draft-ietf-tls-tls13-13.txt   draft-ietf-tls-tls13-14.txt 
Network Working Group E. Rescorla Network Working Group E. Rescorla
Internet-Draft RTFM, Inc. Internet-Draft RTFM, Inc.
Obsoletes: 5077, 5246, 5746 (if May 22, 2016 Obsoletes: 5077, 5246, 5746 (if July 11, 2016
approved) approved)
Updates: 4492, 6066, 6961 (if approved) Updates: 4492, 6066, 6961 (if approved)
Intended status: Standards Track Intended status: Standards Track
Expires: November 23, 2016 Expires: January 12, 2017
The Transport Layer Security (TLS) Protocol Version 1.3 The Transport Layer Security (TLS) Protocol Version 1.3
draft-ietf-tls-tls13-13 draft-ietf-tls-tls13-14
Abstract Abstract
This document specifies Version 1.3 of the Transport Layer Security This document specifies version 1.3 of the Transport Layer Security
(TLS) protocol. The TLS protocol allows client/server applications (TLS) protocol. TLS allows client/server applications to communicate
to communicate over the Internet in a way that is designed to prevent over the Internet in a way that is designed to prevent eavesdropping,
eavesdropping, tampering, and message forgery. tampering, and message forgery.
Status of This Memo Status of This Memo
This Internet-Draft is submitted in full conformance with the This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79. provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts is at http://datatracker.ietf.org/drafts/current/. Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress." material or to cite them other than as "work in progress."
This Internet-Draft will expire on November 23, 2016. This Internet-Draft will expire on January 12, 2017.
Copyright Notice Copyright Notice
Copyright (c) 2016 IETF Trust and the persons identified as the Copyright (c) 2016 IETF Trust and the persons identified as the
document authors. All rights reserved. document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of (http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents publication of this document. Please review these documents
skipping to change at page 2, line 24 skipping to change at page 2, line 24
outside the IETF Standards Process, and derivative works of it may outside the IETF Standards Process, and derivative works of it may
not be created outside the IETF Standards Process, except to format not be created outside the IETF Standards Process, except to format
it for publication as an RFC or to translate it into languages other it for publication as an RFC or to translate it into languages other
than English. than English.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1. Conventions and Terminology . . . . . . . . . . . . . . . 5 1.1. Conventions and Terminology . . . . . . . . . . . . . . . 5
1.2. Major Differences from TLS 1.2 . . . . . . . . . . . . . 6 1.2. Major Differences from TLS 1.2 . . . . . . . . . . . . . 6
2. Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2. Protocol Overview . . . . . . . . . . . . . . . . . . . . . . 10
3. Goals of This Document . . . . . . . . . . . . . . . . . . . 10 2.1. Incorrect DHE Share . . . . . . . . . . . . . . . . . . . 14
4. Presentation Language . . . . . . . . . . . . . . . . . . . . 11 2.2. Resumption and Pre-Shared Key (PSK) . . . . . . . . . . . 15
4.1. Basic Block Size . . . . . . . . . . . . . . . . . . . . 11 2.3. Zero-RTT Data . . . . . . . . . . . . . . . . . . . . . . 17
4.2. Miscellaneous . . . . . . . . . . . . . . . . . . . . . . 11 3. Presentation Language . . . . . . . . . . . . . . . . . . . . 18
4.3. Vectors . . . . . . . . . . . . . . . . . . . . . . . . . 11 3.1. Basic Block Size . . . . . . . . . . . . . . . . . . . . 18
4.4. Numbers . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.2. Miscellaneous . . . . . . . . . . . . . . . . . . . . . . 19
4.5. Enumerateds . . . . . . . . . . . . . . . . . . . . . . . 13 3.3. Vectors . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.6. Constructed Types . . . . . . . . . . . . . . . . . . . . 14 3.4. Numbers . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.6.1. Variants . . . . . . . . . . . . . . . . . . . . . . 14 3.5. Enumerateds . . . . . . . . . . . . . . . . . . . . . . . 20
4.7. Constants . . . . . . . . . . . . . . . . . . . . . . . . 15 3.6. Constructed Types . . . . . . . . . . . . . . . . . . . . 21
4.8. Cryptographic Attributes . . . . . . . . . . . . . . . . 15 3.6.1. Variants . . . . . . . . . . . . . . . . . . . . . . 21
4.8.1. Digital Signing . . . . . . . . . . . . . . . . . . . 16 3.7. Constants . . . . . . . . . . . . . . . . . . . . . . . . 23
4.8.2. Authenticated Encryption with Additional Data (AEAD) 17 4. Handshake Protocol . . . . . . . . . . . . . . . . . . . . . 23
5. The TLS Record Protocol . . . . . . . . . . . . . . . . . . . 17 4.1. Key Exchange Messages . . . . . . . . . . . . . . . . . . 24
5.1. Connection States . . . . . . . . . . . . . . . . . . . . 18 4.1.1. Client Hello . . . . . . . . . . . . . . . . . . . . 25
5.2. Record Layer . . . . . . . . . . . . . . . . . . . . . . 20 4.1.2. Server Hello . . . . . . . . . . . . . . . . . . . . 27
5.2.1. Fragmentation . . . . . . . . . . . . . . . . . . . . 20 4.1.3. Hello Retry Request . . . . . . . . . . . . . . . . . 29
5.2.2. Record Payload Protection . . . . . . . . . . . . . . 22 4.2. Hello Extensions . . . . . . . . . . . . . . . . . . . . 30
5.2.3. Record Padding . . . . . . . . . . . . . . . . . . . 24 4.2.1. Cookie . . . . . . . . . . . . . . . . . . . . . . . 31
6. The TLS Handshaking Protocols . . . . . . . . . . . . . . . . 25 4.2.2. Signature Algorithms . . . . . . . . . . . . . . . . 32
6.1. Alert Protocol . . . . . . . . . . . . . . . . . . . . . 26 4.2.3. Negotiated Groups . . . . . . . . . . . . . . . . . . 35
6.1.1. Closure Alerts . . . . . . . . . . . . . . . . . . . 27 4.2.4. Key Share . . . . . . . . . . . . . . . . . . . . . . 36
6.1.2. Error Alerts . . . . . . . . . . . . . . . . . . . . 29 4.2.5. Pre-Shared Key Extension . . . . . . . . . . . . . . 39
6.2. Handshake Protocol Overview . . . . . . . . . . . . . . . 32 4.2.6. Early Data Indication . . . . . . . . . . . . . . . . 40
6.2.1. Incorrect DHE Share . . . . . . . . . . . . . . . . . 35 4.2.7. OCSP Status Extensions . . . . . . . . . . . . . . . 43
6.2.2. Resumption and Pre-Shared Key (PSK) . . . . . . . . . 36 4.2.8. Encrypted Extensions . . . . . . . . . . . . . . . . 44
6.2.3. Zero-RTT Data . . . . . . . . . . . . . . . . . . . . 38 4.2.9. Certificate Request . . . . . . . . . . . . . . . . . 44
6.3. Handshake Protocol . . . . . . . . . . . . . . . . . . . 39 4.3. Authentication Messages . . . . . . . . . . . . . . . . . 46
6.3.1. Key Exchange Messages . . . . . . . . . . . . . . . . 40 4.3.1. Certificate . . . . . . . . . . . . . . . . . . . . . 47
6.3.2. Hello Extensions . . . . . . . . . . . . . . . . . . 46 4.3.2. Certificate Verify . . . . . . . . . . . . . . . . . 51
6.3.3. Server Parameters . . . . . . . . . . . . . . . . . . 60 4.3.3. Finished . . . . . . . . . . . . . . . . . . . . . . 53
6.3.4. Authentication Messages . . . . . . . . . . . . . . . 63 4.4. Post-Handshake Messages . . . . . . . . . . . . . . . . . 54
6.3.5. Post-Handshake Messages . . . . . . . . . . . . . . . 71 4.4.1. New Session Ticket Message . . . . . . . . . . . . . 54
7. Cryptographic Computations . . . . . . . . . . . . . . . . . 74 4.4.2. Post-Handshake Authentication . . . . . . . . . . . . 56
7.1. Key Schedule . . . . . . . . . . . . . . . . . . . . . . 74 4.4.3. Key and IV Update . . . . . . . . . . . . . . . . . . 57
7.2. Updating Traffic Keys and IVs . . . . . . . . . . . . . . 76 5. Record Protocol . . . . . . . . . . . . . . . . . . . . . . . 58
7.3. Traffic Key Calculation . . . . . . . . . . . . . . . . . 76 5.1. Record Layer . . . . . . . . . . . . . . . . . . . . . . 58
7.3.1. Diffie-Hellman . . . . . . . . . . . . . . . . . . . 77 5.2. Record Payload Protection . . . . . . . . . . . . . . . . 59
7.3.2. Elliptic Curve Diffie-Hellman . . . . . . . . . . . . 78 5.3. Per-Record Nonce . . . . . . . . . . . . . . . . . . . . 61
7.3.3. Exporters . . . . . . . . . . . . . . . . . . . . . . 78 5.4. Record Padding . . . . . . . . . . . . . . . . . . . . . 62
8. Mandatory Algorithms . . . . . . . . . . . . . . . . . . . . 79 5.5. Limits on Key Usage . . . . . . . . . . . . . . . . . . . 63
8.1. MTI Cipher Suites . . . . . . . . . . . . . . . . . . . . 79 6. Alert Protocol . . . . . . . . . . . . . . . . . . . . . . . 63
8.2. MTI Extensions . . . . . . . . . . . . . . . . . . . . . 79 6.1. Closure Alerts . . . . . . . . . . . . . . . . . . . . . 65
9. Application Data Protocol . . . . . . . . . . . . . . . . . . 80 6.2. Error Alerts . . . . . . . . . . . . . . . . . . . . . . 66
10. Security Considerations . . . . . . . . . . . . . . . . . . . 80 7. Cryptographic Computations . . . . . . . . . . . . . . . . . 69
11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 80 7.1. Key Schedule . . . . . . . . . . . . . . . . . . . . . . 69
12. References . . . . . . . . . . . . . . . . . . . . . . . . . 83 7.2. Updating Traffic Keys and IVs . . . . . . . . . . . . . . 72
12.1. Normative References . . . . . . . . . . . . . . . . . . 83 7.3. Traffic Key Calculation . . . . . . . . . . . . . . . . . 72
12.2. Informative References . . . . . . . . . . . . . . . . . 86 7.3.1. Diffie-Hellman . . . . . . . . . . . . . . . . . . . 73
Appendix A. Protocol Data Structures and Constant Values . . . . 92 7.3.2. Elliptic Curve Diffie-Hellman . . . . . . . . . . . . 74
A.1. Record Layer . . . . . . . . . . . . . . . . . . . . . . 92 7.3.3. Exporters . . . . . . . . . . . . . . . . . . . . . . 74
A.2. Alert Messages . . . . . . . . . . . . . . . . . . . . . 92 8. Compliance Requirements . . . . . . . . . . . . . . . . . . . 74
A.3. Handshake Protocol . . . . . . . . . . . . . . . . . . . 94 8.1. MTI Cipher Suites . . . . . . . . . . . . . . . . . . . . 75
A.3.1. Key Exchange Messages . . . . . . . . . . . . . . . . 94 8.2. MTI Extensions . . . . . . . . . . . . . . . . . . . . . 75
A.3.2. Server Parameters Messages . . . . . . . . . . . . . 98 9. Security Considerations . . . . . . . . . . . . . . . . . . . 76
A.3.3. Authentication Messages . . . . . . . . . . . . . . . 99 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 76
A.3.4. Ticket Establishment . . . . . . . . . . . . . . . . 99 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 79
A.4. Cipher Suites . . . . . . . . . . . . . . . . . . . . . . 100 11.1. Normative References . . . . . . . . . . . . . . . . . . 79
A.4.1. Unauthenticated Operation . . . . . . . . . . . . . . 105 11.2. Informative References . . . . . . . . . . . . . . . . . 82
A.5. The Security Parameters . . . . . . . . . . . . . . . . . 105 Appendix A. Protocol Data Structures and Constant Values . . . . 89
A.6. Changes to RFC 4492 . . . . . . . . . . . . . . . . . . . 106 A.1. Record Layer . . . . . . . . . . . . . . . . . . . . . . 89
Appendix B. Implementation Notes . . . . . . . . . . . . . . . . 107 A.2. Alert Messages . . . . . . . . . . . . . . . . . . . . . 89
B.1. Random Number Generation and Seeding . . . . . . . . . . 107 A.3. Handshake Protocol . . . . . . . . . . . . . . . . . . . 91
B.2. Certificates and Authentication . . . . . . . . . . . . . 107 A.3.1. Key Exchange Messages . . . . . . . . . . . . . . . . 91
B.3. Cipher Suite Support . . . . . . . . . . . . . . . . . . 107 A.3.2. Server Parameters Messages . . . . . . . . . . . . . 95
B.4. Implementation Pitfalls . . . . . . . . . . . . . . . . . 107 A.3.3. Authentication Messages . . . . . . . . . . . . . . . 96
B.5. Client Tracking Prevention . . . . . . . . . . . . . . . 109 A.3.4. Ticket Establishment . . . . . . . . . . . . . . . . 96
Appendix C. Backward Compatibility . . . . . . . . . . . . . . . 109 A.4. Cipher Suites . . . . . . . . . . . . . . . . . . . . . . 97
C.1. Negotiating with an older server . . . . . . . . . . . . 110 A.4.1. Unauthenticated Operation . . . . . . . . . . . . . . 102
C.2. Negotiating with an older client . . . . . . . . . . . . 111 Appendix B. Implementation Notes . . . . . . . . . . . . . . . . 102
C.3. Zero-RTT backwards compatibility . . . . . . . . . . . . 111 B.1. Random Number Generation and Seeding . . . . . . . . . . 102
C.4. Backwards Compatibility Security Restrictions . . . . . . 111 B.2. Certificates and Authentication . . . . . . . . . . . . . 103
Appendix D. Security Analysis . . . . . . . . . . . . . . . . . 112 B.3. Cipher Suite Support . . . . . . . . . . . . . . . . . . 103
D.1. Handshake Protocol . . . . . . . . . . . . . . . . . . . 113 B.4. Implementation Pitfalls . . . . . . . . . . . . . . . . . 103
D.1.1. Authentication and Key Exchange . . . . . . . . . . . 113 B.5. Client Tracking Prevention . . . . . . . . . . . . . . . 105
D.1.2. Version Rollback Attacks . . . . . . . . . . . . . . 114
D.1.3. Detecting Attacks Against the Handshake Protocol . . 114 Appendix C. Backward Compatibility . . . . . . . . . . . . . . . 105
D.2. Protecting Application Data . . . . . . . . . . . . . . . 114 C.1. Negotiating with an older server . . . . . . . . . . . . 106
D.3. Denial of Service . . . . . . . . . . . . . . . . . . . . 115 C.2. Negotiating with an older client . . . . . . . . . . . . 106
D.4. Final Notes . . . . . . . . . . . . . . . . . . . . . . . 115 C.3. Zero-RTT backwards compatibility . . . . . . . . . . . . 107
Appendix E. Working Group Information . . . . . . . . . . . . . 115 C.4. Backwards Compatibility Security Restrictions . . . . . . 107
Appendix F. Contributors . . . . . . . . . . . . . . . . . . . . 115 Appendix D. Overview of Security Properties . . . . . . . . . . 108
Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 119 D.1. Handshake . . . . . . . . . . . . . . . . . . . . . . . . 108
D.2. Record Layer . . . . . . . . . . . . . . . . . . . . . . 110
Appendix E. Working Group Information . . . . . . . . . . . . . 112
Appendix F. Contributors . . . . . . . . . . . . . . . . . . . . 112
Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 116
1. Introduction 1. Introduction
DISCLAIMER: This is a WIP draft of TLS 1.3 and has not yet seen DISCLAIMER: This is a WIP draft of TLS 1.3 and has not yet seen
significant security analysis. significant security analysis.
RFC EDITOR: PLEASE REMOVE THE FOLLOWING PARAGRAPH The source for this RFC EDITOR: PLEASE REMOVE THE FOLLOWING PARAGRAPH The source for this
draft is maintained in GitHub. Suggested changes should be submitted draft is maintained in GitHub. Suggested changes should be submitted
as pull requests at https://github.com/tlswg/tls13-spec. as pull requests at https://github.com/tlswg/tls13-spec.
Instructions are on that page as well. Editorial changes can be Instructions are on that page as well. Editorial changes can be
managed in GitHub, but any substantive change should be discussed on managed in GitHub, but any substantive change should be discussed on
the TLS mailing list. the TLS mailing list.
The primary goal of the TLS protocol is to provide privacy and data The primary goal of TLS is to provide a secure channel between two
integrity between two communicating peers. The TLS protocol is communicating peers. Specifically, the channel should provide the
composed of two layers: the TLS Record Protocol and the TLS Handshake following properties.
Protocol. At the lowest level, layered on top of some reliable
transport protocol (e.g., TCP [RFC0793]), is the TLS Record Protocol.
The TLS Record Protocol provides connection security that has two
basic properties:
- The connection is private. Symmetric cryptography is used for - Authentication: The server side of the channel is always
data encryption (e.g., AES [AES]). The keys for this symmetric authenticated; the client side is optionally authenticated.
encryption are generated uniquely for each connection and are Authentication can happen via asymmetric cryptography (e.g., RSA
based on a secret negotiated by another the TLS Handshake [RSA], ECDSA [ECDSA]) or a pre-shared symmetric key.
Protocol.
- The connection is reliable. Messages include an authentication - Confidentiality: Data sent over the channel is not visible to
tag which protects them against modification. attackers.
Note: The TLS Record Protocol can operate in an insecure mode but is - Integrity: Data sent over the channel cannot be modified by
generally only used in this mode while another protocol is using the attackers.
TLS Record Protocol as a transport for negotiating security
parameters.
The TLS Record Protocol is used for encapsulation of various higher- These properties should be true even in the face of an attacker who
level protocols. One such encapsulated protocol, the TLS Handshake has complete control of the network, as described in [RFC3552]. See
Protocol, allows the server and client to authenticate each other and Appendix D for a more complete statement of the relevant security
to negotiate an encryption algorithm and cryptographic keys before properties.
the application protocol transmits or receives its first byte of
data. The TLS Handshake Protocol provides connection security that
has three basic properties:
- The peer's identity can be authenticated using asymmetric (public TLS consists of two primary components:
key) cryptography (e.g., RSA [RSA], ECDSA [ECDSA]) or a pre-shared
symmetric key. The TLS server is always authenticated; client
authentication is optional.
- The negotiation of a shared secret is secure: the negotiated - A handshake protocol (Section 4) which authenticates the
secret is unavailable to eavesdroppers, and for any authenticated communicating parties, negotiates cryptographic modes and
connection the secret cannot be obtained, even by an attacker who parameters, and establishes shared keying material. The handshake
can place himself in the middle of the connection. protocol is designed to resist tampering; an active attacker
should not be able to force the peers to negotiate different
parameters than they would if the connection were not under
attack.
- The negotiation is reliable: no attacker can modify the - A record protocol (Section 5) which uses the parameters
negotiation communication without being detected by the parties to established by the handshake protocol to protect traffic between
the communication. the communicating peers. The record protocol divides traffic up
into a series of records, each of which is independently protected
using the traffic keys.
One advantage of TLS is that it is application protocol independent. TLS is application protocol independent; higher-level protocols can
Higher-level protocols can layer on top of the TLS protocol layer on top of TLS transparently. The TLS standard, however, does
transparently. The TLS standard, however, does not specify how not specify how protocols add security with TLS; the decisions on how
protocols add security with TLS; the decisions on how to initiate TLS to initiate TLS handshaking and how to interpret the authentication
handshaking and how to interpret the authentication certificates certificates exchanged are left to the judgment of the designers and
exchanged are left to the judgment of the designers and implementors implementors of protocols that run on top of TLS.
of protocols that run on top of TLS.
This document defines TLS version 1.3. While TLS 1.3 is not directly
compatible with previous versions, all versions of TLS incorporate a
versioning mechanism which allows clients and servers to
interoperably negotiate a common version if one is supported.
1.1. Conventions and Terminology 1.1. Conventions and Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in RFC "OPTIONAL" in this document are to be interpreted as described in RFC
2119 [RFC2119]. 2119 [RFC2119].
The following terms are used: The following terms are used:
skipping to change at page 6, line 14 skipping to change at page 6, line 14
sender: An endpoint that is transmitting records. sender: An endpoint that is transmitting records.
session: An association between a client and a server resulting from session: An association between a client and a server resulting from
a handshake. a handshake.
server: The endpoint which did not initiate the TLS connection. server: The endpoint which did not initiate the TLS connection.
1.2. Major Differences from TLS 1.2 1.2. Major Differences from TLS 1.2
draft-13 draft-14
- Allow cookies to be longer (*)
- Remove the "context" from EarlyDataIndication as it was undefined
and nobody used it (*)
- Remove 0-RTT EncryptedExtensions and replace the ticket_age
extension with an obfuscated version. Also necessitates a change
to NewSessionTicket (*).
- Move the downgrade sentinel to the end of ServerHello.Random to
accomodate tlsdate (*).
- Define ecdsa_sha1 (*).
- Allow resumption even after fatal alerts. This matches current
practice.
- Remove non-closure warning alerts. Require treating unknown
alerts as fatal.
- Make the rules for accepting 0-RTT less restrictive.
- Clarify 0-RTT backward-compatibility rules.
- Clarify how 0-RTT and PSK identities interact.
- Add a section describing the data limits for each cipher.
- Major editorial restructuring.
- Replace the Security Analysis section with a WIP draft.
(*) indicates changes to the wire protocol which may require
implementations to update.
draft-13
- Allow server to send SupportedGroups. - Allow server to send SupportedGroups.
- Remove 0-RTT client authentication - Remove 0-RTT client authentication
- Remove (EC)DHE 0-RTT. - Remove (EC)DHE 0-RTT.
- Flesh out 0-RTT PSK mode and shrink EarlyDataIndiation - Flesh out 0-RTT PSK mode and shrink EarlyDataIndication
- Turn PSK-resumption response into an index to save room - Turn PSK-resumption response into an index to save room
- Move CertificateStatus to an extension - Move CertificateStatus to an extension
- Extra fields in NewSessionTicket. - Extra fields in NewSessionTicket.
- Restructure key schedule and add a resumption_context value. - Restructure key schedule and add a resumption_context value.
- Require DH public keys and secrets to be zero-padded to the size - Require DH public keys and secrets to be zero-padded to the size
skipping to change at page 8, line 15 skipping to change at page 9, line 5
- Change HKDF labeling to include protocol version and value - Change HKDF labeling to include protocol version and value
lengths. lengths.
- Shift the final decision to abort a handshake due to incompatible - Shift the final decision to abort a handshake due to incompatible
certificates to the client rather than having servers abort early. certificates to the client rather than having servers abort early.
- Deprecate SHA-1 with signatures. - Deprecate SHA-1 with signatures.
- Add MTI algorithms. - Add MTI algorithms.
draft-08
- Remove support for weak and lesser used named curves. - Remove support for weak and lesser used named curves.
- Remove support for MD5 and SHA-224 hashes with signatures. - Remove support for MD5 and SHA-224 hashes with signatures.
- Update lists of available AEAD cipher suites and error alerts. - Update lists of available AEAD cipher suites and error alerts.
- Reduce maximum permitted record expansion for AEAD from 2048 to - Reduce maximum permitted record expansion for AEAD from 2048 to
256 octets. 256 octets.
- Require digital signatures even when a previous configuration is - Require digital signatures even when a previous configuration is
skipping to change at page 10, line 9 skipping to change at page 10, line 48
- Rework handshake to provide 1-RTT mode. - Rework handshake to provide 1-RTT mode.
- Remove custom DHE groups. - Remove custom DHE groups.
- Remove support for compression. - Remove support for compression.
- Remove support for static RSA and DH key exchange. - Remove support for static RSA and DH key exchange.
- Remove support for non-AEAD ciphers. - Remove support for non-AEAD ciphers.
2. Goals 2. Protocol Overview
The goals of the TLS protocol, in order of priority, are as follows: The cryptographic parameters of the session state are produced by the
TLS handshake protocol. When a TLS client and server first start
communicating, they agree on a protocol version, select cryptographic
algorithms, optionally authenticate each other, and establish shared
secret keying material. Once the handshake is complete, the peers
use the established keys to protect application layer traffic.
1. Cryptographic security: TLS should be used to establish a secure TLS supports three basic key exchange modes:
connection between two parties.
2. Interoperability: Independent programmers should be able to - Diffie-Hellman (of both the finite field and elliptic curve
develop applications utilizing TLS that can successfully exchange varieties).
cryptographic parameters without knowledge of one another's code.
3. Extensibility: TLS seeks to provide a framework into which new - A pre-shared symmetric key (PSK)
public key and record protection methods can be incorporated as
necessary. This will also accomplish two sub-goals: preventing
the need to create a new protocol (and risking the introduction
of possible new weaknesses) and avoiding the need to implement an
entire new security library.
4. Relative efficiency: Cryptographic operations tend to be highly - A combination of a symmetric key and Diffie-Hellman
CPU intensive, particularly public key operations. For this
reason, the TLS protocol has incorporated an optional session
caching scheme to reduce the number of connections that need to
be established from scratch. Additionally, care has been taken
to reduce network activity.
3. Goals of This Document Which mode is used depends on the negotiated cipher suite.
Conceptually, the handshake establishes three secrets which are used
to derive all the keys.
This document and the TLS protocol itself have evolved from the SSL Figure 1 below shows the basic full TLS handshake:
3.0 Protocol Specification as published by Netscape. The differences
between this version and previous versions are significant enough
that the various versions of TLS and SSL 3.0 do not interoperate
(although each protocol incorporates a mechanism by which an
implementation can back down to prior versions). This document is
intended primarily for readers who will be implementing the protocol
and for those doing cryptographic analysis of it. The specification
has been written with this in mind, and it is intended to reflect the
needs of those two groups. For that reason, many of the algorithm-
dependent data structures and rules are included in the body of the
text (as opposed to in an appendix), providing easier access to them.
This document is not intended to supply any details of service Client Server
definition or of interface definition, although it does cover select
areas of policy as they are required for the maintenance of solid
security.
4. Presentation Language Key ^ ClientHello
Exch | + key_share*
v + pre_shared_key* -------->
ServerHello ^ Key
+ key_share* | Exch
+ pre_shared_key* v
{EncryptedExtensions} ^ Server
{CertificateRequest*} v Params
{Certificate*} ^
{CertificateVerify*} | Auth
{Finished} v
<-------- [Application Data*]
^ {Certificate*}
Auth | {CertificateVerify*}
v {Finished} -------->
[Application Data] <-------> [Application Data]
+ Indicates extensions sent in the
previously noted message.
* Indicates optional or situation-dependent
messages that are not always sent.
{} Indicates messages protected using keys
derived from handshake_traffic_secret.
[] Indicates messages protected using keys
derived from traffic_secret_N
Figure 1: Message flow for full TLS Handshake
The handshake can be thought of as having three phases, indicated in
the diagram above.
- Key Exchange: Establish shared keying material and select the
cryptographic parameters. Everything after this phase is
encrypted.
- Server Parameters: Establish other handshake parameters. (whether
the client is authenticated, application layer protocol support,
etc.)
- Authentication: Authenticate the server (and optionally the
client) and provide key confirmation and handshake integrity.
In the Key Exchange phase, the client sends the ClientHello
(Section 4.1.1) message, which contains a random nonce
(ClientHello.random), its offered protocol version, cipher suite, and
extensions, and in general either one or more Diffie-Hellman key
shares (in the "key_share" extension Section 4.2.4), one or more pre-
shared key labels (in the "pre_shared_key" extension Section 4.2.5),
or both.
The server processes the ClientHello and determines the appropriate
cryptographic parameters for the connection. It then responds with
its own ServerHello which indicates the negotiated connection
parameters. [Section 4.1.2]. The combination of the ClientHello and
the ServerHello determines the shared keys. If either a pure (EC)DHE
or (EC)DHE-PSK cipher suite is in use, then the ServerHello will
contain a "key_share" extension with the server's ephemeral Diffie-
Hellman share which MUST be in the same group as one of the client's
shares. If a pure PSK or an (EC)DHE-PSK cipher suite is negotiated,
then the ServerHello will contain a "pre_shared_key" extension
indicating which of the client's offered PSKs was selected.
The server then sends two messages to establish the Server
Parameters:
EncryptedExtensions. responses to any extensions which are not
required in order to determine the cryptographic parameters.
[Section 4.2.8]
CertificateRequest. if certificate-based client authentication is
desired, the desired parameters for that certificate. This
message will be omitted if client authentication is not desired.
Finally, the client and server exchange Authentication messages. TLS
uses the same set of messages every time that authentication is
needed. Specifically:
Certificate. the certificate of the endpoint. This message is
omitted if the server is not authenticating with a certificate
(i.e., with PSK or (EC)DHE-PSK cipher suites). Note that if raw
public keys [RFC7250] or the cached information extension
[I-D.ietf-tls-cached-info] are in use, then this message will not
contain a certificate but rather some other value corresponding to
the server's long-term key. [Section 4.3.1]
CertificateVerify. a signature over the entire handshake using the
public key in the Certificate message. This message is omitted if
the server is not authenticating via a certificate (i.e., with PSK
or (EC)DHE-PSK cipher suites). [Section 4.3.2]
Finished. a MAC (Message Authentication Code) over the entire
handshake. This message provides key confirmation, binds the
endpoint's identity to the exchanged keys, and in PSK mode also
authenticates the handshake. [Section 4.3.3]
Upon receiving the server's messages, the client responds with its
Authentication messages, namely Certificate and CertificateVerify (if
requested), and Finished.
At this point, the handshake is complete, and the client and server
may exchange application layer data. Application data MUST NOT be
sent prior to sending the Finished message. Note that while the
server may send application data prior to receiving the client's
Authentication messages, any data sent at that point is, of course,
being sent to an unauthenticated peer.
2.1. Incorrect DHE Share
If the client has not provided a sufficient "key_share" extension
(e.g. it includes only DHE or ECDHE groups unacceptable or
unsupported by the server), the server corrects the mismatch with a
HelloRetryRequest and the client will need to restart the handshake
with an appropriate "key_share" extension, as shown in Figure 2. If
no common cryptographic parameters can be negotiated, the server will
send a "handshake_failure" or "insufficient_security" fatal alert
(see Section 6).
Client Server
ClientHello
+ key_share -------->
<-------- HelloRetryRequest
ClientHello
+ key_share -------->
ServerHello
+ key_share
{EncryptedExtensions}
{CertificateRequest*}
{Certificate*}
{CertificateVerify*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
[Application Data] <-------> [Application Data]
Figure 2: Message flow for a full handshake with mismatched
parameters
Note: The handshake transcript includes the initial ClientHello/
HelloRetryRequest exchange; it is not reset with the new ClientHello.
TLS also allows several optimized variants of the basic handshake, as
described in the following sections.
2.2. Resumption and Pre-Shared Key (PSK)
Although TLS PSKs can be established out of band, PSKs can also be
established in a previous session and then reused ("session
resumption"). Once a handshake has completed, the server can send
the client a PSK identity which corresponds to a key derived from the
initial handshake (See Section 4.4.1). The client can then use that
PSK identity in future handshakes to negotiate use of the PSK. If
the server accepts it, then the security context of the new
connection is tied to the original connection. In TLS 1.2 and below,
this functionality was provided by "session IDs" and "session
tickets" [RFC5077]. Both mechanisms are obsoleted in TLS 1.3.
PSK cipher suites can either use PSK in combination with an (EC)DHE
exchange in order to provide forward secrecy in combination with
shared keys, or can use PSKs alone, at the cost of losing forward
secrecy.
