client authentication by including this certificate type in its CertificateRequest message. 📬 The client must check if it has a certificate that fits the method the server suggested and if it's willing to use it for authentication. 🔍

If these conditions are not met, the client SHOULD send a client Certificate message containing no certificates. 🫠 In this case, the ClientKeyExchange MUST be sent as described in Section 2, and the CertificateVerify MUST NOT be sent. If the server requires client authentication, it may respond with a fatal handshake failure alert. 💀

If the client has an appropriate certificate and is willing to use it for authentication, it must send that certificate in the client's Certificate message (as per Section 5.6) and prove possession of the private key corresponding to the certified key. 🔐 How you figure out which certificate is appropriate, and how you prove you have the private key, is different for each authentication mechanism and is described below. 🍓

NOTE: It is permissible for a server to request (and the client to send) a client certificate of a different type than the server certificate. 👑

3.1. ECDSA_sign

To use this authentication mechanism, the client MUST possess a certificate containing an ECDSA- or EdDSA-capable public key. 🦄

The client proves possession of the private key corresponding to the certified key by including a signature in the CertificateVerify message as described in Section 5.8. ✍️💅

4. TLS Extensions for ECC ✨

Two TLS extensions are defined in this specification: (i) the Supported Elliptic Curves Extension and (ii) the Supported Point Formats Extension. 🎀 These let you negotiate specific curves and point formats (for example, compressed vs. uncompressed, respectively) during a handshake that starts a new session. 🌸 They're especially useful for constrained clients that may only support a limited number of curves or point formats. 🫶 They follow the general approach outlined in [RFC4366]. The message details are in Section 5. The client lists the curves it supports and the point formats it can parse by including the appropriate extensions in its ClientHello message. 📋 The server similarly lists the point formats it can parse by including an extension in its ServerHello message. 🙌

A TLS client that proposes ECC cipher suites in its ClientHello message SHOULD include these extensions. Servers implementing ECC cipher suites MUST support these extensions, and when a client uses these extensions, servers MUST NOT negotiate the use of an ECC cipher suite unless they can complete the handshake while respecting the choice of curves specified by the client. 🛑 This kills the chance that a negotiated ECC handshake gets aborted later because the client can't handle the server's EC key. 😭 Iconic behavior, honestly.

The client MUST NOT include these extensions in the ClientHello message if it does not propose any ECC cipher suites. 🙅‍♀️ A client that proposes ECC cipher suites may choose not to include these extensions. In that case, the server is free to choose any one of the elliptic curves or point formats listed in Section 5. 🍭 That section also goes into the structure and processing of these extensions in more detail. 🧠

In the case of session resumption, the server simply ignores the Supported Elliptic Curves Extension and the Supported Point Formats Extension appearing in the current ClientHello message. 🔄 These extensions only matter during handshakes that negotiate a new session. 💜

5. Data Structures and Computations 🧁

This section specifies the data structures and computations used by the ECC-based key mechanisms specified in the previous three sections. 🎀 The presentation language used here is the same as the one used in TLS. Since this specification extends TLS, these descriptions should be merged with those in the TLS specification and any others that extend TLS. 🫶 This means that enum types may not specify all possible values, and structures with multiple formats chosen with a select() clause may not indicate all possible cases. 🌈

5.1. Client Hello Extensions 🦋

This section specifies two TLS extensions that can be included with the ClientHello message as described in [RFC4366]: the Supported Elliptic Curves Extension and the Supported Point Formats Extension. ✨

When these extensions are sent:

The extensions SHOULD be sent along with any ClientHello message that proposes ECC cipher suites. 📨

Meaning of these extensions:

These extensions let a client enumerate the elliptic curves it supports and/or the point formats it can parse. 🧚

Structure of these extensions:

The general structure of TLS extensions is described in [RFC4366], and this specification adds two types to ExtensionType. 🧩

   enum {
       elliptic_curves(10),
       ec_point_formats(11)
   } ExtensionType;
  • elliptic_curves (Supported Elliptic Curves Extension): Indicates the set of elliptic curves supported by the client. For this extension, the opaque extension_data field contains NamedCurveList. See Section 5.1.1 for details. 🌷
  • ec_point_formats (Supported Point Formats Extension): Indicates the set of point formats that the client can parse. For this extension, the opaque extension_data field contains ECPointFormatList. See Section 5.1.2 for details. 🌸

Actions of the sender:

A client that proposes ECC cipher suites in its ClientHello message appends these extensions (along with any others), enumerating the curves it supports and the point formats it can parse. 📝 Clients SHOULD send both the Supported Elliptic Curves Extension and the Supported Point Formats Extension. If the Supported Point Formats Extension is indeed sent, it MUST contain the value 0 (uncompressed) as one of the items in the list of point formats. 🫧

Actions of the receiver:

A server that receives a ClientHello containing one or both of these extensions MUST use the client's enumerated capabilities to guide its selection of an appropriate cipher suite. 🧭 One of the proposed ECC cipher suites must be negotiated only if the server can successfully complete the handshake while using the curves and point formats supported by the client (cf. Sections 5.3 and 5.4). 🎯

NOTE: A server participating in an ECDHE_ECDSA key exchange may use different curves for the ECDSA or EdDSA key in its certificate and for the ephemeral ECDH key in the ServerKeyExchange message. 🙈 The server MUST consider the extensions in both cases. 👀

If a server does not understand the Supported Elliptic Curves Extension, does not understand the Supported Point Formats Extension, or is unable to complete the ECC handshake while restricting itself to the enumerated curves and point formats, it MUST NOT negotiate the use of an ECC cipher suite. 🚫 Depending on what other cipher suites the client proposes and the server supports, this may result in a fatal handshake failure alert due to the lack of common cipher suites. 😵