Figure 3 shows a pair of handshakes in which the first establishes a
PSK and the second uses it:
Client Server
Initial Handshake:
ClientHello
+ key_share -------->
ServerHello
+ key_share
{EncryptedExtensions}
{CertificateRequest*}
{Certificate*}
{CertificateVerify*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
<-------- [NewSessionTicket]
[Application Data] <-------> [Application Data]
Subsequent Handshake:
ClientHello
+ pre_shared_key
+ key_share* -------->
ServerHello
+ pre_shared_key
+ key_share*
{EncryptedExtensions}
{Finished}
<-------- [Application Data*]
{Finished} -------->
[Application Data] <-------> [Application Data]
Figure 3: Message flow for resumption and PSK
As the server is authenticating via a PSK, it does not send a
Certificate or a CertificateVerify. When a client offers resumption
via PSK it SHOULD also supply a "key_share" extension to the server
as well to allow the server to decline resumption and fall back to a
full handshake, if needed. A "key_share" extension MUST also be sent
if the client is attempting to negotiate an (EC)DHE-PSK cipher suite.
2.3. Zero-RTT Data
When resuming via a PSK with an appropriate ticket (i.e., one with
the "allow_early_data" flag), clients can also send data on their
first flight ("early data"). This data is encrypted solely under
keys derived using the first offered PSK as the static secret. As
shown in Figure 4, the Zero-RTT data is just added to the 1-RTT
handshake in the first flight. The rest of the handshake uses the
same messages.
Client Server
ClientHello
+ early_data
+ pre_shared_key
+ key_share*
(Finished)
(Application Data*)
(end_of_early_data) -------->
ServerHello
+ early_data
+ pre_shared_key
+ key_share*
{EncryptedExtensions}
{CertificateRequest*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
[Application Data] <-------> [Application Data]
* Indicates optional or situation-dependent
messages that are not always sent.
() Indicates messages protected using keys
derived from early_traffic_secret.
{} Indicates messages protected using keys
derived from handshake_traffic_secret.
[] Indicates messages protected using keys
derived from traffic_secret_N
Figure 4: Message flow for a zero round trip handshake
[[OPEN ISSUE: Should it be possible to combine 0-RTT with the server
authenticating via a signature https://github.com/tlswg/tls13-spec/
issues/443]]
IMPORTANT NOTE: The security properties for 0-RTT data (regardless of
the cipher suite) are weaker than those for other kinds of TLS data.
Specifically:
1. This data is not forward secret, because it is encrypted solely
with the PSK.
2. There are no guarantees of non-replay between connections.
Unless the server takes special measures outside those provided
by TLS, the server has no guarantee that the same 0-RTT data was
not transmitted on multiple 0-RTT connections (See
Section 4.2.6.2 for more details). This is especially relevant
if the data is authenticated either with TLS client
authentication or inside the application layer protocol.
However, 0-RTT data cannot be duplicated within a connection
(i.e., the server will not process the same data twice for the
same connection) and an attacker will not be able to make 0-RTT
data appear to be 1-RTT data (because it is protected with
different keys.)
The remainder of this document provides a detailed description of
TLS.
3. Presentation Language
This document deals with the formatting of data in an external This document deals with the formatting of data in an external
representation. The following very basic and somewhat casually representation. The following very basic and somewhat casually
defined presentation syntax will be used. The syntax draws from defined presentation syntax will be used. The syntax draws from
several sources in its structure. Although it resembles the several sources in its structure. Although it resembles the
programming language "C" in its syntax and XDR [RFC4506] in both its programming language "C" in its syntax and XDR [RFC4506] in both its
syntax and intent, it would be risky to draw too many parallels. The syntax and intent, it would be risky to draw too many parallels. The
purpose of this presentation language is to document TLS only; it has purpose of this presentation language is to document TLS only; it has
no general application beyond that particular goal. no general application beyond that particular goal.
4.1. Basic Block Size 3.1. Basic Block Size
The representation of all data items is explicitly specified. The The representation of all data items is explicitly specified. The
basic data block size is one byte (i.e., 8 bits). Multiple byte data basic data block size is one byte (i.e., 8 bits). Multiple byte data
items are concatenations of bytes, from left to right, from top to items are concatenations of bytes, from left to right, from top to
bottom. From the byte stream, a multi-byte item (a numeric in the bottom. From the byte stream, a multi-byte item (a numeric in the
example) is formed (using C notation) by: example) is formed (using C notation) by:
value = (byte[0] << 8*(n-1)) | (byte[1] << 8*(n-2)) | value = (byte[0] << 8*(n-1)) | (byte[1] << 8*(n-2)) |
... | byte[n-1]; ... | byte[n-1];
This byte ordering for multi-byte values is the commonplace network This byte ordering for multi-byte values is the commonplace network
byte order or big-endian format. byte order or big-endian format.
4.2. Miscellaneous 3.2. Miscellaneous
Comments begin with "/*" and end with "*/". Comments begin with "/*" and end with "*/".
Optional components are denoted by enclosing them in "[[ ]]" double Optional components are denoted by enclosing them in "[[ ]]" double
brackets. brackets.
Single-byte entities containing uninterpreted data are of type Single-byte entities containing uninterpreted data are of type
opaque. opaque.
4.3. Vectors 3.3. Vectors
A vector (single-dimensioned array) is a stream of homogeneous data A vector (single-dimensioned array) is a stream of homogeneous data
elements. The size of the vector may be specified at documentation elements. The size of the vector may be specified at documentation
time or left unspecified until runtime. In either case, the length time or left unspecified until runtime. In either case, the length
declares the number of bytes, not the number of elements, in the declares the number of bytes, not the number of elements, in the
vector. The syntax for specifying a new type, T', that is a fixed- vector. The syntax for specifying a new type, T', that is a fixed-
length vector of type T is length vector of type T is
T T'[n]; T T'[n];
skipping to change at page 12, line 29 skipping to change at page 20, line 4
in the byte stream. The length will be in the form of a number in the byte stream. The length will be in the form of a number
consuming as many bytes as required to hold the vector's specified consuming as many bytes as required to hold the vector's specified
maximum (ceiling) length. A variable-length vector with an actual maximum (ceiling) length. A variable-length vector with an actual
length field of zero is referred to as an empty vector. length field of zero is referred to as an empty vector.
T T'<floor..ceiling>; T T'<floor..ceiling>;
In the following example, mandatory is a vector that must contain In the following example, mandatory is a vector that must contain
between 300 and 400 bytes of type opaque. It can never be empty. between 300 and 400 bytes of type opaque. It can never be empty.
The actual length field consumes two bytes, a uint16, which is The actual length field consumes two bytes, a uint16, which is
sufficient to represent the value 400 (see Section 4.4). On the sufficient to represent the value 400 (see Section 3.4). On the
other hand, longer can represent up to 800 bytes of data, or 400 other hand, longer can represent up to 800 bytes of data, or 400
uint16 elements, and it may be empty. Its encoding will include a uint16 elements, and it may be empty. Its encoding will include a
two-byte actual length field prepended to the vector. The length of two-byte actual length field prepended to the vector. The length of
an encoded vector must be an even multiple of the length of a single an encoded vector must be an even multiple of the length of a single
element (for example, a 17-byte vector of uint16 would be illegal). element (for example, a 17-byte vector of uint16 would be illegal).
opaque mandatory<300..400>; opaque mandatory<300..400>;
/* length field is 2 bytes, cannot be empty */ /* length field is 2 bytes, cannot be empty */
uint16 longer<0..800>; uint16 longer<0..800>;
/* zero to 400 16-bit unsigned integers */ /* zero to 400 16-bit unsigned integers */
4.4. Numbers 3.4. Numbers
The basic numeric data type is an unsigned byte (uint8). All larger The basic numeric data type is an unsigned byte (uint8). All larger
numeric data types are formed from fixed-length series of bytes numeric data types are formed from fixed-length series of bytes
concatenated as described in Section 4.1 and are also unsigned. The concatenated as described in Section 3.1 and are also unsigned. The
following numeric types are predefined. following numeric types are predefined.
uint8 uint16[2]; uint8 uint16[2];
uint8 uint24[3]; uint8 uint24[3];
uint8 uint32[4]; uint8 uint32[4];
uint8 uint64[8]; uint8 uint64[8];
All values, here and elsewhere in the specification, are stored in All values, here and elsewhere in the specification, are stored in
network byte (big-endian) order; the uint32 represented by the hex network byte (big-endian) order; the uint32 represented by the hex
bytes 01 02 03 04 is equivalent to the decimal value 16909060. bytes 01 02 03 04 is equivalent to the decimal value 16909060.
Note that in some cases (e.g., DH parameters) it is necessary to Note that in some cases (e.g., DH parameters) it is necessary to
represent integers as opaque vectors. In such cases, they are represent integers as opaque vectors. In such cases, they are
represented as unsigned integers (i.e., additional leading zero represented as unsigned integers (i.e., additional leading zero
octets are not used even if the most significant bit is set). octets are not used even if the most significant bit is set).
4.5. Enumerateds 3.5. Enumerateds
An additional sparse data type is available called enum. A field of An additional sparse data type is available called enum. A field of
type enum can only assume the values declared in the definition. type enum can only assume the values declared in the definition.
Each definition is a different type. Only enumerateds of the same Each definition is a different type. Only enumerateds of the same
type may be assigned or compared. Every element of an enumerated type may be assigned or compared. Every element of an enumerated
must be assigned a value, as demonstrated in the following example. must be assigned a value, as demonstrated in the following example.
Since the elements of the enumerated are not ordered, they can be Since the elements of the enumerated are not ordered, they can be
assigned any unique value, in any order. assigned any unique value, in any order.
enum { e1(v1), e2(v2), ... , en(vn) [[, (n)]] } Te; enum { e1(v1), e2(v2), ... , en(vn) [[, (n)]] } Te;
skipping to change at page 14, line 5 skipping to change at page 21, line 29
specified. specified.
Color color = Color.blue; /* overspecified, legal */ Color color = Color.blue; /* overspecified, legal */
Color color = blue; /* correct, type implicit */ Color color = blue; /* correct, type implicit */
For enumerateds that are never converted to external representation, For enumerateds that are never converted to external representation,
the numerical information may be omitted. the numerical information may be omitted.
enum { low, medium, high } Amount; enum { low, medium, high } Amount;
4.6. Constructed Types 3.6. Constructed Types
Structure types may be constructed from primitive types for Structure types may be constructed from primitive types for
convenience. Each specification declares a new, unique type. The convenience. Each specification declares a new, unique type. The
syntax for definition is much like that of C. syntax for definition is much like that of C.
struct { struct {
T1 f1; T1 f1;
T2 f2; T2 f2;
... ...
Tn fn; Tn fn;
} [[T]]; } [[T]];
The fields within a structure may be qualified using the type's name, The fields within a structure may be qualified using the type's name,
with a syntax much like that available for enumerateds. For example, with a syntax much like that available for enumerateds. For example,
T.f2 refers to the second field of the previous declaration. T.f2 refers to the second field of the previous declaration.
Structure definitions may be embedded. Structure definitions may be embedded.
4.6.1. Variants 3.6.1. Variants
Defined structures may have variants based on some knowledge that is Defined structures may have variants based on some knowledge that is
available within the environment. The selector must be an enumerated available within the environment. The selector must be an enumerated
type that defines the possible variants the structure defines. There type that defines the possible variants the structure defines. There
must be a case arm for every element of the enumeration declared in must be a case arm for every element of the enumeration declared in
the select. Case arms have limited fall-through: if two case arms the select. Case arms have limited fall-through: if two case arms
follow in immediate succession with no fields in between, then they follow in immediate succession with no fields in between, then they
both contain the same fields. Thus, in the example below, "orange" both contain the same fields. Thus, in the example below, "orange"
and "banana" both contain V2. Note that this is a new piece of and "banana" both contain V2. Note that this is a new piece of
syntax in TLS 1.2. syntax in TLS 1.2.
skipping to change at page 15, line 29 skipping to change at page 23, line 5
struct { struct {
select (VariantTag) { /* value of selector is implicit */ select (VariantTag) { /* value of selector is implicit */
case apple: case apple:
V1; /* VariantBody, tag = apple */ V1; /* VariantBody, tag = apple */
case orange: case orange:
case banana: case banana:
V2; /* VariantBody, tag = orange or banana */ V2; /* VariantBody, tag = orange or banana */
} variant_body; /* optional label on variant */ } variant_body; /* optional label on variant */
} VariantRecord; } VariantRecord;
4.7. Constants 3.7. Constants
Typed constants can be defined for purposes of specification by Typed constants can be defined for purposes of specification by
declaring a symbol of the desired type and assigning values to it. declaring a symbol of the desired type and assigning values to it.
Under-specified types (opaque, variable-length vectors, and Under-specified types (opaque, variable-length vectors, and
structures that contain opaque) cannot be assigned values. No fields structures that contain opaque) cannot be assigned values. No fields
of a multi-element structure or vector may be elided. of a multi-element structure or vector may be elided.
For example: For example:
struct { struct {
uint8 f1; uint8 f1;
uint8 f2; uint8 f2;
} Example1; } Example1;
Example1 ex1 = {1, 4}; /* assigns f1 = 1, f2 = 4 */ Example1 ex1 = {1, 4}; /* assigns f1 = 1, f2 = 4 */
4.8. Cryptographic Attributes 4. Handshake Protocol
The two cryptographic operations -- digital signing, and
authenticated encryption with additional data (AEAD) -- are
designated digitally-signed, and aead-ciphered, respectively. A
field's cryptographic processing is specified by prepending an
appropriate key word designation before the field's type
specification. Cryptographic keys are implied by the current session
state (see Section 5.1).
4.8.1. Digital Signing
A digitally-signed element is encoded as a struct DigitallySigned:
struct {
SignatureScheme algorithm;
opaque signature<0..2^16-1>;
} DigitallySigned;
The algorithm field specifies the algorithm used (see Section 6.3.2.2
for the definition of this field). The signature is a digital
signature using those algorithms over the contents of the element.
The contents themselves do not appear on the wire but are simply
calculated. The length of the signature is specified by the signing
algorithm and key.
In previous versions of TLS, the ServerKeyExchange format meant that
attackers can obtain a signature of a message with a chosen, 32-byte
prefix. Because TLS 1.3 servers are likely to also implement prior
versions, the contents of the element always start with 64 bytes of
octet 32 in order to clear that chosen-prefix.
Following that padding is a context string used to disambiguate
signatures for different purposes. The context string will be
specified whenever a digitally-signed element is used. A single 0
byte is appended to the context to act as a separator.
Finally, the specified contents of the digitally-signed structure
follow the 0 byte after the context string. (See the example at the
end of this section.)
The combined input is then fed into the corresponding signature
algorithm to produce the signature value on the wire. See
Section 6.3.2.2 for algorithms defined in this specification.
In the following example
struct {
uint8 field1;
uint8 field2;
digitally-signed opaque {
uint8 field3<0..255>;
uint8 field4;
};
} UserType;
Assume that the context string for the signature was specified as
"Example". The input for the signature/hash algorithm would be:
2020202020202020202020202020202020202020202020202020202020202020
2020202020202020202020202020202020202020202020202020202020202020
4578616d706c6500
followed by the encoding of the inner struct (field3 and field4).
The length of the structure, in bytes, would be equal to two bytes
for field1 and field2, plus two bytes for the signature algorithm,
plus two bytes for the length of the signature, plus the length of
the output of the signing algorithm. The length of the signature is
known because the algorithm and key used for the signing are known
prior to encoding or decoding this structure.
4.8.2. Authenticated Encryption with Additional Data (AEAD)
In AEAD encryption, the plaintext is simultaneously encrypted and
integrity protected. The input may be of any length, and aead-
ciphered output is generally larger than the input in order to
accommodate the integrity check value.
5. The TLS Record Protocol
The TLS Record Protocol takes messages to be transmitted, fragments
the data into manageable blocks, protects the records, and transmits
the result. Received data is decrypted and verified, reassembled,
and then delivered to higher-level clients.
Three protocols that use the TLS Record Protocol are described in
this document: the TLS Handshake Protocol, the Alert Protocol, and
the application data protocol. In order to allow extension of the
TLS protocol, additional record content types can be supported by the
TLS Record Protocol. New record content type values are assigned by
IANA in the TLS Content Type Registry as described in Section 11.
Implementations MUST NOT send record types not defined in this
document unless negotiated by some extension. If a TLS
implementation receives an unexpected record type, it MUST send an
"unexpected_message" alert.
Any protocol designed for use over TLS must be carefully designed to
deal with all possible attacks against it. As a practical matter,
this means that the protocol designer must be aware of what security
properties TLS does and does not provide and cannot safely rely on
the latter.
Note in particular that the length of a record or absence of traffic
itself is not protected by encryption unless the sender uses the
supplied padding mechanism - see Section 5.2.3 for more details.
5.1. Connection States
[[TODO: I plan to totally rewrite or remove this. IT seems like just
cruft.]]
A TLS connection state is the operating environment of the TLS Record
Protocol. It specifies a record protection algorithm and its
parameters as well as the record protection keys and IVs for the
connection in both the read and the write directions. The security
parameters are set by the TLS Handshake Protocol, which also
determines when new cryptographic keys are installed and used for
record protection. The initial current state always specifies that
records are not protected.
The security parameters for a TLS Connection read and write state are
set by providing the following values:
connection end
Whether this entity is considered the "client" or the "server" in
this connection.
Hash algorithm
An algorithm used to generate keys from the appropriate secret
(see Section 7.1 and Section 7.3).
record protection algorithm
The algorithm to be used for record protection. This algorithm
must be of the AEAD type and thus provides integrity and
confidentiality as a single primitive. This specification
includes the key size of this algorithm and of the nonce for the
AEAD algorithm.
master secret
A 48-byte secret shared between the two peers in the connection
and used to generate keys for protecting data.
client random
A 32-byte value provided by the client.
server random
A 32-byte value provided by the server.
These parameters are defined in the presentation language as:
enum { server, client } ConnectionEnd;
enum { tls_kdf_sha256, tls_kdf_sha384 } KDFAlgorithm;
enum { aes_gcm } RecordProtAlgorithm;
/* The algorithms specified in KDFAlgorithm and
RecordProtAlgorithm may be added to. */
struct {
ConnectionEnd entity;
KDFAlgorithm kdf_algorithm;
RecordProtAlgorithm record_prot_algorithm;
uint8 enc_key_length;
uint8 iv_length;
opaque hs_master_secret[48];
opaque master_secret[48];
opaque client_random[32];
opaque server_random[32];
} SecurityParameters;
[TODO: update this to handle new key hierarchy.]
The connection state will use the security parameters to generate the
following four items:
client write key
server write key
client write iv
server write iv
The client write parameters are used by the server when receiving and
processing records and vice versa. The algorithm used for generating
these items from the security parameters is described in Section 7.3.
Once the security parameters have been set and the keys have been
generated, the connection states can be instantiated by making them
the current states. These current states MUST be updated for each
record processed. Each connection state includes the following
elements:
cipher state
The current state of the encryption algorithm. This will consist
of the scheduled key for that connection.
sequence number
Each connection state contains a sequence number, which is
maintained separately for read and write states. The sequence
number is set to zero at the beginning of a connection, and
whenever the key is changed. The sequence number is incremented
after each record: specifically, the first record transmitted
under a particular connection state and record key MUST use
sequence number 0. Sequence numbers are of type uint64 and MUST
NOT exceed 2^64-1. Sequence numbers do not wrap. If a TLS
implementation would need to wrap a sequence number, it MUST
either rekey (Section 6.3.5.3) or terminate the connection.
5.2. Record Layer
The TLS record layer receives uninterpreted data from higher layers
in non-empty blocks of arbitrary size.
5.2.1. Fragmentation
The record layer fragments information blocks into TLSPlaintext
records carrying data in chunks of 2^14 bytes or less. Message
boundaries are not preserved in the record layer (i.e., multiple
messages of the same ContentType MAY be coalesced into a single
TLSPlaintext record, or a single message MAY be fragmented across
several records). Alert messages (Section 6.1) MUST NOT be
fragmented across records.
struct {
uint8 major;
uint8 minor;
} ProtocolVersion;
enum {
alert(21),
handshake(22),
application_data(23)
(255)
} ContentType;
struct {
ContentType type;
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length;
opaque fragment[TLSPlaintext.length];
} TLSPlaintext;
type
The higher-level protocol used to process the enclosed fragment.
record_version
The protocol version the current record is compatible with. This
value MUST be set to { 3, 1 } for all records. This field is
deprecated and MUST be ignored for all purposes.
length
The length (in bytes) of the following TLSPlaintext.fragment. The
length MUST NOT exceed 2^14.
fragment
The application data. This data is transparent and treated as an
independent block to be dealt with by the higher-level protocol
specified by the type field.
This document describes TLS Version 1.3, which uses the version { 3,
4 }. The version value 3.4 is historical, deriving from the use of {
3, 1 } for TLS 1.0 and { 3, 0 } for SSL 3.0. In order to maximize
backwards compatibility, the record layer version identifies as
simply TLS 1.0. Endpoints supporting other versions negotiate the
version to use by following the procedure and requirements in
Appendix C.
Implementations MUST NOT send zero-length fragments of Handshake or
Alert types, even if those fragments contain padding. Zero-length
fragments of Application data MAY be sent as they are potentially
useful as a traffic analysis countermeasure.
When record protection has not yet been engaged, TLSPlaintext
structures are written directly onto the wire. Once record
protection has started, TLSPlaintext records are protected and sent
as described in the following section.
5.2.2. Record Payload Protection
The record protection functions translate a TLSPlaintext structure
into a TLSCiphertext. The deprotection functions reverse the
process. In TLS 1.3 as opposed to previous versions of TLS, all
ciphers are modeled as "Authenticated Encryption with Additional
Data" (AEAD) [RFC5116]. AEAD functions provide a unified encryption
and authentication operation which turns plaintext into authenticated
ciphertext and back again.
AEAD ciphers take as input a single key, a nonce, a plaintext, and
"additional data" to be included in the authentication check, as
described in Section 2.1 of [RFC5116]. The key is either the
client_write_key or the server_write_key and in TLS 1.3 the
additional data input is empty (zero length).
struct {
ContentType opaque_type = application_data(23); /* see fragment.type */
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length;
aead-ciphered struct {
opaque content[TLSPlaintext.length];
ContentType type;
uint8 zeros[length_of_padding];
} fragment;
} TLSCiphertext;
opaque_type
The outer opaque_type field of a TLSCiphertext record is always
set to the value 23 (application_data) for outward compatibility
with middleboxes accustomed to parsing previous versions of TLS.
The actual content type of the record is found in fragment.type
after decryption.
record_version
The record_version field is identical to
TLSPlaintext.record_version and is always { 3, 1 }. Note that the
handshake protocol including the ClientHello and ServerHello
messages authenticates the protocol version, so this value is
redundant.
length
The length (in bytes) of the following TLSCiphertext.fragment.
The length MUST NOT exceed 2^14 + 256. An endpoint that receives
a record that exceeds this length MUST generate a fatal
"record_overflow" alert.
fragment.content
The cleartext of TLSPlaintext.fragment.
fragment.type
The actual content type of the record.
fragment.zeros
An arbitrary-length run of zero-valued bytes may appear in the
cleartext after the type field. This provides an opportunity for
senders to pad any TLS record by a chosen amount as long as the
total stays within record size limits. See Section 5.2.3 for more
details.
fragment
The AEAD encrypted form of TLSPlaintext.fragment +
TLSPlaintext.type + zeros, where "+" denotes concatenation.
The length of the per-record nonce (iv_length) is set to max(8 bytes,
N_MIN) for the AEAD algorithm (see [RFC5116] Section 4). An AEAD
algorithm where N_MAX is less than 8 bytes MUST NOT be used with TLS.
The per-record nonce for the AEAD construction is formed as follows:
1. The 64-bit record sequence number is padded to the left with
zeroes to iv_length.
2. The padded sequence number is XORed with the static
client_write_iv or server_write_iv, depending on the role.
The resulting quantity (of length iv_length) is used as the per-
record nonce.
Note: This is a different construction from that in TLS 1.2, which
specified a partially explicit nonce.
The plaintext is the concatenation of TLSPlaintext.fragment and
TLSPlaintext.type.
The AEAD output consists of the ciphertext output by the AEAD
encryption operation. The length of the plaintext is greater than
TLSPlaintext.length due to the inclusion of TLSPlaintext.type and
however much padding is supplied by the sender. The length of
aead_output will generally be larger than the plaintext, but by an
amount that varies with the AEAD cipher. Since the ciphers might
incorporate padding, the amount of overhead could vary with different
lengths of plaintext. Symbolically,
AEADEncrypted =
AEAD-Encrypt(write_key, nonce, plaintext of fragment)
In order to decrypt and verify, the cipher takes as input the key,
nonce, and the AEADEncrypted value. The output is either the
plaintext or an error indicating that the decryption failed. There
is no separate integrity check. That is:
plaintext of fragment =
AEAD-Decrypt(write_key, nonce, AEADEncrypted)
If the decryption fails, a fatal "bad_record_mac" alert MUST be
generated.
An AEAD cipher MUST NOT produce an expansion of greater than 255
bytes. An endpoint that receives a record from its peer with
TLSCipherText.length larger than 2^14 + 256 octets MUST generate a
fatal "record_overflow" alert. This limit is derived from the
maximum TLSPlaintext length of 2^14 octets + 1 octet for ContentType
+ the maximum AEAD expansion of 255 octets.
5.2.3. Record Padding
All encrypted TLS records can be padded to inflate the size of the
TLSCipherText. This allows the sender to hide the size of the
traffic from an observer.
When generating a TLSCiphertext record, implementations MAY choose to
pad. An unpadded record is just a record with a padding length of
zero. Padding is a string of zero-valued bytes appended to the
ContentType field before encryption. Implementations MUST set the
padding octets to all zeros before encrypting.
Application Data records may contain a zero-length fragment.content
if the sender desires. This permits generation of plausibly-sized
cover traffic in contexts where the presence or absence of activity
may be sensitive. Implementations MUST NOT send Handshake or Alert
records that have a zero-length fragment.content.
The padding sent is automatically verified by the record protection
mechanism: Upon successful decryption of a TLSCiphertext.fragment,
the receiving implementation scans the field from the end toward the
beginning until it finds a non-zero octet. This non-zero octet is
the content type of the message. This padding scheme was selected
because it allows padding of any encrypted TLS record by an arbitrary
size (from zero up to TLS record size limits) without introducing new
content types. The design also enforces all-zero padding octets,
which allows for quick detection of padding errors.
Implementations MUST limit their scanning to the cleartext returned
from the AEAD decryption. If a receiving implementation does not
find a non-zero octet in the cleartext, it should treat the record as
having an unexpected ContentType, sending an "unexpected_message"
alert.
The presence of padding does not change the overall record size
limitations - the full fragment plaintext may not exceed 2^14 octets.
Selecting a padding policy that suggests when and how much to pad is
a complex topic, and is beyond the scope of this specification. If
the application layer protocol atop TLS permits padding, it may be
preferable to pad application_data TLS records within the application
layer. Padding for encrypted handshake and alert TLS records must
still be handled at the TLS layer, though. Later documents may
define padding selection algorithms, or define a padding policy
request mechanism through TLS extensions or some other means.
6. The TLS Handshaking Protocols
TLS has two subprotocols that are used to allow peers to agree upon
security parameters for the record layer, to authenticate themselves,
to instantiate negotiated security parameters, and to report error
conditions to each other.
The TLS Handshake Protocol is responsible for negotiating a session,
which consists of the following items:
peer certificate
X509v3 [RFC5280] certificate of the peer. This element of the
state may be null.
cipher spec
Specifies the authentication and key establishment algorithms, the
hash for use with HKDF to generate keying material, and the record
protection algorithm (See Appendix A.5 for formal definition.)
resumption master secret
a secret shared between the client and server that can be used as
a pre-shared symmetric key (PSK) in future connections.
These items are then used to create security parameters for use by
the record layer when protecting application data. Many connections
can be instantiated using the same session using a PSK established in
an initial handshake.
6.1. Alert Protocol
One of the content types supported by the TLS record layer is the
alert type. Alert messages convey the severity of the message
(warning or fatal) and a description of the alert. Alert messages
with a level of fatal result in the immediate termination of the
connection. In this case, other connections corresponding to the
session may continue, but the session identifier MUST be invalidated,
preventing the failed session from being used to establish new
connections. Like other messages, alert messages are encrypted as
specified by the current connection state.
enum { warning(1), fatal(2), (255) } AlertLevel;
enum {
close_notify(0),
end_of_early_data(1),
unexpected_message(10), /* fatal */
bad_record_mac(20), /* fatal */
record_overflow(22), /* fatal */
handshake_failure(40), /* fatal */
bad_certificate(42),
unsupported_certificate(43),
certificate_revoked(44),
certificate_expired(45),
certificate_unknown(46),
illegal_parameter(47), /* fatal */
unknown_ca(48), /* fatal */
access_denied(49), /* fatal */
decode_error(50), /* fatal */
decrypt_error(51), /* fatal */
protocol_version(70), /* fatal */
insufficient_security(71), /* fatal */
internal_error(80), /* fatal */
inappropriate_fallback(86), /* fatal */
user_canceled(90),
missing_extension(109), /* fatal */
unsupported_extension(110), /* fatal */
certificate_unobtainable(111),
unrecognized_name(112),
bad_certificate_status_response(113), /* fatal */
bad_certificate_hash_value(114), /* fatal */
unknown_psk_identity(115),
(255)
} AlertDescription;
struct {
AlertLevel level;
AlertDescription description;
} Alert;
6.1.1. Closure Alerts
The client and the server must share knowledge that the connection is
ending in order to avoid a truncation attack. Failure to properly
close a connection does not prohibit a session from being resumed.
close_notify
This alert notifies the recipient that the sender will not send
any more messages on this connection. Any data received after a
closure MUST be ignored.
end_of_early_data
This alert is sent by the client to indicate that all 0-RTT
application_data messages have been transmitted (or none will be
sent at all) and that this is the end of the flight. This alert
MUST be at the warning level. Servers MUST NOT send this alert
and clients receiving it MUST terminate the connection with an
"unexpected_message" alert.
user_canceled
This alert notifies the recipient that the sender is canceling the
handshake for some reason unrelated to a protocol failure. If a
user cancels an operation after the handshake is complete, just
closing the connection by sending a "close_notify" is more
appropriate. This alert SHOULD be followed by a "close_notify".
This alert is generally a warning.
Either party MAY initiate a close by sending a "close_notify" alert.
Any data received after a closure alert is ignored. If a transport-
level close is received prior to a "close_notify", the receiver
cannot know that all the data that was sent has been received.
Each party MUST send a "close_notify" alert before closing the write
side of the connection, unless some other fatal alert has been
transmitted. The other party MUST respond with a "close_notify"
alert of its own and close down the connection immediately,
discarding any pending writes. The initiator of the close need not
wait for the responding "close_notify" alert before closing the read
side of the connection.
If the application protocol using TLS provides that any data may be
carried over the underlying transport after the TLS connection is
closed, the TLS implementation must receive the responding
"close_notify" alert before indicating to the application layer that
the TLS connection has ended. If the application protocol will not
transfer any additional data, but will only close the underlying
transport connection, then the implementation MAY choose to close the
transport without waiting for the responding "close_notify". No part
of this standard should be taken to dictate the manner in which a
usage profile for TLS manages its data transport, including when
connections are opened or closed.
Note: It is assumed that closing a connection reliably delivers
pending data before destroying the transport.
6.1.2. Error Alerts
Error handling in the TLS Handshake Protocol is very simple. When an
error is detected, the detecting party sends a message to its peer.
Upon transmission or receipt of a fatal alert message, both parties
immediately close the connection. Servers and clients MUST forget
any session-identifiers, keys, and secrets associated with a failed
connection. Thus, any connection terminated with a fatal alert MUST
NOT be resumed.
Whenever an implementation encounters a condition which is defined as
a fatal alert, it MUST send the appropriate alert prior to closing
the connection. For all errors where an alert level is not
explicitly specified, the sending party MAY determine at its
discretion whether to treat this as a fatal error or not. If the
implementation chooses to send an alert but intends to close the
connection immediately afterwards, it MUST send that alert at the
fatal alert level.
If an alert with a level of warning is sent and received, generally
the connection can continue normally. If the receiving party decides
not to proceed with the connection (e.g., after having received a
"user_canceled" alert that it is not willing to accept), it SHOULD
send a fatal alert to terminate the connection. Given this, the
sending peer cannot, in general, know how the receiving party will
behave. Therefore, warning alerts are not very useful when the
sending party wants to continue the connection, and thus are
sometimes omitted. For example, if a party decides to accept an
expired certificate (perhaps after confirming this with the user) and
wants to continue the connection, it would not generally send a
"certificate_expired" alert.
The following error alerts are defined:
unexpected_message
An inappropriate message was received. This alert is always fatal
and should never be observed in communication between proper
implementations.
bad_record_mac
This alert is returned if a record is received which cannot be
deprotected. Because AEAD algorithms combine decryption and
verification, this alert is used for all deprotection failures.
This alert is always fatal and should never be observed in
communication between proper implementations (except when messages
were corrupted in the network).
record_overflow
A TLSCiphertext record was received that had a length more than
2^14 + 256 bytes, or a record decrypted to a TLSPlaintext record
with more than 2^14 bytes. This alert is always fatal and should
never be observed in communication between proper implementations
(except when messages were corrupted in the network).
handshake_failure
Reception of a "handshake_failure" alert message indicates that
the sender was unable to negotiate an acceptable set of security
parameters given the options available. This alert is always
fatal.
bad_certificate
A certificate was corrupt, contained signatures that did not
verify correctly, etc.
unsupported_certificate
A certificate was of an unsupported type.
certificate_revoked
A certificate was revoked by its signer.
certificate_expired
A certificate has expired or is not currently valid.
certificate_unknown
Some other (unspecified) issue arose in processing the
certificate, rendering it unacceptable.
illegal_parameter
A field in the handshake was out of range or inconsistent with
other fields. This alert is always fatal.
unknown_ca
A valid certificate chain or partial chain was received, but the
certificate was not accepted because the CA certificate could not
be located or couldn't be matched with a known, trusted CA. This
alert is always fatal.
access_denied
A valid certificate or PSK was received, but when access control
was applied, the sender decided not to proceed with negotiation.
This alert is always fatal.
decode_error
A message could not be decoded because some field was out of the
specified range or the length of the message was incorrect. This
alert is always fatal and should never be observed in
communication between proper implementations (except when messages
were corrupted in the network).
decrypt_error
A handshake cryptographic operation failed, including being unable
to correctly verify a signature or validate a Finished message.
This alert is always fatal.
protocol_version
The protocol version the peer has attempted to negotiate is
recognized but not supported. (For example, old protocol versions
might be avoided for security reasons.) This alert is always
fatal.
insufficient_security
Returned instead of "handshake_failure" when a negotiation has
failed specifically because the server requires ciphers more
secure than those supported by the client. This alert is always
fatal.
internal_error
An internal error unrelated to the peer or the correctness of the
protocol (such as a memory allocation failure) makes it impossible
to continue. This alert is always fatal.
inappropriate_fallback
Sent by a server in response to an invalid connection retry
attempt from a client. (see [RFC7507]) This alert is always fatal.
missing_extension
Sent by endpoints that receive a hello message not containing an
extension that is mandatory to send for the offered TLS version.
This message is always fatal. [[TODO: IANA Considerations.]]
unsupported_extension
Sent by endpoints receiving any hello message containing an
extension known to be prohibited for inclusion in the given hello
message, including any extensions in a ServerHello not first
offered in the corresponding ClientHello. This alert is always
fatal.
certificate_unobtainable
Sent by servers when unable to obtain a certificate from a URL
provided by the client via the "client_certificate_url" extension
[RFC6066].
unrecognized_name
Sent by servers when no server exists identified by the name
provided by the client via the "server_name" extension [RFC6066].
bad_certificate_status_response
Sent by clients when an invalid or unacceptable OCSP response is
provided by the server via the "status_request" extension
[RFC6066]. This alert is always fatal.
bad_certificate_hash_value
Sent by servers when a retrieved object does not have the correct
hash provided by the client via the "client_certificate_url"
extension [RFC6066]. This alert is always fatal.
unknown_psk_identity
Sent by servers when a PSK cipher suite is selected but no
acceptable PSK identity is provided by the client. Sending this
alert is OPTIONAL; servers MAY instead choose to send a
"decrypt_error" alert to merely indicate an invalid PSK identity.
New Alert values are assigned by IANA as described in Section 11.
6.2. Handshake Protocol Overview
The cryptographic parameters of the session state are produced by the
TLS Handshake Protocol, which operates on top of the TLS record
layer. When a TLS client and server first start communicating, they
agree on a protocol version, select cryptographic algorithms,
optionally authenticate each other, and establish shared secret
keying material.
TLS supports three basic key exchange modes:
- Diffie-Hellman (of both the finite field and elliptic curve
varieties).
- A pre-shared symmetric key (PSK)
- A combination of a symmetric key and Diffie-Hellman
Which mode is used depends on the negotiated cipher suite.
Conceptually, the handshake establishes three secrets which are used
to derive all the keys.
Figure 1 below shows the basic full TLS handshake.
Client Server
Key ^ ClientHello
Exch | + key_share*
v + pre_shared_key* -------->
ServerHello ^ Key
+ key_share* | Exch
+ pre_shared_key* v
{EncryptedExtensions} ^ Server
{CertificateRequest*} v Params
{Certificate*} ^
{CertificateVerify*} | Auth
{Finished} v
<-------- [Application Data*]
^ {Certificate*}
Auth | {CertificateVerify*}
v {Finished} -------->
[Application Data] <-------> [Application Data]
+ Indicates extensions sent in the
previously noted message.
* Indicates optional or situation-dependent
messages that are not always sent.
{} Indicates messages protected using keys
derived from handshake_traffic_secret.
[] Indicates messages protected using keys
derived from traffic_secret_N
Figure 1: Message flow for full TLS Handshake
The handshake can be thought of as having three phases, indicated in
the diagram above.
Key Exchange: establish shared keying material and select the
cryptographic parameters. Everything after this phase is encrypted.
Server Parameters: establish other handshake parameters (whether the
client is authenticated, application layer protocol support, etc.)
Authentication: authenticate the server (and optionally the client)
and provide key confirmation and handshake integrity.
In the Key Exchange phase, the client sends the ClientHello
(Section 6.3.1.1) message, which contains a random nonce
(ClientHello.random), its offered protocol version, cipher suite, and
extensions, and in general either one or more Diffie-Hellman key
shares (in the "key_share" extension Section 6.3.2.4), one or more
pre-shared key labels (in the "pre_shared_key" extension
Section 6.3.2.5), or both.
The server processes the ClientHello and determines the appropriate
cryptographic parameters for the connection. It then responds with
its own ServerHello which indicates the negotiated connection
parameters. [Section 6.3.1.2]. The combination of the ClientHello
and the ServerHello determines the values of ES and SS, as described
above. If either a pure (EC)DHE or (EC)DHE-PSK cipher suite is in
use, then the ServerHello will contain a "key_share" extension with
the server's ephemeral Diffie-Hellman share which MUST be in the same
group. If a pure PSK or an (EC)DHE-PSK cipher suite is negotiated,
then the ServerHello will contain a "pre_shared_key" extension
indicating which if the client's offered PSKs was selected.
The server then sends two messages to establish the Server
Parameters:
EncryptedExtensions responses to any extensions which are not
required in order to determine the cryptographic parameters.
[Section 6.3.3.1]
CertificateRequest if certificate-based client authentication is
desired, the desired parameters for that certificate. This
message will be omitted if client authentication is not desired.
Finally, the client and server exchange Authentication messages. TLS
uses the same set of messages every time that authentication is
needed. Specifically:
Certificate
the certificate of the endpoint. This message is omitted if the
server is not authenticating with a certificate (i.e., with PSK or
(EC)DHE-PSK cipher suites). Note that if raw public keys
[RFC7250] or the cached information extension
[I-D.ietf-tls-cached-info] are in use, then this message will not
contain a certificate but rather some other value corresponding to
the server's long-term key. [Section 6.3.4.1]
CertificateVerify
a signature over the entire handshake using the public key in the
Certificate message. This message is omitted if the server is not
authenticating via a certificate (i.e., with PSK or (EC)DHE-PSK
cipher suites). [Section 6.3.4.2]
Finished
a MAC over the entire handshake. This message provides key
confirmation, binds the endpoint's identity to the exchanged keys,
and in PSK mode also authenticates the handshake.
[Section 6.3.4.3]
Upon receiving the server's messages, the client responds with its
Authentication messages, namely Certificate and CertificateVerify (if
requested), and Finished.
At this point, the handshake is complete, and the client and server
may exchange application layer data. Application data MUST NOT be
sent prior to sending the Finished message. Note that while the
server may send application data prior to receiving the client's
Authentication messages, any data sent at that point is, of course,
being sent to an unauthenticated peer.
[[TODO: Move this elsewhere? Note that higher layers should not be
overly reliant on whether TLS always negotiates the strongest
possible connection between two endpoints. There are a number of
ways in which a man-in-the-middle attacker can attempt to make two
entities drop down to the least secure method they support (i.e.,
perform a downgrade attack). The TLS protocol has been designed to
minimize this risk, but there are still attacks available: for
example, an attacker could block access to the port a secure service
runs on, or attempt to get the peers to negotiate an unauthenticated
connection. The fundamental rule is that higher levels must be
cognizant of what their security requirements are and never transmit
information over a channel less secure than what they require. The
TLS protocol is secure in that any cipher suite offers its promised
level of security: if you negotiate AES-GCM [GCM] with a 255-bit
ECDHE key exchange with a host whose certificate chain you have
verified, you can expect that to be reasonably "secure" against
algorithmic attacks, at least in the year 2015.]]
6.2.1. Incorrect DHE Share
If the client has not provided an appropriate "key_share" extension
(e.g. it includes only DHE or ECDHE groups unacceptable or
unsupported by the server), the server corrects the mismatch with a
HelloRetryRequest and the client will need to restart the handshake
with an appropriate "key_share" extension, as shown in Figure 2. If
no common cryptographic parameters can be negotiated, the server will
send a "handshake_failure" or "insufficient_security" fatal alert
(see Section 6.1).
Client Server
ClientHello
+ key_share -------->
<-------- HelloRetryRequest
ClientHello
+ key_share -------->
ServerHello
+ key_share
{EncryptedExtensions}
{CertificateRequest*}
{Certificate*}
{CertificateVerify*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
[Application Data] <-------> [Application Data]
Figure 2: Message flow for a full handshake with mismatched
parameters
Note: the handshake transcript includes the initial ClientHello/
HelloRetryRequest exchange. It is not reset with the new
ClientHello.
TLS also allows several optimized variants of the basic handshake, as
described below.
6.2.2. Resumption and Pre-Shared Key (PSK)
Although TLS PSKs can be established out of band, PSKs can also be
established in a previous session and then reused ("session
resumption"). Once a handshake has completed, the server can send
the client a PSK identity which corresponds to a key derived from the
initial handshake (See Section 6.3.5.1). The client can then use
that PSK identity in future handshakes to negotiate use of the PSK;
if the server accepts it, then the security context of the original
connection is tied to the new connection. In TLS 1.2 and below, this
functionality was provided by "session resumption" and "session
tickets" [RFC5077]. Both mechanisms are obsoleted in TLS 1.3.
PSK cipher suites can either use PSK in combination with an (EC)DHE
exchange in order to provide forward secrecy in combination with
shared keys, or can use PSKs alone, at the cost of losing forward
secrecy.
Figure 3 shows a pair of handshakes in which the first establishes a
PSK and the second uses it:
Client Server
Initial Handshake:
ClientHello
+ key_share -------->
ServerHello
+ key_share
{EncryptedExtensions}
{CertificateRequest*}
{Certificate*}
{CertificateVerify*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
<-------- [NewSessionTicket]
[Application Data] <-------> [Application Data]
Subsequent Handshake:
ClientHello
+ key_share
+ pre_shared_key -------->
ServerHello
+ pre_shared_key
+ key_share*
{EncryptedExtensions}
{Finished}
<-------- [Application Data*]
{Finished} -------->
[Application Data] <-------> [Application Data]
Figure 3: Message flow for resumption and PSK
As the server is authenticating via a PSK, it does not send a
Certificate or a CertificateVerify. When a client offers resumption
via PSK it SHOULD also supply a "key_share" extension to the server
as well; this allows server to decline resumption and fall back to a
full handshake. A "key_share" extension MUST also be sent if the
client is attempting to negotiate an (EC)DHE-PSK cipher suite.
6.2.3. Zero-RTT Data
When resuming via a PSK with an appropriate ticket (i.e., one with
the "allow_early_data" flag), clients can also send data on their
first flight ("early data"). This data is encrypted solely under
keys derived using the PSK as the static secret. As shown in
Figure 4, the Zero-RTT data is just added to the 1-RTT handshake in
the first flight, the rest of the handshake uses the same messages.
Client Server
ClientHello
+ early_data
+ key_share*
(EncryptedExtensions)
(Finished)
(Application Data*)
(end_of_early_data) -------->
ServerHello
+ early_data
+ key_share
{EncryptedExtensions}
{CertificateRequest*}
{Finished}
<-------- [Application Data*]
{Certificate*}
{CertificateVerify*}
{Finished} -------->
[Application Data] <-------> [Application Data]
* Indicates optional or situation-dependent
messages that are not always sent.
() Indicates messages protected using keys
derived from early_traffic_secret.
{} Indicates messages protected using keys
derived from handshake_traffic_secret.
[] Indicates messages protected using keys
derived from traffic_secret_N
Figure 4: Message flow for a zero round trip handshake
[[OPEN ISSUE: Should it be possible to combine 0-RTT with the server
authenticating via a signature https://github.com/tlswg/tls13-spec/
issues/443]]
IMPORTANT NOTE: The security properties for 0-RTT data (regardless of
the cipher suite) are weaker than those for other kinds of TLS data.
Specifically:
1. This data is not forward secret, because it is encrypted solely
with the PSK.
2. There are no guarantees of non-replay between connections.
Unless the server takes special measures outside those provided
by TLS (See Section 6.3.2.7.2), the server has no guarantee that
the same 0-RTT data was not transmitted on multiple 0-RTT
connections. This is especially relevant if the data is
authenticated either with TLS client authentication or inside the
application layer protocol. However, 0-RTT data cannot be
duplicated within a connection (i.e., the server will not process
the same data twice for the same connection) and an attacker will
not be able to make 0-RTT data appear to be 1-RTT data (because
it is protected with different keys.)
The contents and significance of each message will be presented in
detail in the following sections.
6.3. Handshake Protocol
The TLS Handshake Protocol is one of the defined higher-level clients The handshake protocol is used to negotiate the secure attributes of
of the TLS Record Protocol. This protocol is used to negotiate the a session. Handshake messages are supplied to the TLS record layer,
secure attributes of a session. Handshake messages are supplied to where they are encapsulated within one or more TLSPlaintext or
the TLS record layer, where they are encapsulated within one or more TLSCiphertext structures, which are processed and transmitted as
TLSPlaintext or TLSCiphertext structures, which are processed and specified by the current active session state.
transmitted as specified by the current active session state.
enum { enum {
client_hello(1), client_hello(1),
server_hello(2), server_hello(2),
session_ticket(4), new_session_ticket(4),
hello_retry_request(6), hello_retry_request(6),
encrypted_extensions(8), encrypted_extensions(8),
certificate(11), certificate(11),
certificate_request(13), certificate_request(13),
certificate_verify(15), certificate_verify(15),
finished(20), finished(20),
key_update(24), key_update(24),
(255) (255)
} HandshakeType; } HandshakeType;
skipping to change at page 40, line 31 skipping to change at page 24, line 31
uint24 length; /* bytes in message */ uint24 length; /* bytes in message */
select (HandshakeType) { select (HandshakeType) {
case client_hello: ClientHello; case client_hello: ClientHello;
case server_hello: ServerHello; case server_hello: ServerHello;
case hello_retry_request: HelloRetryRequest; case hello_retry_request: HelloRetryRequest;
case encrypted_extensions: EncryptedExtensions; case encrypted_extensions: EncryptedExtensions;
case certificate_request: CertificateRequest; case certificate_request: CertificateRequest;
case certificate: Certificate; case certificate: Certificate;
case certificate_verify: CertificateVerify; case certificate_verify: CertificateVerify;
case finished: Finished; case finished: Finished;
case session_ticket: NewSessionTicket; case new_session_ticket: NewSessionTicket;
case key_update: KeyUpdate; case key_update: KeyUpdate;
} body; } body;
} Handshake; } Handshake;
The TLS Handshake Protocol messages are presented below in the order Protocol messages MUST be sent in the order defined below (and shown
they MUST be sent; sending handshake messages in an unexpected order in the diagrams in Section 2). Sending handshake messages in an
results in an "unexpected_message" fatal error. Unneeded handshake unexpected order results in an "unexpected_message" fatal error.
messages can be omitted, however. Unneeded handshake messages are omitted, however.
New handshake message types are assigned by IANA as described in New handshake message types are assigned by IANA as described in
Section 11. Section 10.
6.3.1. Key Exchange Messages 4.1. Key Exchange Messages
The key exchange messages are used to exchange security capabilities The key exchange messages are used to exchange security capabilities
between the client and server and to establish the traffic keys used between the client and server and to establish the traffic keys used
to protect the handshake and the data. to protect the handshake and data.
6.3.1.1. Client Hello 4.1.1. Client Hello
When this message will be sent: When this message will be sent:
When a client first connects to a server, it is required to send When a client first connects to a server, it is required to send
the ClientHello as its first message. The client will also send a the ClientHello as its first message. The client will also send a
ClientHello when the server has responded to its ClientHello with ClientHello when the server has responded to its ClientHello with
a ServerHello that selects cryptographic parameters that don't a ServerHello that selects cryptographic parameters that don't
match the client's "key_share" extension. In that case, the match the client's "key_share" extension. In that case, the
client MUST send the same ClientHello (without modification) client MUST send the same ClientHello (without modification)
except including a new KeyShareEntry as the lowest priority share except:
(i.e., appended to the list of shares in the "key_share"
extension). If a server receives a ClientHello at any other time,
it MUST send a fatal "unexpected_message" alert and close the
connection.
Structure of this message: - Including a new KeyShareEntry as the lowest priority share (i.e.,
appended to the list of shares in the "key_share" extension).
The ClientHello message includes a random structure, which is used - Removing the EarlyDataIndication Section 4.2.6 extension if one
later in the protocol. was present. Early data is not permitted after HelloRetryRequest.
The cipher suite list, passed from the client to the server in the If a server receives a ClientHello at any other time, it MUST send a
ClientHello message, contains the combinations of cryptographic fatal "unexpected_message" alert and close the connection.
algorithms supported by the client in order of the client's
preference (favorite choice first). Each cipher suite defines a key Structure of this message:
exchange algorithm, a record protection algorithm (including secret
key length) and a hash to be used with HKDF. The server will select struct {
a cipher suite or, if no acceptable choices are presented, return a uint8 major;
"handshake_failure" alert and close the connection. If the list uint8 minor;
contains cipher suites the server does not recognize, support, or } ProtocolVersion;
wish to use, the server MUST ignore those cipher suites, and process
the remaining ones as usual.
struct { struct {
opaque random_bytes[32]; opaque random_bytes[32];
} Random; } Random;
uint8 CipherSuite[2]; /* Cryptographic suite selector */ uint8 CipherSuite[2]; /* Cryptographic suite selector */
struct { struct {
ProtocolVersion client_version = { 3, 4 }; /* TLS v1.3 */ ProtocolVersion client_version = { 3, 4 }; /* TLS v1.3 */
Random random; Random random;
skipping to change at page 42, line 15 skipping to change at page 26, line 9
TLS allows extensions to follow the compression_methods field in an TLS allows extensions to follow the compression_methods field in an
extensions block. The presence of extensions can be detected by extensions block. The presence of extensions can be detected by
determining whether there are bytes following the compression_methods determining whether there are bytes following the compression_methods
at the end of the ClientHello. Note that this method of detecting at the end of the ClientHello. Note that this method of detecting
optional data differs from the normal TLS method of having a optional data differs from the normal TLS method of having a
variable-length field, but it is used for compatibility with TLS variable-length field, but it is used for compatibility with TLS
before extensions were defined. As of TLS 1.3, all clients and before extensions were defined. As of TLS 1.3, all clients and
servers will send at least one extension (at least "key_share" or servers will send at least one extension (at least "key_share" or
"pre_shared_key"). "pre_shared_key").
client_version client_version The latest (highest valued) version of the TLS
The version of the TLS protocol by which the client wishes to protocol offered by the client. This SHOULD be the same as the
communicate during this session. This SHOULD be the latest latest version supported. For this version of the specification,
(highest valued) version supported by the client. For this the version will be { 3, 4 }. (See Appendix C for details about
version of the specification, the version will be { 3, 4 }. (See backward compatibility.)
Appendix C for details about backward compatibility.)
random random 32 bytes generated by a secure random number generator. See
32 bytes generated by a secure random number generator. See
Appendix B for additional information. Appendix B for additional information.
legacy_session_id legacy_session_id Versions of TLS before TLS 1.3 supported a session
Versions of TLS before TLS 1.3 supported a session resumption resumption feature which has been merged with Pre-Shared Keys in
feature which has been merged with Pre-Shared Keys in this version this version (see Section 2.2). This field MUST be ignored by a
(see Section 6.2.2). This field MUST be ignored by a server server negotiating TLS 1.3 and SHOULD be set as a zero length
negotiating TLS 1.3 and SHOULD be set as a zero length vector vector (i.e., a single zero byte length field) by clients which do
(i.e., a single zero byte length field) by clients which do not not have a cached session ID set by a pre-TLS 1.3 server.
have a cached session ID set by a pre-TLS 1.3 server.
cipher_suites cipher_suites This is a list of the cryptographic options supported
This is a list of the cryptographic options supported by the by the client, with the client's first preference first. Each
client, with the client's first preference first. Values are cipher suite defines a key exchange algorithm, a record protection
defined in Appendix A.4. algorithm (including secret key length) and a hash to be used with
HKDF. The server will select a cipher suite or, if no acceptable
choices are presented, return a "handshake_failure" alert and
close the connection. If the list contains cipher suites the
server does not recognize, support, or wish to use, the server
MUST ignore those cipher suites, and process the remaining ones as
usual. Values are defined in Appendix A.4.
legacy_compression_methods legacy_compression_methods Versions of TLS before 1.3 supported
Versions of TLS before 1.3 supported compression and the list of compression with the list of supported compression methods being
compression methods was supplied in this field. For any TLS 1.3 sent in this field. For every TLS 1.3 ClientHello, this vector
ClientHello, this vector MUST contain exactly one byte set to MUST contain exactly one byte set to zero, which corresponds to
zero, which corresponds to the "null" compression method in prior the "null" compression method in prior versions of TLS. If a TLS
versions of TLS. If a TLS 1.3 ClientHello is received with any 1.3 ClientHello is received with any other value in this field,
other value in this field, the server MUST generate a fatal the server MUST generate a fatal "illegal_parameter" alert. Note
"illegal_parameter" alert. Note that TLS 1.3 servers might that TLS 1.3 servers might receive TLS 1.2 or prior ClientHellos
receive TLS 1.2 or prior ClientHellos which contain other which contain other compression methods and MUST follow the
compression methods and MUST follow the procedures for the procedures for the appropriate prior version of TLS.
appropriate prior version of TLS.
extensions extensions Clients request extended functionality from servers by
Clients request extended functionality from servers by sending sending data in the extensions field. The actual "Extension"
data in the extensions field. The actual "Extension" format is format is defined in Section 4.2.
defined in Section 6.3.2.
In the event that a client requests additional functionality using In the event that a client requests additional functionality using
extensions, and this functionality is not supplied by the server, the extensions, and this functionality is not supplied by the server, the
client MAY abort the handshake. Note: TLS 1.3 ClientHello messages client MAY abort the handshake. Note that TLS 1.3 ClientHello
MUST always contain extensions, and a TLS 1.3 server MUST respond to messages MUST always contain extensions, and a TLS 1.3 server MUST
any TLS 1.3 ClientHello without extensions with a fatal respond to any TLS 1.3 ClientHello without extensions with a fatal
"decode_error" alert. TLS 1.3 servers may receive TLS 1.2 "decode_error" alert. TLS 1.3 servers may receive TLS 1.2
ClientHello messages without extensions. If negotiating TLS 1.2, a ClientHello messages without extensions. If negotiating TLS 1.2, a
server MUST check that the amount of data in the message precisely server MUST check that the amount of data in the message precisely
matches one of these formats; if not, then it MUST send a fatal matches one of these formats; if not, then it MUST send a fatal
"decode_error" alert. "decode_error" alert.
After sending the ClientHello message, the client waits for a After sending the ClientHello message, the client waits for a
ServerHello or HelloRetryRequest message. ServerHello or HelloRetryRequest message.
6.3.1.2. Server Hello 4.1.2. Server Hello
When this message will be sent: When this message will be sent:
The server will send this message in response to a ClientHello The server will send this message in response to a ClientHello
message when it was able to find an acceptable set of algorithms message when it was able to find an acceptable set of algorithms
and the client's "key_share" extension was acceptable. If the and the client's "key_share" extension was acceptable. If the
client proposed groups are not acceptable by the server, it will client proposed groups are not acceptable by the server, it will
respond with a "handshake_failure" fatal alert. respond with a "handshake_failure" fatal alert.
Structure of this message: Structure of this message:
struct { struct {
ProtocolVersion server_version; ProtocolVersion server_version;
Random random; Random random;
CipherSuite cipher_suite; CipherSuite cipher_suite;
Extension extensions<0..2^16-1>; Extension extensions<0..2^16-1>;
} ServerHello; } ServerHello;
In prior versions of TLS, the extensions field could be omitted server_version This field contains the version of TLS negotiated for
entirely if not needed, similar to ClientHello. As of TLS 1.3, all this session. Servers MUST select the lower of the highest
clients and servers will send at least one extension (at least supported server version and the version offered by the client in
"key_share" or "pre_shared_key"). the ClientHello. In particular, servers MUST accept ClientHello
messages with versions higher than those supported and negotiate
server_version the highest mutually supported version. For this version of the
This field will contain the lower of that suggested by the client specification, the version is { 3, 4 }. (See Appendix C for
in the ClientHello and the highest supported by the server. For details about backward compatibility.)
this version of the specification, the version is { 3, 4 }. (See
Appendix C for details about backward compatibility.)
random random This structure is generated by the server and MUST be
This structure is generated by the server and MUST be generated generated independently of the ClientHello.random.
independently of the ClientHello.random.
cipher_suite cipher_suite The single cipher suite selected by the server from the
The single cipher suite selected by the server from the list in list in ClientHello.cipher_suites. For resumed sessions, this
ClientHello.cipher_suites. For resumed sessions, this field is field is the value from the state of the session being resumed.
the value from the state of the session being resumed. [[TODO: [[TODO: interaction with PSK.]]
interaction with PSK.]]
extensions extensions A list of extensions. Note that only extensions offered
A list of extensions. Note that only extensions offered by the by the client can appear in the server's list. In TLS 1.3, as
client can appear in the server's list. In TLS 1.3 as opposed to opposed to previous versions of TLS, the server's extensions are
previous versions of TLS, the server's extensions are split split between the ServerHello and the EncryptedExtensions
between the ServerHello and the EncryptedExtensions Section 4.2.8 message. The ServerHello MUST only include
Section 6.3.3.1 message. The ServerHello MUST only include
extensions which are required to establish the cryptographic extensions which are required to establish the cryptographic
context. Currently the only such extensions are "key_share", context. Currently the only such extensions are "key_share",
"pre_shared_key", and "early_data". Clients MUST check the "pre_shared_key", and "early_data". Clients MUST check the
ServerHello for the presence of any forbidden extensions and if ServerHello for the presence of any forbidden extensions and if
any are found MUST terminate the handshake with a any are found MUST terminate the handshake with a
"illegal_parameter" alert. "illegal_parameter" alert. In prior versions of TLS, the
extensions field could be omitted entirely if not needed, similar
to ClientHello. As of TLS 1.3, all clients and servers will send
at least one extension (at least "key_share" or "pre_shared_key").
TLS 1.3 server implementations which respond to a ClientHello with a TLS 1.3 has a downgrade protection mechanism embedded in the server's
client_version indicating TLS 1.2 or below MUST set the first eight random value. TLS 1.3 server implementations which respond to a
bytes of their Random value to the bytes: ClientHello with a client_version indicating TLS 1.2 or below MUST
set the last eight bytes of their Random value to the bytes:
44 4F 57 4E 47 52 44 01 44 4F 57 4E 47 52 44 01
TLS 1.2 server implementations which respond to a ClientHello with a TLS 1.2 server implementations which respond to a ClientHello with a
client_version indicating TLS 1.1 or below SHOULD set the first eight client_version indicating TLS 1.1 or below SHOULD set the last eight
bytes of their Random value to the bytes: bytes of their Random value to the bytes:
44 4F 57 4E 47 52 44 00 44 4F 57 4E 47 52 44 00
TLS 1.3 clients receiving a TLS 1.2 or below ServerHello MUST check TLS 1.3 clients receiving a TLS 1.2 or below ServerHello MUST check
that the top eight octets are not equal to either of these values. that the last eight octets are not equal to either of these values.
TLS 1.2 clients SHOULD also perform this check if the ServerHello TLS 1.2 clients SHOULD also perform this check if the ServerHello
indicates TLS 1.1 or below. If a match is found, the client MUST indicates TLS 1.1 or below. If a match is found, the client MUST
abort the handshake with a fatal "illegal_parameter" alert. This abort the handshake with a fatal "illegal_parameter" alert. This
mechanism provides limited protection against downgrade attacks over mechanism provides limited protection against downgrade attacks over
and above that provided by the Finished exchange: because the and above that provided by the Finished exchange: because the
ServerKeyExchange includes a signature over both random values, it is ServerKeyExchange includes a signature over both random values, it is
not possible for an active attacker to modify the randoms without not possible for an active attacker to modify the randoms without
detection as long as ephemeral ciphers are used. It does not provide detection as long as ephemeral ciphers are used. It does not provide
downgrade protection when static RSA is used. downgrade protection when static RSA is used.
Note: This is an update to TLS 1.2 so in practice many TLS 1.2 Note: This is an update to TLS 1.2 so in practice many TLS 1.2
clients and servers will not behave as specified above. clients and servers will not behave as specified above.
Note: Versions of TLS prior to TLS 1.3 used the top 32 bits of the 4.1.3. Hello Retry Request
Random value to encode the time since the UNIX epoch. The sentinel
value above was selected to avoid conflicting with any valid TLS 1.2
Random value and to have a low (2^{-64}) probability of colliding
with randomly selected Random values.
6.3.1.3. Hello Retry Request
When this message will be sent: When this message will be sent:
Servers send this message in response to a ClientHello message Servers send this message in response to a ClientHello message if
when it was able to find an acceptable set of algorithms and they were able to find an acceptable set of algorithms and groups
groups that are mutually supported, but the client's KeyShare did that are mutually supported, but the client's KeyShare did not
not contain an acceptable offer. If it cannot find such a match, contain an acceptable offer. If it cannot find such a match, it
it will respond with a fatal "handshake_failure" alert. will respond with a fatal "handshake_failure" alert.
Structure of this message: Structure of this message:
struct { struct {
ProtocolVersion server_version; ProtocolVersion server_version;
CipherSuite cipher_suite; CipherSuite cipher_suite;
NamedGroup selected_group; NamedGroup selected_group;
Extension extensions<0..2^16-1>; Extension extensions<0..2^16-1>;
} HelloRetryRequest; } HelloRetryRequest;
selected_group selected_group The mutually supported group the server intends to
The mutually supported group the server intends to negotiate and negotiate and is requesting a retried ClientHello/KeyShare for.
is requesting a retried ClientHello/KeyShare for.