5.1.1. Supported Elliptic Curves Extension 🌈

[RFC4366] is not the only thing that's changed here, so let me explain. RFC 4492 defined 25 different curves in the NamedCurve registry (now renamed the "TLS Supported Groups" registry, although the enumeration below is still named NamedCurve) for use in TLS. 📚 Only three have seen much use. 👀 This specification deprecates the rest (with numbers 1-22). It also deprecates the explicit curves with identifiers 0xFF01 and 0xFF02. 🗑️ It also adds the new curves defined in [RFC7748]. The end result is as follows:

        enum {
            deprecated(1..22),
            secp256r1 (23), secp384r1 (24), secp521r1 (25),
            x25519(29), x448(30),
            reserved (0xFE00..0xFEFF),
            deprecated(0xFF01..0xFF02),
            (0xFFFF)
        } NamedCurve;

Note that other specifications have since added other values to this enumeration. Some of those values are not curves at all, but finite field groups. See [RFC7919]. 🧐

secp256r1, etc: Indicates support of the corresponding named curve or groups. The named curves secp256r1, secp384r1, and secp521r1 are specified in SEC 2 [SECG-SEC2]. These curves are also recommended in ANSI X9.62 [ANSI.X9-62.2005] and FIPS 186-4 [FIPS.186-4]. 📖 The rest of this document refers to these three curves as the "NIST curves" because they were originally standardized by the National Institute of Standards and Technology. 🏛️ The curves x25519 and x448 are defined in [RFC7748]. Values 0xFE00 through 0xFEFF are reserved for private use. 🔒

The predecessor of this document also supported explicitly defined prime and char2 curves, but this specification deprecates those. 🫠

The NamedCurve name space (now titled "TLS Supported Groups") is maintained by IANA. See Section 9 for information on how new value assignments are added. 🗂️

        struct {
            NamedCurve named_curve_list<2..2^16-1>
        } NamedCurveList;

Items in named_curve_list are ordered according to the client's preferences (favorite choice first). 👑

As an example, a client that only supports secp256r1 (aka NIST P-256; value 23 = 0x0017) and secp384r1 (aka NIST P-384; value 24 = 0x0018), and prefers secp256r1, would include a TLS extension consisting of the following octets. 🍓 Note that the first two octets indicate the extension type (Supported Elliptic Curves Extension):

        00 0A 00 06 00 04 00 17 00 18

5.1.2. Supported Point Formats Extension 💖

Okay bestie, here's the point format enum and struct, copied exactly:

enum {
    uncompressed (0),
    deprecated (1..2),
    reserved (248..255)
} ECPointFormat;
struct {
    ECPointFormat ec_point_format_list<1..2^8-1>
} ECPointFormatList;

Three point formats are in that ECPointFormat definition, but this spec deprecates everything except uncompressed 🫶. Implementations MUST support the uncompressed format for all of their supported curves, and they MUST NOT support other formats for the curves defined in this spec. For backwards compatibility, the point format list extension MAY still be included, and it can contain exactly one value: the uncompressed point format (0). RFC 4492 said that if this extension is missing, only the uncompressed format is supported, so talking to implementations that support uncompressed works with or without the extension ✨.

Now the drama 🚨. If the client sends the extension and it does NOT contain the uncompressed format, and the client has used the Supported Groups extension to indicate support for any of the curves defined in this spec, then the server MUST abort the handshake and return an illegal_parameter alert. Not it. 💀

The ECPointFormat name space (now titled "TLS EC Point Formats") is maintained by IANA. Check Section 9 for how new value assignments get added 🍓.

A client that's compliant with this spec and supports no other curves MUST send these octets. The first two octets are the extension type (Supported Point Formats Extension):

00 0B 00 02 01 00

5.1.3. The signature_algorithms Extension and EdDSA 🌈

The signature_algorithms extension, defined in Section 7.4.1.4.1 of [RFC5246], advertises the combinations of signature algorithm and hash function that the client supports. The pure (non-prehashed) forms of EdDSA don't hash the data before signing it, so it doesn't make sense to combine them with a hash function in the extension 🙈.

For bits-on-the-wire compatibility with TLS 1.3, we define a new dummy value in the "TLS HashAlgorithm" registry called "Intrinsic" (value 8). It means hashing is intrinsic to the signature algorithm 🧚.

To represent ed25519 and ed448 in the signature_algorithms extension, the value shall be (8,7) and (8,8), respectively 🎀.

5.2. Server Hello Extension ✨

This section specifies a TLS extension that can be included with the ServerHello message, as described in [RFC4366]. It's the Supported Point Formats Extension, so let's get into it 🦋.

When this extension is sent: The Supported Point Formats Extension is included in a ServerHello message, in response to a ClientHello message that contains the Supported Point Formats Extension, when negotiating an ECC cipher suite.

Meaning of this extension: This extension lets a server list the point formats it can parse. That covers the curve that will show up in its ServerKeyExchange message when using the ECDHE_ECDSA, ECDHE_RSA, or ECDH_anon key exchange algorithm 💜.

Structure of this extension: The server's Supported Point Formats Extension has the same structure as the client's (see Section 5.1.2). The items in ec_point_format_list here are ordered by the server's preference, with its favorite first. The server MAY include items that weren't in the client's list. Still, without extensions this spec allows exactly one point format, so there's not really any room for mismatches 🌸.

Actions of the sender: A server that selects an ECC cipher suite in response to a ClientHello that includes a Supported Point Formats Extension appends this extension (along with others) to its ServerHello message, listing the point formats it can parse. When it's used, the Supported Point Formats Extension MUST contain the value 0 (uncompressed) as one of the items in the list of point formats 🔥.

Actions of the receiver: A client that receives a ServerHello message containing a Supported Point Formats Extension MUST respect the server's choice of point formats during the handshake (cf. Sections 5.6 and 5.7). If no Supported Point Formats Extension is received with the ServerHello, that's equivalent to an extension allowing only the uncompressed point format 👑.

5.3. Server Certificate 🍓

When this message is sent: This message is sent in all non-anonymous, ECC-based key exchange algorithms.

Meaning of this message: This message is used to authentically convey the server's static public key to the client. The table below shows the server certificate type for each key exchange algorithm. ECC public keys MUST be encoded in certificates as described in Section 5.9 🦋.

NOTE: The server's Certificate message can carry a chain of certificates. The restrictions mentioned in Table 2 apply only to the server's certificate, which is first in the chain 🫶.

+-------------+-----------------------------------------------------+
| Algorithm   | Server Certificate Type                             |
+-------------+-----------------------------------------------------+
| ECDHE_ECDSA | Certificate MUST contain an ECDSA- or EdDSA-capable |
|             | public key.                                         |
| ECDHE_RSA   | Certificate MUST contain an RSA public key.         |
+-------------+-----------------------------------------------------+

                  Table 2: Server Certificate Types

Structure of this message: Identical to the TLS Certificate format ✨.

Actions of the sender: The server constructs an appropriate certificate chain and conveys it to the client in the Certificate message. If the client has used a Supported Elliptic Curves Extension, the public key in the server's certificate MUST respect the client's choice of elliptic curves. A server that can't satisfy this requirement MUST NOT choose an ECC cipher suite in its ServerHello message.)

Actions of the receiver: The client validates the certificate chain, extracts the server's public key, and checks that the key type is appropriate for the negotiated key exchange algorithm. (A possible reason for a fatal handshake failure is that the client's capabilities for handling elliptic curves and point formats are exceeded; cf. Section 5.1.) 💅

5.4. Server Key Exchange 🎀

When this message is sent:

okay so this message gets sent when you're using the ECDHE_ECDSA, ECDHE_RSA, or ECDH_anon key exchange algorithms 💖

Meaning of this message:

this is how the server hands the client its ephemeral ECDH public key, along with the elliptic curve domain parameters that go with it ✨

The ECCurveType enum used to have values for explicit prime and explicit char2 curves. Those are now deprecated 🥀 so only one value is left standing:

Structure of this message:

enum {
    deprecated (1..2),
    named_curve (3),
    reserved(248..255)
} ECCurveType;

The value named_curve means a named curve is being used. It's now the only format left, and honestly it's giving main character energy 👑

Values 248 through 255 are reserved for private use 🧚

The ECCurveType name space (now titled "TLS EC Curve Types") is maintained by IANA. Check Section 9 to see how new value assignments get added 🌸

RFC 4492 had specs for an ECCurve structure and an ECBasisType structure. Both are gone now because they were only used with the deprecated explicit curves 💅

struct {
    opaque point <1..2^8-1>;
} ECPoint;
  • point: this is the byte string form of an elliptic curve point, following the conversion routine in Section 4.3.6 of [ANSI.X9-62.2005]. The byte string may represent a point in uncompressed, compressed, or hybrid format, but this specification deprecates everything except the uncompressed format 🫶 For the NIST curves, the format gets repeated in Section 5.4.1 so you don't have to go hunting. For X25519 and X448, the only valid representation is the one from [RFC7748]: a 32- or 56-octet representation of the u value of the point ✨ This structure MUST NOT be used with Ed25519 and Ed448 public keys. Not even a little bit 🙅‍♀️
struct {
    ECCurveType    curve_type;
    select (curve_type) {
        case named_curve:
            NamedCurve namedcurve;
    };
} ECParameters;
  • curve_type: says what kind of elliptic curve domain parameters you're looking at 🔥
  • namedCurve: a recommended set of elliptic curve domain parameters. Every NamedCurve value that refers to a curve capable of Diffie-Hellman is allowed. With the explicit curves deprecated, that now means all of the NamedCurve values 🌈
struct {
    ECParameters    curve_params;
    ECPoint         public;
} ServerECDHParams;
  • curve_params: the elliptic curve domain parameters tied to the ECDH public key 💜
  • public: the ephemeral ECDH public key 🦋

The ServerKeyExchange message gets extended like this 🍓

enum {
    ec_diffie_hellman
} KeyExchangeAlgorithm;
  • ec_diffie_hellman: means the ServerKeyExchange message carries an ECDH public key 💖
select (KeyExchangeAlgorithm) {
    case ec_diffie_hellman:
        ServerECDHParams    params;
        Signature           signed_params;
} ServerKeyExchange;
  • params: the ECDH public key plus its associated domain parameters 🎀
  • signed_params: a hash of the params, with the signature that matches that hash applied. The private key that goes with the certified public key in the server's Certificate message is the one used for signing 🔐
enum {
    ecdsa(3),
    ed25519(7)
    ed448(8)
} SignatureAlgorithm;
select (SignatureAlgorithm) {
   case ecdsa:
        digitally-signed struct {
            opaque sha_hash[sha_size];
        };
   case ed25519,ed448:
        digitally-signed struct {
            opaque rawdata[rawdata_size];
        };
} Signature;
ServerKeyExchange.signed_params.sha_hash
    SHA(ClientHello.random + ServerHello.random +
                           ServerKeyExchange.params);
ServerKeyExchange.signed_params.rawdata
    ClientHello.random + ServerHello.random +
                           ServerKeyExchange.params;