The server_version, cipher_suite, and extensions fields have the same The server_version, cipher_suite, and extensions fields have the same
meanings as their corresponding values in the ServerHello. The meanings as their corresponding values in the ServerHello. [[NOTE:
server SHOULD send only the extensions necessary for the client to cipher_suite may disappear. https://github.com/tlswg/tls13-spec/
generate a correct ClientHello pair. As with ServerHello, a issues/528]] The server SHOULD send only the extensions necessary for
HelloRetryRequest MUST NOT contain any extensions that were not first the client to generate a correct ClientHello pair (currently no such
offered by the client in its ClientHello. extensions exist). As with ServerHello, a HelloRetryRequest MUST NOT
contain any extensions that were not first offered by the client in
its ClientHello.
Upon receipt of a HelloRetryRequest, the client MUST first verify Upon receipt of a HelloRetryRequest, the client MUST first verify
that the selected_group field corresponds to a group which was that the selected_group field corresponds to a group which was
provided in the "supported_groups" extension in the original provided in the "supported_groups" extension in the original
ClientHello. It MUST then verify that the selected_group field does ClientHello. It MUST then verify that the selected_group field does
not correspond to a group which was provided in the "key_share" not correspond to a group which was provided in the "key_share"
extension in the original ClientHello. If either of these checks extension in the original ClientHello. If either of these checks
fails, then the client MUST abort the handshake with a fatal fails, then the client MUST abort the handshake with a fatal
"handshake_failure" alert. Clients SHOULD also abort with "handshake_failure" alert. Clients SHOULD also abort with
"handshake_failure" in response to any second HelloRetryRequest which "handshake_failure" in response to any second HelloRetryRequest which
skipping to change at page 46, line 18 skipping to change at page 30, line 11
KeyShare extension to the server. The client MUST append a new KeyShare extension to the server. The client MUST append a new
KeyShareEntry for the group indicated in the selected_group field to KeyShareEntry for the group indicated in the selected_group field to
the groups in its original KeyShare. the groups in its original KeyShare.
Upon re-sending the ClientHello and receiving the server's Upon re-sending the ClientHello and receiving the server's
ServerHello/KeyShare, the client MUST verify that the selected ServerHello/KeyShare, the client MUST verify that the selected
CipherSuite and NamedGroup match that supplied in the CipherSuite and NamedGroup match that supplied in the
HelloRetryRequest. If either of these values differ, the client MUST HelloRetryRequest. If either of these values differ, the client MUST
abort the connection with a fatal "handshake_failure" alert. abort the connection with a fatal "handshake_failure" alert.
6.3.2. Hello Extensions 4.2. Hello Extensions
The extension format is: The extension format is:
struct { struct {
ExtensionType extension_type; ExtensionType extension_type;
opaque extension_data<0..2^16-1>; opaque extension_data<0..2^16-1>;
} Extension; } Extension;
enum { enum {
supported_groups(10), supported_groups(10),
signature_algorithms(13), signature_algorithms(13),
key_share(40), key_share(40),
pre_shared_key(41), pre_shared_key(41),
early_data(42), early_data(42),
ticket_age(43), cookie(44),
cookie (44),
(65535) (65535)
} ExtensionType; } ExtensionType;
Here: Here:
- "extension_type" identifies the particular extension type. - "extension_type" identifies the particular extension type.
- "extension_data" contains information specific to the particular - "extension_data" contains information specific to the particular
extension type. extension type.
The initial set of extensions is defined in [RFC6066]. The list of The initial set of extensions is defined in [RFC6066]. The list of
extension types is maintained by IANA as described in Section 11. extension types is maintained by IANA as described in Section 10.
An extension type MUST NOT appear in the ServerHello or An extension type MUST NOT appear in the ServerHello or
HelloRetryRequest unless the same extension type appeared in the HelloRetryRequest unless the same extension type appeared in the
corresponding ClientHello. If a client receives an extension type in corresponding ClientHello. If a client receives an extension type in
ServerHello or HelloRetryRequest that it did not request in the ServerHello or HelloRetryRequest that it did not request in the
associated ClientHello, it MUST abort the handshake with an associated ClientHello, it MUST abort the handshake with an
"unsupported_extension" fatal alert. "unsupported_extension" fatal alert.
Nonetheless, "server-oriented" extensions may be provided in the Nonetheless, "server-oriented" extensions may be provided within this
future within this framework. Such an extension (say, of type x) framework. Such an extension (say, of type x) would require the
would require the client to first send an extension of type x in a client to first send an extension of type x in a ClientHello with
ClientHello with empty extension_data to indicate that it supports empty extension_data to indicate that it supports the extension type.
the extension type. In this case, the client is offering the In this case, the client is offering the capability to understand the
capability to understand the extension type, and the server is taking extension type, and the server is taking the client up on its offer.
the client up on its offer.
When multiple extensions of different types are present in the When multiple extensions of different types are present in the
ClientHello or ServerHello messages, the extensions MAY appear in any ClientHello or ServerHello messages, the extensions MAY appear in any
order. There MUST NOT be more than one extension of the same type. order. There MUST NOT be more than one extension of the same type.
Finally, note that extensions can be sent both when starting a new Finally, note that extensions can be sent both when starting a new
session and when requesting session resumption or 0-RTT mode. session and when in resumption-PSK mode. A client that requests
Indeed, a client that requests session resumption does not in general session resumption does not in general know whether the server will
know whether the server will accept this request, and therefore it accept this request, and therefore it SHOULD send the same extensions
SHOULD send the same extensions as it would send if it were not as it would send normally.
attempting resumption.
In general, the specification of each extension type needs to In general, the specification of each extension type needs to
describe the effect of the extension both during full handshake and describe the effect of the extension both during full handshake and
session resumption. Most current TLS extensions are relevant only session resumption. Most current TLS extensions are relevant only
when a session is initiated: when an older session is resumed, the when a session is initiated: when an older session is resumed, the
server does not process these extensions in ClientHello, and does not server does not process these extensions in ClientHello, and does not
include them in ServerHello. However, some extensions may specify include them in ServerHello. However, some extensions may specify
different behavior during session resumption. [[TODO: update this different behavior during session resumption. [[TODO: update this
and the previous paragraph to cover PSK-based resumption.]] and the previous paragraph to cover PSK-based resumption.]]
skipping to change at page 48, line 10 skipping to change at page 31, line 47
manipulation of handshake messages. This principle should be manipulation of handshake messages. This principle should be
followed regardless of whether the feature is believed to cause a followed regardless of whether the feature is believed to cause a
security problem. Often the fact that the extension fields are security problem. Often the fact that the extension fields are
included in the inputs to the Finished message hashes will be included in the inputs to the Finished message hashes will be
sufficient, but extreme care is needed when the extension changes sufficient, but extreme care is needed when the extension changes
the meaning of messages sent in the handshake phase. Designers the meaning of messages sent in the handshake phase. Designers
and implementors should be aware of the fact that until the and implementors should be aware of the fact that until the
handshake has been authenticated, active attackers can modify handshake has been authenticated, active attackers can modify
messages and insert, remove, or replace extensions. messages and insert, remove, or replace extensions.
- It would be technically possible to use extensions to change major 4.2.1. Cookie
aspects of the design of TLS; for example, the design of cipher
suite negotiation. This is not recommended; it would be more
appropriate to define a new version of TLS -- particularly since
the TLS handshake algorithms have specific protection against
version rollback attacks based on the version number, and the
possibility of version rollback should be a significant
consideration in any major design change.
6.3.2.1. Cookie
struct { struct {
opaque cookie<0..255>; opaque cookie<0..2^16-1>;
} Cookie; } Cookie;
Cookies serve two primary purposes: Cookies serve two primary purposes:
- Allowing the server to force the client to demonstrate - Allowing the server to force the client to demonstrate
reachability at their apparent network address (thus providing a reachability at their apparent network address (thus providing a
measure of DoS protection). This is primarily useful for non- measure of DoS protection). This is primarily useful for non-
connection-oriented transports (see [RFC6347] for an example of connection-oriented transports (see [RFC6347] for an example of
this). this).
skipping to change at page 48, line 45 skipping to change at page 32, line 25
server does this by pickling that post-ClientHello hash state into server does this by pickling that post-ClientHello hash state into
the cookie (protected with some suitable integrity algorithm). the cookie (protected with some suitable integrity algorithm).
When sending a HelloRetryRequest, the server MAY provide a "cookie" When sending a HelloRetryRequest, the server MAY provide a "cookie"
extension to the client (this is an exception to the usual rule that extension to the client (this is an exception to the usual rule that
the only extensions that may be sent are those that appear in the the only extensions that may be sent are those that appear in the
ClientHello). When sending the new ClientHello, the client MUST echo ClientHello). When sending the new ClientHello, the client MUST echo
the value of the extension. Clients MUST NOT use cookies in the value of the extension. Clients MUST NOT use cookies in
subsequent connections. subsequent connections.
6.3.2.2. Signature Algorithms 4.2.2. Signature Algorithms
The client uses the "signature_algorithms" extension to indicate to The client uses the "signature_algorithms" extension to indicate to
the server which signature algorithms may be used in digital the server which signature algorithms may be used in digital
signatures. signatures.
Clients which offer one or more cipher suites which use certificate Clients which offer one or more cipher suites which use certificate
authentication (i.e., any non-PSK cipher suite) MUST send the authentication (i.e., any non-PSK cipher suite) MUST send the
"signature_algorithms" extension. If this extension is not provided "signature_algorithms" extension. If this extension is not provided
and no alternative cipher suite is available, the server MUST close and no alternative cipher suite is available, the server MUST close
the connection with a fatal "missing_extension" alert. (see the connection with a fatal "missing_extension" alert. (see
Section 8.2) Section 8.2)
The "extension_data" field of this extension contains a The "extension_data" field of this extension contains a
"supported_signature_algorithms" value: "supported_signature_algorithms" value:
enum { enum {
/* RSASSA-PKCS-v1_5 algorithms */ /* RSASSA-PKCS1-v1_5 algorithms */
rsa_pkcs1_sha1 (0x0201), rsa_pkcs1_sha1 (0x0201),
rsa_pkcs1_sha256 (0x0401), rsa_pkcs1_sha256 (0x0401),
rsa_pkcs1_sha384 (0x0501), rsa_pkcs1_sha384 (0x0501),
rsa_pkcs1_sha512 (0x0601), rsa_pkcs1_sha512 (0x0601),
/* ECDSA algorithms */ /* ECDSA algorithms */
ecdsa_secp256r1_sha256 (0x0403), ecdsa_secp256r1_sha256 (0x0403),
ecdsa_secp384r1_sha384 (0x0503), ecdsa_secp384r1_sha384 (0x0503),
ecdsa_secp521r1_sha512 (0x0603), ecdsa_secp521r1_sha512 (0x0603),
skipping to change at page 49, line 43 skipping to change at page 33, line 33
ed25519 (0x0703), ed25519 (0x0703),
ed448 (0x0704), ed448 (0x0704),
/* Reserved Code Points */ /* Reserved Code Points */
private_use (0xFE00..0xFFFF), private_use (0xFE00..0xFFFF),
(0xFFFF) (0xFFFF)
} SignatureScheme; } SignatureScheme;
SignatureScheme supported_signature_algorithms<2..2^16-2>; SignatureScheme supported_signature_algorithms<2..2^16-2>;
Note: This production is named "SignatureScheme" because there is Note: This enum is named "SignatureScheme" because there is already a
already a SignatureAlgorithm type in TLS 1.2. We use the term "SignatureAlgorithm" type in TLS 1.2, which this replaces. We use
"signature algorithm" throughout the text. the term "signature algorithm" throughout the text.
Each SignatureScheme value lists a single signature algorithm that Each SignatureScheme value lists a single signature algorithm that
the client is willing to verify. The values are indicated in the client is willing to verify. The values are indicated in
descending order of preference. Note that a signature algorithm descending order of preference. Note that a signature algorithm
takes as input an arbitrary-length message, rather than a digest. takes as input an arbitrary-length message, rather than a digest.
Algorithms which traditionally act on a digest should be defined in Algorithms which traditionally act on a digest should be defined in
TLS to first hash the input with a specified hash function and then TLS to first hash the input with a specified hash function and then
proceed as usual. The code point groups listed above have the proceed as usual. The code point groups listed above have the
following meanings: following meanings:
RSASSA-PKCS-v1_5 algorithms RSASSA-PKCS1-v1_5 algorithms Indicates a signature algorithm using
Indicates a signature algorithm using RSASSA-PKCS1-v1_5 [RFC3447] RSASSA-PKCS1-v1_5 [RFC3447] with the corresponding hash algorithm
with the corresponding hash algorithm as defined in [SHS]. These as defined in [SHS]. These values refer solely to signatures
values refer solely to signatures which appear in certificates which appear in certificates (see Section 4.3.1.1) and are not
(see Section 6.3.4.1.1) and are not defined for use in signed TLS defined for use in signed TLS handshake messages.
handshake messages (see Section 4.8.1).
ECDSA algorithms ECDSA algorithms Indicates a signature algorithm using ECDSA
Indicates a signature algorithm using ECDSA [ECDSA], the [ECDSA], the corresponding curve as defined in ANSI X9.62 [X962]
corresponding curve as defined in ANSI X9.62 [X962] and FIPS 186-4 and FIPS 186-4 [DSS], and the corresponding hash algorithm as
[DSS], and the corresponding hash algorithm as defined in [SHS]. defined in [SHS]. The signature is represented as a DER-encoded
The signature is represented as a DER-encoded [X690] ECDSA-Sig- [X690] ECDSA-Sig-Value structure.
Value structure.
RSASSA-PSS algorithms RSASSA-PSS algorithms Indicates a signature algorithm using RSASSA-
Indicates a signature algorithm using RSASSA-PSS [RFC3447] with PSS [RFC3447] with MGF1. The digest used in the mask generation
MGF1. The digest used in the mask generation function and the function and the digest being signed are both the corresponding
digest being signed are both the corresponding hash algorithm as hash algorithm as defined in [SHS]. When used in signed TLS
defined in [SHS]. When used in signed TLS handshake messages (see handshake messages, the length of the salt MUST be equal to the
Section 4.8.1), the length of the salt MUST be equal to the length length of the digest output.
of the digest output.
EdDSA algorithms EdDSA algorithms Indicates a signature algorithm using EdDSA as
Indicates a signature algorithm using EdDSA as defined in defined in [I-D.irtf-cfrg-eddsa] or its successors. Note that
[I-D.irtf-cfrg-eddsa] or its successors. Note that these these correspond to the "PureEdDSA" algorithms and not the
correspond to the "PureEdDSA" algorithms and not the "prehash" "prehash" variants.
variants.
The semantics of this extension are somewhat complicated because the The semantics of this extension are somewhat complicated because the
cipher suite adds additional constraints on signature algorithms. cipher suite adds additional constraints on signature algorithms.
Section 6.3.4.1.1 describes the appropriate rules. Section 4.3.1.1 describes the appropriate rules.
rsa_pkcs1_sha1 and dsa_sha1 SHOULD NOT be offered. Clients offering rsa_pkcs1_sha1, dsa_sha1, and ecdsa_sha1 SHOULD NOT be offered.
these values for backwards compatibility MUST list them as the lowest Clients offering these values for backwards compatibility MUST list
priority (listed after all other algorithms in the them as the lowest priority (listed after all other algorithms in the
supported_signature_algorithms vector). TLS 1.3 servers MUST NOT supported_signature_algorithms vector). TLS 1.3 servers MUST NOT
offer a SHA-1 signed certificate unless no valid certificate chain offer a SHA-1 signed certificate unless no valid certificate chain
can be produced without it (see Section 6.3.4.1.1). can be produced without it (see Section 4.3.1.1).
The signatures on certificates that are self-signed or certificates The signatures on certificates that are self-signed or certificates
that are trust anchors are not validated since they begin a that are trust anchors are not validated since they begin a
certification path (see [RFC5280], Section 3.2). A certificate that certification path (see [RFC5280], Section 3.2). A certificate that
begins a certification path MAY use a signature algorithm that is not begins a certification path MAY use a signature algorithm that is not
advertised as being supported in the "signature_algorithms" advertised as being supported in the "signature_algorithms"
extension. extension.
Note that TLS 1.2 defines this extension differently. TLS 1.3 Note that TLS 1.2 defines this extension differently. TLS 1.3
implementations willing to negotiate TLS 1.2 MUST behave in implementations willing to negotiate TLS 1.2 MUST behave in
skipping to change at page 51, line 24 skipping to change at page 35, line 11
been allocated to align with TLS 1.2's encoding. Some legacy been allocated to align with TLS 1.2's encoding. Some legacy
pairs are left unallocated. These algorithms are deprecated as of pairs are left unallocated. These algorithms are deprecated as of
TLS 1.3. They MUST NOT be offered or negotiated by any TLS 1.3. They MUST NOT be offered or negotiated by any
implementation. In particular, MD5 [SLOTH] and SHA-224 MUST NOT implementation. In particular, MD5 [SLOTH] and SHA-224 MUST NOT
be used. be used.
- ecdsa_secp256r1_sha256, etc., align with TLS 1.2's ECDSA hash/ - ecdsa_secp256r1_sha256, etc., align with TLS 1.2's ECDSA hash/
signature pairs. However, the old semantics did not constrain the signature pairs. However, the old semantics did not constrain the
signing curve. signing curve.
6.3.2.3. Negotiated Groups 4.2.3. Negotiated Groups
When sent by the client, the "supported_groups" extension indicates When sent by the client, the "supported_groups" extension indicates
the named groups which the client supports, ordered from most the named groups which the client supports for key exchange, ordered
preferred to least preferred. from most preferred to least preferred.
Note: In versions of TLS prior to TLS 1.3, this extension was named Note: In versions of TLS prior to TLS 1.3, this extension was named
"elliptic_curves" and only contained elliptic curve groups. See "elliptic_curves" and only contained elliptic curve groups. See
[RFC4492] and [I-D.ietf-tls-negotiated-ff-dhe]. This extension was [RFC4492] and [I-D.ietf-tls-negotiated-ff-dhe]. This extension was
also used to negotiate ECDSA curves. Signature algorithms are now also used to negotiate ECDSA curves. Signature algorithms are now
negotiated independently (see Section 6.3.2.2). negotiated independently (see Section 4.2.2).
Clients which offer one or more (EC)DHE cipher suites MUST send at Clients which offer one or more (EC)DHE cipher suites MUST send at
least one supported NamedGroup value and servers MUST NOT negotiate least one supported NamedGroup value and servers MUST NOT negotiate
any of these cipher suites unless a supported value was provided. If any of these cipher suites unless a supported value was provided. If
this extension is not provided and no alternative cipher suite is this extension is not provided and no alternative cipher suite is
available, the server MUST close the connection with a fatal available, the server MUST close the connection with a fatal
"missing_extension" alert. (see Section 8.2) If the extension is "missing_extension" alert. (see Section 8.2) If the extension is
provided, but no compatible group is offered, the server MUST NOT provided, but no compatible group is offered, the server MUST NOT
negotiate a cipher suite of the relevant type. For instance, if a negotiate a cipher suite of the relevant type. For instance, if a
client supplies only ECDHE groups, the server MUST NOT negotiate client supplies only ECDHE groups, the server MUST NOT negotiate
skipping to change at page 52, line 24 skipping to change at page 36, line 24
/* Reserved Code Points */ /* Reserved Code Points */
ffdhe_private_use (0x01FC..0x01FF), ffdhe_private_use (0x01FC..0x01FF),
ecdhe_private_use (0xFE00..0xFEFF), ecdhe_private_use (0xFE00..0xFEFF),
(0xFFFF) (0xFFFF)
} NamedGroup; } NamedGroup;
struct { struct {
NamedGroup named_group_list<1..2^16-1>; NamedGroup named_group_list<1..2^16-1>;
} NamedGroupList; } NamedGroupList;
Elliptic Curve Groups (ECDHE) Elliptic Curve Groups (ECDHE) Indicates support of the corresponding
Indicates support of the corresponding named curve. Note that named curve. Note that some curves are also recommended in ANSI
some curves are also recommended in ANSI X9.62 [X962] and FIPS X9.62 [X962] and FIPS 186-4 [DSS]. Others are recommended in
186-4 [DSS]. Others are recommended in [RFC7748]. Values 0xFE00 [RFC7748]. Values 0xFE00 through 0xFEFF are reserved for private
through 0xFEFF are reserved for private use. use.
Finite Field Groups (DHE) Finite Field Groups (DHE) Indicates support of the corresponding
Indicates support of the corresponding finite field group, defined finite field group, defined in [I-D.ietf-tls-negotiated-ff-dhe].
in [I-D.ietf-tls-negotiated-ff-dhe]. Values 0x01FC through 0x01FF Values 0x01FC through 0x01FF are reserved for private use.
are reserved for private use.
Items in named_group_list are ordered according to the client's Items in named_group_list are ordered according to the client's
preferences (most preferred choice first). preferences (most preferred choice first).
As of TLS 1.3, servers are permitted to send the "supported_groups" As of TLS 1.3, servers are permitted to send the "supported_groups"
extension to the client. If the server has a group it prefers to the extension to the client. If the server has a group it prefers to the
ones in the "key_share" extension but is still willing to accept the ones in the "key_share" extension but is still willing to accept the
ClientHello, it SHOULD send "supported_groups" to update the client's ClientHello, it SHOULD send "supported_groups" to update the client's
view of its preferences. Clients MUST NOT act upon any information view of its preferences. Clients MUST NOT act upon any information
found in "supported_groups" prior to successful completion of the found in "supported_groups" prior to successful completion of the
handshake, but MAY use the information learned from a successfully handshake, but MAY use the information learned from a successfully
completed handshake to change what groups they offer to a server in completed handshake to change what groups they offer to a server in
subsequent connections. subsequent connections.
[[TODO: IANA Considerations.]] 4.2.4. Key Share
6.3.2.4. Key Share
The "key_share" extension contains the endpoint's cryptographic The "key_share" extension contains the endpoint's cryptographic
parameters for non-PSK key establishment methods (currently DHE or parameters for non-PSK key establishment methods (currently DHE or
ECDHE). ECDHE).
Clients which offer one or more (EC)DHE cipher suites MUST send this Clients which offer one or more (EC)DHE cipher suites MUST send this
extension and SHOULD send at least one supported KeyShareEntry value. extension and SHOULD send at least one supported KeyShareEntry value.
Servers MUST NOT negotiate any of these cipher suites unless a Servers MUST NOT negotiate any of these cipher suites unless a
supported value was provided. If this extension is not provided in a supported value was provided. If this extension is not provided in a
ServerHello or ClientHello, and the peer is offering (EC)DHE cipher ServerHello or ClientHello, and the peer is offering (EC)DHE cipher
suites, then the endpoint MUST close the connection with a fatal suites, then the endpoint MUST close the connection with a fatal
"missing_extension" alert. (see Section 8.2) Clients MAY send an "missing_extension" alert. (see Section 8.2) Clients MAY send an
empty client_shares vector in order to request group selection from empty client_shares vector in order to request group selection from
the server at the cost of an additional round trip. (see the server at the cost of an additional round trip. (see
Section 6.3.1.3) Section 4.1.3)
struct { struct {
NamedGroup group; NamedGroup group;
opaque key_exchange<1..2^16-1>; opaque key_exchange<1..2^16-1>;
} KeyShareEntry; } KeyShareEntry;
group group The named group for the key being exchanged. Finite Field
The named group for the key being exchanged. Finite Field Diffie- Diffie-Hellman [DH] parameters are described in Section 4.2.4.1;
Hellman [DH] parameters are described in Section 6.3.2.4.1;
Elliptic Curve Diffie-Hellman parameters are described in Elliptic Curve Diffie-Hellman parameters are described in
Section 6.3.2.4.2. Section 4.2.4.2.
key_exchange key_exchange Key exchange information. The contents of this field
Key exchange information. The contents of this field are are determined by the specified group and its corresponding
determined by the specified group and its corresponding
definition. Endpoints MUST NOT send empty or otherwise invalid definition. Endpoints MUST NOT send empty or otherwise invalid
key_exchange values for any reason. key_exchange values for any reason.
The "extension_data" field of this extension contains a "KeyShare" The "extension_data" field of this extension contains a "KeyShare"
value: value:
struct { struct {
select (role) { select (role) {
case client: case client:
KeyShareEntry client_shares<0..2^16-1>; KeyShareEntry client_shares<0..2^16-1>;
case server: case server:
KeyShareEntry server_share; KeyShareEntry server_share;
} }
} KeyShare; } KeyShare;
client_shares client_shares A list of offered KeyShareEntry values in descending
A list of offered KeyShareEntry values in descending order of order of client preference. This vector MAY be empty if the
client preference. This vector MAY be empty if the client is client is requesting a HelloRetryRequest. The ordering of values
requesting a HelloRetryRequest. The ordering of values here here SHOULD match that of the ordering of offered support in the
SHOULD match that of the ordering of offered support in the
"supported_groups" extension. "supported_groups" extension.
server_share server_share A single KeyShareEntry value for the negotiated cipher
A single KeyShareEntry value for the negotiated cipher suite. suite.
Servers offer exactly one KeyShareEntry value, which corresponds to
the key exchange used for the negotiated cipher suite.
Clients offer an arbitrary number of KeyShareEntry values, each Clients offer an arbitrary number of KeyShareEntry values, each
representing a single set of key exchange parameters. For instance, representing a single set of key exchange parameters. For instance,
a client might offer shares for several elliptic curves or multiple a client might offer shares for several elliptic curves or multiple
FFDHE groups. The key_exchange values for each KeyShareEntry MUST by FFDHE groups. The key_exchange values for each KeyShareEntry MUST by
generated independently. Clients MUST NOT offer multiple generated independently. Clients MUST NOT offer multiple
KeyShareEntry values for the same group and servers receiving KeyShareEntry values for the same group. Clients and MUST NOT offer
multiple KeyShareEntry values for the same group MUST abort the any KeyShareEntry values for groups not listed in the client's
connection with a fatal "illegal_parameter" alert. Clients and "supported_groups" extension.
servers MUST NOT offer or accept any KeyShareEntry values for groups
not listed in the client's "supported_groups" extension. Servers Servers offer exactly one KeyShareEntry value, which corresponds to
MUST NOT offer a KeyShareEntry value for a group not offered by the the key exchange used for the negotiated cipher suite. Servers MUST
client in its corresponding KeyShare. NOT offer a KeyShareEntry value for a group not offered by the client
in its corresponding KeyShare or "supported_groups" extension.
Implementations MAY check for violations of these rules and and MAY
abort the connection with a fatal "illegal_parameter" alert if one is
violated.
If the server selects an (EC)DHE cipher suite and no mutually If the server selects an (EC)DHE cipher suite and no mutually
supported group is available between the two endpoints' KeyShare supported group is available between the two endpoints' KeyShare
offers, yet there is a mutually supported group that can be found via offers, yet there is a mutually supported group that can be found via
the "supported_groups" extension, then the server MUST reply with a the "supported_groups" extension, then the server MUST reply with a
HelloRetryRequest. If there is no mutually supported group at all, HelloRetryRequest. If there is no mutually supported group at all,
the server MUST NOT negotiate an (EC)DHE cipher suite. the server MUST NOT negotiate an (EC)DHE cipher suite.
[[TODO: Recommendation about what the client offers. Presumably [[TODO: Recommendation about what the client offers. Presumably
which integer DH groups and which curves.]] which integer DH groups and which curves.]]
6.3.2.4.1. Diffie-Hellman Parameters 4.2.4.1. Diffie-Hellman Parameters
Diffie-Hellman [DH] parameters for both clients and servers are Diffie-Hellman [DH] parameters for both clients and servers are
encoded in the opaque key_exchange field of a KeyShareEntry in a encoded in the opaque key_exchange field of a KeyShareEntry in a
KeyShare structure. The opaque value contains the Diffie-Hellman KeyShare structure. The opaque value contains the Diffie-Hellman
public value (Y = g^X mod p), encoded as a big-endian integer, padded public value (Y = g^X mod p), encoded as a big-endian integer, padded
with zeros to the size of p. with zeros to the size of p in bytes.
Note: For a given Diffie-Hellman group, the padding results in all Note: For a given Diffie-Hellman group, the padding results in all
public keys having the same length. public keys having the same length.
6.3.2.4.2. ECDHE Parameters Peers SHOULD validate each other's public key Y by ensuring that 1 <
Y < p-1. This check ensures that the remote peer is properly behaved
and isn't forcing the local system into a small subgroup.
4.2.4.2. ECDHE Parameters
ECDHE parameters for both clients and servers are encoded in the the ECDHE parameters for both clients and servers are encoded in the the
opaque key_exchange field of a KeyShareEntry in a KeyShare structure. opaque key_exchange field of a KeyShareEntry in a KeyShare structure.
For secp256r1, secp384r1 and secp521r1, the contents are the byte For secp256r1, secp384r1 and secp521r1, the contents are the byte
string representation of an elliptic curve public value following the string representation of an elliptic curve public value following the
conversion routine in Section 4.3.6 of ANSI X9.62 [X962]. conversion routine in Section 4.3.6 of ANSI X9.62 [X962].
Although X9.62 supports multiple point formats, any given curve MUST Although X9.62 supports multiple point formats, any given curve MUST
specify only a single point format. All curves currently specified specify only a single point format. All curves currently specified
skipping to change at page 55, line 27 skipping to change at page 39, line 22
(the format for all ECDH functions is considered uncompressed). (the format for all ECDH functions is considered uncompressed).
For x25519 and x448, the contents are the byte string inputs and For x25519 and x448, the contents are the byte string inputs and
outputs of the corresponding functions defined in [RFC7748], 32 bytes outputs of the corresponding functions defined in [RFC7748], 32 bytes
for x25519 and 56 bytes for x448. for x25519 and 56 bytes for x448.
Note: Versions of TLS prior to 1.3 permitted point negotiation; TLS Note: Versions of TLS prior to 1.3 permitted point negotiation; TLS
1.3 removes this feature in favor of a single point format for each 1.3 removes this feature in favor of a single point format for each
curve. curve.
6.3.2.5. Pre-Shared Key Extension 4.2.5. Pre-Shared Key Extension
The "pre_shared_key" extension is used to indicate the identity of The "pre_shared_key" extension is used to indicate the identity of
the pre-shared key to be used with a given handshake in association the pre-shared key to be used with a given handshake in association
with a PSK or (EC)DHE-PSK cipher suite (see [RFC4279] for with a PSK or (EC)DHE-PSK cipher suite (see [RFC4279] for
background). background).
Clients which offer one or more PSK cipher suites MUST send at least Clients which offer one or more PSK cipher suites MUST send at least
one supported psk_identity value and servers MUST NOT negotiate any one supported psk_identity value and servers MUST NOT negotiate any
of these cipher suites unless a supported value was provided. If of these cipher suites unless a supported value was provided. If
this extension is not provided and no alternative cipher suite is this extension is not provided and no alternative cipher suite is
skipping to change at page 56, line 17 skipping to change at page 39, line 51
struct { struct {
select (Role) { select (Role) {
case client: case client:
psk_identity identities<2..2^16-1>; psk_identity identities<2..2^16-1>;
case server: case server:
uint16 selected_identity; uint16 selected_identity;
} }
} PreSharedKeyExtension; } PreSharedKeyExtension;
identities identities A list of the identities (labels for keys) that the
A list of the identities (labels for keys) that the client is client is willing to negotiate with the server. If sent alongside
willing to negotiate with the server. the "early_data" extension (see Section 4.2.6), the first identity
is the one used for 0-RTT data.
selected_identity selected_identity The server's chosen identity expressed as a
The server's chosen identity expressed as a (0-based) index into (0-based) index into the identies in the client's list.
the identies in the client's list.