NOTE: SignatureAlgorithm is "rsa" for the ECDHE_RSA key exchange algorithm and "anonymous" for ECDH_anon. Those cases are defined in TLS. For ECDHE_ECDSA, SignatureAlgorithm is "ecdsa" or "eddsa" 🙈

ECDSA signatures are generated and verified the way Section 5.10 describes 💅 In the sha_hash template above, SHA can mean a hash algorithm other than SHA-1. As ANSI X9.62 puts it, an ECDSA signature is a pair of integers, r and s. The digitally-signed element gets encoded as an opaque vector <0..2^16-1>, and what's inside is the DER encoding of this ASN.1 notation:

Ecdsa-Sig-Value ::= SEQUENCE {
    r       INTEGER,
    s       INTEGER
}

EdDSA signatures, both in the protocol and in certificates that conform to [RFC8410], are generated and verified according to [RFC8032]. The digitally-signed element is an opaque vector <0..2^16-1>, and its contents include the octet string output of the EdDSA signing algorithm 🦋

Actions of the sender:

The server picks elliptic curve domain parameters, plus an ephemeral ECDH public key that matches those parameters, using the ECKAS-DH1 scheme from IEEE 1363 [IEEE.P1363]. Then it sends all of that to the client in the ServerKeyExchange message, using the format from above 🎁

Actions of the receiver:

The client verifies the signature (when there is one) and pulls out the server's elliptic curve domain parameters and ephemeral ECDH public key from the ServerKeyExchange message. (A possible reason for a fatal handshake failure is that the client's capabilities for handling elliptic curves and point formats have been exceeded. See Section 5.1 🥺)

5.4.1. Uncompressed Point Format for NIST Curves

okay this is the wire format for ECPoint inside ServerKeyExchange records 🌸 The first octet says which form the point is in: compressed, uncompressed, or hybrid. This specification only supports the uncompressed format for these curves. After that comes the binary X value in "big-endian" or "network" format, then the binary Y value in the same format. There are no internal length markers, so each number takes up as many octets as the curve parameters imply. For P-256, X and Y each use 32 octets, padded with zeros on the left if needed. For P-384 they're 48 octets each, and for P-521 they're 66 octets each 💖

Here's the more formal version:

enum {
    uncompressed(4),
    (255)
  } PointConversionForm;

struct {
    PointConversionForm  form;
    opaque               X[coordinate_length];
    opaque               Y[coordinate_length];
} UncompressedPointRepresentation;

5.5. Certificate Request 🎀

When this message is sent:

this one is sent when the server is requesting client authentication 🔑

Meaning of this message:

the server uses this message to suggest which client authentication methods it finds acceptable ✨

Structure of this message:

The TLS CertificateRequest message gets extended like this 🧚

enum {
    ecdsa_sign(64),
    deprecated1(65),  /* was rsa_fixed_ecdh */
    deprecated2(66),  /* was ecdsa_fixed_ecdh */
    (255)
} ClientCertificateType;
  • ecdsa_sign: tells the client the server would like to use the matching client authentication method from Section 3 🫶

Heads up that RFC 4492 also defined RSA and ECDSA certificates that included a fixed ECDH public key. Those mechanisms saw very little implementation, so this specification deprecates them 🥀

Actions of the sender:

The server decides which client authentication methods it wants to use, then tells the client using the format above 💅

Actions of the receiver:

The client figures out whether it has a suitable certificate for any of the requested methods, and whether it wants to go ahead with client authentication at all 🌈

5.6. Client Certificate 💖

When this message is sent:

this gets sent in response to a CertificateRequest, but only when the client has a suitable certificate and has decided to go ahead with client authentication. (Heads up: if the server used a Supported Point Formats Extension, a certificate only counts as suitable for the ECDSA_sign authentication method if the public key point in it is uncompressed, since that's the only point format still supported 👀)

Meaning of this message:

this is how the client authentically conveys its static public key to the server 🔐 ECC public keys have to be encoded in certificates the way Section 5.9 describes. The certificate MUST contain an ECDSA- or EdDSA-capable public key. No exceptions, bestie 💅

NOTE: the client's Certificate message can carry a whole chain of certificates. The restrictions above only apply to the client's certificate, the first one in the chain 🦋

Structure of this message:

Identical to the TLS client Certificate format 🎁

Actions of the sender:

the client builds an appropriate certificate chain and sends it to the server in the Certificate message ✨

Actions of the receiver:

the TLS server validates the certificate chain, pulls out the client's public key, and checks that the key type fits the client authentication method 🔥

5.7. Client Key Exchange 🌸

When this message is sent:

this message is sent in all key exchange algorithms. It carries the client's ephemeral ECDH public key 💜

Meaning of the message:

it conveys ephemeral data for the client's part of the key exchange, like its ephemeral ECDH public key 🦋

Structure of this message:

The TLS ClientKeyExchange message gets extended like this 🍓

enum {
    implicit,
    explicit
} PublicValueEncoding;
  • implicit, explicit: for ECC cipher suites, this says whether the client's ECDH public key is in the client's certificate (implicit) or sent as an ephemeral ECDH public key in the ClientKeyExchange message (explicit). The implicit encoding is deprecated and only kept here for backward compatibility 🥺
struct {
    ECPoint ecdh_Yc;
} ClientECDiffieHellmanPublic;
  • ecdh_Yc: holds the client's ephemeral ECDH public key as a byte string ECPoint.point, which may represent an elliptic curve point in uncompressed format 🎀
struct {
    select (KeyExchangeAlgorithm) {
        case ec_diffie_hellman: ClientECDiffieHellmanPublic;
    } exchange_keys;
} ClientKeyExchange;

Actions of the sender:

the client picks an ephemeral ECDH public key that matches the parameters it got from the server. The format is the same as in Section 5.4 💖

Actions of the receiver:

the server pulls the client's ephemeral ECDH public key out of the ClientKeyExchange message and checks that it's on the same elliptic curve as the server's ECDH key ✨

5.8. Certificate Verify ✨

Okay bestie, this is the moment the client proves it's really the owner of its certificate. 🔐

When this message is sent:

This message is sent when the client sends a client certificate containing a public key usable for digital signatures. 🦋

Meaning of the message:

This message contains a signature that proves possession of the private key corresponding to the public key in the client's Certificate message. Think of it as showing your ID and then doing the secret handshake 💅

Structure of this message:

The TLS CertificateVerify message and the underlying signature type are defined in the TLS base specifications, and the latter is extended here in Section 5.4. For the "ecdsa" and "eddsa" cases, the signature field in the CertificateVerify message contains an ECDSA or EdDSA (respectively) signature computed over handshake messages exchanged so far, exactly similar to CertificateVerify with other signing algorithms:

        CertificateVerify.signature.sha_hash
            SHA(handshake_messages);
        CertificateVerify.signature.rawdata
            handshake_messages;

ECDSA signatures are computed as described in Section 5.10, and SHA in the above template for sha_hash accordingly may denote a hash algorithm other than SHA-1. 🌸 As per ANSI X9.62, an ECDSA signature consists of a pair of integers, r and s. The digitally-signed element is encoded as an opaque vector <0..2^16-1>, and its contents are the DER encoding [X.690] corresponding to the following ASN.1 notation [X.680]:

        Ecdsa-Sig-Value ::= SEQUENCE {
            r       INTEGER,
            s       INTEGER
        }

EdDSA signatures are generated and verified according to [RFC8032]. The digitally-signed element is encoded as an opaque vector <0..2^16-1>, and its contents include the octet string output of the EdDSA signing algorithm. ✨

Actions of the sender:

The client computes its signature over all handshake messages sent or received starting at client hello and up to but not including this message. It uses the private key corresponding to its certified public key to compute the signature, which is conveyed in the format defined above. 🎀

Actions of the receiver:

The server extracts the client's signature from the CertificateVerify message and verifies the signature using the public key it received in the client's Certificate message. Main character verification energy 👑

5.9. Elliptic Curve Certificates

Okay, certificates with ECC keys have some rules, so listen up 🧚

X.509 certificates containing ECC public keys or signed using ECDSA MUST comply with [RFC3279] or another RFC that replaces or extends it. X.509 certificates containing ECC public keys or signed using EdDSA MUST comply with [RFC8410]. Clients SHOULD use the elliptic curve domain parameters recommended in ANSI X9.62, FIPS 186-4, and SEC 2 [SECG-SEC2], or in [RFC8032].

EdDSA keys using the Ed25519 algorithm MUST use the ed25519 signature algorithm, and Ed448 keys MUST use the ed448 signature algorithm. This document does not define use of Ed25519ph and Ed448ph keys with TLS. Ed25519, Ed25519ph, Ed448, and Ed448ph keys MUST NOT be used with ECDSA. Don't mix them up, bestie 🙈

5.10. ECDH, ECDSA, and RSA Computations

This section is the math-and-rules corner, and it's giving precision 🤓

All ECDH calculations for the NIST curves (including parameter and key generation as well as the shared secret calculation) are performed according to [IEEE.P1363] using the ECKAS-DH1 scheme with the identity map as the Key Derivation Function (KDF) so that the premaster secret is the x-coordinate of the ECDH shared secret elliptic curve point represented as an octet string. Note that this octet string (Z in IEEE 1363 terminology), as output by FE2OSP (Field Element to Octet String Conversion Primitive), has constant length for any given field; leading zeros found in this octet string MUST NOT be truncated. 💀

(Note that this use of the identity KDF is a technicality. The complete picture is that ECDH is employed with a non-trivial KDF because TLS does not directly use the premaster secret for anything other than for computing the master secret. In TLS 1.0 and 1.1, this means that the MD5- and SHA-1-based TLS Pseudorandom Function (PRF) serves as a KDF; in TLS 1.2, the KDF is determined by ciphersuite, and it is conceivable that future TLS versions or new TLS extensions introduced in the future may vary this computation.) 🫶

An ECDHE key exchange using X25519 (curve x25519) goes like this: (1) each party picks a secret key d uniformly at random and computes the corresponding public key x = X25519(d, G); (2) parties exchange their public keys and compute a shared secret as x_S = X25519(d, x_peer); and (3), if either party obtains all-zeroes x_S, it MUST abort the handshake (as required by definition of X25519 and X448). ECDHE for X448 works the same way, replacing X25519 with X448 and x25519 with x448. The derived shared secret is used directly as the premaster secret, which is always exactly 32 bytes when ECDHE with X25519 is used and 56 bytes when ECDHE with X448 is used. 🦋

All ECDSA computations MUST be performed according to ANSI X9.62 or its successors. Data to be signed/verified is hashed, and the result runs directly through the ECDSA algorithm with no additional hashing. A secure hash function such as SHA-256, SHA-384, or SHA-512 from [FIPS.180-4] MUST be used.

All EdDSA computations MUST be performed according to [RFC8032] or its successors. Data to be signed/verified is run through the EdDSA algorithm with no hashing (EdDSA will internally run the data through the "prehash" function PH). The context parameter for Ed448 MUST be set to the empty string.

RFC 4492 anticipated the standardization of a mechanism for specifying the required hash function in the certificate, perhaps in the parameters field of the subjectPublicKeyInfo. Such standardization never took place, and as a result, SHA-1 is used in TLS 1.1 and earlier (except for EdDSA, which uses identity function). TLS 1.2 added a SignatureAndHashAlgorithm parameter to the DigitallySigned struct, thus allowing agility in choosing the signature hash. EdDSA signatures MUST have HashAlgorithm of 8 (Intrinsic). 🌈

All RSA signatures must be generated and verified according to Section 7.2 of [RFC8017].