If no suitable identity is provided, the server MUST NOT negotiate a If no suitable identity is provided, the server MUST NOT negotiate a
PSK cipher suite and MAY respond with an "unknown_psk_identity" alert PSK cipher suite and MAY respond with an "unknown_psk_identity" alert
message. Sending this alert is OPTIONAL; servers MAY instead choose message. Sending this alert is OPTIONAL; servers MAY instead choose
to send a "decrypt_error" alert to merely indicate an invalid PSK to send a "decrypt_error" alert to merely indicate an invalid PSK
identity or instead negotiate use of a non-PSK cipher suite, if identity or instead negotiate use of a non-PSK cipher suite, if
available. available.
If the server selects a PSK cipher suite, it MUST send a If the server selects a PSK cipher suite, it MUST send a
"pre_shared_key" extension with the identity that it selected. The "pre_shared_key" extension with the identity that it selected. The
client MUST verify that the server's selected_identity is within the client MUST verify that the server's selected_identity is within the
range supplied by the client. If any other value is returned, the range supplied by the client. If the server supplies an "early_data"
client MUST generate a fatal "unknown_psk_identity" alert and close extension, the client MUST verify that the server selected the first
the connection. offered identity. If any other value is returned, the client MUST
generate a fatal "unknown_psk_identity" alert and close the
6.3.2.6. OCSP Status Extensions connection.
[RFC6066] and [RFC6961] provide extensions to negotiate the server
sending OCSP responses to the client. In TLS 1.2 and below, the
server sends an empty extension to indicate negotiation of this
extension and the OCSP information is carried in a CertificateStatus
message. In TLS 1.3, the server's OCSP information is carried in an
extension in EncryptedExtensions. Specifically: The body of the
"status_request" or "status_request_v2" extension from the server
MUST be a CertificateStatus structure as defined in [RFC6066] and
[RFC6961] respectively.
Note: this means that the certificate status appears prior to the Note that although 0-RTT data is encrypted with the first PSK
certificates it applies to. This is slightly anomalous but matches identity, the server MAY fall back to 1-RTT and select a different
the existing behavior for SignedCertificateTimestamps [RFC6962], and PSK identity if multiple identities are offered.
is more easily extensible in the handshake state machine.
6.3.2.7. Early Data Indication 4.2.6. Early Data Indication
When PSK resumption is used, the client can send application data in When PSK resumption is used, the client can send application data in
its first flight of messages. If the client opts to do so, it MUST its first flight of messages. If the client opts to do so, it MUST
supply an "early_data" extension as well as the "pre_shared_key" supply an "early_data" extension as well as the "pre_shared_key"
extension. extension.
The "extension_data" field of this extension contains an The "extension_data" field of this extension contains an
"EarlyDataIndication" value: "EarlyDataIndication" value:
struct { struct {
select (Role) { select (Role) {
case client: case client:
opaque context<0..255>; uint32 obfuscated_ticket_age;
case server: case server:
struct {}; struct {};
} }
} EarlyDataIndication; } EarlyDataIndication;
context obfuscated_ticket_age The time since the client learned about the
An optional context value that can be used for anti-replay (see server configuration that it is using, in milliseconds. This
below). value is added modulo 2^32 to with the "ticket_age_add" value that
was included with the ticket, see Section 4.4.1. This addition
prevents passive observers from correlating sessions unless
tickets are reused. Note: because ticket lifetimes are restricted
to a week, 32 bits is enough to represent any plausible age, even
in milliseconds.
All of the parameters for the 0-RTT data (symmetric cipher suite, A server MUST validate that the ticket_age is within a small
ALPN, etc.) MUST be those which were negotiated in the connection tolerance of the time since the ticket was issued (see
Section 4.2.6.2).
The parameters for the 0-RTT data (symmetric cipher suite, ALPN,
etc.) are the same as those which were negotiated in the connection
which established the PSK. The PSK used to encrypt the early data which established the PSK. The PSK used to encrypt the early data
MUST be the first PSK listed in the client's "pre_shared_key" MUST be the first PSK listed in the client's "pre_shared_key"
extension. extension.
0-RTT messages sent in the first flight have the same content types 0-RTT messages sent in the first flight have the same content types
as their corresponding messages sent in other flights (handshake, as their corresponding messages sent in other flights (handshake,
application_data, and alert respectively) but are protected under application_data, and alert respectively) but are protected under
different keys. After all the 0-RTT application data messages (if different keys. After all the 0-RTT application data messages (if
any) have been sent, a "end_of_early_data" alert of type "warning" is any) have been sent, an "end_of_early_data" alert of type "warning"
sent to indicate the end of the flight. 0-RTT MUST always be is sent to indicate the end of the flight. 0-RTT MUST always be
followed by an "end_of_early_data" alert. followed by an "end_of_early_data" alert.
A server which receives an "early_data" extension can behave in one A server which receives an "early_data" extension can behave in one
of two ways: of two ways:
- Ignore the extension and return no response. This indicates that - Ignore the extension and return no response. This indicates that
the server has ignored any early data and an ordinary 1-RTT the server has ignored any early data and an ordinary 1-RTT
handshake is required. handshake is required.
- Return an empty extension, indicating that it intends to process - Return an empty extension, indicating that it intends to process
the early data. It is not possible for the server to accept only the early data. It is not possible for the server to accept only
a subset of the early data messages. a subset of the early data messages.
[[OPEN ISSUE: are the rules below correct? https://github.com/tlswg/ In order to accept early data, the server server MUST have accepted a
tls13-spec/issues/451]] Prior to accepting the "early_data" PSK cipher suite and selected the the first key offered in the
extension, the server MUST validate that the session ticket client's "pre_shared_key" extension. In addition, it MUST verify
parameters are consistent with its current configuration. It MUST that the following values are consistent with those negotiated in the
also validate that the extensions negotiated in the previous connection during which the ticket was established.
connection are identical to those being negotiated in the
ServerHello, with the exception of the following extensions:
- The use of "signed_certificate_timestamp" [RFC6962] MUST be - The TLS version number, symmetric ciphersuite, and the hash for
identical but the server's SCT extension value may differ. HKDF.
- The "padding" extension [RFC7685] MUST be ignored for this - The selected ALPN [RFC7443] value, if any.
purpose.
- The values of "key_share", "pre_shared_key", and "early_data", - The server_name [RFC6066] value provided by the client, if any.
which MUST be as defined in this document.
In addition, it MUST validate that the ticket_age is within a small Future extensions MUST define their interaction with 0-RTT.
tolerance of the time since the ticket was issued (see
Section 6.3.2.7.2).
If any of these checks fail, the server MUST NOT respond with the If any of these checks fail, the server MUST NOT respond with the
extension and must discard all the remaining first flight data (thus extension and must discard all the remaining first flight data (thus
falling back to 1-RTT). If the client attempts a 0-RTT handshake but falling back to 1-RTT). If the client attempts a 0-RTT handshake but
the server rejects it, it will generally not have the 0-RTT record the server rejects it, it will generally not have the 0-RTT record
protection keys and must instead trial decrypt each record with the protection keys and must instead trial decrypt each record with the
1-RTT handshake keys until it finds one that decrypts properly, and 1-RTT handshake keys until it finds one that decrypts properly, and
then pick up the handshake from that point. then pick up the handshake from that point.
If the server chooses to accept the "early_data" extension, then it If the server chooses to accept the "early_data" extension, then it
MUST comply with the same error handling requirements specified for MUST comply with the same error handling requirements specified for
all records when processing early data records. Specifically, all records when processing early data records. Specifically,
decryption failure of any 0-RTT record following an accepted decryption failure of any 0-RTT record following an accepted
"early_data" extension MUST produce a fatal "bad_record_mac" alert as "early_data" extension MUST produce a fatal "bad_record_mac" alert as
per Section 5.2.2. Implementations SHOULD determine the security per Section 5.2.
parameters for the 1-RTT phase of the connection entirely before
processing the EncryptedExtensions and Finished, using those values
solely to determine whether to accept or reject 0-RTT data.
[[TODO: How does the client behave if the indication is rejected.]] If the server rejects the "early_data" extension, the client
application MAY opt to retransmit the data once the handshake has
been completed. TLS stacks SHOULD not do this automatically and
client applications MUST take care that the negotiated parameters are
consistent with those it expected. For example, if the ALPN value
has changed, it is likely unsafe to retransmit the original
application layer data.
6.3.2.7.1. Processing Order 4.2.6.1. Processing Order
Clients are permitted to "stream" 0-RTT data until they receive the Clients are permitted to "stream" 0-RTT data until they receive the
server's Finished, only then sending the "end_of_early_data" alert. server's Finished, only then sending the "end_of_early_data" alert.
In order to avoid deadlock, when accepting "early_data", servers MUST In order to avoid deadlock, when accepting "early_data", servers MUST
process the client's Finished and then immediately send the process the client's Finished and then immediately send the
ServerHello, rather than waiting for the client's "end_of_early_data" ServerHello, rather than waiting for the client's "end_of_early_data"
alert. alert.
6.3.2.7.2. Replay Properties 4.2.6.2. Replay Properties
As noted in Section 6.2.3, TLS provides only a limited inter- As noted in Section 2.3, TLS provides a limited mechanism for replay
connection mechanism for replay protection for data sent by the protection for data sent by the client in the first flight.
client in the first flight.
The "ticket_age" extension sent by the client SHOULD be used by The "obfuscated_ticket_age" parameter in the client's "early_data"
servers to limit the time over which the first flight might be extension SHOULD be used by servers to limit the time over which the
replayed. A server can store the time at which it sends a server first flight might be replayed. A server can store the time at which
configuration to a client, or encode the time in a ticket. Then, it sends a session ticket to the client, or encode the time in the
each time it receives an early_data extension, it can check to see if ticket. Then, each time it receives an "early_data" extension, it
the value used by the client matches its expectations. can subtract the base value and check to see if the value used by the
client matches its expectations.
The "ticket_age" value provided by the client will be shorter than The ticket age (the value with "ticket_age_add" subtracted) provided
the actual time elapsed on the server by a single round trip time. by the client will be shorter than the actual time elapsed on the
This difference is comprised of the delay in sending the server by a single round trip time. This difference is comprised of
NewSessionTicket message to the client, plus the time taken to send the delay in sending the NewSessionTicket message to the client, plus
the ClientHello to the server. For this reason, a server SHOULD the time taken to send the ClientHello to the server. For this
measure the round trip time prior to sending the NewSessionTicket reason, a server SHOULD measure the round trip time prior to sending
message and account for that in the value it saves. the NewSessionTicket message and account for that in the value it
saves.
To properly validate the ticket age, a server needs to save at least
two items:
- The time that the server generated the session ticket and the
estimated round trip time can be added together to form a baseline
time.
- The "ticket_age_add" parameter from the NewSessionTicket is needed
to recover the ticket age from the "obfuscated_ticket_age"
parameter.
There are several potential sources of error that make an exact There are several potential sources of error that make an exact
measurement of time difficult. Variations in client and server measurement of time difficult. Variations in client and server
clocks are likely to be minimal, outside of gross time corrections. clocks are likely to be minimal, outside of gross time corrections.
Network propagation delays are most likely causes of a mismatch in Network propagation delays are most likely causes of a mismatch in
legitimate values for elapsed time. Both the NewSessionTicket and legitimate values for elapsed time. Both the NewSessionTicket and
ClientHello messages might be retransmitted and therefore delayed, ClientHello messages might be retransmitted and therefore delayed,
which might be hidden by TCP. which might be hidden by TCP.
A small allowance for errors in clocks and variations in measurements A small allowance for errors in clocks and variations in measurements
is advisable. However, any allowance also increases the opportunity is advisable. However, any allowance also increases the opportunity
for replay. In this case, it is better to reject early data than to for replay. In this case, it is better to reject early data than to
risk greater exposure to replay attacks. risk greater exposure to replay attacks.
6.3.2.8. Ticket Age 4.2.7. OCSP Status Extensions
struct {
uint32 ticket_age;
} TicketAge;
When the client sends the "early_data" extension, it MUST also send a [RFC6066] and [RFC6961] provide extensions to negotiate the server
"ticket_age" extension in its EncryptedExtensions block. This value sending OCSP responses to the client. In TLS 1.2 and below, the
contains the time elapsed since the client learned about the server server sends an empty extension to indicate negotiation of this
configuration that it is using, in milliseconds. This value can be extension and the OCSP information is carried in a CertificateStatus
used by the server to limit the time over which early data can be message. In TLS 1.3, the server's OCSP information is carried in an
replayed. Note: because ticket lifetimes are restricted to a week, extension in EncryptedExtensions. Specifically: The body of the
32 bits is enough to represent any plausible age, even in "status_request" or "status_request_v2" extension from the server
milliseconds. MUST be a CertificateStatus structure as defined in [RFC6066] and
[RFC6961] respectively.
6.3.3. Server Parameters Note: This means that the certificate status appears prior to the
certificates it applies to. This is slightly anomalous but matches
the existing behavior for SignedCertificateTimestamps [RFC6962], and
is more easily extensible in the handshake state machine.
6.3.3.1. Encrypted Extensions 4.2.8. Encrypted Extensions
When this message will be sent: When this message will be sent:
In all handshakes, the server MUST send the EncryptedExtensions In all handshakes, the server MUST send the EncryptedExtensions
message immediately after the ServerHello message. This is the message immediately after the ServerHello message. This is the
first message that is encrypted under keys derived from first message that is encrypted under keys derived from
handshake_traffic_secret. If the client indicates "early_data" in handshake_traffic_secret.
its ClientHello, it MUST also send EncryptedExtensions immediately
following the ClientHello and immediately prior to the Finished.
Meaning of this message: Meaning of this message:
The EncryptedExtensions message contains any extensions which The EncryptedExtensions message contains any extensions which
should be protected, i.e., any which are not needed to establish should be protected, i.e., any which are not needed to establish
the cryptographic context. the cryptographic context.
The same extension types MUST NOT appear in both the ServerHello and The same extension types MUST NOT appear in both the ServerHello and
EncryptedExtensions. If the same extension appears in both EncryptedExtensions. If the same extension appears in both
locations, the client MUST rely only on the value in the locations, the client MUST rely only on the value in the
EncryptedExtensions block. All server-sent extensions other than EncryptedExtensions block. All server-sent extensions other than
those explicitly listed in Section 6.3.1.2 or designated in the IANA those explicitly listed in Section 4.1.2 or designated in the IANA
registry MUST only appear in EncryptedExtensions. Extensions which registry MUST only appear in EncryptedExtensions. Extensions which
are designated to appear in ServerHello MUST NOT appear in are designated to appear in ServerHello MUST NOT appear in
EncryptedExtensions. Clients MUST check EncryptedExtensions for the EncryptedExtensions. Clients MUST check EncryptedExtensions for the
presence of any forbidden extensions and if any are found MUST presence of any forbidden extensions and if any are found MUST
terminate the handshake with an "illegal_parameter" alert. terminate the handshake with an "illegal_parameter" alert.
The client's EncryptedExtensions apply only to the early data with
which they appear. Servers MUST NOT use them to negotiate the rest
of the handshake. Only those extensions explicitly designated as
being included in 0-RTT Encrypted Extensions in the IANA registry can
be sent in the client's EncryptedExtensions.
Structure of this message: Structure of this message:
struct { struct {
Extension extensions<0..2^16-1>; Extension extensions<0..2^16-1>;
} EncryptedExtensions; } EncryptedExtensions;
extensions extensions A list of extensions.
A list of extensions.
6.3.3.2. Certificate Request 4.2.9. Certificate Request
When this message will be sent: When this message will be sent:
A non-anonymous server can optionally request a certificate from A non-anonymous server can optionally request a certificate from
the client, if appropriate for the selected cipher suite. This the client, if appropriate for the selected cipher suite. This
message, if sent, will follow EncryptedExtensions. message, if sent, will follow EncryptedExtensions.
Structure of this message: Structure of this message:
opaque DistinguishedName<1..2^16-1>; opaque DistinguishedName<1..2^16-1>;
skipping to change at page 61, line 37 skipping to change at page 45, line 20
} CertificateExtension; } CertificateExtension;
struct { struct {
opaque certificate_request_context<0..2^8-1>; opaque certificate_request_context<0..2^8-1>;
SignatureScheme SignatureScheme
supported_signature_algorithms<2..2^16-2>; supported_signature_algorithms<2..2^16-2>;
DistinguishedName certificate_authorities<0..2^16-1>; DistinguishedName certificate_authorities<0..2^16-1>;
CertificateExtension certificate_extensions<0..2^16-1>; CertificateExtension certificate_extensions<0..2^16-1>;
} CertificateRequest; } CertificateRequest;
certificate_request_context certificate_request_context An opaque string which identifies the
An opaque string which identifies the certificate request and certificate request and which will be echoed in the client's
which will be echoed in the client's Certificate message. The Certificate message. The certificate_request_context MUST be
certificate_request_context MUST be unique within the scope of unique within the scope of this connection (thus preventing replay
this connection (thus preventing replay of client of client CertificateVerify messages).
CertificateVerify messages).
supported_signature_algorithms
A list of the signature algorithms that the server is able to
verify, listed in descending order of preference. Any
certificates provided by the client MUST be signed using a
signature algorithm found in supported_signature_algorithms.
certificate_authorities supported_signature_algorithms A list of the signature algorithms
A list of the distinguished names [X501] of acceptable that the server is able to verify, listed in descending order of
certificate_authorities, represented in DER-encoded [X690] format. preference. Any certificates provided by the client MUST be
signed using a signature algorithm found in
supported_signature_algorithms.
These distinguished names may specify a desired distinguished name certificate_authorities A list of the distinguished names [X501] of
for a root CA or for a subordinate CA; thus, this message can be acceptable certificate_authorities, represented in DER-encoded
used to describe known roots as well as a desired authorization [X690] format. These distinguished names may specify a desired
space. If the certificate_authorities list is empty, then the distinguished name for a root CA or for a subordinate CA; thus,
client MAY send any certificate that meets the rest of the this message can be used to describe known roots as well as a
selection criteria in the CertificateRequest, unless there is some desired authorization space. If the certificate_authorities list
external arrangement to the contrary. is empty, then the client MAY send any certificate that meets the
rest of the selection criteria in the CertificateRequest, unless
there is some external arrangement to the contrary.
certificate_extensions certificate_extensions A list of certificate extension OIDs
A list of certificate extension OIDs [RFC5280] with their allowed [RFC5280] with their allowed values, represented in DER-encoded
values, represented in DER-encoded [X690] format. Some [X690] format. Some certificate extension OIDs allow multiple
certificate extension OIDs allow multiple values (e.g. Extended values (e.g. Extended Key Usage). If the server has included a
Key Usage). If the server has included a non-empty non-empty certificate_extensions list, the client certificate MUST
certificate_extensions list, the client certificate MUST contain contain all of the specified extension OIDs that the client
all of the specified extension OIDs that the client recognizes. recognizes. For each extension OID recognized by the client, all
For each extension OID recognized by the client, all of the of the specified values MUST be present in the client certificate
specified values MUST be present in the client certificate (but (but the certificate MAY have other values as well). However, the
the certificate MAY have other values as well). However, the
client MUST ignore and skip any unrecognized certificate extension client MUST ignore and skip any unrecognized certificate extension
OIDs. If the client has ignored some of the required certificate OIDs. If the client has ignored some of the required certificate
extension OIDs, and supplied a certificate that does not satisfy extension OIDs, and supplied a certificate that does not satisfy
the request, the server MAY at its discretion either continue the the request, the server MAY at its discretion either continue the
session without client authentication, or terminate the session session without client authentication, or terminate the session
with a fatal unsupported_certificate alert. PKIX RFCs define a with a fatal unsupported_certificate alert. PKIX RFCs define a
variety of certificate extension OIDs and their corresponding variety of certificate extension OIDs and their corresponding
value types. Depending on the type, matching certificate value types. Depending on the type, matching certificate
extension values are not necessarily bitwise-equal. It is extension values are not necessarily bitwise-equal. It is
expected that TLS implementations will rely on their PKI libraries expected that TLS implementations will rely on their PKI libraries
skipping to change at page 63, line 5 skipping to change at page 46, line 32
also found in the Extended Key Usage certificate extension. also found in the Extended Key Usage certificate extension.
The special anyExtendedKeyUsage OID MUST NOT be used in the The special anyExtendedKeyUsage OID MUST NOT be used in the
request. request.
Separate specifications may define matching rules for other Separate specifications may define matching rules for other
certificate extensions. certificate extensions.
Note: It is a fatal "handshake_failure" alert for an anonymous server Note: It is a fatal "handshake_failure" alert for an anonymous server
to request client authentication. to request client authentication.
6.3.4. Authentication Messages 4.3. Authentication Messages
As discussed in Section 6.2, TLS uses a common set of messages for As discussed in Section 2, TLS uses a common set of messages for
authentication, key confirmation, and handshake integrity: authentication, key confirmation, and handshake integrity:
Certificate, CertificateVerify, and Finished. These messages are Certificate, CertificateVerify, and Finished. These messages are
always sent as the last messages in their handshake flight. The always sent as the last messages in their handshake flight. The
Certificate and CertificateVerify messages are only sent under Certificate and CertificateVerify messages are only sent under
certain circumstances, as defined below. The Finished message is certain circumstances, as defined below. The Finished message is
always sent as part of the Authentication block. always sent as part of the Authentication block.
The computations for the Authentication messages all uniformly take The computations for the Authentication messages all uniformly take
the following inputs: the following inputs:
- The certificate and signing key to be used. - The certificate and signing key to be used.
- A Handshake Context based on the hash of the handshake messages - A Handshake Context based on the hash of the handshake messages
- A base key to be used to compute a MAC key. - A base key to be used to compute a MAC key.
Based on these inputs, the messages then contain: Based on these inputs, the messages then contain:
Certificate Certificate The certificate to be used for authentication and any
The certificate to be used for authentication and any supporting supporting certificates in the chain. Note that certificate-based
certificates in the chain. Note that certificate-based client client authentication is not available in the 0-RTT case.
authentication is not available in the 0-RTT case.
CertificateVerify CertificateVerify A signature over the value Hash(Handshake Context
A signature over the value Hash(Handshake Context + Certificate) + + Certificate) + Hash(resumption_context) See Section 4.4.1 for
Hash(resumption_context) See Section 6.3.5.1 for the definition of the definition of resumption_context.
resumption_context.
Finished Finished A MAC over the value Hash(Handshake Context + Certificate +
A MAC over the value Hash(Handshake Context + Certificate +
CertificateVerify) + Hash(resumption_context) using a MAC key CertificateVerify) + Hash(resumption_context) using a MAC key
derived from the base key. derived from the base key.
Because the CertificateVerify signs the Handshake Context + Because the CertificateVerify signs the Handshake Context +
Certificate and the Finished MACs the Handshake Context + Certificate Certificate and the Finished MACs the Handshake Context + Certificate
+ CertificateVerify, this is mostly equivalent to keeping a running + CertificateVerify, this is mostly equivalent to keeping a running
hash of the handshake messages (exactly so in the pure 1-RTT cases). hash of the handshake messages (exactly so in the pure 1-RTT cases).
Note, however, that subsequent post-handshake authentications do not Note, however, that subsequent post-handshake authentications do not
include each other, just the messages through the end of the main include each other, just the messages through the end of the main
handshake. handshake.
skipping to change at page 64, line 25 skipping to change at page 47, line 48
| 1-RTT | ClientHello ... ServerFinished | handshake_traffic_s | | 1-RTT | ClientHello ... ServerFinished | handshake_traffic_s |
| (Client) | | ecret | | (Client) | | ecret |
| | | | | | | |
| Post- | ClientHello ... ClientFinished | traffic_secret_0 | | Post- | ClientHello ... ClientFinished | traffic_secret_0 |
| Handshake | + CertificateRequest | | | Handshake | + CertificateRequest | |
+------------+--------------------------------+---------------------+ +------------+--------------------------------+---------------------+
Note: The Handshake Context for the last three rows does not include Note: The Handshake Context for the last three rows does not include
any 0-RTT handshake messages, regardless of whether 0-RTT is used. any 0-RTT handshake messages, regardless of whether 0-RTT is used.
6.3.4.1. Certificate 4.3.1. Certificate
When this message will be sent: When this message will be sent:
The server MUST send a Certificate message whenever the agreed- The server MUST send a Certificate message whenever the agreed-
upon key exchange method uses certificates for authentication upon key exchange method uses certificates for authentication
(this includes all key exchange methods defined in this document (this includes all key exchange methods defined in this document
except PSK). except PSK).
The client MUST send a Certificate message if and only if server The client MUST send a Certificate message if and only if server
has requested client authentication via a CertificateRequest has requested client authentication via a CertificateRequest
message (Section 6.3.3.2). If the server requests client message (Section 4.2.9). If the server requests client
authentication but no suitable certificate is available, the authentication but no suitable certificate is available, the
client MUST send a Certificate message containing no certificates client MUST send a Certificate message containing no certificates
(i.e., with the "certificate_list" field having length 0). (i.e., with the "certificate_list" field having length 0).
Meaning of this message: Meaning of this message:
This message conveys the endpoint's certificate chain to the peer. This message conveys the endpoint's certificate chain to the peer.
The certificate MUST be appropriate for the negotiated cipher The certificate MUST be appropriate for the negotiated cipher
suite's key exchange algorithm and any negotiated extensions. suite's authentication algorithm and any negotiated extensions.
Structure of this message: Structure of this message:
opaque ASN1Cert<1..2^24-1>; opaque ASN1Cert<1..2^24-1>;
struct { struct {
opaque certificate_request_context<0..2^8-1>; opaque certificate_request_context<0..2^8-1>;
ASN1Cert certificate_list<0..2^24-1>; ASN1Cert certificate_list<0..2^24-1>;
} Certificate; } Certificate;
certificate_request_context: certificate_request_context If this message is in response to a
If this message is in response to a CertificateRequest, the value CertificateRequest, the value of certificate_request_context in
of certificate_request_context in that message. Otherwise, in the that message. Otherwise, in the case of server authentication
case of server authentication or client authentication in 0-RTT,
this field SHALL be zero length. this field SHALL be zero length.
certificate_list certificate_list This is a sequence (chain) of certificates. The
This is a sequence (chain) of certificates. The sender's sender's certificate MUST come first in the list. Each following
certificate MUST come first in the list. Each following
certificate SHOULD directly certify one preceding it. Because certificate SHOULD directly certify one preceding it. Because
certificate validation requires that trust anchors be distributed certificate validation requires that trust anchors be distributed
independently, a certificate that specifies a trust anchor MAY be independently, a certificate that specifies a trust anchor MAY be
omitted from the chain, provided that supported peers are known to omitted from the chain, provided that supported peers are known to
possess any omitted certificates. possess any omitted certificates.
Note: Prior to TLS 1.3, "certificate_list" ordering required each Note: Prior to TLS 1.3, "certificate_list" ordering required each
certificate to certify the one immediately preceding it, however some certificate to certify the one immediately preceding it, however some
implementations allowed some flexibility. Servers sometimes send implementations allowed some flexibility. Servers sometimes send
both a current and deprecated intermediate for transitional purposes, both a current and deprecated intermediate for transitional purposes,
skipping to change at page 65, line 42 skipping to change at page 49, line 12
nonetheless be validated properly. For maximum compatibility, all nonetheless be validated properly. For maximum compatibility, all
implementations SHOULD be prepared to handle potentially extraneous implementations SHOULD be prepared to handle potentially extraneous
certificates and arbitrary orderings from any TLS version, with the certificates and arbitrary orderings from any TLS version, with the
exception of the end-entity certificate which MUST be first. exception of the end-entity certificate which MUST be first.
The server's certificate list MUST always be non-empty. A client The server's certificate list MUST always be non-empty. A client
will send an empty certificate list if it does not have an will send an empty certificate list if it does not have an
appropriate certificate to send in response to the server's appropriate certificate to send in response to the server's
authentication request. authentication request.
6.3.4.1.1. Server Certificate Selection 4.3.1.1. Server Certificate Selection
The following rules apply to the certificates sent by the server: The following rules apply to the certificates sent by the server:
- The certificate type MUST be X.509v3 [RFC5280], unless explicitly - The certificate type MUST be X.509v3 [RFC5280], unless explicitly
negotiated otherwise (e.g., [RFC5081]). negotiated otherwise (e.g., [RFC5081]).
- The server's end-entity certificate's public key (and associated - The server's end-entity certificate's public key (and associated
restrictions) MUST be compatible with the selected key exchange restrictions) MUST be compatible with the selected authentication
algorithm. algorithm (currently RSA or ECDSA).
+----------------------+---------------------------+
| Key Exchange Alg. | Certificate Key Type |
+----------------------+---------------------------+
| DHE_RSA or ECDHE_RSA | RSA public key |
| | |
| ECDHE_ECDSA | ECDSA or EdDSA public key |
+----------------------+---------------------------+
- The certificate MUST allow the key to be used for signing (i.e., - The certificate MUST allow the key to be used for signing (i.e.,
the digitalSignature bit MUST be set if the Key Usage extension is the digitalSignature bit MUST be set if the Key Usage extension is
present) with a signature scheme indicated in the client's present) with a signature scheme indicated in the client's
"signature_algorithms" extension. "signature_algorithms" extension.
- The "server_name" and "trusted_ca_keys" extensions [RFC6066] are - The "server_name" and "trusted_ca_keys" extensions [RFC6066] are
used to guide certificate selection. As servers MAY require the used to guide certificate selection. As servers MAY require the
presence of the "server_name" extension, clients SHOULD send this presence of the "server_name" extension, clients SHOULD send this
extension. extension, when applicable.
All certificates provided by the server MUST be signed by a signature All certificates provided by the server MUST be signed by a signature
algorithm that appears in the "signature_algorithms" extension algorithm that appears in the "signature_algorithms" extension
provided by the client, if they are able to provide such a chain (see provided by the client, if they are able to provide such a chain (see
Section 6.3.2.2). Certificates that are self-signed or certificates Section 4.2.2). Certificates that are self-signed or certificates
that are expected to be trust anchors are not validated as part of that are expected to be trust anchors are not validated as part of
the chain and therefore MAY be signed with any algorithm. the chain and therefore MAY be signed with any algorithm.
If the server cannot produce a certificate chain that is signed only If the server cannot produce a certificate chain that is signed only
via the indicated supported algorithms, then it SHOULD continue the via the indicated supported algorithms, then it SHOULD continue the
handshake by sending the client a certificate chain of its choice handshake by sending the client a certificate chain of its choice
that may include algorithms that are not known to be supported by the that may include algorithms that are not known to be supported by the
client. This fallback chain MAY use the deprecated SHA-1 hash client. This fallback chain MAY use the deprecated SHA-1 hash
algorithm only if the "signature_algorithms" extension provided by algorithm only if the "signature_algorithms" extension provided by
the client permits it. If the client cannot construct an acceptable the client permits it. If the client cannot construct an acceptable
skipping to change at page 67, line 5 skipping to change at page 50, line 13
message and close the connection. message and close the connection.
If the server has multiple certificates, it chooses one of them based If the server has multiple certificates, it chooses one of them based
on the above-mentioned criteria (in addition to other criteria, such on the above-mentioned criteria (in addition to other criteria, such
as transport layer endpoint, local configuration and preferences). as transport layer endpoint, local configuration and preferences).
As cipher suites that specify new key exchange methods are specified As cipher suites that specify new key exchange methods are specified
for the TLS protocol, they will imply the certificate format and the for the TLS protocol, they will imply the certificate format and the
required encoded keying information. required encoded keying information.
6.3.4.1.2. Client Certificate Selection 4.3.1.2. Client Certificate Selection
The following rules apply to certificates sent by the client: The following rules apply to certificates sent by the client:
In particular: In particular:
- The certificate type MUST be X.509v3 [RFC5280], unless explicitly - The certificate type MUST be X.509v3 [RFC5280], unless explicitly
negotiated otherwise (e.g., [RFC5081]). negotiated otherwise (e.g., [RFC5081]).