5.11. Public Key Validation

Public key validation is the bouncer at the door, and nobody gets in without checks 🚪✨

With the NIST curves, each party MUST validate the public key sent by its peer in the ClientKeyExchange and ServerKeyExchange messages. A receiving party MUST check that the x and y parameters from the peer's public value satisfy the curve equation, y^2 = x^3 + ax + b mod p. See Section 2.3 of [Menezes] for details. Failing to do so allows attackers to gain information about the private key to the point that they may recover the entire private key in a few requests if that key is not really ephemeral. 😭 Not a vibe!

With X25519 and X448, a receiving party MUST check whether the computed premaster secret is the all-zero value and abort the handshake if so, as described in Section 6 of [RFC7748].

Ed25519 and Ed448 internally do public key validation as part of signature verification. Built-in bodyguards 💅

6. Cipher Suites

Let's talk cipher suites, the menu of ECC combos you can order 🍓

The table below defines ECC cipher suites that use the key exchange algorithms specified in Section 2.

    +-----------------------------------------+----------------+
    | CipherSuite                             | Identifier     |
    +-----------------------------------------+----------------+
    | TLS_ECDHE_ECDSA_WITH_NULL_SHA           | { 0xC0, 0x06 } |
    | TLS_ECDHE_ECDSA_WITH_3DES_EDE_CBC_SHA   | { 0xC0, 0x08 } |
    | TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA    | { 0xC0, 0x09 } |
    | TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA    | { 0xC0, 0x0A } |
    | TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 | { 0xC0, 0x2B } |
    | TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 | { 0xC0, 0x2C } |
    |                                         |                |
    | TLS_ECDHE_RSA_WITH_NULL_SHA             | { 0xC0, 0x10 } |
    | TLS_ECDHE_RSA_WITH_3DES_EDE_CBC_SHA     | { 0xC0, 0x12 } |
    | TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA      | { 0xC0, 0x13 } |
    | TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA      | { 0xC0, 0x14 } |
    | TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256   | { 0xC0, 0x2F } |
    | TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384   | { 0xC0, 0x30 } |
    |                                         |                |
    | TLS_ECDH_anon_WITH_NULL_SHA             | { 0xC0, 0x15 } |
    | TLS_ECDH_anon_WITH_3DES_EDE_CBC_SHA     | { 0xC0, 0x17 } |
    | TLS_ECDH_anon_WITH_AES_128_CBC_SHA      | { 0xC0, 0x18 } |
    | TLS_ECDH_anon_WITH_AES_256_CBC_SHA      | { 0xC0, 0x19 } |
    +-----------------------------------------+----------------+

                   Table 3: TLS ECC Cipher Suites

The key exchange method, cipher, and hash algorithm for each of these cipher suites are easily determined by examining the name. Ciphers (other than AES ciphers) and hash algorithms are defined in [RFC2246] and [RFC4346]. AES ciphers are defined in [RFC5246], and AES-GCM ciphersuites are in [RFC5289]. Just read the name, it tells you everything 📖

Server implementations SHOULD support all of the following cipher suites, and client implementations SHOULD support at least one of them:

  • TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
  • TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA
  • TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256
  • TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA

7. Implementation Status

Both ECDHE and ECDSA with the NIST curves are widely implemented and supported in all major browsers and all widely used TLS libraries. ECDHE with Curve25519 is by now implemented in several browsers and several TLS libraries including OpenSSL. Curve448 and EdDSA have working interoperable implementations, but they are not yet as widely deployed. The glow-up is real 🌸

8. Security Considerations

Security issues are discussed throughout this memo. 🚨 Okay, this part is seriously important, so pay attention!

For TLS handshakes using ECC cipher suites, the security considerations in Appendix D of each of the three TLS base documents apply accordingly.

Security discussions specific to ECC can be found in [IEEE.P1363] and [ANSI.X9-62.2005]. One important issue that implementers and users must consider is elliptic curve selection. Guidance on selecting an appropriate elliptic curve size is given in Table 1. Security considerations specific to X25519 and X448 are discussed in Section 7 of [RFC7748].

Beyond elliptic curve size, the main issue is elliptic curve structure. As a general principle, it is more conservative to use elliptic curves with as little algebraic structure as possible. Thus, random curves are more conservative than special curves such as Koblitz curves, and curves over F_p with p random are more conservative than curves over F_p with p of a special form, and curves over F_p with p random are considered more conservative than curves over F_2^m as there is no choice between multiple fields of similar size for characteristic 2. 🧐

Another issue is the potential for catastrophic failures when a single elliptic curve is widely used. In this case, an attack on the elliptic curve might result in the compromise of a large number of keys. Again, this concern may need to be balanced against efficiency and interoperability improvements associated with widely used curves. 💀 A single curve breaking could take out a whole lot of keys at once!

Substantial additional information on elliptic curve choice can be found in [IEEE.P1363], [ANSI.X9-62.2005], and [FIPS.186-4].

The Introduction of [RFC8032] lists the security, performance, and operational advantages of EdDSA signatures over ECDSA signatures using the NIST curves. EdDSA really is the clean girl of signatures 💖

All of the key exchange algorithms defined in this document provide forward secrecy. Some of the deprecated key exchange algorithms do not. ⚠️ Read that twice, babe.

9. IANA Considerations ✨

okay so IANA is the registry keeper for all these numbers 🗂️ and [RFC4492], the predecessor of this document, already set up the IANA registries for these three:

  • Supported Groups (Section 5.1)
  • EC Point Format (Section 5.1)
  • EC Curve Type (Section 5.4)

IANA has prepended "TLS" to the names of these three registries 🎀

For each name space, this document sets the initial value assignments and reserves a range for Private Use: 256 values for NamedCurve, and eight values each for ECPointFormat and ECCurveType 🌸 Any extra assignments follow the Specification Required policy. (RFC 4492 required IETF review, fyi.)

All existing entries in the ExtensionType Values, TLS ClientCertificateType Identifiers, TLS Cipher Suites, TLS Supported Groups, TLS EC Point Format, and TLS EC Curve Type registries that pointed at RFC 4492 now point at this document 💅

IANA gave x25519 the value 29 and x448 the value 30 in the TLS Supported Groups registry ✨

IANA also assigned two values in the TLS SignatureAlgorithm registry for ed25519 (7) and ed448 (8), with this document as the reference. This keeps things compatible with TLS 1.3 🦋

And IANA assigned one value from the TLS HashAlgorithm registry for Intrinsic (8), with DTLS-OK set to true (Y) and this document as the reference. Again, this keeps compatibility with TLS 1.3 🫶

10. References

10.1. Normative References

[ANSI.X9-62.2005]
           American National Standards Institute, "Public Key
           Cryptography for the Financial Services Industry: The
           Elliptic Curve Digital Signature Algorithm (ECDSA)",
           ANSI X9.62, November 2005.

[FIPS.186-4]
           National Institute of Standards and Technology, "Digital
           Signature Standard (DSS)", FIPS PUB 186-4,
           DOI 10.6028/NIST.FIPS.186-4, July 2013,
           <http://nvlpubs.nist.gov/nistpubs/FIPS/
           NIST.FIPS.186-4.pdf>.

[RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
           Requirement Levels", BCP 14, RFC 2119,
           DOI 10.17487/RFC2119, March 1997,
           <https://www.rfc-editor.org/info/rfc2119>.

[RFC2246]  Dierks, T. and C. Allen, "The TLS Protocol Version 1.0",
           RFC 2246, DOI 10.17487/RFC2246, January 1999,
           <https://www.rfc-editor.org/info/rfc2246>.

[RFC3279]  Bassham, L., Polk, W., and R. Housley, "Algorithms and
           Identifiers for the Internet X.509 Public Key
           Infrastructure Certificate and Certificate Revocation List
           (CRL) Profile", RFC 3279, DOI 10.17487/RFC3279, April
           2002, <https://www.rfc-editor.org/info/rfc3279>.

[RFC4346]  Dierks, T. and E. Rescorla, "The Transport Layer Security
           (TLS) Protocol Version 1.1", RFC 4346,
           DOI 10.17487/RFC4346, April 2006,
           <https://www.rfc-editor.org/info/rfc4346>.

[RFC4366]  Blake-Wilson, S., Nystrom, M., Hopwood, D., Mikkelsen, J.,
           and T. Wright, "Transport Layer Security (TLS)
           Extensions", RFC 4366, DOI 10.17487/RFC4366, April 2006,
           <https://www.rfc-editor.org/info/rfc4366>.

[RFC5246]  Dierks, T. and E. Rescorla, "The Transport Layer Security
           (TLS) Protocol Version 1.2", RFC 5246,
           DOI 10.17487/RFC5246, August 2008,
           <https://www.rfc-editor.org/info/rfc5246>.

[RFC5289]  Rescorla, E., "TLS Elliptic Curve Cipher Suites with SHA-
           256/384 and AES Galois Counter Mode (GCM)", RFC 5289,
           DOI 10.17487/RFC5289, August 2008,
           <https://www.rfc-editor.org/info/rfc5289>.

[RFC7748]  Langley, A., Hamburg, M., and S. Turner, "Elliptic Curves
           for Security", RFC 7748, DOI 10.17487/RFC7748, January
           2016, <https://www.rfc-editor.org/info/rfc7748>.

[RFC8017]  Moriarty, K., Ed., Kaliski, B., Jonsson, J., and A. Rusch,
           "PKCS #1: RSA Cryptography Specifications Version 2.2",
           RFC 8017, DOI 10.17487/RFC8017, November 2016,
           <https://www.rfc-editor.org/info/rfc8017>.

[RFC8032]  Josefsson, S. and I. Liusvaara, "Edwards-Curve Digital
           Signature Algorithm (EdDSA)", RFC 8032,
           DOI 10.17487/RFC8032, January 2017,
           <https://www.rfc-editor.org/info/rfc8032>.

[RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
           2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
           May 2017, <https://www.rfc-editor.org/info/rfc8174>.

[RFC8410]  Josefsson, S. and J. Schaad, "Algorithm Identifiers for
           Ed25519, Ed448, X25519 and X448 for Use in the Internet
           X.509 Public Key Infrastructure", RFC 8410,
           DOI 10.17487/RFC8410, August 2018,
           <https://www.rfc-editor.org/info/rfc8410>.

[SECG-SEC2]
           Certicom Research, "SEC 2: Recommended Elliptic Curve
           Domain Parameters", Standards for Efficient Cryptography 2
           (SEC 2), Version 2.0, January 2010,
           <http://www.secg.org/sec2-v2.pdf>.

[X.680]    ITU-T, "Abstract Syntax Notation One (ASN.1):
           Specification of basic notation", ITU-T Recommendation
           X.680, ISO/IEC 8824-1, August 2015.

[X.690]    ITU-T, "Information technology-ASN.1 encoding rules:
           Specification of Basic Encoding Rules (BER), Canonical
           Encoding Rules (CER) and Distinguished Encoding Rules
           (DER)", ITU-T Recommendation X.690, ISO/IEC 8825-1, August
           2015.