- If the certificate_authorities list in the certificate request - If the certificate_authorities list in the certificate request
message was non-empty, one of the certificates in the certificate message was non-empty, one of the certificates in the certificate
chain SHOULD be issued by one of the listed CAs. chain SHOULD be issued by one of the listed CAs.
- The certificates MUST be signed using an acceptable hash/ - The certificates MUST be signed using an acceptable signature
signature algorithm pair, as described in Section 6.3.3.2. Note algorithm, as described in Section 4.2.9. Note that this relaxes
that this relaxes the constraints on certificate-signing the constraints on certificate-signing algorithms found in prior
algorithms found in prior versions of TLS. versions of TLS.
- If the certificate_extensions list in the certificate request - If the certificate_extensions list in the certificate request
message was non-empty, the end-entity certificate MUST match the message was non-empty, the end-entity certificate MUST match the
extension OIDs recognized by the client, as described in extension OIDs recognized by the client, as described in
Section 6.3.3.2. Section 4.2.9.
Note that, as with the server certificate, there are certificates Note that, as with the server certificate, there are certificates
that use algorithm combinations that cannot be currently used with that use algorithm combinations that cannot be currently used with
TLS. TLS.
6.3.4.1.3. Receiving a Certificate Message 4.3.1.3. Receiving a Certificate Message
In general, detailed certificate validation procedures are out of In general, detailed certificate validation procedures are out of
scope for TLS (see [RFC5280]). This section provides TLS-specific scope for TLS (see [RFC5280]). This section provides TLS-specific
requirements. requirements.
If the server supplies an empty Certificate message, the client MUST If the server supplies an empty Certificate message, the client MUST
terminate the handshake with a fatal "decode_error" alert. terminate the handshake with a fatal "decode_error" alert.
If the client does not send any certificates, the server MAY at its If the client does not send any certificates, the server MAY at its
discretion either continue the handshake without client discretion either continue the handshake without client
authentication, or respond with a fatal "handshake_failure" alert. authentication, or respond with a fatal "handshake_failure" alert.
Also, if some aspect of the certificate chain was unacceptable (e.g., Also, if some aspect of the certificate chain was unacceptable (e.g.,
it was not signed by a known, trusted CA), the server MAY at its it was not signed by a known, trusted CA), the server MAY at its
discretion either continue the handshake (considering the client discretion either continue the handshake (considering the client
unauthenticated) or send a fatal alert. unauthenticated) or send a fatal alert.
Any endpoint receiving any certificate signed using any signature Any endpoint receiving any certificate signed using any signature
algorithm using an MD5 hash MUST send a "bad_certificate" alert algorithm using an MD5 hash MUST send a "bad_certificate" alert
message and close the connection. message and close the connection. SHA-1 is deprecated and therefore
NOT RECOMMENDED. All endpoints are RECOMMENDED to transition to
SHA-1 is deprecated and therefore NOT RECOMMENDED. Endpoints that SHA-256 or better as soon as possible to maintain interoperability
reject certification paths due to use of a deprecated hash MUST send with implementations currently in the process of phasing out SHA-1
a fatal "bad_certificate" alert message before closing the
connection. All endpoints are RECOMMENDED to transition to SHA-256
or better as soon as possible to maintain interoperability with
implementations currently in the process of phasing out SHA-1
support. support.
Note that a certificate containing a key for one signature algorithm Note that a certificate containing a key for one signature algorithm
MAY be signed using a different signature algorithm (for instance, an MAY be signed using a different signature algorithm (for instance, an
RSA key signed with an ECDSA key). RSA key signed with an ECDSA key).
6.3.4.2. Certificate Verify Endpoints that reject certification paths due to use of a deprecated
hash MUST send a fatal "bad_certificate" alert message before closing
the connection.
4.3.2. Certificate Verify
When this message will be sent: When this message will be sent:
This message is used to provide explicit proof that an endpoint This message is used to provide explicit proof that an endpoint
possesses the private key corresponding to its certificate and possesses the private key corresponding to its certificate and
also provides integrity for the handshake up to this point. also provides integrity for the handshake up to this point.
Servers MUST send this message when using a cipher suite which is Servers MUST send this message when using a cipher suite which is
authenticated via a certificate. Clients MUST send this message authenticated via a certificate. Clients MUST send this message
whenever authenticating via a Certificate (i.e., when the whenever authenticating via a Certificate (i.e., when the
Certificate message is non-empty). When sent, this message MUST Certificate message is non-empty). When sent, this message MUST
appear immediately after the Certificate Message and immediately appear immediately after the Certificate Message and immediately
prior to the Finished message. prior to the Finished message.
Structure of this message: Structure of this message:
struct { struct {
digitally-signed struct { SignatureScheme algorithm;
opaque hashed_data[hash_length]; opaque signature<0..2^16-1>;
};
} CertificateVerify; } CertificateVerify;
Where hashed_data is the hash output described in Section 6.3.4, The algorithm field specifies the signature algorithm used (see
namely Hash(Handshake Context + Certificate) + Section 4.2.2 for the definition of this field). The signature is a
Hash(resumption_context). For concreteness, this means that the digital signature using that algorithm that covers the hash output
value that is signed is: described in Section 4.3 namely:
padding + context_string + 00 + hashed_data Hash(Handshake Context + Certificate) + Hash(resumption_context)
The context string for a server signature is "TLS 1.3, server In TLS 1.3, the digital signature process takes as input:
CertificateVerify" and for a client signature is "TLS 1.3, client
CertificateVerify". A hash of the handshake messages is signed
rather than the messages themselves because the digitally-signed
format requires padding and context bytes at the beginning of the
input. Thus, by signing a digest of the messages, an
implementation only needs to maintain a single running hash per
hash type for CertificateVerify, Finished and other messages.
If sent by a server, the signature algorithm MUST be one offered - A signing key
in the client's "signature_algorithms" extension unless no valid
certificate chain can be produced without unsupported algorithms
(see Section 6.3.2.2). Note that there is a possibility for
inconsistencies here. For instance, the client might offer
ECDHE_ECDSA key exchange but omit any ECDSA and EdDSA values from
its "signature_algorithms" extension. In order to negotiate
correctly, the server MUST check any candidate cipher suites
against the "signature_algorithms" extension before selecting
them. This is somewhat inelegant but is a compromise designed to
minimize changes to the original cipher suite design.
If sent by a client, the signature algorithm used in the signature - A context string
MUST be one of those present in the supported_signature_algorithms
field of the CertificateRequest message.
In addition, the signature algorithm MUST be compatible with the - The actual content to be signed
key in the sender's end-entity certificate. RSA signatures MUST
use an RSASSA-PSS algorithm, regardless of whether RSASSA-PKCS- The digital signature is then computed using the signing key over the
v1_5 algorithms appear in "signature_algorithms". SHA-1 MUST NOT concatenation of:
be used in any signatures in CertificateVerify. (Note that
rsa_pkcs1_sha1 and dsa_sha1, the only defined SHA-1 signature - 64 bytes of octet 32
algorithms, are undefined for CertificateVerify signatures.)
- The context string
- A single 0 byte which servers as the separator
- The content to be signed
This structure is intended to prevent an attack on previous versions
of previous versions of TLS in which the ServerKeyExchange format
meant that attackers could obtain a signature of a message with a
chosen, 32-byte prefix. The initial 64 byte pad clears that prefix.
The context string for a server signature is "TLS 1.3, server
CertificateVerify" and for a client signature is "TLS 1.3, client
CertificateVerify".
For example, if Hash(Handshake Context + Certificate) was 32 bytes of
01 and Hash(resumption_context) was 32 bytes of 02 (these lengths
would make sense for SHA-256, the input to the final signing process
for a server CertificateVerify would be:
2020202020202020202020202020202020202020202020202020202020202020
2020202020202020202020202020202020202020202020202020202020202020
544c5320312e332c207365727665722043657274696669636174655665726966
79
00
0101010101010101010101010101010101010101010101010101010101010101
0202020202020202020202020202020202020202020202020202020202020202
If sent by a server, the signature algorithm MUST be one offered in
the client's "signature_algorithms" extension unless no valid
certificate chain can be produced without unsupported algorithms (see
Section 4.2.2). Note that there is a possibility for inconsistencies
here. For instance, the client might offer an ECDHE_ECDSA cipher
suite but omit any ECDSA and EdDSA values from its
"signature_algorithms" extension. In order to negotiate correctly,
the server MUST check any candidate cipher suites against the
"signature_algorithms" extension before selecting them. This is
somewhat inelegant but is a compromise designed to minimize changes
to the original cipher suite design.
If sent by a client, the signature algorithm used in the signature
MUST be one of those present in the supported_signature_algorithms
field of the CertificateRequest message.
In addition, the signature algorithm MUST be compatible with the key
in the sender's end-entity certificate. RSA signatures MUST use an
RSASSA-PSS algorithm, regardless of whether RSASSA-PKCS1-v1_5
algorithms appear in "signature_algorithms". SHA-1 MUST NOT be used
in any signatures in CertificateVerify. All SHA-1 signature
algorithms in this specification are defined solely for use in legacy
certificates, and are not valid for CertificateVerify signatures.
Note: When used with non-certificate-based handshakes (e.g., PSK), Note: When used with non-certificate-based handshakes (e.g., PSK),
the client's signature does not cover the server's certificate the client's signature does not cover the server's certificate
directly, although it does cover the server's Finished message, which directly, although it does cover the server's Finished message, which
transitively includes the server's certificate when the PSK derives transitively includes the server's certificate when the PSK derives
from a certificate-authenticated handshake. [PSK-FINISHED] describes from a certificate-authenticated handshake. [PSK-FINISHED] describes
a concrete attack on this mode if the Finished is omitted from the a concrete attack on this mode if the Finished is omitted from the
signature. It is unsafe to use certificate-based client signature. It is unsafe to use certificate-based client
authentication when the client might potentially share the same PSK/ authentication when the client might potentially share the same PSK/
key-id pair with two different endpoints. In order to ensure this, key-id pair with two different endpoints. In order to ensure this,
implementations MUST NOT mix certificate-based client authentication implementations MUST NOT mix certificate-based client authentication
with pure PSK modes (i.e., those where the PSK was not derived from a with pure PSK modes (i.e., those where the PSK was not derived from a
previous non-PSK handshake). previous non-PSK handshake).
6.3.4.3. Finished 4.3.3. Finished
When this message will be sent: When this message will be sent:
The Finished message is the final message in the authentication The Finished message is the final message in the authentication
block. It is essential for providing authentication of the block. It is essential for providing authentication of the
handshake and of the computed keys. handshake and of the computed keys.
Meaning of this message: Meaning of this message:
Recipients of Finished messages MUST verify that the contents are Recipients of Finished messages MUST verify that the contents are
correct. Once a side has sent its Finished message and received correct. Once a side has sent its Finished message and received
and validated the Finished message from its peer, it may begin to and validated the Finished message from its peer, it may begin to
send and receive application data over the connection. send and receive application data over the connection.
The key used to compute the finished message is computed from the The key used to compute the finished message is computed from the
Base key defined in Section 6.3.4 using HKDF (see Section 7.1). Base key defined in Section 4.3 using HKDF (see Section 7.1).
Specifically: Specifically:
client_finished_key = client_finished_key =
HKDF-Expand-Label(BaseKey, "client finished", "", L) HKDF-Expand-Label(BaseKey, "client finished", "", Hash.Length)
server_finished_key = server_finished_key =
HKDF-Expand-Label(BaseKey, "server finished", "", L) HKDF-Expand-Label(BaseKey, "server finished", "", Hash.Length)
Structure of this message: Structure of this message:
struct { struct {
opaque verify_data[verify_data_length]; opaque verify_data[Hash.length];
} Finished; } Finished;
The verify_data value is computed as follows: The verify_data value is computed as follows:
verify_data = verify_data =
HMAC(finished_key, Hash( HMAC(finished_key, Hash(
Handshake Context + Certificate* + CertificateVerify* Handshake Context +
) + Hash(resumption_context) Certificate* +
) CertificateVerify*
) +
Hash(resumption_context)
)
* Only included if present. * Only included if present.
Where HMAC [RFC2104] uses the Hash algorithm for the handshake. As Where HMAC [RFC2104] uses the Hash algorithm for the handshake. As
noted above: the HMAC input can generally be implemented by a running noted above, the HMAC input can generally be implemented by a running
hash, i.e., just the handshake hash at this point. hash, i.e., just the handshake hash at this point.
In previous versions of TLS, the verify_data was always 12 octets In previous versions of TLS, the verify_data was always 12 octets
long. In the current version of TLS, it is the size of the HMAC long. In the current version of TLS, it is the size of the HMAC
output for the Hash used for the handshake. output for the Hash used for the handshake.
Note: Alerts and any other record types are not handshake messages Note: Alerts and any other record types are not handshake messages
and are not included in the hash computations. and are not included in the hash computations.
6.3.5. Post-Handshake Messages 4.4. Post-Handshake Messages
TLS also allows other messages to be sent after the main handshake. TLS also allows other messages to be sent after the main handshake.
These messages use a handshake content type and are encrypted under These messages use a handshake content type and are encrypted under
the application traffic key. the application traffic key.
6.3.5.1. New Session Ticket Message 4.4.1. New Session Ticket Message
At any time after the server has received the client Finished At any time after the server has received the client Finished
message, it MAY send a NewSessionTicket message. This message message, it MAY send a NewSessionTicket message. This message
creates a pre-shared key (PSK) binding between the ticket value and creates a pre-shared key (PSK) binding between the ticket value and
the following two values derived from the resumption master secret: the following two values derived from the resumption master secret:
resumption_psk = HKDF-Expand-Label(resumption_secret, resumption_psk = HKDF-Expand-Label(
"resumption psk", "", L) resumption_secret,
"resumption psk", "", Hash.Length)
resumption_context = HKDF-Expand-Label(resumption_secret, resumption_context = HKDF-Expand-Label(
"resumption context", "", L) resumption_secret,
"resumption context", "", Hash.Length)
The client MAY use this PSK for future handshakes by including the The client MAY use this PSK for future handshakes by including the
ticket value in the "pre_shared_key" extension in its ClientHello ticket value in the "pre_shared_key" extension in its ClientHello
(Section 6.3.2.5) and supplying a suitable PSK cipher suite. Servers (Section 4.2.5) and supplying a suitable PSK cipher suite. Servers
may send multiple tickets on a single connection, for instance after may send multiple tickets on a single connection, for instance after
post-handshake authentication. For handshakes that do not use a post-handshake authentication. For handshakes that do not use a
resumption_psk, the resumption_context is a string of L zeroes. resumption_psk, the resumption_context is a string of Hash.Length
zeroes.
enum { (65535) } TicketExtensionType; enum { (65535) } TicketExtensionType;
struct { struct {
TicketExtensionType extension_type; TicketExtensionType extension_type;
opaque extension_data<0..2^16-1>; opaque extension_data<1..2^16-1>;
} TicketExtension; } TicketExtension;
enum { enum {
allow_early_data(1) allow_early_data(1),
allow_dhe_resumption(2), allow_dhe_resumption(2),
allow_psk_resumption(4) allow_psk_resumption(4)
} TicketFlags; } TicketFlags;
struct { struct {
uint32 ticket_lifetime; uint32 ticket_lifetime;
uint32 flags; uint32 flags;
uint32 ticket_age_add;
TicketExtension extensions<2..2^16-2>; TicketExtension extensions<2..2^16-2>;
opaque ticket<0..2^16-1>; opaque ticket<0..2^16-1>;
} NewSessionTicket; } NewSessionTicket;
flags flags A 32-bit value indicating the ways in which this ticket may be
A 32-bit value indicating the ways in which this ticket may be used (as a bitwise OR of the flags values).
used (as an OR of the flags values).
ticket_lifetime ticket_lifetime Indicates the lifetime in seconds as a 32-bit
Indicates the lifetime in seconds as a 32-bit unsigned integer in unsigned integer in network byte order from the time of ticket
network byte order from the time of ticket issuance. Servers MUST issuance. Servers MUST NOT use any value more than 604800 seconds
NOT use any value more than 604800 seconds (7 days). The value of (7 days). The value of zero indicates that the ticket should be
zero indicates that the ticket should be discarded immediately. discarded immediately. Clients MUST NOT cache session tickets for
Clients MUST NOT cache session tickets for longer than 7 days, longer than 7 days, regardless of the ticket_lifetime. It MAY
regardless of the ticket_lifetime. It MAY delete the ticket delete the ticket earlier based on local policy. A server MAY
earlier based on local policy. A server MAY treat a ticket as treat a ticket as valid for a shorter period of time than what is
valid for a shorter period of time than what is stated in the stated in the ticket_lifetime.
ticket_lifetime.
ticket_extensions ticket_age_add A randomly generated 32-bit value that is used to
A placeholder for extensions in the ticket. Clients MUST ignore obscure the age of the ticket that the client includes in the
unrecognized extensions. "early_data" extension. The actual ticket age is added to this
value modulo 2^32 to obtain the value that is transmitted by the
client.
ticket ticket_extensions A placeholder for extensions in the ticket.
The value of the ticket to be used as the PSK identifier. The Clients MUST ignore unrecognized extensions.
ticket itself is an opaque label. It MAY either be a database
ticket The value of the ticket to be used as the PSK identifier.
The ticket itself is an opaque label. It MAY either be a database
lookup key or a self-encrypted and self-authenticated value. lookup key or a self-encrypted and self-authenticated value.
Section 4 of [RFC5077] describes a recommended ticket construction Section 4 of [RFC5077] describes a recommended ticket construction
mechanism. mechanism.
The meanings of the flags are as follows: The meanings of the flags are as follows:
allow_early_data allow_early_data When resuming with this ticket, the client MAY send
When resuming with this ticket, the client MAY send data in its data in its first flight (early data) encrypted under a key
first flight (early data) encrypted under a key derived from this derived from this PSK.
PSK.
allow_dhe_resumption allow_dhe_resumption This ticket MAY be used with (EC)DHE-PSK cipher
This ticket MAY be used with (EC)DHE-PSK cipher suite suite.
allow_psk_resumption allow_psk_resumption This ticket MAY be used with a pure PSK cipher
This ticket MAY be used with a pure PSK cipher suite. suite.
In all cases, the PSK or (EC)DHE-PSK cipher suites that the client In all cases, the PSK or (EC)DHE-PSK cipher suites that the client
offers/uses MUST have the same symmetric parameters (cipher/hash) as offers/uses MUST have the same symmetric parameters (cipher/hash) as
the cipher suite negotiated for this connection. If no flags are set the cipher suite negotiated for this connection. If no flags are set
that the client recognizes, it MUST ignore the ticket. that the client recognizes, it MUST ignore the ticket.
6.3.5.2. Post-Handshake Authentication 4.4.2. Post-Handshake Authentication
The server is permitted to request client authentication at any time The server is permitted to request client authentication at any time
after the handshake has completed by sending a CertificateRequest after the handshake has completed by sending a CertificateRequest
message. The client SHOULD respond with the appropriate message. The client SHOULD respond with the appropriate
Authentication messages. If the client chooses to authenticate, it Authentication messages. If the client chooses to authenticate, it
MUST send Certificate, CertificateVerify, and Finished. If it MUST send Certificate, CertificateVerify, and Finished. If it
declines, it MUST send a Certificate message containing no declines, it MUST send a Certificate message containing no
certificates followed by Finished. certificates followed by Finished.
Note: Because client authentication may require prompting the user, Note: Because client authentication may require prompting the user,
servers MUST be prepared for some delay, including receiving an servers MUST be prepared for some delay, including receiving an
arbitrary number of other messages between sending the arbitrary number of other messages between sending the
CertificateRequest and receiving a response. In addition, clients CertificateRequest and receiving a response. In addition, clients
which receive multiple CertificateRequests in close succession MAY which receive multiple CertificateRequests in close succession MAY
respond to them in a different order than they were received (the respond to them in a different order than they were received (the
certificate_request_context value allows the server to disambiguate certificate_request_context value allows the server to disambiguate
the responses). the responses).
6.3.5.3. Key and IV Update 4.4.3. Key and IV Update
struct {} KeyUpdate; struct {} KeyUpdate;
The KeyUpdate handshake message is used to indicate that the sender The KeyUpdate handshake message is used to indicate that the sender
is updating its sending cryptographic keys. This message can be sent is updating its sending cryptographic keys. This message can be sent
by the server after sending its first flight and the client after by the server after sending its first flight and the client after
sending its second flight. Implementations that receive a KeyUpdate sending its second flight. Implementations that receive a KeyUpdate
message prior to receiving a Finished message as part of the 1-RTT message prior to receiving a Finished message as part of the 1-RTT
handshake MUST generate a fatal "unexpected_message" alert. After handshake MUST generate a fatal "unexpected_message" alert. After
sending a KeyUpdate message, the sender SHALL send all its traffic sending a KeyUpdate message, the sender SHALL send all its traffic
using the next generation of keys, computed as described in using the next generation of keys, computed as described in
Section 7.2. Upon receiving a KeyUpdate, the receiver MUST update Section 7.2. Upon receiving a KeyUpdate, the receiver MUST update
skipping to change at page 73, line 46 skipping to change at page 57, line 38
number of messages between sending a KeyUpdate and receiving the number of messages between sending a KeyUpdate and receiving the
peer's KeyUpdate because those messages may already be in flight. peer's KeyUpdate because those messages may already be in flight.
Note that if implementations independently send their own KeyUpdates Note that if implementations independently send their own KeyUpdates
and they cross in flight, this only results in an update of one and they cross in flight, this only results in an update of one
generation; when each side receives the other side's update it just generation; when each side receives the other side's update it just
updates its receive keys and notes that the generations match and updates its receive keys and notes that the generations match and
thus no send update is needed. thus no send update is needed.
Note that the side which sends its KeyUpdate first needs to retain Note that the side which sends its KeyUpdate first needs to retain
the traffic keys (though not the traffic secret) for the previous its receive traffic keys (though not the traffic secret) for the
generation of keys until it receives the KeyUpdate from the other previous generation of keys until it receives the KeyUpdate from the
side. other side.
Both sender and receiver MUST encrypt their KeyUpdate messages with Both sender and receiver MUST encrypt their KeyUpdate messages with
the old keys. Additionally, both sides MUST enforce that a KeyUpdate the old keys. Additionally, both sides MUST enforce that a KeyUpdate
with the old key is received before accepting any messages encrypted with the old key is received before accepting any messages encrypted
with the new key. Failure to do so may allow message truncation with the new key. Failure to do so may allow message truncation
attacks. attacks.
5. Record Protocol
The TLS record protocol takes messages to be transmitted, fragments
the data into manageable blocks, protects the records, and transmits
the result. Received data is decrypted and verified, reassembled,
and then delivered to higher-level clients.
TLS records are typed, which allows multiple higher level protocols
to be multiplexed over the same record layer. This document
specifies three content types: handshake, application data, and
alert. Implementations MUST NOT send record types not defined in
this document unless negotiated by some extension. If a TLS
implementation receives an unexpected record type, it MUST send an
"unexpected_message" alert. New record content type values are
assigned by IANA in the TLS Content Type Registry as described in
Section 10.
Application data messages are carried by the record layer and are
fragmented and encrypted as described below. The messages are
treated as transparent data to the record layer.
5.1. Record Layer
The TLS record layer receives uninterpreted data from higher layers
in non-empty blocks of arbitrary size.
The record layer fragments information blocks into TLSPlaintext
records carrying data in chunks of 2^14 bytes or less. Message
boundaries are not preserved in the record layer (i.e., multiple
messages of the same ContentType MAY be coalesced into a single
TLSPlaintext record, or a single message MAY be fragmented across
several records). Alert messages (Section 6) MUST NOT be fragmented
across records.
enum {
alert(21),
handshake(22),
application_data(23)
(255)
} ContentType;
struct {
ContentType type;
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length;
opaque fragment[TLSPlaintext.length];
} TLSPlaintext;
type The higher-level protocol used to process the enclosed
fragment.
record_version The protocol version the current record is compatible
with. This value MUST be set to { 3, 1 } for all records. This
field is deprecated and MUST be ignored for all purposes.
length The length (in bytes) of the following TLSPlaintext.fragment.
The length MUST NOT exceed 2^14.
fragment The data being transmitted. This value transparent and
treated as an independent block to be dealt with by the higher-
level protocol specified by the type field.
This document describes TLS Version 1.3, which uses the version { 3,
4 }. The version value 3.4 is historical, deriving from the use of {
3, 1 } for TLS 1.0 and { 3, 0 } for SSL 3.0. In order to maximize
backwards compatibility, the record layer version identifies as
simply TLS 1.0. Endpoints supporting other versions negotiate the
version to use by following the procedure and requirements in
Appendix C.
Implementations MUST NOT send zero-length fragments of Handshake or
Alert types, even if those fragments contain padding. Zero-length
fragments of Application data MAY be sent as they are potentially
useful as a traffic analysis countermeasure.
When record protection has not yet been engaged, TLSPlaintext
structures are written directly onto the wire. Once record
protection has started, TLSPlaintext records are protected and sent
as described in the following section.
5.2. Record Payload Protection
The record protection functions translate a TLSPlaintext structure
into a TLSCiphertext. The deprotection functions reverse the
process. In TLS 1.3 as opposed to previous versions of TLS, all
ciphers are modeled as "Authenticated Encryption with Additional
Data" (AEAD) [RFC5116]. AEAD functions provide a unified encryption
and authentication operation which turns plaintext into authenticated
ciphertext and back again. Each encrypted record consists of a
plaintext header followed by an encrypted body, which itself contains
a type and optional padding.
struct {
opaque content[TLSPlaintext.length];
ContentType type;
uint8 zeros[length_of_padding];
} TLSInnerPlaintext;
struct {
ContentType opaque_type = application_data(23); /* see fragment.type */
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length;
opaque encrypted_record[length];
} TLSCiphertext;
content The cleartext of TLSPlaintext.fragment.
type The content type of the record.
zeros An arbitrary-length run of zero-valued bytes may appear in the
cleartext after the type field. This provides an opportunity for
senders to pad any TLS record by a chosen amount as long as the
total stays within record size limits. See Section 5.4 for more
details.
opaque_type The outer opaque_type field of a TLSCiphertext record is
always set to the value 23 (application_data) for outward
compatibility with middleboxes accustomed to parsing previous
versions of TLS. The actual content type of the record is found
in fragment.type after decryption.
record_version The record_version field is identical to
TLSPlaintext.record_version and is always { 3, 1 }. Note that the
handshake protocol including the ClientHello and ServerHello
messages authenticates the protocol version, so this value is
redundant.
length The length (in bytes) of the following
TLSCiphertext.fragment, which is the sum of the lengths of the
content and the padding, plus one for the inner content type. The
length MUST NOT exceed 2^14 + 256. An endpoint that receives a
record that exceeds this length MUST generate a fatal
"record_overflow" alert.
encrypted_record The AEAD encrypted form of the serialized
TLSInnerPlaintext structure.
AEAD ciphers take as input a single key, a nonce, a plaintext, and
"additional data" to be included in the authentication check, as
described in Section 2.1 of [RFC5116]. The key is either the
client_write_key or the server_write_key, the nonce is derived from
the sequence number (see Section 5.3) and the client_write_iv or
server_write_iv, and the additional data input is empty (zero
length). Derivation of traffic keys is defined in Section 7.3.
The plaintext is the concatenation of TLSPlaintext.fragment,
TLSPlaintext.type, and any padding bytes (zeros).
The AEAD output consists of the ciphertext output by the AEAD
encryption operation. The length of the plaintext is greater than
TLSPlaintext.length due to the inclusion of TLSPlaintext.type and
however much padding is supplied by the sender. The length of the
AEAD output will generally be larger than the plaintext, but by an
amount that varies with the AEAD cipher. Since the ciphers might
incorporate padding, the amount of overhead could vary with different
lengths of plaintext. Symbolically,
AEADEncrypted =
AEAD-Encrypt(write_key, nonce, plaintext of fragment)
In order to decrypt and verify, the cipher takes as input the key,
nonce, and the AEADEncrypted value. The output is either the
plaintext or an error indicating that the decryption failed. There
is no separate integrity check. That is:
plaintext of fragment =
AEAD-Decrypt(write_key, nonce, AEADEncrypted)
If the decryption fails, a fatal "bad_record_mac" alert MUST be
generated.
An AEAD cipher MUST NOT produce an expansion of greater than 255
bytes. An endpoint that receives a record from its peer with
TLSCipherText.length larger than 2^14 + 256 octets MUST generate a
fatal "record_overflow" alert. This limit is derived from the
maximum TLSPlaintext length of 2^14 octets + 1 octet for ContentType
+ the maximum AEAD expansion of 255 octets.
5.3. Per-Record Nonce
A 64-bit sequence number is maintained separately for reading and
writing records. Each sequence number is set to zero at the
beginning of a connection and whenever the key is changed.
The sequence number is incremented after reading or writing each
record. The first record transmitted under a particular set of
traffic keys record key MUST use sequence number 0.
Sequence numbers do not wrap. If a TLS implementation would need to
wrap a sequence number, it MUST either rekey (Section 4.4.3) or
terminate the connection.
The length of the per-record nonce (iv_length) is set to max(8 bytes,
N_MIN) for the AEAD algorithm (see [RFC5116] Section 4). An AEAD
algorithm where N_MAX is less than 8 bytes MUST NOT be used with TLS.
The per-record nonce for the AEAD construction is formed as follows:
1. The 64-bit record sequence number is padded to the left with
zeroes to iv_length.
2. The padded sequence number is XORed with the static
client_write_iv or server_write_iv, depending on the role.
The resulting quantity (of length iv_length) is used as the per-
record nonce.
Note: This is a different construction from that in TLS 1.2, which
specified a partially explicit nonce.
5.4. Record Padding
All encrypted TLS records can be padded to inflate the size of the
TLSCipherText. This allows the sender to hide the size of the
traffic from an observer.
When generating a TLSCiphertext record, implementations MAY choose to
pad. An unpadded record is just a record with a padding length of
zero. Padding is a string of zero-valued bytes appended to the
ContentType field before encryption. Implementations MUST set the
padding octets to all zeros before encrypting.
Application Data records may contain a zero-length fragment.content
if the sender desires. This permits generation of plausibly-sized
cover traffic in contexts where the presence or absence of activity
may be sensitive. Implementations MUST NOT send Handshake or Alert
records that have a zero-length fragment.content.
The padding sent is automatically verified by the record protection
mechanism: Upon successful decryption of a TLSCiphertext.fragment,
the receiving implementation scans the field from the end toward the
beginning until it finds a non-zero octet. This non-zero octet is
the content type of the message. This padding scheme was selected
because it allows padding of any encrypted TLS record by an arbitrary
size (from zero up to TLS record size limits) without introducing new
content types. The design also enforces all-zero padding octets,
which allows for quick detection of padding errors.
Implementations MUST limit their scanning to the cleartext returned
from the AEAD decryption. If a receiving implementation does not
find a non-zero octet in the cleartext, it should treat the record as
having an unexpected ContentType, sending an "unexpected_message"
alert.
The presence of padding does not change the overall record size
limitations - the full fragment plaintext may not exceed 2^14 octets.
Selecting a padding policy that suggests when and how much to pad is
a complex topic, and is beyond the scope of this specification. If
the application layer protocol atop TLS has its own padding padding,
it may be preferable to pad application_data TLS records within the
application layer. Padding for encrypted handshake and alert TLS
records must still be handled at the TLS layer, though. Later
documents may define padding selection algorithms, or define a
padding policy request mechanism through TLS extensions or some other
means.
5.5. Limits on Key Usage
There are cryptographic limits on the amount of plaintext which can
be safely encrypted under a given set of keys. [AEAD-LIMITS]
provides an analysis of these limits under the assumption that the
underlying primitive (AES or ChaCha20) has no weaknesses.
Implementations SHOULD do a key update Section 4.4.3 prior to
reaching these limits.