10.2. Informative References

[FIPS.180-4]
           National Institute of Standards and Technology, "Secure
           Hash Standard (SHS)", FIPS PUB 180-4, DOI
           10.6028/NIST.FIPS.180-4, August 2015,
           <http://nvlpubs.nist.gov/nistpubs/FIPS/
           NIST.FIPS.180-4.pdf>.

[IEEE.P1363]
           IEEE, "Standard Specifications for Public Key
           Cryptography", IEEE Std P1363,
           <http://ieeexplore.ieee.org/document/891000/>.

[Menezes]  Menezes, A. and B. Ustaoglu, "On reusing ephemeral keys in
           Diffie-Hellman key agreement protocols", International
           Journal of Applied Cryptography, Vol. 2, Issue 2,
           DOI 10.1504/IJACT.2010.038308, January 2010.

[RFC4492]  Blake-Wilson, S., Bolyard, N., Gupta, V., Hawk, C., and B.
           Moeller, "Elliptic Curve Cryptography (ECC) Cipher Suites
           for Transport Layer Security (TLS)", RFC 4492,
           DOI 10.17487/RFC4492, May 2006,
           <https://www.rfc-editor.org/info/rfc4492>.

[RFC7919]  Gillmor, D., "Negotiated Finite Field Diffie-Hellman
           Ephemeral Parameters for Transport Layer Security (TLS)",
           RFC 7919, DOI 10.17487/RFC7919, August 2016,
           <https://www.rfc-editor.org/info/rfc7919>.

[TLS1.3]   Rescorla, E., "The Transport Layer Security (TLS) Protocol
           Version 1.3", Work in Progress, draft-ietf-tls-tls13-28,
           March 2018.

Appendix A. Equivalent Curves (Informative)

All of the NIST curves [FIPS.186-4] and several of the ANSI curves [ANSI.X9-62.2005] are equivalent to curves listed in Section 5.1.1 🦋 Here's the lineup of curve names that different standards orgs picked. If several names share a row, they're aliases for the same curve 💜

              +-----------+------------+------------+
              | SECG      | ANSI X9.62 | NIST       |
              +-----------+------------+------------+
              | sect163k1 |            | NIST K-163 |
              | sect163r1 |            |            |
              | sect163r2 |            | NIST B-163 |
              | sect193r1 |            |            |
              | sect193r2 |            |            |
              | sect233k1 |            | NIST K-233 |
              | sect233r1 |            | NIST B-233 |
              | sect239k1 |            |            |
              | sect283k1 |            | NIST K-283 |
              | sect283r1 |            | NIST B-283 |
              | sect409k1 |            | NIST K-409 |
              | sect409r1 |            | NIST B-409 |
              | sect571k1 |            | NIST K-571 |
              | sect571r1 |            | NIST B-571 |
              | secp160k1 |            |            |
              | secp160r1 |            |            |
              | secp160r2 |            |            |
              | secp192k1 |            |            |
              | secp192r1 | prime192v1 | NIST P-192 |
              | secp224k1 |            |            |
              | secp224r1 |            | NIST P-224 |
              | secp256k1 |            |            |
              | secp256r1 | prime256v1 | NIST P-256 |
              | secp384r1 |            | NIST P-384 |
              | secp521r1 |            | NIST P-521 |
              +-----------+------------+------------+

Table 4: Equivalent Curves Defined by SECG, ANSI, and NIST

Appendix B. Differences from RFC 4492

Okay, so here's what changed from RFC 4492 🌈

  • Renamed EllipticCurveList to NamedCurveList.
  • Added TLS 1.2.
  • Merged errata.
  • Removed the ECDH key exchange algorithms: ECDH_RSA and ECDH_ECDSA
  • Deprecated a bunch of ciphersuites:
    • TLS_ECDH_ECDSA_WITH_NULL_SHA
    • TLS_ECDH_ECDSA_WITH_RC4_128_SHA
    • TLS_ECDH_ECDSA_WITH_3DES_EDE_CBC_SHA
    • TLS_ECDH_ECDSA_WITH_AES_128_CBC_SHA
    • TLS_ECDH_ECDSA_WITH_AES_256_CBC_SHA
    • TLS_ECDH_RSA_WITH_NULL_SHA
    • TLS_ECDH_RSA_WITH_RC4_128_SHA
    • TLS_ECDH_RSA_WITH_3DES_EDE_CBC_SHA
    • TLS_ECDH_RSA_WITH_AES_128_CBC_SHA
    • TLS_ECDH_RSA_WITH_AES_256_CBC_SHA
    • All the other RC4 ciphersuites
  • Removed unused curves and all but the uncompressed point format.
  • Added X25519 and X448. ✨
  • Deprecated explicit curves.
  • Removed restriction on signature algorithm in certificate.

Acknowledgements

Most of the text in this document came from [RFC4492], the predecessor of this document 🫶 The authors of that one were:

  • Simon Blake-Wilson
  • Nelson Bolyard
  • Vipul Gupta
  • Chris Hawk
  • Bodo Moeller

In the predecessor document, the authors thanked Bill Anderson and Tim Dierks for their contributions. The authors of this document would also like to thank Nikos Mavrogiannopoulos, Martin Thomson, and Tanja Lange for their contributions 💖

Authors' Addresses

Yoav Nir
Check Point Software Technologies Ltd.
5 Hasolelim st.
Tel Aviv  6789735
Israel

Email: ynir.ietf@gmail.com


Simon Josefsson
SJD AB

Email: simon@josefsson.org


Manuel Pegourie-Gonnard
ARM

Email: mpg@elzevir.fr
📦 where this page comes from