For AES-GCM, up to 2^24.5 full-size records may be encrypted on a
given connection while keeping a safety margin of approximately 2^-57
for Authenticated Encryption (AE) security. For ChaCha20/Poly1305,
the record sequence number will wrap before the safety limit is
reached.
6. Alert Protocol
One of the content types supported by the TLS record layer is the
alert type. Like other messages, alert messages are encrypted as
specified by the current connection state.
Alert messages convey the severity of the message (warning or fatal)
and a description of the alert. Warning-level messages are used to
indicate orderly closure of the connection (see Section 6.1). Upon
receiving a warning-level alert, the TLS implementation SHOULD
indicate end-of-data to the application and, if appropriate for the
alert type, send a closure alert in response.
Fatal-level messages are used to indicate abortive closure of the
connection (See Section 6.2). Upon receiving a fatal-level alert,
the TLS implementation SHOULD indicate an error to the application
and MUST NOT allow any further data to be sent or received on the
connection. Servers and clients MUST forget keys and secrets
associated with a failed connection. Stateful implementations of
session tickets (as in many clients) SHOULD discard tickets
associated with failed connections.
All the alerts listed in Section 6.2 MUST be sent as fatal and MUST
be treated as fatal regardless of the AlertLevel in the message.
Unknown alert types MUST be treated as fatal.
enum { warning(1), fatal(2), (255) } AlertLevel;
enum {
close_notify(0),
end_of_early_data(1),
unexpected_message(10),
bad_record_mac(20),
record_overflow(22),
handshake_failure(40),
bad_certificate(42),
unsupported_certificate(43),
certificate_revoked(44),
certificate_expired(45),
certificate_unknown(46),
illegal_parameter(47),
unknown_ca(48),
access_denied(49),
decode_error(50),
decrypt_error(51),
protocol_version(70),
insufficient_security(71),
internal_error(80),
inappropriate_fallback(86),
user_canceled(90),
missing_extension(109),
unsupported_extension(110),
certificate_unobtainable(111),
unrecognized_name(112),
bad_certificate_status_response(113),
bad_certificate_hash_value(114),
unknown_psk_identity(115),
(255)
} AlertDescription;
struct {
AlertLevel level;
AlertDescription description;
} Alert;
6.1. Closure Alerts
The client and the server must share knowledge that the connection is
ending in order to avoid a truncation attack. Failure to properly
close a connection does not prohibit a session from being resumed.
close_notify This alert notifies the recipient that the sender will
not send any more messages on this connection. Any data received
after a closure MUST be ignored.
end_of_early_data This alert is sent by the client to indicate that
all 0-RTT application_data messages have been transmitted (or none
will be sent at all) and that this is the end of the flight. This
alert MUST be at the warning level. Servers MUST NOT send this
alert and clients receiving it MUST terminate the connection with
an "unexpected_message" alert.
user_canceled This alert notifies the recipient that the sender is
canceling the handshake for some reason unrelated to a protocol
failure. If a user cancels an operation after the handshake is
complete, just closing the connection by sending a "close_notify"
is more appropriate. This alert SHOULD be followed by a
"close_notify". This alert is generally a warning.
Either party MAY initiate a close by sending a "close_notify" alert.
Any data received after a closure alert is ignored. If a transport-
level close is received prior to a "close_notify", the receiver
cannot know that all the data that was sent has been received.
Each party MUST send a "close_notify" alert before closing the write
side of the connection, unless some other fatal alert has been
transmitted. The other party MUST respond with a "close_notify"
alert of its own and close down the connection immediately,
discarding any pending writes. The initiator of the close need not
wait for the responding "close_notify" alert before closing the read
side of the connection.
If the application protocol using TLS provides that any data may be
carried over the underlying transport after the TLS connection is
closed, the TLS implementation must receive the responding
"close_notify" alert before indicating to the application layer that
the TLS connection has ended. If the application protocol will not
transfer any additional data, but will only close the underlying
transport connection, then the implementation MAY choose to close the
transport without waiting for the responding "close_notify". No part
of this standard should be taken to dictate the manner in which a
usage profile for TLS manages its data transport, including when
connections are opened or closed.
Note: It is assumed that closing a connection reliably delivers
pending data before destroying the transport.
6.2. Error Alerts
Error handling in the TLS Handshake Protocol is very simple. When an
error is detected, the detecting party sends a message to its peer.
Upon transmission or receipt of a fatal alert message, both parties
immediately close the connection. Whenever an implementation
encounters a condition which is defined as a fatal alert, it MUST
send the appropriate alert prior to closing the connection. All
alerts defined in this section below, as well as all unknown alerts
are universally considered fatal as of TLS 1.3 (see Section 6).
The following error alerts are defined:
unexpected_message An inappropriate message was received. This
alert should never be observed in communication between proper
implementations.
bad_record_mac This alert is returned if a record is received which
cannot be deprotected. Because AEAD algorithms combine decryption
and verification, this alert is used for all deprotection
failures. This alert should never be observed in communication
between proper implementations, except when messages were
corrupted in the network.
record_overflow A TLSCiphertext record was received that had a
length more than 2^14 + 256 bytes, or a record decrypted to a
TLSPlaintext record with more than 2^14 bytes. This alert should
never be observed in communication between proper implementations,
except when messages were corrupted in the network.
handshake_failure Reception of a "handshake_failure" alert message
indicates that the sender was unable to negotiate an acceptable
set of security parameters given the options available.
bad_certificate A certificate was corrupt, contained signatures that
did not verify correctly, etc.
unsupported_certificate A certificate was of an unsupported type.
certificate_revoked A certificate was revoked by its signer.
certificate_expired A certificate has expired or is not currently
valid.
certificate_unknown Some other (unspecified) issue arose in
processing the certificate, rendering it unacceptable.
illegal_parameter A field in the handshake was out of range or
inconsistent with other fields.
unknown_ca A valid certificate chain or partial chain was received,
but the certificate was not accepted because the CA certificate
could not be located or couldn't be matched with a known, trusted
CA.
access_denied A valid certificate or PSK was received, but when
access control was applied, the sender decided not to proceed with
negotiation.
decode_error A message could not be decoded because some field was
out of the specified range or the length of the message was
incorrect. This alert should never be observed in communication
between proper implementations, except when messages were
corrupted in the network.
decrypt_error A handshake cryptographic operation failed, including
being unable to correctly verify a signature or validate a
Finished message.
protocol_version The protocol version the peer has attempted to
negotiate is recognized but not supported. (see Appendix C)
insufficient_security Returned instead of "handshake_failure" when a
negotiation has failed specifically because the server requires
ciphers more secure than those supported by the client.
internal_error An internal error unrelated to the peer or the
correctness of the protocol (such as a memory allocation failure)
makes it impossible to continue.
inappropriate_fallback Sent by a server in response to an invalid
connection retry attempt from a client. (see [RFC7507])
missing_extension Sent by endpoints that receive a hello message not
containing an extension that is mandatory to send for the offered
TLS version. [[TODO: IANA Considerations.]]
unsupported_extension Sent by endpoints receiving any hello message
containing an extension known to be prohibited for inclusion in
the given hello message, including any extensions in a ServerHello
not first offered in the corresponding ClientHello.
certificate_unobtainable Sent by servers when unable to obtain a
certificate from a URL provided by the client via the
"client_certificate_url" extension [RFC6066].
unrecognized_name Sent by servers when no server exists identified
by the name provided by the client via the "server_name" extension
[RFC6066].
bad_certificate_status_response Sent by clients when an invalid or
unacceptable OCSP response is provided by the server via the
"status_request" extension [RFC6066]. This alert is always fatal.
bad_certificate_hash_value Sent by servers when a retrieved object
does not have the correct hash provided by the client via the
"client_certificate_url" extension [RFC6066].
unknown_psk_identity Sent by servers when a PSK cipher suite is
selected but no acceptable PSK identity is provided by the client.
Sending this alert is OPTIONAL; servers MAY instead choose to send
a "decrypt_error" alert to merely indicate an invalid PSK
identity.
New Alert values are assigned by IANA as described in Section 10.
7. Cryptographic Computations 7. Cryptographic Computations
In order to begin connection protection, the TLS Record Protocol In order to begin connection protection, the TLS Record Protocol
requires specification of a suite of algorithms, a master secret, and requires specification of a suite of algorithms, a master secret, and
the client and server random values. The authentication, key the client and server random values. The authentication, key
exchange, and record protection algorithms are determined by the exchange, and record protection algorithms are determined by the
cipher_suite selected by the server and revealed in the ServerHello cipher_suite selected by the server and revealed in the ServerHello
message. The random values are exchanged in the hello messages. All message. The random values are exchanged in the hello messages. All
that remains is to calculate the key schedule. that remains is to calculate the key schedule.
7.1. Key Schedule 7.1. Key Schedule
The TLS handshake establishes one or more input secrets which are The TLS handshake establishes one or more input secrets which are
combined to create the actual working keying material, as detailed combined to create the actual working keying material, as detailed
below. The key derivation process makes use of the following below. The key derivation process makes use of the HKDF-Extract and
functions, based on HKDF [RFC5869]: HKDF-Expand functions as defined for HKDF [RFC5869], as well as the
functions defined below:
HKDF-Extract(Salt, IKM) as defined in {{RFC5869}}.
HKDF-Expand-Label(Secret, Label, Messages, Length) = HKDF-Expand-Label(Secret, Label, HashValue, Length) =
HKDF-Expand(Secret, HkdfLabel, Length) HKDF-Expand(Secret, HkdfLabel, Length)
Where HkdfLabel is specified as: Where HkdfLabel is specified as:
struct HkdfLabel { struct HkdfLabel
uint16 length; {
opaque label<9..255>; uint16 length = Length;
opaque hash_value<0..255>; opaque label<9..255> = "TLS 1.3, " + Label;
}; opaque hash_value<0..255> = HashValue;
};
- HkdfLabel.length is Length Derive-Secret(Secret, Label, Messages) =
- HkdfLabel.label is "TLS 1.3, " + Label HKDF-Expand-Label(Secret, Label,
- HkdfLabel.hash_value is HashValue. Hash(Messages) +
Hash(resumption_context), Hash.Length)
Derive-Secret(Secret, Label, Messages) = The Hash function and the HKDF hash are the cipher suite hash
HKDF-Expand-Label(Secret, Label, function. Hash.Length is its output length.
Hash(Messages) + Hash(resumption_context), L))
Given a set of n InputSecrets, the final "master secret" is computed Given a set of n InputSecrets, the final "master secret" is computed
by iteratively invoking HKDF-Extract with InputSecret_1, by iteratively invoking HKDF-Extract with InputSecret_1,
InputSecret_2, etc. The initial secret is simply a string of 0s as InputSecret_2, etc. The initial secret is simply a string of zeroes
long as the size of the Hash that is the basis for the HKDF. as long as the size of the Hash that is the basis for the HKDF.
Concretely, for the present version of TLS 1.3, secrets are added in Concretely, for the present version of TLS 1.3, secrets are added in
the following order: the following order:
- PSK - PSK
- (EC)DHE shared secret - (EC)DHE shared secret
This produces a full key derivation schedule shown in the diagram This produces a full key derivation schedule shown in the diagram
below. In this diagram, the following formatting conventions apply: below. In this diagram, the following formatting conventions apply:
skipping to change at page 75, line 25 skipping to change at page 71, line 11
- Derive-Secret's Secret argument is indicated by the arrow coming - Derive-Secret's Secret argument is indicated by the arrow coming
in from the left. For instance, the Early Secret is the Secret in from the left. For instance, the Early Secret is the Secret
for generating the early_traffic-secret. for generating the early_traffic-secret.
0 0
| |
v v
PSK -> HKDF-Extract PSK -> HKDF-Extract
| |
v v
Early Secret --> Derive-Secret(., "early traffic secret", Early Secret ---> Derive-Secret(., "early traffic secret",
| ClientHello) | ClientHello)
| = early_traffic_secret | = early_traffic_secret
v v
(EC)DHE -> HKDF-Extract (EC)DHE -> HKDF-Extract
| |
v v
Handshake Handshake
Secret -----> Derive-Secret(., "handshake traffic secret", Secret -----> Derive-Secret(., "handshake traffic secret",
| ClientHello + ServerHello) | ClientHello + ServerHello)
| = handshake_traffic_secret | = handshake_traffic_secret
skipping to change at page 76, line 15 skipping to change at page 71, line 50
The general pattern here is that the secrets shown down the left side The general pattern here is that the secrets shown down the left side
of the diagram are just raw entropy without context, whereas the of the diagram are just raw entropy without context, whereas the
secrets down the right side include handshake context and therefore secrets down the right side include handshake context and therefore
can be used to derive working keys without additional context. Note can be used to derive working keys without additional context. Note
that the different calls to Derive-Secret may take different Messages that the different calls to Derive-Secret may take different Messages
arguments, even with the same secret. In a 0-RTT exchange, Derive- arguments, even with the same secret. In a 0-RTT exchange, Derive-
Secret is called with four distinct transcripts; in a 1-RTT only Secret is called with four distinct transcripts; in a 1-RTT only
exchange with three distinct transcripts. exchange with three distinct transcripts.
If a given secret is not available, then the 0-value consisting of a If a given secret is not available, then the 0-value consisting of a
string of L zeroes is used. string of Hash.length zeroes is used. Note that this does not mean
skipping rounds, so if PSK is not in use Early Secret will still be
HKDF-Extract(0, 0).
7.2. Updating Traffic Keys and IVs 7.2. Updating Traffic Keys and IVs
Once the handshake is complete, it is possible for either side to Once the handshake is complete, it is possible for either side to
update its sending traffic keys using the KeyUpdate handshake message update its sending traffic keys using the KeyUpdate handshake message
Section 6.3.5.3. The next generation of traffic keys is computed by defined in Section 4.4.3. The next generation of traffic keys is
generating traffic_secret_N+1 from traffic_secret_N as described in computed by generating traffic_secret_N+1 from traffic_secret_N as
this section then re-deriving the traffic keys as described in described in this section then re-deriving the traffic keys as
Section 7.3. described in Section 7.3.
The next-generation traffic_secret is computed as: The next-generation traffic_secret is computed as:
traffic_secret_N+1 = HKDF-Expand-Label(traffic_secret_N, "application traffic_secret_N+1 = HKDF-Expand-Label(
traffic secret", "", L) traffic_secret_N,
"application traffic secret", "", Hash.Length)
Once traffic_secret_N+1 and its associated traffic keys have been Once traffic_secret_N+1 and its associated traffic keys have been
computed, implementations SHOULD delete traffic_secret_N. Once the computed, implementations SHOULD delete traffic_secret_N. Once the
directional keys are no longer needed, they SHOULD be deleted as directional keys are no longer needed, they SHOULD be deleted as
well. well.
7.3. Traffic Key Calculation 7.3. Traffic Key Calculation
The traffic keying material is generated from the following input The traffic keying material is generated from the following input
values: values:
- A secret value - A secret value
- A phase value indicating the phase of the protocol the keys are - A phase value indicating the phase of the protocol the keys are
being generated for. being generated for
- A purpose value indicating the specific value being generated - A purpose value indicating the specific value being generated
- The length of the key. - The length of the key
The keying material is computed using: The keying material is computed using:
key = HKDF-Expand-Label(Secret, key = HKDF-Expand-Label(Secret,
phase + ", " + purpose, "", phase + ", " + purpose,
"",
key_length) key_length)
The following table describes the inputs to the key calculation for The following table describes the inputs to the key calculation for
each class of traffic keys: each class of traffic keys:
+-------------+--------------------------+--------------------------+ +-------------+--------------------------+--------------------------+
| Record Type | Secret | Phase | | Record Type | Secret | Phase |
+-------------+--------------------------+--------------------------+ +-------------+--------------------------+--------------------------+
| 0-RTT | early_traffic_secret | "early handshake key | | 0-RTT | early_traffic_secret | "early handshake key |
| Handshake | | expansion" | | Handshake | | expansion" |
skipping to change at page 77, line 34 skipping to change at page 73, line 27
| Application | traffic_secret_N | "application data key | | Application | traffic_secret_N | "application data key |
| Data | | expansion" | | Data | | expansion" |
+-------------+--------------------------+--------------------------+ +-------------+--------------------------+--------------------------+
The following table indicates the purpose values for each type of The following table indicates the purpose values for each type of
key: key:
+------------------+--------------------+ +------------------+--------------------+
| Key Type | Purpose | | Key Type | Purpose |
+------------------+--------------------+ +------------------+--------------------+
| Client Write Key | "client write key" | | client_write_key | "client write key" |
| | | | | |
| Server Write Key | "server write key" | | server_write_key | "server write key" |
| | | | | |
| Client Write IV | "client write iv" | | client_write_iv | "client write iv" |
| | | | | |
| Server Write IV | "server write iv" | | server_write_iv | "server write iv" |
+------------------+--------------------+ +------------------+--------------------+
All the traffic keying material is recomputed whenever the underlying All the traffic keying material is recomputed whenever the underlying
Secret changes (e.g., when changing from the handshake to application Secret changes (e.g., when changing from the handshake to application
data keys or upon a key update). data keys or upon a key update).
7.3.1. Diffie-Hellman 7.3.1. Diffie-Hellman
A conventional Diffie-Hellman computation is performed. The A conventional Diffie-Hellman computation is performed. The
negotiated key (Z) is converted to byte string by encoding in big- negotiated key (Z) is converted to byte string by encoding in big-
skipping to change at page 79, line 5 skipping to change at page 74, line 47
7.3.3. Exporters 7.3.3. Exporters
[RFC5705] defines keying material exporters for TLS in terms of the [RFC5705] defines keying material exporters for TLS in terms of the
TLS PRF. This document replaces the PRF with HKDF, thus requiring a TLS PRF. This document replaces the PRF with HKDF, thus requiring a
new construction. The exporter interface remains the same, however new construction. The exporter interface remains the same, however
the value is computed as: the value is computed as:
HKDF-Expand-Label(exporter_secret, HKDF-Expand-Label(exporter_secret,
label, context_value, key_length) label, context_value, key_length)
8. Mandatory Algorithms 8. Compliance Requirements
8.1. MTI Cipher Suites 8.1. MTI Cipher Suites
In the absence of an application profile standard specifying In the absence of an application profile standard specifying
otherwise, a TLS-compliant application MUST implement the following otherwise, a TLS-compliant application MUST implement the following
cipher suites: cipher suites:
TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256
TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
These cipher suites MUST support both digital signatures and key These cipher suites MUST support both digital signatures and key
skipping to change at page 79, line 34 skipping to change at page 75, line 31
TLS_ECDHE_ECDSA_WITH_CHACHA20_POLY1305_SHA256 TLS_ECDHE_ECDSA_WITH_CHACHA20_POLY1305_SHA256
TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384
TLS_ECDHE_RSA_WITH_CHACHA20_POLY1305_SHA256 TLS_ECDHE_RSA_WITH_CHACHA20_POLY1305_SHA256
8.2. MTI Extensions 8.2. MTI Extensions
In the absence of an application profile standard specifying In the absence of an application profile standard specifying
otherwise, a TLS-compliant application MUST implement the following otherwise, a TLS-compliant application MUST implement the following
TLS extensions: TLS extensions:
- Signature Algorithms ("signature_algorithms"; Section 6.3.2.2) - Signature Algorithms ("signature_algorithms"; Section 4.2.2)
- Negotiated Groups ("supported_groups"; Section 6.3.2.3) - Negotiated Groups ("supported_groups"; Section 4.2.3)
- Key Share ("key_share"; Section 6.3.2.4) - Key Share ("key_share"; Section 4.2.4)
- Pre-Shared Key ("pre_shared_key"; Section 6.3.2.5) - Pre-Shared Key ("pre_shared_key"; Section 4.2.5)
- Server Name Indication ("server_name"; Section 3 of [RFC6066]) - Server Name Indication ("server_name"; Section 3 of [RFC6066])
- Cookie ("cookie"; Section 6.3.2.1) - Cookie ("cookie"; Section 4.2.1)
All implementations MUST send and use these extensions when offering All implementations MUST send and use these extensions when offering
applicable cipher suites: applicable cipher suites:
- "signature_algorithms" is REQUIRED for certificate authenticated - "signature_algorithms" is REQUIRED for certificate authenticated
cipher suites cipher suites.
- "supported_groups" and "key_share" are REQUIRED for DHE or ECDHE - "supported_groups" and "key_share" are REQUIRED for DHE or ECDHE
cipher suites cipher suites.
- "pre_shared_key" is REQUIRED for PSK cipher suites - "pre_shared_key" is REQUIRED for PSK cipher suites.
- "cookie" is REQUIRED for all cipher suites. - "cookie" is REQUIRED for all cipher suites.
When negotiating use of applicable cipher suites, endpoints MUST When negotiating use of applicable cipher suites, endpoints MUST
abort the connection with a "missing_extension" alert if the required abort the connection with a "missing_extension" alert if the required
extension was not provided. Any endpoint that receives any invalid extension was not provided. Any endpoint that receives any invalid
combination of cipher suites and extensions MAY abort the connection combination of cipher suites and extensions MAY abort the connection
with a "missing_extension" alert, regardless of negotiated with a "missing_extension" alert, regardless of negotiated
parameters. parameters.
skipping to change at page 80, line 30 skipping to change at page 76, line 27
"server_name" extension with applications capable of using it. "server_name" extension with applications capable of using it.
Servers MAY require clients to send a valid "server_name" extension. Servers MAY require clients to send a valid "server_name" extension.
Servers requiring this extension SHOULD respond to a ClientHello Servers requiring this extension SHOULD respond to a ClientHello
lacking a "server_name" extension with a fatal "missing_extension" lacking a "server_name" extension with a fatal "missing_extension"
alert. alert.
Servers MUST NOT send the "signature_algorithms" extension; if a Servers MUST NOT send the "signature_algorithms" extension; if a
client receives this extension it MUST respond with a fatal client receives this extension it MUST respond with a fatal
"unsupported_extension" alert and close the connection. "unsupported_extension" alert and close the connection.
9. Application Data Protocol 9. Security Considerations
Application data messages are carried by the record layer and are
fragmented and encrypted based on the current connection state. The
messages are treated as transparent data to the record layer.
10. Security Considerations
Security issues are discussed throughout this memo, especially in Security issues are discussed throughout this memo, especially in
Appendices B, C, and D. Appendices B, C, and D.
11. IANA Considerations 10. IANA Considerations
This document uses several registries that were originally created in This document uses several registries that were originally created in
[RFC4346]. IANA has updated these to reference this document. The [RFC4346]. IANA has updated these to reference this document. The
registries and their allocation policies are below: registries and their allocation policies are below:
- TLS Cipher Suite Registry: Values with the first byte in the range - TLS Cipher Suite Registry: Values with the first byte in the range
0-254 (decimal) are assigned via Specification Required [RFC2434]. 0-254 (decimal) are assigned via Specification Required [RFC2434].
Values with the first byte 255 (decimal) are reserved for Private Values with the first byte 255 (decimal) are reserved for Private
Use [RFC2434]. IANA [SHALL add/has added] a "Recommended" column Use [RFC2434]. IANA [SHALL add/has added] a "Recommended" column
to the cipher suite registry. All cipher suites listed in to the cipher suite registry. All cipher suites listed in
skipping to change at page 81, line 19 skipping to change at page 77, line 12
suites marked "No" range from "good" to "bad" from a suites marked "No" range from "good" to "bad" from a
cryptographic standpoint. cryptographic standpoint.
- TLS ContentType Registry: Future values are allocated via - TLS ContentType Registry: Future values are allocated via
Standards Action [RFC2434]. Standards Action [RFC2434].
- TLS Alert Registry: Future values are allocated via Standards - TLS Alert Registry: Future values are allocated via Standards
Action [RFC2434]. Action [RFC2434].
- TLS HandshakeType Registry: Future values are allocated via - TLS HandshakeType Registry: Future values are allocated via
Standards Action [RFC2434]. Standards Action [RFC2434]. IANA [SHALL update/has updated] this
registry to rename item 4 from "NewSessionTicket" to
"new_session_ticket".
This document also uses a registry originally created in [RFC4366]. This document also uses a registry originally created in [RFC4366].
IANA has updated it to reference this document. The registry and its IANA has updated it to reference this document. The registry and its
allocation policy is listed below: allocation policy is listed below:
- TLS ExtensionType Registry: Values with the first byte in the - TLS ExtensionType Registry: Values with the first byte in the
range 0-254 (decimal) are assigned via Specification Required range 0-254 (decimal) are assigned via Specification Required
[RFC2434]. Values with the first byte 255 (decimal) are reserved [RFC2434]. Values with the first byte 255 (decimal) are reserved
for Private Use [RFC2434]. IANA [SHALL update/has updated] this for Private Use [RFC2434]. IANA [SHALL update/has updated] this
registry to include the "key_share", "pre_shared_key", and registry to include the "key_share", "pre_shared_key", and
"early_data" extensions as defined in this document. "early_data" extensions as defined in this document.
IANA [shall update/has updated] this registry to include a "TLS IANA [shall update/has updated] this registry to include a "TLS
1.3" column with the following four values: "Client", indicating 1.3" column with the following four values: "Client", indicating
that the server shall not send them. "Clear", indicating that that the server shall not send them. "Clear", indicating that
they shall be in the ServerHello. "Encrypted", indicating that they shall be in the ServerHello. "Encrypted", indicating that
they shall be in the EncryptedExtensions block, "Early", they shall be in the EncryptedExtensions block, and "No"
indicating that they shall be only in the client's 0-RTT indicating that they are not used in TLS 1.3. This column [shall
EncryptedExtensions block, and "No" indicating that they are not be/has been] initially populated with the values in this document.
used in TLS 1.3. This column [shall be/has been] initially IANA [shall update/has updated] this registry to add a
populated with the values in this document. IANA [shall update/ "Recommended" column. IANA [shall/has] initially populated this
has updated] this registry to add a "Recommended" column. IANA column with the values in the table below. This table has been
[shall/has] initially populated this column with the values in the generated by marking Standards Track RFCs as "Yes" and all others
table below. This table has been generated by marking Standards as "No".
Track RFCs as "Yes" and all others as "No".
+-------------------------------+-----------+-----------------------+ +-------------------------------+-----------+-----------------------+
| Extension | Recommend | TLS 1.3 | | Extension | Recommend | TLS 1.3 |
| | ed | | | | ed | |
+-------------------------------+-----------+-----------------------+ +-------------------------------+-----------+-----------------------+
| server_name [RFC6066] | Yes | Encrypted | | server_name [RFC6066] | Yes | Encrypted |
| | | | | | | |
| max_fragment_length [RFC6066] | Yes | Encrypted | | max_fragment_length [RFC6066] | Yes | Encrypted |
| | | | | | | |
| client_certificate_url | Yes | Encrypted | | client_certificate_url | Yes | Encrypted |
skipping to change at page 83, line 19 skipping to change at page 79, line 12
| | | | | | | |
| SessionTicket TLS [RFC4507] | Yes | No | | SessionTicket TLS [RFC4507] | Yes | No |
| | | | | | | |
| renegotiation_info [RFC5746] | Yes | No | | renegotiation_info [RFC5746] | Yes | No |
| | | | | | | |
| key_share [[this document]] | Yes | Clear | | key_share [[this document]] | Yes | Clear |
| | | | | | | |
| pre_shared_key [[this | Yes | Clear | | pre_shared_key [[this | Yes | Clear |
| document]] | | | | document]] | | |
| | | | | | | |
| early_data [[this document]] | Yes | Clear | | early_data [[this document]] | Yes | Encrypted |
| | | |
| ticket_age [[this document]] | Yes | Early |
| | | | | | | |
| cookie [[this document]] | Yes | Encrypted/HelloRetryR | | cookie [[this document]] | Yes | Encrypted/HelloRetryR |
| | | equest | | | | equest |
+-------------------------------+-----------+-----------------------+ +-------------------------------+-----------+-----------------------+
In addition, this document defines two new registries to be In addition, this document defines two new registries to be
maintained by IANA maintained by IANA
- TLS SignatureScheme Registry: Values with the first byte in the - TLS SignatureScheme Registry: Values with the first byte in the
range 0-254 (decimal) are assigned via Specification Required range 0-254 (decimal) are assigned via Specification Required
[RFC2434]. Values with the first byte 255 (decimal) are reserved [RFC2434]. Values with the first byte 255 (decimal) are reserved
for Private Use [RFC2434]. This registry SHALL have a for Private Use [RFC2434]. This registry SHALL have a
"Recommended" column. The registry [shall be/ has been] initially "Recommended" column. The registry [shall be/ has been] initially
populated with the values described in Section 6.3.2.2. The populated with the values described in Section 4.2.2. The
following values SHALL be marked as "Recommended": following values SHALL be marked as "Recommended":
ecdsa_secp256r1_sha256, ecdsa_secp384r1_sha384, rsa_pss_sha256, ecdsa_secp256r1_sha256, ecdsa_secp384r1_sha384, rsa_pss_sha256,
rsa_pss_sha384, rsa_pss_sha512, ed25519. rsa_pss_sha384, rsa_pss_sha512, ed25519.
12. References 11. References
12.1. Normative References 11.1. Normative References
[AES] National Institute of Standards and Technology, [AES] National Institute of Standards and Technology,
"Specification for the Advanced Encryption Standard "Specification for the Advanced Encryption Standard
(AES)", NIST FIPS 197, November 2001. (AES)", NIST FIPS 197, November 2001.
[DH] Diffie, W. and M. Hellman, "New Directions in [DH] Diffie, W. and M. Hellman, "New Directions in
Cryptography", IEEE Transactions on Information Theory, Cryptography", IEEE Transactions on Information Theory,
V.IT-22 n.6 , June 1977. V.IT-22 n.6 , June 1977.
[I-D.ietf-tls-chacha20-poly1305]
Langley, A., Chang, W., Mavrogiannopoulos, N.,
Strombergson, J., and S. Josefsson, "ChaCha20-Poly1305
Cipher Suites for Transport Layer Security (TLS)", draft-
ietf-tls-chacha20-poly1305-04 (work in progress), December
2015.
[I-D.irtf-cfrg-eddsa] [I-D.irtf-cfrg-eddsa]
Josefsson, S. and I. Liusvaara, "Edwards-curve Digital Josefsson, S. and I. Liusvaara, "Edwards-curve Digital
Signature Algorithm (EdDSA)", draft-irtf-cfrg-eddsa-05 Signature Algorithm (EdDSA)", draft-irtf-cfrg-eddsa-05
(work in progress), March 2016. (work in progress), March 2016.
[I-D.mattsson-tls-ecdhe-psk-aead] [I-D.mattsson-tls-ecdhe-psk-aead]
Mattsson, J. and D. Migault, "ECDHE_PSK with AES-GCM and Mattsson, J. and D. Migault, "ECDHE_PSK with AES-GCM and
AES-CCM Cipher Suites for Transport Layer Security (TLS)", AES-CCM Cipher Suites for Transport Layer Security (TLS)",
draft-mattsson-tls-ecdhe-psk-aead-05 (work in progress), draft-mattsson-tls-ecdhe-psk-aead-05 (work in progress),
April 2016. April 2016.
skipping to change at page 85, line 49 skipping to change at page 81, line 39
[RFC6655] McGrew, D. and D. Bailey, "AES-CCM Cipher Suites for [RFC6655] McGrew, D. and D. Bailey, "AES-CCM Cipher Suites for
Transport Layer Security (TLS)", RFC 6655, Transport Layer Security (TLS)", RFC 6655,
DOI 10.17487/RFC6655, July 2012, DOI 10.17487/RFC6655, July 2012,
<http://www.rfc-editor.org/info/rfc6655>. <http://www.rfc-editor.org/info/rfc6655>.
[RFC6961] Pettersen, Y., "The Transport Layer Security (TLS) [RFC6961] Pettersen, Y., "The Transport Layer Security (TLS)
Multiple Certificate Status Request Extension", RFC 6961, Multiple Certificate Status Request Extension", RFC 6961,
DOI 10.17487/RFC6961, June 2013, DOI 10.17487/RFC6961, June 2013,
<http://www.rfc-editor.org/info/rfc6961>. <http://www.rfc-editor.org/info/rfc6961>.
[RFC6962] Laurie, B., Langley, A., and E. Kasper, "Certificate [RFC6979] Pornin, T., "Deterministic Usage of the Digital Signature
Transparency", RFC 6962, DOI 10.17487/RFC6962, June 2013, Algorithm (DSA) and Elliptic Curve Digital Signature
<http://www.rfc-editor.org/info/rfc6962>. Algorithm (ECDSA)", RFC 6979, DOI 10.17487/RFC6979, August
2013, <http://www.rfc-editor.org/info/rfc6979>.
[RFC7251] McGrew, D., Bailey, D., Campagna, M., and R. Dugal, "AES- [RFC7251] McGrew, D., Bailey, D., Campagna, M., and R. Dugal, "AES-
CCM Elliptic Curve Cryptography (ECC) Cipher Suites for CCM Elliptic Curve Cryptography (ECC) Cipher Suites for
TLS", RFC 7251, DOI 10.17487/RFC7251, June 2014, TLS", RFC 7251, DOI 10.17487/RFC7251, June 2014,
<http://www.rfc-editor.org/info/rfc7251>. <http://www.rfc-editor.org/info/rfc7251>.
[RFC7443] Patil, P., Reddy, T., Salgueiro, G., and M. Petit-
Huguenin, "Application-Layer Protocol Negotiation (ALPN)
Labels for Session Traversal Utilities for NAT (STUN)
Usages", RFC 7443, DOI 10.17487/RFC7443, January 2015,
<http://www.rfc-editor.org/info/rfc7443>.
[RFC7748] Langley, A., Hamburg, M., and S. Turner, "Elliptic Curves [RFC7748] Langley, A., Hamburg, M., and S. Turner, "Elliptic Curves
for Security", RFC 7748, DOI 10.17487/RFC7748, January for Security", RFC 7748, DOI 10.17487/RFC7748, January
2016, <http://www.rfc-editor.org/info/rfc7748>. 2016, <http://www.rfc-editor.org/info/rfc7748>.
[RFC7905] Langley, A., Chang, W., Mavrogiannopoulos, N.,
Strombergson, J., and S. Josefsson, "ChaCha20-Poly1305
Cipher Suites for Transport Layer Security (TLS)",
RFC 7905, DOI 10.17487/RFC7905, June 2016,
<http://www.rfc-editor.org/info/rfc7905>.
[SHS] National Institute of Standards and Technology, U.S. [SHS] National Institute of Standards and Technology, U.S.
Department of Commerce, "Secure Hash Standard", NIST FIPS Department of Commerce, "Secure Hash Standard", NIST FIPS
PUB 180-4, March 2012. PUB 180-4, March 2012.
[X690] ITU-T, "Information technology - ASN.1 encoding Rules: [X690] ITU-T, "Information technology - ASN.1 encoding Rules:
Specification of Basic Encoding Rules (BER), Canonical Specification of Basic Encoding Rules (BER), Canonical
Encoding Rules (CER) and Distinguished Encoding Rules Encoding Rules (CER) and Distinguished Encoding Rules
(DER)", ISO/IEC 8825-1:2002, 2002. (DER)", ISO/IEC 8825-1:2002, 2002.
[X962] ANSI, "Public Key Cryptography For The Financial Services [X962] ANSI, "Public Key Cryptography For The Financial Services
Industry: The Elliptic Curve Digital Signature Algorithm Industry: The Elliptic Curve Digital Signature Algorithm
(ECDSA)", ANSI X9.62, 1998. (ECDSA)", ANSI X9.62, 1998.
12.2. Informative References 11.2. Informative References
[AEAD-LIMITS]
Luykx, A. and K. Paterson, "Limits on Authenticated
Encryption Use in TLS", 2016,
<http://www.isg.rhul.ac.uk/~kp/TLS-AEbounds.pdf>.
[BBFKZG16]
Bhargavan, K., Brzuska, C., Fournet, C., Kohlweiss, M.,
Zanella-Beguelin, S., and M. Green, "Downgrade Resilience
in Key-Exchange Protocols", Proceedings of IEEE Symposium
on Security and Privacy (Oakland) 2016 , 2016.
[CHSV16] Cremers, C., Horvat, M., Scott, S., and T. van der Merwe,
"Automated Analysis and Verification of TLS 1.3: 0-RTT,
Resumption and Delayed Authentication", Proceedings of
IEEE Symposium on Security and Privacy (Oakland) 2016 ,
2016.
[CK01] Canetti, R. and H. Krawczyk, "Analysis of Key-Exchange
Protocols and Their Use for Building Secure Channels",
Proceedings of Eurocrypt 2001 , 2001.
[DOW92] Diffie, W., van Oorschot, P., and M. Wiener,
""Authentication and authenticated key exchanges"",
Designs, Codes and Cryptography , n.d..
[DSS] National Institute of Standards and Technology, U.S. [DSS] National Institute of Standards and Technology, U.S.
Department of Commerce, "Digital Signature Standard, Department of Commerce, "Digital Signature Standard,
version 4", NIST FIPS PUB 186-4, 2013. version 4", NIST FIPS PUB 186-4, 2013.
[ECDSA] American National Standards Institute, "Public Key [ECDSA] American National Standards Institute, "Public Key
Cryptography for the Financial Services Industry: The Cryptography for the Financial Services Industry: The
Elliptic Curve Digital Signature Algorithm (ECDSA)", Elliptic Curve Digital Signature Algorithm (ECDSA)",
ANSI ANS X9.62-2005, November 2005. ANSI ANS X9.62-2005, November 2005.
[FGSW16] Fischlin, M., Guenther, F., Schmidt, B., and B. Warinschi,
"Key Confirmation in Key Exchange: A Formal Treatment and
Implications for TLS 1.3", Proceedings of IEEE Symposium
on Security and Privacy (Oakland) 2016 , 2016.
[FI06] Finney, H., "Bleichenbacher's RSA signature forgery based [FI06] Finney, H., "Bleichenbacher's RSA signature forgery based
on implementation error", August 2006, on implementation error", August 2006,
<https://www.ietf.org/mail-archive/web/openpgp/current/ <https://www.ietf.org/mail-archive/web/openpgp/current/
msg00999.html>. msg00999.html>.
[GCM] Dworkin, M., "Recommendation for Block Cipher Modes of [GCM] Dworkin, M., "Recommendation for Block Cipher Modes of
Operation: Galois/Counter Mode (GCM) and GMAC", Operation: Galois/Counter Mode (GCM) and GMAC",
NIST Special Publication 800-38D, November 2007. NIST Special Publication 800-38D, November 2007.
[I-D.ietf-tls-cached-info] [I-D.ietf-tls-cached-info]
skipping to change at page 87, line 14 skipping to change at page 84, line 5
[I-D.ietf-tls-negotiated-ff-dhe] [I-D.ietf-tls-negotiated-ff-dhe]
Gillmor, D., "Negotiated Finite Field Diffie-Hellman Gillmor, D., "Negotiated Finite Field Diffie-Hellman
Ephemeral Parameters for TLS", draft-ietf-tls-negotiated- Ephemeral Parameters for TLS", draft-ietf-tls-negotiated-
ff-dhe-10 (work in progress), June 2015. ff-dhe-10 (work in progress), June 2015.
[IEEE1363] [IEEE1363]
IEEE, "Standard Specifications for Public Key IEEE, "Standard Specifications for Public Key
Cryptography", IEEE 1363 , 2000. Cryptography", IEEE 1363 , 2000.
[LXZFH16] Li, X., Xu, J., Feng, D., Zhang, Z., and H. Hu, "Multiple
Handshakes Security of TLS 1.3 Candidates", Proceedings of
IEEE Symposium on Security and Privacy (Oakland) 2016 ,
2016.
[PKCS6] RSA Laboratories, "PKCS #6: RSA Extended Certificate [PKCS6] RSA Laboratories, "PKCS #6: RSA Extended Certificate
Syntax Standard, version 1.5", November 1993. Syntax Standard, version 1.5", November 1993.
[PKCS7] RSA Laboratories, "PKCS #7: RSA Cryptographic Message [PKCS7] RSA Laboratories, "PKCS #7: RSA Cryptographic Message
Syntax Standard, version 1.5", November 1993. Syntax Standard, version 1.5", November 1993.
[PSK-FINISHED] [PSK-FINISHED]
Cremers, C., Horvat, M., van der Merwe, T., and S. Scott, Cremers, C., Horvat, M., van der Merwe, T., and S. Scott,
"Revision 10: possible attack if client authentication is "Revision 10: possible attack if client authentication is
allowed during PSK", 2015, <https://www.ietf.org/mail- allowed during PSK", 2015, <https://www.ietf.org/mail-
archive/web/tls/current/msg18215.html>. archive/web/tls/current/msg18215.html>.
[RFC0793] Postel, J., "Transmission Control Protocol", STD 7, [RFC0793] Postel, J., "Transmission Control Protocol", STD 7,
RFC 793, DOI 10.17487/RFC0793, September 1981, RFC 793, DOI 10.17487/RFC0793, September 1981,
<http://www.rfc-editor.org/info/rfc793>. <http://www.rfc-editor.org/info/rfc793>.
[RFC1948] Bellovin, S., "Defending Against Sequence Number Attacks", [RFC1948] Bellovin, S., "Defending Against Sequence Number Attacks",
RFC 1948, DOI 10.17487/RFC1948, May 1996, RFC 1948, DOI 10.17487/RFC1948, May 1996,
<http://www.rfc-editor.org/info/rfc1948>. <http://www.rfc-editor.org/info/rfc1948>.
[RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC
Text on Security Considerations", BCP 72, RFC 3552,
DOI 10.17487/RFC3552, July 2003,
<http://www.rfc-editor.org/info/rfc3552>.
[RFC4086] Eastlake 3rd, D., Schiller, J., and S. Crocker, [RFC4086] Eastlake 3rd, D., Schiller, J., and S. Crocker,
"Randomness Requirements for Security", BCP 106, RFC 4086, "Randomness Requirements for Security", BCP 106, RFC 4086,
DOI 10.17487/RFC4086, June 2005, DOI 10.17487/RFC4086, June 2005,
<http://www.rfc-editor.org/info/rfc4086>. <http://www.rfc-editor.org/info/rfc4086>.
[RFC4279] Eronen, P., Ed. and H. Tschofenig, Ed., "Pre-Shared Key [RFC4279] Eronen, P., Ed. and H. Tschofenig, Ed., "Pre-Shared Key
Ciphersuites for Transport Layer Security (TLS)", Ciphersuites for Transport Layer Security (TLS)",
RFC 4279, DOI 10.17487/RFC4279, December 2005, RFC 4279, DOI 10.17487/RFC4279, December 2005,
<http://www.rfc-editor.org/info/rfc4279>. <http://www.rfc-editor.org/info/rfc4279>.
skipping to change at page 90, line 5 skipping to change at page 87, line 5
[RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer
Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347,
January 2012, <http://www.rfc-editor.org/info/rfc6347>. January 2012, <http://www.rfc-editor.org/info/rfc6347>.
[RFC6520] Seggelmann, R., Tuexen, M., and M. Williams, "Transport [RFC6520] Seggelmann, R., Tuexen, M., and M. Williams, "Transport
Layer Security (TLS) and Datagram Transport Layer Security Layer Security (TLS) and Datagram Transport Layer Security
(DTLS) Heartbeat Extension", RFC 6520, (DTLS) Heartbeat Extension", RFC 6520,
DOI 10.17487/RFC6520, February 2012, DOI 10.17487/RFC6520, February 2012,
<http://www.rfc-editor.org/info/rfc6520>. <http://www.rfc-editor.org/info/rfc6520>.
[RFC6962] Laurie, B., Langley, A., and E. Kasper, "Certificate
Transparency", RFC 6962, DOI 10.17487/RFC6962, June 2013,
<http://www.rfc-editor.org/info/rfc6962>.
[RFC7230] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer [RFC7230] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer
Protocol (HTTP/1.1): Message Syntax and Routing", Protocol (HTTP/1.1): Message Syntax and Routing",
RFC 7230, DOI 10.17487/RFC7230, June 2014, RFC 7230, DOI 10.17487/RFC7230, June 2014,
<http://www.rfc-editor.org/info/rfc7230>. <http://www.rfc-editor.org/info/rfc7230>.
[RFC7250] Wouters, P., Ed., Tschofenig, H., Ed., Gilmore, J., [RFC7250] Wouters, P., Ed., Tschofenig, H., Ed., Gilmore, J.,
Weiler, S., and T. Kivinen, "Using Raw Public Keys in Weiler, S., and T. Kivinen, "Using Raw Public Keys in
Transport Layer Security (TLS) and Datagram Transport Transport Layer Security (TLS) and Datagram Transport
Layer Security (DTLS)", RFC 7250, DOI 10.17487/RFC7250, Layer Security (DTLS)", RFC 7250, DOI 10.17487/RFC7250,
June 2014, <http://www.rfc-editor.org/info/rfc7250>. June 2014, <http://www.rfc-editor.org/info/rfc7250>.
skipping to change at page 91, line 5 skipping to change at page 88, line 5
[RFC7685] Langley, A., "A Transport Layer Security (TLS) ClientHello [RFC7685] Langley, A., "A Transport Layer Security (TLS) ClientHello
Padding Extension", RFC 7685, DOI 10.17487/RFC7685, Padding Extension", RFC 7685, DOI 10.17487/RFC7685,
October 2015, <http://www.rfc-editor.org/info/rfc7685>. October 2015, <http://www.rfc-editor.org/info/rfc7685>.
[RSA] Rivest, R., Shamir, A., and L. Adleman, "A Method for [RSA] Rivest, R., Shamir, A., and L. Adleman, "A Method for
Obtaining Digital Signatures and Public-Key Obtaining Digital Signatures and Public-Key
Cryptosystems", Communications of the ACM v. 21, n. 2, pp. Cryptosystems", Communications of the ACM v. 21, n. 2, pp.
120-126., February 1978. 120-126., February 1978.
[SIGMA] Krawczyk, H., "SIGMA: the 'SIGn-and-MAc' approach to
authenticated Di e-Hellman and its use in the IKE
protocols", Proceedings of CRYPTO 2003 , 2003.
[SLOTH] Bhargavan, K. and G. Leurent, "Transcript Collision [SLOTH] Bhargavan, K. and G. Leurent, "Transcript Collision
Attacks: Breaking Authentication in TLS, IKE, and SSH", Attacks: Breaking Authentication in TLS, IKE, and SSH",
Network and Distributed System Security Symposium (NDSS Network and Distributed System Security Symposium (NDSS
2016) , 2016. 2016) , 2016.
[SSL2] Hickman, K., "The SSL Protocol", February 1995. [SSL2] Hickman, K., "The SSL Protocol", February 1995.
[SSL3] Freier, A., Karlton, P., and P. Kocher, "The SSL 3.0 [SSL3] Freier, A., Karlton, P., and P. Kocher, "The SSL 3.0
Protocol", November 1996. Protocol", November 1996.
[TIMING] Boneh, D. and D. Brumley, "Remote timing attacks are [TIMING] Boneh, D. and D. Brumley, "Remote timing attacks are
practical", USENIX Security Symposium, 2003. practical", USENIX Security Symposium, 2003.
[X501] "Information Technology - Open Systems Interconnection - [X501] "Information Technology - Open Systems Interconnection -
The Directory: Models", ITU-T X.501, 1993. The Directory: Models", ITU-T X.501, 1993.
12.3. URIs 11.3. URIs
[1] mailto:tls@ietf.org [1] mailto:tls@ietf.org
Appendix A. Protocol Data Structures and Constant Values Appendix A. Protocol Data Structures and Constant Values
This section describes protocol types and constants. Values listed This section describes protocol types and constants. Values listed
as _RESERVED were used in previous versions of TLS and are listed as _RESERVED were used in previous versions of TLS and are listed
here for completeness. TLS 1.3 implementations MUST NOT send them here for completeness. TLS 1.3 implementations MUST NOT send them
but might receive them from older TLS implementations. but might receive them from older TLS implementations.
A.1. Record Layer A.1. Record Layer
struct {
uint8 major;
uint8 minor;
} ProtocolVersion;
enum { enum {
invalid_RESERVED(0), invalid_RESERVED(0),
change_cipher_spec_RESERVED(20), change_cipher_spec_RESERVED(20),
alert(21), alert(21),
handshake(22), handshake(22),
application_data(23) application_data(23)
(255) (255)
} ContentType; } ContentType;
struct { struct {
ContentType type; ContentType type;
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */ ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length; uint16 length;
opaque fragment[TLSPlaintext.length]; opaque fragment[TLSPlaintext.length];
} TLSPlaintext; } TLSPlaintext;
struct { struct {
opaque content[TLSPlaintext.length];
ContentType type;
uint8 zeros[length_of_padding];
} TLSInnerPlaintext;
struct {
ContentType opaque_type = application_data(23); /* see fragment.type */ ContentType opaque_type = application_data(23); /* see fragment.type */
ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */ ProtocolVersion record_version = { 3, 1 }; /* TLS v1.x */
uint16 length; uint16 length;
aead-ciphered struct { opaque encrypted_record[length];
opaque content[TLSPlaintext.length];
ContentType type;
uint8 zeros[length_of_padding];
} fragment;
} TLSCiphertext; } TLSCiphertext;
A.2. Alert Messages A.2. Alert Messages
enum { warning(1), fatal(2), (255) } AlertLevel; enum { warning(1), fatal(2), (255) } AlertLevel;
enum { enum {
close_notify(0), close_notify(0),
end_of_early_data(1), end_of_early_data(1),
unexpected_message(10), /* fatal */ unexpected_message(10),
bad_record_mac(20), /* fatal */ bad_record_mac(20),
decryption_failed_RESERVED(21), /* fatal */ decryption_failed_RESERVED(21),
record_overflow(22), /* fatal */ record_overflow(22),
decompression_failure_RESERVED(30), /* fatal */ decompression_failure_RESERVED(30),
handshake_failure(40), /* fatal */ handshake_failure(40),
no_certificate_RESERVED(41), /* fatal */ no_certificate_RESERVED(41),
bad_certificate(42), bad_certificate(42),
unsupported_certificate(43), unsupported_certificate(43),
certificate_revoked(44), certificate_revoked(44),
certificate_expired(45), certificate_expired(45),
certificate_unknown(46), certificate_unknown(46),
illegal_parameter(47), /* fatal */ illegal_parameter(47),
unknown_ca(48), /* fatal */ unknown_ca(48),
access_denied(49), /* fatal */ access_denied(49),
decode_error(50), /* fatal */ decode_error(50),
decrypt_error(51), /* fatal */ decrypt_error(51),
export_restriction_RESERVED(60), /* fatal */ export_restriction_RESERVED(60),
protocol_version(70), /* fatal */ protocol_version(70),
insufficient_security(71), /* fatal */ insufficient_security(71),
internal_error(80), /* fatal */ internal_error(80),
inappropriate_fallback(86), /* fatal */ inappropriate_fallback(86),
user_canceled(90), user_canceled(90),
no_renegotiation_RESERVED(100), /* fatal */ no_renegotiation_RESERVED(100),
missing_extension(109), /* fatal */ missing_extension(109),
unsupported_extension(110), /* fatal */ unsupported_extension(110),
certificate_unobtainable(111), certificate_unobtainable(111),
unrecognized_name(112), unrecognized_name(112),
bad_certificate_status_response(113), /* fatal */ bad_certificate_status_response(113),
bad_certificate_hash_value(114), /* fatal */ bad_certificate_hash_value(114),
unknown_psk_identity(115), unknown_psk_identity(115),
(255) (255)
} AlertDescription; } AlertDescription;
struct { struct {
AlertLevel level; AlertLevel level;
AlertDescription description; AlertDescription description;
} Alert; } Alert;
A.3. Handshake Protocol A.3. Handshake Protocol
enum { enum {
hello_request_RESERVED(0), hello_request_RESERVED(0),
client_hello(1), client_hello(1),
server_hello(2), server_hello(2),
session_ticket(4), new_session_ticket(4),
hello_retry_request(6), hello_retry_request(6),
encrypted_extensions(8), encrypted_extensions(8),
certificate(11), certificate(11),
server_key_exchange_RESERVED(12), server_key_exchange_RESERVED(12),
certificate_request(13), certificate_request(13),
server_hello_done_RESERVED(14), server_hello_done_RESERVED(14),
certificate_verify(15), certificate_verify(15),
client_key_exchange_RESERVED(16), client_key_exchange_RESERVED(16),
finished(20), finished(20),
key_update(24), key_update(24),
skipping to change at page 94, line 37 skipping to change at page 91, line 37
uint24 length; /* bytes in message */ uint24 length; /* bytes in message */
select (HandshakeType) { select (HandshakeType) {
case client_hello: ClientHello; case client_hello: ClientHello;
case server_hello: ServerHello; case server_hello: ServerHello;
case hello_retry_request: HelloRetryRequest; case hello_retry_request: HelloRetryRequest;
case encrypted_extensions: EncryptedExtensions; case encrypted_extensions: EncryptedExtensions;
case certificate_request: CertificateRequest; case certificate_request: CertificateRequest;
case certificate: Certificate; case certificate: Certificate;
case certificate_verify: CertificateVerify; case certificate_verify: CertificateVerify;
case finished: Finished; case finished: Finished;
case session_ticket: NewSessionTicket; case new_session_ticket: NewSessionTicket;
case key_update: KeyUpdate; case key_update: KeyUpdate;
} body; } body;
} Handshake; } Handshake;
A.3.1. Key Exchange Messages A.3.1. Key Exchange Messages
struct { struct {
uint8 major;
uint8 minor;
} ProtocolVersion;
struct {
opaque random_bytes[32]; opaque random_bytes[32];
} Random; } Random;
uint8 CipherSuite[2]; /* Cryptographic suite selector */ uint8 CipherSuite[2]; /* Cryptographic suite selector */
struct { struct {
ProtocolVersion client_version = { 3, 4 }; /* TLS v1.3 */ ProtocolVersion client_version = { 3, 4 }; /* TLS v1.3 */
Random random; Random random;
opaque legacy_session_id<0..32>; opaque legacy_session_id<0..32>;
CipherSuite cipher_suites<2..2^16-2>; CipherSuite cipher_suites<2..2^16-2>;
opaque legacy_compression_methods<1..2^8-1>; opaque legacy_compression_methods<1..2^8-1>;
Extension extensions<0..2^16-1>; Extension extensions<0..2^16-1>;
} ClientHello; } ClientHello;
skipping to change at page 95, line 35 skipping to change at page 92, line 40
ExtensionType extension_type; ExtensionType extension_type;
opaque extension_data<0..2^16-1>; opaque extension_data<0..2^16-1>;
} Extension; } Extension;
enum { enum {
supported_groups(10), supported_groups(10),
signature_algorithms(13), signature_algorithms(13),
key_share(40), key_share(40),
pre_shared_key(41), pre_shared_key(41),
early_data(42), early_data(42),
ticket_age(43), cookie(44),
cookie (44),
(65535) (65535)
} ExtensionType; } ExtensionType;
struct { struct {
NamedGroup group; NamedGroup group;
opaque key_exchange<1..2^16-1>; opaque key_exchange<1..2^16-1>;
} KeyShareEntry; } KeyShareEntry;
struct { struct {
select (role) { select (role) {
skipping to change at page 96, line 23 skipping to change at page 93, line 27
psk_identity identities<2..2^16-1>; psk_identity identities<2..2^16-1>;
case server: case server:
uint16 selected_identity; uint16 selected_identity;
} }
} PreSharedKeyExtension; } PreSharedKeyExtension;
struct { struct {
select (Role) { select (Role) {
case client: case client:
opaque context<0..255>; uint32 obfuscated_ticket_age;
case server: case server:
struct {}; struct {};
} }
} EarlyDataIndication; } EarlyDataIndication;
struct {
uint32 ticket_age;
} TicketAge;
A.3.1.1. Cookie Extension A.3.1.1. Cookie Extension
struct { struct {
opaque cookie<0..255>; opaque cookie<0..2^16-1>;
} Cookie; } Cookie;
A.3.1.2. Signature Algorithm Extension A.3.1.2. Signature Algorithm Extension
enum { enum {
/* RSASSA-PKCS-v1_5 algorithms */ /* RSASSA-PKCS1-v1_5 algorithms */
rsa_pkcs1_sha1 (0x0201), rsa_pkcs1_sha1 (0x0201),
rsa_pkcs1_sha256 (0x0401), rsa_pkcs1_sha256 (0x0401),
rsa_pkcs1_sha384 (0x0501), rsa_pkcs1_sha384 (0x0501),
rsa_pkcs1_sha512 (0x0601), rsa_pkcs1_sha512 (0x0601),
/* ECDSA algorithms */ /* ECDSA algorithms */
ecdsa_secp256r1_sha256 (0x0403), ecdsa_secp256r1_sha256 (0x0403),
ecdsa_secp384r1_sha384 (0x0503), ecdsa_secp384r1_sha384 (0x0503),
ecdsa_secp521r1_sha512 (0x0603), ecdsa_secp521r1_sha512 (0x0603),
skipping to change at page 97, line 30 skipping to change at page 94, line 30
/* EdDSA algorithms */ /* EdDSA algorithms */
ed25519 (0x0703), ed25519 (0x0703),
ed448 (0x0704), ed448 (0x0704),
/* Reserved Code Points */ /* Reserved Code Points */
dsa_sha1_RESERVED (0x0202), dsa_sha1_RESERVED (0x0202),
dsa_sha256_RESERVED (0x0402), dsa_sha256_RESERVED (0x0402),
dsa_sha384_RESERVED (0x0502), dsa_sha384_RESERVED (0x0502),
dsa_sha512_RESERVED (0x0602), dsa_sha512_RESERVED (0x0602),
ecdsa_sha1_RESERVED (0x0203),
obsolete_RESERVED (0x0000..0x0200), obsolete_RESERVED (0x0000..0x0200),
obsolete_RESERVED (0x0203..0x0400), obsolete_RESERVED (0x0204..0x0400),
obsolete_RESERVED (0x0404..0x0500), obsolete_RESERVED (0x0404..0x0500),
obsolete_RESERVED (0x0504..0x0600), obsolete_RESERVED (0x0504..0x0600),
obsolete_RESERVED (0x0604..0x06FF), obsolete_RESERVED (0x0604..0x06FF),
private_use (0xFE00..0xFFFF), private_use (0xFE00..0xFFFF),
(0xFFFF) (0xFFFF)
} SignatureScheme; } SignatureScheme;
SignatureScheme supported_signature_algorithms<2..2^16-2>; SignatureScheme supported_signature_algorithms<2..2^16-2>;
A.3.1.3. Named Group Extension A.3.1.3. Named Group Extension
skipping to change at page 99, line 33 skipping to change at page 96, line 33
A.3.3. Authentication Messages A.3.3. Authentication Messages
opaque ASN1Cert<1..2^24-1>; opaque ASN1Cert<1..2^24-1>;
struct { struct {
opaque certificate_request_context<0..2^8-1>; opaque certificate_request_context<0..2^8-1>;
ASN1Cert certificate_list<0..2^24-1>; ASN1Cert certificate_list<0..2^24-1>;
} Certificate; } Certificate;
struct { struct {
digitally-signed struct { SignatureScheme algorithm;
opaque hashed_data[hash_length]; opaque signature<0..2^16-1>;
};
} CertificateVerify; } CertificateVerify;
struct { struct {
opaque verify_data[verify_data_length]; opaque verify_data[Hash.length];
} Finished; } Finished;
A.3.4. Ticket Establishment A.3.4. Ticket Establishment
enum { (65535) } TicketExtensionType; enum { (65535) } TicketExtensionType;
struct { struct {
TicketExtensionType extension_type; TicketExtensionType extension_type;
opaque extension_data<0..2^16-1>; opaque extension_data<1..2^16-1>;
} TicketExtension; } TicketExtension;
enum { enum {
allow_early_data(1) allow_early_data(1),
allow_dhe_resumption(2), allow_dhe_resumption(2),
allow_psk_resumption(4) allow_psk_resumption(4)
} TicketFlags; } TicketFlags;
struct { struct {
uint32 ticket_lifetime; uint32 ticket_lifetime;
uint32 flags; uint32 flags;
uint32 ticket_age_add;
TicketExtension extensions<2..2^16-2>; TicketExtension extensions<2..2^16-2>;
opaque ticket<0..2^16-1>; opaque ticket<0..2^16-1>;
} NewSessionTicket; } NewSessionTicket;
A.4. Cipher Suites A.4. Cipher Suites
A cipher suite defines a cipher specification supported in TLS and A cipher suite defines a cipher specification supported in TLS and
negotiated via hello messages in the TLS handshake. Cipher suite negotiated via hello messages in the TLS handshake. Cipher suite
names follow a general naming convention composed of a series of names follow a general naming convention composed of a series of
component algorithm names separated by underscores: component algorithm names separated by underscores:
skipping to change at page 102, line 5 skipping to change at page 99, line 5
supported by TLS 1.3. supported by TLS 1.3.
See the definitions of each cipher suite in its specification See the definitions of each cipher suite in its specification
document for the full details of each combination of algorithms that document for the full details of each combination of algorithms that
is specified. is specified.
The following is a list of standards track server-authenticated (and The following is a list of standards track server-authenticated (and
optionally client-authenticated) cipher suites which are currently optionally client-authenticated) cipher suites which are currently
available in TLS 1.3: available in TLS 1.3:
+-------------------------------+----------+------------------------+ +----------------------------------------+-----------+--------------+
| Cipher Suite Name | Value | Specification | | Cipher Suite Name | Value | Specificatio |
+-------------------------------+----------+------------------------+ | | | n |
| TLS_DHE_RSA_WITH_AES_128_GCM_ | {0x00,0x | [RFC5288] | +----------------------------------------+-----------+--------------+
| SHA256 | 9E} | | | TLS_DHE_RSA_WITH_AES_128_GCM_SHA256 | {0x00,0x9 | [RFC5288] |
| | | | | | E} | |
| TLS_DHE_RSA_WITH_AES_256_GCM_ | {0x00,0x | [RFC5288] | | | | |
| SHA384 | 9F} | | | TLS_DHE_RSA_WITH_AES_256_GCM_SHA384 | {0x00,0x9 | [RFC5288] |
| | | | | | F} | |
| TLS_ECDHE_ECDSA_WITH_AES_128_ | {0xC0,0x | [RFC5289] | | | | |
| GCM_SHA256 | 2B} | | | TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA25 | {0xC0,0x2 | [RFC5289] |
| | | | | 6 | B} | |
| TLS_ECDHE_ECDSA_WITH_AES_256_ | {0xC0,0x | [RFC5289] | | | | |
| GCM_SHA384 | 2C} | | | TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA38 | {0xC0,0x2 | [RFC5289] |
| | | | | 4 | C} | |
| TLS_ECDHE_RSA_WITH_AES_128_GC | {0xC0,0x | [RFC5289] | | | | |
| M_SHA256 | 2F} | | | TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 | {0xC0,0x2 | [RFC5289] |
| | | | | | F} | |
| TLS_ECDHE_RSA_WITH_AES_256_GC | {0xC0,0x | [RFC5289] | | | | |
| M_SHA384 | 30} | | | TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 | {0xC0,0x3 | [RFC5289] |
| | | | | | 0} | |
| TLS_DHE_RSA_WITH_AES_128_CCM | {0xC0,0x | [RFC6655] | | | | |
| | 9E} | | | TLS_DHE_RSA_WITH_AES_128_CCM | {0xC0,0x9 | [RFC6655] |
| | | | | | E} | |
| TLS_DHE_RSA_WITH_AES_256_CCM | {0xC0,0x | [RFC6655] | | | | |
| | 9F} | | | TLS_DHE_RSA_WITH_AES_256_CCM | {0xC0,0x9 | [RFC6655] |
| | | | | | F} | |